diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,85359 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999589777249046,
+  "eval_steps": 500,
+  "global_step": 12188,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.204455019075358e-05,
+      "grad_norm": 6.266494143928387,
+      "learning_rate": 0.0,
+      "loss": 0.7985,
+      "step": 1
+    },
+    {
+      "epoch": 0.00016408910038150716,
+      "grad_norm": 5.32342735752047,
+      "learning_rate": 2.7322404371584703e-08,
+      "loss": 0.7623,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002461336505722607,
+      "grad_norm": 6.187765398699765,
+      "learning_rate": 5.4644808743169406e-08,
+      "loss": 0.757,
+      "step": 3
+    },
+    {
+      "epoch": 0.00032817820076301433,
+      "grad_norm": 5.1416991513792745,
+      "learning_rate": 8.19672131147541e-08,
+      "loss": 0.7479,
+      "step": 4
+    },
+    {
+      "epoch": 0.0004102227509537679,
+      "grad_norm": 4.896041914617405,
+      "learning_rate": 1.0928961748633881e-07,
+      "loss": 0.7787,
+      "step": 5
+    },
+    {
+      "epoch": 0.0004922673011445214,
+      "grad_norm": 5.07481063081961,
+      "learning_rate": 1.366120218579235e-07,
+      "loss": 0.8118,
+      "step": 6
+    },
+    {
+      "epoch": 0.000574311851335275,
+      "grad_norm": 5.173268633123172,
+      "learning_rate": 1.639344262295082e-07,
+      "loss": 0.8158,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006563564015260287,
+      "grad_norm": 4.716439772154978,
+      "learning_rate": 1.912568306010929e-07,
+      "loss": 0.7467,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007384009517167823,
+      "grad_norm": 5.262074612254626,
+      "learning_rate": 2.1857923497267762e-07,
+      "loss": 0.7917,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008204455019075358,
+      "grad_norm": 5.339490898319027,
+      "learning_rate": 2.459016393442623e-07,
+      "loss": 0.7487,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009024900520982894,
+      "grad_norm": 4.903293136039212,
+      "learning_rate": 2.73224043715847e-07,
+      "loss": 0.7175,
+      "step": 11
+    },
+    {
+      "epoch": 0.0009845346022890429,
+      "grad_norm": 4.848192521600832,
+      "learning_rate": 3.005464480874317e-07,
+      "loss": 0.752,
+      "step": 12
+    },
+    {
+      "epoch": 0.0010665791524797966,
+      "grad_norm": 5.410313430213829,
+      "learning_rate": 3.278688524590164e-07,
+      "loss": 0.759,
+      "step": 13
+    },
+    {
+      "epoch": 0.00114862370267055,
+      "grad_norm": 4.767157043197228,
+      "learning_rate": 3.551912568306011e-07,
+      "loss": 0.7451,
+      "step": 14
+    },
+    {
+      "epoch": 0.0012306682528613036,
+      "grad_norm": 4.767735351189707,
+      "learning_rate": 3.825136612021858e-07,
+      "loss": 0.7829,
+      "step": 15
+    },
+    {
+      "epoch": 0.0013127128030520573,
+      "grad_norm": 4.787381683190001,
+      "learning_rate": 4.0983606557377047e-07,
+      "loss": 0.7344,
+      "step": 16
+    },
+    {
+      "epoch": 0.0013947573532428108,
+      "grad_norm": 4.605029459841018,
+      "learning_rate": 4.3715846994535524e-07,
+      "loss": 0.7235,
+      "step": 17
+    },
+    {
+      "epoch": 0.0014768019034335645,
+      "grad_norm": 4.721653573484392,
+      "learning_rate": 4.6448087431693996e-07,
+      "loss": 0.7449,
+      "step": 18
+    },
+    {
+      "epoch": 0.001558846453624318,
+      "grad_norm": 3.9254751916372115,
+      "learning_rate": 4.918032786885246e-07,
+      "loss": 0.7197,
+      "step": 19
+    },
+    {
+      "epoch": 0.0016408910038150715,
+      "grad_norm": 3.996290247318457,
+      "learning_rate": 5.191256830601094e-07,
+      "loss": 0.7328,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017229355540058252,
+      "grad_norm": 4.0457784984003045,
+      "learning_rate": 5.46448087431694e-07,
+      "loss": 0.7397,
+      "step": 21
+    },
+    {
+      "epoch": 0.0018049801041965787,
+      "grad_norm": 3.816381637003003,
+      "learning_rate": 5.737704918032787e-07,
+      "loss": 0.7729,
+      "step": 22
+    },
+    {
+      "epoch": 0.0018870246543873322,
+      "grad_norm": 3.5932684645988826,
+      "learning_rate": 6.010928961748634e-07,
+      "loss": 0.6979,
+      "step": 23
+    },
+    {
+      "epoch": 0.0019690692045780857,
+      "grad_norm": 3.5679211998667286,
+      "learning_rate": 6.284153005464482e-07,
+      "loss": 0.7364,
+      "step": 24
+    },
+    {
+      "epoch": 0.0020511137547688397,
+      "grad_norm": 3.637723665750606,
+      "learning_rate": 6.557377049180328e-07,
+      "loss": 0.7886,
+      "step": 25
+    },
+    {
+      "epoch": 0.002133158304959593,
+      "grad_norm": 2.99762130528957,
+      "learning_rate": 6.830601092896176e-07,
+      "loss": 0.7113,
+      "step": 26
+    },
+    {
+      "epoch": 0.0022152028551503467,
+      "grad_norm": 2.387546839084648,
+      "learning_rate": 7.103825136612022e-07,
+      "loss": 0.68,
+      "step": 27
+    },
+    {
+      "epoch": 0.0022972474053411,
+      "grad_norm": 2.429525154972051,
+      "learning_rate": 7.377049180327869e-07,
+      "loss": 0.6626,
+      "step": 28
+    },
+    {
+      "epoch": 0.0023792919555318537,
+      "grad_norm": 2.6305168982601916,
+      "learning_rate": 7.650273224043716e-07,
+      "loss": 0.7016,
+      "step": 29
+    },
+    {
+      "epoch": 0.002461336505722607,
+      "grad_norm": 2.5415775371967095,
+      "learning_rate": 7.923497267759564e-07,
+      "loss": 0.7409,
+      "step": 30
+    },
+    {
+      "epoch": 0.002543381055913361,
+      "grad_norm": 2.252691874380961,
+      "learning_rate": 8.196721311475409e-07,
+      "loss": 0.6854,
+      "step": 31
+    },
+    {
+      "epoch": 0.0026254256061041146,
+      "grad_norm": 2.1897036846870757,
+      "learning_rate": 8.469945355191257e-07,
+      "loss": 0.7208,
+      "step": 32
+    },
+    {
+      "epoch": 0.002707470156294868,
+      "grad_norm": 2.124872376249603,
+      "learning_rate": 8.743169398907105e-07,
+      "loss": 0.6949,
+      "step": 33
+    },
+    {
+      "epoch": 0.0027895147064856216,
+      "grad_norm": 2.2571871851930325,
+      "learning_rate": 9.016393442622952e-07,
+      "loss": 0.7168,
+      "step": 34
+    },
+    {
+      "epoch": 0.002871559256676375,
+      "grad_norm": 2.1045966418835604,
+      "learning_rate": 9.289617486338799e-07,
+      "loss": 0.7352,
+      "step": 35
+    },
+    {
+      "epoch": 0.002953603806867129,
+      "grad_norm": 1.95532013095945,
+      "learning_rate": 9.562841530054645e-07,
+      "loss": 0.6897,
+      "step": 36
+    },
+    {
+      "epoch": 0.0030356483570578826,
+      "grad_norm": 1.9650517839666974,
+      "learning_rate": 9.836065573770493e-07,
+      "loss": 0.6535,
+      "step": 37
+    },
+    {
+      "epoch": 0.003117692907248636,
+      "grad_norm": 1.6796092327237084,
+      "learning_rate": 1.010928961748634e-06,
+      "loss": 0.5977,
+      "step": 38
+    },
+    {
+      "epoch": 0.0031997374574393896,
+      "grad_norm": 1.6448222590189212,
+      "learning_rate": 1.0382513661202188e-06,
+      "loss": 0.679,
+      "step": 39
+    },
+    {
+      "epoch": 0.003281782007630143,
+      "grad_norm": 1.4716278384447392,
+      "learning_rate": 1.0655737704918034e-06,
+      "loss": 0.6433,
+      "step": 40
+    },
+    {
+      "epoch": 0.0033638265578208966,
+      "grad_norm": 1.5523126314586277,
+      "learning_rate": 1.092896174863388e-06,
+      "loss": 0.6698,
+      "step": 41
+    },
+    {
+      "epoch": 0.0034458711080116505,
+      "grad_norm": 1.4096320940338003,
+      "learning_rate": 1.1202185792349727e-06,
+      "loss": 0.646,
+      "step": 42
+    },
+    {
+      "epoch": 0.003527915658202404,
+      "grad_norm": 1.2804099242960334,
+      "learning_rate": 1.1475409836065575e-06,
+      "loss": 0.6442,
+      "step": 43
+    },
+    {
+      "epoch": 0.0036099602083931575,
+      "grad_norm": 1.3057129544831738,
+      "learning_rate": 1.1748633879781422e-06,
+      "loss": 0.6568,
+      "step": 44
+    },
+    {
+      "epoch": 0.003692004758583911,
+      "grad_norm": 1.188954270959499,
+      "learning_rate": 1.2021857923497268e-06,
+      "loss": 0.6428,
+      "step": 45
+    },
+    {
+      "epoch": 0.0037740493087746645,
+      "grad_norm": 1.2591055307310337,
+      "learning_rate": 1.2295081967213116e-06,
+      "loss": 0.6073,
+      "step": 46
+    },
+    {
+      "epoch": 0.0038560938589654184,
+      "grad_norm": 1.2069297325506247,
+      "learning_rate": 1.2568306010928963e-06,
+      "loss": 0.594,
+      "step": 47
+    },
+    {
+      "epoch": 0.0039381384091561715,
+      "grad_norm": 1.3214514535959614,
+      "learning_rate": 1.2841530054644811e-06,
+      "loss": 0.7037,
+      "step": 48
+    },
+    {
+      "epoch": 0.004020182959346925,
+      "grad_norm": 1.3776912000057449,
+      "learning_rate": 1.3114754098360657e-06,
+      "loss": 0.5834,
+      "step": 49
+    },
+    {
+      "epoch": 0.004102227509537679,
+      "grad_norm": 1.0554211703822116,
+      "learning_rate": 1.3387978142076505e-06,
+      "loss": 0.5986,
+      "step": 50
+    },
+    {
+      "epoch": 0.0041842720597284324,
+      "grad_norm": 1.2086440972180963,
+      "learning_rate": 1.3661202185792352e-06,
+      "loss": 0.604,
+      "step": 51
+    },
+    {
+      "epoch": 0.004266316609919186,
+      "grad_norm": 1.3702917120401334,
+      "learning_rate": 1.3934426229508196e-06,
+      "loss": 0.6381,
+      "step": 52
+    },
+    {
+      "epoch": 0.0043483611601099394,
+      "grad_norm": 1.0577331108318386,
+      "learning_rate": 1.4207650273224043e-06,
+      "loss": 0.6388,
+      "step": 53
+    },
+    {
+      "epoch": 0.004430405710300693,
+      "grad_norm": 6.9191466355297395,
+      "learning_rate": 1.4480874316939891e-06,
+      "loss": 0.5692,
+      "step": 54
+    },
+    {
+      "epoch": 0.0045124502604914464,
+      "grad_norm": 0.9762202255589204,
+      "learning_rate": 1.4754098360655739e-06,
+      "loss": 0.5819,
+      "step": 55
+    },
+    {
+      "epoch": 0.0045944948106822,
+      "grad_norm": 1.0484078869988498,
+      "learning_rate": 1.5027322404371587e-06,
+      "loss": 0.6222,
+      "step": 56
+    },
+    {
+      "epoch": 0.004676539360872954,
+      "grad_norm": 1.031760741196525,
+      "learning_rate": 1.5300546448087432e-06,
+      "loss": 0.6438,
+      "step": 57
+    },
+    {
+      "epoch": 0.004758583911063707,
+      "grad_norm": 1.0070658465555649,
+      "learning_rate": 1.557377049180328e-06,
+      "loss": 0.642,
+      "step": 58
+    },
+    {
+      "epoch": 0.004840628461254461,
+      "grad_norm": 0.9572054708843818,
+      "learning_rate": 1.5846994535519128e-06,
+      "loss": 0.5958,
+      "step": 59
+    },
+    {
+      "epoch": 0.004922673011445214,
+      "grad_norm": 1.066797718390194,
+      "learning_rate": 1.6120218579234975e-06,
+      "loss": 0.62,
+      "step": 60
+    },
+    {
+      "epoch": 0.005004717561635968,
+      "grad_norm": 1.154599125092965,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 0.698,
+      "step": 61
+    },
+    {
+      "epoch": 0.005086762111826722,
+      "grad_norm": 0.9740798049139792,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5454,
+      "step": 62
+    },
+    {
+      "epoch": 0.005168806662017475,
+      "grad_norm": 0.9428524504581464,
+      "learning_rate": 1.6939890710382514e-06,
+      "loss": 0.6054,
+      "step": 63
+    },
+    {
+      "epoch": 0.005250851212208229,
+      "grad_norm": 0.9886092552869474,
+      "learning_rate": 1.7213114754098362e-06,
+      "loss": 0.6079,
+      "step": 64
+    },
+    {
+      "epoch": 0.005332895762398982,
+      "grad_norm": 0.8445250375195416,
+      "learning_rate": 1.748633879781421e-06,
+      "loss": 0.5535,
+      "step": 65
+    },
+    {
+      "epoch": 0.005414940312589736,
+      "grad_norm": 0.9178397444285696,
+      "learning_rate": 1.7759562841530055e-06,
+      "loss": 0.6392,
+      "step": 66
+    },
+    {
+      "epoch": 0.00549698486278049,
+      "grad_norm": 0.9473413958976997,
+      "learning_rate": 1.8032786885245903e-06,
+      "loss": 0.5682,
+      "step": 67
+    },
+    {
+      "epoch": 0.005579029412971243,
+      "grad_norm": 0.8040348000715498,
+      "learning_rate": 1.830601092896175e-06,
+      "loss": 0.5532,
+      "step": 68
+    },
+    {
+      "epoch": 0.005661073963161997,
+      "grad_norm": 0.7785719679258585,
+      "learning_rate": 1.8579234972677599e-06,
+      "loss": 0.5506,
+      "step": 69
+    },
+    {
+      "epoch": 0.00574311851335275,
+      "grad_norm": 0.8681680375754203,
+      "learning_rate": 1.8852459016393442e-06,
+      "loss": 0.5929,
+      "step": 70
+    },
+    {
+      "epoch": 0.005825163063543504,
+      "grad_norm": 0.8288211635891115,
+      "learning_rate": 1.912568306010929e-06,
+      "loss": 0.576,
+      "step": 71
+    },
+    {
+      "epoch": 0.005907207613734258,
+      "grad_norm": 0.8409409078712426,
+      "learning_rate": 1.939890710382514e-06,
+      "loss": 0.5882,
+      "step": 72
+    },
+    {
+      "epoch": 0.005989252163925011,
+      "grad_norm": 0.8114591818499827,
+      "learning_rate": 1.9672131147540985e-06,
+      "loss": 0.6139,
+      "step": 73
+    },
+    {
+      "epoch": 0.006071296714115765,
+      "grad_norm": 0.808525241283981,
+      "learning_rate": 1.994535519125683e-06,
+      "loss": 0.5905,
+      "step": 74
+    },
+    {
+      "epoch": 0.006153341264306518,
+      "grad_norm": 0.7894713533720883,
+      "learning_rate": 2.021857923497268e-06,
+      "loss": 0.5822,
+      "step": 75
+    },
+    {
+      "epoch": 0.006235385814497272,
+      "grad_norm": 0.7554413887760214,
+      "learning_rate": 2.0491803278688526e-06,
+      "loss": 0.6153,
+      "step": 76
+    },
+    {
+      "epoch": 0.006317430364688025,
+      "grad_norm": 0.810068435494818,
+      "learning_rate": 2.0765027322404376e-06,
+      "loss": 0.5785,
+      "step": 77
+    },
+    {
+      "epoch": 0.006399474914878779,
+      "grad_norm": 0.8218075222557052,
+      "learning_rate": 2.103825136612022e-06,
+      "loss": 0.5489,
+      "step": 78
+    },
+    {
+      "epoch": 0.006481519465069533,
+      "grad_norm": 0.7766768820882174,
+      "learning_rate": 2.1311475409836067e-06,
+      "loss": 0.6237,
+      "step": 79
+    },
+    {
+      "epoch": 0.006563564015260286,
+      "grad_norm": 0.8614006502657133,
+      "learning_rate": 2.1584699453551913e-06,
+      "loss": 0.5739,
+      "step": 80
+    },
+    {
+      "epoch": 0.00664560856545104,
+      "grad_norm": 0.7128586915540946,
+      "learning_rate": 2.185792349726776e-06,
+      "loss": 0.576,
+      "step": 81
+    },
+    {
+      "epoch": 0.006727653115641793,
+      "grad_norm": 0.7852664186569185,
+      "learning_rate": 2.213114754098361e-06,
+      "loss": 0.6007,
+      "step": 82
+    },
+    {
+      "epoch": 0.006809697665832547,
+      "grad_norm": 0.7315694270843645,
+      "learning_rate": 2.2404371584699454e-06,
+      "loss": 0.5646,
+      "step": 83
+    },
+    {
+      "epoch": 0.006891742216023301,
+      "grad_norm": 0.7338090571728428,
+      "learning_rate": 2.2677595628415304e-06,
+      "loss": 0.5845,
+      "step": 84
+    },
+    {
+      "epoch": 0.006973786766214054,
+      "grad_norm": 0.7797548599203646,
+      "learning_rate": 2.295081967213115e-06,
+      "loss": 0.6037,
+      "step": 85
+    },
+    {
+      "epoch": 0.007055831316404808,
+      "grad_norm": 0.7643951463830156,
+      "learning_rate": 2.3224043715847e-06,
+      "loss": 0.6085,
+      "step": 86
+    },
+    {
+      "epoch": 0.007137875866595561,
+      "grad_norm": 0.7481465252890332,
+      "learning_rate": 2.3497267759562845e-06,
+      "loss": 0.5845,
+      "step": 87
+    },
+    {
+      "epoch": 0.007219920416786315,
+      "grad_norm": 0.7572707341912558,
+      "learning_rate": 2.377049180327869e-06,
+      "loss": 0.5859,
+      "step": 88
+    },
+    {
+      "epoch": 0.007301964966977069,
+      "grad_norm": 0.7492814090628429,
+      "learning_rate": 2.4043715846994536e-06,
+      "loss": 0.5784,
+      "step": 89
+    },
+    {
+      "epoch": 0.007384009517167822,
+      "grad_norm": 0.7552663040151634,
+      "learning_rate": 2.431693989071038e-06,
+      "loss": 0.5947,
+      "step": 90
+    },
+    {
+      "epoch": 0.007466054067358576,
+      "grad_norm": 0.7527265178556873,
+      "learning_rate": 2.459016393442623e-06,
+      "loss": 0.6349,
+      "step": 91
+    },
+    {
+      "epoch": 0.007548098617549329,
+      "grad_norm": 0.7704598114019927,
+      "learning_rate": 2.4863387978142077e-06,
+      "loss": 0.5774,
+      "step": 92
+    },
+    {
+      "epoch": 0.007630143167740083,
+      "grad_norm": 0.7404694632419342,
+      "learning_rate": 2.5136612021857927e-06,
+      "loss": 0.5595,
+      "step": 93
+    },
+    {
+      "epoch": 0.007712187717930837,
+      "grad_norm": 0.7057068463369247,
+      "learning_rate": 2.5409836065573773e-06,
+      "loss": 0.5316,
+      "step": 94
+    },
+    {
+      "epoch": 0.00779423226812159,
+      "grad_norm": 0.7810985469021622,
+      "learning_rate": 2.5683060109289622e-06,
+      "loss": 0.5324,
+      "step": 95
+    },
+    {
+      "epoch": 0.007876276818312343,
+      "grad_norm": 0.7836959543417447,
+      "learning_rate": 2.595628415300547e-06,
+      "loss": 0.5695,
+      "step": 96
+    },
+    {
+      "epoch": 0.007958321368503097,
+      "grad_norm": 0.7059435573011988,
+      "learning_rate": 2.6229508196721314e-06,
+      "loss": 0.5986,
+      "step": 97
+    },
+    {
+      "epoch": 0.00804036591869385,
+      "grad_norm": 0.7165599446303674,
+      "learning_rate": 2.6502732240437163e-06,
+      "loss": 0.5195,
+      "step": 98
+    },
+    {
+      "epoch": 0.008122410468884605,
+      "grad_norm": 0.9708890958776207,
+      "learning_rate": 2.677595628415301e-06,
+      "loss": 0.5574,
+      "step": 99
+    },
+    {
+      "epoch": 0.008204455019075359,
+      "grad_norm": 0.7720714894062014,
+      "learning_rate": 2.704918032786886e-06,
+      "loss": 0.5278,
+      "step": 100
+    },
+    {
+      "epoch": 0.008286499569266111,
+      "grad_norm": 0.7895812343861897,
+      "learning_rate": 2.7322404371584705e-06,
+      "loss": 0.5908,
+      "step": 101
+    },
+    {
+      "epoch": 0.008368544119456865,
+      "grad_norm": 0.9614735702645801,
+      "learning_rate": 2.7595628415300546e-06,
+      "loss": 0.5496,
+      "step": 102
+    },
+    {
+      "epoch": 0.008450588669647619,
+      "grad_norm": 0.7307098529848282,
+      "learning_rate": 2.786885245901639e-06,
+      "loss": 0.5445,
+      "step": 103
+    },
+    {
+      "epoch": 0.008532633219838373,
+      "grad_norm": 0.7387165181019846,
+      "learning_rate": 2.814207650273224e-06,
+      "loss": 0.6102,
+      "step": 104
+    },
+    {
+      "epoch": 0.008614677770029127,
+      "grad_norm": 0.7753005211816525,
+      "learning_rate": 2.8415300546448087e-06,
+      "loss": 0.5578,
+      "step": 105
+    },
+    {
+      "epoch": 0.008696722320219879,
+      "grad_norm": 0.7611701431516128,
+      "learning_rate": 2.8688524590163937e-06,
+      "loss": 0.6299,
+      "step": 106
+    },
+    {
+      "epoch": 0.008778766870410633,
+      "grad_norm": 0.8933202473478205,
+      "learning_rate": 2.8961748633879782e-06,
+      "loss": 0.5987,
+      "step": 107
+    },
+    {
+      "epoch": 0.008860811420601387,
+      "grad_norm": 0.7174368991870086,
+      "learning_rate": 2.923497267759563e-06,
+      "loss": 0.5314,
+      "step": 108
+    },
+    {
+      "epoch": 0.00894285597079214,
+      "grad_norm": 0.7167502434034074,
+      "learning_rate": 2.9508196721311478e-06,
+      "loss": 0.5355,
+      "step": 109
+    },
+    {
+      "epoch": 0.009024900520982893,
+      "grad_norm": 0.73860332646601,
+      "learning_rate": 2.9781420765027323e-06,
+      "loss": 0.5604,
+      "step": 110
+    },
+    {
+      "epoch": 0.009106945071173647,
+      "grad_norm": 0.7221766194271947,
+      "learning_rate": 3.0054644808743173e-06,
+      "loss": 0.5695,
+      "step": 111
+    },
+    {
+      "epoch": 0.0091889896213644,
+      "grad_norm": 0.7172231095738897,
+      "learning_rate": 3.032786885245902e-06,
+      "loss": 0.5111,
+      "step": 112
+    },
+    {
+      "epoch": 0.009271034171555155,
+      "grad_norm": 0.7352308501356551,
+      "learning_rate": 3.0601092896174864e-06,
+      "loss": 0.4964,
+      "step": 113
+    },
+    {
+      "epoch": 0.009353078721745909,
+      "grad_norm": 0.713235393345211,
+      "learning_rate": 3.0874316939890714e-06,
+      "loss": 0.5904,
+      "step": 114
+    },
+    {
+      "epoch": 0.00943512327193666,
+      "grad_norm": 0.7283841882349938,
+      "learning_rate": 3.114754098360656e-06,
+      "loss": 0.5955,
+      "step": 115
+    },
+    {
+      "epoch": 0.009517167822127415,
+      "grad_norm": 0.7272379352954161,
+      "learning_rate": 3.142076502732241e-06,
+      "loss": 0.512,
+      "step": 116
+    },
+    {
+      "epoch": 0.009599212372318169,
+      "grad_norm": 0.704140649234432,
+      "learning_rate": 3.1693989071038255e-06,
+      "loss": 0.5497,
+      "step": 117
+    },
+    {
+      "epoch": 0.009681256922508923,
+      "grad_norm": 0.7403776551067047,
+      "learning_rate": 3.1967213114754105e-06,
+      "loss": 0.5021,
+      "step": 118
+    },
+    {
+      "epoch": 0.009763301472699677,
+      "grad_norm": 0.6228051098562285,
+      "learning_rate": 3.224043715846995e-06,
+      "loss": 0.4729,
+      "step": 119
+    },
+    {
+      "epoch": 0.009845346022890429,
+      "grad_norm": 0.6624910693533389,
+      "learning_rate": 3.2513661202185792e-06,
+      "loss": 0.5042,
+      "step": 120
+    },
+    {
+      "epoch": 0.009927390573081183,
+      "grad_norm": 0.6919127112659911,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 0.5146,
+      "step": 121
+    },
+    {
+      "epoch": 0.010009435123271937,
+      "grad_norm": 0.7149759090075953,
+      "learning_rate": 3.3060109289617488e-06,
+      "loss": 0.549,
+      "step": 122
+    },
+    {
+      "epoch": 0.01009147967346269,
+      "grad_norm": 0.8645606824175845,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5436,
+      "step": 123
+    },
+    {
+      "epoch": 0.010173524223653444,
+      "grad_norm": 0.7160560252655325,
+      "learning_rate": 3.3606557377049183e-06,
+      "loss": 0.5555,
+      "step": 124
+    },
+    {
+      "epoch": 0.010255568773844197,
+      "grad_norm": 0.7095889253209778,
+      "learning_rate": 3.387978142076503e-06,
+      "loss": 0.5538,
+      "step": 125
+    },
+    {
+      "epoch": 0.01033761332403495,
+      "grad_norm": 0.7481914394701366,
+      "learning_rate": 3.4153005464480874e-06,
+      "loss": 0.5698,
+      "step": 126
+    },
+    {
+      "epoch": 0.010419657874225705,
+      "grad_norm": 0.6758026641868979,
+      "learning_rate": 3.4426229508196724e-06,
+      "loss": 0.5439,
+      "step": 127
+    },
+    {
+      "epoch": 0.010501702424416458,
+      "grad_norm": 0.6974401977197344,
+      "learning_rate": 3.469945355191257e-06,
+      "loss": 0.5382,
+      "step": 128
+    },
+    {
+      "epoch": 0.010583746974607212,
+      "grad_norm": 0.7169904057460209,
+      "learning_rate": 3.497267759562842e-06,
+      "loss": 0.5205,
+      "step": 129
+    },
+    {
+      "epoch": 0.010665791524797965,
+      "grad_norm": 0.671314158253721,
+      "learning_rate": 3.5245901639344265e-06,
+      "loss": 0.4966,
+      "step": 130
+    },
+    {
+      "epoch": 0.010747836074988719,
+      "grad_norm": 0.7072336796566324,
+      "learning_rate": 3.551912568306011e-06,
+      "loss": 0.5498,
+      "step": 131
+    },
+    {
+      "epoch": 0.010829880625179472,
+      "grad_norm": 0.7382420926227349,
+      "learning_rate": 3.579234972677596e-06,
+      "loss": 0.525,
+      "step": 132
+    },
+    {
+      "epoch": 0.010911925175370226,
+      "grad_norm": 0.6995398537878451,
+      "learning_rate": 3.6065573770491806e-06,
+      "loss": 0.5442,
+      "step": 133
+    },
+    {
+      "epoch": 0.01099396972556098,
+      "grad_norm": 0.7534950981113788,
+      "learning_rate": 3.6338797814207656e-06,
+      "loss": 0.5198,
+      "step": 134
+    },
+    {
+      "epoch": 0.011076014275751733,
+      "grad_norm": 0.735929378243434,
+      "learning_rate": 3.66120218579235e-06,
+      "loss": 0.555,
+      "step": 135
+    },
+    {
+      "epoch": 0.011158058825942486,
+      "grad_norm": 0.7201764681653368,
+      "learning_rate": 3.6885245901639347e-06,
+      "loss": 0.4907,
+      "step": 136
+    },
+    {
+      "epoch": 0.01124010337613324,
+      "grad_norm": 0.6522720991896332,
+      "learning_rate": 3.7158469945355197e-06,
+      "loss": 0.5263,
+      "step": 137
+    },
+    {
+      "epoch": 0.011322147926323994,
+      "grad_norm": 0.7395099458945698,
+      "learning_rate": 3.7431693989071043e-06,
+      "loss": 0.5452,
+      "step": 138
+    },
+    {
+      "epoch": 0.011404192476514748,
+      "grad_norm": 0.7209513246536812,
+      "learning_rate": 3.7704918032786884e-06,
+      "loss": 0.5395,
+      "step": 139
+    },
+    {
+      "epoch": 0.0114862370267055,
+      "grad_norm": 0.7599055123909271,
+      "learning_rate": 3.7978142076502734e-06,
+      "loss": 0.5753,
+      "step": 140
+    },
+    {
+      "epoch": 0.011568281576896254,
+      "grad_norm": 0.6965762440737608,
+      "learning_rate": 3.825136612021858e-06,
+      "loss": 0.4973,
+      "step": 141
+    },
+    {
+      "epoch": 0.011650326127087008,
+      "grad_norm": 0.7445968423045807,
+      "learning_rate": 3.852459016393443e-06,
+      "loss": 0.497,
+      "step": 142
+    },
+    {
+      "epoch": 0.011732370677277762,
+      "grad_norm": 0.7667796118793652,
+      "learning_rate": 3.879781420765028e-06,
+      "loss": 0.5779,
+      "step": 143
+    },
+    {
+      "epoch": 0.011814415227468516,
+      "grad_norm": 0.7352525692982557,
+      "learning_rate": 3.907103825136612e-06,
+      "loss": 0.5485,
+      "step": 144
+    },
+    {
+      "epoch": 0.011896459777659268,
+      "grad_norm": 0.7388167010484292,
+      "learning_rate": 3.934426229508197e-06,
+      "loss": 0.5416,
+      "step": 145
+    },
+    {
+      "epoch": 0.011978504327850022,
+      "grad_norm": 0.7329334594149585,
+      "learning_rate": 3.961748633879782e-06,
+      "loss": 0.5711,
+      "step": 146
+    },
+    {
+      "epoch": 0.012060548878040776,
+      "grad_norm": 0.7341246080271276,
+      "learning_rate": 3.989071038251366e-06,
+      "loss": 0.5973,
+      "step": 147
+    },
+    {
+      "epoch": 0.01214259342823153,
+      "grad_norm": 0.7063940812731384,
+      "learning_rate": 4.016393442622951e-06,
+      "loss": 0.5293,
+      "step": 148
+    },
+    {
+      "epoch": 0.012224637978422282,
+      "grad_norm": 0.6440475877504753,
+      "learning_rate": 4.043715846994536e-06,
+      "loss": 0.5078,
+      "step": 149
+    },
+    {
+      "epoch": 0.012306682528613036,
+      "grad_norm": 0.7274766928347551,
+      "learning_rate": 4.07103825136612e-06,
+      "loss": 0.5051,
+      "step": 150
+    },
+    {
+      "epoch": 0.01238872707880379,
+      "grad_norm": 0.7772069598712157,
+      "learning_rate": 4.098360655737705e-06,
+      "loss": 0.5666,
+      "step": 151
+    },
+    {
+      "epoch": 0.012470771628994544,
+      "grad_norm": 0.6961686233459559,
+      "learning_rate": 4.12568306010929e-06,
+      "loss": 0.5281,
+      "step": 152
+    },
+    {
+      "epoch": 0.012552816179185298,
+      "grad_norm": 0.7028795014580788,
+      "learning_rate": 4.153005464480875e-06,
+      "loss": 0.5339,
+      "step": 153
+    },
+    {
+      "epoch": 0.01263486072937605,
+      "grad_norm": 0.6977819368837961,
+      "learning_rate": 4.180327868852459e-06,
+      "loss": 0.4681,
+      "step": 154
+    },
+    {
+      "epoch": 0.012716905279566804,
+      "grad_norm": 0.689604405138438,
+      "learning_rate": 4.207650273224044e-06,
+      "loss": 0.5416,
+      "step": 155
+    },
+    {
+      "epoch": 0.012798949829757558,
+      "grad_norm": 0.7329166590311801,
+      "learning_rate": 4.234972677595629e-06,
+      "loss": 0.5395,
+      "step": 156
+    },
+    {
+      "epoch": 0.012880994379948312,
+      "grad_norm": 0.6875105879630147,
+      "learning_rate": 4.2622950819672135e-06,
+      "loss": 0.4868,
+      "step": 157
+    },
+    {
+      "epoch": 0.012963038930139066,
+      "grad_norm": 0.8280596402034371,
+      "learning_rate": 4.289617486338798e-06,
+      "loss": 0.547,
+      "step": 158
+    },
+    {
+      "epoch": 0.013045083480329818,
+      "grad_norm": 0.7251166083954937,
+      "learning_rate": 4.316939890710383e-06,
+      "loss": 0.519,
+      "step": 159
+    },
+    {
+      "epoch": 0.013127128030520572,
+      "grad_norm": 0.6906098353919035,
+      "learning_rate": 4.3442622950819676e-06,
+      "loss": 0.5226,
+      "step": 160
+    },
+    {
+      "epoch": 0.013209172580711326,
+      "grad_norm": 0.7029180259064449,
+      "learning_rate": 4.371584699453552e-06,
+      "loss": 0.4778,
+      "step": 161
+    },
+    {
+      "epoch": 0.01329121713090208,
+      "grad_norm": 0.7212367062954328,
+      "learning_rate": 4.398907103825137e-06,
+      "loss": 0.4836,
+      "step": 162
+    },
+    {
+      "epoch": 0.013373261681092834,
+      "grad_norm": 0.7036493053336564,
+      "learning_rate": 4.426229508196722e-06,
+      "loss": 0.5141,
+      "step": 163
+    },
+    {
+      "epoch": 0.013455306231283586,
+      "grad_norm": 0.6510839308020532,
+      "learning_rate": 4.453551912568307e-06,
+      "loss": 0.4656,
+      "step": 164
+    },
+    {
+      "epoch": 0.01353735078147434,
+      "grad_norm": 0.7026000593685714,
+      "learning_rate": 4.480874316939891e-06,
+      "loss": 0.5375,
+      "step": 165
+    },
+    {
+      "epoch": 0.013619395331665094,
+      "grad_norm": 0.7187224410212275,
+      "learning_rate": 4.508196721311476e-06,
+      "loss": 0.5663,
+      "step": 166
+    },
+    {
+      "epoch": 0.013701439881855848,
+      "grad_norm": 0.7740735840459895,
+      "learning_rate": 4.535519125683061e-06,
+      "loss": 0.5495,
+      "step": 167
+    },
+    {
+      "epoch": 0.013783484432046602,
+      "grad_norm": 0.7174969066563622,
+      "learning_rate": 4.562841530054645e-06,
+      "loss": 0.4883,
+      "step": 168
+    },
+    {
+      "epoch": 0.013865528982237354,
+      "grad_norm": 0.7358742733962577,
+      "learning_rate": 4.59016393442623e-06,
+      "loss": 0.5549,
+      "step": 169
+    },
+    {
+      "epoch": 0.013947573532428108,
+      "grad_norm": 0.7104310931064527,
+      "learning_rate": 4.617486338797815e-06,
+      "loss": 0.5411,
+      "step": 170
+    },
+    {
+      "epoch": 0.014029618082618862,
+      "grad_norm": 0.7431821605877618,
+      "learning_rate": 4.6448087431694e-06,
+      "loss": 0.4588,
+      "step": 171
+    },
+    {
+      "epoch": 0.014111662632809616,
+      "grad_norm": 0.7099686775669957,
+      "learning_rate": 4.672131147540984e-06,
+      "loss": 0.5119,
+      "step": 172
+    },
+    {
+      "epoch": 0.01419370718300037,
+      "grad_norm": 0.6951537547691425,
+      "learning_rate": 4.699453551912569e-06,
+      "loss": 0.5425,
+      "step": 173
+    },
+    {
+      "epoch": 0.014275751733191122,
+      "grad_norm": 0.7089535036082988,
+      "learning_rate": 4.726775956284154e-06,
+      "loss": 0.5526,
+      "step": 174
+    },
+    {
+      "epoch": 0.014357796283381876,
+      "grad_norm": 0.7379642172358976,
+      "learning_rate": 4.754098360655738e-06,
+      "loss": 0.4894,
+      "step": 175
+    },
+    {
+      "epoch": 0.01443984083357263,
+      "grad_norm": 0.7195500398829908,
+      "learning_rate": 4.781420765027322e-06,
+      "loss": 0.5108,
+      "step": 176
+    },
+    {
+      "epoch": 0.014521885383763384,
+      "grad_norm": 0.6953860103437562,
+      "learning_rate": 4.808743169398907e-06,
+      "loss": 0.4737,
+      "step": 177
+    },
+    {
+      "epoch": 0.014603929933954138,
+      "grad_norm": 0.6912848981929625,
+      "learning_rate": 4.836065573770492e-06,
+      "loss": 0.499,
+      "step": 178
+    },
+    {
+      "epoch": 0.01468597448414489,
+      "grad_norm": 0.7475510959967208,
+      "learning_rate": 4.863387978142076e-06,
+      "loss": 0.5133,
+      "step": 179
+    },
+    {
+      "epoch": 0.014768019034335644,
+      "grad_norm": 0.7753831797023494,
+      "learning_rate": 4.890710382513661e-06,
+      "loss": 0.5091,
+      "step": 180
+    },
+    {
+      "epoch": 0.014850063584526398,
+      "grad_norm": 0.6966761826583769,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 0.501,
+      "step": 181
+    },
+    {
+      "epoch": 0.014932108134717152,
+      "grad_norm": 0.7519230472398877,
+      "learning_rate": 4.945355191256831e-06,
+      "loss": 0.4687,
+      "step": 182
+    },
+    {
+      "epoch": 0.015014152684907906,
+      "grad_norm": 0.67827433000988,
+      "learning_rate": 4.9726775956284154e-06,
+      "loss": 0.4911,
+      "step": 183
+    },
+    {
+      "epoch": 0.015096197235098658,
+      "grad_norm": 0.6836673528085871,
+      "learning_rate": 5e-06,
+      "loss": 0.5156,
+      "step": 184
+    },
+    {
+      "epoch": 0.015178241785289412,
+      "grad_norm": 0.8493357596442787,
+      "learning_rate": 5.027322404371585e-06,
+      "loss": 0.4999,
+      "step": 185
+    },
+    {
+      "epoch": 0.015260286335480166,
+      "grad_norm": 0.7251608265186033,
+      "learning_rate": 5.0546448087431695e-06,
+      "loss": 0.4666,
+      "step": 186
+    },
+    {
+      "epoch": 0.01534233088567092,
+      "grad_norm": 0.7256599954067017,
+      "learning_rate": 5.0819672131147545e-06,
+      "loss": 0.5187,
+      "step": 187
+    },
+    {
+      "epoch": 0.015424375435861674,
+      "grad_norm": 0.7515864577718108,
+      "learning_rate": 5.1092896174863395e-06,
+      "loss": 0.4737,
+      "step": 188
+    },
+    {
+      "epoch": 0.015506419986052426,
+      "grad_norm": 0.7139370625006265,
+      "learning_rate": 5.1366120218579245e-06,
+      "loss": 0.5309,
+      "step": 189
+    },
+    {
+      "epoch": 0.01558846453624318,
+      "grad_norm": 0.7433117647073699,
+      "learning_rate": 5.163934426229509e-06,
+      "loss": 0.5027,
+      "step": 190
+    },
+    {
+      "epoch": 0.015670509086433932,
+      "grad_norm": 0.7249128535772957,
+      "learning_rate": 5.191256830601094e-06,
+      "loss": 0.4661,
+      "step": 191
+    },
+    {
+      "epoch": 0.015752553636624686,
+      "grad_norm": 0.7282325284072083,
+      "learning_rate": 5.218579234972679e-06,
+      "loss": 0.5654,
+      "step": 192
+    },
+    {
+      "epoch": 0.01583459818681544,
+      "grad_norm": 0.7199008938094988,
+      "learning_rate": 5.245901639344263e-06,
+      "loss": 0.513,
+      "step": 193
+    },
+    {
+      "epoch": 0.015916642737006194,
+      "grad_norm": 0.7039947581587781,
+      "learning_rate": 5.273224043715848e-06,
+      "loss": 0.5096,
+      "step": 194
+    },
+    {
+      "epoch": 0.015998687287196948,
+      "grad_norm": 0.7110940915833809,
+      "learning_rate": 5.300546448087433e-06,
+      "loss": 0.5375,
+      "step": 195
+    },
+    {
+      "epoch": 0.0160807318373877,
+      "grad_norm": 0.7543700852047978,
+      "learning_rate": 5.327868852459017e-06,
+      "loss": 0.5143,
+      "step": 196
+    },
+    {
+      "epoch": 0.016162776387578456,
+      "grad_norm": 0.7120086353387379,
+      "learning_rate": 5.355191256830602e-06,
+      "loss": 0.5113,
+      "step": 197
+    },
+    {
+      "epoch": 0.01624482093776921,
+      "grad_norm": 0.6862503719751668,
+      "learning_rate": 5.382513661202187e-06,
+      "loss": 0.5447,
+      "step": 198
+    },
+    {
+      "epoch": 0.016326865487959964,
+      "grad_norm": 0.724761928445193,
+      "learning_rate": 5.409836065573772e-06,
+      "loss": 0.5303,
+      "step": 199
+    },
+    {
+      "epoch": 0.016408910038150717,
+      "grad_norm": 0.6763503917362121,
+      "learning_rate": 5.437158469945356e-06,
+      "loss": 0.5113,
+      "step": 200
+    },
+    {
+      "epoch": 0.016490954588341468,
+      "grad_norm": 0.7103630077364631,
+      "learning_rate": 5.464480874316941e-06,
+      "loss": 0.4485,
+      "step": 201
+    },
+    {
+      "epoch": 0.016572999138532222,
+      "grad_norm": 1.470493893164305,
+      "learning_rate": 5.491803278688526e-06,
+      "loss": 0.4781,
+      "step": 202
+    },
+    {
+      "epoch": 0.016655043688722976,
+      "grad_norm": 0.7243588393636814,
+      "learning_rate": 5.519125683060109e-06,
+      "loss": 0.5147,
+      "step": 203
+    },
+    {
+      "epoch": 0.01673708823891373,
+      "grad_norm": 0.6807181004819323,
+      "learning_rate": 5.546448087431694e-06,
+      "loss": 0.4709,
+      "step": 204
+    },
+    {
+      "epoch": 0.016819132789104484,
+      "grad_norm": 0.708060012048527,
+      "learning_rate": 5.573770491803278e-06,
+      "loss": 0.5141,
+      "step": 205
+    },
+    {
+      "epoch": 0.016901177339295238,
+      "grad_norm": 0.68838680333791,
+      "learning_rate": 5.601092896174863e-06,
+      "loss": 0.5155,
+      "step": 206
+    },
+    {
+      "epoch": 0.01698322188948599,
+      "grad_norm": 0.7239270961077369,
+      "learning_rate": 5.628415300546448e-06,
+      "loss": 0.4788,
+      "step": 207
+    },
+    {
+      "epoch": 0.017065266439676745,
+      "grad_norm": 0.7328558873875821,
+      "learning_rate": 5.655737704918033e-06,
+      "loss": 0.4839,
+      "step": 208
+    },
+    {
+      "epoch": 0.0171473109898675,
+      "grad_norm": 0.7760332556150643,
+      "learning_rate": 5.683060109289617e-06,
+      "loss": 0.4979,
+      "step": 209
+    },
+    {
+      "epoch": 0.017229355540058253,
+      "grad_norm": 0.7102887727675635,
+      "learning_rate": 5.710382513661202e-06,
+      "loss": 0.4849,
+      "step": 210
+    },
+    {
+      "epoch": 0.017311400090249004,
+      "grad_norm": 0.7363864553582286,
+      "learning_rate": 5.737704918032787e-06,
+      "loss": 0.4659,
+      "step": 211
+    },
+    {
+      "epoch": 0.017393444640439758,
+      "grad_norm": 0.7035864495655544,
+      "learning_rate": 5.7650273224043715e-06,
+      "loss": 0.4596,
+      "step": 212
+    },
+    {
+      "epoch": 0.01747548919063051,
+      "grad_norm": 0.7457081676983588,
+      "learning_rate": 5.7923497267759565e-06,
+      "loss": 0.487,
+      "step": 213
+    },
+    {
+      "epoch": 0.017557533740821266,
+      "grad_norm": 0.7239763029713772,
+      "learning_rate": 5.8196721311475415e-06,
+      "loss": 0.4861,
+      "step": 214
+    },
+    {
+      "epoch": 0.01763957829101202,
+      "grad_norm": 0.8293037191887754,
+      "learning_rate": 5.846994535519126e-06,
+      "loss": 0.5131,
+      "step": 215
+    },
+    {
+      "epoch": 0.017721622841202773,
+      "grad_norm": 0.713670742599938,
+      "learning_rate": 5.874316939890711e-06,
+      "loss": 0.489,
+      "step": 216
+    },
+    {
+      "epoch": 0.017803667391393527,
+      "grad_norm": 0.6784402301270791,
+      "learning_rate": 5.9016393442622956e-06,
+      "loss": 0.5162,
+      "step": 217
+    },
+    {
+      "epoch": 0.01788571194158428,
+      "grad_norm": 0.7618172919676223,
+      "learning_rate": 5.9289617486338805e-06,
+      "loss": 0.5109,
+      "step": 218
+    },
+    {
+      "epoch": 0.017967756491775035,
+      "grad_norm": 0.7084600876246179,
+      "learning_rate": 5.956284153005465e-06,
+      "loss": 0.5196,
+      "step": 219
+    },
+    {
+      "epoch": 0.018049801041965786,
+      "grad_norm": 0.6975307025246504,
+      "learning_rate": 5.98360655737705e-06,
+      "loss": 0.5351,
+      "step": 220
+    },
+    {
+      "epoch": 0.01813184559215654,
+      "grad_norm": 0.7965883513551822,
+      "learning_rate": 6.010928961748635e-06,
+      "loss": 0.48,
+      "step": 221
+    },
+    {
+      "epoch": 0.018213890142347294,
+      "grad_norm": 0.7124731647151697,
+      "learning_rate": 6.038251366120219e-06,
+      "loss": 0.479,
+      "step": 222
+    },
+    {
+      "epoch": 0.018295934692538048,
+      "grad_norm": 0.6887948060629508,
+      "learning_rate": 6.065573770491804e-06,
+      "loss": 0.4725,
+      "step": 223
+    },
+    {
+      "epoch": 0.0183779792427288,
+      "grad_norm": 0.7955094446195284,
+      "learning_rate": 6.092896174863389e-06,
+      "loss": 0.5118,
+      "step": 224
+    },
+    {
+      "epoch": 0.018460023792919555,
+      "grad_norm": 0.7078808702226231,
+      "learning_rate": 6.120218579234973e-06,
+      "loss": 0.5219,
+      "step": 225
+    },
+    {
+      "epoch": 0.01854206834311031,
+      "grad_norm": 0.752437471063338,
+      "learning_rate": 6.147540983606558e-06,
+      "loss": 0.4865,
+      "step": 226
+    },
+    {
+      "epoch": 0.018624112893301063,
+      "grad_norm": 0.7547184388582712,
+      "learning_rate": 6.174863387978143e-06,
+      "loss": 0.4852,
+      "step": 227
+    },
+    {
+      "epoch": 0.018706157443491817,
+      "grad_norm": 0.7822474394399602,
+      "learning_rate": 6.202185792349728e-06,
+      "loss": 0.5174,
+      "step": 228
+    },
+    {
+      "epoch": 0.01878820199368257,
+      "grad_norm": 0.693832699033056,
+      "learning_rate": 6.229508196721312e-06,
+      "loss": 0.4983,
+      "step": 229
+    },
+    {
+      "epoch": 0.01887024654387332,
+      "grad_norm": 0.7244900380744682,
+      "learning_rate": 6.256830601092897e-06,
+      "loss": 0.4554,
+      "step": 230
+    },
+    {
+      "epoch": 0.018952291094064076,
+      "grad_norm": 0.6734569921178174,
+      "learning_rate": 6.284153005464482e-06,
+      "loss": 0.4778,
+      "step": 231
+    },
+    {
+      "epoch": 0.01903433564425483,
+      "grad_norm": 0.6871351773704322,
+      "learning_rate": 6.311475409836066e-06,
+      "loss": 0.5051,
+      "step": 232
+    },
+    {
+      "epoch": 0.019116380194445583,
+      "grad_norm": 0.8091815082981457,
+      "learning_rate": 6.338797814207651e-06,
+      "loss": 0.5398,
+      "step": 233
+    },
+    {
+      "epoch": 0.019198424744636337,
+      "grad_norm": 0.7724622540458691,
+      "learning_rate": 6.366120218579236e-06,
+      "loss": 0.539,
+      "step": 234
+    },
+    {
+      "epoch": 0.01928046929482709,
+      "grad_norm": 0.7991344883938822,
+      "learning_rate": 6.393442622950821e-06,
+      "loss": 0.5089,
+      "step": 235
+    },
+    {
+      "epoch": 0.019362513845017845,
+      "grad_norm": 0.7783987761834243,
+      "learning_rate": 6.420765027322405e-06,
+      "loss": 0.4716,
+      "step": 236
+    },
+    {
+      "epoch": 0.0194445583952086,
+      "grad_norm": 0.7879664281886021,
+      "learning_rate": 6.44808743169399e-06,
+      "loss": 0.4857,
+      "step": 237
+    },
+    {
+      "epoch": 0.019526602945399353,
+      "grad_norm": 0.7545213205093599,
+      "learning_rate": 6.475409836065575e-06,
+      "loss": 0.4999,
+      "step": 238
+    },
+    {
+      "epoch": 0.019608647495590107,
+      "grad_norm": 0.7485190909060746,
+      "learning_rate": 6.5027322404371584e-06,
+      "loss": 0.4971,
+      "step": 239
+    },
+    {
+      "epoch": 0.019690692045780857,
+      "grad_norm": 0.7114781795927844,
+      "learning_rate": 6.530054644808743e-06,
+      "loss": 0.4441,
+      "step": 240
+    },
+    {
+      "epoch": 0.01977273659597161,
+      "grad_norm": 0.793894220931936,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 0.5011,
+      "step": 241
+    },
+    {
+      "epoch": 0.019854781146162365,
+      "grad_norm": 0.7507841289943263,
+      "learning_rate": 6.5846994535519125e-06,
+      "loss": 0.4542,
+      "step": 242
+    },
+    {
+      "epoch": 0.01993682569635312,
+      "grad_norm": 0.7613787235425983,
+      "learning_rate": 6.6120218579234975e-06,
+      "loss": 0.517,
+      "step": 243
+    },
+    {
+      "epoch": 0.020018870246543873,
+      "grad_norm": 0.7842513155616389,
+      "learning_rate": 6.6393442622950825e-06,
+      "loss": 0.5164,
+      "step": 244
+    },
+    {
+      "epoch": 0.020100914796734627,
+      "grad_norm": 0.7612335417927749,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5188,
+      "step": 245
+    },
+    {
+      "epoch": 0.02018295934692538,
+      "grad_norm": 0.710277948436649,
+      "learning_rate": 6.693989071038252e-06,
+      "loss": 0.5357,
+      "step": 246
+    },
+    {
+      "epoch": 0.020265003897116135,
+      "grad_norm": 0.7593982472109353,
+      "learning_rate": 6.721311475409837e-06,
+      "loss": 0.4947,
+      "step": 247
+    },
+    {
+      "epoch": 0.02034704844730689,
+      "grad_norm": 0.6516468513397349,
+      "learning_rate": 6.748633879781421e-06,
+      "loss": 0.4546,
+      "step": 248
+    },
+    {
+      "epoch": 0.020429092997497643,
+      "grad_norm": 0.6961226163309198,
+      "learning_rate": 6.775956284153006e-06,
+      "loss": 0.4788,
+      "step": 249
+    },
+    {
+      "epoch": 0.020511137547688393,
+      "grad_norm": 0.7856434345730949,
+      "learning_rate": 6.803278688524591e-06,
+      "loss": 0.5056,
+      "step": 250
+    },
+    {
+      "epoch": 0.020593182097879147,
+      "grad_norm": 0.6709913584324687,
+      "learning_rate": 6.830601092896175e-06,
+      "loss": 0.4644,
+      "step": 251
+    },
+    {
+      "epoch": 0.0206752266480699,
+      "grad_norm": 0.725809958662215,
+      "learning_rate": 6.85792349726776e-06,
+      "loss": 0.5257,
+      "step": 252
+    },
+    {
+      "epoch": 0.020757271198260655,
+      "grad_norm": 0.7522902710326914,
+      "learning_rate": 6.885245901639345e-06,
+      "loss": 0.4702,
+      "step": 253
+    },
+    {
+      "epoch": 0.02083931574845141,
+      "grad_norm": 0.7950237553247519,
+      "learning_rate": 6.91256830601093e-06,
+      "loss": 0.4365,
+      "step": 254
+    },
+    {
+      "epoch": 0.020921360298642163,
+      "grad_norm": 0.6950687309001368,
+      "learning_rate": 6.939890710382514e-06,
+      "loss": 0.4841,
+      "step": 255
+    },
+    {
+      "epoch": 0.021003404848832917,
+      "grad_norm": 0.7424707337968431,
+      "learning_rate": 6.967213114754099e-06,
+      "loss": 0.5244,
+      "step": 256
+    },
+    {
+      "epoch": 0.02108544939902367,
+      "grad_norm": 0.7419306692293721,
+      "learning_rate": 6.994535519125684e-06,
+      "loss": 0.4806,
+      "step": 257
+    },
+    {
+      "epoch": 0.021167493949214425,
+      "grad_norm": 0.7515718720294478,
+      "learning_rate": 7.021857923497268e-06,
+      "loss": 0.4922,
+      "step": 258
+    },
+    {
+      "epoch": 0.021249538499405175,
+      "grad_norm": 0.7135712134020837,
+      "learning_rate": 7.049180327868853e-06,
+      "loss": 0.4204,
+      "step": 259
+    },
+    {
+      "epoch": 0.02133158304959593,
+      "grad_norm": 0.7039455465319501,
+      "learning_rate": 7.076502732240438e-06,
+      "loss": 0.5011,
+      "step": 260
+    },
+    {
+      "epoch": 0.021413627599786683,
+      "grad_norm": 0.7521350075014408,
+      "learning_rate": 7.103825136612022e-06,
+      "loss": 0.4777,
+      "step": 261
+    },
+    {
+      "epoch": 0.021495672149977437,
+      "grad_norm": 0.7399889000505141,
+      "learning_rate": 7.131147540983607e-06,
+      "loss": 0.5077,
+      "step": 262
+    },
+    {
+      "epoch": 0.02157771670016819,
+      "grad_norm": 0.6850858476258389,
+      "learning_rate": 7.158469945355192e-06,
+      "loss": 0.464,
+      "step": 263
+    },
+    {
+      "epoch": 0.021659761250358945,
+      "grad_norm": 0.708899608581843,
+      "learning_rate": 7.185792349726777e-06,
+      "loss": 0.4712,
+      "step": 264
+    },
+    {
+      "epoch": 0.0217418058005497,
+      "grad_norm": 0.7209070719253067,
+      "learning_rate": 7.213114754098361e-06,
+      "loss": 0.4798,
+      "step": 265
+    },
+    {
+      "epoch": 0.021823850350740453,
+      "grad_norm": 0.7306374132694864,
+      "learning_rate": 7.240437158469946e-06,
+      "loss": 0.5322,
+      "step": 266
+    },
+    {
+      "epoch": 0.021905894900931207,
+      "grad_norm": 0.7772129021189023,
+      "learning_rate": 7.267759562841531e-06,
+      "loss": 0.5184,
+      "step": 267
+    },
+    {
+      "epoch": 0.02198793945112196,
+      "grad_norm": 0.6765022684123391,
+      "learning_rate": 7.295081967213115e-06,
+      "loss": 0.4622,
+      "step": 268
+    },
+    {
+      "epoch": 0.02206998400131271,
+      "grad_norm": 0.7073564967165935,
+      "learning_rate": 7.3224043715847e-06,
+      "loss": 0.4761,
+      "step": 269
+    },
+    {
+      "epoch": 0.022152028551503465,
+      "grad_norm": 1.0185678136029017,
+      "learning_rate": 7.349726775956285e-06,
+      "loss": 0.4386,
+      "step": 270
+    },
+    {
+      "epoch": 0.02223407310169422,
+      "grad_norm": 0.6974035976598978,
+      "learning_rate": 7.3770491803278695e-06,
+      "loss": 0.4746,
+      "step": 271
+    },
+    {
+      "epoch": 0.022316117651884973,
+      "grad_norm": 0.7338952253335077,
+      "learning_rate": 7.4043715846994544e-06,
+      "loss": 0.5047,
+      "step": 272
+    },
+    {
+      "epoch": 0.022398162202075727,
+      "grad_norm": 0.7456966980847313,
+      "learning_rate": 7.4316939890710394e-06,
+      "loss": 0.5275,
+      "step": 273
+    },
+    {
+      "epoch": 0.02248020675226648,
+      "grad_norm": 0.8026042643549763,
+      "learning_rate": 7.459016393442624e-06,
+      "loss": 0.464,
+      "step": 274
+    },
+    {
+      "epoch": 0.022562251302457235,
+      "grad_norm": 0.7295729206906412,
+      "learning_rate": 7.4863387978142085e-06,
+      "loss": 0.4722,
+      "step": 275
+    },
+    {
+      "epoch": 0.02264429585264799,
+      "grad_norm": 0.7940364942726144,
+      "learning_rate": 7.513661202185793e-06,
+      "loss": 0.4721,
+      "step": 276
+    },
+    {
+      "epoch": 0.022726340402838743,
+      "grad_norm": 0.7486001185681557,
+      "learning_rate": 7.540983606557377e-06,
+      "loss": 0.484,
+      "step": 277
+    },
+    {
+      "epoch": 0.022808384953029497,
+      "grad_norm": 0.7284050307306237,
+      "learning_rate": 7.568306010928962e-06,
+      "loss": 0.5272,
+      "step": 278
+    },
+    {
+      "epoch": 0.022890429503220247,
+      "grad_norm": 0.7807521672555144,
+      "learning_rate": 7.595628415300547e-06,
+      "loss": 0.4841,
+      "step": 279
+    },
+    {
+      "epoch": 0.022972474053411,
+      "grad_norm": 0.7423125440153614,
+      "learning_rate": 7.622950819672132e-06,
+      "loss": 0.5082,
+      "step": 280
+    },
+    {
+      "epoch": 0.023054518603601755,
+      "grad_norm": 0.7202916618864765,
+      "learning_rate": 7.650273224043716e-06,
+      "loss": 0.4698,
+      "step": 281
+    },
+    {
+      "epoch": 0.02313656315379251,
+      "grad_norm": 0.7604901644633639,
+      "learning_rate": 7.677595628415301e-06,
+      "loss": 0.5325,
+      "step": 282
+    },
+    {
+      "epoch": 0.023218607703983263,
+      "grad_norm": 0.7388611534619522,
+      "learning_rate": 7.704918032786886e-06,
+      "loss": 0.4706,
+      "step": 283
+    },
+    {
+      "epoch": 0.023300652254174017,
+      "grad_norm": 0.7478738497567147,
+      "learning_rate": 7.732240437158471e-06,
+      "loss": 0.4919,
+      "step": 284
+    },
+    {
+      "epoch": 0.02338269680436477,
+      "grad_norm": 0.793051303030315,
+      "learning_rate": 7.759562841530056e-06,
+      "loss": 0.5021,
+      "step": 285
+    },
+    {
+      "epoch": 0.023464741354555525,
+      "grad_norm": 0.7300441301912832,
+      "learning_rate": 7.786885245901639e-06,
+      "loss": 0.4747,
+      "step": 286
+    },
+    {
+      "epoch": 0.02354678590474628,
+      "grad_norm": 0.7403184073542309,
+      "learning_rate": 7.814207650273224e-06,
+      "loss": 0.4788,
+      "step": 287
+    },
+    {
+      "epoch": 0.023628830454937032,
+      "grad_norm": 0.7448376481423855,
+      "learning_rate": 7.841530054644809e-06,
+      "loss": 0.4793,
+      "step": 288
+    },
+    {
+      "epoch": 0.023710875005127783,
+      "grad_norm": 0.8004814810395461,
+      "learning_rate": 7.868852459016394e-06,
+      "loss": 0.5112,
+      "step": 289
+    },
+    {
+      "epoch": 0.023792919555318537,
+      "grad_norm": 0.7587413695095064,
+      "learning_rate": 7.896174863387979e-06,
+      "loss": 0.5153,
+      "step": 290
+    },
+    {
+      "epoch": 0.02387496410550929,
+      "grad_norm": 0.7057932396663791,
+      "learning_rate": 7.923497267759564e-06,
+      "loss": 0.4574,
+      "step": 291
+    },
+    {
+      "epoch": 0.023957008655700045,
+      "grad_norm": 0.7346626512251365,
+      "learning_rate": 7.950819672131147e-06,
+      "loss": 0.4482,
+      "step": 292
+    },
+    {
+      "epoch": 0.0240390532058908,
+      "grad_norm": 0.7256235360123868,
+      "learning_rate": 7.978142076502732e-06,
+      "loss": 0.4648,
+      "step": 293
+    },
+    {
+      "epoch": 0.024121097756081553,
+      "grad_norm": 0.7426640793653947,
+      "learning_rate": 8.005464480874317e-06,
+      "loss": 0.4928,
+      "step": 294
+    },
+    {
+      "epoch": 0.024203142306272307,
+      "grad_norm": 0.6830842630027686,
+      "learning_rate": 8.032786885245902e-06,
+      "loss": 0.4773,
+      "step": 295
+    },
+    {
+      "epoch": 0.02428518685646306,
+      "grad_norm": 0.8328130237681087,
+      "learning_rate": 8.060109289617487e-06,
+      "loss": 0.5165,
+      "step": 296
+    },
+    {
+      "epoch": 0.024367231406653814,
+      "grad_norm": 0.6931884993503776,
+      "learning_rate": 8.087431693989072e-06,
+      "loss": 0.4974,
+      "step": 297
+    },
+    {
+      "epoch": 0.024449275956844565,
+      "grad_norm": 0.72777194335979,
+      "learning_rate": 8.114754098360657e-06,
+      "loss": 0.4797,
+      "step": 298
+    },
+    {
+      "epoch": 0.02453132050703532,
+      "grad_norm": 0.7525301581418226,
+      "learning_rate": 8.14207650273224e-06,
+      "loss": 0.4845,
+      "step": 299
+    },
+    {
+      "epoch": 0.024613365057226073,
+      "grad_norm": 0.7243695451603184,
+      "learning_rate": 8.169398907103826e-06,
+      "loss": 0.4896,
+      "step": 300
+    },
+    {
+      "epoch": 0.024695409607416827,
+      "grad_norm": 0.7341444038210161,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 0.4709,
+      "step": 301
+    },
+    {
+      "epoch": 0.02477745415760758,
+      "grad_norm": 0.7498865761223373,
+      "learning_rate": 8.224043715846995e-06,
+      "loss": 0.4876,
+      "step": 302
+    },
+    {
+      "epoch": 0.024859498707798335,
+      "grad_norm": 0.6769756622245103,
+      "learning_rate": 8.25136612021858e-06,
+      "loss": 0.4955,
+      "step": 303
+    },
+    {
+      "epoch": 0.02494154325798909,
+      "grad_norm": 0.7092618677680266,
+      "learning_rate": 8.278688524590165e-06,
+      "loss": 0.5092,
+      "step": 304
+    },
+    {
+      "epoch": 0.025023587808179842,
+      "grad_norm": 0.7740115083798342,
+      "learning_rate": 8.30601092896175e-06,
+      "loss": 0.5194,
+      "step": 305
+    },
+    {
+      "epoch": 0.025105632358370596,
+      "grad_norm": 0.7361606758491017,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.4636,
+      "step": 306
+    },
+    {
+      "epoch": 0.02518767690856135,
+      "grad_norm": 0.7235739656871362,
+      "learning_rate": 8.360655737704919e-06,
+      "loss": 0.4678,
+      "step": 307
+    },
+    {
+      "epoch": 0.0252697214587521,
+      "grad_norm": 0.7200813323160887,
+      "learning_rate": 8.387978142076504e-06,
+      "loss": 0.462,
+      "step": 308
+    },
+    {
+      "epoch": 0.025351766008942855,
+      "grad_norm": 0.7425112449267953,
+      "learning_rate": 8.415300546448089e-06,
+      "loss": 0.4673,
+      "step": 309
+    },
+    {
+      "epoch": 0.02543381055913361,
+      "grad_norm": 0.9021836323143811,
+      "learning_rate": 8.442622950819674e-06,
+      "loss": 0.5408,
+      "step": 310
+    },
+    {
+      "epoch": 0.025515855109324363,
+      "grad_norm": 0.7970247048932755,
+      "learning_rate": 8.469945355191259e-06,
+      "loss": 0.492,
+      "step": 311
+    },
+    {
+      "epoch": 0.025597899659515116,
+      "grad_norm": 0.811538302246817,
+      "learning_rate": 8.497267759562842e-06,
+      "loss": 0.4547,
+      "step": 312
+    },
+    {
+      "epoch": 0.02567994420970587,
+      "grad_norm": 0.7267247239879944,
+      "learning_rate": 8.524590163934427e-06,
+      "loss": 0.4687,
+      "step": 313
+    },
+    {
+      "epoch": 0.025761988759896624,
+      "grad_norm": 0.7752832459631157,
+      "learning_rate": 8.551912568306012e-06,
+      "loss": 0.4897,
+      "step": 314
+    },
+    {
+      "epoch": 0.02584403331008738,
+      "grad_norm": 0.7620499486189858,
+      "learning_rate": 8.579234972677595e-06,
+      "loss": 0.4879,
+      "step": 315
+    },
+    {
+      "epoch": 0.025926077860278132,
+      "grad_norm": 0.7622870026247587,
+      "learning_rate": 8.60655737704918e-06,
+      "loss": 0.4493,
+      "step": 316
+    },
+    {
+      "epoch": 0.026008122410468886,
+      "grad_norm": 0.7038270969801519,
+      "learning_rate": 8.633879781420765e-06,
+      "loss": 0.4959,
+      "step": 317
+    },
+    {
+      "epoch": 0.026090166960659637,
+      "grad_norm": 0.7231370930482404,
+      "learning_rate": 8.66120218579235e-06,
+      "loss": 0.5314,
+      "step": 318
+    },
+    {
+      "epoch": 0.02617221151085039,
+      "grad_norm": 0.7290789659308253,
+      "learning_rate": 8.688524590163935e-06,
+      "loss": 0.451,
+      "step": 319
+    },
+    {
+      "epoch": 0.026254256061041144,
+      "grad_norm": 0.7303738122844974,
+      "learning_rate": 8.71584699453552e-06,
+      "loss": 0.4564,
+      "step": 320
+    },
+    {
+      "epoch": 0.0263363006112319,
+      "grad_norm": 0.6673064334950203,
+      "learning_rate": 8.743169398907103e-06,
+      "loss": 0.4653,
+      "step": 321
+    },
+    {
+      "epoch": 0.026418345161422652,
+      "grad_norm": 0.7680272597113199,
+      "learning_rate": 8.770491803278688e-06,
+      "loss": 0.4914,
+      "step": 322
+    },
+    {
+      "epoch": 0.026500389711613406,
+      "grad_norm": 0.7456630568103154,
+      "learning_rate": 8.797814207650273e-06,
+      "loss": 0.5103,
+      "step": 323
+    },
+    {
+      "epoch": 0.02658243426180416,
+      "grad_norm": 0.6950247147636593,
+      "learning_rate": 8.825136612021858e-06,
+      "loss": 0.4459,
+      "step": 324
+    },
+    {
+      "epoch": 0.026664478811994914,
+      "grad_norm": 0.7095726681711076,
+      "learning_rate": 8.852459016393443e-06,
+      "loss": 0.4509,
+      "step": 325
+    },
+    {
+      "epoch": 0.026746523362185668,
+      "grad_norm": 0.6957022116801302,
+      "learning_rate": 8.879781420765028e-06,
+      "loss": 0.4713,
+      "step": 326
+    },
+    {
+      "epoch": 0.026828567912376422,
+      "grad_norm": 0.7315214323065525,
+      "learning_rate": 8.907103825136613e-06,
+      "loss": 0.5448,
+      "step": 327
+    },
+    {
+      "epoch": 0.026910612462567172,
+      "grad_norm": 0.7872172103370333,
+      "learning_rate": 8.934426229508197e-06,
+      "loss": 0.4836,
+      "step": 328
+    },
+    {
+      "epoch": 0.026992657012757926,
+      "grad_norm": 0.7475204197911525,
+      "learning_rate": 8.961748633879782e-06,
+      "loss": 0.5001,
+      "step": 329
+    },
+    {
+      "epoch": 0.02707470156294868,
+      "grad_norm": 0.7180684338768776,
+      "learning_rate": 8.989071038251367e-06,
+      "loss": 0.5088,
+      "step": 330
+    },
+    {
+      "epoch": 0.027156746113139434,
+      "grad_norm": 0.7620192982877894,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 0.4646,
+      "step": 331
+    },
+    {
+      "epoch": 0.027238790663330188,
+      "grad_norm": 0.6936782731483281,
+      "learning_rate": 9.043715846994537e-06,
+      "loss": 0.4618,
+      "step": 332
+    },
+    {
+      "epoch": 0.027320835213520942,
+      "grad_norm": 0.7699652288427808,
+      "learning_rate": 9.071038251366122e-06,
+      "loss": 0.4737,
+      "step": 333
+    },
+    {
+      "epoch": 0.027402879763711696,
+      "grad_norm": 0.7326266987448204,
+      "learning_rate": 9.098360655737707e-06,
+      "loss": 0.4871,
+      "step": 334
+    },
+    {
+      "epoch": 0.02748492431390245,
+      "grad_norm": 0.7302372275645583,
+      "learning_rate": 9.12568306010929e-06,
+      "loss": 0.493,
+      "step": 335
+    },
+    {
+      "epoch": 0.027566968864093204,
+      "grad_norm": 0.7452132566260702,
+      "learning_rate": 9.153005464480875e-06,
+      "loss": 0.4664,
+      "step": 336
+    },
+    {
+      "epoch": 0.027649013414283958,
+      "grad_norm": 0.7321326686430336,
+      "learning_rate": 9.18032786885246e-06,
+      "loss": 0.4552,
+      "step": 337
+    },
+    {
+      "epoch": 0.02773105796447471,
+      "grad_norm": 0.7091837589791525,
+      "learning_rate": 9.207650273224045e-06,
+      "loss": 0.5116,
+      "step": 338
+    },
+    {
+      "epoch": 0.027813102514665462,
+      "grad_norm": 0.7590251237438369,
+      "learning_rate": 9.23497267759563e-06,
+      "loss": 0.4795,
+      "step": 339
+    },
+    {
+      "epoch": 0.027895147064856216,
+      "grad_norm": 0.7428316985926329,
+      "learning_rate": 9.262295081967215e-06,
+      "loss": 0.4621,
+      "step": 340
+    },
+    {
+      "epoch": 0.02797719161504697,
+      "grad_norm": 0.7650681063577547,
+      "learning_rate": 9.2896174863388e-06,
+      "loss": 0.4822,
+      "step": 341
+    },
+    {
+      "epoch": 0.028059236165237724,
+      "grad_norm": 0.721221484811715,
+      "learning_rate": 9.316939890710383e-06,
+      "loss": 0.4816,
+      "step": 342
+    },
+    {
+      "epoch": 0.028141280715428478,
+      "grad_norm": 0.6887336677008205,
+      "learning_rate": 9.344262295081968e-06,
+      "loss": 0.4697,
+      "step": 343
+    },
+    {
+      "epoch": 0.028223325265619232,
+      "grad_norm": 0.7237681811072515,
+      "learning_rate": 9.371584699453553e-06,
+      "loss": 0.4447,
+      "step": 344
+    },
+    {
+      "epoch": 0.028305369815809986,
+      "grad_norm": 0.7454686846524716,
+      "learning_rate": 9.398907103825138e-06,
+      "loss": 0.4829,
+      "step": 345
+    },
+    {
+      "epoch": 0.02838741436600074,
+      "grad_norm": 0.7202549102789509,
+      "learning_rate": 9.426229508196723e-06,
+      "loss": 0.4868,
+      "step": 346
+    },
+    {
+      "epoch": 0.02846945891619149,
+      "grad_norm": 0.7657389869390842,
+      "learning_rate": 9.453551912568308e-06,
+      "loss": 0.4691,
+      "step": 347
+    },
+    {
+      "epoch": 0.028551503466382244,
+      "grad_norm": 0.6971451694403581,
+      "learning_rate": 9.480874316939891e-06,
+      "loss": 0.5236,
+      "step": 348
+    },
+    {
+      "epoch": 0.028633548016572998,
+      "grad_norm": 0.7778097903864223,
+      "learning_rate": 9.508196721311476e-06,
+      "loss": 0.4497,
+      "step": 349
+    },
+    {
+      "epoch": 0.028715592566763752,
+      "grad_norm": 0.7369490413482781,
+      "learning_rate": 9.535519125683061e-06,
+      "loss": 0.5016,
+      "step": 350
+    },
+    {
+      "epoch": 0.028797637116954506,
+      "grad_norm": 0.8739699109636033,
+      "learning_rate": 9.562841530054644e-06,
+      "loss": 0.4605,
+      "step": 351
+    },
+    {
+      "epoch": 0.02887968166714526,
+      "grad_norm": 0.7029868528561894,
+      "learning_rate": 9.59016393442623e-06,
+      "loss": 0.4723,
+      "step": 352
+    },
+    {
+      "epoch": 0.028961726217336014,
+      "grad_norm": 0.7535751671698064,
+      "learning_rate": 9.617486338797814e-06,
+      "loss": 0.4712,
+      "step": 353
+    },
+    {
+      "epoch": 0.029043770767526768,
+      "grad_norm": 0.7592360519812318,
+      "learning_rate": 9.6448087431694e-06,
+      "loss": 0.4549,
+      "step": 354
+    },
+    {
+      "epoch": 0.029125815317717522,
+      "grad_norm": 0.7535652921010682,
+      "learning_rate": 9.672131147540984e-06,
+      "loss": 0.4632,
+      "step": 355
+    },
+    {
+      "epoch": 0.029207859867908276,
+      "grad_norm": 0.722248001161935,
+      "learning_rate": 9.69945355191257e-06,
+      "loss": 0.4945,
+      "step": 356
+    },
+    {
+      "epoch": 0.029289904418099026,
+      "grad_norm": 0.7282775782224231,
+      "learning_rate": 9.726775956284153e-06,
+      "loss": 0.5234,
+      "step": 357
+    },
+    {
+      "epoch": 0.02937194896828978,
+      "grad_norm": 0.7797206309580986,
+      "learning_rate": 9.754098360655738e-06,
+      "loss": 0.4572,
+      "step": 358
+    },
+    {
+      "epoch": 0.029453993518480534,
+      "grad_norm": 0.7591514259181091,
+      "learning_rate": 9.781420765027323e-06,
+      "loss": 0.4892,
+      "step": 359
+    },
+    {
+      "epoch": 0.029536038068671288,
+      "grad_norm": 0.6880716566756042,
+      "learning_rate": 9.808743169398908e-06,
+      "loss": 0.4553,
+      "step": 360
+    },
+    {
+      "epoch": 0.029618082618862042,
+      "grad_norm": 0.7141878435054724,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 0.4938,
+      "step": 361
+    },
+    {
+      "epoch": 0.029700127169052796,
+      "grad_norm": 0.7280965697748293,
+      "learning_rate": 9.863387978142078e-06,
+      "loss": 0.5462,
+      "step": 362
+    },
+    {
+      "epoch": 0.02978217171924355,
+      "grad_norm": 0.7295001648732062,
+      "learning_rate": 9.890710382513663e-06,
+      "loss": 0.4803,
+      "step": 363
+    },
+    {
+      "epoch": 0.029864216269434304,
+      "grad_norm": 0.6847651753518992,
+      "learning_rate": 9.918032786885246e-06,
+      "loss": 0.4529,
+      "step": 364
+    },
+    {
+      "epoch": 0.029946260819625058,
+      "grad_norm": 0.7370494406656364,
+      "learning_rate": 9.945355191256831e-06,
+      "loss": 0.4819,
+      "step": 365
+    },
+    {
+      "epoch": 0.03002830536981581,
+      "grad_norm": 0.7729949941299822,
+      "learning_rate": 9.972677595628416e-06,
+      "loss": 0.4523,
+      "step": 366
+    },
+    {
+      "epoch": 0.030110349920006562,
+      "grad_norm": 0.775932067067094,
+      "learning_rate": 1e-05,
+      "loss": 0.4881,
+      "step": 367
+    },
+    {
+      "epoch": 0.030192394470197316,
+      "grad_norm": 0.7684759740655762,
+      "learning_rate": 9.999999823454018e-06,
+      "loss": 0.4968,
+      "step": 368
+    },
+    {
+      "epoch": 0.03027443902038807,
+      "grad_norm": 0.7018257469675304,
+      "learning_rate": 9.99999929381608e-06,
+      "loss": 0.4978,
+      "step": 369
+    },
+    {
+      "epoch": 0.030356483570578824,
+      "grad_norm": 0.6964270416468571,
+      "learning_rate": 9.999998411086225e-06,
+      "loss": 0.4373,
+      "step": 370
+    },
+    {
+      "epoch": 0.030438528120769578,
+      "grad_norm": 0.7506393329953867,
+      "learning_rate": 9.999997175264516e-06,
+      "loss": 0.4653,
+      "step": 371
+    },
+    {
+      "epoch": 0.03052057267096033,
+      "grad_norm": 0.7759342224387521,
+      "learning_rate": 9.999995586351038e-06,
+      "loss": 0.5027,
+      "step": 372
+    },
+    {
+      "epoch": 0.030602617221151086,
+      "grad_norm": 0.7612908197879767,
+      "learning_rate": 9.999993644345906e-06,
+      "loss": 0.4916,
+      "step": 373
+    },
+    {
+      "epoch": 0.03068466177134184,
+      "grad_norm": 0.7624981656819343,
+      "learning_rate": 9.999991349249256e-06,
+      "loss": 0.4759,
+      "step": 374
+    },
+    {
+      "epoch": 0.030766706321532594,
+      "grad_norm": 0.8114410658059767,
+      "learning_rate": 9.99998870106125e-06,
+      "loss": 0.4768,
+      "step": 375
+    },
+    {
+      "epoch": 0.030848750871723347,
+      "grad_norm": 0.6690881883040294,
+      "learning_rate": 9.999985699782075e-06,
+      "loss": 0.4571,
+      "step": 376
+    },
+    {
+      "epoch": 0.030930795421914098,
+      "grad_norm": 0.6793569290243348,
+      "learning_rate": 9.999982345411943e-06,
+      "loss": 0.5032,
+      "step": 377
+    },
+    {
+      "epoch": 0.031012839972104852,
+      "grad_norm": 0.7576721180571018,
+      "learning_rate": 9.99997863795109e-06,
+      "loss": 0.4773,
+      "step": 378
+    },
+    {
+      "epoch": 0.031094884522295606,
+      "grad_norm": 0.725333924330082,
+      "learning_rate": 9.99997457739978e-06,
+      "loss": 0.4619,
+      "step": 379
+    },
+    {
+      "epoch": 0.03117692907248636,
+      "grad_norm": 0.6989828858207294,
+      "learning_rate": 9.999970163758296e-06,
+      "loss": 0.5093,
+      "step": 380
+    },
+    {
+      "epoch": 0.031258973622677114,
+      "grad_norm": 0.7171867965154881,
+      "learning_rate": 9.999965397026953e-06,
+      "loss": 0.4287,
+      "step": 381
+    },
+    {
+      "epoch": 0.031341018172867864,
+      "grad_norm": 0.6545167832175092,
+      "learning_rate": 9.999960277206089e-06,
+      "loss": 0.4522,
+      "step": 382
+    },
+    {
+      "epoch": 0.03142306272305862,
+      "grad_norm": 0.7009539190954738,
+      "learning_rate": 9.99995480429606e-06,
+      "loss": 0.4444,
+      "step": 383
+    },
+    {
+      "epoch": 0.03150510727324937,
+      "grad_norm": 1.426807833039473,
+      "learning_rate": 9.999948978297259e-06,
+      "loss": 0.8633,
+      "step": 384
+    },
+    {
+      "epoch": 0.03158715182344013,
+      "grad_norm": 0.796097003615168,
+      "learning_rate": 9.999942799210094e-06,
+      "loss": 0.4892,
+      "step": 385
+    },
+    {
+      "epoch": 0.03166919637363088,
+      "grad_norm": 0.8001708309492076,
+      "learning_rate": 9.999936267035e-06,
+      "loss": 0.4656,
+      "step": 386
+    },
+    {
+      "epoch": 0.03175124092382164,
+      "grad_norm": 0.7411460798090299,
+      "learning_rate": 9.99992938177244e-06,
+      "loss": 0.4845,
+      "step": 387
+    },
+    {
+      "epoch": 0.03183328547401239,
+      "grad_norm": 0.7576570415683321,
+      "learning_rate": 9.999922143422899e-06,
+      "loss": 0.4794,
+      "step": 388
+    },
+    {
+      "epoch": 0.031915330024203145,
+      "grad_norm": 0.8146049452268994,
+      "learning_rate": 9.999914551986891e-06,
+      "loss": 0.4513,
+      "step": 389
+    },
+    {
+      "epoch": 0.031997374574393896,
+      "grad_norm": 0.7409984535414162,
+      "learning_rate": 9.99990660746495e-06,
+      "loss": 0.4848,
+      "step": 390
+    },
+    {
+      "epoch": 0.032079419124584646,
+      "grad_norm": 0.7011404749983224,
+      "learning_rate": 9.999898309857636e-06,
+      "loss": 0.4439,
+      "step": 391
+    },
+    {
+      "epoch": 0.0321614636747754,
+      "grad_norm": 0.6796475389623117,
+      "learning_rate": 9.999889659165538e-06,
+      "loss": 0.4673,
+      "step": 392
+    },
+    {
+      "epoch": 0.032243508224966154,
+      "grad_norm": 0.763895084792315,
+      "learning_rate": 9.999880655389265e-06,
+      "loss": 0.4577,
+      "step": 393
+    },
+    {
+      "epoch": 0.03232555277515691,
+      "grad_norm": 0.7460391447281582,
+      "learning_rate": 9.999871298529454e-06,
+      "loss": 0.4973,
+      "step": 394
+    },
+    {
+      "epoch": 0.03240759732534766,
+      "grad_norm": 0.8026987475580106,
+      "learning_rate": 9.999861588586761e-06,
+      "loss": 0.4855,
+      "step": 395
+    },
+    {
+      "epoch": 0.03248964187553842,
+      "grad_norm": 0.680503150870167,
+      "learning_rate": 9.999851525561878e-06,
+      "loss": 0.4329,
+      "step": 396
+    },
+    {
+      "epoch": 0.03257168642572917,
+      "grad_norm": 0.7736300784584109,
+      "learning_rate": 9.999841109455514e-06,
+      "loss": 0.4767,
+      "step": 397
+    },
+    {
+      "epoch": 0.03265373097591993,
+      "grad_norm": 0.7527041748308959,
+      "learning_rate": 9.999830340268403e-06,
+      "loss": 0.4914,
+      "step": 398
+    },
+    {
+      "epoch": 0.03273577552611068,
+      "grad_norm": 0.709155750640637,
+      "learning_rate": 9.999819218001306e-06,
+      "loss": 0.4505,
+      "step": 399
+    },
+    {
+      "epoch": 0.032817820076301435,
+      "grad_norm": 0.7187531434899711,
+      "learning_rate": 9.999807742655008e-06,
+      "loss": 0.4692,
+      "step": 400
+    },
+    {
+      "epoch": 0.032899864626492185,
+      "grad_norm": 0.7132689878929277,
+      "learning_rate": 9.99979591423032e-06,
+      "loss": 0.4793,
+      "step": 401
+    },
+    {
+      "epoch": 0.032981909176682936,
+      "grad_norm": 0.8172625089056298,
+      "learning_rate": 9.99978373272808e-06,
+      "loss": 0.4277,
+      "step": 402
+    },
+    {
+      "epoch": 0.03306395372687369,
+      "grad_norm": 0.7171741261002212,
+      "learning_rate": 9.999771198149142e-06,
+      "loss": 0.4471,
+      "step": 403
+    },
+    {
+      "epoch": 0.033145998277064444,
+      "grad_norm": 0.7422548952474903,
+      "learning_rate": 9.999758310494396e-06,
+      "loss": 0.4259,
+      "step": 404
+    },
+    {
+      "epoch": 0.0332280428272552,
+      "grad_norm": 0.6974807712110668,
+      "learning_rate": 9.999745069764752e-06,
+      "loss": 0.464,
+      "step": 405
+    },
+    {
+      "epoch": 0.03331008737744595,
+      "grad_norm": 0.8131548089263908,
+      "learning_rate": 9.999731475961144e-06,
+      "loss": 0.5007,
+      "step": 406
+    },
+    {
+      "epoch": 0.03339213192763671,
+      "grad_norm": 0.7143270848396664,
+      "learning_rate": 9.99971752908453e-06,
+      "loss": 0.4886,
+      "step": 407
+    },
+    {
+      "epoch": 0.03347417647782746,
+      "grad_norm": 0.6384133367635114,
+      "learning_rate": 9.999703229135897e-06,
+      "loss": 0.4425,
+      "step": 408
+    },
+    {
+      "epoch": 0.03355622102801822,
+      "grad_norm": 0.7166011710744892,
+      "learning_rate": 9.999688576116255e-06,
+      "loss": 0.4076,
+      "step": 409
+    },
+    {
+      "epoch": 0.03363826557820897,
+      "grad_norm": 0.7432014272460356,
+      "learning_rate": 9.999673570026639e-06,
+      "loss": 0.4692,
+      "step": 410
+    },
+    {
+      "epoch": 0.03372031012839972,
+      "grad_norm": 0.7100299945002516,
+      "learning_rate": 9.999658210868109e-06,
+      "loss": 0.4673,
+      "step": 411
+    },
+    {
+      "epoch": 0.033802354678590475,
+      "grad_norm": 0.7405432300457298,
+      "learning_rate": 9.999642498641747e-06,
+      "loss": 0.4717,
+      "step": 412
+    },
+    {
+      "epoch": 0.033884399228781226,
+      "grad_norm": 0.6661473555433549,
+      "learning_rate": 9.999626433348664e-06,
+      "loss": 0.4695,
+      "step": 413
+    },
+    {
+      "epoch": 0.03396644377897198,
+      "grad_norm": 0.708439477047409,
+      "learning_rate": 9.999610014989995e-06,
+      "loss": 0.4951,
+      "step": 414
+    },
+    {
+      "epoch": 0.034048488329162734,
+      "grad_norm": 0.753318168177001,
+      "learning_rate": 9.9995932435669e-06,
+      "loss": 0.4515,
+      "step": 415
+    },
+    {
+      "epoch": 0.03413053287935349,
+      "grad_norm": 0.695997847162687,
+      "learning_rate": 9.999576119080562e-06,
+      "loss": 0.4521,
+      "step": 416
+    },
+    {
+      "epoch": 0.03421257742954424,
+      "grad_norm": 0.6724188998671643,
+      "learning_rate": 9.999558641532192e-06,
+      "loss": 0.4934,
+      "step": 417
+    },
+    {
+      "epoch": 0.034294621979735,
+      "grad_norm": 0.68020854219338,
+      "learning_rate": 9.99954081092302e-06,
+      "loss": 0.4626,
+      "step": 418
+    },
+    {
+      "epoch": 0.03437666652992575,
+      "grad_norm": 0.71047396486306,
+      "learning_rate": 9.999522627254313e-06,
+      "loss": 0.4403,
+      "step": 419
+    },
+    {
+      "epoch": 0.03445871108011651,
+      "grad_norm": 0.7321064120600358,
+      "learning_rate": 9.999504090527347e-06,
+      "loss": 0.474,
+      "step": 420
+    },
+    {
+      "epoch": 0.03454075563030726,
+      "grad_norm": 0.7364852647853852,
+      "learning_rate": 9.999485200743434e-06,
+      "loss": 0.4488,
+      "step": 421
+    },
+    {
+      "epoch": 0.03462280018049801,
+      "grad_norm": 0.6757232304792319,
+      "learning_rate": 9.999465957903911e-06,
+      "loss": 0.4308,
+      "step": 422
+    },
+    {
+      "epoch": 0.034704844730688765,
+      "grad_norm": 0.724262273632303,
+      "learning_rate": 9.999446362010134e-06,
+      "loss": 0.439,
+      "step": 423
+    },
+    {
+      "epoch": 0.034786889280879515,
+      "grad_norm": 0.7066340834090356,
+      "learning_rate": 9.999426413063486e-06,
+      "loss": 0.4595,
+      "step": 424
+    },
+    {
+      "epoch": 0.03486893383107027,
+      "grad_norm": 0.7992278405540538,
+      "learning_rate": 9.999406111065377e-06,
+      "loss": 0.4283,
+      "step": 425
+    },
+    {
+      "epoch": 0.03495097838126102,
+      "grad_norm": 0.6896584188297401,
+      "learning_rate": 9.999385456017241e-06,
+      "loss": 0.4629,
+      "step": 426
+    },
+    {
+      "epoch": 0.03503302293145178,
+      "grad_norm": 0.7315870806705093,
+      "learning_rate": 9.999364447920536e-06,
+      "loss": 0.4729,
+      "step": 427
+    },
+    {
+      "epoch": 0.03511506748164253,
+      "grad_norm": 0.7124404297869105,
+      "learning_rate": 9.999343086776745e-06,
+      "loss": 0.4621,
+      "step": 428
+    },
+    {
+      "epoch": 0.03519711203183329,
+      "grad_norm": 0.7218482675853214,
+      "learning_rate": 9.999321372587378e-06,
+      "loss": 0.4385,
+      "step": 429
+    },
+    {
+      "epoch": 0.03527915658202404,
+      "grad_norm": 0.6910480264497098,
+      "learning_rate": 9.99929930535397e-06,
+      "loss": 0.4508,
+      "step": 430
+    },
+    {
+      "epoch": 0.03536120113221479,
+      "grad_norm": 0.6795431181855621,
+      "learning_rate": 9.999276885078073e-06,
+      "loss": 0.4644,
+      "step": 431
+    },
+    {
+      "epoch": 0.03544324568240555,
+      "grad_norm": 0.6637696137242858,
+      "learning_rate": 9.999254111761276e-06,
+      "loss": 0.4617,
+      "step": 432
+    },
+    {
+      "epoch": 0.0355252902325963,
+      "grad_norm": 0.7436163218033263,
+      "learning_rate": 9.999230985405186e-06,
+      "loss": 0.4664,
+      "step": 433
+    },
+    {
+      "epoch": 0.035607334782787055,
+      "grad_norm": 0.7233271439596533,
+      "learning_rate": 9.999207506011435e-06,
+      "loss": 0.5239,
+      "step": 434
+    },
+    {
+      "epoch": 0.035689379332977805,
+      "grad_norm": 0.6900246104928095,
+      "learning_rate": 9.999183673581683e-06,
+      "loss": 0.4548,
+      "step": 435
+    },
+    {
+      "epoch": 0.03577142388316856,
+      "grad_norm": 0.6987939441103572,
+      "learning_rate": 9.99915948811761e-06,
+      "loss": 0.5407,
+      "step": 436
+    },
+    {
+      "epoch": 0.03585346843335931,
+      "grad_norm": 0.7687307839914107,
+      "learning_rate": 9.999134949620926e-06,
+      "loss": 0.4643,
+      "step": 437
+    },
+    {
+      "epoch": 0.03593551298355007,
+      "grad_norm": 0.6595324250589515,
+      "learning_rate": 9.999110058093365e-06,
+      "loss": 0.4358,
+      "step": 438
+    },
+    {
+      "epoch": 0.03601755753374082,
+      "grad_norm": 0.7539446339530429,
+      "learning_rate": 9.999084813536682e-06,
+      "loss": 0.4907,
+      "step": 439
+    },
+    {
+      "epoch": 0.03609960208393157,
+      "grad_norm": 0.730052352121622,
+      "learning_rate": 9.999059215952663e-06,
+      "loss": 0.4482,
+      "step": 440
+    },
+    {
+      "epoch": 0.03618164663412233,
+      "grad_norm": 0.6761419069710816,
+      "learning_rate": 9.999033265343112e-06,
+      "loss": 0.4107,
+      "step": 441
+    },
+    {
+      "epoch": 0.03626369118431308,
+      "grad_norm": 0.6594305253554813,
+      "learning_rate": 9.999006961709863e-06,
+      "loss": 0.4147,
+      "step": 442
+    },
+    {
+      "epoch": 0.03634573573450384,
+      "grad_norm": 0.7092653478339592,
+      "learning_rate": 9.998980305054775e-06,
+      "loss": 0.4765,
+      "step": 443
+    },
+    {
+      "epoch": 0.03642778028469459,
+      "grad_norm": 0.7499015791093236,
+      "learning_rate": 9.99895329537973e-06,
+      "loss": 0.4689,
+      "step": 444
+    },
+    {
+      "epoch": 0.036509824834885345,
+      "grad_norm": 0.7978937527632346,
+      "learning_rate": 9.998925932686633e-06,
+      "loss": 0.4946,
+      "step": 445
+    },
+    {
+      "epoch": 0.036591869385076095,
+      "grad_norm": 0.724344233116603,
+      "learning_rate": 9.998898216977418e-06,
+      "loss": 0.4365,
+      "step": 446
+    },
+    {
+      "epoch": 0.03667391393526685,
+      "grad_norm": 0.6736505035140128,
+      "learning_rate": 9.998870148254044e-06,
+      "loss": 0.5005,
+      "step": 447
+    },
+    {
+      "epoch": 0.0367559584854576,
+      "grad_norm": 0.7894997877095115,
+      "learning_rate": 9.99884172651849e-06,
+      "loss": 0.4295,
+      "step": 448
+    },
+    {
+      "epoch": 0.03683800303564836,
+      "grad_norm": 0.746625330966946,
+      "learning_rate": 9.998812951772765e-06,
+      "loss": 0.4315,
+      "step": 449
+    },
+    {
+      "epoch": 0.03692004758583911,
+      "grad_norm": 0.7928060163011839,
+      "learning_rate": 9.998783824018901e-06,
+      "loss": 0.4638,
+      "step": 450
+    },
+    {
+      "epoch": 0.03700209213602986,
+      "grad_norm": 0.6666179772167592,
+      "learning_rate": 9.998754343258954e-06,
+      "loss": 0.4671,
+      "step": 451
+    },
+    {
+      "epoch": 0.03708413668622062,
+      "grad_norm": 0.7480588607900501,
+      "learning_rate": 9.998724509495006e-06,
+      "loss": 0.4291,
+      "step": 452
+    },
+    {
+      "epoch": 0.03716618123641137,
+      "grad_norm": 0.8433380311067008,
+      "learning_rate": 9.998694322729165e-06,
+      "loss": 0.4825,
+      "step": 453
+    },
+    {
+      "epoch": 0.03724822578660213,
+      "grad_norm": 0.7209179605412437,
+      "learning_rate": 9.998663782963562e-06,
+      "loss": 0.4688,
+      "step": 454
+    },
+    {
+      "epoch": 0.03733027033679288,
+      "grad_norm": 0.751286971745102,
+      "learning_rate": 9.998632890200352e-06,
+      "loss": 0.4617,
+      "step": 455
+    },
+    {
+      "epoch": 0.037412314886983634,
+      "grad_norm": 0.7190716216183523,
+      "learning_rate": 9.99860164444172e-06,
+      "loss": 0.4544,
+      "step": 456
+    },
+    {
+      "epoch": 0.037494359437174385,
+      "grad_norm": 0.7178346946916926,
+      "learning_rate": 9.998570045689871e-06,
+      "loss": 0.4537,
+      "step": 457
+    },
+    {
+      "epoch": 0.03757640398736514,
+      "grad_norm": 0.7335286041651922,
+      "learning_rate": 9.998538093947035e-06,
+      "loss": 0.4586,
+      "step": 458
+    },
+    {
+      "epoch": 0.03765844853755589,
+      "grad_norm": 0.7642983229273737,
+      "learning_rate": 9.998505789215469e-06,
+      "loss": 0.4838,
+      "step": 459
+    },
+    {
+      "epoch": 0.03774049308774664,
+      "grad_norm": 0.7215990458463024,
+      "learning_rate": 9.998473131497456e-06,
+      "loss": 0.4531,
+      "step": 460
+    },
+    {
+      "epoch": 0.0378225376379374,
+      "grad_norm": 0.6603331659625503,
+      "learning_rate": 9.998440120795301e-06,
+      "loss": 0.4234,
+      "step": 461
+    },
+    {
+      "epoch": 0.03790458218812815,
+      "grad_norm": 0.6426368880149491,
+      "learning_rate": 9.998406757111334e-06,
+      "loss": 0.4144,
+      "step": 462
+    },
+    {
+      "epoch": 0.03798662673831891,
+      "grad_norm": 0.6710218677280847,
+      "learning_rate": 9.998373040447916e-06,
+      "loss": 0.431,
+      "step": 463
+    },
+    {
+      "epoch": 0.03806867128850966,
+      "grad_norm": 0.7293616851891388,
+      "learning_rate": 9.99833897080742e-06,
+      "loss": 0.4442,
+      "step": 464
+    },
+    {
+      "epoch": 0.038150715838700416,
+      "grad_norm": 0.7582560884861719,
+      "learning_rate": 9.998304548192259e-06,
+      "loss": 0.4618,
+      "step": 465
+    },
+    {
+      "epoch": 0.03823276038889117,
+      "grad_norm": 0.736659797658322,
+      "learning_rate": 9.99826977260486e-06,
+      "loss": 0.4641,
+      "step": 466
+    },
+    {
+      "epoch": 0.038314804939081924,
+      "grad_norm": 0.6772266753444388,
+      "learning_rate": 9.99823464404768e-06,
+      "loss": 0.4714,
+      "step": 467
+    },
+    {
+      "epoch": 0.038396849489272675,
+      "grad_norm": 0.705940911052447,
+      "learning_rate": 9.9981991625232e-06,
+      "loss": 0.457,
+      "step": 468
+    },
+    {
+      "epoch": 0.038478894039463425,
+      "grad_norm": 0.7065678446928986,
+      "learning_rate": 9.998163328033926e-06,
+      "loss": 0.4402,
+      "step": 469
+    },
+    {
+      "epoch": 0.03856093858965418,
+      "grad_norm": 0.7374084184063965,
+      "learning_rate": 9.998127140582387e-06,
+      "loss": 0.442,
+      "step": 470
+    },
+    {
+      "epoch": 0.03864298313984493,
+      "grad_norm": 0.7285734904027434,
+      "learning_rate": 9.99809060017114e-06,
+      "loss": 0.4674,
+      "step": 471
+    },
+    {
+      "epoch": 0.03872502769003569,
+      "grad_norm": 0.6646766165577739,
+      "learning_rate": 9.998053706802765e-06,
+      "loss": 0.4,
+      "step": 472
+    },
+    {
+      "epoch": 0.03880707224022644,
+      "grad_norm": 0.6986617164849681,
+      "learning_rate": 9.998016460479867e-06,
+      "loss": 0.4547,
+      "step": 473
+    },
+    {
+      "epoch": 0.0388891167904172,
+      "grad_norm": 0.7185044850083976,
+      "learning_rate": 9.997978861205077e-06,
+      "loss": 0.4405,
+      "step": 474
+    },
+    {
+      "epoch": 0.03897116134060795,
+      "grad_norm": 0.7216457821711093,
+      "learning_rate": 9.997940908981048e-06,
+      "loss": 0.4611,
+      "step": 475
+    },
+    {
+      "epoch": 0.039053205890798706,
+      "grad_norm": 0.6720259241671865,
+      "learning_rate": 9.997902603810464e-06,
+      "loss": 0.4385,
+      "step": 476
+    },
+    {
+      "epoch": 0.03913525044098946,
+      "grad_norm": 0.7192523710454518,
+      "learning_rate": 9.997863945696028e-06,
+      "loss": 0.4818,
+      "step": 477
+    },
+    {
+      "epoch": 0.039217294991180214,
+      "grad_norm": 0.7951320941996108,
+      "learning_rate": 9.997824934640467e-06,
+      "loss": 0.4588,
+      "step": 478
+    },
+    {
+      "epoch": 0.039299339541370965,
+      "grad_norm": 0.7104438699926404,
+      "learning_rate": 9.99778557064654e-06,
+      "loss": 0.4971,
+      "step": 479
+    },
+    {
+      "epoch": 0.039381384091561715,
+      "grad_norm": 0.7182926253126805,
+      "learning_rate": 9.997745853717027e-06,
+      "loss": 0.4486,
+      "step": 480
+    },
+    {
+      "epoch": 0.03946342864175247,
+      "grad_norm": 0.7253063130456066,
+      "learning_rate": 9.99770578385473e-06,
+      "loss": 0.446,
+      "step": 481
+    },
+    {
+      "epoch": 0.03954547319194322,
+      "grad_norm": 0.7176852772159676,
+      "learning_rate": 9.99766536106248e-06,
+      "loss": 0.4282,
+      "step": 482
+    },
+    {
+      "epoch": 0.03962751774213398,
+      "grad_norm": 0.704270709301284,
+      "learning_rate": 9.997624585343131e-06,
+      "loss": 0.4723,
+      "step": 483
+    },
+    {
+      "epoch": 0.03970956229232473,
+      "grad_norm": 0.731241569059469,
+      "learning_rate": 9.997583456699564e-06,
+      "loss": 0.5358,
+      "step": 484
+    },
+    {
+      "epoch": 0.03979160684251549,
+      "grad_norm": 0.7624455583420627,
+      "learning_rate": 9.997541975134682e-06,
+      "loss": 0.4621,
+      "step": 485
+    },
+    {
+      "epoch": 0.03987365139270624,
+      "grad_norm": 0.7092929275275877,
+      "learning_rate": 9.997500140651415e-06,
+      "loss": 0.4722,
+      "step": 486
+    },
+    {
+      "epoch": 0.039955695942896996,
+      "grad_norm": 0.698531382187613,
+      "learning_rate": 9.997457953252718e-06,
+      "loss": 0.4766,
+      "step": 487
+    },
+    {
+      "epoch": 0.040037740493087746,
+      "grad_norm": 0.6919703903092729,
+      "learning_rate": 9.997415412941568e-06,
+      "loss": 0.4153,
+      "step": 488
+    },
+    {
+      "epoch": 0.0401197850432785,
+      "grad_norm": 0.7150938850897955,
+      "learning_rate": 9.997372519720972e-06,
+      "loss": 0.4678,
+      "step": 489
+    },
+    {
+      "epoch": 0.040201829593469254,
+      "grad_norm": 0.7149162155857924,
+      "learning_rate": 9.997329273593957e-06,
+      "loss": 0.4458,
+      "step": 490
+    },
+    {
+      "epoch": 0.040283874143660005,
+      "grad_norm": 0.730249746093433,
+      "learning_rate": 9.997285674563577e-06,
+      "loss": 0.4338,
+      "step": 491
+    },
+    {
+      "epoch": 0.04036591869385076,
+      "grad_norm": 0.7081732480745071,
+      "learning_rate": 9.997241722632912e-06,
+      "loss": 0.4437,
+      "step": 492
+    },
+    {
+      "epoch": 0.04044796324404151,
+      "grad_norm": 0.6914755232958761,
+      "learning_rate": 9.997197417805064e-06,
+      "loss": 0.4151,
+      "step": 493
+    },
+    {
+      "epoch": 0.04053000779423227,
+      "grad_norm": 0.7576900550207197,
+      "learning_rate": 9.997152760083165e-06,
+      "loss": 0.4264,
+      "step": 494
+    },
+    {
+      "epoch": 0.04061205234442302,
+      "grad_norm": 0.7075658489323291,
+      "learning_rate": 9.997107749470365e-06,
+      "loss": 0.452,
+      "step": 495
+    },
+    {
+      "epoch": 0.04069409689461378,
+      "grad_norm": 0.6762066391470457,
+      "learning_rate": 9.997062385969845e-06,
+      "loss": 0.4725,
+      "step": 496
+    },
+    {
+      "epoch": 0.04077614144480453,
+      "grad_norm": 0.6494608258866502,
+      "learning_rate": 9.997016669584808e-06,
+      "loss": 0.4102,
+      "step": 497
+    },
+    {
+      "epoch": 0.040858185994995286,
+      "grad_norm": 0.6559713877689529,
+      "learning_rate": 9.996970600318481e-06,
+      "loss": 0.4727,
+      "step": 498
+    },
+    {
+      "epoch": 0.040940230545186036,
+      "grad_norm": 0.7685267669463628,
+      "learning_rate": 9.996924178174119e-06,
+      "loss": 0.4341,
+      "step": 499
+    },
+    {
+      "epoch": 0.04102227509537679,
+      "grad_norm": 0.6566814642735063,
+      "learning_rate": 9.996877403155e-06,
+      "loss": 0.4553,
+      "step": 500
+    },
+    {
+      "epoch": 0.041104319645567544,
+      "grad_norm": 0.7156057846306229,
+      "learning_rate": 9.996830275264426e-06,
+      "loss": 0.4826,
+      "step": 501
+    },
+    {
+      "epoch": 0.041186364195758295,
+      "grad_norm": 0.6825683349608124,
+      "learning_rate": 9.996782794505727e-06,
+      "loss": 0.4471,
+      "step": 502
+    },
+    {
+      "epoch": 0.04126840874594905,
+      "grad_norm": 0.7377768936658535,
+      "learning_rate": 9.996734960882254e-06,
+      "loss": 0.4812,
+      "step": 503
+    },
+    {
+      "epoch": 0.0413504532961398,
+      "grad_norm": 0.7078741000048431,
+      "learning_rate": 9.996686774397389e-06,
+      "loss": 0.4646,
+      "step": 504
+    },
+    {
+      "epoch": 0.04143249784633056,
+      "grad_norm": 0.6916688747392733,
+      "learning_rate": 9.996638235054527e-06,
+      "loss": 0.4482,
+      "step": 505
+    },
+    {
+      "epoch": 0.04151454239652131,
+      "grad_norm": 0.7093691406773603,
+      "learning_rate": 9.996589342857104e-06,
+      "loss": 0.4381,
+      "step": 506
+    },
+    {
+      "epoch": 0.04159658694671207,
+      "grad_norm": 0.7107470230584804,
+      "learning_rate": 9.996540097808567e-06,
+      "loss": 0.4375,
+      "step": 507
+    },
+    {
+      "epoch": 0.04167863149690282,
+      "grad_norm": 0.6324066038387123,
+      "learning_rate": 9.996490499912399e-06,
+      "loss": 0.4992,
+      "step": 508
+    },
+    {
+      "epoch": 0.04176067604709357,
+      "grad_norm": 0.7072477675604203,
+      "learning_rate": 9.996440549172096e-06,
+      "loss": 0.4735,
+      "step": 509
+    },
+    {
+      "epoch": 0.041842720597284326,
+      "grad_norm": 0.7177934794210498,
+      "learning_rate": 9.99639024559119e-06,
+      "loss": 0.4492,
+      "step": 510
+    },
+    {
+      "epoch": 0.04192476514747508,
+      "grad_norm": 0.7839993301929503,
+      "learning_rate": 9.996339589173232e-06,
+      "loss": 0.4718,
+      "step": 511
+    },
+    {
+      "epoch": 0.042006809697665834,
+      "grad_norm": 0.6883700832680933,
+      "learning_rate": 9.9962885799218e-06,
+      "loss": 0.4375,
+      "step": 512
+    },
+    {
+      "epoch": 0.042088854247856584,
+      "grad_norm": 0.7538134608751539,
+      "learning_rate": 9.996237217840496e-06,
+      "loss": 0.4539,
+      "step": 513
+    },
+    {
+      "epoch": 0.04217089879804734,
+      "grad_norm": 0.7843009784283251,
+      "learning_rate": 9.996185502932946e-06,
+      "loss": 0.5041,
+      "step": 514
+    },
+    {
+      "epoch": 0.04225294334823809,
+      "grad_norm": 0.710109448349278,
+      "learning_rate": 9.996133435202804e-06,
+      "loss": 0.4285,
+      "step": 515
+    },
+    {
+      "epoch": 0.04233498789842885,
+      "grad_norm": 0.6872200777377337,
+      "learning_rate": 9.996081014653743e-06,
+      "loss": 0.4305,
+      "step": 516
+    },
+    {
+      "epoch": 0.0424170324486196,
+      "grad_norm": 0.7222067037926857,
+      "learning_rate": 9.99602824128947e-06,
+      "loss": 0.4794,
+      "step": 517
+    },
+    {
+      "epoch": 0.04249907699881035,
+      "grad_norm": 0.7659243964102268,
+      "learning_rate": 9.995975115113708e-06,
+      "loss": 0.4832,
+      "step": 518
+    },
+    {
+      "epoch": 0.04258112154900111,
+      "grad_norm": 0.7327686631577982,
+      "learning_rate": 9.99592163613021e-06,
+      "loss": 0.4511,
+      "step": 519
+    },
+    {
+      "epoch": 0.04266316609919186,
+      "grad_norm": 0.6703624799860225,
+      "learning_rate": 9.995867804342754e-06,
+      "loss": 0.4697,
+      "step": 520
+    },
+    {
+      "epoch": 0.042745210649382616,
+      "grad_norm": 0.721676156420074,
+      "learning_rate": 9.99581361975514e-06,
+      "loss": 0.4516,
+      "step": 521
+    },
+    {
+      "epoch": 0.042827255199573366,
+      "grad_norm": 0.6630846745996025,
+      "learning_rate": 9.995759082371192e-06,
+      "loss": 0.417,
+      "step": 522
+    },
+    {
+      "epoch": 0.042909299749764124,
+      "grad_norm": 0.7097672736014867,
+      "learning_rate": 9.995704192194767e-06,
+      "loss": 0.4347,
+      "step": 523
+    },
+    {
+      "epoch": 0.042991344299954874,
+      "grad_norm": 0.6922360012576984,
+      "learning_rate": 9.995648949229737e-06,
+      "loss": 0.4381,
+      "step": 524
+    },
+    {
+      "epoch": 0.04307338885014563,
+      "grad_norm": 0.6634569626058372,
+      "learning_rate": 9.995593353480004e-06,
+      "loss": 0.4401,
+      "step": 525
+    },
+    {
+      "epoch": 0.04315543340033638,
+      "grad_norm": 0.7200049412486922,
+      "learning_rate": 9.995537404949493e-06,
+      "loss": 0.4383,
+      "step": 526
+    },
+    {
+      "epoch": 0.04323747795052714,
+      "grad_norm": 0.6980183502849567,
+      "learning_rate": 9.99548110364216e-06,
+      "loss": 0.4618,
+      "step": 527
+    },
+    {
+      "epoch": 0.04331952250071789,
+      "grad_norm": 0.7411025243528779,
+      "learning_rate": 9.995424449561974e-06,
+      "loss": 0.4553,
+      "step": 528
+    },
+    {
+      "epoch": 0.04340156705090864,
+      "grad_norm": 0.6987726191321368,
+      "learning_rate": 9.995367442712941e-06,
+      "loss": 0.456,
+      "step": 529
+    },
+    {
+      "epoch": 0.0434836116010994,
+      "grad_norm": 0.6758170889261959,
+      "learning_rate": 9.995310083099085e-06,
+      "loss": 0.4667,
+      "step": 530
+    },
+    {
+      "epoch": 0.04356565615129015,
+      "grad_norm": 0.7657406549249167,
+      "learning_rate": 9.995252370724455e-06,
+      "loss": 0.5104,
+      "step": 531
+    },
+    {
+      "epoch": 0.043647700701480906,
+      "grad_norm": 0.6294043636795663,
+      "learning_rate": 9.99519430559313e-06,
+      "loss": 0.3916,
+      "step": 532
+    },
+    {
+      "epoch": 0.043729745251671656,
+      "grad_norm": 0.6853343515399332,
+      "learning_rate": 9.995135887709207e-06,
+      "loss": 0.4489,
+      "step": 533
+    },
+    {
+      "epoch": 0.043811789801862414,
+      "grad_norm": 0.6990569414613095,
+      "learning_rate": 9.995077117076812e-06,
+      "loss": 0.4515,
+      "step": 534
+    },
+    {
+      "epoch": 0.043893834352053164,
+      "grad_norm": 0.7231041042173018,
+      "learning_rate": 9.995017993700099e-06,
+      "loss": 0.4444,
+      "step": 535
+    },
+    {
+      "epoch": 0.04397587890224392,
+      "grad_norm": 0.7278484004519378,
+      "learning_rate": 9.994958517583238e-06,
+      "loss": 0.4881,
+      "step": 536
+    },
+    {
+      "epoch": 0.04405792345243467,
+      "grad_norm": 0.706974439532456,
+      "learning_rate": 9.994898688730432e-06,
+      "loss": 0.4653,
+      "step": 537
+    },
+    {
+      "epoch": 0.04413996800262542,
+      "grad_norm": 0.753641900564539,
+      "learning_rate": 9.994838507145907e-06,
+      "loss": 0.493,
+      "step": 538
+    },
+    {
+      "epoch": 0.04422201255281618,
+      "grad_norm": 0.6452820472005864,
+      "learning_rate": 9.994777972833911e-06,
+      "loss": 0.4719,
+      "step": 539
+    },
+    {
+      "epoch": 0.04430405710300693,
+      "grad_norm": 0.6210280686229739,
+      "learning_rate": 9.994717085798718e-06,
+      "loss": 0.3934,
+      "step": 540
+    },
+    {
+      "epoch": 0.04438610165319769,
+      "grad_norm": 0.7127122239825074,
+      "learning_rate": 9.99465584604463e-06,
+      "loss": 0.4375,
+      "step": 541
+    },
+    {
+      "epoch": 0.04446814620338844,
+      "grad_norm": 0.6866351515332538,
+      "learning_rate": 9.994594253575972e-06,
+      "loss": 0.4679,
+      "step": 542
+    },
+    {
+      "epoch": 0.044550190753579196,
+      "grad_norm": 0.7834169598279863,
+      "learning_rate": 9.99453230839709e-06,
+      "loss": 0.4325,
+      "step": 543
+    },
+    {
+      "epoch": 0.044632235303769946,
+      "grad_norm": 0.859796079968609,
+      "learning_rate": 9.99447001051236e-06,
+      "loss": 0.4925,
+      "step": 544
+    },
+    {
+      "epoch": 0.0447142798539607,
+      "grad_norm": 0.7033137539868488,
+      "learning_rate": 9.994407359926184e-06,
+      "loss": 0.4632,
+      "step": 545
+    },
+    {
+      "epoch": 0.044796324404151454,
+      "grad_norm": 0.6838865216195213,
+      "learning_rate": 9.994344356642982e-06,
+      "loss": 0.4495,
+      "step": 546
+    },
+    {
+      "epoch": 0.04487836895434221,
+      "grad_norm": 0.7027823168201869,
+      "learning_rate": 9.99428100066721e-06,
+      "loss": 0.4408,
+      "step": 547
+    },
+    {
+      "epoch": 0.04496041350453296,
+      "grad_norm": 0.668474461707662,
+      "learning_rate": 9.994217292003334e-06,
+      "loss": 0.4763,
+      "step": 548
+    },
+    {
+      "epoch": 0.04504245805472371,
+      "grad_norm": 0.740202838495107,
+      "learning_rate": 9.994153230655857e-06,
+      "loss": 0.4669,
+      "step": 549
+    },
+    {
+      "epoch": 0.04512450260491447,
+      "grad_norm": 0.6340188489242965,
+      "learning_rate": 9.994088816629303e-06,
+      "loss": 0.4467,
+      "step": 550
+    },
+    {
+      "epoch": 0.04520654715510522,
+      "grad_norm": 0.7530366377950458,
+      "learning_rate": 9.994024049928222e-06,
+      "loss": 0.4393,
+      "step": 551
+    },
+    {
+      "epoch": 0.04528859170529598,
+      "grad_norm": 0.6181649478173531,
+      "learning_rate": 9.993958930557184e-06,
+      "loss": 0.4162,
+      "step": 552
+    },
+    {
+      "epoch": 0.04537063625548673,
+      "grad_norm": 0.6734838772803318,
+      "learning_rate": 9.993893458520791e-06,
+      "loss": 0.4586,
+      "step": 553
+    },
+    {
+      "epoch": 0.045452680805677485,
+      "grad_norm": 0.6868737860674475,
+      "learning_rate": 9.993827633823665e-06,
+      "loss": 0.455,
+      "step": 554
+    },
+    {
+      "epoch": 0.045534725355868236,
+      "grad_norm": 0.7524520698554804,
+      "learning_rate": 9.993761456470454e-06,
+      "loss": 0.4833,
+      "step": 555
+    },
+    {
+      "epoch": 0.04561676990605899,
+      "grad_norm": 0.6954175807615828,
+      "learning_rate": 9.993694926465833e-06,
+      "loss": 0.4339,
+      "step": 556
+    },
+    {
+      "epoch": 0.045698814456249744,
+      "grad_norm": 0.6897275480699461,
+      "learning_rate": 9.993628043814499e-06,
+      "loss": 0.4553,
+      "step": 557
+    },
+    {
+      "epoch": 0.045780859006440494,
+      "grad_norm": 0.7063664026617026,
+      "learning_rate": 9.993560808521174e-06,
+      "loss": 0.4647,
+      "step": 558
+    },
+    {
+      "epoch": 0.04586290355663125,
+      "grad_norm": 0.7381091711503153,
+      "learning_rate": 9.99349322059061e-06,
+      "loss": 0.4605,
+      "step": 559
+    },
+    {
+      "epoch": 0.045944948106822,
+      "grad_norm": 0.7139668062011871,
+      "learning_rate": 9.993425280027576e-06,
+      "loss": 0.441,
+      "step": 560
+    },
+    {
+      "epoch": 0.04602699265701276,
+      "grad_norm": 0.6638210871415727,
+      "learning_rate": 9.993356986836871e-06,
+      "loss": 0.4407,
+      "step": 561
+    },
+    {
+      "epoch": 0.04610903720720351,
+      "grad_norm": 0.6525424575055189,
+      "learning_rate": 9.993288341023317e-06,
+      "loss": 0.4585,
+      "step": 562
+    },
+    {
+      "epoch": 0.04619108175739427,
+      "grad_norm": 0.6742347788315354,
+      "learning_rate": 9.993219342591764e-06,
+      "loss": 0.4252,
+      "step": 563
+    },
+    {
+      "epoch": 0.04627312630758502,
+      "grad_norm": 0.693888837412908,
+      "learning_rate": 9.993149991547085e-06,
+      "loss": 0.4361,
+      "step": 564
+    },
+    {
+      "epoch": 0.046355170857775775,
+      "grad_norm": 0.7246773886954004,
+      "learning_rate": 9.993080287894172e-06,
+      "loss": 0.4401,
+      "step": 565
+    },
+    {
+      "epoch": 0.046437215407966526,
+      "grad_norm": 0.662869896661773,
+      "learning_rate": 9.993010231637954e-06,
+      "loss": 0.4414,
+      "step": 566
+    },
+    {
+      "epoch": 0.046519259958157276,
+      "grad_norm": 0.644278877153484,
+      "learning_rate": 9.992939822783374e-06,
+      "loss": 0.4304,
+      "step": 567
+    },
+    {
+      "epoch": 0.04660130450834803,
+      "grad_norm": 0.7107919870019649,
+      "learning_rate": 9.992869061335406e-06,
+      "loss": 0.4656,
+      "step": 568
+    },
+    {
+      "epoch": 0.046683349058538784,
+      "grad_norm": 0.7295416812611188,
+      "learning_rate": 9.992797947299046e-06,
+      "loss": 0.4581,
+      "step": 569
+    },
+    {
+      "epoch": 0.04676539360872954,
+      "grad_norm": 0.6945077019281005,
+      "learning_rate": 9.992726480679318e-06,
+      "loss": 0.468,
+      "step": 570
+    },
+    {
+      "epoch": 0.04684743815892029,
+      "grad_norm": 0.7060106582479795,
+      "learning_rate": 9.992654661481265e-06,
+      "loss": 0.4237,
+      "step": 571
+    },
+    {
+      "epoch": 0.04692948270911105,
+      "grad_norm": 0.7453770147494599,
+      "learning_rate": 9.992582489709963e-06,
+      "loss": 0.4667,
+      "step": 572
+    },
+    {
+      "epoch": 0.0470115272593018,
+      "grad_norm": 0.6927326292995923,
+      "learning_rate": 9.992509965370507e-06,
+      "loss": 0.455,
+      "step": 573
+    },
+    {
+      "epoch": 0.04709357180949256,
+      "grad_norm": 0.6709403457146476,
+      "learning_rate": 9.992437088468016e-06,
+      "loss": 0.4195,
+      "step": 574
+    },
+    {
+      "epoch": 0.04717561635968331,
+      "grad_norm": 0.7094911587840689,
+      "learning_rate": 9.992363859007642e-06,
+      "loss": 0.4536,
+      "step": 575
+    },
+    {
+      "epoch": 0.047257660909874065,
+      "grad_norm": 0.7971641616069425,
+      "learning_rate": 9.992290276994551e-06,
+      "loss": 0.4519,
+      "step": 576
+    },
+    {
+      "epoch": 0.047339705460064815,
+      "grad_norm": 0.7041060371667365,
+      "learning_rate": 9.992216342433943e-06,
+      "loss": 0.4667,
+      "step": 577
+    },
+    {
+      "epoch": 0.047421750010255566,
+      "grad_norm": 0.6953357617765562,
+      "learning_rate": 9.992142055331035e-06,
+      "loss": 0.4768,
+      "step": 578
+    },
+    {
+      "epoch": 0.04750379456044632,
+      "grad_norm": 0.6928049706538878,
+      "learning_rate": 9.992067415691077e-06,
+      "loss": 0.4236,
+      "step": 579
+    },
+    {
+      "epoch": 0.047585839110637074,
+      "grad_norm": 0.7587579542428657,
+      "learning_rate": 9.991992423519338e-06,
+      "loss": 0.4589,
+      "step": 580
+    },
+    {
+      "epoch": 0.04766788366082783,
+      "grad_norm": 0.7475497620384385,
+      "learning_rate": 9.991917078821113e-06,
+      "loss": 0.4558,
+      "step": 581
+    },
+    {
+      "epoch": 0.04774992821101858,
+      "grad_norm": 0.7949899318799087,
+      "learning_rate": 9.991841381601728e-06,
+      "loss": 0.4632,
+      "step": 582
+    },
+    {
+      "epoch": 0.04783197276120934,
+      "grad_norm": 0.6809040490606169,
+      "learning_rate": 9.99176533186652e-06,
+      "loss": 0.4902,
+      "step": 583
+    },
+    {
+      "epoch": 0.04791401731140009,
+      "grad_norm": 0.6979975772212924,
+      "learning_rate": 9.991688929620869e-06,
+      "loss": 0.4302,
+      "step": 584
+    },
+    {
+      "epoch": 0.04799606186159085,
+      "grad_norm": 0.7085283940553014,
+      "learning_rate": 9.991612174870163e-06,
+      "loss": 0.445,
+      "step": 585
+    },
+    {
+      "epoch": 0.0480781064117816,
+      "grad_norm": 0.6865328729503544,
+      "learning_rate": 9.991535067619824e-06,
+      "loss": 0.4221,
+      "step": 586
+    },
+    {
+      "epoch": 0.04816015096197235,
+      "grad_norm": 0.7027661578641745,
+      "learning_rate": 9.991457607875298e-06,
+      "loss": 0.4329,
+      "step": 587
+    },
+    {
+      "epoch": 0.048242195512163105,
+      "grad_norm": 0.7038563262782168,
+      "learning_rate": 9.991379795642057e-06,
+      "loss": 0.418,
+      "step": 588
+    },
+    {
+      "epoch": 0.048324240062353856,
+      "grad_norm": 0.7123180975594244,
+      "learning_rate": 9.991301630925592e-06,
+      "loss": 0.4249,
+      "step": 589
+    },
+    {
+      "epoch": 0.04840628461254461,
+      "grad_norm": 0.6603136226905367,
+      "learning_rate": 9.991223113731427e-06,
+      "loss": 0.4181,
+      "step": 590
+    },
+    {
+      "epoch": 0.048488329162735364,
+      "grad_norm": 0.7058062451971719,
+      "learning_rate": 9.991144244065104e-06,
+      "loss": 0.4483,
+      "step": 591
+    },
+    {
+      "epoch": 0.04857037371292612,
+      "grad_norm": 0.7727232297860002,
+      "learning_rate": 9.991065021932193e-06,
+      "loss": 0.4556,
+      "step": 592
+    },
+    {
+      "epoch": 0.04865241826311687,
+      "grad_norm": 0.7713043644835836,
+      "learning_rate": 9.990985447338288e-06,
+      "loss": 0.4601,
+      "step": 593
+    },
+    {
+      "epoch": 0.04873446281330763,
+      "grad_norm": 0.7830172006085198,
+      "learning_rate": 9.99090552028901e-06,
+      "loss": 0.466,
+      "step": 594
+    },
+    {
+      "epoch": 0.04881650736349838,
+      "grad_norm": 0.7260927684934873,
+      "learning_rate": 9.990825240790002e-06,
+      "loss": 0.4262,
+      "step": 595
+    },
+    {
+      "epoch": 0.04889855191368913,
+      "grad_norm": 0.7579980410982813,
+      "learning_rate": 9.990744608846936e-06,
+      "loss": 0.457,
+      "step": 596
+    },
+    {
+      "epoch": 0.04898059646387989,
+      "grad_norm": 0.7333740114144691,
+      "learning_rate": 9.990663624465504e-06,
+      "loss": 0.4256,
+      "step": 597
+    },
+    {
+      "epoch": 0.04906264101407064,
+      "grad_norm": 0.6932880433987787,
+      "learning_rate": 9.990582287651423e-06,
+      "loss": 0.4655,
+      "step": 598
+    },
+    {
+      "epoch": 0.049144685564261395,
+      "grad_norm": 0.7302920833272335,
+      "learning_rate": 9.99050059841044e-06,
+      "loss": 0.4624,
+      "step": 599
+    },
+    {
+      "epoch": 0.049226730114452145,
+      "grad_norm": 0.7322994154107735,
+      "learning_rate": 9.990418556748323e-06,
+      "loss": 0.4612,
+      "step": 600
+    },
+    {
+      "epoch": 0.0493087746646429,
+      "grad_norm": 0.723805484599338,
+      "learning_rate": 9.990336162670863e-06,
+      "loss": 0.4543,
+      "step": 601
+    },
+    {
+      "epoch": 0.04939081921483365,
+      "grad_norm": 0.7083508741651389,
+      "learning_rate": 9.990253416183882e-06,
+      "loss": 0.4632,
+      "step": 602
+    },
+    {
+      "epoch": 0.04947286376502441,
+      "grad_norm": 0.6493145592385119,
+      "learning_rate": 9.990170317293224e-06,
+      "loss": 0.4458,
+      "step": 603
+    },
+    {
+      "epoch": 0.04955490831521516,
+      "grad_norm": 0.7114681999241342,
+      "learning_rate": 9.990086866004753e-06,
+      "loss": 0.4775,
+      "step": 604
+    },
+    {
+      "epoch": 0.04963695286540592,
+      "grad_norm": 0.6424219480407392,
+      "learning_rate": 9.990003062324367e-06,
+      "loss": 0.4476,
+      "step": 605
+    },
+    {
+      "epoch": 0.04971899741559667,
+      "grad_norm": 0.7602770197051834,
+      "learning_rate": 9.989918906257979e-06,
+      "loss": 0.4644,
+      "step": 606
+    },
+    {
+      "epoch": 0.04980104196578742,
+      "grad_norm": 0.7408866090717591,
+      "learning_rate": 9.989834397811536e-06,
+      "loss": 0.4429,
+      "step": 607
+    },
+    {
+      "epoch": 0.04988308651597818,
+      "grad_norm": 0.6918586784424133,
+      "learning_rate": 9.989749536991005e-06,
+      "loss": 0.404,
+      "step": 608
+    },
+    {
+      "epoch": 0.04996513106616893,
+      "grad_norm": 0.7595382656960332,
+      "learning_rate": 9.989664323802378e-06,
+      "loss": 0.4535,
+      "step": 609
+    },
+    {
+      "epoch": 0.050047175616359685,
+      "grad_norm": 0.7524547885802731,
+      "learning_rate": 9.989578758251673e-06,
+      "loss": 0.4142,
+      "step": 610
+    },
+    {
+      "epoch": 0.050129220166550435,
+      "grad_norm": 0.6483969395237272,
+      "learning_rate": 9.989492840344934e-06,
+      "loss": 0.4095,
+      "step": 611
+    },
+    {
+      "epoch": 0.05021126471674119,
+      "grad_norm": 0.7086713342832945,
+      "learning_rate": 9.989406570088225e-06,
+      "loss": 0.4926,
+      "step": 612
+    },
+    {
+      "epoch": 0.05029330926693194,
+      "grad_norm": 0.7025032620937127,
+      "learning_rate": 9.989319947487641e-06,
+      "loss": 0.4453,
+      "step": 613
+    },
+    {
+      "epoch": 0.0503753538171227,
+      "grad_norm": 0.6766595841801735,
+      "learning_rate": 9.989232972549296e-06,
+      "loss": 0.436,
+      "step": 614
+    },
+    {
+      "epoch": 0.05045739836731345,
+      "grad_norm": 0.6779705494842774,
+      "learning_rate": 9.989145645279337e-06,
+      "loss": 0.4244,
+      "step": 615
+    },
+    {
+      "epoch": 0.0505394429175042,
+      "grad_norm": 0.6459569642168315,
+      "learning_rate": 9.989057965683926e-06,
+      "loss": 0.4388,
+      "step": 616
+    },
+    {
+      "epoch": 0.05062148746769496,
+      "grad_norm": 0.6416203750432178,
+      "learning_rate": 9.988969933769259e-06,
+      "loss": 0.4486,
+      "step": 617
+    },
+    {
+      "epoch": 0.05070353201788571,
+      "grad_norm": 0.6820101077787526,
+      "learning_rate": 9.98888154954155e-06,
+      "loss": 0.4598,
+      "step": 618
+    },
+    {
+      "epoch": 0.05078557656807647,
+      "grad_norm": 0.7878438515212305,
+      "learning_rate": 9.988792813007042e-06,
+      "loss": 0.4295,
+      "step": 619
+    },
+    {
+      "epoch": 0.05086762111826722,
+      "grad_norm": 0.7037435249883632,
+      "learning_rate": 9.988703724172e-06,
+      "loss": 0.459,
+      "step": 620
+    },
+    {
+      "epoch": 0.050949665668457975,
+      "grad_norm": 0.7377998874700837,
+      "learning_rate": 9.988614283042715e-06,
+      "loss": 0.4247,
+      "step": 621
+    },
+    {
+      "epoch": 0.051031710218648725,
+      "grad_norm": 0.6787425617693291,
+      "learning_rate": 9.988524489625505e-06,
+      "loss": 0.4392,
+      "step": 622
+    },
+    {
+      "epoch": 0.05111375476883948,
+      "grad_norm": 0.6915539089894299,
+      "learning_rate": 9.98843434392671e-06,
+      "loss": 0.4127,
+      "step": 623
+    },
+    {
+      "epoch": 0.05119579931903023,
+      "grad_norm": 0.6327535605777087,
+      "learning_rate": 9.988343845952697e-06,
+      "loss": 0.426,
+      "step": 624
+    },
+    {
+      "epoch": 0.05127784386922099,
+      "grad_norm": 0.6870731267810483,
+      "learning_rate": 9.988252995709855e-06,
+      "loss": 0.4446,
+      "step": 625
+    },
+    {
+      "epoch": 0.05135988841941174,
+      "grad_norm": 0.6956723479089101,
+      "learning_rate": 9.9881617932046e-06,
+      "loss": 0.4818,
+      "step": 626
+    },
+    {
+      "epoch": 0.05144193296960249,
+      "grad_norm": 0.7156039341480762,
+      "learning_rate": 9.988070238443374e-06,
+      "loss": 0.4645,
+      "step": 627
+    },
+    {
+      "epoch": 0.05152397751979325,
+      "grad_norm": 0.6456941083799821,
+      "learning_rate": 9.987978331432643e-06,
+      "loss": 0.4351,
+      "step": 628
+    },
+    {
+      "epoch": 0.051606022069984,
+      "grad_norm": 0.7098513394944423,
+      "learning_rate": 9.987886072178896e-06,
+      "loss": 0.4534,
+      "step": 629
+    },
+    {
+      "epoch": 0.05168806662017476,
+      "grad_norm": 0.6822386545259626,
+      "learning_rate": 9.987793460688647e-06,
+      "loss": 0.4643,
+      "step": 630
+    },
+    {
+      "epoch": 0.05177011117036551,
+      "grad_norm": 0.68402902621502,
+      "learning_rate": 9.987700496968438e-06,
+      "loss": 0.4099,
+      "step": 631
+    },
+    {
+      "epoch": 0.051852155720556264,
+      "grad_norm": 0.646062415245863,
+      "learning_rate": 9.987607181024832e-06,
+      "loss": 0.4236,
+      "step": 632
+    },
+    {
+      "epoch": 0.051934200270747015,
+      "grad_norm": 0.7314396879814337,
+      "learning_rate": 9.98751351286442e-06,
+      "loss": 0.4475,
+      "step": 633
+    },
+    {
+      "epoch": 0.05201624482093777,
+      "grad_norm": 0.6561020453971673,
+      "learning_rate": 9.987419492493818e-06,
+      "loss": 0.4173,
+      "step": 634
+    },
+    {
+      "epoch": 0.05209828937112852,
+      "grad_norm": 0.6540807601222245,
+      "learning_rate": 9.987325119919664e-06,
+      "loss": 0.4194,
+      "step": 635
+    },
+    {
+      "epoch": 0.05218033392131927,
+      "grad_norm": 0.7027635707121694,
+      "learning_rate": 9.987230395148624e-06,
+      "loss": 0.44,
+      "step": 636
+    },
+    {
+      "epoch": 0.05226237847151003,
+      "grad_norm": 0.7142219738381628,
+      "learning_rate": 9.987135318187385e-06,
+      "loss": 0.4238,
+      "step": 637
+    },
+    {
+      "epoch": 0.05234442302170078,
+      "grad_norm": 0.7424527752180415,
+      "learning_rate": 9.98703988904266e-06,
+      "loss": 0.4616,
+      "step": 638
+    },
+    {
+      "epoch": 0.05242646757189154,
+      "grad_norm": 0.7776487531633278,
+      "learning_rate": 9.986944107721194e-06,
+      "loss": 0.4753,
+      "step": 639
+    },
+    {
+      "epoch": 0.05250851212208229,
+      "grad_norm": 0.6908363946190132,
+      "learning_rate": 9.986847974229745e-06,
+      "loss": 0.4089,
+      "step": 640
+    },
+    {
+      "epoch": 0.052590556672273046,
+      "grad_norm": 0.7043532209893897,
+      "learning_rate": 9.986751488575106e-06,
+      "loss": 0.4944,
+      "step": 641
+    },
+    {
+      "epoch": 0.0526726012224638,
+      "grad_norm": 0.7321317571846115,
+      "learning_rate": 9.986654650764086e-06,
+      "loss": 0.4505,
+      "step": 642
+    },
+    {
+      "epoch": 0.052754645772654554,
+      "grad_norm": 0.6529288915412493,
+      "learning_rate": 9.986557460803527e-06,
+      "loss": 0.4499,
+      "step": 643
+    },
+    {
+      "epoch": 0.052836690322845305,
+      "grad_norm": 0.6453874812953793,
+      "learning_rate": 9.986459918700293e-06,
+      "loss": 0.4547,
+      "step": 644
+    },
+    {
+      "epoch": 0.052918734873036055,
+      "grad_norm": 0.6861533265573238,
+      "learning_rate": 9.98636202446127e-06,
+      "loss": 0.4424,
+      "step": 645
+    },
+    {
+      "epoch": 0.05300077942322681,
+      "grad_norm": 0.705464440165486,
+      "learning_rate": 9.986263778093373e-06,
+      "loss": 0.4829,
+      "step": 646
+    },
+    {
+      "epoch": 0.05308282397341756,
+      "grad_norm": 0.8118832790191072,
+      "learning_rate": 9.986165179603538e-06,
+      "loss": 0.4646,
+      "step": 647
+    },
+    {
+      "epoch": 0.05316486852360832,
+      "grad_norm": 0.6981390745350371,
+      "learning_rate": 9.98606622899873e-06,
+      "loss": 0.4385,
+      "step": 648
+    },
+    {
+      "epoch": 0.05324691307379907,
+      "grad_norm": 0.764402786484223,
+      "learning_rate": 9.985966926285935e-06,
+      "loss": 0.4294,
+      "step": 649
+    },
+    {
+      "epoch": 0.05332895762398983,
+      "grad_norm": 0.7151047594318614,
+      "learning_rate": 9.985867271472166e-06,
+      "loss": 0.4996,
+      "step": 650
+    },
+    {
+      "epoch": 0.05341100217418058,
+      "grad_norm": 0.6655314453332335,
+      "learning_rate": 9.985767264564461e-06,
+      "loss": 0.4161,
+      "step": 651
+    },
+    {
+      "epoch": 0.053493046724371336,
+      "grad_norm": 0.6590476819004433,
+      "learning_rate": 9.985666905569882e-06,
+      "loss": 0.4506,
+      "step": 652
+    },
+    {
+      "epoch": 0.05357509127456209,
+      "grad_norm": 0.7214419447852418,
+      "learning_rate": 9.985566194495516e-06,
+      "loss": 0.4394,
+      "step": 653
+    },
+    {
+      "epoch": 0.053657135824752844,
+      "grad_norm": 0.653003124772566,
+      "learning_rate": 9.985465131348475e-06,
+      "loss": 0.4294,
+      "step": 654
+    },
+    {
+      "epoch": 0.053739180374943595,
+      "grad_norm": 0.6576654232270568,
+      "learning_rate": 9.985363716135896e-06,
+      "loss": 0.4116,
+      "step": 655
+    },
+    {
+      "epoch": 0.053821224925134345,
+      "grad_norm": 0.6868513334775568,
+      "learning_rate": 9.985261948864941e-06,
+      "loss": 0.4514,
+      "step": 656
+    },
+    {
+      "epoch": 0.0539032694753251,
+      "grad_norm": 0.7034376887946833,
+      "learning_rate": 9.985159829542798e-06,
+      "loss": 0.4398,
+      "step": 657
+    },
+    {
+      "epoch": 0.05398531402551585,
+      "grad_norm": 0.7063276033071738,
+      "learning_rate": 9.985057358176675e-06,
+      "loss": 0.3851,
+      "step": 658
+    },
+    {
+      "epoch": 0.05406735857570661,
+      "grad_norm": 0.6738112518845624,
+      "learning_rate": 9.984954534773812e-06,
+      "loss": 0.4661,
+      "step": 659
+    },
+    {
+      "epoch": 0.05414940312589736,
+      "grad_norm": 0.7561419868890678,
+      "learning_rate": 9.984851359341469e-06,
+      "loss": 0.4591,
+      "step": 660
+    },
+    {
+      "epoch": 0.05423144767608812,
+      "grad_norm": 0.6850450369343807,
+      "learning_rate": 9.98474783188693e-06,
+      "loss": 0.4568,
+      "step": 661
+    },
+    {
+      "epoch": 0.05431349222627887,
+      "grad_norm": 0.7081006507074291,
+      "learning_rate": 9.98464395241751e-06,
+      "loss": 0.4772,
+      "step": 662
+    },
+    {
+      "epoch": 0.054395536776469626,
+      "grad_norm": 0.6898729857188778,
+      "learning_rate": 9.98453972094054e-06,
+      "loss": 0.4562,
+      "step": 663
+    },
+    {
+      "epoch": 0.054477581326660376,
+      "grad_norm": 0.6316791515209615,
+      "learning_rate": 9.984435137463386e-06,
+      "loss": 0.4227,
+      "step": 664
+    },
+    {
+      "epoch": 0.05455962587685113,
+      "grad_norm": 0.6355303085714438,
+      "learning_rate": 9.984330201993428e-06,
+      "loss": 0.4053,
+      "step": 665
+    },
+    {
+      "epoch": 0.054641670427041884,
+      "grad_norm": 0.6740317890759678,
+      "learning_rate": 9.98422491453808e-06,
+      "loss": 0.4353,
+      "step": 666
+    },
+    {
+      "epoch": 0.054723714977232635,
+      "grad_norm": 0.6656941607213188,
+      "learning_rate": 9.984119275104778e-06,
+      "loss": 0.4351,
+      "step": 667
+    },
+    {
+      "epoch": 0.05480575952742339,
+      "grad_norm": 0.7524910923546507,
+      "learning_rate": 9.984013283700979e-06,
+      "loss": 0.4512,
+      "step": 668
+    },
+    {
+      "epoch": 0.05488780407761414,
+      "grad_norm": 0.6727451923963934,
+      "learning_rate": 9.98390694033417e-06,
+      "loss": 0.4145,
+      "step": 669
+    },
+    {
+      "epoch": 0.0549698486278049,
+      "grad_norm": 0.7223381738399713,
+      "learning_rate": 9.983800245011858e-06,
+      "loss": 0.4755,
+      "step": 670
+    },
+    {
+      "epoch": 0.05505189317799565,
+      "grad_norm": 0.7358819047761049,
+      "learning_rate": 9.983693197741581e-06,
+      "loss": 0.4573,
+      "step": 671
+    },
+    {
+      "epoch": 0.05513393772818641,
+      "grad_norm": 0.7191257984876378,
+      "learning_rate": 9.983585798530898e-06,
+      "loss": 0.4016,
+      "step": 672
+    },
+    {
+      "epoch": 0.05521598227837716,
+      "grad_norm": 0.6577727538401741,
+      "learning_rate": 9.983478047387392e-06,
+      "loss": 0.4873,
+      "step": 673
+    },
+    {
+      "epoch": 0.055298026828567916,
+      "grad_norm": 0.6772138522715679,
+      "learning_rate": 9.983369944318673e-06,
+      "loss": 0.4393,
+      "step": 674
+    },
+    {
+      "epoch": 0.055380071378758666,
+      "grad_norm": 0.6852212196493609,
+      "learning_rate": 9.983261489332375e-06,
+      "loss": 0.4327,
+      "step": 675
+    },
+    {
+      "epoch": 0.05546211592894942,
+      "grad_norm": 0.7127118245753811,
+      "learning_rate": 9.983152682436156e-06,
+      "loss": 0.476,
+      "step": 676
+    },
+    {
+      "epoch": 0.055544160479140174,
+      "grad_norm": 0.7967448536590878,
+      "learning_rate": 9.983043523637703e-06,
+      "loss": 0.4611,
+      "step": 677
+    },
+    {
+      "epoch": 0.055626205029330925,
+      "grad_norm": 0.773059351793155,
+      "learning_rate": 9.98293401294472e-06,
+      "loss": 0.4768,
+      "step": 678
+    },
+    {
+      "epoch": 0.05570824957952168,
+      "grad_norm": 0.6807979706816126,
+      "learning_rate": 9.982824150364943e-06,
+      "loss": 0.4544,
+      "step": 679
+    },
+    {
+      "epoch": 0.05579029412971243,
+      "grad_norm": 0.6893915945536586,
+      "learning_rate": 9.982713935906132e-06,
+      "loss": 0.4793,
+      "step": 680
+    },
+    {
+      "epoch": 0.05587233867990319,
+      "grad_norm": 0.6392432448165112,
+      "learning_rate": 9.982603369576065e-06,
+      "loss": 0.4464,
+      "step": 681
+    },
+    {
+      "epoch": 0.05595438323009394,
+      "grad_norm": 0.6625419182332756,
+      "learning_rate": 9.982492451382554e-06,
+      "loss": 0.4139,
+      "step": 682
+    },
+    {
+      "epoch": 0.0560364277802847,
+      "grad_norm": 0.6811076581172595,
+      "learning_rate": 9.982381181333433e-06,
+      "loss": 0.4636,
+      "step": 683
+    },
+    {
+      "epoch": 0.05611847233047545,
+      "grad_norm": 0.644709537808952,
+      "learning_rate": 9.982269559436557e-06,
+      "loss": 0.3796,
+      "step": 684
+    },
+    {
+      "epoch": 0.0562005168806662,
+      "grad_norm": 0.6925131944653402,
+      "learning_rate": 9.982157585699808e-06,
+      "loss": 0.3892,
+      "step": 685
+    },
+    {
+      "epoch": 0.056282561430856956,
+      "grad_norm": 0.6267594752148631,
+      "learning_rate": 9.982045260131096e-06,
+      "loss": 0.3965,
+      "step": 686
+    },
+    {
+      "epoch": 0.056364605981047707,
+      "grad_norm": 0.7083275003275775,
+      "learning_rate": 9.981932582738352e-06,
+      "loss": 0.4728,
+      "step": 687
+    },
+    {
+      "epoch": 0.056446650531238464,
+      "grad_norm": 0.6775895818256653,
+      "learning_rate": 9.981819553529533e-06,
+      "loss": 0.4177,
+      "step": 688
+    },
+    {
+      "epoch": 0.056528695081429214,
+      "grad_norm": 0.7095397771413141,
+      "learning_rate": 9.98170617251262e-06,
+      "loss": 0.4701,
+      "step": 689
+    },
+    {
+      "epoch": 0.05661073963161997,
+      "grad_norm": 0.7026091176344283,
+      "learning_rate": 9.981592439695622e-06,
+      "loss": 0.4434,
+      "step": 690
+    },
+    {
+      "epoch": 0.05669278418181072,
+      "grad_norm": 0.6823833914164218,
+      "learning_rate": 9.98147835508657e-06,
+      "loss": 0.444,
+      "step": 691
+    },
+    {
+      "epoch": 0.05677482873200148,
+      "grad_norm": 0.6316488009161602,
+      "learning_rate": 9.981363918693518e-06,
+      "loss": 0.4396,
+      "step": 692
+    },
+    {
+      "epoch": 0.05685687328219223,
+      "grad_norm": 0.800062822143366,
+      "learning_rate": 9.981249130524551e-06,
+      "loss": 0.4641,
+      "step": 693
+    },
+    {
+      "epoch": 0.05693891783238298,
+      "grad_norm": 0.6631359653992758,
+      "learning_rate": 9.981133990587773e-06,
+      "loss": 0.4618,
+      "step": 694
+    },
+    {
+      "epoch": 0.05702096238257374,
+      "grad_norm": 0.702374822599577,
+      "learning_rate": 9.981018498891314e-06,
+      "loss": 0.448,
+      "step": 695
+    },
+    {
+      "epoch": 0.05710300693276449,
+      "grad_norm": 0.6494507164843467,
+      "learning_rate": 9.980902655443332e-06,
+      "loss": 0.3994,
+      "step": 696
+    },
+    {
+      "epoch": 0.057185051482955246,
+      "grad_norm": 0.6749110269107862,
+      "learning_rate": 9.980786460252007e-06,
+      "loss": 0.471,
+      "step": 697
+    },
+    {
+      "epoch": 0.057267096033145996,
+      "grad_norm": 0.7157541993889127,
+      "learning_rate": 9.980669913325545e-06,
+      "loss": 0.4364,
+      "step": 698
+    },
+    {
+      "epoch": 0.057349140583336754,
+      "grad_norm": 0.627808033699411,
+      "learning_rate": 9.980553014672175e-06,
+      "loss": 0.4148,
+      "step": 699
+    },
+    {
+      "epoch": 0.057431185133527504,
+      "grad_norm": 0.6538797207534921,
+      "learning_rate": 9.980435764300154e-06,
+      "loss": 0.4673,
+      "step": 700
+    },
+    {
+      "epoch": 0.05751322968371826,
+      "grad_norm": 0.6957043648521234,
+      "learning_rate": 9.980318162217759e-06,
+      "loss": 0.4464,
+      "step": 701
+    },
+    {
+      "epoch": 0.05759527423390901,
+      "grad_norm": 0.6578536128527104,
+      "learning_rate": 9.9802002084333e-06,
+      "loss": 0.4401,
+      "step": 702
+    },
+    {
+      "epoch": 0.05767731878409977,
+      "grad_norm": 0.7163350057026545,
+      "learning_rate": 9.980081902955102e-06,
+      "loss": 0.4278,
+      "step": 703
+    },
+    {
+      "epoch": 0.05775936333429052,
+      "grad_norm": 0.7027717164373313,
+      "learning_rate": 9.979963245791519e-06,
+      "loss": 0.412,
+      "step": 704
+    },
+    {
+      "epoch": 0.05784140788448127,
+      "grad_norm": 0.6812319484730992,
+      "learning_rate": 9.979844236950936e-06,
+      "loss": 0.4644,
+      "step": 705
+    },
+    {
+      "epoch": 0.05792345243467203,
+      "grad_norm": 0.788791247577929,
+      "learning_rate": 9.979724876441751e-06,
+      "loss": 0.4413,
+      "step": 706
+    },
+    {
+      "epoch": 0.05800549698486278,
+      "grad_norm": 0.7049188175294242,
+      "learning_rate": 9.979605164272396e-06,
+      "loss": 0.4693,
+      "step": 707
+    },
+    {
+      "epoch": 0.058087541535053536,
+      "grad_norm": 0.7013926995201676,
+      "learning_rate": 9.979485100451327e-06,
+      "loss": 0.4888,
+      "step": 708
+    },
+    {
+      "epoch": 0.058169586085244286,
+      "grad_norm": 0.6601199182502906,
+      "learning_rate": 9.979364684987018e-06,
+      "loss": 0.4537,
+      "step": 709
+    },
+    {
+      "epoch": 0.058251630635435044,
+      "grad_norm": 0.7274908349236442,
+      "learning_rate": 9.979243917887977e-06,
+      "loss": 0.4211,
+      "step": 710
+    },
+    {
+      "epoch": 0.058333675185625794,
+      "grad_norm": 0.7123192772247481,
+      "learning_rate": 9.979122799162729e-06,
+      "loss": 0.4437,
+      "step": 711
+    },
+    {
+      "epoch": 0.05841571973581655,
+      "grad_norm": 0.6746404604710475,
+      "learning_rate": 9.979001328819828e-06,
+      "loss": 0.4191,
+      "step": 712
+    },
+    {
+      "epoch": 0.0584977642860073,
+      "grad_norm": 0.710890037481162,
+      "learning_rate": 9.978879506867854e-06,
+      "loss": 0.4431,
+      "step": 713
+    },
+    {
+      "epoch": 0.05857980883619805,
+      "grad_norm": 0.6666099100651415,
+      "learning_rate": 9.978757333315406e-06,
+      "loss": 0.4204,
+      "step": 714
+    },
+    {
+      "epoch": 0.05866185338638881,
+      "grad_norm": 0.697978952189993,
+      "learning_rate": 9.978634808171116e-06,
+      "loss": 0.4428,
+      "step": 715
+    },
+    {
+      "epoch": 0.05874389793657956,
+      "grad_norm": 0.6509701629221485,
+      "learning_rate": 9.978511931443636e-06,
+      "loss": 0.4129,
+      "step": 716
+    },
+    {
+      "epoch": 0.05882594248677032,
+      "grad_norm": 0.7071397847671119,
+      "learning_rate": 9.97838870314164e-06,
+      "loss": 0.4998,
+      "step": 717
+    },
+    {
+      "epoch": 0.05890798703696107,
+      "grad_norm": 0.7112688537716723,
+      "learning_rate": 9.978265123273831e-06,
+      "loss": 0.4489,
+      "step": 718
+    },
+    {
+      "epoch": 0.058990031587151825,
+      "grad_norm": 0.6841852797025005,
+      "learning_rate": 9.97814119184894e-06,
+      "loss": 0.5004,
+      "step": 719
+    },
+    {
+      "epoch": 0.059072076137342576,
+      "grad_norm": 0.7096267098111431,
+      "learning_rate": 9.978016908875714e-06,
+      "loss": 0.4212,
+      "step": 720
+    },
+    {
+      "epoch": 0.05915412068753333,
+      "grad_norm": 0.7096661506217141,
+      "learning_rate": 9.977892274362933e-06,
+      "loss": 0.4753,
+      "step": 721
+    },
+    {
+      "epoch": 0.059236165237724084,
+      "grad_norm": 0.6769399770882867,
+      "learning_rate": 9.977767288319396e-06,
+      "loss": 0.416,
+      "step": 722
+    },
+    {
+      "epoch": 0.05931820978791484,
+      "grad_norm": 0.6800702883065445,
+      "learning_rate": 9.97764195075393e-06,
+      "loss": 0.4574,
+      "step": 723
+    },
+    {
+      "epoch": 0.05940025433810559,
+      "grad_norm": 0.6914055545089807,
+      "learning_rate": 9.977516261675389e-06,
+      "loss": 0.4208,
+      "step": 724
+    },
+    {
+      "epoch": 0.05948229888829634,
+      "grad_norm": 0.6876058207583066,
+      "learning_rate": 9.977390221092645e-06,
+      "loss": 0.4323,
+      "step": 725
+    },
+    {
+      "epoch": 0.0595643434384871,
+      "grad_norm": 0.6873223520748643,
+      "learning_rate": 9.977263829014599e-06,
+      "loss": 0.4923,
+      "step": 726
+    },
+    {
+      "epoch": 0.05964638798867785,
+      "grad_norm": 0.7408039917168159,
+      "learning_rate": 9.97713708545018e-06,
+      "loss": 0.4875,
+      "step": 727
+    },
+    {
+      "epoch": 0.05972843253886861,
+      "grad_norm": 0.7329918634377439,
+      "learning_rate": 9.977009990408336e-06,
+      "loss": 0.4908,
+      "step": 728
+    },
+    {
+      "epoch": 0.05981047708905936,
+      "grad_norm": 0.698749954429685,
+      "learning_rate": 9.976882543898043e-06,
+      "loss": 0.4563,
+      "step": 729
+    },
+    {
+      "epoch": 0.059892521639250115,
+      "grad_norm": 1.365501991450519,
+      "learning_rate": 9.976754745928301e-06,
+      "loss": 0.4155,
+      "step": 730
+    },
+    {
+      "epoch": 0.059974566189440866,
+      "grad_norm": 0.7138149946173228,
+      "learning_rate": 9.976626596508133e-06,
+      "loss": 0.4124,
+      "step": 731
+    },
+    {
+      "epoch": 0.06005661073963162,
+      "grad_norm": 0.689960220384126,
+      "learning_rate": 9.976498095646592e-06,
+      "loss": 0.4202,
+      "step": 732
+    },
+    {
+      "epoch": 0.060138655289822374,
+      "grad_norm": 0.6883236046015051,
+      "learning_rate": 9.97636924335275e-06,
+      "loss": 0.4494,
+      "step": 733
+    },
+    {
+      "epoch": 0.060220699840013124,
+      "grad_norm": 0.7124473723343269,
+      "learning_rate": 9.976240039635708e-06,
+      "loss": 0.4205,
+      "step": 734
+    },
+    {
+      "epoch": 0.06030274439020388,
+      "grad_norm": 0.7004946345465758,
+      "learning_rate": 9.976110484504587e-06,
+      "loss": 0.4706,
+      "step": 735
+    },
+    {
+      "epoch": 0.06038478894039463,
+      "grad_norm": 0.7411236802469312,
+      "learning_rate": 9.97598057796854e-06,
+      "loss": 0.4368,
+      "step": 736
+    },
+    {
+      "epoch": 0.06046683349058539,
+      "grad_norm": 0.7287422583409956,
+      "learning_rate": 9.975850320036739e-06,
+      "loss": 0.4196,
+      "step": 737
+    },
+    {
+      "epoch": 0.06054887804077614,
+      "grad_norm": 0.763393132690018,
+      "learning_rate": 9.975719710718384e-06,
+      "loss": 0.3967,
+      "step": 738
+    },
+    {
+      "epoch": 0.0606309225909669,
+      "grad_norm": 0.7430463384461315,
+      "learning_rate": 9.975588750022696e-06,
+      "loss": 0.4315,
+      "step": 739
+    },
+    {
+      "epoch": 0.06071296714115765,
+      "grad_norm": 0.6601706037088947,
+      "learning_rate": 9.975457437958924e-06,
+      "loss": 0.4663,
+      "step": 740
+    },
+    {
+      "epoch": 0.060795011691348405,
+      "grad_norm": 0.7148333770866786,
+      "learning_rate": 9.975325774536342e-06,
+      "loss": 0.4436,
+      "step": 741
+    },
+    {
+      "epoch": 0.060877056241539156,
+      "grad_norm": 1.206851605937381,
+      "learning_rate": 9.975193759764245e-06,
+      "loss": 0.4568,
+      "step": 742
+    },
+    {
+      "epoch": 0.060959100791729906,
+      "grad_norm": 0.7106445755470513,
+      "learning_rate": 9.975061393651961e-06,
+      "loss": 0.411,
+      "step": 743
+    },
+    {
+      "epoch": 0.06104114534192066,
+      "grad_norm": 0.7015856967347889,
+      "learning_rate": 9.974928676208835e-06,
+      "loss": 0.4777,
+      "step": 744
+    },
+    {
+      "epoch": 0.061123189892111414,
+      "grad_norm": 0.7359090101404293,
+      "learning_rate": 9.974795607444236e-06,
+      "loss": 0.3988,
+      "step": 745
+    },
+    {
+      "epoch": 0.06120523444230217,
+      "grad_norm": 0.6918746545964252,
+      "learning_rate": 9.974662187367565e-06,
+      "loss": 0.4337,
+      "step": 746
+    },
+    {
+      "epoch": 0.06128727899249292,
+      "grad_norm": 0.6899487522747162,
+      "learning_rate": 9.974528415988242e-06,
+      "loss": 0.4263,
+      "step": 747
+    },
+    {
+      "epoch": 0.06136932354268368,
+      "grad_norm": 0.8229182671995571,
+      "learning_rate": 9.974394293315715e-06,
+      "loss": 0.4624,
+      "step": 748
+    },
+    {
+      "epoch": 0.06145136809287443,
+      "grad_norm": 0.7915810023953279,
+      "learning_rate": 9.974259819359456e-06,
+      "loss": 0.4167,
+      "step": 749
+    },
+    {
+      "epoch": 0.06153341264306519,
+      "grad_norm": 0.7307976403936562,
+      "learning_rate": 9.974124994128959e-06,
+      "loss": 0.4427,
+      "step": 750
+    },
+    {
+      "epoch": 0.06161545719325594,
+      "grad_norm": 0.8840178619546158,
+      "learning_rate": 9.973989817633746e-06,
+      "loss": 0.4464,
+      "step": 751
+    },
+    {
+      "epoch": 0.061697501743446695,
+      "grad_norm": 0.7681184741000145,
+      "learning_rate": 9.973854289883364e-06,
+      "loss": 0.423,
+      "step": 752
+    },
+    {
+      "epoch": 0.061779546293637445,
+      "grad_norm": 0.7844139231878867,
+      "learning_rate": 9.973718410887383e-06,
+      "loss": 0.4433,
+      "step": 753
+    },
+    {
+      "epoch": 0.061861590843828196,
+      "grad_norm": 1.2242308802650388,
+      "learning_rate": 9.9735821806554e-06,
+      "loss": 0.8212,
+      "step": 754
+    },
+    {
+      "epoch": 0.06194363539401895,
+      "grad_norm": 0.6870410981631561,
+      "learning_rate": 9.973445599197034e-06,
+      "loss": 0.4426,
+      "step": 755
+    },
+    {
+      "epoch": 0.062025679944209704,
+      "grad_norm": 0.7081273307998821,
+      "learning_rate": 9.973308666521929e-06,
+      "loss": 0.4414,
+      "step": 756
+    },
+    {
+      "epoch": 0.06210772449440046,
+      "grad_norm": 1.3391986335129844,
+      "learning_rate": 9.973171382639757e-06,
+      "loss": 0.4665,
+      "step": 757
+    },
+    {
+      "epoch": 0.06218976904459121,
+      "grad_norm": 0.6881160323772392,
+      "learning_rate": 9.97303374756021e-06,
+      "loss": 0.4076,
+      "step": 758
+    },
+    {
+      "epoch": 0.06227181359478197,
+      "grad_norm": 1.0908976755791904,
+      "learning_rate": 9.972895761293012e-06,
+      "loss": 0.4302,
+      "step": 759
+    },
+    {
+      "epoch": 0.06235385814497272,
+      "grad_norm": 1.0851294031523933,
+      "learning_rate": 9.972757423847904e-06,
+      "loss": 0.4751,
+      "step": 760
+    },
+    {
+      "epoch": 0.06243590269516348,
+      "grad_norm": 0.9666763642516082,
+      "learning_rate": 9.972618735234655e-06,
+      "loss": 0.3922,
+      "step": 761
+    },
+    {
+      "epoch": 0.06251794724535423,
+      "grad_norm": 0.7221454177423865,
+      "learning_rate": 9.972479695463061e-06,
+      "loss": 0.466,
+      "step": 762
+    },
+    {
+      "epoch": 0.06259999179554498,
+      "grad_norm": 0.7506869139308555,
+      "learning_rate": 9.972340304542941e-06,
+      "loss": 0.4435,
+      "step": 763
+    },
+    {
+      "epoch": 0.06268203634573573,
+      "grad_norm": 0.704660536991921,
+      "learning_rate": 9.972200562484135e-06,
+      "loss": 0.4597,
+      "step": 764
+    },
+    {
+      "epoch": 0.06276408089592649,
+      "grad_norm": 1.2498355443195492,
+      "learning_rate": 9.972060469296517e-06,
+      "loss": 0.4261,
+      "step": 765
+    },
+    {
+      "epoch": 0.06284612544611724,
+      "grad_norm": 0.9689363988185616,
+      "learning_rate": 9.971920024989974e-06,
+      "loss": 0.4334,
+      "step": 766
+    },
+    {
+      "epoch": 0.062928169996308,
+      "grad_norm": 0.7085182329360842,
+      "learning_rate": 9.971779229574427e-06,
+      "loss": 0.4686,
+      "step": 767
+    },
+    {
+      "epoch": 0.06301021454649874,
+      "grad_norm": 0.7602390024304373,
+      "learning_rate": 9.97163808305982e-06,
+      "loss": 0.4762,
+      "step": 768
+    },
+    {
+      "epoch": 0.0630922590966895,
+      "grad_norm": 0.7386600804349428,
+      "learning_rate": 9.971496585456119e-06,
+      "loss": 0.4486,
+      "step": 769
+    },
+    {
+      "epoch": 0.06317430364688026,
+      "grad_norm": 0.9861261741806931,
+      "learning_rate": 9.971354736773315e-06,
+      "loss": 0.4035,
+      "step": 770
+    },
+    {
+      "epoch": 0.06325634819707102,
+      "grad_norm": 0.8760235492009005,
+      "learning_rate": 9.971212537021428e-06,
+      "loss": 0.4162,
+      "step": 771
+    },
+    {
+      "epoch": 0.06333839274726176,
+      "grad_norm": 0.7876566257276348,
+      "learning_rate": 9.971069986210498e-06,
+      "loss": 0.4445,
+      "step": 772
+    },
+    {
+      "epoch": 0.06342043729745252,
+      "grad_norm": 0.7655248219598937,
+      "learning_rate": 9.970927084350592e-06,
+      "loss": 0.4428,
+      "step": 773
+    },
+    {
+      "epoch": 0.06350248184764327,
+      "grad_norm": 0.7809007178992573,
+      "learning_rate": 9.970783831451803e-06,
+      "loss": 0.4379,
+      "step": 774
+    },
+    {
+      "epoch": 0.06358452639783402,
+      "grad_norm": 1.5540063561559874,
+      "learning_rate": 9.970640227524244e-06,
+      "loss": 0.4298,
+      "step": 775
+    },
+    {
+      "epoch": 0.06366657094802478,
+      "grad_norm": 0.7521697594060276,
+      "learning_rate": 9.970496272578058e-06,
+      "loss": 0.4446,
+      "step": 776
+    },
+    {
+      "epoch": 0.06374861549821553,
+      "grad_norm": 0.7875072055248622,
+      "learning_rate": 9.970351966623413e-06,
+      "loss": 0.4388,
+      "step": 777
+    },
+    {
+      "epoch": 0.06383066004840629,
+      "grad_norm": 0.8167199238283892,
+      "learning_rate": 9.970207309670496e-06,
+      "loss": 0.4379,
+      "step": 778
+    },
+    {
+      "epoch": 0.06391270459859703,
+      "grad_norm": 0.6097628144553788,
+      "learning_rate": 9.970062301729525e-06,
+      "loss": 0.4094,
+      "step": 779
+    },
+    {
+      "epoch": 0.06399474914878779,
+      "grad_norm": 0.8465271046831818,
+      "learning_rate": 9.969916942810738e-06,
+      "loss": 0.4334,
+      "step": 780
+    },
+    {
+      "epoch": 0.06407679369897855,
+      "grad_norm": 1.122569986165872,
+      "learning_rate": 9.969771232924404e-06,
+      "loss": 0.4444,
+      "step": 781
+    },
+    {
+      "epoch": 0.06415883824916929,
+      "grad_norm": 0.712845475178253,
+      "learning_rate": 9.969625172080807e-06,
+      "loss": 0.4448,
+      "step": 782
+    },
+    {
+      "epoch": 0.06424088279936005,
+      "grad_norm": 0.8946826208160347,
+      "learning_rate": 9.969478760290265e-06,
+      "loss": 0.4189,
+      "step": 783
+    },
+    {
+      "epoch": 0.0643229273495508,
+      "grad_norm": 0.839586905995174,
+      "learning_rate": 9.969331997563119e-06,
+      "loss": 0.4371,
+      "step": 784
+    },
+    {
+      "epoch": 0.06440497189974156,
+      "grad_norm": 0.7322923888964609,
+      "learning_rate": 9.96918488390973e-06,
+      "loss": 0.4427,
+      "step": 785
+    },
+    {
+      "epoch": 0.06448701644993231,
+      "grad_norm": 0.9671674751601352,
+      "learning_rate": 9.969037419340488e-06,
+      "loss": 0.435,
+      "step": 786
+    },
+    {
+      "epoch": 0.06456906100012307,
+      "grad_norm": 0.8274759991754652,
+      "learning_rate": 9.968889603865807e-06,
+      "loss": 0.4875,
+      "step": 787
+    },
+    {
+      "epoch": 0.06465110555031382,
+      "grad_norm": 0.700627593526248,
+      "learning_rate": 9.968741437496126e-06,
+      "loss": 0.4168,
+      "step": 788
+    },
+    {
+      "epoch": 0.06473315010050458,
+      "grad_norm": 1.0800517278059598,
+      "learning_rate": 9.968592920241906e-06,
+      "loss": 0.3956,
+      "step": 789
+    },
+    {
+      "epoch": 0.06481519465069532,
+      "grad_norm": 0.7268277124759637,
+      "learning_rate": 9.968444052113637e-06,
+      "loss": 0.4102,
+      "step": 790
+    },
+    {
+      "epoch": 0.06489723920088608,
+      "grad_norm": 0.8215679318681983,
+      "learning_rate": 9.968294833121831e-06,
+      "loss": 0.4927,
+      "step": 791
+    },
+    {
+      "epoch": 0.06497928375107684,
+      "grad_norm": 0.7982039392106447,
+      "learning_rate": 9.968145263277029e-06,
+      "loss": 0.4404,
+      "step": 792
+    },
+    {
+      "epoch": 0.06506132830126758,
+      "grad_norm": 0.7445289226029976,
+      "learning_rate": 9.967995342589787e-06,
+      "loss": 0.418,
+      "step": 793
+    },
+    {
+      "epoch": 0.06514337285145834,
+      "grad_norm": 1.2928868651018985,
+      "learning_rate": 9.967845071070697e-06,
+      "loss": 0.4253,
+      "step": 794
+    },
+    {
+      "epoch": 0.0652254174016491,
+      "grad_norm": 0.8193165161970162,
+      "learning_rate": 9.967694448730369e-06,
+      "loss": 0.4508,
+      "step": 795
+    },
+    {
+      "epoch": 0.06530746195183985,
+      "grad_norm": 1.0708209673453732,
+      "learning_rate": 9.967543475579442e-06,
+      "loss": 0.4454,
+      "step": 796
+    },
+    {
+      "epoch": 0.0653895065020306,
+      "grad_norm": 0.7758366615047031,
+      "learning_rate": 9.967392151628574e-06,
+      "loss": 0.4569,
+      "step": 797
+    },
+    {
+      "epoch": 0.06547155105222136,
+      "grad_norm": 1.5408654914684596,
+      "learning_rate": 9.967240476888453e-06,
+      "loss": 0.4478,
+      "step": 798
+    },
+    {
+      "epoch": 0.06555359560241211,
+      "grad_norm": 0.8709431320997821,
+      "learning_rate": 9.96708845136979e-06,
+      "loss": 0.4511,
+      "step": 799
+    },
+    {
+      "epoch": 0.06563564015260287,
+      "grad_norm": 0.7224564717546419,
+      "learning_rate": 9.966936075083323e-06,
+      "loss": 0.4253,
+      "step": 800
+    },
+    {
+      "epoch": 0.06571768470279361,
+      "grad_norm": 0.9614689114125544,
+      "learning_rate": 9.966783348039808e-06,
+      "loss": 0.4567,
+      "step": 801
+    },
+    {
+      "epoch": 0.06579972925298437,
+      "grad_norm": 0.7498416461943958,
+      "learning_rate": 9.966630270250033e-06,
+      "loss": 0.4244,
+      "step": 802
+    },
+    {
+      "epoch": 0.06588177380317513,
+      "grad_norm": 2.8828206407894488,
+      "learning_rate": 9.966476841724808e-06,
+      "loss": 0.4164,
+      "step": 803
+    },
+    {
+      "epoch": 0.06596381835336587,
+      "grad_norm": 1.5114530211697998,
+      "learning_rate": 9.96632306247497e-06,
+      "loss": 0.4477,
+      "step": 804
+    },
+    {
+      "epoch": 0.06604586290355663,
+      "grad_norm": 1.4176317734140607,
+      "learning_rate": 9.966168932511372e-06,
+      "loss": 0.4696,
+      "step": 805
+    },
+    {
+      "epoch": 0.06612790745374739,
+      "grad_norm": 1.3902368893682386,
+      "learning_rate": 9.966014451844905e-06,
+      "loss": 0.4472,
+      "step": 806
+    },
+    {
+      "epoch": 0.06620995200393814,
+      "grad_norm": 1.2815829654197612,
+      "learning_rate": 9.965859620486476e-06,
+      "loss": 0.4354,
+      "step": 807
+    },
+    {
+      "epoch": 0.06629199655412889,
+      "grad_norm": 1.0073065909922674,
+      "learning_rate": 9.96570443844702e-06,
+      "loss": 0.4239,
+      "step": 808
+    },
+    {
+      "epoch": 0.06637404110431964,
+      "grad_norm": 2.2286609973890084,
+      "learning_rate": 9.965548905737492e-06,
+      "loss": 0.4482,
+      "step": 809
+    },
+    {
+      "epoch": 0.0664560856545104,
+      "grad_norm": 1.0475889602091626,
+      "learning_rate": 9.965393022368878e-06,
+      "loss": 0.4165,
+      "step": 810
+    },
+    {
+      "epoch": 0.06653813020470115,
+      "grad_norm": 3.526510655375477,
+      "learning_rate": 9.96523678835219e-06,
+      "loss": 0.4315,
+      "step": 811
+    },
+    {
+      "epoch": 0.0666201747548919,
+      "grad_norm": 1.1650784318387941,
+      "learning_rate": 9.965080203698453e-06,
+      "loss": 0.4419,
+      "step": 812
+    },
+    {
+      "epoch": 0.06670221930508266,
+      "grad_norm": 1.2897660053375555,
+      "learning_rate": 9.964923268418732e-06,
+      "loss": 0.4243,
+      "step": 813
+    },
+    {
+      "epoch": 0.06678426385527342,
+      "grad_norm": 1.3471857865843913,
+      "learning_rate": 9.964765982524105e-06,
+      "loss": 0.3793,
+      "step": 814
+    },
+    {
+      "epoch": 0.06686630840546416,
+      "grad_norm": 0.7971019248886824,
+      "learning_rate": 9.964608346025681e-06,
+      "loss": 0.4237,
+      "step": 815
+    },
+    {
+      "epoch": 0.06694835295565492,
+      "grad_norm": 0.8009107995151056,
+      "learning_rate": 9.964450358934593e-06,
+      "loss": 0.4356,
+      "step": 816
+    },
+    {
+      "epoch": 0.06703039750584568,
+      "grad_norm": 2.554239154437295,
+      "learning_rate": 9.964292021261996e-06,
+      "loss": 0.4389,
+      "step": 817
+    },
+    {
+      "epoch": 0.06711244205603643,
+      "grad_norm": 0.6872421786623378,
+      "learning_rate": 9.96413333301907e-06,
+      "loss": 0.4122,
+      "step": 818
+    },
+    {
+      "epoch": 0.06719448660622718,
+      "grad_norm": 0.8203768788619519,
+      "learning_rate": 9.963974294217027e-06,
+      "loss": 0.3744,
+      "step": 819
+    },
+    {
+      "epoch": 0.06727653115641793,
+      "grad_norm": 0.7307300269137603,
+      "learning_rate": 9.963814904867092e-06,
+      "loss": 0.4436,
+      "step": 820
+    },
+    {
+      "epoch": 0.06735857570660869,
+      "grad_norm": 0.7441668812092661,
+      "learning_rate": 9.963655164980527e-06,
+      "loss": 0.4175,
+      "step": 821
+    },
+    {
+      "epoch": 0.06744062025679944,
+      "grad_norm": 0.7816769697993257,
+      "learning_rate": 9.963495074568605e-06,
+      "loss": 0.4084,
+      "step": 822
+    },
+    {
+      "epoch": 0.06752266480699019,
+      "grad_norm": 1.2765553223043258,
+      "learning_rate": 9.963334633642637e-06,
+      "loss": 0.3974,
+      "step": 823
+    },
+    {
+      "epoch": 0.06760470935718095,
+      "grad_norm": 0.8527544314673186,
+      "learning_rate": 9.96317384221395e-06,
+      "loss": 0.4698,
+      "step": 824
+    },
+    {
+      "epoch": 0.06768675390737171,
+      "grad_norm": 1.100337221522185,
+      "learning_rate": 9.963012700293902e-06,
+      "loss": 0.4739,
+      "step": 825
+    },
+    {
+      "epoch": 0.06776879845756245,
+      "grad_norm": 1.0121221309547375,
+      "learning_rate": 9.962851207893871e-06,
+      "loss": 0.4414,
+      "step": 826
+    },
+    {
+      "epoch": 0.06785084300775321,
+      "grad_norm": 0.8276096036747058,
+      "learning_rate": 9.962689365025259e-06,
+      "loss": 0.5147,
+      "step": 827
+    },
+    {
+      "epoch": 0.06793288755794397,
+      "grad_norm": 1.1336871801781099,
+      "learning_rate": 9.962527171699498e-06,
+      "loss": 0.4067,
+      "step": 828
+    },
+    {
+      "epoch": 0.06801493210813472,
+      "grad_norm": 0.7388575057541553,
+      "learning_rate": 9.962364627928043e-06,
+      "loss": 0.4279,
+      "step": 829
+    },
+    {
+      "epoch": 0.06809697665832547,
+      "grad_norm": 0.8305770274498132,
+      "learning_rate": 9.962201733722369e-06,
+      "loss": 0.4247,
+      "step": 830
+    },
+    {
+      "epoch": 0.06817902120851622,
+      "grad_norm": 0.7195198782142942,
+      "learning_rate": 9.96203848909398e-06,
+      "loss": 0.3671,
+      "step": 831
+    },
+    {
+      "epoch": 0.06826106575870698,
+      "grad_norm": 0.7518704800600448,
+      "learning_rate": 9.961874894054407e-06,
+      "loss": 0.4114,
+      "step": 832
+    },
+    {
+      "epoch": 0.06834311030889773,
+      "grad_norm": 1.1253583843187456,
+      "learning_rate": 9.9617109486152e-06,
+      "loss": 0.4067,
+      "step": 833
+    },
+    {
+      "epoch": 0.06842515485908848,
+      "grad_norm": 1.0157503255728841,
+      "learning_rate": 9.961546652787937e-06,
+      "loss": 0.4579,
+      "step": 834
+    },
+    {
+      "epoch": 0.06850719940927924,
+      "grad_norm": 0.7638527294071445,
+      "learning_rate": 9.96138200658422e-06,
+      "loss": 0.4966,
+      "step": 835
+    },
+    {
+      "epoch": 0.06858924395947,
+      "grad_norm": 0.7524118615609422,
+      "learning_rate": 9.961217010015679e-06,
+      "loss": 0.3933,
+      "step": 836
+    },
+    {
+      "epoch": 0.06867128850966074,
+      "grad_norm": 0.7171987527369611,
+      "learning_rate": 9.961051663093962e-06,
+      "loss": 0.4711,
+      "step": 837
+    },
+    {
+      "epoch": 0.0687533330598515,
+      "grad_norm": 0.6954227123865457,
+      "learning_rate": 9.960885965830748e-06,
+      "loss": 0.4051,
+      "step": 838
+    },
+    {
+      "epoch": 0.06883537761004226,
+      "grad_norm": 0.7677645402953993,
+      "learning_rate": 9.960719918237735e-06,
+      "loss": 0.3936,
+      "step": 839
+    },
+    {
+      "epoch": 0.06891742216023301,
+      "grad_norm": 1.249771091919413,
+      "learning_rate": 9.960553520326655e-06,
+      "loss": 0.4283,
+      "step": 840
+    },
+    {
+      "epoch": 0.06899946671042376,
+      "grad_norm": 1.086126015497951,
+      "learning_rate": 9.960386772109253e-06,
+      "loss": 0.4499,
+      "step": 841
+    },
+    {
+      "epoch": 0.06908151126061451,
+      "grad_norm": 1.3049367413643127,
+      "learning_rate": 9.960219673597307e-06,
+      "loss": 0.4545,
+      "step": 842
+    },
+    {
+      "epoch": 0.06916355581080527,
+      "grad_norm": 1.14774218072718,
+      "learning_rate": 9.960052224802617e-06,
+      "loss": 0.4245,
+      "step": 843
+    },
+    {
+      "epoch": 0.06924560036099602,
+      "grad_norm": 0.9836341158386622,
+      "learning_rate": 9.959884425737008e-06,
+      "loss": 0.4504,
+      "step": 844
+    },
+    {
+      "epoch": 0.06932764491118677,
+      "grad_norm": 0.8924807008365208,
+      "learning_rate": 9.959716276412327e-06,
+      "loss": 0.4561,
+      "step": 845
+    },
+    {
+      "epoch": 0.06940968946137753,
+      "grad_norm": 0.6936910880640923,
+      "learning_rate": 9.959547776840456e-06,
+      "loss": 0.4868,
+      "step": 846
+    },
+    {
+      "epoch": 0.06949173401156829,
+      "grad_norm": 0.8373245285095935,
+      "learning_rate": 9.959378927033286e-06,
+      "loss": 0.4219,
+      "step": 847
+    },
+    {
+      "epoch": 0.06957377856175903,
+      "grad_norm": 1.2304606499509505,
+      "learning_rate": 9.959209727002743e-06,
+      "loss": 0.4314,
+      "step": 848
+    },
+    {
+      "epoch": 0.06965582311194979,
+      "grad_norm": 0.6850824073037841,
+      "learning_rate": 9.959040176760779e-06,
+      "loss": 0.4004,
+      "step": 849
+    },
+    {
+      "epoch": 0.06973786766214055,
+      "grad_norm": 0.8407589705078783,
+      "learning_rate": 9.958870276319364e-06,
+      "loss": 0.4571,
+      "step": 850
+    },
+    {
+      "epoch": 0.06981991221233129,
+      "grad_norm": 0.7932644766471253,
+      "learning_rate": 9.958700025690496e-06,
+      "loss": 0.4501,
+      "step": 851
+    },
+    {
+      "epoch": 0.06990195676252205,
+      "grad_norm": 0.7388254719841849,
+      "learning_rate": 9.9585294248862e-06,
+      "loss": 0.4562,
+      "step": 852
+    },
+    {
+      "epoch": 0.0699840013127128,
+      "grad_norm": 1.1298381484092663,
+      "learning_rate": 9.958358473918523e-06,
+      "loss": 0.4252,
+      "step": 853
+    },
+    {
+      "epoch": 0.07006604586290356,
+      "grad_norm": 0.7007826742491075,
+      "learning_rate": 9.958187172799536e-06,
+      "loss": 0.4128,
+      "step": 854
+    },
+    {
+      "epoch": 0.0701480904130943,
+      "grad_norm": 1.1858887404203176,
+      "learning_rate": 9.958015521541336e-06,
+      "loss": 0.4343,
+      "step": 855
+    },
+    {
+      "epoch": 0.07023013496328506,
+      "grad_norm": 0.980151521865545,
+      "learning_rate": 9.957843520156047e-06,
+      "loss": 0.4373,
+      "step": 856
+    },
+    {
+      "epoch": 0.07031217951347582,
+      "grad_norm": 0.7310750944470014,
+      "learning_rate": 9.957671168655814e-06,
+      "loss": 0.4744,
+      "step": 857
+    },
+    {
+      "epoch": 0.07039422406366658,
+      "grad_norm": 0.7178513395996499,
+      "learning_rate": 9.957498467052807e-06,
+      "loss": 0.4078,
+      "step": 858
+    },
+    {
+      "epoch": 0.07047626861385732,
+      "grad_norm": 0.7815808397247829,
+      "learning_rate": 9.957325415359225e-06,
+      "loss": 0.4015,
+      "step": 859
+    },
+    {
+      "epoch": 0.07055831316404808,
+      "grad_norm": 0.8509615027554758,
+      "learning_rate": 9.957152013587287e-06,
+      "loss": 0.4721,
+      "step": 860
+    },
+    {
+      "epoch": 0.07064035771423884,
+      "grad_norm": 0.7713550403946904,
+      "learning_rate": 9.956978261749235e-06,
+      "loss": 0.423,
+      "step": 861
+    },
+    {
+      "epoch": 0.07072240226442958,
+      "grad_norm": 0.7062199270119384,
+      "learning_rate": 9.956804159857344e-06,
+      "loss": 0.4462,
+      "step": 862
+    },
+    {
+      "epoch": 0.07080444681462034,
+      "grad_norm": 0.947921175058583,
+      "learning_rate": 9.956629707923906e-06,
+      "loss": 0.3844,
+      "step": 863
+    },
+    {
+      "epoch": 0.0708864913648111,
+      "grad_norm": 0.7237267006884891,
+      "learning_rate": 9.95645490596124e-06,
+      "loss": 0.5079,
+      "step": 864
+    },
+    {
+      "epoch": 0.07096853591500185,
+      "grad_norm": 0.655085148042254,
+      "learning_rate": 9.956279753981696e-06,
+      "loss": 0.4189,
+      "step": 865
+    },
+    {
+      "epoch": 0.0710505804651926,
+      "grad_norm": 0.691980230640483,
+      "learning_rate": 9.956104251997635e-06,
+      "loss": 0.4175,
+      "step": 866
+    },
+    {
+      "epoch": 0.07113262501538335,
+      "grad_norm": 0.6986008512965722,
+      "learning_rate": 9.955928400021457e-06,
+      "loss": 0.4222,
+      "step": 867
+    },
+    {
+      "epoch": 0.07121466956557411,
+      "grad_norm": 0.783787321772155,
+      "learning_rate": 9.955752198065577e-06,
+      "loss": 0.4445,
+      "step": 868
+    },
+    {
+      "epoch": 0.07129671411576487,
+      "grad_norm": 0.9199050845754191,
+      "learning_rate": 9.95557564614244e-06,
+      "loss": 0.4414,
+      "step": 869
+    },
+    {
+      "epoch": 0.07137875866595561,
+      "grad_norm": 1.0221417999723401,
+      "learning_rate": 9.955398744264509e-06,
+      "loss": 0.4309,
+      "step": 870
+    },
+    {
+      "epoch": 0.07146080321614637,
+      "grad_norm": 0.7355849414671902,
+      "learning_rate": 9.955221492444283e-06,
+      "loss": 0.454,
+      "step": 871
+    },
+    {
+      "epoch": 0.07154284776633713,
+      "grad_norm": 0.7046928827529818,
+      "learning_rate": 9.955043890694275e-06,
+      "loss": 0.3999,
+      "step": 872
+    },
+    {
+      "epoch": 0.07162489231652787,
+      "grad_norm": 0.768934199908008,
+      "learning_rate": 9.954865939027028e-06,
+      "loss": 0.413,
+      "step": 873
+    },
+    {
+      "epoch": 0.07170693686671863,
+      "grad_norm": 0.6883138267404174,
+      "learning_rate": 9.95468763745511e-06,
+      "loss": 0.4065,
+      "step": 874
+    },
+    {
+      "epoch": 0.07178898141690938,
+      "grad_norm": 0.7019350962955648,
+      "learning_rate": 9.95450898599111e-06,
+      "loss": 0.4163,
+      "step": 875
+    },
+    {
+      "epoch": 0.07187102596710014,
+      "grad_norm": 0.7219099885019596,
+      "learning_rate": 9.954329984647647e-06,
+      "loss": 0.3995,
+      "step": 876
+    },
+    {
+      "epoch": 0.07195307051729088,
+      "grad_norm": 0.7740303830137161,
+      "learning_rate": 9.95415063343736e-06,
+      "loss": 0.4461,
+      "step": 877
+    },
+    {
+      "epoch": 0.07203511506748164,
+      "grad_norm": 0.7506149195203048,
+      "learning_rate": 9.953970932372915e-06,
+      "loss": 0.4504,
+      "step": 878
+    },
+    {
+      "epoch": 0.0721171596176724,
+      "grad_norm": 0.6841146588360545,
+      "learning_rate": 9.953790881467002e-06,
+      "loss": 0.4351,
+      "step": 879
+    },
+    {
+      "epoch": 0.07219920416786314,
+      "grad_norm": 0.6955057364106116,
+      "learning_rate": 9.953610480732334e-06,
+      "loss": 0.395,
+      "step": 880
+    },
+    {
+      "epoch": 0.0722812487180539,
+      "grad_norm": 0.8472989469116042,
+      "learning_rate": 9.953429730181653e-06,
+      "loss": 0.4523,
+      "step": 881
+    },
+    {
+      "epoch": 0.07236329326824466,
+      "grad_norm": 0.7076974953858242,
+      "learning_rate": 9.953248629827726e-06,
+      "loss": 0.4304,
+      "step": 882
+    },
+    {
+      "epoch": 0.07244533781843542,
+      "grad_norm": 0.7262891227030667,
+      "learning_rate": 9.953067179683336e-06,
+      "loss": 0.4311,
+      "step": 883
+    },
+    {
+      "epoch": 0.07252738236862616,
+      "grad_norm": 1.2932081893528897,
+      "learning_rate": 9.9528853797613e-06,
+      "loss": 0.476,
+      "step": 884
+    },
+    {
+      "epoch": 0.07260942691881692,
+      "grad_norm": 0.6963747973106224,
+      "learning_rate": 9.952703230074454e-06,
+      "loss": 0.4224,
+      "step": 885
+    },
+    {
+      "epoch": 0.07269147146900767,
+      "grad_norm": 0.8955994430951031,
+      "learning_rate": 9.952520730635665e-06,
+      "loss": 0.474,
+      "step": 886
+    },
+    {
+      "epoch": 0.07277351601919843,
+      "grad_norm": 0.8853102345254197,
+      "learning_rate": 9.952337881457819e-06,
+      "loss": 0.4233,
+      "step": 887
+    },
+    {
+      "epoch": 0.07285556056938917,
+      "grad_norm": 0.8841222377669077,
+      "learning_rate": 9.952154682553828e-06,
+      "loss": 0.4335,
+      "step": 888
+    },
+    {
+      "epoch": 0.07293760511957993,
+      "grad_norm": 0.7363315961335914,
+      "learning_rate": 9.951971133936629e-06,
+      "loss": 0.4481,
+      "step": 889
+    },
+    {
+      "epoch": 0.07301964966977069,
+      "grad_norm": 0.6886394101879773,
+      "learning_rate": 9.951787235619183e-06,
+      "loss": 0.4283,
+      "step": 890
+    },
+    {
+      "epoch": 0.07310169421996143,
+      "grad_norm": 0.7342238704539885,
+      "learning_rate": 9.951602987614482e-06,
+      "loss": 0.4201,
+      "step": 891
+    },
+    {
+      "epoch": 0.07318373877015219,
+      "grad_norm": 0.7172350343042992,
+      "learning_rate": 9.951418389935529e-06,
+      "loss": 0.4421,
+      "step": 892
+    },
+    {
+      "epoch": 0.07326578332034295,
+      "grad_norm": 0.6431648471462383,
+      "learning_rate": 9.951233442595366e-06,
+      "loss": 0.3911,
+      "step": 893
+    },
+    {
+      "epoch": 0.0733478278705337,
+      "grad_norm": 0.6478437120304857,
+      "learning_rate": 9.951048145607054e-06,
+      "loss": 0.4236,
+      "step": 894
+    },
+    {
+      "epoch": 0.07342987242072445,
+      "grad_norm": 0.7802761489429634,
+      "learning_rate": 9.950862498983676e-06,
+      "loss": 0.3994,
+      "step": 895
+    },
+    {
+      "epoch": 0.0735119169709152,
+      "grad_norm": 0.7527347537586383,
+      "learning_rate": 9.95067650273834e-06,
+      "loss": 0.4134,
+      "step": 896
+    },
+    {
+      "epoch": 0.07359396152110596,
+      "grad_norm": 0.6890214690531933,
+      "learning_rate": 9.950490156884186e-06,
+      "loss": 0.4113,
+      "step": 897
+    },
+    {
+      "epoch": 0.07367600607129672,
+      "grad_norm": 0.6838558579117131,
+      "learning_rate": 9.95030346143437e-06,
+      "loss": 0.4317,
+      "step": 898
+    },
+    {
+      "epoch": 0.07375805062148746,
+      "grad_norm": 0.678739890794481,
+      "learning_rate": 9.950116416402076e-06,
+      "loss": 0.4778,
+      "step": 899
+    },
+    {
+      "epoch": 0.07384009517167822,
+      "grad_norm": 0.6806451820430354,
+      "learning_rate": 9.949929021800514e-06,
+      "loss": 0.4229,
+      "step": 900
+    },
+    {
+      "epoch": 0.07392213972186898,
+      "grad_norm": 0.6097848204732135,
+      "learning_rate": 9.949741277642917e-06,
+      "loss": 0.4232,
+      "step": 901
+    },
+    {
+      "epoch": 0.07400418427205972,
+      "grad_norm": 0.6592703303046097,
+      "learning_rate": 9.949553183942544e-06,
+      "loss": 0.4055,
+      "step": 902
+    },
+    {
+      "epoch": 0.07408622882225048,
+      "grad_norm": 0.7327266300162744,
+      "learning_rate": 9.949364740712675e-06,
+      "loss": 0.4332,
+      "step": 903
+    },
+    {
+      "epoch": 0.07416827337244124,
+      "grad_norm": 0.6950655072551598,
+      "learning_rate": 9.949175947966623e-06,
+      "loss": 0.4174,
+      "step": 904
+    },
+    {
+      "epoch": 0.074250317922632,
+      "grad_norm": 0.689736571938754,
+      "learning_rate": 9.948986805717715e-06,
+      "loss": 0.4152,
+      "step": 905
+    },
+    {
+      "epoch": 0.07433236247282274,
+      "grad_norm": 0.6697604040175108,
+      "learning_rate": 9.94879731397931e-06,
+      "loss": 0.4393,
+      "step": 906
+    },
+    {
+      "epoch": 0.0744144070230135,
+      "grad_norm": 0.6336772643620869,
+      "learning_rate": 9.948607472764791e-06,
+      "loss": 0.3938,
+      "step": 907
+    },
+    {
+      "epoch": 0.07449645157320425,
+      "grad_norm": 0.6930726943638261,
+      "learning_rate": 9.948417282087561e-06,
+      "loss": 0.4419,
+      "step": 908
+    },
+    {
+      "epoch": 0.074578496123395,
+      "grad_norm": 0.8882878295304286,
+      "learning_rate": 9.948226741961053e-06,
+      "loss": 0.4478,
+      "step": 909
+    },
+    {
+      "epoch": 0.07466054067358575,
+      "grad_norm": 0.7635954101999421,
+      "learning_rate": 9.948035852398722e-06,
+      "loss": 0.421,
+      "step": 910
+    },
+    {
+      "epoch": 0.07474258522377651,
+      "grad_norm": 0.6956740243472812,
+      "learning_rate": 9.94784461341405e-06,
+      "loss": 0.3993,
+      "step": 911
+    },
+    {
+      "epoch": 0.07482462977396727,
+      "grad_norm": 0.6943207412592616,
+      "learning_rate": 9.947653025020541e-06,
+      "loss": 0.42,
+      "step": 912
+    },
+    {
+      "epoch": 0.07490667432415801,
+      "grad_norm": 0.7912731434703135,
+      "learning_rate": 9.947461087231725e-06,
+      "loss": 0.4854,
+      "step": 913
+    },
+    {
+      "epoch": 0.07498871887434877,
+      "grad_norm": 0.7816489574607006,
+      "learning_rate": 9.947268800061153e-06,
+      "loss": 0.4219,
+      "step": 914
+    },
+    {
+      "epoch": 0.07507076342453953,
+      "grad_norm": 0.7449548075945912,
+      "learning_rate": 9.947076163522408e-06,
+      "loss": 0.4564,
+      "step": 915
+    },
+    {
+      "epoch": 0.07515280797473028,
+      "grad_norm": 0.7893298602374255,
+      "learning_rate": 9.946883177629094e-06,
+      "loss": 0.4533,
+      "step": 916
+    },
+    {
+      "epoch": 0.07523485252492103,
+      "grad_norm": 0.6835664232885097,
+      "learning_rate": 9.946689842394836e-06,
+      "loss": 0.4207,
+      "step": 917
+    },
+    {
+      "epoch": 0.07531689707511179,
+      "grad_norm": 0.7261451142975969,
+      "learning_rate": 9.94649615783329e-06,
+      "loss": 0.4659,
+      "step": 918
+    },
+    {
+      "epoch": 0.07539894162530254,
+      "grad_norm": 0.7026996002248749,
+      "learning_rate": 9.94630212395813e-06,
+      "loss": 0.4225,
+      "step": 919
+    },
+    {
+      "epoch": 0.07548098617549329,
+      "grad_norm": 0.6932871787626236,
+      "learning_rate": 9.946107740783063e-06,
+      "loss": 0.4411,
+      "step": 920
+    },
+    {
+      "epoch": 0.07556303072568404,
+      "grad_norm": 0.7353183343163795,
+      "learning_rate": 9.945913008321814e-06,
+      "loss": 0.4409,
+      "step": 921
+    },
+    {
+      "epoch": 0.0756450752758748,
+      "grad_norm": 0.871076373362069,
+      "learning_rate": 9.945717926588133e-06,
+      "loss": 0.4366,
+      "step": 922
+    },
+    {
+      "epoch": 0.07572711982606556,
+      "grad_norm": 0.6988609904423111,
+      "learning_rate": 9.9455224955958e-06,
+      "loss": 0.432,
+      "step": 923
+    },
+    {
+      "epoch": 0.0758091643762563,
+      "grad_norm": 0.6545253604519848,
+      "learning_rate": 9.945326715358612e-06,
+      "loss": 0.4406,
+      "step": 924
+    },
+    {
+      "epoch": 0.07589120892644706,
+      "grad_norm": 0.7083846834587697,
+      "learning_rate": 9.945130585890399e-06,
+      "loss": 0.4461,
+      "step": 925
+    },
+    {
+      "epoch": 0.07597325347663782,
+      "grad_norm": 1.2898628039675772,
+      "learning_rate": 9.944934107205006e-06,
+      "loss": 0.4361,
+      "step": 926
+    },
+    {
+      "epoch": 0.07605529802682857,
+      "grad_norm": 0.697107161232331,
+      "learning_rate": 9.944737279316312e-06,
+      "loss": 0.4573,
+      "step": 927
+    },
+    {
+      "epoch": 0.07613734257701932,
+      "grad_norm": 0.7189401834494062,
+      "learning_rate": 9.944540102238217e-06,
+      "loss": 0.3838,
+      "step": 928
+    },
+    {
+      "epoch": 0.07621938712721008,
+      "grad_norm": 0.7842953603209286,
+      "learning_rate": 9.944342575984643e-06,
+      "loss": 0.4183,
+      "step": 929
+    },
+    {
+      "epoch": 0.07630143167740083,
+      "grad_norm": 0.7738473268711619,
+      "learning_rate": 9.94414470056954e-06,
+      "loss": 0.3732,
+      "step": 930
+    },
+    {
+      "epoch": 0.07638347622759158,
+      "grad_norm": 0.7210879835269495,
+      "learning_rate": 9.943946476006882e-06,
+      "loss": 0.4223,
+      "step": 931
+    },
+    {
+      "epoch": 0.07646552077778233,
+      "grad_norm": 0.6551856613828901,
+      "learning_rate": 9.943747902310666e-06,
+      "loss": 0.3744,
+      "step": 932
+    },
+    {
+      "epoch": 0.07654756532797309,
+      "grad_norm": 0.6885321509046528,
+      "learning_rate": 9.943548979494918e-06,
+      "loss": 0.4319,
+      "step": 933
+    },
+    {
+      "epoch": 0.07662960987816385,
+      "grad_norm": 0.7713390769070645,
+      "learning_rate": 9.943349707573681e-06,
+      "loss": 0.4151,
+      "step": 934
+    },
+    {
+      "epoch": 0.07671165442835459,
+      "grad_norm": 0.7630425826260916,
+      "learning_rate": 9.943150086561031e-06,
+      "loss": 0.4686,
+      "step": 935
+    },
+    {
+      "epoch": 0.07679369897854535,
+      "grad_norm": 0.8875227056794636,
+      "learning_rate": 9.942950116471063e-06,
+      "loss": 0.4256,
+      "step": 936
+    },
+    {
+      "epoch": 0.0768757435287361,
+      "grad_norm": 0.9609378210550329,
+      "learning_rate": 9.9427497973179e-06,
+      "loss": 0.4506,
+      "step": 937
+    },
+    {
+      "epoch": 0.07695778807892685,
+      "grad_norm": 0.7510572080829931,
+      "learning_rate": 9.942549129115686e-06,
+      "loss": 0.5057,
+      "step": 938
+    },
+    {
+      "epoch": 0.07703983262911761,
+      "grad_norm": 0.6752516762993702,
+      "learning_rate": 9.942348111878594e-06,
+      "loss": 0.3878,
+      "step": 939
+    },
+    {
+      "epoch": 0.07712187717930837,
+      "grad_norm": 1.1233156209568713,
+      "learning_rate": 9.94214674562082e-06,
+      "loss": 0.4667,
+      "step": 940
+    },
+    {
+      "epoch": 0.07720392172949912,
+      "grad_norm": 0.7449681283722189,
+      "learning_rate": 9.941945030356584e-06,
+      "loss": 0.4223,
+      "step": 941
+    },
+    {
+      "epoch": 0.07728596627968987,
+      "grad_norm": 0.6985191627577464,
+      "learning_rate": 9.941742966100128e-06,
+      "loss": 0.4223,
+      "step": 942
+    },
+    {
+      "epoch": 0.07736801082988062,
+      "grad_norm": 0.6914461864723972,
+      "learning_rate": 9.941540552865722e-06,
+      "loss": 0.4131,
+      "step": 943
+    },
+    {
+      "epoch": 0.07745005538007138,
+      "grad_norm": 0.7544006376179547,
+      "learning_rate": 9.941337790667662e-06,
+      "loss": 0.447,
+      "step": 944
+    },
+    {
+      "epoch": 0.07753209993026214,
+      "grad_norm": 0.7073026888181757,
+      "learning_rate": 9.941134679520268e-06,
+      "loss": 0.4277,
+      "step": 945
+    },
+    {
+      "epoch": 0.07761414448045288,
+      "grad_norm": 0.643343866459783,
+      "learning_rate": 9.94093121943788e-06,
+      "loss": 0.4177,
+      "step": 946
+    },
+    {
+      "epoch": 0.07769618903064364,
+      "grad_norm": 0.7460106240755561,
+      "learning_rate": 9.940727410434868e-06,
+      "loss": 0.4658,
+      "step": 947
+    },
+    {
+      "epoch": 0.0777782335808344,
+      "grad_norm": 0.7725057675577542,
+      "learning_rate": 9.940523252525623e-06,
+      "loss": 0.447,
+      "step": 948
+    },
+    {
+      "epoch": 0.07786027813102514,
+      "grad_norm": 0.7674120448312616,
+      "learning_rate": 9.940318745724564e-06,
+      "loss": 0.4519,
+      "step": 949
+    },
+    {
+      "epoch": 0.0779423226812159,
+      "grad_norm": 1.2350816033993233,
+      "learning_rate": 9.940113890046133e-06,
+      "loss": 0.4565,
+      "step": 950
+    },
+    {
+      "epoch": 0.07802436723140665,
+      "grad_norm": 0.6770880968202859,
+      "learning_rate": 9.939908685504796e-06,
+      "loss": 0.3869,
+      "step": 951
+    },
+    {
+      "epoch": 0.07810641178159741,
+      "grad_norm": 0.764976404793502,
+      "learning_rate": 9.939703132115045e-06,
+      "loss": 0.4283,
+      "step": 952
+    },
+    {
+      "epoch": 0.07818845633178816,
+      "grad_norm": 0.8101584495342595,
+      "learning_rate": 9.939497229891393e-06,
+      "loss": 0.4451,
+      "step": 953
+    },
+    {
+      "epoch": 0.07827050088197891,
+      "grad_norm": 0.6503397799856954,
+      "learning_rate": 9.939290978848383e-06,
+      "loss": 0.4162,
+      "step": 954
+    },
+    {
+      "epoch": 0.07835254543216967,
+      "grad_norm": 0.6674709624121883,
+      "learning_rate": 9.939084379000582e-06,
+      "loss": 0.4081,
+      "step": 955
+    },
+    {
+      "epoch": 0.07843458998236043,
+      "grad_norm": 0.6719084212635281,
+      "learning_rate": 9.938877430362573e-06,
+      "loss": 0.4328,
+      "step": 956
+    },
+    {
+      "epoch": 0.07851663453255117,
+      "grad_norm": 0.7105399433680906,
+      "learning_rate": 9.938670132948978e-06,
+      "loss": 0.4426,
+      "step": 957
+    },
+    {
+      "epoch": 0.07859867908274193,
+      "grad_norm": 0.7475682885890519,
+      "learning_rate": 9.938462486774432e-06,
+      "loss": 0.4748,
+      "step": 958
+    },
+    {
+      "epoch": 0.07868072363293269,
+      "grad_norm": 0.7357594849524249,
+      "learning_rate": 9.9382544918536e-06,
+      "loss": 0.4261,
+      "step": 959
+    },
+    {
+      "epoch": 0.07876276818312343,
+      "grad_norm": 0.7511356934318226,
+      "learning_rate": 9.938046148201168e-06,
+      "loss": 0.4195,
+      "step": 960
+    },
+    {
+      "epoch": 0.07884481273331419,
+      "grad_norm": 0.8897985060127483,
+      "learning_rate": 9.937837455831852e-06,
+      "loss": 0.4776,
+      "step": 961
+    },
+    {
+      "epoch": 0.07892685728350494,
+      "grad_norm": 0.7313606580044121,
+      "learning_rate": 9.937628414760389e-06,
+      "loss": 0.4493,
+      "step": 962
+    },
+    {
+      "epoch": 0.0790089018336957,
+      "grad_norm": 0.7552914662034828,
+      "learning_rate": 9.937419025001537e-06,
+      "loss": 0.4163,
+      "step": 963
+    },
+    {
+      "epoch": 0.07909094638388645,
+      "grad_norm": 0.6880067089874974,
+      "learning_rate": 9.937209286570089e-06,
+      "loss": 0.4647,
+      "step": 964
+    },
+    {
+      "epoch": 0.0791729909340772,
+      "grad_norm": 0.6681690870529101,
+      "learning_rate": 9.936999199480854e-06,
+      "loss": 0.4104,
+      "step": 965
+    },
+    {
+      "epoch": 0.07925503548426796,
+      "grad_norm": 0.6863014432640218,
+      "learning_rate": 9.936788763748666e-06,
+      "loss": 0.4234,
+      "step": 966
+    },
+    {
+      "epoch": 0.07933708003445872,
+      "grad_norm": 0.6848333446601156,
+      "learning_rate": 9.936577979388387e-06,
+      "loss": 0.3925,
+      "step": 967
+    },
+    {
+      "epoch": 0.07941912458464946,
+      "grad_norm": 0.8358039902304838,
+      "learning_rate": 9.936366846414902e-06,
+      "loss": 0.4238,
+      "step": 968
+    },
+    {
+      "epoch": 0.07950116913484022,
+      "grad_norm": 0.7005072278565283,
+      "learning_rate": 9.936155364843121e-06,
+      "loss": 0.4067,
+      "step": 969
+    },
+    {
+      "epoch": 0.07958321368503098,
+      "grad_norm": 0.7347531023339785,
+      "learning_rate": 9.93594353468798e-06,
+      "loss": 0.4625,
+      "step": 970
+    },
+    {
+      "epoch": 0.07966525823522172,
+      "grad_norm": 0.7996366688989736,
+      "learning_rate": 9.935731355964436e-06,
+      "loss": 0.468,
+      "step": 971
+    },
+    {
+      "epoch": 0.07974730278541248,
+      "grad_norm": 0.7028483575972714,
+      "learning_rate": 9.935518828687473e-06,
+      "loss": 0.3866,
+      "step": 972
+    },
+    {
+      "epoch": 0.07982934733560323,
+      "grad_norm": 0.743354493091255,
+      "learning_rate": 9.935305952872101e-06,
+      "loss": 0.4736,
+      "step": 973
+    },
+    {
+      "epoch": 0.07991139188579399,
+      "grad_norm": 0.9220902604734729,
+      "learning_rate": 9.93509272853335e-06,
+      "loss": 0.4574,
+      "step": 974
+    },
+    {
+      "epoch": 0.07999343643598474,
+      "grad_norm": 1.7502816610519736,
+      "learning_rate": 9.934879155686282e-06,
+      "loss": 0.4388,
+      "step": 975
+    },
+    {
+      "epoch": 0.08007548098617549,
+      "grad_norm": 1.1522665175262483,
+      "learning_rate": 9.934665234345976e-06,
+      "loss": 0.4415,
+      "step": 976
+    },
+    {
+      "epoch": 0.08015752553636625,
+      "grad_norm": 0.7442682661279915,
+      "learning_rate": 9.934450964527536e-06,
+      "loss": 0.3976,
+      "step": 977
+    },
+    {
+      "epoch": 0.080239570086557,
+      "grad_norm": 0.7429787221366082,
+      "learning_rate": 9.9342363462461e-06,
+      "loss": 0.3975,
+      "step": 978
+    },
+    {
+      "epoch": 0.08032161463674775,
+      "grad_norm": 0.7670940570385435,
+      "learning_rate": 9.93402137951682e-06,
+      "loss": 0.464,
+      "step": 979
+    },
+    {
+      "epoch": 0.08040365918693851,
+      "grad_norm": 0.8120672316617124,
+      "learning_rate": 9.933806064354876e-06,
+      "loss": 0.3867,
+      "step": 980
+    },
+    {
+      "epoch": 0.08048570373712927,
+      "grad_norm": 0.9700209124496794,
+      "learning_rate": 9.933590400775476e-06,
+      "loss": 0.4426,
+      "step": 981
+    },
+    {
+      "epoch": 0.08056774828732001,
+      "grad_norm": 0.6219710262227189,
+      "learning_rate": 9.933374388793848e-06,
+      "loss": 0.394,
+      "step": 982
+    },
+    {
+      "epoch": 0.08064979283751077,
+      "grad_norm": 2.3039172189744472,
+      "learning_rate": 9.933158028425247e-06,
+      "loss": 0.4112,
+      "step": 983
+    },
+    {
+      "epoch": 0.08073183738770152,
+      "grad_norm": 0.6837937934377377,
+      "learning_rate": 9.93294131968495e-06,
+      "loss": 0.4364,
+      "step": 984
+    },
+    {
+      "epoch": 0.08081388193789228,
+      "grad_norm": 0.6711868430532517,
+      "learning_rate": 9.932724262588265e-06,
+      "loss": 0.4485,
+      "step": 985
+    },
+    {
+      "epoch": 0.08089592648808303,
+      "grad_norm": 0.7243522554632826,
+      "learning_rate": 9.932506857150515e-06,
+      "loss": 0.4223,
+      "step": 986
+    },
+    {
+      "epoch": 0.08097797103827378,
+      "grad_norm": 0.7230729965485122,
+      "learning_rate": 9.932289103387056e-06,
+      "loss": 0.4707,
+      "step": 987
+    },
+    {
+      "epoch": 0.08106001558846454,
+      "grad_norm": 0.7315737055394317,
+      "learning_rate": 9.932071001313265e-06,
+      "loss": 0.4492,
+      "step": 988
+    },
+    {
+      "epoch": 0.08114206013865528,
+      "grad_norm": 0.7565897513852394,
+      "learning_rate": 9.931852550944545e-06,
+      "loss": 0.4759,
+      "step": 989
+    },
+    {
+      "epoch": 0.08122410468884604,
+      "grad_norm": 0.7617964081738033,
+      "learning_rate": 9.931633752296322e-06,
+      "loss": 0.4148,
+      "step": 990
+    },
+    {
+      "epoch": 0.0813061492390368,
+      "grad_norm": 0.7014227772353403,
+      "learning_rate": 9.931414605384044e-06,
+      "loss": 0.4125,
+      "step": 991
+    },
+    {
+      "epoch": 0.08138819378922756,
+      "grad_norm": 0.7256471475942593,
+      "learning_rate": 9.931195110223193e-06,
+      "loss": 0.4204,
+      "step": 992
+    },
+    {
+      "epoch": 0.0814702383394183,
+      "grad_norm": 0.7433496824395379,
+      "learning_rate": 9.930975266829263e-06,
+      "loss": 0.4198,
+      "step": 993
+    },
+    {
+      "epoch": 0.08155228288960906,
+      "grad_norm": 0.6375470855244235,
+      "learning_rate": 9.930755075217783e-06,
+      "loss": 0.4001,
+      "step": 994
+    },
+    {
+      "epoch": 0.08163432743979981,
+      "grad_norm": 0.6634490102429524,
+      "learning_rate": 9.930534535404302e-06,
+      "loss": 0.4286,
+      "step": 995
+    },
+    {
+      "epoch": 0.08171637198999057,
+      "grad_norm": 0.6929563429283587,
+      "learning_rate": 9.930313647404393e-06,
+      "loss": 0.4167,
+      "step": 996
+    },
+    {
+      "epoch": 0.08179841654018132,
+      "grad_norm": 0.7061253452778037,
+      "learning_rate": 9.930092411233654e-06,
+      "loss": 0.4471,
+      "step": 997
+    },
+    {
+      "epoch": 0.08188046109037207,
+      "grad_norm": 0.7318166001701418,
+      "learning_rate": 9.929870826907711e-06,
+      "loss": 0.4654,
+      "step": 998
+    },
+    {
+      "epoch": 0.08196250564056283,
+      "grad_norm": 0.7116218930335934,
+      "learning_rate": 9.929648894442212e-06,
+      "loss": 0.435,
+      "step": 999
+    },
+    {
+      "epoch": 0.08204455019075357,
+      "grad_norm": 0.7413030730369929,
+      "learning_rate": 9.929426613852827e-06,
+      "loss": 0.4691,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08212659474094433,
+      "grad_norm": 0.7109749000969268,
+      "learning_rate": 9.929203985155252e-06,
+      "loss": 0.4178,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08220863929113509,
+      "grad_norm": 0.7101325678692918,
+      "learning_rate": 9.928981008365213e-06,
+      "loss": 0.4314,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08229068384132585,
+      "grad_norm": 0.8291479379563896,
+      "learning_rate": 9.928757683498456e-06,
+      "loss": 0.4718,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08237272839151659,
+      "grad_norm": 0.6782988467240904,
+      "learning_rate": 9.928534010570747e-06,
+      "loss": 0.4362,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08245477294170735,
+      "grad_norm": 0.6620475997342419,
+      "learning_rate": 9.928309989597887e-06,
+      "loss": 0.4206,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0825368174918981,
+      "grad_norm": 0.6926867878717202,
+      "learning_rate": 9.928085620595692e-06,
+      "loss": 0.4226,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08261886204208885,
+      "grad_norm": 0.8360693859680834,
+      "learning_rate": 9.927860903580007e-06,
+      "loss": 0.4119,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0827009065922796,
+      "grad_norm": 0.6796352438548476,
+      "learning_rate": 9.927635838566703e-06,
+      "loss": 0.4241,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08278295114247036,
+      "grad_norm": 0.7566544162276014,
+      "learning_rate": 9.927410425571674e-06,
+      "loss": 0.422,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08286499569266112,
+      "grad_norm": 0.6886997363985676,
+      "learning_rate": 9.927184664610836e-06,
+      "loss": 0.4216,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08294704024285186,
+      "grad_norm": 0.767230530678925,
+      "learning_rate": 9.926958555700134e-06,
+      "loss": 0.3984,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08302908479304262,
+      "grad_norm": 0.6937348274003537,
+      "learning_rate": 9.926732098855533e-06,
+      "loss": 0.4539,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08311112934323338,
+      "grad_norm": 0.6496219761762128,
+      "learning_rate": 9.92650529409303e-06,
+      "loss": 0.4065,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08319317389342414,
+      "grad_norm": 0.6687353646590134,
+      "learning_rate": 9.926278141428635e-06,
+      "loss": 0.3938,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08327521844361488,
+      "grad_norm": 0.8926425933928338,
+      "learning_rate": 9.926050640878394e-06,
+      "loss": 0.4339,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08335726299380564,
+      "grad_norm": 0.742549667992309,
+      "learning_rate": 9.92582279245837e-06,
+      "loss": 0.4769,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0834393075439964,
+      "grad_norm": 0.7025865279669093,
+      "learning_rate": 9.925594596184654e-06,
+      "loss": 0.4309,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08352135209418714,
+      "grad_norm": 0.7249100770004988,
+      "learning_rate": 9.925366052073361e-06,
+      "loss": 0.4454,
+      "step": 1018
+    },
+    {
+      "epoch": 0.0836033966443779,
+      "grad_norm": 0.7065978333483551,
+      "learning_rate": 9.925137160140632e-06,
+      "loss": 0.411,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08368544119456865,
+      "grad_norm": 0.6706066774815876,
+      "learning_rate": 9.92490792040263e-06,
+      "loss": 0.435,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08376748574475941,
+      "grad_norm": 0.6200450348920604,
+      "learning_rate": 9.924678332875542e-06,
+      "loss": 0.448,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08384953029495015,
+      "grad_norm": 0.7084692607020989,
+      "learning_rate": 9.924448397575582e-06,
+      "loss": 0.4132,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08393157484514091,
+      "grad_norm": 0.6736870831500558,
+      "learning_rate": 9.924218114518988e-06,
+      "loss": 0.4132,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08401361939533167,
+      "grad_norm": 0.6702252747035212,
+      "learning_rate": 9.923987483722023e-06,
+      "loss": 0.4262,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08409566394552243,
+      "grad_norm": 0.7831291570433181,
+      "learning_rate": 9.923756505200973e-06,
+      "loss": 0.4083,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08417770849571317,
+      "grad_norm": 0.6874715800844986,
+      "learning_rate": 9.923525178972149e-06,
+      "loss": 0.3683,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08425975304590393,
+      "grad_norm": 0.6836167814821877,
+      "learning_rate": 9.923293505051885e-06,
+      "loss": 0.3953,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08434179759609468,
+      "grad_norm": 0.6777278079852787,
+      "learning_rate": 9.923061483456547e-06,
+      "loss": 0.409,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08442384214628543,
+      "grad_norm": 0.6862091014435101,
+      "learning_rate": 9.922829114202513e-06,
+      "loss": 0.4148,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08450588669647618,
+      "grad_norm": 0.7424056241180106,
+      "learning_rate": 9.922596397306199e-06,
+      "loss": 0.3975,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08458793124666694,
+      "grad_norm": 0.6391184681213998,
+      "learning_rate": 9.922363332784034e-06,
+      "loss": 0.3885,
+      "step": 1031
+    },
+    {
+      "epoch": 0.0846699757968577,
+      "grad_norm": 0.6992696137506641,
+      "learning_rate": 9.922129920652481e-06,
+      "loss": 0.4218,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08475202034704844,
+      "grad_norm": 0.6558169128359677,
+      "learning_rate": 9.92189616092802e-06,
+      "loss": 0.4229,
+      "step": 1033
+    },
+    {
+      "epoch": 0.0848340648972392,
+      "grad_norm": 0.7000458306021213,
+      "learning_rate": 9.92166205362716e-06,
+      "loss": 0.4245,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08491610944742996,
+      "grad_norm": 0.6711951555984216,
+      "learning_rate": 9.921427598766432e-06,
+      "loss": 0.4409,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0849981539976207,
+      "grad_norm": 0.6955976846783224,
+      "learning_rate": 9.921192796362393e-06,
+      "loss": 0.4256,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08508019854781146,
+      "grad_norm": 0.7166870287698903,
+      "learning_rate": 9.920957646431628e-06,
+      "loss": 0.4265,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08516224309800222,
+      "grad_norm": 0.710983726479255,
+      "learning_rate": 9.920722148990737e-06,
+      "loss": 0.4301,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08524428764819297,
+      "grad_norm": 1.0047287352986747,
+      "learning_rate": 9.920486304056355e-06,
+      "loss": 0.7815,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08532633219838372,
+      "grad_norm": 0.7087288088109175,
+      "learning_rate": 9.920250111645138e-06,
+      "loss": 0.4381,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08540837674857447,
+      "grad_norm": 0.6701075285327227,
+      "learning_rate": 9.92001357177376e-06,
+      "loss": 0.4122,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08549042129876523,
+      "grad_norm": 0.724673079598315,
+      "learning_rate": 9.919776684458931e-06,
+      "loss": 0.4183,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08557246584895599,
+      "grad_norm": 0.6884466272764912,
+      "learning_rate": 9.919539449717373e-06,
+      "loss": 0.4422,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08565451039914673,
+      "grad_norm": 0.7656571946524281,
+      "learning_rate": 9.919301867565845e-06,
+      "loss": 0.4538,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08573655494933749,
+      "grad_norm": 0.6480145113354768,
+      "learning_rate": 9.919063938021123e-06,
+      "loss": 0.4377,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08581859949952825,
+      "grad_norm": 0.6608759176885819,
+      "learning_rate": 9.918825661100009e-06,
+      "loss": 0.4427,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08590064404971899,
+      "grad_norm": 0.6595823057513124,
+      "learning_rate": 9.918587036819328e-06,
+      "loss": 0.4193,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08598268859990975,
+      "grad_norm": 0.6554237050530093,
+      "learning_rate": 9.918348065195934e-06,
+      "loss": 0.4281,
+      "step": 1048
+    },
+    {
+      "epoch": 0.0860647331501005,
+      "grad_norm": 0.7545346461710065,
+      "learning_rate": 9.9181087462467e-06,
+      "loss": 0.4735,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08614677770029126,
+      "grad_norm": 0.6573118995796838,
+      "learning_rate": 9.917869079988529e-06,
+      "loss": 0.4053,
+      "step": 1050
+    },
+    {
+      "epoch": 0.086228822250482,
+      "grad_norm": 0.5527231567433245,
+      "learning_rate": 9.917629066438346e-06,
+      "loss": 0.3884,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08631086680067276,
+      "grad_norm": 0.6358988613837268,
+      "learning_rate": 9.917388705613099e-06,
+      "loss": 0.3765,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08639291135086352,
+      "grad_norm": 0.69036936626371,
+      "learning_rate": 9.917147997529759e-06,
+      "loss": 0.4373,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08647495590105428,
+      "grad_norm": 0.624132952260046,
+      "learning_rate": 9.916906942205328e-06,
+      "loss": 0.4477,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08655700045124502,
+      "grad_norm": 0.6612300601599008,
+      "learning_rate": 9.91666553965683e-06,
+      "loss": 0.4189,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08663904500143578,
+      "grad_norm": 0.7113183075168988,
+      "learning_rate": 9.916423789901312e-06,
+      "loss": 0.4103,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08672108955162654,
+      "grad_norm": 0.6857232483399694,
+      "learning_rate": 9.916181692955841e-06,
+      "loss": 0.4233,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08680313410181728,
+      "grad_norm": 0.7375707281641389,
+      "learning_rate": 9.91593924883752e-06,
+      "loss": 0.4307,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08688517865200804,
+      "grad_norm": 0.6752922173214406,
+      "learning_rate": 9.915696457563469e-06,
+      "loss": 0.3817,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0869672232021988,
+      "grad_norm": 0.6564900849142147,
+      "learning_rate": 9.915453319150829e-06,
+      "loss": 0.4549,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08704926775238955,
+      "grad_norm": 0.7255165723440383,
+      "learning_rate": 9.915209833616773e-06,
+      "loss": 0.4091,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0871313123025803,
+      "grad_norm": 0.6728665627436781,
+      "learning_rate": 9.914966000978497e-06,
+      "loss": 0.4504,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08721335685277105,
+      "grad_norm": 0.694940781485747,
+      "learning_rate": 9.914721821253218e-06,
+      "loss": 0.4279,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08729540140296181,
+      "grad_norm": 0.6721150116638501,
+      "learning_rate": 9.914477294458178e-06,
+      "loss": 0.4097,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08737744595315255,
+      "grad_norm": 0.681894881598639,
+      "learning_rate": 9.914232420610649e-06,
+      "loss": 0.4361,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08745949050334331,
+      "grad_norm": 0.7807027246232714,
+      "learning_rate": 9.91398719972792e-06,
+      "loss": 0.4385,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08754153505353407,
+      "grad_norm": 0.7067243541504444,
+      "learning_rate": 9.913741631827314e-06,
+      "loss": 0.4686,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08762357960372483,
+      "grad_norm": 0.6779428838423512,
+      "learning_rate": 9.913495716926165e-06,
+      "loss": 0.3974,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08770562415391557,
+      "grad_norm": 0.6632642701827999,
+      "learning_rate": 9.913249455041843e-06,
+      "loss": 0.4401,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08778766870410633,
+      "grad_norm": 0.6614633816266263,
+      "learning_rate": 9.913002846191739e-06,
+      "loss": 0.3846,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08786971325429709,
+      "grad_norm": 0.7229895222299344,
+      "learning_rate": 9.912755890393266e-06,
+      "loss": 0.4187,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08795175780448784,
+      "grad_norm": 0.6640987055109133,
+      "learning_rate": 9.912508587663867e-06,
+      "loss": 0.4104,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08803380235467859,
+      "grad_norm": 0.7069488306111321,
+      "learning_rate": 9.912260938021003e-06,
+      "loss": 0.4223,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08811584690486934,
+      "grad_norm": 0.7119552481519055,
+      "learning_rate": 9.912012941482163e-06,
+      "loss": 0.4541,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0881978914550601,
+      "grad_norm": 0.6663159607258136,
+      "learning_rate": 9.911764598064862e-06,
+      "loss": 0.4416,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08827993600525084,
+      "grad_norm": 0.6811971531092029,
+      "learning_rate": 9.911515907786637e-06,
+      "loss": 0.4312,
+      "step": 1076
+    },
+    {
+      "epoch": 0.0883619805554416,
+      "grad_norm": 0.6525403306679165,
+      "learning_rate": 9.911266870665048e-06,
+      "loss": 0.3804,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08844402510563236,
+      "grad_norm": 0.6345804648482422,
+      "learning_rate": 9.911017486717683e-06,
+      "loss": 0.4022,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08852606965582312,
+      "grad_norm": 0.9479582327611155,
+      "learning_rate": 9.910767755962153e-06,
+      "loss": 0.4261,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08860811420601386,
+      "grad_norm": 0.6705686620567799,
+      "learning_rate": 9.910517678416097e-06,
+      "loss": 0.4393,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08869015875620462,
+      "grad_norm": 0.7658878665588302,
+      "learning_rate": 9.910267254097169e-06,
+      "loss": 0.4518,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08877220330639538,
+      "grad_norm": 0.7657366768629393,
+      "learning_rate": 9.910016483023058e-06,
+      "loss": 0.4247,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08885424785658613,
+      "grad_norm": 0.6868276727542265,
+      "learning_rate": 9.90976536521147e-06,
+      "loss": 0.399,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08893629240677688,
+      "grad_norm": 0.7435052617403838,
+      "learning_rate": 9.909513900680141e-06,
+      "loss": 0.4265,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08901833695696763,
+      "grad_norm": 0.7151618488531751,
+      "learning_rate": 9.909262089446828e-06,
+      "loss": 0.4304,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08910038150715839,
+      "grad_norm": 0.6473982113877929,
+      "learning_rate": 9.909009931529313e-06,
+      "loss": 0.4006,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08918242605734913,
+      "grad_norm": 0.6936941613236083,
+      "learning_rate": 9.908757426945403e-06,
+      "loss": 0.4082,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08926447060753989,
+      "grad_norm": 0.7060797548381823,
+      "learning_rate": 9.908504575712932e-06,
+      "loss": 0.4244,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08934651515773065,
+      "grad_norm": 0.7000405717845052,
+      "learning_rate": 9.908251377849752e-06,
+      "loss": 0.4124,
+      "step": 1089
+    },
+    {
+      "epoch": 0.0894285597079214,
+      "grad_norm": 0.7065352789596681,
+      "learning_rate": 9.907997833373747e-06,
+      "loss": 0.3897,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08951060425811215,
+      "grad_norm": 0.7068961064751251,
+      "learning_rate": 9.907743942302818e-06,
+      "loss": 0.4545,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08959264880830291,
+      "grad_norm": 0.6920685948969771,
+      "learning_rate": 9.9074897046549e-06,
+      "loss": 0.42,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08967469335849367,
+      "grad_norm": 0.6110309719032815,
+      "learning_rate": 9.907235120447942e-06,
+      "loss": 0.4235,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08975673790868442,
+      "grad_norm": 0.6402226610263476,
+      "learning_rate": 9.906980189699924e-06,
+      "loss": 0.4058,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08983878245887517,
+      "grad_norm": 0.7503252325187472,
+      "learning_rate": 9.906724912428848e-06,
+      "loss": 0.45,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08992082700906592,
+      "grad_norm": 0.7198361092421223,
+      "learning_rate": 9.906469288652743e-06,
+      "loss": 0.4432,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09000287155925668,
+      "grad_norm": 0.651812856912975,
+      "learning_rate": 9.906213318389658e-06,
+      "loss": 0.3851,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09008491610944742,
+      "grad_norm": 0.6950469907652165,
+      "learning_rate": 9.905957001657673e-06,
+      "loss": 0.43,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09016696065963818,
+      "grad_norm": 0.6822615634633077,
+      "learning_rate": 9.905700338474885e-06,
+      "loss": 0.4734,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09024900520982894,
+      "grad_norm": 0.7505811462089138,
+      "learning_rate": 9.905443328859421e-06,
+      "loss": 0.3991,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0903310497600197,
+      "grad_norm": 0.6829065042436095,
+      "learning_rate": 9.905185972829429e-06,
+      "loss": 0.3894,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09041309431021044,
+      "grad_norm": 0.6323049748513924,
+      "learning_rate": 9.904928270403087e-06,
+      "loss": 0.4165,
+      "step": 1102
+    },
+    {
+      "epoch": 0.0904951388604012,
+      "grad_norm": 0.680080121790264,
+      "learning_rate": 9.90467022159859e-06,
+      "loss": 0.4621,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09057718341059195,
+      "grad_norm": 0.7115796796423162,
+      "learning_rate": 9.904411826434161e-06,
+      "loss": 0.4434,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0906592279607827,
+      "grad_norm": 0.6722889074622413,
+      "learning_rate": 9.90415308492805e-06,
+      "loss": 0.4298,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09074127251097346,
+      "grad_norm": 0.7095079143746601,
+      "learning_rate": 9.903893997098526e-06,
+      "loss": 0.4472,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09082331706116421,
+      "grad_norm": 0.6877716800418657,
+      "learning_rate": 9.903634562963886e-06,
+      "loss": 0.4006,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09090536161135497,
+      "grad_norm": 0.6384142549447768,
+      "learning_rate": 9.903374782542454e-06,
+      "loss": 0.4086,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09098740616154571,
+      "grad_norm": 0.7006394616747653,
+      "learning_rate": 9.90311465585257e-06,
+      "loss": 0.3959,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09106945071173647,
+      "grad_norm": 0.7363399567599227,
+      "learning_rate": 9.902854182912608e-06,
+      "loss": 0.4789,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09115149526192723,
+      "grad_norm": 0.6468754383218632,
+      "learning_rate": 9.90259336374096e-06,
+      "loss": 0.4408,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09123353981211799,
+      "grad_norm": 0.651911083067125,
+      "learning_rate": 9.902332198356045e-06,
+      "loss": 0.4248,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09131558436230873,
+      "grad_norm": 0.6690591422983134,
+      "learning_rate": 9.902070686776308e-06,
+      "loss": 0.4174,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09139762891249949,
+      "grad_norm": 0.6571093204348416,
+      "learning_rate": 9.901808829020215e-06,
+      "loss": 0.4102,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09147967346269024,
+      "grad_norm": 0.6518274895359748,
+      "learning_rate": 9.901546625106255e-06,
+      "loss": 0.4306,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09156171801288099,
+      "grad_norm": 0.6561174758492045,
+      "learning_rate": 9.90128407505295e-06,
+      "loss": 0.4108,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09164376256307175,
+      "grad_norm": 0.7336284310952887,
+      "learning_rate": 9.901021178878835e-06,
+      "loss": 0.4815,
+      "step": 1117
+    },
+    {
+      "epoch": 0.0917258071132625,
+      "grad_norm": 0.6305698193131831,
+      "learning_rate": 9.900757936602481e-06,
+      "loss": 0.4007,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09180785166345326,
+      "grad_norm": 0.6293196847557636,
+      "learning_rate": 9.900494348242475e-06,
+      "loss": 0.3909,
+      "step": 1119
+    },
+    {
+      "epoch": 0.091889896213644,
+      "grad_norm": 0.6997967076587683,
+      "learning_rate": 9.90023041381743e-06,
+      "loss": 0.437,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09197194076383476,
+      "grad_norm": 0.6832521161089257,
+      "learning_rate": 9.899966133345988e-06,
+      "loss": 0.3918,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09205398531402552,
+      "grad_norm": 0.6461341985016275,
+      "learning_rate": 9.89970150684681e-06,
+      "loss": 0.4141,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09213602986421628,
+      "grad_norm": 0.6717038739884312,
+      "learning_rate": 9.899436534338582e-06,
+      "loss": 0.4038,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09221807441440702,
+      "grad_norm": 0.6566158609680398,
+      "learning_rate": 9.899171215840018e-06,
+      "loss": 0.4306,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09230011896459778,
+      "grad_norm": 0.6679261544639572,
+      "learning_rate": 9.898905551369854e-06,
+      "loss": 0.4886,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09238216351478853,
+      "grad_norm": 0.6540523649288594,
+      "learning_rate": 9.89863954094685e-06,
+      "loss": 0.4054,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09246420806497928,
+      "grad_norm": 0.7123965381153851,
+      "learning_rate": 9.898373184589793e-06,
+      "loss": 0.406,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09254625261517004,
+      "grad_norm": 0.9410553449851357,
+      "learning_rate": 9.898106482317491e-06,
+      "loss": 0.4265,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09262829716536079,
+      "grad_norm": 0.6315526830561401,
+      "learning_rate": 9.897839434148779e-06,
+      "loss": 0.4332,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09271034171555155,
+      "grad_norm": 0.64566449805873,
+      "learning_rate": 9.897572040102514e-06,
+      "loss": 0.3979,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0927923862657423,
+      "grad_norm": 0.6902002634118973,
+      "learning_rate": 9.89730430019758e-06,
+      "loss": 0.4131,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09287443081593305,
+      "grad_norm": 0.6604982349479028,
+      "learning_rate": 9.897036214452886e-06,
+      "loss": 0.4281,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09295647536612381,
+      "grad_norm": 0.7318828477310675,
+      "learning_rate": 9.896767782887362e-06,
+      "loss": 0.4215,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09303851991631455,
+      "grad_norm": 0.6330458481580741,
+      "learning_rate": 9.896499005519964e-06,
+      "loss": 0.4172,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09312056446650531,
+      "grad_norm": 0.6927238723795207,
+      "learning_rate": 9.896229882369674e-06,
+      "loss": 0.4224,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09320260901669607,
+      "grad_norm": 0.649251949907775,
+      "learning_rate": 9.895960413455495e-06,
+      "loss": 0.4138,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09328465356688682,
+      "grad_norm": 0.6445516325088413,
+      "learning_rate": 9.895690598796457e-06,
+      "loss": 0.4551,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09336669811707757,
+      "grad_norm": 0.752038870638604,
+      "learning_rate": 9.895420438411616e-06,
+      "loss": 0.4501,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09344874266726833,
+      "grad_norm": 0.6952185683723591,
+      "learning_rate": 9.895149932320048e-06,
+      "loss": 0.4306,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09353078721745908,
+      "grad_norm": 0.7176151175432689,
+      "learning_rate": 9.894879080540856e-06,
+      "loss": 0.4106,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09361283176764984,
+      "grad_norm": 0.628441444353533,
+      "learning_rate": 9.894607883093169e-06,
+      "loss": 0.4344,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09369487631784058,
+      "grad_norm": 0.7381690220771923,
+      "learning_rate": 9.894336339996133e-06,
+      "loss": 0.4168,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09377692086803134,
+      "grad_norm": 0.6450215709130285,
+      "learning_rate": 9.894064451268933e-06,
+      "loss": 0.4015,
+      "step": 1143
+    },
+    {
+      "epoch": 0.0938589654182221,
+      "grad_norm": 0.6903120987355906,
+      "learning_rate": 9.893792216930762e-06,
+      "loss": 0.4122,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09394100996841284,
+      "grad_norm": 0.6353743144411274,
+      "learning_rate": 9.893519637000846e-06,
+      "loss": 0.4192,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0940230545186036,
+      "grad_norm": 0.5969026845530635,
+      "learning_rate": 9.893246711498438e-06,
+      "loss": 0.364,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09410509906879436,
+      "grad_norm": 0.6359288862679797,
+      "learning_rate": 9.892973440442809e-06,
+      "loss": 0.4905,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09418714361898511,
+      "grad_norm": 0.6845632672755336,
+      "learning_rate": 9.892699823853254e-06,
+      "loss": 0.4121,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09426918816917586,
+      "grad_norm": 0.671869539033308,
+      "learning_rate": 9.8924258617491e-06,
+      "loss": 0.3779,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09435123271936662,
+      "grad_norm": 0.6267614607282824,
+      "learning_rate": 9.89215155414969e-06,
+      "loss": 0.4157,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09443327726955737,
+      "grad_norm": 0.7241409693711145,
+      "learning_rate": 9.8918769010744e-06,
+      "loss": 0.4234,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09451532181974813,
+      "grad_norm": 0.6900135158853797,
+      "learning_rate": 9.891601902542621e-06,
+      "loss": 0.4095,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09459736636993887,
+      "grad_norm": 0.6972742051145224,
+      "learning_rate": 9.891326558573774e-06,
+      "loss": 0.4339,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09467941092012963,
+      "grad_norm": 0.6356201051062806,
+      "learning_rate": 9.891050869187305e-06,
+      "loss": 0.4095,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09476145547032039,
+      "grad_norm": 0.6596740323744534,
+      "learning_rate": 9.89077483440268e-06,
+      "loss": 0.4307,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09484350002051113,
+      "grad_norm": 0.6775343311021543,
+      "learning_rate": 9.890498454239396e-06,
+      "loss": 0.4165,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09492554457070189,
+      "grad_norm": 0.6562848619162858,
+      "learning_rate": 9.89022172871697e-06,
+      "loss": 0.3974,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09500758912089265,
+      "grad_norm": 0.6299477003946772,
+      "learning_rate": 9.889944657854937e-06,
+      "loss": 0.4112,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0950896336710834,
+      "grad_norm": 0.7317256411980955,
+      "learning_rate": 9.889667241672872e-06,
+      "loss": 0.4244,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09517167822127415,
+      "grad_norm": 0.7601706411362783,
+      "learning_rate": 9.889389480190361e-06,
+      "loss": 0.4017,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0952537227714649,
+      "grad_norm": 0.6806303852314602,
+      "learning_rate": 9.889111373427021e-06,
+      "loss": 0.446,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09533576732165566,
+      "grad_norm": 0.6397863795597569,
+      "learning_rate": 9.888832921402491e-06,
+      "loss": 0.4371,
+      "step": 1162
+    },
+    {
+      "epoch": 0.0954178118718464,
+      "grad_norm": 0.6946839682036868,
+      "learning_rate": 9.888554124136436e-06,
+      "loss": 0.4147,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09549985642203716,
+      "grad_norm": 0.6542115500813515,
+      "learning_rate": 9.88827498164854e-06,
+      "loss": 0.4094,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09558190097222792,
+      "grad_norm": 0.6603906768724794,
+      "learning_rate": 9.887995493958519e-06,
+      "loss": 0.4558,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09566394552241868,
+      "grad_norm": 0.6075697495063458,
+      "learning_rate": 9.88771566108611e-06,
+      "loss": 0.4243,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09574599007260942,
+      "grad_norm": 0.6880008782152812,
+      "learning_rate": 9.887435483051074e-06,
+      "loss": 0.4392,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09582803462280018,
+      "grad_norm": 0.7317805499474413,
+      "learning_rate": 9.887154959873196e-06,
+      "loss": 0.4381,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09591007917299094,
+      "grad_norm": 0.6771479150337498,
+      "learning_rate": 9.886874091572287e-06,
+      "loss": 0.4285,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0959921237231817,
+      "grad_norm": 0.7051578388913949,
+      "learning_rate": 9.88659287816818e-06,
+      "loss": 0.4417,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09607416827337244,
+      "grad_norm": 0.6506299000712961,
+      "learning_rate": 9.886311319680736e-06,
+      "loss": 0.4003,
+      "step": 1171
+    },
+    {
+      "epoch": 0.0961562128235632,
+      "grad_norm": 0.6609675888923994,
+      "learning_rate": 9.886029416129837e-06,
+      "loss": 0.4143,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09623825737375395,
+      "grad_norm": 0.6666426904436271,
+      "learning_rate": 9.88574716753539e-06,
+      "loss": 0.394,
+      "step": 1173
+    },
+    {
+      "epoch": 0.0963203019239447,
+      "grad_norm": 0.6833645393348772,
+      "learning_rate": 9.885464573917328e-06,
+      "loss": 0.4094,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09640234647413545,
+      "grad_norm": 0.6862142813072858,
+      "learning_rate": 9.885181635295606e-06,
+      "loss": 0.4324,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09648439102432621,
+      "grad_norm": 0.6806571001621979,
+      "learning_rate": 9.884898351690206e-06,
+      "loss": 0.4314,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09656643557451697,
+      "grad_norm": 0.6522613530013103,
+      "learning_rate": 9.884614723121133e-06,
+      "loss": 0.3763,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09664848012470771,
+      "grad_norm": 0.7242211194245459,
+      "learning_rate": 9.884330749608417e-06,
+      "loss": 0.3928,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09673052467489847,
+      "grad_norm": 0.6420712561565053,
+      "learning_rate": 9.88404643117211e-06,
+      "loss": 0.3942,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09681256922508923,
+      "grad_norm": 0.7553004772738057,
+      "learning_rate": 9.883761767832289e-06,
+      "loss": 0.4461,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09689461377527998,
+      "grad_norm": 0.6770050373855221,
+      "learning_rate": 9.883476759609061e-06,
+      "loss": 0.3884,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09697665832547073,
+      "grad_norm": 0.7107389845073202,
+      "learning_rate": 9.883191406522547e-06,
+      "loss": 0.4331,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09705870287566148,
+      "grad_norm": 0.6058284775455465,
+      "learning_rate": 9.882905708592905e-06,
+      "loss": 0.3662,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09714074742585224,
+      "grad_norm": 0.6869063871921786,
+      "learning_rate": 9.882619665840306e-06,
+      "loss": 0.4359,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09722279197604299,
+      "grad_norm": 0.7300115430989915,
+      "learning_rate": 9.882333278284949e-06,
+      "loss": 0.4225,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09730483652623374,
+      "grad_norm": 0.6982908587800185,
+      "learning_rate": 9.88204654594706e-06,
+      "loss": 0.3808,
+      "step": 1186
+    },
+    {
+      "epoch": 0.0973868810764245,
+      "grad_norm": 0.7438174129886317,
+      "learning_rate": 9.881759468846887e-06,
+      "loss": 0.4431,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09746892562661526,
+      "grad_norm": 0.7344692522329485,
+      "learning_rate": 9.881472047004706e-06,
+      "loss": 0.4296,
+      "step": 1188
+    },
+    {
+      "epoch": 0.097550970176806,
+      "grad_norm": 0.6566589764090393,
+      "learning_rate": 9.88118428044081e-06,
+      "loss": 0.4135,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09763301472699676,
+      "grad_norm": 0.723748294584925,
+      "learning_rate": 9.880896169175524e-06,
+      "loss": 0.4198,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09771505927718752,
+      "grad_norm": 0.6595411840386957,
+      "learning_rate": 9.880607713229191e-06,
+      "loss": 0.3826,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09779710382737826,
+      "grad_norm": 0.6960174216244738,
+      "learning_rate": 9.880318912622181e-06,
+      "loss": 0.4208,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09787914837756902,
+      "grad_norm": 0.6477158474312824,
+      "learning_rate": 9.880029767374892e-06,
+      "loss": 0.4485,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09796119292775977,
+      "grad_norm": 0.7296410717560162,
+      "learning_rate": 9.87974027750774e-06,
+      "loss": 0.4481,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09804323747795053,
+      "grad_norm": 0.602611060795991,
+      "learning_rate": 9.879450443041172e-06,
+      "loss": 0.422,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09812528202814128,
+      "grad_norm": 0.7155220574958134,
+      "learning_rate": 9.879160263995652e-06,
+      "loss": 0.444,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09820732657833203,
+      "grad_norm": 0.6447404397894257,
+      "learning_rate": 9.878869740391672e-06,
+      "loss": 0.4362,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09828937112852279,
+      "grad_norm": 0.6476696985064385,
+      "learning_rate": 9.87857887224975e-06,
+      "loss": 0.3968,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09837141567871355,
+      "grad_norm": 0.6354096758634412,
+      "learning_rate": 9.878287659590427e-06,
+      "loss": 0.3666,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09845346022890429,
+      "grad_norm": 0.6573453437167557,
+      "learning_rate": 9.877996102434266e-06,
+      "loss": 0.4291,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09853550477909505,
+      "grad_norm": 0.6571449108824587,
+      "learning_rate": 9.877704200801858e-06,
+      "loss": 0.4116,
+      "step": 1201
+    },
+    {
+      "epoch": 0.0986175493292858,
+      "grad_norm": 0.6554833933456761,
+      "learning_rate": 9.877411954713816e-06,
+      "loss": 0.3901,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09869959387947655,
+      "grad_norm": 0.6681084864792033,
+      "learning_rate": 9.877119364190778e-06,
+      "loss": 0.3981,
+      "step": 1203
+    },
+    {
+      "epoch": 0.0987816384296673,
+      "grad_norm": 0.6976653346389601,
+      "learning_rate": 9.876826429253405e-06,
+      "loss": 0.3619,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09886368297985806,
+      "grad_norm": 0.6366165665130963,
+      "learning_rate": 9.876533149922387e-06,
+      "loss": 0.4409,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09894572753004882,
+      "grad_norm": 0.7170383214243727,
+      "learning_rate": 9.876239526218431e-06,
+      "loss": 0.4304,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09902777208023957,
+      "grad_norm": 0.6558353821819263,
+      "learning_rate": 9.875945558162274e-06,
+      "loss": 0.4218,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09910981663043032,
+      "grad_norm": 0.6379859670242897,
+      "learning_rate": 9.875651245774675e-06,
+      "loss": 0.371,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09919186118062108,
+      "grad_norm": 0.6871401756004878,
+      "learning_rate": 9.87535658907642e-06,
+      "loss": 0.4375,
+      "step": 1209
+    },
+    {
+      "epoch": 0.09927390573081184,
+      "grad_norm": 0.679834868080823,
+      "learning_rate": 9.875061588088316e-06,
+      "loss": 0.4037,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09935595028100258,
+      "grad_norm": 0.7068987327506728,
+      "learning_rate": 9.874766242831193e-06,
+      "loss": 0.4299,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09943799483119334,
+      "grad_norm": 0.6433934867282338,
+      "learning_rate": 9.87447055332591e-06,
+      "loss": 0.4601,
+      "step": 1212
+    },
+    {
+      "epoch": 0.0995200393813841,
+      "grad_norm": 0.6561002884617048,
+      "learning_rate": 9.874174519593349e-06,
+      "loss": 0.3687,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09960208393157484,
+      "grad_norm": 0.6710283268048053,
+      "learning_rate": 9.873878141654414e-06,
+      "loss": 0.391,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0996841284817656,
+      "grad_norm": 0.690449917003757,
+      "learning_rate": 9.873581419530035e-06,
+      "loss": 0.4227,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09976617303195635,
+      "grad_norm": 0.6732269491556577,
+      "learning_rate": 9.873284353241166e-06,
+      "loss": 0.3972,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09984821758214711,
+      "grad_norm": 0.6612178590633094,
+      "learning_rate": 9.872986942808784e-06,
+      "loss": 0.3826,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09993026213233785,
+      "grad_norm": 0.6890354739697394,
+      "learning_rate": 9.872689188253895e-06,
+      "loss": 0.4136,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10001230668252861,
+      "grad_norm": 0.6571427659343753,
+      "learning_rate": 9.872391089597523e-06,
+      "loss": 0.3991,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10009435123271937,
+      "grad_norm": 0.6670393080690409,
+      "learning_rate": 9.872092646860723e-06,
+      "loss": 0.3969,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10017639578291013,
+      "grad_norm": 0.6820732034265237,
+      "learning_rate": 9.871793860064566e-06,
+      "loss": 0.4446,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10025844033310087,
+      "grad_norm": 0.6882090178061364,
+      "learning_rate": 9.871494729230153e-06,
+      "loss": 0.4352,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10034048488329163,
+      "grad_norm": 0.6028845346912963,
+      "learning_rate": 9.87119525437861e-06,
+      "loss": 0.4019,
+      "step": 1223
+    },
+    {
+      "epoch": 0.10042252943348239,
+      "grad_norm": 0.6477729322693746,
+      "learning_rate": 9.870895435531086e-06,
+      "loss": 0.4326,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10050457398367313,
+      "grad_norm": 0.6645127741202622,
+      "learning_rate": 9.870595272708749e-06,
+      "loss": 0.432,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10058661853386389,
+      "grad_norm": 0.711517408515484,
+      "learning_rate": 9.8702947659328e-06,
+      "loss": 0.4591,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10066866308405464,
+      "grad_norm": 0.6512117563742844,
+      "learning_rate": 9.869993915224459e-06,
+      "loss": 0.4157,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1007507076342454,
+      "grad_norm": 0.6449312195160961,
+      "learning_rate": 9.869692720604973e-06,
+      "loss": 0.4189,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10083275218443614,
+      "grad_norm": 0.6481266433904177,
+      "learning_rate": 9.86939118209561e-06,
+      "loss": 0.4308,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1009147967346269,
+      "grad_norm": 0.6737536916479572,
+      "learning_rate": 9.869089299717665e-06,
+      "loss": 0.4255,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10099684128481766,
+      "grad_norm": 0.6957841509529012,
+      "learning_rate": 9.868787073492457e-06,
+      "loss": 0.4291,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1010788858350084,
+      "grad_norm": 0.65006099699815,
+      "learning_rate": 9.868484503441329e-06,
+      "loss": 0.4439,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10116093038519916,
+      "grad_norm": 0.991419344591193,
+      "learning_rate": 9.868181589585644e-06,
+      "loss": 0.7858,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10124297493538992,
+      "grad_norm": 0.685237521233886,
+      "learning_rate": 9.8678783319468e-06,
+      "loss": 0.3862,
+      "step": 1234
+    },
+    {
+      "epoch": 0.10132501948558068,
+      "grad_norm": 0.6602605502174269,
+      "learning_rate": 9.867574730546207e-06,
+      "loss": 0.4262,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10140706403577142,
+      "grad_norm": 0.6771685236067592,
+      "learning_rate": 9.867270785405308e-06,
+      "loss": 0.3976,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10148910858596218,
+      "grad_norm": 0.715946236397249,
+      "learning_rate": 9.866966496545565e-06,
+      "loss": 0.4139,
+      "step": 1237
+    },
+    {
+      "epoch": 0.10157115313615293,
+      "grad_norm": 0.7522121805778265,
+      "learning_rate": 9.866661863988466e-06,
+      "loss": 0.43,
+      "step": 1238
+    },
+    {
+      "epoch": 0.10165319768634369,
+      "grad_norm": 0.6479778470336959,
+      "learning_rate": 9.866356887755529e-06,
+      "loss": 0.3915,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10173524223653443,
+      "grad_norm": 0.6669648557311907,
+      "learning_rate": 9.866051567868284e-06,
+      "loss": 0.3823,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10181728678672519,
+      "grad_norm": 0.6340208922718302,
+      "learning_rate": 9.865745904348296e-06,
+      "loss": 0.4111,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10189933133691595,
+      "grad_norm": 0.6704838790522455,
+      "learning_rate": 9.865439897217151e-06,
+      "loss": 0.42,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10198137588710669,
+      "grad_norm": 0.6852840341495134,
+      "learning_rate": 9.865133546496455e-06,
+      "loss": 0.4024,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10206342043729745,
+      "grad_norm": 0.6786879062856247,
+      "learning_rate": 9.864826852207846e-06,
+      "loss": 0.4576,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10214546498748821,
+      "grad_norm": 0.6719620520994852,
+      "learning_rate": 9.86451981437298e-06,
+      "loss": 0.4398,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10222750953767896,
+      "grad_norm": 0.6665145178916769,
+      "learning_rate": 9.86421243301354e-06,
+      "loss": 0.4325,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10230955408786971,
+      "grad_norm": 0.6511827504146804,
+      "learning_rate": 9.863904708151233e-06,
+      "loss": 0.4072,
+      "step": 1247
+    },
+    {
+      "epoch": 0.10239159863806047,
+      "grad_norm": 0.629715654295075,
+      "learning_rate": 9.86359663980779e-06,
+      "loss": 0.4183,
+      "step": 1248
+    },
+    {
+      "epoch": 0.10247364318825122,
+      "grad_norm": 0.6412015739919525,
+      "learning_rate": 9.863288228004967e-06,
+      "loss": 0.3808,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10255568773844198,
+      "grad_norm": 0.6525704759376479,
+      "learning_rate": 9.862979472764542e-06,
+      "loss": 0.4265,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10263773228863272,
+      "grad_norm": 0.6255580689886147,
+      "learning_rate": 9.862670374108321e-06,
+      "loss": 0.3877,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10271977683882348,
+      "grad_norm": 0.6667134140316704,
+      "learning_rate": 9.862360932058128e-06,
+      "loss": 0.3935,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10280182138901424,
+      "grad_norm": 0.7080637942592976,
+      "learning_rate": 9.86205114663582e-06,
+      "loss": 0.4426,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10288386593920498,
+      "grad_norm": 0.6990533234173746,
+      "learning_rate": 9.86174101786327e-06,
+      "loss": 0.4544,
+      "step": 1254
+    },
+    {
+      "epoch": 0.10296591048939574,
+      "grad_norm": 0.6384122378643747,
+      "learning_rate": 9.861430545762383e-06,
+      "loss": 0.3719,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1030479550395865,
+      "grad_norm": 0.6706019009735944,
+      "learning_rate": 9.86111973035508e-06,
+      "loss": 0.3907,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10312999958977725,
+      "grad_norm": 0.6680371887153953,
+      "learning_rate": 9.860808571663311e-06,
+      "loss": 0.4062,
+      "step": 1257
+    },
+    {
+      "epoch": 0.103212044139968,
+      "grad_norm": 0.6215450697406733,
+      "learning_rate": 9.86049706970905e-06,
+      "loss": 0.4065,
+      "step": 1258
+    },
+    {
+      "epoch": 0.10329408869015876,
+      "grad_norm": 0.6080051932536694,
+      "learning_rate": 9.860185224514295e-06,
+      "loss": 0.4118,
+      "step": 1259
+    },
+    {
+      "epoch": 0.10337613324034951,
+      "grad_norm": 0.6553748716376908,
+      "learning_rate": 9.859873036101069e-06,
+      "loss": 0.4448,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10345817779054026,
+      "grad_norm": 0.6770991146202986,
+      "learning_rate": 9.859560504491417e-06,
+      "loss": 0.4304,
+      "step": 1261
+    },
+    {
+      "epoch": 0.10354022234073101,
+      "grad_norm": 0.6151082114324095,
+      "learning_rate": 9.859247629707407e-06,
+      "loss": 0.4064,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10362226689092177,
+      "grad_norm": 0.6661284667596279,
+      "learning_rate": 9.85893441177114e-06,
+      "loss": 0.4086,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10370431144111253,
+      "grad_norm": 0.7394825309385502,
+      "learning_rate": 9.85862085070473e-06,
+      "loss": 0.4071,
+      "step": 1264
+    },
+    {
+      "epoch": 0.10378635599130327,
+      "grad_norm": 0.7454196505376599,
+      "learning_rate": 9.858306946530321e-06,
+      "loss": 0.44,
+      "step": 1265
+    },
+    {
+      "epoch": 0.10386840054149403,
+      "grad_norm": 0.7613870180313426,
+      "learning_rate": 9.85799269927008e-06,
+      "loss": 0.4294,
+      "step": 1266
+    },
+    {
+      "epoch": 0.10395044509168479,
+      "grad_norm": 0.6909022106121174,
+      "learning_rate": 9.857678108946201e-06,
+      "loss": 0.4128,
+      "step": 1267
+    },
+    {
+      "epoch": 0.10403248964187554,
+      "grad_norm": 0.641527562411698,
+      "learning_rate": 9.8573631755809e-06,
+      "loss": 0.3834,
+      "step": 1268
+    },
+    {
+      "epoch": 0.10411453419206629,
+      "grad_norm": 0.6348015725059497,
+      "learning_rate": 9.857047899196412e-06,
+      "loss": 0.4202,
+      "step": 1269
+    },
+    {
+      "epoch": 0.10419657874225705,
+      "grad_norm": 0.6494075450006449,
+      "learning_rate": 9.856732279815007e-06,
+      "loss": 0.4934,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1042786232924478,
+      "grad_norm": 0.7224195477751449,
+      "learning_rate": 9.85641631745897e-06,
+      "loss": 0.431,
+      "step": 1271
+    },
+    {
+      "epoch": 0.10436066784263855,
+      "grad_norm": 0.652437951040799,
+      "learning_rate": 9.856100012150615e-06,
+      "loss": 0.4091,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1044427123928293,
+      "grad_norm": 0.6565178724976299,
+      "learning_rate": 9.855783363912281e-06,
+      "loss": 0.42,
+      "step": 1273
+    },
+    {
+      "epoch": 0.10452475694302006,
+      "grad_norm": 0.6297142561708148,
+      "learning_rate": 9.855466372766326e-06,
+      "loss": 0.4276,
+      "step": 1274
+    },
+    {
+      "epoch": 0.10460680149321082,
+      "grad_norm": 0.6964022615209245,
+      "learning_rate": 9.855149038735137e-06,
+      "loss": 0.4624,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10468884604340156,
+      "grad_norm": 0.6966913787348221,
+      "learning_rate": 9.854831361841124e-06,
+      "loss": 0.4415,
+      "step": 1276
+    },
+    {
+      "epoch": 0.10477089059359232,
+      "grad_norm": 0.7304323456578327,
+      "learning_rate": 9.854513342106718e-06,
+      "loss": 0.4001,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10485293514378308,
+      "grad_norm": 0.6498924763461473,
+      "learning_rate": 9.85419497955438e-06,
+      "loss": 0.3874,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10493497969397383,
+      "grad_norm": 0.7295932212681431,
+      "learning_rate": 9.853876274206593e-06,
+      "loss": 0.4051,
+      "step": 1279
+    },
+    {
+      "epoch": 0.10501702424416458,
+      "grad_norm": 0.6862922382764395,
+      "learning_rate": 9.853557226085859e-06,
+      "loss": 0.4057,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10509906879435534,
+      "grad_norm": 0.6431361945119496,
+      "learning_rate": 9.853237835214714e-06,
+      "loss": 0.3884,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10518111334454609,
+      "grad_norm": 0.6506842882466967,
+      "learning_rate": 9.85291810161571e-06,
+      "loss": 0.4255,
+      "step": 1282
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.7019330145392001,
+      "learning_rate": 9.852598025311425e-06,
+      "loss": 0.4209,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1053452024449276,
+      "grad_norm": 0.6741459526914287,
+      "learning_rate": 9.852277606324465e-06,
+      "loss": 0.4151,
+      "step": 1284
+    },
+    {
+      "epoch": 0.10542724699511835,
+      "grad_norm": 0.6847475163729722,
+      "learning_rate": 9.851956844677457e-06,
+      "loss": 0.4396,
+      "step": 1285
+    },
+    {
+      "epoch": 0.10550929154530911,
+      "grad_norm": 0.74698406239371,
+      "learning_rate": 9.85163574039305e-06,
+      "loss": 0.4215,
+      "step": 1286
+    },
+    {
+      "epoch": 0.10559133609549985,
+      "grad_norm": 0.7170528134311511,
+      "learning_rate": 9.851314293493923e-06,
+      "loss": 0.4291,
+      "step": 1287
+    },
+    {
+      "epoch": 0.10567338064569061,
+      "grad_norm": 0.6394577654773472,
+      "learning_rate": 9.850992504002775e-06,
+      "loss": 0.4012,
+      "step": 1288
+    },
+    {
+      "epoch": 0.10575542519588137,
+      "grad_norm": 0.6620277445889368,
+      "learning_rate": 9.850670371942328e-06,
+      "loss": 0.4513,
+      "step": 1289
+    },
+    {
+      "epoch": 0.10583746974607211,
+      "grad_norm": 0.669995025232077,
+      "learning_rate": 9.850347897335335e-06,
+      "loss": 0.4243,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10591951429626287,
+      "grad_norm": 0.67796346245835,
+      "learning_rate": 9.850025080204564e-06,
+      "loss": 0.4261,
+      "step": 1291
+    },
+    {
+      "epoch": 0.10600155884645363,
+      "grad_norm": 0.679456152206682,
+      "learning_rate": 9.849701920572816e-06,
+      "loss": 0.4131,
+      "step": 1292
+    },
+    {
+      "epoch": 0.10608360339664438,
+      "grad_norm": 0.7270584145994878,
+      "learning_rate": 9.84937841846291e-06,
+      "loss": 0.4137,
+      "step": 1293
+    },
+    {
+      "epoch": 0.10616564794683513,
+      "grad_norm": 0.6587811123114931,
+      "learning_rate": 9.84905457389769e-06,
+      "loss": 0.4134,
+      "step": 1294
+    },
+    {
+      "epoch": 0.10624769249702588,
+      "grad_norm": 0.6835403346871266,
+      "learning_rate": 9.848730386900027e-06,
+      "loss": 0.3961,
+      "step": 1295
+    },
+    {
+      "epoch": 0.10632973704721664,
+      "grad_norm": 0.6672687991495331,
+      "learning_rate": 9.848405857492814e-06,
+      "loss": 0.4523,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1064117815974074,
+      "grad_norm": 0.6630087195290532,
+      "learning_rate": 9.84808098569897e-06,
+      "loss": 0.3982,
+      "step": 1297
+    },
+    {
+      "epoch": 0.10649382614759814,
+      "grad_norm": 0.6357095712891924,
+      "learning_rate": 9.847755771541433e-06,
+      "loss": 0.4169,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1065758706977889,
+      "grad_norm": 0.7247209470362783,
+      "learning_rate": 9.847430215043175e-06,
+      "loss": 0.4598,
+      "step": 1299
+    },
+    {
+      "epoch": 0.10665791524797966,
+      "grad_norm": 0.7058027856080104,
+      "learning_rate": 9.847104316227183e-06,
+      "loss": 0.3894,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1067399597981704,
+      "grad_norm": 0.7040991162333894,
+      "learning_rate": 9.84677807511647e-06,
+      "loss": 0.4279,
+      "step": 1301
+    },
+    {
+      "epoch": 0.10682200434836116,
+      "grad_norm": 0.6223403987472522,
+      "learning_rate": 9.846451491734077e-06,
+      "loss": 0.4245,
+      "step": 1302
+    },
+    {
+      "epoch": 0.10690404889855191,
+      "grad_norm": 0.6567983767718362,
+      "learning_rate": 9.846124566103065e-06,
+      "loss": 0.3957,
+      "step": 1303
+    },
+    {
+      "epoch": 0.10698609344874267,
+      "grad_norm": 0.7116349186158616,
+      "learning_rate": 9.845797298246525e-06,
+      "loss": 0.4145,
+      "step": 1304
+    },
+    {
+      "epoch": 0.10706813799893342,
+      "grad_norm": 0.717719604744436,
+      "learning_rate": 9.845469688187562e-06,
+      "loss": 0.4525,
+      "step": 1305
+    },
+    {
+      "epoch": 0.10715018254912417,
+      "grad_norm": 0.6919400645963889,
+      "learning_rate": 9.845141735949317e-06,
+      "loss": 0.4612,
+      "step": 1306
+    },
+    {
+      "epoch": 0.10723222709931493,
+      "grad_norm": 0.6930208961648927,
+      "learning_rate": 9.844813441554945e-06,
+      "loss": 0.4008,
+      "step": 1307
+    },
+    {
+      "epoch": 0.10731427164950569,
+      "grad_norm": 0.6634550288776946,
+      "learning_rate": 9.844484805027631e-06,
+      "loss": 0.3841,
+      "step": 1308
+    },
+    {
+      "epoch": 0.10739631619969643,
+      "grad_norm": 0.7076601535907917,
+      "learning_rate": 9.844155826390585e-06,
+      "loss": 0.4226,
+      "step": 1309
+    },
+    {
+      "epoch": 0.10747836074988719,
+      "grad_norm": 0.6904822351915076,
+      "learning_rate": 9.843826505667038e-06,
+      "loss": 0.4451,
+      "step": 1310
+    },
+    {
+      "epoch": 0.10756040530007795,
+      "grad_norm": 0.6489605443167706,
+      "learning_rate": 9.843496842880242e-06,
+      "loss": 0.3895,
+      "step": 1311
+    },
+    {
+      "epoch": 0.10764244985026869,
+      "grad_norm": 0.6303222507067854,
+      "learning_rate": 9.843166838053484e-06,
+      "loss": 0.3857,
+      "step": 1312
+    },
+    {
+      "epoch": 0.10772449440045945,
+      "grad_norm": 0.7154287113044904,
+      "learning_rate": 9.842836491210065e-06,
+      "loss": 0.4262,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1078065389506502,
+      "grad_norm": 0.6895409665767743,
+      "learning_rate": 9.842505802373311e-06,
+      "loss": 0.418,
+      "step": 1314
+    },
+    {
+      "epoch": 0.10788858350084096,
+      "grad_norm": 0.6091491376286569,
+      "learning_rate": 9.84217477156658e-06,
+      "loss": 0.3705,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1079706280510317,
+      "grad_norm": 0.6397348981708049,
+      "learning_rate": 9.841843398813244e-06,
+      "loss": 0.3948,
+      "step": 1316
+    },
+    {
+      "epoch": 0.10805267260122246,
+      "grad_norm": 0.6435997798655205,
+      "learning_rate": 9.841511684136706e-06,
+      "loss": 0.3851,
+      "step": 1317
+    },
+    {
+      "epoch": 0.10813471715141322,
+      "grad_norm": 0.6225858656333272,
+      "learning_rate": 9.841179627560393e-06,
+      "loss": 0.4232,
+      "step": 1318
+    },
+    {
+      "epoch": 0.10821676170160396,
+      "grad_norm": 0.6611355613557116,
+      "learning_rate": 9.840847229107753e-06,
+      "loss": 0.4113,
+      "step": 1319
+    },
+    {
+      "epoch": 0.10829880625179472,
+      "grad_norm": 0.6900657488877701,
+      "learning_rate": 9.840514488802258e-06,
+      "loss": 0.4134,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10838085080198548,
+      "grad_norm": 0.6223353693937551,
+      "learning_rate": 9.840181406667407e-06,
+      "loss": 0.4062,
+      "step": 1321
+    },
+    {
+      "epoch": 0.10846289535217624,
+      "grad_norm": 0.7053855203942552,
+      "learning_rate": 9.839847982726721e-06,
+      "loss": 0.4567,
+      "step": 1322
+    },
+    {
+      "epoch": 0.10854493990236698,
+      "grad_norm": 0.7136399672823662,
+      "learning_rate": 9.839514217003747e-06,
+      "loss": 0.4257,
+      "step": 1323
+    },
+    {
+      "epoch": 0.10862698445255774,
+      "grad_norm": 0.6816131032185019,
+      "learning_rate": 9.839180109522052e-06,
+      "loss": 0.4449,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1087090290027485,
+      "grad_norm": 0.6773406065022538,
+      "learning_rate": 9.838845660305237e-06,
+      "loss": 0.3941,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10879107355293925,
+      "grad_norm": 0.6324905605010055,
+      "learning_rate": 9.838510869376912e-06,
+      "loss": 0.3877,
+      "step": 1326
+    },
+    {
+      "epoch": 0.10887311810313,
+      "grad_norm": 0.6299390018148128,
+      "learning_rate": 9.838175736760723e-06,
+      "loss": 0.399,
+      "step": 1327
+    },
+    {
+      "epoch": 0.10895516265332075,
+      "grad_norm": 0.6428600305383756,
+      "learning_rate": 9.837840262480339e-06,
+      "loss": 0.3952,
+      "step": 1328
+    },
+    {
+      "epoch": 0.10903720720351151,
+      "grad_norm": 0.6942980175110712,
+      "learning_rate": 9.837504446559445e-06,
+      "loss": 0.4701,
+      "step": 1329
+    },
+    {
+      "epoch": 0.10911925175370225,
+      "grad_norm": 0.7372368353410258,
+      "learning_rate": 9.837168289021762e-06,
+      "loss": 0.4266,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10920129630389301,
+      "grad_norm": 0.6314235493988264,
+      "learning_rate": 9.836831789891024e-06,
+      "loss": 0.3995,
+      "step": 1331
+    },
+    {
+      "epoch": 0.10928334085408377,
+      "grad_norm": 0.6904742011891801,
+      "learning_rate": 9.836494949190997e-06,
+      "loss": 0.4373,
+      "step": 1332
+    },
+    {
+      "epoch": 0.10936538540427453,
+      "grad_norm": 0.6735811760990934,
+      "learning_rate": 9.836157766945467e-06,
+      "loss": 0.4556,
+      "step": 1333
+    },
+    {
+      "epoch": 0.10944742995446527,
+      "grad_norm": 0.628138309740677,
+      "learning_rate": 9.835820243178244e-06,
+      "loss": 0.4144,
+      "step": 1334
+    },
+    {
+      "epoch": 0.10952947450465603,
+      "grad_norm": 0.6419761730822787,
+      "learning_rate": 9.835482377913167e-06,
+      "loss": 0.4063,
+      "step": 1335
+    },
+    {
+      "epoch": 0.10961151905484678,
+      "grad_norm": 0.6711330065977402,
+      "learning_rate": 9.835144171174092e-06,
+      "loss": 0.4115,
+      "step": 1336
+    },
+    {
+      "epoch": 0.10969356360503754,
+      "grad_norm": 0.6619015158581834,
+      "learning_rate": 9.834805622984904e-06,
+      "loss": 0.4009,
+      "step": 1337
+    },
+    {
+      "epoch": 0.10977560815522829,
+      "grad_norm": 0.6665592544440841,
+      "learning_rate": 9.834466733369512e-06,
+      "loss": 0.4162,
+      "step": 1338
+    },
+    {
+      "epoch": 0.10985765270541904,
+      "grad_norm": 0.5772220793684119,
+      "learning_rate": 9.834127502351844e-06,
+      "loss": 0.3527,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1099396972556098,
+      "grad_norm": 0.6885532831897582,
+      "learning_rate": 9.83378792995586e-06,
+      "loss": 0.4148,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11002174180580054,
+      "grad_norm": 0.7317675339355514,
+      "learning_rate": 9.833448016205539e-06,
+      "loss": 0.4123,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1101037863559913,
+      "grad_norm": 0.6467775430831076,
+      "learning_rate": 9.833107761124882e-06,
+      "loss": 0.3912,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11018583090618206,
+      "grad_norm": 0.6428868028500705,
+      "learning_rate": 9.832767164737922e-06,
+      "loss": 0.4104,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11026787545637282,
+      "grad_norm": 0.6357336110694325,
+      "learning_rate": 9.83242622706871e-06,
+      "loss": 0.4309,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11034992000656356,
+      "grad_norm": 0.6170507032570505,
+      "learning_rate": 9.83208494814132e-06,
+      "loss": 0.4038,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11043196455675432,
+      "grad_norm": 0.6742377444630175,
+      "learning_rate": 9.831743327979855e-06,
+      "loss": 0.3893,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11051400910694507,
+      "grad_norm": 0.640384665336136,
+      "learning_rate": 9.831401366608438e-06,
+      "loss": 0.4261,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11059605365713583,
+      "grad_norm": 0.6846455386374044,
+      "learning_rate": 9.831059064051221e-06,
+      "loss": 0.3998,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11067809820732658,
+      "grad_norm": 0.6797656807771957,
+      "learning_rate": 9.830716420332373e-06,
+      "loss": 0.3958,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11076014275751733,
+      "grad_norm": 0.6997086949444151,
+      "learning_rate": 9.830373435476092e-06,
+      "loss": 0.4504,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11084218730770809,
+      "grad_norm": 0.7244951685863529,
+      "learning_rate": 9.8300301095066e-06,
+      "loss": 0.4305,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11092423185789883,
+      "grad_norm": 0.6452638699470127,
+      "learning_rate": 9.829686442448141e-06,
+      "loss": 0.4118,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11100627640808959,
+      "grad_norm": 0.6635011012319026,
+      "learning_rate": 9.829342434324986e-06,
+      "loss": 0.4179,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11108832095828035,
+      "grad_norm": 0.5887538087171902,
+      "learning_rate": 9.828998085161428e-06,
+      "loss": 0.3677,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1111703655084711,
+      "grad_norm": 0.6627385616143883,
+      "learning_rate": 9.828653394981781e-06,
+      "loss": 0.4038,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11125241005866185,
+      "grad_norm": 0.6230572814064329,
+      "learning_rate": 9.828308363810392e-06,
+      "loss": 0.4007,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1113344546088526,
+      "grad_norm": 0.6867762837760197,
+      "learning_rate": 9.827962991671622e-06,
+      "loss": 0.3939,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11141649915904336,
+      "grad_norm": 0.6575349774473709,
+      "learning_rate": 9.827617278589864e-06,
+      "loss": 0.3839,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11149854370923411,
+      "grad_norm": 0.6721876841910309,
+      "learning_rate": 9.827271224589528e-06,
+      "loss": 0.4107,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11158058825942486,
+      "grad_norm": 0.7333397532933614,
+      "learning_rate": 9.826924829695055e-06,
+      "loss": 0.4033,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11166263280961562,
+      "grad_norm": 0.6389627815062873,
+      "learning_rate": 9.826578093930904e-06,
+      "loss": 0.4298,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11174467735980638,
+      "grad_norm": 0.6259215557035291,
+      "learning_rate": 9.826231017321564e-06,
+      "loss": 0.4176,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11182672190999712,
+      "grad_norm": 0.7102221340638504,
+      "learning_rate": 9.825883599891543e-06,
+      "loss": 0.4517,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11190876646018788,
+      "grad_norm": 0.6408154444543539,
+      "learning_rate": 9.825535841665374e-06,
+      "loss": 0.4415,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11199081101037864,
+      "grad_norm": 0.6245098562365412,
+      "learning_rate": 9.825187742667619e-06,
+      "loss": 0.4046,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1120728555605694,
+      "grad_norm": 0.6551624449826411,
+      "learning_rate": 9.824839302922858e-06,
+      "loss": 0.4143,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11215490011076014,
+      "grad_norm": 0.6376963090165328,
+      "learning_rate": 9.824490522455695e-06,
+      "loss": 0.3838,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1122369446609509,
+      "grad_norm": 0.6749409300050256,
+      "learning_rate": 9.824141401290763e-06,
+      "loss": 0.4216,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11231898921114165,
+      "grad_norm": 0.6496808878233459,
+      "learning_rate": 9.823791939452717e-06,
+      "loss": 0.3613,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1124010337613324,
+      "grad_norm": 0.6177556392544961,
+      "learning_rate": 9.823442136966234e-06,
+      "loss": 0.4004,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11248307831152315,
+      "grad_norm": 0.653047177829386,
+      "learning_rate": 9.823091993856015e-06,
+      "loss": 0.4003,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11256512286171391,
+      "grad_norm": 0.6656525769510159,
+      "learning_rate": 9.82274151014679e-06,
+      "loss": 0.3979,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11264716741190467,
+      "grad_norm": 0.6715476297560008,
+      "learning_rate": 9.822390685863308e-06,
+      "loss": 0.4233,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11272921196209541,
+      "grad_norm": 0.6703898732264949,
+      "learning_rate": 9.822039521030342e-06,
+      "loss": 0.4037,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11281125651228617,
+      "grad_norm": 0.6647970689150947,
+      "learning_rate": 9.821688015672695e-06,
+      "loss": 0.3937,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11289330106247693,
+      "grad_norm": 0.6394261978115612,
+      "learning_rate": 9.821336169815185e-06,
+      "loss": 0.4208,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11297534561266769,
+      "grad_norm": 0.6379846953967013,
+      "learning_rate": 9.82098398348266e-06,
+      "loss": 0.4107,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11305739016285843,
+      "grad_norm": 0.6124335939708841,
+      "learning_rate": 9.820631456699992e-06,
+      "loss": 0.4129,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11313943471304919,
+      "grad_norm": 0.6813489952821664,
+      "learning_rate": 9.820278589492076e-06,
+      "loss": 0.3802,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11322147926323994,
+      "grad_norm": 0.6049032244510816,
+      "learning_rate": 9.819925381883831e-06,
+      "loss": 0.3946,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11330352381343069,
+      "grad_norm": 0.7205239851034201,
+      "learning_rate": 9.819571833900198e-06,
+      "loss": 0.4383,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11338556836362144,
+      "grad_norm": 0.6731353993097979,
+      "learning_rate": 9.819217945566146e-06,
+      "loss": 0.4588,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1134676129138122,
+      "grad_norm": 0.6701798598264213,
+      "learning_rate": 9.818863716906666e-06,
+      "loss": 0.3678,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11354965746400296,
+      "grad_norm": 0.6712311589289186,
+      "learning_rate": 9.818509147946771e-06,
+      "loss": 0.442,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1136317020141937,
+      "grad_norm": 0.6656612601713228,
+      "learning_rate": 9.818154238711504e-06,
+      "loss": 0.3614,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11371374656438446,
+      "grad_norm": 0.7143717432953391,
+      "learning_rate": 9.817798989225922e-06,
+      "loss": 0.4206,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11379579111457522,
+      "grad_norm": 0.6422093736534835,
+      "learning_rate": 9.817443399515119e-06,
+      "loss": 0.4168,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11387783566476596,
+      "grad_norm": 0.6235270697102051,
+      "learning_rate": 9.817087469604201e-06,
+      "loss": 0.4319,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11395988021495672,
+      "grad_norm": 0.6222532454603311,
+      "learning_rate": 9.816731199518306e-06,
+      "loss": 0.4585,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11404192476514748,
+      "grad_norm": 0.6547530439738983,
+      "learning_rate": 9.816374589282592e-06,
+      "loss": 0.3869,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11412396931533823,
+      "grad_norm": 0.6163793568942376,
+      "learning_rate": 9.816017638922243e-06,
+      "loss": 0.3692,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11420601386552898,
+      "grad_norm": 0.6185712261935417,
+      "learning_rate": 9.815660348462465e-06,
+      "loss": 0.4116,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11428805841571973,
+      "grad_norm": 0.6696028205335488,
+      "learning_rate": 9.815302717928492e-06,
+      "loss": 0.4231,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11437010296591049,
+      "grad_norm": 0.6730500639621144,
+      "learning_rate": 9.814944747345576e-06,
+      "loss": 0.4029,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11445214751610125,
+      "grad_norm": 0.6578081932061403,
+      "learning_rate": 9.814586436738998e-06,
+      "loss": 0.3758,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11453419206629199,
+      "grad_norm": 0.658920807053115,
+      "learning_rate": 9.81422778613406e-06,
+      "loss": 0.3671,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11461623661648275,
+      "grad_norm": 0.6693453606440944,
+      "learning_rate": 9.813868795556092e-06,
+      "loss": 0.4348,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11469828116667351,
+      "grad_norm": 0.6832477109511588,
+      "learning_rate": 9.813509465030442e-06,
+      "loss": 0.4268,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11478032571686425,
+      "grad_norm": 0.6461336999545818,
+      "learning_rate": 9.813149794582489e-06,
+      "loss": 0.4322,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11486237026705501,
+      "grad_norm": 0.688336361764318,
+      "learning_rate": 9.812789784237628e-06,
+      "loss": 0.4207,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11494441481724577,
+      "grad_norm": 0.6078744749811151,
+      "learning_rate": 9.812429434021289e-06,
+      "loss": 0.4233,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11502645936743652,
+      "grad_norm": 0.6287283053473729,
+      "learning_rate": 9.812068743958912e-06,
+      "loss": 0.4001,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11510850391762727,
+      "grad_norm": 0.6520943351966546,
+      "learning_rate": 9.811707714075971e-06,
+      "loss": 0.4051,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11519054846781802,
+      "grad_norm": 0.6333914700691294,
+      "learning_rate": 9.811346344397963e-06,
+      "loss": 0.3892,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11527259301800878,
+      "grad_norm": 0.6776611372803375,
+      "learning_rate": 9.810984634950406e-06,
+      "loss": 0.4501,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11535463756819954,
+      "grad_norm": 0.6677028818057367,
+      "learning_rate": 9.810622585758844e-06,
+      "loss": 0.4473,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11543668211839028,
+      "grad_norm": 0.6388580191757065,
+      "learning_rate": 9.810260196848841e-06,
+      "loss": 0.3942,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11551872666858104,
+      "grad_norm": 0.6693782373647202,
+      "learning_rate": 9.809897468245994e-06,
+      "loss": 0.4678,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1156007712187718,
+      "grad_norm": 0.6823064610913764,
+      "learning_rate": 9.809534399975916e-06,
+      "loss": 0.4456,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11568281576896254,
+      "grad_norm": 0.6713905970540474,
+      "learning_rate": 9.809170992064245e-06,
+      "loss": 0.4172,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1157648603191533,
+      "grad_norm": 0.7041533246712243,
+      "learning_rate": 9.808807244536644e-06,
+      "loss": 0.4536,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11584690486934406,
+      "grad_norm": 0.6731651622997384,
+      "learning_rate": 9.808443157418803e-06,
+      "loss": 0.4421,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11592894941953481,
+      "grad_norm": 0.6840302780175814,
+      "learning_rate": 9.80807873073643e-06,
+      "loss": 0.4346,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11601099396972556,
+      "grad_norm": 0.6310584960138703,
+      "learning_rate": 9.807713964515263e-06,
+      "loss": 0.3943,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11609303851991631,
+      "grad_norm": 0.6456573399494864,
+      "learning_rate": 9.80734885878106e-06,
+      "loss": 0.402,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11617508307010707,
+      "grad_norm": 0.6534126002850384,
+      "learning_rate": 9.806983413559603e-06,
+      "loss": 0.3841,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11625712762029781,
+      "grad_norm": 0.723982129382008,
+      "learning_rate": 9.806617628876701e-06,
+      "loss": 0.4793,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11633917217048857,
+      "grad_norm": 0.7025621449068458,
+      "learning_rate": 9.806251504758184e-06,
+      "loss": 0.4066,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11642121672067933,
+      "grad_norm": 0.6220652795533083,
+      "learning_rate": 9.805885041229909e-06,
+      "loss": 0.384,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11650326127087009,
+      "grad_norm": 0.6515472402253747,
+      "learning_rate": 9.805518238317753e-06,
+      "loss": 0.3838,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11658530582106083,
+      "grad_norm": 0.6491221922491732,
+      "learning_rate": 9.805151096047619e-06,
+      "loss": 0.3793,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11666735037125159,
+      "grad_norm": 0.6596839103435707,
+      "learning_rate": 9.804783614445435e-06,
+      "loss": 0.4167,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11674939492144235,
+      "grad_norm": 0.6512736875104639,
+      "learning_rate": 9.804415793537152e-06,
+      "loss": 0.4025,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1168314394716331,
+      "grad_norm": 0.6386226796780061,
+      "learning_rate": 9.804047633348744e-06,
+      "loss": 0.408,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11691348402182385,
+      "grad_norm": 0.7183694068179068,
+      "learning_rate": 9.80367913390621e-06,
+      "loss": 0.4222,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1169955285720146,
+      "grad_norm": 0.6616504260315532,
+      "learning_rate": 9.803310295235576e-06,
+      "loss": 0.412,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11707757312220536,
+      "grad_norm": 0.6214319187062196,
+      "learning_rate": 9.802941117362882e-06,
+      "loss": 0.4239,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1171596176723961,
+      "grad_norm": 0.6261548881913358,
+      "learning_rate": 9.802571600314205e-06,
+      "loss": 0.3967,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11724166222258686,
+      "grad_norm": 0.6499442118833935,
+      "learning_rate": 9.802201744115637e-06,
+      "loss": 0.4459,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11732370677277762,
+      "grad_norm": 0.6827251354293176,
+      "learning_rate": 9.801831548793298e-06,
+      "loss": 0.4007,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11740575132296838,
+      "grad_norm": 0.6923666467800703,
+      "learning_rate": 9.801461014373328e-06,
+      "loss": 0.3917,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11748779587315912,
+      "grad_norm": 0.7114080587856705,
+      "learning_rate": 9.801090140881897e-06,
+      "loss": 0.4248,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11756984042334988,
+      "grad_norm": 0.6687261576880416,
+      "learning_rate": 9.800718928345195e-06,
+      "loss": 0.4131,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11765188497354064,
+      "grad_norm": 0.6665655245249072,
+      "learning_rate": 9.800347376789432e-06,
+      "loss": 0.3988,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11773392952373139,
+      "grad_norm": 0.7001700538370936,
+      "learning_rate": 9.799975486240852e-06,
+      "loss": 0.4294,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11781597407392214,
+      "grad_norm": 0.637932426326832,
+      "learning_rate": 9.799603256725713e-06,
+      "loss": 0.4121,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1178980186241129,
+      "grad_norm": 0.6733802201200529,
+      "learning_rate": 9.799230688270304e-06,
+      "loss": 0.3966,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11798006317430365,
+      "grad_norm": 0.6587022034130628,
+      "learning_rate": 9.798857780900936e-06,
+      "loss": 0.4052,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1180621077244944,
+      "grad_norm": 0.6823639356109784,
+      "learning_rate": 9.79848453464394e-06,
+      "loss": 0.4305,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11814415227468515,
+      "grad_norm": 0.648886258147364,
+      "learning_rate": 9.798110949525675e-06,
+      "loss": 0.3806,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11822619682487591,
+      "grad_norm": 0.6495668577154488,
+      "learning_rate": 9.797737025572525e-06,
+      "loss": 0.4062,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11830824137506667,
+      "grad_norm": 0.6336336387035173,
+      "learning_rate": 9.797362762810892e-06,
+      "loss": 0.4288,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11839028592525741,
+      "grad_norm": 0.6213891744592384,
+      "learning_rate": 9.79698816126721e-06,
+      "loss": 0.3882,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11847233047544817,
+      "grad_norm": 0.6907357036173748,
+      "learning_rate": 9.79661322096793e-06,
+      "loss": 0.3889,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11855437502563893,
+      "grad_norm": 0.6924070650974781,
+      "learning_rate": 9.796237941939532e-06,
+      "loss": 0.4061,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11863641957582968,
+      "grad_norm": 0.665337839817549,
+      "learning_rate": 9.795862324208516e-06,
+      "loss": 0.4402,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11871846412602043,
+      "grad_norm": 0.6928690713668324,
+      "learning_rate": 9.795486367801407e-06,
+      "loss": 0.3957,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11880050867621118,
+      "grad_norm": 0.6779588140567485,
+      "learning_rate": 9.795110072744756e-06,
+      "loss": 0.4455,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11888255322640194,
+      "grad_norm": 0.6864053047353782,
+      "learning_rate": 9.794733439065135e-06,
+      "loss": 0.4089,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11896459777659268,
+      "grad_norm": 0.701251921500486,
+      "learning_rate": 9.794356466789143e-06,
+      "loss": 0.3751,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11904664232678344,
+      "grad_norm": 0.6246536266412346,
+      "learning_rate": 9.793979155943398e-06,
+      "loss": 0.4091,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1191286868769742,
+      "grad_norm": 0.7385480238135778,
+      "learning_rate": 9.793601506554549e-06,
+      "loss": 0.4156,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11921073142716496,
+      "grad_norm": 0.686089786193087,
+      "learning_rate": 9.793223518649262e-06,
+      "loss": 0.4013,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1192927759773557,
+      "grad_norm": 0.6669097158424796,
+      "learning_rate": 9.792845192254231e-06,
+      "loss": 0.3626,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11937482052754646,
+      "grad_norm": 0.669275619138824,
+      "learning_rate": 9.792466527396174e-06,
+      "loss": 0.3675,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11945686507773721,
+      "grad_norm": 0.6349744483467281,
+      "learning_rate": 9.79208752410183e-06,
+      "loss": 0.4007,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11953890962792796,
+      "grad_norm": 0.7917296463166117,
+      "learning_rate": 9.791708182397962e-06,
+      "loss": 0.4324,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11962095417811872,
+      "grad_norm": 0.696327596787219,
+      "learning_rate": 9.791328502311363e-06,
+      "loss": 0.371,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11970299872830947,
+      "grad_norm": 0.7164068055883751,
+      "learning_rate": 9.790948483868843e-06,
+      "loss": 0.4139,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11978504327850023,
+      "grad_norm": 0.6135485961271476,
+      "learning_rate": 9.790568127097238e-06,
+      "loss": 0.3726,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11986708782869097,
+      "grad_norm": 0.7427009327387653,
+      "learning_rate": 9.790187432023408e-06,
+      "loss": 0.4011,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11994913237888173,
+      "grad_norm": 0.718986600596263,
+      "learning_rate": 9.789806398674238e-06,
+      "loss": 0.4259,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12003117692907249,
+      "grad_norm": 0.643678238979875,
+      "learning_rate": 9.789425027076636e-06,
+      "loss": 0.3611,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12011322147926325,
+      "grad_norm": 0.672378338074051,
+      "learning_rate": 9.789043317257532e-06,
+      "loss": 0.3931,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12019526602945399,
+      "grad_norm": 0.6889385847152725,
+      "learning_rate": 9.788661269243884e-06,
+      "loss": 0.3749,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12027731057964475,
+      "grad_norm": 0.6049584217908063,
+      "learning_rate": 9.788278883062672e-06,
+      "loss": 0.3831,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1203593551298355,
+      "grad_norm": 0.6103393212523193,
+      "learning_rate": 9.787896158740895e-06,
+      "loss": 0.422,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12044139968002625,
+      "grad_norm": 0.6763206537526724,
+      "learning_rate": 9.787513096305587e-06,
+      "loss": 0.395,
+      "step": 1468
+    },
+    {
+      "epoch": 0.120523444230217,
+      "grad_norm": 0.6585679302283872,
+      "learning_rate": 9.787129695783794e-06,
+      "loss": 0.4152,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12060548878040776,
+      "grad_norm": 0.655229914171444,
+      "learning_rate": 9.786745957202593e-06,
+      "loss": 0.3986,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12068753333059852,
+      "grad_norm": 0.7672326689671043,
+      "learning_rate": 9.786361880589084e-06,
+      "loss": 0.3899,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12076957788078926,
+      "grad_norm": 0.8163721734610782,
+      "learning_rate": 9.78597746597039e-06,
+      "loss": 0.4513,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12085162243098002,
+      "grad_norm": 0.724878953553388,
+      "learning_rate": 9.785592713373653e-06,
+      "loss": 0.4271,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12093366698117078,
+      "grad_norm": 0.6928832141730781,
+      "learning_rate": 9.78520762282605e-06,
+      "loss": 0.3971,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12101571153136154,
+      "grad_norm": 0.6655896955464186,
+      "learning_rate": 9.784822194354771e-06,
+      "loss": 0.3817,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12109775608155228,
+      "grad_norm": 0.6966923904541389,
+      "learning_rate": 9.784436427987036e-06,
+      "loss": 0.3726,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12117980063174304,
+      "grad_norm": 0.6672374057189091,
+      "learning_rate": 9.784050323750089e-06,
+      "loss": 0.4058,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1212618451819338,
+      "grad_norm": 0.6636620055452467,
+      "learning_rate": 9.783663881671192e-06,
+      "loss": 0.4166,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12134388973212454,
+      "grad_norm": 0.7162779137978101,
+      "learning_rate": 9.78327710177764e-06,
+      "loss": 0.4372,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1214259342823153,
+      "grad_norm": 0.6573824957186701,
+      "learning_rate": 9.782889984096741e-06,
+      "loss": 0.4075,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12150797883250605,
+      "grad_norm": 0.6421761311508196,
+      "learning_rate": 9.78250252865584e-06,
+      "loss": 0.42,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12159002338269681,
+      "grad_norm": 0.6272964953918427,
+      "learning_rate": 9.78211473548229e-06,
+      "loss": 0.3892,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12167206793288755,
+      "grad_norm": 0.6519749527662803,
+      "learning_rate": 9.781726604603482e-06,
+      "loss": 0.4371,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12175411248307831,
+      "grad_norm": 0.6564008272749982,
+      "learning_rate": 9.781338136046823e-06,
+      "loss": 0.4086,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12183615703326907,
+      "grad_norm": 0.6127098811491234,
+      "learning_rate": 9.780949329839748e-06,
+      "loss": 0.3715,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12191820158345981,
+      "grad_norm": 0.7287987623487143,
+      "learning_rate": 9.78056018600971e-06,
+      "loss": 0.4638,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12200024613365057,
+      "grad_norm": 0.6934663468318079,
+      "learning_rate": 9.780170704584194e-06,
+      "loss": 0.3787,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12208229068384133,
+      "grad_norm": 0.6816676073160215,
+      "learning_rate": 9.779780885590703e-06,
+      "loss": 0.3906,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12216433523403208,
+      "grad_norm": 0.8000506975955545,
+      "learning_rate": 9.779390729056764e-06,
+      "loss": 0.428,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12224637978422283,
+      "grad_norm": 0.6905585108785627,
+      "learning_rate": 9.779000235009932e-06,
+      "loss": 0.4122,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12232842433441359,
+      "grad_norm": 0.7353478250838515,
+      "learning_rate": 9.77860940347778e-06,
+      "loss": 0.4246,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12241046888460434,
+      "grad_norm": 0.6542558113652103,
+      "learning_rate": 9.77821823448791e-06,
+      "loss": 0.4184,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1224925134347951,
+      "grad_norm": 0.6961342811356294,
+      "learning_rate": 9.777826728067945e-06,
+      "loss": 0.3977,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12257455798498584,
+      "grad_norm": 0.7620235386275245,
+      "learning_rate": 9.777434884245533e-06,
+      "loss": 0.4393,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1226566025351766,
+      "grad_norm": 0.6669329335694201,
+      "learning_rate": 9.777042703048345e-06,
+      "loss": 0.3979,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12273864708536736,
+      "grad_norm": 0.6577492281073142,
+      "learning_rate": 9.776650184504075e-06,
+      "loss": 0.4299,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1228206916355581,
+      "grad_norm": 0.6004601414485256,
+      "learning_rate": 9.776257328640444e-06,
+      "loss": 0.3963,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12290273618574886,
+      "grad_norm": 0.6735251535741144,
+      "learning_rate": 9.775864135485194e-06,
+      "loss": 0.3859,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12298478073593962,
+      "grad_norm": 0.6649752938090745,
+      "learning_rate": 9.775470605066091e-06,
+      "loss": 0.4106,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12306682528613037,
+      "grad_norm": 0.7717801592655892,
+      "learning_rate": 9.77507673741093e-06,
+      "loss": 0.427,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12314886983632112,
+      "grad_norm": 0.6541956403553429,
+      "learning_rate": 9.774682532547517e-06,
+      "loss": 0.3843,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12323091438651188,
+      "grad_norm": 0.6331243362936241,
+      "learning_rate": 9.774287990503696e-06,
+      "loss": 0.3947,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12331295893670263,
+      "grad_norm": 0.7189759375048429,
+      "learning_rate": 9.773893111307328e-06,
+      "loss": 0.44,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12339500348689339,
+      "grad_norm": 0.7537063038275403,
+      "learning_rate": 9.773497894986299e-06,
+      "loss": 0.4379,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12347704803708413,
+      "grad_norm": 0.7086106428158296,
+      "learning_rate": 9.773102341568516e-06,
+      "loss": 0.388,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12355909258727489,
+      "grad_norm": 0.6340296492106668,
+      "learning_rate": 9.772706451081916e-06,
+      "loss": 0.3654,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12364113713746565,
+      "grad_norm": 0.6762545384310434,
+      "learning_rate": 9.772310223554453e-06,
+      "loss": 0.4101,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12372318168765639,
+      "grad_norm": 0.7235524084373043,
+      "learning_rate": 9.77191365901411e-06,
+      "loss": 0.3855,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12380522623784715,
+      "grad_norm": 0.8211075387319363,
+      "learning_rate": 9.771516757488893e-06,
+      "loss": 0.4371,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1238872707880379,
+      "grad_norm": 0.6323429640408694,
+      "learning_rate": 9.771119519006827e-06,
+      "loss": 0.4094,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12396931533822866,
+      "grad_norm": 0.7516039792481141,
+      "learning_rate": 9.770721943595964e-06,
+      "loss": 0.4477,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12405135988841941,
+      "grad_norm": 0.6920821834308618,
+      "learning_rate": 9.770324031284384e-06,
+      "loss": 0.379,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12413340443861016,
+      "grad_norm": 0.6751960970228029,
+      "learning_rate": 9.769925782100185e-06,
+      "loss": 0.3733,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12421544898880092,
+      "grad_norm": 0.6437742319098139,
+      "learning_rate": 9.76952719607149e-06,
+      "loss": 0.4277,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12429749353899167,
+      "grad_norm": 0.6640988931469584,
+      "learning_rate": 9.769128273226448e-06,
+      "loss": 0.4092,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12437953808918242,
+      "grad_norm": 0.7973477720412948,
+      "learning_rate": 9.76872901359323e-06,
+      "loss": 0.4295,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12446158263937318,
+      "grad_norm": 0.6922730630024705,
+      "learning_rate": 9.768329417200029e-06,
+      "loss": 0.428,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12454362718956394,
+      "grad_norm": 0.6986213829772566,
+      "learning_rate": 9.767929484075067e-06,
+      "loss": 0.3755,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12462567173975468,
+      "grad_norm": 0.6493608730621981,
+      "learning_rate": 9.767529214246584e-06,
+      "loss": 0.3936,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12470771628994544,
+      "grad_norm": 0.6253465645438284,
+      "learning_rate": 9.767128607742848e-06,
+      "loss": 0.3665,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1247897608401362,
+      "grad_norm": 0.7233093936495785,
+      "learning_rate": 9.766727664592148e-06,
+      "loss": 0.3985,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12487180539032695,
+      "grad_norm": 0.685905721431123,
+      "learning_rate": 9.7663263848228e-06,
+      "loss": 0.4156,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1249538499405177,
+      "grad_norm": 0.6682189641188312,
+      "learning_rate": 9.76592476846314e-06,
+      "loss": 0.3775,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12503589449070845,
+      "grad_norm": 0.6585730601204989,
+      "learning_rate": 9.76552281554153e-06,
+      "loss": 0.4016,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1251179390408992,
+      "grad_norm": 0.6364049782858605,
+      "learning_rate": 9.765120526086353e-06,
+      "loss": 0.4087,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12519998359108997,
+      "grad_norm": 0.6612288014458426,
+      "learning_rate": 9.764717900126024e-06,
+      "loss": 0.4081,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12528202814128073,
+      "grad_norm": 0.6983621467503891,
+      "learning_rate": 9.764314937688968e-06,
+      "loss": 0.3961,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12536407269147146,
+      "grad_norm": 0.604202871290534,
+      "learning_rate": 9.76391163880365e-06,
+      "loss": 0.37,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12544611724166221,
+      "grad_norm": 0.6796838387294488,
+      "learning_rate": 9.763508003498543e-06,
+      "loss": 0.4139,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12552816179185297,
+      "grad_norm": 0.6846546120982818,
+      "learning_rate": 9.763104031802153e-06,
+      "loss": 0.4268,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12561020634204373,
+      "grad_norm": 0.6839907930543507,
+      "learning_rate": 9.76269972374301e-06,
+      "loss": 0.4318,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12569225089223449,
+      "grad_norm": 0.6276950126444646,
+      "learning_rate": 9.762295079349665e-06,
+      "loss": 0.39,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12577429544242524,
+      "grad_norm": 0.6586170956670886,
+      "learning_rate": 9.761890098650693e-06,
+      "loss": 0.4159,
+      "step": 1533
+    },
+    {
+      "epoch": 0.125856339992616,
+      "grad_norm": 0.6274130409913639,
+      "learning_rate": 9.761484781674692e-06,
+      "loss": 0.393,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12593838454280673,
+      "grad_norm": 0.6885647765213672,
+      "learning_rate": 9.761079128450286e-06,
+      "loss": 0.4414,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1260204290929975,
+      "grad_norm": 0.6770569000456611,
+      "learning_rate": 9.76067313900612e-06,
+      "loss": 0.3983,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12610247364318825,
+      "grad_norm": 0.6025019133497252,
+      "learning_rate": 9.760266813370866e-06,
+      "loss": 0.3896,
+      "step": 1537
+    },
+    {
+      "epoch": 0.126184518193379,
+      "grad_norm": 0.672886752232527,
+      "learning_rate": 9.759860151573219e-06,
+      "loss": 0.4056,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12626656274356976,
+      "grad_norm": 0.6814796459924373,
+      "learning_rate": 9.759453153641894e-06,
+      "loss": 0.4026,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12634860729376052,
+      "grad_norm": 0.6230451476365311,
+      "learning_rate": 9.759045819605635e-06,
+      "loss": 0.4298,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12643065184395127,
+      "grad_norm": 0.6122463261793842,
+      "learning_rate": 9.758638149493205e-06,
+      "loss": 0.382,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12651269639414203,
+      "grad_norm": 0.6536640626353912,
+      "learning_rate": 9.758230143333394e-06,
+      "loss": 0.4223,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12659474094433276,
+      "grad_norm": 0.7365533569634384,
+      "learning_rate": 9.757821801155016e-06,
+      "loss": 0.4651,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12667678549452352,
+      "grad_norm": 0.6913869091421551,
+      "learning_rate": 9.757413122986907e-06,
+      "loss": 0.422,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12675883004471428,
+      "grad_norm": 0.6993776107680874,
+      "learning_rate": 9.757004108857924e-06,
+      "loss": 0.3927,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12684087459490503,
+      "grad_norm": 0.763587923762825,
+      "learning_rate": 9.756594758796955e-06,
+      "loss": 0.406,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1269229191450958,
+      "grad_norm": 0.6484544809199568,
+      "learning_rate": 9.756185072832908e-06,
+      "loss": 0.3647,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12700496369528655,
+      "grad_norm": 0.7828990233929114,
+      "learning_rate": 9.755775050994711e-06,
+      "loss": 0.3707,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1270870082454773,
+      "grad_norm": 0.6855145497074832,
+      "learning_rate": 9.75536469331132e-06,
+      "loss": 0.4225,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12716905279566804,
+      "grad_norm": 0.6110355531734428,
+      "learning_rate": 9.754953999811716e-06,
+      "loss": 0.4005,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1272510973458588,
+      "grad_norm": 0.6659901079347503,
+      "learning_rate": 9.754542970524899e-06,
+      "loss": 0.3965,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12733314189604955,
+      "grad_norm": 0.6623992394837738,
+      "learning_rate": 9.754131605479895e-06,
+      "loss": 0.3809,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1274151864462403,
+      "grad_norm": 0.6910033007910751,
+      "learning_rate": 9.753719904705757e-06,
+      "loss": 0.4085,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12749723099643107,
+      "grad_norm": 0.6990140043657594,
+      "learning_rate": 9.753307868231556e-06,
+      "loss": 0.3576,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12757927554662182,
+      "grad_norm": 0.6845144052937195,
+      "learning_rate": 9.75289549608639e-06,
+      "loss": 0.3828,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12766132009681258,
+      "grad_norm": 0.671880369109681,
+      "learning_rate": 9.75248278829938e-06,
+      "loss": 0.4241,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1277433646470033,
+      "grad_norm": 0.6684365876203728,
+      "learning_rate": 9.752069744899672e-06,
+      "loss": 0.376,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12782540919719407,
+      "grad_norm": 0.67072213830453,
+      "learning_rate": 9.751656365916434e-06,
+      "loss": 0.4118,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12790745374738482,
+      "grad_norm": 0.6849174155905544,
+      "learning_rate": 9.751242651378856e-06,
+      "loss": 0.396,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12798949829757558,
+      "grad_norm": 0.6586252790697917,
+      "learning_rate": 9.750828601316155e-06,
+      "loss": 0.3809,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12807154284776634,
+      "grad_norm": 0.6701736534069487,
+      "learning_rate": 9.750414215757571e-06,
+      "loss": 0.4133,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1281535873979571,
+      "grad_norm": 0.6890800721852113,
+      "learning_rate": 9.749999494732368e-06,
+      "loss": 0.4195,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12823563194814785,
+      "grad_norm": 0.641967336951711,
+      "learning_rate": 9.749584438269833e-06,
+      "loss": 0.3707,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12831767649833858,
+      "grad_norm": 0.6677077963114703,
+      "learning_rate": 9.749169046399274e-06,
+      "loss": 0.4076,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12839972104852934,
+      "grad_norm": 0.6692243003723569,
+      "learning_rate": 9.748753319150028e-06,
+      "loss": 0.3893,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1284817655987201,
+      "grad_norm": 0.6573151672467713,
+      "learning_rate": 9.748337256551452e-06,
+      "loss": 0.3962,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12856381014891086,
+      "grad_norm": 0.705883332149516,
+      "learning_rate": 9.747920858632926e-06,
+      "loss": 0.4031,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1286458546991016,
+      "grad_norm": 0.647385654723008,
+      "learning_rate": 9.74750412542386e-06,
+      "loss": 0.4079,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12872789924929237,
+      "grad_norm": 0.6753248391061647,
+      "learning_rate": 9.747087056953679e-06,
+      "loss": 0.4003,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12880994379948313,
+      "grad_norm": 0.6822622235183246,
+      "learning_rate": 9.746669653251837e-06,
+      "loss": 0.4215,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12889198834967389,
+      "grad_norm": 0.6666354553330124,
+      "learning_rate": 9.746251914347809e-06,
+      "loss": 0.4132,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12897403289986462,
+      "grad_norm": 0.6266546383189241,
+      "learning_rate": 9.745833840271098e-06,
+      "loss": 0.455,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12905607745005537,
+      "grad_norm": 0.6721047018914189,
+      "learning_rate": 9.745415431051224e-06,
+      "loss": 0.3886,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12913812200024613,
+      "grad_norm": 0.642027546087138,
+      "learning_rate": 9.744996686717738e-06,
+      "loss": 0.3843,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1292201665504369,
+      "grad_norm": 0.6675568600188481,
+      "learning_rate": 9.74457760730021e-06,
+      "loss": 0.3941,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12930221110062765,
+      "grad_norm": 0.677596289820008,
+      "learning_rate": 9.744158192828232e-06,
+      "loss": 0.3852,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1293842556508184,
+      "grad_norm": 0.7176925654231475,
+      "learning_rate": 9.743738443331426e-06,
+      "loss": 0.431,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12946630020100916,
+      "grad_norm": 0.7021580854790684,
+      "learning_rate": 9.743318358839431e-06,
+      "loss": 0.419,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1295483447511999,
+      "grad_norm": 0.6238288363839153,
+      "learning_rate": 9.742897939381915e-06,
+      "loss": 0.3872,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12963038930139065,
+      "grad_norm": 0.6363286311283066,
+      "learning_rate": 9.742477184988566e-06,
+      "loss": 0.3688,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1297124338515814,
+      "grad_norm": 0.6483499332765814,
+      "learning_rate": 9.7420560956891e-06,
+      "loss": 0.4184,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12979447840177216,
+      "grad_norm": 0.6944101918132317,
+      "learning_rate": 9.741634671513248e-06,
+      "loss": 0.3918,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12987652295196292,
+      "grad_norm": 0.6429135169430492,
+      "learning_rate": 9.741212912490776e-06,
+      "loss": 0.394,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12995856750215368,
+      "grad_norm": 0.677766976561247,
+      "learning_rate": 9.740790818651464e-06,
+      "loss": 0.392,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13004061205234443,
+      "grad_norm": 0.6514365716765502,
+      "learning_rate": 9.740368390025121e-06,
+      "loss": 0.419,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13012265660253516,
+      "grad_norm": 0.6811582557854216,
+      "learning_rate": 9.739945626641579e-06,
+      "loss": 0.4563,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13020470115272592,
+      "grad_norm": 0.6358867166150217,
+      "learning_rate": 9.73952252853069e-06,
+      "loss": 0.4136,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13028674570291668,
+      "grad_norm": 0.6276063124106761,
+      "learning_rate": 9.739099095722338e-06,
+      "loss": 0.4028,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13036879025310744,
+      "grad_norm": 0.6369402329269075,
+      "learning_rate": 9.738675328246421e-06,
+      "loss": 0.3734,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1304508348032982,
+      "grad_norm": 0.6460562635696737,
+      "learning_rate": 9.738251226132864e-06,
+      "loss": 0.408,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13053287935348895,
+      "grad_norm": 0.643767647005558,
+      "learning_rate": 9.73782678941162e-06,
+      "loss": 0.42,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1306149239036797,
+      "grad_norm": 0.7031999774290523,
+      "learning_rate": 9.737402018112658e-06,
+      "loss": 0.4186,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13069696845387044,
+      "grad_norm": 0.6590667109692648,
+      "learning_rate": 9.736976912265978e-06,
+      "loss": 0.4617,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1307790130040612,
+      "grad_norm": 0.6762325585216047,
+      "learning_rate": 9.7365514719016e-06,
+      "loss": 0.4534,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13086105755425195,
+      "grad_norm": 0.6401596533804377,
+      "learning_rate": 9.736125697049565e-06,
+      "loss": 0.3726,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1309431021044427,
+      "grad_norm": 0.6405624940820871,
+      "learning_rate": 9.735699587739943e-06,
+      "loss": 0.3589,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13102514665463347,
+      "grad_norm": 0.656633938705024,
+      "learning_rate": 9.735273144002825e-06,
+      "loss": 0.397,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13110719120482422,
+      "grad_norm": 0.6453958977047897,
+      "learning_rate": 9.734846365868324e-06,
+      "loss": 0.4297,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13118923575501498,
+      "grad_norm": 0.6388074051077589,
+      "learning_rate": 9.734419253366582e-06,
+      "loss": 0.4506,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13127128030520574,
+      "grad_norm": 0.6481145668065146,
+      "learning_rate": 9.733991806527758e-06,
+      "loss": 0.3706,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13135332485539647,
+      "grad_norm": 0.6499197453755916,
+      "learning_rate": 9.733564025382036e-06,
+      "loss": 0.3948,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13143536940558723,
+      "grad_norm": 0.6441508951583312,
+      "learning_rate": 9.73313590995963e-06,
+      "loss": 0.3958,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13151741395577798,
+      "grad_norm": 0.6508959288621353,
+      "learning_rate": 9.73270746029077e-06,
+      "loss": 0.4084,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13159945850596874,
+      "grad_norm": 0.5937972693079825,
+      "learning_rate": 9.732278676405714e-06,
+      "loss": 0.3913,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1316815030561595,
+      "grad_norm": 0.7155186964664231,
+      "learning_rate": 9.731849558334738e-06,
+      "loss": 0.384,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13176354760635026,
+      "grad_norm": 0.7067406114689281,
+      "learning_rate": 9.731420106108152e-06,
+      "loss": 0.4801,
+      "step": 1606
+    },
+    {
+      "epoch": 0.131845592156541,
+      "grad_norm": 0.6873451903847828,
+      "learning_rate": 9.730990319756277e-06,
+      "loss": 0.3753,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13192763670673174,
+      "grad_norm": 0.6487732952602883,
+      "learning_rate": 9.730560199309467e-06,
+      "loss": 0.4314,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1320096812569225,
+      "grad_norm": 0.724534452991904,
+      "learning_rate": 9.730129744798096e-06,
+      "loss": 0.3976,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13209172580711326,
+      "grad_norm": 0.6327196645760851,
+      "learning_rate": 9.729698956252563e-06,
+      "loss": 0.4069,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13217377035730402,
+      "grad_norm": 0.6190690600054813,
+      "learning_rate": 9.729267833703286e-06,
+      "loss": 0.3564,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13225581490749477,
+      "grad_norm": 0.724534896615605,
+      "learning_rate": 9.728836377180715e-06,
+      "loss": 0.3891,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13233785945768553,
+      "grad_norm": 0.6428763179186291,
+      "learning_rate": 9.728404586715317e-06,
+      "loss": 0.451,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1324199040078763,
+      "grad_norm": 0.6915871587078544,
+      "learning_rate": 9.727972462337583e-06,
+      "loss": 0.4033,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13250194855806702,
+      "grad_norm": 0.6728877466538178,
+      "learning_rate": 9.727540004078029e-06,
+      "loss": 0.4272,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13258399310825777,
+      "grad_norm": 0.6377912257148216,
+      "learning_rate": 9.727107211967194e-06,
+      "loss": 0.4146,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13266603765844853,
+      "grad_norm": 0.6878320537648082,
+      "learning_rate": 9.726674086035645e-06,
+      "loss": 0.3974,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1327480822086393,
+      "grad_norm": 0.6743523886057567,
+      "learning_rate": 9.726240626313967e-06,
+      "loss": 0.4037,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13283012675883005,
+      "grad_norm": 0.6580666256219253,
+      "learning_rate": 9.725806832832768e-06,
+      "loss": 0.3944,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1329121713090208,
+      "grad_norm": 0.7360789973516539,
+      "learning_rate": 9.725372705622682e-06,
+      "loss": 0.407,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13299421585921156,
+      "grad_norm": 0.6721686485989398,
+      "learning_rate": 9.724938244714367e-06,
+      "loss": 0.4326,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1330762604094023,
+      "grad_norm": 0.6961730870434042,
+      "learning_rate": 9.724503450138505e-06,
+      "loss": 0.3815,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13315830495959305,
+      "grad_norm": 0.6528871925891847,
+      "learning_rate": 9.724068321925801e-06,
+      "loss": 0.3745,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1332403495097838,
+      "grad_norm": 0.6672025634607378,
+      "learning_rate": 9.72363286010698e-06,
+      "loss": 0.4216,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13332239405997456,
+      "grad_norm": 0.6754758622704397,
+      "learning_rate": 9.723197064712798e-06,
+      "loss": 0.4358,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13340443861016532,
+      "grad_norm": 0.6994786839829311,
+      "learning_rate": 9.722760935774026e-06,
+      "loss": 0.4149,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13348648316035608,
+      "grad_norm": 0.693281162401093,
+      "learning_rate": 9.722324473321464e-06,
+      "loss": 0.4249,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13356852771054684,
+      "grad_norm": 0.5856439581806301,
+      "learning_rate": 9.721887677385936e-06,
+      "loss": 0.3527,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1336505722607376,
+      "grad_norm": 0.6147829204125782,
+      "learning_rate": 9.721450547998285e-06,
+      "loss": 0.3839,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13373261681092832,
+      "grad_norm": 0.6606645073893808,
+      "learning_rate": 9.721013085189384e-06,
+      "loss": 0.4115,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13381466136111908,
+      "grad_norm": 0.6072703632690909,
+      "learning_rate": 9.720575288990121e-06,
+      "loss": 0.3694,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13389670591130984,
+      "grad_norm": 0.5983060404671314,
+      "learning_rate": 9.720137159431418e-06,
+      "loss": 0.3731,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1339787504615006,
+      "grad_norm": 0.6770176955428625,
+      "learning_rate": 9.719698696544211e-06,
+      "loss": 0.3993,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13406079501169135,
+      "grad_norm": 0.603181411966202,
+      "learning_rate": 9.719259900359466e-06,
+      "loss": 0.3581,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1341428395618821,
+      "grad_norm": 0.6591991395281738,
+      "learning_rate": 9.718820770908167e-06,
+      "loss": 0.398,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13422488411207287,
+      "grad_norm": 0.6702028119004705,
+      "learning_rate": 9.718381308221327e-06,
+      "loss": 0.4099,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1343069286622636,
+      "grad_norm": 0.6686821070192852,
+      "learning_rate": 9.71794151232998e-06,
+      "loss": 0.3777,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13438897321245435,
+      "grad_norm": 0.6738732607268347,
+      "learning_rate": 9.717501383265184e-06,
+      "loss": 0.4023,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1344710177626451,
+      "grad_norm": 0.6595127536320008,
+      "learning_rate": 9.717060921058019e-06,
+      "loss": 0.3905,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13455306231283587,
+      "grad_norm": 0.640499927147956,
+      "learning_rate": 9.71662012573959e-06,
+      "loss": 0.4006,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13463510686302663,
+      "grad_norm": 0.628518738795981,
+      "learning_rate": 9.716178997341026e-06,
+      "loss": 0.3536,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13471715141321738,
+      "grad_norm": 0.6405554700067475,
+      "learning_rate": 9.715737535893479e-06,
+      "loss": 0.3729,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13479919596340814,
+      "grad_norm": 0.6601629268467782,
+      "learning_rate": 9.715295741428123e-06,
+      "loss": 0.3958,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13488124051359887,
+      "grad_norm": 0.6828444145399717,
+      "learning_rate": 9.714853613976156e-06,
+      "loss": 0.3621,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13496328506378963,
+      "grad_norm": 0.6578556878862182,
+      "learning_rate": 9.714411153568803e-06,
+      "loss": 0.366,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13504532961398039,
+      "grad_norm": 0.6599618548446035,
+      "learning_rate": 9.71396836023731e-06,
+      "loss": 0.4191,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13512737416417114,
+      "grad_norm": 0.6006643823291847,
+      "learning_rate": 9.713525234012943e-06,
+      "loss": 0.3818,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1352094187143619,
+      "grad_norm": 0.6950765681279477,
+      "learning_rate": 9.713081774926997e-06,
+      "loss": 0.3928,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13529146326455266,
+      "grad_norm": 0.6446073714181629,
+      "learning_rate": 9.712637983010789e-06,
+      "loss": 0.4226,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13537350781474342,
+      "grad_norm": 0.6593469281304896,
+      "learning_rate": 9.712193858295657e-06,
+      "loss": 0.4091,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13545555236493417,
+      "grad_norm": 0.6232536444548172,
+      "learning_rate": 9.711749400812966e-06,
+      "loss": 0.3963,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1355375969151249,
+      "grad_norm": 0.5951451955354329,
+      "learning_rate": 9.711304610594104e-06,
+      "loss": 0.4144,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13561964146531566,
+      "grad_norm": 0.595148825129877,
+      "learning_rate": 9.710859487670478e-06,
+      "loss": 0.4017,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13570168601550642,
+      "grad_norm": 0.6660713741731098,
+      "learning_rate": 9.710414032073524e-06,
+      "loss": 0.4063,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13578373056569717,
+      "grad_norm": 0.6726252111452041,
+      "learning_rate": 9.709968243834698e-06,
+      "loss": 0.4318,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13586577511588793,
+      "grad_norm": 0.6247854459878368,
+      "learning_rate": 9.709522122985482e-06,
+      "loss": 0.4077,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1359478196660787,
+      "grad_norm": 0.7176238050532752,
+      "learning_rate": 9.709075669557381e-06,
+      "loss": 0.4168,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13602986421626945,
+      "grad_norm": 0.9698909817935001,
+      "learning_rate": 9.708628883581921e-06,
+      "loss": 0.7494,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13611190876646018,
+      "grad_norm": 0.6520930136761901,
+      "learning_rate": 9.708181765090654e-06,
+      "loss": 0.3804,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13619395331665093,
+      "grad_norm": 0.6785797200845367,
+      "learning_rate": 9.707734314115157e-06,
+      "loss": 0.4019,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1362759978668417,
+      "grad_norm": 0.5973399387471978,
+      "learning_rate": 9.707286530687025e-06,
+      "loss": 0.3701,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13635804241703245,
+      "grad_norm": 0.6373580329061939,
+      "learning_rate": 9.70683841483788e-06,
+      "loss": 0.3877,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1364400869672232,
+      "grad_norm": 0.648951531138014,
+      "learning_rate": 9.70638996659937e-06,
+      "loss": 0.3784,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13652213151741396,
+      "grad_norm": 0.6753928684898055,
+      "learning_rate": 9.70594118600316e-06,
+      "loss": 0.3964,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13660417606760472,
+      "grad_norm": 0.6433674837475958,
+      "learning_rate": 9.705492073080946e-06,
+      "loss": 0.4197,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13668622061779545,
+      "grad_norm": 0.653519472919519,
+      "learning_rate": 9.70504262786444e-06,
+      "loss": 0.4447,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1367682651679862,
+      "grad_norm": 0.7005015051730782,
+      "learning_rate": 9.704592850385385e-06,
+      "loss": 0.4227,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13685030971817697,
+      "grad_norm": 0.6252080681296962,
+      "learning_rate": 9.704142740675539e-06,
+      "loss": 0.3611,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13693235426836772,
+      "grad_norm": 0.6770551783137584,
+      "learning_rate": 9.703692298766694e-06,
+      "loss": 0.4376,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13701439881855848,
+      "grad_norm": 0.6419697702488025,
+      "learning_rate": 9.703241524690652e-06,
+      "loss": 0.4181,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13709644336874924,
+      "grad_norm": 0.646318590508612,
+      "learning_rate": 9.702790418479252e-06,
+      "loss": 0.3873,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13717848791894,
+      "grad_norm": 0.6494908299913348,
+      "learning_rate": 9.702338980164347e-06,
+      "loss": 0.4188,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13726053246913072,
+      "grad_norm": 0.6730417992048091,
+      "learning_rate": 9.70188720977782e-06,
+      "loss": 0.4262,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13734257701932148,
+      "grad_norm": 0.6436000293832644,
+      "learning_rate": 9.70143510735157e-06,
+      "loss": 0.4075,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13742462156951224,
+      "grad_norm": 0.6493726664551049,
+      "learning_rate": 9.700982672917529e-06,
+      "loss": 0.3738,
+      "step": 1675
+    },
+    {
+      "epoch": 0.137506666119703,
+      "grad_norm": 0.6342064508158133,
+      "learning_rate": 9.700529906507642e-06,
+      "loss": 0.4086,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13758871066989375,
+      "grad_norm": 0.6588596547150405,
+      "learning_rate": 9.700076808153886e-06,
+      "loss": 0.4005,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1376707552200845,
+      "grad_norm": 0.6715754214652936,
+      "learning_rate": 9.699623377888256e-06,
+      "loss": 0.4076,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13775279977027527,
+      "grad_norm": 0.6622775738942146,
+      "learning_rate": 9.699169615742777e-06,
+      "loss": 0.4104,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13783484432046603,
+      "grad_norm": 0.5864352426916448,
+      "learning_rate": 9.698715521749485e-06,
+      "loss": 0.3722,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13791688887065676,
+      "grad_norm": 0.6262017438008677,
+      "learning_rate": 9.698261095940454e-06,
+      "loss": 0.3747,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1379989334208475,
+      "grad_norm": 0.6617649744167831,
+      "learning_rate": 9.697806338347774e-06,
+      "loss": 0.3893,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13808097797103827,
+      "grad_norm": 0.6471035734939772,
+      "learning_rate": 9.697351249003555e-06,
+      "loss": 0.4056,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13816302252122903,
+      "grad_norm": 0.6231307505526998,
+      "learning_rate": 9.69689582793994e-06,
+      "loss": 0.3404,
+      "step": 1684
+    },
+    {
+      "epoch": 0.13824506707141979,
+      "grad_norm": 0.7133051825106753,
+      "learning_rate": 9.696440075189086e-06,
+      "loss": 0.431,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13832711162161054,
+      "grad_norm": 0.6612415628477052,
+      "learning_rate": 9.695983990783182e-06,
+      "loss": 0.4002,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1384091561718013,
+      "grad_norm": 0.7210108547353837,
+      "learning_rate": 9.695527574754431e-06,
+      "loss": 0.4308,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13849120072199203,
+      "grad_norm": 0.6537386152958304,
+      "learning_rate": 9.695070827135067e-06,
+      "loss": 0.3973,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1385732452721828,
+      "grad_norm": 0.6099174232641416,
+      "learning_rate": 9.694613747957344e-06,
+      "loss": 0.3737,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13865528982237355,
+      "grad_norm": 0.6514161090072978,
+      "learning_rate": 9.694156337253541e-06,
+      "loss": 0.379,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1387373343725643,
+      "grad_norm": 0.6785568772167118,
+      "learning_rate": 9.693698595055961e-06,
+      "loss": 0.3715,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13881937892275506,
+      "grad_norm": 0.6684727279853205,
+      "learning_rate": 9.693240521396924e-06,
+      "loss": 0.3882,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13890142347294582,
+      "grad_norm": 0.6824204489363087,
+      "learning_rate": 9.692782116308783e-06,
+      "loss": 0.388,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13898346802313657,
+      "grad_norm": 0.6617457633346283,
+      "learning_rate": 9.692323379823909e-06,
+      "loss": 0.3957,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1390655125733273,
+      "grad_norm": 0.6706517038511693,
+      "learning_rate": 9.691864311974697e-06,
+      "loss": 0.3921,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13914755712351806,
+      "grad_norm": 0.6371419961443001,
+      "learning_rate": 9.691404912793566e-06,
+      "loss": 0.3955,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13922960167370882,
+      "grad_norm": 0.6301764665548193,
+      "learning_rate": 9.690945182312955e-06,
+      "loss": 0.3939,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13931164622389958,
+      "grad_norm": 0.6520518829499622,
+      "learning_rate": 9.690485120565332e-06,
+      "loss": 0.3747,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13939369077409033,
+      "grad_norm": 0.6622015240661902,
+      "learning_rate": 9.690024727583186e-06,
+      "loss": 0.4108,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1394757353242811,
+      "grad_norm": 0.6292592132706053,
+      "learning_rate": 9.68956400339903e-06,
+      "loss": 0.3927,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13955777987447185,
+      "grad_norm": 0.672776217117184,
+      "learning_rate": 9.689102948045398e-06,
+      "loss": 0.3973,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13963982442466258,
+      "grad_norm": 0.6630569994583568,
+      "learning_rate": 9.688641561554849e-06,
+      "loss": 0.3908,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13972186897485334,
+      "grad_norm": 0.70631596259,
+      "learning_rate": 9.688179843959967e-06,
+      "loss": 0.4148,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1398039135250441,
+      "grad_norm": 0.6663008734514986,
+      "learning_rate": 9.687717795293354e-06,
+      "loss": 0.4027,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13988595807523485,
+      "grad_norm": 0.6495239279353695,
+      "learning_rate": 9.687255415587644e-06,
+      "loss": 0.4111,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1399680026254256,
+      "grad_norm": 0.7105889172836348,
+      "learning_rate": 9.686792704875487e-06,
+      "loss": 0.3918,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14005004717561637,
+      "grad_norm": 0.6287773029086667,
+      "learning_rate": 9.686329663189557e-06,
+      "loss": 0.3699,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14013209172580712,
+      "grad_norm": 0.7722245468513048,
+      "learning_rate": 9.685866290562557e-06,
+      "loss": 0.4212,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14021413627599788,
+      "grad_norm": 0.6477937615087848,
+      "learning_rate": 9.685402587027208e-06,
+      "loss": 0.402,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1402961808261886,
+      "grad_norm": 0.6635672581603141,
+      "learning_rate": 9.684938552616257e-06,
+      "loss": 0.4222,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14037822537637937,
+      "grad_norm": 0.6226620166068294,
+      "learning_rate": 9.68447418736247e-06,
+      "loss": 0.4006,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14046026992657012,
+      "grad_norm": 0.711048848837852,
+      "learning_rate": 9.684009491298644e-06,
+      "loss": 0.442,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14054231447676088,
+      "grad_norm": 0.6787796060834663,
+      "learning_rate": 9.683544464457593e-06,
+      "loss": 0.4207,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14062435902695164,
+      "grad_norm": 0.6186356445924795,
+      "learning_rate": 9.683079106872156e-06,
+      "loss": 0.4017,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1407064035771424,
+      "grad_norm": 0.6505845275733174,
+      "learning_rate": 9.682613418575197e-06,
+      "loss": 0.4005,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14078844812733315,
+      "grad_norm": 0.6398391058151615,
+      "learning_rate": 9.6821473995996e-06,
+      "loss": 0.4134,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14087049267752388,
+      "grad_norm": 0.7249152944311217,
+      "learning_rate": 9.681681049978278e-06,
+      "loss": 0.4201,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14095253722771464,
+      "grad_norm": 0.634641928455588,
+      "learning_rate": 9.68121436974416e-06,
+      "loss": 0.3996,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1410345817779054,
+      "grad_norm": 0.7155963158372529,
+      "learning_rate": 9.680747358930205e-06,
+      "loss": 0.3955,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14111662632809616,
+      "grad_norm": 0.6758010921974377,
+      "learning_rate": 9.680280017569393e-06,
+      "loss": 0.4103,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1411986708782869,
+      "grad_norm": 0.6828476544768984,
+      "learning_rate": 9.679812345694725e-06,
+      "loss": 0.3997,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14128071542847767,
+      "grad_norm": 0.6277646974573768,
+      "learning_rate": 9.679344343339228e-06,
+      "loss": 0.4054,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14136275997866843,
+      "grad_norm": 0.6603627347946152,
+      "learning_rate": 9.678876010535949e-06,
+      "loss": 0.3997,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14144480452885916,
+      "grad_norm": 0.7437941080939691,
+      "learning_rate": 9.678407347317967e-06,
+      "loss": 0.4059,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14152684907904992,
+      "grad_norm": 0.6743537465440499,
+      "learning_rate": 9.677938353718372e-06,
+      "loss": 0.3922,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14160889362924067,
+      "grad_norm": 0.6549036919787193,
+      "learning_rate": 9.677469029770286e-06,
+      "loss": 0.4309,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14169093817943143,
+      "grad_norm": 0.7009573379588654,
+      "learning_rate": 9.676999375506854e-06,
+      "loss": 0.4207,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1417729827296222,
+      "grad_norm": 0.6845441925902913,
+      "learning_rate": 9.676529390961237e-06,
+      "loss": 0.3928,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14185502727981295,
+      "grad_norm": 0.6499258243003074,
+      "learning_rate": 9.676059076166632e-06,
+      "loss": 0.3656,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1419370718300037,
+      "grad_norm": 0.6824365474493789,
+      "learning_rate": 9.675588431156245e-06,
+      "loss": 0.4213,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14201911638019443,
+      "grad_norm": 0.6238852895355682,
+      "learning_rate": 9.675117455963317e-06,
+      "loss": 0.3817,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1421011609303852,
+      "grad_norm": 0.6988722704469467,
+      "learning_rate": 9.674646150621103e-06,
+      "loss": 0.3947,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14218320548057595,
+      "grad_norm": 0.6316689640650498,
+      "learning_rate": 9.674174515162889e-06,
+      "loss": 0.4188,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1422652500307667,
+      "grad_norm": 0.7527694544592057,
+      "learning_rate": 9.673702549621982e-06,
+      "loss": 0.3825,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14234729458095746,
+      "grad_norm": 0.6498832592112922,
+      "learning_rate": 9.673230254031708e-06,
+      "loss": 0.406,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14242933913114822,
+      "grad_norm": 0.7018920698467616,
+      "learning_rate": 9.672757628425421e-06,
+      "loss": 0.4209,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14251138368133898,
+      "grad_norm": 0.6595310919546475,
+      "learning_rate": 9.672284672836498e-06,
+      "loss": 0.3989,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14259342823152973,
+      "grad_norm": 0.6429026174372686,
+      "learning_rate": 9.67181138729834e-06,
+      "loss": 0.4054,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14267547278172046,
+      "grad_norm": 0.6347736008827259,
+      "learning_rate": 9.671337771844366e-06,
+      "loss": 0.4029,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14275751733191122,
+      "grad_norm": 0.6878453400054011,
+      "learning_rate": 9.670863826508024e-06,
+      "loss": 0.3938,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14283956188210198,
+      "grad_norm": 0.6861678030259173,
+      "learning_rate": 9.670389551322783e-06,
+      "loss": 0.3605,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14292160643229274,
+      "grad_norm": 0.7586251735867113,
+      "learning_rate": 9.669914946322134e-06,
+      "loss": 0.4065,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1430036509824835,
+      "grad_norm": 0.6357971397754794,
+      "learning_rate": 9.669440011539598e-06,
+      "loss": 0.3458,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14308569553267425,
+      "grad_norm": 0.6584986279337217,
+      "learning_rate": 9.668964747008707e-06,
+      "loss": 0.4003,
+      "step": 1744
+    },
+    {
+      "epoch": 0.143167740082865,
+      "grad_norm": 0.6858484127143393,
+      "learning_rate": 9.668489152763028e-06,
+      "loss": 0.4121,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14324978463305574,
+      "grad_norm": 0.6722117628100942,
+      "learning_rate": 9.668013228836144e-06,
+      "loss": 0.3942,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1433318291832465,
+      "grad_norm": 0.7011677451638232,
+      "learning_rate": 9.667536975261667e-06,
+      "loss": 0.4299,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14341387373343725,
+      "grad_norm": 0.6201871759706097,
+      "learning_rate": 9.667060392073228e-06,
+      "loss": 0.3607,
+      "step": 1748
+    },
+    {
+      "epoch": 0.143495918283628,
+      "grad_norm": 0.6979258554084261,
+      "learning_rate": 9.666583479304483e-06,
+      "loss": 0.3959,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14357796283381877,
+      "grad_norm": 0.8193904705920837,
+      "learning_rate": 9.666106236989107e-06,
+      "loss": 0.4267,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14366000738400952,
+      "grad_norm": 0.6404231965230299,
+      "learning_rate": 9.665628665160809e-06,
+      "loss": 0.4119,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14374205193420028,
+      "grad_norm": 0.6650027157106351,
+      "learning_rate": 9.665150763853307e-06,
+      "loss": 0.3683,
+      "step": 1752
+    },
+    {
+      "epoch": 0.143824096484391,
+      "grad_norm": 0.6180876662121777,
+      "learning_rate": 9.664672533100356e-06,
+      "loss": 0.4485,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14390614103458177,
+      "grad_norm": 0.7126157271517973,
+      "learning_rate": 9.664193972935725e-06,
+      "loss": 0.4076,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14398818558477253,
+      "grad_norm": 0.7670867386860619,
+      "learning_rate": 9.663715083393209e-06,
+      "loss": 0.4359,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14407023013496328,
+      "grad_norm": 0.6196413972544921,
+      "learning_rate": 9.663235864506625e-06,
+      "loss": 0.3962,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14415227468515404,
+      "grad_norm": 0.5985291358578014,
+      "learning_rate": 9.662756316309817e-06,
+      "loss": 0.4024,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1442343192353448,
+      "grad_norm": 0.6000093442977116,
+      "learning_rate": 9.66227643883665e-06,
+      "loss": 0.4028,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14431636378553556,
+      "grad_norm": 0.6677074623666664,
+      "learning_rate": 9.66179623212101e-06,
+      "loss": 0.4273,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14439840833572629,
+      "grad_norm": 0.6287131239746976,
+      "learning_rate": 9.661315696196811e-06,
+      "loss": 0.4394,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14448045288591704,
+      "grad_norm": 0.678570390861692,
+      "learning_rate": 9.660834831097988e-06,
+      "loss": 0.4717,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1445624974361078,
+      "grad_norm": 0.6861028565817904,
+      "learning_rate": 9.660353636858496e-06,
+      "loss": 0.4476,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14464454198629856,
+      "grad_norm": 0.6543150530348537,
+      "learning_rate": 9.659872113512316e-06,
+      "loss": 0.379,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14472658653648932,
+      "grad_norm": 0.6602897603028138,
+      "learning_rate": 9.659390261093456e-06,
+      "loss": 0.3887,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14480863108668007,
+      "grad_norm": 0.7168291464543484,
+      "learning_rate": 9.65890807963594e-06,
+      "loss": 0.4251,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14489067563687083,
+      "grad_norm": 0.6466521788297254,
+      "learning_rate": 9.658425569173822e-06,
+      "loss": 0.3892,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1449727201870616,
+      "grad_norm": 0.6298181834954645,
+      "learning_rate": 9.657942729741174e-06,
+      "loss": 0.3637,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14505476473725232,
+      "grad_norm": 0.7082762343759664,
+      "learning_rate": 9.657459561372094e-06,
+      "loss": 0.4022,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14513680928744307,
+      "grad_norm": 0.6830537250556851,
+      "learning_rate": 9.656976064100703e-06,
+      "loss": 0.3873,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14521885383763383,
+      "grad_norm": 0.6148857806002407,
+      "learning_rate": 9.656492237961143e-06,
+      "loss": 0.3852,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1453008983878246,
+      "grad_norm": 0.6624663264313092,
+      "learning_rate": 9.656008082987584e-06,
+      "loss": 0.3973,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14538294293801535,
+      "grad_norm": 0.5979879900864765,
+      "learning_rate": 9.655523599214215e-06,
+      "loss": 0.3655,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1454649874882061,
+      "grad_norm": 0.6520760542080698,
+      "learning_rate": 9.655038786675247e-06,
+      "loss": 0.4102,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14554703203839686,
+      "grad_norm": 0.7166317240722749,
+      "learning_rate": 9.65455364540492e-06,
+      "loss": 0.4044,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1456290765885876,
+      "grad_norm": 0.6528968878892806,
+      "learning_rate": 9.654068175437492e-06,
+      "loss": 0.4058,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14571112113877835,
+      "grad_norm": 0.7119410303141352,
+      "learning_rate": 9.653582376807248e-06,
+      "loss": 0.3971,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1457931656889691,
+      "grad_norm": 0.6453242069290013,
+      "learning_rate": 9.653096249548493e-06,
+      "loss": 0.417,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14587521023915986,
+      "grad_norm": 0.7337638060073384,
+      "learning_rate": 9.652609793695555e-06,
+      "loss": 0.4442,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14595725478935062,
+      "grad_norm": 0.6402599782927512,
+      "learning_rate": 9.65212300928279e-06,
+      "loss": 0.423,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14603929933954138,
+      "grad_norm": 0.6777987125651905,
+      "learning_rate": 9.651635896344571e-06,
+      "loss": 0.3764,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14612134388973214,
+      "grad_norm": 0.6621402060105367,
+      "learning_rate": 9.6511484549153e-06,
+      "loss": 0.3996,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14620338843992287,
+      "grad_norm": 0.6343833929789129,
+      "learning_rate": 9.650660685029396e-06,
+      "loss": 0.3912,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14628543299011362,
+      "grad_norm": 0.6529944516503842,
+      "learning_rate": 9.650172586721308e-06,
+      "loss": 0.426,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14636747754030438,
+      "grad_norm": 0.7153494145962827,
+      "learning_rate": 9.649684160025503e-06,
+      "loss": 0.4202,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14644952209049514,
+      "grad_norm": 0.6309105266436554,
+      "learning_rate": 9.649195404976471e-06,
+      "loss": 0.4177,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1465315666406859,
+      "grad_norm": 0.6688926430512597,
+      "learning_rate": 9.64870632160873e-06,
+      "loss": 0.4399,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14661361119087665,
+      "grad_norm": 0.6767213274423554,
+      "learning_rate": 9.648216909956818e-06,
+      "loss": 0.3715,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1466956557410674,
+      "grad_norm": 0.8659386219596478,
+      "learning_rate": 9.647727170055294e-06,
+      "loss": 0.3877,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14677770029125814,
+      "grad_norm": 0.6681147541967548,
+      "learning_rate": 9.647237101938748e-06,
+      "loss": 0.4061,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1468597448414489,
+      "grad_norm": 0.6327522875204258,
+      "learning_rate": 9.64674670564178e-06,
+      "loss": 0.3865,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14694178939163965,
+      "grad_norm": 0.6422842129333034,
+      "learning_rate": 9.646255981199027e-06,
+      "loss": 0.3994,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1470238339418304,
+      "grad_norm": 0.6014203982237069,
+      "learning_rate": 9.64576492864514e-06,
+      "loss": 0.3811,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14710587849202117,
+      "grad_norm": 0.6551252819434892,
+      "learning_rate": 9.6452735480148e-06,
+      "loss": 0.4024,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14718792304221193,
+      "grad_norm": 0.6178474541553234,
+      "learning_rate": 9.644781839342706e-06,
+      "loss": 0.4137,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14726996759240268,
+      "grad_norm": 0.6442628465558763,
+      "learning_rate": 9.644289802663578e-06,
+      "loss": 0.4175,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14735201214259344,
+      "grad_norm": 1.8524331741457358,
+      "learning_rate": 9.643797438012166e-06,
+      "loss": 0.404,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14743405669278417,
+      "grad_norm": 0.6706996202233672,
+      "learning_rate": 9.643304745423241e-06,
+      "loss": 0.3913,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14751610124297493,
+      "grad_norm": 0.725673945715908,
+      "learning_rate": 9.642811724931595e-06,
+      "loss": 0.4046,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14759814579316569,
+      "grad_norm": 0.7066561668135817,
+      "learning_rate": 9.642318376572044e-06,
+      "loss": 0.4349,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14768019034335644,
+      "grad_norm": 0.6600143158272705,
+      "learning_rate": 9.641824700379427e-06,
+      "loss": 0.4058,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1477622348935472,
+      "grad_norm": 0.6765257982493408,
+      "learning_rate": 9.64133069638861e-06,
+      "loss": 0.4123,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14784427944373796,
+      "grad_norm": 0.6991802995648302,
+      "learning_rate": 9.640836364634473e-06,
+      "loss": 0.4351,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14792632399392872,
+      "grad_norm": 0.6709754498558876,
+      "learning_rate": 9.640341705151929e-06,
+      "loss": 0.3812,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14800836854411945,
+      "grad_norm": 0.6903113714496548,
+      "learning_rate": 9.639846717975909e-06,
+      "loss": 0.407,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1480904130943102,
+      "grad_norm": 0.6652964021507553,
+      "learning_rate": 9.639351403141368e-06,
+      "loss": 0.4408,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14817245764450096,
+      "grad_norm": 0.699769433439595,
+      "learning_rate": 9.638855760683285e-06,
+      "loss": 0.4323,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14825450219469172,
+      "grad_norm": 0.6422648339841057,
+      "learning_rate": 9.63835979063666e-06,
+      "loss": 0.3943,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14833654674488247,
+      "grad_norm": 0.7378339180262528,
+      "learning_rate": 9.637863493036522e-06,
+      "loss": 0.3784,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14841859129507323,
+      "grad_norm": 0.7142607088329871,
+      "learning_rate": 9.637366867917912e-06,
+      "loss": 0.463,
+      "step": 1809
+    },
+    {
+      "epoch": 0.148500635845264,
+      "grad_norm": 0.7264737187166976,
+      "learning_rate": 9.636869915315905e-06,
+      "loss": 0.397,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14858268039545472,
+      "grad_norm": 0.7381463155556915,
+      "learning_rate": 9.636372635265595e-06,
+      "loss": 0.4573,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14866472494564548,
+      "grad_norm": 0.65705137928414,
+      "learning_rate": 9.635875027802097e-06,
+      "loss": 0.4764,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14874676949583623,
+      "grad_norm": 0.6566487610245988,
+      "learning_rate": 9.635377092960554e-06,
+      "loss": 0.3823,
+      "step": 1813
+    },
+    {
+      "epoch": 0.148828814046027,
+      "grad_norm": 0.6329750521544278,
+      "learning_rate": 9.634878830776128e-06,
+      "loss": 0.36,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14891085859621775,
+      "grad_norm": 0.6058107816356366,
+      "learning_rate": 9.634380241284005e-06,
+      "loss": 0.3671,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1489929031464085,
+      "grad_norm": 0.738036296580335,
+      "learning_rate": 9.633881324519397e-06,
+      "loss": 0.3944,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14907494769659926,
+      "grad_norm": 0.7040162399759994,
+      "learning_rate": 9.633382080517533e-06,
+      "loss": 0.3594,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14915699224679,
+      "grad_norm": 0.7224331403652545,
+      "learning_rate": 9.63288250931367e-06,
+      "loss": 0.4742,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14923903679698075,
+      "grad_norm": 0.7252701211769775,
+      "learning_rate": 9.632382610943088e-06,
+      "loss": 0.3708,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1493210813471715,
+      "grad_norm": 0.7041878827281002,
+      "learning_rate": 9.631882385441088e-06,
+      "loss": 0.4372,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14940312589736227,
+      "grad_norm": 0.6355876792508546,
+      "learning_rate": 9.631381832842996e-06,
+      "loss": 0.4081,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14948517044755302,
+      "grad_norm": 0.7001064516011452,
+      "learning_rate": 9.630880953184161e-06,
+      "loss": 0.4391,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14956721499774378,
+      "grad_norm": 0.722373457072694,
+      "learning_rate": 9.630379746499951e-06,
+      "loss": 0.4541,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14964925954793454,
+      "grad_norm": 0.6498748948699501,
+      "learning_rate": 9.629878212825763e-06,
+      "loss": 0.3512,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1497313040981253,
+      "grad_norm": 0.9635754658889876,
+      "learning_rate": 9.629376352197017e-06,
+      "loss": 0.3915,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14981334864831602,
+      "grad_norm": 0.6493405525469551,
+      "learning_rate": 9.628874164649147e-06,
+      "loss": 0.3937,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14989539319850678,
+      "grad_norm": 0.6654530409628264,
+      "learning_rate": 9.628371650217623e-06,
+      "loss": 0.3827,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14997743774869754,
+      "grad_norm": 0.6455910347418785,
+      "learning_rate": 9.627868808937927e-06,
+      "loss": 0.4175,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1500594822988883,
+      "grad_norm": 0.7116758613133379,
+      "learning_rate": 9.627365640845571e-06,
+      "loss": 0.3985,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15014152684907905,
+      "grad_norm": 0.8186316931604863,
+      "learning_rate": 9.62686214597609e-06,
+      "loss": 0.4053,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1502235713992698,
+      "grad_norm": 1.0202609000165899,
+      "learning_rate": 9.626358324365036e-06,
+      "loss": 0.3873,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15030561594946057,
+      "grad_norm": 0.6350717539560979,
+      "learning_rate": 9.625854176047991e-06,
+      "loss": 0.3751,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1503876604996513,
+      "grad_norm": 0.6728504365355287,
+      "learning_rate": 9.625349701060556e-06,
+      "loss": 0.4188,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15046970504984206,
+      "grad_norm": 2.2083616329065823,
+      "learning_rate": 9.624844899438356e-06,
+      "loss": 0.3537,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1505517496000328,
+      "grad_norm": 0.7334929788608036,
+      "learning_rate": 9.62433977121704e-06,
+      "loss": 0.4625,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15063379415022357,
+      "grad_norm": 0.6828869951091898,
+      "learning_rate": 9.623834316432279e-06,
+      "loss": 0.4404,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15071583870041433,
+      "grad_norm": 0.6644741027399357,
+      "learning_rate": 9.623328535119766e-06,
+      "loss": 0.4194,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15079788325060509,
+      "grad_norm": 0.8682427681674405,
+      "learning_rate": 9.62282242731522e-06,
+      "loss": 0.4013,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15087992780079584,
+      "grad_norm": 0.7849331144475015,
+      "learning_rate": 9.622315993054384e-06,
+      "loss": 0.3943,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15096197235098657,
+      "grad_norm": 0.6791649922155487,
+      "learning_rate": 9.621809232373016e-06,
+      "loss": 0.4078,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15104401690117733,
+      "grad_norm": 0.7717333061775307,
+      "learning_rate": 9.621302145306906e-06,
+      "loss": 0.4087,
+      "step": 1841
+    },
+    {
+      "epoch": 0.1511260614513681,
+      "grad_norm": 0.7599313716690608,
+      "learning_rate": 9.620794731891862e-06,
+      "loss": 0.3889,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15120810600155885,
+      "grad_norm": 0.6480671446379147,
+      "learning_rate": 9.620286992163722e-06,
+      "loss": 0.3743,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1512901505517496,
+      "grad_norm": 0.6341921825821116,
+      "learning_rate": 9.619778926158334e-06,
+      "loss": 0.3982,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15137219510194036,
+      "grad_norm": 0.7934183144315617,
+      "learning_rate": 9.619270533911583e-06,
+      "loss": 0.4124,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15145423965213112,
+      "grad_norm": 0.7257089581445592,
+      "learning_rate": 9.618761815459366e-06,
+      "loss": 0.4084,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15153628420232185,
+      "grad_norm": 0.6669503952731342,
+      "learning_rate": 9.61825277083761e-06,
+      "loss": 0.4134,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1516183287525126,
+      "grad_norm": 0.7200350540924865,
+      "learning_rate": 9.617743400082264e-06,
+      "loss": 0.4127,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15170037330270336,
+      "grad_norm": 0.7789687621364638,
+      "learning_rate": 9.617233703229298e-06,
+      "loss": 0.4624,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15178241785289412,
+      "grad_norm": 0.6477220728483736,
+      "learning_rate": 9.616723680314708e-06,
+      "loss": 0.3985,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15186446240308488,
+      "grad_norm": 0.7313832574885751,
+      "learning_rate": 9.616213331374507e-06,
+      "loss": 0.3988,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15194650695327563,
+      "grad_norm": 0.6859749315589025,
+      "learning_rate": 9.615702656444736e-06,
+      "loss": 0.4094,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1520285515034664,
+      "grad_norm": 0.8889611205508777,
+      "learning_rate": 9.615191655561462e-06,
+      "loss": 0.3777,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15211059605365715,
+      "grad_norm": 0.7412358952295893,
+      "learning_rate": 9.614680328760766e-06,
+      "loss": 0.4071,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15219264060384788,
+      "grad_norm": 0.8093300785912457,
+      "learning_rate": 9.61416867607876e-06,
+      "loss": 0.398,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15227468515403864,
+      "grad_norm": 0.6828719042061052,
+      "learning_rate": 9.613656697551575e-06,
+      "loss": 0.4161,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1523567297042294,
+      "grad_norm": 0.7220147270241172,
+      "learning_rate": 9.613144393215367e-06,
+      "loss": 0.378,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15243877425442015,
+      "grad_norm": 1.019343157328132,
+      "learning_rate": 9.612631763106312e-06,
+      "loss": 0.4123,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1525208188046109,
+      "grad_norm": 0.771487606386382,
+      "learning_rate": 9.612118807260615e-06,
+      "loss": 0.3721,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15260286335480167,
+      "grad_norm": 0.8596852347415576,
+      "learning_rate": 9.611605525714495e-06,
+      "loss": 0.3883,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15268490790499242,
+      "grad_norm": 0.7217531869115089,
+      "learning_rate": 9.611091918504206e-06,
+      "loss": 0.3826,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15276695245518315,
+      "grad_norm": 0.6050684085858685,
+      "learning_rate": 9.61057798566601e-06,
+      "loss": 0.3767,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1528489970053739,
+      "grad_norm": 0.6674442075950593,
+      "learning_rate": 9.610063727236205e-06,
+      "loss": 0.4076,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15293104155556467,
+      "grad_norm": 0.737433006273814,
+      "learning_rate": 9.609549143251107e-06,
+      "loss": 0.4325,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15301308610575542,
+      "grad_norm": 0.7883290143524665,
+      "learning_rate": 9.609034233747055e-06,
+      "loss": 0.416,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15309513065594618,
+      "grad_norm": 1.2490605539637176,
+      "learning_rate": 9.60851899876041e-06,
+      "loss": 0.4163,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15317717520613694,
+      "grad_norm": 0.9313429167704237,
+      "learning_rate": 9.608003438327556e-06,
+      "loss": 0.3909,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1532592197563277,
+      "grad_norm": 0.8001806640224577,
+      "learning_rate": 9.607487552484904e-06,
+      "loss": 0.402,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15334126430651843,
+      "grad_norm": 0.7398905661268993,
+      "learning_rate": 9.606971341268882e-06,
+      "loss": 0.3721,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15342330885670918,
+      "grad_norm": 0.8119327839083773,
+      "learning_rate": 9.606454804715948e-06,
+      "loss": 0.3965,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15350535340689994,
+      "grad_norm": 0.766634374854435,
+      "learning_rate": 9.605937942862574e-06,
+      "loss": 0.421,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1535873979570907,
+      "grad_norm": 0.7143498448980892,
+      "learning_rate": 9.605420755745263e-06,
+      "loss": 0.4006,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15366944250728146,
+      "grad_norm": 0.6709932912971895,
+      "learning_rate": 9.604903243400538e-06,
+      "loss": 0.3955,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1537514870574722,
+      "grad_norm": 0.9725082051373911,
+      "learning_rate": 9.604385405864944e-06,
+      "loss": 0.3954,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15383353160766297,
+      "grad_norm": 0.6831034823191432,
+      "learning_rate": 9.60386724317505e-06,
+      "loss": 0.3812,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1539155761578537,
+      "grad_norm": 0.9279289429372167,
+      "learning_rate": 9.60334875536745e-06,
+      "loss": 0.4231,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15399762070804446,
+      "grad_norm": 0.5682137902193274,
+      "learning_rate": 9.602829942478756e-06,
+      "loss": 0.3686,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15407966525823522,
+      "grad_norm": 0.6609108185261404,
+      "learning_rate": 9.602310804545605e-06,
+      "loss": 0.4023,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15416170980842597,
+      "grad_norm": 0.8081346868549867,
+      "learning_rate": 9.601791341604658e-06,
+      "loss": 0.3952,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15424375435861673,
+      "grad_norm": 0.7868872758547408,
+      "learning_rate": 9.601271553692603e-06,
+      "loss": 0.3963,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1543257989088075,
+      "grad_norm": 0.7453360617918365,
+      "learning_rate": 9.600751440846142e-06,
+      "loss": 0.369,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15440784345899825,
+      "grad_norm": 0.7997994362176901,
+      "learning_rate": 9.600231003102006e-06,
+      "loss": 0.389,
+      "step": 1882
+    },
+    {
+      "epoch": 0.154489888009189,
+      "grad_norm": 0.8364042279113968,
+      "learning_rate": 9.599710240496946e-06,
+      "loss": 0.4285,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15457193255937973,
+      "grad_norm": 0.8366251544825906,
+      "learning_rate": 9.59918915306774e-06,
+      "loss": 0.4329,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1546539771095705,
+      "grad_norm": 0.8616323462080726,
+      "learning_rate": 9.598667740851187e-06,
+      "loss": 0.3685,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15473602165976125,
+      "grad_norm": 0.7639225136960833,
+      "learning_rate": 9.598146003884103e-06,
+      "loss": 0.4375,
+      "step": 1886
+    },
+    {
+      "epoch": 0.154818066209952,
+      "grad_norm": 0.676434123459871,
+      "learning_rate": 9.597623942203337e-06,
+      "loss": 0.4339,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15490011076014276,
+      "grad_norm": 0.6568700793947285,
+      "learning_rate": 9.597101555845755e-06,
+      "loss": 0.4067,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15498215531033352,
+      "grad_norm": 0.6898847525327378,
+      "learning_rate": 9.596578844848248e-06,
+      "loss": 0.4019,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15506419986052428,
+      "grad_norm": 0.7230864807115737,
+      "learning_rate": 9.596055809247727e-06,
+      "loss": 0.4215,
+      "step": 1890
+    },
+    {
+      "epoch": 0.155146244410715,
+      "grad_norm": 0.7121888353139539,
+      "learning_rate": 9.595532449081128e-06,
+      "loss": 0.3965,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15522828896090576,
+      "grad_norm": 0.7329971853674438,
+      "learning_rate": 9.595008764385412e-06,
+      "loss": 0.4081,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15531033351109652,
+      "grad_norm": 0.6991748405418516,
+      "learning_rate": 9.59448475519756e-06,
+      "loss": 0.3938,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15539237806128728,
+      "grad_norm": 0.8053940419421122,
+      "learning_rate": 9.593960421554576e-06,
+      "loss": 0.3644,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15547442261147804,
+      "grad_norm": 1.003890782577942,
+      "learning_rate": 9.593435763493485e-06,
+      "loss": 0.3815,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1555564671616688,
+      "grad_norm": 0.7488656545348362,
+      "learning_rate": 9.592910781051344e-06,
+      "loss": 0.3713,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15563851171185955,
+      "grad_norm": 0.8878539374585147,
+      "learning_rate": 9.592385474265222e-06,
+      "loss": 0.389,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15572055626205028,
+      "grad_norm": 0.8517589865853034,
+      "learning_rate": 9.591859843172215e-06,
+      "loss": 0.3602,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15580260081224104,
+      "grad_norm": 1.0208696141636573,
+      "learning_rate": 9.591333887809441e-06,
+      "loss": 0.3995,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1558846453624318,
+      "grad_norm": 0.9054219441350552,
+      "learning_rate": 9.59080760821405e-06,
+      "loss": 0.4344,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15596668991262255,
+      "grad_norm": 0.8634113871953338,
+      "learning_rate": 9.590281004423196e-06,
+      "loss": 0.391,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1560487344628133,
+      "grad_norm": 0.7641915571996383,
+      "learning_rate": 9.589754076474077e-06,
+      "loss": 0.4101,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15613077901300407,
+      "grad_norm": 0.7413375795144811,
+      "learning_rate": 9.589226824403895e-06,
+      "loss": 0.3846,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15621282356319482,
+      "grad_norm": 0.7542964218527678,
+      "learning_rate": 9.588699248249889e-06,
+      "loss": 0.3898,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15629486811338558,
+      "grad_norm": 0.7647636541827816,
+      "learning_rate": 9.588171348049316e-06,
+      "loss": 0.3518,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1563769126635763,
+      "grad_norm": 1.0625161619101555,
+      "learning_rate": 9.587643123839451e-06,
+      "loss": 0.3923,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15645895721376707,
+      "grad_norm": 1.1278713748856695,
+      "learning_rate": 9.587114575657603e-06,
+      "loss": 0.3979,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15654100176395783,
+      "grad_norm": 0.7843092945758545,
+      "learning_rate": 9.586585703541092e-06,
+      "loss": 0.4106,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15662304631414858,
+      "grad_norm": 0.7049276949793865,
+      "learning_rate": 9.586056507527266e-06,
+      "loss": 0.4328,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15670509086433934,
+      "grad_norm": 0.8602624927156759,
+      "learning_rate": 9.5855269876535e-06,
+      "loss": 0.4095,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1567871354145301,
+      "grad_norm": 1.0080931328601075,
+      "learning_rate": 9.584997143957182e-06,
+      "loss": 0.3934,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15686917996472086,
+      "grad_norm": 0.7455851684078676,
+      "learning_rate": 9.584466976475734e-06,
+      "loss": 0.395,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15695122451491159,
+      "grad_norm": 0.6770224851453617,
+      "learning_rate": 9.583936485246593e-06,
+      "loss": 0.4208,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15703326906510234,
+      "grad_norm": 0.6414189899015822,
+      "learning_rate": 9.583405670307223e-06,
+      "loss": 0.3612,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1571153136152931,
+      "grad_norm": 0.7442098524546827,
+      "learning_rate": 9.582874531695109e-06,
+      "loss": 0.382,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15719735816548386,
+      "grad_norm": 0.9044930523605261,
+      "learning_rate": 9.582343069447758e-06,
+      "loss": 0.4188,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15727940271567462,
+      "grad_norm": 0.9074390520828923,
+      "learning_rate": 9.5818112836027e-06,
+      "loss": 0.4305,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15736144726586537,
+      "grad_norm": 0.7526769108197663,
+      "learning_rate": 9.581279174197491e-06,
+      "loss": 0.4028,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15744349181605613,
+      "grad_norm": 0.7773560750416476,
+      "learning_rate": 9.580746741269707e-06,
+      "loss": 0.3946,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15752553636624686,
+      "grad_norm": 0.6934782680379967,
+      "learning_rate": 9.580213984856947e-06,
+      "loss": 0.3675,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15760758091643762,
+      "grad_norm": 0.7537539249042504,
+      "learning_rate": 9.579680904996836e-06,
+      "loss": 0.3711,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15768962546662837,
+      "grad_norm": 0.6956439661188352,
+      "learning_rate": 9.579147501727018e-06,
+      "loss": 0.3786,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15777167001681913,
+      "grad_norm": 0.7376462050501253,
+      "learning_rate": 9.578613775085158e-06,
+      "loss": 0.3732,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1578537145670099,
+      "grad_norm": 0.649545634175982,
+      "learning_rate": 9.578079725108951e-06,
+      "loss": 0.3887,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15793575911720065,
+      "grad_norm": 0.8313741230397547,
+      "learning_rate": 9.577545351836108e-06,
+      "loss": 0.3704,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1580178036673914,
+      "grad_norm": 1.2730265102510308,
+      "learning_rate": 9.577010655304367e-06,
+      "loss": 0.3832,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15809984821758213,
+      "grad_norm": 0.7024896967487745,
+      "learning_rate": 9.576475635551486e-06,
+      "loss": 0.4487,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1581818927677729,
+      "grad_norm": 0.7347799339838125,
+      "learning_rate": 9.575940292615248e-06,
+      "loss": 0.3531,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15826393731796365,
+      "grad_norm": 0.7592057417886311,
+      "learning_rate": 9.575404626533462e-06,
+      "loss": 0.3638,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1583459818681544,
+      "grad_norm": 2.6982678429240807,
+      "learning_rate": 9.574868637343949e-06,
+      "loss": 0.3853,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15842802641834516,
+      "grad_norm": 0.6649804021728528,
+      "learning_rate": 9.574332325084564e-06,
+      "loss": 0.3601,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15851007096853592,
+      "grad_norm": 0.6979823958217983,
+      "learning_rate": 9.573795689793179e-06,
+      "loss": 0.3744,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15859211551872668,
+      "grad_norm": 0.7959594244142736,
+      "learning_rate": 9.573258731507691e-06,
+      "loss": 0.3906,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15867416006891744,
+      "grad_norm": 0.9877746457939535,
+      "learning_rate": 9.572721450266018e-06,
+      "loss": 0.3743,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15875620461910817,
+      "grad_norm": 1.3764316004798711,
+      "learning_rate": 9.572183846106105e-06,
+      "loss": 0.4141,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15883824916929892,
+      "grad_norm": 0.7754140548419465,
+      "learning_rate": 9.571645919065913e-06,
+      "loss": 0.4327,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15892029371948968,
+      "grad_norm": 1.2003124492932895,
+      "learning_rate": 9.57110766918343e-06,
+      "loss": 0.3809,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15900233826968044,
+      "grad_norm": 1.4499890631415089,
+      "learning_rate": 9.57056909649667e-06,
+      "loss": 0.3597,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1590843828198712,
+      "grad_norm": 0.7119809543584397,
+      "learning_rate": 9.570030201043662e-06,
+      "loss": 0.4325,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15916642737006195,
+      "grad_norm": 0.715413981662324,
+      "learning_rate": 9.569490982862464e-06,
+      "loss": 0.4172,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1592484719202527,
+      "grad_norm": 0.683315246471334,
+      "learning_rate": 9.568951441991154e-06,
+      "loss": 0.3994,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15933051647044344,
+      "grad_norm": 1.0402857033511357,
+      "learning_rate": 9.568411578467834e-06,
+      "loss": 0.3864,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1594125610206342,
+      "grad_norm": 0.9334044394999492,
+      "learning_rate": 9.567871392330627e-06,
+      "loss": 0.4122,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15949460557082495,
+      "grad_norm": 0.9896245713013732,
+      "learning_rate": 9.567330883617682e-06,
+      "loss": 0.4237,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1595766501210157,
+      "grad_norm": 0.8477232654325453,
+      "learning_rate": 9.566790052367167e-06,
+      "loss": 0.414,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15965869467120647,
+      "grad_norm": 1.745163961525163,
+      "learning_rate": 9.566248898617277e-06,
+      "loss": 0.4091,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15974073922139723,
+      "grad_norm": 0.7186767083130713,
+      "learning_rate": 9.565707422406227e-06,
+      "loss": 0.4262,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15982278377158798,
+      "grad_norm": 0.9946328572321541,
+      "learning_rate": 9.565165623772252e-06,
+      "loss": 0.399,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1599048283217787,
+      "grad_norm": 0.9051702325789885,
+      "learning_rate": 9.564623502753617e-06,
+      "loss": 0.4196,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15998687287196947,
+      "grad_norm": 0.7329518808790392,
+      "learning_rate": 9.5640810593886e-06,
+      "loss": 0.4241,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16006891742216023,
+      "grad_norm": 0.6059202459362284,
+      "learning_rate": 9.563538293715515e-06,
+      "loss": 0.3533,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16015096197235099,
+      "grad_norm": 0.6470734947006833,
+      "learning_rate": 9.562995205772686e-06,
+      "loss": 0.3665,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16023300652254174,
+      "grad_norm": 0.6333585598290781,
+      "learning_rate": 9.562451795598469e-06,
+      "loss": 0.4003,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1603150510727325,
+      "grad_norm": 0.6459968971572155,
+      "learning_rate": 9.561908063231234e-06,
+      "loss": 0.4443,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16039709562292326,
+      "grad_norm": 0.6384250451743513,
+      "learning_rate": 9.561364008709382e-06,
+      "loss": 0.4447,
+      "step": 1955
+    },
+    {
+      "epoch": 0.160479140173114,
+      "grad_norm": 1.4414832387891947,
+      "learning_rate": 9.560819632071332e-06,
+      "loss": 0.4164,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16056118472330475,
+      "grad_norm": 0.8374636712224814,
+      "learning_rate": 9.560274933355526e-06,
+      "loss": 0.4052,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1606432292734955,
+      "grad_norm": 0.7260656505741745,
+      "learning_rate": 9.559729912600431e-06,
+      "loss": 0.4082,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16072527382368626,
+      "grad_norm": 0.6847775676553601,
+      "learning_rate": 9.559184569844537e-06,
+      "loss": 0.3835,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16080731837387702,
+      "grad_norm": 0.9222455941052295,
+      "learning_rate": 9.558638905126352e-06,
+      "loss": 0.3743,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16088936292406777,
+      "grad_norm": 1.4606496656768977,
+      "learning_rate": 9.558092918484412e-06,
+      "loss": 0.4028,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16097140747425853,
+      "grad_norm": 0.8013660666629057,
+      "learning_rate": 9.557546609957274e-06,
+      "loss": 0.3955,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1610534520244493,
+      "grad_norm": 0.7190178220713034,
+      "learning_rate": 9.556999979583514e-06,
+      "loss": 0.4281,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16113549657464002,
+      "grad_norm": 0.7247228636456992,
+      "learning_rate": 9.55645302740174e-06,
+      "loss": 0.4166,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16121754112483078,
+      "grad_norm": 0.7074941284200869,
+      "learning_rate": 9.555905753450572e-06,
+      "loss": 0.357,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16129958567502153,
+      "grad_norm": 0.9809315066710028,
+      "learning_rate": 9.55535815776866e-06,
+      "loss": 0.3985,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1613816302252123,
+      "grad_norm": 0.9563937998682773,
+      "learning_rate": 9.554810240394674e-06,
+      "loss": 0.3668,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16146367477540305,
+      "grad_norm": 2.8834831451153793,
+      "learning_rate": 9.554262001367305e-06,
+      "loss": 0.354,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1615457193255938,
+      "grad_norm": 0.7654506604058557,
+      "learning_rate": 9.553713440725271e-06,
+      "loss": 0.4132,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16162776387578456,
+      "grad_norm": 0.7483664777073545,
+      "learning_rate": 9.55316455850731e-06,
+      "loss": 0.3702,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1617098084259753,
+      "grad_norm": 0.7046036929955434,
+      "learning_rate": 9.552615354752184e-06,
+      "loss": 0.4018,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16179185297616605,
+      "grad_norm": 0.8118892450052483,
+      "learning_rate": 9.552065829498676e-06,
+      "loss": 0.4123,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1618738975263568,
+      "grad_norm": 0.65021370054662,
+      "learning_rate": 9.55151598278559e-06,
+      "loss": 0.3889,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16195594207654757,
+      "grad_norm": 0.8943213531808847,
+      "learning_rate": 9.550965814651762e-06,
+      "loss": 0.3633,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16203798662673832,
+      "grad_norm": 0.7446667355558567,
+      "learning_rate": 9.550415325136038e-06,
+      "loss": 0.4457,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16212003117692908,
+      "grad_norm": 0.6142091809001924,
+      "learning_rate": 9.549864514277294e-06,
+      "loss": 0.3568,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16220207572711984,
+      "grad_norm": 0.6659297831608174,
+      "learning_rate": 9.549313382114427e-06,
+      "loss": 0.4088,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16228412027731057,
+      "grad_norm": 0.6707498096605559,
+      "learning_rate": 9.54876192868636e-06,
+      "loss": 0.3599,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16236616482750132,
+      "grad_norm": 0.711571949258598,
+      "learning_rate": 9.548210154032032e-06,
+      "loss": 0.4261,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16244820937769208,
+      "grad_norm": 0.6908732747200182,
+      "learning_rate": 9.547658058190412e-06,
+      "loss": 0.3817,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16253025392788284,
+      "grad_norm": 0.6305463124520796,
+      "learning_rate": 9.547105641200484e-06,
+      "loss": 0.3689,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1626122984780736,
+      "grad_norm": 0.6179914685371531,
+      "learning_rate": 9.546552903101263e-06,
+      "loss": 0.3505,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16269434302826435,
+      "grad_norm": 0.7049669203361869,
+      "learning_rate": 9.545999843931782e-06,
+      "loss": 0.3852,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1627763875784551,
+      "grad_norm": 0.6228365107223868,
+      "learning_rate": 9.545446463731092e-06,
+      "loss": 0.3744,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16285843212864584,
+      "grad_norm": 0.6367485527420049,
+      "learning_rate": 9.544892762538279e-06,
+      "loss": 0.4101,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1629404766788366,
+      "grad_norm": 0.7522629680204149,
+      "learning_rate": 9.544338740392438e-06,
+      "loss": 0.438,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16302252122902736,
+      "grad_norm": 0.6541765516724902,
+      "learning_rate": 9.543784397332698e-06,
+      "loss": 0.3621,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1631045657792181,
+      "grad_norm": 0.6450138292144433,
+      "learning_rate": 9.543229733398203e-06,
+      "loss": 0.3893,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16318661032940887,
+      "grad_norm": 0.6404737565847795,
+      "learning_rate": 9.542674748628125e-06,
+      "loss": 0.4026,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16326865487959963,
+      "grad_norm": 0.8092023355456673,
+      "learning_rate": 9.542119443061652e-06,
+      "loss": 0.3734,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16335069942979039,
+      "grad_norm": 0.690983312456753,
+      "learning_rate": 9.541563816738005e-06,
+      "loss": 0.4017,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16343274397998114,
+      "grad_norm": 0.6182705725217754,
+      "learning_rate": 9.541007869696415e-06,
+      "loss": 0.4275,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16351478853017187,
+      "grad_norm": 0.6857874932346671,
+      "learning_rate": 9.540451601976147e-06,
+      "loss": 0.3895,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16359683308036263,
+      "grad_norm": 0.7411404060654745,
+      "learning_rate": 9.53989501361648e-06,
+      "loss": 0.3872,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1636788776305534,
+      "grad_norm": 0.6319110773086042,
+      "learning_rate": 9.539338104656722e-06,
+      "loss": 0.3674,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16376092218074415,
+      "grad_norm": 0.7146678155642402,
+      "learning_rate": 9.538780875136197e-06,
+      "loss": 0.4,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1638429667309349,
+      "grad_norm": 0.6481602861396486,
+      "learning_rate": 9.538223325094263e-06,
+      "loss": 0.4144,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16392501128112566,
+      "grad_norm": 0.7141142156759234,
+      "learning_rate": 9.537665454570287e-06,
+      "loss": 0.3759,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16400705583131642,
+      "grad_norm": 0.6643782637755888,
+      "learning_rate": 9.537107263603666e-06,
+      "loss": 0.3976,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16408910038150715,
+      "grad_norm": 0.6615330866869896,
+      "learning_rate": 9.536548752233822e-06,
+      "loss": 0.3754,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1641711449316979,
+      "grad_norm": 0.6454987640312385,
+      "learning_rate": 9.535989920500192e-06,
+      "loss": 0.3779,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16425318948188866,
+      "grad_norm": 0.666721224712006,
+      "learning_rate": 9.535430768442243e-06,
+      "loss": 0.368,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16433523403207942,
+      "grad_norm": 0.6389178572387327,
+      "learning_rate": 9.534871296099457e-06,
+      "loss": 0.3944,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16441727858227018,
+      "grad_norm": 0.7317108104882305,
+      "learning_rate": 9.534311503511349e-06,
+      "loss": 0.4254,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16449932313246093,
+      "grad_norm": 0.6762410647157744,
+      "learning_rate": 9.533751390717445e-06,
+      "loss": 0.413,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1645813676826517,
+      "grad_norm": 0.6444889923973044,
+      "learning_rate": 9.533190957757304e-06,
+      "loss": 0.34,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16466341223284242,
+      "grad_norm": 0.6644876794096636,
+      "learning_rate": 9.532630204670498e-06,
+      "loss": 0.3852,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16474545678303318,
+      "grad_norm": 0.6724701353146899,
+      "learning_rate": 9.53206913149663e-06,
+      "loss": 0.3824,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16482750133322394,
+      "grad_norm": 0.8050560220734574,
+      "learning_rate": 9.531507738275322e-06,
+      "loss": 0.3677,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1649095458834147,
+      "grad_norm": 0.697766078198689,
+      "learning_rate": 9.53094602504622e-06,
+      "loss": 0.403,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16499159043360545,
+      "grad_norm": 0.6465449553753231,
+      "learning_rate": 9.530383991848986e-06,
+      "loss": 0.4098,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1650736349837962,
+      "grad_norm": 0.6301280259493615,
+      "learning_rate": 9.529821638723316e-06,
+      "loss": 0.3874,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16515567953398697,
+      "grad_norm": 0.6671963271781229,
+      "learning_rate": 9.529258965708916e-06,
+      "loss": 0.3944,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1652377240841777,
+      "grad_norm": 0.6370267983084557,
+      "learning_rate": 9.528695972845528e-06,
+      "loss": 0.3976,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16531976863436845,
+      "grad_norm": 0.6772344206516141,
+      "learning_rate": 9.528132660172904e-06,
+      "loss": 0.3799,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1654018131845592,
+      "grad_norm": 0.6947358861704559,
+      "learning_rate": 9.527569027730827e-06,
+      "loss": 0.3982,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16548385773474997,
+      "grad_norm": 0.6575675545886158,
+      "learning_rate": 9.5270050755591e-06,
+      "loss": 0.3853,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16556590228494072,
+      "grad_norm": 0.645696202976445,
+      "learning_rate": 9.526440803697548e-06,
+      "loss": 0.4106,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16564794683513148,
+      "grad_norm": 1.0451687351123693,
+      "learning_rate": 9.525876212186018e-06,
+      "loss": 0.3854,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16572999138532224,
+      "grad_norm": 0.6746753921879453,
+      "learning_rate": 9.525311301064382e-06,
+      "loss": 0.4157,
+      "step": 2020
+    },
+    {
+      "epoch": 0.165812035935513,
+      "grad_norm": 0.650482297441164,
+      "learning_rate": 9.524746070372532e-06,
+      "loss": 0.396,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16589408048570373,
+      "grad_norm": 0.6073364183742793,
+      "learning_rate": 9.524180520150383e-06,
+      "loss": 0.3997,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16597612503589448,
+      "grad_norm": 0.6592839151429709,
+      "learning_rate": 9.523614650437876e-06,
+      "loss": 0.3648,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16605816958608524,
+      "grad_norm": 0.5890399739925504,
+      "learning_rate": 9.523048461274968e-06,
+      "loss": 0.3719,
+      "step": 2024
+    },
+    {
+      "epoch": 0.166140214136276,
+      "grad_norm": 0.6456595556435875,
+      "learning_rate": 9.522481952701646e-06,
+      "loss": 0.3832,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16622225868646676,
+      "grad_norm": 0.635424696776647,
+      "learning_rate": 9.521915124757915e-06,
+      "loss": 0.3805,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1663043032366575,
+      "grad_norm": 0.7108972816665662,
+      "learning_rate": 9.521347977483802e-06,
+      "loss": 0.3903,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16638634778684827,
+      "grad_norm": 0.6790390324754764,
+      "learning_rate": 9.520780510919358e-06,
+      "loss": 0.3893,
+      "step": 2028
+    },
+    {
+      "epoch": 0.166468392337039,
+      "grad_norm": 0.6623001563476419,
+      "learning_rate": 9.52021272510466e-06,
+      "loss": 0.4347,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16655043688722976,
+      "grad_norm": 0.6756134243952981,
+      "learning_rate": 9.5196446200798e-06,
+      "loss": 0.3786,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16663248143742052,
+      "grad_norm": 0.6638170965607121,
+      "learning_rate": 9.519076195884898e-06,
+      "loss": 0.3826,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16671452598761127,
+      "grad_norm": 0.6593180172801646,
+      "learning_rate": 9.518507452560098e-06,
+      "loss": 0.3677,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16679657053780203,
+      "grad_norm": 0.6647054837471289,
+      "learning_rate": 9.517938390145557e-06,
+      "loss": 0.4122,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1668786150879928,
+      "grad_norm": 0.5681682591886665,
+      "learning_rate": 9.51736900868147e-06,
+      "loss": 0.3664,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16696065963818355,
+      "grad_norm": 1.4093112600287505,
+      "learning_rate": 9.516799308208038e-06,
+      "loss": 0.745,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16704270418837427,
+      "grad_norm": 0.6760218521077411,
+      "learning_rate": 9.516229288765497e-06,
+      "loss": 0.3723,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16712474873856503,
+      "grad_norm": 0.6352752496245146,
+      "learning_rate": 9.515658950394099e-06,
+      "loss": 0.3821,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1672067932887558,
+      "grad_norm": 0.690411557528179,
+      "learning_rate": 9.51508829313412e-06,
+      "loss": 0.4022,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16728883783894655,
+      "grad_norm": 0.6465252725924353,
+      "learning_rate": 9.514517317025862e-06,
+      "loss": 0.3712,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1673708823891373,
+      "grad_norm": 0.6041117795907213,
+      "learning_rate": 9.513946022109643e-06,
+      "loss": 0.3591,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16745292693932806,
+      "grad_norm": 0.6371733511640165,
+      "learning_rate": 9.513374408425809e-06,
+      "loss": 0.3989,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16753497148951882,
+      "grad_norm": 0.6645386425129249,
+      "learning_rate": 9.512802476014724e-06,
+      "loss": 0.3784,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16761701603970955,
+      "grad_norm": 0.6806526969839312,
+      "learning_rate": 9.512230224916779e-06,
+      "loss": 0.3816,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1676990605899003,
+      "grad_norm": 0.60840087199747,
+      "learning_rate": 9.511657655172387e-06,
+      "loss": 0.3879,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16778110514009106,
+      "grad_norm": 0.6853069552833762,
+      "learning_rate": 9.511084766821977e-06,
+      "loss": 0.3946,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16786314969028182,
+      "grad_norm": 0.6296207821916263,
+      "learning_rate": 9.51051155990601e-06,
+      "loss": 0.4033,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16794519424047258,
+      "grad_norm": 0.6188812020140116,
+      "learning_rate": 9.509938034464963e-06,
+      "loss": 0.3913,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16802723879066334,
+      "grad_norm": 0.6801244804332838,
+      "learning_rate": 9.509364190539337e-06,
+      "loss": 0.4168,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1681092833408541,
+      "grad_norm": 0.6499705079230528,
+      "learning_rate": 9.508790028169658e-06,
+      "loss": 0.4011,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16819132789104485,
+      "grad_norm": 0.5877648594634254,
+      "learning_rate": 9.508215547396471e-06,
+      "loss": 0.3923,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16827337244123558,
+      "grad_norm": 0.7362608193593966,
+      "learning_rate": 9.507640748260343e-06,
+      "loss": 0.4277,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16835541699142634,
+      "grad_norm": 0.6698426997993757,
+      "learning_rate": 9.50706563080187e-06,
+      "loss": 0.3793,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1684374615416171,
+      "grad_norm": 0.6148110396362916,
+      "learning_rate": 9.506490195061661e-06,
+      "loss": 0.4107,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16851950609180785,
+      "grad_norm": 0.6616398134284277,
+      "learning_rate": 9.505914441080357e-06,
+      "loss": 0.3866,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1686015506419986,
+      "grad_norm": 0.6520781994815675,
+      "learning_rate": 9.505338368898613e-06,
+      "loss": 0.3608,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16868359519218937,
+      "grad_norm": 0.5870593293940386,
+      "learning_rate": 9.504761978557114e-06,
+      "loss": 0.3479,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16876563974238012,
+      "grad_norm": 0.6978838600959042,
+      "learning_rate": 9.504185270096559e-06,
+      "loss": 0.4008,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16884768429257085,
+      "grad_norm": 0.7996099399003268,
+      "learning_rate": 9.503608243557677e-06,
+      "loss": 0.4062,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1689297288427616,
+      "grad_norm": 0.655076642423272,
+      "learning_rate": 9.503030898981218e-06,
+      "loss": 0.4023,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16901177339295237,
+      "grad_norm": 0.6682407743629372,
+      "learning_rate": 9.50245323640795e-06,
+      "loss": 0.456,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16909381794314313,
+      "grad_norm": 0.698400430902904,
+      "learning_rate": 9.501875255878668e-06,
+      "loss": 0.4156,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16917586249333388,
+      "grad_norm": 0.6414445837166823,
+      "learning_rate": 9.501296957434189e-06,
+      "loss": 0.4155,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16925790704352464,
+      "grad_norm": 0.7093097218178649,
+      "learning_rate": 9.500718341115351e-06,
+      "loss": 0.3927,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1693399515937154,
+      "grad_norm": 0.6625421664414551,
+      "learning_rate": 9.500139406963015e-06,
+      "loss": 0.3605,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16942199614390613,
+      "grad_norm": 0.6051510953902565,
+      "learning_rate": 9.499560155018063e-06,
+      "loss": 0.3838,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16950404069409689,
+      "grad_norm": 0.6208827235103149,
+      "learning_rate": 9.498980585321402e-06,
+      "loss": 0.4194,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16958608524428764,
+      "grad_norm": 0.6385513863172552,
+      "learning_rate": 9.498400697913961e-06,
+      "loss": 0.3624,
+      "step": 2067
+    },
+    {
+      "epoch": 0.1696681297944784,
+      "grad_norm": 0.6617461944382855,
+      "learning_rate": 9.497820492836688e-06,
+      "loss": 0.388,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16975017434466916,
+      "grad_norm": 0.7383865685022031,
+      "learning_rate": 9.497239970130561e-06,
+      "loss": 0.3684,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16983221889485992,
+      "grad_norm": 0.6631743716555812,
+      "learning_rate": 9.496659129836572e-06,
+      "loss": 0.3983,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16991426344505067,
+      "grad_norm": 0.6863668626219841,
+      "learning_rate": 9.49607797199574e-06,
+      "loss": 0.4097,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1699963079952414,
+      "grad_norm": 0.6754321401413568,
+      "learning_rate": 9.495496496649103e-06,
+      "loss": 0.391,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17007835254543216,
+      "grad_norm": 0.6535940494846231,
+      "learning_rate": 9.494914703837727e-06,
+      "loss": 0.4039,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17016039709562292,
+      "grad_norm": 0.6671388477676423,
+      "learning_rate": 9.494332593602698e-06,
+      "loss": 0.3832,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17024244164581367,
+      "grad_norm": 0.634333440070248,
+      "learning_rate": 9.493750165985119e-06,
+      "loss": 0.3663,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17032448619600443,
+      "grad_norm": 0.6674061082982202,
+      "learning_rate": 9.493167421026125e-06,
+      "loss": 0.3835,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1704065307461952,
+      "grad_norm": 0.6720198206757878,
+      "learning_rate": 9.492584358766865e-06,
+      "loss": 0.3846,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17048857529638595,
+      "grad_norm": 0.6622905305338308,
+      "learning_rate": 9.492000979248517e-06,
+      "loss": 0.3724,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1705706198465767,
+      "grad_norm": 0.6485427453735093,
+      "learning_rate": 9.491417282512276e-06,
+      "loss": 0.3771,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17065266439676743,
+      "grad_norm": 0.6374465809153983,
+      "learning_rate": 9.490833268599363e-06,
+      "loss": 0.353,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1707347089469582,
+      "grad_norm": 0.653688377009864,
+      "learning_rate": 9.49024893755102e-06,
+      "loss": 0.3435,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17081675349714895,
+      "grad_norm": 0.6814514445925304,
+      "learning_rate": 9.48966428940851e-06,
+      "loss": 0.4117,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1708987980473397,
+      "grad_norm": 0.6250403365020534,
+      "learning_rate": 9.489079324213122e-06,
+      "loss": 0.3967,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17098084259753046,
+      "grad_norm": 0.655347769870268,
+      "learning_rate": 9.488494042006164e-06,
+      "loss": 0.3759,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17106288714772122,
+      "grad_norm": 0.6684577211007403,
+      "learning_rate": 9.487908442828968e-06,
+      "loss": 0.4143,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17114493169791198,
+      "grad_norm": 0.6494042171804018,
+      "learning_rate": 9.487322526722888e-06,
+      "loss": 0.3549,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1712269762481027,
+      "grad_norm": 0.6639681353878347,
+      "learning_rate": 9.486736293729302e-06,
+      "loss": 0.3758,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17130902079829347,
+      "grad_norm": 0.6904734339139861,
+      "learning_rate": 9.486149743889606e-06,
+      "loss": 0.4048,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17139106534848422,
+      "grad_norm": 0.6399937429851352,
+      "learning_rate": 9.485562877245222e-06,
+      "loss": 0.3737,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17147310989867498,
+      "grad_norm": 0.6288069678474457,
+      "learning_rate": 9.484975693837594e-06,
+      "loss": 0.3873,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17155515444886574,
+      "grad_norm": 0.6198744146856983,
+      "learning_rate": 9.48438819370819e-06,
+      "loss": 0.3889,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1716371989990565,
+      "grad_norm": 0.6378273296083161,
+      "learning_rate": 9.483800376898496e-06,
+      "loss": 0.3749,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17171924354924725,
+      "grad_norm": 0.7231097687565606,
+      "learning_rate": 9.483212243450024e-06,
+      "loss": 0.4155,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17180128809943798,
+      "grad_norm": 0.7032520856798052,
+      "learning_rate": 9.482623793404304e-06,
+      "loss": 0.4249,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17188333264962874,
+      "grad_norm": 0.6243037195802119,
+      "learning_rate": 9.482035026802896e-06,
+      "loss": 0.402,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1719653771998195,
+      "grad_norm": 0.6718219616355926,
+      "learning_rate": 9.481445943687372e-06,
+      "loss": 0.4025,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17204742175001025,
+      "grad_norm": 0.6486948240579343,
+      "learning_rate": 9.48085654409934e-06,
+      "loss": 0.3775,
+      "step": 2097
+    },
+    {
+      "epoch": 0.172129466300201,
+      "grad_norm": 0.8442830609225564,
+      "learning_rate": 9.480266828080414e-06,
+      "loss": 0.3754,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17221151085039177,
+      "grad_norm": 0.6240556700854979,
+      "learning_rate": 9.479676795672243e-06,
+      "loss": 0.3985,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17229355540058253,
+      "grad_norm": 0.6103422874355373,
+      "learning_rate": 9.479086446916496e-06,
+      "loss": 0.3587,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17237559995077326,
+      "grad_norm": 0.6459806798684672,
+      "learning_rate": 9.47849578185486e-06,
+      "loss": 0.3651,
+      "step": 2101
+    },
+    {
+      "epoch": 0.172457644500964,
+      "grad_norm": 0.6284468720633719,
+      "learning_rate": 9.477904800529045e-06,
+      "loss": 0.3688,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17253968905115477,
+      "grad_norm": 0.6297547861872408,
+      "learning_rate": 9.477313502980789e-06,
+      "loss": 0.377,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17262173360134553,
+      "grad_norm": 0.6147160731840557,
+      "learning_rate": 9.476721889251847e-06,
+      "loss": 0.3884,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17270377815153629,
+      "grad_norm": 0.6443563456447778,
+      "learning_rate": 9.476129959383997e-06,
+      "loss": 0.3772,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17278582270172704,
+      "grad_norm": 0.6551181382776884,
+      "learning_rate": 9.47553771341904e-06,
+      "loss": 0.4077,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1728678672519178,
+      "grad_norm": 0.6528804253001774,
+      "learning_rate": 9.474945151398802e-06,
+      "loss": 0.4435,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17294991180210856,
+      "grad_norm": 0.653221198244301,
+      "learning_rate": 9.474352273365126e-06,
+      "loss": 0.3888,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1730319563522993,
+      "grad_norm": 0.6765208052350777,
+      "learning_rate": 9.47375907935988e-06,
+      "loss": 0.3662,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17311400090249005,
+      "grad_norm": 0.6940520653946005,
+      "learning_rate": 9.473165569424957e-06,
+      "loss": 0.4267,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1731960454526808,
+      "grad_norm": 0.5960889305488823,
+      "learning_rate": 9.472571743602268e-06,
+      "loss": 0.3425,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17327809000287156,
+      "grad_norm": 0.6229313662134756,
+      "learning_rate": 9.471977601933748e-06,
+      "loss": 0.377,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17336013455306232,
+      "grad_norm": 0.6157005510837348,
+      "learning_rate": 9.471383144461355e-06,
+      "loss": 0.3672,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17344217910325307,
+      "grad_norm": 0.6352097445820374,
+      "learning_rate": 9.470788371227066e-06,
+      "loss": 0.396,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17352422365344383,
+      "grad_norm": 0.6359464422477211,
+      "learning_rate": 9.470193282272886e-06,
+      "loss": 0.3726,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17360626820363456,
+      "grad_norm": 0.6466230855266695,
+      "learning_rate": 9.46959787764084e-06,
+      "loss": 0.3869,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17368831275382532,
+      "grad_norm": 0.6937326723104547,
+      "learning_rate": 9.469002157372971e-06,
+      "loss": 0.4248,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17377035730401608,
+      "grad_norm": 0.7227652417622037,
+      "learning_rate": 9.46840612151135e-06,
+      "loss": 0.4502,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17385240185420683,
+      "grad_norm": 0.9368103418132826,
+      "learning_rate": 9.467809770098069e-06,
+      "loss": 0.3789,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1739344464043976,
+      "grad_norm": 0.6559765320590938,
+      "learning_rate": 9.467213103175239e-06,
+      "loss": 0.3969,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17401649095458835,
+      "grad_norm": 0.7130207267757129,
+      "learning_rate": 9.466616120784996e-06,
+      "loss": 0.4373,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1740985355047791,
+      "grad_norm": 0.6231641242689382,
+      "learning_rate": 9.466018822969502e-06,
+      "loss": 0.4026,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17418058005496984,
+      "grad_norm": 0.6071613195841269,
+      "learning_rate": 9.465421209770931e-06,
+      "loss": 0.3914,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1742626246051606,
+      "grad_norm": 0.7122064943758393,
+      "learning_rate": 9.464823281231489e-06,
+      "loss": 0.4304,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17434466915535135,
+      "grad_norm": 0.6888877215274384,
+      "learning_rate": 9.4642250373934e-06,
+      "loss": 0.3896,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1744267137055421,
+      "grad_norm": 0.6670143879958849,
+      "learning_rate": 9.463626478298912e-06,
+      "loss": 0.4259,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17450875825573287,
+      "grad_norm": 0.6746645275724326,
+      "learning_rate": 9.463027603990293e-06,
+      "loss": 0.3686,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17459080280592362,
+      "grad_norm": 0.6903876575486265,
+      "learning_rate": 9.462428414509834e-06,
+      "loss": 0.3707,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17467284735611438,
+      "grad_norm": 0.6903866236498873,
+      "learning_rate": 9.461828909899851e-06,
+      "loss": 0.4032,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1747548919063051,
+      "grad_norm": 0.7190446531792516,
+      "learning_rate": 9.46122909020268e-06,
+      "loss": 0.3992,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17483693645649587,
+      "grad_norm": 0.6599878379913392,
+      "learning_rate": 9.460628955460676e-06,
+      "loss": 0.4284,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17491898100668662,
+      "grad_norm": 0.7220997676346619,
+      "learning_rate": 9.460028505716223e-06,
+      "loss": 0.3577,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17500102555687738,
+      "grad_norm": 0.6562808994793623,
+      "learning_rate": 9.459427741011722e-06,
+      "loss": 0.3605,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17508307010706814,
+      "grad_norm": 0.6011486287540881,
+      "learning_rate": 9.458826661389598e-06,
+      "loss": 0.3965,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1751651146572589,
+      "grad_norm": 0.6578600105842128,
+      "learning_rate": 9.4582252668923e-06,
+      "loss": 0.3958,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17524715920744965,
+      "grad_norm": 0.6875046296203334,
+      "learning_rate": 9.457623557562294e-06,
+      "loss": 0.3844,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1753292037576404,
+      "grad_norm": 0.6714086925935329,
+      "learning_rate": 9.457021533442077e-06,
+      "loss": 0.4127,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17541124830783114,
+      "grad_norm": 0.6432822982080353,
+      "learning_rate": 9.456419194574158e-06,
+      "loss": 0.4163,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1754932928580219,
+      "grad_norm": 0.6333231796941724,
+      "learning_rate": 9.455816541001076e-06,
+      "loss": 0.3708,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17557533740821266,
+      "grad_norm": 0.6589952753261158,
+      "learning_rate": 9.45521357276539e-06,
+      "loss": 0.371,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1756573819584034,
+      "grad_norm": 0.6704613264384965,
+      "learning_rate": 9.454610289909677e-06,
+      "loss": 0.3917,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17573942650859417,
+      "grad_norm": 0.635807791993477,
+      "learning_rate": 9.454006692476542e-06,
+      "loss": 0.369,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17582147105878493,
+      "grad_norm": 0.6616100436931275,
+      "learning_rate": 9.453402780508613e-06,
+      "loss": 0.3799,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17590351560897569,
+      "grad_norm": 0.6607526549309787,
+      "learning_rate": 9.452798554048533e-06,
+      "loss": 0.3963,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17598556015916642,
+      "grad_norm": 0.6887396052447419,
+      "learning_rate": 9.452194013138973e-06,
+      "loss": 0.3866,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17606760470935717,
+      "grad_norm": 0.7929722795295286,
+      "learning_rate": 9.451589157822625e-06,
+      "loss": 0.3962,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17614964925954793,
+      "grad_norm": 0.6341056491494356,
+      "learning_rate": 9.450983988142204e-06,
+      "loss": 0.3814,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1762316938097387,
+      "grad_norm": 0.6542684426685678,
+      "learning_rate": 9.450378504140442e-06,
+      "loss": 0.353,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17631373835992944,
+      "grad_norm": 0.6509365656240226,
+      "learning_rate": 9.449772705860103e-06,
+      "loss": 0.3583,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1763957829101202,
+      "grad_norm": 0.618367100191869,
+      "learning_rate": 9.449166593343963e-06,
+      "loss": 0.3478,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17647782746031096,
+      "grad_norm": 0.662083463356088,
+      "learning_rate": 9.448560166634825e-06,
+      "loss": 0.3607,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1765598720105017,
+      "grad_norm": 0.655070480347313,
+      "learning_rate": 9.447953425775517e-06,
+      "loss": 0.3857,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17664191656069245,
+      "grad_norm": 0.6623319019243954,
+      "learning_rate": 9.447346370808886e-06,
+      "loss": 0.3754,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1767239611108832,
+      "grad_norm": 0.6269102976922356,
+      "learning_rate": 9.446739001777797e-06,
+      "loss": 0.3936,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17680600566107396,
+      "grad_norm": 0.7102098290008985,
+      "learning_rate": 9.446131318725144e-06,
+      "loss": 0.4305,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17688805021126472,
+      "grad_norm": 0.6228675869218296,
+      "learning_rate": 9.445523321693841e-06,
+      "loss": 0.3979,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17697009476145548,
+      "grad_norm": 0.6466026366331475,
+      "learning_rate": 9.444915010726823e-06,
+      "loss": 0.4103,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17705213931164623,
+      "grad_norm": 0.6033933674811555,
+      "learning_rate": 9.444306385867048e-06,
+      "loss": 0.3648,
+      "step": 2158
+    },
+    {
+      "epoch": 0.177134183861837,
+      "grad_norm": 0.6566147770884809,
+      "learning_rate": 9.443697447157498e-06,
+      "loss": 0.4029,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17721622841202772,
+      "grad_norm": 0.7060382086212109,
+      "learning_rate": 9.443088194641173e-06,
+      "loss": 0.4132,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17729827296221848,
+      "grad_norm": 0.6485147074556397,
+      "learning_rate": 9.442478628361098e-06,
+      "loss": 0.3591,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17738031751240924,
+      "grad_norm": 0.6344716263267414,
+      "learning_rate": 9.441868748360318e-06,
+      "loss": 0.3788,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1774623620626,
+      "grad_norm": 0.736493312309787,
+      "learning_rate": 9.441258554681905e-06,
+      "loss": 0.4053,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17754440661279075,
+      "grad_norm": 0.5898464570465856,
+      "learning_rate": 9.440648047368947e-06,
+      "loss": 0.3871,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1776264511629815,
+      "grad_norm": 0.6842687449141741,
+      "learning_rate": 9.44003722646456e-06,
+      "loss": 0.3933,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17770849571317227,
+      "grad_norm": 0.6376928603520313,
+      "learning_rate": 9.439426092011877e-06,
+      "loss": 0.4111,
+      "step": 2166
+    },
+    {
+      "epoch": 0.177790540263363,
+      "grad_norm": 0.6949546092846008,
+      "learning_rate": 9.438814644054055e-06,
+      "loss": 0.4367,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17787258481355375,
+      "grad_norm": 0.6211866752560201,
+      "learning_rate": 9.438202882634275e-06,
+      "loss": 0.3709,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1779546293637445,
+      "grad_norm": 0.6129535788611599,
+      "learning_rate": 9.437590807795739e-06,
+      "loss": 0.368,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17803667391393527,
+      "grad_norm": 0.6605104676798506,
+      "learning_rate": 9.436978419581668e-06,
+      "loss": 0.3937,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17811871846412602,
+      "grad_norm": 0.6647572239361348,
+      "learning_rate": 9.436365718035309e-06,
+      "loss": 0.3722,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17820076301431678,
+      "grad_norm": 0.6277074233442591,
+      "learning_rate": 9.435752703199933e-06,
+      "loss": 0.3793,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17828280756450754,
+      "grad_norm": 1.824599905107891,
+      "learning_rate": 9.435139375118828e-06,
+      "loss": 0.7329,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17836485211469827,
+      "grad_norm": 0.7229824637649559,
+      "learning_rate": 9.434525733835303e-06,
+      "loss": 0.4478,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17844689666488903,
+      "grad_norm": 0.6637279868888083,
+      "learning_rate": 9.433911779392697e-06,
+      "loss": 0.4268,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17852894121507978,
+      "grad_norm": 0.6900268003365817,
+      "learning_rate": 9.433297511834365e-06,
+      "loss": 0.4217,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17861098576527054,
+      "grad_norm": 0.6622097783072212,
+      "learning_rate": 9.432682931203685e-06,
+      "loss": 0.3694,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1786930303154613,
+      "grad_norm": 0.6724116575076883,
+      "learning_rate": 9.432068037544056e-06,
+      "loss": 0.446,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17877507486565206,
+      "grad_norm": 0.6439994823881346,
+      "learning_rate": 9.431452830898906e-06,
+      "loss": 0.3767,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1788571194158428,
+      "grad_norm": 0.7168670934860905,
+      "learning_rate": 9.430837311311675e-06,
+      "loss": 0.3905,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17893916396603354,
+      "grad_norm": 0.7205890888156126,
+      "learning_rate": 9.430221478825832e-06,
+      "loss": 0.4314,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1790212085162243,
+      "grad_norm": 0.723417091544534,
+      "learning_rate": 9.429605333484866e-06,
+      "loss": 0.4139,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17910325306641506,
+      "grad_norm": 0.6645417683186658,
+      "learning_rate": 9.428988875332287e-06,
+      "loss": 0.3668,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17918529761660582,
+      "grad_norm": 0.6882091372275944,
+      "learning_rate": 9.428372104411632e-06,
+      "loss": 0.3724,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17926734216679657,
+      "grad_norm": 0.5792270665112401,
+      "learning_rate": 9.427755020766451e-06,
+      "loss": 0.3494,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17934938671698733,
+      "grad_norm": 0.6295093796940383,
+      "learning_rate": 9.427137624440325e-06,
+      "loss": 0.4406,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1794314312671781,
+      "grad_norm": 0.630594763016925,
+      "learning_rate": 9.426519915476854e-06,
+      "loss": 0.3901,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17951347581736884,
+      "grad_norm": 0.6791929483766356,
+      "learning_rate": 9.425901893919658e-06,
+      "loss": 0.4095,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17959552036755957,
+      "grad_norm": 0.6178873734258511,
+      "learning_rate": 9.425283559812379e-06,
+      "loss": 0.3679,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17967756491775033,
+      "grad_norm": 0.6370518953313021,
+      "learning_rate": 9.424664913198687e-06,
+      "loss": 0.4019,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1797596094679411,
+      "grad_norm": 0.6632717450535445,
+      "learning_rate": 9.424045954122267e-06,
+      "loss": 0.3964,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17984165401813185,
+      "grad_norm": 0.6920225652980718,
+      "learning_rate": 9.423426682626828e-06,
+      "loss": 0.3768,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1799236985683226,
+      "grad_norm": 0.6499908122947389,
+      "learning_rate": 9.422807098756105e-06,
+      "loss": 0.374,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18000574311851336,
+      "grad_norm": 0.6196275280355542,
+      "learning_rate": 9.42218720255385e-06,
+      "loss": 0.4067,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18008778766870412,
+      "grad_norm": 0.6190436076150186,
+      "learning_rate": 9.42156699406384e-06,
+      "loss": 0.3716,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18016983221889485,
+      "grad_norm": 0.6346424346874966,
+      "learning_rate": 9.420946473329872e-06,
+      "loss": 0.3987,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1802518767690856,
+      "grad_norm": 0.67746073696769,
+      "learning_rate": 9.420325640395767e-06,
+      "loss": 0.3642,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18033392131927636,
+      "grad_norm": 0.6017433317658702,
+      "learning_rate": 9.419704495305367e-06,
+      "loss": 0.3832,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18041596586946712,
+      "grad_norm": 0.7753335315988652,
+      "learning_rate": 9.419083038102537e-06,
+      "loss": 0.3594,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18049801041965788,
+      "grad_norm": 0.6837616948780358,
+      "learning_rate": 9.418461268831162e-06,
+      "loss": 0.396,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18058005496984864,
+      "grad_norm": 0.6502643563205313,
+      "learning_rate": 9.417839187535151e-06,
+      "loss": 0.3588,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1806620995200394,
+      "grad_norm": 0.6523448146213707,
+      "learning_rate": 9.417216794258435e-06,
+      "loss": 0.3561,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18074414407023012,
+      "grad_norm": 0.6685912226017888,
+      "learning_rate": 9.416594089044966e-06,
+      "loss": 0.3936,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18082618862042088,
+      "grad_norm": 0.6547343545914671,
+      "learning_rate": 9.415971071938717e-06,
+      "loss": 0.4101,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18090823317061164,
+      "grad_norm": 1.0548170795569114,
+      "learning_rate": 9.415347742983686e-06,
+      "loss": 0.3808,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1809902777208024,
+      "grad_norm": 0.7316710816740809,
+      "learning_rate": 9.41472410222389e-06,
+      "loss": 0.4106,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18107232227099315,
+      "grad_norm": 0.6603312746399724,
+      "learning_rate": 9.414100149703373e-06,
+      "loss": 0.3663,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1811543668211839,
+      "grad_norm": 0.7073172738802865,
+      "learning_rate": 9.413475885466195e-06,
+      "loss": 0.4044,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18123641137137467,
+      "grad_norm": 0.6663641540504266,
+      "learning_rate": 9.41285130955644e-06,
+      "loss": 0.3783,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1813184559215654,
+      "grad_norm": 0.6881091429795879,
+      "learning_rate": 9.412226422018215e-06,
+      "loss": 0.3783,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18140050047175615,
+      "grad_norm": 0.6733045992143493,
+      "learning_rate": 9.41160122289565e-06,
+      "loss": 0.394,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1814825450219469,
+      "grad_norm": 0.6146932072189493,
+      "learning_rate": 9.410975712232894e-06,
+      "loss": 0.3714,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18156458957213767,
+      "grad_norm": 0.632481681787621,
+      "learning_rate": 9.41034989007412e-06,
+      "loss": 0.3502,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18164663412232843,
+      "grad_norm": 0.6569777079337834,
+      "learning_rate": 9.409723756463521e-06,
+      "loss": 0.374,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18172867867251918,
+      "grad_norm": 0.6611920731383617,
+      "learning_rate": 9.409097311445317e-06,
+      "loss": 0.4016,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18181072322270994,
+      "grad_norm": 0.6467827985571128,
+      "learning_rate": 9.408470555063744e-06,
+      "loss": 0.3624,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1818927677729007,
+      "grad_norm": 0.646655663101853,
+      "learning_rate": 9.407843487363064e-06,
+      "loss": 0.3792,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18197481232309143,
+      "grad_norm": 0.6274145628844251,
+      "learning_rate": 9.407216108387556e-06,
+      "loss": 0.3488,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18205685687328219,
+      "grad_norm": 0.6220082768281531,
+      "learning_rate": 9.406588418181531e-06,
+      "loss": 0.3752,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18213890142347294,
+      "grad_norm": 0.6155965760256709,
+      "learning_rate": 9.405960416789307e-06,
+      "loss": 0.3534,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1822209459736637,
+      "grad_norm": 0.6823560486937664,
+      "learning_rate": 9.40533210425524e-06,
+      "loss": 0.4078,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18230299052385446,
+      "grad_norm": 0.6457696871254243,
+      "learning_rate": 9.404703480623697e-06,
+      "loss": 0.323,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18238503507404522,
+      "grad_norm": 0.6896538390069588,
+      "learning_rate": 9.404074545939073e-06,
+      "loss": 0.3688,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18246707962423597,
+      "grad_norm": 0.6195564738461724,
+      "learning_rate": 9.403445300245777e-06,
+      "loss": 0.3777,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1825491241744267,
+      "grad_norm": 0.6848346227998103,
+      "learning_rate": 9.40281574358825e-06,
+      "loss": 0.3837,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18263116872461746,
+      "grad_norm": 0.6576468898410852,
+      "learning_rate": 9.402185876010949e-06,
+      "loss": 0.4466,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18271321327480822,
+      "grad_norm": 0.6528249401591845,
+      "learning_rate": 9.401555697558353e-06,
+      "loss": 0.3872,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18279525782499897,
+      "grad_norm": 0.6796720253416653,
+      "learning_rate": 9.400925208274966e-06,
+      "loss": 0.388,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18287730237518973,
+      "grad_norm": 0.5902597180970633,
+      "learning_rate": 9.400294408205312e-06,
+      "loss": 0.3669,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1829593469253805,
+      "grad_norm": 0.6834302477887529,
+      "learning_rate": 9.399663297393937e-06,
+      "loss": 0.3968,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18304139147557125,
+      "grad_norm": 0.7071145276524656,
+      "learning_rate": 9.399031875885408e-06,
+      "loss": 0.3809,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18312343602576198,
+      "grad_norm": 0.6554898320757157,
+      "learning_rate": 9.398400143724313e-06,
+      "loss": 0.3522,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18320548057595273,
+      "grad_norm": 0.6567173127897523,
+      "learning_rate": 9.39776810095527e-06,
+      "loss": 0.4006,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1832875251261435,
+      "grad_norm": 0.5944488131242966,
+      "learning_rate": 9.397135747622907e-06,
+      "loss": 0.3469,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18336956967633425,
+      "grad_norm": 0.6637165438887703,
+      "learning_rate": 9.396503083771883e-06,
+      "loss": 0.3798,
+      "step": 2235
+    },
+    {
+      "epoch": 0.183451614226525,
+      "grad_norm": 0.6279468266552583,
+      "learning_rate": 9.395870109446874e-06,
+      "loss": 0.3495,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18353365877671576,
+      "grad_norm": 0.6158115677833944,
+      "learning_rate": 9.395236824692582e-06,
+      "loss": 0.4,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18361570332690652,
+      "grad_norm": 0.6619265761041224,
+      "learning_rate": 9.394603229553724e-06,
+      "loss": 0.3646,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18369774787709725,
+      "grad_norm": 0.6076963455303073,
+      "learning_rate": 9.39396932407505e-06,
+      "loss": 0.397,
+      "step": 2239
+    },
+    {
+      "epoch": 0.183779792427288,
+      "grad_norm": 0.699870375233772,
+      "learning_rate": 9.393335108301319e-06,
+      "loss": 0.3688,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18386183697747877,
+      "grad_norm": 0.6770508207131117,
+      "learning_rate": 9.392700582277322e-06,
+      "loss": 0.38,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18394388152766952,
+      "grad_norm": 0.6935104706403638,
+      "learning_rate": 9.392065746047867e-06,
+      "loss": 0.3692,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18402592607786028,
+      "grad_norm": 0.6858485591562076,
+      "learning_rate": 9.391430599657786e-06,
+      "loss": 0.3927,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18410797062805104,
+      "grad_norm": 0.7478212998293969,
+      "learning_rate": 9.390795143151932e-06,
+      "loss": 0.4174,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1841900151782418,
+      "grad_norm": 0.6434551146699471,
+      "learning_rate": 9.390159376575178e-06,
+      "loss": 0.3991,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18427205972843255,
+      "grad_norm": 0.6727178614128846,
+      "learning_rate": 9.389523299972422e-06,
+      "loss": 0.3999,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18435410427862328,
+      "grad_norm": 0.6407423966574375,
+      "learning_rate": 9.388886913388584e-06,
+      "loss": 0.3895,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18443614882881404,
+      "grad_norm": 0.6421328858081942,
+      "learning_rate": 9.388250216868605e-06,
+      "loss": 0.3763,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1845181933790048,
+      "grad_norm": 0.6595486341259511,
+      "learning_rate": 9.387613210457442e-06,
+      "loss": 0.3918,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18460023792919555,
+      "grad_norm": 0.6248532575258672,
+      "learning_rate": 9.386975894200085e-06,
+      "loss": 0.3681,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1846822824793863,
+      "grad_norm": 0.6479105870233962,
+      "learning_rate": 9.38633826814154e-06,
+      "loss": 0.4039,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18476432702957707,
+      "grad_norm": 0.9624248925461987,
+      "learning_rate": 9.385700332326833e-06,
+      "loss": 0.3808,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18484637157976783,
+      "grad_norm": 0.856002467069955,
+      "learning_rate": 9.385062086801013e-06,
+      "loss": 0.4139,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18492841612995856,
+      "grad_norm": 0.6268356099297339,
+      "learning_rate": 9.384423531609154e-06,
+      "loss": 0.3781,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1850104606801493,
+      "grad_norm": 0.6955185856448675,
+      "learning_rate": 9.383784666796351e-06,
+      "loss": 0.3924,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18509250523034007,
+      "grad_norm": 0.6887244874005455,
+      "learning_rate": 9.383145492407716e-06,
+      "loss": 0.3746,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18517454978053083,
+      "grad_norm": 0.7884194179745071,
+      "learning_rate": 9.38250600848839e-06,
+      "loss": 0.4097,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18525659433072159,
+      "grad_norm": 0.6243083294238491,
+      "learning_rate": 9.38186621508353e-06,
+      "loss": 0.3694,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18533863888091234,
+      "grad_norm": 0.8161845516123343,
+      "learning_rate": 9.38122611223832e-06,
+      "loss": 0.4328,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1854206834311031,
+      "grad_norm": 0.6605651214233957,
+      "learning_rate": 9.380585699997957e-06,
+      "loss": 0.386,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18550272798129383,
+      "grad_norm": 0.6252312026972461,
+      "learning_rate": 9.379944978407673e-06,
+      "loss": 0.3917,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1855847725314846,
+      "grad_norm": 0.6145930999132546,
+      "learning_rate": 9.37930394751271e-06,
+      "loss": 0.3685,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18566681708167534,
+      "grad_norm": 0.6367830376133443,
+      "learning_rate": 9.378662607358337e-06,
+      "loss": 0.388,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1857488616318661,
+      "grad_norm": 0.7091900224133999,
+      "learning_rate": 9.378020957989849e-06,
+      "loss": 0.4639,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18583090618205686,
+      "grad_norm": 0.6675615097627332,
+      "learning_rate": 9.377378999452553e-06,
+      "loss": 0.3668,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18591295073224762,
+      "grad_norm": 0.7273131848756691,
+      "learning_rate": 9.376736731791785e-06,
+      "loss": 0.401,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18599499528243837,
+      "grad_norm": 0.6356609381217574,
+      "learning_rate": 9.3760941550529e-06,
+      "loss": 0.3752,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1860770398326291,
+      "grad_norm": 0.7384666900295588,
+      "learning_rate": 9.375451269281276e-06,
+      "loss": 0.3688,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18615908438281986,
+      "grad_norm": 0.6932047602509236,
+      "learning_rate": 9.374808074522313e-06,
+      "loss": 0.3995,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18624112893301062,
+      "grad_norm": 0.6834944352460937,
+      "learning_rate": 9.374164570821433e-06,
+      "loss": 0.3805,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18632317348320138,
+      "grad_norm": 0.7277940075119779,
+      "learning_rate": 9.37352075822408e-06,
+      "loss": 0.4145,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18640521803339213,
+      "grad_norm": 0.6531854889773601,
+      "learning_rate": 9.372876636775716e-06,
+      "loss": 0.3835,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1864872625835829,
+      "grad_norm": 0.7024824429776769,
+      "learning_rate": 9.372232206521832e-06,
+      "loss": 0.3808,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18656930713377365,
+      "grad_norm": 1.000322208161805,
+      "learning_rate": 9.37158746750793e-06,
+      "loss": 0.3643,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1866513516839644,
+      "grad_norm": 0.6118338887591332,
+      "learning_rate": 9.370942419779547e-06,
+      "loss": 0.3461,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18673339623415514,
+      "grad_norm": 0.6173971655520525,
+      "learning_rate": 9.370297063382235e-06,
+      "loss": 0.3838,
+      "step": 2276
+    },
+    {
+      "epoch": 0.1868154407843459,
+      "grad_norm": 0.6511222196648068,
+      "learning_rate": 9.369651398361562e-06,
+      "loss": 0.4239,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18689748533453665,
+      "grad_norm": 0.678468967577286,
+      "learning_rate": 9.369005424763128e-06,
+      "loss": 0.4515,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1869795298847274,
+      "grad_norm": 0.6392780539789913,
+      "learning_rate": 9.368359142632553e-06,
+      "loss": 0.4065,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18706157443491817,
+      "grad_norm": 0.6127404658287932,
+      "learning_rate": 9.367712552015471e-06,
+      "loss": 0.3816,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18714361898510892,
+      "grad_norm": 0.6449515390467405,
+      "learning_rate": 9.367065652957546e-06,
+      "loss": 0.3731,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18722566353529968,
+      "grad_norm": 0.670470114390444,
+      "learning_rate": 9.366418445504461e-06,
+      "loss": 0.4252,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1873077080854904,
+      "grad_norm": 0.677813016240396,
+      "learning_rate": 9.365770929701924e-06,
+      "loss": 0.3954,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18738975263568117,
+      "grad_norm": 0.636291147168594,
+      "learning_rate": 9.365123105595655e-06,
+      "loss": 0.3835,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18747179718587192,
+      "grad_norm": 0.6837155224780271,
+      "learning_rate": 9.364474973231404e-06,
+      "loss": 0.3433,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18755384173606268,
+      "grad_norm": 0.6220463901083001,
+      "learning_rate": 9.363826532654945e-06,
+      "loss": 0.3896,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18763588628625344,
+      "grad_norm": 0.6691651982497119,
+      "learning_rate": 9.363177783912065e-06,
+      "loss": 0.3947,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1877179308364442,
+      "grad_norm": 0.6240319823007835,
+      "learning_rate": 9.362528727048581e-06,
+      "loss": 0.4017,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18779997538663495,
+      "grad_norm": 0.6493122935859134,
+      "learning_rate": 9.361879362110329e-06,
+      "loss": 0.3781,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18788201993682568,
+      "grad_norm": 0.6149076413458776,
+      "learning_rate": 9.361229689143162e-06,
+      "loss": 0.3995,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18796406448701644,
+      "grad_norm": 0.6733927035094591,
+      "learning_rate": 9.360579708192962e-06,
+      "loss": 0.3527,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1880461090372072,
+      "grad_norm": 0.6866269485961236,
+      "learning_rate": 9.359929419305628e-06,
+      "loss": 0.4123,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18812815358739796,
+      "grad_norm": 0.7172255950777638,
+      "learning_rate": 9.359278822527085e-06,
+      "loss": 0.4001,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1882101981375887,
+      "grad_norm": 0.656060969524315,
+      "learning_rate": 9.358627917903273e-06,
+      "loss": 0.3792,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18829224268777947,
+      "grad_norm": 0.6836726409283856,
+      "learning_rate": 9.357976705480162e-06,
+      "loss": 0.3411,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18837428723797023,
+      "grad_norm": 0.6703446505008837,
+      "learning_rate": 9.357325185303737e-06,
+      "loss": 0.4449,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18845633178816096,
+      "grad_norm": 0.7571872760995528,
+      "learning_rate": 9.356673357420007e-06,
+      "loss": 0.3811,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18853837633835172,
+      "grad_norm": 0.617408696652861,
+      "learning_rate": 9.356021221875008e-06,
+      "loss": 0.386,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18862042088854247,
+      "grad_norm": 0.6564785386615719,
+      "learning_rate": 9.355368778714784e-06,
+      "loss": 0.3741,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18870246543873323,
+      "grad_norm": 0.5501074663949519,
+      "learning_rate": 9.354716027985415e-06,
+      "loss": 0.3598,
+      "step": 2300
+    },
+    {
+      "epoch": 0.188784509988924,
+      "grad_norm": 0.6452070966471581,
+      "learning_rate": 9.354062969732998e-06,
+      "loss": 0.3644,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18886655453911474,
+      "grad_norm": 0.6993197176344447,
+      "learning_rate": 9.353409604003649e-06,
+      "loss": 0.4031,
+      "step": 2302
+    },
+    {
+      "epoch": 0.1889485990893055,
+      "grad_norm": 0.6750433056768211,
+      "learning_rate": 9.35275593084351e-06,
+      "loss": 0.3776,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18903064363949626,
+      "grad_norm": 0.6521842431324694,
+      "learning_rate": 9.352101950298738e-06,
+      "loss": 0.4258,
+      "step": 2304
+    },
+    {
+      "epoch": 0.189112688189687,
+      "grad_norm": 0.6692348451482892,
+      "learning_rate": 9.351447662415518e-06,
+      "loss": 0.4106,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18919473273987775,
+      "grad_norm": 0.6219592900216891,
+      "learning_rate": 9.350793067240057e-06,
+      "loss": 0.3824,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1892767772900685,
+      "grad_norm": 0.676789578999462,
+      "learning_rate": 9.350138164818579e-06,
+      "loss": 0.3978,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18935882184025926,
+      "grad_norm": 0.6263081341237644,
+      "learning_rate": 9.349482955197332e-06,
+      "loss": 0.3579,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18944086639045002,
+      "grad_norm": 0.6162221441845724,
+      "learning_rate": 9.348827438422588e-06,
+      "loss": 0.4295,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18952291094064078,
+      "grad_norm": 0.6318641776423526,
+      "learning_rate": 9.348171614540635e-06,
+      "loss": 0.3619,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18960495549083153,
+      "grad_norm": 0.6654406810458581,
+      "learning_rate": 9.34751548359779e-06,
+      "loss": 0.4025,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18968700004102226,
+      "grad_norm": 0.6344833745203432,
+      "learning_rate": 9.346859045640386e-06,
+      "loss": 0.3862,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18976904459121302,
+      "grad_norm": 0.6446421225136829,
+      "learning_rate": 9.346202300714781e-06,
+      "loss": 0.3531,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18985108914140378,
+      "grad_norm": 0.6313985964154566,
+      "learning_rate": 9.345545248867352e-06,
+      "loss": 0.3829,
+      "step": 2314
+    },
+    {
+      "epoch": 0.18993313369159454,
+      "grad_norm": 0.6200295342403026,
+      "learning_rate": 9.344887890144499e-06,
+      "loss": 0.3748,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1900151782417853,
+      "grad_norm": 1.105055352895138,
+      "learning_rate": 9.344230224592642e-06,
+      "loss": 0.3631,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19009722279197605,
+      "grad_norm": 0.6157647470344825,
+      "learning_rate": 9.343572252258228e-06,
+      "loss": 0.3969,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1901792673421668,
+      "grad_norm": 0.6661407728938531,
+      "learning_rate": 9.34291397318772e-06,
+      "loss": 0.4063,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19026131189235754,
+      "grad_norm": 0.6703460683806264,
+      "learning_rate": 9.342255387427604e-06,
+      "loss": 0.3727,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1903433564425483,
+      "grad_norm": 0.6617965736360988,
+      "learning_rate": 9.34159649502439e-06,
+      "loss": 0.379,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19042540099273905,
+      "grad_norm": 0.6466904126275137,
+      "learning_rate": 9.340937296024605e-06,
+      "loss": 0.3582,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1905074455429298,
+      "grad_norm": 0.6241798643953657,
+      "learning_rate": 9.340277790474804e-06,
+      "loss": 0.3548,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19058949009312057,
+      "grad_norm": 0.6142666862581382,
+      "learning_rate": 9.339617978421558e-06,
+      "loss": 0.4049,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19067153464331132,
+      "grad_norm": 0.6103447822310806,
+      "learning_rate": 9.338957859911462e-06,
+      "loss": 0.4203,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19075357919350208,
+      "grad_norm": 0.7373588364415218,
+      "learning_rate": 9.338297434991134e-06,
+      "loss": 0.3641,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1908356237436928,
+      "grad_norm": 0.6487574283133665,
+      "learning_rate": 9.33763670370721e-06,
+      "loss": 0.3988,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19091766829388357,
+      "grad_norm": 0.6258202448480539,
+      "learning_rate": 9.336975666106353e-06,
+      "loss": 0.3709,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19099971284407433,
+      "grad_norm": 0.6526014575649657,
+      "learning_rate": 9.33631432223524e-06,
+      "loss": 0.3848,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19108175739426508,
+      "grad_norm": 0.6077719393611979,
+      "learning_rate": 9.335652672140577e-06,
+      "loss": 0.3999,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19116380194445584,
+      "grad_norm": 0.6530432971866037,
+      "learning_rate": 9.33499071586909e-06,
+      "loss": 0.435,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1912458464946466,
+      "grad_norm": 0.6852933491871285,
+      "learning_rate": 9.334328453467521e-06,
+      "loss": 0.3653,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19132789104483736,
+      "grad_norm": 0.5951626098623948,
+      "learning_rate": 9.333665884982642e-06,
+      "loss": 0.3554,
+      "step": 2332
+    },
+    {
+      "epoch": 0.1914099355950281,
+      "grad_norm": 0.6894792253496951,
+      "learning_rate": 9.333003010461238e-06,
+      "loss": 0.4332,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19149198014521884,
+      "grad_norm": 0.6670601335952702,
+      "learning_rate": 9.332339829950126e-06,
+      "loss": 0.4233,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1915740246954096,
+      "grad_norm": 0.6337882942508766,
+      "learning_rate": 9.331676343496135e-06,
+      "loss": 0.4109,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19165606924560036,
+      "grad_norm": 0.6501163921099293,
+      "learning_rate": 9.331012551146118e-06,
+      "loss": 0.4293,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19173811379579112,
+      "grad_norm": 0.6429138918948967,
+      "learning_rate": 9.330348452946954e-06,
+      "loss": 0.3442,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19182015834598187,
+      "grad_norm": 0.6513355939967174,
+      "learning_rate": 9.329684048945539e-06,
+      "loss": 0.3874,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19190220289617263,
+      "grad_norm": 0.6791273565491414,
+      "learning_rate": 9.329019339188794e-06,
+      "loss": 0.431,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1919842474463634,
+      "grad_norm": 0.6672639844251973,
+      "learning_rate": 9.328354323723659e-06,
+      "loss": 0.425,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19206629199655412,
+      "grad_norm": 0.6235551051809817,
+      "learning_rate": 9.327689002597093e-06,
+      "loss": 0.371,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19214833654674487,
+      "grad_norm": 0.6012436315626537,
+      "learning_rate": 9.327023375856085e-06,
+      "loss": 0.3625,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19223038109693563,
+      "grad_norm": 0.645288267973353,
+      "learning_rate": 9.326357443547636e-06,
+      "loss": 0.3514,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1923124256471264,
+      "grad_norm": 0.6363989557059667,
+      "learning_rate": 9.325691205718777e-06,
+      "loss": 0.3767,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19239447019731715,
+      "grad_norm": 0.660560340225214,
+      "learning_rate": 9.325024662416553e-06,
+      "loss": 0.3657,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1924765147475079,
+      "grad_norm": 0.6721760596889514,
+      "learning_rate": 9.324357813688037e-06,
+      "loss": 0.37,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19255855929769866,
+      "grad_norm": 0.6834240895159485,
+      "learning_rate": 9.32369065958032e-06,
+      "loss": 0.4077,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1926406038478894,
+      "grad_norm": 0.6075540351343719,
+      "learning_rate": 9.323023200140516e-06,
+      "loss": 0.3588,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19272264839808015,
+      "grad_norm": 0.5984516730170012,
+      "learning_rate": 9.322355435415758e-06,
+      "loss": 0.3796,
+      "step": 2349
+    },
+    {
+      "epoch": 0.1928046929482709,
+      "grad_norm": 0.7086851352381045,
+      "learning_rate": 9.321687365453202e-06,
+      "loss": 0.4025,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19288673749846166,
+      "grad_norm": 0.6986547697987913,
+      "learning_rate": 9.32101899030003e-06,
+      "loss": 0.384,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19296878204865242,
+      "grad_norm": 0.6773073365420988,
+      "learning_rate": 9.320350310003438e-06,
+      "loss": 0.4185,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19305082659884318,
+      "grad_norm": 0.6432950407371643,
+      "learning_rate": 9.31968132461065e-06,
+      "loss": 0.3788,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19313287114903394,
+      "grad_norm": 0.6075904401925407,
+      "learning_rate": 9.319012034168905e-06,
+      "loss": 0.3801,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19321491569922467,
+      "grad_norm": 0.6439171969927543,
+      "learning_rate": 9.318342438725469e-06,
+      "loss": 0.3972,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19329696024941542,
+      "grad_norm": 0.6469349463673484,
+      "learning_rate": 9.31767253832763e-06,
+      "loss": 0.4119,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19337900479960618,
+      "grad_norm": 0.6343675366354676,
+      "learning_rate": 9.317002333022692e-06,
+      "loss": 0.4078,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19346104934979694,
+      "grad_norm": 0.5995391965320391,
+      "learning_rate": 9.316331822857984e-06,
+      "loss": 0.3951,
+      "step": 2358
+    },
+    {
+      "epoch": 0.1935430938999877,
+      "grad_norm": 0.6118923486469612,
+      "learning_rate": 9.315661007880858e-06,
+      "loss": 0.3713,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19362513845017845,
+      "grad_norm": 0.6073245728495696,
+      "learning_rate": 9.314989888138687e-06,
+      "loss": 0.3838,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1937071830003692,
+      "grad_norm": 0.6018354857740531,
+      "learning_rate": 9.31431846367886e-06,
+      "loss": 0.3575,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19378922755055997,
+      "grad_norm": 0.6245305670153241,
+      "learning_rate": 9.313646734548797e-06,
+      "loss": 0.3992,
+      "step": 2362
+    },
+    {
+      "epoch": 0.1938712721007507,
+      "grad_norm": 0.666903812854981,
+      "learning_rate": 9.312974700795932e-06,
+      "loss": 0.3795,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19395331665094145,
+      "grad_norm": 0.6582225024748427,
+      "learning_rate": 9.31230236246772e-06,
+      "loss": 0.3955,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1940353612011322,
+      "grad_norm": 0.6691318717690604,
+      "learning_rate": 9.311629719611645e-06,
+      "loss": 0.3921,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19411740575132297,
+      "grad_norm": 0.6072426756093179,
+      "learning_rate": 9.310956772275208e-06,
+      "loss": 0.3727,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19419945030151373,
+      "grad_norm": 0.66605317716138,
+      "learning_rate": 9.310283520505927e-06,
+      "loss": 0.3929,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19428149485170448,
+      "grad_norm": 0.636609286105985,
+      "learning_rate": 9.30960996435135e-06,
+      "loss": 0.3729,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19436353940189524,
+      "grad_norm": 0.6751185362791333,
+      "learning_rate": 9.308936103859041e-06,
+      "loss": 0.3946,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19444558395208597,
+      "grad_norm": 0.6366011292783129,
+      "learning_rate": 9.308261939076587e-06,
+      "loss": 0.3851,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19452762850227673,
+      "grad_norm": 0.6875852001728366,
+      "learning_rate": 9.3075874700516e-06,
+      "loss": 0.3858,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19460967305246749,
+      "grad_norm": 0.6399271283474355,
+      "learning_rate": 9.306912696831702e-06,
+      "loss": 0.393,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19469171760265824,
+      "grad_norm": 0.6527654766933841,
+      "learning_rate": 9.306237619464551e-06,
+      "loss": 0.3749,
+      "step": 2373
+    },
+    {
+      "epoch": 0.194773762152849,
+      "grad_norm": 0.6206295270930102,
+      "learning_rate": 9.305562237997819e-06,
+      "loss": 0.3581,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19485580670303976,
+      "grad_norm": 0.684269813194962,
+      "learning_rate": 9.304886552479196e-06,
+      "loss": 0.3984,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19493785125323052,
+      "grad_norm": 0.5669441430237632,
+      "learning_rate": 9.304210562956403e-06,
+      "loss": 0.4198,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19501989580342124,
+      "grad_norm": 0.6005508642710212,
+      "learning_rate": 9.303534269477174e-06,
+      "loss": 0.3938,
+      "step": 2377
+    },
+    {
+      "epoch": 0.195101940353612,
+      "grad_norm": 0.6283032062849251,
+      "learning_rate": 9.302857672089272e-06,
+      "loss": 0.3942,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19518398490380276,
+      "grad_norm": 0.6652222170410326,
+      "learning_rate": 9.302180770840471e-06,
+      "loss": 0.3635,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19526602945399352,
+      "grad_norm": 0.7343818972158671,
+      "learning_rate": 9.301503565778578e-06,
+      "loss": 0.3863,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19534807400418427,
+      "grad_norm": 0.6742437382952083,
+      "learning_rate": 9.300826056951412e-06,
+      "loss": 0.402,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19543011855437503,
+      "grad_norm": 0.6968039245131766,
+      "learning_rate": 9.30014824440682e-06,
+      "loss": 0.3632,
+      "step": 2382
+    },
+    {
+      "epoch": 0.1955121631045658,
+      "grad_norm": 0.6279828665406602,
+      "learning_rate": 9.299470128192671e-06,
+      "loss": 0.3414,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19559420765475652,
+      "grad_norm": 0.6983501511485233,
+      "learning_rate": 9.298791708356845e-06,
+      "loss": 0.3907,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19567625220494728,
+      "grad_norm": 0.6623804825395393,
+      "learning_rate": 9.298112984947258e-06,
+      "loss": 0.3933,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19575829675513803,
+      "grad_norm": 0.589850216098204,
+      "learning_rate": 9.297433958011836e-06,
+      "loss": 0.377,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1958403413053288,
+      "grad_norm": 0.6997192198506152,
+      "learning_rate": 9.296754627598533e-06,
+      "loss": 0.351,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19592238585551955,
+      "grad_norm": 0.6148879349719744,
+      "learning_rate": 9.296074993755321e-06,
+      "loss": 0.4118,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1960044304057103,
+      "grad_norm": 0.6140273467884046,
+      "learning_rate": 9.295395056530195e-06,
+      "loss": 0.3765,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19608647495590106,
+      "grad_norm": 0.6433561116710074,
+      "learning_rate": 9.294714815971171e-06,
+      "loss": 0.3664,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19616851950609182,
+      "grad_norm": 0.6878520688476624,
+      "learning_rate": 9.294034272126286e-06,
+      "loss": 0.393,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19625056405628255,
+      "grad_norm": 0.6625706750801742,
+      "learning_rate": 9.293353425043602e-06,
+      "loss": 0.3712,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1963326086064733,
+      "grad_norm": 0.6302656156108761,
+      "learning_rate": 9.292672274771195e-06,
+      "loss": 0.3489,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19641465315666407,
+      "grad_norm": 0.6205742578498861,
+      "learning_rate": 9.291990821357169e-06,
+      "loss": 0.3996,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19649669770685482,
+      "grad_norm": 0.6851257416146276,
+      "learning_rate": 9.291309064849647e-06,
+      "loss": 0.3935,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19657874225704558,
+      "grad_norm": 0.6407958944446134,
+      "learning_rate": 9.290627005296774e-06,
+      "loss": 0.3346,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19666078680723634,
+      "grad_norm": 0.6542314536246738,
+      "learning_rate": 9.289944642746716e-06,
+      "loss": 0.3925,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1967428313574271,
+      "grad_norm": 0.6741693286973981,
+      "learning_rate": 9.289261977247657e-06,
+      "loss": 0.38,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19682487590761782,
+      "grad_norm": 0.6600185965876197,
+      "learning_rate": 9.288579008847808e-06,
+      "loss": 0.3942,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19690692045780858,
+      "grad_norm": 0.6500004989772018,
+      "learning_rate": 9.287895737595402e-06,
+      "loss": 0.3668,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19698896500799934,
+      "grad_norm": 0.6578815761528988,
+      "learning_rate": 9.287212163538688e-06,
+      "loss": 0.3987,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1970710095581901,
+      "grad_norm": 0.597979129506302,
+      "learning_rate": 9.286528286725938e-06,
+      "loss": 0.3674,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19715305410838085,
+      "grad_norm": 0.7644720592103794,
+      "learning_rate": 9.285844107205449e-06,
+      "loss": 0.3876,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1972350986585716,
+      "grad_norm": 1.6390271640978227,
+      "learning_rate": 9.285159625025532e-06,
+      "loss": 0.7827,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19731714320876237,
+      "grad_norm": 0.5894560510348452,
+      "learning_rate": 9.284474840234528e-06,
+      "loss": 0.3266,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1973991877589531,
+      "grad_norm": 0.6591760439526413,
+      "learning_rate": 9.283789752880794e-06,
+      "loss": 0.3875,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19748123230914386,
+      "grad_norm": 0.6727596787251703,
+      "learning_rate": 9.283104363012712e-06,
+      "loss": 0.389,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1975632768593346,
+      "grad_norm": 0.6894187916010971,
+      "learning_rate": 9.282418670678679e-06,
+      "loss": 0.3925,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19764532140952537,
+      "grad_norm": 0.9338308791000948,
+      "learning_rate": 9.28173267592712e-06,
+      "loss": 0.4233,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19772736595971613,
+      "grad_norm": 0.6866554370071849,
+      "learning_rate": 9.28104637880648e-06,
+      "loss": 0.3758,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19780941050990689,
+      "grad_norm": 0.8298722722337927,
+      "learning_rate": 9.280359779365223e-06,
+      "loss": 0.3763,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19789145506009764,
+      "grad_norm": 0.6099956718290087,
+      "learning_rate": 9.279672877651833e-06,
+      "loss": 0.3843,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1979734996102884,
+      "grad_norm": 0.7471474555886637,
+      "learning_rate": 9.278985673714822e-06,
+      "loss": 0.3958,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19805554416047913,
+      "grad_norm": 0.7018924458043543,
+      "learning_rate": 9.278298167602716e-06,
+      "loss": 0.4429,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1981375887106699,
+      "grad_norm": 0.634952444708645,
+      "learning_rate": 9.277610359364068e-06,
+      "loss": 0.3703,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19821963326086064,
+      "grad_norm": 0.6768139062394007,
+      "learning_rate": 9.276922249047449e-06,
+      "loss": 0.4016,
+      "step": 2416
+    },
+    {
+      "epoch": 0.1983016778110514,
+      "grad_norm": 0.7026517447161629,
+      "learning_rate": 9.276233836701452e-06,
+      "loss": 0.4129,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19838372236124216,
+      "grad_norm": 0.7045179946729204,
+      "learning_rate": 9.275545122374691e-06,
+      "loss": 0.3915,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19846576691143292,
+      "grad_norm": 0.6301978495651493,
+      "learning_rate": 9.274856106115804e-06,
+      "loss": 0.419,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19854781146162367,
+      "grad_norm": 0.6612030207215627,
+      "learning_rate": 9.274166787973447e-06,
+      "loss": 0.397,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1986298560118144,
+      "grad_norm": 0.6404455848892218,
+      "learning_rate": 9.273477167996297e-06,
+      "loss": 0.3711,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19871190056200516,
+      "grad_norm": 0.6372516432575472,
+      "learning_rate": 9.272787246233055e-06,
+      "loss": 0.3486,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19879394511219592,
+      "grad_norm": 0.6923049597965895,
+      "learning_rate": 9.272097022732444e-06,
+      "loss": 0.431,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19887598966238668,
+      "grad_norm": 0.6595526316435263,
+      "learning_rate": 9.271406497543203e-06,
+      "loss": 0.4186,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19895803421257743,
+      "grad_norm": 0.640709629763413,
+      "learning_rate": 9.2707156707141e-06,
+      "loss": 0.3756,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1990400787627682,
+      "grad_norm": 0.7600656638226995,
+      "learning_rate": 9.270024542293917e-06,
+      "loss": 0.4083,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19912212331295895,
+      "grad_norm": 0.5768666273826845,
+      "learning_rate": 9.26933311233146e-06,
+      "loss": 0.3588,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19920416786314968,
+      "grad_norm": 0.6530369856226637,
+      "learning_rate": 9.268641380875558e-06,
+      "loss": 0.4078,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19928621241334044,
+      "grad_norm": 0.7542791332480379,
+      "learning_rate": 9.26794934797506e-06,
+      "loss": 0.3885,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1993682569635312,
+      "grad_norm": 0.6093758525387879,
+      "learning_rate": 9.267257013678838e-06,
+      "loss": 0.3776,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19945030151372195,
+      "grad_norm": 0.728821713930413,
+      "learning_rate": 9.266564378035778e-06,
+      "loss": 0.4059,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1995323460639127,
+      "grad_norm": 0.7797850821447573,
+      "learning_rate": 9.2658714410948e-06,
+      "loss": 0.3757,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19961439061410347,
+      "grad_norm": 0.76934715671421,
+      "learning_rate": 9.265178202904832e-06,
+      "loss": 0.4198,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19969643516429422,
+      "grad_norm": 0.9146614702592466,
+      "learning_rate": 9.264484663514832e-06,
+      "loss": 0.3661,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19977847971448495,
+      "grad_norm": 1.6643746278538945,
+      "learning_rate": 9.263790822973778e-06,
+      "loss": 0.3776,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1998605242646757,
+      "grad_norm": 7.341172149670472,
+      "learning_rate": 9.263096681330665e-06,
+      "loss": 0.3903,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19994256881486647,
+      "grad_norm": 2.723318094343984,
+      "learning_rate": 9.262402238634514e-06,
+      "loss": 0.3742,
+      "step": 2437
+    },
+    {
+      "epoch": 0.20002461336505722,
+      "grad_norm": 0.6625932061133696,
+      "learning_rate": 9.261707494934365e-06,
+      "loss": 0.3904,
+      "step": 2438
+    },
+    {
+      "epoch": 0.20010665791524798,
+      "grad_norm": 0.68402025879098,
+      "learning_rate": 9.26101245027928e-06,
+      "loss": 0.431,
+      "step": 2439
+    },
+    {
+      "epoch": 0.20018870246543874,
+      "grad_norm": 1.4648379743506206,
+      "learning_rate": 9.260317104718341e-06,
+      "loss": 0.3804,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2002707470156295,
+      "grad_norm": 0.6333649909430112,
+      "learning_rate": 9.259621458300653e-06,
+      "loss": 0.3737,
+      "step": 2441
+    },
+    {
+      "epoch": 0.20035279156582025,
+      "grad_norm": 0.6133769772102693,
+      "learning_rate": 9.258925511075342e-06,
+      "loss": 0.4065,
+      "step": 2442
+    },
+    {
+      "epoch": 0.20043483611601098,
+      "grad_norm": 0.6167993518472847,
+      "learning_rate": 9.258229263091554e-06,
+      "loss": 0.3661,
+      "step": 2443
+    },
+    {
+      "epoch": 0.20051688066620174,
+      "grad_norm": 0.5962270933531933,
+      "learning_rate": 9.257532714398457e-06,
+      "loss": 0.3607,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2005989252163925,
+      "grad_norm": 0.6119506230590925,
+      "learning_rate": 9.256835865045237e-06,
+      "loss": 0.3544,
+      "step": 2445
+    },
+    {
+      "epoch": 0.20068096976658326,
+      "grad_norm": 0.6984681789431609,
+      "learning_rate": 9.25613871508111e-06,
+      "loss": 0.3635,
+      "step": 2446
+    },
+    {
+      "epoch": 0.200763014316774,
+      "grad_norm": 0.6734108091549181,
+      "learning_rate": 9.255441264555303e-06,
+      "loss": 0.4047,
+      "step": 2447
+    },
+    {
+      "epoch": 0.20084505886696477,
+      "grad_norm": 0.7021954445146551,
+      "learning_rate": 9.254743513517072e-06,
+      "loss": 0.3714,
+      "step": 2448
+    },
+    {
+      "epoch": 0.20092710341715553,
+      "grad_norm": 0.6817109934966334,
+      "learning_rate": 9.254045462015691e-06,
+      "loss": 0.4064,
+      "step": 2449
+    },
+    {
+      "epoch": 0.20100914796734626,
+      "grad_norm": 0.6599785957237947,
+      "learning_rate": 9.253347110100453e-06,
+      "loss": 0.3578,
+      "step": 2450
+    },
+    {
+      "epoch": 0.20109119251753702,
+      "grad_norm": 0.650214581742734,
+      "learning_rate": 9.252648457820676e-06,
+      "loss": 0.4133,
+      "step": 2451
+    },
+    {
+      "epoch": 0.20117323706772777,
+      "grad_norm": 0.6743089185953098,
+      "learning_rate": 9.251949505225698e-06,
+      "loss": 0.3503,
+      "step": 2452
+    },
+    {
+      "epoch": 0.20125528161791853,
+      "grad_norm": 0.6210238181287334,
+      "learning_rate": 9.251250252364873e-06,
+      "loss": 0.3698,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2013373261681093,
+      "grad_norm": 0.6748403622548241,
+      "learning_rate": 9.25055069928759e-06,
+      "loss": 0.4017,
+      "step": 2454
+    },
+    {
+      "epoch": 0.20141937071830004,
+      "grad_norm": 0.619584121209874,
+      "learning_rate": 9.249850846043244e-06,
+      "loss": 0.3748,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2015014152684908,
+      "grad_norm": 0.6702978661475986,
+      "learning_rate": 9.24915069268126e-06,
+      "loss": 0.3908,
+      "step": 2456
+    },
+    {
+      "epoch": 0.20158345981868153,
+      "grad_norm": 0.6149596959206236,
+      "learning_rate": 9.24845023925108e-06,
+      "loss": 0.3639,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2016655043688723,
+      "grad_norm": 0.6870720441157786,
+      "learning_rate": 9.24774948580217e-06,
+      "loss": 0.3702,
+      "step": 2458
+    },
+    {
+      "epoch": 0.20174754891906305,
+      "grad_norm": 0.603141165689622,
+      "learning_rate": 9.247048432384015e-06,
+      "loss": 0.3689,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2018295934692538,
+      "grad_norm": 0.8007614036681425,
+      "learning_rate": 9.246347079046124e-06,
+      "loss": 0.4183,
+      "step": 2460
+    },
+    {
+      "epoch": 0.20191163801944456,
+      "grad_norm": 0.6421298279910205,
+      "learning_rate": 9.245645425838025e-06,
+      "loss": 0.4201,
+      "step": 2461
+    },
+    {
+      "epoch": 0.20199368256963532,
+      "grad_norm": 0.6448477097517213,
+      "learning_rate": 9.244943472809267e-06,
+      "loss": 0.3631,
+      "step": 2462
+    },
+    {
+      "epoch": 0.20207572711982608,
+      "grad_norm": 0.5993308014050052,
+      "learning_rate": 9.24424122000942e-06,
+      "loss": 0.3648,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2021577716700168,
+      "grad_norm": 0.6428438234885919,
+      "learning_rate": 9.243538667488076e-06,
+      "loss": 0.3703,
+      "step": 2464
+    },
+    {
+      "epoch": 0.20223981622020756,
+      "grad_norm": 0.6198728433843121,
+      "learning_rate": 9.242835815294851e-06,
+      "loss": 0.3719,
+      "step": 2465
+    },
+    {
+      "epoch": 0.20232186077039832,
+      "grad_norm": 0.599305199704726,
+      "learning_rate": 9.242132663479378e-06,
+      "loss": 0.3663,
+      "step": 2466
+    },
+    {
+      "epoch": 0.20240390532058908,
+      "grad_norm": 0.6643431379381497,
+      "learning_rate": 9.24142921209131e-06,
+      "loss": 0.3765,
+      "step": 2467
+    },
+    {
+      "epoch": 0.20248594987077984,
+      "grad_norm": 0.6445376697161935,
+      "learning_rate": 9.240725461180327e-06,
+      "loss": 0.3914,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2025679944209706,
+      "grad_norm": 0.5847350731032054,
+      "learning_rate": 9.240021410796125e-06,
+      "loss": 0.3351,
+      "step": 2469
+    },
+    {
+      "epoch": 0.20265003897116135,
+      "grad_norm": 0.6145280301010878,
+      "learning_rate": 9.239317060988422e-06,
+      "loss": 0.3793,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2027320835213521,
+      "grad_norm": 0.6004232800873174,
+      "learning_rate": 9.238612411806959e-06,
+      "loss": 0.3767,
+      "step": 2471
+    },
+    {
+      "epoch": 0.20281412807154284,
+      "grad_norm": 0.6358736405047657,
+      "learning_rate": 9.237907463301499e-06,
+      "loss": 0.3914,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2028961726217336,
+      "grad_norm": 0.7328805760166421,
+      "learning_rate": 9.237202215521822e-06,
+      "loss": 0.4055,
+      "step": 2473
+    },
+    {
+      "epoch": 0.20297821717192435,
+      "grad_norm": 0.6417973166510093,
+      "learning_rate": 9.236496668517731e-06,
+      "loss": 0.4082,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2030602617221151,
+      "grad_norm": 0.6276167169465922,
+      "learning_rate": 9.235790822339052e-06,
+      "loss": 0.3863,
+      "step": 2475
+    },
+    {
+      "epoch": 0.20314230627230587,
+      "grad_norm": 0.6697461871997861,
+      "learning_rate": 9.23508467703563e-06,
+      "loss": 0.3645,
+      "step": 2476
+    },
+    {
+      "epoch": 0.20322435082249662,
+      "grad_norm": 0.6311349176838916,
+      "learning_rate": 9.234378232657334e-06,
+      "loss": 0.3573,
+      "step": 2477
+    },
+    {
+      "epoch": 0.20330639537268738,
+      "grad_norm": 0.642336113201591,
+      "learning_rate": 9.233671489254048e-06,
+      "loss": 0.4076,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2033884399228781,
+      "grad_norm": 0.654791174020213,
+      "learning_rate": 9.232964446875685e-06,
+      "loss": 0.3905,
+      "step": 2479
+    },
+    {
+      "epoch": 0.20347048447306887,
+      "grad_norm": 0.6770208824107404,
+      "learning_rate": 9.232257105572173e-06,
+      "loss": 0.3989,
+      "step": 2480
+    },
+    {
+      "epoch": 0.20355252902325963,
+      "grad_norm": 0.6129336277312075,
+      "learning_rate": 9.231549465393466e-06,
+      "loss": 0.3804,
+      "step": 2481
+    },
+    {
+      "epoch": 0.20363457357345038,
+      "grad_norm": 0.6025250742520959,
+      "learning_rate": 9.230841526389532e-06,
+      "loss": 0.3753,
+      "step": 2482
+    },
+    {
+      "epoch": 0.20371661812364114,
+      "grad_norm": 0.6456490879843146,
+      "learning_rate": 9.230133288610366e-06,
+      "loss": 0.4086,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2037986626738319,
+      "grad_norm": 0.6841928238489919,
+      "learning_rate": 9.229424752105986e-06,
+      "loss": 0.365,
+      "step": 2484
+    },
+    {
+      "epoch": 0.20388070722402266,
+      "grad_norm": 0.6626579538188243,
+      "learning_rate": 9.228715916926423e-06,
+      "loss": 0.3812,
+      "step": 2485
+    },
+    {
+      "epoch": 0.20396275177421339,
+      "grad_norm": 0.6651270462316579,
+      "learning_rate": 9.228006783121737e-06,
+      "loss": 0.4164,
+      "step": 2486
+    },
+    {
+      "epoch": 0.20404479632440414,
+      "grad_norm": 0.6520317542262685,
+      "learning_rate": 9.227297350742004e-06,
+      "loss": 0.3731,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2041268408745949,
+      "grad_norm": 0.6420033718758106,
+      "learning_rate": 9.226587619837323e-06,
+      "loss": 0.398,
+      "step": 2488
+    },
+    {
+      "epoch": 0.20420888542478566,
+      "grad_norm": 0.6478761712165978,
+      "learning_rate": 9.225877590457817e-06,
+      "loss": 0.4116,
+      "step": 2489
+    },
+    {
+      "epoch": 0.20429092997497642,
+      "grad_norm": 0.6877741147755323,
+      "learning_rate": 9.225167262653623e-06,
+      "loss": 0.4156,
+      "step": 2490
+    },
+    {
+      "epoch": 0.20437297452516717,
+      "grad_norm": 0.6486003228025788,
+      "learning_rate": 9.224456636474906e-06,
+      "loss": 0.3969,
+      "step": 2491
+    },
+    {
+      "epoch": 0.20445501907535793,
+      "grad_norm": 0.6723038590546425,
+      "learning_rate": 9.223745711971849e-06,
+      "loss": 0.3686,
+      "step": 2492
+    },
+    {
+      "epoch": 0.20453706362554866,
+      "grad_norm": 0.6094318863015463,
+      "learning_rate": 9.223034489194655e-06,
+      "loss": 0.348,
+      "step": 2493
+    },
+    {
+      "epoch": 0.20461910817573942,
+      "grad_norm": 0.6318595968907784,
+      "learning_rate": 9.222322968193551e-06,
+      "loss": 0.3856,
+      "step": 2494
+    },
+    {
+      "epoch": 0.20470115272593017,
+      "grad_norm": 0.674235166876527,
+      "learning_rate": 9.221611149018781e-06,
+      "loss": 0.4001,
+      "step": 2495
+    },
+    {
+      "epoch": 0.20478319727612093,
+      "grad_norm": 0.650487171079544,
+      "learning_rate": 9.220899031720615e-06,
+      "loss": 0.3767,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2048652418263117,
+      "grad_norm": 0.6301554493117935,
+      "learning_rate": 9.220186616349342e-06,
+      "loss": 0.3631,
+      "step": 2497
+    },
+    {
+      "epoch": 0.20494728637650245,
+      "grad_norm": 0.6805967023242344,
+      "learning_rate": 9.21947390295527e-06,
+      "loss": 0.3981,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2050293309266932,
+      "grad_norm": 0.6273462263520281,
+      "learning_rate": 9.21876089158873e-06,
+      "loss": 0.3815,
+      "step": 2499
+    },
+    {
+      "epoch": 0.20511137547688396,
+      "grad_norm": 0.6386554608069779,
+      "learning_rate": 9.218047582300071e-06,
+      "loss": 0.3715,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2051934200270747,
+      "grad_norm": 0.6343396638094907,
+      "learning_rate": 9.217333975139673e-06,
+      "loss": 0.3657,
+      "step": 2501
+    },
+    {
+      "epoch": 0.20527546457726545,
+      "grad_norm": 0.5935044333434655,
+      "learning_rate": 9.216620070157922e-06,
+      "loss": 0.3536,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2053575091274562,
+      "grad_norm": 0.7905891102033439,
+      "learning_rate": 9.215905867405238e-06,
+      "loss": 0.3975,
+      "step": 2503
+    },
+    {
+      "epoch": 0.20543955367764696,
+      "grad_norm": 0.6765764073846644,
+      "learning_rate": 9.215191366932055e-06,
+      "loss": 0.406,
+      "step": 2504
+    },
+    {
+      "epoch": 0.20552159822783772,
+      "grad_norm": 0.6288024935616046,
+      "learning_rate": 9.214476568788828e-06,
+      "loss": 0.355,
+      "step": 2505
+    },
+    {
+      "epoch": 0.20560364277802848,
+      "grad_norm": 0.62627135514007,
+      "learning_rate": 9.213761473026039e-06,
+      "loss": 0.3579,
+      "step": 2506
+    },
+    {
+      "epoch": 0.20568568732821924,
+      "grad_norm": 0.662085036278497,
+      "learning_rate": 9.213046079694183e-06,
+      "loss": 0.4228,
+      "step": 2507
+    },
+    {
+      "epoch": 0.20576773187840997,
+      "grad_norm": 0.6473849995493279,
+      "learning_rate": 9.212330388843782e-06,
+      "loss": 0.4412,
+      "step": 2508
+    },
+    {
+      "epoch": 0.20584977642860072,
+      "grad_norm": 0.6610397084958671,
+      "learning_rate": 9.211614400525377e-06,
+      "loss": 0.4032,
+      "step": 2509
+    },
+    {
+      "epoch": 0.20593182097879148,
+      "grad_norm": 0.679636987949905,
+      "learning_rate": 9.21089811478953e-06,
+      "loss": 0.3949,
+      "step": 2510
+    },
+    {
+      "epoch": 0.20601386552898224,
+      "grad_norm": 0.7091923837415262,
+      "learning_rate": 9.210181531686823e-06,
+      "loss": 0.3605,
+      "step": 2511
+    },
+    {
+      "epoch": 0.206095910079173,
+      "grad_norm": 0.6925181332355043,
+      "learning_rate": 9.209464651267862e-06,
+      "loss": 0.4311,
+      "step": 2512
+    },
+    {
+      "epoch": 0.20617795462936375,
+      "grad_norm": 0.6346061250075556,
+      "learning_rate": 9.208747473583267e-06,
+      "loss": 0.3675,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2062599991795545,
+      "grad_norm": 0.6408648447434393,
+      "learning_rate": 9.208029998683688e-06,
+      "loss": 0.3859,
+      "step": 2514
+    },
+    {
+      "epoch": 0.20634204372974524,
+      "grad_norm": 0.6021887007768129,
+      "learning_rate": 9.207312226619793e-06,
+      "loss": 0.377,
+      "step": 2515
+    },
+    {
+      "epoch": 0.206424088279936,
+      "grad_norm": 0.5969534385443728,
+      "learning_rate": 9.206594157442267e-06,
+      "loss": 0.403,
+      "step": 2516
+    },
+    {
+      "epoch": 0.20650613283012675,
+      "grad_norm": 0.6075791250564491,
+      "learning_rate": 9.20587579120182e-06,
+      "loss": 0.3432,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2065881773803175,
+      "grad_norm": 0.6845160426906952,
+      "learning_rate": 9.205157127949182e-06,
+      "loss": 0.3683,
+      "step": 2518
+    },
+    {
+      "epoch": 0.20667022193050827,
+      "grad_norm": 0.6959926461691346,
+      "learning_rate": 9.204438167735104e-06,
+      "loss": 0.3804,
+      "step": 2519
+    },
+    {
+      "epoch": 0.20675226648069903,
+      "grad_norm": 0.6190919122064135,
+      "learning_rate": 9.203718910610358e-06,
+      "loss": 0.4126,
+      "step": 2520
+    },
+    {
+      "epoch": 0.20683431103088978,
+      "grad_norm": 0.7201391503267754,
+      "learning_rate": 9.202999356625735e-06,
+      "loss": 0.3795,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2069163555810805,
+      "grad_norm": 0.6647600323815233,
+      "learning_rate": 9.202279505832053e-06,
+      "loss": 0.3816,
+      "step": 2522
+    },
+    {
+      "epoch": 0.20699840013127127,
+      "grad_norm": 0.668155898495432,
+      "learning_rate": 9.20155935828014e-06,
+      "loss": 0.3703,
+      "step": 2523
+    },
+    {
+      "epoch": 0.20708044468146203,
+      "grad_norm": 0.6732578208363885,
+      "learning_rate": 9.200838914020857e-06,
+      "loss": 0.3843,
+      "step": 2524
+    },
+    {
+      "epoch": 0.20716248923165279,
+      "grad_norm": 0.6299295989507886,
+      "learning_rate": 9.200118173105079e-06,
+      "loss": 0.3802,
+      "step": 2525
+    },
+    {
+      "epoch": 0.20724453378184354,
+      "grad_norm": 0.6865258250479377,
+      "learning_rate": 9.199397135583704e-06,
+      "loss": 0.4092,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2073265783320343,
+      "grad_norm": 0.6831176545914609,
+      "learning_rate": 9.198675801507651e-06,
+      "loss": 0.3956,
+      "step": 2527
+    },
+    {
+      "epoch": 0.20740862288222506,
+      "grad_norm": 0.6603954295539408,
+      "learning_rate": 9.197954170927857e-06,
+      "loss": 0.4013,
+      "step": 2528
+    },
+    {
+      "epoch": 0.20749066743241582,
+      "grad_norm": 0.7352828335180958,
+      "learning_rate": 9.197232243895285e-06,
+      "loss": 0.3513,
+      "step": 2529
+    },
+    {
+      "epoch": 0.20757271198260654,
+      "grad_norm": 0.6752157676155963,
+      "learning_rate": 9.196510020460914e-06,
+      "loss": 0.3702,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2076547565327973,
+      "grad_norm": 1.0903409929835717,
+      "learning_rate": 9.195787500675748e-06,
+      "loss": 0.4048,
+      "step": 2531
+    },
+    {
+      "epoch": 0.20773680108298806,
+      "grad_norm": 0.6621229346795444,
+      "learning_rate": 9.19506468459081e-06,
+      "loss": 0.3688,
+      "step": 2532
+    },
+    {
+      "epoch": 0.20781884563317882,
+      "grad_norm": 0.6329987018266661,
+      "learning_rate": 9.194341572257145e-06,
+      "loss": 0.3971,
+      "step": 2533
+    },
+    {
+      "epoch": 0.20790089018336957,
+      "grad_norm": 0.6731275703819781,
+      "learning_rate": 9.193618163725814e-06,
+      "loss": 0.3677,
+      "step": 2534
+    },
+    {
+      "epoch": 0.20798293473356033,
+      "grad_norm": 0.6400521967933274,
+      "learning_rate": 9.192894459047908e-06,
+      "loss": 0.3542,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2080649792837511,
+      "grad_norm": 0.6219788260337382,
+      "learning_rate": 9.192170458274533e-06,
+      "loss": 0.3835,
+      "step": 2536
+    },
+    {
+      "epoch": 0.20814702383394182,
+      "grad_norm": 0.5602779745400707,
+      "learning_rate": 9.191446161456811e-06,
+      "loss": 0.3277,
+      "step": 2537
+    },
+    {
+      "epoch": 0.20822906838413258,
+      "grad_norm": 0.6387629806017365,
+      "learning_rate": 9.1907215686459e-06,
+      "loss": 0.3597,
+      "step": 2538
+    },
+    {
+      "epoch": 0.20831111293432333,
+      "grad_norm": 0.6194926518947934,
+      "learning_rate": 9.189996679892963e-06,
+      "loss": 0.3822,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2083931574845141,
+      "grad_norm": 0.6619313825063131,
+      "learning_rate": 9.189271495249191e-06,
+      "loss": 0.355,
+      "step": 2540
+    },
+    {
+      "epoch": 0.20847520203470485,
+      "grad_norm": 1.2650541102066193,
+      "learning_rate": 9.188546014765798e-06,
+      "loss": 0.3658,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2085572465848956,
+      "grad_norm": 0.6112369919232287,
+      "learning_rate": 9.187820238494014e-06,
+      "loss": 0.387,
+      "step": 2542
+    },
+    {
+      "epoch": 0.20863929113508636,
+      "grad_norm": 0.6176682179302964,
+      "learning_rate": 9.187094166485094e-06,
+      "loss": 0.3782,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2087213356852771,
+      "grad_norm": 0.6404696741333546,
+      "learning_rate": 9.18636779879031e-06,
+      "loss": 0.3674,
+      "step": 2544
+    },
+    {
+      "epoch": 0.20880338023546785,
+      "grad_norm": 0.7025096095138386,
+      "learning_rate": 9.185641135460957e-06,
+      "loss": 0.435,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2088854247856586,
+      "grad_norm": 0.6440880757634656,
+      "learning_rate": 9.184914176548354e-06,
+      "loss": 0.3854,
+      "step": 2546
+    },
+    {
+      "epoch": 0.20896746933584937,
+      "grad_norm": 0.680057351973999,
+      "learning_rate": 9.184186922103835e-06,
+      "loss": 0.3947,
+      "step": 2547
+    },
+    {
+      "epoch": 0.20904951388604012,
+      "grad_norm": 0.6926670499330875,
+      "learning_rate": 9.183459372178758e-06,
+      "loss": 0.411,
+      "step": 2548
+    },
+    {
+      "epoch": 0.20913155843623088,
+      "grad_norm": 0.7354107719483063,
+      "learning_rate": 9.1827315268245e-06,
+      "loss": 0.3501,
+      "step": 2549
+    },
+    {
+      "epoch": 0.20921360298642164,
+      "grad_norm": 0.6166892422702375,
+      "learning_rate": 9.182003386092462e-06,
+      "loss": 0.3634,
+      "step": 2550
+    },
+    {
+      "epoch": 0.20929564753661237,
+      "grad_norm": 0.6494766954642243,
+      "learning_rate": 9.181274950034065e-06,
+      "loss": 0.376,
+      "step": 2551
+    },
+    {
+      "epoch": 0.20937769208680312,
+      "grad_norm": 0.6723158082749306,
+      "learning_rate": 9.180546218700748e-06,
+      "loss": 0.3715,
+      "step": 2552
+    },
+    {
+      "epoch": 0.20945973663699388,
+      "grad_norm": 0.6886700081101815,
+      "learning_rate": 9.179817192143974e-06,
+      "loss": 0.376,
+      "step": 2553
+    },
+    {
+      "epoch": 0.20954178118718464,
+      "grad_norm": 0.6720438526297107,
+      "learning_rate": 9.179087870415225e-06,
+      "loss": 0.4191,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2096238257373754,
+      "grad_norm": 0.6352527444170173,
+      "learning_rate": 9.178358253566005e-06,
+      "loss": 0.3662,
+      "step": 2555
+    },
+    {
+      "epoch": 0.20970587028756615,
+      "grad_norm": 0.6769711962910481,
+      "learning_rate": 9.177628341647837e-06,
+      "loss": 0.4056,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2097879148377569,
+      "grad_norm": 0.6114019555803968,
+      "learning_rate": 9.176898134712269e-06,
+      "loss": 0.3612,
+      "step": 2557
+    },
+    {
+      "epoch": 0.20986995938794767,
+      "grad_norm": 0.6933371607834712,
+      "learning_rate": 9.176167632810864e-06,
+      "loss": 0.3866,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2099520039381384,
+      "grad_norm": 0.7733077719020665,
+      "learning_rate": 9.17543683599521e-06,
+      "loss": 0.3299,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21003404848832916,
+      "grad_norm": 0.7758775083614357,
+      "learning_rate": 9.174705744316917e-06,
+      "loss": 0.3755,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2101160930385199,
+      "grad_norm": 0.872009775262656,
+      "learning_rate": 9.173974357827611e-06,
+      "loss": 0.3961,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21019813758871067,
+      "grad_norm": 0.733150187218619,
+      "learning_rate": 9.173242676578941e-06,
+      "loss": 0.4486,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21028018213890143,
+      "grad_norm": 0.6947389945253467,
+      "learning_rate": 9.172510700622579e-06,
+      "loss": 0.3502,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21036222668909219,
+      "grad_norm": 0.7525518369283369,
+      "learning_rate": 9.171778430010213e-06,
+      "loss": 0.3826,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21044427123928294,
+      "grad_norm": 0.7253784658528457,
+      "learning_rate": 9.17104586479356e-06,
+      "loss": 0.3621,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.8221396166899103,
+      "learning_rate": 9.170313005024347e-06,
+      "loss": 0.4276,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21060836033966443,
+      "grad_norm": 0.6995004447663062,
+      "learning_rate": 9.16957985075433e-06,
+      "loss": 0.4181,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2106904048898552,
+      "grad_norm": 0.6161706890809652,
+      "learning_rate": 9.168846402035283e-06,
+      "loss": 0.3648,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21077244944004594,
+      "grad_norm": 0.6669988110754128,
+      "learning_rate": 9.168112658918999e-06,
+      "loss": 0.4007,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2108544939902367,
+      "grad_norm": 0.6396015474469983,
+      "learning_rate": 9.167378621457299e-06,
+      "loss": 0.3466,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21093653854042746,
+      "grad_norm": 0.6025791702638675,
+      "learning_rate": 9.166644289702014e-06,
+      "loss": 0.3673,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21101858309061822,
+      "grad_norm": 0.6380567832519853,
+      "learning_rate": 9.165909663705003e-06,
+      "loss": 0.3587,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21110062764080895,
+      "grad_norm": 0.6344471060072117,
+      "learning_rate": 9.165174743518147e-06,
+      "loss": 0.3862,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2111826721909997,
+      "grad_norm": 0.9000524314741505,
+      "learning_rate": 9.16443952919334e-06,
+      "loss": 0.4191,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21126471674119046,
+      "grad_norm": 0.680214900401495,
+      "learning_rate": 9.163704020782507e-06,
+      "loss": 0.3857,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21134676129138122,
+      "grad_norm": 0.6629947457461781,
+      "learning_rate": 9.162968218337583e-06,
+      "loss": 0.3643,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21142880584157198,
+      "grad_norm": 0.6978901261857845,
+      "learning_rate": 9.162232121910533e-06,
+      "loss": 0.3956,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21151085039176273,
+      "grad_norm": 0.8236785799541977,
+      "learning_rate": 9.161495731553339e-06,
+      "loss": 0.3834,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2115928949419535,
+      "grad_norm": 0.8487063045216624,
+      "learning_rate": 9.160759047318e-06,
+      "loss": 0.3672,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21167493949214422,
+      "grad_norm": 0.7979938079659996,
+      "learning_rate": 9.160022069256542e-06,
+      "loss": 0.4268,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21175698404233498,
+      "grad_norm": 0.6242965821688784,
+      "learning_rate": 9.159284797421008e-06,
+      "loss": 0.3602,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21183902859252574,
+      "grad_norm": 0.7426535469099641,
+      "learning_rate": 9.158547231863467e-06,
+      "loss": 0.3797,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2119210731427165,
+      "grad_norm": 0.5976764249646294,
+      "learning_rate": 9.157809372636e-06,
+      "loss": 0.3661,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21200311769290725,
+      "grad_norm": 0.6650128016650478,
+      "learning_rate": 9.157071219790716e-06,
+      "loss": 0.3893,
+      "step": 2584
+    },
+    {
+      "epoch": 0.212085162243098,
+      "grad_norm": 0.6348002417281701,
+      "learning_rate": 9.156332773379741e-06,
+      "loss": 0.4041,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21216720679328877,
+      "grad_norm": 0.837936166012152,
+      "learning_rate": 9.155594033455223e-06,
+      "loss": 0.3645,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21224925134347952,
+      "grad_norm": 0.9041537207802659,
+      "learning_rate": 9.15485500006933e-06,
+      "loss": 0.3684,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21233129589367025,
+      "grad_norm": 0.9448503132247393,
+      "learning_rate": 9.154115673274254e-06,
+      "loss": 0.36,
+      "step": 2588
+    },
+    {
+      "epoch": 0.212413340443861,
+      "grad_norm": 0.9388235586690427,
+      "learning_rate": 9.153376053122202e-06,
+      "loss": 0.3647,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21249538499405177,
+      "grad_norm": 0.7908265667135228,
+      "learning_rate": 9.152636139665406e-06,
+      "loss": 0.3952,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21257742954424252,
+      "grad_norm": 0.6552883589284766,
+      "learning_rate": 9.151895932956118e-06,
+      "loss": 0.3594,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21265947409443328,
+      "grad_norm": 1.4012111163643108,
+      "learning_rate": 9.151155433046609e-06,
+      "loss": 0.3815,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21274151864462404,
+      "grad_norm": 0.6775788641429835,
+      "learning_rate": 9.150414639989173e-06,
+      "loss": 0.4062,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2128235631948148,
+      "grad_norm": 0.6120355009000017,
+      "learning_rate": 9.149673553836124e-06,
+      "loss": 0.3591,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21290560774500553,
+      "grad_norm": 0.6974272858927714,
+      "learning_rate": 9.148932174639795e-06,
+      "loss": 0.3783,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21298765229519628,
+      "grad_norm": 0.6795583856768193,
+      "learning_rate": 9.14819050245254e-06,
+      "loss": 0.3897,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21306969684538704,
+      "grad_norm": 0.6272430445407098,
+      "learning_rate": 9.14744853732674e-06,
+      "loss": 0.3813,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2131517413955778,
+      "grad_norm": 0.6800431062910558,
+      "learning_rate": 9.146706279314786e-06,
+      "loss": 0.4413,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21323378594576856,
+      "grad_norm": 0.6591762708312267,
+      "learning_rate": 9.145963728469094e-06,
+      "loss": 0.4259,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2133158304959593,
+      "grad_norm": 0.7199618632597287,
+      "learning_rate": 9.145220884842108e-06,
+      "loss": 0.4142,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21339787504615007,
+      "grad_norm": 0.642801519987645,
+      "learning_rate": 9.144477748486282e-06,
+      "loss": 0.3779,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2134799195963408,
+      "grad_norm": 0.7541864731717199,
+      "learning_rate": 9.143734319454097e-06,
+      "loss": 0.3833,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21356196414653156,
+      "grad_norm": 0.6772462637293893,
+      "learning_rate": 9.142990597798049e-06,
+      "loss": 0.4494,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21364400869672232,
+      "grad_norm": 0.7399981041059068,
+      "learning_rate": 9.142246583570662e-06,
+      "loss": 0.3303,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21372605324691307,
+      "grad_norm": 0.6548829404872172,
+      "learning_rate": 9.141502276824478e-06,
+      "loss": 0.4207,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21380809779710383,
+      "grad_norm": 0.8806390532265409,
+      "learning_rate": 9.140757677612056e-06,
+      "loss": 0.3586,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2138901423472946,
+      "grad_norm": 1.4090753518738564,
+      "learning_rate": 9.140012785985978e-06,
+      "loss": 0.3622,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21397218689748534,
+      "grad_norm": 0.9061500651481044,
+      "learning_rate": 9.139267601998852e-06,
+      "loss": 0.4111,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21405423144767607,
+      "grad_norm": 0.6736693473272303,
+      "learning_rate": 9.138522125703295e-06,
+      "loss": 0.4333,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21413627599786683,
+      "grad_norm": 0.6506427335828158,
+      "learning_rate": 9.137776357151956e-06,
+      "loss": 0.3415,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2142183205480576,
+      "grad_norm": 0.652079053853386,
+      "learning_rate": 9.137030296397499e-06,
+      "loss": 0.3635,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21430036509824835,
+      "grad_norm": 0.8667769584999436,
+      "learning_rate": 9.136283943492608e-06,
+      "loss": 0.4273,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2143824096484391,
+      "grad_norm": 1.332103470810094,
+      "learning_rate": 9.135537298489991e-06,
+      "loss": 0.4297,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21446445419862986,
+      "grad_norm": 0.6843070182332299,
+      "learning_rate": 9.134790361442374e-06,
+      "loss": 0.3675,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21454649874882062,
+      "grad_norm": 0.8169727244478298,
+      "learning_rate": 9.134043132402505e-06,
+      "loss": 0.4276,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21462854329901138,
+      "grad_norm": 0.6558182686860287,
+      "learning_rate": 9.133295611423151e-06,
+      "loss": 0.3974,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2147105878492021,
+      "grad_norm": 0.7150293464944485,
+      "learning_rate": 9.132547798557102e-06,
+      "loss": 0.3799,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21479263239939286,
+      "grad_norm": 0.635076352796439,
+      "learning_rate": 9.131799693857166e-06,
+      "loss": 0.3392,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21487467694958362,
+      "grad_norm": 0.708377456032219,
+      "learning_rate": 9.131051297376177e-06,
+      "loss": 0.3774,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21495672149977438,
+      "grad_norm": 2.2925515389135036,
+      "learning_rate": 9.130302609166978e-06,
+      "loss": 0.4082,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21503876604996514,
+      "grad_norm": 0.6557596775279755,
+      "learning_rate": 9.129553629282448e-06,
+      "loss": 0.3554,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2151208106001559,
+      "grad_norm": 0.8698250247340976,
+      "learning_rate": 9.128804357775474e-06,
+      "loss": 0.3667,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21520285515034665,
+      "grad_norm": 0.7021177093500526,
+      "learning_rate": 9.12805479469897e-06,
+      "loss": 0.3783,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21528489970053738,
+      "grad_norm": 0.6871955603836548,
+      "learning_rate": 9.127304940105869e-06,
+      "loss": 0.3519,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21536694425072814,
+      "grad_norm": 0.8616898177224022,
+      "learning_rate": 9.126554794049123e-06,
+      "loss": 0.3883,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2154489888009189,
+      "grad_norm": 1.148924209720728,
+      "learning_rate": 9.125804356581708e-06,
+      "loss": 0.3411,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21553103335110965,
+      "grad_norm": 0.66376179724882,
+      "learning_rate": 9.125053627756619e-06,
+      "loss": 0.3902,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2156130779013004,
+      "grad_norm": 0.8208795214390936,
+      "learning_rate": 9.124302607626868e-06,
+      "loss": 0.3734,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21569512245149117,
+      "grad_norm": 0.7665815928861057,
+      "learning_rate": 9.123551296245494e-06,
+      "loss": 0.3514,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21577716700168192,
+      "grad_norm": 0.7372555374851242,
+      "learning_rate": 9.122799693665554e-06,
+      "loss": 0.3771,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21585921155187265,
+      "grad_norm": 0.7559201047176727,
+      "learning_rate": 9.122047799940121e-06,
+      "loss": 0.3798,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2159412561020634,
+      "grad_norm": 0.6249792617513545,
+      "learning_rate": 9.121295615122297e-06,
+      "loss": 0.3952,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21602330065225417,
+      "grad_norm": 0.7896245110270724,
+      "learning_rate": 9.120543139265196e-06,
+      "loss": 0.3692,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21610534520244493,
+      "grad_norm": 0.6755468976710258,
+      "learning_rate": 9.11979037242196e-06,
+      "loss": 0.3731,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21618738975263568,
+      "grad_norm": 0.7859251495163223,
+      "learning_rate": 9.119037314645745e-06,
+      "loss": 0.425,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21626943430282644,
+      "grad_norm": 0.6945618870320187,
+      "learning_rate": 9.118283965989735e-06,
+      "loss": 0.3335,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2163514788530172,
+      "grad_norm": 0.6707223388070855,
+      "learning_rate": 9.117530326507128e-06,
+      "loss": 0.3932,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21643352340320793,
+      "grad_norm": 0.7516661768433814,
+      "learning_rate": 9.116776396251142e-06,
+      "loss": 0.3865,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21651556795339869,
+      "grad_norm": 0.6168859964025974,
+      "learning_rate": 9.116022175275024e-06,
+      "loss": 0.3771,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21659761250358944,
+      "grad_norm": 0.7417190074416967,
+      "learning_rate": 9.11526766363203e-06,
+      "loss": 0.3855,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2166796570537802,
+      "grad_norm": 0.6922904410950621,
+      "learning_rate": 9.114512861375446e-06,
+      "loss": 0.3561,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21676170160397096,
+      "grad_norm": 0.6007934887304395,
+      "learning_rate": 9.113757768558575e-06,
+      "loss": 0.3837,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21684374615416172,
+      "grad_norm": 0.6703042890440778,
+      "learning_rate": 9.113002385234741e-06,
+      "loss": 0.429,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21692579070435247,
+      "grad_norm": 0.6436295562014803,
+      "learning_rate": 9.112246711457284e-06,
+      "loss": 0.3762,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21700783525454323,
+      "grad_norm": 0.6731962763196602,
+      "learning_rate": 9.111490747279573e-06,
+      "loss": 0.3673,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21708987980473396,
+      "grad_norm": 0.6581652784353491,
+      "learning_rate": 9.11073449275499e-06,
+      "loss": 0.3567,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21717192435492472,
+      "grad_norm": 0.6521725079754717,
+      "learning_rate": 9.109977947936943e-06,
+      "loss": 0.4305,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21725396890511547,
+      "grad_norm": 0.6362530975603055,
+      "learning_rate": 9.109221112878855e-06,
+      "loss": 0.4119,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21733601345530623,
+      "grad_norm": 0.6604054075935757,
+      "learning_rate": 9.108463987634174e-06,
+      "loss": 0.3932,
+      "step": 2649
+    },
+    {
+      "epoch": 0.217418058005497,
+      "grad_norm": 0.661250339433179,
+      "learning_rate": 9.107706572256368e-06,
+      "loss": 0.3693,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21750010255568775,
+      "grad_norm": 0.6886547209841107,
+      "learning_rate": 9.10694886679892e-06,
+      "loss": 0.3923,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2175821471058785,
+      "grad_norm": 0.628713327526805,
+      "learning_rate": 9.106190871315346e-06,
+      "loss": 0.3752,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21766419165606923,
+      "grad_norm": 0.6850177586994058,
+      "learning_rate": 9.105432585859166e-06,
+      "loss": 0.3616,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21774623620626,
+      "grad_norm": 0.6515376344267589,
+      "learning_rate": 9.104674010483935e-06,
+      "loss": 0.3732,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21782828075645075,
+      "grad_norm": 0.6437056902115268,
+      "learning_rate": 9.10391514524322e-06,
+      "loss": 0.3699,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2179103253066415,
+      "grad_norm": 0.6286454480122333,
+      "learning_rate": 9.10315599019061e-06,
+      "loss": 0.3854,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21799236985683226,
+      "grad_norm": 0.6269492408316368,
+      "learning_rate": 9.102396545379717e-06,
+      "loss": 0.3854,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21807441440702302,
+      "grad_norm": 0.7169312575402552,
+      "learning_rate": 9.101636810864169e-06,
+      "loss": 0.3743,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21815645895721378,
+      "grad_norm": 0.6514459619909193,
+      "learning_rate": 9.10087678669762e-06,
+      "loss": 0.4122,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2182385035074045,
+      "grad_norm": 0.6978395549137809,
+      "learning_rate": 9.100116472933743e-06,
+      "loss": 0.4425,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21832054805759527,
+      "grad_norm": 0.5843856624211788,
+      "learning_rate": 9.099355869626228e-06,
+      "loss": 0.3585,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21840259260778602,
+      "grad_norm": 0.6822448143616036,
+      "learning_rate": 9.098594976828784e-06,
+      "loss": 0.3512,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21848463715797678,
+      "grad_norm": 0.6099835789565957,
+      "learning_rate": 9.09783379459515e-06,
+      "loss": 0.3477,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21856668170816754,
+      "grad_norm": 0.6966032252966249,
+      "learning_rate": 9.097072322979077e-06,
+      "loss": 0.4139,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2186487262583583,
+      "grad_norm": 0.650800889925391,
+      "learning_rate": 9.09631056203434e-06,
+      "loss": 0.3694,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21873077080854905,
+      "grad_norm": 0.7547165291255985,
+      "learning_rate": 9.095548511814731e-06,
+      "loss": 0.4044,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2188128153587398,
+      "grad_norm": 0.6952654734821276,
+      "learning_rate": 9.094786172374066e-06,
+      "loss": 0.3549,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21889485990893054,
+      "grad_norm": 0.7156367162993398,
+      "learning_rate": 9.094023543766181e-06,
+      "loss": 0.3856,
+      "step": 2668
+    },
+    {
+      "epoch": 0.2189769044591213,
+      "grad_norm": 1.2960325982966123,
+      "learning_rate": 9.09326062604493e-06,
+      "loss": 0.3989,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21905894900931205,
+      "grad_norm": 0.6882163007574524,
+      "learning_rate": 9.09249741926419e-06,
+      "loss": 0.377,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2191409935595028,
+      "grad_norm": 0.6728774465760998,
+      "learning_rate": 9.091733923477859e-06,
+      "loss": 0.4067,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21922303810969357,
+      "grad_norm": 0.9819559185838568,
+      "learning_rate": 9.09097013873985e-06,
+      "loss": 0.3981,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21930508265988433,
+      "grad_norm": 0.6877635741797897,
+      "learning_rate": 9.090206065104104e-06,
+      "loss": 0.3487,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21938712721007508,
+      "grad_norm": 0.6416883078539304,
+      "learning_rate": 9.089441702624577e-06,
+      "loss": 0.3669,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2194691717602658,
+      "grad_norm": 0.6218478402361466,
+      "learning_rate": 9.088677051355245e-06,
+      "loss": 0.3395,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21955121631045657,
+      "grad_norm": 0.694439705721062,
+      "learning_rate": 9.087912111350112e-06,
+      "loss": 0.4198,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21963326086064733,
+      "grad_norm": 0.6097141767785935,
+      "learning_rate": 9.08714688266319e-06,
+      "loss": 0.3804,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21971530541083809,
+      "grad_norm": 0.614045430071774,
+      "learning_rate": 9.086381365348525e-06,
+      "loss": 0.3828,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21979734996102884,
+      "grad_norm": 0.6774049162390624,
+      "learning_rate": 9.085615559460172e-06,
+      "loss": 0.4111,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2198793945112196,
+      "grad_norm": 0.6448850209351373,
+      "learning_rate": 9.08484946505221e-06,
+      "loss": 0.3865,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21996143906141036,
+      "grad_norm": 0.6264325804312416,
+      "learning_rate": 9.084083082178745e-06,
+      "loss": 0.3769,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2200434836116011,
+      "grad_norm": 0.6133743105482133,
+      "learning_rate": 9.083316410893892e-06,
+      "loss": 0.3766,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22012552816179184,
+      "grad_norm": 0.651752604561037,
+      "learning_rate": 9.082549451251796e-06,
+      "loss": 0.3629,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2202075727119826,
+      "grad_norm": 0.6227683259350826,
+      "learning_rate": 9.081782203306616e-06,
+      "loss": 0.368,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22028961726217336,
+      "grad_norm": 0.7116011986603388,
+      "learning_rate": 9.081014667112532e-06,
+      "loss": 0.4102,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22037166181236412,
+      "grad_norm": 0.6681303963890041,
+      "learning_rate": 9.080246842723753e-06,
+      "loss": 0.3742,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22045370636255487,
+      "grad_norm": 0.7333917392920957,
+      "learning_rate": 9.079478730194496e-06,
+      "loss": 0.3987,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22053575091274563,
+      "grad_norm": 0.6272940318010276,
+      "learning_rate": 9.078710329579006e-06,
+      "loss": 0.3475,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22061779546293636,
+      "grad_norm": 1.2575045153798088,
+      "learning_rate": 9.077941640931544e-06,
+      "loss": 0.3592,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22069984001312712,
+      "grad_norm": 0.6259641010492106,
+      "learning_rate": 9.077172664306396e-06,
+      "loss": 0.3685,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22078188456331788,
+      "grad_norm": 0.5779463112069402,
+      "learning_rate": 9.076403399757865e-06,
+      "loss": 0.3198,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22086392911350863,
+      "grad_norm": 0.6446623898945679,
+      "learning_rate": 9.075633847340275e-06,
+      "loss": 0.382,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2209459736636994,
+      "grad_norm": 0.674686611828484,
+      "learning_rate": 9.07486400710797e-06,
+      "loss": 0.4045,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22102801821389015,
+      "grad_norm": 0.655929095543329,
+      "learning_rate": 9.074093879115315e-06,
+      "loss": 0.3674,
+      "step": 2694
+    },
+    {
+      "epoch": 0.2211100627640809,
+      "grad_norm": 0.8133682189270502,
+      "learning_rate": 9.073323463416699e-06,
+      "loss": 0.3914,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22119210731427166,
+      "grad_norm": 0.6622689135724624,
+      "learning_rate": 9.072552760066522e-06,
+      "loss": 0.3774,
+      "step": 2696
+    },
+    {
+      "epoch": 0.2212741518644624,
+      "grad_norm": 0.6801801679045957,
+      "learning_rate": 9.071781769119211e-06,
+      "loss": 0.3682,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22135619641465315,
+      "grad_norm": 0.6157309758313722,
+      "learning_rate": 9.071010490629216e-06,
+      "loss": 0.3935,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2214382409648439,
+      "grad_norm": 0.6569473336696454,
+      "learning_rate": 9.070238924650998e-06,
+      "loss": 0.3681,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22152028551503467,
+      "grad_norm": 0.6018524020942407,
+      "learning_rate": 9.069467071239049e-06,
+      "loss": 0.3589,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22160233006522542,
+      "grad_norm": 0.6880986379218741,
+      "learning_rate": 9.068694930447871e-06,
+      "loss": 0.3608,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22168437461541618,
+      "grad_norm": 0.5745336980322849,
+      "learning_rate": 9.067922502331997e-06,
+      "loss": 0.3577,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22176641916560694,
+      "grad_norm": 0.637974002127837,
+      "learning_rate": 9.067149786945967e-06,
+      "loss": 0.3414,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22184846371579767,
+      "grad_norm": 0.6709654782255665,
+      "learning_rate": 9.066376784344357e-06,
+      "loss": 0.3526,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22193050826598842,
+      "grad_norm": 0.6620388541482407,
+      "learning_rate": 9.065603494581752e-06,
+      "loss": 0.3646,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22201255281617918,
+      "grad_norm": 0.6462595507172947,
+      "learning_rate": 9.064829917712759e-06,
+      "loss": 0.3768,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22209459736636994,
+      "grad_norm": 0.6175889811584091,
+      "learning_rate": 9.064056053792006e-06,
+      "loss": 0.3195,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2221766419165607,
+      "grad_norm": 0.631373771349987,
+      "learning_rate": 9.063281902874146e-06,
+      "loss": 0.368,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22225868646675145,
+      "grad_norm": 0.6600114793452201,
+      "learning_rate": 9.062507465013845e-06,
+      "loss": 0.3838,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2223407310169422,
+      "grad_norm": 0.5923089891289999,
+      "learning_rate": 9.061732740265797e-06,
+      "loss": 0.3748,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22242277556713294,
+      "grad_norm": 0.6513483666234801,
+      "learning_rate": 9.060957728684705e-06,
+      "loss": 0.4136,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2225048201173237,
+      "grad_norm": 0.5919377915623482,
+      "learning_rate": 9.060182430325303e-06,
+      "loss": 0.4011,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22258686466751446,
+      "grad_norm": 0.7807396400069077,
+      "learning_rate": 9.059406845242343e-06,
+      "loss": 0.3745,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2226689092177052,
+      "grad_norm": 0.670477961255953,
+      "learning_rate": 9.058630973490591e-06,
+      "loss": 0.4042,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22275095376789597,
+      "grad_norm": 0.7370705555511812,
+      "learning_rate": 9.057854815124841e-06,
+      "loss": 0.4088,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22283299831808673,
+      "grad_norm": 0.6664938483987727,
+      "learning_rate": 9.057078370199905e-06,
+      "loss": 0.3666,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22291504286827749,
+      "grad_norm": 0.6113593669993602,
+      "learning_rate": 9.05630163877061e-06,
+      "loss": 0.3536,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22299708741846822,
+      "grad_norm": 0.7384710363433474,
+      "learning_rate": 9.055524620891814e-06,
+      "loss": 0.4004,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22307913196865897,
+      "grad_norm": 0.6171335961026468,
+      "learning_rate": 9.054747316618381e-06,
+      "loss": 0.4312,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22316117651884973,
+      "grad_norm": 0.8163005832165359,
+      "learning_rate": 9.053969726005209e-06,
+      "loss": 0.3913,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2232432210690405,
+      "grad_norm": 0.6367867062695093,
+      "learning_rate": 9.053191849107208e-06,
+      "loss": 0.3974,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22332526561923124,
+      "grad_norm": 0.6194505909454521,
+      "learning_rate": 9.052413685979311e-06,
+      "loss": 0.3919,
+      "step": 2722
+    },
+    {
+      "epoch": 0.223407310169422,
+      "grad_norm": 0.6594909840340127,
+      "learning_rate": 9.05163523667647e-06,
+      "loss": 0.409,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22348935471961276,
+      "grad_norm": 0.5836912829189559,
+      "learning_rate": 9.050856501253657e-06,
+      "loss": 0.3652,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22357139926980352,
+      "grad_norm": 0.6611595599297265,
+      "learning_rate": 9.050077479765867e-06,
+      "loss": 0.4276,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22365344381999425,
+      "grad_norm": 0.674963760868661,
+      "learning_rate": 9.049298172268114e-06,
+      "loss": 0.3891,
+      "step": 2726
+    },
+    {
+      "epoch": 0.223735488370185,
+      "grad_norm": 0.6179739663466204,
+      "learning_rate": 9.048518578815427e-06,
+      "loss": 0.4089,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22381753292037576,
+      "grad_norm": 0.6425962326109272,
+      "learning_rate": 9.047738699462863e-06,
+      "loss": 0.3508,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22389957747056652,
+      "grad_norm": 0.6310368700182063,
+      "learning_rate": 9.046958534265495e-06,
+      "loss": 0.3915,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22398162202075728,
+      "grad_norm": 0.633905099077027,
+      "learning_rate": 9.046178083278417e-06,
+      "loss": 0.3455,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22406366657094803,
+      "grad_norm": 0.6175438725589734,
+      "learning_rate": 9.045397346556744e-06,
+      "loss": 0.3773,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2241457111211388,
+      "grad_norm": 0.6382434164535213,
+      "learning_rate": 9.044616324155611e-06,
+      "loss": 0.3724,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22422775567132952,
+      "grad_norm": 0.6966640756446734,
+      "learning_rate": 9.04383501613017e-06,
+      "loss": 0.3727,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22430980022152028,
+      "grad_norm": 0.6033039677245693,
+      "learning_rate": 9.043053422535597e-06,
+      "loss": 0.3849,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22439184477171104,
+      "grad_norm": 0.6555502016461975,
+      "learning_rate": 9.042271543427088e-06,
+      "loss": 0.3997,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2244738893219018,
+      "grad_norm": 0.6239332558612243,
+      "learning_rate": 9.041489378859856e-06,
+      "loss": 0.3656,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22455593387209255,
+      "grad_norm": 0.6357515889087331,
+      "learning_rate": 9.040706928889138e-06,
+      "loss": 0.4045,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2246379784222833,
+      "grad_norm": 0.5831629079831314,
+      "learning_rate": 9.039924193570188e-06,
+      "loss": 0.3727,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22472002297247406,
+      "grad_norm": 0.7500176258821311,
+      "learning_rate": 9.039141172958282e-06,
+      "loss": 0.3786,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2248020675226648,
+      "grad_norm": 0.899032325871651,
+      "learning_rate": 9.038357867108716e-06,
+      "loss": 0.4163,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22488411207285555,
+      "grad_norm": 0.6352166783017423,
+      "learning_rate": 9.037574276076806e-06,
+      "loss": 0.3603,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2249661566230463,
+      "grad_norm": 0.7693239250401079,
+      "learning_rate": 9.036790399917886e-06,
+      "loss": 0.4335,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22504820117323707,
+      "grad_norm": 0.6321867568959859,
+      "learning_rate": 9.036006238687316e-06,
+      "loss": 0.3917,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22513024572342782,
+      "grad_norm": 0.6040845374030762,
+      "learning_rate": 9.035221792440468e-06,
+      "loss": 0.3541,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22521229027361858,
+      "grad_norm": 0.6118227147740725,
+      "learning_rate": 9.03443706123274e-06,
+      "loss": 0.3492,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22529433482380934,
+      "grad_norm": 0.6347230243745255,
+      "learning_rate": 9.03365204511955e-06,
+      "loss": 0.3966,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22537637937400007,
+      "grad_norm": 0.798327795822871,
+      "learning_rate": 9.032866744156332e-06,
+      "loss": 0.3907,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22545842392419083,
+      "grad_norm": 0.6620000572059247,
+      "learning_rate": 9.032081158398542e-06,
+      "loss": 0.3541,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22554046847438158,
+      "grad_norm": 0.6650270675360587,
+      "learning_rate": 9.031295287901663e-06,
+      "loss": 0.4061,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22562251302457234,
+      "grad_norm": 0.5998919718947743,
+      "learning_rate": 9.030509132721183e-06,
+      "loss": 0.3795,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2257045575747631,
+      "grad_norm": 0.7178732843558149,
+      "learning_rate": 9.029722692912625e-06,
+      "loss": 0.337,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22578660212495386,
+      "grad_norm": 0.6415540256019558,
+      "learning_rate": 9.028935968531528e-06,
+      "loss": 0.3633,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2258686466751446,
+      "grad_norm": 0.604806225408011,
+      "learning_rate": 9.028148959633442e-06,
+      "loss": 0.3447,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22595069122533537,
+      "grad_norm": 0.7699195815096239,
+      "learning_rate": 9.027361666273949e-06,
+      "loss": 0.4174,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2260327357755261,
+      "grad_norm": 0.6496747233912341,
+      "learning_rate": 9.026574088508646e-06,
+      "loss": 0.389,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22611478032571686,
+      "grad_norm": 0.7474543134083539,
+      "learning_rate": 9.02578622639315e-06,
+      "loss": 0.3565,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22619682487590761,
+      "grad_norm": 0.6095781669110364,
+      "learning_rate": 9.024998079983099e-06,
+      "loss": 0.3409,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22627886942609837,
+      "grad_norm": 0.728958368822159,
+      "learning_rate": 9.02420964933415e-06,
+      "loss": 0.385,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22636091397628913,
+      "grad_norm": 0.6731824457966341,
+      "learning_rate": 9.023420934501981e-06,
+      "loss": 0.4326,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2264429585264799,
+      "grad_norm": 0.6878003562322927,
+      "learning_rate": 9.02263193554229e-06,
+      "loss": 0.3581,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22652500307667064,
+      "grad_norm": 0.8819676549349127,
+      "learning_rate": 9.021842652510793e-06,
+      "loss": 0.4097,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22660704762686137,
+      "grad_norm": 0.6645278369742283,
+      "learning_rate": 9.021053085463231e-06,
+      "loss": 0.3616,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22668909217705213,
+      "grad_norm": 0.5963076260764391,
+      "learning_rate": 9.02026323445536e-06,
+      "loss": 0.3513,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2267711367272429,
+      "grad_norm": 0.723715950065725,
+      "learning_rate": 9.01947309954296e-06,
+      "loss": 0.3929,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22685318127743365,
+      "grad_norm": 0.6911059081842901,
+      "learning_rate": 9.018682680781825e-06,
+      "loss": 0.3639,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2269352258276244,
+      "grad_norm": 0.6270442530743773,
+      "learning_rate": 9.017891978227776e-06,
+      "loss": 0.3757,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22701727037781516,
+      "grad_norm": 0.7141614357024026,
+      "learning_rate": 9.017100991936651e-06,
+      "loss": 0.4123,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22709931492800592,
+      "grad_norm": 0.6321003809007437,
+      "learning_rate": 9.01630972196431e-06,
+      "loss": 0.3881,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22718135947819665,
+      "grad_norm": 0.7103376453755822,
+      "learning_rate": 9.015518168366626e-06,
+      "loss": 0.417,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2272634040283874,
+      "grad_norm": 0.73348481405324,
+      "learning_rate": 9.014726331199501e-06,
+      "loss": 0.3991,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22734544857857816,
+      "grad_norm": 0.6520513209000508,
+      "learning_rate": 9.013934210518853e-06,
+      "loss": 0.3605,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22742749312876892,
+      "grad_norm": 0.6090894415052664,
+      "learning_rate": 9.013141806380621e-06,
+      "loss": 0.3736,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22750953767895968,
+      "grad_norm": 0.6508748349767601,
+      "learning_rate": 9.012349118840761e-06,
+      "loss": 0.3704,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22759158222915044,
+      "grad_norm": 1.0448578923778737,
+      "learning_rate": 9.011556147955255e-06,
+      "loss": 0.3858,
+      "step": 2774
+    },
+    {
+      "epoch": 0.2276736267793412,
+      "grad_norm": 0.630351923945559,
+      "learning_rate": 9.010762893780095e-06,
+      "loss": 0.3773,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22775567132953192,
+      "grad_norm": 0.6260563594943098,
+      "learning_rate": 9.009969356371307e-06,
+      "loss": 0.3763,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22783771587972268,
+      "grad_norm": 0.7453042150792981,
+      "learning_rate": 9.009175535784923e-06,
+      "loss": 0.3764,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22791976042991344,
+      "grad_norm": 0.6417839437446449,
+      "learning_rate": 9.008381432077005e-06,
+      "loss": 0.4028,
+      "step": 2778
+    },
+    {
+      "epoch": 0.2280018049801042,
+      "grad_norm": 0.7028196683040475,
+      "learning_rate": 9.007587045303628e-06,
+      "loss": 0.3717,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22808384953029495,
+      "grad_norm": 0.6891279795857909,
+      "learning_rate": 9.006792375520895e-06,
+      "loss": 0.4177,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2281658940804857,
+      "grad_norm": 0.62558348181125,
+      "learning_rate": 9.005997422784922e-06,
+      "loss": 0.3787,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22824793863067647,
+      "grad_norm": 0.6433918327057836,
+      "learning_rate": 9.005202187151845e-06,
+      "loss": 0.3491,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22832998318086722,
+      "grad_norm": 0.6546822753837355,
+      "learning_rate": 9.004406668677825e-06,
+      "loss": 0.3795,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22841202773105795,
+      "grad_norm": 0.6236442273667719,
+      "learning_rate": 9.003610867419042e-06,
+      "loss": 0.4068,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2284940722812487,
+      "grad_norm": 0.5872768155829498,
+      "learning_rate": 9.00281478343169e-06,
+      "loss": 0.3475,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22857611683143947,
+      "grad_norm": 0.7259445017063803,
+      "learning_rate": 9.00201841677199e-06,
+      "loss": 0.3715,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22865816138163023,
+      "grad_norm": 0.6291178896770515,
+      "learning_rate": 9.001221767496178e-06,
+      "loss": 0.3767,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22874020593182098,
+      "grad_norm": 0.6377587814826681,
+      "learning_rate": 9.000424835660514e-06,
+      "loss": 0.3723,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22882225048201174,
+      "grad_norm": 0.6443038619724775,
+      "learning_rate": 8.999627621321274e-06,
+      "loss": 0.4119,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2289042950322025,
+      "grad_norm": 0.6489885307546357,
+      "learning_rate": 8.998830124534758e-06,
+      "loss": 0.337,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22898633958239323,
+      "grad_norm": 1.0471705587245788,
+      "learning_rate": 8.998032345357284e-06,
+      "loss": 0.3927,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22906838413258399,
+      "grad_norm": 0.5895130236301753,
+      "learning_rate": 8.997234283845188e-06,
+      "loss": 0.3309,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22915042868277474,
+      "grad_norm": 0.6090335629479631,
+      "learning_rate": 8.996435940054829e-06,
+      "loss": 0.324,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2292324732329655,
+      "grad_norm": 0.6632852336243181,
+      "learning_rate": 8.995637314042585e-06,
+      "loss": 0.3827,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22931451778315626,
+      "grad_norm": 0.692373258105404,
+      "learning_rate": 8.994838405864854e-06,
+      "loss": 0.3982,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22939656233334701,
+      "grad_norm": 0.6578694611168083,
+      "learning_rate": 8.994039215578053e-06,
+      "loss": 0.3748,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22947860688353777,
+      "grad_norm": 0.7081116026236615,
+      "learning_rate": 8.993239743238618e-06,
+      "loss": 0.3795,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2295606514337285,
+      "grad_norm": 0.67094636764655,
+      "learning_rate": 8.99243998890301e-06,
+      "loss": 0.394,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22964269598391926,
+      "grad_norm": 0.7124296318099466,
+      "learning_rate": 8.991639952627702e-06,
+      "loss": 0.3863,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22972474053411002,
+      "grad_norm": 0.5959499314579229,
+      "learning_rate": 8.990839634469194e-06,
+      "loss": 0.372,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22980678508430077,
+      "grad_norm": 0.6508109055578482,
+      "learning_rate": 8.990039034484004e-06,
+      "loss": 0.3419,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22988882963449153,
+      "grad_norm": 0.6590394882262447,
+      "learning_rate": 8.989238152728667e-06,
+      "loss": 0.3353,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2299708741846823,
+      "grad_norm": 0.6411953722561791,
+      "learning_rate": 8.98843698925974e-06,
+      "loss": 0.3787,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23005291873487305,
+      "grad_norm": 0.6850047222999149,
+      "learning_rate": 8.987635544133804e-06,
+      "loss": 0.3991,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23013496328506378,
+      "grad_norm": 0.700834438885956,
+      "learning_rate": 8.98683381740745e-06,
+      "loss": 0.375,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23021700783525453,
+      "grad_norm": 0.7847630452590935,
+      "learning_rate": 8.9860318091373e-06,
+      "loss": 0.3783,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2302990523854453,
+      "grad_norm": 0.6738351275018104,
+      "learning_rate": 8.985229519379984e-06,
+      "loss": 0.3917,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23038109693563605,
+      "grad_norm": 0.7807312940810197,
+      "learning_rate": 8.984426948192164e-06,
+      "loss": 0.3781,
+      "step": 2808
+    },
+    {
+      "epoch": 0.2304631414858268,
+      "grad_norm": 3.2430297971100543,
+      "learning_rate": 8.983624095630513e-06,
+      "loss": 0.4168,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23054518603601756,
+      "grad_norm": 0.6497912653789684,
+      "learning_rate": 8.98282096175173e-06,
+      "loss": 0.3607,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23062723058620832,
+      "grad_norm": 0.7326014909954742,
+      "learning_rate": 8.982017546612529e-06,
+      "loss": 0.3749,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23070927513639908,
+      "grad_norm": 0.5858238930873262,
+      "learning_rate": 8.981213850269645e-06,
+      "loss": 0.3338,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2307913196865898,
+      "grad_norm": 0.6748765958187032,
+      "learning_rate": 8.980409872779838e-06,
+      "loss": 0.412,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23087336423678056,
+      "grad_norm": 0.7213546333103097,
+      "learning_rate": 8.979605614199878e-06,
+      "loss": 0.3787,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23095540878697132,
+      "grad_norm": 0.6648633834793602,
+      "learning_rate": 8.978801074586566e-06,
+      "loss": 0.3987,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23103745333716208,
+      "grad_norm": 0.6460710982567687,
+      "learning_rate": 8.977996253996712e-06,
+      "loss": 0.3745,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23111949788735284,
+      "grad_norm": 0.8365663242920363,
+      "learning_rate": 8.977191152487155e-06,
+      "loss": 0.3791,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2312015424375436,
+      "grad_norm": 0.6010725332122804,
+      "learning_rate": 8.97638577011475e-06,
+      "loss": 0.3525,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23128358698773435,
+      "grad_norm": 0.621051755278034,
+      "learning_rate": 8.975580106936368e-06,
+      "loss": 0.4169,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23136563153792508,
+      "grad_norm": 0.7093269690785807,
+      "learning_rate": 8.974774163008909e-06,
+      "loss": 0.3749,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23144767608811584,
+      "grad_norm": 0.6383281248654004,
+      "learning_rate": 8.973967938389281e-06,
+      "loss": 0.34,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2315297206383066,
+      "grad_norm": 0.6103467727392008,
+      "learning_rate": 8.973161433134423e-06,
+      "loss": 0.371,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23161176518849735,
+      "grad_norm": 0.6588769591285334,
+      "learning_rate": 8.972354647301289e-06,
+      "loss": 0.3534,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2316938097386881,
+      "grad_norm": 0.6201438268161813,
+      "learning_rate": 8.971547580946851e-06,
+      "loss": 0.4028,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23177585428887887,
+      "grad_norm": 0.6847087866969229,
+      "learning_rate": 8.970740234128104e-06,
+      "loss": 0.3699,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23185789883906963,
+      "grad_norm": 0.6758868976319113,
+      "learning_rate": 8.969932606902062e-06,
+      "loss": 0.3642,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23193994338926036,
+      "grad_norm": 0.6431867451199983,
+      "learning_rate": 8.969124699325757e-06,
+      "loss": 0.4264,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2320219879394511,
+      "grad_norm": 0.6684884343355303,
+      "learning_rate": 8.968316511456241e-06,
+      "loss": 0.3734,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23210403248964187,
+      "grad_norm": 0.7499984103468563,
+      "learning_rate": 8.967508043350591e-06,
+      "loss": 0.4023,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23218607703983263,
+      "grad_norm": 0.670011303186805,
+      "learning_rate": 8.966699295065896e-06,
+      "loss": 0.3411,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23226812159002339,
+      "grad_norm": 0.6148100098038527,
+      "learning_rate": 8.96589026665927e-06,
+      "loss": 0.3738,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23235016614021414,
+      "grad_norm": 0.6563853724912911,
+      "learning_rate": 8.965080958187844e-06,
+      "loss": 0.3834,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2324322106904049,
+      "grad_norm": 0.6587105449591535,
+      "learning_rate": 8.964271369708772e-06,
+      "loss": 0.3907,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23251425524059563,
+      "grad_norm": 0.6166844155017971,
+      "learning_rate": 8.963461501279226e-06,
+      "loss": 0.3823,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2325962997907864,
+      "grad_norm": 0.6491379102190407,
+      "learning_rate": 8.962651352956395e-06,
+      "loss": 0.3807,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23267834434097714,
+      "grad_norm": 0.6411009912319582,
+      "learning_rate": 8.961840924797494e-06,
+      "loss": 0.4077,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2327603888911679,
+      "grad_norm": 0.6495545750076355,
+      "learning_rate": 8.96103021685975e-06,
+      "loss": 0.4024,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23284243344135866,
+      "grad_norm": 0.6506503987888818,
+      "learning_rate": 8.960219229200415e-06,
+      "loss": 0.4101,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23292447799154942,
+      "grad_norm": 0.698790370202475,
+      "learning_rate": 8.959407961876763e-06,
+      "loss": 0.3732,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23300652254174017,
+      "grad_norm": 0.6518050638924852,
+      "learning_rate": 8.95859641494608e-06,
+      "loss": 0.3973,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23308856709193093,
+      "grad_norm": 0.6810023785928286,
+      "learning_rate": 8.957784588465681e-06,
+      "loss": 0.4023,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23317061164212166,
+      "grad_norm": 0.6320474202454498,
+      "learning_rate": 8.956972482492892e-06,
+      "loss": 0.3423,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23325265619231242,
+      "grad_norm": 0.6874450829400771,
+      "learning_rate": 8.956160097085063e-06,
+      "loss": 0.4018,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23333470074250318,
+      "grad_norm": 0.6514770069923067,
+      "learning_rate": 8.955347432299564e-06,
+      "loss": 0.3577,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23341674529269393,
+      "grad_norm": 0.7935894061753583,
+      "learning_rate": 8.954534488193784e-06,
+      "loss": 0.3955,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2334987898428847,
+      "grad_norm": 0.6653489701000806,
+      "learning_rate": 8.95372126482513e-06,
+      "loss": 0.3888,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23358083439307545,
+      "grad_norm": 0.6743321486495167,
+      "learning_rate": 8.952907762251035e-06,
+      "loss": 0.3807,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2336628789432662,
+      "grad_norm": 0.6493644849579343,
+      "learning_rate": 8.952093980528944e-06,
+      "loss": 0.3891,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23374492349345694,
+      "grad_norm": 0.6834333134738156,
+      "learning_rate": 8.951279919716327e-06,
+      "loss": 0.3986,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2338269680436477,
+      "grad_norm": 0.6344322860802195,
+      "learning_rate": 8.95046557987067e-06,
+      "loss": 0.3737,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23390901259383845,
+      "grad_norm": 0.6223410778258939,
+      "learning_rate": 8.949650961049479e-06,
+      "loss": 0.3864,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2339910571440292,
+      "grad_norm": 0.6323982676845853,
+      "learning_rate": 8.948836063310282e-06,
+      "loss": 0.3711,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23407310169421996,
+      "grad_norm": 0.6470325707757156,
+      "learning_rate": 8.948020886710628e-06,
+      "loss": 0.4038,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23415514624441072,
+      "grad_norm": 0.5837683183602504,
+      "learning_rate": 8.94720543130808e-06,
+      "loss": 0.359,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23423719079460148,
+      "grad_norm": 0.6088437018672478,
+      "learning_rate": 8.94638969716023e-06,
+      "loss": 0.3559,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2343192353447922,
+      "grad_norm": 0.6512790594952518,
+      "learning_rate": 8.945573684324678e-06,
+      "loss": 0.4436,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23440127989498297,
+      "grad_norm": 0.6660856469952046,
+      "learning_rate": 8.944757392859051e-06,
+      "loss": 0.3641,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23448332444517372,
+      "grad_norm": 1.1417493521102622,
+      "learning_rate": 8.943940822820995e-06,
+      "loss": 0.3773,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23456536899536448,
+      "grad_norm": 0.72666207395507,
+      "learning_rate": 8.943123974268176e-06,
+      "loss": 0.3777,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23464741354555524,
+      "grad_norm": 0.7084710293200186,
+      "learning_rate": 8.942306847258274e-06,
+      "loss": 0.4124,
+      "step": 2860
+    },
+    {
+      "epoch": 0.234729458095746,
+      "grad_norm": 0.6067822316659417,
+      "learning_rate": 8.941489441848998e-06,
+      "loss": 0.3679,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23481150264593675,
+      "grad_norm": 0.6698828780806523,
+      "learning_rate": 8.94067175809807e-06,
+      "loss": 0.3877,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23489354719612748,
+      "grad_norm": 0.6556076532304588,
+      "learning_rate": 8.939853796063233e-06,
+      "loss": 0.3781,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23497559174631824,
+      "grad_norm": 0.6578205863896038,
+      "learning_rate": 8.939035555802252e-06,
+      "loss": 0.4005,
+      "step": 2864
+    },
+    {
+      "epoch": 0.235057636296509,
+      "grad_norm": 0.5811174706792741,
+      "learning_rate": 8.938217037372906e-06,
+      "loss": 0.3574,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23513968084669976,
+      "grad_norm": 0.637130252938261,
+      "learning_rate": 8.937398240833003e-06,
+      "loss": 0.3604,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2352217253968905,
+      "grad_norm": 0.5871356700218356,
+      "learning_rate": 8.936579166240361e-06,
+      "loss": 0.3601,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23530376994708127,
+      "grad_norm": 0.6021555393083228,
+      "learning_rate": 8.935759813652823e-06,
+      "loss": 0.3564,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23538581449727203,
+      "grad_norm": 0.6446333887410599,
+      "learning_rate": 8.93494018312825e-06,
+      "loss": 0.41,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23546785904746279,
+      "grad_norm": 0.674192447075651,
+      "learning_rate": 8.934120274724524e-06,
+      "loss": 0.3846,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23554990359765351,
+      "grad_norm": 0.6215204071546431,
+      "learning_rate": 8.933300088499544e-06,
+      "loss": 0.3838,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23563194814784427,
+      "grad_norm": 0.5919991432779403,
+      "learning_rate": 8.932479624511233e-06,
+      "loss": 0.3645,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23571399269803503,
+      "grad_norm": 0.6591273631512962,
+      "learning_rate": 8.931658882817526e-06,
+      "loss": 0.3705,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2357960372482258,
+      "grad_norm": 0.6149007667897357,
+      "learning_rate": 8.930837863476386e-06,
+      "loss": 0.3678,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23587808179841654,
+      "grad_norm": 0.6451157429586025,
+      "learning_rate": 8.930016566545791e-06,
+      "loss": 0.3868,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2359601263486073,
+      "grad_norm": 0.6347826799097732,
+      "learning_rate": 8.929194992083742e-06,
+      "loss": 0.3437,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23604217089879806,
+      "grad_norm": 0.6163991769587065,
+      "learning_rate": 8.928373140148256e-06,
+      "loss": 0.3827,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2361242154489888,
+      "grad_norm": 0.5732857456768983,
+      "learning_rate": 8.927551010797368e-06,
+      "loss": 0.3644,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23620625999917955,
+      "grad_norm": 0.7247437718389891,
+      "learning_rate": 8.926728604089138e-06,
+      "loss": 0.3375,
+      "step": 2879
+    },
+    {
+      "epoch": 0.2362883045493703,
+      "grad_norm": 0.6174532971840703,
+      "learning_rate": 8.925905920081644e-06,
+      "loss": 0.3893,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23637034909956106,
+      "grad_norm": 0.5518504980687496,
+      "learning_rate": 8.92508295883298e-06,
+      "loss": 0.3274,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23645239364975182,
+      "grad_norm": 0.6621168245382867,
+      "learning_rate": 8.924259720401266e-06,
+      "loss": 0.403,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23653443819994258,
+      "grad_norm": 0.6469375177910959,
+      "learning_rate": 8.923436204844634e-06,
+      "loss": 0.364,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23661648275013333,
+      "grad_norm": 0.6693256929742419,
+      "learning_rate": 8.92261241222124e-06,
+      "loss": 0.3943,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23669852730032406,
+      "grad_norm": 0.6434986567676918,
+      "learning_rate": 8.921788342589259e-06,
+      "loss": 0.3697,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23678057185051482,
+      "grad_norm": 0.619912294918691,
+      "learning_rate": 8.920963996006886e-06,
+      "loss": 0.3608,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23686261640070558,
+      "grad_norm": 0.5919551511495076,
+      "learning_rate": 8.920139372532337e-06,
+      "loss": 0.3582,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23694466095089634,
+      "grad_norm": 0.6508571365693582,
+      "learning_rate": 8.919314472223844e-06,
+      "loss": 0.3795,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2370267055010871,
+      "grad_norm": 0.6314271336521535,
+      "learning_rate": 8.918489295139658e-06,
+      "loss": 0.3873,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23710875005127785,
+      "grad_norm": 0.6708615986499037,
+      "learning_rate": 8.917663841338053e-06,
+      "loss": 0.4051,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2371907946014686,
+      "grad_norm": 0.699905471308614,
+      "learning_rate": 8.916838110877324e-06,
+      "loss": 0.4113,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23727283915165936,
+      "grad_norm": 0.5887714025289545,
+      "learning_rate": 8.916012103815779e-06,
+      "loss": 0.3643,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2373548837018501,
+      "grad_norm": 0.6251878422598868,
+      "learning_rate": 8.915185820211752e-06,
+      "loss": 0.3687,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23743692825204085,
+      "grad_norm": 0.6364729293854063,
+      "learning_rate": 8.914359260123591e-06,
+      "loss": 0.365,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2375189728022316,
+      "grad_norm": 0.6662931769749016,
+      "learning_rate": 8.913532423609669e-06,
+      "loss": 0.3432,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23760101735242237,
+      "grad_norm": 0.6280052768572216,
+      "learning_rate": 8.912705310728374e-06,
+      "loss": 0.3628,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23768306190261312,
+      "grad_norm": 0.6036907500157082,
+      "learning_rate": 8.911877921538117e-06,
+      "loss": 0.3462,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23776510645280388,
+      "grad_norm": 0.7014054968638831,
+      "learning_rate": 8.911050256097325e-06,
+      "loss": 0.4233,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23784715100299464,
+      "grad_norm": 0.6778717261016785,
+      "learning_rate": 8.91022231446445e-06,
+      "loss": 0.3563,
+      "step": 2899
+    },
+    {
+      "epoch": 0.23792919555318537,
+      "grad_norm": 0.7435442881354691,
+      "learning_rate": 8.909394096697954e-06,
+      "loss": 0.3837,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23801124010337613,
+      "grad_norm": 0.6541384230453,
+      "learning_rate": 8.908565602856331e-06,
+      "loss": 0.4098,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23809328465356688,
+      "grad_norm": 0.6082014240568439,
+      "learning_rate": 8.907736832998084e-06,
+      "loss": 0.3767,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23817532920375764,
+      "grad_norm": 0.6247864116744205,
+      "learning_rate": 8.906907787181738e-06,
+      "loss": 0.3683,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2382573737539484,
+      "grad_norm": 0.65853141758166,
+      "learning_rate": 8.906078465465842e-06,
+      "loss": 0.3957,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23833941830413916,
+      "grad_norm": 0.5928337123809184,
+      "learning_rate": 8.90524886790896e-06,
+      "loss": 0.3742,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2384214628543299,
+      "grad_norm": 0.6558475490514107,
+      "learning_rate": 8.90441899456968e-06,
+      "loss": 0.3687,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23850350740452064,
+      "grad_norm": 0.6722866966202312,
+      "learning_rate": 8.903588845506602e-06,
+      "loss": 0.3963,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2385855519547114,
+      "grad_norm": 0.6362781646396646,
+      "learning_rate": 8.90275842077835e-06,
+      "loss": 0.3627,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23866759650490216,
+      "grad_norm": 0.624801228882643,
+      "learning_rate": 8.90192772044357e-06,
+      "loss": 0.3721,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23874964105509291,
+      "grad_norm": 0.692194649705521,
+      "learning_rate": 8.901096744560922e-06,
+      "loss": 0.3883,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23883168560528367,
+      "grad_norm": 0.6189195145213632,
+      "learning_rate": 8.900265493189088e-06,
+      "loss": 0.372,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23891373015547443,
+      "grad_norm": 0.6637043928086939,
+      "learning_rate": 8.899433966386774e-06,
+      "loss": 0.3389,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2389957747056652,
+      "grad_norm": 0.8684461174639787,
+      "learning_rate": 8.898602164212699e-06,
+      "loss": 0.6889,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23907781925585592,
+      "grad_norm": 0.6435246012276016,
+      "learning_rate": 8.8977700867256e-06,
+      "loss": 0.376,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23915986380604667,
+      "grad_norm": 0.6559197268086102,
+      "learning_rate": 8.896937733984241e-06,
+      "loss": 0.3339,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23924190835623743,
+      "grad_norm": 0.6232006263404782,
+      "learning_rate": 8.896105106047402e-06,
+      "loss": 0.3403,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2393239529064282,
+      "grad_norm": 0.6134458403535092,
+      "learning_rate": 8.895272202973878e-06,
+      "loss": 0.3551,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23940599745661895,
+      "grad_norm": 0.6260810664758947,
+      "learning_rate": 8.894439024822488e-06,
+      "loss": 0.3553,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2394880420068097,
+      "grad_norm": 0.6969805421703364,
+      "learning_rate": 8.893605571652073e-06,
+      "loss": 0.3716,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23957008655700046,
+      "grad_norm": 0.6146901423571102,
+      "learning_rate": 8.892771843521487e-06,
+      "loss": 0.3738,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23965213110719122,
+      "grad_norm": 0.6435564088146203,
+      "learning_rate": 8.89193784048961e-06,
+      "loss": 0.3798,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23973417565738195,
+      "grad_norm": 0.6699994576962305,
+      "learning_rate": 8.891103562615332e-06,
+      "loss": 0.3708,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2398162202075727,
+      "grad_norm": 0.651367205231876,
+      "learning_rate": 8.890269009957572e-06,
+      "loss": 0.3599,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23989826475776346,
+      "grad_norm": 0.6854914543971296,
+      "learning_rate": 8.889434182575268e-06,
+      "loss": 0.3848,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23998030930795422,
+      "grad_norm": 0.6424518915258374,
+      "learning_rate": 8.888599080527369e-06,
+      "loss": 0.4014,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24006235385814498,
+      "grad_norm": 0.6328570802454909,
+      "learning_rate": 8.88776370387285e-06,
+      "loss": 0.3518,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24014439840833574,
+      "grad_norm": 0.6178950341582466,
+      "learning_rate": 8.886928052670705e-06,
+      "loss": 0.3725,
+      "step": 2927
+    },
+    {
+      "epoch": 0.2402264429585265,
+      "grad_norm": 0.5799861657001326,
+      "learning_rate": 8.886092126979947e-06,
+      "loss": 0.3417,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24030848750871722,
+      "grad_norm": 0.6375117269402045,
+      "learning_rate": 8.885255926859604e-06,
+      "loss": 0.4144,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24039053205890798,
+      "grad_norm": 0.6888364022092032,
+      "learning_rate": 8.884419452368731e-06,
+      "loss": 0.3917,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24047257660909874,
+      "grad_norm": 0.5967390926033186,
+      "learning_rate": 8.883582703566396e-06,
+      "loss": 0.3653,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2405546211592895,
+      "grad_norm": 0.637205329497468,
+      "learning_rate": 8.882745680511691e-06,
+      "loss": 0.366,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24063666570948025,
+      "grad_norm": 0.7285719588182957,
+      "learning_rate": 8.881908383263724e-06,
+      "loss": 0.3495,
+      "step": 2933
+    },
+    {
+      "epoch": 0.240718710259671,
+      "grad_norm": 0.62935659474105,
+      "learning_rate": 8.881070811881625e-06,
+      "loss": 0.3619,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24080075480986177,
+      "grad_norm": 0.59605408510328,
+      "learning_rate": 8.880232966424538e-06,
+      "loss": 0.4114,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2408827993600525,
+      "grad_norm": 0.6019998863293899,
+      "learning_rate": 8.879394846951634e-06,
+      "loss": 0.3787,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24096484391024325,
+      "grad_norm": 0.665972488237809,
+      "learning_rate": 8.8785564535221e-06,
+      "loss": 0.376,
+      "step": 2937
+    },
+    {
+      "epoch": 0.241046888460434,
+      "grad_norm": 0.5781642148684826,
+      "learning_rate": 8.877717786195139e-06,
+      "loss": 0.3302,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24112893301062477,
+      "grad_norm": 3.7848056234878324,
+      "learning_rate": 8.87687884502998e-06,
+      "loss": 0.3911,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24121097756081553,
+      "grad_norm": 0.6025707738219767,
+      "learning_rate": 8.876039630085864e-06,
+      "loss": 0.3639,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24129302211100628,
+      "grad_norm": 0.6217113902553154,
+      "learning_rate": 8.875200141422055e-06,
+      "loss": 0.3565,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24137506666119704,
+      "grad_norm": 0.6725469128729444,
+      "learning_rate": 8.874360379097842e-06,
+      "loss": 0.3816,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24145711121138777,
+      "grad_norm": 0.6701578198273084,
+      "learning_rate": 8.87352034317252e-06,
+      "loss": 0.378,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24153915576157853,
+      "grad_norm": 0.6525777360426738,
+      "learning_rate": 8.872680033705417e-06,
+      "loss": 0.4025,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24162120031176929,
+      "grad_norm": 0.8122632291807959,
+      "learning_rate": 8.871839450755871e-06,
+      "loss": 0.3682,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24170324486196004,
+      "grad_norm": 0.7128929761661555,
+      "learning_rate": 8.870998594383243e-06,
+      "loss": 0.3429,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2417852894121508,
+      "grad_norm": 0.6766362915481693,
+      "learning_rate": 8.870157464646914e-06,
+      "loss": 0.3743,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24186733396234156,
+      "grad_norm": 0.658855529971069,
+      "learning_rate": 8.869316061606284e-06,
+      "loss": 0.3982,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24194937851253231,
+      "grad_norm": 1.8583630306099317,
+      "learning_rate": 8.868474385320768e-06,
+      "loss": 0.3816,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24203142306272307,
+      "grad_norm": 0.6813865636272286,
+      "learning_rate": 8.867632435849807e-06,
+      "loss": 0.3779,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2421134676129138,
+      "grad_norm": 0.7115334542252896,
+      "learning_rate": 8.866790213252857e-06,
+      "loss": 0.3634,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24219551216310456,
+      "grad_norm": 1.0315392331445858,
+      "learning_rate": 8.865947717589395e-06,
+      "loss": 0.4151,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24227755671329532,
+      "grad_norm": 1.3275459719613265,
+      "learning_rate": 8.865104948918919e-06,
+      "loss": 0.3617,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24235960126348607,
+      "grad_norm": 0.819239676945428,
+      "learning_rate": 8.864261907300938e-06,
+      "loss": 0.3903,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24244164581367683,
+      "grad_norm": 0.5979350611811772,
+      "learning_rate": 8.86341859279499e-06,
+      "loss": 0.3733,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2425236903638676,
+      "grad_norm": 1.0125180520667478,
+      "learning_rate": 8.862575005460627e-06,
+      "loss": 0.3803,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24260573491405835,
+      "grad_norm": 0.6980702206783775,
+      "learning_rate": 8.861731145357425e-06,
+      "loss": 0.3559,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24268777946424908,
+      "grad_norm": 0.6846648321188114,
+      "learning_rate": 8.860887012544973e-06,
+      "loss": 0.3581,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24276982401443983,
+      "grad_norm": 3.8836180252393167,
+      "learning_rate": 8.860042607082884e-06,
+      "loss": 0.407,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2428518685646306,
+      "grad_norm": 0.7166740670075609,
+      "learning_rate": 8.859197929030787e-06,
+      "loss": 0.353,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24293391311482135,
+      "grad_norm": 68.18688749380681,
+      "learning_rate": 8.858352978448332e-06,
+      "loss": 0.3527,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2430159576650121,
+      "grad_norm": 0.6286895626728577,
+      "learning_rate": 8.857507755395189e-06,
+      "loss": 0.3684,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24309800221520286,
+      "grad_norm": 0.7742085822591436,
+      "learning_rate": 8.856662259931044e-06,
+      "loss": 0.344,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24318004676539362,
+      "grad_norm": 0.6397985939186634,
+      "learning_rate": 8.85581649211561e-06,
+      "loss": 0.3798,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24326209131558435,
+      "grad_norm": 0.7391188921762428,
+      "learning_rate": 8.854970452008607e-06,
+      "loss": 0.4344,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2433441358657751,
+      "grad_norm": 0.8616388630966433,
+      "learning_rate": 8.854124139669786e-06,
+      "loss": 0.3522,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24342618041596586,
+      "grad_norm": 0.8023915485466961,
+      "learning_rate": 8.85327755515891e-06,
+      "loss": 0.3494,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24350822496615662,
+      "grad_norm": 0.6581434876308332,
+      "learning_rate": 8.852430698535763e-06,
+      "loss": 0.3308,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24359026951634738,
+      "grad_norm": 0.8750521533295929,
+      "learning_rate": 8.851583569860152e-06,
+      "loss": 0.3884,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24367231406653814,
+      "grad_norm": 0.7191204995461383,
+      "learning_rate": 8.850736169191894e-06,
+      "loss": 0.348,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2437543586167289,
+      "grad_norm": 1.0052298242852982,
+      "learning_rate": 8.849888496590836e-06,
+      "loss": 0.3829,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24383640316691962,
+      "grad_norm": 0.8419282681299323,
+      "learning_rate": 8.849040552116838e-06,
+      "loss": 0.4099,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24391844771711038,
+      "grad_norm": 0.8925208940562116,
+      "learning_rate": 8.848192335829779e-06,
+      "loss": 0.3996,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24400049226730114,
+      "grad_norm": 1.0241052704304432,
+      "learning_rate": 8.847343847789561e-06,
+      "loss": 0.3568,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2440825368174919,
+      "grad_norm": 1.137870100440833,
+      "learning_rate": 8.846495088056102e-06,
+      "loss": 0.355,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24416458136768265,
+      "grad_norm": 1.6860574513736974,
+      "learning_rate": 8.845646056689338e-06,
+      "loss": 0.3371,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2442466259178734,
+      "grad_norm": 1.5749898552406878,
+      "learning_rate": 8.84479675374923e-06,
+      "loss": 0.3681,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24432867046806417,
+      "grad_norm": 0.6993195843086124,
+      "learning_rate": 8.84394717929575e-06,
+      "loss": 0.4115,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24441071501825493,
+      "grad_norm": 1.33168649437014,
+      "learning_rate": 8.843097333388899e-06,
+      "loss": 0.3997,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24449275956844566,
+      "grad_norm": 0.6674422972469435,
+      "learning_rate": 8.842247216088686e-06,
+      "loss": 0.3524,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2445748041186364,
+      "grad_norm": 1.011506674669749,
+      "learning_rate": 8.841396827455148e-06,
+      "loss": 0.3737,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24465684866882717,
+      "grad_norm": 0.9208283085501211,
+      "learning_rate": 8.84054616754834e-06,
+      "loss": 0.3936,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24473889321901793,
+      "grad_norm": 1.7350299210082005,
+      "learning_rate": 8.83969523642833e-06,
+      "loss": 0.3905,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24482093776920869,
+      "grad_norm": 0.6723485109809039,
+      "learning_rate": 8.838844034155212e-06,
+      "loss": 0.3621,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24490298231939944,
+      "grad_norm": 1.3872921989885618,
+      "learning_rate": 8.837992560789097e-06,
+      "loss": 0.4276,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2449850268695902,
+      "grad_norm": 0.7373517874793873,
+      "learning_rate": 8.83714081639011e-06,
+      "loss": 0.3551,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24506707141978093,
+      "grad_norm": 1.0603806876606234,
+      "learning_rate": 8.836288801018407e-06,
+      "loss": 0.3752,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2451491159699717,
+      "grad_norm": 0.9135588159242656,
+      "learning_rate": 8.83543651473415e-06,
+      "loss": 0.3816,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24523116052016244,
+      "grad_norm": 1.0892331224806229,
+      "learning_rate": 8.83458395759753e-06,
+      "loss": 0.3511,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2453132050703532,
+      "grad_norm": 1.0953726587840913,
+      "learning_rate": 8.83373112966875e-06,
+      "loss": 0.4146,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24539524962054396,
+      "grad_norm": 1.6219048080624463,
+      "learning_rate": 8.83287803100804e-06,
+      "loss": 0.4426,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24547729417073472,
+      "grad_norm": 0.7944427017362852,
+      "learning_rate": 8.832024661675638e-06,
+      "loss": 0.382,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24555933872092547,
+      "grad_norm": 1.2735612607266662,
+      "learning_rate": 8.831171021731813e-06,
+      "loss": 0.3677,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2456413832711162,
+      "grad_norm": 0.7496066643507194,
+      "learning_rate": 8.830317111236845e-06,
+      "loss": 0.4076,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24572342782130696,
+      "grad_norm": 1.0939212315275408,
+      "learning_rate": 8.829462930251036e-06,
+      "loss": 0.3595,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24580547237149772,
+      "grad_norm": 0.7084763321717339,
+      "learning_rate": 8.828608478834709e-06,
+      "loss": 0.373,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24588751692168848,
+      "grad_norm": 1.791695889600847,
+      "learning_rate": 8.8277537570482e-06,
+      "loss": 0.3579,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24596956147187923,
+      "grad_norm": 1.6731616090838826,
+      "learning_rate": 8.826898764951873e-06,
+      "loss": 0.3456,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24605160602207,
+      "grad_norm": 1.3987558829463436,
+      "learning_rate": 8.826043502606102e-06,
+      "loss": 0.3767,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24613365057226075,
+      "grad_norm": 0.7047340385992438,
+      "learning_rate": 8.825187970071285e-06,
+      "loss": 0.4165,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24621569512245148,
+      "grad_norm": 0.6892800029957916,
+      "learning_rate": 8.824332167407841e-06,
+      "loss": 0.3756,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24629773967264224,
+      "grad_norm": 0.8804511525739181,
+      "learning_rate": 8.823476094676204e-06,
+      "loss": 0.3971,
+      "step": 3002
+    },
+    {
+      "epoch": 0.246379784222833,
+      "grad_norm": 0.6948173451149168,
+      "learning_rate": 8.822619751936827e-06,
+      "loss": 0.3518,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24646182877302375,
+      "grad_norm": 1.4107031175128668,
+      "learning_rate": 8.821763139250185e-06,
+      "loss": 0.4015,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2465438733232145,
+      "grad_norm": 1.5328218913982867,
+      "learning_rate": 8.820906256676769e-06,
+      "loss": 0.3935,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24662591787340526,
+      "grad_norm": 3.1632452557076234,
+      "learning_rate": 8.820049104277094e-06,
+      "loss": 0.3642,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24670796242359602,
+      "grad_norm": 0.9080465456772747,
+      "learning_rate": 8.819191682111688e-06,
+      "loss": 0.367,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24679000697378678,
+      "grad_norm": 0.7985931729804,
+      "learning_rate": 8.8183339902411e-06,
+      "loss": 0.3701,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2468720515239775,
+      "grad_norm": 0.7448904294284471,
+      "learning_rate": 8.8174760287259e-06,
+      "loss": 0.3739,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24695409607416827,
+      "grad_norm": 0.913598843814343,
+      "learning_rate": 8.816617797626679e-06,
+      "loss": 0.3319,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24703614062435902,
+      "grad_norm": 1.1745708919610058,
+      "learning_rate": 8.815759297004038e-06,
+      "loss": 0.3957,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24711818517454978,
+      "grad_norm": 0.9024060409119015,
+      "learning_rate": 8.814900526918608e-06,
+      "loss": 0.3859,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24720022972474054,
+      "grad_norm": 1.417596370361803,
+      "learning_rate": 8.814041487431031e-06,
+      "loss": 0.3912,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2472822742749313,
+      "grad_norm": 0.639688275575993,
+      "learning_rate": 8.81318217860197e-06,
+      "loss": 0.4082,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24736431882512205,
+      "grad_norm": 0.7488936303860247,
+      "learning_rate": 8.812322600492114e-06,
+      "loss": 0.4034,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24744636337531278,
+      "grad_norm": 0.6703740975490076,
+      "learning_rate": 8.811462753162157e-06,
+      "loss": 0.3761,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24752840792550354,
+      "grad_norm": 0.8422298300499784,
+      "learning_rate": 8.810602636672827e-06,
+      "loss": 0.3513,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2476104524756943,
+      "grad_norm": 0.6918097269493877,
+      "learning_rate": 8.80974225108486e-06,
+      "loss": 0.3922,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24769249702588506,
+      "grad_norm": 0.676674827611269,
+      "learning_rate": 8.808881596459015e-06,
+      "loss": 0.3694,
+      "step": 3019
+    },
+    {
+      "epoch": 0.2477745415760758,
+      "grad_norm": 0.7210611531257359,
+      "learning_rate": 8.808020672856072e-06,
+      "loss": 0.3683,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24785658612626657,
+      "grad_norm": 0.6295419739809693,
+      "learning_rate": 8.807159480336827e-06,
+      "loss": 0.305,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24793863067645733,
+      "grad_norm": 0.6162706266320641,
+      "learning_rate": 8.806298018962097e-06,
+      "loss": 0.3492,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24802067522664806,
+      "grad_norm": 0.6312782944741793,
+      "learning_rate": 8.805436288792714e-06,
+      "loss": 0.4032,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24810271977683881,
+      "grad_norm": 0.6296488597631836,
+      "learning_rate": 8.804574289889536e-06,
+      "loss": 0.3882,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24818476432702957,
+      "grad_norm": 0.7166549387625955,
+      "learning_rate": 8.803712022313433e-06,
+      "loss": 0.3929,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24826680887722033,
+      "grad_norm": 0.7556890375643721,
+      "learning_rate": 8.8028494861253e-06,
+      "loss": 0.4198,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2483488534274111,
+      "grad_norm": 0.8018425809417878,
+      "learning_rate": 8.801986681386044e-06,
+      "loss": 0.3403,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24843089797760184,
+      "grad_norm": 0.6280914350240412,
+      "learning_rate": 8.801123608156597e-06,
+      "loss": 0.4029,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2485129425277926,
+      "grad_norm": 0.9391231158467825,
+      "learning_rate": 8.800260266497909e-06,
+      "loss": 0.4109,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24859498707798333,
+      "grad_norm": 0.7106335902286239,
+      "learning_rate": 8.799396656470946e-06,
+      "loss": 0.3843,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2486770316281741,
+      "grad_norm": 0.9968615290258042,
+      "learning_rate": 8.798532778136697e-06,
+      "loss": 0.3999,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24875907617836485,
+      "grad_norm": 0.7407498450785749,
+      "learning_rate": 8.797668631556163e-06,
+      "loss": 0.3768,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2488411207285556,
+      "grad_norm": 0.6316564959525762,
+      "learning_rate": 8.796804216790373e-06,
+      "loss": 0.3668,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24892316527874636,
+      "grad_norm": 0.892639849407937,
+      "learning_rate": 8.79593953390037e-06,
+      "loss": 0.3728,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24900520982893712,
+      "grad_norm": 0.6797198220073875,
+      "learning_rate": 8.795074582947214e-06,
+      "loss": 0.3583,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24908725437912788,
+      "grad_norm": 0.7928915643944122,
+      "learning_rate": 8.79420936399199e-06,
+      "loss": 0.3953,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24916929892931863,
+      "grad_norm": 1.0304504985137364,
+      "learning_rate": 8.793343877095795e-06,
+      "loss": 0.3469,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24925134347950936,
+      "grad_norm": 0.9216749731040303,
+      "learning_rate": 8.792478122319753e-06,
+      "loss": 0.3715,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24933338802970012,
+      "grad_norm": 0.7851333497720723,
+      "learning_rate": 8.791612099724996e-06,
+      "loss": 0.3803,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24941543257989088,
+      "grad_norm": 1.2000388047521169,
+      "learning_rate": 8.790745809372685e-06,
+      "loss": 0.3381,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24949747713008164,
+      "grad_norm": 1.1040804094062369,
+      "learning_rate": 8.789879251323994e-06,
+      "loss": 0.3312,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2495795216802724,
+      "grad_norm": 0.6510548362952556,
+      "learning_rate": 8.78901242564012e-06,
+      "loss": 0.349,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24966156623046315,
+      "grad_norm": 0.748662987209478,
+      "learning_rate": 8.788145332382277e-06,
+      "loss": 0.4081,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2497436107806539,
+      "grad_norm": 0.9527244100926405,
+      "learning_rate": 8.787277971611697e-06,
+      "loss": 0.3902,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24982565533084464,
+      "grad_norm": 1.146875550439966,
+      "learning_rate": 8.78641034338963e-06,
+      "loss": 0.3714,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2499076998810354,
+      "grad_norm": 0.675962389119729,
+      "learning_rate": 8.78554244777735e-06,
+      "loss": 0.3854,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24998974443122615,
+      "grad_norm": 0.702898146711128,
+      "learning_rate": 8.784674284836141e-06,
+      "loss": 0.3719,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2500717889814169,
+      "grad_norm": 0.5893852862835779,
+      "learning_rate": 8.783805854627317e-06,
+      "loss": 0.33,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25015383353160764,
+      "grad_norm": 0.9111823938331519,
+      "learning_rate": 8.782937157212201e-06,
+      "loss": 0.3741,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2502358780817984,
+      "grad_norm": 0.828993825505094,
+      "learning_rate": 8.782068192652143e-06,
+      "loss": 0.397,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25031792263198915,
+      "grad_norm": 0.9204120507503143,
+      "learning_rate": 8.781198961008504e-06,
+      "loss": 0.386,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25039996718217994,
+      "grad_norm": 0.6389945631281025,
+      "learning_rate": 8.780329462342669e-06,
+      "loss": 0.3753,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25048201173237067,
+      "grad_norm": 1.1067516642738267,
+      "learning_rate": 8.779459696716042e-06,
+      "loss": 0.359,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25056405628256145,
+      "grad_norm": 0.9991267696738696,
+      "learning_rate": 8.778589664190043e-06,
+      "loss": 0.3847,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2506461008327522,
+      "grad_norm": 0.642300597176655,
+      "learning_rate": 8.777719364826112e-06,
+      "loss": 0.3386,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2507281453829429,
+      "grad_norm": 0.7457795379850656,
+      "learning_rate": 8.77684879868571e-06,
+      "loss": 0.3581,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2508101899331337,
+      "grad_norm": 0.7290220715012876,
+      "learning_rate": 8.775977965830312e-06,
+      "loss": 0.3859,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25089223448332443,
+      "grad_norm": 0.7150292323788718,
+      "learning_rate": 8.775106866321419e-06,
+      "loss": 0.3536,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2509742790335152,
+      "grad_norm": 0.6151029625758356,
+      "learning_rate": 8.774235500220542e-06,
+      "loss": 0.3955,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25105632358370594,
+      "grad_norm": 0.8500418380247492,
+      "learning_rate": 8.773363867589219e-06,
+      "loss": 0.3846,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2511383681338967,
+      "grad_norm": 0.7340326277096766,
+      "learning_rate": 8.772491968489003e-06,
+      "loss": 0.3553,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25122041268408746,
+      "grad_norm": 0.9137741147682489,
+      "learning_rate": 8.771619802981463e-06,
+      "loss": 0.378,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2513024572342782,
+      "grad_norm": 0.6801281100191163,
+      "learning_rate": 8.770747371128195e-06,
+      "loss": 0.3893,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25138450178446897,
+      "grad_norm": 0.6934673904657431,
+      "learning_rate": 8.769874672990802e-06,
+      "loss": 0.3761,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2514665463346597,
+      "grad_norm": 0.704497505988441,
+      "learning_rate": 8.76900170863092e-06,
+      "loss": 0.4173,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2515485908848505,
+      "grad_norm": 0.649337551494044,
+      "learning_rate": 8.76812847811019e-06,
+      "loss": 0.3981,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2516306354350412,
+      "grad_norm": 0.7441478229283116,
+      "learning_rate": 8.767254981490282e-06,
+      "loss": 0.3991,
+      "step": 3067
+    },
+    {
+      "epoch": 0.251712679985232,
+      "grad_norm": 0.7252684379492031,
+      "learning_rate": 8.76638121883288e-06,
+      "loss": 0.3713,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25179472453542273,
+      "grad_norm": 0.6737842172951855,
+      "learning_rate": 8.765507190199687e-06,
+      "loss": 0.3546,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25187676908561346,
+      "grad_norm": 0.6950362241773563,
+      "learning_rate": 8.764632895652425e-06,
+      "loss": 0.3407,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25195881363580425,
+      "grad_norm": 0.8296314664343977,
+      "learning_rate": 8.763758335252838e-06,
+      "loss": 0.3639,
+      "step": 3071
+    },
+    {
+      "epoch": 0.252040858185995,
+      "grad_norm": 0.7643804841735021,
+      "learning_rate": 8.762883509062684e-06,
+      "loss": 0.3629,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25212290273618576,
+      "grad_norm": 0.8889505624952746,
+      "learning_rate": 8.762008417143742e-06,
+      "loss": 0.4237,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2522049472863765,
+      "grad_norm": 0.6238035289740788,
+      "learning_rate": 8.761133059557809e-06,
+      "loss": 0.3777,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2522869918365673,
+      "grad_norm": 0.76396875095185,
+      "learning_rate": 8.760257436366703e-06,
+      "loss": 0.3724,
+      "step": 3075
+    },
+    {
+      "epoch": 0.252369036386758,
+      "grad_norm": 0.6519233319380319,
+      "learning_rate": 8.759381547632258e-06,
+      "loss": 0.3901,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2524510809369488,
+      "grad_norm": 0.682121211004433,
+      "learning_rate": 8.758505393416327e-06,
+      "loss": 0.3526,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2525331254871395,
+      "grad_norm": 0.7626716712707696,
+      "learning_rate": 8.757628973780785e-06,
+      "loss": 0.3923,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25261517003733025,
+      "grad_norm": 0.7330884134346776,
+      "learning_rate": 8.75675228878752e-06,
+      "loss": 0.4167,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25269721458752104,
+      "grad_norm": 0.6871625114297178,
+      "learning_rate": 8.755875338498445e-06,
+      "loss": 0.3531,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25277925913771176,
+      "grad_norm": 0.653881730237364,
+      "learning_rate": 8.754998122975489e-06,
+      "loss": 0.3419,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25286130368790255,
+      "grad_norm": 0.7733595193509784,
+      "learning_rate": 8.754120642280597e-06,
+      "loss": 0.4616,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2529433482380933,
+      "grad_norm": 0.7332296891829597,
+      "learning_rate": 8.753242896475737e-06,
+      "loss": 0.3927,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25302539278828406,
+      "grad_norm": 1.1917109692431973,
+      "learning_rate": 8.752364885622895e-06,
+      "loss": 0.3788,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2531074373384748,
+      "grad_norm": 0.6938713731334266,
+      "learning_rate": 8.751486609784073e-06,
+      "loss": 0.3564,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2531894818886655,
+      "grad_norm": 0.6756648714475658,
+      "learning_rate": 8.750608069021292e-06,
+      "loss": 0.3866,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2532715264388563,
+      "grad_norm": 0.654213081297272,
+      "learning_rate": 8.749729263396597e-06,
+      "loss": 0.3389,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25335357098904704,
+      "grad_norm": 0.8856446479266715,
+      "learning_rate": 8.748850192972046e-06,
+      "loss": 0.3685,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2534356155392378,
+      "grad_norm": 1.0602983304927052,
+      "learning_rate": 8.747970857809715e-06,
+      "loss": 0.3648,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25351766008942855,
+      "grad_norm": 0.7159911380712197,
+      "learning_rate": 8.747091257971706e-06,
+      "loss": 0.3416,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25359970463961934,
+      "grad_norm": 0.9648275976423778,
+      "learning_rate": 8.746211393520131e-06,
+      "loss": 0.3737,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25368174918981007,
+      "grad_norm": 0.6486625779394859,
+      "learning_rate": 8.745331264517126e-06,
+      "loss": 0.3688,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2537637937400008,
+      "grad_norm": 0.644147888213152,
+      "learning_rate": 8.744450871024846e-06,
+      "loss": 0.4088,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2538458382901916,
+      "grad_norm": 0.8956896047725762,
+      "learning_rate": 8.74357021310546e-06,
+      "loss": 0.3909,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2539278828403823,
+      "grad_norm": 1.026362358628755,
+      "learning_rate": 8.74268929082116e-06,
+      "loss": 0.3455,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2540099273905731,
+      "grad_norm": 0.6290164191936498,
+      "learning_rate": 8.741808104234157e-06,
+      "loss": 0.4243,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2540919719407638,
+      "grad_norm": 0.988097534016537,
+      "learning_rate": 8.740926653406676e-06,
+      "loss": 0.3837,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2541740164909546,
+      "grad_norm": 0.6539478522582054,
+      "learning_rate": 8.740044938400965e-06,
+      "loss": 0.3659,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25425606104114534,
+      "grad_norm": 0.6980028278389073,
+      "learning_rate": 8.73916295927929e-06,
+      "loss": 0.3777,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25433810559133607,
+      "grad_norm": 0.7182057234326134,
+      "learning_rate": 8.738280716103933e-06,
+      "loss": 0.3894,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25442015014152686,
+      "grad_norm": 0.7106198030103167,
+      "learning_rate": 8.7373982089372e-06,
+      "loss": 0.3854,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2545021946917176,
+      "grad_norm": 0.6866009263146211,
+      "learning_rate": 8.736515437841408e-06,
+      "loss": 0.3821,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25458423924190837,
+      "grad_norm": 0.6293881911369105,
+      "learning_rate": 8.7356324028789e-06,
+      "loss": 0.4356,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2546662837920991,
+      "grad_norm": 0.6421007613734946,
+      "learning_rate": 8.734749104112032e-06,
+      "loss": 0.382,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2547483283422899,
+      "grad_norm": 0.7882450782968005,
+      "learning_rate": 8.733865541603185e-06,
+      "loss": 0.4328,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2548303728924806,
+      "grad_norm": 0.6729069518455423,
+      "learning_rate": 8.73298171541475e-06,
+      "loss": 0.4106,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25491241744267135,
+      "grad_norm": 0.6719987722859339,
+      "learning_rate": 8.732097625609145e-06,
+      "loss": 0.3882,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25499446199286213,
+      "grad_norm": 0.78185861093442,
+      "learning_rate": 8.731213272248803e-06,
+      "loss": 0.387,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25507650654305286,
+      "grad_norm": 0.7203179324118741,
+      "learning_rate": 8.730328655396171e-06,
+      "loss": 0.3944,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25515855109324365,
+      "grad_norm": 0.6432028957022919,
+      "learning_rate": 8.729443775113727e-06,
+      "loss": 0.3573,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2552405956434344,
+      "grad_norm": 0.9474889377858284,
+      "learning_rate": 8.728558631463951e-06,
+      "loss": 0.3882,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25532264019362516,
+      "grad_norm": 0.8867881523997386,
+      "learning_rate": 8.727673224509358e-06,
+      "loss": 0.3503,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2554046847438159,
+      "grad_norm": 0.747558353303817,
+      "learning_rate": 8.72678755431247e-06,
+      "loss": 0.332,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2554867292940066,
+      "grad_norm": 0.7508825298846448,
+      "learning_rate": 8.725901620935832e-06,
+      "loss": 0.4018,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2555687738441974,
+      "grad_norm": 0.7161126731747893,
+      "learning_rate": 8.725015424442008e-06,
+      "loss": 0.4386,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25565081839438814,
+      "grad_norm": 0.6145420502089158,
+      "learning_rate": 8.724128964893577e-06,
+      "loss": 0.3746,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2557328629445789,
+      "grad_norm": 0.6279896416480096,
+      "learning_rate": 8.723242242353144e-06,
+      "loss": 0.3473,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25581490749476965,
+      "grad_norm": 0.8217677090022822,
+      "learning_rate": 8.722355256883327e-06,
+      "loss": 0.3553,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25589695204496044,
+      "grad_norm": 0.7874414825887118,
+      "learning_rate": 8.721468008546761e-06,
+      "loss": 0.3634,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25597899659515116,
+      "grad_norm": 1.28294732132005,
+      "learning_rate": 8.720580497406103e-06,
+      "loss": 0.4106,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2560610411453419,
+      "grad_norm": 0.651098870704053,
+      "learning_rate": 8.719692723524028e-06,
+      "loss": 0.3524,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2561430856955327,
+      "grad_norm": 0.6494304494663434,
+      "learning_rate": 8.718804686963229e-06,
+      "loss": 0.3786,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2562251302457234,
+      "grad_norm": 0.6631088836328631,
+      "learning_rate": 8.717916387786417e-06,
+      "loss": 0.3768,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2563071747959142,
+      "grad_norm": 0.6573655231048589,
+      "learning_rate": 8.717027826056324e-06,
+      "loss": 0.3747,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2563892193461049,
+      "grad_norm": 0.7067813693421778,
+      "learning_rate": 8.716139001835698e-06,
+      "loss": 0.3587,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2564712638962957,
+      "grad_norm": 0.623039176360185,
+      "learning_rate": 8.715249915187305e-06,
+      "loss": 0.344,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25655330844648644,
+      "grad_norm": 0.6425020434895568,
+      "learning_rate": 8.714360566173932e-06,
+      "loss": 0.3606,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25663535299667717,
+      "grad_norm": 0.7357136308599259,
+      "learning_rate": 8.713470954858385e-06,
+      "loss": 0.4179,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25671739754686795,
+      "grad_norm": 0.8416991652523412,
+      "learning_rate": 8.712581081303482e-06,
+      "loss": 0.4084,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2567994420970587,
+      "grad_norm": 0.6194575037564201,
+      "learning_rate": 8.71169094557207e-06,
+      "loss": 0.326,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25688148664724947,
+      "grad_norm": 0.6477605472957103,
+      "learning_rate": 8.710800547727008e-06,
+      "loss": 0.356,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2569635311974402,
+      "grad_norm": 0.6478904926986937,
+      "learning_rate": 8.709909887831172e-06,
+      "loss": 0.3759,
+      "step": 3132
+    },
+    {
+      "epoch": 0.257045575747631,
+      "grad_norm": 0.7428250866988848,
+      "learning_rate": 8.70901896594746e-06,
+      "loss": 0.3539,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2571276202978217,
+      "grad_norm": 0.8011847400343893,
+      "learning_rate": 8.708127782138788e-06,
+      "loss": 0.3959,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2572096648480125,
+      "grad_norm": 0.7458113687547491,
+      "learning_rate": 8.707236336468089e-06,
+      "loss": 0.3474,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2572917093982032,
+      "grad_norm": 0.6841930482702712,
+      "learning_rate": 8.706344628998315e-06,
+      "loss": 0.3868,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25737375394839396,
+      "grad_norm": 0.6558018591547164,
+      "learning_rate": 8.705452659792439e-06,
+      "loss": 0.4093,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25745579849858474,
+      "grad_norm": 0.7643772301741847,
+      "learning_rate": 8.704560428913451e-06,
+      "loss": 0.3612,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25753784304877547,
+      "grad_norm": 0.7424897260193716,
+      "learning_rate": 8.703667936424357e-06,
+      "loss": 0.3534,
+      "step": 3139
+    },
+    {
+      "epoch": 0.25761988759896626,
+      "grad_norm": 0.7388482348297393,
+      "learning_rate": 8.702775182388182e-06,
+      "loss": 0.3565,
+      "step": 3140
+    },
+    {
+      "epoch": 0.257701932149157,
+      "grad_norm": 0.6605223023147049,
+      "learning_rate": 8.701882166867974e-06,
+      "loss": 0.3425,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25778397669934777,
+      "grad_norm": 0.970269352791597,
+      "learning_rate": 8.700988889926793e-06,
+      "loss": 0.3897,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2578660212495385,
+      "grad_norm": 0.7797093430693782,
+      "learning_rate": 8.700095351627725e-06,
+      "loss": 0.3687,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25794806579972923,
+      "grad_norm": 0.7057463372552539,
+      "learning_rate": 8.699201552033866e-06,
+      "loss": 0.4083,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25803011034992,
+      "grad_norm": 0.679278765238125,
+      "learning_rate": 8.698307491208338e-06,
+      "loss": 0.3924,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25811215490011075,
+      "grad_norm": 0.6240637632897383,
+      "learning_rate": 8.697413169214275e-06,
+      "loss": 0.3554,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25819419945030153,
+      "grad_norm": 0.6908611847725763,
+      "learning_rate": 8.696518586114836e-06,
+      "loss": 0.3619,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25827624400049226,
+      "grad_norm": 0.8104880066789462,
+      "learning_rate": 8.695623741973194e-06,
+      "loss": 0.378,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25835828855068305,
+      "grad_norm": 0.8373296263428138,
+      "learning_rate": 8.694728636852538e-06,
+      "loss": 0.3875,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2584403331008738,
+      "grad_norm": 0.6428865257449289,
+      "learning_rate": 8.693833270816083e-06,
+      "loss": 0.3816,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2585223776510645,
+      "grad_norm": 0.7097985361668929,
+      "learning_rate": 8.692937643927058e-06,
+      "loss": 0.3391,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2586044222012553,
+      "grad_norm": 0.9167344511255734,
+      "learning_rate": 8.69204175624871e-06,
+      "loss": 0.3716,
+      "step": 3152
+    },
+    {
+      "epoch": 0.258686466751446,
+      "grad_norm": 0.8911211227606003,
+      "learning_rate": 8.691145607844304e-06,
+      "loss": 0.4019,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2587685113016368,
+      "grad_norm": 0.6503747754520841,
+      "learning_rate": 8.690249198777126e-06,
+      "loss": 0.3571,
+      "step": 3154
+    },
+    {
+      "epoch": 0.25885055585182754,
+      "grad_norm": 0.7816476818370983,
+      "learning_rate": 8.689352529110478e-06,
+      "loss": 0.3868,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2589326004020183,
+      "grad_norm": 0.6697908427987581,
+      "learning_rate": 8.688455598907681e-06,
+      "loss": 0.3968,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25901464495220905,
+      "grad_norm": 0.7453234468912611,
+      "learning_rate": 8.687558408232076e-06,
+      "loss": 0.3452,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2590966895023998,
+      "grad_norm": 0.6713603638243494,
+      "learning_rate": 8.686660957147022e-06,
+      "loss": 0.3612,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25917873405259056,
+      "grad_norm": 0.7281823224805904,
+      "learning_rate": 8.685763245715892e-06,
+      "loss": 0.3878,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2592607786027813,
+      "grad_norm": 0.6185378397304611,
+      "learning_rate": 8.684865274002085e-06,
+      "loss": 0.3436,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2593428231529721,
+      "grad_norm": 0.6650556484057979,
+      "learning_rate": 8.683967042069013e-06,
+      "loss": 0.3665,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2594248677031628,
+      "grad_norm": 0.6584477396312579,
+      "learning_rate": 8.683068549980106e-06,
+      "loss": 0.3885,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2595069122533536,
+      "grad_norm": 0.7620441465304493,
+      "learning_rate": 8.682169797798815e-06,
+      "loss": 0.3702,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2595889568035443,
+      "grad_norm": 0.9322986109480751,
+      "learning_rate": 8.681270785588609e-06,
+      "loss": 0.3781,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25967100135373505,
+      "grad_norm": 0.6439292904015512,
+      "learning_rate": 8.680371513412975e-06,
+      "loss": 0.3609,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25975304590392584,
+      "grad_norm": 0.7562510900172674,
+      "learning_rate": 8.679471981335418e-06,
+      "loss": 0.3757,
+      "step": 3166
+    },
+    {
+      "epoch": 0.25983509045411657,
+      "grad_norm": 0.7864026859619301,
+      "learning_rate": 8.678572189419461e-06,
+      "loss": 0.3785,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25991713500430735,
+      "grad_norm": 0.6676929479268652,
+      "learning_rate": 8.677672137728645e-06,
+      "loss": 0.3741,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2599991795544981,
+      "grad_norm": 0.7591932369625426,
+      "learning_rate": 8.676771826326533e-06,
+      "loss": 0.3844,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26008122410468887,
+      "grad_norm": 0.676637261864924,
+      "learning_rate": 8.6758712552767e-06,
+      "loss": 0.3573,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2601632686548796,
+      "grad_norm": 0.6862974032385996,
+      "learning_rate": 8.674970424642745e-06,
+      "loss": 0.3446,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2602453132050703,
+      "grad_norm": 0.8217609971756128,
+      "learning_rate": 8.674069334488284e-06,
+      "loss": 0.3766,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2603273577552611,
+      "grad_norm": 0.872000909382775,
+      "learning_rate": 8.67316798487695e-06,
+      "loss": 0.4219,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26040940230545184,
+      "grad_norm": 0.6317096560431704,
+      "learning_rate": 8.672266375872392e-06,
+      "loss": 0.3462,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2604914468556426,
+      "grad_norm": 0.8354059107714799,
+      "learning_rate": 8.671364507538284e-06,
+      "loss": 0.4294,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26057349140583336,
+      "grad_norm": 0.6842325914013869,
+      "learning_rate": 8.670462379938313e-06,
+      "loss": 0.3531,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26065553595602414,
+      "grad_norm": 0.6745021993281622,
+      "learning_rate": 8.669559993136185e-06,
+      "loss": 0.3542,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26073758050621487,
+      "grad_norm": 0.6275608482114347,
+      "learning_rate": 8.668657347195627e-06,
+      "loss": 0.3627,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2608196250564056,
+      "grad_norm": 0.6408103104894711,
+      "learning_rate": 8.66775444218038e-06,
+      "loss": 0.3808,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2609016696065964,
+      "grad_norm": 0.6971596452561198,
+      "learning_rate": 8.666851278154208e-06,
+      "loss": 0.3655,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2609837141567871,
+      "grad_norm": 0.9113848394650199,
+      "learning_rate": 8.665947855180889e-06,
+      "loss": 0.3779,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2610657587069779,
+      "grad_norm": 0.9659987441869748,
+      "learning_rate": 8.665044173324225e-06,
+      "loss": 0.4087,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26114780325716863,
+      "grad_norm": 0.7461397821440396,
+      "learning_rate": 8.664140232648026e-06,
+      "loss": 0.4262,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2612298478073594,
+      "grad_norm": 0.6998101999176618,
+      "learning_rate": 8.663236033216133e-06,
+      "loss": 0.3317,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26131189235755015,
+      "grad_norm": 0.8824764177083688,
+      "learning_rate": 8.662331575092396e-06,
+      "loss": 0.3755,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2613939369077409,
+      "grad_norm": 0.6955518195546772,
+      "learning_rate": 8.661426858340687e-06,
+      "loss": 0.375,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26147598145793166,
+      "grad_norm": 0.6840782445293266,
+      "learning_rate": 8.660521883024895e-06,
+      "loss": 0.4199,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2615580260081224,
+      "grad_norm": 0.726703061227362,
+      "learning_rate": 8.65961664920893e-06,
+      "loss": 0.345,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2616400705583132,
+      "grad_norm": 0.6660433052332396,
+      "learning_rate": 8.658711156956719e-06,
+      "loss": 0.4116,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2617221151085039,
+      "grad_norm": 0.6362847054808488,
+      "learning_rate": 8.6578054063322e-06,
+      "loss": 0.3496,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2618041596586947,
+      "grad_norm": 0.5719256410813679,
+      "learning_rate": 8.656899397399343e-06,
+      "loss": 0.3423,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2618862042088854,
+      "grad_norm": 0.821187213726844,
+      "learning_rate": 8.655993130222124e-06,
+      "loss": 0.3846,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2619682487590762,
+      "grad_norm": 0.6994188008868596,
+      "learning_rate": 8.655086604864545e-06,
+      "loss": 0.3809,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26205029330926694,
+      "grad_norm": 0.6576833782345283,
+      "learning_rate": 8.65417982139062e-06,
+      "loss": 0.3407,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26213233785945766,
+      "grad_norm": 0.6064850972496457,
+      "learning_rate": 8.65327277986439e-06,
+      "loss": 0.3649,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26221438240964845,
+      "grad_norm": 0.7598725275017515,
+      "learning_rate": 8.652365480349904e-06,
+      "loss": 0.3849,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2622964269598392,
+      "grad_norm": 0.8441698055133459,
+      "learning_rate": 8.651457922911237e-06,
+      "loss": 0.4065,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26237847151002996,
+      "grad_norm": 0.8925089476833528,
+      "learning_rate": 8.650550107612476e-06,
+      "loss": 0.3854,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2624605160602207,
+      "grad_norm": 0.7144866912545919,
+      "learning_rate": 8.649642034517734e-06,
+      "loss": 0.375,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2625425606104115,
+      "grad_norm": 0.7182152168181775,
+      "learning_rate": 8.648733703691135e-06,
+      "loss": 0.3604,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2626246051606022,
+      "grad_norm": 0.930773357163344,
+      "learning_rate": 8.647825115196822e-06,
+      "loss": 0.3798,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26270664971079294,
+      "grad_norm": 0.6590250366615306,
+      "learning_rate": 8.646916269098961e-06,
+      "loss": 0.3639,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2627886942609837,
+      "grad_norm": 0.867241860071896,
+      "learning_rate": 8.646007165461732e-06,
+      "loss": 0.3522,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26287073881117445,
+      "grad_norm": 0.6315540145595606,
+      "learning_rate": 8.645097804349334e-06,
+      "loss": 0.3576,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26295278336136524,
+      "grad_norm": 0.9136374261081334,
+      "learning_rate": 8.644188185825986e-06,
+      "loss": 0.387,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26303482791155597,
+      "grad_norm": 0.663583004873792,
+      "learning_rate": 8.643278309955924e-06,
+      "loss": 0.3479,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26311687246174675,
+      "grad_norm": 0.6824957562695569,
+      "learning_rate": 8.642368176803399e-06,
+      "loss": 0.3226,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2631989170119375,
+      "grad_norm": 0.6437995995973135,
+      "learning_rate": 8.641457786432687e-06,
+      "loss": 0.3384,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2632809615621282,
+      "grad_norm": 0.970874036938425,
+      "learning_rate": 8.640547138908077e-06,
+      "loss": 0.3533,
+      "step": 3209
+    },
+    {
+      "epoch": 0.263363006112319,
+      "grad_norm": 0.7048179393632471,
+      "learning_rate": 8.639636234293878e-06,
+      "loss": 0.3936,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2634450506625097,
+      "grad_norm": 0.671542169402969,
+      "learning_rate": 8.638725072654413e-06,
+      "loss": 0.3713,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2635270952127005,
+      "grad_norm": 0.7769028436988911,
+      "learning_rate": 8.63781365405403e-06,
+      "loss": 0.3858,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26360913976289124,
+      "grad_norm": 0.6528740141544908,
+      "learning_rate": 8.636901978557092e-06,
+      "loss": 0.3863,
+      "step": 3213
+    },
+    {
+      "epoch": 0.263691184313082,
+      "grad_norm": 0.6207298013852083,
+      "learning_rate": 8.63599004622798e-06,
+      "loss": 0.3804,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26377322886327276,
+      "grad_norm": 0.6433253737367097,
+      "learning_rate": 8.635077857131091e-06,
+      "loss": 0.4064,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2638552734134635,
+      "grad_norm": 0.6828524581270953,
+      "learning_rate": 8.634165411330845e-06,
+      "loss": 0.3751,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26393731796365427,
+      "grad_norm": 0.6689959753054014,
+      "learning_rate": 8.633252708891677e-06,
+      "loss": 0.3538,
+      "step": 3217
+    },
+    {
+      "epoch": 0.264019362513845,
+      "grad_norm": 0.6667264992057949,
+      "learning_rate": 8.632339749878038e-06,
+      "loss": 0.3694,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2641014070640358,
+      "grad_norm": 0.6772850431914896,
+      "learning_rate": 8.631426534354404e-06,
+      "loss": 0.3445,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2641834516142265,
+      "grad_norm": 0.6227125483324175,
+      "learning_rate": 8.63051306238526e-06,
+      "loss": 0.3829,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2642654961644173,
+      "grad_norm": 0.7325950010051868,
+      "learning_rate": 8.629599334035119e-06,
+      "loss": 0.348,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26434754071460803,
+      "grad_norm": 0.6511875370197818,
+      "learning_rate": 8.628685349368502e-06,
+      "loss": 0.4011,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26442958526479876,
+      "grad_norm": 0.7485857068953476,
+      "learning_rate": 8.627771108449956e-06,
+      "loss": 0.4098,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26451162981498955,
+      "grad_norm": 0.6082163498104752,
+      "learning_rate": 8.626856611344043e-06,
+      "loss": 0.3611,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2645936743651803,
+      "grad_norm": 0.6395266340263815,
+      "learning_rate": 8.625941858115344e-06,
+      "loss": 0.4444,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26467571891537106,
+      "grad_norm": 0.6451303156266769,
+      "learning_rate": 8.625026848828456e-06,
+      "loss": 0.334,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2647577634655618,
+      "grad_norm": 0.6061972021851753,
+      "learning_rate": 8.624111583547997e-06,
+      "loss": 0.3552,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2648398080157526,
+      "grad_norm": 0.6181049341189951,
+      "learning_rate": 8.623196062338599e-06,
+      "loss": 0.3342,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2649218525659433,
+      "grad_norm": 0.6557250700634831,
+      "learning_rate": 8.622280285264917e-06,
+      "loss": 0.3994,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26500389711613404,
+      "grad_norm": 0.6832133785576164,
+      "learning_rate": 8.62136425239162e-06,
+      "loss": 0.365,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2650859416663248,
+      "grad_norm": 0.6636789581709824,
+      "learning_rate": 8.620447963783398e-06,
+      "loss": 0.357,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26516798621651555,
+      "grad_norm": 0.6344144547596535,
+      "learning_rate": 8.619531419504959e-06,
+      "loss": 0.3683,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26525003076670634,
+      "grad_norm": 0.6767139979645451,
+      "learning_rate": 8.618614619621024e-06,
+      "loss": 0.3758,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26533207531689706,
+      "grad_norm": 0.6772993783532404,
+      "learning_rate": 8.61769756419634e-06,
+      "loss": 0.3774,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26541411986708785,
+      "grad_norm": 0.6378409991335119,
+      "learning_rate": 8.616780253295666e-06,
+      "loss": 0.3194,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2654961644172786,
+      "grad_norm": 0.642644099401376,
+      "learning_rate": 8.61586268698378e-06,
+      "loss": 0.3561,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2655782089674693,
+      "grad_norm": 0.6939394635748598,
+      "learning_rate": 8.614944865325482e-06,
+      "loss": 0.3833,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2656602535176601,
+      "grad_norm": 0.6501407092802053,
+      "learning_rate": 8.614026788385586e-06,
+      "loss": 0.3602,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2657422980678508,
+      "grad_norm": 0.6402038742112107,
+      "learning_rate": 8.613108456228922e-06,
+      "loss": 0.359,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2658243426180416,
+      "grad_norm": 0.6639549790861202,
+      "learning_rate": 8.612189868920345e-06,
+      "loss": 0.3583,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26590638716823234,
+      "grad_norm": 0.6661981839701105,
+      "learning_rate": 8.611271026524724e-06,
+      "loss": 0.4388,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2659884317184231,
+      "grad_norm": 0.6761006437455896,
+      "learning_rate": 8.610351929106944e-06,
+      "loss": 0.3364,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26607047626861385,
+      "grad_norm": 0.689577757976695,
+      "learning_rate": 8.609432576731912e-06,
+      "loss": 0.3576,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2661525208188046,
+      "grad_norm": 0.7360804602585995,
+      "learning_rate": 8.608512969464548e-06,
+      "loss": 0.3409,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26623456536899537,
+      "grad_norm": 0.6630668349970025,
+      "learning_rate": 8.607593107369798e-06,
+      "loss": 0.3601,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2663166099191861,
+      "grad_norm": 0.7216620656135864,
+      "learning_rate": 8.606672990512619e-06,
+      "loss": 0.367,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2663986544693769,
+      "grad_norm": 0.6438797040617212,
+      "learning_rate": 8.605752618957986e-06,
+      "loss": 0.4011,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2664806990195676,
+      "grad_norm": 0.6051501087297476,
+      "learning_rate": 8.604831992770898e-06,
+      "loss": 0.3679,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2665627435697584,
+      "grad_norm": 0.6236186445758509,
+      "learning_rate": 8.603911112016366e-06,
+      "loss": 0.356,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2666447881199491,
+      "grad_norm": 0.6243825523224746,
+      "learning_rate": 8.60298997675942e-06,
+      "loss": 0.3765,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2667268326701399,
+      "grad_norm": 0.6747955268758146,
+      "learning_rate": 8.602068587065111e-06,
+      "loss": 0.3342,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26680887722033064,
+      "grad_norm": 0.6334535669023696,
+      "learning_rate": 8.601146942998506e-06,
+      "loss": 0.3899,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26689092177052137,
+      "grad_norm": 0.6158139843527525,
+      "learning_rate": 8.60022504462469e-06,
+      "loss": 0.3998,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26697296632071216,
+      "grad_norm": 0.6477846139878669,
+      "learning_rate": 8.599302892008765e-06,
+      "loss": 0.3967,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2670550108709029,
+      "grad_norm": 0.6559324470935723,
+      "learning_rate": 8.598380485215854e-06,
+      "loss": 0.3745,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26713705542109367,
+      "grad_norm": 0.6997829725696408,
+      "learning_rate": 8.597457824311093e-06,
+      "loss": 0.4038,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2672190999712844,
+      "grad_norm": 0.5901643689256617,
+      "learning_rate": 8.59653490935964e-06,
+      "loss": 0.3629,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2673011445214752,
+      "grad_norm": 0.6495234022896821,
+      "learning_rate": 8.59561174042667e-06,
+      "loss": 0.3813,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2673831890716659,
+      "grad_norm": 0.5914827102542379,
+      "learning_rate": 8.594688317577375e-06,
+      "loss": 0.3316,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26746523362185665,
+      "grad_norm": 0.6962460000048584,
+      "learning_rate": 8.593764640876967e-06,
+      "loss": 0.4206,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26754727817204743,
+      "grad_norm": 0.7182113195914129,
+      "learning_rate": 8.592840710390674e-06,
+      "loss": 0.3868,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26762932272223816,
+      "grad_norm": 0.6832580505229849,
+      "learning_rate": 8.591916526183743e-06,
+      "loss": 0.384,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26771136727242895,
+      "grad_norm": 0.7030113162990322,
+      "learning_rate": 8.590992088321438e-06,
+      "loss": 0.3886,
+      "step": 3263
+    },
+    {
+      "epoch": 0.2677934118226197,
+      "grad_norm": 0.6731790011494098,
+      "learning_rate": 8.59006739686904e-06,
+      "loss": 0.3665,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26787545637281046,
+      "grad_norm": 0.7641656423183154,
+      "learning_rate": 8.589142451891849e-06,
+      "loss": 0.3397,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2679575009230012,
+      "grad_norm": 0.658157963939966,
+      "learning_rate": 8.588217253455187e-06,
+      "loss": 0.4146,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2680395454731919,
+      "grad_norm": 0.6907026811123446,
+      "learning_rate": 8.587291801624387e-06,
+      "loss": 0.3569,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2681215900233827,
+      "grad_norm": 0.6059179382545469,
+      "learning_rate": 8.586366096464803e-06,
+      "loss": 0.3522,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26820363457357344,
+      "grad_norm": 0.646071605463082,
+      "learning_rate": 8.585440138041807e-06,
+      "loss": 0.3944,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2682856791237642,
+      "grad_norm": 0.6028200119786715,
+      "learning_rate": 8.58451392642079e-06,
+      "loss": 0.4019,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26836772367395495,
+      "grad_norm": 0.6630482470963482,
+      "learning_rate": 8.583587461667156e-06,
+      "loss": 0.3661,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26844976822414574,
+      "grad_norm": 0.6248627487806411,
+      "learning_rate": 8.582660743846335e-06,
+      "loss": 0.3999,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26853181277433646,
+      "grad_norm": 0.5994540075743592,
+      "learning_rate": 8.581733773023769e-06,
+      "loss": 0.3748,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2686138573245272,
+      "grad_norm": 0.6203738218863328,
+      "learning_rate": 8.580806549264915e-06,
+      "loss": 0.3655,
+      "step": 3274
+    },
+    {
+      "epoch": 0.268695901874718,
+      "grad_norm": 0.6724758946829833,
+      "learning_rate": 8.579879072635257e-06,
+      "loss": 0.3964,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2687779464249087,
+      "grad_norm": 0.5905579702210505,
+      "learning_rate": 8.578951343200293e-06,
+      "loss": 0.3463,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2688599909750995,
+      "grad_norm": 0.591717821279884,
+      "learning_rate": 8.578023361025533e-06,
+      "loss": 0.3459,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2689420355252902,
+      "grad_norm": 0.6325227740071763,
+      "learning_rate": 8.57709512617651e-06,
+      "loss": 0.3399,
+      "step": 3278
+    },
+    {
+      "epoch": 0.269024080075481,
+      "grad_norm": 0.6526223794903492,
+      "learning_rate": 8.57616663871878e-06,
+      "loss": 0.3703,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26910612462567174,
+      "grad_norm": 1.3812211300760728,
+      "learning_rate": 8.575237898717905e-06,
+      "loss": 0.3836,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26918816917586247,
+      "grad_norm": 0.6976767785011899,
+      "learning_rate": 8.574308906239475e-06,
+      "loss": 0.4266,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26927021372605325,
+      "grad_norm": 0.6271205132920827,
+      "learning_rate": 8.57337966134909e-06,
+      "loss": 0.3767,
+      "step": 3282
+    },
+    {
+      "epoch": 0.269352258276244,
+      "grad_norm": 0.6611365503519531,
+      "learning_rate": 8.572450164112377e-06,
+      "loss": 0.3802,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26943430282643477,
+      "grad_norm": 0.6627138270800723,
+      "learning_rate": 8.571520414594974e-06,
+      "loss": 0.3906,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2695163473766255,
+      "grad_norm": 0.6145021144521077,
+      "learning_rate": 8.570590412862535e-06,
+      "loss": 0.3566,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2695983919268163,
+      "grad_norm": 0.7041261819722647,
+      "learning_rate": 8.56966015898074e-06,
+      "loss": 0.3698,
+      "step": 3286
+    },
+    {
+      "epoch": 0.269680436477007,
+      "grad_norm": 0.693337245368687,
+      "learning_rate": 8.568729653015278e-06,
+      "loss": 0.4043,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26976248102719774,
+      "grad_norm": 0.6878196959572115,
+      "learning_rate": 8.56779889503186e-06,
+      "loss": 0.4143,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2698445255773885,
+      "grad_norm": 0.6051001866848589,
+      "learning_rate": 8.566867885096217e-06,
+      "loss": 0.3846,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26992657012757926,
+      "grad_norm": 0.678229354167882,
+      "learning_rate": 8.565936623274096e-06,
+      "loss": 0.3513,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27000861467777004,
+      "grad_norm": 0.6759791011212622,
+      "learning_rate": 8.56500510963126e-06,
+      "loss": 0.3742,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27009065922796077,
+      "grad_norm": 0.6370254385721991,
+      "learning_rate": 8.56407334423349e-06,
+      "loss": 0.363,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27017270377815156,
+      "grad_norm": 0.6251773747632747,
+      "learning_rate": 8.563141327146586e-06,
+      "loss": 0.3487,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2702547483283423,
+      "grad_norm": 0.6429476500384653,
+      "learning_rate": 8.562209058436366e-06,
+      "loss": 0.4081,
+      "step": 3294
+    },
+    {
+      "epoch": 0.270336792878533,
+      "grad_norm": 0.6321418894108022,
+      "learning_rate": 8.561276538168666e-06,
+      "loss": 0.3806,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2704188374287238,
+      "grad_norm": 0.6085740262330057,
+      "learning_rate": 8.560343766409339e-06,
+      "loss": 0.3779,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27050088197891453,
+      "grad_norm": 0.6387869076648358,
+      "learning_rate": 8.559410743224254e-06,
+      "loss": 0.3494,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2705829265291053,
+      "grad_norm": 0.6360997949940853,
+      "learning_rate": 8.558477468679303e-06,
+      "loss": 0.3409,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27066497107929605,
+      "grad_norm": 0.6635218062822279,
+      "learning_rate": 8.557543942840386e-06,
+      "loss": 0.3864,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27074701562948683,
+      "grad_norm": 0.6510219920529147,
+      "learning_rate": 8.556610165773436e-06,
+      "loss": 0.3743,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27082906017967756,
+      "grad_norm": 0.6299717362146108,
+      "learning_rate": 8.55567613754439e-06,
+      "loss": 0.3394,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27091110472986835,
+      "grad_norm": 0.6361141886600025,
+      "learning_rate": 8.554741858219205e-06,
+      "loss": 0.4078,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2709931492800591,
+      "grad_norm": 0.798120255751697,
+      "learning_rate": 8.553807327863864e-06,
+      "loss": 0.3635,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2710751938302498,
+      "grad_norm": 0.6722980772604644,
+      "learning_rate": 8.552872546544356e-06,
+      "loss": 0.3276,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2711572383804406,
+      "grad_norm": 0.6831843945940786,
+      "learning_rate": 8.5519375143267e-06,
+      "loss": 0.3752,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2712392829306313,
+      "grad_norm": 0.6899635942031966,
+      "learning_rate": 8.55100223127692e-06,
+      "loss": 0.3834,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2713213274808221,
+      "grad_norm": 0.6365657990269605,
+      "learning_rate": 8.55006669746107e-06,
+      "loss": 0.3726,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27140337203101284,
+      "grad_norm": 0.6530643775922542,
+      "learning_rate": 8.549130912945214e-06,
+      "loss": 0.3791,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2714854165812036,
+      "grad_norm": 0.6860332496796889,
+      "learning_rate": 8.548194877795433e-06,
+      "loss": 0.3789,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27156746113139435,
+      "grad_norm": 0.6187877077363435,
+      "learning_rate": 8.547258592077831e-06,
+      "loss": 0.3747,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2716495056815851,
+      "grad_norm": 0.6512878945588138,
+      "learning_rate": 8.546322055858526e-06,
+      "loss": 0.4424,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27173155023177586,
+      "grad_norm": 0.6325318838917804,
+      "learning_rate": 8.545385269203656e-06,
+      "loss": 0.3891,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2718135947819666,
+      "grad_norm": 0.6080239914391311,
+      "learning_rate": 8.544448232179375e-06,
+      "loss": 0.3843,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2718956393321574,
+      "grad_norm": 0.6358470331793431,
+      "learning_rate": 8.543510944851852e-06,
+      "loss": 0.3867,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2719776838823481,
+      "grad_norm": 0.7294541476765931,
+      "learning_rate": 8.542573407287283e-06,
+      "loss": 0.3879,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2720597284325389,
+      "grad_norm": 0.6377637961037489,
+      "learning_rate": 8.541635619551868e-06,
+      "loss": 0.3374,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2721417729827296,
+      "grad_norm": 0.7269738113743257,
+      "learning_rate": 8.540697581711836e-06,
+      "loss": 0.3872,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27222381753292035,
+      "grad_norm": 0.6948339737960392,
+      "learning_rate": 8.539759293833431e-06,
+      "loss": 0.3528,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27230586208311114,
+      "grad_norm": 0.6531878288613396,
+      "learning_rate": 8.538820755982911e-06,
+      "loss": 0.3908,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27238790663330187,
+      "grad_norm": 0.5979804721472852,
+      "learning_rate": 8.537881968226553e-06,
+      "loss": 0.344,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27246995118349265,
+      "grad_norm": 0.7834091711478649,
+      "learning_rate": 8.536942930630657e-06,
+      "loss": 0.6843,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2725519957336834,
+      "grad_norm": 0.6236577235821512,
+      "learning_rate": 8.536003643261532e-06,
+      "loss": 0.3417,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27263404028387417,
+      "grad_norm": 0.6656137243275213,
+      "learning_rate": 8.535064106185513e-06,
+      "loss": 0.4064,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2727160848340649,
+      "grad_norm": 0.6377039487094848,
+      "learning_rate": 8.534124319468945e-06,
+      "loss": 0.3657,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2727981293842556,
+      "grad_norm": 0.6537862992972946,
+      "learning_rate": 8.533184283178196e-06,
+      "loss": 0.3684,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2728801739344464,
+      "grad_norm": 0.5822828533886745,
+      "learning_rate": 8.532243997379647e-06,
+      "loss": 0.3532,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27296221848463714,
+      "grad_norm": 0.7044216761035875,
+      "learning_rate": 8.531303462139705e-06,
+      "loss": 0.3785,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2730442630348279,
+      "grad_norm": 0.6858242909864383,
+      "learning_rate": 8.530362677524783e-06,
+      "loss": 0.3549,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27312630758501866,
+      "grad_norm": 0.6618576640803222,
+      "learning_rate": 8.529421643601323e-06,
+      "loss": 0.4035,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27320835213520944,
+      "grad_norm": 0.6205153655815526,
+      "learning_rate": 8.528480360435777e-06,
+      "loss": 0.3526,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27329039668540017,
+      "grad_norm": 0.6376699275993787,
+      "learning_rate": 8.527538828094617e-06,
+      "loss": 0.3309,
+      "step": 3331
+    },
+    {
+      "epoch": 0.2733724412355909,
+      "grad_norm": 0.6912427793885435,
+      "learning_rate": 8.526597046644332e-06,
+      "loss": 0.4104,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2734544857857817,
+      "grad_norm": 0.6558498897207347,
+      "learning_rate": 8.525655016151428e-06,
+      "loss": 0.3731,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2735365303359724,
+      "grad_norm": 0.6215413890339537,
+      "learning_rate": 8.524712736682433e-06,
+      "loss": 0.3609,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2736185748861632,
+      "grad_norm": 0.6374270102442289,
+      "learning_rate": 8.523770208303885e-06,
+      "loss": 0.3979,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27370061943635393,
+      "grad_norm": 0.7079729830548607,
+      "learning_rate": 8.52282743108235e-06,
+      "loss": 0.3652,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2737826639865447,
+      "grad_norm": 0.6275660259556283,
+      "learning_rate": 8.521884405084398e-06,
+      "loss": 0.3391,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27386470853673545,
+      "grad_norm": 0.8776549787787862,
+      "learning_rate": 8.52094113037663e-06,
+      "loss": 0.3739,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2739467530869262,
+      "grad_norm": 0.6327790724193015,
+      "learning_rate": 8.519997607025657e-06,
+      "loss": 0.3382,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27402879763711696,
+      "grad_norm": 0.669778428687306,
+      "learning_rate": 8.519053835098105e-06,
+      "loss": 0.382,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2741108421873077,
+      "grad_norm": 0.6975543529543528,
+      "learning_rate": 8.518109814660627e-06,
+      "loss": 0.3626,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2741928867374985,
+      "grad_norm": 0.6836501746316828,
+      "learning_rate": 8.517165545779885e-06,
+      "loss": 0.3522,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2742749312876892,
+      "grad_norm": 0.7087949164875523,
+      "learning_rate": 8.516221028522565e-06,
+      "loss": 0.3857,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27435697583788,
+      "grad_norm": 0.6347795408501671,
+      "learning_rate": 8.515276262955363e-06,
+      "loss": 0.3783,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2744390203880707,
+      "grad_norm": 0.7653179499710488,
+      "learning_rate": 8.514331249145001e-06,
+      "loss": 0.3607,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27452106493826145,
+      "grad_norm": 0.6663665370776249,
+      "learning_rate": 8.513385987158212e-06,
+      "loss": 0.3674,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27460310948845223,
+      "grad_norm": 0.6473514611552825,
+      "learning_rate": 8.512440477061747e-06,
+      "loss": 0.3249,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27468515403864296,
+      "grad_norm": 0.68759291184105,
+      "learning_rate": 8.511494718922382e-06,
+      "loss": 0.347,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27476719858883375,
+      "grad_norm": 0.6310365479977752,
+      "learning_rate": 8.5105487128069e-06,
+      "loss": 0.3803,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2748492431390245,
+      "grad_norm": 0.6294616534516116,
+      "learning_rate": 8.509602458782107e-06,
+      "loss": 0.354,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27493128768921526,
+      "grad_norm": 0.6462303507706147,
+      "learning_rate": 8.508655956914829e-06,
+      "loss": 0.3849,
+      "step": 3351
+    },
+    {
+      "epoch": 0.275013332239406,
+      "grad_norm": 0.6311783833147525,
+      "learning_rate": 8.507709207271903e-06,
+      "loss": 0.377,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2750953767895967,
+      "grad_norm": 0.7053697975881968,
+      "learning_rate": 8.506762209920188e-06,
+      "loss": 0.386,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2751774213397875,
+      "grad_norm": 0.6493725175115188,
+      "learning_rate": 8.50581496492656e-06,
+      "loss": 0.3424,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27525946588997824,
+      "grad_norm": 0.6316487499796051,
+      "learning_rate": 8.504867472357913e-06,
+      "loss": 0.3618,
+      "step": 3355
+    },
+    {
+      "epoch": 0.275341510440169,
+      "grad_norm": 0.6118402492890171,
+      "learning_rate": 8.503919732281156e-06,
+      "loss": 0.3457,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27542355499035975,
+      "grad_norm": 0.6609012135000731,
+      "learning_rate": 8.502971744763216e-06,
+      "loss": 0.3507,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27550559954055054,
+      "grad_norm": 0.8407371670649131,
+      "learning_rate": 8.502023509871041e-06,
+      "loss": 0.3977,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27558764409074127,
+      "grad_norm": 0.623832364964405,
+      "learning_rate": 8.50107502767159e-06,
+      "loss": 0.3604,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27566968864093205,
+      "grad_norm": 0.6515255850079706,
+      "learning_rate": 8.500126298231845e-06,
+      "loss": 0.3458,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2757517331911228,
+      "grad_norm": 0.7450669627688858,
+      "learning_rate": 8.499177321618805e-06,
+      "loss": 0.3778,
+      "step": 3361
+    },
+    {
+      "epoch": 0.2758337777413135,
+      "grad_norm": 0.6729363127328544,
+      "learning_rate": 8.498228097899486e-06,
+      "loss": 0.3769,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2759158222915043,
+      "grad_norm": 1.1330233747119909,
+      "learning_rate": 8.497278627140916e-06,
+      "loss": 0.36,
+      "step": 3363
+    },
+    {
+      "epoch": 0.275997866841695,
+      "grad_norm": 0.6319496811155692,
+      "learning_rate": 8.496328909410151e-06,
+      "loss": 0.3552,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2760799113918858,
+      "grad_norm": 1.0820283663958843,
+      "learning_rate": 8.495378944774256e-06,
+      "loss": 0.3748,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27616195594207654,
+      "grad_norm": 0.5875028667287997,
+      "learning_rate": 8.494428733300313e-06,
+      "loss": 0.3565,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2762440004922673,
+      "grad_norm": 0.6618674861782741,
+      "learning_rate": 8.49347827505543e-06,
+      "loss": 0.3768,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27632604504245806,
+      "grad_norm": 0.6847948746588862,
+      "learning_rate": 8.49252757010672e-06,
+      "loss": 0.372,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2764080895926488,
+      "grad_norm": 0.8060639400014415,
+      "learning_rate": 8.491576618521328e-06,
+      "loss": 0.3473,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27649013414283957,
+      "grad_norm": 0.658399531407268,
+      "learning_rate": 8.490625420366405e-06,
+      "loss": 0.3548,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2765721786930303,
+      "grad_norm": 0.6051827670882558,
+      "learning_rate": 8.489673975709121e-06,
+      "loss": 0.3689,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2766542232432211,
+      "grad_norm": 0.7296156157263644,
+      "learning_rate": 8.488722284616668e-06,
+      "loss": 0.3731,
+      "step": 3372
+    },
+    {
+      "epoch": 0.2767362677934118,
+      "grad_norm": 0.577687537422088,
+      "learning_rate": 8.487770347156252e-06,
+      "loss": 0.3621,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2768183123436026,
+      "grad_norm": 0.6496900867000808,
+      "learning_rate": 8.486818163395097e-06,
+      "loss": 0.3528,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27690035689379333,
+      "grad_norm": 0.6969801250723261,
+      "learning_rate": 8.485865733400447e-06,
+      "loss": 0.3201,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27698240144398406,
+      "grad_norm": 0.6474267464803716,
+      "learning_rate": 8.484913057239557e-06,
+      "loss": 0.3469,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27706444599417485,
+      "grad_norm": 0.7517927009950771,
+      "learning_rate": 8.483960134979707e-06,
+      "loss": 0.4058,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2771464905443656,
+      "grad_norm": 0.6678379043823098,
+      "learning_rate": 8.48300696668819e-06,
+      "loss": 0.3767,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27722853509455636,
+      "grad_norm": 0.7054996287842821,
+      "learning_rate": 8.482053552432318e-06,
+      "loss": 0.3778,
+      "step": 3379
+    },
+    {
+      "epoch": 0.2773105796447471,
+      "grad_norm": 0.6708040946719929,
+      "learning_rate": 8.481099892279418e-06,
+      "loss": 0.378,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2773926241949379,
+      "grad_norm": 0.6506917862917554,
+      "learning_rate": 8.480145986296834e-06,
+      "loss": 0.3774,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2774746687451286,
+      "grad_norm": 0.6295702248851965,
+      "learning_rate": 8.479191834551934e-06,
+      "loss": 0.355,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27755671329531933,
+      "grad_norm": 0.6238009396916685,
+      "learning_rate": 8.478237437112095e-06,
+      "loss": 0.3666,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2776387578455101,
+      "grad_norm": 0.6464341248163621,
+      "learning_rate": 8.47728279404472e-06,
+      "loss": 0.3632,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27772080239570085,
+      "grad_norm": 0.5885182905808428,
+      "learning_rate": 8.476327905417217e-06,
+      "loss": 0.3501,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27780284694589163,
+      "grad_norm": 0.6623730728426084,
+      "learning_rate": 8.475372771297023e-06,
+      "loss": 0.3661,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27788489149608236,
+      "grad_norm": 0.6765234131901048,
+      "learning_rate": 8.474417391751587e-06,
+      "loss": 0.3962,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27796693604627315,
+      "grad_norm": 0.6559950876048656,
+      "learning_rate": 8.473461766848378e-06,
+      "loss": 0.4239,
+      "step": 3388
+    },
+    {
+      "epoch": 0.2780489805964639,
+      "grad_norm": 0.6153620751399828,
+      "learning_rate": 8.472505896654881e-06,
+      "loss": 0.4008,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2781310251466546,
+      "grad_norm": 0.5955904603928537,
+      "learning_rate": 8.471549781238595e-06,
+      "loss": 0.3645,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2782130696968454,
+      "grad_norm": 0.76005689601423,
+      "learning_rate": 8.47059342066704e-06,
+      "loss": 0.3304,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2782951142470361,
+      "grad_norm": 0.673465866803607,
+      "learning_rate": 8.469636815007756e-06,
+      "loss": 0.3218,
+      "step": 3392
+    },
+    {
+      "epoch": 0.2783771587972269,
+      "grad_norm": 0.6123056958932955,
+      "learning_rate": 8.468679964328293e-06,
+      "loss": 0.3468,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27845920334741764,
+      "grad_norm": 0.7724704085941378,
+      "learning_rate": 8.467722868696222e-06,
+      "loss": 0.3887,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2785412478976084,
+      "grad_norm": 1.403768095558883,
+      "learning_rate": 8.466765528179135e-06,
+      "loss": 0.3514,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27862329244779915,
+      "grad_norm": 0.6627194637876966,
+      "learning_rate": 8.465807942844637e-06,
+      "loss": 0.3526,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2787053369979899,
+      "grad_norm": 0.6272940627822977,
+      "learning_rate": 8.46485011276035e-06,
+      "loss": 0.3354,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27878738154818067,
+      "grad_norm": 0.7228014886934863,
+      "learning_rate": 8.463892037993915e-06,
+      "loss": 0.3945,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2788694260983714,
+      "grad_norm": 0.6409645453786755,
+      "learning_rate": 8.462933718612988e-06,
+      "loss": 0.3498,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2789514706485622,
+      "grad_norm": 0.6669332810482087,
+      "learning_rate": 8.461975154685246e-06,
+      "loss": 0.3393,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2790335151987529,
+      "grad_norm": 0.6532874670131701,
+      "learning_rate": 8.461016346278381e-06,
+      "loss": 0.3655,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2791155597489437,
+      "grad_norm": 0.6270392277342277,
+      "learning_rate": 8.460057293460102e-06,
+      "loss": 0.3775,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2791976042991344,
+      "grad_norm": 0.7323421314357743,
+      "learning_rate": 8.459097996298137e-06,
+      "loss": 0.3888,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27927964884932516,
+      "grad_norm": 0.6620425194669889,
+      "learning_rate": 8.458138454860227e-06,
+      "loss": 0.3855,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27936169339951594,
+      "grad_norm": 0.6279296975635633,
+      "learning_rate": 8.457178669214137e-06,
+      "loss": 0.3656,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27944373794970667,
+      "grad_norm": 0.63499609208425,
+      "learning_rate": 8.456218639427643e-06,
+      "loss": 0.3788,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27952578249989746,
+      "grad_norm": 0.6449587567636315,
+      "learning_rate": 8.455258365568541e-06,
+      "loss": 0.3912,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2796078270500882,
+      "grad_norm": 0.6490175705752423,
+      "learning_rate": 8.454297847704644e-06,
+      "loss": 0.363,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27968987160027897,
+      "grad_norm": 0.6842078052063821,
+      "learning_rate": 8.453337085903785e-06,
+      "loss": 0.3572,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2797719161504697,
+      "grad_norm": 0.6101541451739401,
+      "learning_rate": 8.452376080233808e-06,
+      "loss": 0.3358,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27985396070066043,
+      "grad_norm": 0.6370458046077488,
+      "learning_rate": 8.45141483076258e-06,
+      "loss": 0.3491,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2799360052508512,
+      "grad_norm": 0.6358033865353049,
+      "learning_rate": 8.45045333755798e-06,
+      "loss": 0.3679,
+      "step": 3412
+    },
+    {
+      "epoch": 0.28001804980104195,
+      "grad_norm": 0.769570310923854,
+      "learning_rate": 8.449491600687909e-06,
+      "loss": 0.3796,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28010009435123273,
+      "grad_norm": 0.6325342257872946,
+      "learning_rate": 8.448529620220284e-06,
+      "loss": 0.3296,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28018213890142346,
+      "grad_norm": 0.6321208902726518,
+      "learning_rate": 8.447567396223037e-06,
+      "loss": 0.3878,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28026418345161425,
+      "grad_norm": 0.6896559586463067,
+      "learning_rate": 8.446604928764121e-06,
+      "loss": 0.4271,
+      "step": 3416
+    },
+    {
+      "epoch": 0.280346228001805,
+      "grad_norm": 0.7026683021735592,
+      "learning_rate": 8.4456422179115e-06,
+      "loss": 0.3719,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28042827255199576,
+      "grad_norm": 0.6830953539661754,
+      "learning_rate": 8.444679263733163e-06,
+      "loss": 0.3801,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2805103171021865,
+      "grad_norm": 0.6450555241584502,
+      "learning_rate": 8.44371606629711e-06,
+      "loss": 0.3532,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2805923616523772,
+      "grad_norm": 0.6298693397151094,
+      "learning_rate": 8.442752625671363e-06,
+      "loss": 0.3962,
+      "step": 3420
+    },
+    {
+      "epoch": 0.280674406202568,
+      "grad_norm": 0.6614103169215835,
+      "learning_rate": 8.441788941923954e-06,
+      "loss": 0.3671,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28075645075275873,
+      "grad_norm": 0.6970377061730759,
+      "learning_rate": 8.44082501512294e-06,
+      "loss": 0.3528,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2808384953029495,
+      "grad_norm": 0.693982340722826,
+      "learning_rate": 8.439860845336394e-06,
+      "loss": 0.3758,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28092053985314025,
+      "grad_norm": 0.626164248601781,
+      "learning_rate": 8.4388964326324e-06,
+      "loss": 0.3558,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28100258440333103,
+      "grad_norm": 0.6609691032453204,
+      "learning_rate": 8.437931777079065e-06,
+      "loss": 0.3773,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28108462895352176,
+      "grad_norm": 0.5921268813373416,
+      "learning_rate": 8.43696687874451e-06,
+      "loss": 0.3359,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2811666735037125,
+      "grad_norm": 0.6835237060351199,
+      "learning_rate": 8.436001737696877e-06,
+      "loss": 0.4001,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2812487180539033,
+      "grad_norm": 0.6879833360292036,
+      "learning_rate": 8.435036354004322e-06,
+      "loss": 0.3606,
+      "step": 3428
+    },
+    {
+      "epoch": 0.281330762604094,
+      "grad_norm": 0.6110808685597865,
+      "learning_rate": 8.43407072773502e-06,
+      "loss": 0.363,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2814128071542848,
+      "grad_norm": 0.6871048456141328,
+      "learning_rate": 8.433104858957157e-06,
+      "loss": 0.399,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2814948517044755,
+      "grad_norm": 0.6593151752335106,
+      "learning_rate": 8.432138747738947e-06,
+      "loss": 0.3722,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2815768962546663,
+      "grad_norm": 0.7421636622724503,
+      "learning_rate": 8.431172394148613e-06,
+      "loss": 0.4073,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28165894080485704,
+      "grad_norm": 0.6100168483276671,
+      "learning_rate": 8.430205798254396e-06,
+      "loss": 0.3799,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28174098535504777,
+      "grad_norm": 0.6465055845783861,
+      "learning_rate": 8.429238960124557e-06,
+      "loss": 0.3709,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28182302990523855,
+      "grad_norm": 0.6128445637201315,
+      "learning_rate": 8.42827187982737e-06,
+      "loss": 0.3363,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2819050744554293,
+      "grad_norm": 0.7204802265835862,
+      "learning_rate": 8.427304557431135e-06,
+      "loss": 0.3675,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28198711900562007,
+      "grad_norm": 0.6114231300497421,
+      "learning_rate": 8.426336993004155e-06,
+      "loss": 0.3406,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2820691635558108,
+      "grad_norm": 0.6423786510935567,
+      "learning_rate": 8.425369186614763e-06,
+      "loss": 0.3973,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2821512081060016,
+      "grad_norm": 0.6621577404937086,
+      "learning_rate": 8.424401138331302e-06,
+      "loss": 0.3714,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2822332526561923,
+      "grad_norm": 0.6844631466382674,
+      "learning_rate": 8.423432848222135e-06,
+      "loss": 0.4082,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28231529720638304,
+      "grad_norm": 0.5845605096291969,
+      "learning_rate": 8.422464316355638e-06,
+      "loss": 0.3556,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2823973417565738,
+      "grad_norm": 0.5664860811533896,
+      "learning_rate": 8.421495542800214e-06,
+      "loss": 0.3147,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28247938630676456,
+      "grad_norm": 0.6937066025830424,
+      "learning_rate": 8.42052652762427e-06,
+      "loss": 0.3905,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28256143085695534,
+      "grad_norm": 0.7190235995480226,
+      "learning_rate": 8.419557270896235e-06,
+      "loss": 0.3685,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28264347540714607,
+      "grad_norm": 0.7161857989304208,
+      "learning_rate": 8.418587772684561e-06,
+      "loss": 0.3657,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28272551995733686,
+      "grad_norm": 0.628734744876603,
+      "learning_rate": 8.417618033057714e-06,
+      "loss": 0.3379,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2828075645075276,
+      "grad_norm": 0.6664825407616755,
+      "learning_rate": 8.416648052084169e-06,
+      "loss": 0.3773,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2828896090577183,
+      "grad_norm": 0.6903571784101009,
+      "learning_rate": 8.415677829832429e-06,
+      "loss": 0.3801,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2829716536079091,
+      "grad_norm": 0.6567758677848439,
+      "learning_rate": 8.414707366371006e-06,
+      "loss": 0.3861,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28305369815809983,
+      "grad_norm": 0.6230621965021098,
+      "learning_rate": 8.41373666176844e-06,
+      "loss": 0.374,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2831357427082906,
+      "grad_norm": 0.6431197041430349,
+      "learning_rate": 8.412765716093273e-06,
+      "loss": 0.355,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28321778725848135,
+      "grad_norm": 0.6381677380991532,
+      "learning_rate": 8.411794529414073e-06,
+      "loss": 0.3709,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28329983180867213,
+      "grad_norm": 0.6641998288457245,
+      "learning_rate": 8.410823101799425e-06,
+      "loss": 0.3531,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28338187635886286,
+      "grad_norm": 0.634658670723116,
+      "learning_rate": 8.409851433317929e-06,
+      "loss": 0.4028,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2834639209090536,
+      "grad_norm": 0.6645372835341995,
+      "learning_rate": 8.408879524038204e-06,
+      "loss": 0.3618,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2835459654592444,
+      "grad_norm": 0.6267727785310939,
+      "learning_rate": 8.407907374028886e-06,
+      "loss": 0.3963,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2836280100094351,
+      "grad_norm": 0.6618094087849503,
+      "learning_rate": 8.406934983358621e-06,
+      "loss": 0.3541,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2837100545596259,
+      "grad_norm": 0.6779732944354131,
+      "learning_rate": 8.405962352096082e-06,
+      "loss": 0.36,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2837920991098166,
+      "grad_norm": 0.6264761716738891,
+      "learning_rate": 8.404989480309955e-06,
+      "loss": 0.3634,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2838741436600074,
+      "grad_norm": 0.6722487147354522,
+      "learning_rate": 8.404016368068941e-06,
+      "loss": 0.4085,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28395618821019813,
+      "grad_norm": 0.7238723605679896,
+      "learning_rate": 8.40304301544176e-06,
+      "loss": 0.3647,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28403823276038886,
+      "grad_norm": 0.6691336897521227,
+      "learning_rate": 8.402069422497148e-06,
+      "loss": 0.3633,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28412027731057965,
+      "grad_norm": 0.7057453073392254,
+      "learning_rate": 8.40109558930386e-06,
+      "loss": 0.3413,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2842023218607704,
+      "grad_norm": 0.735376815807539,
+      "learning_rate": 8.400121515930665e-06,
+      "loss": 0.3736,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28428436641096116,
+      "grad_norm": 0.6801466633729751,
+      "learning_rate": 8.399147202446352e-06,
+      "loss": 0.3825,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2843664109611519,
+      "grad_norm": 0.6192933094015607,
+      "learning_rate": 8.398172648919724e-06,
+      "loss": 0.3204,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2844484555113427,
+      "grad_norm": 0.6991923620971567,
+      "learning_rate": 8.397197855419603e-06,
+      "loss": 0.374,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2845305000615334,
+      "grad_norm": 0.670913021001082,
+      "learning_rate": 8.39622282201483e-06,
+      "loss": 0.3732,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28461254461172414,
+      "grad_norm": 0.6376052709364729,
+      "learning_rate": 8.395247548774255e-06,
+      "loss": 0.3893,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2846945891619149,
+      "grad_norm": 0.7322522515072981,
+      "learning_rate": 8.394272035766754e-06,
+      "loss": 0.3789,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28477663371210565,
+      "grad_norm": 0.7335034628385203,
+      "learning_rate": 8.393296283061213e-06,
+      "loss": 0.3518,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28485867826229644,
+      "grad_norm": 0.6617254072638449,
+      "learning_rate": 8.392320290726543e-06,
+      "loss": 0.3244,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28494072281248717,
+      "grad_norm": 0.6625476628455924,
+      "learning_rate": 8.391344058831664e-06,
+      "loss": 0.352,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28502276736267795,
+      "grad_norm": 0.6545296910639843,
+      "learning_rate": 8.390367587445516e-06,
+      "loss": 0.391,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2851048119128687,
+      "grad_norm": 0.6996455541033507,
+      "learning_rate": 8.389390876637056e-06,
+      "loss": 0.3997,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28518685646305947,
+      "grad_norm": 0.6541767356483655,
+      "learning_rate": 8.388413926475256e-06,
+      "loss": 0.3899,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2852689010132502,
+      "grad_norm": 0.6118599665210402,
+      "learning_rate": 8.387436737029111e-06,
+      "loss": 0.3645,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2853509455634409,
+      "grad_norm": 0.6867180316668599,
+      "learning_rate": 8.386459308367624e-06,
+      "loss": 0.3549,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2854329901136317,
+      "grad_norm": 0.6895838337569483,
+      "learning_rate": 8.385481640559823e-06,
+      "loss": 0.3498,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28551503466382244,
+      "grad_norm": 0.6319090426101271,
+      "learning_rate": 8.384503733674745e-06,
+      "loss": 0.3645,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2855970792140132,
+      "grad_norm": 0.7374892079836192,
+      "learning_rate": 8.383525587781453e-06,
+      "loss": 0.3356,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28567912376420396,
+      "grad_norm": 0.7509844764605954,
+      "learning_rate": 8.382547202949021e-06,
+      "loss": 0.3791,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28576116831439474,
+      "grad_norm": 0.6634690427660834,
+      "learning_rate": 8.381568579246539e-06,
+      "loss": 0.3521,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28584321286458547,
+      "grad_norm": 0.6340079404616704,
+      "learning_rate": 8.380589716743117e-06,
+      "loss": 0.3586,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2859252574147762,
+      "grad_norm": 0.6270316508620339,
+      "learning_rate": 8.37961061550788e-06,
+      "loss": 0.3764,
+      "step": 3485
+    },
+    {
+      "epoch": 0.286007301964967,
+      "grad_norm": 0.6476940181205523,
+      "learning_rate": 8.37863127560997e-06,
+      "loss": 0.3465,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2860893465151577,
+      "grad_norm": 0.6804904247372864,
+      "learning_rate": 8.37765169711855e-06,
+      "loss": 0.3991,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2861713910653485,
+      "grad_norm": 0.8023637783743709,
+      "learning_rate": 8.376671880102794e-06,
+      "loss": 0.382,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28625343561553923,
+      "grad_norm": 0.7086712056635153,
+      "learning_rate": 8.375691824631891e-06,
+      "loss": 0.3902,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28633548016573,
+      "grad_norm": 0.7252786448155563,
+      "learning_rate": 8.374711530775058e-06,
+      "loss": 0.3422,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28641752471592075,
+      "grad_norm": 0.6895012477776933,
+      "learning_rate": 8.373730998601519e-06,
+      "loss": 0.3448,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2864995692661115,
+      "grad_norm": 0.7032771358433052,
+      "learning_rate": 8.372750228180515e-06,
+      "loss": 0.3633,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28658161381630226,
+      "grad_norm": 0.7209193396089784,
+      "learning_rate": 8.37176921958131e-06,
+      "loss": 0.3668,
+      "step": 3493
+    },
+    {
+      "epoch": 0.286663658366493,
+      "grad_norm": 0.8552011541825834,
+      "learning_rate": 8.37078797287318e-06,
+      "loss": 0.3679,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2867457029166838,
+      "grad_norm": 0.6608783433490023,
+      "learning_rate": 8.369806488125418e-06,
+      "loss": 0.3915,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2868277474668745,
+      "grad_norm": 0.7707222522135421,
+      "learning_rate": 8.368824765407336e-06,
+      "loss": 0.4089,
+      "step": 3496
+    },
+    {
+      "epoch": 0.2869097920170653,
+      "grad_norm": 0.6867488796221437,
+      "learning_rate": 8.367842804788263e-06,
+      "loss": 0.3967,
+      "step": 3497
+    },
+    {
+      "epoch": 0.286991836567256,
+      "grad_norm": 0.6309949907602602,
+      "learning_rate": 8.36686060633754e-06,
+      "loss": 0.3586,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28707388111744675,
+      "grad_norm": 0.6732659043466915,
+      "learning_rate": 8.36587817012453e-06,
+      "loss": 0.3695,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28715592566763753,
+      "grad_norm": 0.6678061687797274,
+      "learning_rate": 8.364895496218612e-06,
+      "loss": 0.3591,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28723797021782826,
+      "grad_norm": 0.6426574810324396,
+      "learning_rate": 8.363912584689182e-06,
+      "loss": 0.3564,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28732001476801905,
+      "grad_norm": 0.6745034370897223,
+      "learning_rate": 8.362929435605648e-06,
+      "loss": 0.3571,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2874020593182098,
+      "grad_norm": 0.7565708504492948,
+      "learning_rate": 8.361946049037439e-06,
+      "loss": 0.378,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28748410386840056,
+      "grad_norm": 0.6369445780274209,
+      "learning_rate": 8.360962425054004e-06,
+      "loss": 0.3442,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2875661484185913,
+      "grad_norm": 0.618788914373755,
+      "learning_rate": 8.359978563724802e-06,
+      "loss": 0.4055,
+      "step": 3505
+    },
+    {
+      "epoch": 0.287648192968782,
+      "grad_norm": 0.7177517411269513,
+      "learning_rate": 8.358994465119311e-06,
+      "loss": 0.3324,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2877302375189728,
+      "grad_norm": 0.7088866973686898,
+      "learning_rate": 8.358010129307027e-06,
+      "loss": 0.3564,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28781228206916354,
+      "grad_norm": 0.8180993131154556,
+      "learning_rate": 8.357025556357464e-06,
+      "loss": 0.3494,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2878943266193543,
+      "grad_norm": 0.6899736316426106,
+      "learning_rate": 8.356040746340151e-06,
+      "loss": 0.3121,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28797637116954505,
+      "grad_norm": 0.7738044875567018,
+      "learning_rate": 8.35505569932463e-06,
+      "loss": 0.3582,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28805841571973584,
+      "grad_norm": 0.6817082341127795,
+      "learning_rate": 8.354070415380466e-06,
+      "loss": 0.3849,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28814046026992657,
+      "grad_norm": 0.8160073497138232,
+      "learning_rate": 8.353084894577238e-06,
+      "loss": 0.3716,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2882225048201173,
+      "grad_norm": 0.6391476567442804,
+      "learning_rate": 8.352099136984541e-06,
+      "loss": 0.3606,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2883045493703081,
+      "grad_norm": 0.6259213988616056,
+      "learning_rate": 8.351113142671991e-06,
+      "loss": 0.3624,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2883865939204988,
+      "grad_norm": 0.6443945932579539,
+      "learning_rate": 8.350126911709215e-06,
+      "loss": 0.3751,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2884686384706896,
+      "grad_norm": 0.6382517412307317,
+      "learning_rate": 8.349140444165857e-06,
+      "loss": 0.3429,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2885506830208803,
+      "grad_norm": 0.7389245938388516,
+      "learning_rate": 8.348153740111582e-06,
+      "loss": 0.3701,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2886327275710711,
+      "grad_norm": 0.7365417457639529,
+      "learning_rate": 8.347166799616069e-06,
+      "loss": 0.3717,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28871477212126184,
+      "grad_norm": 0.695274385938636,
+      "learning_rate": 8.346179622749015e-06,
+      "loss": 0.3779,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28879681667145257,
+      "grad_norm": 0.6428473080611645,
+      "learning_rate": 8.345192209580131e-06,
+      "loss": 0.3758,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28887886122164336,
+      "grad_norm": 0.6083095534741604,
+      "learning_rate": 8.344204560179148e-06,
+      "loss": 0.3961,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2889609057718341,
+      "grad_norm": 0.6904731948838451,
+      "learning_rate": 8.343216674615813e-06,
+      "loss": 0.3291,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28904295032202487,
+      "grad_norm": 0.6996120779563204,
+      "learning_rate": 8.342228552959887e-06,
+      "loss": 0.3703,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2891249948722156,
+      "grad_norm": 0.6246277186539626,
+      "learning_rate": 8.341240195281149e-06,
+      "loss": 0.3286,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2892070394224064,
+      "grad_norm": 0.6469852677872273,
+      "learning_rate": 8.340251601649398e-06,
+      "loss": 0.3994,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2892890839725971,
+      "grad_norm": 0.687514337487327,
+      "learning_rate": 8.339262772134444e-06,
+      "loss": 0.3475,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2893711285227879,
+      "grad_norm": 0.6986047844914042,
+      "learning_rate": 8.33827370680612e-06,
+      "loss": 0.376,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28945317307297863,
+      "grad_norm": 0.6738940212270224,
+      "learning_rate": 8.337284405734269e-06,
+      "loss": 0.3603,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28953521762316936,
+      "grad_norm": 0.7794610627358677,
+      "learning_rate": 8.336294868988755e-06,
+      "loss": 0.3203,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28961726217336015,
+      "grad_norm": 0.7051687526182074,
+      "learning_rate": 8.335305096639458e-06,
+      "loss": 0.3674,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2896993067235509,
+      "grad_norm": 0.6496163141892362,
+      "learning_rate": 8.334315088756273e-06,
+      "loss": 0.3561,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28978135127374166,
+      "grad_norm": 0.6738057931525337,
+      "learning_rate": 8.333324845409115e-06,
+      "loss": 0.3653,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2898633958239324,
+      "grad_norm": 0.6964536277050751,
+      "learning_rate": 8.33233436666791e-06,
+      "loss": 0.3835,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2899454403741232,
+      "grad_norm": 0.6216354018666645,
+      "learning_rate": 8.331343652602606e-06,
+      "loss": 0.3331,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2900274849243139,
+      "grad_norm": 0.7019787455914381,
+      "learning_rate": 8.330352703283166e-06,
+      "loss": 0.3845,
+      "step": 3535
+    },
+    {
+      "epoch": 0.29010952947450463,
+      "grad_norm": 0.6371413596532413,
+      "learning_rate": 8.329361518779569e-06,
+      "loss": 0.3849,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2901915740246954,
+      "grad_norm": 0.6467245350760537,
+      "learning_rate": 8.328370099161811e-06,
+      "loss": 0.3528,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29027361857488615,
+      "grad_norm": 1.1760046002296607,
+      "learning_rate": 8.327378444499901e-06,
+      "loss": 0.3766,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29035566312507693,
+      "grad_norm": 0.6343895279065744,
+      "learning_rate": 8.326386554863874e-06,
+      "loss": 0.336,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29043770767526766,
+      "grad_norm": 0.6772792426794163,
+      "learning_rate": 8.325394430323773e-06,
+      "loss": 0.3537,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29051975222545845,
+      "grad_norm": 0.7064613644768047,
+      "learning_rate": 8.324402070949658e-06,
+      "loss": 0.3831,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2906017967756492,
+      "grad_norm": 0.7803266441814598,
+      "learning_rate": 8.323409476811612e-06,
+      "loss": 0.3753,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2906838413258399,
+      "grad_norm": 0.6637101815904007,
+      "learning_rate": 8.322416647979726e-06,
+      "loss": 0.3438,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2907658858760307,
+      "grad_norm": 0.6063149112767192,
+      "learning_rate": 8.321423584524116e-06,
+      "loss": 0.3457,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2908479304262214,
+      "grad_norm": 0.6974808107826455,
+      "learning_rate": 8.320430286514908e-06,
+      "loss": 0.4038,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2909299749764122,
+      "grad_norm": 0.6469397991586194,
+      "learning_rate": 8.319436754022247e-06,
+      "loss": 0.3586,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29101201952660294,
+      "grad_norm": 0.617980119528497,
+      "learning_rate": 8.318442987116296e-06,
+      "loss": 0.3505,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2910940640767937,
+      "grad_norm": 0.6046028939962098,
+      "learning_rate": 8.317448985867233e-06,
+      "loss": 0.3281,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29117610862698445,
+      "grad_norm": 0.7121508278179322,
+      "learning_rate": 8.316454750345253e-06,
+      "loss": 0.3916,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2912581531771752,
+      "grad_norm": 0.7500961409996909,
+      "learning_rate": 8.315460280620566e-06,
+      "loss": 0.3737,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29134019772736597,
+      "grad_norm": 0.6311074624831574,
+      "learning_rate": 8.314465576763401e-06,
+      "loss": 0.3293,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2914222422775567,
+      "grad_norm": 0.7224201970041401,
+      "learning_rate": 8.313470638844003e-06,
+      "loss": 0.3803,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2915042868277475,
+      "grad_norm": 0.6560335227467121,
+      "learning_rate": 8.312475466932632e-06,
+      "loss": 0.3687,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2915863313779382,
+      "grad_norm": 0.599542053535244,
+      "learning_rate": 8.311480061099563e-06,
+      "loss": 0.3307,
+      "step": 3554
+    },
+    {
+      "epoch": 0.291668375928129,
+      "grad_norm": 0.6814476287367497,
+      "learning_rate": 8.310484421415094e-06,
+      "loss": 0.3297,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2917504204783197,
+      "grad_norm": 0.6296149373829947,
+      "learning_rate": 8.309488547949535e-06,
+      "loss": 0.3624,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29183246502851046,
+      "grad_norm": 0.743423712621411,
+      "learning_rate": 8.30849244077321e-06,
+      "loss": 0.371,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29191450957870124,
+      "grad_norm": 0.7483993195634401,
+      "learning_rate": 8.307496099956465e-06,
+      "loss": 0.3324,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29199655412889197,
+      "grad_norm": 0.6953397358823545,
+      "learning_rate": 8.306499525569661e-06,
+      "loss": 0.3549,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29207859867908276,
+      "grad_norm": 0.6467697636796358,
+      "learning_rate": 8.30550271768317e-06,
+      "loss": 0.3628,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2921606432292735,
+      "grad_norm": 0.6781305445741366,
+      "learning_rate": 8.304505676367388e-06,
+      "loss": 0.3719,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29224268777946427,
+      "grad_norm": 0.7031623478808917,
+      "learning_rate": 8.303508401692725e-06,
+      "loss": 0.3798,
+      "step": 3562
+    },
+    {
+      "epoch": 0.292324732329655,
+      "grad_norm": 0.6640199017729365,
+      "learning_rate": 8.302510893729607e-06,
+      "loss": 0.4223,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29240677687984573,
+      "grad_norm": 0.6352306380597317,
+      "learning_rate": 8.301513152548474e-06,
+      "loss": 0.3779,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2924888214300365,
+      "grad_norm": 0.6374775824993386,
+      "learning_rate": 8.300515178219788e-06,
+      "loss": 0.3455,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29257086598022725,
+      "grad_norm": 0.6491430989425522,
+      "learning_rate": 8.299516970814022e-06,
+      "loss": 0.3933,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29265291053041803,
+      "grad_norm": 0.6661220140790688,
+      "learning_rate": 8.298518530401668e-06,
+      "loss": 0.3681,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29273495508060876,
+      "grad_norm": 0.6884004275185375,
+      "learning_rate": 8.297519857053235e-06,
+      "loss": 0.3622,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29281699963079955,
+      "grad_norm": 0.6065103522155028,
+      "learning_rate": 8.29652095083925e-06,
+      "loss": 0.3783,
+      "step": 3569
+    },
+    {
+      "epoch": 0.2928990441809903,
+      "grad_norm": 0.659886239340308,
+      "learning_rate": 8.295521811830249e-06,
+      "loss": 0.348,
+      "step": 3570
+    },
+    {
+      "epoch": 0.292981088731181,
+      "grad_norm": 0.6111673564009654,
+      "learning_rate": 8.294522440096793e-06,
+      "loss": 0.3756,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2930631332813718,
+      "grad_norm": 0.726655664192741,
+      "learning_rate": 8.293522835709455e-06,
+      "loss": 0.3572,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2931451778315625,
+      "grad_norm": 0.6398650703459081,
+      "learning_rate": 8.292522998738827e-06,
+      "loss": 0.3457,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2932272223817533,
+      "grad_norm": 0.6190328780506671,
+      "learning_rate": 8.291522929255513e-06,
+      "loss": 0.3844,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29330926693194403,
+      "grad_norm": 0.6851343580567714,
+      "learning_rate": 8.29052262733014e-06,
+      "loss": 0.3989,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2933913114821348,
+      "grad_norm": 0.6307561256957867,
+      "learning_rate": 8.289522093033345e-06,
+      "loss": 0.3372,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29347335603232555,
+      "grad_norm": 0.634359172510698,
+      "learning_rate": 8.288521326435784e-06,
+      "loss": 0.366,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2935554005825163,
+      "grad_norm": 0.6503487387943174,
+      "learning_rate": 8.28752032760813e-06,
+      "loss": 0.3786,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29363744513270706,
+      "grad_norm": 0.6432810925259413,
+      "learning_rate": 8.286519096621073e-06,
+      "loss": 0.3478,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2937194896828978,
+      "grad_norm": 0.6029727401342846,
+      "learning_rate": 8.285517633545317e-06,
+      "loss": 0.3611,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2938015342330886,
+      "grad_norm": 0.6089744103165244,
+      "learning_rate": 8.284515938451586e-06,
+      "loss": 0.3099,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2938835787832793,
+      "grad_norm": 0.6665730547037353,
+      "learning_rate": 8.283514011410616e-06,
+      "loss": 0.3587,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2939656233334701,
+      "grad_norm": 0.6442164303776041,
+      "learning_rate": 8.28251185249316e-06,
+      "loss": 0.3445,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2940476678836608,
+      "grad_norm": 0.7230755398038096,
+      "learning_rate": 8.281509461769992e-06,
+      "loss": 0.3463,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2941297124338516,
+      "grad_norm": 0.631047798262599,
+      "learning_rate": 8.280506839311898e-06,
+      "loss": 0.3754,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29421175698404234,
+      "grad_norm": 0.6642789640860559,
+      "learning_rate": 8.27950398518968e-06,
+      "loss": 0.3921,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29429380153423307,
+      "grad_norm": 0.6923855876030156,
+      "learning_rate": 8.278500899474162e-06,
+      "loss": 0.3421,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29437584608442385,
+      "grad_norm": 0.6224219452471373,
+      "learning_rate": 8.277497582236177e-06,
+      "loss": 0.3528,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2944578906346146,
+      "grad_norm": 0.6973986957952036,
+      "learning_rate": 8.276494033546579e-06,
+      "loss": 0.3926,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29453993518480537,
+      "grad_norm": 0.6838799524426902,
+      "learning_rate": 8.275490253476236e-06,
+      "loss": 0.3567,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2946219797349961,
+      "grad_norm": 0.6817997931166102,
+      "learning_rate": 8.274486242096032e-06,
+      "loss": 0.392,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2947040242851869,
+      "grad_norm": 0.7401633047104399,
+      "learning_rate": 8.273481999476871e-06,
+      "loss": 0.3694,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2947860688353776,
+      "grad_norm": 0.5708210793289804,
+      "learning_rate": 8.272477525689672e-06,
+      "loss": 0.3645,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29486811338556834,
+      "grad_norm": 0.7897511779879776,
+      "learning_rate": 8.271472820805367e-06,
+      "loss": 0.3941,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2949501579357591,
+      "grad_norm": 0.6444597126779538,
+      "learning_rate": 8.270467884894908e-06,
+      "loss": 0.3368,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29503220248594986,
+      "grad_norm": 0.6621667758677521,
+      "learning_rate": 8.26946271802926e-06,
+      "loss": 0.3869,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29511424703614064,
+      "grad_norm": 0.7496557959018151,
+      "learning_rate": 8.268457320279408e-06,
+      "loss": 0.3869,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29519629158633137,
+      "grad_norm": 0.6657602796353397,
+      "learning_rate": 8.267451691716352e-06,
+      "loss": 0.3469,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29527833613652216,
+      "grad_norm": 0.5883409943559883,
+      "learning_rate": 8.266445832411108e-06,
+      "loss": 0.3816,
+      "step": 3599
+    },
+    {
+      "epoch": 0.2953603806867129,
+      "grad_norm": 0.6000569286096555,
+      "learning_rate": 8.265439742434706e-06,
+      "loss": 0.364,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2954424252369036,
+      "grad_norm": 0.6274879973377376,
+      "learning_rate": 8.264433421858195e-06,
+      "loss": 0.3605,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2955244697870944,
+      "grad_norm": 0.7000890237919551,
+      "learning_rate": 8.26342687075264e-06,
+      "loss": 0.3655,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29560651433728513,
+      "grad_norm": 0.6375886239238504,
+      "learning_rate": 8.262420089189124e-06,
+      "loss": 0.3435,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2956885588874759,
+      "grad_norm": 0.6244918847281141,
+      "learning_rate": 8.261413077238743e-06,
+      "loss": 0.3549,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29577060343766665,
+      "grad_norm": 0.7191676372032562,
+      "learning_rate": 8.260405834972611e-06,
+      "loss": 0.3432,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29585264798785743,
+      "grad_norm": 0.7332926621092345,
+      "learning_rate": 8.259398362461855e-06,
+      "loss": 0.3376,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29593469253804816,
+      "grad_norm": 0.5967374039342035,
+      "learning_rate": 8.258390659777625e-06,
+      "loss": 0.3207,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2960167370882389,
+      "grad_norm": 0.6269103921117167,
+      "learning_rate": 8.257382726991081e-06,
+      "loss": 0.3454,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2960987816384297,
+      "grad_norm": 0.6311135144322346,
+      "learning_rate": 8.256374564173401e-06,
+      "loss": 0.3975,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2961808261886204,
+      "grad_norm": 0.7973713007874345,
+      "learning_rate": 8.255366171395783e-06,
+      "loss": 0.3774,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2962628707388112,
+      "grad_norm": 0.7319646606256761,
+      "learning_rate": 8.254357548729435e-06,
+      "loss": 0.4193,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2963449152890019,
+      "grad_norm": 0.6740943599425254,
+      "learning_rate": 8.253348696245586e-06,
+      "loss": 0.377,
+      "step": 3612
+    },
+    {
+      "epoch": 0.2964269598391927,
+      "grad_norm": 0.6090005937134895,
+      "learning_rate": 8.252339614015478e-06,
+      "loss": 0.3358,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29650900438938343,
+      "grad_norm": 0.659427167593826,
+      "learning_rate": 8.251330302110373e-06,
+      "loss": 0.3478,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29659104893957416,
+      "grad_norm": 0.650235064877028,
+      "learning_rate": 8.250320760601544e-06,
+      "loss": 0.4057,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29667309348976495,
+      "grad_norm": 0.6365065767178258,
+      "learning_rate": 8.249310989560286e-06,
+      "loss": 0.3339,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2967551380399557,
+      "grad_norm": 0.8126977743513947,
+      "learning_rate": 8.248300989057906e-06,
+      "loss": 0.366,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29683718259014646,
+      "grad_norm": 0.6695079211895333,
+      "learning_rate": 8.247290759165731e-06,
+      "loss": 0.3888,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2969192271403372,
+      "grad_norm": 0.6238956985646354,
+      "learning_rate": 8.246280299955096e-06,
+      "loss": 0.3922,
+      "step": 3619
+    },
+    {
+      "epoch": 0.297001271690528,
+      "grad_norm": 0.6102053837218336,
+      "learning_rate": 8.245269611497365e-06,
+      "loss": 0.3545,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2970833162407187,
+      "grad_norm": 0.7108778581671541,
+      "learning_rate": 8.244258693863907e-06,
+      "loss": 0.3391,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29716536079090944,
+      "grad_norm": 0.6563424455626311,
+      "learning_rate": 8.243247547126112e-06,
+      "loss": 0.3457,
+      "step": 3622
+    },
+    {
+      "epoch": 0.2972474053411002,
+      "grad_norm": 0.638259877486403,
+      "learning_rate": 8.242236171355386e-06,
+      "loss": 0.3618,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29732944989129095,
+      "grad_norm": 0.6771981051182538,
+      "learning_rate": 8.24122456662315e-06,
+      "loss": 0.3657,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29741149444148174,
+      "grad_norm": 0.6416815625004777,
+      "learning_rate": 8.240212733000844e-06,
+      "loss": 0.3566,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29749353899167247,
+      "grad_norm": 0.6898132822501467,
+      "learning_rate": 8.23920067055992e-06,
+      "loss": 0.3551,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29757558354186325,
+      "grad_norm": 0.7824184177715442,
+      "learning_rate": 8.23818837937185e-06,
+      "loss": 0.4062,
+      "step": 3627
+    },
+    {
+      "epoch": 0.297657628092054,
+      "grad_norm": 0.7533786266343193,
+      "learning_rate": 8.237175859508116e-06,
+      "loss": 0.3706,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2977396726422447,
+      "grad_norm": 0.6345265818044278,
+      "learning_rate": 8.236163111040227e-06,
+      "loss": 0.3697,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2978217171924355,
+      "grad_norm": 0.6691662625736022,
+      "learning_rate": 8.235150134039696e-06,
+      "loss": 0.4033,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2979037617426262,
+      "grad_norm": 0.6442645788282831,
+      "learning_rate": 8.23413692857806e-06,
+      "loss": 0.3525,
+      "step": 3631
+    },
+    {
+      "epoch": 0.297985806292817,
+      "grad_norm": 0.6552997179892466,
+      "learning_rate": 8.233123494726873e-06,
+      "loss": 0.3751,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29806785084300774,
+      "grad_norm": 0.6922764908281266,
+      "learning_rate": 8.232109832557696e-06,
+      "loss": 0.4455,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2981498953931985,
+      "grad_norm": 0.5897808517142249,
+      "learning_rate": 8.231095942142117e-06,
+      "loss": 0.3425,
+      "step": 3634
+    },
+    {
+      "epoch": 0.29823193994338926,
+      "grad_norm": 0.7653187201466285,
+      "learning_rate": 8.230081823551734e-06,
+      "loss": 0.3404,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29831398449358,
+      "grad_norm": 0.7159661955567239,
+      "learning_rate": 8.229067476858162e-06,
+      "loss": 0.3615,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29839602904377077,
+      "grad_norm": 0.7861930509592386,
+      "learning_rate": 8.228052902133032e-06,
+      "loss": 0.3661,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2984780735939615,
+      "grad_norm": 0.7115417254867795,
+      "learning_rate": 8.227038099447992e-06,
+      "loss": 0.3647,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2985601181441523,
+      "grad_norm": 0.6302883119751456,
+      "learning_rate": 8.226023068874707e-06,
+      "loss": 0.3542,
+      "step": 3639
+    },
+    {
+      "epoch": 0.298642162694343,
+      "grad_norm": 0.6434313628920583,
+      "learning_rate": 8.225007810484857e-06,
+      "loss": 0.367,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2987242072445338,
+      "grad_norm": 0.6547364429458058,
+      "learning_rate": 8.223992324350136e-06,
+      "loss": 0.355,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29880625179472453,
+      "grad_norm": 0.6556560205467691,
+      "learning_rate": 8.222976610542255e-06,
+      "loss": 0.3503,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2988882963449153,
+      "grad_norm": 0.6343337182429735,
+      "learning_rate": 8.221960669132946e-06,
+      "loss": 0.3697,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29897034089510605,
+      "grad_norm": 0.6660543025783155,
+      "learning_rate": 8.22094450019395e-06,
+      "loss": 0.3538,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2990523854452968,
+      "grad_norm": 0.6323255188268146,
+      "learning_rate": 8.219928103797027e-06,
+      "loss": 0.3676,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29913442999548756,
+      "grad_norm": 0.7331387710855694,
+      "learning_rate": 8.218911480013956e-06,
+      "loss": 0.3597,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2992164745456783,
+      "grad_norm": 0.7358359268865273,
+      "learning_rate": 8.217894628916529e-06,
+      "loss": 0.3579,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2992985190958691,
+      "grad_norm": 0.7247127536082223,
+      "learning_rate": 8.216877550576552e-06,
+      "loss": 0.3817,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2993805636460598,
+      "grad_norm": 0.6026506284120019,
+      "learning_rate": 8.21586024506585e-06,
+      "loss": 0.3536,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2994626081962506,
+      "grad_norm": 0.7646876979700887,
+      "learning_rate": 8.214842712456266e-06,
+      "loss": 0.3631,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2995446527464413,
+      "grad_norm": 0.6205685296821226,
+      "learning_rate": 8.213824952819653e-06,
+      "loss": 0.3635,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29962669729663205,
+      "grad_norm": 0.654455213911147,
+      "learning_rate": 8.212806966227884e-06,
+      "loss": 0.3738,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29970874184682283,
+      "grad_norm": 0.6389881875571665,
+      "learning_rate": 8.211788752752851e-06,
+      "loss": 0.3719,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29979078639701356,
+      "grad_norm": 0.6059923394151949,
+      "learning_rate": 8.210770312466457e-06,
+      "loss": 0.3548,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29987283094720435,
+      "grad_norm": 0.6299163066578031,
+      "learning_rate": 8.20975164544062e-06,
+      "loss": 0.3524,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2999548754973951,
+      "grad_norm": 0.6392035697540956,
+      "learning_rate": 8.208732751747281e-06,
+      "loss": 0.3598,
+      "step": 3656
+    },
+    {
+      "epoch": 0.30003692004758586,
+      "grad_norm": 0.6513946533404006,
+      "learning_rate": 8.20771363145839e-06,
+      "loss": 0.3731,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3001189645977766,
+      "grad_norm": 0.6456693432294469,
+      "learning_rate": 8.206694284645914e-06,
+      "loss": 0.3583,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3002010091479673,
+      "grad_norm": 0.6453263295597355,
+      "learning_rate": 8.205674711381841e-06,
+      "loss": 0.366,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3002830536981581,
+      "grad_norm": 0.7561994766024434,
+      "learning_rate": 8.204654911738169e-06,
+      "loss": 0.3768,
+      "step": 3660
+    },
+    {
+      "epoch": 0.30036509824834884,
+      "grad_norm": 0.7452132925867124,
+      "learning_rate": 8.203634885786916e-06,
+      "loss": 0.3586,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3004471427985396,
+      "grad_norm": 0.5761056620366076,
+      "learning_rate": 8.202614633600116e-06,
+      "loss": 0.356,
+      "step": 3662
+    },
+    {
+      "epoch": 0.30052918734873035,
+      "grad_norm": 0.6947477641576664,
+      "learning_rate": 8.201594155249814e-06,
+      "loss": 0.3407,
+      "step": 3663
+    },
+    {
+      "epoch": 0.30061123189892114,
+      "grad_norm": 0.6271732766471348,
+      "learning_rate": 8.200573450808079e-06,
+      "loss": 0.3675,
+      "step": 3664
+    },
+    {
+      "epoch": 0.30069327644911187,
+      "grad_norm": 0.715113388129879,
+      "learning_rate": 8.199552520346985e-06,
+      "loss": 0.4249,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3007753209993026,
+      "grad_norm": 0.6778471170928092,
+      "learning_rate": 8.198531363938635e-06,
+      "loss": 0.3456,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3008573655494934,
+      "grad_norm": 0.5704621991035866,
+      "learning_rate": 8.19750998165514e-06,
+      "loss": 0.3327,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3009394100996841,
+      "grad_norm": 0.6890884398981182,
+      "learning_rate": 8.196488373568627e-06,
+      "loss": 0.3395,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3010214546498749,
+      "grad_norm": 0.6650477986619512,
+      "learning_rate": 8.195466539751237e-06,
+      "loss": 0.3454,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3011034992000656,
+      "grad_norm": 0.7072168083765741,
+      "learning_rate": 8.194444480275138e-06,
+      "loss": 0.375,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3011855437502564,
+      "grad_norm": 0.6267027951103582,
+      "learning_rate": 8.1934221952125e-06,
+      "loss": 0.3524,
+      "step": 3671
+    },
+    {
+      "epoch": 0.30126758830044714,
+      "grad_norm": 0.7339586146156185,
+      "learning_rate": 8.192399684635518e-06,
+      "loss": 0.3607,
+      "step": 3672
+    },
+    {
+      "epoch": 0.30134963285063787,
+      "grad_norm": 0.6401070773412432,
+      "learning_rate": 8.1913769486164e-06,
+      "loss": 0.3138,
+      "step": 3673
+    },
+    {
+      "epoch": 0.30143167740082866,
+      "grad_norm": 0.6819634284938385,
+      "learning_rate": 8.190353987227368e-06,
+      "loss": 0.3801,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3015137219510194,
+      "grad_norm": 0.6332827764577323,
+      "learning_rate": 8.189330800540662e-06,
+      "loss": 0.3467,
+      "step": 3675
+    },
+    {
+      "epoch": 0.30159576650121017,
+      "grad_norm": 0.6428811281422284,
+      "learning_rate": 8.188307388628542e-06,
+      "loss": 0.3484,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3016778110514009,
+      "grad_norm": 0.6849333939129765,
+      "learning_rate": 8.187283751563275e-06,
+      "loss": 0.3824,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3017598556015917,
+      "grad_norm": 0.616381359430852,
+      "learning_rate": 8.186259889417149e-06,
+      "loss": 0.3735,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3018419001517824,
+      "grad_norm": 0.7332272667209161,
+      "learning_rate": 8.18523580226247e-06,
+      "loss": 0.3857,
+      "step": 3679
+    },
+    {
+      "epoch": 0.30192394470197315,
+      "grad_norm": 0.6849108715925101,
+      "learning_rate": 8.184211490171557e-06,
+      "loss": 0.3666,
+      "step": 3680
+    },
+    {
+      "epoch": 0.30200598925216393,
+      "grad_norm": 0.6704428751675263,
+      "learning_rate": 8.183186953216743e-06,
+      "loss": 0.3506,
+      "step": 3681
+    },
+    {
+      "epoch": 0.30208803380235466,
+      "grad_norm": 0.714640095208256,
+      "learning_rate": 8.18216219147038e-06,
+      "loss": 0.3634,
+      "step": 3682
+    },
+    {
+      "epoch": 0.30217007835254545,
+      "grad_norm": 0.7114754792366482,
+      "learning_rate": 8.181137205004836e-06,
+      "loss": 0.3489,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3022521229027362,
+      "grad_norm": 0.6891030127240041,
+      "learning_rate": 8.180111993892494e-06,
+      "loss": 0.3387,
+      "step": 3684
+    },
+    {
+      "epoch": 0.30233416745292696,
+      "grad_norm": 0.695446166347455,
+      "learning_rate": 8.179086558205754e-06,
+      "loss": 0.386,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3024162120031177,
+      "grad_norm": 0.6646124895496505,
+      "learning_rate": 8.178060898017027e-06,
+      "loss": 0.327,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3024982565533084,
+      "grad_norm": 0.6406450692977516,
+      "learning_rate": 8.177035013398744e-06,
+      "loss": 0.386,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3025803011034992,
+      "grad_norm": 0.6293813815558957,
+      "learning_rate": 8.176008904423354e-06,
+      "loss": 0.4364,
+      "step": 3688
+    },
+    {
+      "epoch": 0.30266234565368993,
+      "grad_norm": 0.6246980516282074,
+      "learning_rate": 8.17498257116332e-06,
+      "loss": 0.3655,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3027443902038807,
+      "grad_norm": 0.6515939557908829,
+      "learning_rate": 8.173956013691115e-06,
+      "loss": 0.359,
+      "step": 3690
+    },
+    {
+      "epoch": 0.30282643475407145,
+      "grad_norm": 0.6174616355625498,
+      "learning_rate": 8.172929232079237e-06,
+      "loss": 0.3768,
+      "step": 3691
+    },
+    {
+      "epoch": 0.30290847930426223,
+      "grad_norm": 0.6659977087219559,
+      "learning_rate": 8.171902226400192e-06,
+      "loss": 0.3671,
+      "step": 3692
+    },
+    {
+      "epoch": 0.30299052385445296,
+      "grad_norm": 0.6985377160200529,
+      "learning_rate": 8.17087499672651e-06,
+      "loss": 0.3907,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3030725684046437,
+      "grad_norm": 0.6654448702284389,
+      "learning_rate": 8.16984754313073e-06,
+      "loss": 0.3594,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3031546129548345,
+      "grad_norm": 0.6833930174490663,
+      "learning_rate": 8.16881986568541e-06,
+      "loss": 0.3623,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3032366575050252,
+      "grad_norm": 0.7487952615884995,
+      "learning_rate": 8.167791964463119e-06,
+      "loss": 0.324,
+      "step": 3696
+    },
+    {
+      "epoch": 0.303318702055216,
+      "grad_norm": 0.6104002003052371,
+      "learning_rate": 8.166763839536451e-06,
+      "loss": 0.3214,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3034007466054067,
+      "grad_norm": 0.676941931357917,
+      "learning_rate": 8.165735490978009e-06,
+      "loss": 0.36,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3034827911555975,
+      "grad_norm": 0.6338090419478866,
+      "learning_rate": 8.164706918860413e-06,
+      "loss": 0.366,
+      "step": 3699
+    },
+    {
+      "epoch": 0.30356483570578824,
+      "grad_norm": 0.6839412925567715,
+      "learning_rate": 8.163678123256298e-06,
+      "loss": 0.3573,
+      "step": 3700
+    },
+    {
+      "epoch": 0.303646880255979,
+      "grad_norm": 0.6483757337686723,
+      "learning_rate": 8.162649104238317e-06,
+      "loss": 0.448,
+      "step": 3701
+    },
+    {
+      "epoch": 0.30372892480616975,
+      "grad_norm": 0.639655469631495,
+      "learning_rate": 8.161619861879136e-06,
+      "loss": 0.4177,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3038109693563605,
+      "grad_norm": 0.6622430278969967,
+      "learning_rate": 8.160590396251442e-06,
+      "loss": 0.3982,
+      "step": 3703
+    },
+    {
+      "epoch": 0.30389301390655127,
+      "grad_norm": 0.8296519444821848,
+      "learning_rate": 8.159560707427931e-06,
+      "loss": 0.3722,
+      "step": 3704
+    },
+    {
+      "epoch": 0.303975058456742,
+      "grad_norm": 0.65312055420101,
+      "learning_rate": 8.158530795481318e-06,
+      "loss": 0.3618,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3040571030069328,
+      "grad_norm": 0.6396627037384319,
+      "learning_rate": 8.157500660484338e-06,
+      "loss": 0.389,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3041391475571235,
+      "grad_norm": 0.8488563043330178,
+      "learning_rate": 8.156470302509732e-06,
+      "loss": 0.363,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3042211921073143,
+      "grad_norm": 0.8949899371459598,
+      "learning_rate": 8.155439721630265e-06,
+      "loss": 0.3911,
+      "step": 3708
+    },
+    {
+      "epoch": 0.304303236657505,
+      "grad_norm": 0.6586773593839367,
+      "learning_rate": 8.154408917918715e-06,
+      "loss": 0.3508,
+      "step": 3709
+    },
+    {
+      "epoch": 0.30438528120769576,
+      "grad_norm": 0.6683063922685982,
+      "learning_rate": 8.153377891447873e-06,
+      "loss": 0.3872,
+      "step": 3710
+    },
+    {
+      "epoch": 0.30446732575788654,
+      "grad_norm": 0.625530565985483,
+      "learning_rate": 8.152346642290553e-06,
+      "loss": 0.3415,
+      "step": 3711
+    },
+    {
+      "epoch": 0.30454937030807727,
+      "grad_norm": 0.6196339556766683,
+      "learning_rate": 8.151315170519576e-06,
+      "loss": 0.3516,
+      "step": 3712
+    },
+    {
+      "epoch": 0.30463141485826806,
+      "grad_norm": 0.6582202616014271,
+      "learning_rate": 8.150283476207785e-06,
+      "loss": 0.3592,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3047134594084588,
+      "grad_norm": 0.6471003772103989,
+      "learning_rate": 8.149251559428037e-06,
+      "loss": 0.3764,
+      "step": 3714
+    },
+    {
+      "epoch": 0.30479550395864957,
+      "grad_norm": 0.6000829175740287,
+      "learning_rate": 8.148219420253204e-06,
+      "loss": 0.3443,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3048775485088403,
+      "grad_norm": 0.6820618672812747,
+      "learning_rate": 8.147187058756173e-06,
+      "loss": 0.3725,
+      "step": 3716
+    },
+    {
+      "epoch": 0.30495959305903103,
+      "grad_norm": 0.6464593079106565,
+      "learning_rate": 8.146154475009849e-06,
+      "loss": 0.4129,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3050416376092218,
+      "grad_norm": 0.5990658193462356,
+      "learning_rate": 8.145121669087148e-06,
+      "loss": 0.3981,
+      "step": 3718
+    },
+    {
+      "epoch": 0.30512368215941255,
+      "grad_norm": 0.7368837391827804,
+      "learning_rate": 8.14408864106101e-06,
+      "loss": 0.3669,
+      "step": 3719
+    },
+    {
+      "epoch": 0.30520572670960333,
+      "grad_norm": 0.6330419826420343,
+      "learning_rate": 8.143055391004383e-06,
+      "loss": 0.3596,
+      "step": 3720
+    },
+    {
+      "epoch": 0.30528777125979406,
+      "grad_norm": 0.6796482450217874,
+      "learning_rate": 8.142021918990234e-06,
+      "loss": 0.3457,
+      "step": 3721
+    },
+    {
+      "epoch": 0.30536981580998485,
+      "grad_norm": 0.6527959583661851,
+      "learning_rate": 8.140988225091544e-06,
+      "loss": 0.3774,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3054518603601756,
+      "grad_norm": 0.7193023778973627,
+      "learning_rate": 8.139954309381314e-06,
+      "loss": 0.3872,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3055339049103663,
+      "grad_norm": 0.633028113186812,
+      "learning_rate": 8.138920171932554e-06,
+      "loss": 0.3431,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3056159494605571,
+      "grad_norm": 0.7903542669249172,
+      "learning_rate": 8.137885812818296e-06,
+      "loss": 0.3861,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3056979940107478,
+      "grad_norm": 0.6718901445630933,
+      "learning_rate": 8.136851232111582e-06,
+      "loss": 0.3152,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3057800385609386,
+      "grad_norm": 0.6319407351078121,
+      "learning_rate": 8.135816429885473e-06,
+      "loss": 0.3538,
+      "step": 3727
+    },
+    {
+      "epoch": 0.30586208311112933,
+      "grad_norm": 0.7186039003497519,
+      "learning_rate": 8.134781406213048e-06,
+      "loss": 0.4108,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3059441276613201,
+      "grad_norm": 0.697040414967708,
+      "learning_rate": 8.133746161167396e-06,
+      "loss": 0.3745,
+      "step": 3729
+    },
+    {
+      "epoch": 0.30602617221151085,
+      "grad_norm": 0.5967516965760348,
+      "learning_rate": 8.132710694821625e-06,
+      "loss": 0.355,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3061082167617016,
+      "grad_norm": 0.641143424217867,
+      "learning_rate": 8.131675007248854e-06,
+      "loss": 0.3742,
+      "step": 3731
+    },
+    {
+      "epoch": 0.30619026131189236,
+      "grad_norm": 0.7042387759902232,
+      "learning_rate": 8.130639098522229e-06,
+      "loss": 0.3438,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3062723058620831,
+      "grad_norm": 0.6760606517123672,
+      "learning_rate": 8.1296029687149e-06,
+      "loss": 0.352,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3063543504122739,
+      "grad_norm": 0.6904951491455673,
+      "learning_rate": 8.128566617900038e-06,
+      "loss": 0.3445,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3064363949624646,
+      "grad_norm": 0.7478448105713729,
+      "learning_rate": 8.127530046150827e-06,
+      "loss": 0.3246,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3065184395126554,
+      "grad_norm": 0.699408515137096,
+      "learning_rate": 8.126493253540468e-06,
+      "loss": 0.3942,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3066004840628461,
+      "grad_norm": 0.6245814278527828,
+      "learning_rate": 8.12545624014218e-06,
+      "loss": 0.38,
+      "step": 3737
+    },
+    {
+      "epoch": 0.30668252861303685,
+      "grad_norm": 0.6399073847357779,
+      "learning_rate": 8.124419006029194e-06,
+      "loss": 0.3814,
+      "step": 3738
+    },
+    {
+      "epoch": 0.30676457316322764,
+      "grad_norm": 0.6484955172867246,
+      "learning_rate": 8.123381551274757e-06,
+      "loss": 0.3326,
+      "step": 3739
+    },
+    {
+      "epoch": 0.30684661771341837,
+      "grad_norm": 0.6563902836181652,
+      "learning_rate": 8.122343875952132e-06,
+      "loss": 0.3491,
+      "step": 3740
+    },
+    {
+      "epoch": 0.30692866226360915,
+      "grad_norm": 0.6947986187680709,
+      "learning_rate": 8.1213059801346e-06,
+      "loss": 0.3461,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3070107068137999,
+      "grad_norm": 0.6270537823736456,
+      "learning_rate": 8.120267863895454e-06,
+      "loss": 0.3455,
+      "step": 3742
+    },
+    {
+      "epoch": 0.30709275136399067,
+      "grad_norm": 0.7256888674252525,
+      "learning_rate": 8.119229527308006e-06,
+      "loss": 0.3619,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3071747959141814,
+      "grad_norm": 0.5854846940049835,
+      "learning_rate": 8.11819097044558e-06,
+      "loss": 0.3207,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3072568404643721,
+      "grad_norm": 0.6820854094251013,
+      "learning_rate": 8.117152193381514e-06,
+      "loss": 0.3376,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3073388850145629,
+      "grad_norm": 0.6077357388188029,
+      "learning_rate": 8.116113196189172e-06,
+      "loss": 0.3452,
+      "step": 3746
+    },
+    {
+      "epoch": 0.30742092956475364,
+      "grad_norm": 0.6191818946951178,
+      "learning_rate": 8.11507397894192e-06,
+      "loss": 0.3638,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3075029741149444,
+      "grad_norm": 0.6307666598701795,
+      "learning_rate": 8.114034541713152e-06,
+      "loss": 0.3527,
+      "step": 3748
+    },
+    {
+      "epoch": 0.30758501866513516,
+      "grad_norm": 0.6897836657645112,
+      "learning_rate": 8.112994884576263e-06,
+      "loss": 0.4309,
+      "step": 3749
+    },
+    {
+      "epoch": 0.30766706321532594,
+      "grad_norm": 0.764316386195445,
+      "learning_rate": 8.111955007604679e-06,
+      "loss": 0.3502,
+      "step": 3750
+    },
+    {
+      "epoch": 0.30774910776551667,
+      "grad_norm": 0.6944829968747057,
+      "learning_rate": 8.110914910871834e-06,
+      "loss": 0.337,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3078311523157074,
+      "grad_norm": 0.6152311676733283,
+      "learning_rate": 8.109874594451172e-06,
+      "loss": 0.3402,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3079131968658982,
+      "grad_norm": 0.6325560970700014,
+      "learning_rate": 8.108834058416167e-06,
+      "loss": 0.3577,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3079952414160889,
+      "grad_norm": 0.6649117388536481,
+      "learning_rate": 8.107793302840292e-06,
+      "loss": 0.3716,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3080772859662797,
+      "grad_norm": 0.7025638345168482,
+      "learning_rate": 8.106752327797047e-06,
+      "loss": 0.3472,
+      "step": 3755
+    },
+    {
+      "epoch": 0.30815933051647043,
+      "grad_norm": 0.690686430472669,
+      "learning_rate": 8.105711133359946e-06,
+      "loss": 0.3924,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3082413750666612,
+      "grad_norm": 0.7638037162775908,
+      "learning_rate": 8.104669719602514e-06,
+      "loss": 0.3874,
+      "step": 3757
+    },
+    {
+      "epoch": 0.30832341961685195,
+      "grad_norm": 0.7063535891478551,
+      "learning_rate": 8.103628086598293e-06,
+      "loss": 0.3263,
+      "step": 3758
+    },
+    {
+      "epoch": 0.30840546416704273,
+      "grad_norm": 0.6930317178207955,
+      "learning_rate": 8.102586234420843e-06,
+      "loss": 0.3306,
+      "step": 3759
+    },
+    {
+      "epoch": 0.30848750871723346,
+      "grad_norm": 0.6589628128630135,
+      "learning_rate": 8.10154416314374e-06,
+      "loss": 0.3927,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3085695532674242,
+      "grad_norm": 0.8679995197507775,
+      "learning_rate": 8.10050187284057e-06,
+      "loss": 0.3527,
+      "step": 3761
+    },
+    {
+      "epoch": 0.308651597817615,
+      "grad_norm": 0.6500420558104189,
+      "learning_rate": 8.099459363584939e-06,
+      "loss": 0.3506,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3087336423678057,
+      "grad_norm": 0.6820795028662977,
+      "learning_rate": 8.098416635450467e-06,
+      "loss": 0.323,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3088156869179965,
+      "grad_norm": 0.8172453296134877,
+      "learning_rate": 8.097373688510791e-06,
+      "loss": 0.3589,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3088977314681872,
+      "grad_norm": 0.690002560494531,
+      "learning_rate": 8.096330522839562e-06,
+      "loss": 0.3796,
+      "step": 3765
+    },
+    {
+      "epoch": 0.308979776018378,
+      "grad_norm": 0.6437657353090358,
+      "learning_rate": 8.095287138510444e-06,
+      "loss": 0.3449,
+      "step": 3766
+    },
+    {
+      "epoch": 0.30906182056856873,
+      "grad_norm": 0.631329022331102,
+      "learning_rate": 8.094243535597122e-06,
+      "loss": 0.332,
+      "step": 3767
+    },
+    {
+      "epoch": 0.30914386511875946,
+      "grad_norm": 0.6632217754130587,
+      "learning_rate": 8.093199714173295e-06,
+      "loss": 0.3609,
+      "step": 3768
+    },
+    {
+      "epoch": 0.30922590966895025,
+      "grad_norm": 0.6190182755280317,
+      "learning_rate": 8.092155674312672e-06,
+      "loss": 0.3401,
+      "step": 3769
+    },
+    {
+      "epoch": 0.309307954219141,
+      "grad_norm": 0.6700915707663538,
+      "learning_rate": 8.091111416088982e-06,
+      "loss": 0.3343,
+      "step": 3770
+    },
+    {
+      "epoch": 0.30938999876933176,
+      "grad_norm": 0.7041549628336048,
+      "learning_rate": 8.090066939575972e-06,
+      "loss": 0.3691,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3094720433195225,
+      "grad_norm": 0.6827893084338273,
+      "learning_rate": 8.089022244847398e-06,
+      "loss": 0.3459,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3095540878697133,
+      "grad_norm": 0.6050722508117824,
+      "learning_rate": 8.087977331977037e-06,
+      "loss": 0.3187,
+      "step": 3773
+    },
+    {
+      "epoch": 0.309636132419904,
+      "grad_norm": 0.6622323771789739,
+      "learning_rate": 8.086932201038677e-06,
+      "loss": 0.3493,
+      "step": 3774
+    },
+    {
+      "epoch": 0.30971817697009474,
+      "grad_norm": 0.7444730415871988,
+      "learning_rate": 8.085886852106126e-06,
+      "loss": 0.3159,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3098002215202855,
+      "grad_norm": 0.7822435732768103,
+      "learning_rate": 8.084841285253202e-06,
+      "loss": 0.3635,
+      "step": 3776
+    },
+    {
+      "epoch": 0.30988226607047625,
+      "grad_norm": 0.6193612202493416,
+      "learning_rate": 8.08379550055374e-06,
+      "loss": 0.3542,
+      "step": 3777
+    },
+    {
+      "epoch": 0.30996431062066704,
+      "grad_norm": 0.6409272255557549,
+      "learning_rate": 8.0827494980816e-06,
+      "loss": 0.3353,
+      "step": 3778
+    },
+    {
+      "epoch": 0.31004635517085777,
+      "grad_norm": 0.7003568573702437,
+      "learning_rate": 8.08170327791064e-06,
+      "loss": 0.3792,
+      "step": 3779
+    },
+    {
+      "epoch": 0.31012839972104855,
+      "grad_norm": 0.7305430668852296,
+      "learning_rate": 8.080656840114746e-06,
+      "loss": 0.3601,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3102104442712393,
+      "grad_norm": 0.6506541728102117,
+      "learning_rate": 8.079610184767817e-06,
+      "loss": 0.3502,
+      "step": 3781
+    },
+    {
+      "epoch": 0.31029248882143,
+      "grad_norm": 0.663606491894116,
+      "learning_rate": 8.078563311943762e-06,
+      "loss": 0.3728,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3103745333716208,
+      "grad_norm": 0.8156997176488361,
+      "learning_rate": 8.077516221716512e-06,
+      "loss": 0.3428,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3104565779218115,
+      "grad_norm": 0.6055363967819456,
+      "learning_rate": 8.076468914160012e-06,
+      "loss": 0.3577,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3105386224720023,
+      "grad_norm": 0.7158928624816983,
+      "learning_rate": 8.07542138934822e-06,
+      "loss": 0.4011,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31062066702219304,
+      "grad_norm": 0.6401908618143217,
+      "learning_rate": 8.074373647355111e-06,
+      "loss": 0.3922,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3107027115723838,
+      "grad_norm": 0.7037931875162666,
+      "learning_rate": 8.073325688254674e-06,
+      "loss": 0.4126,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31078475612257456,
+      "grad_norm": 0.6478800700869541,
+      "learning_rate": 8.072277512120913e-06,
+      "loss": 0.3507,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3108668006727653,
+      "grad_norm": 1.0708590596761325,
+      "learning_rate": 8.071229119027854e-06,
+      "loss": 0.3482,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31094884522295607,
+      "grad_norm": 0.6377954153801891,
+      "learning_rate": 8.070180509049525e-06,
+      "loss": 0.369,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3110308897731468,
+      "grad_norm": 0.6242527972454538,
+      "learning_rate": 8.069131682259982e-06,
+      "loss": 0.3742,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3111129343233376,
+      "grad_norm": 0.7138263359445549,
+      "learning_rate": 8.068082638733292e-06,
+      "loss": 0.3454,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3111949788735283,
+      "grad_norm": 0.6598161329568597,
+      "learning_rate": 8.067033378543533e-06,
+      "loss": 0.4027,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3112770234237191,
+      "grad_norm": 0.6560553662045475,
+      "learning_rate": 8.065983901764807e-06,
+      "loss": 0.3512,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31135906797390983,
+      "grad_norm": 0.6808179655597337,
+      "learning_rate": 8.064934208471223e-06,
+      "loss": 0.3316,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31144111252410056,
+      "grad_norm": 0.6281325668961318,
+      "learning_rate": 8.063884298736908e-06,
+      "loss": 0.3901,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31152315707429135,
+      "grad_norm": 0.6122081147300408,
+      "learning_rate": 8.062834172636007e-06,
+      "loss": 0.3572,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3116052016244821,
+      "grad_norm": 0.6220295723603877,
+      "learning_rate": 8.061783830242677e-06,
+      "loss": 0.3482,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31168724617467286,
+      "grad_norm": 0.7981226439039777,
+      "learning_rate": 8.060733271631094e-06,
+      "loss": 0.349,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3117692907248636,
+      "grad_norm": 0.6820320767824207,
+      "learning_rate": 8.059682496875443e-06,
+      "loss": 0.3519,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3118513352750544,
+      "grad_norm": 0.617233903042285,
+      "learning_rate": 8.05863150604993e-06,
+      "loss": 0.3247,
+      "step": 3801
+    },
+    {
+      "epoch": 0.3119333798252451,
+      "grad_norm": 0.6494984493154473,
+      "learning_rate": 8.057580299228775e-06,
+      "loss": 0.3706,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31201542437543583,
+      "grad_norm": 0.6241086261413347,
+      "learning_rate": 8.05652887648621e-06,
+      "loss": 0.3651,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3120974689256266,
+      "grad_norm": 0.5974246493782609,
+      "learning_rate": 8.055477237896488e-06,
+      "loss": 0.3227,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31217951347581735,
+      "grad_norm": 0.6988472281649473,
+      "learning_rate": 8.054425383533873e-06,
+      "loss": 0.3577,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31226155802600813,
+      "grad_norm": 0.6419930526703052,
+      "learning_rate": 8.053373313472643e-06,
+      "loss": 0.381,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31234360257619886,
+      "grad_norm": 0.6063727361863772,
+      "learning_rate": 8.052321027787095e-06,
+      "loss": 0.3447,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31242564712638965,
+      "grad_norm": 0.637572784861728,
+      "learning_rate": 8.051268526551541e-06,
+      "loss": 0.3519,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3125076916765804,
+      "grad_norm": 0.6499384224182844,
+      "learning_rate": 8.050215809840307e-06,
+      "loss": 0.3295,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31258973622677116,
+      "grad_norm": 0.6854629663562677,
+      "learning_rate": 8.049162877727732e-06,
+      "loss": 0.3717,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3126717807769619,
+      "grad_norm": 0.6281481369074717,
+      "learning_rate": 8.048109730288173e-06,
+      "loss": 0.3662,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3127538253271526,
+      "grad_norm": 0.6438323364392188,
+      "learning_rate": 8.047056367596003e-06,
+      "loss": 0.3404,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3128358698773434,
+      "grad_norm": 0.624793658813486,
+      "learning_rate": 8.046002789725606e-06,
+      "loss": 0.3546,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31291791442753414,
+      "grad_norm": 0.6209641221180479,
+      "learning_rate": 8.044948996751387e-06,
+      "loss": 0.3255,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3129999589777249,
+      "grad_norm": 0.6678031841070188,
+      "learning_rate": 8.043894988747761e-06,
+      "loss": 0.3807,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31308200352791565,
+      "grad_norm": 0.662185711859975,
+      "learning_rate": 8.042840765789164e-06,
+      "loss": 0.3638,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31316404807810644,
+      "grad_norm": 0.6921928054538047,
+      "learning_rate": 8.041786327950037e-06,
+      "loss": 0.3848,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31324609262829717,
+      "grad_norm": 0.7540856690383578,
+      "learning_rate": 8.04073167530485e-06,
+      "loss": 0.369,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3133281371784879,
+      "grad_norm": 0.6574172757735359,
+      "learning_rate": 8.039676807928073e-06,
+      "loss": 0.3467,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3134101817286787,
+      "grad_norm": 0.5956322600090069,
+      "learning_rate": 8.038621725894209e-06,
+      "loss": 0.3587,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3134922262788694,
+      "grad_norm": 0.6224785116937196,
+      "learning_rate": 8.037566429277757e-06,
+      "loss": 0.3415,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3135742708290602,
+      "grad_norm": 0.6359340468732328,
+      "learning_rate": 8.036510918153244e-06,
+      "loss": 0.3622,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3136563153792509,
+      "grad_norm": 0.6230590231680748,
+      "learning_rate": 8.035455192595209e-06,
+      "loss": 0.3157,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3137383599294417,
+      "grad_norm": 0.6290988356913858,
+      "learning_rate": 8.034399252678207e-06,
+      "loss": 0.3419,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31382040447963244,
+      "grad_norm": 0.6246612551505695,
+      "learning_rate": 8.033343098476802e-06,
+      "loss": 0.337,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31390244902982317,
+      "grad_norm": 0.6795281513406113,
+      "learning_rate": 8.032286730065582e-06,
+      "loss": 0.3738,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31398449358001396,
+      "grad_norm": 0.6818992681653913,
+      "learning_rate": 8.031230147519145e-06,
+      "loss": 0.3629,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3140665381302047,
+      "grad_norm": 0.6546878633104358,
+      "learning_rate": 8.030173350912105e-06,
+      "loss": 0.3467,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31414858268039547,
+      "grad_norm": 0.5899564776507595,
+      "learning_rate": 8.029116340319091e-06,
+      "loss": 0.3624,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3142306272305862,
+      "grad_norm": 0.5992658169904254,
+      "learning_rate": 8.028059115814749e-06,
+      "loss": 0.3347,
+      "step": 3830
+    },
+    {
+      "epoch": 0.314312671780777,
+      "grad_norm": 0.6803001263093066,
+      "learning_rate": 8.027001677473737e-06,
+      "loss": 0.4064,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3143947163309677,
+      "grad_norm": 0.6681520034845131,
+      "learning_rate": 8.025944025370728e-06,
+      "loss": 0.3475,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31447676088115845,
+      "grad_norm": 0.600463022313911,
+      "learning_rate": 8.024886159580414e-06,
+      "loss": 0.3547,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31455880543134923,
+      "grad_norm": 0.687896949499154,
+      "learning_rate": 8.023828080177502e-06,
+      "loss": 0.3577,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31464084998153996,
+      "grad_norm": 0.6882198647869333,
+      "learning_rate": 8.022769787236706e-06,
+      "loss": 0.42,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31472289453173075,
+      "grad_norm": 0.7998958883850691,
+      "learning_rate": 8.021711280832764e-06,
+      "loss": 0.3649,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3148049390819215,
+      "grad_norm": 2.232759765918411,
+      "learning_rate": 8.02065256104043e-06,
+      "loss": 0.3397,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31488698363211226,
+      "grad_norm": 0.6549891573351758,
+      "learning_rate": 8.019593627934461e-06,
+      "loss": 0.3586,
+      "step": 3838
+    },
+    {
+      "epoch": 0.314969028182303,
+      "grad_norm": 0.7309199502236531,
+      "learning_rate": 8.018534481589641e-06,
+      "loss": 0.3984,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3150510727324937,
+      "grad_norm": 0.6345951053777079,
+      "learning_rate": 8.017475122080767e-06,
+      "loss": 0.3519,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3151331172826845,
+      "grad_norm": 0.6259662674338453,
+      "learning_rate": 8.016415549482649e-06,
+      "loss": 0.3293,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31521516183287523,
+      "grad_norm": 0.675270323793165,
+      "learning_rate": 8.015355763870109e-06,
+      "loss": 0.371,
+      "step": 3842
+    },
+    {
+      "epoch": 0.315297206383066,
+      "grad_norm": 0.7594717595707051,
+      "learning_rate": 8.014295765317992e-06,
+      "loss": 0.6751,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31537925093325675,
+      "grad_norm": 0.6816959317020318,
+      "learning_rate": 8.013235553901149e-06,
+      "loss": 0.3725,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31546129548344753,
+      "grad_norm": 0.6515114770327648,
+      "learning_rate": 8.012175129694454e-06,
+      "loss": 0.3706,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31554334003363826,
+      "grad_norm": 0.6635054714850697,
+      "learning_rate": 8.01111449277279e-06,
+      "loss": 0.3641,
+      "step": 3846
+    },
+    {
+      "epoch": 0.315625384583829,
+      "grad_norm": 0.6145524251649618,
+      "learning_rate": 8.010053643211056e-06,
+      "loss": 0.3571,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3157074291340198,
+      "grad_norm": 0.7002745133826439,
+      "learning_rate": 8.008992581084172e-06,
+      "loss": 0.405,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.7038948568120698,
+      "learning_rate": 8.007931306467066e-06,
+      "loss": 0.397,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3158715182344013,
+      "grad_norm": 0.5908020036463014,
+      "learning_rate": 8.006869819434684e-06,
+      "loss": 0.3299,
+      "step": 3850
+    },
+    {
+      "epoch": 0.315953562784592,
+      "grad_norm": 0.6706173909670387,
+      "learning_rate": 8.005808120061986e-06,
+      "loss": 0.3188,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3160356073347828,
+      "grad_norm": 0.6428900123328335,
+      "learning_rate": 8.004746208423948e-06,
+      "loss": 0.3235,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31611765188497354,
+      "grad_norm": 0.6461992661214845,
+      "learning_rate": 8.003684084595558e-06,
+      "loss": 0.3747,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31619969643516427,
+      "grad_norm": 0.6195166271713635,
+      "learning_rate": 8.002621748651827e-06,
+      "loss": 0.3651,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31628174098535505,
+      "grad_norm": 0.6760714440685509,
+      "learning_rate": 8.00155920066777e-06,
+      "loss": 0.3224,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3163637855355458,
+      "grad_norm": 0.6483215539218088,
+      "learning_rate": 8.000496440718427e-06,
+      "loss": 0.3621,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31644583008573657,
+      "grad_norm": 0.6982664926721688,
+      "learning_rate": 7.999433468878844e-06,
+      "loss": 0.3465,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3165278746359273,
+      "grad_norm": 0.6114354151398582,
+      "learning_rate": 7.998370285224087e-06,
+      "loss": 0.3568,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3166099191861181,
+      "grad_norm": 0.6678153007574715,
+      "learning_rate": 7.99730688982924e-06,
+      "loss": 0.4011,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3166919637363088,
+      "grad_norm": 0.6299205262030573,
+      "learning_rate": 7.996243282769395e-06,
+      "loss": 0.3659,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31677400828649954,
+      "grad_norm": 0.6304206408387993,
+      "learning_rate": 7.995179464119664e-06,
+      "loss": 0.3681,
+      "step": 3861
+    },
+    {
+      "epoch": 0.3168560528366903,
+      "grad_norm": 0.6683101221777121,
+      "learning_rate": 7.994115433955172e-06,
+      "loss": 0.3705,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31693809738688106,
+      "grad_norm": 0.7072235990484781,
+      "learning_rate": 7.993051192351056e-06,
+      "loss": 0.354,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31702014193707184,
+      "grad_norm": 0.6030812241140354,
+      "learning_rate": 7.991986739382475e-06,
+      "loss": 0.3551,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31710218648726257,
+      "grad_norm": 0.5870045189766122,
+      "learning_rate": 7.990922075124599e-06,
+      "loss": 0.3355,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31718423103745336,
+      "grad_norm": 0.7266934047669696,
+      "learning_rate": 7.989857199652608e-06,
+      "loss": 0.3762,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3172662755876441,
+      "grad_norm": 0.6124220172321695,
+      "learning_rate": 7.988792113041707e-06,
+      "loss": 0.3386,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31734832013783487,
+      "grad_norm": 0.6461141768756277,
+      "learning_rate": 7.98772681536711e-06,
+      "loss": 0.3061,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3174303646880256,
+      "grad_norm": 0.6401013453169033,
+      "learning_rate": 7.986661306704045e-06,
+      "loss": 0.3657,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31751240923821633,
+      "grad_norm": 0.68054047891886,
+      "learning_rate": 7.985595587127755e-06,
+      "loss": 0.4089,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3175944537884071,
+      "grad_norm": 0.6576787243461274,
+      "learning_rate": 7.984529656713503e-06,
+      "loss": 0.3899,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31767649833859785,
+      "grad_norm": 0.6481058653469716,
+      "learning_rate": 7.983463515536562e-06,
+      "loss": 0.3569,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31775854288878863,
+      "grad_norm": 0.720920466891122,
+      "learning_rate": 7.982397163672219e-06,
+      "loss": 0.3876,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31784058743897936,
+      "grad_norm": 0.6564787598634939,
+      "learning_rate": 7.981330601195781e-06,
+      "loss": 0.353,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31792263198917015,
+      "grad_norm": 0.5570633336826369,
+      "learning_rate": 7.980263828182565e-06,
+      "loss": 0.32,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3180046765393609,
+      "grad_norm": 0.5964074854975797,
+      "learning_rate": 7.979196844707907e-06,
+      "loss": 0.4008,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3180867210895516,
+      "grad_norm": 0.621998611967159,
+      "learning_rate": 7.978129650847154e-06,
+      "loss": 0.3723,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3181687656397424,
+      "grad_norm": 0.6144461491455195,
+      "learning_rate": 7.977062246675669e-06,
+      "loss": 0.2993,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3182508101899331,
+      "grad_norm": 0.6643811657342329,
+      "learning_rate": 7.975994632268833e-06,
+      "loss": 0.3477,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3183328547401239,
+      "grad_norm": 0.6082270234265359,
+      "learning_rate": 7.974926807702034e-06,
+      "loss": 0.3331,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31841489929031463,
+      "grad_norm": 0.6153515968846742,
+      "learning_rate": 7.973858773050686e-06,
+      "loss": 0.334,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3184969438405054,
+      "grad_norm": 0.6608487269506883,
+      "learning_rate": 7.972790528390208e-06,
+      "loss": 0.3638,
+      "step": 3882
+    },
+    {
+      "epoch": 0.31857898839069615,
+      "grad_norm": 0.666448065729024,
+      "learning_rate": 7.971722073796042e-06,
+      "loss": 0.3578,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3186610329408869,
+      "grad_norm": 0.6552916247319747,
+      "learning_rate": 7.970653409343633e-06,
+      "loss": 0.361,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31874307749107766,
+      "grad_norm": 0.6335559448652389,
+      "learning_rate": 7.969584535108457e-06,
+      "loss": 0.3348,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3188251220412684,
+      "grad_norm": 0.6502294675694414,
+      "learning_rate": 7.96851545116599e-06,
+      "loss": 0.3577,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3189071665914592,
+      "grad_norm": 0.6053667198179379,
+      "learning_rate": 7.967446157591732e-06,
+      "loss": 0.395,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3189892111416499,
+      "grad_norm": 0.6506269232638298,
+      "learning_rate": 7.966376654461192e-06,
+      "loss": 0.3986,
+      "step": 3888
+    },
+    {
+      "epoch": 0.3190712556918407,
+      "grad_norm": 0.6162088358801006,
+      "learning_rate": 7.965306941849901e-06,
+      "loss": 0.358,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3191533002420314,
+      "grad_norm": 0.630536779916773,
+      "learning_rate": 7.964237019833397e-06,
+      "loss": 0.3686,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31923534479222215,
+      "grad_norm": 0.6663980459815173,
+      "learning_rate": 7.963166888487237e-06,
+      "loss": 0.3995,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31931738934241294,
+      "grad_norm": 0.6279850895461475,
+      "learning_rate": 7.962096547886992e-06,
+      "loss": 0.3262,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31939943389260367,
+      "grad_norm": 0.6623162330401523,
+      "learning_rate": 7.961025998108247e-06,
+      "loss": 0.383,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31948147844279445,
+      "grad_norm": 0.6641405726539766,
+      "learning_rate": 7.959955239226604e-06,
+      "loss": 0.4196,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3195635229929852,
+      "grad_norm": 0.6153395383701022,
+      "learning_rate": 7.958884271317678e-06,
+      "loss": 0.3412,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31964556754317597,
+      "grad_norm": 0.641901559617204,
+      "learning_rate": 7.957813094457098e-06,
+      "loss": 0.3681,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3197276120933667,
+      "grad_norm": 0.5966580130589388,
+      "learning_rate": 7.956741708720507e-06,
+      "loss": 0.3541,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3198096566435574,
+      "grad_norm": 0.60311918378115,
+      "learning_rate": 7.955670114183572e-06,
+      "loss": 0.3703,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3198917011937482,
+      "grad_norm": 0.6177892301247501,
+      "learning_rate": 7.954598310921959e-06,
+      "loss": 0.3696,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31997374574393894,
+      "grad_norm": 0.6829771448698114,
+      "learning_rate": 7.95352629901136e-06,
+      "loss": 0.3844,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3200557902941297,
+      "grad_norm": 0.6275109024662275,
+      "learning_rate": 7.95245407852748e-06,
+      "loss": 0.3862,
+      "step": 3901
+    },
+    {
+      "epoch": 0.32013783484432046,
+      "grad_norm": 0.6787389459321248,
+      "learning_rate": 7.951381649546036e-06,
+      "loss": 0.2989,
+      "step": 3902
+    },
+    {
+      "epoch": 0.32021987939451124,
+      "grad_norm": 0.678526754782796,
+      "learning_rate": 7.950309012142763e-06,
+      "loss": 0.3774,
+      "step": 3903
+    },
+    {
+      "epoch": 0.32030192394470197,
+      "grad_norm": 0.6307138439151928,
+      "learning_rate": 7.949236166393405e-06,
+      "loss": 0.3743,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3203839684948927,
+      "grad_norm": 0.6175280852850934,
+      "learning_rate": 7.948163112373728e-06,
+      "loss": 0.3558,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3204660130450835,
+      "grad_norm": 0.6316745226954891,
+      "learning_rate": 7.94708985015951e-06,
+      "loss": 0.3645,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3205480575952742,
+      "grad_norm": 0.6482204479996003,
+      "learning_rate": 7.946016379826542e-06,
+      "loss": 0.4113,
+      "step": 3907
+    },
+    {
+      "epoch": 0.320630102145465,
+      "grad_norm": 0.6192578593041019,
+      "learning_rate": 7.944942701450629e-06,
+      "loss": 0.3799,
+      "step": 3908
+    },
+    {
+      "epoch": 0.32071214669565573,
+      "grad_norm": 0.5493903776218557,
+      "learning_rate": 7.943868815107594e-06,
+      "loss": 0.311,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3207941912458465,
+      "grad_norm": 0.6638494056051444,
+      "learning_rate": 7.942794720873273e-06,
+      "loss": 0.3777,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32087623579603725,
+      "grad_norm": 0.657238398184785,
+      "learning_rate": 7.941720418823517e-06,
+      "loss": 0.3832,
+      "step": 3911
+    },
+    {
+      "epoch": 0.320958280346228,
+      "grad_norm": 0.613085975305512,
+      "learning_rate": 7.94064590903419e-06,
+      "loss": 0.3209,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32104032489641876,
+      "grad_norm": 0.6096749809994939,
+      "learning_rate": 7.939571191581175e-06,
+      "loss": 0.3504,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3211223694466095,
+      "grad_norm": 0.6712996160730604,
+      "learning_rate": 7.938496266540365e-06,
+      "loss": 0.3715,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3212044139968003,
+      "grad_norm": 0.6355582959568326,
+      "learning_rate": 7.937421133987668e-06,
+      "loss": 0.3424,
+      "step": 3915
+    },
+    {
+      "epoch": 0.321286458546991,
+      "grad_norm": 0.5959060926435421,
+      "learning_rate": 7.936345793999012e-06,
+      "loss": 0.3256,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3213685030971818,
+      "grad_norm": 0.6116077434544214,
+      "learning_rate": 7.935270246650332e-06,
+      "loss": 0.3531,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3214505476473725,
+      "grad_norm": 0.5896952478749731,
+      "learning_rate": 7.934194492017586e-06,
+      "loss": 0.3492,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32153259219756325,
+      "grad_norm": 0.5928023536210993,
+      "learning_rate": 7.933118530176737e-06,
+      "loss": 0.3496,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32161463674775403,
+      "grad_norm": 0.6142095992498606,
+      "learning_rate": 7.93204236120377e-06,
+      "loss": 0.3371,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32169668129794476,
+      "grad_norm": 0.5970911669404029,
+      "learning_rate": 7.930965985174682e-06,
+      "loss": 0.3442,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32177872584813555,
+      "grad_norm": 0.6333611954280999,
+      "learning_rate": 7.929889402165485e-06,
+      "loss": 0.3437,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3218607703983263,
+      "grad_norm": 0.668037308259036,
+      "learning_rate": 7.928812612252206e-06,
+      "loss": 0.3413,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32194281494851706,
+      "grad_norm": 0.6198391666417926,
+      "learning_rate": 7.927735615510885e-06,
+      "loss": 0.3825,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3220248594987078,
+      "grad_norm": 0.5677706508354115,
+      "learning_rate": 7.926658412017578e-06,
+      "loss": 0.3455,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3221069040488986,
+      "grad_norm": 0.6068829409841491,
+      "learning_rate": 7.925581001848359e-06,
+      "loss": 0.3438,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3221889485990893,
+      "grad_norm": 0.5849193731907085,
+      "learning_rate": 7.924503385079308e-06,
+      "loss": 0.3759,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32227099314928004,
+      "grad_norm": 0.6251760323296239,
+      "learning_rate": 7.923425561786526e-06,
+      "loss": 0.3561,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3223530376994708,
+      "grad_norm": 0.6302712915761736,
+      "learning_rate": 7.922347532046128e-06,
+      "loss": 0.3453,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32243508224966155,
+      "grad_norm": 0.615151293597076,
+      "learning_rate": 7.921269295934243e-06,
+      "loss": 0.35,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32251712679985234,
+      "grad_norm": 0.6171327380729658,
+      "learning_rate": 7.920190853527015e-06,
+      "loss": 0.3537,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32259917135004307,
+      "grad_norm": 0.6034391738772455,
+      "learning_rate": 7.919112204900597e-06,
+      "loss": 0.365,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32268121590023385,
+      "grad_norm": 0.6325137612226737,
+      "learning_rate": 7.918033350131167e-06,
+      "loss": 0.3327,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3227632604504246,
+      "grad_norm": 0.6159699568866887,
+      "learning_rate": 7.916954289294911e-06,
+      "loss": 0.3503,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3228453050006153,
+      "grad_norm": 0.6142462815445192,
+      "learning_rate": 7.915875022468028e-06,
+      "loss": 0.3287,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3229273495508061,
+      "grad_norm": 0.6512363324862457,
+      "learning_rate": 7.914795549726736e-06,
+      "loss": 0.322,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3230093941009968,
+      "grad_norm": 0.6393467987430286,
+      "learning_rate": 7.913715871147265e-06,
+      "loss": 0.3492,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3230914386511876,
+      "grad_norm": 0.6283074626994584,
+      "learning_rate": 7.91263598680586e-06,
+      "loss": 0.3404,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32317348320137834,
+      "grad_norm": 0.6387035459125884,
+      "learning_rate": 7.911555896778783e-06,
+      "loss": 0.3643,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3232555277515691,
+      "grad_norm": 0.6354296279015863,
+      "learning_rate": 7.910475601142305e-06,
+      "loss": 0.337,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32333757230175986,
+      "grad_norm": 0.6532515328017992,
+      "learning_rate": 7.909395099972716e-06,
+      "loss": 0.3715,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3234196168519506,
+      "grad_norm": 0.622138297593448,
+      "learning_rate": 7.90831439334632e-06,
+      "loss": 0.4139,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32350166140214137,
+      "grad_norm": 0.630292571790943,
+      "learning_rate": 7.907233481339433e-06,
+      "loss": 0.3879,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3235837059523321,
+      "grad_norm": 0.5981378479419311,
+      "learning_rate": 7.90615236402839e-06,
+      "loss": 0.3535,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3236657505025229,
+      "grad_norm": 0.6188383266273982,
+      "learning_rate": 7.905071041489536e-06,
+      "loss": 0.3324,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3237477950527136,
+      "grad_norm": 0.67888223330206,
+      "learning_rate": 7.903989513799231e-06,
+      "loss": 0.3721,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3238298396029044,
+      "grad_norm": 0.6591709735704394,
+      "learning_rate": 7.902907781033852e-06,
+      "loss": 0.3288,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32391188415309513,
+      "grad_norm": 0.7320782829009737,
+      "learning_rate": 7.901825843269791e-06,
+      "loss": 0.3832,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32399392870328586,
+      "grad_norm": 0.6374901607662895,
+      "learning_rate": 7.90074370058345e-06,
+      "loss": 0.3644,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32407597325347665,
+      "grad_norm": 0.6755055834030602,
+      "learning_rate": 7.89966135305125e-06,
+      "loss": 0.3264,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3241580178036674,
+      "grad_norm": 0.6265540477858386,
+      "learning_rate": 7.898578800749623e-06,
+      "loss": 0.3506,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32424006235385816,
+      "grad_norm": 0.6594968579770331,
+      "learning_rate": 7.89749604375502e-06,
+      "loss": 0.3591,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3243221069040489,
+      "grad_norm": 0.7697289669851164,
+      "learning_rate": 7.896413082143898e-06,
+      "loss": 0.3501,
+      "step": 3953
+    },
+    {
+      "epoch": 0.3244041514542397,
+      "grad_norm": 0.6623318197396063,
+      "learning_rate": 7.89532991599274e-06,
+      "loss": 0.351,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3244861960044304,
+      "grad_norm": 0.5904079543831577,
+      "learning_rate": 7.894246545378037e-06,
+      "loss": 0.3473,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32456824055462113,
+      "grad_norm": 0.6185702569174376,
+      "learning_rate": 7.89316297037629e-06,
+      "loss": 0.3656,
+      "step": 3956
+    },
+    {
+      "epoch": 0.3246502851048119,
+      "grad_norm": 0.6074512031440279,
+      "learning_rate": 7.892079191064022e-06,
+      "loss": 0.3511,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32473232965500265,
+      "grad_norm": 0.6520430268789338,
+      "learning_rate": 7.89099520751777e-06,
+      "loss": 0.3837,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32481437420519343,
+      "grad_norm": 0.5914351613926238,
+      "learning_rate": 7.88991101981408e-06,
+      "loss": 0.3539,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32489641875538416,
+      "grad_norm": 0.6579085554748094,
+      "learning_rate": 7.888826628029515e-06,
+      "loss": 0.3488,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32497846330557495,
+      "grad_norm": 0.6413105503159083,
+      "learning_rate": 7.887742032240658e-06,
+      "loss": 0.3692,
+      "step": 3961
+    },
+    {
+      "epoch": 0.3250605078557657,
+      "grad_norm": 0.5844974551054252,
+      "learning_rate": 7.886657232524097e-06,
+      "loss": 0.3224,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3251425524059564,
+      "grad_norm": 0.6154527558402214,
+      "learning_rate": 7.885572228956439e-06,
+      "loss": 0.3554,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3252245969561472,
+      "grad_norm": 0.5936541588595473,
+      "learning_rate": 7.884487021614307e-06,
+      "loss": 0.3457,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3253066415063379,
+      "grad_norm": 0.6693585047527416,
+      "learning_rate": 7.883401610574338e-06,
+      "loss": 0.3615,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3253886860565287,
+      "grad_norm": 0.6224291688012054,
+      "learning_rate": 7.882315995913177e-06,
+      "loss": 0.3536,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32547073060671944,
+      "grad_norm": 0.6607569437566597,
+      "learning_rate": 7.88123017770749e-06,
+      "loss": 0.324,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3255527751569102,
+      "grad_norm": 0.6005491087793292,
+      "learning_rate": 7.88014415603396e-06,
+      "loss": 0.3587,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32563481970710095,
+      "grad_norm": 0.675947071963376,
+      "learning_rate": 7.879057930969275e-06,
+      "loss": 0.3933,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3257168642572917,
+      "grad_norm": 0.7883656661439067,
+      "learning_rate": 7.877971502590146e-06,
+      "loss": 0.3611,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32579890880748247,
+      "grad_norm": 0.6727351513975756,
+      "learning_rate": 7.876884870973292e-06,
+      "loss": 0.3327,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3258809533576732,
+      "grad_norm": 0.6800786748191792,
+      "learning_rate": 7.875798036195452e-06,
+      "loss": 0.3465,
+      "step": 3972
+    },
+    {
+      "epoch": 0.325962997907864,
+      "grad_norm": 0.5973085298349345,
+      "learning_rate": 7.874710998333374e-06,
+      "loss": 0.3507,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3260450424580547,
+      "grad_norm": 0.5858483849414303,
+      "learning_rate": 7.873623757463823e-06,
+      "loss": 0.3184,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3261270870082455,
+      "grad_norm": 0.6881163131693194,
+      "learning_rate": 7.87253631366358e-06,
+      "loss": 0.3642,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3262091315584362,
+      "grad_norm": 0.6488852563146476,
+      "learning_rate": 7.871448667009439e-06,
+      "loss": 0.3553,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32629117610862696,
+      "grad_norm": 0.6228542435813113,
+      "learning_rate": 7.870360817578204e-06,
+      "loss": 0.3338,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32637322065881774,
+      "grad_norm": 0.6572146851318225,
+      "learning_rate": 7.869272765446701e-06,
+      "loss": 0.3881,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32645526520900847,
+      "grad_norm": 0.6616244550715745,
+      "learning_rate": 7.868184510691764e-06,
+      "loss": 0.3479,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32653730975919926,
+      "grad_norm": 0.6349049333045897,
+      "learning_rate": 7.867096053390245e-06,
+      "loss": 0.3682,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32661935430939,
+      "grad_norm": 0.6403071683367509,
+      "learning_rate": 7.86600739361901e-06,
+      "loss": 0.4019,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32670139885958077,
+      "grad_norm": 0.6607855779424673,
+      "learning_rate": 7.864918531454938e-06,
+      "loss": 0.3528,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3267834434097715,
+      "grad_norm": 0.6601597929430983,
+      "learning_rate": 7.863829466974922e-06,
+      "loss": 0.3394,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3268654879599623,
+      "grad_norm": 0.6296103185973102,
+      "learning_rate": 7.86274020025587e-06,
+      "loss": 0.3424,
+      "step": 3984
+    },
+    {
+      "epoch": 0.326947532510153,
+      "grad_norm": 0.7159922709197738,
+      "learning_rate": 7.861650731374704e-06,
+      "loss": 0.3425,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32702957706034375,
+      "grad_norm": 0.6561541873508885,
+      "learning_rate": 7.860561060408361e-06,
+      "loss": 0.3573,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32711162161053453,
+      "grad_norm": 0.5897368470674484,
+      "learning_rate": 7.85947118743379e-06,
+      "loss": 0.3192,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32719366616072526,
+      "grad_norm": 0.6279699200658293,
+      "learning_rate": 7.85838111252796e-06,
+      "loss": 0.3378,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32727571071091605,
+      "grad_norm": 0.6824885757754671,
+      "learning_rate": 7.857290835767847e-06,
+      "loss": 0.3511,
+      "step": 3989
+    },
+    {
+      "epoch": 0.3273577552611068,
+      "grad_norm": 0.671245271637452,
+      "learning_rate": 7.856200357230448e-06,
+      "loss": 0.331,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32743979981129756,
+      "grad_norm": 0.6078551324477071,
+      "learning_rate": 7.855109676992767e-06,
+      "loss": 0.39,
+      "step": 3991
+    },
+    {
+      "epoch": 0.3275218443614883,
+      "grad_norm": 0.5885925827695659,
+      "learning_rate": 7.854018795131827e-06,
+      "loss": 0.3096,
+      "step": 3992
+    },
+    {
+      "epoch": 0.327603888911679,
+      "grad_norm": 0.7341562081010585,
+      "learning_rate": 7.852927711724665e-06,
+      "loss": 0.3459,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3276859334618698,
+      "grad_norm": 0.63300688310773,
+      "learning_rate": 7.851836426848334e-06,
+      "loss": 0.3648,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32776797801206053,
+      "grad_norm": 0.632139299535845,
+      "learning_rate": 7.850744940579894e-06,
+      "loss": 0.3589,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3278500225622513,
+      "grad_norm": 0.6821673304043518,
+      "learning_rate": 7.849653252996426e-06,
+      "loss": 0.3908,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32793206711244205,
+      "grad_norm": 0.6075171025690768,
+      "learning_rate": 7.848561364175024e-06,
+      "loss": 0.3577,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32801411166263283,
+      "grad_norm": 0.6932375966761548,
+      "learning_rate": 7.847469274192797e-06,
+      "loss": 0.3702,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32809615621282356,
+      "grad_norm": 0.653687992792476,
+      "learning_rate": 7.846376983126864e-06,
+      "loss": 0.3884,
+      "step": 3999
+    },
+    {
+      "epoch": 0.3281782007630143,
+      "grad_norm": 0.6518183150477391,
+      "learning_rate": 7.845284491054362e-06,
+      "loss": 0.3588,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3282602453132051,
+      "grad_norm": 0.6375284604467827,
+      "learning_rate": 7.844191798052438e-06,
+      "loss": 0.3371,
+      "step": 4001
+    },
+    {
+      "epoch": 0.3283422898633958,
+      "grad_norm": 0.6265765863855154,
+      "learning_rate": 7.84309890419826e-06,
+      "loss": 0.3363,
+      "step": 4002
+    },
+    {
+      "epoch": 0.3284243344135866,
+      "grad_norm": 0.6280739753315693,
+      "learning_rate": 7.842005809569006e-06,
+      "loss": 0.3869,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3285063789637773,
+      "grad_norm": 0.6228477258996614,
+      "learning_rate": 7.840912514241867e-06,
+      "loss": 0.3304,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3285884235139681,
+      "grad_norm": 0.6045811158920567,
+      "learning_rate": 7.83981901829405e-06,
+      "loss": 0.3623,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32867046806415884,
+      "grad_norm": 0.602334073140482,
+      "learning_rate": 7.838725321802777e-06,
+      "loss": 0.3749,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32875251261434957,
+      "grad_norm": 0.6356108172209286,
+      "learning_rate": 7.837631424845283e-06,
+      "loss": 0.3639,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32883455716454035,
+      "grad_norm": 0.586167161747632,
+      "learning_rate": 7.836537327498816e-06,
+      "loss": 0.3501,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3289166017147311,
+      "grad_norm": 0.5955764241018714,
+      "learning_rate": 7.835443029840641e-06,
+      "loss": 0.3478,
+      "step": 4009
+    },
+    {
+      "epoch": 0.32899864626492187,
+      "grad_norm": 0.6310876489378392,
+      "learning_rate": 7.834348531948033e-06,
+      "loss": 0.3623,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3290806908151126,
+      "grad_norm": 0.6013697302010771,
+      "learning_rate": 7.833253833898287e-06,
+      "loss": 0.335,
+      "step": 4011
+    },
+    {
+      "epoch": 0.3291627353653034,
+      "grad_norm": 0.6137811050045451,
+      "learning_rate": 7.832158935768708e-06,
+      "loss": 0.3763,
+      "step": 4012
+    },
+    {
+      "epoch": 0.3292447799154941,
+      "grad_norm": 0.5985477812643535,
+      "learning_rate": 7.831063837636613e-06,
+      "loss": 0.3547,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32932682446568484,
+      "grad_norm": 0.7077727514563682,
+      "learning_rate": 7.82996853957934e-06,
+      "loss": 0.3409,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3294088690158756,
+      "grad_norm": 0.6619767304476943,
+      "learning_rate": 7.828873041674234e-06,
+      "loss": 0.341,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32949091356606636,
+      "grad_norm": 0.7033161918899659,
+      "learning_rate": 7.827777343998659e-06,
+      "loss": 0.3627,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32957295811625714,
+      "grad_norm": 0.6036607667061773,
+      "learning_rate": 7.826681446629994e-06,
+      "loss": 0.3629,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32965500266644787,
+      "grad_norm": 0.612501694517382,
+      "learning_rate": 7.825585349645623e-06,
+      "loss": 0.3247,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32973704721663866,
+      "grad_norm": 0.5988654229207765,
+      "learning_rate": 7.824489053122955e-06,
+      "loss": 0.3841,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3298190917668294,
+      "grad_norm": 0.6305634578819034,
+      "learning_rate": 7.82339255713941e-06,
+      "loss": 0.3651,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3299011363170201,
+      "grad_norm": 0.6637110888019846,
+      "learning_rate": 7.822295861772416e-06,
+      "loss": 0.3804,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3299831808672109,
+      "grad_norm": 0.6562247816091767,
+      "learning_rate": 7.821198967099424e-06,
+      "loss": 0.3629,
+      "step": 4022
+    },
+    {
+      "epoch": 0.33006522541740163,
+      "grad_norm": 0.6088352522734316,
+      "learning_rate": 7.820101873197893e-06,
+      "loss": 0.3293,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3301472699675924,
+      "grad_norm": 0.6499286980283207,
+      "learning_rate": 7.819004580145298e-06,
+      "loss": 0.3341,
+      "step": 4024
+    },
+    {
+      "epoch": 0.33022931451778315,
+      "grad_norm": 0.7052735316645613,
+      "learning_rate": 7.81790708801913e-06,
+      "loss": 0.3614,
+      "step": 4025
+    },
+    {
+      "epoch": 0.33031135906797393,
+      "grad_norm": 0.6467460278317205,
+      "learning_rate": 7.81680939689689e-06,
+      "loss": 0.3569,
+      "step": 4026
+    },
+    {
+      "epoch": 0.33039340361816466,
+      "grad_norm": 0.6204054104661068,
+      "learning_rate": 7.815711506856097e-06,
+      "loss": 0.3707,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3304754481683554,
+      "grad_norm": 0.6090689313770575,
+      "learning_rate": 7.81461341797428e-06,
+      "loss": 0.344,
+      "step": 4028
+    },
+    {
+      "epoch": 0.3305574927185462,
+      "grad_norm": 0.7342942312196887,
+      "learning_rate": 7.813515130328985e-06,
+      "loss": 0.3664,
+      "step": 4029
+    },
+    {
+      "epoch": 0.3306395372687369,
+      "grad_norm": 0.657810736799398,
+      "learning_rate": 7.812416643997773e-06,
+      "loss": 0.3671,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3307215818189277,
+      "grad_norm": 0.639405861422662,
+      "learning_rate": 7.811317959058217e-06,
+      "loss": 0.3305,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3308036263691184,
+      "grad_norm": 0.6006220226280254,
+      "learning_rate": 7.810219075587901e-06,
+      "loss": 0.3788,
+      "step": 4032
+    },
+    {
+      "epoch": 0.3308856709193092,
+      "grad_norm": 0.6099278227892865,
+      "learning_rate": 7.80911999366443e-06,
+      "loss": 0.332,
+      "step": 4033
+    },
+    {
+      "epoch": 0.33096771546949993,
+      "grad_norm": 0.6408729358525758,
+      "learning_rate": 7.808020713365419e-06,
+      "loss": 0.3415,
+      "step": 4034
+    },
+    {
+      "epoch": 0.3310497600196907,
+      "grad_norm": 0.6436575764418772,
+      "learning_rate": 7.806921234768497e-06,
+      "loss": 0.3057,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33113180456988145,
+      "grad_norm": 0.6303219447073176,
+      "learning_rate": 7.805821557951305e-06,
+      "loss": 0.3736,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3312138491200722,
+      "grad_norm": 0.6990575678022085,
+      "learning_rate": 7.804721682991503e-06,
+      "loss": 0.3513,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33129589367026296,
+      "grad_norm": 0.6149638125793881,
+      "learning_rate": 7.803621609966762e-06,
+      "loss": 0.3205,
+      "step": 4038
+    },
+    {
+      "epoch": 0.3313779382204537,
+      "grad_norm": 0.6345414947727697,
+      "learning_rate": 7.802521338954767e-06,
+      "loss": 0.3986,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3314599827706445,
+      "grad_norm": 0.6178439259006434,
+      "learning_rate": 7.801420870033218e-06,
+      "loss": 0.3485,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3315420273208352,
+      "grad_norm": 0.6558184768911091,
+      "learning_rate": 7.800320203279829e-06,
+      "loss": 0.3457,
+      "step": 4041
+    },
+    {
+      "epoch": 0.331624071871026,
+      "grad_norm": 0.6082820914053809,
+      "learning_rate": 7.799219338772327e-06,
+      "loss": 0.3359,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3317061164212167,
+      "grad_norm": 0.6546210030911372,
+      "learning_rate": 7.79811827658845e-06,
+      "loss": 0.3854,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33178816097140745,
+      "grad_norm": 0.616362556344265,
+      "learning_rate": 7.797017016805957e-06,
+      "loss": 0.3901,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33187020552159824,
+      "grad_norm": 0.6166494035329391,
+      "learning_rate": 7.795915559502616e-06,
+      "loss": 0.3417,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33195225007178897,
+      "grad_norm": 0.6185169507440403,
+      "learning_rate": 7.794813904756209e-06,
+      "loss": 0.3257,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33203429462197975,
+      "grad_norm": 0.599565054802429,
+      "learning_rate": 7.793712052644535e-06,
+      "loss": 0.3504,
+      "step": 4047
+    },
+    {
+      "epoch": 0.3321163391721705,
+      "grad_norm": 0.6340181327639354,
+      "learning_rate": 7.792610003245406e-06,
+      "loss": 0.359,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33219838372236127,
+      "grad_norm": 0.6888545278701614,
+      "learning_rate": 7.791507756636643e-06,
+      "loss": 0.3843,
+      "step": 4049
+    },
+    {
+      "epoch": 0.332280428272552,
+      "grad_norm": 0.6355573283236872,
+      "learning_rate": 7.790405312896088e-06,
+      "loss": 0.4168,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3323624728227427,
+      "grad_norm": 0.6484295051338416,
+      "learning_rate": 7.789302672101593e-06,
+      "loss": 0.374,
+      "step": 4051
+    },
+    {
+      "epoch": 0.3324445173729335,
+      "grad_norm": 0.6278443509416608,
+      "learning_rate": 7.788199834331026e-06,
+      "loss": 0.3776,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33252656192312424,
+      "grad_norm": 0.5815208244978146,
+      "learning_rate": 7.787096799662263e-06,
+      "loss": 0.3498,
+      "step": 4053
+    },
+    {
+      "epoch": 0.332608606473315,
+      "grad_norm": 0.6377875177696042,
+      "learning_rate": 7.785993568173205e-06,
+      "loss": 0.393,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33269065102350576,
+      "grad_norm": 0.5828083675328058,
+      "learning_rate": 7.784890139941755e-06,
+      "loss": 0.3443,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33277269557369654,
+      "grad_norm": 0.6476602545527373,
+      "learning_rate": 7.783786515045839e-06,
+      "loss": 0.3836,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33285474012388727,
+      "grad_norm": 0.6024946614834383,
+      "learning_rate": 7.782682693563391e-06,
+      "loss": 0.3328,
+      "step": 4057
+    },
+    {
+      "epoch": 0.332936784674078,
+      "grad_norm": 0.6640445670082776,
+      "learning_rate": 7.781578675572364e-06,
+      "loss": 0.3641,
+      "step": 4058
+    },
+    {
+      "epoch": 0.3330188292242688,
+      "grad_norm": 0.6438062692831585,
+      "learning_rate": 7.780474461150715e-06,
+      "loss": 0.368,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3331008737744595,
+      "grad_norm": 0.6562827502869143,
+      "learning_rate": 7.77937005037643e-06,
+      "loss": 0.3574,
+      "step": 4060
+    },
+    {
+      "epoch": 0.3331829183246503,
+      "grad_norm": 0.6530491319729351,
+      "learning_rate": 7.7782654433275e-06,
+      "loss": 0.3301,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33326496287484103,
+      "grad_norm": 0.6147073991186335,
+      "learning_rate": 7.777160640081925e-06,
+      "loss": 0.3093,
+      "step": 4062
+    },
+    {
+      "epoch": 0.3333470074250318,
+      "grad_norm": 0.6432917382446715,
+      "learning_rate": 7.776055640717727e-06,
+      "loss": 0.3753,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33342905197522255,
+      "grad_norm": 0.613698436362954,
+      "learning_rate": 7.77495044531294e-06,
+      "loss": 0.3178,
+      "step": 4064
+    },
+    {
+      "epoch": 0.3335110965254133,
+      "grad_norm": 0.645327953473634,
+      "learning_rate": 7.773845053945612e-06,
+      "loss": 0.3436,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33359314107560406,
+      "grad_norm": 0.6568562900111395,
+      "learning_rate": 7.772739466693801e-06,
+      "loss": 0.3525,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3336751856257948,
+      "grad_norm": 0.663047764817123,
+      "learning_rate": 7.771633683635586e-06,
+      "loss": 0.3735,
+      "step": 4067
+    },
+    {
+      "epoch": 0.3337572301759856,
+      "grad_norm": 0.6614313133898073,
+      "learning_rate": 7.770527704849052e-06,
+      "loss": 0.3498,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3338392747261763,
+      "grad_norm": 0.7726066030644256,
+      "learning_rate": 7.769421530412302e-06,
+      "loss": 0.3798,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3339213192763671,
+      "grad_norm": 0.6200755293614089,
+      "learning_rate": 7.768315160403453e-06,
+      "loss": 0.3466,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3340033638265578,
+      "grad_norm": 0.6208324990411024,
+      "learning_rate": 7.767208594900635e-06,
+      "loss": 0.3509,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33408540837674855,
+      "grad_norm": 0.6549414557634783,
+      "learning_rate": 7.766101833981994e-06,
+      "loss": 0.3378,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33416745292693933,
+      "grad_norm": 0.6385141172821575,
+      "learning_rate": 7.764994877725684e-06,
+      "loss": 0.3714,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33424949747713006,
+      "grad_norm": 0.6358442985048072,
+      "learning_rate": 7.763887726209877e-06,
+      "loss": 0.3369,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33433154202732085,
+      "grad_norm": 0.6683018011152858,
+      "learning_rate": 7.762780379512761e-06,
+      "loss": 0.37,
+      "step": 4075
+    },
+    {
+      "epoch": 0.3344135865775116,
+      "grad_norm": 0.6034171788126358,
+      "learning_rate": 7.761672837712533e-06,
+      "loss": 0.3279,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33449563112770236,
+      "grad_norm": 0.6232956628194204,
+      "learning_rate": 7.760565100887405e-06,
+      "loss": 0.3432,
+      "step": 4077
+    },
+    {
+      "epoch": 0.3345776756778931,
+      "grad_norm": 0.6731138592028492,
+      "learning_rate": 7.759457169115606e-06,
+      "loss": 0.3623,
+      "step": 4078
+    },
+    {
+      "epoch": 0.3346597202280838,
+      "grad_norm": 0.6312498985376971,
+      "learning_rate": 7.758349042475374e-06,
+      "loss": 0.341,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3347417647782746,
+      "grad_norm": 0.630887739453927,
+      "learning_rate": 7.757240721044967e-06,
+      "loss": 0.3406,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33482380932846534,
+      "grad_norm": 0.6639877117524499,
+      "learning_rate": 7.756132204902648e-06,
+      "loss": 0.3717,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3349058538786561,
+      "grad_norm": 0.6267533022672894,
+      "learning_rate": 7.7550234941267e-06,
+      "loss": 0.4127,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33498789842884685,
+      "grad_norm": 0.6539414151688071,
+      "learning_rate": 7.753914588795421e-06,
+      "loss": 0.3944,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33506994297903764,
+      "grad_norm": 0.6668430120363532,
+      "learning_rate": 7.752805488987116e-06,
+      "loss": 0.3831,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33515198752922837,
+      "grad_norm": 0.6295090830388266,
+      "learning_rate": 7.751696194780112e-06,
+      "loss": 0.364,
+      "step": 4085
+    },
+    {
+      "epoch": 0.3352340320794191,
+      "grad_norm": 0.589225661483559,
+      "learning_rate": 7.750586706252743e-06,
+      "loss": 0.3553,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3353160766296099,
+      "grad_norm": 0.6464329866054863,
+      "learning_rate": 7.749477023483361e-06,
+      "loss": 0.3762,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3353981211798006,
+      "grad_norm": 0.6765773449165166,
+      "learning_rate": 7.748367146550327e-06,
+      "loss": 0.3589,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3354801657299914,
+      "grad_norm": 0.6593691413743069,
+      "learning_rate": 7.747257075532022e-06,
+      "loss": 0.3731,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3355622102801821,
+      "grad_norm": 0.6520867097679971,
+      "learning_rate": 7.746146810506834e-06,
+      "loss": 0.3604,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3356442548303729,
+      "grad_norm": 0.5790589471127424,
+      "learning_rate": 7.745036351553172e-06,
+      "loss": 0.3081,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33572629938056364,
+      "grad_norm": 0.6418662798914643,
+      "learning_rate": 7.743925698749454e-06,
+      "loss": 0.345,
+      "step": 4092
+    },
+    {
+      "epoch": 0.3358083439307544,
+      "grad_norm": 0.6261020494174188,
+      "learning_rate": 7.742814852174112e-06,
+      "loss": 0.3472,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33589038848094516,
+      "grad_norm": 0.6709153771679117,
+      "learning_rate": 7.74170381190559e-06,
+      "loss": 0.3448,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3359724330311359,
+      "grad_norm": 0.6936349398173108,
+      "learning_rate": 7.74059257802235e-06,
+      "loss": 0.3752,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33605447758132667,
+      "grad_norm": 0.5967813711468822,
+      "learning_rate": 7.739481150602865e-06,
+      "loss": 0.328,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3361365221315174,
+      "grad_norm": 0.648522402259355,
+      "learning_rate": 7.738369529725624e-06,
+      "loss": 0.391,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3362185666817082,
+      "grad_norm": 0.6156427963440132,
+      "learning_rate": 7.737257715469124e-06,
+      "loss": 0.3164,
+      "step": 4098
+    },
+    {
+      "epoch": 0.3363006112318989,
+      "grad_norm": 0.6527680347219319,
+      "learning_rate": 7.736145707911884e-06,
+      "loss": 0.3758,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3363826557820897,
+      "grad_norm": 0.6290622617897808,
+      "learning_rate": 7.735033507132429e-06,
+      "loss": 0.3881,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33646470033228043,
+      "grad_norm": 0.6294631981361496,
+      "learning_rate": 7.733921113209302e-06,
+      "loss": 0.3054,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33654674488247116,
+      "grad_norm": 0.6418479001775654,
+      "learning_rate": 7.732808526221058e-06,
+      "loss": 0.3554,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33662878943266195,
+      "grad_norm": 0.5891680660121622,
+      "learning_rate": 7.731695746246268e-06,
+      "loss": 0.3461,
+      "step": 4103
+    },
+    {
+      "epoch": 0.3367108339828527,
+      "grad_norm": 0.6694743769865772,
+      "learning_rate": 7.730582773363511e-06,
+      "loss": 0.3717,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33679287853304346,
+      "grad_norm": 0.6848407776811216,
+      "learning_rate": 7.729469607651387e-06,
+      "loss": 0.3498,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3368749230832342,
+      "grad_norm": 0.7085383850063207,
+      "learning_rate": 7.728356249188504e-06,
+      "loss": 0.3621,
+      "step": 4106
+    },
+    {
+      "epoch": 0.336956967633425,
+      "grad_norm": 0.61248451804519,
+      "learning_rate": 7.727242698053488e-06,
+      "loss": 0.3564,
+      "step": 4107
+    },
+    {
+      "epoch": 0.3370390121836157,
+      "grad_norm": 0.6298196608028203,
+      "learning_rate": 7.726128954324971e-06,
+      "loss": 0.3324,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33712105673380643,
+      "grad_norm": 0.5492919697206993,
+      "learning_rate": 7.72501501808161e-06,
+      "loss": 0.32,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3372031012839972,
+      "grad_norm": 0.6137360893697085,
+      "learning_rate": 7.723900889402065e-06,
+      "loss": 0.366,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33728514583418795,
+      "grad_norm": 0.6873326873730783,
+      "learning_rate": 7.722786568365015e-06,
+      "loss": 0.3422,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33736719038437873,
+      "grad_norm": 0.6930406786073827,
+      "learning_rate": 7.721672055049153e-06,
+      "loss": 0.3658,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33744923493456946,
+      "grad_norm": 0.6476472333231771,
+      "learning_rate": 7.720557349533183e-06,
+      "loss": 0.3838,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33753127948476025,
+      "grad_norm": 0.6709359484316791,
+      "learning_rate": 7.719442451895824e-06,
+      "loss": 0.3708,
+      "step": 4114
+    },
+    {
+      "epoch": 0.337613324034951,
+      "grad_norm": 0.6476366832798613,
+      "learning_rate": 7.718327362215809e-06,
+      "loss": 0.3335,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3376953685851417,
+      "grad_norm": 0.6338101260424004,
+      "learning_rate": 7.71721208057188e-06,
+      "loss": 0.3734,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3377774131353325,
+      "grad_norm": 0.6085437108899204,
+      "learning_rate": 7.716096607042802e-06,
+      "loss": 0.3609,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3378594576855232,
+      "grad_norm": 0.6470512411841841,
+      "learning_rate": 7.714980941707345e-06,
+      "loss": 0.3793,
+      "step": 4118
+    },
+    {
+      "epoch": 0.337941502235714,
+      "grad_norm": 0.6156501974438027,
+      "learning_rate": 7.713865084644297e-06,
+      "loss": 0.3576,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33802354678590474,
+      "grad_norm": 0.6189923049066208,
+      "learning_rate": 7.712749035932454e-06,
+      "loss": 0.3717,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3381055913360955,
+      "grad_norm": 0.6072185223982953,
+      "learning_rate": 7.711632795650634e-06,
+      "loss": 0.3518,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33818763588628625,
+      "grad_norm": 0.6605101238515693,
+      "learning_rate": 7.710516363877662e-06,
+      "loss": 0.3452,
+      "step": 4122
+    },
+    {
+      "epoch": 0.338269680436477,
+      "grad_norm": 0.5995008222933018,
+      "learning_rate": 7.70939974069238e-06,
+      "loss": 0.3757,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33835172498666777,
+      "grad_norm": 0.6325787971708328,
+      "learning_rate": 7.70828292617364e-06,
+      "loss": 0.3451,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3384337695368585,
+      "grad_norm": 0.6303034516816596,
+      "learning_rate": 7.707165920400314e-06,
+      "loss": 0.3514,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3385158140870493,
+      "grad_norm": 0.6497728565504431,
+      "learning_rate": 7.706048723451276e-06,
+      "loss": 0.3717,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33859785863724,
+      "grad_norm": 0.622618222911655,
+      "learning_rate": 7.704931335405428e-06,
+      "loss": 0.3204,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3386799031874308,
+      "grad_norm": 0.6401061890837538,
+      "learning_rate": 7.703813756341673e-06,
+      "loss": 0.3722,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3387619477376215,
+      "grad_norm": 0.6710295571931958,
+      "learning_rate": 7.702695986338935e-06,
+      "loss": 0.3515,
+      "step": 4129
+    },
+    {
+      "epoch": 0.33884399228781226,
+      "grad_norm": 0.6752807833759652,
+      "learning_rate": 7.70157802547615e-06,
+      "loss": 0.3586,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33892603683800304,
+      "grad_norm": 0.5822108647807972,
+      "learning_rate": 7.700459873832263e-06,
+      "loss": 0.3639,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33900808138819377,
+      "grad_norm": 0.6149556060115998,
+      "learning_rate": 7.69934153148624e-06,
+      "loss": 0.3152,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33909012593838456,
+      "grad_norm": 0.6012387432917506,
+      "learning_rate": 7.698222998517054e-06,
+      "loss": 0.375,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3391721704885753,
+      "grad_norm": 0.6794910920560392,
+      "learning_rate": 7.697104275003696e-06,
+      "loss": 0.3522,
+      "step": 4134
+    },
+    {
+      "epoch": 0.33925421503876607,
+      "grad_norm": 0.6533480119663367,
+      "learning_rate": 7.695985361025164e-06,
+      "loss": 0.3406,
+      "step": 4135
+    },
+    {
+      "epoch": 0.3393362595889568,
+      "grad_norm": 0.6722957442307633,
+      "learning_rate": 7.694866256660482e-06,
+      "loss": 0.3654,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33941830413914753,
+      "grad_norm": 0.588981785937436,
+      "learning_rate": 7.693746961988672e-06,
+      "loss": 0.2935,
+      "step": 4137
+    },
+    {
+      "epoch": 0.3395003486893383,
+      "grad_norm": 0.617751406931135,
+      "learning_rate": 7.69262747708878e-06,
+      "loss": 0.3421,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33958239323952905,
+      "grad_norm": 0.6383900746965064,
+      "learning_rate": 7.691507802039861e-06,
+      "loss": 0.3209,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33966443778971983,
+      "grad_norm": 0.6619485120387361,
+      "learning_rate": 7.690387936920984e-06,
+      "loss": 0.3545,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33974648233991056,
+      "grad_norm": 0.6438548151848984,
+      "learning_rate": 7.689267881811235e-06,
+      "loss": 0.3588,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33982852689010135,
+      "grad_norm": 0.6581337681313286,
+      "learning_rate": 7.688147636789709e-06,
+      "loss": 0.4125,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3399105714402921,
+      "grad_norm": 0.6564239670323103,
+      "learning_rate": 7.687027201935516e-06,
+      "loss": 0.3367,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3399926159904828,
+      "grad_norm": 0.6922833063232003,
+      "learning_rate": 7.685906577327779e-06,
+      "loss": 0.3831,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3400746605406736,
+      "grad_norm": 0.6170678755364921,
+      "learning_rate": 7.684785763045635e-06,
+      "loss": 0.3485,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3401567050908643,
+      "grad_norm": 0.6527918436732315,
+      "learning_rate": 7.683664759168233e-06,
+      "loss": 0.3471,
+      "step": 4146
+    },
+    {
+      "epoch": 0.3402387496410551,
+      "grad_norm": 0.6218469196777309,
+      "learning_rate": 7.682543565774736e-06,
+      "loss": 0.3523,
+      "step": 4147
+    },
+    {
+      "epoch": 0.34032079419124583,
+      "grad_norm": 0.6448057374013446,
+      "learning_rate": 7.681422182944325e-06,
+      "loss": 0.3537,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3404028387414366,
+      "grad_norm": 0.5570291417623058,
+      "learning_rate": 7.680300610756185e-06,
+      "loss": 0.33,
+      "step": 4149
+    },
+    {
+      "epoch": 0.34048488329162735,
+      "grad_norm": 0.6301632843778358,
+      "learning_rate": 7.679178849289523e-06,
+      "loss": 0.3506,
+      "step": 4150
+    },
+    {
+      "epoch": 0.34056692784181813,
+      "grad_norm": 0.6515931428324109,
+      "learning_rate": 7.678056898623556e-06,
+      "loss": 0.3474,
+      "step": 4151
+    },
+    {
+      "epoch": 0.34064897239200886,
+      "grad_norm": 0.651364428947656,
+      "learning_rate": 7.676934758837513e-06,
+      "loss": 0.3388,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3407310169421996,
+      "grad_norm": 0.5890293957823008,
+      "learning_rate": 7.675812430010636e-06,
+      "loss": 0.3152,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3408130614923904,
+      "grad_norm": 0.6886443607233144,
+      "learning_rate": 7.674689912222186e-06,
+      "loss": 0.4014,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3408951060425811,
+      "grad_norm": 0.6471019604599654,
+      "learning_rate": 7.673567205551434e-06,
+      "loss": 0.3663,
+      "step": 4155
+    },
+    {
+      "epoch": 0.3409771505927719,
+      "grad_norm": 0.6147573590327262,
+      "learning_rate": 7.672444310077657e-06,
+      "loss": 0.3554,
+      "step": 4156
+    },
+    {
+      "epoch": 0.3410591951429626,
+      "grad_norm": 0.6371172640234267,
+      "learning_rate": 7.67132122588016e-06,
+      "loss": 0.3357,
+      "step": 4157
+    },
+    {
+      "epoch": 0.3411412396931534,
+      "grad_norm": 0.582293503074691,
+      "learning_rate": 7.670197953038248e-06,
+      "loss": 0.341,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34122328424334414,
+      "grad_norm": 0.6394112499375746,
+      "learning_rate": 7.669074491631246e-06,
+      "loss": 0.3403,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34130532879353487,
+      "grad_norm": 0.5969793174572812,
+      "learning_rate": 7.667950841738493e-06,
+      "loss": 0.3681,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34138737334372565,
+      "grad_norm": 0.6218478816256013,
+      "learning_rate": 7.666827003439338e-06,
+      "loss": 0.3344,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3414694178939164,
+      "grad_norm": 0.5986552307935588,
+      "learning_rate": 7.665702976813142e-06,
+      "loss": 0.3421,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34155146244410717,
+      "grad_norm": 0.6265208808661847,
+      "learning_rate": 7.664578761939288e-06,
+      "loss": 0.3358,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3416335069942979,
+      "grad_norm": 0.636826290104893,
+      "learning_rate": 7.663454358897161e-06,
+      "loss": 0.3469,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3417155515444887,
+      "grad_norm": 0.6127432665563871,
+      "learning_rate": 7.662329767766168e-06,
+      "loss": 0.3385,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3417975960946794,
+      "grad_norm": 0.6296551919136815,
+      "learning_rate": 7.661204988625724e-06,
+      "loss": 0.3345,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34187964064487014,
+      "grad_norm": 0.6334706555844278,
+      "learning_rate": 7.660080021555255e-06,
+      "loss": 0.3377,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3419616851950609,
+      "grad_norm": 0.6872982067060177,
+      "learning_rate": 7.658954866634213e-06,
+      "loss": 0.3077,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34204372974525166,
+      "grad_norm": 0.6120626683400578,
+      "learning_rate": 7.657829523942049e-06,
+      "loss": 0.3748,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34212577429544244,
+      "grad_norm": 0.6209229222010534,
+      "learning_rate": 7.656703993558235e-06,
+      "loss": 0.3135,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34220781884563317,
+      "grad_norm": 0.6169271841310977,
+      "learning_rate": 7.655578275562252e-06,
+      "loss": 0.3305,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34228986339582396,
+      "grad_norm": 0.6003132310032544,
+      "learning_rate": 7.654452370033598e-06,
+      "loss": 0.311,
+      "step": 4172
+    },
+    {
+      "epoch": 0.3423719079460147,
+      "grad_norm": 0.5966932265723529,
+      "learning_rate": 7.653326277051783e-06,
+      "loss": 0.3602,
+      "step": 4173
+    },
+    {
+      "epoch": 0.3424539524962054,
+      "grad_norm": 0.6353966093247563,
+      "learning_rate": 7.652199996696329e-06,
+      "loss": 0.3529,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3425359970463962,
+      "grad_norm": 0.6131613583334949,
+      "learning_rate": 7.651073529046772e-06,
+      "loss": 0.4009,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34261804159658693,
+      "grad_norm": 0.6376003007183152,
+      "learning_rate": 7.64994687418266e-06,
+      "loss": 0.3296,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3427000861467777,
+      "grad_norm": 0.6772421087192332,
+      "learning_rate": 7.648820032183559e-06,
+      "loss": 0.3408,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34278213069696845,
+      "grad_norm": 0.673260777300277,
+      "learning_rate": 7.647693003129044e-06,
+      "loss": 0.3422,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34286417524715923,
+      "grad_norm": 0.6276107990810416,
+      "learning_rate": 7.646565787098702e-06,
+      "loss": 0.3851,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34294621979734996,
+      "grad_norm": 0.6091362958104939,
+      "learning_rate": 7.645438384172134e-06,
+      "loss": 0.366,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3430282643475407,
+      "grad_norm": 0.6840879137211848,
+      "learning_rate": 7.64431079442896e-06,
+      "loss": 0.3538,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3431103088977315,
+      "grad_norm": 0.615641688926424,
+      "learning_rate": 7.643183017948806e-06,
+      "loss": 0.3369,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3431923534479222,
+      "grad_norm": 0.6300762101745881,
+      "learning_rate": 7.642055054811315e-06,
+      "loss": 0.3505,
+      "step": 4183
+    },
+    {
+      "epoch": 0.343274397998113,
+      "grad_norm": 0.5953211574660715,
+      "learning_rate": 7.640926905096139e-06,
+      "loss": 0.3981,
+      "step": 4184
+    },
+    {
+      "epoch": 0.3433564425483037,
+      "grad_norm": 0.6658403625323033,
+      "learning_rate": 7.639798568882947e-06,
+      "loss": 0.3463,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3434384870984945,
+      "grad_norm": 0.6727834958858018,
+      "learning_rate": 7.638670046251424e-06,
+      "loss": 0.3937,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34352053164868523,
+      "grad_norm": 0.6599046868929475,
+      "learning_rate": 7.63754133728126e-06,
+      "loss": 0.3892,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34360257619887596,
+      "grad_norm": 0.6569295012268115,
+      "learning_rate": 7.636412442052163e-06,
+      "loss": 0.3281,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34368462074906675,
+      "grad_norm": 0.7205105339357653,
+      "learning_rate": 7.635283360643857e-06,
+      "loss": 0.3301,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3437666652992575,
+      "grad_norm": 0.6436054931278368,
+      "learning_rate": 7.634154093136073e-06,
+      "loss": 0.4032,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34384870984944826,
+      "grad_norm": 0.6553668616523388,
+      "learning_rate": 7.63302463960856e-06,
+      "loss": 0.3274,
+      "step": 4191
+    },
+    {
+      "epoch": 0.343930754399639,
+      "grad_norm": 0.6285064547994039,
+      "learning_rate": 7.631895000141077e-06,
+      "loss": 0.3352,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3440127989498298,
+      "grad_norm": 0.629241353024002,
+      "learning_rate": 7.630765174813398e-06,
+      "loss": 0.3699,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3440948435000205,
+      "grad_norm": 0.6807690662923291,
+      "learning_rate": 7.629635163705307e-06,
+      "loss": 0.3567,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34417688805021124,
+      "grad_norm": 0.6306659626364883,
+      "learning_rate": 7.628504966896608e-06,
+      "loss": 0.3614,
+      "step": 4195
+    },
+    {
+      "epoch": 0.344258932600402,
+      "grad_norm": 0.6552450636938627,
+      "learning_rate": 7.62737458446711e-06,
+      "loss": 0.3799,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34434097715059275,
+      "grad_norm": 0.6453503104225492,
+      "learning_rate": 7.62624401649664e-06,
+      "loss": 0.4039,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34442302170078354,
+      "grad_norm": 0.6009334128294239,
+      "learning_rate": 7.625113263065037e-06,
+      "loss": 0.3201,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34450506625097427,
+      "grad_norm": 0.59830809119082,
+      "learning_rate": 7.623982324252152e-06,
+      "loss": 0.3475,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34458711080116505,
+      "grad_norm": 0.7291623311275677,
+      "learning_rate": 7.622851200137853e-06,
+      "loss": 0.3448,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3446691553513558,
+      "grad_norm": 0.6338732573196608,
+      "learning_rate": 7.621719890802015e-06,
+      "loss": 0.3698,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3447511999015465,
+      "grad_norm": 0.655434616838403,
+      "learning_rate": 7.620588396324531e-06,
+      "loss": 0.3837,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3448332444517373,
+      "grad_norm": 0.6686584962886666,
+      "learning_rate": 7.619456716785304e-06,
+      "loss": 0.3702,
+      "step": 4203
+    },
+    {
+      "epoch": 0.344915289001928,
+      "grad_norm": 0.6593125026740291,
+      "learning_rate": 7.618324852264252e-06,
+      "loss": 0.3422,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3449973335521188,
+      "grad_norm": 0.6511905035731084,
+      "learning_rate": 7.617192802841307e-06,
+      "loss": 0.3741,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34507937810230954,
+      "grad_norm": 0.6315667385825471,
+      "learning_rate": 7.61606056859641e-06,
+      "loss": 0.343,
+      "step": 4206
+    },
+    {
+      "epoch": 0.3451614226525003,
+      "grad_norm": 0.5971311019704033,
+      "learning_rate": 7.61492814960952e-06,
+      "loss": 0.3476,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34524346720269106,
+      "grad_norm": 0.6594458193600418,
+      "learning_rate": 7.613795545960602e-06,
+      "loss": 0.3678,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34532551175288184,
+      "grad_norm": 0.6374718536019456,
+      "learning_rate": 7.6126627577296455e-06,
+      "loss": 0.3944,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34540755630307257,
+      "grad_norm": 0.646288074249693,
+      "learning_rate": 7.6115297849966405e-06,
+      "loss": 0.3706,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3454896008532633,
+      "grad_norm": 0.673964731650604,
+      "learning_rate": 7.610396627841599e-06,
+      "loss": 0.3935,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3455716454034541,
+      "grad_norm": 0.6316129950973488,
+      "learning_rate": 7.609263286344542e-06,
+      "loss": 0.3414,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3456536899536448,
+      "grad_norm": 0.6192726674243323,
+      "learning_rate": 7.6081297605855035e-06,
+      "loss": 0.3683,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3457357345038356,
+      "grad_norm": 0.6197332532087737,
+      "learning_rate": 7.60699605064453e-06,
+      "loss": 0.3569,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34581777905402633,
+      "grad_norm": 0.606347068074032,
+      "learning_rate": 7.605862156601685e-06,
+      "loss": 0.3019,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3458998236042171,
+      "grad_norm": 0.6653438710702703,
+      "learning_rate": 7.604728078537041e-06,
+      "loss": 0.3779,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34598186815440785,
+      "grad_norm": 0.6223921772266564,
+      "learning_rate": 7.603593816530683e-06,
+      "loss": 0.343,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3460639127045986,
+      "grad_norm": 0.6070428411815881,
+      "learning_rate": 7.602459370662716e-06,
+      "loss": 0.3797,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34614595725478936,
+      "grad_norm": 0.659610433892064,
+      "learning_rate": 7.601324741013247e-06,
+      "loss": 0.3526,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3462280018049801,
+      "grad_norm": 0.683372606055429,
+      "learning_rate": 7.600189927662407e-06,
+      "loss": 0.685,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3463100463551709,
+      "grad_norm": 0.6192853574908286,
+      "learning_rate": 7.599054930690328e-06,
+      "loss": 0.3287,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3463920909053616,
+      "grad_norm": 0.6189136393817631,
+      "learning_rate": 7.597919750177168e-06,
+      "loss": 0.3488,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3464741354555524,
+      "grad_norm": 0.6529379163612716,
+      "learning_rate": 7.596784386203089e-06,
+      "loss": 0.3754,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3465561800057431,
+      "grad_norm": 0.6145389381502726,
+      "learning_rate": 7.595648838848269e-06,
+      "loss": 0.3712,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34663822455593385,
+      "grad_norm": 0.6094683308529003,
+      "learning_rate": 7.594513108192896e-06,
+      "loss": 0.3185,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34672026910612463,
+      "grad_norm": 0.688650902428089,
+      "learning_rate": 7.593377194317178e-06,
+      "loss": 0.3275,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34680231365631536,
+      "grad_norm": 0.6061120998345038,
+      "learning_rate": 7.59224109730133e-06,
+      "loss": 0.3672,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34688435820650615,
+      "grad_norm": 0.6143612657691596,
+      "learning_rate": 7.5911048172255785e-06,
+      "loss": 0.3452,
+      "step": 4228
+    },
+    {
+      "epoch": 0.3469664027566969,
+      "grad_norm": 0.6615846798131626,
+      "learning_rate": 7.589968354170168e-06,
+      "loss": 0.357,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34704844730688766,
+      "grad_norm": 0.6151145594145898,
+      "learning_rate": 7.588831708215352e-06,
+      "loss": 0.3215,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3471304918570784,
+      "grad_norm": 0.635932913103609,
+      "learning_rate": 7.5876948794414015e-06,
+      "loss": 0.3497,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3472125364072691,
+      "grad_norm": 0.6743044942594593,
+      "learning_rate": 7.586557867928596e-06,
+      "loss": 0.3477,
+      "step": 4232
+    },
+    {
+      "epoch": 0.3472945809574599,
+      "grad_norm": 0.6361215922338673,
+      "learning_rate": 7.58542067375723e-06,
+      "loss": 0.3275,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34737662550765064,
+      "grad_norm": 0.6550805951389614,
+      "learning_rate": 7.584283297007608e-06,
+      "loss": 0.3426,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3474586700578414,
+      "grad_norm": 0.6977456053747091,
+      "learning_rate": 7.583145737760053e-06,
+      "loss": 0.3532,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34754071460803215,
+      "grad_norm": 0.6217961920617204,
+      "learning_rate": 7.5820079960948975e-06,
+      "loss": 0.3903,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34762275915822294,
+      "grad_norm": 0.6339588251558873,
+      "learning_rate": 7.5808700720924845e-06,
+      "loss": 0.3584,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34770480370841367,
+      "grad_norm": 0.6122623741229729,
+      "learning_rate": 7.579731965833172e-06,
+      "loss": 0.3531,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3477868482586044,
+      "grad_norm": 0.5825195083240912,
+      "learning_rate": 7.578593677397334e-06,
+      "loss": 0.3446,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3478688928087952,
+      "grad_norm": 0.6077396104394803,
+      "learning_rate": 7.577455206865354e-06,
+      "loss": 0.3391,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3479509373589859,
+      "grad_norm": 0.6763596497391418,
+      "learning_rate": 7.5763165543176275e-06,
+      "loss": 0.3142,
+      "step": 4241
+    },
+    {
+      "epoch": 0.3480329819091767,
+      "grad_norm": 0.5871255350160196,
+      "learning_rate": 7.5751777198345656e-06,
+      "loss": 0.3494,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3481150264593674,
+      "grad_norm": 0.6341308724082138,
+      "learning_rate": 7.574038703496589e-06,
+      "loss": 0.3236,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3481970710095582,
+      "grad_norm": 0.5868923729965149,
+      "learning_rate": 7.5728995053841355e-06,
+      "loss": 0.3692,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34827911555974894,
+      "grad_norm": 0.663033824117739,
+      "learning_rate": 7.571760125577653e-06,
+      "loss": 0.3362,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34836116010993967,
+      "grad_norm": 0.6292374048588996,
+      "learning_rate": 7.570620564157603e-06,
+      "loss": 0.3924,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34844320466013046,
+      "grad_norm": 0.6369732549947911,
+      "learning_rate": 7.569480821204457e-06,
+      "loss": 0.3407,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3485252492103212,
+      "grad_norm": 0.5970315724878045,
+      "learning_rate": 7.568340896798705e-06,
+      "loss": 0.3915,
+      "step": 4248
+    },
+    {
+      "epoch": 0.34860729376051197,
+      "grad_norm": 0.6252682069447442,
+      "learning_rate": 7.567200791020845e-06,
+      "loss": 0.3391,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3486893383107027,
+      "grad_norm": 0.6495134321887972,
+      "learning_rate": 7.566060503951389e-06,
+      "loss": 0.3385,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3487713828608935,
+      "grad_norm": 0.6204995685189649,
+      "learning_rate": 7.564920035670863e-06,
+      "loss": 0.3514,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3488534274110842,
+      "grad_norm": 0.6980589131079669,
+      "learning_rate": 7.563779386259804e-06,
+      "loss": 0.3767,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34893547196127495,
+      "grad_norm": 0.637951603121941,
+      "learning_rate": 7.562638555798764e-06,
+      "loss": 0.3569,
+      "step": 4253
+    },
+    {
+      "epoch": 0.34901751651146573,
+      "grad_norm": 0.6354906729350202,
+      "learning_rate": 7.561497544368309e-06,
+      "loss": 0.3533,
+      "step": 4254
+    },
+    {
+      "epoch": 0.34909956106165646,
+      "grad_norm": 0.6343082681165116,
+      "learning_rate": 7.5603563520490116e-06,
+      "loss": 0.3527,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34918160561184725,
+      "grad_norm": 0.5956771182146808,
+      "learning_rate": 7.559214978921461e-06,
+      "loss": 0.3281,
+      "step": 4256
+    },
+    {
+      "epoch": 0.349263650162038,
+      "grad_norm": 0.6124560071680589,
+      "learning_rate": 7.558073425066261e-06,
+      "loss": 0.3419,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34934569471222876,
+      "grad_norm": 0.6352194252186627,
+      "learning_rate": 7.556931690564025e-06,
+      "loss": 0.3378,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3494277392624195,
+      "grad_norm": 0.5691971213388486,
+      "learning_rate": 7.555789775495381e-06,
+      "loss": 0.317,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3495097838126102,
+      "grad_norm": 0.6216878617862772,
+      "learning_rate": 7.554647679940971e-06,
+      "loss": 0.3314,
+      "step": 4260
+    },
+    {
+      "epoch": 0.349591828362801,
+      "grad_norm": 0.6576566668282812,
+      "learning_rate": 7.553505403981445e-06,
+      "loss": 0.3349,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34967387291299173,
+      "grad_norm": 0.6662283184890773,
+      "learning_rate": 7.552362947697469e-06,
+      "loss": 0.3554,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3497559174631825,
+      "grad_norm": 0.6864579523716543,
+      "learning_rate": 7.551220311169724e-06,
+      "loss": 0.3696,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34983796201337325,
+      "grad_norm": 0.635300120223772,
+      "learning_rate": 7.550077494478898e-06,
+      "loss": 0.3601,
+      "step": 4264
+    },
+    {
+      "epoch": 0.34992000656356403,
+      "grad_norm": 0.628518939642808,
+      "learning_rate": 7.548934497705698e-06,
+      "loss": 0.3835,
+      "step": 4265
+    },
+    {
+      "epoch": 0.35000205111375476,
+      "grad_norm": 0.6394347960841288,
+      "learning_rate": 7.5477913209308365e-06,
+      "loss": 0.3208,
+      "step": 4266
+    },
+    {
+      "epoch": 0.35008409566394555,
+      "grad_norm": 0.7221784985659809,
+      "learning_rate": 7.546647964235046e-06,
+      "loss": 0.329,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3501661402141363,
+      "grad_norm": 0.6275033853648736,
+      "learning_rate": 7.545504427699071e-06,
+      "loss": 0.3652,
+      "step": 4268
+    },
+    {
+      "epoch": 0.350248184764327,
+      "grad_norm": 0.5861022241181714,
+      "learning_rate": 7.54436071140366e-06,
+      "loss": 0.3448,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3503302293145178,
+      "grad_norm": 0.6659024782185522,
+      "learning_rate": 7.543216815429583e-06,
+      "loss": 0.3488,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3504122738647085,
+      "grad_norm": 0.6463002426729056,
+      "learning_rate": 7.542072739857621e-06,
+      "loss": 0.3226,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3504943184148993,
+      "grad_norm": 0.6379978281913091,
+      "learning_rate": 7.540928484768566e-06,
+      "loss": 0.3967,
+      "step": 4272
+    },
+    {
+      "epoch": 0.35057636296509004,
+      "grad_norm": 0.6194832703490142,
+      "learning_rate": 7.539784050243225e-06,
+      "loss": 0.3261,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3506584075152808,
+      "grad_norm": 0.6853577658157123,
+      "learning_rate": 7.538639436362414e-06,
+      "loss": 0.4138,
+      "step": 4274
+    },
+    {
+      "epoch": 0.35074045206547155,
+      "grad_norm": 0.6276949067194048,
+      "learning_rate": 7.537494643206965e-06,
+      "loss": 0.3619,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3508224966156623,
+      "grad_norm": 0.6646120184757417,
+      "learning_rate": 7.536349670857721e-06,
+      "loss": 0.347,
+      "step": 4276
+    },
+    {
+      "epoch": 0.35090454116585307,
+      "grad_norm": 0.6345240435841458,
+      "learning_rate": 7.535204519395538e-06,
+      "loss": 0.3262,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3509865857160438,
+      "grad_norm": 0.6236557023572784,
+      "learning_rate": 7.534059188901286e-06,
+      "loss": 0.3397,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3510686302662346,
+      "grad_norm": 0.6173829657257821,
+      "learning_rate": 7.532913679455842e-06,
+      "loss": 0.3372,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3511506748164253,
+      "grad_norm": 0.6587159208071837,
+      "learning_rate": 7.531767991140106e-06,
+      "loss": 0.3563,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3512327193666161,
+      "grad_norm": 0.6597391712111631,
+      "learning_rate": 7.530622124034983e-06,
+      "loss": 0.3168,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3513147639168068,
+      "grad_norm": 0.6436967924359435,
+      "learning_rate": 7.52947607822139e-06,
+      "loss": 0.3475,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35139680846699756,
+      "grad_norm": 0.5977351127833163,
+      "learning_rate": 7.528329853780261e-06,
+      "loss": 0.3255,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35147885301718834,
+      "grad_norm": 0.5776958894034172,
+      "learning_rate": 7.527183450792539e-06,
+      "loss": 0.3368,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35156089756737907,
+      "grad_norm": 0.6889093339830665,
+      "learning_rate": 7.526036869339182e-06,
+      "loss": 0.4177,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35164294211756986,
+      "grad_norm": 0.6388676181149929,
+      "learning_rate": 7.52489010950116e-06,
+      "loss": 0.3585,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3517249866677606,
+      "grad_norm": 0.6778103223709722,
+      "learning_rate": 7.523743171359454e-06,
+      "loss": 0.3376,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35180703121795137,
+      "grad_norm": 0.6581683102443368,
+      "learning_rate": 7.522596054995061e-06,
+      "loss": 0.3885,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3518890757681421,
+      "grad_norm": 0.6258515781265952,
+      "learning_rate": 7.521448760488987e-06,
+      "loss": 0.335,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35197112031833283,
+      "grad_norm": 0.6263623170867634,
+      "learning_rate": 7.520301287922254e-06,
+      "loss": 0.3821,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3520531648685236,
+      "grad_norm": 0.5717338245276864,
+      "learning_rate": 7.519153637375892e-06,
+      "loss": 0.3077,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35213520941871435,
+      "grad_norm": 0.6026086667051354,
+      "learning_rate": 7.518005808930946e-06,
+      "loss": 0.3629,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35221725396890513,
+      "grad_norm": 0.6245381727471431,
+      "learning_rate": 7.5168578026684776e-06,
+      "loss": 0.3657,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35229929851909586,
+      "grad_norm": 0.6257045745573946,
+      "learning_rate": 7.515709618669553e-06,
+      "loss": 0.3465,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35238134306928665,
+      "grad_norm": 0.6488353078263582,
+      "learning_rate": 7.514561257015258e-06,
+      "loss": 0.3571,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3524633876194774,
+      "grad_norm": 0.6493271002323215,
+      "learning_rate": 7.513412717786689e-06,
+      "loss": 0.3343,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3525454321696681,
+      "grad_norm": 0.6314440378527651,
+      "learning_rate": 7.512264001064948e-06,
+      "loss": 0.3412,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3526274767198589,
+      "grad_norm": 0.7105644261300093,
+      "learning_rate": 7.5111151069311614e-06,
+      "loss": 0.336,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3527095212700496,
+      "grad_norm": 0.6230930593812091,
+      "learning_rate": 7.509966035466461e-06,
+      "loss": 0.3211,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3527915658202404,
+      "grad_norm": 0.6395672974985945,
+      "learning_rate": 7.508816786751991e-06,
+      "loss": 0.3422,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35287361037043113,
+      "grad_norm": 0.5982304442826913,
+      "learning_rate": 7.50766736086891e-06,
+      "loss": 0.3211,
+      "step": 4301
+    },
+    {
+      "epoch": 0.3529556549206219,
+      "grad_norm": 0.6296563833320001,
+      "learning_rate": 7.50651775789839e-06,
+      "loss": 0.3325,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35303769947081265,
+      "grad_norm": 0.6063132479136442,
+      "learning_rate": 7.505367977921611e-06,
+      "loss": 0.3186,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3531197440210034,
+      "grad_norm": 0.6617527989235938,
+      "learning_rate": 7.504218021019773e-06,
+      "loss": 0.3686,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35320178857119416,
+      "grad_norm": 0.6366464504170396,
+      "learning_rate": 7.503067887274081e-06,
+      "loss": 0.3787,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3532838331213849,
+      "grad_norm": 0.6223089949692734,
+      "learning_rate": 7.501917576765754e-06,
+      "loss": 0.3503,
+      "step": 4306
+    },
+    {
+      "epoch": 0.3533658776715757,
+      "grad_norm": 0.6255287942167347,
+      "learning_rate": 7.500767089576029e-06,
+      "loss": 0.3318,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3534479222217664,
+      "grad_norm": 0.6440984228788312,
+      "learning_rate": 7.499616425786151e-06,
+      "loss": 0.3448,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3535299667719572,
+      "grad_norm": 0.6433205251759967,
+      "learning_rate": 7.498465585477376e-06,
+      "loss": 0.3575,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3536120113221479,
+      "grad_norm": 0.6174748092036719,
+      "learning_rate": 7.497314568730975e-06,
+      "loss": 0.3615,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35369405587233865,
+      "grad_norm": 0.7916555641271639,
+      "learning_rate": 7.496163375628232e-06,
+      "loss": 0.3662,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35377610042252944,
+      "grad_norm": 0.6586975818400768,
+      "learning_rate": 7.495012006250441e-06,
+      "loss": 0.3306,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35385814497272017,
+      "grad_norm": 0.6452707928966408,
+      "learning_rate": 7.493860460678913e-06,
+      "loss": 0.3375,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35394018952291095,
+      "grad_norm": 0.6728464603494987,
+      "learning_rate": 7.492708738994963e-06,
+      "loss": 0.3553,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3540222340731017,
+      "grad_norm": 0.6272506664444788,
+      "learning_rate": 7.491556841279927e-06,
+      "loss": 0.3402,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35410427862329247,
+      "grad_norm": 0.6311330295476362,
+      "learning_rate": 7.490404767615151e-06,
+      "loss": 0.3618,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3541863231734832,
+      "grad_norm": 0.6400532587798117,
+      "learning_rate": 7.489252518081992e-06,
+      "loss": 0.3784,
+      "step": 4317
+    },
+    {
+      "epoch": 0.354268367723674,
+      "grad_norm": 0.642829908089807,
+      "learning_rate": 7.488100092761818e-06,
+      "loss": 0.3485,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3543504122738647,
+      "grad_norm": 0.6694575577293749,
+      "learning_rate": 7.486947491736014e-06,
+      "loss": 0.3537,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35443245682405544,
+      "grad_norm": 0.6751649659868225,
+      "learning_rate": 7.485794715085972e-06,
+      "loss": 0.3723,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3545145013742462,
+      "grad_norm": 0.6565596275845019,
+      "learning_rate": 7.484641762893103e-06,
+      "loss": 0.3433,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35459654592443696,
+      "grad_norm": 0.6475155859450633,
+      "learning_rate": 7.4834886352388235e-06,
+      "loss": 0.36,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35467859047462774,
+      "grad_norm": 0.6567303792492567,
+      "learning_rate": 7.482335332204568e-06,
+      "loss": 0.3622,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35476063502481847,
+      "grad_norm": 0.6658836652188527,
+      "learning_rate": 7.481181853871779e-06,
+      "loss": 0.3891,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35484267957500926,
+      "grad_norm": 0.6317212592335821,
+      "learning_rate": 7.4800282003219125e-06,
+      "loss": 0.3469,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3549247241252,
+      "grad_norm": 0.6475511231525789,
+      "learning_rate": 7.47887437163644e-06,
+      "loss": 0.3809,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3550067686753907,
+      "grad_norm": 0.6627038432445448,
+      "learning_rate": 7.477720367896841e-06,
+      "loss": 0.3408,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3550888132255815,
+      "grad_norm": 0.6629144233435889,
+      "learning_rate": 7.476566189184611e-06,
+      "loss": 0.3403,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35517085777577223,
+      "grad_norm": 0.6182701765851676,
+      "learning_rate": 7.4754118355812555e-06,
+      "loss": 0.3314,
+      "step": 4329
+    },
+    {
+      "epoch": 0.355252902325963,
+      "grad_norm": 0.6460695654600577,
+      "learning_rate": 7.474257307168294e-06,
+      "loss": 0.335,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35533494687615375,
+      "grad_norm": 0.7087393806172034,
+      "learning_rate": 7.473102604027256e-06,
+      "loss": 0.3828,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35541699142634453,
+      "grad_norm": 0.6405880082173135,
+      "learning_rate": 7.471947726239686e-06,
+      "loss": 0.3568,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35549903597653526,
+      "grad_norm": 0.5982838278312579,
+      "learning_rate": 7.470792673887137e-06,
+      "loss": 0.3264,
+      "step": 4333
+    },
+    {
+      "epoch": 0.355581080526726,
+      "grad_norm": 0.610301576918131,
+      "learning_rate": 7.4696374470511816e-06,
+      "loss": 0.3313,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3556631250769168,
+      "grad_norm": 0.6011528764884647,
+      "learning_rate": 7.468482045813397e-06,
+      "loss": 0.3257,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3557451696271075,
+      "grad_norm": 0.5929875644428972,
+      "learning_rate": 7.467326470255375e-06,
+      "loss": 0.3608,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3558272141772983,
+      "grad_norm": 0.6236428836693166,
+      "learning_rate": 7.466170720458723e-06,
+      "loss": 0.3795,
+      "step": 4337
+    },
+    {
+      "epoch": 0.355909258727489,
+      "grad_norm": 0.6106339630704036,
+      "learning_rate": 7.4650147965050565e-06,
+      "loss": 0.3186,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3559913032776798,
+      "grad_norm": 0.6369288925068904,
+      "learning_rate": 7.463858698476005e-06,
+      "loss": 0.3726,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35607334782787053,
+      "grad_norm": 0.6862770160706607,
+      "learning_rate": 7.462702426453211e-06,
+      "loss": 0.3364,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35615539237806126,
+      "grad_norm": 0.5938372492364701,
+      "learning_rate": 7.461545980518329e-06,
+      "loss": 0.305,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35623743692825205,
+      "grad_norm": 0.6212205515244796,
+      "learning_rate": 7.460389360753022e-06,
+      "loss": 0.3328,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3563194814784428,
+      "grad_norm": 0.6656243008759053,
+      "learning_rate": 7.459232567238974e-06,
+      "loss": 0.3401,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35640152602863356,
+      "grad_norm": 0.6547177870983455,
+      "learning_rate": 7.458075600057872e-06,
+      "loss": 0.3283,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3564835705788243,
+      "grad_norm": 0.6582335482137988,
+      "learning_rate": 7.456918459291422e-06,
+      "loss": 0.3486,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3565656151290151,
+      "grad_norm": 0.6324710227916306,
+      "learning_rate": 7.455761145021335e-06,
+      "loss": 0.3478,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3566476596792058,
+      "grad_norm": 0.6414716709773297,
+      "learning_rate": 7.454603657329343e-06,
+      "loss": 0.3496,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35672970422939654,
+      "grad_norm": 0.6507135276064299,
+      "learning_rate": 7.453445996297184e-06,
+      "loss": 0.3227,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3568117487795873,
+      "grad_norm": 0.664473897068423,
+      "learning_rate": 7.452288162006611e-06,
+      "loss": 0.3678,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35689379332977805,
+      "grad_norm": 0.6689909548618833,
+      "learning_rate": 7.451130154539388e-06,
+      "loss": 0.3529,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35697583787996884,
+      "grad_norm": 0.653677999501801,
+      "learning_rate": 7.44997197397729e-06,
+      "loss": 0.373,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35705788243015957,
+      "grad_norm": 0.6383447491320662,
+      "learning_rate": 7.448813620402109e-06,
+      "loss": 0.3737,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35713992698035035,
+      "grad_norm": 0.591868408217995,
+      "learning_rate": 7.4476550938956445e-06,
+      "loss": 0.3309,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3572219715305411,
+      "grad_norm": 0.6815575533534833,
+      "learning_rate": 7.446496394539708e-06,
+      "loss": 0.365,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3573040160807318,
+      "grad_norm": 0.6733698593491041,
+      "learning_rate": 7.445337522416129e-06,
+      "loss": 0.3402,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3573860606309226,
+      "grad_norm": 0.6394825126233591,
+      "learning_rate": 7.44417847760674e-06,
+      "loss": 0.3617,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3574681051811133,
+      "grad_norm": 0.6048246932515794,
+      "learning_rate": 7.443019260193396e-06,
+      "loss": 0.3555,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3575501497313041,
+      "grad_norm": 0.578084433260552,
+      "learning_rate": 7.441859870257958e-06,
+      "loss": 0.3421,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35763219428149484,
+      "grad_norm": 0.615063443666004,
+      "learning_rate": 7.4407003078822984e-06,
+      "loss": 0.3185,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3577142388316856,
+      "grad_norm": 0.6187348676420912,
+      "learning_rate": 7.439540573148301e-06,
+      "loss": 0.3352,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35779628338187636,
+      "grad_norm": 0.6151791871930327,
+      "learning_rate": 7.438380666137871e-06,
+      "loss": 0.3592,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3578783279320671,
+      "grad_norm": 0.6242779793229818,
+      "learning_rate": 7.437220586932915e-06,
+      "loss": 0.3489,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35796037248225787,
+      "grad_norm": 0.6512260872191291,
+      "learning_rate": 7.436060335615357e-06,
+      "loss": 0.3432,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3580424170324486,
+      "grad_norm": 0.7132103600101476,
+      "learning_rate": 7.434899912267132e-06,
+      "loss": 0.3329,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3581244615826394,
+      "grad_norm": 0.6285569652877508,
+      "learning_rate": 7.433739316970188e-06,
+      "loss": 0.3987,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3582065061328301,
+      "grad_norm": 0.601290249993666,
+      "learning_rate": 7.432578549806482e-06,
+      "loss": 0.3439,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3582885506830209,
+      "grad_norm": 0.6607392114742283,
+      "learning_rate": 7.4314176108579875e-06,
+      "loss": 0.3425,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35837059523321163,
+      "grad_norm": 0.6942147642375738,
+      "learning_rate": 7.430256500206687e-06,
+      "loss": 0.3558,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35845263978340236,
+      "grad_norm": 0.6560675961717752,
+      "learning_rate": 7.429095217934578e-06,
+      "loss": 0.3652,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35853468433359315,
+      "grad_norm": 0.6638894798983005,
+      "learning_rate": 7.427933764123667e-06,
+      "loss": 0.3835,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3586167288837839,
+      "grad_norm": 0.7119980051202276,
+      "learning_rate": 7.426772138855974e-06,
+      "loss": 0.3454,
+      "step": 4371
+    },
+    {
+      "epoch": 0.35869877343397466,
+      "grad_norm": 0.5553583297506532,
+      "learning_rate": 7.425610342213534e-06,
+      "loss": 0.3467,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3587808179841654,
+      "grad_norm": 0.5962110984757653,
+      "learning_rate": 7.424448374278386e-06,
+      "loss": 0.3389,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3588628625343562,
+      "grad_norm": 0.7044429099197352,
+      "learning_rate": 7.4232862351325885e-06,
+      "loss": 0.3326,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3589449070845469,
+      "grad_norm": 0.5930890692841625,
+      "learning_rate": 7.4221239248582125e-06,
+      "loss": 0.3325,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3590269516347377,
+      "grad_norm": 0.6969737741824142,
+      "learning_rate": 7.420961443537335e-06,
+      "loss": 0.3475,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3591089961849284,
+      "grad_norm": 0.6080119804313153,
+      "learning_rate": 7.419798791252051e-06,
+      "loss": 0.3478,
+      "step": 4377
+    },
+    {
+      "epoch": 0.35919104073511915,
+      "grad_norm": 0.6461650894661735,
+      "learning_rate": 7.418635968084462e-06,
+      "loss": 0.3428,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35927308528530993,
+      "grad_norm": 0.8317194172719057,
+      "learning_rate": 7.417472974116688e-06,
+      "loss": 0.3669,
+      "step": 4379
+    },
+    {
+      "epoch": 0.35935512983550066,
+      "grad_norm": 0.6679939149786575,
+      "learning_rate": 7.416309809430857e-06,
+      "loss": 0.3627,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35943717438569145,
+      "grad_norm": 0.625165080669201,
+      "learning_rate": 7.41514647410911e-06,
+      "loss": 0.3205,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3595192189358822,
+      "grad_norm": 0.6056413106731233,
+      "learning_rate": 7.413982968233597e-06,
+      "loss": 0.367,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35960126348607296,
+      "grad_norm": 0.6560467031513265,
+      "learning_rate": 7.412819291886486e-06,
+      "loss": 0.3552,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3596833080362637,
+      "grad_norm": 0.6689812545031468,
+      "learning_rate": 7.411655445149953e-06,
+      "loss": 0.3432,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3597653525864544,
+      "grad_norm": 0.6232025185902996,
+      "learning_rate": 7.410491428106188e-06,
+      "loss": 0.3501,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3598473971366452,
+      "grad_norm": 0.6476007272917023,
+      "learning_rate": 7.409327240837392e-06,
+      "loss": 0.363,
+      "step": 4386
+    },
+    {
+      "epoch": 0.35992944168683594,
+      "grad_norm": 0.6284790928942839,
+      "learning_rate": 7.4081628834257745e-06,
+      "loss": 0.32,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3600114862370267,
+      "grad_norm": 0.6758515113212977,
+      "learning_rate": 7.406998355953566e-06,
+      "loss": 0.337,
+      "step": 4388
+    },
+    {
+      "epoch": 0.36009353078721745,
+      "grad_norm": 0.6420267954780257,
+      "learning_rate": 7.405833658502999e-06,
+      "loss": 0.3443,
+      "step": 4389
+    },
+    {
+      "epoch": 0.36017557533740824,
+      "grad_norm": 0.7113750503142612,
+      "learning_rate": 7.404668791156325e-06,
+      "loss": 0.3227,
+      "step": 4390
+    },
+    {
+      "epoch": 0.36025761988759897,
+      "grad_norm": 0.6829374111887803,
+      "learning_rate": 7.403503753995802e-06,
+      "loss": 0.3534,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3603396644377897,
+      "grad_norm": 0.708935876759268,
+      "learning_rate": 7.402338547103708e-06,
+      "loss": 0.3495,
+      "step": 4392
+    },
+    {
+      "epoch": 0.3604217089879805,
+      "grad_norm": 0.599165683619013,
+      "learning_rate": 7.401173170562324e-06,
+      "loss": 0.3713,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3605037535381712,
+      "grad_norm": 0.5973927137372558,
+      "learning_rate": 7.4000076244539495e-06,
+      "loss": 0.3245,
+      "step": 4394
+    },
+    {
+      "epoch": 0.360585798088362,
+      "grad_norm": 0.7043930248607715,
+      "learning_rate": 7.398841908860893e-06,
+      "loss": 0.3717,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3606678426385527,
+      "grad_norm": 0.6934282240285471,
+      "learning_rate": 7.397676023865473e-06,
+      "loss": 0.3847,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3607498871887435,
+      "grad_norm": 0.6652166108221608,
+      "learning_rate": 7.396509969550025e-06,
+      "loss": 0.3872,
+      "step": 4397
+    },
+    {
+      "epoch": 0.36083193173893424,
+      "grad_norm": 0.626383766579676,
+      "learning_rate": 7.395343745996895e-06,
+      "loss": 0.342,
+      "step": 4398
+    },
+    {
+      "epoch": 0.36091397628912497,
+      "grad_norm": 0.6164468244190888,
+      "learning_rate": 7.3941773532884365e-06,
+      "loss": 0.3204,
+      "step": 4399
+    },
+    {
+      "epoch": 0.36099602083931576,
+      "grad_norm": 0.6170625821940035,
+      "learning_rate": 7.3930107915070204e-06,
+      "loss": 0.3592,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3610780653895065,
+      "grad_norm": 0.7417326315126752,
+      "learning_rate": 7.391844060735026e-06,
+      "loss": 0.3633,
+      "step": 4401
+    },
+    {
+      "epoch": 0.36116010993969727,
+      "grad_norm": 0.6469829770265636,
+      "learning_rate": 7.390677161054848e-06,
+      "loss": 0.3243,
+      "step": 4402
+    },
+    {
+      "epoch": 0.361242154489888,
+      "grad_norm": 0.700525160582892,
+      "learning_rate": 7.389510092548889e-06,
+      "loss": 0.3415,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3613241990400788,
+      "grad_norm": 0.6618879518988219,
+      "learning_rate": 7.388342855299566e-06,
+      "loss": 0.343,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3614062435902695,
+      "grad_norm": 1.0733349969483126,
+      "learning_rate": 7.387175449389308e-06,
+      "loss": 0.3693,
+      "step": 4405
+    },
+    {
+      "epoch": 0.36148828814046025,
+      "grad_norm": 0.5982129933780846,
+      "learning_rate": 7.386007874900556e-06,
+      "loss": 0.3309,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36157033269065103,
+      "grad_norm": 0.64261813241084,
+      "learning_rate": 7.38484013191576e-06,
+      "loss": 0.3797,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36165237724084176,
+      "grad_norm": 0.6678708534609936,
+      "learning_rate": 7.383672220517387e-06,
+      "loss": 0.3772,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36173442179103255,
+      "grad_norm": 0.6664458332867247,
+      "learning_rate": 7.3825041407879094e-06,
+      "loss": 0.3233,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3618164663412233,
+      "grad_norm": 0.6395899140390399,
+      "learning_rate": 7.381335892809819e-06,
+      "loss": 0.3686,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36189851089141406,
+      "grad_norm": 0.6009256617712884,
+      "learning_rate": 7.380167476665613e-06,
+      "loss": 0.3672,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3619805554416048,
+      "grad_norm": 0.6845035211521446,
+      "learning_rate": 7.378998892437803e-06,
+      "loss": 0.404,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3620625999917955,
+      "grad_norm": 0.6101624247854002,
+      "learning_rate": 7.377830140208915e-06,
+      "loss": 0.335,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3621446445419863,
+      "grad_norm": 0.6294084226537625,
+      "learning_rate": 7.3766612200614814e-06,
+      "loss": 0.3766,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36222668909217703,
+      "grad_norm": 0.6350153693266466,
+      "learning_rate": 7.375492132078051e-06,
+      "loss": 0.3418,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3623087336423678,
+      "grad_norm": 0.6374429854646978,
+      "learning_rate": 7.374322876341184e-06,
+      "loss": 0.3359,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36239077819255855,
+      "grad_norm": 0.6407877733643291,
+      "learning_rate": 7.373153452933449e-06,
+      "loss": 0.3736,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36247282274274933,
+      "grad_norm": 0.6499455093746933,
+      "learning_rate": 7.37198386193743e-06,
+      "loss": 0.3309,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36255486729294006,
+      "grad_norm": 0.6763926816479953,
+      "learning_rate": 7.370814103435722e-06,
+      "loss": 0.333,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3626369118431308,
+      "grad_norm": 0.5998832387869221,
+      "learning_rate": 7.3696441775109305e-06,
+      "loss": 0.3264,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3627189563933216,
+      "grad_norm": 0.7041598595714932,
+      "learning_rate": 7.368474084245676e-06,
+      "loss": 0.3545,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3628010009435123,
+      "grad_norm": 0.6501930347990397,
+      "learning_rate": 7.367303823722585e-06,
+      "loss": 0.3685,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3628830454937031,
+      "grad_norm": 0.6476972186666273,
+      "learning_rate": 7.366133396024301e-06,
+      "loss": 0.3748,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3629650900438938,
+      "grad_norm": 0.6601205350797885,
+      "learning_rate": 7.3649628012334785e-06,
+      "loss": 0.3647,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3630471345940846,
+      "grad_norm": 0.6339175031374529,
+      "learning_rate": 7.363792039432783e-06,
+      "loss": 0.3406,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36312917914427534,
+      "grad_norm": 0.6765117464761441,
+      "learning_rate": 7.362621110704892e-06,
+      "loss": 0.3586,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36321122369446607,
+      "grad_norm": 0.6970669120099977,
+      "learning_rate": 7.361450015132493e-06,
+      "loss": 0.3411,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36329326824465685,
+      "grad_norm": 0.7909817181359752,
+      "learning_rate": 7.360278752798289e-06,
+      "loss": 0.3177,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3633753127948476,
+      "grad_norm": 0.6819152842513383,
+      "learning_rate": 7.3591073237849894e-06,
+      "loss": 0.3733,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36345735734503837,
+      "grad_norm": 0.6436183425199609,
+      "learning_rate": 7.357935728175323e-06,
+      "loss": 0.3401,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3635394018952291,
+      "grad_norm": 0.6731149291024439,
+      "learning_rate": 7.356763966052023e-06,
+      "loss": 0.3513,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3636214464454199,
+      "grad_norm": 0.6218670400972143,
+      "learning_rate": 7.355592037497837e-06,
+      "loss": 0.3549,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3637034909956106,
+      "grad_norm": 0.6380267825389364,
+      "learning_rate": 7.354419942595528e-06,
+      "loss": 0.345,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3637855355458014,
+      "grad_norm": 0.6609139292955213,
+      "learning_rate": 7.353247681427864e-06,
+      "loss": 0.3622,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3638675800959921,
+      "grad_norm": 0.6727447036845013,
+      "learning_rate": 7.352075254077631e-06,
+      "loss": 0.3459,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36394962464618286,
+      "grad_norm": 0.6554882436607857,
+      "learning_rate": 7.350902660627619e-06,
+      "loss": 0.3563,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36403166919637364,
+      "grad_norm": 0.5994054344432587,
+      "learning_rate": 7.349729901160641e-06,
+      "loss": 0.3668,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36411371374656437,
+      "grad_norm": 0.6348533510354611,
+      "learning_rate": 7.348556975759512e-06,
+      "loss": 0.3352,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36419575829675516,
+      "grad_norm": 0.7251521769048809,
+      "learning_rate": 7.3473838845070624e-06,
+      "loss": 0.3252,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3642778028469459,
+      "grad_norm": 0.6441605967026591,
+      "learning_rate": 7.346210627486135e-06,
+      "loss": 0.3476,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36435984739713667,
+      "grad_norm": 0.6176229748268995,
+      "learning_rate": 7.345037204779582e-06,
+      "loss": 0.3373,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3644418919473274,
+      "grad_norm": 0.6269304182499377,
+      "learning_rate": 7.343863616470268e-06,
+      "loss": 0.3176,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36452393649751813,
+      "grad_norm": 0.6006855761953981,
+      "learning_rate": 7.3426898626410725e-06,
+      "loss": 0.3429,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3646059810477089,
+      "grad_norm": 0.6547044309892195,
+      "learning_rate": 7.341515943374883e-06,
+      "loss": 0.339,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36468802559789965,
+      "grad_norm": 0.6725113197848159,
+      "learning_rate": 7.340341858754599e-06,
+      "loss": 0.3869,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36477007014809043,
+      "grad_norm": 0.6147373243573265,
+      "learning_rate": 7.339167608863131e-06,
+      "loss": 0.3664,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36485211469828116,
+      "grad_norm": 0.6375180721352509,
+      "learning_rate": 7.337993193783408e-06,
+      "loss": 0.343,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36493415924847195,
+      "grad_norm": 0.6147827627242128,
+      "learning_rate": 7.336818613598361e-06,
+      "loss": 0.3417,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3650162037986627,
+      "grad_norm": 0.6020835506477966,
+      "learning_rate": 7.335643868390938e-06,
+      "loss": 0.3715,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3650982483488534,
+      "grad_norm": 0.6260595375567823,
+      "learning_rate": 7.3344689582440955e-06,
+      "loss": 0.3353,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3651802928990442,
+      "grad_norm": 0.6366009324746107,
+      "learning_rate": 7.333293883240808e-06,
+      "loss": 0.3632,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3652623374492349,
+      "grad_norm": 0.600599193340005,
+      "learning_rate": 7.332118643464053e-06,
+      "loss": 0.368,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3653443819994257,
+      "grad_norm": 0.6096312564988314,
+      "learning_rate": 7.330943238996828e-06,
+      "loss": 0.3594,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36542642654961643,
+      "grad_norm": 0.5725190989470458,
+      "learning_rate": 7.329767669922137e-06,
+      "loss": 0.3798,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3655084710998072,
+      "grad_norm": 0.6150599889606586,
+      "learning_rate": 7.328591936322995e-06,
+      "loss": 0.373,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36559051564999795,
+      "grad_norm": 0.6424906307449497,
+      "learning_rate": 7.327416038282431e-06,
+      "loss": 0.3719,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3656725602001887,
+      "grad_norm": 0.6309647227282019,
+      "learning_rate": 7.326239975883487e-06,
+      "loss": 0.3472,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36575460475037946,
+      "grad_norm": 0.8904847635425932,
+      "learning_rate": 7.325063749209212e-06,
+      "loss": 0.3594,
+      "step": 4458
+    },
+    {
+      "epoch": 0.3658366493005702,
+      "grad_norm": 0.6238159381357332,
+      "learning_rate": 7.323887358342672e-06,
+      "loss": 0.3492,
+      "step": 4459
+    },
+    {
+      "epoch": 0.365918693850761,
+      "grad_norm": 0.6487635344976724,
+      "learning_rate": 7.3227108033669394e-06,
+      "loss": 0.3564,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3660007384009517,
+      "grad_norm": 0.6284846517162695,
+      "learning_rate": 7.321534084365101e-06,
+      "loss": 0.3225,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3660827829511425,
+      "grad_norm": 0.7125498405467358,
+      "learning_rate": 7.320357201420257e-06,
+      "loss": 0.3191,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3661648275013332,
+      "grad_norm": 0.7099434655516514,
+      "learning_rate": 7.319180154615515e-06,
+      "loss": 0.3757,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36624687205152395,
+      "grad_norm": 0.5976533147164482,
+      "learning_rate": 7.318002944033995e-06,
+      "loss": 0.3424,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36632891660171474,
+      "grad_norm": 0.6213979599963461,
+      "learning_rate": 7.316825569758833e-06,
+      "loss": 0.3418,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36641096115190547,
+      "grad_norm": 0.6734566194971553,
+      "learning_rate": 7.31564803187317e-06,
+      "loss": 0.3735,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36649300570209625,
+      "grad_norm": 0.6388291634179296,
+      "learning_rate": 7.314470330460164e-06,
+      "loss": 0.4059,
+      "step": 4467
+    },
+    {
+      "epoch": 0.366575050252287,
+      "grad_norm": 0.6581903859379095,
+      "learning_rate": 7.313292465602983e-06,
+      "loss": 0.4027,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36665709480247777,
+      "grad_norm": 0.6446268187717511,
+      "learning_rate": 7.312114437384804e-06,
+      "loss": 0.3409,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3667391393526685,
+      "grad_norm": 0.6318880958570701,
+      "learning_rate": 7.310936245888818e-06,
+      "loss": 0.3481,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3668211839028592,
+      "grad_norm": 0.6527347820093603,
+      "learning_rate": 7.309757891198227e-06,
+      "loss": 0.3836,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36690322845305,
+      "grad_norm": 0.6569228036941724,
+      "learning_rate": 7.308579373396244e-06,
+      "loss": 0.3233,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36698527300324074,
+      "grad_norm": 0.6539466341023623,
+      "learning_rate": 7.307400692566096e-06,
+      "loss": 0.3439,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3670673175534315,
+      "grad_norm": 0.6297798781630164,
+      "learning_rate": 7.306221848791016e-06,
+      "loss": 0.3494,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36714936210362226,
+      "grad_norm": 0.612803921478661,
+      "learning_rate": 7.305042842154256e-06,
+      "loss": 0.3523,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36723140665381304,
+      "grad_norm": 0.7093418937619914,
+      "learning_rate": 7.3038636727390735e-06,
+      "loss": 0.3521,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36731345120400377,
+      "grad_norm": 0.6346448344765775,
+      "learning_rate": 7.302684340628741e-06,
+      "loss": 0.3627,
+      "step": 4477
+    },
+    {
+      "epoch": 0.3673954957541945,
+      "grad_norm": 0.6525679523046496,
+      "learning_rate": 7.301504845906539e-06,
+      "loss": 0.3586,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3674775403043853,
+      "grad_norm": 0.5715750574327396,
+      "learning_rate": 7.300325188655762e-06,
+      "loss": 0.3343,
+      "step": 4479
+    },
+    {
+      "epoch": 0.367559584854576,
+      "grad_norm": 0.6305398437290901,
+      "learning_rate": 7.299145368959716e-06,
+      "loss": 0.318,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3676416294047668,
+      "grad_norm": 0.6584126996205908,
+      "learning_rate": 7.297965386901717e-06,
+      "loss": 0.374,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36772367395495753,
+      "grad_norm": 0.6037557676825892,
+      "learning_rate": 7.296785242565096e-06,
+      "loss": 0.3263,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3678057185051483,
+      "grad_norm": 0.8780942747403893,
+      "learning_rate": 7.295604936033191e-06,
+      "loss": 0.3379,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36788776305533905,
+      "grad_norm": 0.6140191241266739,
+      "learning_rate": 7.294424467389354e-06,
+      "loss": 0.3244,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3679698076055298,
+      "grad_norm": 0.6342944379965634,
+      "learning_rate": 7.293243836716946e-06,
+      "loss": 0.3721,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36805185215572056,
+      "grad_norm": 0.632325675135138,
+      "learning_rate": 7.292063044099342e-06,
+      "loss": 0.3548,
+      "step": 4486
+    },
+    {
+      "epoch": 0.3681338967059113,
+      "grad_norm": 0.6303741355678775,
+      "learning_rate": 7.29088208961993e-06,
+      "loss": 0.4026,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3682159412561021,
+      "grad_norm": 0.6377357492070529,
+      "learning_rate": 7.289700973362105e-06,
+      "loss": 0.3916,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3682979858062928,
+      "grad_norm": 0.7126510202866146,
+      "learning_rate": 7.288519695409276e-06,
+      "loss": 0.3407,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3683800303564836,
+      "grad_norm": 0.6605418186646366,
+      "learning_rate": 7.287338255844862e-06,
+      "loss": 0.3575,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3684620749066743,
+      "grad_norm": 0.6037327318495305,
+      "learning_rate": 7.286156654752297e-06,
+      "loss": 0.3122,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3685441194568651,
+      "grad_norm": 0.676753591528034,
+      "learning_rate": 7.284974892215021e-06,
+      "loss": 0.3427,
+      "step": 4492
+    },
+    {
+      "epoch": 0.36862616400705583,
+      "grad_norm": 0.62955802748041,
+      "learning_rate": 7.28379296831649e-06,
+      "loss": 0.3321,
+      "step": 4493
+    },
+    {
+      "epoch": 0.36870820855724656,
+      "grad_norm": 0.6434062289016449,
+      "learning_rate": 7.282610883140166e-06,
+      "loss": 0.3526,
+      "step": 4494
+    },
+    {
+      "epoch": 0.36879025310743735,
+      "grad_norm": 0.6123325529613416,
+      "learning_rate": 7.2814286367695315e-06,
+      "loss": 0.356,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3688722976576281,
+      "grad_norm": 0.645541094768656,
+      "learning_rate": 7.280246229288073e-06,
+      "loss": 0.3197,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36895434220781886,
+      "grad_norm": 0.6481948085789633,
+      "learning_rate": 7.2790636607792875e-06,
+      "loss": 0.3468,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3690363867580096,
+      "grad_norm": 0.6816806057956345,
+      "learning_rate": 7.2778809313266885e-06,
+      "loss": 0.4115,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3691184313082004,
+      "grad_norm": 0.7121846019766345,
+      "learning_rate": 7.276698041013797e-06,
+      "loss": 0.3893,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3692004758583911,
+      "grad_norm": 0.6462996314843406,
+      "learning_rate": 7.275514989924147e-06,
+      "loss": 0.3491,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36928252040858184,
+      "grad_norm": 0.6637095542908347,
+      "learning_rate": 7.274331778141286e-06,
+      "loss": 0.3506,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3693645649587726,
+      "grad_norm": 0.6533285104217263,
+      "learning_rate": 7.273148405748768e-06,
+      "loss": 0.3178,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36944660950896335,
+      "grad_norm": 0.6283698751010459,
+      "learning_rate": 7.271964872830161e-06,
+      "loss": 0.3931,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36952865405915414,
+      "grad_norm": 0.6415766094456171,
+      "learning_rate": 7.2707811794690465e-06,
+      "loss": 0.3405,
+      "step": 4504
+    },
+    {
+      "epoch": 0.36961069860934487,
+      "grad_norm": 0.6485447339297795,
+      "learning_rate": 7.269597325749013e-06,
+      "loss": 0.3474,
+      "step": 4505
+    },
+    {
+      "epoch": 0.36969274315953565,
+      "grad_norm": 0.6423858484568417,
+      "learning_rate": 7.268413311753661e-06,
+      "loss": 0.3251,
+      "step": 4506
+    },
+    {
+      "epoch": 0.3697747877097264,
+      "grad_norm": 0.6332851366570621,
+      "learning_rate": 7.267229137566607e-06,
+      "loss": 0.3576,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3698568322599171,
+      "grad_norm": 0.6080020232116587,
+      "learning_rate": 7.266044803271473e-06,
+      "loss": 0.3521,
+      "step": 4508
+    },
+    {
+      "epoch": 0.3699388768101079,
+      "grad_norm": 0.65694068293177,
+      "learning_rate": 7.264860308951897e-06,
+      "loss": 0.3442,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3700209213602986,
+      "grad_norm": 0.646239747107836,
+      "learning_rate": 7.263675654691524e-06,
+      "loss": 0.3203,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3701029659104894,
+      "grad_norm": 0.6700003462171,
+      "learning_rate": 7.262490840574013e-06,
+      "loss": 0.3475,
+      "step": 4511
+    },
+    {
+      "epoch": 0.37018501046068014,
+      "grad_norm": 0.6159476581278434,
+      "learning_rate": 7.261305866683034e-06,
+      "loss": 0.3263,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3702670550108709,
+      "grad_norm": 0.6213180382365286,
+      "learning_rate": 7.260120733102267e-06,
+      "loss": 0.3784,
+      "step": 4513
+    },
+    {
+      "epoch": 0.37034909956106166,
+      "grad_norm": 0.6830069200113369,
+      "learning_rate": 7.258935439915407e-06,
+      "loss": 0.3949,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3704311441112524,
+      "grad_norm": 0.6471402647261272,
+      "learning_rate": 7.257749987206154e-06,
+      "loss": 0.3532,
+      "step": 4515
+    },
+    {
+      "epoch": 0.37051318866144317,
+      "grad_norm": 0.6441136377526595,
+      "learning_rate": 7.2565643750582235e-06,
+      "loss": 0.3475,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3705952332116339,
+      "grad_norm": 0.623062700862877,
+      "learning_rate": 7.255378603555344e-06,
+      "loss": 0.3481,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3706772777618247,
+      "grad_norm": 0.6467878762655555,
+      "learning_rate": 7.25419267278125e-06,
+      "loss": 0.3687,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3707593223120154,
+      "grad_norm": 0.612524860095522,
+      "learning_rate": 7.253006582819692e-06,
+      "loss": 0.3392,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3708413668622062,
+      "grad_norm": 0.6119624838557122,
+      "learning_rate": 7.251820333754428e-06,
+      "loss": 0.3838,
+      "step": 4520
+    },
+    {
+      "epoch": 0.37092341141239693,
+      "grad_norm": 0.6299302769517757,
+      "learning_rate": 7.2506339256692295e-06,
+      "loss": 0.3159,
+      "step": 4521
+    },
+    {
+      "epoch": 0.37100545596258766,
+      "grad_norm": 0.6485023649539763,
+      "learning_rate": 7.24944735864788e-06,
+      "loss": 0.3354,
+      "step": 4522
+    },
+    {
+      "epoch": 0.37108750051277845,
+      "grad_norm": 0.6171984771550416,
+      "learning_rate": 7.248260632774171e-06,
+      "loss": 0.3481,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3711695450629692,
+      "grad_norm": 0.6228146720681734,
+      "learning_rate": 7.247073748131911e-06,
+      "loss": 0.32,
+      "step": 4524
+    },
+    {
+      "epoch": 0.37125158961315996,
+      "grad_norm": 0.6221656835005879,
+      "learning_rate": 7.245886704804911e-06,
+      "loss": 0.353,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3713336341633507,
+      "grad_norm": 0.6792378864310321,
+      "learning_rate": 7.244699502877e-06,
+      "loss": 0.3556,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3714156787135415,
+      "grad_norm": 0.5551696875630208,
+      "learning_rate": 7.243512142432017e-06,
+      "loss": 0.3294,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3714977232637322,
+      "grad_norm": 0.6636198359313248,
+      "learning_rate": 7.242324623553809e-06,
+      "loss": 0.4055,
+      "step": 4528
+    },
+    {
+      "epoch": 0.37157976781392293,
+      "grad_norm": 0.6103341228489678,
+      "learning_rate": 7.2411369463262415e-06,
+      "loss": 0.3515,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3716618123641137,
+      "grad_norm": 0.6229334169396119,
+      "learning_rate": 7.239949110833182e-06,
+      "loss": 0.3808,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37174385691430445,
+      "grad_norm": 0.6076721264717494,
+      "learning_rate": 7.238761117158516e-06,
+      "loss": 0.3578,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37182590146449523,
+      "grad_norm": 0.5840020162022859,
+      "learning_rate": 7.237572965386136e-06,
+      "loss": 0.3459,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37190794601468596,
+      "grad_norm": 0.6344823705302016,
+      "learning_rate": 7.236384655599948e-06,
+      "loss": 0.3802,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37198999056487675,
+      "grad_norm": 0.6337986228321436,
+      "learning_rate": 7.235196187883869e-06,
+      "loss": 0.3626,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3720720351150675,
+      "grad_norm": 0.6160054787549931,
+      "learning_rate": 7.234007562321826e-06,
+      "loss": 0.3452,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3721540796652582,
+      "grad_norm": 0.6027258657964668,
+      "learning_rate": 7.232818778997759e-06,
+      "loss": 0.3343,
+      "step": 4536
+    },
+    {
+      "epoch": 0.372236124215449,
+      "grad_norm": 0.6025708834732685,
+      "learning_rate": 7.231629837995616e-06,
+      "loss": 0.3879,
+      "step": 4537
+    },
+    {
+      "epoch": 0.3723181687656397,
+      "grad_norm": 0.6347990496239765,
+      "learning_rate": 7.230440739399361e-06,
+      "loss": 0.3697,
+      "step": 4538
+    },
+    {
+      "epoch": 0.3724002133158305,
+      "grad_norm": 0.6455741405465947,
+      "learning_rate": 7.229251483292963e-06,
+      "loss": 0.3507,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37248225786602124,
+      "grad_norm": 0.6236592684346532,
+      "learning_rate": 7.228062069760407e-06,
+      "loss": 0.3182,
+      "step": 4540
+    },
+    {
+      "epoch": 0.372564302416212,
+      "grad_norm": 0.7002629926765556,
+      "learning_rate": 7.226872498885687e-06,
+      "loss": 0.3389,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37264634696640275,
+      "grad_norm": 0.6246921403067516,
+      "learning_rate": 7.225682770752809e-06,
+      "loss": 0.3435,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37272839151659354,
+      "grad_norm": 0.629490115673483,
+      "learning_rate": 7.22449288544579e-06,
+      "loss": 0.3249,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37281043606678427,
+      "grad_norm": 0.6187782526746847,
+      "learning_rate": 7.223302843048657e-06,
+      "loss": 0.3396,
+      "step": 4544
+    },
+    {
+      "epoch": 0.372892480616975,
+      "grad_norm": 0.5641704033534676,
+      "learning_rate": 7.222112643645449e-06,
+      "loss": 0.3389,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3729745251671658,
+      "grad_norm": 0.6602707855972512,
+      "learning_rate": 7.220922287320218e-06,
+      "loss": 0.357,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3730565697173565,
+      "grad_norm": 0.6680970688849711,
+      "learning_rate": 7.219731774157021e-06,
+      "loss": 0.3829,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3731386142675473,
+      "grad_norm": 0.6242002413107802,
+      "learning_rate": 7.218541104239932e-06,
+      "loss": 0.293,
+      "step": 4548
+    },
+    {
+      "epoch": 0.373220658817738,
+      "grad_norm": 0.6588583150388905,
+      "learning_rate": 7.217350277653037e-06,
+      "loss": 0.3369,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3733027033679288,
+      "grad_norm": 0.6812972169835476,
+      "learning_rate": 7.216159294480426e-06,
+      "loss": 0.3709,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37338474791811954,
+      "grad_norm": 0.6230288394072228,
+      "learning_rate": 7.214968154806207e-06,
+      "loss": 0.3355,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37346679246831027,
+      "grad_norm": 0.6557079677485985,
+      "learning_rate": 7.213776858714496e-06,
+      "loss": 0.3133,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37354883701850106,
+      "grad_norm": 0.6520820953296452,
+      "learning_rate": 7.2125854062894184e-06,
+      "loss": 0.3159,
+      "step": 4553
+    },
+    {
+      "epoch": 0.3736308815686918,
+      "grad_norm": 0.654679771493733,
+      "learning_rate": 7.211393797615115e-06,
+      "loss": 0.3675,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37371292611888257,
+      "grad_norm": 0.6717372182310656,
+      "learning_rate": 7.2102020327757346e-06,
+      "loss": 0.3959,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3737949706690733,
+      "grad_norm": 0.6248080509617369,
+      "learning_rate": 7.209010111855438e-06,
+      "loss": 0.372,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3738770152192641,
+      "grad_norm": 0.670813213592887,
+      "learning_rate": 7.2078180349383954e-06,
+      "loss": 0.339,
+      "step": 4557
+    },
+    {
+      "epoch": 0.3739590597694548,
+      "grad_norm": 0.6319216197094404,
+      "learning_rate": 7.206625802108791e-06,
+      "loss": 0.3603,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37404110431964555,
+      "grad_norm": 0.6414289807226805,
+      "learning_rate": 7.205433413450818e-06,
+      "loss": 0.3521,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37412314886983633,
+      "grad_norm": 0.6101991972449019,
+      "learning_rate": 7.20424086904868e-06,
+      "loss": 0.3246,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37420519342002706,
+      "grad_norm": 0.6387235177142778,
+      "learning_rate": 7.203048168986594e-06,
+      "loss": 0.3409,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37428723797021785,
+      "grad_norm": 0.5927854054360885,
+      "learning_rate": 7.201855313348783e-06,
+      "loss": 0.3499,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3743692825204086,
+      "grad_norm": 0.6323178301152805,
+      "learning_rate": 7.20066230221949e-06,
+      "loss": 0.3365,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37445132707059936,
+      "grad_norm": 0.6396998104837491,
+      "learning_rate": 7.199469135682961e-06,
+      "loss": 0.3565,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3745333716207901,
+      "grad_norm": 0.6136634813664233,
+      "learning_rate": 7.1982758138234545e-06,
+      "loss": 0.3449,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3746154161709808,
+      "grad_norm": 0.6681420354874046,
+      "learning_rate": 7.197082336725242e-06,
+      "loss": 0.3506,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3746974607211716,
+      "grad_norm": 0.6622153404679221,
+      "learning_rate": 7.195888704472604e-06,
+      "loss": 0.349,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37477950527136233,
+      "grad_norm": 0.6131075384216026,
+      "learning_rate": 7.194694917149835e-06,
+      "loss": 0.3279,
+      "step": 4568
+    },
+    {
+      "epoch": 0.3748615498215531,
+      "grad_norm": 0.5840939478889764,
+      "learning_rate": 7.193500974841237e-06,
+      "loss": 0.3396,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37494359437174385,
+      "grad_norm": 0.5936414153069233,
+      "learning_rate": 7.192306877631123e-06,
+      "loss": 0.3113,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37502563892193463,
+      "grad_norm": 0.6243617257217913,
+      "learning_rate": 7.19111262560382e-06,
+      "loss": 0.3389,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37510768347212536,
+      "grad_norm": 0.7087472663658257,
+      "learning_rate": 7.1899182188436645e-06,
+      "loss": 0.3553,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3751897280223161,
+      "grad_norm": 0.6266643024705528,
+      "learning_rate": 7.1887236574350014e-06,
+      "loss": 0.3163,
+      "step": 4573
+    },
+    {
+      "epoch": 0.3752717725725069,
+      "grad_norm": 0.6284290215169902,
+      "learning_rate": 7.187528941462192e-06,
+      "loss": 0.3632,
+      "step": 4574
+    },
+    {
+      "epoch": 0.3753538171226976,
+      "grad_norm": 0.5999578501799753,
+      "learning_rate": 7.186334071009601e-06,
+      "loss": 0.3751,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3754358616728884,
+      "grad_norm": 0.6131897245003688,
+      "learning_rate": 7.185139046161611e-06,
+      "loss": 0.3704,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3755179062230791,
+      "grad_norm": 0.636853065338729,
+      "learning_rate": 7.183943867002613e-06,
+      "loss": 0.3467,
+      "step": 4577
+    },
+    {
+      "epoch": 0.3755999507732699,
+      "grad_norm": 0.669880208949198,
+      "learning_rate": 7.182748533617008e-06,
+      "loss": 0.3285,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37568199532346064,
+      "grad_norm": 0.6513621866812953,
+      "learning_rate": 7.181553046089208e-06,
+      "loss": 0.3704,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37576403987365137,
+      "grad_norm": 0.618763331950323,
+      "learning_rate": 7.180357404503637e-06,
+      "loss": 0.3433,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37584608442384215,
+      "grad_norm": 0.6738839395112313,
+      "learning_rate": 7.179161608944729e-06,
+      "loss": 0.3711,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3759281289740329,
+      "grad_norm": 0.6056175326675547,
+      "learning_rate": 7.1779656594969294e-06,
+      "loss": 0.3398,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37601017352422367,
+      "grad_norm": 0.6698336359043113,
+      "learning_rate": 7.176769556244694e-06,
+      "loss": 0.3642,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3760922180744144,
+      "grad_norm": 0.6457454653779154,
+      "learning_rate": 7.175573299272489e-06,
+      "loss": 0.3521,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3761742626246052,
+      "grad_norm": 0.5851184146503637,
+      "learning_rate": 7.174376888664793e-06,
+      "loss": 0.3695,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3762563071747959,
+      "grad_norm": 0.650207789781782,
+      "learning_rate": 7.1731803245060975e-06,
+      "loss": 0.3899,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37633835172498664,
+      "grad_norm": 0.6546627464387019,
+      "learning_rate": 7.171983606880897e-06,
+      "loss": 0.3701,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3764203962751774,
+      "grad_norm": 0.6245754979339774,
+      "learning_rate": 7.1707867358737025e-06,
+      "loss": 0.3422,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37650244082536816,
+      "grad_norm": 0.6183531432787688,
+      "learning_rate": 7.169589711569036e-06,
+      "loss": 0.3834,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37658448537555894,
+      "grad_norm": 0.6762340236268533,
+      "learning_rate": 7.168392534051432e-06,
+      "loss": 0.3893,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37666652992574967,
+      "grad_norm": 0.6109926256002997,
+      "learning_rate": 7.16719520340543e-06,
+      "loss": 0.344,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37674857447594046,
+      "grad_norm": 0.6408347195822748,
+      "learning_rate": 7.165997719715585e-06,
+      "loss": 0.3359,
+      "step": 4592
+    },
+    {
+      "epoch": 0.3768306190261312,
+      "grad_norm": 0.6196138029302344,
+      "learning_rate": 7.164800083066461e-06,
+      "loss": 0.2755,
+      "step": 4593
+    },
+    {
+      "epoch": 0.3769126635763219,
+      "grad_norm": 0.6547185257254503,
+      "learning_rate": 7.163602293542632e-06,
+      "loss": 0.356,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3769947081265127,
+      "grad_norm": 0.6583762322497764,
+      "learning_rate": 7.162404351228687e-06,
+      "loss": 0.3398,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37707675267670343,
+      "grad_norm": 0.6327589236397315,
+      "learning_rate": 7.161206256209219e-06,
+      "loss": 0.3357,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3771587972268942,
+      "grad_norm": 0.6769617055291326,
+      "learning_rate": 7.160008008568838e-06,
+      "loss": 0.3504,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37724084177708495,
+      "grad_norm": 0.6675015343935601,
+      "learning_rate": 7.158809608392162e-06,
+      "loss": 0.3649,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37732288632727573,
+      "grad_norm": 0.6589232453663924,
+      "learning_rate": 7.15761105576382e-06,
+      "loss": 0.3186,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37740493087746646,
+      "grad_norm": 0.6698742371630652,
+      "learning_rate": 7.156412350768453e-06,
+      "loss": 0.3453,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37748697542765725,
+      "grad_norm": 0.6214184043187801,
+      "learning_rate": 7.155213493490708e-06,
+      "loss": 0.3538,
+      "step": 4601
+    },
+    {
+      "epoch": 0.377569019977848,
+      "grad_norm": 0.5598535660592915,
+      "learning_rate": 7.154014484015249e-06,
+      "loss": 0.3201,
+      "step": 4602
+    },
+    {
+      "epoch": 0.3776510645280387,
+      "grad_norm": 0.6053298589460756,
+      "learning_rate": 7.152815322426749e-06,
+      "loss": 0.3025,
+      "step": 4603
+    },
+    {
+      "epoch": 0.3777331090782295,
+      "grad_norm": 0.6043484274724334,
+      "learning_rate": 7.151616008809889e-06,
+      "loss": 0.2942,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3778151536284202,
+      "grad_norm": 0.6384230919752857,
+      "learning_rate": 7.150416543249364e-06,
+      "loss": 0.3584,
+      "step": 4605
+    },
+    {
+      "epoch": 0.377897198178611,
+      "grad_norm": 0.6594304640773905,
+      "learning_rate": 7.149216925829876e-06,
+      "loss": 0.3166,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37797924272880173,
+      "grad_norm": 0.6263459003309572,
+      "learning_rate": 7.148017156636144e-06,
+      "loss": 0.3441,
+      "step": 4607
+    },
+    {
+      "epoch": 0.3780612872789925,
+      "grad_norm": 0.6649856142974321,
+      "learning_rate": 7.14681723575289e-06,
+      "loss": 0.3392,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37814333182918325,
+      "grad_norm": 0.6288700692100119,
+      "learning_rate": 7.145617163264851e-06,
+      "loss": 0.3648,
+      "step": 4609
+    },
+    {
+      "epoch": 0.378225376379374,
+      "grad_norm": 0.6722480482214914,
+      "learning_rate": 7.144416939256777e-06,
+      "loss": 0.3992,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37830742092956476,
+      "grad_norm": 0.6676482889913766,
+      "learning_rate": 7.143216563813422e-06,
+      "loss": 0.3349,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3783894654797555,
+      "grad_norm": 0.6058845333353462,
+      "learning_rate": 7.142016037019558e-06,
+      "loss": 0.3092,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3784715100299463,
+      "grad_norm": 0.6059316919339756,
+      "learning_rate": 7.140815358959963e-06,
+      "loss": 0.3279,
+      "step": 4613
+    },
+    {
+      "epoch": 0.378553554580137,
+      "grad_norm": 0.6179206172006814,
+      "learning_rate": 7.139614529719427e-06,
+      "loss": 0.3244,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3786355991303278,
+      "grad_norm": 0.7038102112651489,
+      "learning_rate": 7.13841354938275e-06,
+      "loss": 0.3445,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3787176436805185,
+      "grad_norm": 0.610168456491401,
+      "learning_rate": 7.137212418034744e-06,
+      "loss": 0.3611,
+      "step": 4616
+    },
+    {
+      "epoch": 0.37879968823070925,
+      "grad_norm": 0.6398608324819129,
+      "learning_rate": 7.13601113576023e-06,
+      "loss": 0.3417,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37888173278090004,
+      "grad_norm": 0.615174691070354,
+      "learning_rate": 7.134809702644042e-06,
+      "loss": 0.3388,
+      "step": 4618
+    },
+    {
+      "epoch": 0.37896377733109077,
+      "grad_norm": 0.6362750750975755,
+      "learning_rate": 7.1336081187710225e-06,
+      "loss": 0.3019,
+      "step": 4619
+    },
+    {
+      "epoch": 0.37904582188128155,
+      "grad_norm": 0.618682997169944,
+      "learning_rate": 7.132406384226026e-06,
+      "loss": 0.3546,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3791278664314723,
+      "grad_norm": 0.6554493974317134,
+      "learning_rate": 7.1312044990939155e-06,
+      "loss": 0.3695,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37920991098166307,
+      "grad_norm": 0.6160912395014112,
+      "learning_rate": 7.130002463459569e-06,
+      "loss": 0.3193,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3792919555318538,
+      "grad_norm": 0.6168319962086525,
+      "learning_rate": 7.128800277407869e-06,
+      "loss": 0.3464,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3793740000820445,
+      "grad_norm": 0.6209990516895828,
+      "learning_rate": 7.127597941023714e-06,
+      "loss": 0.3756,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3794560446322353,
+      "grad_norm": 0.6359743173956799,
+      "learning_rate": 7.12639545439201e-06,
+      "loss": 0.3605,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37953808918242604,
+      "grad_norm": 0.6420938389637031,
+      "learning_rate": 7.125192817597677e-06,
+      "loss": 0.3739,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3796201337326168,
+      "grad_norm": 0.686949918105225,
+      "learning_rate": 7.123990030725641e-06,
+      "loss": 0.345,
+      "step": 4627
+    },
+    {
+      "epoch": 0.37970217828280756,
+      "grad_norm": 0.7394284048511102,
+      "learning_rate": 7.1227870938608424e-06,
+      "loss": 0.3623,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37978422283299834,
+      "grad_norm": 0.6078096466837737,
+      "learning_rate": 7.121584007088227e-06,
+      "loss": 0.3478,
+      "step": 4629
+    },
+    {
+      "epoch": 0.37986626738318907,
+      "grad_norm": 0.6767129742039698,
+      "learning_rate": 7.12038077049276e-06,
+      "loss": 0.3822,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3799483119333798,
+      "grad_norm": 0.6466603985766054,
+      "learning_rate": 7.119177384159408e-06,
+      "loss": 0.3241,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3800303564835706,
+      "grad_norm": 0.6224227990810629,
+      "learning_rate": 7.117973848173155e-06,
+      "loss": 0.3597,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3801124010337613,
+      "grad_norm": 0.6316224364934603,
+      "learning_rate": 7.116770162618992e-06,
+      "loss": 0.3546,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3801944455839521,
+      "grad_norm": 0.6153778948050093,
+      "learning_rate": 7.11556632758192e-06,
+      "loss": 0.3539,
+      "step": 4634
+    },
+    {
+      "epoch": 0.38027649013414283,
+      "grad_norm": 0.6288405494355122,
+      "learning_rate": 7.114362343146953e-06,
+      "loss": 0.3301,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3803585346843336,
+      "grad_norm": 0.69984218722341,
+      "learning_rate": 7.113158209399115e-06,
+      "loss": 0.3596,
+      "step": 4636
+    },
+    {
+      "epoch": 0.38044057923452435,
+      "grad_norm": 0.581289559179693,
+      "learning_rate": 7.111953926423439e-06,
+      "loss": 0.3449,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3805226237847151,
+      "grad_norm": 0.6611346918409942,
+      "learning_rate": 7.11074949430497e-06,
+      "loss": 0.356,
+      "step": 4638
+    },
+    {
+      "epoch": 0.38060466833490586,
+      "grad_norm": 0.7413926830620532,
+      "learning_rate": 7.109544913128761e-06,
+      "loss": 0.3722,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3806867128850966,
+      "grad_norm": 0.6509113468802402,
+      "learning_rate": 7.108340182979881e-06,
+      "loss": 0.3465,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3807687574352874,
+      "grad_norm": 0.6649248102714362,
+      "learning_rate": 7.107135303943406e-06,
+      "loss": 0.3556,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3808508019854781,
+      "grad_norm": 0.6376762570057808,
+      "learning_rate": 7.10593027610442e-06,
+      "loss": 0.3384,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3809328465356689,
+      "grad_norm": 0.6916527744300611,
+      "learning_rate": 7.10472509954802e-06,
+      "loss": 0.343,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3810148910858596,
+      "grad_norm": 0.6830960883510018,
+      "learning_rate": 7.103519774359316e-06,
+      "loss": 0.3233,
+      "step": 4644
+    },
+    {
+      "epoch": 0.38109693563605035,
+      "grad_norm": 0.7235524853796136,
+      "learning_rate": 7.102314300623425e-06,
+      "loss": 0.3613,
+      "step": 4645
+    },
+    {
+      "epoch": 0.38117898018624113,
+      "grad_norm": 0.6483337368142531,
+      "learning_rate": 7.1011086784254765e-06,
+      "loss": 0.3053,
+      "step": 4646
+    },
+    {
+      "epoch": 0.38126102473643186,
+      "grad_norm": 0.6685637610906855,
+      "learning_rate": 7.0999029078506075e-06,
+      "loss": 0.3502,
+      "step": 4647
+    },
+    {
+      "epoch": 0.38134306928662265,
+      "grad_norm": 0.640235014361567,
+      "learning_rate": 7.098696988983971e-06,
+      "loss": 0.3457,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3814251138368134,
+      "grad_norm": 0.6574630403766186,
+      "learning_rate": 7.097490921910724e-06,
+      "loss": 0.338,
+      "step": 4649
+    },
+    {
+      "epoch": 0.38150715838700416,
+      "grad_norm": 0.6313317937295632,
+      "learning_rate": 7.096284706716037e-06,
+      "loss": 0.3408,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3815892029371949,
+      "grad_norm": 0.6500644362236917,
+      "learning_rate": 7.095078343485092e-06,
+      "loss": 0.3596,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3816712474873856,
+      "grad_norm": 0.6383067557191806,
+      "learning_rate": 7.0938718323030805e-06,
+      "loss": 0.3444,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3817532920375764,
+      "grad_norm": 0.6456607901486484,
+      "learning_rate": 7.092665173255205e-06,
+      "loss": 0.3381,
+      "step": 4653
+    },
+    {
+      "epoch": 0.38183533658776714,
+      "grad_norm": 0.6902482327208974,
+      "learning_rate": 7.0914583664266765e-06,
+      "loss": 0.3765,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3819173811379579,
+      "grad_norm": 0.6401415544028507,
+      "learning_rate": 7.090251411902719e-06,
+      "loss": 0.3832,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38199942568814865,
+      "grad_norm": 0.7117352057134897,
+      "learning_rate": 7.089044309768565e-06,
+      "loss": 0.3108,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38208147023833944,
+      "grad_norm": 0.7013307059044207,
+      "learning_rate": 7.087837060109457e-06,
+      "loss": 0.3717,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38216351478853017,
+      "grad_norm": 0.6539154226430143,
+      "learning_rate": 7.086629663010651e-06,
+      "loss": 0.3414,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38224555933872095,
+      "grad_norm": 0.6114126278442333,
+      "learning_rate": 7.085422118557411e-06,
+      "loss": 0.358,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3823276038889117,
+      "grad_norm": 0.5974801472897218,
+      "learning_rate": 7.08421442683501e-06,
+      "loss": 0.3345,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3824096484391024,
+      "grad_norm": 0.6527621455724205,
+      "learning_rate": 7.083006587928736e-06,
+      "loss": 0.3188,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3824916929892932,
+      "grad_norm": 0.6575883877036462,
+      "learning_rate": 7.081798601923883e-06,
+      "loss": 0.3597,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3825737375394839,
+      "grad_norm": 0.6449736472558852,
+      "learning_rate": 7.08059046890576e-06,
+      "loss": 0.3206,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3826557820896747,
+      "grad_norm": 0.6796440147154437,
+      "learning_rate": 7.079382188959677e-06,
+      "loss": 0.3823,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38273782663986544,
+      "grad_norm": 0.6072123729904314,
+      "learning_rate": 7.078173762170966e-06,
+      "loss": 0.371,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3828198711900562,
+      "grad_norm": 0.6496195573907408,
+      "learning_rate": 7.076965188624963e-06,
+      "loss": 0.3496,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38290191574024696,
+      "grad_norm": 0.5814918922101804,
+      "learning_rate": 7.075756468407014e-06,
+      "loss": 0.3441,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3829839602904377,
+      "grad_norm": 0.6598857082746105,
+      "learning_rate": 7.074547601602479e-06,
+      "loss": 0.3161,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38306600484062847,
+      "grad_norm": 0.5799553260545802,
+      "learning_rate": 7.073338588296725e-06,
+      "loss": 0.3405,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3831480493908192,
+      "grad_norm": 0.6362750059433098,
+      "learning_rate": 7.072129428575132e-06,
+      "loss": 0.3361,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38323009394101,
+      "grad_norm": 0.6853901222767201,
+      "learning_rate": 7.070920122523088e-06,
+      "loss": 0.3813,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3833121384912007,
+      "grad_norm": 0.6801754449818809,
+      "learning_rate": 7.069710670225991e-06,
+      "loss": 0.3949,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3833941830413915,
+      "grad_norm": 0.6987238566839218,
+      "learning_rate": 7.068501071769253e-06,
+      "loss": 0.3957,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38347622759158223,
+      "grad_norm": 0.7144356373853752,
+      "learning_rate": 7.06729132723829e-06,
+      "loss": 0.3625,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38355827214177296,
+      "grad_norm": 0.7051687642855712,
+      "learning_rate": 7.066081436718537e-06,
+      "loss": 0.3564,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38364031669196375,
+      "grad_norm": 0.6146888317710476,
+      "learning_rate": 7.0648714002954336e-06,
+      "loss": 0.3516,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3837223612421545,
+      "grad_norm": 0.6692583695599219,
+      "learning_rate": 7.063661218054427e-06,
+      "loss": 0.3566,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38380440579234526,
+      "grad_norm": 0.6948625367916874,
+      "learning_rate": 7.0624508900809815e-06,
+      "loss": 0.3447,
+      "step": 4678
+    },
+    {
+      "epoch": 0.383886450342536,
+      "grad_norm": 1.1970696329767296,
+      "learning_rate": 7.061240416460568e-06,
+      "loss": 0.3924,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3839684948927268,
+      "grad_norm": 0.7092447645197281,
+      "learning_rate": 7.060029797278667e-06,
+      "loss": 0.3454,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3840505394429175,
+      "grad_norm": 0.6523569088896445,
+      "learning_rate": 7.058819032620773e-06,
+      "loss": 0.3643,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38413258399310823,
+      "grad_norm": 0.6093158175579021,
+      "learning_rate": 7.0576081225723855e-06,
+      "loss": 0.3435,
+      "step": 4682
+    },
+    {
+      "epoch": 0.384214628543299,
+      "grad_norm": 0.6535634717504762,
+      "learning_rate": 7.0563970672190185e-06,
+      "loss": 0.329,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38429667309348975,
+      "grad_norm": 0.6168477146793439,
+      "learning_rate": 7.055185866646195e-06,
+      "loss": 0.31,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38437871764368053,
+      "grad_norm": 0.6869006389820373,
+      "learning_rate": 7.053974520939446e-06,
+      "loss": 0.3285,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38446076219387126,
+      "grad_norm": 0.6826708065766045,
+      "learning_rate": 7.052763030184317e-06,
+      "loss": 0.3384,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38454280674406205,
+      "grad_norm": 0.6670061317108733,
+      "learning_rate": 7.05155139446636e-06,
+      "loss": 0.3969,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3846248512942528,
+      "grad_norm": 0.6452404811563304,
+      "learning_rate": 7.050339613871141e-06,
+      "loss": 0.365,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3847068958444435,
+      "grad_norm": 0.6088569946308204,
+      "learning_rate": 7.049127688484232e-06,
+      "loss": 0.3779,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3847889403946343,
+      "grad_norm": 0.717471665958166,
+      "learning_rate": 7.047915618391219e-06,
+      "loss": 0.3481,
+      "step": 4690
+    },
+    {
+      "epoch": 0.384870984944825,
+      "grad_norm": 0.6385994861818417,
+      "learning_rate": 7.0467034036776945e-06,
+      "loss": 0.324,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3849530294950158,
+      "grad_norm": 0.597639356115617,
+      "learning_rate": 7.0454910444292634e-06,
+      "loss": 0.3428,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38503507404520654,
+      "grad_norm": 0.7258858202217039,
+      "learning_rate": 7.044278540731542e-06,
+      "loss": 0.3789,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3851171185953973,
+      "grad_norm": 0.6327011780901894,
+      "learning_rate": 7.043065892670154e-06,
+      "loss": 0.3718,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38519916314558805,
+      "grad_norm": 0.6614184155300179,
+      "learning_rate": 7.041853100330733e-06,
+      "loss": 0.3357,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3852812076957788,
+      "grad_norm": 0.6409723023988451,
+      "learning_rate": 7.04064016379893e-06,
+      "loss": 0.3501,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38536325224596957,
+      "grad_norm": 0.6299453587421379,
+      "learning_rate": 7.039427083160395e-06,
+      "loss": 0.3331,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3854452967961603,
+      "grad_norm": 0.6209352033359549,
+      "learning_rate": 7.038213858500796e-06,
+      "loss": 0.3469,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3855273413463511,
+      "grad_norm": 0.6283319564526384,
+      "learning_rate": 7.0370004899058095e-06,
+      "loss": 0.33,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3856093858965418,
+      "grad_norm": 0.6136776872930106,
+      "learning_rate": 7.035786977461122e-06,
+      "loss": 0.3207,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3856914304467326,
+      "grad_norm": 0.6189912581070234,
+      "learning_rate": 7.034573321252427e-06,
+      "loss": 0.3137,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3857734749969233,
+      "grad_norm": 0.6604956207653381,
+      "learning_rate": 7.033359521365434e-06,
+      "loss": 0.3614,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38585551954711406,
+      "grad_norm": 0.5955116122108765,
+      "learning_rate": 7.032145577885858e-06,
+      "loss": 0.3467,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38593756409730484,
+      "grad_norm": 0.7040972731573512,
+      "learning_rate": 7.030931490899427e-06,
+      "loss": 0.4079,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38601960864749557,
+      "grad_norm": 0.6333611660811831,
+      "learning_rate": 7.029717260491876e-06,
+      "loss": 0.3484,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38610165319768636,
+      "grad_norm": 0.6336255809191544,
+      "learning_rate": 7.0285028867489535e-06,
+      "loss": 0.34,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3861836977478771,
+      "grad_norm": 0.6700850190754825,
+      "learning_rate": 7.0272883697564175e-06,
+      "loss": 0.3914,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38626574229806787,
+      "grad_norm": 0.6159136465534256,
+      "learning_rate": 7.026073709600032e-06,
+      "loss": 0.3595,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3863477868482586,
+      "grad_norm": 0.6063251467769445,
+      "learning_rate": 7.024858906365577e-06,
+      "loss": 0.3712,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38642983139844933,
+      "grad_norm": 0.6525735740234816,
+      "learning_rate": 7.02364396013884e-06,
+      "loss": 0.339,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3865118759486401,
+      "grad_norm": 0.6193555781656128,
+      "learning_rate": 7.022428871005617e-06,
+      "loss": 0.3366,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38659392049883085,
+      "grad_norm": 0.6214093320753337,
+      "learning_rate": 7.021213639051716e-06,
+      "loss": 0.3427,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38667596504902163,
+      "grad_norm": 0.6449756539075978,
+      "learning_rate": 7.019998264362955e-06,
+      "loss": 0.3657,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38675800959921236,
+      "grad_norm": 0.6193760337652469,
+      "learning_rate": 7.018782747025161e-06,
+      "loss": 0.366,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38684005414940315,
+      "grad_norm": 0.6606956222650053,
+      "learning_rate": 7.017567087124175e-06,
+      "loss": 0.3666,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3869220986995939,
+      "grad_norm": 0.6263676130528987,
+      "learning_rate": 7.016351284745842e-06,
+      "loss": 0.3492,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38700414324978466,
+      "grad_norm": 0.6822459280734642,
+      "learning_rate": 7.015135339976021e-06,
+      "loss": 0.3978,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3870861877999754,
+      "grad_norm": 0.7894163223922203,
+      "learning_rate": 7.0139192529005805e-06,
+      "loss": 0.3681,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3871682323501661,
+      "grad_norm": 0.6662535739087185,
+      "learning_rate": 7.012703023605396e-06,
+      "loss": 0.3601,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3872502769003569,
+      "grad_norm": 0.6702994622159525,
+      "learning_rate": 7.011486652176359e-06,
+      "loss": 0.3638,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38733232145054763,
+      "grad_norm": 0.6953643805799282,
+      "learning_rate": 7.010270138699367e-06,
+      "loss": 0.3292,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3874143660007384,
+      "grad_norm": 0.6415383661506523,
+      "learning_rate": 7.0090534832603266e-06,
+      "loss": 0.3601,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38749641055092915,
+      "grad_norm": 0.6269068794905951,
+      "learning_rate": 7.007836685945155e-06,
+      "loss": 0.3396,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38757845510111993,
+      "grad_norm": 0.6914735298129079,
+      "learning_rate": 7.006619746839787e-06,
+      "loss": 0.3381,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38766049965131066,
+      "grad_norm": 0.6384224010476345,
+      "learning_rate": 7.0054026660301535e-06,
+      "loss": 0.3681,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3877425442015014,
+      "grad_norm": 0.7097910223925299,
+      "learning_rate": 7.004185443602206e-06,
+      "loss": 0.3402,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3878245887516922,
+      "grad_norm": 0.6776094609596632,
+      "learning_rate": 7.002968079641902e-06,
+      "loss": 0.3071,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3879066333018829,
+      "grad_norm": 0.7001188141864748,
+      "learning_rate": 7.00175057423521e-06,
+      "loss": 0.3253,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3879886778520737,
+      "grad_norm": 0.6861247189476901,
+      "learning_rate": 7.000532927468109e-06,
+      "loss": 0.3597,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3880707224022644,
+      "grad_norm": 0.6300299473712486,
+      "learning_rate": 6.999315139426588e-06,
+      "loss": 0.3774,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3881527669524552,
+      "grad_norm": 0.6691013538576442,
+      "learning_rate": 6.998097210196643e-06,
+      "loss": 0.3369,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38823481150264594,
+      "grad_norm": 0.5851513393778214,
+      "learning_rate": 6.996879139864283e-06,
+      "loss": 0.3175,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38831685605283667,
+      "grad_norm": 0.6482868573356558,
+      "learning_rate": 6.995660928515527e-06,
+      "loss": 0.391,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38839890060302745,
+      "grad_norm": 0.6588137734290144,
+      "learning_rate": 6.994442576236401e-06,
+      "loss": 0.3695,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3884809451532182,
+      "grad_norm": 0.6316762165263078,
+      "learning_rate": 6.993224083112946e-06,
+      "loss": 0.3642,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38856298970340897,
+      "grad_norm": 0.6999030550700606,
+      "learning_rate": 6.9920054492312086e-06,
+      "loss": 0.3422,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3886450342535997,
+      "grad_norm": 0.6425112603935651,
+      "learning_rate": 6.990786674677246e-06,
+      "loss": 0.3355,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3887270788037905,
+      "grad_norm": 0.658297009888357,
+      "learning_rate": 6.989567759537128e-06,
+      "loss": 0.3662,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3888091233539812,
+      "grad_norm": 0.6259424830299274,
+      "learning_rate": 6.98834870389693e-06,
+      "loss": 0.3432,
+      "step": 4739
+    },
+    {
+      "epoch": 0.38889116790417194,
+      "grad_norm": 0.6633535411600685,
+      "learning_rate": 6.987129507842742e-06,
+      "loss": 0.3664,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3889732124543627,
+      "grad_norm": 0.6214158514972004,
+      "learning_rate": 6.985910171460659e-06,
+      "loss": 0.3364,
+      "step": 4741
+    },
+    {
+      "epoch": 0.38905525700455346,
+      "grad_norm": 0.6755323131997197,
+      "learning_rate": 6.984690694836792e-06,
+      "loss": 0.3664,
+      "step": 4742
+    },
+    {
+      "epoch": 0.38913730155474424,
+      "grad_norm": 0.6475969520223085,
+      "learning_rate": 6.9834710780572565e-06,
+      "loss": 0.358,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38921934610493497,
+      "grad_norm": 0.6320926644903382,
+      "learning_rate": 6.98225132120818e-06,
+      "loss": 0.3778,
+      "step": 4744
+    },
+    {
+      "epoch": 0.38930139065512576,
+      "grad_norm": 0.6399108928703082,
+      "learning_rate": 6.981031424375701e-06,
+      "loss": 0.3304,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3893834352053165,
+      "grad_norm": 0.6263686519654121,
+      "learning_rate": 6.979811387645964e-06,
+      "loss": 0.335,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3894654797555072,
+      "grad_norm": 0.6164670206560524,
+      "learning_rate": 6.978591211105128e-06,
+      "loss": 0.3162,
+      "step": 4747
+    },
+    {
+      "epoch": 0.389547524305698,
+      "grad_norm": 0.6424928012027573,
+      "learning_rate": 6.9773708948393605e-06,
+      "loss": 0.3472,
+      "step": 4748
+    },
+    {
+      "epoch": 0.38962956885588873,
+      "grad_norm": 0.619558839836847,
+      "learning_rate": 6.976150438934837e-06,
+      "loss": 0.3429,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3897116134060795,
+      "grad_norm": 0.6625317416390896,
+      "learning_rate": 6.9749298434777425e-06,
+      "loss": 0.3336,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38979365795627025,
+      "grad_norm": 0.6275629529314904,
+      "learning_rate": 6.973709108554277e-06,
+      "loss": 0.3472,
+      "step": 4751
+    },
+    {
+      "epoch": 0.38987570250646103,
+      "grad_norm": 0.6502481414453936,
+      "learning_rate": 6.972488234250646e-06,
+      "loss": 0.3204,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38995774705665176,
+      "grad_norm": 0.6165199309615962,
+      "learning_rate": 6.971267220653064e-06,
+      "loss": 0.3423,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3900397916068425,
+      "grad_norm": 0.599839199175915,
+      "learning_rate": 6.970046067847758e-06,
+      "loss": 0.3577,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3901218361570333,
+      "grad_norm": 0.6247209257298942,
+      "learning_rate": 6.9688247759209635e-06,
+      "loss": 0.3403,
+      "step": 4755
+    },
+    {
+      "epoch": 0.390203880707224,
+      "grad_norm": 0.6106956200584188,
+      "learning_rate": 6.967603344958926e-06,
+      "loss": 0.3382,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3902859252574148,
+      "grad_norm": 0.5773732816578326,
+      "learning_rate": 6.9663817750479035e-06,
+      "loss": 0.3525,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3903679698076055,
+      "grad_norm": 0.6149089354400228,
+      "learning_rate": 6.965160066274157e-06,
+      "loss": 0.3369,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3904500143577963,
+      "grad_norm": 0.5844371998220882,
+      "learning_rate": 6.963938218723966e-06,
+      "loss": 0.3427,
+      "step": 4759
+    },
+    {
+      "epoch": 0.39053205890798703,
+      "grad_norm": 0.6510851788114717,
+      "learning_rate": 6.962716232483612e-06,
+      "loss": 0.3502,
+      "step": 4760
+    },
+    {
+      "epoch": 0.39061410345817776,
+      "grad_norm": 0.7047810583237964,
+      "learning_rate": 6.961494107639391e-06,
+      "loss": 0.3629,
+      "step": 4761
+    },
+    {
+      "epoch": 0.39069614800836855,
+      "grad_norm": 0.6973023946479312,
+      "learning_rate": 6.960271844277608e-06,
+      "loss": 0.3544,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3907781925585593,
+      "grad_norm": 0.6829735770051724,
+      "learning_rate": 6.959049442484577e-06,
+      "loss": 0.3348,
+      "step": 4763
+    },
+    {
+      "epoch": 0.39086023710875006,
+      "grad_norm": 0.6371529124581703,
+      "learning_rate": 6.95782690234662e-06,
+      "loss": 0.3728,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3909422816589408,
+      "grad_norm": 0.7052422258959135,
+      "learning_rate": 6.956604223950074e-06,
+      "loss": 0.3873,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3910243262091316,
+      "grad_norm": 0.6304363967385775,
+      "learning_rate": 6.955381407381283e-06,
+      "loss": 0.3528,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3911063707593223,
+      "grad_norm": 0.6293273669895987,
+      "learning_rate": 6.9541584527265975e-06,
+      "loss": 0.3481,
+      "step": 4767
+    },
+    {
+      "epoch": 0.39118841530951304,
+      "grad_norm": 0.6685522247591779,
+      "learning_rate": 6.952935360072379e-06,
+      "loss": 0.3682,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3912704598597038,
+      "grad_norm": 0.602957969040934,
+      "learning_rate": 6.9517121295050046e-06,
+      "loss": 0.3385,
+      "step": 4769
+    },
+    {
+      "epoch": 0.39135250440989455,
+      "grad_norm": 0.7145231343047503,
+      "learning_rate": 6.950488761110856e-06,
+      "loss": 0.3659,
+      "step": 4770
+    },
+    {
+      "epoch": 0.39143454896008534,
+      "grad_norm": 0.6596480377413894,
+      "learning_rate": 6.949265254976325e-06,
+      "loss": 0.3547,
+      "step": 4771
+    },
+    {
+      "epoch": 0.39151659351027607,
+      "grad_norm": 0.643969330439757,
+      "learning_rate": 6.948041611187812e-06,
+      "loss": 0.3227,
+      "step": 4772
+    },
+    {
+      "epoch": 0.39159863806046685,
+      "grad_norm": 0.6557457633520409,
+      "learning_rate": 6.9468178298317325e-06,
+      "loss": 0.3337,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3916806826106576,
+      "grad_norm": 0.6349849378808968,
+      "learning_rate": 6.945593910994503e-06,
+      "loss": 0.3514,
+      "step": 4774
+    },
+    {
+      "epoch": 0.39176272716084837,
+      "grad_norm": 0.7014558201236325,
+      "learning_rate": 6.944369854762559e-06,
+      "loss": 0.3942,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3918447717110391,
+      "grad_norm": 0.6589549774992697,
+      "learning_rate": 6.94314566122234e-06,
+      "loss": 0.3435,
+      "step": 4776
+    },
+    {
+      "epoch": 0.3919268162612298,
+      "grad_norm": 0.6048995604885627,
+      "learning_rate": 6.941921330460295e-06,
+      "loss": 0.3748,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3920088608114206,
+      "grad_norm": 0.672930565852243,
+      "learning_rate": 6.940696862562886e-06,
+      "loss": 0.3267,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39209090536161134,
+      "grad_norm": 0.6350505959890812,
+      "learning_rate": 6.939472257616583e-06,
+      "loss": 0.3312,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3921729499118021,
+      "grad_norm": 0.6809206674827769,
+      "learning_rate": 6.938247515707866e-06,
+      "loss": 0.3337,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39225499446199286,
+      "grad_norm": 0.6100557447931385,
+      "learning_rate": 6.937022636923222e-06,
+      "loss": 0.2992,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39233703901218364,
+      "grad_norm": 0.6306807270934633,
+      "learning_rate": 6.935797621349151e-06,
+      "loss": 0.3131,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39241908356237437,
+      "grad_norm": 0.6546533574034628,
+      "learning_rate": 6.934572469072163e-06,
+      "loss": 0.3275,
+      "step": 4783
+    },
+    {
+      "epoch": 0.3925011281125651,
+      "grad_norm": 0.6672860831985065,
+      "learning_rate": 6.933347180178775e-06,
+      "loss": 0.3665,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3925831726627559,
+      "grad_norm": 0.6451134182690936,
+      "learning_rate": 6.932121754755516e-06,
+      "loss": 0.3563,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3926652172129466,
+      "grad_norm": 0.6725702723296628,
+      "learning_rate": 6.930896192888923e-06,
+      "loss": 0.3449,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3927472617631374,
+      "grad_norm": 0.706390511414826,
+      "learning_rate": 6.929670494665542e-06,
+      "loss": 0.4329,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39282930631332813,
+      "grad_norm": 0.6637343015918635,
+      "learning_rate": 6.928444660171931e-06,
+      "loss": 0.3809,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3929113508635189,
+      "grad_norm": 1.1373887246239056,
+      "learning_rate": 6.927218689494657e-06,
+      "loss": 0.3605,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39299339541370965,
+      "grad_norm": 0.6113426572839179,
+      "learning_rate": 6.9259925827202955e-06,
+      "loss": 0.3306,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3930754399639004,
+      "grad_norm": 0.5987631070187168,
+      "learning_rate": 6.924766339935431e-06,
+      "loss": 0.3479,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39315748451409116,
+      "grad_norm": 0.6379477389840761,
+      "learning_rate": 6.92353996122666e-06,
+      "loss": 0.3136,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3932395290642819,
+      "grad_norm": 0.6161290627595079,
+      "learning_rate": 6.92231344668059e-06,
+      "loss": 0.3278,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3933215736144727,
+      "grad_norm": 0.6313101456642353,
+      "learning_rate": 6.921086796383831e-06,
+      "loss": 0.3379,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3934036181646634,
+      "grad_norm": 0.6288439578552399,
+      "learning_rate": 6.919860010423008e-06,
+      "loss": 0.3331,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3934856627148542,
+      "grad_norm": 0.6746071299893998,
+      "learning_rate": 6.918633088884756e-06,
+      "loss": 0.3397,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3935677072650449,
+      "grad_norm": 0.6575442256598377,
+      "learning_rate": 6.917406031855719e-06,
+      "loss": 0.3237,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39364975181523565,
+      "grad_norm": 0.6483836202881701,
+      "learning_rate": 6.916178839422548e-06,
+      "loss": 0.3471,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39373179636542643,
+      "grad_norm": 0.622209958183057,
+      "learning_rate": 6.914951511671905e-06,
+      "loss": 0.312,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39381384091561716,
+      "grad_norm": 0.5983020345672395,
+      "learning_rate": 6.913724048690464e-06,
+      "loss": 0.3532,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39389588546580795,
+      "grad_norm": 0.6672562687153197,
+      "learning_rate": 6.9124964505649065e-06,
+      "loss": 0.3299,
+      "step": 4801
+    },
+    {
+      "epoch": 0.3939779300159987,
+      "grad_norm": 0.6716284113644152,
+      "learning_rate": 6.9112687173819205e-06,
+      "loss": 0.3389,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39405997456618946,
+      "grad_norm": 0.6253787213730002,
+      "learning_rate": 6.910040849228208e-06,
+      "loss": 0.3645,
+      "step": 4803
+    },
+    {
+      "epoch": 0.3941420191163802,
+      "grad_norm": 0.64971855966765,
+      "learning_rate": 6.908812846190479e-06,
+      "loss": 0.3676,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3942240636665709,
+      "grad_norm": 0.6239684570595055,
+      "learning_rate": 6.9075847083554546e-06,
+      "loss": 0.3762,
+      "step": 4805
+    },
+    {
+      "epoch": 0.3943061082167617,
+      "grad_norm": 0.6583754040666003,
+      "learning_rate": 6.9063564358098636e-06,
+      "loss": 0.3199,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39438815276695244,
+      "grad_norm": 0.5919608032701172,
+      "learning_rate": 6.905128028640443e-06,
+      "loss": 0.309,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3944701973171432,
+      "grad_norm": 0.6201950000374403,
+      "learning_rate": 6.903899486933943e-06,
+      "loss": 0.3529,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39455224186733395,
+      "grad_norm": 0.6081802865466093,
+      "learning_rate": 6.902670810777119e-06,
+      "loss": 0.3537,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39463428641752474,
+      "grad_norm": 0.6771028884233132,
+      "learning_rate": 6.9014420002567395e-06,
+      "loss": 0.4054,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39471633096771547,
+      "grad_norm": 0.6657438417185029,
+      "learning_rate": 6.900213055459581e-06,
+      "loss": 0.3766,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3947983755179062,
+      "grad_norm": 0.6046718300919297,
+      "learning_rate": 6.898983976472429e-06,
+      "loss": 0.3366,
+      "step": 4812
+    },
+    {
+      "epoch": 0.394880420068097,
+      "grad_norm": 0.6376820944795345,
+      "learning_rate": 6.8977547633820785e-06,
+      "loss": 0.3521,
+      "step": 4813
+    },
+    {
+      "epoch": 0.3949624646182877,
+      "grad_norm": 0.6022045292853576,
+      "learning_rate": 6.896525416275337e-06,
+      "loss": 0.2957,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3950445091684785,
+      "grad_norm": 0.6370485730439917,
+      "learning_rate": 6.895295935239017e-06,
+      "loss": 0.3373,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3951265537186692,
+      "grad_norm": 0.6220236795135651,
+      "learning_rate": 6.8940663203599445e-06,
+      "loss": 0.3514,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39520859826886,
+      "grad_norm": 0.6437827855758046,
+      "learning_rate": 6.89283657172495e-06,
+      "loss": 0.3398,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39529064281905074,
+      "grad_norm": 0.6493984153714338,
+      "learning_rate": 6.891606689420877e-06,
+      "loss": 0.354,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39537268736924147,
+      "grad_norm": 0.6583168527825255,
+      "learning_rate": 6.89037667353458e-06,
+      "loss": 0.3824,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39545473191943226,
+      "grad_norm": 0.6669034828484723,
+      "learning_rate": 6.8891465241529195e-06,
+      "loss": 0.396,
+      "step": 4820
+    },
+    {
+      "epoch": 0.395536776469623,
+      "grad_norm": 0.6516830117199849,
+      "learning_rate": 6.887916241362766e-06,
+      "loss": 0.3405,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39561882101981377,
+      "grad_norm": 0.714631361800471,
+      "learning_rate": 6.886685825251002e-06,
+      "loss": 0.368,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3957008655700045,
+      "grad_norm": 0.6644434239798798,
+      "learning_rate": 6.885455275904516e-06,
+      "loss": 0.3683,
+      "step": 4823
+    },
+    {
+      "epoch": 0.3957829101201953,
+      "grad_norm": 0.6514158241121741,
+      "learning_rate": 6.884224593410207e-06,
+      "loss": 0.3574,
+      "step": 4824
+    },
+    {
+      "epoch": 0.395864954670386,
+      "grad_norm": 0.6222904696722746,
+      "learning_rate": 6.882993777854985e-06,
+      "loss": 0.3546,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3959469992205768,
+      "grad_norm": 0.6122790374083659,
+      "learning_rate": 6.881762829325767e-06,
+      "loss": 0.3269,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39602904377076753,
+      "grad_norm": 0.6073385437775973,
+      "learning_rate": 6.88053174790948e-06,
+      "loss": 0.3217,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39611108832095826,
+      "grad_norm": 0.6754852127835899,
+      "learning_rate": 6.879300533693065e-06,
+      "loss": 0.3361,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39619313287114905,
+      "grad_norm": 0.6467397096926741,
+      "learning_rate": 6.878069186763466e-06,
+      "loss": 0.3297,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3962751774213398,
+      "grad_norm": 0.64847346593117,
+      "learning_rate": 6.876837707207635e-06,
+      "loss": 0.3191,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39635722197153056,
+      "grad_norm": 0.6649960380245264,
+      "learning_rate": 6.8756060951125415e-06,
+      "loss": 0.3554,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3964392665217213,
+      "grad_norm": 0.6499942308292885,
+      "learning_rate": 6.874374350565161e-06,
+      "loss": 0.3956,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3965213110719121,
+      "grad_norm": 0.6958343652897404,
+      "learning_rate": 6.873142473652473e-06,
+      "loss": 0.379,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3966033556221028,
+      "grad_norm": 0.6619277088540484,
+      "learning_rate": 6.871910464461475e-06,
+      "loss": 0.347,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39668540017229353,
+      "grad_norm": 0.6680023795387767,
+      "learning_rate": 6.870678323079165e-06,
+      "loss": 0.3473,
+      "step": 4835
+    },
+    {
+      "epoch": 0.3967674447224843,
+      "grad_norm": 0.7416131761287764,
+      "learning_rate": 6.86944604959256e-06,
+      "loss": 0.3321,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39684948927267505,
+      "grad_norm": 0.6455632793979039,
+      "learning_rate": 6.868213644088677e-06,
+      "loss": 0.3369,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39693153382286583,
+      "grad_norm": 0.6222302120172022,
+      "learning_rate": 6.866981106654548e-06,
+      "loss": 0.3319,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39701357837305656,
+      "grad_norm": 0.5885309838529184,
+      "learning_rate": 6.865748437377213e-06,
+      "loss": 0.3403,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39709562292324735,
+      "grad_norm": 0.6347703526263198,
+      "learning_rate": 6.86451563634372e-06,
+      "loss": 0.3566,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3971776674734381,
+      "grad_norm": 0.6281177777811634,
+      "learning_rate": 6.863282703641129e-06,
+      "loss": 0.3469,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3972597120236288,
+      "grad_norm": 0.5944925392232031,
+      "learning_rate": 6.862049639356508e-06,
+      "loss": 0.3452,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3973417565738196,
+      "grad_norm": 0.6633579281609273,
+      "learning_rate": 6.860816443576932e-06,
+      "loss": 0.3568,
+      "step": 4843
+    },
+    {
+      "epoch": 0.3974238011240103,
+      "grad_norm": 0.6644578254553546,
+      "learning_rate": 6.859583116389489e-06,
+      "loss": 0.3628,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3975058456742011,
+      "grad_norm": 0.6139552453709749,
+      "learning_rate": 6.858349657881273e-06,
+      "loss": 0.3526,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39758789022439184,
+      "grad_norm": 0.6349654333355003,
+      "learning_rate": 6.85711606813939e-06,
+      "loss": 0.3519,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3976699347745826,
+      "grad_norm": 0.6324761051290656,
+      "learning_rate": 6.855882347250955e-06,
+      "loss": 0.3697,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39775197932477335,
+      "grad_norm": 0.6588464490470071,
+      "learning_rate": 6.8546484953030885e-06,
+      "loss": 0.3567,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3978340238749641,
+      "grad_norm": 0.6405467792662278,
+      "learning_rate": 6.853414512382928e-06,
+      "loss": 0.3395,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39791606842515487,
+      "grad_norm": 0.7078641007661302,
+      "learning_rate": 6.85218039857761e-06,
+      "loss": 0.3396,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3979981129753456,
+      "grad_norm": 0.6946260115631758,
+      "learning_rate": 6.8509461539742894e-06,
+      "loss": 0.4045,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3980801575255364,
+      "grad_norm": 0.6280702622477675,
+      "learning_rate": 6.849711778660124e-06,
+      "loss": 0.2994,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3981622020757271,
+      "grad_norm": 0.6111243966309688,
+      "learning_rate": 6.848477272722284e-06,
+      "loss": 0.2859,
+      "step": 4853
+    },
+    {
+      "epoch": 0.3982442466259179,
+      "grad_norm": 0.619966357181228,
+      "learning_rate": 6.84724263624795e-06,
+      "loss": 0.3441,
+      "step": 4854
+    },
+    {
+      "epoch": 0.3983262911761086,
+      "grad_norm": 0.6343226024788606,
+      "learning_rate": 6.846007869324308e-06,
+      "loss": 0.3093,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39840833572629936,
+      "grad_norm": 0.6261851108471559,
+      "learning_rate": 6.844772972038558e-06,
+      "loss": 0.3639,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39849038027649014,
+      "grad_norm": 0.6844061494218175,
+      "learning_rate": 6.843537944477904e-06,
+      "loss": 0.3334,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39857242482668087,
+      "grad_norm": 0.6459717334733264,
+      "learning_rate": 6.84230278672956e-06,
+      "loss": 0.3591,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39865446937687166,
+      "grad_norm": 0.7299925733092899,
+      "learning_rate": 6.841067498880754e-06,
+      "loss": 0.4015,
+      "step": 4859
+    },
+    {
+      "epoch": 0.3987365139270624,
+      "grad_norm": 0.7161113260341332,
+      "learning_rate": 6.83983208101872e-06,
+      "loss": 0.3497,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39881855847725317,
+      "grad_norm": 0.6425760829149234,
+      "learning_rate": 6.838596533230701e-06,
+      "loss": 0.3539,
+      "step": 4861
+    },
+    {
+      "epoch": 0.3989006030274439,
+      "grad_norm": 0.6301688325566874,
+      "learning_rate": 6.837360855603947e-06,
+      "loss": 0.3403,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39898264757763463,
+      "grad_norm": 0.6805520913414699,
+      "learning_rate": 6.836125048225723e-06,
+      "loss": 0.3553,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3990646921278254,
+      "grad_norm": 0.7063855063989007,
+      "learning_rate": 6.8348891111832975e-06,
+      "loss": 0.3283,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39914673667801615,
+      "grad_norm": 0.6702787159612048,
+      "learning_rate": 6.83365304456395e-06,
+      "loss": 0.3691,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39922878122820693,
+      "grad_norm": 0.6634140536573462,
+      "learning_rate": 6.832416848454971e-06,
+      "loss": 0.3403,
+      "step": 4866
+    },
+    {
+      "epoch": 0.39931082577839766,
+      "grad_norm": 0.6489512292160619,
+      "learning_rate": 6.831180522943658e-06,
+      "loss": 0.38,
+      "step": 4867
+    },
+    {
+      "epoch": 0.39939287032858845,
+      "grad_norm": 0.6282844421674947,
+      "learning_rate": 6.829944068117319e-06,
+      "loss": 0.3634,
+      "step": 4868
+    },
+    {
+      "epoch": 0.3994749148787792,
+      "grad_norm": 0.6422245918564127,
+      "learning_rate": 6.828707484063269e-06,
+      "loss": 0.3291,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3995569594289699,
+      "grad_norm": 0.5705389063779657,
+      "learning_rate": 6.827470770868837e-06,
+      "loss": 0.3092,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3996390039791607,
+      "grad_norm": 0.6389025387197856,
+      "learning_rate": 6.826233928621353e-06,
+      "loss": 0.3865,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3997210485293514,
+      "grad_norm": 0.5898315226321365,
+      "learning_rate": 6.824996957408163e-06,
+      "loss": 0.3782,
+      "step": 4872
+    },
+    {
+      "epoch": 0.3998030930795422,
+      "grad_norm": 0.6102978692535678,
+      "learning_rate": 6.823759857316621e-06,
+      "loss": 0.3421,
+      "step": 4873
+    },
+    {
+      "epoch": 0.39988513762973293,
+      "grad_norm": 0.6207864855069712,
+      "learning_rate": 6.822522628434087e-06,
+      "loss": 0.3993,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3999671821799237,
+      "grad_norm": 0.7155506330570192,
+      "learning_rate": 6.821285270847934e-06,
+      "loss": 0.335,
+      "step": 4875
+    },
+    {
+      "epoch": 0.40004922673011445,
+      "grad_norm": 0.6233017277162617,
+      "learning_rate": 6.82004778464554e-06,
+      "loss": 0.3161,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4001312712803052,
+      "grad_norm": 0.6416443613860174,
+      "learning_rate": 6.818810169914297e-06,
+      "loss": 0.371,
+      "step": 4877
+    },
+    {
+      "epoch": 0.40021331583049596,
+      "grad_norm": 0.6492735355893094,
+      "learning_rate": 6.817572426741601e-06,
+      "loss": 0.3355,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4002953603806867,
+      "grad_norm": 0.664958748799679,
+      "learning_rate": 6.8163345552148604e-06,
+      "loss": 0.3633,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4003774049308775,
+      "grad_norm": 0.6508217351597321,
+      "learning_rate": 6.815096555421492e-06,
+      "loss": 0.3262,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4004594494810682,
+      "grad_norm": 0.670913281212954,
+      "learning_rate": 6.813858427448921e-06,
+      "loss": 0.3608,
+      "step": 4881
+    },
+    {
+      "epoch": 0.400541494031259,
+      "grad_norm": 0.7059654924394448,
+      "learning_rate": 6.812620171384582e-06,
+      "loss": 0.3262,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4006235385814497,
+      "grad_norm": 0.6663320920632044,
+      "learning_rate": 6.811381787315919e-06,
+      "loss": 0.3661,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4007055831316405,
+      "grad_norm": 0.6265749732680048,
+      "learning_rate": 6.810143275330385e-06,
+      "loss": 0.3308,
+      "step": 4884
+    },
+    {
+      "epoch": 0.40078762768183124,
+      "grad_norm": 0.5916058372343015,
+      "learning_rate": 6.80890463551544e-06,
+      "loss": 0.3341,
+      "step": 4885
+    },
+    {
+      "epoch": 0.40086967223202197,
+      "grad_norm": 0.6847261485240617,
+      "learning_rate": 6.807665867958555e-06,
+      "loss": 0.3298,
+      "step": 4886
+    },
+    {
+      "epoch": 0.40095171678221275,
+      "grad_norm": 0.6553643312483408,
+      "learning_rate": 6.806426972747212e-06,
+      "loss": 0.3624,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4010337613324035,
+      "grad_norm": 0.6949898098669737,
+      "learning_rate": 6.805187949968899e-06,
+      "loss": 0.3608,
+      "step": 4888
+    },
+    {
+      "epoch": 0.40111580588259427,
+      "grad_norm": 0.62603060969543,
+      "learning_rate": 6.803948799711111e-06,
+      "loss": 0.3373,
+      "step": 4889
+    },
+    {
+      "epoch": 0.401197850432785,
+      "grad_norm": 0.6373425409426963,
+      "learning_rate": 6.80270952206136e-06,
+      "loss": 0.3608,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4012798949829758,
+      "grad_norm": 0.6592823659941973,
+      "learning_rate": 6.801470117107157e-06,
+      "loss": 0.3556,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4013619395331665,
+      "grad_norm": 0.6622851611266923,
+      "learning_rate": 6.8002305849360295e-06,
+      "loss": 0.3837,
+      "step": 4892
+    },
+    {
+      "epoch": 0.40144398408335724,
+      "grad_norm": 0.6237076160689363,
+      "learning_rate": 6.798990925635509e-06,
+      "loss": 0.3093,
+      "step": 4893
+    },
+    {
+      "epoch": 0.401526028633548,
+      "grad_norm": 0.649112335353178,
+      "learning_rate": 6.79775113929314e-06,
+      "loss": 0.3547,
+      "step": 4894
+    },
+    {
+      "epoch": 0.40160807318373876,
+      "grad_norm": 0.6644178723690366,
+      "learning_rate": 6.796511225996474e-06,
+      "loss": 0.3594,
+      "step": 4895
+    },
+    {
+      "epoch": 0.40169011773392954,
+      "grad_norm": 0.6294704190966411,
+      "learning_rate": 6.795271185833072e-06,
+      "loss": 0.3474,
+      "step": 4896
+    },
+    {
+      "epoch": 0.40177216228412027,
+      "grad_norm": 0.6685170968277875,
+      "learning_rate": 6.794031018890504e-06,
+      "loss": 0.3695,
+      "step": 4897
+    },
+    {
+      "epoch": 0.40185420683431106,
+      "grad_norm": 0.7006229860418428,
+      "learning_rate": 6.792790725256347e-06,
+      "loss": 0.4255,
+      "step": 4898
+    },
+    {
+      "epoch": 0.4019362513845018,
+      "grad_norm": 0.6581623093745641,
+      "learning_rate": 6.791550305018188e-06,
+      "loss": 0.3622,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4020182959346925,
+      "grad_norm": 0.6084159103938195,
+      "learning_rate": 6.7903097582636255e-06,
+      "loss": 0.3351,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4021003404848833,
+      "grad_norm": 0.6190081771103121,
+      "learning_rate": 6.789069085080265e-06,
+      "loss": 0.3318,
+      "step": 4901
+    },
+    {
+      "epoch": 0.40218238503507403,
+      "grad_norm": 0.6191747396434278,
+      "learning_rate": 6.78782828555572e-06,
+      "loss": 0.3512,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4022644295852648,
+      "grad_norm": 0.6348349053785265,
+      "learning_rate": 6.786587359777612e-06,
+      "loss": 0.3659,
+      "step": 4903
+    },
+    {
+      "epoch": 0.40234647413545555,
+      "grad_norm": 0.6677494050538845,
+      "learning_rate": 6.785346307833578e-06,
+      "loss": 0.3836,
+      "step": 4904
+    },
+    {
+      "epoch": 0.40242851868564633,
+      "grad_norm": 0.6264477351309535,
+      "learning_rate": 6.7841051298112545e-06,
+      "loss": 0.3825,
+      "step": 4905
+    },
+    {
+      "epoch": 0.40251056323583706,
+      "grad_norm": 0.6635287889364051,
+      "learning_rate": 6.782863825798294e-06,
+      "loss": 0.366,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4025926077860278,
+      "grad_norm": 0.7060649071338949,
+      "learning_rate": 6.781622395882352e-06,
+      "loss": 0.3138,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4026746523362186,
+      "grad_norm": 0.622463429764897,
+      "learning_rate": 6.780380840151101e-06,
+      "loss": 0.3406,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4027566968864093,
+      "grad_norm": 0.6267225274029479,
+      "learning_rate": 6.779139158692215e-06,
+      "loss": 0.3277,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4028387414366001,
+      "grad_norm": 0.6497938163112323,
+      "learning_rate": 6.77789735159338e-06,
+      "loss": 0.3376,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4029207859867908,
+      "grad_norm": 0.6626605057549633,
+      "learning_rate": 6.776655418942291e-06,
+      "loss": 0.3137,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4030028305369816,
+      "grad_norm": 0.6520541863309564,
+      "learning_rate": 6.7754133608266505e-06,
+      "loss": 0.3762,
+      "step": 4912
+    },
+    {
+      "epoch": 0.40308487508717233,
+      "grad_norm": 0.6073489758230328,
+      "learning_rate": 6.77417117733417e-06,
+      "loss": 0.3518,
+      "step": 4913
+    },
+    {
+      "epoch": 0.40316691963736306,
+      "grad_norm": 0.6675746167649793,
+      "learning_rate": 6.772928868552572e-06,
+      "loss": 0.3527,
+      "step": 4914
+    },
+    {
+      "epoch": 0.40324896418755385,
+      "grad_norm": 0.6793227520611388,
+      "learning_rate": 6.771686434569587e-06,
+      "loss": 0.3784,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4033310087377446,
+      "grad_norm": 0.620309365696505,
+      "learning_rate": 6.77044387547295e-06,
+      "loss": 0.3506,
+      "step": 4916
+    },
+    {
+      "epoch": 0.40341305328793536,
+      "grad_norm": 0.5700276757841729,
+      "learning_rate": 6.769201191350412e-06,
+      "loss": 0.3133,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4034950978381261,
+      "grad_norm": 0.6252086081463765,
+      "learning_rate": 6.767958382289729e-06,
+      "loss": 0.3494,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4035771423883169,
+      "grad_norm": 0.627933679215801,
+      "learning_rate": 6.7667154483786645e-06,
+      "loss": 0.3346,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4036591869385076,
+      "grad_norm": 0.6810044449668229,
+      "learning_rate": 6.765472389704994e-06,
+      "loss": 0.3481,
+      "step": 4920
+    },
+    {
+      "epoch": 0.40374123148869834,
+      "grad_norm": 0.6484788462990652,
+      "learning_rate": 6.764229206356498e-06,
+      "loss": 0.3563,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4038232760388891,
+      "grad_norm": 0.6539835020266082,
+      "learning_rate": 6.762985898420973e-06,
+      "loss": 0.3809,
+      "step": 4922
+    },
+    {
+      "epoch": 0.40390532058907985,
+      "grad_norm": 0.6122648808648451,
+      "learning_rate": 6.761742465986214e-06,
+      "loss": 0.3464,
+      "step": 4923
+    },
+    {
+      "epoch": 0.40398736513927064,
+      "grad_norm": 0.6382028844695699,
+      "learning_rate": 6.760498909140034e-06,
+      "loss": 0.352,
+      "step": 4924
+    },
+    {
+      "epoch": 0.40406940968946137,
+      "grad_norm": 0.6599365357192872,
+      "learning_rate": 6.759255227970248e-06,
+      "loss": 0.3321,
+      "step": 4925
+    },
+    {
+      "epoch": 0.40415145423965215,
+      "grad_norm": 0.6080118933263182,
+      "learning_rate": 6.758011422564685e-06,
+      "loss": 0.3214,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4042334987898429,
+      "grad_norm": 0.597284155653588,
+      "learning_rate": 6.756767493011181e-06,
+      "loss": 0.3121,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4043155433400336,
+      "grad_norm": 0.6791713275425657,
+      "learning_rate": 6.755523439397578e-06,
+      "loss": 0.3643,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4043975878902244,
+      "grad_norm": 0.5831110253192245,
+      "learning_rate": 6.75427926181173e-06,
+      "loss": 0.3281,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4044796324404151,
+      "grad_norm": 0.6978067900902315,
+      "learning_rate": 6.753034960341498e-06,
+      "loss": 0.3425,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4045616769906059,
+      "grad_norm": 0.6576029636761913,
+      "learning_rate": 6.7517905350747544e-06,
+      "loss": 0.352,
+      "step": 4931
+    },
+    {
+      "epoch": 0.40464372154079664,
+      "grad_norm": 0.6796491203545519,
+      "learning_rate": 6.7505459860993796e-06,
+      "loss": 0.3601,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4047257660909874,
+      "grad_norm": 0.6274827500540757,
+      "learning_rate": 6.749301313503258e-06,
+      "loss": 0.3403,
+      "step": 4933
+    },
+    {
+      "epoch": 0.40480781064117816,
+      "grad_norm": 0.7194811571477671,
+      "learning_rate": 6.748056517374288e-06,
+      "loss": 0.3555,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4048898551913689,
+      "grad_norm": 0.6457035504708221,
+      "learning_rate": 6.746811597800376e-06,
+      "loss": 0.3415,
+      "step": 4935
+    },
+    {
+      "epoch": 0.40497189974155967,
+      "grad_norm": 0.6430443794883016,
+      "learning_rate": 6.7455665548694334e-06,
+      "loss": 0.3328,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4050539442917504,
+      "grad_norm": 0.6355811629364211,
+      "learning_rate": 6.744321388669388e-06,
+      "loss": 0.3417,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4051359888419412,
+      "grad_norm": 0.6714736337505299,
+      "learning_rate": 6.743076099288167e-06,
+      "loss": 0.3642,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4052180333921319,
+      "grad_norm": 0.6015327803643041,
+      "learning_rate": 6.7418306868137125e-06,
+      "loss": 0.3517,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4053000779423227,
+      "grad_norm": 0.6167796926026554,
+      "learning_rate": 6.7405851513339736e-06,
+      "loss": 0.3309,
+      "step": 4940
+    },
+    {
+      "epoch": 0.40538212249251343,
+      "grad_norm": 0.6452911913742196,
+      "learning_rate": 6.739339492936907e-06,
+      "loss": 0.3334,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4054641670427042,
+      "grad_norm": 0.6667441642782671,
+      "learning_rate": 6.7380937117104816e-06,
+      "loss": 0.3242,
+      "step": 4942
+    },
+    {
+      "epoch": 0.40554621159289495,
+      "grad_norm": 0.6251854783080425,
+      "learning_rate": 6.736847807742668e-06,
+      "loss": 0.3536,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4056282561430857,
+      "grad_norm": 0.6386899039256704,
+      "learning_rate": 6.735601781121454e-06,
+      "loss": 0.3572,
+      "step": 4944
+    },
+    {
+      "epoch": 0.40571030069327646,
+      "grad_norm": 0.6249290442377724,
+      "learning_rate": 6.734355631934832e-06,
+      "loss": 0.3194,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4057923452434672,
+      "grad_norm": 0.6591062480166139,
+      "learning_rate": 6.733109360270802e-06,
+      "loss": 0.3299,
+      "step": 4946
+    },
+    {
+      "epoch": 0.405874389793658,
+      "grad_norm": 0.6193065056615554,
+      "learning_rate": 6.731862966217372e-06,
+      "loss": 0.3294,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4059564343438487,
+      "grad_norm": 0.6898001411562877,
+      "learning_rate": 6.730616449862562e-06,
+      "loss": 0.3869,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4060384788940395,
+      "grad_norm": 0.6936790894746817,
+      "learning_rate": 6.7293698112943975e-06,
+      "loss": 0.3253,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4061205234442302,
+      "grad_norm": 0.6120317013108436,
+      "learning_rate": 6.728123050600917e-06,
+      "loss": 0.2939,
+      "step": 4950
+    },
+    {
+      "epoch": 0.40620256799442095,
+      "grad_norm": 0.6872433011366046,
+      "learning_rate": 6.726876167870162e-06,
+      "loss": 0.3795,
+      "step": 4951
+    },
+    {
+      "epoch": 0.40628461254461173,
+      "grad_norm": 0.6585342409400938,
+      "learning_rate": 6.725629163190188e-06,
+      "loss": 0.3783,
+      "step": 4952
+    },
+    {
+      "epoch": 0.40636665709480246,
+      "grad_norm": 0.6426335152317297,
+      "learning_rate": 6.724382036649054e-06,
+      "loss": 0.338,
+      "step": 4953
+    },
+    {
+      "epoch": 0.40644870164499325,
+      "grad_norm": 0.6992265907225961,
+      "learning_rate": 6.723134788334831e-06,
+      "loss": 0.3709,
+      "step": 4954
+    },
+    {
+      "epoch": 0.406530746195184,
+      "grad_norm": 0.6826438677400899,
+      "learning_rate": 6.721887418335598e-06,
+      "loss": 0.307,
+      "step": 4955
+    },
+    {
+      "epoch": 0.40661279074537476,
+      "grad_norm": 0.6567698816605938,
+      "learning_rate": 6.720639926739442e-06,
+      "loss": 0.3862,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4066948352955655,
+      "grad_norm": 0.5949663256800234,
+      "learning_rate": 6.719392313634457e-06,
+      "loss": 0.3511,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4067768798457562,
+      "grad_norm": 0.6143614122653783,
+      "learning_rate": 6.718144579108751e-06,
+      "loss": 0.342,
+      "step": 4958
+    },
+    {
+      "epoch": 0.406858924395947,
+      "grad_norm": 0.6490121842429755,
+      "learning_rate": 6.716896723250435e-06,
+      "loss": 0.3445,
+      "step": 4959
+    },
+    {
+      "epoch": 0.40694096894613774,
+      "grad_norm": 0.6077458181808136,
+      "learning_rate": 6.715648746147632e-06,
+      "loss": 0.3343,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4070230134963285,
+      "grad_norm": 0.6388800332236465,
+      "learning_rate": 6.714400647888468e-06,
+      "loss": 0.3286,
+      "step": 4961
+    },
+    {
+      "epoch": 0.40710505804651925,
+      "grad_norm": 0.6001355746285076,
+      "learning_rate": 6.7131524285610875e-06,
+      "loss": 0.3201,
+      "step": 4962
+    },
+    {
+      "epoch": 0.40718710259671004,
+      "grad_norm": 0.5746255611035749,
+      "learning_rate": 6.7119040882536335e-06,
+      "loss": 0.3063,
+      "step": 4963
+    },
+    {
+      "epoch": 0.40726914714690077,
+      "grad_norm": 0.6608332809172109,
+      "learning_rate": 6.7106556270542635e-06,
+      "loss": 0.3646,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4073511916970915,
+      "grad_norm": 0.6506754507602338,
+      "learning_rate": 6.709407045051141e-06,
+      "loss": 0.3538,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4074332362472823,
+      "grad_norm": 0.651193599222472,
+      "learning_rate": 6.70815834233244e-06,
+      "loss": 0.382,
+      "step": 4966
+    },
+    {
+      "epoch": 0.407515280797473,
+      "grad_norm": 0.6207900176991058,
+      "learning_rate": 6.706909518986341e-06,
+      "loss": 0.337,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4075973253476638,
+      "grad_norm": 0.6355801083281849,
+      "learning_rate": 6.705660575101034e-06,
+      "loss": 0.3199,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4076793698978545,
+      "grad_norm": 0.6704655640104885,
+      "learning_rate": 6.704411510764718e-06,
+      "loss": 0.3804,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4077614144480453,
+      "grad_norm": 0.6744296722482246,
+      "learning_rate": 6.703162326065597e-06,
+      "loss": 0.3711,
+      "step": 4970
+    },
+    {
+      "epoch": 0.40784345899823604,
+      "grad_norm": 0.6615010622717705,
+      "learning_rate": 6.701913021091891e-06,
+      "loss": 0.3353,
+      "step": 4971
+    },
+    {
+      "epoch": 0.40792550354842677,
+      "grad_norm": 0.6648781605942418,
+      "learning_rate": 6.700663595931822e-06,
+      "loss": 0.3358,
+      "step": 4972
+    },
+    {
+      "epoch": 0.40800754809861756,
+      "grad_norm": 0.6660883913773978,
+      "learning_rate": 6.699414050673623e-06,
+      "loss": 0.3722,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4080895926488083,
+      "grad_norm": 0.6683334757855927,
+      "learning_rate": 6.6981643854055335e-06,
+      "loss": 0.3409,
+      "step": 4974
+    },
+    {
+      "epoch": 0.40817163719899907,
+      "grad_norm": 0.6360582508547494,
+      "learning_rate": 6.696914600215803e-06,
+      "loss": 0.3192,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4082536817491898,
+      "grad_norm": 0.642414970025541,
+      "learning_rate": 6.695664695192689e-06,
+      "loss": 0.3673,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4083357262993806,
+      "grad_norm": 0.6967030888198229,
+      "learning_rate": 6.694414670424458e-06,
+      "loss": 0.3637,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4084177708495713,
+      "grad_norm": 0.6573774065649983,
+      "learning_rate": 6.693164525999387e-06,
+      "loss": 0.2912,
+      "step": 4978
+    },
+    {
+      "epoch": 0.40849981539976205,
+      "grad_norm": 0.651196500176776,
+      "learning_rate": 6.691914262005756e-06,
+      "loss": 0.3448,
+      "step": 4979
+    },
+    {
+      "epoch": 0.40858185994995283,
+      "grad_norm": 0.6452126886409756,
+      "learning_rate": 6.69066387853186e-06,
+      "loss": 0.3124,
+      "step": 4980
+    },
+    {
+      "epoch": 0.40866390450014356,
+      "grad_norm": 0.6640241667349844,
+      "learning_rate": 6.689413375665996e-06,
+      "loss": 0.3063,
+      "step": 4981
+    },
+    {
+      "epoch": 0.40874594905033435,
+      "grad_norm": 0.6369922072037174,
+      "learning_rate": 6.6881627534964745e-06,
+      "loss": 0.3167,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4088279936005251,
+      "grad_norm": 0.6106326831269132,
+      "learning_rate": 6.68691201211161e-06,
+      "loss": 0.3355,
+      "step": 4983
+    },
+    {
+      "epoch": 0.40891003815071586,
+      "grad_norm": 0.5821905421685835,
+      "learning_rate": 6.68566115159973e-06,
+      "loss": 0.3474,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4089920827009066,
+      "grad_norm": 0.6815429786556128,
+      "learning_rate": 6.6844101720491685e-06,
+      "loss": 0.3776,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4090741272510973,
+      "grad_norm": 0.6074691499205703,
+      "learning_rate": 6.6831590735482665e-06,
+      "loss": 0.3455,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4091561718012881,
+      "grad_norm": 0.680081132596233,
+      "learning_rate": 6.681907856185375e-06,
+      "loss": 0.3429,
+      "step": 4987
+    },
+    {
+      "epoch": 0.40923821635147883,
+      "grad_norm": 0.6639730929872217,
+      "learning_rate": 6.680656520048852e-06,
+      "loss": 0.3808,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4093202609016696,
+      "grad_norm": 0.6237737987327995,
+      "learning_rate": 6.679405065227067e-06,
+      "loss": 0.3411,
+      "step": 4989
+    },
+    {
+      "epoch": 0.40940230545186035,
+      "grad_norm": 0.6226136934287759,
+      "learning_rate": 6.678153491808394e-06,
+      "loss": 0.3599,
+      "step": 4990
+    },
+    {
+      "epoch": 0.40948435000205113,
+      "grad_norm": 0.6661661717329688,
+      "learning_rate": 6.6769017998812165e-06,
+      "loss": 0.3316,
+      "step": 4991
+    },
+    {
+      "epoch": 0.40956639455224186,
+      "grad_norm": 0.6748625486470552,
+      "learning_rate": 6.67564998953393e-06,
+      "loss": 0.3457,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4096484391024326,
+      "grad_norm": 0.6366100156841996,
+      "learning_rate": 6.674398060854931e-06,
+      "loss": 0.4017,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4097304836526234,
+      "grad_norm": 0.6742550067557929,
+      "learning_rate": 6.673146013932633e-06,
+      "loss": 0.3469,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4098125282028141,
+      "grad_norm": 0.6709160177844344,
+      "learning_rate": 6.671893848855451e-06,
+      "loss": 0.3896,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4098945727530049,
+      "grad_norm": 0.6649031515274719,
+      "learning_rate": 6.670641565711812e-06,
+      "loss": 0.3275,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4099766173031956,
+      "grad_norm": 0.6263667572897673,
+      "learning_rate": 6.6693891645901475e-06,
+      "loss": 0.3169,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4100586618533864,
+      "grad_norm": 0.6015805321313127,
+      "learning_rate": 6.668136645578904e-06,
+      "loss": 0.3566,
+      "step": 4998
+    },
+    {
+      "epoch": 0.41014070640357714,
+      "grad_norm": 0.630318823135579,
+      "learning_rate": 6.666884008766532e-06,
+      "loss": 0.3246,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4102227509537679,
+      "grad_norm": 0.6277027701259991,
+      "learning_rate": 6.665631254241488e-06,
+      "loss": 0.3493,
+      "step": 5000
+    },
+    {
+      "epoch": 0.41030479550395865,
+      "grad_norm": 0.6388392988352803,
+      "learning_rate": 6.66437838209224e-06,
+      "loss": 0.3632,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4103868400541494,
+      "grad_norm": 0.6497015136664338,
+      "learning_rate": 6.663125392407266e-06,
+      "loss": 0.396,
+      "step": 5002
+    },
+    {
+      "epoch": 0.41046888460434017,
+      "grad_norm": 0.7111299049647726,
+      "learning_rate": 6.661872285275048e-06,
+      "loss": 0.3835,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4105509291545309,
+      "grad_norm": 0.6585530007357562,
+      "learning_rate": 6.66061906078408e-06,
+      "loss": 0.3417,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4106329737047217,
+      "grad_norm": 0.6142745560125971,
+      "learning_rate": 6.659365719022862e-06,
+      "loss": 0.3412,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4107150182549124,
+      "grad_norm": 0.5960536422366214,
+      "learning_rate": 6.6581122600799005e-06,
+      "loss": 0.3561,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4107970628051032,
+      "grad_norm": 0.587578705635884,
+      "learning_rate": 6.656858684043718e-06,
+      "loss": 0.3006,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4108791073552939,
+      "grad_norm": 0.6779802107705379,
+      "learning_rate": 6.655604991002838e-06,
+      "loss": 0.3511,
+      "step": 5008
+    },
+    {
+      "epoch": 0.41096115190548466,
+      "grad_norm": 0.7216523391797738,
+      "learning_rate": 6.654351181045791e-06,
+      "loss": 0.3556,
+      "step": 5009
+    },
+    {
+      "epoch": 0.41104319645567544,
+      "grad_norm": 0.6486286644516935,
+      "learning_rate": 6.653097254261123e-06,
+      "loss": 0.3757,
+      "step": 5010
+    },
+    {
+      "epoch": 0.41112524100586617,
+      "grad_norm": 0.677709925093962,
+      "learning_rate": 6.6518432107373805e-06,
+      "loss": 0.3588,
+      "step": 5011
+    },
+    {
+      "epoch": 0.41120728555605696,
+      "grad_norm": 0.6544873266376132,
+      "learning_rate": 6.6505890505631255e-06,
+      "loss": 0.3364,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4112893301062477,
+      "grad_norm": 0.6174123392378338,
+      "learning_rate": 6.649334773826924e-06,
+      "loss": 0.3445,
+      "step": 5013
+    },
+    {
+      "epoch": 0.41137137465643847,
+      "grad_norm": 0.6772261338339999,
+      "learning_rate": 6.648080380617351e-06,
+      "loss": 0.3777,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4114534192066292,
+      "grad_norm": 0.6383983032210293,
+      "learning_rate": 6.646825871022989e-06,
+      "loss": 0.3612,
+      "step": 5015
+    },
+    {
+      "epoch": 0.41153546375681993,
+      "grad_norm": 0.7600111205851326,
+      "learning_rate": 6.6455712451324295e-06,
+      "loss": 0.3566,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4116175083070107,
+      "grad_norm": 0.5990190901779086,
+      "learning_rate": 6.644316503034274e-06,
+      "loss": 0.2982,
+      "step": 5017
+    },
+    {
+      "epoch": 0.41169955285720145,
+      "grad_norm": 0.6501365718313655,
+      "learning_rate": 6.643061644817127e-06,
+      "loss": 0.324,
+      "step": 5018
+    },
+    {
+      "epoch": 0.41178159740739223,
+      "grad_norm": 0.6286857421259688,
+      "learning_rate": 6.641806670569607e-06,
+      "loss": 0.3626,
+      "step": 5019
+    },
+    {
+      "epoch": 0.41186364195758296,
+      "grad_norm": 0.6314497638014579,
+      "learning_rate": 6.640551580380337e-06,
+      "loss": 0.3174,
+      "step": 5020
+    },
+    {
+      "epoch": 0.41194568650777375,
+      "grad_norm": 0.6422253268157427,
+      "learning_rate": 6.639296374337952e-06,
+      "loss": 0.3286,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4120277310579645,
+      "grad_norm": 0.6065883730198437,
+      "learning_rate": 6.63804105253109e-06,
+      "loss": 0.3562,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4121097756081552,
+      "grad_norm": 0.6899660047667886,
+      "learning_rate": 6.6367856150484e-06,
+      "loss": 0.3163,
+      "step": 5023
+    },
+    {
+      "epoch": 0.412191820158346,
+      "grad_norm": 0.6657463011718957,
+      "learning_rate": 6.635530061978539e-06,
+      "loss": 0.3677,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4122738647085367,
+      "grad_norm": 0.6462178390387865,
+      "learning_rate": 6.634274393410174e-06,
+      "loss": 0.3708,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4123559092587275,
+      "grad_norm": 0.6612016531120649,
+      "learning_rate": 6.633018609431976e-06,
+      "loss": 0.3744,
+      "step": 5026
+    },
+    {
+      "epoch": 0.41243795380891823,
+      "grad_norm": 0.5999474416529031,
+      "learning_rate": 6.631762710132628e-06,
+      "loss": 0.3386,
+      "step": 5027
+    },
+    {
+      "epoch": 0.412519998359109,
+      "grad_norm": 0.6407497962099511,
+      "learning_rate": 6.630506695600819e-06,
+      "loss": 0.3648,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41260204290929975,
+      "grad_norm": 0.603437618416668,
+      "learning_rate": 6.6292505659252475e-06,
+      "loss": 0.3323,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4126840874594905,
+      "grad_norm": 0.657752716366616,
+      "learning_rate": 6.627994321194618e-06,
+      "loss": 0.3898,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41276613200968126,
+      "grad_norm": 0.6296077181788612,
+      "learning_rate": 6.626737961497645e-06,
+      "loss": 0.316,
+      "step": 5031
+    },
+    {
+      "epoch": 0.412848176559872,
+      "grad_norm": 0.6754773090990753,
+      "learning_rate": 6.62548148692305e-06,
+      "loss": 0.3247,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4129302211100628,
+      "grad_norm": 0.6908093135626663,
+      "learning_rate": 6.6242248975595636e-06,
+      "loss": 0.315,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4130122656602535,
+      "grad_norm": 0.6147823994390029,
+      "learning_rate": 6.622968193495925e-06,
+      "loss": 0.3549,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4130943102104443,
+      "grad_norm": 0.7033291754510901,
+      "learning_rate": 6.621711374820881e-06,
+      "loss": 0.3135,
+      "step": 5035
+    },
+    {
+      "epoch": 0.413176354760635,
+      "grad_norm": 0.6143944810929215,
+      "learning_rate": 6.6204544416231865e-06,
+      "loss": 0.3328,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41325839931082575,
+      "grad_norm": 0.6697546729965775,
+      "learning_rate": 6.619197393991601e-06,
+      "loss": 0.3777,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41334044386101654,
+      "grad_norm": 0.6410777126923382,
+      "learning_rate": 6.617940232014896e-06,
+      "loss": 0.3746,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41342248841120727,
+      "grad_norm": 0.6843113135794406,
+      "learning_rate": 6.616682955781853e-06,
+      "loss": 0.334,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41350453296139805,
+      "grad_norm": 0.6478038706513792,
+      "learning_rate": 6.615425565381255e-06,
+      "loss": 0.339,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4135865775115888,
+      "grad_norm": 0.6197252842322732,
+      "learning_rate": 6.614168060901901e-06,
+      "loss": 0.3611,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41366862206177957,
+      "grad_norm": 0.580822767388757,
+      "learning_rate": 6.612910442432592e-06,
+      "loss": 0.3298,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4137506666119703,
+      "grad_norm": 0.6841798946756973,
+      "learning_rate": 6.611652710062138e-06,
+      "loss": 0.3833,
+      "step": 5043
+    },
+    {
+      "epoch": 0.413832711162161,
+      "grad_norm": 0.6156749782527996,
+      "learning_rate": 6.610394863879358e-06,
+      "loss": 0.3596,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4139147557123518,
+      "grad_norm": 0.6755142132743711,
+      "learning_rate": 6.609136903973081e-06,
+      "loss": 0.3503,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41399680026254254,
+      "grad_norm": 0.6917881531617888,
+      "learning_rate": 6.6078788304321405e-06,
+      "loss": 0.3385,
+      "step": 5046
+    },
+    {
+      "epoch": 0.4140788448127333,
+      "grad_norm": 0.618165841089808,
+      "learning_rate": 6.60662064334538e-06,
+      "loss": 0.3457,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41416088936292406,
+      "grad_norm": 0.6222639769583619,
+      "learning_rate": 6.6053623428016535e-06,
+      "loss": 0.3335,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41424293391311484,
+      "grad_norm": 0.7582516056953096,
+      "learning_rate": 6.604103928889816e-06,
+      "loss": 0.3955,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41432497846330557,
+      "grad_norm": 0.6507232094527547,
+      "learning_rate": 6.602845401698736e-06,
+      "loss": 0.3274,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41440702301349636,
+      "grad_norm": 0.6236274576318941,
+      "learning_rate": 6.601586761317289e-06,
+      "loss": 0.3589,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4144890675636871,
+      "grad_norm": 0.7162303836735988,
+      "learning_rate": 6.600328007834358e-06,
+      "loss": 0.3786,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4145711121138778,
+      "grad_norm": 0.631251380470334,
+      "learning_rate": 6.599069141338834e-06,
+      "loss": 0.3226,
+      "step": 5053
+    },
+    {
+      "epoch": 0.4146531566640686,
+      "grad_norm": 0.6179039658323782,
+      "learning_rate": 6.5978101619196165e-06,
+      "loss": 0.3167,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41473520121425933,
+      "grad_norm": 0.6523356349842412,
+      "learning_rate": 6.596551069665611e-06,
+      "loss": 0.34,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4148172457644501,
+      "grad_norm": 0.6976366814114234,
+      "learning_rate": 6.595291864665737e-06,
+      "loss": 0.3442,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41489929031464085,
+      "grad_norm": 0.6326465448514241,
+      "learning_rate": 6.594032547008913e-06,
+      "loss": 0.3597,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41498133486483163,
+      "grad_norm": 0.6511219360017153,
+      "learning_rate": 6.592773116784072e-06,
+      "loss": 0.3465,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41506337941502236,
+      "grad_norm": 0.611247221964164,
+      "learning_rate": 6.591513574080152e-06,
+      "loss": 0.3667,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4151454239652131,
+      "grad_norm": 0.6396968802643342,
+      "learning_rate": 6.590253918986099e-06,
+      "loss": 0.3254,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4152274685154039,
+      "grad_norm": 0.6376107730003008,
+      "learning_rate": 6.58899415159087e-06,
+      "loss": 0.3091,
+      "step": 5061
+    },
+    {
+      "epoch": 0.4153095130655946,
+      "grad_norm": 0.6324296699526295,
+      "learning_rate": 6.587734271983429e-06,
+      "loss": 0.3304,
+      "step": 5062
+    },
+    {
+      "epoch": 0.4153915576157854,
+      "grad_norm": 0.7193730777701783,
+      "learning_rate": 6.586474280252744e-06,
+      "loss": 0.3677,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4154736021659761,
+      "grad_norm": 0.6201864146979013,
+      "learning_rate": 6.585214176487793e-06,
+      "loss": 0.3661,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4155556467161669,
+      "grad_norm": 0.651058540542821,
+      "learning_rate": 6.583953960777563e-06,
+      "loss": 0.36,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41563769126635763,
+      "grad_norm": 0.6437902598990766,
+      "learning_rate": 6.58269363321105e-06,
+      "loss": 0.3426,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41571973581654836,
+      "grad_norm": 0.6506312834425303,
+      "learning_rate": 6.581433193877257e-06,
+      "loss": 0.3053,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41580178036673915,
+      "grad_norm": 0.6872231678645984,
+      "learning_rate": 6.5801726428651906e-06,
+      "loss": 0.3376,
+      "step": 5068
+    },
+    {
+      "epoch": 0.4158838249169299,
+      "grad_norm": 0.6255709427305678,
+      "learning_rate": 6.578911980263872e-06,
+      "loss": 0.3261,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41596586946712066,
+      "grad_norm": 0.6313467790526498,
+      "learning_rate": 6.577651206162325e-06,
+      "loss": 0.3184,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4160479140173114,
+      "grad_norm": 0.6790914053549727,
+      "learning_rate": 6.576390320649586e-06,
+      "loss": 0.3434,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4161299585675022,
+      "grad_norm": 0.6375229931221533,
+      "learning_rate": 6.575129323814694e-06,
+      "loss": 0.3344,
+      "step": 5072
+    },
+    {
+      "epoch": 0.4162120031176929,
+      "grad_norm": 0.6706996688206027,
+      "learning_rate": 6.5738682157467e-06,
+      "loss": 0.3808,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41629404766788364,
+      "grad_norm": 0.6148390229047223,
+      "learning_rate": 6.572606996534661e-06,
+      "loss": 0.3403,
+      "step": 5074
+    },
+    {
+      "epoch": 0.4163760922180744,
+      "grad_norm": 0.6213202164534833,
+      "learning_rate": 6.5713456662676424e-06,
+      "loss": 0.3357,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41645813676826515,
+      "grad_norm": 0.6127336167850783,
+      "learning_rate": 6.570084225034717e-06,
+      "loss": 0.339,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41654018131845594,
+      "grad_norm": 0.6499561991437104,
+      "learning_rate": 6.568822672924968e-06,
+      "loss": 0.3329,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41662222586864667,
+      "grad_norm": 0.6159956260017914,
+      "learning_rate": 6.567561010027481e-06,
+      "loss": 0.3441,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41670427041883745,
+      "grad_norm": 0.6096562550524244,
+      "learning_rate": 6.566299236431355e-06,
+      "loss": 0.3712,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4167863149690282,
+      "grad_norm": 0.692489448834971,
+      "learning_rate": 6.565037352225692e-06,
+      "loss": 0.3151,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4168683595192189,
+      "grad_norm": 0.7263974375177631,
+      "learning_rate": 6.563775357499606e-06,
+      "loss": 0.4025,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4169504040694097,
+      "grad_norm": 0.6362986087024752,
+      "learning_rate": 6.562513252342216e-06,
+      "loss": 0.3153,
+      "step": 5082
+    },
+    {
+      "epoch": 0.4170324486196004,
+      "grad_norm": 0.706697048482703,
+      "learning_rate": 6.561251036842652e-06,
+      "loss": 0.355,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4171144931697912,
+      "grad_norm": 0.6565062854086267,
+      "learning_rate": 6.559988711090048e-06,
+      "loss": 0.3376,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41719653771998194,
+      "grad_norm": 0.7246218485710798,
+      "learning_rate": 6.558726275173548e-06,
+      "loss": 0.3452,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4172785822701727,
+      "grad_norm": 0.6654911888190091,
+      "learning_rate": 6.5574637291823005e-06,
+      "loss": 0.3611,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41736062682036346,
+      "grad_norm": 0.6649737998123993,
+      "learning_rate": 6.556201073205468e-06,
+      "loss": 0.3565,
+      "step": 5087
+    },
+    {
+      "epoch": 0.4174426713705542,
+      "grad_norm": 0.6448825415273444,
+      "learning_rate": 6.554938307332216e-06,
+      "loss": 0.3523,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41752471592074497,
+      "grad_norm": 0.8023174810400285,
+      "learning_rate": 6.553675431651718e-06,
+      "loss": 0.3372,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4176067604709357,
+      "grad_norm": 0.6361580947397938,
+      "learning_rate": 6.552412446253159e-06,
+      "loss": 0.3904,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4176888050211265,
+      "grad_norm": 0.6700893647650392,
+      "learning_rate": 6.5511493512257265e-06,
+      "loss": 0.3862,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4177708495713172,
+      "grad_norm": 0.9775117940306062,
+      "learning_rate": 6.549886146658618e-06,
+      "loss": 0.3222,
+      "step": 5092
+    },
+    {
+      "epoch": 0.417852894121508,
+      "grad_norm": 0.7059258508229287,
+      "learning_rate": 6.5486228326410406e-06,
+      "loss": 0.3338,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41793493867169873,
+      "grad_norm": 0.6431229113939733,
+      "learning_rate": 6.547359409262207e-06,
+      "loss": 0.3261,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41801698322188946,
+      "grad_norm": 0.6608087662238608,
+      "learning_rate": 6.546095876611338e-06,
+      "loss": 0.3203,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41809902777208025,
+      "grad_norm": 0.7981759953077381,
+      "learning_rate": 6.54483223477766e-06,
+      "loss": 0.3525,
+      "step": 5096
+    },
+    {
+      "epoch": 0.418181072322271,
+      "grad_norm": 0.6395807150998258,
+      "learning_rate": 6.543568483850414e-06,
+      "loss": 0.3531,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41826311687246176,
+      "grad_norm": 0.6429427814445986,
+      "learning_rate": 6.542304623918841e-06,
+      "loss": 0.3836,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4183451614226525,
+      "grad_norm": 0.677750762060539,
+      "learning_rate": 6.541040655072195e-06,
+      "loss": 0.3467,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4184272059728433,
+      "grad_norm": 0.5798273680252334,
+      "learning_rate": 6.5397765773997315e-06,
+      "loss": 0.3002,
+      "step": 5100
+    },
+    {
+      "epoch": 0.418509250523034,
+      "grad_norm": 0.666127162760314,
+      "learning_rate": 6.53851239099072e-06,
+      "loss": 0.3816,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41859129507322473,
+      "grad_norm": 0.6350125306014183,
+      "learning_rate": 6.537248095934436e-06,
+      "loss": 0.3108,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4186733396234155,
+      "grad_norm": 0.6715250123708563,
+      "learning_rate": 6.535983692320161e-06,
+      "loss": 0.3428,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41875538417360625,
+      "grad_norm": 0.6551200243017691,
+      "learning_rate": 6.534719180237185e-06,
+      "loss": 0.3455,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41883742872379703,
+      "grad_norm": 0.622514392963851,
+      "learning_rate": 6.5334545597748075e-06,
+      "loss": 0.3265,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41891947327398776,
+      "grad_norm": 0.7169768199589607,
+      "learning_rate": 6.532189831022332e-06,
+      "loss": 0.3632,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41900151782417855,
+      "grad_norm": 0.6779559285649769,
+      "learning_rate": 6.530924994069072e-06,
+      "loss": 0.3615,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4190835623743693,
+      "grad_norm": 0.6429137550679064,
+      "learning_rate": 6.529660049004349e-06,
+      "loss": 0.3844,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41916560692456006,
+      "grad_norm": 0.62625198760087,
+      "learning_rate": 6.52839499591749e-06,
+      "loss": 0.3534,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4192476514747508,
+      "grad_norm": 0.6522556194565455,
+      "learning_rate": 6.527129834897833e-06,
+      "loss": 0.3173,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4193296960249415,
+      "grad_norm": 0.7877024825378831,
+      "learning_rate": 6.52586456603472e-06,
+      "loss": 0.3848,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4194117405751323,
+      "grad_norm": 0.646900802180497,
+      "learning_rate": 6.524599189417503e-06,
+      "loss": 0.3464,
+      "step": 5112
+    },
+    {
+      "epoch": 0.41949378512532304,
+      "grad_norm": 0.6524993889663047,
+      "learning_rate": 6.523333705135542e-06,
+      "loss": 0.3238,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4195758296755138,
+      "grad_norm": 0.6584697092461032,
+      "learning_rate": 6.522068113278199e-06,
+      "loss": 0.3497,
+      "step": 5114
+    },
+    {
+      "epoch": 0.41965787422570455,
+      "grad_norm": 0.6140846600838934,
+      "learning_rate": 6.520802413934855e-06,
+      "loss": 0.3496,
+      "step": 5115
+    },
+    {
+      "epoch": 0.41973991877589534,
+      "grad_norm": 0.6866999449748009,
+      "learning_rate": 6.519536607194885e-06,
+      "loss": 0.3352,
+      "step": 5116
+    },
+    {
+      "epoch": 0.41982196332608607,
+      "grad_norm": 0.6670290636997397,
+      "learning_rate": 6.518270693147682e-06,
+      "loss": 0.3775,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4199040078762768,
+      "grad_norm": 0.6682563661796914,
+      "learning_rate": 6.517004671882643e-06,
+      "loss": 0.3735,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4199860524264676,
+      "grad_norm": 0.6304420809434466,
+      "learning_rate": 6.51573854348917e-06,
+      "loss": 0.3201,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4200680969766583,
+      "grad_norm": 0.6003627795122776,
+      "learning_rate": 6.514472308056677e-06,
+      "loss": 0.3347,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4201501415268491,
+      "grad_norm": 0.6834723478763514,
+      "learning_rate": 6.513205965674583e-06,
+      "loss": 0.3292,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4202321860770398,
+      "grad_norm": 0.5904014297528934,
+      "learning_rate": 6.511939516432313e-06,
+      "loss": 0.3207,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4203142306272306,
+      "grad_norm": 0.6241085154423279,
+      "learning_rate": 6.5106729604193045e-06,
+      "loss": 0.3202,
+      "step": 5123
+    },
+    {
+      "epoch": 0.42039627517742134,
+      "grad_norm": 0.6858295611488902,
+      "learning_rate": 6.5094062977249985e-06,
+      "loss": 0.4185,
+      "step": 5124
+    },
+    {
+      "epoch": 0.42047831972761207,
+      "grad_norm": 0.6722858420586281,
+      "learning_rate": 6.508139528438846e-06,
+      "loss": 0.3608,
+      "step": 5125
+    },
+    {
+      "epoch": 0.42056036427780286,
+      "grad_norm": 0.6686253691235785,
+      "learning_rate": 6.5068726526503026e-06,
+      "loss": 0.339,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4206424088279936,
+      "grad_norm": 0.6490139173847522,
+      "learning_rate": 6.505605670448832e-06,
+      "loss": 0.3603,
+      "step": 5127
+    },
+    {
+      "epoch": 0.42072445337818437,
+      "grad_norm": 0.6919949925551347,
+      "learning_rate": 6.5043385819239095e-06,
+      "loss": 0.3696,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4208064979283751,
+      "grad_norm": 0.6706837027821417,
+      "learning_rate": 6.503071387165012e-06,
+      "loss": 0.3218,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4208885424785659,
+      "grad_norm": 0.6421837180712053,
+      "learning_rate": 6.50180408626163e-06,
+      "loss": 0.3382,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4209705870287566,
+      "grad_norm": 0.6425590303012557,
+      "learning_rate": 6.500536679303254e-06,
+      "loss": 0.333,
+      "step": 5131
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.7879399601202434,
+      "learning_rate": 6.499269166379389e-06,
+      "loss": 0.3298,
+      "step": 5132
+    },
+    {
+      "epoch": 0.42113467612913813,
+      "grad_norm": 0.6388955790303262,
+      "learning_rate": 6.498001547579545e-06,
+      "loss": 0.3613,
+      "step": 5133
+    },
+    {
+      "epoch": 0.42121672067932886,
+      "grad_norm": 0.6727981746768384,
+      "learning_rate": 6.4967338229932385e-06,
+      "loss": 0.3557,
+      "step": 5134
+    },
+    {
+      "epoch": 0.42129876522951965,
+      "grad_norm": 0.5999735811434077,
+      "learning_rate": 6.495465992709994e-06,
+      "loss": 0.312,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4213808097797104,
+      "grad_norm": 0.6235710902549307,
+      "learning_rate": 6.494198056819343e-06,
+      "loss": 0.3185,
+      "step": 5136
+    },
+    {
+      "epoch": 0.42146285432990116,
+      "grad_norm": 0.6217402296585255,
+      "learning_rate": 6.492930015410826e-06,
+      "loss": 0.3368,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4215448988800919,
+      "grad_norm": 0.6526706913681306,
+      "learning_rate": 6.49166186857399e-06,
+      "loss": 0.3484,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4216269434302826,
+      "grad_norm": 0.6168835642621397,
+      "learning_rate": 6.4903936163983895e-06,
+      "loss": 0.3651,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4217089879804734,
+      "grad_norm": 0.6231982961736996,
+      "learning_rate": 6.4891252589735875e-06,
+      "loss": 0.308,
+      "step": 5140
+    },
+    {
+      "epoch": 0.42179103253066413,
+      "grad_norm": 0.6414335877794801,
+      "learning_rate": 6.48785679638915e-06,
+      "loss": 0.3389,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4218730770808549,
+      "grad_norm": 0.6389372010337849,
+      "learning_rate": 6.486588228734658e-06,
+      "loss": 0.3737,
+      "step": 5142
+    },
+    {
+      "epoch": 0.42195512163104565,
+      "grad_norm": 0.6157694245536742,
+      "learning_rate": 6.485319556099693e-06,
+      "loss": 0.3407,
+      "step": 5143
+    },
+    {
+      "epoch": 0.42203716618123643,
+      "grad_norm": 0.6538066859436552,
+      "learning_rate": 6.484050778573847e-06,
+      "loss": 0.3109,
+      "step": 5144
+    },
+    {
+      "epoch": 0.42211921073142716,
+      "grad_norm": 0.5982155101437681,
+      "learning_rate": 6.482781896246718e-06,
+      "loss": 0.3179,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4222012552816179,
+      "grad_norm": 0.6409813849173083,
+      "learning_rate": 6.481512909207915e-06,
+      "loss": 0.353,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4222832998318087,
+      "grad_norm": 0.6606571796837007,
+      "learning_rate": 6.480243817547051e-06,
+      "loss": 0.3309,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4223653443819994,
+      "grad_norm": 0.6050975223031237,
+      "learning_rate": 6.4789746213537475e-06,
+      "loss": 0.3258,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4224473889321902,
+      "grad_norm": 0.6336891024507462,
+      "learning_rate": 6.477705320717631e-06,
+      "loss": 0.3382,
+      "step": 5149
+    },
+    {
+      "epoch": 0.4225294334823809,
+      "grad_norm": 0.6128775665976234,
+      "learning_rate": 6.4764359157283385e-06,
+      "loss": 0.375,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4226114780325717,
+      "grad_norm": 0.63272285733197,
+      "learning_rate": 6.475166406475515e-06,
+      "loss": 0.3185,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42269352258276244,
+      "grad_norm": 0.745781090368366,
+      "learning_rate": 6.473896793048808e-06,
+      "loss": 0.3299,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42277556713295317,
+      "grad_norm": 0.6163465611759131,
+      "learning_rate": 6.472627075537879e-06,
+      "loss": 0.38,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42285761168314395,
+      "grad_norm": 0.6115683748262448,
+      "learning_rate": 6.471357254032394e-06,
+      "loss": 0.343,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4229396562333347,
+      "grad_norm": 0.5769321836362016,
+      "learning_rate": 6.47008732862202e-06,
+      "loss": 0.3002,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42302170078352547,
+      "grad_norm": 0.693161448636482,
+      "learning_rate": 6.468817299396442e-06,
+      "loss": 0.3307,
+      "step": 5156
+    },
+    {
+      "epoch": 0.4231037453337162,
+      "grad_norm": 0.7334249379087152,
+      "learning_rate": 6.467547166445347e-06,
+      "loss": 0.3691,
+      "step": 5157
+    },
+    {
+      "epoch": 0.423185789883907,
+      "grad_norm": 0.6765202913300221,
+      "learning_rate": 6.466276929858428e-06,
+      "loss": 0.3268,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4232678344340977,
+      "grad_norm": 0.6789276754008495,
+      "learning_rate": 6.465006589725386e-06,
+      "loss": 0.3849,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42334987898428844,
+      "grad_norm": 0.6364127168570713,
+      "learning_rate": 6.463736146135935e-06,
+      "loss": 0.3176,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4234319235344792,
+      "grad_norm": 0.6616168869162704,
+      "learning_rate": 6.46246559917979e-06,
+      "loss": 0.3129,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42351396808466996,
+      "grad_norm": 0.7708904148942255,
+      "learning_rate": 6.461194948946672e-06,
+      "loss": 0.3736,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42359601263486074,
+      "grad_norm": 0.6767272494905269,
+      "learning_rate": 6.4599241955263145e-06,
+      "loss": 0.3678,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42367805718505147,
+      "grad_norm": 0.6596296536584656,
+      "learning_rate": 6.458653339008456e-06,
+      "loss": 0.3258,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42376010173524226,
+      "grad_norm": 0.6686029693003641,
+      "learning_rate": 6.457382379482842e-06,
+      "loss": 0.3672,
+      "step": 5165
+    },
+    {
+      "epoch": 0.423842146285433,
+      "grad_norm": 0.6871668801814365,
+      "learning_rate": 6.456111317039226e-06,
+      "loss": 0.3326,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42392419083562377,
+      "grad_norm": 0.6224213201921864,
+      "learning_rate": 6.454840151767369e-06,
+      "loss": 0.339,
+      "step": 5167
+    },
+    {
+      "epoch": 0.4240062353858145,
+      "grad_norm": 0.6426333242406136,
+      "learning_rate": 6.453568883757038e-06,
+      "loss": 0.3913,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42408827993600523,
+      "grad_norm": 0.7206447243653405,
+      "learning_rate": 6.452297513098006e-06,
+      "loss": 0.3366,
+      "step": 5169
+    },
+    {
+      "epoch": 0.424170324486196,
+      "grad_norm": 0.6780156884610461,
+      "learning_rate": 6.451026039880059e-06,
+      "loss": 0.3342,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42425236903638675,
+      "grad_norm": 0.6392570460908237,
+      "learning_rate": 6.449754464192983e-06,
+      "loss": 0.3214,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42433441358657753,
+      "grad_norm": 0.700224876897939,
+      "learning_rate": 6.448482786126575e-06,
+      "loss": 0.3014,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42441645813676826,
+      "grad_norm": 0.633758341427693,
+      "learning_rate": 6.447211005770641e-06,
+      "loss": 0.333,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42449850268695904,
+      "grad_norm": 0.702374111251984,
+      "learning_rate": 6.445939123214991e-06,
+      "loss": 0.3676,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4245805472371498,
+      "grad_norm": 0.6512145731144774,
+      "learning_rate": 6.444667138549444e-06,
+      "loss": 0.3216,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4246625917873405,
+      "grad_norm": 0.625002260547807,
+      "learning_rate": 6.443395051863824e-06,
+      "loss": 0.319,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4247446363375313,
+      "grad_norm": 0.676068702594816,
+      "learning_rate": 6.442122863247962e-06,
+      "loss": 0.3188,
+      "step": 5177
+    },
+    {
+      "epoch": 0.424826680887722,
+      "grad_norm": 0.654211660610666,
+      "learning_rate": 6.440850572791703e-06,
+      "loss": 0.3539,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4249087254379128,
+      "grad_norm": 0.7322512914799778,
+      "learning_rate": 6.439578180584891e-06,
+      "loss": 0.3434,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42499076998810353,
+      "grad_norm": 0.6878009154107956,
+      "learning_rate": 6.438305686717381e-06,
+      "loss": 0.3609,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4250728145382943,
+      "grad_norm": 0.6275083242932019,
+      "learning_rate": 6.437033091279035e-06,
+      "loss": 0.3369,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42515485908848505,
+      "grad_norm": 0.6498015079937223,
+      "learning_rate": 6.435760394359719e-06,
+      "loss": 0.3525,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4252369036386758,
+      "grad_norm": 0.647568894711256,
+      "learning_rate": 6.434487596049313e-06,
+      "loss": 0.3524,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42531894818886656,
+      "grad_norm": 0.6024460731169333,
+      "learning_rate": 6.433214696437695e-06,
+      "loss": 0.3179,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4254009927390573,
+      "grad_norm": 0.6468147835518474,
+      "learning_rate": 6.43194169561476e-06,
+      "loss": 0.3426,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4254830372892481,
+      "grad_norm": 0.6444681840094403,
+      "learning_rate": 6.4306685936704e-06,
+      "loss": 0.3155,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4255650818394388,
+      "grad_norm": 0.6125312370168582,
+      "learning_rate": 6.429395390694525e-06,
+      "loss": 0.3206,
+      "step": 5187
+    },
+    {
+      "epoch": 0.4256471263896296,
+      "grad_norm": 0.6719795895467638,
+      "learning_rate": 6.428122086777044e-06,
+      "loss": 0.3694,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4257291709398203,
+      "grad_norm": 0.7072844437884663,
+      "learning_rate": 6.426848682007876e-06,
+      "loss": 0.3365,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42581121549001105,
+      "grad_norm": 0.7203318760489152,
+      "learning_rate": 6.425575176476944e-06,
+      "loss": 0.3437,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42589326004020184,
+      "grad_norm": 0.6255127061873847,
+      "learning_rate": 6.4243015702741854e-06,
+      "loss": 0.3792,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42597530459039257,
+      "grad_norm": 0.641770814738569,
+      "learning_rate": 6.423027863489539e-06,
+      "loss": 0.3282,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42605734914058335,
+      "grad_norm": 0.718779275273746,
+      "learning_rate": 6.421754056212949e-06,
+      "loss": 0.3463,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4261393936907741,
+      "grad_norm": 0.701082155676886,
+      "learning_rate": 6.420480148534373e-06,
+      "loss": 0.3853,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42622143824096487,
+      "grad_norm": 0.5919362274482954,
+      "learning_rate": 6.41920614054377e-06,
+      "loss": 0.3361,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4263034827911556,
+      "grad_norm": 0.7038698476091733,
+      "learning_rate": 6.417932032331111e-06,
+      "loss": 0.3252,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4263855273413463,
+      "grad_norm": 0.6273004494594092,
+      "learning_rate": 6.41665782398637e-06,
+      "loss": 0.3302,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4264675718915371,
+      "grad_norm": 0.6305240533975135,
+      "learning_rate": 6.415383515599528e-06,
+      "loss": 0.3222,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42654961644172784,
+      "grad_norm": 0.6324784597258244,
+      "learning_rate": 6.414109107260576e-06,
+      "loss": 0.3294,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4266316609919186,
+      "grad_norm": 0.6603074343241788,
+      "learning_rate": 6.412834599059512e-06,
+      "loss": 0.2929,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42671370554210936,
+      "grad_norm": 0.6752060611461368,
+      "learning_rate": 6.411559991086338e-06,
+      "loss": 0.3752,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42679575009230014,
+      "grad_norm": 0.6881531338194561,
+      "learning_rate": 6.410285283431066e-06,
+      "loss": 0.3427,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42687779464249087,
+      "grad_norm": 0.7001727066164868,
+      "learning_rate": 6.409010476183713e-06,
+      "loss": 0.3226,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4269598391926816,
+      "grad_norm": 0.6160462829846428,
+      "learning_rate": 6.407735569434304e-06,
+      "loss": 0.3159,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4270418837428724,
+      "grad_norm": 0.6241510678399531,
+      "learning_rate": 6.40646056327287e-06,
+      "loss": 0.3503,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4271239282930631,
+      "grad_norm": 0.6557599440524775,
+      "learning_rate": 6.40518545778945e-06,
+      "loss": 0.3115,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4272059728432539,
+      "grad_norm": 0.683439408575065,
+      "learning_rate": 6.403910253074091e-06,
+      "loss": 0.3305,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42728801739344463,
+      "grad_norm": 0.6602701443845505,
+      "learning_rate": 6.402634949216846e-06,
+      "loss": 0.352,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4273700619436354,
+      "grad_norm": 0.6793466778938498,
+      "learning_rate": 6.401359546307775e-06,
+      "loss": 0.3538,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42745210649382614,
+      "grad_norm": 0.5902600077754572,
+      "learning_rate": 6.400084044436944e-06,
+      "loss": 0.3032,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4275341510440169,
+      "grad_norm": 0.6239050078336231,
+      "learning_rate": 6.398808443694426e-06,
+      "loss": 0.3491,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42761619559420766,
+      "grad_norm": 0.6897131744938616,
+      "learning_rate": 6.397532744170305e-06,
+      "loss": 0.3431,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4276982401443984,
+      "grad_norm": 0.6630845692764604,
+      "learning_rate": 6.3962569459546655e-06,
+      "loss": 0.3699,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4277802846945892,
+      "grad_norm": 0.6256370385863597,
+      "learning_rate": 6.394981049137603e-06,
+      "loss": 0.3188,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4278623292447799,
+      "grad_norm": 0.7294911274223399,
+      "learning_rate": 6.393705053809221e-06,
+      "loss": 0.3347,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4279443737949707,
+      "grad_norm": 0.6852816425336488,
+      "learning_rate": 6.392428960059628e-06,
+      "loss": 0.3857,
+      "step": 5216
+    },
+    {
+      "epoch": 0.4280264183451614,
+      "grad_norm": 0.6149539419000086,
+      "learning_rate": 6.391152767978939e-06,
+      "loss": 0.3518,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42810846289535215,
+      "grad_norm": 0.6854941505993525,
+      "learning_rate": 6.389876477657275e-06,
+      "loss": 0.3342,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42819050744554293,
+      "grad_norm": 0.625936470953801,
+      "learning_rate": 6.388600089184769e-06,
+      "loss": 0.3526,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42827255199573366,
+      "grad_norm": 0.6232577211406911,
+      "learning_rate": 6.387323602651554e-06,
+      "loss": 0.3551,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42835459654592445,
+      "grad_norm": 0.6811953951797977,
+      "learning_rate": 6.386047018147776e-06,
+      "loss": 0.3686,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4284366410961152,
+      "grad_norm": 0.6883477694345871,
+      "learning_rate": 6.384770335763584e-06,
+      "loss": 0.3562,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42851868564630596,
+      "grad_norm": 0.6093799108535922,
+      "learning_rate": 6.383493555589136e-06,
+      "loss": 0.3524,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4286007301964967,
+      "grad_norm": 0.613873978620886,
+      "learning_rate": 6.382216677714597e-06,
+      "loss": 0.3644,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4286827747466875,
+      "grad_norm": 0.6743667402912125,
+      "learning_rate": 6.380939702230136e-06,
+      "loss": 0.3734,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4287648192968782,
+      "grad_norm": 0.6420071399206235,
+      "learning_rate": 6.37966262922593e-06,
+      "loss": 0.3528,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42884686384706894,
+      "grad_norm": 0.6370514626860275,
+      "learning_rate": 6.3783854587921666e-06,
+      "loss": 0.36,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4289289083972597,
+      "grad_norm": 0.6053231780139077,
+      "learning_rate": 6.377108191019038e-06,
+      "loss": 0.3167,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42901095294745045,
+      "grad_norm": 0.6346567151370308,
+      "learning_rate": 6.3758308259967404e-06,
+      "loss": 0.3151,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42909299749764124,
+      "grad_norm": 0.6578174319008205,
+      "learning_rate": 6.374553363815481e-06,
+      "loss": 0.3243,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42917504204783197,
+      "grad_norm": 0.6905181950605965,
+      "learning_rate": 6.373275804565471e-06,
+      "loss": 0.3484,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42925708659802275,
+      "grad_norm": 0.6332042465236195,
+      "learning_rate": 6.371998148336931e-06,
+      "loss": 0.3443,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4293391311482135,
+      "grad_norm": 0.6408808795961548,
+      "learning_rate": 6.370720395220085e-06,
+      "loss": 0.3554,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4294211756984042,
+      "grad_norm": 0.6695121960595787,
+      "learning_rate": 6.3694425453051665e-06,
+      "loss": 0.3516,
+      "step": 5234
+    },
+    {
+      "epoch": 0.429503220248595,
+      "grad_norm": 0.6653009485268677,
+      "learning_rate": 6.3681645986824155e-06,
+      "loss": 0.3729,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4295852647987857,
+      "grad_norm": 0.6750830139740812,
+      "learning_rate": 6.36688655544208e-06,
+      "loss": 0.3292,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4296673093489765,
+      "grad_norm": 0.6450712443154315,
+      "learning_rate": 6.365608415674412e-06,
+      "loss": 0.3401,
+      "step": 5237
+    },
+    {
+      "epoch": 0.42974935389916724,
+      "grad_norm": 0.655978074450412,
+      "learning_rate": 6.364330179469671e-06,
+      "loss": 0.3192,
+      "step": 5238
+    },
+    {
+      "epoch": 0.429831398449358,
+      "grad_norm": 0.6342942639745143,
+      "learning_rate": 6.3630518469181246e-06,
+      "loss": 0.3406,
+      "step": 5239
+    },
+    {
+      "epoch": 0.42991344299954876,
+      "grad_norm": 0.6067398283571768,
+      "learning_rate": 6.361773418110046e-06,
+      "loss": 0.3603,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4299954875497395,
+      "grad_norm": 0.6253228209171883,
+      "learning_rate": 6.360494893135718e-06,
+      "loss": 0.3078,
+      "step": 5241
+    },
+    {
+      "epoch": 0.43007753209993027,
+      "grad_norm": 0.6563721232056683,
+      "learning_rate": 6.359216272085426e-06,
+      "loss": 0.3243,
+      "step": 5242
+    },
+    {
+      "epoch": 0.430159576650121,
+      "grad_norm": 0.6335055832961313,
+      "learning_rate": 6.357937555049465e-06,
+      "loss": 0.3427,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4302416212003118,
+      "grad_norm": 0.6821082798169849,
+      "learning_rate": 6.356658742118135e-06,
+      "loss": 0.3433,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4303236657505025,
+      "grad_norm": 0.6256579622817607,
+      "learning_rate": 6.355379833381744e-06,
+      "loss": 0.3238,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4304057103006933,
+      "grad_norm": 0.6024407717261073,
+      "learning_rate": 6.354100828930607e-06,
+      "loss": 0.2994,
+      "step": 5246
+    },
+    {
+      "epoch": 0.43048775485088403,
+      "grad_norm": 0.6819714793577115,
+      "learning_rate": 6.3528217288550455e-06,
+      "loss": 0.3601,
+      "step": 5247
+    },
+    {
+      "epoch": 0.43056979940107476,
+      "grad_norm": 0.5916509097785213,
+      "learning_rate": 6.351542533245385e-06,
+      "loss": 0.3581,
+      "step": 5248
+    },
+    {
+      "epoch": 0.43065184395126554,
+      "grad_norm": 0.6486083863506376,
+      "learning_rate": 6.350263242191963e-06,
+      "loss": 0.2981,
+      "step": 5249
+    },
+    {
+      "epoch": 0.4307338885014563,
+      "grad_norm": 0.6399318135844196,
+      "learning_rate": 6.348983855785122e-06,
+      "loss": 0.3507,
+      "step": 5250
+    },
+    {
+      "epoch": 0.43081593305164706,
+      "grad_norm": 0.6341026956448524,
+      "learning_rate": 6.347704374115208e-06,
+      "loss": 0.2958,
+      "step": 5251
+    },
+    {
+      "epoch": 0.4308979776018378,
+      "grad_norm": 0.6167459001987715,
+      "learning_rate": 6.346424797272576e-06,
+      "loss": 0.3689,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4309800221520286,
+      "grad_norm": 0.5881058918400779,
+      "learning_rate": 6.345145125347587e-06,
+      "loss": 0.3271,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4310620667022193,
+      "grad_norm": 0.635261452801603,
+      "learning_rate": 6.343865358430611e-06,
+      "loss": 0.3167,
+      "step": 5254
+    },
+    {
+      "epoch": 0.43114411125241003,
+      "grad_norm": 0.6750927587839273,
+      "learning_rate": 6.342585496612021e-06,
+      "loss": 0.2825,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4312261558026008,
+      "grad_norm": 0.6997605604316285,
+      "learning_rate": 6.341305539982203e-06,
+      "loss": 0.362,
+      "step": 5256
+    },
+    {
+      "epoch": 0.43130820035279155,
+      "grad_norm": 0.5874013812022385,
+      "learning_rate": 6.340025488631541e-06,
+      "loss": 0.2927,
+      "step": 5257
+    },
+    {
+      "epoch": 0.43139024490298233,
+      "grad_norm": 0.6542020108547706,
+      "learning_rate": 6.338745342650433e-06,
+      "loss": 0.3588,
+      "step": 5258
+    },
+    {
+      "epoch": 0.43147228945317306,
+      "grad_norm": 0.6666251142393067,
+      "learning_rate": 6.337465102129279e-06,
+      "loss": 0.322,
+      "step": 5259
+    },
+    {
+      "epoch": 0.43155433400336385,
+      "grad_norm": 0.732614064284012,
+      "learning_rate": 6.336184767158488e-06,
+      "loss": 0.3807,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4316363785535546,
+      "grad_norm": 0.6070683078518448,
+      "learning_rate": 6.334904337828477e-06,
+      "loss": 0.3334,
+      "step": 5261
+    },
+    {
+      "epoch": 0.4317184231037453,
+      "grad_norm": 0.6579420488954592,
+      "learning_rate": 6.3336238142296645e-06,
+      "loss": 0.3647,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4318004676539361,
+      "grad_norm": 0.6768781855207855,
+      "learning_rate": 6.3323431964524815e-06,
+      "loss": 0.3433,
+      "step": 5263
+    },
+    {
+      "epoch": 0.4318825122041268,
+      "grad_norm": 0.6514030112478398,
+      "learning_rate": 6.331062484587362e-06,
+      "loss": 0.3189,
+      "step": 5264
+    },
+    {
+      "epoch": 0.4319645567543176,
+      "grad_norm": 0.6781344724585752,
+      "learning_rate": 6.329781678724752e-06,
+      "loss": 0.3528,
+      "step": 5265
+    },
+    {
+      "epoch": 0.43204660130450834,
+      "grad_norm": 0.6584637535953564,
+      "learning_rate": 6.328500778955091e-06,
+      "loss": 0.3363,
+      "step": 5266
+    },
+    {
+      "epoch": 0.4321286458546991,
+      "grad_norm": 0.7171365596956452,
+      "learning_rate": 6.3272197853688425e-06,
+      "loss": 0.3715,
+      "step": 5267
+    },
+    {
+      "epoch": 0.43221069040488985,
+      "grad_norm": 0.6402552462497081,
+      "learning_rate": 6.325938698056464e-06,
+      "loss": 0.3302,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4322927349550806,
+      "grad_norm": 0.616058905299238,
+      "learning_rate": 6.324657517108426e-06,
+      "loss": 0.3178,
+      "step": 5269
+    },
+    {
+      "epoch": 0.43237477950527137,
+      "grad_norm": 0.574571618644265,
+      "learning_rate": 6.323376242615201e-06,
+      "loss": 0.2976,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4324568240554621,
+      "grad_norm": 0.6546633250482502,
+      "learning_rate": 6.322094874667273e-06,
+      "loss": 0.302,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4325388686056529,
+      "grad_norm": 0.6229092716736692,
+      "learning_rate": 6.320813413355129e-06,
+      "loss": 0.3333,
+      "step": 5272
+    },
+    {
+      "epoch": 0.4326209131558436,
+      "grad_norm": 0.6908233474323751,
+      "learning_rate": 6.319531858769263e-06,
+      "loss": 0.368,
+      "step": 5273
+    },
+    {
+      "epoch": 0.4327029577060344,
+      "grad_norm": 0.6268276713180859,
+      "learning_rate": 6.318250211000179e-06,
+      "loss": 0.3384,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4327850022562251,
+      "grad_norm": 0.7456314594038563,
+      "learning_rate": 6.316968470138379e-06,
+      "loss": 0.3693,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43286704680641586,
+      "grad_norm": 0.6370785744866168,
+      "learning_rate": 6.315686636274385e-06,
+      "loss": 0.3489,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43294909135660664,
+      "grad_norm": 0.6631063861949101,
+      "learning_rate": 6.314404709498714e-06,
+      "loss": 0.319,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43303113590679737,
+      "grad_norm": 0.6447495565504858,
+      "learning_rate": 6.313122689901894e-06,
+      "loss": 0.3217,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43311318045698816,
+      "grad_norm": 0.6555301576265714,
+      "learning_rate": 6.31184057757446e-06,
+      "loss": 0.3062,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4331952250071789,
+      "grad_norm": 0.6250268482552348,
+      "learning_rate": 6.3105583726069495e-06,
+      "loss": 0.3201,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43327726955736967,
+      "grad_norm": 0.6099854655690351,
+      "learning_rate": 6.309276075089914e-06,
+      "loss": 0.3413,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4333593141075604,
+      "grad_norm": 0.7599188757773451,
+      "learning_rate": 6.307993685113907e-06,
+      "loss": 0.3151,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4334413586577512,
+      "grad_norm": 0.6068302141474277,
+      "learning_rate": 6.306711202769485e-06,
+      "loss": 0.3119,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4335234032079419,
+      "grad_norm": 0.6377773356715855,
+      "learning_rate": 6.305428628147218e-06,
+      "loss": 0.3686,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43360544775813264,
+      "grad_norm": 0.655997009829089,
+      "learning_rate": 6.3041459613376795e-06,
+      "loss": 0.3139,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43368749230832343,
+      "grad_norm": 0.7499527684634661,
+      "learning_rate": 6.3028632024314485e-06,
+      "loss": 0.329,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43376953685851416,
+      "grad_norm": 0.621286575168743,
+      "learning_rate": 6.3015803515191096e-06,
+      "loss": 0.34,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43385158140870494,
+      "grad_norm": 0.7517773430299831,
+      "learning_rate": 6.30029740869126e-06,
+      "loss": 0.3451,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4339336259588957,
+      "grad_norm": 0.6129273578187738,
+      "learning_rate": 6.299014374038493e-06,
+      "loss": 0.3163,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43401567050908646,
+      "grad_norm": 0.6470323353587519,
+      "learning_rate": 6.297731247651419e-06,
+      "loss": 0.3384,
+      "step": 5290
+    },
+    {
+      "epoch": 0.4340977150592772,
+      "grad_norm": 0.6573035819230301,
+      "learning_rate": 6.296448029620652e-06,
+      "loss": 0.3352,
+      "step": 5291
+    },
+    {
+      "epoch": 0.4341797596094679,
+      "grad_norm": 0.6727398589516524,
+      "learning_rate": 6.295164720036806e-06,
+      "loss": 0.3438,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4342618041596587,
+      "grad_norm": 0.655232979788338,
+      "learning_rate": 6.293881318990509e-06,
+      "loss": 0.358,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43434384870984943,
+      "grad_norm": 0.6838524687406651,
+      "learning_rate": 6.292597826572391e-06,
+      "loss": 0.3627,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4344258932600402,
+      "grad_norm": 0.6519562107258645,
+      "learning_rate": 6.291314242873092e-06,
+      "loss": 0.36,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43450793781023095,
+      "grad_norm": 0.6901950835938988,
+      "learning_rate": 6.2900305679832565e-06,
+      "loss": 0.3533,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43458998236042173,
+      "grad_norm": 0.6102133299352966,
+      "learning_rate": 6.288746801993533e-06,
+      "loss": 0.3305,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43467202691061246,
+      "grad_norm": 0.6412059354167786,
+      "learning_rate": 6.287462944994582e-06,
+      "loss": 0.3245,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4347540714608032,
+      "grad_norm": 0.6150182740332532,
+      "learning_rate": 6.286178997077066e-06,
+      "loss": 0.342,
+      "step": 5299
+    },
+    {
+      "epoch": 0.434836116010994,
+      "grad_norm": 0.6251794313137558,
+      "learning_rate": 6.284894958331657e-06,
+      "loss": 0.3341,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4349181605611847,
+      "grad_norm": 0.5997117985570801,
+      "learning_rate": 6.283610828849028e-06,
+      "loss": 0.3193,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4350002051113755,
+      "grad_norm": 0.6738481040632917,
+      "learning_rate": 6.282326608719865e-06,
+      "loss": 0.3607,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4350822496615662,
+      "grad_norm": 0.714878430444649,
+      "learning_rate": 6.281042298034858e-06,
+      "loss": 0.3524,
+      "step": 5303
+    },
+    {
+      "epoch": 0.435164294211757,
+      "grad_norm": 0.6673566456773349,
+      "learning_rate": 6.2797578968847015e-06,
+      "loss": 0.3304,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43524633876194774,
+      "grad_norm": 0.6948467177238463,
+      "learning_rate": 6.278473405360099e-06,
+      "loss": 0.3837,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43532838331213847,
+      "grad_norm": 0.6676519832474107,
+      "learning_rate": 6.2771888235517586e-06,
+      "loss": 0.3641,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43541042786232925,
+      "grad_norm": 0.5897993247575718,
+      "learning_rate": 6.275904151550396e-06,
+      "loss": 0.3239,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43549247241252,
+      "grad_norm": 0.6379250598981987,
+      "learning_rate": 6.274619389446731e-06,
+      "loss": 0.3515,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43557451696271077,
+      "grad_norm": 0.6398214982428917,
+      "learning_rate": 6.273334537331494e-06,
+      "loss": 0.3436,
+      "step": 5309
+    },
+    {
+      "epoch": 0.4356565615129015,
+      "grad_norm": 0.7026309962756463,
+      "learning_rate": 6.272049595295416e-06,
+      "loss": 0.3174,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4357386060630923,
+      "grad_norm": 0.6775524466114633,
+      "learning_rate": 6.2707645634292395e-06,
+      "loss": 0.3017,
+      "step": 5311
+    },
+    {
+      "epoch": 0.435820650613283,
+      "grad_norm": 0.6719736914165287,
+      "learning_rate": 6.269479441823712e-06,
+      "loss": 0.3588,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43590269516347374,
+      "grad_norm": 0.6371205173884819,
+      "learning_rate": 6.2681942305695865e-06,
+      "loss": 0.313,
+      "step": 5313
+    },
+    {
+      "epoch": 0.4359847397136645,
+      "grad_norm": 0.6553271533414169,
+      "learning_rate": 6.26690892975762e-06,
+      "loss": 0.3552,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43606678426385526,
+      "grad_norm": 0.6545600423530018,
+      "learning_rate": 6.265623539478583e-06,
+      "loss": 0.3326,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43614882881404604,
+      "grad_norm": 0.6598676411412718,
+      "learning_rate": 6.264338059823242e-06,
+      "loss": 0.3481,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43623087336423677,
+      "grad_norm": 0.5962628065136057,
+      "learning_rate": 6.26305249088238e-06,
+      "loss": 0.3289,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43631291791442756,
+      "grad_norm": 0.6710013027884152,
+      "learning_rate": 6.26176683274678e-06,
+      "loss": 0.3662,
+      "step": 5318
+    },
+    {
+      "epoch": 0.4363949624646183,
+      "grad_norm": 0.6034649334824211,
+      "learning_rate": 6.2604810855072325e-06,
+      "loss": 0.314,
+      "step": 5319
+    },
+    {
+      "epoch": 0.436477007014809,
+      "grad_norm": 0.6717090793088603,
+      "learning_rate": 6.259195249254539e-06,
+      "loss": 0.3052,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4365590515649998,
+      "grad_norm": 0.6637976030011434,
+      "learning_rate": 6.257909324079498e-06,
+      "loss": 0.3569,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43664109611519053,
+      "grad_norm": 0.6703190297065894,
+      "learning_rate": 6.25662331007292e-06,
+      "loss": 0.3665,
+      "step": 5322
+    },
+    {
+      "epoch": 0.4367231406653813,
+      "grad_norm": 0.6521997040013581,
+      "learning_rate": 6.255337207325625e-06,
+      "loss": 0.3779,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43680518521557204,
+      "grad_norm": 0.6736351670757265,
+      "learning_rate": 6.254051015928433e-06,
+      "loss": 0.3324,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43688722976576283,
+      "grad_norm": 0.6614256394923991,
+      "learning_rate": 6.252764735972172e-06,
+      "loss": 0.3336,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43696927431595356,
+      "grad_norm": 0.6690544532802983,
+      "learning_rate": 6.251478367547679e-06,
+      "loss": 0.3253,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4370513188661443,
+      "grad_norm": 0.6272554073108739,
+      "learning_rate": 6.250191910745794e-06,
+      "loss": 0.2928,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4371333634163351,
+      "grad_norm": 0.7031796625952058,
+      "learning_rate": 6.248905365657365e-06,
+      "loss": 0.3073,
+      "step": 5328
+    },
+    {
+      "epoch": 0.4372154079665258,
+      "grad_norm": 0.6672313095533291,
+      "learning_rate": 6.247618732373245e-06,
+      "loss": 0.3561,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4372974525167166,
+      "grad_norm": 0.6558791794554417,
+      "learning_rate": 6.246332010984295e-06,
+      "loss": 0.3488,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4373794970669073,
+      "grad_norm": 0.6361873991588707,
+      "learning_rate": 6.245045201581381e-06,
+      "loss": 0.3661,
+      "step": 5331
+    },
+    {
+      "epoch": 0.4374615416170981,
+      "grad_norm": 0.7005031721817526,
+      "learning_rate": 6.243758304255375e-06,
+      "loss": 0.3361,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43754358616728883,
+      "grad_norm": 0.6000547211030636,
+      "learning_rate": 6.242471319097156e-06,
+      "loss": 0.3223,
+      "step": 5333
+    },
+    {
+      "epoch": 0.4376256307174796,
+      "grad_norm": 0.6240332426985268,
+      "learning_rate": 6.241184246197609e-06,
+      "loss": 0.327,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43770767526767035,
+      "grad_norm": 0.6993525984398733,
+      "learning_rate": 6.239897085647624e-06,
+      "loss": 0.3261,
+      "step": 5335
+    },
+    {
+      "epoch": 0.4377897198178611,
+      "grad_norm": 0.7350183782995722,
+      "learning_rate": 6.238609837538099e-06,
+      "loss": 0.3707,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43787176436805186,
+      "grad_norm": 0.7592780315708043,
+      "learning_rate": 6.237322501959937e-06,
+      "loss": 0.3561,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4379538089182426,
+      "grad_norm": 0.6872676705366995,
+      "learning_rate": 6.236035079004047e-06,
+      "loss": 0.3356,
+      "step": 5338
+    },
+    {
+      "epoch": 0.4380358534684334,
+      "grad_norm": 0.6475948826660268,
+      "learning_rate": 6.234747568761347e-06,
+      "loss": 0.3419,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4381178980186241,
+      "grad_norm": 0.6508953879648159,
+      "learning_rate": 6.233459971322757e-06,
+      "loss": 0.3427,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4381999425688149,
+      "grad_norm": 0.6594518732305046,
+      "learning_rate": 6.232172286779207e-06,
+      "loss": 0.3476,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4382819871190056,
+      "grad_norm": 0.6715762156548872,
+      "learning_rate": 6.230884515221628e-06,
+      "loss": 0.3932,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43836403166919635,
+      "grad_norm": 0.5937519173037982,
+      "learning_rate": 6.229596656740962e-06,
+      "loss": 0.3401,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43844607621938714,
+      "grad_norm": 0.6036184991162397,
+      "learning_rate": 6.228308711428156e-06,
+      "loss": 0.3442,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43852812076957787,
+      "grad_norm": 0.634207534747967,
+      "learning_rate": 6.227020679374162e-06,
+      "loss": 0.3576,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43861016531976865,
+      "grad_norm": 0.70698607066324,
+      "learning_rate": 6.2257325606699394e-06,
+      "loss": 0.351,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4386922098699594,
+      "grad_norm": 0.6321356520616177,
+      "learning_rate": 6.224444355406452e-06,
+      "loss": 0.3271,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43877425442015017,
+      "grad_norm": 0.6668518541302388,
+      "learning_rate": 6.223156063674672e-06,
+      "loss": 0.3268,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4388562989703409,
+      "grad_norm": 0.6513172028904911,
+      "learning_rate": 6.221867685565577e-06,
+      "loss": 0.3382,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4389383435205316,
+      "grad_norm": 0.6554140386226012,
+      "learning_rate": 6.220579221170148e-06,
+      "loss": 0.6937,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4390203880707224,
+      "grad_norm": 0.6285572082764198,
+      "learning_rate": 6.219290670579376e-06,
+      "loss": 0.3069,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43910243262091314,
+      "grad_norm": 0.6492774289287315,
+      "learning_rate": 6.218002033884254e-06,
+      "loss": 0.3218,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4391844771711039,
+      "grad_norm": 0.6931704557852036,
+      "learning_rate": 6.216713311175786e-06,
+      "loss": 0.331,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43926652172129466,
+      "grad_norm": 0.6532870118840011,
+      "learning_rate": 6.21542450254498e-06,
+      "loss": 0.3133,
+      "step": 5354
+    },
+    {
+      "epoch": 0.43934856627148544,
+      "grad_norm": 0.6782259028936538,
+      "learning_rate": 6.214135608082848e-06,
+      "loss": 0.3723,
+      "step": 5355
+    },
+    {
+      "epoch": 0.43943061082167617,
+      "grad_norm": 0.6921848804338662,
+      "learning_rate": 6.212846627880409e-06,
+      "loss": 0.3301,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4395126553718669,
+      "grad_norm": 0.7392834677133558,
+      "learning_rate": 6.211557562028691e-06,
+      "loss": 0.3313,
+      "step": 5357
+    },
+    {
+      "epoch": 0.4395946999220577,
+      "grad_norm": 0.6345976340888854,
+      "learning_rate": 6.210268410618723e-06,
+      "loss": 0.3286,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4396767444722484,
+      "grad_norm": 0.8178809391581554,
+      "learning_rate": 6.208979173741545e-06,
+      "loss": 0.3254,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4397587890224392,
+      "grad_norm": 0.611873512663777,
+      "learning_rate": 6.2076898514882e-06,
+      "loss": 0.3434,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43984083357262993,
+      "grad_norm": 0.6874242602483885,
+      "learning_rate": 6.206400443949739e-06,
+      "loss": 0.3747,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4399228781228207,
+      "grad_norm": 0.6396422364427905,
+      "learning_rate": 6.205110951217216e-06,
+      "loss": 0.3635,
+      "step": 5362
+    },
+    {
+      "epoch": 0.44000492267301144,
+      "grad_norm": 0.6430157115814371,
+      "learning_rate": 6.2038213733816935e-06,
+      "loss": 0.3339,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4400869672232022,
+      "grad_norm": 0.6083142396681176,
+      "learning_rate": 6.202531710534241e-06,
+      "loss": 0.3469,
+      "step": 5364
+    },
+    {
+      "epoch": 0.44016901177339296,
+      "grad_norm": 0.67922160678454,
+      "learning_rate": 6.201241962765928e-06,
+      "loss": 0.3578,
+      "step": 5365
+    },
+    {
+      "epoch": 0.4402510563235837,
+      "grad_norm": 0.6564507068943601,
+      "learning_rate": 6.19995213016784e-06,
+      "loss": 0.3451,
+      "step": 5366
+    },
+    {
+      "epoch": 0.4403331008737745,
+      "grad_norm": 0.6100774607205361,
+      "learning_rate": 6.198662212831061e-06,
+      "loss": 0.3225,
+      "step": 5367
+    },
+    {
+      "epoch": 0.4404151454239652,
+      "grad_norm": 0.5742598233271644,
+      "learning_rate": 6.197372210846681e-06,
+      "loss": 0.3409,
+      "step": 5368
+    },
+    {
+      "epoch": 0.440497189974156,
+      "grad_norm": 0.7064212911105872,
+      "learning_rate": 6.196082124305801e-06,
+      "loss": 0.3735,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4405792345243467,
+      "grad_norm": 0.6742039074284577,
+      "learning_rate": 6.1947919532995205e-06,
+      "loss": 0.3386,
+      "step": 5370
+    },
+    {
+      "epoch": 0.44066127907453745,
+      "grad_norm": 0.6566202615542254,
+      "learning_rate": 6.1935016979189534e-06,
+      "loss": 0.3543,
+      "step": 5371
+    },
+    {
+      "epoch": 0.44074332362472823,
+      "grad_norm": 0.675050608112663,
+      "learning_rate": 6.192211358255213e-06,
+      "loss": 0.3411,
+      "step": 5372
+    },
+    {
+      "epoch": 0.44082536817491896,
+      "grad_norm": 0.6442812659564909,
+      "learning_rate": 6.190920934399423e-06,
+      "loss": 0.336,
+      "step": 5373
+    },
+    {
+      "epoch": 0.44090741272510975,
+      "grad_norm": 0.7323866584048604,
+      "learning_rate": 6.189630426442709e-06,
+      "loss": 0.3703,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4409894572753005,
+      "grad_norm": 0.6389016759967228,
+      "learning_rate": 6.188339834476207e-06,
+      "loss": 0.3361,
+      "step": 5375
+    },
+    {
+      "epoch": 0.44107150182549126,
+      "grad_norm": 0.6495036343423697,
+      "learning_rate": 6.1870491585910545e-06,
+      "loss": 0.3378,
+      "step": 5376
+    },
+    {
+      "epoch": 0.441153546375682,
+      "grad_norm": 0.6938208198054447,
+      "learning_rate": 6.185758398878396e-06,
+      "loss": 0.3379,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4412355909258727,
+      "grad_norm": 0.6063668198884578,
+      "learning_rate": 6.184467555429386e-06,
+      "loss": 0.3179,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4413176354760635,
+      "grad_norm": 0.6661746800269878,
+      "learning_rate": 6.183176628335179e-06,
+      "loss": 0.3554,
+      "step": 5379
+    },
+    {
+      "epoch": 0.44139968002625424,
+      "grad_norm": 0.6352408718499332,
+      "learning_rate": 6.18188561768694e-06,
+      "loss": 0.3389,
+      "step": 5380
+    },
+    {
+      "epoch": 0.441481724576445,
+      "grad_norm": 0.6357264294956044,
+      "learning_rate": 6.180594523575838e-06,
+      "loss": 0.3384,
+      "step": 5381
+    },
+    {
+      "epoch": 0.44156376912663575,
+      "grad_norm": 0.6396849797399508,
+      "learning_rate": 6.179303346093048e-06,
+      "loss": 0.3256,
+      "step": 5382
+    },
+    {
+      "epoch": 0.44164581367682654,
+      "grad_norm": 0.6211383089939199,
+      "learning_rate": 6.17801208532975e-06,
+      "loss": 0.3614,
+      "step": 5383
+    },
+    {
+      "epoch": 0.44172785822701727,
+      "grad_norm": 0.7194091869629602,
+      "learning_rate": 6.17672074137713e-06,
+      "loss": 0.343,
+      "step": 5384
+    },
+    {
+      "epoch": 0.441809902777208,
+      "grad_norm": 0.6548287812933139,
+      "learning_rate": 6.175429314326383e-06,
+      "loss": 0.3445,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4418919473273988,
+      "grad_norm": 0.729251482815104,
+      "learning_rate": 6.174137804268706e-06,
+      "loss": 0.3635,
+      "step": 5386
+    },
+    {
+      "epoch": 0.4419739918775895,
+      "grad_norm": 0.6912858075840108,
+      "learning_rate": 6.172846211295303e-06,
+      "loss": 0.3328,
+      "step": 5387
+    },
+    {
+      "epoch": 0.4420560364277803,
+      "grad_norm": 0.6744932942351691,
+      "learning_rate": 6.171554535497385e-06,
+      "loss": 0.3629,
+      "step": 5388
+    },
+    {
+      "epoch": 0.442138080977971,
+      "grad_norm": 0.6744976915004357,
+      "learning_rate": 6.1702627769661695e-06,
+      "loss": 0.3569,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4422201255281618,
+      "grad_norm": 0.6441835271963754,
+      "learning_rate": 6.168970935792876e-06,
+      "loss": 0.3328,
+      "step": 5390
+    },
+    {
+      "epoch": 0.44230217007835254,
+      "grad_norm": 0.62192963929581,
+      "learning_rate": 6.167679012068732e-06,
+      "loss": 0.3183,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4423842146285433,
+      "grad_norm": 0.5923058733546206,
+      "learning_rate": 6.1663870058849725e-06,
+      "loss": 0.3238,
+      "step": 5392
+    },
+    {
+      "epoch": 0.44246625917873406,
+      "grad_norm": 0.6994600127018021,
+      "learning_rate": 6.165094917332838e-06,
+      "loss": 0.3218,
+      "step": 5393
+    },
+    {
+      "epoch": 0.4425483037289248,
+      "grad_norm": 0.6849920904922439,
+      "learning_rate": 6.163802746503571e-06,
+      "loss": 0.3566,
+      "step": 5394
+    },
+    {
+      "epoch": 0.44263034827911557,
+      "grad_norm": 0.6742896100282577,
+      "learning_rate": 6.162510493488425e-06,
+      "loss": 0.338,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4427123928293063,
+      "grad_norm": 0.654505191935681,
+      "learning_rate": 6.161218158378655e-06,
+      "loss": 0.6943,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4427944373794971,
+      "grad_norm": 0.6302788695247225,
+      "learning_rate": 6.159925741265525e-06,
+      "loss": 0.3174,
+      "step": 5397
+    },
+    {
+      "epoch": 0.4428764819296878,
+      "grad_norm": 0.6445960262525062,
+      "learning_rate": 6.158633242240301e-06,
+      "loss": 0.3397,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4429585264798786,
+      "grad_norm": 0.6230789221401076,
+      "learning_rate": 6.15734066139426e-06,
+      "loss": 0.3253,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44304057103006933,
+      "grad_norm": 0.6254752156079448,
+      "learning_rate": 6.156047998818681e-06,
+      "loss": 0.3251,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44312261558026006,
+      "grad_norm": 0.6310018888023665,
+      "learning_rate": 6.154755254604848e-06,
+      "loss": 0.3625,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44320466013045084,
+      "grad_norm": 0.6421411667108851,
+      "learning_rate": 6.153462428844056e-06,
+      "loss": 0.3492,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4432867046806416,
+      "grad_norm": 0.6639064492878384,
+      "learning_rate": 6.1521695216276015e-06,
+      "loss": 0.3727,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44336874923083236,
+      "grad_norm": 0.6873662441543628,
+      "learning_rate": 6.150876533046784e-06,
+      "loss": 0.3447,
+      "step": 5404
+    },
+    {
+      "epoch": 0.4434507937810231,
+      "grad_norm": 0.6619976750739308,
+      "learning_rate": 6.149583463192916e-06,
+      "loss": 0.369,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4435328383312139,
+      "grad_norm": 0.6833161685079882,
+      "learning_rate": 6.14829031215731e-06,
+      "loss": 0.3469,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4436148828814046,
+      "grad_norm": 0.8072335641257796,
+      "learning_rate": 6.146997080031289e-06,
+      "loss": 0.355,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44369692743159533,
+      "grad_norm": 0.6771199350216374,
+      "learning_rate": 6.145703766906175e-06,
+      "loss": 0.3181,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4437789719817861,
+      "grad_norm": 0.6122944274195639,
+      "learning_rate": 6.144410372873302e-06,
+      "loss": 0.3259,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44386101653197685,
+      "grad_norm": 0.6369071005966971,
+      "learning_rate": 6.143116898024007e-06,
+      "loss": 0.3418,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44394306108216763,
+      "grad_norm": 0.7273487089284443,
+      "learning_rate": 6.141823342449634e-06,
+      "loss": 0.3086,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44402510563235836,
+      "grad_norm": 0.6331981987262111,
+      "learning_rate": 6.140529706241531e-06,
+      "loss": 0.3596,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44410715018254915,
+      "grad_norm": 0.6697665422086079,
+      "learning_rate": 6.1392359894910526e-06,
+      "loss": 0.3482,
+      "step": 5413
+    },
+    {
+      "epoch": 0.4441891947327399,
+      "grad_norm": 0.6560347256514527,
+      "learning_rate": 6.137942192289559e-06,
+      "loss": 0.3457,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4442712392829306,
+      "grad_norm": 0.6824475554018098,
+      "learning_rate": 6.136648314728414e-06,
+      "loss": 0.3357,
+      "step": 5415
+    },
+    {
+      "epoch": 0.4443532838331214,
+      "grad_norm": 0.7012592175959917,
+      "learning_rate": 6.135354356898992e-06,
+      "loss": 0.319,
+      "step": 5416
+    },
+    {
+      "epoch": 0.4444353283833121,
+      "grad_norm": 0.5875426265377571,
+      "learning_rate": 6.134060318892672e-06,
+      "loss": 0.3341,
+      "step": 5417
+    },
+    {
+      "epoch": 0.4445173729335029,
+      "grad_norm": 0.649209060668112,
+      "learning_rate": 6.1327662008008326e-06,
+      "loss": 0.3381,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44459941748369364,
+      "grad_norm": 0.666071561830033,
+      "learning_rate": 6.131472002714863e-06,
+      "loss": 0.3455,
+      "step": 5419
+    },
+    {
+      "epoch": 0.4446814620338844,
+      "grad_norm": 0.6720492815756898,
+      "learning_rate": 6.13017772472616e-06,
+      "loss": 0.3122,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44476350658407515,
+      "grad_norm": 0.6318987706302928,
+      "learning_rate": 6.12888336692612e-06,
+      "loss": 0.3542,
+      "step": 5421
+    },
+    {
+      "epoch": 0.4448455511342659,
+      "grad_norm": 0.6599169190255967,
+      "learning_rate": 6.127588929406152e-06,
+      "loss": 0.3127,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44492759568445667,
+      "grad_norm": 1.0434883365326493,
+      "learning_rate": 6.1262944122576655e-06,
+      "loss": 0.3431,
+      "step": 5423
+    },
+    {
+      "epoch": 0.4450096402346474,
+      "grad_norm": 0.6541377258581761,
+      "learning_rate": 6.124999815572077e-06,
+      "loss": 0.3456,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4450916847848382,
+      "grad_norm": 0.6178473486251103,
+      "learning_rate": 6.123705139440807e-06,
+      "loss": 0.325,
+      "step": 5425
+    },
+    {
+      "epoch": 0.4451737293350289,
+      "grad_norm": 0.6498704162262888,
+      "learning_rate": 6.122410383955288e-06,
+      "loss": 0.3356,
+      "step": 5426
+    },
+    {
+      "epoch": 0.4452557738852197,
+      "grad_norm": 0.639599276148696,
+      "learning_rate": 6.12111554920695e-06,
+      "loss": 0.2974,
+      "step": 5427
+    },
+    {
+      "epoch": 0.4453378184354104,
+      "grad_norm": 0.6102688000203615,
+      "learning_rate": 6.119820635287231e-06,
+      "loss": 0.3546,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44541986298560116,
+      "grad_norm": 0.6216565122007832,
+      "learning_rate": 6.118525642287581e-06,
+      "loss": 0.3726,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44550190753579194,
+      "grad_norm": 0.6597454353936569,
+      "learning_rate": 6.117230570299446e-06,
+      "loss": 0.3145,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44558395208598267,
+      "grad_norm": 0.6212861501415113,
+      "learning_rate": 6.115935419414284e-06,
+      "loss": 0.2929,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44566599663617346,
+      "grad_norm": 0.6250602318848004,
+      "learning_rate": 6.114640189723554e-06,
+      "loss": 0.3031,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4457480411863642,
+      "grad_norm": 0.6612160527293325,
+      "learning_rate": 6.113344881318726e-06,
+      "loss": 0.3666,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44583008573655497,
+      "grad_norm": 0.6437606733392837,
+      "learning_rate": 6.11204949429127e-06,
+      "loss": 0.3936,
+      "step": 5434
+    },
+    {
+      "epoch": 0.4459121302867457,
+      "grad_norm": 0.6634903613112948,
+      "learning_rate": 6.110754028732666e-06,
+      "loss": 0.3737,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44599417483693643,
+      "grad_norm": 0.613275901211568,
+      "learning_rate": 6.109458484734397e-06,
+      "loss": 0.3448,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4460762193871272,
+      "grad_norm": 0.6585234191371513,
+      "learning_rate": 6.108162862387952e-06,
+      "loss": 0.353,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44615826393731794,
+      "grad_norm": 0.6988412191583517,
+      "learning_rate": 6.1068671617848266e-06,
+      "loss": 0.3503,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44624030848750873,
+      "grad_norm": 0.647452712607638,
+      "learning_rate": 6.105571383016521e-06,
+      "loss": 0.3477,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44632235303769946,
+      "grad_norm": 0.7047629677777232,
+      "learning_rate": 6.104275526174539e-06,
+      "loss": 0.334,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44640439758789024,
+      "grad_norm": 0.6175094214760531,
+      "learning_rate": 6.102979591350393e-06,
+      "loss": 0.3157,
+      "step": 5441
+    },
+    {
+      "epoch": 0.446486442138081,
+      "grad_norm": 0.6316573000828313,
+      "learning_rate": 6.101683578635602e-06,
+      "loss": 0.3212,
+      "step": 5442
+    },
+    {
+      "epoch": 0.4465684866882717,
+      "grad_norm": 0.6276254095547698,
+      "learning_rate": 6.100387488121686e-06,
+      "loss": 0.3241,
+      "step": 5443
+    },
+    {
+      "epoch": 0.4466505312384625,
+      "grad_norm": 0.6532827843811483,
+      "learning_rate": 6.099091319900174e-06,
+      "loss": 0.3488,
+      "step": 5444
+    },
+    {
+      "epoch": 0.4467325757886532,
+      "grad_norm": 0.6597437889245041,
+      "learning_rate": 6.0977950740626e-06,
+      "loss": 0.3516,
+      "step": 5445
+    },
+    {
+      "epoch": 0.446814620338844,
+      "grad_norm": 0.673465416955106,
+      "learning_rate": 6.0964987507004995e-06,
+      "loss": 0.3677,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44689666488903473,
+      "grad_norm": 0.6929750532437164,
+      "learning_rate": 6.09520234990542e-06,
+      "loss": 0.328,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4469787094392255,
+      "grad_norm": 0.6602300092819677,
+      "learning_rate": 6.09390587176891e-06,
+      "loss": 0.3201,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44706075398941625,
+      "grad_norm": 0.6363767547081828,
+      "learning_rate": 6.092609316382524e-06,
+      "loss": 0.3889,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44714279853960703,
+      "grad_norm": 0.6799568811921192,
+      "learning_rate": 6.091312683837823e-06,
+      "loss": 0.3335,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44722484308979776,
+      "grad_norm": 0.6139865195769476,
+      "learning_rate": 6.090015974226376e-06,
+      "loss": 0.3101,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4473068876399885,
+      "grad_norm": 0.7330366935895841,
+      "learning_rate": 6.088719187639751e-06,
+      "loss": 0.3574,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4473889321901793,
+      "grad_norm": 0.7326798797040889,
+      "learning_rate": 6.0874223241695265e-06,
+      "loss": 0.3307,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44747097674037,
+      "grad_norm": 0.6421716446731747,
+      "learning_rate": 6.0861253839072844e-06,
+      "loss": 0.4055,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4475530212905608,
+      "grad_norm": 0.6470539648463657,
+      "learning_rate": 6.084828366944611e-06,
+      "loss": 0.3494,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4476350658407515,
+      "grad_norm": 0.6684281477838185,
+      "learning_rate": 6.0835312733731025e-06,
+      "loss": 0.3372,
+      "step": 5456
+    },
+    {
+      "epoch": 0.4477171103909423,
+      "grad_norm": 0.687067021851575,
+      "learning_rate": 6.0822341032843554e-06,
+      "loss": 0.3504,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44779915494113304,
+      "grad_norm": 0.7155176179976332,
+      "learning_rate": 6.080936856769977e-06,
+      "loss": 0.3251,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44788119949132377,
+      "grad_norm": 0.6532729819069926,
+      "learning_rate": 6.079639533921573e-06,
+      "loss": 0.3273,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44796324404151455,
+      "grad_norm": 0.6634569288662667,
+      "learning_rate": 6.078342134830759e-06,
+      "loss": 0.339,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4480452885917053,
+      "grad_norm": 0.6400452476836332,
+      "learning_rate": 6.077044659589157e-06,
+      "loss": 0.3306,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44812733314189607,
+      "grad_norm": 0.6785973744575335,
+      "learning_rate": 6.075747108288391e-06,
+      "loss": 0.3254,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4482093776920868,
+      "grad_norm": 0.6695437828303891,
+      "learning_rate": 6.0744494810200925e-06,
+      "loss": 0.3627,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4482914222422776,
+      "grad_norm": 0.6005774429460095,
+      "learning_rate": 6.073151777875898e-06,
+      "loss": 0.3386,
+      "step": 5464
+    },
+    {
+      "epoch": 0.4483734667924683,
+      "grad_norm": 0.6887394624298431,
+      "learning_rate": 6.07185399894745e-06,
+      "loss": 0.3207,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44845551134265904,
+      "grad_norm": 0.6782027417523246,
+      "learning_rate": 6.070556144326394e-06,
+      "loss": 0.3432,
+      "step": 5466
+    },
+    {
+      "epoch": 0.4485375558928498,
+      "grad_norm": 0.6624356347499887,
+      "learning_rate": 6.069258214104383e-06,
+      "loss": 0.3349,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44861960044304056,
+      "grad_norm": 0.6940690266691288,
+      "learning_rate": 6.067960208373074e-06,
+      "loss": 0.3596,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44870164499323134,
+      "grad_norm": 0.6894500474263449,
+      "learning_rate": 6.066662127224131e-06,
+      "loss": 0.3138,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44878368954342207,
+      "grad_norm": 0.6504563324575957,
+      "learning_rate": 6.065363970749223e-06,
+      "loss": 0.325,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44886573409361286,
+      "grad_norm": 0.740866555438256,
+      "learning_rate": 6.064065739040024e-06,
+      "loss": 0.3411,
+      "step": 5471
+    },
+    {
+      "epoch": 0.4489477786438036,
+      "grad_norm": 0.8863520606927088,
+      "learning_rate": 6.06276743218821e-06,
+      "loss": 0.3306,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4490298231939943,
+      "grad_norm": 0.6621580863688217,
+      "learning_rate": 6.061469050285469e-06,
+      "loss": 0.3427,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4491118677441851,
+      "grad_norm": 0.701148129491801,
+      "learning_rate": 6.060170593423488e-06,
+      "loss": 0.3613,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44919391229437583,
+      "grad_norm": 0.6384967786335821,
+      "learning_rate": 6.058872061693963e-06,
+      "loss": 0.3568,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4492759568445666,
+      "grad_norm": 0.6421604082407371,
+      "learning_rate": 6.0575734551885955e-06,
+      "loss": 0.3393,
+      "step": 5476
+    },
+    {
+      "epoch": 0.44935800139475734,
+      "grad_norm": 0.7392197567798597,
+      "learning_rate": 6.056274773999088e-06,
+      "loss": 0.3557,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44944004594494813,
+      "grad_norm": 0.6439159539738014,
+      "learning_rate": 6.0549760182171535e-06,
+      "loss": 0.3534,
+      "step": 5478
+    },
+    {
+      "epoch": 0.44952209049513886,
+      "grad_norm": 0.6426759723779124,
+      "learning_rate": 6.053677187934509e-06,
+      "loss": 0.3498,
+      "step": 5479
+    },
+    {
+      "epoch": 0.4496041350453296,
+      "grad_norm": 0.6179923165702758,
+      "learning_rate": 6.052378283242873e-06,
+      "loss": 0.3295,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4496861795955204,
+      "grad_norm": 0.6080740011818413,
+      "learning_rate": 6.051079304233974e-06,
+      "loss": 0.327,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4497682241457111,
+      "grad_norm": 0.6710371434961294,
+      "learning_rate": 6.0497802509995415e-06,
+      "loss": 0.3068,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4498502686959019,
+      "grad_norm": 0.7139480876138421,
+      "learning_rate": 6.0484811236313165e-06,
+      "loss": 0.328,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4499323132460926,
+      "grad_norm": 0.6533839053368865,
+      "learning_rate": 6.047181922221038e-06,
+      "loss": 0.289,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4500143577962834,
+      "grad_norm": 0.6357370025907507,
+      "learning_rate": 6.045882646860455e-06,
+      "loss": 0.3382,
+      "step": 5485
+    },
+    {
+      "epoch": 0.45009640234647413,
+      "grad_norm": 0.6510978201072027,
+      "learning_rate": 6.0445832976413195e-06,
+      "loss": 0.3609,
+      "step": 5486
+    },
+    {
+      "epoch": 0.45017844689666486,
+      "grad_norm": 0.6520876902952883,
+      "learning_rate": 6.04328387465539e-06,
+      "loss": 0.3342,
+      "step": 5487
+    },
+    {
+      "epoch": 0.45026049144685565,
+      "grad_norm": 0.6416852186222428,
+      "learning_rate": 6.04198437799443e-06,
+      "loss": 0.3222,
+      "step": 5488
+    },
+    {
+      "epoch": 0.4503425359970464,
+      "grad_norm": 0.6485373823844366,
+      "learning_rate": 6.040684807750207e-06,
+      "loss": 0.3155,
+      "step": 5489
+    },
+    {
+      "epoch": 0.45042458054723716,
+      "grad_norm": 0.6870121342363632,
+      "learning_rate": 6.039385164014494e-06,
+      "loss": 0.3424,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4505066250974279,
+      "grad_norm": 0.6473386424868797,
+      "learning_rate": 6.038085446879069e-06,
+      "loss": 0.2995,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4505886696476187,
+      "grad_norm": 0.6426299492137905,
+      "learning_rate": 6.0367856564357205e-06,
+      "loss": 0.3298,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4506707141978094,
+      "grad_norm": 0.664097347495453,
+      "learning_rate": 6.035485792776236e-06,
+      "loss": 0.3878,
+      "step": 5493
+    },
+    {
+      "epoch": 0.45075275874800014,
+      "grad_norm": 0.6483958455480544,
+      "learning_rate": 6.034185855992406e-06,
+      "loss": 0.3499,
+      "step": 5494
+    },
+    {
+      "epoch": 0.4508348032981909,
+      "grad_norm": 0.6619851940036816,
+      "learning_rate": 6.032885846176031e-06,
+      "loss": 0.3048,
+      "step": 5495
+    },
+    {
+      "epoch": 0.45091684784838165,
+      "grad_norm": 0.6051700979335092,
+      "learning_rate": 6.031585763418919e-06,
+      "loss": 0.3173,
+      "step": 5496
+    },
+    {
+      "epoch": 0.45099889239857244,
+      "grad_norm": 0.6241741447112886,
+      "learning_rate": 6.030285607812878e-06,
+      "loss": 0.3204,
+      "step": 5497
+    },
+    {
+      "epoch": 0.45108093694876317,
+      "grad_norm": 0.6544476841076127,
+      "learning_rate": 6.028985379449721e-06,
+      "loss": 0.3851,
+      "step": 5498
+    },
+    {
+      "epoch": 0.45116298149895395,
+      "grad_norm": 0.6947382906989665,
+      "learning_rate": 6.027685078421271e-06,
+      "loss": 0.3285,
+      "step": 5499
+    },
+    {
+      "epoch": 0.4512450260491447,
+      "grad_norm": 0.7023779343859339,
+      "learning_rate": 6.026384704819353e-06,
+      "loss": 0.3633,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4513270705993354,
+      "grad_norm": 0.6799815687256484,
+      "learning_rate": 6.025084258735794e-06,
+      "loss": 0.3175,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4514091151495262,
+      "grad_norm": 0.6775441842805292,
+      "learning_rate": 6.0237837402624325e-06,
+      "loss": 0.3391,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4514911596997169,
+      "grad_norm": 0.6366694024553469,
+      "learning_rate": 6.022483149491108e-06,
+      "loss": 0.3049,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4515732042499077,
+      "grad_norm": 0.7432419452890565,
+      "learning_rate": 6.021182486513666e-06,
+      "loss": 0.3402,
+      "step": 5504
+    },
+    {
+      "epoch": 0.45165524880009844,
+      "grad_norm": 0.6213128043872371,
+      "learning_rate": 6.019881751421956e-06,
+      "loss": 0.3601,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4517372933502892,
+      "grad_norm": 0.6488952127113141,
+      "learning_rate": 6.018580944307838e-06,
+      "loss": 0.316,
+      "step": 5506
+    },
+    {
+      "epoch": 0.45181933790047996,
+      "grad_norm": 0.7075582753752307,
+      "learning_rate": 6.0172800652631706e-06,
+      "loss": 0.3317,
+      "step": 5507
+    },
+    {
+      "epoch": 0.45190138245067074,
+      "grad_norm": 0.5957253564481122,
+      "learning_rate": 6.015979114379817e-06,
+      "loss": 0.321,
+      "step": 5508
+    },
+    {
+      "epoch": 0.45198342700086147,
+      "grad_norm": 0.6973502544827489,
+      "learning_rate": 6.014678091749652e-06,
+      "loss": 0.3421,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4520654715510522,
+      "grad_norm": 0.6462989343087144,
+      "learning_rate": 6.0133769974645495e-06,
+      "loss": 0.3239,
+      "step": 5510
+    },
+    {
+      "epoch": 0.452147516101243,
+      "grad_norm": 0.7080401988445989,
+      "learning_rate": 6.0120758316163915e-06,
+      "loss": 0.394,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4522295606514337,
+      "grad_norm": 0.6743565106731608,
+      "learning_rate": 6.0107745942970655e-06,
+      "loss": 0.3469,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4523116052016245,
+      "grad_norm": 0.658687660630437,
+      "learning_rate": 6.00947328559846e-06,
+      "loss": 0.3688,
+      "step": 5513
+    },
+    {
+      "epoch": 0.45239364975181523,
+      "grad_norm": 0.6555819766162051,
+      "learning_rate": 6.008171905612475e-06,
+      "loss": 0.3694,
+      "step": 5514
+    },
+    {
+      "epoch": 0.452475694302006,
+      "grad_norm": 0.6513326651265298,
+      "learning_rate": 6.0068704544310074e-06,
+      "loss": 0.3225,
+      "step": 5515
+    },
+    {
+      "epoch": 0.45255773885219674,
+      "grad_norm": 0.6541615310077519,
+      "learning_rate": 6.005568932145968e-06,
+      "loss": 0.3221,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4526397834023875,
+      "grad_norm": 0.9898334595807405,
+      "learning_rate": 6.004267338849266e-06,
+      "loss": 0.334,
+      "step": 5517
+    },
+    {
+      "epoch": 0.45272182795257826,
+      "grad_norm": 0.6413067132671195,
+      "learning_rate": 6.0029656746328155e-06,
+      "loss": 0.2889,
+      "step": 5518
+    },
+    {
+      "epoch": 0.452803872502769,
+      "grad_norm": 0.6633380460388496,
+      "learning_rate": 6.0016639395885424e-06,
+      "loss": 0.2928,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4528859170529598,
+      "grad_norm": 0.6124229370074592,
+      "learning_rate": 6.000362133808371e-06,
+      "loss": 0.3268,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4529679616031505,
+      "grad_norm": 0.6309807644197025,
+      "learning_rate": 5.999060257384234e-06,
+      "loss": 0.3388,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4530500061533413,
+      "grad_norm": 0.6166869416788537,
+      "learning_rate": 5.997758310408065e-06,
+      "loss": 0.3302,
+      "step": 5522
+    },
+    {
+      "epoch": 0.453132050703532,
+      "grad_norm": 0.651285078681108,
+      "learning_rate": 5.996456292971807e-06,
+      "loss": 0.3519,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45321409525372275,
+      "grad_norm": 0.6562100223669893,
+      "learning_rate": 5.995154205167408e-06,
+      "loss": 0.3842,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45329613980391353,
+      "grad_norm": 0.6484352067715783,
+      "learning_rate": 5.993852047086816e-06,
+      "loss": 0.3048,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45337818435410426,
+      "grad_norm": 0.6188144797394131,
+      "learning_rate": 5.992549818821991e-06,
+      "loss": 0.3338,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45346022890429505,
+      "grad_norm": 0.6319155617119969,
+      "learning_rate": 5.991247520464892e-06,
+      "loss": 0.3621,
+      "step": 5527
+    },
+    {
+      "epoch": 0.4535422734544858,
+      "grad_norm": 0.6383471081712955,
+      "learning_rate": 5.9899451521074854e-06,
+      "loss": 0.3174,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45362431800467656,
+      "grad_norm": 0.656940014252948,
+      "learning_rate": 5.988642713841743e-06,
+      "loss": 0.3361,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4537063625548673,
+      "grad_norm": 0.640893938208602,
+      "learning_rate": 5.987340205759641e-06,
+      "loss": 0.3479,
+      "step": 5530
+    },
+    {
+      "epoch": 0.453788407105058,
+      "grad_norm": 1.8795527842096598,
+      "learning_rate": 5.986037627953159e-06,
+      "loss": 0.3455,
+      "step": 5531
+    },
+    {
+      "epoch": 0.4538704516552488,
+      "grad_norm": 0.6461924422753108,
+      "learning_rate": 5.984734980514283e-06,
+      "loss": 0.3148,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45395249620543954,
+      "grad_norm": 0.6443675541267752,
+      "learning_rate": 5.983432263535006e-06,
+      "loss": 0.3656,
+      "step": 5533
+    },
+    {
+      "epoch": 0.4540345407556303,
+      "grad_norm": 0.637461552770478,
+      "learning_rate": 5.9821294771073224e-06,
+      "loss": 0.343,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45411658530582105,
+      "grad_norm": 0.6723983285669854,
+      "learning_rate": 5.980826621323235e-06,
+      "loss": 0.3589,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45419862985601184,
+      "grad_norm": 0.6321509849578475,
+      "learning_rate": 5.979523696274744e-06,
+      "loss": 0.3093,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45428067440620257,
+      "grad_norm": 0.6408188034387523,
+      "learning_rate": 5.978220702053865e-06,
+      "loss": 0.3054,
+      "step": 5537
+    },
+    {
+      "epoch": 0.4543627189563933,
+      "grad_norm": 0.9813101505324792,
+      "learning_rate": 5.976917638752611e-06,
+      "loss": 0.3456,
+      "step": 5538
+    },
+    {
+      "epoch": 0.4544447635065841,
+      "grad_norm": 0.6623833307308827,
+      "learning_rate": 5.975614506463003e-06,
+      "loss": 0.3391,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4545268080567748,
+      "grad_norm": 0.6857947812692796,
+      "learning_rate": 5.9743113052770665e-06,
+      "loss": 0.3189,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4546088526069656,
+      "grad_norm": 0.6528317641019266,
+      "learning_rate": 5.9730080352868315e-06,
+      "loss": 0.2922,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4546908971571563,
+      "grad_norm": 0.6609991345309001,
+      "learning_rate": 5.971704696584332e-06,
+      "loss": 0.3745,
+      "step": 5542
+    },
+    {
+      "epoch": 0.4547729417073471,
+      "grad_norm": 0.7061369639985658,
+      "learning_rate": 5.970401289261607e-06,
+      "loss": 0.3435,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45485498625753784,
+      "grad_norm": 0.6373312215475151,
+      "learning_rate": 5.969097813410702e-06,
+      "loss": 0.3824,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45493703080772857,
+      "grad_norm": 0.642361281823492,
+      "learning_rate": 5.967794269123667e-06,
+      "loss": 0.318,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45501907535791936,
+      "grad_norm": 0.7159739929018204,
+      "learning_rate": 5.966490656492555e-06,
+      "loss": 0.3526,
+      "step": 5546
+    },
+    {
+      "epoch": 0.4551011199081101,
+      "grad_norm": 0.7661485544414552,
+      "learning_rate": 5.9651869756094274e-06,
+      "loss": 0.3753,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45518316445830087,
+      "grad_norm": 0.6932638530063452,
+      "learning_rate": 5.9638832265663446e-06,
+      "loss": 0.3169,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4552652090084916,
+      "grad_norm": 0.6207358728507886,
+      "learning_rate": 5.962579409455378e-06,
+      "loss": 0.3205,
+      "step": 5549
+    },
+    {
+      "epoch": 0.4553472535586824,
+      "grad_norm": 0.6888717439657758,
+      "learning_rate": 5.9612755243686e-06,
+      "loss": 0.3437,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4554292981088731,
+      "grad_norm": 0.6173028113380666,
+      "learning_rate": 5.959971571398088e-06,
+      "loss": 0.6786,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45551134265906384,
+      "grad_norm": 0.7108502933887647,
+      "learning_rate": 5.958667550635926e-06,
+      "loss": 0.348,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45559338720925463,
+      "grad_norm": 0.7920827249362714,
+      "learning_rate": 5.957363462174201e-06,
+      "loss": 0.3234,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45567543175944536,
+      "grad_norm": 0.769886138428182,
+      "learning_rate": 5.9560593061050075e-06,
+      "loss": 0.3292,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45575747630963614,
+      "grad_norm": 0.714548230853448,
+      "learning_rate": 5.954755082520443e-06,
+      "loss": 0.348,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4558395208598269,
+      "grad_norm": 0.7138511905171131,
+      "learning_rate": 5.953450791512607e-06,
+      "loss": 0.3608,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45592156541001766,
+      "grad_norm": 1.3480332531131842,
+      "learning_rate": 5.952146433173608e-06,
+      "loss": 0.3158,
+      "step": 5557
+    },
+    {
+      "epoch": 0.4560036099602084,
+      "grad_norm": 0.6542702664985423,
+      "learning_rate": 5.950842007595558e-06,
+      "loss": 0.377,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4560856545103992,
+      "grad_norm": 0.6411625265599863,
+      "learning_rate": 5.949537514870573e-06,
+      "loss": 0.3151,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4561676990605899,
+      "grad_norm": 0.6577815833295287,
+      "learning_rate": 5.948232955090773e-06,
+      "loss": 0.3038,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45624974361078063,
+      "grad_norm": 0.7073627133841413,
+      "learning_rate": 5.946928328348287e-06,
+      "loss": 0.2997,
+      "step": 5561
+    },
+    {
+      "epoch": 0.4563317881609714,
+      "grad_norm": 0.658333878368744,
+      "learning_rate": 5.945623634735243e-06,
+      "loss": 0.3348,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45641383271116215,
+      "grad_norm": 0.6239675523211403,
+      "learning_rate": 5.944318874343777e-06,
+      "loss": 0.322,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45649587726135293,
+      "grad_norm": 0.6304928883848269,
+      "learning_rate": 5.9430140472660305e-06,
+      "loss": 0.3432,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45657792181154366,
+      "grad_norm": 0.9102701876462636,
+      "learning_rate": 5.941709153594146e-06,
+      "loss": 0.3441,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45665996636173445,
+      "grad_norm": 0.6753951424661508,
+      "learning_rate": 5.940404193420274e-06,
+      "loss": 0.3211,
+      "step": 5566
+    },
+    {
+      "epoch": 0.4567420109119252,
+      "grad_norm": 0.6465607232752106,
+      "learning_rate": 5.939099166836568e-06,
+      "loss": 0.352,
+      "step": 5567
+    },
+    {
+      "epoch": 0.4568240554621159,
+      "grad_norm": 0.7658171727854373,
+      "learning_rate": 5.937794073935188e-06,
+      "loss": 0.348,
+      "step": 5568
+    },
+    {
+      "epoch": 0.4569061000123067,
+      "grad_norm": 0.7196852633028321,
+      "learning_rate": 5.936488914808299e-06,
+      "loss": 0.3189,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4569881445624974,
+      "grad_norm": 0.6637642728572852,
+      "learning_rate": 5.935183689548067e-06,
+      "loss": 0.3257,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4570701891126882,
+      "grad_norm": 0.7421734716874182,
+      "learning_rate": 5.933878398246664e-06,
+      "loss": 0.3266,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45715223366287894,
+      "grad_norm": 0.7003584444984671,
+      "learning_rate": 5.932573040996269e-06,
+      "loss": 0.3643,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4572342782130697,
+      "grad_norm": 0.6692826823807438,
+      "learning_rate": 5.931267617889064e-06,
+      "loss": 0.3074,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45731632276326045,
+      "grad_norm": 0.6864573590187877,
+      "learning_rate": 5.929962129017237e-06,
+      "loss": 0.3879,
+      "step": 5574
+    },
+    {
+      "epoch": 0.4573983673134512,
+      "grad_norm": 0.6641824542863001,
+      "learning_rate": 5.928656574472977e-06,
+      "loss": 0.308,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45748041186364197,
+      "grad_norm": 0.8164192912212634,
+      "learning_rate": 5.9273509543484846e-06,
+      "loss": 0.3722,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4575624564138327,
+      "grad_norm": 0.6556268800038538,
+      "learning_rate": 5.926045268735955e-06,
+      "loss": 0.3268,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4576445009640235,
+      "grad_norm": 0.7011062163856382,
+      "learning_rate": 5.924739517727598e-06,
+      "loss": 0.3621,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4577265455142142,
+      "grad_norm": 0.7564384318443695,
+      "learning_rate": 5.9234337014156205e-06,
+      "loss": 0.3263,
+      "step": 5579
+    },
+    {
+      "epoch": 0.457808590064405,
+      "grad_norm": 0.6506223738886315,
+      "learning_rate": 5.92212781989224e-06,
+      "loss": 0.3329,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4578906346145957,
+      "grad_norm": 0.6349843433646007,
+      "learning_rate": 5.9208218732496725e-06,
+      "loss": 0.3117,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45797267916478646,
+      "grad_norm": 0.6894768939509959,
+      "learning_rate": 5.919515861580145e-06,
+      "loss": 0.326,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45805472371497724,
+      "grad_norm": 0.6721422338320288,
+      "learning_rate": 5.9182097849758856e-06,
+      "loss": 0.3664,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45813676826516797,
+      "grad_norm": 0.7495446309772632,
+      "learning_rate": 5.916903643529125e-06,
+      "loss": 0.3321,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45821881281535876,
+      "grad_norm": 0.5928276583448506,
+      "learning_rate": 5.915597437332101e-06,
+      "loss": 0.3153,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4583008573655495,
+      "grad_norm": 1.0960855704915726,
+      "learning_rate": 5.91429116647706e-06,
+      "loss": 0.3579,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45838290191574027,
+      "grad_norm": 0.6672754623195776,
+      "learning_rate": 5.9129848310562435e-06,
+      "loss": 0.3499,
+      "step": 5587
+    },
+    {
+      "epoch": 0.458464946465931,
+      "grad_norm": 0.8529967749865908,
+      "learning_rate": 5.911678431161907e-06,
+      "loss": 0.3432,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45854699101612173,
+      "grad_norm": 0.6246207412681954,
+      "learning_rate": 5.910371966886303e-06,
+      "loss": 0.6318,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4586290355663125,
+      "grad_norm": 0.746595676133107,
+      "learning_rate": 5.909065438321693e-06,
+      "loss": 0.3445,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45871108011650324,
+      "grad_norm": 0.6708414689120072,
+      "learning_rate": 5.907758845560343e-06,
+      "loss": 0.3478,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45879312466669403,
+      "grad_norm": 0.7225947969439656,
+      "learning_rate": 5.906452188694523e-06,
+      "loss": 0.3318,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45887516921688476,
+      "grad_norm": 0.6931856774318916,
+      "learning_rate": 5.905145467816504e-06,
+      "loss": 0.3359,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45895721376707554,
+      "grad_norm": 0.6366168677805925,
+      "learning_rate": 5.903838683018566e-06,
+      "loss": 0.3351,
+      "step": 5594
+    },
+    {
+      "epoch": 0.4590392583172663,
+      "grad_norm": 0.7174774533206663,
+      "learning_rate": 5.902531834392993e-06,
+      "loss": 0.302,
+      "step": 5595
+    },
+    {
+      "epoch": 0.459121302867457,
+      "grad_norm": 0.7812284102424977,
+      "learning_rate": 5.901224922032072e-06,
+      "loss": 0.3023,
+      "step": 5596
+    },
+    {
+      "epoch": 0.4592033474176478,
+      "grad_norm": 0.7642154106020173,
+      "learning_rate": 5.899917946028095e-06,
+      "loss": 0.3376,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4592853919678385,
+      "grad_norm": 0.6766804570651881,
+      "learning_rate": 5.8986109064733586e-06,
+      "loss": 0.3361,
+      "step": 5598
+    },
+    {
+      "epoch": 0.4593674365180293,
+      "grad_norm": 0.6543002974320388,
+      "learning_rate": 5.897303803460162e-06,
+      "loss": 0.3427,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45944948106822003,
+      "grad_norm": 0.6157343494189152,
+      "learning_rate": 5.895996637080814e-06,
+      "loss": 0.317,
+      "step": 5600
+    },
+    {
+      "epoch": 0.4595315256184108,
+      "grad_norm": 0.6364521858715172,
+      "learning_rate": 5.894689407427623e-06,
+      "loss": 0.3243,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45961357016860155,
+      "grad_norm": 0.6713622292610186,
+      "learning_rate": 5.893382114592904e-06,
+      "loss": 0.3467,
+      "step": 5602
+    },
+    {
+      "epoch": 0.4596956147187923,
+      "grad_norm": 0.7004919687277371,
+      "learning_rate": 5.892074758668974e-06,
+      "loss": 0.3088,
+      "step": 5603
+    },
+    {
+      "epoch": 0.45977765926898306,
+      "grad_norm": 0.7669819596501203,
+      "learning_rate": 5.890767339748159e-06,
+      "loss": 0.3451,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4598597038191738,
+      "grad_norm": 0.6059230749650479,
+      "learning_rate": 5.889459857922786e-06,
+      "loss": 0.3406,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4599417483693646,
+      "grad_norm": 0.709874464505298,
+      "learning_rate": 5.888152313285185e-06,
+      "loss": 0.3461,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4600237929195553,
+      "grad_norm": 0.67836227331605,
+      "learning_rate": 5.886844705927696e-06,
+      "loss": 0.3904,
+      "step": 5607
+    },
+    {
+      "epoch": 0.4601058374697461,
+      "grad_norm": 0.7102065968634225,
+      "learning_rate": 5.885537035942657e-06,
+      "loss": 0.3432,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4601878820199368,
+      "grad_norm": 0.7212216636617652,
+      "learning_rate": 5.884229303422417e-06,
+      "loss": 0.3316,
+      "step": 5609
+    },
+    {
+      "epoch": 0.46026992657012755,
+      "grad_norm": 0.7813451457418568,
+      "learning_rate": 5.882921508459323e-06,
+      "loss": 0.3351,
+      "step": 5610
+    },
+    {
+      "epoch": 0.46035197112031834,
+      "grad_norm": 0.6929304489473179,
+      "learning_rate": 5.881613651145732e-06,
+      "loss": 0.3468,
+      "step": 5611
+    },
+    {
+      "epoch": 0.46043401567050907,
+      "grad_norm": 0.8338467280648812,
+      "learning_rate": 5.880305731574001e-06,
+      "loss": 0.3404,
+      "step": 5612
+    },
+    {
+      "epoch": 0.46051606022069985,
+      "grad_norm": 0.6442309183856274,
+      "learning_rate": 5.878997749836493e-06,
+      "loss": 0.3348,
+      "step": 5613
+    },
+    {
+      "epoch": 0.4605981047708906,
+      "grad_norm": 0.80036370569721,
+      "learning_rate": 5.8776897060255755e-06,
+      "loss": 0.3526,
+      "step": 5614
+    },
+    {
+      "epoch": 0.46068014932108137,
+      "grad_norm": 0.6978485357394113,
+      "learning_rate": 5.876381600233623e-06,
+      "loss": 0.3713,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4607621938712721,
+      "grad_norm": 0.6930847977548901,
+      "learning_rate": 5.8750734325530075e-06,
+      "loss": 0.3621,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4608442384214629,
+      "grad_norm": 0.8867558998264928,
+      "learning_rate": 5.873765203076115e-06,
+      "loss": 0.3496,
+      "step": 5617
+    },
+    {
+      "epoch": 0.4609262829716536,
+      "grad_norm": 0.6412492395153948,
+      "learning_rate": 5.872456911895326e-06,
+      "loss": 0.326,
+      "step": 5618
+    },
+    {
+      "epoch": 0.46100832752184434,
+      "grad_norm": 0.7939947686776085,
+      "learning_rate": 5.8711485591030315e-06,
+      "loss": 0.3536,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4610903720720351,
+      "grad_norm": 0.672878074509171,
+      "learning_rate": 5.869840144791625e-06,
+      "loss": 0.3515,
+      "step": 5620
+    },
+    {
+      "epoch": 0.46117241662222586,
+      "grad_norm": 0.643158830987762,
+      "learning_rate": 5.868531669053506e-06,
+      "loss": 0.3127,
+      "step": 5621
+    },
+    {
+      "epoch": 0.46125446117241664,
+      "grad_norm": 0.5946572766363513,
+      "learning_rate": 5.867223131981075e-06,
+      "loss": 0.3229,
+      "step": 5622
+    },
+    {
+      "epoch": 0.46133650572260737,
+      "grad_norm": 0.7946946190496376,
+      "learning_rate": 5.86591453366674e-06,
+      "loss": 0.3309,
+      "step": 5623
+    },
+    {
+      "epoch": 0.46141855027279816,
+      "grad_norm": 0.6093890352942969,
+      "learning_rate": 5.864605874202912e-06,
+      "loss": 0.3575,
+      "step": 5624
+    },
+    {
+      "epoch": 0.4615005948229889,
+      "grad_norm": 0.6775713046048214,
+      "learning_rate": 5.863297153682006e-06,
+      "loss": 0.3469,
+      "step": 5625
+    },
+    {
+      "epoch": 0.4615826393731796,
+      "grad_norm": 0.6487274084713462,
+      "learning_rate": 5.861988372196442e-06,
+      "loss": 0.3565,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4616646839233704,
+      "grad_norm": 0.6698768711130284,
+      "learning_rate": 5.860679529838645e-06,
+      "loss": 0.3039,
+      "step": 5627
+    },
+    {
+      "epoch": 0.46174672847356113,
+      "grad_norm": 0.6427905418362132,
+      "learning_rate": 5.859370626701042e-06,
+      "loss": 0.3211,
+      "step": 5628
+    },
+    {
+      "epoch": 0.4618287730237519,
+      "grad_norm": 0.6576717514325471,
+      "learning_rate": 5.858061662876066e-06,
+      "loss": 0.3253,
+      "step": 5629
+    },
+    {
+      "epoch": 0.46191081757394264,
+      "grad_norm": 0.701705075690251,
+      "learning_rate": 5.8567526384561526e-06,
+      "loss": 0.352,
+      "step": 5630
+    },
+    {
+      "epoch": 0.46199286212413343,
+      "grad_norm": 0.6837494800102436,
+      "learning_rate": 5.855443553533746e-06,
+      "loss": 0.3322,
+      "step": 5631
+    },
+    {
+      "epoch": 0.46207490667432416,
+      "grad_norm": 0.6678776428796558,
+      "learning_rate": 5.85413440820129e-06,
+      "loss": 0.354,
+      "step": 5632
+    },
+    {
+      "epoch": 0.4621569512245149,
+      "grad_norm": 0.6652860079576262,
+      "learning_rate": 5.852825202551234e-06,
+      "loss": 0.3578,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4622389957747057,
+      "grad_norm": 0.7046878807144938,
+      "learning_rate": 5.851515936676031e-06,
+      "loss": 0.3361,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4623210403248964,
+      "grad_norm": 0.6846089112196712,
+      "learning_rate": 5.8502066106681425e-06,
+      "loss": 0.3289,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4624030848750872,
+      "grad_norm": 0.638872189056295,
+      "learning_rate": 5.848897224620027e-06,
+      "loss": 0.3361,
+      "step": 5636
+    },
+    {
+      "epoch": 0.4624851294252779,
+      "grad_norm": 0.6041738858037254,
+      "learning_rate": 5.847587778624155e-06,
+      "loss": 0.334,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4625671739754687,
+      "grad_norm": 0.6968250852540273,
+      "learning_rate": 5.846278272772996e-06,
+      "loss": 0.3233,
+      "step": 5638
+    },
+    {
+      "epoch": 0.46264921852565943,
+      "grad_norm": 0.6672849828130197,
+      "learning_rate": 5.8449687071590245e-06,
+      "loss": 0.3109,
+      "step": 5639
+    },
+    {
+      "epoch": 0.46273126307585016,
+      "grad_norm": 0.6853404389504869,
+      "learning_rate": 5.8436590818747205e-06,
+      "loss": 0.3153,
+      "step": 5640
+    },
+    {
+      "epoch": 0.46281330762604095,
+      "grad_norm": 0.6261927617724908,
+      "learning_rate": 5.842349397012568e-06,
+      "loss": 0.3177,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4628953521762317,
+      "grad_norm": 0.6432644086075544,
+      "learning_rate": 5.841039652665055e-06,
+      "loss": 0.335,
+      "step": 5642
+    },
+    {
+      "epoch": 0.46297739672642246,
+      "grad_norm": 0.7157849052037343,
+      "learning_rate": 5.83972984892467e-06,
+      "loss": 0.3563,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4630594412766132,
+      "grad_norm": 0.7580150230718008,
+      "learning_rate": 5.838419985883914e-06,
+      "loss": 0.3382,
+      "step": 5644
+    },
+    {
+      "epoch": 0.463141485826804,
+      "grad_norm": 0.6554450690139668,
+      "learning_rate": 5.837110063635288e-06,
+      "loss": 0.3799,
+      "step": 5645
+    },
+    {
+      "epoch": 0.4632235303769947,
+      "grad_norm": 0.7418496893573319,
+      "learning_rate": 5.835800082271292e-06,
+      "loss": 0.3459,
+      "step": 5646
+    },
+    {
+      "epoch": 0.46330557492718544,
+      "grad_norm": 0.6391218671549351,
+      "learning_rate": 5.8344900418844375e-06,
+      "loss": 0.3339,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4633876194773762,
+      "grad_norm": 0.636662898049728,
+      "learning_rate": 5.833179942567236e-06,
+      "loss": 0.3193,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46346966402756695,
+      "grad_norm": 0.6783243008827441,
+      "learning_rate": 5.831869784412205e-06,
+      "loss": 0.3508,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46355170857775774,
+      "grad_norm": 0.6515139392200288,
+      "learning_rate": 5.830559567511867e-06,
+      "loss": 0.3467,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46363375312794847,
+      "grad_norm": 0.6605002003121275,
+      "learning_rate": 5.8292492919587475e-06,
+      "loss": 0.6518,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46371579767813925,
+      "grad_norm": 0.6749060190482958,
+      "learning_rate": 5.827938957845376e-06,
+      "loss": 0.3614,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46379784222833,
+      "grad_norm": 0.8329708839216042,
+      "learning_rate": 5.826628565264284e-06,
+      "loss": 0.3442,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4638798867785207,
+      "grad_norm": 0.6813145527459288,
+      "learning_rate": 5.825318114308012e-06,
+      "loss": 0.3607,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4639619313287115,
+      "grad_norm": 0.9540762572197363,
+      "learning_rate": 5.824007605069099e-06,
+      "loss": 0.2884,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4640439758789022,
+      "grad_norm": 0.7936298281020341,
+      "learning_rate": 5.822697037640094e-06,
+      "loss": 0.3373,
+      "step": 5656
+    },
+    {
+      "epoch": 0.464126020429093,
+      "grad_norm": 0.7327339545283205,
+      "learning_rate": 5.821386412113546e-06,
+      "loss": 0.3574,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46420806497928374,
+      "grad_norm": 1.2593623174008202,
+      "learning_rate": 5.820075728582008e-06,
+      "loss": 0.3662,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4642901095294745,
+      "grad_norm": 0.656581754887785,
+      "learning_rate": 5.818764987138041e-06,
+      "loss": 0.3832,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46437215407966526,
+      "grad_norm": 0.6446310404647804,
+      "learning_rate": 5.817454187874206e-06,
+      "loss": 0.3032,
+      "step": 5660
+    },
+    {
+      "epoch": 0.464454198629856,
+      "grad_norm": 1.0469953252484536,
+      "learning_rate": 5.816143330883069e-06,
+      "loss": 0.3358,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46453624318004677,
+      "grad_norm": 1.2437371318533903,
+      "learning_rate": 5.814832416257201e-06,
+      "loss": 0.3331,
+      "step": 5662
+    },
+    {
+      "epoch": 0.4646182877302375,
+      "grad_norm": 0.632829735665257,
+      "learning_rate": 5.813521444089177e-06,
+      "loss": 0.3291,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4647003322804283,
+      "grad_norm": 0.7060543187605239,
+      "learning_rate": 5.812210414471576e-06,
+      "loss": 0.3427,
+      "step": 5664
+    },
+    {
+      "epoch": 0.464782376830619,
+      "grad_norm": 0.6473358016596947,
+      "learning_rate": 5.810899327496982e-06,
+      "loss": 0.3479,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4648644213808098,
+      "grad_norm": 0.6975671980332018,
+      "learning_rate": 5.8095881832579785e-06,
+      "loss": 0.3771,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46494646593100053,
+      "grad_norm": 0.7281542023061512,
+      "learning_rate": 5.808276981847158e-06,
+      "loss": 0.3399,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46502851048119126,
+      "grad_norm": 0.7497563457750749,
+      "learning_rate": 5.806965723357117e-06,
+      "loss": 0.3496,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46511055503138204,
+      "grad_norm": 0.7702950270217217,
+      "learning_rate": 5.805654407880451e-06,
+      "loss": 0.3884,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4651925995815728,
+      "grad_norm": 0.6546537022932486,
+      "learning_rate": 5.804343035509766e-06,
+      "loss": 0.3447,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46527464413176356,
+      "grad_norm": 0.6172072317935141,
+      "learning_rate": 5.803031606337668e-06,
+      "loss": 0.3713,
+      "step": 5671
+    },
+    {
+      "epoch": 0.4653566886819543,
+      "grad_norm": 0.6415853456386006,
+      "learning_rate": 5.8017201204567685e-06,
+      "loss": 0.3009,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4654387332321451,
+      "grad_norm": 0.6181269219598822,
+      "learning_rate": 5.800408577959683e-06,
+      "loss": 0.3373,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4655207777823358,
+      "grad_norm": 0.7037319719523585,
+      "learning_rate": 5.799096978939029e-06,
+      "loss": 0.3781,
+      "step": 5674
+    },
+    {
+      "epoch": 0.4656028223325266,
+      "grad_norm": 0.6682989150927433,
+      "learning_rate": 5.797785323487429e-06,
+      "loss": 0.3532,
+      "step": 5675
+    },
+    {
+      "epoch": 0.4656848668827173,
+      "grad_norm": 0.8469001526364878,
+      "learning_rate": 5.7964736116975126e-06,
+      "loss": 0.3532,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46576691143290805,
+      "grad_norm": 0.881641625244505,
+      "learning_rate": 5.7951618436619094e-06,
+      "loss": 0.3427,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46584895598309883,
+      "grad_norm": 0.7798728802414348,
+      "learning_rate": 5.793850019473254e-06,
+      "loss": 0.3002,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46593100053328956,
+      "grad_norm": 0.6481697704564919,
+      "learning_rate": 5.7925381392241855e-06,
+      "loss": 0.3422,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46601304508348035,
+      "grad_norm": 0.6889872903662249,
+      "learning_rate": 5.791226203007346e-06,
+      "loss": 0.3148,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4660950896336711,
+      "grad_norm": 0.6506388174528905,
+      "learning_rate": 5.789914210915385e-06,
+      "loss": 0.3078,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46617713418386186,
+      "grad_norm": 0.6869316472456992,
+      "learning_rate": 5.788602163040951e-06,
+      "loss": 0.3261,
+      "step": 5682
+    },
+    {
+      "epoch": 0.4662591787340526,
+      "grad_norm": 0.6950803822493425,
+      "learning_rate": 5.787290059476699e-06,
+      "loss": 0.3468,
+      "step": 5683
+    },
+    {
+      "epoch": 0.4663412232842433,
+      "grad_norm": 0.8056980343966856,
+      "learning_rate": 5.785977900315287e-06,
+      "loss": 0.3185,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4664232678344341,
+      "grad_norm": 0.7406628628059071,
+      "learning_rate": 5.7846656856493784e-06,
+      "loss": 0.3216,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46650531238462484,
+      "grad_norm": 0.702193114892242,
+      "learning_rate": 5.783353415571642e-06,
+      "loss": 0.3674,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4665873569348156,
+      "grad_norm": 1.219078887485048,
+      "learning_rate": 5.782041090174745e-06,
+      "loss": 0.327,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46666940148500635,
+      "grad_norm": 0.7726659248602945,
+      "learning_rate": 5.780728709551361e-06,
+      "loss": 0.3176,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46675144603519714,
+      "grad_norm": 0.8934145421651492,
+      "learning_rate": 5.7794162737941715e-06,
+      "loss": 0.3111,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46683349058538787,
+      "grad_norm": 0.9569924369941886,
+      "learning_rate": 5.778103782995857e-06,
+      "loss": 0.3354,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4669155351355786,
+      "grad_norm": 0.628013597399316,
+      "learning_rate": 5.776791237249103e-06,
+      "loss": 0.3133,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4669975796857694,
+      "grad_norm": 0.8258790252714707,
+      "learning_rate": 5.7754786366465996e-06,
+      "loss": 0.3387,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4670796242359601,
+      "grad_norm": 0.7334875212838678,
+      "learning_rate": 5.774165981281041e-06,
+      "loss": 0.3187,
+      "step": 5693
+    },
+    {
+      "epoch": 0.4671616687861509,
+      "grad_norm": 0.7243823709729214,
+      "learning_rate": 5.772853271245124e-06,
+      "loss": 0.3077,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4672437133363416,
+      "grad_norm": 0.6562767205730785,
+      "learning_rate": 5.771540506631552e-06,
+      "loss": 0.3512,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4673257578865324,
+      "grad_norm": 0.747020049775407,
+      "learning_rate": 5.770227687533029e-06,
+      "loss": 0.3513,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46740780243672314,
+      "grad_norm": 0.6871990735792035,
+      "learning_rate": 5.768914814042262e-06,
+      "loss": 0.3472,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46748984698691387,
+      "grad_norm": 0.7200514102697401,
+      "learning_rate": 5.767601886251968e-06,
+      "loss": 0.3024,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46757189153710466,
+      "grad_norm": 0.641699382158038,
+      "learning_rate": 5.766288904254862e-06,
+      "loss": 0.3462,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4676539360872954,
+      "grad_norm": 0.9742008458407339,
+      "learning_rate": 5.764975868143664e-06,
+      "loss": 0.3418,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46773598063748617,
+      "grad_norm": 0.6996650625797898,
+      "learning_rate": 5.763662778011101e-06,
+      "loss": 0.3854,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4678180251876769,
+      "grad_norm": 0.670192844227175,
+      "learning_rate": 5.762349633949897e-06,
+      "loss": 0.3339,
+      "step": 5702
+    },
+    {
+      "epoch": 0.4679000697378677,
+      "grad_norm": 0.6712138851434098,
+      "learning_rate": 5.761036436052788e-06,
+      "loss": 0.3095,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4679821142880584,
+      "grad_norm": 0.9976963519196268,
+      "learning_rate": 5.759723184412509e-06,
+      "loss": 0.3419,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46806415883824914,
+      "grad_norm": 0.7349988692175181,
+      "learning_rate": 5.758409879121799e-06,
+      "loss": 0.3535,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46814620338843993,
+      "grad_norm": 0.6598762360214864,
+      "learning_rate": 5.7570965202734e-06,
+      "loss": 0.3507,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46822824793863066,
+      "grad_norm": 0.660756998655975,
+      "learning_rate": 5.755783107960064e-06,
+      "loss": 0.3347,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46831029248882144,
+      "grad_norm": 0.7869587727039768,
+      "learning_rate": 5.754469642274538e-06,
+      "loss": 0.3292,
+      "step": 5708
+    },
+    {
+      "epoch": 0.4683923370390122,
+      "grad_norm": 0.6709887625052827,
+      "learning_rate": 5.753156123309578e-06,
+      "loss": 0.3084,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46847438158920296,
+      "grad_norm": 0.6717123093823959,
+      "learning_rate": 5.7518425511579424e-06,
+      "loss": 0.3139,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4685564261393937,
+      "grad_norm": 0.7086692278963683,
+      "learning_rate": 5.7505289259123934e-06,
+      "loss": 0.3481,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4686384706895844,
+      "grad_norm": 0.6657687860660522,
+      "learning_rate": 5.7492152476656975e-06,
+      "loss": 0.2985,
+      "step": 5712
+    },
+    {
+      "epoch": 0.4687205152397752,
+      "grad_norm": 0.7575455536671059,
+      "learning_rate": 5.747901516510624e-06,
+      "loss": 0.3719,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46880255978996593,
+      "grad_norm": 0.6506929167519427,
+      "learning_rate": 5.746587732539949e-06,
+      "loss": 0.3594,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4688846043401567,
+      "grad_norm": 0.6178091799936627,
+      "learning_rate": 5.745273895846447e-06,
+      "loss": 0.3269,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46896664889034745,
+      "grad_norm": 0.6607663252540589,
+      "learning_rate": 5.743960006522899e-06,
+      "loss": 0.3214,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46904869344053823,
+      "grad_norm": 1.161593604589931,
+      "learning_rate": 5.742646064662092e-06,
+      "loss": 0.3547,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46913073799072896,
+      "grad_norm": 0.6820126070263495,
+      "learning_rate": 5.741332070356812e-06,
+      "loss": 0.3121,
+      "step": 5718
+    },
+    {
+      "epoch": 0.4692127825409197,
+      "grad_norm": 0.7454690114338495,
+      "learning_rate": 5.7400180236998524e-06,
+      "loss": 0.3402,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4692948270911105,
+      "grad_norm": 0.6594328291796367,
+      "learning_rate": 5.738703924784008e-06,
+      "loss": 0.3143,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4693768716413012,
+      "grad_norm": 0.651740142039005,
+      "learning_rate": 5.7373897737020814e-06,
+      "loss": 0.3351,
+      "step": 5721
+    },
+    {
+      "epoch": 0.469458916191492,
+      "grad_norm": 0.6183046494310948,
+      "learning_rate": 5.7360755705468725e-06,
+      "loss": 0.3441,
+      "step": 5722
+    },
+    {
+      "epoch": 0.4695409607416827,
+      "grad_norm": 0.709894475386862,
+      "learning_rate": 5.734761315411189e-06,
+      "loss": 0.2999,
+      "step": 5723
+    },
+    {
+      "epoch": 0.4696230052918735,
+      "grad_norm": 0.8278949263630512,
+      "learning_rate": 5.733447008387842e-06,
+      "loss": 0.3097,
+      "step": 5724
+    },
+    {
+      "epoch": 0.46970504984206424,
+      "grad_norm": 0.6035581250595632,
+      "learning_rate": 5.7321326495696465e-06,
+      "loss": 0.3573,
+      "step": 5725
+    },
+    {
+      "epoch": 0.46978709439225497,
+      "grad_norm": 0.6042139354667709,
+      "learning_rate": 5.7308182390494185e-06,
+      "loss": 0.3536,
+      "step": 5726
+    },
+    {
+      "epoch": 0.46986913894244575,
+      "grad_norm": 0.6485460533640659,
+      "learning_rate": 5.729503776919981e-06,
+      "loss": 0.3423,
+      "step": 5727
+    },
+    {
+      "epoch": 0.4699511834926365,
+      "grad_norm": 0.6630573915486866,
+      "learning_rate": 5.7281892632741585e-06,
+      "loss": 0.3282,
+      "step": 5728
+    },
+    {
+      "epoch": 0.47003322804282727,
+      "grad_norm": 0.6011098123159836,
+      "learning_rate": 5.726874698204781e-06,
+      "loss": 0.3217,
+      "step": 5729
+    },
+    {
+      "epoch": 0.470115272593018,
+      "grad_norm": 0.6220820824827509,
+      "learning_rate": 5.72556008180468e-06,
+      "loss": 0.32,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4701973171432088,
+      "grad_norm": 0.7302613623823453,
+      "learning_rate": 5.724245414166692e-06,
+      "loss": 0.3327,
+      "step": 5731
+    },
+    {
+      "epoch": 0.4702793616933995,
+      "grad_norm": 0.742562949137512,
+      "learning_rate": 5.722930695383655e-06,
+      "loss": 0.3485,
+      "step": 5732
+    },
+    {
+      "epoch": 0.4703614062435903,
+      "grad_norm": 0.7109741433953998,
+      "learning_rate": 5.721615925548414e-06,
+      "loss": 0.3191,
+      "step": 5733
+    },
+    {
+      "epoch": 0.470443450793781,
+      "grad_norm": 0.7236661873428676,
+      "learning_rate": 5.720301104753816e-06,
+      "loss": 0.3453,
+      "step": 5734
+    },
+    {
+      "epoch": 0.47052549534397176,
+      "grad_norm": 0.6135740881429658,
+      "learning_rate": 5.718986233092713e-06,
+      "loss": 0.3246,
+      "step": 5735
+    },
+    {
+      "epoch": 0.47060753989416254,
+      "grad_norm": 0.6425593800666708,
+      "learning_rate": 5.717671310657956e-06,
+      "loss": 0.3707,
+      "step": 5736
+    },
+    {
+      "epoch": 0.47068958444435327,
+      "grad_norm": 1.537081774256029,
+      "learning_rate": 5.716356337542404e-06,
+      "loss": 0.3158,
+      "step": 5737
+    },
+    {
+      "epoch": 0.47077162899454406,
+      "grad_norm": 0.7010697705213512,
+      "learning_rate": 5.715041313838917e-06,
+      "loss": 0.3266,
+      "step": 5738
+    },
+    {
+      "epoch": 0.4708536735447348,
+      "grad_norm": 0.6005854304673508,
+      "learning_rate": 5.713726239640363e-06,
+      "loss": 0.3134,
+      "step": 5739
+    },
+    {
+      "epoch": 0.47093571809492557,
+      "grad_norm": 0.6688207478880569,
+      "learning_rate": 5.712411115039609e-06,
+      "loss": 0.3275,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4710177626451163,
+      "grad_norm": 0.6581403551233702,
+      "learning_rate": 5.711095940129525e-06,
+      "loss": 0.3471,
+      "step": 5741
+    },
+    {
+      "epoch": 0.47109980719530703,
+      "grad_norm": 0.6992679011650181,
+      "learning_rate": 5.70978071500299e-06,
+      "loss": 0.3664,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4711818517454978,
+      "grad_norm": 0.6251148743905443,
+      "learning_rate": 5.7084654397528795e-06,
+      "loss": 0.3584,
+      "step": 5743
+    },
+    {
+      "epoch": 0.47126389629568854,
+      "grad_norm": 0.6954666603326526,
+      "learning_rate": 5.707150114472079e-06,
+      "loss": 0.3381,
+      "step": 5744
+    },
+    {
+      "epoch": 0.47134594084587933,
+      "grad_norm": 0.75920640573586,
+      "learning_rate": 5.705834739253473e-06,
+      "loss": 0.3326,
+      "step": 5745
+    },
+    {
+      "epoch": 0.47142798539607006,
+      "grad_norm": 0.7270033558351008,
+      "learning_rate": 5.704519314189953e-06,
+      "loss": 0.3178,
+      "step": 5746
+    },
+    {
+      "epoch": 0.47151002994626084,
+      "grad_norm": 0.656186741127401,
+      "learning_rate": 5.703203839374408e-06,
+      "loss": 0.3358,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4715920744964516,
+      "grad_norm": 0.6747194347399383,
+      "learning_rate": 5.70188831489974e-06,
+      "loss": 0.3732,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4716741190466423,
+      "grad_norm": 0.688141510426695,
+      "learning_rate": 5.700572740858847e-06,
+      "loss": 0.3189,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4717561635968331,
+      "grad_norm": 0.673985480759074,
+      "learning_rate": 5.699257117344632e-06,
+      "loss": 0.3405,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4718382081470238,
+      "grad_norm": 0.6891613811334615,
+      "learning_rate": 5.697941444450002e-06,
+      "loss": 0.3568,
+      "step": 5751
+    },
+    {
+      "epoch": 0.4719202526972146,
+      "grad_norm": 0.8787203493200316,
+      "learning_rate": 5.696625722267869e-06,
+      "loss": 0.3354,
+      "step": 5752
+    },
+    {
+      "epoch": 0.47200229724740533,
+      "grad_norm": 0.8022905553881532,
+      "learning_rate": 5.695309950891147e-06,
+      "loss": 0.3622,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4720843417975961,
+      "grad_norm": 0.6561820099233279,
+      "learning_rate": 5.693994130412753e-06,
+      "loss": 0.2926,
+      "step": 5754
+    },
+    {
+      "epoch": 0.47216638634778685,
+      "grad_norm": 0.649180701805331,
+      "learning_rate": 5.692678260925609e-06,
+      "loss": 0.325,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4722484308979776,
+      "grad_norm": 0.6210872341638767,
+      "learning_rate": 5.691362342522639e-06,
+      "loss": 0.3444,
+      "step": 5756
+    },
+    {
+      "epoch": 0.47233047544816836,
+      "grad_norm": 0.6481060080526465,
+      "learning_rate": 5.6900463752967715e-06,
+      "loss": 0.3338,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4724125199983591,
+      "grad_norm": 0.6462421742731278,
+      "learning_rate": 5.6887303593409375e-06,
+      "loss": 0.3502,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4724945645485499,
+      "grad_norm": 0.7013442654604444,
+      "learning_rate": 5.687414294748072e-06,
+      "loss": 0.3269,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4725766090987406,
+      "grad_norm": 0.6560393469616381,
+      "learning_rate": 5.686098181611112e-06,
+      "loss": 0.3688,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4726586536489314,
+      "grad_norm": 0.6925767556787797,
+      "learning_rate": 5.684782020023002e-06,
+      "loss": 0.3295,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4727406981991221,
+      "grad_norm": 0.7258193454719536,
+      "learning_rate": 5.683465810076686e-06,
+      "loss": 0.3199,
+      "step": 5762
+    },
+    {
+      "epoch": 0.47282274274931285,
+      "grad_norm": 1.0716855582095814,
+      "learning_rate": 5.682149551865114e-06,
+      "loss": 0.3787,
+      "step": 5763
+    },
+    {
+      "epoch": 0.47290478729950364,
+      "grad_norm": 0.7570152423512436,
+      "learning_rate": 5.680833245481234e-06,
+      "loss": 0.351,
+      "step": 5764
+    },
+    {
+      "epoch": 0.47298683184969437,
+      "grad_norm": 0.7124877417694819,
+      "learning_rate": 5.6795168910180045e-06,
+      "loss": 0.3551,
+      "step": 5765
+    },
+    {
+      "epoch": 0.47306887639988515,
+      "grad_norm": 0.6239413665155625,
+      "learning_rate": 5.678200488568384e-06,
+      "loss": 0.3591,
+      "step": 5766
+    },
+    {
+      "epoch": 0.4731509209500759,
+      "grad_norm": 0.6014923795694451,
+      "learning_rate": 5.676884038225335e-06,
+      "loss": 0.3151,
+      "step": 5767
+    },
+    {
+      "epoch": 0.47323296550026667,
+      "grad_norm": 0.638810439489679,
+      "learning_rate": 5.675567540081823e-06,
+      "loss": 0.3274,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4733150100504574,
+      "grad_norm": 0.5906772921991192,
+      "learning_rate": 5.674250994230816e-06,
+      "loss": 0.3512,
+      "step": 5769
+    },
+    {
+      "epoch": 0.4733970546006481,
+      "grad_norm": 0.7051188601196372,
+      "learning_rate": 5.672934400765289e-06,
+      "loss": 0.3427,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4734790991508389,
+      "grad_norm": 0.7389448390714702,
+      "learning_rate": 5.671617759778214e-06,
+      "loss": 0.3666,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47356114370102964,
+      "grad_norm": 0.6917660656409399,
+      "learning_rate": 5.6703010713625715e-06,
+      "loss": 0.3497,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4736431882512204,
+      "grad_norm": 0.6387875719699064,
+      "learning_rate": 5.668984335611345e-06,
+      "loss": 0.3567,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47372523280141116,
+      "grad_norm": 1.543371687212305,
+      "learning_rate": 5.667667552617518e-06,
+      "loss": 0.3362,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47380727735160194,
+      "grad_norm": 0.6860792655535065,
+      "learning_rate": 5.666350722474083e-06,
+      "loss": 0.2958,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47388932190179267,
+      "grad_norm": 0.7064669655489002,
+      "learning_rate": 5.66503384527403e-06,
+      "loss": 0.3145,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4739713664519834,
+      "grad_norm": 0.751933844920491,
+      "learning_rate": 5.663716921110357e-06,
+      "loss": 0.3522,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4740534110021742,
+      "grad_norm": 0.6775041354192025,
+      "learning_rate": 5.662399950076059e-06,
+      "loss": 0.3345,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4741354555523649,
+      "grad_norm": 0.9536142358821604,
+      "learning_rate": 5.6610829322641415e-06,
+      "loss": 0.3535,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4742175001025557,
+      "grad_norm": 0.7068279170781496,
+      "learning_rate": 5.65976586776761e-06,
+      "loss": 0.3551,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47429954465274643,
+      "grad_norm": 0.6769394456325483,
+      "learning_rate": 5.658448756679472e-06,
+      "loss": 0.3549,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4743815892029372,
+      "grad_norm": 0.6785427069445297,
+      "learning_rate": 5.657131599092741e-06,
+      "loss": 0.3322,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47446363375312794,
+      "grad_norm": 0.6581315422419188,
+      "learning_rate": 5.655814395100432e-06,
+      "loss": 0.3088,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47454567830331873,
+      "grad_norm": 0.64586113633901,
+      "learning_rate": 5.654497144795564e-06,
+      "loss": 0.3326,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47462772285350946,
+      "grad_norm": 0.6640640874584368,
+      "learning_rate": 5.65317984827116e-06,
+      "loss": 0.3164,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4747097674037002,
+      "grad_norm": 0.6547447252127988,
+      "learning_rate": 5.651862505620245e-06,
+      "loss": 0.3026,
+      "step": 5786
+    },
+    {
+      "epoch": 0.474791811953891,
+      "grad_norm": 0.8712893439372147,
+      "learning_rate": 5.650545116935845e-06,
+      "loss": 0.3004,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4748738565040817,
+      "grad_norm": 0.6824468082435192,
+      "learning_rate": 5.649227682310995e-06,
+      "loss": 0.3166,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4749559010542725,
+      "grad_norm": 0.7256639738935461,
+      "learning_rate": 5.6479102018387284e-06,
+      "loss": 0.3046,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4750379456044632,
+      "grad_norm": 0.735702267371449,
+      "learning_rate": 5.6465926756120856e-06,
+      "loss": 0.331,
+      "step": 5790
+    },
+    {
+      "epoch": 0.475119990154654,
+      "grad_norm": 0.7384671705556669,
+      "learning_rate": 5.645275103724108e-06,
+      "loss": 0.3535,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47520203470484473,
+      "grad_norm": 1.092933937263611,
+      "learning_rate": 5.643957486267837e-06,
+      "loss": 0.3426,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47528407925503546,
+      "grad_norm": 0.784418465334993,
+      "learning_rate": 5.642639823336325e-06,
+      "loss": 0.303,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47536612380522625,
+      "grad_norm": 0.6320436418934405,
+      "learning_rate": 5.641322115022619e-06,
+      "loss": 0.3057,
+      "step": 5794
+    },
+    {
+      "epoch": 0.475448168355417,
+      "grad_norm": 0.7801603886915517,
+      "learning_rate": 5.640004361419776e-06,
+      "loss": 0.3259,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47553021290560776,
+      "grad_norm": 0.6574120730517724,
+      "learning_rate": 5.638686562620853e-06,
+      "loss": 0.3129,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4756122574557985,
+      "grad_norm": 0.7998188746925676,
+      "learning_rate": 5.637368718718913e-06,
+      "loss": 0.3657,
+      "step": 5797
+    },
+    {
+      "epoch": 0.4756943020059893,
+      "grad_norm": 0.7717443659888918,
+      "learning_rate": 5.636050829807017e-06,
+      "loss": 0.3462,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47577634655618,
+      "grad_norm": 0.6558847518562111,
+      "learning_rate": 5.634732895978233e-06,
+      "loss": 0.3196,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47585839110637074,
+      "grad_norm": 0.6558742209136228,
+      "learning_rate": 5.6334149173256324e-06,
+      "loss": 0.329,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4759404356565615,
+      "grad_norm": 0.6110398874714734,
+      "learning_rate": 5.632096893942286e-06,
+      "loss": 0.2868,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47602248020675225,
+      "grad_norm": 0.9916059847666463,
+      "learning_rate": 5.630778825921274e-06,
+      "loss": 0.3336,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47610452475694304,
+      "grad_norm": 0.7993609454073255,
+      "learning_rate": 5.629460713355674e-06,
+      "loss": 0.2869,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47618656930713377,
+      "grad_norm": 0.7433008900036907,
+      "learning_rate": 5.628142556338572e-06,
+      "loss": 0.3754,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47626861385732455,
+      "grad_norm": 0.9001305751367051,
+      "learning_rate": 5.6268243549630495e-06,
+      "loss": 0.3244,
+      "step": 5805
+    },
+    {
+      "epoch": 0.4763506584075153,
+      "grad_norm": 0.7408216421156409,
+      "learning_rate": 5.625506109322198e-06,
+      "loss": 0.3246,
+      "step": 5806
+    },
+    {
+      "epoch": 0.476432702957706,
+      "grad_norm": 0.6499653493560755,
+      "learning_rate": 5.624187819509111e-06,
+      "loss": 0.3568,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4765147475078968,
+      "grad_norm": 1.0310101885836804,
+      "learning_rate": 5.622869485616884e-06,
+      "loss": 0.3413,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4765967920580875,
+      "grad_norm": 0.6170460325650196,
+      "learning_rate": 5.621551107738612e-06,
+      "loss": 0.3397,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4766788366082783,
+      "grad_norm": 0.645885724607222,
+      "learning_rate": 5.620232685967401e-06,
+      "loss": 0.3724,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47676088115846904,
+      "grad_norm": 0.6686065530027615,
+      "learning_rate": 5.6189142203963555e-06,
+      "loss": 0.3249,
+      "step": 5811
+    },
+    {
+      "epoch": 0.4768429257086598,
+      "grad_norm": 0.7316135058939681,
+      "learning_rate": 5.617595711118581e-06,
+      "loss": 0.3382,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47692497025885056,
+      "grad_norm": 0.6691424404557327,
+      "learning_rate": 5.616277158227189e-06,
+      "loss": 0.3342,
+      "step": 5813
+    },
+    {
+      "epoch": 0.4770070148090413,
+      "grad_norm": 0.7619711302672337,
+      "learning_rate": 5.614958561815293e-06,
+      "loss": 0.3235,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47708905935923207,
+      "grad_norm": 0.7216352436105262,
+      "learning_rate": 5.613639921976014e-06,
+      "loss": 0.3029,
+      "step": 5815
+    },
+    {
+      "epoch": 0.4771711039094228,
+      "grad_norm": 0.6899617764570439,
+      "learning_rate": 5.612321238802469e-06,
+      "loss": 0.3414,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4772531484596136,
+      "grad_norm": 0.706466634904955,
+      "learning_rate": 5.611002512387781e-06,
+      "loss": 0.3708,
+      "step": 5817
+    },
+    {
+      "epoch": 0.4773351930098043,
+      "grad_norm": 0.6879024478873225,
+      "learning_rate": 5.609683742825078e-06,
+      "loss": 0.3158,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4774172375599951,
+      "grad_norm": 0.6377933801793106,
+      "learning_rate": 5.608364930207488e-06,
+      "loss": 0.3534,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47749928211018583,
+      "grad_norm": 0.8269170948875589,
+      "learning_rate": 5.607046074628144e-06,
+      "loss": 0.37,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47758132666037656,
+      "grad_norm": 0.6810750274631096,
+      "learning_rate": 5.605727176180182e-06,
+      "loss": 0.269,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47766337121056734,
+      "grad_norm": 0.7048896358941374,
+      "learning_rate": 5.604408234956741e-06,
+      "loss": 0.3528,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4777454157607581,
+      "grad_norm": 0.6635748010775592,
+      "learning_rate": 5.603089251050958e-06,
+      "loss": 0.3169,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47782746031094886,
+      "grad_norm": 0.7324038241864584,
+      "learning_rate": 5.601770224555985e-06,
+      "loss": 0.3119,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4779095048611396,
+      "grad_norm": 0.713291944180992,
+      "learning_rate": 5.600451155564964e-06,
+      "loss": 0.3264,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4779915494113304,
+      "grad_norm": 0.6870959972392929,
+      "learning_rate": 5.599132044171047e-06,
+      "loss": 0.3446,
+      "step": 5826
+    },
+    {
+      "epoch": 0.4780735939615211,
+      "grad_norm": 0.9590619932074397,
+      "learning_rate": 5.597812890467387e-06,
+      "loss": 0.3445,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47815563851171183,
+      "grad_norm": 0.7831003764761274,
+      "learning_rate": 5.596493694547142e-06,
+      "loss": 0.3105,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4782376830619026,
+      "grad_norm": 0.6982972386664962,
+      "learning_rate": 5.59517445650347e-06,
+      "loss": 0.353,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47831972761209335,
+      "grad_norm": 0.7226919706431887,
+      "learning_rate": 5.593855176429535e-06,
+      "loss": 0.3403,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47840177216228413,
+      "grad_norm": 0.6806593624694443,
+      "learning_rate": 5.5925358544185e-06,
+      "loss": 0.34,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47848381671247486,
+      "grad_norm": 0.7041799469926452,
+      "learning_rate": 5.591216490563537e-06,
+      "loss": 0.3336,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47856586126266565,
+      "grad_norm": 0.8710597323594446,
+      "learning_rate": 5.5898970849578135e-06,
+      "loss": 0.3592,
+      "step": 5833
+    },
+    {
+      "epoch": 0.4786479058128564,
+      "grad_norm": 0.8489065654683358,
+      "learning_rate": 5.588577637694506e-06,
+      "loss": 0.3008,
+      "step": 5834
+    },
+    {
+      "epoch": 0.4787299503630471,
+      "grad_norm": 0.7072048441702786,
+      "learning_rate": 5.5872581488667935e-06,
+      "loss": 0.3698,
+      "step": 5835
+    },
+    {
+      "epoch": 0.4788119949132379,
+      "grad_norm": 0.7130703194595901,
+      "learning_rate": 5.58593861856785e-06,
+      "loss": 0.3776,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4788940394634286,
+      "grad_norm": 0.839146751117855,
+      "learning_rate": 5.584619046890866e-06,
+      "loss": 0.3675,
+      "step": 5837
+    },
+    {
+      "epoch": 0.4789760840136194,
+      "grad_norm": 0.8106871431317682,
+      "learning_rate": 5.583299433929023e-06,
+      "loss": 0.3299,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47905812856381014,
+      "grad_norm": 0.7427592582468495,
+      "learning_rate": 5.581979779775512e-06,
+      "loss": 0.3512,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4791401731140009,
+      "grad_norm": 0.8105843207119885,
+      "learning_rate": 5.5806600845235225e-06,
+      "loss": 0.3211,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47922221766419165,
+      "grad_norm": 0.5993572619192716,
+      "learning_rate": 5.579340348266251e-06,
+      "loss": 0.3255,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47930426221438244,
+      "grad_norm": 0.7188691705772143,
+      "learning_rate": 5.578020571096895e-06,
+      "loss": 0.3417,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47938630676457317,
+      "grad_norm": 0.6662376603542941,
+      "learning_rate": 5.576700753108656e-06,
+      "loss": 0.3328,
+      "step": 5843
+    },
+    {
+      "epoch": 0.4794683513147639,
+      "grad_norm": 0.6739622428307548,
+      "learning_rate": 5.5753808943947355e-06,
+      "loss": 0.3628,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4795503958649547,
+      "grad_norm": 0.7478133579230836,
+      "learning_rate": 5.574060995048341e-06,
+      "loss": 0.3498,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4796324404151454,
+      "grad_norm": 0.6330563441168316,
+      "learning_rate": 5.572741055162681e-06,
+      "loss": 0.3314,
+      "step": 5846
+    },
+    {
+      "epoch": 0.4797144849653362,
+      "grad_norm": 0.6975282743506287,
+      "learning_rate": 5.571421074830967e-06,
+      "loss": 0.3084,
+      "step": 5847
+    },
+    {
+      "epoch": 0.4797965295155269,
+      "grad_norm": 0.8517975633126432,
+      "learning_rate": 5.5701010541464165e-06,
+      "loss": 0.3149,
+      "step": 5848
+    },
+    {
+      "epoch": 0.4798785740657177,
+      "grad_norm": 0.596449391657487,
+      "learning_rate": 5.568780993202244e-06,
+      "loss": 0.3355,
+      "step": 5849
+    },
+    {
+      "epoch": 0.47996061861590844,
+      "grad_norm": 0.7185568980164502,
+      "learning_rate": 5.5674608920916715e-06,
+      "loss": 0.3047,
+      "step": 5850
+    },
+    {
+      "epoch": 0.48004266316609917,
+      "grad_norm": 0.7168009685333881,
+      "learning_rate": 5.566140750907923e-06,
+      "loss": 0.3071,
+      "step": 5851
+    },
+    {
+      "epoch": 0.48012470771628996,
+      "grad_norm": 0.6643973589376521,
+      "learning_rate": 5.564820569744225e-06,
+      "loss": 0.3431,
+      "step": 5852
+    },
+    {
+      "epoch": 0.4802067522664807,
+      "grad_norm": 0.7653069856383651,
+      "learning_rate": 5.5635003486938066e-06,
+      "loss": 0.3651,
+      "step": 5853
+    },
+    {
+      "epoch": 0.48028879681667147,
+      "grad_norm": 0.6724736592952175,
+      "learning_rate": 5.562180087849897e-06,
+      "loss": 0.3351,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4803708413668622,
+      "grad_norm": 0.8399929545713878,
+      "learning_rate": 5.560859787305733e-06,
+      "loss": 0.3323,
+      "step": 5855
+    },
+    {
+      "epoch": 0.480452885917053,
+      "grad_norm": 0.6956448956703046,
+      "learning_rate": 5.559539447154552e-06,
+      "loss": 0.3318,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4805349304672437,
+      "grad_norm": 1.0041623515145273,
+      "learning_rate": 5.558219067489594e-06,
+      "loss": 0.3016,
+      "step": 5857
+    },
+    {
+      "epoch": 0.48061697501743444,
+      "grad_norm": 0.7011019679016505,
+      "learning_rate": 5.556898648404104e-06,
+      "loss": 0.3077,
+      "step": 5858
+    },
+    {
+      "epoch": 0.48069901956762523,
+      "grad_norm": 0.7646111348580772,
+      "learning_rate": 5.555578189991324e-06,
+      "loss": 0.3542,
+      "step": 5859
+    },
+    {
+      "epoch": 0.48078106411781596,
+      "grad_norm": 0.6639121648571633,
+      "learning_rate": 5.554257692344507e-06,
+      "loss": 0.3368,
+      "step": 5860
+    },
+    {
+      "epoch": 0.48086310866800674,
+      "grad_norm": 0.6606963317400342,
+      "learning_rate": 5.552937155556901e-06,
+      "loss": 0.3368,
+      "step": 5861
+    },
+    {
+      "epoch": 0.4809451532181975,
+      "grad_norm": 0.7987879916261934,
+      "learning_rate": 5.5516165797217615e-06,
+      "loss": 0.3209,
+      "step": 5862
+    },
+    {
+      "epoch": 0.48102719776838826,
+      "grad_norm": 0.627885450814271,
+      "learning_rate": 5.550295964932344e-06,
+      "loss": 0.3148,
+      "step": 5863
+    },
+    {
+      "epoch": 0.481109242318579,
+      "grad_norm": 0.7064505181226096,
+      "learning_rate": 5.548975311281911e-06,
+      "loss": 0.3854,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4811912868687697,
+      "grad_norm": 0.6972311013154606,
+      "learning_rate": 5.5476546188637235e-06,
+      "loss": 0.3293,
+      "step": 5865
+    },
+    {
+      "epoch": 0.4812733314189605,
+      "grad_norm": 0.794610204621401,
+      "learning_rate": 5.546333887771047e-06,
+      "loss": 0.3202,
+      "step": 5866
+    },
+    {
+      "epoch": 0.48135537596915123,
+      "grad_norm": 0.8544085075226078,
+      "learning_rate": 5.54501311809715e-06,
+      "loss": 0.311,
+      "step": 5867
+    },
+    {
+      "epoch": 0.481437420519342,
+      "grad_norm": 0.6520828599499673,
+      "learning_rate": 5.5436923099353005e-06,
+      "loss": 0.3479,
+      "step": 5868
+    },
+    {
+      "epoch": 0.48151946506953275,
+      "grad_norm": 0.8547429970933862,
+      "learning_rate": 5.542371463378773e-06,
+      "loss": 0.3493,
+      "step": 5869
+    },
+    {
+      "epoch": 0.48160150961972353,
+      "grad_norm": 0.6397048134741544,
+      "learning_rate": 5.541050578520846e-06,
+      "loss": 0.3414,
+      "step": 5870
+    },
+    {
+      "epoch": 0.48168355416991426,
+      "grad_norm": 0.7562019039907907,
+      "learning_rate": 5.539729655454795e-06,
+      "loss": 0.3602,
+      "step": 5871
+    },
+    {
+      "epoch": 0.481765598720105,
+      "grad_norm": 0.7418216912404008,
+      "learning_rate": 5.538408694273902e-06,
+      "loss": 0.3282,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4818476432702958,
+      "grad_norm": 0.6934666654962214,
+      "learning_rate": 5.5370876950714555e-06,
+      "loss": 0.3899,
+      "step": 5873
+    },
+    {
+      "epoch": 0.4819296878204865,
+      "grad_norm": 0.7334228867111794,
+      "learning_rate": 5.535766657940737e-06,
+      "loss": 0.3281,
+      "step": 5874
+    },
+    {
+      "epoch": 0.4820117323706773,
+      "grad_norm": 1.0688747931956895,
+      "learning_rate": 5.534445582975038e-06,
+      "loss": 0.3277,
+      "step": 5875
+    },
+    {
+      "epoch": 0.482093776920868,
+      "grad_norm": 0.6972486673632109,
+      "learning_rate": 5.53312447026765e-06,
+      "loss": 0.3554,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4821758214710588,
+      "grad_norm": 0.686434028121366,
+      "learning_rate": 5.531803319911868e-06,
+      "loss": 0.3091,
+      "step": 5877
+    },
+    {
+      "epoch": 0.48225786602124954,
+      "grad_norm": 1.3017686310936936,
+      "learning_rate": 5.530482132000991e-06,
+      "loss": 0.3454,
+      "step": 5878
+    },
+    {
+      "epoch": 0.48233991057144027,
+      "grad_norm": 0.6762384144058724,
+      "learning_rate": 5.529160906628318e-06,
+      "loss": 0.3926,
+      "step": 5879
+    },
+    {
+      "epoch": 0.48242195512163105,
+      "grad_norm": 0.7474371827719036,
+      "learning_rate": 5.5278396438871516e-06,
+      "loss": 0.322,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4825039996718218,
+      "grad_norm": 0.6838525117016092,
+      "learning_rate": 5.526518343870798e-06,
+      "loss": 0.309,
+      "step": 5881
+    },
+    {
+      "epoch": 0.48258604422201257,
+      "grad_norm": 0.6911322879658597,
+      "learning_rate": 5.525197006672565e-06,
+      "loss": 0.3199,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4826680887722033,
+      "grad_norm": 0.6640910159533756,
+      "learning_rate": 5.523875632385762e-06,
+      "loss": 0.3242,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4827501333223941,
+      "grad_norm": 0.7051032973983383,
+      "learning_rate": 5.522554221103705e-06,
+      "loss": 0.3282,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4828321778725848,
+      "grad_norm": 0.6869162977231653,
+      "learning_rate": 5.521232772919708e-06,
+      "loss": 0.3372,
+      "step": 5885
+    },
+    {
+      "epoch": 0.48291422242277554,
+      "grad_norm": 0.879541939780528,
+      "learning_rate": 5.519911287927089e-06,
+      "loss": 0.3368,
+      "step": 5886
+    },
+    {
+      "epoch": 0.4829962669729663,
+      "grad_norm": 0.7021724220981687,
+      "learning_rate": 5.518589766219173e-06,
+      "loss": 0.3092,
+      "step": 5887
+    },
+    {
+      "epoch": 0.48307831152315706,
+      "grad_norm": 0.6642643657634221,
+      "learning_rate": 5.517268207889278e-06,
+      "loss": 0.332,
+      "step": 5888
+    },
+    {
+      "epoch": 0.48316035607334784,
+      "grad_norm": 0.7098895361922243,
+      "learning_rate": 5.515946613030734e-06,
+      "loss": 0.3297,
+      "step": 5889
+    },
+    {
+      "epoch": 0.48324240062353857,
+      "grad_norm": 0.7526916630547397,
+      "learning_rate": 5.514624981736868e-06,
+      "loss": 0.3105,
+      "step": 5890
+    },
+    {
+      "epoch": 0.48332444517372936,
+      "grad_norm": 0.7446035483210831,
+      "learning_rate": 5.513303314101014e-06,
+      "loss": 0.3462,
+      "step": 5891
+    },
+    {
+      "epoch": 0.4834064897239201,
+      "grad_norm": 0.6721148830955618,
+      "learning_rate": 5.511981610216505e-06,
+      "loss": 0.3346,
+      "step": 5892
+    },
+    {
+      "epoch": 0.4834885342741108,
+      "grad_norm": 0.6791299065039735,
+      "learning_rate": 5.510659870176677e-06,
+      "loss": 0.3466,
+      "step": 5893
+    },
+    {
+      "epoch": 0.4835705788243016,
+      "grad_norm": 0.6835186766180341,
+      "learning_rate": 5.509338094074869e-06,
+      "loss": 0.3289,
+      "step": 5894
+    },
+    {
+      "epoch": 0.48365262337449233,
+      "grad_norm": 0.6944111435609946,
+      "learning_rate": 5.508016282004422e-06,
+      "loss": 0.3206,
+      "step": 5895
+    },
+    {
+      "epoch": 0.4837346679246831,
+      "grad_norm": 0.6597400884066319,
+      "learning_rate": 5.506694434058683e-06,
+      "loss": 0.2957,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48381671247487384,
+      "grad_norm": 0.6556973159083284,
+      "learning_rate": 5.505372550330997e-06,
+      "loss": 0.3239,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48389875702506463,
+      "grad_norm": 0.6709615202792696,
+      "learning_rate": 5.504050630914714e-06,
+      "loss": 0.3156,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48398080157525536,
+      "grad_norm": 0.6653448554622636,
+      "learning_rate": 5.5027286759031825e-06,
+      "loss": 0.3133,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48406284612544614,
+      "grad_norm": 0.6404608068769828,
+      "learning_rate": 5.501406685389762e-06,
+      "loss": 0.375,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4841448906756369,
+      "grad_norm": 0.6393399155251961,
+      "learning_rate": 5.500084659467807e-06,
+      "loss": 0.3295,
+      "step": 5901
+    },
+    {
+      "epoch": 0.4842269352258276,
+      "grad_norm": 0.6715515814150057,
+      "learning_rate": 5.4987625982306765e-06,
+      "loss": 0.3078,
+      "step": 5902
+    },
+    {
+      "epoch": 0.4843089797760184,
+      "grad_norm": 0.9020843978606343,
+      "learning_rate": 5.497440501771732e-06,
+      "loss": 0.3252,
+      "step": 5903
+    },
+    {
+      "epoch": 0.4843910243262091,
+      "grad_norm": 0.848469533981476,
+      "learning_rate": 5.496118370184339e-06,
+      "loss": 0.353,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4844730688763999,
+      "grad_norm": 0.8853815169686621,
+      "learning_rate": 5.4947962035618626e-06,
+      "loss": 0.3244,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48455511342659063,
+      "grad_norm": 0.6816405599103,
+      "learning_rate": 5.493474001997675e-06,
+      "loss": 0.344,
+      "step": 5906
+    },
+    {
+      "epoch": 0.4846371579767814,
+      "grad_norm": 0.6499035666270017,
+      "learning_rate": 5.4921517655851456e-06,
+      "loss": 0.355,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48471920252697215,
+      "grad_norm": 0.6871782261652233,
+      "learning_rate": 5.490829494417649e-06,
+      "loss": 0.3263,
+      "step": 5908
+    },
+    {
+      "epoch": 0.4848012470771629,
+      "grad_norm": 0.8548358096578865,
+      "learning_rate": 5.489507188588562e-06,
+      "loss": 0.3002,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48488329162735366,
+      "grad_norm": 0.7273205159467313,
+      "learning_rate": 5.488184848191265e-06,
+      "loss": 0.3347,
+      "step": 5910
+    },
+    {
+      "epoch": 0.4849653361775444,
+      "grad_norm": 0.7029351643126349,
+      "learning_rate": 5.486862473319137e-06,
+      "loss": 0.3407,
+      "step": 5911
+    },
+    {
+      "epoch": 0.4850473807277352,
+      "grad_norm": 0.7020731817113454,
+      "learning_rate": 5.485540064065563e-06,
+      "loss": 0.3269,
+      "step": 5912
+    },
+    {
+      "epoch": 0.4851294252779259,
+      "grad_norm": 0.7006233774815915,
+      "learning_rate": 5.48421762052393e-06,
+      "loss": 0.3602,
+      "step": 5913
+    },
+    {
+      "epoch": 0.4852114698281167,
+      "grad_norm": 0.6114954601302944,
+      "learning_rate": 5.482895142787626e-06,
+      "loss": 0.301,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4852935143783074,
+      "grad_norm": 0.670885207925429,
+      "learning_rate": 5.481572630950046e-06,
+      "loss": 0.3744,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48537555892849815,
+      "grad_norm": 0.6989005313306871,
+      "learning_rate": 5.480250085104578e-06,
+      "loss": 0.3327,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48545760347868894,
+      "grad_norm": 0.6759939145153354,
+      "learning_rate": 5.47892750534462e-06,
+      "loss": 0.3786,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48553964802887967,
+      "grad_norm": 0.7453123616951471,
+      "learning_rate": 5.477604891763573e-06,
+      "loss": 0.3225,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48562169257907045,
+      "grad_norm": 0.8373622799802783,
+      "learning_rate": 5.476282244454835e-06,
+      "loss": 0.3339,
+      "step": 5919
+    },
+    {
+      "epoch": 0.4857037371292612,
+      "grad_norm": 0.6536431992704383,
+      "learning_rate": 5.4749595635118095e-06,
+      "loss": 0.3734,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48578578167945197,
+      "grad_norm": 0.6208002677950261,
+      "learning_rate": 5.4736368490279035e-06,
+      "loss": 0.3177,
+      "step": 5921
+    },
+    {
+      "epoch": 0.4858678262296427,
+      "grad_norm": 0.6303830401580974,
+      "learning_rate": 5.472314101096523e-06,
+      "loss": 0.3508,
+      "step": 5922
+    },
+    {
+      "epoch": 0.4859498707798334,
+      "grad_norm": 0.860686955689994,
+      "learning_rate": 5.470991319811081e-06,
+      "loss": 0.3233,
+      "step": 5923
+    },
+    {
+      "epoch": 0.4860319153300242,
+      "grad_norm": 0.7058584164138638,
+      "learning_rate": 5.469668505264988e-06,
+      "loss": 0.3408,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48611395988021494,
+      "grad_norm": 0.6960648084882163,
+      "learning_rate": 5.468345657551659e-06,
+      "loss": 0.3257,
+      "step": 5925
+    },
+    {
+      "epoch": 0.4861960044304057,
+      "grad_norm": 0.6950950337237048,
+      "learning_rate": 5.467022776764512e-06,
+      "loss": 0.3633,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48627804898059646,
+      "grad_norm": 0.7283644322978563,
+      "learning_rate": 5.4656998629969685e-06,
+      "loss": 0.3764,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48636009353078724,
+      "grad_norm": 0.72475490794421,
+      "learning_rate": 5.464376916342447e-06,
+      "loss": 0.3174,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48644213808097797,
+      "grad_norm": 0.6742302293385076,
+      "learning_rate": 5.463053936894377e-06,
+      "loss": 0.3265,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4865241826311687,
+      "grad_norm": 0.6821001875094584,
+      "learning_rate": 5.461730924746181e-06,
+      "loss": 0.3421,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4866062271813595,
+      "grad_norm": 0.9026259083443373,
+      "learning_rate": 5.460407879991285e-06,
+      "loss": 0.3472,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4866882717315502,
+      "grad_norm": 0.7191333723644797,
+      "learning_rate": 5.459084802723128e-06,
+      "loss": 0.3015,
+      "step": 5932
+    },
+    {
+      "epoch": 0.486770316281741,
+      "grad_norm": 0.8367875688776854,
+      "learning_rate": 5.457761693035139e-06,
+      "loss": 0.3291,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48685236083193173,
+      "grad_norm": 0.712999670393838,
+      "learning_rate": 5.456438551020754e-06,
+      "loss": 0.3091,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4869344053821225,
+      "grad_norm": 0.6511819782837653,
+      "learning_rate": 5.455115376773412e-06,
+      "loss": 0.3512,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48701644993231324,
+      "grad_norm": 0.8562991141173595,
+      "learning_rate": 5.453792170386554e-06,
+      "loss": 0.3502,
+      "step": 5936
+    },
+    {
+      "epoch": 0.487098494482504,
+      "grad_norm": 0.6309868091011173,
+      "learning_rate": 5.452468931953622e-06,
+      "loss": 0.3086,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48718053903269476,
+      "grad_norm": 0.9311776174353888,
+      "learning_rate": 5.45114566156806e-06,
+      "loss": 0.3688,
+      "step": 5938
+    },
+    {
+      "epoch": 0.4872625835828855,
+      "grad_norm": 0.9003723581613301,
+      "learning_rate": 5.449822359323317e-06,
+      "loss": 0.3358,
+      "step": 5939
+    },
+    {
+      "epoch": 0.4873446281330763,
+      "grad_norm": 1.0572113387189237,
+      "learning_rate": 5.44849902531284e-06,
+      "loss": 0.312,
+      "step": 5940
+    },
+    {
+      "epoch": 0.487426672683267,
+      "grad_norm": 0.9263548033955961,
+      "learning_rate": 5.447175659630084e-06,
+      "loss": 0.3191,
+      "step": 5941
+    },
+    {
+      "epoch": 0.4875087172334578,
+      "grad_norm": 1.7929792260933854,
+      "learning_rate": 5.445852262368501e-06,
+      "loss": 0.3307,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4875907617836485,
+      "grad_norm": 0.6852857198901435,
+      "learning_rate": 5.444528833621549e-06,
+      "loss": 0.3412,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48767280633383925,
+      "grad_norm": 0.6981393990983255,
+      "learning_rate": 5.443205373482683e-06,
+      "loss": 0.3342,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48775485088403003,
+      "grad_norm": 0.9595926302983052,
+      "learning_rate": 5.441881882045366e-06,
+      "loss": 0.3375,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48783689543422076,
+      "grad_norm": 0.6792422498394335,
+      "learning_rate": 5.440558359403061e-06,
+      "loss": 0.3429,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48791893998441155,
+      "grad_norm": 0.7303786146464281,
+      "learning_rate": 5.439234805649232e-06,
+      "loss": 0.3149,
+      "step": 5947
+    },
+    {
+      "epoch": 0.4880009845346023,
+      "grad_norm": 0.7264856885020243,
+      "learning_rate": 5.437911220877346e-06,
+      "loss": 0.3182,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48808302908479306,
+      "grad_norm": 0.8591064176375152,
+      "learning_rate": 5.436587605180875e-06,
+      "loss": 0.3537,
+      "step": 5949
+    },
+    {
+      "epoch": 0.4881650736349838,
+      "grad_norm": 0.747396185425496,
+      "learning_rate": 5.435263958653287e-06,
+      "loss": 0.3361,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4882471181851745,
+      "grad_norm": 0.6936321637994561,
+      "learning_rate": 5.433940281388059e-06,
+      "loss": 0.3152,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4883291627353653,
+      "grad_norm": 0.7316760709677798,
+      "learning_rate": 5.432616573478665e-06,
+      "loss": 0.3399,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48841120728555604,
+      "grad_norm": 1.07565946485485,
+      "learning_rate": 5.4312928350185826e-06,
+      "loss": 0.3426,
+      "step": 5953
+    },
+    {
+      "epoch": 0.4884932518357468,
+      "grad_norm": 0.6438669454318987,
+      "learning_rate": 5.429969066101294e-06,
+      "loss": 0.3276,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48857529638593755,
+      "grad_norm": 0.6394467971435926,
+      "learning_rate": 5.428645266820281e-06,
+      "loss": 0.3163,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48865734093612834,
+      "grad_norm": 0.679423844398113,
+      "learning_rate": 5.427321437269027e-06,
+      "loss": 0.3226,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48873938548631907,
+      "grad_norm": 0.6868827336728175,
+      "learning_rate": 5.425997577541022e-06,
+      "loss": 0.275,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48882143003650985,
+      "grad_norm": 0.7780450410307794,
+      "learning_rate": 5.42467368772975e-06,
+      "loss": 0.3214,
+      "step": 5958
+    },
+    {
+      "epoch": 0.4889034745867006,
+      "grad_norm": 0.6668393994346512,
+      "learning_rate": 5.423349767928705e-06,
+      "loss": 0.3481,
+      "step": 5959
+    },
+    {
+      "epoch": 0.4889855191368913,
+      "grad_norm": 0.6641863159636754,
+      "learning_rate": 5.42202581823138e-06,
+      "loss": 0.3523,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4890675636870821,
+      "grad_norm": 0.7153028275308563,
+      "learning_rate": 5.4207018387312695e-06,
+      "loss": 0.3364,
+      "step": 5961
+    },
+    {
+      "epoch": 0.4891496082372728,
+      "grad_norm": 0.7092205182071456,
+      "learning_rate": 5.41937782952187e-06,
+      "loss": 0.3291,
+      "step": 5962
+    },
+    {
+      "epoch": 0.4892316527874636,
+      "grad_norm": 0.8007451602906069,
+      "learning_rate": 5.418053790696685e-06,
+      "loss": 0.2997,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48931369733765434,
+      "grad_norm": 0.666871152415175,
+      "learning_rate": 5.41672972234921e-06,
+      "loss": 0.3428,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4893957418878451,
+      "grad_norm": 0.8906622876586006,
+      "learning_rate": 5.415405624572953e-06,
+      "loss": 0.3542,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48947778643803586,
+      "grad_norm": 0.6123522129579323,
+      "learning_rate": 5.4140814974614175e-06,
+      "loss": 0.3435,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4895598309882266,
+      "grad_norm": 0.9263203127455067,
+      "learning_rate": 5.412757341108111e-06,
+      "loss": 0.3283,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48964187553841737,
+      "grad_norm": 0.6919976046941833,
+      "learning_rate": 5.4114331556065455e-06,
+      "loss": 0.3652,
+      "step": 5968
+    },
+    {
+      "epoch": 0.4897239200886081,
+      "grad_norm": 0.6362444826933681,
+      "learning_rate": 5.410108941050232e-06,
+      "loss": 0.3371,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4898059646387989,
+      "grad_norm": 0.7361893077733512,
+      "learning_rate": 5.4087846975326826e-06,
+      "loss": 0.2928,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4898880091889896,
+      "grad_norm": 0.6638887859206367,
+      "learning_rate": 5.407460425147416e-06,
+      "loss": 0.3301,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4899700537391804,
+      "grad_norm": 0.6296255596777681,
+      "learning_rate": 5.406136123987948e-06,
+      "loss": 0.3398,
+      "step": 5972
+    },
+    {
+      "epoch": 0.49005209828937113,
+      "grad_norm": 0.7566191332273655,
+      "learning_rate": 5.404811794147801e-06,
+      "loss": 0.3445,
+      "step": 5973
+    },
+    {
+      "epoch": 0.49013414283956186,
+      "grad_norm": 0.6308876097661891,
+      "learning_rate": 5.403487435720493e-06,
+      "loss": 0.3201,
+      "step": 5974
+    },
+    {
+      "epoch": 0.49021618738975264,
+      "grad_norm": 0.7961837940172054,
+      "learning_rate": 5.402163048799553e-06,
+      "loss": 0.3183,
+      "step": 5975
+    },
+    {
+      "epoch": 0.4902982319399434,
+      "grad_norm": 1.2456270148612139,
+      "learning_rate": 5.400838633478502e-06,
+      "loss": 0.3384,
+      "step": 5976
+    },
+    {
+      "epoch": 0.49038027649013416,
+      "grad_norm": 0.6237004484913997,
+      "learning_rate": 5.399514189850873e-06,
+      "loss": 0.3142,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4904623210403249,
+      "grad_norm": 0.7043272324007849,
+      "learning_rate": 5.398189718010193e-06,
+      "loss": 0.3045,
+      "step": 5978
+    },
+    {
+      "epoch": 0.4905443655905157,
+      "grad_norm": 0.6931562139930062,
+      "learning_rate": 5.396865218049995e-06,
+      "loss": 0.3334,
+      "step": 5979
+    },
+    {
+      "epoch": 0.4906264101407064,
+      "grad_norm": 0.6762414404526099,
+      "learning_rate": 5.395540690063811e-06,
+      "loss": 0.3479,
+      "step": 5980
+    },
+    {
+      "epoch": 0.49070845469089713,
+      "grad_norm": 0.6066276635957596,
+      "learning_rate": 5.39421613414518e-06,
+      "loss": 0.2953,
+      "step": 5981
+    },
+    {
+      "epoch": 0.4907904992410879,
+      "grad_norm": 0.6695412709247373,
+      "learning_rate": 5.39289155038764e-06,
+      "loss": 0.3545,
+      "step": 5982
+    },
+    {
+      "epoch": 0.49087254379127865,
+      "grad_norm": 0.6784075495552226,
+      "learning_rate": 5.3915669388847305e-06,
+      "loss": 0.3228,
+      "step": 5983
+    },
+    {
+      "epoch": 0.49095458834146943,
+      "grad_norm": 0.6353730754061591,
+      "learning_rate": 5.390242299729991e-06,
+      "loss": 0.3234,
+      "step": 5984
+    },
+    {
+      "epoch": 0.49103663289166016,
+      "grad_norm": 0.6869652184640813,
+      "learning_rate": 5.388917633016968e-06,
+      "loss": 0.3452,
+      "step": 5985
+    },
+    {
+      "epoch": 0.49111867744185095,
+      "grad_norm": 0.6333295978881687,
+      "learning_rate": 5.387592938839207e-06,
+      "loss": 0.2839,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4912007219920417,
+      "grad_norm": 0.7424279641263108,
+      "learning_rate": 5.386268217290256e-06,
+      "loss": 0.3754,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4912827665422324,
+      "grad_norm": 0.791712620397989,
+      "learning_rate": 5.384943468463663e-06,
+      "loss": 0.3714,
+      "step": 5988
+    },
+    {
+      "epoch": 0.4913648110924232,
+      "grad_norm": 0.6742295628890763,
+      "learning_rate": 5.383618692452979e-06,
+      "loss": 0.35,
+      "step": 5989
+    },
+    {
+      "epoch": 0.4914468556426139,
+      "grad_norm": 0.6275968551617577,
+      "learning_rate": 5.382293889351762e-06,
+      "loss": 0.3309,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4915289001928047,
+      "grad_norm": 0.6600562515652123,
+      "learning_rate": 5.380969059253565e-06,
+      "loss": 0.401,
+      "step": 5991
+    },
+    {
+      "epoch": 0.49161094474299544,
+      "grad_norm": 0.6837979519144922,
+      "learning_rate": 5.379644202251943e-06,
+      "loss": 0.3613,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4916929892931862,
+      "grad_norm": 0.7230604063156064,
+      "learning_rate": 5.378319318440458e-06,
+      "loss": 0.3221,
+      "step": 5993
+    },
+    {
+      "epoch": 0.49177503384337695,
+      "grad_norm": 0.6651753322637123,
+      "learning_rate": 5.376994407912671e-06,
+      "loss": 0.3156,
+      "step": 5994
+    },
+    {
+      "epoch": 0.4918570783935677,
+      "grad_norm": 0.6181499771255744,
+      "learning_rate": 5.375669470762145e-06,
+      "loss": 0.3165,
+      "step": 5995
+    },
+    {
+      "epoch": 0.49193912294375847,
+      "grad_norm": 0.6357819407552753,
+      "learning_rate": 5.374344507082444e-06,
+      "loss": 0.3436,
+      "step": 5996
+    },
+    {
+      "epoch": 0.4920211674939492,
+      "grad_norm": 0.7150247881569385,
+      "learning_rate": 5.373019516967135e-06,
+      "loss": 0.3548,
+      "step": 5997
+    },
+    {
+      "epoch": 0.49210321204414,
+      "grad_norm": 0.7155639376005716,
+      "learning_rate": 5.371694500509787e-06,
+      "loss": 0.3496,
+      "step": 5998
+    },
+    {
+      "epoch": 0.4921852565943307,
+      "grad_norm": 0.6764483861459452,
+      "learning_rate": 5.370369457803971e-06,
+      "loss": 0.3601,
+      "step": 5999
+    },
+    {
+      "epoch": 0.4922673011445215,
+      "grad_norm": 0.8160193188475613,
+      "learning_rate": 5.3690443889432595e-06,
+      "loss": 0.3279,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4923493456947122,
+      "grad_norm": 0.7434928325330301,
+      "learning_rate": 5.367719294021225e-06,
+      "loss": 0.3345,
+      "step": 6001
+    },
+    {
+      "epoch": 0.49243139024490296,
+      "grad_norm": 0.6335670062549674,
+      "learning_rate": 5.366394173131445e-06,
+      "loss": 0.3089,
+      "step": 6002
+    },
+    {
+      "epoch": 0.49251343479509374,
+      "grad_norm": 0.8943111010896452,
+      "learning_rate": 5.365069026367498e-06,
+      "loss": 0.3449,
+      "step": 6003
+    },
+    {
+      "epoch": 0.49259547934528447,
+      "grad_norm": 0.6281513543899186,
+      "learning_rate": 5.363743853822962e-06,
+      "loss": 0.2912,
+      "step": 6004
+    },
+    {
+      "epoch": 0.49267752389547526,
+      "grad_norm": 0.8872501393460088,
+      "learning_rate": 5.3624186555914205e-06,
+      "loss": 0.3822,
+      "step": 6005
+    },
+    {
+      "epoch": 0.492759568445666,
+      "grad_norm": 0.6798766969777158,
+      "learning_rate": 5.361093431766455e-06,
+      "loss": 0.3354,
+      "step": 6006
+    },
+    {
+      "epoch": 0.49284161299585677,
+      "grad_norm": 0.7502490522867507,
+      "learning_rate": 5.3597681824416505e-06,
+      "loss": 0.3282,
+      "step": 6007
+    },
+    {
+      "epoch": 0.4929236575460475,
+      "grad_norm": 0.6931439827934208,
+      "learning_rate": 5.3584429077105955e-06,
+      "loss": 0.3571,
+      "step": 6008
+    },
+    {
+      "epoch": 0.49300570209623823,
+      "grad_norm": 0.790713187084373,
+      "learning_rate": 5.35711760766688e-06,
+      "loss": 0.3238,
+      "step": 6009
+    },
+    {
+      "epoch": 0.493087746646429,
+      "grad_norm": 0.8015063182705268,
+      "learning_rate": 5.355792282404091e-06,
+      "loss": 0.3426,
+      "step": 6010
+    },
+    {
+      "epoch": 0.49316979119661974,
+      "grad_norm": 0.7555985162589158,
+      "learning_rate": 5.3544669320158235e-06,
+      "loss": 0.3506,
+      "step": 6011
+    },
+    {
+      "epoch": 0.49325183574681053,
+      "grad_norm": 0.7224424865220608,
+      "learning_rate": 5.35314155659567e-06,
+      "loss": 0.3349,
+      "step": 6012
+    },
+    {
+      "epoch": 0.49333388029700126,
+      "grad_norm": 0.6535880746583861,
+      "learning_rate": 5.351816156237228e-06,
+      "loss": 0.3392,
+      "step": 6013
+    },
+    {
+      "epoch": 0.49341592484719204,
+      "grad_norm": 0.7131664715374354,
+      "learning_rate": 5.350490731034093e-06,
+      "loss": 0.3306,
+      "step": 6014
+    },
+    {
+      "epoch": 0.4934979693973828,
+      "grad_norm": 0.6359298459988126,
+      "learning_rate": 5.349165281079868e-06,
+      "loss": 0.3237,
+      "step": 6015
+    },
+    {
+      "epoch": 0.49358001394757356,
+      "grad_norm": 0.7492323854568221,
+      "learning_rate": 5.347839806468149e-06,
+      "loss": 0.3374,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4936620584977643,
+      "grad_norm": 0.6952182991216241,
+      "learning_rate": 5.346514307292543e-06,
+      "loss": 0.3747,
+      "step": 6017
+    },
+    {
+      "epoch": 0.493744103047955,
+      "grad_norm": 0.6790413177039417,
+      "learning_rate": 5.345188783646654e-06,
+      "loss": 0.3366,
+      "step": 6018
+    },
+    {
+      "epoch": 0.4938261475981458,
+      "grad_norm": 0.7100749776983348,
+      "learning_rate": 5.343863235624088e-06,
+      "loss": 0.3563,
+      "step": 6019
+    },
+    {
+      "epoch": 0.49390819214833653,
+      "grad_norm": 0.6670286443983946,
+      "learning_rate": 5.342537663318451e-06,
+      "loss": 0.3169,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4939902366985273,
+      "grad_norm": 0.8260964332137584,
+      "learning_rate": 5.341212066823356e-06,
+      "loss": 0.3216,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49407228124871805,
+      "grad_norm": 0.6363881917118904,
+      "learning_rate": 5.339886446232412e-06,
+      "loss": 0.2935,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49415432579890883,
+      "grad_norm": 0.7044754723806042,
+      "learning_rate": 5.338560801639234e-06,
+      "loss": 0.3475,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49423637034909956,
+      "grad_norm": 0.7695122932742788,
+      "learning_rate": 5.337235133137436e-06,
+      "loss": 0.3749,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4943184148992903,
+      "grad_norm": 0.6463844295631681,
+      "learning_rate": 5.335909440820635e-06,
+      "loss": 0.3241,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4944004594494811,
+      "grad_norm": 0.7110365207689999,
+      "learning_rate": 5.334583724782449e-06,
+      "loss": 0.3307,
+      "step": 6026
+    },
+    {
+      "epoch": 0.4944825039996718,
+      "grad_norm": 0.7488808193618336,
+      "learning_rate": 5.3332579851164965e-06,
+      "loss": 0.3584,
+      "step": 6027
+    },
+    {
+      "epoch": 0.4945645485498626,
+      "grad_norm": 0.8584224193852777,
+      "learning_rate": 5.331932221916401e-06,
+      "loss": 0.3467,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4946465931000533,
+      "grad_norm": 0.6370878492845486,
+      "learning_rate": 5.330606435275785e-06,
+      "loss": 0.3598,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4947286376502441,
+      "grad_norm": 0.6633770087693207,
+      "learning_rate": 5.329280625288274e-06,
+      "loss": 0.3565,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49481068220043484,
+      "grad_norm": 0.6111294751533995,
+      "learning_rate": 5.327954792047495e-06,
+      "loss": 0.301,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49489272675062557,
+      "grad_norm": 0.6279158939069744,
+      "learning_rate": 5.326628935647076e-06,
+      "loss": 0.3212,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49497477130081635,
+      "grad_norm": 0.6738451226721919,
+      "learning_rate": 5.3253030561806464e-06,
+      "loss": 0.3208,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4950568158510071,
+      "grad_norm": 0.760700321997078,
+      "learning_rate": 5.323977153741835e-06,
+      "loss": 0.33,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49513886040119787,
+      "grad_norm": 0.8221194738493767,
+      "learning_rate": 5.322651228424279e-06,
+      "loss": 0.3201,
+      "step": 6035
+    },
+    {
+      "epoch": 0.4952209049513886,
+      "grad_norm": 0.6850611954102807,
+      "learning_rate": 5.321325280321613e-06,
+      "loss": 0.3229,
+      "step": 6036
+    },
+    {
+      "epoch": 0.4953029495015794,
+      "grad_norm": 0.6375108284043189,
+      "learning_rate": 5.31999930952747e-06,
+      "loss": 0.344,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4953849940517701,
+      "grad_norm": 0.6556045785511098,
+      "learning_rate": 5.3186733161354895e-06,
+      "loss": 0.3061,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49546703860196084,
+      "grad_norm": 0.691163575082186,
+      "learning_rate": 5.317347300239314e-06,
+      "loss": 0.3571,
+      "step": 6039
+    },
+    {
+      "epoch": 0.4955490831521516,
+      "grad_norm": 0.7378008109128339,
+      "learning_rate": 5.316021261932581e-06,
+      "loss": 0.2941,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49563112770234236,
+      "grad_norm": 0.9292790021999193,
+      "learning_rate": 5.314695201308934e-06,
+      "loss": 0.3237,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49571317225253314,
+      "grad_norm": 0.6124287426459596,
+      "learning_rate": 5.313369118462017e-06,
+      "loss": 0.3354,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49579521680272387,
+      "grad_norm": 0.7187552392399367,
+      "learning_rate": 5.312043013485476e-06,
+      "loss": 0.3311,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49587726135291466,
+      "grad_norm": 1.077032724995278,
+      "learning_rate": 5.31071688647296e-06,
+      "loss": 0.3156,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4959593059031054,
+      "grad_norm": 0.7838118918770594,
+      "learning_rate": 5.309390737518115e-06,
+      "loss": 0.3155,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4960413504532961,
+      "grad_norm": 0.6868417580825388,
+      "learning_rate": 5.308064566714595e-06,
+      "loss": 0.363,
+      "step": 6046
+    },
+    {
+      "epoch": 0.4961233950034869,
+      "grad_norm": 0.6960748943465731,
+      "learning_rate": 5.30673837415605e-06,
+      "loss": 0.3402,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49620543955367763,
+      "grad_norm": 0.6476635917157717,
+      "learning_rate": 5.305412159936133e-06,
+      "loss": 0.3004,
+      "step": 6048
+    },
+    {
+      "epoch": 0.4962874841038684,
+      "grad_norm": 0.6425369787855844,
+      "learning_rate": 5.3040859241485e-06,
+      "loss": 0.3004,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49636952865405914,
+      "grad_norm": 0.6373243866723795,
+      "learning_rate": 5.302759666886807e-06,
+      "loss": 0.2777,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49645157320424993,
+      "grad_norm": 0.6353924235572437,
+      "learning_rate": 5.301433388244715e-06,
+      "loss": 0.3521,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49653361775444066,
+      "grad_norm": 0.7039078644660405,
+      "learning_rate": 5.300107088315878e-06,
+      "loss": 0.3887,
+      "step": 6052
+    },
+    {
+      "epoch": 0.4966156623046314,
+      "grad_norm": 0.6733312497511752,
+      "learning_rate": 5.298780767193964e-06,
+      "loss": 0.2903,
+      "step": 6053
+    },
+    {
+      "epoch": 0.4966977068548222,
+      "grad_norm": 0.6115749732579427,
+      "learning_rate": 5.297454424972631e-06,
+      "loss": 0.3258,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4967797514050129,
+      "grad_norm": 0.8446017076890843,
+      "learning_rate": 5.296128061745545e-06,
+      "loss": 0.4006,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4968617959552037,
+      "grad_norm": 0.686798132379945,
+      "learning_rate": 5.294801677606369e-06,
+      "loss": 0.3393,
+      "step": 6056
+    },
+    {
+      "epoch": 0.4969438405053944,
+      "grad_norm": 0.6349385925853959,
+      "learning_rate": 5.2934752726487725e-06,
+      "loss": 0.3137,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4970258850555852,
+      "grad_norm": 0.6350511441674974,
+      "learning_rate": 5.292148846966425e-06,
+      "loss": 0.3416,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49710792960577593,
+      "grad_norm": 0.6640242806455906,
+      "learning_rate": 5.290822400652995e-06,
+      "loss": 0.3311,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49718997415596666,
+      "grad_norm": 0.6768722331842102,
+      "learning_rate": 5.289495933802155e-06,
+      "loss": 0.3434,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49727201870615745,
+      "grad_norm": 0.6645244591261519,
+      "learning_rate": 5.288169446507576e-06,
+      "loss": 0.333,
+      "step": 6061
+    },
+    {
+      "epoch": 0.4973540632563482,
+      "grad_norm": 0.7775140509106834,
+      "learning_rate": 5.286842938862934e-06,
+      "loss": 0.3452,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49743610780653896,
+      "grad_norm": 0.646546074532065,
+      "learning_rate": 5.285516410961905e-06,
+      "loss": 0.3732,
+      "step": 6063
+    },
+    {
+      "epoch": 0.4975181523567297,
+      "grad_norm": 0.6508937368436576,
+      "learning_rate": 5.284189862898166e-06,
+      "loss": 0.3498,
+      "step": 6064
+    },
+    {
+      "epoch": 0.4976001969069205,
+      "grad_norm": 0.6286957009574038,
+      "learning_rate": 5.282863294765393e-06,
+      "loss": 0.3146,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4976822414571112,
+      "grad_norm": 0.6457312570015231,
+      "learning_rate": 5.281536706657271e-06,
+      "loss": 0.6718,
+      "step": 6066
+    },
+    {
+      "epoch": 0.497764286007302,
+      "grad_norm": 0.6606843219297923,
+      "learning_rate": 5.28021009866748e-06,
+      "loss": 0.3529,
+      "step": 6067
+    },
+    {
+      "epoch": 0.4978463305574927,
+      "grad_norm": 0.6759966810409561,
+      "learning_rate": 5.2788834708897005e-06,
+      "loss": 0.3671,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49792837510768345,
+      "grad_norm": 0.7135059270032462,
+      "learning_rate": 5.277556823417619e-06,
+      "loss": 0.3436,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49801041965787424,
+      "grad_norm": 1.0043934256061646,
+      "learning_rate": 5.2762301563449195e-06,
+      "loss": 0.2979,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49809246420806497,
+      "grad_norm": 0.6689288599105845,
+      "learning_rate": 5.27490346976529e-06,
+      "loss": 0.3467,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49817450875825575,
+      "grad_norm": 0.7216974845186587,
+      "learning_rate": 5.27357676377242e-06,
+      "loss": 0.36,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4982565533084465,
+      "grad_norm": 0.6752512900699371,
+      "learning_rate": 5.272250038459999e-06,
+      "loss": 0.3358,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49833859785863727,
+      "grad_norm": 0.6881755531452278,
+      "learning_rate": 5.270923293921716e-06,
+      "loss": 0.3421,
+      "step": 6074
+    },
+    {
+      "epoch": 0.498420642408828,
+      "grad_norm": 0.6706942321771636,
+      "learning_rate": 5.269596530251267e-06,
+      "loss": 0.3106,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4985026869590187,
+      "grad_norm": 0.65081123660065,
+      "learning_rate": 5.268269747542343e-06,
+      "loss": 0.3386,
+      "step": 6076
+    },
+    {
+      "epoch": 0.4985847315092095,
+      "grad_norm": 1.0453441021184628,
+      "learning_rate": 5.2669429458886415e-06,
+      "loss": 0.3632,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49866677605940024,
+      "grad_norm": 0.6282639210651904,
+      "learning_rate": 5.265616125383857e-06,
+      "loss": 0.3454,
+      "step": 6078
+    },
+    {
+      "epoch": 0.498748820609591,
+      "grad_norm": 0.6043044372036198,
+      "learning_rate": 5.264289286121687e-06,
+      "loss": 0.2814,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49883086515978176,
+      "grad_norm": 0.7034529729790224,
+      "learning_rate": 5.262962428195835e-06,
+      "loss": 0.3288,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49891290970997254,
+      "grad_norm": 0.7010725495055713,
+      "learning_rate": 5.2616355516999995e-06,
+      "loss": 0.3389,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49899495426016327,
+      "grad_norm": 0.612323161141405,
+      "learning_rate": 5.2603086567278795e-06,
+      "loss": 0.3071,
+      "step": 6082
+    },
+    {
+      "epoch": 0.499076998810354,
+      "grad_norm": 0.6226115477420574,
+      "learning_rate": 5.25898174337318e-06,
+      "loss": 0.3294,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4991590433605448,
+      "grad_norm": 0.8545483617258917,
+      "learning_rate": 5.257654811729606e-06,
+      "loss": 0.3317,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4992410879107355,
+      "grad_norm": 0.7164052387965758,
+      "learning_rate": 5.256327861890865e-06,
+      "loss": 0.3168,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4993231324609263,
+      "grad_norm": 0.8476147705197375,
+      "learning_rate": 5.255000893950662e-06,
+      "loss": 0.3114,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49940517701111703,
+      "grad_norm": 0.810483009763398,
+      "learning_rate": 5.2536739080027046e-06,
+      "loss": 0.3489,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4994872215613078,
+      "grad_norm": 0.6383311921764868,
+      "learning_rate": 5.252346904140704e-06,
+      "loss": 0.317,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49956926611149854,
+      "grad_norm": 1.002087390856705,
+      "learning_rate": 5.2510198824583695e-06,
+      "loss": 0.3349,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4996513106616893,
+      "grad_norm": 0.606586052818834,
+      "learning_rate": 5.2496928430494145e-06,
+      "loss": 0.3452,
+      "step": 6090
+    },
+    {
+      "epoch": 0.49973335521188006,
+      "grad_norm": 0.7224670415550207,
+      "learning_rate": 5.248365786007552e-06,
+      "loss": 0.3315,
+      "step": 6091
+    },
+    {
+      "epoch": 0.4998153997620708,
+      "grad_norm": 0.7267695598330209,
+      "learning_rate": 5.247038711426497e-06,
+      "loss": 0.3436,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4998974443122616,
+      "grad_norm": 0.6721029504995465,
+      "learning_rate": 5.245711619399965e-06,
+      "loss": 0.3165,
+      "step": 6093
+    },
+    {
+      "epoch": 0.4999794888624523,
+      "grad_norm": 0.5530515338948787,
+      "learning_rate": 5.244384510021673e-06,
+      "loss": 0.3014,
+      "step": 6094
+    },
+    {
+      "epoch": 0.500061533412643,
+      "grad_norm": 0.7304962526315347,
+      "learning_rate": 5.2430573833853415e-06,
+      "loss": 0.3249,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5001435779628338,
+      "grad_norm": 0.6809740755112174,
+      "learning_rate": 5.241730239584686e-06,
+      "loss": 0.3327,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5002256225130246,
+      "grad_norm": 0.6863533613800127,
+      "learning_rate": 5.240403078713429e-06,
+      "loss": 0.3257,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5003076670632153,
+      "grad_norm": 0.6644444321415502,
+      "learning_rate": 5.239075900865294e-06,
+      "loss": 0.3237,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5003897116134061,
+      "grad_norm": 0.6651939029240997,
+      "learning_rate": 5.2377487061340024e-06,
+      "loss": 0.3238,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5004717561635968,
+      "grad_norm": 0.7601981224910895,
+      "learning_rate": 5.23642149461328e-06,
+      "loss": 0.3755,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5005538007137876,
+      "grad_norm": 0.6735649494984945,
+      "learning_rate": 5.2350942663968496e-06,
+      "loss": 0.3276,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5006358452639783,
+      "grad_norm": 0.8946368250686952,
+      "learning_rate": 5.233767021578442e-06,
+      "loss": 0.3105,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5007178898141691,
+      "grad_norm": 0.6717597790159182,
+      "learning_rate": 5.232439760251782e-06,
+      "loss": 0.324,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5007999343643599,
+      "grad_norm": 0.6796383681163426,
+      "learning_rate": 5.2311124825106e-06,
+      "loss": 0.3095,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5008819789145506,
+      "grad_norm": 0.7823821330063024,
+      "learning_rate": 5.2297851884486254e-06,
+      "loss": 0.3466,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5009640234647413,
+      "grad_norm": 1.1330972701635542,
+      "learning_rate": 5.2284578781595885e-06,
+      "loss": 0.3372,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5010460680149321,
+      "grad_norm": 0.6431825378604005,
+      "learning_rate": 5.227130551737224e-06,
+      "loss": 0.3162,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5011281125651229,
+      "grad_norm": 0.7269666222040047,
+      "learning_rate": 5.225803209275266e-06,
+      "loss": 0.2962,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5012101571153136,
+      "grad_norm": 0.7105314655865291,
+      "learning_rate": 5.224475850867449e-06,
+      "loss": 0.338,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5012922016655044,
+      "grad_norm": 0.6877798325762804,
+      "learning_rate": 5.223148476607506e-06,
+      "loss": 0.3407,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5013742462156952,
+      "grad_norm": 0.971232972345301,
+      "learning_rate": 5.221821086589176e-06,
+      "loss": 0.343,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5014562907658858,
+      "grad_norm": 0.7262258997798167,
+      "learning_rate": 5.2204936809061975e-06,
+      "loss": 0.3254,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5015383353160766,
+      "grad_norm": 0.6388258225955606,
+      "learning_rate": 5.219166259652311e-06,
+      "loss": 0.3207,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5016203798662674,
+      "grad_norm": 0.6673940797577163,
+      "learning_rate": 5.217838822921255e-06,
+      "loss": 0.3249,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5017024244164582,
+      "grad_norm": 0.6802134076996373,
+      "learning_rate": 5.21651137080677e-06,
+      "loss": 0.2824,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5017844689666489,
+      "grad_norm": 0.6749672700726236,
+      "learning_rate": 5.215183903402601e-06,
+      "loss": 0.3092,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5018665135168396,
+      "grad_norm": 0.6427560518181888,
+      "learning_rate": 5.21385642080249e-06,
+      "loss": 0.2893,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5019485580670304,
+      "grad_norm": 0.6187419348946307,
+      "learning_rate": 5.212528923100183e-06,
+      "loss": 0.3318,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5020306026172211,
+      "grad_norm": 0.6597933244129496,
+      "learning_rate": 5.2112014103894226e-06,
+      "loss": 0.3029,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5021126471674119,
+      "grad_norm": 0.6134542537933947,
+      "learning_rate": 5.209873882763959e-06,
+      "loss": 0.3144,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5021946917176027,
+      "grad_norm": 0.6581416592121989,
+      "learning_rate": 5.208546340317539e-06,
+      "loss": 0.3388,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5022767362677935,
+      "grad_norm": 0.671938226723893,
+      "learning_rate": 5.2072187831439125e-06,
+      "loss": 0.3539,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5023587808179841,
+      "grad_norm": 0.607841869863722,
+      "learning_rate": 5.20589121133683e-06,
+      "loss": 0.2956,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5024408253681749,
+      "grad_norm": 0.9811062177914502,
+      "learning_rate": 5.204563624990038e-06,
+      "loss": 0.3664,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5025228699183657,
+      "grad_norm": 0.6680250200906155,
+      "learning_rate": 5.2032360241972916e-06,
+      "loss": 0.2916,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5026049144685564,
+      "grad_norm": 0.6361794479100825,
+      "learning_rate": 5.2019084090523445e-06,
+      "loss": 0.3366,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5026869590187472,
+      "grad_norm": 0.6635264089343025,
+      "learning_rate": 5.20058077964895e-06,
+      "loss": 0.3612,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5027690035689379,
+      "grad_norm": 0.8000638880623595,
+      "learning_rate": 5.199253136080862e-06,
+      "loss": 0.343,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5028510481191287,
+      "grad_norm": 0.6817371977583552,
+      "learning_rate": 5.19792547844184e-06,
+      "loss": 0.3286,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5029330926693194,
+      "grad_norm": 0.6885598833452434,
+      "learning_rate": 5.196597806825637e-06,
+      "loss": 0.3014,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5030151372195102,
+      "grad_norm": 0.7114739252013976,
+      "learning_rate": 5.195270121326012e-06,
+      "loss": 0.3224,
+      "step": 6131
+    },
+    {
+      "epoch": 0.503097181769701,
+      "grad_norm": 0.7254350535657679,
+      "learning_rate": 5.193942422036726e-06,
+      "loss": 0.3552,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5031792263198916,
+      "grad_norm": 0.7806033168023406,
+      "learning_rate": 5.192614709051538e-06,
+      "loss": 0.3452,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5032612708700824,
+      "grad_norm": 0.7238003794760175,
+      "learning_rate": 5.191286982464207e-06,
+      "loss": 0.3847,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5033433154202732,
+      "grad_norm": 0.6336882295382987,
+      "learning_rate": 5.189959242368499e-06,
+      "loss": 0.3659,
+      "step": 6135
+    },
+    {
+      "epoch": 0.503425359970464,
+      "grad_norm": 0.752440758848439,
+      "learning_rate": 5.188631488858173e-06,
+      "loss": 0.3092,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5035074045206547,
+      "grad_norm": 0.7384322382732299,
+      "learning_rate": 5.187303722026995e-06,
+      "loss": 0.3293,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5035894490708455,
+      "grad_norm": 0.7672182394032006,
+      "learning_rate": 5.1859759419687286e-06,
+      "loss": 0.3287,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5036714936210362,
+      "grad_norm": 0.6849687048215778,
+      "learning_rate": 5.184648148777139e-06,
+      "loss": 0.3323,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5037535381712269,
+      "grad_norm": 0.6765254134421824,
+      "learning_rate": 5.183320342545995e-06,
+      "loss": 0.3428,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5038355827214177,
+      "grad_norm": 0.7376746911460872,
+      "learning_rate": 5.181992523369064e-06,
+      "loss": 0.3526,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5039176272716085,
+      "grad_norm": 0.6712988919490659,
+      "learning_rate": 5.180664691340112e-06,
+      "loss": 0.3248,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5039996718217993,
+      "grad_norm": 0.6331452828502557,
+      "learning_rate": 5.179336846552911e-06,
+      "loss": 0.291,
+      "step": 6143
+    },
+    {
+      "epoch": 0.50408171637199,
+      "grad_norm": 0.6468665051261759,
+      "learning_rate": 5.178008989101229e-06,
+      "loss": 0.2954,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5041637609221807,
+      "grad_norm": 0.6748069814018698,
+      "learning_rate": 5.176681119078838e-06,
+      "loss": 0.3292,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5042458054723715,
+      "grad_norm": 1.0139923215753477,
+      "learning_rate": 5.175353236579509e-06,
+      "loss": 0.3697,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5043278500225622,
+      "grad_norm": 0.6394764480785601,
+      "learning_rate": 5.174025341697018e-06,
+      "loss": 0.2978,
+      "step": 6147
+    },
+    {
+      "epoch": 0.504409894572753,
+      "grad_norm": 0.6454529008929889,
+      "learning_rate": 5.172697434525136e-06,
+      "loss": 0.3038,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5044919391229438,
+      "grad_norm": 0.7035923675926891,
+      "learning_rate": 5.171369515157639e-06,
+      "loss": 0.3219,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5045739836731346,
+      "grad_norm": 0.6229819244659677,
+      "learning_rate": 5.1700415836883034e-06,
+      "loss": 0.3329,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5046560282233252,
+      "grad_norm": 0.7576197160093016,
+      "learning_rate": 5.168713640210903e-06,
+      "loss": 0.328,
+      "step": 6151
+    },
+    {
+      "epoch": 0.504738072773516,
+      "grad_norm": 0.9278884792271498,
+      "learning_rate": 5.167385684819216e-06,
+      "loss": 0.3364,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5048201173237068,
+      "grad_norm": 0.6605430298094092,
+      "learning_rate": 5.166057717607023e-06,
+      "loss": 0.2881,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5049021618738976,
+      "grad_norm": 0.6708160407303695,
+      "learning_rate": 5.164729738668098e-06,
+      "loss": 0.3363,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5049842064240883,
+      "grad_norm": 0.679088197898457,
+      "learning_rate": 5.163401748096224e-06,
+      "loss": 0.2832,
+      "step": 6155
+    },
+    {
+      "epoch": 0.505066250974279,
+      "grad_norm": 0.6301317710988825,
+      "learning_rate": 5.162073745985182e-06,
+      "loss": 0.3135,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5051482955244698,
+      "grad_norm": 0.610203024306466,
+      "learning_rate": 5.1607457324287535e-06,
+      "loss": 0.307,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5052303400746605,
+      "grad_norm": 0.6311895412556618,
+      "learning_rate": 5.159417707520719e-06,
+      "loss": 0.3189,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5053123846248513,
+      "grad_norm": 0.6576134376547381,
+      "learning_rate": 5.158089671354861e-06,
+      "loss": 0.33,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5053944291750421,
+      "grad_norm": 0.5872235571849912,
+      "learning_rate": 5.1567616240249655e-06,
+      "loss": 0.3368,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5054764737252329,
+      "grad_norm": 0.7507417306661818,
+      "learning_rate": 5.155433565624814e-06,
+      "loss": 0.3559,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5055585182754235,
+      "grad_norm": 0.7007223960328275,
+      "learning_rate": 5.154105496248196e-06,
+      "loss": 0.3499,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5056405628256143,
+      "grad_norm": 0.6541155434215719,
+      "learning_rate": 5.152777415988894e-06,
+      "loss": 0.3079,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5057226073758051,
+      "grad_norm": 0.7279026130041839,
+      "learning_rate": 5.151449324940698e-06,
+      "loss": 0.3095,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5058046519259958,
+      "grad_norm": 0.7815473601294264,
+      "learning_rate": 5.150121223197393e-06,
+      "loss": 0.32,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5058866964761866,
+      "grad_norm": 0.654943063189209,
+      "learning_rate": 5.1487931108527686e-06,
+      "loss": 0.3422,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5059687410263773,
+      "grad_norm": 0.7286314511506342,
+      "learning_rate": 5.147464988000615e-06,
+      "loss": 0.365,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5060507855765681,
+      "grad_norm": 0.6113384915054785,
+      "learning_rate": 5.146136854734719e-06,
+      "loss": 0.2866,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5061328301267588,
+      "grad_norm": 0.6917692618273381,
+      "learning_rate": 5.144808711148872e-06,
+      "loss": 0.318,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5062148746769496,
+      "grad_norm": 0.6516525931459465,
+      "learning_rate": 5.143480557336869e-06,
+      "loss": 0.3925,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5062969192271404,
+      "grad_norm": 0.8592050565309056,
+      "learning_rate": 5.142152393392499e-06,
+      "loss": 0.2956,
+      "step": 6171
+    },
+    {
+      "epoch": 0.506378963777331,
+      "grad_norm": 0.6259381974774835,
+      "learning_rate": 5.140824219409555e-06,
+      "loss": 0.2993,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5064610083275218,
+      "grad_norm": 0.6818489587906601,
+      "learning_rate": 5.13949603548183e-06,
+      "loss": 0.3278,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5065430528777126,
+      "grad_norm": 0.6438460373275238,
+      "learning_rate": 5.13816784170312e-06,
+      "loss": 0.3218,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5066250974279034,
+      "grad_norm": 0.6610028154511484,
+      "learning_rate": 5.136839638167219e-06,
+      "loss": 0.3078,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5067071419780941,
+      "grad_norm": 0.9542628258127317,
+      "learning_rate": 5.1355114249679225e-06,
+      "loss": 0.3285,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5067891865282849,
+      "grad_norm": 0.7669595738407505,
+      "learning_rate": 5.134183202199028e-06,
+      "loss": 0.3343,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5068712310784756,
+      "grad_norm": 0.6710097812059521,
+      "learning_rate": 5.132854969954329e-06,
+      "loss": 0.326,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5069532756286663,
+      "grad_norm": 0.6466377759646041,
+      "learning_rate": 5.131526728327627e-06,
+      "loss": 0.315,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5070353201788571,
+      "grad_norm": 0.7170198091970342,
+      "learning_rate": 5.130198477412719e-06,
+      "loss": 0.3392,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5071173647290479,
+      "grad_norm": 0.7072778987114329,
+      "learning_rate": 5.128870217303403e-06,
+      "loss": 0.3671,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5071994092792387,
+      "grad_norm": 0.6427575916807178,
+      "learning_rate": 5.12754194809348e-06,
+      "loss": 0.3192,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5072814538294294,
+      "grad_norm": 0.7516173533643669,
+      "learning_rate": 5.126213669876748e-06,
+      "loss": 0.3052,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5073634983796201,
+      "grad_norm": 0.9898676602378883,
+      "learning_rate": 5.1248853827470114e-06,
+      "loss": 0.2967,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5074455429298109,
+      "grad_norm": 0.6431237449698677,
+      "learning_rate": 5.12355708679807e-06,
+      "loss": 0.3023,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5075275874800016,
+      "grad_norm": 0.7381107318182555,
+      "learning_rate": 5.122228782123723e-06,
+      "loss": 0.3353,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5076096320301924,
+      "grad_norm": 0.7174998033190988,
+      "learning_rate": 5.120900468817776e-06,
+      "loss": 0.3553,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5076916765803832,
+      "grad_norm": 0.6642493760029903,
+      "learning_rate": 5.119572146974034e-06,
+      "loss": 0.3152,
+      "step": 6188
+    },
+    {
+      "epoch": 0.507773721130574,
+      "grad_norm": 0.7121320457337056,
+      "learning_rate": 5.118243816686299e-06,
+      "loss": 0.3455,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5078557656807646,
+      "grad_norm": 0.6859361077688829,
+      "learning_rate": 5.1169154780483754e-06,
+      "loss": 0.3145,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5079378102309554,
+      "grad_norm": 0.7089908239378433,
+      "learning_rate": 5.115587131154068e-06,
+      "loss": 0.3659,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5080198547811462,
+      "grad_norm": 0.705611738136857,
+      "learning_rate": 5.114258776097184e-06,
+      "loss": 0.3283,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5081018993313369,
+      "grad_norm": 0.6671148688924149,
+      "learning_rate": 5.1129304129715286e-06,
+      "loss": 0.3047,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5081839438815277,
+      "grad_norm": 0.7301396883911623,
+      "learning_rate": 5.111602041870908e-06,
+      "loss": 0.3008,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5082659884317184,
+      "grad_norm": 0.5897363383155861,
+      "learning_rate": 5.11027366288913e-06,
+      "loss": 0.2951,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5083480329819092,
+      "grad_norm": 0.6798093046651194,
+      "learning_rate": 5.108945276120004e-06,
+      "loss": 0.3355,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5084300775320999,
+      "grad_norm": 0.6272955743271569,
+      "learning_rate": 5.107616881657338e-06,
+      "loss": 0.3548,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5085121220822907,
+      "grad_norm": 0.6130594209051445,
+      "learning_rate": 5.10628847959494e-06,
+      "loss": 0.3481,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5085941666324815,
+      "grad_norm": 1.110403727353935,
+      "learning_rate": 5.104960070026622e-06,
+      "loss": 0.3199,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5086762111826721,
+      "grad_norm": 0.7343184496706434,
+      "learning_rate": 5.1036316530461905e-06,
+      "loss": 0.3579,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5087582557328629,
+      "grad_norm": 0.6242034287661231,
+      "learning_rate": 5.102303228747458e-06,
+      "loss": 0.3069,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5088403002830537,
+      "grad_norm": 0.6458645484761398,
+      "learning_rate": 5.1009747972242375e-06,
+      "loss": 0.3241,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5089223448332445,
+      "grad_norm": 0.7527947975521535,
+      "learning_rate": 5.099646358570338e-06,
+      "loss": 0.2891,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5090043893834352,
+      "grad_norm": 0.6817048773540239,
+      "learning_rate": 5.098317912879572e-06,
+      "loss": 0.3434,
+      "step": 6204
+    },
+    {
+      "epoch": 0.509086433933626,
+      "grad_norm": 0.6483723377012615,
+      "learning_rate": 5.0969894602457556e-06,
+      "loss": 0.3221,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5091684784838167,
+      "grad_norm": 0.6614716431191098,
+      "learning_rate": 5.0956610007626974e-06,
+      "loss": 0.3762,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5092505230340074,
+      "grad_norm": 0.9211816286022725,
+      "learning_rate": 5.094332534524214e-06,
+      "loss": 0.3118,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5093325675841982,
+      "grad_norm": 0.724486710768803,
+      "learning_rate": 5.093004061624118e-06,
+      "loss": 0.324,
+      "step": 6208
+    },
+    {
+      "epoch": 0.509414612134389,
+      "grad_norm": 0.7396572594024775,
+      "learning_rate": 5.091675582156224e-06,
+      "loss": 0.303,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5094966566845798,
+      "grad_norm": 0.6968157565988639,
+      "learning_rate": 5.0903470962143476e-06,
+      "loss": 0.324,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5095787012347704,
+      "grad_norm": 0.6859223151100055,
+      "learning_rate": 5.089018603892305e-06,
+      "loss": 0.3405,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5096607457849612,
+      "grad_norm": 0.6750635162055268,
+      "learning_rate": 5.0876901052839115e-06,
+      "loss": 0.3202,
+      "step": 6212
+    },
+    {
+      "epoch": 0.509742790335152,
+      "grad_norm": 0.7263142033875682,
+      "learning_rate": 5.086361600482984e-06,
+      "loss": 0.3703,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5098248348853427,
+      "grad_norm": 0.5729729372028806,
+      "learning_rate": 5.0850330895833385e-06,
+      "loss": 0.2845,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5099068794355335,
+      "grad_norm": 0.6779079813263039,
+      "learning_rate": 5.083704572678794e-06,
+      "loss": 0.2988,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5099889239857243,
+      "grad_norm": 0.8180727108863056,
+      "learning_rate": 5.082376049863166e-06,
+      "loss": 0.3237,
+      "step": 6216
+    },
+    {
+      "epoch": 0.510070968535915,
+      "grad_norm": 0.8385886029679234,
+      "learning_rate": 5.081047521230274e-06,
+      "loss": 0.3479,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5101530130861057,
+      "grad_norm": 0.7284224010811763,
+      "learning_rate": 5.079718986873935e-06,
+      "loss": 0.3906,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5102350576362965,
+      "grad_norm": 0.6945855185027167,
+      "learning_rate": 5.0783904468879715e-06,
+      "loss": 0.3216,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5103171021864873,
+      "grad_norm": 0.6469474235374474,
+      "learning_rate": 5.077061901366199e-06,
+      "loss": 0.3294,
+      "step": 6220
+    },
+    {
+      "epoch": 0.510399146736678,
+      "grad_norm": 0.7084266807241337,
+      "learning_rate": 5.075733350402439e-06,
+      "loss": 0.3226,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5104811912868688,
+      "grad_norm": 0.6446478842255121,
+      "learning_rate": 5.074404794090512e-06,
+      "loss": 0.3309,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5105632358370595,
+      "grad_norm": 0.6269334413714738,
+      "learning_rate": 5.073076232524238e-06,
+      "loss": 0.3295,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5106452803872503,
+      "grad_norm": 0.731116716220615,
+      "learning_rate": 5.071747665797437e-06,
+      "loss": 0.3217,
+      "step": 6224
+    },
+    {
+      "epoch": 0.510727324937441,
+      "grad_norm": 0.7243773795701205,
+      "learning_rate": 5.0704190940039326e-06,
+      "loss": 0.3192,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5108093694876318,
+      "grad_norm": 0.682465446981013,
+      "learning_rate": 5.069090517237544e-06,
+      "loss": 0.2768,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5108914140378226,
+      "grad_norm": 0.6564719848395066,
+      "learning_rate": 5.0677619355920935e-06,
+      "loss": 0.3457,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5109734585880132,
+      "grad_norm": 0.5996189534031003,
+      "learning_rate": 5.066433349161405e-06,
+      "loss": 0.2752,
+      "step": 6228
+    },
+    {
+      "epoch": 0.511055503138204,
+      "grad_norm": 0.7120756382029867,
+      "learning_rate": 5.0651047580392995e-06,
+      "loss": 0.3568,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5111375476883948,
+      "grad_norm": 0.834137465398037,
+      "learning_rate": 5.063776162319601e-06,
+      "loss": 0.39,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5112195922385856,
+      "grad_norm": 0.6918539259726886,
+      "learning_rate": 5.062447562096131e-06,
+      "loss": 0.3118,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5113016367887763,
+      "grad_norm": 0.6992027917321748,
+      "learning_rate": 5.061118957462716e-06,
+      "loss": 0.3645,
+      "step": 6232
+    },
+    {
+      "epoch": 0.511383681338967,
+      "grad_norm": 1.0101645571755509,
+      "learning_rate": 5.0597903485131795e-06,
+      "loss": 0.331,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5114657258891578,
+      "grad_norm": 0.6906606048212824,
+      "learning_rate": 5.058461735341342e-06,
+      "loss": 0.3358,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5115477704393485,
+      "grad_norm": 0.7143036148928786,
+      "learning_rate": 5.057133118041033e-06,
+      "loss": 0.3831,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5116298149895393,
+      "grad_norm": 0.7242335836530975,
+      "learning_rate": 5.055804496706072e-06,
+      "loss": 0.3466,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5117118595397301,
+      "grad_norm": 0.7070541799017314,
+      "learning_rate": 5.05447587143029e-06,
+      "loss": 0.3942,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5117939040899209,
+      "grad_norm": 0.6821295853247645,
+      "learning_rate": 5.053147242307508e-06,
+      "loss": 0.3674,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5118759486401115,
+      "grad_norm": 0.6843494275280468,
+      "learning_rate": 5.051818609431554e-06,
+      "loss": 0.3219,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5119579931903023,
+      "grad_norm": 0.7109406089346123,
+      "learning_rate": 5.050489972896253e-06,
+      "loss": 0.3468,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5120400377404931,
+      "grad_norm": 0.768381150568973,
+      "learning_rate": 5.049161332795431e-06,
+      "loss": 0.3237,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5121220822906838,
+      "grad_norm": 0.8030191724350119,
+      "learning_rate": 5.047832689222914e-06,
+      "loss": 0.2992,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5122041268408746,
+      "grad_norm": 0.6491892115890736,
+      "learning_rate": 5.046504042272531e-06,
+      "loss": 0.3215,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5122861713910654,
+      "grad_norm": 0.642599302460327,
+      "learning_rate": 5.0451753920381056e-06,
+      "loss": 0.2979,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5123682159412561,
+      "grad_norm": 0.6780806420679877,
+      "learning_rate": 5.043846738613466e-06,
+      "loss": 0.3284,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5124502604914468,
+      "grad_norm": 0.8205128788453121,
+      "learning_rate": 5.0425180820924415e-06,
+      "loss": 0.3117,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5125323050416376,
+      "grad_norm": 0.6159462650171361,
+      "learning_rate": 5.04118942256886e-06,
+      "loss": 0.3284,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5126143495918284,
+      "grad_norm": 0.8531120567255412,
+      "learning_rate": 5.039860760136546e-06,
+      "loss": 0.2981,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5126963941420191,
+      "grad_norm": 0.6213923323638715,
+      "learning_rate": 5.038532094889327e-06,
+      "loss": 0.3255,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5127784386922098,
+      "grad_norm": 0.6827982171350587,
+      "learning_rate": 5.037203426921036e-06,
+      "loss": 0.3914,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5128604832424006,
+      "grad_norm": 0.6146004958505239,
+      "learning_rate": 5.035874756325498e-06,
+      "loss": 0.3141,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5129425277925914,
+      "grad_norm": 0.7532862560973611,
+      "learning_rate": 5.034546083196542e-06,
+      "loss": 0.2765,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5130245723427821,
+      "grad_norm": 0.6448313375069078,
+      "learning_rate": 5.033217407627998e-06,
+      "loss": 0.3355,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5131066168929729,
+      "grad_norm": 0.6473600323359883,
+      "learning_rate": 5.031888729713693e-06,
+      "loss": 0.3032,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5131886614431637,
+      "grad_norm": 0.6991865539790592,
+      "learning_rate": 5.0305600495474586e-06,
+      "loss": 0.3696,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5132707059933543,
+      "grad_norm": 0.6864863230702278,
+      "learning_rate": 5.029231367223122e-06,
+      "loss": 0.3458,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5133527505435451,
+      "grad_norm": 0.7014872669767792,
+      "learning_rate": 5.027902682834512e-06,
+      "loss": 0.3153,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5134347950937359,
+      "grad_norm": 0.7295244391843319,
+      "learning_rate": 5.02657399647546e-06,
+      "loss": 0.3879,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5135168396439267,
+      "grad_norm": 0.6534739534666046,
+      "learning_rate": 5.025245308239795e-06,
+      "loss": 0.3071,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5135988841941174,
+      "grad_norm": 0.6978070273663871,
+      "learning_rate": 5.0239166182213486e-06,
+      "loss": 0.3068,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5136809287443082,
+      "grad_norm": 0.9725471909719373,
+      "learning_rate": 5.022587926513946e-06,
+      "loss": 0.3056,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5137629732944989,
+      "grad_norm": 0.6218270088833704,
+      "learning_rate": 5.021259233211424e-06,
+      "loss": 0.3625,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5138450178446896,
+      "grad_norm": 0.8256569135673905,
+      "learning_rate": 5.019930538407607e-06,
+      "loss": 0.3255,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5139270623948804,
+      "grad_norm": 0.6987161452173056,
+      "learning_rate": 5.018601842196326e-06,
+      "loss": 0.3526,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5140091069450712,
+      "grad_norm": 0.778872577205239,
+      "learning_rate": 5.017273144671415e-06,
+      "loss": 0.3315,
+      "step": 6265
+    },
+    {
+      "epoch": 0.514091151495262,
+      "grad_norm": 0.7352479083218652,
+      "learning_rate": 5.015944445926702e-06,
+      "loss": 0.3479,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5141731960454526,
+      "grad_norm": 0.7055458171348938,
+      "learning_rate": 5.014615746056017e-06,
+      "loss": 0.3366,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5142552405956434,
+      "grad_norm": 0.6644579199081649,
+      "learning_rate": 5.013287045153192e-06,
+      "loss": 0.311,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5143372851458342,
+      "grad_norm": 0.6713774660761762,
+      "learning_rate": 5.011958343312058e-06,
+      "loss": 0.3352,
+      "step": 6269
+    },
+    {
+      "epoch": 0.514419329696025,
+      "grad_norm": 0.6676976407555815,
+      "learning_rate": 5.010629640626444e-06,
+      "loss": 0.359,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5145013742462157,
+      "grad_norm": 0.9587958720083599,
+      "learning_rate": 5.0093009371901826e-06,
+      "loss": 0.3221,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5145834187964065,
+      "grad_norm": 0.6589971149427153,
+      "learning_rate": 5.007972233097103e-06,
+      "loss": 0.3432,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5146654633465972,
+      "grad_norm": 0.7016904098763822,
+      "learning_rate": 5.0066435284410365e-06,
+      "loss": 0.324,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5147475078967879,
+      "grad_norm": 0.8720232162755597,
+      "learning_rate": 5.005314823315816e-06,
+      "loss": 0.3209,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5148295524469787,
+      "grad_norm": 0.8086267527476155,
+      "learning_rate": 5.00398611781527e-06,
+      "loss": 0.364,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5149115969971695,
+      "grad_norm": 0.6901219185681543,
+      "learning_rate": 5.002657412033233e-06,
+      "loss": 0.3058,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5149936415473603,
+      "grad_norm": 0.6324082962090889,
+      "learning_rate": 5.001328706063532e-06,
+      "loss": 0.3287,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5150756860975509,
+      "grad_norm": 0.9034650092530822,
+      "learning_rate": 5e-06,
+      "loss": 0.3665,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5151577306477417,
+      "grad_norm": 0.739017032251046,
+      "learning_rate": 4.998671293936468e-06,
+      "loss": 0.3869,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5152397751979325,
+      "grad_norm": 0.6165443663891559,
+      "learning_rate": 4.997342587966769e-06,
+      "loss": 0.2691,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5153218197481232,
+      "grad_norm": 0.7064045367105056,
+      "learning_rate": 4.9960138821847305e-06,
+      "loss": 0.3172,
+      "step": 6281
+    },
+    {
+      "epoch": 0.515403864298314,
+      "grad_norm": 0.8055243615107298,
+      "learning_rate": 4.994685176684185e-06,
+      "loss": 0.3295,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5154859088485048,
+      "grad_norm": 0.7320038235611215,
+      "learning_rate": 4.993356471558965e-06,
+      "loss": 0.3043,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5155679533986955,
+      "grad_norm": 0.6684841915904202,
+      "learning_rate": 4.992027766902898e-06,
+      "loss": 0.3277,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5156499979488862,
+      "grad_norm": 0.6564097209868054,
+      "learning_rate": 4.99069906280982e-06,
+      "loss": 0.3365,
+      "step": 6285
+    },
+    {
+      "epoch": 0.515732042499077,
+      "grad_norm": 0.651512792248995,
+      "learning_rate": 4.989370359373557e-06,
+      "loss": 0.3306,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5158140870492678,
+      "grad_norm": 0.8571551357555078,
+      "learning_rate": 4.988041656687944e-06,
+      "loss": 0.3132,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5158961315994585,
+      "grad_norm": 0.7146762298809404,
+      "learning_rate": 4.986712954846809e-06,
+      "loss": 0.3208,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5159781761496492,
+      "grad_norm": 0.6814279219341846,
+      "learning_rate": 4.985384253943984e-06,
+      "loss": 0.3071,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51606022069984,
+      "grad_norm": 0.8345226437784572,
+      "learning_rate": 4.9840555540733e-06,
+      "loss": 0.3688,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5161422652500308,
+      "grad_norm": 0.6076427478298173,
+      "learning_rate": 4.982726855328586e-06,
+      "loss": 0.3574,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5162243098002215,
+      "grad_norm": 0.7878642144527467,
+      "learning_rate": 4.981398157803674e-06,
+      "loss": 0.3792,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5163063543504123,
+      "grad_norm": 0.7653276010091864,
+      "learning_rate": 4.980069461592395e-06,
+      "loss": 0.3199,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5163883989006031,
+      "grad_norm": 0.664061097255933,
+      "learning_rate": 4.978740766788578e-06,
+      "loss": 0.3165,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5164704434507937,
+      "grad_norm": 0.6399794163433625,
+      "learning_rate": 4.977412073486055e-06,
+      "loss": 0.3535,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5165524880009845,
+      "grad_norm": 0.7518753587977639,
+      "learning_rate": 4.976083381778654e-06,
+      "loss": 0.3359,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5166345325511753,
+      "grad_norm": 0.673333460748579,
+      "learning_rate": 4.974754691760205e-06,
+      "loss": 0.3142,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5167165771013661,
+      "grad_norm": 0.7170494221395433,
+      "learning_rate": 4.973426003524541e-06,
+      "loss": 0.3245,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5167986216515568,
+      "grad_norm": 0.6217869748416132,
+      "learning_rate": 4.972097317165489e-06,
+      "loss": 0.3562,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5168806662017476,
+      "grad_norm": 0.752618322253961,
+      "learning_rate": 4.9707686327768805e-06,
+      "loss": 0.3154,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5169627107519383,
+      "grad_norm": 0.6328261260313488,
+      "learning_rate": 4.969439950452543e-06,
+      "loss": 0.2773,
+      "step": 6301
+    },
+    {
+      "epoch": 0.517044755302129,
+      "grad_norm": 0.7757745420528865,
+      "learning_rate": 4.968111270286308e-06,
+      "loss": 0.3389,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5171267998523198,
+      "grad_norm": 0.7710245049494594,
+      "learning_rate": 4.966782592372003e-06,
+      "loss": 0.3156,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5172088444025106,
+      "grad_norm": 0.7154113956241509,
+      "learning_rate": 4.965453916803459e-06,
+      "loss": 0.3219,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5172908889527014,
+      "grad_norm": 0.6573657399873225,
+      "learning_rate": 4.964125243674502e-06,
+      "loss": 0.3432,
+      "step": 6305
+    },
+    {
+      "epoch": 0.517372933502892,
+      "grad_norm": 0.7014073976857707,
+      "learning_rate": 4.962796573078966e-06,
+      "loss": 0.3824,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5174549780530828,
+      "grad_norm": 0.6476866273111834,
+      "learning_rate": 4.961467905110673e-06,
+      "loss": 0.326,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5175370226032736,
+      "grad_norm": 0.637243367970202,
+      "learning_rate": 4.960139239863458e-06,
+      "loss": 0.3437,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5176190671534643,
+      "grad_norm": 0.5901485027418826,
+      "learning_rate": 4.958810577431142e-06,
+      "loss": 0.326,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5177011117036551,
+      "grad_norm": 0.6084238453345372,
+      "learning_rate": 4.957481917907559e-06,
+      "loss": 0.3224,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5177831562538459,
+      "grad_norm": 0.7538280705024395,
+      "learning_rate": 4.956153261386536e-06,
+      "loss": 0.3297,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5178652008040366,
+      "grad_norm": 0.6288244443165543,
+      "learning_rate": 4.954824607961895e-06,
+      "loss": 0.3247,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5179472453542273,
+      "grad_norm": 0.7070583202430462,
+      "learning_rate": 4.953495957727472e-06,
+      "loss": 0.3396,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5180292899044181,
+      "grad_norm": 0.7218404667711075,
+      "learning_rate": 4.952167310777087e-06,
+      "loss": 0.3838,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5181113344546089,
+      "grad_norm": 0.6236963148593145,
+      "learning_rate": 4.95083866720457e-06,
+      "loss": 0.3088,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5181933790047996,
+      "grad_norm": 0.6849636057744553,
+      "learning_rate": 4.949510027103749e-06,
+      "loss": 0.3399,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5182754235549903,
+      "grad_norm": 0.6489221616527864,
+      "learning_rate": 4.948181390568447e-06,
+      "loss": 0.2916,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5183574681051811,
+      "grad_norm": 0.7550642511114839,
+      "learning_rate": 4.946852757692494e-06,
+      "loss": 0.3278,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5184395126553719,
+      "grad_norm": 0.6464516370877077,
+      "learning_rate": 4.945524128569712e-06,
+      "loss": 0.2805,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5185215572055626,
+      "grad_norm": 0.6392095533866036,
+      "learning_rate": 4.9441955032939275e-06,
+      "loss": 0.279,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5186036017557534,
+      "grad_norm": 0.6864196083079667,
+      "learning_rate": 4.94286688195897e-06,
+      "loss": 0.3509,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5186856463059442,
+      "grad_norm": 0.6718820991223714,
+      "learning_rate": 4.941538264658659e-06,
+      "loss": 0.3471,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5187676908561348,
+      "grad_norm": 0.6652029419513638,
+      "learning_rate": 4.940209651486822e-06,
+      "loss": 0.38,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5188497354063256,
+      "grad_norm": 0.7021689227344364,
+      "learning_rate": 4.938881042537286e-06,
+      "loss": 0.3133,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5189317799565164,
+      "grad_norm": 0.6169948946190759,
+      "learning_rate": 4.937552437903869e-06,
+      "loss": 0.2822,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5190138245067072,
+      "grad_norm": 0.701789398738632,
+      "learning_rate": 4.9362238376804015e-06,
+      "loss": 0.3325,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5190958690568979,
+      "grad_norm": 0.6309096926400584,
+      "learning_rate": 4.934895241960702e-06,
+      "loss": 0.3166,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5191779136070886,
+      "grad_norm": 0.6661901582167571,
+      "learning_rate": 4.933566650838595e-06,
+      "loss": 0.3396,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5192599581572794,
+      "grad_norm": 0.7968513977758986,
+      "learning_rate": 4.932238064407907e-06,
+      "loss": 0.3293,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5193420027074701,
+      "grad_norm": 0.6419164698388885,
+      "learning_rate": 4.930909482762458e-06,
+      "loss": 0.3132,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5194240472576609,
+      "grad_norm": 0.7189913715014496,
+      "learning_rate": 4.92958090599607e-06,
+      "loss": 0.2994,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5195060918078517,
+      "grad_norm": 0.6951910135068428,
+      "learning_rate": 4.9282523342025645e-06,
+      "loss": 0.3371,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5195881363580425,
+      "grad_norm": 0.6342317546682834,
+      "learning_rate": 4.926923767475763e-06,
+      "loss": 0.3108,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5196701809082331,
+      "grad_norm": 0.757014494470952,
+      "learning_rate": 4.92559520590949e-06,
+      "loss": 0.322,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5197522254584239,
+      "grad_norm": 0.9938035981200666,
+      "learning_rate": 4.9242666495975615e-06,
+      "loss": 0.3314,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5198342700086147,
+      "grad_norm": 0.6736032170948458,
+      "learning_rate": 4.922938098633803e-06,
+      "loss": 0.357,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5199163145588054,
+      "grad_norm": 0.6678352462565611,
+      "learning_rate": 4.92160955311203e-06,
+      "loss": 0.3101,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5199983591089962,
+      "grad_norm": 0.64403409561554,
+      "learning_rate": 4.920281013126065e-06,
+      "loss": 0.3097,
+      "step": 6338
+    },
+    {
+      "epoch": 0.520080403659187,
+      "grad_norm": 0.6596206512378826,
+      "learning_rate": 4.918952478769729e-06,
+      "loss": 0.3137,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5201624482093777,
+      "grad_norm": 0.693519761800403,
+      "learning_rate": 4.917623950136836e-06,
+      "loss": 0.3372,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5202444927595684,
+      "grad_norm": 0.6603338032978666,
+      "learning_rate": 4.916295427321206e-06,
+      "loss": 0.3222,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5203265373097592,
+      "grad_norm": 0.6911408422880482,
+      "learning_rate": 4.914966910416662e-06,
+      "loss": 0.3387,
+      "step": 6342
+    },
+    {
+      "epoch": 0.52040858185995,
+      "grad_norm": 0.6372374205318857,
+      "learning_rate": 4.913638399517017e-06,
+      "loss": 0.3105,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5204906264101407,
+      "grad_norm": 0.6352722677233603,
+      "learning_rate": 4.912309894716089e-06,
+      "loss": 0.3053,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5205726709603314,
+      "grad_norm": 0.6819068592537884,
+      "learning_rate": 4.910981396107696e-06,
+      "loss": 0.3227,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5206547155105222,
+      "grad_norm": 0.7162950137761486,
+      "learning_rate": 4.909652903785652e-06,
+      "loss": 0.3586,
+      "step": 6346
+    },
+    {
+      "epoch": 0.520736760060713,
+      "grad_norm": 0.7020961943452084,
+      "learning_rate": 4.908324417843779e-06,
+      "loss": 0.3012,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5208188046109037,
+      "grad_norm": 0.6784077769535696,
+      "learning_rate": 4.906995938375884e-06,
+      "loss": 0.3273,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5209008491610945,
+      "grad_norm": 0.6296269709300761,
+      "learning_rate": 4.905667465475789e-06,
+      "loss": 0.3519,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5209828937112853,
+      "grad_norm": 0.7541140198850645,
+      "learning_rate": 4.904338999237303e-06,
+      "loss": 0.3322,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5210649382614759,
+      "grad_norm": 0.689158457438998,
+      "learning_rate": 4.903010539754246e-06,
+      "loss": 0.302,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5211469828116667,
+      "grad_norm": 0.6152989466676231,
+      "learning_rate": 4.901682087120429e-06,
+      "loss": 0.3239,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5212290273618575,
+      "grad_norm": 0.6654498422021737,
+      "learning_rate": 4.900353641429663e-06,
+      "loss": 0.3458,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5213110719120483,
+      "grad_norm": 0.6425409938862446,
+      "learning_rate": 4.899025202775765e-06,
+      "loss": 0.302,
+      "step": 6354
+    },
+    {
+      "epoch": 0.521393116462239,
+      "grad_norm": 0.7024515261389408,
+      "learning_rate": 4.897696771252543e-06,
+      "loss": 0.3154,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5214751610124297,
+      "grad_norm": 0.6598882299864881,
+      "learning_rate": 4.8963683469538095e-06,
+      "loss": 0.3307,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5215572055626205,
+      "grad_norm": 0.7969777381462393,
+      "learning_rate": 4.895039929973381e-06,
+      "loss": 0.3035,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5216392501128112,
+      "grad_norm": 0.7548871745875803,
+      "learning_rate": 4.893711520405061e-06,
+      "loss": 0.3377,
+      "step": 6358
+    },
+    {
+      "epoch": 0.521721294663002,
+      "grad_norm": 0.6078093195995952,
+      "learning_rate": 4.892383118342663e-06,
+      "loss": 0.2833,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5218033392131928,
+      "grad_norm": 0.7518478963485022,
+      "learning_rate": 4.891054723879998e-06,
+      "loss": 0.3262,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5218853837633836,
+      "grad_norm": 0.6482356783980824,
+      "learning_rate": 4.889726337110871e-06,
+      "loss": 0.342,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5219674283135742,
+      "grad_norm": 0.7061376133433687,
+      "learning_rate": 4.888397958129095e-06,
+      "loss": 0.3086,
+      "step": 6362
+    },
+    {
+      "epoch": 0.522049472863765,
+      "grad_norm": 0.702450328092684,
+      "learning_rate": 4.887069587028473e-06,
+      "loss": 0.3416,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5221315174139558,
+      "grad_norm": 0.6740275121571017,
+      "learning_rate": 4.8857412239028175e-06,
+      "loss": 0.3235,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5222135619641465,
+      "grad_norm": 0.8964045097714142,
+      "learning_rate": 4.884412868845933e-06,
+      "loss": 0.317,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5222956065143373,
+      "grad_norm": 0.7341031871500424,
+      "learning_rate": 4.883084521951626e-06,
+      "loss": 0.3235,
+      "step": 6366
+    },
+    {
+      "epoch": 0.522377651064528,
+      "grad_norm": 0.7362464078370211,
+      "learning_rate": 4.881756183313703e-06,
+      "loss": 0.3277,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5224596956147188,
+      "grad_norm": 0.6641655652038256,
+      "learning_rate": 4.880427853025967e-06,
+      "loss": 0.3305,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5225417401649095,
+      "grad_norm": 0.6678186090981628,
+      "learning_rate": 4.879099531182223e-06,
+      "loss": 0.2948,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5226237847151003,
+      "grad_norm": 0.7494080055314063,
+      "learning_rate": 4.877771217876279e-06,
+      "loss": 0.3661,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5227058292652911,
+      "grad_norm": 0.61908503674112,
+      "learning_rate": 4.876442913201932e-06,
+      "loss": 0.2807,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5227878738154818,
+      "grad_norm": 0.6093869483284496,
+      "learning_rate": 4.87511461725299e-06,
+      "loss": 0.3129,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5228699183656725,
+      "grad_norm": 0.657294667694786,
+      "learning_rate": 4.873786330123253e-06,
+      "loss": 0.3254,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5229519629158633,
+      "grad_norm": 0.6756210517533582,
+      "learning_rate": 4.872458051906521e-06,
+      "loss": 0.3258,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5230340074660541,
+      "grad_norm": 0.6909801876245921,
+      "learning_rate": 4.8711297826965996e-06,
+      "loss": 0.3471,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5231160520162448,
+      "grad_norm": 0.7064117982188312,
+      "learning_rate": 4.869801522587282e-06,
+      "loss": 0.3342,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5231980965664356,
+      "grad_norm": 0.7598319086873879,
+      "learning_rate": 4.868473271672374e-06,
+      "loss": 0.3307,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5232801411166264,
+      "grad_norm": 0.6514840329763004,
+      "learning_rate": 4.867145030045672e-06,
+      "loss": 0.3222,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5233621856668171,
+      "grad_norm": 0.6268027955429467,
+      "learning_rate": 4.865816797800974e-06,
+      "loss": 0.3252,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5234442302170078,
+      "grad_norm": 0.7017156524124344,
+      "learning_rate": 4.86448857503208e-06,
+      "loss": 0.2722,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5235262747671986,
+      "grad_norm": 0.6696823561115274,
+      "learning_rate": 4.8631603618327824e-06,
+      "loss": 0.322,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5236083193173894,
+      "grad_norm": 0.6366954617613553,
+      "learning_rate": 4.8618321582968806e-06,
+      "loss": 0.3235,
+      "step": 6382
+    },
+    {
+      "epoch": 0.52369036386758,
+      "grad_norm": 0.7129473512633275,
+      "learning_rate": 4.860503964518171e-06,
+      "loss": 0.3222,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5237724084177708,
+      "grad_norm": 0.7019331808949657,
+      "learning_rate": 4.859175780590446e-06,
+      "loss": 0.3072,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5238544529679616,
+      "grad_norm": 0.882216062870113,
+      "learning_rate": 4.8578476066075035e-06,
+      "loss": 0.3453,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5239364975181524,
+      "grad_norm": 0.6619102282699266,
+      "learning_rate": 4.856519442663134e-06,
+      "loss": 0.3303,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5240185420683431,
+      "grad_norm": 0.6584526859196712,
+      "learning_rate": 4.855191288851129e-06,
+      "loss": 0.309,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5241005866185339,
+      "grad_norm": 0.7261233000103132,
+      "learning_rate": 4.853863145265284e-06,
+      "loss": 0.344,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5241826311687247,
+      "grad_norm": 0.7276654752376331,
+      "learning_rate": 4.852535011999388e-06,
+      "loss": 0.3054,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5242646757189153,
+      "grad_norm": 0.7114448045587529,
+      "learning_rate": 4.851206889147234e-06,
+      "loss": 0.2971,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5243467202691061,
+      "grad_norm": 0.6645113191519174,
+      "learning_rate": 4.849878776802609e-06,
+      "loss": 0.2946,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5244287648192969,
+      "grad_norm": 0.7028239825751901,
+      "learning_rate": 4.848550675059303e-06,
+      "loss": 0.3117,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5245108093694877,
+      "grad_norm": 0.6322231255375499,
+      "learning_rate": 4.847222584011107e-06,
+      "loss": 0.2907,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5245928539196784,
+      "grad_norm": 0.6500429532246079,
+      "learning_rate": 4.845894503751805e-06,
+      "loss": 0.3305,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5246748984698691,
+      "grad_norm": 0.6808992291614967,
+      "learning_rate": 4.844566434375186e-06,
+      "loss": 0.2809,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5247569430200599,
+      "grad_norm": 1.7701173490131832,
+      "learning_rate": 4.843238375975037e-06,
+      "loss": 0.304,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5248389875702506,
+      "grad_norm": 0.6536991270006799,
+      "learning_rate": 4.84191032864514e-06,
+      "loss": 0.3552,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5249210321204414,
+      "grad_norm": 0.6519010251564082,
+      "learning_rate": 4.840582292479283e-06,
+      "loss": 0.3394,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5250030766706322,
+      "grad_norm": 0.625415893996803,
+      "learning_rate": 4.839254267571247e-06,
+      "loss": 0.3551,
+      "step": 6399
+    },
+    {
+      "epoch": 0.525085121220823,
+      "grad_norm": 0.6804610950058226,
+      "learning_rate": 4.8379262540148185e-06,
+      "loss": 0.2994,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5251671657710136,
+      "grad_norm": 0.6509451454120031,
+      "learning_rate": 4.836598251903777e-06,
+      "loss": 0.3268,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5252492103212044,
+      "grad_norm": 0.6725567493374132,
+      "learning_rate": 4.835270261331903e-06,
+      "loss": 0.3102,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5253312548713952,
+      "grad_norm": 0.6365206690782139,
+      "learning_rate": 4.83394228239298e-06,
+      "loss": 0.3142,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5254132994215859,
+      "grad_norm": 0.6417063370056676,
+      "learning_rate": 4.832614315180784e-06,
+      "loss": 0.3073,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5254953439717767,
+      "grad_norm": 0.7313683857154759,
+      "learning_rate": 4.8312863597890976e-06,
+      "loss": 0.3328,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5255773885219674,
+      "grad_norm": 0.615524609236577,
+      "learning_rate": 4.829958416311698e-06,
+      "loss": 0.3005,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5256594330721582,
+      "grad_norm": 0.6963599218588755,
+      "learning_rate": 4.828630484842362e-06,
+      "loss": 0.3149,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5257414776223489,
+      "grad_norm": 0.7160141157120016,
+      "learning_rate": 4.827302565474866e-06,
+      "loss": 0.3809,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5258235221725397,
+      "grad_norm": 0.6645684251276942,
+      "learning_rate": 4.825974658302984e-06,
+      "loss": 0.2878,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5259055667227305,
+      "grad_norm": 0.6274811153207897,
+      "learning_rate": 4.8246467634204916e-06,
+      "loss": 0.3479,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5259876112729212,
+      "grad_norm": 0.7657754740911851,
+      "learning_rate": 4.823318880921165e-06,
+      "loss": 0.2831,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5260696558231119,
+      "grad_norm": 0.6163848627368349,
+      "learning_rate": 4.821991010898773e-06,
+      "loss": 0.2879,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5261517003733027,
+      "grad_norm": 0.7123570557573818,
+      "learning_rate": 4.820663153447091e-06,
+      "loss": 0.3251,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5262337449234935,
+      "grad_norm": 0.7338435296933814,
+      "learning_rate": 4.819335308659891e-06,
+      "loss": 0.3327,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6856323477542919,
+      "learning_rate": 4.818007476630938e-06,
+      "loss": 0.2974,
+      "step": 6415
+    },
+    {
+      "epoch": 0.526397834023875,
+      "grad_norm": 0.702840335745122,
+      "learning_rate": 4.8166796574540065e-06,
+      "loss": 0.303,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5264798785740658,
+      "grad_norm": 0.6640547801932544,
+      "learning_rate": 4.815351851222862e-06,
+      "loss": 0.3496,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5265619231242564,
+      "grad_norm": 0.6682107735190768,
+      "learning_rate": 4.814024058031272e-06,
+      "loss": 0.3195,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5266439676744472,
+      "grad_norm": 0.6431278805081704,
+      "learning_rate": 4.812696277973007e-06,
+      "loss": 0.3512,
+      "step": 6419
+    },
+    {
+      "epoch": 0.526726012224638,
+      "grad_norm": 0.7081851912923351,
+      "learning_rate": 4.811368511141828e-06,
+      "loss": 0.3327,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5268080567748288,
+      "grad_norm": 0.707561699292353,
+      "learning_rate": 4.810040757631503e-06,
+      "loss": 0.3266,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5268901013250195,
+      "grad_norm": 0.7785197868042092,
+      "learning_rate": 4.8087130175357936e-06,
+      "loss": 0.3123,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5269721458752102,
+      "grad_norm": 0.6024005872219144,
+      "learning_rate": 4.807385290948463e-06,
+      "loss": 0.2971,
+      "step": 6423
+    },
+    {
+      "epoch": 0.527054190425401,
+      "grad_norm": 0.6142711236342112,
+      "learning_rate": 4.806057577963276e-06,
+      "loss": 0.348,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5271362349755917,
+      "grad_norm": 0.9379075418931083,
+      "learning_rate": 4.8047298786739884e-06,
+      "loss": 0.3416,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5272182795257825,
+      "grad_norm": 0.6953238841419617,
+      "learning_rate": 4.803402193174365e-06,
+      "loss": 0.3437,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5273003240759733,
+      "grad_norm": 0.6938141037203264,
+      "learning_rate": 4.802074521558162e-06,
+      "loss": 0.2941,
+      "step": 6427
+    },
+    {
+      "epoch": 0.527382368626164,
+      "grad_norm": 0.7172131393005544,
+      "learning_rate": 4.800746863919138e-06,
+      "loss": 0.3465,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5274644131763547,
+      "grad_norm": 0.6955343225795614,
+      "learning_rate": 4.799419220351053e-06,
+      "loss": 0.3182,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5275464577265455,
+      "grad_norm": 0.6074139835570509,
+      "learning_rate": 4.798091590947657e-06,
+      "loss": 0.3217,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5276285022767363,
+      "grad_norm": 0.716395281215371,
+      "learning_rate": 4.796763975802708e-06,
+      "loss": 0.2989,
+      "step": 6431
+    },
+    {
+      "epoch": 0.527710546826927,
+      "grad_norm": 0.6560262884724528,
+      "learning_rate": 4.795436375009964e-06,
+      "loss": 0.3189,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5277925913771178,
+      "grad_norm": 0.663509697600703,
+      "learning_rate": 4.794108788663172e-06,
+      "loss": 0.3411,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5278746359273085,
+      "grad_norm": 0.842701441298497,
+      "learning_rate": 4.792781216856088e-06,
+      "loss": 0.3217,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5279566804774993,
+      "grad_norm": 0.709771154165741,
+      "learning_rate": 4.7914536596824616e-06,
+      "loss": 0.3277,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52803872502769,
+      "grad_norm": 0.6592317606034984,
+      "learning_rate": 4.790126117236041e-06,
+      "loss": 0.3161,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5281207695778808,
+      "grad_norm": 0.6472412099415618,
+      "learning_rate": 4.788798589610579e-06,
+      "loss": 0.3362,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5282028141280716,
+      "grad_norm": 0.676737385965375,
+      "learning_rate": 4.787471076899819e-06,
+      "loss": 0.3103,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5282848586782622,
+      "grad_norm": 0.6695112928863937,
+      "learning_rate": 4.7861435791975124e-06,
+      "loss": 0.338,
+      "step": 6439
+    },
+    {
+      "epoch": 0.528366903228453,
+      "grad_norm": 0.6286101399178482,
+      "learning_rate": 4.784816096597401e-06,
+      "loss": 0.3553,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5284489477786438,
+      "grad_norm": 0.7030778956387379,
+      "learning_rate": 4.783488629193231e-06,
+      "loss": 0.2978,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5285309923288346,
+      "grad_norm": 0.7620871776306524,
+      "learning_rate": 4.782161177078748e-06,
+      "loss": 0.3334,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5286130368790253,
+      "grad_norm": 0.7411217968885891,
+      "learning_rate": 4.78083374034769e-06,
+      "loss": 0.3274,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5286950814292161,
+      "grad_norm": 0.7322676438938259,
+      "learning_rate": 4.779506319093804e-06,
+      "loss": 0.3532,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5287771259794068,
+      "grad_norm": 0.7935104611200501,
+      "learning_rate": 4.778178913410826e-06,
+      "loss": 0.302,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5288591705295975,
+      "grad_norm": 0.6204027958975259,
+      "learning_rate": 4.776851523392495e-06,
+      "loss": 0.3022,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5289412150797883,
+      "grad_norm": 0.7408853705289103,
+      "learning_rate": 4.7755241491325536e-06,
+      "loss": 0.3323,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5290232596299791,
+      "grad_norm": 0.5895372929375333,
+      "learning_rate": 4.774196790724735e-06,
+      "loss": 0.3191,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5291053041801699,
+      "grad_norm": 0.7025894922565487,
+      "learning_rate": 4.772869448262776e-06,
+      "loss": 0.3158,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5291873487303606,
+      "grad_norm": 0.6337893609081927,
+      "learning_rate": 4.771542121840413e-06,
+      "loss": 0.324,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5292693932805513,
+      "grad_norm": 0.6056857251647226,
+      "learning_rate": 4.770214811551377e-06,
+      "loss": 0.3016,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5293514378307421,
+      "grad_norm": 0.6091318508832806,
+      "learning_rate": 4.7688875174894035e-06,
+      "loss": 0.3214,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5294334823809328,
+      "grad_norm": 0.8672441463235769,
+      "learning_rate": 4.76756023974822e-06,
+      "loss": 0.3523,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5295155269311236,
+      "grad_norm": 0.6125493616888918,
+      "learning_rate": 4.76623297842156e-06,
+      "loss": 0.3208,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5295975714813144,
+      "grad_norm": 0.6982249271655526,
+      "learning_rate": 4.764905733603151e-06,
+      "loss": 0.2997,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5296796160315052,
+      "grad_norm": 0.6417057356670137,
+      "learning_rate": 4.763578505386722e-06,
+      "loss": 0.3105,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5297616605816958,
+      "grad_norm": 0.678905644299935,
+      "learning_rate": 4.762251293866e-06,
+      "loss": 0.3518,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5298437051318866,
+      "grad_norm": 0.6803009456785772,
+      "learning_rate": 4.760924099134708e-06,
+      "loss": 0.317,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5299257496820774,
+      "grad_norm": 0.6239532405047689,
+      "learning_rate": 4.759596921286571e-06,
+      "loss": 0.3127,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5300077942322681,
+      "grad_norm": 0.6602102923728154,
+      "learning_rate": 4.758269760415317e-06,
+      "loss": 0.2847,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5300898387824589,
+      "grad_norm": 0.6936822315608264,
+      "learning_rate": 4.75694261661466e-06,
+      "loss": 0.3183,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5301718833326496,
+      "grad_norm": 0.6688800452140071,
+      "learning_rate": 4.755615489978328e-06,
+      "loss": 0.3271,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5302539278828404,
+      "grad_norm": 0.705572487552552,
+      "learning_rate": 4.754288380600037e-06,
+      "loss": 0.3488,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5303359724330311,
+      "grad_norm": 0.6460349626106814,
+      "learning_rate": 4.752961288573504e-06,
+      "loss": 0.3581,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5304180169832219,
+      "grad_norm": 0.6225280767979392,
+      "learning_rate": 4.75163421399245e-06,
+      "loss": 0.3,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5305000615334127,
+      "grad_norm": 0.6577282656142559,
+      "learning_rate": 4.750307156950586e-06,
+      "loss": 0.3001,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5305821060836033,
+      "grad_norm": 0.6517177485691309,
+      "learning_rate": 4.7489801175416305e-06,
+      "loss": 0.3113,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5306641506337941,
+      "grad_norm": 0.7370052328886373,
+      "learning_rate": 4.747653095859298e-06,
+      "loss": 0.341,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5307461951839849,
+      "grad_norm": 0.7229813876197587,
+      "learning_rate": 4.746326091997297e-06,
+      "loss": 0.3708,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5308282397341757,
+      "grad_norm": 0.6779431675034842,
+      "learning_rate": 4.744999106049341e-06,
+      "loss": 0.3044,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5309102842843664,
+      "grad_norm": 0.6884349835089357,
+      "learning_rate": 4.743672138109136e-06,
+      "loss": 0.3136,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5309923288345572,
+      "grad_norm": 0.6184939685569363,
+      "learning_rate": 4.742345188270393e-06,
+      "loss": 0.3163,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5310743733847479,
+      "grad_norm": 0.6301402176384234,
+      "learning_rate": 4.741018256626821e-06,
+      "loss": 0.3012,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5311564179349386,
+      "grad_norm": 0.7017685512897761,
+      "learning_rate": 4.739691343272121e-06,
+      "loss": 0.3291,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5312384624851294,
+      "grad_norm": 0.6581269562647446,
+      "learning_rate": 4.738364448300004e-06,
+      "loss": 0.3226,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5313205070353202,
+      "grad_norm": 0.6736222748257695,
+      "learning_rate": 4.737037571804166e-06,
+      "loss": 0.3251,
+      "step": 6476
+    },
+    {
+      "epoch": 0.531402551585511,
+      "grad_norm": 0.7475652488462076,
+      "learning_rate": 4.735710713878313e-06,
+      "loss": 0.3327,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5314845961357016,
+      "grad_norm": 0.7146624250264249,
+      "learning_rate": 4.734383874616145e-06,
+      "loss": 0.3462,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5315666406858924,
+      "grad_norm": 0.6966893663007123,
+      "learning_rate": 4.73305705411136e-06,
+      "loss": 0.3432,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5316486852360832,
+      "grad_norm": 0.6488278528807537,
+      "learning_rate": 4.731730252457657e-06,
+      "loss": 0.3322,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5317307297862739,
+      "grad_norm": 0.6617696136848813,
+      "learning_rate": 4.730403469748735e-06,
+      "loss": 0.3172,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5318127743364647,
+      "grad_norm": 1.0078933072343452,
+      "learning_rate": 4.729076706078285e-06,
+      "loss": 0.3917,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5318948188866555,
+      "grad_norm": 0.6818596879176274,
+      "learning_rate": 4.727749961540003e-06,
+      "loss": 0.3124,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5319768634368462,
+      "grad_norm": 0.809636544880875,
+      "learning_rate": 4.726423236227581e-06,
+      "loss": 0.3217,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5320589079870369,
+      "grad_norm": 1.1610846092119096,
+      "learning_rate": 4.72509653023471e-06,
+      "loss": 0.3148,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5321409525372277,
+      "grad_norm": 0.6721915097275346,
+      "learning_rate": 4.723769843655083e-06,
+      "loss": 0.3295,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5322229970874185,
+      "grad_norm": 0.7532798096731711,
+      "learning_rate": 4.722443176582383e-06,
+      "loss": 0.3311,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5323050416376092,
+      "grad_norm": 0.6648754405508817,
+      "learning_rate": 4.721116529110302e-06,
+      "loss": 0.3459,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5323870861878,
+      "grad_norm": 0.7693017134025625,
+      "learning_rate": 4.719789901332521e-06,
+      "loss": 0.3442,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5324691307379907,
+      "grad_norm": 0.7285608214125033,
+      "learning_rate": 4.718463293342729e-06,
+      "loss": 0.3164,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5325511752881815,
+      "grad_norm": 0.8108392229016047,
+      "learning_rate": 4.7171367052346076e-06,
+      "loss": 0.3215,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5326332198383722,
+      "grad_norm": 0.6796359953806506,
+      "learning_rate": 4.715810137101836e-06,
+      "loss": 0.3741,
+      "step": 6492
+    },
+    {
+      "epoch": 0.532715264388563,
+      "grad_norm": 1.10414478984892,
+      "learning_rate": 4.714483589038098e-06,
+      "loss": 0.3303,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5327973089387538,
+      "grad_norm": 0.6663845799007458,
+      "learning_rate": 4.713157061137067e-06,
+      "loss": 0.3229,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5328793534889446,
+      "grad_norm": 0.651591871167615,
+      "learning_rate": 4.711830553492424e-06,
+      "loss": 0.316,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5329613980391352,
+      "grad_norm": 0.6107987405653001,
+      "learning_rate": 4.710504066197847e-06,
+      "loss": 0.304,
+      "step": 6496
+    },
+    {
+      "epoch": 0.533043442589326,
+      "grad_norm": 0.7442249931363393,
+      "learning_rate": 4.709177599347006e-06,
+      "loss": 0.3392,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5331254871395168,
+      "grad_norm": 0.65793924813501,
+      "learning_rate": 4.707851153033575e-06,
+      "loss": 0.3421,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5332075316897075,
+      "grad_norm": 0.6970254263873301,
+      "learning_rate": 4.706524727351229e-06,
+      "loss": 0.3172,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5332895762398983,
+      "grad_norm": 0.7515797034929957,
+      "learning_rate": 4.705198322393632e-06,
+      "loss": 0.3254,
+      "step": 6500
+    },
+    {
+      "epoch": 0.533371620790089,
+      "grad_norm": 0.7248466507801031,
+      "learning_rate": 4.7038719382544585e-06,
+      "loss": 0.325,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5334536653402798,
+      "grad_norm": 0.7208976210227123,
+      "learning_rate": 4.702545575027371e-06,
+      "loss": 0.3282,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5335357098904705,
+      "grad_norm": 0.7295543088469668,
+      "learning_rate": 4.701219232806038e-06,
+      "loss": 0.2974,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5336177544406613,
+      "grad_norm": 0.714143762944556,
+      "learning_rate": 4.699892911684123e-06,
+      "loss": 0.3058,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5336997989908521,
+      "grad_norm": 0.879078116163491,
+      "learning_rate": 4.698566611755287e-06,
+      "loss": 0.3171,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5337818435410427,
+      "grad_norm": 0.7111569102057732,
+      "learning_rate": 4.697240333113195e-06,
+      "loss": 0.3098,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5338638880912335,
+      "grad_norm": 0.6382265217833307,
+      "learning_rate": 4.695914075851502e-06,
+      "loss": 0.3087,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5339459326414243,
+      "grad_norm": 0.6500691072382634,
+      "learning_rate": 4.694587840063868e-06,
+      "loss": 0.2996,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5340279771916151,
+      "grad_norm": 0.6669166899945345,
+      "learning_rate": 4.693261625843952e-06,
+      "loss": 0.2822,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5341100217418058,
+      "grad_norm": 0.6764701344101464,
+      "learning_rate": 4.691935433285407e-06,
+      "loss": 0.3334,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5341920662919966,
+      "grad_norm": 0.8404237818386513,
+      "learning_rate": 4.690609262481886e-06,
+      "loss": 0.3078,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5342741108421873,
+      "grad_norm": 0.7661198897625121,
+      "learning_rate": 4.689283113527042e-06,
+      "loss": 0.3397,
+      "step": 6512
+    },
+    {
+      "epoch": 0.534356155392378,
+      "grad_norm": 0.6550702307412601,
+      "learning_rate": 4.6879569865145245e-06,
+      "loss": 0.288,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5344381999425688,
+      "grad_norm": 0.6691834909071003,
+      "learning_rate": 4.686630881537985e-06,
+      "loss": 0.3292,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5345202444927596,
+      "grad_norm": 0.7010073174193141,
+      "learning_rate": 4.685304798691068e-06,
+      "loss": 0.3123,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5346022890429504,
+      "grad_norm": 0.6515762964319776,
+      "learning_rate": 4.6839787380674206e-06,
+      "loss": 0.3479,
+      "step": 6516
+    },
+    {
+      "epoch": 0.534684333593141,
+      "grad_norm": 0.6485335477155159,
+      "learning_rate": 4.682652699760687e-06,
+      "loss": 0.315,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5347663781433318,
+      "grad_norm": 0.729533470897309,
+      "learning_rate": 4.68132668386451e-06,
+      "loss": 0.3246,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5348484226935226,
+      "grad_norm": 0.7350256596165342,
+      "learning_rate": 4.680000690472532e-06,
+      "loss": 0.2955,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5349304672437133,
+      "grad_norm": 0.711989495172929,
+      "learning_rate": 4.678674719678389e-06,
+      "loss": 0.3118,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5350125117939041,
+      "grad_norm": 0.7244283047526591,
+      "learning_rate": 4.677348771575721e-06,
+      "loss": 0.3387,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5350945563440949,
+      "grad_norm": 0.7815322995558449,
+      "learning_rate": 4.676022846258166e-06,
+      "loss": 0.3385,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5351766008942856,
+      "grad_norm": 0.6570595883675087,
+      "learning_rate": 4.674696943819355e-06,
+      "loss": 0.3742,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5352586454444763,
+      "grad_norm": 0.7111641548414062,
+      "learning_rate": 4.6733710643529265e-06,
+      "loss": 0.3134,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5353406899946671,
+      "grad_norm": 0.6547611261046177,
+      "learning_rate": 4.672045207952506e-06,
+      "loss": 0.3367,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5354227345448579,
+      "grad_norm": 1.0412867820418297,
+      "learning_rate": 4.6707193747117256e-06,
+      "loss": 0.301,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5355047790950486,
+      "grad_norm": 0.9507639819477764,
+      "learning_rate": 4.669393564724216e-06,
+      "loss": 0.3129,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5355868236452394,
+      "grad_norm": 0.6806636921318503,
+      "learning_rate": 4.6680677780836005e-06,
+      "loss": 0.3181,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5356688681954301,
+      "grad_norm": 0.6765024457244816,
+      "learning_rate": 4.666742014883506e-06,
+      "loss": 0.337,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5357509127456209,
+      "grad_norm": 0.7837383895325591,
+      "learning_rate": 4.665416275217554e-06,
+      "loss": 0.3472,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5358329572958116,
+      "grad_norm": 0.6243802344764031,
+      "learning_rate": 4.664090559179367e-06,
+      "loss": 0.3277,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5359150018460024,
+      "grad_norm": 0.7030767575038154,
+      "learning_rate": 4.6627648668625666e-06,
+      "loss": 0.3217,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5359970463961932,
+      "grad_norm": 0.6322518123806369,
+      "learning_rate": 4.6614391983607675e-06,
+      "loss": 0.3468,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5360790909463838,
+      "grad_norm": 0.6966745329001286,
+      "learning_rate": 4.660113553767588e-06,
+      "loss": 0.3459,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5361611354965746,
+      "grad_norm": 0.7485775039415573,
+      "learning_rate": 4.6587879331766465e-06,
+      "loss": 0.3711,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5362431800467654,
+      "grad_norm": 0.672108897422256,
+      "learning_rate": 4.6574623366815494e-06,
+      "loss": 0.3291,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5363252245969562,
+      "grad_norm": 0.6180321161867419,
+      "learning_rate": 4.656136764375914e-06,
+      "loss": 0.3142,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5364072691471469,
+      "grad_norm": 0.6491090521966368,
+      "learning_rate": 4.654811216353347e-06,
+      "loss": 0.3124,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5364893136973377,
+      "grad_norm": 0.6842240603278845,
+      "learning_rate": 4.653485692707457e-06,
+      "loss": 0.3202,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5365713582475284,
+      "grad_norm": 0.6129491544914593,
+      "learning_rate": 4.652160193531852e-06,
+      "loss": 0.3081,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5366534027977191,
+      "grad_norm": 0.6478515178515564,
+      "learning_rate": 4.650834718920134e-06,
+      "loss": 0.3336,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5367354473479099,
+      "grad_norm": 0.67484862092707,
+      "learning_rate": 4.649509268965908e-06,
+      "loss": 0.3212,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5368174918981007,
+      "grad_norm": 4.4803802416506855,
+      "learning_rate": 4.648183843762773e-06,
+      "loss": 0.3317,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5368995364482915,
+      "grad_norm": 0.726755809089181,
+      "learning_rate": 4.646858443404331e-06,
+      "loss": 0.3203,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5369815809984821,
+      "grad_norm": 0.6387678064555097,
+      "learning_rate": 4.645533067984178e-06,
+      "loss": 0.3412,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5370636255486729,
+      "grad_norm": 0.7488566910173665,
+      "learning_rate": 4.64420771759591e-06,
+      "loss": 0.2924,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5371456700988637,
+      "grad_norm": 0.7414996183187145,
+      "learning_rate": 4.6428823923331225e-06,
+      "loss": 0.3257,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5372277146490544,
+      "grad_norm": 0.7456565504621376,
+      "learning_rate": 4.641557092289405e-06,
+      "loss": 0.3606,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5373097591992452,
+      "grad_norm": 0.8825089360456805,
+      "learning_rate": 4.6402318175583495e-06,
+      "loss": 0.3527,
+      "step": 6549
+    },
+    {
+      "epoch": 0.537391803749436,
+      "grad_norm": 0.6386515963420552,
+      "learning_rate": 4.638906568233548e-06,
+      "loss": 0.2902,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5374738482996267,
+      "grad_norm": 0.8115718639219985,
+      "learning_rate": 4.637581344408581e-06,
+      "loss": 0.2994,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5375558928498174,
+      "grad_norm": 0.7312280253803255,
+      "learning_rate": 4.636256146177039e-06,
+      "loss": 0.2908,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5376379374000082,
+      "grad_norm": 0.6575124584852915,
+      "learning_rate": 4.634930973632504e-06,
+      "loss": 0.3019,
+      "step": 6553
+    },
+    {
+      "epoch": 0.537719981950199,
+      "grad_norm": 0.7336090946429098,
+      "learning_rate": 4.633605826868556e-06,
+      "loss": 0.3425,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5378020265003897,
+      "grad_norm": 0.6883905358299838,
+      "learning_rate": 4.632280705978777e-06,
+      "loss": 0.2908,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5378840710505804,
+      "grad_norm": 1.0709887603175448,
+      "learning_rate": 4.630955611056742e-06,
+      "loss": 0.3061,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5379661156007712,
+      "grad_norm": 0.6883446794617726,
+      "learning_rate": 4.6296305421960286e-06,
+      "loss": 0.3204,
+      "step": 6557
+    },
+    {
+      "epoch": 0.538048160150962,
+      "grad_norm": 1.1456171379238453,
+      "learning_rate": 4.6283054994902135e-06,
+      "loss": 0.3582,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5381302047011527,
+      "grad_norm": 0.7795959618777809,
+      "learning_rate": 4.626980483032866e-06,
+      "loss": 0.3285,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5382122492513435,
+      "grad_norm": 0.6073082827983459,
+      "learning_rate": 4.625655492917559e-06,
+      "loss": 0.3006,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5382942938015343,
+      "grad_norm": 0.7089134435193638,
+      "learning_rate": 4.624330529237857e-06,
+      "loss": 0.3397,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5383763383517249,
+      "grad_norm": 0.6551710636253657,
+      "learning_rate": 4.6230055920873295e-06,
+      "loss": 0.3846,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5384583829019157,
+      "grad_norm": 0.6850136017563303,
+      "learning_rate": 4.621680681559543e-06,
+      "loss": 0.3452,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5385404274521065,
+      "grad_norm": 0.6970733164515353,
+      "learning_rate": 4.6203557977480575e-06,
+      "loss": 0.3129,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5386224720022973,
+      "grad_norm": 0.7425674882342087,
+      "learning_rate": 4.619030940746438e-06,
+      "loss": 0.324,
+      "step": 6565
+    },
+    {
+      "epoch": 0.538704516552488,
+      "grad_norm": 0.8505817339332054,
+      "learning_rate": 4.6177061106482404e-06,
+      "loss": 0.331,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5387865611026788,
+      "grad_norm": 0.6994951237472067,
+      "learning_rate": 4.616381307547021e-06,
+      "loss": 0.3174,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5388686056528695,
+      "grad_norm": 0.6965895524513226,
+      "learning_rate": 4.61505653153634e-06,
+      "loss": 0.3248,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5389506502030602,
+      "grad_norm": 0.7480340637392122,
+      "learning_rate": 4.613731782709746e-06,
+      "loss": 0.3008,
+      "step": 6569
+    },
+    {
+      "epoch": 0.539032694753251,
+      "grad_norm": 0.730880122438488,
+      "learning_rate": 4.612407061160793e-06,
+      "loss": 0.3428,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5391147393034418,
+      "grad_norm": 0.7724574919827628,
+      "learning_rate": 4.611082366983033e-06,
+      "loss": 0.3063,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5391967838536326,
+      "grad_norm": 0.6688265468478232,
+      "learning_rate": 4.60975770027001e-06,
+      "loss": 0.3285,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5392788284038232,
+      "grad_norm": 0.760268734959991,
+      "learning_rate": 4.608433061115272e-06,
+      "loss": 0.3257,
+      "step": 6573
+    },
+    {
+      "epoch": 0.539360872954014,
+      "grad_norm": 1.0790434683636105,
+      "learning_rate": 4.607108449612361e-06,
+      "loss": 0.3367,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5394429175042048,
+      "grad_norm": 0.6800845461920116,
+      "learning_rate": 4.605783865854819e-06,
+      "loss": 0.3351,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5395249620543955,
+      "grad_norm": 0.8670793073886153,
+      "learning_rate": 4.60445930993619e-06,
+      "loss": 0.3228,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5396070066045863,
+      "grad_norm": 0.8218375835160058,
+      "learning_rate": 4.603134781950007e-06,
+      "loss": 0.3014,
+      "step": 6577
+    },
+    {
+      "epoch": 0.539689051154777,
+      "grad_norm": 0.657048397441641,
+      "learning_rate": 4.60181028198981e-06,
+      "loss": 0.331,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5397710957049678,
+      "grad_norm": 0.7033570566290194,
+      "learning_rate": 4.600485810149128e-06,
+      "loss": 0.6581,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5398531402551585,
+      "grad_norm": 0.7051134799529278,
+      "learning_rate": 4.5991613665214985e-06,
+      "loss": 0.3325,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5399351848053493,
+      "grad_norm": 0.6797760777122236,
+      "learning_rate": 4.59783695120045e-06,
+      "loss": 0.2983,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5400172293555401,
+      "grad_norm": 0.6543910861638469,
+      "learning_rate": 4.5965125642795075e-06,
+      "loss": 0.3232,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5400992739057308,
+      "grad_norm": 1.0036728124221554,
+      "learning_rate": 4.5951882058522015e-06,
+      "loss": 0.3177,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5401813184559215,
+      "grad_norm": 0.6023475955121932,
+      "learning_rate": 4.5938638760120524e-06,
+      "loss": 0.3096,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5402633630061123,
+      "grad_norm": 0.6636462637436831,
+      "learning_rate": 4.592539574852584e-06,
+      "loss": 0.3265,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5403454075563031,
+      "grad_norm": 0.661711445246962,
+      "learning_rate": 4.591215302467318e-06,
+      "loss": 0.3298,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5404274521064938,
+      "grad_norm": 0.7273373996880295,
+      "learning_rate": 4.589891058949769e-06,
+      "loss": 0.3418,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5405094966566846,
+      "grad_norm": 0.7810569987425767,
+      "learning_rate": 4.5885668443934544e-06,
+      "loss": 0.2985,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5405915412068754,
+      "grad_norm": 1.1424902343710133,
+      "learning_rate": 4.58724265889189e-06,
+      "loss": 0.3831,
+      "step": 6589
+    },
+    {
+      "epoch": 0.540673585757066,
+      "grad_norm": 0.8354976116993118,
+      "learning_rate": 4.585918502538583e-06,
+      "loss": 0.3396,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5407556303072568,
+      "grad_norm": 0.6796246701671677,
+      "learning_rate": 4.584594375427049e-06,
+      "loss": 0.3104,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5408376748574476,
+      "grad_norm": 1.23516468452257,
+      "learning_rate": 4.583270277650791e-06,
+      "loss": 0.3233,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5409197194076384,
+      "grad_norm": 0.8085945991237256,
+      "learning_rate": 4.581946209303318e-06,
+      "loss": 0.3337,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5410017639578291,
+      "grad_norm": 0.7133299997737675,
+      "learning_rate": 4.580622170478131e-06,
+      "loss": 0.3398,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5410838085080198,
+      "grad_norm": 0.6575642525959369,
+      "learning_rate": 4.579298161268732e-06,
+      "loss": 0.3169,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5411658530582106,
+      "grad_norm": 0.6837877709936927,
+      "learning_rate": 4.5779741817686226e-06,
+      "loss": 0.323,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5412478976084013,
+      "grad_norm": 0.6941174723096086,
+      "learning_rate": 4.576650232071296e-06,
+      "loss": 0.332,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5413299421585921,
+      "grad_norm": 0.7722575688044551,
+      "learning_rate": 4.575326312270251e-06,
+      "loss": 0.3179,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5414119867087829,
+      "grad_norm": 0.8621510285309543,
+      "learning_rate": 4.57400242245898e-06,
+      "loss": 0.3227,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5414940312589737,
+      "grad_norm": 0.6503589754631622,
+      "learning_rate": 4.5726785627309736e-06,
+      "loss": 0.3319,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5415760758091643,
+      "grad_norm": 0.6422732477065941,
+      "learning_rate": 4.57135473317972e-06,
+      "loss": 0.3159,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5416581203593551,
+      "grad_norm": 0.9469102030730522,
+      "learning_rate": 4.570030933898708e-06,
+      "loss": 0.3438,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5417401649095459,
+      "grad_norm": 0.616392985354242,
+      "learning_rate": 4.568707164981417e-06,
+      "loss": 0.3023,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5418222094597367,
+      "grad_norm": 0.6457926439133652,
+      "learning_rate": 4.567383426521337e-06,
+      "loss": 0.3018,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5419042540099274,
+      "grad_norm": 0.6529895197234888,
+      "learning_rate": 4.566059718611942e-06,
+      "loss": 0.2994,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5419862985601182,
+      "grad_norm": 0.7317614076614148,
+      "learning_rate": 4.564736041346714e-06,
+      "loss": 0.3364,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5420683431103089,
+      "grad_norm": 0.6580076291083967,
+      "learning_rate": 4.563412394819126e-06,
+      "loss": 0.3427,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5421503876604996,
+      "grad_norm": 0.6977790868885471,
+      "learning_rate": 4.5620887791226546e-06,
+      "loss": 0.3396,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5422324322106904,
+      "grad_norm": 0.7756527046015252,
+      "learning_rate": 4.56076519435077e-06,
+      "loss": 0.3381,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5423144767608812,
+      "grad_norm": 0.8152142357488587,
+      "learning_rate": 4.559441640596941e-06,
+      "loss": 0.3317,
+      "step": 6610
+    },
+    {
+      "epoch": 0.542396521311072,
+      "grad_norm": 0.7048124926049529,
+      "learning_rate": 4.558118117954634e-06,
+      "loss": 0.3102,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5424785658612626,
+      "grad_norm": 0.8137955297289524,
+      "learning_rate": 4.5567946265173185e-06,
+      "loss": 0.3187,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5425606104114534,
+      "grad_norm": 0.679643449161701,
+      "learning_rate": 4.555471166378452e-06,
+      "loss": 0.3079,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5426426549616442,
+      "grad_norm": 0.7122022694526992,
+      "learning_rate": 4.5541477376315e-06,
+      "loss": 0.2909,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5427246995118349,
+      "grad_norm": 0.899945066050088,
+      "learning_rate": 4.5528243403699175e-06,
+      "loss": 0.3066,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5428067440620257,
+      "grad_norm": 0.8288936387967779,
+      "learning_rate": 4.5515009746871605e-06,
+      "loss": 0.3192,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5428887886122165,
+      "grad_norm": 0.6461095103289181,
+      "learning_rate": 4.550177640676686e-06,
+      "loss": 0.3096,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5429708331624072,
+      "grad_norm": 0.9888061342687005,
+      "learning_rate": 4.548854338431942e-06,
+      "loss": 0.304,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5430528777125979,
+      "grad_norm": 0.9851580699458711,
+      "learning_rate": 4.5475310680463815e-06,
+      "loss": 0.3494,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5431349222627887,
+      "grad_norm": 0.6952172240892996,
+      "learning_rate": 4.546207829613448e-06,
+      "loss": 0.3334,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5432169668129795,
+      "grad_norm": 0.7383570424046088,
+      "learning_rate": 4.544884623226589e-06,
+      "loss": 0.312,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5432990113631702,
+      "grad_norm": 1.114491713813638,
+      "learning_rate": 4.543561448979249e-06,
+      "loss": 0.3298,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5433810559133609,
+      "grad_norm": 0.9719120930088114,
+      "learning_rate": 4.542238306964863e-06,
+      "loss": 0.3605,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5434631004635517,
+      "grad_norm": 0.9356547788043873,
+      "learning_rate": 4.540915197276873e-06,
+      "loss": 0.3392,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5435451450137425,
+      "grad_norm": 1.3961063832115832,
+      "learning_rate": 4.539592120008716e-06,
+      "loss": 0.3542,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5436271895639332,
+      "grad_norm": 0.6554546061804806,
+      "learning_rate": 4.538269075253822e-06,
+      "loss": 0.341,
+      "step": 6626
+    },
+    {
+      "epoch": 0.543709234114124,
+      "grad_norm": 1.2053652059598776,
+      "learning_rate": 4.5369460631056255e-06,
+      "loss": 0.3164,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5437912786643148,
+      "grad_norm": 0.7000678570825389,
+      "learning_rate": 4.535623083657554e-06,
+      "loss": 0.3176,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5438733232145054,
+      "grad_norm": 0.6400194719206977,
+      "learning_rate": 4.534300137003032e-06,
+      "loss": 0.3071,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5439553677646962,
+      "grad_norm": 0.8617863792600595,
+      "learning_rate": 4.532977223235489e-06,
+      "loss": 0.3218,
+      "step": 6630
+    },
+    {
+      "epoch": 0.544037412314887,
+      "grad_norm": 0.685124944521495,
+      "learning_rate": 4.531654342448342e-06,
+      "loss": 0.3369,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5441194568650778,
+      "grad_norm": 0.7418338856400996,
+      "learning_rate": 4.530331494735014e-06,
+      "loss": 0.3061,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5442015014152685,
+      "grad_norm": 0.9614572526358146,
+      "learning_rate": 4.529008680188921e-06,
+      "loss": 0.2882,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5442835459654592,
+      "grad_norm": 0.6761184297497992,
+      "learning_rate": 4.5276858989034785e-06,
+      "loss": 0.3072,
+      "step": 6634
+    },
+    {
+      "epoch": 0.54436559051565,
+      "grad_norm": 0.6617353268971472,
+      "learning_rate": 4.526363150972098e-06,
+      "loss": 0.3204,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5444476350658407,
+      "grad_norm": 0.6431886015653834,
+      "learning_rate": 4.525040436488192e-06,
+      "loss": 0.2978,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5445296796160315,
+      "grad_norm": 0.6921977004461656,
+      "learning_rate": 4.523717755545169e-06,
+      "loss": 0.3148,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5446117241662223,
+      "grad_norm": 0.6389362104460716,
+      "learning_rate": 4.5223951082364295e-06,
+      "loss": 0.304,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5446937687164131,
+      "grad_norm": 0.6537827326513636,
+      "learning_rate": 4.52107249465538e-06,
+      "loss": 0.3308,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5447758132666037,
+      "grad_norm": 0.6661523847902422,
+      "learning_rate": 4.5197499148954245e-06,
+      "loss": 0.3246,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5448578578167945,
+      "grad_norm": 1.0730574770050287,
+      "learning_rate": 4.518427369049955e-06,
+      "loss": 0.3222,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5449399023669853,
+      "grad_norm": 0.6608812420357539,
+      "learning_rate": 4.517104857212373e-06,
+      "loss": 0.3254,
+      "step": 6642
+    },
+    {
+      "epoch": 0.545021946917176,
+      "grad_norm": 0.739663139133177,
+      "learning_rate": 4.515782379476071e-06,
+      "loss": 0.3266,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5451039914673668,
+      "grad_norm": 0.7093361686896349,
+      "learning_rate": 4.514459935934438e-06,
+      "loss": 0.3537,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5451860360175576,
+      "grad_norm": 0.8330795751905851,
+      "learning_rate": 4.513137526680865e-06,
+      "loss": 0.2995,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5452680805677483,
+      "grad_norm": 0.9528699593062254,
+      "learning_rate": 4.511815151808737e-06,
+      "loss": 0.3136,
+      "step": 6646
+    },
+    {
+      "epoch": 0.545350125117939,
+      "grad_norm": 0.8866474588382535,
+      "learning_rate": 4.510492811411438e-06,
+      "loss": 0.301,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5454321696681298,
+      "grad_norm": 0.6826434203615505,
+      "learning_rate": 4.509170505582353e-06,
+      "loss": 0.3511,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5455142142183206,
+      "grad_norm": 0.8345956901151385,
+      "learning_rate": 4.507848234414856e-06,
+      "loss": 0.3189,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5455962587685113,
+      "grad_norm": 0.7823195407302365,
+      "learning_rate": 4.506525998002327e-06,
+      "loss": 0.3262,
+      "step": 6650
+    },
+    {
+      "epoch": 0.545678303318702,
+      "grad_norm": 0.711771287877461,
+      "learning_rate": 4.505203796438138e-06,
+      "loss": 0.3309,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5457603478688928,
+      "grad_norm": 0.7251761480061573,
+      "learning_rate": 4.503881629815662e-06,
+      "loss": 0.3357,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5458423924190836,
+      "grad_norm": 0.6801426136197581,
+      "learning_rate": 4.50255949822827e-06,
+      "loss": 0.3374,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5459244369692743,
+      "grad_norm": 0.7118957993259311,
+      "learning_rate": 4.501237401769324e-06,
+      "loss": 0.3106,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5460064815194651,
+      "grad_norm": 0.6855557638236704,
+      "learning_rate": 4.499915340532195e-06,
+      "loss": 0.3644,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5460885260696559,
+      "grad_norm": 0.7206692307428746,
+      "learning_rate": 4.4985933146102395e-06,
+      "loss": 0.3346,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5461705706198465,
+      "grad_norm": 0.6730257820120866,
+      "learning_rate": 4.497271324096817e-06,
+      "loss": 0.3399,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5462526151700373,
+      "grad_norm": 0.6609151538220458,
+      "learning_rate": 4.495949369085289e-06,
+      "loss": 0.3473,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5463346597202281,
+      "grad_norm": 0.654405449148218,
+      "learning_rate": 4.494627449669004e-06,
+      "loss": 0.3148,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5464167042704189,
+      "grad_norm": 1.0082717496658828,
+      "learning_rate": 4.4933055659413175e-06,
+      "loss": 0.2668,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5464987488206096,
+      "grad_norm": 0.744989151729109,
+      "learning_rate": 4.4919837179955784e-06,
+      "loss": 0.3631,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5465807933708003,
+      "grad_norm": 0.6493334677128881,
+      "learning_rate": 4.490661905925133e-06,
+      "loss": 0.3362,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5466628379209911,
+      "grad_norm": 0.6868609117274248,
+      "learning_rate": 4.489340129823325e-06,
+      "loss": 0.3197,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5467448824711818,
+      "grad_norm": 0.8072744691443682,
+      "learning_rate": 4.488018389783497e-06,
+      "loss": 0.3479,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5468269270213726,
+      "grad_norm": 0.6777937321916154,
+      "learning_rate": 4.486696685898986e-06,
+      "loss": 0.3326,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5469089715715634,
+      "grad_norm": 0.784364048118098,
+      "learning_rate": 4.485375018263133e-06,
+      "loss": 0.3503,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5469910161217542,
+      "grad_norm": 0.8761956018731701,
+      "learning_rate": 4.484053386969267e-06,
+      "loss": 0.3167,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5470730606719448,
+      "grad_norm": 0.724389682123803,
+      "learning_rate": 4.4827317921107246e-06,
+      "loss": 0.3478,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5471551052221356,
+      "grad_norm": 0.6632040821972258,
+      "learning_rate": 4.48141023378083e-06,
+      "loss": 0.29,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5472371497723264,
+      "grad_norm": 0.6922198650856087,
+      "learning_rate": 4.480088712072911e-06,
+      "loss": 0.3183,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5473191943225171,
+      "grad_norm": 0.7466822495656646,
+      "learning_rate": 4.4787672270802945e-06,
+      "loss": 0.3325,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5474012388727079,
+      "grad_norm": 0.6615232563839297,
+      "learning_rate": 4.477445778896296e-06,
+      "loss": 0.2939,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5474832834228986,
+      "grad_norm": 0.9557263149265696,
+      "learning_rate": 4.476124367614237e-06,
+      "loss": 0.3059,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5475653279730894,
+      "grad_norm": 0.7560678458323987,
+      "learning_rate": 4.474802993327437e-06,
+      "loss": 0.2868,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5476473725232801,
+      "grad_norm": 0.708755279442068,
+      "learning_rate": 4.473481656129202e-06,
+      "loss": 0.3744,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5477294170734709,
+      "grad_norm": 0.6570744530048801,
+      "learning_rate": 4.472160356112849e-06,
+      "loss": 0.2864,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5478114616236617,
+      "grad_norm": 0.731198301230586,
+      "learning_rate": 4.470839093371684e-06,
+      "loss": 0.3409,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5478935061738524,
+      "grad_norm": 0.6965629560438585,
+      "learning_rate": 4.469517867999009e-06,
+      "loss": 0.3246,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5479755507240431,
+      "grad_norm": 0.8183525515971455,
+      "learning_rate": 4.4681966800881334e-06,
+      "loss": 0.3393,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5480575952742339,
+      "grad_norm": 0.9071918057208048,
+      "learning_rate": 4.4668755297323515e-06,
+      "loss": 0.3185,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5481396398244247,
+      "grad_norm": 0.741037494976712,
+      "learning_rate": 4.465554417024965e-06,
+      "loss": 0.3855,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5482216843746154,
+      "grad_norm": 0.8050012637202707,
+      "learning_rate": 4.464233342059264e-06,
+      "loss": 0.3299,
+      "step": 6682
+    },
+    {
+      "epoch": 0.5483037289248062,
+      "grad_norm": 0.6740747319463376,
+      "learning_rate": 4.462912304928546e-06,
+      "loss": 0.2782,
+      "step": 6683
+    },
+    {
+      "epoch": 0.548385773474997,
+      "grad_norm": 0.6674335800787131,
+      "learning_rate": 4.4615913057260986e-06,
+      "loss": 0.3174,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5484678180251876,
+      "grad_norm": 0.8078423242041717,
+      "learning_rate": 4.460270344545206e-06,
+      "loss": 0.3634,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5485498625753784,
+      "grad_norm": 0.645242762389889,
+      "learning_rate": 4.458949421479157e-06,
+      "loss": 0.3336,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5486319071255692,
+      "grad_norm": 3.593996931493718,
+      "learning_rate": 4.457628536621228e-06,
+      "loss": 0.3468,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54871395167576,
+      "grad_norm": 0.659171277611106,
+      "learning_rate": 4.4563076900647e-06,
+      "loss": 0.2979,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5487959962259507,
+      "grad_norm": 0.7168289246697148,
+      "learning_rate": 4.454986881902853e-06,
+      "loss": 0.3139,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5488780407761414,
+      "grad_norm": 0.7853451511678374,
+      "learning_rate": 4.453666112228955e-06,
+      "loss": 0.3161,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5489600853263322,
+      "grad_norm": 0.9900074292994447,
+      "learning_rate": 4.4523453811362765e-06,
+      "loss": 0.3105,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5490421298765229,
+      "grad_norm": 0.8189521398955933,
+      "learning_rate": 4.45102468871809e-06,
+      "loss": 0.3039,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5491241744267137,
+      "grad_norm": 0.7061003194962877,
+      "learning_rate": 4.449704035067656e-06,
+      "loss": 0.3203,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5492062189769045,
+      "grad_norm": 0.9435476808302292,
+      "learning_rate": 4.448383420278241e-06,
+      "loss": 0.346,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5492882635270953,
+      "grad_norm": 0.690581851337164,
+      "learning_rate": 4.4470628444431e-06,
+      "loss": 0.3018,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5493703080772859,
+      "grad_norm": 0.6472165663482872,
+      "learning_rate": 4.445742307655495e-06,
+      "loss": 0.3643,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5494523526274767,
+      "grad_norm": 0.7541689710722681,
+      "learning_rate": 4.4444218100086765e-06,
+      "loss": 0.3207,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5495343971776675,
+      "grad_norm": 0.7686037909618954,
+      "learning_rate": 4.443101351595897e-06,
+      "loss": 0.3338,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5496164417278582,
+      "grad_norm": 0.6885075155305106,
+      "learning_rate": 4.441780932510408e-06,
+      "loss": 0.3798,
+      "step": 6699
+    },
+    {
+      "epoch": 0.549698486278049,
+      "grad_norm": 0.7860218327135049,
+      "learning_rate": 4.440460552845449e-06,
+      "loss": 0.3105,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5497805308282397,
+      "grad_norm": 0.663656667139125,
+      "learning_rate": 4.439140212694267e-06,
+      "loss": 0.3337,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5498625753784305,
+      "grad_norm": 0.6483270298335215,
+      "learning_rate": 4.437819912150106e-06,
+      "loss": 0.3377,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5499446199286212,
+      "grad_norm": 0.7096493246330439,
+      "learning_rate": 4.436499651306195e-06,
+      "loss": 0.3249,
+      "step": 6703
+    },
+    {
+      "epoch": 0.550026664478812,
+      "grad_norm": 1.3572510588438054,
+      "learning_rate": 4.4351794302557765e-06,
+      "loss": 0.3385,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5501087090290028,
+      "grad_norm": 0.6547384999574851,
+      "learning_rate": 4.4338592490920775e-06,
+      "loss": 0.3372,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5501907535791934,
+      "grad_norm": 0.715939070206566,
+      "learning_rate": 4.432539107908329e-06,
+      "loss": 0.3305,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5502727981293842,
+      "grad_norm": 0.6517453038810681,
+      "learning_rate": 4.431219006797758e-06,
+      "loss": 0.3398,
+      "step": 6707
+    },
+    {
+      "epoch": 0.550354842679575,
+      "grad_norm": 0.7254858949900578,
+      "learning_rate": 4.429898945853585e-06,
+      "loss": 0.3452,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5504368872297658,
+      "grad_norm": 0.6930645994314034,
+      "learning_rate": 4.428578925169033e-06,
+      "loss": 0.306,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5505189317799565,
+      "grad_norm": 0.7031778482582536,
+      "learning_rate": 4.427258944837321e-06,
+      "loss": 0.3174,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5506009763301473,
+      "grad_norm": 1.2189381894609415,
+      "learning_rate": 4.425939004951661e-06,
+      "loss": 0.3447,
+      "step": 6711
+    },
+    {
+      "epoch": 0.550683020880338,
+      "grad_norm": 0.6703564210035865,
+      "learning_rate": 4.424619105605267e-06,
+      "loss": 0.3316,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5507650654305287,
+      "grad_norm": 0.6344969656559866,
+      "learning_rate": 4.423299246891346e-06,
+      "loss": 0.3066,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5508471099807195,
+      "grad_norm": 0.6972037378701752,
+      "learning_rate": 4.4219794289031045e-06,
+      "loss": 0.3092,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5509291545309103,
+      "grad_norm": 0.8036992025463134,
+      "learning_rate": 4.420659651733751e-06,
+      "loss": 0.2955,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5510111990811011,
+      "grad_norm": 0.7039491486923808,
+      "learning_rate": 4.419339915476478e-06,
+      "loss": 0.2755,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5510932436312918,
+      "grad_norm": 0.6365439227696155,
+      "learning_rate": 4.418020220224491e-06,
+      "loss": 0.3328,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5511752881814825,
+      "grad_norm": 0.9468601774605638,
+      "learning_rate": 4.416700566070979e-06,
+      "loss": 0.3123,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5512573327316733,
+      "grad_norm": 0.6884106387113409,
+      "learning_rate": 4.415380953109135e-06,
+      "loss": 0.2963,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5513393772818641,
+      "grad_norm": 0.7056247085879479,
+      "learning_rate": 4.4140613814321505e-06,
+      "loss": 0.3256,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5514214218320548,
+      "grad_norm": 1.1215063428427814,
+      "learning_rate": 4.412741851133209e-06,
+      "loss": 0.3119,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5515034663822456,
+      "grad_norm": 0.7166992462544531,
+      "learning_rate": 4.411422362305495e-06,
+      "loss": 0.3552,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5515855109324364,
+      "grad_norm": 0.6369228781484115,
+      "learning_rate": 4.410102915042187e-06,
+      "loss": 0.3421,
+      "step": 6723
+    },
+    {
+      "epoch": 0.551667555482627,
+      "grad_norm": 0.9434176081791087,
+      "learning_rate": 4.4087835094364645e-06,
+      "loss": 0.3248,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5517496000328178,
+      "grad_norm": 0.6279072983283516,
+      "learning_rate": 4.407464145581501e-06,
+      "loss": 0.3074,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5518316445830086,
+      "grad_norm": 0.9138458016388096,
+      "learning_rate": 4.406144823570467e-06,
+      "loss": 0.3321,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5519136891331994,
+      "grad_norm": 0.6866479284255324,
+      "learning_rate": 4.4048255434965305e-06,
+      "loss": 0.3223,
+      "step": 6727
+    },
+    {
+      "epoch": 0.55199573368339,
+      "grad_norm": 0.83961677554453,
+      "learning_rate": 4.40350630545286e-06,
+      "loss": 0.3182,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5520777782335808,
+      "grad_norm": 0.6870461811563392,
+      "learning_rate": 4.402187109532613e-06,
+      "loss": 0.3191,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5521598227837716,
+      "grad_norm": 0.8083947243482414,
+      "learning_rate": 4.400867955828955e-06,
+      "loss": 0.3499,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5522418673339623,
+      "grad_norm": 1.092232046994438,
+      "learning_rate": 4.399548844435037e-06,
+      "loss": 0.3221,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5523239118841531,
+      "grad_norm": 0.7652380165480627,
+      "learning_rate": 4.398229775444016e-06,
+      "loss": 0.3299,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5524059564343439,
+      "grad_norm": 0.7414088503027958,
+      "learning_rate": 4.396910748949043e-06,
+      "loss": 0.3469,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5524880009845347,
+      "grad_norm": 0.713425595024947,
+      "learning_rate": 4.395591765043261e-06,
+      "loss": 0.3324,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5525700455347253,
+      "grad_norm": 0.6979651575013641,
+      "learning_rate": 4.39427282381982e-06,
+      "loss": 0.2983,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5526520900849161,
+      "grad_norm": 0.6900268606288174,
+      "learning_rate": 4.3929539253718565e-06,
+      "loss": 0.3251,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5527341346351069,
+      "grad_norm": 0.7035175844752444,
+      "learning_rate": 4.391635069792512e-06,
+      "loss": 0.3161,
+      "step": 6737
+    },
+    {
+      "epoch": 0.5528161791852976,
+      "grad_norm": 0.6874203651610159,
+      "learning_rate": 4.3903162571749234e-06,
+      "loss": 0.3362,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5528982237354884,
+      "grad_norm": 0.650741466047761,
+      "learning_rate": 4.388997487612221e-06,
+      "loss": 0.3305,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5529802682856791,
+      "grad_norm": 0.8554196030194147,
+      "learning_rate": 4.387678761197534e-06,
+      "loss": 0.3141,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5530623128358699,
+      "grad_norm": 0.7414369907438002,
+      "learning_rate": 4.386360078023987e-06,
+      "loss": 0.3335,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5531443573860606,
+      "grad_norm": 0.7447946106280197,
+      "learning_rate": 4.385041438184706e-06,
+      "loss": 0.3137,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5532264019362514,
+      "grad_norm": 0.6206666292183329,
+      "learning_rate": 4.383722841772813e-06,
+      "loss": 0.3175,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5533084464864422,
+      "grad_norm": 0.6664525873469053,
+      "learning_rate": 4.38240428888142e-06,
+      "loss": 0.3135,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5533904910366328,
+      "grad_norm": 0.8143191440183999,
+      "learning_rate": 4.381085779603645e-06,
+      "loss": 0.335,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5534725355868236,
+      "grad_norm": 0.7150881547476697,
+      "learning_rate": 4.3797673140325995e-06,
+      "loss": 0.3784,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5535545801370144,
+      "grad_norm": 0.746344164664304,
+      "learning_rate": 4.378448892261388e-06,
+      "loss": 0.3179,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5536366246872052,
+      "grad_norm": 0.7474174221355758,
+      "learning_rate": 4.3771305143831184e-06,
+      "loss": 0.3047,
+      "step": 6748
+    },
+    {
+      "epoch": 0.5537186692373959,
+      "grad_norm": 0.6753173043945567,
+      "learning_rate": 4.37581218049089e-06,
+      "loss": 0.3299,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5538007137875867,
+      "grad_norm": 0.6552664402630239,
+      "learning_rate": 4.3744938906778016e-06,
+      "loss": 0.316,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5538827583377774,
+      "grad_norm": 0.7701367013280838,
+      "learning_rate": 4.373175645036951e-06,
+      "loss": 0.3181,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5539648028879681,
+      "grad_norm": 0.767636563079807,
+      "learning_rate": 4.3718574436614306e-06,
+      "loss": 0.315,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5540468474381589,
+      "grad_norm": 0.6390116525099653,
+      "learning_rate": 4.3705392866443264e-06,
+      "loss": 0.2963,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5541288919883497,
+      "grad_norm": 0.783074806003217,
+      "learning_rate": 4.369221174078727e-06,
+      "loss": 0.3182,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5542109365385405,
+      "grad_norm": 0.7439565534637474,
+      "learning_rate": 4.3679031060577145e-06,
+      "loss": 0.3236,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5542929810887312,
+      "grad_norm": 4.544919592270904,
+      "learning_rate": 4.366585082674371e-06,
+      "loss": 0.3414,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5543750256389219,
+      "grad_norm": 0.6669524268176537,
+      "learning_rate": 4.365267104021768e-06,
+      "loss": 0.296,
+      "step": 6757
+    },
+    {
+      "epoch": 0.5544570701891127,
+      "grad_norm": 0.6783630995061904,
+      "learning_rate": 4.3639491701929856e-06,
+      "loss": 0.3042,
+      "step": 6758
+    },
+    {
+      "epoch": 0.5545391147393034,
+      "grad_norm": 0.7120198766094002,
+      "learning_rate": 4.3626312812810885e-06,
+      "loss": 0.3391,
+      "step": 6759
+    },
+    {
+      "epoch": 0.5546211592894942,
+      "grad_norm": 0.6151703650681069,
+      "learning_rate": 4.361313437379147e-06,
+      "loss": 0.3553,
+      "step": 6760
+    },
+    {
+      "epoch": 0.554703203839685,
+      "grad_norm": 0.6554463651048029,
+      "learning_rate": 4.359995638580226e-06,
+      "loss": 0.2872,
+      "step": 6761
+    },
+    {
+      "epoch": 0.5547852483898758,
+      "grad_norm": 0.6758930887896354,
+      "learning_rate": 4.358677884977382e-06,
+      "loss": 0.3308,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5548672929400664,
+      "grad_norm": 0.7306655508260328,
+      "learning_rate": 4.357360176663676e-06,
+      "loss": 0.3286,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5549493374902572,
+      "grad_norm": 0.7184733613961963,
+      "learning_rate": 4.356042513732164e-06,
+      "loss": 0.3444,
+      "step": 6764
+    },
+    {
+      "epoch": 0.555031382040448,
+      "grad_norm": 0.7504463210300375,
+      "learning_rate": 4.354724896275893e-06,
+      "loss": 0.3307,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5551134265906387,
+      "grad_norm": 0.6768283398430935,
+      "learning_rate": 4.353407324387915e-06,
+      "loss": 0.3178,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5551954711408295,
+      "grad_norm": 0.9057608164981166,
+      "learning_rate": 4.352089798161272e-06,
+      "loss": 0.3232,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5552775156910202,
+      "grad_norm": 0.6368800374256393,
+      "learning_rate": 4.350772317689006e-06,
+      "loss": 0.3121,
+      "step": 6768
+    },
+    {
+      "epoch": 0.555359560241211,
+      "grad_norm": 0.8171091814062101,
+      "learning_rate": 4.349454883064157e-06,
+      "loss": 0.2913,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5554416047914017,
+      "grad_norm": 0.7450638831837604,
+      "learning_rate": 4.348137494379757e-06,
+      "loss": 0.358,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5555236493415925,
+      "grad_norm": 1.012944999711811,
+      "learning_rate": 4.346820151728843e-06,
+      "loss": 0.3299,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5556056938917833,
+      "grad_norm": 0.6605896848650747,
+      "learning_rate": 4.3455028552044365e-06,
+      "loss": 0.341,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5556877384419739,
+      "grad_norm": 0.8711210511661212,
+      "learning_rate": 4.344185604899569e-06,
+      "loss": 0.384,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5557697829921647,
+      "grad_norm": 0.8862944423937736,
+      "learning_rate": 4.342868400907261e-06,
+      "loss": 0.3487,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5558518275423555,
+      "grad_norm": 0.7183026345314314,
+      "learning_rate": 4.3415512433205295e-06,
+      "loss": 0.3279,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5559338720925463,
+      "grad_norm": 0.663103725787822,
+      "learning_rate": 4.340234132232393e-06,
+      "loss": 0.3365,
+      "step": 6776
+    },
+    {
+      "epoch": 0.556015916642737,
+      "grad_norm": 0.7660237666187014,
+      "learning_rate": 4.33891706773586e-06,
+      "loss": 0.3376,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5560979611929278,
+      "grad_norm": 0.7957621717766968,
+      "learning_rate": 4.337600049923941e-06,
+      "loss": 0.319,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5561800057431185,
+      "grad_norm": 0.7014654055321247,
+      "learning_rate": 4.336283078889646e-06,
+      "loss": 0.341,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5562620502933092,
+      "grad_norm": 0.796448260916141,
+      "learning_rate": 4.334966154725971e-06,
+      "loss": 0.3983,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5563440948435,
+      "grad_norm": 1.6259367917739354,
+      "learning_rate": 4.333649277525917e-06,
+      "loss": 0.332,
+      "step": 6781
+    },
+    {
+      "epoch": 0.5564261393936908,
+      "grad_norm": 0.7014969440594874,
+      "learning_rate": 4.3323324473824825e-06,
+      "loss": 0.3212,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5565081839438816,
+      "grad_norm": 0.8028131508440614,
+      "learning_rate": 4.331015664388656e-06,
+      "loss": 0.336,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5565902284940722,
+      "grad_norm": 0.8017512184672216,
+      "learning_rate": 4.32969892863743e-06,
+      "loss": 0.3464,
+      "step": 6784
+    },
+    {
+      "epoch": 0.556672273044263,
+      "grad_norm": 0.6308143361162645,
+      "learning_rate": 4.328382240221788e-06,
+      "loss": 0.3036,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5567543175944538,
+      "grad_norm": 0.6973326866573275,
+      "learning_rate": 4.327065599234713e-06,
+      "loss": 0.3725,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5568363621446445,
+      "grad_norm": 6.689959803520651,
+      "learning_rate": 4.325749005769184e-06,
+      "loss": 0.3482,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5569184066948353,
+      "grad_norm": 0.7276283989345558,
+      "learning_rate": 4.324432459918178e-06,
+      "loss": 0.3129,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5570004512450261,
+      "grad_norm": 0.6653614519250868,
+      "learning_rate": 4.3231159617746675e-06,
+      "loss": 0.3262,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5570824957952168,
+      "grad_norm": 0.7263986862842099,
+      "learning_rate": 4.321799511431617e-06,
+      "loss": 0.2929,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5571645403454075,
+      "grad_norm": 0.6412680460596641,
+      "learning_rate": 4.320483108981996e-06,
+      "loss": 0.279,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5572465848955983,
+      "grad_norm": 0.8053925148021125,
+      "learning_rate": 4.319166754518768e-06,
+      "loss": 0.3388,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5573286294457891,
+      "grad_norm": 0.7116719583350045,
+      "learning_rate": 4.317850448134888e-06,
+      "loss": 0.3327,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5574106739959798,
+      "grad_norm": 0.6647307520934768,
+      "learning_rate": 4.316534189923315e-06,
+      "loss": 0.3127,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5574927185461706,
+      "grad_norm": 0.746945178475027,
+      "learning_rate": 4.315217979976999e-06,
+      "loss": 0.3092,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5575747630963613,
+      "grad_norm": 0.661592463072802,
+      "learning_rate": 4.313901818388888e-06,
+      "loss": 0.312,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5576568076465521,
+      "grad_norm": 0.7398650548993734,
+      "learning_rate": 4.312585705251931e-06,
+      "loss": 0.2852,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5577388521967428,
+      "grad_norm": 0.6654821850530835,
+      "learning_rate": 4.311269640659063e-06,
+      "loss": 0.3138,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5578208967469336,
+      "grad_norm": 0.6355507872322262,
+      "learning_rate": 4.3099536247032285e-06,
+      "loss": 0.3464,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5579029412971244,
+      "grad_norm": 0.6732619581316067,
+      "learning_rate": 4.308637657477362e-06,
+      "loss": 0.3324,
+      "step": 6800
+    },
+    {
+      "epoch": 0.557984985847315,
+      "grad_norm": 0.907297030361115,
+      "learning_rate": 4.307321739074392e-06,
+      "loss": 0.2885,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5580670303975058,
+      "grad_norm": 0.7759586826718257,
+      "learning_rate": 4.306005869587249e-06,
+      "loss": 0.3095,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5581490749476966,
+      "grad_norm": 0.7250309118361513,
+      "learning_rate": 4.304690049108854e-06,
+      "loss": 0.306,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5582311194978874,
+      "grad_norm": 0.7047117870624263,
+      "learning_rate": 4.303374277732131e-06,
+      "loss": 0.3857,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5583131640480781,
+      "grad_norm": 0.6799472203320427,
+      "learning_rate": 4.302058555549999e-06,
+      "loss": 0.3229,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5583952085982689,
+      "grad_norm": 0.6501043474394151,
+      "learning_rate": 4.300742882655369e-06,
+      "loss": 0.2889,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5584772531484596,
+      "grad_norm": 0.7627032870964416,
+      "learning_rate": 4.299427259141155e-06,
+      "loss": 0.3136,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5585592976986503,
+      "grad_norm": 0.7238211399602147,
+      "learning_rate": 4.298111685100262e-06,
+      "loss": 0.3396,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5586413422488411,
+      "grad_norm": 0.6236377983464992,
+      "learning_rate": 4.296796160625593e-06,
+      "loss": 0.2945,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5587233867990319,
+      "grad_norm": 0.6442571263173138,
+      "learning_rate": 4.2954806858100506e-06,
+      "loss": 0.3386,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5588054313492227,
+      "grad_norm": 0.6680561874282331,
+      "learning_rate": 4.294165260746528e-06,
+      "loss": 0.3083,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5588874758994133,
+      "grad_norm": 0.6603990690462882,
+      "learning_rate": 4.292849885527924e-06,
+      "loss": 0.3257,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5589695204496041,
+      "grad_norm": 0.7796874091313046,
+      "learning_rate": 4.291534560247121e-06,
+      "loss": 0.3294,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5590515649997949,
+      "grad_norm": 0.7084962300668426,
+      "learning_rate": 4.290219284997012e-06,
+      "loss": 0.3124,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5591336095499856,
+      "grad_norm": 0.6404764298377381,
+      "learning_rate": 4.288904059870476e-06,
+      "loss": 0.355,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5592156541001764,
+      "grad_norm": 0.7699553145228769,
+      "learning_rate": 4.287588884960393e-06,
+      "loss": 0.3508,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5592976986503672,
+      "grad_norm": 0.6833341987595396,
+      "learning_rate": 4.286273760359637e-06,
+      "loss": 0.3005,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5593797432005579,
+      "grad_norm": 0.7518361291469017,
+      "learning_rate": 4.2849586861610835e-06,
+      "loss": 0.2987,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5594617877507486,
+      "grad_norm": 0.8195673866346911,
+      "learning_rate": 4.283643662457597e-06,
+      "loss": 0.3073,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5595438323009394,
+      "grad_norm": 0.6874986056079481,
+      "learning_rate": 4.282328689342046e-06,
+      "loss": 0.3117,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5596258768511302,
+      "grad_norm": 0.5572900015973375,
+      "learning_rate": 4.281013766907288e-06,
+      "loss": 0.6142,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5597079214013209,
+      "grad_norm": 0.6699929839338844,
+      "learning_rate": 4.279698895246184e-06,
+      "loss": 0.3297,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5597899659515116,
+      "grad_norm": 0.6959382836610806,
+      "learning_rate": 4.278384074451587e-06,
+      "loss": 0.297,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5598720105017024,
+      "grad_norm": 0.796813582460204,
+      "learning_rate": 4.277069304616346e-06,
+      "loss": 0.3282,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5599540550518932,
+      "grad_norm": 1.145189394838764,
+      "learning_rate": 4.275754585833311e-06,
+      "loss": 0.3316,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5600360996020839,
+      "grad_norm": 0.6550184053677593,
+      "learning_rate": 4.274439918195322e-06,
+      "loss": 0.314,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5601181441522747,
+      "grad_norm": 0.6903948870748502,
+      "learning_rate": 4.273125301795219e-06,
+      "loss": 0.3135,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5602001887024655,
+      "grad_norm": 0.6704743738984165,
+      "learning_rate": 4.271810736725842e-06,
+      "loss": 0.3444,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5602822332526562,
+      "grad_norm": 0.6603022642647622,
+      "learning_rate": 4.2704962230800205e-06,
+      "loss": 0.3236,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5603642778028469,
+      "grad_norm": 0.8605633271129273,
+      "learning_rate": 4.269181760950584e-06,
+      "loss": 0.3255,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5604463223530377,
+      "grad_norm": 0.8923613014761076,
+      "learning_rate": 4.267867350430356e-06,
+      "loss": 0.3363,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5605283669032285,
+      "grad_norm": 0.775499787407987,
+      "learning_rate": 4.2665529916121585e-06,
+      "loss": 0.2877,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5606104114534192,
+      "grad_norm": 0.6656153135256929,
+      "learning_rate": 4.265238684588813e-06,
+      "loss": 0.3081,
+      "step": 6833
+    },
+    {
+      "epoch": 0.56069245600361,
+      "grad_norm": 0.6543311238046932,
+      "learning_rate": 4.263924429453129e-06,
+      "loss": 0.3363,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5607745005538007,
+      "grad_norm": 0.800162234181209,
+      "learning_rate": 4.26261022629792e-06,
+      "loss": 0.3476,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5608565451039915,
+      "grad_norm": 0.8446547909133654,
+      "learning_rate": 4.261296075215994e-06,
+      "loss": 0.3768,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5609385896541822,
+      "grad_norm": 0.7075802852997959,
+      "learning_rate": 4.259981976300149e-06,
+      "loss": 0.3521,
+      "step": 6837
+    },
+    {
+      "epoch": 0.561020634204373,
+      "grad_norm": 0.6712423595523062,
+      "learning_rate": 4.258667929643191e-06,
+      "loss": 0.2995,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5611026787545638,
+      "grad_norm": 0.755454797816162,
+      "learning_rate": 4.25735393533791e-06,
+      "loss": 0.3248,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5611847233047544,
+      "grad_norm": 0.8129621324314413,
+      "learning_rate": 4.256039993477102e-06,
+      "loss": 0.3102,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5612667678549452,
+      "grad_norm": 0.6959783034399186,
+      "learning_rate": 4.254726104153555e-06,
+      "loss": 0.372,
+      "step": 6841
+    },
+    {
+      "epoch": 0.561348812405136,
+      "grad_norm": 0.7261323471314025,
+      "learning_rate": 4.253412267460053e-06,
+      "loss": 0.3085,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5614308569553268,
+      "grad_norm": 0.7336270795569216,
+      "learning_rate": 4.252098483489377e-06,
+      "loss": 0.3174,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5615129015055175,
+      "grad_norm": 0.6248281053073925,
+      "learning_rate": 4.250784752334303e-06,
+      "loss": 0.2941,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5615949460557083,
+      "grad_norm": 0.6276555505295975,
+      "learning_rate": 4.249471074087607e-06,
+      "loss": 0.3168,
+      "step": 6845
+    },
+    {
+      "epoch": 0.561676990605899,
+      "grad_norm": 0.7456016462780767,
+      "learning_rate": 4.24815744884206e-06,
+      "loss": 0.3379,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5617590351560897,
+      "grad_norm": 0.8509309905542873,
+      "learning_rate": 4.246843876690423e-06,
+      "loss": 0.3355,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5618410797062805,
+      "grad_norm": 0.6835580400363951,
+      "learning_rate": 4.245530357725464e-06,
+      "loss": 0.3186,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5619231242564713,
+      "grad_norm": 0.6613420701434356,
+      "learning_rate": 4.244216892039937e-06,
+      "loss": 0.2995,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5620051688066621,
+      "grad_norm": 0.723793586389631,
+      "learning_rate": 4.2429034797266e-06,
+      "loss": 0.3255,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5620872133568527,
+      "grad_norm": 0.7019117594071697,
+      "learning_rate": 4.241590120878204e-06,
+      "loss": 0.3016,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5621692579070435,
+      "grad_norm": 0.6741312451406342,
+      "learning_rate": 4.240276815587492e-06,
+      "loss": 0.363,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5622513024572343,
+      "grad_norm": 0.646643911130416,
+      "learning_rate": 4.238963563947212e-06,
+      "loss": 0.3056,
+      "step": 6853
+    },
+    {
+      "epoch": 0.562333347007425,
+      "grad_norm": 0.6816958224016055,
+      "learning_rate": 4.2376503660501035e-06,
+      "loss": 0.3576,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5624153915576158,
+      "grad_norm": 0.7768009292638507,
+      "learning_rate": 4.2363372219889e-06,
+      "loss": 0.2989,
+      "step": 6855
+    },
+    {
+      "epoch": 0.5624974361078066,
+      "grad_norm": 0.6870493202698912,
+      "learning_rate": 4.235024131856336e-06,
+      "loss": 0.3517,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5625794806579973,
+      "grad_norm": 0.6580774915842689,
+      "learning_rate": 4.233711095745139e-06,
+      "loss": 0.3102,
+      "step": 6857
+    },
+    {
+      "epoch": 0.562661525208188,
+      "grad_norm": 0.723627538302717,
+      "learning_rate": 4.2323981137480315e-06,
+      "loss": 0.3206,
+      "step": 6858
+    },
+    {
+      "epoch": 0.5627435697583788,
+      "grad_norm": 0.7320721801722306,
+      "learning_rate": 4.231085185957739e-06,
+      "loss": 0.3133,
+      "step": 6859
+    },
+    {
+      "epoch": 0.5628256143085696,
+      "grad_norm": 0.677016491155214,
+      "learning_rate": 4.229772312466973e-06,
+      "loss": 0.3313,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5629076588587603,
+      "grad_norm": 0.6540499773395221,
+      "learning_rate": 4.22845949336845e-06,
+      "loss": 0.3027,
+      "step": 6861
+    },
+    {
+      "epoch": 0.562989703408951,
+      "grad_norm": 0.9353417541471516,
+      "learning_rate": 4.2271467287548764e-06,
+      "loss": 0.3297,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5630717479591418,
+      "grad_norm": 0.6869921656538439,
+      "learning_rate": 4.22583401871896e-06,
+      "loss": 0.3339,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5631537925093326,
+      "grad_norm": 0.7050705132277877,
+      "learning_rate": 4.224521363353403e-06,
+      "loss": 0.3293,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5632358370595233,
+      "grad_norm": 0.6537196622100582,
+      "learning_rate": 4.223208762750899e-06,
+      "loss": 0.3051,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5633178816097141,
+      "grad_norm": 0.7178002306083943,
+      "learning_rate": 4.221896217004143e-06,
+      "loss": 0.3001,
+      "step": 6866
+    },
+    {
+      "epoch": 0.5633999261599049,
+      "grad_norm": 0.6697008677137798,
+      "learning_rate": 4.220583726205829e-06,
+      "loss": 0.3096,
+      "step": 6867
+    },
+    {
+      "epoch": 0.5634819707100955,
+      "grad_norm": 0.708251380170302,
+      "learning_rate": 4.219271290448639e-06,
+      "loss": 0.3101,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5635640152602863,
+      "grad_norm": 0.7153291831326698,
+      "learning_rate": 4.2179589098252574e-06,
+      "loss": 0.2861,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5636460598104771,
+      "grad_norm": 0.6764377801170567,
+      "learning_rate": 4.21664658442836e-06,
+      "loss": 0.3323,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5637281043606679,
+      "grad_norm": 0.6311258819375247,
+      "learning_rate": 4.2153343143506215e-06,
+      "loss": 0.3132,
+      "step": 6871
+    },
+    {
+      "epoch": 0.5638101489108586,
+      "grad_norm": 0.7133626537922685,
+      "learning_rate": 4.214022099684715e-06,
+      "loss": 0.298,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5638921934610494,
+      "grad_norm": 0.7559332432795586,
+      "learning_rate": 4.2127099405233024e-06,
+      "loss": 0.3317,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5639742380112401,
+      "grad_norm": 0.7314242472005377,
+      "learning_rate": 4.211397836959052e-06,
+      "loss": 0.3096,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5640562825614308,
+      "grad_norm": 0.888065920744925,
+      "learning_rate": 4.210085789084617e-06,
+      "loss": 0.3152,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5641383271116216,
+      "grad_norm": 0.7019143323088833,
+      "learning_rate": 4.2087737969926545e-06,
+      "loss": 0.3327,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5642203716618124,
+      "grad_norm": 0.7038400384767666,
+      "learning_rate": 4.207461860775816e-06,
+      "loss": 0.3145,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5643024162120032,
+      "grad_norm": 0.6438697621708375,
+      "learning_rate": 4.206149980526748e-06,
+      "loss": 0.3502,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5643844607621938,
+      "grad_norm": 0.8791847500223677,
+      "learning_rate": 4.204838156338094e-06,
+      "loss": 0.352,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5644665053123846,
+      "grad_norm": 0.7755602228064423,
+      "learning_rate": 4.203526388302489e-06,
+      "loss": 0.2976,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5645485498625754,
+      "grad_norm": 0.662208172669629,
+      "learning_rate": 4.202214676512571e-06,
+      "loss": 0.3323,
+      "step": 6881
+    },
+    {
+      "epoch": 0.5646305944127661,
+      "grad_norm": 0.8372389688983864,
+      "learning_rate": 4.200903021060973e-06,
+      "loss": 0.2878,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5647126389629569,
+      "grad_norm": 0.7091300357324504,
+      "learning_rate": 4.199591422040318e-06,
+      "loss": 0.2904,
+      "step": 6883
+    },
+    {
+      "epoch": 0.5647946835131477,
+      "grad_norm": 0.7113278452611494,
+      "learning_rate": 4.198279879543232e-06,
+      "loss": 0.3367,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5648767280633384,
+      "grad_norm": 0.885381811246509,
+      "learning_rate": 4.196968393662334e-06,
+      "loss": 0.3408,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5649587726135291,
+      "grad_norm": 0.6701482847558223,
+      "learning_rate": 4.195656964490235e-06,
+      "loss": 0.3268,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5650408171637199,
+      "grad_norm": 0.6736994841787683,
+      "learning_rate": 4.194345592119551e-06,
+      "loss": 0.3137,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5651228617139107,
+      "grad_norm": 0.6911840987200915,
+      "learning_rate": 4.193034276642886e-06,
+      "loss": 0.3034,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5652049062641014,
+      "grad_norm": 0.7531252199774918,
+      "learning_rate": 4.191723018152842e-06,
+      "loss": 0.3113,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5652869508142921,
+      "grad_norm": 0.6781339419909869,
+      "learning_rate": 4.190411816742024e-06,
+      "loss": 0.2737,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5653689953644829,
+      "grad_norm": 0.6484822494139602,
+      "learning_rate": 4.18910067250302e-06,
+      "loss": 0.3245,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5654510399146737,
+      "grad_norm": 0.7038148379510518,
+      "learning_rate": 4.1877895855284255e-06,
+      "loss": 0.322,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5655330844648644,
+      "grad_norm": 0.7705895973782804,
+      "learning_rate": 4.186478555910824e-06,
+      "loss": 0.3679,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5656151290150552,
+      "grad_norm": 0.7240639894024602,
+      "learning_rate": 4.185167583742799e-06,
+      "loss": 0.3173,
+      "step": 6894
+    },
+    {
+      "epoch": 0.565697173565246,
+      "grad_norm": 0.7358884946933573,
+      "learning_rate": 4.183856669116933e-06,
+      "loss": 0.3579,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5657792181154366,
+      "grad_norm": 0.7207568017553144,
+      "learning_rate": 4.182545812125795e-06,
+      "loss": 0.318,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5658612626656274,
+      "grad_norm": 0.6984776736592815,
+      "learning_rate": 4.1812350128619605e-06,
+      "loss": 0.3582,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5659433072158182,
+      "grad_norm": 0.731863391933108,
+      "learning_rate": 4.179924271417993e-06,
+      "loss": 0.3442,
+      "step": 6898
+    },
+    {
+      "epoch": 0.566025351766009,
+      "grad_norm": 0.7750780146730002,
+      "learning_rate": 4.178613587886455e-06,
+      "loss": 0.3533,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5661073963161997,
+      "grad_norm": 0.5947962493815266,
+      "learning_rate": 4.177302962359908e-06,
+      "loss": 0.2948,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5661894408663904,
+      "grad_norm": 0.6723880569938533,
+      "learning_rate": 4.175992394930902e-06,
+      "loss": 0.2942,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5662714854165812,
+      "grad_norm": 0.7090711666992854,
+      "learning_rate": 4.174681885691989e-06,
+      "loss": 0.3304,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5663535299667719,
+      "grad_norm": 0.7069521877799458,
+      "learning_rate": 4.173371434735718e-06,
+      "loss": 0.3085,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5664355745169627,
+      "grad_norm": 0.576566070297212,
+      "learning_rate": 4.172061042154626e-06,
+      "loss": 0.3164,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5665176190671535,
+      "grad_norm": 0.711439052194926,
+      "learning_rate": 4.170750708041253e-06,
+      "loss": 0.3311,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5665996636173443,
+      "grad_norm": 0.6593520400133956,
+      "learning_rate": 4.169440432488134e-06,
+      "loss": 0.3126,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5666817081675349,
+      "grad_norm": 0.736135436896953,
+      "learning_rate": 4.168130215587795e-06,
+      "loss": 0.3554,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5667637527177257,
+      "grad_norm": 0.7145187738397223,
+      "learning_rate": 4.166820057432767e-06,
+      "loss": 0.3354,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5668457972679165,
+      "grad_norm": 0.6576719242064437,
+      "learning_rate": 4.165509958115564e-06,
+      "loss": 0.316,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5669278418181072,
+      "grad_norm": 0.7495493828894344,
+      "learning_rate": 4.16419991772871e-06,
+      "loss": 0.3234,
+      "step": 6910
+    },
+    {
+      "epoch": 0.567009886368298,
+      "grad_norm": 0.7936187555852469,
+      "learning_rate": 4.162889936364713e-06,
+      "loss": 0.3458,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5670919309184888,
+      "grad_norm": 0.6911874097994237,
+      "learning_rate": 4.161580014116086e-06,
+      "loss": 0.3164,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5671739754686795,
+      "grad_norm": 0.6841731024244019,
+      "learning_rate": 4.1602701510753305e-06,
+      "loss": 0.3163,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5672560200188702,
+      "grad_norm": 0.7566167260289158,
+      "learning_rate": 4.158960347334947e-06,
+      "loss": 0.3153,
+      "step": 6914
+    },
+    {
+      "epoch": 0.567338064569061,
+      "grad_norm": 0.6763104197820017,
+      "learning_rate": 4.157650602987434e-06,
+      "loss": 0.3125,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5674201091192518,
+      "grad_norm": 0.766683665438428,
+      "learning_rate": 4.15634091812528e-06,
+      "loss": 0.3126,
+      "step": 6916
+    },
+    {
+      "epoch": 0.5675021536694425,
+      "grad_norm": 0.6628930395981967,
+      "learning_rate": 4.1550312928409754e-06,
+      "loss": 0.3062,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5675841982196332,
+      "grad_norm": 0.6865321714799644,
+      "learning_rate": 4.153721727227005e-06,
+      "loss": 0.3246,
+      "step": 6918
+    },
+    {
+      "epoch": 0.567666242769824,
+      "grad_norm": 0.7734844515937673,
+      "learning_rate": 4.152412221375846e-06,
+      "loss": 0.2982,
+      "step": 6919
+    },
+    {
+      "epoch": 0.5677482873200148,
+      "grad_norm": 0.6653850646122772,
+      "learning_rate": 4.151102775379973e-06,
+      "loss": 0.3015,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5678303318702055,
+      "grad_norm": 0.6200181714559303,
+      "learning_rate": 4.14979338933186e-06,
+      "loss": 0.2821,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5679123764203963,
+      "grad_norm": 0.6666429496710033,
+      "learning_rate": 4.148484063323969e-06,
+      "loss": 0.3039,
+      "step": 6922
+    },
+    {
+      "epoch": 0.567994420970587,
+      "grad_norm": 0.7225994028029408,
+      "learning_rate": 4.147174797448769e-06,
+      "loss": 0.3632,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5680764655207777,
+      "grad_norm": 0.5811358968776329,
+      "learning_rate": 4.145865591798712e-06,
+      "loss": 0.2901,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5681585100709685,
+      "grad_norm": 0.7187757374627715,
+      "learning_rate": 4.144556446466255e-06,
+      "loss": 0.2988,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5682405546211593,
+      "grad_norm": 0.6075443490821545,
+      "learning_rate": 4.14324736154385e-06,
+      "loss": 0.2925,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5683225991713501,
+      "grad_norm": 0.6104094431953117,
+      "learning_rate": 4.141938337123936e-06,
+      "loss": 0.2939,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5684046437215408,
+      "grad_norm": 0.7326633965073347,
+      "learning_rate": 4.1406293732989615e-06,
+      "loss": 0.3577,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5684866882717315,
+      "grad_norm": 0.6428367028747346,
+      "learning_rate": 4.139320470161357e-06,
+      "loss": 0.3879,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5685687328219223,
+      "grad_norm": 0.9053085205224067,
+      "learning_rate": 4.138011627803557e-06,
+      "loss": 0.3467,
+      "step": 6930
+    },
+    {
+      "epoch": 0.568650777372113,
+      "grad_norm": 0.7136946742105611,
+      "learning_rate": 4.136702846317996e-06,
+      "loss": 0.3158,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5687328219223038,
+      "grad_norm": 0.6836885316216437,
+      "learning_rate": 4.135394125797089e-06,
+      "loss": 0.3085,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5688148664724946,
+      "grad_norm": 0.6801341001971227,
+      "learning_rate": 4.134085466333262e-06,
+      "loss": 0.3274,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5688969110226854,
+      "grad_norm": 0.7308131235519495,
+      "learning_rate": 4.132776868018927e-06,
+      "loss": 0.3294,
+      "step": 6934
+    },
+    {
+      "epoch": 0.568978955572876,
+      "grad_norm": 0.6641076769699115,
+      "learning_rate": 4.131468330946495e-06,
+      "loss": 0.3052,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5690610001230668,
+      "grad_norm": 0.6992090043868859,
+      "learning_rate": 4.130159855208377e-06,
+      "loss": 0.2946,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5691430446732576,
+      "grad_norm": 0.6337122386852062,
+      "learning_rate": 4.12885144089697e-06,
+      "loss": 0.2886,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5692250892234483,
+      "grad_norm": 0.7146726445488698,
+      "learning_rate": 4.127543088104676e-06,
+      "loss": 0.3234,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5693071337736391,
+      "grad_norm": 0.6709743314915877,
+      "learning_rate": 4.126234796923887e-06,
+      "loss": 0.2682,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5693891783238298,
+      "grad_norm": 0.6554346451388557,
+      "learning_rate": 4.124926567446993e-06,
+      "loss": 0.3243,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5694712228740206,
+      "grad_norm": 0.6816801438719239,
+      "learning_rate": 4.1236183997663795e-06,
+      "loss": 0.3005,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5695532674242113,
+      "grad_norm": 0.6616986328437117,
+      "learning_rate": 4.122310293974425e-06,
+      "loss": 0.3179,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5696353119744021,
+      "grad_norm": 0.7536716282860246,
+      "learning_rate": 4.121002250163507e-06,
+      "loss": 0.3499,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5697173565245929,
+      "grad_norm": 0.6910769457336239,
+      "learning_rate": 4.119694268426001e-06,
+      "loss": 0.3245,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5697994010747837,
+      "grad_norm": 0.7199650672932433,
+      "learning_rate": 4.1183863488542686e-06,
+      "loss": 0.3003,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5698814456249743,
+      "grad_norm": 0.6855904293951374,
+      "learning_rate": 4.117078491540678e-06,
+      "loss": 0.2947,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5699634901751651,
+      "grad_norm": 0.6601123027333226,
+      "learning_rate": 4.115770696577584e-06,
+      "loss": 0.3219,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5700455347253559,
+      "grad_norm": 0.6852906784233334,
+      "learning_rate": 4.114462964057343e-06,
+      "loss": 0.3212,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5701275792755466,
+      "grad_norm": 0.7233307258523697,
+      "learning_rate": 4.113155294072307e-06,
+      "loss": 0.3261,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5702096238257374,
+      "grad_norm": 0.5619957158477011,
+      "learning_rate": 4.111847686714816e-06,
+      "loss": 0.3033,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5702916683759282,
+      "grad_norm": 0.746567132301524,
+      "learning_rate": 4.110540142077217e-06,
+      "loss": 0.3665,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5703737129261189,
+      "grad_norm": 0.716510758884508,
+      "learning_rate": 4.109232660251842e-06,
+      "loss": 0.3481,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5704557574763096,
+      "grad_norm": 0.6596294566299962,
+      "learning_rate": 4.107925241331027e-06,
+      "loss": 0.342,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5705378020265004,
+      "grad_norm": 0.7476415645316282,
+      "learning_rate": 4.106617885407099e-06,
+      "loss": 0.3229,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5706198465766912,
+      "grad_norm": 0.6956286973617859,
+      "learning_rate": 4.105310592572378e-06,
+      "loss": 0.329,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5707018911268819,
+      "grad_norm": 0.7440644060725291,
+      "learning_rate": 4.104003362919186e-06,
+      "loss": 0.3247,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5707839356770726,
+      "grad_norm": 0.7435992787321252,
+      "learning_rate": 4.102696196539839e-06,
+      "loss": 0.3096,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5708659802272634,
+      "grad_norm": 0.6947084227824596,
+      "learning_rate": 4.101389093526643e-06,
+      "loss": 0.3009,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5709480247774542,
+      "grad_norm": 0.6657235810689347,
+      "learning_rate": 4.1000820539719076e-06,
+      "loss": 0.295,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5710300693276449,
+      "grad_norm": 0.658010379340932,
+      "learning_rate": 4.0987750779679305e-06,
+      "loss": 0.298,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5711121138778357,
+      "grad_norm": 0.6071285705719381,
+      "learning_rate": 4.097468165607008e-06,
+      "loss": 0.2924,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5711941584280265,
+      "grad_norm": 0.6667954722478523,
+      "learning_rate": 4.0961613169814365e-06,
+      "loss": 0.3278,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5712762029782171,
+      "grad_norm": 0.6744910234241758,
+      "learning_rate": 4.094854532183498e-06,
+      "loss": 0.3577,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5713582475284079,
+      "grad_norm": 0.7833084025449972,
+      "learning_rate": 4.0935478113054805e-06,
+      "loss": 0.2983,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5714402920785987,
+      "grad_norm": 0.6330834318486096,
+      "learning_rate": 4.092241154439658e-06,
+      "loss": 0.3252,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5715223366287895,
+      "grad_norm": 0.6654606160316849,
+      "learning_rate": 4.090934561678308e-06,
+      "loss": 0.3394,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5716043811789802,
+      "grad_norm": 0.7222240377196809,
+      "learning_rate": 4.089628033113699e-06,
+      "loss": 0.2824,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5716864257291709,
+      "grad_norm": 0.7605795957027262,
+      "learning_rate": 4.088321568838095e-06,
+      "loss": 0.3171,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5717684702793617,
+      "grad_norm": 0.7628074233366285,
+      "learning_rate": 4.087015168943759e-06,
+      "loss": 0.3372,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5718505148295524,
+      "grad_norm": 0.645841160089767,
+      "learning_rate": 4.085708833522942e-06,
+      "loss": 0.3267,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5719325593797432,
+      "grad_norm": 0.6849193617573982,
+      "learning_rate": 4.084402562667898e-06,
+      "loss": 0.3543,
+      "step": 6971
+    },
+    {
+      "epoch": 0.572014603929934,
+      "grad_norm": 0.6168335123791162,
+      "learning_rate": 4.083096356470877e-06,
+      "loss": 0.3176,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5720966484801248,
+      "grad_norm": 0.6995823518767246,
+      "learning_rate": 4.081790215024116e-06,
+      "loss": 0.296,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5721786930303154,
+      "grad_norm": 0.8358496185520664,
+      "learning_rate": 4.080484138419856e-06,
+      "loss": 0.3123,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5722607375805062,
+      "grad_norm": 0.9605081009523275,
+      "learning_rate": 4.079178126750329e-06,
+      "loss": 0.31,
+      "step": 6975
+    },
+    {
+      "epoch": 0.572342782130697,
+      "grad_norm": 0.6915835142002024,
+      "learning_rate": 4.077872180107762e-06,
+      "loss": 0.3147,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5724248266808877,
+      "grad_norm": 0.6954388810117653,
+      "learning_rate": 4.076566298584382e-06,
+      "loss": 0.3087,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5725068712310785,
+      "grad_norm": 0.6288023146294207,
+      "learning_rate": 4.075260482272404e-06,
+      "loss": 0.3109,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5725889157812692,
+      "grad_norm": 0.8400815968245532,
+      "learning_rate": 4.073954731264045e-06,
+      "loss": 0.3522,
+      "step": 6979
+    },
+    {
+      "epoch": 0.57267096033146,
+      "grad_norm": 0.7037062298508322,
+      "learning_rate": 4.072649045651518e-06,
+      "loss": 0.341,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5727530048816507,
+      "grad_norm": 0.6589907890843151,
+      "learning_rate": 4.071343425527024e-06,
+      "loss": 0.3487,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5728350494318415,
+      "grad_norm": 0.7206084481598822,
+      "learning_rate": 4.070037870982766e-06,
+      "loss": 0.3173,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5729170939820323,
+      "grad_norm": 0.6795358871422833,
+      "learning_rate": 4.068732382110937e-06,
+      "loss": 0.2939,
+      "step": 6983
+    },
+    {
+      "epoch": 0.572999138532223,
+      "grad_norm": 0.6345399339562025,
+      "learning_rate": 4.0674269590037315e-06,
+      "loss": 0.2776,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5730811830824137,
+      "grad_norm": 1.0180950559369122,
+      "learning_rate": 4.066121601753338e-06,
+      "loss": 0.3241,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5731632276326045,
+      "grad_norm": 0.6647794320071855,
+      "learning_rate": 4.064816310451934e-06,
+      "loss": 0.293,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5732452721827953,
+      "grad_norm": 0.6806961351540692,
+      "learning_rate": 4.0635110851917035e-06,
+      "loss": 0.2854,
+      "step": 6987
+    },
+    {
+      "epoch": 0.573327316732986,
+      "grad_norm": 0.7096645670227227,
+      "learning_rate": 4.062205926064813e-06,
+      "loss": 0.3296,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5734093612831768,
+      "grad_norm": 0.6824147303065012,
+      "learning_rate": 4.060900833163433e-06,
+      "loss": 0.2866,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5734914058333676,
+      "grad_norm": 0.6827812499750032,
+      "learning_rate": 4.059595806579728e-06,
+      "loss": 0.3199,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5735734503835582,
+      "grad_norm": 0.6479100534871851,
+      "learning_rate": 4.058290846405856e-06,
+      "loss": 0.3038,
+      "step": 6991
+    },
+    {
+      "epoch": 0.573655494933749,
+      "grad_norm": 0.7497434832635859,
+      "learning_rate": 4.05698595273397e-06,
+      "loss": 0.3439,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5737375394839398,
+      "grad_norm": 0.6370568116620402,
+      "learning_rate": 4.0556811256562235e-06,
+      "loss": 0.3234,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5738195840341306,
+      "grad_norm": 0.7355222179703254,
+      "learning_rate": 4.054376365264758e-06,
+      "loss": 0.32,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5739016285843213,
+      "grad_norm": 0.7589568853042064,
+      "learning_rate": 4.053071671651715e-06,
+      "loss": 0.3236,
+      "step": 6995
+    },
+    {
+      "epoch": 0.573983673134512,
+      "grad_norm": 0.860671963086332,
+      "learning_rate": 4.051767044909227e-06,
+      "loss": 0.3645,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5740657176847028,
+      "grad_norm": 0.746475904544963,
+      "learning_rate": 4.050462485129428e-06,
+      "loss": 0.316,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5741477622348935,
+      "grad_norm": 0.6466842210021542,
+      "learning_rate": 4.049157992404444e-06,
+      "loss": 0.3011,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5742298067850843,
+      "grad_norm": 0.6256971774156157,
+      "learning_rate": 4.047853566826393e-06,
+      "loss": 0.2971,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5743118513352751,
+      "grad_norm": 0.6873831217710347,
+      "learning_rate": 4.046549208487395e-06,
+      "loss": 0.359,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5743938958854659,
+      "grad_norm": 0.6357290531075114,
+      "learning_rate": 4.045244917479559e-06,
+      "loss": 0.3272,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5744759404356565,
+      "grad_norm": 0.7207354426785008,
+      "learning_rate": 4.043940693894993e-06,
+      "loss": 0.3027,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5745579849858473,
+      "grad_norm": 0.6742783413324004,
+      "learning_rate": 4.042636537825801e-06,
+      "loss": 0.317,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5746400295360381,
+      "grad_norm": 0.6687598670234369,
+      "learning_rate": 4.041332449364076e-06,
+      "loss": 0.2893,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5747220740862288,
+      "grad_norm": 0.7331421099940713,
+      "learning_rate": 4.040028428601916e-06,
+      "loss": 0.3662,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5748041186364196,
+      "grad_norm": 0.9300866609790351,
+      "learning_rate": 4.038724475631403e-06,
+      "loss": 0.3248,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5748861631866103,
+      "grad_norm": 0.6436819288037349,
+      "learning_rate": 4.037420590544623e-06,
+      "loss": 0.321,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5749682077368011,
+      "grad_norm": 0.6619033706271414,
+      "learning_rate": 4.036116773433657e-06,
+      "loss": 0.3304,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5750502522869918,
+      "grad_norm": 0.7017569035422633,
+      "learning_rate": 4.034813024390575e-06,
+      "loss": 0.332,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5751322968371826,
+      "grad_norm": 0.749426171129015,
+      "learning_rate": 4.033509343507445e-06,
+      "loss": 0.3104,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5752143413873734,
+      "grad_norm": 0.8400309910631586,
+      "learning_rate": 4.032205730876335e-06,
+      "loss": 0.3182,
+      "step": 7011
+    },
+    {
+      "epoch": 0.575296385937564,
+      "grad_norm": 0.88225610856447,
+      "learning_rate": 4.0309021865892985e-06,
+      "loss": 0.3758,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5753784304877548,
+      "grad_norm": 0.6354609103955794,
+      "learning_rate": 4.029598710738395e-06,
+      "loss": 0.2901,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5754604750379456,
+      "grad_norm": 0.6277512216747303,
+      "learning_rate": 4.02829530341567e-06,
+      "loss": 0.2934,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5755425195881364,
+      "grad_norm": 0.7473433064487568,
+      "learning_rate": 4.02699196471317e-06,
+      "loss": 0.3461,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5756245641383271,
+      "grad_norm": 0.6917624863315063,
+      "learning_rate": 4.025688694722936e-06,
+      "loss": 0.3272,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5757066086885179,
+      "grad_norm": 0.7544928460092258,
+      "learning_rate": 4.0243854935369975e-06,
+      "loss": 0.2854,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5757886532387086,
+      "grad_norm": 0.6628049061848498,
+      "learning_rate": 4.0230823612473916e-06,
+      "loss": 0.3671,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5758706977888993,
+      "grad_norm": 0.7068680598034355,
+      "learning_rate": 4.021779297946137e-06,
+      "loss": 0.353,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5759527423390901,
+      "grad_norm": 0.6805002595505221,
+      "learning_rate": 4.020476303725257e-06,
+      "loss": 0.3116,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5760347868892809,
+      "grad_norm": 0.7295191922547828,
+      "learning_rate": 4.0191733786767685e-06,
+      "loss": 0.3437,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5761168314394717,
+      "grad_norm": 0.6288543326865177,
+      "learning_rate": 4.017870522892679e-06,
+      "loss": 0.3048,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5761988759896624,
+      "grad_norm": 0.6479572051075171,
+      "learning_rate": 4.016567736464995e-06,
+      "loss": 0.3329,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5762809205398531,
+      "grad_norm": 0.6885586219155,
+      "learning_rate": 4.015265019485718e-06,
+      "loss": 0.3069,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5763629650900439,
+      "grad_norm": 0.714750423369962,
+      "learning_rate": 4.013962372046842e-06,
+      "loss": 0.3157,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5764450096402346,
+      "grad_norm": 0.7019730291284866,
+      "learning_rate": 4.012659794240362e-06,
+      "loss": 0.2985,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5765270541904254,
+      "grad_norm": 0.6602756867442732,
+      "learning_rate": 4.011357286158258e-06,
+      "loss": 0.3145,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5766090987406162,
+      "grad_norm": 0.6827533306835487,
+      "learning_rate": 4.010054847892515e-06,
+      "loss": 0.3042,
+      "step": 7028
+    },
+    {
+      "epoch": 0.576691143290807,
+      "grad_norm": 0.7716707439859777,
+      "learning_rate": 4.008752479535109e-06,
+      "loss": 0.3248,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5767731878409976,
+      "grad_norm": 0.6530558880017374,
+      "learning_rate": 4.00745018117801e-06,
+      "loss": 0.3599,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5768552323911884,
+      "grad_norm": 0.6113233852803779,
+      "learning_rate": 4.006147952913185e-06,
+      "loss": 0.3188,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5769372769413792,
+      "grad_norm": 0.683861227760472,
+      "learning_rate": 4.004845794832594e-06,
+      "loss": 0.3092,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5770193214915699,
+      "grad_norm": 0.7185269744110251,
+      "learning_rate": 4.003543707028193e-06,
+      "loss": 0.3116,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5771013660417607,
+      "grad_norm": 0.7496553611210701,
+      "learning_rate": 4.002241689591937e-06,
+      "loss": 0.3306,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5771834105919514,
+      "grad_norm": 0.635405816814588,
+      "learning_rate": 4.000939742615767e-06,
+      "loss": 0.3232,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5772654551421422,
+      "grad_norm": 0.6533167004693939,
+      "learning_rate": 3.999637866191631e-06,
+      "loss": 0.2893,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5773474996923329,
+      "grad_norm": 0.6371600003158373,
+      "learning_rate": 3.998336060411459e-06,
+      "loss": 0.3186,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5774295442425237,
+      "grad_norm": 0.6784854497463945,
+      "learning_rate": 3.997034325367185e-06,
+      "loss": 0.321,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5775115887927145,
+      "grad_norm": 0.6205135586072251,
+      "learning_rate": 3.9957326611507375e-06,
+      "loss": 0.2935,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5775936333429051,
+      "grad_norm": 0.6615267459066195,
+      "learning_rate": 3.994431067854034e-06,
+      "loss": 0.3089,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5776756778930959,
+      "grad_norm": 0.6345931054594947,
+      "learning_rate": 3.9931295455689925e-06,
+      "loss": 0.3345,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5777577224432867,
+      "grad_norm": 0.7363490729105718,
+      "learning_rate": 3.991828094387527e-06,
+      "loss": 0.2974,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5778397669934775,
+      "grad_norm": 0.6242427207801468,
+      "learning_rate": 3.9905267144015405e-06,
+      "loss": 0.3293,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5779218115436682,
+      "grad_norm": 0.6696329665190742,
+      "learning_rate": 3.989225405702937e-06,
+      "loss": 0.3325,
+      "step": 7044
+    },
+    {
+      "epoch": 0.578003856093859,
+      "grad_norm": 0.7508968003214228,
+      "learning_rate": 3.987924168383609e-06,
+      "loss": 0.3091,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5780859006440497,
+      "grad_norm": 0.9584492564019172,
+      "learning_rate": 3.986623002535451e-06,
+      "loss": 0.328,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5781679451942404,
+      "grad_norm": 0.6467305190817915,
+      "learning_rate": 3.98532190825035e-06,
+      "loss": 0.3152,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5782499897444312,
+      "grad_norm": 0.6832835532038457,
+      "learning_rate": 3.984020885620184e-06,
+      "loss": 0.331,
+      "step": 7048
+    },
+    {
+      "epoch": 0.578332034294622,
+      "grad_norm": 0.6133288643854968,
+      "learning_rate": 3.982719934736832e-06,
+      "loss": 0.3017,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5784140788448128,
+      "grad_norm": 0.6444804286684577,
+      "learning_rate": 3.981419055692163e-06,
+      "loss": 0.3205,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5784961233950034,
+      "grad_norm": 0.7362409470660659,
+      "learning_rate": 3.980118248578044e-06,
+      "loss": 0.3191,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5785781679451942,
+      "grad_norm": 0.7469665419351849,
+      "learning_rate": 3.978817513486336e-06,
+      "loss": 0.3563,
+      "step": 7052
+    },
+    {
+      "epoch": 0.578660212495385,
+      "grad_norm": 0.6283500330473225,
+      "learning_rate": 3.9775168505088935e-06,
+      "loss": 0.3278,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5787422570455758,
+      "grad_norm": 0.6822687519853526,
+      "learning_rate": 3.97621625973757e-06,
+      "loss": 0.3186,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5788243015957665,
+      "grad_norm": 0.8770899120585389,
+      "learning_rate": 3.974915741264207e-06,
+      "loss": 0.3376,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5789063461459573,
+      "grad_norm": 0.8526754569490341,
+      "learning_rate": 3.97361529518065e-06,
+      "loss": 0.3627,
+      "step": 7056
+    },
+    {
+      "epoch": 0.578988390696148,
+      "grad_norm": 0.6296284779829309,
+      "learning_rate": 3.97231492157873e-06,
+      "loss": 0.3134,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5790704352463387,
+      "grad_norm": 0.782384296001675,
+      "learning_rate": 3.97101462055028e-06,
+      "loss": 0.3327,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5791524797965295,
+      "grad_norm": 0.6436598779039309,
+      "learning_rate": 3.969714392187123e-06,
+      "loss": 0.3049,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5792345243467203,
+      "grad_norm": 0.711515625217179,
+      "learning_rate": 3.968414236581083e-06,
+      "loss": 0.3307,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5793165688969111,
+      "grad_norm": 0.6833762085966762,
+      "learning_rate": 3.967114153823969e-06,
+      "loss": 0.3295,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5793986134471018,
+      "grad_norm": 0.6715673261747686,
+      "learning_rate": 3.965814144007597e-06,
+      "loss": 0.323,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5794806579972925,
+      "grad_norm": 0.7265214758726792,
+      "learning_rate": 3.964514207223766e-06,
+      "loss": 0.3178,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5795627025474833,
+      "grad_norm": 0.7637256399583046,
+      "learning_rate": 3.9632143435642794e-06,
+      "loss": 0.3052,
+      "step": 7064
+    },
+    {
+      "epoch": 0.579644747097674,
+      "grad_norm": 0.6637979998593568,
+      "learning_rate": 3.961914553120932e-06,
+      "loss": 0.31,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5797267916478648,
+      "grad_norm": 0.6439715002950562,
+      "learning_rate": 3.960614835985507e-06,
+      "loss": 0.3239,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5798088361980556,
+      "grad_norm": 0.6749593373751774,
+      "learning_rate": 3.959315192249796e-06,
+      "loss": 0.3155,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5798908807482464,
+      "grad_norm": 0.6231114511518994,
+      "learning_rate": 3.958015622005572e-06,
+      "loss": 0.3224,
+      "step": 7068
+    },
+    {
+      "epoch": 0.579972925298437,
+      "grad_norm": 0.6667287078216324,
+      "learning_rate": 3.95671612534461e-06,
+      "loss": 0.3417,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5800549698486278,
+      "grad_norm": 0.6213101741666504,
+      "learning_rate": 3.955416702358681e-06,
+      "loss": 0.3114,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5801370143988186,
+      "grad_norm": 0.6324006110994637,
+      "learning_rate": 3.954117353139546e-06,
+      "loss": 0.2947,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5802190589490093,
+      "grad_norm": 0.6232425363301958,
+      "learning_rate": 3.9528180777789644e-06,
+      "loss": 0.3101,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5803011034992,
+      "grad_norm": 0.6341785803235916,
+      "learning_rate": 3.951518876368685e-06,
+      "loss": 0.3208,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5803831480493908,
+      "grad_norm": 0.6669559432445702,
+      "learning_rate": 3.950219749000458e-06,
+      "loss": 0.3114,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5804651925995816,
+      "grad_norm": 0.7841255045747927,
+      "learning_rate": 3.9489206957660284e-06,
+      "loss": 0.3159,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5805472371497723,
+      "grad_norm": 0.7047042714882914,
+      "learning_rate": 3.947621716757128e-06,
+      "loss": 0.2985,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5806292816999631,
+      "grad_norm": 0.6534534466867709,
+      "learning_rate": 3.9463228120654926e-06,
+      "loss": 0.3296,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5807113262501539,
+      "grad_norm": 0.6608091541839569,
+      "learning_rate": 3.945023981782848e-06,
+      "loss": 0.3322,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5807933708003445,
+      "grad_norm": 0.689433687168061,
+      "learning_rate": 3.943725226000913e-06,
+      "loss": 0.3473,
+      "step": 7079
+    },
+    {
+      "epoch": 0.5808754153505353,
+      "grad_norm": 0.7239795294334047,
+      "learning_rate": 3.942426544811407e-06,
+      "loss": 0.2979,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5809574599007261,
+      "grad_norm": 0.7406484587023205,
+      "learning_rate": 3.941127938306038e-06,
+      "loss": 0.3562,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5810395044509169,
+      "grad_norm": 0.6190174065976299,
+      "learning_rate": 3.939829406576512e-06,
+      "loss": 0.3113,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5811215490011076,
+      "grad_norm": 0.6450454865140072,
+      "learning_rate": 3.938530949714533e-06,
+      "loss": 0.3177,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5812035935512984,
+      "grad_norm": 0.6309644916438866,
+      "learning_rate": 3.9372325678117916e-06,
+      "loss": 0.2971,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5812856381014891,
+      "grad_norm": 0.7088865937371482,
+      "learning_rate": 3.935934260959978e-06,
+      "loss": 0.3042,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5813676826516798,
+      "grad_norm": 0.6539693795584257,
+      "learning_rate": 3.934636029250778e-06,
+      "loss": 0.3041,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5814497272018706,
+      "grad_norm": 0.7274804773765053,
+      "learning_rate": 3.933337872775869e-06,
+      "loss": 0.3267,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5815317717520614,
+      "grad_norm": 0.6465491574664447,
+      "learning_rate": 3.932039791626928e-06,
+      "loss": 0.3116,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5816138163022522,
+      "grad_norm": 0.6351010343525784,
+      "learning_rate": 3.9307417858956185e-06,
+      "loss": 0.3119,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5816958608524428,
+      "grad_norm": 0.6892604857598336,
+      "learning_rate": 3.9294438556736084e-06,
+      "loss": 0.3027,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5817779054026336,
+      "grad_norm": 0.6797616343118371,
+      "learning_rate": 3.928146001052552e-06,
+      "loss": 0.3267,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5818599499528244,
+      "grad_norm": 0.6173215327651256,
+      "learning_rate": 3.926848222124103e-06,
+      "loss": 0.2928,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5819419945030151,
+      "grad_norm": 0.625427632986777,
+      "learning_rate": 3.92555051897991e-06,
+      "loss": 0.3225,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5820240390532059,
+      "grad_norm": 0.6490537748896813,
+      "learning_rate": 3.92425289171161e-06,
+      "loss": 0.3225,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5821060836033967,
+      "grad_norm": 0.6650790892971045,
+      "learning_rate": 3.922955340410844e-06,
+      "loss": 0.3089,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5821881281535874,
+      "grad_norm": 0.7575538917799592,
+      "learning_rate": 3.921657865169242e-06,
+      "loss": 0.3746,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5822701727037781,
+      "grad_norm": 0.7238834980676823,
+      "learning_rate": 3.920360466078428e-06,
+      "loss": 0.3307,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5823522172539689,
+      "grad_norm": 0.6192677796303321,
+      "learning_rate": 3.9190631432300255e-06,
+      "loss": 0.3156,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5824342618041597,
+      "grad_norm": 0.8259690397297306,
+      "learning_rate": 3.917765896715645e-06,
+      "loss": 0.323,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5825163063543504,
+      "grad_norm": 0.6816451998344789,
+      "learning_rate": 3.916468726626898e-06,
+      "loss": 0.3143,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5825983509045412,
+      "grad_norm": 0.7236855927731777,
+      "learning_rate": 3.915171633055391e-06,
+      "loss": 0.3161,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5826803954547319,
+      "grad_norm": 0.694377700001339,
+      "learning_rate": 3.913874616092718e-06,
+      "loss": 0.2982,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5827624400049227,
+      "grad_norm": 0.6870768707494438,
+      "learning_rate": 3.912577675830476e-06,
+      "loss": 0.3415,
+      "step": 7103
+    },
+    {
+      "epoch": 0.5828444845551134,
+      "grad_norm": 0.6981205608860701,
+      "learning_rate": 3.91128081236025e-06,
+      "loss": 0.3232,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5829265291053042,
+      "grad_norm": 1.1734905690978235,
+      "learning_rate": 3.909984025773625e-06,
+      "loss": 0.3612,
+      "step": 7105
+    },
+    {
+      "epoch": 0.583008573655495,
+      "grad_norm": 0.7140004986468282,
+      "learning_rate": 3.908687316162178e-06,
+      "loss": 0.3415,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5830906182056856,
+      "grad_norm": 0.7059338097929733,
+      "learning_rate": 3.907390683617477e-06,
+      "loss": 0.2806,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5831726627558764,
+      "grad_norm": 0.6112723902393721,
+      "learning_rate": 3.906094128231093e-06,
+      "loss": 0.3235,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5832547073060672,
+      "grad_norm": 0.6883349668439013,
+      "learning_rate": 3.904797650094581e-06,
+      "loss": 0.3138,
+      "step": 7109
+    },
+    {
+      "epoch": 0.583336751856258,
+      "grad_norm": 0.6540419296387142,
+      "learning_rate": 3.9035012492995004e-06,
+      "loss": 0.3022,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5834187964064487,
+      "grad_norm": 0.6660529412527436,
+      "learning_rate": 3.902204925937402e-06,
+      "loss": 0.3211,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5835008409566395,
+      "grad_norm": 0.790437207113533,
+      "learning_rate": 3.9009086800998266e-06,
+      "loss": 0.2987,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5835828855068302,
+      "grad_norm": 0.7444876432058228,
+      "learning_rate": 3.899612511878313e-06,
+      "loss": 0.3447,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5836649300570209,
+      "grad_norm": 0.69691914385101,
+      "learning_rate": 3.898316421364398e-06,
+      "loss": 0.3272,
+      "step": 7114
+    },
+    {
+      "epoch": 0.5837469746072117,
+      "grad_norm": 0.6349887809909757,
+      "learning_rate": 3.897020408649607e-06,
+      "loss": 0.2771,
+      "step": 7115
+    },
+    {
+      "epoch": 0.5838290191574025,
+      "grad_norm": 0.6498870848821703,
+      "learning_rate": 3.895724473825463e-06,
+      "loss": 0.32,
+      "step": 7116
+    },
+    {
+      "epoch": 0.5839110637075933,
+      "grad_norm": 0.6826724446987634,
+      "learning_rate": 3.8944286169834815e-06,
+      "loss": 0.3028,
+      "step": 7117
+    },
+    {
+      "epoch": 0.5839931082577839,
+      "grad_norm": 0.6565625120493439,
+      "learning_rate": 3.893132838215174e-06,
+      "loss": 0.3063,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5840751528079747,
+      "grad_norm": 0.7564100313443195,
+      "learning_rate": 3.891837137612049e-06,
+      "loss": 0.2833,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5841571973581655,
+      "grad_norm": 0.6695459789247694,
+      "learning_rate": 3.890541515265604e-06,
+      "loss": 0.3379,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5842392419083562,
+      "grad_norm": 0.6947277115320083,
+      "learning_rate": 3.889245971267336e-06,
+      "loss": 0.2943,
+      "step": 7121
+    },
+    {
+      "epoch": 0.584321286458547,
+      "grad_norm": 0.616039547900988,
+      "learning_rate": 3.887950505708731e-06,
+      "loss": 0.3136,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5844033310087378,
+      "grad_norm": 0.6726608817780806,
+      "learning_rate": 3.886655118681275e-06,
+      "loss": 0.308,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5844853755589285,
+      "grad_norm": 0.6417471240453699,
+      "learning_rate": 3.885359810276448e-06,
+      "loss": 0.2767,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5845674201091192,
+      "grad_norm": 0.760363229724296,
+      "learning_rate": 3.884064580585717e-06,
+      "loss": 0.3206,
+      "step": 7125
+    },
+    {
+      "epoch": 0.58464946465931,
+      "grad_norm": 0.8210428058699445,
+      "learning_rate": 3.882769429700556e-06,
+      "loss": 0.3161,
+      "step": 7126
+    },
+    {
+      "epoch": 0.5847315092095008,
+      "grad_norm": 0.7185227175375883,
+      "learning_rate": 3.88147435771242e-06,
+      "loss": 0.3051,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5848135537596915,
+      "grad_norm": 0.7842419390327371,
+      "learning_rate": 3.880179364712769e-06,
+      "loss": 0.3244,
+      "step": 7128
+    },
+    {
+      "epoch": 0.5848955983098822,
+      "grad_norm": 0.6804311368808922,
+      "learning_rate": 3.878884450793053e-06,
+      "loss": 0.318,
+      "step": 7129
+    },
+    {
+      "epoch": 0.584977642860073,
+      "grad_norm": 0.7789190918684386,
+      "learning_rate": 3.8775896160447136e-06,
+      "loss": 0.3307,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5850596874102638,
+      "grad_norm": 0.6863130311748864,
+      "learning_rate": 3.8762948605591926e-06,
+      "loss": 0.3063,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5851417319604545,
+      "grad_norm": 0.7247532335649363,
+      "learning_rate": 3.875000184427925e-06,
+      "loss": 0.3153,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5852237765106453,
+      "grad_norm": 0.7518642801682311,
+      "learning_rate": 3.873705587742336e-06,
+      "loss": 0.2994,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5853058210608361,
+      "grad_norm": 0.709024819946285,
+      "learning_rate": 3.87241107059385e-06,
+      "loss": 0.3177,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5853878656110267,
+      "grad_norm": 0.6694662924412195,
+      "learning_rate": 3.8711166330738805e-06,
+      "loss": 0.3515,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5854699101612175,
+      "grad_norm": 0.6176459839849961,
+      "learning_rate": 3.869822275273841e-06,
+      "loss": 0.3029,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5855519547114083,
+      "grad_norm": 0.6611098381144256,
+      "learning_rate": 3.8685279972851385e-06,
+      "loss": 0.3123,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5856339992615991,
+      "grad_norm": 0.6500870580510151,
+      "learning_rate": 3.867233799199169e-06,
+      "loss": 0.3179,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5857160438117898,
+      "grad_norm": 0.9137029007399061,
+      "learning_rate": 3.86593968110733e-06,
+      "loss": 0.3217,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5857980883619806,
+      "grad_norm": 0.5885311209301591,
+      "learning_rate": 3.8646456431010086e-06,
+      "loss": 0.3135,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5858801329121713,
+      "grad_norm": 0.6493091590779807,
+      "learning_rate": 3.863351685271586e-06,
+      "loss": 0.3795,
+      "step": 7141
+    },
+    {
+      "epoch": 0.585962177462362,
+      "grad_norm": 0.7428204571022734,
+      "learning_rate": 3.8620578077104444e-06,
+      "loss": 0.2989,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5860442220125528,
+      "grad_norm": 0.7102592299900693,
+      "learning_rate": 3.860764010508949e-06,
+      "loss": 0.3308,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5861262665627436,
+      "grad_norm": 0.6891525533758004,
+      "learning_rate": 3.859470293758471e-06,
+      "loss": 0.2897,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5862083111129344,
+      "grad_norm": 0.7048727913594738,
+      "learning_rate": 3.858176657550367e-06,
+      "loss": 0.3378,
+      "step": 7145
+    },
+    {
+      "epoch": 0.586290355663125,
+      "grad_norm": 0.6760663331657984,
+      "learning_rate": 3.856883101975994e-06,
+      "loss": 0.2758,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5863724002133158,
+      "grad_norm": 0.6783123061345858,
+      "learning_rate": 3.855589627126699e-06,
+      "loss": 0.3089,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5864544447635066,
+      "grad_norm": 0.6376470609416038,
+      "learning_rate": 3.8542962330938266e-06,
+      "loss": 0.2854,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5865364893136973,
+      "grad_norm": 0.7530538550280989,
+      "learning_rate": 3.853002919968712e-06,
+      "loss": 0.304,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5866185338638881,
+      "grad_norm": 0.6427119968910059,
+      "learning_rate": 3.851709687842691e-06,
+      "loss": 0.31,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5867005784140789,
+      "grad_norm": 0.7937034792899227,
+      "learning_rate": 3.850416536807085e-06,
+      "loss": 0.3257,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5867826229642696,
+      "grad_norm": 0.7397425394525285,
+      "learning_rate": 3.849123466953217e-06,
+      "loss": 0.3101,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5868646675144603,
+      "grad_norm": 0.8411283318933851,
+      "learning_rate": 3.8478304783724e-06,
+      "loss": 0.3587,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5869467120646511,
+      "grad_norm": 0.6581912306524748,
+      "learning_rate": 3.846537571155944e-06,
+      "loss": 0.344,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5870287566148419,
+      "grad_norm": 0.5586088349081809,
+      "learning_rate": 3.845244745395153e-06,
+      "loss": 0.2758,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5871108011650326,
+      "grad_norm": 0.6179859722235006,
+      "learning_rate": 3.84395200118132e-06,
+      "loss": 0.3222,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5871928457152233,
+      "grad_norm": 0.7441055697535667,
+      "learning_rate": 3.842659338605742e-06,
+      "loss": 0.3251,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5872748902654141,
+      "grad_norm": 0.8716257247250763,
+      "learning_rate": 3.8413667577597e-06,
+      "loss": 0.3388,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5873569348156049,
+      "grad_norm": 0.5978289656440595,
+      "learning_rate": 3.840074258734476e-06,
+      "loss": 0.3172,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5874389793657956,
+      "grad_norm": 0.709926111460126,
+      "learning_rate": 3.838781841621347e-06,
+      "loss": 0.3446,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5875210239159864,
+      "grad_norm": 0.721083949680846,
+      "learning_rate": 3.8374895065115765e-06,
+      "loss": 0.2931,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5876030684661772,
+      "grad_norm": 0.651152241032542,
+      "learning_rate": 3.836197253496431e-06,
+      "loss": 0.2835,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5876851130163678,
+      "grad_norm": 0.5880737518424691,
+      "learning_rate": 3.834905082667164e-06,
+      "loss": 0.2712,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5877671575665586,
+      "grad_norm": 0.6524215039231293,
+      "learning_rate": 3.8336129941150274e-06,
+      "loss": 0.3508,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5878492021167494,
+      "grad_norm": 0.6378763689157145,
+      "learning_rate": 3.832320987931269e-06,
+      "loss": 0.3079,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5879312466669402,
+      "grad_norm": 0.7584176422430177,
+      "learning_rate": 3.831029064207126e-06,
+      "loss": 0.3329,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5880132912171309,
+      "grad_norm": 0.6382100276205213,
+      "learning_rate": 3.829737223033832e-06,
+      "loss": 0.3289,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5880953357673216,
+      "grad_norm": 0.8783902679919763,
+      "learning_rate": 3.828445464502616e-06,
+      "loss": 0.3038,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5881773803175124,
+      "grad_norm": 0.6519960633993306,
+      "learning_rate": 3.8271537887046976e-06,
+      "loss": 0.3337,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5882594248677032,
+      "grad_norm": 0.6430223353774616,
+      "learning_rate": 3.825862195731297e-06,
+      "loss": 0.3087,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5883414694178939,
+      "grad_norm": 0.6713441669531096,
+      "learning_rate": 3.824570685673618e-06,
+      "loss": 0.3297,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5884235139680847,
+      "grad_norm": 0.7800091339648594,
+      "learning_rate": 3.82327925862287e-06,
+      "loss": 0.3528,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5885055585182755,
+      "grad_norm": 0.7228640424780296,
+      "learning_rate": 3.821987914670252e-06,
+      "loss": 0.3244,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5885876030684661,
+      "grad_norm": 0.6573493540278731,
+      "learning_rate": 3.820696653906954e-06,
+      "loss": 0.2868,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5886696476186569,
+      "grad_norm": 0.6841963190315057,
+      "learning_rate": 3.819405476424164e-06,
+      "loss": 0.3645,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5887516921688477,
+      "grad_norm": 0.7212799106043073,
+      "learning_rate": 3.8181143823130615e-06,
+      "loss": 0.3216,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5888337367190385,
+      "grad_norm": 0.675665847980701,
+      "learning_rate": 3.8168233716648215e-06,
+      "loss": 0.2825,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5889157812692292,
+      "grad_norm": 0.7248134786581449,
+      "learning_rate": 3.815532444570617e-06,
+      "loss": 0.3227,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58899782581942,
+      "grad_norm": 0.7021444340943291,
+      "learning_rate": 3.814241601121605e-06,
+      "loss": 0.3194,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5890798703696107,
+      "grad_norm": 0.7230799719223179,
+      "learning_rate": 3.812950841408949e-06,
+      "loss": 0.3261,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5891619149198014,
+      "grad_norm": 0.7352006059283654,
+      "learning_rate": 3.811660165523795e-06,
+      "loss": 0.2793,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5892439594699922,
+      "grad_norm": 0.6542192293466013,
+      "learning_rate": 3.8103695735572922e-06,
+      "loss": 0.3106,
+      "step": 7182
+    },
+    {
+      "epoch": 0.589326004020183,
+      "grad_norm": 0.7345820411028887,
+      "learning_rate": 3.8090790656005792e-06,
+      "loss": 0.3226,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5894080485703738,
+      "grad_norm": 0.6408745021308299,
+      "learning_rate": 3.807788641744788e-06,
+      "loss": 0.3176,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5894900931205644,
+      "grad_norm": 0.6917106406397342,
+      "learning_rate": 3.806498302081047e-06,
+      "loss": 0.3265,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5895721376707552,
+      "grad_norm": 0.6997254398933781,
+      "learning_rate": 3.8052080467004808e-06,
+      "loss": 0.2873,
+      "step": 7186
+    },
+    {
+      "epoch": 0.589654182220946,
+      "grad_norm": 0.7957212310715522,
+      "learning_rate": 3.803917875694201e-06,
+      "loss": 0.2908,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5897362267711367,
+      "grad_norm": 0.6818145975230553,
+      "learning_rate": 3.8026277891533203e-06,
+      "loss": 0.3145,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5898182713213275,
+      "grad_norm": 0.7735560625190547,
+      "learning_rate": 3.8013377871689406e-06,
+      "loss": 0.3348,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5899003158715183,
+      "grad_norm": 0.7839031778841977,
+      "learning_rate": 3.80004786983216e-06,
+      "loss": 0.2822,
+      "step": 7190
+    },
+    {
+      "epoch": 0.589982360421709,
+      "grad_norm": 0.7351420586940812,
+      "learning_rate": 3.7987580372340726e-06,
+      "loss": 0.2991,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5900644049718997,
+      "grad_norm": 0.6654759582515442,
+      "learning_rate": 3.797468289465761e-06,
+      "loss": 0.3236,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5901464495220905,
+      "grad_norm": 0.8712808455352973,
+      "learning_rate": 3.7961786266183086e-06,
+      "loss": 0.3095,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5902284940722813,
+      "grad_norm": 0.648544529532138,
+      "learning_rate": 3.7948890487827854e-06,
+      "loss": 0.3305,
+      "step": 7194
+    },
+    {
+      "epoch": 0.590310538622472,
+      "grad_norm": 0.6892681753941027,
+      "learning_rate": 3.7935995560502626e-06,
+      "loss": 0.2968,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5903925831726627,
+      "grad_norm": 0.5878559502861292,
+      "learning_rate": 3.792310148511802e-06,
+      "loss": 0.2912,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5904746277228535,
+      "grad_norm": 0.9695804095482218,
+      "learning_rate": 3.791020826258456e-06,
+      "loss": 0.3502,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5905566722730443,
+      "grad_norm": 0.7348195773644431,
+      "learning_rate": 3.7897315893812796e-06,
+      "loss": 0.3114,
+      "step": 7198
+    },
+    {
+      "epoch": 0.590638716823235,
+      "grad_norm": 0.6952812838149849,
+      "learning_rate": 3.7884424379713114e-06,
+      "loss": 0.3188,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5907207613734258,
+      "grad_norm": 0.6467866273126414,
+      "learning_rate": 3.787153372119592e-06,
+      "loss": 0.3111,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5908028059236166,
+      "grad_norm": 0.6691183368201904,
+      "learning_rate": 3.7858643919171543e-06,
+      "loss": 0.3166,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5908848504738072,
+      "grad_norm": 0.6963407064591152,
+      "learning_rate": 3.784575497455022e-06,
+      "loss": 0.3301,
+      "step": 7202
+    },
+    {
+      "epoch": 0.590966895023998,
+      "grad_norm": 0.7391155352090247,
+      "learning_rate": 3.783286688824214e-06,
+      "loss": 0.2939,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5910489395741888,
+      "grad_norm": 0.6799494405852019,
+      "learning_rate": 3.781997966115748e-06,
+      "loss": 0.3503,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5911309841243796,
+      "grad_norm": 0.6702364502460492,
+      "learning_rate": 3.780709329420626e-06,
+      "loss": 0.305,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5912130286745703,
+      "grad_norm": 0.7591670734751982,
+      "learning_rate": 3.7794207788298547e-06,
+      "loss": 0.313,
+      "step": 7206
+    },
+    {
+      "epoch": 0.591295073224761,
+      "grad_norm": 0.6639823950118168,
+      "learning_rate": 3.7781323144344246e-06,
+      "loss": 0.2943,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5913771177749518,
+      "grad_norm": 0.7266195361518968,
+      "learning_rate": 3.7768439363253285e-06,
+      "loss": 0.3411,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5914591623251425,
+      "grad_norm": 0.625461482222722,
+      "learning_rate": 3.7755556445935485e-06,
+      "loss": 0.3175,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5915412068753333,
+      "grad_norm": 0.7111427365721458,
+      "learning_rate": 3.7742674393300614e-06,
+      "loss": 0.295,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5916232514255241,
+      "grad_norm": 0.7113071724881922,
+      "learning_rate": 3.77297932062584e-06,
+      "loss": 0.3206,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5917052959757149,
+      "grad_norm": 0.9696151828553659,
+      "learning_rate": 3.7716912885718455e-06,
+      "loss": 0.3243,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5917873405259055,
+      "grad_norm": 0.7139980241878989,
+      "learning_rate": 3.7704033432590387e-06,
+      "loss": 0.3116,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5918693850760963,
+      "grad_norm": 0.6688096625615642,
+      "learning_rate": 3.769115484778374e-06,
+      "loss": 0.3132,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5919514296262871,
+      "grad_norm": 0.6656830971446304,
+      "learning_rate": 3.7678277132207946e-06,
+      "loss": 0.3165,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5920334741764778,
+      "grad_norm": 0.651971566506506,
+      "learning_rate": 3.766540028677244e-06,
+      "loss": 0.3179,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5921155187266686,
+      "grad_norm": 0.6897508569575664,
+      "learning_rate": 3.7652524312386546e-06,
+      "loss": 0.3221,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5921975632768594,
+      "grad_norm": 0.7918441778213916,
+      "learning_rate": 3.763964920995953e-06,
+      "loss": 0.3172,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5922796078270501,
+      "grad_norm": 0.6512561117097561,
+      "learning_rate": 3.7626774980400654e-06,
+      "loss": 0.319,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5923616523772408,
+      "grad_norm": 0.7032321815420306,
+      "learning_rate": 3.7613901624619025e-06,
+      "loss": 0.3072,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5924436969274316,
+      "grad_norm": 0.6759050204132572,
+      "learning_rate": 3.7601029143523767e-06,
+      "loss": 0.2682,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5925257414776224,
+      "grad_norm": 0.6700895978768787,
+      "learning_rate": 3.758815753802393e-06,
+      "loss": 0.2848,
+      "step": 7222
+    },
+    {
+      "epoch": 0.592607786027813,
+      "grad_norm": 0.6680421135136808,
+      "learning_rate": 3.7575286809028455e-06,
+      "loss": 0.3027,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5926898305780038,
+      "grad_norm": 0.6888500843828618,
+      "learning_rate": 3.7562416957446274e-06,
+      "loss": 0.3104,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5927718751281946,
+      "grad_norm": 0.7389586104534029,
+      "learning_rate": 3.7549547984186204e-06,
+      "loss": 0.328,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5928539196783854,
+      "grad_norm": 0.6806113925253936,
+      "learning_rate": 3.7536679890157052e-06,
+      "loss": 0.2966,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5929359642285761,
+      "grad_norm": 0.7077353145315458,
+      "learning_rate": 3.7523812676267563e-06,
+      "loss": 0.3078,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5930180087787669,
+      "grad_norm": 0.7184138541277988,
+      "learning_rate": 3.7510946343426356e-06,
+      "loss": 0.3537,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5931000533289577,
+      "grad_norm": 0.7184675785292883,
+      "learning_rate": 3.749808089254208e-06,
+      "loss": 0.2996,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5931820978791483,
+      "grad_norm": 0.6692012974499022,
+      "learning_rate": 3.748521632452323e-06,
+      "loss": 0.2841,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5932641424293391,
+      "grad_norm": 0.6773130516246277,
+      "learning_rate": 3.7472352640278287e-06,
+      "loss": 0.3123,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5933461869795299,
+      "grad_norm": 0.7392915702466201,
+      "learning_rate": 3.7459489840715694e-06,
+      "loss": 0.357,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5934282315297207,
+      "grad_norm": 0.6782726310269451,
+      "learning_rate": 3.7446627926743763e-06,
+      "loss": 0.2948,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5935102760799114,
+      "grad_norm": 0.6809573472011539,
+      "learning_rate": 3.7433766899270797e-06,
+      "loss": 0.3875,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5935923206301021,
+      "grad_norm": 0.6859074512711146,
+      "learning_rate": 3.7420906759205045e-06,
+      "loss": 0.3547,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5936743651802929,
+      "grad_norm": 0.666873904757583,
+      "learning_rate": 3.740804750745464e-06,
+      "loss": 0.3576,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5937564097304836,
+      "grad_norm": 0.6701863723744211,
+      "learning_rate": 3.739518914492768e-06,
+      "loss": 0.2936,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5938384542806744,
+      "grad_norm": 0.679773633658077,
+      "learning_rate": 3.738233167253222e-06,
+      "loss": 0.3125,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5939204988308652,
+      "grad_norm": 0.7524534966551839,
+      "learning_rate": 3.736947509117621e-06,
+      "loss": 0.3495,
+      "step": 7239
+    },
+    {
+      "epoch": 0.594002543381056,
+      "grad_norm": 0.7770783319244988,
+      "learning_rate": 3.73566194017676e-06,
+      "loss": 0.323,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5940845879312466,
+      "grad_norm": 0.7501839180103946,
+      "learning_rate": 3.7343764605214194e-06,
+      "loss": 0.3385,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5941666324814374,
+      "grad_norm": 0.6742068067919476,
+      "learning_rate": 3.7330910702423818e-06,
+      "loss": 0.3021,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5942486770316282,
+      "grad_norm": 0.8434091326360804,
+      "learning_rate": 3.7318057694304156e-06,
+      "loss": 0.3354,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5943307215818189,
+      "grad_norm": 0.7366218374897413,
+      "learning_rate": 3.7305205581762895e-06,
+      "loss": 0.2835,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5944127661320097,
+      "grad_norm": 1.0855631549824982,
+      "learning_rate": 3.729235436570762e-06,
+      "loss": 0.3066,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5944948106822004,
+      "grad_norm": 0.7120864891154057,
+      "learning_rate": 3.7279504047045855e-06,
+      "loss": 0.3609,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5945768552323912,
+      "grad_norm": 0.7061888123626341,
+      "learning_rate": 3.726665462668509e-06,
+      "loss": 0.3057,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5946588997825819,
+      "grad_norm": 0.6759962244523561,
+      "learning_rate": 3.7253806105532698e-06,
+      "loss": 0.3482,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5947409443327727,
+      "grad_norm": 0.7363265492753155,
+      "learning_rate": 3.724095848449604e-06,
+      "loss": 0.3179,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5948229888829635,
+      "grad_norm": 0.686769550183009,
+      "learning_rate": 3.7228111764482423e-06,
+      "loss": 0.3806,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5949050334331542,
+      "grad_norm": 0.6977129454091969,
+      "learning_rate": 3.7215265946399016e-06,
+      "loss": 0.3323,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5949870779833449,
+      "grad_norm": 0.7801505859272524,
+      "learning_rate": 3.720242103115298e-06,
+      "loss": 0.323,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5950691225335357,
+      "grad_norm": 0.7650211342392836,
+      "learning_rate": 3.7189577019651433e-06,
+      "loss": 0.2681,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5951511670837265,
+      "grad_norm": 0.772891127440572,
+      "learning_rate": 3.717673391280135e-06,
+      "loss": 0.3297,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5952332116339172,
+      "grad_norm": 0.6864706621340303,
+      "learning_rate": 3.7163891711509738e-06,
+      "loss": 0.3283,
+      "step": 7255
+    },
+    {
+      "epoch": 0.595315256184108,
+      "grad_norm": 0.665090747074034,
+      "learning_rate": 3.7151050416683456e-06,
+      "loss": 0.3024,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5953973007342988,
+      "grad_norm": 0.7113259825586545,
+      "learning_rate": 3.713821002922935e-06,
+      "loss": 0.346,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5954793452844894,
+      "grad_norm": 0.6749758276169923,
+      "learning_rate": 3.71253705500542e-06,
+      "loss": 0.3088,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5955613898346802,
+      "grad_norm": 0.7706345084268255,
+      "learning_rate": 3.711253198006468e-06,
+      "loss": 0.3078,
+      "step": 7259
+    },
+    {
+      "epoch": 0.595643434384871,
+      "grad_norm": 0.7358529936646714,
+      "learning_rate": 3.709969432016747e-06,
+      "loss": 0.3454,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5957254789350618,
+      "grad_norm": 0.6368017247116778,
+      "learning_rate": 3.7086857571269097e-06,
+      "loss": 0.3133,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5958075234852525,
+      "grad_norm": 0.6985597627254309,
+      "learning_rate": 3.707402173427609e-06,
+      "loss": 0.3158,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5958895680354432,
+      "grad_norm": 0.734438450284418,
+      "learning_rate": 3.706118681009493e-06,
+      "loss": 0.3168,
+      "step": 7263
+    },
+    {
+      "epoch": 0.595971612585634,
+      "grad_norm": 0.6853146413745365,
+      "learning_rate": 3.7048352799631957e-06,
+      "loss": 0.3236,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5960536571358247,
+      "grad_norm": 0.7608783916951116,
+      "learning_rate": 3.7035519703793497e-06,
+      "loss": 0.3315,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5961357016860155,
+      "grad_norm": 0.8123752372381187,
+      "learning_rate": 3.702268752348581e-06,
+      "loss": 0.3328,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5962177462362063,
+      "grad_norm": 0.6622995668208891,
+      "learning_rate": 3.7009856259615074e-06,
+      "loss": 0.303,
+      "step": 7267
+    },
+    {
+      "epoch": 0.596299790786397,
+      "grad_norm": 0.7810517645253268,
+      "learning_rate": 3.699702591308743e-06,
+      "loss": 0.2835,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5963818353365877,
+      "grad_norm": 0.6686856025464344,
+      "learning_rate": 3.698419648480891e-06,
+      "loss": 0.3245,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5964638798867785,
+      "grad_norm": 0.6346466841490684,
+      "learning_rate": 3.6971367975685536e-06,
+      "loss": 0.2829,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5965459244369693,
+      "grad_norm": 0.6981589267238595,
+      "learning_rate": 3.695854038662322e-06,
+      "loss": 0.3305,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59662796898716,
+      "grad_norm": 0.6838059610423393,
+      "learning_rate": 3.6945713718527832e-06,
+      "loss": 0.3408,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5967100135373508,
+      "grad_norm": 0.7070381234745083,
+      "learning_rate": 3.6932887972305177e-06,
+      "loss": 0.3159,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5967920580875415,
+      "grad_norm": 0.6690927082684344,
+      "learning_rate": 3.6920063148860954e-06,
+      "loss": 0.308,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5968741026377323,
+      "grad_norm": 0.7190410605437944,
+      "learning_rate": 3.6907239249100857e-06,
+      "loss": 0.3302,
+      "step": 7275
+    },
+    {
+      "epoch": 0.596956147187923,
+      "grad_norm": 0.6518797660523141,
+      "learning_rate": 3.689441627393051e-06,
+      "loss": 0.2848,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5970381917381138,
+      "grad_norm": 0.7090811376041319,
+      "learning_rate": 3.6881594224255414e-06,
+      "loss": 0.3346,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5971202362883046,
+      "grad_norm": 0.6489565772951977,
+      "learning_rate": 3.6868773100981083e-06,
+      "loss": 0.3044,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5972022808384952,
+      "grad_norm": 0.6510453408622608,
+      "learning_rate": 3.685595290501288e-06,
+      "loss": 0.3256,
+      "step": 7279
+    },
+    {
+      "epoch": 0.597284325388686,
+      "grad_norm": 1.148534745385316,
+      "learning_rate": 3.6843133637256158e-06,
+      "loss": 0.292,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5973663699388768,
+      "grad_norm": 0.7341198833679929,
+      "learning_rate": 3.6830315298616215e-06,
+      "loss": 0.3216,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5974484144890676,
+      "grad_norm": 0.658285468798146,
+      "learning_rate": 3.6817497889998233e-06,
+      "loss": 0.3459,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5975304590392583,
+      "grad_norm": 0.6997172678820773,
+      "learning_rate": 3.680468141230739e-06,
+      "loss": 0.2866,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5976125035894491,
+      "grad_norm": 0.6481588571625304,
+      "learning_rate": 3.679186586644873e-06,
+      "loss": 0.3101,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5976945481396398,
+      "grad_norm": 0.7178630366960127,
+      "learning_rate": 3.677905125332728e-06,
+      "loss": 0.3507,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5977765926898306,
+      "grad_norm": 0.6576814291414941,
+      "learning_rate": 3.676623757384801e-06,
+      "loss": 0.3162,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5978586372400213,
+      "grad_norm": 0.6416342409749265,
+      "learning_rate": 3.675342482891576e-06,
+      "loss": 0.3362,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5979406817902121,
+      "grad_norm": 0.6588084403267093,
+      "learning_rate": 3.6740613019435356e-06,
+      "loss": 0.3184,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5980227263404029,
+      "grad_norm": 0.9191869639003687,
+      "learning_rate": 3.672780214631159e-06,
+      "loss": 0.3224,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5981047708905936,
+      "grad_norm": 0.768504225771457,
+      "learning_rate": 3.6714992210449084e-06,
+      "loss": 0.3089,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5981868154407843,
+      "grad_norm": 0.6488379547114949,
+      "learning_rate": 3.6702183212752513e-06,
+      "loss": 0.3014,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5982688599909751,
+      "grad_norm": 0.7222790851309039,
+      "learning_rate": 3.6689375154126384e-06,
+      "loss": 0.3621,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5983509045411659,
+      "grad_norm": 0.6700465501766817,
+      "learning_rate": 3.667656803547519e-06,
+      "loss": 0.37,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5984329490913566,
+      "grad_norm": 0.7438818332623135,
+      "learning_rate": 3.666376185770337e-06,
+      "loss": 0.3051,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5985149936415474,
+      "grad_norm": 0.7261410157720272,
+      "learning_rate": 3.6650956621715246e-06,
+      "loss": 0.3341,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5985970381917382,
+      "grad_norm": 0.6803243049794409,
+      "learning_rate": 3.6638152328415134e-06,
+      "loss": 0.307,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5986790827419288,
+      "grad_norm": 0.7715378664035378,
+      "learning_rate": 3.662534897870722e-06,
+      "loss": 0.312,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5987611272921196,
+      "grad_norm": 0.6351767010365732,
+      "learning_rate": 3.661254657349568e-06,
+      "loss": 0.3708,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5988431718423104,
+      "grad_norm": 0.7757449025974851,
+      "learning_rate": 3.65997451136846e-06,
+      "loss": 0.3405,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5989252163925012,
+      "grad_norm": 0.7040245667837078,
+      "learning_rate": 3.6586944600177987e-06,
+      "loss": 0.3275,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5990072609426919,
+      "grad_norm": 0.7331961300491002,
+      "learning_rate": 3.6574145033879806e-06,
+      "loss": 0.2972,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5990893054928826,
+      "grad_norm": 0.709826309170206,
+      "learning_rate": 3.656134641569391e-06,
+      "loss": 0.3326,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5991713500430734,
+      "grad_norm": 0.6540036991545352,
+      "learning_rate": 3.654854874652414e-06,
+      "loss": 0.3218,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5992533945932641,
+      "grad_norm": 0.614396156043077,
+      "learning_rate": 3.653575202727427e-06,
+      "loss": 0.302,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5993354391434549,
+      "grad_norm": 0.7327629423396594,
+      "learning_rate": 3.652295625884793e-06,
+      "loss": 0.2887,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5994174836936457,
+      "grad_norm": 0.6628188838512677,
+      "learning_rate": 3.6510161442148783e-06,
+      "loss": 0.3229,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5994995282438365,
+      "grad_norm": 0.7430644297537885,
+      "learning_rate": 3.649736757808038e-06,
+      "loss": 0.3521,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5995815727940271,
+      "grad_norm": 0.7156335800722952,
+      "learning_rate": 3.6484574667546157e-06,
+      "loss": 0.3319,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5996636173442179,
+      "grad_norm": 0.6669503415210385,
+      "learning_rate": 3.647178271144958e-06,
+      "loss": 0.3207,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5997456618944087,
+      "grad_norm": 0.644421249431716,
+      "learning_rate": 3.6458991710693946e-06,
+      "loss": 0.3435,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5998277064445994,
+      "grad_norm": 0.6302408643465133,
+      "learning_rate": 3.644620166618257e-06,
+      "loss": 0.2999,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5999097509947902,
+      "grad_norm": 0.6843770928777246,
+      "learning_rate": 3.643341257881867e-06,
+      "loss": 0.2902,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5999917955449809,
+      "grad_norm": 0.6340774882562198,
+      "learning_rate": 3.642062444950537e-06,
+      "loss": 0.3288,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6000738400951717,
+      "grad_norm": 0.6893527728179772,
+      "learning_rate": 3.6407837279145765e-06,
+      "loss": 0.3466,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6001558846453624,
+      "grad_norm": 0.6645323397570695,
+      "learning_rate": 3.6395051068642833e-06,
+      "loss": 0.318,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6002379291955532,
+      "grad_norm": 0.8062779474890003,
+      "learning_rate": 3.6382265818899537e-06,
+      "loss": 0.296,
+      "step": 7316
+    },
+    {
+      "epoch": 0.600319973745744,
+      "grad_norm": 0.7757773214514152,
+      "learning_rate": 3.6369481530818775e-06,
+      "loss": 0.324,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6004020182959346,
+      "grad_norm": 0.6655574221349635,
+      "learning_rate": 3.63566982053033e-06,
+      "loss": 0.318,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6004840628461254,
+      "grad_norm": 0.7160850135958331,
+      "learning_rate": 3.6343915843255904e-06,
+      "loss": 0.3651,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6005661073963162,
+      "grad_norm": 0.6853297709204458,
+      "learning_rate": 3.633113444557922e-06,
+      "loss": 0.3211,
+      "step": 7320
+    },
+    {
+      "epoch": 0.600648151946507,
+      "grad_norm": 0.7659765860091136,
+      "learning_rate": 3.631835401317585e-06,
+      "loss": 0.3283,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6007301964966977,
+      "grad_norm": 0.6896435234190944,
+      "learning_rate": 3.630557454694835e-06,
+      "loss": 0.3084,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6008122410468885,
+      "grad_norm": 0.7480456168807056,
+      "learning_rate": 3.629279604779917e-06,
+      "loss": 0.3106,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6008942855970792,
+      "grad_norm": 0.8452089257104003,
+      "learning_rate": 3.62800185166307e-06,
+      "loss": 0.3128,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6009763301472699,
+      "grad_norm": 0.6763056392831096,
+      "learning_rate": 3.62672419543453e-06,
+      "loss": 0.3479,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6010583746974607,
+      "grad_norm": 0.7025491851576205,
+      "learning_rate": 3.6254466361845204e-06,
+      "loss": 0.3362,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6011404192476515,
+      "grad_norm": 0.7572833314009506,
+      "learning_rate": 3.624169174003261e-06,
+      "loss": 0.3244,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6012224637978423,
+      "grad_norm": 0.7037637387428657,
+      "learning_rate": 3.622891808980964e-06,
+      "loss": 0.3187,
+      "step": 7328
+    },
+    {
+      "epoch": 0.601304508348033,
+      "grad_norm": 0.7168570824766943,
+      "learning_rate": 3.621614541207833e-06,
+      "loss": 0.327,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6013865528982237,
+      "grad_norm": 0.8246001669015819,
+      "learning_rate": 3.6203373707740718e-06,
+      "loss": 0.3085,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6014685974484145,
+      "grad_norm": 0.6674883812313396,
+      "learning_rate": 3.619060297769866e-06,
+      "loss": 0.3188,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6015506419986052,
+      "grad_norm": 0.6254276309516091,
+      "learning_rate": 3.6177833222854062e-06,
+      "loss": 0.3376,
+      "step": 7332
+    },
+    {
+      "epoch": 0.601632686548796,
+      "grad_norm": 0.7095592486625008,
+      "learning_rate": 3.616506444410865e-06,
+      "loss": 0.3022,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6017147310989868,
+      "grad_norm": 0.6069013594220055,
+      "learning_rate": 3.6152296642364172e-06,
+      "loss": 0.2981,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6017967756491776,
+      "grad_norm": 0.6275576457744964,
+      "learning_rate": 3.613952981852226e-06,
+      "loss": 0.308,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6018788201993682,
+      "grad_norm": 0.7505411042577407,
+      "learning_rate": 3.612676397348447e-06,
+      "loss": 0.3095,
+      "step": 7336
+    },
+    {
+      "epoch": 0.601960864749559,
+      "grad_norm": 0.6736870628585179,
+      "learning_rate": 3.6113999108152342e-06,
+      "loss": 0.3242,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6020429092997498,
+      "grad_norm": 0.683029094823687,
+      "learning_rate": 3.6101235223427256e-06,
+      "loss": 0.3014,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6021249538499405,
+      "grad_norm": 0.7495402782139019,
+      "learning_rate": 3.608847232021062e-06,
+      "loss": 0.3014,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6022069984001313,
+      "grad_norm": 0.6903116043766065,
+      "learning_rate": 3.607571039940373e-06,
+      "loss": 0.3013,
+      "step": 7340
+    },
+    {
+      "epoch": 0.602289042950322,
+      "grad_norm": 0.7961319355121266,
+      "learning_rate": 3.606294946190779e-06,
+      "loss": 0.2938,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6023710875005128,
+      "grad_norm": 0.6546309733361485,
+      "learning_rate": 3.6050189508623966e-06,
+      "loss": 0.2821,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6024531320507035,
+      "grad_norm": 0.7017238792098635,
+      "learning_rate": 3.6037430540453366e-06,
+      "loss": 0.2985,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6025351766008943,
+      "grad_norm": 0.6168873353332147,
+      "learning_rate": 3.6024672558296964e-06,
+      "loss": 0.3305,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6026172211510851,
+      "grad_norm": 0.7085224191700044,
+      "learning_rate": 3.6011915563055754e-06,
+      "loss": 0.3209,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6026992657012757,
+      "grad_norm": 0.6956024028922172,
+      "learning_rate": 3.599915955563057e-06,
+      "loss": 0.3095,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6027813102514665,
+      "grad_norm": 0.6364544491451555,
+      "learning_rate": 3.598640453692226e-06,
+      "loss": 0.2898,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6028633548016573,
+      "grad_norm": 0.6877773279864221,
+      "learning_rate": 3.5973650507831558e-06,
+      "loss": 0.2909,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6029453993518481,
+      "grad_norm": 0.7181730749608688,
+      "learning_rate": 3.596089746925909e-06,
+      "loss": 0.3627,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6030274439020388,
+      "grad_norm": 0.7186457767734491,
+      "learning_rate": 3.594814542210552e-06,
+      "loss": 0.3182,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6031094884522296,
+      "grad_norm": 0.8326075545963394,
+      "learning_rate": 3.593539436727132e-06,
+      "loss": 0.3444,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6031915330024203,
+      "grad_norm": 1.0161111494374915,
+      "learning_rate": 3.592264430565697e-06,
+      "loss": 0.3156,
+      "step": 7352
+    },
+    {
+      "epoch": 0.603273577552611,
+      "grad_norm": 0.7077551312696579,
+      "learning_rate": 3.590989523816289e-06,
+      "loss": 0.3349,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6033556221028018,
+      "grad_norm": 0.6711663347979142,
+      "learning_rate": 3.5897147165689357e-06,
+      "loss": 0.3,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6034376666529926,
+      "grad_norm": 0.7659970535314726,
+      "learning_rate": 3.5884400089136628e-06,
+      "loss": 0.2805,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6035197112031834,
+      "grad_norm": 0.7126632227606037,
+      "learning_rate": 3.5871654009404895e-06,
+      "loss": 0.3552,
+      "step": 7356
+    },
+    {
+      "epoch": 0.603601755753374,
+      "grad_norm": 0.687222799555183,
+      "learning_rate": 3.5858908927394244e-06,
+      "loss": 0.3275,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6036838003035648,
+      "grad_norm": 0.6752854574759234,
+      "learning_rate": 3.584616484400474e-06,
+      "loss": 0.2757,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6037658448537556,
+      "grad_norm": 0.7002739935750442,
+      "learning_rate": 3.5833421760136323e-06,
+      "loss": 0.2999,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6038478894039463,
+      "grad_norm": 0.6637448876898706,
+      "learning_rate": 3.5820679676688904e-06,
+      "loss": 0.3214,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6039299339541371,
+      "grad_norm": 0.8143801171647592,
+      "learning_rate": 3.580793859456231e-06,
+      "loss": 0.3226,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6040119785043279,
+      "grad_norm": 0.6652568050805074,
+      "learning_rate": 3.5795198514656287e-06,
+      "loss": 0.3285,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6040940230545186,
+      "grad_norm": 0.9996737845530881,
+      "learning_rate": 3.578245943787053e-06,
+      "loss": 0.3471,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6041760676047093,
+      "grad_norm": 0.6835515666886643,
+      "learning_rate": 3.5769721365104635e-06,
+      "loss": 0.307,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6042581121549001,
+      "grad_norm": 0.6319480167506479,
+      "learning_rate": 3.5756984297258145e-06,
+      "loss": 0.3251,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6043401567050909,
+      "grad_norm": 0.7633612609637127,
+      "learning_rate": 3.5744248235230572e-06,
+      "loss": 0.322,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6044222012552816,
+      "grad_norm": 0.6983044596098176,
+      "learning_rate": 3.5731513179921257e-06,
+      "loss": 0.3401,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6045042458054724,
+      "grad_norm": 0.7699145972715643,
+      "learning_rate": 3.571877913222958e-06,
+      "loss": 0.2962,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6045862903556631,
+      "grad_norm": 0.9534719818931351,
+      "learning_rate": 3.5706046093054768e-06,
+      "loss": 0.3094,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6046683349058539,
+      "grad_norm": 0.7654749159009464,
+      "learning_rate": 3.5693314063296005e-06,
+      "loss": 0.2778,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6047503794560446,
+      "grad_norm": 0.778779999629643,
+      "learning_rate": 3.5680583043852434e-06,
+      "loss": 0.3069,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6048324240062354,
+      "grad_norm": 0.7072589181535862,
+      "learning_rate": 3.566785303562306e-06,
+      "loss": 0.345,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6049144685564262,
+      "grad_norm": 0.6008922772180731,
+      "learning_rate": 3.5655124039506905e-06,
+      "loss": 0.3203,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6049965131066168,
+      "grad_norm": 0.6814734053594317,
+      "learning_rate": 3.5642396056402816e-06,
+      "loss": 0.2876,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6050785576568076,
+      "grad_norm": 0.7088797464969373,
+      "learning_rate": 3.5629669087209668e-06,
+      "loss": 0.3022,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6051606022069984,
+      "grad_norm": 0.7059994934401025,
+      "learning_rate": 3.56169431328262e-06,
+      "loss": 0.3324,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6052426467571892,
+      "grad_norm": 0.7001294319173045,
+      "learning_rate": 3.5604218194151096e-06,
+      "loss": 0.3459,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6053246913073799,
+      "grad_norm": 0.7019431379833707,
+      "learning_rate": 3.5591494272082966e-06,
+      "loss": 0.3012,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6054067358575707,
+      "grad_norm": 0.7905428869275991,
+      "learning_rate": 3.557877136752038e-06,
+      "loss": 0.3043,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6054887804077614,
+      "grad_norm": 0.7470098514730407,
+      "learning_rate": 3.5566049481361776e-06,
+      "loss": 0.3339,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6055708249579521,
+      "grad_norm": 0.6408536521263387,
+      "learning_rate": 3.555332861450559e-06,
+      "loss": 0.3019,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6056528695081429,
+      "grad_norm": 0.6479048377891653,
+      "learning_rate": 3.5540608767850106e-06,
+      "loss": 0.3498,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6057349140583337,
+      "grad_norm": 0.6623904995532996,
+      "learning_rate": 3.5527889942293594e-06,
+      "loss": 0.3344,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6058169586085245,
+      "grad_norm": 0.7484588443887322,
+      "learning_rate": 3.5515172138734268e-06,
+      "loss": 0.3091,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6058990031587151,
+      "grad_norm": 0.7673271814812999,
+      "learning_rate": 3.5502455358070187e-06,
+      "loss": 0.2999,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6059810477089059,
+      "grad_norm": 0.7269810305330422,
+      "learning_rate": 3.548973960119944e-06,
+      "loss": 0.3507,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6060630922590967,
+      "grad_norm": 0.6443759220490902,
+      "learning_rate": 3.547702486901995e-06,
+      "loss": 0.2723,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6061451368092874,
+      "grad_norm": 0.669675957277011,
+      "learning_rate": 3.5464311162429645e-06,
+      "loss": 0.3219,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6062271813594782,
+      "grad_norm": 0.7973761569122617,
+      "learning_rate": 3.545159848232632e-06,
+      "loss": 0.3506,
+      "step": 7389
+    },
+    {
+      "epoch": 0.606309225909669,
+      "grad_norm": 0.633215419330747,
+      "learning_rate": 3.543888682960775e-06,
+      "loss": 0.2881,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6063912704598597,
+      "grad_norm": 0.6959843482386928,
+      "learning_rate": 3.5426176205171603e-06,
+      "loss": 0.3364,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6064733150100504,
+      "grad_norm": 0.651852076149581,
+      "learning_rate": 3.541346660991546e-06,
+      "loss": 0.3129,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6065553595602412,
+      "grad_norm": 0.6559761226515187,
+      "learning_rate": 3.540075804473686e-06,
+      "loss": 0.3443,
+      "step": 7393
+    },
+    {
+      "epoch": 0.606637404110432,
+      "grad_norm": 0.6328817715962887,
+      "learning_rate": 3.53880505105333e-06,
+      "loss": 0.3248,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6067194486606228,
+      "grad_norm": 0.6834376408223483,
+      "learning_rate": 3.5375344008202115e-06,
+      "loss": 0.3231,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6068014932108134,
+      "grad_norm": 0.6197874129690384,
+      "learning_rate": 3.5362638538640646e-06,
+      "loss": 0.6706,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6068835377610042,
+      "grad_norm": 0.646493787883927,
+      "learning_rate": 3.534993410274615e-06,
+      "loss": 0.3261,
+      "step": 7397
+    },
+    {
+      "epoch": 0.606965582311195,
+      "grad_norm": 0.6174009631790393,
+      "learning_rate": 3.5337230701415737e-06,
+      "loss": 0.322,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6070476268613857,
+      "grad_norm": 0.6772883884459907,
+      "learning_rate": 3.5324528335546558e-06,
+      "loss": 0.2744,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6071296714115765,
+      "grad_norm": 0.6253233793249897,
+      "learning_rate": 3.5311827006035594e-06,
+      "loss": 0.306,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6072117159617673,
+      "grad_norm": 0.7211769706894017,
+      "learning_rate": 3.52991267137798e-06,
+      "loss": 0.3157,
+      "step": 7401
+    },
+    {
+      "epoch": 0.607293760511958,
+      "grad_norm": 0.6746080028012422,
+      "learning_rate": 3.528642745967609e-06,
+      "loss": 0.312,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6073758050621487,
+      "grad_norm": 0.7681626608829278,
+      "learning_rate": 3.527372924462121e-06,
+      "loss": 0.3309,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6074578496123395,
+      "grad_norm": 0.6757909572389573,
+      "learning_rate": 3.5261032069511926e-06,
+      "loss": 0.3441,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6075398941625303,
+      "grad_norm": 0.622804551905367,
+      "learning_rate": 3.524833593524487e-06,
+      "loss": 0.3003,
+      "step": 7405
+    },
+    {
+      "epoch": 0.607621938712721,
+      "grad_norm": 0.6992014377809824,
+      "learning_rate": 3.523564084271662e-06,
+      "loss": 0.3158,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6077039832629118,
+      "grad_norm": 0.6359507508842861,
+      "learning_rate": 3.5222946792823707e-06,
+      "loss": 0.3125,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6077860278131025,
+      "grad_norm": 0.7471791525278911,
+      "learning_rate": 3.5210253786462537e-06,
+      "loss": 0.3001,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6078680723632933,
+      "grad_norm": 0.7387088881620653,
+      "learning_rate": 3.5197561824529503e-06,
+      "loss": 0.3093,
+      "step": 7409
+    },
+    {
+      "epoch": 0.607950116913484,
+      "grad_norm": 0.6420529881391406,
+      "learning_rate": 3.5184870907920866e-06,
+      "loss": 0.3016,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6080321614636748,
+      "grad_norm": 0.6181070648408427,
+      "learning_rate": 3.5172181037532826e-06,
+      "loss": 0.3001,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6081142060138656,
+      "grad_norm": 0.6755679450463142,
+      "learning_rate": 3.515949221426156e-06,
+      "loss": 0.3226,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6081962505640562,
+      "grad_norm": 0.9543806981890995,
+      "learning_rate": 3.514680443900309e-06,
+      "loss": 0.3158,
+      "step": 7413
+    },
+    {
+      "epoch": 0.608278295114247,
+      "grad_norm": 0.6282908880820016,
+      "learning_rate": 3.5134117712653425e-06,
+      "loss": 0.3036,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6083603396644378,
+      "grad_norm": 0.695835357858873,
+      "learning_rate": 3.5121432036108504e-06,
+      "loss": 0.2856,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6084423842146286,
+      "grad_norm": 0.6845448399419068,
+      "learning_rate": 3.510874741026414e-06,
+      "loss": 0.3136,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6085244287648193,
+      "grad_norm": 0.6454885469309778,
+      "learning_rate": 3.5096063836016113e-06,
+      "loss": 0.2954,
+      "step": 7417
+    },
+    {
+      "epoch": 0.60860647331501,
+      "grad_norm": 0.6798801698306891,
+      "learning_rate": 3.5083381314260113e-06,
+      "loss": 0.3448,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6086885178652008,
+      "grad_norm": 0.7067141761099209,
+      "learning_rate": 3.5070699845891743e-06,
+      "loss": 0.2999,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6087705624153915,
+      "grad_norm": 0.6431631379377862,
+      "learning_rate": 3.505801943180659e-06,
+      "loss": 0.3059,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6088526069655823,
+      "grad_norm": 0.8082708547298518,
+      "learning_rate": 3.504534007290007e-06,
+      "loss": 0.338,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6089346515157731,
+      "grad_norm": 0.7065092196493253,
+      "learning_rate": 3.503266177006763e-06,
+      "loss": 0.2908,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6090166960659639,
+      "grad_norm": 0.6720289898318801,
+      "learning_rate": 3.5019984524204553e-06,
+      "loss": 0.2872,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6090987406161545,
+      "grad_norm": 0.7306292578149894,
+      "learning_rate": 3.5007308336206114e-06,
+      "loss": 0.3256,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6091807851663453,
+      "grad_norm": 0.7041077903241723,
+      "learning_rate": 3.4994633206967478e-06,
+      "loss": 0.3007,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6092628297165361,
+      "grad_norm": 0.6919639399491972,
+      "learning_rate": 3.498195913738372e-06,
+      "loss": 0.3182,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6093448742667268,
+      "grad_norm": 0.6869441225456919,
+      "learning_rate": 3.4969286128349877e-06,
+      "loss": 0.3279,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6094269188169176,
+      "grad_norm": 0.6909316262050522,
+      "learning_rate": 3.4956614180760918e-06,
+      "loss": 0.3076,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6095089633671084,
+      "grad_norm": 0.6583538539303995,
+      "learning_rate": 3.4943943295511678e-06,
+      "loss": 0.2888,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6095910079172991,
+      "grad_norm": 0.6848451776811375,
+      "learning_rate": 3.4931273473496995e-06,
+      "loss": 0.294,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6096730524674898,
+      "grad_norm": 0.6289519283637874,
+      "learning_rate": 3.4918604715611558e-06,
+      "loss": 0.2974,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6097550970176806,
+      "grad_norm": 0.6235288617880282,
+      "learning_rate": 3.4905937022750014e-06,
+      "loss": 0.6493,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6098371415678714,
+      "grad_norm": 1.8024893571225273,
+      "learning_rate": 3.489327039580697e-06,
+      "loss": 0.3126,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6099191861180621,
+      "grad_norm": 0.7206758566358886,
+      "learning_rate": 3.488060483567688e-06,
+      "loss": 0.3033,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6100012306682528,
+      "grad_norm": 0.6641295792764199,
+      "learning_rate": 3.4867940343254203e-06,
+      "loss": 0.3402,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6100832752184436,
+      "grad_norm": 0.6278915720931426,
+      "learning_rate": 3.4855276919433246e-06,
+      "loss": 0.2968,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6101653197686344,
+      "grad_norm": 0.7529146756421482,
+      "learning_rate": 3.4842614565108314e-06,
+      "loss": 0.3346,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6102473643188251,
+      "grad_norm": 0.7489740627491908,
+      "learning_rate": 3.482995328117359e-06,
+      "loss": 0.3299,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6103294088690159,
+      "grad_norm": 0.7229818550327285,
+      "learning_rate": 3.4817293068523196e-06,
+      "loss": 0.364,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6104114534192067,
+      "grad_norm": 0.6300617620884528,
+      "learning_rate": 3.4804633928051173e-06,
+      "loss": 0.2755,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6104934979693973,
+      "grad_norm": 0.6527034700295054,
+      "learning_rate": 3.4791975860651474e-06,
+      "loss": 0.3117,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6105755425195881,
+      "grad_norm": 0.620131097526273,
+      "learning_rate": 3.4779318867218004e-06,
+      "loss": 0.2655,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6106575870697789,
+      "grad_norm": 0.6670204604813074,
+      "learning_rate": 3.4766662948644602e-06,
+      "loss": 0.2992,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6107396316199697,
+      "grad_norm": 0.6110338225225188,
+      "learning_rate": 3.475400810582498e-06,
+      "loss": 0.3386,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6108216761701604,
+      "grad_norm": 0.6933367853672178,
+      "learning_rate": 3.47413543396528e-06,
+      "loss": 0.3218,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6109037207203512,
+      "grad_norm": 0.7006785064662148,
+      "learning_rate": 3.472870165102169e-06,
+      "loss": 0.3159,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6109857652705419,
+      "grad_norm": 0.696240299575665,
+      "learning_rate": 3.47160500408251e-06,
+      "loss": 0.3279,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6110678098207326,
+      "grad_norm": 0.7230563246017028,
+      "learning_rate": 3.4703399509956533e-06,
+      "loss": 0.3028,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6111498543709234,
+      "grad_norm": 0.6770126729315632,
+      "learning_rate": 3.4690750059309284e-06,
+      "loss": 0.3289,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6112318989211142,
+      "grad_norm": 0.6830809862984007,
+      "learning_rate": 3.467810168977669e-06,
+      "loss": 0.3088,
+      "step": 7450
+    },
+    {
+      "epoch": 0.611313943471305,
+      "grad_norm": 0.7175110397263255,
+      "learning_rate": 3.4665454402251937e-06,
+      "loss": 0.3088,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6113959880214956,
+      "grad_norm": 0.742760267955176,
+      "learning_rate": 3.465280819762815e-06,
+      "loss": 0.2988,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6114780325716864,
+      "grad_norm": 0.7525410556688317,
+      "learning_rate": 3.464016307679841e-06,
+      "loss": 0.2996,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6115600771218772,
+      "grad_norm": 0.6883906958773317,
+      "learning_rate": 3.4627519040655656e-06,
+      "loss": 0.3052,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6116421216720679,
+      "grad_norm": 1.0023920113721536,
+      "learning_rate": 3.4614876090092797e-06,
+      "loss": 0.3287,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6117241662222587,
+      "grad_norm": 0.6433948661045505,
+      "learning_rate": 3.4602234226002706e-06,
+      "loss": 0.2963,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6118062107724495,
+      "grad_norm": 0.645875729502254,
+      "learning_rate": 3.458959344927807e-06,
+      "loss": 0.2897,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6118882553226402,
+      "grad_norm": 0.6018480969393027,
+      "learning_rate": 3.4576953760811603e-06,
+      "loss": 0.2987,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6119702998728309,
+      "grad_norm": 0.7510275580880507,
+      "learning_rate": 3.456431516149587e-06,
+      "loss": 0.3365,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6120523444230217,
+      "grad_norm": 0.6750041180836782,
+      "learning_rate": 3.455167765222339e-06,
+      "loss": 0.312,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6121343889732125,
+      "grad_norm": 0.6409432525354124,
+      "learning_rate": 3.4539041233886648e-06,
+      "loss": 0.2921,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6122164335234032,
+      "grad_norm": 0.6351739840636988,
+      "learning_rate": 3.4526405907377948e-06,
+      "loss": 0.3173,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6122984780735939,
+      "grad_norm": 0.7773492453751744,
+      "learning_rate": 3.451377167358959e-06,
+      "loss": 0.3241,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6123805226237847,
+      "grad_norm": 0.7419255598687203,
+      "learning_rate": 3.450113853341383e-06,
+      "loss": 0.3136,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6124625671739755,
+      "grad_norm": 0.7356078327938028,
+      "learning_rate": 3.4488506487742756e-06,
+      "loss": 0.3171,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6125446117241662,
+      "grad_norm": 0.7504075311888254,
+      "learning_rate": 3.447587553746843e-06,
+      "loss": 0.3315,
+      "step": 7466
+    },
+    {
+      "epoch": 0.612626656274357,
+      "grad_norm": 0.7249368532136704,
+      "learning_rate": 3.4463245683482827e-06,
+      "loss": 0.3245,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6127087008245478,
+      "grad_norm": 0.6920558309913993,
+      "learning_rate": 3.445061692667785e-06,
+      "loss": 0.3239,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6127907453747384,
+      "grad_norm": 0.7091783837963869,
+      "learning_rate": 3.443798926794534e-06,
+      "loss": 0.307,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6128727899249292,
+      "grad_norm": 0.6501924505780903,
+      "learning_rate": 3.4425362708177003e-06,
+      "loss": 0.3083,
+      "step": 7470
+    },
+    {
+      "epoch": 0.61295483447512,
+      "grad_norm": 0.69317474788538,
+      "learning_rate": 3.441273724826455e-06,
+      "loss": 0.297,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6130368790253108,
+      "grad_norm": 0.8518564800791698,
+      "learning_rate": 3.440011288909953e-06,
+      "loss": 0.3183,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6131189235755015,
+      "grad_norm": 0.7338758390007157,
+      "learning_rate": 3.438748963157349e-06,
+      "loss": 0.324,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6132009681256922,
+      "grad_norm": 0.6083147501084353,
+      "learning_rate": 3.437486747657785e-06,
+      "loss": 0.3253,
+      "step": 7474
+    },
+    {
+      "epoch": 0.613283012675883,
+      "grad_norm": 0.8624364006846237,
+      "learning_rate": 3.4362246425003954e-06,
+      "loss": 0.3134,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6133650572260737,
+      "grad_norm": 0.7195063370826048,
+      "learning_rate": 3.4349626477743104e-06,
+      "loss": 0.2999,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6134471017762645,
+      "grad_norm": 0.6450392048144609,
+      "learning_rate": 3.433700763568647e-06,
+      "loss": 0.3104,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6135291463264553,
+      "grad_norm": 0.8769068910389266,
+      "learning_rate": 3.4324389899725203e-06,
+      "loss": 0.2998,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6136111908766461,
+      "grad_norm": 0.6331468860262652,
+      "learning_rate": 3.431177327075034e-06,
+      "loss": 0.2992,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6136932354268367,
+      "grad_norm": 0.7048234377565898,
+      "learning_rate": 3.4299157749652832e-06,
+      "loss": 0.33,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6137752799770275,
+      "grad_norm": 0.7951126305980978,
+      "learning_rate": 3.4286543337323584e-06,
+      "loss": 0.3177,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6138573245272183,
+      "grad_norm": 0.712567008164625,
+      "learning_rate": 3.4273930034653414e-06,
+      "loss": 0.2944,
+      "step": 7482
+    },
+    {
+      "epoch": 0.613939369077409,
+      "grad_norm": 0.5737334118663148,
+      "learning_rate": 3.4261317842533016e-06,
+      "loss": 0.2765,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6140214136275998,
+      "grad_norm": 0.6680393775927773,
+      "learning_rate": 3.424870676185309e-06,
+      "loss": 0.3313,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6141034581777906,
+      "grad_norm": 0.6515628958148214,
+      "learning_rate": 3.423609679350416e-06,
+      "loss": 0.2828,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6141855027279813,
+      "grad_norm": 0.6802206110074748,
+      "learning_rate": 3.422348793837676e-06,
+      "loss": 0.2846,
+      "step": 7486
+    },
+    {
+      "epoch": 0.614267547278172,
+      "grad_norm": 0.7138411303084394,
+      "learning_rate": 3.4210880197361305e-06,
+      "loss": 0.3341,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6143495918283628,
+      "grad_norm": 0.7254740000325566,
+      "learning_rate": 3.4198273571348107e-06,
+      "loss": 0.3099,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6144316363785536,
+      "grad_norm": 0.6959369156437583,
+      "learning_rate": 3.4185668061227463e-06,
+      "loss": 0.29,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6145136809287443,
+      "grad_norm": 0.7098949857246906,
+      "learning_rate": 3.4173063667889505e-06,
+      "loss": 0.3054,
+      "step": 7490
+    },
+    {
+      "epoch": 0.614595725478935,
+      "grad_norm": 0.6561666675152363,
+      "learning_rate": 3.416046039222437e-06,
+      "loss": 0.2996,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6146777700291258,
+      "grad_norm": 0.6689815466432188,
+      "learning_rate": 3.4147858235122087e-06,
+      "loss": 0.3199,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6147598145793166,
+      "grad_norm": 0.6238478336796217,
+      "learning_rate": 3.4135257197472582e-06,
+      "loss": 0.3382,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6148418591295073,
+      "grad_norm": 0.6319646711504817,
+      "learning_rate": 3.4122657280165727e-06,
+      "loss": 0.2958,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6149239036796981,
+      "grad_norm": 0.7131580924523873,
+      "learning_rate": 3.41100584840913e-06,
+      "loss": 0.3177,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6150059482298889,
+      "grad_norm": 0.6338243893421869,
+      "learning_rate": 3.4097460810139006e-06,
+      "loss": 0.2986,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6150879927800795,
+      "grad_norm": 0.7117382840690364,
+      "learning_rate": 3.40848642591985e-06,
+      "loss": 0.3124,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6151700373302703,
+      "grad_norm": 0.817825824829663,
+      "learning_rate": 3.4072268832159294e-06,
+      "loss": 0.2907,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6152520818804611,
+      "grad_norm": 0.6734279965836947,
+      "learning_rate": 3.405967452991088e-06,
+      "loss": 0.3108,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6153341264306519,
+      "grad_norm": 0.7198714446456803,
+      "learning_rate": 3.4047081353342644e-06,
+      "loss": 0.3165,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6154161709808426,
+      "grad_norm": 0.6746380845039942,
+      "learning_rate": 3.403448930334389e-06,
+      "loss": 0.3224,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6154982155310333,
+      "grad_norm": 0.7217752117853371,
+      "learning_rate": 3.4021898380803864e-06,
+      "loss": 0.3082,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6155802600812241,
+      "grad_norm": 0.6252536658084653,
+      "learning_rate": 3.400930858661168e-06,
+      "loss": 0.2966,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6156623046314148,
+      "grad_norm": 0.6275194096570474,
+      "learning_rate": 3.399671992165643e-06,
+      "loss": 0.3134,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6157443491816056,
+      "grad_norm": 0.6870729858520332,
+      "learning_rate": 3.3984132386827127e-06,
+      "loss": 0.3187,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6158263937317964,
+      "grad_norm": 0.8176781372763191,
+      "learning_rate": 3.397154598301266e-06,
+      "loss": 0.2744,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6159084382819872,
+      "grad_norm": 0.6984125376860711,
+      "learning_rate": 3.395896071110186e-06,
+      "loss": 0.3639,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6159904828321778,
+      "grad_norm": 0.7620619248704,
+      "learning_rate": 3.3946376571983486e-06,
+      "loss": 0.2937,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6160725273823686,
+      "grad_norm": 0.6772811405414948,
+      "learning_rate": 3.3933793566546193e-06,
+      "loss": 0.31,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6161545719325594,
+      "grad_norm": 0.6386768741304198,
+      "learning_rate": 3.3921211695678608e-06,
+      "loss": 0.331,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6162366164827502,
+      "grad_norm": 0.748035218966803,
+      "learning_rate": 3.39086309602692e-06,
+      "loss": 0.3233,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6163186610329409,
+      "grad_norm": 0.7288109461551766,
+      "learning_rate": 3.389605136120644e-06,
+      "loss": 0.3016,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6164007055831316,
+      "grad_norm": 0.6421720516001497,
+      "learning_rate": 3.3883472899378643e-06,
+      "loss": 0.3024,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6164827501333224,
+      "grad_norm": 0.6455920982216345,
+      "learning_rate": 3.3870895575674103e-06,
+      "loss": 0.3378,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6165647946835131,
+      "grad_norm": 0.6869153777451609,
+      "learning_rate": 3.3858319390981014e-06,
+      "loss": 0.3463,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6166468392337039,
+      "grad_norm": 0.788799335802774,
+      "learning_rate": 3.384574434618746e-06,
+      "loss": 0.3234,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6167288837838947,
+      "grad_norm": 0.7791786799299534,
+      "learning_rate": 3.3833170442181485e-06,
+      "loss": 0.3329,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6168109283340855,
+      "grad_norm": 0.613976765340438,
+      "learning_rate": 3.3820597679851053e-06,
+      "loss": 0.2854,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6168929728842761,
+      "grad_norm": 0.7032968891791315,
+      "learning_rate": 3.380802606008401e-06,
+      "loss": 0.3454,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6169750174344669,
+      "grad_norm": 0.7295843461391748,
+      "learning_rate": 3.379545558376816e-06,
+      "loss": 0.3236,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6170570619846577,
+      "grad_norm": 1.2074533913747016,
+      "learning_rate": 3.3782886251791204e-06,
+      "loss": 0.2907,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6171391065348484,
+      "grad_norm": 0.7382944400292841,
+      "learning_rate": 3.377031806504075e-06,
+      "loss": 0.3241,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6172211510850392,
+      "grad_norm": 0.7284580475125412,
+      "learning_rate": 3.3757751024404373e-06,
+      "loss": 0.2874,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61730319563523,
+      "grad_norm": 0.6354568757800065,
+      "learning_rate": 3.374518513076951e-06,
+      "loss": 0.3065,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6173852401854207,
+      "grad_norm": 0.6301469655514427,
+      "learning_rate": 3.3732620385023584e-06,
+      "loss": 0.3303,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6174672847356114,
+      "grad_norm": 0.6657003331054802,
+      "learning_rate": 3.372005678805384e-06,
+      "loss": 0.3075,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6175493292858022,
+      "grad_norm": 0.7616381962015818,
+      "learning_rate": 3.370749434074755e-06,
+      "loss": 0.3191,
+      "step": 7527
+    },
+    {
+      "epoch": 0.617631373835993,
+      "grad_norm": 0.6790141388108708,
+      "learning_rate": 3.3694933043991824e-06,
+      "loss": 0.3154,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6177134183861837,
+      "grad_norm": 0.6506717660850097,
+      "learning_rate": 3.368237289867373e-06,
+      "loss": 0.3312,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6177954629363744,
+      "grad_norm": 0.800079320486423,
+      "learning_rate": 3.3669813905680264e-06,
+      "loss": 0.292,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6178775074865652,
+      "grad_norm": 0.6692564168117523,
+      "learning_rate": 3.365725606589828e-06,
+      "loss": 0.318,
+      "step": 7531
+    },
+    {
+      "epoch": 0.617959552036756,
+      "grad_norm": 0.6935169853576456,
+      "learning_rate": 3.364469938021462e-06,
+      "loss": 0.308,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6180415965869467,
+      "grad_norm": 0.7284017149940257,
+      "learning_rate": 3.3632143849516025e-06,
+      "loss": 0.3047,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6181236411371375,
+      "grad_norm": 0.650959195818578,
+      "learning_rate": 3.361958947468912e-06,
+      "loss": 0.318,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6182056856873283,
+      "grad_norm": 0.6906755832591244,
+      "learning_rate": 3.3607036256620496e-06,
+      "loss": 0.335,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6182877302375189,
+      "grad_norm": 0.7791541392989837,
+      "learning_rate": 3.3594484196196643e-06,
+      "loss": 0.3276,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6183697747877097,
+      "grad_norm": 0.6979164033222154,
+      "learning_rate": 3.358193329430394e-06,
+      "loss": 0.3419,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6184518193379005,
+      "grad_norm": 0.6687142447717542,
+      "learning_rate": 3.3569383551828756e-06,
+      "loss": 0.2964,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6185338638880913,
+      "grad_norm": 0.7235690064184076,
+      "learning_rate": 3.3556834969657283e-06,
+      "loss": 0.3332,
+      "step": 7539
+    },
+    {
+      "epoch": 0.618615908438282,
+      "grad_norm": 0.6426933415746802,
+      "learning_rate": 3.3544287548675713e-06,
+      "loss": 0.28,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6186979529884727,
+      "grad_norm": 0.7403596840122827,
+      "learning_rate": 3.3531741289770124e-06,
+      "loss": 0.3405,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6187799975386635,
+      "grad_norm": 0.7097018396584912,
+      "learning_rate": 3.3519196193826503e-06,
+      "loss": 0.3233,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6188620420888542,
+      "grad_norm": 0.6987950128497856,
+      "learning_rate": 3.350665226173078e-06,
+      "loss": 0.3066,
+      "step": 7543
+    },
+    {
+      "epoch": 0.618944086639045,
+      "grad_norm": 0.6328997098871462,
+      "learning_rate": 3.3494109494368753e-06,
+      "loss": 0.2846,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6190261311892358,
+      "grad_norm": 0.835539530691368,
+      "learning_rate": 3.34815678926262e-06,
+      "loss": 0.3115,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6191081757394266,
+      "grad_norm": 0.6209099137644976,
+      "learning_rate": 3.34690274573888e-06,
+      "loss": 0.2769,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6191902202896172,
+      "grad_norm": 0.6281132850778627,
+      "learning_rate": 3.34564881895421e-06,
+      "loss": 0.3585,
+      "step": 7547
+    },
+    {
+      "epoch": 0.619272264839808,
+      "grad_norm": 0.6575364683374894,
+      "learning_rate": 3.3443950089971655e-06,
+      "loss": 0.2997,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6193543093899988,
+      "grad_norm": 0.7499722401530369,
+      "learning_rate": 3.343141315956283e-06,
+      "loss": 0.3297,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6194363539401895,
+      "grad_norm": 1.2048050150843221,
+      "learning_rate": 3.3418877399200987e-06,
+      "loss": 0.3458,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6195183984903803,
+      "grad_norm": 0.6541952783052477,
+      "learning_rate": 3.3406342809771407e-06,
+      "loss": 0.3076,
+      "step": 7551
+    },
+    {
+      "epoch": 0.619600443040571,
+      "grad_norm": 0.6911348009584296,
+      "learning_rate": 3.339380939215921e-06,
+      "loss": 0.3001,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6196824875907618,
+      "grad_norm": 0.644953118718961,
+      "learning_rate": 3.338127714724951e-06,
+      "loss": 0.308,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6197645321409525,
+      "grad_norm": 0.6901075093458734,
+      "learning_rate": 3.336874607592735e-06,
+      "loss": 0.2494,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6198465766911433,
+      "grad_norm": 0.7143753501988912,
+      "learning_rate": 3.3356216179077604e-06,
+      "loss": 0.3062,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6199286212413341,
+      "grad_norm": 0.7048752312064426,
+      "learning_rate": 3.334368745758514e-06,
+      "loss": 0.332,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6200106657915248,
+      "grad_norm": 0.623070057404465,
+      "learning_rate": 3.33311599123347e-06,
+      "loss": 0.3262,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6200927103417155,
+      "grad_norm": 0.7795394318507216,
+      "learning_rate": 3.3318633544210953e-06,
+      "loss": 0.3306,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6201747548919063,
+      "grad_norm": 0.6540709748484987,
+      "learning_rate": 3.3306108354098534e-06,
+      "loss": 0.2938,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6202567994420971,
+      "grad_norm": 0.6545650471626023,
+      "learning_rate": 3.329358434288189e-06,
+      "loss": 0.3333,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6203388439922878,
+      "grad_norm": 0.8073739079506727,
+      "learning_rate": 3.328106151144551e-06,
+      "loss": 0.3139,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6204208885424786,
+      "grad_norm": 0.6950337823452881,
+      "learning_rate": 3.3268539860673675e-06,
+      "loss": 0.3244,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6205029330926694,
+      "grad_norm": 0.6574677846062877,
+      "learning_rate": 3.3256019391450696e-06,
+      "loss": 0.2973,
+      "step": 7563
+    },
+    {
+      "epoch": 0.62058497764286,
+      "grad_norm": 0.6538032688180658,
+      "learning_rate": 3.324350010466073e-06,
+      "loss": 0.3483,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6206670221930508,
+      "grad_norm": 0.6681391071691531,
+      "learning_rate": 3.3230982001187843e-06,
+      "loss": 0.3002,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6207490667432416,
+      "grad_norm": 0.7462950808152023,
+      "learning_rate": 3.321846508191609e-06,
+      "loss": 0.2912,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6208311112934324,
+      "grad_norm": 0.6103705353176,
+      "learning_rate": 3.320594934772935e-06,
+      "loss": 0.3219,
+      "step": 7567
+    },
+    {
+      "epoch": 0.620913155843623,
+      "grad_norm": 0.6582805026074374,
+      "learning_rate": 3.319343479951148e-06,
+      "loss": 0.3014,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6209952003938138,
+      "grad_norm": 0.6845052462658069,
+      "learning_rate": 3.3180921438146276e-06,
+      "loss": 0.2941,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6210772449440046,
+      "grad_norm": 0.639110799443921,
+      "learning_rate": 3.3168409264517356e-06,
+      "loss": 0.3279,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6211592894941953,
+      "grad_norm": 0.736165339355737,
+      "learning_rate": 3.3155898279508323e-06,
+      "loss": 0.3127,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6212413340443861,
+      "grad_norm": 0.6917465023308873,
+      "learning_rate": 3.3143388484002714e-06,
+      "loss": 0.336,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6213233785945769,
+      "grad_norm": 0.7103985329294301,
+      "learning_rate": 3.3130879878883904e-06,
+      "loss": 0.2795,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6214054231447677,
+      "grad_norm": 0.6904405776453764,
+      "learning_rate": 3.311837246503528e-06,
+      "loss": 0.3117,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6214874676949583,
+      "grad_norm": 0.6184240555202316,
+      "learning_rate": 3.3105866243340045e-06,
+      "loss": 0.3028,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6215695122451491,
+      "grad_norm": 0.65555204952658,
+      "learning_rate": 3.3093361214681406e-06,
+      "loss": 0.3217,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6216515567953399,
+      "grad_norm": 0.6871698603735688,
+      "learning_rate": 3.308085737994245e-06,
+      "loss": 0.3377,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6217336013455306,
+      "grad_norm": 0.7405368004831698,
+      "learning_rate": 3.3068354740006137e-06,
+      "loss": 0.2967,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6218156458957214,
+      "grad_norm": 0.6917438130066078,
+      "learning_rate": 3.3055853295755434e-06,
+      "loss": 0.3238,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6218976904459121,
+      "grad_norm": 0.8261027281497243,
+      "learning_rate": 3.304335304807312e-06,
+      "loss": 0.3166,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6219797349961029,
+      "grad_norm": 0.6637225985660443,
+      "learning_rate": 3.303085399784198e-06,
+      "loss": 0.3027,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6220617795462936,
+      "grad_norm": 0.6545811016796896,
+      "learning_rate": 3.3018356145944686e-06,
+      "loss": 0.3363,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6221438240964844,
+      "grad_norm": 0.6565268784618141,
+      "learning_rate": 3.300585949326379e-06,
+      "loss": 0.3368,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6222258686466752,
+      "grad_norm": 0.7128030550777129,
+      "learning_rate": 3.2993364040681788e-06,
+      "loss": 0.3563,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6223079131968658,
+      "grad_norm": 0.6559060072883954,
+      "learning_rate": 3.2980869789081093e-06,
+      "loss": 0.3172,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6223899577470566,
+      "grad_norm": 0.6230723638420859,
+      "learning_rate": 3.296837673934403e-06,
+      "loss": 0.279,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6224720022972474,
+      "grad_norm": 1.23137226802982,
+      "learning_rate": 3.2955884892352853e-06,
+      "loss": 0.3002,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6225540468474382,
+      "grad_norm": 0.6773851318725477,
+      "learning_rate": 3.2943394248989673e-06,
+      "loss": 0.317,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6226360913976289,
+      "grad_norm": 0.7142611252247938,
+      "learning_rate": 3.293090481013661e-06,
+      "loss": 0.2885,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6227181359478197,
+      "grad_norm": 0.6565767927977971,
+      "learning_rate": 3.2918416576675617e-06,
+      "loss": 0.3272,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6228001804980104,
+      "grad_norm": 0.6510429075026453,
+      "learning_rate": 3.2905929549488602e-06,
+      "loss": 0.2938,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6228822250482011,
+      "grad_norm": 0.6644473141626085,
+      "learning_rate": 3.28934437294574e-06,
+      "loss": 0.3014,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6229642695983919,
+      "grad_norm": 0.8018310771218468,
+      "learning_rate": 3.2880959117463686e-06,
+      "loss": 0.3112,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6230463141485827,
+      "grad_norm": 0.6270969929719568,
+      "learning_rate": 3.2868475714389137e-06,
+      "loss": 0.6607,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6231283586987735,
+      "grad_norm": 0.6733238654310781,
+      "learning_rate": 3.285599352111533e-06,
+      "loss": 0.2765,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6232104032489642,
+      "grad_norm": 0.7574502687870791,
+      "learning_rate": 3.2843512538523696e-06,
+      "loss": 0.3275,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6232924477991549,
+      "grad_norm": 0.6421632014890413,
+      "learning_rate": 3.283103276749566e-06,
+      "loss": 0.3045,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6233744923493457,
+      "grad_norm": 0.6485635386393759,
+      "learning_rate": 3.2818554208912505e-06,
+      "loss": 0.2928,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6234565368995364,
+      "grad_norm": 0.6389389984988755,
+      "learning_rate": 3.2806076863655435e-06,
+      "loss": 0.3587,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6235385814497272,
+      "grad_norm": 0.6447973687662982,
+      "learning_rate": 3.279360073260561e-06,
+      "loss": 0.3386,
+      "step": 7600
+    },
+    {
+      "epoch": 0.623620625999918,
+      "grad_norm": 0.6761695781447997,
+      "learning_rate": 3.278112581664403e-06,
+      "loss": 0.3004,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6237026705501088,
+      "grad_norm": 0.6870705675501044,
+      "learning_rate": 3.276865211665171e-06,
+      "loss": 0.3443,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6237847151002994,
+      "grad_norm": 0.6716941941358516,
+      "learning_rate": 3.275617963350948e-06,
+      "loss": 0.2804,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6238667596504902,
+      "grad_norm": 0.7187412283971988,
+      "learning_rate": 3.274370836809814e-06,
+      "loss": 0.3137,
+      "step": 7604
+    },
+    {
+      "epoch": 0.623948804200681,
+      "grad_norm": 0.656826475728549,
+      "learning_rate": 3.2731238321298397e-06,
+      "loss": 0.289,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6240308487508717,
+      "grad_norm": 0.7163573044219866,
+      "learning_rate": 3.2718769493990843e-06,
+      "loss": 0.3184,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6241128933010625,
+      "grad_norm": 0.6632536538800159,
+      "learning_rate": 3.2706301887056025e-06,
+      "loss": 0.3289,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6241949378512532,
+      "grad_norm": 0.6727117814889959,
+      "learning_rate": 3.2693835501374407e-06,
+      "loss": 0.3029,
+      "step": 7608
+    },
+    {
+      "epoch": 0.624276982401444,
+      "grad_norm": 0.6460300073450925,
+      "learning_rate": 3.2681370337826297e-06,
+      "loss": 0.303,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6243590269516347,
+      "grad_norm": 0.6183842907455513,
+      "learning_rate": 3.266890639729201e-06,
+      "loss": 0.3117,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6244410715018255,
+      "grad_norm": 0.6544970210342772,
+      "learning_rate": 3.2656443680651696e-06,
+      "loss": 0.3635,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6245231160520163,
+      "grad_norm": 0.6656578802042,
+      "learning_rate": 3.2643982188785457e-06,
+      "loss": 0.3563,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6246051606022069,
+      "grad_norm": 0.6313341102642193,
+      "learning_rate": 3.263152192257333e-06,
+      "loss": 0.3003,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6246872051523977,
+      "grad_norm": 0.7374257615090007,
+      "learning_rate": 3.26190628828952e-06,
+      "loss": 0.3229,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6247692497025885,
+      "grad_norm": 0.6675910000584917,
+      "learning_rate": 3.2606605070630946e-06,
+      "loss": 0.2967,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6248512942527793,
+      "grad_norm": 0.7093522486851205,
+      "learning_rate": 3.2594148486660277e-06,
+      "loss": 0.3059,
+      "step": 7616
+    },
+    {
+      "epoch": 0.62493333880297,
+      "grad_norm": 0.685210899494753,
+      "learning_rate": 3.258169313186289e-06,
+      "loss": 0.2957,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6250153833531608,
+      "grad_norm": 0.6685619717718588,
+      "learning_rate": 3.2569239007118346e-06,
+      "loss": 0.3335,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6250974279033515,
+      "grad_norm": 0.699181741198157,
+      "learning_rate": 3.255678611330614e-06,
+      "loss": 0.3061,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6251794724535423,
+      "grad_norm": 0.6974272675691382,
+      "learning_rate": 3.254433445130566e-06,
+      "loss": 0.3031,
+      "step": 7620
+    },
+    {
+      "epoch": 0.625261517003733,
+      "grad_norm": 0.6536429926825554,
+      "learning_rate": 3.2531884021996263e-06,
+      "loss": 0.3021,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6253435615539238,
+      "grad_norm": 0.6727457332862126,
+      "learning_rate": 3.2519434826257123e-06,
+      "loss": 0.2956,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6254256061041146,
+      "grad_norm": 0.743191989580852,
+      "learning_rate": 3.250698686496744e-06,
+      "loss": 0.3317,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6255076506543052,
+      "grad_norm": 0.7527587992987506,
+      "learning_rate": 3.2494540139006212e-06,
+      "loss": 0.2967,
+      "step": 7624
+    },
+    {
+      "epoch": 0.625589695204496,
+      "grad_norm": 0.728557167924075,
+      "learning_rate": 3.248209464925245e-06,
+      "loss": 0.2954,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6256717397546868,
+      "grad_norm": 0.6967469050914696,
+      "learning_rate": 3.246965039658503e-06,
+      "loss": 0.319,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6257537843048776,
+      "grad_norm": 0.7708228526220536,
+      "learning_rate": 3.2457207381882714e-06,
+      "loss": 0.3555,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6258358288550683,
+      "grad_norm": 0.7294127456556637,
+      "learning_rate": 3.2444765606024244e-06,
+      "loss": 0.2876,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6259178734052591,
+      "grad_norm": 0.63903714863856,
+      "learning_rate": 3.2432325069888204e-06,
+      "loss": 0.3045,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6259999179554498,
+      "grad_norm": 0.8141833580975513,
+      "learning_rate": 3.2419885774353144e-06,
+      "loss": 0.3178,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6260819625056405,
+      "grad_norm": 0.6944052889103182,
+      "learning_rate": 3.2407447720297524e-06,
+      "loss": 0.3161,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6261640070558313,
+      "grad_norm": 0.6207601443192018,
+      "learning_rate": 3.239501090859968e-06,
+      "loss": 0.307,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6262460516060221,
+      "grad_norm": 0.6592349664340437,
+      "learning_rate": 3.238257534013788e-06,
+      "loss": 0.2881,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6263280961562129,
+      "grad_norm": 0.6456070353234807,
+      "learning_rate": 3.237014101579029e-06,
+      "loss": 0.3209,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6264101407064036,
+      "grad_norm": 0.6182980263764035,
+      "learning_rate": 3.2357707936435013e-06,
+      "loss": 0.2968,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6264921852565943,
+      "grad_norm": 0.658675778388259,
+      "learning_rate": 3.2345276102950086e-06,
+      "loss": 0.322,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6265742298067851,
+      "grad_norm": 0.6815429627143189,
+      "learning_rate": 3.2332845516213368e-06,
+      "loss": 0.2973,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6266562743569758,
+      "grad_norm": 0.7217139583548198,
+      "learning_rate": 3.2320416177102725e-06,
+      "loss": 0.3086,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6267383189071666,
+      "grad_norm": 0.6669321005894749,
+      "learning_rate": 3.23079880864959e-06,
+      "loss": 0.263,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6268203634573574,
+      "grad_norm": 0.6620074679976954,
+      "learning_rate": 3.229556124527051e-06,
+      "loss": 0.3444,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6269024080075482,
+      "grad_norm": 0.723476903374681,
+      "learning_rate": 3.2283135654304165e-06,
+      "loss": 0.3424,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6269844525577388,
+      "grad_norm": 0.8254437797499928,
+      "learning_rate": 3.227071131447429e-06,
+      "loss": 0.2933,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6270664971079296,
+      "grad_norm": 0.6396991208070041,
+      "learning_rate": 3.22582882266583e-06,
+      "loss": 0.3034,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6271485416581204,
+      "grad_norm": 0.6429618311097878,
+      "learning_rate": 3.2245866391733516e-06,
+      "loss": 0.3121,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6272305862083111,
+      "grad_norm": 0.6878728754207939,
+      "learning_rate": 3.223344581057711e-06,
+      "loss": 0.3563,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6273126307585019,
+      "grad_norm": 0.6489634218843509,
+      "learning_rate": 3.222102648406621e-06,
+      "loss": 0.2872,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6273946753086926,
+      "grad_norm": 0.6645297569114453,
+      "learning_rate": 3.2208608413077865e-06,
+      "loss": 0.3136,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6274767198588834,
+      "grad_norm": 0.6827274864113,
+      "learning_rate": 3.2196191598488996e-06,
+      "loss": 0.323,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6275587644090741,
+      "grad_norm": 0.6544370381231942,
+      "learning_rate": 3.2183776041176495e-06,
+      "loss": 0.3134,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6276408089592649,
+      "grad_norm": 0.6502903720963082,
+      "learning_rate": 3.2171361742017084e-06,
+      "loss": 0.3184,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6277228535094557,
+      "grad_norm": 0.6931191995602433,
+      "learning_rate": 3.2158948701887484e-06,
+      "loss": 0.3025,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6278048980596463,
+      "grad_norm": 0.6193038305977315,
+      "learning_rate": 3.2146536921664234e-06,
+      "loss": 0.3185,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6278869426098371,
+      "grad_norm": 0.738392871207813,
+      "learning_rate": 3.213412640222388e-06,
+      "loss": 0.3355,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6279689871600279,
+      "grad_norm": 0.6890647582385838,
+      "learning_rate": 3.2121717144442828e-06,
+      "loss": 0.2981,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6280510317102187,
+      "grad_norm": 0.6616803790921515,
+      "learning_rate": 3.2109309149197365e-06,
+      "loss": 0.3058,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6281330762604094,
+      "grad_norm": 0.6973855099040438,
+      "learning_rate": 3.2096902417363745e-06,
+      "loss": 0.3215,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6282151208106002,
+      "grad_norm": 0.7461142555205744,
+      "learning_rate": 3.2084496949818134e-06,
+      "loss": 0.3331,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6282971653607909,
+      "grad_norm": 0.710797054548336,
+      "learning_rate": 3.2072092747436546e-06,
+      "loss": 0.3056,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6283792099109816,
+      "grad_norm": 0.7217819045092094,
+      "learning_rate": 3.2059689811094984e-06,
+      "loss": 0.3216,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6284612544611724,
+      "grad_norm": 0.6268763619627743,
+      "learning_rate": 3.2047288141669296e-06,
+      "loss": 0.3264,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6285432990113632,
+      "grad_norm": 0.7107719857504035,
+      "learning_rate": 3.203488774003526e-06,
+      "loss": 0.3371,
+      "step": 7661
+    },
+    {
+      "epoch": 0.628625343561554,
+      "grad_norm": 0.6763816441474215,
+      "learning_rate": 3.2022488607068615e-06,
+      "loss": 0.2917,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6287073881117446,
+      "grad_norm": 0.7244356410442343,
+      "learning_rate": 3.201009074364492e-06,
+      "loss": 0.3552,
+      "step": 7663
+    },
+    {
+      "epoch": 0.6287894326619354,
+      "grad_norm": 0.728125741078538,
+      "learning_rate": 3.1997694150639734e-06,
+      "loss": 0.3357,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6288714772121262,
+      "grad_norm": 0.6842723631256924,
+      "learning_rate": 3.1985298828928444e-06,
+      "loss": 0.2853,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6289535217623169,
+      "grad_norm": 0.7293547749459033,
+      "learning_rate": 3.1972904779386416e-06,
+      "loss": 0.3366,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6290355663125077,
+      "grad_norm": 0.6171349751698371,
+      "learning_rate": 3.19605120028889e-06,
+      "loss": 0.2904,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6291176108626985,
+      "grad_norm": 0.6399289102073317,
+      "learning_rate": 3.1948120500311032e-06,
+      "loss": 0.2989,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6291996554128892,
+      "grad_norm": 0.7181250489063933,
+      "learning_rate": 3.1935730272527897e-06,
+      "loss": 0.2985,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6292816999630799,
+      "grad_norm": 0.598959733121871,
+      "learning_rate": 3.192334132041446e-06,
+      "loss": 0.2658,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6293637445132707,
+      "grad_norm": 0.6150304238491153,
+      "learning_rate": 3.1910953644845612e-06,
+      "loss": 0.3406,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6294457890634615,
+      "grad_norm": 0.6830495151726997,
+      "learning_rate": 3.189856724669617e-06,
+      "loss": 0.3034,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6295278336136522,
+      "grad_norm": 0.6142015012868625,
+      "learning_rate": 3.1886182126840827e-06,
+      "loss": 0.3275,
+      "step": 7673
+    },
+    {
+      "epoch": 0.629609878163843,
+      "grad_norm": 0.7146770626128951,
+      "learning_rate": 3.187379828615418e-06,
+      "loss": 0.3036,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6296919227140337,
+      "grad_norm": 0.820344375086415,
+      "learning_rate": 3.1861415725510804e-06,
+      "loss": 0.3227,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6297739672642245,
+      "grad_norm": 0.6395680631808929,
+      "learning_rate": 3.1849034445785087e-06,
+      "loss": 0.3004,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6298560118144152,
+      "grad_norm": 0.6582495958386495,
+      "learning_rate": 3.1836654447851412e-06,
+      "loss": 0.3521,
+      "step": 7677
+    },
+    {
+      "epoch": 0.629938056364606,
+      "grad_norm": 0.664814553175427,
+      "learning_rate": 3.1824275732584e-06,
+      "loss": 0.322,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6300201009147968,
+      "grad_norm": 0.7123673891155939,
+      "learning_rate": 3.181189830085704e-06,
+      "loss": 0.2988,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6301021454649874,
+      "grad_norm": 0.6552665439888942,
+      "learning_rate": 3.1799522153544604e-06,
+      "loss": 0.3034,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6301841900151782,
+      "grad_norm": 0.7884622893745143,
+      "learning_rate": 3.1787147291520675e-06,
+      "loss": 0.3425,
+      "step": 7681
+    },
+    {
+      "epoch": 0.630266234565369,
+      "grad_norm": 0.6886132983028815,
+      "learning_rate": 3.1774773715659152e-06,
+      "loss": 0.3202,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6303482791155598,
+      "grad_norm": 0.8753437990173845,
+      "learning_rate": 3.1762401426833806e-06,
+      "loss": 0.3597,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6304303236657505,
+      "grad_norm": 0.7235016535126797,
+      "learning_rate": 3.175003042591837e-06,
+      "loss": 0.3036,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6305123682159413,
+      "grad_norm": 0.6932519814010062,
+      "learning_rate": 3.173766071378649e-06,
+      "loss": 0.3485,
+      "step": 7685
+    },
+    {
+      "epoch": 0.630594412766132,
+      "grad_norm": 0.6230443462360661,
+      "learning_rate": 3.1725292291311645e-06,
+      "loss": 0.3177,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6306764573163227,
+      "grad_norm": 0.6553346560339011,
+      "learning_rate": 3.1712925159367313e-06,
+      "loss": 0.3016,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6307585018665135,
+      "grad_norm": 0.7329492457946207,
+      "learning_rate": 3.1700559318826823e-06,
+      "loss": 0.3042,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6308405464167043,
+      "grad_norm": 0.770709513148705,
+      "learning_rate": 3.1688194770563424e-06,
+      "loss": 0.3289,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6309225909668951,
+      "grad_norm": 0.647648193195911,
+      "learning_rate": 3.1675831515450305e-06,
+      "loss": 0.3103,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6310046355170857,
+      "grad_norm": 0.6848508568075091,
+      "learning_rate": 3.166346955436051e-06,
+      "loss": 0.3458,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6310866800672765,
+      "grad_norm": 0.659630068089105,
+      "learning_rate": 3.1651108888167033e-06,
+      "loss": 0.2987,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6311687246174673,
+      "grad_norm": 0.6067012017581452,
+      "learning_rate": 3.1638749517742785e-06,
+      "loss": 0.3089,
+      "step": 7693
+    },
+    {
+      "epoch": 0.631250769167658,
+      "grad_norm": 0.6415237323407528,
+      "learning_rate": 3.1626391443960537e-06,
+      "loss": 0.2849,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6313328137178488,
+      "grad_norm": 0.6965538761496733,
+      "learning_rate": 3.1614034667693016e-06,
+      "loss": 0.2893,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6314148582680396,
+      "grad_norm": 0.7015444175306825,
+      "learning_rate": 3.1601679189812806e-06,
+      "loss": 0.3199,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6314969028182303,
+      "grad_norm": 0.5915769295369177,
+      "learning_rate": 3.1589325011192457e-06,
+      "loss": 0.298,
+      "step": 7697
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.7857930812724384,
+      "learning_rate": 3.1576972132704413e-06,
+      "loss": 0.3881,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6316609919186118,
+      "grad_norm": 0.7164104071056457,
+      "learning_rate": 3.1564620555220983e-06,
+      "loss": 0.3189,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6317430364688026,
+      "grad_norm": 0.730883707832101,
+      "learning_rate": 3.155227027961444e-06,
+      "loss": 0.3185,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6318250810189933,
+      "grad_norm": 0.6656107309884524,
+      "learning_rate": 3.153992130675693e-06,
+      "loss": 0.3004,
+      "step": 7701
+    },
+    {
+      "epoch": 0.631907125569184,
+      "grad_norm": 0.704859230151463,
+      "learning_rate": 3.15275736375205e-06,
+      "loss": 0.321,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6319891701193748,
+      "grad_norm": 0.6766159215327779,
+      "learning_rate": 3.1515227272777177e-06,
+      "loss": 0.3049,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6320712146695656,
+      "grad_norm": 0.6955054441538157,
+      "learning_rate": 3.1502882213398776e-06,
+      "loss": 0.2859,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6321532592197563,
+      "grad_norm": 0.6607658627429287,
+      "learning_rate": 3.1490538460257135e-06,
+      "loss": 0.3289,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6322353037699471,
+      "grad_norm": 0.7047971494036114,
+      "learning_rate": 3.1478196014223916e-06,
+      "loss": 0.2979,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6323173483201379,
+      "grad_norm": 0.7931731283742318,
+      "learning_rate": 3.146585487617074e-06,
+      "loss": 0.3168,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6323993928703285,
+      "grad_norm": 0.6838545553218616,
+      "learning_rate": 3.145351504696912e-06,
+      "loss": 0.327,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6324814374205193,
+      "grad_norm": 0.6779178034041179,
+      "learning_rate": 3.144117652749047e-06,
+      "loss": 0.305,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6325634819707101,
+      "grad_norm": 0.6598304439869621,
+      "learning_rate": 3.14288393186061e-06,
+      "loss": 0.3143,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6326455265209009,
+      "grad_norm": 0.7357075906900127,
+      "learning_rate": 3.141650342118728e-06,
+      "loss": 0.2768,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6327275710710916,
+      "grad_norm": 0.6646151843771846,
+      "learning_rate": 3.140416883610512e-06,
+      "loss": 0.2891,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6328096156212824,
+      "grad_norm": 0.6805452976530105,
+      "learning_rate": 3.13918355642307e-06,
+      "loss": 0.2797,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6328916601714731,
+      "grad_norm": 0.6857112013066167,
+      "learning_rate": 3.1379503606434937e-06,
+      "loss": 0.3643,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6329737047216638,
+      "grad_norm": 0.738476705711589,
+      "learning_rate": 3.1367172963588714e-06,
+      "loss": 0.3197,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6330557492718546,
+      "grad_norm": 0.7177204449897122,
+      "learning_rate": 3.1354843636562816e-06,
+      "loss": 0.3529,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6331377938220454,
+      "grad_norm": 0.619961436495956,
+      "learning_rate": 3.1342515626227886e-06,
+      "loss": 0.28,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6332198383722362,
+      "grad_norm": 0.6652415108340368,
+      "learning_rate": 3.133018893345455e-06,
+      "loss": 0.3123,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6333018829224268,
+      "grad_norm": 0.6395013176946704,
+      "learning_rate": 3.131786355911325e-06,
+      "loss": 0.3269,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6333839274726176,
+      "grad_norm": 0.7052746151754038,
+      "learning_rate": 3.1305539504074412e-06,
+      "loss": 0.3483,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6334659720228084,
+      "grad_norm": 0.6820688966662933,
+      "learning_rate": 3.1293216769208355e-06,
+      "loss": 0.3009,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6335480165729991,
+      "grad_norm": 0.6788033104167813,
+      "learning_rate": 3.128089535538527e-06,
+      "loss": 0.2909,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6336300611231899,
+      "grad_norm": 0.7172173955801459,
+      "learning_rate": 3.126857526347529e-06,
+      "loss": 0.3278,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6337121056733807,
+      "grad_norm": 0.6525387882591213,
+      "learning_rate": 3.1256256494348413e-06,
+      "loss": 0.3086,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6337941502235714,
+      "grad_norm": 0.6472449377875968,
+      "learning_rate": 3.1243939048874585e-06,
+      "loss": 0.3071,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6338761947737621,
+      "grad_norm": 0.7278765571012226,
+      "learning_rate": 3.1231622927923667e-06,
+      "loss": 0.3211,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6339582393239529,
+      "grad_norm": 0.6914740744667582,
+      "learning_rate": 3.1219308132365365e-06,
+      "loss": 0.3654,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6340402838741437,
+      "grad_norm": 0.6941591636512837,
+      "learning_rate": 3.1206994663069364e-06,
+      "loss": 0.3042,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6341223284243344,
+      "grad_norm": 0.7532209788762007,
+      "learning_rate": 3.1194682520905207e-06,
+      "loss": 0.3257,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6342043729745251,
+      "grad_norm": 0.6448576594540297,
+      "learning_rate": 3.1182371706742344e-06,
+      "loss": 0.3422,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6342864175247159,
+      "grad_norm": 0.6535292216186267,
+      "learning_rate": 3.1170062221450183e-06,
+      "loss": 0.332,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6343684620749067,
+      "grad_norm": 0.686492618120573,
+      "learning_rate": 3.1157754065897944e-06,
+      "loss": 0.2966,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6344505066250974,
+      "grad_norm": 0.7569979609088898,
+      "learning_rate": 3.1145447240954848e-06,
+      "loss": 0.2853,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6345325511752882,
+      "grad_norm": 0.6035540458827029,
+      "learning_rate": 3.1133141747489993e-06,
+      "loss": 0.2925,
+      "step": 7734
+    },
+    {
+      "epoch": 0.634614595725479,
+      "grad_norm": 0.7413393955543822,
+      "learning_rate": 3.1120837586372343e-06,
+      "loss": 0.3177,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6346966402756697,
+      "grad_norm": 0.6515873071705026,
+      "learning_rate": 3.1108534758470817e-06,
+      "loss": 0.3307,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6347786848258604,
+      "grad_norm": 0.629822621986909,
+      "learning_rate": 3.109623326465421e-06,
+      "loss": 0.3069,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6348607293760512,
+      "grad_norm": 0.7625268411358558,
+      "learning_rate": 3.108393310579123e-06,
+      "loss": 0.3209,
+      "step": 7738
+    },
+    {
+      "epoch": 0.634942773926242,
+      "grad_norm": 0.6747100024333176,
+      "learning_rate": 3.1071634282750523e-06,
+      "loss": 0.3164,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6350248184764327,
+      "grad_norm": 0.8496436443397873,
+      "learning_rate": 3.1059336796400576e-06,
+      "loss": 0.2987,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6351068630266234,
+      "grad_norm": 0.5937131282410163,
+      "learning_rate": 3.104704064760985e-06,
+      "loss": 0.6918,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6351889075768142,
+      "grad_norm": 0.8217064341260074,
+      "learning_rate": 3.1034745837246636e-06,
+      "loss": 0.3256,
+      "step": 7742
+    },
+    {
+      "epoch": 0.635270952127005,
+      "grad_norm": 0.6213375639938397,
+      "learning_rate": 3.1022452366179223e-06,
+      "loss": 0.2879,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6353529966771957,
+      "grad_norm": 0.7154280654282967,
+      "learning_rate": 3.1010160235275743e-06,
+      "loss": 0.2929,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6354350412273865,
+      "grad_norm": 0.66271222534041,
+      "learning_rate": 3.0997869445404214e-06,
+      "loss": 0.3335,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6355170857775773,
+      "grad_norm": 0.7180636539183726,
+      "learning_rate": 3.0985579997432613e-06,
+      "loss": 0.3431,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6355991303277679,
+      "grad_norm": 0.7687384237103397,
+      "learning_rate": 3.097329189222883e-06,
+      "loss": 0.3505,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6356811748779587,
+      "grad_norm": 0.6185108877599653,
+      "learning_rate": 3.0961005130660584e-06,
+      "loss": 0.3323,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6357632194281495,
+      "grad_norm": 0.6843305741287432,
+      "learning_rate": 3.094871971359558e-06,
+      "loss": 0.2964,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6358452639783403,
+      "grad_norm": 0.6878783804828956,
+      "learning_rate": 3.093643564190138e-06,
+      "loss": 0.2877,
+      "step": 7750
+    },
+    {
+      "epoch": 0.635927308528531,
+      "grad_norm": 0.6648808518767417,
+      "learning_rate": 3.092415291644545e-06,
+      "loss": 0.3461,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6360093530787218,
+      "grad_norm": 0.676705409848407,
+      "learning_rate": 3.0911871538095217e-06,
+      "loss": 0.3038,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6360913976289125,
+      "grad_norm": 0.8349058550922067,
+      "learning_rate": 3.089959150771793e-06,
+      "loss": 0.3079,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6361734421791032,
+      "grad_norm": 0.6958876339775097,
+      "learning_rate": 3.088731282618083e-06,
+      "loss": 0.315,
+      "step": 7754
+    },
+    {
+      "epoch": 0.636255486729294,
+      "grad_norm": 0.6853061478553536,
+      "learning_rate": 3.087503549435096e-06,
+      "loss": 0.3498,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6363375312794848,
+      "grad_norm": 0.7244817580127989,
+      "learning_rate": 3.0862759513095364e-06,
+      "loss": 0.3164,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6364195758296756,
+      "grad_norm": 1.24767176170229,
+      "learning_rate": 3.0850484883280963e-06,
+      "loss": 0.2983,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6365016203798662,
+      "grad_norm": 0.7757573417954177,
+      "learning_rate": 3.083821160577453e-06,
+      "loss": 0.309,
+      "step": 7758
+    },
+    {
+      "epoch": 0.636583664930057,
+      "grad_norm": 0.6753151552296048,
+      "learning_rate": 3.082593968144283e-06,
+      "loss": 0.3169,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6366657094802478,
+      "grad_norm": 0.6986615028157744,
+      "learning_rate": 3.0813669111152445e-06,
+      "loss": 0.3258,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6367477540304385,
+      "grad_norm": 0.7586305099647502,
+      "learning_rate": 3.0801399895769917e-06,
+      "loss": 0.3198,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6368297985806293,
+      "grad_norm": 0.6816405953987477,
+      "learning_rate": 3.0789132036161707e-06,
+      "loss": 0.3182,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63691184313082,
+      "grad_norm": 0.7083136586293868,
+      "learning_rate": 3.0776865533194123e-06,
+      "loss": 0.3193,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6369938876810108,
+      "grad_norm": 0.7942095722553699,
+      "learning_rate": 3.076460038773339e-06,
+      "loss": 0.3188,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6370759322312015,
+      "grad_norm": 0.6728994197255231,
+      "learning_rate": 3.0752336600645705e-06,
+      "loss": 0.3151,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6371579767813923,
+      "grad_norm": 0.697624331311182,
+      "learning_rate": 3.0740074172797058e-06,
+      "loss": 0.3092,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6372400213315831,
+      "grad_norm": 0.6604789170946354,
+      "learning_rate": 3.072781310505345e-06,
+      "loss": 0.282,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6373220658817738,
+      "grad_norm": 0.6655589846856483,
+      "learning_rate": 3.0715553398280694e-06,
+      "loss": 0.2862,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6374041104319645,
+      "grad_norm": 0.6688355026495787,
+      "learning_rate": 3.0703295053344597e-06,
+      "loss": 0.2953,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6374861549821553,
+      "grad_norm": 0.7118784222864066,
+      "learning_rate": 3.069103807111079e-06,
+      "loss": 0.3093,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6375681995323461,
+      "grad_norm": 0.6926295731579223,
+      "learning_rate": 3.0678782452444855e-06,
+      "loss": 0.3562,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6376502440825368,
+      "grad_norm": 0.6750179132320691,
+      "learning_rate": 3.066652819821227e-06,
+      "loss": 0.3312,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6377322886327276,
+      "grad_norm": 0.691101915798543,
+      "learning_rate": 3.0654275309278382e-06,
+      "loss": 0.2796,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6378143331829184,
+      "grad_norm": 0.7363899930994864,
+      "learning_rate": 3.064202378650849e-06,
+      "loss": 0.3119,
+      "step": 7774
+    },
+    {
+      "epoch": 0.637896377733109,
+      "grad_norm": 0.6756953799120246,
+      "learning_rate": 3.06297736307678e-06,
+      "loss": 0.3189,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6379784222832998,
+      "grad_norm": 0.67328994396926,
+      "learning_rate": 3.0617524842921354e-06,
+      "loss": 0.296,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6380604668334906,
+      "grad_norm": 0.6513208351645545,
+      "learning_rate": 3.060527742383419e-06,
+      "loss": 0.2807,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6381425113836814,
+      "grad_norm": 0.6876881168599975,
+      "learning_rate": 3.059303137437115e-06,
+      "loss": 0.3171,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6382245559338721,
+      "grad_norm": 0.699789160308309,
+      "learning_rate": 3.0580786695397057e-06,
+      "loss": 0.308,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6383066004840628,
+      "grad_norm": 0.7496674615646777,
+      "learning_rate": 3.0568543387776623e-06,
+      "loss": 0.3489,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6383886450342536,
+      "grad_norm": 0.7099805082503451,
+      "learning_rate": 3.0556301452374415e-06,
+      "loss": 0.2831,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6384706895844443,
+      "grad_norm": 0.7109047582319586,
+      "learning_rate": 3.054406089005496e-06,
+      "loss": 0.3291,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6385527341346351,
+      "grad_norm": 0.6554576447115273,
+      "learning_rate": 3.0531821701682696e-06,
+      "loss": 0.3232,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6386347786848259,
+      "grad_norm": 0.6664733612039719,
+      "learning_rate": 3.0519583888121885e-06,
+      "loss": 0.3047,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6387168232350167,
+      "grad_norm": 0.6347806729101962,
+      "learning_rate": 3.0507347450236775e-06,
+      "loss": 0.3297,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6387988677852073,
+      "grad_norm": 0.6950434945027444,
+      "learning_rate": 3.0495112388891446e-06,
+      "loss": 0.3029,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6388809123353981,
+      "grad_norm": 0.5846720657151786,
+      "learning_rate": 3.0482878704949946e-06,
+      "loss": 0.262,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6389629568855889,
+      "grad_norm": 0.6891519841248843,
+      "learning_rate": 3.047064639927622e-06,
+      "loss": 0.3323,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6390450014357796,
+      "grad_norm": 0.6781163009345883,
+      "learning_rate": 3.045841547273405e-06,
+      "loss": 0.3263,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6391270459859704,
+      "grad_norm": 0.6517064844601173,
+      "learning_rate": 3.04461859261872e-06,
+      "loss": 0.3153,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6392090905361612,
+      "grad_norm": 0.618788669369427,
+      "learning_rate": 3.0433957760499265e-06,
+      "loss": 0.2853,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6392911350863519,
+      "grad_norm": 0.7313311813652529,
+      "learning_rate": 3.0421730976533805e-06,
+      "loss": 0.3541,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6393731796365426,
+      "grad_norm": 0.6790154831189619,
+      "learning_rate": 3.0409505575154254e-06,
+      "loss": 0.3461,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6394552241867334,
+      "grad_norm": 0.7438733473960171,
+      "learning_rate": 3.0397281557223934e-06,
+      "loss": 0.3104,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6395372687369242,
+      "grad_norm": 0.6348597939359599,
+      "learning_rate": 3.0385058923606107e-06,
+      "loss": 0.3002,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6396193132871149,
+      "grad_norm": 0.6868727191320944,
+      "learning_rate": 3.037283767516389e-06,
+      "loss": 0.323,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6397013578373056,
+      "grad_norm": 0.6779647267327232,
+      "learning_rate": 3.0360617812760355e-06,
+      "loss": 0.319,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6397834023874964,
+      "grad_norm": 0.709800112679197,
+      "learning_rate": 3.0348399337258438e-06,
+      "loss": 0.3242,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6398654469376872,
+      "grad_norm": 0.6612482934549028,
+      "learning_rate": 3.0336182249520986e-06,
+      "loss": 0.3224,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6399474914878779,
+      "grad_norm": 0.6069927803760065,
+      "learning_rate": 3.0323966550410737e-06,
+      "loss": 0.3117,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6400295360380687,
+      "grad_norm": 0.679845061045507,
+      "learning_rate": 3.031175224079038e-06,
+      "loss": 0.2906,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6401115805882595,
+      "grad_norm": 0.6665156258387546,
+      "learning_rate": 3.029953932152243e-06,
+      "loss": 0.2786,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6401936251384501,
+      "grad_norm": 0.6579561914797685,
+      "learning_rate": 3.0287327793469382e-06,
+      "loss": 0.3308,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6402756696886409,
+      "grad_norm": 0.6834986973438799,
+      "learning_rate": 3.0275117657493557e-06,
+      "loss": 0.3221,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6403577142388317,
+      "grad_norm": 0.6616660118849739,
+      "learning_rate": 3.0262908914457234e-06,
+      "loss": 0.3041,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6404397587890225,
+      "grad_norm": 0.7628630461521478,
+      "learning_rate": 3.0250701565222596e-06,
+      "loss": 0.2915,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6405218033392132,
+      "grad_norm": 0.6493632150315023,
+      "learning_rate": 3.023849561065165e-06,
+      "loss": 0.322,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6406038478894039,
+      "grad_norm": 0.7488627330299291,
+      "learning_rate": 3.022629105160643e-06,
+      "loss": 0.3434,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6406858924395947,
+      "grad_norm": 0.7153161835055938,
+      "learning_rate": 3.021408788894873e-06,
+      "loss": 0.3328,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6407679369897854,
+      "grad_norm": 0.7138867980168832,
+      "learning_rate": 3.0201886123540366e-06,
+      "loss": 0.3318,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6408499815399762,
+      "grad_norm": 0.7233469574992816,
+      "learning_rate": 3.0189685756243012e-06,
+      "loss": 0.3156,
+      "step": 7811
+    },
+    {
+      "epoch": 0.640932026090167,
+      "grad_norm": 0.6515274295096345,
+      "learning_rate": 3.017748678791821e-06,
+      "loss": 0.2981,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6410140706403578,
+      "grad_norm": 0.6917718847818376,
+      "learning_rate": 3.016528921942744e-06,
+      "loss": 0.3314,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6410961151905484,
+      "grad_norm": 0.6666062574319265,
+      "learning_rate": 3.015309305163209e-06,
+      "loss": 0.2963,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6411781597407392,
+      "grad_norm": 0.8424561173968661,
+      "learning_rate": 3.014089828539341e-06,
+      "loss": 0.3162,
+      "step": 7815
+    },
+    {
+      "epoch": 0.64126020429093,
+      "grad_norm": 0.6944531215458454,
+      "learning_rate": 3.0128704921572607e-06,
+      "loss": 0.3513,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6413422488411207,
+      "grad_norm": 0.7274371724564016,
+      "learning_rate": 3.011651296103071e-06,
+      "loss": 0.2923,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6414242933913115,
+      "grad_norm": 0.6679301792728677,
+      "learning_rate": 3.010432240462874e-06,
+      "loss": 0.3127,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6415063379415022,
+      "grad_norm": 0.7433672337294885,
+      "learning_rate": 3.0092133253227563e-06,
+      "loss": 0.3174,
+      "step": 7819
+    },
+    {
+      "epoch": 0.641588382491693,
+      "grad_norm": 0.9149632976963986,
+      "learning_rate": 3.007994550768793e-06,
+      "loss": 0.3346,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6416704270418837,
+      "grad_norm": 0.7154898192118904,
+      "learning_rate": 3.0067759168870566e-06,
+      "loss": 0.2996,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6417524715920745,
+      "grad_norm": 0.7200449344034872,
+      "learning_rate": 3.0055574237635997e-06,
+      "loss": 0.3198,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6418345161422653,
+      "grad_norm": 0.7371101845080279,
+      "learning_rate": 3.0043390714844746e-06,
+      "loss": 0.3297,
+      "step": 7823
+    },
+    {
+      "epoch": 0.641916560692456,
+      "grad_norm": 0.6837785193002267,
+      "learning_rate": 3.0031208601357188e-06,
+      "loss": 0.3058,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6419986052426467,
+      "grad_norm": 0.6016999241048406,
+      "learning_rate": 3.0019027898033594e-06,
+      "loss": 0.2884,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6420806497928375,
+      "grad_norm": 0.611405779326475,
+      "learning_rate": 3.000684860573414e-06,
+      "loss": 0.2937,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6421626943430283,
+      "grad_norm": 0.5889521605850816,
+      "learning_rate": 2.999467072531892e-06,
+      "loss": 0.3005,
+      "step": 7827
+    },
+    {
+      "epoch": 0.642244738893219,
+      "grad_norm": 0.6577390017864173,
+      "learning_rate": 2.9982494257647903e-06,
+      "loss": 0.3009,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6423267834434098,
+      "grad_norm": 0.7395957496905823,
+      "learning_rate": 2.9970319203581e-06,
+      "loss": 0.3279,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6424088279936006,
+      "grad_norm": 0.7035972326524855,
+      "learning_rate": 2.995814556397796e-06,
+      "loss": 0.3037,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6424908725437912,
+      "grad_norm": 0.7044619277474561,
+      "learning_rate": 2.994597333969848e-06,
+      "loss": 0.3048,
+      "step": 7831
+    },
+    {
+      "epoch": 0.642572917093982,
+      "grad_norm": 0.6342716978508968,
+      "learning_rate": 2.9933802531602153e-06,
+      "loss": 0.341,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6426549616441728,
+      "grad_norm": 0.6157272951290039,
+      "learning_rate": 2.9921633140548446e-06,
+      "loss": 0.2818,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6427370061943636,
+      "grad_norm": 0.6977014090429889,
+      "learning_rate": 2.9909465167396764e-06,
+      "loss": 0.3282,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6428190507445543,
+      "grad_norm": 0.661562347123577,
+      "learning_rate": 2.9897298613006346e-06,
+      "loss": 0.2971,
+      "step": 7835
+    },
+    {
+      "epoch": 0.642901095294745,
+      "grad_norm": 0.6559564512619949,
+      "learning_rate": 2.988513347823641e-06,
+      "loss": 0.3206,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6429831398449358,
+      "grad_norm": 0.6438994911304657,
+      "learning_rate": 2.9872969763946045e-06,
+      "loss": 0.3195,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6430651843951265,
+      "grad_norm": 0.7088444982372757,
+      "learning_rate": 2.9860807470994203e-06,
+      "loss": 0.3718,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6431472289453173,
+      "grad_norm": 0.6304333538233449,
+      "learning_rate": 2.9848646600239805e-06,
+      "loss": 0.2905,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6432292734955081,
+      "grad_norm": 0.6902480685911379,
+      "learning_rate": 2.9836487152541593e-06,
+      "loss": 0.2884,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6433113180456989,
+      "grad_norm": 0.6342560081869444,
+      "learning_rate": 2.9824329128758257e-06,
+      "loss": 0.3262,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6433933625958895,
+      "grad_norm": 0.7409480168897551,
+      "learning_rate": 2.9812172529748395e-06,
+      "loss": 0.3202,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6434754071460803,
+      "grad_norm": 0.6636728576391258,
+      "learning_rate": 2.9800017356370457e-06,
+      "loss": 0.2765,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6435574516962711,
+      "grad_norm": 0.6973791975289033,
+      "learning_rate": 2.9787863609482862e-06,
+      "loss": 0.3094,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6436394962464619,
+      "grad_norm": 0.6687263276299026,
+      "learning_rate": 2.977571128994385e-06,
+      "loss": 0.3233,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6437215407966526,
+      "grad_norm": 0.6771592618999389,
+      "learning_rate": 2.9763560398611618e-06,
+      "loss": 0.331,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6438035853468433,
+      "grad_norm": 0.7503503454025147,
+      "learning_rate": 2.975141093634425e-06,
+      "loss": 0.3536,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6438856298970341,
+      "grad_norm": 0.677093752647987,
+      "learning_rate": 2.973926290399969e-06,
+      "loss": 0.3146,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6439676744472248,
+      "grad_norm": 0.6566253071439456,
+      "learning_rate": 2.9727116302435834e-06,
+      "loss": 0.3506,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6440497189974156,
+      "grad_norm": 0.6356163072797812,
+      "learning_rate": 2.9714971132510473e-06,
+      "loss": 0.3011,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6441317635476064,
+      "grad_norm": 0.658584148918262,
+      "learning_rate": 2.9702827395081236e-06,
+      "loss": 0.296,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6442138080977972,
+      "grad_norm": 0.729503951814414,
+      "learning_rate": 2.9690685091005742e-06,
+      "loss": 0.3408,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6442958526479878,
+      "grad_norm": 0.7007203427978271,
+      "learning_rate": 2.9678544221141432e-06,
+      "loss": 0.3345,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6443778971981786,
+      "grad_norm": 0.666225500184316,
+      "learning_rate": 2.9666404786345664e-06,
+      "loss": 0.2903,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6444599417483694,
+      "grad_norm": 0.6968577743421285,
+      "learning_rate": 2.9654266787475744e-06,
+      "loss": 0.3443,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6445419862985601,
+      "grad_norm": 0.6240277656376868,
+      "learning_rate": 2.9642130225388793e-06,
+      "loss": 0.2961,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6446240308487509,
+      "grad_norm": 0.7328977962303808,
+      "learning_rate": 2.962999510094192e-06,
+      "loss": 0.3084,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6447060753989416,
+      "grad_norm": 0.706366268960322,
+      "learning_rate": 2.9617861414992046e-06,
+      "loss": 0.3164,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6447881199491324,
+      "grad_norm": 0.6691378155724207,
+      "learning_rate": 2.9605729168396063e-06,
+      "loss": 0.3407,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6448701644993231,
+      "grad_norm": 0.6678505134259365,
+      "learning_rate": 2.959359836201072e-06,
+      "loss": 0.2963,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6449522090495139,
+      "grad_norm": 0.6345569345666814,
+      "learning_rate": 2.9581468996692675e-06,
+      "loss": 0.3135,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6450342535997047,
+      "grad_norm": 0.6565842502155249,
+      "learning_rate": 2.9569341073298496e-06,
+      "loss": 0.3053,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6451162981498954,
+      "grad_norm": 0.7232982111333305,
+      "learning_rate": 2.9557214592684603e-06,
+      "loss": 0.3093,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6451983427000861,
+      "grad_norm": 0.6796880659720962,
+      "learning_rate": 2.9545089555707374e-06,
+      "loss": 0.2812,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6452803872502769,
+      "grad_norm": 0.6307854246649054,
+      "learning_rate": 2.9532965963223076e-06,
+      "loss": 0.3,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6453624318004677,
+      "grad_norm": 0.6285542849770379,
+      "learning_rate": 2.9520843816087817e-06,
+      "loss": 0.3121,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6454444763506584,
+      "grad_norm": 0.6785440817723829,
+      "learning_rate": 2.9508723115157677e-06,
+      "loss": 0.3797,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6455265209008492,
+      "grad_norm": 0.6327937214673227,
+      "learning_rate": 2.94966038612886e-06,
+      "loss": 0.2778,
+      "step": 7868
+    },
+    {
+      "epoch": 0.64560856545104,
+      "grad_norm": 0.6718698155949103,
+      "learning_rate": 2.9484486055336397e-06,
+      "loss": 0.3013,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6456906100012306,
+      "grad_norm": 0.6398283718344047,
+      "learning_rate": 2.9472369698156848e-06,
+      "loss": 0.3026,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6457726545514214,
+      "grad_norm": 0.6965766624832384,
+      "learning_rate": 2.946025479060555e-06,
+      "loss": 0.3308,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6458546991016122,
+      "grad_norm": 0.6433599274318208,
+      "learning_rate": 2.944814133353806e-06,
+      "loss": 0.3116,
+      "step": 7872
+    },
+    {
+      "epoch": 0.645936743651803,
+      "grad_norm": 0.7067240700004009,
+      "learning_rate": 2.9436029327809827e-06,
+      "loss": 0.3231,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6460187882019937,
+      "grad_norm": 0.6604995479018346,
+      "learning_rate": 2.942391877427616e-06,
+      "loss": 0.3172,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6461008327521844,
+      "grad_norm": 0.7459003909179422,
+      "learning_rate": 2.94118096737923e-06,
+      "loss": 0.3214,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6461828773023752,
+      "grad_norm": 0.681660223001046,
+      "learning_rate": 2.939970202721334e-06,
+      "loss": 0.3048,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6462649218525659,
+      "grad_norm": 0.6096752889595506,
+      "learning_rate": 2.9387595835394323e-06,
+      "loss": 0.3279,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6463469664027567,
+      "grad_norm": 0.6597694198613703,
+      "learning_rate": 2.93754910991902e-06,
+      "loss": 0.328,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6464290109529475,
+      "grad_norm": 1.047558522241843,
+      "learning_rate": 2.9363387819455735e-06,
+      "loss": 0.2535,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6465110555031383,
+      "grad_norm": 0.7894562138932683,
+      "learning_rate": 2.9351285997045685e-06,
+      "loss": 0.3104,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6465931000533289,
+      "grad_norm": 0.735184208502519,
+      "learning_rate": 2.9339185632814636e-06,
+      "loss": 0.3273,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6466751446035197,
+      "grad_norm": 0.6554458016933145,
+      "learning_rate": 2.932708672761709e-06,
+      "loss": 0.2951,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6467571891537105,
+      "grad_norm": 0.6439430117587528,
+      "learning_rate": 2.9314989282307493e-06,
+      "loss": 0.2818,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6468392337039012,
+      "grad_norm": 0.7122216663510784,
+      "learning_rate": 2.9302893297740097e-06,
+      "loss": 0.3076,
+      "step": 7884
+    },
+    {
+      "epoch": 0.646921278254092,
+      "grad_norm": 0.6549923454334292,
+      "learning_rate": 2.9290798774769124e-06,
+      "loss": 0.3004,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6470033228042827,
+      "grad_norm": 0.6549517521545923,
+      "learning_rate": 2.927870571424869e-06,
+      "loss": 0.3283,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6470853673544735,
+      "grad_norm": 0.6562485794043544,
+      "learning_rate": 2.9266614117032755e-06,
+      "loss": 0.2558,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6471674119046642,
+      "grad_norm": 0.7253378547406737,
+      "learning_rate": 2.9254523983975224e-06,
+      "loss": 0.3139,
+      "step": 7888
+    },
+    {
+      "epoch": 0.647249456454855,
+      "grad_norm": 0.7418256418922389,
+      "learning_rate": 2.9242435315929874e-06,
+      "loss": 0.3126,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6473315010050458,
+      "grad_norm": 0.7778057691949133,
+      "learning_rate": 2.923034811375038e-06,
+      "loss": 0.3037,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6474135455552364,
+      "grad_norm": 0.6774589806639804,
+      "learning_rate": 2.9218262378290364e-06,
+      "loss": 0.3286,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6474955901054272,
+      "grad_norm": 0.725690248453789,
+      "learning_rate": 2.920617811040324e-06,
+      "loss": 0.2966,
+      "step": 7892
+    },
+    {
+      "epoch": 0.647577634655618,
+      "grad_norm": 0.6693992701819582,
+      "learning_rate": 2.9194095310942438e-06,
+      "loss": 0.2718,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6476596792058088,
+      "grad_norm": 0.6873170391610772,
+      "learning_rate": 2.9182013980761183e-06,
+      "loss": 0.3021,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6477417237559995,
+      "grad_norm": 0.7516683444515576,
+      "learning_rate": 2.9169934120712647e-06,
+      "loss": 0.3586,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6478237683061903,
+      "grad_norm": 0.6757927350022866,
+      "learning_rate": 2.9157855731649908e-06,
+      "loss": 0.272,
+      "step": 7896
+    },
+    {
+      "epoch": 0.647905812856381,
+      "grad_norm": 0.6537422001065344,
+      "learning_rate": 2.9145778814425905e-06,
+      "loss": 0.3111,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6479878574065717,
+      "grad_norm": 0.7044753779089296,
+      "learning_rate": 2.913370336989351e-06,
+      "loss": 0.3189,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6480699019567625,
+      "grad_norm": 0.7060317684342294,
+      "learning_rate": 2.912162939890544e-06,
+      "loss": 0.3178,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6481519465069533,
+      "grad_norm": 0.7287152758109418,
+      "learning_rate": 2.9109556902314362e-06,
+      "loss": 0.3055,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6482339910571441,
+      "grad_norm": 0.6478130774693637,
+      "learning_rate": 2.9097485880972825e-06,
+      "loss": 0.3016,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6483160356073348,
+      "grad_norm": 0.7247778474563674,
+      "learning_rate": 2.908541633573324e-06,
+      "loss": 0.3207,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6483980801575255,
+      "grad_norm": 0.6485163481098979,
+      "learning_rate": 2.9073348267447944e-06,
+      "loss": 0.3258,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6484801247077163,
+      "grad_norm": 0.6307363673488708,
+      "learning_rate": 2.90612816769692e-06,
+      "loss": 0.2776,
+      "step": 7904
+    },
+    {
+      "epoch": 0.648562169257907,
+      "grad_norm": 0.7484924495848486,
+      "learning_rate": 2.9049216565149098e-06,
+      "loss": 0.3375,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6486442138080978,
+      "grad_norm": 0.6933302123289165,
+      "learning_rate": 2.9037152932839662e-06,
+      "loss": 0.3316,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6487262583582886,
+      "grad_norm": 0.851775352987921,
+      "learning_rate": 2.902509078089279e-06,
+      "loss": 0.2963,
+      "step": 7907
+    },
+    {
+      "epoch": 0.6488083029084794,
+      "grad_norm": 0.7364113791189463,
+      "learning_rate": 2.9013030110160306e-06,
+      "loss": 0.3573,
+      "step": 7908
+    },
+    {
+      "epoch": 0.64889034745867,
+      "grad_norm": 0.7421336792863152,
+      "learning_rate": 2.900097092149394e-06,
+      "loss": 0.3298,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6489723920088608,
+      "grad_norm": 0.6615825562775031,
+      "learning_rate": 2.8988913215745247e-06,
+      "loss": 0.3246,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6490544365590516,
+      "grad_norm": 0.7023137444472821,
+      "learning_rate": 2.8976856993765766e-06,
+      "loss": 0.3229,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6491364811092423,
+      "grad_norm": 0.7331880495541492,
+      "learning_rate": 2.896480225640685e-06,
+      "loss": 0.3264,
+      "step": 7912
+    },
+    {
+      "epoch": 0.649218525659433,
+      "grad_norm": 0.678827914168673,
+      "learning_rate": 2.89527490045198e-06,
+      "loss": 0.2926,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6493005702096238,
+      "grad_norm": 0.655881840378925,
+      "learning_rate": 2.8940697238955823e-06,
+      "loss": 0.2762,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6493826147598146,
+      "grad_norm": 0.6444815807167679,
+      "learning_rate": 2.8928646960565955e-06,
+      "loss": 0.3001,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6494646593100053,
+      "grad_norm": 0.6411143770305764,
+      "learning_rate": 2.89165981702012e-06,
+      "loss": 0.2932,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6495467038601961,
+      "grad_norm": 0.7127931672417895,
+      "learning_rate": 2.890455086871239e-06,
+      "loss": 0.2803,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6496287484103869,
+      "grad_norm": 0.6605409883897024,
+      "learning_rate": 2.889250505695033e-06,
+      "loss": 0.2894,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6497107929605775,
+      "grad_norm": 0.641909636316265,
+      "learning_rate": 2.8880460735765625e-06,
+      "loss": 0.3059,
+      "step": 7919
+    },
+    {
+      "epoch": 0.6497928375107683,
+      "grad_norm": 0.6724373149057047,
+      "learning_rate": 2.886841790600887e-06,
+      "loss": 0.3103,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6498748820609591,
+      "grad_norm": 0.6568583675555307,
+      "learning_rate": 2.8856376568530475e-06,
+      "loss": 0.3144,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6499569266111499,
+      "grad_norm": 0.6946086189934464,
+      "learning_rate": 2.884433672418082e-06,
+      "loss": 0.3071,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6500389711613406,
+      "grad_norm": 0.6914409892436414,
+      "learning_rate": 2.8832298373810097e-06,
+      "loss": 0.2741,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6501210157115314,
+      "grad_norm": 0.6285796170141753,
+      "learning_rate": 2.8820261518268467e-06,
+      "loss": 0.2988,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6502030602617221,
+      "grad_norm": 0.6772474665238627,
+      "learning_rate": 2.880822615840593e-06,
+      "loss": 0.3656,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6502851048119128,
+      "grad_norm": 0.7547696272140891,
+      "learning_rate": 2.8796192295072407e-06,
+      "loss": 0.3081,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6503671493621036,
+      "grad_norm": 0.6443260390877436,
+      "learning_rate": 2.878415992911774e-06,
+      "loss": 0.3194,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6504491939122944,
+      "grad_norm": 0.6618860837241534,
+      "learning_rate": 2.8772129061391597e-06,
+      "loss": 0.2978,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6505312384624852,
+      "grad_norm": 0.6992622902608143,
+      "learning_rate": 2.8760099692743605e-06,
+      "loss": 0.3161,
+      "step": 7929
+    },
+    {
+      "epoch": 0.6506132830126758,
+      "grad_norm": 0.7030085896258407,
+      "learning_rate": 2.8748071824023237e-06,
+      "loss": 0.3239,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6506953275628666,
+      "grad_norm": 0.6825206759098832,
+      "learning_rate": 2.8736045456079886e-06,
+      "loss": 0.3135,
+      "step": 7931
+    },
+    {
+      "epoch": 0.6507773721130574,
+      "grad_norm": 0.6858175432462865,
+      "learning_rate": 2.872402058976287e-06,
+      "loss": 0.318,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6508594166632481,
+      "grad_norm": 0.7338919818855838,
+      "learning_rate": 2.871199722592133e-06,
+      "loss": 0.3307,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6509414612134389,
+      "grad_norm": 0.6396812443565536,
+      "learning_rate": 2.869997536540435e-06,
+      "loss": 0.3363,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6510235057636297,
+      "grad_norm": 0.6121799695825972,
+      "learning_rate": 2.8687955009060865e-06,
+      "loss": 0.3123,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6511055503138204,
+      "grad_norm": 0.6414855678259733,
+      "learning_rate": 2.867593615773976e-06,
+      "loss": 0.2856,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6511875948640111,
+      "grad_norm": 0.6352504731407603,
+      "learning_rate": 2.86639188122898e-06,
+      "loss": 0.2706,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6512696394142019,
+      "grad_norm": 0.6499272350889234,
+      "learning_rate": 2.8651902973559596e-06,
+      "loss": 0.3444,
+      "step": 7938
+    },
+    {
+      "epoch": 0.6513516839643927,
+      "grad_norm": 0.6564125904487961,
+      "learning_rate": 2.8639888642397707e-06,
+      "loss": 0.3157,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6514337285145834,
+      "grad_norm": 0.6814807608950137,
+      "learning_rate": 2.8627875819652577e-06,
+      "loss": 0.297,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6515157730647742,
+      "grad_norm": 0.8397677299349381,
+      "learning_rate": 2.8615864506172507e-06,
+      "loss": 0.3329,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6515978176149649,
+      "grad_norm": 0.6801672702668141,
+      "learning_rate": 2.860385470280575e-06,
+      "loss": 0.3101,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6516798621651557,
+      "grad_norm": 0.6765277364038924,
+      "learning_rate": 2.8591846410400377e-06,
+      "loss": 0.2953,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6517619067153464,
+      "grad_norm": 0.7350482531954359,
+      "learning_rate": 2.857983962980441e-06,
+      "loss": 0.297,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6518439512655372,
+      "grad_norm": 0.7049472355009058,
+      "learning_rate": 2.856783436186578e-06,
+      "loss": 0.3192,
+      "step": 7945
+    },
+    {
+      "epoch": 0.651925995815728,
+      "grad_norm": 0.6982644096731664,
+      "learning_rate": 2.8555830607432254e-06,
+      "loss": 0.3155,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6520080403659186,
+      "grad_norm": 0.6958384888013761,
+      "learning_rate": 2.85438283673515e-06,
+      "loss": 0.2905,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6520900849161094,
+      "grad_norm": 0.7398985371162089,
+      "learning_rate": 2.853182764247113e-06,
+      "loss": 0.3291,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6521721294663002,
+      "grad_norm": 0.7179507957991076,
+      "learning_rate": 2.8519828433638588e-06,
+      "loss": 0.2979,
+      "step": 7949
+    },
+    {
+      "epoch": 0.652254174016491,
+      "grad_norm": 0.6863901151014533,
+      "learning_rate": 2.850783074170126e-06,
+      "loss": 0.3057,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6523362185666817,
+      "grad_norm": 0.8325304341959844,
+      "learning_rate": 2.849583456750639e-06,
+      "loss": 0.3358,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6524182631168725,
+      "grad_norm": 0.6687921145996351,
+      "learning_rate": 2.8483839911901133e-06,
+      "loss": 0.2829,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6525003076670632,
+      "grad_norm": 0.6373976290196698,
+      "learning_rate": 2.8471846775732527e-06,
+      "loss": 0.3292,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6525823522172539,
+      "grad_norm": 0.7797742406055792,
+      "learning_rate": 2.8459855159847516e-06,
+      "loss": 0.3366,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6526643967674447,
+      "grad_norm": 0.8092259947762614,
+      "learning_rate": 2.844786506509294e-06,
+      "loss": 0.3123,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6527464413176355,
+      "grad_norm": 0.7036453370856735,
+      "learning_rate": 2.8435876492315493e-06,
+      "loss": 0.2888,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6528284858678263,
+      "grad_norm": 0.698869530473606,
+      "learning_rate": 2.8423889442361797e-06,
+      "loss": 0.2994,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6529105304180169,
+      "grad_norm": 0.6699415265517371,
+      "learning_rate": 2.841190391607839e-06,
+      "loss": 0.3229,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6529925749682077,
+      "grad_norm": 0.6516444464727982,
+      "learning_rate": 2.839991991431162e-06,
+      "loss": 0.3456,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6530746195183985,
+      "grad_norm": 0.6406807813078994,
+      "learning_rate": 2.838793743790782e-06,
+      "loss": 0.3328,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6531566640685893,
+      "grad_norm": 0.6863490521841865,
+      "learning_rate": 2.8375956487713163e-06,
+      "loss": 0.3361,
+      "step": 7961
+    },
+    {
+      "epoch": 0.65323870861878,
+      "grad_norm": 0.6987511602099447,
+      "learning_rate": 2.836397706457369e-06,
+      "loss": 0.3028,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6533207531689708,
+      "grad_norm": 0.6928213109037088,
+      "learning_rate": 2.8351999169335422e-06,
+      "loss": 0.2993,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6534027977191615,
+      "grad_norm": 0.8466721831532721,
+      "learning_rate": 2.834002280284417e-06,
+      "loss": 0.274,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6534848422693522,
+      "grad_norm": 0.7094650687726095,
+      "learning_rate": 2.8328047965945725e-06,
+      "loss": 0.3053,
+      "step": 7965
+    },
+    {
+      "epoch": 0.653566886819543,
+      "grad_norm": 0.6711204301341248,
+      "learning_rate": 2.8316074659485693e-06,
+      "loss": 0.2822,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6536489313697338,
+      "grad_norm": 0.682633691559265,
+      "learning_rate": 2.8304102884309637e-06,
+      "loss": 0.3133,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6537309759199246,
+      "grad_norm": 0.6958498081477276,
+      "learning_rate": 2.829213264126299e-06,
+      "loss": 0.313,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6538130204701152,
+      "grad_norm": 0.6507135030912758,
+      "learning_rate": 2.828016393119105e-06,
+      "loss": 0.3232,
+      "step": 7969
+    },
+    {
+      "epoch": 0.653895065020306,
+      "grad_norm": 0.6409495066158053,
+      "learning_rate": 2.8268196754939054e-06,
+      "loss": 0.2933,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6539771095704968,
+      "grad_norm": 0.622404800664217,
+      "learning_rate": 2.8256231113352067e-06,
+      "loss": 0.27,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6540591541206875,
+      "grad_norm": 0.8880670867118448,
+      "learning_rate": 2.8244267007275106e-06,
+      "loss": 0.3337,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6541411986708783,
+      "grad_norm": 0.7027143735024193,
+      "learning_rate": 2.8232304437553072e-06,
+      "loss": 0.3173,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6542232432210691,
+      "grad_norm": 0.5823489809875999,
+      "learning_rate": 2.822034340503073e-06,
+      "loss": 0.3077,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6543052877712598,
+      "grad_norm": 0.6921982287229088,
+      "learning_rate": 2.820838391055273e-06,
+      "loss": 0.3387,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6543873323214505,
+      "grad_norm": 0.7136520763691966,
+      "learning_rate": 2.8196425954963657e-06,
+      "loss": 0.3023,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6544693768716413,
+      "grad_norm": 0.6629077762227502,
+      "learning_rate": 2.8184469539107933e-06,
+      "loss": 0.2964,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6545514214218321,
+      "grad_norm": 0.6983121782900509,
+      "learning_rate": 2.817251466382995e-06,
+      "loss": 0.3259,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6546334659720228,
+      "grad_norm": 0.6128416116034732,
+      "learning_rate": 2.8160561329973886e-06,
+      "loss": 0.3194,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6547155105222136,
+      "grad_norm": 0.7288143720187644,
+      "learning_rate": 2.814860953838389e-06,
+      "loss": 0.3319,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6547975550724043,
+      "grad_norm": 0.6391696619735456,
+      "learning_rate": 2.8136659289904012e-06,
+      "loss": 0.3289,
+      "step": 7981
+    },
+    {
+      "epoch": 0.6548795996225951,
+      "grad_norm": 0.7139327133318862,
+      "learning_rate": 2.81247105853781e-06,
+      "loss": 0.3177,
+      "step": 7982
+    },
+    {
+      "epoch": 0.6549616441727858,
+      "grad_norm": 0.6587489679516033,
+      "learning_rate": 2.8112763425650002e-06,
+      "loss": 0.3302,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6550436887229766,
+      "grad_norm": 0.7120362546668921,
+      "learning_rate": 2.8100817811563376e-06,
+      "loss": 0.3285,
+      "step": 7984
+    },
+    {
+      "epoch": 0.6551257332731674,
+      "grad_norm": 0.6389258287537181,
+      "learning_rate": 2.80888737439618e-06,
+      "loss": 0.2508,
+      "step": 7985
+    },
+    {
+      "epoch": 0.655207777823358,
+      "grad_norm": 0.6811103802674385,
+      "learning_rate": 2.8076931223688787e-06,
+      "loss": 0.3056,
+      "step": 7986
+    },
+    {
+      "epoch": 0.6552898223735488,
+      "grad_norm": 0.6912444692850849,
+      "learning_rate": 2.8064990251587644e-06,
+      "loss": 0.3078,
+      "step": 7987
+    },
+    {
+      "epoch": 0.6553718669237396,
+      "grad_norm": 0.6641887956839314,
+      "learning_rate": 2.8053050828501666e-06,
+      "loss": 0.2717,
+      "step": 7988
+    },
+    {
+      "epoch": 0.6554539114739304,
+      "grad_norm": 0.7065666428448144,
+      "learning_rate": 2.8041112955273974e-06,
+      "loss": 0.3161,
+      "step": 7989
+    },
+    {
+      "epoch": 0.6555359560241211,
+      "grad_norm": 0.5942516017808813,
+      "learning_rate": 2.8029176632747594e-06,
+      "loss": 0.2766,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6556180005743119,
+      "grad_norm": 0.6699655914191199,
+      "learning_rate": 2.801724186176548e-06,
+      "loss": 0.2786,
+      "step": 7991
+    },
+    {
+      "epoch": 0.6557000451245026,
+      "grad_norm": 0.7360553589527367,
+      "learning_rate": 2.800530864317041e-06,
+      "loss": 0.322,
+      "step": 7992
+    },
+    {
+      "epoch": 0.6557820896746933,
+      "grad_norm": 0.7357586256751101,
+      "learning_rate": 2.7993376977805102e-06,
+      "loss": 0.2811,
+      "step": 7993
+    },
+    {
+      "epoch": 0.6558641342248841,
+      "grad_norm": 0.8908884198167593,
+      "learning_rate": 2.7981446866512176e-06,
+      "loss": 0.37,
+      "step": 7994
+    },
+    {
+      "epoch": 0.6559461787750749,
+      "grad_norm": 0.7362893708896101,
+      "learning_rate": 2.796951831013408e-06,
+      "loss": 0.274,
+      "step": 7995
+    },
+    {
+      "epoch": 0.6560282233252657,
+      "grad_norm": 0.6876768496629768,
+      "learning_rate": 2.795759130951322e-06,
+      "loss": 0.3401,
+      "step": 7996
+    },
+    {
+      "epoch": 0.6561102678754563,
+      "grad_norm": 0.7091011055176648,
+      "learning_rate": 2.794566586549183e-06,
+      "loss": 0.3029,
+      "step": 7997
+    },
+    {
+      "epoch": 0.6561923124256471,
+      "grad_norm": 0.6829291494477921,
+      "learning_rate": 2.793374197891209e-06,
+      "loss": 0.3239,
+      "step": 7998
+    },
+    {
+      "epoch": 0.6562743569758379,
+      "grad_norm": 0.6451047204750673,
+      "learning_rate": 2.7921819650616054e-06,
+      "loss": 0.297,
+      "step": 7999
+    },
+    {
+      "epoch": 0.6563564015260286,
+      "grad_norm": 0.6540013213232526,
+      "learning_rate": 2.7909898881445625e-06,
+      "loss": 0.3156,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6564384460762194,
+      "grad_norm": 0.6751906687975074,
+      "learning_rate": 2.7897979672242663e-06,
+      "loss": 0.3168,
+      "step": 8001
+    },
+    {
+      "epoch": 0.6565204906264102,
+      "grad_norm": 0.7238674827712859,
+      "learning_rate": 2.788606202384887e-06,
+      "loss": 0.3037,
+      "step": 8002
+    },
+    {
+      "epoch": 0.6566025351766009,
+      "grad_norm": 0.7093944862321852,
+      "learning_rate": 2.787414593710583e-06,
+      "loss": 0.3233,
+      "step": 8003
+    },
+    {
+      "epoch": 0.6566845797267916,
+      "grad_norm": 0.6740494199761177,
+      "learning_rate": 2.786223141285507e-06,
+      "loss": 0.3176,
+      "step": 8004
+    },
+    {
+      "epoch": 0.6567666242769824,
+      "grad_norm": 0.6353185441552085,
+      "learning_rate": 2.785031845193794e-06,
+      "loss": 0.3659,
+      "step": 8005
+    },
+    {
+      "epoch": 0.6568486688271732,
+      "grad_norm": 0.7453400095871374,
+      "learning_rate": 2.7838407055195743e-06,
+      "loss": 0.3003,
+      "step": 8006
+    },
+    {
+      "epoch": 0.6569307133773639,
+      "grad_norm": 0.7113638454707971,
+      "learning_rate": 2.782649722346965e-06,
+      "loss": 0.2979,
+      "step": 8007
+    },
+    {
+      "epoch": 0.6570127579275546,
+      "grad_norm": 0.6926472896882117,
+      "learning_rate": 2.781458895760068e-06,
+      "loss": 0.3404,
+      "step": 8008
+    },
+    {
+      "epoch": 0.6570948024777454,
+      "grad_norm": 0.6486150653778961,
+      "learning_rate": 2.7802682258429813e-06,
+      "loss": 0.3247,
+      "step": 8009
+    },
+    {
+      "epoch": 0.6571768470279362,
+      "grad_norm": 0.6364815364482068,
+      "learning_rate": 2.779077712679784e-06,
+      "loss": 0.2866,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6572588915781269,
+      "grad_norm": 0.6713799371571609,
+      "learning_rate": 2.7778873563545506e-06,
+      "loss": 0.3416,
+      "step": 8011
+    },
+    {
+      "epoch": 0.6573409361283177,
+      "grad_norm": 0.6786738301281661,
+      "learning_rate": 2.7766971569513445e-06,
+      "loss": 0.3248,
+      "step": 8012
+    },
+    {
+      "epoch": 0.6574229806785085,
+      "grad_norm": 0.6532163379205073,
+      "learning_rate": 2.77550711455421e-06,
+      "loss": 0.308,
+      "step": 8013
+    },
+    {
+      "epoch": 0.6575050252286991,
+      "grad_norm": 0.6327741146465581,
+      "learning_rate": 2.774317229247192e-06,
+      "loss": 0.3195,
+      "step": 8014
+    },
+    {
+      "epoch": 0.6575870697788899,
+      "grad_norm": 0.6610881328573429,
+      "learning_rate": 2.7731275011143133e-06,
+      "loss": 0.3567,
+      "step": 8015
+    },
+    {
+      "epoch": 0.6576691143290807,
+      "grad_norm": 0.6643218407675162,
+      "learning_rate": 2.7719379302395954e-06,
+      "loss": 0.3266,
+      "step": 8016
+    },
+    {
+      "epoch": 0.6577511588792715,
+      "grad_norm": 0.7222147017967254,
+      "learning_rate": 2.7707485167070402e-06,
+      "loss": 0.3086,
+      "step": 8017
+    },
+    {
+      "epoch": 0.6578332034294622,
+      "grad_norm": 0.6528689719801548,
+      "learning_rate": 2.769559260600642e-06,
+      "loss": 0.2753,
+      "step": 8018
+    },
+    {
+      "epoch": 0.657915247979653,
+      "grad_norm": 0.7107512390642204,
+      "learning_rate": 2.768370162004387e-06,
+      "loss": 0.2904,
+      "step": 8019
+    },
+    {
+      "epoch": 0.6579972925298437,
+      "grad_norm": 0.6799752773150332,
+      "learning_rate": 2.7671812210022433e-06,
+      "loss": 0.2886,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6580793370800344,
+      "grad_norm": 0.7773458414795124,
+      "learning_rate": 2.7659924376781742e-06,
+      "loss": 0.3416,
+      "step": 8021
+    },
+    {
+      "epoch": 0.6581613816302252,
+      "grad_norm": 0.6843294505511738,
+      "learning_rate": 2.764803812116133e-06,
+      "loss": 0.309,
+      "step": 8022
+    },
+    {
+      "epoch": 0.658243426180416,
+      "grad_norm": 0.7143500021761804,
+      "learning_rate": 2.763615344400053e-06,
+      "loss": 0.2935,
+      "step": 8023
+    },
+    {
+      "epoch": 0.6583254707306068,
+      "grad_norm": 0.7239624082823232,
+      "learning_rate": 2.7624270346138643e-06,
+      "loss": 0.3247,
+      "step": 8024
+    },
+    {
+      "epoch": 0.6584075152807974,
+      "grad_norm": 0.6059016395654344,
+      "learning_rate": 2.761238882841486e-06,
+      "loss": 0.2665,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6584895598309882,
+      "grad_norm": 0.7159362278412904,
+      "learning_rate": 2.760050889166818e-06,
+      "loss": 0.3197,
+      "step": 8026
+    },
+    {
+      "epoch": 0.658571604381179,
+      "grad_norm": 0.6687780116060105,
+      "learning_rate": 2.7588630536737606e-06,
+      "loss": 0.3167,
+      "step": 8027
+    },
+    {
+      "epoch": 0.6586536489313697,
+      "grad_norm": 0.6872441246826941,
+      "learning_rate": 2.7576753764461906e-06,
+      "loss": 0.3274,
+      "step": 8028
+    },
+    {
+      "epoch": 0.6587356934815605,
+      "grad_norm": 0.678388006870216,
+      "learning_rate": 2.7564878575679852e-06,
+      "loss": 0.2803,
+      "step": 8029
+    },
+    {
+      "epoch": 0.6588177380317513,
+      "grad_norm": 0.6670497135557879,
+      "learning_rate": 2.7553004971230037e-06,
+      "loss": 0.3009,
+      "step": 8030
+    },
+    {
+      "epoch": 0.658899782581942,
+      "grad_norm": 0.6603155764043044,
+      "learning_rate": 2.7541132951950917e-06,
+      "loss": 0.3108,
+      "step": 8031
+    },
+    {
+      "epoch": 0.6589818271321327,
+      "grad_norm": 0.7870375908753914,
+      "learning_rate": 2.7529262518680927e-06,
+      "loss": 0.3361,
+      "step": 8032
+    },
+    {
+      "epoch": 0.6590638716823235,
+      "grad_norm": 0.6620973308066272,
+      "learning_rate": 2.7517393672258296e-06,
+      "loss": 0.2932,
+      "step": 8033
+    },
+    {
+      "epoch": 0.6591459162325143,
+      "grad_norm": 0.7327253398425994,
+      "learning_rate": 2.7505526413521205e-06,
+      "loss": 0.3436,
+      "step": 8034
+    },
+    {
+      "epoch": 0.659227960782705,
+      "grad_norm": 0.7331772807342967,
+      "learning_rate": 2.7493660743307717e-06,
+      "loss": 0.328,
+      "step": 8035
+    },
+    {
+      "epoch": 0.6593100053328957,
+      "grad_norm": 0.6653254454261436,
+      "learning_rate": 2.748179666245573e-06,
+      "loss": 0.2983,
+      "step": 8036
+    },
+    {
+      "epoch": 0.6593920498830865,
+      "grad_norm": 0.7663685244853932,
+      "learning_rate": 2.7469934171803103e-06,
+      "loss": 0.3031,
+      "step": 8037
+    },
+    {
+      "epoch": 0.6594740944332773,
+      "grad_norm": 0.6694959074221706,
+      "learning_rate": 2.745807327218751e-06,
+      "loss": 0.3748,
+      "step": 8038
+    },
+    {
+      "epoch": 0.659556138983468,
+      "grad_norm": 0.6426470712764738,
+      "learning_rate": 2.744621396444656e-06,
+      "loss": 0.2847,
+      "step": 8039
+    },
+    {
+      "epoch": 0.6596381835336588,
+      "grad_norm": 0.6915273324327135,
+      "learning_rate": 2.743435624941777e-06,
+      "loss": 0.2902,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6597202280838496,
+      "grad_norm": 0.677242015945906,
+      "learning_rate": 2.742250012793847e-06,
+      "loss": 0.3233,
+      "step": 8041
+    },
+    {
+      "epoch": 0.6598022726340402,
+      "grad_norm": 0.6561433765768877,
+      "learning_rate": 2.741064560084595e-06,
+      "loss": 0.2813,
+      "step": 8042
+    },
+    {
+      "epoch": 0.659884317184231,
+      "grad_norm": 0.6646038766648574,
+      "learning_rate": 2.7398792668977327e-06,
+      "loss": 0.3167,
+      "step": 8043
+    },
+    {
+      "epoch": 0.6599663617344218,
+      "grad_norm": 0.6384142758739458,
+      "learning_rate": 2.7386941333169677e-06,
+      "loss": 0.3367,
+      "step": 8044
+    },
+    {
+      "epoch": 0.6600484062846126,
+      "grad_norm": 0.641680118670966,
+      "learning_rate": 2.7375091594259893e-06,
+      "loss": 0.2859,
+      "step": 8045
+    },
+    {
+      "epoch": 0.6601304508348033,
+      "grad_norm": 0.6510985085255568,
+      "learning_rate": 2.7363243453084776e-06,
+      "loss": 0.3098,
+      "step": 8046
+    },
+    {
+      "epoch": 0.660212495384994,
+      "grad_norm": 0.7038564196400859,
+      "learning_rate": 2.735139691048104e-06,
+      "loss": 0.3223,
+      "step": 8047
+    },
+    {
+      "epoch": 0.6602945399351848,
+      "grad_norm": 0.666672380930649,
+      "learning_rate": 2.7339551967285285e-06,
+      "loss": 0.3299,
+      "step": 8048
+    },
+    {
+      "epoch": 0.6603765844853755,
+      "grad_norm": 0.6422938782666655,
+      "learning_rate": 2.7327708624333936e-06,
+      "loss": 0.2857,
+      "step": 8049
+    },
+    {
+      "epoch": 0.6604586290355663,
+      "grad_norm": 0.6591156552037318,
+      "learning_rate": 2.7315866882463403e-06,
+      "loss": 0.3167,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6605406735857571,
+      "grad_norm": 0.6285949551324369,
+      "learning_rate": 2.7304026742509893e-06,
+      "loss": 0.3038,
+      "step": 8051
+    },
+    {
+      "epoch": 0.6606227181359479,
+      "grad_norm": 0.691286203714858,
+      "learning_rate": 2.7292188205309543e-06,
+      "loss": 0.2915,
+      "step": 8052
+    },
+    {
+      "epoch": 0.6607047626861385,
+      "grad_norm": 0.7118300678313455,
+      "learning_rate": 2.7280351271698403e-06,
+      "loss": 0.3058,
+      "step": 8053
+    },
+    {
+      "epoch": 0.6607868072363293,
+      "grad_norm": 0.6980213872462306,
+      "learning_rate": 2.726851594251233e-06,
+      "loss": 0.2844,
+      "step": 8054
+    },
+    {
+      "epoch": 0.6608688517865201,
+      "grad_norm": 0.688073249989975,
+      "learning_rate": 2.725668221858716e-06,
+      "loss": 0.325,
+      "step": 8055
+    },
+    {
+      "epoch": 0.6609508963367108,
+      "grad_norm": 0.6672051519276098,
+      "learning_rate": 2.7244850100758535e-06,
+      "loss": 0.3347,
+      "step": 8056
+    },
+    {
+      "epoch": 0.6610329408869016,
+      "grad_norm": 0.6684148634700576,
+      "learning_rate": 2.7233019589862055e-06,
+      "loss": 0.2969,
+      "step": 8057
+    },
+    {
+      "epoch": 0.6611149854370924,
+      "grad_norm": 0.6890590419296401,
+      "learning_rate": 2.7221190686733157e-06,
+      "loss": 0.2878,
+      "step": 8058
+    },
+    {
+      "epoch": 0.6611970299872831,
+      "grad_norm": 0.8440488719906014,
+      "learning_rate": 2.720936339220715e-06,
+      "loss": 0.328,
+      "step": 8059
+    },
+    {
+      "epoch": 0.6612790745374738,
+      "grad_norm": 0.6718076373570552,
+      "learning_rate": 2.7197537707119287e-06,
+      "loss": 0.3511,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6613611190876646,
+      "grad_norm": 0.6307489280416285,
+      "learning_rate": 2.71857136323047e-06,
+      "loss": 0.2959,
+      "step": 8061
+    },
+    {
+      "epoch": 0.6614431636378554,
+      "grad_norm": 0.6632069809781064,
+      "learning_rate": 2.7173891168598332e-06,
+      "loss": 0.2798,
+      "step": 8062
+    },
+    {
+      "epoch": 0.661525208188046,
+      "grad_norm": 0.7118375188753067,
+      "learning_rate": 2.7162070316835125e-06,
+      "loss": 0.3409,
+      "step": 8063
+    },
+    {
+      "epoch": 0.6616072527382368,
+      "grad_norm": 0.738878922979594,
+      "learning_rate": 2.7150251077849797e-06,
+      "loss": 0.3121,
+      "step": 8064
+    },
+    {
+      "epoch": 0.6616892972884276,
+      "grad_norm": 0.7342211310168653,
+      "learning_rate": 2.713843345247703e-06,
+      "loss": 0.362,
+      "step": 8065
+    },
+    {
+      "epoch": 0.6617713418386184,
+      "grad_norm": 0.7126169143099944,
+      "learning_rate": 2.7126617441551383e-06,
+      "loss": 0.2911,
+      "step": 8066
+    },
+    {
+      "epoch": 0.6618533863888091,
+      "grad_norm": 0.7788238433970327,
+      "learning_rate": 2.7114803045907245e-06,
+      "loss": 0.3224,
+      "step": 8067
+    },
+    {
+      "epoch": 0.6619354309389999,
+      "grad_norm": 0.69345805511402,
+      "learning_rate": 2.710299026637896e-06,
+      "loss": 0.3011,
+      "step": 8068
+    },
+    {
+      "epoch": 0.6620174754891907,
+      "grad_norm": 0.993037419205405,
+      "learning_rate": 2.70911791038007e-06,
+      "loss": 0.3064,
+      "step": 8069
+    },
+    {
+      "epoch": 0.6620995200393814,
+      "grad_norm": 0.727668006912345,
+      "learning_rate": 2.707936955900659e-06,
+      "loss": 0.3159,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6621815645895721,
+      "grad_norm": 0.6852722117056528,
+      "learning_rate": 2.706756163283055e-06,
+      "loss": 0.291,
+      "step": 8071
+    },
+    {
+      "epoch": 0.6622636091397629,
+      "grad_norm": 0.7162188748850601,
+      "learning_rate": 2.705575532610649e-06,
+      "loss": 0.278,
+      "step": 8072
+    },
+    {
+      "epoch": 0.6623456536899537,
+      "grad_norm": 0.6809288264383753,
+      "learning_rate": 2.7043950639668126e-06,
+      "loss": 0.3506,
+      "step": 8073
+    },
+    {
+      "epoch": 0.6624276982401444,
+      "grad_norm": 0.7429844150807912,
+      "learning_rate": 2.703214757434906e-06,
+      "loss": 0.3179,
+      "step": 8074
+    },
+    {
+      "epoch": 0.6625097427903351,
+      "grad_norm": 0.6839896720523092,
+      "learning_rate": 2.702034613098283e-06,
+      "loss": 0.3333,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6625917873405259,
+      "grad_norm": 0.6741242199754676,
+      "learning_rate": 2.7008546310402867e-06,
+      "loss": 0.2937,
+      "step": 8076
+    },
+    {
+      "epoch": 0.6626738318907167,
+      "grad_norm": 0.6544130251368766,
+      "learning_rate": 2.6996748113442397e-06,
+      "loss": 0.2924,
+      "step": 8077
+    },
+    {
+      "epoch": 0.6627558764409074,
+      "grad_norm": 0.6803490741008016,
+      "learning_rate": 2.698495154093462e-06,
+      "loss": 0.2917,
+      "step": 8078
+    },
+    {
+      "epoch": 0.6628379209910982,
+      "grad_norm": 0.6952327704071696,
+      "learning_rate": 2.6973156593712615e-06,
+      "loss": 0.3186,
+      "step": 8079
+    },
+    {
+      "epoch": 0.662919965541289,
+      "grad_norm": 0.6880336821437825,
+      "learning_rate": 2.6961363272609265e-06,
+      "loss": 0.2919,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6630020100914796,
+      "grad_norm": 0.6841302263248249,
+      "learning_rate": 2.694957157845745e-06,
+      "loss": 0.2745,
+      "step": 8081
+    },
+    {
+      "epoch": 0.6630840546416704,
+      "grad_norm": 0.690849634849999,
+      "learning_rate": 2.6937781512089838e-06,
+      "loss": 0.3249,
+      "step": 8082
+    },
+    {
+      "epoch": 0.6631660991918612,
+      "grad_norm": 0.6929730180761545,
+      "learning_rate": 2.692599307433904e-06,
+      "loss": 0.2799,
+      "step": 8083
+    },
+    {
+      "epoch": 0.663248143742052,
+      "grad_norm": 0.7435123354123031,
+      "learning_rate": 2.691420626603757e-06,
+      "loss": 0.2873,
+      "step": 8084
+    },
+    {
+      "epoch": 0.6633301882922427,
+      "grad_norm": 0.6681196669253539,
+      "learning_rate": 2.690242108801775e-06,
+      "loss": 0.3156,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6634122328424334,
+      "grad_norm": 0.7421668639915527,
+      "learning_rate": 2.689063754111185e-06,
+      "loss": 0.2843,
+      "step": 8086
+    },
+    {
+      "epoch": 0.6634942773926242,
+      "grad_norm": 0.6365789694547688,
+      "learning_rate": 2.687885562615198e-06,
+      "loss": 0.2767,
+      "step": 8087
+    },
+    {
+      "epoch": 0.6635763219428149,
+      "grad_norm": 0.6645573967068177,
+      "learning_rate": 2.686707534397018e-06,
+      "loss": 0.3533,
+      "step": 8088
+    },
+    {
+      "epoch": 0.6636583664930057,
+      "grad_norm": 0.719401861918022,
+      "learning_rate": 2.685529669539837e-06,
+      "loss": 0.3514,
+      "step": 8089
+    },
+    {
+      "epoch": 0.6637404110431965,
+      "grad_norm": 0.6827517104173965,
+      "learning_rate": 2.6843519681268304e-06,
+      "loss": 0.3072,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6638224555933873,
+      "grad_norm": 0.65661880335195,
+      "learning_rate": 2.6831744302411695e-06,
+      "loss": 0.3218,
+      "step": 8091
+    },
+    {
+      "epoch": 0.6639045001435779,
+      "grad_norm": 0.6856756902771224,
+      "learning_rate": 2.681997055966006e-06,
+      "loss": 0.2907,
+      "step": 8092
+    },
+    {
+      "epoch": 0.6639865446937687,
+      "grad_norm": 0.6480347547467532,
+      "learning_rate": 2.6808198453844857e-06,
+      "loss": 0.3129,
+      "step": 8093
+    },
+    {
+      "epoch": 0.6640685892439595,
+      "grad_norm": 0.7273533159580978,
+      "learning_rate": 2.6796427985797447e-06,
+      "loss": 0.3535,
+      "step": 8094
+    },
+    {
+      "epoch": 0.6641506337941502,
+      "grad_norm": 0.8172638076533322,
+      "learning_rate": 2.678465915634899e-06,
+      "loss": 0.2896,
+      "step": 8095
+    },
+    {
+      "epoch": 0.664232678344341,
+      "grad_norm": 0.6362608093654536,
+      "learning_rate": 2.6772891966330605e-06,
+      "loss": 0.274,
+      "step": 8096
+    },
+    {
+      "epoch": 0.6643147228945318,
+      "grad_norm": 0.6555354696046777,
+      "learning_rate": 2.676112641657329e-06,
+      "loss": 0.3302,
+      "step": 8097
+    },
+    {
+      "epoch": 0.6643967674447225,
+      "grad_norm": 0.7080644017992007,
+      "learning_rate": 2.6749362507907895e-06,
+      "loss": 0.3054,
+      "step": 8098
+    },
+    {
+      "epoch": 0.6644788119949132,
+      "grad_norm": 0.7173359113883363,
+      "learning_rate": 2.673760024116514e-06,
+      "loss": 0.2839,
+      "step": 8099
+    },
+    {
+      "epoch": 0.664560856545104,
+      "grad_norm": 0.7291809372131586,
+      "learning_rate": 2.672583961717571e-06,
+      "loss": 0.3083,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6646429010952948,
+      "grad_norm": 0.5962906250061993,
+      "learning_rate": 2.671408063677007e-06,
+      "loss": 0.3062,
+      "step": 8101
+    },
+    {
+      "epoch": 0.6647249456454855,
+      "grad_norm": 0.5737824387099492,
+      "learning_rate": 2.670232330077866e-06,
+      "loss": 0.2805,
+      "step": 8102
+    },
+    {
+      "epoch": 0.6648069901956762,
+      "grad_norm": 0.7161265067709608,
+      "learning_rate": 2.669056761003173e-06,
+      "loss": 0.3254,
+      "step": 8103
+    },
+    {
+      "epoch": 0.664889034745867,
+      "grad_norm": 0.7163349336592878,
+      "learning_rate": 2.6678813565359487e-06,
+      "loss": 0.3061,
+      "step": 8104
+    },
+    {
+      "epoch": 0.6649710792960578,
+      "grad_norm": 0.6891721573356774,
+      "learning_rate": 2.666706116759194e-06,
+      "loss": 0.3364,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6650531238462485,
+      "grad_norm": 0.7091457857780976,
+      "learning_rate": 2.665531041755905e-06,
+      "loss": 0.3168,
+      "step": 8106
+    },
+    {
+      "epoch": 0.6651351683964393,
+      "grad_norm": 0.7166241820556125,
+      "learning_rate": 2.664356131609065e-06,
+      "loss": 0.2898,
+      "step": 8107
+    },
+    {
+      "epoch": 0.66521721294663,
+      "grad_norm": 0.6610305757150101,
+      "learning_rate": 2.66318138640164e-06,
+      "loss": 0.285,
+      "step": 8108
+    },
+    {
+      "epoch": 0.6652992574968207,
+      "grad_norm": 0.7043420100687419,
+      "learning_rate": 2.662006806216594e-06,
+      "loss": 0.2955,
+      "step": 8109
+    },
+    {
+      "epoch": 0.6653813020470115,
+      "grad_norm": 0.692499879363533,
+      "learning_rate": 2.660832391136868e-06,
+      "loss": 0.3145,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6654633465972023,
+      "grad_norm": 0.6932960613536591,
+      "learning_rate": 2.6596581412454013e-06,
+      "loss": 0.2916,
+      "step": 8111
+    },
+    {
+      "epoch": 0.6655453911473931,
+      "grad_norm": 0.6837357807637505,
+      "learning_rate": 2.658484056625118e-06,
+      "loss": 0.2893,
+      "step": 8112
+    },
+    {
+      "epoch": 0.6656274356975838,
+      "grad_norm": 0.7119390310635075,
+      "learning_rate": 2.6573101373589296e-06,
+      "loss": 0.3025,
+      "step": 8113
+    },
+    {
+      "epoch": 0.6657094802477745,
+      "grad_norm": 0.8727110300646701,
+      "learning_rate": 2.6561363835297325e-06,
+      "loss": 0.3393,
+      "step": 8114
+    },
+    {
+      "epoch": 0.6657915247979653,
+      "grad_norm": 0.6731182348400044,
+      "learning_rate": 2.6549627952204202e-06,
+      "loss": 0.3295,
+      "step": 8115
+    },
+    {
+      "epoch": 0.665873569348156,
+      "grad_norm": 0.6174508455107487,
+      "learning_rate": 2.6537893725138664e-06,
+      "loss": 0.298,
+      "step": 8116
+    },
+    {
+      "epoch": 0.6659556138983468,
+      "grad_norm": 0.6857399510701339,
+      "learning_rate": 2.6526161154929396e-06,
+      "loss": 0.2744,
+      "step": 8117
+    },
+    {
+      "epoch": 0.6660376584485376,
+      "grad_norm": 0.6875080184029904,
+      "learning_rate": 2.651443024240489e-06,
+      "loss": 0.339,
+      "step": 8118
+    },
+    {
+      "epoch": 0.6661197029987284,
+      "grad_norm": 0.6508821962647262,
+      "learning_rate": 2.6502700988393593e-06,
+      "loss": 0.2997,
+      "step": 8119
+    },
+    {
+      "epoch": 0.666201747548919,
+      "grad_norm": 0.6884870745907083,
+      "learning_rate": 2.649097339372382e-06,
+      "loss": 0.3013,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6662837920991098,
+      "grad_norm": 0.6236580612074822,
+      "learning_rate": 2.647924745922371e-06,
+      "loss": 0.3206,
+      "step": 8121
+    },
+    {
+      "epoch": 0.6663658366493006,
+      "grad_norm": 0.7228484502580459,
+      "learning_rate": 2.6467523185721376e-06,
+      "loss": 0.3167,
+      "step": 8122
+    },
+    {
+      "epoch": 0.6664478811994913,
+      "grad_norm": 0.639749980162052,
+      "learning_rate": 2.645580057404473e-06,
+      "loss": 0.2964,
+      "step": 8123
+    },
+    {
+      "epoch": 0.6665299257496821,
+      "grad_norm": 0.6677747266995528,
+      "learning_rate": 2.6444079625021625e-06,
+      "loss": 0.3123,
+      "step": 8124
+    },
+    {
+      "epoch": 0.6666119702998728,
+      "grad_norm": 0.6450231882670014,
+      "learning_rate": 2.6432360339479786e-06,
+      "loss": 0.3718,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6666940148500636,
+      "grad_norm": 0.5965623434140845,
+      "learning_rate": 2.642064271824679e-06,
+      "loss": 0.3112,
+      "step": 8126
+    },
+    {
+      "epoch": 0.6667760594002543,
+      "grad_norm": 0.7153228657172461,
+      "learning_rate": 2.640892676215011e-06,
+      "loss": 0.3085,
+      "step": 8127
+    },
+    {
+      "epoch": 0.6668581039504451,
+      "grad_norm": 0.837820386523976,
+      "learning_rate": 2.639721247201714e-06,
+      "loss": 0.3469,
+      "step": 8128
+    },
+    {
+      "epoch": 0.6669401485006359,
+      "grad_norm": 0.6615683063390081,
+      "learning_rate": 2.6385499848675086e-06,
+      "loss": 0.3192,
+      "step": 8129
+    },
+    {
+      "epoch": 0.6670221930508266,
+      "grad_norm": 0.6727767448893986,
+      "learning_rate": 2.637378889295111e-06,
+      "loss": 0.3394,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6671042376010173,
+      "grad_norm": 0.5941384548812534,
+      "learning_rate": 2.6362079605672185e-06,
+      "loss": 0.2839,
+      "step": 8131
+    },
+    {
+      "epoch": 0.6671862821512081,
+      "grad_norm": 0.7181424550604965,
+      "learning_rate": 2.635037198766522e-06,
+      "loss": 0.3001,
+      "step": 8132
+    },
+    {
+      "epoch": 0.6672683267013989,
+      "grad_norm": 0.6692103986756982,
+      "learning_rate": 2.633866603975701e-06,
+      "loss": 0.3265,
+      "step": 8133
+    },
+    {
+      "epoch": 0.6673503712515896,
+      "grad_norm": 0.6806407853006319,
+      "learning_rate": 2.6326961762774167e-06,
+      "loss": 0.311,
+      "step": 8134
+    },
+    {
+      "epoch": 0.6674324158017804,
+      "grad_norm": 0.7249538073675089,
+      "learning_rate": 2.631525915754327e-06,
+      "loss": 0.363,
+      "step": 8135
+    },
+    {
+      "epoch": 0.6675144603519712,
+      "grad_norm": 0.6999870269766665,
+      "learning_rate": 2.6303558224890704e-06,
+      "loss": 0.3169,
+      "step": 8136
+    },
+    {
+      "epoch": 0.6675965049021618,
+      "grad_norm": 0.7131143692699918,
+      "learning_rate": 2.629185896564278e-06,
+      "loss": 0.3301,
+      "step": 8137
+    },
+    {
+      "epoch": 0.6676785494523526,
+      "grad_norm": 0.6797668714849202,
+      "learning_rate": 2.6280161380625707e-06,
+      "loss": 0.3105,
+      "step": 8138
+    },
+    {
+      "epoch": 0.6677605940025434,
+      "grad_norm": 0.7323832287144254,
+      "learning_rate": 2.626846547066551e-06,
+      "loss": 0.2923,
+      "step": 8139
+    },
+    {
+      "epoch": 0.6678426385527342,
+      "grad_norm": 0.6723776864239038,
+      "learning_rate": 2.6256771236588174e-06,
+      "loss": 0.3439,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6679246831029249,
+      "grad_norm": 0.6410461778248537,
+      "learning_rate": 2.6245078679219503e-06,
+      "loss": 0.3353,
+      "step": 8141
+    },
+    {
+      "epoch": 0.6680067276531156,
+      "grad_norm": 0.6347722817095229,
+      "learning_rate": 2.6233387799385194e-06,
+      "loss": 0.3173,
+      "step": 8142
+    },
+    {
+      "epoch": 0.6680887722033064,
+      "grad_norm": 1.4644522030650402,
+      "learning_rate": 2.622169859791088e-06,
+      "loss": 0.3042,
+      "step": 8143
+    },
+    {
+      "epoch": 0.6681708167534971,
+      "grad_norm": 0.6440476632582055,
+      "learning_rate": 2.6210011075621977e-06,
+      "loss": 0.2929,
+      "step": 8144
+    },
+    {
+      "epoch": 0.6682528613036879,
+      "grad_norm": 0.6524827453287786,
+      "learning_rate": 2.61983252333439e-06,
+      "loss": 0.2444,
+      "step": 8145
+    },
+    {
+      "epoch": 0.6683349058538787,
+      "grad_norm": 0.6379372349024204,
+      "learning_rate": 2.618664107190183e-06,
+      "loss": 0.3231,
+      "step": 8146
+    },
+    {
+      "epoch": 0.6684169504040695,
+      "grad_norm": 0.6432079303479796,
+      "learning_rate": 2.617495859212091e-06,
+      "loss": 0.3257,
+      "step": 8147
+    },
+    {
+      "epoch": 0.6684989949542601,
+      "grad_norm": 0.6480653517664382,
+      "learning_rate": 2.616327779482615e-06,
+      "loss": 0.3313,
+      "step": 8148
+    },
+    {
+      "epoch": 0.6685810395044509,
+      "grad_norm": 0.6662150036632537,
+      "learning_rate": 2.6151598680842405e-06,
+      "loss": 0.3132,
+      "step": 8149
+    },
+    {
+      "epoch": 0.6686630840546417,
+      "grad_norm": 0.6639229354896464,
+      "learning_rate": 2.6139921250994438e-06,
+      "loss": 0.3563,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6687451286048324,
+      "grad_norm": 0.6562889436331024,
+      "learning_rate": 2.6128245506106927e-06,
+      "loss": 0.2982,
+      "step": 8151
+    },
+    {
+      "epoch": 0.6688271731550232,
+      "grad_norm": 0.74076463421231,
+      "learning_rate": 2.6116571447004335e-06,
+      "loss": 0.3378,
+      "step": 8152
+    },
+    {
+      "epoch": 0.6689092177052139,
+      "grad_norm": 0.6782853852071222,
+      "learning_rate": 2.6104899074511125e-06,
+      "loss": 0.2845,
+      "step": 8153
+    },
+    {
+      "epoch": 0.6689912622554047,
+      "grad_norm": 0.6513678402579531,
+      "learning_rate": 2.6093228389451547e-06,
+      "loss": 0.3147,
+      "step": 8154
+    },
+    {
+      "epoch": 0.6690733068055954,
+      "grad_norm": 0.5925118463421216,
+      "learning_rate": 2.6081559392649754e-06,
+      "loss": 0.2905,
+      "step": 8155
+    },
+    {
+      "epoch": 0.6691553513557862,
+      "grad_norm": 0.5985026531199384,
+      "learning_rate": 2.6069892084929825e-06,
+      "loss": 0.2791,
+      "step": 8156
+    },
+    {
+      "epoch": 0.669237395905977,
+      "grad_norm": 0.663936716330973,
+      "learning_rate": 2.6058226467115647e-06,
+      "loss": 0.2912,
+      "step": 8157
+    },
+    {
+      "epoch": 0.6693194404561676,
+      "grad_norm": 0.7304583404517128,
+      "learning_rate": 2.604656254003108e-06,
+      "loss": 0.3382,
+      "step": 8158
+    },
+    {
+      "epoch": 0.6694014850063584,
+      "grad_norm": 0.9091436404186509,
+      "learning_rate": 2.6034900304499754e-06,
+      "loss": 0.2794,
+      "step": 8159
+    },
+    {
+      "epoch": 0.6694835295565492,
+      "grad_norm": 0.7208340150312141,
+      "learning_rate": 2.6023239761345277e-06,
+      "loss": 0.3047,
+      "step": 8160
+    },
+    {
+      "epoch": 0.66956557410674,
+      "grad_norm": 0.6796498641753095,
+      "learning_rate": 2.6011580911391094e-06,
+      "loss": 0.317,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6696476186569307,
+      "grad_norm": 0.6915819096663688,
+      "learning_rate": 2.599992375546051e-06,
+      "loss": 0.3281,
+      "step": 8162
+    },
+    {
+      "epoch": 0.6697296632071215,
+      "grad_norm": 0.6747360055943659,
+      "learning_rate": 2.5988268294376774e-06,
+      "loss": 0.3392,
+      "step": 8163
+    },
+    {
+      "epoch": 0.6698117077573122,
+      "grad_norm": 0.7102160457581349,
+      "learning_rate": 2.597661452896293e-06,
+      "loss": 0.2997,
+      "step": 8164
+    },
+    {
+      "epoch": 0.6698937523075029,
+      "grad_norm": 0.6779858898004372,
+      "learning_rate": 2.5964962460041977e-06,
+      "loss": 0.3091,
+      "step": 8165
+    },
+    {
+      "epoch": 0.6699757968576937,
+      "grad_norm": 0.6742282714108595,
+      "learning_rate": 2.595331208843678e-06,
+      "loss": 0.3005,
+      "step": 8166
+    },
+    {
+      "epoch": 0.6700578414078845,
+      "grad_norm": 0.668907631913854,
+      "learning_rate": 2.594166341497002e-06,
+      "loss": 0.3125,
+      "step": 8167
+    },
+    {
+      "epoch": 0.6701398859580753,
+      "grad_norm": 0.6787387641859602,
+      "learning_rate": 2.593001644046437e-06,
+      "loss": 0.2862,
+      "step": 8168
+    },
+    {
+      "epoch": 0.670221930508266,
+      "grad_norm": 0.6780458308364831,
+      "learning_rate": 2.5918371165742272e-06,
+      "loss": 0.288,
+      "step": 8169
+    },
+    {
+      "epoch": 0.6703039750584567,
+      "grad_norm": 0.6590150781016358,
+      "learning_rate": 2.59067275916261e-06,
+      "loss": 0.3268,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6703860196086475,
+      "grad_norm": 0.6654381309264227,
+      "learning_rate": 2.589508571893814e-06,
+      "loss": 0.3257,
+      "step": 8171
+    },
+    {
+      "epoch": 0.6704680641588382,
+      "grad_norm": 0.6708101594709509,
+      "learning_rate": 2.5883445548500474e-06,
+      "loss": 0.3038,
+      "step": 8172
+    },
+    {
+      "epoch": 0.670550108709029,
+      "grad_norm": 0.7123476594689931,
+      "learning_rate": 2.587180708113514e-06,
+      "loss": 0.2787,
+      "step": 8173
+    },
+    {
+      "epoch": 0.6706321532592198,
+      "grad_norm": 0.6151728797771059,
+      "learning_rate": 2.5860170317664047e-06,
+      "loss": 0.2833,
+      "step": 8174
+    },
+    {
+      "epoch": 0.6707141978094106,
+      "grad_norm": 0.7211218917174692,
+      "learning_rate": 2.5848535258908918e-06,
+      "loss": 0.327,
+      "step": 8175
+    },
+    {
+      "epoch": 0.6707962423596012,
+      "grad_norm": 0.7359921239262016,
+      "learning_rate": 2.583690190569145e-06,
+      "loss": 0.3239,
+      "step": 8176
+    },
+    {
+      "epoch": 0.670878286909792,
+      "grad_norm": 0.7271387913716978,
+      "learning_rate": 2.5825270258833126e-06,
+      "loss": 0.3216,
+      "step": 8177
+    },
+    {
+      "epoch": 0.6709603314599828,
+      "grad_norm": 0.8147457991253308,
+      "learning_rate": 2.5813640319155376e-06,
+      "loss": 0.2763,
+      "step": 8178
+    },
+    {
+      "epoch": 0.6710423760101735,
+      "grad_norm": 0.5741600279950695,
+      "learning_rate": 2.5802012087479513e-06,
+      "loss": 0.2495,
+      "step": 8179
+    },
+    {
+      "epoch": 0.6711244205603643,
+      "grad_norm": 0.6848158198473198,
+      "learning_rate": 2.5790385564626654e-06,
+      "loss": 0.3317,
+      "step": 8180
+    },
+    {
+      "epoch": 0.671206465110555,
+      "grad_norm": 0.6551076964876796,
+      "learning_rate": 2.5778760751417896e-06,
+      "loss": 0.3014,
+      "step": 8181
+    },
+    {
+      "epoch": 0.6712885096607458,
+      "grad_norm": 0.7117376843015984,
+      "learning_rate": 2.5767137648674136e-06,
+      "loss": 0.303,
+      "step": 8182
+    },
+    {
+      "epoch": 0.6713705542109365,
+      "grad_norm": 0.8192909981509287,
+      "learning_rate": 2.5755516257216163e-06,
+      "loss": 0.2972,
+      "step": 8183
+    },
+    {
+      "epoch": 0.6714525987611273,
+      "grad_norm": 0.6741981710184765,
+      "learning_rate": 2.57438965778647e-06,
+      "loss": 0.3085,
+      "step": 8184
+    },
+    {
+      "epoch": 0.6715346433113181,
+      "grad_norm": 0.8436251740587363,
+      "learning_rate": 2.5732278611440265e-06,
+      "loss": 0.3015,
+      "step": 8185
+    },
+    {
+      "epoch": 0.6716166878615089,
+      "grad_norm": 0.635303833837762,
+      "learning_rate": 2.572066235876334e-06,
+      "loss": 0.3022,
+      "step": 8186
+    },
+    {
+      "epoch": 0.6716987324116995,
+      "grad_norm": 0.6415965000562331,
+      "learning_rate": 2.5709047820654236e-06,
+      "loss": 0.2989,
+      "step": 8187
+    },
+    {
+      "epoch": 0.6717807769618903,
+      "grad_norm": 0.7510122053697051,
+      "learning_rate": 2.5697434997933133e-06,
+      "loss": 0.3178,
+      "step": 8188
+    },
+    {
+      "epoch": 0.6718628215120811,
+      "grad_norm": 0.7341908339375776,
+      "learning_rate": 2.5685823891420146e-06,
+      "loss": 0.3213,
+      "step": 8189
+    },
+    {
+      "epoch": 0.6719448660622718,
+      "grad_norm": 0.8088667559578648,
+      "learning_rate": 2.5674214501935198e-06,
+      "loss": 0.3004,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6720269106124626,
+      "grad_norm": 0.6572969113792508,
+      "learning_rate": 2.566260683029813e-06,
+      "loss": 0.3028,
+      "step": 8191
+    },
+    {
+      "epoch": 0.6721089551626533,
+      "grad_norm": 0.6721398141587366,
+      "learning_rate": 2.5651000877328692e-06,
+      "loss": 0.3234,
+      "step": 8192
+    },
+    {
+      "epoch": 0.6721909997128441,
+      "grad_norm": 0.705007418271208,
+      "learning_rate": 2.5639396643846436e-06,
+      "loss": 0.3231,
+      "step": 8193
+    },
+    {
+      "epoch": 0.6722730442630348,
+      "grad_norm": 0.6736570582587528,
+      "learning_rate": 2.5627794130670863e-06,
+      "loss": 0.2888,
+      "step": 8194
+    },
+    {
+      "epoch": 0.6723550888132256,
+      "grad_norm": 0.7109293122804601,
+      "learning_rate": 2.56161933386213e-06,
+      "loss": 0.3121,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6724371333634164,
+      "grad_norm": 0.6639248152571967,
+      "learning_rate": 2.5604594268516995e-06,
+      "loss": 0.283,
+      "step": 8196
+    },
+    {
+      "epoch": 0.672519177913607,
+      "grad_norm": 1.785582809651136,
+      "learning_rate": 2.559299692117706e-06,
+      "loss": 0.3138,
+      "step": 8197
+    },
+    {
+      "epoch": 0.6726012224637978,
+      "grad_norm": 0.7236312110453947,
+      "learning_rate": 2.5581401297420444e-06,
+      "loss": 0.3103,
+      "step": 8198
+    },
+    {
+      "epoch": 0.6726832670139886,
+      "grad_norm": 0.715287777226239,
+      "learning_rate": 2.5569807398066036e-06,
+      "loss": 0.2955,
+      "step": 8199
+    },
+    {
+      "epoch": 0.6727653115641794,
+      "grad_norm": 0.6514545843591932,
+      "learning_rate": 2.55582152239326e-06,
+      "loss": 0.2984,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6728473561143701,
+      "grad_norm": 0.6473567894613749,
+      "learning_rate": 2.5546624775838725e-06,
+      "loss": 0.2914,
+      "step": 8201
+    },
+    {
+      "epoch": 0.6729294006645609,
+      "grad_norm": 0.8032997522720652,
+      "learning_rate": 2.5535036054602934e-06,
+      "loss": 0.3097,
+      "step": 8202
+    },
+    {
+      "epoch": 0.6730114452147516,
+      "grad_norm": 0.6235762450323863,
+      "learning_rate": 2.552344906104357e-06,
+      "loss": 0.2859,
+      "step": 8203
+    },
+    {
+      "epoch": 0.6730934897649423,
+      "grad_norm": 0.6449381710364568,
+      "learning_rate": 2.551186379597891e-06,
+      "loss": 0.3,
+      "step": 8204
+    },
+    {
+      "epoch": 0.6731755343151331,
+      "grad_norm": 0.6697603070893805,
+      "learning_rate": 2.5500280260227106e-06,
+      "loss": 0.3433,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6732575788653239,
+      "grad_norm": 0.8431303902489062,
+      "learning_rate": 2.5488698454606128e-06,
+      "loss": 0.3476,
+      "step": 8206
+    },
+    {
+      "epoch": 0.6733396234155147,
+      "grad_norm": 0.6174061456902428,
+      "learning_rate": 2.5477118379933903e-06,
+      "loss": 0.3316,
+      "step": 8207
+    },
+    {
+      "epoch": 0.6734216679657054,
+      "grad_norm": 0.8034581054843295,
+      "learning_rate": 2.546554003702816e-06,
+      "loss": 0.307,
+      "step": 8208
+    },
+    {
+      "epoch": 0.6735037125158961,
+      "grad_norm": 0.6859813351186242,
+      "learning_rate": 2.5453963426706585e-06,
+      "loss": 0.2902,
+      "step": 8209
+    },
+    {
+      "epoch": 0.6735857570660869,
+      "grad_norm": 1.0068267857392428,
+      "learning_rate": 2.5442388549786668e-06,
+      "loss": 0.3024,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6736678016162776,
+      "grad_norm": 0.6491049922210534,
+      "learning_rate": 2.5430815407085805e-06,
+      "loss": 0.3785,
+      "step": 8211
+    },
+    {
+      "epoch": 0.6737498461664684,
+      "grad_norm": 0.6263892646848009,
+      "learning_rate": 2.5419243999421305e-06,
+      "loss": 0.3333,
+      "step": 8212
+    },
+    {
+      "epoch": 0.6738318907166592,
+      "grad_norm": 0.6745267031320171,
+      "learning_rate": 2.5407674327610277e-06,
+      "loss": 0.2915,
+      "step": 8213
+    },
+    {
+      "epoch": 0.67391393526685,
+      "grad_norm": 0.6850156417416312,
+      "learning_rate": 2.539610639246978e-06,
+      "loss": 0.3128,
+      "step": 8214
+    },
+    {
+      "epoch": 0.6739959798170406,
+      "grad_norm": 0.6136271227679502,
+      "learning_rate": 2.5384540194816743e-06,
+      "loss": 0.2878,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6740780243672314,
+      "grad_norm": 0.7313377768000188,
+      "learning_rate": 2.5372975735467905e-06,
+      "loss": 0.3051,
+      "step": 8216
+    },
+    {
+      "epoch": 0.6741600689174222,
+      "grad_norm": 0.6609957878579671,
+      "learning_rate": 2.5361413015239957e-06,
+      "loss": 0.3068,
+      "step": 8217
+    },
+    {
+      "epoch": 0.6742421134676129,
+      "grad_norm": 0.661754539804971,
+      "learning_rate": 2.534985203494945e-06,
+      "loss": 0.3046,
+      "step": 8218
+    },
+    {
+      "epoch": 0.6743241580178037,
+      "grad_norm": 0.6672268619070145,
+      "learning_rate": 2.533829279541278e-06,
+      "loss": 0.3158,
+      "step": 8219
+    },
+    {
+      "epoch": 0.6744062025679944,
+      "grad_norm": 0.7162871584891964,
+      "learning_rate": 2.5326735297446257e-06,
+      "loss": 0.3112,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6744882471181852,
+      "grad_norm": 0.660995699968139,
+      "learning_rate": 2.5315179541866043e-06,
+      "loss": 0.3193,
+      "step": 8221
+    },
+    {
+      "epoch": 0.6745702916683759,
+      "grad_norm": 0.7019934165488235,
+      "learning_rate": 2.5303625529488197e-06,
+      "loss": 0.3027,
+      "step": 8222
+    },
+    {
+      "epoch": 0.6746523362185667,
+      "grad_norm": 0.675697052054468,
+      "learning_rate": 2.5292073261128624e-06,
+      "loss": 0.2886,
+      "step": 8223
+    },
+    {
+      "epoch": 0.6747343807687575,
+      "grad_norm": 0.7165303628992089,
+      "learning_rate": 2.5280522737603162e-06,
+      "loss": 0.3056,
+      "step": 8224
+    },
+    {
+      "epoch": 0.6748164253189481,
+      "grad_norm": 0.6725406526331622,
+      "learning_rate": 2.5268973959727473e-06,
+      "loss": 0.2957,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6748984698691389,
+      "grad_norm": 0.7227347210732469,
+      "learning_rate": 2.525742692831708e-06,
+      "loss": 0.299,
+      "step": 8226
+    },
+    {
+      "epoch": 0.6749805144193297,
+      "grad_norm": 0.6558634734980996,
+      "learning_rate": 2.5245881644187453e-06,
+      "loss": 0.2889,
+      "step": 8227
+    },
+    {
+      "epoch": 0.6750625589695205,
+      "grad_norm": 0.7054439296638866,
+      "learning_rate": 2.523433810815391e-06,
+      "loss": 0.3245,
+      "step": 8228
+    },
+    {
+      "epoch": 0.6751446035197112,
+      "grad_norm": 0.66711674945728,
+      "learning_rate": 2.52227963210316e-06,
+      "loss": 0.2927,
+      "step": 8229
+    },
+    {
+      "epoch": 0.675226648069902,
+      "grad_norm": 0.7161721965383536,
+      "learning_rate": 2.5211256283635627e-06,
+      "loss": 0.3437,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6753086926200927,
+      "grad_norm": 0.8293361245471647,
+      "learning_rate": 2.519971799678089e-06,
+      "loss": 0.32,
+      "step": 8231
+    },
+    {
+      "epoch": 0.6753907371702834,
+      "grad_norm": 0.7001075490823231,
+      "learning_rate": 2.5188181461282225e-06,
+      "loss": 0.3292,
+      "step": 8232
+    },
+    {
+      "epoch": 0.6754727817204742,
+      "grad_norm": 0.6901065922471759,
+      "learning_rate": 2.517664667795434e-06,
+      "loss": 0.2902,
+      "step": 8233
+    },
+    {
+      "epoch": 0.675554826270665,
+      "grad_norm": 0.6896446589808451,
+      "learning_rate": 2.516511364761176e-06,
+      "loss": 0.2951,
+      "step": 8234
+    },
+    {
+      "epoch": 0.6756368708208558,
+      "grad_norm": 0.6462385772246437,
+      "learning_rate": 2.5153582371068965e-06,
+      "loss": 0.3208,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6757189153710464,
+      "grad_norm": 0.638789993251734,
+      "learning_rate": 2.5142052849140274e-06,
+      "loss": 0.315,
+      "step": 8236
+    },
+    {
+      "epoch": 0.6758009599212372,
+      "grad_norm": 0.7292259846351746,
+      "learning_rate": 2.5130525082639878e-06,
+      "loss": 0.287,
+      "step": 8237
+    },
+    {
+      "epoch": 0.675883004471428,
+      "grad_norm": 0.7912604541990902,
+      "learning_rate": 2.5118999072381846e-06,
+      "loss": 0.3028,
+      "step": 8238
+    },
+    {
+      "epoch": 0.6759650490216187,
+      "grad_norm": 0.6707389489770805,
+      "learning_rate": 2.5107474819180098e-06,
+      "loss": 0.306,
+      "step": 8239
+    },
+    {
+      "epoch": 0.6760470935718095,
+      "grad_norm": 0.7062855747666732,
+      "learning_rate": 2.5095952323848496e-06,
+      "loss": 0.3425,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6761291381220003,
+      "grad_norm": 0.7212468414076025,
+      "learning_rate": 2.508443158720074e-06,
+      "loss": 0.3226,
+      "step": 8241
+    },
+    {
+      "epoch": 0.676211182672191,
+      "grad_norm": 0.7197356678351364,
+      "learning_rate": 2.5072912610050382e-06,
+      "loss": 0.3321,
+      "step": 8242
+    },
+    {
+      "epoch": 0.6762932272223817,
+      "grad_norm": 0.6512645794465217,
+      "learning_rate": 2.5061395393210907e-06,
+      "loss": 0.3264,
+      "step": 8243
+    },
+    {
+      "epoch": 0.6763752717725725,
+      "grad_norm": 0.7122606400667696,
+      "learning_rate": 2.5049879937495603e-06,
+      "loss": 0.3274,
+      "step": 8244
+    },
+    {
+      "epoch": 0.6764573163227633,
+      "grad_norm": 0.6417451242013785,
+      "learning_rate": 2.503836624371768e-06,
+      "loss": 0.31,
+      "step": 8245
+    },
+    {
+      "epoch": 0.676539360872954,
+      "grad_norm": 0.6739431558362823,
+      "learning_rate": 2.502685431269026e-06,
+      "loss": 0.3123,
+      "step": 8246
+    },
+    {
+      "epoch": 0.6766214054231448,
+      "grad_norm": 0.6415563845534185,
+      "learning_rate": 2.501534414522625e-06,
+      "loss": 0.3104,
+      "step": 8247
+    },
+    {
+      "epoch": 0.6767034499733355,
+      "grad_norm": 0.6993839504681887,
+      "learning_rate": 2.5003835742138506e-06,
+      "loss": 0.2937,
+      "step": 8248
+    },
+    {
+      "epoch": 0.6767854945235263,
+      "grad_norm": 0.6860711794763299,
+      "learning_rate": 2.499232910423971e-06,
+      "loss": 0.2786,
+      "step": 8249
+    },
+    {
+      "epoch": 0.676867539073717,
+      "grad_norm": 0.7113135738025319,
+      "learning_rate": 2.498082423234247e-06,
+      "loss": 0.3131,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6769495836239078,
+      "grad_norm": 0.7683167025474629,
+      "learning_rate": 2.496932112725921e-06,
+      "loss": 0.341,
+      "step": 8251
+    },
+    {
+      "epoch": 0.6770316281740986,
+      "grad_norm": 0.6897940571162184,
+      "learning_rate": 2.495781978980229e-06,
+      "loss": 0.3062,
+      "step": 8252
+    },
+    {
+      "epoch": 0.6771136727242892,
+      "grad_norm": 0.7448288131618244,
+      "learning_rate": 2.4946320220783892e-06,
+      "loss": 0.3074,
+      "step": 8253
+    },
+    {
+      "epoch": 0.67719571727448,
+      "grad_norm": 0.6609809283583811,
+      "learning_rate": 2.4934822421016123e-06,
+      "loss": 0.329,
+      "step": 8254
+    },
+    {
+      "epoch": 0.6772777618246708,
+      "grad_norm": 0.7144552945103181,
+      "learning_rate": 2.4923326391310904e-06,
+      "loss": 0.2834,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6773598063748616,
+      "grad_norm": 0.6395652210267223,
+      "learning_rate": 2.4911832132480108e-06,
+      "loss": 0.3153,
+      "step": 8256
+    },
+    {
+      "epoch": 0.6774418509250523,
+      "grad_norm": 0.665143188424082,
+      "learning_rate": 2.4900339645335404e-06,
+      "loss": 0.3203,
+      "step": 8257
+    },
+    {
+      "epoch": 0.677523895475243,
+      "grad_norm": 0.7379739600239136,
+      "learning_rate": 2.488884893068838e-06,
+      "loss": 0.2978,
+      "step": 8258
+    },
+    {
+      "epoch": 0.6776059400254338,
+      "grad_norm": 0.5987309129474184,
+      "learning_rate": 2.4877359989350526e-06,
+      "loss": 0.2903,
+      "step": 8259
+    },
+    {
+      "epoch": 0.6776879845756245,
+      "grad_norm": 0.7430598032867024,
+      "learning_rate": 2.4865872822133124e-06,
+      "loss": 0.3018,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6777700291258153,
+      "grad_norm": 0.6865215727581864,
+      "learning_rate": 2.485438742984743e-06,
+      "loss": 0.3034,
+      "step": 8261
+    },
+    {
+      "epoch": 0.6778520736760061,
+      "grad_norm": 0.680879291394232,
+      "learning_rate": 2.4842903813304464e-06,
+      "loss": 0.3559,
+      "step": 8262
+    },
+    {
+      "epoch": 0.6779341182261969,
+      "grad_norm": 0.6773931085708029,
+      "learning_rate": 2.4831421973315216e-06,
+      "loss": 0.296,
+      "step": 8263
+    },
+    {
+      "epoch": 0.6780161627763875,
+      "grad_norm": 0.8358115910114993,
+      "learning_rate": 2.481994191069054e-06,
+      "loss": 0.366,
+      "step": 8264
+    },
+    {
+      "epoch": 0.6780982073265783,
+      "grad_norm": 0.6394689583586362,
+      "learning_rate": 2.4808463626241103e-06,
+      "loss": 0.3225,
+      "step": 8265
+    },
+    {
+      "epoch": 0.6781802518767691,
+      "grad_norm": 0.6676622303499525,
+      "learning_rate": 2.4796987120777493e-06,
+      "loss": 0.3243,
+      "step": 8266
+    },
+    {
+      "epoch": 0.6782622964269598,
+      "grad_norm": 0.649058000072509,
+      "learning_rate": 2.4785512395110147e-06,
+      "loss": 0.2885,
+      "step": 8267
+    },
+    {
+      "epoch": 0.6783443409771506,
+      "grad_norm": 0.6662107373386024,
+      "learning_rate": 2.47740394500494e-06,
+      "loss": 0.2994,
+      "step": 8268
+    },
+    {
+      "epoch": 0.6784263855273414,
+      "grad_norm": 0.6457700428301126,
+      "learning_rate": 2.4762568286405475e-06,
+      "loss": 0.3244,
+      "step": 8269
+    },
+    {
+      "epoch": 0.6785084300775321,
+      "grad_norm": 0.6756173437320291,
+      "learning_rate": 2.4751098904988413e-06,
+      "loss": 0.2841,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6785904746277228,
+      "grad_norm": 0.6296142258765764,
+      "learning_rate": 2.4739631306608185e-06,
+      "loss": 0.3177,
+      "step": 8271
+    },
+    {
+      "epoch": 0.6786725191779136,
+      "grad_norm": 0.6099434432353263,
+      "learning_rate": 2.472816549207463e-06,
+      "loss": 0.2949,
+      "step": 8272
+    },
+    {
+      "epoch": 0.6787545637281044,
+      "grad_norm": 0.8500895683048043,
+      "learning_rate": 2.47167014621974e-06,
+      "loss": 0.3717,
+      "step": 8273
+    },
+    {
+      "epoch": 0.6788366082782951,
+      "grad_norm": 0.6416406401933127,
+      "learning_rate": 2.4705239217786115e-06,
+      "loss": 0.2876,
+      "step": 8274
+    },
+    {
+      "epoch": 0.6789186528284858,
+      "grad_norm": 0.7369892876207508,
+      "learning_rate": 2.4693778759650178e-06,
+      "loss": 0.2747,
+      "step": 8275
+    },
+    {
+      "epoch": 0.6790006973786766,
+      "grad_norm": 0.7712906588034872,
+      "learning_rate": 2.4682320088598926e-06,
+      "loss": 0.2993,
+      "step": 8276
+    },
+    {
+      "epoch": 0.6790827419288674,
+      "grad_norm": 0.689659605421786,
+      "learning_rate": 2.4670863205441577e-06,
+      "loss": 0.2835,
+      "step": 8277
+    },
+    {
+      "epoch": 0.6791647864790581,
+      "grad_norm": 0.6933565038496964,
+      "learning_rate": 2.465940811098717e-06,
+      "loss": 0.3425,
+      "step": 8278
+    },
+    {
+      "epoch": 0.6792468310292489,
+      "grad_norm": 0.7709572247643542,
+      "learning_rate": 2.4647954806044633e-06,
+      "loss": 0.2979,
+      "step": 8279
+    },
+    {
+      "epoch": 0.6793288755794397,
+      "grad_norm": 0.6841993816813264,
+      "learning_rate": 2.4636503291422816e-06,
+      "loss": 0.3126,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6794109201296303,
+      "grad_norm": 0.6881894555542035,
+      "learning_rate": 2.4625053567930364e-06,
+      "loss": 0.3288,
+      "step": 8281
+    },
+    {
+      "epoch": 0.6794929646798211,
+      "grad_norm": 0.7192565503368499,
+      "learning_rate": 2.461360563637588e-06,
+      "loss": 0.2854,
+      "step": 8282
+    },
+    {
+      "epoch": 0.6795750092300119,
+      "grad_norm": 0.6658175153647469,
+      "learning_rate": 2.4602159497567764e-06,
+      "loss": 0.2769,
+      "step": 8283
+    },
+    {
+      "epoch": 0.6796570537802027,
+      "grad_norm": 0.7703067144280286,
+      "learning_rate": 2.4590715152314352e-06,
+      "loss": 0.2703,
+      "step": 8284
+    },
+    {
+      "epoch": 0.6797390983303934,
+      "grad_norm": 0.7249328450059153,
+      "learning_rate": 2.4579272601423803e-06,
+      "loss": 0.3166,
+      "step": 8285
+    },
+    {
+      "epoch": 0.6798211428805842,
+      "grad_norm": 0.7724776628245068,
+      "learning_rate": 2.4567831845704177e-06,
+      "loss": 0.2821,
+      "step": 8286
+    },
+    {
+      "epoch": 0.6799031874307749,
+      "grad_norm": 0.6687423778271976,
+      "learning_rate": 2.4556392885963426e-06,
+      "loss": 0.3484,
+      "step": 8287
+    },
+    {
+      "epoch": 0.6799852319809656,
+      "grad_norm": 0.6896025349966145,
+      "learning_rate": 2.4544955723009313e-06,
+      "loss": 0.2938,
+      "step": 8288
+    },
+    {
+      "epoch": 0.6800672765311564,
+      "grad_norm": 0.801192311137692,
+      "learning_rate": 2.4533520357649528e-06,
+      "loss": 0.3174,
+      "step": 8289
+    },
+    {
+      "epoch": 0.6801493210813472,
+      "grad_norm": 0.661192701486842,
+      "learning_rate": 2.452208679069164e-06,
+      "loss": 0.2843,
+      "step": 8290
+    },
+    {
+      "epoch": 0.680231365631538,
+      "grad_norm": 0.6773581270464557,
+      "learning_rate": 2.451065502294303e-06,
+      "loss": 0.301,
+      "step": 8291
+    },
+    {
+      "epoch": 0.6803134101817286,
+      "grad_norm": 0.7270677730668718,
+      "learning_rate": 2.449922505521103e-06,
+      "loss": 0.2986,
+      "step": 8292
+    },
+    {
+      "epoch": 0.6803954547319194,
+      "grad_norm": 0.6585108581971382,
+      "learning_rate": 2.4487796888302784e-06,
+      "loss": 0.3237,
+      "step": 8293
+    },
+    {
+      "epoch": 0.6804774992821102,
+      "grad_norm": 0.6715504412531779,
+      "learning_rate": 2.4476370523025316e-06,
+      "loss": 0.2963,
+      "step": 8294
+    },
+    {
+      "epoch": 0.680559543832301,
+      "grad_norm": 0.6940328781968662,
+      "learning_rate": 2.4464945960185577e-06,
+      "loss": 0.2947,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6806415883824917,
+      "grad_norm": 0.6813385539949587,
+      "learning_rate": 2.4453523200590307e-06,
+      "loss": 0.3164,
+      "step": 8296
+    },
+    {
+      "epoch": 0.6807236329326825,
+      "grad_norm": 0.6987162707073588,
+      "learning_rate": 2.444210224504621e-06,
+      "loss": 0.2829,
+      "step": 8297
+    },
+    {
+      "epoch": 0.6808056774828732,
+      "grad_norm": 0.7519715415455768,
+      "learning_rate": 2.4430683094359764e-06,
+      "loss": 0.3186,
+      "step": 8298
+    },
+    {
+      "epoch": 0.6808877220330639,
+      "grad_norm": 0.6260598682698914,
+      "learning_rate": 2.4419265749337394e-06,
+      "loss": 0.2832,
+      "step": 8299
+    },
+    {
+      "epoch": 0.6809697665832547,
+      "grad_norm": 0.6999118022502496,
+      "learning_rate": 2.4407850210785405e-06,
+      "loss": 0.3361,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6810518111334455,
+      "grad_norm": 0.6431132331080818,
+      "learning_rate": 2.43964364795099e-06,
+      "loss": 0.3165,
+      "step": 8301
+    },
+    {
+      "epoch": 0.6811338556836363,
+      "grad_norm": 0.6697292638389045,
+      "learning_rate": 2.4385024556316928e-06,
+      "loss": 0.2946,
+      "step": 8302
+    },
+    {
+      "epoch": 0.6812159002338269,
+      "grad_norm": 0.7075546407755652,
+      "learning_rate": 2.4373614442012356e-06,
+      "loss": 0.3252,
+      "step": 8303
+    },
+    {
+      "epoch": 0.6812979447840177,
+      "grad_norm": 0.6902556591508883,
+      "learning_rate": 2.4362206137401956e-06,
+      "loss": 0.3209,
+      "step": 8304
+    },
+    {
+      "epoch": 0.6813799893342085,
+      "grad_norm": 0.6760724596583676,
+      "learning_rate": 2.4350799643291383e-06,
+      "loss": 0.3171,
+      "step": 8305
+    },
+    {
+      "epoch": 0.6814620338843992,
+      "grad_norm": 0.7628540191955996,
+      "learning_rate": 2.4339394960486136e-06,
+      "loss": 0.3278,
+      "step": 8306
+    },
+    {
+      "epoch": 0.68154407843459,
+      "grad_norm": 0.6583426595518629,
+      "learning_rate": 2.432799208979157e-06,
+      "loss": 0.3259,
+      "step": 8307
+    },
+    {
+      "epoch": 0.6816261229847808,
+      "grad_norm": 0.7027886147142653,
+      "learning_rate": 2.4316591032012977e-06,
+      "loss": 0.2874,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6817081675349715,
+      "grad_norm": 0.750736202561974,
+      "learning_rate": 2.4305191787955445e-06,
+      "loss": 0.3335,
+      "step": 8309
+    },
+    {
+      "epoch": 0.6817902120851622,
+      "grad_norm": 0.7219383590482706,
+      "learning_rate": 2.4293794358424e-06,
+      "loss": 0.3215,
+      "step": 8310
+    },
+    {
+      "epoch": 0.681872256635353,
+      "grad_norm": 0.8043112464291104,
+      "learning_rate": 2.4282398744223483e-06,
+      "loss": 0.2966,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6819543011855438,
+      "grad_norm": 0.6941623164994203,
+      "learning_rate": 2.427100494615865e-06,
+      "loss": 0.3059,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6820363457357345,
+      "grad_norm": 0.6199747127897641,
+      "learning_rate": 2.4259612965034124e-06,
+      "loss": 0.2899,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6821183902859252,
+      "grad_norm": 0.7325916337038134,
+      "learning_rate": 2.4248222801654357e-06,
+      "loss": 0.3023,
+      "step": 8314
+    },
+    {
+      "epoch": 0.682200434836116,
+      "grad_norm": 0.668263593361036,
+      "learning_rate": 2.423683445682374e-06,
+      "loss": 0.3416,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6822824793863068,
+      "grad_norm": 0.609868872710549,
+      "learning_rate": 2.4225447931346476e-06,
+      "loss": 0.3151,
+      "step": 8316
+    },
+    {
+      "epoch": 0.6823645239364975,
+      "grad_norm": 0.6834415896438135,
+      "learning_rate": 2.4214063226026657e-06,
+      "loss": 0.2854,
+      "step": 8317
+    },
+    {
+      "epoch": 0.6824465684866883,
+      "grad_norm": 0.6928299052283201,
+      "learning_rate": 2.4202680341668285e-06,
+      "loss": 0.2777,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6825286130368791,
+      "grad_norm": 0.7521309887366998,
+      "learning_rate": 2.4191299279075163e-06,
+      "loss": 0.3643,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6826106575870697,
+      "grad_norm": 0.6764106214986015,
+      "learning_rate": 2.417992003905104e-06,
+      "loss": 0.2724,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6826927021372605,
+      "grad_norm": 0.6855029205994069,
+      "learning_rate": 2.4168542622399483e-06,
+      "loss": 0.2776,
+      "step": 8321
+    },
+    {
+      "epoch": 0.6827747466874513,
+      "grad_norm": 0.7360611363073749,
+      "learning_rate": 2.415716702992392e-06,
+      "loss": 0.3122,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6828567912376421,
+      "grad_norm": 0.8333760979942669,
+      "learning_rate": 2.4145793262427727e-06,
+      "loss": 0.2844,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6829388357878328,
+      "grad_norm": 0.7318740927046636,
+      "learning_rate": 2.4134421320714053e-06,
+      "loss": 0.3397,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6830208803380236,
+      "grad_norm": 0.7597235428267722,
+      "learning_rate": 2.412305120558599e-06,
+      "loss": 0.2862,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6831029248882143,
+      "grad_norm": 0.7132244207181545,
+      "learning_rate": 2.411168291784649e-06,
+      "loss": 0.3016,
+      "step": 8326
+    },
+    {
+      "epoch": 0.683184969438405,
+      "grad_norm": 0.7374787822613273,
+      "learning_rate": 2.410031645829834e-06,
+      "loss": 0.326,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6832670139885958,
+      "grad_norm": 0.7120550674775281,
+      "learning_rate": 2.4088951827744245e-06,
+      "loss": 0.3067,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6833490585387866,
+      "grad_norm": 0.6674611224736968,
+      "learning_rate": 2.4077589026986724e-06,
+      "loss": 0.2983,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6834311030889774,
+      "grad_norm": 0.7619781782714684,
+      "learning_rate": 2.4066228056828216e-06,
+      "loss": 0.3079,
+      "step": 8330
+    },
+    {
+      "epoch": 0.683513147639168,
+      "grad_norm": 0.728214271941948,
+      "learning_rate": 2.405486891807104e-06,
+      "loss": 0.3062,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6835951921893588,
+      "grad_norm": 0.6574215077108687,
+      "learning_rate": 2.4043511611517315e-06,
+      "loss": 0.2637,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6836772367395496,
+      "grad_norm": 0.6976390764169955,
+      "learning_rate": 2.4032156137969122e-06,
+      "loss": 0.3235,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6837592812897403,
+      "grad_norm": 0.6724593117775969,
+      "learning_rate": 2.4020802498228333e-06,
+      "loss": 0.3288,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6838413258399311,
+      "grad_norm": 0.7293970126635487,
+      "learning_rate": 2.400945069309672e-06,
+      "loss": 0.2874,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6839233703901219,
+      "grad_norm": 0.7539126545064425,
+      "learning_rate": 2.399810072337596e-06,
+      "loss": 0.3252,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6840054149403126,
+      "grad_norm": 0.6846004949489112,
+      "learning_rate": 2.3986752589867534e-06,
+      "loss": 0.2769,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6840874594905033,
+      "grad_norm": 0.7011799536026369,
+      "learning_rate": 2.3975406293372866e-06,
+      "loss": 0.3026,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6841695040406941,
+      "grad_norm": 0.7119134902843873,
+      "learning_rate": 2.3964061834693174e-06,
+      "loss": 0.2741,
+      "step": 8339
+    },
+    {
+      "epoch": 0.6842515485908849,
+      "grad_norm": 0.8849286288090471,
+      "learning_rate": 2.3952719214629605e-06,
+      "loss": 0.3138,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6843335931410756,
+      "grad_norm": 4.984012995523921,
+      "learning_rate": 2.3941378433983175e-06,
+      "loss": 0.2903,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6844156376912663,
+      "grad_norm": 0.628296612426894,
+      "learning_rate": 2.3930039493554714e-06,
+      "loss": 0.2969,
+      "step": 8342
+    },
+    {
+      "epoch": 0.6844976822414571,
+      "grad_norm": 0.6539941355604384,
+      "learning_rate": 2.391870239414498e-06,
+      "loss": 0.2891,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6845797267916479,
+      "grad_norm": 0.6857007650979102,
+      "learning_rate": 2.39073671365546e-06,
+      "loss": 0.2819,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6846617713418386,
+      "grad_norm": 0.7300873636236981,
+      "learning_rate": 2.389603372158401e-06,
+      "loss": 0.2777,
+      "step": 8345
+    },
+    {
+      "epoch": 0.6847438158920294,
+      "grad_norm": 0.6492972252259237,
+      "learning_rate": 2.38847021500336e-06,
+      "loss": 0.3513,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6848258604422202,
+      "grad_norm": 0.729383544593263,
+      "learning_rate": 2.387337242270355e-06,
+      "loss": 0.2978,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6849079049924108,
+      "grad_norm": 0.6444616493094725,
+      "learning_rate": 2.386204454039398e-06,
+      "loss": 0.3251,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6849899495426016,
+      "grad_norm": 0.7346164978447326,
+      "learning_rate": 2.3850718503904835e-06,
+      "loss": 0.3178,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6850719940927924,
+      "grad_norm": 0.7114100551734897,
+      "learning_rate": 2.3839394314035915e-06,
+      "loss": 0.2803,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6851540386429832,
+      "grad_norm": 0.687441984013582,
+      "learning_rate": 2.382807197158696e-06,
+      "loss": 0.3296,
+      "step": 8351
+    },
+    {
+      "epoch": 0.6852360831931739,
+      "grad_norm": 0.7256977952435328,
+      "learning_rate": 2.381675147735749e-06,
+      "loss": 0.3173,
+      "step": 8352
+    },
+    {
+      "epoch": 0.6853181277433646,
+      "grad_norm": 0.7534423490819239,
+      "learning_rate": 2.380543283214697e-06,
+      "loss": 0.3507,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6854001722935554,
+      "grad_norm": 0.6052849423802877,
+      "learning_rate": 2.379411603675471e-06,
+      "loss": 0.6187,
+      "step": 8354
+    },
+    {
+      "epoch": 0.6854822168437461,
+      "grad_norm": 0.6693732846388519,
+      "learning_rate": 2.378280109197986e-06,
+      "loss": 0.3184,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6855642613939369,
+      "grad_norm": 0.6991019614950318,
+      "learning_rate": 2.377148799862149e-06,
+      "loss": 0.3161,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6856463059441277,
+      "grad_norm": 0.6758131587398727,
+      "learning_rate": 2.376017675747848e-06,
+      "loss": 0.2946,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6857283504943185,
+      "grad_norm": 0.7505729893028888,
+      "learning_rate": 2.374886736934963e-06,
+      "loss": 0.3114,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6858103950445091,
+      "grad_norm": 0.6154745280762708,
+      "learning_rate": 2.3737559835033613e-06,
+      "loss": 0.2873,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6858924395946999,
+      "grad_norm": 0.7488275407516267,
+      "learning_rate": 2.372625415532891e-06,
+      "loss": 0.2975,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6859744841448907,
+      "grad_norm": 0.6363393887657729,
+      "learning_rate": 2.3714950331033937e-06,
+      "loss": 0.3004,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6860565286950814,
+      "grad_norm": 0.7129044508133185,
+      "learning_rate": 2.3703648362946948e-06,
+      "loss": 0.3199,
+      "step": 8362
+    },
+    {
+      "epoch": 0.6861385732452722,
+      "grad_norm": 0.7511311743378948,
+      "learning_rate": 2.3692348251866045e-06,
+      "loss": 0.329,
+      "step": 8363
+    },
+    {
+      "epoch": 0.686220617795463,
+      "grad_norm": 0.755186784984683,
+      "learning_rate": 2.3681049998589256e-06,
+      "loss": 0.2817,
+      "step": 8364
+    },
+    {
+      "epoch": 0.6863026623456537,
+      "grad_norm": 0.6646631003339217,
+      "learning_rate": 2.3669753603914414e-06,
+      "loss": 0.2845,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6863847068958444,
+      "grad_norm": 0.7699812523270269,
+      "learning_rate": 2.365845906863927e-06,
+      "loss": 0.2842,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6864667514460352,
+      "grad_norm": 0.7009009421981954,
+      "learning_rate": 2.3647166393561442e-06,
+      "loss": 0.2977,
+      "step": 8367
+    },
+    {
+      "epoch": 0.686548795996226,
+      "grad_norm": 0.6868733216544639,
+      "learning_rate": 2.3635875579478373e-06,
+      "loss": 0.2798,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6866308405464167,
+      "grad_norm": 0.6179613650778729,
+      "learning_rate": 2.362458662718743e-06,
+      "loss": 0.2611,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6867128850966074,
+      "grad_norm": 0.6522440106612896,
+      "learning_rate": 2.361329953748578e-06,
+      "loss": 0.3048,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6867949296467982,
+      "grad_norm": 0.6273403478604974,
+      "learning_rate": 2.3602014311170524e-06,
+      "loss": 0.3088,
+      "step": 8371
+    },
+    {
+      "epoch": 0.686876974196989,
+      "grad_norm": 0.6395300013210247,
+      "learning_rate": 2.3590730949038625e-06,
+      "loss": 0.3126,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6869590187471797,
+      "grad_norm": 0.6721268394649139,
+      "learning_rate": 2.357944945188686e-06,
+      "loss": 0.3115,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6870410632973705,
+      "grad_norm": 0.6369462091427557,
+      "learning_rate": 2.356816982051195e-06,
+      "loss": 0.3129,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6871231078475613,
+      "grad_norm": 0.6774249195154234,
+      "learning_rate": 2.3556892055710394e-06,
+      "loss": 0.3119,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6872051523977519,
+      "grad_norm": 0.5859881242659029,
+      "learning_rate": 2.354561615827866e-06,
+      "loss": 0.6369,
+      "step": 8376
+    },
+    {
+      "epoch": 0.6872871969479427,
+      "grad_norm": 0.6629160921274511,
+      "learning_rate": 2.353434212901301e-06,
+      "loss": 0.3126,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6873692414981335,
+      "grad_norm": 0.7177084658276051,
+      "learning_rate": 2.3523069968709577e-06,
+      "loss": 0.3234,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6874512860483243,
+      "grad_norm": 0.6750213548446535,
+      "learning_rate": 2.351179967816441e-06,
+      "loss": 0.2955,
+      "step": 8379
+    },
+    {
+      "epoch": 0.687533330598515,
+      "grad_norm": 0.6712082416814205,
+      "learning_rate": 2.350053125817341e-06,
+      "loss": 0.2809,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6876153751487057,
+      "grad_norm": 0.7075057205270392,
+      "learning_rate": 2.3489264709532294e-06,
+      "loss": 0.3291,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6876974196988965,
+      "grad_norm": 0.6456892834411921,
+      "learning_rate": 2.3478000033036735e-06,
+      "loss": 0.2874,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6877794642490872,
+      "grad_norm": 0.6280025663581629,
+      "learning_rate": 2.3466737229482183e-06,
+      "loss": 0.2762,
+      "step": 8383
+    },
+    {
+      "epoch": 0.687861508799278,
+      "grad_norm": 0.6841597653661252,
+      "learning_rate": 2.345547629966402e-06,
+      "loss": 0.2637,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6879435533494688,
+      "grad_norm": 0.6836937858614035,
+      "learning_rate": 2.344421724437749e-06,
+      "loss": 0.284,
+      "step": 8385
+    },
+    {
+      "epoch": 0.6880255978996596,
+      "grad_norm": 0.6717997290561433,
+      "learning_rate": 2.343296006441766e-06,
+      "loss": 0.3222,
+      "step": 8386
+    },
+    {
+      "epoch": 0.6881076424498502,
+      "grad_norm": 0.7721185283749203,
+      "learning_rate": 2.342170476057952e-06,
+      "loss": 0.3077,
+      "step": 8387
+    },
+    {
+      "epoch": 0.688189687000041,
+      "grad_norm": 0.655219148359471,
+      "learning_rate": 2.3410451333657876e-06,
+      "loss": 0.331,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6882717315502318,
+      "grad_norm": 0.6939437390465069,
+      "learning_rate": 2.3399199784447453e-06,
+      "loss": 0.3083,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6883537761004225,
+      "grad_norm": 0.7185900073546676,
+      "learning_rate": 2.3387950113742803e-06,
+      "loss": 0.3202,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6884358206506133,
+      "grad_norm": 0.8192364709046628,
+      "learning_rate": 2.3376702322338346e-06,
+      "loss": 0.2883,
+      "step": 8391
+    },
+    {
+      "epoch": 0.688517865200804,
+      "grad_norm": 0.6060552913251196,
+      "learning_rate": 2.3365456411028394e-06,
+      "loss": 0.2902,
+      "step": 8392
+    },
+    {
+      "epoch": 0.6885999097509948,
+      "grad_norm": 0.7096432116413979,
+      "learning_rate": 2.335421238060714e-06,
+      "loss": 0.3075,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6886819543011855,
+      "grad_norm": 0.6702899098792274,
+      "learning_rate": 2.3342970231868573e-06,
+      "loss": 0.2883,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6887639988513763,
+      "grad_norm": 0.7406051329852473,
+      "learning_rate": 2.333172996560664e-06,
+      "loss": 0.3203,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6888460434015671,
+      "grad_norm": 0.6734142752386051,
+      "learning_rate": 2.332049158261508e-06,
+      "loss": 0.3061,
+      "step": 8396
+    },
+    {
+      "epoch": 0.6889280879517578,
+      "grad_norm": 0.6591955959097712,
+      "learning_rate": 2.330925508368754e-06,
+      "loss": 0.2941,
+      "step": 8397
+    },
+    {
+      "epoch": 0.6890101325019485,
+      "grad_norm": 0.6907204613312792,
+      "learning_rate": 2.329802046961753e-06,
+      "loss": 0.2739,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6890921770521393,
+      "grad_norm": 0.6543796618349068,
+      "learning_rate": 2.328678774119841e-06,
+      "loss": 0.2744,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6891742216023301,
+      "grad_norm": 0.6984399513185185,
+      "learning_rate": 2.3275556899223435e-06,
+      "loss": 0.3246,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6892562661525208,
+      "grad_norm": 0.6654122757956571,
+      "learning_rate": 2.3264327944485675e-06,
+      "loss": 0.3134,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6893383107027116,
+      "grad_norm": 0.6521645585444783,
+      "learning_rate": 2.3253100877778145e-06,
+      "loss": 0.3486,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6894203552529024,
+      "grad_norm": 0.6964565028404341,
+      "learning_rate": 2.3241875699893635e-06,
+      "loss": 0.3125,
+      "step": 8403
+    },
+    {
+      "epoch": 0.689502399803093,
+      "grad_norm": 0.7321331046639769,
+      "learning_rate": 2.323065241162489e-06,
+      "loss": 0.2954,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6895844443532838,
+      "grad_norm": 0.6860738215173128,
+      "learning_rate": 2.321943101376447e-06,
+      "loss": 0.3058,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6896664889034746,
+      "grad_norm": 0.652300870071461,
+      "learning_rate": 2.3208211507104787e-06,
+      "loss": 0.2894,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6897485334536654,
+      "grad_norm": 0.733998015983069,
+      "learning_rate": 2.3196993892438165e-06,
+      "loss": 0.311,
+      "step": 8407
+    },
+    {
+      "epoch": 0.689830578003856,
+      "grad_norm": 0.6914687688186681,
+      "learning_rate": 2.3185778170556785e-06,
+      "loss": 0.3284,
+      "step": 8408
+    },
+    {
+      "epoch": 0.6899126225540468,
+      "grad_norm": 0.7245804164525625,
+      "learning_rate": 2.3174564342252646e-06,
+      "loss": 0.2931,
+      "step": 8409
+    },
+    {
+      "epoch": 0.6899946671042376,
+      "grad_norm": 0.7101541334730583,
+      "learning_rate": 2.3163352408317687e-06,
+      "loss": 0.3069,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6900767116544284,
+      "grad_norm": 0.6322689495445846,
+      "learning_rate": 2.315214236954367e-06,
+      "loss": 0.2983,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6901587562046191,
+      "grad_norm": 0.6784758959593733,
+      "learning_rate": 2.3140934226722215e-06,
+      "loss": 0.3223,
+      "step": 8412
+    },
+    {
+      "epoch": 0.6902408007548099,
+      "grad_norm": 0.7165070797637035,
+      "learning_rate": 2.312972798064485e-06,
+      "loss": 0.2881,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6903228453050007,
+      "grad_norm": 0.9795618506814592,
+      "learning_rate": 2.3118523632102907e-06,
+      "loss": 0.3018,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6904048898551913,
+      "grad_norm": 0.6737161617842534,
+      "learning_rate": 2.310732118188764e-06,
+      "loss": 0.281,
+      "step": 8415
+    },
+    {
+      "epoch": 0.6904869344053821,
+      "grad_norm": 0.7252962922188236,
+      "learning_rate": 2.309612063079015e-06,
+      "loss": 0.3409,
+      "step": 8416
+    },
+    {
+      "epoch": 0.6905689789555729,
+      "grad_norm": 0.7405399947997914,
+      "learning_rate": 2.308492197960141e-06,
+      "loss": 0.3023,
+      "step": 8417
+    },
+    {
+      "epoch": 0.6906510235057637,
+      "grad_norm": 0.6482852553494243,
+      "learning_rate": 2.307372522911223e-06,
+      "loss": 0.3002,
+      "step": 8418
+    },
+    {
+      "epoch": 0.6907330680559544,
+      "grad_norm": 0.7761340383932952,
+      "learning_rate": 2.3062530380113303e-06,
+      "loss": 0.2766,
+      "step": 8419
+    },
+    {
+      "epoch": 0.6908151126061451,
+      "grad_norm": 0.6888861654444001,
+      "learning_rate": 2.30513374333952e-06,
+      "loss": 0.335,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6908971571563359,
+      "grad_norm": 0.6917846413032996,
+      "learning_rate": 2.3040146389748365e-06,
+      "loss": 0.296,
+      "step": 8421
+    },
+    {
+      "epoch": 0.6909792017065266,
+      "grad_norm": 0.6051934748856868,
+      "learning_rate": 2.3028957249963057e-06,
+      "loss": 0.2984,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6910612462567174,
+      "grad_norm": 0.6999567361788783,
+      "learning_rate": 2.301777001482948e-06,
+      "loss": 0.3285,
+      "step": 8423
+    },
+    {
+      "epoch": 0.6911432908069082,
+      "grad_norm": 0.774632852509146,
+      "learning_rate": 2.300658468513762e-06,
+      "loss": 0.3186,
+      "step": 8424
+    },
+    {
+      "epoch": 0.691225335357099,
+      "grad_norm": 0.7590620454978356,
+      "learning_rate": 2.299540126167737e-06,
+      "loss": 0.3571,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6913073799072896,
+      "grad_norm": 0.6274022148554714,
+      "learning_rate": 2.2984219745238524e-06,
+      "loss": 0.3173,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6913894244574804,
+      "grad_norm": 0.7065588446296904,
+      "learning_rate": 2.2973040136610652e-06,
+      "loss": 0.2884,
+      "step": 8427
+    },
+    {
+      "epoch": 0.6914714690076712,
+      "grad_norm": 0.7137615471729695,
+      "learning_rate": 2.2961862436583266e-06,
+      "loss": 0.2797,
+      "step": 8428
+    },
+    {
+      "epoch": 0.6915535135578619,
+      "grad_norm": 0.6907114399849358,
+      "learning_rate": 2.295068664594573e-06,
+      "loss": 0.3216,
+      "step": 8429
+    },
+    {
+      "epoch": 0.6916355581080527,
+      "grad_norm": 0.7028024497611337,
+      "learning_rate": 2.2939512765487247e-06,
+      "loss": 0.3036,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6917176026582434,
+      "grad_norm": 0.7238378974480834,
+      "learning_rate": 2.292834079599688e-06,
+      "loss": 0.2868,
+      "step": 8431
+    },
+    {
+      "epoch": 0.6917996472084342,
+      "grad_norm": 0.6343720710879647,
+      "learning_rate": 2.2917170738263607e-06,
+      "loss": 0.2673,
+      "step": 8432
+    },
+    {
+      "epoch": 0.6918816917586249,
+      "grad_norm": 0.6703199126717517,
+      "learning_rate": 2.290600259307621e-06,
+      "loss": 0.2961,
+      "step": 8433
+    },
+    {
+      "epoch": 0.6919637363088157,
+      "grad_norm": 0.7031926368665418,
+      "learning_rate": 2.2894836361223393e-06,
+      "loss": 0.3334,
+      "step": 8434
+    },
+    {
+      "epoch": 0.6920457808590065,
+      "grad_norm": 0.6585703058593593,
+      "learning_rate": 2.2883672043493672e-06,
+      "loss": 0.3132,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6921278254091972,
+      "grad_norm": 0.6557232623371865,
+      "learning_rate": 2.2872509640675476e-06,
+      "loss": 0.3253,
+      "step": 8436
+    },
+    {
+      "epoch": 0.6922098699593879,
+      "grad_norm": 0.6947883515319264,
+      "learning_rate": 2.2861349153557057e-06,
+      "loss": 0.3443,
+      "step": 8437
+    },
+    {
+      "epoch": 0.6922919145095787,
+      "grad_norm": 0.7435085581347918,
+      "learning_rate": 2.285019058292655e-06,
+      "loss": 0.3291,
+      "step": 8438
+    },
+    {
+      "epoch": 0.6923739590597695,
+      "grad_norm": 0.7536326158277764,
+      "learning_rate": 2.283903392957199e-06,
+      "loss": 0.3012,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6924560036099602,
+      "grad_norm": 0.7515719463836891,
+      "learning_rate": 2.2827879194281196e-06,
+      "loss": 0.29,
+      "step": 8440
+    },
+    {
+      "epoch": 0.692538048160151,
+      "grad_norm": 0.6797721049335307,
+      "learning_rate": 2.2816726377841935e-06,
+      "loss": 0.3321,
+      "step": 8441
+    },
+    {
+      "epoch": 0.6926200927103418,
+      "grad_norm": 0.7365241092248542,
+      "learning_rate": 2.280557548104177e-06,
+      "loss": 0.3128,
+      "step": 8442
+    },
+    {
+      "epoch": 0.6927021372605324,
+      "grad_norm": 0.6538619279453959,
+      "learning_rate": 2.2794426504668164e-06,
+      "loss": 0.3186,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6927841818107232,
+      "grad_norm": 0.6663924923873927,
+      "learning_rate": 2.278327944950848e-06,
+      "loss": 0.2943,
+      "step": 8444
+    },
+    {
+      "epoch": 0.692866226360914,
+      "grad_norm": 0.6800300791901293,
+      "learning_rate": 2.2772134316349864e-06,
+      "loss": 0.3265,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6929482709111048,
+      "grad_norm": 0.6473853602246318,
+      "learning_rate": 2.2760991105979363e-06,
+      "loss": 0.3363,
+      "step": 8446
+    },
+    {
+      "epoch": 0.6930303154612955,
+      "grad_norm": 0.6513515515546079,
+      "learning_rate": 2.274984981918393e-06,
+      "loss": 0.3162,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6931123600114862,
+      "grad_norm": 0.6736381327219939,
+      "learning_rate": 2.2738710456750296e-06,
+      "loss": 0.2829,
+      "step": 8448
+    },
+    {
+      "epoch": 0.693194404561677,
+      "grad_norm": 0.6504726319282755,
+      "learning_rate": 2.2727573019465154e-06,
+      "loss": 0.2897,
+      "step": 8449
+    },
+    {
+      "epoch": 0.6932764491118677,
+      "grad_norm": 0.8195302209823659,
+      "learning_rate": 2.271643750811497e-06,
+      "loss": 0.3368,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6933584936620585,
+      "grad_norm": 0.7639111801703793,
+      "learning_rate": 2.270530392348613e-06,
+      "loss": 0.3076,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6934405382122493,
+      "grad_norm": 0.6496517855364619,
+      "learning_rate": 2.26941722663649e-06,
+      "loss": 0.3032,
+      "step": 8452
+    },
+    {
+      "epoch": 0.69352258276244,
+      "grad_norm": 0.6814060996710656,
+      "learning_rate": 2.268304253753733e-06,
+      "loss": 0.3076,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6936046273126307,
+      "grad_norm": 0.5991610833811132,
+      "learning_rate": 2.267191473778943e-06,
+      "loss": 0.2723,
+      "step": 8454
+    },
+    {
+      "epoch": 0.6936866718628215,
+      "grad_norm": 0.6676611007981537,
+      "learning_rate": 2.2660788867906985e-06,
+      "loss": 0.2835,
+      "step": 8455
+    },
+    {
+      "epoch": 0.6937687164130123,
+      "grad_norm": 0.7134101157471192,
+      "learning_rate": 2.264966492867571e-06,
+      "loss": 0.3461,
+      "step": 8456
+    },
+    {
+      "epoch": 0.693850760963203,
+      "grad_norm": 0.7130902454564787,
+      "learning_rate": 2.2638542920881175e-06,
+      "loss": 0.2799,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6939328055133938,
+      "grad_norm": 0.6772103767263459,
+      "learning_rate": 2.262742284530877e-06,
+      "loss": 0.2928,
+      "step": 8458
+    },
+    {
+      "epoch": 0.6940148500635845,
+      "grad_norm": 0.6487930454348104,
+      "learning_rate": 2.261630470274378e-06,
+      "loss": 0.3334,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6940968946137753,
+      "grad_norm": 0.6493656966667773,
+      "learning_rate": 2.2605188493971363e-06,
+      "loss": 0.3043,
+      "step": 8460
+    },
+    {
+      "epoch": 0.694178939163966,
+      "grad_norm": 0.742167988955624,
+      "learning_rate": 2.2594074219776514e-06,
+      "loss": 0.3085,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6942609837141568,
+      "grad_norm": 0.6371695928703363,
+      "learning_rate": 2.2582961880944126e-06,
+      "loss": 0.3264,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6943430282643476,
+      "grad_norm": 0.6403792179120508,
+      "learning_rate": 2.2571851478258903e-06,
+      "loss": 0.2927,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6944250728145382,
+      "grad_norm": 0.5979113380143684,
+      "learning_rate": 2.256074301250546e-06,
+      "loss": 0.2462,
+      "step": 8464
+    },
+    {
+      "epoch": 0.694507117364729,
+      "grad_norm": 0.6881220846485586,
+      "learning_rate": 2.2549636484468284e-06,
+      "loss": 0.3035,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6945891619149198,
+      "grad_norm": 0.6805016274598356,
+      "learning_rate": 2.2538531894931655e-06,
+      "loss": 0.321,
+      "step": 8466
+    },
+    {
+      "epoch": 0.6946712064651106,
+      "grad_norm": 0.6710819876545808,
+      "learning_rate": 2.2527429244679805e-06,
+      "loss": 0.3034,
+      "step": 8467
+    },
+    {
+      "epoch": 0.6947532510153013,
+      "grad_norm": 0.7554629913269382,
+      "learning_rate": 2.251632853449674e-06,
+      "loss": 0.2992,
+      "step": 8468
+    },
+    {
+      "epoch": 0.6948352955654921,
+      "grad_norm": 0.7146282863423231,
+      "learning_rate": 2.2505229765166403e-06,
+      "loss": 0.3071,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6949173401156828,
+      "grad_norm": 0.7353264341498161,
+      "learning_rate": 2.249413293747258e-06,
+      "loss": 0.2962,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6949993846658735,
+      "grad_norm": 0.6393813188790107,
+      "learning_rate": 2.2483038052198888e-06,
+      "loss": 0.2817,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6950814292160643,
+      "grad_norm": 0.7686730147652518,
+      "learning_rate": 2.2471945110128846e-06,
+      "loss": 0.2963,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6951634737662551,
+      "grad_norm": 0.6804731470868303,
+      "learning_rate": 2.2460854112045823e-06,
+      "loss": 0.3127,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6952455183164459,
+      "grad_norm": 0.6693717412334199,
+      "learning_rate": 2.244976505873301e-06,
+      "loss": 0.2909,
+      "step": 8474
+    },
+    {
+      "epoch": 0.6953275628666366,
+      "grad_norm": 0.6333412071259485,
+      "learning_rate": 2.2438677950973546e-06,
+      "loss": 0.2962,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6954096074168273,
+      "grad_norm": 0.6845210240954418,
+      "learning_rate": 2.242759278955035e-06,
+      "loss": 0.2844,
+      "step": 8476
+    },
+    {
+      "epoch": 0.6954916519670181,
+      "grad_norm": 0.6632446044707546,
+      "learning_rate": 2.2416509575246276e-06,
+      "loss": 0.2745,
+      "step": 8477
+    },
+    {
+      "epoch": 0.6955736965172088,
+      "grad_norm": 0.6412999220113647,
+      "learning_rate": 2.2405428308843953e-06,
+      "loss": 0.3352,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6956557410673996,
+      "grad_norm": 0.6412125454679405,
+      "learning_rate": 2.2394348991125954e-06,
+      "loss": 0.3118,
+      "step": 8479
+    },
+    {
+      "epoch": 0.6957377856175904,
+      "grad_norm": 0.6568649426558497,
+      "learning_rate": 2.2383271622874693e-06,
+      "loss": 0.298,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6958198301677812,
+      "grad_norm": 0.6955852841164623,
+      "learning_rate": 2.23721962048724e-06,
+      "loss": 0.3034,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6959018747179718,
+      "grad_norm": 0.936172494490829,
+      "learning_rate": 2.2361122737901225e-06,
+      "loss": 0.3451,
+      "step": 8482
+    },
+    {
+      "epoch": 0.6959839192681626,
+      "grad_norm": 0.6543080035420153,
+      "learning_rate": 2.2350051222743175e-06,
+      "loss": 0.2968,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6960659638183534,
+      "grad_norm": 0.9893957896938688,
+      "learning_rate": 2.2338981660180065e-06,
+      "loss": 0.3095,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6961480083685441,
+      "grad_norm": 0.6410898296729268,
+      "learning_rate": 2.2327914050993647e-06,
+      "loss": 0.2894,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6962300529187349,
+      "grad_norm": 0.6008335056655068,
+      "learning_rate": 2.2316848395965483e-06,
+      "loss": 0.2924,
+      "step": 8486
+    },
+    {
+      "epoch": 0.6963120974689256,
+      "grad_norm": 0.6943207379181008,
+      "learning_rate": 2.230578469587699e-06,
+      "loss": 0.3126,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6963941420191164,
+      "grad_norm": 0.6685655355411761,
+      "learning_rate": 2.2294722951509514e-06,
+      "loss": 0.2937,
+      "step": 8488
+    },
+    {
+      "epoch": 0.6964761865693071,
+      "grad_norm": 0.6608278371609908,
+      "learning_rate": 2.228366316364416e-06,
+      "loss": 0.3042,
+      "step": 8489
+    },
+    {
+      "epoch": 0.6965582311194979,
+      "grad_norm": 0.6772698193010623,
+      "learning_rate": 2.227260533306201e-06,
+      "loss": 0.2967,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6966402756696887,
+      "grad_norm": 0.7581986358014706,
+      "learning_rate": 2.2261549460543903e-06,
+      "loss": 0.2964,
+      "step": 8491
+    },
+    {
+      "epoch": 0.6967223202198793,
+      "grad_norm": 0.7106067005898171,
+      "learning_rate": 2.2250495546870605e-06,
+      "loss": 0.3177,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6968043647700701,
+      "grad_norm": 0.7422588591987173,
+      "learning_rate": 2.223944359282275e-06,
+      "loss": 0.315,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6968864093202609,
+      "grad_norm": 0.714669480799354,
+      "learning_rate": 2.222839359918077e-06,
+      "loss": 0.3072,
+      "step": 8494
+    },
+    {
+      "epoch": 0.6969684538704517,
+      "grad_norm": 0.6523179932380464,
+      "learning_rate": 2.2217345566725033e-06,
+      "loss": 0.2943,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6970504984206424,
+      "grad_norm": 0.719719935610785,
+      "learning_rate": 2.2206299496235696e-06,
+      "loss": 0.3108,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6971325429708332,
+      "grad_norm": 0.6682463017930667,
+      "learning_rate": 2.2195255388492837e-06,
+      "loss": 0.2686,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6972145875210239,
+      "grad_norm": 0.6776076471543394,
+      "learning_rate": 2.2184213244276386e-06,
+      "loss": 0.2927,
+      "step": 8498
+    },
+    {
+      "epoch": 0.6972966320712146,
+      "grad_norm": 0.6837747519286048,
+      "learning_rate": 2.2173173064366093e-06,
+      "loss": 0.3096,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6973786766214054,
+      "grad_norm": 0.6753766162029626,
+      "learning_rate": 2.216213484954162e-06,
+      "loss": 0.2963,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6974607211715962,
+      "grad_norm": 0.7072782753978214,
+      "learning_rate": 2.215109860058247e-06,
+      "loss": 0.2913,
+      "step": 8501
+    },
+    {
+      "epoch": 0.697542765721787,
+      "grad_norm": 0.6762272123048779,
+      "learning_rate": 2.2140064318267974e-06,
+      "loss": 0.2769,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6976248102719776,
+      "grad_norm": 0.7134288471854526,
+      "learning_rate": 2.212903200337739e-06,
+      "loss": 0.3082,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6977068548221684,
+      "grad_norm": 0.7065478549039748,
+      "learning_rate": 2.211800165668977e-06,
+      "loss": 0.3,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6977888993723592,
+      "grad_norm": 0.6534283602810218,
+      "learning_rate": 2.210697327898407e-06,
+      "loss": 0.304,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6978709439225499,
+      "grad_norm": 0.6397686726802237,
+      "learning_rate": 2.209594687103914e-06,
+      "loss": 0.3041,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6979529884727407,
+      "grad_norm": 0.6954571643098861,
+      "learning_rate": 2.2084922433633577e-06,
+      "loss": 0.3063,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6980350330229315,
+      "grad_norm": 0.688651714856069,
+      "learning_rate": 2.2073899967545966e-06,
+      "loss": 0.3022,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6981170775731222,
+      "grad_norm": 0.7298141847918963,
+      "learning_rate": 2.2062879473554654e-06,
+      "loss": 0.3123,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6981991221233129,
+      "grad_norm": 0.7462393783912802,
+      "learning_rate": 2.205186095243791e-06,
+      "loss": 0.3072,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6982811666735037,
+      "grad_norm": 0.7829643131397663,
+      "learning_rate": 2.204084440497386e-06,
+      "loss": 0.3285,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6983632112236945,
+      "grad_norm": 0.6600021830884221,
+      "learning_rate": 2.202982983194044e-06,
+      "loss": 0.284,
+      "step": 8512
+    },
+    {
+      "epoch": 0.6984452557738852,
+      "grad_norm": 0.6595529175556403,
+      "learning_rate": 2.201881723411552e-06,
+      "loss": 0.3056,
+      "step": 8513
+    },
+    {
+      "epoch": 0.698527300324076,
+      "grad_norm": 0.6382586815899334,
+      "learning_rate": 2.2007806612276767e-06,
+      "loss": 0.2816,
+      "step": 8514
+    },
+    {
+      "epoch": 0.6986093448742667,
+      "grad_norm": 0.6837743251660042,
+      "learning_rate": 2.1996797967201727e-06,
+      "loss": 0.2863,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6986913894244575,
+      "grad_norm": 0.6650046922017776,
+      "learning_rate": 2.1985791299667835e-06,
+      "loss": 0.2798,
+      "step": 8516
+    },
+    {
+      "epoch": 0.6987734339746482,
+      "grad_norm": 0.694503330612488,
+      "learning_rate": 2.1974786610452342e-06,
+      "loss": 0.3267,
+      "step": 8517
+    },
+    {
+      "epoch": 0.698855478524839,
+      "grad_norm": 0.6449413886258419,
+      "learning_rate": 2.1963783900332384e-06,
+      "loss": 0.3338,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6989375230750298,
+      "grad_norm": 0.6837157264866787,
+      "learning_rate": 2.195278317008499e-06,
+      "loss": 0.2931,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6990195676252204,
+      "grad_norm": 0.6671018391937721,
+      "learning_rate": 2.194178442048697e-06,
+      "loss": 0.2965,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6991016121754112,
+      "grad_norm": 0.7151557565524083,
+      "learning_rate": 2.1930787652315066e-06,
+      "loss": 0.3071,
+      "step": 8521
+    },
+    {
+      "epoch": 0.699183656725602,
+      "grad_norm": 0.6695839138099848,
+      "learning_rate": 2.1919792866345822e-06,
+      "loss": 0.3096,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6992657012757928,
+      "grad_norm": 0.7455062299650645,
+      "learning_rate": 2.19088000633557e-06,
+      "loss": 0.2999,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6993477458259835,
+      "grad_norm": 0.873165421170161,
+      "learning_rate": 2.1897809244120998e-06,
+      "loss": 0.3169,
+      "step": 8524
+    },
+    {
+      "epoch": 0.6994297903761743,
+      "grad_norm": 0.6198386032555964,
+      "learning_rate": 2.188682040941784e-06,
+      "loss": 0.2908,
+      "step": 8525
+    },
+    {
+      "epoch": 0.699511834926365,
+      "grad_norm": 0.7065028096416244,
+      "learning_rate": 2.187583356002228e-06,
+      "loss": 0.3248,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6995938794765558,
+      "grad_norm": 0.6640600448042938,
+      "learning_rate": 2.1864848696710146e-06,
+      "loss": 0.3166,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6996759240267465,
+      "grad_norm": 0.7561087207564882,
+      "learning_rate": 2.1853865820257215e-06,
+      "loss": 0.3687,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6997579685769373,
+      "grad_norm": 0.6366687769359334,
+      "learning_rate": 2.1842884931439055e-06,
+      "loss": 0.2809,
+      "step": 8529
+    },
+    {
+      "epoch": 0.6998400131271281,
+      "grad_norm": 0.6452675215451865,
+      "learning_rate": 2.1831906031031115e-06,
+      "loss": 0.2979,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6999220576773187,
+      "grad_norm": 0.6693408975177367,
+      "learning_rate": 2.1820929119808727e-06,
+      "loss": 0.3022,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7000041022275095,
+      "grad_norm": 0.6955174710510156,
+      "learning_rate": 2.180995419854703e-06,
+      "loss": 0.3174,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7000861467777003,
+      "grad_norm": 0.7109368201493697,
+      "learning_rate": 2.179898126802108e-06,
+      "loss": 0.2856,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7001681913278911,
+      "grad_norm": 0.7141792313164439,
+      "learning_rate": 2.1788010329005783e-06,
+      "loss": 0.2984,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7002502358780818,
+      "grad_norm": 0.6545095203426954,
+      "learning_rate": 2.1777041382275854e-06,
+      "loss": 0.3253,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7003322804282726,
+      "grad_norm": 0.6443531288297351,
+      "learning_rate": 2.1766074428605916e-06,
+      "loss": 0.3164,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7004143249784633,
+      "grad_norm": 0.6668816268413645,
+      "learning_rate": 2.175510946877046e-06,
+      "loss": 0.303,
+      "step": 8537
+    },
+    {
+      "epoch": 0.700496369528654,
+      "grad_norm": 0.7366697625616684,
+      "learning_rate": 2.1744146503543777e-06,
+      "loss": 0.3294,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7005784140788448,
+      "grad_norm": 0.8364004962291854,
+      "learning_rate": 2.1733185533700093e-06,
+      "loss": 0.324,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7006604586290356,
+      "grad_norm": 0.6884578951158553,
+      "learning_rate": 2.172222656001341e-06,
+      "loss": 0.3335,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7007425031792264,
+      "grad_norm": 0.6753598187263437,
+      "learning_rate": 2.171126958325767e-06,
+      "loss": 0.2811,
+      "step": 8541
+    },
+    {
+      "epoch": 0.700824547729417,
+      "grad_norm": 0.7736914807118115,
+      "learning_rate": 2.170031460420663e-06,
+      "loss": 0.3325,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7009065922796078,
+      "grad_norm": 0.6886556350711053,
+      "learning_rate": 2.168936162363388e-06,
+      "loss": 0.3052,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7009886368297986,
+      "grad_norm": 0.7372744653935149,
+      "learning_rate": 2.167841064231295e-06,
+      "loss": 0.2616,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7010706813799893,
+      "grad_norm": 0.6259341958917373,
+      "learning_rate": 2.166746166101714e-06,
+      "loss": 0.3188,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7011527259301801,
+      "grad_norm": 0.6589217726028992,
+      "learning_rate": 2.1656514680519676e-06,
+      "loss": 0.309,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7012347704803709,
+      "grad_norm": 0.8973042663716344,
+      "learning_rate": 2.1645569701593616e-06,
+      "loss": 0.3005,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7013168150305616,
+      "grad_norm": 0.7434541644038816,
+      "learning_rate": 2.1634626725011854e-06,
+      "loss": 0.3012,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7013988595807523,
+      "grad_norm": 0.7137574172371192,
+      "learning_rate": 2.1623685751547198e-06,
+      "loss": 0.3128,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7014809041309431,
+      "grad_norm": 0.6927715548597561,
+      "learning_rate": 2.1612746781972242e-06,
+      "loss": 0.2869,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7015629486811339,
+      "grad_norm": 0.6632173784691282,
+      "learning_rate": 2.1601809817059503e-06,
+      "loss": 0.2964,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7016449932313246,
+      "grad_norm": 0.6520517211952185,
+      "learning_rate": 2.159087485758135e-06,
+      "loss": 0.3139,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7017270377815154,
+      "grad_norm": 0.6854977554721713,
+      "learning_rate": 2.157994190430995e-06,
+      "loss": 0.2816,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7018090823317061,
+      "grad_norm": 0.646080073005483,
+      "learning_rate": 2.1569010958017416e-06,
+      "loss": 0.2847,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7018911268818969,
+      "grad_norm": 0.6770645417625125,
+      "learning_rate": 2.155808201947563e-06,
+      "loss": 0.2978,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7019731714320876,
+      "grad_norm": 0.7065153339421542,
+      "learning_rate": 2.1547155089456414e-06,
+      "loss": 0.3168,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7020552159822784,
+      "grad_norm": 0.6585356291492391,
+      "learning_rate": 2.1536230168731394e-06,
+      "loss": 0.2939,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7021372605324692,
+      "grad_norm": 0.6750161653423995,
+      "learning_rate": 2.1525307258072047e-06,
+      "loss": 0.3013,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7022193050826598,
+      "grad_norm": 0.6237912163783761,
+      "learning_rate": 2.1514386358249757e-06,
+      "loss": 0.3369,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7023013496328506,
+      "grad_norm": 0.6496107490718095,
+      "learning_rate": 2.1503467470035747e-06,
+      "loss": 0.3152,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7023833941830414,
+      "grad_norm": 0.646157194935363,
+      "learning_rate": 2.149255059420107e-06,
+      "loss": 0.3084,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7024654387332322,
+      "grad_norm": 0.7425241077056324,
+      "learning_rate": 2.1481635731516692e-06,
+      "loss": 0.3064,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7025474832834229,
+      "grad_norm": 0.7584092472321631,
+      "learning_rate": 2.1470722882753353e-06,
+      "loss": 0.3111,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7026295278336137,
+      "grad_norm": 0.6368848972417931,
+      "learning_rate": 2.1459812048681735e-06,
+      "loss": 0.2878,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7027115723838044,
+      "grad_norm": 0.714295060602985,
+      "learning_rate": 2.1448903230072355e-06,
+      "loss": 0.3604,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7027936169339951,
+      "grad_norm": 0.7393786785790648,
+      "learning_rate": 2.1437996427695535e-06,
+      "loss": 0.3261,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7028756614841859,
+      "grad_norm": 0.8928150089549641,
+      "learning_rate": 2.142709164232152e-06,
+      "loss": 0.3391,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7029577060343767,
+      "grad_norm": 0.7013511494444823,
+      "learning_rate": 2.1416188874720404e-06,
+      "loss": 0.32,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7030397505845675,
+      "grad_norm": 0.6503896294082573,
+      "learning_rate": 2.140528812566211e-06,
+      "loss": 0.3006,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7031217951347581,
+      "grad_norm": 0.775166729926355,
+      "learning_rate": 2.1394389395916408e-06,
+      "loss": 0.296,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7032038396849489,
+      "grad_norm": 0.7215355820531719,
+      "learning_rate": 2.1383492686252987e-06,
+      "loss": 0.2984,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7032858842351397,
+      "grad_norm": 0.6847644576726902,
+      "learning_rate": 2.137259799744132e-06,
+      "loss": 0.2802,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7033679287853304,
+      "grad_norm": 0.6889697259182342,
+      "learning_rate": 2.1361705330250804e-06,
+      "loss": 0.3293,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7034499733355212,
+      "grad_norm": 0.6728321408944494,
+      "learning_rate": 2.1350814685450634e-06,
+      "loss": 0.2833,
+      "step": 8574
+    },
+    {
+      "epoch": 0.703532017885712,
+      "grad_norm": 0.7645750363873952,
+      "learning_rate": 2.133992606380991e-06,
+      "loss": 0.2993,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7036140624359027,
+      "grad_norm": 0.6381488662192882,
+      "learning_rate": 2.132903946609756e-06,
+      "loss": 0.3307,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7036961069860934,
+      "grad_norm": 0.6234267222148875,
+      "learning_rate": 2.1318154893082364e-06,
+      "loss": 0.3162,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7037781515362842,
+      "grad_norm": 0.6905260233112792,
+      "learning_rate": 2.130727234553301e-06,
+      "loss": 0.2913,
+      "step": 8578
+    },
+    {
+      "epoch": 0.703860196086475,
+      "grad_norm": 0.6630373067931943,
+      "learning_rate": 2.129639182421797e-06,
+      "loss": 0.3088,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7039422406366657,
+      "grad_norm": 0.6499555416004164,
+      "learning_rate": 2.128551332990564e-06,
+      "loss": 0.3376,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7040242851868564,
+      "grad_norm": 0.6223806653756995,
+      "learning_rate": 2.1274636863364205e-06,
+      "loss": 0.295,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7041063297370472,
+      "grad_norm": 0.7018646338870215,
+      "learning_rate": 2.1263762425361787e-06,
+      "loss": 0.3223,
+      "step": 8582
+    },
+    {
+      "epoch": 0.704188374287238,
+      "grad_norm": 0.6578896978934261,
+      "learning_rate": 2.1252890016666277e-06,
+      "loss": 0.2682,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7042704188374287,
+      "grad_norm": 0.632824383505475,
+      "learning_rate": 2.124201963804551e-06,
+      "loss": 0.2697,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7043524633876195,
+      "grad_norm": 0.683482550054492,
+      "learning_rate": 2.123115129026709e-06,
+      "loss": 0.33,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7044345079378103,
+      "grad_norm": 0.6845386097422029,
+      "learning_rate": 2.1220284974098565e-06,
+      "loss": 0.3386,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7045165524880009,
+      "grad_norm": 0.6936743228858807,
+      "learning_rate": 2.120942069030726e-06,
+      "loss": 0.3211,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7045985970381917,
+      "grad_norm": 0.7075021439256101,
+      "learning_rate": 2.119855843966042e-06,
+      "loss": 0.3039,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7046806415883825,
+      "grad_norm": 0.788423344287218,
+      "learning_rate": 2.11876982229251e-06,
+      "loss": 0.2762,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7047626861385733,
+      "grad_norm": 0.6253524183535496,
+      "learning_rate": 2.117684004086824e-06,
+      "loss": 0.2945,
+      "step": 8590
+    },
+    {
+      "epoch": 0.704844730688764,
+      "grad_norm": 0.719585562886671,
+      "learning_rate": 2.1165983894256647e-06,
+      "loss": 0.3646,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7049267752389548,
+      "grad_norm": 0.6574490280840557,
+      "learning_rate": 2.1155129783856932e-06,
+      "loss": 0.3157,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7050088197891455,
+      "grad_norm": 0.7062569352158091,
+      "learning_rate": 2.1144277710435617e-06,
+      "loss": 0.2851,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7050908643393362,
+      "grad_norm": 0.7274064035696508,
+      "learning_rate": 2.1133427674759037e-06,
+      "loss": 0.3096,
+      "step": 8594
+    },
+    {
+      "epoch": 0.705172908889527,
+      "grad_norm": 0.7072660723551519,
+      "learning_rate": 2.112257967759342e-06,
+      "loss": 0.2882,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7052549534397178,
+      "grad_norm": 0.6647384663281616,
+      "learning_rate": 2.1111733719704845e-06,
+      "loss": 0.2971,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7053369979899086,
+      "grad_norm": 0.7695134364176482,
+      "learning_rate": 2.1100889801859226e-06,
+      "loss": 0.3086,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7054190425400992,
+      "grad_norm": 0.6973748908148096,
+      "learning_rate": 2.1090047924822316e-06,
+      "loss": 0.32,
+      "step": 8598
+    },
+    {
+      "epoch": 0.70550108709029,
+      "grad_norm": 0.7050904764916824,
+      "learning_rate": 2.1079208089359797e-06,
+      "loss": 0.2943,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7055831316404808,
+      "grad_norm": 0.6523975802715528,
+      "learning_rate": 2.1068370296237117e-06,
+      "loss": 0.2749,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7056651761906715,
+      "grad_norm": 0.7179107212436023,
+      "learning_rate": 2.105753454621966e-06,
+      "loss": 0.3291,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7057472207408623,
+      "grad_norm": 0.6513103401148286,
+      "learning_rate": 2.10467008400726e-06,
+      "loss": 0.2828,
+      "step": 8602
+    },
+    {
+      "epoch": 0.705829265291053,
+      "grad_norm": 0.6915826246057426,
+      "learning_rate": 2.103586917856101e-06,
+      "loss": 0.3315,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7059113098412438,
+      "grad_norm": 0.6645930655339649,
+      "learning_rate": 2.1025039562449824e-06,
+      "loss": 0.3229,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7059933543914345,
+      "grad_norm": 0.7391944029637704,
+      "learning_rate": 2.101421199250377e-06,
+      "loss": 0.3175,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7060753989416253,
+      "grad_norm": 0.6701453082421517,
+      "learning_rate": 2.1003386469487513e-06,
+      "loss": 0.3059,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7061574434918161,
+      "grad_norm": 0.6045579160206003,
+      "learning_rate": 2.09925629941655e-06,
+      "loss": 0.3225,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7062394880420068,
+      "grad_norm": 0.7827825975514608,
+      "learning_rate": 2.0981741567302084e-06,
+      "loss": 0.3163,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7063215325921975,
+      "grad_norm": 0.6930913204775817,
+      "learning_rate": 2.0970922189661475e-06,
+      "loss": 0.3031,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7064035771423883,
+      "grad_norm": 0.6337495686533312,
+      "learning_rate": 2.0960104862007706e-06,
+      "loss": 0.2792,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7064856216925791,
+      "grad_norm": 0.6191412334291371,
+      "learning_rate": 2.0949289585104655e-06,
+      "loss": 0.3319,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7065676662427698,
+      "grad_norm": 0.7641686552484275,
+      "learning_rate": 2.0938476359716115e-06,
+      "loss": 0.2816,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7066497107929606,
+      "grad_norm": 0.6727252629073038,
+      "learning_rate": 2.092766518660567e-06,
+      "loss": 0.3068,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7067317553431514,
+      "grad_norm": 1.4520640939228255,
+      "learning_rate": 2.091685606653682e-06,
+      "loss": 0.309,
+      "step": 8614
+    },
+    {
+      "epoch": 0.706813799893342,
+      "grad_norm": 0.7086534737415748,
+      "learning_rate": 2.090604900027284e-06,
+      "loss": 0.3267,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7068958444435328,
+      "grad_norm": 0.6872606021659673,
+      "learning_rate": 2.089524398857697e-06,
+      "loss": 0.2968,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7069778889937236,
+      "grad_norm": 0.7231767888040673,
+      "learning_rate": 2.088444103221218e-06,
+      "loss": 0.3018,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7070599335439144,
+      "grad_norm": 0.6904562247404605,
+      "learning_rate": 2.087364013194139e-06,
+      "loss": 0.6553,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7071419780941051,
+      "grad_norm": 0.6901026159567332,
+      "learning_rate": 2.0862841288527363e-06,
+      "loss": 0.2992,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7072240226442958,
+      "grad_norm": 0.6744698657419953,
+      "learning_rate": 2.085204450273265e-06,
+      "loss": 0.2967,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7073060671944866,
+      "grad_norm": 0.6669800242291101,
+      "learning_rate": 2.0841249775319722e-06,
+      "loss": 0.3266,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7073881117446773,
+      "grad_norm": 0.7015904203779522,
+      "learning_rate": 2.0830457107050904e-06,
+      "loss": 0.3171,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7074701562948681,
+      "grad_norm": 0.6388042671821151,
+      "learning_rate": 2.081966649868833e-06,
+      "loss": 0.2541,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7075522008450589,
+      "grad_norm": 0.684127986464583,
+      "learning_rate": 2.0808877950994037e-06,
+      "loss": 0.3031,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7076342453952497,
+      "grad_norm": 0.703868487044287,
+      "learning_rate": 2.0798091464729886e-06,
+      "loss": 0.3348,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7077162899454403,
+      "grad_norm": 0.6097310766540152,
+      "learning_rate": 2.0787307040657582e-06,
+      "loss": 0.3289,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7077983344956311,
+      "grad_norm": 0.6737255907636763,
+      "learning_rate": 2.077652467953874e-06,
+      "loss": 0.3063,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7078803790458219,
+      "grad_norm": 0.6787429739979552,
+      "learning_rate": 2.076574438213475e-06,
+      "loss": 0.3153,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7079624235960126,
+      "grad_norm": 0.7677446726213017,
+      "learning_rate": 2.0754966149206952e-06,
+      "loss": 0.3157,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7080444681462034,
+      "grad_norm": 0.7663581638404202,
+      "learning_rate": 2.074418998151643e-06,
+      "loss": 0.293,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7081265126963942,
+      "grad_norm": 0.6829683934271861,
+      "learning_rate": 2.0733415879824213e-06,
+      "loss": 0.3343,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7082085572465849,
+      "grad_norm": 0.7480549315936282,
+      "learning_rate": 2.0722643844891167e-06,
+      "loss": 0.3146,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7082906017967756,
+      "grad_norm": 0.6122445800884466,
+      "learning_rate": 2.0711873877477957e-06,
+      "loss": 0.2876,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7083726463469664,
+      "grad_norm": 0.7085462572151328,
+      "learning_rate": 2.070110597834517e-06,
+      "loss": 0.3389,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7084546908971572,
+      "grad_norm": 0.6893541053557867,
+      "learning_rate": 2.0690340148253196e-06,
+      "loss": 0.3409,
+      "step": 8635
+    },
+    {
+      "epoch": 0.708536735447348,
+      "grad_norm": 0.6745415862694574,
+      "learning_rate": 2.0679576387962307e-06,
+      "loss": 0.276,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7086187799975386,
+      "grad_norm": 0.6980706060344365,
+      "learning_rate": 2.066881469823265e-06,
+      "loss": 0.3321,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7087008245477294,
+      "grad_norm": 0.7141504450978707,
+      "learning_rate": 2.0658055079824167e-06,
+      "loss": 0.3199,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7087828690979202,
+      "grad_norm": 0.7035931744296074,
+      "learning_rate": 2.064729753349668e-06,
+      "loss": 0.3051,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7088649136481109,
+      "grad_norm": 0.6499129031372103,
+      "learning_rate": 2.0636542060009897e-06,
+      "loss": 0.2715,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7089469581983017,
+      "grad_norm": 0.6794466306456004,
+      "learning_rate": 2.062578866012332e-06,
+      "loss": 0.3146,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7090290027484925,
+      "grad_norm": 0.7167829474667018,
+      "learning_rate": 2.0615037334596377e-06,
+      "loss": 0.3284,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7091110472986832,
+      "grad_norm": 0.6963038745728191,
+      "learning_rate": 2.060428808418826e-06,
+      "loss": 0.2994,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7091930918488739,
+      "grad_norm": 0.6659537093115858,
+      "learning_rate": 2.05935409096581e-06,
+      "loss": 0.3293,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7092751363990647,
+      "grad_norm": 0.6719761217651491,
+      "learning_rate": 2.058279581176485e-06,
+      "loss": 0.3052,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7093571809492555,
+      "grad_norm": 0.6293109030020089,
+      "learning_rate": 2.057205279126728e-06,
+      "loss": 0.2674,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7094392254994462,
+      "grad_norm": 0.6587291431481554,
+      "learning_rate": 2.0561311848924082e-06,
+      "loss": 0.3511,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7095212700496369,
+      "grad_norm": 0.7303783115476247,
+      "learning_rate": 2.0550572985493724e-06,
+      "loss": 0.3403,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7096033145998277,
+      "grad_norm": 0.663198649420936,
+      "learning_rate": 2.053983620173459e-06,
+      "loss": 0.336,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7096853591500185,
+      "grad_norm": 0.7144309861740779,
+      "learning_rate": 2.0529101498404906e-06,
+      "loss": 0.2837,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7097674037002092,
+      "grad_norm": 0.6648116836064685,
+      "learning_rate": 2.0518368876262712e-06,
+      "loss": 0.3496,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7098494482504,
+      "grad_norm": 0.6552274783472861,
+      "learning_rate": 2.0507638336065953e-06,
+      "loss": 0.3138,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7099314928005908,
+      "grad_norm": 0.8878190199519229,
+      "learning_rate": 2.04969098785724e-06,
+      "loss": 0.3028,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7100135373507814,
+      "grad_norm": 0.6292290673177873,
+      "learning_rate": 2.0486183504539652e-06,
+      "loss": 0.2742,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7100955819009722,
+      "grad_norm": 0.6906193201845551,
+      "learning_rate": 2.0475459214725223e-06,
+      "loss": 0.2819,
+      "step": 8655
+    },
+    {
+      "epoch": 0.710177626451163,
+      "grad_norm": 0.678123085780532,
+      "learning_rate": 2.046473700988641e-06,
+      "loss": 0.3258,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7102596710013538,
+      "grad_norm": 0.6559265506632429,
+      "learning_rate": 2.0454016890780423e-06,
+      "loss": 0.3356,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7103417155515445,
+      "grad_norm": 0.6850375380027687,
+      "learning_rate": 2.044329885816431e-06,
+      "loss": 0.315,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7104237601017352,
+      "grad_norm": 0.6914699283162741,
+      "learning_rate": 2.0432582912794916e-06,
+      "loss": 0.3025,
+      "step": 8659
+    },
+    {
+      "epoch": 0.710505804651926,
+      "grad_norm": 0.6802687249595705,
+      "learning_rate": 2.0421869055429044e-06,
+      "loss": 0.3373,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7105878492021167,
+      "grad_norm": 0.722721890471728,
+      "learning_rate": 2.0411157286823237e-06,
+      "loss": 0.3373,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7106698937523075,
+      "grad_norm": 0.7099993155931762,
+      "learning_rate": 2.0400447607733962e-06,
+      "loss": 0.3004,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7107519383024983,
+      "grad_norm": 0.876898054551594,
+      "learning_rate": 2.038974001891754e-06,
+      "loss": 0.2818,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7108339828526891,
+      "grad_norm": 1.2347318125693012,
+      "learning_rate": 2.0379034521130093e-06,
+      "loss": 0.3145,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7109160274028797,
+      "grad_norm": 0.6385410499012626,
+      "learning_rate": 2.036833111512765e-06,
+      "loss": 0.2851,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7109980719530705,
+      "grad_norm": 0.7102085080115476,
+      "learning_rate": 2.035762980166604e-06,
+      "loss": 0.3082,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7110801165032613,
+      "grad_norm": 0.7533600263936452,
+      "learning_rate": 2.034693058150101e-06,
+      "loss": 0.3128,
+      "step": 8667
+    },
+    {
+      "epoch": 0.711162161053452,
+      "grad_norm": 0.6247757596974288,
+      "learning_rate": 2.0336233455388097e-06,
+      "loss": 0.293,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7112442056036428,
+      "grad_norm": 0.8117502949322851,
+      "learning_rate": 2.03255384240827e-06,
+      "loss": 0.323,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7113262501538336,
+      "grad_norm": 0.6540934059914936,
+      "learning_rate": 2.031484548834013e-06,
+      "loss": 0.3117,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7114082947040243,
+      "grad_norm": 0.7446605484919162,
+      "learning_rate": 2.0304154648915452e-06,
+      "loss": 0.3314,
+      "step": 8671
+    },
+    {
+      "epoch": 0.711490339254215,
+      "grad_norm": 0.6850156051706658,
+      "learning_rate": 2.029346590656367e-06,
+      "loss": 0.2876,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7115723838044058,
+      "grad_norm": 0.6307427336412192,
+      "learning_rate": 2.0282779262039608e-06,
+      "loss": 0.3344,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7116544283545966,
+      "grad_norm": 0.7288090746991917,
+      "learning_rate": 2.0272094716097917e-06,
+      "loss": 0.3224,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7117364729047873,
+      "grad_norm": 0.6428652685926236,
+      "learning_rate": 2.026141226949314e-06,
+      "loss": 0.3062,
+      "step": 8675
+    },
+    {
+      "epoch": 0.711818517454978,
+      "grad_norm": 0.7068907874430505,
+      "learning_rate": 2.025073192297966e-06,
+      "loss": 0.276,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7119005620051688,
+      "grad_norm": 0.7248724682405944,
+      "learning_rate": 2.0240053677311683e-06,
+      "loss": 0.3168,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7119826065553596,
+      "grad_norm": 0.7382768100006186,
+      "learning_rate": 2.0229377533243317e-06,
+      "loss": 0.308,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7120646511055503,
+      "grad_norm": 0.677035662653609,
+      "learning_rate": 2.0218703491528462e-06,
+      "loss": 0.3284,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7121466956557411,
+      "grad_norm": 0.6104823608357483,
+      "learning_rate": 2.020803155292094e-06,
+      "loss": 0.3298,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7122287402059319,
+      "grad_norm": 0.6689857680469772,
+      "learning_rate": 2.0197361718174363e-06,
+      "loss": 0.3155,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7123107847561225,
+      "grad_norm": 0.7152276409855604,
+      "learning_rate": 2.0186693988042205e-06,
+      "loss": 0.2814,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7123928293063133,
+      "grad_norm": 0.7218468617828654,
+      "learning_rate": 2.017602836327783e-06,
+      "loss": 0.3383,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7124748738565041,
+      "grad_norm": 0.7001040040371089,
+      "learning_rate": 2.0165364844634404e-06,
+      "loss": 0.3082,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7125569184066949,
+      "grad_norm": 0.6575820168464382,
+      "learning_rate": 2.015470343286498e-06,
+      "loss": 0.2882,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7126389629568856,
+      "grad_norm": 0.7161489013148323,
+      "learning_rate": 2.014404412872246e-06,
+      "loss": 0.3182,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7127210075070763,
+      "grad_norm": 0.638023874611153,
+      "learning_rate": 2.013338693295957e-06,
+      "loss": 0.3177,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7128030520572671,
+      "grad_norm": 0.7007224798658402,
+      "learning_rate": 2.012273184632892e-06,
+      "loss": 0.282,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7128850966074578,
+      "grad_norm": 0.6854037190226966,
+      "learning_rate": 2.0112078869582928e-06,
+      "loss": 0.3066,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7129671411576486,
+      "grad_norm": 0.7137127874513172,
+      "learning_rate": 2.010142800347391e-06,
+      "loss": 0.2975,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7130491857078394,
+      "grad_norm": 0.8343738300029073,
+      "learning_rate": 2.009077924875403e-06,
+      "loss": 0.3086,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7131312302580302,
+      "grad_norm": 0.6924204660856113,
+      "learning_rate": 2.0080132606175246e-06,
+      "loss": 0.3088,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7132132748082208,
+      "grad_norm": 0.7011752927079282,
+      "learning_rate": 2.0069488076489445e-06,
+      "loss": 0.3329,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7132953193584116,
+      "grad_norm": 0.640779827868074,
+      "learning_rate": 2.005884566044829e-06,
+      "loss": 0.3159,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7133773639086024,
+      "grad_norm": 0.7162119165662145,
+      "learning_rate": 2.0048205358803373e-06,
+      "loss": 0.3455,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7134594084587931,
+      "grad_norm": 0.6472380773934624,
+      "learning_rate": 2.0037567172306066e-06,
+      "loss": 0.308,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7135414530089839,
+      "grad_norm": 0.6770951335438661,
+      "learning_rate": 2.0026931101707615e-06,
+      "loss": 0.3011,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7136234975591746,
+      "grad_norm": 0.6686354215630953,
+      "learning_rate": 2.0016297147759136e-06,
+      "loss": 0.3177,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7137055421093654,
+      "grad_norm": 0.7171758847929613,
+      "learning_rate": 2.0005665311211594e-06,
+      "loss": 0.2776,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7137875866595561,
+      "grad_norm": 0.7161995053915972,
+      "learning_rate": 1.9995035592815753e-06,
+      "loss": 0.3125,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7138696312097469,
+      "grad_norm": 0.6738731762597415,
+      "learning_rate": 1.998440799332232e-06,
+      "loss": 0.279,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7139516757599377,
+      "grad_norm": 0.657572797442084,
+      "learning_rate": 1.9973782513481747e-06,
+      "loss": 0.3076,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7140337203101284,
+      "grad_norm": 0.7017426929913089,
+      "learning_rate": 1.9963159154044414e-06,
+      "loss": 0.2998,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7141157648603191,
+      "grad_norm": 0.6529636914744505,
+      "learning_rate": 1.9952537915760544e-06,
+      "loss": 0.3086,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7141978094105099,
+      "grad_norm": 0.6681122031332561,
+      "learning_rate": 1.994191879938015e-06,
+      "loss": 0.2933,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7142798539607007,
+      "grad_norm": 0.6860049645809341,
+      "learning_rate": 1.993130180565318e-06,
+      "loss": 0.3108,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7143618985108914,
+      "grad_norm": 0.6476169988695959,
+      "learning_rate": 1.9920686935329347e-06,
+      "loss": 0.3017,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7144439430610822,
+      "grad_norm": 0.6572155225668831,
+      "learning_rate": 1.9910074189158295e-06,
+      "loss": 0.2927,
+      "step": 8708
+    },
+    {
+      "epoch": 0.714525987611273,
+      "grad_norm": 0.7628580280745553,
+      "learning_rate": 1.989946356788946e-06,
+      "loss": 0.3477,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7146080321614636,
+      "grad_norm": 0.6860071728193782,
+      "learning_rate": 1.988885507227213e-06,
+      "loss": 0.2732,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7146900767116544,
+      "grad_norm": 0.6996029770844724,
+      "learning_rate": 1.987824870305548e-06,
+      "loss": 0.3391,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7147721212618452,
+      "grad_norm": 0.660160426182279,
+      "learning_rate": 1.9867644460988527e-06,
+      "loss": 0.3119,
+      "step": 8712
+    },
+    {
+      "epoch": 0.714854165812036,
+      "grad_norm": 0.6392902697044213,
+      "learning_rate": 1.9857042346820095e-06,
+      "loss": 0.3331,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7149362103622267,
+      "grad_norm": 0.6745678469968336,
+      "learning_rate": 1.9846442361298916e-06,
+      "loss": 0.318,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7150182549124174,
+      "grad_norm": 0.9354851523593529,
+      "learning_rate": 1.9835844505173526e-06,
+      "loss": 0.2706,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7151002994626082,
+      "grad_norm": 0.6911507397676763,
+      "learning_rate": 1.9825248779192323e-06,
+      "loss": 0.3268,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7151823440127989,
+      "grad_norm": 0.6895240946049852,
+      "learning_rate": 1.981465518410359e-06,
+      "loss": 0.3156,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7152643885629897,
+      "grad_norm": 0.72304840128854,
+      "learning_rate": 1.9804063720655404e-06,
+      "loss": 0.3152,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7153464331131805,
+      "grad_norm": 0.6359872403477259,
+      "learning_rate": 1.9793474389595733e-06,
+      "loss": 0.2636,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7154284776633713,
+      "grad_norm": 0.6843448035964024,
+      "learning_rate": 1.9782887191672357e-06,
+      "loss": 0.2934,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7155105222135619,
+      "grad_norm": 0.673813910342241,
+      "learning_rate": 1.977230212763296e-06,
+      "loss": 0.2875,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7155925667637527,
+      "grad_norm": 0.638958724490516,
+      "learning_rate": 1.9761719198225e-06,
+      "loss": 0.2906,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7156746113139435,
+      "grad_norm": 0.7310878297815753,
+      "learning_rate": 1.975113840419587e-06,
+      "loss": 0.3185,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7157566558641342,
+      "grad_norm": 0.7664561912875966,
+      "learning_rate": 1.9740559746292744e-06,
+      "loss": 0.3168,
+      "step": 8724
+    },
+    {
+      "epoch": 0.715838700414325,
+      "grad_norm": 0.634288471385577,
+      "learning_rate": 1.9729983225262654e-06,
+      "loss": 0.2936,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7159207449645157,
+      "grad_norm": 0.7653694470899749,
+      "learning_rate": 1.971940884185252e-06,
+      "loss": 0.2903,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7160027895147065,
+      "grad_norm": 0.6796593335799707,
+      "learning_rate": 1.97088365968091e-06,
+      "loss": 0.288,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7160848340648972,
+      "grad_norm": 0.7491139810162085,
+      "learning_rate": 1.9698266490878955e-06,
+      "loss": 0.281,
+      "step": 8728
+    },
+    {
+      "epoch": 0.716166878615088,
+      "grad_norm": 0.6520808907302309,
+      "learning_rate": 1.9687698524808547e-06,
+      "loss": 0.2883,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7162489231652788,
+      "grad_norm": 0.6743231590253416,
+      "learning_rate": 1.9677132699344188e-06,
+      "loss": 0.2651,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7163309677154694,
+      "grad_norm": 0.629026936139408,
+      "learning_rate": 1.966656901523198e-06,
+      "loss": 0.3065,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7164130122656602,
+      "grad_norm": 0.6785248266049204,
+      "learning_rate": 1.965600747321795e-06,
+      "loss": 0.3132,
+      "step": 8732
+    },
+    {
+      "epoch": 0.716495056815851,
+      "grad_norm": 0.688124078553897,
+      "learning_rate": 1.9645448074047907e-06,
+      "loss": 0.3075,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7165771013660418,
+      "grad_norm": 0.6889873357072427,
+      "learning_rate": 1.9634890818467573e-06,
+      "loss": 0.3034,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7166591459162325,
+      "grad_norm": 0.728606433632499,
+      "learning_rate": 1.9624335707222443e-06,
+      "loss": 0.2945,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7167411904664233,
+      "grad_norm": 0.7079211978264676,
+      "learning_rate": 1.961378274105794e-06,
+      "loss": 0.3341,
+      "step": 8736
+    },
+    {
+      "epoch": 0.716823235016614,
+      "grad_norm": 0.7233542747930684,
+      "learning_rate": 1.9603231920719283e-06,
+      "loss": 0.2771,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7169052795668047,
+      "grad_norm": 0.7164533425825862,
+      "learning_rate": 1.959268324695153e-06,
+      "loss": 0.3354,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7169873241169955,
+      "grad_norm": 0.68748022090083,
+      "learning_rate": 1.958213672049964e-06,
+      "loss": 0.3173,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7170693686671863,
+      "grad_norm": 0.7181280183541695,
+      "learning_rate": 1.95715923421084e-06,
+      "loss": 0.3336,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7171514132173771,
+      "grad_norm": 0.8573974534019405,
+      "learning_rate": 1.9561050112522395e-06,
+      "loss": 0.2971,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7172334577675678,
+      "grad_norm": 0.7790223370821655,
+      "learning_rate": 1.955051003248615e-06,
+      "loss": 0.3248,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7173155023177585,
+      "grad_norm": 0.6863212878848213,
+      "learning_rate": 1.9539972102743956e-06,
+      "loss": 0.2707,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7173975468679493,
+      "grad_norm": 0.6587152733547417,
+      "learning_rate": 1.9529436324039986e-06,
+      "loss": 0.318,
+      "step": 8744
+    },
+    {
+      "epoch": 0.71747959141814,
+      "grad_norm": 0.7203639571840522,
+      "learning_rate": 1.951890269711829e-06,
+      "loss": 0.3183,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7175616359683308,
+      "grad_norm": 0.6521604234764918,
+      "learning_rate": 1.9508371222722692e-06,
+      "loss": 0.2922,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7176436805185216,
+      "grad_norm": 0.704289090940811,
+      "learning_rate": 1.9497841901596932e-06,
+      "loss": 0.2971,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7177257250687124,
+      "grad_norm": 0.6604199605908143,
+      "learning_rate": 1.9487314734484593e-06,
+      "loss": 0.3074,
+      "step": 8748
+    },
+    {
+      "epoch": 0.717807769618903,
+      "grad_norm": 0.6903273163845,
+      "learning_rate": 1.947678972212906e-06,
+      "loss": 0.2814,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7178898141690938,
+      "grad_norm": 0.6657522473314494,
+      "learning_rate": 1.9466266865273582e-06,
+      "loss": 0.2732,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7179718587192846,
+      "grad_norm": 0.701178558704494,
+      "learning_rate": 1.9455746164661297e-06,
+      "loss": 0.3266,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7180539032694754,
+      "grad_norm": 0.7660913107805865,
+      "learning_rate": 1.944522762103513e-06,
+      "loss": 0.2972,
+      "step": 8752
+    },
+    {
+      "epoch": 0.718135947819666,
+      "grad_norm": 0.6209015876509123,
+      "learning_rate": 1.9434711235137916e-06,
+      "loss": 0.3152,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7182179923698568,
+      "grad_norm": 0.7092796810917408,
+      "learning_rate": 1.9424197007712266e-06,
+      "loss": 0.3183,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7183000369200476,
+      "grad_norm": 0.7252326654755689,
+      "learning_rate": 1.941368493950072e-06,
+      "loss": 0.3158,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7183820814702383,
+      "grad_norm": 0.7715873946787926,
+      "learning_rate": 1.9403175031245587e-06,
+      "loss": 0.275,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7184641260204291,
+      "grad_norm": 1.132520071578745,
+      "learning_rate": 1.9392667283689072e-06,
+      "loss": 0.3053,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7185461705706199,
+      "grad_norm": 0.6698474728264862,
+      "learning_rate": 1.9382161697573237e-06,
+      "loss": 0.2822,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7186282151208107,
+      "grad_norm": 0.7406507509093772,
+      "learning_rate": 1.937165827363994e-06,
+      "loss": 0.3059,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7187102596710013,
+      "grad_norm": 0.6894411225948008,
+      "learning_rate": 1.9361157012630923e-06,
+      "loss": 0.2942,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7187923042211921,
+      "grad_norm": 1.1969129996877022,
+      "learning_rate": 1.935065791528779e-06,
+      "loss": 0.2856,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7188743487713829,
+      "grad_norm": 0.632286798168785,
+      "learning_rate": 1.9340160982351937e-06,
+      "loss": 0.3067,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7189563933215736,
+      "grad_norm": 0.6432015936173333,
+      "learning_rate": 1.932966621456467e-06,
+      "loss": 0.2648,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7190384378717644,
+      "grad_norm": 0.7118834529773037,
+      "learning_rate": 1.93191736126671e-06,
+      "loss": 0.2824,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7191204824219551,
+      "grad_norm": 0.6410752759581183,
+      "learning_rate": 1.9308683177400186e-06,
+      "loss": 0.3038,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7192025269721459,
+      "grad_norm": 1.8205168462918362,
+      "learning_rate": 1.9298194909504767e-06,
+      "loss": 0.316,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7192845715223366,
+      "grad_norm": 0.684275314577548,
+      "learning_rate": 1.9287708809721485e-06,
+      "loss": 0.284,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7193666160725274,
+      "grad_norm": 0.7816438680756649,
+      "learning_rate": 1.9277224878790878e-06,
+      "loss": 0.2787,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7194486606227182,
+      "grad_norm": 0.6854293826266691,
+      "learning_rate": 1.9266743117453274e-06,
+      "loss": 0.3194,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7195307051729088,
+      "grad_norm": 0.7250519662747927,
+      "learning_rate": 1.925626352644889e-06,
+      "loss": 0.2856,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7196127497230996,
+      "grad_norm": 0.6510055363592454,
+      "learning_rate": 1.9245786106517807e-06,
+      "loss": 0.2962,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7196947942732904,
+      "grad_norm": 0.6589763456106594,
+      "learning_rate": 1.923531085839988e-06,
+      "loss": 0.2975,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7197768388234812,
+      "grad_norm": 0.6592441791699473,
+      "learning_rate": 1.922483778283489e-06,
+      "loss": 0.2739,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7198588833736719,
+      "grad_norm": 0.8440492478225594,
+      "learning_rate": 1.921436688056239e-06,
+      "loss": 0.3165,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7199409279238627,
+      "grad_norm": 0.742855498356306,
+      "learning_rate": 1.9203898152321843e-06,
+      "loss": 0.3171,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7200229724740534,
+      "grad_norm": 0.6867595493401953,
+      "learning_rate": 1.919343159885255e-06,
+      "loss": 0.3478,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7201050170242441,
+      "grad_norm": 0.6686506992712807,
+      "learning_rate": 1.9182967220893616e-06,
+      "loss": 0.3137,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7201870615744349,
+      "grad_norm": 0.6435774257645878,
+      "learning_rate": 1.917250501918402e-06,
+      "loss": 0.2907,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7202691061246257,
+      "grad_norm": 0.7749148526204607,
+      "learning_rate": 1.9162044994462596e-06,
+      "loss": 0.3361,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7203511506748165,
+      "grad_norm": 0.6810214077202398,
+      "learning_rate": 1.9151587147468e-06,
+      "loss": 0.3098,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7204331952250072,
+      "grad_norm": 0.6964930425998914,
+      "learning_rate": 1.9141131478938772e-06,
+      "loss": 0.3063,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7205152397751979,
+      "grad_norm": 0.7224259021735384,
+      "learning_rate": 1.9130677989613246e-06,
+      "loss": 0.2793,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7205972843253887,
+      "grad_norm": 0.6525144532151577,
+      "learning_rate": 1.912022668022964e-06,
+      "loss": 0.3158,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7206793288755794,
+      "grad_norm": 0.7436195919808097,
+      "learning_rate": 1.910977755152604e-06,
+      "loss": 0.2779,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7207613734257702,
+      "grad_norm": 0.6706583210356017,
+      "learning_rate": 1.909933060424029e-06,
+      "loss": 0.2821,
+      "step": 8785
+    },
+    {
+      "epoch": 0.720843417975961,
+      "grad_norm": 0.6462332027876059,
+      "learning_rate": 1.9088885839110193e-06,
+      "loss": 0.2963,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7209254625261518,
+      "grad_norm": 0.7037894605398226,
+      "learning_rate": 1.90784432568733e-06,
+      "loss": 0.2698,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7210075070763424,
+      "grad_norm": 0.724560577537143,
+      "learning_rate": 1.906800285826706e-06,
+      "loss": 0.3464,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7210895516265332,
+      "grad_norm": 0.6801639046272174,
+      "learning_rate": 1.9057564644028782e-06,
+      "loss": 0.2959,
+      "step": 8789
+    },
+    {
+      "epoch": 0.721171596176724,
+      "grad_norm": 0.7347541256140651,
+      "learning_rate": 1.9047128614895555e-06,
+      "loss": 0.3193,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7212536407269147,
+      "grad_norm": 0.7298803171739334,
+      "learning_rate": 1.9036694771604397e-06,
+      "loss": 0.2798,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7213356852771055,
+      "grad_norm": 0.7748411029504626,
+      "learning_rate": 1.9026263114892107e-06,
+      "loss": 0.3361,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7214177298272962,
+      "grad_norm": 0.8015120180030773,
+      "learning_rate": 1.9015833645495335e-06,
+      "loss": 0.3447,
+      "step": 8793
+    },
+    {
+      "epoch": 0.721499774377487,
+      "grad_norm": 0.7041721852177341,
+      "learning_rate": 1.900540636415063e-06,
+      "loss": 0.2863,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7215818189276777,
+      "grad_norm": 0.6573191328663264,
+      "learning_rate": 1.899498127159431e-06,
+      "loss": 0.3455,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7216638634778685,
+      "grad_norm": 0.6557364538772836,
+      "learning_rate": 1.89845583685626e-06,
+      "loss": 0.2918,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7217459080280593,
+      "grad_norm": 0.6895814119431082,
+      "learning_rate": 1.897413765579157e-06,
+      "loss": 0.3124,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7218279525782499,
+      "grad_norm": 0.6782863682122618,
+      "learning_rate": 1.8963719134017073e-06,
+      "loss": 0.2775,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7219099971284407,
+      "grad_norm": 0.6879685570018319,
+      "learning_rate": 1.8953302803974882e-06,
+      "loss": 0.2949,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7219920416786315,
+      "grad_norm": 0.7131996602404125,
+      "learning_rate": 1.894288866640055e-06,
+      "loss": 0.2946,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7220740862288223,
+      "grad_norm": 0.6942787193283043,
+      "learning_rate": 1.893247672202952e-06,
+      "loss": 0.332,
+      "step": 8801
+    },
+    {
+      "epoch": 0.722156130779013,
+      "grad_norm": 0.6642793689093893,
+      "learning_rate": 1.8922066971597092e-06,
+      "loss": 0.3175,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7222381753292038,
+      "grad_norm": 0.7020480501135085,
+      "learning_rate": 1.8911659415838346e-06,
+      "loss": 0.301,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7223202198793945,
+      "grad_norm": 0.7275655733368539,
+      "learning_rate": 1.8901254055488283e-06,
+      "loss": 0.3107,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7224022644295852,
+      "grad_norm": 0.6744625258378093,
+      "learning_rate": 1.8890850891281693e-06,
+      "loss": 0.3366,
+      "step": 8805
+    },
+    {
+      "epoch": 0.722484308979776,
+      "grad_norm": 0.6854078898044887,
+      "learning_rate": 1.8880449923953214e-06,
+      "loss": 0.3114,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7225663535299668,
+      "grad_norm": 0.66035487417567,
+      "learning_rate": 1.8870051154237385e-06,
+      "loss": 0.3014,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7226483980801576,
+      "grad_norm": 0.6738764906199928,
+      "learning_rate": 1.8859654582868508e-06,
+      "loss": 0.2733,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7227304426303482,
+      "grad_norm": 0.7079058559144006,
+      "learning_rate": 1.884926021058081e-06,
+      "loss": 0.2863,
+      "step": 8809
+    },
+    {
+      "epoch": 0.722812487180539,
+      "grad_norm": 0.6862166718276848,
+      "learning_rate": 1.8838868038108292e-06,
+      "loss": 0.299,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7228945317307298,
+      "grad_norm": 0.7066066093008425,
+      "learning_rate": 1.8828478066184852e-06,
+      "loss": 0.3156,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7229765762809205,
+      "grad_norm": 0.719283480678453,
+      "learning_rate": 1.8818090295544228e-06,
+      "loss": 0.2898,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7230586208311113,
+      "grad_norm": 0.726069435139811,
+      "learning_rate": 1.880770472691995e-06,
+      "loss": 0.3154,
+      "step": 8813
+    },
+    {
+      "epoch": 0.7231406653813021,
+      "grad_norm": 0.7196329281084355,
+      "learning_rate": 1.8797321361045451e-06,
+      "loss": 0.309,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7232227099314928,
+      "grad_norm": 0.6199155559817806,
+      "learning_rate": 1.8786940198654001e-06,
+      "loss": 0.3324,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7233047544816835,
+      "grad_norm": 0.6772160791625833,
+      "learning_rate": 1.8776561240478674e-06,
+      "loss": 0.3084,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7233867990318743,
+      "grad_norm": 0.6595909659357778,
+      "learning_rate": 1.8766184487252442e-06,
+      "loss": 0.3177,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7234688435820651,
+      "grad_norm": 0.92626414332437,
+      "learning_rate": 1.8755809939708063e-06,
+      "loss": 0.3213,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7235508881322558,
+      "grad_norm": 0.7045815192081347,
+      "learning_rate": 1.8745437598578208e-06,
+      "loss": 0.3065,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7236329326824466,
+      "grad_norm": 0.6899888741322161,
+      "learning_rate": 1.8735067464595336e-06,
+      "loss": 0.2787,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7237149772326373,
+      "grad_norm": 0.6683851997327388,
+      "learning_rate": 1.8724699538491747e-06,
+      "loss": 0.3126,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7237970217828281,
+      "grad_norm": 0.6732871196471715,
+      "learning_rate": 1.8714333820999647e-06,
+      "loss": 0.3083,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7238790663330188,
+      "grad_norm": 0.7209480050052699,
+      "learning_rate": 1.8703970312851016e-06,
+      "loss": 0.295,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7239611108832096,
+      "grad_norm": 0.648852140855837,
+      "learning_rate": 1.8693609014777719e-06,
+      "loss": 0.3002,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7240431554334004,
+      "grad_norm": 0.670943264790055,
+      "learning_rate": 1.8683249927511471e-06,
+      "loss": 0.3117,
+      "step": 8825
+    },
+    {
+      "epoch": 0.724125199983591,
+      "grad_norm": 0.5881688932547685,
+      "learning_rate": 1.867289305178378e-06,
+      "loss": 0.2963,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7242072445337818,
+      "grad_norm": 0.6883807119407828,
+      "learning_rate": 1.8662538388326074e-06,
+      "loss": 0.2829,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7242892890839726,
+      "grad_norm": 0.6867261931041033,
+      "learning_rate": 1.865218593786954e-06,
+      "loss": 0.2672,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7243713336341634,
+      "grad_norm": 0.6898355720169969,
+      "learning_rate": 1.8641835701145267e-06,
+      "loss": 0.3104,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7244533781843541,
+      "grad_norm": 0.8012323758290845,
+      "learning_rate": 1.8631487678884196e-06,
+      "loss": 0.3265,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7245354227345449,
+      "grad_norm": 0.7210025851788807,
+      "learning_rate": 1.862114187181705e-06,
+      "loss": 0.3007,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7246174672847356,
+      "grad_norm": 0.6642923313824369,
+      "learning_rate": 1.8610798280674468e-06,
+      "loss": 0.2949,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7246995118349263,
+      "grad_norm": 0.7176748480234693,
+      "learning_rate": 1.860045690618688e-06,
+      "loss": 0.3256,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7247815563851171,
+      "grad_norm": 0.7618713645007494,
+      "learning_rate": 1.8590117749084562e-06,
+      "loss": 0.3596,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7248636009353079,
+      "grad_norm": 0.6727099422397614,
+      "learning_rate": 1.8579780810097687e-06,
+      "loss": 0.3079,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7249456454854987,
+      "grad_norm": 0.7170220558154219,
+      "learning_rate": 1.8569446089956184e-06,
+      "loss": 0.3098,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7250276900356893,
+      "grad_norm": 0.688475215638465,
+      "learning_rate": 1.855911358938991e-06,
+      "loss": 0.3258,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7251097345858801,
+      "grad_norm": 0.7101931463553803,
+      "learning_rate": 1.854878330912853e-06,
+      "loss": 0.3223,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7251917791360709,
+      "grad_norm": 0.6490739159102257,
+      "learning_rate": 1.8538455249901533e-06,
+      "loss": 0.3055,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7252738236862616,
+      "grad_norm": 0.6356292577143471,
+      "learning_rate": 1.852812941243829e-06,
+      "loss": 0.2649,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7253558682364524,
+      "grad_norm": 0.576820300034229,
+      "learning_rate": 1.851780579746797e-06,
+      "loss": 0.6167,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7254379127866432,
+      "grad_norm": 1.151273974447519,
+      "learning_rate": 1.8507484405719623e-06,
+      "loss": 0.3023,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7255199573368339,
+      "grad_norm": 0.720250226410731,
+      "learning_rate": 1.8497165237922148e-06,
+      "loss": 0.2998,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7256020018870246,
+      "grad_norm": 0.6752643949968303,
+      "learning_rate": 1.8486848294804239e-06,
+      "loss": 0.3377,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7256840464372154,
+      "grad_norm": 0.70824646626698,
+      "learning_rate": 1.8476533577094485e-06,
+      "loss": 0.3051,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7257660909874062,
+      "grad_norm": 0.7173601736217289,
+      "learning_rate": 1.846622108552127e-06,
+      "loss": 0.2908,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7258481355375969,
+      "grad_norm": 0.6838449103432068,
+      "learning_rate": 1.8455910820812872e-06,
+      "loss": 0.2934,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7259301800877876,
+      "grad_norm": 0.6691856123248048,
+      "learning_rate": 1.8445602783697375e-06,
+      "loss": 0.2566,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7260122246379784,
+      "grad_norm": 0.7029712969586496,
+      "learning_rate": 1.84352969749027e-06,
+      "loss": 0.3177,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7260942691881692,
+      "grad_norm": 0.6513278725103478,
+      "learning_rate": 1.8424993395156633e-06,
+      "loss": 0.3055,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7261763137383599,
+      "grad_norm": 0.7263258346916427,
+      "learning_rate": 1.841469204518682e-06,
+      "loss": 0.2892,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7262583582885507,
+      "grad_norm": 0.6714821043164698,
+      "learning_rate": 1.84043929257207e-06,
+      "loss": 0.2947,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7263404028387415,
+      "grad_norm": 1.0108410497668356,
+      "learning_rate": 1.8394096037485604e-06,
+      "loss": 0.3477,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7264224473889321,
+      "grad_norm": 0.6404947943563707,
+      "learning_rate": 1.8383801381208644e-06,
+      "loss": 0.3351,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7265044919391229,
+      "grad_norm": 0.6967441097947858,
+      "learning_rate": 1.8373508957616843e-06,
+      "loss": 0.315,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7265865364893137,
+      "grad_norm": 0.7669770029555887,
+      "learning_rate": 1.836321876743704e-06,
+      "loss": 0.2767,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7266685810395045,
+      "grad_norm": 0.6313170131867067,
+      "learning_rate": 1.835293081139588e-06,
+      "loss": 0.3033,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7267506255896952,
+      "grad_norm": 0.7159420022833884,
+      "learning_rate": 1.8342645090219924e-06,
+      "loss": 0.2883,
+      "step": 8858
+    },
+    {
+      "epoch": 0.726832670139886,
+      "grad_norm": 0.6840793940218085,
+      "learning_rate": 1.8332361604635491e-06,
+      "loss": 0.325,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7269147146900767,
+      "grad_norm": 0.66849172317979,
+      "learning_rate": 1.8322080355368821e-06,
+      "loss": 0.3033,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7269967592402675,
+      "grad_norm": 0.6453845677082082,
+      "learning_rate": 1.831180134314594e-06,
+      "loss": 0.3054,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7270788037904582,
+      "grad_norm": 0.7115284373847901,
+      "learning_rate": 1.830152456869272e-06,
+      "loss": 0.3026,
+      "step": 8862
+    },
+    {
+      "epoch": 0.727160848340649,
+      "grad_norm": 0.6796470649480088,
+      "learning_rate": 1.8291250032734925e-06,
+      "loss": 0.3003,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7272428928908398,
+      "grad_norm": 0.752368135748544,
+      "learning_rate": 1.828097773599809e-06,
+      "loss": 0.2911,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7273249374410304,
+      "grad_norm": 0.6595396116087031,
+      "learning_rate": 1.827070767920765e-06,
+      "loss": 0.334,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7274069819912212,
+      "grad_norm": 0.69135065156498,
+      "learning_rate": 1.8260439863088874e-06,
+      "loss": 0.2705,
+      "step": 8866
+    },
+    {
+      "epoch": 0.727489026541412,
+      "grad_norm": 0.7032325819164614,
+      "learning_rate": 1.8250174288366822e-06,
+      "loss": 0.3486,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7275710710916028,
+      "grad_norm": 0.5885878301203463,
+      "learning_rate": 1.8239910955766454e-06,
+      "loss": 0.2697,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7276531156417935,
+      "grad_norm": 0.6762828059136307,
+      "learning_rate": 1.8229649866012562e-06,
+      "loss": 0.2792,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7277351601919843,
+      "grad_norm": 0.662750488862776,
+      "learning_rate": 1.821939101982974e-06,
+      "loss": 0.3229,
+      "step": 8870
+    },
+    {
+      "epoch": 0.727817204742175,
+      "grad_norm": 0.8036815161052486,
+      "learning_rate": 1.8209134417942481e-06,
+      "loss": 0.3026,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7278992492923657,
+      "grad_norm": 0.7045508618350358,
+      "learning_rate": 1.8198880061075053e-06,
+      "loss": 0.3069,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7279812938425565,
+      "grad_norm": 0.673816806565814,
+      "learning_rate": 1.8188627949951642e-06,
+      "loss": 0.3223,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7280633383927473,
+      "grad_norm": 0.604424077920739,
+      "learning_rate": 1.81783780852962e-06,
+      "loss": 0.2913,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7281453829429381,
+      "grad_norm": 0.6919733695558113,
+      "learning_rate": 1.816813046783259e-06,
+      "loss": 0.332,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7282274274931287,
+      "grad_norm": 0.6404197063407413,
+      "learning_rate": 1.815788509828446e-06,
+      "loss": 0.2793,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7283094720433195,
+      "grad_norm": 0.7311882236607521,
+      "learning_rate": 1.8147641977375313e-06,
+      "loss": 0.3092,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7283915165935103,
+      "grad_norm": 0.6903424810501663,
+      "learning_rate": 1.8137401105828518e-06,
+      "loss": 0.305,
+      "step": 8878
+    },
+    {
+      "epoch": 0.728473561143701,
+      "grad_norm": 0.6863868323309782,
+      "learning_rate": 1.8127162484367277e-06,
+      "loss": 0.3507,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7285556056938918,
+      "grad_norm": 0.6607153036055354,
+      "learning_rate": 1.81169261137146e-06,
+      "loss": 0.2663,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7286376502440826,
+      "grad_norm": 0.7087174897990361,
+      "learning_rate": 1.810669199459339e-06,
+      "loss": 0.3382,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7287196947942733,
+      "grad_norm": 0.6754172499892703,
+      "learning_rate": 1.809646012772634e-06,
+      "loss": 0.2933,
+      "step": 8882
+    },
+    {
+      "epoch": 0.728801739344464,
+      "grad_norm": 0.7126754303822509,
+      "learning_rate": 1.8086230513836011e-06,
+      "loss": 0.3223,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7288837838946548,
+      "grad_norm": 0.6832834666812259,
+      "learning_rate": 1.8076003153644834e-06,
+      "loss": 0.2824,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7289658284448456,
+      "grad_norm": 0.6641371437820992,
+      "learning_rate": 1.8065778047875005e-06,
+      "loss": 0.3303,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7290478729950363,
+      "grad_norm": 0.6633316168850799,
+      "learning_rate": 1.805555519724862e-06,
+      "loss": 0.2874,
+      "step": 8886
+    },
+    {
+      "epoch": 0.729129917545227,
+      "grad_norm": 0.6734861520315042,
+      "learning_rate": 1.8045334602487624e-06,
+      "loss": 0.2818,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7292119620954178,
+      "grad_norm": 0.654402076792195,
+      "learning_rate": 1.8035116264313763e-06,
+      "loss": 0.2929,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7292940066456086,
+      "grad_norm": 0.6836213374338516,
+      "learning_rate": 1.802490018344863e-06,
+      "loss": 0.298,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7293760511957993,
+      "grad_norm": 0.6369187636363243,
+      "learning_rate": 1.8014686360613658e-06,
+      "loss": 0.3219,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7294580957459901,
+      "grad_norm": 0.688213524768224,
+      "learning_rate": 1.8004474796530153e-06,
+      "loss": 0.3244,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7295401402961809,
+      "grad_norm": 0.7308210060859348,
+      "learning_rate": 1.7994265491919245e-06,
+      "loss": 0.3174,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7296221848463715,
+      "grad_norm": 0.7350746576441208,
+      "learning_rate": 1.7984058447501868e-06,
+      "loss": 0.3102,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7297042293965623,
+      "grad_norm": 0.735404779461389,
+      "learning_rate": 1.797385366399887e-06,
+      "loss": 0.3041,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7297862739467531,
+      "grad_norm": 0.684381450203058,
+      "learning_rate": 1.796365114213085e-06,
+      "loss": 0.2814,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7298683184969439,
+      "grad_norm": 0.7836529694760743,
+      "learning_rate": 1.7953450882618317e-06,
+      "loss": 0.3104,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7299503630471346,
+      "grad_norm": 0.7729638970670301,
+      "learning_rate": 1.7943252886181612e-06,
+      "loss": 0.3471,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7300324075973254,
+      "grad_norm": 0.6717011820125501,
+      "learning_rate": 1.7933057153540868e-06,
+      "loss": 0.295,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7301144521475161,
+      "grad_norm": 0.7619550357171175,
+      "learning_rate": 1.7922863685416126e-06,
+      "loss": 0.3566,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7301964966977068,
+      "grad_norm": 0.7363395153225991,
+      "learning_rate": 1.79126724825272e-06,
+      "loss": 0.3232,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7302785412478976,
+      "grad_norm": 0.6647112492675581,
+      "learning_rate": 1.7902483545593807e-06,
+      "loss": 0.3232,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7303605857980884,
+      "grad_norm": 0.7532844624103986,
+      "learning_rate": 1.7892296875335435e-06,
+      "loss": 0.2782,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7304426303482792,
+      "grad_norm": 0.7170199058415745,
+      "learning_rate": 1.7882112472471497e-06,
+      "loss": 0.3261,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7305246748984698,
+      "grad_norm": 0.6442763958926027,
+      "learning_rate": 1.7871930337721155e-06,
+      "loss": 0.3151,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7306067194486606,
+      "grad_norm": 0.6994705146422289,
+      "learning_rate": 1.7861750471803491e-06,
+      "loss": 0.3029,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7306887639988514,
+      "grad_norm": 1.061815164668363,
+      "learning_rate": 1.7851572875437361e-06,
+      "loss": 0.3057,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7307708085490421,
+      "grad_norm": 0.6499582689089072,
+      "learning_rate": 1.7841397549341516e-06,
+      "loss": 0.3464,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7308528530992329,
+      "grad_norm": 0.6848945673492225,
+      "learning_rate": 1.7831224494234495e-06,
+      "loss": 0.3036,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7309348976494237,
+      "grad_norm": 0.6979204182866767,
+      "learning_rate": 1.7821053710834719e-06,
+      "loss": 0.2836,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7310169421996144,
+      "grad_norm": 0.7221067366929906,
+      "learning_rate": 1.7810885199860446e-06,
+      "loss": 0.2944,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7310989867498051,
+      "grad_norm": 0.7049993819394008,
+      "learning_rate": 1.7800718962029728e-06,
+      "loss": 0.295,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7311810312999959,
+      "grad_norm": 0.693329785297274,
+      "learning_rate": 1.779055499806052e-06,
+      "loss": 0.2942,
+      "step": 8912
+    },
+    {
+      "epoch": 0.7312630758501867,
+      "grad_norm": 0.6479530375806514,
+      "learning_rate": 1.7780393308670552e-06,
+      "loss": 0.2853,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7313451204003774,
+      "grad_norm": 0.6904452901421244,
+      "learning_rate": 1.7770233894577448e-06,
+      "loss": 0.3365,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7314271649505681,
+      "grad_norm": 0.6661733277300433,
+      "learning_rate": 1.7760076756498657e-06,
+      "loss": 0.2979,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7315092095007589,
+      "grad_norm": 0.8167327706738282,
+      "learning_rate": 1.774992189515145e-06,
+      "loss": 0.3208,
+      "step": 8916
+    },
+    {
+      "epoch": 0.7315912540509497,
+      "grad_norm": 0.7220385676207921,
+      "learning_rate": 1.7739769311252947e-06,
+      "loss": 0.3107,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7316732986011404,
+      "grad_norm": 0.6396824200105194,
+      "learning_rate": 1.7729619005520087e-06,
+      "loss": 0.2924,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7317553431513312,
+      "grad_norm": 0.7343301429303518,
+      "learning_rate": 1.771947097866969e-06,
+      "loss": 0.2752,
+      "step": 8919
+    },
+    {
+      "epoch": 0.731837387701522,
+      "grad_norm": 0.6880298948833085,
+      "learning_rate": 1.77093252314184e-06,
+      "loss": 0.2803,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7319194322517126,
+      "grad_norm": 0.7414191287463398,
+      "learning_rate": 1.769918176448267e-06,
+      "loss": 0.3172,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7320014768019034,
+      "grad_norm": 0.7070686578963368,
+      "learning_rate": 1.7689040578578825e-06,
+      "loss": 0.3283,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7320835213520942,
+      "grad_norm": 0.691445984996701,
+      "learning_rate": 1.7678901674423044e-06,
+      "loss": 0.3026,
+      "step": 8923
+    },
+    {
+      "epoch": 0.732165565902285,
+      "grad_norm": 0.7313608467639027,
+      "learning_rate": 1.7668765052731285e-06,
+      "loss": 0.2935,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7322476104524757,
+      "grad_norm": 0.7598804432429502,
+      "learning_rate": 1.7658630714219404e-06,
+      "loss": 0.3024,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7323296550026664,
+      "grad_norm": 0.7699039986152302,
+      "learning_rate": 1.7648498659603047e-06,
+      "loss": 0.2745,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7324116995528572,
+      "grad_norm": 0.6646087977043496,
+      "learning_rate": 1.763836888959774e-06,
+      "loss": 0.3045,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7324937441030479,
+      "grad_norm": 0.6772943909103356,
+      "learning_rate": 1.7628241404918845e-06,
+      "loss": 0.3101,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7325757886532387,
+      "grad_norm": 0.717110732598274,
+      "learning_rate": 1.7618116206281533e-06,
+      "loss": 0.3794,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7326578332034295,
+      "grad_norm": 0.8752566589886425,
+      "learning_rate": 1.760799329440081e-06,
+      "loss": 0.2986,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7327398777536203,
+      "grad_norm": 0.697619512163851,
+      "learning_rate": 1.7597872669991578e-06,
+      "loss": 0.2935,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7328219223038109,
+      "grad_norm": 0.723166215594604,
+      "learning_rate": 1.7587754333768504e-06,
+      "loss": 0.3178,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7329039668540017,
+      "grad_norm": 0.6806498271495274,
+      "learning_rate": 1.7577638286446158e-06,
+      "loss": 0.2758,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7329860114041925,
+      "grad_norm": 0.6375739852103074,
+      "learning_rate": 1.756752452873889e-06,
+      "loss": 0.2929,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7330680559543832,
+      "grad_norm": 0.6625624968328604,
+      "learning_rate": 1.755741306136095e-06,
+      "loss": 0.3255,
+      "step": 8935
+    },
+    {
+      "epoch": 0.733150100504574,
+      "grad_norm": 0.6516767951843688,
+      "learning_rate": 1.7547303885026361e-06,
+      "loss": 0.2697,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7332321450547648,
+      "grad_norm": 0.6874143873338475,
+      "learning_rate": 1.753719700044903e-06,
+      "loss": 0.2974,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7333141896049555,
+      "grad_norm": 0.6427651416263919,
+      "learning_rate": 1.7527092408342712e-06,
+      "loss": 0.2978,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7333962341551462,
+      "grad_norm": 0.7193072133496274,
+      "learning_rate": 1.7516990109420934e-06,
+      "loss": 0.3151,
+      "step": 8939
+    },
+    {
+      "epoch": 0.733478278705337,
+      "grad_norm": 0.7257478945672733,
+      "learning_rate": 1.7506890104397134e-06,
+      "loss": 0.3384,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7335603232555278,
+      "grad_norm": 0.8368089288440411,
+      "learning_rate": 1.7496792393984563e-06,
+      "loss": 0.3168,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7336423678057185,
+      "grad_norm": 0.692273552873225,
+      "learning_rate": 1.7486696978896279e-06,
+      "loss": 0.2704,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7337244123559092,
+      "grad_norm": 0.6691674953327194,
+      "learning_rate": 1.7476603859845227e-06,
+      "loss": 0.2698,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7338064569061,
+      "grad_norm": 0.6435502995135963,
+      "learning_rate": 1.7466513037544165e-06,
+      "loss": 0.3032,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7338885014562908,
+      "grad_norm": 0.7893630471725237,
+      "learning_rate": 1.745642451270566e-06,
+      "loss": 0.3487,
+      "step": 8945
+    },
+    {
+      "epoch": 0.7339705460064815,
+      "grad_norm": 1.2463368076762091,
+      "learning_rate": 1.7446338286042196e-06,
+      "loss": 0.276,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7340525905566723,
+      "grad_norm": 0.7804812275390927,
+      "learning_rate": 1.7436254358265997e-06,
+      "loss": 0.2777,
+      "step": 8947
+    },
+    {
+      "epoch": 0.734134635106863,
+      "grad_norm": 0.742250539996497,
+      "learning_rate": 1.742617273008922e-06,
+      "loss": 0.2869,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7342166796570537,
+      "grad_norm": 0.6867341013687283,
+      "learning_rate": 1.7416093402223773e-06,
+      "loss": 0.2692,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7342987242072445,
+      "grad_norm": 0.6848250055449742,
+      "learning_rate": 1.7406016375381452e-06,
+      "loss": 0.2837,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7343807687574353,
+      "grad_norm": 0.7034547194232116,
+      "learning_rate": 1.7395941650273913e-06,
+      "loss": 0.2958,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7344628133076261,
+      "grad_norm": 0.7147571910108416,
+      "learning_rate": 1.7385869227612573e-06,
+      "loss": 0.3139,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7345448578578168,
+      "grad_norm": 0.7400548666269325,
+      "learning_rate": 1.7375799108108749e-06,
+      "loss": 0.334,
+      "step": 8953
+    },
+    {
+      "epoch": 0.7346269024080075,
+      "grad_norm": 0.7084713464067481,
+      "learning_rate": 1.7365731292473597e-06,
+      "loss": 0.3386,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7347089469581983,
+      "grad_norm": 0.6504436962409069,
+      "learning_rate": 1.7355665781418052e-06,
+      "loss": 0.2708,
+      "step": 8955
+    },
+    {
+      "epoch": 0.734790991508389,
+      "grad_norm": 0.7592680454740324,
+      "learning_rate": 1.7345602575652958e-06,
+      "loss": 0.3001,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7348730360585798,
+      "grad_norm": 0.7167414232461782,
+      "learning_rate": 1.7335541675888945e-06,
+      "loss": 0.3033,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7349550806087706,
+      "grad_norm": 0.6916089563364819,
+      "learning_rate": 1.7325483082836487e-06,
+      "loss": 0.2957,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7350371251589614,
+      "grad_norm": 0.7135597541464364,
+      "learning_rate": 1.7315426797205931e-06,
+      "loss": 0.2678,
+      "step": 8959
+    },
+    {
+      "epoch": 0.735119169709152,
+      "grad_norm": 0.6501630956375256,
+      "learning_rate": 1.730537281970741e-06,
+      "loss": 0.2595,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7352012142593428,
+      "grad_norm": 0.6338480386787156,
+      "learning_rate": 1.7295321151050948e-06,
+      "loss": 0.2914,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7352832588095336,
+      "grad_norm": 0.7994384223978753,
+      "learning_rate": 1.7285271791946345e-06,
+      "loss": 0.3293,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7353653033597243,
+      "grad_norm": 0.6830473290270371,
+      "learning_rate": 1.7275224743103287e-06,
+      "loss": 0.3253,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7354473479099151,
+      "grad_norm": 0.7557528314024115,
+      "learning_rate": 1.7265180005231296e-06,
+      "loss": 0.3051,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7355293924601058,
+      "grad_norm": 0.6710443543356633,
+      "learning_rate": 1.7255137579039689e-06,
+      "loss": 0.3169,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7356114370102966,
+      "grad_norm": 0.6888025772896211,
+      "learning_rate": 1.724509746523767e-06,
+      "loss": 0.2839,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7356934815604873,
+      "grad_norm": 0.7697590119075425,
+      "learning_rate": 1.7235059664534226e-06,
+      "loss": 0.2923,
+      "step": 8967
+    },
+    {
+      "epoch": 0.7357755261106781,
+      "grad_norm": 0.8336188151826045,
+      "learning_rate": 1.7225024177638234e-06,
+      "loss": 0.3128,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7358575706608689,
+      "grad_norm": 0.6347785848017541,
+      "learning_rate": 1.7214991005258386e-06,
+      "loss": 0.2845,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7359396152110596,
+      "grad_norm": 0.5588556452840895,
+      "learning_rate": 1.720496014810319e-06,
+      "loss": 0.6286,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7360216597612503,
+      "grad_norm": 0.7183268785900047,
+      "learning_rate": 1.7194931606881033e-06,
+      "loss": 0.2659,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7361037043114411,
+      "grad_norm": 0.6543951534712729,
+      "learning_rate": 1.7184905382300098e-06,
+      "loss": 0.2702,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7361857488616319,
+      "grad_norm": 0.7225607439981256,
+      "learning_rate": 1.7174881475068412e-06,
+      "loss": 0.2899,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7362677934118226,
+      "grad_norm": 0.794948771972903,
+      "learning_rate": 1.7164859885893875e-06,
+      "loss": 0.3077,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7363498379620134,
+      "grad_norm": 0.6813247503058396,
+      "learning_rate": 1.715484061548416e-06,
+      "loss": 0.3069,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7364318825122042,
+      "grad_norm": 0.75262455094905,
+      "learning_rate": 1.7144823664546829e-06,
+      "loss": 0.3193,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7365139270623949,
+      "grad_norm": 0.7053356726146659,
+      "learning_rate": 1.7134809033789285e-06,
+      "loss": 0.3285,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7365959716125856,
+      "grad_norm": 0.6587535464649084,
+      "learning_rate": 1.7124796723918708e-06,
+      "loss": 0.2979,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7366780161627764,
+      "grad_norm": 0.9174572515908077,
+      "learning_rate": 1.711478673564218e-06,
+      "loss": 0.2977,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7367600607129672,
+      "grad_norm": 0.6728239703671753,
+      "learning_rate": 1.7104779069666565e-06,
+      "loss": 0.2702,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.7129829182312416,
+      "learning_rate": 1.7094773726698605e-06,
+      "loss": 0.2901,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7369241498133486,
+      "grad_norm": 0.8046584577556191,
+      "learning_rate": 1.7084770707444876e-06,
+      "loss": 0.3488,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7370061943635394,
+      "grad_norm": 0.7243545647415289,
+      "learning_rate": 1.7074770012611736e-06,
+      "loss": 0.2948,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7370882389137302,
+      "grad_norm": 0.780839927326685,
+      "learning_rate": 1.706477164290546e-06,
+      "loss": 0.3312,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7371702834639209,
+      "grad_norm": 0.7429924567306323,
+      "learning_rate": 1.7054775599032093e-06,
+      "loss": 0.3302,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7372523280141117,
+      "grad_norm": 0.7682139587949723,
+      "learning_rate": 1.7044781881697526e-06,
+      "loss": 0.3101,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7373343725643025,
+      "grad_norm": 0.7167572322321712,
+      "learning_rate": 1.7034790491607533e-06,
+      "loss": 0.2977,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7374164171144931,
+      "grad_norm": 0.6640243546463714,
+      "learning_rate": 1.7024801429467653e-06,
+      "loss": 0.2996,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7374984616646839,
+      "grad_norm": 0.6545899786588583,
+      "learning_rate": 1.7014814695983324e-06,
+      "loss": 0.3297,
+      "step": 8989
+    },
+    {
+      "epoch": 0.7375805062148747,
+      "grad_norm": 0.6700403408978004,
+      "learning_rate": 1.7004830291859797e-06,
+      "loss": 0.2844,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7376625507650655,
+      "grad_norm": 0.661776999602463,
+      "learning_rate": 1.699484821780213e-06,
+      "loss": 0.3215,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7377445953152562,
+      "grad_norm": 0.6081439294094818,
+      "learning_rate": 1.698486847451527e-06,
+      "loss": 0.276,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7378266398654469,
+      "grad_norm": 0.7495937262079024,
+      "learning_rate": 1.697489106270394e-06,
+      "loss": 0.2995,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7379086844156377,
+      "grad_norm": 0.6192607642735171,
+      "learning_rate": 1.696491598307275e-06,
+      "loss": 0.3206,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7379907289658284,
+      "grad_norm": 0.7284624271437059,
+      "learning_rate": 1.6954943236326127e-06,
+      "loss": 0.2961,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7380727735160192,
+      "grad_norm": 0.6603833228740623,
+      "learning_rate": 1.694497282316831e-06,
+      "loss": 0.3212,
+      "step": 8996
+    },
+    {
+      "epoch": 0.73815481806621,
+      "grad_norm": 0.6337974809722191,
+      "learning_rate": 1.6935004744303419e-06,
+      "loss": 0.2835,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7382368626164008,
+      "grad_norm": 0.6747330689742984,
+      "learning_rate": 1.6925039000435356e-06,
+      "loss": 0.316,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7383189071665914,
+      "grad_norm": 0.6682804784636275,
+      "learning_rate": 1.691507559226791e-06,
+      "loss": 0.2917,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7384009517167822,
+      "grad_norm": 0.6691646917769626,
+      "learning_rate": 1.690511452050468e-06,
+      "loss": 0.3171,
+      "step": 9000
+    },
+    {
+      "epoch": 0.738482996266973,
+      "grad_norm": 0.6693448218331087,
+      "learning_rate": 1.6895155785849066e-06,
+      "loss": 0.2842,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7385650408171637,
+      "grad_norm": 0.7286526513964698,
+      "learning_rate": 1.6885199389004387e-06,
+      "loss": 0.287,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7386470853673545,
+      "grad_norm": 0.6859951615063671,
+      "learning_rate": 1.6875245330673701e-06,
+      "loss": 0.302,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7387291299175452,
+      "grad_norm": 0.6861150342408416,
+      "learning_rate": 1.686529361155998e-06,
+      "loss": 0.279,
+      "step": 9004
+    },
+    {
+      "epoch": 0.738811174467736,
+      "grad_norm": 0.7437497223916295,
+      "learning_rate": 1.6855344232366e-06,
+      "loss": 0.2809,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7388932190179267,
+      "grad_norm": 0.6820149275668013,
+      "learning_rate": 1.6845397193794338e-06,
+      "loss": 0.3023,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7389752635681175,
+      "grad_norm": 0.6712613527344765,
+      "learning_rate": 1.6835452496547467e-06,
+      "loss": 0.3067,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7390573081183083,
+      "grad_norm": 0.6617028263513157,
+      "learning_rate": 1.6825510141327673e-06,
+      "loss": 0.2918,
+      "step": 9008
+    },
+    {
+      "epoch": 0.739139352668499,
+      "grad_norm": 0.7340516303826488,
+      "learning_rate": 1.6815570128837034e-06,
+      "loss": 0.2908,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7392213972186897,
+      "grad_norm": 0.6883543136285418,
+      "learning_rate": 1.6805632459777539e-06,
+      "loss": 0.3361,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7393034417688805,
+      "grad_norm": 0.7268936700782411,
+      "learning_rate": 1.6795697134850925e-06,
+      "loss": 0.2991,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7393854863190713,
+      "grad_norm": 0.6684433573046916,
+      "learning_rate": 1.6785764154758856e-06,
+      "loss": 0.2994,
+      "step": 9012
+    },
+    {
+      "epoch": 0.739467530869262,
+      "grad_norm": 0.6656360420262505,
+      "learning_rate": 1.6775833520202756e-06,
+      "loss": 0.2698,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7395495754194528,
+      "grad_norm": 0.68109686522161,
+      "learning_rate": 1.67659052318839e-06,
+      "loss": 0.3117,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7396316199696436,
+      "grad_norm": 0.654135378793226,
+      "learning_rate": 1.6755979290503437e-06,
+      "loss": 0.2954,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7397136645198342,
+      "grad_norm": 0.7649677015364339,
+      "learning_rate": 1.6746055696762292e-06,
+      "loss": 0.3263,
+      "step": 9016
+    },
+    {
+      "epoch": 0.739795709070025,
+      "grad_norm": 0.6597918718731217,
+      "learning_rate": 1.6736134451361264e-06,
+      "loss": 0.3162,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7398777536202158,
+      "grad_norm": 0.72792292049969,
+      "learning_rate": 1.6726215555000997e-06,
+      "loss": 0.3238,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7399597981704066,
+      "grad_norm": 0.6711972828716992,
+      "learning_rate": 1.6716299008381908e-06,
+      "loss": 0.2826,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7400418427205973,
+      "grad_norm": 0.6949884953702924,
+      "learning_rate": 1.670638481220433e-06,
+      "loss": 0.2973,
+      "step": 9020
+    },
+    {
+      "epoch": 0.740123887270788,
+      "grad_norm": 0.6754048676465267,
+      "learning_rate": 1.6696472967168348e-06,
+      "loss": 0.323,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7402059318209788,
+      "grad_norm": 0.7436001414419489,
+      "learning_rate": 1.6686563473973937e-06,
+      "loss": 0.32,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7402879763711695,
+      "grad_norm": 0.7062226280030225,
+      "learning_rate": 1.6676656333320912e-06,
+      "loss": 0.3131,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7403700209213603,
+      "grad_norm": 0.6823053108515451,
+      "learning_rate": 1.666675154590886e-06,
+      "loss": 0.3152,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7404520654715511,
+      "grad_norm": 0.6805744809170506,
+      "learning_rate": 1.6656849112437278e-06,
+      "loss": 0.3385,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7405341100217419,
+      "grad_norm": 0.7000163162123353,
+      "learning_rate": 1.6646949033605425e-06,
+      "loss": 0.3227,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7406161545719325,
+      "grad_norm": 0.7023632559256614,
+      "learning_rate": 1.6637051310112462e-06,
+      "loss": 0.299,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7406981991221233,
+      "grad_norm": 0.7470093413941307,
+      "learning_rate": 1.662715594265733e-06,
+      "loss": 0.3062,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7407802436723141,
+      "grad_norm": 0.7117175104360663,
+      "learning_rate": 1.6617262931938815e-06,
+      "loss": 0.2716,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7408622882225048,
+      "grad_norm": 0.6106436893754139,
+      "learning_rate": 1.660737227865556e-06,
+      "loss": 0.2849,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7409443327726956,
+      "grad_norm": 0.6950027555523081,
+      "learning_rate": 1.6597483983506034e-06,
+      "loss": 0.3308,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7410263773228863,
+      "grad_norm": 0.6920913257190062,
+      "learning_rate": 1.6587598047188514e-06,
+      "loss": 0.3164,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7411084218730771,
+      "grad_norm": 0.664407442500741,
+      "learning_rate": 1.6577714470401158e-06,
+      "loss": 0.2828,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7411904664232678,
+      "grad_norm": 0.6637088553775715,
+      "learning_rate": 1.6567833253841887e-06,
+      "loss": 0.3127,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7412725109734586,
+      "grad_norm": 0.6408015633200501,
+      "learning_rate": 1.655795439820852e-06,
+      "loss": 0.2963,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7413545555236494,
+      "grad_norm": 0.7681641666305644,
+      "learning_rate": 1.6548077904198701e-06,
+      "loss": 0.2745,
+      "step": 9036
+    },
+    {
+      "epoch": 0.74143660007384,
+      "grad_norm": 0.716099262657564,
+      "learning_rate": 1.6538203772509859e-06,
+      "loss": 0.2798,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7415186446240308,
+      "grad_norm": 0.666989265656943,
+      "learning_rate": 1.6528332003839325e-06,
+      "loss": 0.3099,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7416006891742216,
+      "grad_norm": 0.8014097227963686,
+      "learning_rate": 1.6518462598884188e-06,
+      "loss": 0.3356,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7416827337244124,
+      "grad_norm": 0.6874789384405119,
+      "learning_rate": 1.650859555834145e-06,
+      "loss": 0.2803,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7417647782746031,
+      "grad_norm": 0.6474085209659657,
+      "learning_rate": 1.6498730882907882e-06,
+      "loss": 0.3006,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7418468228247939,
+      "grad_norm": 0.6574283308067319,
+      "learning_rate": 1.6488868573280104e-06,
+      "loss": 0.3083,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7419288673749846,
+      "grad_norm": 0.7212466880378862,
+      "learning_rate": 1.6479008630154585e-06,
+      "loss": 0.305,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7420109119251753,
+      "grad_norm": 0.73971582578225,
+      "learning_rate": 1.6469151054227638e-06,
+      "loss": 0.3089,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7420929564753661,
+      "grad_norm": 0.7088399068870913,
+      "learning_rate": 1.6459295846195355e-06,
+      "loss": 0.3056,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7421750010255569,
+      "grad_norm": 0.6880613062290395,
+      "learning_rate": 1.644944300675373e-06,
+      "loss": 0.3123,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7422570455757477,
+      "grad_norm": 0.6840190205935308,
+      "learning_rate": 1.6439592536598515e-06,
+      "loss": 0.2876,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7423390901259384,
+      "grad_norm": 0.7338737151358825,
+      "learning_rate": 1.642974443642536e-06,
+      "loss": 0.2641,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7424211346761291,
+      "grad_norm": 0.6698407893525709,
+      "learning_rate": 1.6419898706929731e-06,
+      "loss": 0.3392,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7425031792263199,
+      "grad_norm": 0.7885291647291128,
+      "learning_rate": 1.6410055348806898e-06,
+      "loss": 0.3305,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7425852237765106,
+      "grad_norm": 0.7070536118361729,
+      "learning_rate": 1.6400214362751997e-06,
+      "loss": 0.2924,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7426672683267014,
+      "grad_norm": 0.7108653088893122,
+      "learning_rate": 1.6390375749459963e-06,
+      "loss": 0.2873,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7427493128768922,
+      "grad_norm": 0.6730435541998144,
+      "learning_rate": 1.638053950962561e-06,
+      "loss": 0.3028,
+      "step": 9053
+    },
+    {
+      "epoch": 0.742831357427083,
+      "grad_norm": 0.6995730457940071,
+      "learning_rate": 1.637070564394353e-06,
+      "loss": 0.3013,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7429134019772736,
+      "grad_norm": 0.6617124658625,
+      "learning_rate": 1.6360874153108203e-06,
+      "loss": 0.259,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7429954465274644,
+      "grad_norm": 0.7101098321956643,
+      "learning_rate": 1.6351045037813895e-06,
+      "loss": 0.2898,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7430774910776552,
+      "grad_norm": 1.1596632965734603,
+      "learning_rate": 1.6341218298754713e-06,
+      "loss": 0.3137,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7431595356278459,
+      "grad_norm": 0.6895124250304071,
+      "learning_rate": 1.6331393936624612e-06,
+      "loss": 0.296,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7432415801780367,
+      "grad_norm": 0.7193980938670601,
+      "learning_rate": 1.6321571952117399e-06,
+      "loss": 0.3094,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7433236247282274,
+      "grad_norm": 0.7389389889552638,
+      "learning_rate": 1.631175234592665e-06,
+      "loss": 0.2791,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7434056692784182,
+      "grad_norm": 0.657766107814324,
+      "learning_rate": 1.6301935118745826e-06,
+      "loss": 0.3114,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7434877138286089,
+      "grad_norm": 0.6327310497753086,
+      "learning_rate": 1.629212027126822e-06,
+      "loss": 0.3239,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7435697583787997,
+      "grad_norm": 0.6610957480588027,
+      "learning_rate": 1.628230780418691e-06,
+      "loss": 0.3223,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7436518029289905,
+      "grad_norm": 0.7037077924357004,
+      "learning_rate": 1.6272497718194863e-06,
+      "loss": 0.2825,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7437338474791811,
+      "grad_norm": 0.639935767389984,
+      "learning_rate": 1.6262690013984827e-06,
+      "loss": 0.3086,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7438158920293719,
+      "grad_norm": 0.6922715914259635,
+      "learning_rate": 1.6252884692249416e-06,
+      "loss": 0.3214,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7438979365795627,
+      "grad_norm": 0.6428565356409127,
+      "learning_rate": 1.6243081753681088e-06,
+      "loss": 0.2684,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7439799811297535,
+      "grad_norm": 0.7023888436896288,
+      "learning_rate": 1.6233281198972095e-06,
+      "loss": 0.3274,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7440620256799442,
+      "grad_norm": 0.6829619565194667,
+      "learning_rate": 1.622348302881453e-06,
+      "loss": 0.3132,
+      "step": 9069
+    },
+    {
+      "epoch": 0.744144070230135,
+      "grad_norm": 0.7811628012862532,
+      "learning_rate": 1.6213687243900306e-06,
+      "loss": 0.319,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7442261147803257,
+      "grad_norm": 0.7735306476549021,
+      "learning_rate": 1.6203893844921215e-06,
+      "loss": 0.3003,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7443081593305164,
+      "grad_norm": 0.7287048584364672,
+      "learning_rate": 1.6194102832568852e-06,
+      "loss": 0.2874,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7443902038807072,
+      "grad_norm": 0.7999219326742384,
+      "learning_rate": 1.6184314207534624e-06,
+      "loss": 0.3279,
+      "step": 9073
+    },
+    {
+      "epoch": 0.744472248430898,
+      "grad_norm": 0.6848472309301447,
+      "learning_rate": 1.6174527970509812e-06,
+      "loss": 0.2949,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7445542929810888,
+      "grad_norm": 0.7135893622328953,
+      "learning_rate": 1.6164744122185472e-06,
+      "loss": 0.3241,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7446363375312794,
+      "grad_norm": 0.7029037648166645,
+      "learning_rate": 1.6154962663252543e-06,
+      "loss": 0.29,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7447183820814702,
+      "grad_norm": 0.6356269516502361,
+      "learning_rate": 1.6145183594401793e-06,
+      "loss": 0.3,
+      "step": 9077
+    },
+    {
+      "epoch": 0.744800426631661,
+      "grad_norm": 0.6321280015944296,
+      "learning_rate": 1.6135406916323764e-06,
+      "loss": 0.2687,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7448824711818517,
+      "grad_norm": 0.6959042801415447,
+      "learning_rate": 1.6125632629708892e-06,
+      "loss": 0.2869,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7449645157320425,
+      "grad_norm": 0.7464582286894883,
+      "learning_rate": 1.611586073524744e-06,
+      "loss": 0.3254,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7450465602822333,
+      "grad_norm": 0.7316517299580773,
+      "learning_rate": 1.6106091233629462e-06,
+      "loss": 0.2801,
+      "step": 9081
+    },
+    {
+      "epoch": 0.745128604832424,
+      "grad_norm": 0.6623635947431125,
+      "learning_rate": 1.6096324125544854e-06,
+      "loss": 0.3219,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7452106493826147,
+      "grad_norm": 0.6838392583825101,
+      "learning_rate": 1.6086559411683377e-06,
+      "loss": 0.2839,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7452926939328055,
+      "grad_norm": 0.6617853389863254,
+      "learning_rate": 1.6076797092734575e-06,
+      "loss": 0.3036,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7453747384829963,
+      "grad_norm": 0.7252353481745579,
+      "learning_rate": 1.6067037169387873e-06,
+      "loss": 0.3375,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7454567830331871,
+      "grad_norm": 0.6463708449647348,
+      "learning_rate": 1.6057279642332474e-06,
+      "loss": 0.3005,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7455388275833778,
+      "grad_norm": 0.6820459681107165,
+      "learning_rate": 1.6047524512257473e-06,
+      "loss": 0.288,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7456208721335685,
+      "grad_norm": 0.6699024414363287,
+      "learning_rate": 1.6037771779851725e-06,
+      "loss": 0.3233,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7457029166837593,
+      "grad_norm": 0.6777868346103995,
+      "learning_rate": 1.6028021445803966e-06,
+      "loss": 0.277,
+      "step": 9089
+    },
+    {
+      "epoch": 0.74578496123395,
+      "grad_norm": 0.7163815163750522,
+      "learning_rate": 1.6018273510802768e-06,
+      "loss": 0.3053,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7458670057841408,
+      "grad_norm": 0.7725826844059878,
+      "learning_rate": 1.6008527975536487e-06,
+      "loss": 0.3098,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7459490503343316,
+      "grad_norm": 0.7302938381148406,
+      "learning_rate": 1.599878484069336e-06,
+      "loss": 0.3266,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7460310948845223,
+      "grad_norm": 0.7460955195747975,
+      "learning_rate": 1.598904410696141e-06,
+      "loss": 0.2938,
+      "step": 9093
+    },
+    {
+      "epoch": 0.746113139434713,
+      "grad_norm": 0.7715438324768245,
+      "learning_rate": 1.597930577502852e-06,
+      "loss": 0.3311,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7461951839849038,
+      "grad_norm": 0.6516948515067847,
+      "learning_rate": 1.596956984558241e-06,
+      "loss": 0.3054,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7462772285350946,
+      "grad_norm": 0.7050684527006341,
+      "learning_rate": 1.5959836319310612e-06,
+      "loss": 0.2943,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7463592730852853,
+      "grad_norm": 0.8172788034452727,
+      "learning_rate": 1.595010519690046e-06,
+      "loss": 0.2795,
+      "step": 9097
+    },
+    {
+      "epoch": 0.746441317635476,
+      "grad_norm": 0.6636126186161512,
+      "learning_rate": 1.5940376479039195e-06,
+      "loss": 0.3427,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7465233621856668,
+      "grad_norm": 0.6788968251317297,
+      "learning_rate": 1.5930650166413803e-06,
+      "loss": 0.2735,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7466054067358576,
+      "grad_norm": 0.6733406646098854,
+      "learning_rate": 1.5920926259711173e-06,
+      "loss": 0.3065,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7466874512860483,
+      "grad_norm": 0.7055709475257307,
+      "learning_rate": 1.5911204759617965e-06,
+      "loss": 0.2967,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7467694958362391,
+      "grad_norm": 0.6426294599051983,
+      "learning_rate": 1.590148566682071e-06,
+      "loss": 0.2851,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7468515403864299,
+      "grad_norm": 0.7890754221302223,
+      "learning_rate": 1.5891768982005768e-06,
+      "loss": 0.2807,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7469335849366205,
+      "grad_norm": 0.6381561749967412,
+      "learning_rate": 1.588205470585929e-06,
+      "loss": 0.3681,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7470156294868113,
+      "grad_norm": 0.8827305939835662,
+      "learning_rate": 1.5872342839067305e-06,
+      "loss": 0.2941,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7470976740370021,
+      "grad_norm": 0.6961776691810667,
+      "learning_rate": 1.5862633382315622e-06,
+      "loss": 0.3189,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7471797185871929,
+      "grad_norm": 0.6687360181074119,
+      "learning_rate": 1.5852926336289926e-06,
+      "loss": 0.3217,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7472617631373836,
+      "grad_norm": 0.6920541067043184,
+      "learning_rate": 1.5843221701675725e-06,
+      "loss": 0.2934,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7473438076875744,
+      "grad_norm": 0.7193549224101764,
+      "learning_rate": 1.5833519479158332e-06,
+      "loss": 0.2742,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7474258522377651,
+      "grad_norm": 0.6981266211374404,
+      "learning_rate": 1.5823819669422885e-06,
+      "loss": 0.3075,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7475078967879558,
+      "grad_norm": 0.6577638170988356,
+      "learning_rate": 1.5814122273154404e-06,
+      "loss": 0.315,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7475899413381466,
+      "grad_norm": 0.6672922898165287,
+      "learning_rate": 1.580442729103766e-06,
+      "loss": 0.2608,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7476719858883374,
+      "grad_norm": 0.6890477040278038,
+      "learning_rate": 1.5794734723757343e-06,
+      "loss": 0.2981,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7477540304385282,
+      "grad_norm": 0.6868272641599467,
+      "learning_rate": 1.5785044571997888e-06,
+      "loss": 0.3159,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7478360749887188,
+      "grad_norm": 0.6698799138406278,
+      "learning_rate": 1.5775356836443611e-06,
+      "loss": 0.2923,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7479181195389096,
+      "grad_norm": 0.7404772768838462,
+      "learning_rate": 1.5765671517778668e-06,
+      "loss": 0.3282,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7480001640891004,
+      "grad_norm": 0.7378404966605135,
+      "learning_rate": 1.5755988616686984e-06,
+      "loss": 0.3242,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7480822086392911,
+      "grad_norm": 0.6980366553636397,
+      "learning_rate": 1.5746308133852384e-06,
+      "loss": 0.3083,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7481642531894819,
+      "grad_norm": 0.6466009930587999,
+      "learning_rate": 1.5736630069958453e-06,
+      "loss": 0.303,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7482462977396727,
+      "grad_norm": 0.6567802313823907,
+      "learning_rate": 1.5726954425688662e-06,
+      "loss": 0.3549,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7483283422898634,
+      "grad_norm": 0.7170366919959754,
+      "learning_rate": 1.5717281201726298e-06,
+      "loss": 0.2957,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7484103868400541,
+      "grad_norm": 0.7037708334887701,
+      "learning_rate": 1.5707610398754442e-06,
+      "loss": 0.299,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7484924313902449,
+      "grad_norm": 0.6572472390398075,
+      "learning_rate": 1.5697942017456059e-06,
+      "loss": 0.3175,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7485744759404357,
+      "grad_norm": 0.7696857817497345,
+      "learning_rate": 1.5688276058513897e-06,
+      "loss": 0.296,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7486565204906264,
+      "grad_norm": 0.657748095710143,
+      "learning_rate": 1.5678612522610542e-06,
+      "loss": 0.2717,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7487385650408172,
+      "grad_norm": 0.6299949777815126,
+      "learning_rate": 1.5668951410428447e-06,
+      "loss": 0.2988,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7488206095910079,
+      "grad_norm": 0.6419552673765891,
+      "learning_rate": 1.5659292722649826e-06,
+      "loss": 0.2738,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7489026541411987,
+      "grad_norm": 0.6809964567822552,
+      "learning_rate": 1.5649636459956796e-06,
+      "loss": 0.3521,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7489846986913894,
+      "grad_norm": 0.6799680960985561,
+      "learning_rate": 1.563998262303124e-06,
+      "loss": 0.2921,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7490667432415802,
+      "grad_norm": 0.6376174101864279,
+      "learning_rate": 1.5630331212554906e-06,
+      "loss": 0.3103,
+      "step": 9130
+    },
+    {
+      "epoch": 0.749148787791771,
+      "grad_norm": 0.7094791927007649,
+      "learning_rate": 1.5620682229209378e-06,
+      "loss": 0.2923,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7492308323419616,
+      "grad_norm": 0.6719685136341286,
+      "learning_rate": 1.5611035673676018e-06,
+      "loss": 0.3147,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7493128768921524,
+      "grad_norm": 0.7136717596468931,
+      "learning_rate": 1.560139154663607e-06,
+      "loss": 0.3013,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7493949214423432,
+      "grad_norm": 0.8333775271396712,
+      "learning_rate": 1.5591749848770604e-06,
+      "loss": 0.3401,
+      "step": 9134
+    },
+    {
+      "epoch": 0.749476965992534,
+      "grad_norm": 0.65425805564909,
+      "learning_rate": 1.5582110580760462e-06,
+      "loss": 0.2878,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7495590105427247,
+      "grad_norm": 0.7290318435649271,
+      "learning_rate": 1.5572473743286393e-06,
+      "loss": 0.324,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7496410550929155,
+      "grad_norm": 0.7011110421180631,
+      "learning_rate": 1.5562839337028917e-06,
+      "loss": 0.2935,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7497230996431062,
+      "grad_norm": 0.6967666027685071,
+      "learning_rate": 1.5553207362668383e-06,
+      "loss": 0.3613,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7498051441932969,
+      "grad_norm": 0.7275714332667291,
+      "learning_rate": 1.5543577820885014e-06,
+      "loss": 0.298,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7498871887434877,
+      "grad_norm": 0.6880846212841614,
+      "learning_rate": 1.553395071235881e-06,
+      "loss": 0.2626,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7499692332936785,
+      "grad_norm": 0.7519216195318449,
+      "learning_rate": 1.5524326037769649e-06,
+      "loss": 0.2985,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7500512778438693,
+      "grad_norm": 0.6984129768591841,
+      "learning_rate": 1.5514703797797177e-06,
+      "loss": 0.2964,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7501333223940599,
+      "grad_norm": 0.6465152585095775,
+      "learning_rate": 1.5505083993120917e-06,
+      "loss": 0.3062,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7502153669442507,
+      "grad_norm": 0.6880619507049293,
+      "learning_rate": 1.5495466624420218e-06,
+      "loss": 0.3032,
+      "step": 9144
+    },
+    {
+      "epoch": 0.7502974114944415,
+      "grad_norm": 0.67201332833827,
+      "learning_rate": 1.548585169237422e-06,
+      "loss": 0.3225,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7503794560446322,
+      "grad_norm": 0.6992739144946865,
+      "learning_rate": 1.5476239197661918e-06,
+      "loss": 0.2992,
+      "step": 9146
+    },
+    {
+      "epoch": 0.750461500594823,
+      "grad_norm": 1.0077875140743224,
+      "learning_rate": 1.5466629140962158e-06,
+      "loss": 0.315,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7505435451450138,
+      "grad_norm": 0.5851851201151765,
+      "learning_rate": 1.5457021522953547e-06,
+      "loss": 0.2805,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7506255896952045,
+      "grad_norm": 0.6428382535898999,
+      "learning_rate": 1.5447416344314598e-06,
+      "loss": 0.3161,
+      "step": 9149
+    },
+    {
+      "epoch": 0.7507076342453952,
+      "grad_norm": 0.7978582464426802,
+      "learning_rate": 1.5437813605723578e-06,
+      "loss": 0.3383,
+      "step": 9150
+    },
+    {
+      "epoch": 0.750789678795586,
+      "grad_norm": 0.672175060471377,
+      "learning_rate": 1.5428213307858646e-06,
+      "loss": 0.2977,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7508717233457768,
+      "grad_norm": 0.7050464612210299,
+      "learning_rate": 1.5418615451397746e-06,
+      "loss": 0.2896,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7509537678959675,
+      "grad_norm": 0.6549473822720497,
+      "learning_rate": 1.5409020037018652e-06,
+      "loss": 0.3337,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7510358124461582,
+      "grad_norm": 0.742869157980379,
+      "learning_rate": 1.5399427065398998e-06,
+      "loss": 0.3025,
+      "step": 9154
+    },
+    {
+      "epoch": 0.751117856996349,
+      "grad_norm": 0.6795245045191878,
+      "learning_rate": 1.53898365372162e-06,
+      "loss": 0.3813,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7511999015465398,
+      "grad_norm": 0.6938287127007983,
+      "learning_rate": 1.5380248453147544e-06,
+      "loss": 0.3106,
+      "step": 9156
+    },
+    {
+      "epoch": 0.7512819460967305,
+      "grad_norm": 0.7278392247149372,
+      "learning_rate": 1.5370662813870134e-06,
+      "loss": 0.2927,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7513639906469213,
+      "grad_norm": 0.7879069809375612,
+      "learning_rate": 1.5361079620060864e-06,
+      "loss": 0.2963,
+      "step": 9158
+    },
+    {
+      "epoch": 0.7514460351971121,
+      "grad_norm": 0.7320851860239854,
+      "learning_rate": 1.535149887239652e-06,
+      "loss": 0.2692,
+      "step": 9159
+    },
+    {
+      "epoch": 0.7515280797473027,
+      "grad_norm": 0.6476565785774194,
+      "learning_rate": 1.5341920571553636e-06,
+      "loss": 0.2845,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7516101242974935,
+      "grad_norm": 0.7039356798482055,
+      "learning_rate": 1.533234471820864e-06,
+      "loss": 0.305,
+      "step": 9161
+    },
+    {
+      "epoch": 0.7516921688476843,
+      "grad_norm": 1.4241033597206696,
+      "learning_rate": 1.5322771313037783e-06,
+      "loss": 0.3031,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7517742133978751,
+      "grad_norm": 0.7121918807748865,
+      "learning_rate": 1.5313200356717083e-06,
+      "loss": 0.2742,
+      "step": 9163
+    },
+    {
+      "epoch": 0.7518562579480658,
+      "grad_norm": 0.7679116034375515,
+      "learning_rate": 1.5303631849922462e-06,
+      "loss": 0.3166,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7519383024982565,
+      "grad_norm": 0.6560395073685127,
+      "learning_rate": 1.5294065793329616e-06,
+      "loss": 0.2844,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7520203470484473,
+      "grad_norm": 0.6807875724073007,
+      "learning_rate": 1.528450218761407e-06,
+      "loss": 0.283,
+      "step": 9166
+    },
+    {
+      "epoch": 0.752102391598638,
+      "grad_norm": 0.7183227643857413,
+      "learning_rate": 1.5274941033451218e-06,
+      "loss": 0.3223,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7521844361488288,
+      "grad_norm": 0.7271724516487399,
+      "learning_rate": 1.5265382331516227e-06,
+      "loss": 0.2892,
+      "step": 9168
+    },
+    {
+      "epoch": 0.7522664806990196,
+      "grad_norm": 0.657986725264206,
+      "learning_rate": 1.5255826082484126e-06,
+      "loss": 0.2698,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7523485252492104,
+      "grad_norm": 0.6534391545751441,
+      "learning_rate": 1.5246272287029783e-06,
+      "loss": 0.3045,
+      "step": 9170
+    },
+    {
+      "epoch": 0.752430569799401,
+      "grad_norm": 0.6509778658661204,
+      "learning_rate": 1.5236720945827844e-06,
+      "loss": 0.2911,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7525126143495918,
+      "grad_norm": 0.6690868549058394,
+      "learning_rate": 1.522717205955283e-06,
+      "loss": 0.3421,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7525946588997826,
+      "grad_norm": 0.6525673204976727,
+      "learning_rate": 1.5217625628879045e-06,
+      "loss": 0.3096,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7526767034499733,
+      "grad_norm": 0.7396936634119159,
+      "learning_rate": 1.520808165448066e-06,
+      "loss": 0.3003,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7527587480001641,
+      "grad_norm": 0.6636148684495844,
+      "learning_rate": 1.5198540137031665e-06,
+      "loss": 0.3414,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7528407925503549,
+      "grad_norm": 3.191919589800228,
+      "learning_rate": 1.5189001077205835e-06,
+      "loss": 0.3574,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7529228371005456,
+      "grad_norm": 0.7116793624624108,
+      "learning_rate": 1.5179464475676835e-06,
+      "loss": 0.3041,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7530048816507363,
+      "grad_norm": 0.7608789438966971,
+      "learning_rate": 1.5169930333118094e-06,
+      "loss": 0.2626,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7530869262009271,
+      "grad_norm": 0.674784568345504,
+      "learning_rate": 1.5160398650202935e-06,
+      "loss": 0.3084,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7531689707511179,
+      "grad_norm": 0.7321304923826125,
+      "learning_rate": 1.5150869427604448e-06,
+      "loss": 0.314,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7532510153013086,
+      "grad_norm": 0.6520031688040314,
+      "learning_rate": 1.5141342665995552e-06,
+      "loss": 0.2986,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7533330598514993,
+      "grad_norm": 0.8030035799669342,
+      "learning_rate": 1.5131818366049034e-06,
+      "loss": 0.2686,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7534151044016901,
+      "grad_norm": 0.7384619431204582,
+      "learning_rate": 1.5122296528437502e-06,
+      "loss": 0.3318,
+      "step": 9183
+    },
+    {
+      "epoch": 0.7534971489518809,
+      "grad_norm": 0.6754281930775009,
+      "learning_rate": 1.5112777153833336e-06,
+      "loss": 0.3043,
+      "step": 9184
+    },
+    {
+      "epoch": 0.7535791935020716,
+      "grad_norm": 0.6846731212773788,
+      "learning_rate": 1.5103260242908812e-06,
+      "loss": 0.2937,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7536612380522624,
+      "grad_norm": 0.7290069618841798,
+      "learning_rate": 1.5093745796335969e-06,
+      "loss": 0.2814,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7537432826024532,
+      "grad_norm": 0.7203253214957259,
+      "learning_rate": 1.5084233814786715e-06,
+      "loss": 0.3239,
+      "step": 9187
+    },
+    {
+      "epoch": 0.7538253271526438,
+      "grad_norm": 0.7225019351014466,
+      "learning_rate": 1.507472429893279e-06,
+      "loss": 0.3111,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7539073717028346,
+      "grad_norm": 0.7164073393245418,
+      "learning_rate": 1.5065217249445706e-06,
+      "loss": 0.2787,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7539894162530254,
+      "grad_norm": 0.7213558175888527,
+      "learning_rate": 1.5055712666996875e-06,
+      "loss": 0.3627,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7540714608032162,
+      "grad_norm": 0.7129719831062662,
+      "learning_rate": 1.504621055225745e-06,
+      "loss": 0.2952,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7541535053534069,
+      "grad_norm": 0.754976649542303,
+      "learning_rate": 1.5036710905898494e-06,
+      "loss": 0.2986,
+      "step": 9192
+    },
+    {
+      "epoch": 0.7542355499035976,
+      "grad_norm": 0.660755308835193,
+      "learning_rate": 1.5027213728590846e-06,
+      "loss": 0.2905,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7543175944537884,
+      "grad_norm": 0.7750331331952097,
+      "learning_rate": 1.5017719021005156e-06,
+      "loss": 0.3214,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7543996390039791,
+      "grad_norm": 0.642645921460962,
+      "learning_rate": 1.5008226783811964e-06,
+      "loss": 0.2795,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7544816835541699,
+      "grad_norm": 0.6746223973324793,
+      "learning_rate": 1.4998737017681564e-06,
+      "loss": 0.3048,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7545637281043607,
+      "grad_norm": 0.6803201156766523,
+      "learning_rate": 1.4989249723284115e-06,
+      "loss": 0.2843,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7546457726545515,
+      "grad_norm": 0.7619729009169417,
+      "learning_rate": 1.4979764901289622e-06,
+      "loss": 0.2908,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7547278172047421,
+      "grad_norm": 0.6138998504762339,
+      "learning_rate": 1.4970282552367854e-06,
+      "loss": 0.301,
+      "step": 9199
+    },
+    {
+      "epoch": 0.7548098617549329,
+      "grad_norm": 0.7329022478954274,
+      "learning_rate": 1.4960802677188447e-06,
+      "loss": 0.275,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7548919063051237,
+      "grad_norm": 0.6981227559973097,
+      "learning_rate": 1.495132527642088e-06,
+      "loss": 0.3231,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7549739508553145,
+      "grad_norm": 0.7140241735386723,
+      "learning_rate": 1.4941850350734393e-06,
+      "loss": 0.31,
+      "step": 9202
+    },
+    {
+      "epoch": 0.7550559954055052,
+      "grad_norm": 0.6953399072965489,
+      "learning_rate": 1.4932377900798128e-06,
+      "loss": 0.3064,
+      "step": 9203
+    },
+    {
+      "epoch": 0.755138039955696,
+      "grad_norm": 0.6942969378417492,
+      "learning_rate": 1.4922907927280978e-06,
+      "loss": 0.3226,
+      "step": 9204
+    },
+    {
+      "epoch": 0.7552200845058867,
+      "grad_norm": 0.7214004599936166,
+      "learning_rate": 1.4913440430851734e-06,
+      "loss": 0.2913,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7553021290560774,
+      "grad_norm": 0.8525021562435114,
+      "learning_rate": 1.4903975412178933e-06,
+      "loss": 0.2865,
+      "step": 9206
+    },
+    {
+      "epoch": 0.7553841736062682,
+      "grad_norm": 0.6993586753442246,
+      "learning_rate": 1.4894512871931022e-06,
+      "loss": 0.3277,
+      "step": 9207
+    },
+    {
+      "epoch": 0.755466218156459,
+      "grad_norm": 0.7134090588152315,
+      "learning_rate": 1.4885052810776213e-06,
+      "loss": 0.3351,
+      "step": 9208
+    },
+    {
+      "epoch": 0.7555482627066498,
+      "grad_norm": 0.7073667559518289,
+      "learning_rate": 1.4875595229382538e-06,
+      "loss": 0.3263,
+      "step": 9209
+    },
+    {
+      "epoch": 0.7556303072568404,
+      "grad_norm": 0.7541154750864597,
+      "learning_rate": 1.48661401284179e-06,
+      "loss": 0.3188,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7557123518070312,
+      "grad_norm": 0.698358085869162,
+      "learning_rate": 1.4856687508550012e-06,
+      "loss": 0.3132,
+      "step": 9211
+    },
+    {
+      "epoch": 0.755794396357222,
+      "grad_norm": 0.6661904949246041,
+      "learning_rate": 1.4847237370446378e-06,
+      "loss": 0.2778,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7558764409074127,
+      "grad_norm": 1.1116093088864225,
+      "learning_rate": 1.4837789714774375e-06,
+      "loss": 0.293,
+      "step": 9213
+    },
+    {
+      "epoch": 0.7559584854576035,
+      "grad_norm": 0.756842306798306,
+      "learning_rate": 1.4828344542201155e-06,
+      "loss": 0.2865,
+      "step": 9214
+    },
+    {
+      "epoch": 0.7560405300077943,
+      "grad_norm": 0.7275980221026159,
+      "learning_rate": 1.4818901853393735e-06,
+      "loss": 0.2822,
+      "step": 9215
+    },
+    {
+      "epoch": 0.756122574557985,
+      "grad_norm": 0.7555158607721775,
+      "learning_rate": 1.4809461649018964e-06,
+      "loss": 0.2974,
+      "step": 9216
+    },
+    {
+      "epoch": 0.7562046191081757,
+      "grad_norm": 0.661055058867387,
+      "learning_rate": 1.4800023929743452e-06,
+      "loss": 0.2957,
+      "step": 9217
+    },
+    {
+      "epoch": 0.7562866636583665,
+      "grad_norm": 0.694163802668386,
+      "learning_rate": 1.4790588696233698e-06,
+      "loss": 0.3114,
+      "step": 9218
+    },
+    {
+      "epoch": 0.7563687082085573,
+      "grad_norm": 0.9195982955353118,
+      "learning_rate": 1.4781155949156023e-06,
+      "loss": 0.2758,
+      "step": 9219
+    },
+    {
+      "epoch": 0.756450752758748,
+      "grad_norm": 0.7221824786586951,
+      "learning_rate": 1.4771725689176524e-06,
+      "loss": 0.3351,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7565327973089387,
+      "grad_norm": 0.7003842661623756,
+      "learning_rate": 1.4762297916961166e-06,
+      "loss": 0.3426,
+      "step": 9221
+    },
+    {
+      "epoch": 0.7566148418591295,
+      "grad_norm": 0.6579569838708601,
+      "learning_rate": 1.4752872633175691e-06,
+      "loss": 0.2569,
+      "step": 9222
+    },
+    {
+      "epoch": 0.7566968864093203,
+      "grad_norm": 0.7392316193781011,
+      "learning_rate": 1.4743449838485729e-06,
+      "loss": 0.2797,
+      "step": 9223
+    },
+    {
+      "epoch": 0.756778930959511,
+      "grad_norm": 0.6812342101443275,
+      "learning_rate": 1.473402953355671e-06,
+      "loss": 0.272,
+      "step": 9224
+    },
+    {
+      "epoch": 0.7568609755097018,
+      "grad_norm": 0.7261298334809048,
+      "learning_rate": 1.472461171905385e-06,
+      "loss": 0.2773,
+      "step": 9225
+    },
+    {
+      "epoch": 0.7569430200598926,
+      "grad_norm": 0.6840227282079538,
+      "learning_rate": 1.471519639564225e-06,
+      "loss": 0.3235,
+      "step": 9226
+    },
+    {
+      "epoch": 0.7570250646100832,
+      "grad_norm": 0.7364400679685211,
+      "learning_rate": 1.4705783563986774e-06,
+      "loss": 0.3362,
+      "step": 9227
+    },
+    {
+      "epoch": 0.757107109160274,
+      "grad_norm": 0.7055406173826272,
+      "learning_rate": 1.4696373224752163e-06,
+      "loss": 0.3003,
+      "step": 9228
+    },
+    {
+      "epoch": 0.7571891537104648,
+      "grad_norm": 0.7200501568089207,
+      "learning_rate": 1.468696537860297e-06,
+      "loss": 0.2691,
+      "step": 9229
+    },
+    {
+      "epoch": 0.7572711982606556,
+      "grad_norm": 0.6776057596254301,
+      "learning_rate": 1.467756002620353e-06,
+      "loss": 0.2906,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7573532428108463,
+      "grad_norm": 0.7835559404156061,
+      "learning_rate": 1.4668157168218067e-06,
+      "loss": 0.319,
+      "step": 9231
+    },
+    {
+      "epoch": 0.757435287361037,
+      "grad_norm": 0.6822266091205285,
+      "learning_rate": 1.4658756805310565e-06,
+      "loss": 0.3026,
+      "step": 9232
+    },
+    {
+      "epoch": 0.7575173319112278,
+      "grad_norm": 0.6799446615307962,
+      "learning_rate": 1.4649358938144891e-06,
+      "loss": 0.3285,
+      "step": 9233
+    },
+    {
+      "epoch": 0.7575993764614185,
+      "grad_norm": 0.7425403576103549,
+      "learning_rate": 1.4639963567384679e-06,
+      "loss": 0.3014,
+      "step": 9234
+    },
+    {
+      "epoch": 0.7576814210116093,
+      "grad_norm": 0.7040123270424096,
+      "learning_rate": 1.4630570693693447e-06,
+      "loss": 0.3076,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7577634655618001,
+      "grad_norm": 0.682921712997639,
+      "learning_rate": 1.4621180317734468e-06,
+      "loss": 0.3202,
+      "step": 9236
+    },
+    {
+      "epoch": 0.7578455101119909,
+      "grad_norm": 0.7054652397741124,
+      "learning_rate": 1.4611792440170914e-06,
+      "loss": 0.3043,
+      "step": 9237
+    },
+    {
+      "epoch": 0.7579275546621815,
+      "grad_norm": 0.718690519577948,
+      "learning_rate": 1.4602407061665702e-06,
+      "loss": 0.3274,
+      "step": 9238
+    },
+    {
+      "epoch": 0.7580095992123723,
+      "grad_norm": 0.6644921582658266,
+      "learning_rate": 1.4593024182881655e-06,
+      "loss": 0.3045,
+      "step": 9239
+    },
+    {
+      "epoch": 0.7580916437625631,
+      "grad_norm": 0.7162049255688311,
+      "learning_rate": 1.4583643804481334e-06,
+      "loss": 0.3112,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7581736883127538,
+      "grad_norm": 0.6988928740569651,
+      "learning_rate": 1.457426592712719e-06,
+      "loss": 0.3114,
+      "step": 9241
+    },
+    {
+      "epoch": 0.7582557328629446,
+      "grad_norm": 0.7530640242009705,
+      "learning_rate": 1.4564890551481481e-06,
+      "loss": 0.2637,
+      "step": 9242
+    },
+    {
+      "epoch": 0.7583377774131353,
+      "grad_norm": 0.6548753628879006,
+      "learning_rate": 1.455551767820626e-06,
+      "loss": 0.2433,
+      "step": 9243
+    },
+    {
+      "epoch": 0.7584198219633261,
+      "grad_norm": 0.6187969151752795,
+      "learning_rate": 1.4546147307963449e-06,
+      "loss": 0.2849,
+      "step": 9244
+    },
+    {
+      "epoch": 0.7585018665135168,
+      "grad_norm": 0.6977057251403874,
+      "learning_rate": 1.453677944141474e-06,
+      "loss": 0.2881,
+      "step": 9245
+    },
+    {
+      "epoch": 0.7585839110637076,
+      "grad_norm": 0.6382797864891625,
+      "learning_rate": 1.4527414079221686e-06,
+      "loss": 0.2968,
+      "step": 9246
+    },
+    {
+      "epoch": 0.7586659556138984,
+      "grad_norm": 0.6696849898780775,
+      "learning_rate": 1.4518051222045675e-06,
+      "loss": 0.2798,
+      "step": 9247
+    },
+    {
+      "epoch": 0.758748000164089,
+      "grad_norm": 0.6320923057655288,
+      "learning_rate": 1.4508690870547888e-06,
+      "loss": 0.3067,
+      "step": 9248
+    },
+    {
+      "epoch": 0.7588300447142798,
+      "grad_norm": 0.7498836072215277,
+      "learning_rate": 1.4499333025389323e-06,
+      "loss": 0.2941,
+      "step": 9249
+    },
+    {
+      "epoch": 0.7589120892644706,
+      "grad_norm": 0.6826060213679302,
+      "learning_rate": 1.448997768723081e-06,
+      "loss": 0.3312,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7589941338146614,
+      "grad_norm": 1.0009028410180587,
+      "learning_rate": 1.4480624856733022e-06,
+      "loss": 0.3097,
+      "step": 9251
+    },
+    {
+      "epoch": 0.7590761783648521,
+      "grad_norm": 0.7371326961357251,
+      "learning_rate": 1.4471274534556457e-06,
+      "loss": 0.3221,
+      "step": 9252
+    },
+    {
+      "epoch": 0.7591582229150429,
+      "grad_norm": 0.6499480125468649,
+      "learning_rate": 1.4461926721361386e-06,
+      "loss": 0.2887,
+      "step": 9253
+    },
+    {
+      "epoch": 0.7592402674652337,
+      "grad_norm": 0.7694362692281264,
+      "learning_rate": 1.4452581417807954e-06,
+      "loss": 0.3513,
+      "step": 9254
+    },
+    {
+      "epoch": 0.7593223120154243,
+      "grad_norm": 0.7282890715211622,
+      "learning_rate": 1.4443238624556127e-06,
+      "loss": 0.3238,
+      "step": 9255
+    },
+    {
+      "epoch": 0.7594043565656151,
+      "grad_norm": 0.7098687955274144,
+      "learning_rate": 1.4433898342265646e-06,
+      "loss": 0.2972,
+      "step": 9256
+    },
+    {
+      "epoch": 0.7594864011158059,
+      "grad_norm": 0.6558852745064866,
+      "learning_rate": 1.4424560571596135e-06,
+      "loss": 0.301,
+      "step": 9257
+    },
+    {
+      "epoch": 0.7595684456659967,
+      "grad_norm": 1.1541714737177515,
+      "learning_rate": 1.4415225313206987e-06,
+      "loss": 0.3011,
+      "step": 9258
+    },
+    {
+      "epoch": 0.7596504902161874,
+      "grad_norm": 0.7142716166552527,
+      "learning_rate": 1.4405892567757456e-06,
+      "loss": 0.2909,
+      "step": 9259
+    },
+    {
+      "epoch": 0.7597325347663781,
+      "grad_norm": 0.6808382242791372,
+      "learning_rate": 1.4396562335906622e-06,
+      "loss": 0.2945,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7598145793165689,
+      "grad_norm": 0.6825213403184629,
+      "learning_rate": 1.4387234618313357e-06,
+      "loss": 0.3007,
+      "step": 9261
+    },
+    {
+      "epoch": 0.7598966238667596,
+      "grad_norm": 0.7037418250309235,
+      "learning_rate": 1.4377909415636349e-06,
+      "loss": 0.329,
+      "step": 9262
+    },
+    {
+      "epoch": 0.7599786684169504,
+      "grad_norm": 0.6354895164170925,
+      "learning_rate": 1.4368586728534162e-06,
+      "loss": 0.2629,
+      "step": 9263
+    },
+    {
+      "epoch": 0.7600607129671412,
+      "grad_norm": 0.7174339256110237,
+      "learning_rate": 1.435926655766512e-06,
+      "loss": 0.2928,
+      "step": 9264
+    },
+    {
+      "epoch": 0.760142757517332,
+      "grad_norm": 0.7573908535082903,
+      "learning_rate": 1.4349948903687428e-06,
+      "loss": 0.3195,
+      "step": 9265
+    },
+    {
+      "epoch": 0.7602248020675226,
+      "grad_norm": 1.0978568632099286,
+      "learning_rate": 1.434063376725905e-06,
+      "loss": 0.3274,
+      "step": 9266
+    },
+    {
+      "epoch": 0.7603068466177134,
+      "grad_norm": 0.6960448399426167,
+      "learning_rate": 1.433132114903784e-06,
+      "loss": 0.2964,
+      "step": 9267
+    },
+    {
+      "epoch": 0.7603888911679042,
+      "grad_norm": 0.7238711963858693,
+      "learning_rate": 1.432201104968141e-06,
+      "loss": 0.3088,
+      "step": 9268
+    },
+    {
+      "epoch": 0.7604709357180949,
+      "grad_norm": 0.7778066039005469,
+      "learning_rate": 1.4312703469847238e-06,
+      "loss": 0.284,
+      "step": 9269
+    },
+    {
+      "epoch": 0.7605529802682857,
+      "grad_norm": 0.7202979909856309,
+      "learning_rate": 1.430339841019263e-06,
+      "loss": 0.2737,
+      "step": 9270
+    },
+    {
+      "epoch": 0.7606350248184764,
+      "grad_norm": 0.7202120969311447,
+      "learning_rate": 1.4294095871374658e-06,
+      "loss": 0.2908,
+      "step": 9271
+    },
+    {
+      "epoch": 0.7607170693686672,
+      "grad_norm": 0.6440082928498808,
+      "learning_rate": 1.4284795854050265e-06,
+      "loss": 0.2975,
+      "step": 9272
+    },
+    {
+      "epoch": 0.7607991139188579,
+      "grad_norm": 0.7076982961292944,
+      "learning_rate": 1.4275498358876227e-06,
+      "loss": 0.3035,
+      "step": 9273
+    },
+    {
+      "epoch": 0.7608811584690487,
+      "grad_norm": 0.6965634754577341,
+      "learning_rate": 1.4266203386509087e-06,
+      "loss": 0.2918,
+      "step": 9274
+    },
+    {
+      "epoch": 0.7609632030192395,
+      "grad_norm": 0.7274097777569987,
+      "learning_rate": 1.4256910937605263e-06,
+      "loss": 0.2737,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7610452475694301,
+      "grad_norm": 0.6886811359824375,
+      "learning_rate": 1.4247621012820967e-06,
+      "loss": 0.2845,
+      "step": 9276
+    },
+    {
+      "epoch": 0.7611272921196209,
+      "grad_norm": 0.8710762267617796,
+      "learning_rate": 1.4238333612812215e-06,
+      "loss": 0.2935,
+      "step": 9277
+    },
+    {
+      "epoch": 0.7612093366698117,
+      "grad_norm": 0.7857801119031241,
+      "learning_rate": 1.4229048738234907e-06,
+      "loss": 0.2946,
+      "step": 9278
+    },
+    {
+      "epoch": 0.7612913812200025,
+      "grad_norm": 0.6691777722420632,
+      "learning_rate": 1.421976638974469e-06,
+      "loss": 0.33,
+      "step": 9279
+    },
+    {
+      "epoch": 0.7613734257701932,
+      "grad_norm": 0.674172836804011,
+      "learning_rate": 1.42104865679971e-06,
+      "loss": 0.3319,
+      "step": 9280
+    },
+    {
+      "epoch": 0.761455470320384,
+      "grad_norm": 0.6273823089372486,
+      "learning_rate": 1.4201209273647431e-06,
+      "loss": 0.3113,
+      "step": 9281
+    },
+    {
+      "epoch": 0.7615375148705747,
+      "grad_norm": 0.6674981642152833,
+      "learning_rate": 1.4191934507350851e-06,
+      "loss": 0.2718,
+      "step": 9282
+    },
+    {
+      "epoch": 0.7616195594207654,
+      "grad_norm": 0.684316638559796,
+      "learning_rate": 1.4182662269762342e-06,
+      "loss": 0.3036,
+      "step": 9283
+    },
+    {
+      "epoch": 0.7617016039709562,
+      "grad_norm": 0.6308855713960391,
+      "learning_rate": 1.4173392561536665e-06,
+      "loss": 0.2942,
+      "step": 9284
+    },
+    {
+      "epoch": 0.761783648521147,
+      "grad_norm": 0.6566180033423976,
+      "learning_rate": 1.4164125383328453e-06,
+      "loss": 0.2872,
+      "step": 9285
+    },
+    {
+      "epoch": 0.7618656930713378,
+      "grad_norm": 0.6715570915998335,
+      "learning_rate": 1.415486073579212e-06,
+      "loss": 0.2804,
+      "step": 9286
+    },
+    {
+      "epoch": 0.7619477376215285,
+      "grad_norm": 0.7300449007503895,
+      "learning_rate": 1.4145598619581935e-06,
+      "loss": 0.3425,
+      "step": 9287
+    },
+    {
+      "epoch": 0.7620297821717192,
+      "grad_norm": 0.6540424568609088,
+      "learning_rate": 1.4136339035351988e-06,
+      "loss": 0.2896,
+      "step": 9288
+    },
+    {
+      "epoch": 0.76211182672191,
+      "grad_norm": 1.3556891082007572,
+      "learning_rate": 1.4127081983756152e-06,
+      "loss": 0.3045,
+      "step": 9289
+    },
+    {
+      "epoch": 0.7621938712721007,
+      "grad_norm": 0.8025580035668011,
+      "learning_rate": 1.4117827465448142e-06,
+      "loss": 0.3407,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7622759158222915,
+      "grad_norm": 0.6905415791545629,
+      "learning_rate": 1.4108575481081522e-06,
+      "loss": 0.3283,
+      "step": 9291
+    },
+    {
+      "epoch": 0.7623579603724823,
+      "grad_norm": 0.6857240416789837,
+      "learning_rate": 1.4099326031309617e-06,
+      "loss": 0.3114,
+      "step": 9292
+    },
+    {
+      "epoch": 0.762440004922673,
+      "grad_norm": 0.6934297143092084,
+      "learning_rate": 1.4090079116785649e-06,
+      "loss": 0.2726,
+      "step": 9293
+    },
+    {
+      "epoch": 0.7625220494728637,
+      "grad_norm": 0.686659466308641,
+      "learning_rate": 1.4080834738162585e-06,
+      "loss": 0.3368,
+      "step": 9294
+    },
+    {
+      "epoch": 0.7626040940230545,
+      "grad_norm": 0.7335486166023579,
+      "learning_rate": 1.4071592896093261e-06,
+      "loss": 0.2933,
+      "step": 9295
+    },
+    {
+      "epoch": 0.7626861385732453,
+      "grad_norm": 0.7483965767782312,
+      "learning_rate": 1.4062353591230343e-06,
+      "loss": 0.2923,
+      "step": 9296
+    },
+    {
+      "epoch": 0.762768183123436,
+      "grad_norm": 0.6863755778645716,
+      "learning_rate": 1.4053116824226253e-06,
+      "loss": 0.3003,
+      "step": 9297
+    },
+    {
+      "epoch": 0.7628502276736268,
+      "grad_norm": 0.712397095393335,
+      "learning_rate": 1.4043882595733322e-06,
+      "loss": 0.3218,
+      "step": 9298
+    },
+    {
+      "epoch": 0.7629322722238175,
+      "grad_norm": 0.6932781518656108,
+      "learning_rate": 1.4034650906403618e-06,
+      "loss": 0.2921,
+      "step": 9299
+    },
+    {
+      "epoch": 0.7630143167740083,
+      "grad_norm": 0.6949008051618677,
+      "learning_rate": 1.402542175688908e-06,
+      "loss": 0.2639,
+      "step": 9300
+    },
+    {
+      "epoch": 0.763096361324199,
+      "grad_norm": 0.6556130747892462,
+      "learning_rate": 1.4016195147841483e-06,
+      "loss": 0.3029,
+      "step": 9301
+    },
+    {
+      "epoch": 0.7631784058743898,
+      "grad_norm": 0.6825410212042119,
+      "learning_rate": 1.4006971079912352e-06,
+      "loss": 0.3411,
+      "step": 9302
+    },
+    {
+      "epoch": 0.7632604504245806,
+      "grad_norm": 0.6624327897868582,
+      "learning_rate": 1.399774955375311e-06,
+      "loss": 0.2944,
+      "step": 9303
+    },
+    {
+      "epoch": 0.7633424949747712,
+      "grad_norm": 0.6817195041752729,
+      "learning_rate": 1.3988530570014952e-06,
+      "loss": 0.2981,
+      "step": 9304
+    },
+    {
+      "epoch": 0.763424539524962,
+      "grad_norm": 0.7031522364979518,
+      "learning_rate": 1.3979314129348898e-06,
+      "loss": 0.2702,
+      "step": 9305
+    },
+    {
+      "epoch": 0.7635065840751528,
+      "grad_norm": 0.7062412021405754,
+      "learning_rate": 1.3970100232405819e-06,
+      "loss": 0.3045,
+      "step": 9306
+    },
+    {
+      "epoch": 0.7635886286253436,
+      "grad_norm": 0.7586326197753792,
+      "learning_rate": 1.396088887983636e-06,
+      "loss": 0.2753,
+      "step": 9307
+    },
+    {
+      "epoch": 0.7636706731755343,
+      "grad_norm": 0.778740765910696,
+      "learning_rate": 1.395168007229103e-06,
+      "loss": 0.3013,
+      "step": 9308
+    },
+    {
+      "epoch": 0.7637527177257251,
+      "grad_norm": 0.6947346984143363,
+      "learning_rate": 1.3942473810420153e-06,
+      "loss": 0.2717,
+      "step": 9309
+    },
+    {
+      "epoch": 0.7638347622759158,
+      "grad_norm": 0.7747584240322248,
+      "learning_rate": 1.3933270094873824e-06,
+      "loss": 0.2853,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7639168068261066,
+      "grad_norm": 0.7252900860257016,
+      "learning_rate": 1.3924068926302038e-06,
+      "loss": 0.2776,
+      "step": 9311
+    },
+    {
+      "epoch": 0.7639988513762973,
+      "grad_norm": 0.6829652120912202,
+      "learning_rate": 1.3914870305354522e-06,
+      "loss": 0.2963,
+      "step": 9312
+    },
+    {
+      "epoch": 0.7640808959264881,
+      "grad_norm": 0.6835369330934753,
+      "learning_rate": 1.3905674232680893e-06,
+      "loss": 0.2821,
+      "step": 9313
+    },
+    {
+      "epoch": 0.7641629404766789,
+      "grad_norm": 0.6865494180502284,
+      "learning_rate": 1.3896480708930576e-06,
+      "loss": 0.3163,
+      "step": 9314
+    },
+    {
+      "epoch": 0.7642449850268695,
+      "grad_norm": 0.6969587291620375,
+      "learning_rate": 1.3887289734752769e-06,
+      "loss": 0.336,
+      "step": 9315
+    },
+    {
+      "epoch": 0.7643270295770603,
+      "grad_norm": 0.7505836128439293,
+      "learning_rate": 1.387810131079656e-06,
+      "loss": 0.3252,
+      "step": 9316
+    },
+    {
+      "epoch": 0.7644090741272511,
+      "grad_norm": 0.7104277433832354,
+      "learning_rate": 1.38689154377108e-06,
+      "loss": 0.2938,
+      "step": 9317
+    },
+    {
+      "epoch": 0.7644911186774419,
+      "grad_norm": 0.653500720812662,
+      "learning_rate": 1.3859732116144164e-06,
+      "loss": 0.2919,
+      "step": 9318
+    },
+    {
+      "epoch": 0.7645731632276326,
+      "grad_norm": 0.9718745621658281,
+      "learning_rate": 1.3850551346745207e-06,
+      "loss": 0.3165,
+      "step": 9319
+    },
+    {
+      "epoch": 0.7646552077778234,
+      "grad_norm": 0.6731560548997546,
+      "learning_rate": 1.384137313016221e-06,
+      "loss": 0.2818,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7647372523280141,
+      "grad_norm": 0.6556983461936733,
+      "learning_rate": 1.3832197467043367e-06,
+      "loss": 0.2611,
+      "step": 9321
+    },
+    {
+      "epoch": 0.7648192968782048,
+      "grad_norm": 0.7291156751604972,
+      "learning_rate": 1.3823024358036618e-06,
+      "loss": 0.3211,
+      "step": 9322
+    },
+    {
+      "epoch": 0.7649013414283956,
+      "grad_norm": 0.6923057278527324,
+      "learning_rate": 1.3813853803789767e-06,
+      "loss": 0.3045,
+      "step": 9323
+    },
+    {
+      "epoch": 0.7649833859785864,
+      "grad_norm": 0.7221946034797314,
+      "learning_rate": 1.3804685804950434e-06,
+      "loss": 0.3323,
+      "step": 9324
+    },
+    {
+      "epoch": 0.7650654305287772,
+      "grad_norm": 0.7103821002211725,
+      "learning_rate": 1.3795520362166026e-06,
+      "loss": 0.2973,
+      "step": 9325
+    },
+    {
+      "epoch": 0.7651474750789679,
+      "grad_norm": 0.7061782110803665,
+      "learning_rate": 1.3786357476083801e-06,
+      "loss": 0.2868,
+      "step": 9326
+    },
+    {
+      "epoch": 0.7652295196291586,
+      "grad_norm": 0.7517841206040957,
+      "learning_rate": 1.3777197147350845e-06,
+      "loss": 0.3523,
+      "step": 9327
+    },
+    {
+      "epoch": 0.7653115641793494,
+      "grad_norm": 0.6944571579305808,
+      "learning_rate": 1.3768039376614017e-06,
+      "loss": 0.3216,
+      "step": 9328
+    },
+    {
+      "epoch": 0.7653936087295401,
+      "grad_norm": 0.752504510625696,
+      "learning_rate": 1.3758884164520047e-06,
+      "loss": 0.3345,
+      "step": 9329
+    },
+    {
+      "epoch": 0.7654756532797309,
+      "grad_norm": 0.7890981420748432,
+      "learning_rate": 1.3749731511715442e-06,
+      "loss": 0.263,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7655576978299217,
+      "grad_norm": 0.7316067507835474,
+      "learning_rate": 1.374058141884657e-06,
+      "loss": 0.325,
+      "step": 9331
+    },
+    {
+      "epoch": 0.7656397423801125,
+      "grad_norm": 0.7272804731283538,
+      "learning_rate": 1.373143388655958e-06,
+      "loss": 0.2994,
+      "step": 9332
+    },
+    {
+      "epoch": 0.7657217869303031,
+      "grad_norm": 0.7155462294976511,
+      "learning_rate": 1.3722288915500448e-06,
+      "loss": 0.3079,
+      "step": 9333
+    },
+    {
+      "epoch": 0.7658038314804939,
+      "grad_norm": 0.6591412198448748,
+      "learning_rate": 1.3713146506315006e-06,
+      "loss": 0.2754,
+      "step": 9334
+    },
+    {
+      "epoch": 0.7658858760306847,
+      "grad_norm": 0.6672666749803127,
+      "learning_rate": 1.3704006659648838e-06,
+      "loss": 0.3128,
+      "step": 9335
+    },
+    {
+      "epoch": 0.7659679205808754,
+      "grad_norm": 0.6936766321046661,
+      "learning_rate": 1.3694869376147406e-06,
+      "loss": 0.3052,
+      "step": 9336
+    },
+    {
+      "epoch": 0.7660499651310662,
+      "grad_norm": 0.785297340586933,
+      "learning_rate": 1.368573465645599e-06,
+      "loss": 0.2985,
+      "step": 9337
+    },
+    {
+      "epoch": 0.7661320096812569,
+      "grad_norm": 0.7609096255149898,
+      "learning_rate": 1.3676602501219626e-06,
+      "loss": 0.3156,
+      "step": 9338
+    },
+    {
+      "epoch": 0.7662140542314477,
+      "grad_norm": 0.7181366738404591,
+      "learning_rate": 1.366747291108324e-06,
+      "loss": 0.309,
+      "step": 9339
+    },
+    {
+      "epoch": 0.7662960987816384,
+      "grad_norm": 0.6566488796788624,
+      "learning_rate": 1.365834588669156e-06,
+      "loss": 0.2892,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7663781433318292,
+      "grad_norm": 0.6779402952106534,
+      "learning_rate": 1.3649221428689092e-06,
+      "loss": 0.3337,
+      "step": 9341
+    },
+    {
+      "epoch": 0.76646018788202,
+      "grad_norm": 0.7451083715443181,
+      "learning_rate": 1.364009953772022e-06,
+      "loss": 0.2757,
+      "step": 9342
+    },
+    {
+      "epoch": 0.7665422324322106,
+      "grad_norm": 0.646782846340555,
+      "learning_rate": 1.3630980214429085e-06,
+      "loss": 0.3188,
+      "step": 9343
+    },
+    {
+      "epoch": 0.7666242769824014,
+      "grad_norm": 0.6756493058840775,
+      "learning_rate": 1.362186345945971e-06,
+      "loss": 0.2896,
+      "step": 9344
+    },
+    {
+      "epoch": 0.7667063215325922,
+      "grad_norm": 0.6609077241433914,
+      "learning_rate": 1.3612749273455894e-06,
+      "loss": 0.2944,
+      "step": 9345
+    },
+    {
+      "epoch": 0.766788366082783,
+      "grad_norm": 0.8457189772739062,
+      "learning_rate": 1.3603637657061247e-06,
+      "loss": 0.3194,
+      "step": 9346
+    },
+    {
+      "epoch": 0.7668704106329737,
+      "grad_norm": 0.8069969628645388,
+      "learning_rate": 1.359452861091925e-06,
+      "loss": 0.2576,
+      "step": 9347
+    },
+    {
+      "epoch": 0.7669524551831645,
+      "grad_norm": 0.7138507536526255,
+      "learning_rate": 1.3585422135673131e-06,
+      "loss": 0.3191,
+      "step": 9348
+    },
+    {
+      "epoch": 0.7670344997333552,
+      "grad_norm": 0.8496686394766987,
+      "learning_rate": 1.3576318231966006e-06,
+      "loss": 0.3207,
+      "step": 9349
+    },
+    {
+      "epoch": 0.7671165442835459,
+      "grad_norm": 0.5973518418274022,
+      "learning_rate": 1.3567216900440777e-06,
+      "loss": 0.291,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7671985888337367,
+      "grad_norm": 0.8317819840425363,
+      "learning_rate": 1.3558118141740145e-06,
+      "loss": 0.3019,
+      "step": 9351
+    },
+    {
+      "epoch": 0.7672806333839275,
+      "grad_norm": 0.7024110115706885,
+      "learning_rate": 1.3549021956506676e-06,
+      "loss": 0.3004,
+      "step": 9352
+    },
+    {
+      "epoch": 0.7673626779341183,
+      "grad_norm": 0.7109664548391333,
+      "learning_rate": 1.3539928345382697e-06,
+      "loss": 0.3285,
+      "step": 9353
+    },
+    {
+      "epoch": 0.767444722484309,
+      "grad_norm": 0.7473452780012032,
+      "learning_rate": 1.3530837309010403e-06,
+      "loss": 0.311,
+      "step": 9354
+    },
+    {
+      "epoch": 0.7675267670344997,
+      "grad_norm": 0.6158085615505112,
+      "learning_rate": 1.3521748848031796e-06,
+      "loss": 0.3094,
+      "step": 9355
+    },
+    {
+      "epoch": 0.7676088115846905,
+      "grad_norm": 0.6702463432259008,
+      "learning_rate": 1.3512662963088668e-06,
+      "loss": 0.3272,
+      "step": 9356
+    },
+    {
+      "epoch": 0.7676908561348812,
+      "grad_norm": 0.7585281084946524,
+      "learning_rate": 1.3503579654822674e-06,
+      "loss": 0.3149,
+      "step": 9357
+    },
+    {
+      "epoch": 0.767772900685072,
+      "grad_norm": 0.7387432017688836,
+      "learning_rate": 1.3494498923875238e-06,
+      "loss": 0.3214,
+      "step": 9358
+    },
+    {
+      "epoch": 0.7678549452352628,
+      "grad_norm": 0.8247021751889948,
+      "learning_rate": 1.3485420770887648e-06,
+      "loss": 0.3284,
+      "step": 9359
+    },
+    {
+      "epoch": 0.7679369897854535,
+      "grad_norm": 0.6649404160714328,
+      "learning_rate": 1.3476345196500978e-06,
+      "loss": 0.2745,
+      "step": 9360
+    },
+    {
+      "epoch": 0.7680190343356442,
+      "grad_norm": 0.6826433773316647,
+      "learning_rate": 1.3467272201356113e-06,
+      "loss": 0.2822,
+      "step": 9361
+    },
+    {
+      "epoch": 0.768101078885835,
+      "grad_norm": 0.6793193905525505,
+      "learning_rate": 1.3458201786093795e-06,
+      "loss": 0.288,
+      "step": 9362
+    },
+    {
+      "epoch": 0.7681831234360258,
+      "grad_norm": 0.6255687785677982,
+      "learning_rate": 1.3449133951354571e-06,
+      "loss": 0.2547,
+      "step": 9363
+    },
+    {
+      "epoch": 0.7682651679862165,
+      "grad_norm": 0.7031132752867626,
+      "learning_rate": 1.3440068697778774e-06,
+      "loss": 0.3047,
+      "step": 9364
+    },
+    {
+      "epoch": 0.7683472125364073,
+      "grad_norm": 0.6868850121429243,
+      "learning_rate": 1.3431006026006594e-06,
+      "loss": 0.2971,
+      "step": 9365
+    },
+    {
+      "epoch": 0.768429257086598,
+      "grad_norm": 0.7057823064420108,
+      "learning_rate": 1.3421945936678004e-06,
+      "loss": 0.3363,
+      "step": 9366
+    },
+    {
+      "epoch": 0.7685113016367888,
+      "grad_norm": 0.6931151615092911,
+      "learning_rate": 1.341288843043283e-06,
+      "loss": 0.2953,
+      "step": 9367
+    },
+    {
+      "epoch": 0.7685933461869795,
+      "grad_norm": 0.7710520587645033,
+      "learning_rate": 1.3403833507910702e-06,
+      "loss": 0.2819,
+      "step": 9368
+    },
+    {
+      "epoch": 0.7686753907371703,
+      "grad_norm": 0.696594459815586,
+      "learning_rate": 1.3394781169751042e-06,
+      "loss": 0.2783,
+      "step": 9369
+    },
+    {
+      "epoch": 0.7687574352873611,
+      "grad_norm": 0.6967644646711653,
+      "learning_rate": 1.3385731416593144e-06,
+      "loss": 0.3044,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7688394798375517,
+      "grad_norm": 0.6980653422448246,
+      "learning_rate": 1.337668424907605e-06,
+      "loss": 0.3003,
+      "step": 9371
+    },
+    {
+      "epoch": 0.7689215243877425,
+      "grad_norm": 0.7561156476694205,
+      "learning_rate": 1.3367639667838689e-06,
+      "loss": 0.345,
+      "step": 9372
+    },
+    {
+      "epoch": 0.7690035689379333,
+      "grad_norm": 0.7417897505483574,
+      "learning_rate": 1.3358597673519757e-06,
+      "loss": 0.2931,
+      "step": 9373
+    },
+    {
+      "epoch": 0.7690856134881241,
+      "grad_norm": 0.7093466237549991,
+      "learning_rate": 1.334955826675778e-06,
+      "loss": 0.2689,
+      "step": 9374
+    },
+    {
+      "epoch": 0.7691676580383148,
+      "grad_norm": 0.7438925505412906,
+      "learning_rate": 1.3340521448191113e-06,
+      "loss": 0.3385,
+      "step": 9375
+    },
+    {
+      "epoch": 0.7692497025885056,
+      "grad_norm": 0.7077094086519436,
+      "learning_rate": 1.3331487218457933e-06,
+      "loss": 0.3485,
+      "step": 9376
+    },
+    {
+      "epoch": 0.7693317471386963,
+      "grad_norm": 0.6520126156043501,
+      "learning_rate": 1.3322455578196203e-06,
+      "loss": 0.3141,
+      "step": 9377
+    },
+    {
+      "epoch": 0.769413791688887,
+      "grad_norm": 0.6889673779878093,
+      "learning_rate": 1.3313426528043749e-06,
+      "loss": 0.2895,
+      "step": 9378
+    },
+    {
+      "epoch": 0.7694958362390778,
+      "grad_norm": 0.6496286862406304,
+      "learning_rate": 1.3304400068638157e-06,
+      "loss": 0.304,
+      "step": 9379
+    },
+    {
+      "epoch": 0.7695778807892686,
+      "grad_norm": 0.6387111222490852,
+      "learning_rate": 1.3295376200616871e-06,
+      "loss": 0.3139,
+      "step": 9380
+    },
+    {
+      "epoch": 0.7696599253394594,
+      "grad_norm": 0.707088296227551,
+      "learning_rate": 1.3286354924617167e-06,
+      "loss": 0.3096,
+      "step": 9381
+    },
+    {
+      "epoch": 0.76974196988965,
+      "grad_norm": 0.7108748668204429,
+      "learning_rate": 1.3277336241276079e-06,
+      "loss": 0.292,
+      "step": 9382
+    },
+    {
+      "epoch": 0.7698240144398408,
+      "grad_norm": 0.7344199469186936,
+      "learning_rate": 1.3268320151230518e-06,
+      "loss": 0.2793,
+      "step": 9383
+    },
+    {
+      "epoch": 0.7699060589900316,
+      "grad_norm": 0.7168324449805693,
+      "learning_rate": 1.325930665511716e-06,
+      "loss": 0.3076,
+      "step": 9384
+    },
+    {
+      "epoch": 0.7699881035402223,
+      "grad_norm": 0.7180154158553844,
+      "learning_rate": 1.3250295753572557e-06,
+      "loss": 0.3114,
+      "step": 9385
+    },
+    {
+      "epoch": 0.7700701480904131,
+      "grad_norm": 0.7014788992058931,
+      "learning_rate": 1.3241287447233003e-06,
+      "loss": 0.3278,
+      "step": 9386
+    },
+    {
+      "epoch": 0.7701521926406039,
+      "grad_norm": 0.7275701801128694,
+      "learning_rate": 1.3232281736734692e-06,
+      "loss": 0.3398,
+      "step": 9387
+    },
+    {
+      "epoch": 0.7702342371907946,
+      "grad_norm": 0.7323965117513407,
+      "learning_rate": 1.322327862271357e-06,
+      "loss": 0.3409,
+      "step": 9388
+    },
+    {
+      "epoch": 0.7703162817409853,
+      "grad_norm": 0.6675906045315905,
+      "learning_rate": 1.3214278105805412e-06,
+      "loss": 0.294,
+      "step": 9389
+    },
+    {
+      "epoch": 0.7703983262911761,
+      "grad_norm": 0.840671616209331,
+      "learning_rate": 1.320528018664583e-06,
+      "loss": 0.2849,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7704803708413669,
+      "grad_norm": 0.7387514484854402,
+      "learning_rate": 1.319628486587027e-06,
+      "loss": 0.3495,
+      "step": 9391
+    },
+    {
+      "epoch": 0.7705624153915576,
+      "grad_norm": 0.7641005465468312,
+      "learning_rate": 1.3187292144113917e-06,
+      "loss": 0.2962,
+      "step": 9392
+    },
+    {
+      "epoch": 0.7706444599417483,
+      "grad_norm": 0.8074664655158822,
+      "learning_rate": 1.3178302022011851e-06,
+      "loss": 0.3069,
+      "step": 9393
+    },
+    {
+      "epoch": 0.7707265044919391,
+      "grad_norm": 0.7184463673683307,
+      "learning_rate": 1.316931450019896e-06,
+      "loss": 0.2689,
+      "step": 9394
+    },
+    {
+      "epoch": 0.7708085490421299,
+      "grad_norm": 0.7760776415987473,
+      "learning_rate": 1.3160329579309883e-06,
+      "loss": 0.3082,
+      "step": 9395
+    },
+    {
+      "epoch": 0.7708905935923206,
+      "grad_norm": 0.6749117352712403,
+      "learning_rate": 1.3151347259979164e-06,
+      "loss": 0.2838,
+      "step": 9396
+    },
+    {
+      "epoch": 0.7709726381425114,
+      "grad_norm": 0.7440487473978237,
+      "learning_rate": 1.3142367542841084e-06,
+      "loss": 0.3461,
+      "step": 9397
+    },
+    {
+      "epoch": 0.7710546826927022,
+      "grad_norm": 0.6528065481952806,
+      "learning_rate": 1.3133390428529784e-06,
+      "loss": 0.2895,
+      "step": 9398
+    },
+    {
+      "epoch": 0.7711367272428928,
+      "grad_norm": 0.7695461405298927,
+      "learning_rate": 1.3124415917679244e-06,
+      "loss": 0.3429,
+      "step": 9399
+    },
+    {
+      "epoch": 0.7712187717930836,
+      "grad_norm": 0.6728451974861641,
+      "learning_rate": 1.3115444010923205e-06,
+      "loss": 0.2903,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7713008163432744,
+      "grad_norm": 0.6418926491579545,
+      "learning_rate": 1.310647470889525e-06,
+      "loss": 0.6374,
+      "step": 9401
+    },
+    {
+      "epoch": 0.7713828608934652,
+      "grad_norm": 0.7664640728376813,
+      "learning_rate": 1.3097508012228765e-06,
+      "loss": 0.2858,
+      "step": 9402
+    },
+    {
+      "epoch": 0.7714649054436559,
+      "grad_norm": 0.6469811520339486,
+      "learning_rate": 1.3088543921556974e-06,
+      "loss": 0.3262,
+      "step": 9403
+    },
+    {
+      "epoch": 0.7715469499938467,
+      "grad_norm": 0.6262552537846414,
+      "learning_rate": 1.3079582437512922e-06,
+      "loss": 0.2768,
+      "step": 9404
+    },
+    {
+      "epoch": 0.7716289945440374,
+      "grad_norm": 0.6552189960237065,
+      "learning_rate": 1.3070623560729429e-06,
+      "loss": 0.3297,
+      "step": 9405
+    },
+    {
+      "epoch": 0.7717110390942281,
+      "grad_norm": 0.6605208826202545,
+      "learning_rate": 1.3061667291839182e-06,
+      "loss": 0.31,
+      "step": 9406
+    },
+    {
+      "epoch": 0.7717930836444189,
+      "grad_norm": 0.740143318292247,
+      "learning_rate": 1.305271363147463e-06,
+      "loss": 0.2988,
+      "step": 9407
+    },
+    {
+      "epoch": 0.7718751281946097,
+      "grad_norm": 0.7348978931929337,
+      "learning_rate": 1.3043762580268078e-06,
+      "loss": 0.311,
+      "step": 9408
+    },
+    {
+      "epoch": 0.7719571727448005,
+      "grad_norm": 0.730445448272583,
+      "learning_rate": 1.3034814138851654e-06,
+      "loss": 0.3125,
+      "step": 9409
+    },
+    {
+      "epoch": 0.7720392172949911,
+      "grad_norm": 0.6369258452095655,
+      "learning_rate": 1.3025868307857252e-06,
+      "loss": 0.2949,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7721212618451819,
+      "grad_norm": 0.6451568375437282,
+      "learning_rate": 1.3016925087916627e-06,
+      "loss": 0.3227,
+      "step": 9411
+    },
+    {
+      "epoch": 0.7722033063953727,
+      "grad_norm": 0.6861094263420312,
+      "learning_rate": 1.3007984479661352e-06,
+      "loss": 0.2864,
+      "step": 9412
+    },
+    {
+      "epoch": 0.7722853509455634,
+      "grad_norm": 0.6545808396235185,
+      "learning_rate": 1.2999046483722777e-06,
+      "loss": 0.3043,
+      "step": 9413
+    },
+    {
+      "epoch": 0.7723673954957542,
+      "grad_norm": 0.7259407683990599,
+      "learning_rate": 1.2990111100732078e-06,
+      "loss": 0.3225,
+      "step": 9414
+    },
+    {
+      "epoch": 0.772449440045945,
+      "grad_norm": 0.7213591794388743,
+      "learning_rate": 1.2981178331320288e-06,
+      "loss": 0.2945,
+      "step": 9415
+    },
+    {
+      "epoch": 0.7725314845961357,
+      "grad_norm": 0.7241461177743153,
+      "learning_rate": 1.2972248176118197e-06,
+      "loss": 0.33,
+      "step": 9416
+    },
+    {
+      "epoch": 0.7726135291463264,
+      "grad_norm": 0.7361909852431496,
+      "learning_rate": 1.296332063575646e-06,
+      "loss": 0.2775,
+      "step": 9417
+    },
+    {
+      "epoch": 0.7726955736965172,
+      "grad_norm": 0.7173016389039947,
+      "learning_rate": 1.2954395710865502e-06,
+      "loss": 0.3556,
+      "step": 9418
+    },
+    {
+      "epoch": 0.772777618246708,
+      "grad_norm": 0.6691832644342685,
+      "learning_rate": 1.294547340207562e-06,
+      "loss": 0.3079,
+      "step": 9419
+    },
+    {
+      "epoch": 0.7728596627968987,
+      "grad_norm": 0.7093142633535597,
+      "learning_rate": 1.2936553710016858e-06,
+      "loss": 0.2907,
+      "step": 9420
+    },
+    {
+      "epoch": 0.7729417073470894,
+      "grad_norm": 0.7136708321757962,
+      "learning_rate": 1.2927636635319125e-06,
+      "loss": 0.3241,
+      "step": 9421
+    },
+    {
+      "epoch": 0.7730237518972802,
+      "grad_norm": 0.8103657550204946,
+      "learning_rate": 1.2918722178612142e-06,
+      "loss": 0.2953,
+      "step": 9422
+    },
+    {
+      "epoch": 0.773105796447471,
+      "grad_norm": 0.7692145795847605,
+      "learning_rate": 1.2909810340525408e-06,
+      "loss": 0.293,
+      "step": 9423
+    },
+    {
+      "epoch": 0.7731878409976617,
+      "grad_norm": 1.090311628755862,
+      "learning_rate": 1.2900901121688298e-06,
+      "loss": 0.3599,
+      "step": 9424
+    },
+    {
+      "epoch": 0.7732698855478525,
+      "grad_norm": 0.7996196852048367,
+      "learning_rate": 1.2891994522729928e-06,
+      "loss": 0.3059,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7733519300980433,
+      "grad_norm": 0.6816599368407352,
+      "learning_rate": 1.2883090544279291e-06,
+      "loss": 0.303,
+      "step": 9426
+    },
+    {
+      "epoch": 0.773433974648234,
+      "grad_norm": 0.8980641814284908,
+      "learning_rate": 1.2874189186965181e-06,
+      "loss": 0.2899,
+      "step": 9427
+    },
+    {
+      "epoch": 0.7735160191984247,
+      "grad_norm": 0.8026988289833895,
+      "learning_rate": 1.2865290451416179e-06,
+      "loss": 0.2761,
+      "step": 9428
+    },
+    {
+      "epoch": 0.7735980637486155,
+      "grad_norm": 0.7183965355871295,
+      "learning_rate": 1.2856394338260691e-06,
+      "loss": 0.2782,
+      "step": 9429
+    },
+    {
+      "epoch": 0.7736801082988063,
+      "grad_norm": 0.6741752848912461,
+      "learning_rate": 1.2847500848126976e-06,
+      "loss": 0.2586,
+      "step": 9430
+    },
+    {
+      "epoch": 0.773762152848997,
+      "grad_norm": 0.6133092552228466,
+      "learning_rate": 1.2838609981643041e-06,
+      "loss": 0.2549,
+      "step": 9431
+    },
+    {
+      "epoch": 0.7738441973991877,
+      "grad_norm": 0.7071532718167487,
+      "learning_rate": 1.2829721739436786e-06,
+      "loss": 0.2713,
+      "step": 9432
+    },
+    {
+      "epoch": 0.7739262419493785,
+      "grad_norm": 0.6697536128321161,
+      "learning_rate": 1.2820836122135843e-06,
+      "loss": 0.2644,
+      "step": 9433
+    },
+    {
+      "epoch": 0.7740082864995693,
+      "grad_norm": 0.6947180714669959,
+      "learning_rate": 1.2811953130367722e-06,
+      "loss": 0.2849,
+      "step": 9434
+    },
+    {
+      "epoch": 0.77409033104976,
+      "grad_norm": 0.7133139634168204,
+      "learning_rate": 1.2803072764759738e-06,
+      "loss": 0.2729,
+      "step": 9435
+    },
+    {
+      "epoch": 0.7741723755999508,
+      "grad_norm": 0.7723794815153069,
+      "learning_rate": 1.2794195025938982e-06,
+      "loss": 0.3161,
+      "step": 9436
+    },
+    {
+      "epoch": 0.7742544201501416,
+      "grad_norm": 0.7741032722507116,
+      "learning_rate": 1.2785319914532408e-06,
+      "loss": 0.3426,
+      "step": 9437
+    },
+    {
+      "epoch": 0.7743364647003322,
+      "grad_norm": 0.7396300172100171,
+      "learning_rate": 1.2776447431166738e-06,
+      "loss": 0.2933,
+      "step": 9438
+    },
+    {
+      "epoch": 0.774418509250523,
+      "grad_norm": 0.7948135966218227,
+      "learning_rate": 1.2767577576468549e-06,
+      "loss": 0.2964,
+      "step": 9439
+    },
+    {
+      "epoch": 0.7745005538007138,
+      "grad_norm": 0.6840064269370589,
+      "learning_rate": 1.275871035106423e-06,
+      "loss": 0.3096,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7745825983509046,
+      "grad_norm": 0.6495447543273203,
+      "learning_rate": 1.2749845755579947e-06,
+      "loss": 0.3113,
+      "step": 9441
+    },
+    {
+      "epoch": 0.7746646429010953,
+      "grad_norm": 0.6766981049614856,
+      "learning_rate": 1.2740983790641697e-06,
+      "loss": 0.3225,
+      "step": 9442
+    },
+    {
+      "epoch": 0.774746687451286,
+      "grad_norm": 0.6725192646461425,
+      "learning_rate": 1.2732124456875327e-06,
+      "loss": 0.3076,
+      "step": 9443
+    },
+    {
+      "epoch": 0.7748287320014768,
+      "grad_norm": 0.682036072345719,
+      "learning_rate": 1.2723267754906437e-06,
+      "loss": 0.2807,
+      "step": 9444
+    },
+    {
+      "epoch": 0.7749107765516675,
+      "grad_norm": 0.716233458917682,
+      "learning_rate": 1.2714413685360505e-06,
+      "loss": 0.2893,
+      "step": 9445
+    },
+    {
+      "epoch": 0.7749928211018583,
+      "grad_norm": 0.7850519489264427,
+      "learning_rate": 1.2705562248862757e-06,
+      "loss": 0.3299,
+      "step": 9446
+    },
+    {
+      "epoch": 0.7750748656520491,
+      "grad_norm": 0.6852465421585731,
+      "learning_rate": 1.2696713446038283e-06,
+      "loss": 0.311,
+      "step": 9447
+    },
+    {
+      "epoch": 0.7751569102022399,
+      "grad_norm": 0.6860776514836957,
+      "learning_rate": 1.2687867277511995e-06,
+      "loss": 0.3067,
+      "step": 9448
+    },
+    {
+      "epoch": 0.7752389547524305,
+      "grad_norm": 0.6685275864465627,
+      "learning_rate": 1.2679023743908552e-06,
+      "loss": 0.2795,
+      "step": 9449
+    },
+    {
+      "epoch": 0.7753209993026213,
+      "grad_norm": 0.7912960154317725,
+      "learning_rate": 1.2670182845852508e-06,
+      "loss": 0.3318,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7754030438528121,
+      "grad_norm": 0.6713730890441729,
+      "learning_rate": 1.266134458396816e-06,
+      "loss": 0.3361,
+      "step": 9451
+    },
+    {
+      "epoch": 0.7754850884030028,
+      "grad_norm": 0.6979109112831349,
+      "learning_rate": 1.2652508958879671e-06,
+      "loss": 0.3145,
+      "step": 9452
+    },
+    {
+      "epoch": 0.7755671329531936,
+      "grad_norm": 0.6790347865772011,
+      "learning_rate": 1.2643675971211012e-06,
+      "loss": 0.3001,
+      "step": 9453
+    },
+    {
+      "epoch": 0.7756491775033844,
+      "grad_norm": 0.7405207579929024,
+      "learning_rate": 1.2634845621585923e-06,
+      "loss": 0.2778,
+      "step": 9454
+    },
+    {
+      "epoch": 0.7757312220535751,
+      "grad_norm": 0.6927074388562803,
+      "learning_rate": 1.2626017910628018e-06,
+      "loss": 0.3462,
+      "step": 9455
+    },
+    {
+      "epoch": 0.7758132666037658,
+      "grad_norm": 0.8591569598593007,
+      "learning_rate": 1.2617192838960684e-06,
+      "loss": 0.3152,
+      "step": 9456
+    },
+    {
+      "epoch": 0.7758953111539566,
+      "grad_norm": 0.7373819944703198,
+      "learning_rate": 1.2608370407207115e-06,
+      "loss": 0.3784,
+      "step": 9457
+    },
+    {
+      "epoch": 0.7759773557041474,
+      "grad_norm": 0.7196285011302676,
+      "learning_rate": 1.2599550615990369e-06,
+      "loss": 0.294,
+      "step": 9458
+    },
+    {
+      "epoch": 0.7760594002543381,
+      "grad_norm": 0.6804280689339713,
+      "learning_rate": 1.2590733465933252e-06,
+      "loss": 0.3251,
+      "step": 9459
+    },
+    {
+      "epoch": 0.7761414448045288,
+      "grad_norm": 0.6619595791729395,
+      "learning_rate": 1.2581918957658457e-06,
+      "loss": 0.3242,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7762234893547196,
+      "grad_norm": 0.7578610422221058,
+      "learning_rate": 1.2573107091788406e-06,
+      "loss": 0.2944,
+      "step": 9461
+    },
+    {
+      "epoch": 0.7763055339049104,
+      "grad_norm": 0.7157448508944692,
+      "learning_rate": 1.2564297868945407e-06,
+      "loss": 0.3009,
+      "step": 9462
+    },
+    {
+      "epoch": 0.7763875784551011,
+      "grad_norm": 0.6860195461033018,
+      "learning_rate": 1.255549128975156e-06,
+      "loss": 0.3039,
+      "step": 9463
+    },
+    {
+      "epoch": 0.7764696230052919,
+      "grad_norm": 0.6741390977208243,
+      "learning_rate": 1.2546687354828747e-06,
+      "loss": 0.3089,
+      "step": 9464
+    },
+    {
+      "epoch": 0.7765516675554827,
+      "grad_norm": 0.8194516120997102,
+      "learning_rate": 1.2537886064798693e-06,
+      "loss": 0.2754,
+      "step": 9465
+    },
+    {
+      "epoch": 0.7766337121056733,
+      "grad_norm": 0.6659650894760413,
+      "learning_rate": 1.2529087420282958e-06,
+      "loss": 0.3347,
+      "step": 9466
+    },
+    {
+      "epoch": 0.7767157566558641,
+      "grad_norm": 0.7339893043553106,
+      "learning_rate": 1.2520291421902852e-06,
+      "loss": 0.3267,
+      "step": 9467
+    },
+    {
+      "epoch": 0.7767978012060549,
+      "grad_norm": 0.6718396223390917,
+      "learning_rate": 1.2511498070279564e-06,
+      "loss": 0.2818,
+      "step": 9468
+    },
+    {
+      "epoch": 0.7768798457562457,
+      "grad_norm": 0.6862738211009264,
+      "learning_rate": 1.250270736603405e-06,
+      "loss": 0.3246,
+      "step": 9469
+    },
+    {
+      "epoch": 0.7769618903064364,
+      "grad_norm": 0.6631596749794709,
+      "learning_rate": 1.2493919309787089e-06,
+      "loss": 0.2899,
+      "step": 9470
+    },
+    {
+      "epoch": 0.7770439348566271,
+      "grad_norm": 0.6670080221615525,
+      "learning_rate": 1.2485133902159297e-06,
+      "loss": 0.2881,
+      "step": 9471
+    },
+    {
+      "epoch": 0.7771259794068179,
+      "grad_norm": 0.727715766402037,
+      "learning_rate": 1.2476351143771065e-06,
+      "loss": 0.3239,
+      "step": 9472
+    },
+    {
+      "epoch": 0.7772080239570086,
+      "grad_norm": 0.6432346317706314,
+      "learning_rate": 1.2467571035242643e-06,
+      "loss": 0.3075,
+      "step": 9473
+    },
+    {
+      "epoch": 0.7772900685071994,
+      "grad_norm": 0.7043256229264242,
+      "learning_rate": 1.245879357719404e-06,
+      "loss": 0.2902,
+      "step": 9474
+    },
+    {
+      "epoch": 0.7773721130573902,
+      "grad_norm": 0.8501645342264728,
+      "learning_rate": 1.245001877024512e-06,
+      "loss": 0.3166,
+      "step": 9475
+    },
+    {
+      "epoch": 0.777454157607581,
+      "grad_norm": 0.6555193967848882,
+      "learning_rate": 1.2441246615015562e-06,
+      "loss": 0.3197,
+      "step": 9476
+    },
+    {
+      "epoch": 0.7775362021577716,
+      "grad_norm": 0.7536958757469431,
+      "learning_rate": 1.2432477112124807e-06,
+      "loss": 0.3138,
+      "step": 9477
+    },
+    {
+      "epoch": 0.7776182467079624,
+      "grad_norm": 0.6704368430312582,
+      "learning_rate": 1.242371026219218e-06,
+      "loss": 0.2837,
+      "step": 9478
+    },
+    {
+      "epoch": 0.7777002912581532,
+      "grad_norm": 0.7237558744286343,
+      "learning_rate": 1.2414946065836741e-06,
+      "loss": 0.303,
+      "step": 9479
+    },
+    {
+      "epoch": 0.7777823358083439,
+      "grad_norm": 0.7114135496973234,
+      "learning_rate": 1.2406184523677434e-06,
+      "loss": 0.2831,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7778643803585347,
+      "grad_norm": 0.6625244997079408,
+      "learning_rate": 1.2397425636332989e-06,
+      "loss": 0.3124,
+      "step": 9481
+    },
+    {
+      "epoch": 0.7779464249087255,
+      "grad_norm": 0.6470667678297466,
+      "learning_rate": 1.2388669404421915e-06,
+      "loss": 0.2637,
+      "step": 9482
+    },
+    {
+      "epoch": 0.7780284694589162,
+      "grad_norm": 0.6486086360448046,
+      "learning_rate": 1.2379915828562599e-06,
+      "loss": 0.2775,
+      "step": 9483
+    },
+    {
+      "epoch": 0.7781105140091069,
+      "grad_norm": 0.6606486354089672,
+      "learning_rate": 1.2371164909373185e-06,
+      "loss": 0.2685,
+      "step": 9484
+    },
+    {
+      "epoch": 0.7781925585592977,
+      "grad_norm": 0.6849682620149726,
+      "learning_rate": 1.2362416647471636e-06,
+      "loss": 0.3336,
+      "step": 9485
+    },
+    {
+      "epoch": 0.7782746031094885,
+      "grad_norm": 0.6531808298074107,
+      "learning_rate": 1.2353671043475767e-06,
+      "loss": 0.3066,
+      "step": 9486
+    },
+    {
+      "epoch": 0.7783566476596792,
+      "grad_norm": 0.6544563530183068,
+      "learning_rate": 1.2344928098003151e-06,
+      "loss": 0.3151,
+      "step": 9487
+    },
+    {
+      "epoch": 0.7784386922098699,
+      "grad_norm": 0.7358278499249498,
+      "learning_rate": 1.2336187811671213e-06,
+      "loss": 0.3289,
+      "step": 9488
+    },
+    {
+      "epoch": 0.7785207367600607,
+      "grad_norm": 0.7484987615520949,
+      "learning_rate": 1.23274501850972e-06,
+      "loss": 0.2896,
+      "step": 9489
+    },
+    {
+      "epoch": 0.7786027813102515,
+      "grad_norm": 0.6700905868323748,
+      "learning_rate": 1.2318715218898108e-06,
+      "loss": 0.2864,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7786848258604422,
+      "grad_norm": 0.691787996276545,
+      "learning_rate": 1.2309982913690827e-06,
+      "loss": 0.3174,
+      "step": 9491
+    },
+    {
+      "epoch": 0.778766870410633,
+      "grad_norm": 0.6547584822390023,
+      "learning_rate": 1.2301253270091983e-06,
+      "loss": 0.339,
+      "step": 9492
+    },
+    {
+      "epoch": 0.7788489149608238,
+      "grad_norm": 0.6824202785269682,
+      "learning_rate": 1.2292526288718065e-06,
+      "loss": 0.3063,
+      "step": 9493
+    },
+    {
+      "epoch": 0.7789309595110144,
+      "grad_norm": 0.6281799792090326,
+      "learning_rate": 1.2283801970185378e-06,
+      "loss": 0.3247,
+      "step": 9494
+    },
+    {
+      "epoch": 0.7790130040612052,
+      "grad_norm": 0.7389851795904141,
+      "learning_rate": 1.2275080315109984e-06,
+      "loss": 0.3144,
+      "step": 9495
+    },
+    {
+      "epoch": 0.779095048611396,
+      "grad_norm": 0.6473730926574172,
+      "learning_rate": 1.2266361324107818e-06,
+      "loss": 0.313,
+      "step": 9496
+    },
+    {
+      "epoch": 0.7791770931615868,
+      "grad_norm": 0.6537587758075698,
+      "learning_rate": 1.2257644997794593e-06,
+      "loss": 0.3364,
+      "step": 9497
+    },
+    {
+      "epoch": 0.7792591377117775,
+      "grad_norm": 0.6852850177916169,
+      "learning_rate": 1.224893133678583e-06,
+      "loss": 0.316,
+      "step": 9498
+    },
+    {
+      "epoch": 0.7793411822619682,
+      "grad_norm": 0.6699206881801049,
+      "learning_rate": 1.2240220341696896e-06,
+      "loss": 0.3221,
+      "step": 9499
+    },
+    {
+      "epoch": 0.779423226812159,
+      "grad_norm": 0.6932171531900697,
+      "learning_rate": 1.2231512013142921e-06,
+      "loss": 0.3059,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7795052713623497,
+      "grad_norm": 0.6711948490161517,
+      "learning_rate": 1.2222806351738885e-06,
+      "loss": 0.2886,
+      "step": 9501
+    },
+    {
+      "epoch": 0.7795873159125405,
+      "grad_norm": 0.6411665335639382,
+      "learning_rate": 1.2214103358099593e-06,
+      "loss": 0.2734,
+      "step": 9502
+    },
+    {
+      "epoch": 0.7796693604627313,
+      "grad_norm": 0.751426593137183,
+      "learning_rate": 1.2205403032839597e-06,
+      "loss": 0.3354,
+      "step": 9503
+    },
+    {
+      "epoch": 0.7797514050129221,
+      "grad_norm": 0.7318218321182025,
+      "learning_rate": 1.2196705376573326e-06,
+      "loss": 0.2891,
+      "step": 9504
+    },
+    {
+      "epoch": 0.7798334495631127,
+      "grad_norm": 0.6728346409093117,
+      "learning_rate": 1.2188010389914978e-06,
+      "loss": 0.2669,
+      "step": 9505
+    },
+    {
+      "epoch": 0.7799154941133035,
+      "grad_norm": 0.6613771662348439,
+      "learning_rate": 1.2179318073478579e-06,
+      "loss": 0.2737,
+      "step": 9506
+    },
+    {
+      "epoch": 0.7799975386634943,
+      "grad_norm": 0.7315991156499347,
+      "learning_rate": 1.2170628427877996e-06,
+      "loss": 0.2927,
+      "step": 9507
+    },
+    {
+      "epoch": 0.780079583213685,
+      "grad_norm": 0.7411721107028592,
+      "learning_rate": 1.2161941453726838e-06,
+      "loss": 0.3061,
+      "step": 9508
+    },
+    {
+      "epoch": 0.7801616277638758,
+      "grad_norm": 0.6841386781224476,
+      "learning_rate": 1.2153257151638598e-06,
+      "loss": 0.2879,
+      "step": 9509
+    },
+    {
+      "epoch": 0.7802436723140665,
+      "grad_norm": 0.6894908969356803,
+      "learning_rate": 1.214457552222652e-06,
+      "loss": 0.2844,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7803257168642573,
+      "grad_norm": 0.7304056016719347,
+      "learning_rate": 1.2135896566103712e-06,
+      "loss": 0.3343,
+      "step": 9511
+    },
+    {
+      "epoch": 0.780407761414448,
+      "grad_norm": 0.7341963287857188,
+      "learning_rate": 1.2127220283883057e-06,
+      "loss": 0.3022,
+      "step": 9512
+    },
+    {
+      "epoch": 0.7804898059646388,
+      "grad_norm": 0.6623593461984122,
+      "learning_rate": 1.211854667617724e-06,
+      "loss": 0.2996,
+      "step": 9513
+    },
+    {
+      "epoch": 0.7805718505148296,
+      "grad_norm": 0.7024091227599516,
+      "learning_rate": 1.2109875743598814e-06,
+      "loss": 0.3292,
+      "step": 9514
+    },
+    {
+      "epoch": 0.7806538950650203,
+      "grad_norm": 0.6508531247474956,
+      "learning_rate": 1.210120748676007e-06,
+      "loss": 0.293,
+      "step": 9515
+    },
+    {
+      "epoch": 0.780735939615211,
+      "grad_norm": 0.7992203546869692,
+      "learning_rate": 1.2092541906273165e-06,
+      "loss": 0.3489,
+      "step": 9516
+    },
+    {
+      "epoch": 0.7808179841654018,
+      "grad_norm": 0.7098234383916726,
+      "learning_rate": 1.2083879002750065e-06,
+      "loss": 0.3141,
+      "step": 9517
+    },
+    {
+      "epoch": 0.7809000287155926,
+      "grad_norm": 1.2456837757467565,
+      "learning_rate": 1.2075218776802493e-06,
+      "loss": 0.3324,
+      "step": 9518
+    },
+    {
+      "epoch": 0.7809820732657833,
+      "grad_norm": 0.7673006978192884,
+      "learning_rate": 1.2066561229042046e-06,
+      "loss": 0.2937,
+      "step": 9519
+    },
+    {
+      "epoch": 0.7810641178159741,
+      "grad_norm": 0.7037646672558758,
+      "learning_rate": 1.2057906360080113e-06,
+      "loss": 0.3111,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7811461623661649,
+      "grad_norm": 0.7326650897633415,
+      "learning_rate": 1.2049254170527857e-06,
+      "loss": 0.2989,
+      "step": 9521
+    },
+    {
+      "epoch": 0.7812282069163555,
+      "grad_norm": 0.6770416018904425,
+      "learning_rate": 1.2040604660996318e-06,
+      "loss": 0.2934,
+      "step": 9522
+    },
+    {
+      "epoch": 0.7813102514665463,
+      "grad_norm": 0.6667713105486341,
+      "learning_rate": 1.2031957832096276e-06,
+      "loss": 0.3039,
+      "step": 9523
+    },
+    {
+      "epoch": 0.7813922960167371,
+      "grad_norm": 0.8132795404644281,
+      "learning_rate": 1.202331368443838e-06,
+      "loss": 0.2835,
+      "step": 9524
+    },
+    {
+      "epoch": 0.7814743405669279,
+      "grad_norm": 0.7202512886037464,
+      "learning_rate": 1.2014672218633062e-06,
+      "loss": 0.3141,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7815563851171186,
+      "grad_norm": 0.9278549009467425,
+      "learning_rate": 1.2006033435290548e-06,
+      "loss": 0.3338,
+      "step": 9526
+    },
+    {
+      "epoch": 0.7816384296673093,
+      "grad_norm": 0.7530893852172502,
+      "learning_rate": 1.1997397335020927e-06,
+      "loss": 0.3096,
+      "step": 9527
+    },
+    {
+      "epoch": 0.7817204742175001,
+      "grad_norm": 0.7527084011929829,
+      "learning_rate": 1.1988763918434037e-06,
+      "loss": 0.3499,
+      "step": 9528
+    },
+    {
+      "epoch": 0.7818025187676908,
+      "grad_norm": 0.6764922012509791,
+      "learning_rate": 1.198013318613957e-06,
+      "loss": 0.3211,
+      "step": 9529
+    },
+    {
+      "epoch": 0.7818845633178816,
+      "grad_norm": 0.6975693328404183,
+      "learning_rate": 1.1971505138747024e-06,
+      "loss": 0.2737,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7819666078680724,
+      "grad_norm": 0.7265654872547649,
+      "learning_rate": 1.1962879776865676e-06,
+      "loss": 0.2882,
+      "step": 9531
+    },
+    {
+      "epoch": 0.7820486524182632,
+      "grad_norm": 0.6864282363043378,
+      "learning_rate": 1.1954257101104644e-06,
+      "loss": 0.3108,
+      "step": 9532
+    },
+    {
+      "epoch": 0.7821306969684538,
+      "grad_norm": 0.7613611775507109,
+      "learning_rate": 1.1945637112072867e-06,
+      "loss": 0.3079,
+      "step": 9533
+    },
+    {
+      "epoch": 0.7822127415186446,
+      "grad_norm": 0.7548069337842737,
+      "learning_rate": 1.1937019810379047e-06,
+      "loss": 0.3008,
+      "step": 9534
+    },
+    {
+      "epoch": 0.7822947860688354,
+      "grad_norm": 0.6414297007594181,
+      "learning_rate": 1.1928405196631742e-06,
+      "loss": 0.6056,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7823768306190261,
+      "grad_norm": 0.7193756768965877,
+      "learning_rate": 1.191979327143929e-06,
+      "loss": 0.2992,
+      "step": 9536
+    },
+    {
+      "epoch": 0.7824588751692169,
+      "grad_norm": 0.7702850571528689,
+      "learning_rate": 1.1911184035409863e-06,
+      "loss": 0.2944,
+      "step": 9537
+    },
+    {
+      "epoch": 0.7825409197194076,
+      "grad_norm": 0.7663066410445181,
+      "learning_rate": 1.1902577489151413e-06,
+      "loss": 0.3154,
+      "step": 9538
+    },
+    {
+      "epoch": 0.7826229642695984,
+      "grad_norm": 0.76165944330163,
+      "learning_rate": 1.189397363327175e-06,
+      "loss": 0.2865,
+      "step": 9539
+    },
+    {
+      "epoch": 0.7827050088197891,
+      "grad_norm": 0.7094536144078059,
+      "learning_rate": 1.1885372468378443e-06,
+      "loss": 0.3179,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7827870533699799,
+      "grad_norm": 0.7416343860459857,
+      "learning_rate": 1.1876773995078883e-06,
+      "loss": 0.3155,
+      "step": 9541
+    },
+    {
+      "epoch": 0.7828690979201707,
+      "grad_norm": 0.7200250986599923,
+      "learning_rate": 1.1868178213980296e-06,
+      "loss": 0.2969,
+      "step": 9542
+    },
+    {
+      "epoch": 0.7829511424703615,
+      "grad_norm": 0.6837374063455417,
+      "learning_rate": 1.1859585125689715e-06,
+      "loss": 0.2719,
+      "step": 9543
+    },
+    {
+      "epoch": 0.7830331870205521,
+      "grad_norm": 0.6569531242541486,
+      "learning_rate": 1.1850994730813937e-06,
+      "loss": 0.2618,
+      "step": 9544
+    },
+    {
+      "epoch": 0.7831152315707429,
+      "grad_norm": 0.7087149532167772,
+      "learning_rate": 1.1842407029959634e-06,
+      "loss": 0.3115,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7831972761209337,
+      "grad_norm": 0.9693731831225075,
+      "learning_rate": 1.1833822023733232e-06,
+      "loss": 0.3078,
+      "step": 9546
+    },
+    {
+      "epoch": 0.7832793206711244,
+      "grad_norm": 0.6441756557316005,
+      "learning_rate": 1.1825239712741e-06,
+      "loss": 0.3045,
+      "step": 9547
+    },
+    {
+      "epoch": 0.7833613652213152,
+      "grad_norm": 0.6354188624408107,
+      "learning_rate": 1.1816660097589016e-06,
+      "loss": 0.2776,
+      "step": 9548
+    },
+    {
+      "epoch": 0.783443409771506,
+      "grad_norm": 0.7530952722059109,
+      "learning_rate": 1.180808317888314e-06,
+      "loss": 0.2865,
+      "step": 9549
+    },
+    {
+      "epoch": 0.7835254543216967,
+      "grad_norm": 0.7084901157532715,
+      "learning_rate": 1.1799508957229066e-06,
+      "loss": 0.3113,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7836074988718874,
+      "grad_norm": 0.7005974571266372,
+      "learning_rate": 1.1790937433232313e-06,
+      "loss": 0.2975,
+      "step": 9551
+    },
+    {
+      "epoch": 0.7836895434220782,
+      "grad_norm": 0.6356441288606516,
+      "learning_rate": 1.1782368607498174e-06,
+      "loss": 0.274,
+      "step": 9552
+    },
+    {
+      "epoch": 0.783771587972269,
+      "grad_norm": 0.7201159154842803,
+      "learning_rate": 1.1773802480631758e-06,
+      "loss": 0.3287,
+      "step": 9553
+    },
+    {
+      "epoch": 0.7838536325224597,
+      "grad_norm": 0.7272091857902244,
+      "learning_rate": 1.1765239053237981e-06,
+      "loss": 0.3462,
+      "step": 9554
+    },
+    {
+      "epoch": 0.7839356770726504,
+      "grad_norm": 0.670226406921644,
+      "learning_rate": 1.1756678325921595e-06,
+      "loss": 0.3048,
+      "step": 9555
+    },
+    {
+      "epoch": 0.7840177216228412,
+      "grad_norm": 0.6642360082764814,
+      "learning_rate": 1.1748120299287158e-06,
+      "loss": 0.306,
+      "step": 9556
+    },
+    {
+      "epoch": 0.784099766173032,
+      "grad_norm": 0.7324084746116774,
+      "learning_rate": 1.1739564973938993e-06,
+      "loss": 0.3254,
+      "step": 9557
+    },
+    {
+      "epoch": 0.7841818107232227,
+      "grad_norm": 0.8029795081658178,
+      "learning_rate": 1.173101235048129e-06,
+      "loss": 0.2937,
+      "step": 9558
+    },
+    {
+      "epoch": 0.7842638552734135,
+      "grad_norm": 0.6933923616129102,
+      "learning_rate": 1.1722462429518006e-06,
+      "loss": 0.3076,
+      "step": 9559
+    },
+    {
+      "epoch": 0.7843458998236043,
+      "grad_norm": 0.668708195886214,
+      "learning_rate": 1.171391521165292e-06,
+      "loss": 0.3008,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7844279443737949,
+      "grad_norm": 0.735302710110953,
+      "learning_rate": 1.170537069748965e-06,
+      "loss": 0.3521,
+      "step": 9561
+    },
+    {
+      "epoch": 0.7845099889239857,
+      "grad_norm": 0.7321737197421424,
+      "learning_rate": 1.1696828887631557e-06,
+      "loss": 0.2928,
+      "step": 9562
+    },
+    {
+      "epoch": 0.7845920334741765,
+      "grad_norm": 0.7154487768907141,
+      "learning_rate": 1.1688289782681889e-06,
+      "loss": 0.2955,
+      "step": 9563
+    },
+    {
+      "epoch": 0.7846740780243673,
+      "grad_norm": 0.7703276047921173,
+      "learning_rate": 1.1679753383243626e-06,
+      "loss": 0.2761,
+      "step": 9564
+    },
+    {
+      "epoch": 0.784756122574558,
+      "grad_norm": 0.673878144513089,
+      "learning_rate": 1.1671219689919627e-06,
+      "loss": 0.3004,
+      "step": 9565
+    },
+    {
+      "epoch": 0.7848381671247487,
+      "grad_norm": 0.7039668937665268,
+      "learning_rate": 1.1662688703312503e-06,
+      "loss": 0.2869,
+      "step": 9566
+    },
+    {
+      "epoch": 0.7849202116749395,
+      "grad_norm": 0.7647827275431048,
+      "learning_rate": 1.1654160424024718e-06,
+      "loss": 0.2991,
+      "step": 9567
+    },
+    {
+      "epoch": 0.7850022562251302,
+      "grad_norm": 0.7630141987047323,
+      "learning_rate": 1.1645634852658504e-06,
+      "loss": 0.2682,
+      "step": 9568
+    },
+    {
+      "epoch": 0.785084300775321,
+      "grad_norm": 0.6368847503120773,
+      "learning_rate": 1.1637111989815953e-06,
+      "loss": 0.3007,
+      "step": 9569
+    },
+    {
+      "epoch": 0.7851663453255118,
+      "grad_norm": 0.7464328940369462,
+      "learning_rate": 1.1628591836098902e-06,
+      "loss": 0.277,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7852483898757026,
+      "grad_norm": 0.6080051610608936,
+      "learning_rate": 1.162007439210906e-06,
+      "loss": 0.3065,
+      "step": 9571
+    },
+    {
+      "epoch": 0.7853304344258932,
+      "grad_norm": 0.7341045821291764,
+      "learning_rate": 1.161155965844789e-06,
+      "loss": 0.313,
+      "step": 9572
+    },
+    {
+      "epoch": 0.785412478976084,
+      "grad_norm": 0.6337592222787742,
+      "learning_rate": 1.1603047635716701e-06,
+      "loss": 0.2805,
+      "step": 9573
+    },
+    {
+      "epoch": 0.7854945235262748,
+      "grad_norm": 0.7330662729832812,
+      "learning_rate": 1.1594538324516614e-06,
+      "loss": 0.3042,
+      "step": 9574
+    },
+    {
+      "epoch": 0.7855765680764655,
+      "grad_norm": 0.6693634646028594,
+      "learning_rate": 1.1586031725448515e-06,
+      "loss": 0.3004,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7856586126266563,
+      "grad_norm": 0.6746861486514295,
+      "learning_rate": 1.157752783911315e-06,
+      "loss": 0.2923,
+      "step": 9576
+    },
+    {
+      "epoch": 0.785740657176847,
+      "grad_norm": 0.6815858081018751,
+      "learning_rate": 1.1569026666111021e-06,
+      "loss": 0.2789,
+      "step": 9577
+    },
+    {
+      "epoch": 0.7858227017270378,
+      "grad_norm": 0.7394440059884214,
+      "learning_rate": 1.1560528207042492e-06,
+      "loss": 0.2855,
+      "step": 9578
+    },
+    {
+      "epoch": 0.7859047462772285,
+      "grad_norm": 0.5993192816594258,
+      "learning_rate": 1.1552032462507712e-06,
+      "loss": 0.6453,
+      "step": 9579
+    },
+    {
+      "epoch": 0.7859867908274193,
+      "grad_norm": 0.6191756958040064,
+      "learning_rate": 1.1543539433106631e-06,
+      "loss": 0.3525,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7860688353776101,
+      "grad_norm": 0.6808883229411661,
+      "learning_rate": 1.1535049119439007e-06,
+      "loss": 0.2945,
+      "step": 9581
+    },
+    {
+      "epoch": 0.7861508799278007,
+      "grad_norm": 0.6699051196553077,
+      "learning_rate": 1.1526561522104406e-06,
+      "loss": 0.2733,
+      "step": 9582
+    },
+    {
+      "epoch": 0.7862329244779915,
+      "grad_norm": 0.8801298633030374,
+      "learning_rate": 1.1518076641702214e-06,
+      "loss": 0.3015,
+      "step": 9583
+    },
+    {
+      "epoch": 0.7863149690281823,
+      "grad_norm": 0.7134631934537354,
+      "learning_rate": 1.1509594478831642e-06,
+      "loss": 0.2728,
+      "step": 9584
+    },
+    {
+      "epoch": 0.7863970135783731,
+      "grad_norm": 0.7158844308637997,
+      "learning_rate": 1.1501115034091647e-06,
+      "loss": 0.3179,
+      "step": 9585
+    },
+    {
+      "epoch": 0.7864790581285638,
+      "grad_norm": 0.7415499902566032,
+      "learning_rate": 1.149263830808106e-06,
+      "loss": 0.3134,
+      "step": 9586
+    },
+    {
+      "epoch": 0.7865611026787546,
+      "grad_norm": 0.8060991717623939,
+      "learning_rate": 1.1484164301398504e-06,
+      "loss": 0.2881,
+      "step": 9587
+    },
+    {
+      "epoch": 0.7866431472289453,
+      "grad_norm": 0.6895199003234622,
+      "learning_rate": 1.147569301464237e-06,
+      "loss": 0.2619,
+      "step": 9588
+    },
+    {
+      "epoch": 0.786725191779136,
+      "grad_norm": 0.9962710405300726,
+      "learning_rate": 1.1467224448410913e-06,
+      "loss": 0.3277,
+      "step": 9589
+    },
+    {
+      "epoch": 0.7868072363293268,
+      "grad_norm": 0.8268005161459283,
+      "learning_rate": 1.1458758603302145e-06,
+      "loss": 0.2992,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7868892808795176,
+      "grad_norm": 0.6775982666155976,
+      "learning_rate": 1.1450295479913925e-06,
+      "loss": 0.2964,
+      "step": 9591
+    },
+    {
+      "epoch": 0.7869713254297084,
+      "grad_norm": 0.700604773012843,
+      "learning_rate": 1.1441835078843915e-06,
+      "loss": 0.3728,
+      "step": 9592
+    },
+    {
+      "epoch": 0.787053369979899,
+      "grad_norm": 0.736130714002193,
+      "learning_rate": 1.1433377400689566e-06,
+      "loss": 0.2657,
+      "step": 9593
+    },
+    {
+      "epoch": 0.7871354145300898,
+      "grad_norm": 0.7079197295691078,
+      "learning_rate": 1.1424922446048125e-06,
+      "loss": 0.2901,
+      "step": 9594
+    },
+    {
+      "epoch": 0.7872174590802806,
+      "grad_norm": 0.6508326455707649,
+      "learning_rate": 1.1416470215516707e-06,
+      "loss": 0.257,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7872995036304713,
+      "grad_norm": 0.8100833227573343,
+      "learning_rate": 1.1408020709692152e-06,
+      "loss": 0.3288,
+      "step": 9596
+    },
+    {
+      "epoch": 0.7873815481806621,
+      "grad_norm": 0.708226885548412,
+      "learning_rate": 1.1399573929171187e-06,
+      "loss": 0.3002,
+      "step": 9597
+    },
+    {
+      "epoch": 0.7874635927308529,
+      "grad_norm": 0.647181427796152,
+      "learning_rate": 1.1391129874550278e-06,
+      "loss": 0.3062,
+      "step": 9598
+    },
+    {
+      "epoch": 0.7875456372810437,
+      "grad_norm": 0.9709615496345413,
+      "learning_rate": 1.1382688546425768e-06,
+      "loss": 0.3453,
+      "step": 9599
+    },
+    {
+      "epoch": 0.7876276818312343,
+      "grad_norm": 0.6234264724080593,
+      "learning_rate": 1.1374249945393727e-06,
+      "loss": 0.2702,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7877097263814251,
+      "grad_norm": 0.6976645788127996,
+      "learning_rate": 1.1365814072050108e-06,
+      "loss": 0.3346,
+      "step": 9601
+    },
+    {
+      "epoch": 0.7877917709316159,
+      "grad_norm": 0.6663091475818247,
+      "learning_rate": 1.1357380926990636e-06,
+      "loss": 0.3223,
+      "step": 9602
+    },
+    {
+      "epoch": 0.7878738154818066,
+      "grad_norm": 0.7201820501034318,
+      "learning_rate": 1.1348950510810825e-06,
+      "loss": 0.2922,
+      "step": 9603
+    },
+    {
+      "epoch": 0.7879558600319974,
+      "grad_norm": 0.7490777800008356,
+      "learning_rate": 1.1340522824106032e-06,
+      "loss": 0.2983,
+      "step": 9604
+    },
+    {
+      "epoch": 0.7880379045821881,
+      "grad_norm": 0.8471111119197429,
+      "learning_rate": 1.133209786747142e-06,
+      "loss": 0.2769,
+      "step": 9605
+    },
+    {
+      "epoch": 0.7881199491323789,
+      "grad_norm": 0.734753515454393,
+      "learning_rate": 1.1323675641501919e-06,
+      "loss": 0.2788,
+      "step": 9606
+    },
+    {
+      "epoch": 0.7882019936825696,
+      "grad_norm": 0.6764297272628402,
+      "learning_rate": 1.131525614679232e-06,
+      "loss": 0.3096,
+      "step": 9607
+    },
+    {
+      "epoch": 0.7882840382327604,
+      "grad_norm": 0.668225661054991,
+      "learning_rate": 1.1306839383937179e-06,
+      "loss": 0.3379,
+      "step": 9608
+    },
+    {
+      "epoch": 0.7883660827829512,
+      "grad_norm": 0.7188311963385298,
+      "learning_rate": 1.129842535353086e-06,
+      "loss": 0.2873,
+      "step": 9609
+    },
+    {
+      "epoch": 0.7884481273331418,
+      "grad_norm": 0.8488591155965975,
+      "learning_rate": 1.1290014056167586e-06,
+      "loss": 0.2976,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7885301718833326,
+      "grad_norm": 0.7889555593915561,
+      "learning_rate": 1.1281605492441306e-06,
+      "loss": 0.3101,
+      "step": 9611
+    },
+    {
+      "epoch": 0.7886122164335234,
+      "grad_norm": 0.7818326159633858,
+      "learning_rate": 1.1273199662945856e-06,
+      "loss": 0.2883,
+      "step": 9612
+    },
+    {
+      "epoch": 0.7886942609837142,
+      "grad_norm": 0.6880137389730324,
+      "learning_rate": 1.1264796568274811e-06,
+      "loss": 0.2862,
+      "step": 9613
+    },
+    {
+      "epoch": 0.7887763055339049,
+      "grad_norm": 0.6954698362464777,
+      "learning_rate": 1.1256396209021598e-06,
+      "loss": 0.2882,
+      "step": 9614
+    },
+    {
+      "epoch": 0.7888583500840957,
+      "grad_norm": 0.7560182165262116,
+      "learning_rate": 1.1247998585779451e-06,
+      "loss": 0.3076,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7889403946342864,
+      "grad_norm": 0.6900567105347082,
+      "learning_rate": 1.1239603699141377e-06,
+      "loss": 0.2975,
+      "step": 9616
+    },
+    {
+      "epoch": 0.7890224391844771,
+      "grad_norm": 0.7728564504956428,
+      "learning_rate": 1.1231211549700227e-06,
+      "loss": 0.3029,
+      "step": 9617
+    },
+    {
+      "epoch": 0.7891044837346679,
+      "grad_norm": 0.6634796530616577,
+      "learning_rate": 1.1222822138048617e-06,
+      "loss": 0.2806,
+      "step": 9618
+    },
+    {
+      "epoch": 0.7891865282848587,
+      "grad_norm": 0.6901813758868665,
+      "learning_rate": 1.1214435464779006e-06,
+      "loss": 0.3308,
+      "step": 9619
+    },
+    {
+      "epoch": 0.7892685728350495,
+      "grad_norm": 0.6414720997310577,
+      "learning_rate": 1.1206051530483663e-06,
+      "loss": 0.2672,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7893506173852401,
+      "grad_norm": 0.6968771996115808,
+      "learning_rate": 1.1197670335754634e-06,
+      "loss": 0.3325,
+      "step": 9621
+    },
+    {
+      "epoch": 0.7894326619354309,
+      "grad_norm": 0.7160687899750834,
+      "learning_rate": 1.1189291881183773e-06,
+      "loss": 0.3108,
+      "step": 9622
+    },
+    {
+      "epoch": 0.7895147064856217,
+      "grad_norm": 0.7174421178921622,
+      "learning_rate": 1.1180916167362777e-06,
+      "loss": 0.3565,
+      "step": 9623
+    },
+    {
+      "epoch": 0.7895967510358124,
+      "grad_norm": 1.0023055964435794,
+      "learning_rate": 1.1172543194883095e-06,
+      "loss": 0.2998,
+      "step": 9624
+    },
+    {
+      "epoch": 0.7896787955860032,
+      "grad_norm": 0.7142564740686341,
+      "learning_rate": 1.1164172964336051e-06,
+      "loss": 0.3101,
+      "step": 9625
+    },
+    {
+      "epoch": 0.789760840136194,
+      "grad_norm": 0.7832361962179645,
+      "learning_rate": 1.1155805476312704e-06,
+      "loss": 0.3007,
+      "step": 9626
+    },
+    {
+      "epoch": 0.7898428846863847,
+      "grad_norm": 0.7394870364968618,
+      "learning_rate": 1.114744073140397e-06,
+      "loss": 0.3194,
+      "step": 9627
+    },
+    {
+      "epoch": 0.7899249292365754,
+      "grad_norm": 0.7105820926959094,
+      "learning_rate": 1.113907873020056e-06,
+      "loss": 0.3123,
+      "step": 9628
+    },
+    {
+      "epoch": 0.7900069737867662,
+      "grad_norm": 0.6868488385048797,
+      "learning_rate": 1.1130719473292957e-06,
+      "loss": 0.2948,
+      "step": 9629
+    },
+    {
+      "epoch": 0.790089018336957,
+      "grad_norm": 0.6640233362843462,
+      "learning_rate": 1.1122362961271516e-06,
+      "loss": 0.3023,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7901710628871477,
+      "grad_norm": 0.7933732629800669,
+      "learning_rate": 1.1114009194726322e-06,
+      "loss": 0.2822,
+      "step": 9631
+    },
+    {
+      "epoch": 0.7902531074373385,
+      "grad_norm": 0.667246267229541,
+      "learning_rate": 1.110565817424733e-06,
+      "loss": 0.2864,
+      "step": 9632
+    },
+    {
+      "epoch": 0.7903351519875292,
+      "grad_norm": 0.7866296897359484,
+      "learning_rate": 1.109730990042428e-06,
+      "loss": 0.2852,
+      "step": 9633
+    },
+    {
+      "epoch": 0.79041719653772,
+      "grad_norm": 0.7036022507244026,
+      "learning_rate": 1.108896437384669e-06,
+      "loss": 0.3399,
+      "step": 9634
+    },
+    {
+      "epoch": 0.7904992410879107,
+      "grad_norm": 0.6794852446045092,
+      "learning_rate": 1.108062159510393e-06,
+      "loss": 0.3134,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7905812856381015,
+      "grad_norm": 0.6825751327571795,
+      "learning_rate": 1.1072281564785148e-06,
+      "loss": 0.3213,
+      "step": 9636
+    },
+    {
+      "epoch": 0.7906633301882923,
+      "grad_norm": 0.7254060269083571,
+      "learning_rate": 1.106394428347929e-06,
+      "loss": 0.3154,
+      "step": 9637
+    },
+    {
+      "epoch": 0.7907453747384829,
+      "grad_norm": 0.6607859319258728,
+      "learning_rate": 1.1055609751775138e-06,
+      "loss": 0.2829,
+      "step": 9638
+    },
+    {
+      "epoch": 0.7908274192886737,
+      "grad_norm": 0.7334292326628956,
+      "learning_rate": 1.1047277970261245e-06,
+      "loss": 0.3366,
+      "step": 9639
+    },
+    {
+      "epoch": 0.7909094638388645,
+      "grad_norm": 0.7321685100489616,
+      "learning_rate": 1.1038948939526e-06,
+      "loss": 0.3506,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7909915083890553,
+      "grad_norm": 0.7448389856182153,
+      "learning_rate": 1.1030622660157597e-06,
+      "loss": 0.2744,
+      "step": 9641
+    },
+    {
+      "epoch": 0.791073552939246,
+      "grad_norm": 0.6792834921590607,
+      "learning_rate": 1.1022299132744001e-06,
+      "loss": 0.2894,
+      "step": 9642
+    },
+    {
+      "epoch": 0.7911555974894368,
+      "grad_norm": 0.7078649760291994,
+      "learning_rate": 1.1013978357873034e-06,
+      "loss": 0.2781,
+      "step": 9643
+    },
+    {
+      "epoch": 0.7912376420396275,
+      "grad_norm": 0.6628914769666017,
+      "learning_rate": 1.1005660336132263e-06,
+      "loss": 0.3854,
+      "step": 9644
+    },
+    {
+      "epoch": 0.7913196865898182,
+      "grad_norm": 0.7726802502795505,
+      "learning_rate": 1.0997345068109104e-06,
+      "loss": 0.3007,
+      "step": 9645
+    },
+    {
+      "epoch": 0.791401731140009,
+      "grad_norm": 0.6595926134940389,
+      "learning_rate": 1.0989032554390795e-06,
+      "loss": 0.267,
+      "step": 9646
+    },
+    {
+      "epoch": 0.7914837756901998,
+      "grad_norm": 0.7806418853715403,
+      "learning_rate": 1.0980722795564319e-06,
+      "loss": 0.307,
+      "step": 9647
+    },
+    {
+      "epoch": 0.7915658202403906,
+      "grad_norm": 0.8020984070586159,
+      "learning_rate": 1.0972415792216518e-06,
+      "loss": 0.3115,
+      "step": 9648
+    },
+    {
+      "epoch": 0.7916478647905812,
+      "grad_norm": 0.691690440127166,
+      "learning_rate": 1.0964111544934014e-06,
+      "loss": 0.2651,
+      "step": 9649
+    },
+    {
+      "epoch": 0.791729909340772,
+      "grad_norm": 0.7003450064553851,
+      "learning_rate": 1.0955810054303223e-06,
+      "loss": 0.3319,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7918119538909628,
+      "grad_norm": 0.6732345953659882,
+      "learning_rate": 1.0947511320910409e-06,
+      "loss": 0.2973,
+      "step": 9651
+    },
+    {
+      "epoch": 0.7918939984411536,
+      "grad_norm": 0.6165510530489406,
+      "learning_rate": 1.093921534534159e-06,
+      "loss": 0.2689,
+      "step": 9652
+    },
+    {
+      "epoch": 0.7919760429913443,
+      "grad_norm": 0.7020258272228573,
+      "learning_rate": 1.0930922128182642e-06,
+      "loss": 0.2785,
+      "step": 9653
+    },
+    {
+      "epoch": 0.7920580875415351,
+      "grad_norm": 0.6081867175406243,
+      "learning_rate": 1.092263167001919e-06,
+      "loss": 0.2921,
+      "step": 9654
+    },
+    {
+      "epoch": 0.7921401320917258,
+      "grad_norm": 0.6603779971246475,
+      "learning_rate": 1.0914343971436703e-06,
+      "loss": 0.2917,
+      "step": 9655
+    },
+    {
+      "epoch": 0.7922221766419165,
+      "grad_norm": 0.6464052579368882,
+      "learning_rate": 1.0906059033020466e-06,
+      "loss": 0.2856,
+      "step": 9656
+    },
+    {
+      "epoch": 0.7923042211921073,
+      "grad_norm": 0.6367161417174462,
+      "learning_rate": 1.0897776855355518e-06,
+      "loss": 0.2971,
+      "step": 9657
+    },
+    {
+      "epoch": 0.7923862657422981,
+      "grad_norm": 0.6756396375477429,
+      "learning_rate": 1.0889497439026747e-06,
+      "loss": 0.2669,
+      "step": 9658
+    },
+    {
+      "epoch": 0.7924683102924889,
+      "grad_norm": 0.7057180937742957,
+      "learning_rate": 1.088122078461884e-06,
+      "loss": 0.2734,
+      "step": 9659
+    },
+    {
+      "epoch": 0.7925503548426795,
+      "grad_norm": 0.7171233817355803,
+      "learning_rate": 1.0872946892716264e-06,
+      "loss": 0.34,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7926323993928703,
+      "grad_norm": 0.70969892002735,
+      "learning_rate": 1.0864675763903327e-06,
+      "loss": 0.337,
+      "step": 9661
+    },
+    {
+      "epoch": 0.7927144439430611,
+      "grad_norm": 0.869901671489642,
+      "learning_rate": 1.0856407398764096e-06,
+      "loss": 0.3369,
+      "step": 9662
+    },
+    {
+      "epoch": 0.7927964884932518,
+      "grad_norm": 0.671226220348087,
+      "learning_rate": 1.08481417978825e-06,
+      "loss": 0.2679,
+      "step": 9663
+    },
+    {
+      "epoch": 0.7928785330434426,
+      "grad_norm": 0.733161868407346,
+      "learning_rate": 1.0839878961842232e-06,
+      "loss": 0.2902,
+      "step": 9664
+    },
+    {
+      "epoch": 0.7929605775936334,
+      "grad_norm": 0.6999938149780257,
+      "learning_rate": 1.083161889122678e-06,
+      "loss": 0.2892,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7930426221438241,
+      "grad_norm": 0.8170445810528847,
+      "learning_rate": 1.0823361586619485e-06,
+      "loss": 0.3044,
+      "step": 9666
+    },
+    {
+      "epoch": 0.7931246666940148,
+      "grad_norm": 0.7042764366602295,
+      "learning_rate": 1.081510704860344e-06,
+      "loss": 0.2974,
+      "step": 9667
+    },
+    {
+      "epoch": 0.7932067112442056,
+      "grad_norm": 0.7379183772248418,
+      "learning_rate": 1.0806855277761579e-06,
+      "loss": 0.2473,
+      "step": 9668
+    },
+    {
+      "epoch": 0.7932887557943964,
+      "grad_norm": 0.7500410434590674,
+      "learning_rate": 1.0798606274676643e-06,
+      "loss": 0.2993,
+      "step": 9669
+    },
+    {
+      "epoch": 0.7933708003445871,
+      "grad_norm": 0.7105970930085642,
+      "learning_rate": 1.0790360039931142e-06,
+      "loss": 0.2817,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7934528448947779,
+      "grad_norm": 0.7052506218025753,
+      "learning_rate": 1.0782116574107427e-06,
+      "loss": 0.3287,
+      "step": 9671
+    },
+    {
+      "epoch": 0.7935348894449686,
+      "grad_norm": 0.6981995014063488,
+      "learning_rate": 1.0773875877787621e-06,
+      "loss": 0.2996,
+      "step": 9672
+    },
+    {
+      "epoch": 0.7936169339951594,
+      "grad_norm": 0.6888654153079027,
+      "learning_rate": 1.0765637951553675e-06,
+      "loss": 0.2903,
+      "step": 9673
+    },
+    {
+      "epoch": 0.7936989785453501,
+      "grad_norm": 0.7922690748918657,
+      "learning_rate": 1.0757402795987359e-06,
+      "loss": 0.3131,
+      "step": 9674
+    },
+    {
+      "epoch": 0.7937810230955409,
+      "grad_norm": 0.8828637389576498,
+      "learning_rate": 1.07491704116702e-06,
+      "loss": 0.286,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7938630676457317,
+      "grad_norm": 0.6181893881132039,
+      "learning_rate": 1.0740940799183574e-06,
+      "loss": 0.2613,
+      "step": 9676
+    },
+    {
+      "epoch": 0.7939451121959223,
+      "grad_norm": 0.7878245225219478,
+      "learning_rate": 1.073271395910863e-06,
+      "loss": 0.3034,
+      "step": 9677
+    },
+    {
+      "epoch": 0.7940271567461131,
+      "grad_norm": 0.6454530264699171,
+      "learning_rate": 1.0724489892026336e-06,
+      "loss": 0.2827,
+      "step": 9678
+    },
+    {
+      "epoch": 0.7941092012963039,
+      "grad_norm": 0.6667697941905488,
+      "learning_rate": 1.071626859851747e-06,
+      "loss": 0.3018,
+      "step": 9679
+    },
+    {
+      "epoch": 0.7941912458464947,
+      "grad_norm": 0.6446193549550975,
+      "learning_rate": 1.070805007916259e-06,
+      "loss": 0.288,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7942732903966854,
+      "grad_norm": 0.6495007156037393,
+      "learning_rate": 1.0699834334542086e-06,
+      "loss": 0.3024,
+      "step": 9681
+    },
+    {
+      "epoch": 0.7943553349468762,
+      "grad_norm": 0.6785546445731079,
+      "learning_rate": 1.0691621365236154e-06,
+      "loss": 0.2931,
+      "step": 9682
+    },
+    {
+      "epoch": 0.7944373794970669,
+      "grad_norm": 0.6681383966409612,
+      "learning_rate": 1.0683411171824754e-06,
+      "loss": 0.2706,
+      "step": 9683
+    },
+    {
+      "epoch": 0.7945194240472576,
+      "grad_norm": 0.6187231342283733,
+      "learning_rate": 1.0675203754887703e-06,
+      "loss": 0.3156,
+      "step": 9684
+    },
+    {
+      "epoch": 0.7946014685974484,
+      "grad_norm": 0.6518623346990626,
+      "learning_rate": 1.0666999115004572e-06,
+      "loss": 0.2989,
+      "step": 9685
+    },
+    {
+      "epoch": 0.7946835131476392,
+      "grad_norm": 0.9841495842413076,
+      "learning_rate": 1.0658797252754766e-06,
+      "loss": 0.2827,
+      "step": 9686
+    },
+    {
+      "epoch": 0.79476555769783,
+      "grad_norm": 0.7088717682164004,
+      "learning_rate": 1.0650598168717507e-06,
+      "loss": 0.2974,
+      "step": 9687
+    },
+    {
+      "epoch": 0.7948476022480206,
+      "grad_norm": 0.7341065697917314,
+      "learning_rate": 1.0642401863471774e-06,
+      "loss": 0.2664,
+      "step": 9688
+    },
+    {
+      "epoch": 0.7949296467982114,
+      "grad_norm": 0.7534556582570283,
+      "learning_rate": 1.0634208337596403e-06,
+      "loss": 0.33,
+      "step": 9689
+    },
+    {
+      "epoch": 0.7950116913484022,
+      "grad_norm": 0.7532805179725515,
+      "learning_rate": 1.0626017591669973e-06,
+      "loss": 0.303,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7950937358985929,
+      "grad_norm": 0.6552303849006605,
+      "learning_rate": 1.0617829626270942e-06,
+      "loss": 0.3067,
+      "step": 9691
+    },
+    {
+      "epoch": 0.7951757804487837,
+      "grad_norm": 0.7465083557138046,
+      "learning_rate": 1.0609644441977507e-06,
+      "loss": 0.28,
+      "step": 9692
+    },
+    {
+      "epoch": 0.7952578249989745,
+      "grad_norm": 0.748594285574528,
+      "learning_rate": 1.0601462039367683e-06,
+      "loss": 0.3572,
+      "step": 9693
+    },
+    {
+      "epoch": 0.7953398695491652,
+      "grad_norm": 0.788036845522253,
+      "learning_rate": 1.0593282419019313e-06,
+      "loss": 0.2815,
+      "step": 9694
+    },
+    {
+      "epoch": 0.7954219140993559,
+      "grad_norm": 0.6623159516768088,
+      "learning_rate": 1.058510558151004e-06,
+      "loss": 0.2841,
+      "step": 9695
+    },
+    {
+      "epoch": 0.7955039586495467,
+      "grad_norm": 0.7489118829644695,
+      "learning_rate": 1.0576931527417273e-06,
+      "loss": 0.3332,
+      "step": 9696
+    },
+    {
+      "epoch": 0.7955860031997375,
+      "grad_norm": 0.698726292461689,
+      "learning_rate": 1.0568760257318273e-06,
+      "loss": 0.3024,
+      "step": 9697
+    },
+    {
+      "epoch": 0.7956680477499282,
+      "grad_norm": 0.6397241519608899,
+      "learning_rate": 1.0560591771790062e-06,
+      "loss": 0.3332,
+      "step": 9698
+    },
+    {
+      "epoch": 0.795750092300119,
+      "grad_norm": 0.752680867159014,
+      "learning_rate": 1.0552426071409494e-06,
+      "loss": 0.3158,
+      "step": 9699
+    },
+    {
+      "epoch": 0.7958321368503097,
+      "grad_norm": 0.7987157795931096,
+      "learning_rate": 1.0544263156753238e-06,
+      "loss": 0.3108,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7959141814005005,
+      "grad_norm": 0.7782564202505178,
+      "learning_rate": 1.053610302839771e-06,
+      "loss": 0.3388,
+      "step": 9701
+    },
+    {
+      "epoch": 0.7959962259506912,
+      "grad_norm": 0.6426458923844272,
+      "learning_rate": 1.0527945686919195e-06,
+      "loss": 0.2796,
+      "step": 9702
+    },
+    {
+      "epoch": 0.796078270500882,
+      "grad_norm": 0.6623357372305415,
+      "learning_rate": 1.0519791132893725e-06,
+      "loss": 0.2846,
+      "step": 9703
+    },
+    {
+      "epoch": 0.7961603150510728,
+      "grad_norm": 0.6929786531561708,
+      "learning_rate": 1.0511639366897193e-06,
+      "loss": 0.3051,
+      "step": 9704
+    },
+    {
+      "epoch": 0.7962423596012634,
+      "grad_norm": 0.6555037390926991,
+      "learning_rate": 1.0503490389505244e-06,
+      "loss": 0.3073,
+      "step": 9705
+    },
+    {
+      "epoch": 0.7963244041514542,
+      "grad_norm": 0.7660649132045481,
+      "learning_rate": 1.0495344201293335e-06,
+      "loss": 0.3027,
+      "step": 9706
+    },
+    {
+      "epoch": 0.796406448701645,
+      "grad_norm": 0.6562603273878431,
+      "learning_rate": 1.0487200802836762e-06,
+      "loss": 0.2586,
+      "step": 9707
+    },
+    {
+      "epoch": 0.7964884932518358,
+      "grad_norm": 0.7136175444693272,
+      "learning_rate": 1.0479060194710571e-06,
+      "loss": 0.2796,
+      "step": 9708
+    },
+    {
+      "epoch": 0.7965705378020265,
+      "grad_norm": 0.7177127369910634,
+      "learning_rate": 1.0470922377489656e-06,
+      "loss": 0.2859,
+      "step": 9709
+    },
+    {
+      "epoch": 0.7966525823522173,
+      "grad_norm": 0.6590590022644492,
+      "learning_rate": 1.0462787351748705e-06,
+      "loss": 0.3061,
+      "step": 9710
+    },
+    {
+      "epoch": 0.796734626902408,
+      "grad_norm": 0.7430890328813978,
+      "learning_rate": 1.045465511806218e-06,
+      "loss": 0.3081,
+      "step": 9711
+    },
+    {
+      "epoch": 0.7968166714525987,
+      "grad_norm": 0.6915932012984729,
+      "learning_rate": 1.044652567700437e-06,
+      "loss": 0.2829,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7968987160027895,
+      "grad_norm": 0.6820506918777985,
+      "learning_rate": 1.0438399029149393e-06,
+      "loss": 0.3248,
+      "step": 9713
+    },
+    {
+      "epoch": 0.7969807605529803,
+      "grad_norm": 0.6511126983495595,
+      "learning_rate": 1.0430275175071097e-06,
+      "loss": 0.2783,
+      "step": 9714
+    },
+    {
+      "epoch": 0.7970628051031711,
+      "grad_norm": 0.8600290328424022,
+      "learning_rate": 1.0422154115343207e-06,
+      "loss": 0.2699,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7971448496533617,
+      "grad_norm": 0.6847340755619638,
+      "learning_rate": 1.0414035850539194e-06,
+      "loss": 0.2882,
+      "step": 9716
+    },
+    {
+      "epoch": 0.7972268942035525,
+      "grad_norm": 0.7312363803572416,
+      "learning_rate": 1.0405920381232382e-06,
+      "loss": 0.3033,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7973089387537433,
+      "grad_norm": 0.6554196538556682,
+      "learning_rate": 1.039780770799585e-06,
+      "loss": 0.2836,
+      "step": 9718
+    },
+    {
+      "epoch": 0.797390983303934,
+      "grad_norm": 0.6830131626699958,
+      "learning_rate": 1.0389697831402518e-06,
+      "loss": 0.3014,
+      "step": 9719
+    },
+    {
+      "epoch": 0.7974730278541248,
+      "grad_norm": 0.6956710290113618,
+      "learning_rate": 1.0381590752025094e-06,
+      "loss": 0.3046,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7975550724043156,
+      "grad_norm": 0.7238205543691911,
+      "learning_rate": 1.037348647043606e-06,
+      "loss": 0.3071,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7976371169545063,
+      "grad_norm": 0.7460662999460426,
+      "learning_rate": 1.036538498720775e-06,
+      "loss": 0.3051,
+      "step": 9722
+    },
+    {
+      "epoch": 0.797719161504697,
+      "grad_norm": 0.6759811409598785,
+      "learning_rate": 1.035728630291229e-06,
+      "loss": 0.3039,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7978012060548878,
+      "grad_norm": 0.6811990365699983,
+      "learning_rate": 1.0349190418121563e-06,
+      "loss": 0.2984,
+      "step": 9724
+    },
+    {
+      "epoch": 0.7978832506050786,
+      "grad_norm": 0.6908456411908239,
+      "learning_rate": 1.0341097333407307e-06,
+      "loss": 0.3394,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7979652951552693,
+      "grad_norm": 0.6911655693450823,
+      "learning_rate": 1.0333007049341053e-06,
+      "loss": 0.2837,
+      "step": 9726
+    },
+    {
+      "epoch": 0.79804733970546,
+      "grad_norm": 0.6789092634400771,
+      "learning_rate": 1.0324919566494097e-06,
+      "loss": 0.3057,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7981293842556508,
+      "grad_norm": 0.7388128327476113,
+      "learning_rate": 1.0316834885437594e-06,
+      "loss": 0.3197,
+      "step": 9728
+    },
+    {
+      "epoch": 0.7982114288058416,
+      "grad_norm": 0.7292884806052073,
+      "learning_rate": 1.0308753006742439e-06,
+      "loss": 0.3152,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7982934733560323,
+      "grad_norm": 0.6879670056739093,
+      "learning_rate": 1.0300673930979383e-06,
+      "loss": 0.2904,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7983755179062231,
+      "grad_norm": 0.6319502062091653,
+      "learning_rate": 1.0292597658718962e-06,
+      "loss": 0.2659,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7984575624564139,
+      "grad_norm": 0.7951552401218481,
+      "learning_rate": 1.0284524190531497e-06,
+      "loss": 0.2863,
+      "step": 9732
+    },
+    {
+      "epoch": 0.7985396070066045,
+      "grad_norm": 0.841256180801021,
+      "learning_rate": 1.0276453526987112e-06,
+      "loss": 0.2983,
+      "step": 9733
+    },
+    {
+      "epoch": 0.7986216515567953,
+      "grad_norm": 0.7267228141893712,
+      "learning_rate": 1.0268385668655779e-06,
+      "loss": 0.3159,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7987036961069861,
+      "grad_norm": 0.6461241297753051,
+      "learning_rate": 1.0260320616107194e-06,
+      "loss": 0.2752,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7987857406571769,
+      "grad_norm": 0.6905544255990491,
+      "learning_rate": 1.0252258369910938e-06,
+      "loss": 0.3103,
+      "step": 9736
+    },
+    {
+      "epoch": 0.7988677852073676,
+      "grad_norm": 0.6801312996012808,
+      "learning_rate": 1.0244198930636329e-06,
+      "loss": 0.3508,
+      "step": 9737
+    },
+    {
+      "epoch": 0.7989498297575583,
+      "grad_norm": 0.6862616144605563,
+      "learning_rate": 1.0236142298852524e-06,
+      "loss": 0.2697,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7990318743077491,
+      "grad_norm": 0.7361527138373231,
+      "learning_rate": 1.0228088475128456e-06,
+      "loss": 0.3299,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7991139188579398,
+      "grad_norm": 0.691151808806186,
+      "learning_rate": 1.0220037460032877e-06,
+      "loss": 0.3312,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7991959634081306,
+      "grad_norm": 0.7222641393571314,
+      "learning_rate": 1.0211989254134358e-06,
+      "loss": 0.2748,
+      "step": 9741
+    },
+    {
+      "epoch": 0.7992780079583214,
+      "grad_norm": 0.6630360571728401,
+      "learning_rate": 1.020394385800122e-06,
+      "loss": 0.3343,
+      "step": 9742
+    },
+    {
+      "epoch": 0.7993600525085122,
+      "grad_norm": 0.6258425836827777,
+      "learning_rate": 1.0195901272201625e-06,
+      "loss": 0.2926,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7994420970587028,
+      "grad_norm": 0.6305658484180072,
+      "learning_rate": 1.0187861497303547e-06,
+      "loss": 0.3141,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7995241416088936,
+      "grad_norm": 0.7113479767628508,
+      "learning_rate": 1.0179824533874733e-06,
+      "loss": 0.3054,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7996061861590844,
+      "grad_norm": 0.703694630264181,
+      "learning_rate": 1.0171790382482717e-06,
+      "loss": 0.2958,
+      "step": 9746
+    },
+    {
+      "epoch": 0.7996882307092751,
+      "grad_norm": 0.7581092566660904,
+      "learning_rate": 1.0163759043694887e-06,
+      "loss": 0.3153,
+      "step": 9747
+    },
+    {
+      "epoch": 0.7997702752594659,
+      "grad_norm": 0.6707064729862992,
+      "learning_rate": 1.0155730518078378e-06,
+      "loss": 0.3186,
+      "step": 9748
+    },
+    {
+      "epoch": 0.7998523198096567,
+      "grad_norm": 0.6809130719065786,
+      "learning_rate": 1.0147704806200182e-06,
+      "loss": 0.2884,
+      "step": 9749
+    },
+    {
+      "epoch": 0.7999343643598474,
+      "grad_norm": 0.6590389507740514,
+      "learning_rate": 1.013968190862703e-06,
+      "loss": 0.2676,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8000164089100381,
+      "grad_norm": 0.6489121474698356,
+      "learning_rate": 1.013166182592551e-06,
+      "loss": 0.2947,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8000984534602289,
+      "grad_norm": 0.7200977102546422,
+      "learning_rate": 1.0123644558661972e-06,
+      "loss": 0.3535,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8001804980104197,
+      "grad_norm": 0.7723103388491903,
+      "learning_rate": 1.0115630107402586e-06,
+      "loss": 0.2934,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8002625425606104,
+      "grad_norm": 0.7459319688889009,
+      "learning_rate": 1.0107618472713338e-06,
+      "loss": 0.3064,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8003445871108011,
+      "grad_norm": 0.7017361125236685,
+      "learning_rate": 1.0099609655159964e-06,
+      "loss": 0.3157,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8004266316609919,
+      "grad_norm": 0.7324281092465356,
+      "learning_rate": 1.0091603655308069e-06,
+      "loss": 0.33,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8005086762111827,
+      "grad_norm": 0.716559584055515,
+      "learning_rate": 1.0083600473722987e-06,
+      "loss": 0.2643,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8005907207613734,
+      "grad_norm": 0.7103255845120597,
+      "learning_rate": 1.0075600110969914e-06,
+      "loss": 0.3452,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8006727653115642,
+      "grad_norm": 0.8175115105202304,
+      "learning_rate": 1.0067602567613832e-06,
+      "loss": 0.2946,
+      "step": 9759
+    },
+    {
+      "epoch": 0.800754809861755,
+      "grad_norm": 0.7094279332298987,
+      "learning_rate": 1.00596078442195e-06,
+      "loss": 0.2888,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8008368544119456,
+      "grad_norm": 0.7292073297580015,
+      "learning_rate": 1.0051615941351472e-06,
+      "loss": 0.2917,
+      "step": 9761
+    },
+    {
+      "epoch": 0.8009188989621364,
+      "grad_norm": 0.8867568727181516,
+      "learning_rate": 1.0043626859574167e-06,
+      "loss": 0.3051,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8010009435123272,
+      "grad_norm": 0.6566991392409652,
+      "learning_rate": 1.0035640599451719e-06,
+      "loss": 0.3091,
+      "step": 9763
+    },
+    {
+      "epoch": 0.801082988062518,
+      "grad_norm": 0.6280236108057966,
+      "learning_rate": 1.0027657161548144e-06,
+      "loss": 0.3019,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8011650326127087,
+      "grad_norm": 1.0925294907366019,
+      "learning_rate": 1.001967654642718e-06,
+      "loss": 0.315,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8012470771628994,
+      "grad_norm": 0.9080477333923538,
+      "learning_rate": 1.0011698754652427e-06,
+      "loss": 0.2989,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8013291217130902,
+      "grad_norm": 0.6806791517242843,
+      "learning_rate": 1.0003723786787272e-06,
+      "loss": 0.3237,
+      "step": 9767
+    },
+    {
+      "epoch": 0.801411166263281,
+      "grad_norm": 0.7621514496859785,
+      "learning_rate": 9.995751643394879e-07,
+      "loss": 0.2868,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8014932108134717,
+      "grad_norm": 0.699591762040652,
+      "learning_rate": 9.987782325038242e-07,
+      "loss": 0.3308,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8015752553636625,
+      "grad_norm": 0.7437175704568343,
+      "learning_rate": 9.979815832280115e-07,
+      "loss": 0.3174,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8016572999138533,
+      "grad_norm": 0.7265005103162883,
+      "learning_rate": 9.971852165683105e-07,
+      "loss": 0.3402,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8017393444640439,
+      "grad_norm": 0.6999851254031201,
+      "learning_rate": 9.963891325809594e-07,
+      "loss": 0.3318,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8018213890142347,
+      "grad_norm": 0.7205715110463815,
+      "learning_rate": 9.955933313221754e-07,
+      "loss": 0.2958,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8019034335644255,
+      "grad_norm": 0.705778142913246,
+      "learning_rate": 9.947978128481556e-07,
+      "loss": 0.3198,
+      "step": 9774
+    },
+    {
+      "epoch": 0.8019854781146163,
+      "grad_norm": 0.731759531252536,
+      "learning_rate": 9.940025772150808e-07,
+      "loss": 0.3035,
+      "step": 9775
+    },
+    {
+      "epoch": 0.802067522664807,
+      "grad_norm": 0.6596676324798115,
+      "learning_rate": 9.932076244791062e-07,
+      "loss": 0.308,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8021495672149977,
+      "grad_norm": 0.7233696228057126,
+      "learning_rate": 9.924129546963735e-07,
+      "loss": 0.2988,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8022316117651885,
+      "grad_norm": 0.7561149816515415,
+      "learning_rate": 9.916185679229973e-07,
+      "loss": 0.2711,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8023136563153792,
+      "grad_norm": 0.6980271607191585,
+      "learning_rate": 9.908244642150783e-07,
+      "loss": 0.3393,
+      "step": 9779
+    },
+    {
+      "epoch": 0.80239570086557,
+      "grad_norm": 1.0591153948226608,
+      "learning_rate": 9.900306436286962e-07,
+      "loss": 0.2953,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8024777454157608,
+      "grad_norm": 0.6753955976136956,
+      "learning_rate": 9.892371062199051e-07,
+      "loss": 0.264,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8025597899659516,
+      "grad_norm": 0.6872355797388457,
+      "learning_rate": 9.88443852044748e-07,
+      "loss": 0.2871,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8026418345161422,
+      "grad_norm": 0.8323191142725243,
+      "learning_rate": 9.876508811592395e-07,
+      "loss": 0.301,
+      "step": 9783
+    },
+    {
+      "epoch": 0.802723879066333,
+      "grad_norm": 0.7195435139664191,
+      "learning_rate": 9.868581936193794e-07,
+      "loss": 0.2998,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8028059236165238,
+      "grad_norm": 0.7074759420834195,
+      "learning_rate": 9.860657894811476e-07,
+      "loss": 0.3412,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8028879681667145,
+      "grad_norm": 0.7110081082637354,
+      "learning_rate": 9.852736688004988e-07,
+      "loss": 0.3054,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8029700127169053,
+      "grad_norm": 0.8953186980153235,
+      "learning_rate": 9.844818316333754e-07,
+      "loss": 0.3092,
+      "step": 9787
+    },
+    {
+      "epoch": 0.803052057267096,
+      "grad_norm": 0.7050655151403508,
+      "learning_rate": 9.836902780356932e-07,
+      "loss": 0.2654,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8031341018172868,
+      "grad_norm": 0.7524818726373084,
+      "learning_rate": 9.8289900806335e-07,
+      "loss": 0.2896,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8032161463674775,
+      "grad_norm": 0.6950350403705858,
+      "learning_rate": 9.821080217722262e-07,
+      "loss": 0.2994,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8032981909176683,
+      "grad_norm": 0.6470198952528978,
+      "learning_rate": 9.81317319218177e-07,
+      "loss": 0.3016,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8033802354678591,
+      "grad_norm": 0.6838308419005166,
+      "learning_rate": 9.805269004570435e-07,
+      "loss": 0.2856,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8034622800180498,
+      "grad_norm": 0.7965745518742563,
+      "learning_rate": 9.797367655446415e-07,
+      "loss": 0.3389,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8035443245682405,
+      "grad_norm": 0.6563726677985214,
+      "learning_rate": 9.789469145367696e-07,
+      "loss": 0.2707,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8036263691184313,
+      "grad_norm": 0.6478316969374982,
+      "learning_rate": 9.78157347489208e-07,
+      "loss": 0.3275,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8037084136686221,
+      "grad_norm": 0.6528130855580764,
+      "learning_rate": 9.773680644577116e-07,
+      "loss": 0.2956,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8037904582188128,
+      "grad_norm": 0.6895641087024712,
+      "learning_rate": 9.765790654980195e-07,
+      "loss": 0.3274,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8038725027690036,
+      "grad_norm": 0.8196761825169925,
+      "learning_rate": 9.75790350665851e-07,
+      "loss": 0.2996,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8039545473191944,
+      "grad_norm": 0.685141756807335,
+      "learning_rate": 9.750019200169014e-07,
+      "loss": 0.3097,
+      "step": 9799
+    },
+    {
+      "epoch": 0.804036591869385,
+      "grad_norm": 0.6889135550307863,
+      "learning_rate": 9.742137736068508e-07,
+      "loss": 0.3302,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8041186364195758,
+      "grad_norm": 0.7059144562459329,
+      "learning_rate": 9.734259114913557e-07,
+      "loss": 0.3068,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8042006809697666,
+      "grad_norm": 0.7223391291343132,
+      "learning_rate": 9.72638333726052e-07,
+      "loss": 0.2966,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8042827255199574,
+      "grad_norm": 0.6691725662348752,
+      "learning_rate": 9.718510403665599e-07,
+      "loss": 0.2635,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8043647700701481,
+      "grad_norm": 0.7167795288995025,
+      "learning_rate": 9.710640314684744e-07,
+      "loss": 0.3324,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8044468146203388,
+      "grad_norm": 0.6387484427288932,
+      "learning_rate": 9.702773070873755e-07,
+      "loss": 0.2543,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8045288591705296,
+      "grad_norm": 0.6481747222188106,
+      "learning_rate": 9.694908672788178e-07,
+      "loss": 0.3054,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8046109037207203,
+      "grad_norm": 0.6796133006510948,
+      "learning_rate": 9.687047120983396e-07,
+      "loss": 0.3419,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8046929482709111,
+      "grad_norm": 0.7343041708307997,
+      "learning_rate": 9.679188416014585e-07,
+      "loss": 0.2962,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8047749928211019,
+      "grad_norm": 0.6893584309659683,
+      "learning_rate": 9.671332558436697e-07,
+      "loss": 0.2848,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8048570373712927,
+      "grad_norm": 0.6793687699884136,
+      "learning_rate": 9.663479548804527e-07,
+      "loss": 0.2922,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8049390819214833,
+      "grad_norm": 0.7364870950316772,
+      "learning_rate": 9.65562938767261e-07,
+      "loss": 0.3048,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8050211264716741,
+      "grad_norm": 0.6809929878339055,
+      "learning_rate": 9.64778207559533e-07,
+      "loss": 0.2861,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8051031710218649,
+      "grad_norm": 0.7160690586838426,
+      "learning_rate": 9.63993761312686e-07,
+      "loss": 0.3133,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8051852155720556,
+      "grad_norm": 0.6466841562524096,
+      "learning_rate": 9.63209600082114e-07,
+      "loss": 0.3424,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8052672601222464,
+      "grad_norm": 0.719344173188712,
+      "learning_rate": 9.624257239231955e-07,
+      "loss": 0.3226,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8053493046724371,
+      "grad_norm": 0.6558356641966975,
+      "learning_rate": 9.616421328912862e-07,
+      "loss": 0.28,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8054313492226279,
+      "grad_norm": 0.7098968242875349,
+      "learning_rate": 9.608588270417196e-07,
+      "loss": 0.3131,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8055133937728186,
+      "grad_norm": 0.6844156285407984,
+      "learning_rate": 9.600758064298144e-07,
+      "loss": 0.3156,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8055954383230094,
+      "grad_norm": 0.7284685029721311,
+      "learning_rate": 9.59293071110864e-07,
+      "loss": 0.3178,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8056774828732002,
+      "grad_norm": 0.7212034662941212,
+      "learning_rate": 9.58510621140145e-07,
+      "loss": 0.2838,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8057595274233909,
+      "grad_norm": 0.7437386228851424,
+      "learning_rate": 9.577284565729139e-07,
+      "loss": 0.3164,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8058415719735816,
+      "grad_norm": 0.7196098185849947,
+      "learning_rate": 9.569465774644038e-07,
+      "loss": 0.3057,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8059236165237724,
+      "grad_norm": 0.7792155867896248,
+      "learning_rate": 9.561649838698317e-07,
+      "loss": 0.2954,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8060056610739632,
+      "grad_norm": 0.7217605930956181,
+      "learning_rate": 9.553836758443908e-07,
+      "loss": 0.3113,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8060877056241539,
+      "grad_norm": 0.6278993455399428,
+      "learning_rate": 9.546026534432563e-07,
+      "loss": 0.2928,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8061697501743447,
+      "grad_norm": 0.7425513582935067,
+      "learning_rate": 9.53821916721584e-07,
+      "loss": 0.2959,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8062517947245355,
+      "grad_norm": 0.7999019379598489,
+      "learning_rate": 9.530414657345061e-07,
+      "loss": 0.2818,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8063338392747261,
+      "grad_norm": 0.9870979950903441,
+      "learning_rate": 9.522613005371389e-07,
+      "loss": 0.3447,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8064158838249169,
+      "grad_norm": 0.6895303205989711,
+      "learning_rate": 9.514814211845747e-07,
+      "loss": 0.2856,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8064979283751077,
+      "grad_norm": 0.6956771369815642,
+      "learning_rate": 9.507018277318896e-07,
+      "loss": 0.2869,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8065799729252985,
+      "grad_norm": 0.9245295892625602,
+      "learning_rate": 9.49922520234135e-07,
+      "loss": 0.3447,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8066620174754892,
+      "grad_norm": 0.7074187552951492,
+      "learning_rate": 9.491434987463444e-07,
+      "loss": 0.2959,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8067440620256799,
+      "grad_norm": 0.6812669031564845,
+      "learning_rate": 9.483647633235315e-07,
+      "loss": 0.3102,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8068261065758707,
+      "grad_norm": 0.9703420755628027,
+      "learning_rate": 9.47586314020691e-07,
+      "loss": 0.3327,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8069081511260614,
+      "grad_norm": 0.7958605027791359,
+      "learning_rate": 9.468081508927923e-07,
+      "loss": 0.313,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8069901956762522,
+      "grad_norm": 0.7759277178364582,
+      "learning_rate": 9.460302739947924e-07,
+      "loss": 0.3255,
+      "step": 9836
+    },
+    {
+      "epoch": 0.807072240226443,
+      "grad_norm": 0.6975689778285528,
+      "learning_rate": 9.452526833816194e-07,
+      "loss": 0.2878,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8071542847766338,
+      "grad_norm": 0.6620731945567515,
+      "learning_rate": 9.444753791081874e-07,
+      "loss": 0.2918,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8072363293268244,
+      "grad_norm": 0.7303074673590996,
+      "learning_rate": 9.436983612293898e-07,
+      "loss": 0.2874,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8073183738770152,
+      "grad_norm": 0.6660505469852952,
+      "learning_rate": 9.429216298000959e-07,
+      "loss": 0.2557,
+      "step": 9840
+    },
+    {
+      "epoch": 0.807400418427206,
+      "grad_norm": 0.717309954091727,
+      "learning_rate": 9.421451848751595e-07,
+      "loss": 0.3294,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8074824629773967,
+      "grad_norm": 0.7040660663612565,
+      "learning_rate": 9.413690265094094e-07,
+      "loss": 0.3145,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8075645075275875,
+      "grad_norm": 0.7458908649745646,
+      "learning_rate": 9.405931547576591e-07,
+      "loss": 0.3258,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8076465520777782,
+      "grad_norm": 0.8413307116599421,
+      "learning_rate": 9.398175696746986e-07,
+      "loss": 0.273,
+      "step": 9844
+    },
+    {
+      "epoch": 0.807728596627969,
+      "grad_norm": 0.6743870438271684,
+      "learning_rate": 9.39042271315297e-07,
+      "loss": 0.3069,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8078106411781597,
+      "grad_norm": 0.6947556637756624,
+      "learning_rate": 9.382672597342063e-07,
+      "loss": 0.3375,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8078926857283505,
+      "grad_norm": 0.672092148224181,
+      "learning_rate": 9.374925349861552e-07,
+      "loss": 0.2917,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8079747302785413,
+      "grad_norm": 0.6860567183996107,
+      "learning_rate": 9.367180971258543e-07,
+      "loss": 0.338,
+      "step": 9848
+    },
+    {
+      "epoch": 0.808056774828732,
+      "grad_norm": 0.7367611415321796,
+      "learning_rate": 9.35943946207995e-07,
+      "loss": 0.3064,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8081388193789227,
+      "grad_norm": 0.7073835079459354,
+      "learning_rate": 9.351700822872428e-07,
+      "loss": 0.3013,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8082208639291135,
+      "grad_norm": 0.6490301534496024,
+      "learning_rate": 9.343965054182491e-07,
+      "loss": 0.3242,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8083029084793043,
+      "grad_norm": 0.7157015323278011,
+      "learning_rate": 9.336232156556435e-07,
+      "loss": 0.2992,
+      "step": 9852
+    },
+    {
+      "epoch": 0.808384953029495,
+      "grad_norm": 0.7316506111072479,
+      "learning_rate": 9.328502130540324e-07,
+      "loss": 0.278,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8084669975796858,
+      "grad_norm": 0.7172763416520466,
+      "learning_rate": 9.320774976680058e-07,
+      "loss": 0.3066,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8085490421298765,
+      "grad_norm": 0.7223924647716998,
+      "learning_rate": 9.313050695521292e-07,
+      "loss": 0.2985,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8086310866800672,
+      "grad_norm": 0.6851662807725613,
+      "learning_rate": 9.305329287609533e-07,
+      "loss": 0.2617,
+      "step": 9856
+    },
+    {
+      "epoch": 0.808713131230258,
+      "grad_norm": 0.6681299895446859,
+      "learning_rate": 9.297610753490027e-07,
+      "loss": 0.2811,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8087951757804488,
+      "grad_norm": 0.6649945691432954,
+      "learning_rate": 9.289895093707862e-07,
+      "loss": 0.2927,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8088772203306396,
+      "grad_norm": 0.7004445350191841,
+      "learning_rate": 9.282182308807907e-07,
+      "loss": 0.3014,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8089592648808303,
+      "grad_norm": 0.7171687005866211,
+      "learning_rate": 9.274472399334805e-07,
+      "loss": 0.3512,
+      "step": 9860
+    },
+    {
+      "epoch": 0.809041309431021,
+      "grad_norm": 0.7877707717277538,
+      "learning_rate": 9.266765365833031e-07,
+      "loss": 0.3,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8091233539812118,
+      "grad_norm": 0.6831735686691142,
+      "learning_rate": 9.259061208846859e-07,
+      "loss": 0.2935,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8092053985314025,
+      "grad_norm": 0.6926548792955355,
+      "learning_rate": 9.251359928920312e-07,
+      "loss": 0.2576,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8092874430815933,
+      "grad_norm": 0.7440237065165932,
+      "learning_rate": 9.243661526597275e-07,
+      "loss": 0.3123,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8093694876317841,
+      "grad_norm": 0.6703671856287735,
+      "learning_rate": 9.235966002421365e-07,
+      "loss": 0.2947,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8094515321819749,
+      "grad_norm": 0.7255419388336735,
+      "learning_rate": 9.228273356936046e-07,
+      "loss": 0.3001,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8095335767321655,
+      "grad_norm": 0.8571055350460317,
+      "learning_rate": 9.220583590684567e-07,
+      "loss": 0.3168,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8096156212823563,
+      "grad_norm": 0.6783254430930901,
+      "learning_rate": 9.21289670420995e-07,
+      "loss": 0.3125,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8096976658325471,
+      "grad_norm": 0.780027348645489,
+      "learning_rate": 9.205212698055049e-07,
+      "loss": 0.2761,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8097797103827378,
+      "grad_norm": 0.6556621257300823,
+      "learning_rate": 9.197531572762475e-07,
+      "loss": 0.3425,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8098617549329286,
+      "grad_norm": 0.6986245338494164,
+      "learning_rate": 9.189853328874676e-07,
+      "loss": 0.3224,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8099437994831193,
+      "grad_norm": 0.6812308031330732,
+      "learning_rate": 9.182177966933869e-07,
+      "loss": 0.3106,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8100258440333101,
+      "grad_norm": 0.7546753668351757,
+      "learning_rate": 9.174505487482066e-07,
+      "loss": 0.304,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8101078885835008,
+      "grad_norm": 0.6545160176254974,
+      "learning_rate": 9.166835891061088e-07,
+      "loss": 0.3117,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8101899331336916,
+      "grad_norm": 0.7390978281438431,
+      "learning_rate": 9.159169178212574e-07,
+      "loss": 0.3069,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8102719776838824,
+      "grad_norm": 0.8205958777880177,
+      "learning_rate": 9.151505349477901e-07,
+      "loss": 0.2822,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8103540222340732,
+      "grad_norm": 0.6660293619681592,
+      "learning_rate": 9.143844405398306e-07,
+      "loss": 0.2976,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8104360667842638,
+      "grad_norm": 0.6099372105089564,
+      "learning_rate": 9.136186346514769e-07,
+      "loss": 0.3019,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8105181113344546,
+      "grad_norm": 0.6860702532859134,
+      "learning_rate": 9.128531173368094e-07,
+      "loss": 0.3162,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8106001558846454,
+      "grad_norm": 0.8590212415022558,
+      "learning_rate": 9.120878886498897e-07,
+      "loss": 0.3108,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8106822004348361,
+      "grad_norm": 0.7603906380285638,
+      "learning_rate": 9.113229486447545e-07,
+      "loss": 0.3263,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8107642449850269,
+      "grad_norm": 0.7054667666563984,
+      "learning_rate": 9.105582973754252e-07,
+      "loss": 0.3142,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8108462895352176,
+      "grad_norm": 0.7248225064392722,
+      "learning_rate": 9.097939348958968e-07,
+      "loss": 0.3142,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8109283340854084,
+      "grad_norm": 0.658489341883135,
+      "learning_rate": 9.090298612601511e-07,
+      "loss": 0.3522,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8110103786355991,
+      "grad_norm": 0.7097749804285083,
+      "learning_rate": 9.082660765221424e-07,
+      "loss": 0.2932,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8110924231857899,
+      "grad_norm": 0.7008237798854138,
+      "learning_rate": 9.075025807358107e-07,
+      "loss": 0.333,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8111744677359807,
+      "grad_norm": 0.9484944051553038,
+      "learning_rate": 9.067393739550707e-07,
+      "loss": 0.2997,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8112565122861713,
+      "grad_norm": 0.7391896560312154,
+      "learning_rate": 9.059764562338208e-07,
+      "loss": 0.3213,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8113385568363621,
+      "grad_norm": 0.711461187236298,
+      "learning_rate": 9.052138276259348e-07,
+      "loss": 0.288,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8114206013865529,
+      "grad_norm": 0.6671805054266858,
+      "learning_rate": 9.044514881852706e-07,
+      "loss": 0.2761,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8115026459367437,
+      "grad_norm": 0.7388556586752176,
+      "learning_rate": 9.036894379656613e-07,
+      "loss": 0.3041,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8115846904869344,
+      "grad_norm": 0.7704781022605657,
+      "learning_rate": 9.029276770209228e-07,
+      "loss": 0.2976,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8116667350371252,
+      "grad_norm": 0.6979649702597693,
+      "learning_rate": 9.021662054048502e-07,
+      "loss": 0.3117,
+      "step": 9893
+    },
+    {
+      "epoch": 0.811748779587316,
+      "grad_norm": 0.68914797568284,
+      "learning_rate": 9.014050231712157e-07,
+      "loss": 0.292,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8118308241375066,
+      "grad_norm": 0.6217146797714456,
+      "learning_rate": 9.006441303737746e-07,
+      "loss": 0.302,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8119128686876974,
+      "grad_norm": 0.6392786375343115,
+      "learning_rate": 8.998835270662576e-07,
+      "loss": 0.2756,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8119949132378882,
+      "grad_norm": 1.1686975315029873,
+      "learning_rate": 8.991232133023798e-07,
+      "loss": 0.3099,
+      "step": 9897
+    },
+    {
+      "epoch": 0.812076957788079,
+      "grad_norm": 0.728643639357711,
+      "learning_rate": 8.983631891358308e-07,
+      "loss": 0.2938,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8121590023382697,
+      "grad_norm": 0.8272818163655001,
+      "learning_rate": 8.97603454620285e-07,
+      "loss": 0.2738,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8122410468884604,
+      "grad_norm": 0.6824884633625767,
+      "learning_rate": 8.968440098093922e-07,
+      "loss": 0.309,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8123230914386512,
+      "grad_norm": 0.8406569236029452,
+      "learning_rate": 8.960848547567819e-07,
+      "loss": 0.2889,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8124051359888419,
+      "grad_norm": 0.6826955907718023,
+      "learning_rate": 8.95325989516066e-07,
+      "loss": 0.2996,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8124871805390327,
+      "grad_norm": 0.677333772562814,
+      "learning_rate": 8.94567414140835e-07,
+      "loss": 0.2625,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8125692250892235,
+      "grad_norm": 0.6481731982853305,
+      "learning_rate": 8.938091286846562e-07,
+      "loss": 0.2882,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8126512696394143,
+      "grad_norm": 0.7414621503123551,
+      "learning_rate": 8.930511332010794e-07,
+      "loss": 0.2928,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8127333141896049,
+      "grad_norm": 0.7275096625833021,
+      "learning_rate": 8.922934277436346e-07,
+      "loss": 0.3072,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8128153587397957,
+      "grad_norm": 0.6796553651016553,
+      "learning_rate": 8.915360123658273e-07,
+      "loss": 0.3029,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8128974032899865,
+      "grad_norm": 0.7408342001434438,
+      "learning_rate": 8.90778887121147e-07,
+      "loss": 0.296,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8129794478401772,
+      "grad_norm": 0.6336706381750581,
+      "learning_rate": 8.900220520630586e-07,
+      "loss": 0.3074,
+      "step": 9909
+    },
+    {
+      "epoch": 0.813061492390368,
+      "grad_norm": 0.7411025171443586,
+      "learning_rate": 8.892655072450101e-07,
+      "loss": 0.3174,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8131435369405587,
+      "grad_norm": 0.6784130267520196,
+      "learning_rate": 8.885092527204281e-07,
+      "loss": 0.3298,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8132255814907495,
+      "grad_norm": 0.6137762171306191,
+      "learning_rate": 8.87753288542717e-07,
+      "loss": 0.2731,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8133076260409402,
+      "grad_norm": 0.8206709166171217,
+      "learning_rate": 8.869976147652603e-07,
+      "loss": 0.2933,
+      "step": 9913
+    },
+    {
+      "epoch": 0.813389670591131,
+      "grad_norm": 0.6911130162651796,
+      "learning_rate": 8.86242231441426e-07,
+      "loss": 0.3228,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8134717151413218,
+      "grad_norm": 0.734282301224056,
+      "learning_rate": 8.854871386245539e-07,
+      "loss": 0.3027,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8135537596915124,
+      "grad_norm": 0.768857852188212,
+      "learning_rate": 8.847323363679716e-07,
+      "loss": 0.2883,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8136358042417032,
+      "grad_norm": 0.679680829598957,
+      "learning_rate": 8.839778247249781e-07,
+      "loss": 0.336,
+      "step": 9917
+    },
+    {
+      "epoch": 0.813717848791894,
+      "grad_norm": 0.6807649305761625,
+      "learning_rate": 8.832236037488584e-07,
+      "loss": 0.2689,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8137998933420848,
+      "grad_norm": 0.6770006535684625,
+      "learning_rate": 8.824696734928745e-07,
+      "loss": 0.3684,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8138819378922755,
+      "grad_norm": 0.6626749203148308,
+      "learning_rate": 8.817160340102659e-07,
+      "loss": 0.3038,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8139639824424663,
+      "grad_norm": 0.6823917293015234,
+      "learning_rate": 8.809626853542558e-07,
+      "loss": 0.3032,
+      "step": 9921
+    },
+    {
+      "epoch": 0.814046026992657,
+      "grad_norm": 0.8178245280121827,
+      "learning_rate": 8.802096275780414e-07,
+      "loss": 0.3231,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8141280715428477,
+      "grad_norm": 0.639351637384253,
+      "learning_rate": 8.794568607348042e-07,
+      "loss": 0.3289,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8142101160930385,
+      "grad_norm": 0.7451168787616431,
+      "learning_rate": 8.787043848777049e-07,
+      "loss": 0.3344,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8142921606432293,
+      "grad_norm": 0.7479108138895423,
+      "learning_rate": 8.779522000598794e-07,
+      "loss": 0.3038,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8143742051934201,
+      "grad_norm": 0.6805849312540139,
+      "learning_rate": 8.772003063344481e-07,
+      "loss": 0.2921,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8144562497436107,
+      "grad_norm": 0.7070605704215939,
+      "learning_rate": 8.764487037545071e-07,
+      "loss": 0.2982,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8145382942938015,
+      "grad_norm": 0.809582963118306,
+      "learning_rate": 8.756973923731326e-07,
+      "loss": 0.335,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8146203388439923,
+      "grad_norm": 0.6360728126724817,
+      "learning_rate": 8.749463722433838e-07,
+      "loss": 0.291,
+      "step": 9929
+    },
+    {
+      "epoch": 0.814702383394183,
+      "grad_norm": 0.6553919895057515,
+      "learning_rate": 8.74195643418293e-07,
+      "loss": 0.311,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8147844279443738,
+      "grad_norm": 0.7368048887751896,
+      "learning_rate": 8.734452059508786e-07,
+      "loss": 0.2944,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8148664724945646,
+      "grad_norm": 0.7350025076938042,
+      "learning_rate": 8.726950598941325e-07,
+      "loss": 0.2753,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8149485170447553,
+      "grad_norm": 0.6843042222402481,
+      "learning_rate": 8.719452053010308e-07,
+      "loss": 0.2963,
+      "step": 9933
+    },
+    {
+      "epoch": 0.815030561594946,
+      "grad_norm": 0.6986262444140938,
+      "learning_rate": 8.711956422245271e-07,
+      "loss": 0.3402,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8151126061451368,
+      "grad_norm": 0.714887100739989,
+      "learning_rate": 8.704463707175526e-07,
+      "loss": 0.2738,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8151946506953276,
+      "grad_norm": 0.6962748875407442,
+      "learning_rate": 8.696973908330209e-07,
+      "loss": 0.3228,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8152766952455183,
+      "grad_norm": 0.8817106829733244,
+      "learning_rate": 8.689487026238247e-07,
+      "loss": 0.294,
+      "step": 9937
+    },
+    {
+      "epoch": 0.815358739795709,
+      "grad_norm": 0.6780573148245364,
+      "learning_rate": 8.682003061428334e-07,
+      "loss": 0.3068,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8154407843458998,
+      "grad_norm": 0.7113575215364757,
+      "learning_rate": 8.674522014428988e-07,
+      "loss": 0.3074,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8155228288960906,
+      "grad_norm": 0.7277756280233242,
+      "learning_rate": 8.667043885768505e-07,
+      "loss": 0.3185,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8156048734462813,
+      "grad_norm": 0.7787673477250345,
+      "learning_rate": 8.659568675974967e-07,
+      "loss": 0.2946,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8156869179964721,
+      "grad_norm": 0.7514232969988364,
+      "learning_rate": 8.652096385576281e-07,
+      "loss": 0.2877,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8157689625466629,
+      "grad_norm": 0.734604795464531,
+      "learning_rate": 8.644627015100105e-07,
+      "loss": 0.3074,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8158510070968535,
+      "grad_norm": 0.6800082317226397,
+      "learning_rate": 8.637160565073938e-07,
+      "loss": 0.2918,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8159330516470443,
+      "grad_norm": 0.6962407807194965,
+      "learning_rate": 8.629697036025025e-07,
+      "loss": 0.3225,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8160150961972351,
+      "grad_norm": 0.7272088598278299,
+      "learning_rate": 8.622236428480441e-07,
+      "loss": 0.2897,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8160971407474259,
+      "grad_norm": 0.7433580985666725,
+      "learning_rate": 8.614778742967056e-07,
+      "loss": 0.2466,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8161791852976166,
+      "grad_norm": 0.70744928593701,
+      "learning_rate": 8.607323980011495e-07,
+      "loss": 0.2847,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8162612298478074,
+      "grad_norm": 0.6811113633596454,
+      "learning_rate": 8.59987214014022e-07,
+      "loss": 0.2442,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8163432743979981,
+      "grad_norm": 0.6340678426392529,
+      "learning_rate": 8.592423223879448e-07,
+      "loss": 0.2619,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8164253189481888,
+      "grad_norm": 0.6765765729763842,
+      "learning_rate": 8.584977231755226e-07,
+      "loss": 0.3127,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8165073634983796,
+      "grad_norm": 0.6887212684088112,
+      "learning_rate": 8.577534164293383e-07,
+      "loss": 0.3216,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8165894080485704,
+      "grad_norm": 0.7667301631689426,
+      "learning_rate": 8.570094022019515e-07,
+      "loss": 0.2925,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8166714525987612,
+      "grad_norm": 0.707468684324811,
+      "learning_rate": 8.562656805459057e-07,
+      "loss": 0.285,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8167534971489518,
+      "grad_norm": 0.6605149187691395,
+      "learning_rate": 8.555222515137201e-07,
+      "loss": 0.2869,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8168355416991426,
+      "grad_norm": 0.7041623193061963,
+      "learning_rate": 8.547791151578933e-07,
+      "loss": 0.3007,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8169175862493334,
+      "grad_norm": 0.6740889398139586,
+      "learning_rate": 8.54036271530907e-07,
+      "loss": 0.3048,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8169996307995241,
+      "grad_norm": 0.8495202292083532,
+      "learning_rate": 8.532937206852165e-07,
+      "loss": 0.2891,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8170816753497149,
+      "grad_norm": 0.9066434042320891,
+      "learning_rate": 8.525514626732617e-07,
+      "loss": 0.3479,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8171637198999057,
+      "grad_norm": 0.6955111977068543,
+      "learning_rate": 8.5180949754746e-07,
+      "loss": 0.259,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8172457644500964,
+      "grad_norm": 0.6574939419062198,
+      "learning_rate": 8.510678253602061e-07,
+      "loss": 0.2901,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8173278090002871,
+      "grad_norm": 0.6245118111989612,
+      "learning_rate": 8.503264461638783e-07,
+      "loss": 0.2818,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8174098535504779,
+      "grad_norm": 0.6716286979834424,
+      "learning_rate": 8.495853600108278e-07,
+      "loss": 0.2742,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8174918981006687,
+      "grad_norm": 0.7684944941848302,
+      "learning_rate": 8.488445669533918e-07,
+      "loss": 0.3205,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8175739426508594,
+      "grad_norm": 0.7316136487369692,
+      "learning_rate": 8.481040670438839e-07,
+      "loss": 0.3245,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8176559872010501,
+      "grad_norm": 0.6745086820083381,
+      "learning_rate": 8.473638603345952e-07,
+      "loss": 0.307,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8177380317512409,
+      "grad_norm": 0.7065048150368797,
+      "learning_rate": 8.466239468777998e-07,
+      "loss": 0.3028,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8178200763014317,
+      "grad_norm": 0.7277758451816031,
+      "learning_rate": 8.458843267257489e-07,
+      "loss": 0.32,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8179021208516224,
+      "grad_norm": 0.6858935315355583,
+      "learning_rate": 8.45144999930671e-07,
+      "loss": 0.2868,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8179841654018132,
+      "grad_norm": 0.6927879787167,
+      "learning_rate": 8.444059665447796e-07,
+      "loss": 0.3003,
+      "step": 9970
+    },
+    {
+      "epoch": 0.818066209952004,
+      "grad_norm": 0.6417163029946833,
+      "learning_rate": 8.436672266202611e-07,
+      "loss": 0.2542,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8181482545021946,
+      "grad_norm": 0.7768328978909367,
+      "learning_rate": 8.42928780209285e-07,
+      "loss": 0.3429,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8182302990523854,
+      "grad_norm": 0.7237764144885434,
+      "learning_rate": 8.421906273640012e-07,
+      "loss": 0.3201,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8183123436025762,
+      "grad_norm": 0.8205095427411543,
+      "learning_rate": 8.414527681365342e-07,
+      "loss": 0.308,
+      "step": 9974
+    },
+    {
+      "epoch": 0.818394388152767,
+      "grad_norm": 0.7719295719181825,
+      "learning_rate": 8.407152025789922e-07,
+      "loss": 0.3017,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8184764327029577,
+      "grad_norm": 0.7414626548789739,
+      "learning_rate": 8.399779307434591e-07,
+      "loss": 0.3314,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8185584772531485,
+      "grad_norm": 0.7309156240620299,
+      "learning_rate": 8.392409526820011e-07,
+      "loss": 0.2811,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8186405218033392,
+      "grad_norm": 0.7815557466730577,
+      "learning_rate": 8.385042684466638e-07,
+      "loss": 0.3244,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8187225663535299,
+      "grad_norm": 0.7580607462806421,
+      "learning_rate": 8.377678780894671e-07,
+      "loss": 0.2914,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8188046109037207,
+      "grad_norm": 0.6933233868138898,
+      "learning_rate": 8.370317816624174e-07,
+      "loss": 0.2902,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8188866554539115,
+      "grad_norm": 0.7408589891522513,
+      "learning_rate": 8.362959792174941e-07,
+      "loss": 0.2923,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8189687000041023,
+      "grad_norm": 0.6956005918620577,
+      "learning_rate": 8.3556047080666e-07,
+      "loss": 0.2863,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8190507445542929,
+      "grad_norm": 0.6670572976685084,
+      "learning_rate": 8.348252564818549e-07,
+      "loss": 0.3439,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8191327891044837,
+      "grad_norm": 0.7449673275683398,
+      "learning_rate": 8.34090336294997e-07,
+      "loss": 0.3135,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8192148336546745,
+      "grad_norm": 0.7185149164108385,
+      "learning_rate": 8.333557102979878e-07,
+      "loss": 0.3106,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8192968782048652,
+      "grad_norm": 0.7295754044235199,
+      "learning_rate": 8.326213785427028e-07,
+      "loss": 0.2813,
+      "step": 9986
+    },
+    {
+      "epoch": 0.819378922755056,
+      "grad_norm": 0.6717306364870657,
+      "learning_rate": 8.318873410810008e-07,
+      "loss": 0.3144,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8194609673052468,
+      "grad_norm": 0.7455687987278119,
+      "learning_rate": 8.31153597964719e-07,
+      "loss": 0.3005,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8195430118554375,
+      "grad_norm": 0.7306154155515208,
+      "learning_rate": 8.304201492456715e-07,
+      "loss": 0.3392,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8196250564056282,
+      "grad_norm": 0.6679741066361008,
+      "learning_rate": 8.296869949756542e-07,
+      "loss": 0.2926,
+      "step": 9990
+    },
+    {
+      "epoch": 0.819707100955819,
+      "grad_norm": 0.6854579841453268,
+      "learning_rate": 8.289541352064423e-07,
+      "loss": 0.3514,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8197891455060098,
+      "grad_norm": 0.6929100028244656,
+      "learning_rate": 8.282215699897867e-07,
+      "loss": 0.2877,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8198711900562006,
+      "grad_norm": 0.7848311685635613,
+      "learning_rate": 8.27489299377423e-07,
+      "loss": 0.2878,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8199532346063912,
+      "grad_norm": 0.9385172372893661,
+      "learning_rate": 8.267573234210596e-07,
+      "loss": 0.2812,
+      "step": 9994
+    },
+    {
+      "epoch": 0.820035279156582,
+      "grad_norm": 0.9220113878802888,
+      "learning_rate": 8.260256421723905e-07,
+      "loss": 0.3147,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8201173237067728,
+      "grad_norm": 0.6513279481313993,
+      "learning_rate": 8.25294255683085e-07,
+      "loss": 0.3002,
+      "step": 9996
+    },
+    {
+      "epoch": 0.8201993682569635,
+      "grad_norm": 0.6686831763993318,
+      "learning_rate": 8.245631640047902e-07,
+      "loss": 0.2812,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8202814128071543,
+      "grad_norm": 0.6545384111288217,
+      "learning_rate": 8.23832367189138e-07,
+      "loss": 0.3124,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8203634573573451,
+      "grad_norm": 0.6723783770365839,
+      "learning_rate": 8.231018652877332e-07,
+      "loss": 0.3299,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8204455019075358,
+      "grad_norm": 0.6863357748478488,
+      "learning_rate": 8.223716583521641e-07,
+      "loss": 0.2985,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8205275464577265,
+      "grad_norm": 0.7485574266577424,
+      "learning_rate": 8.216417464339977e-07,
+      "loss": 0.3116,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8206095910079173,
+      "grad_norm": 0.6633523955826437,
+      "learning_rate": 8.209121295847766e-07,
+      "loss": 0.3157,
+      "step": 10002
+    },
+    {
+      "epoch": 0.8206916355581081,
+      "grad_norm": 0.698386958331463,
+      "learning_rate": 8.201828078560281e-07,
+      "loss": 0.2921,
+      "step": 10003
+    },
+    {
+      "epoch": 0.8207736801082988,
+      "grad_norm": 0.7620711634879784,
+      "learning_rate": 8.194537812992531e-07,
+      "loss": 0.2935,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8208557246584895,
+      "grad_norm": 0.7383258538397767,
+      "learning_rate": 8.187250499659361e-07,
+      "loss": 0.2746,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8209377692086803,
+      "grad_norm": 0.7333519450400655,
+      "learning_rate": 8.17996613907539e-07,
+      "loss": 0.3199,
+      "step": 10006
+    },
+    {
+      "epoch": 0.8210198137588711,
+      "grad_norm": 0.8454515591647837,
+      "learning_rate": 8.17268473175501e-07,
+      "loss": 0.3057,
+      "step": 10007
+    },
+    {
+      "epoch": 0.8211018583090618,
+      "grad_norm": 0.7658575961553649,
+      "learning_rate": 8.165406278212445e-07,
+      "loss": 0.3095,
+      "step": 10008
+    },
+    {
+      "epoch": 0.8211839028592526,
+      "grad_norm": 0.7384774402027977,
+      "learning_rate": 8.158130778961665e-07,
+      "loss": 0.3005,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8212659474094434,
+      "grad_norm": 0.7009785883860883,
+      "learning_rate": 8.150858234516474e-07,
+      "loss": 0.2932,
+      "step": 10010
+    },
+    {
+      "epoch": 0.821347991959634,
+      "grad_norm": 0.7157308573620244,
+      "learning_rate": 8.143588645390443e-07,
+      "loss": 0.2876,
+      "step": 10011
+    },
+    {
+      "epoch": 0.8214300365098248,
+      "grad_norm": 0.7148652542951242,
+      "learning_rate": 8.13632201209692e-07,
+      "loss": 0.2815,
+      "step": 10012
+    },
+    {
+      "epoch": 0.8215120810600156,
+      "grad_norm": 0.8476973258384153,
+      "learning_rate": 8.129058335149076e-07,
+      "loss": 0.2743,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8215941256102064,
+      "grad_norm": 0.699560603765606,
+      "learning_rate": 8.121797615059878e-07,
+      "loss": 0.3039,
+      "step": 10014
+    },
+    {
+      "epoch": 0.8216761701603971,
+      "grad_norm": 0.749944012344955,
+      "learning_rate": 8.114539852342035e-07,
+      "loss": 0.3237,
+      "step": 10015
+    },
+    {
+      "epoch": 0.8217582147105879,
+      "grad_norm": 0.6741289180561448,
+      "learning_rate": 8.107285047508106e-07,
+      "loss": 0.2766,
+      "step": 10016
+    },
+    {
+      "epoch": 0.8218402592607786,
+      "grad_norm": 0.6751374560792478,
+      "learning_rate": 8.10003320107039e-07,
+      "loss": 0.3108,
+      "step": 10017
+    },
+    {
+      "epoch": 0.8219223038109693,
+      "grad_norm": 0.7084130364762172,
+      "learning_rate": 8.092784313541013e-07,
+      "loss": 0.3285,
+      "step": 10018
+    },
+    {
+      "epoch": 0.8220043483611601,
+      "grad_norm": 0.7069734549006289,
+      "learning_rate": 8.085538385431885e-07,
+      "loss": 0.2842,
+      "step": 10019
+    },
+    {
+      "epoch": 0.8220863929113509,
+      "grad_norm": 0.7181980498869606,
+      "learning_rate": 8.078295417254689e-07,
+      "loss": 0.266,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8221684374615417,
+      "grad_norm": 0.7125202934871537,
+      "learning_rate": 8.071055409520928e-07,
+      "loss": 0.2757,
+      "step": 10021
+    },
+    {
+      "epoch": 0.8222504820117323,
+      "grad_norm": 0.793066012266633,
+      "learning_rate": 8.063818362741859e-07,
+      "loss": 0.3002,
+      "step": 10022
+    },
+    {
+      "epoch": 0.8223325265619231,
+      "grad_norm": 0.7507675310819187,
+      "learning_rate": 8.056584277428576e-07,
+      "loss": 0.3077,
+      "step": 10023
+    },
+    {
+      "epoch": 0.8224145711121139,
+      "grad_norm": 0.6894603618893044,
+      "learning_rate": 8.049353154091921e-07,
+      "loss": 0.3202,
+      "step": 10024
+    },
+    {
+      "epoch": 0.8224966156623046,
+      "grad_norm": 0.7105392210491025,
+      "learning_rate": 8.042124993242534e-07,
+      "loss": 0.3014,
+      "step": 10025
+    },
+    {
+      "epoch": 0.8225786602124954,
+      "grad_norm": 0.7151488470550648,
+      "learning_rate": 8.034899795390871e-07,
+      "loss": 0.3021,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8226607047626862,
+      "grad_norm": 0.7257186107197838,
+      "learning_rate": 8.027677561047176e-07,
+      "loss": 0.2998,
+      "step": 10027
+    },
+    {
+      "epoch": 0.8227427493128769,
+      "grad_norm": 0.7068769679270152,
+      "learning_rate": 8.020458290721445e-07,
+      "loss": 0.3002,
+      "step": 10028
+    },
+    {
+      "epoch": 0.8228247938630676,
+      "grad_norm": 0.7466931655054855,
+      "learning_rate": 8.013241984923514e-07,
+      "loss": 0.2743,
+      "step": 10029
+    },
+    {
+      "epoch": 0.8229068384132584,
+      "grad_norm": 0.6412637509317874,
+      "learning_rate": 8.006028644162966e-07,
+      "loss": 0.2753,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8229888829634492,
+      "grad_norm": 0.6714349268387522,
+      "learning_rate": 7.99881826894921e-07,
+      "loss": 0.2867,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8230709275136399,
+      "grad_norm": 0.6845160274729521,
+      "learning_rate": 7.991610859791438e-07,
+      "loss": 0.2739,
+      "step": 10032
+    },
+    {
+      "epoch": 0.8231529720638306,
+      "grad_norm": 0.6878365454070263,
+      "learning_rate": 7.984406417198604e-07,
+      "loss": 0.2986,
+      "step": 10033
+    },
+    {
+      "epoch": 0.8232350166140214,
+      "grad_norm": 1.0182290434375763,
+      "learning_rate": 7.977204941679495e-07,
+      "loss": 0.3034,
+      "step": 10034
+    },
+    {
+      "epoch": 0.8233170611642122,
+      "grad_norm": 1.2106947018653997,
+      "learning_rate": 7.97000643374265e-07,
+      "loss": 0.2869,
+      "step": 10035
+    },
+    {
+      "epoch": 0.8233991057144029,
+      "grad_norm": 0.6512151233628316,
+      "learning_rate": 7.962810893896433e-07,
+      "loss": 0.3051,
+      "step": 10036
+    },
+    {
+      "epoch": 0.8234811502645937,
+      "grad_norm": 0.6869083870096141,
+      "learning_rate": 7.955618322648961e-07,
+      "loss": 0.3238,
+      "step": 10037
+    },
+    {
+      "epoch": 0.8235631948147845,
+      "grad_norm": 0.7423027061028878,
+      "learning_rate": 7.948428720508189e-07,
+      "loss": 0.3334,
+      "step": 10038
+    },
+    {
+      "epoch": 0.8236452393649751,
+      "grad_norm": 0.7145317009147976,
+      "learning_rate": 7.941242087981816e-07,
+      "loss": 0.3143,
+      "step": 10039
+    },
+    {
+      "epoch": 0.8237272839151659,
+      "grad_norm": 0.747925423864394,
+      "learning_rate": 7.93405842557734e-07,
+      "loss": 0.288,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8238093284653567,
+      "grad_norm": 0.8128863579951375,
+      "learning_rate": 7.92687773380208e-07,
+      "loss": 0.322,
+      "step": 10041
+    },
+    {
+      "epoch": 0.8238913730155475,
+      "grad_norm": 0.6694067836607709,
+      "learning_rate": 7.919700013163128e-07,
+      "loss": 0.2659,
+      "step": 10042
+    },
+    {
+      "epoch": 0.8239734175657382,
+      "grad_norm": 0.6568838029651366,
+      "learning_rate": 7.912525264167342e-07,
+      "loss": 0.3081,
+      "step": 10043
+    },
+    {
+      "epoch": 0.824055462115929,
+      "grad_norm": 0.7027557550168749,
+      "learning_rate": 7.905353487321404e-07,
+      "loss": 0.304,
+      "step": 10044
+    },
+    {
+      "epoch": 0.8241375066661197,
+      "grad_norm": 0.7014344359743985,
+      "learning_rate": 7.898184683131782e-07,
+      "loss": 0.3283,
+      "step": 10045
+    },
+    {
+      "epoch": 0.8242195512163104,
+      "grad_norm": 0.6989999338091103,
+      "learning_rate": 7.891018852104709e-07,
+      "loss": 0.3075,
+      "step": 10046
+    },
+    {
+      "epoch": 0.8243015957665012,
+      "grad_norm": 0.7039863882998292,
+      "learning_rate": 7.883855994746237e-07,
+      "loss": 0.3055,
+      "step": 10047
+    },
+    {
+      "epoch": 0.824383640316692,
+      "grad_norm": 0.6379750470291076,
+      "learning_rate": 7.876696111562182e-07,
+      "loss": 0.281,
+      "step": 10048
+    },
+    {
+      "epoch": 0.8244656848668828,
+      "grad_norm": 0.7108308013156872,
+      "learning_rate": 7.869539203058169e-07,
+      "loss": 0.3306,
+      "step": 10049
+    },
+    {
+      "epoch": 0.8245477294170734,
+      "grad_norm": 0.6493965404672631,
+      "learning_rate": 7.862385269739625e-07,
+      "loss": 0.2741,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8246297739672642,
+      "grad_norm": 0.652930706621255,
+      "learning_rate": 7.855234312111732e-07,
+      "loss": 0.3131,
+      "step": 10051
+    },
+    {
+      "epoch": 0.824711818517455,
+      "grad_norm": 0.6301630730606906,
+      "learning_rate": 7.848086330679483e-07,
+      "loss": 0.3051,
+      "step": 10052
+    },
+    {
+      "epoch": 0.8247938630676457,
+      "grad_norm": 0.7162958636734652,
+      "learning_rate": 7.840941325947637e-07,
+      "loss": 0.3087,
+      "step": 10053
+    },
+    {
+      "epoch": 0.8248759076178365,
+      "grad_norm": 0.7232080751071915,
+      "learning_rate": 7.833799298420786e-07,
+      "loss": 0.2685,
+      "step": 10054
+    },
+    {
+      "epoch": 0.8249579521680273,
+      "grad_norm": 0.6881345156765095,
+      "learning_rate": 7.826660248603296e-07,
+      "loss": 0.2823,
+      "step": 10055
+    },
+    {
+      "epoch": 0.825039996718218,
+      "grad_norm": 0.6818179070096816,
+      "learning_rate": 7.819524176999288e-07,
+      "loss": 0.3118,
+      "step": 10056
+    },
+    {
+      "epoch": 0.8251220412684087,
+      "grad_norm": 0.7040230251822188,
+      "learning_rate": 7.812391084112731e-07,
+      "loss": 0.3044,
+      "step": 10057
+    },
+    {
+      "epoch": 0.8252040858185995,
+      "grad_norm": 0.6719861774037399,
+      "learning_rate": 7.80526097044732e-07,
+      "loss": 0.2977,
+      "step": 10058
+    },
+    {
+      "epoch": 0.8252861303687903,
+      "grad_norm": 0.8154609876319086,
+      "learning_rate": 7.798133836506588e-07,
+      "loss": 0.3021,
+      "step": 10059
+    },
+    {
+      "epoch": 0.825368174918981,
+      "grad_norm": 0.848608694969285,
+      "learning_rate": 7.791009682793855e-07,
+      "loss": 0.3367,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8254502194691717,
+      "grad_norm": 0.694968435894115,
+      "learning_rate": 7.783888509812193e-07,
+      "loss": 0.3616,
+      "step": 10061
+    },
+    {
+      "epoch": 0.8255322640193625,
+      "grad_norm": 0.6289032230201995,
+      "learning_rate": 7.776770318064497e-07,
+      "loss": 0.3076,
+      "step": 10062
+    },
+    {
+      "epoch": 0.8256143085695533,
+      "grad_norm": 0.6826286933538148,
+      "learning_rate": 7.769655108053459e-07,
+      "loss": 0.2942,
+      "step": 10063
+    },
+    {
+      "epoch": 0.825696353119744,
+      "grad_norm": 0.6840664119753669,
+      "learning_rate": 7.762542880281526e-07,
+      "loss": 0.2828,
+      "step": 10064
+    },
+    {
+      "epoch": 0.8257783976699348,
+      "grad_norm": 0.7495870532947759,
+      "learning_rate": 7.755433635250948e-07,
+      "loss": 0.2736,
+      "step": 10065
+    },
+    {
+      "epoch": 0.8258604422201256,
+      "grad_norm": 0.7602460278865553,
+      "learning_rate": 7.748327373463782e-07,
+      "loss": 0.3046,
+      "step": 10066
+    },
+    {
+      "epoch": 0.8259424867703162,
+      "grad_norm": 0.6808044905587926,
+      "learning_rate": 7.741224095421845e-07,
+      "loss": 0.2927,
+      "step": 10067
+    },
+    {
+      "epoch": 0.826024531320507,
+      "grad_norm": 0.7652004776886638,
+      "learning_rate": 7.734123801626781e-07,
+      "loss": 0.3394,
+      "step": 10068
+    },
+    {
+      "epoch": 0.8261065758706978,
+      "grad_norm": 0.6639111288147781,
+      "learning_rate": 7.727026492579976e-07,
+      "loss": 0.3235,
+      "step": 10069
+    },
+    {
+      "epoch": 0.8261886204208886,
+      "grad_norm": 0.7489950562831769,
+      "learning_rate": 7.719932168782656e-07,
+      "loss": 0.265,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8262706649710793,
+      "grad_norm": 0.7723629101589876,
+      "learning_rate": 7.712840830735785e-07,
+      "loss": 0.2962,
+      "step": 10071
+    },
+    {
+      "epoch": 0.82635270952127,
+      "grad_norm": 0.768541027722128,
+      "learning_rate": 7.705752478940154e-07,
+      "loss": 0.3172,
+      "step": 10072
+    },
+    {
+      "epoch": 0.8264347540714608,
+      "grad_norm": 0.7522553418699148,
+      "learning_rate": 7.698667113896346e-07,
+      "loss": 0.2957,
+      "step": 10073
+    },
+    {
+      "epoch": 0.8265167986216515,
+      "grad_norm": 0.7673432749302894,
+      "learning_rate": 7.69158473610469e-07,
+      "loss": 0.3384,
+      "step": 10074
+    },
+    {
+      "epoch": 0.8265988431718423,
+      "grad_norm": 0.8018378664664306,
+      "learning_rate": 7.684505346065363e-07,
+      "loss": 0.3171,
+      "step": 10075
+    },
+    {
+      "epoch": 0.8266808877220331,
+      "grad_norm": 0.7109265133681996,
+      "learning_rate": 7.677428944278271e-07,
+      "loss": 0.3173,
+      "step": 10076
+    },
+    {
+      "epoch": 0.8267629322722239,
+      "grad_norm": 0.6898189432998911,
+      "learning_rate": 7.670355531243145e-07,
+      "loss": 0.2705,
+      "step": 10077
+    },
+    {
+      "epoch": 0.8268449768224145,
+      "grad_norm": 0.7093954286130688,
+      "learning_rate": 7.663285107459517e-07,
+      "loss": 0.2925,
+      "step": 10078
+    },
+    {
+      "epoch": 0.8269270213726053,
+      "grad_norm": 0.7574675657732958,
+      "learning_rate": 7.65621767342668e-07,
+      "loss": 0.2763,
+      "step": 10079
+    },
+    {
+      "epoch": 0.8270090659227961,
+      "grad_norm": 0.7124400894932973,
+      "learning_rate": 7.649153229643708e-07,
+      "loss": 0.312,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8270911104729868,
+      "grad_norm": 1.0847845415445831,
+      "learning_rate": 7.6420917766095e-07,
+      "loss": 0.3325,
+      "step": 10081
+    },
+    {
+      "epoch": 0.8271731550231776,
+      "grad_norm": 0.7129825287262971,
+      "learning_rate": 7.63503331482271e-07,
+      "loss": 0.306,
+      "step": 10082
+    },
+    {
+      "epoch": 0.8272551995733683,
+      "grad_norm": 0.7040769495675181,
+      "learning_rate": 7.627977844781815e-07,
+      "loss": 0.2968,
+      "step": 10083
+    },
+    {
+      "epoch": 0.8273372441235591,
+      "grad_norm": 0.8182252910162848,
+      "learning_rate": 7.620925366985033e-07,
+      "loss": 0.3002,
+      "step": 10084
+    },
+    {
+      "epoch": 0.8274192886737498,
+      "grad_norm": 0.7862939754966901,
+      "learning_rate": 7.613875881930416e-07,
+      "loss": 0.271,
+      "step": 10085
+    },
+    {
+      "epoch": 0.8275013332239406,
+      "grad_norm": 0.7124290871620932,
+      "learning_rate": 7.606829390115799e-07,
+      "loss": 0.2987,
+      "step": 10086
+    },
+    {
+      "epoch": 0.8275833777741314,
+      "grad_norm": 0.9886649651976749,
+      "learning_rate": 7.599785892038764e-07,
+      "loss": 0.28,
+      "step": 10087
+    },
+    {
+      "epoch": 0.827665422324322,
+      "grad_norm": 0.6634874606142072,
+      "learning_rate": 7.592745388196748e-07,
+      "loss": 0.2761,
+      "step": 10088
+    },
+    {
+      "epoch": 0.8277474668745128,
+      "grad_norm": 0.6629123360003163,
+      "learning_rate": 7.585707879086901e-07,
+      "loss": 0.2782,
+      "step": 10089
+    },
+    {
+      "epoch": 0.8278295114247036,
+      "grad_norm": 0.6812177497979843,
+      "learning_rate": 7.578673365206224e-07,
+      "loss": 0.3056,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8279115559748944,
+      "grad_norm": 0.669282223082241,
+      "learning_rate": 7.571641847051492e-07,
+      "loss": 0.3095,
+      "step": 10091
+    },
+    {
+      "epoch": 0.8279936005250851,
+      "grad_norm": 0.6524228563022577,
+      "learning_rate": 7.564613325119241e-07,
+      "loss": 0.3004,
+      "step": 10092
+    },
+    {
+      "epoch": 0.8280756450752759,
+      "grad_norm": 0.7139426120716742,
+      "learning_rate": 7.557587799905813e-07,
+      "loss": 0.2753,
+      "step": 10093
+    },
+    {
+      "epoch": 0.8281576896254667,
+      "grad_norm": 0.6932378947427449,
+      "learning_rate": 7.550565271907357e-07,
+      "loss": 0.2803,
+      "step": 10094
+    },
+    {
+      "epoch": 0.8282397341756573,
+      "grad_norm": 0.7388890800971697,
+      "learning_rate": 7.543545741619762e-07,
+      "loss": 0.2995,
+      "step": 10095
+    },
+    {
+      "epoch": 0.8283217787258481,
+      "grad_norm": 0.6703445300258937,
+      "learning_rate": 7.536529209538773e-07,
+      "loss": 0.2605,
+      "step": 10096
+    },
+    {
+      "epoch": 0.8284038232760389,
+      "grad_norm": 0.5759045610828644,
+      "learning_rate": 7.52951567615986e-07,
+      "loss": 0.2862,
+      "step": 10097
+    },
+    {
+      "epoch": 0.8284858678262297,
+      "grad_norm": 0.7425209229759971,
+      "learning_rate": 7.522505141978309e-07,
+      "loss": 0.2747,
+      "step": 10098
+    },
+    {
+      "epoch": 0.8285679123764204,
+      "grad_norm": 0.7247908946256503,
+      "learning_rate": 7.515497607489213e-07,
+      "loss": 0.282,
+      "step": 10099
+    },
+    {
+      "epoch": 0.8286499569266111,
+      "grad_norm": 0.6360200951867712,
+      "learning_rate": 7.508493073187411e-07,
+      "loss": 0.3032,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8287320014768019,
+      "grad_norm": 0.7779863330978688,
+      "learning_rate": 7.501491539567574e-07,
+      "loss": 0.3148,
+      "step": 10101
+    },
+    {
+      "epoch": 0.8288140460269927,
+      "grad_norm": 0.7275033725327077,
+      "learning_rate": 7.494493007124109e-07,
+      "loss": 0.319,
+      "step": 10102
+    },
+    {
+      "epoch": 0.8288960905771834,
+      "grad_norm": 0.6948678017211177,
+      "learning_rate": 7.487497476351258e-07,
+      "loss": 0.2968,
+      "step": 10103
+    },
+    {
+      "epoch": 0.8289781351273742,
+      "grad_norm": 0.7281302609553751,
+      "learning_rate": 7.480504947743044e-07,
+      "loss": 0.3555,
+      "step": 10104
+    },
+    {
+      "epoch": 0.829060179677565,
+      "grad_norm": 0.6426411694290217,
+      "learning_rate": 7.473515421793248e-07,
+      "loss": 0.3123,
+      "step": 10105
+    },
+    {
+      "epoch": 0.8291422242277556,
+      "grad_norm": 0.6975479428697758,
+      "learning_rate": 7.466528898995479e-07,
+      "loss": 0.2862,
+      "step": 10106
+    },
+    {
+      "epoch": 0.8292242687779464,
+      "grad_norm": 0.7622877372846111,
+      "learning_rate": 7.459545379843108e-07,
+      "loss": 0.3074,
+      "step": 10107
+    },
+    {
+      "epoch": 0.8293063133281372,
+      "grad_norm": 0.6771001742578777,
+      "learning_rate": 7.452564864829281e-07,
+      "loss": 0.2927,
+      "step": 10108
+    },
+    {
+      "epoch": 0.829388357878328,
+      "grad_norm": 0.653229240738029,
+      "learning_rate": 7.445587354446975e-07,
+      "loss": 0.3006,
+      "step": 10109
+    },
+    {
+      "epoch": 0.8294704024285187,
+      "grad_norm": 0.7326996366078978,
+      "learning_rate": 7.438612849188915e-07,
+      "loss": 0.3452,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8295524469787094,
+      "grad_norm": 0.6628722309791439,
+      "learning_rate": 7.43164134954763e-07,
+      "loss": 0.3539,
+      "step": 10111
+    },
+    {
+      "epoch": 0.8296344915289002,
+      "grad_norm": 0.6728660195973715,
+      "learning_rate": 7.424672856015458e-07,
+      "loss": 0.2953,
+      "step": 10112
+    },
+    {
+      "epoch": 0.8297165360790909,
+      "grad_norm": 0.6936255560562267,
+      "learning_rate": 7.417707369084476e-07,
+      "loss": 0.3142,
+      "step": 10113
+    },
+    {
+      "epoch": 0.8297985806292817,
+      "grad_norm": 0.7684826138767574,
+      "learning_rate": 7.41074488924659e-07,
+      "loss": 0.3179,
+      "step": 10114
+    },
+    {
+      "epoch": 0.8298806251794725,
+      "grad_norm": 0.703692531396334,
+      "learning_rate": 7.40378541699347e-07,
+      "loss": 0.3237,
+      "step": 10115
+    },
+    {
+      "epoch": 0.8299626697296633,
+      "grad_norm": 0.6643505674678118,
+      "learning_rate": 7.396828952816587e-07,
+      "loss": 0.3019,
+      "step": 10116
+    },
+    {
+      "epoch": 0.8300447142798539,
+      "grad_norm": 0.6803382851713176,
+      "learning_rate": 7.389875497207205e-07,
+      "loss": 0.2678,
+      "step": 10117
+    },
+    {
+      "epoch": 0.8301267588300447,
+      "grad_norm": 0.7918378310505687,
+      "learning_rate": 7.382925050656348e-07,
+      "loss": 0.2899,
+      "step": 10118
+    },
+    {
+      "epoch": 0.8302088033802355,
+      "grad_norm": 0.7038283593298607,
+      "learning_rate": 7.375977613654861e-07,
+      "loss": 0.2977,
+      "step": 10119
+    },
+    {
+      "epoch": 0.8302908479304262,
+      "grad_norm": 0.7707435404365823,
+      "learning_rate": 7.369033186693359e-07,
+      "loss": 0.3079,
+      "step": 10120
+    },
+    {
+      "epoch": 0.830372892480617,
+      "grad_norm": 1.3381445347956011,
+      "learning_rate": 7.362091770262231e-07,
+      "loss": 0.3017,
+      "step": 10121
+    },
+    {
+      "epoch": 0.8304549370308077,
+      "grad_norm": 0.7178564864557997,
+      "learning_rate": 7.355153364851686e-07,
+      "loss": 0.2912,
+      "step": 10122
+    },
+    {
+      "epoch": 0.8305369815809985,
+      "grad_norm": 0.6662846206995062,
+      "learning_rate": 7.348217970951687e-07,
+      "loss": 0.2565,
+      "step": 10123
+    },
+    {
+      "epoch": 0.8306190261311892,
+      "grad_norm": 0.6524936958891756,
+      "learning_rate": 7.341285589052022e-07,
+      "loss": 0.3226,
+      "step": 10124
+    },
+    {
+      "epoch": 0.83070107068138,
+      "grad_norm": 0.7301771531681354,
+      "learning_rate": 7.334356219642219e-07,
+      "loss": 0.3199,
+      "step": 10125
+    },
+    {
+      "epoch": 0.8307831152315708,
+      "grad_norm": 0.7783946217917983,
+      "learning_rate": 7.327429863211633e-07,
+      "loss": 0.341,
+      "step": 10126
+    },
+    {
+      "epoch": 0.8308651597817615,
+      "grad_norm": 0.7004370143193185,
+      "learning_rate": 7.320506520249404e-07,
+      "loss": 0.2657,
+      "step": 10127
+    },
+    {
+      "epoch": 0.8309472043319522,
+      "grad_norm": 0.7011859681950081,
+      "learning_rate": 7.313586191244421e-07,
+      "loss": 0.2872,
+      "step": 10128
+    },
+    {
+      "epoch": 0.831029248882143,
+      "grad_norm": 0.6743235056052237,
+      "learning_rate": 7.306668876685402e-07,
+      "loss": 0.3042,
+      "step": 10129
+    },
+    {
+      "epoch": 0.8311112934323338,
+      "grad_norm": 0.6976056330847829,
+      "learning_rate": 7.299754577060847e-07,
+      "loss": 0.3172,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8311933379825245,
+      "grad_norm": 0.6358573930576686,
+      "learning_rate": 7.292843292859009e-07,
+      "loss": 0.2958,
+      "step": 10131
+    },
+    {
+      "epoch": 0.8312753825327153,
+      "grad_norm": 0.7716399434031991,
+      "learning_rate": 7.285935024567975e-07,
+      "loss": 0.2991,
+      "step": 10132
+    },
+    {
+      "epoch": 0.831357427082906,
+      "grad_norm": 1.1320617780928306,
+      "learning_rate": 7.279029772675572e-07,
+      "loss": 0.2716,
+      "step": 10133
+    },
+    {
+      "epoch": 0.8314394716330967,
+      "grad_norm": 0.5641867941224109,
+      "learning_rate": 7.27212753766946e-07,
+      "loss": 0.6362,
+      "step": 10134
+    },
+    {
+      "epoch": 0.8315215161832875,
+      "grad_norm": 0.7337317000117262,
+      "learning_rate": 7.265228320037054e-07,
+      "loss": 0.293,
+      "step": 10135
+    },
+    {
+      "epoch": 0.8316035607334783,
+      "grad_norm": 0.6698936042868271,
+      "learning_rate": 7.258332120265554e-07,
+      "loss": 0.2928,
+      "step": 10136
+    },
+    {
+      "epoch": 0.8316856052836691,
+      "grad_norm": 0.7031422690883417,
+      "learning_rate": 7.251438938841981e-07,
+      "loss": 0.3145,
+      "step": 10137
+    },
+    {
+      "epoch": 0.8317676498338598,
+      "grad_norm": 0.8151297897587783,
+      "learning_rate": 7.244548776253102e-07,
+      "loss": 0.3281,
+      "step": 10138
+    },
+    {
+      "epoch": 0.8318496943840505,
+      "grad_norm": 0.7276805542950563,
+      "learning_rate": 7.237661632985493e-07,
+      "loss": 0.3111,
+      "step": 10139
+    },
+    {
+      "epoch": 0.8319317389342413,
+      "grad_norm": 0.7388024152547777,
+      "learning_rate": 7.230777509525527e-07,
+      "loss": 0.2746,
+      "step": 10140
+    },
+    {
+      "epoch": 0.832013783484432,
+      "grad_norm": 0.7212528969898304,
+      "learning_rate": 7.223896406359326e-07,
+      "loss": 0.2893,
+      "step": 10141
+    },
+    {
+      "epoch": 0.8320958280346228,
+      "grad_norm": 0.9707466372236226,
+      "learning_rate": 7.217018323972852e-07,
+      "loss": 0.2512,
+      "step": 10142
+    },
+    {
+      "epoch": 0.8321778725848136,
+      "grad_norm": 0.8014674469865452,
+      "learning_rate": 7.210143262851793e-07,
+      "loss": 0.321,
+      "step": 10143
+    },
+    {
+      "epoch": 0.8322599171350044,
+      "grad_norm": 0.6816029383724682,
+      "learning_rate": 7.203271223481672e-07,
+      "loss": 0.3207,
+      "step": 10144
+    },
+    {
+      "epoch": 0.832341961685195,
+      "grad_norm": 0.6965551256628942,
+      "learning_rate": 7.196402206347792e-07,
+      "loss": 0.2921,
+      "step": 10145
+    },
+    {
+      "epoch": 0.8324240062353858,
+      "grad_norm": 0.7939449091237148,
+      "learning_rate": 7.189536211935205e-07,
+      "loss": 0.2561,
+      "step": 10146
+    },
+    {
+      "epoch": 0.8325060507855766,
+      "grad_norm": 0.6677011593365014,
+      "learning_rate": 7.182673240728804e-07,
+      "loss": 0.3272,
+      "step": 10147
+    },
+    {
+      "epoch": 0.8325880953357673,
+      "grad_norm": 0.6756560295688715,
+      "learning_rate": 7.175813293213224e-07,
+      "loss": 0.3304,
+      "step": 10148
+    },
+    {
+      "epoch": 0.8326701398859581,
+      "grad_norm": 0.6114765150952517,
+      "learning_rate": 7.168956369872898e-07,
+      "loss": 0.301,
+      "step": 10149
+    },
+    {
+      "epoch": 0.8327521844361488,
+      "grad_norm": 0.6974623863487545,
+      "learning_rate": 7.162102471192067e-07,
+      "loss": 0.2921,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8328342289863396,
+      "grad_norm": 0.8010168194010194,
+      "learning_rate": 7.155251597654727e-07,
+      "loss": 0.293,
+      "step": 10151
+    },
+    {
+      "epoch": 0.8329162735365303,
+      "grad_norm": 0.834032971046517,
+      "learning_rate": 7.148403749744687e-07,
+      "loss": 0.2765,
+      "step": 10152
+    },
+    {
+      "epoch": 0.8329983180867211,
+      "grad_norm": 0.6820917789203933,
+      "learning_rate": 7.141558927945536e-07,
+      "loss": 0.3135,
+      "step": 10153
+    },
+    {
+      "epoch": 0.8330803626369119,
+      "grad_norm": 0.7355230399211167,
+      "learning_rate": 7.134717132740626e-07,
+      "loss": 0.2845,
+      "step": 10154
+    },
+    {
+      "epoch": 0.8331624071871025,
+      "grad_norm": 0.6493570167816212,
+      "learning_rate": 7.127878364613133e-07,
+      "loss": 0.2536,
+      "step": 10155
+    },
+    {
+      "epoch": 0.8332444517372933,
+      "grad_norm": 0.7500649343570959,
+      "learning_rate": 7.121042624045981e-07,
+      "loss": 0.2658,
+      "step": 10156
+    },
+    {
+      "epoch": 0.8333264962874841,
+      "grad_norm": 0.7040977962139067,
+      "learning_rate": 7.114209911521907e-07,
+      "loss": 0.2956,
+      "step": 10157
+    },
+    {
+      "epoch": 0.8334085408376749,
+      "grad_norm": 0.7294570626054767,
+      "learning_rate": 7.107380227523442e-07,
+      "loss": 0.2752,
+      "step": 10158
+    },
+    {
+      "epoch": 0.8334905853878656,
+      "grad_norm": 0.8923738164751359,
+      "learning_rate": 7.100553572532859e-07,
+      "loss": 0.281,
+      "step": 10159
+    },
+    {
+      "epoch": 0.8335726299380564,
+      "grad_norm": 0.6646100208774294,
+      "learning_rate": 7.093729947032274e-07,
+      "loss": 0.2748,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8336546744882471,
+      "grad_norm": 0.829037310564732,
+      "learning_rate": 7.086909351503529e-07,
+      "loss": 0.2878,
+      "step": 10161
+    },
+    {
+      "epoch": 0.8337367190384378,
+      "grad_norm": 1.0992092868213428,
+      "learning_rate": 7.080091786428317e-07,
+      "loss": 0.2835,
+      "step": 10162
+    },
+    {
+      "epoch": 0.8338187635886286,
+      "grad_norm": 0.6915997049603271,
+      "learning_rate": 7.073277252288063e-07,
+      "loss": 0.3411,
+      "step": 10163
+    },
+    {
+      "epoch": 0.8339008081388194,
+      "grad_norm": 0.7210457337712466,
+      "learning_rate": 7.066465749563994e-07,
+      "loss": 0.3374,
+      "step": 10164
+    },
+    {
+      "epoch": 0.8339828526890102,
+      "grad_norm": 0.7003786014633521,
+      "learning_rate": 7.059657278737136e-07,
+      "loss": 0.2668,
+      "step": 10165
+    },
+    {
+      "epoch": 0.8340648972392009,
+      "grad_norm": 0.6499426191429359,
+      "learning_rate": 7.052851840288299e-07,
+      "loss": 0.2682,
+      "step": 10166
+    },
+    {
+      "epoch": 0.8341469417893916,
+      "grad_norm": 0.6751932423601864,
+      "learning_rate": 7.04604943469806e-07,
+      "loss": 0.309,
+      "step": 10167
+    },
+    {
+      "epoch": 0.8342289863395824,
+      "grad_norm": 0.6869485470497234,
+      "learning_rate": 7.039250062446806e-07,
+      "loss": 0.3118,
+      "step": 10168
+    },
+    {
+      "epoch": 0.8343110308897731,
+      "grad_norm": 0.7337821041991711,
+      "learning_rate": 7.032453724014681e-07,
+      "loss": 0.286,
+      "step": 10169
+    },
+    {
+      "epoch": 0.8343930754399639,
+      "grad_norm": 0.7107508338183295,
+      "learning_rate": 7.025660419881641e-07,
+      "loss": 0.2962,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8344751199901547,
+      "grad_norm": 0.6776453510353874,
+      "learning_rate": 7.01887015052743e-07,
+      "loss": 0.3197,
+      "step": 10171
+    },
+    {
+      "epoch": 0.8345571645403455,
+      "grad_norm": 0.7209207195731273,
+      "learning_rate": 7.012082916431545e-07,
+      "loss": 0.3357,
+      "step": 10172
+    },
+    {
+      "epoch": 0.8346392090905361,
+      "grad_norm": 0.740024882074249,
+      "learning_rate": 7.005298718073311e-07,
+      "loss": 0.2901,
+      "step": 10173
+    },
+    {
+      "epoch": 0.8347212536407269,
+      "grad_norm": 0.6511401777789159,
+      "learning_rate": 6.998517555931788e-07,
+      "loss": 0.3031,
+      "step": 10174
+    },
+    {
+      "epoch": 0.8348032981909177,
+      "grad_norm": 0.7572224158075594,
+      "learning_rate": 6.991739430485883e-07,
+      "loss": 0.2739,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8348853427411084,
+      "grad_norm": 0.7519064400896487,
+      "learning_rate": 6.984964342214245e-07,
+      "loss": 0.294,
+      "step": 10176
+    },
+    {
+      "epoch": 0.8349673872912992,
+      "grad_norm": 0.7613563822467044,
+      "learning_rate": 6.978192291595304e-07,
+      "loss": 0.3085,
+      "step": 10177
+    },
+    {
+      "epoch": 0.8350494318414899,
+      "grad_norm": 0.6597066740857469,
+      "learning_rate": 6.971423279107309e-07,
+      "loss": 0.3209,
+      "step": 10178
+    },
+    {
+      "epoch": 0.8351314763916807,
+      "grad_norm": 0.6438441697989762,
+      "learning_rate": 6.964657305228262e-07,
+      "loss": 0.2921,
+      "step": 10179
+    },
+    {
+      "epoch": 0.8352135209418714,
+      "grad_norm": 0.6830474227918166,
+      "learning_rate": 6.95789437043598e-07,
+      "loss": 0.302,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8352955654920622,
+      "grad_norm": 0.7511126229780828,
+      "learning_rate": 6.951134475208049e-07,
+      "loss": 0.3461,
+      "step": 10181
+    },
+    {
+      "epoch": 0.835377610042253,
+      "grad_norm": 0.737813412751922,
+      "learning_rate": 6.944377620021831e-07,
+      "loss": 0.3097,
+      "step": 10182
+    },
+    {
+      "epoch": 0.8354596545924436,
+      "grad_norm": 0.6898218545085553,
+      "learning_rate": 6.937623805354493e-07,
+      "loss": 0.258,
+      "step": 10183
+    },
+    {
+      "epoch": 0.8355416991426344,
+      "grad_norm": 0.7236619561899384,
+      "learning_rate": 6.930873031682983e-07,
+      "loss": 0.3484,
+      "step": 10184
+    },
+    {
+      "epoch": 0.8356237436928252,
+      "grad_norm": 0.7648567579991399,
+      "learning_rate": 6.924125299484014e-07,
+      "loss": 0.2968,
+      "step": 10185
+    },
+    {
+      "epoch": 0.835705788243016,
+      "grad_norm": 0.690540892504803,
+      "learning_rate": 6.917380609234125e-07,
+      "loss": 0.347,
+      "step": 10186
+    },
+    {
+      "epoch": 0.8357878327932067,
+      "grad_norm": 0.6509096370960392,
+      "learning_rate": 6.910638961409583e-07,
+      "loss": 0.3005,
+      "step": 10187
+    },
+    {
+      "epoch": 0.8358698773433975,
+      "grad_norm": 0.7037512192230222,
+      "learning_rate": 6.903900356486504e-07,
+      "loss": 0.2729,
+      "step": 10188
+    },
+    {
+      "epoch": 0.8359519218935882,
+      "grad_norm": 0.8647689936651005,
+      "learning_rate": 6.89716479494073e-07,
+      "loss": 0.3303,
+      "step": 10189
+    },
+    {
+      "epoch": 0.8360339664437789,
+      "grad_norm": 0.7290872424603703,
+      "learning_rate": 6.890432277247943e-07,
+      "loss": 0.2835,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8361160109939697,
+      "grad_norm": 0.774309148978802,
+      "learning_rate": 6.883702803883563e-07,
+      "loss": 0.3077,
+      "step": 10191
+    },
+    {
+      "epoch": 0.8361980555441605,
+      "grad_norm": 1.8388038896457768,
+      "learning_rate": 6.876976375322808e-07,
+      "loss": 0.3095,
+      "step": 10192
+    },
+    {
+      "epoch": 0.8362801000943513,
+      "grad_norm": 0.6676410756918433,
+      "learning_rate": 6.870252992040705e-07,
+      "loss": 0.2953,
+      "step": 10193
+    },
+    {
+      "epoch": 0.836362144644542,
+      "grad_norm": 0.7858091253424574,
+      "learning_rate": 6.86353265451205e-07,
+      "loss": 0.2985,
+      "step": 10194
+    },
+    {
+      "epoch": 0.8364441891947327,
+      "grad_norm": 0.6986036351078516,
+      "learning_rate": 6.856815363211399e-07,
+      "loss": 0.2967,
+      "step": 10195
+    },
+    {
+      "epoch": 0.8365262337449235,
+      "grad_norm": 0.6350717831710674,
+      "learning_rate": 6.85010111861315e-07,
+      "loss": 0.2983,
+      "step": 10196
+    },
+    {
+      "epoch": 0.8366082782951142,
+      "grad_norm": 0.6593579941528923,
+      "learning_rate": 6.843389921191423e-07,
+      "loss": 0.2706,
+      "step": 10197
+    },
+    {
+      "epoch": 0.836690322845305,
+      "grad_norm": 0.7062049924558031,
+      "learning_rate": 6.836681771420162e-07,
+      "loss": 0.2811,
+      "step": 10198
+    },
+    {
+      "epoch": 0.8367723673954958,
+      "grad_norm": 0.7642103969668868,
+      "learning_rate": 6.829976669773098e-07,
+      "loss": 0.2923,
+      "step": 10199
+    },
+    {
+      "epoch": 0.8368544119456865,
+      "grad_norm": 0.6955342630759509,
+      "learning_rate": 6.823274616723707e-07,
+      "loss": 0.2972,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8369364564958772,
+      "grad_norm": 0.9057425193114538,
+      "learning_rate": 6.816575612745302e-07,
+      "loss": 0.2995,
+      "step": 10201
+    },
+    {
+      "epoch": 0.837018501046068,
+      "grad_norm": 0.7061158963511436,
+      "learning_rate": 6.809879658310953e-07,
+      "loss": 0.3408,
+      "step": 10202
+    },
+    {
+      "epoch": 0.8371005455962588,
+      "grad_norm": 0.6479562047233145,
+      "learning_rate": 6.803186753893515e-07,
+      "loss": 0.315,
+      "step": 10203
+    },
+    {
+      "epoch": 0.8371825901464495,
+      "grad_norm": 0.7740617291458284,
+      "learning_rate": 6.796496899965627e-07,
+      "loss": 0.3141,
+      "step": 10204
+    },
+    {
+      "epoch": 0.8372646346966403,
+      "grad_norm": 0.7641798931849958,
+      "learning_rate": 6.78981009699971e-07,
+      "loss": 0.3159,
+      "step": 10205
+    },
+    {
+      "epoch": 0.837346679246831,
+      "grad_norm": 0.9493460362929982,
+      "learning_rate": 6.783126345467978e-07,
+      "loss": 0.3465,
+      "step": 10206
+    },
+    {
+      "epoch": 0.8374287237970218,
+      "grad_norm": 0.6876856084368813,
+      "learning_rate": 6.776445645842439e-07,
+      "loss": 0.3212,
+      "step": 10207
+    },
+    {
+      "epoch": 0.8375107683472125,
+      "grad_norm": 0.734811536242437,
+      "learning_rate": 6.769767998594857e-07,
+      "loss": 0.3074,
+      "step": 10208
+    },
+    {
+      "epoch": 0.8375928128974033,
+      "grad_norm": 0.7194913839187138,
+      "learning_rate": 6.763093404196808e-07,
+      "loss": 0.3095,
+      "step": 10209
+    },
+    {
+      "epoch": 0.8376748574475941,
+      "grad_norm": 0.7223729239127515,
+      "learning_rate": 6.756421863119634e-07,
+      "loss": 0.3914,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8377569019977847,
+      "grad_norm": 0.7826423586015759,
+      "learning_rate": 6.749753375834467e-07,
+      "loss": 0.2938,
+      "step": 10211
+    },
+    {
+      "epoch": 0.8378389465479755,
+      "grad_norm": 0.893940623294124,
+      "learning_rate": 6.743087942812243e-07,
+      "loss": 0.2913,
+      "step": 10212
+    },
+    {
+      "epoch": 0.8379209910981663,
+      "grad_norm": 0.7056750580241833,
+      "learning_rate": 6.736425564523641e-07,
+      "loss": 0.3151,
+      "step": 10213
+    },
+    {
+      "epoch": 0.8380030356483571,
+      "grad_norm": 0.6953198958850048,
+      "learning_rate": 6.729766241439167e-07,
+      "loss": 0.2715,
+      "step": 10214
+    },
+    {
+      "epoch": 0.8380850801985478,
+      "grad_norm": 0.6675656748795157,
+      "learning_rate": 6.723109974029074e-07,
+      "loss": 0.3057,
+      "step": 10215
+    },
+    {
+      "epoch": 0.8381671247487386,
+      "grad_norm": 0.8431128488633739,
+      "learning_rate": 6.716456762763435e-07,
+      "loss": 0.2714,
+      "step": 10216
+    },
+    {
+      "epoch": 0.8382491692989293,
+      "grad_norm": 0.6893376576716229,
+      "learning_rate": 6.709806608112068e-07,
+      "loss": 0.2916,
+      "step": 10217
+    },
+    {
+      "epoch": 0.8383312138491201,
+      "grad_norm": 0.6924132231097415,
+      "learning_rate": 6.703159510544616e-07,
+      "loss": 0.2742,
+      "step": 10218
+    },
+    {
+      "epoch": 0.8384132583993108,
+      "grad_norm": 0.7231155240965993,
+      "learning_rate": 6.696515470530468e-07,
+      "loss": 0.3335,
+      "step": 10219
+    },
+    {
+      "epoch": 0.8384953029495016,
+      "grad_norm": 0.7395204117987156,
+      "learning_rate": 6.689874488538833e-07,
+      "loss": 0.3031,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8385773474996924,
+      "grad_norm": 0.727386393875644,
+      "learning_rate": 6.683236565038676e-07,
+      "loss": 0.2939,
+      "step": 10221
+    },
+    {
+      "epoch": 0.838659392049883,
+      "grad_norm": 0.6554513890567404,
+      "learning_rate": 6.676601700498764e-07,
+      "loss": 0.3085,
+      "step": 10222
+    },
+    {
+      "epoch": 0.8387414366000738,
+      "grad_norm": 0.7210751120333388,
+      "learning_rate": 6.669969895387623e-07,
+      "loss": 0.2651,
+      "step": 10223
+    },
+    {
+      "epoch": 0.8388234811502646,
+      "grad_norm": 0.7324475833451857,
+      "learning_rate": 6.663341150173597e-07,
+      "loss": 0.3309,
+      "step": 10224
+    },
+    {
+      "epoch": 0.8389055257004554,
+      "grad_norm": 0.6648655411337359,
+      "learning_rate": 6.656715465324803e-07,
+      "loss": 0.3043,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8389875702506461,
+      "grad_norm": 0.7687340552298876,
+      "learning_rate": 6.650092841309114e-07,
+      "loss": 0.3246,
+      "step": 10226
+    },
+    {
+      "epoch": 0.8390696148008369,
+      "grad_norm": 0.6655645707128116,
+      "learning_rate": 6.643473278594231e-07,
+      "loss": 0.3062,
+      "step": 10227
+    },
+    {
+      "epoch": 0.8391516593510276,
+      "grad_norm": 0.6491278791839943,
+      "learning_rate": 6.636856777647599e-07,
+      "loss": 0.2922,
+      "step": 10228
+    },
+    {
+      "epoch": 0.8392337039012183,
+      "grad_norm": 0.729117388229119,
+      "learning_rate": 6.630243338936476e-07,
+      "loss": 0.2797,
+      "step": 10229
+    },
+    {
+      "epoch": 0.8393157484514091,
+      "grad_norm": 0.6996409689435872,
+      "learning_rate": 6.623632962927895e-07,
+      "loss": 0.2939,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8393977930015999,
+      "grad_norm": 0.7057184095854309,
+      "learning_rate": 6.617025650088671e-07,
+      "loss": 0.3009,
+      "step": 10231
+    },
+    {
+      "epoch": 0.8394798375517907,
+      "grad_norm": 0.6979267139394612,
+      "learning_rate": 6.610421400885392e-07,
+      "loss": 0.2634,
+      "step": 10232
+    },
+    {
+      "epoch": 0.8395618821019813,
+      "grad_norm": 0.7838761242298083,
+      "learning_rate": 6.603820215784429e-07,
+      "loss": 0.3229,
+      "step": 10233
+    },
+    {
+      "epoch": 0.8396439266521721,
+      "grad_norm": 0.665905678110749,
+      "learning_rate": 6.597222095251965e-07,
+      "loss": 0.2931,
+      "step": 10234
+    },
+    {
+      "epoch": 0.8397259712023629,
+      "grad_norm": 0.7848516458245218,
+      "learning_rate": 6.590627039753955e-07,
+      "loss": 0.2809,
+      "step": 10235
+    },
+    {
+      "epoch": 0.8398080157525536,
+      "grad_norm": 0.70395026637664,
+      "learning_rate": 6.58403504975611e-07,
+      "loss": 0.3371,
+      "step": 10236
+    },
+    {
+      "epoch": 0.8398900603027444,
+      "grad_norm": 0.7122698149128062,
+      "learning_rate": 6.577446125723958e-07,
+      "loss": 0.282,
+      "step": 10237
+    },
+    {
+      "epoch": 0.8399721048529352,
+      "grad_norm": 0.7738840564919226,
+      "learning_rate": 6.570860268122808e-07,
+      "loss": 0.3083,
+      "step": 10238
+    },
+    {
+      "epoch": 0.840054149403126,
+      "grad_norm": 0.7162122719846237,
+      "learning_rate": 6.564277477417719e-07,
+      "loss": 0.2652,
+      "step": 10239
+    },
+    {
+      "epoch": 0.8401361939533166,
+      "grad_norm": 1.2170997431810169,
+      "learning_rate": 6.557697754073577e-07,
+      "loss": 0.298,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8402182385035074,
+      "grad_norm": 0.6879762572292433,
+      "learning_rate": 6.551121098555019e-07,
+      "loss": 0.2817,
+      "step": 10241
+    },
+    {
+      "epoch": 0.8403002830536982,
+      "grad_norm": 0.6111233599219296,
+      "learning_rate": 6.544547511326482e-07,
+      "loss": 0.2827,
+      "step": 10242
+    },
+    {
+      "epoch": 0.8403823276038889,
+      "grad_norm": 0.7159289250032476,
+      "learning_rate": 6.537976992852196e-07,
+      "loss": 0.3527,
+      "step": 10243
+    },
+    {
+      "epoch": 0.8404643721540797,
+      "grad_norm": 0.7099313418466957,
+      "learning_rate": 6.531409543596146e-07,
+      "loss": 0.3003,
+      "step": 10244
+    },
+    {
+      "epoch": 0.8405464167042704,
+      "grad_norm": 0.8923925295168889,
+      "learning_rate": 6.524845164022104e-07,
+      "loss": 0.2825,
+      "step": 10245
+    },
+    {
+      "epoch": 0.8406284612544612,
+      "grad_norm": 0.6949366196049648,
+      "learning_rate": 6.51828385459366e-07,
+      "loss": 0.3114,
+      "step": 10246
+    },
+    {
+      "epoch": 0.8407105058046519,
+      "grad_norm": 0.662161572951465,
+      "learning_rate": 6.511725615774139e-07,
+      "loss": 0.3042,
+      "step": 10247
+    },
+    {
+      "epoch": 0.8407925503548427,
+      "grad_norm": 0.7045084844036463,
+      "learning_rate": 6.505170448026699e-07,
+      "loss": 0.2698,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8408745949050335,
+      "grad_norm": 0.890645398862087,
+      "learning_rate": 6.498618351814229e-07,
+      "loss": 0.3218,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8409566394552241,
+      "grad_norm": 0.6394767603357617,
+      "learning_rate": 6.492069327599454e-07,
+      "loss": 0.2785,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8410386840054149,
+      "grad_norm": 0.6532410427622589,
+      "learning_rate": 6.485523375844826e-07,
+      "loss": 0.3092,
+      "step": 10251
+    },
+    {
+      "epoch": 0.8411207285556057,
+      "grad_norm": 0.7020893510834382,
+      "learning_rate": 6.478980497012632e-07,
+      "loss": 0.3298,
+      "step": 10252
+    },
+    {
+      "epoch": 0.8412027731057965,
+      "grad_norm": 0.6828575942090274,
+      "learning_rate": 6.472440691564924e-07,
+      "loss": 0.3063,
+      "step": 10253
+    },
+    {
+      "epoch": 0.8412848176559872,
+      "grad_norm": 0.6446680350918014,
+      "learning_rate": 6.46590395996351e-07,
+      "loss": 0.3167,
+      "step": 10254
+    },
+    {
+      "epoch": 0.841366862206178,
+      "grad_norm": 0.688361935611369,
+      "learning_rate": 6.459370302670015e-07,
+      "loss": 0.3073,
+      "step": 10255
+    },
+    {
+      "epoch": 0.8414489067563687,
+      "grad_norm": 0.6498375464490341,
+      "learning_rate": 6.452839720145848e-07,
+      "loss": 0.2817,
+      "step": 10256
+    },
+    {
+      "epoch": 0.8415309513065594,
+      "grad_norm": 0.7979966927003347,
+      "learning_rate": 6.446312212852162e-07,
+      "loss": 0.2654,
+      "step": 10257
+    },
+    {
+      "epoch": 0.8416129958567502,
+      "grad_norm": 0.6837185797665534,
+      "learning_rate": 6.439787781249945e-07,
+      "loss": 0.3026,
+      "step": 10258
+    },
+    {
+      "epoch": 0.841695040406941,
+      "grad_norm": 0.7426755400218565,
+      "learning_rate": 6.433266425799933e-07,
+      "loss": 0.3094,
+      "step": 10259
+    },
+    {
+      "epoch": 0.8417770849571318,
+      "grad_norm": 0.6445027987026241,
+      "learning_rate": 6.426748146962635e-07,
+      "loss": 0.2921,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8418591295073224,
+      "grad_norm": 0.6501413723204233,
+      "learning_rate": 6.420232945198395e-07,
+      "loss": 0.3285,
+      "step": 10261
+    },
+    {
+      "epoch": 0.8419411740575132,
+      "grad_norm": 0.8758095594013837,
+      "learning_rate": 6.413720820967273e-07,
+      "loss": 0.3055,
+      "step": 10262
+    },
+    {
+      "epoch": 0.842023218607704,
+      "grad_norm": 0.6965286097996044,
+      "learning_rate": 6.407211774729171e-07,
+      "loss": 0.2843,
+      "step": 10263
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.733394441636145,
+      "learning_rate": 6.400705806943724e-07,
+      "loss": 0.2721,
+      "step": 10264
+    },
+    {
+      "epoch": 0.8421873077080855,
+      "grad_norm": 0.7608593737810193,
+      "learning_rate": 6.394202918070391e-07,
+      "loss": 0.3014,
+      "step": 10265
+    },
+    {
+      "epoch": 0.8422693522582763,
+      "grad_norm": 1.022102653375809,
+      "learning_rate": 6.387703108568394e-07,
+      "loss": 0.3153,
+      "step": 10266
+    },
+    {
+      "epoch": 0.842351396808467,
+      "grad_norm": 0.6790884344565646,
+      "learning_rate": 6.381206378896726e-07,
+      "loss": 0.3076,
+      "step": 10267
+    },
+    {
+      "epoch": 0.8424334413586577,
+      "grad_norm": 0.6890011406566402,
+      "learning_rate": 6.374712729514198e-07,
+      "loss": 0.2759,
+      "step": 10268
+    },
+    {
+      "epoch": 0.8425154859088485,
+      "grad_norm": 0.7077517902088416,
+      "learning_rate": 6.368222160879356e-07,
+      "loss": 0.2974,
+      "step": 10269
+    },
+    {
+      "epoch": 0.8425975304590393,
+      "grad_norm": 0.6986708272733356,
+      "learning_rate": 6.361734673450565e-07,
+      "loss": 0.3037,
+      "step": 10270
+    },
+    {
+      "epoch": 0.84267957500923,
+      "grad_norm": 0.8075717981915276,
+      "learning_rate": 6.35525026768597e-07,
+      "loss": 0.3123,
+      "step": 10271
+    },
+    {
+      "epoch": 0.8427616195594207,
+      "grad_norm": 0.7442355903139858,
+      "learning_rate": 6.348768944043482e-07,
+      "loss": 0.2987,
+      "step": 10272
+    },
+    {
+      "epoch": 0.8428436641096115,
+      "grad_norm": 0.6939547713249683,
+      "learning_rate": 6.342290702980785e-07,
+      "loss": 0.2798,
+      "step": 10273
+    },
+    {
+      "epoch": 0.8429257086598023,
+      "grad_norm": 0.6816121374682521,
+      "learning_rate": 6.335815544955392e-07,
+      "loss": 0.3072,
+      "step": 10274
+    },
+    {
+      "epoch": 0.843007753209993,
+      "grad_norm": 0.7406183269751925,
+      "learning_rate": 6.329343470424542e-07,
+      "loss": 0.2844,
+      "step": 10275
+    },
+    {
+      "epoch": 0.8430897977601838,
+      "grad_norm": 0.6923676368617356,
+      "learning_rate": 6.322874479845309e-07,
+      "loss": 0.3091,
+      "step": 10276
+    },
+    {
+      "epoch": 0.8431718423103746,
+      "grad_norm": 0.9901397099828627,
+      "learning_rate": 6.316408573674492e-07,
+      "loss": 0.3176,
+      "step": 10277
+    },
+    {
+      "epoch": 0.8432538868605652,
+      "grad_norm": 0.6914159267111583,
+      "learning_rate": 6.309945752368718e-07,
+      "loss": 0.2983,
+      "step": 10278
+    },
+    {
+      "epoch": 0.843335931410756,
+      "grad_norm": 0.7349655859005626,
+      "learning_rate": 6.303486016384392e-07,
+      "loss": 0.3283,
+      "step": 10279
+    },
+    {
+      "epoch": 0.8434179759609468,
+      "grad_norm": 0.7534559330448495,
+      "learning_rate": 6.29702936617767e-07,
+      "loss": 0.3117,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8435000205111376,
+      "grad_norm": 0.6551607270601425,
+      "learning_rate": 6.290575802204535e-07,
+      "loss": 0.3211,
+      "step": 10281
+    },
+    {
+      "epoch": 0.8435820650613283,
+      "grad_norm": 0.6970938798989922,
+      "learning_rate": 6.284125324920698e-07,
+      "loss": 0.3198,
+      "step": 10282
+    },
+    {
+      "epoch": 0.843664109611519,
+      "grad_norm": 0.7216633505705045,
+      "learning_rate": 6.277677934781695e-07,
+      "loss": 0.3265,
+      "step": 10283
+    },
+    {
+      "epoch": 0.8437461541617098,
+      "grad_norm": 1.1987060374566454,
+      "learning_rate": 6.27123363224284e-07,
+      "loss": 0.2849,
+      "step": 10284
+    },
+    {
+      "epoch": 0.8438281987119005,
+      "grad_norm": 0.7602938170790171,
+      "learning_rate": 6.264792417759202e-07,
+      "loss": 0.3265,
+      "step": 10285
+    },
+    {
+      "epoch": 0.8439102432620913,
+      "grad_norm": 0.685773503286683,
+      "learning_rate": 6.258354291785668e-07,
+      "loss": 0.2614,
+      "step": 10286
+    },
+    {
+      "epoch": 0.8439922878122821,
+      "grad_norm": 0.6988137470106264,
+      "learning_rate": 6.251919254776878e-07,
+      "loss": 0.2756,
+      "step": 10287
+    },
+    {
+      "epoch": 0.8440743323624729,
+      "grad_norm": 0.7133321491927573,
+      "learning_rate": 6.245487307187253e-07,
+      "loss": 0.3025,
+      "step": 10288
+    },
+    {
+      "epoch": 0.8441563769126635,
+      "grad_norm": 0.743906865402098,
+      "learning_rate": 6.239058449471025e-07,
+      "loss": 0.3323,
+      "step": 10289
+    },
+    {
+      "epoch": 0.8442384214628543,
+      "grad_norm": 0.7435357778197605,
+      "learning_rate": 6.232632682082174e-07,
+      "loss": 0.3001,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8443204660130451,
+      "grad_norm": 0.6647423165651076,
+      "learning_rate": 6.226210005474486e-07,
+      "loss": 0.293,
+      "step": 10291
+    },
+    {
+      "epoch": 0.8444025105632358,
+      "grad_norm": 0.7287679971082337,
+      "learning_rate": 6.219790420101529e-07,
+      "loss": 0.3116,
+      "step": 10292
+    },
+    {
+      "epoch": 0.8444845551134266,
+      "grad_norm": 0.9025145684348604,
+      "learning_rate": 6.213373926416627e-07,
+      "loss": 0.2803,
+      "step": 10293
+    },
+    {
+      "epoch": 0.8445665996636174,
+      "grad_norm": 0.8546085431023722,
+      "learning_rate": 6.206960524872913e-07,
+      "loss": 0.2755,
+      "step": 10294
+    },
+    {
+      "epoch": 0.8446486442138081,
+      "grad_norm": 0.8039055169055911,
+      "learning_rate": 6.200550215923284e-07,
+      "loss": 0.3034,
+      "step": 10295
+    },
+    {
+      "epoch": 0.8447306887639988,
+      "grad_norm": 0.8359804913013861,
+      "learning_rate": 6.194143000020425e-07,
+      "loss": 0.3032,
+      "step": 10296
+    },
+    {
+      "epoch": 0.8448127333141896,
+      "grad_norm": 0.6258373534076496,
+      "learning_rate": 6.187738877616822e-07,
+      "loss": 0.2814,
+      "step": 10297
+    },
+    {
+      "epoch": 0.8448947778643804,
+      "grad_norm": 0.6695353604951522,
+      "learning_rate": 6.181337849164699e-07,
+      "loss": 0.2908,
+      "step": 10298
+    },
+    {
+      "epoch": 0.8449768224145711,
+      "grad_norm": 0.6907029047057565,
+      "learning_rate": 6.174939915116107e-07,
+      "loss": 0.3345,
+      "step": 10299
+    },
+    {
+      "epoch": 0.8450588669647618,
+      "grad_norm": 0.7291701633824489,
+      "learning_rate": 6.168545075922844e-07,
+      "loss": 0.3124,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8451409115149526,
+      "grad_norm": 0.6742337020395687,
+      "learning_rate": 6.162153332036503e-07,
+      "loss": 0.2846,
+      "step": 10301
+    },
+    {
+      "epoch": 0.8452229560651434,
+      "grad_norm": 0.6936166187537371,
+      "learning_rate": 6.155764683908466e-07,
+      "loss": 0.3045,
+      "step": 10302
+    },
+    {
+      "epoch": 0.8453050006153341,
+      "grad_norm": 0.7617879341955055,
+      "learning_rate": 6.14937913198988e-07,
+      "loss": 0.3142,
+      "step": 10303
+    },
+    {
+      "epoch": 0.8453870451655249,
+      "grad_norm": 0.7246605274396433,
+      "learning_rate": 6.142996676731688e-07,
+      "loss": 0.2975,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8454690897157157,
+      "grad_norm": 0.7766419081135274,
+      "learning_rate": 6.13661731858462e-07,
+      "loss": 0.3127,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8455511342659063,
+      "grad_norm": 0.78784351010387,
+      "learning_rate": 6.130241057999153e-07,
+      "loss": 0.3382,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8456331788160971,
+      "grad_norm": 0.844308813639729,
+      "learning_rate": 6.123867895425589e-07,
+      "loss": 0.3133,
+      "step": 10307
+    },
+    {
+      "epoch": 0.8457152233662879,
+      "grad_norm": 0.7863849601896831,
+      "learning_rate": 6.117497831313973e-07,
+      "loss": 0.2955,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8457972679164787,
+      "grad_norm": 0.6551853648601094,
+      "learning_rate": 6.111130866114162e-07,
+      "loss": 0.293,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8458793124666694,
+      "grad_norm": 0.7142099757333334,
+      "learning_rate": 6.10476700027578e-07,
+      "loss": 0.343,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8459613570168601,
+      "grad_norm": 0.6994995556827385,
+      "learning_rate": 6.098406234248222e-07,
+      "loss": 0.261,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8460434015670509,
+      "grad_norm": 0.7475540525736598,
+      "learning_rate": 6.092048568480697e-07,
+      "loss": 0.2922,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8461254461172416,
+      "grad_norm": 0.6934878152567915,
+      "learning_rate": 6.085694003422144e-07,
+      "loss": 0.2885,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8462074906674324,
+      "grad_norm": 0.6816507521102593,
+      "learning_rate": 6.079342539521338e-07,
+      "loss": 0.2935,
+      "step": 10314
+    },
+    {
+      "epoch": 0.8462895352176232,
+      "grad_norm": 0.7049120298113404,
+      "learning_rate": 6.072994177226793e-07,
+      "loss": 0.2822,
+      "step": 10315
+    },
+    {
+      "epoch": 0.846371579767814,
+      "grad_norm": 0.8020367411369395,
+      "learning_rate": 6.066648916986823e-07,
+      "loss": 0.3207,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8464536243180046,
+      "grad_norm": 0.7433878846565632,
+      "learning_rate": 6.060306759249529e-07,
+      "loss": 0.2915,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8465356688681954,
+      "grad_norm": 0.7149124049257364,
+      "learning_rate": 6.053967704462766e-07,
+      "loss": 0.2819,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8466177134183862,
+      "grad_norm": 0.9652129689450527,
+      "learning_rate": 6.047631753074195e-07,
+      "loss": 0.2758,
+      "step": 10319
+    },
+    {
+      "epoch": 0.8466997579685769,
+      "grad_norm": 0.7035546117939983,
+      "learning_rate": 6.041298905531273e-07,
+      "loss": 0.327,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8467818025187677,
+      "grad_norm": 0.7037272814514137,
+      "learning_rate": 6.034969162281184e-07,
+      "loss": 0.3167,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8468638470689585,
+      "grad_norm": 0.6551487556160562,
+      "learning_rate": 6.028642523770934e-07,
+      "loss": 0.3054,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8469458916191492,
+      "grad_norm": 0.6663918732611702,
+      "learning_rate": 6.022318990447318e-07,
+      "loss": 0.2795,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8470279361693399,
+      "grad_norm": 0.6816384223488243,
+      "learning_rate": 6.01599856275687e-07,
+      "loss": 0.294,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8471099807195307,
+      "grad_norm": 0.6912108846437859,
+      "learning_rate": 6.009681241145943e-07,
+      "loss": 0.3302,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8471920252697215,
+      "grad_norm": 0.778602054609131,
+      "learning_rate": 6.003367026060647e-07,
+      "loss": 0.2836,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8472740698199123,
+      "grad_norm": 0.7408990623509364,
+      "learning_rate": 5.997055917946893e-07,
+      "loss": 0.2832,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8473561143701029,
+      "grad_norm": 0.7280357938147507,
+      "learning_rate": 5.99074791725035e-07,
+      "loss": 0.2934,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8474381589202937,
+      "grad_norm": 0.684753840239408,
+      "learning_rate": 5.984443024416476e-07,
+      "loss": 0.2546,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8475202034704845,
+      "grad_norm": 0.8556288922223203,
+      "learning_rate": 5.978141239890528e-07,
+      "loss": 0.3155,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8476022480206752,
+      "grad_norm": 0.6669321023419676,
+      "learning_rate": 5.971842564117513e-07,
+      "loss": 0.3302,
+      "step": 10331
+    },
+    {
+      "epoch": 0.847684292570866,
+      "grad_norm": 0.6822213727521311,
+      "learning_rate": 5.965546997542238e-07,
+      "loss": 0.275,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8477663371210568,
+      "grad_norm": 0.6899421848887478,
+      "learning_rate": 5.959254540609294e-07,
+      "loss": 0.3444,
+      "step": 10333
+    },
+    {
+      "epoch": 0.8478483816712475,
+      "grad_norm": 0.7191092170344996,
+      "learning_rate": 5.952965193763028e-07,
+      "loss": 0.2757,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8479304262214382,
+      "grad_norm": 0.7316979220250501,
+      "learning_rate": 5.946678957447605e-07,
+      "loss": 0.3228,
+      "step": 10335
+    },
+    {
+      "epoch": 0.848012470771629,
+      "grad_norm": 0.6639212430139187,
+      "learning_rate": 5.940395832106926e-07,
+      "loss": 0.3067,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8480945153218198,
+      "grad_norm": 0.7035053871295954,
+      "learning_rate": 5.93411581818471e-07,
+      "loss": 0.3332,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8481765598720105,
+      "grad_norm": 0.6563462668635291,
+      "learning_rate": 5.927838916124445e-07,
+      "loss": 0.3233,
+      "step": 10338
+    },
+    {
+      "epoch": 0.8482586044222012,
+      "grad_norm": 0.7112914480458847,
+      "learning_rate": 5.921565126369378e-07,
+      "loss": 0.314,
+      "step": 10339
+    },
+    {
+      "epoch": 0.848340648972392,
+      "grad_norm": 0.7807210717509075,
+      "learning_rate": 5.915294449362574e-07,
+      "loss": 0.3006,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8484226935225828,
+      "grad_norm": 0.6784822425853597,
+      "learning_rate": 5.909026885546837e-07,
+      "loss": 0.3064,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8485047380727735,
+      "grad_norm": 0.6208071693002277,
+      "learning_rate": 5.902762435364795e-07,
+      "loss": 0.3106,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8485867826229643,
+      "grad_norm": 0.6817413556593023,
+      "learning_rate": 5.896501099258822e-07,
+      "loss": 0.2893,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8486688271731551,
+      "grad_norm": 0.7574218045612284,
+      "learning_rate": 5.890242877671077e-07,
+      "loss": 0.2735,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8487508717233457,
+      "grad_norm": 0.6194824475603423,
+      "learning_rate": 5.883987771043509e-07,
+      "loss": 0.2623,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8488329162735365,
+      "grad_norm": 0.683529713255001,
+      "learning_rate": 5.877735779817861e-07,
+      "loss": 0.2878,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8489149608237273,
+      "grad_norm": 0.7073907179791113,
+      "learning_rate": 5.871486904435608e-07,
+      "loss": 0.3136,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8489970053739181,
+      "grad_norm": 0.6483325646988438,
+      "learning_rate": 5.865241145338063e-07,
+      "loss": 0.2579,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8490790499241088,
+      "grad_norm": 0.7347414503717626,
+      "learning_rate": 5.858998502966273e-07,
+      "loss": 0.2872,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8491610944742995,
+      "grad_norm": 0.7583775566664461,
+      "learning_rate": 5.852758977761091e-07,
+      "loss": 0.289,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8492431390244903,
+      "grad_norm": 0.7980671512134331,
+      "learning_rate": 5.846522570163155e-07,
+      "loss": 0.3286,
+      "step": 10351
+    },
+    {
+      "epoch": 0.849325183574681,
+      "grad_norm": 0.7833204960407256,
+      "learning_rate": 5.840289280612837e-07,
+      "loss": 0.3206,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8494072281248718,
+      "grad_norm": 0.6685430292974563,
+      "learning_rate": 5.83405910955036e-07,
+      "loss": 0.252,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8494892726750626,
+      "grad_norm": 0.5750977496340008,
+      "learning_rate": 5.827832057415661e-07,
+      "loss": 0.6414,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8495713172252534,
+      "grad_norm": 0.6691806867516952,
+      "learning_rate": 5.821608124648504e-07,
+      "loss": 0.2967,
+      "step": 10355
+    },
+    {
+      "epoch": 0.849653361775444,
+      "grad_norm": 0.676831371654559,
+      "learning_rate": 5.815387311688398e-07,
+      "loss": 0.304,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8497354063256348,
+      "grad_norm": 0.6890346521888092,
+      "learning_rate": 5.809169618974647e-07,
+      "loss": 0.3053,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8498174508758256,
+      "grad_norm": 0.7019609771137237,
+      "learning_rate": 5.802955046946335e-07,
+      "loss": 0.2852,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8498994954260163,
+      "grad_norm": 0.7099042120611966,
+      "learning_rate": 5.796743596042342e-07,
+      "loss": 0.2961,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8499815399762071,
+      "grad_norm": 0.6415029034264929,
+      "learning_rate": 5.790535266701291e-07,
+      "loss": 0.2426,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8500635845263979,
+      "grad_norm": 0.7275649537648752,
+      "learning_rate": 5.78433005936162e-07,
+      "loss": 0.3098,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8501456290765886,
+      "grad_norm": 0.7431540623331732,
+      "learning_rate": 5.778127974461511e-07,
+      "loss": 0.2771,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8502276736267793,
+      "grad_norm": 0.9558456150636936,
+      "learning_rate": 5.77192901243896e-07,
+      "loss": 0.2751,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8503097181769701,
+      "grad_norm": 0.6944672573083066,
+      "learning_rate": 5.765733173731731e-07,
+      "loss": 0.3185,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8503917627271609,
+      "grad_norm": 0.7766038985934067,
+      "learning_rate": 5.759540458777346e-07,
+      "loss": 0.3429,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8504738072773516,
+      "grad_norm": 0.6437450085160427,
+      "learning_rate": 5.753350868013147e-07,
+      "loss": 0.2682,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8505558518275423,
+      "grad_norm": 0.7044979105242638,
+      "learning_rate": 5.747164401876215e-07,
+      "loss": 0.3437,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8506378963777331,
+      "grad_norm": 0.7378429335326553,
+      "learning_rate": 5.740981060803441e-07,
+      "loss": 0.2864,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8507199409279239,
+      "grad_norm": 0.7265763038268253,
+      "learning_rate": 5.73480084523147e-07,
+      "loss": 0.3108,
+      "step": 10369
+    },
+    {
+      "epoch": 0.8508019854781146,
+      "grad_norm": 0.7162800187874928,
+      "learning_rate": 5.728623755596757e-07,
+      "loss": 0.2912,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8508840300283054,
+      "grad_norm": 0.7142703942875657,
+      "learning_rate": 5.722449792335505e-07,
+      "loss": 0.3397,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8509660745784962,
+      "grad_norm": 0.6741223663355562,
+      "learning_rate": 5.716278955883703e-07,
+      "loss": 0.2969,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8510481191286868,
+      "grad_norm": 0.7336860449656302,
+      "learning_rate": 5.71011124667713e-07,
+      "loss": 0.2748,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8511301636788776,
+      "grad_norm": 0.7284255807436877,
+      "learning_rate": 5.703946665151356e-07,
+      "loss": 0.2712,
+      "step": 10374
+    },
+    {
+      "epoch": 0.8512122082290684,
+      "grad_norm": 0.9120423496459216,
+      "learning_rate": 5.697785211741691e-07,
+      "loss": 0.2902,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8512942527792592,
+      "grad_norm": 0.6264882114520404,
+      "learning_rate": 5.691626886883261e-07,
+      "loss": 0.3097,
+      "step": 10376
+    },
+    {
+      "epoch": 0.8513762973294499,
+      "grad_norm": 0.8081314332262372,
+      "learning_rate": 5.685471691010958e-07,
+      "loss": 0.2955,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8514583418796406,
+      "grad_norm": 0.7532676656918502,
+      "learning_rate": 5.679319624559443e-07,
+      "loss": 0.2899,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8515403864298314,
+      "grad_norm": 0.774682559067121,
+      "learning_rate": 5.673170687963175e-07,
+      "loss": 0.3062,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8516224309800221,
+      "grad_norm": 0.6964211804306668,
+      "learning_rate": 5.667024881656369e-07,
+      "loss": 0.28,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8517044755302129,
+      "grad_norm": 0.720734413559343,
+      "learning_rate": 5.660882206073037e-07,
+      "loss": 0.2868,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8517865200804037,
+      "grad_norm": 0.7430745315309292,
+      "learning_rate": 5.654742661646978e-07,
+      "loss": 0.3166,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8518685646305945,
+      "grad_norm": 0.6647366089461092,
+      "learning_rate": 5.64860624881175e-07,
+      "loss": 0.2939,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8519506091807851,
+      "grad_norm": 0.682267264688231,
+      "learning_rate": 5.64247296800069e-07,
+      "loss": 0.2798,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8520326537309759,
+      "grad_norm": 0.7289661049191652,
+      "learning_rate": 5.636342819646912e-07,
+      "loss": 0.3102,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8521146982811667,
+      "grad_norm": 0.7389550224448883,
+      "learning_rate": 5.630215804183325e-07,
+      "loss": 0.3056,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8521967428313574,
+      "grad_norm": 0.7154451437812105,
+      "learning_rate": 5.624091922042629e-07,
+      "loss": 0.3318,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8522787873815482,
+      "grad_norm": 0.6851803718290775,
+      "learning_rate": 5.617971173657255e-07,
+      "loss": 0.3027,
+      "step": 10388
+    },
+    {
+      "epoch": 0.852360831931739,
+      "grad_norm": 0.7103246153239989,
+      "learning_rate": 5.611853559459457e-07,
+      "loss": 0.2833,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8524428764819297,
+      "grad_norm": 0.6632762634872797,
+      "learning_rate": 5.60573907988124e-07,
+      "loss": 0.28,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8525249210321204,
+      "grad_norm": 0.900563916184925,
+      "learning_rate": 5.599627735354408e-07,
+      "loss": 0.3189,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8526069655823112,
+      "grad_norm": 0.7171355958033043,
+      "learning_rate": 5.593519526310532e-07,
+      "loss": 0.2766,
+      "step": 10392
+    },
+    {
+      "epoch": 0.852689010132502,
+      "grad_norm": 0.6805815590547873,
+      "learning_rate": 5.587414453180956e-07,
+      "loss": 0.2562,
+      "step": 10393
+    },
+    {
+      "epoch": 0.8527710546826927,
+      "grad_norm": 0.7324951347957896,
+      "learning_rate": 5.58131251639682e-07,
+      "loss": 0.3056,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8528530992328834,
+      "grad_norm": 0.7294097034867921,
+      "learning_rate": 5.575213716389039e-07,
+      "loss": 0.2914,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8529351437830742,
+      "grad_norm": 0.7813700489017142,
+      "learning_rate": 5.569118053588291e-07,
+      "loss": 0.3298,
+      "step": 10396
+    },
+    {
+      "epoch": 0.853017188333265,
+      "grad_norm": 0.6723269165323361,
+      "learning_rate": 5.563025528425031e-07,
+      "loss": 0.2931,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8530992328834557,
+      "grad_norm": 0.6449872372789548,
+      "learning_rate": 5.556936141329521e-07,
+      "loss": 0.3207,
+      "step": 10398
+    },
+    {
+      "epoch": 0.8531812774336465,
+      "grad_norm": 0.6689077863087264,
+      "learning_rate": 5.550849892731774e-07,
+      "loss": 0.2865,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8532633219838373,
+      "grad_norm": 0.6809660313600899,
+      "learning_rate": 5.5447667830616e-07,
+      "loss": 0.3031,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8533453665340279,
+      "grad_norm": 0.706072518033087,
+      "learning_rate": 5.538686812748567e-07,
+      "loss": 0.3002,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8534274110842187,
+      "grad_norm": 0.7415221494135418,
+      "learning_rate": 5.532609982222048e-07,
+      "loss": 0.2969,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8535094556344095,
+      "grad_norm": 0.6943602068261671,
+      "learning_rate": 5.526536291911161e-07,
+      "loss": 0.3231,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8535915001846003,
+      "grad_norm": 0.7152807001278461,
+      "learning_rate": 5.520465742244824e-07,
+      "loss": 0.3091,
+      "step": 10404
+    },
+    {
+      "epoch": 0.853673544734791,
+      "grad_norm": 0.7147983428794491,
+      "learning_rate": 5.51439833365175e-07,
+      "loss": 0.3184,
+      "step": 10405
+    },
+    {
+      "epoch": 0.8537555892849817,
+      "grad_norm": 0.6403205842887822,
+      "learning_rate": 5.508334066560377e-07,
+      "loss": 0.2367,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8538376338351725,
+      "grad_norm": 0.6868125756271176,
+      "learning_rate": 5.50227294139899e-07,
+      "loss": 0.3062,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8539196783853632,
+      "grad_norm": 0.7533856128457211,
+      "learning_rate": 5.496214958595581e-07,
+      "loss": 0.2897,
+      "step": 10408
+    },
+    {
+      "epoch": 0.854001722935554,
+      "grad_norm": 0.6284149822149206,
+      "learning_rate": 5.490160118577969e-07,
+      "loss": 0.2767,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8540837674857448,
+      "grad_norm": 0.637918027421002,
+      "learning_rate": 5.484108421773754e-07,
+      "loss": 0.2732,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8541658120359356,
+      "grad_norm": 0.7400805943978099,
+      "learning_rate": 5.47805986861028e-07,
+      "loss": 0.3079,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8542478565861262,
+      "grad_norm": 0.6425002800818335,
+      "learning_rate": 5.47201445951468e-07,
+      "loss": 0.3107,
+      "step": 10412
+    },
+    {
+      "epoch": 0.854329901136317,
+      "grad_norm": 0.6761175538803809,
+      "learning_rate": 5.465972194913882e-07,
+      "loss": 0.2913,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8544119456865078,
+      "grad_norm": 0.6820315769990121,
+      "learning_rate": 5.459933075234575e-07,
+      "loss": 0.2725,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8544939902366985,
+      "grad_norm": 0.7562915005474405,
+      "learning_rate": 5.453897100903244e-07,
+      "loss": 0.3374,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8545760347868893,
+      "grad_norm": 0.7034198556967811,
+      "learning_rate": 5.447864272346121e-07,
+      "loss": 0.3039,
+      "step": 10416
+    },
+    {
+      "epoch": 0.85465807933708,
+      "grad_norm": 0.6787052994827447,
+      "learning_rate": 5.441834589989242e-07,
+      "loss": 0.3333,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8547401238872708,
+      "grad_norm": 0.6785082215551858,
+      "learning_rate": 5.435808054258429e-07,
+      "loss": 0.2746,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8548221684374615,
+      "grad_norm": 0.6339662603247014,
+      "learning_rate": 5.42978466557924e-07,
+      "loss": 0.2977,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8549042129876523,
+      "grad_norm": 0.8221744072154719,
+      "learning_rate": 5.423764424377065e-07,
+      "loss": 0.3074,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8549862575378431,
+      "grad_norm": 0.6326773679031907,
+      "learning_rate": 5.417747331077017e-07,
+      "loss": 0.2799,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8550683020880337,
+      "grad_norm": 0.6602045663749527,
+      "learning_rate": 5.411733386104029e-07,
+      "loss": 0.3198,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8551503466382245,
+      "grad_norm": 1.1426886383101764,
+      "learning_rate": 5.4057225898828e-07,
+      "loss": 0.2626,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8552323911884153,
+      "grad_norm": 0.6565848822833811,
+      "learning_rate": 5.399714942837792e-07,
+      "loss": 0.3035,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8553144357386061,
+      "grad_norm": 0.6191210434127241,
+      "learning_rate": 5.393710445393252e-07,
+      "loss": 0.3172,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8553964802887968,
+      "grad_norm": 0.7002029530069326,
+      "learning_rate": 5.387709097973226e-07,
+      "loss": 0.2795,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8554785248389876,
+      "grad_norm": 0.7428679571900466,
+      "learning_rate": 5.381710901001497e-07,
+      "loss": 0.3342,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8555605693891783,
+      "grad_norm": 0.6557150723226425,
+      "learning_rate": 5.37571585490167e-07,
+      "loss": 0.2916,
+      "step": 10428
+    },
+    {
+      "epoch": 0.855642613939369,
+      "grad_norm": 0.6448007748588743,
+      "learning_rate": 5.369723960097084e-07,
+      "loss": 0.3143,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8557246584895598,
+      "grad_norm": 0.8302972814292012,
+      "learning_rate": 5.36373521701089e-07,
+      "loss": 0.324,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8558067030397506,
+      "grad_norm": 0.7029523578995263,
+      "learning_rate": 5.357749626066011e-07,
+      "loss": 0.2972,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8558887475899414,
+      "grad_norm": 0.6873694549125294,
+      "learning_rate": 5.351767187685114e-07,
+      "loss": 0.2868,
+      "step": 10432
+    },
+    {
+      "epoch": 0.855970792140132,
+      "grad_norm": 0.6357403908278147,
+      "learning_rate": 5.345787902290706e-07,
+      "loss": 0.2672,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8560528366903228,
+      "grad_norm": 0.775093932960794,
+      "learning_rate": 5.339811770304992e-07,
+      "loss": 0.2942,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8561348812405136,
+      "grad_norm": 0.7823288491058169,
+      "learning_rate": 5.333838792150026e-07,
+      "loss": 0.27,
+      "step": 10435
+    },
+    {
+      "epoch": 0.8562169257907043,
+      "grad_norm": 0.7250026856526447,
+      "learning_rate": 5.327868968247613e-07,
+      "loss": 0.2824,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8562989703408951,
+      "grad_norm": 0.726616111753451,
+      "learning_rate": 5.32190229901931e-07,
+      "loss": 0.2918,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8563810148910859,
+      "grad_norm": 0.7043334229192754,
+      "learning_rate": 5.315938784886499e-07,
+      "loss": 0.2862,
+      "step": 10438
+    },
+    {
+      "epoch": 0.8564630594412767,
+      "grad_norm": 0.694805681065564,
+      "learning_rate": 5.309978426270296e-07,
+      "loss": 0.2819,
+      "step": 10439
+    },
+    {
+      "epoch": 0.8565451039914673,
+      "grad_norm": 0.705325667850079,
+      "learning_rate": 5.30402122359161e-07,
+      "loss": 0.3204,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8566271485416581,
+      "grad_norm": 0.8596593154287862,
+      "learning_rate": 5.298067177271144e-07,
+      "loss": 0.2666,
+      "step": 10441
+    },
+    {
+      "epoch": 0.8567091930918489,
+      "grad_norm": 0.743331309529327,
+      "learning_rate": 5.292116287729348e-07,
+      "loss": 0.2972,
+      "step": 10442
+    },
+    {
+      "epoch": 0.8567912376420397,
+      "grad_norm": 0.740275600333234,
+      "learning_rate": 5.286168555386478e-07,
+      "loss": 0.2848,
+      "step": 10443
+    },
+    {
+      "epoch": 0.8568732821922304,
+      "grad_norm": 0.77122509682468,
+      "learning_rate": 5.280223980662535e-07,
+      "loss": 0.3113,
+      "step": 10444
+    },
+    {
+      "epoch": 0.8569553267424211,
+      "grad_norm": 0.7866464086927509,
+      "learning_rate": 5.274282563977328e-07,
+      "loss": 0.3407,
+      "step": 10445
+    },
+    {
+      "epoch": 0.8570373712926119,
+      "grad_norm": 0.7472908306023793,
+      "learning_rate": 5.268344305750439e-07,
+      "loss": 0.264,
+      "step": 10446
+    },
+    {
+      "epoch": 0.8571194158428026,
+      "grad_norm": 0.6482782979690823,
+      "learning_rate": 5.262409206401198e-07,
+      "loss": 0.2551,
+      "step": 10447
+    },
+    {
+      "epoch": 0.8572014603929934,
+      "grad_norm": 0.8025204862231836,
+      "learning_rate": 5.256477266348747e-07,
+      "loss": 0.3268,
+      "step": 10448
+    },
+    {
+      "epoch": 0.8572835049431842,
+      "grad_norm": 0.673269270159495,
+      "learning_rate": 5.250548486011992e-07,
+      "loss": 0.3102,
+      "step": 10449
+    },
+    {
+      "epoch": 0.857365549493375,
+      "grad_norm": 0.7138217367168923,
+      "learning_rate": 5.244622865809596e-07,
+      "loss": 0.2879,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8574475940435656,
+      "grad_norm": 0.6954340638861432,
+      "learning_rate": 5.238700406160036e-07,
+      "loss": 0.3506,
+      "step": 10451
+    },
+    {
+      "epoch": 0.8575296385937564,
+      "grad_norm": 0.6844201280390024,
+      "learning_rate": 5.232781107481544e-07,
+      "loss": 0.2699,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8576116831439472,
+      "grad_norm": 0.676398154557605,
+      "learning_rate": 5.226864970192114e-07,
+      "loss": 0.2571,
+      "step": 10453
+    },
+    {
+      "epoch": 0.8576937276941379,
+      "grad_norm": 0.6679555759989788,
+      "learning_rate": 5.220951994709555e-07,
+      "loss": 0.2765,
+      "step": 10454
+    },
+    {
+      "epoch": 0.8577757722443287,
+      "grad_norm": 0.6921434902122404,
+      "learning_rate": 5.215042181451418e-07,
+      "loss": 0.2639,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8578578167945194,
+      "grad_norm": 0.8025424103986046,
+      "learning_rate": 5.209135530835053e-07,
+      "loss": 0.3034,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8579398613447102,
+      "grad_norm": 0.7148671966293527,
+      "learning_rate": 5.203232043277568e-07,
+      "loss": 0.3283,
+      "step": 10457
+    },
+    {
+      "epoch": 0.8580219058949009,
+      "grad_norm": 1.0134099101786544,
+      "learning_rate": 5.197331719195864e-07,
+      "loss": 0.3211,
+      "step": 10458
+    },
+    {
+      "epoch": 0.8581039504450917,
+      "grad_norm": 0.7256680676321415,
+      "learning_rate": 5.191434559006625e-07,
+      "loss": 0.2531,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8581859949952825,
+      "grad_norm": 0.7182383618776872,
+      "learning_rate": 5.185540563126274e-07,
+      "loss": 0.3272,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8582680395454731,
+      "grad_norm": 0.7496723603065563,
+      "learning_rate": 5.179649731971059e-07,
+      "loss": 0.3385,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8583500840956639,
+      "grad_norm": 0.6891523729107191,
+      "learning_rate": 5.173762065956967e-07,
+      "loss": 0.2811,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8584321286458547,
+      "grad_norm": 0.6614636908200474,
+      "learning_rate": 5.167877565499774e-07,
+      "loss": 0.3019,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8585141731960455,
+      "grad_norm": 0.7819332830741355,
+      "learning_rate": 5.161996231015049e-07,
+      "loss": 0.2909,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8585962177462362,
+      "grad_norm": 0.7788197101768075,
+      "learning_rate": 5.156118062918098e-07,
+      "loss": 0.3067,
+      "step": 10465
+    },
+    {
+      "epoch": 0.858678262296427,
+      "grad_norm": 0.659613550161262,
+      "learning_rate": 5.150243061624055e-07,
+      "loss": 0.3223,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8587603068466177,
+      "grad_norm": 0.6426622950436653,
+      "learning_rate": 5.144371227547795e-07,
+      "loss": 0.2631,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8588423513968084,
+      "grad_norm": 0.6551515609050332,
+      "learning_rate": 5.138502561103959e-07,
+      "loss": 0.284,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8589243959469992,
+      "grad_norm": 0.7121862042590171,
+      "learning_rate": 5.132637062707008e-07,
+      "loss": 0.2899,
+      "step": 10469
+    },
+    {
+      "epoch": 0.85900644049719,
+      "grad_norm": 0.7242889722574,
+      "learning_rate": 5.126774732771129e-07,
+      "loss": 0.2746,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8590884850473808,
+      "grad_norm": 0.7314597998236294,
+      "learning_rate": 5.120915571710328e-07,
+      "loss": 0.3179,
+      "step": 10471
+    },
+    {
+      "epoch": 0.8591705295975715,
+      "grad_norm": 0.6882312893698928,
+      "learning_rate": 5.115059579938375e-07,
+      "loss": 0.2807,
+      "step": 10472
+    },
+    {
+      "epoch": 0.8592525741477622,
+      "grad_norm": 0.6724404733925822,
+      "learning_rate": 5.10920675786879e-07,
+      "loss": 0.3045,
+      "step": 10473
+    },
+    {
+      "epoch": 0.859334618697953,
+      "grad_norm": 0.6899778077785419,
+      "learning_rate": 5.103357105914913e-07,
+      "loss": 0.3121,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8594166632481437,
+      "grad_norm": 0.6872568677008554,
+      "learning_rate": 5.097510624489816e-07,
+      "loss": 0.3075,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8594987077983345,
+      "grad_norm": 0.6987369269253456,
+      "learning_rate": 5.091667314006371e-07,
+      "loss": 0.3149,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8595807523485253,
+      "grad_norm": 0.6898171363510426,
+      "learning_rate": 5.085827174877245e-07,
+      "loss": 0.301,
+      "step": 10477
+    },
+    {
+      "epoch": 0.859662796898716,
+      "grad_norm": 0.7249993497516516,
+      "learning_rate": 5.079990207514835e-07,
+      "loss": 0.2811,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8597448414489067,
+      "grad_norm": 0.6349448895992833,
+      "learning_rate": 5.074156412331354e-07,
+      "loss": 0.3054,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8598268859990975,
+      "grad_norm": 0.7958607152716314,
+      "learning_rate": 5.068325789738771e-07,
+      "loss": 0.3121,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8599089305492883,
+      "grad_norm": 0.6541176178380216,
+      "learning_rate": 5.062498340148819e-07,
+      "loss": 0.279,
+      "step": 10481
+    },
+    {
+      "epoch": 0.859990975099479,
+      "grad_norm": 0.6699619987391173,
+      "learning_rate": 5.05667406397305e-07,
+      "loss": 0.2694,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8600730196496698,
+      "grad_norm": 0.6631662651713326,
+      "learning_rate": 5.050852961622738e-07,
+      "loss": 0.2864,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8601550641998605,
+      "grad_norm": 0.6733560806983567,
+      "learning_rate": 5.045035033508977e-07,
+      "loss": 0.2884,
+      "step": 10484
+    },
+    {
+      "epoch": 0.8602371087500513,
+      "grad_norm": 0.753403254868042,
+      "learning_rate": 5.039220280042623e-07,
+      "loss": 0.3074,
+      "step": 10485
+    },
+    {
+      "epoch": 0.860319153300242,
+      "grad_norm": 0.7152986149465859,
+      "learning_rate": 5.03340870163429e-07,
+      "loss": 0.2917,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8604011978504328,
+      "grad_norm": 0.7349895724552785,
+      "learning_rate": 5.027600298694397e-07,
+      "loss": 0.3146,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8604832424006236,
+      "grad_norm": 0.6238752080764198,
+      "learning_rate": 5.021795071633113e-07,
+      "loss": 0.2829,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8605652869508142,
+      "grad_norm": 0.758593555709412,
+      "learning_rate": 5.015993020860394e-07,
+      "loss": 0.3033,
+      "step": 10489
+    },
+    {
+      "epoch": 0.860647331501005,
+      "grad_norm": 0.7564498645106432,
+      "learning_rate": 5.010194146785985e-07,
+      "loss": 0.2973,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8607293760511958,
+      "grad_norm": 0.6663847303224121,
+      "learning_rate": 5.004398449819376e-07,
+      "loss": 0.2763,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8608114206013866,
+      "grad_norm": 0.7854330484093242,
+      "learning_rate": 4.998605930369865e-07,
+      "loss": 0.378,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8608934651515773,
+      "grad_norm": 0.6984120660348335,
+      "learning_rate": 4.992816588846495e-07,
+      "loss": 0.313,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8609755097017681,
+      "grad_norm": 0.6882376371826255,
+      "learning_rate": 4.987030425658118e-07,
+      "loss": 0.3145,
+      "step": 10494
+    },
+    {
+      "epoch": 0.8610575542519588,
+      "grad_norm": 0.6997104342234349,
+      "learning_rate": 4.981247441213333e-07,
+      "loss": 0.2971,
+      "step": 10495
+    },
+    {
+      "epoch": 0.8611395988021495,
+      "grad_norm": 0.63931761248081,
+      "learning_rate": 4.975467635920517e-07,
+      "loss": 0.3301,
+      "step": 10496
+    },
+    {
+      "epoch": 0.8612216433523403,
+      "grad_norm": 1.1531016897906137,
+      "learning_rate": 4.969691010187838e-07,
+      "loss": 0.2896,
+      "step": 10497
+    },
+    {
+      "epoch": 0.8613036879025311,
+      "grad_norm": 0.6868647661993389,
+      "learning_rate": 4.963917564423243e-07,
+      "loss": 0.3208,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8613857324527219,
+      "grad_norm": 0.6673085592368974,
+      "learning_rate": 4.958147299034421e-07,
+      "loss": 0.2809,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8614677770029125,
+      "grad_norm": 0.6787717353470205,
+      "learning_rate": 4.952380214428886e-07,
+      "loss": 0.2981,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8615498215531033,
+      "grad_norm": 0.6513231179327369,
+      "learning_rate": 4.946616311013875e-07,
+      "loss": 0.3033,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8616318661032941,
+      "grad_norm": 1.2266266786498807,
+      "learning_rate": 4.940855589196436e-07,
+      "loss": 0.3027,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8617139106534848,
+      "grad_norm": 0.8163264517860556,
+      "learning_rate": 4.935098049383391e-07,
+      "loss": 0.3193,
+      "step": 10503
+    },
+    {
+      "epoch": 0.8617959552036756,
+      "grad_norm": 0.7063359504369663,
+      "learning_rate": 4.929343691981309e-07,
+      "loss": 0.3101,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8618779997538664,
+      "grad_norm": 0.7150605256071769,
+      "learning_rate": 4.923592517396575e-07,
+      "loss": 0.3024,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8619600443040571,
+      "grad_norm": 0.7466792144958719,
+      "learning_rate": 4.917844526035304e-07,
+      "loss": 0.2988,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8620420888542478,
+      "grad_norm": 0.6692988897640089,
+      "learning_rate": 4.912099718303437e-07,
+      "loss": 0.3001,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8621241334044386,
+      "grad_norm": 0.6891191662524733,
+      "learning_rate": 4.906358094606645e-07,
+      "loss": 0.2895,
+      "step": 10508
+    },
+    {
+      "epoch": 0.8622061779546294,
+      "grad_norm": 0.7160723844988542,
+      "learning_rate": 4.900619655350386e-07,
+      "loss": 0.2814,
+      "step": 10509
+    },
+    {
+      "epoch": 0.8622882225048201,
+      "grad_norm": 0.6645773065970483,
+      "learning_rate": 4.89488440093992e-07,
+      "loss": 0.3042,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8623702670550109,
+      "grad_norm": 0.7181509670597426,
+      "learning_rate": 4.889152331780245e-07,
+      "loss": 0.2725,
+      "step": 10511
+    },
+    {
+      "epoch": 0.8624523116052016,
+      "grad_norm": 0.7054924422197454,
+      "learning_rate": 4.883423448276154e-07,
+      "loss": 0.3154,
+      "step": 10512
+    },
+    {
+      "epoch": 0.8625343561553924,
+      "grad_norm": 0.6474819174556822,
+      "learning_rate": 4.877697750832222e-07,
+      "loss": 0.2696,
+      "step": 10513
+    },
+    {
+      "epoch": 0.8626164007055831,
+      "grad_norm": 0.6936041080635829,
+      "learning_rate": 4.871975239852766e-07,
+      "loss": 0.2755,
+      "step": 10514
+    },
+    {
+      "epoch": 0.8626984452557739,
+      "grad_norm": 0.7249458561088006,
+      "learning_rate": 4.866255915741924e-07,
+      "loss": 0.3272,
+      "step": 10515
+    },
+    {
+      "epoch": 0.8627804898059647,
+      "grad_norm": 0.739793704839267,
+      "learning_rate": 4.860539778903579e-07,
+      "loss": 0.3214,
+      "step": 10516
+    },
+    {
+      "epoch": 0.8628625343561553,
+      "grad_norm": 0.6743627336815983,
+      "learning_rate": 4.854826829741388e-07,
+      "loss": 0.2791,
+      "step": 10517
+    },
+    {
+      "epoch": 0.8629445789063461,
+      "grad_norm": 0.8658713750638113,
+      "learning_rate": 4.849117068658799e-07,
+      "loss": 0.2941,
+      "step": 10518
+    },
+    {
+      "epoch": 0.8630266234565369,
+      "grad_norm": 0.6522476186278326,
+      "learning_rate": 4.84341049605902e-07,
+      "loss": 0.2717,
+      "step": 10519
+    },
+    {
+      "epoch": 0.8631086680067277,
+      "grad_norm": 0.7976895857172894,
+      "learning_rate": 4.837707112345047e-07,
+      "loss": 0.3045,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8631907125569184,
+      "grad_norm": 0.7330548131610825,
+      "learning_rate": 4.832006917919629e-07,
+      "loss": 0.3186,
+      "step": 10521
+    },
+    {
+      "epoch": 0.8632727571071092,
+      "grad_norm": 0.6931152034899195,
+      "learning_rate": 4.826309913185323e-07,
+      "loss": 0.2735,
+      "step": 10522
+    },
+    {
+      "epoch": 0.8633548016572999,
+      "grad_norm": 0.6936469542269598,
+      "learning_rate": 4.820616098544439e-07,
+      "loss": 0.2942,
+      "step": 10523
+    },
+    {
+      "epoch": 0.8634368462074906,
+      "grad_norm": 0.6911985445388491,
+      "learning_rate": 4.814925474399046e-07,
+      "loss": 0.2855,
+      "step": 10524
+    },
+    {
+      "epoch": 0.8635188907576814,
+      "grad_norm": 0.6528171547816409,
+      "learning_rate": 4.809238041151021e-07,
+      "loss": 0.2784,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8636009353078722,
+      "grad_norm": 0.7140323963227849,
+      "learning_rate": 4.803553799202016e-07,
+      "loss": 0.3233,
+      "step": 10526
+    },
+    {
+      "epoch": 0.863682979858063,
+      "grad_norm": 0.6922333382148136,
+      "learning_rate": 4.797872748953414e-07,
+      "loss": 0.296,
+      "step": 10527
+    },
+    {
+      "epoch": 0.8637650244082536,
+      "grad_norm": 0.755751191371869,
+      "learning_rate": 4.792194890806423e-07,
+      "loss": 0.3159,
+      "step": 10528
+    },
+    {
+      "epoch": 0.8638470689584444,
+      "grad_norm": 0.7010512023219468,
+      "learning_rate": 4.786520225161989e-07,
+      "loss": 0.3629,
+      "step": 10529
+    },
+    {
+      "epoch": 0.8639291135086352,
+      "grad_norm": 0.6106439593008155,
+      "learning_rate": 4.780848752420858e-07,
+      "loss": 0.2855,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8640111580588259,
+      "grad_norm": 0.6927639054312313,
+      "learning_rate": 4.775180472983549e-07,
+      "loss": 0.3256,
+      "step": 10531
+    },
+    {
+      "epoch": 0.8640932026090167,
+      "grad_norm": 0.7454284309097802,
+      "learning_rate": 4.769515387250317e-07,
+      "loss": 0.3056,
+      "step": 10532
+    },
+    {
+      "epoch": 0.8641752471592075,
+      "grad_norm": 0.7037712285585387,
+      "learning_rate": 4.763853495621251e-07,
+      "loss": 0.3228,
+      "step": 10533
+    },
+    {
+      "epoch": 0.8642572917093982,
+      "grad_norm": 0.7106786644762823,
+      "learning_rate": 4.75819479849618e-07,
+      "loss": 0.3025,
+      "step": 10534
+    },
+    {
+      "epoch": 0.8643393362595889,
+      "grad_norm": 0.6866205349653082,
+      "learning_rate": 4.7525392962746997e-07,
+      "loss": 0.2879,
+      "step": 10535
+    },
+    {
+      "epoch": 0.8644213808097797,
+      "grad_norm": 0.6732242113803065,
+      "learning_rate": 4.746886989356203e-07,
+      "loss": 0.2761,
+      "step": 10536
+    },
+    {
+      "epoch": 0.8645034253599705,
+      "grad_norm": 0.692972879479816,
+      "learning_rate": 4.741237878139837e-07,
+      "loss": 0.2824,
+      "step": 10537
+    },
+    {
+      "epoch": 0.8645854699101612,
+      "grad_norm": 0.7355698215011806,
+      "learning_rate": 4.735591963024533e-07,
+      "loss": 0.3321,
+      "step": 10538
+    },
+    {
+      "epoch": 0.864667514460352,
+      "grad_norm": 0.7520244934976308,
+      "learning_rate": 4.72994924440901e-07,
+      "loss": 0.3283,
+      "step": 10539
+    },
+    {
+      "epoch": 0.8647495590105427,
+      "grad_norm": 0.7449849003558008,
+      "learning_rate": 4.7243097226917334e-07,
+      "loss": 0.3028,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8648316035607335,
+      "grad_norm": 0.7487393427162734,
+      "learning_rate": 4.718673398270973e-07,
+      "loss": 0.3012,
+      "step": 10541
+    },
+    {
+      "epoch": 0.8649136481109242,
+      "grad_norm": 0.7006890708276878,
+      "learning_rate": 4.7130402715447386e-07,
+      "loss": 0.3218,
+      "step": 10542
+    },
+    {
+      "epoch": 0.864995692661115,
+      "grad_norm": 0.6759176719636786,
+      "learning_rate": 4.7074103429108397e-07,
+      "loss": 0.2836,
+      "step": 10543
+    },
+    {
+      "epoch": 0.8650777372113058,
+      "grad_norm": 0.720447192177282,
+      "learning_rate": 4.7017836127668636e-07,
+      "loss": 0.3086,
+      "step": 10544
+    },
+    {
+      "epoch": 0.8651597817614964,
+      "grad_norm": 0.7512951729160456,
+      "learning_rate": 4.696160081510143e-07,
+      "loss": 0.2904,
+      "step": 10545
+    },
+    {
+      "epoch": 0.8652418263116872,
+      "grad_norm": 0.7252883439424053,
+      "learning_rate": 4.69053974953782e-07,
+      "loss": 0.3021,
+      "step": 10546
+    },
+    {
+      "epoch": 0.865323870861878,
+      "grad_norm": 0.6633344323919425,
+      "learning_rate": 4.684922617246773e-07,
+      "loss": 0.2788,
+      "step": 10547
+    },
+    {
+      "epoch": 0.8654059154120688,
+      "grad_norm": 0.7353065783758499,
+      "learning_rate": 4.679308685033701e-07,
+      "loss": 0.3082,
+      "step": 10548
+    },
+    {
+      "epoch": 0.8654879599622595,
+      "grad_norm": 0.7351368057104488,
+      "learning_rate": 4.67369795329502e-07,
+      "loss": 0.3033,
+      "step": 10549
+    },
+    {
+      "epoch": 0.8655700045124503,
+      "grad_norm": 1.0513677658768643,
+      "learning_rate": 4.6680904224269797e-07,
+      "loss": 0.3084,
+      "step": 10550
+    },
+    {
+      "epoch": 0.865652049062641,
+      "grad_norm": 0.7167034793927923,
+      "learning_rate": 4.6624860928255576e-07,
+      "loss": 0.3034,
+      "step": 10551
+    },
+    {
+      "epoch": 0.8657340936128317,
+      "grad_norm": 0.7631000666520177,
+      "learning_rate": 4.6568849648865323e-07,
+      "loss": 0.3209,
+      "step": 10552
+    },
+    {
+      "epoch": 0.8658161381630225,
+      "grad_norm": 0.7019222147190464,
+      "learning_rate": 4.651287039005431e-07,
+      "loss": 0.2922,
+      "step": 10553
+    },
+    {
+      "epoch": 0.8658981827132133,
+      "grad_norm": 0.9611881331111695,
+      "learning_rate": 4.6456923155775934e-07,
+      "loss": 0.2878,
+      "step": 10554
+    },
+    {
+      "epoch": 0.8659802272634041,
+      "grad_norm": 0.6771159758782442,
+      "learning_rate": 4.640100794998087e-07,
+      "loss": 0.2824,
+      "step": 10555
+    },
+    {
+      "epoch": 0.8660622718135947,
+      "grad_norm": 0.7251380706202112,
+      "learning_rate": 4.6345124776617847e-07,
+      "loss": 0.2724,
+      "step": 10556
+    },
+    {
+      "epoch": 0.8661443163637855,
+      "grad_norm": 0.726816715949741,
+      "learning_rate": 4.628927363963337e-07,
+      "loss": 0.3169,
+      "step": 10557
+    },
+    {
+      "epoch": 0.8662263609139763,
+      "grad_norm": 0.6412088339476956,
+      "learning_rate": 4.623345454297135e-07,
+      "loss": 0.2807,
+      "step": 10558
+    },
+    {
+      "epoch": 0.8663084054641671,
+      "grad_norm": 0.6671857490547741,
+      "learning_rate": 4.617766749057384e-07,
+      "loss": 0.2518,
+      "step": 10559
+    },
+    {
+      "epoch": 0.8663904500143578,
+      "grad_norm": 0.8439287938511166,
+      "learning_rate": 4.61219124863802e-07,
+      "loss": 0.2681,
+      "step": 10560
+    },
+    {
+      "epoch": 0.8664724945645486,
+      "grad_norm": 0.6899404171727658,
+      "learning_rate": 4.606618953432795e-07,
+      "loss": 0.2994,
+      "step": 10561
+    },
+    {
+      "epoch": 0.8665545391147393,
+      "grad_norm": 0.7152171473913713,
+      "learning_rate": 4.601049863835216e-07,
+      "loss": 0.2993,
+      "step": 10562
+    },
+    {
+      "epoch": 0.86663658366493,
+      "grad_norm": 0.6949952352821602,
+      "learning_rate": 4.5954839802385575e-07,
+      "loss": 0.2816,
+      "step": 10563
+    },
+    {
+      "epoch": 0.8667186282151208,
+      "grad_norm": 0.6881221182313383,
+      "learning_rate": 4.5899213030358657e-07,
+      "loss": 0.2801,
+      "step": 10564
+    },
+    {
+      "epoch": 0.8668006727653116,
+      "grad_norm": 0.7326630980514716,
+      "learning_rate": 4.5843618326199713e-07,
+      "loss": 0.2714,
+      "step": 10565
+    },
+    {
+      "epoch": 0.8668827173155024,
+      "grad_norm": 0.7083870209158682,
+      "learning_rate": 4.578805569383482e-07,
+      "loss": 0.3231,
+      "step": 10566
+    },
+    {
+      "epoch": 0.866964761865693,
+      "grad_norm": 0.7200189939072313,
+      "learning_rate": 4.5732525137187733e-07,
+      "loss": 0.3037,
+      "step": 10567
+    },
+    {
+      "epoch": 0.8670468064158838,
+      "grad_norm": 0.6893692600349182,
+      "learning_rate": 4.567702666017976e-07,
+      "loss": 0.2819,
+      "step": 10568
+    },
+    {
+      "epoch": 0.8671288509660746,
+      "grad_norm": 0.720809896418474,
+      "learning_rate": 4.562156026673031e-07,
+      "loss": 0.2975,
+      "step": 10569
+    },
+    {
+      "epoch": 0.8672108955162653,
+      "grad_norm": 0.6870960957809871,
+      "learning_rate": 4.5566125960756324e-07,
+      "loss": 0.316,
+      "step": 10570
+    },
+    {
+      "epoch": 0.8672929400664561,
+      "grad_norm": 0.7107302822751841,
+      "learning_rate": 4.5510723746172326e-07,
+      "loss": 0.3079,
+      "step": 10571
+    },
+    {
+      "epoch": 0.8673749846166469,
+      "grad_norm": 0.7397025860408882,
+      "learning_rate": 4.545535362689091e-07,
+      "loss": 0.3149,
+      "step": 10572
+    },
+    {
+      "epoch": 0.8674570291668376,
+      "grad_norm": 0.6810261807759497,
+      "learning_rate": 4.5400015606822003e-07,
+      "loss": 0.3236,
+      "step": 10573
+    },
+    {
+      "epoch": 0.8675390737170283,
+      "grad_norm": 0.7161431316830457,
+      "learning_rate": 4.5344709689873703e-07,
+      "loss": 0.2845,
+      "step": 10574
+    },
+    {
+      "epoch": 0.8676211182672191,
+      "grad_norm": 0.7221352003997565,
+      "learning_rate": 4.5289435879951605e-07,
+      "loss": 0.2914,
+      "step": 10575
+    },
+    {
+      "epoch": 0.8677031628174099,
+      "grad_norm": 0.7101817454258936,
+      "learning_rate": 4.523419418095898e-07,
+      "loss": 0.3302,
+      "step": 10576
+    },
+    {
+      "epoch": 0.8677852073676006,
+      "grad_norm": 0.7840770268799913,
+      "learning_rate": 4.5178984596796815e-07,
+      "loss": 0.3252,
+      "step": 10577
+    },
+    {
+      "epoch": 0.8678672519177913,
+      "grad_norm": 0.7050016158176936,
+      "learning_rate": 4.5123807131364164e-07,
+      "loss": 0.2897,
+      "step": 10578
+    },
+    {
+      "epoch": 0.8679492964679821,
+      "grad_norm": 0.6984791569458373,
+      "learning_rate": 4.5068661788557345e-07,
+      "loss": 0.3033,
+      "step": 10579
+    },
+    {
+      "epoch": 0.8680313410181729,
+      "grad_norm": 0.743683896425943,
+      "learning_rate": 4.5013548572270806e-07,
+      "loss": 0.3031,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8681133855683636,
+      "grad_norm": 1.5472108259906454,
+      "learning_rate": 4.495846748639643e-07,
+      "loss": 0.2773,
+      "step": 10581
+    },
+    {
+      "epoch": 0.8681954301185544,
+      "grad_norm": 0.6688203088750657,
+      "learning_rate": 4.4903418534824053e-07,
+      "loss": 0.2793,
+      "step": 10582
+    },
+    {
+      "epoch": 0.8682774746687452,
+      "grad_norm": 1.228571693378322,
+      "learning_rate": 4.4848401721440947e-07,
+      "loss": 0.2884,
+      "step": 10583
+    },
+    {
+      "epoch": 0.8683595192189358,
+      "grad_norm": 0.6936879299512503,
+      "learning_rate": 4.4793417050132513e-07,
+      "loss": 0.3061,
+      "step": 10584
+    },
+    {
+      "epoch": 0.8684415637691266,
+      "grad_norm": 0.7254082983107917,
+      "learning_rate": 4.47384645247817e-07,
+      "loss": 0.2913,
+      "step": 10585
+    },
+    {
+      "epoch": 0.8685236083193174,
+      "grad_norm": 0.724802576672017,
+      "learning_rate": 4.468354414926901e-07,
+      "loss": 0.2865,
+      "step": 10586
+    },
+    {
+      "epoch": 0.8686056528695082,
+      "grad_norm": 0.6828551772827268,
+      "learning_rate": 4.4628655927472896e-07,
+      "loss": 0.2884,
+      "step": 10587
+    },
+    {
+      "epoch": 0.8686876974196989,
+      "grad_norm": 0.6960262516182695,
+      "learning_rate": 4.45737998632696e-07,
+      "loss": 0.2846,
+      "step": 10588
+    },
+    {
+      "epoch": 0.8687697419698897,
+      "grad_norm": 0.7135449035546079,
+      "learning_rate": 4.4518975960532786e-07,
+      "loss": 0.3265,
+      "step": 10589
+    },
+    {
+      "epoch": 0.8688517865200804,
+      "grad_norm": 1.056814135135268,
+      "learning_rate": 4.4464184223134143e-07,
+      "loss": 0.292,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8689338310702711,
+      "grad_norm": 0.6749196516553498,
+      "learning_rate": 4.4409424654942967e-07,
+      "loss": 0.2887,
+      "step": 10591
+    },
+    {
+      "epoch": 0.8690158756204619,
+      "grad_norm": 0.7316441142407271,
+      "learning_rate": 4.4354697259826165e-07,
+      "loss": 0.3343,
+      "step": 10592
+    },
+    {
+      "epoch": 0.8690979201706527,
+      "grad_norm": 0.7654972043902065,
+      "learning_rate": 4.4300002041648695e-07,
+      "loss": 0.3147,
+      "step": 10593
+    },
+    {
+      "epoch": 0.8691799647208435,
+      "grad_norm": 1.7993050146163159,
+      "learning_rate": 4.42453390042728e-07,
+      "loss": 0.2986,
+      "step": 10594
+    },
+    {
+      "epoch": 0.8692620092710341,
+      "grad_norm": 0.7362171802380615,
+      "learning_rate": 4.419070815155896e-07,
+      "loss": 0.2919,
+      "step": 10595
+    },
+    {
+      "epoch": 0.8693440538212249,
+      "grad_norm": 0.6977073269685569,
+      "learning_rate": 4.413610948736491e-07,
+      "loss": 0.2772,
+      "step": 10596
+    },
+    {
+      "epoch": 0.8694260983714157,
+      "grad_norm": 0.6555892124998586,
+      "learning_rate": 4.408154301554646e-07,
+      "loss": 0.3194,
+      "step": 10597
+    },
+    {
+      "epoch": 0.8695081429216064,
+      "grad_norm": 0.6676008948002067,
+      "learning_rate": 4.4027008739956964e-07,
+      "loss": 0.2853,
+      "step": 10598
+    },
+    {
+      "epoch": 0.8695901874717972,
+      "grad_norm": 0.7200263632362686,
+      "learning_rate": 4.397250666444747e-07,
+      "loss": 0.2989,
+      "step": 10599
+    },
+    {
+      "epoch": 0.869672232021988,
+      "grad_norm": 0.8289390322000114,
+      "learning_rate": 4.3918036792866993e-07,
+      "loss": 0.3017,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8697542765721787,
+      "grad_norm": 0.6833222668963299,
+      "learning_rate": 4.3863599129061905e-07,
+      "loss": 0.2589,
+      "step": 10601
+    },
+    {
+      "epoch": 0.8698363211223694,
+      "grad_norm": 0.7155244130616424,
+      "learning_rate": 4.3809193676876584e-07,
+      "loss": 0.2739,
+      "step": 10602
+    },
+    {
+      "epoch": 0.8699183656725602,
+      "grad_norm": 0.7834651468330477,
+      "learning_rate": 4.375482044015322e-07,
+      "loss": 0.2955,
+      "step": 10603
+    },
+    {
+      "epoch": 0.870000410222751,
+      "grad_norm": 0.9084427281564397,
+      "learning_rate": 4.3700479422731424e-07,
+      "loss": 0.3072,
+      "step": 10604
+    },
+    {
+      "epoch": 0.8700824547729417,
+      "grad_norm": 0.663524552238768,
+      "learning_rate": 4.3646170628448504e-07,
+      "loss": 0.2721,
+      "step": 10605
+    },
+    {
+      "epoch": 0.8701644993231324,
+      "grad_norm": 0.6918935414082584,
+      "learning_rate": 4.3591894061140005e-07,
+      "loss": 0.3156,
+      "step": 10606
+    },
+    {
+      "epoch": 0.8702465438733232,
+      "grad_norm": 0.7090698210945209,
+      "learning_rate": 4.353764972463853e-07,
+      "loss": 0.3273,
+      "step": 10607
+    },
+    {
+      "epoch": 0.870328588423514,
+      "grad_norm": 0.7140739468175915,
+      "learning_rate": 4.348343762277496e-07,
+      "loss": 0.2862,
+      "step": 10608
+    },
+    {
+      "epoch": 0.8704106329737047,
+      "grad_norm": 0.6979832360782587,
+      "learning_rate": 4.3429257759377506e-07,
+      "loss": 0.2887,
+      "step": 10609
+    },
+    {
+      "epoch": 0.8704926775238955,
+      "grad_norm": 0.7263228993907205,
+      "learning_rate": 4.3375110138272337e-07,
+      "loss": 0.2967,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8705747220740863,
+      "grad_norm": 0.67710521563214,
+      "learning_rate": 4.332099476328333e-07,
+      "loss": 0.3013,
+      "step": 10611
+    },
+    {
+      "epoch": 0.8706567666242769,
+      "grad_norm": 0.7328070585413515,
+      "learning_rate": 4.326691163823188e-07,
+      "loss": 0.3073,
+      "step": 10612
+    },
+    {
+      "epoch": 0.8707388111744677,
+      "grad_norm": 0.7924673684201058,
+      "learning_rate": 4.3212860766937427e-07,
+      "loss": 0.2947,
+      "step": 10613
+    },
+    {
+      "epoch": 0.8708208557246585,
+      "grad_norm": 0.8114629513791897,
+      "learning_rate": 4.315884215321675e-07,
+      "loss": 0.2962,
+      "step": 10614
+    },
+    {
+      "epoch": 0.8709029002748493,
+      "grad_norm": 0.6754396453354852,
+      "learning_rate": 4.3104855800884694e-07,
+      "loss": 0.2877,
+      "step": 10615
+    },
+    {
+      "epoch": 0.87098494482504,
+      "grad_norm": 0.7284284207042937,
+      "learning_rate": 4.3050901713753755e-07,
+      "loss": 0.2948,
+      "step": 10616
+    },
+    {
+      "epoch": 0.8710669893752307,
+      "grad_norm": 0.7421136730259835,
+      "learning_rate": 4.2996979895633894e-07,
+      "loss": 0.3047,
+      "step": 10617
+    },
+    {
+      "epoch": 0.8711490339254215,
+      "grad_norm": 0.6316575763486557,
+      "learning_rate": 4.294309035033317e-07,
+      "loss": 0.2435,
+      "step": 10618
+    },
+    {
+      "epoch": 0.8712310784756122,
+      "grad_norm": 0.6951348284731078,
+      "learning_rate": 4.28892330816571e-07,
+      "loss": 0.3093,
+      "step": 10619
+    },
+    {
+      "epoch": 0.871313123025803,
+      "grad_norm": 0.6626748250250697,
+      "learning_rate": 4.283540809340886e-07,
+      "loss": 0.2898,
+      "step": 10620
+    },
+    {
+      "epoch": 0.8713951675759938,
+      "grad_norm": 1.1529074973790994,
+      "learning_rate": 4.278161538938974e-07,
+      "loss": 0.3053,
+      "step": 10621
+    },
+    {
+      "epoch": 0.8714772121261846,
+      "grad_norm": 0.6437496492020085,
+      "learning_rate": 4.2727854973398267e-07,
+      "loss": 0.2997,
+      "step": 10622
+    },
+    {
+      "epoch": 0.8715592566763752,
+      "grad_norm": 0.7982568386200748,
+      "learning_rate": 4.2674126849231e-07,
+      "loss": 0.3159,
+      "step": 10623
+    },
+    {
+      "epoch": 0.871641301226566,
+      "grad_norm": 0.7414261453370182,
+      "learning_rate": 4.2620431020682253e-07,
+      "loss": 0.3257,
+      "step": 10624
+    },
+    {
+      "epoch": 0.8717233457767568,
+      "grad_norm": 0.6557920158752574,
+      "learning_rate": 4.2566767491543706e-07,
+      "loss": 0.2789,
+      "step": 10625
+    },
+    {
+      "epoch": 0.8718053903269475,
+      "grad_norm": 0.7003205963533056,
+      "learning_rate": 4.251313626560521e-07,
+      "loss": 0.2913,
+      "step": 10626
+    },
+    {
+      "epoch": 0.8718874348771383,
+      "grad_norm": 0.8025136704571632,
+      "learning_rate": 4.245953734665398e-07,
+      "loss": 0.2778,
+      "step": 10627
+    },
+    {
+      "epoch": 0.871969479427329,
+      "grad_norm": 0.7380933981335916,
+      "learning_rate": 4.240597073847508e-07,
+      "loss": 0.302,
+      "step": 10628
+    },
+    {
+      "epoch": 0.8720515239775198,
+      "grad_norm": 0.6786398302034437,
+      "learning_rate": 4.235243644485149e-07,
+      "loss": 0.2763,
+      "step": 10629
+    },
+    {
+      "epoch": 0.8721335685277105,
+      "grad_norm": 0.7512990840693271,
+      "learning_rate": 4.22989344695634e-07,
+      "loss": 0.2697,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8722156130779013,
+      "grad_norm": 0.7049210528384273,
+      "learning_rate": 4.2245464816389357e-07,
+      "loss": 0.3049,
+      "step": 10631
+    },
+    {
+      "epoch": 0.8722976576280921,
+      "grad_norm": 0.7021090333958492,
+      "learning_rate": 4.2192027489105113e-07,
+      "loss": 0.3026,
+      "step": 10632
+    },
+    {
+      "epoch": 0.8723797021782828,
+      "grad_norm": 0.7142141995932553,
+      "learning_rate": 4.213862249148432e-07,
+      "loss": 0.2896,
+      "step": 10633
+    },
+    {
+      "epoch": 0.8724617467284735,
+      "grad_norm": 0.748923600544931,
+      "learning_rate": 4.2085249827298445e-07,
+      "loss": 0.3189,
+      "step": 10634
+    },
+    {
+      "epoch": 0.8725437912786643,
+      "grad_norm": 0.7348382269439483,
+      "learning_rate": 4.2031909500316493e-07,
+      "loss": 0.2909,
+      "step": 10635
+    },
+    {
+      "epoch": 0.8726258358288551,
+      "grad_norm": 0.8239588125250739,
+      "learning_rate": 4.197860151430533e-07,
+      "loss": 0.3168,
+      "step": 10636
+    },
+    {
+      "epoch": 0.8727078803790458,
+      "grad_norm": 0.6639404600069776,
+      "learning_rate": 4.1925325873029387e-07,
+      "loss": 0.2635,
+      "step": 10637
+    },
+    {
+      "epoch": 0.8727899249292366,
+      "grad_norm": 0.7675647381931301,
+      "learning_rate": 4.1872082580250994e-07,
+      "loss": 0.3381,
+      "step": 10638
+    },
+    {
+      "epoch": 0.8728719694794274,
+      "grad_norm": 0.7166562378061995,
+      "learning_rate": 4.18188716397302e-07,
+      "loss": 0.2895,
+      "step": 10639
+    },
+    {
+      "epoch": 0.872954014029618,
+      "grad_norm": 0.7194992925719045,
+      "learning_rate": 4.1765693055224434e-07,
+      "loss": 0.3045,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8730360585798088,
+      "grad_norm": 0.7060631154904886,
+      "learning_rate": 4.1712546830489253e-07,
+      "loss": 0.2952,
+      "step": 10641
+    },
+    {
+      "epoch": 0.8731181031299996,
+      "grad_norm": 0.6875745875014164,
+      "learning_rate": 4.165943296927777e-07,
+      "loss": 0.2971,
+      "step": 10642
+    },
+    {
+      "epoch": 0.8732001476801904,
+      "grad_norm": 0.7203223504252293,
+      "learning_rate": 4.1606351475340703e-07,
+      "loss": 0.3062,
+      "step": 10643
+    },
+    {
+      "epoch": 0.8732821922303811,
+      "grad_norm": 0.7209818029433145,
+      "learning_rate": 4.1553302352426715e-07,
+      "loss": 0.296,
+      "step": 10644
+    },
+    {
+      "epoch": 0.8733642367805718,
+      "grad_norm": 0.7590210156898725,
+      "learning_rate": 4.1500285604281877e-07,
+      "loss": 0.2522,
+      "step": 10645
+    },
+    {
+      "epoch": 0.8734462813307626,
+      "grad_norm": 0.7440886849404142,
+      "learning_rate": 4.1447301234650297e-07,
+      "loss": 0.3256,
+      "step": 10646
+    },
+    {
+      "epoch": 0.8735283258809533,
+      "grad_norm": 0.6372723465122311,
+      "learning_rate": 4.139434924727359e-07,
+      "loss": 0.2653,
+      "step": 10647
+    },
+    {
+      "epoch": 0.8736103704311441,
+      "grad_norm": 0.6998607932217216,
+      "learning_rate": 4.134142964589105e-07,
+      "loss": 0.3177,
+      "step": 10648
+    },
+    {
+      "epoch": 0.8736924149813349,
+      "grad_norm": 0.6869760696114136,
+      "learning_rate": 4.128854243423991e-07,
+      "loss": 0.2889,
+      "step": 10649
+    },
+    {
+      "epoch": 0.8737744595315257,
+      "grad_norm": 0.7119880325951097,
+      "learning_rate": 4.1235687616054897e-07,
+      "loss": 0.3026,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8738565040817163,
+      "grad_norm": 1.3967181444891934,
+      "learning_rate": 4.1182865195068535e-07,
+      "loss": 0.2956,
+      "step": 10651
+    },
+    {
+      "epoch": 0.8739385486319071,
+      "grad_norm": 0.7178712852753787,
+      "learning_rate": 4.1130075175011163e-07,
+      "loss": 0.2908,
+      "step": 10652
+    },
+    {
+      "epoch": 0.8740205931820979,
+      "grad_norm": 0.7067391923393348,
+      "learning_rate": 4.1077317559610584e-07,
+      "loss": 0.2826,
+      "step": 10653
+    },
+    {
+      "epoch": 0.8741026377322886,
+      "grad_norm": 0.6614099364316512,
+      "learning_rate": 4.10245923525926e-07,
+      "loss": 0.2823,
+      "step": 10654
+    },
+    {
+      "epoch": 0.8741846822824794,
+      "grad_norm": 0.7000989753498861,
+      "learning_rate": 4.097189955768038e-07,
+      "loss": 0.2903,
+      "step": 10655
+    },
+    {
+      "epoch": 0.8742667268326701,
+      "grad_norm": 0.6334926866750158,
+      "learning_rate": 4.0919239178595194e-07,
+      "loss": 0.2663,
+      "step": 10656
+    },
+    {
+      "epoch": 0.8743487713828609,
+      "grad_norm": 0.667058669409739,
+      "learning_rate": 4.0866611219055827e-07,
+      "loss": 0.2528,
+      "step": 10657
+    },
+    {
+      "epoch": 0.8744308159330516,
+      "grad_norm": 0.7234879226126119,
+      "learning_rate": 4.081401568277865e-07,
+      "loss": 0.3028,
+      "step": 10658
+    },
+    {
+      "epoch": 0.8745128604832424,
+      "grad_norm": 0.7869074609457788,
+      "learning_rate": 4.076145257347802e-07,
+      "loss": 0.3001,
+      "step": 10659
+    },
+    {
+      "epoch": 0.8745949050334332,
+      "grad_norm": 0.7341273122882264,
+      "learning_rate": 4.0708921894865747e-07,
+      "loss": 0.3043,
+      "step": 10660
+    },
+    {
+      "epoch": 0.8746769495836239,
+      "grad_norm": 0.6914212212415541,
+      "learning_rate": 4.0656423650651477e-07,
+      "loss": 0.3301,
+      "step": 10661
+    },
+    {
+      "epoch": 0.8747589941338146,
+      "grad_norm": 0.7247265731778423,
+      "learning_rate": 4.060395784454263e-07,
+      "loss": 0.3299,
+      "step": 10662
+    },
+    {
+      "epoch": 0.8748410386840054,
+      "grad_norm": 2.2362299951657123,
+      "learning_rate": 4.055152448024413e-07,
+      "loss": 0.3124,
+      "step": 10663
+    },
+    {
+      "epoch": 0.8749230832341962,
+      "grad_norm": 0.7219909891989075,
+      "learning_rate": 4.0499123561458795e-07,
+      "loss": 0.2703,
+      "step": 10664
+    },
+    {
+      "epoch": 0.8750051277843869,
+      "grad_norm": 0.7583775927541824,
+      "learning_rate": 4.044675509188722e-07,
+      "loss": 0.2915,
+      "step": 10665
+    },
+    {
+      "epoch": 0.8750871723345777,
+      "grad_norm": 0.7759714110303447,
+      "learning_rate": 4.03944190752274e-07,
+      "loss": 0.2964,
+      "step": 10666
+    },
+    {
+      "epoch": 0.8751692168847685,
+      "grad_norm": 0.7176308524455763,
+      "learning_rate": 4.0342115515175363e-07,
+      "loss": 0.2844,
+      "step": 10667
+    },
+    {
+      "epoch": 0.8752512614349592,
+      "grad_norm": 0.8027920439018081,
+      "learning_rate": 4.0289844415424497e-07,
+      "loss": 0.3159,
+      "step": 10668
+    },
+    {
+      "epoch": 0.8753333059851499,
+      "grad_norm": 0.6576497989066044,
+      "learning_rate": 4.023760577966629e-07,
+      "loss": 0.2604,
+      "step": 10669
+    },
+    {
+      "epoch": 0.8754153505353407,
+      "grad_norm": 0.8224573565194759,
+      "learning_rate": 4.0185399611589794e-07,
+      "loss": 0.2712,
+      "step": 10670
+    },
+    {
+      "epoch": 0.8754973950855315,
+      "grad_norm": 0.6502921676583979,
+      "learning_rate": 4.01332259148815e-07,
+      "loss": 0.3019,
+      "step": 10671
+    },
+    {
+      "epoch": 0.8755794396357222,
+      "grad_norm": 0.7263592227251334,
+      "learning_rate": 4.008108469322608e-07,
+      "loss": 0.2785,
+      "step": 10672
+    },
+    {
+      "epoch": 0.8756614841859129,
+      "grad_norm": 0.7668094858550537,
+      "learning_rate": 4.0028975950305416e-07,
+      "loss": 0.2694,
+      "step": 10673
+    },
+    {
+      "epoch": 0.8757435287361037,
+      "grad_norm": 0.7435504728344147,
+      "learning_rate": 3.997689968979962e-07,
+      "loss": 0.2815,
+      "step": 10674
+    },
+    {
+      "epoch": 0.8758255732862945,
+      "grad_norm": 0.7305170140901994,
+      "learning_rate": 3.992485591538603e-07,
+      "loss": 0.3246,
+      "step": 10675
+    },
+    {
+      "epoch": 0.8759076178364852,
+      "grad_norm": 0.6870790677421496,
+      "learning_rate": 3.987284463073987e-07,
+      "loss": 0.2947,
+      "step": 10676
+    },
+    {
+      "epoch": 0.875989662386676,
+      "grad_norm": 0.6804007488173737,
+      "learning_rate": 3.9820865839534206e-07,
+      "loss": 0.2932,
+      "step": 10677
+    },
+    {
+      "epoch": 0.8760717069368668,
+      "grad_norm": 0.8290679873803678,
+      "learning_rate": 3.976891954543971e-07,
+      "loss": 0.3088,
+      "step": 10678
+    },
+    {
+      "epoch": 0.8761537514870574,
+      "grad_norm": 0.6807698564904336,
+      "learning_rate": 3.971700575212462e-07,
+      "loss": 0.2991,
+      "step": 10679
+    },
+    {
+      "epoch": 0.8762357960372482,
+      "grad_norm": 0.6788653068082897,
+      "learning_rate": 3.966512446325521e-07,
+      "loss": 0.3092,
+      "step": 10680
+    },
+    {
+      "epoch": 0.876317840587439,
+      "grad_norm": 0.7734537263480676,
+      "learning_rate": 3.9613275682494956e-07,
+      "loss": 0.3056,
+      "step": 10681
+    },
+    {
+      "epoch": 0.8763998851376298,
+      "grad_norm": 0.686923600301431,
+      "learning_rate": 3.956145941350559e-07,
+      "loss": 0.3364,
+      "step": 10682
+    },
+    {
+      "epoch": 0.8764819296878205,
+      "grad_norm": 0.6605550307942503,
+      "learning_rate": 3.9509675659946234e-07,
+      "loss": 0.311,
+      "step": 10683
+    },
+    {
+      "epoch": 0.8765639742380112,
+      "grad_norm": 0.7246043584538278,
+      "learning_rate": 3.9457924425473694e-07,
+      "loss": 0.2954,
+      "step": 10684
+    },
+    {
+      "epoch": 0.876646018788202,
+      "grad_norm": 0.6823903288313646,
+      "learning_rate": 3.940620571374271e-07,
+      "loss": 0.3182,
+      "step": 10685
+    },
+    {
+      "epoch": 0.8767280633383927,
+      "grad_norm": 0.6908398203335342,
+      "learning_rate": 3.9354519528405357e-07,
+      "loss": 0.3079,
+      "step": 10686
+    },
+    {
+      "epoch": 0.8768101078885835,
+      "grad_norm": 0.7915320044742858,
+      "learning_rate": 3.9302865873111884e-07,
+      "loss": 0.3612,
+      "step": 10687
+    },
+    {
+      "epoch": 0.8768921524387743,
+      "grad_norm": 0.7123590176082698,
+      "learning_rate": 3.925124475150982e-07,
+      "loss": 0.3011,
+      "step": 10688
+    },
+    {
+      "epoch": 0.8769741969889651,
+      "grad_norm": 0.6846521126708313,
+      "learning_rate": 3.9199656167244525e-07,
+      "loss": 0.2757,
+      "step": 10689
+    },
+    {
+      "epoch": 0.8770562415391557,
+      "grad_norm": 0.7301866643949786,
+      "learning_rate": 3.914810012395914e-07,
+      "loss": 0.2789,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8771382860893465,
+      "grad_norm": 0.6826605498046311,
+      "learning_rate": 3.909657662529465e-07,
+      "loss": 0.2961,
+      "step": 10691
+    },
+    {
+      "epoch": 0.8772203306395373,
+      "grad_norm": 0.6870922636630897,
+      "learning_rate": 3.904508567488935e-07,
+      "loss": 0.2722,
+      "step": 10692
+    },
+    {
+      "epoch": 0.877302375189728,
+      "grad_norm": 0.7247235565707109,
+      "learning_rate": 3.8993627276379564e-07,
+      "loss": 0.3161,
+      "step": 10693
+    },
+    {
+      "epoch": 0.8773844197399188,
+      "grad_norm": 0.6809995361048672,
+      "learning_rate": 3.894220143339905e-07,
+      "loss": 0.2818,
+      "step": 10694
+    },
+    {
+      "epoch": 0.8774664642901095,
+      "grad_norm": 0.6936624301387705,
+      "learning_rate": 3.8890808149579573e-07,
+      "loss": 0.3349,
+      "step": 10695
+    },
+    {
+      "epoch": 0.8775485088403003,
+      "grad_norm": 0.694886775239229,
+      "learning_rate": 3.883944742855045e-07,
+      "loss": 0.2882,
+      "step": 10696
+    },
+    {
+      "epoch": 0.877630553390491,
+      "grad_norm": 0.6652252252059668,
+      "learning_rate": 3.8788119273938606e-07,
+      "loss": 0.285,
+      "step": 10697
+    },
+    {
+      "epoch": 0.8777125979406818,
+      "grad_norm": 0.7631551902480868,
+      "learning_rate": 3.8736823689368817e-07,
+      "loss": 0.2916,
+      "step": 10698
+    },
+    {
+      "epoch": 0.8777946424908726,
+      "grad_norm": 0.9621070572959409,
+      "learning_rate": 3.8685560678463405e-07,
+      "loss": 0.2804,
+      "step": 10699
+    },
+    {
+      "epoch": 0.8778766870410633,
+      "grad_norm": 0.8423089197440218,
+      "learning_rate": 3.863433024484259e-07,
+      "loss": 0.279,
+      "step": 10700
+    },
+    {
+      "epoch": 0.877958731591254,
+      "grad_norm": 0.7400695924889888,
+      "learning_rate": 3.858313239212408e-07,
+      "loss": 0.2649,
+      "step": 10701
+    },
+    {
+      "epoch": 0.8780407761414448,
+      "grad_norm": 0.6973139319185422,
+      "learning_rate": 3.8531967123923486e-07,
+      "loss": 0.2968,
+      "step": 10702
+    },
+    {
+      "epoch": 0.8781228206916356,
+      "grad_norm": 0.7302795208692383,
+      "learning_rate": 3.8480834443854034e-07,
+      "loss": 0.3061,
+      "step": 10703
+    },
+    {
+      "epoch": 0.8782048652418263,
+      "grad_norm": 1.0640676297326044,
+      "learning_rate": 3.8429734355526437e-07,
+      "loss": 0.3082,
+      "step": 10704
+    },
+    {
+      "epoch": 0.8782869097920171,
+      "grad_norm": 0.6836291557835369,
+      "learning_rate": 3.837866686254943e-07,
+      "loss": 0.309,
+      "step": 10705
+    },
+    {
+      "epoch": 0.8783689543422079,
+      "grad_norm": 0.6966795541215912,
+      "learning_rate": 3.83276319685294e-07,
+      "loss": 0.3026,
+      "step": 10706
+    },
+    {
+      "epoch": 0.8784509988923985,
+      "grad_norm": 1.011743993973967,
+      "learning_rate": 3.827662967707019e-07,
+      "loss": 0.3713,
+      "step": 10707
+    },
+    {
+      "epoch": 0.8785330434425893,
+      "grad_norm": 0.7028130329346055,
+      "learning_rate": 3.822565999177358e-07,
+      "loss": 0.313,
+      "step": 10708
+    },
+    {
+      "epoch": 0.8786150879927801,
+      "grad_norm": 0.736021558052495,
+      "learning_rate": 3.817472291623908e-07,
+      "loss": 0.2755,
+      "step": 10709
+    },
+    {
+      "epoch": 0.8786971325429709,
+      "grad_norm": 0.6604024487802794,
+      "learning_rate": 3.812381845406354e-07,
+      "loss": 0.2604,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8787791770931616,
+      "grad_norm": 0.7207801358365983,
+      "learning_rate": 3.807294660884198e-07,
+      "loss": 0.3089,
+      "step": 10711
+    },
+    {
+      "epoch": 0.8788612216433523,
+      "grad_norm": 0.7794356061790533,
+      "learning_rate": 3.8022107384166676e-07,
+      "loss": 0.2759,
+      "step": 10712
+    },
+    {
+      "epoch": 0.8789432661935431,
+      "grad_norm": 0.6653694485352035,
+      "learning_rate": 3.7971300783627987e-07,
+      "loss": 0.2913,
+      "step": 10713
+    },
+    {
+      "epoch": 0.8790253107437338,
+      "grad_norm": 0.8153220995742125,
+      "learning_rate": 3.7920526810813763e-07,
+      "loss": 0.2942,
+      "step": 10714
+    },
+    {
+      "epoch": 0.8791073552939246,
+      "grad_norm": 0.7063440013487704,
+      "learning_rate": 3.7869785469309526e-07,
+      "loss": 0.3034,
+      "step": 10715
+    },
+    {
+      "epoch": 0.8791893998441154,
+      "grad_norm": 0.7301637231725094,
+      "learning_rate": 3.781907676269864e-07,
+      "loss": 0.3243,
+      "step": 10716
+    },
+    {
+      "epoch": 0.8792714443943062,
+      "grad_norm": 0.7150024263217332,
+      "learning_rate": 3.776840069456189e-07,
+      "loss": 0.3103,
+      "step": 10717
+    },
+    {
+      "epoch": 0.8793534889444968,
+      "grad_norm": 0.6970414625109904,
+      "learning_rate": 3.7717757268478037e-07,
+      "loss": 0.2607,
+      "step": 10718
+    },
+    {
+      "epoch": 0.8794355334946876,
+      "grad_norm": 0.7893684246532853,
+      "learning_rate": 3.7667146488023554e-07,
+      "loss": 0.3136,
+      "step": 10719
+    },
+    {
+      "epoch": 0.8795175780448784,
+      "grad_norm": 0.8530051678659988,
+      "learning_rate": 3.76165683567723e-07,
+      "loss": 0.2747,
+      "step": 10720
+    },
+    {
+      "epoch": 0.8795996225950691,
+      "grad_norm": 0.6857964361731975,
+      "learning_rate": 3.7566022878296194e-07,
+      "loss": 0.3298,
+      "step": 10721
+    },
+    {
+      "epoch": 0.8796816671452599,
+      "grad_norm": 0.7261320442303525,
+      "learning_rate": 3.7515510056164493e-07,
+      "loss": 0.292,
+      "step": 10722
+    },
+    {
+      "epoch": 0.8797637116954506,
+      "grad_norm": 0.7336511778846666,
+      "learning_rate": 3.746502989394446e-07,
+      "loss": 0.2958,
+      "step": 10723
+    },
+    {
+      "epoch": 0.8798457562456414,
+      "grad_norm": 1.0639408803500643,
+      "learning_rate": 3.741458239520096e-07,
+      "loss": 0.2805,
+      "step": 10724
+    },
+    {
+      "epoch": 0.8799278007958321,
+      "grad_norm": 0.6847913795603848,
+      "learning_rate": 3.736416756349642e-07,
+      "loss": 0.3131,
+      "step": 10725
+    },
+    {
+      "epoch": 0.8800098453460229,
+      "grad_norm": 0.8334359429853826,
+      "learning_rate": 3.7313785402391054e-07,
+      "loss": 0.3199,
+      "step": 10726
+    },
+    {
+      "epoch": 0.8800918898962137,
+      "grad_norm": 0.8548442373018286,
+      "learning_rate": 3.7263435915442846e-07,
+      "loss": 0.2781,
+      "step": 10727
+    },
+    {
+      "epoch": 0.8801739344464043,
+      "grad_norm": 0.8357344627894595,
+      "learning_rate": 3.72131191062074e-07,
+      "loss": 0.2887,
+      "step": 10728
+    },
+    {
+      "epoch": 0.8802559789965951,
+      "grad_norm": 0.7295745280643023,
+      "learning_rate": 3.716283497823786e-07,
+      "loss": 0.3509,
+      "step": 10729
+    },
+    {
+      "epoch": 0.8803380235467859,
+      "grad_norm": 0.9860546020344879,
+      "learning_rate": 3.71125835350854e-07,
+      "loss": 0.3097,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8804200680969767,
+      "grad_norm": 0.6936643963418342,
+      "learning_rate": 3.706236478029851e-07,
+      "loss": 0.3319,
+      "step": 10731
+    },
+    {
+      "epoch": 0.8805021126471674,
+      "grad_norm": 0.7390591493959974,
+      "learning_rate": 3.7012178717423744e-07,
+      "loss": 0.3393,
+      "step": 10732
+    },
+    {
+      "epoch": 0.8805841571973582,
+      "grad_norm": 0.7715508560890467,
+      "learning_rate": 3.6962025350004983e-07,
+      "loss": 0.3297,
+      "step": 10733
+    },
+    {
+      "epoch": 0.880666201747549,
+      "grad_norm": 0.7429980174607828,
+      "learning_rate": 3.691190468158412e-07,
+      "loss": 0.2773,
+      "step": 10734
+    },
+    {
+      "epoch": 0.8807482462977396,
+      "grad_norm": 0.6973019725490931,
+      "learning_rate": 3.686181671570049e-07,
+      "loss": 0.2796,
+      "step": 10735
+    },
+    {
+      "epoch": 0.8808302908479304,
+      "grad_norm": 0.7462494354335634,
+      "learning_rate": 3.681176145589127e-07,
+      "loss": 0.2818,
+      "step": 10736
+    },
+    {
+      "epoch": 0.8809123353981212,
+      "grad_norm": 0.6786958739291334,
+      "learning_rate": 3.676173890569135e-07,
+      "loss": 0.2759,
+      "step": 10737
+    },
+    {
+      "epoch": 0.880994379948312,
+      "grad_norm": 0.7345476786957887,
+      "learning_rate": 3.671174906863312e-07,
+      "loss": 0.2502,
+      "step": 10738
+    },
+    {
+      "epoch": 0.8810764244985027,
+      "grad_norm": 1.0668949393998937,
+      "learning_rate": 3.666179194824693e-07,
+      "loss": 0.2828,
+      "step": 10739
+    },
+    {
+      "epoch": 0.8811584690486934,
+      "grad_norm": 0.6721842476438258,
+      "learning_rate": 3.661186754806051e-07,
+      "loss": 0.3308,
+      "step": 10740
+    },
+    {
+      "epoch": 0.8812405135988842,
+      "grad_norm": 0.7577266404010189,
+      "learning_rate": 3.6561975871599487e-07,
+      "loss": 0.2771,
+      "step": 10741
+    },
+    {
+      "epoch": 0.8813225581490749,
+      "grad_norm": 0.690618449153438,
+      "learning_rate": 3.6512116922387264e-07,
+      "loss": 0.3157,
+      "step": 10742
+    },
+    {
+      "epoch": 0.8814046026992657,
+      "grad_norm": 0.7361288135796112,
+      "learning_rate": 3.646229070394464e-07,
+      "loss": 0.3076,
+      "step": 10743
+    },
+    {
+      "epoch": 0.8814866472494565,
+      "grad_norm": 0.6923240710525015,
+      "learning_rate": 3.641249721979029e-07,
+      "loss": 0.3274,
+      "step": 10744
+    },
+    {
+      "epoch": 0.8815686917996473,
+      "grad_norm": 0.7332616816696536,
+      "learning_rate": 3.636273647344063e-07,
+      "loss": 0.2959,
+      "step": 10745
+    },
+    {
+      "epoch": 0.8816507363498379,
+      "grad_norm": 0.7502622705181693,
+      "learning_rate": 3.631300846840957e-07,
+      "loss": 0.318,
+      "step": 10746
+    },
+    {
+      "epoch": 0.8817327809000287,
+      "grad_norm": 0.7269814144457117,
+      "learning_rate": 3.626331320820897e-07,
+      "loss": 0.3042,
+      "step": 10747
+    },
+    {
+      "epoch": 0.8818148254502195,
+      "grad_norm": 0.6895535375539081,
+      "learning_rate": 3.6213650696348025e-07,
+      "loss": 0.3479,
+      "step": 10748
+    },
+    {
+      "epoch": 0.8818968700004102,
+      "grad_norm": 0.6874793263742873,
+      "learning_rate": 3.616402093633398e-07,
+      "loss": 0.304,
+      "step": 10749
+    },
+    {
+      "epoch": 0.881978914550601,
+      "grad_norm": 0.6883945431464021,
+      "learning_rate": 3.611442393167158e-07,
+      "loss": 0.305,
+      "step": 10750
+    },
+    {
+      "epoch": 0.8820609591007917,
+      "grad_norm": 0.9562056490178452,
+      "learning_rate": 3.606485968586326e-07,
+      "loss": 0.2877,
+      "step": 10751
+    },
+    {
+      "epoch": 0.8821430036509825,
+      "grad_norm": 0.8587805072096744,
+      "learning_rate": 3.6015328202409216e-07,
+      "loss": 0.3039,
+      "step": 10752
+    },
+    {
+      "epoch": 0.8822250482011732,
+      "grad_norm": 0.7115586194417716,
+      "learning_rate": 3.5965829484807203e-07,
+      "loss": 0.3048,
+      "step": 10753
+    },
+    {
+      "epoch": 0.882307092751364,
+      "grad_norm": 0.7637356696278053,
+      "learning_rate": 3.5916363536552754e-07,
+      "loss": 0.303,
+      "step": 10754
+    },
+    {
+      "epoch": 0.8823891373015548,
+      "grad_norm": 0.6947611021865987,
+      "learning_rate": 3.5866930361139196e-07,
+      "loss": 0.2714,
+      "step": 10755
+    },
+    {
+      "epoch": 0.8824711818517454,
+      "grad_norm": 0.6537068901520802,
+      "learning_rate": 3.581752996205734e-07,
+      "loss": 0.3006,
+      "step": 10756
+    },
+    {
+      "epoch": 0.8825532264019362,
+      "grad_norm": 0.7451305678572027,
+      "learning_rate": 3.5768162342795675e-07,
+      "loss": 0.2762,
+      "step": 10757
+    },
+    {
+      "epoch": 0.882635270952127,
+      "grad_norm": 0.737356940450702,
+      "learning_rate": 3.571882750684058e-07,
+      "loss": 0.3037,
+      "step": 10758
+    },
+    {
+      "epoch": 0.8827173155023178,
+      "grad_norm": 0.6808656778979371,
+      "learning_rate": 3.5669525457675935e-07,
+      "loss": 0.2877,
+      "step": 10759
+    },
+    {
+      "epoch": 0.8827993600525085,
+      "grad_norm": 0.6268649540584085,
+      "learning_rate": 3.5620256198783454e-07,
+      "loss": 0.2792,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8828814046026993,
+      "grad_norm": 0.621480331346605,
+      "learning_rate": 3.55710197336423e-07,
+      "loss": 0.2849,
+      "step": 10761
+    },
+    {
+      "epoch": 0.88296344915289,
+      "grad_norm": 0.7585613234337817,
+      "learning_rate": 3.5521816065729574e-07,
+      "loss": 0.2931,
+      "step": 10762
+    },
+    {
+      "epoch": 0.8830454937030807,
+      "grad_norm": 0.738800977325464,
+      "learning_rate": 3.5472645198520064e-07,
+      "loss": 0.2945,
+      "step": 10763
+    },
+    {
+      "epoch": 0.8831275382532715,
+      "grad_norm": 0.7934219732037373,
+      "learning_rate": 3.5423507135485925e-07,
+      "loss": 0.3066,
+      "step": 10764
+    },
+    {
+      "epoch": 0.8832095828034623,
+      "grad_norm": 0.6746253694082947,
+      "learning_rate": 3.537440188009744e-07,
+      "loss": 0.3045,
+      "step": 10765
+    },
+    {
+      "epoch": 0.8832916273536531,
+      "grad_norm": 0.8015085959015974,
+      "learning_rate": 3.5325329435822066e-07,
+      "loss": 0.2505,
+      "step": 10766
+    },
+    {
+      "epoch": 0.8833736719038437,
+      "grad_norm": 0.685996084563905,
+      "learning_rate": 3.5276289806125406e-07,
+      "loss": 0.2569,
+      "step": 10767
+    },
+    {
+      "epoch": 0.8834557164540345,
+      "grad_norm": 0.6969686687118699,
+      "learning_rate": 3.522728299447059e-07,
+      "loss": 0.3141,
+      "step": 10768
+    },
+    {
+      "epoch": 0.8835377610042253,
+      "grad_norm": 0.9511167012630165,
+      "learning_rate": 3.517830900431829e-07,
+      "loss": 0.2994,
+      "step": 10769
+    },
+    {
+      "epoch": 0.883619805554416,
+      "grad_norm": 0.6500912035148473,
+      "learning_rate": 3.512936783912707e-07,
+      "loss": 0.2906,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8837018501046068,
+      "grad_norm": 0.7456250717301549,
+      "learning_rate": 3.5080459502353003e-07,
+      "loss": 0.3249,
+      "step": 10771
+    },
+    {
+      "epoch": 0.8837838946547976,
+      "grad_norm": 1.1623343120179837,
+      "learning_rate": 3.5031583997449883e-07,
+      "loss": 0.3338,
+      "step": 10772
+    },
+    {
+      "epoch": 0.8838659392049883,
+      "grad_norm": 0.7067606295232163,
+      "learning_rate": 3.498274132786933e-07,
+      "loss": 0.2832,
+      "step": 10773
+    },
+    {
+      "epoch": 0.883947983755179,
+      "grad_norm": 0.7127179051156542,
+      "learning_rate": 3.4933931497060423e-07,
+      "loss": 0.311,
+      "step": 10774
+    },
+    {
+      "epoch": 0.8840300283053698,
+      "grad_norm": 0.7043432131062769,
+      "learning_rate": 3.4885154508470186e-07,
+      "loss": 0.2581,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8841120728555606,
+      "grad_norm": 0.7875058153744926,
+      "learning_rate": 3.4836410365542974e-07,
+      "loss": 0.3328,
+      "step": 10776
+    },
+    {
+      "epoch": 0.8841941174057513,
+      "grad_norm": 0.667966867731291,
+      "learning_rate": 3.478769907172108e-07,
+      "loss": 0.3069,
+      "step": 10777
+    },
+    {
+      "epoch": 0.884276161955942,
+      "grad_norm": 0.7013046435438002,
+      "learning_rate": 3.47390206304446e-07,
+      "loss": 0.2775,
+      "step": 10778
+    },
+    {
+      "epoch": 0.8843582065061328,
+      "grad_norm": 0.70448123844531,
+      "learning_rate": 3.469037504515083e-07,
+      "loss": 0.2966,
+      "step": 10779
+    },
+    {
+      "epoch": 0.8844402510563236,
+      "grad_norm": 0.7737390272255159,
+      "learning_rate": 3.464176231927524e-07,
+      "loss": 0.2978,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8845222956065143,
+      "grad_norm": 0.7429176697699668,
+      "learning_rate": 3.459318245625082e-07,
+      "loss": 0.2495,
+      "step": 10781
+    },
+    {
+      "epoch": 0.8846043401567051,
+      "grad_norm": 0.7027949863914136,
+      "learning_rate": 3.4544635459508037e-07,
+      "loss": 0.3106,
+      "step": 10782
+    },
+    {
+      "epoch": 0.8846863847068959,
+      "grad_norm": 0.724606518913892,
+      "learning_rate": 3.4496121332475375e-07,
+      "loss": 0.2674,
+      "step": 10783
+    },
+    {
+      "epoch": 0.8847684292570867,
+      "grad_norm": 0.6511675759282761,
+      "learning_rate": 3.444764007857876e-07,
+      "loss": 0.2879,
+      "step": 10784
+    },
+    {
+      "epoch": 0.8848504738072773,
+      "grad_norm": 0.7180564514628679,
+      "learning_rate": 3.439919170124173e-07,
+      "loss": 0.2952,
+      "step": 10785
+    },
+    {
+      "epoch": 0.8849325183574681,
+      "grad_norm": 0.8639292665906375,
+      "learning_rate": 3.435077620388577e-07,
+      "loss": 0.2665,
+      "step": 10786
+    },
+    {
+      "epoch": 0.8850145629076589,
+      "grad_norm": 0.6946270690230265,
+      "learning_rate": 3.4302393589929874e-07,
+      "loss": 0.2915,
+      "step": 10787
+    },
+    {
+      "epoch": 0.8850966074578496,
+      "grad_norm": 0.721429132461902,
+      "learning_rate": 3.425404386279074e-07,
+      "loss": 0.2817,
+      "step": 10788
+    },
+    {
+      "epoch": 0.8851786520080404,
+      "grad_norm": 0.6974236888033727,
+      "learning_rate": 3.420572702588276e-07,
+      "loss": 0.262,
+      "step": 10789
+    },
+    {
+      "epoch": 0.8852606965582311,
+      "grad_norm": 0.6782112756880199,
+      "learning_rate": 3.4157443082617924e-07,
+      "loss": 0.2933,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8853427411084219,
+      "grad_norm": 0.7301630418897865,
+      "learning_rate": 3.4109192036406114e-07,
+      "loss": 0.3178,
+      "step": 10791
+    },
+    {
+      "epoch": 0.8854247856586126,
+      "grad_norm": 0.7478433319223032,
+      "learning_rate": 3.4060973890654546e-07,
+      "loss": 0.322,
+      "step": 10792
+    },
+    {
+      "epoch": 0.8855068302088034,
+      "grad_norm": 0.730328110149417,
+      "learning_rate": 3.401278864876856e-07,
+      "loss": 0.3334,
+      "step": 10793
+    },
+    {
+      "epoch": 0.8855888747589942,
+      "grad_norm": 0.9220129310455696,
+      "learning_rate": 3.396463631415065e-07,
+      "loss": 0.3766,
+      "step": 10794
+    },
+    {
+      "epoch": 0.8856709193091848,
+      "grad_norm": 0.6975834435000785,
+      "learning_rate": 3.391651689020137e-07,
+      "loss": 0.2923,
+      "step": 10795
+    },
+    {
+      "epoch": 0.8857529638593756,
+      "grad_norm": 0.6750498972903078,
+      "learning_rate": 3.386843038031895e-07,
+      "loss": 0.3319,
+      "step": 10796
+    },
+    {
+      "epoch": 0.8858350084095664,
+      "grad_norm": 0.7288450013664576,
+      "learning_rate": 3.3820376787899013e-07,
+      "loss": 0.273,
+      "step": 10797
+    },
+    {
+      "epoch": 0.8859170529597572,
+      "grad_norm": 0.6835775441308786,
+      "learning_rate": 3.3772356116335113e-07,
+      "loss": 0.2655,
+      "step": 10798
+    },
+    {
+      "epoch": 0.8859990975099479,
+      "grad_norm": 0.6762043350106368,
+      "learning_rate": 3.3724368369018435e-07,
+      "loss": 0.2913,
+      "step": 10799
+    },
+    {
+      "epoch": 0.8860811420601387,
+      "grad_norm": 0.7049489005875917,
+      "learning_rate": 3.367641354933765e-07,
+      "loss": 0.2896,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8861631866103294,
+      "grad_norm": 0.7142339764913342,
+      "learning_rate": 3.3628491660679385e-07,
+      "loss": 0.2972,
+      "step": 10801
+    },
+    {
+      "epoch": 0.8862452311605201,
+      "grad_norm": 0.661293627677091,
+      "learning_rate": 3.3580602706427654e-07,
+      "loss": 0.3084,
+      "step": 10802
+    },
+    {
+      "epoch": 0.8863272757107109,
+      "grad_norm": 0.6734843865113084,
+      "learning_rate": 3.353274668996442e-07,
+      "loss": 0.2907,
+      "step": 10803
+    },
+    {
+      "epoch": 0.8864093202609017,
+      "grad_norm": 0.7657539653435671,
+      "learning_rate": 3.348492361466932e-07,
+      "loss": 0.2917,
+      "step": 10804
+    },
+    {
+      "epoch": 0.8864913648110925,
+      "grad_norm": 0.7203982782120762,
+      "learning_rate": 3.343713348391925e-07,
+      "loss": 0.271,
+      "step": 10805
+    },
+    {
+      "epoch": 0.8865734093612831,
+      "grad_norm": 0.6871958868198255,
+      "learning_rate": 3.338937630108929e-07,
+      "loss": 0.2812,
+      "step": 10806
+    },
+    {
+      "epoch": 0.8866554539114739,
+      "grad_norm": 0.6835127584938527,
+      "learning_rate": 3.3341652069551866e-07,
+      "loss": 0.3037,
+      "step": 10807
+    },
+    {
+      "epoch": 0.8867374984616647,
+      "grad_norm": 0.7147492660687821,
+      "learning_rate": 3.329396079267727e-07,
+      "loss": 0.278,
+      "step": 10808
+    },
+    {
+      "epoch": 0.8868195430118554,
+      "grad_norm": 0.7048103901016645,
+      "learning_rate": 3.324630247383337e-07,
+      "loss": 0.2995,
+      "step": 10809
+    },
+    {
+      "epoch": 0.8869015875620462,
+      "grad_norm": 0.6910702457580282,
+      "learning_rate": 3.319867711638558e-07,
+      "loss": 0.3434,
+      "step": 10810
+    },
+    {
+      "epoch": 0.886983632112237,
+      "grad_norm": 0.6647738203091207,
+      "learning_rate": 3.315108472369738e-07,
+      "loss": 0.2917,
+      "step": 10811
+    },
+    {
+      "epoch": 0.8870656766624277,
+      "grad_norm": 1.1548288157474993,
+      "learning_rate": 3.310352529912952e-07,
+      "loss": 0.3218,
+      "step": 10812
+    },
+    {
+      "epoch": 0.8871477212126184,
+      "grad_norm": 0.7032198116176429,
+      "learning_rate": 3.3055998846040436e-07,
+      "loss": 0.3267,
+      "step": 10813
+    },
+    {
+      "epoch": 0.8872297657628092,
+      "grad_norm": 0.6764647008162368,
+      "learning_rate": 3.3008505367786656e-07,
+      "loss": 0.3198,
+      "step": 10814
+    },
+    {
+      "epoch": 0.887311810313,
+      "grad_norm": 0.804983849648139,
+      "learning_rate": 3.296104486772184e-07,
+      "loss": 0.2765,
+      "step": 10815
+    },
+    {
+      "epoch": 0.8873938548631907,
+      "grad_norm": 0.6618731503917012,
+      "learning_rate": 3.291361734919768e-07,
+      "loss": 0.3047,
+      "step": 10816
+    },
+    {
+      "epoch": 0.8874758994133815,
+      "grad_norm": 0.7289259368006815,
+      "learning_rate": 3.2866222815563565e-07,
+      "loss": 0.2744,
+      "step": 10817
+    },
+    {
+      "epoch": 0.8875579439635722,
+      "grad_norm": 0.7998006761433493,
+      "learning_rate": 3.2818861270166147e-07,
+      "loss": 0.3841,
+      "step": 10818
+    },
+    {
+      "epoch": 0.887639988513763,
+      "grad_norm": 0.7098420379514835,
+      "learning_rate": 3.277153271635025e-07,
+      "loss": 0.3319,
+      "step": 10819
+    },
+    {
+      "epoch": 0.8877220330639537,
+      "grad_norm": 0.681547246443237,
+      "learning_rate": 3.272423715745798e-07,
+      "loss": 0.2845,
+      "step": 10820
+    },
+    {
+      "epoch": 0.8878040776141445,
+      "grad_norm": 0.7385974405659341,
+      "learning_rate": 3.2676974596829337e-07,
+      "loss": 0.2664,
+      "step": 10821
+    },
+    {
+      "epoch": 0.8878861221643353,
+      "grad_norm": 0.7484189541667893,
+      "learning_rate": 3.262974503780203e-07,
+      "loss": 0.3044,
+      "step": 10822
+    },
+    {
+      "epoch": 0.8879681667145259,
+      "grad_norm": 0.7259241535319886,
+      "learning_rate": 3.2582548483711173e-07,
+      "loss": 0.3149,
+      "step": 10823
+    },
+    {
+      "epoch": 0.8880502112647167,
+      "grad_norm": 0.6641245514347613,
+      "learning_rate": 3.2535384937889824e-07,
+      "loss": 0.2885,
+      "step": 10824
+    },
+    {
+      "epoch": 0.8881322558149075,
+      "grad_norm": 0.7989539124930556,
+      "learning_rate": 3.2488254403668484e-07,
+      "loss": 0.3422,
+      "step": 10825
+    },
+    {
+      "epoch": 0.8882143003650983,
+      "grad_norm": 0.7227075894439258,
+      "learning_rate": 3.2441156884375604e-07,
+      "loss": 0.2629,
+      "step": 10826
+    },
+    {
+      "epoch": 0.888296344915289,
+      "grad_norm": 0.7780915202268474,
+      "learning_rate": 3.239409238333702e-07,
+      "loss": 0.3243,
+      "step": 10827
+    },
+    {
+      "epoch": 0.8883783894654798,
+      "grad_norm": 0.7817101946948053,
+      "learning_rate": 3.2347060903876235e-07,
+      "loss": 0.3161,
+      "step": 10828
+    },
+    {
+      "epoch": 0.8884604340156705,
+      "grad_norm": 0.7229107946789612,
+      "learning_rate": 3.230006244931483e-07,
+      "loss": 0.2923,
+      "step": 10829
+    },
+    {
+      "epoch": 0.8885424785658612,
+      "grad_norm": 0.6513375806001469,
+      "learning_rate": 3.225309702297147e-07,
+      "loss": 0.3105,
+      "step": 10830
+    },
+    {
+      "epoch": 0.888624523116052,
+      "grad_norm": 0.7888002315441996,
+      "learning_rate": 3.2206164628162893e-07,
+      "loss": 0.2985,
+      "step": 10831
+    },
+    {
+      "epoch": 0.8887065676662428,
+      "grad_norm": 0.7340768330967434,
+      "learning_rate": 3.215926526820351e-07,
+      "loss": 0.3168,
+      "step": 10832
+    },
+    {
+      "epoch": 0.8887886122164336,
+      "grad_norm": 0.7683702079895828,
+      "learning_rate": 3.211239894640511e-07,
+      "loss": 0.29,
+      "step": 10833
+    },
+    {
+      "epoch": 0.8888706567666242,
+      "grad_norm": 0.6202136575031338,
+      "learning_rate": 3.2065565666077324e-07,
+      "loss": 0.2691,
+      "step": 10834
+    },
+    {
+      "epoch": 0.888952701316815,
+      "grad_norm": 0.7388909356314423,
+      "learning_rate": 3.201876543052762e-07,
+      "loss": 0.3325,
+      "step": 10835
+    },
+    {
+      "epoch": 0.8890347458670058,
+      "grad_norm": 0.718295075180934,
+      "learning_rate": 3.197199824306074e-07,
+      "loss": 0.2956,
+      "step": 10836
+    },
+    {
+      "epoch": 0.8891167904171965,
+      "grad_norm": 0.6917692684898977,
+      "learning_rate": 3.1925264106979493e-07,
+      "loss": 0.2764,
+      "step": 10837
+    },
+    {
+      "epoch": 0.8891988349673873,
+      "grad_norm": 0.8300058216577257,
+      "learning_rate": 3.187856302558395e-07,
+      "loss": 0.2844,
+      "step": 10838
+    },
+    {
+      "epoch": 0.8892808795175781,
+      "grad_norm": 0.6710197629683972,
+      "learning_rate": 3.183189500217232e-07,
+      "loss": 0.3091,
+      "step": 10839
+    },
+    {
+      "epoch": 0.8893629240677688,
+      "grad_norm": 0.6816802181732426,
+      "learning_rate": 3.178526004004007e-07,
+      "loss": 0.2807,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8894449686179595,
+      "grad_norm": 0.6796246200064088,
+      "learning_rate": 3.173865814248045e-07,
+      "loss": 0.3214,
+      "step": 10841
+    },
+    {
+      "epoch": 0.8895270131681503,
+      "grad_norm": 0.6826435758656089,
+      "learning_rate": 3.1692089312784556e-07,
+      "loss": 0.3168,
+      "step": 10842
+    },
+    {
+      "epoch": 0.8896090577183411,
+      "grad_norm": 0.7431158064158394,
+      "learning_rate": 3.1645553554240815e-07,
+      "loss": 0.3163,
+      "step": 10843
+    },
+    {
+      "epoch": 0.8896911022685318,
+      "grad_norm": 0.6796578400321,
+      "learning_rate": 3.159905087013565e-07,
+      "loss": 0.2861,
+      "step": 10844
+    },
+    {
+      "epoch": 0.8897731468187225,
+      "grad_norm": 0.6371220897235447,
+      "learning_rate": 3.155258126375299e-07,
+      "loss": 0.2832,
+      "step": 10845
+    },
+    {
+      "epoch": 0.8898551913689133,
+      "grad_norm": 0.8072115762476729,
+      "learning_rate": 3.150614473837438e-07,
+      "loss": 0.2754,
+      "step": 10846
+    },
+    {
+      "epoch": 0.8899372359191041,
+      "grad_norm": 0.7069170150577162,
+      "learning_rate": 3.1459741297279247e-07,
+      "loss": 0.3087,
+      "step": 10847
+    },
+    {
+      "epoch": 0.8900192804692948,
+      "grad_norm": 0.7477296898236298,
+      "learning_rate": 3.1413370943744305e-07,
+      "loss": 0.28,
+      "step": 10848
+    },
+    {
+      "epoch": 0.8901013250194856,
+      "grad_norm": 0.7056635402775728,
+      "learning_rate": 3.136703368104421e-07,
+      "loss": 0.279,
+      "step": 10849
+    },
+    {
+      "epoch": 0.8901833695696764,
+      "grad_norm": 0.7100749965445801,
+      "learning_rate": 3.1320729512451466e-07,
+      "loss": 0.3128,
+      "step": 10850
+    },
+    {
+      "epoch": 0.890265414119867,
+      "grad_norm": 0.702652039833649,
+      "learning_rate": 3.1274458441235664e-07,
+      "loss": 0.3042,
+      "step": 10851
+    },
+    {
+      "epoch": 0.8903474586700578,
+      "grad_norm": 0.6515513907012971,
+      "learning_rate": 3.122822047066465e-07,
+      "loss": 0.2651,
+      "step": 10852
+    },
+    {
+      "epoch": 0.8904295032202486,
+      "grad_norm": 0.6916139181300489,
+      "learning_rate": 3.118201560400347e-07,
+      "loss": 0.2706,
+      "step": 10853
+    },
+    {
+      "epoch": 0.8905115477704394,
+      "grad_norm": 0.749606295017447,
+      "learning_rate": 3.113584384451518e-07,
+      "loss": 0.318,
+      "step": 10854
+    },
+    {
+      "epoch": 0.8905935923206301,
+      "grad_norm": 0.7513571833940018,
+      "learning_rate": 3.108970519546034e-07,
+      "loss": 0.3122,
+      "step": 10855
+    },
+    {
+      "epoch": 0.8906756368708209,
+      "grad_norm": 0.7729236461561649,
+      "learning_rate": 3.1043599660097125e-07,
+      "loss": 0.298,
+      "step": 10856
+    },
+    {
+      "epoch": 0.8907576814210116,
+      "grad_norm": 0.8043280504603794,
+      "learning_rate": 3.0997527241681425e-07,
+      "loss": 0.2722,
+      "step": 10857
+    },
+    {
+      "epoch": 0.8908397259712023,
+      "grad_norm": 0.7234542783669548,
+      "learning_rate": 3.095148794346692e-07,
+      "loss": 0.3459,
+      "step": 10858
+    },
+    {
+      "epoch": 0.8909217705213931,
+      "grad_norm": 0.6494417818751382,
+      "learning_rate": 3.090548176870467e-07,
+      "loss": 0.2673,
+      "step": 10859
+    },
+    {
+      "epoch": 0.8910038150715839,
+      "grad_norm": 0.923083291231742,
+      "learning_rate": 3.0859508720643693e-07,
+      "loss": 0.2884,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8910858596217747,
+      "grad_norm": 0.7296758404531482,
+      "learning_rate": 3.0813568802530445e-07,
+      "loss": 0.3125,
+      "step": 10861
+    },
+    {
+      "epoch": 0.8911679041719653,
+      "grad_norm": 0.7297117054463833,
+      "learning_rate": 3.0767662017609167e-07,
+      "loss": 0.3145,
+      "step": 10862
+    },
+    {
+      "epoch": 0.8912499487221561,
+      "grad_norm": 0.7537340140946273,
+      "learning_rate": 3.0721788369121765e-07,
+      "loss": 0.3232,
+      "step": 10863
+    },
+    {
+      "epoch": 0.8913319932723469,
+      "grad_norm": 0.6456330037516474,
+      "learning_rate": 3.0675947860307643e-07,
+      "loss": 0.2958,
+      "step": 10864
+    },
+    {
+      "epoch": 0.8914140378225376,
+      "grad_norm": 0.8455813428629453,
+      "learning_rate": 3.0630140494404103e-07,
+      "loss": 0.3221,
+      "step": 10865
+    },
+    {
+      "epoch": 0.8914960823727284,
+      "grad_norm": 0.6894925920450609,
+      "learning_rate": 3.0584366274645947e-07,
+      "loss": 0.2828,
+      "step": 10866
+    },
+    {
+      "epoch": 0.8915781269229192,
+      "grad_norm": 0.6622760221088576,
+      "learning_rate": 3.05386252042657e-07,
+      "loss": 0.3264,
+      "step": 10867
+    },
+    {
+      "epoch": 0.8916601714731099,
+      "grad_norm": 0.7005191868594292,
+      "learning_rate": 3.0492917286493507e-07,
+      "loss": 0.2818,
+      "step": 10868
+    },
+    {
+      "epoch": 0.8917422160233006,
+      "grad_norm": 0.7306751557316571,
+      "learning_rate": 3.044724252455711e-07,
+      "loss": 0.2988,
+      "step": 10869
+    },
+    {
+      "epoch": 0.8918242605734914,
+      "grad_norm": 0.6551756081781487,
+      "learning_rate": 3.0401600921681986e-07,
+      "loss": 0.3026,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8919063051236822,
+      "grad_norm": 0.7067005477370957,
+      "learning_rate": 3.0355992481091503e-07,
+      "loss": 0.2804,
+      "step": 10871
+    },
+    {
+      "epoch": 0.8919883496738729,
+      "grad_norm": 0.6572516276315826,
+      "learning_rate": 3.0310417206006136e-07,
+      "loss": 0.3135,
+      "step": 10872
+    },
+    {
+      "epoch": 0.8920703942240636,
+      "grad_norm": 0.6504505158131968,
+      "learning_rate": 3.0264875099644595e-07,
+      "loss": 0.2807,
+      "step": 10873
+    },
+    {
+      "epoch": 0.8921524387742544,
+      "grad_norm": 0.683768418428627,
+      "learning_rate": 3.021936616522281e-07,
+      "loss": 0.3161,
+      "step": 10874
+    },
+    {
+      "epoch": 0.8922344833244452,
+      "grad_norm": 0.7130306072157914,
+      "learning_rate": 3.017389040595464e-07,
+      "loss": 0.2956,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8923165278746359,
+      "grad_norm": 0.8021579139877233,
+      "learning_rate": 3.0128447825051587e-07,
+      "loss": 0.2744,
+      "step": 10876
+    },
+    {
+      "epoch": 0.8923985724248267,
+      "grad_norm": 0.651657430444367,
+      "learning_rate": 3.008303842572252e-07,
+      "loss": 0.2701,
+      "step": 10877
+    },
+    {
+      "epoch": 0.8924806169750175,
+      "grad_norm": 0.6484546190247283,
+      "learning_rate": 3.0037662211174437e-07,
+      "loss": 0.3164,
+      "step": 10878
+    },
+    {
+      "epoch": 0.8925626615252081,
+      "grad_norm": 0.814262189076426,
+      "learning_rate": 2.9992319184611485e-07,
+      "loss": 0.2935,
+      "step": 10879
+    },
+    {
+      "epoch": 0.8926447060753989,
+      "grad_norm": 0.7715402655729438,
+      "learning_rate": 2.994700934923589e-07,
+      "loss": 0.3104,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8927267506255897,
+      "grad_norm": 0.9164684457541452,
+      "learning_rate": 2.9901732708247256e-07,
+      "loss": 0.3122,
+      "step": 10881
+    },
+    {
+      "epoch": 0.8928087951757805,
+      "grad_norm": 0.7076797274152618,
+      "learning_rate": 2.9856489264843025e-07,
+      "loss": 0.2945,
+      "step": 10882
+    },
+    {
+      "epoch": 0.8928908397259712,
+      "grad_norm": 0.6913351967974697,
+      "learning_rate": 2.981127902221809e-07,
+      "loss": 0.3228,
+      "step": 10883
+    },
+    {
+      "epoch": 0.892972884276162,
+      "grad_norm": 0.7297479767556889,
+      "learning_rate": 2.976610198356533e-07,
+      "loss": 0.2701,
+      "step": 10884
+    },
+    {
+      "epoch": 0.8930549288263527,
+      "grad_norm": 0.6976886066877085,
+      "learning_rate": 2.972095815207487e-07,
+      "loss": 0.2997,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8931369733765434,
+      "grad_norm": 0.8138496560127182,
+      "learning_rate": 2.9675847530934887e-07,
+      "loss": 0.2832,
+      "step": 10886
+    },
+    {
+      "epoch": 0.8932190179267342,
+      "grad_norm": 0.5769117514713917,
+      "learning_rate": 2.963077012333082e-07,
+      "loss": 0.6372,
+      "step": 10887
+    },
+    {
+      "epoch": 0.893301062476925,
+      "grad_norm": 0.6925629277208385,
+      "learning_rate": 2.9585725932446075e-07,
+      "loss": 0.2919,
+      "step": 10888
+    },
+    {
+      "epoch": 0.8933831070271158,
+      "grad_norm": 0.676507696219157,
+      "learning_rate": 2.95407149614616e-07,
+      "loss": 0.318,
+      "step": 10889
+    },
+    {
+      "epoch": 0.8934651515773064,
+      "grad_norm": 0.7065558525680767,
+      "learning_rate": 2.9495737213555977e-07,
+      "loss": 0.3093,
+      "step": 10890
+    },
+    {
+      "epoch": 0.8935471961274972,
+      "grad_norm": 0.8475060699495368,
+      "learning_rate": 2.9450792691905537e-07,
+      "loss": 0.2826,
+      "step": 10891
+    },
+    {
+      "epoch": 0.893629240677688,
+      "grad_norm": 0.6690612937246283,
+      "learning_rate": 2.940588139968403e-07,
+      "loss": 0.2845,
+      "step": 10892
+    },
+    {
+      "epoch": 0.8937112852278788,
+      "grad_norm": 2.205660001416486,
+      "learning_rate": 2.9361003340063087e-07,
+      "loss": 0.3006,
+      "step": 10893
+    },
+    {
+      "epoch": 0.8937933297780695,
+      "grad_norm": 0.9023827879085891,
+      "learning_rate": 2.9316158516212056e-07,
+      "loss": 0.2748,
+      "step": 10894
+    },
+    {
+      "epoch": 0.8938753743282603,
+      "grad_norm": 0.8342565568165216,
+      "learning_rate": 2.9271346931297684e-07,
+      "loss": 0.3195,
+      "step": 10895
+    },
+    {
+      "epoch": 0.893957418878451,
+      "grad_norm": 0.6998111525184544,
+      "learning_rate": 2.9226568588484437e-07,
+      "loss": 0.3137,
+      "step": 10896
+    },
+    {
+      "epoch": 0.8940394634286417,
+      "grad_norm": 0.6423626466166217,
+      "learning_rate": 2.918182349093468e-07,
+      "loss": 0.2918,
+      "step": 10897
+    },
+    {
+      "epoch": 0.8941215079788325,
+      "grad_norm": 0.7210907142927755,
+      "learning_rate": 2.913711164180799e-07,
+      "loss": 0.3087,
+      "step": 10898
+    },
+    {
+      "epoch": 0.8942035525290233,
+      "grad_norm": 0.642432259392805,
+      "learning_rate": 2.909243304426207e-07,
+      "loss": 0.2908,
+      "step": 10899
+    },
+    {
+      "epoch": 0.8942855970792141,
+      "grad_norm": 0.7141665408817929,
+      "learning_rate": 2.904778770145189e-07,
+      "loss": 0.2667,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8943676416294047,
+      "grad_norm": 0.7209895640347819,
+      "learning_rate": 2.9003175616530264e-07,
+      "loss": 0.3078,
+      "step": 10901
+    },
+    {
+      "epoch": 0.8944496861795955,
+      "grad_norm": 0.7486289720917149,
+      "learning_rate": 2.895859679264779e-07,
+      "loss": 0.2733,
+      "step": 10902
+    },
+    {
+      "epoch": 0.8945317307297863,
+      "grad_norm": 0.7739089336241555,
+      "learning_rate": 2.8914051232952325e-07,
+      "loss": 0.3171,
+      "step": 10903
+    },
+    {
+      "epoch": 0.894613775279977,
+      "grad_norm": 0.7738067177352186,
+      "learning_rate": 2.88695389405898e-07,
+      "loss": 0.2889,
+      "step": 10904
+    },
+    {
+      "epoch": 0.8946958198301678,
+      "grad_norm": 1.7362854566028287,
+      "learning_rate": 2.882505991870338e-07,
+      "loss": 0.2793,
+      "step": 10905
+    },
+    {
+      "epoch": 0.8947778643803586,
+      "grad_norm": 0.74298095180728,
+      "learning_rate": 2.8780614170434264e-07,
+      "loss": 0.3691,
+      "step": 10906
+    },
+    {
+      "epoch": 0.8948599089305493,
+      "grad_norm": 0.6595627315675903,
+      "learning_rate": 2.873620169892116e-07,
+      "loss": 0.2924,
+      "step": 10907
+    },
+    {
+      "epoch": 0.89494195348074,
+      "grad_norm": 0.7658306968004694,
+      "learning_rate": 2.86918225073004e-07,
+      "loss": 0.3197,
+      "step": 10908
+    },
+    {
+      "epoch": 0.8950239980309308,
+      "grad_norm": 0.6921074793945114,
+      "learning_rate": 2.8647476598705816e-07,
+      "loss": 0.2923,
+      "step": 10909
+    },
+    {
+      "epoch": 0.8951060425811216,
+      "grad_norm": 0.9151162271055862,
+      "learning_rate": 2.860316397626922e-07,
+      "loss": 0.3353,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8951880871313123,
+      "grad_norm": 0.7508579105971467,
+      "learning_rate": 2.855888464311973e-07,
+      "loss": 0.2737,
+      "step": 10911
+    },
+    {
+      "epoch": 0.895270131681503,
+      "grad_norm": 0.7504671463251783,
+      "learning_rate": 2.851463860238446e-07,
+      "loss": 0.324,
+      "step": 10912
+    },
+    {
+      "epoch": 0.8953521762316938,
+      "grad_norm": 0.7243239263991109,
+      "learning_rate": 2.8470425857187847e-07,
+      "loss": 0.3118,
+      "step": 10913
+    },
+    {
+      "epoch": 0.8954342207818846,
+      "grad_norm": 0.7150438048171645,
+      "learning_rate": 2.842624641065228e-07,
+      "loss": 0.2962,
+      "step": 10914
+    },
+    {
+      "epoch": 0.8955162653320753,
+      "grad_norm": 0.7407985933130281,
+      "learning_rate": 2.8382100265897437e-07,
+      "loss": 0.2932,
+      "step": 10915
+    },
+    {
+      "epoch": 0.8955983098822661,
+      "grad_norm": 0.7708787008633821,
+      "learning_rate": 2.833798742604099e-07,
+      "loss": 0.3219,
+      "step": 10916
+    },
+    {
+      "epoch": 0.8956803544324569,
+      "grad_norm": 2.4149476541879062,
+      "learning_rate": 2.829390789419817e-07,
+      "loss": 0.2786,
+      "step": 10917
+    },
+    {
+      "epoch": 0.8957623989826475,
+      "grad_norm": 0.7934096649680883,
+      "learning_rate": 2.824986167348165e-07,
+      "loss": 0.3296,
+      "step": 10918
+    },
+    {
+      "epoch": 0.8958444435328383,
+      "grad_norm": 0.7026113829093992,
+      "learning_rate": 2.820584876700194e-07,
+      "loss": 0.2902,
+      "step": 10919
+    },
+    {
+      "epoch": 0.8959264880830291,
+      "grad_norm": 0.647293004548505,
+      "learning_rate": 2.8161869177867285e-07,
+      "loss": 0.2689,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8960085326332199,
+      "grad_norm": 0.6738017166603993,
+      "learning_rate": 2.81179229091833e-07,
+      "loss": 0.3168,
+      "step": 10921
+    },
+    {
+      "epoch": 0.8960905771834106,
+      "grad_norm": 0.8033554741463365,
+      "learning_rate": 2.8074009964053574e-07,
+      "loss": 0.3056,
+      "step": 10922
+    },
+    {
+      "epoch": 0.8961726217336013,
+      "grad_norm": 0.7634524289278235,
+      "learning_rate": 2.8030130345578997e-07,
+      "loss": 0.2768,
+      "step": 10923
+    },
+    {
+      "epoch": 0.8962546662837921,
+      "grad_norm": 0.6642941821060885,
+      "learning_rate": 2.798628405685827e-07,
+      "loss": 0.302,
+      "step": 10924
+    },
+    {
+      "epoch": 0.8963367108339828,
+      "grad_norm": 0.6712780798478656,
+      "learning_rate": 2.794247110098791e-07,
+      "loss": 0.3015,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8964187553841736,
+      "grad_norm": 0.7125069174108606,
+      "learning_rate": 2.789869148106172e-07,
+      "loss": 0.3048,
+      "step": 10926
+    },
+    {
+      "epoch": 0.8965007999343644,
+      "grad_norm": 0.8065920775372195,
+      "learning_rate": 2.785494520017157e-07,
+      "loss": 0.316,
+      "step": 10927
+    },
+    {
+      "epoch": 0.8965828444845552,
+      "grad_norm": 0.7001826256501802,
+      "learning_rate": 2.7811232261406527e-07,
+      "loss": 0.3054,
+      "step": 10928
+    },
+    {
+      "epoch": 0.8966648890347458,
+      "grad_norm": 0.7347243611439116,
+      "learning_rate": 2.7767552667853634e-07,
+      "loss": 0.2996,
+      "step": 10929
+    },
+    {
+      "epoch": 0.8967469335849366,
+      "grad_norm": 0.6446090531981381,
+      "learning_rate": 2.7723906422597537e-07,
+      "loss": 0.2927,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8968289781351274,
+      "grad_norm": 0.7144332651038321,
+      "learning_rate": 2.7680293528720314e-07,
+      "loss": 0.2815,
+      "step": 10931
+    },
+    {
+      "epoch": 0.8969110226853181,
+      "grad_norm": 0.7079309331088491,
+      "learning_rate": 2.763671398930201e-07,
+      "loss": 0.3065,
+      "step": 10932
+    },
+    {
+      "epoch": 0.8969930672355089,
+      "grad_norm": 0.6719583319505184,
+      "learning_rate": 2.7593167807419994e-07,
+      "loss": 0.3027,
+      "step": 10933
+    },
+    {
+      "epoch": 0.8970751117856997,
+      "grad_norm": 0.7360533531029342,
+      "learning_rate": 2.754965498614948e-07,
+      "loss": 0.3078,
+      "step": 10934
+    },
+    {
+      "epoch": 0.8971571563358904,
+      "grad_norm": 0.724447040794507,
+      "learning_rate": 2.7506175528563385e-07,
+      "loss": 0.2778,
+      "step": 10935
+    },
+    {
+      "epoch": 0.8972392008860811,
+      "grad_norm": 0.62649123889358,
+      "learning_rate": 2.746272943773198e-07,
+      "loss": 0.2748,
+      "step": 10936
+    },
+    {
+      "epoch": 0.8973212454362719,
+      "grad_norm": 0.6895103352013905,
+      "learning_rate": 2.741931671672343e-07,
+      "loss": 0.3277,
+      "step": 10937
+    },
+    {
+      "epoch": 0.8974032899864627,
+      "grad_norm": 0.6735100633379151,
+      "learning_rate": 2.73759373686035e-07,
+      "loss": 0.3073,
+      "step": 10938
+    },
+    {
+      "epoch": 0.8974853345366534,
+      "grad_norm": 0.7134542516678272,
+      "learning_rate": 2.7332591396435513e-07,
+      "loss": 0.2849,
+      "step": 10939
+    },
+    {
+      "epoch": 0.8975673790868441,
+      "grad_norm": 0.7787113215650179,
+      "learning_rate": 2.7289278803280586e-07,
+      "loss": 0.2637,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8976494236370349,
+      "grad_norm": 0.7699192986227466,
+      "learning_rate": 2.724599959219726e-07,
+      "loss": 0.2975,
+      "step": 10941
+    },
+    {
+      "epoch": 0.8977314681872257,
+      "grad_norm": 0.771421543802181,
+      "learning_rate": 2.720275376624187e-07,
+      "loss": 0.3316,
+      "step": 10942
+    },
+    {
+      "epoch": 0.8978135127374164,
+      "grad_norm": 0.7227235619714899,
+      "learning_rate": 2.7159541328468487e-07,
+      "loss": 0.2846,
+      "step": 10943
+    },
+    {
+      "epoch": 0.8978955572876072,
+      "grad_norm": 0.701707327276715,
+      "learning_rate": 2.7116362281928543e-07,
+      "loss": 0.2707,
+      "step": 10944
+    },
+    {
+      "epoch": 0.897977601837798,
+      "grad_norm": 0.7226938173283675,
+      "learning_rate": 2.707321662967144e-07,
+      "loss": 0.2667,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8980596463879886,
+      "grad_norm": 0.6611351276893066,
+      "learning_rate": 2.703010437474385e-07,
+      "loss": 0.3228,
+      "step": 10946
+    },
+    {
+      "epoch": 0.8981416909381794,
+      "grad_norm": 0.7475286167604361,
+      "learning_rate": 2.698702552019045e-07,
+      "loss": 0.3169,
+      "step": 10947
+    },
+    {
+      "epoch": 0.8982237354883702,
+      "grad_norm": 0.6542295495301631,
+      "learning_rate": 2.694398006905341e-07,
+      "loss": 0.2995,
+      "step": 10948
+    },
+    {
+      "epoch": 0.898305780038561,
+      "grad_norm": 0.7516680005061922,
+      "learning_rate": 2.690096802437242e-07,
+      "loss": 0.3349,
+      "step": 10949
+    },
+    {
+      "epoch": 0.8983878245887517,
+      "grad_norm": 0.7851926248255171,
+      "learning_rate": 2.685798938918505e-07,
+      "loss": 0.2791,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8984698691389424,
+      "grad_norm": 0.7053432169495597,
+      "learning_rate": 2.6815044166526247e-07,
+      "loss": 0.3244,
+      "step": 10951
+    },
+    {
+      "epoch": 0.8985519136891332,
+      "grad_norm": 0.6376767734444132,
+      "learning_rate": 2.677213235942877e-07,
+      "loss": 0.2944,
+      "step": 10952
+    },
+    {
+      "epoch": 0.8986339582393239,
+      "grad_norm": 0.7108949010498005,
+      "learning_rate": 2.672925397092313e-07,
+      "loss": 0.3169,
+      "step": 10953
+    },
+    {
+      "epoch": 0.8987160027895147,
+      "grad_norm": 0.6986416690263251,
+      "learning_rate": 2.668640900403707e-07,
+      "loss": 0.2825,
+      "step": 10954
+    },
+    {
+      "epoch": 0.8987980473397055,
+      "grad_norm": 0.6474496300534479,
+      "learning_rate": 2.66435974617964e-07,
+      "loss": 0.2928,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8988800918898963,
+      "grad_norm": 0.6756780977715422,
+      "learning_rate": 2.6600819347224416e-07,
+      "loss": 0.2945,
+      "step": 10956
+    },
+    {
+      "epoch": 0.8989621364400869,
+      "grad_norm": 0.6594106954472303,
+      "learning_rate": 2.6558074663341926e-07,
+      "loss": 0.2742,
+      "step": 10957
+    },
+    {
+      "epoch": 0.8990441809902777,
+      "grad_norm": 0.8316726060370224,
+      "learning_rate": 2.6515363413167626e-07,
+      "loss": 0.3332,
+      "step": 10958
+    },
+    {
+      "epoch": 0.8991262255404685,
+      "grad_norm": 0.7406249165534987,
+      "learning_rate": 2.647268559971761e-07,
+      "loss": 0.3304,
+      "step": 10959
+    },
+    {
+      "epoch": 0.8992082700906592,
+      "grad_norm": 0.6714476557904773,
+      "learning_rate": 2.643004122600573e-07,
+      "loss": 0.2861,
+      "step": 10960
+    },
+    {
+      "epoch": 0.89929031464085,
+      "grad_norm": 0.6820222224041373,
+      "learning_rate": 2.638743029504359e-07,
+      "loss": 0.2592,
+      "step": 10961
+    },
+    {
+      "epoch": 0.8993723591910407,
+      "grad_norm": 0.8658625017456942,
+      "learning_rate": 2.6344852809840114e-07,
+      "loss": 0.292,
+      "step": 10962
+    },
+    {
+      "epoch": 0.8994544037412315,
+      "grad_norm": 0.6881513902831928,
+      "learning_rate": 2.630230877340223e-07,
+      "loss": 0.2853,
+      "step": 10963
+    },
+    {
+      "epoch": 0.8995364482914222,
+      "grad_norm": 0.759072456430159,
+      "learning_rate": 2.625979818873425e-07,
+      "loss": 0.2659,
+      "step": 10964
+    },
+    {
+      "epoch": 0.899618492841613,
+      "grad_norm": 0.7259277008942375,
+      "learning_rate": 2.621732105883812e-07,
+      "loss": 0.3088,
+      "step": 10965
+    },
+    {
+      "epoch": 0.8997005373918038,
+      "grad_norm": 0.6871182070206967,
+      "learning_rate": 2.617487738671365e-07,
+      "loss": 0.2914,
+      "step": 10966
+    },
+    {
+      "epoch": 0.8997825819419945,
+      "grad_norm": 0.6590745968776163,
+      "learning_rate": 2.6132467175358003e-07,
+      "loss": 0.3164,
+      "step": 10967
+    },
+    {
+      "epoch": 0.8998646264921852,
+      "grad_norm": 0.7841070715738664,
+      "learning_rate": 2.609009042776628e-07,
+      "loss": 0.3279,
+      "step": 10968
+    },
+    {
+      "epoch": 0.899946671042376,
+      "grad_norm": 0.6487785954714603,
+      "learning_rate": 2.6047747146930924e-07,
+      "loss": 0.3056,
+      "step": 10969
+    },
+    {
+      "epoch": 0.9000287155925668,
+      "grad_norm": 0.6473659161272229,
+      "learning_rate": 2.6005437335842155e-07,
+      "loss": 0.2593,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9001107601427575,
+      "grad_norm": 0.6755981882315768,
+      "learning_rate": 2.5963160997487967e-07,
+      "loss": 0.2923,
+      "step": 10971
+    },
+    {
+      "epoch": 0.9001928046929483,
+      "grad_norm": 0.7177838404799121,
+      "learning_rate": 2.59209181348537e-07,
+      "loss": 0.2931,
+      "step": 10972
+    },
+    {
+      "epoch": 0.900274849243139,
+      "grad_norm": 0.754206304384468,
+      "learning_rate": 2.587870875092252e-07,
+      "loss": 0.3062,
+      "step": 10973
+    },
+    {
+      "epoch": 0.9003568937933297,
+      "grad_norm": 0.7032384668202015,
+      "learning_rate": 2.5836532848675267e-07,
+      "loss": 0.3349,
+      "step": 10974
+    },
+    {
+      "epoch": 0.9004389383435205,
+      "grad_norm": 0.6621521152012244,
+      "learning_rate": 2.5794390431090166e-07,
+      "loss": 0.2934,
+      "step": 10975
+    },
+    {
+      "epoch": 0.9005209828937113,
+      "grad_norm": 0.7231851557628248,
+      "learning_rate": 2.575228150114345e-07,
+      "loss": 0.3363,
+      "step": 10976
+    },
+    {
+      "epoch": 0.9006030274439021,
+      "grad_norm": 0.6897300109261915,
+      "learning_rate": 2.571020606180857e-07,
+      "loss": 0.2837,
+      "step": 10977
+    },
+    {
+      "epoch": 0.9006850719940928,
+      "grad_norm": 0.8102525494178103,
+      "learning_rate": 2.5668164116057046e-07,
+      "loss": 0.3332,
+      "step": 10978
+    },
+    {
+      "epoch": 0.9007671165442835,
+      "grad_norm": 0.643120276051315,
+      "learning_rate": 2.562615566685767e-07,
+      "loss": 0.2991,
+      "step": 10979
+    },
+    {
+      "epoch": 0.9008491610944743,
+      "grad_norm": 0.7357569171138059,
+      "learning_rate": 2.5584180717177e-07,
+      "loss": 0.3097,
+      "step": 10980
+    },
+    {
+      "epoch": 0.900931205644665,
+      "grad_norm": 0.660804791571219,
+      "learning_rate": 2.5542239269979283e-07,
+      "loss": 0.2992,
+      "step": 10981
+    },
+    {
+      "epoch": 0.9010132501948558,
+      "grad_norm": 0.7149469169297925,
+      "learning_rate": 2.5500331328226324e-07,
+      "loss": 0.307,
+      "step": 10982
+    },
+    {
+      "epoch": 0.9010952947450466,
+      "grad_norm": 0.7380529770841104,
+      "learning_rate": 2.5458456894877637e-07,
+      "loss": 0.2695,
+      "step": 10983
+    },
+    {
+      "epoch": 0.9011773392952374,
+      "grad_norm": 0.673349333058996,
+      "learning_rate": 2.5416615972890414e-07,
+      "loss": 0.2887,
+      "step": 10984
+    },
+    {
+      "epoch": 0.901259383845428,
+      "grad_norm": 0.675525054496975,
+      "learning_rate": 2.5374808565219175e-07,
+      "loss": 0.2867,
+      "step": 10985
+    },
+    {
+      "epoch": 0.9013414283956188,
+      "grad_norm": 0.7528596778034459,
+      "learning_rate": 2.533303467481646e-07,
+      "loss": 0.2845,
+      "step": 10986
+    },
+    {
+      "epoch": 0.9014234729458096,
+      "grad_norm": 0.7320094743479311,
+      "learning_rate": 2.529129430463223e-07,
+      "loss": 0.3083,
+      "step": 10987
+    },
+    {
+      "epoch": 0.9015055174960003,
+      "grad_norm": 0.6617059159259602,
+      "learning_rate": 2.524958745761408e-07,
+      "loss": 0.2839,
+      "step": 10988
+    },
+    {
+      "epoch": 0.9015875620461911,
+      "grad_norm": 0.7034405300692741,
+      "learning_rate": 2.5207914136707366e-07,
+      "loss": 0.3485,
+      "step": 10989
+    },
+    {
+      "epoch": 0.9016696065963818,
+      "grad_norm": 0.8115476318846994,
+      "learning_rate": 2.516627434485491e-07,
+      "loss": 0.2519,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9017516511465726,
+      "grad_norm": 0.7472083797993223,
+      "learning_rate": 2.5124668084997306e-07,
+      "loss": 0.3277,
+      "step": 10991
+    },
+    {
+      "epoch": 0.9018336956967633,
+      "grad_norm": 0.634162211539971,
+      "learning_rate": 2.50830953600727e-07,
+      "loss": 0.2863,
+      "step": 10992
+    },
+    {
+      "epoch": 0.9019157402469541,
+      "grad_norm": 0.6929967429918459,
+      "learning_rate": 2.50415561730169e-07,
+      "loss": 0.3163,
+      "step": 10993
+    },
+    {
+      "epoch": 0.9019977847971449,
+      "grad_norm": 0.7273296070690904,
+      "learning_rate": 2.5000050526763353e-07,
+      "loss": 0.3284,
+      "step": 10994
+    },
+    {
+      "epoch": 0.9020798293473355,
+      "grad_norm": 0.7347387676900763,
+      "learning_rate": 2.4958578424242984e-07,
+      "loss": 0.2737,
+      "step": 10995
+    },
+    {
+      "epoch": 0.9021618738975263,
+      "grad_norm": 0.6833167916048214,
+      "learning_rate": 2.491713986838462e-07,
+      "loss": 0.2862,
+      "step": 10996
+    },
+    {
+      "epoch": 0.9022439184477171,
+      "grad_norm": 0.6904592393916806,
+      "learning_rate": 2.4875734862114644e-07,
+      "loss": 0.268,
+      "step": 10997
+    },
+    {
+      "epoch": 0.9023259629979079,
+      "grad_norm": 0.6655998790747366,
+      "learning_rate": 2.4834363408356823e-07,
+      "loss": 0.293,
+      "step": 10998
+    },
+    {
+      "epoch": 0.9024080075480986,
+      "grad_norm": 0.6612945124174688,
+      "learning_rate": 2.479302551003293e-07,
+      "loss": 0.2879,
+      "step": 10999
+    },
+    {
+      "epoch": 0.9024900520982894,
+      "grad_norm": 0.6554125010626053,
+      "learning_rate": 2.4751721170062026e-07,
+      "loss": 0.2612,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9025720966484801,
+      "grad_norm": 0.6837216428100591,
+      "learning_rate": 2.471045039136105e-07,
+      "loss": 0.3143,
+      "step": 11001
+    },
+    {
+      "epoch": 0.9026541411986708,
+      "grad_norm": 0.7060102261572091,
+      "learning_rate": 2.4669213176844507e-07,
+      "loss": 0.3461,
+      "step": 11002
+    },
+    {
+      "epoch": 0.9027361857488616,
+      "grad_norm": 0.756044293292788,
+      "learning_rate": 2.4628009529424403e-07,
+      "loss": 0.2944,
+      "step": 11003
+    },
+    {
+      "epoch": 0.9028182302990524,
+      "grad_norm": 0.6860545519864056,
+      "learning_rate": 2.4586839452010567e-07,
+      "loss": 0.3373,
+      "step": 11004
+    },
+    {
+      "epoch": 0.9029002748492432,
+      "grad_norm": 0.6932370561818255,
+      "learning_rate": 2.454570294751024e-07,
+      "loss": 0.3194,
+      "step": 11005
+    },
+    {
+      "epoch": 0.9029823193994339,
+      "grad_norm": 0.6502664657086499,
+      "learning_rate": 2.4504600018828593e-07,
+      "loss": 0.2667,
+      "step": 11006
+    },
+    {
+      "epoch": 0.9030643639496246,
+      "grad_norm": 0.7525412180783677,
+      "learning_rate": 2.446353066886814e-07,
+      "loss": 0.2903,
+      "step": 11007
+    },
+    {
+      "epoch": 0.9031464084998154,
+      "grad_norm": 0.7767236522190514,
+      "learning_rate": 2.4422494900529057e-07,
+      "loss": 0.3,
+      "step": 11008
+    },
+    {
+      "epoch": 0.9032284530500062,
+      "grad_norm": 0.6809199235657354,
+      "learning_rate": 2.438149271670931e-07,
+      "loss": 0.3033,
+      "step": 11009
+    },
+    {
+      "epoch": 0.9033104976001969,
+      "grad_norm": 0.6689647860182254,
+      "learning_rate": 2.434052412030452e-07,
+      "loss": 0.2632,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9033925421503877,
+      "grad_norm": 0.698770739017651,
+      "learning_rate": 2.42995891142076e-07,
+      "loss": 0.2805,
+      "step": 11011
+    },
+    {
+      "epoch": 0.9034745867005785,
+      "grad_norm": 0.7491888259008203,
+      "learning_rate": 2.4258687701309513e-07,
+      "loss": 0.3114,
+      "step": 11012
+    },
+    {
+      "epoch": 0.9035566312507691,
+      "grad_norm": 0.7461443866255235,
+      "learning_rate": 2.421781988449851e-07,
+      "loss": 0.2977,
+      "step": 11013
+    },
+    {
+      "epoch": 0.9036386758009599,
+      "grad_norm": 0.6547502988139429,
+      "learning_rate": 2.417698566666066e-07,
+      "loss": 0.3065,
+      "step": 11014
+    },
+    {
+      "epoch": 0.9037207203511507,
+      "grad_norm": 0.7170168191335047,
+      "learning_rate": 2.413618505067966e-07,
+      "loss": 0.3136,
+      "step": 11015
+    },
+    {
+      "epoch": 0.9038027649013415,
+      "grad_norm": 0.7530850800747291,
+      "learning_rate": 2.409541803943666e-07,
+      "loss": 0.306,
+      "step": 11016
+    },
+    {
+      "epoch": 0.9038848094515322,
+      "grad_norm": 0.7945131902821084,
+      "learning_rate": 2.4054684635810733e-07,
+      "loss": 0.3244,
+      "step": 11017
+    },
+    {
+      "epoch": 0.9039668540017229,
+      "grad_norm": 0.6959933746603186,
+      "learning_rate": 2.401398484267825e-07,
+      "loss": 0.3108,
+      "step": 11018
+    },
+    {
+      "epoch": 0.9040488985519137,
+      "grad_norm": 0.7754895433121995,
+      "learning_rate": 2.397331866291347e-07,
+      "loss": 0.2976,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9041309431021044,
+      "grad_norm": 0.6970995374692873,
+      "learning_rate": 2.3932686099388094e-07,
+      "loss": 0.2905,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9042129876522952,
+      "grad_norm": 0.6701163578447653,
+      "learning_rate": 2.3892087154971597e-07,
+      "loss": 0.3075,
+      "step": 11021
+    },
+    {
+      "epoch": 0.904295032202486,
+      "grad_norm": 0.8290086014854928,
+      "learning_rate": 2.385152183253103e-07,
+      "loss": 0.316,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9043770767526768,
+      "grad_norm": 0.7137747188170519,
+      "learning_rate": 2.3810990134930878e-07,
+      "loss": 0.2898,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9044591213028674,
+      "grad_norm": 0.6632761642528989,
+      "learning_rate": 2.3770492065033567e-07,
+      "loss": 0.2887,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9045411658530582,
+      "grad_norm": 0.9469760754777237,
+      "learning_rate": 2.3730027625699036e-07,
+      "loss": 0.3123,
+      "step": 11025
+    },
+    {
+      "epoch": 0.904623210403249,
+      "grad_norm": 0.7391621208614291,
+      "learning_rate": 2.3689596819784723e-07,
+      "loss": 0.2892,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9047052549534397,
+      "grad_norm": 0.6746043885206133,
+      "learning_rate": 2.3649199650145836e-07,
+      "loss": 0.2842,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9047872995036305,
+      "grad_norm": 0.8493418297404041,
+      "learning_rate": 2.3608836119635214e-07,
+      "loss": 0.2877,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9048693440538212,
+      "grad_norm": 0.7312575074067926,
+      "learning_rate": 2.356850623110313e-07,
+      "loss": 0.2986,
+      "step": 11029
+    },
+    {
+      "epoch": 0.904951388604012,
+      "grad_norm": 0.6827590215957832,
+      "learning_rate": 2.3528209987397798e-07,
+      "loss": 0.2691,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9050334331542027,
+      "grad_norm": 0.7010946062291399,
+      "learning_rate": 2.3487947391364673e-07,
+      "loss": 0.3005,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9051154777043935,
+      "grad_norm": 0.7109617996154037,
+      "learning_rate": 2.3447718445847145e-07,
+      "loss": 0.2648,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9051975222545843,
+      "grad_norm": 0.7005979114952052,
+      "learning_rate": 2.3407523153686108e-07,
+      "loss": 0.3336,
+      "step": 11033
+    },
+    {
+      "epoch": 0.905279566804775,
+      "grad_norm": 0.654897908838917,
+      "learning_rate": 2.336736151772012e-07,
+      "loss": 0.2938,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9053616113549657,
+      "grad_norm": 0.7047492158491926,
+      "learning_rate": 2.3327233540785255e-07,
+      "loss": 0.3572,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9054436559051565,
+      "grad_norm": 0.8762317752050633,
+      "learning_rate": 2.3287139225715294e-07,
+      "loss": 0.2801,
+      "step": 11036
+    },
+    {
+      "epoch": 0.9055257004553473,
+      "grad_norm": 0.9623297676840968,
+      "learning_rate": 2.3247078575341697e-07,
+      "loss": 0.3457,
+      "step": 11037
+    },
+    {
+      "epoch": 0.905607745005538,
+      "grad_norm": 0.7242520921475294,
+      "learning_rate": 2.3207051592493478e-07,
+      "loss": 0.3226,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9056897895557288,
+      "grad_norm": 0.6963732155299343,
+      "learning_rate": 2.3167058279997156e-07,
+      "loss": 0.3094,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9057718341059195,
+      "grad_norm": 0.6867845384631519,
+      "learning_rate": 2.312709864067719e-07,
+      "loss": 0.2795,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9058538786561102,
+      "grad_norm": 0.6977438502993485,
+      "learning_rate": 2.3087172677355275e-07,
+      "loss": 0.2961,
+      "step": 11041
+    },
+    {
+      "epoch": 0.905935923206301,
+      "grad_norm": 1.208515892182109,
+      "learning_rate": 2.3047280392851035e-07,
+      "loss": 0.3296,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9060179677564918,
+      "grad_norm": 0.7595794961611299,
+      "learning_rate": 2.300742178998161e-07,
+      "loss": 0.2983,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9061000123066826,
+      "grad_norm": 0.801648114763078,
+      "learning_rate": 2.2967596871561694e-07,
+      "loss": 0.2384,
+      "step": 11044
+    },
+    {
+      "epoch": 0.9061820568568733,
+      "grad_norm": 0.6198628507488491,
+      "learning_rate": 2.2927805640403645e-07,
+      "loss": 0.2722,
+      "step": 11045
+    },
+    {
+      "epoch": 0.906264101407064,
+      "grad_norm": 0.6470893711774793,
+      "learning_rate": 2.2888048099317495e-07,
+      "loss": 0.2886,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9063461459572548,
+      "grad_norm": 0.7378369003078725,
+      "learning_rate": 2.2848324251110942e-07,
+      "loss": 0.2957,
+      "step": 11047
+    },
+    {
+      "epoch": 0.9064281905074455,
+      "grad_norm": 0.722455821366662,
+      "learning_rate": 2.2808634098589022e-07,
+      "loss": 0.3379,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9065102350576363,
+      "grad_norm": 0.719580295825653,
+      "learning_rate": 2.2768977644554767e-07,
+      "loss": 0.2714,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9065922796078271,
+      "grad_norm": 0.6969653727960582,
+      "learning_rate": 2.2729354891808498e-07,
+      "loss": 0.3101,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9066743241580179,
+      "grad_norm": 0.7058104348975818,
+      "learning_rate": 2.268976584314847e-07,
+      "loss": 0.2841,
+      "step": 11051
+    },
+    {
+      "epoch": 0.9067563687082085,
+      "grad_norm": 0.6937810204266641,
+      "learning_rate": 2.265021050137023e-07,
+      "loss": 0.3022,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9068384132583993,
+      "grad_norm": 0.6349471734156742,
+      "learning_rate": 2.2610688869267318e-07,
+      "loss": 0.2955,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9069204578085901,
+      "grad_norm": 0.6163104236466069,
+      "learning_rate": 2.2571200949630446e-07,
+      "loss": 0.2923,
+      "step": 11054
+    },
+    {
+      "epoch": 0.9070025023587808,
+      "grad_norm": 0.7184975274483083,
+      "learning_rate": 2.2531746745248384e-07,
+      "loss": 0.3048,
+      "step": 11055
+    },
+    {
+      "epoch": 0.9070845469089716,
+      "grad_norm": 0.710219991700766,
+      "learning_rate": 2.2492326258907293e-07,
+      "loss": 0.2761,
+      "step": 11056
+    },
+    {
+      "epoch": 0.9071665914591623,
+      "grad_norm": 0.7387458967560435,
+      "learning_rate": 2.2452939493390892e-07,
+      "loss": 0.3268,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9072486360093531,
+      "grad_norm": 0.8146173277574303,
+      "learning_rate": 2.241358645148062e-07,
+      "loss": 0.2919,
+      "step": 11058
+    },
+    {
+      "epoch": 0.9073306805595438,
+      "grad_norm": 0.6777887596749318,
+      "learning_rate": 2.2374267135955695e-07,
+      "loss": 0.2836,
+      "step": 11059
+    },
+    {
+      "epoch": 0.9074127251097346,
+      "grad_norm": 0.6824686972532733,
+      "learning_rate": 2.2334981549592627e-07,
+      "loss": 0.2807,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9074947696599254,
+      "grad_norm": 0.6892679681296934,
+      "learning_rate": 2.229572969516569e-07,
+      "loss": 0.291,
+      "step": 11061
+    },
+    {
+      "epoch": 0.907576814210116,
+      "grad_norm": 0.7593601270513842,
+      "learning_rate": 2.2256511575446837e-07,
+      "loss": 0.3515,
+      "step": 11062
+    },
+    {
+      "epoch": 0.9076588587603068,
+      "grad_norm": 0.7320659277101225,
+      "learning_rate": 2.2217327193205572e-07,
+      "loss": 0.3026,
+      "step": 11063
+    },
+    {
+      "epoch": 0.9077409033104976,
+      "grad_norm": 0.8519021292311116,
+      "learning_rate": 2.217817655120913e-07,
+      "loss": 0.306,
+      "step": 11064
+    },
+    {
+      "epoch": 0.9078229478606884,
+      "grad_norm": 0.637939939296928,
+      "learning_rate": 2.2139059652222074e-07,
+      "loss": 0.2959,
+      "step": 11065
+    },
+    {
+      "epoch": 0.9079049924108791,
+      "grad_norm": 0.7213678893578991,
+      "learning_rate": 2.209997649900697e-07,
+      "loss": 0.3024,
+      "step": 11066
+    },
+    {
+      "epoch": 0.9079870369610699,
+      "grad_norm": 0.6767659548133198,
+      "learning_rate": 2.2060927094323613e-07,
+      "loss": 0.2627,
+      "step": 11067
+    },
+    {
+      "epoch": 0.9080690815112606,
+      "grad_norm": 0.7002209969781469,
+      "learning_rate": 2.2021911440929745e-07,
+      "loss": 0.2575,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9081511260614513,
+      "grad_norm": 0.7313252830457433,
+      "learning_rate": 2.1982929541580656e-07,
+      "loss": 0.3176,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9082331706116421,
+      "grad_norm": 0.6518779735900359,
+      "learning_rate": 2.194398139902898e-07,
+      "loss": 0.2958,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9083152151618329,
+      "grad_norm": 0.7611166918004683,
+      "learning_rate": 2.1905067016025404e-07,
+      "loss": 0.295,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9083972597120237,
+      "grad_norm": 0.8815634139595656,
+      "learning_rate": 2.1866186395317735e-07,
+      "loss": 0.2879,
+      "step": 11072
+    },
+    {
+      "epoch": 0.9084793042622143,
+      "grad_norm": 0.7182525432083767,
+      "learning_rate": 2.1827339539651882e-07,
+      "loss": 0.3243,
+      "step": 11073
+    },
+    {
+      "epoch": 0.9085613488124051,
+      "grad_norm": 0.7921510005677898,
+      "learning_rate": 2.1788526451771096e-07,
+      "loss": 0.2919,
+      "step": 11074
+    },
+    {
+      "epoch": 0.9086433933625959,
+      "grad_norm": 0.7742245181975927,
+      "learning_rate": 2.1749747134416242e-07,
+      "loss": 0.3046,
+      "step": 11075
+    },
+    {
+      "epoch": 0.9087254379127866,
+      "grad_norm": 0.7998605355840976,
+      "learning_rate": 2.1711001590325854e-07,
+      "loss": 0.3477,
+      "step": 11076
+    },
+    {
+      "epoch": 0.9088074824629774,
+      "grad_norm": 0.8127318742578645,
+      "learning_rate": 2.1672289822236126e-07,
+      "loss": 0.3224,
+      "step": 11077
+    },
+    {
+      "epoch": 0.9088895270131682,
+      "grad_norm": 0.8252353584170111,
+      "learning_rate": 2.1633611832880762e-07,
+      "loss": 0.3198,
+      "step": 11078
+    },
+    {
+      "epoch": 0.908971571563359,
+      "grad_norm": 0.730718029338126,
+      "learning_rate": 2.1594967624991248e-07,
+      "loss": 0.2716,
+      "step": 11079
+    },
+    {
+      "epoch": 0.9090536161135496,
+      "grad_norm": 0.7069483997991824,
+      "learning_rate": 2.1556357201296396e-07,
+      "loss": 0.2873,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9091356606637404,
+      "grad_norm": 0.7812522729731106,
+      "learning_rate": 2.1517780564522973e-07,
+      "loss": 0.2634,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9092177052139312,
+      "grad_norm": 0.7971623667188922,
+      "learning_rate": 2.1479237717395185e-07,
+      "loss": 0.3153,
+      "step": 11082
+    },
+    {
+      "epoch": 0.9092997497641219,
+      "grad_norm": 0.7474683658524715,
+      "learning_rate": 2.1440728662634746e-07,
+      "loss": 0.269,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9093817943143127,
+      "grad_norm": 0.9141033943705338,
+      "learning_rate": 2.1402253402961316e-07,
+      "loss": 0.271,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9094638388645034,
+      "grad_norm": 0.6829019139237121,
+      "learning_rate": 2.136381194109166e-07,
+      "loss": 0.3032,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9095458834146942,
+      "grad_norm": 0.7163204016491467,
+      "learning_rate": 2.1325404279740725e-07,
+      "loss": 0.3078,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9096279279648849,
+      "grad_norm": 0.6296729594540968,
+      "learning_rate": 2.1287030421620669e-07,
+      "loss": 0.2721,
+      "step": 11087
+    },
+    {
+      "epoch": 0.9097099725150757,
+      "grad_norm": 0.8056880158734023,
+      "learning_rate": 2.1248690369441382e-07,
+      "loss": 0.3215,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9097920170652665,
+      "grad_norm": 0.7422725043726729,
+      "learning_rate": 2.1210384125910477e-07,
+      "loss": 0.2826,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9098740616154571,
+      "grad_norm": 0.7596159849539639,
+      "learning_rate": 2.1172111693733012e-07,
+      "loss": 0.3075,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9099561061656479,
+      "grad_norm": 0.6958568945036104,
+      "learning_rate": 2.1133873075611654e-07,
+      "loss": 0.2995,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9100381507158387,
+      "grad_norm": 0.6610804583999377,
+      "learning_rate": 2.1095668274246862e-07,
+      "loss": 0.273,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9101201952660295,
+      "grad_norm": 0.7562142176782358,
+      "learning_rate": 2.1057497292336525e-07,
+      "loss": 0.3206,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9102022398162202,
+      "grad_norm": 0.674812575518776,
+      "learning_rate": 2.1019360132576272e-07,
+      "loss": 0.3304,
+      "step": 11094
+    },
+    {
+      "epoch": 0.910284284366411,
+      "grad_norm": 0.6944436958949896,
+      "learning_rate": 2.0981256797659277e-07,
+      "loss": 0.3174,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9103663289166017,
+      "grad_norm": 0.6963497239104594,
+      "learning_rate": 2.094318729027628e-07,
+      "loss": 0.3354,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9104483734667924,
+      "grad_norm": 0.7484602828551651,
+      "learning_rate": 2.0905151613115792e-07,
+      "loss": 0.2678,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9105304180169832,
+      "grad_norm": 0.8689835891728948,
+      "learning_rate": 2.0867149768863726e-07,
+      "loss": 0.3225,
+      "step": 11098
+    },
+    {
+      "epoch": 0.910612462567174,
+      "grad_norm": 0.6695250180169582,
+      "learning_rate": 2.0829181760203766e-07,
+      "loss": 0.2939,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9106945071173648,
+      "grad_norm": 0.7672080529831762,
+      "learning_rate": 2.0791247589817155e-07,
+      "loss": 0.3361,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9107765516675554,
+      "grad_norm": 0.6908880266610972,
+      "learning_rate": 2.0753347260382695e-07,
+      "loss": 0.3023,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9108585962177462,
+      "grad_norm": 0.7335737962371754,
+      "learning_rate": 2.0715480774576967e-07,
+      "loss": 0.2793,
+      "step": 11102
+    },
+    {
+      "epoch": 0.910940640767937,
+      "grad_norm": 0.8605776566532689,
+      "learning_rate": 2.0677648135073948e-07,
+      "loss": 0.334,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9110226853181277,
+      "grad_norm": 0.8453429866548898,
+      "learning_rate": 2.063984934454527e-07,
+      "loss": 0.3084,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9111047298683185,
+      "grad_norm": 0.8217292028223222,
+      "learning_rate": 2.0602084405660306e-07,
+      "loss": 0.299,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9111867744185093,
+      "grad_norm": 0.7279077637493089,
+      "learning_rate": 2.0564353321085918e-07,
+      "loss": 0.3069,
+      "step": 11106
+    },
+    {
+      "epoch": 0.9112688189687,
+      "grad_norm": 0.7902321574938087,
+      "learning_rate": 2.052665609348664e-07,
+      "loss": 0.3025,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9113508635188907,
+      "grad_norm": 0.6983258835738846,
+      "learning_rate": 2.048899272552457e-07,
+      "loss": 0.2598,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9114329080690815,
+      "grad_norm": 0.6930793038930163,
+      "learning_rate": 2.045136321985941e-07,
+      "loss": 0.3122,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9115149526192723,
+      "grad_norm": 0.6965650341963195,
+      "learning_rate": 2.0413767579148536e-07,
+      "loss": 0.2605,
+      "step": 11110
+    },
+    {
+      "epoch": 0.911596997169463,
+      "grad_norm": 0.7330287493965384,
+      "learning_rate": 2.0376205806046878e-07,
+      "loss": 0.2955,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9116790417196537,
+      "grad_norm": 0.6682015558805867,
+      "learning_rate": 2.0338677903206982e-07,
+      "loss": 0.3331,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9117610862698445,
+      "grad_norm": 0.7768337484211976,
+      "learning_rate": 2.0301183873279116e-07,
+      "loss": 0.3355,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9118431308200353,
+      "grad_norm": 0.6786458739914953,
+      "learning_rate": 2.0263723718910831e-07,
+      "loss": 0.293,
+      "step": 11114
+    },
+    {
+      "epoch": 0.911925175370226,
+      "grad_norm": 0.7218622278293266,
+      "learning_rate": 2.0226297442747733e-07,
+      "loss": 0.336,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9120072199204168,
+      "grad_norm": 0.6616774691738434,
+      "learning_rate": 2.0188905047432594e-07,
+      "loss": 0.2783,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9120892644706076,
+      "grad_norm": 0.7835064503616289,
+      "learning_rate": 2.0151546535606193e-07,
+      "loss": 0.3324,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9121713090207983,
+      "grad_norm": 0.7490123273432472,
+      "learning_rate": 2.0114221909906584e-07,
+      "loss": 0.2902,
+      "step": 11118
+    },
+    {
+      "epoch": 0.912253353570989,
+      "grad_norm": 0.7477377020337048,
+      "learning_rate": 2.0076931172969605e-07,
+      "loss": 0.2862,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9123353981211798,
+      "grad_norm": 0.7082522973919244,
+      "learning_rate": 2.003967432742876e-07,
+      "loss": 0.2622,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9124174426713706,
+      "grad_norm": 0.900628046817474,
+      "learning_rate": 2.0002451375914944e-07,
+      "loss": 0.3278,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9124994872215613,
+      "grad_norm": 0.8143182244016149,
+      "learning_rate": 1.9965262321056832e-07,
+      "loss": 0.3041,
+      "step": 11122
+    },
+    {
+      "epoch": 0.912581531771752,
+      "grad_norm": 0.7381831746901284,
+      "learning_rate": 1.9928107165480715e-07,
+      "loss": 0.2957,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9126635763219428,
+      "grad_norm": 0.6880101939996542,
+      "learning_rate": 1.9890985911810322e-07,
+      "loss": 0.2839,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9127456208721336,
+      "grad_norm": 0.6611435576517619,
+      "learning_rate": 1.9853898562667174e-07,
+      "loss": 0.2792,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9128276654223243,
+      "grad_norm": 0.7651843177811886,
+      "learning_rate": 1.981684512067028e-07,
+      "loss": 0.2889,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9129097099725151,
+      "grad_norm": 0.7063201955641065,
+      "learning_rate": 1.9779825588436276e-07,
+      "loss": 0.3165,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9129917545227059,
+      "grad_norm": 0.7166086200810546,
+      "learning_rate": 1.9742839968579562e-07,
+      "loss": 0.2831,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9130737990728965,
+      "grad_norm": 0.7272325939136827,
+      "learning_rate": 1.9705888263711837e-07,
+      "loss": 0.2795,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9131558436230873,
+      "grad_norm": 0.7363807413938573,
+      "learning_rate": 1.9668970476442617e-07,
+      "loss": 0.3198,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9132378881732781,
+      "grad_norm": 0.8685586385926733,
+      "learning_rate": 1.9632086609379041e-07,
+      "loss": 0.2825,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9133199327234689,
+      "grad_norm": 0.6729068293537376,
+      "learning_rate": 1.9595236665125694e-07,
+      "loss": 0.3364,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9134019772736596,
+      "grad_norm": 0.7141158952086742,
+      "learning_rate": 1.9558420646284937e-07,
+      "loss": 0.2847,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9134840218238504,
+      "grad_norm": 0.671388230888706,
+      "learning_rate": 1.952163855545658e-07,
+      "loss": 0.3148,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9135660663740411,
+      "grad_norm": 0.7163166672678231,
+      "learning_rate": 1.9484890395238155e-07,
+      "loss": 0.308,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9136481109242318,
+      "grad_norm": 0.6620913463068657,
+      "learning_rate": 1.9448176168224863e-07,
+      "loss": 0.2926,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9137301554744226,
+      "grad_norm": 0.7310237576023542,
+      "learning_rate": 1.941149587700919e-07,
+      "loss": 0.3276,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9138122000246134,
+      "grad_norm": 0.6719753184940123,
+      "learning_rate": 1.9374849524181617e-07,
+      "loss": 0.3061,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9138942445748042,
+      "grad_norm": 0.7110357693715192,
+      "learning_rate": 1.933823711232996e-07,
+      "loss": 0.2876,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9139762891249948,
+      "grad_norm": 0.7633738260756525,
+      "learning_rate": 1.9301658644039712e-07,
+      "loss": 0.2736,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9140583336751856,
+      "grad_norm": 0.6973669268543403,
+      "learning_rate": 1.9265114121894135e-07,
+      "loss": 0.2947,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9141403782253764,
+      "grad_norm": 0.7265682684983477,
+      "learning_rate": 1.922860354847378e-07,
+      "loss": 0.312,
+      "step": 11142
+    },
+    {
+      "epoch": 0.9142224227755671,
+      "grad_norm": 1.3213079836829484,
+      "learning_rate": 1.919212692635708e-07,
+      "loss": 0.3032,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9143044673257579,
+      "grad_norm": 0.7311004340338896,
+      "learning_rate": 1.915568425811981e-07,
+      "loss": 0.2869,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9143865118759487,
+      "grad_norm": 0.7224139881290137,
+      "learning_rate": 1.9119275546335637e-07,
+      "loss": 0.2788,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9144685564261394,
+      "grad_norm": 0.7465645922246069,
+      "learning_rate": 1.9082900793575665e-07,
+      "loss": 0.27,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9145506009763301,
+      "grad_norm": 0.7400633987061468,
+      "learning_rate": 1.904656000240851e-07,
+      "loss": 0.278,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9146326455265209,
+      "grad_norm": 0.7230530685052629,
+      "learning_rate": 1.901025317540056e-07,
+      "loss": 0.3184,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9147146900767117,
+      "grad_norm": 0.675960299376606,
+      "learning_rate": 1.897398031511588e-07,
+      "loss": 0.3331,
+      "step": 11149
+    },
+    {
+      "epoch": 0.9147967346269024,
+      "grad_norm": 0.7848120208910211,
+      "learning_rate": 1.8937741424115751e-07,
+      "loss": 0.2967,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9148787791770931,
+      "grad_norm": 0.6717026894063174,
+      "learning_rate": 1.8901536504959516e-07,
+      "loss": 0.2932,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9149608237272839,
+      "grad_norm": 0.664932907836154,
+      "learning_rate": 1.8865365560203797e-07,
+      "loss": 0.3189,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9150428682774747,
+      "grad_norm": 0.690621836871945,
+      "learning_rate": 1.8829228592402938e-07,
+      "loss": 0.2904,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9151249128276654,
+      "grad_norm": 1.0322625913187795,
+      "learning_rate": 1.879312560410901e-07,
+      "loss": 0.291,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9152069573778562,
+      "grad_norm": 0.7454802152204937,
+      "learning_rate": 1.8757056597871305e-07,
+      "loss": 0.2977,
+      "step": 11155
+    },
+    {
+      "epoch": 0.915289001928047,
+      "grad_norm": 0.7651793497618505,
+      "learning_rate": 1.8721021576237175e-07,
+      "loss": 0.3649,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9153710464782376,
+      "grad_norm": 0.7272753752748902,
+      "learning_rate": 1.8685020541751197e-07,
+      "loss": 0.2707,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9154530910284284,
+      "grad_norm": 0.705189845612724,
+      "learning_rate": 1.8649053496955838e-07,
+      "loss": 0.2764,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9155351355786192,
+      "grad_norm": 0.6965057558910952,
+      "learning_rate": 1.8613120444390952e-07,
+      "loss": 0.3137,
+      "step": 11159
+    },
+    {
+      "epoch": 0.91561718012881,
+      "grad_norm": 0.7121474424237527,
+      "learning_rate": 1.857722138659407e-07,
+      "loss": 0.3672,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9156992246790007,
+      "grad_norm": 0.7304445228206803,
+      "learning_rate": 1.8541356326100436e-07,
+      "loss": 0.2866,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9157812692291915,
+      "grad_norm": 0.7228311666864288,
+      "learning_rate": 1.850552526544258e-07,
+      "loss": 0.309,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9158633137793822,
+      "grad_norm": 0.6861195223137565,
+      "learning_rate": 1.8469728207150982e-07,
+      "loss": 0.279,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9159453583295729,
+      "grad_norm": 0.7630807386893559,
+      "learning_rate": 1.8433965153753563e-07,
+      "loss": 0.2917,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9160274028797637,
+      "grad_norm": 0.753962437814923,
+      "learning_rate": 1.8398236107775802e-07,
+      "loss": 0.3243,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9161094474299545,
+      "grad_norm": 0.7292182178479899,
+      "learning_rate": 1.8362541071740846e-07,
+      "loss": 0.3038,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9161914919801453,
+      "grad_norm": 0.6704179740482594,
+      "learning_rate": 1.832688004816946e-07,
+      "loss": 0.3021,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9162735365303359,
+      "grad_norm": 0.6983363922874645,
+      "learning_rate": 1.8291253039579905e-07,
+      "loss": 0.2643,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9163555810805267,
+      "grad_norm": 0.8452411253206148,
+      "learning_rate": 1.8255660048488223e-07,
+      "loss": 0.2979,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9164376256307175,
+      "grad_norm": 0.8109625656939726,
+      "learning_rate": 1.8220101077407738e-07,
+      "loss": 0.3348,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9165196701809082,
+      "grad_norm": 0.7071663711852987,
+      "learning_rate": 1.8184576128849773e-07,
+      "loss": 0.2865,
+      "step": 11171
+    },
+    {
+      "epoch": 0.916601714731099,
+      "grad_norm": 0.6619586319323217,
+      "learning_rate": 1.814908520532288e-07,
+      "loss": 0.2674,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9166837592812898,
+      "grad_norm": 0.8184110438987262,
+      "learning_rate": 1.8113628309333497e-07,
+      "loss": 0.2723,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9167658038314805,
+      "grad_norm": 0.6975408489152973,
+      "learning_rate": 1.8078205443385456e-07,
+      "loss": 0.2919,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9168478483816712,
+      "grad_norm": 0.667828990298193,
+      "learning_rate": 1.804281660998025e-07,
+      "loss": 0.2725,
+      "step": 11175
+    },
+    {
+      "epoch": 0.916929892931862,
+      "grad_norm": 0.6990916261594601,
+      "learning_rate": 1.8007461811617e-07,
+      "loss": 0.3599,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9170119374820528,
+      "grad_norm": 0.6924810515610663,
+      "learning_rate": 1.797214105079248e-07,
+      "loss": 0.2979,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9170939820322435,
+      "grad_norm": 0.7043245378203861,
+      "learning_rate": 1.793685433000081e-07,
+      "loss": 0.3303,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9171760265824342,
+      "grad_norm": 0.7385129743358945,
+      "learning_rate": 1.7901601651734101e-07,
+      "loss": 0.3328,
+      "step": 11179
+    },
+    {
+      "epoch": 0.917258071132625,
+      "grad_norm": 0.6921634011960124,
+      "learning_rate": 1.7866383018481704e-07,
+      "loss": 0.2887,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9173401156828158,
+      "grad_norm": 0.6844293706559036,
+      "learning_rate": 1.7831198432730678e-07,
+      "loss": 0.2873,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9174221602330065,
+      "grad_norm": 0.6796220099640102,
+      "learning_rate": 1.7796047896965817e-07,
+      "loss": 0.2746,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9175042047831973,
+      "grad_norm": 0.6921495037304204,
+      "learning_rate": 1.7760931413669303e-07,
+      "loss": 0.2972,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9175862493333881,
+      "grad_norm": 0.7212800982725015,
+      "learning_rate": 1.7725848985321037e-07,
+      "loss": 0.3128,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9176682938835787,
+      "grad_norm": 0.6882332817040915,
+      "learning_rate": 1.7690800614398483e-07,
+      "loss": 0.2881,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9177503384337695,
+      "grad_norm": 0.6861585968057536,
+      "learning_rate": 1.7655786303376775e-07,
+      "loss": 0.3187,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9178323829839603,
+      "grad_norm": 0.6483065416102405,
+      "learning_rate": 1.7620806054728434e-07,
+      "loss": 0.2767,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9179144275341511,
+      "grad_norm": 0.6764015887033898,
+      "learning_rate": 1.758585987092376e-07,
+      "loss": 0.2947,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9179964720843418,
+      "grad_norm": 0.6933261171485027,
+      "learning_rate": 1.7550947754430615e-07,
+      "loss": 0.2881,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9180785166345325,
+      "grad_norm": 0.7525985161569397,
+      "learning_rate": 1.751606970771441e-07,
+      "loss": 0.3162,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9181605611847233,
+      "grad_norm": 0.716739042668065,
+      "learning_rate": 1.7481225733238184e-07,
+      "loss": 0.272,
+      "step": 11191
+    },
+    {
+      "epoch": 0.918242605734914,
+      "grad_norm": 0.7365274492026737,
+      "learning_rate": 1.7446415833462625e-07,
+      "loss": 0.2625,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9183246502851048,
+      "grad_norm": 0.8168158587579464,
+      "learning_rate": 1.7411640010845832e-07,
+      "loss": 0.3011,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9184066948352956,
+      "grad_norm": 0.6940720419308746,
+      "learning_rate": 1.7376898267843722e-07,
+      "loss": 0.2795,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9184887393854864,
+      "grad_norm": 0.8798386951359392,
+      "learning_rate": 1.734219060690967e-07,
+      "loss": 0.2928,
+      "step": 11195
+    },
+    {
+      "epoch": 0.918570783935677,
+      "grad_norm": 0.678240931096757,
+      "learning_rate": 1.730751703049466e-07,
+      "loss": 0.3227,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9186528284858678,
+      "grad_norm": 0.7234938157357826,
+      "learning_rate": 1.7272877541047284e-07,
+      "loss": 0.3271,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9187348730360586,
+      "grad_norm": 0.7028889109201336,
+      "learning_rate": 1.7238272141013702e-07,
+      "loss": 0.2736,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9188169175862493,
+      "grad_norm": 0.6874845792514387,
+      "learning_rate": 1.7203700832837844e-07,
+      "loss": 0.3238,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9188989621364401,
+      "grad_norm": 0.6628352257407423,
+      "learning_rate": 1.7169163618960817e-07,
+      "loss": 0.2907,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9189810066866309,
+      "grad_norm": 0.7002645128753517,
+      "learning_rate": 1.713466050182183e-07,
+      "loss": 0.3021,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9190630512368216,
+      "grad_norm": 0.8354052123605867,
+      "learning_rate": 1.7100191483857275e-07,
+      "loss": 0.299,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9191450957870123,
+      "grad_norm": 0.722330505346524,
+      "learning_rate": 1.7065756567501424e-07,
+      "loss": 0.3119,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9192271403372031,
+      "grad_norm": 0.6628553907140592,
+      "learning_rate": 1.7031355755185886e-07,
+      "loss": 0.2916,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9193091848873939,
+      "grad_norm": 0.5992450247589688,
+      "learning_rate": 1.6996989049340106e-07,
+      "loss": 0.7059,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9193912294375846,
+      "grad_norm": 0.7592193550353118,
+      "learning_rate": 1.6962656452390925e-07,
+      "loss": 0.3046,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9194732739877753,
+      "grad_norm": 0.7455404640776357,
+      "learning_rate": 1.6928357966762843e-07,
+      "loss": 0.3171,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9195553185379661,
+      "grad_norm": 0.6989159411021488,
+      "learning_rate": 1.689409359487809e-07,
+      "loss": 0.2875,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9196373630881569,
+      "grad_norm": 0.6905813105663163,
+      "learning_rate": 1.6859863339156235e-07,
+      "loss": 0.3047,
+      "step": 11209
+    },
+    {
+      "epoch": 0.9197194076383476,
+      "grad_norm": 0.7935923729944225,
+      "learning_rate": 1.6825667202014617e-07,
+      "loss": 0.2719,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9198014521885384,
+      "grad_norm": 0.7806284687107553,
+      "learning_rate": 1.6791505185868085e-07,
+      "loss": 0.2826,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9198834967387292,
+      "grad_norm": 0.6759202247344804,
+      "learning_rate": 1.675737729312915e-07,
+      "loss": 0.2872,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9199655412889198,
+      "grad_norm": 1.1288871029894665,
+      "learning_rate": 1.6723283526207833e-07,
+      "loss": 0.3049,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9200475858391106,
+      "grad_norm": 0.6856095783914257,
+      "learning_rate": 1.6689223887511818e-07,
+      "loss": 0.2898,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9201296303893014,
+      "grad_norm": 0.7153111065605168,
+      "learning_rate": 1.6655198379446347e-07,
+      "loss": 0.3139,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9202116749394922,
+      "grad_norm": 0.7191451634516014,
+      "learning_rate": 1.6621207004414108e-07,
+      "loss": 0.3033,
+      "step": 11216
+    },
+    {
+      "epoch": 0.9202937194896829,
+      "grad_norm": 0.719124035669612,
+      "learning_rate": 1.6587249764815628e-07,
+      "loss": 0.2821,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9203757640398736,
+      "grad_norm": 1.452160518598789,
+      "learning_rate": 1.6553326663048986e-07,
+      "loss": 0.3018,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9204578085900644,
+      "grad_norm": 0.9117506253789769,
+      "learning_rate": 1.6519437701509654e-07,
+      "loss": 0.3129,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9205398531402551,
+      "grad_norm": 0.680658085115046,
+      "learning_rate": 1.6485582882590833e-07,
+      "loss": 0.275,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9206218976904459,
+      "grad_norm": 0.7052165392069647,
+      "learning_rate": 1.6451762208683387e-07,
+      "loss": 0.3184,
+      "step": 11221
+    },
+    {
+      "epoch": 0.9207039422406367,
+      "grad_norm": 0.7901783793045781,
+      "learning_rate": 1.6417975682175568e-07,
+      "loss": 0.2786,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9207859867908275,
+      "grad_norm": 0.7054444668489732,
+      "learning_rate": 1.6384223305453417e-07,
+      "loss": 0.3001,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9208680313410181,
+      "grad_norm": 0.7214858180495974,
+      "learning_rate": 1.6350505080900358e-07,
+      "loss": 0.2887,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9209500758912089,
+      "grad_norm": 0.826739836995761,
+      "learning_rate": 1.63168210108976e-07,
+      "loss": 0.3368,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9210321204413997,
+      "grad_norm": 0.7318023373426352,
+      "learning_rate": 1.6283171097823903e-07,
+      "loss": 0.3034,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9211141649915904,
+      "grad_norm": 0.729646299806955,
+      "learning_rate": 1.6249555344055533e-07,
+      "loss": 0.3065,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9211962095417812,
+      "grad_norm": 0.6913356817962049,
+      "learning_rate": 1.6215973751966262e-07,
+      "loss": 0.3196,
+      "step": 11228
+    },
+    {
+      "epoch": 0.921278254091972,
+      "grad_norm": 0.6960751253149262,
+      "learning_rate": 1.6182426323927747e-07,
+      "loss": 0.279,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9213602986421627,
+      "grad_norm": 0.8618209069150877,
+      "learning_rate": 1.6148913062308923e-07,
+      "loss": 0.2934,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9214423431923534,
+      "grad_norm": 0.6406518716860364,
+      "learning_rate": 1.6115433969476567e-07,
+      "loss": 0.252,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9215243877425442,
+      "grad_norm": 0.6392102293422011,
+      "learning_rate": 1.6081989047794732e-07,
+      "loss": 0.2432,
+      "step": 11232
+    },
+    {
+      "epoch": 0.921606432292735,
+      "grad_norm": 0.836608443596888,
+      "learning_rate": 1.6048578299625472e-07,
+      "loss": 0.2976,
+      "step": 11233
+    },
+    {
+      "epoch": 0.9216884768429258,
+      "grad_norm": 0.6492510640608065,
+      "learning_rate": 1.6015201727328012e-07,
+      "loss": 0.2776,
+      "step": 11234
+    },
+    {
+      "epoch": 0.9217705213931164,
+      "grad_norm": 0.9783766342457163,
+      "learning_rate": 1.598185933325941e-07,
+      "loss": 0.2824,
+      "step": 11235
+    },
+    {
+      "epoch": 0.9218525659433072,
+      "grad_norm": 0.7894503230086999,
+      "learning_rate": 1.5948551119774336e-07,
+      "loss": 0.3565,
+      "step": 11236
+    },
+    {
+      "epoch": 0.921934610493498,
+      "grad_norm": 0.7539995594651171,
+      "learning_rate": 1.59152770892248e-07,
+      "loss": 0.279,
+      "step": 11237
+    },
+    {
+      "epoch": 0.9220166550436887,
+      "grad_norm": 0.6715661158807812,
+      "learning_rate": 1.5882037243960702e-07,
+      "loss": 0.326,
+      "step": 11238
+    },
+    {
+      "epoch": 0.9220986995938795,
+      "grad_norm": 0.9482750293672452,
+      "learning_rate": 1.5848831586329382e-07,
+      "loss": 0.3096,
+      "step": 11239
+    },
+    {
+      "epoch": 0.9221807441440703,
+      "grad_norm": 0.7705029835906967,
+      "learning_rate": 1.581566011867569e-07,
+      "loss": 0.315,
+      "step": 11240
+    },
+    {
+      "epoch": 0.922262788694261,
+      "grad_norm": 0.8560409064645318,
+      "learning_rate": 1.5782522843342197e-07,
+      "loss": 0.2815,
+      "step": 11241
+    },
+    {
+      "epoch": 0.9223448332444517,
+      "grad_norm": 0.7716120465924932,
+      "learning_rate": 1.574941976266897e-07,
+      "loss": 0.2798,
+      "step": 11242
+    },
+    {
+      "epoch": 0.9224268777946425,
+      "grad_norm": 0.7046781578762867,
+      "learning_rate": 1.5716350878993703e-07,
+      "loss": 0.3104,
+      "step": 11243
+    },
+    {
+      "epoch": 0.9225089223448333,
+      "grad_norm": 0.705683153091046,
+      "learning_rate": 1.5683316194651743e-07,
+      "loss": 0.2715,
+      "step": 11244
+    },
+    {
+      "epoch": 0.922590966895024,
+      "grad_norm": 0.7384033892575611,
+      "learning_rate": 1.5650315711975783e-07,
+      "loss": 0.3314,
+      "step": 11245
+    },
+    {
+      "epoch": 0.9226730114452147,
+      "grad_norm": 0.7079310372752333,
+      "learning_rate": 1.5617349433296402e-07,
+      "loss": 0.3598,
+      "step": 11246
+    },
+    {
+      "epoch": 0.9227550559954055,
+      "grad_norm": 0.7258826758902082,
+      "learning_rate": 1.5584417360941572e-07,
+      "loss": 0.3518,
+      "step": 11247
+    },
+    {
+      "epoch": 0.9228371005455963,
+      "grad_norm": 0.7216110467071996,
+      "learning_rate": 1.5551519497236877e-07,
+      "loss": 0.3026,
+      "step": 11248
+    },
+    {
+      "epoch": 0.922919145095787,
+      "grad_norm": 0.6977563838547306,
+      "learning_rate": 1.5518655844505627e-07,
+      "loss": 0.2899,
+      "step": 11249
+    },
+    {
+      "epoch": 0.9230011896459778,
+      "grad_norm": 0.703545816390553,
+      "learning_rate": 1.548582640506846e-07,
+      "loss": 0.2956,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9230832341961686,
+      "grad_norm": 0.6288482227595057,
+      "learning_rate": 1.5453031181243804e-07,
+      "loss": 0.3022,
+      "step": 11251
+    },
+    {
+      "epoch": 0.9231652787463592,
+      "grad_norm": 0.6555838979655301,
+      "learning_rate": 1.5420270175347696e-07,
+      "loss": 0.2765,
+      "step": 11252
+    },
+    {
+      "epoch": 0.92324732329655,
+      "grad_norm": 0.7514918857230738,
+      "learning_rate": 1.538754338969345e-07,
+      "loss": 0.2784,
+      "step": 11253
+    },
+    {
+      "epoch": 0.9233293678467408,
+      "grad_norm": 0.7172409743248405,
+      "learning_rate": 1.5354850826592383e-07,
+      "loss": 0.2793,
+      "step": 11254
+    },
+    {
+      "epoch": 0.9234114123969316,
+      "grad_norm": 0.7536980807505492,
+      "learning_rate": 1.532219248835315e-07,
+      "loss": 0.2637,
+      "step": 11255
+    },
+    {
+      "epoch": 0.9234934569471223,
+      "grad_norm": 1.4270163908946427,
+      "learning_rate": 1.5289568377281905e-07,
+      "loss": 0.2922,
+      "step": 11256
+    },
+    {
+      "epoch": 0.923575501497313,
+      "grad_norm": 0.6829870913262469,
+      "learning_rate": 1.525697849568264e-07,
+      "loss": 0.3226,
+      "step": 11257
+    },
+    {
+      "epoch": 0.9236575460475038,
+      "grad_norm": 0.7590322842287303,
+      "learning_rate": 1.5224422845856678e-07,
+      "loss": 0.3126,
+      "step": 11258
+    },
+    {
+      "epoch": 0.9237395905976945,
+      "grad_norm": 0.68215068527405,
+      "learning_rate": 1.5191901430103185e-07,
+      "loss": 0.2657,
+      "step": 11259
+    },
+    {
+      "epoch": 0.9238216351478853,
+      "grad_norm": 0.8290426973810083,
+      "learning_rate": 1.5159414250718706e-07,
+      "loss": 0.3424,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9239036796980761,
+      "grad_norm": 0.7024612883770815,
+      "learning_rate": 1.5126961309997355e-07,
+      "loss": 0.2938,
+      "step": 11261
+    },
+    {
+      "epoch": 0.9239857242482669,
+      "grad_norm": 0.7086143270946227,
+      "learning_rate": 1.5094542610231133e-07,
+      "loss": 0.2898,
+      "step": 11262
+    },
+    {
+      "epoch": 0.9240677687984575,
+      "grad_norm": 0.6406344220471059,
+      "learning_rate": 1.506215815370915e-07,
+      "loss": 0.3182,
+      "step": 11263
+    },
+    {
+      "epoch": 0.9241498133486483,
+      "grad_norm": 0.7933529067986125,
+      "learning_rate": 1.5029807942718466e-07,
+      "loss": 0.2859,
+      "step": 11264
+    },
+    {
+      "epoch": 0.9242318578988391,
+      "grad_norm": 0.794534790606814,
+      "learning_rate": 1.4997491979543587e-07,
+      "loss": 0.3222,
+      "step": 11265
+    },
+    {
+      "epoch": 0.9243139024490298,
+      "grad_norm": 0.6750508994025037,
+      "learning_rate": 1.496521026646658e-07,
+      "loss": 0.3196,
+      "step": 11266
+    },
+    {
+      "epoch": 0.9243959469992206,
+      "grad_norm": 0.7593303539514339,
+      "learning_rate": 1.4932962805767225e-07,
+      "loss": 0.2977,
+      "step": 11267
+    },
+    {
+      "epoch": 0.9244779915494113,
+      "grad_norm": 0.7179146474938451,
+      "learning_rate": 1.4900749599722653e-07,
+      "loss": 0.2873,
+      "step": 11268
+    },
+    {
+      "epoch": 0.9245600360996021,
+      "grad_norm": 0.7850343304713295,
+      "learning_rate": 1.4868570650607816e-07,
+      "loss": 0.2741,
+      "step": 11269
+    },
+    {
+      "epoch": 0.9246420806497928,
+      "grad_norm": 0.6978921782379626,
+      "learning_rate": 1.4836425960695123e-07,
+      "loss": 0.2909,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9247241251999836,
+      "grad_norm": 0.6979728814980153,
+      "learning_rate": 1.4804315532254475e-07,
+      "loss": 0.3262,
+      "step": 11271
+    },
+    {
+      "epoch": 0.9248061697501744,
+      "grad_norm": 0.6985883908306517,
+      "learning_rate": 1.4772239367553564e-07,
+      "loss": 0.2944,
+      "step": 11272
+    },
+    {
+      "epoch": 0.924888214300365,
+      "grad_norm": 0.6659904539890675,
+      "learning_rate": 1.474019746885752e-07,
+      "loss": 0.2947,
+      "step": 11273
+    },
+    {
+      "epoch": 0.9249702588505558,
+      "grad_norm": 0.6817429909123156,
+      "learning_rate": 1.4708189838429087e-07,
+      "loss": 0.3079,
+      "step": 11274
+    },
+    {
+      "epoch": 0.9250523034007466,
+      "grad_norm": 0.7549974121390611,
+      "learning_rate": 1.4676216478528683e-07,
+      "loss": 0.3075,
+      "step": 11275
+    },
+    {
+      "epoch": 0.9251343479509374,
+      "grad_norm": 0.9295201700457826,
+      "learning_rate": 1.4644277391414107e-07,
+      "loss": 0.2986,
+      "step": 11276
+    },
+    {
+      "epoch": 0.9252163925011281,
+      "grad_norm": 0.718340537534761,
+      "learning_rate": 1.4612372579340893e-07,
+      "loss": 0.2789,
+      "step": 11277
+    },
+    {
+      "epoch": 0.9252984370513189,
+      "grad_norm": 0.659254255382194,
+      "learning_rate": 1.4580502044562016e-07,
+      "loss": 0.2897,
+      "step": 11278
+    },
+    {
+      "epoch": 0.9253804816015097,
+      "grad_norm": 0.797120424569661,
+      "learning_rate": 1.454866578932823e-07,
+      "loss": 0.3121,
+      "step": 11279
+    },
+    {
+      "epoch": 0.9254625261517003,
+      "grad_norm": 0.6847949039391912,
+      "learning_rate": 1.4516863815887795e-07,
+      "loss": 0.2828,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9255445707018911,
+      "grad_norm": 0.67594605535415,
+      "learning_rate": 1.4485096126486355e-07,
+      "loss": 0.3091,
+      "step": 11281
+    },
+    {
+      "epoch": 0.9256266152520819,
+      "grad_norm": 0.6732869434447714,
+      "learning_rate": 1.4453362723367449e-07,
+      "loss": 0.271,
+      "step": 11282
+    },
+    {
+      "epoch": 0.9257086598022727,
+      "grad_norm": 0.7183071249117103,
+      "learning_rate": 1.4421663608772009e-07,
+      "loss": 0.3131,
+      "step": 11283
+    },
+    {
+      "epoch": 0.9257907043524634,
+      "grad_norm": 0.6752971073416759,
+      "learning_rate": 1.4389998784938465e-07,
+      "loss": 0.2573,
+      "step": 11284
+    },
+    {
+      "epoch": 0.9258727489026541,
+      "grad_norm": 0.6944810955715095,
+      "learning_rate": 1.4358368254103084e-07,
+      "loss": 0.2857,
+      "step": 11285
+    },
+    {
+      "epoch": 0.9259547934528449,
+      "grad_norm": 0.724692445678581,
+      "learning_rate": 1.4326772018499414e-07,
+      "loss": 0.3511,
+      "step": 11286
+    },
+    {
+      "epoch": 0.9260368380030356,
+      "grad_norm": 0.7117560893041518,
+      "learning_rate": 1.4295210080358835e-07,
+      "loss": 0.2837,
+      "step": 11287
+    },
+    {
+      "epoch": 0.9261188825532264,
+      "grad_norm": 0.7005462828353944,
+      "learning_rate": 1.4263682441910233e-07,
+      "loss": 0.2705,
+      "step": 11288
+    },
+    {
+      "epoch": 0.9262009271034172,
+      "grad_norm": 0.8498327730214074,
+      "learning_rate": 1.4232189105379935e-07,
+      "loss": 0.3493,
+      "step": 11289
+    },
+    {
+      "epoch": 0.926282971653608,
+      "grad_norm": 0.684546161141997,
+      "learning_rate": 1.4200730072991996e-07,
+      "loss": 0.2779,
+      "step": 11290
+    },
+    {
+      "epoch": 0.9263650162037986,
+      "grad_norm": 0.7368485610691786,
+      "learning_rate": 1.4169305346968033e-07,
+      "loss": 0.3232,
+      "step": 11291
+    },
+    {
+      "epoch": 0.9264470607539894,
+      "grad_norm": 0.6715791684176645,
+      "learning_rate": 1.4137914929527097e-07,
+      "loss": 0.2545,
+      "step": 11292
+    },
+    {
+      "epoch": 0.9265291053041802,
+      "grad_norm": 0.6863598838757133,
+      "learning_rate": 1.410655882288614e-07,
+      "loss": 0.2502,
+      "step": 11293
+    },
+    {
+      "epoch": 0.9266111498543709,
+      "grad_norm": 0.8010982568166047,
+      "learning_rate": 1.4075237029259227e-07,
+      "loss": 0.3092,
+      "step": 11294
+    },
+    {
+      "epoch": 0.9266931944045617,
+      "grad_norm": 0.6705657571860355,
+      "learning_rate": 1.404394955085847e-07,
+      "loss": 0.2874,
+      "step": 11295
+    },
+    {
+      "epoch": 0.9267752389547524,
+      "grad_norm": 0.7316853808713365,
+      "learning_rate": 1.401269638989322e-07,
+      "loss": 0.2775,
+      "step": 11296
+    },
+    {
+      "epoch": 0.9268572835049432,
+      "grad_norm": 0.7139781543064865,
+      "learning_rate": 1.3981477548570544e-07,
+      "loss": 0.2723,
+      "step": 11297
+    },
+    {
+      "epoch": 0.9269393280551339,
+      "grad_norm": 0.7314803747348139,
+      "learning_rate": 1.395029302909512e-07,
+      "loss": 0.3046,
+      "step": 11298
+    },
+    {
+      "epoch": 0.9270213726053247,
+      "grad_norm": 0.7283965355999754,
+      "learning_rate": 1.391914283366902e-07,
+      "loss": 0.2616,
+      "step": 11299
+    },
+    {
+      "epoch": 0.9271034171555155,
+      "grad_norm": 0.6958124189152547,
+      "learning_rate": 1.3888026964492208e-07,
+      "loss": 0.2656,
+      "step": 11300
+    },
+    {
+      "epoch": 0.9271854617057061,
+      "grad_norm": 0.7295573936436791,
+      "learning_rate": 1.3856945423761813e-07,
+      "loss": 0.3163,
+      "step": 11301
+    },
+    {
+      "epoch": 0.9272675062558969,
+      "grad_norm": 0.763953193318596,
+      "learning_rate": 1.3825898213672918e-07,
+      "loss": 0.3275,
+      "step": 11302
+    },
+    {
+      "epoch": 0.9273495508060877,
+      "grad_norm": 0.82011987365467,
+      "learning_rate": 1.379488533641804e-07,
+      "loss": 0.3109,
+      "step": 11303
+    },
+    {
+      "epoch": 0.9274315953562785,
+      "grad_norm": 0.7064057225881357,
+      "learning_rate": 1.376390679418721e-07,
+      "loss": 0.2938,
+      "step": 11304
+    },
+    {
+      "epoch": 0.9275136399064692,
+      "grad_norm": 0.593066555050762,
+      "learning_rate": 1.373296258916801e-07,
+      "loss": 0.6324,
+      "step": 11305
+    },
+    {
+      "epoch": 0.92759568445666,
+      "grad_norm": 0.7032499203370332,
+      "learning_rate": 1.3702052723545856e-07,
+      "loss": 0.2559,
+      "step": 11306
+    },
+    {
+      "epoch": 0.9276777290068507,
+      "grad_norm": 0.6756299047239213,
+      "learning_rate": 1.3671177199503338e-07,
+      "loss": 0.2982,
+      "step": 11307
+    },
+    {
+      "epoch": 0.9277597735570414,
+      "grad_norm": 0.6210576385567462,
+      "learning_rate": 1.364033601922099e-07,
+      "loss": 0.6589,
+      "step": 11308
+    },
+    {
+      "epoch": 0.9278418181072322,
+      "grad_norm": 0.6989787731096335,
+      "learning_rate": 1.360952918487668e-07,
+      "loss": 0.2986,
+      "step": 11309
+    },
+    {
+      "epoch": 0.927923862657423,
+      "grad_norm": 0.5604770703059284,
+      "learning_rate": 1.3578756698646056e-07,
+      "loss": 0.6251,
+      "step": 11310
+    },
+    {
+      "epoch": 0.9280059072076138,
+      "grad_norm": 0.7911659236298193,
+      "learning_rate": 1.3548018562702104e-07,
+      "loss": 0.33,
+      "step": 11311
+    },
+    {
+      "epoch": 0.9280879517578045,
+      "grad_norm": 0.6867741797946453,
+      "learning_rate": 1.3517314779215473e-07,
+      "loss": 0.3322,
+      "step": 11312
+    },
+    {
+      "epoch": 0.9281699963079952,
+      "grad_norm": 0.7257845619484413,
+      "learning_rate": 1.3486645350354599e-07,
+      "loss": 0.284,
+      "step": 11313
+    },
+    {
+      "epoch": 0.928252040858186,
+      "grad_norm": 0.6726026381122251,
+      "learning_rate": 1.345601027828508e-07,
+      "loss": 0.2507,
+      "step": 11314
+    },
+    {
+      "epoch": 0.9283340854083767,
+      "grad_norm": 0.6951390267442393,
+      "learning_rate": 1.342540956517041e-07,
+      "loss": 0.2973,
+      "step": 11315
+    },
+    {
+      "epoch": 0.9284161299585675,
+      "grad_norm": 1.1983671059225978,
+      "learning_rate": 1.3394843213171638e-07,
+      "loss": 0.2904,
+      "step": 11316
+    },
+    {
+      "epoch": 0.9284981745087583,
+      "grad_norm": 0.7013899192290893,
+      "learning_rate": 1.33643112244472e-07,
+      "loss": 0.3144,
+      "step": 11317
+    },
+    {
+      "epoch": 0.928580219058949,
+      "grad_norm": 0.7720146195306696,
+      "learning_rate": 1.3333813601153322e-07,
+      "loss": 0.2923,
+      "step": 11318
+    },
+    {
+      "epoch": 0.9286622636091397,
+      "grad_norm": 0.6717292895047386,
+      "learning_rate": 1.330335034544361e-07,
+      "loss": 0.2968,
+      "step": 11319
+    },
+    {
+      "epoch": 0.9287443081593305,
+      "grad_norm": 0.6662895003714538,
+      "learning_rate": 1.3272921459469345e-07,
+      "loss": 0.2755,
+      "step": 11320
+    },
+    {
+      "epoch": 0.9288263527095213,
+      "grad_norm": 0.7679706841656804,
+      "learning_rate": 1.3242526945379363e-07,
+      "loss": 0.2975,
+      "step": 11321
+    },
+    {
+      "epoch": 0.928908397259712,
+      "grad_norm": 0.7331630502770061,
+      "learning_rate": 1.3212166805320114e-07,
+      "loss": 0.2891,
+      "step": 11322
+    },
+    {
+      "epoch": 0.9289904418099028,
+      "grad_norm": 0.6796957127717965,
+      "learning_rate": 1.3181841041435605e-07,
+      "loss": 0.315,
+      "step": 11323
+    },
+    {
+      "epoch": 0.9290724863600935,
+      "grad_norm": 0.7261566472812229,
+      "learning_rate": 1.315154965586729e-07,
+      "loss": 0.3208,
+      "step": 11324
+    },
+    {
+      "epoch": 0.9291545309102843,
+      "grad_norm": 0.7273980639672608,
+      "learning_rate": 1.3121292650754402e-07,
+      "loss": 0.3047,
+      "step": 11325
+    },
+    {
+      "epoch": 0.929236575460475,
+      "grad_norm": 0.6963380002685616,
+      "learning_rate": 1.3091070028233622e-07,
+      "loss": 0.315,
+      "step": 11326
+    },
+    {
+      "epoch": 0.9293186200106658,
+      "grad_norm": 0.7288534285840331,
+      "learning_rate": 1.306088179043913e-07,
+      "loss": 0.3001,
+      "step": 11327
+    },
+    {
+      "epoch": 0.9294006645608566,
+      "grad_norm": 0.7935116915332042,
+      "learning_rate": 1.303072793950283e-07,
+      "loss": 0.3039,
+      "step": 11328
+    },
+    {
+      "epoch": 0.9294827091110472,
+      "grad_norm": 0.7171768424513504,
+      "learning_rate": 1.3000608477554188e-07,
+      "loss": 0.3166,
+      "step": 11329
+    },
+    {
+      "epoch": 0.929564753661238,
+      "grad_norm": 0.7153761866589071,
+      "learning_rate": 1.2970523406720114e-07,
+      "loss": 0.2849,
+      "step": 11330
+    },
+    {
+      "epoch": 0.9296467982114288,
+      "grad_norm": 0.6922826824402796,
+      "learning_rate": 1.2940472729125241e-07,
+      "loss": 0.2641,
+      "step": 11331
+    },
+    {
+      "epoch": 0.9297288427616196,
+      "grad_norm": 0.6696058243899249,
+      "learning_rate": 1.291045644689165e-07,
+      "loss": 0.3188,
+      "step": 11332
+    },
+    {
+      "epoch": 0.9298108873118103,
+      "grad_norm": 0.7074256875536896,
+      "learning_rate": 1.2880474562139033e-07,
+      "loss": 0.3465,
+      "step": 11333
+    },
+    {
+      "epoch": 0.9298929318620011,
+      "grad_norm": 0.6877398087876817,
+      "learning_rate": 1.28505270769847e-07,
+      "loss": 0.2549,
+      "step": 11334
+    },
+    {
+      "epoch": 0.9299749764121918,
+      "grad_norm": 0.7149710787690073,
+      "learning_rate": 1.2820613993543452e-07,
+      "loss": 0.2629,
+      "step": 11335
+    },
+    {
+      "epoch": 0.9300570209623825,
+      "grad_norm": 0.7229540418757125,
+      "learning_rate": 1.2790735313927825e-07,
+      "loss": 0.2976,
+      "step": 11336
+    },
+    {
+      "epoch": 0.9301390655125733,
+      "grad_norm": 0.6982596910270458,
+      "learning_rate": 1.2760891040247637e-07,
+      "loss": 0.2569,
+      "step": 11337
+    },
+    {
+      "epoch": 0.9302211100627641,
+      "grad_norm": 0.6272067417141023,
+      "learning_rate": 1.2731081174610526e-07,
+      "loss": 0.2692,
+      "step": 11338
+    },
+    {
+      "epoch": 0.9303031546129549,
+      "grad_norm": 0.6975718961443422,
+      "learning_rate": 1.270130571912165e-07,
+      "loss": 0.3427,
+      "step": 11339
+    },
+    {
+      "epoch": 0.9303851991631455,
+      "grad_norm": 0.6826704821554437,
+      "learning_rate": 1.2671564675883595e-07,
+      "loss": 0.309,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9304672437133363,
+      "grad_norm": 0.6865336270516733,
+      "learning_rate": 1.2641858046996636e-07,
+      "loss": 0.26,
+      "step": 11341
+    },
+    {
+      "epoch": 0.9305492882635271,
+      "grad_norm": 0.7539957717422371,
+      "learning_rate": 1.261218583455881e-07,
+      "loss": 0.3107,
+      "step": 11342
+    },
+    {
+      "epoch": 0.9306313328137179,
+      "grad_norm": 0.7005539685016577,
+      "learning_rate": 1.2582548040665276e-07,
+      "loss": 0.3072,
+      "step": 11343
+    },
+    {
+      "epoch": 0.9307133773639086,
+      "grad_norm": 0.6049307188471116,
+      "learning_rate": 1.2552944667409139e-07,
+      "loss": 0.2712,
+      "step": 11344
+    },
+    {
+      "epoch": 0.9307954219140994,
+      "grad_norm": 0.692870470887097,
+      "learning_rate": 1.2523375716880836e-07,
+      "loss": 0.3012,
+      "step": 11345
+    },
+    {
+      "epoch": 0.9308774664642901,
+      "grad_norm": 0.728680672057194,
+      "learning_rate": 1.2493841191168587e-07,
+      "loss": 0.2719,
+      "step": 11346
+    },
+    {
+      "epoch": 0.9309595110144808,
+      "grad_norm": 0.6899178005010467,
+      "learning_rate": 1.2464341092358113e-07,
+      "loss": 0.3029,
+      "step": 11347
+    },
+    {
+      "epoch": 0.9310415555646716,
+      "grad_norm": 0.7627259455355251,
+      "learning_rate": 1.2434875422532467e-07,
+      "loss": 0.3118,
+      "step": 11348
+    },
+    {
+      "epoch": 0.9311236001148624,
+      "grad_norm": 0.7068565581650353,
+      "learning_rate": 1.2405444183772654e-07,
+      "loss": 0.3078,
+      "step": 11349
+    },
+    {
+      "epoch": 0.9312056446650532,
+      "grad_norm": 0.6601198203041769,
+      "learning_rate": 1.2376047378156952e-07,
+      "loss": 0.2544,
+      "step": 11350
+    },
+    {
+      "epoch": 0.9312876892152439,
+      "grad_norm": 0.7434556273183656,
+      "learning_rate": 1.2346685007761428e-07,
+      "loss": 0.2985,
+      "step": 11351
+    },
+    {
+      "epoch": 0.9313697337654346,
+      "grad_norm": 0.6686697344028023,
+      "learning_rate": 1.2317357074659476e-07,
+      "loss": 0.3105,
+      "step": 11352
+    },
+    {
+      "epoch": 0.9314517783156254,
+      "grad_norm": 0.7620720448654159,
+      "learning_rate": 1.2288063580922272e-07,
+      "loss": 0.2637,
+      "step": 11353
+    },
+    {
+      "epoch": 0.9315338228658161,
+      "grad_norm": 0.9563181249997663,
+      "learning_rate": 1.2258804528618495e-07,
+      "loss": 0.3004,
+      "step": 11354
+    },
+    {
+      "epoch": 0.9316158674160069,
+      "grad_norm": 1.1432227354352245,
+      "learning_rate": 1.2229579919814217e-07,
+      "loss": 0.2715,
+      "step": 11355
+    },
+    {
+      "epoch": 0.9316979119661977,
+      "grad_norm": 0.7305368604705774,
+      "learning_rate": 1.2200389756573395e-07,
+      "loss": 0.2876,
+      "step": 11356
+    },
+    {
+      "epoch": 0.9317799565163885,
+      "grad_norm": 0.7160069013692295,
+      "learning_rate": 1.2171234040957324e-07,
+      "loss": 0.2401,
+      "step": 11357
+    },
+    {
+      "epoch": 0.9318620010665791,
+      "grad_norm": 0.7472259082719492,
+      "learning_rate": 1.2142112775024972e-07,
+      "loss": 0.3396,
+      "step": 11358
+    },
+    {
+      "epoch": 0.9319440456167699,
+      "grad_norm": 0.673307116597146,
+      "learning_rate": 1.2113025960832748e-07,
+      "loss": 0.3047,
+      "step": 11359
+    },
+    {
+      "epoch": 0.9320260901669607,
+      "grad_norm": 0.719249184461802,
+      "learning_rate": 1.2083973600434894e-07,
+      "loss": 0.2928,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9321081347171514,
+      "grad_norm": 0.6908989841745159,
+      "learning_rate": 1.205495569588283e-07,
+      "loss": 0.289,
+      "step": 11361
+    },
+    {
+      "epoch": 0.9321901792673422,
+      "grad_norm": 0.699792260463337,
+      "learning_rate": 1.2025972249225914e-07,
+      "loss": 0.2945,
+      "step": 11362
+    },
+    {
+      "epoch": 0.9322722238175329,
+      "grad_norm": 0.6724413265549596,
+      "learning_rate": 1.1997023262510788e-07,
+      "loss": 0.2979,
+      "step": 11363
+    },
+    {
+      "epoch": 0.9323542683677237,
+      "grad_norm": 1.2135802262445003,
+      "learning_rate": 1.196810873778187e-07,
+      "loss": 0.3047,
+      "step": 11364
+    },
+    {
+      "epoch": 0.9324363129179144,
+      "grad_norm": 0.7713717069788307,
+      "learning_rate": 1.1939228677081027e-07,
+      "loss": 0.3267,
+      "step": 11365
+    },
+    {
+      "epoch": 0.9325183574681052,
+      "grad_norm": 0.7822713932341598,
+      "learning_rate": 1.1910383082447797e-07,
+      "loss": 0.2718,
+      "step": 11366
+    },
+    {
+      "epoch": 0.932600402018296,
+      "grad_norm": 0.6675976532896847,
+      "learning_rate": 1.1881571955919102e-07,
+      "loss": 0.307,
+      "step": 11367
+    },
+    {
+      "epoch": 0.9326824465684866,
+      "grad_norm": 0.6540416551397882,
+      "learning_rate": 1.1852795299529485e-07,
+      "loss": 0.2597,
+      "step": 11368
+    },
+    {
+      "epoch": 0.9327644911186774,
+      "grad_norm": 0.7188410466220364,
+      "learning_rate": 1.1824053115311263e-07,
+      "loss": 0.2947,
+      "step": 11369
+    },
+    {
+      "epoch": 0.9328465356688682,
+      "grad_norm": 0.6572633976137058,
+      "learning_rate": 1.179534540529409e-07,
+      "loss": 0.2798,
+      "step": 11370
+    },
+    {
+      "epoch": 0.932928580219059,
+      "grad_norm": 0.7203672925118564,
+      "learning_rate": 1.1766672171505289e-07,
+      "loss": 0.2835,
+      "step": 11371
+    },
+    {
+      "epoch": 0.9330106247692497,
+      "grad_norm": 0.6943549169379132,
+      "learning_rate": 1.1738033415969686e-07,
+      "loss": 0.3188,
+      "step": 11372
+    },
+    {
+      "epoch": 0.9330926693194405,
+      "grad_norm": 0.7696880228382366,
+      "learning_rate": 1.1709429140709661e-07,
+      "loss": 0.2835,
+      "step": 11373
+    },
+    {
+      "epoch": 0.9331747138696312,
+      "grad_norm": 0.6965586799562371,
+      "learning_rate": 1.1680859347745266e-07,
+      "loss": 0.3169,
+      "step": 11374
+    },
+    {
+      "epoch": 0.9332567584198219,
+      "grad_norm": 0.6910341328217541,
+      "learning_rate": 1.1652324039094054e-07,
+      "loss": 0.2869,
+      "step": 11375
+    },
+    {
+      "epoch": 0.9333388029700127,
+      "grad_norm": 0.82551054963208,
+      "learning_rate": 1.1623823216771135e-07,
+      "loss": 0.2889,
+      "step": 11376
+    },
+    {
+      "epoch": 0.9334208475202035,
+      "grad_norm": 0.7152022241461462,
+      "learning_rate": 1.1595356882789122e-07,
+      "loss": 0.2892,
+      "step": 11377
+    },
+    {
+      "epoch": 0.9335028920703943,
+      "grad_norm": 0.6126444972274933,
+      "learning_rate": 1.1566925039158405e-07,
+      "loss": 0.3125,
+      "step": 11378
+    },
+    {
+      "epoch": 0.933584936620585,
+      "grad_norm": 0.7202586611361574,
+      "learning_rate": 1.1538527687886714e-07,
+      "loss": 0.2835,
+      "step": 11379
+    },
+    {
+      "epoch": 0.9336669811707757,
+      "grad_norm": 0.8195151975378883,
+      "learning_rate": 1.1510164830979387e-07,
+      "loss": 0.2519,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9337490257209665,
+      "grad_norm": 0.7090745014823181,
+      "learning_rate": 1.1481836470439433e-07,
+      "loss": 0.2732,
+      "step": 11381
+    },
+    {
+      "epoch": 0.9338310702711572,
+      "grad_norm": 0.6423483536903929,
+      "learning_rate": 1.145354260826731e-07,
+      "loss": 0.2937,
+      "step": 11382
+    },
+    {
+      "epoch": 0.933913114821348,
+      "grad_norm": 0.6896616274033623,
+      "learning_rate": 1.1425283246461083e-07,
+      "loss": 0.3156,
+      "step": 11383
+    },
+    {
+      "epoch": 0.9339951593715388,
+      "grad_norm": 0.6448863618577595,
+      "learning_rate": 1.139705838701638e-07,
+      "loss": 0.3098,
+      "step": 11384
+    },
+    {
+      "epoch": 0.9340772039217295,
+      "grad_norm": 0.7789249096004623,
+      "learning_rate": 1.1368868031926494e-07,
+      "loss": 0.2751,
+      "step": 11385
+    },
+    {
+      "epoch": 0.9341592484719202,
+      "grad_norm": 0.6936033014838413,
+      "learning_rate": 1.1340712183182002e-07,
+      "loss": 0.2917,
+      "step": 11386
+    },
+    {
+      "epoch": 0.934241293022111,
+      "grad_norm": 0.7659509136121095,
+      "learning_rate": 1.1312590842771365e-07,
+      "loss": 0.2998,
+      "step": 11387
+    },
+    {
+      "epoch": 0.9343233375723018,
+      "grad_norm": 0.7350386997255511,
+      "learning_rate": 1.1284504012680442e-07,
+      "loss": 0.3175,
+      "step": 11388
+    },
+    {
+      "epoch": 0.9344053821224925,
+      "grad_norm": 0.8958812115886376,
+      "learning_rate": 1.1256451694892645e-07,
+      "loss": 0.2637,
+      "step": 11389
+    },
+    {
+      "epoch": 0.9344874266726833,
+      "grad_norm": 0.7508145405763,
+      "learning_rate": 1.1228433891389056e-07,
+      "loss": 0.2717,
+      "step": 11390
+    },
+    {
+      "epoch": 0.934569471222874,
+      "grad_norm": 0.6647608854048891,
+      "learning_rate": 1.1200450604148095e-07,
+      "loss": 0.2851,
+      "step": 11391
+    },
+    {
+      "epoch": 0.9346515157730648,
+      "grad_norm": 0.66358182941827,
+      "learning_rate": 1.1172501835146065e-07,
+      "loss": 0.3381,
+      "step": 11392
+    },
+    {
+      "epoch": 0.9347335603232555,
+      "grad_norm": 0.740793170584464,
+      "learning_rate": 1.1144587586356614e-07,
+      "loss": 0.2757,
+      "step": 11393
+    },
+    {
+      "epoch": 0.9348156048734463,
+      "grad_norm": 0.6512646350553803,
+      "learning_rate": 1.1116707859750941e-07,
+      "loss": 0.2922,
+      "step": 11394
+    },
+    {
+      "epoch": 0.9348976494236371,
+      "grad_norm": 0.6850469949552235,
+      "learning_rate": 1.1088862657297916e-07,
+      "loss": 0.2677,
+      "step": 11395
+    },
+    {
+      "epoch": 0.9349796939738277,
+      "grad_norm": 0.7005914661754405,
+      "learning_rate": 1.106105198096391e-07,
+      "loss": 0.3162,
+      "step": 11396
+    },
+    {
+      "epoch": 0.9350617385240185,
+      "grad_norm": 0.7571390339721051,
+      "learning_rate": 1.1033275832712853e-07,
+      "loss": 0.2827,
+      "step": 11397
+    },
+    {
+      "epoch": 0.9351437830742093,
+      "grad_norm": 0.7069939723087038,
+      "learning_rate": 1.1005534214506342e-07,
+      "loss": 0.2982,
+      "step": 11398
+    },
+    {
+      "epoch": 0.9352258276244001,
+      "grad_norm": 0.6404881203116829,
+      "learning_rate": 1.0977827128303309e-07,
+      "loss": 0.3063,
+      "step": 11399
+    },
+    {
+      "epoch": 0.9353078721745908,
+      "grad_norm": 0.723347912509273,
+      "learning_rate": 1.0950154576060413e-07,
+      "loss": 0.2818,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9353899167247816,
+      "grad_norm": 0.7542077419544452,
+      "learning_rate": 1.0922516559731978e-07,
+      "loss": 0.2668,
+      "step": 11401
+    },
+    {
+      "epoch": 0.9354719612749723,
+      "grad_norm": 0.7177162248985793,
+      "learning_rate": 1.0894913081269608e-07,
+      "loss": 0.3269,
+      "step": 11402
+    },
+    {
+      "epoch": 0.935554005825163,
+      "grad_norm": 0.6966283999230296,
+      "learning_rate": 1.0867344142622693e-07,
+      "loss": 0.3109,
+      "step": 11403
+    },
+    {
+      "epoch": 0.9356360503753538,
+      "grad_norm": 0.7108086248359303,
+      "learning_rate": 1.0839809745738061e-07,
+      "loss": 0.3031,
+      "step": 11404
+    },
+    {
+      "epoch": 0.9357180949255446,
+      "grad_norm": 0.756789845449846,
+      "learning_rate": 1.0812309892560157e-07,
+      "loss": 0.2796,
+      "step": 11405
+    },
+    {
+      "epoch": 0.9358001394757354,
+      "grad_norm": 0.6312500303703533,
+      "learning_rate": 1.078484458503104e-07,
+      "loss": 0.3226,
+      "step": 11406
+    },
+    {
+      "epoch": 0.935882184025926,
+      "grad_norm": 0.6705309153607277,
+      "learning_rate": 1.0757413825090212e-07,
+      "loss": 0.2677,
+      "step": 11407
+    },
+    {
+      "epoch": 0.9359642285761168,
+      "grad_norm": 0.7069090425686048,
+      "learning_rate": 1.0730017614674736e-07,
+      "loss": 0.2916,
+      "step": 11408
+    },
+    {
+      "epoch": 0.9360462731263076,
+      "grad_norm": 0.7021047243998828,
+      "learning_rate": 1.07026559557194e-07,
+      "loss": 0.3257,
+      "step": 11409
+    },
+    {
+      "epoch": 0.9361283176764983,
+      "grad_norm": 0.6482833492790715,
+      "learning_rate": 1.0675328850156319e-07,
+      "loss": 0.2897,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9362103622266891,
+      "grad_norm": 1.0958555761546935,
+      "learning_rate": 1.0648036299915399e-07,
+      "loss": 0.3561,
+      "step": 11411
+    },
+    {
+      "epoch": 0.9362924067768799,
+      "grad_norm": 1.3956406613874863,
+      "learning_rate": 1.0620778306923929e-07,
+      "loss": 0.3077,
+      "step": 11412
+    },
+    {
+      "epoch": 0.9363744513270706,
+      "grad_norm": 0.7270396975116507,
+      "learning_rate": 1.059355487310687e-07,
+      "loss": 0.3324,
+      "step": 11413
+    },
+    {
+      "epoch": 0.9364564958772613,
+      "grad_norm": 0.6539490933305117,
+      "learning_rate": 1.0566366000386685e-07,
+      "loss": 0.2647,
+      "step": 11414
+    },
+    {
+      "epoch": 0.9365385404274521,
+      "grad_norm": 0.7342834202437024,
+      "learning_rate": 1.0539211690683337e-07,
+      "loss": 0.2955,
+      "step": 11415
+    },
+    {
+      "epoch": 0.9366205849776429,
+      "grad_norm": 0.6738348399562187,
+      "learning_rate": 1.0512091945914571e-07,
+      "loss": 0.2851,
+      "step": 11416
+    },
+    {
+      "epoch": 0.9367026295278336,
+      "grad_norm": 0.6454627459951423,
+      "learning_rate": 1.0485006767995354e-07,
+      "loss": 0.3022,
+      "step": 11417
+    },
+    {
+      "epoch": 0.9367846740780243,
+      "grad_norm": 0.6810733174296697,
+      "learning_rate": 1.0457956158838545e-07,
+      "loss": 0.3342,
+      "step": 11418
+    },
+    {
+      "epoch": 0.9368667186282151,
+      "grad_norm": 0.7096071201988584,
+      "learning_rate": 1.0430940120354393e-07,
+      "loss": 0.3137,
+      "step": 11419
+    },
+    {
+      "epoch": 0.9369487631784059,
+      "grad_norm": 0.7091124758561269,
+      "learning_rate": 1.0403958654450652e-07,
+      "loss": 0.3235,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9370308077285966,
+      "grad_norm": 0.716195530791351,
+      "learning_rate": 1.037701176303274e-07,
+      "loss": 0.296,
+      "step": 11421
+    },
+    {
+      "epoch": 0.9371128522787874,
+      "grad_norm": 0.6801368675866272,
+      "learning_rate": 1.0350099448003692e-07,
+      "loss": 0.2847,
+      "step": 11422
+    },
+    {
+      "epoch": 0.9371948968289782,
+      "grad_norm": 0.6666004055091737,
+      "learning_rate": 1.0323221711263876e-07,
+      "loss": 0.3201,
+      "step": 11423
+    },
+    {
+      "epoch": 0.9372769413791688,
+      "grad_norm": 0.7503059771870692,
+      "learning_rate": 1.0296378554711439e-07,
+      "loss": 0.3032,
+      "step": 11424
+    },
+    {
+      "epoch": 0.9373589859293596,
+      "grad_norm": 0.6609817933238784,
+      "learning_rate": 1.0269569980241978e-07,
+      "loss": 0.2723,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9374410304795504,
+      "grad_norm": 0.6680656245403223,
+      "learning_rate": 1.02427959897487e-07,
+      "loss": 0.3184,
+      "step": 11426
+    },
+    {
+      "epoch": 0.9375230750297412,
+      "grad_norm": 0.7027542145849075,
+      "learning_rate": 1.0216056585122258e-07,
+      "loss": 0.2599,
+      "step": 11427
+    },
+    {
+      "epoch": 0.9376051195799319,
+      "grad_norm": 0.7597473867667347,
+      "learning_rate": 1.0189351768250977e-07,
+      "loss": 0.3173,
+      "step": 11428
+    },
+    {
+      "epoch": 0.9376871641301227,
+      "grad_norm": 0.7763884027728617,
+      "learning_rate": 1.0162681541020791e-07,
+      "loss": 0.2952,
+      "step": 11429
+    },
+    {
+      "epoch": 0.9377692086803134,
+      "grad_norm": 0.67041235146651,
+      "learning_rate": 1.0136045905315028e-07,
+      "loss": 0.3512,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9378512532305041,
+      "grad_norm": 0.6778152558807613,
+      "learning_rate": 1.0109444863014684e-07,
+      "loss": 0.2799,
+      "step": 11431
+    },
+    {
+      "epoch": 0.9379332977806949,
+      "grad_norm": 0.8274101173392185,
+      "learning_rate": 1.0082878415998309e-07,
+      "loss": 0.2935,
+      "step": 11432
+    },
+    {
+      "epoch": 0.9380153423308857,
+      "grad_norm": 0.7236163054366673,
+      "learning_rate": 1.0056346566141906e-07,
+      "loss": 0.3289,
+      "step": 11433
+    },
+    {
+      "epoch": 0.9380973868810765,
+      "grad_norm": 0.68777586512269,
+      "learning_rate": 1.0029849315319196e-07,
+      "loss": 0.2991,
+      "step": 11434
+    },
+    {
+      "epoch": 0.9381794314312671,
+      "grad_norm": 0.7060564727549526,
+      "learning_rate": 1.0003386665401293e-07,
+      "loss": 0.3161,
+      "step": 11435
+    },
+    {
+      "epoch": 0.9382614759814579,
+      "grad_norm": 0.6792646972898028,
+      "learning_rate": 9.976958618256982e-08,
+      "loss": 0.2877,
+      "step": 11436
+    },
+    {
+      "epoch": 0.9383435205316487,
+      "grad_norm": 0.8881067833235698,
+      "learning_rate": 9.950565175752602e-08,
+      "loss": 0.3436,
+      "step": 11437
+    },
+    {
+      "epoch": 0.9384255650818394,
+      "grad_norm": 0.9293869187017928,
+      "learning_rate": 9.92420633975194e-08,
+      "loss": 0.3157,
+      "step": 11438
+    },
+    {
+      "epoch": 0.9385076096320302,
+      "grad_norm": 0.6377382813319581,
+      "learning_rate": 9.897882112116509e-08,
+      "loss": 0.2773,
+      "step": 11439
+    },
+    {
+      "epoch": 0.938589654182221,
+      "grad_norm": 0.6836583759668082,
+      "learning_rate": 9.871592494705206e-08,
+      "loss": 0.3107,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9386716987324117,
+      "grad_norm": 0.7308620994963569,
+      "learning_rate": 9.845337489374551e-08,
+      "loss": 0.2921,
+      "step": 11441
+    },
+    {
+      "epoch": 0.9387537432826024,
+      "grad_norm": 0.9451983149243877,
+      "learning_rate": 9.819117097978726e-08,
+      "loss": 0.3062,
+      "step": 11442
+    },
+    {
+      "epoch": 0.9388357878327932,
+      "grad_norm": 1.127779596808752,
+      "learning_rate": 9.792931322369304e-08,
+      "loss": 0.2911,
+      "step": 11443
+    },
+    {
+      "epoch": 0.938917832382984,
+      "grad_norm": 0.7039047667305782,
+      "learning_rate": 9.766780164395472e-08,
+      "loss": 0.2839,
+      "step": 11444
+    },
+    {
+      "epoch": 0.9389998769331747,
+      "grad_norm": 0.7241812376798402,
+      "learning_rate": 9.740663625904034e-08,
+      "loss": 0.2885,
+      "step": 11445
+    },
+    {
+      "epoch": 0.9390819214833654,
+      "grad_norm": 0.7773796437347924,
+      "learning_rate": 9.714581708739235e-08,
+      "loss": 0.3162,
+      "step": 11446
+    },
+    {
+      "epoch": 0.9391639660335562,
+      "grad_norm": 0.7057798068506712,
+      "learning_rate": 9.688534414743045e-08,
+      "loss": 0.2962,
+      "step": 11447
+    },
+    {
+      "epoch": 0.939246010583747,
+      "grad_norm": 0.7265873181916543,
+      "learning_rate": 9.662521745754716e-08,
+      "loss": 0.3274,
+      "step": 11448
+    },
+    {
+      "epoch": 0.9393280551339377,
+      "grad_norm": 0.7377738352545727,
+      "learning_rate": 9.63654370361139e-08,
+      "loss": 0.2748,
+      "step": 11449
+    },
+    {
+      "epoch": 0.9394100996841285,
+      "grad_norm": 0.6887898273397809,
+      "learning_rate": 9.610600290147487e-08,
+      "loss": 0.2875,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9394921442343193,
+      "grad_norm": 0.7268902331079897,
+      "learning_rate": 9.584691507195098e-08,
+      "loss": 0.3198,
+      "step": 11451
+    },
+    {
+      "epoch": 0.9395741887845099,
+      "grad_norm": 0.7756066819228182,
+      "learning_rate": 9.558817356583926e-08,
+      "loss": 0.3372,
+      "step": 11452
+    },
+    {
+      "epoch": 0.9396562333347007,
+      "grad_norm": 0.7487825956126792,
+      "learning_rate": 9.532977840141123e-08,
+      "loss": 0.2736,
+      "step": 11453
+    },
+    {
+      "epoch": 0.9397382778848915,
+      "grad_norm": 0.7279215288386756,
+      "learning_rate": 9.507172959691336e-08,
+      "loss": 0.2774,
+      "step": 11454
+    },
+    {
+      "epoch": 0.9398203224350823,
+      "grad_norm": 0.7684900591157463,
+      "learning_rate": 9.481402717057053e-08,
+      "loss": 0.3307,
+      "step": 11455
+    },
+    {
+      "epoch": 0.939902366985273,
+      "grad_norm": 0.7222079140562472,
+      "learning_rate": 9.455667114057988e-08,
+      "loss": 0.2962,
+      "step": 11456
+    },
+    {
+      "epoch": 0.9399844115354637,
+      "grad_norm": 0.9023910511683461,
+      "learning_rate": 9.429966152511627e-08,
+      "loss": 0.3188,
+      "step": 11457
+    },
+    {
+      "epoch": 0.9400664560856545,
+      "grad_norm": 0.7323408846256841,
+      "learning_rate": 9.404299834232855e-08,
+      "loss": 0.2626,
+      "step": 11458
+    },
+    {
+      "epoch": 0.9401485006358453,
+      "grad_norm": 0.678084038314316,
+      "learning_rate": 9.378668161034222e-08,
+      "loss": 0.2925,
+      "step": 11459
+    },
+    {
+      "epoch": 0.940230545186036,
+      "grad_norm": 0.6773146536634166,
+      "learning_rate": 9.353071134725888e-08,
+      "loss": 0.2676,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9403125897362268,
+      "grad_norm": 0.7191334700738516,
+      "learning_rate": 9.3275087571153e-08,
+      "loss": 0.2996,
+      "step": 11461
+    },
+    {
+      "epoch": 0.9403946342864176,
+      "grad_norm": 0.7069648059663971,
+      "learning_rate": 9.301981030007734e-08,
+      "loss": 0.2865,
+      "step": 11462
+    },
+    {
+      "epoch": 0.9404766788366082,
+      "grad_norm": 0.8005826709223106,
+      "learning_rate": 9.27648795520597e-08,
+      "loss": 0.2637,
+      "step": 11463
+    },
+    {
+      "epoch": 0.940558723386799,
+      "grad_norm": 0.6631779215825939,
+      "learning_rate": 9.251029534510126e-08,
+      "loss": 0.2903,
+      "step": 11464
+    },
+    {
+      "epoch": 0.9406407679369898,
+      "grad_norm": 0.7086006727272716,
+      "learning_rate": 9.225605769718149e-08,
+      "loss": 0.3092,
+      "step": 11465
+    },
+    {
+      "epoch": 0.9407228124871806,
+      "grad_norm": 0.664737686550547,
+      "learning_rate": 9.200216662625439e-08,
+      "loss": 0.2579,
+      "step": 11466
+    },
+    {
+      "epoch": 0.9408048570373713,
+      "grad_norm": 0.7508361526648413,
+      "learning_rate": 9.174862215024837e-08,
+      "loss": 0.2873,
+      "step": 11467
+    },
+    {
+      "epoch": 0.940886901587562,
+      "grad_norm": 0.6725443568134285,
+      "learning_rate": 9.14954242870697e-08,
+      "loss": 0.2839,
+      "step": 11468
+    },
+    {
+      "epoch": 0.9409689461377528,
+      "grad_norm": 0.7709590716724483,
+      "learning_rate": 9.124257305459738e-08,
+      "loss": 0.2685,
+      "step": 11469
+    },
+    {
+      "epoch": 0.9410509906879435,
+      "grad_norm": 0.7748988585923726,
+      "learning_rate": 9.099006847068826e-08,
+      "loss": 0.2846,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9411330352381343,
+      "grad_norm": 0.9226660646977534,
+      "learning_rate": 9.07379105531736e-08,
+      "loss": 0.3106,
+      "step": 11471
+    },
+    {
+      "epoch": 0.9412150797883251,
+      "grad_norm": 0.6781604278001975,
+      "learning_rate": 9.04860993198603e-08,
+      "loss": 0.3191,
+      "step": 11472
+    },
+    {
+      "epoch": 0.9412971243385159,
+      "grad_norm": 0.6706630661581654,
+      "learning_rate": 9.023463478853134e-08,
+      "loss": 0.3073,
+      "step": 11473
+    },
+    {
+      "epoch": 0.9413791688887065,
+      "grad_norm": 0.6976323791989976,
+      "learning_rate": 8.998351697694363e-08,
+      "loss": 0.3022,
+      "step": 11474
+    },
+    {
+      "epoch": 0.9414612134388973,
+      "grad_norm": 0.8137035472646597,
+      "learning_rate": 8.973274590283188e-08,
+      "loss": 0.285,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9415432579890881,
+      "grad_norm": 0.6854038704070666,
+      "learning_rate": 8.948232158390468e-08,
+      "loss": 0.3242,
+      "step": 11476
+    },
+    {
+      "epoch": 0.9416253025392788,
+      "grad_norm": 0.6933678160368562,
+      "learning_rate": 8.923224403784681e-08,
+      "loss": 0.2893,
+      "step": 11477
+    },
+    {
+      "epoch": 0.9417073470894696,
+      "grad_norm": 0.6508860653228502,
+      "learning_rate": 8.898251328231799e-08,
+      "loss": 0.2696,
+      "step": 11478
+    },
+    {
+      "epoch": 0.9417893916396604,
+      "grad_norm": 0.66387184311223,
+      "learning_rate": 8.873312933495359e-08,
+      "loss": 0.2825,
+      "step": 11479
+    },
+    {
+      "epoch": 0.9418714361898511,
+      "grad_norm": 0.6254492848877901,
+      "learning_rate": 8.848409221336452e-08,
+      "loss": 0.2608,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9419534807400418,
+      "grad_norm": 0.7001094807201339,
+      "learning_rate": 8.823540193513891e-08,
+      "loss": 0.3053,
+      "step": 11481
+    },
+    {
+      "epoch": 0.9420355252902326,
+      "grad_norm": 0.7367446995456215,
+      "learning_rate": 8.798705851783717e-08,
+      "loss": 0.298,
+      "step": 11482
+    },
+    {
+      "epoch": 0.9421175698404234,
+      "grad_norm": 0.7234311453615313,
+      "learning_rate": 8.773906197899861e-08,
+      "loss": 0.2882,
+      "step": 11483
+    },
+    {
+      "epoch": 0.9421996143906141,
+      "grad_norm": 0.697165993964216,
+      "learning_rate": 8.749141233613422e-08,
+      "loss": 0.261,
+      "step": 11484
+    },
+    {
+      "epoch": 0.9422816589408048,
+      "grad_norm": 0.7468877288977702,
+      "learning_rate": 8.72441096067339e-08,
+      "loss": 0.2992,
+      "step": 11485
+    },
+    {
+      "epoch": 0.9423637034909956,
+      "grad_norm": 0.6881004219747189,
+      "learning_rate": 8.699715380826256e-08,
+      "loss": 0.3019,
+      "step": 11486
+    },
+    {
+      "epoch": 0.9424457480411864,
+      "grad_norm": 0.7322255727381941,
+      "learning_rate": 8.675054495815793e-08,
+      "loss": 0.3304,
+      "step": 11487
+    },
+    {
+      "epoch": 0.9425277925913771,
+      "grad_norm": 0.7822209792718776,
+      "learning_rate": 8.650428307383663e-08,
+      "loss": 0.2816,
+      "step": 11488
+    },
+    {
+      "epoch": 0.9426098371415679,
+      "grad_norm": 0.7055538434914514,
+      "learning_rate": 8.625836817268806e-08,
+      "loss": 0.2995,
+      "step": 11489
+    },
+    {
+      "epoch": 0.9426918816917587,
+      "grad_norm": 0.7191450127452647,
+      "learning_rate": 8.601280027208004e-08,
+      "loss": 0.2738,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9427739262419493,
+      "grad_norm": 0.6942053505925784,
+      "learning_rate": 8.576757938935254e-08,
+      "loss": 0.3143,
+      "step": 11491
+    },
+    {
+      "epoch": 0.9428559707921401,
+      "grad_norm": 0.8665718632841662,
+      "learning_rate": 8.552270554182284e-08,
+      "loss": 0.3037,
+      "step": 11492
+    },
+    {
+      "epoch": 0.9429380153423309,
+      "grad_norm": 0.7018224538944521,
+      "learning_rate": 8.52781787467849e-08,
+      "loss": 0.2282,
+      "step": 11493
+    },
+    {
+      "epoch": 0.9430200598925217,
+      "grad_norm": 0.6553122668049707,
+      "learning_rate": 8.503399902150488e-08,
+      "loss": 0.2605,
+      "step": 11494
+    },
+    {
+      "epoch": 0.9431021044427124,
+      "grad_norm": 0.7406374258371962,
+      "learning_rate": 8.479016638322791e-08,
+      "loss": 0.3325,
+      "step": 11495
+    },
+    {
+      "epoch": 0.9431841489929031,
+      "grad_norm": 0.7510378505301605,
+      "learning_rate": 8.45466808491724e-08,
+      "loss": 0.2998,
+      "step": 11496
+    },
+    {
+      "epoch": 0.9432661935430939,
+      "grad_norm": 0.7087174324181473,
+      "learning_rate": 8.430354243653294e-08,
+      "loss": 0.3286,
+      "step": 11497
+    },
+    {
+      "epoch": 0.9433482380932846,
+      "grad_norm": 0.6404028587349707,
+      "learning_rate": 8.40607511624797e-08,
+      "loss": 0.3223,
+      "step": 11498
+    },
+    {
+      "epoch": 0.9434302826434754,
+      "grad_norm": 0.7128122833506783,
+      "learning_rate": 8.381830704415839e-08,
+      "loss": 0.2705,
+      "step": 11499
+    },
+    {
+      "epoch": 0.9435123271936662,
+      "grad_norm": 0.6431078259141513,
+      "learning_rate": 8.357621009868921e-08,
+      "loss": 0.3267,
+      "step": 11500
+    },
+    {
+      "epoch": 0.943594371743857,
+      "grad_norm": 0.6609161649635266,
+      "learning_rate": 8.333446034317017e-08,
+      "loss": 0.2753,
+      "step": 11501
+    },
+    {
+      "epoch": 0.9436764162940476,
+      "grad_norm": 0.7160954714356189,
+      "learning_rate": 8.309305779467147e-08,
+      "loss": 0.2952,
+      "step": 11502
+    },
+    {
+      "epoch": 0.9437584608442384,
+      "grad_norm": 0.7711993144787835,
+      "learning_rate": 8.285200247024172e-08,
+      "loss": 0.3393,
+      "step": 11503
+    },
+    {
+      "epoch": 0.9438405053944292,
+      "grad_norm": 0.6648171728000005,
+      "learning_rate": 8.26112943869034e-08,
+      "loss": 0.2964,
+      "step": 11504
+    },
+    {
+      "epoch": 0.9439225499446199,
+      "grad_norm": 0.6710913119965035,
+      "learning_rate": 8.237093356165571e-08,
+      "loss": 0.2638,
+      "step": 11505
+    },
+    {
+      "epoch": 0.9440045944948107,
+      "grad_norm": 0.6680039168548512,
+      "learning_rate": 8.213092001147117e-08,
+      "loss": 0.2986,
+      "step": 11506
+    },
+    {
+      "epoch": 0.9440866390450015,
+      "grad_norm": 0.8254647749936619,
+      "learning_rate": 8.189125375330009e-08,
+      "loss": 0.3145,
+      "step": 11507
+    },
+    {
+      "epoch": 0.9441686835951922,
+      "grad_norm": 0.7193503975875055,
+      "learning_rate": 8.16519348040673e-08,
+      "loss": 0.3112,
+      "step": 11508
+    },
+    {
+      "epoch": 0.9442507281453829,
+      "grad_norm": 0.7609394532573432,
+      "learning_rate": 8.14129631806726e-08,
+      "loss": 0.3289,
+      "step": 11509
+    },
+    {
+      "epoch": 0.9443327726955737,
+      "grad_norm": 0.7261217644712346,
+      "learning_rate": 8.117433889999248e-08,
+      "loss": 0.2862,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9444148172457645,
+      "grad_norm": 0.7346362788092017,
+      "learning_rate": 8.093606197887793e-08,
+      "loss": 0.2902,
+      "step": 11511
+    },
+    {
+      "epoch": 0.9444968617959552,
+      "grad_norm": 0.685341725041379,
+      "learning_rate": 8.069813243415492e-08,
+      "loss": 0.3116,
+      "step": 11512
+    },
+    {
+      "epoch": 0.9445789063461459,
+      "grad_norm": 0.7197600106547691,
+      "learning_rate": 8.046055028262667e-08,
+      "loss": 0.3236,
+      "step": 11513
+    },
+    {
+      "epoch": 0.9446609508963367,
+      "grad_norm": 0.7767876637259552,
+      "learning_rate": 8.022331554107088e-08,
+      "loss": 0.2995,
+      "step": 11514
+    },
+    {
+      "epoch": 0.9447429954465275,
+      "grad_norm": 0.7167543509770662,
+      "learning_rate": 7.998642822624025e-08,
+      "loss": 0.3159,
+      "step": 11515
+    },
+    {
+      "epoch": 0.9448250399967182,
+      "grad_norm": 0.8018397332447394,
+      "learning_rate": 7.974988835486308e-08,
+      "loss": 0.2962,
+      "step": 11516
+    },
+    {
+      "epoch": 0.944907084546909,
+      "grad_norm": 0.7198641076539729,
+      "learning_rate": 7.951369594364433e-08,
+      "loss": 0.2928,
+      "step": 11517
+    },
+    {
+      "epoch": 0.9449891290970998,
+      "grad_norm": 0.7531539889308981,
+      "learning_rate": 7.927785100926288e-08,
+      "loss": 0.3084,
+      "step": 11518
+    },
+    {
+      "epoch": 0.9450711736472904,
+      "grad_norm": 0.7228436373021899,
+      "learning_rate": 7.904235356837376e-08,
+      "loss": 0.3057,
+      "step": 11519
+    },
+    {
+      "epoch": 0.9451532181974812,
+      "grad_norm": 0.6561254685139658,
+      "learning_rate": 7.880720363760753e-08,
+      "loss": 0.2591,
+      "step": 11520
+    },
+    {
+      "epoch": 0.945235262747672,
+      "grad_norm": 0.7648048534035438,
+      "learning_rate": 7.85724012335698e-08,
+      "loss": 0.3089,
+      "step": 11521
+    },
+    {
+      "epoch": 0.9453173072978628,
+      "grad_norm": 0.6633559754210834,
+      "learning_rate": 7.833794637284232e-08,
+      "loss": 0.3155,
+      "step": 11522
+    },
+    {
+      "epoch": 0.9453993518480535,
+      "grad_norm": 0.6130192923942714,
+      "learning_rate": 7.810383907198182e-08,
+      "loss": 0.2582,
+      "step": 11523
+    },
+    {
+      "epoch": 0.9454813963982442,
+      "grad_norm": 1.3514072613305377,
+      "learning_rate": 7.787007934752067e-08,
+      "loss": 0.2699,
+      "step": 11524
+    },
+    {
+      "epoch": 0.945563440948435,
+      "grad_norm": 0.6541022912331392,
+      "learning_rate": 7.763666721596674e-08,
+      "loss": 0.2654,
+      "step": 11525
+    },
+    {
+      "epoch": 0.9456454854986257,
+      "grad_norm": 0.7482947846810067,
+      "learning_rate": 7.740360269380243e-08,
+      "loss": 0.3299,
+      "step": 11526
+    },
+    {
+      "epoch": 0.9457275300488165,
+      "grad_norm": 0.6834654716709402,
+      "learning_rate": 7.717088579748733e-08,
+      "loss": 0.3014,
+      "step": 11527
+    },
+    {
+      "epoch": 0.9458095745990073,
+      "grad_norm": 0.7910266505529446,
+      "learning_rate": 7.693851654345497e-08,
+      "loss": 0.3105,
+      "step": 11528
+    },
+    {
+      "epoch": 0.9458916191491981,
+      "grad_norm": 0.7208158964372585,
+      "learning_rate": 7.670649494811555e-08,
+      "loss": 0.3013,
+      "step": 11529
+    },
+    {
+      "epoch": 0.9459736636993887,
+      "grad_norm": 0.7179922695152416,
+      "learning_rate": 7.647482102785264e-08,
+      "loss": 0.3015,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9460557082495795,
+      "grad_norm": 1.053499872628843,
+      "learning_rate": 7.62434947990287e-08,
+      "loss": 0.3441,
+      "step": 11531
+    },
+    {
+      "epoch": 0.9461377527997703,
+      "grad_norm": 0.7241879030965517,
+      "learning_rate": 7.60125162779779e-08,
+      "loss": 0.3129,
+      "step": 11532
+    },
+    {
+      "epoch": 0.946219797349961,
+      "grad_norm": 0.7899809994912884,
+      "learning_rate": 7.578188548101217e-08,
+      "loss": 0.3062,
+      "step": 11533
+    },
+    {
+      "epoch": 0.9463018419001518,
+      "grad_norm": 0.7352672876745031,
+      "learning_rate": 7.55516024244185e-08,
+      "loss": 0.3173,
+      "step": 11534
+    },
+    {
+      "epoch": 0.9463838864503425,
+      "grad_norm": 0.6668356450437638,
+      "learning_rate": 7.532166712445888e-08,
+      "loss": 0.3205,
+      "step": 11535
+    },
+    {
+      "epoch": 0.9464659310005333,
+      "grad_norm": 0.755542719906116,
+      "learning_rate": 7.509207959737086e-08,
+      "loss": 0.3063,
+      "step": 11536
+    },
+    {
+      "epoch": 0.946547975550724,
+      "grad_norm": 0.681026534198614,
+      "learning_rate": 7.486283985936815e-08,
+      "loss": 0.3142,
+      "step": 11537
+    },
+    {
+      "epoch": 0.9466300201009148,
+      "grad_norm": 0.6237727887729317,
+      "learning_rate": 7.463394792663891e-08,
+      "loss": 0.2736,
+      "step": 11538
+    },
+    {
+      "epoch": 0.9467120646511056,
+      "grad_norm": 0.7638572748357146,
+      "learning_rate": 7.440540381534633e-08,
+      "loss": 0.2974,
+      "step": 11539
+    },
+    {
+      "epoch": 0.9467941092012963,
+      "grad_norm": 0.7131684810039436,
+      "learning_rate": 7.417720754163138e-08,
+      "loss": 0.3289,
+      "step": 11540
+    },
+    {
+      "epoch": 0.946876153751487,
+      "grad_norm": 0.730520466899204,
+      "learning_rate": 7.394935912160728e-08,
+      "loss": 0.3095,
+      "step": 11541
+    },
+    {
+      "epoch": 0.9469581983016778,
+      "grad_norm": 0.6413106345185771,
+      "learning_rate": 7.372185857136616e-08,
+      "loss": 0.2549,
+      "step": 11542
+    },
+    {
+      "epoch": 0.9470402428518686,
+      "grad_norm": 0.7914003476836075,
+      "learning_rate": 7.349470590697183e-08,
+      "loss": 0.3206,
+      "step": 11543
+    },
+    {
+      "epoch": 0.9471222874020593,
+      "grad_norm": 0.8292503670826639,
+      "learning_rate": 7.326790114446647e-08,
+      "loss": 0.2841,
+      "step": 11544
+    },
+    {
+      "epoch": 0.9472043319522501,
+      "grad_norm": 0.7766804640250751,
+      "learning_rate": 7.30414442998667e-08,
+      "loss": 0.3013,
+      "step": 11545
+    },
+    {
+      "epoch": 0.9472863765024409,
+      "grad_norm": 0.7569266832408227,
+      "learning_rate": 7.281533538916474e-08,
+      "loss": 0.2861,
+      "step": 11546
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.7073713388128995,
+      "learning_rate": 7.258957442832726e-08,
+      "loss": 0.3521,
+      "step": 11547
+    },
+    {
+      "epoch": 0.9474504656028223,
+      "grad_norm": 0.7069615044443939,
+      "learning_rate": 7.236416143329761e-08,
+      "loss": 0.2795,
+      "step": 11548
+    },
+    {
+      "epoch": 0.9475325101530131,
+      "grad_norm": 0.7593124967028164,
+      "learning_rate": 7.213909641999361e-08,
+      "loss": 0.3164,
+      "step": 11549
+    },
+    {
+      "epoch": 0.9476145547032039,
+      "grad_norm": 0.7030275355544322,
+      "learning_rate": 7.191437940431034e-08,
+      "loss": 0.2631,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9476965992533946,
+      "grad_norm": 0.7047416719048204,
+      "learning_rate": 7.16900104021151e-08,
+      "loss": 0.3065,
+      "step": 11551
+    },
+    {
+      "epoch": 0.9477786438035853,
+      "grad_norm": 0.70642068740684,
+      "learning_rate": 7.146598942925353e-08,
+      "loss": 0.3081,
+      "step": 11552
+    },
+    {
+      "epoch": 0.9478606883537761,
+      "grad_norm": 0.6977539222829745,
+      "learning_rate": 7.124231650154633e-08,
+      "loss": 0.2826,
+      "step": 11553
+    },
+    {
+      "epoch": 0.9479427329039668,
+      "grad_norm": 0.7352318816392833,
+      "learning_rate": 7.101899163478698e-08,
+      "loss": 0.2692,
+      "step": 11554
+    },
+    {
+      "epoch": 0.9480247774541576,
+      "grad_norm": 0.6485382777894068,
+      "learning_rate": 7.079601484474841e-08,
+      "loss": 0.3286,
+      "step": 11555
+    },
+    {
+      "epoch": 0.9481068220043484,
+      "grad_norm": 0.7489884914580902,
+      "learning_rate": 7.057338614717523e-08,
+      "loss": 0.2876,
+      "step": 11556
+    },
+    {
+      "epoch": 0.9481888665545392,
+      "grad_norm": 0.7037828799217508,
+      "learning_rate": 7.03511055577899e-08,
+      "loss": 0.2981,
+      "step": 11557
+    },
+    {
+      "epoch": 0.9482709111047298,
+      "grad_norm": 0.670332478639947,
+      "learning_rate": 7.012917309228984e-08,
+      "loss": 0.2707,
+      "step": 11558
+    },
+    {
+      "epoch": 0.9483529556549206,
+      "grad_norm": 0.7155784798390297,
+      "learning_rate": 6.990758876634695e-08,
+      "loss": 0.3036,
+      "step": 11559
+    },
+    {
+      "epoch": 0.9484350002051114,
+      "grad_norm": 0.7863270248620298,
+      "learning_rate": 6.968635259560874e-08,
+      "loss": 0.2846,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9485170447553021,
+      "grad_norm": 0.8666710398026094,
+      "learning_rate": 6.946546459569991e-08,
+      "loss": 0.2535,
+      "step": 11561
+    },
+    {
+      "epoch": 0.9485990893054929,
+      "grad_norm": 0.6496157312306675,
+      "learning_rate": 6.9244924782218e-08,
+      "loss": 0.273,
+      "step": 11562
+    },
+    {
+      "epoch": 0.9486811338556836,
+      "grad_norm": 0.7464031494779539,
+      "learning_rate": 6.902473317073776e-08,
+      "loss": 0.2996,
+      "step": 11563
+    },
+    {
+      "epoch": 0.9487631784058744,
+      "grad_norm": 0.7743515508692638,
+      "learning_rate": 6.880488977680899e-08,
+      "loss": 0.2922,
+      "step": 11564
+    },
+    {
+      "epoch": 0.9488452229560651,
+      "grad_norm": 0.6872132288112274,
+      "learning_rate": 6.858539461595593e-08,
+      "loss": 0.2996,
+      "step": 11565
+    },
+    {
+      "epoch": 0.9489272675062559,
+      "grad_norm": 0.6323569730298406,
+      "learning_rate": 6.83662477036795e-08,
+      "loss": 0.2837,
+      "step": 11566
+    },
+    {
+      "epoch": 0.9490093120564467,
+      "grad_norm": 0.7146995360155519,
+      "learning_rate": 6.814744905545512e-08,
+      "loss": 0.3127,
+      "step": 11567
+    },
+    {
+      "epoch": 0.9490913566066375,
+      "grad_norm": 0.674149265133856,
+      "learning_rate": 6.792899868673487e-08,
+      "loss": 0.2711,
+      "step": 11568
+    },
+    {
+      "epoch": 0.9491734011568281,
+      "grad_norm": 0.7050749829592777,
+      "learning_rate": 6.771089661294417e-08,
+      "loss": 0.3186,
+      "step": 11569
+    },
+    {
+      "epoch": 0.9492554457070189,
+      "grad_norm": 0.7080718004137615,
+      "learning_rate": 6.749314284948571e-08,
+      "loss": 0.2412,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9493374902572097,
+      "grad_norm": 0.7257453845881495,
+      "learning_rate": 6.727573741173721e-08,
+      "loss": 0.3474,
+      "step": 11571
+    },
+    {
+      "epoch": 0.9494195348074004,
+      "grad_norm": 0.7507020156443684,
+      "learning_rate": 6.705868031505025e-08,
+      "loss": 0.2653,
+      "step": 11572
+    },
+    {
+      "epoch": 0.9495015793575912,
+      "grad_norm": 0.7152057311059684,
+      "learning_rate": 6.68419715747548e-08,
+      "loss": 0.2788,
+      "step": 11573
+    },
+    {
+      "epoch": 0.949583623907782,
+      "grad_norm": 0.7336896137863117,
+      "learning_rate": 6.662561120615363e-08,
+      "loss": 0.3003,
+      "step": 11574
+    },
+    {
+      "epoch": 0.9496656684579727,
+      "grad_norm": 0.7131525452173534,
+      "learning_rate": 6.640959922452506e-08,
+      "loss": 0.3523,
+      "step": 11575
+    },
+    {
+      "epoch": 0.9497477130081634,
+      "grad_norm": 0.6911459016783086,
+      "learning_rate": 6.619393564512466e-08,
+      "loss": 0.3117,
+      "step": 11576
+    },
+    {
+      "epoch": 0.9498297575583542,
+      "grad_norm": 0.6441670966913938,
+      "learning_rate": 6.597862048318193e-08,
+      "loss": 0.3102,
+      "step": 11577
+    },
+    {
+      "epoch": 0.949911802108545,
+      "grad_norm": 0.708464484553245,
+      "learning_rate": 6.576365375390137e-08,
+      "loss": 0.2816,
+      "step": 11578
+    },
+    {
+      "epoch": 0.9499938466587357,
+      "grad_norm": 0.8331564780285896,
+      "learning_rate": 6.554903547246416e-08,
+      "loss": 0.2918,
+      "step": 11579
+    },
+    {
+      "epoch": 0.9500758912089264,
+      "grad_norm": 0.7211721795479534,
+      "learning_rate": 6.533476565402653e-08,
+      "loss": 0.2862,
+      "step": 11580
+    },
+    {
+      "epoch": 0.9501579357591172,
+      "grad_norm": 0.6665651631275498,
+      "learning_rate": 6.512084431371968e-08,
+      "loss": 0.3288,
+      "step": 11581
+    },
+    {
+      "epoch": 0.950239980309308,
+      "grad_norm": 0.7779070189916172,
+      "learning_rate": 6.490727146664988e-08,
+      "loss": 0.2972,
+      "step": 11582
+    },
+    {
+      "epoch": 0.9503220248594987,
+      "grad_norm": 0.7662630165966777,
+      "learning_rate": 6.469404712790062e-08,
+      "loss": 0.2875,
+      "step": 11583
+    },
+    {
+      "epoch": 0.9504040694096895,
+      "grad_norm": 0.8093046489790418,
+      "learning_rate": 6.448117131252762e-08,
+      "loss": 0.293,
+      "step": 11584
+    },
+    {
+      "epoch": 0.9504861139598803,
+      "grad_norm": 0.6786021819703811,
+      "learning_rate": 6.426864403556499e-08,
+      "loss": 0.2507,
+      "step": 11585
+    },
+    {
+      "epoch": 0.9505681585100709,
+      "grad_norm": 0.6268263005150039,
+      "learning_rate": 6.405646531202126e-08,
+      "loss": 0.3109,
+      "step": 11586
+    },
+    {
+      "epoch": 0.9506502030602617,
+      "grad_norm": 0.6555697327105452,
+      "learning_rate": 6.384463515687945e-08,
+      "loss": 0.2632,
+      "step": 11587
+    },
+    {
+      "epoch": 0.9507322476104525,
+      "grad_norm": 0.9631445474291255,
+      "learning_rate": 6.363315358509926e-08,
+      "loss": 0.3086,
+      "step": 11588
+    },
+    {
+      "epoch": 0.9508142921606433,
+      "grad_norm": 0.6926316201432011,
+      "learning_rate": 6.342202061161485e-08,
+      "loss": 0.2961,
+      "step": 11589
+    },
+    {
+      "epoch": 0.950896336710834,
+      "grad_norm": 0.7800170289582182,
+      "learning_rate": 6.32112362513354e-08,
+      "loss": 0.2958,
+      "step": 11590
+    },
+    {
+      "epoch": 0.9509783812610247,
+      "grad_norm": 0.7498247219864855,
+      "learning_rate": 6.300080051914792e-08,
+      "loss": 0.2451,
+      "step": 11591
+    },
+    {
+      "epoch": 0.9510604258112155,
+      "grad_norm": 0.7490158684690564,
+      "learning_rate": 6.279071342991106e-08,
+      "loss": 0.3026,
+      "step": 11592
+    },
+    {
+      "epoch": 0.9511424703614062,
+      "grad_norm": 0.7708519609944678,
+      "learning_rate": 6.258097499846238e-08,
+      "loss": 0.2806,
+      "step": 11593
+    },
+    {
+      "epoch": 0.951224514911597,
+      "grad_norm": 0.6763920187398628,
+      "learning_rate": 6.237158523961284e-08,
+      "loss": 0.3029,
+      "step": 11594
+    },
+    {
+      "epoch": 0.9513065594617878,
+      "grad_norm": 0.8800498508458161,
+      "learning_rate": 6.216254416814837e-08,
+      "loss": 0.3113,
+      "step": 11595
+    },
+    {
+      "epoch": 0.9513886040119786,
+      "grad_norm": 0.6632601102159557,
+      "learning_rate": 6.195385179883218e-08,
+      "loss": 0.2862,
+      "step": 11596
+    },
+    {
+      "epoch": 0.9514706485621692,
+      "grad_norm": 0.7765305025720091,
+      "learning_rate": 6.174550814640135e-08,
+      "loss": 0.3493,
+      "step": 11597
+    },
+    {
+      "epoch": 0.95155269311236,
+      "grad_norm": 0.7248034626895643,
+      "learning_rate": 6.153751322556855e-08,
+      "loss": 0.2722,
+      "step": 11598
+    },
+    {
+      "epoch": 0.9516347376625508,
+      "grad_norm": 0.7189031701545107,
+      "learning_rate": 6.132986705102317e-08,
+      "loss": 0.2857,
+      "step": 11599
+    },
+    {
+      "epoch": 0.9517167822127415,
+      "grad_norm": 1.0673086684779385,
+      "learning_rate": 6.112256963742735e-08,
+      "loss": 0.2774,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9517988267629323,
+      "grad_norm": 0.7306987898428707,
+      "learning_rate": 6.091562099942106e-08,
+      "loss": 0.2914,
+      "step": 11601
+    },
+    {
+      "epoch": 0.951880871313123,
+      "grad_norm": 0.7177292746121716,
+      "learning_rate": 6.070902115161814e-08,
+      "loss": 0.3061,
+      "step": 11602
+    },
+    {
+      "epoch": 0.9519629158633138,
+      "grad_norm": 0.7630713359627296,
+      "learning_rate": 6.050277010860861e-08,
+      "loss": 0.3246,
+      "step": 11603
+    },
+    {
+      "epoch": 0.9520449604135045,
+      "grad_norm": 0.7469175423030573,
+      "learning_rate": 6.029686788495748e-08,
+      "loss": 0.2489,
+      "step": 11604
+    },
+    {
+      "epoch": 0.9521270049636953,
+      "grad_norm": 0.6803354094364792,
+      "learning_rate": 6.009131449520534e-08,
+      "loss": 0.3053,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9522090495138861,
+      "grad_norm": 0.683176537428937,
+      "learning_rate": 5.988610995386779e-08,
+      "loss": 0.2994,
+      "step": 11606
+    },
+    {
+      "epoch": 0.9522910940640767,
+      "grad_norm": 0.6690136663173144,
+      "learning_rate": 5.968125427543658e-08,
+      "loss": 0.2792,
+      "step": 11607
+    },
+    {
+      "epoch": 0.9523731386142675,
+      "grad_norm": 0.7392215599824304,
+      "learning_rate": 5.947674747437793e-08,
+      "loss": 0.3085,
+      "step": 11608
+    },
+    {
+      "epoch": 0.9524551831644583,
+      "grad_norm": 1.7299677084087792,
+      "learning_rate": 5.9272589565133595e-08,
+      "loss": 0.2701,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9525372277146491,
+      "grad_norm": 0.7019272451742088,
+      "learning_rate": 5.906878056212151e-08,
+      "loss": 0.3,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9526192722648398,
+      "grad_norm": 0.7861135960358739,
+      "learning_rate": 5.886532047973348e-08,
+      "loss": 0.3042,
+      "step": 11611
+    },
+    {
+      "epoch": 0.9527013168150306,
+      "grad_norm": 0.702293766411594,
+      "learning_rate": 5.866220933233802e-08,
+      "loss": 0.3319,
+      "step": 11612
+    },
+    {
+      "epoch": 0.9527833613652213,
+      "grad_norm": 0.688562641923607,
+      "learning_rate": 5.8459447134278645e-08,
+      "loss": 0.2687,
+      "step": 11613
+    },
+    {
+      "epoch": 0.952865405915412,
+      "grad_norm": 0.6767244198083797,
+      "learning_rate": 5.825703389987392e-08,
+      "loss": 0.2735,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9529474504656028,
+      "grad_norm": 0.8225906034779811,
+      "learning_rate": 5.80549696434185e-08,
+      "loss": 0.3149,
+      "step": 11615
+    },
+    {
+      "epoch": 0.9530294950157936,
+      "grad_norm": 0.6629842734259151,
+      "learning_rate": 5.785325437918044e-08,
+      "loss": 0.2937,
+      "step": 11616
+    },
+    {
+      "epoch": 0.9531115395659844,
+      "grad_norm": 0.7516852340698679,
+      "learning_rate": 5.765188812140554e-08,
+      "loss": 0.3258,
+      "step": 11617
+    },
+    {
+      "epoch": 0.953193584116175,
+      "grad_norm": 0.8139492164037053,
+      "learning_rate": 5.745087088431411e-08,
+      "loss": 0.3003,
+      "step": 11618
+    },
+    {
+      "epoch": 0.9532756286663658,
+      "grad_norm": 1.1031558597044164,
+      "learning_rate": 5.7250202682100907e-08,
+      "loss": 0.2819,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9533576732165566,
+      "grad_norm": 0.7883789065832931,
+      "learning_rate": 5.704988352893792e-08,
+      "loss": 0.3065,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9534397177667473,
+      "grad_norm": 0.7862213952373404,
+      "learning_rate": 5.684991343896995e-08,
+      "loss": 0.3044,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9535217623169381,
+      "grad_norm": 0.6734445861737006,
+      "learning_rate": 5.665029242632014e-08,
+      "loss": 0.264,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9536038068671289,
+      "grad_norm": 0.8753518522241668,
+      "learning_rate": 5.645102050508389e-08,
+      "loss": 0.2858,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9536858514173197,
+      "grad_norm": 0.7141115081875615,
+      "learning_rate": 5.625209768933437e-08,
+      "loss": 0.2703,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9537678959675103,
+      "grad_norm": 0.7637739955258056,
+      "learning_rate": 5.60535239931187e-08,
+      "loss": 0.2898,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9538499405177011,
+      "grad_norm": 0.7914097604703239,
+      "learning_rate": 5.585529943046064e-08,
+      "loss": 0.2963,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9539319850678919,
+      "grad_norm": 0.6955174703033478,
+      "learning_rate": 5.5657424015357895e-08,
+      "loss": 0.2853,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9540140296180826,
+      "grad_norm": 0.6882487890453047,
+      "learning_rate": 5.545989776178373e-08,
+      "loss": 0.2996,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9540960741682734,
+      "grad_norm": 0.8945239570390398,
+      "learning_rate": 5.5262720683688096e-08,
+      "loss": 0.2914,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9541781187184641,
+      "grad_norm": 0.6845077893942489,
+      "learning_rate": 5.506589279499486e-08,
+      "loss": 0.2989,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9542601632686549,
+      "grad_norm": 0.7263728681554206,
+      "learning_rate": 5.486941410960289e-08,
+      "loss": 0.2862,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9543422078188456,
+      "grad_norm": 0.6468141374464808,
+      "learning_rate": 5.467328464138888e-08,
+      "loss": 0.3251,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9544242523690364,
+      "grad_norm": 0.7236304806659882,
+      "learning_rate": 5.4477504404201185e-08,
+      "loss": 0.3204,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9545062969192272,
+      "grad_norm": 0.7226330231593685,
+      "learning_rate": 5.4282073411867087e-08,
+      "loss": 0.3264,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9545883414694178,
+      "grad_norm": 0.7916578027205069,
+      "learning_rate": 5.4086991678187206e-08,
+      "loss": 0.3173,
+      "step": 11635
+    },
+    {
+      "epoch": 0.9546703860196086,
+      "grad_norm": 0.7151703863281624,
+      "learning_rate": 5.389225921693775e-08,
+      "loss": 0.2749,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9547524305697994,
+      "grad_norm": 0.7256015837028933,
+      "learning_rate": 5.369787604186993e-08,
+      "loss": 0.2862,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9548344751199902,
+      "grad_norm": 0.7275220701334022,
+      "learning_rate": 5.350384216671167e-08,
+      "loss": 0.2799,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9549165196701809,
+      "grad_norm": 0.7378517868581264,
+      "learning_rate": 5.331015760516478e-08,
+      "loss": 0.2978,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9549985642203717,
+      "grad_norm": 0.7260809333066194,
+      "learning_rate": 5.3116822370907206e-08,
+      "loss": 0.2944,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9550806087705624,
+      "grad_norm": 0.6293781908901763,
+      "learning_rate": 5.2923836477591916e-08,
+      "loss": 0.2761,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9551626533207531,
+      "grad_norm": 0.7715753417416791,
+      "learning_rate": 5.273119993884745e-08,
+      "loss": 0.2983,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9552446978709439,
+      "grad_norm": 0.7726584181603239,
+      "learning_rate": 5.2538912768276826e-08,
+      "loss": 0.2858,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9553267424211347,
+      "grad_norm": 0.7737672234062729,
+      "learning_rate": 5.234697497945973e-08,
+      "loss": 0.2791,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9554087869713255,
+      "grad_norm": 0.6925338263432202,
+      "learning_rate": 5.215538658595032e-08,
+      "loss": 0.2955,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9554908315215161,
+      "grad_norm": 0.7254993147516373,
+      "learning_rate": 5.196414760127777e-08,
+      "loss": 0.3223,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9555728760717069,
+      "grad_norm": 0.6967789586918134,
+      "learning_rate": 5.1773258038947395e-08,
+      "loss": 0.3158,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9556549206218977,
+      "grad_norm": 0.7081755538608755,
+      "learning_rate": 5.1582717912440076e-08,
+      "loss": 0.286,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9557369651720884,
+      "grad_norm": 0.6586506599722132,
+      "learning_rate": 5.13925272352106e-08,
+      "loss": 0.2612,
+      "step": 11649
+    },
+    {
+      "epoch": 0.9558190097222792,
+      "grad_norm": 0.7118524064473264,
+      "learning_rate": 5.120268602069101e-08,
+      "loss": 0.3236,
+      "step": 11650
+    },
+    {
+      "epoch": 0.95590105427247,
+      "grad_norm": 0.7846537587869404,
+      "learning_rate": 5.101319428228613e-08,
+      "loss": 0.3203,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9559830988226607,
+      "grad_norm": 0.7188852118761534,
+      "learning_rate": 5.082405203337859e-08,
+      "loss": 0.3012,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9560651433728514,
+      "grad_norm": 0.7022949622889041,
+      "learning_rate": 5.0635259287324934e-08,
+      "loss": 0.3367,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9561471879230422,
+      "grad_norm": 0.6692906195496103,
+      "learning_rate": 5.044681605745727e-08,
+      "loss": 0.2658,
+      "step": 11654
+    },
+    {
+      "epoch": 0.956229232473233,
+      "grad_norm": 0.6930987571676941,
+      "learning_rate": 5.025872235708384e-08,
+      "loss": 0.3102,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9563112770234237,
+      "grad_norm": 0.7179901071047708,
+      "learning_rate": 5.0070978199486806e-08,
+      "loss": 0.3282,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9563933215736145,
+      "grad_norm": 0.7337719477967835,
+      "learning_rate": 4.9883583597925e-08,
+      "loss": 0.3543,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9564753661238052,
+      "grad_norm": 0.6785531997986926,
+      "learning_rate": 4.9696538565631724e-08,
+      "loss": 0.2752,
+      "step": 11658
+    },
+    {
+      "epoch": 0.956557410673996,
+      "grad_norm": 0.7133175359435705,
+      "learning_rate": 4.9509843115814746e-08,
+      "loss": 0.2954,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9566394552241867,
+      "grad_norm": 0.6255715757558338,
+      "learning_rate": 4.9323497261659635e-08,
+      "loss": 0.2908,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9567214997743775,
+      "grad_norm": 0.7061489897289401,
+      "learning_rate": 4.9137501016325305e-08,
+      "loss": 0.2956,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9568035443245683,
+      "grad_norm": 0.7133243185531801,
+      "learning_rate": 4.8951854392946254e-08,
+      "loss": 0.2978,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9568855888747589,
+      "grad_norm": 1.372606863428361,
+      "learning_rate": 4.8766557404633095e-08,
+      "loss": 0.3192,
+      "step": 11663
+    },
+    {
+      "epoch": 0.9569676334249497,
+      "grad_norm": 0.6569735329067793,
+      "learning_rate": 4.858161006447038e-08,
+      "loss": 0.2794,
+      "step": 11664
+    },
+    {
+      "epoch": 0.9570496779751405,
+      "grad_norm": 0.7142479706436962,
+      "learning_rate": 4.8397012385519856e-08,
+      "loss": 0.2788,
+      "step": 11665
+    },
+    {
+      "epoch": 0.9571317225253313,
+      "grad_norm": 0.7414638337743316,
+      "learning_rate": 4.8212764380816676e-08,
+      "loss": 0.286,
+      "step": 11666
+    },
+    {
+      "epoch": 0.957213767075522,
+      "grad_norm": 0.6712293816685777,
+      "learning_rate": 4.802886606337209e-08,
+      "loss": 0.2693,
+      "step": 11667
+    },
+    {
+      "epoch": 0.9572958116257128,
+      "grad_norm": 0.730692845492134,
+      "learning_rate": 4.7845317446174044e-08,
+      "loss": 0.3289,
+      "step": 11668
+    },
+    {
+      "epoch": 0.9573778561759035,
+      "grad_norm": 0.6382819405406593,
+      "learning_rate": 4.766211854218217e-08,
+      "loss": 0.2954,
+      "step": 11669
+    },
+    {
+      "epoch": 0.9574599007260942,
+      "grad_norm": 0.6661417151995015,
+      "learning_rate": 4.747926936433611e-08,
+      "loss": 0.2632,
+      "step": 11670
+    },
+    {
+      "epoch": 0.957541945276285,
+      "grad_norm": 0.7053995824009092,
+      "learning_rate": 4.729676992554666e-08,
+      "loss": 0.2823,
+      "step": 11671
+    },
+    {
+      "epoch": 0.9576239898264758,
+      "grad_norm": 0.6989509434702533,
+      "learning_rate": 4.711462023870239e-08,
+      "loss": 0.3135,
+      "step": 11672
+    },
+    {
+      "epoch": 0.9577060343766666,
+      "grad_norm": 0.6798632915550349,
+      "learning_rate": 4.693282031666579e-08,
+      "loss": 0.2947,
+      "step": 11673
+    },
+    {
+      "epoch": 0.9577880789268572,
+      "grad_norm": 0.7092252390311168,
+      "learning_rate": 4.6751370172276044e-08,
+      "loss": 0.2924,
+      "step": 11674
+    },
+    {
+      "epoch": 0.957870123477048,
+      "grad_norm": 0.714456566258489,
+      "learning_rate": 4.657026981834623e-08,
+      "loss": 0.2891,
+      "step": 11675
+    },
+    {
+      "epoch": 0.9579521680272388,
+      "grad_norm": 0.6729717548896826,
+      "learning_rate": 4.6389519267666107e-08,
+      "loss": 0.2984,
+      "step": 11676
+    },
+    {
+      "epoch": 0.9580342125774295,
+      "grad_norm": 0.7285435134129544,
+      "learning_rate": 4.6209118532999365e-08,
+      "loss": 0.3122,
+      "step": 11677
+    },
+    {
+      "epoch": 0.9581162571276203,
+      "grad_norm": 0.6540854137882476,
+      "learning_rate": 4.602906762708526e-08,
+      "loss": 0.2646,
+      "step": 11678
+    },
+    {
+      "epoch": 0.9581983016778111,
+      "grad_norm": 0.8733006772219206,
+      "learning_rate": 4.584936656264027e-08,
+      "loss": 0.3085,
+      "step": 11679
+    },
+    {
+      "epoch": 0.9582803462280018,
+      "grad_norm": 0.6558224495342561,
+      "learning_rate": 4.56700153523526e-08,
+      "loss": 0.2843,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9583623907781925,
+      "grad_norm": 0.7102350170094677,
+      "learning_rate": 4.549101400888933e-08,
+      "loss": 0.2926,
+      "step": 11681
+    },
+    {
+      "epoch": 0.9584444353283833,
+      "grad_norm": 0.6936138810247945,
+      "learning_rate": 4.531236254489035e-08,
+      "loss": 0.3222,
+      "step": 11682
+    },
+    {
+      "epoch": 0.9585264798785741,
+      "grad_norm": 0.69741783499237,
+      "learning_rate": 4.513406097297224e-08,
+      "loss": 0.2823,
+      "step": 11683
+    },
+    {
+      "epoch": 0.9586085244287649,
+      "grad_norm": 0.6971399339603158,
+      "learning_rate": 4.495610930572603e-08,
+      "loss": 0.3291,
+      "step": 11684
+    },
+    {
+      "epoch": 0.9586905689789555,
+      "grad_norm": 0.7021632965185957,
+      "learning_rate": 4.4778507555718886e-08,
+      "loss": 0.2693,
+      "step": 11685
+    },
+    {
+      "epoch": 0.9587726135291463,
+      "grad_norm": 0.697082191378743,
+      "learning_rate": 4.46012557354919e-08,
+      "loss": 0.3289,
+      "step": 11686
+    },
+    {
+      "epoch": 0.9588546580793371,
+      "grad_norm": 0.7462864331191027,
+      "learning_rate": 4.442435385756283e-08,
+      "loss": 0.2774,
+      "step": 11687
+    },
+    {
+      "epoch": 0.9589367026295278,
+      "grad_norm": 0.666601498078659,
+      "learning_rate": 4.4247801934423904e-08,
+      "loss": 0.2806,
+      "step": 11688
+    },
+    {
+      "epoch": 0.9590187471797186,
+      "grad_norm": 0.7529304183853978,
+      "learning_rate": 4.407159997854349e-08,
+      "loss": 0.3414,
+      "step": 11689
+    },
+    {
+      "epoch": 0.9591007917299094,
+      "grad_norm": 0.6770672095762997,
+      "learning_rate": 4.3895748002364404e-08,
+      "loss": 0.2951,
+      "step": 11690
+    },
+    {
+      "epoch": 0.9591828362801001,
+      "grad_norm": 0.7452582678387977,
+      "learning_rate": 4.372024601830449e-08,
+      "loss": 0.3152,
+      "step": 11691
+    },
+    {
+      "epoch": 0.9592648808302908,
+      "grad_norm": 0.7028363539932324,
+      "learning_rate": 4.354509403875884e-08,
+      "loss": 0.2903,
+      "step": 11692
+    },
+    {
+      "epoch": 0.9593469253804816,
+      "grad_norm": 0.6892106612488396,
+      "learning_rate": 4.337029207609478e-08,
+      "loss": 0.2656,
+      "step": 11693
+    },
+    {
+      "epoch": 0.9594289699306724,
+      "grad_norm": 0.7650994410200068,
+      "learning_rate": 4.319584014265743e-08,
+      "loss": 0.2977,
+      "step": 11694
+    },
+    {
+      "epoch": 0.9595110144808631,
+      "grad_norm": 0.6584422767267282,
+      "learning_rate": 4.302173825076639e-08,
+      "loss": 0.2971,
+      "step": 11695
+    },
+    {
+      "epoch": 0.9595930590310539,
+      "grad_norm": 0.7125131309122773,
+      "learning_rate": 4.28479864127157e-08,
+      "loss": 0.3184,
+      "step": 11696
+    },
+    {
+      "epoch": 0.9596751035812446,
+      "grad_norm": 0.7090560816354273,
+      "learning_rate": 4.2674584640776676e-08,
+      "loss": 0.3027,
+      "step": 11697
+    },
+    {
+      "epoch": 0.9597571481314354,
+      "grad_norm": 0.6974034583373845,
+      "learning_rate": 4.250153294719339e-08,
+      "loss": 0.2867,
+      "step": 11698
+    },
+    {
+      "epoch": 0.9598391926816261,
+      "grad_norm": 0.6316702881271453,
+      "learning_rate": 4.2328831344187196e-08,
+      "loss": 0.2717,
+      "step": 11699
+    },
+    {
+      "epoch": 0.9599212372318169,
+      "grad_norm": 0.6908986167203036,
+      "learning_rate": 4.215647984395388e-08,
+      "loss": 0.3003,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9600032817820077,
+      "grad_norm": 0.6703579958197343,
+      "learning_rate": 4.1984478458664265e-08,
+      "loss": 0.2606,
+      "step": 11701
+    },
+    {
+      "epoch": 0.9600853263321983,
+      "grad_norm": 0.7046171578916499,
+      "learning_rate": 4.181282720046531e-08,
+      "loss": 0.2838,
+      "step": 11702
+    },
+    {
+      "epoch": 0.9601673708823891,
+      "grad_norm": 0.8037567147224653,
+      "learning_rate": 4.164152608147842e-08,
+      "loss": 0.3228,
+      "step": 11703
+    },
+    {
+      "epoch": 0.9602494154325799,
+      "grad_norm": 0.8229693507182569,
+      "learning_rate": 4.147057511380115e-08,
+      "loss": 0.2708,
+      "step": 11704
+    },
+    {
+      "epoch": 0.9603314599827707,
+      "grad_norm": 0.7746577918682171,
+      "learning_rate": 4.129997430950494e-08,
+      "loss": 0.3492,
+      "step": 11705
+    },
+    {
+      "epoch": 0.9604135045329614,
+      "grad_norm": 0.6767792692808359,
+      "learning_rate": 4.112972368063794e-08,
+      "loss": 0.3111,
+      "step": 11706
+    },
+    {
+      "epoch": 0.9604955490831522,
+      "grad_norm": 0.7372446315421658,
+      "learning_rate": 4.095982323922332e-08,
+      "loss": 0.2873,
+      "step": 11707
+    },
+    {
+      "epoch": 0.9605775936333429,
+      "grad_norm": 0.7454254549261135,
+      "learning_rate": 4.079027299725757e-08,
+      "loss": 0.3324,
+      "step": 11708
+    },
+    {
+      "epoch": 0.9606596381835336,
+      "grad_norm": 0.6539531795984255,
+      "learning_rate": 4.0621072966716135e-08,
+      "loss": 0.2367,
+      "step": 11709
+    },
+    {
+      "epoch": 0.9607416827337244,
+      "grad_norm": 0.687637724582446,
+      "learning_rate": 4.045222315954611e-08,
+      "loss": 0.2888,
+      "step": 11710
+    },
+    {
+      "epoch": 0.9608237272839152,
+      "grad_norm": 0.7488082251629669,
+      "learning_rate": 4.0283723587672405e-08,
+      "loss": 0.2818,
+      "step": 11711
+    },
+    {
+      "epoch": 0.960905771834106,
+      "grad_norm": 0.6651036890354699,
+      "learning_rate": 4.0115574262993284e-08,
+      "loss": 0.2917,
+      "step": 11712
+    },
+    {
+      "epoch": 0.9609878163842966,
+      "grad_norm": 0.732288267477677,
+      "learning_rate": 3.994777519738424e-08,
+      "loss": 0.2852,
+      "step": 11713
+    },
+    {
+      "epoch": 0.9610698609344874,
+      "grad_norm": 1.6232059362851552,
+      "learning_rate": 3.9780326402694135e-08,
+      "loss": 0.3334,
+      "step": 11714
+    },
+    {
+      "epoch": 0.9611519054846782,
+      "grad_norm": 0.7130187670321327,
+      "learning_rate": 3.961322789074795e-08,
+      "loss": 0.326,
+      "step": 11715
+    },
+    {
+      "epoch": 0.9612339500348689,
+      "grad_norm": 0.75702686297671,
+      "learning_rate": 3.944647967334625e-08,
+      "loss": 0.27,
+      "step": 11716
+    },
+    {
+      "epoch": 0.9613159945850597,
+      "grad_norm": 0.7002604275805014,
+      "learning_rate": 3.928008176226461e-08,
+      "loss": 0.3516,
+      "step": 11717
+    },
+    {
+      "epoch": 0.9613980391352505,
+      "grad_norm": 0.7685242146912565,
+      "learning_rate": 3.911403416925308e-08,
+      "loss": 0.3241,
+      "step": 11718
+    },
+    {
+      "epoch": 0.9614800836854412,
+      "grad_norm": 0.8400916865389251,
+      "learning_rate": 3.894833690603839e-08,
+      "loss": 0.3502,
+      "step": 11719
+    },
+    {
+      "epoch": 0.9615621282356319,
+      "grad_norm": 0.7574854575223104,
+      "learning_rate": 3.878298998432228e-08,
+      "loss": 0.263,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9616441727858227,
+      "grad_norm": 0.7051589145922943,
+      "learning_rate": 3.8617993415779876e-08,
+      "loss": 0.2635,
+      "step": 11721
+    },
+    {
+      "epoch": 0.9617262173360135,
+      "grad_norm": 0.7526586435886291,
+      "learning_rate": 3.8453347212064064e-08,
+      "loss": 0.3232,
+      "step": 11722
+    },
+    {
+      "epoch": 0.9618082618862042,
+      "grad_norm": 0.664154980328035,
+      "learning_rate": 3.82890513848011e-08,
+      "loss": 0.2632,
+      "step": 11723
+    },
+    {
+      "epoch": 0.961890306436395,
+      "grad_norm": 0.6282909310540896,
+      "learning_rate": 3.8125105945593935e-08,
+      "loss": 0.2822,
+      "step": 11724
+    },
+    {
+      "epoch": 0.9619723509865857,
+      "grad_norm": 0.7261969774121017,
+      "learning_rate": 3.7961510906020516e-08,
+      "loss": 0.2882,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9620543955367765,
+      "grad_norm": 0.8195329686789037,
+      "learning_rate": 3.779826627763272e-08,
+      "loss": 0.3189,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9621364400869672,
+      "grad_norm": 0.862099040234421,
+      "learning_rate": 3.763537207195855e-08,
+      "loss": 0.2856,
+      "step": 11727
+    },
+    {
+      "epoch": 0.962218484637158,
+      "grad_norm": 0.7463895037789964,
+      "learning_rate": 3.747282830050214e-08,
+      "loss": 0.2851,
+      "step": 11728
+    },
+    {
+      "epoch": 0.9623005291873488,
+      "grad_norm": 0.8460156573202277,
+      "learning_rate": 3.731063497474152e-08,
+      "loss": 0.2818,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9623825737375394,
+      "grad_norm": 0.7264107458137816,
+      "learning_rate": 3.714879210613087e-08,
+      "loss": 0.3083,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9624646182877302,
+      "grad_norm": 0.635855808899441,
+      "learning_rate": 3.698729970609882e-08,
+      "loss": 0.2858,
+      "step": 11731
+    },
+    {
+      "epoch": 0.962546662837921,
+      "grad_norm": 0.675665825486257,
+      "learning_rate": 3.6826157786050144e-08,
+      "loss": 0.2846,
+      "step": 11732
+    },
+    {
+      "epoch": 0.9626287073881118,
+      "grad_norm": 0.6619130338425246,
+      "learning_rate": 3.666536635736406e-08,
+      "loss": 0.2941,
+      "step": 11733
+    },
+    {
+      "epoch": 0.9627107519383025,
+      "grad_norm": 0.7752647428431868,
+      "learning_rate": 3.6504925431395946e-08,
+      "loss": 0.2843,
+      "step": 11734
+    },
+    {
+      "epoch": 0.9627927964884933,
+      "grad_norm": 0.6998958968531642,
+      "learning_rate": 3.634483501947561e-08,
+      "loss": 0.3075,
+      "step": 11735
+    },
+    {
+      "epoch": 0.962874841038684,
+      "grad_norm": 0.6616961092635475,
+      "learning_rate": 3.618509513290791e-08,
+      "loss": 0.2953,
+      "step": 11736
+    },
+    {
+      "epoch": 0.9629568855888747,
+      "grad_norm": 0.6868409984402457,
+      "learning_rate": 3.602570578297382e-08,
+      "loss": 0.2862,
+      "step": 11737
+    },
+    {
+      "epoch": 0.9630389301390655,
+      "grad_norm": 0.7933610086532551,
+      "learning_rate": 3.5866666980929334e-08,
+      "loss": 0.2826,
+      "step": 11738
+    },
+    {
+      "epoch": 0.9631209746892563,
+      "grad_norm": 0.6622524124392871,
+      "learning_rate": 3.5707978738005464e-08,
+      "loss": 0.2536,
+      "step": 11739
+    },
+    {
+      "epoch": 0.9632030192394471,
+      "grad_norm": 0.774683476027425,
+      "learning_rate": 3.554964106540826e-08,
+      "loss": 0.2984,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9632850637896377,
+      "grad_norm": 0.6862867604699145,
+      "learning_rate": 3.539165397431932e-08,
+      "loss": 0.3343,
+      "step": 11741
+    },
+    {
+      "epoch": 0.9633671083398285,
+      "grad_norm": 0.6733606598884274,
+      "learning_rate": 3.5234017475895276e-08,
+      "loss": 0.2592,
+      "step": 11742
+    },
+    {
+      "epoch": 0.9634491528900193,
+      "grad_norm": 0.7301835799093065,
+      "learning_rate": 3.5076731581268896e-08,
+      "loss": 0.2944,
+      "step": 11743
+    },
+    {
+      "epoch": 0.96353119744021,
+      "grad_norm": 0.6643222751507747,
+      "learning_rate": 3.491979630154685e-08,
+      "loss": 0.2583,
+      "step": 11744
+    },
+    {
+      "epoch": 0.9636132419904008,
+      "grad_norm": 0.7248130730780148,
+      "learning_rate": 3.4763211647811377e-08,
+      "loss": 0.3023,
+      "step": 11745
+    },
+    {
+      "epoch": 0.9636952865405916,
+      "grad_norm": 0.685152509358323,
+      "learning_rate": 3.460697763112142e-08,
+      "loss": 0.3034,
+      "step": 11746
+    },
+    {
+      "epoch": 0.9637773310907823,
+      "grad_norm": 0.7160678202968648,
+      "learning_rate": 3.44510942625087e-08,
+      "loss": 0.2614,
+      "step": 11747
+    },
+    {
+      "epoch": 0.963859375640973,
+      "grad_norm": 0.6697934647932404,
+      "learning_rate": 3.429556155298219e-08,
+      "loss": 0.334,
+      "step": 11748
+    },
+    {
+      "epoch": 0.9639414201911638,
+      "grad_norm": 0.7463842793250114,
+      "learning_rate": 3.414037951352478e-08,
+      "loss": 0.3067,
+      "step": 11749
+    },
+    {
+      "epoch": 0.9640234647413546,
+      "grad_norm": 0.7444659406878115,
+      "learning_rate": 3.398554815509547e-08,
+      "loss": 0.3208,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9641055092915453,
+      "grad_norm": 0.6459098384704235,
+      "learning_rate": 3.38310674886283e-08,
+      "loss": 0.2908,
+      "step": 11751
+    },
+    {
+      "epoch": 0.964187553841736,
+      "grad_norm": 0.7263679514888431,
+      "learning_rate": 3.3676937525032314e-08,
+      "loss": 0.2929,
+      "step": 11752
+    },
+    {
+      "epoch": 0.9642695983919268,
+      "grad_norm": 0.6551739102594666,
+      "learning_rate": 3.35231582751927e-08,
+      "loss": 0.3357,
+      "step": 11753
+    },
+    {
+      "epoch": 0.9643516429421176,
+      "grad_norm": 0.6933551625860638,
+      "learning_rate": 3.336972974996799e-08,
+      "loss": 0.3413,
+      "step": 11754
+    },
+    {
+      "epoch": 0.9644336874923083,
+      "grad_norm": 0.6829270029722005,
+      "learning_rate": 3.321665196019286e-08,
+      "loss": 0.2654,
+      "step": 11755
+    },
+    {
+      "epoch": 0.9645157320424991,
+      "grad_norm": 0.6991471688637967,
+      "learning_rate": 3.306392491667865e-08,
+      "loss": 0.289,
+      "step": 11756
+    },
+    {
+      "epoch": 0.9645977765926899,
+      "grad_norm": 0.7392448930717189,
+      "learning_rate": 3.291154863021007e-08,
+      "loss": 0.3025,
+      "step": 11757
+    },
+    {
+      "epoch": 0.9646798211428805,
+      "grad_norm": 0.7546942825520505,
+      "learning_rate": 3.27595231115474e-08,
+      "loss": 0.2715,
+      "step": 11758
+    },
+    {
+      "epoch": 0.9647618656930713,
+      "grad_norm": 0.7072459610522315,
+      "learning_rate": 3.260784837142705e-08,
+      "loss": 0.2704,
+      "step": 11759
+    },
+    {
+      "epoch": 0.9648439102432621,
+      "grad_norm": 0.6661378087313085,
+      "learning_rate": 3.2456524420559354e-08,
+      "loss": 0.313,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9649259547934529,
+      "grad_norm": 0.7157668773873078,
+      "learning_rate": 3.230555126963131e-08,
+      "loss": 0.3355,
+      "step": 11761
+    },
+    {
+      "epoch": 0.9650079993436436,
+      "grad_norm": 0.8191401978842636,
+      "learning_rate": 3.215492892930383e-08,
+      "loss": 0.3107,
+      "step": 11762
+    },
+    {
+      "epoch": 0.9650900438938343,
+      "grad_norm": 0.6897975781707293,
+      "learning_rate": 3.200465741021341e-08,
+      "loss": 0.301,
+      "step": 11763
+    },
+    {
+      "epoch": 0.9651720884440251,
+      "grad_norm": 0.843358889834847,
+      "learning_rate": 3.185473672297323e-08,
+      "loss": 0.3408,
+      "step": 11764
+    },
+    {
+      "epoch": 0.9652541329942158,
+      "grad_norm": 0.7510466055212545,
+      "learning_rate": 3.170516687816871e-08,
+      "loss": 0.3402,
+      "step": 11765
+    },
+    {
+      "epoch": 0.9653361775444066,
+      "grad_norm": 0.6529991969263421,
+      "learning_rate": 3.1555947886363626e-08,
+      "loss": 0.2778,
+      "step": 11766
+    },
+    {
+      "epoch": 0.9654182220945974,
+      "grad_norm": 0.7107735699270215,
+      "learning_rate": 3.1407079758095646e-08,
+      "loss": 0.3404,
+      "step": 11767
+    },
+    {
+      "epoch": 0.9655002666447882,
+      "grad_norm": 0.6966059195690419,
+      "learning_rate": 3.125856250387638e-08,
+      "loss": 0.3534,
+      "step": 11768
+    },
+    {
+      "epoch": 0.9655823111949788,
+      "grad_norm": 0.7103916350377413,
+      "learning_rate": 3.111039613419464e-08,
+      "loss": 0.2756,
+      "step": 11769
+    },
+    {
+      "epoch": 0.9656643557451696,
+      "grad_norm": 0.6936868108420756,
+      "learning_rate": 3.096258065951319e-08,
+      "loss": 0.3267,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9657464002953604,
+      "grad_norm": 0.662406752577065,
+      "learning_rate": 3.081511609027144e-08,
+      "loss": 0.2776,
+      "step": 11771
+    },
+    {
+      "epoch": 0.9658284448455511,
+      "grad_norm": 0.7209581296349444,
+      "learning_rate": 3.066800243688273e-08,
+      "loss": 0.2985,
+      "step": 11772
+    },
+    {
+      "epoch": 0.9659104893957419,
+      "grad_norm": 0.7102255419906799,
+      "learning_rate": 3.052123970973542e-08,
+      "loss": 0.292,
+      "step": 11773
+    },
+    {
+      "epoch": 0.9659925339459327,
+      "grad_norm": 0.7185573342960094,
+      "learning_rate": 3.0374827919193994e-08,
+      "loss": 0.3165,
+      "step": 11774
+    },
+    {
+      "epoch": 0.9660745784961234,
+      "grad_norm": 0.6726121937627337,
+      "learning_rate": 3.022876707559796e-08,
+      "loss": 0.3118,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9661566230463141,
+      "grad_norm": 0.6976511000687883,
+      "learning_rate": 3.008305718926241e-08,
+      "loss": 0.2794,
+      "step": 11776
+    },
+    {
+      "epoch": 0.9662386675965049,
+      "grad_norm": 0.7442440970151887,
+      "learning_rate": 2.9937698270476324e-08,
+      "loss": 0.3299,
+      "step": 11777
+    },
+    {
+      "epoch": 0.9663207121466957,
+      "grad_norm": 0.7935176306709048,
+      "learning_rate": 2.979269032950427e-08,
+      "loss": 0.2967,
+      "step": 11778
+    },
+    {
+      "epoch": 0.9664027566968864,
+      "grad_norm": 0.7844202432870817,
+      "learning_rate": 2.9648033376588058e-08,
+      "loss": 0.3232,
+      "step": 11779
+    },
+    {
+      "epoch": 0.9664848012470771,
+      "grad_norm": 0.6478186246615673,
+      "learning_rate": 2.950372742194174e-08,
+      "loss": 0.3054,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9665668457972679,
+      "grad_norm": 0.6835916286926742,
+      "learning_rate": 2.9359772475757164e-08,
+      "loss": 0.2981,
+      "step": 11781
+    },
+    {
+      "epoch": 0.9666488903474587,
+      "grad_norm": 0.7080681723579166,
+      "learning_rate": 2.9216168548198975e-08,
+      "loss": 0.2975,
+      "step": 11782
+    },
+    {
+      "epoch": 0.9667309348976494,
+      "grad_norm": 0.6719710670099752,
+      "learning_rate": 2.9072915649408505e-08,
+      "loss": 0.2796,
+      "step": 11783
+    },
+    {
+      "epoch": 0.9668129794478402,
+      "grad_norm": 0.6566857922021959,
+      "learning_rate": 2.893001378950322e-08,
+      "loss": 0.3149,
+      "step": 11784
+    },
+    {
+      "epoch": 0.966895023998031,
+      "grad_norm": 0.7279379061059654,
+      "learning_rate": 2.8787462978572823e-08,
+      "loss": 0.277,
+      "step": 11785
+    },
+    {
+      "epoch": 0.9669770685482216,
+      "grad_norm": 0.6858089075669881,
+      "learning_rate": 2.864526322668537e-08,
+      "loss": 0.2878,
+      "step": 11786
+    },
+    {
+      "epoch": 0.9670591130984124,
+      "grad_norm": 0.654258835322203,
+      "learning_rate": 2.850341454388228e-08,
+      "loss": 0.2985,
+      "step": 11787
+    },
+    {
+      "epoch": 0.9671411576486032,
+      "grad_norm": 0.8938497180491137,
+      "learning_rate": 2.8361916940180534e-08,
+      "loss": 0.288,
+      "step": 11788
+    },
+    {
+      "epoch": 0.967223202198794,
+      "grad_norm": 1.037729649311417,
+      "learning_rate": 2.8220770425573253e-08,
+      "loss": 0.2971,
+      "step": 11789
+    },
+    {
+      "epoch": 0.9673052467489847,
+      "grad_norm": 0.6881692190811911,
+      "learning_rate": 2.8079975010026904e-08,
+      "loss": 0.3006,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9673872912991754,
+      "grad_norm": 0.7383589309523635,
+      "learning_rate": 2.7939530703484652e-08,
+      "loss": 0.3022,
+      "step": 11791
+    },
+    {
+      "epoch": 0.9674693358493662,
+      "grad_norm": 0.7378144101709202,
+      "learning_rate": 2.7799437515864668e-08,
+      "loss": 0.2711,
+      "step": 11792
+    },
+    {
+      "epoch": 0.9675513803995569,
+      "grad_norm": 0.6836823333985783,
+      "learning_rate": 2.76596954570596e-08,
+      "loss": 0.2814,
+      "step": 11793
+    },
+    {
+      "epoch": 0.9676334249497477,
+      "grad_norm": 0.7107867160544855,
+      "learning_rate": 2.752030453693877e-08,
+      "loss": 0.2936,
+      "step": 11794
+    },
+    {
+      "epoch": 0.9677154694999385,
+      "grad_norm": 0.6976072985645755,
+      "learning_rate": 2.7381264765344862e-08,
+      "loss": 0.2852,
+      "step": 11795
+    },
+    {
+      "epoch": 0.9677975140501293,
+      "grad_norm": 0.6754633412006988,
+      "learning_rate": 2.7242576152097245e-08,
+      "loss": 0.2965,
+      "step": 11796
+    },
+    {
+      "epoch": 0.9678795586003199,
+      "grad_norm": 0.6905375990808529,
+      "learning_rate": 2.7104238706989194e-08,
+      "loss": 0.2951,
+      "step": 11797
+    },
+    {
+      "epoch": 0.9679616031505107,
+      "grad_norm": 0.7672391623296835,
+      "learning_rate": 2.696625243979012e-08,
+      "loss": 0.3041,
+      "step": 11798
+    },
+    {
+      "epoch": 0.9680436477007015,
+      "grad_norm": 0.635035930184306,
+      "learning_rate": 2.6828617360244448e-08,
+      "loss": 0.2554,
+      "step": 11799
+    },
+    {
+      "epoch": 0.9681256922508923,
+      "grad_norm": 0.703098335597188,
+      "learning_rate": 2.669133347807218e-08,
+      "loss": 0.2935,
+      "step": 11800
+    },
+    {
+      "epoch": 0.968207736801083,
+      "grad_norm": 0.7716559218070024,
+      "learning_rate": 2.6554400802967785e-08,
+      "loss": 0.297,
+      "step": 11801
+    },
+    {
+      "epoch": 0.9682897813512737,
+      "grad_norm": 0.7600061280752313,
+      "learning_rate": 2.6417819344600747e-08,
+      "loss": 0.3381,
+      "step": 11802
+    },
+    {
+      "epoch": 0.9683718259014645,
+      "grad_norm": 0.7281152095304223,
+      "learning_rate": 2.628158911261669e-08,
+      "loss": 0.3026,
+      "step": 11803
+    },
+    {
+      "epoch": 0.9684538704516552,
+      "grad_norm": 0.878117359464687,
+      "learning_rate": 2.6145710116636246e-08,
+      "loss": 0.3277,
+      "step": 11804
+    },
+    {
+      "epoch": 0.968535915001846,
+      "grad_norm": 0.7045401183836395,
+      "learning_rate": 2.6010182366254523e-08,
+      "loss": 0.2962,
+      "step": 11805
+    },
+    {
+      "epoch": 0.9686179595520368,
+      "grad_norm": 0.8132685235714541,
+      "learning_rate": 2.5875005871042192e-08,
+      "loss": 0.267,
+      "step": 11806
+    },
+    {
+      "epoch": 0.9687000041022276,
+      "grad_norm": 0.6492678706933637,
+      "learning_rate": 2.574018064054551e-08,
+      "loss": 0.2926,
+      "step": 11807
+    },
+    {
+      "epoch": 0.9687820486524182,
+      "grad_norm": 0.8165854682513639,
+      "learning_rate": 2.5605706684285747e-08,
+      "loss": 0.2945,
+      "step": 11808
+    },
+    {
+      "epoch": 0.968864093202609,
+      "grad_norm": 0.7117295351403994,
+      "learning_rate": 2.5471584011758645e-08,
+      "loss": 0.3012,
+      "step": 11809
+    },
+    {
+      "epoch": 0.9689461377527998,
+      "grad_norm": 0.8033442978484137,
+      "learning_rate": 2.533781263243662e-08,
+      "loss": 0.349,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9690281823029905,
+      "grad_norm": 0.7181463202292038,
+      "learning_rate": 2.5204392555765456e-08,
+      "loss": 0.311,
+      "step": 11811
+    },
+    {
+      "epoch": 0.9691102268531813,
+      "grad_norm": 0.6781159597992863,
+      "learning_rate": 2.5071323791167058e-08,
+      "loss": 0.3055,
+      "step": 11812
+    },
+    {
+      "epoch": 0.969192271403372,
+      "grad_norm": 0.87114516363029,
+      "learning_rate": 2.4938606348040017e-08,
+      "loss": 0.282,
+      "step": 11813
+    },
+    {
+      "epoch": 0.9692743159535628,
+      "grad_norm": 1.1970908837802041,
+      "learning_rate": 2.4806240235754618e-08,
+      "loss": 0.3029,
+      "step": 11814
+    },
+    {
+      "epoch": 0.9693563605037535,
+      "grad_norm": 0.6641540085242239,
+      "learning_rate": 2.4674225463659495e-08,
+      "loss": 0.2972,
+      "step": 11815
+    },
+    {
+      "epoch": 0.9694384050539443,
+      "grad_norm": 0.6413399175726497,
+      "learning_rate": 2.454256204107719e-08,
+      "loss": 0.2947,
+      "step": 11816
+    },
+    {
+      "epoch": 0.9695204496041351,
+      "grad_norm": 0.6620351853207994,
+      "learning_rate": 2.4411249977305264e-08,
+      "loss": 0.2843,
+      "step": 11817
+    },
+    {
+      "epoch": 0.9696024941543258,
+      "grad_norm": 0.7497840617845648,
+      "learning_rate": 2.4280289281617407e-08,
+      "loss": 0.3096,
+      "step": 11818
+    },
+    {
+      "epoch": 0.9696845387045165,
+      "grad_norm": 0.7379717583049524,
+      "learning_rate": 2.414967996326123e-08,
+      "loss": 0.2855,
+      "step": 11819
+    },
+    {
+      "epoch": 0.9697665832547073,
+      "grad_norm": 0.7237847515912483,
+      "learning_rate": 2.401942203146046e-08,
+      "loss": 0.2874,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9698486278048981,
+      "grad_norm": 1.020435977137287,
+      "learning_rate": 2.3889515495413297e-08,
+      "loss": 0.2883,
+      "step": 11821
+    },
+    {
+      "epoch": 0.9699306723550888,
+      "grad_norm": 0.729852470527345,
+      "learning_rate": 2.3759960364294067e-08,
+      "loss": 0.2779,
+      "step": 11822
+    },
+    {
+      "epoch": 0.9700127169052796,
+      "grad_norm": 0.6880037615420559,
+      "learning_rate": 2.363075664725156e-08,
+      "loss": 0.2896,
+      "step": 11823
+    },
+    {
+      "epoch": 0.9700947614554704,
+      "grad_norm": 1.5304901313436763,
+      "learning_rate": 2.3501904353409598e-08,
+      "loss": 0.2952,
+      "step": 11824
+    },
+    {
+      "epoch": 0.970176806005661,
+      "grad_norm": 0.7313967453619797,
+      "learning_rate": 2.3373403491867562e-08,
+      "loss": 0.3152,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9702588505558518,
+      "grad_norm": 0.7401796146652507,
+      "learning_rate": 2.3245254071700418e-08,
+      "loss": 0.2711,
+      "step": 11826
+    },
+    {
+      "epoch": 0.9703408951060426,
+      "grad_norm": 0.7694185180377083,
+      "learning_rate": 2.3117456101958148e-08,
+      "loss": 0.332,
+      "step": 11827
+    },
+    {
+      "epoch": 0.9704229396562334,
+      "grad_norm": 0.6548918451824549,
+      "learning_rate": 2.2990009591664642e-08,
+      "loss": 0.3057,
+      "step": 11828
+    },
+    {
+      "epoch": 0.9705049842064241,
+      "grad_norm": 0.7018499720590083,
+      "learning_rate": 2.286291454982048e-08,
+      "loss": 0.3194,
+      "step": 11829
+    },
+    {
+      "epoch": 0.9705870287566148,
+      "grad_norm": 0.6994378354522104,
+      "learning_rate": 2.2736170985401262e-08,
+      "loss": 0.2878,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9706690733068056,
+      "grad_norm": 0.7262117096782068,
+      "learning_rate": 2.260977890735705e-08,
+      "loss": 0.2994,
+      "step": 11831
+    },
+    {
+      "epoch": 0.9707511178569963,
+      "grad_norm": 0.727453945882473,
+      "learning_rate": 2.2483738324612924e-08,
+      "loss": 0.3073,
+      "step": 11832
+    },
+    {
+      "epoch": 0.9708331624071871,
+      "grad_norm": 0.7953501504770691,
+      "learning_rate": 2.2358049246070658e-08,
+      "loss": 0.3249,
+      "step": 11833
+    },
+    {
+      "epoch": 0.9709152069573779,
+      "grad_norm": 0.6890075219046383,
+      "learning_rate": 2.2232711680605368e-08,
+      "loss": 0.2744,
+      "step": 11834
+    },
+    {
+      "epoch": 0.9709972515075687,
+      "grad_norm": 0.7290735435767984,
+      "learning_rate": 2.210772563706942e-08,
+      "loss": 0.3285,
+      "step": 11835
+    },
+    {
+      "epoch": 0.9710792960577593,
+      "grad_norm": 0.6560437490470241,
+      "learning_rate": 2.1983091124287426e-08,
+      "loss": 0.3188,
+      "step": 11836
+    },
+    {
+      "epoch": 0.9711613406079501,
+      "grad_norm": 0.7312078717881314,
+      "learning_rate": 2.185880815106234e-08,
+      "loss": 0.2816,
+      "step": 11837
+    },
+    {
+      "epoch": 0.9712433851581409,
+      "grad_norm": 0.646126352171407,
+      "learning_rate": 2.1734876726169918e-08,
+      "loss": 0.3005,
+      "step": 11838
+    },
+    {
+      "epoch": 0.9713254297083316,
+      "grad_norm": 0.6586355376686953,
+      "learning_rate": 2.1611296858362052e-08,
+      "loss": 0.3003,
+      "step": 11839
+    },
+    {
+      "epoch": 0.9714074742585224,
+      "grad_norm": 0.7295545429498,
+      "learning_rate": 2.1488068556366205e-08,
+      "loss": 0.2981,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9714895188087131,
+      "grad_norm": 0.6969707480221857,
+      "learning_rate": 2.1365191828884857e-08,
+      "loss": 0.2922,
+      "step": 11841
+    },
+    {
+      "epoch": 0.9715715633589039,
+      "grad_norm": 0.6974483829908057,
+      "learning_rate": 2.1242666684594405e-08,
+      "loss": 0.3069,
+      "step": 11842
+    },
+    {
+      "epoch": 0.9716536079090946,
+      "grad_norm": 0.7203085632174782,
+      "learning_rate": 2.1120493132147924e-08,
+      "loss": 0.314,
+      "step": 11843
+    },
+    {
+      "epoch": 0.9717356524592854,
+      "grad_norm": 0.7138719014747184,
+      "learning_rate": 2.0998671180172957e-08,
+      "loss": 0.3043,
+      "step": 11844
+    },
+    {
+      "epoch": 0.9718176970094762,
+      "grad_norm": 0.7434399579216556,
+      "learning_rate": 2.0877200837272626e-08,
+      "loss": 0.3082,
+      "step": 11845
+    },
+    {
+      "epoch": 0.9718997415596669,
+      "grad_norm": 0.8450178906984308,
+      "learning_rate": 2.075608211202451e-08,
+      "loss": 0.2937,
+      "step": 11846
+    },
+    {
+      "epoch": 0.9719817861098576,
+      "grad_norm": 0.7231087220832549,
+      "learning_rate": 2.0635315012982325e-08,
+      "loss": 0.3024,
+      "step": 11847
+    },
+    {
+      "epoch": 0.9720638306600484,
+      "grad_norm": 0.687489386650773,
+      "learning_rate": 2.051489954867425e-08,
+      "loss": 0.2691,
+      "step": 11848
+    },
+    {
+      "epoch": 0.9721458752102392,
+      "grad_norm": 0.78895999041588,
+      "learning_rate": 2.039483572760348e-08,
+      "loss": 0.3105,
+      "step": 11849
+    },
+    {
+      "epoch": 0.9722279197604299,
+      "grad_norm": 0.777166501845725,
+      "learning_rate": 2.02751235582499e-08,
+      "loss": 0.3303,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9723099643106207,
+      "grad_norm": 0.8006298078922864,
+      "learning_rate": 2.015576304906619e-08,
+      "loss": 0.2776,
+      "step": 11851
+    },
+    {
+      "epoch": 0.9723920088608115,
+      "grad_norm": 0.7090074049924731,
+      "learning_rate": 2.003675420848117e-08,
+      "loss": 0.3158,
+      "step": 11852
+    },
+    {
+      "epoch": 0.9724740534110021,
+      "grad_norm": 0.6808212569354767,
+      "learning_rate": 1.991809704490033e-08,
+      "loss": 0.2907,
+      "step": 11853
+    },
+    {
+      "epoch": 0.9725560979611929,
+      "grad_norm": 0.6495796948369779,
+      "learning_rate": 1.9799791566701975e-08,
+      "loss": 0.3237,
+      "step": 11854
+    },
+    {
+      "epoch": 0.9726381425113837,
+      "grad_norm": 0.6703849856282391,
+      "learning_rate": 1.9681837782241086e-08,
+      "loss": 0.3467,
+      "step": 11855
+    },
+    {
+      "epoch": 0.9727201870615745,
+      "grad_norm": 0.7127639592683117,
+      "learning_rate": 1.9564235699847666e-08,
+      "loss": 0.3046,
+      "step": 11856
+    },
+    {
+      "epoch": 0.9728022316117652,
+      "grad_norm": 0.7255804908107256,
+      "learning_rate": 1.944698532782563e-08,
+      "loss": 0.2791,
+      "step": 11857
+    },
+    {
+      "epoch": 0.9728842761619559,
+      "grad_norm": 0.7625606994101352,
+      "learning_rate": 1.9330086674456128e-08,
+      "loss": 0.3304,
+      "step": 11858
+    },
+    {
+      "epoch": 0.9729663207121467,
+      "grad_norm": 0.734686387055305,
+      "learning_rate": 1.921353974799367e-08,
+      "loss": 0.3244,
+      "step": 11859
+    },
+    {
+      "epoch": 0.9730483652623374,
+      "grad_norm": 0.7131614917844517,
+      "learning_rate": 1.9097344556668894e-08,
+      "loss": 0.2697,
+      "step": 11860
+    },
+    {
+      "epoch": 0.9731304098125282,
+      "grad_norm": 0.7140473088537427,
+      "learning_rate": 1.8981501108686907e-08,
+      "loss": 0.3029,
+      "step": 11861
+    },
+    {
+      "epoch": 0.973212454362719,
+      "grad_norm": 0.7806517525876355,
+      "learning_rate": 1.8866009412228937e-08,
+      "loss": 0.3004,
+      "step": 11862
+    },
+    {
+      "epoch": 0.9732944989129098,
+      "grad_norm": 0.6570360109473535,
+      "learning_rate": 1.8750869475450682e-08,
+      "loss": 0.291,
+      "step": 11863
+    },
+    {
+      "epoch": 0.9733765434631004,
+      "grad_norm": 0.7406834634261138,
+      "learning_rate": 1.8636081306482866e-08,
+      "loss": 0.2657,
+      "step": 11864
+    },
+    {
+      "epoch": 0.9734585880132912,
+      "grad_norm": 0.7697500186974211,
+      "learning_rate": 1.852164491343178e-08,
+      "loss": 0.2748,
+      "step": 11865
+    },
+    {
+      "epoch": 0.973540632563482,
+      "grad_norm": 0.7281351805770805,
+      "learning_rate": 1.8407560304378736e-08,
+      "loss": 0.2939,
+      "step": 11866
+    },
+    {
+      "epoch": 0.9736226771136727,
+      "grad_norm": 0.9983206095968886,
+      "learning_rate": 1.8293827487380623e-08,
+      "loss": 0.3123,
+      "step": 11867
+    },
+    {
+      "epoch": 0.9737047216638635,
+      "grad_norm": 0.7781674308254124,
+      "learning_rate": 1.818044647046824e-08,
+      "loss": 0.284,
+      "step": 11868
+    },
+    {
+      "epoch": 0.9737867662140542,
+      "grad_norm": 0.7511739858716409,
+      "learning_rate": 1.8067417261649066e-08,
+      "loss": 0.2806,
+      "step": 11869
+    },
+    {
+      "epoch": 0.973868810764245,
+      "grad_norm": 0.7295571222003446,
+      "learning_rate": 1.795473986890506e-08,
+      "loss": 0.2991,
+      "step": 11870
+    },
+    {
+      "epoch": 0.9739508553144357,
+      "grad_norm": 0.814416039664701,
+      "learning_rate": 1.7842414300192624e-08,
+      "loss": 0.3044,
+      "step": 11871
+    },
+    {
+      "epoch": 0.9740328998646265,
+      "grad_norm": 0.9101091764155838,
+      "learning_rate": 1.7730440563444307e-08,
+      "loss": 0.3206,
+      "step": 11872
+    },
+    {
+      "epoch": 0.9741149444148173,
+      "grad_norm": 0.9266574771763675,
+      "learning_rate": 1.7618818666568226e-08,
+      "loss": 0.2856,
+      "step": 11873
+    },
+    {
+      "epoch": 0.974196988965008,
+      "grad_norm": 0.7075300691028347,
+      "learning_rate": 1.7507548617445857e-08,
+      "loss": 0.2697,
+      "step": 11874
+    },
+    {
+      "epoch": 0.9742790335151987,
+      "grad_norm": 0.7199256206308239,
+      "learning_rate": 1.7396630423935356e-08,
+      "loss": 0.2871,
+      "step": 11875
+    },
+    {
+      "epoch": 0.9743610780653895,
+      "grad_norm": 0.725469958253232,
+      "learning_rate": 1.7286064093869902e-08,
+      "loss": 0.2917,
+      "step": 11876
+    },
+    {
+      "epoch": 0.9744431226155803,
+      "grad_norm": 0.8444639601217719,
+      "learning_rate": 1.7175849635057694e-08,
+      "loss": 0.3423,
+      "step": 11877
+    },
+    {
+      "epoch": 0.974525167165771,
+      "grad_norm": 0.9004385072924513,
+      "learning_rate": 1.7065987055280842e-08,
+      "loss": 0.2774,
+      "step": 11878
+    },
+    {
+      "epoch": 0.9746072117159618,
+      "grad_norm": 0.7917666042840839,
+      "learning_rate": 1.6956476362298692e-08,
+      "loss": 0.2841,
+      "step": 11879
+    },
+    {
+      "epoch": 0.9746892562661525,
+      "grad_norm": 0.7170991396043128,
+      "learning_rate": 1.6847317563843946e-08,
+      "loss": 0.3306,
+      "step": 11880
+    },
+    {
+      "epoch": 0.9747713008163432,
+      "grad_norm": 0.9585170618591542,
+      "learning_rate": 1.6738510667625997e-08,
+      "loss": 0.2918,
+      "step": 11881
+    },
+    {
+      "epoch": 0.974853345366534,
+      "grad_norm": 0.8131578028725107,
+      "learning_rate": 1.6630055681327582e-08,
+      "loss": 0.2686,
+      "step": 11882
+    },
+    {
+      "epoch": 0.9749353899167248,
+      "grad_norm": 0.6415692703131848,
+      "learning_rate": 1.6521952612608693e-08,
+      "loss": 0.2787,
+      "step": 11883
+    },
+    {
+      "epoch": 0.9750174344669156,
+      "grad_norm": 0.7351592343705473,
+      "learning_rate": 1.6414201469102664e-08,
+      "loss": 0.3161,
+      "step": 11884
+    },
+    {
+      "epoch": 0.9750994790171063,
+      "grad_norm": 0.6988460262753082,
+      "learning_rate": 1.630680225841952e-08,
+      "loss": 0.2714,
+      "step": 11885
+    },
+    {
+      "epoch": 0.975181523567297,
+      "grad_norm": 0.6823040237528494,
+      "learning_rate": 1.6199754988142635e-08,
+      "loss": 0.3092,
+      "step": 11886
+    },
+    {
+      "epoch": 0.9752635681174878,
+      "grad_norm": 0.7139887048084075,
+      "learning_rate": 1.609305966583208e-08,
+      "loss": 0.3051,
+      "step": 11887
+    },
+    {
+      "epoch": 0.9753456126676785,
+      "grad_norm": 0.6997460777205247,
+      "learning_rate": 1.5986716299022375e-08,
+      "loss": 0.2728,
+      "step": 11888
+    },
+    {
+      "epoch": 0.9754276572178693,
+      "grad_norm": 0.6983695816708014,
+      "learning_rate": 1.5880724895223077e-08,
+      "loss": 0.306,
+      "step": 11889
+    },
+    {
+      "epoch": 0.9755097017680601,
+      "grad_norm": 0.6234624448295786,
+      "learning_rate": 1.577508546191986e-08,
+      "loss": 0.2654,
+      "step": 11890
+    },
+    {
+      "epoch": 0.9755917463182509,
+      "grad_norm": 0.7629390906917347,
+      "learning_rate": 1.5669798006572313e-08,
+      "loss": 0.335,
+      "step": 11891
+    },
+    {
+      "epoch": 0.9756737908684415,
+      "grad_norm": 0.6330950119907233,
+      "learning_rate": 1.5564862536615598e-08,
+      "loss": 0.2741,
+      "step": 11892
+    },
+    {
+      "epoch": 0.9757558354186323,
+      "grad_norm": 0.7726551055890479,
+      "learning_rate": 1.5460279059459903e-08,
+      "loss": 0.2891,
+      "step": 11893
+    },
+    {
+      "epoch": 0.9758378799688231,
+      "grad_norm": 0.714112646427943,
+      "learning_rate": 1.535604758249154e-08,
+      "loss": 0.2943,
+      "step": 11894
+    },
+    {
+      "epoch": 0.9759199245190138,
+      "grad_norm": 0.9619888988796041,
+      "learning_rate": 1.5252168113070177e-08,
+      "loss": 0.3189,
+      "step": 11895
+    },
+    {
+      "epoch": 0.9760019690692046,
+      "grad_norm": 0.6562409572957998,
+      "learning_rate": 1.5148640658532164e-08,
+      "loss": 0.3331,
+      "step": 11896
+    },
+    {
+      "epoch": 0.9760840136193953,
+      "grad_norm": 0.7818313746567441,
+      "learning_rate": 1.5045465226188882e-08,
+      "loss": 0.2717,
+      "step": 11897
+    },
+    {
+      "epoch": 0.9761660581695861,
+      "grad_norm": 0.795521041301965,
+      "learning_rate": 1.4942641823325056e-08,
+      "loss": 0.2935,
+      "step": 11898
+    },
+    {
+      "epoch": 0.9762481027197768,
+      "grad_norm": 0.6990926916574886,
+      "learning_rate": 1.4840170457203206e-08,
+      "loss": 0.2782,
+      "step": 11899
+    },
+    {
+      "epoch": 0.9763301472699676,
+      "grad_norm": 0.6658979252999624,
+      "learning_rate": 1.4738051135059217e-08,
+      "loss": 0.2798,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9764121918201584,
+      "grad_norm": 0.7487380992600067,
+      "learning_rate": 1.463628386410454e-08,
+      "loss": 0.335,
+      "step": 11901
+    },
+    {
+      "epoch": 0.976494236370349,
+      "grad_norm": 0.6744265241032749,
+      "learning_rate": 1.4534868651526202e-08,
+      "loss": 0.2726,
+      "step": 11902
+    },
+    {
+      "epoch": 0.9765762809205398,
+      "grad_norm": 0.6973889820684239,
+      "learning_rate": 1.4433805504485699e-08,
+      "loss": 0.2932,
+      "step": 11903
+    },
+    {
+      "epoch": 0.9766583254707306,
+      "grad_norm": 0.7433379230425784,
+      "learning_rate": 1.4333094430119542e-08,
+      "loss": 0.3023,
+      "step": 11904
+    },
+    {
+      "epoch": 0.9767403700209214,
+      "grad_norm": 0.6511351598282097,
+      "learning_rate": 1.4232735435540378e-08,
+      "loss": 0.2788,
+      "step": 11905
+    },
+    {
+      "epoch": 0.9768224145711121,
+      "grad_norm": 0.6976879494252395,
+      "learning_rate": 1.4132728527835315e-08,
+      "loss": 0.3196,
+      "step": 11906
+    },
+    {
+      "epoch": 0.9769044591213029,
+      "grad_norm": 0.777043327488366,
+      "learning_rate": 1.4033073714065926e-08,
+      "loss": 0.3284,
+      "step": 11907
+    },
+    {
+      "epoch": 0.9769865036714936,
+      "grad_norm": 0.7183288543642025,
+      "learning_rate": 1.3933771001271023e-08,
+      "loss": 0.3072,
+      "step": 11908
+    },
+    {
+      "epoch": 0.9770685482216844,
+      "grad_norm": 0.69828211192818,
+      "learning_rate": 1.3834820396462223e-08,
+      "loss": 0.3209,
+      "step": 11909
+    },
+    {
+      "epoch": 0.9771505927718751,
+      "grad_norm": 0.6745426172229101,
+      "learning_rate": 1.3736221906627822e-08,
+      "loss": 0.2628,
+      "step": 11910
+    },
+    {
+      "epoch": 0.9772326373220659,
+      "grad_norm": 0.8232738625304331,
+      "learning_rate": 1.363797553873003e-08,
+      "loss": 0.2834,
+      "step": 11911
+    },
+    {
+      "epoch": 0.9773146818722567,
+      "grad_norm": 0.9397519763465695,
+      "learning_rate": 1.3540081299707187e-08,
+      "loss": 0.2882,
+      "step": 11912
+    },
+    {
+      "epoch": 0.9773967264224473,
+      "grad_norm": 1.2200544403635472,
+      "learning_rate": 1.3442539196472647e-08,
+      "loss": 0.2662,
+      "step": 11913
+    },
+    {
+      "epoch": 0.9774787709726381,
+      "grad_norm": 0.75353702828581,
+      "learning_rate": 1.334534923591424e-08,
+      "loss": 0.3048,
+      "step": 11914
+    },
+    {
+      "epoch": 0.9775608155228289,
+      "grad_norm": 0.6671476460249791,
+      "learning_rate": 1.324851142489536e-08,
+      "loss": 0.2824,
+      "step": 11915
+    },
+    {
+      "epoch": 0.9776428600730197,
+      "grad_norm": 0.801213933887738,
+      "learning_rate": 1.3152025770255539e-08,
+      "loss": 0.3191,
+      "step": 11916
+    },
+    {
+      "epoch": 0.9777249046232104,
+      "grad_norm": 0.639686531346677,
+      "learning_rate": 1.3055892278807103e-08,
+      "loss": 0.2768,
+      "step": 11917
+    },
+    {
+      "epoch": 0.9778069491734012,
+      "grad_norm": 0.6653936719897269,
+      "learning_rate": 1.2960110957339623e-08,
+      "loss": 0.285,
+      "step": 11918
+    },
+    {
+      "epoch": 0.977888993723592,
+      "grad_norm": 0.7087604380738076,
+      "learning_rate": 1.2864681812616575e-08,
+      "loss": 0.2962,
+      "step": 11919
+    },
+    {
+      "epoch": 0.9779710382737826,
+      "grad_norm": 0.733306386948836,
+      "learning_rate": 1.276960485137757e-08,
+      "loss": 0.3032,
+      "step": 11920
+    },
+    {
+      "epoch": 0.9780530828239734,
+      "grad_norm": 0.6921642552887733,
+      "learning_rate": 1.2674880080336682e-08,
+      "loss": 0.2738,
+      "step": 11921
+    },
+    {
+      "epoch": 0.9781351273741642,
+      "grad_norm": 0.7573325721611717,
+      "learning_rate": 1.258050750618245e-08,
+      "loss": 0.2815,
+      "step": 11922
+    },
+    {
+      "epoch": 0.978217171924355,
+      "grad_norm": 0.6569582869595058,
+      "learning_rate": 1.2486487135580094e-08,
+      "loss": 0.2838,
+      "step": 11923
+    },
+    {
+      "epoch": 0.9782992164745457,
+      "grad_norm": 0.7360793381463095,
+      "learning_rate": 1.2392818975168752e-08,
+      "loss": 0.2938,
+      "step": 11924
+    },
+    {
+      "epoch": 0.9783812610247364,
+      "grad_norm": 0.7944275878684655,
+      "learning_rate": 1.2299503031563687e-08,
+      "loss": 0.2823,
+      "step": 11925
+    },
+    {
+      "epoch": 0.9784633055749272,
+      "grad_norm": 0.7019107344168205,
+      "learning_rate": 1.2206539311354071e-08,
+      "loss": 0.3217,
+      "step": 11926
+    },
+    {
+      "epoch": 0.9785453501251179,
+      "grad_norm": 0.699565143031579,
+      "learning_rate": 1.211392782110521e-08,
+      "loss": 0.2606,
+      "step": 11927
+    },
+    {
+      "epoch": 0.9786273946753087,
+      "grad_norm": 0.6841079694710712,
+      "learning_rate": 1.2021668567357425e-08,
+      "loss": 0.2987,
+      "step": 11928
+    },
+    {
+      "epoch": 0.9787094392254995,
+      "grad_norm": 0.7990467512646007,
+      "learning_rate": 1.1929761556625507e-08,
+      "loss": 0.2913,
+      "step": 11929
+    },
+    {
+      "epoch": 0.9787914837756903,
+      "grad_norm": 0.6730102157035833,
+      "learning_rate": 1.183820679539982e-08,
+      "loss": 0.3169,
+      "step": 11930
+    },
+    {
+      "epoch": 0.9788735283258809,
+      "grad_norm": 0.7186480504521766,
+      "learning_rate": 1.1747004290145747e-08,
+      "loss": 0.303,
+      "step": 11931
+    },
+    {
+      "epoch": 0.9789555728760717,
+      "grad_norm": 0.6954729337214143,
+      "learning_rate": 1.1656154047303691e-08,
+      "loss": 0.2681,
+      "step": 11932
+    },
+    {
+      "epoch": 0.9790376174262625,
+      "grad_norm": 0.6971602777615542,
+      "learning_rate": 1.1565656073290188e-08,
+      "loss": 0.3109,
+      "step": 11933
+    },
+    {
+      "epoch": 0.9791196619764532,
+      "grad_norm": 0.7639642774163216,
+      "learning_rate": 1.147551037449568e-08,
+      "loss": 0.2941,
+      "step": 11934
+    },
+    {
+      "epoch": 0.979201706526644,
+      "grad_norm": 0.6885622107146153,
+      "learning_rate": 1.138571695728563e-08,
+      "loss": 0.2882,
+      "step": 11935
+    },
+    {
+      "epoch": 0.9792837510768347,
+      "grad_norm": 0.7194506006520661,
+      "learning_rate": 1.1296275828001635e-08,
+      "loss": 0.3063,
+      "step": 11936
+    },
+    {
+      "epoch": 0.9793657956270255,
+      "grad_norm": 0.7237481981748485,
+      "learning_rate": 1.1207186992959197e-08,
+      "loss": 0.3195,
+      "step": 11937
+    },
+    {
+      "epoch": 0.9794478401772162,
+      "grad_norm": 0.672828819250741,
+      "learning_rate": 1.1118450458450503e-08,
+      "loss": 0.2589,
+      "step": 11938
+    },
+    {
+      "epoch": 0.979529884727407,
+      "grad_norm": 0.7242980807925332,
+      "learning_rate": 1.1030066230741099e-08,
+      "loss": 0.2757,
+      "step": 11939
+    },
+    {
+      "epoch": 0.9796119292775978,
+      "grad_norm": 0.7017461368699519,
+      "learning_rate": 1.094203431607377e-08,
+      "loss": 0.2929,
+      "step": 11940
+    },
+    {
+      "epoch": 0.9796939738277884,
+      "grad_norm": 0.6646447107136422,
+      "learning_rate": 1.0854354720664095e-08,
+      "loss": 0.2953,
+      "step": 11941
+    },
+    {
+      "epoch": 0.9797760183779792,
+      "grad_norm": 0.7031329425762648,
+      "learning_rate": 1.0767027450703793e-08,
+      "loss": 0.2753,
+      "step": 11942
+    },
+    {
+      "epoch": 0.97985806292817,
+      "grad_norm": 0.7620013092535042,
+      "learning_rate": 1.0680052512360706e-08,
+      "loss": 0.2774,
+      "step": 11943
+    },
+    {
+      "epoch": 0.9799401074783608,
+      "grad_norm": 0.6592802830515809,
+      "learning_rate": 1.0593429911776588e-08,
+      "loss": 0.2972,
+      "step": 11944
+    },
+    {
+      "epoch": 0.9800221520285515,
+      "grad_norm": 0.7607778865242127,
+      "learning_rate": 1.0507159655067656e-08,
+      "loss": 0.2933,
+      "step": 11945
+    },
+    {
+      "epoch": 0.9801041965787423,
+      "grad_norm": 0.6456257486821463,
+      "learning_rate": 1.0421241748327371e-08,
+      "loss": 0.2998,
+      "step": 11946
+    },
+    {
+      "epoch": 0.980186241128933,
+      "grad_norm": 0.8039681822834317,
+      "learning_rate": 1.0335676197622546e-08,
+      "loss": 0.3422,
+      "step": 11947
+    },
+    {
+      "epoch": 0.9802682856791237,
+      "grad_norm": 0.7150968206029614,
+      "learning_rate": 1.025046300899557e-08,
+      "loss": 0.292,
+      "step": 11948
+    },
+    {
+      "epoch": 0.9803503302293145,
+      "grad_norm": 0.6942269447264758,
+      "learning_rate": 1.0165602188464408e-08,
+      "loss": 0.2822,
+      "step": 11949
+    },
+    {
+      "epoch": 0.9804323747795053,
+      "grad_norm": 0.7019116579158207,
+      "learning_rate": 1.0081093742021486e-08,
+      "loss": 0.3014,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9805144193296961,
+      "grad_norm": 0.7108022306725496,
+      "learning_rate": 9.996937675635365e-09,
+      "loss": 0.2905,
+      "step": 11951
+    },
+    {
+      "epoch": 0.9805964638798867,
+      "grad_norm": 0.6839908054594752,
+      "learning_rate": 9.913133995247959e-09,
+      "loss": 0.281,
+      "step": 11952
+    },
+    {
+      "epoch": 0.9806785084300775,
+      "grad_norm": 0.7359122627917103,
+      "learning_rate": 9.829682706777866e-09,
+      "loss": 0.3505,
+      "step": 11953
+    },
+    {
+      "epoch": 0.9807605529802683,
+      "grad_norm": 0.7109664223038876,
+      "learning_rate": 9.74658381611815e-09,
+      "loss": 0.3262,
+      "step": 11954
+    },
+    {
+      "epoch": 0.980842597530459,
+      "grad_norm": 0.7206691906665511,
+      "learning_rate": 9.66383732913745e-09,
+      "loss": 0.3187,
+      "step": 11955
+    },
+    {
+      "epoch": 0.9809246420806498,
+      "grad_norm": 0.7883568229475316,
+      "learning_rate": 9.581443251678869e-09,
+      "loss": 0.307,
+      "step": 11956
+    },
+    {
+      "epoch": 0.9810066866308406,
+      "grad_norm": 0.6727231035555945,
+      "learning_rate": 9.499401589561085e-09,
+      "loss": 0.282,
+      "step": 11957
+    },
+    {
+      "epoch": 0.9810887311810313,
+      "grad_norm": 0.6683989602523375,
+      "learning_rate": 9.4177123485778e-09,
+      "loss": 0.2794,
+      "step": 11958
+    },
+    {
+      "epoch": 0.981170775731222,
+      "grad_norm": 0.7199804268409076,
+      "learning_rate": 9.336375534497732e-09,
+      "loss": 0.2798,
+      "step": 11959
+    },
+    {
+      "epoch": 0.9812528202814128,
+      "grad_norm": 0.709818663578977,
+      "learning_rate": 9.255391153064065e-09,
+      "loss": 0.2903,
+      "step": 11960
+    },
+    {
+      "epoch": 0.9813348648316036,
+      "grad_norm": 0.6882524483479812,
+      "learning_rate": 9.17475920999722e-09,
+      "loss": 0.2794,
+      "step": 11961
+    },
+    {
+      "epoch": 0.9814169093817943,
+      "grad_norm": 0.7533908860506451,
+      "learning_rate": 9.094479710990422e-09,
+      "loss": 0.2733,
+      "step": 11962
+    },
+    {
+      "epoch": 0.981498953931985,
+      "grad_norm": 0.685076286795829,
+      "learning_rate": 9.014552661712473e-09,
+      "loss": 0.2885,
+      "step": 11963
+    },
+    {
+      "epoch": 0.9815809984821758,
+      "grad_norm": 0.6583766793320132,
+      "learning_rate": 8.934978067808852e-09,
+      "loss": 0.2695,
+      "step": 11964
+    },
+    {
+      "epoch": 0.9816630430323666,
+      "grad_norm": 0.7061661698360902,
+      "learning_rate": 8.855755934897847e-09,
+      "loss": 0.281,
+      "step": 11965
+    },
+    {
+      "epoch": 0.9817450875825573,
+      "grad_norm": 0.7284648064067792,
+      "learning_rate": 8.776886268574424e-09,
+      "loss": 0.3184,
+      "step": 11966
+    },
+    {
+      "epoch": 0.9818271321327481,
+      "grad_norm": 0.6782327568288081,
+      "learning_rate": 8.698369074408575e-09,
+      "loss": 0.2848,
+      "step": 11967
+    },
+    {
+      "epoch": 0.9819091766829389,
+      "grad_norm": 0.7069161449791997,
+      "learning_rate": 8.620204357944195e-09,
+      "loss": 0.3112,
+      "step": 11968
+    },
+    {
+      "epoch": 0.9819912212331295,
+      "grad_norm": 0.7367157240263079,
+      "learning_rate": 8.542392124702426e-09,
+      "loss": 0.2931,
+      "step": 11969
+    },
+    {
+      "epoch": 0.9820732657833203,
+      "grad_norm": 1.6756490422597825,
+      "learning_rate": 8.464932380177204e-09,
+      "loss": 0.2877,
+      "step": 11970
+    },
+    {
+      "epoch": 0.9821553103335111,
+      "grad_norm": 0.7001808207141774,
+      "learning_rate": 8.387825129839155e-09,
+      "loss": 0.2961,
+      "step": 11971
+    },
+    {
+      "epoch": 0.9822373548837019,
+      "grad_norm": 0.6612142447106969,
+      "learning_rate": 8.311070379132812e-09,
+      "loss": 0.2538,
+      "step": 11972
+    },
+    {
+      "epoch": 0.9823193994338926,
+      "grad_norm": 0.6903208120527958,
+      "learning_rate": 8.234668133479395e-09,
+      "loss": 0.2616,
+      "step": 11973
+    },
+    {
+      "epoch": 0.9824014439840834,
+      "grad_norm": 0.6968877377107683,
+      "learning_rate": 8.158618398273477e-09,
+      "loss": 0.31,
+      "step": 11974
+    },
+    {
+      "epoch": 0.9824834885342741,
+      "grad_norm": 0.6426872937035143,
+      "learning_rate": 8.082921178886316e-09,
+      "loss": 0.2771,
+      "step": 11975
+    },
+    {
+      "epoch": 0.9825655330844648,
+      "grad_norm": 0.7671479234049666,
+      "learning_rate": 8.007576480663082e-09,
+      "loss": 0.2678,
+      "step": 11976
+    },
+    {
+      "epoch": 0.9826475776346556,
+      "grad_norm": 0.6527261330528418,
+      "learning_rate": 7.932584308923962e-09,
+      "loss": 0.298,
+      "step": 11977
+    },
+    {
+      "epoch": 0.9827296221848464,
+      "grad_norm": 0.8489005547606931,
+      "learning_rate": 7.857944668965833e-09,
+      "loss": 0.3018,
+      "step": 11978
+    },
+    {
+      "epoch": 0.9828116667350372,
+      "grad_norm": 0.7409085178766682,
+      "learning_rate": 7.783657566058922e-09,
+      "loss": 0.3196,
+      "step": 11979
+    },
+    {
+      "epoch": 0.9828937112852278,
+      "grad_norm": 0.7353703539712663,
+      "learning_rate": 7.709723005450143e-09,
+      "loss": 0.294,
+      "step": 11980
+    },
+    {
+      "epoch": 0.9829757558354186,
+      "grad_norm": 0.7040919664915308,
+      "learning_rate": 7.636140992359209e-09,
+      "loss": 0.2989,
+      "step": 11981
+    },
+    {
+      "epoch": 0.9830578003856094,
+      "grad_norm": 0.7680694340337009,
+      "learning_rate": 7.56291153198363e-09,
+      "loss": 0.2807,
+      "step": 11982
+    },
+    {
+      "epoch": 0.9831398449358001,
+      "grad_norm": 0.7252608993747147,
+      "learning_rate": 7.490034629494269e-09,
+      "loss": 0.2743,
+      "step": 11983
+    },
+    {
+      "epoch": 0.9832218894859909,
+      "grad_norm": 0.729418249350231,
+      "learning_rate": 7.417510290037011e-09,
+      "loss": 0.3252,
+      "step": 11984
+    },
+    {
+      "epoch": 0.9833039340361817,
+      "grad_norm": 0.7734436958455039,
+      "learning_rate": 7.345338518734979e-09,
+      "loss": 0.2875,
+      "step": 11985
+    },
+    {
+      "epoch": 0.9833859785863724,
+      "grad_norm": 0.6889613779948398,
+      "learning_rate": 7.273519320682987e-09,
+      "loss": 0.3242,
+      "step": 11986
+    },
+    {
+      "epoch": 0.9834680231365631,
+      "grad_norm": 0.6677423937012755,
+      "learning_rate": 7.202052700954198e-09,
+      "loss": 0.2895,
+      "step": 11987
+    },
+    {
+      "epoch": 0.9835500676867539,
+      "grad_norm": 0.6561968681979079,
+      "learning_rate": 7.130938664594578e-09,
+      "loss": 0.2835,
+      "step": 11988
+    },
+    {
+      "epoch": 0.9836321122369447,
+      "grad_norm": 0.7069317137126904,
+      "learning_rate": 7.060177216626218e-09,
+      "loss": 0.2781,
+      "step": 11989
+    },
+    {
+      "epoch": 0.9837141567871354,
+      "grad_norm": 0.7195869348752281,
+      "learning_rate": 6.9897683620467894e-09,
+      "loss": 0.3068,
+      "step": 11990
+    },
+    {
+      "epoch": 0.9837962013373261,
+      "grad_norm": 0.7430286496472137,
+      "learning_rate": 6.91971210582787e-09,
+      "loss": 0.2804,
+      "step": 11991
+    },
+    {
+      "epoch": 0.9838782458875169,
+      "grad_norm": 0.7079335504836638,
+      "learning_rate": 6.850008452916612e-09,
+      "loss": 0.3033,
+      "step": 11992
+    },
+    {
+      "epoch": 0.9839602904377077,
+      "grad_norm": 0.7553456280106791,
+      "learning_rate": 6.780657408236302e-09,
+      "loss": 0.3334,
+      "step": 11993
+    },
+    {
+      "epoch": 0.9840423349878984,
+      "grad_norm": 0.6713499276859001,
+      "learning_rate": 6.711658976683022e-09,
+      "loss": 0.2927,
+      "step": 11994
+    },
+    {
+      "epoch": 0.9841243795380892,
+      "grad_norm": 0.7015279011415045,
+      "learning_rate": 6.643013163130651e-09,
+      "loss": 0.2894,
+      "step": 11995
+    },
+    {
+      "epoch": 0.98420642408828,
+      "grad_norm": 0.6781556960230659,
+      "learning_rate": 6.574719972425869e-09,
+      "loss": 0.2852,
+      "step": 11996
+    },
+    {
+      "epoch": 0.9842884686384706,
+      "grad_norm": 0.7498155541617796,
+      "learning_rate": 6.506779409392039e-09,
+      "loss": 0.3053,
+      "step": 11997
+    },
+    {
+      "epoch": 0.9843705131886614,
+      "grad_norm": 0.7190099338177652,
+      "learning_rate": 6.439191478826434e-09,
+      "loss": 0.2795,
+      "step": 11998
+    },
+    {
+      "epoch": 0.9844525577388522,
+      "grad_norm": 0.6293185521186591,
+      "learning_rate": 6.3719561855024595e-09,
+      "loss": 0.2767,
+      "step": 11999
+    },
+    {
+      "epoch": 0.984534602289043,
+      "grad_norm": 0.7543702907098014,
+      "learning_rate": 6.305073534168538e-09,
+      "loss": 0.2692,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9846166468392337,
+      "grad_norm": 0.6992228380559786,
+      "learning_rate": 6.238543529547003e-09,
+      "loss": 0.345,
+      "step": 12001
+    },
+    {
+      "epoch": 0.9846986913894245,
+      "grad_norm": 0.7118215211067631,
+      "learning_rate": 6.17236617633632e-09,
+      "loss": 0.3011,
+      "step": 12002
+    },
+    {
+      "epoch": 0.9847807359396152,
+      "grad_norm": 0.7276285837542504,
+      "learning_rate": 6.106541479209971e-09,
+      "loss": 0.3074,
+      "step": 12003
+    },
+    {
+      "epoch": 0.9848627804898059,
+      "grad_norm": 0.7598208439108629,
+      "learning_rate": 6.041069442816461e-09,
+      "loss": 0.2936,
+      "step": 12004
+    },
+    {
+      "epoch": 0.9849448250399967,
+      "grad_norm": 0.7337695879847373,
+      "learning_rate": 5.975950071779313e-09,
+      "loss": 0.3152,
+      "step": 12005
+    },
+    {
+      "epoch": 0.9850268695901875,
+      "grad_norm": 0.6880338602086229,
+      "learning_rate": 5.911183370697071e-09,
+      "loss": 0.2858,
+      "step": 12006
+    },
+    {
+      "epoch": 0.9851089141403783,
+      "grad_norm": 0.6567432551314125,
+      "learning_rate": 5.846769344143299e-09,
+      "loss": 0.2967,
+      "step": 12007
+    },
+    {
+      "epoch": 0.9851909586905689,
+      "grad_norm": 0.6585994226388779,
+      "learning_rate": 5.782707996667136e-09,
+      "loss": 0.2986,
+      "step": 12008
+    },
+    {
+      "epoch": 0.9852730032407597,
+      "grad_norm": 0.6983495807601585,
+      "learning_rate": 5.718999332792185e-09,
+      "loss": 0.2846,
+      "step": 12009
+    },
+    {
+      "epoch": 0.9853550477909505,
+      "grad_norm": 0.6921996599131651,
+      "learning_rate": 5.655643357017071e-09,
+      "loss": 0.2941,
+      "step": 12010
+    },
+    {
+      "epoch": 0.9854370923411412,
+      "grad_norm": 0.6906143781070033,
+      "learning_rate": 5.592640073817102e-09,
+      "loss": 0.2975,
+      "step": 12011
+    },
+    {
+      "epoch": 0.985519136891332,
+      "grad_norm": 0.6740027001708184,
+      "learning_rate": 5.529989487640386e-09,
+      "loss": 0.2665,
+      "step": 12012
+    },
+    {
+      "epoch": 0.9856011814415228,
+      "grad_norm": 0.7287744002681045,
+      "learning_rate": 5.467691602912273e-09,
+      "loss": 0.3458,
+      "step": 12013
+    },
+    {
+      "epoch": 0.9856832259917135,
+      "grad_norm": 0.7373008380270086,
+      "learning_rate": 5.405746424030356e-09,
+      "loss": 0.2781,
+      "step": 12014
+    },
+    {
+      "epoch": 0.9857652705419042,
+      "grad_norm": 0.6401282439819119,
+      "learning_rate": 5.344153955371134e-09,
+      "loss": 0.3089,
+      "step": 12015
+    },
+    {
+      "epoch": 0.985847315092095,
+      "grad_norm": 0.7552793115058015,
+      "learning_rate": 5.282914201282796e-09,
+      "loss": 0.3103,
+      "step": 12016
+    },
+    {
+      "epoch": 0.9859293596422858,
+      "grad_norm": 0.6774891491421445,
+      "learning_rate": 5.2220271660907705e-09,
+      "loss": 0.3303,
+      "step": 12017
+    },
+    {
+      "epoch": 0.9860114041924765,
+      "grad_norm": 0.7028793595978021,
+      "learning_rate": 5.1614928540938416e-09,
+      "loss": 0.2804,
+      "step": 12018
+    },
+    {
+      "epoch": 0.9860934487426672,
+      "grad_norm": 0.6837449675988525,
+      "learning_rate": 5.101311269568032e-09,
+      "loss": 0.3067,
+      "step": 12019
+    },
+    {
+      "epoch": 0.986175493292858,
+      "grad_norm": 0.7156130867120302,
+      "learning_rate": 5.041482416762167e-09,
+      "loss": 0.3127,
+      "step": 12020
+    },
+    {
+      "epoch": 0.9862575378430488,
+      "grad_norm": 0.8594639654237166,
+      "learning_rate": 4.982006299902309e-09,
+      "loss": 0.2974,
+      "step": 12021
+    },
+    {
+      "epoch": 0.9863395823932395,
+      "grad_norm": 0.7942213068926728,
+      "learning_rate": 4.922882923187877e-09,
+      "loss": 0.3293,
+      "step": 12022
+    },
+    {
+      "epoch": 0.9864216269434303,
+      "grad_norm": 0.8053987759660884,
+      "learning_rate": 4.864112290793865e-09,
+      "loss": 0.3286,
+      "step": 12023
+    },
+    {
+      "epoch": 0.9865036714936211,
+      "grad_norm": 0.7248724264506464,
+      "learning_rate": 4.805694406871397e-09,
+      "loss": 0.2819,
+      "step": 12024
+    },
+    {
+      "epoch": 0.9865857160438118,
+      "grad_norm": 0.8473175290064431,
+      "learning_rate": 4.747629275545507e-09,
+      "loss": 0.2862,
+      "step": 12025
+    },
+    {
+      "epoch": 0.9866677605940025,
+      "grad_norm": 0.7456595169778325,
+      "learning_rate": 4.689916900916247e-09,
+      "loss": 0.273,
+      "step": 12026
+    },
+    {
+      "epoch": 0.9867498051441933,
+      "grad_norm": 1.1427773337262186,
+      "learning_rate": 4.632557287059803e-09,
+      "loss": 0.3328,
+      "step": 12027
+    },
+    {
+      "epoch": 0.9868318496943841,
+      "grad_norm": 0.6375337520383748,
+      "learning_rate": 4.575550438026266e-09,
+      "loss": 0.3261,
+      "step": 12028
+    },
+    {
+      "epoch": 0.9869138942445748,
+      "grad_norm": 0.6728120066841088,
+      "learning_rate": 4.518896357841307e-09,
+      "loss": 0.3231,
+      "step": 12029
+    },
+    {
+      "epoch": 0.9869959387947655,
+      "grad_norm": 0.8318011187278251,
+      "learning_rate": 4.462595050506724e-09,
+      "loss": 0.2764,
+      "step": 12030
+    },
+    {
+      "epoch": 0.9870779833449563,
+      "grad_norm": 0.7232094921558797,
+      "learning_rate": 4.406646519997115e-09,
+      "loss": 0.2899,
+      "step": 12031
+    },
+    {
+      "epoch": 0.9871600278951471,
+      "grad_norm": 0.6923838487527949,
+      "learning_rate": 4.35105077026432e-09,
+      "loss": 0.3222,
+      "step": 12032
+    },
+    {
+      "epoch": 0.9872420724453378,
+      "grad_norm": 0.7349402932591843,
+      "learning_rate": 4.295807805234087e-09,
+      "loss": 0.3039,
+      "step": 12033
+    },
+    {
+      "epoch": 0.9873241169955286,
+      "grad_norm": 0.7544504098259628,
+      "learning_rate": 4.240917628808294e-09,
+      "loss": 0.2741,
+      "step": 12034
+    },
+    {
+      "epoch": 0.9874061615457194,
+      "grad_norm": 0.6678888435465737,
+      "learning_rate": 4.186380244862176e-09,
+      "loss": 0.2482,
+      "step": 12035
+    },
+    {
+      "epoch": 0.98748820609591,
+      "grad_norm": 0.6985729390892562,
+      "learning_rate": 4.132195657247095e-09,
+      "loss": 0.2808,
+      "step": 12036
+    },
+    {
+      "epoch": 0.9875702506461008,
+      "grad_norm": 0.705351190784906,
+      "learning_rate": 4.078363869790547e-09,
+      "loss": 0.3165,
+      "step": 12037
+    },
+    {
+      "epoch": 0.9876522951962916,
+      "grad_norm": 0.7253546330145214,
+      "learning_rate": 4.024884886292823e-09,
+      "loss": 0.3137,
+      "step": 12038
+    },
+    {
+      "epoch": 0.9877343397464824,
+      "grad_norm": 0.6897963417624674,
+      "learning_rate": 3.971758710531459e-09,
+      "loss": 0.333,
+      "step": 12039
+    },
+    {
+      "epoch": 0.9878163842966731,
+      "grad_norm": 0.6692845286912422,
+      "learning_rate": 3.9189853462573424e-09,
+      "loss": 0.2962,
+      "step": 12040
+    },
+    {
+      "epoch": 0.9878984288468639,
+      "grad_norm": 0.8251751968423249,
+      "learning_rate": 3.866564797198047e-09,
+      "loss": 0.2942,
+      "step": 12041
+    },
+    {
+      "epoch": 0.9879804733970546,
+      "grad_norm": 0.7123383575000505,
+      "learning_rate": 3.814497067055056e-09,
+      "loss": 0.3263,
+      "step": 12042
+    },
+    {
+      "epoch": 0.9880625179472453,
+      "grad_norm": 0.7174445090833846,
+      "learning_rate": 3.762782159505429e-09,
+      "loss": 0.3231,
+      "step": 12043
+    },
+    {
+      "epoch": 0.9881445624974361,
+      "grad_norm": 0.6632632721833733,
+      "learning_rate": 3.7114200782006894e-09,
+      "loss": 0.3116,
+      "step": 12044
+    },
+    {
+      "epoch": 0.9882266070476269,
+      "grad_norm": 0.7923863440655751,
+      "learning_rate": 3.6604108267684903e-09,
+      "loss": 0.3023,
+      "step": 12045
+    },
+    {
+      "epoch": 0.9883086515978177,
+      "grad_norm": 0.6229596018928063,
+      "learning_rate": 3.6097544088103954e-09,
+      "loss": 0.2814,
+      "step": 12046
+    },
+    {
+      "epoch": 0.9883906961480083,
+      "grad_norm": 0.6859287291961409,
+      "learning_rate": 3.5594508279046534e-09,
+      "loss": 0.2983,
+      "step": 12047
+    },
+    {
+      "epoch": 0.9884727406981991,
+      "grad_norm": 0.7371465586414937,
+      "learning_rate": 3.5095000876028683e-09,
+      "loss": 0.2452,
+      "step": 12048
+    },
+    {
+      "epoch": 0.9885547852483899,
+      "grad_norm": 0.8800825505106061,
+      "learning_rate": 3.4599021914327737e-09,
+      "loss": 0.3394,
+      "step": 12049
+    },
+    {
+      "epoch": 0.9886368297985806,
+      "grad_norm": 0.7524285786067918,
+      "learning_rate": 3.4106571428971223e-09,
+      "loss": 0.3093,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9887188743487714,
+      "grad_norm": 0.6982090757300221,
+      "learning_rate": 3.361764945473134e-09,
+      "loss": 0.2953,
+      "step": 12051
+    },
+    {
+      "epoch": 0.9888009188989622,
+      "grad_norm": 0.7205202202714382,
+      "learning_rate": 3.313225602613046e-09,
+      "loss": 0.3091,
+      "step": 12052
+    },
+    {
+      "epoch": 0.9888829634491529,
+      "grad_norm": 0.688901246974096,
+      "learning_rate": 3.265039117745783e-09,
+      "loss": 0.31,
+      "step": 12053
+    },
+    {
+      "epoch": 0.9889650079993436,
+      "grad_norm": 0.6611143822378295,
+      "learning_rate": 3.217205494273623e-09,
+      "loss": 0.3482,
+      "step": 12054
+    },
+    {
+      "epoch": 0.9890470525495344,
+      "grad_norm": 0.7112119219029972,
+      "learning_rate": 3.1697247355738646e-09,
+      "loss": 0.2945,
+      "step": 12055
+    },
+    {
+      "epoch": 0.9891290970997252,
+      "grad_norm": 0.7204285057128671,
+      "learning_rate": 3.1225968450004916e-09,
+      "loss": 0.2424,
+      "step": 12056
+    },
+    {
+      "epoch": 0.9892111416499159,
+      "grad_norm": 0.717939218646702,
+      "learning_rate": 3.0758218258813977e-09,
+      "loss": 0.2909,
+      "step": 12057
+    },
+    {
+      "epoch": 0.9892931862001066,
+      "grad_norm": 0.7425115451097501,
+      "learning_rate": 3.0293996815194958e-09,
+      "loss": 0.2972,
+      "step": 12058
+    },
+    {
+      "epoch": 0.9893752307502974,
+      "grad_norm": 0.6955731571090201,
+      "learning_rate": 2.98333041519272e-09,
+      "loss": 0.2639,
+      "step": 12059
+    },
+    {
+      "epoch": 0.9894572753004882,
+      "grad_norm": 0.8105589684058404,
+      "learning_rate": 2.9376140301556887e-09,
+      "loss": 0.288,
+      "step": 12060
+    },
+    {
+      "epoch": 0.9895393198506789,
+      "grad_norm": 0.659765959788263,
+      "learning_rate": 2.8922505296352654e-09,
+      "loss": 0.282,
+      "step": 12061
+    },
+    {
+      "epoch": 0.9896213644008697,
+      "grad_norm": 1.1717919775445684,
+      "learning_rate": 2.847239916836109e-09,
+      "loss": 0.3102,
+      "step": 12062
+    },
+    {
+      "epoch": 0.9897034089510605,
+      "grad_norm": 0.7173302947912172,
+      "learning_rate": 2.8025821949356768e-09,
+      "loss": 0.2917,
+      "step": 12063
+    },
+    {
+      "epoch": 0.9897854535012511,
+      "grad_norm": 0.7177206529740717,
+      "learning_rate": 2.7582773670886685e-09,
+      "loss": 0.2656,
+      "step": 12064
+    },
+    {
+      "epoch": 0.9898674980514419,
+      "grad_norm": 0.7606218725820786,
+      "learning_rate": 2.7143254364236925e-09,
+      "loss": 0.2992,
+      "step": 12065
+    },
+    {
+      "epoch": 0.9899495426016327,
+      "grad_norm": 0.740253589915343,
+      "learning_rate": 2.670726406043822e-09,
+      "loss": 0.2852,
+      "step": 12066
+    },
+    {
+      "epoch": 0.9900315871518235,
+      "grad_norm": 0.9811550190074441,
+      "learning_rate": 2.6274802790288156e-09,
+      "loss": 0.2978,
+      "step": 12067
+    },
+    {
+      "epoch": 0.9901136317020142,
+      "grad_norm": 0.6954606214539052,
+      "learning_rate": 2.5845870584317865e-09,
+      "loss": 0.3133,
+      "step": 12068
+    },
+    {
+      "epoch": 0.990195676252205,
+      "grad_norm": 0.679555082822226,
+      "learning_rate": 2.542046747282534e-09,
+      "loss": 0.3143,
+      "step": 12069
+    },
+    {
+      "epoch": 0.9902777208023957,
+      "grad_norm": 0.7651483017920145,
+      "learning_rate": 2.4998593485853204e-09,
+      "loss": 0.3224,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9903597653525864,
+      "grad_norm": 0.7078481093095091,
+      "learning_rate": 2.4580248653183204e-09,
+      "loss": 0.2761,
+      "step": 12071
+    },
+    {
+      "epoch": 0.9904418099027772,
+      "grad_norm": 0.7464613919920773,
+      "learning_rate": 2.4165433004363914e-09,
+      "loss": 0.2863,
+      "step": 12072
+    },
+    {
+      "epoch": 0.990523854452968,
+      "grad_norm": 0.7267098821003252,
+      "learning_rate": 2.3754146568694124e-09,
+      "loss": 0.3269,
+      "step": 12073
+    },
+    {
+      "epoch": 0.9906058990031588,
+      "grad_norm": 0.8344591811903043,
+      "learning_rate": 2.334638937521172e-09,
+      "loss": 0.3132,
+      "step": 12074
+    },
+    {
+      "epoch": 0.9906879435533494,
+      "grad_norm": 0.7306083269006632,
+      "learning_rate": 2.2942161452715883e-09,
+      "loss": 0.286,
+      "step": 12075
+    },
+    {
+      "epoch": 0.9907699881035402,
+      "grad_norm": 0.7903321938869761,
+      "learning_rate": 2.25414628297449e-09,
+      "loss": 0.3116,
+      "step": 12076
+    },
+    {
+      "epoch": 0.990852032653731,
+      "grad_norm": 0.7234864551561773,
+      "learning_rate": 2.21442935346039e-09,
+      "loss": 0.2966,
+      "step": 12077
+    },
+    {
+      "epoch": 0.9909340772039217,
+      "grad_norm": 0.7118018459937393,
+      "learning_rate": 2.1750653595337125e-09,
+      "loss": 0.3035,
+      "step": 12078
+    },
+    {
+      "epoch": 0.9910161217541125,
+      "grad_norm": 0.789888007347094,
+      "learning_rate": 2.136054303974455e-09,
+      "loss": 0.2954,
+      "step": 12079
+    },
+    {
+      "epoch": 0.9910981663043033,
+      "grad_norm": 1.0986616055415148,
+      "learning_rate": 2.097396189537637e-09,
+      "loss": 0.3254,
+      "step": 12080
+    },
+    {
+      "epoch": 0.991180210854494,
+      "grad_norm": 0.6773704384309996,
+      "learning_rate": 2.059091018952186e-09,
+      "loss": 0.3207,
+      "step": 12081
+    },
+    {
+      "epoch": 0.9912622554046847,
+      "grad_norm": 0.6383976740267153,
+      "learning_rate": 2.0211387949242712e-09,
+      "loss": 0.3174,
+      "step": 12082
+    },
+    {
+      "epoch": 0.9913442999548755,
+      "grad_norm": 0.7096954231602314,
+      "learning_rate": 1.98353952013397e-09,
+      "loss": 0.3213,
+      "step": 12083
+    },
+    {
+      "epoch": 0.9914263445050663,
+      "grad_norm": 0.7140533195408271,
+      "learning_rate": 1.9462931972358268e-09,
+      "loss": 0.2922,
+      "step": 12084
+    },
+    {
+      "epoch": 0.991508389055257,
+      "grad_norm": 0.7201562744898937,
+      "learning_rate": 1.9093998288605144e-09,
+      "loss": 0.297,
+      "step": 12085
+    },
+    {
+      "epoch": 0.9915904336054477,
+      "grad_norm": 0.7756326626687771,
+      "learning_rate": 1.8728594176131707e-09,
+      "loss": 0.2931,
+      "step": 12086
+    },
+    {
+      "epoch": 0.9916724781556385,
+      "grad_norm": 1.2652358533238566,
+      "learning_rate": 1.8366719660750653e-09,
+      "loss": 0.3243,
+      "step": 12087
+    },
+    {
+      "epoch": 0.9917545227058293,
+      "grad_norm": 0.7715954884269426,
+      "learning_rate": 1.8008374768002657e-09,
+      "loss": 0.3127,
+      "step": 12088
+    },
+    {
+      "epoch": 0.99183656725602,
+      "grad_norm": 0.6839474039602714,
+      "learning_rate": 1.7653559523206353e-09,
+      "loss": 0.2975,
+      "step": 12089
+    },
+    {
+      "epoch": 0.9919186118062108,
+      "grad_norm": 0.649725165773083,
+      "learning_rate": 1.7302273951408376e-09,
+      "loss": 0.2967,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9920006563564016,
+      "grad_norm": 0.6807266999467324,
+      "learning_rate": 1.695451807742221e-09,
+      "loss": 0.3123,
+      "step": 12091
+    },
+    {
+      "epoch": 0.9920827009065922,
+      "grad_norm": 0.7490932968864299,
+      "learning_rate": 1.661029192580599e-09,
+      "loss": 0.2549,
+      "step": 12092
+    },
+    {
+      "epoch": 0.992164745456783,
+      "grad_norm": 0.77747696663745,
+      "learning_rate": 1.6269595520862492e-09,
+      "loss": 0.2846,
+      "step": 12093
+    },
+    {
+      "epoch": 0.9922467900069738,
+      "grad_norm": 0.7280692182470332,
+      "learning_rate": 1.5932428886661355e-09,
+      "loss": 0.2826,
+      "step": 12094
+    },
+    {
+      "epoch": 0.9923288345571646,
+      "grad_norm": 0.8378323751021993,
+      "learning_rate": 1.5598792047000212e-09,
+      "loss": 0.269,
+      "step": 12095
+    },
+    {
+      "epoch": 0.9924108791073553,
+      "grad_norm": 0.7147663434836368,
+      "learning_rate": 1.5268685025449093e-09,
+      "loss": 0.2851,
+      "step": 12096
+    },
+    {
+      "epoch": 0.992492923657546,
+      "grad_norm": 0.7127686052171058,
+      "learning_rate": 1.4942107845317132e-09,
+      "loss": 0.3436,
+      "step": 12097
+    },
+    {
+      "epoch": 0.9925749682077368,
+      "grad_norm": 0.7049640916240326,
+      "learning_rate": 1.4619060529663664e-09,
+      "loss": 0.3168,
+      "step": 12098
+    },
+    {
+      "epoch": 0.9926570127579275,
+      "grad_norm": 0.9930834411311188,
+      "learning_rate": 1.4299543101309321e-09,
+      "loss": 0.2859,
+      "step": 12099
+    },
+    {
+      "epoch": 0.9927390573081183,
+      "grad_norm": 0.7763617665508145,
+      "learning_rate": 1.3983555582808283e-09,
+      "loss": 0.3169,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9928211018583091,
+      "grad_norm": 0.6932619135201484,
+      "learning_rate": 1.3671097996487137e-09,
+      "loss": 0.3018,
+      "step": 12101
+    },
+    {
+      "epoch": 0.9929031464084999,
+      "grad_norm": 0.6887769570773651,
+      "learning_rate": 1.336217036439491e-09,
+      "loss": 0.2779,
+      "step": 12102
+    },
+    {
+      "epoch": 0.9929851909586905,
+      "grad_norm": 0.7589839286393315,
+      "learning_rate": 1.305677270836414e-09,
+      "loss": 0.3101,
+      "step": 12103
+    },
+    {
+      "epoch": 0.9930672355088813,
+      "grad_norm": 0.6364487346079452,
+      "learning_rate": 1.2754905049949807e-09,
+      "loss": 0.2337,
+      "step": 12104
+    },
+    {
+      "epoch": 0.9931492800590721,
+      "grad_norm": 0.6893315274832005,
+      "learning_rate": 1.2456567410473742e-09,
+      "loss": 0.3294,
+      "step": 12105
+    },
+    {
+      "epoch": 0.9932313246092628,
+      "grad_norm": 0.8215547828397164,
+      "learning_rate": 1.2161759811002428e-09,
+      "loss": 0.2799,
+      "step": 12106
+    },
+    {
+      "epoch": 0.9933133691594536,
+      "grad_norm": 0.9293954220756048,
+      "learning_rate": 1.1870482272358096e-09,
+      "loss": 0.2985,
+      "step": 12107
+    },
+    {
+      "epoch": 0.9933954137096443,
+      "grad_norm": 0.6881899743896706,
+      "learning_rate": 1.1582734815107633e-09,
+      "loss": 0.3037,
+      "step": 12108
+    },
+    {
+      "epoch": 0.9934774582598351,
+      "grad_norm": 0.7734113944130752,
+      "learning_rate": 1.1298517459573666e-09,
+      "loss": 0.3005,
+      "step": 12109
+    },
+    {
+      "epoch": 0.9935595028100258,
+      "grad_norm": 0.7079005660307379,
+      "learning_rate": 1.1017830225823478e-09,
+      "loss": 0.3145,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9936415473602166,
+      "grad_norm": 0.6497412952769558,
+      "learning_rate": 1.0740673133680102e-09,
+      "loss": 0.2905,
+      "step": 12111
+    },
+    {
+      "epoch": 0.9937235919104074,
+      "grad_norm": 0.7442311138812682,
+      "learning_rate": 1.046704620271677e-09,
+      "loss": 0.2636,
+      "step": 12112
+    },
+    {
+      "epoch": 0.993805636460598,
+      "grad_norm": 0.7497516587316622,
+      "learning_rate": 1.019694945225691e-09,
+      "loss": 0.3034,
+      "step": 12113
+    },
+    {
+      "epoch": 0.9938876810107888,
+      "grad_norm": 0.686756718975904,
+      "learning_rate": 9.930382901374157e-10,
+      "loss": 0.3236,
+      "step": 12114
+    },
+    {
+      "epoch": 0.9939697255609796,
+      "grad_norm": 0.6955790036871125,
+      "learning_rate": 9.667346568892344e-10,
+      "loss": 0.2918,
+      "step": 12115
+    },
+    {
+      "epoch": 0.9940517701111704,
+      "grad_norm": 0.7442942191475674,
+      "learning_rate": 9.407840473385499e-10,
+      "loss": 0.2902,
+      "step": 12116
+    },
+    {
+      "epoch": 0.9941338146613611,
+      "grad_norm": 0.8365599911759198,
+      "learning_rate": 9.151864633183405e-10,
+      "loss": 0.3222,
+      "step": 12117
+    },
+    {
+      "epoch": 0.9942158592115519,
+      "grad_norm": 0.7276170706019655,
+      "learning_rate": 8.899419066360492e-10,
+      "loss": 0.3172,
+      "step": 12118
+    },
+    {
+      "epoch": 0.9942979037617427,
+      "grad_norm": 0.7372365016067817,
+      "learning_rate": 8.650503790741394e-10,
+      "loss": 0.3018,
+      "step": 12119
+    },
+    {
+      "epoch": 0.9943799483119333,
+      "grad_norm": 0.7323286009602681,
+      "learning_rate": 8.405118823906489e-10,
+      "loss": 0.2741,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9944619928621241,
+      "grad_norm": 0.6801148325772978,
+      "learning_rate": 8.163264183186359e-10,
+      "loss": 0.2788,
+      "step": 12121
+    },
+    {
+      "epoch": 0.9945440374123149,
+      "grad_norm": 0.7320228669586563,
+      "learning_rate": 7.924939885656235e-10,
+      "loss": 0.3247,
+      "step": 12122
+    },
+    {
+      "epoch": 0.9946260819625057,
+      "grad_norm": 0.6998227784924308,
+      "learning_rate": 7.690145948152649e-10,
+      "loss": 0.2984,
+      "step": 12123
+    },
+    {
+      "epoch": 0.9947081265126964,
+      "grad_norm": 0.6822689190777981,
+      "learning_rate": 7.458882387245681e-10,
+      "loss": 0.3119,
+      "step": 12124
+    },
+    {
+      "epoch": 0.9947901710628871,
+      "grad_norm": 1.1432553741525204,
+      "learning_rate": 7.231149219277811e-10,
+      "loss": 0.3086,
+      "step": 12125
+    },
+    {
+      "epoch": 0.9948722156130779,
+      "grad_norm": 0.8174961107312826,
+      "learning_rate": 7.00694646032507e-10,
+      "loss": 0.2956,
+      "step": 12126
+    },
+    {
+      "epoch": 0.9949542601632686,
+      "grad_norm": 0.8813890706848693,
+      "learning_rate": 6.78627412622479e-10,
+      "loss": 0.3507,
+      "step": 12127
+    },
+    {
+      "epoch": 0.9950363047134594,
+      "grad_norm": 0.7459278609136889,
+      "learning_rate": 6.569132232553399e-10,
+      "loss": 0.3085,
+      "step": 12128
+    },
+    {
+      "epoch": 0.9951183492636502,
+      "grad_norm": 0.6936859796155047,
+      "learning_rate": 6.35552079465418e-10,
+      "loss": 0.2807,
+      "step": 12129
+    },
+    {
+      "epoch": 0.995200393813841,
+      "grad_norm": 0.7066791681711961,
+      "learning_rate": 6.145439827603961e-10,
+      "loss": 0.3011,
+      "step": 12130
+    },
+    {
+      "epoch": 0.9952824383640316,
+      "grad_norm": 0.6654944860608583,
+      "learning_rate": 5.938889346240873e-10,
+      "loss": 0.3282,
+      "step": 12131
+    },
+    {
+      "epoch": 0.9953644829142224,
+      "grad_norm": 0.7424028604245797,
+      "learning_rate": 5.735869365153247e-10,
+      "loss": 0.2819,
+      "step": 12132
+    },
+    {
+      "epoch": 0.9954465274644132,
+      "grad_norm": 0.6457238467336138,
+      "learning_rate": 5.536379898674061e-10,
+      "loss": 0.2749,
+      "step": 12133
+    },
+    {
+      "epoch": 0.995528572014604,
+      "grad_norm": 0.697404311749881,
+      "learning_rate": 5.340420960897597e-10,
+      "loss": 0.3021,
+      "step": 12134
+    },
+    {
+      "epoch": 0.9956106165647947,
+      "grad_norm": 0.735622371137498,
+      "learning_rate": 5.147992565657234e-10,
+      "loss": 0.3546,
+      "step": 12135
+    },
+    {
+      "epoch": 0.9956926611149854,
+      "grad_norm": 0.8620438060987387,
+      "learning_rate": 4.959094726542102e-10,
+      "loss": 0.3078,
+      "step": 12136
+    },
+    {
+      "epoch": 0.9957747056651762,
+      "grad_norm": 0.7490479748944667,
+      "learning_rate": 4.77372745689153e-10,
+      "loss": 0.2913,
+      "step": 12137
+    },
+    {
+      "epoch": 0.9958567502153669,
+      "grad_norm": 0.7997867436209518,
+      "learning_rate": 4.591890769795049e-10,
+      "loss": 0.2541,
+      "step": 12138
+    },
+    {
+      "epoch": 0.9959387947655577,
+      "grad_norm": 0.6768657521444497,
+      "learning_rate": 4.4135846780979373e-10,
+      "loss": 0.3164,
+      "step": 12139
+    },
+    {
+      "epoch": 0.9960208393157485,
+      "grad_norm": 0.6971419056316803,
+      "learning_rate": 4.2388091943901257e-10,
+      "loss": 0.2878,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9961028838659393,
+      "grad_norm": 0.7185564820010338,
+      "learning_rate": 4.0675643310117417e-10,
+      "loss": 0.292,
+      "step": 12141
+    },
+    {
+      "epoch": 0.9961849284161299,
+      "grad_norm": 0.7245075954323954,
+      "learning_rate": 3.8998501000586664e-10,
+      "loss": 0.2532,
+      "step": 12142
+    },
+    {
+      "epoch": 0.9962669729663207,
+      "grad_norm": 0.7062481657985381,
+      "learning_rate": 3.735666513371428e-10,
+      "loss": 0.3324,
+      "step": 12143
+    },
+    {
+      "epoch": 0.9963490175165115,
+      "grad_norm": 0.7019017600811804,
+      "learning_rate": 3.575013582546305e-10,
+      "loss": 0.3247,
+      "step": 12144
+    },
+    {
+      "epoch": 0.9964310620667022,
+      "grad_norm": 0.7453614855340213,
+      "learning_rate": 3.417891318929778e-10,
+      "loss": 0.2827,
+      "step": 12145
+    },
+    {
+      "epoch": 0.996513106616893,
+      "grad_norm": 0.7269582015069586,
+      "learning_rate": 3.2642997336185257e-10,
+      "loss": 0.337,
+      "step": 12146
+    },
+    {
+      "epoch": 0.9965951511670837,
+      "grad_norm": 0.6626479973421284,
+      "learning_rate": 3.114238837453876e-10,
+      "loss": 0.2907,
+      "step": 12147
+    },
+    {
+      "epoch": 0.9966771957172745,
+      "grad_norm": 0.7682351476336463,
+      "learning_rate": 2.967708641032907e-10,
+      "loss": 0.3096,
+      "step": 12148
+    },
+    {
+      "epoch": 0.9967592402674652,
+      "grad_norm": 0.7327147814162173,
+      "learning_rate": 2.8247091547084496e-10,
+      "loss": 0.2565,
+      "step": 12149
+    },
+    {
+      "epoch": 0.996841284817656,
+      "grad_norm": 0.7149256545468203,
+      "learning_rate": 2.685240388577981e-10,
+      "loss": 0.3118,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9969233293678468,
+      "grad_norm": 0.9139052722347522,
+      "learning_rate": 2.54930235248918e-10,
+      "loss": 0.3033,
+      "step": 12151
+    },
+    {
+      "epoch": 0.9970053739180375,
+      "grad_norm": 0.7119349281319355,
+      "learning_rate": 2.4168950560399254e-10,
+      "loss": 0.2875,
+      "step": 12152
+    },
+    {
+      "epoch": 0.9970874184682282,
+      "grad_norm": 0.6788835895940576,
+      "learning_rate": 2.2880185085838447e-10,
+      "loss": 0.2964,
+      "step": 12153
+    },
+    {
+      "epoch": 0.997169463018419,
+      "grad_norm": 0.6719440568678516,
+      "learning_rate": 2.1626727192192164e-10,
+      "loss": 0.2829,
+      "step": 12154
+    },
+    {
+      "epoch": 0.9972515075686098,
+      "grad_norm": 0.7205698199116506,
+      "learning_rate": 2.0408576968000693e-10,
+      "loss": 0.291,
+      "step": 12155
+    },
+    {
+      "epoch": 0.9973335521188005,
+      "grad_norm": 0.7520700516753814,
+      "learning_rate": 1.9225734499250802e-10,
+      "loss": 0.318,
+      "step": 12156
+    },
+    {
+      "epoch": 0.9974155966689913,
+      "grad_norm": 0.7307349560232842,
+      "learning_rate": 1.807819986948678e-10,
+      "loss": 0.3024,
+      "step": 12157
+    },
+    {
+      "epoch": 0.997497641219182,
+      "grad_norm": 0.6858342788771874,
+      "learning_rate": 1.6965973159810413e-10,
+      "loss": 0.3087,
+      "step": 12158
+    },
+    {
+      "epoch": 0.9975796857693727,
+      "grad_norm": 0.6726860341918907,
+      "learning_rate": 1.5889054448658957e-10,
+      "loss": 0.3009,
+      "step": 12159
+    },
+    {
+      "epoch": 0.9976617303195635,
+      "grad_norm": 0.6838097611739047,
+      "learning_rate": 1.4847443812193717e-10,
+      "loss": 0.298,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9977437748697543,
+      "grad_norm": 0.680738351778282,
+      "learning_rate": 1.3841141323855946e-10,
+      "loss": 0.2895,
+      "step": 12161
+    },
+    {
+      "epoch": 0.9978258194199451,
+      "grad_norm": 0.6847825400400924,
+      "learning_rate": 1.2870147054810933e-10,
+      "loss": 0.3093,
+      "step": 12162
+    },
+    {
+      "epoch": 0.9979078639701358,
+      "grad_norm": 0.6637383014754411,
+      "learning_rate": 1.1934461073559444e-10,
+      "loss": 0.2859,
+      "step": 12163
+    },
+    {
+      "epoch": 0.9979899085203265,
+      "grad_norm": 0.6680863702533094,
+      "learning_rate": 1.103408344621526e-10,
+      "loss": 0.2753,
+      "step": 12164
+    },
+    {
+      "epoch": 0.9980719530705173,
+      "grad_norm": 0.6740179372925819,
+      "learning_rate": 1.0169014236394159e-10,
+      "loss": 0.3463,
+      "step": 12165
+    },
+    {
+      "epoch": 0.998153997620708,
+      "grad_norm": 0.7270722781893997,
+      "learning_rate": 9.339253505102896e-11,
+      "loss": 0.2506,
+      "step": 12166
+    },
+    {
+      "epoch": 0.9982360421708988,
+      "grad_norm": 0.8367408345914051,
+      "learning_rate": 8.544801310961248e-11,
+      "loss": 0.3332,
+      "step": 12167
+    },
+    {
+      "epoch": 0.9983180867210896,
+      "grad_norm": 0.7450982306246592,
+      "learning_rate": 7.785657710090988e-11,
+      "loss": 0.2769,
+      "step": 12168
+    },
+    {
+      "epoch": 0.9984001312712804,
+      "grad_norm": 0.6848980772909937,
+      "learning_rate": 7.061822756115888e-11,
+      "loss": 0.3022,
+      "step": 12169
+    },
+    {
+      "epoch": 0.998482175821471,
+      "grad_norm": 0.7260283720764462,
+      "learning_rate": 6.373296500106207e-11,
+      "loss": 0.2884,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9985642203716618,
+      "grad_norm": 0.7171866644126272,
+      "learning_rate": 5.720078990745226e-11,
+      "loss": 0.2959,
+      "step": 12171
+    },
+    {
+      "epoch": 0.9986462649218526,
+      "grad_norm": 0.7008026972218931,
+      "learning_rate": 5.10217027416271e-11,
+      "loss": 0.2751,
+      "step": 12172
+    },
+    {
+      "epoch": 0.9987283094720433,
+      "grad_norm": 0.7313592627555455,
+      "learning_rate": 4.519570393879402e-11,
+      "loss": 0.2885,
+      "step": 12173
+    },
+    {
+      "epoch": 0.9988103540222341,
+      "grad_norm": 0.6808936275331859,
+      "learning_rate": 3.972279391195599e-11,
+      "loss": 0.3299,
+      "step": 12174
+    },
+    {
+      "epoch": 0.9988923985724248,
+      "grad_norm": 0.7229704094762887,
+      "learning_rate": 3.4602973046360396e-11,
+      "loss": 0.2828,
+      "step": 12175
+    },
+    {
+      "epoch": 0.9989744431226156,
+      "grad_norm": 0.6721469598558081,
+      "learning_rate": 2.983624170393995e-11,
+      "loss": 0.2741,
+      "step": 12176
+    },
+    {
+      "epoch": 0.9990564876728063,
+      "grad_norm": 0.73856770978565,
+      "learning_rate": 2.5422600221647332e-11,
+      "loss": 0.3132,
+      "step": 12177
+    },
+    {
+      "epoch": 0.9991385322229971,
+      "grad_norm": 0.7158585408689695,
+      "learning_rate": 2.1362048910900103e-11,
+      "loss": 0.2754,
+      "step": 12178
+    },
+    {
+      "epoch": 0.9992205767731879,
+      "grad_norm": 0.6213247263896023,
+      "learning_rate": 1.7654588058690914e-11,
+      "loss": 0.2786,
+      "step": 12179
+    },
+    {
+      "epoch": 0.9993026213233785,
+      "grad_norm": 0.7056029790445787,
+      "learning_rate": 1.4300217926477289e-11,
+      "loss": 0.3241,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9993846658735693,
+      "grad_norm": 0.7338911248734207,
+      "learning_rate": 1.1298938751291843e-11,
+      "loss": 0.3309,
+      "step": 12181
+    },
+    {
+      "epoch": 0.9994667104237601,
+      "grad_norm": 0.730576473900203,
+      "learning_rate": 8.650750745187175e-12,
+      "loss": 0.3106,
+      "step": 12182
+    },
+    {
+      "epoch": 0.9995487549739509,
+      "grad_norm": 0.744057119528094,
+      "learning_rate": 6.355654094680752e-12,
+      "loss": 0.301,
+      "step": 12183
+    },
+    {
+      "epoch": 0.9996307995241416,
+      "grad_norm": 0.697079369277063,
+      "learning_rate": 4.413648962975358e-12,
+      "loss": 0.2749,
+      "step": 12184
+    },
+    {
+      "epoch": 0.9997128440743324,
+      "grad_norm": 0.6239705306645849,
+      "learning_rate": 2.8247354855182043e-12,
+      "loss": 0.2758,
+      "step": 12185
+    },
+    {
+      "epoch": 0.9997948886245231,
+      "grad_norm": 0.8171981682247088,
+      "learning_rate": 1.5889137761071482e-12,
+      "loss": 0.3068,
+      "step": 12186
+    },
+    {
+      "epoch": 0.9998769331747138,
+      "grad_norm": 0.6937437990715053,
+      "learning_rate": 7.061839213395871e-13,
+      "loss": 0.3017,
+      "step": 12187
+    },
+    {
+      "epoch": 0.9999589777249046,
+      "grad_norm": 0.793574577878814,
+      "learning_rate": 1.765459833880101e-13,
+      "loss": 0.2951,
+      "step": 12188
+    },
+    {
+      "epoch": 0.9999589777249046,
+      "step": 12188,
+      "total_flos": 8.074170064909107e+16,
+      "train_loss": 0.02913342710690769,
+      "train_runtime": 9597.8769,
+      "train_samples_per_second": 650.242,
+      "train_steps_per_second": 1.27
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12188,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.074170064909107e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}