{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2539,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009846396218983852,
      "grad_norm": 5.9883294105529785,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.2518,
      "step": 25
    },
    {
      "epoch": 0.019692792437967704,
      "grad_norm": 5.031759262084961,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.285,
      "step": 50
    },
    {
      "epoch": 0.029539188656951557,
      "grad_norm": 4.0184102058410645,
      "learning_rate": 1.5e-06,
      "loss": 0.2444,
      "step": 75
    },
    {
      "epoch": 0.03938558487593541,
      "grad_norm": 4.6933512687683105,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.2471,
      "step": 100
    },
    {
      "epoch": 0.04923198109491926,
      "grad_norm": 6.389063835144043,
      "learning_rate": 2.5e-06,
      "loss": 0.2567,
      "step": 125
    },
    {
      "epoch": 0.05907837731390311,
      "grad_norm": 5.830894947052002,
      "learning_rate": 3e-06,
      "loss": 0.2274,
      "step": 150
    },
    {
      "epoch": 0.06892477353288697,
      "grad_norm": 3.234384536743164,
      "learning_rate": 3.5e-06,
      "loss": 0.2415,
      "step": 175
    },
    {
      "epoch": 0.07877116975187082,
      "grad_norm": 4.506255149841309,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2362,
      "step": 200
    },
    {
      "epoch": 0.08861756597085467,
      "grad_norm": 4.6171674728393555,
      "learning_rate": 4.5e-06,
      "loss": 0.207,
      "step": 225
    },
    {
      "epoch": 0.09846396218983852,
      "grad_norm": 3.5775146484375,
      "learning_rate": 5e-06,
      "loss": 0.209,
      "step": 250
    },
    {
      "epoch": 0.10831035840882237,
      "grad_norm": 4.521059989929199,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.2349,
      "step": 275
    },
    {
      "epoch": 0.11815675462780623,
      "grad_norm": 4.866330623626709,
      "learning_rate": 6e-06,
      "loss": 0.2528,
      "step": 300
    },
    {
      "epoch": 0.12800315084679006,
      "grad_norm": 4.609464645385742,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.1845,
      "step": 325
    },
    {
      "epoch": 0.13784954706577393,
      "grad_norm": 4.313091278076172,
      "learning_rate": 7e-06,
      "loss": 0.2509,
      "step": 350
    },
    {
      "epoch": 0.14769594328475777,
      "grad_norm": 4.451571941375732,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.2397,
      "step": 375
    },
    {
      "epoch": 0.15754233950374164,
      "grad_norm": 4.453555107116699,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2273,
      "step": 400
    },
    {
      "epoch": 0.16738873572272547,
      "grad_norm": 3.1133830547332764,
      "learning_rate": 8.5e-06,
      "loss": 0.2333,
      "step": 425
    },
    {
      "epoch": 0.17723513194170934,
      "grad_norm": 3.7271251678466797,
      "learning_rate": 9e-06,
      "loss": 0.2183,
      "step": 450
    },
    {
      "epoch": 0.18708152816069318,
      "grad_norm": 3.5686957836151123,
      "learning_rate": 9.5e-06,
      "loss": 0.2217,
      "step": 475
    },
    {
      "epoch": 0.19692792437967704,
      "grad_norm": 3.4622321128845215,
      "learning_rate": 1e-05,
      "loss": 0.2243,
      "step": 500
    },
    {
      "epoch": 0.20677432059866088,
      "grad_norm": 5.357515811920166,
      "learning_rate": 9.877390877881315e-06,
      "loss": 0.216,
      "step": 525
    },
    {
      "epoch": 0.21662071681764475,
      "grad_norm": 4.189847469329834,
      "learning_rate": 9.75478175576263e-06,
      "loss": 0.2293,
      "step": 550
    },
    {
      "epoch": 0.2264671130366286,
      "grad_norm": 5.216653347015381,
      "learning_rate": 9.632172633643944e-06,
      "loss": 0.2119,
      "step": 575
    },
    {
      "epoch": 0.23631350925561245,
      "grad_norm": 3.989429235458374,
      "learning_rate": 9.509563511525259e-06,
      "loss": 0.2028,
      "step": 600
    },
    {
      "epoch": 0.2461599054745963,
      "grad_norm": 4.413074970245361,
      "learning_rate": 9.386954389406573e-06,
      "loss": 0.2317,
      "step": 625
    },
    {
      "epoch": 0.25600630169358013,
      "grad_norm": 4.636322021484375,
      "learning_rate": 9.264345267287887e-06,
      "loss": 0.2295,
      "step": 650
    },
    {
      "epoch": 0.265852697912564,
      "grad_norm": 4.068817615509033,
      "learning_rate": 9.141736145169202e-06,
      "loss": 0.2168,
      "step": 675
    },
    {
      "epoch": 0.27569909413154786,
      "grad_norm": 3.76188588142395,
      "learning_rate": 9.019127023050516e-06,
      "loss": 0.2085,
      "step": 700
    },
    {
      "epoch": 0.2855454903505317,
      "grad_norm": 5.307705879211426,
      "learning_rate": 8.89651790093183e-06,
      "loss": 0.2234,
      "step": 725
    },
    {
      "epoch": 0.29539188656951554,
      "grad_norm": 5.800079822540283,
      "learning_rate": 8.773908778813145e-06,
      "loss": 0.2501,
      "step": 750
    },
    {
      "epoch": 0.3052382827884994,
      "grad_norm": 3.7706942558288574,
      "learning_rate": 8.651299656694458e-06,
      "loss": 0.2425,
      "step": 775
    },
    {
      "epoch": 0.31508467900748327,
      "grad_norm": 4.2593512535095215,
      "learning_rate": 8.528690534575772e-06,
      "loss": 0.215,
      "step": 800
    },
    {
      "epoch": 0.32493107522646714,
      "grad_norm": 5.396867752075195,
      "learning_rate": 8.406081412457088e-06,
      "loss": 0.1831,
      "step": 825
    },
    {
      "epoch": 0.33477747144545095,
      "grad_norm": 3.5747408866882324,
      "learning_rate": 8.283472290338403e-06,
      "loss": 0.1976,
      "step": 850
    },
    {
      "epoch": 0.3446238676644348,
      "grad_norm": 5.190290451049805,
      "learning_rate": 8.160863168219716e-06,
      "loss": 0.2329,
      "step": 875
    },
    {
      "epoch": 0.3544702638834187,
      "grad_norm": 6.033483505249023,
      "learning_rate": 8.043158410985778e-06,
      "loss": 0.2233,
      "step": 900
    },
    {
      "epoch": 0.36431666010240255,
      "grad_norm": 3.0238583087921143,
      "learning_rate": 7.920549288867092e-06,
      "loss": 0.2168,
      "step": 925
    },
    {
      "epoch": 0.37416305632138636,
      "grad_norm": 4.231501579284668,
      "learning_rate": 7.797940166748407e-06,
      "loss": 0.2065,
      "step": 950
    },
    {
      "epoch": 0.3840094525403702,
      "grad_norm": 3.6126151084899902,
      "learning_rate": 7.675331044629721e-06,
      "loss": 0.2126,
      "step": 975
    },
    {
      "epoch": 0.3938558487593541,
      "grad_norm": 4.870268821716309,
      "learning_rate": 7.552721922511036e-06,
      "loss": 0.2269,
      "step": 1000
    },
    {
      "epoch": 0.40370224497833795,
      "grad_norm": 4.647860527038574,
      "learning_rate": 7.430112800392349e-06,
      "loss": 0.1892,
      "step": 1025
    },
    {
      "epoch": 0.41354864119732176,
      "grad_norm": 4.208144187927246,
      "learning_rate": 7.307503678273664e-06,
      "loss": 0.181,
      "step": 1050
    },
    {
      "epoch": 0.42339503741630563,
      "grad_norm": 4.470357894897461,
      "learning_rate": 7.184894556154979e-06,
      "loss": 0.1931,
      "step": 1075
    },
    {
      "epoch": 0.4332414336352895,
      "grad_norm": 3.936370372772217,
      "learning_rate": 7.062285434036293e-06,
      "loss": 0.1884,
      "step": 1100
    },
    {
      "epoch": 0.44308782985427336,
      "grad_norm": 2.863593816757202,
      "learning_rate": 6.939676311917607e-06,
      "loss": 0.2054,
      "step": 1125
    },
    {
      "epoch": 0.4529342260732572,
      "grad_norm": 3.7073140144348145,
      "learning_rate": 6.817067189798921e-06,
      "loss": 0.202,
      "step": 1150
    },
    {
      "epoch": 0.46278062229224104,
      "grad_norm": 3.3471767902374268,
      "learning_rate": 6.6944580676802366e-06,
      "loss": 0.1962,
      "step": 1175
    },
    {
      "epoch": 0.4726270185112249,
      "grad_norm": 2.521800994873047,
      "learning_rate": 6.571848945561551e-06,
      "loss": 0.1967,
      "step": 1200
    },
    {
      "epoch": 0.4824734147302087,
      "grad_norm": 4.672091007232666,
      "learning_rate": 6.4492398234428646e-06,
      "loss": 0.1866,
      "step": 1225
    },
    {
      "epoch": 0.4923198109491926,
      "grad_norm": 4.11766242980957,
      "learning_rate": 6.326630701324179e-06,
      "loss": 0.2203,
      "step": 1250
    },
    {
      "epoch": 0.5021662071681764,
      "grad_norm": 3.7128915786743164,
      "learning_rate": 6.204021579205493e-06,
      "loss": 0.1754,
      "step": 1275
    },
    {
      "epoch": 0.5120126033871603,
      "grad_norm": 2.663398265838623,
      "learning_rate": 6.081412457086808e-06,
      "loss": 0.1737,
      "step": 1300
    },
    {
      "epoch": 0.5218589996061441,
      "grad_norm": 2.656521797180176,
      "learning_rate": 5.9588033349681214e-06,
      "loss": 0.2085,
      "step": 1325
    },
    {
      "epoch": 0.531705395825128,
      "grad_norm": 4.869094371795654,
      "learning_rate": 5.836194212849437e-06,
      "loss": 0.1943,
      "step": 1350
    },
    {
      "epoch": 0.5415517920441119,
      "grad_norm": 4.049020290374756,
      "learning_rate": 5.713585090730751e-06,
      "loss": 0.2035,
      "step": 1375
    },
    {
      "epoch": 0.5513981882630957,
      "grad_norm": 3.473222255706787,
      "learning_rate": 5.5909759686120656e-06,
      "loss": 0.1774,
      "step": 1400
    },
    {
      "epoch": 0.5612445844820796,
      "grad_norm": 4.270164489746094,
      "learning_rate": 5.468366846493379e-06,
      "loss": 0.202,
      "step": 1425
    },
    {
      "epoch": 0.5710909807010635,
      "grad_norm": 2.7588560581207275,
      "learning_rate": 5.345757724374694e-06,
      "loss": 0.2143,
      "step": 1450
    },
    {
      "epoch": 0.5809373769200472,
      "grad_norm": 4.859555721282959,
      "learning_rate": 5.223148602256009e-06,
      "loss": 0.2009,
      "step": 1475
    },
    {
      "epoch": 0.5907837731390311,
      "grad_norm": 5.985973358154297,
      "learning_rate": 5.100539480137323e-06,
      "loss": 0.2334,
      "step": 1500
    },
    {
      "epoch": 0.6006301693580149,
      "grad_norm": 4.282456398010254,
      "learning_rate": 4.977930358018637e-06,
      "loss": 0.2082,
      "step": 1525
    },
    {
      "epoch": 0.6104765655769988,
      "grad_norm": 3.605886697769165,
      "learning_rate": 4.855321235899951e-06,
      "loss": 0.2204,
      "step": 1550
    },
    {
      "epoch": 0.6203229617959827,
      "grad_norm": 4.4538044929504395,
      "learning_rate": 4.732712113781266e-06,
      "loss": 0.1548,
      "step": 1575
    },
    {
      "epoch": 0.6301693580149665,
      "grad_norm": 2.746072769165039,
      "learning_rate": 4.61010299166258e-06,
      "loss": 0.2109,
      "step": 1600
    },
    {
      "epoch": 0.6400157542339504,
      "grad_norm": 4.928244113922119,
      "learning_rate": 4.487493869543895e-06,
      "loss": 0.18,
      "step": 1625
    },
    {
      "epoch": 0.6498621504529343,
      "grad_norm": 3.3110663890838623,
      "learning_rate": 4.364884747425209e-06,
      "loss": 0.1707,
      "step": 1650
    },
    {
      "epoch": 0.659708546671918,
      "grad_norm": 2.6316301822662354,
      "learning_rate": 4.2422756253065234e-06,
      "loss": 0.2035,
      "step": 1675
    },
    {
      "epoch": 0.6695549428909019,
      "grad_norm": 4.228478908538818,
      "learning_rate": 4.119666503187837e-06,
      "loss": 0.201,
      "step": 1700
    },
    {
      "epoch": 0.6794013391098858,
      "grad_norm": 4.702217102050781,
      "learning_rate": 3.997057381069152e-06,
      "loss": 0.1919,
      "step": 1725
    },
    {
      "epoch": 0.6892477353288696,
      "grad_norm": 4.416632652282715,
      "learning_rate": 3.874448258950466e-06,
      "loss": 0.1837,
      "step": 1750
    },
    {
      "epoch": 0.6990941315478535,
      "grad_norm": 4.304840087890625,
      "learning_rate": 3.7518391368317807e-06,
      "loss": 0.1866,
      "step": 1775
    },
    {
      "epoch": 0.7089405277668374,
      "grad_norm": 2.635377883911133,
      "learning_rate": 3.6292300147130947e-06,
      "loss": 0.1949,
      "step": 1800
    },
    {
      "epoch": 0.7187869239858212,
      "grad_norm": 4.77269172668457,
      "learning_rate": 3.5066208925944096e-06,
      "loss": 0.202,
      "step": 1825
    },
    {
      "epoch": 0.7286333202048051,
      "grad_norm": 3.0173046588897705,
      "learning_rate": 3.3840117704757236e-06,
      "loss": 0.1867,
      "step": 1850
    },
    {
      "epoch": 0.7384797164237888,
      "grad_norm": 3.118154764175415,
      "learning_rate": 3.261402648357038e-06,
      "loss": 0.1681,
      "step": 1875
    },
    {
      "epoch": 0.7483261126427727,
      "grad_norm": 4.375317573547363,
      "learning_rate": 3.1387935262383525e-06,
      "loss": 0.1782,
      "step": 1900
    },
    {
      "epoch": 0.7581725088617566,
      "grad_norm": 5.0968122482299805,
      "learning_rate": 3.016184404119667e-06,
      "loss": 0.1862,
      "step": 1925
    },
    {
      "epoch": 0.7680189050807404,
      "grad_norm": 4.492552280426025,
      "learning_rate": 2.893575282000981e-06,
      "loss": 0.1951,
      "step": 1950
    },
    {
      "epoch": 0.7778653012997243,
      "grad_norm": 3.4022319316864014,
      "learning_rate": 2.7709661598822958e-06,
      "loss": 0.1873,
      "step": 1975
    },
    {
      "epoch": 0.7877116975187082,
      "grad_norm": 3.6453022956848145,
      "learning_rate": 2.6483570377636098e-06,
      "loss": 0.2134,
      "step": 2000
    },
    {
      "epoch": 0.797558093737692,
      "grad_norm": 2.7721006870269775,
      "learning_rate": 2.525747915644924e-06,
      "loss": 0.1731,
      "step": 2025
    },
    {
      "epoch": 0.8074044899566759,
      "grad_norm": 3.6965274810791016,
      "learning_rate": 2.4031387935262386e-06,
      "loss": 0.1581,
      "step": 2050
    },
    {
      "epoch": 0.8172508861756597,
      "grad_norm": 3.9200284481048584,
      "learning_rate": 2.2805296714075526e-06,
      "loss": 0.2,
      "step": 2075
    },
    {
      "epoch": 0.8270972823946435,
      "grad_norm": 3.353593587875366,
      "learning_rate": 2.157920549288867e-06,
      "loss": 0.1836,
      "step": 2100
    },
    {
      "epoch": 0.8369436786136274,
      "grad_norm": 3.3680572509765625,
      "learning_rate": 2.0353114271701815e-06,
      "loss": 0.2145,
      "step": 2125
    },
    {
      "epoch": 0.8467900748326113,
      "grad_norm": 2.7071352005004883,
      "learning_rate": 1.912702305051496e-06,
      "loss": 0.1837,
      "step": 2150
    },
    {
      "epoch": 0.8566364710515951,
      "grad_norm": 3.3706490993499756,
      "learning_rate": 1.7900931829328103e-06,
      "loss": 0.1642,
      "step": 2175
    },
    {
      "epoch": 0.866482867270579,
      "grad_norm": 3.4134597778320312,
      "learning_rate": 1.6674840608141246e-06,
      "loss": 0.1829,
      "step": 2200
    },
    {
      "epoch": 0.8763292634895629,
      "grad_norm": 4.351123809814453,
      "learning_rate": 1.544874938695439e-06,
      "loss": 0.1766,
      "step": 2225
    },
    {
      "epoch": 0.8861756597085467,
      "grad_norm": 4.491402626037598,
      "learning_rate": 1.4222658165767534e-06,
      "loss": 0.1793,
      "step": 2250
    },
    {
      "epoch": 0.8960220559275305,
      "grad_norm": 4.368827819824219,
      "learning_rate": 1.2996566944580676e-06,
      "loss": 0.1871,
      "step": 2275
    },
    {
      "epoch": 0.9058684521465143,
      "grad_norm": 3.653308868408203,
      "learning_rate": 1.1770475723393823e-06,
      "loss": 0.1685,
      "step": 2300
    },
    {
      "epoch": 0.9157148483654982,
      "grad_norm": 4.542746067047119,
      "learning_rate": 1.0544384502206965e-06,
      "loss": 0.1921,
      "step": 2325
    },
    {
      "epoch": 0.9255612445844821,
      "grad_norm": 3.869722366333008,
      "learning_rate": 9.318293281020109e-07,
      "loss": 0.1985,
      "step": 2350
    },
    {
      "epoch": 0.935407640803466,
      "grad_norm": 5.539394378662109,
      "learning_rate": 8.092202059833253e-07,
      "loss": 0.233,
      "step": 2375
    },
    {
      "epoch": 0.9452540370224498,
      "grad_norm": 4.2061872482299805,
      "learning_rate": 6.866110838646396e-07,
      "loss": 0.1844,
      "step": 2400
    },
    {
      "epoch": 0.9551004332414337,
      "grad_norm": 5.669102191925049,
      "learning_rate": 5.64001961745954e-07,
      "loss": 0.208,
      "step": 2425
    },
    {
      "epoch": 0.9649468294604174,
      "grad_norm": 2.2702748775482178,
      "learning_rate": 4.4139283962726833e-07,
      "loss": 0.188,
      "step": 2450
    },
    {
      "epoch": 0.9747932256794013,
      "grad_norm": 2.7447268962860107,
      "learning_rate": 3.187837175085827e-07,
      "loss": 0.1868,
      "step": 2475
    },
    {
      "epoch": 0.9846396218983852,
      "grad_norm": 4.356093406677246,
      "learning_rate": 1.9617459538989703e-07,
      "loss": 0.1993,
      "step": 2500
    },
    {
      "epoch": 0.994486018117369,
      "grad_norm": 3.6656863689422607,
      "learning_rate": 7.356547327121139e-08,
      "loss": 0.1833,
      "step": 2525
    }
  ],
  "logging_steps": 25,
  "max_steps": 2539,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.172262754639872e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}