{"optimizer_step": 1000, "train/token_ce": 6.9117332091555, "train/token_acc1": 0.00834147158820997, "dev/token_ce": 6.94883628655225, "dev/token_acc1": 0.005981445487122983, "train/acc_by_position": [0.017578125, 0.013671875, 0.03515625, 0.013671875, 0.060546875, 0.0, 0.0, 0.001953125, 0.0, 0.013671875, 0.021484375, 0.00390625, 0.005859375, 0.013671875, 0.04296875, 0.02734375, 0.005859375, 0.005859375, 0.00390625, 0.0, 0.009765625, 0.00390625, 0.0078125, 0.009765625, 0.009765625, 0.00390625, 0.001953125, 0.005859375, 0.00390625, 0.00390625, 0.005859375, 0.0, 0.001953125, 0.0078125, 0.0, 0.0, 0.005859375, 0.00390625, 0.0, 0.0078125, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.001953125, 0.005859375, 0.00390625, 0.001953125], "dev/acc_by_position": [0.015625, 0.0078125, 0.017578125, 0.013671875, 0.03125, 0.0, 0.0, 0.0, 0.0, 0.005859375, 0.017578125, 0.005859375, 0.005859375, 0.009765625, 0.0390625, 0.01953125, 0.0, 0.001953125, 0.001953125, 0.0, 0.001953125, 0.009765625, 0.001953125, 0.0078125, 0.0, 0.005859375, 0.005859375, 0.0078125, 0.005859375, 0.005859375, 0.005859375, 0.00390625, 0.001953125, 0.0, 0.0, 0.00390625, 0.0, 0.0078125, 0.001953125, 0.00390625, 0.001953125, 0.0, 0.00390625, 0.001953125, 0.00390625, 0.0, 0.0, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 2000, "train/token_ce": 6.382958302274346, "train/token_acc1": 0.02518717510974966, "dev/token_ce": 6.611551518552005, "dev/token_acc1": 0.011555989913176745, "train/acc_by_position": [0.06640625, 0.056640625, 0.068359375, 0.068359375, 0.078125, 0.048828125, 0.068359375, 0.017578125, 0.021484375, 0.046875, 0.06640625, 0.05859375, 0.08984375, 0.060546875, 0.080078125, 0.05859375, 0.01171875, 0.009765625, 0.009765625, 0.015625, 0.025390625, 0.009765625, 0.013671875, 0.017578125, 0.013671875, 0.009765625, 0.013671875, 0.01171875, 0.0078125, 0.01171875, 0.0078125, 0.005859375, 0.005859375, 0.00390625, 0.005859375, 0.005859375, 0.001953125, 0.001953125, 0.00390625, 0.001953125, 0.001953125, 0.001953125, 0.005859375, 0.001953125, 0.001953125, 0.0, 0.0078125, 0.005859375], "dev/acc_by_position": [0.0234375, 0.015625, 0.03125, 0.021484375, 0.060546875, 0.021484375, 0.041015625, 0.01171875, 0.005859375, 0.013671875, 0.029296875, 0.017578125, 0.04296875, 0.03515625, 0.052734375, 0.03125, 0.005859375, 0.0, 0.005859375, 0.005859375, 0.0078125, 0.0078125, 0.01171875, 0.009765625, 0.009765625, 0.001953125, 0.00390625, 0.00390625, 0.0, 0.00390625, 0.00390625, 0.00390625, 0.001953125, 0.0, 0.0, 0.0, 0.001953125, 0.0, 0.001953125, 0.0, 0.0, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.001953125, 0.0, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 3000, "train/token_ce": 6.023266343399882, "train/token_acc1": 0.07088216334886965, "dev/token_ce": 6.751016986556351, "dev/token_acc1": 0.013549805033107987, "train/acc_by_position": [0.203125, 0.1875, 0.205078125, 0.212890625, 0.244140625, 0.1484375, 0.216796875, 0.1328125, 0.099609375, 0.0390625, 0.115234375, 0.130859375, 0.169921875, 0.154296875, 0.150390625, 0.1640625, 0.044921875, 0.05078125, 0.0390625, 0.064453125, 0.05859375, 0.04296875, 0.044921875, 0.0546875, 0.05859375, 0.0390625, 0.0703125, 0.05078125, 0.033203125, 0.02734375, 0.04296875, 0.02734375, 0.01171875, 0.005859375, 0.0078125, 0.005859375, 0.009765625, 0.0, 0.001953125, 0.00390625, 0.0, 0.0, 0.0078125, 0.001953125, 0.005859375, 0.001953125, 0.00390625, 0.009765625], "dev/acc_by_position": [0.03515625, 0.0234375, 0.029296875, 0.02734375, 0.068359375, 0.033203125, 0.0546875, 0.01953125, 0.015625, 0.01953125, 0.025390625, 0.017578125, 0.041015625, 0.048828125, 0.029296875, 0.037109375, 0.001953125, 0.005859375, 0.0, 0.00390625, 0.009765625, 0.00390625, 0.0078125, 0.00390625, 0.00390625, 0.005859375, 0.0, 0.0078125, 0.00390625, 0.0078125, 0.00390625, 0.01171875, 0.005859375, 0.001953125, 0.0, 0.00390625, 0.00390625, 0.0, 0.001953125, 0.001953125, 0.001953125, 0.001953125, 0.005859375, 0.00390625, 0.001953125, 0.0, 0.00390625, 0.00390625], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 4000, "train/token_ce": 5.658540292643011, "train/token_acc1": 0.12955729529858218, "dev/token_ce": 7.045392878353596, "dev/token_acc1": 0.012898763347038766, "train/acc_by_position": [0.33984375, 0.345703125, 0.353515625, 0.3515625, 0.423828125, 0.28515625, 0.376953125, 0.34375, 0.302734375, 0.22265625, 0.32421875, 0.2421875, 0.275390625, 0.2578125, 0.259765625, 0.244140625, 0.09765625, 0.072265625, 0.0625, 0.08203125, 0.078125, 0.064453125, 0.068359375, 0.095703125, 0.07421875, 0.0703125, 0.091796875, 0.068359375, 0.0625, 0.060546875, 0.056640625, 0.044921875, 0.017578125, 0.01171875, 0.00390625, 0.005859375, 0.00390625, 0.0, 0.005859375, 0.005859375, 0.009765625, 0.0078125, 0.009765625, 0.005859375, 0.015625, 0.00390625, 0.001953125, 0.009765625], "dev/acc_by_position": [0.025390625, 0.0078125, 0.017578125, 0.009765625, 0.052734375, 0.033203125, 0.046875, 0.0234375, 0.03125, 0.02734375, 0.029296875, 0.037109375, 0.048828125, 0.03125, 0.0390625, 0.052734375, 0.00390625, 0.0078125, 0.001953125, 0.009765625, 0.0078125, 0.0, 0.0078125, 0.001953125, 0.005859375, 0.005859375, 0.001953125, 0.001953125, 0.005859375, 0.001953125, 0.005859375, 0.001953125, 0.009765625, 0.0, 0.0, 0.0, 0.001953125, 0.001953125, 0.0, 0.00390625, 0.001953125, 0.0, 0.00390625, 0.001953125, 0.0, 0.001953125, 0.005859375, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 5000, "train/token_ce": 5.982596336398274, "train/token_acc1": 0.15323893656022847, "dev/token_ce": 7.77852673176676, "dev/token_acc1": 0.011678060214762809, "train/acc_by_position": [0.435546875, 0.451171875, 0.478515625, 0.48046875, 0.50390625, 0.4609375, 0.470703125, 0.392578125, 0.3515625, 0.099609375, 0.11328125, 0.1796875, 0.369140625, 0.35546875, 0.328125, 0.349609375, 0.11328125, 0.078125, 0.05859375, 0.107421875, 0.087890625, 0.078125, 0.080078125, 0.08984375, 0.0859375, 0.1015625, 0.1328125, 0.1171875, 0.0625, 0.080078125, 0.080078125, 0.07421875, 0.013671875, 0.00390625, 0.00390625, 0.013671875, 0.009765625, 0.001953125, 0.00390625, 0.009765625, 0.00390625, 0.009765625, 0.00390625, 0.0078125, 0.001953125, 0.00390625, 0.005859375, 0.009765625], "dev/acc_by_position": [0.025390625, 0.025390625, 0.02734375, 0.005859375, 0.05078125, 0.037109375, 0.056640625, 0.021484375, 0.029296875, 0.015625, 0.005859375, 0.009765625, 0.0390625, 0.02734375, 0.02734375, 0.037109375, 0.001953125, 0.001953125, 0.0, 0.0078125, 0.005859375, 0.005859375, 0.0078125, 0.001953125, 0.0078125, 0.001953125, 0.001953125, 0.0078125, 0.00390625, 0.00390625, 0.005859375, 0.00390625, 0.005859375, 0.001953125, 0.00390625, 0.001953125, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.0078125, 0.00390625, 0.001953125, 0.001953125, 0.00390625, 0.001953125, 0.00390625, 0.00390625], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 6000, "train/token_ce": 6.721743047237396, "train/token_acc1": 0.1538085983011115, "dev/token_ce": 8.688436035998166, "dev/token_acc1": 0.009806315378227737, "train/acc_by_position": [0.49609375, 0.490234375, 0.515625, 0.52734375, 0.5703125, 0.494140625, 0.39453125, 0.185546875, 0.046875, 0.015625, 0.16015625, 0.34765625, 0.4140625, 0.408203125, 0.36328125, 0.353515625, 0.109375, 0.08984375, 0.076171875, 0.11328125, 0.10546875, 0.080078125, 0.107421875, 0.09375, 0.1015625, 0.115234375, 0.111328125, 0.123046875, 0.072265625, 0.072265625, 0.064453125, 0.068359375, 0.013671875, 0.009765625, 0.001953125, 0.017578125, 0.005859375, 0.001953125, 0.005859375, 0.001953125, 0.00390625, 0.0078125, 0.005859375, 0.0078125, 0.001953125, 0.001953125, 0.00390625, 0.00390625], "dev/acc_by_position": [0.0234375, 0.021484375, 0.02734375, 0.005859375, 0.044921875, 0.03125, 0.02734375, 0.013671875, 0.001953125, 0.00390625, 0.015625, 0.021484375, 0.02734375, 0.037109375, 0.037109375, 0.03515625, 0.001953125, 0.001953125, 0.005859375, 0.00390625, 0.0078125, 0.00390625, 0.0078125, 0.005859375, 0.00390625, 0.0, 0.0, 0.00390625, 0.00390625, 0.0, 0.005859375, 0.0078125, 0.001953125, 0.001953125, 0.005859375, 0.001953125, 0.0, 0.0, 0.001953125, 0.001953125, 0.005859375, 0.001953125, 0.0, 0.0, 0.0, 0.001953125, 0.00390625, 0.001953125], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 7000, "train/token_ce": 6.3942762785591185, "train/token_acc1": 0.2008870499994373, "dev/token_ce": 8.886069516651332, "dev/token_acc1": 0.011800130505434936, "train/acc_by_position": [0.505859375, 0.529296875, 0.5625, 0.58984375, 0.59375, 0.529296875, 0.5703125, 0.54296875, 0.46484375, 0.310546875, 0.51171875, 0.447265625, 0.455078125, 0.439453125, 0.42578125, 0.40625, 0.1171875, 0.087890625, 0.08203125, 0.1015625, 0.095703125, 0.083984375, 0.11328125, 0.123046875, 0.095703125, 0.119140625, 0.14453125, 0.12109375, 0.076171875, 0.078125, 0.09375, 0.0859375, 0.001953125, 0.01171875, 0.00390625, 0.01171875, 0.017578125, 0.00390625, 0.005859375, 0.0078125, 0.01171875, 0.0078125, 0.013671875, 0.009765625, 0.01171875, 0.00390625, 0.005859375, 0.009765625], "dev/acc_by_position": [0.03125, 0.021484375, 0.0234375, 0.015625, 0.048828125, 0.0234375, 0.05078125, 0.017578125, 0.009765625, 0.013671875, 0.01953125, 0.021484375, 0.04296875, 0.0234375, 0.0390625, 0.046875, 0.005859375, 0.001953125, 0.001953125, 0.0078125, 0.0078125, 0.0078125, 0.015625, 0.001953125, 0.001953125, 0.0, 0.001953125, 0.005859375, 0.0, 0.001953125, 0.0078125, 0.009765625, 0.00390625, 0.005859375, 0.0, 0.0, 0.001953125, 0.00390625, 0.0, 0.001953125, 0.00390625, 0.0, 0.001953125, 0.001953125, 0.0, 0.0078125, 0.00390625, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 8000, "train/token_ce": 6.68649094318971, "train/token_acc1": 0.21459961509390268, "dev/token_ce": 9.359243780374527, "dev/token_acc1": 0.011149088841193588, "train/acc_by_position": [0.533203125, 0.564453125, 0.564453125, 0.6328125, 0.623046875, 0.578125, 0.591796875, 0.556640625, 0.513671875, 0.3515625, 0.52734375, 0.423828125, 0.5, 0.494140625, 0.443359375, 0.443359375, 0.171875, 0.09765625, 0.08203125, 0.12109375, 0.1171875, 0.09375, 0.119140625, 0.119140625, 0.10546875, 0.109375, 0.166015625, 0.138671875, 0.076171875, 0.111328125, 0.09765625, 0.095703125, 0.01953125, 0.0078125, 0.0078125, 0.015625, 0.00390625, 0.0, 0.00390625, 0.0078125, 0.009765625, 0.013671875, 0.017578125, 0.01171875, 0.00390625, 0.001953125, 0.00390625, 0.0078125], "dev/acc_by_position": [0.025390625, 0.015625, 0.03515625, 0.013671875, 0.041015625, 0.029296875, 0.033203125, 0.017578125, 0.03125, 0.017578125, 0.017578125, 0.017578125, 0.044921875, 0.025390625, 0.025390625, 0.037109375, 0.001953125, 0.001953125, 0.0, 0.00390625, 0.005859375, 0.005859375, 0.015625, 0.005859375, 0.009765625, 0.001953125, 0.0, 0.00390625, 0.0, 0.0, 0.005859375, 0.0078125, 0.01171875, 0.001953125, 0.005859375, 0.001953125, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.0, 0.005859375, 0.001953125, 0.0, 0.0, 0.001953125, 0.0, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 9000, "train/token_ce": 6.858140524476767, "train/token_acc1": 0.2294515041721752, "dev/token_ce": 9.725704299286008, "dev/token_acc1": 0.011393229473469546, "train/acc_by_position": [0.5390625, 0.5625, 0.59375, 0.646484375, 0.6484375, 0.603515625, 0.634765625, 0.626953125, 0.599609375, 0.552734375, 0.58984375, 0.529296875, 0.53125, 0.50390625, 0.46875, 0.447265625, 0.126953125, 0.09375, 0.08984375, 0.13671875, 0.1171875, 0.087890625, 0.125, 0.12109375, 0.08203125, 0.111328125, 0.16796875, 0.125, 0.087890625, 0.09375, 0.08203125, 0.109375, 0.013671875, 0.013671875, 0.013671875, 0.01953125, 0.013671875, 0.0078125, 0.00390625, 0.013671875, 0.009765625, 0.015625, 0.021484375, 0.0078125, 0.0078125, 0.001953125, 0.001953125, 0.01171875], "dev/acc_by_position": [0.029296875, 0.017578125, 0.017578125, 0.013671875, 0.052734375, 0.02734375, 0.052734375, 0.025390625, 0.01171875, 0.03125, 0.017578125, 0.02734375, 0.04296875, 0.021484375, 0.03125, 0.03515625, 0.001953125, 0.0, 0.00390625, 0.001953125, 0.001953125, 0.001953125, 0.009765625, 0.001953125, 0.001953125, 0.00390625, 0.009765625, 0.009765625, 0.00390625, 0.00390625, 0.001953125, 0.005859375, 0.00390625, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.0, 0.00390625, 0.005859375, 0.00390625, 0.0, 0.0, 0.001953125, 0.0, 0.001953125, 0.0, 0.0], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 10000, "train/token_ce": 7.249734256416559, "train/token_acc1": 0.22208659448369872, "dev/token_ce": 10.135110894218087, "dev/token_acc1": 0.011759440407331567, "train/acc_by_position": [0.5390625, 0.56640625, 0.599609375, 0.65234375, 0.689453125, 0.6328125, 0.662109375, 0.6171875, 0.556640625, 0.251953125, 0.4140625, 0.46875, 0.525390625, 0.54296875, 0.470703125, 0.462890625, 0.138671875, 0.107421875, 0.095703125, 0.115234375, 0.1328125, 0.09765625, 0.11328125, 0.1171875, 0.115234375, 0.12890625, 0.154296875, 0.150390625, 0.087890625, 0.103515625, 0.09765625, 0.08984375, 0.015625, 0.009765625, 0.001953125, 0.02734375, 0.001953125, 0.009765625, 0.005859375, 0.0234375, 0.0078125, 0.021484375, 0.01171875, 0.009765625, 0.005859375, 0.005859375, 0.001953125, 0.001953125], "dev/acc_by_position": [0.01953125, 0.015625, 0.03515625, 0.01953125, 0.0390625, 0.03125, 0.048828125, 0.02734375, 0.025390625, 0.017578125, 0.015625, 0.015625, 0.041015625, 0.02734375, 0.025390625, 0.037109375, 0.0, 0.0, 0.001953125, 0.005859375, 0.013671875, 0.00390625, 0.01171875, 0.001953125, 0.001953125, 0.0078125, 0.0, 0.005859375, 0.001953125, 0.0078125, 0.00390625, 0.0078125, 0.005859375, 0.001953125, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.009765625, 0.005859375, 0.001953125, 0.00390625, 0.001953125, 0.0, 0.0, 0.0, 0.001953125, 0.0078125], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true} {"optimizer_step": 10000, "train/token_ce": 7.249734256416559, "train/token_acc1": 0.22208659448369872, "dev/token_ce": 10.135110894218087, "dev/token_acc1": 0.011759440407331567, "train/acc_by_position": [0.5390625, 0.56640625, 0.599609375, 0.65234375, 0.689453125, 0.6328125, 0.662109375, 0.6171875, 0.556640625, 0.251953125, 0.4140625, 0.46875, 0.525390625, 0.54296875, 0.470703125, 0.462890625, 0.138671875, 0.107421875, 0.095703125, 0.115234375, 0.1328125, 0.09765625, 0.11328125, 0.1171875, 0.115234375, 0.12890625, 0.154296875, 0.150390625, 0.087890625, 0.103515625, 0.09765625, 0.08984375, 0.015625, 0.009765625, 0.001953125, 0.02734375, 0.001953125, 0.009765625, 0.005859375, 0.0234375, 0.0078125, 0.021484375, 0.01171875, 0.009765625, 0.005859375, 0.005859375, 0.001953125, 0.001953125], "dev/acc_by_position": [0.01953125, 0.015625, 0.03515625, 0.01953125, 0.0390625, 0.03125, 0.048828125, 0.02734375, 0.025390625, 0.017578125, 0.015625, 0.015625, 0.041015625, 0.02734375, 0.025390625, 0.037109375, 0.0, 0.0, 0.001953125, 0.005859375, 0.013671875, 0.00390625, 0.01171875, 0.001953125, 0.001953125, 0.0078125, 0.0, 0.005859375, 0.001953125, 0.0078125, 0.00390625, 0.0078125, 0.005859375, 0.001953125, 0.001953125, 0.001953125, 0.001953125, 0.0, 0.009765625, 0.005859375, 0.001953125, 0.00390625, 0.001953125, 0.0, 0.0, 0.0, 0.001953125, 0.0078125], "finite_loss": true, "finite_logits": true, "finite_parameters": true, "n_parameters_scanned": 877, "skipped_optimizer_steps": 0, "scaler_scale": null, "finite_grad_norm": true}