{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.6246719160104988, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "ce_loss": 23.0, "distill_loss": 1.8828125, "epoch": 0, "explain_loss": 6.125, "loss": 33.5, "ref_ce_loss": 2.078125, "stage_align_loss": 0.546875, "step": 0 }, { "ce_loss": 36.0, "distill_loss": 2.5, "epoch": 0, "explain_loss": 6.0, "loss": 47.5, "ref_ce_loss": 2.421875, "stage_align_loss": 0.53515625, "step": 0 }, { "ce_loss": 25.625, "distill_loss": 1.75, "epoch": 0, "explain_loss": 6.09375, "loss": 36.5, "ref_ce_loss": 2.578125, "stage_align_loss": 0.52734375, "step": 0 }, { "ce_loss": 26.25, "distill_loss": 1.921875, "epoch": 0, "explain_loss": 5.78125, "loss": 37.25, "ref_ce_loss": 2.890625, "stage_align_loss": 0.5390625, "step": 0 }, { "epoch": 0.004374453193350831, "grad_norm": 10022.642578125, "learning_rate": 5.247813411078718e-06, "loss": 38.69375, "step": 10 }, { "ce_loss": 18.375, "distill_loss": 1.9375, "epoch": 0.004374453193350831, "explain_loss": 5.90625, "loss": 29.375, "ref_ce_loss": 2.765625, "stage_align_loss": 0.52734375, "step": 10 }, { "ce_loss": 22.625, "distill_loss": 2.296875, "epoch": 0.004374453193350831, "explain_loss": 5.875, "loss": 34.0, "ref_ce_loss": 2.921875, "stage_align_loss": 0.53515625, "step": 10 }, { "ce_loss": 22.75, "distill_loss": 2.21875, "epoch": 0.004374453193350831, "explain_loss": 5.9375, "loss": 34.25, "ref_ce_loss": 2.703125, "stage_align_loss": 0.5546875, "step": 10 }, { "ce_loss": 21.5, "distill_loss": 1.53125, "epoch": 0.004374453193350831, "explain_loss": 5.75, "loss": 32.5, "ref_ce_loss": 2.984375, "stage_align_loss": 0.578125, "step": 10 }, { "epoch": 0.008748906386701663, "grad_norm": 366631.0625, "learning_rate": 1.1078717201166182e-05, "loss": 38.059375, "step": 20 }, { "ce_loss": 25.375, "distill_loss": 1.953125, "epoch": 0.008748906386701663, "explain_loss": 5.3125, "loss": 36.25, "ref_ce_loss": 2.984375, "stage_align_loss": 0.515625, "step": 20 }, { "ce_loss": 24.625, "distill_loss": 2.046875, "epoch": 0.008748906386701663, "explain_loss": 5.25, "loss": 35.0, "ref_ce_loss": 2.46875, "stage_align_loss": 0.54296875, "step": 20 }, { "ce_loss": 20.625, "distill_loss": 3.1875, "epoch": 0.008748906386701663, "explain_loss": 5.9375, "loss": 32.75, "ref_ce_loss": 2.359375, "stage_align_loss": 0.609375, "step": 20 }, { "ce_loss": 24.125, "distill_loss": 2.46875, "epoch": 0.008748906386701663, "explain_loss": 5.625, "loss": 35.75, "ref_ce_loss": 3.046875, "stage_align_loss": 0.5546875, "step": 20 }, { "epoch": 0.013123359580052493, "grad_norm": 3336.931640625, "learning_rate": 1.6909620991253645e-05, "loss": 37.5, "step": 30 }, { "ce_loss": 38.75, "distill_loss": 2.65625, "epoch": 0.013123359580052493, "explain_loss": 5.125, "loss": 50.0, "ref_ce_loss": 2.796875, "stage_align_loss": 0.55078125, "step": 30 }, { "ce_loss": 20.375, "distill_loss": 2.328125, "epoch": 0.013123359580052493, "explain_loss": 5.25, "loss": 31.25, "ref_ce_loss": 2.8125, "stage_align_loss": 0.5625, "step": 30 }, { "ce_loss": 33.0, "distill_loss": 2.453125, "epoch": 0.013123359580052493, "explain_loss": 5.8125, "loss": 44.25, "ref_ce_loss": 2.609375, "stage_align_loss": 0.51953125, "step": 30 }, { "ce_loss": 22.875, "distill_loss": 2.890625, "epoch": 0.013123359580052493, "explain_loss": 5.6875, "loss": 34.75, "ref_ce_loss": 2.78125, "stage_align_loss": 0.56640625, "step": 30 }, { "epoch": 0.017497812773403325, "grad_norm": 2828.241943359375, "learning_rate": 2.2740524781341106e-05, "loss": 37.6734375, "step": 40 }, { "ce_loss": 31.125, "distill_loss": 2.3125, "epoch": 0.017497812773403325, "explain_loss": 5.03125, "loss": 41.75, "ref_ce_loss": 2.796875, "stage_align_loss": 0.5625, "step": 40 }, { "ce_loss": 35.25, "distill_loss": 2.5625, "epoch": 0.017497812773403325, "explain_loss": 4.875, "loss": 46.0, "ref_ce_loss": 2.8125, "stage_align_loss": 0.49609375, "step": 40 }, { "ce_loss": 33.0, "distill_loss": 3.0, "epoch": 0.017497812773403325, "explain_loss": 5.0, "loss": 44.0, "ref_ce_loss": 2.609375, "stage_align_loss": 0.546875, "step": 40 }, { "ce_loss": 38.5, "distill_loss": 2.09375, "epoch": 0.017497812773403325, "explain_loss": 4.9375, "loss": 48.25, "ref_ce_loss": 2.328125, "stage_align_loss": 0.484375, "step": 40 }, { "epoch": 0.021872265966754154, "grad_norm": 449.33642578125, "learning_rate": 2.857142857142857e-05, "loss": 34.190625, "step": 50 }, { "ce_loss": 20.5, "distill_loss": 2.25, "epoch": 0.021872265966754154, "explain_loss": 4.5625, "loss": 30.75, "ref_ce_loss": 3.046875, "stage_align_loss": 0.53125, "step": 50 }, { "ce_loss": 29.0, "distill_loss": 2.375, "epoch": 0.021872265966754154, "explain_loss": 4.625, "loss": 38.5, "ref_ce_loss": 2.1875, "stage_align_loss": 0.478515625, "step": 50 }, { "ce_loss": 15.875, "distill_loss": 2.046875, "epoch": 0.021872265966754154, "explain_loss": 4.375, "loss": 25.125, "ref_ce_loss": 2.328125, "stage_align_loss": 0.51953125, "step": 50 }, { "ce_loss": 26.25, "distill_loss": 2.328125, "epoch": 0.021872265966754154, "explain_loss": 4.59375, "loss": 35.75, "ref_ce_loss": 1.984375, "stage_align_loss": 0.5234375, "step": 50 }, { "epoch": 0.026246719160104987, "grad_norm": 7396.8779296875, "learning_rate": 3.4402332361516035e-05, "loss": 25.878125, "step": 60 }, { "ce_loss": 11.5625, "distill_loss": 1.78125, "epoch": 0.026246719160104987, "explain_loss": 3.953125, "loss": 20.5, "ref_ce_loss": 2.625, "stage_align_loss": 0.458984375, "step": 60 }, { "ce_loss": 16.125, "distill_loss": 3.203125, "epoch": 0.026246719160104987, "explain_loss": 4.1875, "loss": 26.5, "ref_ce_loss": 2.515625, "stage_align_loss": 0.435546875, "step": 60 }, { "ce_loss": 13.1875, "distill_loss": 2.21875, "epoch": 0.026246719160104987, "explain_loss": 4.34375, "loss": 22.0, "ref_ce_loss": 1.9296875, "stage_align_loss": 0.412109375, "step": 60 }, { "ce_loss": 13.5, "distill_loss": 2.03125, "epoch": 0.026246719160104987, "explain_loss": 4.125, "loss": 22.25, "ref_ce_loss": 2.21875, "stage_align_loss": 0.41015625, "step": 60 }, { "epoch": 0.03062117235345582, "grad_norm": 14051.982421875, "learning_rate": 4.02332361516035e-05, "loss": 18.46875, "step": 70 }, { "ce_loss": 5.875, "distill_loss": 2.40625, "epoch": 0.03062117235345582, "explain_loss": 4.125, "loss": 15.0, "ref_ce_loss": 2.015625, "stage_align_loss": 0.65234375, "step": 70 }, { "ce_loss": 4.46875, "distill_loss": 2.578125, "epoch": 0.03062117235345582, "explain_loss": 4.40625, "loss": 14.0, "ref_ce_loss": 1.96875, "stage_align_loss": 0.6171875, "step": 70 }, { "ce_loss": 4.90625, "distill_loss": 2.21875, "epoch": 0.03062117235345582, "explain_loss": 4.03125, "loss": 14.0, "ref_ce_loss": 2.1875, "stage_align_loss": 0.625, "step": 70 }, { "ce_loss": 6.5625, "distill_loss": 2.40625, "epoch": 0.03062117235345582, "explain_loss": 4.0625, "loss": 15.6875, "ref_ce_loss": 1.9296875, "stage_align_loss": 0.703125, "step": 70 }, { "epoch": 0.03499562554680665, "grad_norm": 23.066892623901367, "learning_rate": 4.6064139941690965e-05, "loss": 13.5984375, "step": 80 }, { "ce_loss": 1.328125, "distill_loss": 2.21875, "epoch": 0.03499562554680665, "explain_loss": 3.828125, "loss": 9.875, "ref_ce_loss": 1.859375, "stage_align_loss": 0.62890625, "step": 80 }, { "ce_loss": 1.75, "distill_loss": 2.65625, "epoch": 0.03499562554680665, "explain_loss": 4.0, "loss": 11.0, "ref_ce_loss": 1.9765625, "stage_align_loss": 0.6484375, "step": 80 }, { "ce_loss": 2.15625, "distill_loss": 2.046875, "epoch": 0.03499562554680665, "explain_loss": 3.78125, "loss": 10.6875, "ref_ce_loss": 2.046875, "stage_align_loss": 0.64453125, "step": 80 }, { "ce_loss": 1.7734375, "distill_loss": 2.359375, "epoch": 0.03499562554680665, "explain_loss": 4.0, "loss": 10.375, "ref_ce_loss": 1.6328125, "stage_align_loss": 0.63671875, "step": 80 }, { "epoch": 0.03937007874015748, "grad_norm": 144564.34375, "learning_rate": 5.189504373177842e-05, "loss": 9.126171875, "step": 90 }, { "ce_loss": 1.25, "distill_loss": 1.96875, "epoch": 0.03937007874015748, "explain_loss": 4.0, "loss": 9.5, "ref_ce_loss": 1.78125, "stage_align_loss": 0.48046875, "step": 90 }, { "ce_loss": 0.828125, "distill_loss": 1.8046875, "epoch": 0.03937007874015748, "explain_loss": 4.09375, "loss": 9.125, "ref_ce_loss": 1.8984375, "stage_align_loss": 0.4921875, "step": 90 }, { "ce_loss": 1.875, "distill_loss": 1.71875, "epoch": 0.03937007874015748, "explain_loss": 3.46875, "loss": 9.25, "ref_ce_loss": 1.7109375, "stage_align_loss": 0.482421875, "step": 90 }, { "ce_loss": 1.140625, "distill_loss": 1.71875, "epoch": 0.03937007874015748, "explain_loss": 3.5625, "loss": 8.625, "ref_ce_loss": 1.6953125, "stage_align_loss": 0.474609375, "step": 90 }, { "epoch": 0.04374453193350831, "grad_norm": 15.442902565002441, "learning_rate": 5.77259475218659e-05, "loss": 9.25546875, "step": 100 }, { "ce_loss": 1.7109375, "distill_loss": 1.28125, "epoch": 0.04374453193350831, "explain_loss": 3.03125, "loss": 7.875, "ref_ce_loss": 1.4453125, "stage_align_loss": 0.41796875, "step": 100 }, { "ce_loss": 0.1611328125, "distill_loss": 1.5703125, "epoch": 0.04374453193350831, "explain_loss": 3.21875, "loss": 6.8125, "ref_ce_loss": 1.453125, "stage_align_loss": 0.416015625, "step": 100 }, { "ce_loss": 0.059814453125, "distill_loss": 1.5546875, "epoch": 0.04374453193350831, "explain_loss": 3.59375, "loss": 7.28125, "ref_ce_loss": 1.6484375, "stage_align_loss": 0.423828125, "step": 100 }, { "ce_loss": 1.1640625, "distill_loss": 1.453125, "epoch": 0.04374453193350831, "explain_loss": 3.03125, "loss": 7.5, "ref_ce_loss": 1.4140625, "stage_align_loss": 0.431640625, "step": 100 }, { "epoch": 0.048118985126859144, "grad_norm": 4.415060997009277, "learning_rate": 6.355685131195336e-05, "loss": 6.5921875, "step": 110 }, { "ce_loss": 1.0703125, "distill_loss": 0.98046875, "epoch": 0.048118985126859144, "explain_loss": 2.328125, "loss": 5.875, "ref_ce_loss": 1.1328125, "stage_align_loss": 0.359375, "step": 110 }, { "ce_loss": 1.5859375, "distill_loss": 0.921875, "epoch": 0.048118985126859144, "explain_loss": 2.8125, "loss": 6.75, "ref_ce_loss": 1.015625, "stage_align_loss": 0.404296875, "step": 110 }, { "ce_loss": 0.6640625, "distill_loss": 0.953125, "epoch": 0.048118985126859144, "explain_loss": 2.546875, "loss": 5.78125, "ref_ce_loss": 1.2265625, "stage_align_loss": 0.392578125, "step": 110 }, { "ce_loss": 1.0, "distill_loss": 0.84375, "epoch": 0.048118985126859144, "explain_loss": 2.28125, "loss": 5.875, "ref_ce_loss": 1.359375, "stage_align_loss": 0.404296875, "step": 110 }, { "epoch": 0.05249343832020997, "grad_norm": 3.83837628364563, "learning_rate": 6.938775510204082e-05, "loss": 5.13125, "step": 120 }, { "ce_loss": 1.5546875, "distill_loss": 0.890625, "epoch": 0.05249343832020997, "explain_loss": 2.09375, "loss": 5.53125, "ref_ce_loss": 0.59375, "stage_align_loss": 0.40234375, "step": 120 }, { "ce_loss": 1.9140625, "distill_loss": 0.734375, "epoch": 0.05249343832020997, "explain_loss": 2.046875, "loss": 5.9375, "ref_ce_loss": 0.8359375, "stage_align_loss": 0.41015625, "step": 120 }, { "ce_loss": 1.015625, "distill_loss": 0.59375, "epoch": 0.05249343832020997, "explain_loss": 1.984375, "loss": 4.8125, "ref_ce_loss": 0.80859375, "stage_align_loss": 0.42578125, "step": 120 }, { "ce_loss": 1.1015625, "distill_loss": 0.8046875, "epoch": 0.05249343832020997, "explain_loss": 2.0625, "loss": 5.1875, "ref_ce_loss": 0.78125, "stage_align_loss": 0.42578125, "step": 120 }, { "epoch": 0.0568678915135608, "grad_norm": 5.532820701599121, "learning_rate": 7.521865889212827e-05, "loss": 4.3158203125, "step": 130 }, { "ce_loss": 1.125, "distill_loss": 0.421875, "epoch": 0.0568678915135608, "explain_loss": 1.71875, "loss": 4.5625, "ref_ce_loss": 0.921875, "stage_align_loss": 0.357421875, "step": 130 }, { "ce_loss": 2.859375, "distill_loss": 0.625, "epoch": 0.0568678915135608, "explain_loss": 1.828125, "loss": 6.3125, "ref_ce_loss": 0.65625, "stage_align_loss": 0.353515625, "step": 130 }, { "ce_loss": 0.7578125, "distill_loss": 0.71875, "epoch": 0.0568678915135608, "explain_loss": 1.7421875, "loss": 4.1875, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.359375, "step": 130 }, { "ce_loss": 0.044677734375, "distill_loss": 0.71875, "epoch": 0.0568678915135608, "explain_loss": 1.8515625, "loss": 3.71875, "ref_ce_loss": 0.73828125, "stage_align_loss": 0.375, "step": 130 }, { "epoch": 0.06124234470691164, "grad_norm": 2.1909165382385254, "learning_rate": 8.104956268221575e-05, "loss": 3.9537109375, "step": 140 }, { "ce_loss": 0.9453125, "distill_loss": 0.4921875, "epoch": 0.06124234470691164, "explain_loss": 1.53125, "loss": 3.796875, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.267578125, "step": 140 }, { "ce_loss": 0.078125, "distill_loss": 0.3359375, "epoch": 0.06124234470691164, "explain_loss": 1.46875, "loss": 2.59375, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.2578125, "step": 140 }, { "ce_loss": 0.76953125, "distill_loss": 0.498046875, "epoch": 0.06124234470691164, "explain_loss": 1.3203125, "loss": 3.625, "ref_ce_loss": 0.7578125, "stage_align_loss": 0.2890625, "step": 140 }, { "ce_loss": 0.64453125, "distill_loss": 0.53125, "epoch": 0.06124234470691164, "explain_loss": 1.4375, "loss": 3.484375, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.259765625, "step": 140 }, { "epoch": 0.06561679790026247, "grad_norm": 3.1651899814605713, "learning_rate": 8.688046647230322e-05, "loss": 3.537890625, "step": 150 }, { "ce_loss": 0.09130859375, "distill_loss": 0.50390625, "epoch": 0.06561679790026247, "explain_loss": 1.7578125, "loss": 3.6875, "ref_ce_loss": 1.1171875, "stage_align_loss": 0.2099609375, "step": 150 }, { "ce_loss": 0.359375, "distill_loss": 0.41796875, "epoch": 0.06561679790026247, "explain_loss": 1.703125, "loss": 3.40625, "ref_ce_loss": 0.7265625, "stage_align_loss": 0.2001953125, "step": 150 }, { "ce_loss": 0.482421875, "distill_loss": 0.4375, "epoch": 0.06561679790026247, "explain_loss": 1.46875, "loss": 2.96875, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.2099609375, "step": 150 }, { "ce_loss": 0.68359375, "distill_loss": 0.41796875, "epoch": 0.06561679790026247, "explain_loss": 1.7421875, "loss": 3.8125, "ref_ce_loss": 0.7265625, "stage_align_loss": 0.232421875, "step": 150 }, { "epoch": 0.0699912510936133, "grad_norm": 1.7378454208374023, "learning_rate": 9.271137026239067e-05, "loss": 3.356640625, "step": 160 }, { "ce_loss": 0.2734375, "distill_loss": 0.4453125, "epoch": 0.0699912510936133, "explain_loss": 1.671875, "loss": 3.21875, "ref_ce_loss": 0.65234375, "stage_align_loss": 0.1787109375, "step": 160 }, { "ce_loss": 1.453125, "distill_loss": 0.37109375, "epoch": 0.0699912510936133, "explain_loss": 1.6875, "loss": 4.21875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.1640625, "step": 160 }, { "ce_loss": 0.259765625, "distill_loss": 0.3359375, "epoch": 0.0699912510936133, "explain_loss": 1.421875, "loss": 2.84375, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.1953125, "step": 160 }, { "ce_loss": 0.40625, "distill_loss": 0.451171875, "epoch": 0.0699912510936133, "explain_loss": 1.53125, "loss": 3.375, "ref_ce_loss": 0.796875, "stage_align_loss": 0.1943359375, "step": 160 }, { "epoch": 0.07436570428696412, "grad_norm": 3.130913019180298, "learning_rate": 9.854227405247813e-05, "loss": 3.051171875, "step": 170 }, { "ce_loss": 0.052001953125, "distill_loss": 0.16796875, "epoch": 0.07436570428696412, "explain_loss": 1.578125, "loss": 2.53125, "ref_ce_loss": 0.625, "stage_align_loss": 0.11669921875, "step": 170 }, { "ce_loss": 0.1494140625, "distill_loss": 0.21875, "epoch": 0.07436570428696412, "explain_loss": 1.7578125, "loss": 2.65625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.1259765625, "step": 170 }, { "ce_loss": 2.234375, "distill_loss": 0.353515625, "epoch": 0.07436570428696412, "explain_loss": 2.3125, "loss": 5.59375, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.138671875, "step": 170 }, { "ce_loss": 1.7265625, "distill_loss": 0.306640625, "epoch": 0.07436570428696412, "explain_loss": 1.078125, "loss": 3.6875, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.146484375, "step": 170 }, { "epoch": 0.07874015748031496, "grad_norm": 3.472876787185669, "learning_rate": 0.0001043731778425656, "loss": 2.887890625, "step": 180 }, { "ce_loss": 0.3828125, "distill_loss": 0.3515625, "epoch": 0.07874015748031496, "explain_loss": 1.453125, "loss": 2.875, "ref_ce_loss": 0.5625, "stage_align_loss": 0.123046875, "step": 180 }, { "ce_loss": 0.65234375, "distill_loss": 0.287109375, "epoch": 0.07874015748031496, "explain_loss": 1.484375, "loss": 3.125, "ref_ce_loss": 0.53125, "stage_align_loss": 0.16796875, "step": 180 }, { "ce_loss": 0.6953125, "distill_loss": 0.322265625, "epoch": 0.07874015748031496, "explain_loss": 1.890625, "loss": 3.671875, "ref_ce_loss": 0.625, "stage_align_loss": 0.1396484375, "step": 180 }, { "ce_loss": 0.5390625, "distill_loss": 0.240234375, "epoch": 0.07874015748031496, "explain_loss": 1.25, "loss": 2.78125, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.142578125, "step": 180 }, { "epoch": 0.0831146106736658, "grad_norm": 1.9977803230285645, "learning_rate": 0.00011020408163265306, "loss": 2.92080078125, "step": 190 }, { "ce_loss": 0.033447265625, "distill_loss": 0.2890625, "epoch": 0.0831146106736658, "explain_loss": 1.546875, "loss": 2.671875, "ref_ce_loss": 0.7265625, "stage_align_loss": 0.078125, "step": 190 }, { "ce_loss": 0.302734375, "distill_loss": 0.1611328125, "epoch": 0.0831146106736658, "explain_loss": 1.2109375, "loss": 2.1875, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.06689453125, "step": 190 }, { "ce_loss": 0.9765625, "distill_loss": 0.29296875, "epoch": 0.0831146106736658, "explain_loss": 1.2421875, "loss": 3.15625, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.07958984375, "step": 190 }, { "ce_loss": 0.07763671875, "distill_loss": 0.18359375, "epoch": 0.0831146106736658, "explain_loss": 1.3515625, "loss": 2.203125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.06640625, "step": 190 }, { "epoch": 0.08748906386701662, "grad_norm": 1.4570120573043823, "learning_rate": 0.00011603498542274054, "loss": 2.67373046875, "step": 200 }, { "ce_loss": 0.208984375, "distill_loss": 0.240234375, "epoch": 0.08748906386701662, "explain_loss": 1.2734375, "loss": 2.46875, "ref_ce_loss": 0.65625, "stage_align_loss": 0.080078125, "step": 200 }, { "ce_loss": 0.2470703125, "distill_loss": 0.29296875, "epoch": 0.08748906386701662, "explain_loss": 1.4453125, "loss": 2.671875, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.0859375, "step": 200 }, { "ce_loss": 0.67578125, "distill_loss": 0.146484375, "epoch": 0.08748906386701662, "explain_loss": 1.71875, "loss": 3.25, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.06689453125, "step": 200 }, { "ce_loss": 0.75, "distill_loss": 0.185546875, "epoch": 0.08748906386701662, "explain_loss": 1.453125, "loss": 3.078125, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.06298828125, "step": 200 }, { "epoch": 0.09186351706036745, "grad_norm": 6.860328674316406, "learning_rate": 0.000121865889212828, "loss": 2.7029296875, "step": 210 }, { "ce_loss": 1.96875, "distill_loss": 0.1572265625, "epoch": 0.09186351706036745, "explain_loss": 1.359375, "loss": 4.25, "ref_ce_loss": 0.70703125, "stage_align_loss": 0.08056640625, "step": 210 }, { "ce_loss": 1.3671875, "distill_loss": 0.1669921875, "epoch": 0.09186351706036745, "explain_loss": 1.1953125, "loss": 3.15625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0791015625, "step": 210 }, { "ce_loss": 0.859375, "distill_loss": 0.1533203125, "epoch": 0.09186351706036745, "explain_loss": 1.265625, "loss": 2.84375, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.0771484375, "step": 210 }, { "ce_loss": 0.8359375, "distill_loss": 0.228515625, "epoch": 0.09186351706036745, "explain_loss": 1.1875, "loss": 2.890625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.0751953125, "step": 210 }, { "epoch": 0.09623797025371829, "grad_norm": 5.691582202911377, "learning_rate": 0.00012769679300291545, "loss": 2.76171875, "step": 220 }, { "ce_loss": 1.5234375, "distill_loss": 0.166015625, "epoch": 0.09623797025371829, "explain_loss": 1.34375, "loss": 3.8125, "ref_ce_loss": 0.68359375, "stage_align_loss": 0.087890625, "step": 220 }, { "ce_loss": 0.41015625, "distill_loss": 0.203125, "epoch": 0.09623797025371829, "explain_loss": 1.3984375, "loss": 2.71875, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.0830078125, "step": 220 }, { "ce_loss": 1.28125, "distill_loss": 0.1640625, "epoch": 0.09623797025371829, "explain_loss": 1.5234375, "loss": 3.90625, "ref_ce_loss": 0.83203125, "stage_align_loss": 0.0869140625, "step": 220 }, { "ce_loss": 0.625, "distill_loss": 0.1669921875, "epoch": 0.09623797025371829, "explain_loss": 1.2890625, "loss": 2.625, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0712890625, "step": 220 }, { "epoch": 0.10061242344706911, "grad_norm": 5.802473068237305, "learning_rate": 0.00013352769679300292, "loss": 2.93662109375, "step": 230 }, { "ce_loss": 0.7265625, "distill_loss": 0.2255859375, "epoch": 0.10061242344706911, "explain_loss": 1.59375, "loss": 3.15625, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0732421875, "step": 230 }, { "ce_loss": 0.6171875, "distill_loss": 0.16796875, "epoch": 0.10061242344706911, "explain_loss": 1.3984375, "loss": 3.015625, "ref_ce_loss": 0.74609375, "stage_align_loss": 0.0830078125, "step": 230 }, { "ce_loss": 0.37890625, "distill_loss": 0.212890625, "epoch": 0.10061242344706911, "explain_loss": 1.0234375, "loss": 2.3125, "ref_ce_loss": 0.59375, "stage_align_loss": 0.09716796875, "step": 230 }, { "ce_loss": 0.134765625, "distill_loss": 0.19921875, "epoch": 0.10061242344706911, "explain_loss": 1.796875, "loss": 2.734375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.07275390625, "step": 230 }, { "epoch": 0.10498687664041995, "grad_norm": 2.501568555831909, "learning_rate": 0.00013935860058309038, "loss": 3.2146484375, "step": 240 }, { "ce_loss": 0.45703125, "distill_loss": 0.1953125, "epoch": 0.10498687664041995, "explain_loss": 1.2421875, "loss": 2.5, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.09326171875, "step": 240 }, { "ce_loss": 0.265625, "distill_loss": 0.181640625, "epoch": 0.10498687664041995, "explain_loss": 1.25, "loss": 2.59375, "ref_ce_loss": 0.81640625, "stage_align_loss": 0.07861328125, "step": 240 }, { "ce_loss": 0.09765625, "distill_loss": 0.16015625, "epoch": 0.10498687664041995, "explain_loss": 1.0859375, "loss": 1.8046875, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0673828125, "step": 240 }, { "ce_loss": 0.357421875, "distill_loss": 0.1650390625, "epoch": 0.10498687664041995, "explain_loss": 1.28125, "loss": 2.3125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.0693359375, "step": 240 }, { "epoch": 0.10936132983377078, "grad_norm": 2.214406728744507, "learning_rate": 0.00014518950437317785, "loss": 2.67490234375, "step": 250 }, { "ce_loss": 0.5, "distill_loss": 0.150390625, "epoch": 0.10936132983377078, "explain_loss": 1.1015625, "loss": 2.21875, "ref_ce_loss": 0.421875, "stage_align_loss": 0.04052734375, "step": 250 }, { "ce_loss": 0.416015625, "distill_loss": 0.162109375, "epoch": 0.10936132983377078, "explain_loss": 1.2734375, "loss": 2.53125, "ref_ce_loss": 0.640625, "stage_align_loss": 0.036376953125, "step": 250 }, { "ce_loss": 0.84765625, "distill_loss": 0.146484375, "epoch": 0.10936132983377078, "explain_loss": 1.6015625, "loss": 3.28125, "ref_ce_loss": 0.640625, "stage_align_loss": 0.043212890625, "step": 250 }, { "ce_loss": 0.380859375, "distill_loss": 0.181640625, "epoch": 0.10936132983377078, "explain_loss": 1.6796875, "loss": 2.734375, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.037109375, "step": 250 }, { "epoch": 0.1137357830271216, "grad_norm": 2.339704751968384, "learning_rate": 0.0001510204081632653, "loss": 2.51435546875, "step": 260 }, { "ce_loss": 1.4921875, "distill_loss": 0.134765625, "epoch": 0.1137357830271216, "explain_loss": 1.265625, "loss": 3.703125, "ref_ce_loss": 0.7734375, "stage_align_loss": 0.038818359375, "step": 260 }, { "ce_loss": 0.376953125, "distill_loss": 0.1640625, "epoch": 0.1137357830271216, "explain_loss": 1.109375, "loss": 2.171875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.046875, "step": 260 }, { "ce_loss": 0.43359375, "distill_loss": 0.12890625, "epoch": 0.1137357830271216, "explain_loss": 1.484375, "loss": 2.625, "ref_ce_loss": 0.53125, "stage_align_loss": 0.037353515625, "step": 260 }, { "ce_loss": 1.4609375, "distill_loss": 0.12890625, "epoch": 0.1137357830271216, "explain_loss": 1.390625, "loss": 3.65625, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.03759765625, "step": 260 }, { "epoch": 0.11811023622047244, "grad_norm": 1.396722674369812, "learning_rate": 0.00015685131195335278, "loss": 2.54287109375, "step": 270 }, { "ce_loss": 0.171875, "distill_loss": 0.1748046875, "epoch": 0.11811023622047244, "explain_loss": 1.453125, "loss": 2.3125, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.040771484375, "step": 270 }, { "ce_loss": 0.07373046875, "distill_loss": 0.1767578125, "epoch": 0.11811023622047244, "explain_loss": 1.25, "loss": 1.9375, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.040771484375, "step": 270 }, { "ce_loss": 0.1328125, "distill_loss": 0.19921875, "epoch": 0.11811023622047244, "explain_loss": 1.453125, "loss": 2.5625, "ref_ce_loss": 0.71875, "stage_align_loss": 0.051025390625, "step": 270 }, { "ce_loss": 0.216796875, "distill_loss": 0.146484375, "epoch": 0.11811023622047244, "explain_loss": 1.171875, "loss": 2.21875, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.040283203125, "step": 270 }, { "epoch": 0.12248468941382328, "grad_norm": 1.339150309562683, "learning_rate": 0.00016268221574344024, "loss": 2.63623046875, "step": 280 }, { "ce_loss": 2.390625, "distill_loss": 0.19140625, "epoch": 0.12248468941382328, "explain_loss": 1.3125, "loss": 4.5, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.0361328125, "step": 280 }, { "ce_loss": 3.15625, "distill_loss": 0.185546875, "epoch": 0.12248468941382328, "explain_loss": 1.21875, "loss": 5.125, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.039794921875, "step": 280 }, { "ce_loss": 0.79296875, "distill_loss": 0.181640625, "epoch": 0.12248468941382328, "explain_loss": 1.296875, "loss": 2.984375, "ref_ce_loss": 0.66015625, "stage_align_loss": 0.04345703125, "step": 280 }, { "ce_loss": 9.625, "distill_loss": 0.21875, "epoch": 0.12248468941382328, "explain_loss": 1.4140625, "loss": 12.125, "ref_ce_loss": 0.7578125, "stage_align_loss": 0.0380859375, "step": 280 }, { "epoch": 0.1268591426071741, "grad_norm": 1.4034538269042969, "learning_rate": 0.0001685131195335277, "loss": 3.00224609375, "step": 290 }, { "ce_loss": 0.20703125, "distill_loss": 0.1796875, "epoch": 0.1268591426071741, "explain_loss": 1.6015625, "loss": 2.453125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.0322265625, "step": 290 }, { "ce_loss": 0.36328125, "distill_loss": 0.171875, "epoch": 0.1268591426071741, "explain_loss": 1.640625, "loss": 2.765625, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.0283203125, "step": 290 }, { "ce_loss": 0.1376953125, "distill_loss": 0.20703125, "epoch": 0.1268591426071741, "explain_loss": 1.640625, "loss": 2.40625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.0294189453125, "step": 290 }, { "ce_loss": 0.5859375, "distill_loss": 0.166015625, "epoch": 0.1268591426071741, "explain_loss": 1.03125, "loss": 2.5, "ref_ce_loss": 0.6796875, "stage_align_loss": 0.03125, "step": 290 }, { "epoch": 0.13123359580052493, "grad_norm": 2.6279213428497314, "learning_rate": 0.00017434402332361517, "loss": 2.474609375, "step": 300 }, { "ce_loss": 0.53125, "distill_loss": 0.140625, "epoch": 0.13123359580052493, "explain_loss": 1.6484375, "loss": 3.0625, "ref_ce_loss": 0.71484375, "stage_align_loss": 0.018310546875, "step": 300 }, { "ce_loss": 0.054931640625, "distill_loss": 0.146484375, "epoch": 0.13123359580052493, "explain_loss": 1.0703125, "loss": 1.6875, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.0230712890625, "step": 300 }, { "ce_loss": 0.1904296875, "distill_loss": 0.130859375, "epoch": 0.13123359580052493, "explain_loss": 1.09375, "loss": 1.984375, "ref_ce_loss": 0.546875, "stage_align_loss": 0.02197265625, "step": 300 }, { "ce_loss": 0.07763671875, "distill_loss": 0.1484375, "epoch": 0.13123359580052493, "explain_loss": 1.1328125, "loss": 1.8515625, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0225830078125, "step": 300 }, { "epoch": 0.13560804899387577, "grad_norm": 1.2991151809692383, "learning_rate": 0.00018017492711370264, "loss": 2.44521484375, "step": 310 }, { "ce_loss": 0.279296875, "distill_loss": 0.126953125, "epoch": 0.13560804899387577, "explain_loss": 1.1796875, "loss": 1.921875, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.0185546875, "step": 310 }, { "ce_loss": 0.66796875, "distill_loss": 0.1220703125, "epoch": 0.13560804899387577, "explain_loss": 1.6328125, "loss": 2.953125, "ref_ce_loss": 0.515625, "stage_align_loss": 0.016845703125, "step": 310 }, { "ce_loss": 1.2265625, "distill_loss": 0.130859375, "epoch": 0.13560804899387577, "explain_loss": 1.0078125, "loss": 3.21875, "ref_ce_loss": 0.8125, "stage_align_loss": 0.0255126953125, "step": 310 }, { "ce_loss": 1.09375, "distill_loss": 0.12890625, "epoch": 0.13560804899387577, "explain_loss": 1.171875, "loss": 2.96875, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.0225830078125, "step": 310 }, { "epoch": 0.1399825021872266, "grad_norm": 1.059230923652649, "learning_rate": 0.0001860058309037901, "loss": 2.4361328125, "step": 320 }, { "ce_loss": 0.443359375, "distill_loss": 0.1083984375, "epoch": 0.1399825021872266, "explain_loss": 1.6640625, "loss": 2.84375, "ref_ce_loss": 0.609375, "stage_align_loss": 0.0260009765625, "step": 320 }, { "ce_loss": 0.244140625, "distill_loss": 0.109375, "epoch": 0.1399825021872266, "explain_loss": 1.203125, "loss": 2.0, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.021484375, "step": 320 }, { "ce_loss": 0.2431640625, "distill_loss": 0.138671875, "epoch": 0.1399825021872266, "explain_loss": 1.625, "loss": 2.859375, "ref_ce_loss": 0.828125, "stage_align_loss": 0.026611328125, "step": 320 }, { "ce_loss": 0.6015625, "distill_loss": 0.103515625, "epoch": 0.1399825021872266, "explain_loss": 1.3046875, "loss": 2.53125, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.020263671875, "step": 320 }, { "epoch": 0.14435695538057744, "grad_norm": 1.315498948097229, "learning_rate": 0.00019183673469387756, "loss": 2.38466796875, "step": 330 }, { "ce_loss": 0.0400390625, "distill_loss": 0.0927734375, "epoch": 0.14435695538057744, "explain_loss": 1.296875, "loss": 1.9375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.01397705078125, "step": 330 }, { "ce_loss": 0.71484375, "distill_loss": 0.10498046875, "epoch": 0.14435695538057744, "explain_loss": 1.0390625, "loss": 2.46875, "ref_ce_loss": 0.578125, "stage_align_loss": 0.0213623046875, "step": 330 }, { "ce_loss": 0.5625, "distill_loss": 0.1171875, "epoch": 0.14435695538057744, "explain_loss": 1.4375, "loss": 2.921875, "ref_ce_loss": 0.77734375, "stage_align_loss": 0.033203125, "step": 330 }, { "ce_loss": 0.181640625, "distill_loss": 0.1005859375, "epoch": 0.14435695538057744, "explain_loss": 1.21875, "loss": 2.03125, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0164794921875, "step": 330 }, { "epoch": 0.14873140857392825, "grad_norm": 2.1174120903015137, "learning_rate": 0.00019766763848396503, "loss": 2.35732421875, "step": 340 }, { "ce_loss": 0.79296875, "distill_loss": 0.09326171875, "epoch": 0.14873140857392825, "explain_loss": 1.2734375, "loss": 2.578125, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.01251220703125, "step": 340 }, { "ce_loss": 1.1484375, "distill_loss": 0.1025390625, "epoch": 0.14873140857392825, "explain_loss": 1.34375, "loss": 3.1875, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.01904296875, "step": 340 }, { "ce_loss": 0.057861328125, "distill_loss": 0.10009765625, "epoch": 0.14873140857392825, "explain_loss": 1.0625, "loss": 1.6875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.01104736328125, "step": 340 }, { "ce_loss": 0.103515625, "distill_loss": 0.0986328125, "epoch": 0.14873140857392825, "explain_loss": 1.4765625, "loss": 2.34375, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.0115966796875, "step": 340 }, { "epoch": 0.15310586176727908, "grad_norm": 1.3244109153747559, "learning_rate": 0.00019999958145408515, "loss": 2.4576171875, "step": 350 }, { "ce_loss": 0.55859375, "distill_loss": 0.091796875, "epoch": 0.15310586176727908, "explain_loss": 0.9609375, "loss": 2.1875, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.028076171875, "step": 350 }, { "ce_loss": 0.73046875, "distill_loss": 0.091796875, "epoch": 0.15310586176727908, "explain_loss": 0.9921875, "loss": 2.34375, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0203857421875, "step": 350 }, { "ce_loss": 0.208984375, "distill_loss": 0.083984375, "epoch": 0.15310586176727908, "explain_loss": 1.328125, "loss": 2.125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.017822265625, "step": 350 }, { "ce_loss": 0.3828125, "distill_loss": 0.08349609375, "epoch": 0.15310586176727908, "explain_loss": 1.40625, "loss": 2.421875, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.020751953125, "step": 350 }, { "epoch": 0.15748031496062992, "grad_norm": 1186.6708984375, "learning_rate": 0.00019999702368618226, "loss": 2.54970703125, "step": 360 }, { "ce_loss": 0.7734375, "distill_loss": 0.123046875, "epoch": 0.15748031496062992, "explain_loss": 1.171875, "loss": 3.265625, "ref_ce_loss": 0.71484375, "stage_align_loss": 0.486328125, "step": 360 }, { "ce_loss": 0.045654296875, "distill_loss": 0.1484375, "epoch": 0.15748031496062992, "explain_loss": 1.2890625, "loss": 2.40625, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.4765625, "step": 360 }, { "ce_loss": 0.310546875, "distill_loss": 0.1328125, "epoch": 0.15748031496062992, "explain_loss": 1.1796875, "loss": 2.625, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.41796875, "step": 360 }, { "ce_loss": 2.125, "distill_loss": 0.11474609375, "epoch": 0.15748031496062992, "explain_loss": 1.203125, "loss": 4.3125, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.419921875, "step": 360 }, { "epoch": 0.16185476815398075, "grad_norm": 2.0178771018981934, "learning_rate": 0.00019999214073528724, "loss": 2.63662109375, "step": 370 }, { "ce_loss": 0.036865234375, "distill_loss": 0.080078125, "epoch": 0.16185476815398075, "explain_loss": 1.5, "loss": 2.0625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.023193359375, "step": 370 }, { "ce_loss": 0.07177734375, "distill_loss": 0.0810546875, "epoch": 0.16185476815398075, "explain_loss": 1.40625, "loss": 2.03125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0267333984375, "step": 370 }, { "ce_loss": 1.03125, "distill_loss": 0.080078125, "epoch": 0.16185476815398075, "explain_loss": 1.3359375, "loss": 2.9375, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.0244140625, "step": 370 }, { "ce_loss": 0.72265625, "distill_loss": 0.09326171875, "epoch": 0.16185476815398075, "explain_loss": 1.2734375, "loss": 2.65625, "ref_ce_loss": 0.546875, "stage_align_loss": 0.032470703125, "step": 370 }, { "epoch": 0.1662292213473316, "grad_norm": 1.7496424913406372, "learning_rate": 0.00019998493271494098, "loss": 2.4212890625, "step": 380 }, { "ce_loss": 0.103515625, "distill_loss": 0.083984375, "epoch": 0.1662292213473316, "explain_loss": 1.734375, "loss": 2.484375, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0264892578125, "step": 380 }, { "ce_loss": 0.09033203125, "distill_loss": 0.09326171875, "epoch": 0.1662292213473316, "explain_loss": 1.0234375, "loss": 1.796875, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.019287109375, "step": 380 }, { "ce_loss": 0.73046875, "distill_loss": 0.09814453125, "epoch": 0.1662292213473316, "explain_loss": 1.28125, "loss": 2.703125, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.02197265625, "step": 380 }, { "ce_loss": 0.365234375, "distill_loss": 0.0810546875, "epoch": 0.1662292213473316, "explain_loss": 1.15625, "loss": 2.09375, "ref_ce_loss": 0.453125, "stage_align_loss": 0.031494140625, "step": 380 }, { "epoch": 0.17060367454068243, "grad_norm": 1.2791790962219238, "learning_rate": 0.00019997539979274816, "loss": 2.2859375, "step": 390 }, { "ce_loss": 1.046875, "distill_loss": 0.0771484375, "epoch": 0.17060367454068243, "explain_loss": 1.203125, "loss": 2.84375, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.02099609375, "step": 390 }, { "ce_loss": 0.8671875, "distill_loss": 0.08544921875, "epoch": 0.17060367454068243, "explain_loss": 1.484375, "loss": 2.96875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.020263671875, "step": 390 }, { "ce_loss": 0.8984375, "distill_loss": 0.107421875, "epoch": 0.17060367454068243, "explain_loss": 0.8515625, "loss": 2.421875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.031494140625, "step": 390 }, { "ce_loss": 0.625, "distill_loss": 0.08251953125, "epoch": 0.17060367454068243, "explain_loss": 1.4375, "loss": 2.75, "ref_ce_loss": 0.578125, "stage_align_loss": 0.033935546875, "step": 390 }, { "epoch": 0.17497812773403323, "grad_norm": 0.9093512892723083, "learning_rate": 0.00019996354219037324, "loss": 2.3376953125, "step": 400 }, { "ce_loss": 1.21875, "distill_loss": 0.0654296875, "epoch": 0.17497812773403323, "explain_loss": 1.046875, "loss": 2.8125, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.0186767578125, "step": 400 }, { "ce_loss": 0.9296875, "distill_loss": 0.07373046875, "epoch": 0.17497812773403323, "explain_loss": 1.703125, "loss": 3.25, "ref_ce_loss": 0.53125, "stage_align_loss": 0.019287109375, "step": 400 }, { "ce_loss": 0.84375, "distill_loss": 0.080078125, "epoch": 0.17497812773403323, "explain_loss": 1.078125, "loss": 2.71875, "ref_ce_loss": 0.6875, "stage_align_loss": 0.032470703125, "step": 400 }, { "ce_loss": 0.06640625, "distill_loss": 0.0693359375, "epoch": 0.17497812773403323, "explain_loss": 1.5, "loss": 2.1875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.01507568359375, "step": 400 }, { "epoch": 0.17935258092738407, "grad_norm": 1.2753005027770996, "learning_rate": 0.0001999493601835354, "loss": 2.380078125, "step": 410 }, { "ce_loss": 0.828125, "distill_loss": 0.06884765625, "epoch": 0.17935258092738407, "explain_loss": 1.25, "loss": 2.703125, "ref_ce_loss": 0.515625, "stage_align_loss": 0.03515625, "step": 410 }, { "ce_loss": 1.21875, "distill_loss": 0.072265625, "epoch": 0.17935258092738407, "explain_loss": 1.296875, "loss": 3.25, "ref_ce_loss": 0.609375, "stage_align_loss": 0.051025390625, "step": 410 }, { "ce_loss": 0.0703125, "distill_loss": 0.0751953125, "epoch": 0.17935258092738407, "explain_loss": 0.8515625, "loss": 1.515625, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.017333984375, "step": 410 }, { "ce_loss": 0.1396484375, "distill_loss": 0.0712890625, "epoch": 0.17935258092738407, "explain_loss": 1.25, "loss": 1.984375, "ref_ce_loss": 0.5, "stage_align_loss": 0.0179443359375, "step": 410 }, { "epoch": 0.1837270341207349, "grad_norm": 46.293758392333984, "learning_rate": 0.000199932854102002, "loss": 2.41875, "step": 420 }, { "ce_loss": 0.44921875, "distill_loss": 0.06103515625, "epoch": 0.1837270341207349, "explain_loss": 1.0625, "loss": 2.21875, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0218505859375, "step": 420 }, { "ce_loss": 0.69921875, "distill_loss": 0.07666015625, "epoch": 0.1837270341207349, "explain_loss": 1.09375, "loss": 2.5, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.02294921875, "step": 420 }, { "ce_loss": 0.28515625, "distill_loss": 0.0556640625, "epoch": 0.1837270341207349, "explain_loss": 0.96484375, "loss": 1.96875, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.02001953125, "step": 420 }, { "ce_loss": 0.41015625, "distill_loss": 0.0654296875, "epoch": 0.1837270341207349, "explain_loss": 1.21875, "loss": 2.09375, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.021728515625, "step": 420 }, { "epoch": 0.18810148731408574, "grad_norm": 1.029591679573059, "learning_rate": 0.00019991402432958117, "loss": 2.3115234375, "step": 430 }, { "ce_loss": 0.11572265625, "distill_loss": 0.05517578125, "epoch": 0.18810148731408574, "explain_loss": 1.4296875, "loss": 2.046875, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.019287109375, "step": 430 }, { "ce_loss": 0.068359375, "distill_loss": 0.0576171875, "epoch": 0.18810148731408574, "explain_loss": 1.65625, "loss": 2.3125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.014404296875, "step": 430 }, { "ce_loss": 0.380859375, "distill_loss": 0.05224609375, "epoch": 0.18810148731408574, "explain_loss": 1.1484375, "loss": 2.25, "ref_ce_loss": 0.65625, "stage_align_loss": 0.018310546875, "step": 430 }, { "ce_loss": 1.6484375, "distill_loss": 0.054931640625, "epoch": 0.18810148731408574, "explain_loss": 1.203125, "loss": 3.546875, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.0198974609375, "step": 430 }, { "epoch": 0.19247594050743658, "grad_norm": 1.2342251539230347, "learning_rate": 0.0001998928713041125, "loss": 2.29970703125, "step": 440 }, { "ce_loss": 0.1181640625, "distill_loss": 0.0634765625, "epoch": 0.19247594050743658, "explain_loss": 1.1484375, "loss": 2.0, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.026123046875, "step": 440 }, { "ce_loss": 0.357421875, "distill_loss": 0.0556640625, "epoch": 0.19247594050743658, "explain_loss": 1.203125, "loss": 2.125, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.0157470703125, "step": 440 }, { "ce_loss": 0.255859375, "distill_loss": 0.064453125, "epoch": 0.19247594050743658, "explain_loss": 1.7578125, "loss": 2.40625, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.01483154296875, "step": 440 }, { "ce_loss": 0.1943359375, "distill_loss": 0.057373046875, "epoch": 0.19247594050743658, "explain_loss": 1.0390625, "loss": 1.796875, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.01611328125, "step": 440 }, { "epoch": 0.1968503937007874, "grad_norm": 0.6972479224205017, "learning_rate": 0.00019986939551745728, "loss": 2.22734375, "step": 450 }, { "ce_loss": 0.09765625, "distill_loss": 0.05517578125, "epoch": 0.1968503937007874, "explain_loss": 1.2734375, "loss": 2.09375, "ref_ce_loss": 0.65234375, "stage_align_loss": 0.0106201171875, "step": 450 }, { "ce_loss": 0.34765625, "distill_loss": 0.0693359375, "epoch": 0.1968503937007874, "explain_loss": 1.0625, "loss": 2.015625, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.01153564453125, "step": 450 }, { "ce_loss": 0.5546875, "distill_loss": 0.06201171875, "epoch": 0.1968503937007874, "explain_loss": 1.3125, "loss": 2.546875, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.0167236328125, "step": 450 }, { "ce_loss": 0.0211181640625, "distill_loss": 0.0498046875, "epoch": 0.1968503937007874, "explain_loss": 0.9609375, "loss": 1.53125, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0108642578125, "step": 450 }, { "epoch": 0.20122484689413822, "grad_norm": 0.7481743693351746, "learning_rate": 0.0001998435975154867, "loss": 2.3076171875, "step": 460 }, { "ce_loss": 0.2197265625, "distill_loss": 0.060546875, "epoch": 0.20122484689413822, "explain_loss": 1.046875, "loss": 1.765625, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.01123046875, "step": 460 }, { "ce_loss": 0.392578125, "distill_loss": 0.0595703125, "epoch": 0.20122484689413822, "explain_loss": 1.0390625, "loss": 2.15625, "ref_ce_loss": 0.640625, "stage_align_loss": 0.01239013671875, "step": 460 }, { "ce_loss": 0.3203125, "distill_loss": 0.0517578125, "epoch": 0.20122484689413822, "explain_loss": 1.484375, "loss": 2.34375, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.0128173828125, "step": 460 }, { "ce_loss": 0.1259765625, "distill_loss": 0.0654296875, "epoch": 0.20122484689413822, "explain_loss": 1.484375, "loss": 2.171875, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0107421875, "step": 460 }, { "epoch": 0.20559930008748906, "grad_norm": 1.3833484649658203, "learning_rate": 0.00019981547789806937, "loss": 2.207421875, "step": 470 }, { "ce_loss": 0.2734375, "distill_loss": 0.0576171875, "epoch": 0.20559930008748906, "explain_loss": 1.359375, "loss": 2.0625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.01226806640625, "step": 470 }, { "ce_loss": 0.0390625, "distill_loss": 0.05419921875, "epoch": 0.20559930008748906, "explain_loss": 1.421875, "loss": 2.0, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.01190185546875, "step": 470 }, { "ce_loss": 0.54296875, "distill_loss": 0.056396484375, "epoch": 0.20559930008748906, "explain_loss": 1.0625, "loss": 2.21875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.01190185546875, "step": 470 }, { "ce_loss": 1.125, "distill_loss": 0.05126953125, "epoch": 0.20559930008748906, "explain_loss": 1.34375, "loss": 3.046875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.012451171875, "step": 470 }, { "epoch": 0.2099737532808399, "grad_norm": 1.5563236474990845, "learning_rate": 0.00019978503731905733, "loss": 2.28203125, "step": 480 }, { "ce_loss": 0.365234375, "distill_loss": 0.0458984375, "epoch": 0.2099737532808399, "explain_loss": 1.25, "loss": 2.21875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.023681640625, "step": 480 }, { "ce_loss": 0.6953125, "distill_loss": 0.04931640625, "epoch": 0.2099737532808399, "explain_loss": 1.46875, "loss": 3.09375, "ref_ce_loss": 0.85546875, "stage_align_loss": 0.033447265625, "step": 480 }, { "ce_loss": 0.625, "distill_loss": 0.04443359375, "epoch": 0.2099737532808399, "explain_loss": 1.015625, "loss": 2.421875, "ref_ce_loss": 0.7265625, "stage_align_loss": 0.0179443359375, "step": 480 }, { "ce_loss": 0.330078125, "distill_loss": 0.0458984375, "epoch": 0.2099737532808399, "explain_loss": 1.1015625, "loss": 2.28125, "ref_ce_loss": 0.7734375, "stage_align_loss": 0.024658203125, "step": 480 }, { "epoch": 0.21434820647419073, "grad_norm": 0.6766189932823181, "learning_rate": 0.0001997522764862708, "loss": 2.490234375, "step": 490 }, { "ce_loss": 0.1806640625, "distill_loss": 0.040283203125, "epoch": 0.21434820647419073, "explain_loss": 1.125, "loss": 1.6953125, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.0089111328125, "step": 490 }, { "ce_loss": 0.65625, "distill_loss": 0.0458984375, "epoch": 0.21434820647419073, "explain_loss": 1.46875, "loss": 2.6875, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.01458740234375, "step": 490 }, { "ce_loss": 0.1162109375, "distill_loss": 0.05029296875, "epoch": 0.21434820647419073, "explain_loss": 1.5, "loss": 2.171875, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.0184326171875, "step": 490 }, { "ce_loss": 0.3515625, "distill_loss": 0.06396484375, "epoch": 0.21434820647419073, "explain_loss": 1.515625, "loss": 2.75, "ref_ce_loss": 0.77734375, "stage_align_loss": 0.0296630859375, "step": 490 }, { "epoch": 0.21872265966754156, "grad_norm": 0.833185613155365, "learning_rate": 0.00019971719616148175, "loss": 2.25693359375, "step": 500 }, { "ce_loss": 1.125, "distill_loss": 0.052490234375, "epoch": 0.21872265966754156, "explain_loss": 1.1640625, "loss": 2.890625, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.0238037109375, "step": 500 }, { "ce_loss": 0.11962890625, "distill_loss": 0.0556640625, "epoch": 0.21872265966754156, "explain_loss": 1.25, "loss": 2.046875, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.04150390625, "step": 500 }, { "ce_loss": 0.07177734375, "distill_loss": 0.054931640625, "epoch": 0.21872265966754156, "explain_loss": 0.96875, "loss": 1.515625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.025146484375, "step": 500 }, { "ce_loss": 0.01434326171875, "distill_loss": 0.049072265625, "epoch": 0.21872265966754156, "explain_loss": 1.3828125, "loss": 1.984375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0118408203125, "step": 500 }, { "epoch": 0.2230971128608924, "grad_norm": 0.886531412601471, "learning_rate": 0.00019967979716039623, "loss": 2.26474609375, "step": 510 }, { "ce_loss": 0.115234375, "distill_loss": 0.045166015625, "epoch": 0.2230971128608924, "explain_loss": 1.03125, "loss": 1.5078125, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.0206298828125, "step": 510 }, { "ce_loss": 0.384765625, "distill_loss": 0.0517578125, "epoch": 0.2230971128608924, "explain_loss": 1.4453125, "loss": 2.421875, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.0252685546875, "step": 510 }, { "ce_loss": 0.045654296875, "distill_loss": 0.050048828125, "epoch": 0.2230971128608924, "explain_loss": 1.265625, "loss": 1.6875, "ref_ce_loss": 0.302734375, "stage_align_loss": 0.0203857421875, "step": 510 }, { "ce_loss": 0.1376953125, "distill_loss": 0.0546875, "epoch": 0.2230971128608924, "explain_loss": 1.5, "loss": 2.3125, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.0234375, "step": 510 }, { "epoch": 0.2274715660542432, "grad_norm": 1.0455466508865356, "learning_rate": 0.0001996400803526353, "loss": 2.15478515625, "step": 520 }, { "ce_loss": 1.421875, "distill_loss": 0.06640625, "epoch": 0.2274715660542432, "explain_loss": 1.265625, "loss": 3.359375, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.0164794921875, "step": 520 }, { "ce_loss": 0.462890625, "distill_loss": 0.0517578125, "epoch": 0.2274715660542432, "explain_loss": 1.59375, "loss": 2.71875, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.055908203125, "step": 520 }, { "ce_loss": 0.5078125, "distill_loss": 0.053955078125, "epoch": 0.2274715660542432, "explain_loss": 1.0625, "loss": 2.0625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.049072265625, "step": 520 }, { "ce_loss": 0.349609375, "distill_loss": 0.055908203125, "epoch": 0.2274715660542432, "explain_loss": 1.09375, "loss": 2.125, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.0203857421875, "step": 520 }, { "epoch": 0.23184601924759404, "grad_norm": 1.3697181940078735, "learning_rate": 0.000199598046661715, "loss": 2.21796875, "step": 530 }, { "ce_loss": 0.625, "distill_loss": 0.0556640625, "epoch": 0.23184601924759404, "explain_loss": 1.59375, "loss": 2.859375, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.043701171875, "step": 530 }, { "ce_loss": 0.05908203125, "distill_loss": 0.06103515625, "epoch": 0.23184601924759404, "explain_loss": 1.078125, "loss": 1.703125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.01531982421875, "step": 530 }, { "ce_loss": 0.267578125, "distill_loss": 0.05078125, "epoch": 0.23184601924759404, "explain_loss": 1.109375, "loss": 1.84375, "ref_ce_loss": 0.390625, "stage_align_loss": 0.020751953125, "step": 530 }, { "ce_loss": 0.04443359375, "distill_loss": 0.053466796875, "epoch": 0.23184601924759404, "explain_loss": 1.3828125, "loss": 2.0625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01312255859375, "step": 530 }, { "epoch": 0.23622047244094488, "grad_norm": 1.044211506843567, "learning_rate": 0.0001995536970650245, "loss": 2.33203125, "step": 540 }, { "ce_loss": 1.453125, "distill_loss": 0.047607421875, "epoch": 0.23622047244094488, "explain_loss": 1.234375, "loss": 3.5, "ref_ce_loss": 0.74609375, "stage_align_loss": 0.0107421875, "step": 540 }, { "ce_loss": 0.1884765625, "distill_loss": 0.0478515625, "epoch": 0.23622047244094488, "explain_loss": 1.015625, "loss": 1.671875, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.027099609375, "step": 540 }, { "ce_loss": 0.07666015625, "distill_loss": 0.0458984375, "epoch": 0.23622047244094488, "explain_loss": 1.15625, "loss": 1.9609375, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.01904296875, "step": 540 }, { "ce_loss": 0.1416015625, "distill_loss": 0.05908203125, "epoch": 0.23622047244094488, "explain_loss": 1.6953125, "loss": 2.46875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.033203125, "step": 540 }, { "epoch": 0.24059492563429571, "grad_norm": 0.680176317691803, "learning_rate": 0.0001995070325938039, "loss": 2.21083984375, "step": 550 }, { "ce_loss": 0.023681640625, "distill_loss": 0.05078125, "epoch": 0.24059492563429571, "explain_loss": 1.1796875, "loss": 1.8515625, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.01055908203125, "step": 550 }, { "ce_loss": 1.0625, "distill_loss": 0.052490234375, "epoch": 0.24059492563429571, "explain_loss": 1.046875, "loss": 2.625, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0250244140625, "step": 550 }, { "ce_loss": 0.064453125, "distill_loss": 0.0458984375, "epoch": 0.24059492563429571, "explain_loss": 1.265625, "loss": 1.78125, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.01953125, "step": 550 }, { "ce_loss": 0.12109375, "distill_loss": 0.047607421875, "epoch": 0.24059492563429571, "explain_loss": 1.2265625, "loss": 1.890625, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.018798828125, "step": 550 }, { "epoch": 0.24496937882764655, "grad_norm": 1.782778024673462, "learning_rate": 0.0001994580543331198, "loss": 2.1056640625, "step": 560 }, { "ce_loss": 1.234375, "distill_loss": 0.04541015625, "epoch": 0.24496937882764655, "explain_loss": 1.046875, "loss": 2.828125, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.01470947265625, "step": 560 }, { "ce_loss": 0.21484375, "distill_loss": 0.043701171875, "epoch": 0.24496937882764655, "explain_loss": 1.1875, "loss": 1.9296875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0115966796875, "step": 560 }, { "ce_loss": 0.90234375, "distill_loss": 0.0380859375, "epoch": 0.24496937882764655, "explain_loss": 1.421875, "loss": 2.953125, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.013916015625, "step": 560 }, { "ce_loss": 0.326171875, "distill_loss": 0.053955078125, "epoch": 0.24496937882764655, "explain_loss": 1.328125, "loss": 2.28125, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.0137939453125, "step": 560 }, { "epoch": 0.24934383202099739, "grad_norm": 1.3406211137771606, "learning_rate": 0.00019940676342184027, "loss": 2.166015625, "step": 570 }, { "ce_loss": 0.0810546875, "distill_loss": 0.036865234375, "epoch": 0.24934383202099739, "explain_loss": 1.2421875, "loss": 1.96875, "ref_ce_loss": 0.59375, "stage_align_loss": 0.0101318359375, "step": 570 }, { "ce_loss": 0.1025390625, "distill_loss": 0.0498046875, "epoch": 0.24934383202099739, "explain_loss": 0.9921875, "loss": 1.5625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0186767578125, "step": 570 }, { "ce_loss": 0.10986328125, "distill_loss": 0.053466796875, "epoch": 0.24934383202099739, "explain_loss": 1.2578125, "loss": 1.96875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.01287841796875, "step": 570 }, { "ce_loss": 0.09228515625, "distill_loss": 0.048828125, "epoch": 0.24934383202099739, "explain_loss": 1.078125, "loss": 1.6953125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.01129150390625, "step": 570 }, { "epoch": 0.2537182852143482, "grad_norm": 17.25648307800293, "learning_rate": 0.00019935316105260836, "loss": 2.2587890625, "step": 580 }, { "ce_loss": 0.90625, "distill_loss": 0.036376953125, "epoch": 0.2537182852143482, "explain_loss": 1.90625, "loss": 3.484375, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.01220703125, "step": 580 }, { "ce_loss": 0.94921875, "distill_loss": 0.04345703125, "epoch": 0.2537182852143482, "explain_loss": 1.3203125, "loss": 2.84375, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.0113525390625, "step": 580 }, { "ce_loss": 3.78125, "distill_loss": 0.0546875, "epoch": 0.2537182852143482, "explain_loss": 1.2265625, "loss": 5.8125, "ref_ce_loss": 0.7578125, "stage_align_loss": 0.010498046875, "step": 580 }, { "ce_loss": 0.1240234375, "distill_loss": 0.041259765625, "epoch": 0.2537182852143482, "explain_loss": 1.28125, "loss": 1.9140625, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.01239013671875, "step": 580 }, { "epoch": 0.25809273840769903, "grad_norm": 0.8300268054008484, "learning_rate": 0.00019929724847181436, "loss": 2.426953125, "step": 590 }, { "ce_loss": 0.52734375, "distill_loss": 0.0634765625, "epoch": 0.25809273840769903, "explain_loss": 1.375, "loss": 2.46875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.01141357421875, "step": 590 }, { "ce_loss": 0.392578125, "distill_loss": 0.05322265625, "epoch": 0.25809273840769903, "explain_loss": 0.984375, "loss": 1.9921875, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.010009765625, "step": 590 }, { "ce_loss": 0.39453125, "distill_loss": 0.046630859375, "epoch": 0.25809273840769903, "explain_loss": 1.328125, "loss": 2.3125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.009765625, "step": 590 }, { "ce_loss": 0.341796875, "distill_loss": 0.05126953125, "epoch": 0.25809273840769903, "explain_loss": 0.875, "loss": 1.5546875, "ref_ce_loss": 0.28125, "stage_align_loss": 0.00970458984375, "step": 590 }, { "epoch": 0.26246719160104987, "grad_norm": 3.5695483684539795, "learning_rate": 0.00019923902697956675, "loss": 2.11455078125, "step": 600 }, { "ce_loss": 0.03369140625, "distill_loss": 0.04931640625, "epoch": 0.26246719160104987, "explain_loss": 1.265625, "loss": 1.78125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.00921630859375, "step": 600 }, { "ce_loss": 0.039794921875, "distill_loss": 0.04052734375, "epoch": 0.26246719160104987, "explain_loss": 0.91015625, "loss": 1.453125, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.01190185546875, "step": 600 }, { "ce_loss": 0.251953125, "distill_loss": 0.05322265625, "epoch": 0.26246719160104987, "explain_loss": 1.203125, "loss": 2.03125, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.015869140625, "step": 600 }, { "ce_loss": 0.06982421875, "distill_loss": 0.05126953125, "epoch": 0.26246719160104987, "explain_loss": 1.390625, "loss": 2.3125, "ref_ce_loss": 0.796875, "stage_align_loss": 0.01129150390625, "step": 600 }, { "epoch": 0.2668416447944007, "grad_norm": 0.6545160412788391, "learning_rate": 0.00019917849792966207, "loss": 2.1326171875, "step": 610 }, { "ce_loss": 0.345703125, "distill_loss": 0.0478515625, "epoch": 0.2668416447944007, "explain_loss": 0.98828125, "loss": 1.8359375, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.01324462890625, "step": 610 }, { "ce_loss": 0.043212890625, "distill_loss": 0.045166015625, "epoch": 0.2668416447944007, "explain_loss": 1.3125, "loss": 1.890625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.01043701171875, "step": 610 }, { "ce_loss": 0.306640625, "distill_loss": 0.0478515625, "epoch": 0.2668416447944007, "explain_loss": 1.078125, "loss": 1.875, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0089111328125, "step": 610 }, { "ce_loss": 0.07666015625, "distill_loss": 0.04931640625, "epoch": 0.2668416447944007, "explain_loss": 0.9296875, "loss": 1.4375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.01806640625, "step": 610 }, { "epoch": 0.27121609798775154, "grad_norm": 0.7463482618331909, "learning_rate": 0.00019911566272955345, "loss": 1.95849609375, "step": 620 }, { "ce_loss": 0.216796875, "distill_loss": 0.042724609375, "epoch": 0.27121609798775154, "explain_loss": 1.4140625, "loss": 2.34375, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.00921630859375, "step": 620 }, { "ce_loss": 0.53125, "distill_loss": 0.0498046875, "epoch": 0.27121609798775154, "explain_loss": 1.140625, "loss": 2.265625, "ref_ce_loss": 0.515625, "stage_align_loss": 0.0274658203125, "step": 620 }, { "ce_loss": 0.0458984375, "distill_loss": 0.04248046875, "epoch": 0.27121609798775154, "explain_loss": 1.3828125, "loss": 2.03125, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.00970458984375, "step": 620 }, { "ce_loss": 0.08642578125, "distill_loss": 0.03662109375, "epoch": 0.27121609798775154, "explain_loss": 1.390625, "loss": 1.9296875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0079345703125, "step": 620 }, { "epoch": 0.2755905511811024, "grad_norm": 0.6461924314498901, "learning_rate": 0.00019905052284031763, "loss": 2.082421875, "step": 630 }, { "ce_loss": 0.353515625, "distill_loss": 0.09765625, "epoch": 0.2755905511811024, "explain_loss": 1.390625, "loss": 2.375, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.017333984375, "step": 630 }, { "ce_loss": 0.1787109375, "distill_loss": 0.0615234375, "epoch": 0.2755905511811024, "explain_loss": 1.46875, "loss": 2.1875, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.00909423828125, "step": 630 }, { "ce_loss": 0.28125, "distill_loss": 0.07861328125, "epoch": 0.2755905511811024, "explain_loss": 1.3671875, "loss": 2.1875, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.01348876953125, "step": 630 }, { "ce_loss": 0.06298828125, "distill_loss": 0.06494140625, "epoch": 0.2755905511811024, "explain_loss": 1.515625, "loss": 2.265625, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.008544921875, "step": 630 }, { "epoch": 0.2799650043744532, "grad_norm": 14.378064155578613, "learning_rate": 0.00019898307977662144, "loss": 4.371875, "step": 640 }, { "ce_loss": 2.296875, "distill_loss": 0.0673828125, "epoch": 0.2799650043744532, "explain_loss": 1.125, "loss": 4.0, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.00775146484375, "step": 640 }, { "ce_loss": 2.578125, "distill_loss": 0.0693359375, "epoch": 0.2799650043744532, "explain_loss": 1.2578125, "loss": 4.46875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0205078125, "step": 640 }, { "ce_loss": 0.1142578125, "distill_loss": 0.10986328125, "epoch": 0.2799650043744532, "explain_loss": 1.140625, "loss": 1.8828125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.005859375, "step": 640 }, { "ce_loss": 2.515625, "distill_loss": 0.05322265625, "epoch": 0.2799650043744532, "explain_loss": 0.984375, "loss": 4.25, "ref_ce_loss": 0.6796875, "stage_align_loss": 0.0128173828125, "step": 640 }, { "epoch": 0.28433945756780404, "grad_norm": 1.2668070793151855, "learning_rate": 0.00019891333510668613, "loss": 2.72802734375, "step": 650 }, { "ce_loss": 0.045654296875, "distill_loss": 0.0693359375, "epoch": 0.28433945756780404, "explain_loss": 1.3359375, "loss": 1.921875, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.01153564453125, "step": 650 }, { "ce_loss": 0.1875, "distill_loss": 0.07177734375, "epoch": 0.28433945756780404, "explain_loss": 1.3359375, "loss": 2.0, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.01226806640625, "step": 650 }, { "ce_loss": 0.392578125, "distill_loss": 0.05126953125, "epoch": 0.28433945756780404, "explain_loss": 1.2265625, "loss": 2.3125, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0115966796875, "step": 650 }, { "ce_loss": 0.11279296875, "distill_loss": 0.05078125, "epoch": 0.28433945756780404, "explain_loss": 1.0625, "loss": 1.5625, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.0135498046875, "step": 650 }, { "epoch": 0.2887139107611549, "grad_norm": 26.046934127807617, "learning_rate": 0.0001988412904522512, "loss": 9.74677734375, "step": 660 }, { "ce_loss": 3.65625, "distill_loss": 0.04833984375, "epoch": 0.2887139107611549, "explain_loss": 4.96875, "loss": 14.125, "ref_ce_loss": 5.4375, "stage_align_loss": 0.01220703125, "step": 660 }, { "ce_loss": 3.484375, "distill_loss": 0.0595703125, "epoch": 0.2887139107611549, "explain_loss": 4.59375, "loss": 12.625, "ref_ce_loss": 4.5, "stage_align_loss": 0.0106201171875, "step": 660 }, { "ce_loss": 4.40625, "distill_loss": 0.0703125, "epoch": 0.2887139107611549, "explain_loss": 4.75, "loss": 14.25, "ref_ce_loss": 5.0, "stage_align_loss": 0.01373291015625, "step": 660 }, { "ce_loss": 3.78125, "distill_loss": 0.059814453125, "epoch": 0.2887139107611549, "explain_loss": 4.78125, "loss": 14.375, "ref_ce_loss": 5.8125, "stage_align_loss": 0.0106201171875, "step": 660 }, { "epoch": 0.29308836395450566, "grad_norm": 15.79896354675293, "learning_rate": 0.00019876694748853654, "loss": 20.22890625, "step": 670 }, { "ce_loss": 2.453125, "distill_loss": 0.0654296875, "epoch": 0.29308836395450566, "explain_loss": 5.25, "loss": 15.625, "ref_ce_loss": 7.875, "stage_align_loss": 0.00054168701171875, "step": 670 }, { "ce_loss": 5.125, "distill_loss": 0.0634765625, "epoch": 0.29308836395450566, "explain_loss": 4.625, "loss": 16.375, "ref_ce_loss": 6.59375, "stage_align_loss": 0.00057220458984375, "step": 670 }, { "ce_loss": 5.78125, "distill_loss": 0.08544921875, "epoch": 0.29308836395450566, "explain_loss": 4.9375, "loss": 17.0, "ref_ce_loss": 6.28125, "stage_align_loss": 0.0013427734375, "step": 670 }, { "ce_loss": 2.4375, "distill_loss": 0.07177734375, "epoch": 0.29308836395450566, "explain_loss": 5.0, "loss": 14.625, "ref_ce_loss": 7.125, "stage_align_loss": 0.000820159912109375, "step": 670 }, { "epoch": 0.2974628171478565, "grad_norm": 11.70031452178955, "learning_rate": 0.00019869030794420357, "loss": 14.06015625, "step": 680 }, { "ce_loss": 2.140625, "distill_loss": 0.07080078125, "epoch": 0.2974628171478565, "explain_loss": 3.03125, "loss": 11.125, "ref_ce_loss": 5.84375, "stage_align_loss": 0.0028228759765625, "step": 680 }, { "ce_loss": 2.09375, "distill_loss": 0.0830078125, "epoch": 0.2974628171478565, "explain_loss": 3.46875, "loss": 10.75, "ref_ce_loss": 5.125, "stage_align_loss": 0.00191497802734375, "step": 680 }, { "ce_loss": 1.9609375, "distill_loss": 0.0771484375, "epoch": 0.2974628171478565, "explain_loss": 3.78125, "loss": 10.25, "ref_ce_loss": 4.4375, "stage_align_loss": 0.00225830078125, "step": 680 }, { "ce_loss": 2.046875, "distill_loss": 0.0595703125, "epoch": 0.2974628171478565, "explain_loss": 3.921875, "loss": 11.4375, "ref_ce_loss": 5.1875, "stage_align_loss": 0.2001953125, "step": 680 }, { "epoch": 0.30183727034120733, "grad_norm": 9.112813949584961, "learning_rate": 0.000198611373601315, "loss": 10.5375, "step": 690 }, { "ce_loss": 1.828125, "distill_loss": 0.0751953125, "epoch": 0.30183727034120733, "explain_loss": 2.84375, "loss": 9.125, "ref_ce_loss": 4.375, "stage_align_loss": 0.0027618408203125, "step": 690 }, { "ce_loss": 1.9375, "distill_loss": 0.091796875, "epoch": 0.30183727034120733, "explain_loss": 3.28125, "loss": 9.8125, "ref_ce_loss": 4.5, "stage_align_loss": 0.00146484375, "step": 690 }, { "ce_loss": 1.9375, "distill_loss": 0.08251953125, "epoch": 0.30183727034120733, "explain_loss": 3.015625, "loss": 9.625, "ref_ce_loss": 4.59375, "stage_align_loss": 0.0027313232421875, "step": 690 }, { "ce_loss": 1.96875, "distill_loss": 0.0927734375, "epoch": 0.30183727034120733, "explain_loss": 3.1875, "loss": 8.5, "ref_ce_loss": 3.234375, "stage_align_loss": 0.00077056884765625, "step": 690 }, { "epoch": 0.30621172353455817, "grad_norm": 9.88357925415039, "learning_rate": 0.00019853014629529333, "loss": 8.35, "step": 700 }, { "ce_loss": 0.38671875, "distill_loss": 0.095703125, "epoch": 0.30621172353455817, "explain_loss": 2.484375, "loss": 7.6875, "ref_ce_loss": 4.71875, "stage_align_loss": 0.00421142578125, "step": 700 }, { "ce_loss": 1.9296875, "distill_loss": 0.087890625, "epoch": 0.30621172353455817, "explain_loss": 2.375, "loss": 8.125, "ref_ce_loss": 3.71875, "stage_align_loss": 0.0026702880859375, "step": 700 }, { "ce_loss": 2.140625, "distill_loss": 0.09765625, "epoch": 0.30621172353455817, "explain_loss": 2.625, "loss": 8.9375, "ref_ce_loss": 4.09375, "stage_align_loss": 0.0030364990234375, "step": 700 }, { "ce_loss": 2.046875, "distill_loss": 0.095703125, "epoch": 0.30621172353455817, "explain_loss": 2.6875, "loss": 9.125, "ref_ce_loss": 4.3125, "stage_align_loss": 0.0037994384765625, "step": 700 }, { "epoch": 0.310586176727909, "grad_norm": 6.795778274536133, "learning_rate": 0.00019844662791487826, "loss": 7.934375, "step": 710 }, { "ce_loss": 0.28125, "distill_loss": 0.0908203125, "epoch": 0.310586176727909, "explain_loss": 2.40625, "loss": 6.625, "ref_ce_loss": 3.703125, "stage_align_loss": 0.1650390625, "step": 710 }, { "ce_loss": 1.0390625, "distill_loss": 0.11181640625, "epoch": 0.310586176727909, "explain_loss": 2.25, "loss": 6.5625, "ref_ce_loss": 3.171875, "stage_align_loss": 0.0031890869140625, "step": 710 }, { "ce_loss": 1.0, "distill_loss": 0.0888671875, "epoch": 0.310586176727909, "explain_loss": 2.21875, "loss": 6.59375, "ref_ce_loss": 3.28125, "stage_align_loss": 0.0017852783203125, "step": 710 }, { "ce_loss": 1.6640625, "distill_loss": 0.0869140625, "epoch": 0.310586176727909, "explain_loss": 2.3125, "loss": 7.34375, "ref_ce_loss": 3.09375, "stage_align_loss": 0.1787109375, "step": 710 }, { "epoch": 0.31496062992125984, "grad_norm": 6.763725280761719, "learning_rate": 0.00019836082040208285, "loss": 7.13125, "step": 720 }, { "ce_loss": 1.0078125, "distill_loss": 0.10986328125, "epoch": 0.31496062992125984, "explain_loss": 2.4375, "loss": 7.28125, "ref_ce_loss": 3.328125, "stage_align_loss": 0.4140625, "step": 720 }, { "ce_loss": 0.3671875, "distill_loss": 0.1103515625, "epoch": 0.31496062992125984, "explain_loss": 2.546875, "loss": 7.25, "ref_ce_loss": 4.0625, "stage_align_loss": 0.193359375, "step": 720 }, { "ce_loss": 0.51171875, "distill_loss": 0.109375, "epoch": 0.31496062992125984, "explain_loss": 2.0625, "loss": 6.53125, "ref_ce_loss": 3.65625, "stage_align_loss": 0.1875, "step": 720 }, { "ce_loss": 0.361328125, "distill_loss": 0.1064453125, "epoch": 0.31496062992125984, "explain_loss": 2.3125, "loss": 6.5625, "ref_ce_loss": 3.359375, "stage_align_loss": 0.40625, "step": 720 }, { "epoch": 0.3193350831146107, "grad_norm": 4.200126647949219, "learning_rate": 0.00019827272575214812, "loss": 6.651953125, "step": 730 }, { "ce_loss": 0.64453125, "distill_loss": 0.1279296875, "epoch": 0.3193350831146107, "explain_loss": 1.8359375, "loss": 5.78125, "ref_ce_loss": 2.625, "stage_align_loss": 0.546875, "step": 730 }, { "ce_loss": 0.2138671875, "distill_loss": 0.11328125, "epoch": 0.3193350831146107, "explain_loss": 1.71875, "loss": 4.375, "ref_ce_loss": 1.921875, "stage_align_loss": 0.390625, "step": 730 }, { "ce_loss": 0.189453125, "distill_loss": 0.123046875, "epoch": 0.3193350831146107, "explain_loss": 1.5703125, "loss": 4.1875, "ref_ce_loss": 1.9140625, "stage_align_loss": 0.396484375, "step": 730 }, { "ce_loss": 0.859375, "distill_loss": 0.1005859375, "epoch": 0.3193350831146107, "explain_loss": 2.09375, "loss": 5.8125, "ref_ce_loss": 2.34375, "stage_align_loss": 0.38671875, "step": 730 }, { "epoch": 0.3237095363079615, "grad_norm": 2.175652027130127, "learning_rate": 0.00019818234601349686, "loss": 4.1162109375, "step": 740 }, { "ce_loss": 0.095703125, "distill_loss": 0.099609375, "epoch": 0.3237095363079615, "explain_loss": 1.875, "loss": 3.46875, "ref_ce_loss": 1.3984375, "stage_align_loss": 0.00390625, "step": 740 }, { "ce_loss": 0.365234375, "distill_loss": 0.095703125, "epoch": 0.3237095363079615, "explain_loss": 1.765625, "loss": 3.4375, "ref_ce_loss": 1.203125, "stage_align_loss": 0.0042724609375, "step": 740 }, { "ce_loss": 0.220703125, "distill_loss": 0.1533203125, "epoch": 0.3237095363079615, "explain_loss": 1.3984375, "loss": 3.25, "ref_ce_loss": 1.484375, "stage_align_loss": 0.00341796875, "step": 740 }, { "ce_loss": 0.28125, "distill_loss": 0.091796875, "epoch": 0.3237095363079615, "explain_loss": 1.9296875, "loss": 3.5625, "ref_ce_loss": 1.2421875, "stage_align_loss": 0.003936767578125, "step": 740 }, { "epoch": 0.32808398950131235, "grad_norm": 2.2784175872802734, "learning_rate": 0.00019808968328768593, "loss": 2.95556640625, "step": 750 }, { "ce_loss": 0.318359375, "distill_loss": 0.09716796875, "epoch": 0.32808398950131235, "explain_loss": 1.0625, "loss": 2.5625, "ref_ce_loss": 0.73046875, "stage_align_loss": 0.359375, "step": 750 }, { "ce_loss": 0.419921875, "distill_loss": 0.125, "epoch": 0.32808398950131235, "explain_loss": 1.4375, "loss": 2.9375, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.333984375, "step": 750 }, { "ce_loss": 0.34375, "distill_loss": 0.10791015625, "epoch": 0.32808398950131235, "explain_loss": 1.3125, "loss": 2.84375, "ref_ce_loss": 0.75, "stage_align_loss": 0.337890625, "step": 750 }, { "ce_loss": 0.67578125, "distill_loss": 0.1611328125, "epoch": 0.32808398950131235, "explain_loss": 1.25, "loss": 3.265625, "ref_ce_loss": 0.8359375, "stage_align_loss": 0.337890625, "step": 750 }, { "epoch": 0.3324584426946632, "grad_norm": 3.835927963256836, "learning_rate": 0.00019799473972935741, "loss": 3.1591796875, "step": 760 }, { "ce_loss": 0.072265625, "distill_loss": 0.72265625, "epoch": 0.3324584426946632, "explain_loss": 1.6171875, "loss": 3.40625, "ref_ce_loss": 0.640625, "stage_align_loss": 0.345703125, "step": 760 }, { "ce_loss": 0.2265625, "distill_loss": 0.796875, "epoch": 0.3324584426946632, "explain_loss": 0.9609375, "loss": 3.0, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.34765625, "step": 760 }, { "ce_loss": 0.06494140625, "distill_loss": 0.7734375, "epoch": 0.3324584426946632, "explain_loss": 0.9765625, "loss": 3.0, "ref_ce_loss": 0.83984375, "stage_align_loss": 0.34375, "step": 760 }, { "ce_loss": 0.271484375, "distill_loss": 0.859375, "epoch": 0.3324584426946632, "explain_loss": 1.390625, "loss": 3.765625, "ref_ce_loss": 1.0859375, "stage_align_loss": 0.16015625, "step": 760 }, { "epoch": 0.336832895888014, "grad_norm": 4.402361869812012, "learning_rate": 0.0001978975175461885, "loss": 3.0578125, "step": 770 }, { "ce_loss": 1.0234375, "distill_loss": 0.11669921875, "epoch": 0.336832895888014, "explain_loss": 1.515625, "loss": 3.34375, "ref_ce_loss": 0.671875, "stage_align_loss": 0.01263427734375, "step": 770 }, { "ce_loss": 0.59375, "distill_loss": 0.1044921875, "epoch": 0.336832895888014, "explain_loss": 1.390625, "loss": 2.90625, "ref_ce_loss": 0.81640625, "stage_align_loss": 0.0107421875, "step": 770 }, { "ce_loss": 0.06884765625, "distill_loss": 0.08984375, "epoch": 0.336832895888014, "explain_loss": 1.25, "loss": 2.015625, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.011474609375, "step": 770 }, { "ce_loss": 0.205078125, "distill_loss": 0.1201171875, "epoch": 0.336832895888014, "explain_loss": 1.25, "loss": 2.234375, "ref_ce_loss": 0.65234375, "stage_align_loss": 0.01141357421875, "step": 770 }, { "epoch": 0.34120734908136485, "grad_norm": 1.070021629333496, "learning_rate": 0.0001977980189988401, "loss": 2.7716796875, "step": 780 }, { "ce_loss": 0.142578125, "distill_loss": 0.103515625, "epoch": 0.34120734908136485, "explain_loss": 1.2890625, "loss": 2.46875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.3984375, "step": 780 }, { "ce_loss": 1.0, "distill_loss": 0.10546875, "epoch": 0.34120734908136485, "explain_loss": 1.21875, "loss": 3.5, "ref_ce_loss": 0.7734375, "stage_align_loss": 0.39453125, "step": 780 }, { "ce_loss": 0.103515625, "distill_loss": 0.138671875, "epoch": 0.34120734908136485, "explain_loss": 1.109375, "loss": 2.5, "ref_ce_loss": 0.76171875, "stage_align_loss": 0.380859375, "step": 780 }, { "ce_loss": 0.041015625, "distill_loss": 0.0869140625, "epoch": 0.34120734908136485, "explain_loss": 1.203125, "loss": 2.28125, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.392578125, "step": 780 }, { "epoch": 0.3455818022747157, "grad_norm": 1.1287102699279785, "learning_rate": 0.0001976962464009044, "loss": 2.70703125, "step": 790 }, { "ce_loss": 0.255859375, "distill_loss": 0.091796875, "epoch": 0.3455818022747157, "explain_loss": 1.0703125, "loss": 2.40625, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.35546875, "step": 790 }, { "ce_loss": 0.466796875, "distill_loss": 0.09814453125, "epoch": 0.3455818022747157, "explain_loss": 1.25, "loss": 2.71875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.3515625, "step": 790 }, { "ce_loss": 0.369140625, "distill_loss": 0.08251953125, "epoch": 0.3455818022747157, "explain_loss": 0.93359375, "loss": 2.21875, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.345703125, "step": 790 }, { "ce_loss": 0.1572265625, "distill_loss": 0.0986328125, "epoch": 0.3455818022747157, "explain_loss": 1.484375, "loss": 2.6875, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.34765625, "step": 790 }, { "epoch": 0.34995625546806647, "grad_norm": 3.9079782962799072, "learning_rate": 0.00019759220211885097, "loss": 2.58779296875, "step": 800 }, { "ce_loss": 0.09033203125, "distill_loss": 0.1796875, "epoch": 0.34995625546806647, "explain_loss": 0.9453125, "loss": 1.640625, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.0810546875, "step": 800 }, { "ce_loss": 0.439453125, "distill_loss": 0.22265625, "epoch": 0.34995625546806647, "explain_loss": 1.015625, "loss": 2.4375, "ref_ce_loss": 0.69921875, "stage_align_loss": 0.07080078125, "step": 800 }, { "ce_loss": 0.23828125, "distill_loss": 0.134765625, "epoch": 0.34995625546806647, "explain_loss": 1.234375, "loss": 2.203125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.05517578125, "step": 800 }, { "ce_loss": 0.0703125, "distill_loss": 0.2099609375, "epoch": 0.34995625546806647, "explain_loss": 1.2109375, "loss": 2.046875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0595703125, "step": 800 }, { "epoch": 0.3543307086614173, "grad_norm": 1.7363102436065674, "learning_rate": 0.00019748588857197174, "loss": 2.1865234375, "step": 810 }, { "ce_loss": 0.23046875, "distill_loss": 0.123046875, "epoch": 0.3543307086614173, "explain_loss": 1.3359375, "loss": 2.1875, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.0179443359375, "step": 810 }, { "ce_loss": 0.060791015625, "distill_loss": 0.1875, "epoch": 0.3543307086614173, "explain_loss": 1.078125, "loss": 1.984375, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.0203857421875, "step": 810 }, { "ce_loss": 0.56640625, "distill_loss": 0.1875, "epoch": 0.3543307086614173, "explain_loss": 0.984375, "loss": 2.15625, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.0257568359375, "step": 810 }, { "ce_loss": 0.035400390625, "distill_loss": 0.162109375, "epoch": 0.3543307086614173, "explain_loss": 1.421875, "loss": 2.0625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.01904296875, "step": 810 }, { "epoch": 0.35870516185476814, "grad_norm": 5.508694648742676, "learning_rate": 0.00019737730823232488, "loss": 2.1984375, "step": 820 }, { "ce_loss": 0.09619140625, "distill_loss": 0.11669921875, "epoch": 0.35870516185476814, "explain_loss": 1.359375, "loss": 2.25, "ref_ce_loss": 0.6796875, "stage_align_loss": 0.00897216796875, "step": 820 }, { "ce_loss": 0.08837890625, "distill_loss": 0.181640625, "epoch": 0.35870516185476814, "explain_loss": 1.1171875, "loss": 1.953125, "ref_ce_loss": 0.5625, "stage_align_loss": 0.00897216796875, "step": 820 }, { "ce_loss": 0.05908203125, "distill_loss": 0.1064453125, "epoch": 0.35870516185476814, "explain_loss": 1.3125, "loss": 2.09375, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.01507568359375, "step": 820 }, { "ce_loss": 0.30859375, "distill_loss": 0.1337890625, "epoch": 0.35870516185476814, "explain_loss": 0.921875, "loss": 1.8125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.01507568359375, "step": 820 }, { "epoch": 0.363079615048119, "grad_norm": 1.0342780351638794, "learning_rate": 0.000197266463624677, "loss": 2.20625, "step": 830 }, { "ce_loss": 0.19921875, "distill_loss": 0.11181640625, "epoch": 0.363079615048119, "explain_loss": 0.9375, "loss": 1.7890625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.01458740234375, "step": 830 }, { "ce_loss": 0.1767578125, "distill_loss": 0.1005859375, "epoch": 0.363079615048119, "explain_loss": 1.140625, "loss": 1.9375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.01312255859375, "step": 830 }, { "ce_loss": 0.47265625, "distill_loss": 0.10498046875, "epoch": 0.363079615048119, "explain_loss": 1.3671875, "loss": 2.34375, "ref_ce_loss": 0.375, "stage_align_loss": 0.0159912109375, "step": 830 }, { "ce_loss": 0.271484375, "distill_loss": 0.1279296875, "epoch": 0.363079615048119, "explain_loss": 1.015625, "loss": 1.9296875, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.017822265625, "step": 830 }, { "epoch": 0.3674540682414698, "grad_norm": 0.9518575072288513, "learning_rate": 0.0001971533573264449, "loss": 2.2419921875, "step": 840 }, { "ce_loss": 0.357421875, "distill_loss": 0.10693359375, "epoch": 0.3674540682414698, "explain_loss": 1.1328125, "loss": 2.21875, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.0179443359375, "step": 840 }, { "ce_loss": 0.0208740234375, "distill_loss": 0.103515625, "epoch": 0.3674540682414698, "explain_loss": 1.09375, "loss": 1.703125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.01544189453125, "step": 840 }, { "ce_loss": 0.07275390625, "distill_loss": 0.1025390625, "epoch": 0.3674540682414698, "explain_loss": 0.77734375, "loss": 1.4140625, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0194091796875, "step": 840 }, { "ce_loss": 0.0986328125, "distill_loss": 0.0947265625, "epoch": 0.3674540682414698, "explain_loss": 1.65625, "loss": 2.4375, "ref_ce_loss": 0.5625, "stage_align_loss": 0.015869140625, "step": 840 }, { "epoch": 0.37182852143482065, "grad_norm": 1.6504946947097778, "learning_rate": 0.00019703799196763515, "loss": 2.1138671875, "step": 850 }, { "ce_loss": 0.01275634765625, "distill_loss": 0.09765625, "epoch": 0.37182852143482065, "explain_loss": 1.5859375, "loss": 2.21875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0245361328125, "step": 850 }, { "ce_loss": 0.203125, "distill_loss": 0.0908203125, "epoch": 0.37182852143482065, "explain_loss": 1.125, "loss": 2.09375, "ref_ce_loss": 0.640625, "stage_align_loss": 0.03076171875, "step": 850 }, { "ce_loss": 0.78515625, "distill_loss": 0.103515625, "epoch": 0.37182852143482065, "explain_loss": 1.2109375, "loss": 2.71875, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.042236328125, "step": 850 }, { "ce_loss": 0.6875, "distill_loss": 0.103515625, "epoch": 0.37182852143482065, "explain_loss": 1.0390625, "loss": 2.4375, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.031982421875, "step": 850 }, { "epoch": 0.3762029746281715, "grad_norm": 4.888132095336914, "learning_rate": 0.0001969203702307834, "loss": 2.3873046875, "step": 860 }, { "ce_loss": 0.02001953125, "distill_loss": 0.103515625, "epoch": 0.3762029746281715, "explain_loss": 1.328125, "loss": 2.203125, "ref_ce_loss": 0.72265625, "stage_align_loss": 0.021728515625, "step": 860 }, { "ce_loss": 0.004364013671875, "distill_loss": 0.111328125, "epoch": 0.3762029746281715, "explain_loss": 1.5703125, "loss": 2.375, "ref_ce_loss": 0.671875, "stage_align_loss": 0.0206298828125, "step": 860 }, { "ce_loss": 0.0011444091796875, "distill_loss": 0.111328125, "epoch": 0.3762029746281715, "explain_loss": 1.25, "loss": 1.71875, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.0203857421875, "step": 860 }, { "ce_loss": 0.205078125, "distill_loss": 0.107421875, "epoch": 0.3762029746281715, "explain_loss": 1.671875, "loss": 2.5, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.04248046875, "step": 860 }, { "epoch": 0.3805774278215223, "grad_norm": 0.8729460835456848, "learning_rate": 0.0001968004948508916, "loss": 2.2201171875, "step": 870 }, { "ce_loss": 0.07373046875, "distill_loss": 0.1279296875, "epoch": 0.3805774278215223, "explain_loss": 1.2734375, "loss": 1.984375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.0274658203125, "step": 870 }, { "ce_loss": 0.68359375, "distill_loss": 0.1123046875, "epoch": 0.3805774278215223, "explain_loss": 1.171875, "loss": 2.59375, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.025634765625, "step": 870 }, { "ce_loss": 0.06494140625, "distill_loss": 0.111328125, "epoch": 0.3805774278215223, "explain_loss": 1.1171875, "loss": 1.75, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0267333984375, "step": 870 }, { "ce_loss": 0.48046875, "distill_loss": 0.115234375, "epoch": 0.3805774278215223, "explain_loss": 1.515625, "loss": 2.828125, "ref_ce_loss": 0.69140625, "stage_align_loss": 0.032958984375, "step": 870 }, { "epoch": 0.38495188101487315, "grad_norm": 0.7181954979896545, "learning_rate": 0.00019667836861536468, "loss": 2.1873046875, "step": 880 }, { "ce_loss": 1.6015625, "distill_loss": 0.12109375, "epoch": 0.38495188101487315, "explain_loss": 1.1875, "loss": 3.328125, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0225830078125, "step": 880 }, { "ce_loss": 0.07861328125, "distill_loss": 0.103515625, "epoch": 0.38495188101487315, "explain_loss": 1.171875, "loss": 1.765625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.019775390625, "step": 880 }, { "ce_loss": 0.1953125, "distill_loss": 0.091796875, "epoch": 0.38495188101487315, "explain_loss": 1.203125, "loss": 2.03125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.0194091796875, "step": 880 }, { "ce_loss": 0.45703125, "distill_loss": 0.09326171875, "epoch": 0.38495188101487315, "explain_loss": 1.21875, "loss": 2.3125, "ref_ce_loss": 0.53125, "stage_align_loss": 0.017822265625, "step": 880 }, { "epoch": 0.389326334208224, "grad_norm": 1.2711408138275146, "learning_rate": 0.00019655399436394565, "loss": 2.11591796875, "step": 890 }, { "ce_loss": 0.030517578125, "distill_loss": 0.0947265625, "epoch": 0.389326334208224, "explain_loss": 1.0703125, "loss": 1.703125, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.01806640625, "step": 890 }, { "ce_loss": 0.0556640625, "distill_loss": 0.0986328125, "epoch": 0.389326334208224, "explain_loss": 0.953125, "loss": 1.671875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.018310546875, "step": 890 }, { "ce_loss": 0.020751953125, "distill_loss": 0.083984375, "epoch": 0.389326334208224, "explain_loss": 1.21875, "loss": 1.7578125, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.01397705078125, "step": 890 }, { "ce_loss": 0.1376953125, "distill_loss": 0.091796875, "epoch": 0.389326334208224, "explain_loss": 1.234375, "loss": 1.84375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0152587890625, "step": 890 }, { "epoch": 0.3937007874015748, "grad_norm": 1.7024356126785278, "learning_rate": 0.00019642737498864947, "loss": 2.00712890625, "step": 900 }, { "ce_loss": 0.1298828125, "distill_loss": 0.10693359375, "epoch": 0.3937007874015748, "explain_loss": 0.953125, "loss": 1.4453125, "ref_ce_loss": 0.244140625, "stage_align_loss": 0.0125732421875, "step": 900 }, { "ce_loss": 1.125, "distill_loss": 0.09814453125, "epoch": 0.3937007874015748, "explain_loss": 1.0546875, "loss": 2.796875, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0294189453125, "step": 900 }, { "ce_loss": 0.05029296875, "distill_loss": 0.0830078125, "epoch": 0.3937007874015748, "explain_loss": 1.484375, "loss": 1.9609375, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.01263427734375, "step": 900 }, { "ce_loss": 0.020751953125, "distill_loss": 0.076171875, "epoch": 0.3937007874015748, "explain_loss": 1.21875, "loss": 1.6953125, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.01171875, "step": 900 }, { "epoch": 0.39807524059492566, "grad_norm": 0.8078297972679138, "learning_rate": 0.00019629851343369597, "loss": 2.12900390625, "step": 910 }, { "ce_loss": 0.76953125, "distill_loss": 0.076171875, "epoch": 0.39807524059492566, "explain_loss": 1.15625, "loss": 2.5625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.01287841796875, "step": 910 }, { "ce_loss": 0.486328125, "distill_loss": 0.0751953125, "epoch": 0.39807524059492566, "explain_loss": 0.6953125, "loss": 1.7109375, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.01446533203125, "step": 910 }, { "ce_loss": 0.65234375, "distill_loss": 0.0771484375, "epoch": 0.39807524059492566, "explain_loss": 0.9921875, "loss": 2.171875, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.0147705078125, "step": 910 }, { "ce_loss": 0.026123046875, "distill_loss": 0.07080078125, "epoch": 0.39807524059492566, "explain_loss": 1.234375, "loss": 1.953125, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.01019287109375, "step": 910 }, { "epoch": 0.40244969378827644, "grad_norm": 0.9429155588150024, "learning_rate": 0.00019616741269544126, "loss": 2.14130859375, "step": 920 }, { "ce_loss": 0.369140625, "distill_loss": 0.07666015625, "epoch": 0.40244969378827644, "explain_loss": 1.46875, "loss": 2.421875, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.015380859375, "step": 920 }, { "ce_loss": 0.2734375, "distill_loss": 0.0771484375, "epoch": 0.40244969378827644, "explain_loss": 1.2109375, "loss": 2.21875, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.0120849609375, "step": 920 }, { "ce_loss": 0.46484375, "distill_loss": 0.0751953125, "epoch": 0.40244969378827644, "explain_loss": 1.2109375, "loss": 2.375, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.01171875, "step": 920 }, { "ce_loss": 0.310546875, "distill_loss": 0.0693359375, "epoch": 0.40244969378827644, "explain_loss": 0.859375, "loss": 1.7109375, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0147705078125, "step": 920 }, { "epoch": 0.4068241469816273, "grad_norm": 1.3593651056289673, "learning_rate": 0.00019603407582230824, "loss": 2.059375, "step": 930 }, { "ce_loss": 0.29296875, "distill_loss": 0.0654296875, "epoch": 0.4068241469816273, "explain_loss": 0.8046875, "loss": 1.515625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.013671875, "step": 930 }, { "ce_loss": 0.7265625, "distill_loss": 0.07373046875, "epoch": 0.4068241469816273, "explain_loss": 1.3125, "loss": 2.5625, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.00982666015625, "step": 930 }, { "ce_loss": 0.052978515625, "distill_loss": 0.08544921875, "epoch": 0.4068241469816273, "explain_loss": 1.2265625, "loss": 1.7578125, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.01190185546875, "step": 930 }, { "ce_loss": 0.0299072265625, "distill_loss": 0.07275390625, "epoch": 0.4068241469816273, "explain_loss": 1.4375, "loss": 1.9765625, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0111083984375, "step": 930 }, { "epoch": 0.4111986001749781, "grad_norm": 0.9881080985069275, "learning_rate": 0.00019589850591471538, "loss": 2.09970703125, "step": 940 }, { "ce_loss": 0.330078125, "distill_loss": 0.1025390625, "epoch": 0.4111986001749781, "explain_loss": 0.83203125, "loss": 1.6796875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.01611328125, "step": 940 }, { "ce_loss": 0.318359375, "distill_loss": 0.1083984375, "epoch": 0.4111986001749781, "explain_loss": 1.265625, "loss": 2.171875, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.01318359375, "step": 940 }, { "ce_loss": 0.4765625, "distill_loss": 0.11328125, "epoch": 0.4111986001749781, "explain_loss": 1.03125, "loss": 2.140625, "ref_ce_loss": 0.5, "stage_align_loss": 0.01495361328125, "step": 940 }, { "ce_loss": 0.390625, "distill_loss": 0.349609375, "epoch": 0.4111986001749781, "explain_loss": 0.8125, "loss": 1.921875, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.016357421875, "step": 940 }, { "epoch": 0.41557305336832895, "grad_norm": 0.7105141282081604, "learning_rate": 0.000195760706125005, "loss": 2.1625, "step": 950 }, { "ce_loss": 0.2890625, "distill_loss": 0.091796875, "epoch": 0.41557305336832895, "explain_loss": 1.1171875, "loss": 2.09375, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.012939453125, "step": 950 }, { "ce_loss": 0.16015625, "distill_loss": 0.08984375, "epoch": 0.41557305336832895, "explain_loss": 1.2734375, "loss": 2.015625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.011474609375, "step": 950 }, { "ce_loss": 0.11328125, "distill_loss": 0.09033203125, "epoch": 0.41557305336832895, "explain_loss": 1.0546875, "loss": 1.9375, "ref_ce_loss": 0.66015625, "stage_align_loss": 0.01458740234375, "step": 950 }, { "ce_loss": 0.053466796875, "distill_loss": 0.0751953125, "epoch": 0.41557305336832895, "explain_loss": 0.875, "loss": 1.515625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0098876953125, "step": 950 }, { "epoch": 0.4199475065616798, "grad_norm": 0.6534632444381714, "learning_rate": 0.00019562067965736966, "loss": 2.034765625, "step": 960 }, { "ce_loss": 0.81640625, "distill_loss": 0.0703125, "epoch": 0.4199475065616798, "explain_loss": 1.375, "loss": 2.75, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.01129150390625, "step": 960 }, { "ce_loss": 0.185546875, "distill_loss": 0.0751953125, "epoch": 0.4199475065616798, "explain_loss": 1.2734375, "loss": 2.15625, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.010498046875, "step": 960 }, { "ce_loss": 0.01251220703125, "distill_loss": 0.0810546875, "epoch": 0.4199475065616798, "explain_loss": 1.3046875, "loss": 1.875, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.0089111328125, "step": 960 }, { "ce_loss": 0.11962890625, "distill_loss": 0.07275390625, "epoch": 0.4199475065616798, "explain_loss": 1.3125, "loss": 2.0625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0111083984375, "step": 960 }, { "epoch": 0.4243219597550306, "grad_norm": 0.8484123945236206, "learning_rate": 0.000195478429767778, "loss": 2.0353515625, "step": 970 }, { "ce_loss": 0.212890625, "distill_loss": 0.142578125, "epoch": 0.4243219597550306, "explain_loss": 1.5390625, "loss": 2.40625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.010498046875, "step": 970 }, { "ce_loss": 0.447265625, "distill_loss": 0.0693359375, "epoch": 0.4243219597550306, "explain_loss": 1.1328125, "loss": 2.3125, "ref_ce_loss": 0.640625, "stage_align_loss": 0.0181884765625, "step": 970 }, { "ce_loss": 0.404296875, "distill_loss": 0.07080078125, "epoch": 0.4243219597550306, "explain_loss": 1.09375, "loss": 2.15625, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.0184326171875, "step": 970 }, { "ce_loss": 0.00653076171875, "distill_loss": 0.076171875, "epoch": 0.4243219597550306, "explain_loss": 1.3125, "loss": 1.8828125, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.0081787109375, "step": 970 }, { "epoch": 0.42869641294838146, "grad_norm": 0.9739748239517212, "learning_rate": 0.0001953339597638986, "loss": 2.136328125, "step": 980 }, { "ce_loss": 0.875, "distill_loss": 0.08984375, "epoch": 0.42869641294838146, "explain_loss": 0.97265625, "loss": 2.53125, "ref_ce_loss": 0.5625, "stage_align_loss": 0.025146484375, "step": 980 }, { "ce_loss": 0.197265625, "distill_loss": 0.150390625, "epoch": 0.42869641294838146, "explain_loss": 1.375, "loss": 2.015625, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.009521484375, "step": 980 }, { "ce_loss": 0.50390625, "distill_loss": 0.1865234375, "epoch": 0.42869641294838146, "explain_loss": 0.796875, "loss": 1.8984375, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0255126953125, "step": 980 }, { "ce_loss": 0.072265625, "distill_loss": 0.07373046875, "epoch": 0.42869641294838146, "explain_loss": 1.359375, "loss": 1.8671875, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.0084228515625, "step": 980 }, { "epoch": 0.4330708661417323, "grad_norm": 0.6864160299301147, "learning_rate": 0.00019518727300502344, "loss": 2.06259765625, "step": 990 }, { "ce_loss": 0.0478515625, "distill_loss": 0.064453125, "epoch": 0.4330708661417323, "explain_loss": 1.0078125, "loss": 1.4765625, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.009765625, "step": 990 }, { "ce_loss": 0.138671875, "distill_loss": 0.0751953125, "epoch": 0.4330708661417323, "explain_loss": 1.234375, "loss": 1.890625, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.01177978515625, "step": 990 }, { "ce_loss": 0.11865234375, "distill_loss": 0.07275390625, "epoch": 0.4330708661417323, "explain_loss": 1.421875, "loss": 2.25, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.007476806640625, "step": 990 }, { "ce_loss": 0.330078125, "distill_loss": 0.0810546875, "epoch": 0.4330708661417323, "explain_loss": 1.28125, "loss": 2.234375, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.025146484375, "step": 990 }, { "epoch": 0.4374453193350831, "grad_norm": 0.47726380825042725, "learning_rate": 0.0001950383729019896, "loss": 2.0474609375, "step": 1000 }, { "ce_loss": 0.349609375, "distill_loss": 0.060546875, "epoch": 0.4374453193350831, "explain_loss": 1.2109375, "loss": 2.25, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.015625, "step": 1000 }, { "ce_loss": 0.1162109375, "distill_loss": 0.06591796875, "epoch": 0.4374453193350831, "explain_loss": 1.03125, "loss": 1.75, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.01116943359375, "step": 1000 }, { "ce_loss": 0.90625, "distill_loss": 0.15625, "epoch": 0.4374453193350831, "explain_loss": 1.234375, "loss": 2.8125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0277099609375, "step": 1000 }, { "ce_loss": 0.2490234375, "distill_loss": 0.06396484375, "epoch": 0.4374453193350831, "explain_loss": 1.03125, "loss": 2.03125, "ref_ce_loss": 0.671875, "stage_align_loss": 0.0118408203125, "step": 1000 }, { "epoch": 0.44181977252843396, "grad_norm": 0.7931193113327026, "learning_rate": 0.00019488726291710002, "loss": 2.02373046875, "step": 1010 }, { "ce_loss": 0.0269775390625, "distill_loss": 0.05859375, "epoch": 0.44181977252843396, "explain_loss": 1.34375, "loss": 2.03125, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.00787353515625, "step": 1010 }, { "ce_loss": 0.26953125, "distill_loss": 0.06298828125, "epoch": 0.44181977252843396, "explain_loss": 1.15625, "loss": 2.03125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.01055908203125, "step": 1010 }, { "ce_loss": 0.036376953125, "distill_loss": 0.06591796875, "epoch": 0.44181977252843396, "explain_loss": 1.5546875, "loss": 2.296875, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0106201171875, "step": 1010 }, { "ce_loss": 0.08837890625, "distill_loss": 0.060791015625, "epoch": 0.44181977252843396, "explain_loss": 1.1171875, "loss": 1.671875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.00982666015625, "step": 1010 }, { "epoch": 0.4461942257217848, "grad_norm": 0.9042448997497559, "learning_rate": 0.00019473394656404293, "loss": 2.10615234375, "step": 1020 }, { "ce_loss": 0.67578125, "distill_loss": 0.07421875, "epoch": 0.4461942257217848, "explain_loss": 1.25, "loss": 2.453125, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0654296875, "step": 1020 }, { "ce_loss": 0.4296875, "distill_loss": 0.11962890625, "epoch": 0.4461942257217848, "explain_loss": 1.2890625, "loss": 2.3125, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.01251220703125, "step": 1020 }, { "ce_loss": 0.02392578125, "distill_loss": 0.078125, "epoch": 0.4461942257217848, "explain_loss": 1.703125, "loss": 2.25, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.01141357421875, "step": 1020 }, { "ce_loss": 0.380859375, "distill_loss": 0.11474609375, "epoch": 0.4461942257217848, "explain_loss": 0.796875, "loss": 1.65625, "ref_ce_loss": 0.3125, "stage_align_loss": 0.05517578125, "step": 1020 }, { "epoch": 0.45056867891513563, "grad_norm": 0.7094565629959106, "learning_rate": 0.0001945784274078102, "loss": 2.03427734375, "step": 1030 }, { "ce_loss": 0.66796875, "distill_loss": 0.11181640625, "epoch": 0.45056867891513563, "explain_loss": 1.1796875, "loss": 2.4375, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.037353515625, "step": 1030 }, { "ce_loss": 0.34375, "distill_loss": 0.107421875, "epoch": 0.45056867891513563, "explain_loss": 0.87890625, "loss": 1.78125, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0123291015625, "step": 1030 }, { "ce_loss": 0.70703125, "distill_loss": 0.0751953125, "epoch": 0.45056867891513563, "explain_loss": 0.75390625, "loss": 2.078125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0107421875, "step": 1030 }, { "ce_loss": 0.056640625, "distill_loss": 0.0791015625, "epoch": 0.45056867891513563, "explain_loss": 1.1796875, "loss": 1.9453125, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.026123046875, "step": 1030 }, { "epoch": 0.4549431321084864, "grad_norm": 0.8471792340278625, "learning_rate": 0.00019442070906461445, "loss": 2.05712890625, "step": 1040 }, { "ce_loss": 0.5703125, "distill_loss": 0.10009765625, "epoch": 0.4549431321084864, "explain_loss": 1.015625, "loss": 1.96875, "ref_ce_loss": 0.2255859375, "stage_align_loss": 0.049560546875, "step": 1040 }, { "ce_loss": 0.1416015625, "distill_loss": 0.10009765625, "epoch": 0.4549431321084864, "explain_loss": 0.92578125, "loss": 1.7421875, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01275634765625, "step": 1040 }, { "ce_loss": 0.107421875, "distill_loss": 0.0986328125, "epoch": 0.4549431321084864, "explain_loss": 1.234375, "loss": 1.84375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.009765625, "step": 1040 }, { "ce_loss": 0.359375, "distill_loss": 0.09814453125, "epoch": 0.4549431321084864, "explain_loss": 0.921875, "loss": 1.8046875, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.03173828125, "step": 1040 }, { "epoch": 0.45931758530183725, "grad_norm": 0.8587881326675415, "learning_rate": 0.0001942607952018049, "loss": 2.12333984375, "step": 1050 }, { "ce_loss": 0.00860595703125, "distill_loss": 0.07763671875, "epoch": 0.45931758530183725, "explain_loss": 0.96875, "loss": 1.3984375, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.00823974609375, "step": 1050 }, { "ce_loss": 0.078125, "distill_loss": 0.08251953125, "epoch": 0.45931758530183725, "explain_loss": 1.15625, "loss": 1.7421875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.00726318359375, "step": 1050 }, { "ce_loss": 0.73046875, "distill_loss": 0.07177734375, "epoch": 0.45931758530183725, "explain_loss": 1.4921875, "loss": 2.78125, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.03369140625, "step": 1050 }, { "ce_loss": 0.50390625, "distill_loss": 0.0810546875, "epoch": 0.45931758530183725, "explain_loss": 1.171875, "loss": 2.25, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0230712890625, "step": 1050 }, { "epoch": 0.4636920384951881, "grad_norm": 0.7530121207237244, "learning_rate": 0.00019409868953778218, "loss": 2.07724609375, "step": 1060 }, { "ce_loss": 0.021240234375, "distill_loss": 0.087890625, "epoch": 0.4636920384951881, "explain_loss": 1.2109375, "loss": 1.703125, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.01055908203125, "step": 1060 }, { "ce_loss": 0.2294921875, "distill_loss": 0.0859375, "epoch": 0.4636920384951881, "explain_loss": 1.234375, "loss": 1.9453125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.02294921875, "step": 1060 }, { "ce_loss": 0.0084228515625, "distill_loss": 0.11328125, "epoch": 0.4636920384951881, "explain_loss": 1.40625, "loss": 1.9296875, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0120849609375, "step": 1060 }, { "ce_loss": 0.271484375, "distill_loss": 0.1435546875, "epoch": 0.4636920384951881, "explain_loss": 1.21875, "loss": 2.078125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.0201416015625, "step": 1060 }, { "epoch": 0.4680664916885389, "grad_norm": 1.068654179573059, "learning_rate": 0.00019393439584191176, "loss": 2.078125, "step": 1070 }, { "ce_loss": 0.5859375, "distill_loss": 0.091796875, "epoch": 0.4680664916885389, "explain_loss": 1.1484375, "loss": 2.375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.01123046875, "step": 1070 }, { "ce_loss": 0.2109375, "distill_loss": 0.0869140625, "epoch": 0.4680664916885389, "explain_loss": 1.265625, "loss": 2.09375, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.009033203125, "step": 1070 }, { "ce_loss": 0.267578125, "distill_loss": 0.0810546875, "epoch": 0.4680664916885389, "explain_loss": 1.1953125, "loss": 1.9375, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.012451171875, "step": 1070 }, { "ce_loss": 0.07861328125, "distill_loss": 0.099609375, "epoch": 0.4680664916885389, "explain_loss": 0.9296875, "loss": 1.6640625, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0172119140625, "step": 1070 }, { "epoch": 0.47244094488188976, "grad_norm": 0.9611126780509949, "learning_rate": 0.00019376791793443646, "loss": 2.02275390625, "step": 1080 }, { "ce_loss": 0.38671875, "distill_loss": 0.08056640625, "epoch": 0.47244094488188976, "explain_loss": 1.21875, "loss": 2.28125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.052001953125, "step": 1080 }, { "ce_loss": 0.09130859375, "distill_loss": 0.07666015625, "epoch": 0.47244094488188976, "explain_loss": 0.7578125, "loss": 1.2578125, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.042236328125, "step": 1080 }, { "ce_loss": 0.44140625, "distill_loss": 0.0791015625, "epoch": 0.47244094488188976, "explain_loss": 1.3984375, "loss": 2.3125, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.01953125, "step": 1080 }, { "ce_loss": 0.00848388671875, "distill_loss": 0.07763671875, "epoch": 0.47244094488188976, "explain_loss": 1.453125, "loss": 1.9765625, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.00946044921875, "step": 1080 }, { "epoch": 0.4768153980752406, "grad_norm": 0.9106386303901672, "learning_rate": 0.0001935992596863875, "loss": 2.04150390625, "step": 1090 }, { "ce_loss": 0.19921875, "distill_loss": 0.08837890625, "epoch": 0.4768153980752406, "explain_loss": 1.203125, "loss": 2.1875, "ref_ce_loss": 0.6796875, "stage_align_loss": 0.01043701171875, "step": 1090 }, { "ce_loss": 0.033203125, "distill_loss": 0.0751953125, "epoch": 0.4768153980752406, "explain_loss": 1.3828125, "loss": 1.8671875, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.0087890625, "step": 1090 }, { "ce_loss": 1.3046875, "distill_loss": 0.07666015625, "epoch": 0.4768153980752406, "explain_loss": 1.4375, "loss": 3.703125, "ref_ce_loss": 0.81640625, "stage_align_loss": 0.062255859375, "step": 1090 }, { "ce_loss": 0.0128173828125, "distill_loss": 0.087890625, "epoch": 0.4768153980752406, "explain_loss": 1.53125, "loss": 2.1875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.038818359375, "step": 1090 }, { "epoch": 0.48118985126859143, "grad_norm": 1.3736907243728638, "learning_rate": 0.00019342842501949447, "loss": 2.0861328125, "step": 1100 }, { "ce_loss": 0.1640625, "distill_loss": 0.06201171875, "epoch": 0.48118985126859143, "explain_loss": 1.5625, "loss": 2.28125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.009765625, "step": 1100 }, { "ce_loss": 0.06005859375, "distill_loss": 0.0634765625, "epoch": 0.48118985126859143, "explain_loss": 1.109375, "loss": 1.6640625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.01019287109375, "step": 1100 }, { "ce_loss": 0.041748046875, "distill_loss": 0.08251953125, "epoch": 0.48118985126859143, "explain_loss": 0.984375, "loss": 1.625, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.007049560546875, "step": 1100 }, { "ce_loss": 0.017822265625, "distill_loss": 0.0751953125, "epoch": 0.48118985126859143, "explain_loss": 0.984375, "loss": 1.4765625, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.00640869140625, "step": 1100 }, { "epoch": 0.48556430446194226, "grad_norm": 0.5742623209953308, "learning_rate": 0.00019325541790609423, "loss": 1.9087890625, "step": 1110 }, { "ce_loss": 0.12255859375, "distill_loss": 0.08984375, "epoch": 0.48556430446194226, "explain_loss": 1.125, "loss": 1.9296875, "ref_ce_loss": 0.578125, "stage_align_loss": 0.01556396484375, "step": 1110 }, { "ce_loss": 0.78125, "distill_loss": 0.08251953125, "epoch": 0.48556430446194226, "explain_loss": 1.5234375, "loss": 3.1875, "ref_ce_loss": 0.7734375, "stage_align_loss": 0.0240478515625, "step": 1110 }, { "ce_loss": 0.04443359375, "distill_loss": 0.087890625, "epoch": 0.48556430446194226, "explain_loss": 1.21875, "loss": 1.75, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.007049560546875, "step": 1110 }, { "ce_loss": 0.38671875, "distill_loss": 0.0751953125, "epoch": 0.48556430446194226, "explain_loss": 1.7109375, "loss": 2.71875, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.02880859375, "step": 1110 }, { "epoch": 0.4899387576552931, "grad_norm": 1.206512212753296, "learning_rate": 0.00019308024236903846, "loss": 2.003125, "step": 1120 }, { "ce_loss": 0.26953125, "distill_loss": 0.0673828125, "epoch": 0.4899387576552931, "explain_loss": 1.5859375, "loss": 2.46875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.01318359375, "step": 1120 }, { "ce_loss": 0.01708984375, "distill_loss": 0.0830078125, "epoch": 0.4899387576552931, "explain_loss": 1.109375, "loss": 1.671875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0128173828125, "step": 1120 }, { "ce_loss": 0.51171875, "distill_loss": 0.0615234375, "epoch": 0.4899387576552931, "explain_loss": 1.1640625, "loss": 2.3125, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0303955078125, "step": 1120 }, { "ce_loss": 0.11962890625, "distill_loss": 0.1123046875, "epoch": 0.4899387576552931, "explain_loss": 0.828125, "loss": 1.53125, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.021728515625, "step": 1120 }, { "epoch": 0.49431321084864394, "grad_norm": 0.5838649868965149, "learning_rate": 0.00019290290248160022, "loss": 1.9322265625, "step": 1130 }, { "ce_loss": 0.2373046875, "distill_loss": 0.0595703125, "epoch": 0.49431321084864394, "explain_loss": 1.171875, "loss": 1.9921875, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.119140625, "step": 1130 }, { "ce_loss": 0.18359375, "distill_loss": 0.06689453125, "epoch": 0.49431321084864394, "explain_loss": 1.59375, "loss": 2.375, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.091796875, "step": 1130 }, { "ce_loss": 0.03759765625, "distill_loss": 0.05810546875, "epoch": 0.49431321084864394, "explain_loss": 0.95703125, "loss": 1.640625, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.0147705078125, "step": 1130 }, { "ce_loss": 0.373046875, "distill_loss": 0.076171875, "epoch": 0.49431321084864394, "explain_loss": 1.453125, "loss": 2.5625, "ref_ce_loss": 0.625, "stage_align_loss": 0.037841796875, "step": 1130 }, { "epoch": 0.49868766404199477, "grad_norm": 0.559495747089386, "learning_rate": 0.00019272340236737917, "loss": 2.18037109375, "step": 1140 }, { "ce_loss": 0.185546875, "distill_loss": 0.0595703125, "epoch": 0.49868766404199477, "explain_loss": 0.97265625, "loss": 1.671875, "ref_ce_loss": 0.390625, "stage_align_loss": 0.06298828125, "step": 1140 }, { "ce_loss": 0.296875, "distill_loss": 0.05908203125, "epoch": 0.49868766404199477, "explain_loss": 0.94921875, "loss": 1.703125, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.06689453125, "step": 1140 }, { "ce_loss": 0.90234375, "distill_loss": 0.061279296875, "epoch": 0.49868766404199477, "explain_loss": 1.2109375, "loss": 2.5625, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.07470703125, "step": 1140 }, { "ce_loss": 0.482421875, "distill_loss": 0.0576171875, "epoch": 0.49868766404199477, "explain_loss": 1.171875, "loss": 2.234375, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.09765625, "step": 1140 }, { "epoch": 0.5030621172353456, "grad_norm": 0.6255845427513123, "learning_rate": 0.0001925417462002056, "loss": 1.93876953125, "step": 1150 }, { "ce_loss": 0.033935546875, "distill_loss": 0.05712890625, "epoch": 0.5030621172353456, "explain_loss": 0.70703125, "loss": 1.2734375, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.006500244140625, "step": 1150 }, { "ce_loss": 0.12353515625, "distill_loss": 0.0556640625, "epoch": 0.5030621172353456, "explain_loss": 1.28125, "loss": 1.90625, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.0167236328125, "step": 1150 }, { "ce_loss": 0.4765625, "distill_loss": 0.0625, "epoch": 0.5030621172353456, "explain_loss": 1.1875, "loss": 2.40625, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.0159912109375, "step": 1150 }, { "ce_loss": 0.030029296875, "distill_loss": 0.05029296875, "epoch": 0.5030621172353456, "explain_loss": 1.40625, "loss": 1.890625, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0155029296875, "step": 1150 }, { "epoch": 0.5074365704286964, "grad_norm": 0.6101519465446472, "learning_rate": 0.00019235793820404355, "loss": 2.04638671875, "step": 1160 }, { "ce_loss": 0.1787109375, "distill_loss": 0.061279296875, "epoch": 0.5074365704286964, "explain_loss": 1.203125, "loss": 1.9375, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.017333984375, "step": 1160 }, { "ce_loss": 0.0157470703125, "distill_loss": 0.057373046875, "epoch": 0.5074365704286964, "explain_loss": 1.2265625, "loss": 1.84375, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.00738525390625, "step": 1160 }, { "ce_loss": 0.08251953125, "distill_loss": 0.06884765625, "epoch": 0.5074365704286964, "explain_loss": 1.03125, "loss": 1.6796875, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.007293701171875, "step": 1160 }, { "ce_loss": 0.470703125, "distill_loss": 0.0634765625, "epoch": 0.5074365704286964, "explain_loss": 1.25, "loss": 2.3125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.038818359375, "step": 1160 }, { "epoch": 0.5118110236220472, "grad_norm": 0.7154414653778076, "learning_rate": 0.0001921719826528925, "loss": 1.991796875, "step": 1170 }, { "ce_loss": 0.671875, "distill_loss": 0.058349609375, "epoch": 0.5118110236220472, "explain_loss": 1.109375, "loss": 2.3125, "ref_ce_loss": 0.453125, "stage_align_loss": 0.007598876953125, "step": 1170 }, { "ce_loss": 0.0849609375, "distill_loss": 0.051025390625, "epoch": 0.5118110236220472, "explain_loss": 1.359375, "loss": 2.0, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.00830078125, "step": 1170 }, { "ce_loss": 0.0703125, "distill_loss": 0.045166015625, "epoch": 0.5118110236220472, "explain_loss": 1.0859375, "loss": 1.625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.005035400390625, "step": 1170 }, { "ce_loss": 0.024169921875, "distill_loss": 0.054931640625, "epoch": 0.5118110236220472, "explain_loss": 1.015625, "loss": 1.6953125, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.004608154296875, "step": 1170 }, { "epoch": 0.5161854768153981, "grad_norm": 0.7044707536697388, "learning_rate": 0.00019198388387068802, "loss": 1.9783203125, "step": 1180 }, { "ce_loss": 0.259765625, "distill_loss": 0.078125, "epoch": 0.5161854768153981, "explain_loss": 0.89453125, "loss": 1.75, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0062255859375, "step": 1180 }, { "ce_loss": 0.671875, "distill_loss": 0.0595703125, "epoch": 0.5161854768153981, "explain_loss": 1.40625, "loss": 2.5, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.00726318359375, "step": 1180 }, { "ce_loss": 0.03515625, "distill_loss": 0.06103515625, "epoch": 0.5161854768153981, "explain_loss": 1.484375, "loss": 1.96875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00634765625, "step": 1180 }, { "ce_loss": 1.78125, "distill_loss": 0.0498046875, "epoch": 0.5161854768153981, "explain_loss": 1.15625, "loss": 3.546875, "ref_ce_loss": 0.546875, "stage_align_loss": 0.01153564453125, "step": 1180 }, { "epoch": 0.5205599300087489, "grad_norm": 0.6813885569572449, "learning_rate": 0.00019179364623120114, "loss": 2.0076171875, "step": 1190 }, { "ce_loss": 0.01385498046875, "distill_loss": 0.0654296875, "epoch": 0.5205599300087489, "explain_loss": 1.1640625, "loss": 1.640625, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.004730224609375, "step": 1190 }, { "ce_loss": 0.30078125, "distill_loss": 0.056640625, "epoch": 0.5205599300087489, "explain_loss": 0.890625, "loss": 1.75, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.00634765625, "step": 1190 }, { "ce_loss": 0.1865234375, "distill_loss": 0.056396484375, "epoch": 0.5205599300087489, "explain_loss": 1.109375, "loss": 1.7890625, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0107421875, "step": 1190 }, { "ce_loss": 0.11767578125, "distill_loss": 0.04541015625, "epoch": 0.5205599300087489, "explain_loss": 0.9140625, "loss": 1.390625, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.0052490234375, "step": 1190 }, { "epoch": 0.5249343832020997, "grad_norm": 0.9167773723602295, "learning_rate": 0.00019160127415793676, "loss": 1.9306640625, "step": 1200 }, { "ce_loss": 0.240234375, "distill_loss": 0.058349609375, "epoch": 0.5249343832020997, "explain_loss": 1.40625, "loss": 2.34375, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.00677490234375, "step": 1200 }, { "ce_loss": 0.390625, "distill_loss": 0.05126953125, "epoch": 0.5249343832020997, "explain_loss": 1.2421875, "loss": 2.171875, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.0220947265625, "step": 1200 }, { "ce_loss": 1.1953125, "distill_loss": 0.058837890625, "epoch": 0.5249343832020997, "explain_loss": 1.046875, "loss": 2.828125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.03955078125, "step": 1200 }, { "ce_loss": 0.1591796875, "distill_loss": 0.0595703125, "epoch": 0.5249343832020997, "explain_loss": 0.9765625, "loss": 1.734375, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.01904296875, "step": 1200 }, { "epoch": 0.5293088363954506, "grad_norm": 0.7403603196144104, "learning_rate": 0.00019140677212403079, "loss": 1.9642578125, "step": 1210 }, { "ce_loss": 0.248046875, "distill_loss": 0.06591796875, "epoch": 0.5293088363954506, "explain_loss": 1.171875, "loss": 2.15625, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.017578125, "step": 1210 }, { "ce_loss": 0.28125, "distill_loss": 0.05029296875, "epoch": 0.5293088363954506, "explain_loss": 1.40625, "loss": 2.203125, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.01104736328125, "step": 1210 }, { "ce_loss": 0.5, "distill_loss": 0.047607421875, "epoch": 0.5293088363954506, "explain_loss": 1.109375, "loss": 2.234375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0274658203125, "step": 1210 }, { "ce_loss": 0.1845703125, "distill_loss": 0.05712890625, "epoch": 0.5293088363954506, "explain_loss": 1.34375, "loss": 2.046875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.01434326171875, "step": 1210 }, { "epoch": 0.5336832895888014, "grad_norm": 0.6152716279029846, "learning_rate": 0.00019121014465214598, "loss": 1.984765625, "step": 1220 }, { "ce_loss": 1.3125, "distill_loss": 0.0537109375, "epoch": 0.5336832895888014, "explain_loss": 1.25, "loss": 3.21875, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0634765625, "step": 1220 }, { "ce_loss": 0.04931640625, "distill_loss": 0.0439453125, "epoch": 0.5336832895888014, "explain_loss": 1.3671875, "loss": 1.859375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.00872802734375, "step": 1220 }, { "ce_loss": 0.05322265625, "distill_loss": 0.0517578125, "epoch": 0.5336832895888014, "explain_loss": 1.1875, "loss": 1.9296875, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.01470947265625, "step": 1220 }, { "ce_loss": 0.154296875, "distill_loss": 0.05419921875, "epoch": 0.5336832895888014, "explain_loss": 1.421875, "loss": 2.125, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.00823974609375, "step": 1220 }, { "epoch": 0.5380577427821522, "grad_norm": 0.7350424528121948, "learning_rate": 0.00019101139631436698, "loss": 2.04033203125, "step": 1230 }, { "ce_loss": 0.06982421875, "distill_loss": 0.060546875, "epoch": 0.5380577427821522, "explain_loss": 1.203125, "loss": 1.8359375, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.007110595703125, "step": 1230 }, { "ce_loss": 0.0625, "distill_loss": 0.0673828125, "epoch": 0.5380577427821522, "explain_loss": 1.03125, "loss": 1.5625, "ref_ce_loss": 0.390625, "stage_align_loss": 0.006683349609375, "step": 1230 }, { "ce_loss": 0.298828125, "distill_loss": 0.05322265625, "epoch": 0.5380577427821522, "explain_loss": 0.921875, "loss": 1.6640625, "ref_ce_loss": 0.375, "stage_align_loss": 0.015625, "step": 1230 }, { "ce_loss": 0.345703125, "distill_loss": 0.047607421875, "epoch": 0.5380577427821522, "explain_loss": 0.921875, "loss": 1.671875, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.02734375, "step": 1230 }, { "epoch": 0.5424321959755031, "grad_norm": 1.0593407154083252, "learning_rate": 0.0001908105317320939, "loss": 1.9806640625, "step": 1240 }, { "ce_loss": 0.71875, "distill_loss": 0.07177734375, "epoch": 0.5424321959755031, "explain_loss": 1.125, "loss": 2.5625, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0130615234375, "step": 1240 }, { "ce_loss": 0.62890625, "distill_loss": 0.0908203125, "epoch": 0.5424321959755031, "explain_loss": 1.1953125, "loss": 2.4375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0169677734375, "step": 1240 }, { "ce_loss": 0.376953125, "distill_loss": 0.0595703125, "epoch": 0.5424321959755031, "explain_loss": 1.5703125, "loss": 2.53125, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.048583984375, "step": 1240 }, { "ce_loss": 0.412109375, "distill_loss": 0.0673828125, "epoch": 0.5424321959755031, "explain_loss": 1.359375, "loss": 2.4375, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.031982421875, "step": 1240 }, { "epoch": 0.5468066491688539, "grad_norm": 0.6777040958404541, "learning_rate": 0.00019060755557593479, "loss": 2.1650390625, "step": 1250 }, { "ce_loss": 0.375, "distill_loss": 0.087890625, "epoch": 0.5468066491688539, "explain_loss": 1.0390625, "loss": 2.03125, "ref_ce_loss": 0.515625, "stage_align_loss": 0.0238037109375, "step": 1250 }, { "ce_loss": 1.1953125, "distill_loss": 0.06298828125, "epoch": 0.5468066491688539, "explain_loss": 1.296875, "loss": 3.21875, "ref_ce_loss": 0.640625, "stage_align_loss": 0.0159912109375, "step": 1250 }, { "ce_loss": 0.037353515625, "distill_loss": 0.06787109375, "epoch": 0.5468066491688539, "explain_loss": 1.0703125, "loss": 1.4609375, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.007293701171875, "step": 1250 }, { "ce_loss": 0.0024871826171875, "distill_loss": 0.1162109375, "epoch": 0.5468066491688539, "explain_loss": 1.15625, "loss": 1.703125, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.00506591796875, "step": 1250 }, { "epoch": 0.5511811023622047, "grad_norm": 1.2155237197875977, "learning_rate": 0.00019040247256559727, "loss": 2.03095703125, "step": 1260 }, { "ce_loss": 0.314453125, "distill_loss": 0.06689453125, "epoch": 0.5511811023622047, "explain_loss": 1.1875, "loss": 2.0625, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.02197265625, "step": 1260 }, { "ce_loss": 0.171875, "distill_loss": 0.08740234375, "epoch": 0.5511811023622047, "explain_loss": 1.265625, "loss": 2.15625, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.0145263671875, "step": 1260 }, { "ce_loss": 0.09375, "distill_loss": 0.06005859375, "epoch": 0.5511811023622047, "explain_loss": 1.2421875, "loss": 1.984375, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.01324462890625, "step": 1260 }, { "ce_loss": 0.0576171875, "distill_loss": 0.0576171875, "epoch": 0.5511811023622047, "explain_loss": 1.3984375, "loss": 1.890625, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.00714111328125, "step": 1260 }, { "epoch": 0.5555555555555556, "grad_norm": 0.6100040078163147, "learning_rate": 0.0001901952874697785, "loss": 2.12236328125, "step": 1270 }, { "ce_loss": 0.19140625, "distill_loss": 0.0654296875, "epoch": 0.5555555555555556, "explain_loss": 1.171875, "loss": 1.75, "ref_ce_loss": 0.302734375, "stage_align_loss": 0.01031494140625, "step": 1270 }, { "ce_loss": 0.1650390625, "distill_loss": 0.0576171875, "epoch": 0.5555555555555556, "explain_loss": 1.2421875, "loss": 1.9375, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0081787109375, "step": 1270 }, { "ce_loss": 0.2578125, "distill_loss": 0.0732421875, "epoch": 0.5555555555555556, "explain_loss": 0.9921875, "loss": 1.8125, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0147705078125, "step": 1270 }, { "ce_loss": 0.169921875, "distill_loss": 0.0654296875, "epoch": 0.5555555555555556, "explain_loss": 0.984375, "loss": 1.5703125, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.006500244140625, "step": 1270 }, { "epoch": 0.5599300087489064, "grad_norm": 0.6778023838996887, "learning_rate": 0.0001899860051060545, "loss": 1.96416015625, "step": 1280 }, { "ce_loss": 0.0145263671875, "distill_loss": 0.0634765625, "epoch": 0.5599300087489064, "explain_loss": 1.375, "loss": 1.8359375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.007415771484375, "step": 1280 }, { "ce_loss": 0.037841796875, "distill_loss": 0.07177734375, "epoch": 0.5599300087489064, "explain_loss": 1.1328125, "loss": 1.734375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.00921630859375, "step": 1280 }, { "ce_loss": 0.0289306640625, "distill_loss": 0.06396484375, "epoch": 0.5599300087489064, "explain_loss": 1.125, "loss": 1.75, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.005645751953125, "step": 1280 }, { "ce_loss": 0.0283203125, "distill_loss": 0.05859375, "epoch": 0.5599300087489064, "explain_loss": 1.4140625, "loss": 2.03125, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.007598876953125, "step": 1280 }, { "epoch": 0.5643044619422573, "grad_norm": 1.0976567268371582, "learning_rate": 0.000189774630340768, "loss": 1.941796875, "step": 1290 }, { "ce_loss": 0.2734375, "distill_loss": 0.05712890625, "epoch": 0.5643044619422573, "explain_loss": 0.6953125, "loss": 1.609375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.04150390625, "step": 1290 }, { "ce_loss": 0.034912109375, "distill_loss": 0.05126953125, "epoch": 0.5643044619422573, "explain_loss": 1.5078125, "loss": 2.03125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.007354736328125, "step": 1290 }, { "ce_loss": 0.046875, "distill_loss": 0.046630859375, "epoch": 0.5643044619422573, "explain_loss": 0.9609375, "loss": 1.578125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0089111328125, "step": 1290 }, { "ce_loss": 0.1484375, "distill_loss": 0.0673828125, "epoch": 0.5643044619422573, "explain_loss": 0.8828125, "loss": 1.453125, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0126953125, "step": 1290 }, { "epoch": 0.5686789151356081, "grad_norm": 0.9156106114387512, "learning_rate": 0.0001895611680889154, "loss": 1.9869140625, "step": 1300 }, { "ce_loss": 1.296875, "distill_loss": 0.0537109375, "epoch": 0.5686789151356081, "explain_loss": 1.25, "loss": 3.125, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0252685546875, "step": 1300 }, { "ce_loss": 0.044189453125, "distill_loss": 0.064453125, "epoch": 0.5686789151356081, "explain_loss": 1.4375, "loss": 2.046875, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.008056640625, "step": 1300 }, { "ce_loss": 0.0400390625, "distill_loss": 0.058349609375, "epoch": 0.5686789151356081, "explain_loss": 1.1328125, "loss": 1.546875, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.0140380859375, "step": 1300 }, { "ce_loss": 0.03515625, "distill_loss": 0.05615234375, "epoch": 0.5686789151356081, "explain_loss": 1.5, "loss": 2.0, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.005340576171875, "step": 1300 }, { "epoch": 0.5730533683289589, "grad_norm": 0.7551911473274231, "learning_rate": 0.00018934562331403244, "loss": 2.00400390625, "step": 1310 }, { "ce_loss": 0.07421875, "distill_loss": 0.07275390625, "epoch": 0.5730533683289589, "explain_loss": 0.9609375, "loss": 1.421875, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.00531005859375, "step": 1310 }, { "ce_loss": 0.365234375, "distill_loss": 0.0546875, "epoch": 0.5730533683289589, "explain_loss": 0.9296875, "loss": 1.953125, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.022705078125, "step": 1310 }, { "ce_loss": 0.33984375, "distill_loss": 0.05810546875, "epoch": 0.5730533683289589, "explain_loss": 1.3671875, "loss": 2.109375, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.00762939453125, "step": 1310 }, { "ce_loss": 0.181640625, "distill_loss": 0.058837890625, "epoch": 0.5730533683289589, "explain_loss": 1.015625, "loss": 1.75, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.010986328125, "step": 1310 }, { "epoch": 0.5774278215223098, "grad_norm": 0.642061710357666, "learning_rate": 0.00018912800102807866, "loss": 1.95107421875, "step": 1320 }, { "ce_loss": 0.7890625, "distill_loss": 0.06298828125, "epoch": 0.5774278215223098, "explain_loss": 1.25, "loss": 2.46875, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0111083984375, "step": 1320 }, { "ce_loss": 0.404296875, "distill_loss": 0.06298828125, "epoch": 0.5774278215223098, "explain_loss": 1.515625, "loss": 2.46875, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0244140625, "step": 1320 }, { "ce_loss": 0.73046875, "distill_loss": 0.07666015625, "epoch": 0.5774278215223098, "explain_loss": 1.1171875, "loss": 2.375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.052490234375, "step": 1320 }, { "ce_loss": 0.1611328125, "distill_loss": 0.060791015625, "epoch": 0.5774278215223098, "explain_loss": 1.296875, "loss": 2.015625, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.03125, "step": 1320 }, { "epoch": 0.5818022747156606, "grad_norm": 0.5706647038459778, "learning_rate": 0.0001889083062913211, "loss": 2.004296875, "step": 1330 }, { "ce_loss": 0.7734375, "distill_loss": 0.103515625, "epoch": 0.5818022747156606, "explain_loss": 1.234375, "loss": 2.609375, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0439453125, "step": 1330 }, { "ce_loss": 0.400390625, "distill_loss": 0.0595703125, "epoch": 0.5818022747156606, "explain_loss": 1.578125, "loss": 2.515625, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.06298828125, "step": 1330 }, { "ce_loss": 0.052490234375, "distill_loss": 0.06689453125, "epoch": 0.5818022747156606, "explain_loss": 1.015625, "loss": 1.671875, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.01544189453125, "step": 1330 }, { "ce_loss": 0.08740234375, "distill_loss": 0.0615234375, "epoch": 0.5818022747156606, "explain_loss": 0.8828125, "loss": 1.609375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0400390625, "step": 1330 }, { "epoch": 0.5861767279090113, "grad_norm": 0.8078960180282593, "learning_rate": 0.00018868654421221644, "loss": 1.88642578125, "step": 1340 }, { "ce_loss": 0.330078125, "distill_loss": 0.051025390625, "epoch": 0.5861767279090113, "explain_loss": 1.359375, "loss": 2.28125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.007476806640625, "step": 1340 }, { "ce_loss": 1.7421875, "distill_loss": 0.057373046875, "epoch": 0.5861767279090113, "explain_loss": 1.3515625, "loss": 3.75, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0712890625, "step": 1340 }, { "ce_loss": 0.01092529296875, "distill_loss": 0.0537109375, "epoch": 0.5861767279090113, "explain_loss": 1.2890625, "loss": 1.953125, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.005340576171875, "step": 1340 }, { "ce_loss": 1.1484375, "distill_loss": 0.04931640625, "epoch": 0.5861767279090113, "explain_loss": 1.1015625, "loss": 3.09375, "ref_ce_loss": 0.70703125, "stage_align_loss": 0.07861328125, "step": 1340 }, { "epoch": 0.5905511811023622, "grad_norm": 0.8823025822639465, "learning_rate": 0.00018846271994729235, "loss": 2.00361328125, "step": 1350 }, { "ce_loss": 0.0166015625, "distill_loss": 0.062255859375, "epoch": 0.5905511811023622, "explain_loss": 1.3359375, "loss": 1.859375, "ref_ce_loss": 0.4375, "stage_align_loss": 0.00408935546875, "step": 1350 }, { "ce_loss": 0.34765625, "distill_loss": 0.0517578125, "epoch": 0.5905511811023622, "explain_loss": 1.203125, "loss": 1.953125, "ref_ce_loss": 0.34375, "stage_align_loss": 0.004974365234375, "step": 1350 }, { "ce_loss": 0.038818359375, "distill_loss": 0.0634765625, "epoch": 0.5905511811023622, "explain_loss": 0.828125, "loss": 1.28125, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.002960205078125, "step": 1350 }, { "ce_loss": 0.765625, "distill_loss": 0.0634765625, "epoch": 0.5905511811023622, "explain_loss": 1.4375, "loss": 2.75, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.00421142578125, "step": 1350 }, { "epoch": 0.594925634295713, "grad_norm": 0.7198610901832581, "learning_rate": 0.0001882368387010274, "loss": 1.96015625, "step": 1360 }, { "ce_loss": 0.024658203125, "distill_loss": 0.05859375, "epoch": 0.594925634295713, "explain_loss": 1.09375, "loss": 1.53125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.005584716796875, "step": 1360 }, { "ce_loss": 0.27734375, "distill_loss": 0.06591796875, "epoch": 0.594925634295713, "explain_loss": 1.0390625, "loss": 1.96875, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.0113525390625, "step": 1360 }, { "ce_loss": 0.03125, "distill_loss": 0.0576171875, "epoch": 0.594925634295713, "explain_loss": 0.84765625, "loss": 1.421875, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0034637451171875, "step": 1360 }, { "ce_loss": 0.45703125, "distill_loss": 0.0673828125, "epoch": 0.594925634295713, "explain_loss": 0.7578125, "loss": 1.8046875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0140380859375, "step": 1360 }, { "epoch": 0.5993000874890638, "grad_norm": 1.433884620666504, "learning_rate": 0.00018800890572573026, "loss": 1.9576171875, "step": 1370 }, { "ce_loss": 1.015625, "distill_loss": 0.06298828125, "epoch": 0.5993000874890638, "explain_loss": 0.8828125, "loss": 2.34375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.0147705078125, "step": 1370 }, { "ce_loss": 0.298828125, "distill_loss": 0.050048828125, "epoch": 0.5993000874890638, "explain_loss": 1.0703125, "loss": 1.890625, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.0216064453125, "step": 1370 }, { "ce_loss": 1.671875, "distill_loss": 0.05419921875, "epoch": 0.5993000874890638, "explain_loss": 1.46875, "loss": 3.8125, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.0133056640625, "step": 1370 }, { "ce_loss": 0.010498046875, "distill_loss": 0.0537109375, "epoch": 0.5993000874890638, "explain_loss": 0.9375, "loss": 1.3671875, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.004058837890625, "step": 1370 }, { "epoch": 0.6036745406824147, "grad_norm": 0.8038689494132996, "learning_rate": 0.00018777892632141748, "loss": 1.963330078125, "step": 1380 }, { "ce_loss": 0.00762939453125, "distill_loss": 0.0732421875, "epoch": 0.6036745406824147, "explain_loss": 1.078125, "loss": 1.7109375, "ref_ce_loss": 0.546875, "stage_align_loss": 0.0048828125, "step": 1380 }, { "ce_loss": 0.11572265625, "distill_loss": 0.0556640625, "epoch": 0.6036745406824147, "explain_loss": 1.21875, "loss": 1.828125, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0054931640625, "step": 1380 }, { "ce_loss": 1.5390625, "distill_loss": 0.05078125, "epoch": 0.6036745406824147, "explain_loss": 1.2109375, "loss": 3.125, "ref_ce_loss": 0.302734375, "stage_align_loss": 0.01275634765625, "step": 1380 }, { "ce_loss": 0.326171875, "distill_loss": 0.0654296875, "epoch": 0.6036745406824147, "explain_loss": 1.2734375, "loss": 2.1875, "ref_ce_loss": 0.515625, "stage_align_loss": 0.006256103515625, "step": 1380 }, { "epoch": 0.6080489938757655, "grad_norm": 0.8771425485610962, "learning_rate": 0.0001875469058356902, "loss": 2.2517578125, "step": 1390 }, { "ce_loss": 0.1826171875, "distill_loss": 0.06103515625, "epoch": 0.6080489938757655, "explain_loss": 1.1015625, "loss": 2.15625, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.373046875, "step": 1390 }, { "ce_loss": 0.91015625, "distill_loss": 0.06494140625, "epoch": 0.6080489938757655, "explain_loss": 1.5625, "loss": 3.390625, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.380859375, "step": 1390 }, { "ce_loss": 0.66015625, "distill_loss": 0.0595703125, "epoch": 0.6080489938757655, "explain_loss": 0.9921875, "loss": 2.4375, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.357421875, "step": 1390 }, { "ce_loss": 0.01019287109375, "distill_loss": 0.06787109375, "epoch": 0.6080489938757655, "explain_loss": 1.25, "loss": 2.203125, "ref_ce_loss": 0.53125, "stage_align_loss": 0.34375, "step": 1390 }, { "epoch": 0.6124234470691163, "grad_norm": 0.8339081406593323, "learning_rate": 0.0001873128496636098, "loss": 2.15849609375, "step": 1400 }, { "ce_loss": 0.07373046875, "distill_loss": 0.06884765625, "epoch": 0.6124234470691163, "explain_loss": 1.3515625, "loss": 1.875, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.01611328125, "step": 1400 }, { "ce_loss": 0.3125, "distill_loss": 0.060546875, "epoch": 0.6124234470691163, "explain_loss": 1.25, "loss": 1.953125, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.031005859375, "step": 1400 }, { "ce_loss": 0.0751953125, "distill_loss": 0.062255859375, "epoch": 0.6124234470691163, "explain_loss": 1.203125, "loss": 1.8515625, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.0162353515625, "step": 1400 }, { "ce_loss": 0.88671875, "distill_loss": 0.05517578125, "epoch": 0.6124234470691163, "explain_loss": 1.046875, "loss": 2.390625, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.01483154296875, "step": 1400 }, { "epoch": 0.6167979002624672, "grad_norm": 0.6924841403961182, "learning_rate": 0.0001870767632475727, "loss": 1.96162109375, "step": 1410 }, { "ce_loss": 0.1171875, "distill_loss": 0.0576171875, "epoch": 0.6167979002624672, "explain_loss": 0.8046875, "loss": 1.4765625, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0255126953125, "step": 1410 }, { "ce_loss": 1.1015625, "distill_loss": 0.0537109375, "epoch": 0.6167979002624672, "explain_loss": 1.09375, "loss": 2.9375, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.03125, "step": 1410 }, { "ce_loss": 0.06884765625, "distill_loss": 0.05322265625, "epoch": 0.6167979002624672, "explain_loss": 1.125, "loss": 1.625, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.015869140625, "step": 1410 }, { "ce_loss": 0.1591796875, "distill_loss": 0.0595703125, "epoch": 0.6167979002624672, "explain_loss": 1.140625, "loss": 1.7265625, "ref_ce_loss": 0.34375, "stage_align_loss": 0.02294921875, "step": 1410 }, { "epoch": 0.621172353455818, "grad_norm": 0.7544463872909546, "learning_rate": 0.00018683865207718336, "loss": 1.928564453125, "step": 1420 }, { "ce_loss": 1.2109375, "distill_loss": 0.06201171875, "epoch": 0.621172353455818, "explain_loss": 1.09375, "loss": 3.0, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.044189453125, "step": 1420 }, { "ce_loss": 0.00823974609375, "distill_loss": 0.062255859375, "epoch": 0.621172353455818, "explain_loss": 0.96875, "loss": 1.625, "ref_ce_loss": 0.578125, "stage_align_loss": 0.01300048828125, "step": 1420 }, { "ce_loss": 0.76171875, "distill_loss": 0.064453125, "epoch": 0.621172353455818, "explain_loss": 1.0703125, "loss": 2.46875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.048583984375, "step": 1420 }, { "ce_loss": 0.404296875, "distill_loss": 0.0615234375, "epoch": 0.621172353455818, "explain_loss": 1.4375, "loss": 2.375, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0203857421875, "step": 1420 }, { "epoch": 0.6255468066491688, "grad_norm": 0.8189056515693665, "learning_rate": 0.00018659852168912707, "loss": 1.99169921875, "step": 1430 }, { "ce_loss": 0.00421142578125, "distill_loss": 0.07080078125, "epoch": 0.6255468066491688, "explain_loss": 1.390625, "loss": 1.921875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.01251220703125, "step": 1430 }, { "ce_loss": 0.0269775390625, "distill_loss": 0.05615234375, "epoch": 0.6255468066491688, "explain_loss": 1.296875, "loss": 1.8125, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.01483154296875, "step": 1430 }, { "ce_loss": 0.1298828125, "distill_loss": 0.0625, "epoch": 0.6255468066491688, "explain_loss": 1.5625, "loss": 2.203125, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.0302734375, "step": 1430 }, { "ce_loss": 0.02587890625, "distill_loss": 0.05859375, "epoch": 0.6255468066491688, "explain_loss": 1.234375, "loss": 1.8984375, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01409912109375, "step": 1430 }, { "epoch": 0.6299212598425197, "grad_norm": 0.5336226224899292, "learning_rate": 0.00018635637766704094, "loss": 1.95654296875, "step": 1440 }, { "ce_loss": 0.1455078125, "distill_loss": 0.0654296875, "epoch": 0.6299212598425197, "explain_loss": 1.1640625, "loss": 1.8515625, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.023193359375, "step": 1440 }, { "ce_loss": 0.2138671875, "distill_loss": 0.06005859375, "epoch": 0.6299212598425197, "explain_loss": 0.828125, "loss": 1.578125, "ref_ce_loss": 0.453125, "stage_align_loss": 0.02685546875, "step": 1440 }, { "ce_loss": 0.8984375, "distill_loss": 0.06494140625, "epoch": 0.6299212598425197, "explain_loss": 1.1875, "loss": 2.78125, "ref_ce_loss": 0.609375, "stage_align_loss": 0.02392578125, "step": 1440 }, { "ce_loss": 0.043212890625, "distill_loss": 0.05517578125, "epoch": 0.6299212598425197, "explain_loss": 0.90234375, "loss": 1.40625, "ref_ce_loss": 0.390625, "stage_align_loss": 0.01251220703125, "step": 1440 }, { "epoch": 0.6342957130358705, "grad_norm": 0.47256165742874146, "learning_rate": 0.00018611222564138418, "loss": 1.8572265625, "step": 1450 }, { "ce_loss": 0.43359375, "distill_loss": 0.058837890625, "epoch": 0.6342957130358705, "explain_loss": 0.984375, "loss": 2.03125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0419921875, "step": 1450 }, { "ce_loss": 0.158203125, "distill_loss": 0.06591796875, "epoch": 0.6342957130358705, "explain_loss": 1.15625, "loss": 1.765625, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.019775390625, "step": 1450 }, { "ce_loss": 0.1396484375, "distill_loss": 0.06005859375, "epoch": 0.6342957130358705, "explain_loss": 0.84765625, "loss": 1.4296875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.01361083984375, "step": 1450 }, { "ce_loss": 0.55078125, "distill_loss": 0.056396484375, "epoch": 0.6342957130358705, "explain_loss": 1.15625, "loss": 2.171875, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0233154296875, "step": 1450 }, { "epoch": 0.6386701662292213, "grad_norm": 1.414615273475647, "learning_rate": 0.00018586607128930714, "loss": 2.02470703125, "step": 1460 }, { "ce_loss": 0.0181884765625, "distill_loss": 0.06396484375, "epoch": 0.6386701662292213, "explain_loss": 0.765625, "loss": 1.3046875, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0252685546875, "step": 1460 }, { "ce_loss": 0.0654296875, "distill_loss": 0.064453125, "epoch": 0.6386701662292213, "explain_loss": 1.3125, "loss": 1.9375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.00872802734375, "step": 1460 }, { "ce_loss": 0.0341796875, "distill_loss": 0.07373046875, "epoch": 0.6386701662292213, "explain_loss": 1.0546875, "loss": 1.5625, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0089111328125, "step": 1460 }, { "ce_loss": 0.1806640625, "distill_loss": 0.060791015625, "epoch": 0.6386701662292213, "explain_loss": 1.1484375, "loss": 2.0625, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.0201416015625, "step": 1460 }, { "epoch": 0.6430446194225722, "grad_norm": 0.6524248123168945, "learning_rate": 0.00018561792033451924, "loss": 1.9837890625, "step": 1470 }, { "ce_loss": 0.04248046875, "distill_loss": 0.05517578125, "epoch": 0.6430446194225722, "explain_loss": 1.09375, "loss": 1.828125, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.01104736328125, "step": 1470 }, { "ce_loss": 0.26953125, "distill_loss": 0.052734375, "epoch": 0.6430446194225722, "explain_loss": 1.0625, "loss": 1.90625, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.0198974609375, "step": 1470 }, { "ce_loss": 0.154296875, "distill_loss": 0.064453125, "epoch": 0.6430446194225722, "explain_loss": 1.0234375, "loss": 1.796875, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.026611328125, "step": 1470 }, { "ce_loss": 0.052978515625, "distill_loss": 0.04833984375, "epoch": 0.6430446194225722, "explain_loss": 1.21875, "loss": 1.7421875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.01007080078125, "step": 1470 }, { "epoch": 0.647419072615923, "grad_norm": 1.489967942237854, "learning_rate": 0.0001853677785471561, "loss": 2.11923828125, "step": 1480 }, { "ce_loss": 0.357421875, "distill_loss": 0.06884765625, "epoch": 0.647419072615923, "explain_loss": 1.1015625, "loss": 2.15625, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.01434326171875, "step": 1480 }, { "ce_loss": 0.01318359375, "distill_loss": 0.0615234375, "epoch": 0.647419072615923, "explain_loss": 1.2421875, "loss": 1.7265625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.01129150390625, "step": 1480 }, { "ce_loss": 0.028076171875, "distill_loss": 0.058837890625, "epoch": 0.647419072615923, "explain_loss": 1.25, "loss": 1.9296875, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.006988525390625, "step": 1480 }, { "ce_loss": 0.36328125, "distill_loss": 0.08544921875, "epoch": 0.647419072615923, "explain_loss": 1.3125, "loss": 2.265625, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0150146484375, "step": 1480 }, { "epoch": 0.6517935258092739, "grad_norm": 0.7365946173667908, "learning_rate": 0.00018511565174364507, "loss": 2.06904296875, "step": 1490 }, { "ce_loss": 0.08203125, "distill_loss": 0.05859375, "epoch": 0.6517935258092739, "explain_loss": 0.9609375, "loss": 1.5546875, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.006378173828125, "step": 1490 }, { "ce_loss": 0.1708984375, "distill_loss": 0.0556640625, "epoch": 0.6517935258092739, "explain_loss": 1.2421875, "loss": 1.9453125, "ref_ce_loss": 0.46875, "stage_align_loss": 0.00665283203125, "step": 1490 }, { "ce_loss": 0.234375, "distill_loss": 0.05517578125, "epoch": 0.6517935258092739, "explain_loss": 1.25, "loss": 1.9296875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.009765625, "step": 1490 }, { "ce_loss": 0.1484375, "distill_loss": 0.05908203125, "epoch": 0.6517935258092739, "explain_loss": 1.40625, "loss": 2.1875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0279541015625, "step": 1490 }, { "epoch": 0.6561679790026247, "grad_norm": 0.4648372530937195, "learning_rate": 0.00018486154578657024, "loss": 1.95859375, "step": 1500 }, { "ce_loss": 0.0595703125, "distill_loss": 0.054931640625, "epoch": 0.6561679790026247, "explain_loss": 1.171875, "loss": 1.78125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0069580078125, "step": 1500 }, { "ce_loss": 0.0283203125, "distill_loss": 0.0537109375, "epoch": 0.6561679790026247, "explain_loss": 1.328125, "loss": 1.875, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0086669921875, "step": 1500 }, { "ce_loss": 0.78125, "distill_loss": 0.0654296875, "epoch": 0.6561679790026247, "explain_loss": 1.1015625, "loss": 2.53125, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.01129150390625, "step": 1500 }, { "ce_loss": 0.125, "distill_loss": 0.061279296875, "epoch": 0.6561679790026247, "explain_loss": 0.94921875, "loss": 1.78125, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.0311279296875, "step": 1500 }, { "epoch": 0.6605424321959755, "grad_norm": 0.9448956847190857, "learning_rate": 0.00018460546658453594, "loss": 2.009765625, "step": 1510 }, { "ce_loss": 0.111328125, "distill_loss": 0.05419921875, "epoch": 0.6605424321959755, "explain_loss": 1.15625, "loss": 1.6640625, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.0164794921875, "step": 1510 }, { "ce_loss": 0.7421875, "distill_loss": 0.049072265625, "epoch": 0.6605424321959755, "explain_loss": 1.453125, "loss": 2.75, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.008544921875, "step": 1510 }, { "ce_loss": 0.318359375, "distill_loss": 0.0458984375, "epoch": 0.6605424321959755, "explain_loss": 1.2578125, "loss": 1.984375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.00897216796875, "step": 1510 }, { "ce_loss": 0.06494140625, "distill_loss": 0.046875, "epoch": 0.6605424321959755, "explain_loss": 1.203125, "loss": 1.6875, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.009521484375, "step": 1510 }, { "epoch": 0.6649168853893264, "grad_norm": 0.6521678566932678, "learning_rate": 0.00018434742009202943, "loss": 1.9296875, "step": 1520 }, { "ce_loss": 0.0115966796875, "distill_loss": 0.0615234375, "epoch": 0.6649168853893264, "explain_loss": 1.265625, "loss": 1.796875, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.006103515625, "step": 1520 }, { "ce_loss": 0.73828125, "distill_loss": 0.04931640625, "epoch": 0.6649168853893264, "explain_loss": 1.0625, "loss": 2.609375, "ref_ce_loss": 0.71875, "stage_align_loss": 0.04052734375, "step": 1520 }, { "ce_loss": 0.875, "distill_loss": 0.05859375, "epoch": 0.6649168853893264, "explain_loss": 1.328125, "loss": 2.8125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.03271484375, "step": 1520 }, { "ce_loss": 0.2294921875, "distill_loss": 0.052490234375, "epoch": 0.6649168853893264, "explain_loss": 1.421875, "loss": 2.171875, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.01507568359375, "step": 1520 }, { "epoch": 0.6692913385826772, "grad_norm": 0.7979750633239746, "learning_rate": 0.0001840874123092825, "loss": 2.04755859375, "step": 1530 }, { "ce_loss": 0.2890625, "distill_loss": 0.06494140625, "epoch": 0.6692913385826772, "explain_loss": 1.2265625, "loss": 1.953125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.01708984375, "step": 1530 }, { "ce_loss": 0.162109375, "distill_loss": 0.05712890625, "epoch": 0.6692913385826772, "explain_loss": 1.3046875, "loss": 1.8203125, "ref_ce_loss": 0.2734375, "stage_align_loss": 0.0223388671875, "step": 1530 }, { "ce_loss": 0.02783203125, "distill_loss": 0.0634765625, "epoch": 0.6692913385826772, "explain_loss": 1.0390625, "loss": 1.4453125, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.01080322265625, "step": 1530 }, { "ce_loss": 0.10595703125, "distill_loss": 0.0625, "epoch": 0.6692913385826772, "explain_loss": 1.0078125, "loss": 1.6875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.01055908203125, "step": 1530 }, { "epoch": 0.673665791776028, "grad_norm": 0.7226892113685608, "learning_rate": 0.00018382544928213177, "loss": 1.85302734375, "step": 1540 }, { "ce_loss": 0.058837890625, "distill_loss": 0.0693359375, "epoch": 0.673665791776028, "explain_loss": 1.203125, "loss": 1.765625, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.0162353515625, "step": 1540 }, { "ce_loss": 0.5, "distill_loss": 0.05029296875, "epoch": 0.673665791776028, "explain_loss": 0.8671875, "loss": 1.8359375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.02490234375, "step": 1540 }, { "ce_loss": 0.01708984375, "distill_loss": 0.0576171875, "epoch": 0.673665791776028, "explain_loss": 1.390625, "loss": 2.03125, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01177978515625, "step": 1540 }, { "ce_loss": 0.2373046875, "distill_loss": 0.06396484375, "epoch": 0.673665791776028, "explain_loss": 1.390625, "loss": 2.375, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.014892578125, "step": 1540 }, { "epoch": 0.6780402449693789, "grad_norm": 0.677794873714447, "learning_rate": 0.00018356153710187833, "loss": 1.88173828125, "step": 1550 }, { "ce_loss": 0.0162353515625, "distill_loss": 0.06787109375, "epoch": 0.6780402449693789, "explain_loss": 1.21875, "loss": 1.6875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00830078125, "step": 1550 }, { "ce_loss": 0.376953125, "distill_loss": 0.0478515625, "epoch": 0.6780402449693789, "explain_loss": 1.0, "loss": 2.03125, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.0145263671875, "step": 1550 }, { "ce_loss": 0.0128173828125, "distill_loss": 0.05859375, "epoch": 0.6780402449693789, "explain_loss": 1.21875, "loss": 1.7109375, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.006256103515625, "step": 1550 }, { "ce_loss": 0.2216796875, "distill_loss": 0.0537109375, "epoch": 0.6780402449693789, "explain_loss": 0.7890625, "loss": 1.5625, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0162353515625, "step": 1550 }, { "epoch": 0.6824146981627297, "grad_norm": 0.6361228227615356, "learning_rate": 0.00018329568190514597, "loss": 1.96181640625, "step": 1560 }, { "ce_loss": 0.09228515625, "distill_loss": 0.0576171875, "epoch": 0.6824146981627297, "explain_loss": 1.28125, "loss": 1.921875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0098876953125, "step": 1560 }, { "ce_loss": 0.16796875, "distill_loss": 0.0498046875, "epoch": 0.6824146981627297, "explain_loss": 1.265625, "loss": 1.984375, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.01446533203125, "step": 1560 }, { "ce_loss": 0.396484375, "distill_loss": 0.0498046875, "epoch": 0.6824146981627297, "explain_loss": 1.3125, "loss": 2.0625, "ref_ce_loss": 0.28125, "stage_align_loss": 0.014404296875, "step": 1560 }, { "ce_loss": 0.1064453125, "distill_loss": 0.0595703125, "epoch": 0.6824146981627297, "explain_loss": 1.09375, "loss": 1.7421875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0096435546875, "step": 1560 }, { "epoch": 0.6867891513560805, "grad_norm": 73.33589172363281, "learning_rate": 0.00018302788987373847, "loss": 2.2162109375, "step": 1570 }, { "ce_loss": 0.7890625, "distill_loss": 0.08349609375, "epoch": 0.6867891513560805, "explain_loss": 0.9453125, "loss": 2.40625, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.154296875, "step": 1570 }, { "ce_loss": 0.373046875, "distill_loss": 0.0732421875, "epoch": 0.6867891513560805, "explain_loss": 1.515625, "loss": 2.9375, "ref_ce_loss": 0.609375, "stage_align_loss": 0.359375, "step": 1570 }, { "ce_loss": 0.1865234375, "distill_loss": 0.07421875, "epoch": 0.6867891513560805, "explain_loss": 1.125, "loss": 2.21875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.3671875, "step": 1570 }, { "ce_loss": 0.9609375, "distill_loss": 0.0751953125, "epoch": 0.6867891513560805, "explain_loss": 1.234375, "loss": 3.0, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.279296875, "step": 1570 }, { "epoch": 0.6911636045494314, "grad_norm": 190.2582550048828, "learning_rate": 0.00018275816723449593, "loss": 2.66064453125, "step": 1580 }, { "ce_loss": 0.423828125, "distill_loss": 0.0576171875, "epoch": 0.6911636045494314, "explain_loss": 0.8984375, "loss": 2.03125, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.28515625, "step": 1580 }, { "ce_loss": 0.53515625, "distill_loss": 0.06103515625, "epoch": 0.6911636045494314, "explain_loss": 1.234375, "loss": 2.59375, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.28515625, "step": 1580 }, { "ce_loss": 1.0234375, "distill_loss": 0.08984375, "epoch": 0.6911636045494314, "explain_loss": 1.0859375, "loss": 3.3125, "ref_ce_loss": 0.65625, "stage_align_loss": 0.447265625, "step": 1580 }, { "ce_loss": 0.0712890625, "distill_loss": 0.06005859375, "epoch": 0.6911636045494314, "explain_loss": 1.046875, "loss": 1.9296875, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.302734375, "step": 1580 }, { "epoch": 0.6955380577427821, "grad_norm": 0.9107314348220825, "learning_rate": 0.00018248652025914996, "loss": 2.349609375, "step": 1590 }, { "ce_loss": 4.65625, "distill_loss": 0.0673828125, "epoch": 0.6955380577427821, "explain_loss": 1.4921875, "loss": 7.125, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.30859375, "step": 1590 }, { "ce_loss": 0.09912109375, "distill_loss": 0.056396484375, "epoch": 0.6955380577427821, "explain_loss": 1.6796875, "loss": 2.546875, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.310546875, "step": 1590 }, { "ce_loss": 4.96875, "distill_loss": 0.07177734375, "epoch": 0.6955380577427821, "explain_loss": 1.53125, "loss": 7.46875, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.306640625, "step": 1590 }, { "ce_loss": 17.625, "distill_loss": 0.08740234375, "epoch": 0.6955380577427821, "explain_loss": 1.5078125, "loss": 20.125, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.30859375, "step": 1590 }, { "epoch": 0.6999125109361329, "grad_norm": 0.678905189037323, "learning_rate": 0.0001822129552641778, "loss": 3.19482421875, "step": 1600 }, { "ce_loss": 0.119140625, "distill_loss": 0.0595703125, "epoch": 0.6999125109361329, "explain_loss": 1.0390625, "loss": 1.9375, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.296875, "step": 1600 }, { "ce_loss": 0.5078125, "distill_loss": 0.080078125, "epoch": 0.6999125109361329, "explain_loss": 0.9921875, "loss": 2.46875, "ref_ce_loss": 0.59375, "stage_align_loss": 0.3046875, "step": 1600 }, { "ce_loss": 0.2431640625, "distill_loss": 0.058837890625, "epoch": 0.6999125109361329, "explain_loss": 1.1875, "loss": 2.296875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.30859375, "step": 1600 }, { "ce_loss": 0.103515625, "distill_loss": 0.07080078125, "epoch": 0.6999125109361329, "explain_loss": 1.5546875, "loss": 2.46875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.302734375, "step": 1600 }, { "epoch": 0.7042869641294838, "grad_norm": 0.6150801181793213, "learning_rate": 0.00018193747861065557, "loss": 2.22431640625, "step": 1610 }, { "ce_loss": 0.0260009765625, "distill_loss": 0.058349609375, "epoch": 0.7042869641294838, "explain_loss": 0.828125, "loss": 1.609375, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.296875, "step": 1610 }, { "ce_loss": 0.87890625, "distill_loss": 0.06787109375, "epoch": 0.7042869641294838, "explain_loss": 1.3125, "loss": 3.015625, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.30859375, "step": 1610 }, { "ce_loss": 0.0205078125, "distill_loss": 0.06396484375, "epoch": 0.7042869641294838, "explain_loss": 1.1640625, "loss": 1.96875, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.291015625, "step": 1610 }, { "ce_loss": 0.029541015625, "distill_loss": 0.06689453125, "epoch": 0.7042869641294838, "explain_loss": 1.09375, "loss": 1.9375, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.298828125, "step": 1610 }, { "epoch": 0.7086614173228346, "grad_norm": 1.395228624343872, "learning_rate": 0.00018166009670411018, "loss": 2.29892578125, "step": 1620 }, { "ce_loss": 0.205078125, "distill_loss": 0.049072265625, "epoch": 0.7086614173228346, "explain_loss": 1.0390625, "loss": 2.0625, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.2734375, "step": 1620 }, { "ce_loss": 0.65625, "distill_loss": 0.05078125, "epoch": 0.7086614173228346, "explain_loss": 0.95703125, "loss": 2.46875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.28125, "step": 1620 }, { "ce_loss": 0.20703125, "distill_loss": 0.058349609375, "epoch": 0.7086614173228346, "explain_loss": 0.71484375, "loss": 1.53125, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.283203125, "step": 1620 }, { "ce_loss": 0.032958984375, "distill_loss": 0.0576171875, "epoch": 0.7086614173228346, "explain_loss": 0.94921875, "loss": 1.7734375, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.27734375, "step": 1620 }, { "epoch": 0.7130358705161854, "grad_norm": 1.1090033054351807, "learning_rate": 0.00018138081599437046, "loss": 2.18095703125, "step": 1630 }, { "ce_loss": 0.0101318359375, "distill_loss": 0.04833984375, "epoch": 0.7130358705161854, "explain_loss": 0.8125, "loss": 1.3984375, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.26171875, "step": 1630 }, { "ce_loss": 0.01068115234375, "distill_loss": 0.054931640625, "epoch": 0.7130358705161854, "explain_loss": 1.203125, "loss": 2.0, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.26171875, "step": 1630 }, { "ce_loss": 1.5703125, "distill_loss": 0.0537109375, "epoch": 0.7130358705161854, "explain_loss": 1.0234375, "loss": 3.4375, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.2734375, "step": 1630 }, { "ce_loss": 0.16796875, "distill_loss": 0.06787109375, "epoch": 0.7130358705161854, "explain_loss": 1.625, "loss": 2.515625, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.279296875, "step": 1630 }, { "epoch": 0.7174103237095363, "grad_norm": 0.8422850966453552, "learning_rate": 0.00018109964297541734, "loss": 2.15400390625, "step": 1640 }, { "ce_loss": 0.296875, "distill_loss": 0.0712890625, "epoch": 0.7174103237095363, "explain_loss": 1.21875, "loss": 2.28125, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.251953125, "step": 1640 }, { "ce_loss": 0.84375, "distill_loss": 0.06640625, "epoch": 0.7174103237095363, "explain_loss": 1.234375, "loss": 3.0, "ref_ce_loss": 0.609375, "stage_align_loss": 0.255859375, "step": 1640 }, { "ce_loss": 0.361328125, "distill_loss": 0.050048828125, "epoch": 0.7174103237095363, "explain_loss": 1.53125, "loss": 2.5625, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.251953125, "step": 1640 }, { "ce_loss": 0.0634765625, "distill_loss": 0.06201171875, "epoch": 0.7174103237095363, "explain_loss": 1.2109375, "loss": 1.9140625, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.236328125, "step": 1640 }, { "epoch": 0.7217847769028871, "grad_norm": 3.6855292320251465, "learning_rate": 0.0001808165841852326, "loss": 2.1857421875, "step": 1650 }, { "ce_loss": 0.2060546875, "distill_loss": 0.06640625, "epoch": 0.7217847769028871, "explain_loss": 0.9765625, "loss": 2.34375, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.56640625, "step": 1650 }, { "ce_loss": 1.1171875, "distill_loss": 0.0673828125, "epoch": 0.7217847769028871, "explain_loss": 1.234375, "loss": 3.53125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.5625, "step": 1650 }, { "ce_loss": 0.06494140625, "distill_loss": 0.05712890625, "epoch": 0.7217847769028871, "explain_loss": 1.203125, "loss": 2.265625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.5546875, "step": 1650 }, { "ce_loss": 1.6875, "distill_loss": 0.053955078125, "epoch": 0.7217847769028871, "explain_loss": 1.21875, "loss": 4.375, "ref_ce_loss": 0.68359375, "stage_align_loss": 0.7421875, "step": 1650 }, { "epoch": 0.726159230096238, "grad_norm": 4.52638578414917, "learning_rate": 0.00018053164620564696, "loss": 2.20625, "step": 1660 }, { "ce_loss": 0.36328125, "distill_loss": 0.057373046875, "epoch": 0.726159230096238, "explain_loss": 0.9453125, "loss": 2.0, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.30859375, "step": 1660 }, { "ce_loss": 0.0128173828125, "distill_loss": 0.0537109375, "epoch": 0.726159230096238, "explain_loss": 1.46875, "loss": 2.09375, "ref_ce_loss": 0.5, "stage_align_loss": 0.0498046875, "step": 1660 }, { "ce_loss": 0.02734375, "distill_loss": 0.0595703125, "epoch": 0.726159230096238, "explain_loss": 1.2578125, "loss": 1.734375, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.057861328125, "step": 1660 }, { "ce_loss": 0.17578125, "distill_loss": 0.0576171875, "epoch": 0.726159230096238, "explain_loss": 1.109375, "loss": 2.03125, "ref_ce_loss": 0.40625, "stage_align_loss": 0.2890625, "step": 1660 }, { "epoch": 0.7305336832895888, "grad_norm": 1.2378263473510742, "learning_rate": 0.00018024483566218708, "loss": 2.2337890625, "step": 1670 }, { "ce_loss": 0.09521484375, "distill_loss": 0.07177734375, "epoch": 0.7305336832895888, "explain_loss": 1.2421875, "loss": 2.5625, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.6328125, "step": 1670 }, { "ce_loss": 0.10888671875, "distill_loss": 0.0751953125, "epoch": 0.7305336832895888, "explain_loss": 0.6953125, "loss": 2.046875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.625, "step": 1670 }, { "ce_loss": 0.82421875, "distill_loss": 0.07177734375, "epoch": 0.7305336832895888, "explain_loss": 1.1015625, "loss": 3.125, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.625, "step": 1670 }, { "ce_loss": 0.1376953125, "distill_loss": 0.05126953125, "epoch": 0.7305336832895888, "explain_loss": 1.015625, "loss": 2.28125, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.62890625, "step": 1670 }, { "epoch": 0.7349081364829396, "grad_norm": 1.4709298610687256, "learning_rate": 0.0001799561592239214, "loss": 2.45205078125, "step": 1680 }, { "ce_loss": 0.68359375, "distill_loss": 0.06005859375, "epoch": 0.7349081364829396, "explain_loss": 1.046875, "loss": 2.71875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.494140625, "step": 1680 }, { "ce_loss": 0.4921875, "distill_loss": 0.07080078125, "epoch": 0.7349081364829396, "explain_loss": 0.96484375, "loss": 2.4375, "ref_ce_loss": 0.421875, "stage_align_loss": 0.490234375, "step": 1680 }, { "ce_loss": 0.040771484375, "distill_loss": 0.058837890625, "epoch": 0.7349081364829396, "explain_loss": 1.2109375, "loss": 2.28125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.48828125, "step": 1680 }, { "ce_loss": 0.4453125, "distill_loss": 0.05615234375, "epoch": 0.7349081364829396, "explain_loss": 1.109375, "loss": 2.546875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.49609375, "step": 1680 }, { "epoch": 0.7392825896762905, "grad_norm": 2.3642683029174805, "learning_rate": 0.0001796656236033052, "loss": 2.44384765625, "step": 1690 }, { "ce_loss": 0.255859375, "distill_loss": 0.09228515625, "epoch": 0.7392825896762905, "explain_loss": 1.1171875, "loss": 2.1875, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.244140625, "step": 1690 }, { "ce_loss": 0.6328125, "distill_loss": 0.049072265625, "epoch": 0.7392825896762905, "explain_loss": 1.328125, "loss": 2.6875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.2392578125, "step": 1690 }, { "ce_loss": 0.061767578125, "distill_loss": 0.0517578125, "epoch": 0.7392825896762905, "explain_loss": 0.78515625, "loss": 1.609375, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.236328125, "step": 1690 }, { "ce_loss": 0.169921875, "distill_loss": 0.06298828125, "epoch": 0.7392825896762905, "explain_loss": 0.9453125, "loss": 1.8984375, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.2333984375, "step": 1690 }, { "epoch": 0.7436570428696413, "grad_norm": 1.0252982378005981, "learning_rate": 0.0001793732355560243, "loss": 2.1865234375, "step": 1700 }, { "ce_loss": 1.2421875, "distill_loss": 0.0703125, "epoch": 0.7436570428696413, "explain_loss": 1.25, "loss": 3.4375, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.427734375, "step": 1700 }, { "ce_loss": 0.18359375, "distill_loss": 0.06494140625, "epoch": 0.7436570428696413, "explain_loss": 1.2890625, "loss": 2.21875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.1728515625, "step": 1700 }, { "ce_loss": 0.138671875, "distill_loss": 0.08056640625, "epoch": 0.7436570428696413, "explain_loss": 0.9609375, "loss": 2.34375, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.70703125, "step": 1700 }, { "ce_loss": 0.1484375, "distill_loss": 0.0654296875, "epoch": 0.7436570428696413, "explain_loss": 1.390625, "loss": 2.75, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.671875, "step": 1700 }, { "epoch": 0.7480314960629921, "grad_norm": 10.935608863830566, "learning_rate": 0.00017907900188083827, "loss": 2.153125, "step": 1710 }, { "ce_loss": 0.263671875, "distill_loss": 0.08984375, "epoch": 0.7480314960629921, "explain_loss": 1.2421875, "loss": 2.25, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.057373046875, "step": 1710 }, { "ce_loss": 0.54296875, "distill_loss": 0.078125, "epoch": 0.7480314960629921, "explain_loss": 1.2421875, "loss": 2.46875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0732421875, "step": 1710 }, { "ce_loss": 1.09375, "distill_loss": 0.052490234375, "epoch": 0.7480314960629921, "explain_loss": 1.140625, "loss": 2.703125, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.08154296875, "step": 1710 }, { "ce_loss": 0.138671875, "distill_loss": 0.06787109375, "epoch": 0.7480314960629921, "explain_loss": 0.74609375, "loss": 1.3984375, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.10400390625, "step": 1710 }, { "epoch": 0.752405949256343, "grad_norm": 1.067238450050354, "learning_rate": 0.00017878292941942212, "loss": 2.11611328125, "step": 1720 }, { "ce_loss": 8.8125, "distill_loss": 0.07080078125, "epoch": 0.752405949256343, "explain_loss": 1.9375, "loss": 11.5625, "ref_ce_loss": 0.671875, "stage_align_loss": 0.050537109375, "step": 1720 }, { "ce_loss": 2.453125, "distill_loss": 0.072265625, "epoch": 0.752405949256343, "explain_loss": 1.28125, "loss": 4.65625, "ref_ce_loss": 0.80078125, "stage_align_loss": 0.045654296875, "step": 1720 }, { "ce_loss": 7.8125, "distill_loss": 0.06640625, "epoch": 0.752405949256343, "explain_loss": 1.171875, "loss": 9.875, "ref_ce_loss": 0.765625, "stage_align_loss": 0.042236328125, "step": 1720 }, { "ce_loss": 5.5625, "distill_loss": 0.1171875, "epoch": 0.752405949256343, "explain_loss": 1.53125, "loss": 7.9375, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.049072265625, "step": 1720 }, { "epoch": 0.7567804024496938, "grad_norm": 0.9859330654144287, "learning_rate": 0.00017848502505620723, "loss": 2.4646484375, "step": 1730 }, { "ce_loss": 0.490234375, "distill_loss": 0.0693359375, "epoch": 0.7567804024496938, "explain_loss": 0.921875, "loss": 2.0, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.0277099609375, "step": 1730 }, { "ce_loss": 0.06640625, "distill_loss": 0.06494140625, "epoch": 0.7567804024496938, "explain_loss": 1.0, "loss": 1.640625, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.027099609375, "step": 1730 }, { "ce_loss": 0.3671875, "distill_loss": 0.062255859375, "epoch": 0.7567804024496938, "explain_loss": 1.0625, "loss": 1.9375, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.029541015625, "step": 1730 }, { "ce_loss": 0.439453125, "distill_loss": 0.083984375, "epoch": 0.7567804024496938, "explain_loss": 1.03125, "loss": 2.078125, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.031494140625, "step": 1730 }, { "epoch": 0.7611548556430446, "grad_norm": 1.089994192123413, "learning_rate": 0.00017818529571822137, "loss": 2.03896484375, "step": 1740 }, { "ce_loss": 0.2236328125, "distill_loss": 0.062255859375, "epoch": 0.7611548556430446, "explain_loss": 1.140625, "loss": 1.8828125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.023681640625, "step": 1740 }, { "ce_loss": 0.130859375, "distill_loss": 0.0771484375, "epoch": 0.7611548556430446, "explain_loss": 1.203125, "loss": 1.703125, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.0206298828125, "step": 1740 }, { "ce_loss": 0.29296875, "distill_loss": 0.051025390625, "epoch": 0.7611548556430446, "explain_loss": 0.84375, "loss": 1.5078125, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.025390625, "step": 1740 }, { "ce_loss": 0.09375, "distill_loss": 0.0595703125, "epoch": 0.7611548556430446, "explain_loss": 0.96875, "loss": 1.609375, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0206298828125, "step": 1740 }, { "epoch": 0.7655293088363955, "grad_norm": 0.9664937853813171, "learning_rate": 0.00017788374837492756, "loss": 2.01630859375, "step": 1750 }, { "ce_loss": 1.4765625, "distill_loss": 0.083984375, "epoch": 0.7655293088363955, "explain_loss": 1.34375, "loss": 3.421875, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.0274658203125, "step": 1750 }, { "ce_loss": 0.470703125, "distill_loss": 0.06494140625, "epoch": 0.7655293088363955, "explain_loss": 0.9375, "loss": 1.875, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.029052734375, "step": 1750 }, { "ce_loss": 0.423828125, "distill_loss": 0.056396484375, "epoch": 0.7655293088363955, "explain_loss": 1.0859375, "loss": 2.09375, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0252685546875, "step": 1750 }, { "ce_loss": 0.052978515625, "distill_loss": 0.0673828125, "epoch": 0.7655293088363955, "explain_loss": 0.90625, "loss": 1.6328125, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.0224609375, "step": 1750 }, { "epoch": 0.7699037620297463, "grad_norm": 1.4918144941329956, "learning_rate": 0.00017758039003806206, "loss": 2.038671875, "step": 1760 }, { "ce_loss": 0.0230712890625, "distill_loss": 0.0615234375, "epoch": 0.7699037620297463, "explain_loss": 0.96875, "loss": 1.4921875, "ref_ce_loss": 0.421875, "stage_align_loss": 0.01611328125, "step": 1760 }, { "ce_loss": 0.06640625, "distill_loss": 0.056396484375, "epoch": 0.7699037620297463, "explain_loss": 1.234375, "loss": 1.78125, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0189208984375, "step": 1760 }, { "ce_loss": 0.443359375, "distill_loss": 0.0654296875, "epoch": 0.7699037620297463, "explain_loss": 0.875, "loss": 1.765625, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.03759765625, "step": 1760 }, { "ce_loss": 0.203125, "distill_loss": 0.06396484375, "epoch": 0.7699037620297463, "explain_loss": 1.3828125, "loss": 2.125, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0240478515625, "step": 1760 }, { "epoch": 0.7742782152230971, "grad_norm": 0.8334062099456787, "learning_rate": 0.00017727522776147124, "loss": 1.97578125, "step": 1770 }, { "ce_loss": 0.0361328125, "distill_loss": 0.087890625, "epoch": 0.7742782152230971, "explain_loss": 1.15625, "loss": 1.7890625, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.0181884765625, "step": 1770 }, { "ce_loss": 0.03173828125, "distill_loss": 0.0517578125, "epoch": 0.7742782152230971, "explain_loss": 1.75, "loss": 2.3125, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0167236328125, "step": 1770 }, { "ce_loss": 0.69140625, "distill_loss": 0.054931640625, "epoch": 0.7742782152230971, "explain_loss": 1.203125, "loss": 2.453125, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0478515625, "step": 1770 }, { "ce_loss": 0.099609375, "distill_loss": 0.0751953125, "epoch": 0.7742782152230971, "explain_loss": 1.3046875, "loss": 1.9296875, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0216064453125, "step": 1770 }, { "epoch": 0.778652668416448, "grad_norm": 0.5033002495765686, "learning_rate": 0.0001769682686409476, "loss": 1.8890625, "step": 1780 }, { "ce_loss": 0.09130859375, "distill_loss": 0.0693359375, "epoch": 0.778652668416448, "explain_loss": 1.0234375, "loss": 1.5625, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0291748046875, "step": 1780 }, { "ce_loss": 0.55078125, "distill_loss": 0.05810546875, "epoch": 0.778652668416448, "explain_loss": 1.015625, "loss": 2.15625, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.0269775390625, "step": 1780 }, { "ce_loss": 0.0257568359375, "distill_loss": 0.0712890625, "epoch": 0.778652668416448, "explain_loss": 1.078125, "loss": 1.78125, "ref_ce_loss": 0.578125, "stage_align_loss": 0.021728515625, "step": 1780 }, { "ce_loss": 0.19921875, "distill_loss": 0.0703125, "epoch": 0.778652668416448, "explain_loss": 1.015625, "loss": 1.734375, "ref_ce_loss": 0.4375, "stage_align_loss": 0.01519775390625, "step": 1780 }, { "epoch": 0.7830271216097988, "grad_norm": 0.66920405626297, "learning_rate": 0.00017665951981406494, "loss": 1.9478515625, "step": 1790 }, { "ce_loss": 0.0888671875, "distill_loss": 0.058837890625, "epoch": 0.7830271216097988, "explain_loss": 0.953125, "loss": 1.515625, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0380859375, "step": 1790 }, { "ce_loss": 0.2431640625, "distill_loss": 0.051513671875, "epoch": 0.7830271216097988, "explain_loss": 0.8203125, "loss": 1.5546875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0303955078125, "step": 1790 }, { "ce_loss": 0.021484375, "distill_loss": 0.054931640625, "epoch": 0.7830271216097988, "explain_loss": 1.1875, "loss": 1.8046875, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.0166015625, "step": 1790 }, { "ce_loss": 0.5625, "distill_loss": 0.061279296875, "epoch": 0.7830271216097988, "explain_loss": 1.125, "loss": 2.28125, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.10107421875, "step": 1790 }, { "epoch": 0.7874015748031497, "grad_norm": 0.8417300581932068, "learning_rate": 0.00017634898846001206, "loss": 1.95322265625, "step": 1800 }, { "ce_loss": 0.1416015625, "distill_loss": 0.05029296875, "epoch": 0.7874015748031497, "explain_loss": 1.078125, "loss": 1.96875, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.047607421875, "step": 1800 }, { "ce_loss": 0.025146484375, "distill_loss": 0.07763671875, "epoch": 0.7874015748031497, "explain_loss": 1.0625, "loss": 1.484375, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.007598876953125, "step": 1800 }, { "ce_loss": 0.0869140625, "distill_loss": 0.06640625, "epoch": 0.7874015748031497, "explain_loss": 0.76953125, "loss": 1.2890625, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.010498046875, "step": 1800 }, { "ce_loss": 0.01031494140625, "distill_loss": 0.049072265625, "epoch": 0.7874015748031497, "explain_loss": 1.078125, "loss": 1.5859375, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.01422119140625, "step": 1800 }, { "epoch": 0.7917760279965005, "grad_norm": 0.7631195187568665, "learning_rate": 0.00017603668179942615, "loss": 1.95849609375, "step": 1810 }, { "ce_loss": 0.02880859375, "distill_loss": 0.06640625, "epoch": 0.7917760279965005, "explain_loss": 1.28125, "loss": 1.859375, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0087890625, "step": 1810 }, { "ce_loss": 0.0810546875, "distill_loss": 0.0576171875, "epoch": 0.7917760279965005, "explain_loss": 1.125, "loss": 1.6328125, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0123291015625, "step": 1810 }, { "ce_loss": 0.1533203125, "distill_loss": 0.0625, "epoch": 0.7917760279965005, "explain_loss": 1.0859375, "loss": 1.859375, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0262451171875, "step": 1810 }, { "ce_loss": 0.034912109375, "distill_loss": 0.05517578125, "epoch": 0.7917760279965005, "explain_loss": 1.2421875, "loss": 1.953125, "ref_ce_loss": 0.609375, "stage_align_loss": 0.0093994140625, "step": 1810 }, { "epoch": 0.7961504811898513, "grad_norm": 0.5242951512336731, "learning_rate": 0.0001757226070942247, "loss": 1.9384765625, "step": 1820 }, { "ce_loss": 0.6953125, "distill_loss": 0.05859375, "epoch": 0.7961504811898513, "explain_loss": 0.8671875, "loss": 2.09375, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.038330078125, "step": 1820 }, { "ce_loss": 0.064453125, "distill_loss": 0.0615234375, "epoch": 0.7961504811898513, "explain_loss": 1.09375, "loss": 1.78125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.03515625, "step": 1820 }, { "ce_loss": 0.76953125, "distill_loss": 0.05322265625, "epoch": 0.7961504811898513, "explain_loss": 1.3359375, "loss": 2.53125, "ref_ce_loss": 0.34375, "stage_align_loss": 0.03515625, "step": 1820 }, { "ce_loss": 0.007476806640625, "distill_loss": 0.0673828125, "epoch": 0.7961504811898513, "explain_loss": 1.390625, "loss": 2.03125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0439453125, "step": 1820 }, { "epoch": 0.800524934383202, "grad_norm": 0.7546286582946777, "learning_rate": 0.0001754067716474367, "loss": 1.9240234375, "step": 1830 }, { "ce_loss": 0.044189453125, "distill_loss": 0.058837890625, "epoch": 0.800524934383202, "explain_loss": 1.0859375, "loss": 1.5625, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.01220703125, "step": 1830 }, { "ce_loss": 0.10107421875, "distill_loss": 0.05615234375, "epoch": 0.800524934383202, "explain_loss": 1.203125, "loss": 1.84375, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.01904296875, "step": 1830 }, { "ce_loss": 0.0216064453125, "distill_loss": 0.060791015625, "epoch": 0.800524934383202, "explain_loss": 1.15625, "loss": 1.71875, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.01275634765625, "step": 1830 }, { "ce_loss": 0.94140625, "distill_loss": 0.051025390625, "epoch": 0.800524934383202, "explain_loss": 1.2890625, "loss": 2.796875, "ref_ce_loss": 0.5, "stage_align_loss": 0.01312255859375, "step": 1830 }, { "epoch": 0.8048993875765529, "grad_norm": 1.6206483840942383, "learning_rate": 0.00017508918280303277, "loss": 1.91982421875, "step": 1840 }, { "ce_loss": 0.08544921875, "distill_loss": 0.056396484375, "epoch": 0.8048993875765529, "explain_loss": 1.1875, "loss": 1.6796875, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.01300048828125, "step": 1840 }, { "ce_loss": 0.2333984375, "distill_loss": 0.0576171875, "epoch": 0.8048993875765529, "explain_loss": 1.0234375, "loss": 1.7109375, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.0260009765625, "step": 1840 }, { "ce_loss": 0.1494140625, "distill_loss": 0.0439453125, "epoch": 0.8048993875765529, "explain_loss": 1.15625, "loss": 1.90625, "ref_ce_loss": 0.546875, "stage_align_loss": 0.01361083984375, "step": 1840 }, { "ce_loss": 0.31640625, "distill_loss": 0.0654296875, "epoch": 0.8048993875765529, "explain_loss": 1.140625, "loss": 2.15625, "ref_ce_loss": 0.625, "stage_align_loss": 0.0101318359375, "step": 1840 }, { "epoch": 0.8092738407699037, "grad_norm": 0.9712022542953491, "learning_rate": 0.00017476984794575454, "loss": 1.9021484375, "step": 1850 }, { "ce_loss": 0.3984375, "distill_loss": 0.0615234375, "epoch": 0.8092738407699037, "explain_loss": 1.0, "loss": 1.8671875, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.033203125, "step": 1850 }, { "ce_loss": 0.0291748046875, "distill_loss": 0.0654296875, "epoch": 0.8092738407699037, "explain_loss": 0.72265625, "loss": 1.2109375, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.0142822265625, "step": 1850 }, { "ce_loss": 0.36328125, "distill_loss": 0.0517578125, "epoch": 0.8092738407699037, "explain_loss": 1.1640625, "loss": 2.0625, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0206298828125, "step": 1850 }, { "ce_loss": 0.1650390625, "distill_loss": 0.072265625, "epoch": 0.8092738407699037, "explain_loss": 1.234375, "loss": 2.1875, "ref_ce_loss": 0.6875, "stage_align_loss": 0.0230712890625, "step": 1850 }, { "epoch": 0.8136482939632546, "grad_norm": 0.5178813338279724, "learning_rate": 0.00017444877450094283, "loss": 1.979296875, "step": 1860 }, { "ce_loss": 0.265625, "distill_loss": 0.046875, "epoch": 0.8136482939632546, "explain_loss": 1.171875, "loss": 1.875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0115966796875, "step": 1860 }, { "ce_loss": 0.546875, "distill_loss": 0.0751953125, "epoch": 0.8136482939632546, "explain_loss": 1.125, "loss": 2.3125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.035400390625, "step": 1860 }, { "ce_loss": 0.035888671875, "distill_loss": 0.06298828125, "epoch": 0.8136482939632546, "explain_loss": 1.2421875, "loss": 1.765625, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.01373291015625, "step": 1860 }, { "ce_loss": 0.31640625, "distill_loss": 0.051025390625, "epoch": 0.8136482939632546, "explain_loss": 1.171875, "loss": 1.84375, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.015625, "step": 1860 }, { "epoch": 0.8180227471566054, "grad_norm": 0.8133080005645752, "learning_rate": 0.0001741259699343649, "loss": 1.9248046875, "step": 1870 }, { "ce_loss": 0.26953125, "distill_loss": 0.05712890625, "epoch": 0.8180227471566054, "explain_loss": 1.21875, "loss": 2.125, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.00994873046875, "step": 1870 }, { "ce_loss": 0.10009765625, "distill_loss": 0.064453125, "epoch": 0.8180227471566054, "explain_loss": 1.1484375, "loss": 1.8125, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0108642578125, "step": 1870 }, { "ce_loss": 0.01708984375, "distill_loss": 0.0517578125, "epoch": 0.8180227471566054, "explain_loss": 1.234375, "loss": 1.8515625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.008544921875, "step": 1870 }, { "ce_loss": 0.08984375, "distill_loss": 0.0693359375, "epoch": 0.8180227471566054, "explain_loss": 0.91796875, "loss": 1.578125, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0103759765625, "step": 1870 }, { "epoch": 0.8223972003499562, "grad_norm": 0.7855452299118042, "learning_rate": 0.0001738014417520411, "loss": 1.904296875, "step": 1880 }, { "ce_loss": 0.007659912109375, "distill_loss": 0.0576171875, "epoch": 0.8223972003499562, "explain_loss": 2.15625, "loss": 2.6875, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.00653076171875, "step": 1880 }, { "ce_loss": 0.435546875, "distill_loss": 0.0712890625, "epoch": 0.8223972003499562, "explain_loss": 1.109375, "loss": 2.296875, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.046142578125, "step": 1880 }, { "ce_loss": 1.1640625, "distill_loss": 0.064453125, "epoch": 0.8223972003499562, "explain_loss": 1.015625, "loss": 2.84375, "ref_ce_loss": 0.546875, "stage_align_loss": 0.052001953125, "step": 1880 }, { "ce_loss": 0.06591796875, "distill_loss": 0.072265625, "epoch": 0.8223972003499562, "explain_loss": 0.984375, "loss": 1.484375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.01104736328125, "step": 1880 }, { "epoch": 0.8267716535433071, "grad_norm": 0.6049206852912903, "learning_rate": 0.0001734751975000701, "loss": 1.90185546875, "step": 1890 }, { "ce_loss": 0.06787109375, "distill_loss": 0.0556640625, "epoch": 0.8267716535433071, "explain_loss": 1.484375, "loss": 2.015625, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.010498046875, "step": 1890 }, { "ce_loss": 0.1669921875, "distill_loss": 0.07421875, "epoch": 0.8267716535433071, "explain_loss": 1.234375, "loss": 1.8984375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.01373291015625, "step": 1890 }, { "ce_loss": 0.125, "distill_loss": 0.0634765625, "epoch": 0.8267716535433071, "explain_loss": 0.875, "loss": 1.34375, "ref_ce_loss": 0.267578125, "stage_align_loss": 0.01190185546875, "step": 1890 }, { "ce_loss": 0.0303955078125, "distill_loss": 0.07275390625, "epoch": 0.8267716535433071, "explain_loss": 1.078125, "loss": 1.671875, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.007476806640625, "step": 1890 }, { "epoch": 0.8311461067366579, "grad_norm": 0.6109294295310974, "learning_rate": 0.00017314724476445354, "loss": 1.9875, "step": 1900 }, { "ce_loss": 0.068359375, "distill_loss": 0.061279296875, "epoch": 0.8311461067366579, "explain_loss": 1.359375, "loss": 1.9140625, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.01226806640625, "step": 1900 }, { "ce_loss": 0.00457763671875, "distill_loss": 0.08544921875, "epoch": 0.8311461067366579, "explain_loss": 1.2109375, "loss": 1.703125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.00927734375, "step": 1900 }, { "ce_loss": 0.55078125, "distill_loss": 0.0576171875, "epoch": 0.8311461067366579, "explain_loss": 1.2421875, "loss": 2.3125, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.01348876953125, "step": 1900 }, { "ce_loss": 0.0167236328125, "distill_loss": 0.0712890625, "epoch": 0.8311461067366579, "explain_loss": 1.25, "loss": 2.046875, "ref_ce_loss": 0.7109375, "stage_align_loss": 0.00531005859375, "step": 1900 }, { "epoch": 0.8355205599300087, "grad_norm": 0.8560067415237427, "learning_rate": 0.0001728175911709196, "loss": 1.9466796875, "step": 1910 }, { "ce_loss": 0.43359375, "distill_loss": 0.0634765625, "epoch": 0.8355205599300087, "explain_loss": 0.984375, "loss": 1.9375, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0712890625, "step": 1910 }, { "ce_loss": 0.90625, "distill_loss": 0.06494140625, "epoch": 0.8355205599300087, "explain_loss": 0.94140625, "loss": 2.5625, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.080078125, "step": 1910 }, { "ce_loss": 0.1474609375, "distill_loss": 0.0986328125, "epoch": 0.8355205599300087, "explain_loss": 1.328125, "loss": 2.03125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.018798828125, "step": 1910 }, { "ce_loss": 0.796875, "distill_loss": 0.0703125, "epoch": 0.8355205599300087, "explain_loss": 1.0703125, "loss": 2.421875, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0096435546875, "step": 1910 }, { "epoch": 0.8398950131233596, "grad_norm": 0.6663600206375122, "learning_rate": 0.0001724862443847458, "loss": 1.94580078125, "step": 1920 }, { "ce_loss": 0.1640625, "distill_loss": 0.0712890625, "epoch": 0.8398950131233596, "explain_loss": 0.75, "loss": 1.375, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00628662109375, "step": 1920 }, { "ce_loss": 0.07275390625, "distill_loss": 0.06591796875, "epoch": 0.8398950131233596, "explain_loss": 1.3125, "loss": 1.859375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0098876953125, "step": 1920 }, { "ce_loss": 0.1669921875, "distill_loss": 0.0478515625, "epoch": 0.8398950131233596, "explain_loss": 1.1171875, "loss": 1.90625, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.028564453125, "step": 1920 }, { "ce_loss": 0.1494140625, "distill_loss": 0.0693359375, "epoch": 0.8398950131233596, "explain_loss": 1.125, "loss": 1.890625, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0098876953125, "step": 1920 }, { "epoch": 0.8442694663167104, "grad_norm": 7.366171836853027, "learning_rate": 0.00017215321211058043, "loss": 1.99482421875, "step": 1930 }, { "ce_loss": 0.30859375, "distill_loss": 0.06591796875, "epoch": 0.8442694663167104, "explain_loss": 1.078125, "loss": 2.0625, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.031494140625, "step": 1930 }, { "ce_loss": 0.341796875, "distill_loss": 0.06396484375, "epoch": 0.8442694663167104, "explain_loss": 1.359375, "loss": 2.25, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.00958251953125, "step": 1930 }, { "ce_loss": 0.3359375, "distill_loss": 0.06298828125, "epoch": 0.8442694663167104, "explain_loss": 1.15625, "loss": 2.0625, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.016845703125, "step": 1930 }, { "ce_loss": 0.06494140625, "distill_loss": 0.06640625, "epoch": 0.8442694663167104, "explain_loss": 1.09375, "loss": 1.6171875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.01318359375, "step": 1930 }, { "epoch": 0.8486439195100612, "grad_norm": 0.7392945885658264, "learning_rate": 0.00017181850209226386, "loss": 2.06953125, "step": 1940 }, { "ce_loss": 0.05859375, "distill_loss": 0.07177734375, "epoch": 0.8486439195100612, "explain_loss": 1.28125, "loss": 2.03125, "ref_ce_loss": 0.609375, "stage_align_loss": 0.004730224609375, "step": 1940 }, { "ce_loss": 0.2412109375, "distill_loss": 0.06640625, "epoch": 0.8486439195100612, "explain_loss": 1.0390625, "loss": 1.8984375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.00958251953125, "step": 1940 }, { "ce_loss": 0.28125, "distill_loss": 0.0693359375, "epoch": 0.8486439195100612, "explain_loss": 1.296875, "loss": 2.03125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.02294921875, "step": 1940 }, { "ce_loss": 0.7734375, "distill_loss": 0.07275390625, "epoch": 0.8486439195100612, "explain_loss": 1.2109375, "loss": 2.53125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.0084228515625, "step": 1940 }, { "epoch": 0.8530183727034121, "grad_norm": 0.5502923727035522, "learning_rate": 0.00017148212211264815, "loss": 1.9470703125, "step": 1950 }, { "ce_loss": 0.047119140625, "distill_loss": 0.076171875, "epoch": 0.8530183727034121, "explain_loss": 1.3046875, "loss": 1.828125, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.0050048828125, "step": 1950 }, { "ce_loss": 0.4921875, "distill_loss": 0.06591796875, "epoch": 0.8530183727034121, "explain_loss": 1.171875, "loss": 2.421875, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.10009765625, "step": 1950 }, { "ce_loss": 0.61328125, "distill_loss": 0.059814453125, "epoch": 0.8530183727034121, "explain_loss": 1.09375, "loss": 2.1875, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.0084228515625, "step": 1950 }, { "ce_loss": 0.275390625, "distill_loss": 0.080078125, "epoch": 0.8530183727034121, "explain_loss": 0.921875, "loss": 1.890625, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.029296875, "step": 1950 }, { "epoch": 0.8573928258967629, "grad_norm": 0.7845126986503601, "learning_rate": 0.00017114407999341616, "loss": 1.89599609375, "step": 1960 }, { "ce_loss": 0.10205078125, "distill_loss": 0.056640625, "epoch": 0.8573928258967629, "explain_loss": 1.1875, "loss": 1.796875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.011474609375, "step": 1960 }, { "ce_loss": 0.0032806396484375, "distill_loss": 0.06787109375, "epoch": 0.8573928258967629, "explain_loss": 1.21875, "loss": 1.6796875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00543212890625, "step": 1960 }, { "ce_loss": 0.00185394287109375, "distill_loss": 0.06787109375, "epoch": 0.8573928258967629, "explain_loss": 0.9609375, "loss": 1.53125, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.00653076171875, "step": 1960 }, { "ce_loss": 0.00148773193359375, "distill_loss": 0.08251953125, "epoch": 0.8573928258967629, "explain_loss": 0.96875, "loss": 1.453125, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.006378173828125, "step": 1960 }, { "epoch": 0.8617672790901137, "grad_norm": 0.5797335505485535, "learning_rate": 0.00017080438359489967, "loss": 1.88818359375, "step": 1970 }, { "ce_loss": 0.006011962890625, "distill_loss": 0.0654296875, "epoch": 0.8617672790901137, "explain_loss": 1.046875, "loss": 1.421875, "ref_ce_loss": 0.296875, "stage_align_loss": 0.0081787109375, "step": 1970 }, { "ce_loss": 0.1953125, "distill_loss": 0.08984375, "epoch": 0.8617672790901137, "explain_loss": 1.0, "loss": 1.6875, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0101318359375, "step": 1970 }, { "ce_loss": 0.1806640625, "distill_loss": 0.0947265625, "epoch": 0.8617672790901137, "explain_loss": 1.0, "loss": 1.7265625, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0145263671875, "step": 1970 }, { "ce_loss": 0.1708984375, "distill_loss": 0.0673828125, "epoch": 0.8617672790901137, "explain_loss": 1.2421875, "loss": 2.015625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.007354736328125, "step": 1970 }, { "epoch": 0.8661417322834646, "grad_norm": 0.7001150846481323, "learning_rate": 0.0001704630408158967, "loss": 1.881884765625, "step": 1980 }, { "ce_loss": 0.2255859375, "distill_loss": 0.080078125, "epoch": 0.8661417322834646, "explain_loss": 1.0390625, "loss": 1.828125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.0179443359375, "step": 1980 }, { "ce_loss": 0.2021484375, "distill_loss": 0.08056640625, "epoch": 0.8661417322834646, "explain_loss": 1.3125, "loss": 2.109375, "ref_ce_loss": 0.5, "stage_align_loss": 0.0152587890625, "step": 1980 }, { "ce_loss": 0.0673828125, "distill_loss": 0.0478515625, "epoch": 0.8661417322834646, "explain_loss": 0.765625, "loss": 1.2890625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0091552734375, "step": 1980 }, { "ce_loss": 0.53515625, "distill_loss": 0.05419921875, "epoch": 0.8661417322834646, "explain_loss": 1.125, "loss": 2.34375, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.050048828125, "step": 1980 }, { "epoch": 0.8705161854768154, "grad_norm": 0.6437104344367981, "learning_rate": 0.0001701200595934877, "loss": 1.92265625, "step": 1990 }, { "ce_loss": 0.0142822265625, "distill_loss": 0.07275390625, "epoch": 0.8705161854768154, "explain_loss": 1.1875, "loss": 1.8359375, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.007568359375, "step": 1990 }, { "ce_loss": 0.091796875, "distill_loss": 0.06396484375, "epoch": 0.8705161854768154, "explain_loss": 1.4375, "loss": 1.9296875, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.0194091796875, "step": 1990 }, { "ce_loss": 0.318359375, "distill_loss": 0.05615234375, "epoch": 0.8705161854768154, "explain_loss": 1.1484375, "loss": 2.015625, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0341796875, "step": 1990 }, { "ce_loss": 0.9609375, "distill_loss": 0.05908203125, "epoch": 0.8705161854768154, "explain_loss": 1.359375, "loss": 2.984375, "ref_ce_loss": 0.546875, "stage_align_loss": 0.057861328125, "step": 1990 }, { "epoch": 0.8748906386701663, "grad_norm": 0.5112404227256775, "learning_rate": 0.00016977544790285114, "loss": 2.0146484375, "step": 2000 }, { "ce_loss": 0.1298828125, "distill_loss": 0.0615234375, "epoch": 0.8748906386701663, "explain_loss": 1.0625, "loss": 1.7734375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.00823974609375, "step": 2000 }, { "ce_loss": 0.2109375, "distill_loss": 0.064453125, "epoch": 0.8748906386701663, "explain_loss": 1.0625, "loss": 1.78125, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.017333984375, "step": 2000 }, { "ce_loss": 0.275390625, "distill_loss": 0.0517578125, "epoch": 0.8748906386701663, "explain_loss": 1.25, "loss": 1.96875, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.0146484375, "step": 2000 }, { "ce_loss": 0.095703125, "distill_loss": 0.055908203125, "epoch": 0.8748906386701663, "explain_loss": 1.0546875, "loss": 1.625, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.01806640625, "step": 2000 }, { "epoch": 0.8792650918635171, "grad_norm": 0.7234396934509277, "learning_rate": 0.00016942921375707784, "loss": 1.92158203125, "step": 2010 }, { "ce_loss": 0.388671875, "distill_loss": 0.078125, "epoch": 0.8792650918635171, "explain_loss": 1.3203125, "loss": 2.15625, "ref_ce_loss": 0.359375, "stage_align_loss": 0.00946044921875, "step": 2010 }, { "ce_loss": 0.1796875, "distill_loss": 0.06396484375, "epoch": 0.8792650918635171, "explain_loss": 1.3046875, "loss": 1.828125, "ref_ce_loss": 0.26953125, "stage_align_loss": 0.00689697265625, "step": 2010 }, { "ce_loss": 0.0361328125, "distill_loss": 0.07275390625, "epoch": 0.8792650918635171, "explain_loss": 0.8046875, "loss": 1.28125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00634765625, "step": 2010 }, { "ce_loss": 0.2099609375, "distill_loss": 0.078125, "epoch": 0.8792650918635171, "explain_loss": 1.046875, "loss": 1.78125, "ref_ce_loss": 0.421875, "stage_align_loss": 0.02880859375, "step": 2010 }, { "epoch": 0.8836395450568679, "grad_norm": 0.7055696845054626, "learning_rate": 0.00016908136520698498, "loss": 1.9787109375, "step": 2020 }, { "ce_loss": 0.357421875, "distill_loss": 0.06298828125, "epoch": 0.8836395450568679, "explain_loss": 1.046875, "loss": 1.8828125, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.0230712890625, "step": 2020 }, { "ce_loss": 0.6328125, "distill_loss": 0.076171875, "epoch": 0.8836395450568679, "explain_loss": 1.1953125, "loss": 2.40625, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.04052734375, "step": 2020 }, { "ce_loss": 0.298828125, "distill_loss": 0.062255859375, "epoch": 0.8836395450568679, "explain_loss": 0.9765625, "loss": 1.6875, "ref_ce_loss": 0.3125, "stage_align_loss": 0.03369140625, "step": 2020 }, { "ce_loss": 0.32421875, "distill_loss": 0.06201171875, "epoch": 0.8836395450568679, "explain_loss": 1.125, "loss": 1.6875, "ref_ce_loss": 0.166015625, "stage_align_loss": 0.01324462890625, "step": 2020 }, { "epoch": 0.8880139982502188, "grad_norm": 1.0296227931976318, "learning_rate": 0.00016873191034092852, "loss": 1.998046875, "step": 2030 }, { "ce_loss": 0.057861328125, "distill_loss": 0.05419921875, "epoch": 0.8880139982502188, "explain_loss": 1.234375, "loss": 1.734375, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.0093994140625, "step": 2030 }, { "ce_loss": 0.267578125, "distill_loss": 0.0849609375, "epoch": 0.8880139982502188, "explain_loss": 1.046875, "loss": 2.0625, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.033203125, "step": 2030 }, { "ce_loss": 0.625, "distill_loss": 0.059814453125, "epoch": 0.8880139982502188, "explain_loss": 1.609375, "loss": 2.65625, "ref_ce_loss": 0.302734375, "stage_align_loss": 0.0625, "step": 2030 }, { "ce_loss": 0.1962890625, "distill_loss": 0.07080078125, "epoch": 0.8880139982502188, "explain_loss": 1.2421875, "loss": 2.125, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.01263427734375, "step": 2030 }, { "epoch": 0.8923884514435696, "grad_norm": 0.6358256340026855, "learning_rate": 0.0001683808572846155, "loss": 2.0140625, "step": 2040 }, { "ce_loss": 0.036865234375, "distill_loss": 0.06787109375, "epoch": 0.8923884514435696, "explain_loss": 1.109375, "loss": 1.703125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.00811767578125, "step": 2040 }, { "ce_loss": 0.357421875, "distill_loss": 0.0595703125, "epoch": 0.8923884514435696, "explain_loss": 1.15625, "loss": 2.015625, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.005523681640625, "step": 2040 }, { "ce_loss": 0.34375, "distill_loss": 0.07861328125, "epoch": 0.8923884514435696, "explain_loss": 0.90625, "loss": 1.78125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.018798828125, "step": 2040 }, { "ce_loss": 0.01220703125, "distill_loss": 0.1005859375, "epoch": 0.8923884514435696, "explain_loss": 1.234375, "loss": 1.734375, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00518798828125, "step": 2040 }, { "epoch": 0.8967629046369204, "grad_norm": 8.303781509399414, "learning_rate": 0.00016802821420091482, "loss": 1.87255859375, "step": 2050 }, { "ce_loss": 0.71484375, "distill_loss": 0.0654296875, "epoch": 0.8967629046369204, "explain_loss": 1.046875, "loss": 2.375, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.01251220703125, "step": 2050 }, { "ce_loss": 1.0078125, "distill_loss": 0.11328125, "epoch": 0.8967629046369204, "explain_loss": 1.5, "loss": 3.125, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.03857421875, "step": 2050 }, { "ce_loss": 1.5078125, "distill_loss": 0.06396484375, "epoch": 0.8967629046369204, "explain_loss": 1.390625, "loss": 3.390625, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.05908203125, "step": 2050 }, { "ce_loss": 0.193359375, "distill_loss": 0.103515625, "epoch": 0.8967629046369204, "explain_loss": 1.1640625, "loss": 2.109375, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.0311279296875, "step": 2050 }, { "epoch": 0.9011373578302713, "grad_norm": 2.553049325942993, "learning_rate": 0.0001676739892896675, "loss": 2.17138671875, "step": 2060 }, { "ce_loss": 0.099609375, "distill_loss": 0.0810546875, "epoch": 0.9011373578302713, "explain_loss": 0.9765625, "loss": 1.53125, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.00555419921875, "step": 2060 }, { "ce_loss": 0.33203125, "distill_loss": 0.0869140625, "epoch": 0.9011373578302713, "explain_loss": 1.0, "loss": 1.859375, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.036376953125, "step": 2060 }, { "ce_loss": 0.0111083984375, "distill_loss": 0.09765625, "epoch": 0.9011373578302713, "explain_loss": 0.7578125, "loss": 1.2890625, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.003875732421875, "step": 2060 }, { "ce_loss": 0.041748046875, "distill_loss": 0.0673828125, "epoch": 0.9011373578302713, "explain_loss": 1.03125, "loss": 1.4296875, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.007354736328125, "step": 2060 }, { "epoch": 0.905511811023622, "grad_norm": 0.7098190188407898, "learning_rate": 0.00016731819078749614, "loss": 2.03154296875, "step": 2070 }, { "ce_loss": 0.078125, "distill_loss": 0.07666015625, "epoch": 0.905511811023622, "explain_loss": 1.546875, "loss": 2.265625, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.0052490234375, "step": 2070 }, { "ce_loss": 0.00994873046875, "distill_loss": 0.083984375, "epoch": 0.905511811023622, "explain_loss": 1.65625, "loss": 2.328125, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.004547119140625, "step": 2070 }, { "ce_loss": 0.0247802734375, "distill_loss": 0.08056640625, "epoch": 0.905511811023622, "explain_loss": 0.85546875, "loss": 1.2890625, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.007080078125, "step": 2070 }, { "ce_loss": 0.023193359375, "distill_loss": 0.09326171875, "epoch": 0.905511811023622, "explain_loss": 1.0, "loss": 1.6328125, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.00616455078125, "step": 2070 }, { "epoch": 0.9098862642169728, "grad_norm": 0.6659917235374451, "learning_rate": 0.00016696082696761314, "loss": 1.87626953125, "step": 2080 }, { "ce_loss": 0.326171875, "distill_loss": 0.07666015625, "epoch": 0.9098862642169728, "explain_loss": 1.28125, "loss": 2.25, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.048095703125, "step": 2080 }, { "ce_loss": 0.002685546875, "distill_loss": 0.0673828125, "epoch": 0.9098862642169728, "explain_loss": 0.71875, "loss": 1.21875, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.00439453125, "step": 2080 }, { "ce_loss": 0.06396484375, "distill_loss": 0.06201171875, "epoch": 0.9098862642169728, "explain_loss": 1.421875, "loss": 1.9765625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00628662109375, "step": 2080 }, { "ce_loss": 0.318359375, "distill_loss": 0.0693359375, "epoch": 0.9098862642169728, "explain_loss": 0.9296875, "loss": 1.8828125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0302734375, "step": 2080 }, { "epoch": 0.9142607174103237, "grad_norm": 0.6997997164726257, "learning_rate": 0.0001666019061396287, "loss": 1.8466796875, "step": 2090 }, { "ce_loss": 0.007293701171875, "distill_loss": 0.06396484375, "epoch": 0.9142607174103237, "explain_loss": 1.28125, "loss": 1.703125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.00360107421875, "step": 2090 }, { "ce_loss": 0.030517578125, "distill_loss": 0.064453125, "epoch": 0.9142607174103237, "explain_loss": 0.9375, "loss": 1.484375, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.0048828125, "step": 2090 }, { "ce_loss": 0.1416015625, "distill_loss": 0.06298828125, "epoch": 0.9142607174103237, "explain_loss": 1.5, "loss": 2.171875, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.00775146484375, "step": 2090 }, { "ce_loss": 0.67578125, "distill_loss": 0.06640625, "epoch": 0.9142607174103237, "explain_loss": 1.0703125, "loss": 2.40625, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.0169677734375, "step": 2090 }, { "epoch": 0.9186351706036745, "grad_norm": 0.5873184204101562, "learning_rate": 0.00016624143664935716, "loss": 1.91259765625, "step": 2100 }, { "ce_loss": 0.240234375, "distill_loss": 0.0615234375, "epoch": 0.9186351706036745, "explain_loss": 0.9765625, "loss": 1.625, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.0079345703125, "step": 2100 }, { "ce_loss": 0.02685546875, "distill_loss": 0.060791015625, "epoch": 0.9186351706036745, "explain_loss": 1.28125, "loss": 1.8359375, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.005859375, "step": 2100 }, { "ce_loss": 0.138671875, "distill_loss": 0.0478515625, "epoch": 0.9186351706036745, "explain_loss": 1.234375, "loss": 1.7734375, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.005950927734375, "step": 2100 }, { "ce_loss": 0.470703125, "distill_loss": 0.07373046875, "epoch": 0.9186351706036745, "explain_loss": 1.15625, "loss": 2.125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.005889892578125, "step": 2100 }, { "epoch": 0.9230096237970253, "grad_norm": 0.678631067276001, "learning_rate": 0.00016587942687862334, "loss": 1.83896484375, "step": 2110 }, { "ce_loss": 0.07763671875, "distill_loss": 0.060546875, "epoch": 0.9230096237970253, "explain_loss": 1.0078125, "loss": 1.671875, "ref_ce_loss": 0.515625, "stage_align_loss": 0.005523681640625, "step": 2110 }, { "ce_loss": 0.45703125, "distill_loss": 0.076171875, "epoch": 0.9230096237970253, "explain_loss": 1.1171875, "loss": 2.3125, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.037109375, "step": 2110 }, { "ce_loss": 0.2255859375, "distill_loss": 0.06640625, "epoch": 0.9230096237970253, "explain_loss": 1.34375, "loss": 2.03125, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.00921630859375, "step": 2110 }, { "ce_loss": 1.078125, "distill_loss": 0.05126953125, "epoch": 0.9230096237970253, "explain_loss": 1.1875, "loss": 2.84375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0054931640625, "step": 2110 }, { "epoch": 0.9273840769903762, "grad_norm": 0.7841988205909729, "learning_rate": 0.00016551588524506736, "loss": 2.02255859375, "step": 2120 }, { "ce_loss": 0.3828125, "distill_loss": 0.08349609375, "epoch": 0.9273840769903762, "explain_loss": 1.109375, "loss": 2.03125, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.040771484375, "step": 2120 }, { "ce_loss": 0.02001953125, "distill_loss": 0.060546875, "epoch": 0.9273840769903762, "explain_loss": 0.9765625, "loss": 1.625, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.004547119140625, "step": 2120 }, { "ce_loss": 0.197265625, "distill_loss": 0.055908203125, "epoch": 0.9273840769903762, "explain_loss": 1.484375, "loss": 2.359375, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.00579833984375, "step": 2120 }, { "ce_loss": 0.0888671875, "distill_loss": 0.08349609375, "epoch": 0.9273840769903762, "explain_loss": 1.328125, "loss": 1.921875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.005859375, "step": 2120 }, { "epoch": 0.931758530183727, "grad_norm": 1.022707223892212, "learning_rate": 0.00016515082020194903, "loss": 1.84248046875, "step": 2130 }, { "ce_loss": 0.1337890625, "distill_loss": 0.0576171875, "epoch": 0.931758530183727, "explain_loss": 0.83984375, "loss": 1.421875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.01031494140625, "step": 2130 }, { "ce_loss": 0.01129150390625, "distill_loss": 0.07421875, "epoch": 0.931758530183727, "explain_loss": 0.95703125, "loss": 1.671875, "ref_ce_loss": 0.59375, "stage_align_loss": 0.03125, "step": 2130 }, { "ce_loss": 0.09765625, "distill_loss": 0.05859375, "epoch": 0.931758530183727, "explain_loss": 1.421875, "loss": 1.9296875, "ref_ce_loss": 0.34375, "stage_align_loss": 0.00640869140625, "step": 2130 }, { "ce_loss": 0.006561279296875, "distill_loss": 0.060791015625, "epoch": 0.931758530183727, "explain_loss": 1.265625, "loss": 1.8046875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.004669189453125, "step": 2130 }, { "epoch": 0.9361329833770778, "grad_norm": 5.4409260749816895, "learning_rate": 0.00016478424023795137, "loss": 2.19697265625, "step": 2140 }, { "ce_loss": 0.236328125, "distill_loss": 0.0654296875, "epoch": 0.9361329833770778, "explain_loss": 1.265625, "loss": 2.84375, "ref_ce_loss": 1.265625, "stage_align_loss": 0.00921630859375, "step": 2140 }, { "ce_loss": 0.474609375, "distill_loss": 0.06884765625, "epoch": 0.9361329833770778, "explain_loss": 1.53125, "loss": 3.3125, "ref_ce_loss": 1.2265625, "stage_align_loss": 0.0257568359375, "step": 2140 }, { "ce_loss": 0.65625, "distill_loss": 0.064453125, "epoch": 0.9361329833770778, "explain_loss": 1.2890625, "loss": 3.15625, "ref_ce_loss": 1.09375, "stage_align_loss": 0.05517578125, "step": 2140 }, { "ce_loss": 0.0224609375, "distill_loss": 0.06689453125, "epoch": 0.9361329833770778, "explain_loss": 1.140625, "loss": 2.84375, "ref_ce_loss": 1.6171875, "stage_align_loss": 0.00335693359375, "step": 2140 }, { "epoch": 0.9405074365704287, "grad_norm": 35.448116302490234, "learning_rate": 0.00016441615387698295, "loss": 2.72509765625, "step": 2150 }, { "ce_loss": 0.09326171875, "distill_loss": 0.10498046875, "epoch": 0.9405074365704287, "explain_loss": 1.1875, "loss": 3.90625, "ref_ce_loss": 2.515625, "stage_align_loss": 0.00494384765625, "step": 2150 }, { "ce_loss": 0.1875, "distill_loss": 0.08740234375, "epoch": 0.9405074365704287, "explain_loss": 1.40625, "loss": 2.9375, "ref_ce_loss": 1.25, "stage_align_loss": 0.007598876953125, "step": 2150 }, { "ce_loss": 0.796875, "distill_loss": 0.0771484375, "epoch": 0.9405074365704287, "explain_loss": 1.1875, "loss": 4.375, "ref_ce_loss": 2.28125, "stage_align_loss": 0.016357421875, "step": 2150 }, { "ce_loss": 0.58203125, "distill_loss": 0.06787109375, "epoch": 0.9405074365704287, "explain_loss": 1.2265625, "loss": 3.875, "ref_ce_loss": 1.9921875, "stage_align_loss": 0.0135498046875, "step": 2150 }, { "epoch": 0.9448818897637795, "grad_norm": 286.4322204589844, "learning_rate": 0.00016404656967798005, "loss": 17.41953125, "step": 2160 }, { "ce_loss": 2.0625, "distill_loss": 0.1767578125, "epoch": 0.9448818897637795, "explain_loss": 3.9375, "loss": 17.25, "ref_ce_loss": 11.0625, "stage_align_loss": 0.01275634765625, "step": 2160 }, { "ce_loss": 16.25, "distill_loss": 0.138671875, "epoch": 0.9448818897637795, "explain_loss": 3.84375, "loss": 31.125, "ref_ce_loss": 10.875, "stage_align_loss": 0.0230712890625, "step": 2160 }, { "ce_loss": 2.046875, "distill_loss": 0.19140625, "epoch": 0.9448818897637795, "explain_loss": 4.375, "loss": 18.0, "ref_ce_loss": 11.1875, "stage_align_loss": 0.19140625, "step": 2160 }, { "ce_loss": 11.75, "distill_loss": 0.12451171875, "epoch": 0.9448818897637795, "explain_loss": 3.875, "loss": 27.25, "ref_ce_loss": 11.5, "stage_align_loss": 0.0224609375, "step": 2160 }, { "epoch": 0.9492563429571304, "grad_norm": 21.92824363708496, "learning_rate": 0.00016367549623470736, "loss": 16.99765625, "step": 2170 }, { "ce_loss": 0.216796875, "distill_loss": 0.14453125, "epoch": 0.9492563429571304, "explain_loss": 3.078125, "loss": 13.5, "ref_ce_loss": 9.75, "stage_align_loss": 0.326171875, "step": 2170 }, { "ce_loss": 2.296875, "distill_loss": 0.11328125, "epoch": 0.9492563429571304, "explain_loss": 3.53125, "loss": 16.25, "ref_ce_loss": 10.0625, "stage_align_loss": 0.169921875, "step": 2170 }, { "ce_loss": 2.625, "distill_loss": 0.1015625, "epoch": 0.9492563429571304, "explain_loss": 3.1875, "loss": 16.375, "ref_ce_loss": 10.25, "stage_align_loss": 0.18359375, "step": 2170 }, { "ce_loss": 2.375, "distill_loss": 0.107421875, "epoch": 0.9492563429571304, "explain_loss": 3.515625, "loss": 16.375, "ref_ce_loss": 10.375, "stage_align_loss": 0.00946044921875, "step": 2170 }, { "epoch": 0.9536307961504812, "grad_norm": 32.231571197509766, "learning_rate": 0.00016330294217555826, "loss": 13.51796875, "step": 2180 }, { "ce_loss": 0.71875, "distill_loss": 0.15234375, "epoch": 0.9536307961504812, "explain_loss": 3.0625, "loss": 12.5, "ref_ce_loss": 7.71875, "stage_align_loss": 0.8515625, "step": 2180 }, { "ce_loss": 0.13671875, "distill_loss": 0.154296875, "epoch": 0.9536307961504812, "explain_loss": 2.8125, "loss": 10.1875, "ref_ce_loss": 6.28125, "stage_align_loss": 0.8203125, "step": 2180 }, { "ce_loss": 2.875, "distill_loss": 0.1201171875, "epoch": 0.9536307961504812, "explain_loss": 2.390625, "loss": 13.625, "ref_ce_loss": 7.625, "stage_align_loss": 0.59375, "step": 2180 }, { "ce_loss": 2.6875, "distill_loss": 0.1474609375, "epoch": 0.9536307961504812, "explain_loss": 2.5, "loss": 13.5, "ref_ce_loss": 7.28125, "stage_align_loss": 0.87890625, "step": 2180 }, { "epoch": 0.958005249343832, "grad_norm": 5.419594764709473, "learning_rate": 0.00016292891616335423, "loss": 11.5390625, "step": 2190 }, { "ce_loss": 2.046875, "distill_loss": 0.11669921875, "epoch": 0.958005249343832, "explain_loss": 2.5, "loss": 11.375, "ref_ce_loss": 6.65625, "stage_align_loss": 0.08984375, "step": 2190 }, { "ce_loss": 2.03125, "distill_loss": 0.109375, "epoch": 0.958005249343832, "explain_loss": 3.0, "loss": 12.5, "ref_ce_loss": 7.21875, "stage_align_loss": 0.1337890625, "step": 2190 }, { "ce_loss": 0.359375, "distill_loss": 0.1357421875, "epoch": 0.958005249343832, "explain_loss": 2.578125, "loss": 9.8125, "ref_ce_loss": 6.65625, "stage_align_loss": 0.08740234375, "step": 2190 }, { "ce_loss": 2.09375, "distill_loss": 0.107421875, "epoch": 0.958005249343832, "explain_loss": 2.09375, "loss": 10.25, "ref_ce_loss": 5.84375, "stage_align_loss": 0.1298828125, "step": 2190 }, { "epoch": 0.9623797025371829, "grad_norm": 4.884639263153076, "learning_rate": 0.00016255342689514333, "loss": 10.8171875, "step": 2200 }, { "ce_loss": 1.4609375, "distill_loss": 0.1240234375, "epoch": 0.9623797025371829, "explain_loss": 2.25, "loss": 9.6875, "ref_ce_loss": 5.8125, "stage_align_loss": 0.0196533203125, "step": 2200 }, { "ce_loss": 1.59375, "distill_loss": 0.107421875, "epoch": 0.9623797025371829, "explain_loss": 2.15625, "loss": 10.25, "ref_ce_loss": 6.40625, "stage_align_loss": 0.0135498046875, "step": 2200 }, { "ce_loss": 0.322265625, "distill_loss": 0.109375, "epoch": 0.9623797025371829, "explain_loss": 2.09375, "loss": 7.9375, "ref_ce_loss": 5.375, "stage_align_loss": 0.0169677734375, "step": 2200 }, { "ce_loss": 0.1611328125, "distill_loss": 0.115234375, "epoch": 0.9623797025371829, "explain_loss": 2.28125, "loss": 8.375, "ref_ce_loss": 5.75, "stage_align_loss": 0.0224609375, "step": 2200 }, { "epoch": 0.9667541557305337, "grad_norm": 17.130685806274414, "learning_rate": 0.00016217648310199804, "loss": 9.474609375, "step": 2210 }, { "ce_loss": 0.0140380859375, "distill_loss": 0.087890625, "epoch": 0.9667541557305337, "explain_loss": 2.15625, "loss": 8.0, "ref_ce_loss": 5.71875, "stage_align_loss": 0.023193359375, "step": 2210 }, { "ce_loss": 1.65625, "distill_loss": 0.087890625, "epoch": 0.9667541557305337, "explain_loss": 1.984375, "loss": 7.5625, "ref_ce_loss": 3.78125, "stage_align_loss": 0.03759765625, "step": 2210 }, { "ce_loss": 1.90625, "distill_loss": 0.08349609375, "epoch": 0.9667541557305337, "explain_loss": 2.375, "loss": 9.8125, "ref_ce_loss": 5.40625, "stage_align_loss": 0.0260009765625, "step": 2210 }, { "ce_loss": 1.7109375, "distill_loss": 0.09814453125, "epoch": 0.9667541557305337, "explain_loss": 1.90625, "loss": 9.125, "ref_ce_loss": 5.40625, "stage_align_loss": 0.01953125, "step": 2210 }, { "epoch": 0.9711286089238845, "grad_norm": 3.848235607147217, "learning_rate": 0.00016179809354881214, "loss": 7.843359375, "step": 2220 }, { "ce_loss": 1.2890625, "distill_loss": 0.091796875, "epoch": 0.9711286089238845, "explain_loss": 1.203125, "loss": 6.5625, "ref_ce_loss": 3.953125, "stage_align_loss": 0.0120849609375, "step": 2220 }, { "ce_loss": 1.3359375, "distill_loss": 0.09228515625, "epoch": 0.9711286089238845, "explain_loss": 1.453125, "loss": 7.125, "ref_ce_loss": 4.21875, "stage_align_loss": 0.01409912109375, "step": 2220 }, { "ce_loss": 1.1015625, "distill_loss": 0.09375, "epoch": 0.9711286089238845, "explain_loss": 1.5, "loss": 6.46875, "ref_ce_loss": 3.765625, "stage_align_loss": 0.0130615234375, "step": 2220 }, { "ce_loss": 1.0390625, "distill_loss": 0.08349609375, "epoch": 0.9711286089238845, "explain_loss": 1.140625, "loss": 6.375, "ref_ce_loss": 4.0625, "stage_align_loss": 0.0299072265625, "step": 2220 }, { "epoch": 0.9755030621172354, "grad_norm": 14.601493835449219, "learning_rate": 0.00016141826703409712, "loss": 6.09921875, "step": 2230 }, { "ce_loss": 0.205078125, "distill_loss": 0.1416015625, "epoch": 0.9755030621172354, "explain_loss": 1.1875, "loss": 6.3125, "ref_ce_loss": 4.75, "stage_align_loss": 0.0341796875, "step": 2230 }, { "ce_loss": 1.3359375, "distill_loss": 0.1064453125, "epoch": 0.9755030621172354, "explain_loss": 1.21875, "loss": 7.375, "ref_ce_loss": 4.6875, "stage_align_loss": 0.034912109375, "step": 2230 }, { "ce_loss": 0.9375, "distill_loss": 0.09521484375, "epoch": 0.9755030621172354, "explain_loss": 1.4453125, "loss": 6.65625, "ref_ce_loss": 4.15625, "stage_align_loss": 0.0255126953125, "step": 2230 }, { "ce_loss": 0.83203125, "distill_loss": 0.08251953125, "epoch": 0.9755030621172354, "explain_loss": 1.3671875, "loss": 6.75, "ref_ce_loss": 4.4375, "stage_align_loss": 0.0203857421875, "step": 2230 }, { "epoch": 0.9798775153105862, "grad_norm": 4.137810707092285, "learning_rate": 0.00016103701238977735, "loss": 7.246875, "step": 2240 }, { "ce_loss": 0.0274658203125, "distill_loss": 0.099609375, "epoch": 0.9798775153105862, "explain_loss": 1.6796875, "loss": 6.84375, "ref_ce_loss": 5.03125, "stage_align_loss": 0.01007080078125, "step": 2240 }, { "ce_loss": 0.271484375, "distill_loss": 0.09521484375, "epoch": 0.9798775153105862, "explain_loss": 1.1328125, "loss": 7.03125, "ref_ce_loss": 5.53125, "stage_align_loss": 0.012939453125, "step": 2240 }, { "ce_loss": 0.01239013671875, "distill_loss": 0.1220703125, "epoch": 0.9798775153105862, "explain_loss": 1.46875, "loss": 6.6875, "ref_ce_loss": 5.09375, "stage_align_loss": 0.0103759765625, "step": 2240 }, { "ce_loss": 1.15625, "distill_loss": 0.07177734375, "epoch": 0.9798775153105862, "explain_loss": 1.75, "loss": 9.8125, "ref_ce_loss": 6.84375, "stage_align_loss": 0.009765625, "step": 2240 }, { "epoch": 0.984251968503937, "grad_norm": 84.7074203491211, "learning_rate": 0.0001606543384809848, "loss": 7.327734375, "step": 2250 }, { "ce_loss": 1.078125, "distill_loss": 0.0888671875, "epoch": 0.984251968503937, "explain_loss": 1.109375, "loss": 6.875, "ref_ce_loss": 4.5625, "stage_align_loss": 0.0341796875, "step": 2250 }, { "ce_loss": 0.76953125, "distill_loss": 0.0927734375, "epoch": 0.984251968503937, "explain_loss": 1.65625, "loss": 8.4375, "ref_ce_loss": 5.90625, "stage_align_loss": 0.01043701171875, "step": 2250 }, { "ce_loss": 1.3671875, "distill_loss": 0.076171875, "epoch": 0.984251968503937, "explain_loss": 1.4296875, "loss": 7.125, "ref_ce_loss": 4.21875, "stage_align_loss": 0.046142578125, "step": 2250 }, { "ce_loss": 0.455078125, "distill_loss": 0.080078125, "epoch": 0.984251968503937, "explain_loss": 1.25, "loss": 6.21875, "ref_ce_loss": 4.4375, "stage_align_loss": 0.00982666015625, "step": 2250 }, { "epoch": 0.9886264216972879, "grad_norm": 2.6714277267456055, "learning_rate": 0.00016027025420585296, "loss": 6.2578125, "step": 2260 }, { "ce_loss": 0.36328125, "distill_loss": 0.07080078125, "epoch": 0.9886264216972879, "explain_loss": 1.0, "loss": 5.3125, "ref_ce_loss": 3.84375, "stage_align_loss": 0.0169677734375, "step": 2260 }, { "ce_loss": 0.59765625, "distill_loss": 0.07763671875, "epoch": 0.9886264216972879, "explain_loss": 1.140625, "loss": 4.84375, "ref_ce_loss": 3.0, "stage_align_loss": 0.0380859375, "step": 2260 }, { "ce_loss": 0.0791015625, "distill_loss": 0.0908203125, "epoch": 0.9886264216972879, "explain_loss": 1.484375, "loss": 5.3125, "ref_ce_loss": 3.640625, "stage_align_loss": 0.007293701171875, "step": 2260 }, { "ce_loss": 0.06103515625, "distill_loss": 0.07177734375, "epoch": 0.9886264216972879, "explain_loss": 1.1171875, "loss": 4.8125, "ref_ce_loss": 3.546875, "stage_align_loss": 0.00933837890625, "step": 2260 }, { "epoch": 0.9930008748906387, "grad_norm": 43.705726623535156, "learning_rate": 0.0001598847684953099, "loss": 4.105078125, "step": 2270 }, { "ce_loss": 0.0145263671875, "distill_loss": 0.0712890625, "epoch": 0.9930008748906387, "explain_loss": 1.2734375, "loss": 3.28125, "ref_ce_loss": 1.90625, "stage_align_loss": 0.00933837890625, "step": 2270 }, { "ce_loss": 1.125, "distill_loss": 0.076171875, "epoch": 0.9930008748906387, "explain_loss": 1.015625, "loss": 4.0625, "ref_ce_loss": 1.75, "stage_align_loss": 0.0810546875, "step": 2270 }, { "ce_loss": 0.010498046875, "distill_loss": 0.0888671875, "epoch": 0.9930008748906387, "explain_loss": 1.1640625, "loss": 2.78125, "ref_ce_loss": 1.515625, "stage_align_loss": 0.00994873046875, "step": 2270 }, { "ce_loss": 0.236328125, "distill_loss": 0.078125, "epoch": 0.9930008748906387, "explain_loss": 0.9921875, "loss": 2.46875, "ref_ce_loss": 1.140625, "stage_align_loss": 0.033203125, "step": 2270 }, { "epoch": 0.9973753280839895, "grad_norm": 41.866371154785156, "learning_rate": 0.00015949789031287053, "loss": 2.684765625, "step": 2280 }, { "ce_loss": 0.09326171875, "distill_loss": 0.0693359375, "epoch": 0.9973753280839895, "explain_loss": 1.046875, "loss": 2.28125, "ref_ce_loss": 1.046875, "stage_align_loss": 0.0166015625, "step": 2280 }, { "ce_loss": 0.63671875, "distill_loss": 0.072265625, "epoch": 0.9973753280839895, "explain_loss": 1.1484375, "loss": 2.90625, "ref_ce_loss": 1.0, "stage_align_loss": 0.057861328125, "step": 2280 }, { "ce_loss": 0.2490234375, "distill_loss": 0.0859375, "epoch": 0.9973753280839895, "explain_loss": 1.1796875, "loss": 2.375, "ref_ce_loss": 0.84375, "stage_align_loss": 0.019287109375, "step": 2280 }, { "ce_loss": 0.0341796875, "distill_loss": 0.0732421875, "epoch": 0.9973753280839895, "explain_loss": 0.90625, "loss": 2.09375, "ref_ce_loss": 1.046875, "stage_align_loss": 0.021728515625, "step": 2280 }, { "epoch": 1.0017497812773404, "grad_norm": 1.71124267578125, "learning_rate": 0.00015910962865442834, "loss": 3.1806640625, "step": 2290 }, { "ce_loss": 0.00946044921875, "distill_loss": 0.064453125, "epoch": 1.0017497812773404, "explain_loss": 1.4375, "loss": 3.390625, "ref_ce_loss": 1.875, "stage_align_loss": 0.006927490234375, "step": 2290 }, { "ce_loss": 0.1904296875, "distill_loss": 0.0673828125, "epoch": 1.0017497812773404, "explain_loss": 0.890625, "loss": 2.375, "ref_ce_loss": 1.203125, "stage_align_loss": 0.0234375, "step": 2290 }, { "ce_loss": 0.1708984375, "distill_loss": 0.06640625, "epoch": 1.0017497812773404, "explain_loss": 1.296875, "loss": 2.96875, "ref_ce_loss": 1.421875, "stage_align_loss": 0.00799560546875, "step": 2290 }, { "ce_loss": 0.1064453125, "distill_loss": 0.0732421875, "epoch": 1.0017497812773404, "explain_loss": 0.9765625, "loss": 2.53125, "ref_ce_loss": 1.3671875, "stage_align_loss": 0.00933837890625, "step": 2290 }, { "epoch": 1.006124234470691, "grad_norm": 14.627141952514648, "learning_rate": 0.00015871999254804607, "loss": 3.0734375, "step": 2300 }, { "ce_loss": 0.6328125, "distill_loss": 0.080078125, "epoch": 1.006124234470691, "explain_loss": 1.078125, "loss": 3.25, "ref_ce_loss": 1.453125, "stage_align_loss": 0.01483154296875, "step": 2300 }, { "ce_loss": 0.7734375, "distill_loss": 0.06884765625, "epoch": 1.006124234470691, "explain_loss": 1.125, "loss": 3.765625, "ref_ce_loss": 1.7890625, "stage_align_loss": 0.01385498046875, "step": 2300 }, { "ce_loss": 1.6953125, "distill_loss": 0.080078125, "epoch": 1.006124234470691, "explain_loss": 1.4140625, "loss": 6.03125, "ref_ce_loss": 2.828125, "stage_align_loss": 0.0181884765625, "step": 2300 }, { "ce_loss": 0.734375, "distill_loss": 0.0634765625, "epoch": 1.006124234470691, "explain_loss": 1.1171875, "loss": 3.625, "ref_ce_loss": 1.6640625, "stage_align_loss": 0.031005859375, "step": 2300 }, { "epoch": 1.010498687664042, "grad_norm": 6.362780570983887, "learning_rate": 0.00015832899105374586, "loss": 3.240625, "step": 2310 }, { "ce_loss": 0.400390625, "distill_loss": 0.076171875, "epoch": 1.010498687664042, "explain_loss": 1.390625, "loss": 2.875, "ref_ce_loss": 0.984375, "stage_align_loss": 0.01239013671875, "step": 2310 }, { "ce_loss": 0.1923828125, "distill_loss": 0.08349609375, "epoch": 1.010498687664042, "explain_loss": 1.296875, "loss": 2.9375, "ref_ce_loss": 1.359375, "stage_align_loss": 0.0126953125, "step": 2310 }, { "ce_loss": 0.193359375, "distill_loss": 0.08349609375, "epoch": 1.010498687664042, "explain_loss": 0.921875, "loss": 2.484375, "ref_ce_loss": 1.265625, "stage_align_loss": 0.015380859375, "step": 2310 }, { "ce_loss": 0.10986328125, "distill_loss": 0.07763671875, "epoch": 1.010498687664042, "explain_loss": 1.234375, "loss": 3.4375, "ref_ce_loss": 2.015625, "stage_align_loss": 0.007659912109375, "step": 2310 }, { "epoch": 1.0148731408573928, "grad_norm": 7.720058441162109, "learning_rate": 0.00015793663326329853, "loss": 2.65546875, "step": 2320 }, { "ce_loss": 0.49609375, "distill_loss": 0.0654296875, "epoch": 1.0148731408573928, "explain_loss": 1.28125, "loss": 2.625, "ref_ce_loss": 0.77734375, "stage_align_loss": 0.00738525390625, "step": 2320 }, { "ce_loss": 0.494140625, "distill_loss": 0.0693359375, "epoch": 1.0148731408573928, "explain_loss": 1.15625, "loss": 2.5, "ref_ce_loss": 0.7734375, "stage_align_loss": 0.00982666015625, "step": 2320 }, { "ce_loss": 0.625, "distill_loss": 0.064453125, "epoch": 1.0148731408573928, "explain_loss": 1.34375, "loss": 2.859375, "ref_ce_loss": 0.78125, "stage_align_loss": 0.049072265625, "step": 2320 }, { "ce_loss": 0.2158203125, "distill_loss": 0.0732421875, "epoch": 1.0148731408573928, "explain_loss": 1.1328125, "loss": 2.15625, "ref_ce_loss": 0.72265625, "stage_align_loss": 0.0091552734375, "step": 2320 }, { "epoch": 1.0192475940507437, "grad_norm": 4.929980278015137, "learning_rate": 0.00015754292830001226, "loss": 2.194921875, "step": 2330 }, { "ce_loss": 0.169921875, "distill_loss": 0.08056640625, "epoch": 1.0192475940507437, "explain_loss": 0.921875, "loss": 1.671875, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.0150146484375, "step": 2330 }, { "ce_loss": 0.040771484375, "distill_loss": 0.0888671875, "epoch": 1.0192475940507437, "explain_loss": 0.921875, "loss": 1.6015625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.025146484375, "step": 2330 }, { "ce_loss": 0.2421875, "distill_loss": 0.091796875, "epoch": 1.0192475940507437, "explain_loss": 1.171875, "loss": 2.125, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.0135498046875, "step": 2330 }, { "ce_loss": 0.1767578125, "distill_loss": 0.0927734375, "epoch": 1.0192475940507437, "explain_loss": 1.0390625, "loss": 2.03125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.1728515625, "step": 2330 }, { "epoch": 1.0236220472440944, "grad_norm": 1.752362847328186, "learning_rate": 0.00015714788531852035, "loss": 2.78876953125, "step": 2340 }, { "ce_loss": 0.0947265625, "distill_loss": 0.126953125, "epoch": 1.0236220472440944, "explain_loss": 1.4140625, "loss": 2.1875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.00848388671875, "step": 2340 }, { "ce_loss": 0.00567626953125, "distill_loss": 0.99609375, "epoch": 1.0236220472440944, "explain_loss": 0.71484375, "loss": 2.171875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.0113525390625, "step": 2340 }, { "ce_loss": 0.07373046875, "distill_loss": 0.1171875, "epoch": 1.0236220472440944, "explain_loss": 1.2265625, "loss": 2.0, "ref_ce_loss": 0.578125, "stage_align_loss": 0.010009765625, "step": 2340 }, { "ce_loss": 0.0125732421875, "distill_loss": 0.10693359375, "epoch": 1.0236220472440944, "explain_loss": 1.3125, "loss": 1.9140625, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.01019287109375, "step": 2340 }, { "epoch": 1.0279965004374454, "grad_norm": 0.9185921549797058, "learning_rate": 0.00015675151350456838, "loss": 2.13115234375, "step": 2350 }, { "ce_loss": 0.265625, "distill_loss": 0.0654296875, "epoch": 1.0279965004374454, "explain_loss": 0.80078125, "loss": 1.609375, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.01470947265625, "step": 2350 }, { "ce_loss": 0.263671875, "distill_loss": 0.06494140625, "epoch": 1.0279965004374454, "explain_loss": 1.265625, "loss": 1.9375, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.01434326171875, "step": 2350 }, { "ce_loss": 0.05126953125, "distill_loss": 0.283203125, "epoch": 1.0279965004374454, "explain_loss": 0.9609375, "loss": 1.7734375, "ref_ce_loss": 0.453125, "stage_align_loss": 0.022705078125, "step": 2350 }, { "ce_loss": 0.2451171875, "distill_loss": 0.0693359375, "epoch": 1.0279965004374454, "explain_loss": 1.5390625, "loss": 2.5625, "ref_ce_loss": 0.68359375, "stage_align_loss": 0.014404296875, "step": 2350 }, { "epoch": 1.0323709536307961, "grad_norm": 0.8845518827438354, "learning_rate": 0.00015635382207480073, "loss": 1.95439453125, "step": 2360 }, { "ce_loss": 0.1611328125, "distill_loss": 0.06201171875, "epoch": 1.0323709536307961, "explain_loss": 1.328125, "loss": 1.84375, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.0101318359375, "step": 2360 }, { "ce_loss": 0.373046875, "distill_loss": 0.1435546875, "epoch": 1.0323709536307961, "explain_loss": 0.9765625, "loss": 2.25, "ref_ce_loss": 0.734375, "stage_align_loss": 0.02392578125, "step": 2360 }, { "ce_loss": 0.00994873046875, "distill_loss": 0.0615234375, "epoch": 1.0323709536307961, "explain_loss": 1.015625, "loss": 1.515625, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0146484375, "step": 2360 }, { "ce_loss": 0.326171875, "distill_loss": 0.18359375, "epoch": 1.0323709536307961, "explain_loss": 1.171875, "loss": 2.359375, "ref_ce_loss": 0.65625, "stage_align_loss": 0.0262451171875, "step": 2360 }, { "epoch": 1.036745406824147, "grad_norm": 1.5183933973312378, "learning_rate": 0.0001559548202765461, "loss": 1.9140625, "step": 2370 }, { "ce_loss": 0.1259765625, "distill_loss": 0.0673828125, "epoch": 1.036745406824147, "explain_loss": 1.328125, "loss": 2.171875, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.0286865234375, "step": 2370 }, { "ce_loss": 0.03271484375, "distill_loss": 0.064453125, "epoch": 1.036745406824147, "explain_loss": 1.5625, "loss": 2.15625, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.01025390625, "step": 2370 }, { "ce_loss": 0.07080078125, "distill_loss": 0.06201171875, "epoch": 1.036745406824147, "explain_loss": 1.15625, "loss": 1.6328125, "ref_ce_loss": 0.328125, "stage_align_loss": 0.0145263671875, "step": 2370 }, { "ce_loss": 0.1591796875, "distill_loss": 0.09814453125, "epoch": 1.036745406824147, "explain_loss": 1.1171875, "loss": 1.8125, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.010009765625, "step": 2370 }, { "epoch": 1.0411198600174978, "grad_norm": 0.9602667093276978, "learning_rate": 0.00015555451738760256, "loss": 2.00517578125, "step": 2380 }, { "ce_loss": 0.2392578125, "distill_loss": 0.061279296875, "epoch": 1.0411198600174978, "explain_loss": 1.4765625, "loss": 2.25, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.01043701171875, "step": 2380 }, { "ce_loss": 0.69921875, "distill_loss": 0.0625, "epoch": 1.0411198600174978, "explain_loss": 0.85546875, "loss": 2.171875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0228271484375, "step": 2380 }, { "ce_loss": 0.51171875, "distill_loss": 0.0888671875, "epoch": 1.0411198600174978, "explain_loss": 1.2265625, "loss": 2.390625, "ref_ce_loss": 0.53125, "stage_align_loss": 0.031982421875, "step": 2380 }, { "ce_loss": 0.1357421875, "distill_loss": 0.0869140625, "epoch": 1.0411198600174978, "explain_loss": 0.9765625, "loss": 1.609375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.01519775390625, "step": 2380 }, { "epoch": 1.0454943132108487, "grad_norm": 1.8282670974731445, "learning_rate": 0.00015515292271602187, "loss": 1.9564453125, "step": 2390 }, { "ce_loss": 0.140625, "distill_loss": 0.138671875, "epoch": 1.0454943132108487, "explain_loss": 1.171875, "loss": 2.0625, "ref_ce_loss": 0.59375, "stage_align_loss": 0.00982666015625, "step": 2390 }, { "ce_loss": 0.5, "distill_loss": 0.0537109375, "epoch": 1.0454943132108487, "explain_loss": 0.86328125, "loss": 1.96875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0089111328125, "step": 2390 }, { "ce_loss": 0.171875, "distill_loss": 0.0830078125, "epoch": 1.0454943132108487, "explain_loss": 1.328125, "loss": 1.921875, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.012939453125, "step": 2390 }, { "ce_loss": 0.609375, "distill_loss": 0.059814453125, "epoch": 1.0454943132108487, "explain_loss": 1.15625, "loss": 2.25, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.0115966796875, "step": 2390 }, { "epoch": 1.0498687664041995, "grad_norm": 1.4587677717208862, "learning_rate": 0.00015475004559989291, "loss": 1.933203125, "step": 2400 }, { "ce_loss": 0.0177001953125, "distill_loss": 0.0654296875, "epoch": 1.0498687664041995, "explain_loss": 0.890625, "loss": 1.453125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.01025390625, "step": 2400 }, { "ce_loss": 0.2060546875, "distill_loss": 0.072265625, "epoch": 1.0498687664041995, "explain_loss": 1.03125, "loss": 1.640625, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.009765625, "step": 2400 }, { "ce_loss": 0.01129150390625, "distill_loss": 0.0966796875, "epoch": 1.0498687664041995, "explain_loss": 1.140625, "loss": 1.703125, "ref_ce_loss": 0.4375, "stage_align_loss": 0.010009765625, "step": 2400 }, { "ce_loss": 0.408203125, "distill_loss": 0.0751953125, "epoch": 1.0498687664041995, "explain_loss": 1.015625, "loss": 1.8984375, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.015869140625, "step": 2400 }, { "epoch": 1.0542432195975504, "grad_norm": 1.1029112339019775, "learning_rate": 0.0001543458954071247, "loss": 1.955859375, "step": 2410 }, { "ce_loss": 0.064453125, "distill_loss": 0.0703125, "epoch": 1.0542432195975504, "explain_loss": 0.81640625, "loss": 1.390625, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.01055908203125, "step": 2410 }, { "ce_loss": 0.03125, "distill_loss": 0.07763671875, "epoch": 1.0542432195975504, "explain_loss": 1.0546875, "loss": 1.8125, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.01806640625, "step": 2410 }, { "ce_loss": 0.42578125, "distill_loss": 0.0751953125, "epoch": 1.0542432195975504, "explain_loss": 1.0390625, "loss": 1.828125, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.00958251953125, "step": 2410 }, { "ce_loss": 0.06396484375, "distill_loss": 0.08837890625, "epoch": 1.0542432195975504, "explain_loss": 1.3125, "loss": 1.890625, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0106201171875, "step": 2410 }, { "epoch": 1.0586176727909011, "grad_norm": 0.8968416452407837, "learning_rate": 0.00015394048153522855, "loss": 1.93125, "step": 2420 }, { "ce_loss": 0.018310546875, "distill_loss": 0.07275390625, "epoch": 1.0586176727909011, "explain_loss": 1.1640625, "loss": 1.6640625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.0098876953125, "step": 2420 }, { "ce_loss": 0.0517578125, "distill_loss": 0.0888671875, "epoch": 1.0586176727909011, "explain_loss": 1.1796875, "loss": 1.890625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01080322265625, "step": 2420 }, { "ce_loss": 0.01177978515625, "distill_loss": 0.0791015625, "epoch": 1.0586176727909011, "explain_loss": 0.99609375, "loss": 1.5625, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.00787353515625, "step": 2420 }, { "ce_loss": 0.01361083984375, "distill_loss": 0.078125, "epoch": 1.0586176727909011, "explain_loss": 1.078125, "loss": 1.46875, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.00732421875, "step": 2420 }, { "epoch": 1.0629921259842519, "grad_norm": 1.582795262336731, "learning_rate": 0.00015353381341109935, "loss": 1.85576171875, "step": 2430 }, { "ce_loss": 0.006317138671875, "distill_loss": 0.064453125, "epoch": 1.0629921259842519, "explain_loss": 0.8046875, "loss": 1.2421875, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00762939453125, "step": 2430 }, { "ce_loss": 0.30859375, "distill_loss": 0.07666015625, "epoch": 1.0629921259842519, "explain_loss": 1.359375, "loss": 2.375, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.007659912109375, "step": 2430 }, { "ce_loss": 0.004974365234375, "distill_loss": 0.09033203125, "epoch": 1.0629921259842519, "explain_loss": 1.3125, "loss": 1.84375, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0067138671875, "step": 2430 }, { "ce_loss": 0.1796875, "distill_loss": 0.0791015625, "epoch": 1.0629921259842519, "explain_loss": 1.3828125, "loss": 1.96875, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.0120849609375, "step": 2430 }, { "epoch": 1.0673665791776028, "grad_norm": 0.6114404201507568, "learning_rate": 0.00015312590049079662, "loss": 1.80302734375, "step": 2440 }, { "ce_loss": 1.7734375, "distill_loss": 0.08056640625, "epoch": 1.0673665791776028, "explain_loss": 1.171875, "loss": 3.53125, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.01953125, "step": 2440 }, { "ce_loss": 0.039306640625, "distill_loss": 0.05419921875, "epoch": 1.0673665791776028, "explain_loss": 1.0, "loss": 1.390625, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.010498046875, "step": 2440 }, { "ce_loss": 0.1416015625, "distill_loss": 0.07666015625, "epoch": 1.0673665791776028, "explain_loss": 1.46875, "loss": 2.125, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.006439208984375, "step": 2440 }, { "ce_loss": 0.71484375, "distill_loss": 0.060791015625, "epoch": 1.0673665791776028, "explain_loss": 1.265625, "loss": 2.421875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.005523681640625, "step": 2440 }, { "epoch": 1.0717410323709535, "grad_norm": 0.590105414390564, "learning_rate": 0.0001527167522593245, "loss": 1.823046875, "step": 2450 }, { "ce_loss": 0.035400390625, "distill_loss": 0.06005859375, "epoch": 1.0717410323709535, "explain_loss": 0.91796875, "loss": 1.515625, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0152587890625, "step": 2450 }, { "ce_loss": 0.58984375, "distill_loss": 0.0595703125, "epoch": 1.0717410323709535, "explain_loss": 0.86328125, "loss": 2.03125, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0155029296875, "step": 2450 }, { "ce_loss": 0.263671875, "distill_loss": 0.0693359375, "epoch": 1.0717410323709535, "explain_loss": 1.5625, "loss": 2.265625, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0126953125, "step": 2450 }, { "ce_loss": 0.462890625, "distill_loss": 0.080078125, "epoch": 1.0717410323709535, "explain_loss": 1.1875, "loss": 2.4375, "ref_ce_loss": 0.69140625, "stage_align_loss": 0.018310546875, "step": 2450 }, { "epoch": 1.0761154855643045, "grad_norm": 1.2475793361663818, "learning_rate": 0.00015230637823041124, "loss": 1.79306640625, "step": 2460 }, { "ce_loss": 0.00689697265625, "distill_loss": 0.055908203125, "epoch": 1.0761154855643045, "explain_loss": 1.109375, "loss": 1.5, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.006744384765625, "step": 2460 }, { "ce_loss": 1.640625, "distill_loss": 0.06884765625, "epoch": 1.0761154855643045, "explain_loss": 1.03125, "loss": 3.25, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.0439453125, "step": 2460 }, { "ce_loss": 0.019775390625, "distill_loss": 0.07373046875, "epoch": 1.0761154855643045, "explain_loss": 1.0625, "loss": 1.4375, "ref_ce_loss": 0.2734375, "stage_align_loss": 0.00408935546875, "step": 2460 }, { "ce_loss": 0.07080078125, "distill_loss": 0.050048828125, "epoch": 1.0761154855643045, "explain_loss": 1.078125, "loss": 1.6875, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.005035400390625, "step": 2460 }, { "epoch": 1.0804899387576552, "grad_norm": 0.531653881072998, "learning_rate": 0.00015189478794628802, "loss": 1.8345703125, "step": 2470 }, { "ce_loss": 0.1982421875, "distill_loss": 0.05810546875, "epoch": 1.0804899387576552, "explain_loss": 0.90625, "loss": 1.515625, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.00885009765625, "step": 2470 }, { "ce_loss": 0.003875732421875, "distill_loss": 0.0703125, "epoch": 1.0804899387576552, "explain_loss": 1.46875, "loss": 2.03125, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.00433349609375, "step": 2470 }, { "ce_loss": 0.400390625, "distill_loss": 0.06494140625, "epoch": 1.0804899387576552, "explain_loss": 1.1171875, "loss": 2.046875, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.0184326171875, "step": 2470 }, { "ce_loss": 0.0225830078125, "distill_loss": 0.0556640625, "epoch": 1.0804899387576552, "explain_loss": 0.7109375, "loss": 1.15625, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.004150390625, "step": 2470 }, { "epoch": 1.0848643919510061, "grad_norm": 1.6473695039749146, "learning_rate": 0.00015148199097746694, "loss": 2.013671875, "step": 2480 }, { "ce_loss": 0.52734375, "distill_loss": 0.12158203125, "epoch": 1.0848643919510061, "explain_loss": 1.0703125, "loss": 2.0625, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.031494140625, "step": 2480 }, { "ce_loss": 0.045166015625, "distill_loss": 0.07373046875, "epoch": 1.0848643919510061, "explain_loss": 0.984375, "loss": 1.484375, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.0057373046875, "step": 2480 }, { "ce_loss": 0.130859375, "distill_loss": 0.12255859375, "epoch": 1.0848643919510061, "explain_loss": 1.203125, "loss": 1.875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.01483154296875, "step": 2480 }, { "ce_loss": 0.048583984375, "distill_loss": 0.062255859375, "epoch": 1.0848643919510061, "explain_loss": 1.4375, "loss": 1.984375, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.006103515625, "step": 2480 }, { "epoch": 1.0892388451443569, "grad_norm": 1.1668928861618042, "learning_rate": 0.00015106799692251853, "loss": 1.87431640625, "step": 2490 }, { "ce_loss": 0.06689453125, "distill_loss": 0.07080078125, "epoch": 1.0892388451443569, "explain_loss": 1.375, "loss": 2.125, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.009033203125, "step": 2490 }, { "ce_loss": 0.080078125, "distill_loss": 0.07763671875, "epoch": 1.0892388451443569, "explain_loss": 0.96875, "loss": 1.625, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.0087890625, "step": 2490 }, { "ce_loss": 0.004730224609375, "distill_loss": 0.1103515625, "epoch": 1.0892388451443569, "explain_loss": 1.328125, "loss": 1.859375, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.00390625, "step": 2490 }, { "ce_loss": 0.10009765625, "distill_loss": 0.06396484375, "epoch": 1.0892388451443569, "explain_loss": 1.765625, "loss": 2.3125, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00738525390625, "step": 2490 }, { "epoch": 1.0936132983377078, "grad_norm": 1.13765287399292, "learning_rate": 0.00015065281540784878, "loss": 1.8578125, "step": 2500 }, { "ce_loss": 0.1884765625, "distill_loss": 0.0751953125, "epoch": 1.0936132983377078, "explain_loss": 0.83203125, "loss": 1.625, "ref_ce_loss": 0.5, "stage_align_loss": 0.023681640625, "step": 2500 }, { "ce_loss": 0.083984375, "distill_loss": 0.064453125, "epoch": 1.0936132983377078, "explain_loss": 1.34375, "loss": 1.734375, "ref_ce_loss": 0.2373046875, "stage_align_loss": 0.0084228515625, "step": 2500 }, { "ce_loss": 0.1416015625, "distill_loss": 0.0751953125, "epoch": 1.0936132983377078, "explain_loss": 0.9765625, "loss": 1.7265625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.00921630859375, "step": 2500 }, { "ce_loss": 0.001556396484375, "distill_loss": 0.197265625, "epoch": 1.0936132983377078, "explain_loss": 1.1328125, "loss": 1.796875, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.004425048828125, "step": 2500 }, { "epoch": 1.0979877515310585, "grad_norm": 0.519006073474884, "learning_rate": 0.00015023645608747493, "loss": 1.805078125, "step": 2510 }, { "ce_loss": 0.06787109375, "distill_loss": 0.0634765625, "epoch": 1.0979877515310585, "explain_loss": 1.140625, "loss": 1.7265625, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.006591796875, "step": 2510 }, { "ce_loss": 0.003265380859375, "distill_loss": 0.0712890625, "epoch": 1.0979877515310585, "explain_loss": 1.1953125, "loss": 1.890625, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.00555419921875, "step": 2510 }, { "ce_loss": 0.00323486328125, "distill_loss": 0.08837890625, "epoch": 1.0979877515310585, "explain_loss": 1.2265625, "loss": 1.7421875, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.0050048828125, "step": 2510 }, { "ce_loss": 0.0040283203125, "distill_loss": 0.0771484375, "epoch": 1.0979877515310585, "explain_loss": 0.8984375, "loss": 1.40625, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.006561279296875, "step": 2510 }, { "epoch": 1.1023622047244095, "grad_norm": 0.8585734367370605, "learning_rate": 0.0001498189286428012, "loss": 1.823046875, "step": 2520 }, { "ce_loss": 0.09716796875, "distill_loss": 0.064453125, "epoch": 1.1023622047244095, "explain_loss": 1.265625, "loss": 1.78125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.00543212890625, "step": 2520 }, { "ce_loss": 0.004638671875, "distill_loss": 0.060791015625, "epoch": 1.1023622047244095, "explain_loss": 0.875, "loss": 1.328125, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00531005859375, "step": 2520 }, { "ce_loss": 0.09033203125, "distill_loss": 0.07275390625, "epoch": 1.1023622047244095, "explain_loss": 1.015625, "loss": 1.59375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00555419921875, "step": 2520 }, { "ce_loss": 0.00921630859375, "distill_loss": 0.16015625, "epoch": 1.1023622047244095, "explain_loss": 1.109375, "loss": 1.5703125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.00567626953125, "step": 2520 }, { "epoch": 1.1067366579177602, "grad_norm": 0.5464480519294739, "learning_rate": 0.00014940024278239372, "loss": 1.82431640625, "step": 2530 }, { "ce_loss": 0.06396484375, "distill_loss": 0.080078125, "epoch": 1.1067366579177602, "explain_loss": 1.0859375, "loss": 1.59375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.005218505859375, "step": 2530 }, { "ce_loss": 0.019775390625, "distill_loss": 0.07373046875, "epoch": 1.1067366579177602, "explain_loss": 0.92578125, "loss": 1.6328125, "ref_ce_loss": 0.609375, "stage_align_loss": 0.005523681640625, "step": 2530 }, { "ce_loss": 0.5390625, "distill_loss": 0.0556640625, "epoch": 1.1067366579177602, "explain_loss": 0.96484375, "loss": 2.015625, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.033203125, "step": 2530 }, { "ce_loss": 0.00927734375, "distill_loss": 0.0625, "epoch": 1.1067366579177602, "explain_loss": 1.375, "loss": 1.90625, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.006103515625, "step": 2530 }, { "epoch": 1.1111111111111112, "grad_norm": 0.5857027769088745, "learning_rate": 0.00014898040824175464, "loss": 1.77109375, "step": 2540 }, { "ce_loss": 0.030029296875, "distill_loss": 0.0673828125, "epoch": 1.1111111111111112, "explain_loss": 1.2421875, "loss": 1.8125, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.00433349609375, "step": 2540 }, { "ce_loss": 0.23046875, "distill_loss": 0.060546875, "epoch": 1.1111111111111112, "explain_loss": 0.9609375, "loss": 1.7109375, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.00927734375, "step": 2540 }, { "ce_loss": 0.126953125, "distill_loss": 0.0673828125, "epoch": 1.1111111111111112, "explain_loss": 0.82421875, "loss": 1.40625, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00567626953125, "step": 2540 }, { "ce_loss": 0.005859375, "distill_loss": 0.08251953125, "epoch": 1.1111111111111112, "explain_loss": 1.2734375, "loss": 1.828125, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.003936767578125, "step": 2540 }, { "epoch": 1.1154855643044619, "grad_norm": 1.3760970830917358, "learning_rate": 0.00014855943478309585, "loss": 1.82998046875, "step": 2550 }, { "ce_loss": 0.0020904541015625, "distill_loss": 0.091796875, "epoch": 1.1154855643044619, "explain_loss": 1.09375, "loss": 1.5703125, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.004119873046875, "step": 2550 }, { "ce_loss": 0.07373046875, "distill_loss": 0.07763671875, "epoch": 1.1154855643044619, "explain_loss": 0.921875, "loss": 1.4375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.00537109375, "step": 2550 }, { "ce_loss": 0.01446533203125, "distill_loss": 0.07861328125, "epoch": 1.1154855643044619, "explain_loss": 0.9921875, "loss": 1.34375, "ref_ce_loss": 0.25, "stage_align_loss": 0.00518798828125, "step": 2550 }, { "ce_loss": 0.0015106201171875, "distill_loss": 0.0751953125, "epoch": 1.1154855643044619, "explain_loss": 1.1640625, "loss": 1.7265625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.003326416015625, "step": 2550 }, { "epoch": 1.1198600174978128, "grad_norm": 1.578573226928711, "learning_rate": 0.00014813733219511198, "loss": 1.8498046875, "step": 2560 }, { "ce_loss": 0.02685546875, "distill_loss": 0.0703125, "epoch": 1.1198600174978128, "explain_loss": 1.0390625, "loss": 1.4609375, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.0047607421875, "step": 2560 }, { "ce_loss": 0.059326171875, "distill_loss": 0.0751953125, "epoch": 1.1198600174978128, "explain_loss": 1.4453125, "loss": 1.8515625, "ref_ce_loss": 0.267578125, "stage_align_loss": 0.00567626953125, "step": 2560 }, { "ce_loss": 0.0242919921875, "distill_loss": 0.083984375, "epoch": 1.1198600174978128, "explain_loss": 1.1875, "loss": 1.59375, "ref_ce_loss": 0.296875, "stage_align_loss": 0.00421142578125, "step": 2560 }, { "ce_loss": 0.0240478515625, "distill_loss": 0.0927734375, "epoch": 1.1198600174978128, "explain_loss": 1.078125, "loss": 1.640625, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.004425048828125, "step": 2560 }, { "epoch": 1.1242344706911636, "grad_norm": 0.4986472725868225, "learning_rate": 0.00014771411029275276, "loss": 1.76357421875, "step": 2570 }, { "ce_loss": 0.0185546875, "distill_loss": 0.0673828125, "epoch": 1.1242344706911636, "explain_loss": 1.140625, "loss": 1.625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.00372314453125, "step": 2570 }, { "ce_loss": 0.005828857421875, "distill_loss": 0.06298828125, "epoch": 1.1242344706911636, "explain_loss": 0.78125, "loss": 1.234375, "ref_ce_loss": 0.375, "stage_align_loss": 0.006195068359375, "step": 2570 }, { "ce_loss": 0.0074462890625, "distill_loss": 0.06640625, "epoch": 1.1242344706911636, "explain_loss": 1.3828125, "loss": 1.890625, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.0040283203125, "step": 2570 }, { "ce_loss": 0.314453125, "distill_loss": 0.0693359375, "epoch": 1.1242344706911636, "explain_loss": 1.3359375, "loss": 2.203125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.03173828125, "step": 2570 }, { "epoch": 1.1286089238845145, "grad_norm": 0.6862332820892334, "learning_rate": 0.00014728977891699472, "loss": 1.81337890625, "step": 2580 }, { "ce_loss": 0.0294189453125, "distill_loss": 0.052734375, "epoch": 1.1286089238845145, "explain_loss": 1.328125, "loss": 1.7265625, "ref_ce_loss": 0.310546875, "stage_align_loss": 0.005523681640625, "step": 2580 }, { "ce_loss": 0.00921630859375, "distill_loss": 0.06494140625, "epoch": 1.1286089238845145, "explain_loss": 1.0, "loss": 1.484375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.003875732421875, "step": 2580 }, { "ce_loss": 0.0654296875, "distill_loss": 0.083984375, "epoch": 1.1286089238845145, "explain_loss": 1.4375, "loss": 2.09375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.00506591796875, "step": 2580 }, { "ce_loss": 0.494140625, "distill_loss": 0.07080078125, "epoch": 1.1286089238845145, "explain_loss": 0.9921875, "loss": 2.015625, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.03173828125, "step": 2580 }, { "epoch": 1.1329833770778652, "grad_norm": 1.1194524765014648, "learning_rate": 0.0001468643479346125, "loss": 1.82646484375, "step": 2590 }, { "ce_loss": 0.11474609375, "distill_loss": 0.0751953125, "epoch": 1.1329833770778652, "explain_loss": 1.0546875, "loss": 1.6171875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0086669921875, "step": 2590 }, { "ce_loss": 0.024169921875, "distill_loss": 0.10302734375, "epoch": 1.1329833770778652, "explain_loss": 0.98046875, "loss": 1.5703125, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.0052490234375, "step": 2590 }, { "ce_loss": 0.40625, "distill_loss": 0.05810546875, "epoch": 1.1329833770778652, "explain_loss": 0.87890625, "loss": 1.8125, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.034912109375, "step": 2590 }, { "ce_loss": 0.1064453125, "distill_loss": 0.06640625, "epoch": 1.1329833770778652, "explain_loss": 1.359375, "loss": 1.875, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.0184326171875, "step": 2590 }, { "epoch": 1.1373578302712162, "grad_norm": 0.5321169495582581, "learning_rate": 0.00014643782723794947, "loss": 1.816015625, "step": 2600 }, { "ce_loss": 0.0400390625, "distill_loss": 0.06640625, "epoch": 1.1373578302712162, "explain_loss": 0.6953125, "loss": 1.1953125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.004058837890625, "step": 2600 }, { "ce_loss": 0.189453125, "distill_loss": 0.08251953125, "epoch": 1.1373578302712162, "explain_loss": 1.109375, "loss": 1.8046875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.007110595703125, "step": 2600 }, { "ce_loss": 0.0240478515625, "distill_loss": 0.06005859375, "epoch": 1.1373578302712162, "explain_loss": 0.9375, "loss": 1.34375, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.00506591796875, "step": 2600 }, { "ce_loss": 0.0147705078125, "distill_loss": 0.0634765625, "epoch": 1.1373578302712162, "explain_loss": 0.66015625, "loss": 1.03125, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.00445556640625, "step": 2600 }, { "epoch": 1.141732283464567, "grad_norm": 0.7438850402832031, "learning_rate": 0.00014601022674468742, "loss": 1.8224609375, "step": 2610 }, { "ce_loss": 0.236328125, "distill_loss": 0.078125, "epoch": 1.141732283464567, "explain_loss": 1.1953125, "loss": 1.96875, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0172119140625, "step": 2610 }, { "ce_loss": 0.515625, "distill_loss": 0.0703125, "epoch": 1.141732283464567, "explain_loss": 0.9921875, "loss": 2.0, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0147705078125, "step": 2610 }, { "ce_loss": 0.059814453125, "distill_loss": 0.0673828125, "epoch": 1.141732283464567, "explain_loss": 0.953125, "loss": 1.53125, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0128173828125, "step": 2610 }, { "ce_loss": 0.1806640625, "distill_loss": 0.0703125, "epoch": 1.141732283464567, "explain_loss": 1.3046875, "loss": 2.1875, "ref_ce_loss": 0.609375, "stage_align_loss": 0.0145263671875, "step": 2610 }, { "epoch": 1.1461067366579178, "grad_norm": 0.6410822868347168, "learning_rate": 0.00014558155639761623, "loss": 1.84384765625, "step": 2620 }, { "ce_loss": 0.2392578125, "distill_loss": 0.060791015625, "epoch": 1.1461067366579178, "explain_loss": 1.1640625, "loss": 1.78125, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.005950927734375, "step": 2620 }, { "ce_loss": 0.96484375, "distill_loss": 0.11328125, "epoch": 1.1461067366579178, "explain_loss": 1.03125, "loss": 2.75, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.041015625, "step": 2620 }, { "ce_loss": 0.146484375, "distill_loss": 0.0595703125, "epoch": 1.1461067366579178, "explain_loss": 1.2421875, "loss": 1.8125, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.00750732421875, "step": 2620 }, { "ce_loss": 0.703125, "distill_loss": 0.05517578125, "epoch": 1.1461067366579178, "explain_loss": 0.65234375, "loss": 1.765625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.01324462890625, "step": 2620 }, { "epoch": 1.1504811898512686, "grad_norm": 0.7677959203720093, "learning_rate": 0.0001451518261644026, "loss": 1.8087890625, "step": 2630 }, { "ce_loss": 0.0230712890625, "distill_loss": 0.0654296875, "epoch": 1.1504811898512686, "explain_loss": 0.953125, "loss": 1.5, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0038604736328125, "step": 2630 }, { "ce_loss": 0.177734375, "distill_loss": 0.083984375, "epoch": 1.1504811898512686, "explain_loss": 0.96875, "loss": 1.859375, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.01806640625, "step": 2630 }, { "ce_loss": 0.034423828125, "distill_loss": 0.06298828125, "epoch": 1.1504811898512686, "explain_loss": 1.1015625, "loss": 1.703125, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.00567626953125, "step": 2630 }, { "ce_loss": 0.005157470703125, "distill_loss": 0.09033203125, "epoch": 1.1504811898512686, "explain_loss": 1.28125, "loss": 1.796875, "ref_ce_loss": 0.421875, "stage_align_loss": 0.004302978515625, "step": 2630 }, { "epoch": 1.1548556430446195, "grad_norm": 1.2415165901184082, "learning_rate": 0.00014472104603735817, "loss": 1.8966796875, "step": 2640 }, { "ce_loss": 0.60546875, "distill_loss": 0.0537109375, "epoch": 1.1548556430446195, "explain_loss": 0.9453125, "loss": 2.1875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0306396484375, "step": 2640 }, { "ce_loss": 0.2275390625, "distill_loss": 0.08544921875, "epoch": 1.1548556430446195, "explain_loss": 1.140625, "loss": 1.796875, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.047119140625, "step": 2640 }, { "ce_loss": 0.003387451171875, "distill_loss": 0.061279296875, "epoch": 1.1548556430446195, "explain_loss": 0.9609375, "loss": 1.328125, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.005950927734375, "step": 2640 }, { "ce_loss": 0.048828125, "distill_loss": 0.06201171875, "epoch": 1.1548556430446195, "explain_loss": 0.8046875, "loss": 1.3046875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.005401611328125, "step": 2640 }, { "epoch": 1.1592300962379702, "grad_norm": 1.0761642456054688, "learning_rate": 0.00014428922603320742, "loss": 1.75, "step": 2650 }, { "ce_loss": 0.022705078125, "distill_loss": 0.0849609375, "epoch": 1.1592300962379702, "explain_loss": 1.0390625, "loss": 1.65625, "ref_ce_loss": 0.5, "stage_align_loss": 0.003448486328125, "step": 2650 }, { "ce_loss": 0.0027923583984375, "distill_loss": 0.103515625, "epoch": 1.1592300962379702, "explain_loss": 1.390625, "loss": 1.9765625, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0030364990234375, "step": 2650 }, { "ce_loss": 0.0556640625, "distill_loss": 0.08837890625, "epoch": 1.1592300962379702, "explain_loss": 0.85546875, "loss": 1.34375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.007354736328125, "step": 2650 }, { "ce_loss": 0.11669921875, "distill_loss": 0.095703125, "epoch": 1.1592300962379702, "explain_loss": 0.95703125, "loss": 1.625, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.00799560546875, "step": 2650 }, { "epoch": 1.1636045494313212, "grad_norm": 14.011808395385742, "learning_rate": 0.0001438563761928544, "loss": 1.92470703125, "step": 2660 }, { "ce_loss": 0.01043701171875, "distill_loss": 0.25390625, "epoch": 1.1636045494313212, "explain_loss": 0.98828125, "loss": 1.59375, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.00616455078125, "step": 2660 }, { "ce_loss": 0.02001953125, "distill_loss": 0.15625, "epoch": 1.1636045494313212, "explain_loss": 1.4375, "loss": 1.90625, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.005157470703125, "step": 2660 }, { "ce_loss": 0.0029449462890625, "distill_loss": 0.16015625, "epoch": 1.1636045494313212, "explain_loss": 0.5625, "loss": 1.140625, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00408935546875, "step": 2660 }, { "ce_loss": 0.044189453125, "distill_loss": 0.11669921875, "epoch": 1.1636045494313212, "explain_loss": 1.34375, "loss": 1.9296875, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0078125, "step": 2660 }, { "epoch": 1.167979002624672, "grad_norm": 5.616036415100098, "learning_rate": 0.0001434225065811496, "loss": 2.8650390625, "step": 2670 }, { "ce_loss": 1.5390625, "distill_loss": 0.19140625, "epoch": 1.167979002624672, "explain_loss": 0.9375, "loss": 2.96875, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.01092529296875, "step": 2670 }, { "ce_loss": 0.006866455078125, "distill_loss": 0.134765625, "epoch": 1.167979002624672, "explain_loss": 1.0546875, "loss": 1.6015625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.005035400390625, "step": 2670 }, { "ce_loss": 0.003631591796875, "distill_loss": 0.18359375, "epoch": 1.167979002624672, "explain_loss": 1.3359375, "loss": 2.09375, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.0033416748046875, "step": 2670 }, { "ce_loss": 1.3359375, "distill_loss": 0.142578125, "epoch": 1.167979002624672, "explain_loss": 0.90625, "loss": 2.84375, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0057373046875, "step": 2670 }, { "epoch": 1.1723534558180226, "grad_norm": 2.613694906234741, "learning_rate": 0.00014298762728665566, "loss": 2.5599609375, "step": 2680 }, { "ce_loss": 0.76171875, "distill_loss": 0.0751953125, "epoch": 1.1723534558180226, "explain_loss": 0.62890625, "loss": 1.875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.01153564453125, "step": 2680 }, { "ce_loss": 0.0419921875, "distill_loss": 0.173828125, "epoch": 1.1723534558180226, "explain_loss": 1.1328125, "loss": 1.75, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.004669189453125, "step": 2680 }, { "ce_loss": 0.81640625, "distill_loss": 0.08251953125, "epoch": 1.1723534558180226, "explain_loss": 1.1328125, "loss": 2.5625, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0184326171875, "step": 2680 }, { "ce_loss": 0.59765625, "distill_loss": 0.1435546875, "epoch": 1.1723534558180226, "explain_loss": 1.28125, "loss": 2.578125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0157470703125, "step": 2680 }, { "epoch": 1.1767279090113736, "grad_norm": 1.6731935739517212, "learning_rate": 0.00014255174842141296, "loss": 2.1763671875, "step": 2690 }, { "ce_loss": 0.291015625, "distill_loss": 0.14453125, "epoch": 1.1767279090113736, "explain_loss": 1.4375, "loss": 2.34375, "ref_ce_loss": 0.46875, "stage_align_loss": 0.007080078125, "step": 2690 }, { "ce_loss": 0.021240234375, "distill_loss": 0.08203125, "epoch": 1.1767279090113736, "explain_loss": 1.03125, "loss": 1.640625, "ref_ce_loss": 0.5, "stage_align_loss": 0.004486083984375, "step": 2690 }, { "ce_loss": 0.01275634765625, "distill_loss": 0.2060546875, "epoch": 1.1767279090113736, "explain_loss": 1.25, "loss": 1.90625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.004364013671875, "step": 2690 }, { "ce_loss": 0.00122833251953125, "distill_loss": 0.142578125, "epoch": 1.1767279090113736, "explain_loss": 1.0390625, "loss": 1.5625, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.00347900390625, "step": 2690 }, { "epoch": 1.1811023622047245, "grad_norm": 1.5384821891784668, "learning_rate": 0.0001421148801207044, "loss": 1.9537109375, "step": 2700 }, { "ce_loss": 0.00168609619140625, "distill_loss": 0.111328125, "epoch": 1.1811023622047245, "explain_loss": 0.6875, "loss": 1.0703125, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.0057373046875, "step": 2700 }, { "ce_loss": 0.005889892578125, "distill_loss": 0.20703125, "epoch": 1.1811023622047245, "explain_loss": 1.375, "loss": 1.984375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0031890869140625, "step": 2700 }, { "ce_loss": 0.002227783203125, "distill_loss": 0.1865234375, "epoch": 1.1811023622047245, "explain_loss": 1.078125, "loss": 1.703125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.00469970703125, "step": 2700 }, { "ce_loss": 0.072265625, "distill_loss": 0.06640625, "epoch": 1.1811023622047245, "explain_loss": 0.96875, "loss": 1.5, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0078125, "step": 2700 }, { "epoch": 1.1854768153980753, "grad_norm": 1.1831530332565308, "learning_rate": 0.00014167703254281975, "loss": 1.9673828125, "step": 2710 }, { "ce_loss": 0.1279296875, "distill_loss": 0.16796875, "epoch": 1.1854768153980753, "explain_loss": 1.328125, "loss": 2.09375, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.0146484375, "step": 2710 }, { "ce_loss": 0.50390625, "distill_loss": 0.087890625, "epoch": 1.1854768153980753, "explain_loss": 1.09375, "loss": 2.15625, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.041015625, "step": 2710 }, { "ce_loss": 0.10986328125, "distill_loss": 0.11279296875, "epoch": 1.1854768153980753, "explain_loss": 1.296875, "loss": 2.109375, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.00750732421875, "step": 2710 }, { "ce_loss": 0.216796875, "distill_loss": 0.07373046875, "epoch": 1.1854768153980753, "explain_loss": 1.0703125, "loss": 1.75, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.00872802734375, "step": 2710 }, { "epoch": 1.189851268591426, "grad_norm": 1.6848722696304321, "learning_rate": 0.00014123821586881945, "loss": 1.897265625, "step": 2720 }, { "ce_loss": 0.486328125, "distill_loss": 0.06005859375, "epoch": 1.189851268591426, "explain_loss": 0.9921875, "loss": 1.984375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.03857421875, "step": 2720 }, { "ce_loss": 0.0184326171875, "distill_loss": 0.07275390625, "epoch": 1.189851268591426, "explain_loss": 0.7109375, "loss": 1.21875, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.005828857421875, "step": 2720 }, { "ce_loss": 0.068359375, "distill_loss": 0.109375, "epoch": 1.189851268591426, "explain_loss": 1.484375, "loss": 2.1875, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.005035400390625, "step": 2720 }, { "ce_loss": 0.455078125, "distill_loss": 0.130859375, "epoch": 1.189851268591426, "explain_loss": 1.25, "loss": 2.4375, "ref_ce_loss": 0.578125, "stage_align_loss": 0.031494140625, "step": 2720 }, { "epoch": 1.194225721784777, "grad_norm": 0.7515852451324463, "learning_rate": 0.00014079844030229785, "loss": 1.98466796875, "step": 2730 }, { "ce_loss": 1.390625, "distill_loss": 0.19921875, "epoch": 1.194225721784777, "explain_loss": 0.97265625, "loss": 3.125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.032470703125, "step": 2730 }, { "ce_loss": 0.89453125, "distill_loss": 0.058837890625, "epoch": 1.194225721784777, "explain_loss": 1.375, "loss": 3.0, "ref_ce_loss": 0.62890625, "stage_align_loss": 0.037841796875, "step": 2730 }, { "ce_loss": 1.609375, "distill_loss": 0.1474609375, "epoch": 1.194225721784777, "explain_loss": 1.171875, "loss": 3.59375, "ref_ce_loss": 0.578125, "stage_align_loss": 0.07275390625, "step": 2730 }, { "ce_loss": 0.0927734375, "distill_loss": 0.058349609375, "epoch": 1.194225721784777, "explain_loss": 0.859375, "loss": 1.4140625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.00482177734375, "step": 2730 }, { "epoch": 1.1986001749781277, "grad_norm": 1.16350257396698, "learning_rate": 0.00014035771606914595, "loss": 1.92490234375, "step": 2740 }, { "ce_loss": 0.095703125, "distill_loss": 0.06298828125, "epoch": 1.1986001749781277, "explain_loss": 1.078125, "loss": 1.65625, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.009765625, "step": 2740 }, { "ce_loss": 0.07763671875, "distill_loss": 0.0400390625, "epoch": 1.1986001749781277, "explain_loss": 1.203125, "loss": 1.890625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.01068115234375, "step": 2740 }, { "ce_loss": 0.5703125, "distill_loss": 0.0693359375, "epoch": 1.1986001749781277, "explain_loss": 1.046875, "loss": 2.25, "ref_ce_loss": 0.546875, "stage_align_loss": 0.0167236328125, "step": 2740 }, { "ce_loss": 0.049560546875, "distill_loss": 0.1044921875, "epoch": 1.1986001749781277, "explain_loss": 1.015625, "loss": 1.734375, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.00836181640625, "step": 2740 }, { "epoch": 1.2029746281714786, "grad_norm": 0.6703765392303467, "learning_rate": 0.00013991605341731375, "loss": 1.9453125, "step": 2750 }, { "ce_loss": 0.006378173828125, "distill_loss": 0.08984375, "epoch": 1.2029746281714786, "explain_loss": 1.3671875, "loss": 1.8125, "ref_ce_loss": 0.34375, "stage_align_loss": 0.00457763671875, "step": 2750 }, { "ce_loss": 0.171875, "distill_loss": 0.296875, "epoch": 1.2029746281714786, "explain_loss": 1.046875, "loss": 1.9921875, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.03857421875, "step": 2750 }, { "ce_loss": 0.5625, "distill_loss": 0.035888671875, "epoch": 1.2029746281714786, "explain_loss": 1.3359375, "loss": 2.515625, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.00604248046875, "step": 2750 }, { "ce_loss": 0.142578125, "distill_loss": 0.0888671875, "epoch": 1.2029746281714786, "explain_loss": 1.171875, "loss": 1.84375, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.005340576171875, "step": 2750 }, { "epoch": 1.2073490813648293, "grad_norm": 1.299296259880066, "learning_rate": 0.0001394734626165717, "loss": 1.97216796875, "step": 2760 }, { "ce_loss": 0.19140625, "distill_loss": 0.05419921875, "epoch": 1.2073490813648293, "explain_loss": 0.99609375, "loss": 1.5703125, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.0057373046875, "step": 2760 }, { "ce_loss": 0.02783203125, "distill_loss": 0.0830078125, "epoch": 1.2073490813648293, "explain_loss": 1.03125, "loss": 1.5859375, "ref_ce_loss": 0.4375, "stage_align_loss": 0.0087890625, "step": 2760 }, { "ce_loss": 0.003509521484375, "distill_loss": 0.1279296875, "epoch": 1.2073490813648293, "explain_loss": 0.9921875, "loss": 1.3984375, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.0040283203125, "step": 2760 }, { "ce_loss": 1.1015625, "distill_loss": 0.1376953125, "epoch": 1.2073490813648293, "explain_loss": 0.97265625, "loss": 2.703125, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.076171875, "step": 2760 }, { "epoch": 1.2117235345581803, "grad_norm": 1.9784820079803467, "learning_rate": 0.00013902995395827217, "loss": 1.9546875, "step": 2770 }, { "ce_loss": 0.38671875, "distill_loss": 0.119140625, "epoch": 1.2117235345581803, "explain_loss": 1.0546875, "loss": 2.125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.03955078125, "step": 2770 }, { "ce_loss": 0.234375, "distill_loss": 0.1328125, "epoch": 1.2117235345581803, "explain_loss": 0.9921875, "loss": 1.765625, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0185546875, "step": 2770 }, { "ce_loss": 0.08154296875, "distill_loss": 0.09375, "epoch": 1.2117235345581803, "explain_loss": 1.21875, "loss": 1.75, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.00408935546875, "step": 2770 }, { "ce_loss": 0.060546875, "distill_loss": 0.076171875, "epoch": 1.2117235345581803, "explain_loss": 0.80078125, "loss": 1.2421875, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.00494384765625, "step": 2770 }, { "epoch": 1.216097987751531, "grad_norm": 1.5358965396881104, "learning_rate": 0.00013858553775510993, "loss": 1.80224609375, "step": 2780 }, { "ce_loss": 0.03759765625, "distill_loss": 0.1357421875, "epoch": 1.216097987751531, "explain_loss": 0.6171875, "loss": 1.09375, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.004119873046875, "step": 2780 }, { "ce_loss": 0.06494140625, "distill_loss": 0.1796875, "epoch": 1.216097987751531, "explain_loss": 1.3359375, "loss": 1.9453125, "ref_ce_loss": 0.359375, "stage_align_loss": 0.007080078125, "step": 2780 }, { "ce_loss": 0.0341796875, "distill_loss": 0.1875, "epoch": 1.216097987751531, "explain_loss": 1.4765625, "loss": 2.1875, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0040283203125, "step": 2780 }, { "ce_loss": 0.2138671875, "distill_loss": 0.109375, "epoch": 1.216097987751531, "explain_loss": 0.9921875, "loss": 1.703125, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.0439453125, "step": 2780 }, { "epoch": 1.220472440944882, "grad_norm": 1.3133505582809448, "learning_rate": 0.0001381402243408826, "loss": 1.8619140625, "step": 2790 }, { "ce_loss": 0.038818359375, "distill_loss": 0.0712890625, "epoch": 1.220472440944882, "explain_loss": 1.265625, "loss": 1.8515625, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.040771484375, "step": 2790 }, { "ce_loss": 0.259765625, "distill_loss": 0.09521484375, "epoch": 1.220472440944882, "explain_loss": 0.88671875, "loss": 1.7421875, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.0181884765625, "step": 2790 }, { "ce_loss": 0.6171875, "distill_loss": 0.201171875, "epoch": 1.220472440944882, "explain_loss": 1.1875, "loss": 2.5625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.032958984375, "step": 2790 }, { "ce_loss": 0.05517578125, "distill_loss": 0.169921875, "epoch": 1.220472440944882, "explain_loss": 0.984375, "loss": 1.5546875, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.006103515625, "step": 2790 }, { "epoch": 1.2248468941382327, "grad_norm": 2.3574743270874023, "learning_rate": 0.0001376940240702501, "loss": 1.88525390625, "step": 2800 }, { "ce_loss": 0.0032196044921875, "distill_loss": 0.15234375, "epoch": 1.2248468941382327, "explain_loss": 0.9921875, "loss": 1.609375, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.006622314453125, "step": 2800 }, { "ce_loss": 0.0032196044921875, "distill_loss": 0.134765625, "epoch": 1.2248468941382327, "explain_loss": 1.1171875, "loss": 1.84375, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.005401611328125, "step": 2800 }, { "ce_loss": 0.013427734375, "distill_loss": 0.142578125, "epoch": 1.2248468941382327, "explain_loss": 0.921875, "loss": 1.3359375, "ref_ce_loss": 0.2490234375, "stage_align_loss": 0.005035400390625, "step": 2800 }, { "ce_loss": 0.103515625, "distill_loss": 0.24609375, "epoch": 1.2248468941382327, "explain_loss": 1.046875, "loss": 1.8203125, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.03466796875, "step": 2800 }, { "epoch": 1.2292213473315836, "grad_norm": 0.5550914406776428, "learning_rate": 0.00013724694731849401, "loss": 1.90224609375, "step": 2810 }, { "ce_loss": 0.00421142578125, "distill_loss": 0.095703125, "epoch": 1.2292213473315836, "explain_loss": 0.9375, "loss": 1.390625, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0047607421875, "step": 2810 }, { "ce_loss": 0.00213623046875, "distill_loss": 0.134765625, "epoch": 1.2292213473315836, "explain_loss": 1.328125, "loss": 1.921875, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.0036773681640625, "step": 2810 }, { "ce_loss": 0.318359375, "distill_loss": 0.099609375, "epoch": 1.2292213473315836, "explain_loss": 0.73046875, "loss": 1.5, "ref_ce_loss": 0.3125, "stage_align_loss": 0.033935546875, "step": 2810 }, { "ce_loss": 0.007110595703125, "distill_loss": 0.150390625, "epoch": 1.2292213473315836, "explain_loss": 1.140625, "loss": 1.75, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.0036468505859375, "step": 2810 }, { "epoch": 1.2335958005249343, "grad_norm": 0.7426575422286987, "learning_rate": 0.00013679900448127638, "loss": 1.88779296875, "step": 2820 }, { "ce_loss": 0.2470703125, "distill_loss": 0.10546875, "epoch": 1.2335958005249343, "explain_loss": 1.078125, "loss": 1.9609375, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.00592041015625, "step": 2820 }, { "ce_loss": 0.0185546875, "distill_loss": 0.058349609375, "epoch": 1.2335958005249343, "explain_loss": 1.28125, "loss": 1.875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.00494384765625, "step": 2820 }, { "ce_loss": 0.1669921875, "distill_loss": 0.046875, "epoch": 1.2335958005249343, "explain_loss": 1.1875, "loss": 2.109375, "ref_ce_loss": 0.68359375, "stage_align_loss": 0.0247802734375, "step": 2820 }, { "ce_loss": 0.006072998046875, "distill_loss": 0.2431640625, "epoch": 1.2335958005249343, "explain_loss": 1.28125, "loss": 1.9375, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.004302978515625, "step": 2820 }, { "epoch": 1.2379702537182853, "grad_norm": 1.550355076789856, "learning_rate": 0.00013635020597439786, "loss": 1.90234375, "step": 2830 }, { "ce_loss": 0.251953125, "distill_loss": 0.259765625, "epoch": 1.2379702537182853, "explain_loss": 1.09375, "loss": 2.171875, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.0191650390625, "step": 2830 }, { "ce_loss": 0.25, "distill_loss": 0.08984375, "epoch": 1.2379702537182853, "explain_loss": 0.9609375, "loss": 1.671875, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0089111328125, "step": 2830 }, { "ce_loss": 0.765625, "distill_loss": 0.046142578125, "epoch": 1.2379702537182853, "explain_loss": 1.46875, "loss": 2.78125, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.027587890625, "step": 2830 }, { "ce_loss": 0.19140625, "distill_loss": 0.123046875, "epoch": 1.2379702537182853, "explain_loss": 0.83984375, "loss": 1.578125, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0064697265625, "step": 2830 }, { "epoch": 1.242344706911636, "grad_norm": 1.0522243976593018, "learning_rate": 0.00013590056223355555, "loss": 1.82890625, "step": 2840 }, { "ce_loss": 0.02001953125, "distill_loss": 0.27734375, "epoch": 1.242344706911636, "explain_loss": 1.3515625, "loss": 2.0, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.005126953125, "step": 2840 }, { "ce_loss": 0.06494140625, "distill_loss": 0.03271484375, "epoch": 1.242344706911636, "explain_loss": 0.96484375, "loss": 1.390625, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.00555419921875, "step": 2840 }, { "ce_loss": 1.5625, "distill_loss": 0.0673828125, "epoch": 1.242344706911636, "explain_loss": 1.28125, "loss": 3.5, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.025634765625, "step": 2840 }, { "ce_loss": 0.271484375, "distill_loss": 0.031494140625, "epoch": 1.242344706911636, "explain_loss": 1.3125, "loss": 1.9609375, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.0206298828125, "step": 2840 }, { "epoch": 1.246719160104987, "grad_norm": 1.4725275039672852, "learning_rate": 0.00013545008371410052, "loss": 1.8521484375, "step": 2850 }, { "ce_loss": 0.01177978515625, "distill_loss": 0.10205078125, "epoch": 1.246719160104987, "explain_loss": 1.34375, "loss": 1.8203125, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0050048828125, "step": 2850 }, { "ce_loss": 0.01025390625, "distill_loss": 0.14453125, "epoch": 1.246719160104987, "explain_loss": 1.03125, "loss": 1.5546875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0042724609375, "step": 2850 }, { "ce_loss": 0.0201416015625, "distill_loss": 0.06591796875, "epoch": 1.246719160104987, "explain_loss": 1.546875, "loss": 2.125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0035552978515625, "step": 2850 }, { "ce_loss": 0.00775146484375, "distill_loss": 0.052734375, "epoch": 1.246719160104987, "explain_loss": 0.84765625, "loss": 1.09375, "ref_ce_loss": 0.1787109375, "stage_align_loss": 0.0047607421875, "step": 2850 }, { "epoch": 1.2510936132983377, "grad_norm": 4.027642726898193, "learning_rate": 0.00013499878089079435, "loss": 2.5548828125, "step": 2860 }, { "ce_loss": 0.07177734375, "distill_loss": 0.3671875, "epoch": 1.2510936132983377, "explain_loss": 0.9921875, "loss": 1.84375, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0081787109375, "step": 2860 }, { "ce_loss": 0.0130615234375, "distill_loss": 0.0712890625, "epoch": 1.2510936132983377, "explain_loss": 1.5, "loss": 2.03125, "ref_ce_loss": 0.4375, "stage_align_loss": 0.003265380859375, "step": 2860 }, { "ce_loss": 0.0654296875, "distill_loss": 0.134765625, "epoch": 1.2510936132983377, "explain_loss": 1.0859375, "loss": 1.796875, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0050048828125, "step": 2860 }, { "ce_loss": 1.0234375, "distill_loss": 0.1015625, "epoch": 1.2510936132983377, "explain_loss": 1.203125, "loss": 2.875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.00506591796875, "step": 2860 }, { "epoch": 1.2554680664916886, "grad_norm": 1.2574095726013184, "learning_rate": 0.00013454666425756598, "loss": 1.9365234375, "step": 2870 }, { "ce_loss": 0.006622314453125, "distill_loss": 0.05615234375, "epoch": 1.2554680664916886, "explain_loss": 0.9453125, "loss": 1.4375, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0036773681640625, "step": 2870 }, { "ce_loss": 0.390625, "distill_loss": 0.142578125, "epoch": 1.2554680664916886, "explain_loss": 1.640625, "loss": 2.75, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.06494140625, "step": 2870 }, { "ce_loss": 0.2294921875, "distill_loss": 0.0634765625, "epoch": 1.2554680664916886, "explain_loss": 1.015625, "loss": 1.640625, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.012451171875, "step": 2870 }, { "ce_loss": 3.234375, "distill_loss": 0.06640625, "epoch": 1.2554680664916886, "explain_loss": 1.171875, "loss": 4.9375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.00494384765625, "step": 2870 }, { "epoch": 1.2598425196850394, "grad_norm": 1.0732839107513428, "learning_rate": 0.00013409374432726723, "loss": 2.077734375, "step": 2880 }, { "ce_loss": 0.423828125, "distill_loss": 0.12255859375, "epoch": 1.2598425196850394, "explain_loss": 1.484375, "loss": 2.59375, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.003936767578125, "step": 2880 }, { "ce_loss": 0.51953125, "distill_loss": 0.0908203125, "epoch": 1.2598425196850394, "explain_loss": 0.75, "loss": 1.8515625, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.07763671875, "step": 2880 }, { "ce_loss": 0.26171875, "distill_loss": 0.09521484375, "epoch": 1.2598425196850394, "explain_loss": 1.0, "loss": 1.796875, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0322265625, "step": 2880 }, { "ce_loss": 0.0634765625, "distill_loss": 0.0869140625, "epoch": 1.2598425196850394, "explain_loss": 1.2109375, "loss": 2.015625, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.0034332275390625, "step": 2880 }, { "epoch": 1.26421697287839, "grad_norm": 1.62621009349823, "learning_rate": 0.00013364003163142884, "loss": 1.9375, "step": 2890 }, { "ce_loss": 0.22265625, "distill_loss": 0.064453125, "epoch": 1.26421697287839, "explain_loss": 0.80078125, "loss": 1.59375, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.053955078125, "step": 2890 }, { "ce_loss": 0.1630859375, "distill_loss": 0.03955078125, "epoch": 1.26421697287839, "explain_loss": 0.9765625, "loss": 1.65625, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.00384521484375, "step": 2890 }, { "ce_loss": 0.00823974609375, "distill_loss": 0.228515625, "epoch": 1.26421697287839, "explain_loss": 1.609375, "loss": 2.359375, "ref_ce_loss": 0.515625, "stage_align_loss": 0.002227783203125, "step": 2890 }, { "ce_loss": 0.0927734375, "distill_loss": 0.08056640625, "epoch": 1.26421697287839, "explain_loss": 0.9609375, "loss": 1.640625, "ref_ce_loss": 0.5, "stage_align_loss": 0.0113525390625, "step": 2890 }, { "epoch": 1.268591426071741, "grad_norm": 0.5034669041633606, "learning_rate": 0.00013318553672001519, "loss": 2.02294921875, "step": 2900 }, { "ce_loss": 0.51953125, "distill_loss": 0.2041015625, "epoch": 1.268591426071741, "explain_loss": 0.90625, "loss": 2.03125, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00531005859375, "step": 2900 }, { "ce_loss": 0.07275390625, "distill_loss": 0.17578125, "epoch": 1.268591426071741, "explain_loss": 1.390625, "loss": 2.34375, "ref_ce_loss": 0.703125, "stage_align_loss": 0.0032196044921875, "step": 2900 }, { "ce_loss": 0.326171875, "distill_loss": 0.18359375, "epoch": 1.268591426071741, "explain_loss": 1.3046875, "loss": 2.3125, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.00555419921875, "step": 2900 }, { "ce_loss": 0.19140625, "distill_loss": 0.049072265625, "epoch": 1.268591426071741, "explain_loss": 1.3515625, "loss": 2.09375, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.039306640625, "step": 2900 }, { "epoch": 1.272965879265092, "grad_norm": 1.3098713159561157, "learning_rate": 0.00013273027016117922, "loss": 1.959375, "step": 2910 }, { "ce_loss": 0.032958984375, "distill_loss": 0.1474609375, "epoch": 1.272965879265092, "explain_loss": 0.890625, "loss": 1.5078125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.00360107421875, "step": 2910 }, { "ce_loss": 0.003997802734375, "distill_loss": 0.099609375, "epoch": 1.272965879265092, "explain_loss": 0.796875, "loss": 1.140625, "ref_ce_loss": 0.240234375, "stage_align_loss": 0.004150390625, "step": 2910 }, { "ce_loss": 0.0166015625, "distill_loss": 0.08544921875, "epoch": 1.272965879265092, "explain_loss": 1.125, "loss": 1.578125, "ref_ce_loss": 0.34375, "stage_align_loss": 0.00408935546875, "step": 2910 }, { "ce_loss": 0.09521484375, "distill_loss": 0.142578125, "epoch": 1.272965879265092, "explain_loss": 0.8984375, "loss": 1.5, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.010009765625, "step": 2910 }, { "epoch": 1.2773403324584427, "grad_norm": 1.090000033378601, "learning_rate": 0.0001322742425410166, "loss": 1.96220703125, "step": 2920 }, { "ce_loss": 0.003692626953125, "distill_loss": 0.099609375, "epoch": 1.2773403324584427, "explain_loss": 0.796875, "loss": 1.3125, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.003143310546875, "step": 2920 }, { "ce_loss": 0.010009765625, "distill_loss": 0.14453125, "epoch": 1.2773403324584427, "explain_loss": 0.85546875, "loss": 1.4140625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0031890869140625, "step": 2920 }, { "ce_loss": 0.026123046875, "distill_loss": 0.083984375, "epoch": 1.2773403324584427, "explain_loss": 1.1484375, "loss": 1.671875, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.00347900390625, "step": 2920 }, { "ce_loss": 0.1484375, "distill_loss": 0.0712890625, "epoch": 1.2773403324584427, "explain_loss": 0.921875, "loss": 1.78125, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.00787353515625, "step": 2920 }, { "epoch": 1.2817147856517934, "grad_norm": 1.114511251449585, "learning_rate": 0.0001318174644633196, "loss": 1.84794921875, "step": 2930 }, { "ce_loss": 0.12060546875, "distill_loss": 0.083984375, "epoch": 1.2817147856517934, "explain_loss": 0.8984375, "loss": 1.375, "ref_ce_loss": 0.26953125, "stage_align_loss": 0.003631591796875, "step": 2930 }, { "ce_loss": 0.0220947265625, "distill_loss": 0.0947265625, "epoch": 1.2817147856517934, "explain_loss": 1.203125, "loss": 1.8125, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.007049560546875, "step": 2930 }, { "ce_loss": 0.4453125, "distill_loss": 0.1650390625, "epoch": 1.2817147856517934, "explain_loss": 1.1171875, "loss": 2.15625, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.040283203125, "step": 2930 }, { "ce_loss": 0.0184326171875, "distill_loss": 0.1640625, "epoch": 1.2817147856517934, "explain_loss": 1.421875, "loss": 2.171875, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.0037078857421875, "step": 2930 }, { "epoch": 1.2860892388451444, "grad_norm": 1.4715040922164917, "learning_rate": 0.00013135994654933056, "loss": 1.8494140625, "step": 2940 }, { "ce_loss": 0.08349609375, "distill_loss": 0.146484375, "epoch": 1.2860892388451444, "explain_loss": 0.9375, "loss": 1.578125, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00421142578125, "step": 2940 }, { "ce_loss": 0.2158203125, "distill_loss": 0.0712890625, "epoch": 1.2860892388451444, "explain_loss": 0.9140625, "loss": 1.5625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.003997802734375, "step": 2940 }, { "ce_loss": 0.61328125, "distill_loss": 0.0732421875, "epoch": 1.2860892388451444, "explain_loss": 1.359375, "loss": 2.515625, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.0076904296875, "step": 2940 }, { "ce_loss": 0.4296875, "distill_loss": 0.07080078125, "epoch": 1.2860892388451444, "explain_loss": 0.7109375, "loss": 1.640625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.042236328125, "step": 2940 }, { "epoch": 1.2904636920384953, "grad_norm": 0.8619838953018188, "learning_rate": 0.00013090169943749476, "loss": 1.8689453125, "step": 2950 }, { "ce_loss": 0.1455078125, "distill_loss": 0.154296875, "epoch": 1.2904636920384953, "explain_loss": 0.8671875, "loss": 1.625, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.0101318359375, "step": 2950 }, { "ce_loss": 0.72265625, "distill_loss": 0.1953125, "epoch": 1.2904636920384953, "explain_loss": 1.3203125, "loss": 2.8125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.06640625, "step": 2950 }, { "ce_loss": 0.017578125, "distill_loss": 0.08837890625, "epoch": 1.2904636920384953, "explain_loss": 0.9453125, "loss": 1.359375, "ref_ce_loss": 0.302734375, "stage_align_loss": 0.0037841796875, "step": 2950 }, { "ce_loss": 0.2109375, "distill_loss": 0.06787109375, "epoch": 1.2904636920384953, "explain_loss": 1.265625, "loss": 2.0625, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.03564453125, "step": 2950 }, { "epoch": 1.294838145231846, "grad_norm": 6.179172515869141, "learning_rate": 0.00013044273378321326, "loss": 2.04521484375, "step": 2960 }, { "ce_loss": 0.026611328125, "distill_loss": 0.2294921875, "epoch": 1.294838145231846, "explain_loss": 1.1953125, "loss": 1.875, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0038909912109375, "step": 2960 }, { "ce_loss": 0.431640625, "distill_loss": 0.140625, "epoch": 1.294838145231846, "explain_loss": 0.6953125, "loss": 1.8125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.00457763671875, "step": 2960 }, { "ce_loss": 0.09326171875, "distill_loss": 0.10546875, "epoch": 1.294838145231846, "explain_loss": 0.8828125, "loss": 1.421875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.00469970703125, "step": 2960 }, { "ce_loss": 0.06494140625, "distill_loss": 0.173828125, "epoch": 1.294838145231846, "explain_loss": 0.6484375, "loss": 1.171875, "ref_ce_loss": 0.27734375, "stage_align_loss": 0.004425048828125, "step": 2960 }, { "epoch": 1.2992125984251968, "grad_norm": 52.51261901855469, "learning_rate": 0.00012998306025859506, "loss": 2.3064453125, "step": 2970 }, { "ce_loss": 0.24609375, "distill_loss": 0.62109375, "epoch": 1.2992125984251968, "explain_loss": 1.390625, "loss": 2.71875, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.004150390625, "step": 2970 }, { "ce_loss": 0.1513671875, "distill_loss": 2.015625, "epoch": 1.2992125984251968, "explain_loss": 1.125, "loss": 3.671875, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.0038299560546875, "step": 2970 }, { "ce_loss": 0.1689453125, "distill_loss": 0.8359375, "epoch": 1.2992125984251968, "explain_loss": 1.1171875, "loss": 2.53125, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.00384521484375, "step": 2970 }, { "ce_loss": 0.00396728515625, "distill_loss": 1.46875, "epoch": 1.2992125984251968, "explain_loss": 0.80859375, "loss": 2.546875, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.0026092529296875, "step": 2970 }, { "epoch": 1.3035870516185477, "grad_norm": 7.305479526519775, "learning_rate": 0.00012952268955220887, "loss": 2.38564453125, "step": 2980 }, { "ce_loss": 0.027099609375, "distill_loss": 0.392578125, "epoch": 1.3035870516185477, "explain_loss": 1.015625, "loss": 1.96875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0037384033203125, "step": 2980 }, { "ce_loss": 0.146484375, "distill_loss": 0.138671875, "epoch": 1.3035870516185477, "explain_loss": 0.93359375, "loss": 1.671875, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0291748046875, "step": 2980 }, { "ce_loss": 0.051513671875, "distill_loss": 0.19140625, "epoch": 1.3035870516185477, "explain_loss": 1.765625, "loss": 2.34375, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.0029296875, "step": 2980 }, { "ce_loss": 0.7265625, "distill_loss": 0.21484375, "epoch": 1.3035870516185477, "explain_loss": 1.40625, "loss": 2.84375, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.060546875, "step": 2980 }, { "epoch": 1.3079615048118984, "grad_norm": 1.2797749042510986, "learning_rate": 0.0001290616323688347, "loss": 1.97333984375, "step": 2990 }, { "ce_loss": 0.38671875, "distill_loss": 0.10693359375, "epoch": 1.3079615048118984, "explain_loss": 1.3515625, "loss": 2.34375, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.06005859375, "step": 2990 }, { "ce_loss": 0.412109375, "distill_loss": 0.08251953125, "epoch": 1.3079615048118984, "explain_loss": 1.171875, "loss": 2.078125, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.0286865234375, "step": 2990 }, { "ce_loss": 0.0203857421875, "distill_loss": 0.12890625, "epoch": 1.3079615048118984, "explain_loss": 1.0625, "loss": 1.6328125, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.00689697265625, "step": 2990 }, { "ce_loss": 0.0693359375, "distill_loss": 0.09033203125, "epoch": 1.3079615048118984, "explain_loss": 1.2421875, "loss": 1.796875, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.00445556640625, "step": 2990 }, { "epoch": 1.3123359580052494, "grad_norm": 0.836749255657196, "learning_rate": 0.00012859989942921483, "loss": 1.8109375, "step": 3000 }, { "ce_loss": 1.1015625, "distill_loss": 0.0615234375, "epoch": 1.3123359580052494, "explain_loss": 1.125, "loss": 2.859375, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.06591796875, "step": 3000 }, { "ce_loss": 0.72265625, "distill_loss": 0.08740234375, "epoch": 1.3123359580052494, "explain_loss": 1.09375, "loss": 2.28125, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0045166015625, "step": 3000 }, { "ce_loss": 1.453125, "distill_loss": 0.08349609375, "epoch": 1.3123359580052494, "explain_loss": 0.953125, "loss": 2.84375, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.005340576171875, "step": 3000 }, { "ce_loss": 0.1875, "distill_loss": 0.09228515625, "epoch": 1.3123359580052494, "explain_loss": 0.96875, "loss": 1.7421875, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.01708984375, "step": 3000 }, { "epoch": 1.3167104111986, "grad_norm": 0.7936205863952637, "learning_rate": 0.00012813750146980455, "loss": 1.87314453125, "step": 3010 }, { "ce_loss": 0.04248046875, "distill_loss": 0.06640625, "epoch": 1.3167104111986, "explain_loss": 1.015625, "loss": 1.46875, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.00335693359375, "step": 3010 }, { "ce_loss": 0.00136566162109375, "distill_loss": 0.0888671875, "epoch": 1.3167104111986, "explain_loss": 1.109375, "loss": 1.5625, "ref_ce_loss": 0.359375, "stage_align_loss": 0.00323486328125, "step": 3010 }, { "ce_loss": 0.010498046875, "distill_loss": 0.1025390625, "epoch": 1.3167104111986, "explain_loss": 1.125, "loss": 1.53125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0032806396484375, "step": 3010 }, { "ce_loss": 0.0034027099609375, "distill_loss": 0.091796875, "epoch": 1.3167104111986, "explain_loss": 1.140625, "loss": 1.71875, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0028228759765625, "step": 3010 }, { "epoch": 1.321084864391951, "grad_norm": 0.7742398381233215, "learning_rate": 0.00012767444924252264, "loss": 1.91689453125, "step": 3020 }, { "ce_loss": 0.251953125, "distill_loss": 0.07666015625, "epoch": 1.321084864391951, "explain_loss": 1.484375, "loss": 2.4375, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.004791259765625, "step": 3020 }, { "ce_loss": 0.0196533203125, "distill_loss": 0.08056640625, "epoch": 1.321084864391951, "explain_loss": 1.0625, "loss": 1.5859375, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.00201416015625, "step": 3020 }, { "ce_loss": 0.0019073486328125, "distill_loss": 0.08984375, "epoch": 1.321084864391951, "explain_loss": 1.0234375, "loss": 1.4375, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.0032958984375, "step": 3020 }, { "ce_loss": 0.0400390625, "distill_loss": 0.083984375, "epoch": 1.321084864391951, "explain_loss": 1.234375, "loss": 1.796875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.021240234375, "step": 3020 }, { "epoch": 1.3254593175853018, "grad_norm": 1.793794870376587, "learning_rate": 0.00012721075351450114, "loss": 2.4875, "step": 3030 }, { "ce_loss": 0.087890625, "distill_loss": 0.107421875, "epoch": 1.3254593175853018, "explain_loss": 1.296875, "loss": 2.0, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0206298828125, "step": 3030 }, { "ce_loss": 0.022216796875, "distill_loss": 0.150390625, "epoch": 1.3254593175853018, "explain_loss": 0.90234375, "loss": 1.328125, "ref_ce_loss": 0.2451171875, "stage_align_loss": 0.00323486328125, "step": 3030 }, { "ce_loss": 0.515625, "distill_loss": 0.1337890625, "epoch": 1.3254593175853018, "explain_loss": 1.09375, "loss": 2.375, "ref_ce_loss": 0.578125, "stage_align_loss": 0.060546875, "step": 3030 }, { "ce_loss": 0.0703125, "distill_loss": 0.31640625, "epoch": 1.3254593175853018, "explain_loss": 1.2578125, "loss": 2.109375, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.0036773681640625, "step": 3030 }, { "epoch": 1.3298337707786527, "grad_norm": 1.0808335542678833, "learning_rate": 0.00012674642506783524, "loss": 1.9623046875, "step": 3040 }, { "ce_loss": 0.11474609375, "distill_loss": 0.11279296875, "epoch": 1.3298337707786527, "explain_loss": 1.140625, "loss": 1.8984375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0040283203125, "step": 3040 }, { "ce_loss": 0.01513671875, "distill_loss": 0.1005859375, "epoch": 1.3298337707786527, "explain_loss": 1.328125, "loss": 2.078125, "ref_ce_loss": 0.625, "stage_align_loss": 0.002899169921875, "step": 3040 }, { "ce_loss": 0.1201171875, "distill_loss": 0.08056640625, "epoch": 1.3298337707786527, "explain_loss": 0.96484375, "loss": 1.625, "ref_ce_loss": 0.453125, "stage_align_loss": 0.003875732421875, "step": 3040 }, { "ce_loss": 0.32421875, "distill_loss": 0.0869140625, "epoch": 1.3298337707786527, "explain_loss": 1.3203125, "loss": 2.078125, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.0120849609375, "step": 3040 }, { "epoch": 1.3342082239720034, "grad_norm": 0.7669210433959961, "learning_rate": 0.00012628147469933226, "loss": 1.85322265625, "step": 3050 }, { "ce_loss": 0.3125, "distill_loss": 0.08056640625, "epoch": 1.3342082239720034, "explain_loss": 1.1875, "loss": 2.03125, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.003814697265625, "step": 3050 }, { "ce_loss": 0.484375, "distill_loss": 0.08203125, "epoch": 1.3342082239720034, "explain_loss": 1.28125, "loss": 2.40625, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.06494140625, "step": 3050 }, { "ce_loss": 0.11328125, "distill_loss": 0.103515625, "epoch": 1.3342082239720034, "explain_loss": 1.1875, "loss": 1.8046875, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.01470947265625, "step": 3050 }, { "ce_loss": 0.048095703125, "distill_loss": 0.1416015625, "epoch": 1.3342082239720034, "explain_loss": 1.1171875, "loss": 1.7578125, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.00567626953125, "step": 3050 }, { "epoch": 1.3385826771653544, "grad_norm": 0.9342232346534729, "learning_rate": 0.00012581591322026093, "loss": 1.822265625, "step": 3060 }, { "ce_loss": 0.1318359375, "distill_loss": 0.177734375, "epoch": 1.3385826771653544, "explain_loss": 1.1484375, "loss": 2.0, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.00787353515625, "step": 3060 }, { "ce_loss": 0.0072021484375, "distill_loss": 0.09765625, "epoch": 1.3385826771653544, "explain_loss": 0.77734375, "loss": 1.125, "ref_ce_loss": 0.2373046875, "stage_align_loss": 0.004486083984375, "step": 3060 }, { "ce_loss": 0.01336669921875, "distill_loss": 0.1123046875, "epoch": 1.3385826771653544, "explain_loss": 0.96875, "loss": 1.7421875, "ref_ce_loss": 0.64453125, "stage_align_loss": 0.0038299560546875, "step": 3060 }, { "ce_loss": 0.154296875, "distill_loss": 0.119140625, "epoch": 1.3385826771653544, "explain_loss": 0.890625, "loss": 1.578125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.046142578125, "step": 3060 }, { "epoch": 1.3429571303587051, "grad_norm": 1.0518914461135864, "learning_rate": 0.00012534975145609976, "loss": 1.89248046875, "step": 3070 }, { "ce_loss": 0.0849609375, "distill_loss": 0.08837890625, "epoch": 1.3429571303587051, "explain_loss": 1.03125, "loss": 1.59375, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.003814697265625, "step": 3070 }, { "ce_loss": 0.0361328125, "distill_loss": 0.109375, "epoch": 1.3429571303587051, "explain_loss": 1.09375, "loss": 1.671875, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.00518798828125, "step": 3070 }, { "ce_loss": 0.671875, "distill_loss": 0.1123046875, "epoch": 1.3429571303587051, "explain_loss": 0.9375, "loss": 2.203125, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.005889892578125, "step": 3070 }, { "ce_loss": 0.267578125, "distill_loss": 0.10693359375, "epoch": 1.3429571303587051, "explain_loss": 1.1171875, "loss": 2.0625, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.0098876953125, "step": 3070 }, { "epoch": 1.347331583552056, "grad_norm": 0.7683847546577454, "learning_rate": 0.00012488300024628544, "loss": 1.78046875, "step": 3080 }, { "ce_loss": 0.14453125, "distill_loss": 0.1279296875, "epoch": 1.347331583552056, "explain_loss": 0.77734375, "loss": 1.5, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.017578125, "step": 3080 }, { "ce_loss": 0.33203125, "distill_loss": 0.142578125, "epoch": 1.347331583552056, "explain_loss": 1.1015625, "loss": 1.921875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.0052490234375, "step": 3080 }, { "ce_loss": 0.061279296875, "distill_loss": 0.08544921875, "epoch": 1.347331583552056, "explain_loss": 1.03125, "loss": 1.640625, "ref_ce_loss": 0.46484375, "stage_align_loss": 0.00360107421875, "step": 3080 }, { "ce_loss": 0.009521484375, "distill_loss": 0.08984375, "epoch": 1.347331583552056, "explain_loss": 1.3125, "loss": 1.890625, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0022735595703125, "step": 3080 }, { "epoch": 1.3517060367454068, "grad_norm": 0.8836283087730408, "learning_rate": 0.00012441567044396072, "loss": 1.76162109375, "step": 3090 }, { "ce_loss": 0.205078125, "distill_loss": 0.146484375, "epoch": 1.3517060367454068, "explain_loss": 0.94921875, "loss": 1.703125, "ref_ce_loss": 0.375, "stage_align_loss": 0.023193359375, "step": 3090 }, { "ce_loss": 0.00494384765625, "distill_loss": 0.083984375, "epoch": 1.3517060367454068, "explain_loss": 1.1875, "loss": 1.6484375, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.0028839111328125, "step": 3090 }, { "ce_loss": 0.041015625, "distill_loss": 0.0966796875, "epoch": 1.3517060367454068, "explain_loss": 0.9921875, "loss": 1.6875, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.0031585693359375, "step": 3090 }, { "ce_loss": 0.07470703125, "distill_loss": 0.10986328125, "epoch": 1.3517060367454068, "explain_loss": 1.1015625, "loss": 1.546875, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.004547119140625, "step": 3090 }, { "epoch": 1.3560804899387577, "grad_norm": 4.35458517074585, "learning_rate": 0.00012394777291572213, "loss": 1.93994140625, "step": 3100 }, { "ce_loss": 0.369140625, "distill_loss": 0.31640625, "epoch": 1.3560804899387577, "explain_loss": 0.859375, "loss": 1.9296875, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.026611328125, "step": 3100 }, { "ce_loss": 0.007354736328125, "distill_loss": 0.12451171875, "epoch": 1.3560804899387577, "explain_loss": 0.9140625, "loss": 1.453125, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.005035400390625, "step": 3100 }, { "ce_loss": 1.5625, "distill_loss": 0.0927734375, "epoch": 1.3560804899387577, "explain_loss": 1.5, "loss": 3.84375, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.09375, "step": 3100 }, { "ce_loss": 0.0021209716796875, "distill_loss": 0.130859375, "epoch": 1.3560804899387577, "explain_loss": 1.203125, "loss": 1.875, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.004638671875, "step": 3100 }, { "epoch": 1.3604549431321085, "grad_norm": 2.146939516067505, "learning_rate": 0.00012347931854136718, "loss": 2.07392578125, "step": 3110 }, { "ce_loss": 0.029541015625, "distill_loss": 0.1806640625, "epoch": 1.3604549431321085, "explain_loss": 1.359375, "loss": 2.140625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.003265380859375, "step": 3110 }, { "ce_loss": 0.08935546875, "distill_loss": 0.1845703125, "epoch": 1.3604549431321085, "explain_loss": 0.8359375, "loss": 1.5703125, "ref_ce_loss": 0.421875, "stage_align_loss": 0.037841796875, "step": 3110 }, { "ce_loss": 0.50390625, "distill_loss": 0.16015625, "epoch": 1.3604549431321085, "explain_loss": 0.97265625, "loss": 2.5625, "ref_ce_loss": 0.88671875, "stage_align_loss": 0.048583984375, "step": 3110 }, { "ce_loss": 0.1962890625, "distill_loss": 0.203125, "epoch": 1.3604549431321085, "explain_loss": 1.328125, "loss": 2.34375, "ref_ce_loss": 0.6015625, "stage_align_loss": 0.017822265625, "step": 3110 }, { "epoch": 1.3648293963254594, "grad_norm": 1.89054274559021, "learning_rate": 0.00012301031821364158, "loss": 1.98974609375, "step": 3120 }, { "ce_loss": 0.051025390625, "distill_loss": 0.05859375, "epoch": 1.3648293963254594, "explain_loss": 1.1015625, "loss": 1.578125, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.00421142578125, "step": 3120 }, { "ce_loss": 0.048095703125, "distill_loss": 0.0908203125, "epoch": 1.3648293963254594, "explain_loss": 0.94140625, "loss": 1.4375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.005035400390625, "step": 3120 }, { "ce_loss": 0.162109375, "distill_loss": 0.10009765625, "epoch": 1.3648293963254594, "explain_loss": 0.78515625, "loss": 1.234375, "ref_ce_loss": 0.1796875, "stage_align_loss": 0.004150390625, "step": 3120 }, { "ce_loss": 0.00335693359375, "distill_loss": 0.2392578125, "epoch": 1.3648293963254594, "explain_loss": 1.1484375, "loss": 1.78125, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.00762939453125, "step": 3120 }, { "epoch": 1.3692038495188101, "grad_norm": 0.7518616914749146, "learning_rate": 0.0001225407828379857, "loss": 1.82822265625, "step": 3130 }, { "ce_loss": 0.2421875, "distill_loss": 0.134765625, "epoch": 1.3692038495188101, "explain_loss": 1.4296875, "loss": 2.234375, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.00677490234375, "step": 3130 }, { "ce_loss": 0.00421142578125, "distill_loss": 0.1767578125, "epoch": 1.3692038495188101, "explain_loss": 0.76171875, "loss": 1.25, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.00567626953125, "step": 3130 }, { "ce_loss": 0.10302734375, "distill_loss": 0.25390625, "epoch": 1.3692038495188101, "explain_loss": 1.40625, "loss": 2.34375, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.006988525390625, "step": 3130 }, { "ce_loss": 0.00225830078125, "distill_loss": 0.09033203125, "epoch": 1.3692038495188101, "explain_loss": 0.9375, "loss": 1.40625, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.003173828125, "step": 3130 }, { "epoch": 1.3735783027121609, "grad_norm": 5.93734884262085, "learning_rate": 0.00012207072333228128, "loss": 1.916455078125, "step": 3140 }, { "ce_loss": 0.5078125, "distill_loss": 0.09033203125, "epoch": 1.3735783027121609, "explain_loss": 1.0078125, "loss": 2.078125, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.007080078125, "step": 3140 }, { "ce_loss": 1.4921875, "distill_loss": 0.08984375, "epoch": 1.3735783027121609, "explain_loss": 1.203125, "loss": 3.234375, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.0029144287109375, "step": 3140 }, { "ce_loss": 0.014892578125, "distill_loss": 0.0595703125, "epoch": 1.3735783027121609, "explain_loss": 1.3359375, "loss": 2.078125, "ref_ce_loss": 0.6640625, "stage_align_loss": 0.0037689208984375, "step": 3140 }, { "ce_loss": 0.04345703125, "distill_loss": 0.16796875, "epoch": 1.3735783027121609, "explain_loss": 1.5, "loss": 2.171875, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.0031585693359375, "step": 3140 }, { "epoch": 1.3779527559055118, "grad_norm": 2.602163076400757, "learning_rate": 0.0001216001506265972, "loss": 1.9068359375, "step": 3150 }, { "ce_loss": 0.00897216796875, "distill_loss": 0.07421875, "epoch": 1.3779527559055118, "explain_loss": 1.078125, "loss": 1.671875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0037689208984375, "step": 3150 }, { "ce_loss": 0.32421875, "distill_loss": 0.134765625, "epoch": 1.3779527559055118, "explain_loss": 1.15625, "loss": 2.09375, "ref_ce_loss": 0.46875, "stage_align_loss": 0.006622314453125, "step": 3150 }, { "ce_loss": 0.41015625, "distill_loss": 0.1298828125, "epoch": 1.3779527559055118, "explain_loss": 1.28125, "loss": 2.296875, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.006805419921875, "step": 3150 }, { "ce_loss": 0.0260009765625, "distill_loss": 0.06787109375, "epoch": 1.3779527559055118, "explain_loss": 0.7421875, "loss": 1.21875, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.0030517578125, "step": 3150 }, { "epoch": 1.3823272090988628, "grad_norm": 0.9618291258811951, "learning_rate": 0.0001211290756629357, "loss": 1.86826171875, "step": 3160 }, { "ce_loss": 0.07666015625, "distill_loss": 0.0830078125, "epoch": 1.3823272090988628, "explain_loss": 1.3046875, "loss": 2.046875, "ref_ce_loss": 0.578125, "stage_align_loss": 0.005615234375, "step": 3160 }, { "ce_loss": 0.00830078125, "distill_loss": 0.15234375, "epoch": 1.3823272090988628, "explain_loss": 1.0, "loss": 1.5, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.0026702880859375, "step": 3160 }, { "ce_loss": 0.14453125, "distill_loss": 0.1171875, "epoch": 1.3823272090988628, "explain_loss": 1.171875, "loss": 1.9921875, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.004119873046875, "step": 3160 }, { "ce_loss": 0.0498046875, "distill_loss": 0.09521484375, "epoch": 1.3823272090988628, "explain_loss": 0.9296875, "loss": 1.515625, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0035552978515625, "step": 3160 }, { "epoch": 1.3867016622922135, "grad_norm": 2.1852755546569824, "learning_rate": 0.00012065750939497776, "loss": 1.850390625, "step": 3170 }, { "ce_loss": 0.271484375, "distill_loss": 0.07373046875, "epoch": 1.3867016622922135, "explain_loss": 1.1015625, "loss": 1.765625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.01092529296875, "step": 3170 }, { "ce_loss": 0.37109375, "distill_loss": 0.0751953125, "epoch": 1.3867016622922135, "explain_loss": 0.875, "loss": 1.65625, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0166015625, "step": 3170 }, { "ce_loss": 0.002716064453125, "distill_loss": 0.16015625, "epoch": 1.3867016622922135, "explain_loss": 1.4140625, "loss": 2.015625, "ref_ce_loss": 0.4375, "stage_align_loss": 0.00360107421875, "step": 3170 }, { "ce_loss": 0.5078125, "distill_loss": 0.1376953125, "epoch": 1.3867016622922135, "explain_loss": 1.171875, "loss": 2.1875, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0198974609375, "step": 3170 }, { "epoch": 1.3910761154855642, "grad_norm": 0.741684079170227, "learning_rate": 0.00012018546278782831, "loss": 1.80703125, "step": 3180 }, { "ce_loss": 0.02734375, "distill_loss": 0.10693359375, "epoch": 1.3910761154855642, "explain_loss": 0.9921875, "loss": 1.59375, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.00347900390625, "step": 3180 }, { "ce_loss": 0.72265625, "distill_loss": 0.1376953125, "epoch": 1.3910761154855642, "explain_loss": 0.96875, "loss": 2.375, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.09912109375, "step": 3180 }, { "ce_loss": 0.419921875, "distill_loss": 0.0595703125, "epoch": 1.3910761154855642, "explain_loss": 0.93359375, "loss": 1.953125, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.083984375, "step": 3180 }, { "ce_loss": 0.0052490234375, "distill_loss": 0.10986328125, "epoch": 1.3910761154855642, "explain_loss": 1.015625, "loss": 1.546875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.0036163330078125, "step": 3180 }, { "epoch": 1.3954505686789151, "grad_norm": 0.6729668378829956, "learning_rate": 0.00011971294681776157, "loss": 1.91083984375, "step": 3190 }, { "ce_loss": 0.3359375, "distill_loss": 0.0986328125, "epoch": 1.3954505686789151, "explain_loss": 1.140625, "loss": 1.90625, "ref_ce_loss": 0.328125, "stage_align_loss": 0.004608154296875, "step": 3190 }, { "ce_loss": 0.12060546875, "distill_loss": 0.2890625, "epoch": 1.3954505686789151, "explain_loss": 0.91796875, "loss": 1.90625, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.03857421875, "step": 3190 }, { "ce_loss": 0.05810546875, "distill_loss": 0.0751953125, "epoch": 1.3954505686789151, "explain_loss": 1.046875, "loss": 1.703125, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.005767822265625, "step": 3190 }, { "ce_loss": 0.0228271484375, "distill_loss": 0.16796875, "epoch": 1.3954505686789151, "explain_loss": 0.91015625, "loss": 1.4375, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0037384033203125, "step": 3190 }, { "epoch": 1.399825021872266, "grad_norm": 1.5154081583023071, "learning_rate": 0.00011923997247196545, "loss": 1.90888671875, "step": 3200 }, { "ce_loss": 0.078125, "distill_loss": 0.138671875, "epoch": 1.399825021872266, "explain_loss": 0.9375, "loss": 1.3203125, "ref_ce_loss": 0.1611328125, "stage_align_loss": 0.004852294921875, "step": 3200 }, { "ce_loss": 0.08837890625, "distill_loss": 0.05810546875, "epoch": 1.399825021872266, "explain_loss": 1.4375, "loss": 2.0, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0140380859375, "step": 3200 }, { "ce_loss": 0.1328125, "distill_loss": 0.1171875, "epoch": 1.399825021872266, "explain_loss": 0.8515625, "loss": 1.4140625, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.006866455078125, "step": 3200 }, { "ce_loss": 0.0225830078125, "distill_loss": 0.1328125, "epoch": 1.399825021872266, "explain_loss": 0.921875, "loss": 1.5, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00360107421875, "step": 3200 }, { "epoch": 1.4041994750656168, "grad_norm": 1.7288461923599243, "learning_rate": 0.00011876655074828635, "loss": 1.86376953125, "step": 3210 }, { "ce_loss": 0.08056640625, "distill_loss": 0.23046875, "epoch": 1.4041994750656168, "explain_loss": 1.15625, "loss": 1.796875, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.003662109375, "step": 3210 }, { "ce_loss": 0.00933837890625, "distill_loss": 0.04833984375, "epoch": 1.4041994750656168, "explain_loss": 1.2421875, "loss": 1.6484375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.01507568359375, "step": 3210 }, { "ce_loss": 0.2451171875, "distill_loss": 0.115234375, "epoch": 1.4041994750656168, "explain_loss": 0.8046875, "loss": 1.7734375, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.0205078125, "step": 3210 }, { "ce_loss": 0.0091552734375, "distill_loss": 0.0703125, "epoch": 1.4041994750656168, "explain_loss": 0.94140625, "loss": 1.4375, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.00372314453125, "step": 3210 }, { "epoch": 1.4085739282589675, "grad_norm": 0.8299067616462708, "learning_rate": 0.00011829269265497332, "loss": 1.83759765625, "step": 3220 }, { "ce_loss": 0.416015625, "distill_loss": 0.17578125, "epoch": 1.4085739282589675, "explain_loss": 0.7578125, "loss": 1.78125, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.03515625, "step": 3220 }, { "ce_loss": 0.0181884765625, "distill_loss": 0.1181640625, "epoch": 1.4085739282589675, "explain_loss": 1.3203125, "loss": 1.8046875, "ref_ce_loss": 0.34375, "stage_align_loss": 0.0027618408203125, "step": 3220 }, { "ce_loss": 0.55078125, "distill_loss": 0.06396484375, "epoch": 1.4085739282589675, "explain_loss": 1.171875, "loss": 2.34375, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.026611328125, "step": 3220 }, { "ce_loss": 0.07177734375, "distill_loss": 0.126953125, "epoch": 1.4085739282589675, "explain_loss": 1.0078125, "loss": 1.625, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.005340576171875, "step": 3220 }, { "epoch": 1.4129483814523185, "grad_norm": 1.6496254205703735, "learning_rate": 0.00011781840921042206, "loss": 1.880859375, "step": 3230 }, { "ce_loss": 0.302734375, "distill_loss": 0.08740234375, "epoch": 1.4129483814523185, "explain_loss": 1.2578125, "loss": 2.015625, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0054931640625, "step": 3230 }, { "ce_loss": 0.228515625, "distill_loss": 0.051025390625, "epoch": 1.4129483814523185, "explain_loss": 1.109375, "loss": 1.921875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0311279296875, "step": 3230 }, { "ce_loss": 0.44921875, "distill_loss": 0.1142578125, "epoch": 1.4129483814523185, "explain_loss": 0.6171875, "loss": 1.6875, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.0216064453125, "step": 3230 }, { "ce_loss": 0.076171875, "distill_loss": 0.1103515625, "epoch": 1.4129483814523185, "explain_loss": 0.9921875, "loss": 1.4375, "ref_ce_loss": 0.2421875, "stage_align_loss": 0.01953125, "step": 3230 }, { "epoch": 1.4173228346456692, "grad_norm": 1.8360990285873413, "learning_rate": 0.00011734371144291882, "loss": 1.8046875, "step": 3240 }, { "ce_loss": 0.02734375, "distill_loss": 0.1083984375, "epoch": 1.4173228346456692, "explain_loss": 0.8125, "loss": 1.2578125, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.00439453125, "step": 3240 }, { "ce_loss": 0.00823974609375, "distill_loss": 0.1484375, "epoch": 1.4173228346456692, "explain_loss": 1.234375, "loss": 1.796875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.004302978515625, "step": 3240 }, { "ce_loss": 0.099609375, "distill_loss": 0.058837890625, "epoch": 1.4173228346456692, "explain_loss": 1.1640625, "loss": 1.96875, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.005218505859375, "step": 3240 }, { "ce_loss": 0.01458740234375, "distill_loss": 0.07763671875, "epoch": 1.4173228346456692, "explain_loss": 0.953125, "loss": 1.5390625, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.004302978515625, "step": 3240 }, { "epoch": 1.4216972878390202, "grad_norm": 2.761064052581787, "learning_rate": 0.00011686861039038386, "loss": 1.7916015625, "step": 3250 }, { "ce_loss": 0.091796875, "distill_loss": 0.11962890625, "epoch": 1.4216972878390202, "explain_loss": 1.1328125, "loss": 1.84375, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.0164794921875, "step": 3250 }, { "ce_loss": 0.03515625, "distill_loss": 0.054931640625, "epoch": 1.4216972878390202, "explain_loss": 1.125, "loss": 1.625, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.003814697265625, "step": 3250 }, { "ce_loss": 0.0019989013671875, "distill_loss": 0.166015625, "epoch": 1.4216972878390202, "explain_loss": 0.984375, "loss": 1.484375, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.003448486328125, "step": 3250 }, { "ce_loss": 0.10205078125, "distill_loss": 0.091796875, "epoch": 1.4216972878390202, "explain_loss": 0.80859375, "loss": 1.25, "ref_ce_loss": 0.2451171875, "stage_align_loss": 0.004730224609375, "step": 3250 }, { "epoch": 1.4260717410323709, "grad_norm": 0.7047882676124573, "learning_rate": 0.00011639311710011489, "loss": 1.8083984375, "step": 3260 }, { "ce_loss": 2.3125, "distill_loss": 0.09228515625, "epoch": 1.4260717410323709, "explain_loss": 1.2421875, "loss": 4.3125, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.08056640625, "step": 3260 }, { "ce_loss": 0.89453125, "distill_loss": 0.203125, "epoch": 1.4260717410323709, "explain_loss": 1.15625, "loss": 2.703125, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.03515625, "step": 3260 }, { "ce_loss": 0.00555419921875, "distill_loss": 0.06005859375, "epoch": 1.4260717410323709, "explain_loss": 0.875, "loss": 1.34375, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0030364990234375, "step": 3260 }, { "ce_loss": 0.0018463134765625, "distill_loss": 0.054931640625, "epoch": 1.4260717410323709, "explain_loss": 0.8984375, "loss": 1.5, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0034027099609375, "step": 3260 }, { "epoch": 1.4304461942257218, "grad_norm": 1.3956198692321777, "learning_rate": 0.00011591724262853008, "loss": 1.84853515625, "step": 3270 }, { "ce_loss": 0.1318359375, "distill_loss": 0.115234375, "epoch": 1.4304461942257218, "explain_loss": 1.28125, "loss": 1.921875, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0038604736328125, "step": 3270 }, { "ce_loss": 0.0206298828125, "distill_loss": 0.177734375, "epoch": 1.4304461942257218, "explain_loss": 1.421875, "loss": 2.21875, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.002349853515625, "step": 3270 }, { "ce_loss": 0.0185546875, "distill_loss": 0.2294921875, "epoch": 1.4304461942257218, "explain_loss": 0.828125, "loss": 1.484375, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.003082275390625, "step": 3270 }, { "ce_loss": 0.97265625, "distill_loss": 0.1083984375, "epoch": 1.4304461942257218, "explain_loss": 0.62109375, "loss": 2.140625, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.09033203125, "step": 3270 }, { "epoch": 1.4348206474190726, "grad_norm": 1.5825932025909424, "learning_rate": 0.00011544099804091102, "loss": 1.8126953125, "step": 3280 }, { "ce_loss": 0.00341796875, "distill_loss": 0.11181640625, "epoch": 1.4348206474190726, "explain_loss": 1.078125, "loss": 1.65625, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.003692626953125, "step": 3280 }, { "ce_loss": 0.484375, "distill_loss": 0.091796875, "epoch": 1.4348206474190726, "explain_loss": 0.90625, "loss": 1.9453125, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.07470703125, "step": 3280 }, { "ce_loss": 0.002471923828125, "distill_loss": 0.087890625, "epoch": 1.4348206474190726, "explain_loss": 0.9765625, "loss": 1.4375, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0032958984375, "step": 3280 }, { "ce_loss": 0.38671875, "distill_loss": 0.057373046875, "epoch": 1.4348206474190726, "explain_loss": 0.84375, "loss": 1.7265625, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.021484375, "step": 3280 }, { "epoch": 1.4391951006124235, "grad_norm": 0.9937505722045898, "learning_rate": 0.00011496439441114549, "loss": 1.7396484375, "step": 3290 }, { "ce_loss": 0.09912109375, "distill_loss": 0.142578125, "epoch": 1.4391951006124235, "explain_loss": 1.078125, "loss": 1.5390625, "ref_ce_loss": 0.2158203125, "stage_align_loss": 0.005340576171875, "step": 3290 }, { "ce_loss": 0.021240234375, "distill_loss": 0.111328125, "epoch": 1.4391951006124235, "explain_loss": 1.21875, "loss": 1.765625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.003692626953125, "step": 3290 }, { "ce_loss": 0.0107421875, "distill_loss": 0.11474609375, "epoch": 1.4391951006124235, "explain_loss": 1.078125, "loss": 1.640625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.003082275390625, "step": 3290 }, { "ce_loss": 0.0546875, "distill_loss": 0.1376953125, "epoch": 1.4391951006124235, "explain_loss": 1.1796875, "loss": 1.96875, "ref_ce_loss": 0.59765625, "stage_align_loss": 0.0037384033203125, "step": 3290 }, { "epoch": 1.4435695538057742, "grad_norm": 1.4117013216018677, "learning_rate": 0.00011448744282146982, "loss": 1.877734375, "step": 3300 }, { "ce_loss": 0.0010833740234375, "distill_loss": 0.0712890625, "epoch": 1.4435695538057742, "explain_loss": 1.0234375, "loss": 1.515625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.00286865234375, "step": 3300 }, { "ce_loss": 0.041259765625, "distill_loss": 0.11181640625, "epoch": 1.4435695538057742, "explain_loss": 0.9296875, "loss": 1.46875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.004119873046875, "step": 3300 }, { "ce_loss": 0.038818359375, "distill_loss": 0.099609375, "epoch": 1.4435695538057742, "explain_loss": 1.359375, "loss": 1.984375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.0028533935546875, "step": 3300 }, { "ce_loss": 0.087890625, "distill_loss": 0.0576171875, "epoch": 1.4435695538057742, "explain_loss": 0.9453125, "loss": 1.453125, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.005706787109375, "step": 3300 }, { "epoch": 1.4479440069991252, "grad_norm": 0.8935161232948303, "learning_rate": 0.00011401015436221138, "loss": 1.82373046875, "step": 3310 }, { "ce_loss": 0.0130615234375, "distill_loss": 0.12451171875, "epoch": 1.4479440069991252, "explain_loss": 1.078125, "loss": 1.515625, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.0031890869140625, "step": 3310 }, { "ce_loss": 0.12109375, "distill_loss": 0.1259765625, "epoch": 1.4479440069991252, "explain_loss": 1.0, "loss": 1.6328125, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.007293701171875, "step": 3310 }, { "ce_loss": 0.0634765625, "distill_loss": 0.050048828125, "epoch": 1.4479440069991252, "explain_loss": 0.9765625, "loss": 1.5234375, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.0191650390625, "step": 3310 }, { "ce_loss": 0.0869140625, "distill_loss": 0.107421875, "epoch": 1.4479440069991252, "explain_loss": 1.078125, "loss": 1.890625, "ref_ce_loss": 0.609375, "stage_align_loss": 0.005859375, "step": 3310 }, { "epoch": 1.452318460192476, "grad_norm": 0.9439171552658081, "learning_rate": 0.00011353254013153059, "loss": 1.79365234375, "step": 3320 }, { "ce_loss": 0.018798828125, "distill_loss": 0.1103515625, "epoch": 1.452318460192476, "explain_loss": 0.72265625, "loss": 1.1484375, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.004425048828125, "step": 3320 }, { "ce_loss": 0.62890625, "distill_loss": 0.0712890625, "epoch": 1.452318460192476, "explain_loss": 0.8984375, "loss": 2.109375, "ref_ce_loss": 0.5, "stage_align_loss": 0.004974365234375, "step": 3320 }, { "ce_loss": 0.1259765625, "distill_loss": 0.059814453125, "epoch": 1.452318460192476, "explain_loss": 1.203125, "loss": 1.71875, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.0076904296875, "step": 3320 }, { "ce_loss": 0.12353515625, "distill_loss": 0.103515625, "epoch": 1.452318460192476, "explain_loss": 1.0, "loss": 1.625, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0123291015625, "step": 3320 }, { "epoch": 1.4566929133858268, "grad_norm": 0.8958713412284851, "learning_rate": 0.00011305461123516279, "loss": 1.830322265625, "step": 3330 }, { "ce_loss": 0.0035247802734375, "distill_loss": 0.134765625, "epoch": 1.4566929133858268, "explain_loss": 0.9453125, "loss": 1.546875, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.00445556640625, "step": 3330 }, { "ce_loss": 0.05224609375, "distill_loss": 0.07861328125, "epoch": 1.4566929133858268, "explain_loss": 1.421875, "loss": 2.015625, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0032196044921875, "step": 3330 }, { "ce_loss": 0.014892578125, "distill_loss": 0.095703125, "epoch": 1.4566929133858268, "explain_loss": 0.8828125, "loss": 1.2265625, "ref_ce_loss": 0.2314453125, "stage_align_loss": 0.003021240234375, "step": 3330 }, { "ce_loss": 0.42578125, "distill_loss": 0.09326171875, "epoch": 1.4566929133858268, "explain_loss": 1.0, "loss": 2.03125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.04150390625, "step": 3330 }, { "epoch": 1.4610673665791776, "grad_norm": 2.1053948402404785, "learning_rate": 0.00011257637878616018, "loss": 1.85078125, "step": 3340 }, { "ce_loss": 0.01092529296875, "distill_loss": 0.09814453125, "epoch": 1.4610673665791776, "explain_loss": 1.09375, "loss": 1.59375, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.003692626953125, "step": 3340 }, { "ce_loss": 0.455078125, "distill_loss": 0.0732421875, "epoch": 1.4610673665791776, "explain_loss": 1.140625, "loss": 1.96875, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.01190185546875, "step": 3340 }, { "ce_loss": 0.09326171875, "distill_loss": 0.11767578125, "epoch": 1.4610673665791776, "explain_loss": 1.4140625, "loss": 1.9453125, "ref_ce_loss": 0.310546875, "stage_align_loss": 0.006195068359375, "step": 3340 }, { "ce_loss": 0.0150146484375, "distill_loss": 0.1328125, "epoch": 1.4610673665791776, "explain_loss": 1.375, "loss": 1.9140625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.00335693359375, "step": 3340 }, { "epoch": 1.4654418197725283, "grad_norm": 1.0743649005889893, "learning_rate": 0.0001120978539046333, "loss": 1.834765625, "step": 3350 }, { "ce_loss": 0.055419921875, "distill_loss": 0.051513671875, "epoch": 1.4654418197725283, "explain_loss": 0.72265625, "loss": 1.1328125, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.00628662109375, "step": 3350 }, { "ce_loss": 0.29296875, "distill_loss": 0.181640625, "epoch": 1.4654418197725283, "explain_loss": 0.8671875, "loss": 1.8359375, "ref_ce_loss": 0.484375, "stage_align_loss": 0.00732421875, "step": 3350 }, { "ce_loss": 0.02490234375, "distill_loss": 0.119140625, "epoch": 1.4654418197725283, "explain_loss": 1.078125, "loss": 1.8203125, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.044677734375, "step": 3350 }, { "ce_loss": 0.1435546875, "distill_loss": 0.1259765625, "epoch": 1.4654418197725283, "explain_loss": 0.953125, "loss": 1.6796875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.0174560546875, "step": 3350 }, { "epoch": 1.4698162729658792, "grad_norm": 0.9060342907905579, "learning_rate": 0.0001116190477174925, "loss": 1.93525390625, "step": 3360 }, { "ce_loss": 0.046875, "distill_loss": 0.07373046875, "epoch": 1.4698162729658792, "explain_loss": 1.0546875, "loss": 1.625, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.0035247802734375, "step": 3360 }, { "ce_loss": 0.59765625, "distill_loss": 0.1083984375, "epoch": 1.4698162729658792, "explain_loss": 1.0625, "loss": 2.296875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.10546875, "step": 3360 }, { "ce_loss": 0.171875, "distill_loss": 0.255859375, "epoch": 1.4698162729658792, "explain_loss": 0.7734375, "loss": 1.59375, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.03564453125, "step": 3360 }, { "ce_loss": 0.1123046875, "distill_loss": 0.072265625, "epoch": 1.4698162729658792, "explain_loss": 1.140625, "loss": 1.859375, "ref_ce_loss": 0.53125, "stage_align_loss": 0.009033203125, "step": 3360 }, { "epoch": 1.4741907261592302, "grad_norm": 0.6835344433784485, "learning_rate": 0.0001111399713581891, "loss": 1.9048828125, "step": 3370 }, { "ce_loss": 0.01513671875, "distill_loss": 0.126953125, "epoch": 1.4741907261592302, "explain_loss": 1.265625, "loss": 1.71875, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.0059814453125, "step": 3370 }, { "ce_loss": 0.076171875, "distill_loss": 0.1123046875, "epoch": 1.4741907261592302, "explain_loss": 0.97265625, "loss": 1.625, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.006256103515625, "step": 3370 }, { "ce_loss": 0.12060546875, "distill_loss": 0.04443359375, "epoch": 1.4741907261592302, "explain_loss": 0.984375, "loss": 1.5390625, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00726318359375, "step": 3370 }, { "ce_loss": 0.08203125, "distill_loss": 0.05126953125, "epoch": 1.4741907261592302, "explain_loss": 1.265625, "loss": 1.7734375, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.004913330078125, "step": 3370 }, { "epoch": 1.478565179352581, "grad_norm": 1.3698726892471313, "learning_rate": 0.00011066063596645677, "loss": 1.8400390625, "step": 3380 }, { "ce_loss": 0.291015625, "distill_loss": 0.1865234375, "epoch": 1.478565179352581, "explain_loss": 1.296875, "loss": 2.140625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.01007080078125, "step": 3380 }, { "ce_loss": 0.11474609375, "distill_loss": 0.1259765625, "epoch": 1.478565179352581, "explain_loss": 1.109375, "loss": 1.78125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.004974365234375, "step": 3380 }, { "ce_loss": 0.10546875, "distill_loss": 0.146484375, "epoch": 1.478565179352581, "explain_loss": 0.79296875, "loss": 1.4140625, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.013427734375, "step": 3380 }, { "ce_loss": 0.310546875, "distill_loss": 0.10498046875, "epoch": 1.478565179352581, "explain_loss": 0.8515625, "loss": 1.65625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.058837890625, "step": 3380 }, { "epoch": 1.4829396325459316, "grad_norm": 0.86570143699646, "learning_rate": 0.00011018105268805218, "loss": 1.800537109375, "step": 3390 }, { "ce_loss": 0.0142822265625, "distill_loss": 0.06201171875, "epoch": 1.4829396325459316, "explain_loss": 1.0078125, "loss": 1.4921875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.0035400390625, "step": 3390 }, { "ce_loss": 0.46484375, "distill_loss": 0.09326171875, "epoch": 1.4829396325459316, "explain_loss": 0.984375, "loss": 2.09375, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.0634765625, "step": 3390 }, { "ce_loss": 0.1806640625, "distill_loss": 0.0751953125, "epoch": 1.4829396325459316, "explain_loss": 1.2578125, "loss": 1.875, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.01116943359375, "step": 3390 }, { "ce_loss": 0.0020599365234375, "distill_loss": 0.06005859375, "epoch": 1.4829396325459316, "explain_loss": 0.9765625, "loss": 1.5625, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.005767822265625, "step": 3390 }, { "epoch": 1.4873140857392826, "grad_norm": 1.1995186805725098, "learning_rate": 0.0001097012326744961, "loss": 1.8693359375, "step": 3400 }, { "ce_loss": 0.00142669677734375, "distill_loss": 0.04541015625, "epoch": 1.4873140857392826, "explain_loss": 1.0, "loss": 1.4765625, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0029754638671875, "step": 3400 }, { "ce_loss": 1.3671875, "distill_loss": 0.0458984375, "epoch": 1.4873140857392826, "explain_loss": 1.046875, "loss": 3.078125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0732421875, "step": 3400 }, { "ce_loss": 0.166015625, "distill_loss": 0.0693359375, "epoch": 1.4873140857392826, "explain_loss": 1.1796875, "loss": 1.71875, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.005096435546875, "step": 3400 }, { "ce_loss": 0.2080078125, "distill_loss": 0.0869140625, "epoch": 1.4873140857392826, "explain_loss": 1.8203125, "loss": 2.625, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.006134033203125, "step": 3400 }, { "epoch": 1.4916885389326335, "grad_norm": 8.765689849853516, "learning_rate": 0.00010922118708281395, "loss": 1.89521484375, "step": 3410 }, { "ce_loss": 0.01214599609375, "distill_loss": 0.103515625, "epoch": 1.4916885389326335, "explain_loss": 1.015625, "loss": 1.53125, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0027618408203125, "step": 3410 }, { "ce_loss": 0.61328125, "distill_loss": 0.12451171875, "epoch": 1.4916885389326335, "explain_loss": 0.921875, "loss": 2.078125, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.05859375, "step": 3410 }, { "ce_loss": 0.1865234375, "distill_loss": 0.076171875, "epoch": 1.4916885389326335, "explain_loss": 1.28125, "loss": 2.203125, "ref_ce_loss": 0.65625, "stage_align_loss": 0.005279541015625, "step": 3410 }, { "ce_loss": 0.09130859375, "distill_loss": 0.203125, "epoch": 1.4916885389326335, "explain_loss": 0.79296875, "loss": 1.5625, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.06103515625, "step": 3410 }, { "epoch": 1.4960629921259843, "grad_norm": 2.630976438522339, "learning_rate": 0.00010874092707527636, "loss": 1.9166015625, "step": 3420 }, { "ce_loss": 0.007659912109375, "distill_loss": 0.080078125, "epoch": 1.4960629921259843, "explain_loss": 1.0234375, "loss": 1.609375, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0030364990234375, "step": 3420 }, { "ce_loss": 0.2138671875, "distill_loss": 0.115234375, "epoch": 1.4960629921259843, "explain_loss": 1.03125, "loss": 1.796875, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0205078125, "step": 3420 }, { "ce_loss": 0.90625, "distill_loss": 0.095703125, "epoch": 1.4960629921259843, "explain_loss": 1.203125, "loss": 2.765625, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.06884765625, "step": 3420 }, { "ce_loss": 0.11376953125, "distill_loss": 0.3359375, "epoch": 1.4960629921259843, "explain_loss": 1.265625, "loss": 2.0, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.004486083984375, "step": 3420 }, { "epoch": 1.500437445319335, "grad_norm": 1.7340924739837646, "learning_rate": 0.00010826046381913975, "loss": 1.9515625, "step": 3430 }, { "ce_loss": 0.08056640625, "distill_loss": 0.1171875, "epoch": 1.500437445319335, "explain_loss": 1.0703125, "loss": 1.6875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0032806396484375, "step": 3430 }, { "ce_loss": 0.4609375, "distill_loss": 0.06884765625, "epoch": 1.500437445319335, "explain_loss": 1.1796875, "loss": 2.3125, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.003082275390625, "step": 3430 }, { "ce_loss": 0.037841796875, "distill_loss": 0.1123046875, "epoch": 1.500437445319335, "explain_loss": 0.7734375, "loss": 1.171875, "ref_ce_loss": 0.248046875, "stage_align_loss": 0.0034027099609375, "step": 3430 }, { "ce_loss": 0.10498046875, "distill_loss": 0.080078125, "epoch": 1.500437445319335, "explain_loss": 0.75, "loss": 1.28125, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.00726318359375, "step": 3430 }, { "epoch": 1.504811898512686, "grad_norm": 1.1343210935592651, "learning_rate": 0.00010777980848638656, "loss": 1.84296875, "step": 3440 }, { "ce_loss": 0.014892578125, "distill_loss": 0.09033203125, "epoch": 1.504811898512686, "explain_loss": 1.015625, "loss": 1.484375, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.004425048828125, "step": 3440 }, { "ce_loss": 0.054931640625, "distill_loss": 0.0947265625, "epoch": 1.504811898512686, "explain_loss": 1.3515625, "loss": 1.9140625, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.003631591796875, "step": 3440 }, { "ce_loss": 0.00726318359375, "distill_loss": 0.08740234375, "epoch": 1.504811898512686, "explain_loss": 1.078125, "loss": 1.5, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0037689208984375, "step": 3440 }, { "ce_loss": 0.1025390625, "distill_loss": 0.0849609375, "epoch": 1.504811898512686, "explain_loss": 0.92578125, "loss": 1.578125, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.00921630859375, "step": 3440 }, { "epoch": 1.5091863517060369, "grad_norm": 0.9387580752372742, "learning_rate": 0.00010729897225346544, "loss": 1.7873046875, "step": 3450 }, { "ce_loss": 0.00872802734375, "distill_loss": 0.080078125, "epoch": 1.5091863517060369, "explain_loss": 1.25, "loss": 1.703125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00372314453125, "step": 3450 }, { "ce_loss": 0.004913330078125, "distill_loss": 0.1025390625, "epoch": 1.5091863517060369, "explain_loss": 1.21875, "loss": 1.65625, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.00360107421875, "step": 3450 }, { "ce_loss": 0.0966796875, "distill_loss": 0.09765625, "epoch": 1.5091863517060369, "explain_loss": 1.09375, "loss": 1.6875, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.006622314453125, "step": 3450 }, { "ce_loss": 0.24609375, "distill_loss": 0.099609375, "epoch": 1.5091863517060369, "explain_loss": 1.03125, "loss": 1.90625, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.019287109375, "step": 3450 }, { "epoch": 1.5135608048993876, "grad_norm": 0.7131326198577881, "learning_rate": 0.00010681796630103152, "loss": 1.8365234375, "step": 3460 }, { "ce_loss": 0.232421875, "distill_loss": 0.1572265625, "epoch": 1.5135608048993876, "explain_loss": 1.171875, "loss": 2.078125, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.004486083984375, "step": 3460 }, { "ce_loss": 0.119140625, "distill_loss": 0.1005859375, "epoch": 1.5135608048993876, "explain_loss": 0.921875, "loss": 1.53125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.00335693359375, "step": 3460 }, { "ce_loss": 0.009765625, "distill_loss": 0.07763671875, "epoch": 1.5135608048993876, "explain_loss": 1.1015625, "loss": 1.625, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.004669189453125, "step": 3460 }, { "ce_loss": 0.00732421875, "distill_loss": 0.083984375, "epoch": 1.5135608048993876, "explain_loss": 1.234375, "loss": 1.8046875, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.0035247802734375, "step": 3460 }, { "epoch": 1.5179352580927383, "grad_norm": 0.7531045079231262, "learning_rate": 0.00010633680181368621, "loss": 1.77490234375, "step": 3470 }, { "ce_loss": 0.1416015625, "distill_loss": 0.091796875, "epoch": 1.5179352580927383, "explain_loss": 0.8828125, "loss": 1.609375, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.004974365234375, "step": 3470 }, { "ce_loss": 0.1962890625, "distill_loss": 0.07373046875, "epoch": 1.5179352580927383, "explain_loss": 1.375, "loss": 2.140625, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0029449462890625, "step": 3470 }, { "ce_loss": 0.048583984375, "distill_loss": 0.08544921875, "epoch": 1.5179352580927383, "explain_loss": 1.1640625, "loss": 1.671875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0034637451171875, "step": 3470 }, { "ce_loss": 0.00970458984375, "distill_loss": 0.091796875, "epoch": 1.5179352580927383, "explain_loss": 1.1171875, "loss": 1.6875, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.0025177001953125, "step": 3470 }, { "epoch": 1.5223097112860893, "grad_norm": 1.0230947732925415, "learning_rate": 0.00010585548997971744, "loss": 1.73974609375, "step": 3480 }, { "ce_loss": 0.0076904296875, "distill_loss": 0.107421875, "epoch": 1.5223097112860893, "explain_loss": 1.28125, "loss": 1.875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.003173828125, "step": 3480 }, { "ce_loss": 0.0101318359375, "distill_loss": 0.0888671875, "epoch": 1.5223097112860893, "explain_loss": 1.3046875, "loss": 1.875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0030059814453125, "step": 3480 }, { "ce_loss": 0.208984375, "distill_loss": 0.091796875, "epoch": 1.5223097112860893, "explain_loss": 0.6640625, "loss": 1.3359375, "ref_ce_loss": 0.34375, "stage_align_loss": 0.026611328125, "step": 3480 }, { "ce_loss": 0.302734375, "distill_loss": 0.0732421875, "epoch": 1.5223097112860893, "explain_loss": 0.859375, "loss": 1.59375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0033111572265625, "step": 3480 }, { "epoch": 1.5266841644794402, "grad_norm": 0.7249631881713867, "learning_rate": 0.00010537404199083916, "loss": 1.820458984375, "step": 3490 }, { "ce_loss": 0.09228515625, "distill_loss": 0.0869140625, "epoch": 1.5266841644794402, "explain_loss": 1.09375, "loss": 1.640625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.0030670166015625, "step": 3490 }, { "ce_loss": 0.01068115234375, "distill_loss": 0.09326171875, "epoch": 1.5266841644794402, "explain_loss": 1.203125, "loss": 1.828125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0025482177734375, "step": 3490 }, { "ce_loss": 0.404296875, "distill_loss": 0.107421875, "epoch": 1.5266841644794402, "explain_loss": 1.2265625, "loss": 2.359375, "ref_ce_loss": 0.5859375, "stage_align_loss": 0.035400390625, "step": 3490 }, { "ce_loss": 0.71484375, "distill_loss": 0.09521484375, "epoch": 1.5266841644794402, "explain_loss": 1.078125, "loss": 2.3125, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.047607421875, "step": 3490 }, { "epoch": 1.531058617672791, "grad_norm": 0.8871873021125793, "learning_rate": 0.00010489246904193127, "loss": 1.861328125, "step": 3500 }, { "ce_loss": 0.138671875, "distill_loss": 0.0810546875, "epoch": 1.531058617672791, "explain_loss": 1.0, "loss": 1.75, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.00396728515625, "step": 3500 }, { "ce_loss": 0.06591796875, "distill_loss": 0.07421875, "epoch": 1.531058617672791, "explain_loss": 0.984375, "loss": 1.484375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0128173828125, "step": 3500 }, { "ce_loss": 0.37109375, "distill_loss": 0.11767578125, "epoch": 1.531058617672791, "explain_loss": 1.2109375, "loss": 2.15625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.0162353515625, "step": 3500 }, { "ce_loss": 0.003814697265625, "distill_loss": 0.0703125, "epoch": 1.531058617672791, "explain_loss": 1.1875, "loss": 1.640625, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.002655029296875, "step": 3500 }, { "epoch": 1.5354330708661417, "grad_norm": 1.4642235040664673, "learning_rate": 0.00010441078233077943, "loss": 1.83095703125, "step": 3510 }, { "ce_loss": 0.04736328125, "distill_loss": 0.07763671875, "epoch": 1.5354330708661417, "explain_loss": 0.96484375, "loss": 1.4453125, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0020751953125, "step": 3510 }, { "ce_loss": 0.37890625, "distill_loss": 0.10498046875, "epoch": 1.5354330708661417, "explain_loss": 1.390625, "loss": 2.375, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.0045166015625, "step": 3510 }, { "ce_loss": 0.353515625, "distill_loss": 0.10693359375, "epoch": 1.5354330708661417, "explain_loss": 0.859375, "loss": 1.703125, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.0213623046875, "step": 3510 }, { "ce_loss": 0.640625, "distill_loss": 0.091796875, "epoch": 1.5354330708661417, "explain_loss": 0.8359375, "loss": 2.0625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0031585693359375, "step": 3510 }, { "epoch": 1.5398075240594926, "grad_norm": 0.7872414588928223, "learning_rate": 0.00010392899305781443, "loss": 1.7849609375, "step": 3520 }, { "ce_loss": 0.283203125, "distill_loss": 0.1162109375, "epoch": 1.5398075240594926, "explain_loss": 1.484375, "loss": 2.1875, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.00970458984375, "step": 3520 }, { "ce_loss": 0.166015625, "distill_loss": 0.072265625, "epoch": 1.5398075240594926, "explain_loss": 0.72265625, "loss": 1.3515625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.00457763671875, "step": 3520 }, { "ce_loss": 0.1767578125, "distill_loss": 0.10986328125, "epoch": 1.5398075240594926, "explain_loss": 1.0078125, "loss": 1.6796875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0078125, "step": 3520 }, { "ce_loss": 0.5078125, "distill_loss": 0.107421875, "epoch": 1.5398075240594926, "explain_loss": 0.94921875, "loss": 1.90625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.033935546875, "step": 3520 }, { "epoch": 1.5441819772528433, "grad_norm": 0.7630012631416321, "learning_rate": 0.000103447112425852, "loss": 1.793994140625, "step": 3530 }, { "ce_loss": 0.9140625, "distill_loss": 0.10546875, "epoch": 1.5441819772528433, "explain_loss": 0.83984375, "loss": 2.25, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.00848388671875, "step": 3530 }, { "ce_loss": 0.1064453125, "distill_loss": 0.0966796875, "epoch": 1.5441819772528433, "explain_loss": 0.8515625, "loss": 1.5, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.00396728515625, "step": 3530 }, { "ce_loss": 0.33984375, "distill_loss": 0.07666015625, "epoch": 1.5441819772528433, "explain_loss": 0.9375, "loss": 1.7109375, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.007080078125, "step": 3530 }, { "ce_loss": 0.6484375, "distill_loss": 0.0859375, "epoch": 1.5441819772528433, "explain_loss": 1.046875, "loss": 2.25, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.037109375, "step": 3530 }, { "epoch": 1.5485564304461943, "grad_norm": 1.2666566371917725, "learning_rate": 0.00010296515163983203, "loss": 1.8634765625, "step": 3540 }, { "ce_loss": 0.390625, "distill_loss": 0.10986328125, "epoch": 1.5485564304461943, "explain_loss": 1.203125, "loss": 2.09375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.00823974609375, "step": 3540 }, { "ce_loss": 0.057373046875, "distill_loss": 0.07861328125, "epoch": 1.5485564304461943, "explain_loss": 0.828125, "loss": 1.3203125, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.004974365234375, "step": 3540 }, { "ce_loss": 0.024169921875, "distill_loss": 0.11669921875, "epoch": 1.5485564304461943, "explain_loss": 1.0078125, "loss": 1.515625, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0029449462890625, "step": 3540 }, { "ce_loss": 0.00909423828125, "distill_loss": 0.103515625, "epoch": 1.5485564304461943, "explain_loss": 0.96875, "loss": 1.375, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.00194549560546875, "step": 3540 }, { "epoch": 1.552930883639545, "grad_norm": 1.7435907125473022, "learning_rate": 0.00010248312190655838, "loss": 3.084375, "step": 3550 }, { "ce_loss": 1.3515625, "distill_loss": 0.09375, "epoch": 1.552930883639545, "explain_loss": 0.875, "loss": 2.734375, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.030517578125, "step": 3550 }, { "ce_loss": 0.6015625, "distill_loss": 0.099609375, "epoch": 1.552930883639545, "explain_loss": 1.4296875, "loss": 2.5625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0228271484375, "step": 3550 }, { "ce_loss": 0.5625, "distill_loss": 0.09228515625, "epoch": 1.552930883639545, "explain_loss": 1.2734375, "loss": 2.421875, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.00811767578125, "step": 3550 }, { "ce_loss": 0.26171875, "distill_loss": 0.1767578125, "epoch": 1.552930883639545, "explain_loss": 1.328125, "loss": 2.21875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.0030517578125, "step": 3550 }, { "epoch": 1.5573053368328957, "grad_norm": 4.881060600280762, "learning_rate": 0.00010200103443443801, "loss": 2.18525390625, "step": 3560 }, { "ce_loss": 0.7890625, "distill_loss": 0.1240234375, "epoch": 1.5573053368328957, "explain_loss": 1.203125, "loss": 2.625, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0252685546875, "step": 3560 }, { "ce_loss": 0.67578125, "distill_loss": 0.1474609375, "epoch": 1.5573053368328957, "explain_loss": 1.046875, "loss": 2.28125, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.038330078125, "step": 3560 }, { "ce_loss": 0.0908203125, "distill_loss": 0.0849609375, "epoch": 1.5573053368328957, "explain_loss": 0.82421875, "loss": 1.4375, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.00433349609375, "step": 3560 }, { "ce_loss": 0.0244140625, "distill_loss": 0.1162109375, "epoch": 1.5573053368328957, "explain_loss": 1.1484375, "loss": 1.625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.003204345703125, "step": 3560 }, { "epoch": 1.5616797900262467, "grad_norm": 1.5080201625823975, "learning_rate": 0.0001015189004332204, "loss": 1.86513671875, "step": 3570 }, { "ce_loss": 0.150390625, "distill_loss": 0.134765625, "epoch": 1.5616797900262467, "explain_loss": 1.296875, "loss": 1.96875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.007537841796875, "step": 3570 }, { "ce_loss": 0.01141357421875, "distill_loss": 0.16015625, "epoch": 1.5616797900262467, "explain_loss": 1.25, "loss": 1.859375, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.003265380859375, "step": 3570 }, { "ce_loss": 0.055419921875, "distill_loss": 0.138671875, "epoch": 1.5616797900262467, "explain_loss": 1.1875, "loss": 1.9140625, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.005157470703125, "step": 3570 }, { "ce_loss": 0.00174713134765625, "distill_loss": 0.1005859375, "epoch": 1.5616797900262467, "explain_loss": 1.09375, "loss": 1.5078125, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.00311279296875, "step": 3570 }, { "epoch": 1.5660542432195976, "grad_norm": 1.0758044719696045, "learning_rate": 0.00010103673111373712, "loss": 2.313671875, "step": 3580 }, { "ce_loss": 0.00738525390625, "distill_loss": 0.111328125, "epoch": 1.5660542432195976, "explain_loss": 1.0859375, "loss": 1.75, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.003021240234375, "step": 3580 }, { "ce_loss": 0.201171875, "distill_loss": 0.12890625, "epoch": 1.5660542432195976, "explain_loss": 1.109375, "loss": 1.75, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.0262451171875, "step": 3580 }, { "ce_loss": 0.005157470703125, "distill_loss": 0.103515625, "epoch": 1.5660542432195976, "explain_loss": 1.25, "loss": 1.796875, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0026397705078125, "step": 3580 }, { "ce_loss": 0.03125, "distill_loss": 0.130859375, "epoch": 1.5660542432195976, "explain_loss": 0.890625, "loss": 1.515625, "ref_ce_loss": 0.453125, "stage_align_loss": 0.00262451171875, "step": 3580 }, { "epoch": 1.5704286964129484, "grad_norm": 0.728141188621521, "learning_rate": 0.00010055453768764083, "loss": 1.8171875, "step": 3590 }, { "ce_loss": 0.00665283203125, "distill_loss": 0.1162109375, "epoch": 1.5704286964129484, "explain_loss": 1.078125, "loss": 1.625, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0028228759765625, "step": 3590 }, { "ce_loss": 0.006683349609375, "distill_loss": 0.08740234375, "epoch": 1.5704286964129484, "explain_loss": 1.0625, "loss": 1.59375, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.002960205078125, "step": 3590 }, { "ce_loss": 0.01116943359375, "distill_loss": 0.09716796875, "epoch": 1.5704286964129484, "explain_loss": 0.92578125, "loss": 1.46875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.002288818359375, "step": 3590 }, { "ce_loss": 0.76953125, "distill_loss": 0.091796875, "epoch": 1.5704286964129484, "explain_loss": 1.0703125, "loss": 2.375, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.028076171875, "step": 3590 }, { "epoch": 1.574803149606299, "grad_norm": 0.735720157623291, "learning_rate": 0.00010007233136714472, "loss": 1.85009765625, "step": 3600 }, { "ce_loss": 0.0546875, "distill_loss": 0.12890625, "epoch": 1.574803149606299, "explain_loss": 0.83984375, "loss": 1.5234375, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.026611328125, "step": 3600 }, { "ce_loss": 0.0020599365234375, "distill_loss": 0.109375, "epoch": 1.574803149606299, "explain_loss": 1.234375, "loss": 1.78125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0028076171875, "step": 3600 }, { "ce_loss": 0.018310546875, "distill_loss": 0.123046875, "epoch": 1.574803149606299, "explain_loss": 0.9921875, "loss": 1.6484375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0022125244140625, "step": 3600 }, { "ce_loss": 0.1689453125, "distill_loss": 0.1416015625, "epoch": 1.574803149606299, "explain_loss": 1.265625, "loss": 1.921875, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.0205078125, "step": 3600 }, { "epoch": 1.57917760279965, "grad_norm": 1.7681901454925537, "learning_rate": 9.959012336476193e-05, "loss": 1.78779296875, "step": 3610 }, { "ce_loss": 0.002899169921875, "distill_loss": 0.111328125, "epoch": 1.57917760279965, "explain_loss": 1.4453125, "loss": 1.90625, "ref_ce_loss": 0.34375, "stage_align_loss": 0.00225830078125, "step": 3610 }, { "ce_loss": 0.039306640625, "distill_loss": 0.1005859375, "epoch": 1.57917760279965, "explain_loss": 0.96484375, "loss": 1.390625, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.003387451171875, "step": 3610 }, { "ce_loss": 0.310546875, "distill_loss": 0.08056640625, "epoch": 1.57917760279965, "explain_loss": 1.2265625, "loss": 2.046875, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0203857421875, "step": 3610 }, { "ce_loss": 0.08251953125, "distill_loss": 0.0966796875, "epoch": 1.57917760279965, "explain_loss": 1.515625, "loss": 2.125, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.003997802734375, "step": 3610 }, { "epoch": 1.583552055993001, "grad_norm": 1.4025565385818481, "learning_rate": 9.910792489304458e-05, "loss": 1.82900390625, "step": 3620 }, { "ce_loss": 0.45703125, "distill_loss": 0.119140625, "epoch": 1.583552055993001, "explain_loss": 1.2890625, "loss": 2.296875, "ref_ce_loss": 0.375, "stage_align_loss": 0.058349609375, "step": 3620 }, { "ce_loss": 0.091796875, "distill_loss": 0.10693359375, "epoch": 1.583552055993001, "explain_loss": 1.109375, "loss": 1.734375, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.006927490234375, "step": 3620 }, { "ce_loss": 0.7109375, "distill_loss": 0.08544921875, "epoch": 1.583552055993001, "explain_loss": 1.15625, "loss": 2.46875, "ref_ce_loss": 0.4375, "stage_align_loss": 0.08349609375, "step": 3620 }, { "ce_loss": 0.043701171875, "distill_loss": 0.1103515625, "epoch": 1.583552055993001, "explain_loss": 0.7890625, "loss": 1.3125, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0033721923828125, "step": 3620 }, { "epoch": 1.5879265091863517, "grad_norm": 1.001179575920105, "learning_rate": 9.862574716432333e-05, "loss": 1.805859375, "step": 3630 }, { "ce_loss": 0.478515625, "distill_loss": 0.1015625, "epoch": 1.5879265091863517, "explain_loss": 1.0546875, "loss": 1.9609375, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.061279296875, "step": 3630 }, { "ce_loss": 0.0179443359375, "distill_loss": 0.087890625, "epoch": 1.5879265091863517, "explain_loss": 1.265625, "loss": 1.6328125, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.0026397705078125, "step": 3630 }, { "ce_loss": 0.25390625, "distill_loss": 0.12890625, "epoch": 1.5879265091863517, "explain_loss": 0.96875, "loss": 1.84375, "ref_ce_loss": 0.46875, "stage_align_loss": 0.02783203125, "step": 3630 }, { "ce_loss": 0.01336669921875, "distill_loss": 0.08544921875, "epoch": 1.5879265091863517, "explain_loss": 0.90625, "loss": 1.328125, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.00323486328125, "step": 3630 }, { "epoch": 1.5923009623797024, "grad_norm": 0.7788344025611877, "learning_rate": 9.814360139044631e-05, "loss": 1.82265625, "step": 3640 }, { "ce_loss": 0.0216064453125, "distill_loss": 0.1123046875, "epoch": 1.5923009623797024, "explain_loss": 0.86328125, "loss": 1.359375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.004547119140625, "step": 3640 }, { "ce_loss": 0.232421875, "distill_loss": 0.087890625, "epoch": 1.5923009623797024, "explain_loss": 1.2734375, "loss": 2.0625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.031005859375, "step": 3640 }, { "ce_loss": 0.09912109375, "distill_loss": 0.095703125, "epoch": 1.5923009623797024, "explain_loss": 0.75, "loss": 1.3125, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.0284423828125, "step": 3640 }, { "ce_loss": 0.01806640625, "distill_loss": 0.154296875, "epoch": 1.5923009623797024, "explain_loss": 0.984375, "loss": 1.671875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0033416748046875, "step": 3640 }, { "epoch": 1.5966754155730534, "grad_norm": 1.0144033432006836, "learning_rate": 9.766149878251881e-05, "loss": 1.883984375, "step": 3650 }, { "ce_loss": 0.2080078125, "distill_loss": 0.09716796875, "epoch": 1.5966754155730534, "explain_loss": 0.859375, "loss": 1.515625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.01092529296875, "step": 3650 }, { "ce_loss": 0.031982421875, "distill_loss": 0.0888671875, "epoch": 1.5966754155730534, "explain_loss": 1.21875, "loss": 1.640625, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.002655029296875, "step": 3650 }, { "ce_loss": 0.10595703125, "distill_loss": 0.1171875, "epoch": 1.5966754155730534, "explain_loss": 1.140625, "loss": 2.0, "ref_ce_loss": 0.6171875, "stage_align_loss": 0.014404296875, "step": 3650 }, { "ce_loss": 0.006072998046875, "distill_loss": 0.158203125, "epoch": 1.5966754155730534, "explain_loss": 1.0234375, "loss": 1.625, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.002166748046875, "step": 3650 }, { "epoch": 1.6010498687664043, "grad_norm": 1.3602244853973389, "learning_rate": 9.717945055064227e-05, "loss": 1.7853515625, "step": 3660 }, { "ce_loss": 0.01129150390625, "distill_loss": 0.103515625, "epoch": 1.6010498687664043, "explain_loss": 1.234375, "loss": 1.8125, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0026702880859375, "step": 3660 }, { "ce_loss": 0.54296875, "distill_loss": 0.138671875, "epoch": 1.6010498687664043, "explain_loss": 1.3125, "loss": 2.453125, "ref_ce_loss": 0.455078125, "stage_align_loss": 0.00421142578125, "step": 3660 }, { "ce_loss": 0.002044677734375, "distill_loss": 0.138671875, "epoch": 1.6010498687664043, "explain_loss": 1.25, "loss": 1.8125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.002471923828125, "step": 3660 }, { "ce_loss": 0.034423828125, "distill_loss": 0.103515625, "epoch": 1.6010498687664043, "explain_loss": 0.890625, "loss": 1.4296875, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.00311279296875, "step": 3660 }, { "epoch": 1.605424321959755, "grad_norm": 1.354155421257019, "learning_rate": 9.669746790365385e-05, "loss": 1.80478515625, "step": 3670 }, { "ce_loss": 0.1533203125, "distill_loss": 0.138671875, "epoch": 1.605424321959755, "explain_loss": 1.2578125, "loss": 1.921875, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.003265380859375, "step": 3670 }, { "ce_loss": 0.0186767578125, "distill_loss": 0.1259765625, "epoch": 1.605424321959755, "explain_loss": 1.1953125, "loss": 1.890625, "ref_ce_loss": 0.546875, "stage_align_loss": 0.0034942626953125, "step": 3670 }, { "ce_loss": 0.0888671875, "distill_loss": 0.1611328125, "epoch": 1.605424321959755, "explain_loss": 0.87890625, "loss": 1.5390625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.00445556640625, "step": 3670 }, { "ce_loss": 0.11083984375, "distill_loss": 0.1044921875, "epoch": 1.605424321959755, "explain_loss": 1.0078125, "loss": 1.6484375, "ref_ce_loss": 0.421875, "stage_align_loss": 0.003173828125, "step": 3670 }, { "epoch": 1.6097987751531058, "grad_norm": 1.068560004234314, "learning_rate": 9.621556204886559e-05, "loss": 1.8646484375, "step": 3680 }, { "ce_loss": 0.07470703125, "distill_loss": 0.107421875, "epoch": 1.6097987751531058, "explain_loss": 0.765625, "loss": 1.515625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.033203125, "step": 3680 }, { "ce_loss": 0.052734375, "distill_loss": 0.10791015625, "epoch": 1.6097987751531058, "explain_loss": 0.99609375, "loss": 1.6015625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.0035858154296875, "step": 3680 }, { "ce_loss": 0.072265625, "distill_loss": 0.19140625, "epoch": 1.6097987751531058, "explain_loss": 1.328125, "loss": 2.3125, "ref_ce_loss": 0.71484375, "stage_align_loss": 0.0032196044921875, "step": 3680 }, { "ce_loss": 0.7734375, "distill_loss": 0.0927734375, "epoch": 1.6097987751531058, "explain_loss": 1.1875, "loss": 2.4375, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.031982421875, "step": 3680 }, { "epoch": 1.6141732283464567, "grad_norm": 1.0732598304748535, "learning_rate": 9.573374419180403e-05, "loss": 1.806640625, "step": 3690 }, { "ce_loss": 0.0439453125, "distill_loss": 0.1767578125, "epoch": 1.6141732283464567, "explain_loss": 0.859375, "loss": 1.4296875, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.00457763671875, "step": 3690 }, { "ce_loss": 0.1767578125, "distill_loss": 0.1220703125, "epoch": 1.6141732283464567, "explain_loss": 1.015625, "loss": 1.84375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.00537109375, "step": 3690 }, { "ce_loss": 0.0281982421875, "distill_loss": 0.0966796875, "epoch": 1.6141732283464567, "explain_loss": 1.046875, "loss": 1.5, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.0030517578125, "step": 3690 }, { "ce_loss": 0.033203125, "distill_loss": 0.0732421875, "epoch": 1.6141732283464567, "explain_loss": 1.40625, "loss": 1.9921875, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.01007080078125, "step": 3690 }, { "epoch": 1.6185476815398077, "grad_norm": 4.846541404724121, "learning_rate": 9.525202553594946e-05, "loss": 1.77666015625, "step": 3700 }, { "ce_loss": 0.0252685546875, "distill_loss": 0.10205078125, "epoch": 1.6185476815398077, "explain_loss": 0.9765625, "loss": 1.5078125, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.0068359375, "step": 3700 }, { "ce_loss": 0.035400390625, "distill_loss": 0.15625, "epoch": 1.6185476815398077, "explain_loss": 0.9453125, "loss": 1.546875, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.0029144287109375, "step": 3700 }, { "ce_loss": 0.10302734375, "distill_loss": 0.1669921875, "epoch": 1.6185476815398077, "explain_loss": 0.8671875, "loss": 1.59375, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.003265380859375, "step": 3700 }, { "ce_loss": 0.051513671875, "distill_loss": 0.1708984375, "epoch": 1.6185476815398077, "explain_loss": 0.9453125, "loss": 1.6796875, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.003173828125, "step": 3700 }, { "epoch": 1.6229221347331584, "grad_norm": 0.6091433167457581, "learning_rate": 9.477041728247552e-05, "loss": 1.73779296875, "step": 3710 }, { "ce_loss": 0.81640625, "distill_loss": 0.1279296875, "epoch": 1.6229221347331584, "explain_loss": 1.1875, "loss": 2.75, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.004058837890625, "step": 3710 }, { "ce_loss": 0.00958251953125, "distill_loss": 0.0849609375, "epoch": 1.6229221347331584, "explain_loss": 1.59375, "loss": 2.1875, "ref_ce_loss": 0.490234375, "stage_align_loss": 0.00445556640625, "step": 3710 }, { "ce_loss": 0.1064453125, "distill_loss": 0.0908203125, "epoch": 1.6229221347331584, "explain_loss": 1.0546875, "loss": 1.6875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0269775390625, "step": 3710 }, { "ce_loss": 0.2578125, "distill_loss": 0.076171875, "epoch": 1.6229221347331584, "explain_loss": 1.078125, "loss": 2.28125, "ref_ce_loss": 0.8125, "stage_align_loss": 0.052978515625, "step": 3710 }, { "epoch": 1.627296587926509, "grad_norm": 0.8809022903442383, "learning_rate": 9.42889306299887e-05, "loss": 1.78076171875, "step": 3720 }, { "ce_loss": 0.380859375, "distill_loss": 0.10791015625, "epoch": 1.627296587926509, "explain_loss": 1.171875, "loss": 2.15625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.01007080078125, "step": 3720 }, { "ce_loss": 0.349609375, "distill_loss": 0.0693359375, "epoch": 1.627296587926509, "explain_loss": 0.98046875, "loss": 1.6875, "ref_ce_loss": 0.251953125, "stage_align_loss": 0.037841796875, "step": 3720 }, { "ce_loss": 0.0120849609375, "distill_loss": 0.08251953125, "epoch": 1.627296587926509, "explain_loss": 1.65625, "loss": 2.109375, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.002960205078125, "step": 3720 }, { "ce_loss": 0.05908203125, "distill_loss": 0.08203125, "epoch": 1.627296587926509, "explain_loss": 0.875, "loss": 1.3984375, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00567626953125, "step": 3720 }, { "epoch": 1.63167104111986, "grad_norm": 0.8581141829490662, "learning_rate": 9.380757677426804e-05, "loss": 1.80537109375, "step": 3730 }, { "ce_loss": 0.2890625, "distill_loss": 0.11181640625, "epoch": 1.63167104111986, "explain_loss": 1.375, "loss": 2.25, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.0301513671875, "step": 3730 }, { "ce_loss": 0.1162109375, "distill_loss": 0.0986328125, "epoch": 1.63167104111986, "explain_loss": 0.9921875, "loss": 1.71875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0439453125, "step": 3730 }, { "ce_loss": 0.10888671875, "distill_loss": 0.111328125, "epoch": 1.63167104111986, "explain_loss": 1.46875, "loss": 2.0625, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.01904296875, "step": 3730 }, { "ce_loss": 0.044921875, "distill_loss": 0.06884765625, "epoch": 1.63167104111986, "explain_loss": 0.86328125, "loss": 1.34375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0084228515625, "step": 3730 }, { "epoch": 1.636045494313211, "grad_norm": 0.8698899149894714, "learning_rate": 9.332636690800453e-05, "loss": 1.79326171875, "step": 3740 }, { "ce_loss": 0.8828125, "distill_loss": 0.08984375, "epoch": 1.636045494313211, "explain_loss": 1.0625, "loss": 2.546875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0615234375, "step": 3740 }, { "ce_loss": 0.0223388671875, "distill_loss": 0.076171875, "epoch": 1.636045494313211, "explain_loss": 1.375, "loss": 1.734375, "ref_ce_loss": 0.2412109375, "stage_align_loss": 0.0238037109375, "step": 3740 }, { "ce_loss": 0.9140625, "distill_loss": 0.142578125, "epoch": 1.636045494313211, "explain_loss": 1.390625, "loss": 2.90625, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.056884765625, "step": 3740 }, { "ce_loss": 0.34375, "distill_loss": 0.1171875, "epoch": 1.636045494313211, "explain_loss": 0.7578125, "loss": 1.828125, "ref_ce_loss": 0.59375, "stage_align_loss": 0.0181884765625, "step": 3740 }, { "epoch": 1.6404199475065617, "grad_norm": 0.51591956615448, "learning_rate": 9.28453122205413e-05, "loss": 1.71591796875, "step": 3750 }, { "ce_loss": 0.11767578125, "distill_loss": 0.080078125, "epoch": 1.6404199475065617, "explain_loss": 1.1875, "loss": 1.890625, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.004638671875, "step": 3750 }, { "ce_loss": 0.203125, "distill_loss": 0.12255859375, "epoch": 1.6404199475065617, "explain_loss": 0.828125, "loss": 1.4375, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.004119873046875, "step": 3750 }, { "ce_loss": 0.0498046875, "distill_loss": 0.06787109375, "epoch": 1.6404199475065617, "explain_loss": 1.375, "loss": 1.8203125, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.004150390625, "step": 3750 }, { "ce_loss": 0.5234375, "distill_loss": 0.1435546875, "epoch": 1.6404199475065617, "explain_loss": 1.1171875, "loss": 2.4375, "ref_ce_loss": 0.58203125, "stage_align_loss": 0.06103515625, "step": 3750 }, { "epoch": 1.6447944006999125, "grad_norm": 0.8214882016181946, "learning_rate": 9.236442389761296e-05, "loss": 1.8267578125, "step": 3760 }, { "ce_loss": 0.0927734375, "distill_loss": 0.0654296875, "epoch": 1.6447944006999125, "explain_loss": 0.96484375, "loss": 1.46875, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.017578125, "step": 3760 }, { "ce_loss": 0.04638671875, "distill_loss": 0.11181640625, "epoch": 1.6447944006999125, "explain_loss": 0.83203125, "loss": 1.328125, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0030059814453125, "step": 3760 }, { "ce_loss": 1.28125, "distill_loss": 0.09814453125, "epoch": 1.6447944006999125, "explain_loss": 1.203125, "loss": 3.03125, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.06689453125, "step": 3760 }, { "ce_loss": 0.01019287109375, "distill_loss": 0.0966796875, "epoch": 1.6447944006999125, "explain_loss": 1.296875, "loss": 1.8125, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.0019989013671875, "step": 3760 }, { "epoch": 1.6491688538932632, "grad_norm": 0.498764306306839, "learning_rate": 9.188371312108582e-05, "loss": 1.79912109375, "step": 3770 }, { "ce_loss": 0.75390625, "distill_loss": 0.0908203125, "epoch": 1.6491688538932632, "explain_loss": 0.75, "loss": 2.296875, "ref_ce_loss": 0.65625, "stage_align_loss": 0.04736328125, "step": 3770 }, { "ce_loss": 0.1123046875, "distill_loss": 0.0751953125, "epoch": 1.6491688538932632, "explain_loss": 1.28125, "loss": 1.8203125, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.003143310546875, "step": 3770 }, { "ce_loss": 0.012451171875, "distill_loss": 0.10546875, "epoch": 1.6491688538932632, "explain_loss": 1.03125, "loss": 1.5546875, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.004425048828125, "step": 3770 }, { "ce_loss": 0.498046875, "distill_loss": 0.07080078125, "epoch": 1.6491688538932632, "explain_loss": 1.203125, "loss": 2.28125, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0400390625, "step": 3770 }, { "epoch": 1.6535433070866141, "grad_norm": 0.8709380030632019, "learning_rate": 9.140319106869787e-05, "loss": 1.83662109375, "step": 3780 }, { "ce_loss": 0.10009765625, "distill_loss": 0.07177734375, "epoch": 1.6535433070866141, "explain_loss": 1.1796875, "loss": 1.84375, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.005035400390625, "step": 3780 }, { "ce_loss": 0.00872802734375, "distill_loss": 0.0751953125, "epoch": 1.6535433070866141, "explain_loss": 1.4921875, "loss": 2.0, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0033721923828125, "step": 3780 }, { "ce_loss": 0.310546875, "distill_loss": 0.0927734375, "epoch": 1.6535433070866141, "explain_loss": 0.921875, "loss": 1.796875, "ref_ce_loss": 0.46875, "stage_align_loss": 0.00738525390625, "step": 3780 }, { "ce_loss": 0.0712890625, "distill_loss": 0.0732421875, "epoch": 1.6535433070866141, "explain_loss": 0.9375, "loss": 1.546875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.059326171875, "step": 3780 }, { "epoch": 1.657917760279965, "grad_norm": 0.8110426664352417, "learning_rate": 9.09228689137986e-05, "loss": 1.76396484375, "step": 3790 }, { "ce_loss": 0.005615234375, "distill_loss": 0.053955078125, "epoch": 1.657917760279965, "explain_loss": 1.015625, "loss": 1.5625, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.003173828125, "step": 3790 }, { "ce_loss": 0.1064453125, "distill_loss": 0.09375, "epoch": 1.657917760279965, "explain_loss": 0.94140625, "loss": 1.546875, "ref_ce_loss": 0.375, "stage_align_loss": 0.037109375, "step": 3790 }, { "ce_loss": 0.17578125, "distill_loss": 0.091796875, "epoch": 1.657917760279965, "explain_loss": 0.890625, "loss": 1.546875, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.0034027099609375, "step": 3790 }, { "ce_loss": 0.0306396484375, "distill_loss": 0.0751953125, "epoch": 1.657917760279965, "explain_loss": 1.5078125, "loss": 2.1875, "ref_ce_loss": 0.578125, "stage_align_loss": 0.0037078857421875, "step": 3790 }, { "epoch": 1.6622922134733158, "grad_norm": 0.9564551711082458, "learning_rate": 9.044275782508952e-05, "loss": 1.709033203125, "step": 3800 }, { "ce_loss": 0.016357421875, "distill_loss": 0.07861328125, "epoch": 1.6622922134733158, "explain_loss": 1.203125, "loss": 1.6640625, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0024261474609375, "step": 3800 }, { "ce_loss": 0.52734375, "distill_loss": 0.08837890625, "epoch": 1.6622922134733158, "explain_loss": 1.1328125, "loss": 2.234375, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.07275390625, "step": 3800 }, { "ce_loss": 0.1875, "distill_loss": 0.111328125, "epoch": 1.6622922134733158, "explain_loss": 1.1328125, "loss": 1.796875, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.00262451171875, "step": 3800 }, { "ce_loss": 0.1376953125, "distill_loss": 0.09228515625, "epoch": 1.6622922134733158, "explain_loss": 1.2265625, "loss": 1.7734375, "ref_ce_loss": 0.3125, "stage_align_loss": 0.003326416015625, "step": 3800 }, { "epoch": 1.6666666666666665, "grad_norm": 0.8560087084770203, "learning_rate": 8.996286896636425e-05, "loss": 1.801953125, "step": 3810 }, { "ce_loss": 0.09716796875, "distill_loss": 0.0703125, "epoch": 1.6666666666666665, "explain_loss": 0.984375, "loss": 1.75, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.03173828125, "step": 3810 }, { "ce_loss": 0.052001953125, "distill_loss": 0.0693359375, "epoch": 1.6666666666666665, "explain_loss": 1.359375, "loss": 1.9453125, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.006256103515625, "step": 3810 }, { "ce_loss": 0.0208740234375, "distill_loss": 0.080078125, "epoch": 1.6666666666666665, "explain_loss": 1.1171875, "loss": 1.5859375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0023345947265625, "step": 3810 }, { "ce_loss": 0.026123046875, "distill_loss": 0.06884765625, "epoch": 1.6666666666666665, "explain_loss": 1.1484375, "loss": 1.6796875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0023651123046875, "step": 3810 }, { "epoch": 1.6710411198600175, "grad_norm": 0.5842794179916382, "learning_rate": 8.948321349624909e-05, "loss": 1.76328125, "step": 3820 }, { "ce_loss": 0.0234375, "distill_loss": 0.080078125, "epoch": 1.6710411198600175, "explain_loss": 0.66796875, "loss": 1.1328125, "ref_ce_loss": 0.359375, "stage_align_loss": 0.002655029296875, "step": 3820 }, { "ce_loss": 0.007080078125, "distill_loss": 0.07666015625, "epoch": 1.6710411198600175, "explain_loss": 0.85546875, "loss": 1.5703125, "ref_ce_loss": 0.625, "stage_align_loss": 0.004119873046875, "step": 3820 }, { "ce_loss": 0.51953125, "distill_loss": 0.0537109375, "epoch": 1.6710411198600175, "explain_loss": 1.734375, "loss": 2.6875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00860595703125, "step": 3820 }, { "ce_loss": 0.2373046875, "distill_loss": 0.091796875, "epoch": 1.6710411198600175, "explain_loss": 0.9921875, "loss": 1.7421875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.0038604736328125, "step": 3820 }, { "epoch": 1.6754155730533684, "grad_norm": 0.7635750770568848, "learning_rate": 8.900380256794334e-05, "loss": 1.73349609375, "step": 3830 }, { "ce_loss": 0.0184326171875, "distill_loss": 0.09814453125, "epoch": 1.6754155730533684, "explain_loss": 1.1015625, "loss": 1.84375, "ref_ce_loss": 0.625, "stage_align_loss": 0.00360107421875, "step": 3830 }, { "ce_loss": 0.002685546875, "distill_loss": 0.072265625, "epoch": 1.6754155730533684, "explain_loss": 0.9609375, "loss": 1.25, "ref_ce_loss": 0.2109375, "stage_align_loss": 0.0028076171875, "step": 3830 }, { "ce_loss": 0.54296875, "distill_loss": 0.0751953125, "epoch": 1.6754155730533684, "explain_loss": 1.3046875, "loss": 2.515625, "ref_ce_loss": 0.5625, "stage_align_loss": 0.03076171875, "step": 3830 }, { "ce_loss": 0.275390625, "distill_loss": 0.083984375, "epoch": 1.6754155730533684, "explain_loss": 1.0078125, "loss": 1.6953125, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.006744384765625, "step": 3830 }, { "epoch": 1.6797900262467191, "grad_norm": 0.5631242990493774, "learning_rate": 8.852464732896019e-05, "loss": 1.78037109375, "step": 3840 }, { "ce_loss": 0.10791015625, "distill_loss": 0.0595703125, "epoch": 1.6797900262467191, "explain_loss": 1.0390625, "loss": 1.84375, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0026702880859375, "step": 3840 }, { "ce_loss": 0.248046875, "distill_loss": 0.09521484375, "epoch": 1.6797900262467191, "explain_loss": 1.046875, "loss": 1.8046875, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0179443359375, "step": 3840 }, { "ce_loss": 0.002685546875, "distill_loss": 0.103515625, "epoch": 1.6797900262467191, "explain_loss": 1.0390625, "loss": 1.5, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.001922607421875, "step": 3840 }, { "ce_loss": 0.0223388671875, "distill_loss": 0.1220703125, "epoch": 1.6797900262467191, "explain_loss": 1.4921875, "loss": 1.984375, "ref_ce_loss": 0.34375, "stage_align_loss": 0.003448486328125, "step": 3840 }, { "epoch": 1.6841644794400699, "grad_norm": 0.765076756477356, "learning_rate": 8.804575892086735e-05, "loss": 1.73095703125, "step": 3850 }, { "ce_loss": 0.2138671875, "distill_loss": 0.0810546875, "epoch": 1.6841644794400699, "explain_loss": 1.109375, "loss": 1.921875, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.026611328125, "step": 3850 }, { "ce_loss": 0.15625, "distill_loss": 0.08251953125, "epoch": 1.6841644794400699, "explain_loss": 1.0546875, "loss": 1.671875, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.0038909912109375, "step": 3850 }, { "ce_loss": 0.265625, "distill_loss": 0.08544921875, "epoch": 1.6841644794400699, "explain_loss": 0.98046875, "loss": 1.796875, "ref_ce_loss": 0.4375, "stage_align_loss": 0.0262451171875, "step": 3850 }, { "ce_loss": 0.003021240234375, "distill_loss": 0.1337890625, "epoch": 1.6841644794400699, "explain_loss": 0.9453125, "loss": 1.609375, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.0019683837890625, "step": 3850 }, { "epoch": 1.6885389326334208, "grad_norm": 0.8092917203903198, "learning_rate": 8.756714847902809e-05, "loss": 1.78857421875, "step": 3860 }, { "ce_loss": 0.0125732421875, "distill_loss": 0.115234375, "epoch": 1.6885389326334208, "explain_loss": 0.75, "loss": 1.359375, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.002777099609375, "step": 3860 }, { "ce_loss": 0.01153564453125, "distill_loss": 0.05810546875, "epoch": 1.6885389326334208, "explain_loss": 0.9375, "loss": 1.25, "ref_ce_loss": 0.2392578125, "stage_align_loss": 0.0029449462890625, "step": 3860 }, { "ce_loss": 0.388671875, "distill_loss": 0.0751953125, "epoch": 1.6885389326334208, "explain_loss": 1.484375, "loss": 2.4375, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.029541015625, "step": 3860 }, { "ce_loss": 0.435546875, "distill_loss": 0.08056640625, "epoch": 1.6885389326334208, "explain_loss": 1.28125, "loss": 2.234375, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.080078125, "step": 3860 }, { "epoch": 1.6929133858267718, "grad_norm": 0.610823929309845, "learning_rate": 8.708882713234218e-05, "loss": 1.7431640625, "step": 3870 }, { "ce_loss": 0.39453125, "distill_loss": 0.115234375, "epoch": 1.6929133858267718, "explain_loss": 0.953125, "loss": 1.78125, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.0031585693359375, "step": 3870 }, { "ce_loss": 0.333984375, "distill_loss": 0.08740234375, "epoch": 1.6929133858267718, "explain_loss": 1.3984375, "loss": 2.1875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0032196044921875, "step": 3870 }, { "ce_loss": 0.11181640625, "distill_loss": 0.0693359375, "epoch": 1.6929133858267718, "explain_loss": 1.078125, "loss": 1.6953125, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.07080078125, "step": 3870 }, { "ce_loss": 0.1025390625, "distill_loss": 0.08251953125, "epoch": 1.6929133858267718, "explain_loss": 1.2734375, "loss": 1.90625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.0233154296875, "step": 3870 }, { "epoch": 1.6972878390201225, "grad_norm": 0.6839799284934998, "learning_rate": 8.661080600298736e-05, "loss": 1.720703125, "step": 3880 }, { "ce_loss": 0.384765625, "distill_loss": 0.0849609375, "epoch": 1.6972878390201225, "explain_loss": 1.1171875, "loss": 1.90625, "ref_ce_loss": 0.2392578125, "stage_align_loss": 0.07666015625, "step": 3880 }, { "ce_loss": 0.0027008056640625, "distill_loss": 0.078125, "epoch": 1.6972878390201225, "explain_loss": 0.875, "loss": 1.2421875, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.002349853515625, "step": 3880 }, { "ce_loss": 0.0205078125, "distill_loss": 0.0869140625, "epoch": 1.6972878390201225, "explain_loss": 1.15625, "loss": 1.78125, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.004669189453125, "step": 3880 }, { "ce_loss": 0.73828125, "distill_loss": 0.09716796875, "epoch": 1.6972878390201225, "explain_loss": 1.046875, "loss": 2.296875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.07421875, "step": 3880 }, { "epoch": 1.7016622922134732, "grad_norm": 1.1477733850479126, "learning_rate": 8.613309620616029e-05, "loss": 1.796875, "step": 3890 }, { "ce_loss": 0.003692626953125, "distill_loss": 0.087890625, "epoch": 1.7016622922134732, "explain_loss": 1.328125, "loss": 1.8515625, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0023956298828125, "step": 3890 }, { "ce_loss": 0.003570556640625, "distill_loss": 0.095703125, "epoch": 1.7016622922134732, "explain_loss": 0.8828125, "loss": 1.421875, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0028228759765625, "step": 3890 }, { "ce_loss": 0.2890625, "distill_loss": 0.134765625, "epoch": 1.7016622922134732, "explain_loss": 0.89453125, "loss": 1.828125, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0205078125, "step": 3890 }, { "ce_loss": 0.00640869140625, "distill_loss": 0.0732421875, "epoch": 1.7016622922134732, "explain_loss": 0.98828125, "loss": 1.390625, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.002471923828125, "step": 3890 }, { "epoch": 1.7060367454068242, "grad_norm": 0.9185165166854858, "learning_rate": 8.565570884981868e-05, "loss": 1.774755859375, "step": 3900 }, { "ce_loss": 0.043701171875, "distill_loss": 0.0791015625, "epoch": 1.7060367454068242, "explain_loss": 0.6015625, "loss": 1.0859375, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.00555419921875, "step": 3900 }, { "ce_loss": 0.2470703125, "distill_loss": 0.0751953125, "epoch": 1.7060367454068242, "explain_loss": 1.328125, "loss": 2.125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.052001953125, "step": 3900 }, { "ce_loss": 0.6875, "distill_loss": 0.1015625, "epoch": 1.7060367454068242, "explain_loss": 1.328125, "loss": 2.5625, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.00653076171875, "step": 3900 }, { "ce_loss": 0.455078125, "distill_loss": 0.07421875, "epoch": 1.7060367454068242, "explain_loss": 0.8984375, "loss": 1.84375, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.0033416748046875, "step": 3900 }, { "epoch": 1.710411198600175, "grad_norm": 3.3759360313415527, "learning_rate": 8.51786550344224e-05, "loss": 1.81904296875, "step": 3910 }, { "ce_loss": 0.76953125, "distill_loss": 0.119140625, "epoch": 1.710411198600175, "explain_loss": 1.34375, "loss": 2.75, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.07861328125, "step": 3910 }, { "ce_loss": 0.005340576171875, "distill_loss": 0.146484375, "epoch": 1.710411198600175, "explain_loss": 1.15625, "loss": 1.640625, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.0035552978515625, "step": 3910 }, { "ce_loss": 0.05810546875, "distill_loss": 0.40625, "epoch": 1.710411198600175, "explain_loss": 1.25, "loss": 2.171875, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.00469970703125, "step": 3910 }, { "ce_loss": 0.298828125, "distill_loss": 0.064453125, "epoch": 1.710411198600175, "explain_loss": 0.8125, "loss": 1.5, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.00433349609375, "step": 3910 }, { "epoch": 1.7147856517935258, "grad_norm": 2.290956974029541, "learning_rate": 8.470194585267587e-05, "loss": 1.8236328125, "step": 3920 }, { "ce_loss": 0.03466796875, "distill_loss": 0.119140625, "epoch": 1.7147856517935258, "explain_loss": 1.3671875, "loss": 1.859375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.00390625, "step": 3920 }, { "ce_loss": 0.0262451171875, "distill_loss": 0.1181640625, "epoch": 1.7147856517935258, "explain_loss": 1.1171875, "loss": 1.4140625, "ref_ce_loss": 0.1474609375, "stage_align_loss": 0.002777099609375, "step": 3920 }, { "ce_loss": 0.1318359375, "distill_loss": 0.08544921875, "epoch": 1.7147856517935258, "explain_loss": 1.234375, "loss": 1.828125, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.00360107421875, "step": 3920 }, { "ce_loss": 0.8046875, "distill_loss": 0.0927734375, "epoch": 1.7147856517935258, "explain_loss": 1.359375, "loss": 3.03125, "ref_ce_loss": 0.69140625, "stage_align_loss": 0.07421875, "step": 3920 }, { "epoch": 1.7191601049868765, "grad_norm": 0.9610370993614197, "learning_rate": 8.422559238926984e-05, "loss": 1.8056640625, "step": 3930 }, { "ce_loss": 0.0361328125, "distill_loss": 0.0869140625, "epoch": 1.7191601049868765, "explain_loss": 0.98828125, "loss": 1.453125, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.00433349609375, "step": 3930 }, { "ce_loss": 0.0791015625, "distill_loss": 0.0908203125, "epoch": 1.7191601049868765, "explain_loss": 0.98828125, "loss": 1.6484375, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.003021240234375, "step": 3930 }, { "ce_loss": 0.07958984375, "distill_loss": 0.08740234375, "epoch": 1.7191601049868765, "explain_loss": 0.875, "loss": 1.4609375, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0299072265625, "step": 3930 }, { "ce_loss": 0.0017547607421875, "distill_loss": 0.08837890625, "epoch": 1.7191601049868765, "explain_loss": 1.046875, "loss": 1.515625, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0035552978515625, "step": 3930 }, { "epoch": 1.7235345581802275, "grad_norm": 1.2480878829956055, "learning_rate": 8.374960572062359e-05, "loss": 1.88154296875, "step": 3940 }, { "ce_loss": 0.0024871826171875, "distill_loss": 0.1474609375, "epoch": 1.7235345581802275, "explain_loss": 0.91015625, "loss": 1.4453125, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00299072265625, "step": 3940 }, { "ce_loss": 0.007537841796875, "distill_loss": 0.06884765625, "epoch": 1.7235345581802275, "explain_loss": 1.1640625, "loss": 1.71875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.004150390625, "step": 3940 }, { "ce_loss": 0.04736328125, "distill_loss": 0.1845703125, "epoch": 1.7235345581802275, "explain_loss": 1.4375, "loss": 2.015625, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0023193359375, "step": 3940 }, { "ce_loss": 0.007110595703125, "distill_loss": 0.14453125, "epoch": 1.7235345581802275, "explain_loss": 1.5625, "loss": 2.328125, "ref_ce_loss": 0.609375, "stage_align_loss": 0.0020599365234375, "step": 3940 }, { "epoch": 1.7279090113735784, "grad_norm": 0.8008632063865662, "learning_rate": 8.327399691462771e-05, "loss": 1.8099609375, "step": 3950 }, { "ce_loss": 0.08935546875, "distill_loss": 0.115234375, "epoch": 1.7279090113735784, "explain_loss": 1.1875, "loss": 1.765625, "ref_ce_loss": 0.375, "stage_align_loss": 0.002532958984375, "step": 3950 }, { "ce_loss": 0.1953125, "distill_loss": 0.1240234375, "epoch": 1.7279090113735784, "explain_loss": 0.89453125, "loss": 1.5078125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0028839111328125, "step": 3950 }, { "ce_loss": 0.1953125, "distill_loss": 0.130859375, "epoch": 1.7279090113735784, "explain_loss": 0.75, "loss": 1.46875, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.005035400390625, "step": 3950 }, { "ce_loss": 0.049072265625, "distill_loss": 0.09228515625, "epoch": 1.7279090113735784, "explain_loss": 1.3046875, "loss": 1.859375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.003570556640625, "step": 3950 }, { "epoch": 1.7322834645669292, "grad_norm": 0.8852497935295105, "learning_rate": 8.279877703038634e-05, "loss": 1.790380859375, "step": 3960 }, { "ce_loss": 0.039794921875, "distill_loss": 0.1416015625, "epoch": 1.7322834645669292, "explain_loss": 1.140625, "loss": 1.765625, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.002899169921875, "step": 3960 }, { "ce_loss": 0.9609375, "distill_loss": 0.09765625, "epoch": 1.7322834645669292, "explain_loss": 1.09375, "loss": 2.625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.045166015625, "step": 3960 }, { "ce_loss": 0.458984375, "distill_loss": 0.1171875, "epoch": 1.7322834645669292, "explain_loss": 1.1640625, "loss": 2.140625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0034332275390625, "step": 3960 }, { "ce_loss": 0.69140625, "distill_loss": 0.0849609375, "epoch": 1.7322834645669292, "explain_loss": 1.2265625, "loss": 2.59375, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.021728515625, "step": 3960 }, { "epoch": 1.73665791776028, "grad_norm": 0.8173248171806335, "learning_rate": 8.232395711796034e-05, "loss": 1.85576171875, "step": 3970 }, { "ce_loss": 0.388671875, "distill_loss": 0.0947265625, "epoch": 1.73665791776028, "explain_loss": 1.4140625, "loss": 2.53125, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.01708984375, "step": 3970 }, { "ce_loss": 0.00372314453125, "distill_loss": 0.099609375, "epoch": 1.73665791776028, "explain_loss": 1.5, "loss": 1.984375, "ref_ce_loss": 0.375, "stage_align_loss": 0.0019683837890625, "step": 3970 }, { "ce_loss": 0.0693359375, "distill_loss": 0.0654296875, "epoch": 1.73665791776028, "explain_loss": 0.875, "loss": 1.25, "ref_ce_loss": 0.2392578125, "stage_align_loss": 0.002227783203125, "step": 3970 }, { "ce_loss": 0.1328125, "distill_loss": 0.072265625, "epoch": 1.73665791776028, "explain_loss": 0.9296875, "loss": 1.4375, "ref_ce_loss": 0.296875, "stage_align_loss": 0.0029296875, "step": 3970 }, { "epoch": 1.7410323709536308, "grad_norm": 1.538632869720459, "learning_rate": 8.184954821811013e-05, "loss": 1.71796875, "step": 3980 }, { "ce_loss": 0.00531005859375, "distill_loss": 0.0751953125, "epoch": 1.7410323709536308, "explain_loss": 1.03125, "loss": 1.46875, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0017852783203125, "step": 3980 }, { "ce_loss": 0.0167236328125, "distill_loss": 0.0810546875, "epoch": 1.7410323709536308, "explain_loss": 0.921875, "loss": 1.53125, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.002288818359375, "step": 3980 }, { "ce_loss": 0.080078125, "distill_loss": 0.126953125, "epoch": 1.7410323709536308, "explain_loss": 1.0703125, "loss": 1.65625, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.004608154296875, "step": 3980 }, { "ce_loss": 1.046875, "distill_loss": 0.091796875, "epoch": 1.7410323709536308, "explain_loss": 1.2109375, "loss": 2.90625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0576171875, "step": 3980 }, { "epoch": 1.7454068241469818, "grad_norm": 0.854701578617096, "learning_rate": 8.13755613620391e-05, "loss": 1.75439453125, "step": 3990 }, { "ce_loss": 0.51171875, "distill_loss": 0.1474609375, "epoch": 1.7454068241469818, "explain_loss": 0.7578125, "loss": 1.8828125, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.002777099609375, "step": 3990 }, { "ce_loss": 0.007232666015625, "distill_loss": 0.103515625, "epoch": 1.7454068241469818, "explain_loss": 0.98828125, "loss": 1.4609375, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.00189208984375, "step": 3990 }, { "ce_loss": 0.00616455078125, "distill_loss": 0.111328125, "epoch": 1.7454068241469818, "explain_loss": 0.8671875, "loss": 1.3515625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.002777099609375, "step": 3990 }, { "ce_loss": 0.05859375, "distill_loss": 0.09326171875, "epoch": 1.7454068241469818, "explain_loss": 1.0859375, "loss": 1.65625, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.004241943359375, "step": 3990 }, { "epoch": 1.7497812773403325, "grad_norm": 0.6830964088439941, "learning_rate": 8.090200757113703e-05, "loss": 1.83369140625, "step": 4000 }, { "ce_loss": 0.47265625, "distill_loss": 0.09814453125, "epoch": 1.7497812773403325, "explain_loss": 1.265625, "loss": 2.21875, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0244140625, "step": 4000 }, { "ce_loss": 0.01214599609375, "distill_loss": 0.177734375, "epoch": 1.7497812773403325, "explain_loss": 1.1328125, "loss": 1.671875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.002685546875, "step": 4000 }, { "ce_loss": 0.146484375, "distill_loss": 0.0859375, "epoch": 1.7497812773403325, "explain_loss": 1.2890625, "loss": 2.015625, "ref_ce_loss": 0.49609375, "stage_align_loss": 0.004180908203125, "step": 4000 }, { "ce_loss": 0.60546875, "distill_loss": 0.09228515625, "epoch": 1.7497812773403325, "explain_loss": 1.2734375, "loss": 2.65625, "ref_ce_loss": 0.60546875, "stage_align_loss": 0.0771484375, "step": 4000 }, { "epoch": 1.7541557305336832, "grad_norm": 0.9258520603179932, "learning_rate": 8.042889785672389e-05, "loss": 1.7775390625, "step": 4010 }, { "ce_loss": 0.232421875, "distill_loss": 0.1416015625, "epoch": 1.7541557305336832, "explain_loss": 0.85546875, "loss": 1.5234375, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.003326416015625, "step": 4010 }, { "ce_loss": 0.44140625, "distill_loss": 0.1806640625, "epoch": 1.7541557305336832, "explain_loss": 1.0546875, "loss": 2.1875, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.059326171875, "step": 4010 }, { "ce_loss": 0.1650390625, "distill_loss": 0.0927734375, "epoch": 1.7541557305336832, "explain_loss": 0.91796875, "loss": 1.46875, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.005950927734375, "step": 4010 }, { "ce_loss": 0.1748046875, "distill_loss": 0.1201171875, "epoch": 1.7541557305336832, "explain_loss": 1.0703125, "loss": 2.03125, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.00836181640625, "step": 4010 }, { "epoch": 1.758530183727034, "grad_norm": 0.7121061682701111, "learning_rate": 7.995624321979368e-05, "loss": 1.8205078125, "step": 4020 }, { "ce_loss": 0.26171875, "distill_loss": 0.251953125, "epoch": 1.758530183727034, "explain_loss": 1.09375, "loss": 2.171875, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.04931640625, "step": 4020 }, { "ce_loss": 0.003509521484375, "distill_loss": 0.140625, "epoch": 1.758530183727034, "explain_loss": 1.15625, "loss": 1.6015625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.002471923828125, "step": 4020 }, { "ce_loss": 0.279296875, "distill_loss": 0.11669921875, "epoch": 1.758530183727034, "explain_loss": 1.234375, "loss": 2.125, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.024658203125, "step": 4020 }, { "ce_loss": 0.037353515625, "distill_loss": 0.10498046875, "epoch": 1.758530183727034, "explain_loss": 1.390625, "loss": 1.96875, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0024871826171875, "step": 4020 }, { "epoch": 1.762904636920385, "grad_norm": 0.7156880497932434, "learning_rate": 7.948405465075876e-05, "loss": 1.7498046875, "step": 4030 }, { "ce_loss": 0.013427734375, "distill_loss": 0.10302734375, "epoch": 1.762904636920385, "explain_loss": 1.1953125, "loss": 1.65625, "ref_ce_loss": 0.34375, "stage_align_loss": 0.0026702880859375, "step": 4030 }, { "ce_loss": 0.041259765625, "distill_loss": 0.06787109375, "epoch": 1.762904636920385, "explain_loss": 1.015625, "loss": 1.484375, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0242919921875, "step": 4030 }, { "ce_loss": 0.455078125, "distill_loss": 0.1259765625, "epoch": 1.762904636920385, "explain_loss": 1.546875, "loss": 2.765625, "ref_ce_loss": 0.63671875, "stage_align_loss": 0.0033721923828125, "step": 4030 }, { "ce_loss": 0.00927734375, "distill_loss": 0.07666015625, "epoch": 1.762904636920385, "explain_loss": 1.15625, "loss": 1.53125, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.003662109375, "step": 4030 }, { "epoch": 1.7672790901137359, "grad_norm": 0.6914727687835693, "learning_rate": 7.90123431291942e-05, "loss": 1.77744140625, "step": 4040 }, { "ce_loss": 0.00836181640625, "distill_loss": 0.177734375, "epoch": 1.7672790901137359, "explain_loss": 1.234375, "loss": 1.9375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.0017242431640625, "step": 4040 }, { "ce_loss": 0.0673828125, "distill_loss": 0.126953125, "epoch": 1.7672790901137359, "explain_loss": 1.15625, "loss": 1.671875, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0028533935546875, "step": 4040 }, { "ce_loss": 0.004241943359375, "distill_loss": 0.0986328125, "epoch": 1.7672790901137359, "explain_loss": 0.671875, "loss": 1.15625, "ref_ce_loss": 0.375, "stage_align_loss": 0.002471923828125, "step": 4040 }, { "ce_loss": 0.058349609375, "distill_loss": 0.11669921875, "epoch": 1.7672790901137359, "explain_loss": 0.9453125, "loss": 1.546875, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0037384033203125, "step": 4040 }, { "epoch": 1.7716535433070866, "grad_norm": 1.0442255735397339, "learning_rate": 7.854111962358257e-05, "loss": 1.801513671875, "step": 4050 }, { "ce_loss": 0.000461578369140625, "distill_loss": 0.09033203125, "epoch": 1.7716535433070866, "explain_loss": 1.0859375, "loss": 1.4609375, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.0023345947265625, "step": 4050 }, { "ce_loss": 0.004364013671875, "distill_loss": 0.10009765625, "epoch": 1.7716535433070866, "explain_loss": 1.109375, "loss": 1.5625, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.0016632080078125, "step": 4050 }, { "ce_loss": 0.140625, "distill_loss": 0.11474609375, "epoch": 1.7716535433070866, "explain_loss": 1.09375, "loss": 1.6015625, "ref_ce_loss": 0.248046875, "stage_align_loss": 0.0052490234375, "step": 4050 }, { "ce_loss": 0.009033203125, "distill_loss": 0.1279296875, "epoch": 1.7716535433070866, "explain_loss": 1.296875, "loss": 1.90625, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0021209716796875, "step": 4050 }, { "epoch": 1.7760279965004373, "grad_norm": 0.7354423999786377, "learning_rate": 7.807039509105871e-05, "loss": 1.86259765625, "step": 4060 }, { "ce_loss": 0.09765625, "distill_loss": 0.11767578125, "epoch": 1.7760279965004373, "explain_loss": 0.78515625, "loss": 1.3359375, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.0035552978515625, "step": 4060 }, { "ce_loss": 0.0208740234375, "distill_loss": 0.0966796875, "epoch": 1.7760279965004373, "explain_loss": 1.3125, "loss": 1.765625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.0029296875, "step": 4060 }, { "ce_loss": 0.048583984375, "distill_loss": 0.0830078125, "epoch": 1.7760279965004373, "explain_loss": 0.6796875, "loss": 1.046875, "ref_ce_loss": 0.232421875, "stage_align_loss": 0.0030670166015625, "step": 4060 }, { "ce_loss": 0.03271484375, "distill_loss": 0.109375, "epoch": 1.7760279965004373, "explain_loss": 1.359375, "loss": 2.015625, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.0026397705078125, "step": 4060 }, { "epoch": 1.7804024496937882, "grad_norm": 0.8768242597579956, "learning_rate": 7.760018047715524e-05, "loss": 1.84794921875, "step": 4070 }, { "ce_loss": 0.1708984375, "distill_loss": 0.1044921875, "epoch": 1.7804024496937882, "explain_loss": 0.90234375, "loss": 1.5625, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0198974609375, "step": 4070 }, { "ce_loss": 0.1865234375, "distill_loss": 0.06591796875, "epoch": 1.7804024496937882, "explain_loss": 1.015625, "loss": 1.6640625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.032470703125, "step": 4070 }, { "ce_loss": 0.1142578125, "distill_loss": 0.1162109375, "epoch": 1.7804024496937882, "explain_loss": 0.890625, "loss": 1.6484375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.015869140625, "step": 4070 }, { "ce_loss": 0.044189453125, "distill_loss": 0.078125, "epoch": 1.7804024496937882, "explain_loss": 0.953125, "loss": 1.6015625, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.004119873046875, "step": 4070 }, { "epoch": 1.7847769028871392, "grad_norm": 0.9829733967781067, "learning_rate": 7.713048671554782e-05, "loss": 1.7880859375, "step": 4080 }, { "ce_loss": 0.01165771484375, "distill_loss": 0.20703125, "epoch": 1.7847769028871392, "explain_loss": 0.765625, "loss": 1.390625, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.00162506103515625, "step": 4080 }, { "ce_loss": 0.005126953125, "distill_loss": 0.123046875, "epoch": 1.7847769028871392, "explain_loss": 0.8359375, "loss": 1.2890625, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0023651123046875, "step": 4080 }, { "ce_loss": 0.01104736328125, "distill_loss": 0.10693359375, "epoch": 1.7847769028871392, "explain_loss": 1.0625, "loss": 1.5859375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00147247314453125, "step": 4080 }, { "ce_loss": 0.11376953125, "distill_loss": 0.1025390625, "epoch": 1.7847769028871392, "explain_loss": 0.890625, "loss": 1.484375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.004058837890625, "step": 4080 }, { "epoch": 1.78915135608049, "grad_norm": 1.064321756362915, "learning_rate": 7.666132472780087e-05, "loss": 1.80966796875, "step": 4090 }, { "ce_loss": 0.09814453125, "distill_loss": 0.1123046875, "epoch": 1.78915135608049, "explain_loss": 1.109375, "loss": 1.515625, "ref_ce_loss": 0.1923828125, "stage_align_loss": 0.0023956298828125, "step": 4090 }, { "ce_loss": 0.44140625, "distill_loss": 0.091796875, "epoch": 1.78915135608049, "explain_loss": 1.0, "loss": 1.828125, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.0025177001953125, "step": 4090 }, { "ce_loss": 0.0078125, "distill_loss": 0.107421875, "epoch": 1.78915135608049, "explain_loss": 1.2734375, "loss": 1.703125, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0022125244140625, "step": 4090 }, { "ce_loss": 0.080078125, "distill_loss": 0.12890625, "epoch": 1.78915135608049, "explain_loss": 1.1640625, "loss": 1.6796875, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.00170135498046875, "step": 4090 }, { "epoch": 1.7935258092738406, "grad_norm": 0.7191616892814636, "learning_rate": 7.619270542311396e-05, "loss": 1.74267578125, "step": 4100 }, { "ce_loss": 0.0145263671875, "distill_loss": 0.08740234375, "epoch": 1.7935258092738406, "explain_loss": 1.265625, "loss": 1.875, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0027923583984375, "step": 4100 }, { "ce_loss": 0.01043701171875, "distill_loss": 0.072265625, "epoch": 1.7935258092738406, "explain_loss": 1.1875, "loss": 1.53125, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.0028228759765625, "step": 4100 }, { "ce_loss": 0.2216796875, "distill_loss": 0.1162109375, "epoch": 1.7935258092738406, "explain_loss": 1.046875, "loss": 2.015625, "ref_ce_loss": 0.5703125, "stage_align_loss": 0.059814453125, "step": 4100 }, { "ce_loss": 0.05126953125, "distill_loss": 0.09033203125, "epoch": 1.7935258092738406, "explain_loss": 0.9453125, "loss": 1.53125, "ref_ce_loss": 0.4375, "stage_align_loss": 0.004241943359375, "step": 4100 }, { "epoch": 1.7979002624671916, "grad_norm": 1.004927635192871, "learning_rate": 7.572463969806767e-05, "loss": 1.77099609375, "step": 4110 }, { "ce_loss": 0.08203125, "distill_loss": 0.07177734375, "epoch": 1.7979002624671916, "explain_loss": 1.4609375, "loss": 2.125, "ref_ce_loss": 0.5, "stage_align_loss": 0.0019073486328125, "step": 4110 }, { "ce_loss": 0.01068115234375, "distill_loss": 0.1015625, "epoch": 1.7979002624671916, "explain_loss": 1.3515625, "loss": 1.7578125, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.00173187255859375, "step": 4110 }, { "ce_loss": 1.46875, "distill_loss": 0.1123046875, "epoch": 1.7979002624671916, "explain_loss": 1.2265625, "loss": 3.28125, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0732421875, "step": 4110 }, { "ce_loss": 0.1123046875, "distill_loss": 0.10009765625, "epoch": 1.7979002624671916, "explain_loss": 1.390625, "loss": 1.96875, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.003204345703125, "step": 4110 }, { "epoch": 1.8022747156605425, "grad_norm": 0.9244388937950134, "learning_rate": 7.525713843637064e-05, "loss": 1.7880859375, "step": 4120 }, { "ce_loss": 0.0238037109375, "distill_loss": 0.08740234375, "epoch": 1.8022747156605425, "explain_loss": 0.7578125, "loss": 1.265625, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.005340576171875, "step": 4120 }, { "ce_loss": 0.06640625, "distill_loss": 0.09326171875, "epoch": 1.8022747156605425, "explain_loss": 1.4765625, "loss": 2.0625, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.0032806396484375, "step": 4120 }, { "ce_loss": 0.010009765625, "distill_loss": 0.0859375, "epoch": 1.8022747156605425, "explain_loss": 1.421875, "loss": 2.0, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.0019989013671875, "step": 4120 }, { "ce_loss": 0.01220703125, "distill_loss": 0.1025390625, "epoch": 1.8022747156605425, "explain_loss": 0.96875, "loss": 1.421875, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0027313232421875, "step": 4120 }, { "epoch": 1.8066491688538933, "grad_norm": 0.7084280252456665, "learning_rate": 7.479021250860618e-05, "loss": 1.80791015625, "step": 4130 }, { "ce_loss": 0.349609375, "distill_loss": 0.09326171875, "epoch": 1.8066491688538933, "explain_loss": 0.70703125, "loss": 1.5078125, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.08056640625, "step": 4130 }, { "ce_loss": 0.06689453125, "distill_loss": 0.091796875, "epoch": 1.8066491688538933, "explain_loss": 1.2734375, "loss": 1.890625, "ref_ce_loss": 0.453125, "stage_align_loss": 0.0027618408203125, "step": 4130 }, { "ce_loss": 0.158203125, "distill_loss": 0.0869140625, "epoch": 1.8066491688538933, "explain_loss": 0.9140625, "loss": 1.78125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.07373046875, "step": 4130 }, { "ce_loss": 0.25390625, "distill_loss": 0.16796875, "epoch": 1.8066491688538933, "explain_loss": 1.1953125, "loss": 2.09375, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0284423828125, "step": 4130 }, { "epoch": 1.811023622047244, "grad_norm": 0.89706951379776, "learning_rate": 7.432387277197971e-05, "loss": 1.77841796875, "step": 4140 }, { "ce_loss": 0.0703125, "distill_loss": 0.1435546875, "epoch": 1.811023622047244, "explain_loss": 0.96875, "loss": 1.625, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0022735595703125, "step": 4140 }, { "ce_loss": 0.026123046875, "distill_loss": 0.07861328125, "epoch": 1.811023622047244, "explain_loss": 0.8203125, "loss": 1.359375, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.003509521484375, "step": 4140 }, { "ce_loss": 0.002349853515625, "distill_loss": 0.07666015625, "epoch": 1.811023622047244, "explain_loss": 0.92578125, "loss": 1.328125, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.002960205078125, "step": 4140 }, { "ce_loss": 0.38671875, "distill_loss": 0.08251953125, "epoch": 1.811023622047244, "explain_loss": 1.140625, "loss": 2.171875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.01434326171875, "step": 4140 }, { "epoch": 1.815398075240595, "grad_norm": 0.68935626745224, "learning_rate": 7.38581300700662e-05, "loss": 1.7595703125, "step": 4150 }, { "ce_loss": 0.1982421875, "distill_loss": 0.061279296875, "epoch": 1.815398075240595, "explain_loss": 0.69921875, "loss": 1.34375, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.01055908203125, "step": 4150 }, { "ce_loss": 0.10693359375, "distill_loss": 0.08740234375, "epoch": 1.815398075240595, "explain_loss": 1.140625, "loss": 1.8125, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.034423828125, "step": 4150 }, { "ce_loss": 0.035400390625, "distill_loss": 0.1025390625, "epoch": 1.815398075240595, "explain_loss": 0.9609375, "loss": 1.421875, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.002044677734375, "step": 4150 }, { "ce_loss": 0.2119140625, "distill_loss": 0.1083984375, "epoch": 1.815398075240595, "explain_loss": 1.46875, "loss": 2.125, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0028533935546875, "step": 4150 }, { "epoch": 1.8197725284339459, "grad_norm": 0.7946755886077881, "learning_rate": 7.339299523255814e-05, "loss": 1.73056640625, "step": 4160 }, { "ce_loss": 0.035400390625, "distill_loss": 0.1083984375, "epoch": 1.8197725284339459, "explain_loss": 1.015625, "loss": 1.703125, "ref_ce_loss": 0.53515625, "stage_align_loss": 0.0036773681640625, "step": 4160 }, { "ce_loss": 0.001373291015625, "distill_loss": 0.103515625, "epoch": 1.8197725284339459, "explain_loss": 1.1015625, "loss": 1.546875, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.00225830078125, "step": 4160 }, { "ce_loss": 0.08056640625, "distill_loss": 0.078125, "epoch": 1.8197725284339459, "explain_loss": 1.21875, "loss": 1.8203125, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.040771484375, "step": 4160 }, { "ce_loss": 0.054443359375, "distill_loss": 0.10986328125, "epoch": 1.8197725284339459, "explain_loss": 0.83984375, "loss": 1.515625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0021820068359375, "step": 4160 }, { "epoch": 1.8241469816272966, "grad_norm": 0.8103623986244202, "learning_rate": 7.292847907501346e-05, "loss": 1.705712890625, "step": 4170 }, { "ce_loss": 0.8046875, "distill_loss": 0.107421875, "epoch": 1.8241469816272966, "explain_loss": 1.1953125, "loss": 2.5625, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.0026092529296875, "step": 4170 }, { "ce_loss": 0.033935546875, "distill_loss": 0.09716796875, "epoch": 1.8241469816272966, "explain_loss": 0.9140625, "loss": 1.34375, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.0031890869140625, "step": 4170 }, { "ce_loss": 0.01611328125, "distill_loss": 0.0654296875, "epoch": 1.8241469816272966, "explain_loss": 1.2109375, "loss": 1.5390625, "ref_ce_loss": 0.244140625, "stage_align_loss": 0.002685546875, "step": 4170 }, { "ce_loss": 0.0007476806640625, "distill_loss": 0.1103515625, "epoch": 1.8241469816272966, "explain_loss": 1.125, "loss": 1.671875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.00194549560546875, "step": 4170 }, { "epoch": 1.8285214348206473, "grad_norm": 1.1458008289337158, "learning_rate": 7.24645923986044e-05, "loss": 1.766796875, "step": 4180 }, { "ce_loss": 0.00531005859375, "distill_loss": 0.10205078125, "epoch": 1.8285214348206473, "explain_loss": 1.1640625, "loss": 1.578125, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.0118408203125, "step": 4180 }, { "ce_loss": 0.83984375, "distill_loss": 0.062255859375, "epoch": 1.8285214348206473, "explain_loss": 1.171875, "loss": 2.515625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.0439453125, "step": 4180 }, { "ce_loss": 0.177734375, "distill_loss": 0.083984375, "epoch": 1.8285214348206473, "explain_loss": 1.203125, "loss": 1.8671875, "ref_ce_loss": 0.375, "stage_align_loss": 0.0274658203125, "step": 4180 }, { "ce_loss": 0.038330078125, "distill_loss": 0.111328125, "epoch": 1.8285214348206473, "explain_loss": 1.078125, "loss": 1.640625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0031585693359375, "step": 4180 }, { "epoch": 1.8328958880139983, "grad_norm": 0.6341211199760437, "learning_rate": 7.200134598986606e-05, "loss": 1.71279296875, "step": 4190 }, { "ce_loss": 0.189453125, "distill_loss": 0.05859375, "epoch": 1.8328958880139983, "explain_loss": 0.96484375, "loss": 1.609375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.004302978515625, "step": 4190 }, { "ce_loss": 0.029296875, "distill_loss": 0.062255859375, "epoch": 1.8328958880139983, "explain_loss": 1.1796875, "loss": 1.6875, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0025177001953125, "step": 4190 }, { "ce_loss": 0.01068115234375, "distill_loss": 0.10693359375, "epoch": 1.8328958880139983, "explain_loss": 0.8046875, "loss": 1.375, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.002899169921875, "step": 4190 }, { "ce_loss": 0.85546875, "distill_loss": 0.091796875, "epoch": 1.8328958880139983, "explain_loss": 1.3046875, "loss": 2.703125, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.068359375, "step": 4190 }, { "epoch": 1.8372703412073492, "grad_norm": 0.6689817309379578, "learning_rate": 7.15387506204458e-05, "loss": 1.7572265625, "step": 4200 }, { "ce_loss": 0.00799560546875, "distill_loss": 0.11279296875, "epoch": 1.8372703412073492, "explain_loss": 1.375, "loss": 1.9921875, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.00225830078125, "step": 4200 }, { "ce_loss": 0.0732421875, "distill_loss": 0.06103515625, "epoch": 1.8372703412073492, "explain_loss": 0.7890625, "loss": 1.328125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.035888671875, "step": 4200 }, { "ce_loss": 0.020751953125, "distill_loss": 0.1455078125, "epoch": 1.8372703412073492, "explain_loss": 1.1015625, "loss": 1.609375, "ref_ce_loss": 0.34375, "stage_align_loss": 0.004058837890625, "step": 4200 }, { "ce_loss": 0.0242919921875, "distill_loss": 0.0625, "epoch": 1.8372703412073492, "explain_loss": 0.84765625, "loss": 1.34375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00244140625, "step": 4200 }, { "epoch": 1.8416447944007, "grad_norm": 1.0261813402175903, "learning_rate": 7.107681704685256e-05, "loss": 1.7486328125, "step": 4210 }, { "ce_loss": 0.404296875, "distill_loss": 0.083984375, "epoch": 1.8416447944007, "explain_loss": 1.1875, "loss": 2.1875, "ref_ce_loss": 0.515625, "stage_align_loss": 0.002227783203125, "step": 4210 }, { "ce_loss": 0.00396728515625, "distill_loss": 0.080078125, "epoch": 1.8416447944007, "explain_loss": 1.1171875, "loss": 1.46875, "ref_ce_loss": 0.265625, "stage_align_loss": 0.002655029296875, "step": 4210 }, { "ce_loss": 0.318359375, "distill_loss": 0.0927734375, "epoch": 1.8416447944007, "explain_loss": 1.0546875, "loss": 1.9453125, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.10009765625, "step": 4210 }, { "ce_loss": 0.06298828125, "distill_loss": 0.08544921875, "epoch": 1.8416447944007, "explain_loss": 1.34375, "loss": 1.859375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0034637451171875, "step": 4210 }, { "epoch": 1.8460192475940507, "grad_norm": 1.029557228088379, "learning_rate": 7.061555601020695e-05, "loss": 1.74326171875, "step": 4220 }, { "ce_loss": 0.33984375, "distill_loss": 0.1328125, "epoch": 1.8460192475940507, "explain_loss": 1.1171875, "loss": 2.125, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.046630859375, "step": 4220 }, { "ce_loss": 0.6484375, "distill_loss": 0.0751953125, "epoch": 1.8460192475940507, "explain_loss": 1.2265625, "loss": 2.5, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0021514892578125, "step": 4220 }, { "ce_loss": 0.040771484375, "distill_loss": 0.0693359375, "epoch": 1.8460192475940507, "explain_loss": 1.40625, "loss": 1.8984375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.00323486328125, "step": 4220 }, { "ce_loss": 0.0245361328125, "distill_loss": 0.07861328125, "epoch": 1.8460192475940507, "explain_loss": 0.8828125, "loss": 1.40625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.0029754638671875, "step": 4220 }, { "epoch": 1.8503937007874016, "grad_norm": 1.1461853981018066, "learning_rate": 7.015497823599138e-05, "loss": 1.7900390625, "step": 4230 }, { "ce_loss": 0.10400390625, "distill_loss": 0.0947265625, "epoch": 1.8503937007874016, "explain_loss": 1.171875, "loss": 1.796875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.01239013671875, "step": 4230 }, { "ce_loss": 0.04833984375, "distill_loss": 0.0693359375, "epoch": 1.8503937007874016, "explain_loss": 1.1953125, "loss": 1.7109375, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.00439453125, "step": 4230 }, { "ce_loss": 0.0966796875, "distill_loss": 0.125, "epoch": 1.8503937007874016, "explain_loss": 1.015625, "loss": 1.78125, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0111083984375, "step": 4230 }, { "ce_loss": 0.76953125, "distill_loss": 0.0654296875, "epoch": 1.8503937007874016, "explain_loss": 1.0, "loss": 2.28125, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.00408935546875, "step": 4230 }, { "epoch": 1.8547681539807526, "grad_norm": 1.0040833950042725, "learning_rate": 6.969509443380064e-05, "loss": 1.78408203125, "step": 4240 }, { "ce_loss": 0.0037384033203125, "distill_loss": 0.0751953125, "epoch": 1.8547681539807526, "explain_loss": 1.5234375, "loss": 2.109375, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.003265380859375, "step": 4240 }, { "ce_loss": 0.58984375, "distill_loss": 0.083984375, "epoch": 1.8547681539807526, "explain_loss": 0.98046875, "loss": 2.078125, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.06787109375, "step": 4240 }, { "ce_loss": 0.62890625, "distill_loss": 0.107421875, "epoch": 1.8547681539807526, "explain_loss": 0.671875, "loss": 1.84375, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.07763671875, "step": 4240 }, { "ce_loss": 0.388671875, "distill_loss": 0.109375, "epoch": 1.8547681539807526, "explain_loss": 1.484375, "loss": 2.53125, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0087890625, "step": 4240 }, { "epoch": 1.8591426071741033, "grad_norm": 0.7895421981811523, "learning_rate": 6.923591529709301e-05, "loss": 1.7564453125, "step": 4250 }, { "ce_loss": 0.01470947265625, "distill_loss": 0.0859375, "epoch": 1.8591426071741033, "explain_loss": 0.984375, "loss": 1.5625, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.002685546875, "step": 4250 }, { "ce_loss": 0.0167236328125, "distill_loss": 0.0673828125, "epoch": 1.8591426071741033, "explain_loss": 0.88671875, "loss": 1.3515625, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.0024871826171875, "step": 4250 }, { "ce_loss": 0.06396484375, "distill_loss": 0.12158203125, "epoch": 1.8591426071741033, "explain_loss": 0.78125, "loss": 1.2578125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0025177001953125, "step": 4250 }, { "ce_loss": 0.042724609375, "distill_loss": 0.087890625, "epoch": 1.8591426071741033, "explain_loss": 1.390625, "loss": 2.109375, "ref_ce_loss": 0.58984375, "stage_align_loss": 0.005035400390625, "step": 4250 }, { "epoch": 1.863517060367454, "grad_norm": 0.8274617791175842, "learning_rate": 6.877745150294135e-05, "loss": 1.7978515625, "step": 4260 }, { "ce_loss": 0.37890625, "distill_loss": 0.09033203125, "epoch": 1.863517060367454, "explain_loss": 1.1953125, "loss": 2.09375, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.08837890625, "step": 4260 }, { "ce_loss": 0.01904296875, "distill_loss": 0.083984375, "epoch": 1.863517060367454, "explain_loss": 0.78515625, "loss": 1.1015625, "ref_ce_loss": 0.208984375, "stage_align_loss": 0.003021240234375, "step": 4260 }, { "ce_loss": 0.29296875, "distill_loss": 0.0673828125, "epoch": 1.863517060367454, "explain_loss": 0.7109375, "loss": 1.3515625, "ref_ce_loss": 0.27734375, "stage_align_loss": 0.0025177001953125, "step": 4260 }, { "ce_loss": 0.0264892578125, "distill_loss": 0.08349609375, "epoch": 1.863517060367454, "explain_loss": 1.21875, "loss": 1.78125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0034637451171875, "step": 4260 }, { "epoch": 1.8678915135608047, "grad_norm": 1.0122042894363403, "learning_rate": 6.831971371178526e-05, "loss": 1.79365234375, "step": 4270 }, { "ce_loss": 0.234375, "distill_loss": 0.07275390625, "epoch": 1.8678915135608047, "explain_loss": 1.203125, "loss": 1.96875, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.02099609375, "step": 4270 }, { "ce_loss": 0.005859375, "distill_loss": 0.08837890625, "epoch": 1.8678915135608047, "explain_loss": 0.79296875, "loss": 1.3359375, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.0028533935546875, "step": 4270 }, { "ce_loss": 0.017822265625, "distill_loss": 0.087890625, "epoch": 1.8678915135608047, "explain_loss": 1.1171875, "loss": 1.6875, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.0018463134765625, "step": 4270 }, { "ce_loss": 0.0189208984375, "distill_loss": 0.0791015625, "epoch": 1.8678915135608047, "explain_loss": 0.94921875, "loss": 1.3515625, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.002899169921875, "step": 4270 }, { "epoch": 1.8722659667541557, "grad_norm": 0.6800698041915894, "learning_rate": 6.786271256718268e-05, "loss": 1.761279296875, "step": 4280 }, { "ce_loss": 0.00982666015625, "distill_loss": 0.1005859375, "epoch": 1.8722659667541557, "explain_loss": 1.203125, "loss": 1.671875, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0025634765625, "step": 4280 }, { "ce_loss": 0.3125, "distill_loss": 0.08056640625, "epoch": 1.8722659667541557, "explain_loss": 0.875, "loss": 1.859375, "ref_ce_loss": 0.53125, "stage_align_loss": 0.058349609375, "step": 4280 }, { "ce_loss": 0.29296875, "distill_loss": 0.095703125, "epoch": 1.8722659667541557, "explain_loss": 1.28125, "loss": 2.015625, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.05810546875, "step": 4280 }, { "ce_loss": 0.01318359375, "distill_loss": 0.0654296875, "epoch": 1.8722659667541557, "explain_loss": 0.9375, "loss": 1.328125, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.002349853515625, "step": 4280 }, { "epoch": 1.8766404199475066, "grad_norm": 0.7536885738372803, "learning_rate": 6.740645869556284e-05, "loss": 1.7591796875, "step": 4290 }, { "ce_loss": 0.0021209716796875, "distill_loss": 0.126953125, "epoch": 1.8766404199475066, "explain_loss": 1.3203125, "loss": 1.90625, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.0019989013671875, "step": 4290 }, { "ce_loss": 0.047607421875, "distill_loss": 0.091796875, "epoch": 1.8766404199475066, "explain_loss": 1.21875, "loss": 1.921875, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.00885009765625, "step": 4290 }, { "ce_loss": 0.015380859375, "distill_loss": 0.0888671875, "epoch": 1.8766404199475066, "explain_loss": 0.9453125, "loss": 1.421875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.0024566650390625, "step": 4290 }, { "ce_loss": 0.10107421875, "distill_loss": 0.16796875, "epoch": 1.8766404199475066, "explain_loss": 0.859375, "loss": 1.65625, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.0059814453125, "step": 4290 }, { "epoch": 1.8810148731408574, "grad_norm": 0.6638011932373047, "learning_rate": 6.695096270597886e-05, "loss": 1.74736328125, "step": 4300 }, { "ce_loss": 0.10205078125, "distill_loss": 0.091796875, "epoch": 1.8810148731408574, "explain_loss": 1.0703125, "loss": 2.015625, "ref_ce_loss": 0.6484375, "stage_align_loss": 0.10595703125, "step": 4300 }, { "ce_loss": 0.027099609375, "distill_loss": 0.0830078125, "epoch": 1.8810148731408574, "explain_loss": 0.890625, "loss": 1.3984375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0027008056640625, "step": 4300 }, { "ce_loss": 0.0159912109375, "distill_loss": 0.062255859375, "epoch": 1.8810148731408574, "explain_loss": 0.953125, "loss": 1.421875, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.00433349609375, "step": 4300 }, { "ce_loss": 0.008544921875, "distill_loss": 0.0654296875, "epoch": 1.8810148731408574, "explain_loss": 1.2109375, "loss": 1.78125, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.003936767578125, "step": 4300 }, { "epoch": 1.885389326334208, "grad_norm": 0.8736134767532349, "learning_rate": 6.649623518986133e-05, "loss": 1.69921875, "step": 4310 }, { "ce_loss": 0.046630859375, "distill_loss": 0.06396484375, "epoch": 1.885389326334208, "explain_loss": 1.0703125, "loss": 1.6171875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.003387451171875, "step": 4310 }, { "ce_loss": 0.051025390625, "distill_loss": 0.083984375, "epoch": 1.885389326334208, "explain_loss": 0.796875, "loss": 1.3359375, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.004119873046875, "step": 4310 }, { "ce_loss": 0.004608154296875, "distill_loss": 0.1044921875, "epoch": 1.885389326334208, "explain_loss": 1.234375, "loss": 1.8203125, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.00274658203125, "step": 4310 }, { "ce_loss": 0.0172119140625, "distill_loss": 0.095703125, "epoch": 1.885389326334208, "explain_loss": 0.703125, "loss": 1.1015625, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.003082275390625, "step": 4310 }, { "epoch": 1.889763779527559, "grad_norm": 0.8447951674461365, "learning_rate": 6.604228672077175e-05, "loss": 1.7732421875, "step": 4320 }, { "ce_loss": 0.23828125, "distill_loss": 0.0859375, "epoch": 1.889763779527559, "explain_loss": 1.0, "loss": 1.78125, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.0250244140625, "step": 4320 }, { "ce_loss": 0.71484375, "distill_loss": 0.0673828125, "epoch": 1.889763779527559, "explain_loss": 1.0234375, "loss": 2.171875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.002899169921875, "step": 4320 }, { "ce_loss": 0.0123291015625, "distill_loss": 0.07421875, "epoch": 1.889763779527559, "explain_loss": 1.1953125, "loss": 1.5859375, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.01324462890625, "step": 4320 }, { "ce_loss": 0.3203125, "distill_loss": 0.09326171875, "epoch": 1.889763779527559, "explain_loss": 0.90625, "loss": 1.734375, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.072265625, "step": 4320 }, { "epoch": 1.89413823272091, "grad_norm": 1.0848275423049927, "learning_rate": 6.558912785415693e-05, "loss": 1.74072265625, "step": 4330 }, { "ce_loss": 0.048583984375, "distill_loss": 0.087890625, "epoch": 1.89413823272091, "explain_loss": 1.328125, "loss": 1.921875, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.0023345947265625, "step": 4330 }, { "ce_loss": 0.002593994140625, "distill_loss": 0.1044921875, "epoch": 1.89413823272091, "explain_loss": 1.078125, "loss": 1.5, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.0020599365234375, "step": 4330 }, { "ce_loss": 0.458984375, "distill_loss": 0.09765625, "epoch": 1.89413823272091, "explain_loss": 0.875, "loss": 1.921875, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0986328125, "step": 4330 }, { "ce_loss": 1.1015625, "distill_loss": 0.07177734375, "epoch": 1.89413823272091, "explain_loss": 0.953125, "loss": 2.46875, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.0030975341796875, "step": 4330 }, { "epoch": 1.8985126859142607, "grad_norm": 0.8560370802879333, "learning_rate": 6.513676912710334e-05, "loss": 1.80986328125, "step": 4340 }, { "ce_loss": 0.12255859375, "distill_loss": 0.1259765625, "epoch": 1.8985126859142607, "explain_loss": 1.140625, "loss": 1.8125, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.00665283203125, "step": 4340 }, { "ce_loss": 0.005157470703125, "distill_loss": 0.21484375, "epoch": 1.8985126859142607, "explain_loss": 1.2421875, "loss": 1.84375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.0026092529296875, "step": 4340 }, { "ce_loss": 0.1123046875, "distill_loss": 0.06298828125, "epoch": 1.8985126859142607, "explain_loss": 1.125, "loss": 1.9375, "ref_ce_loss": 0.6328125, "stage_align_loss": 0.0032806396484375, "step": 4340 }, { "ce_loss": 0.451171875, "distill_loss": 0.15625, "epoch": 1.8985126859142607, "explain_loss": 0.90625, "loss": 2.0625, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.0032196044921875, "step": 4340 }, { "epoch": 1.9028871391076114, "grad_norm": 0.9078378081321716, "learning_rate": 6.468522105809228e-05, "loss": 1.83193359375, "step": 4350 }, { "ce_loss": 0.189453125, "distill_loss": 0.1142578125, "epoch": 1.9028871391076114, "explain_loss": 1.3515625, "loss": 2.03125, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.03271484375, "step": 4350 }, { "ce_loss": 0.50390625, "distill_loss": 0.078125, "epoch": 1.9028871391076114, "explain_loss": 1.375, "loss": 2.375, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0220947265625, "step": 4350 }, { "ce_loss": 0.068359375, "distill_loss": 0.095703125, "epoch": 1.9028871391076114, "explain_loss": 1.125, "loss": 1.65625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.002777099609375, "step": 4350 }, { "ce_loss": 1.3125, "distill_loss": 0.07421875, "epoch": 1.9028871391076114, "explain_loss": 0.9140625, "loss": 2.71875, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.08740234375, "step": 4350 }, { "epoch": 1.9072615923009624, "grad_norm": 0.9439929127693176, "learning_rate": 6.423449414675509e-05, "loss": 1.7998046875, "step": 4360 }, { "ce_loss": 0.5078125, "distill_loss": 0.07177734375, "epoch": 1.9072615923009624, "explain_loss": 1.09375, "loss": 2.0, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.0296630859375, "step": 4360 }, { "ce_loss": 0.03759765625, "distill_loss": 0.1181640625, "epoch": 1.9072615923009624, "explain_loss": 1.15625, "loss": 1.703125, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.00244140625, "step": 4360 }, { "ce_loss": 0.2099609375, "distill_loss": 0.06689453125, "epoch": 1.9072615923009624, "explain_loss": 0.921875, "loss": 1.65625, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.01092529296875, "step": 4360 }, { "ce_loss": 0.0157470703125, "distill_loss": 0.1357421875, "epoch": 1.9072615923009624, "explain_loss": 0.98828125, "loss": 1.5625, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0023956298828125, "step": 4360 }, { "epoch": 1.9116360454943133, "grad_norm": 0.7410246729850769, "learning_rate": 6.378459887362922e-05, "loss": 1.78623046875, "step": 4370 }, { "ce_loss": 0.1650390625, "distill_loss": 0.1435546875, "epoch": 1.9116360454943133, "explain_loss": 0.9921875, "loss": 1.796875, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0830078125, "step": 4370 }, { "ce_loss": 0.75, "distill_loss": 0.099609375, "epoch": 1.9116360454943133, "explain_loss": 0.8984375, "loss": 2.125, "ref_ce_loss": 0.359375, "stage_align_loss": 0.00860595703125, "step": 4370 }, { "ce_loss": 0.21484375, "distill_loss": 0.146484375, "epoch": 1.9116360454943133, "explain_loss": 0.890625, "loss": 1.78125, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.08349609375, "step": 4370 }, { "ce_loss": 0.12890625, "distill_loss": 0.1015625, "epoch": 1.9116360454943133, "explain_loss": 1.0390625, "loss": 1.71875, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.004425048828125, "step": 4370 }, { "epoch": 1.916010498687664, "grad_norm": 0.7117178440093994, "learning_rate": 6.333554569991442e-05, "loss": 1.7283203125, "step": 4380 }, { "ce_loss": 0.1923828125, "distill_loss": 0.07177734375, "epoch": 1.916010498687664, "explain_loss": 0.765625, "loss": 1.5, "ref_ce_loss": 0.46875, "stage_align_loss": 0.0031280517578125, "step": 4380 }, { "ce_loss": 0.0159912109375, "distill_loss": 0.09814453125, "epoch": 1.916010498687664, "explain_loss": 0.9453125, "loss": 1.4375, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.002227783203125, "step": 4380 }, { "ce_loss": 0.08251953125, "distill_loss": 0.076171875, "epoch": 1.916010498687664, "explain_loss": 0.76171875, "loss": 1.296875, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.00396728515625, "step": 4380 }, { "ce_loss": 0.030029296875, "distill_loss": 0.09326171875, "epoch": 1.916010498687664, "explain_loss": 1.0703125, "loss": 1.5078125, "ref_ce_loss": 0.3125, "stage_align_loss": 0.00323486328125, "step": 4380 }, { "epoch": 1.9203849518810148, "grad_norm": 0.7916120290756226, "learning_rate": 6.288734506722947e-05, "loss": 1.705859375, "step": 4390 }, { "ce_loss": 0.1767578125, "distill_loss": 0.08837890625, "epoch": 1.9203849518810148, "explain_loss": 1.0078125, "loss": 1.71875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.038330078125, "step": 4390 }, { "ce_loss": 0.0159912109375, "distill_loss": 0.0703125, "epoch": 1.9203849518810148, "explain_loss": 0.7578125, "loss": 1.21875, "ref_ce_loss": 0.375, "stage_align_loss": 0.0020294189453125, "step": 4390 }, { "ce_loss": 0.000942230224609375, "distill_loss": 0.0947265625, "epoch": 1.9203849518810148, "explain_loss": 0.953125, "loss": 1.328125, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.002349853515625, "step": 4390 }, { "ce_loss": 0.013671875, "distill_loss": 0.0859375, "epoch": 1.9203849518810148, "explain_loss": 1.015625, "loss": 1.4375, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0024261474609375, "step": 4390 }, { "epoch": 1.9247594050743657, "grad_norm": 0.8493144512176514, "learning_rate": 6.244000739736942e-05, "loss": 1.7166015625, "step": 4400 }, { "ce_loss": 0.04736328125, "distill_loss": 0.1025390625, "epoch": 1.9247594050743657, "explain_loss": 0.9375, "loss": 1.625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.00262451171875, "step": 4400 }, { "ce_loss": 0.01611328125, "distill_loss": 0.09326171875, "epoch": 1.9247594050743657, "explain_loss": 1.1015625, "loss": 1.625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0034027099609375, "step": 4400 }, { "ce_loss": 0.06396484375, "distill_loss": 0.07861328125, "epoch": 1.9247594050743657, "explain_loss": 1.0078125, "loss": 1.5625, "ref_ce_loss": 0.40625, "stage_align_loss": 0.00360107421875, "step": 4400 }, { "ce_loss": 0.431640625, "distill_loss": 0.10693359375, "epoch": 1.9247594050743657, "explain_loss": 1.21875, "loss": 2.25, "ref_ce_loss": 0.484375, "stage_align_loss": 0.008544921875, "step": 4400 }, { "epoch": 1.9291338582677167, "grad_norm": 0.5076186656951904, "learning_rate": 6.199354309206329e-05, "loss": 1.75615234375, "step": 4410 }, { "ce_loss": 1.046875, "distill_loss": 0.10009765625, "epoch": 1.9291338582677167, "explain_loss": 1.0703125, "loss": 2.703125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.056640625, "step": 4410 }, { "ce_loss": 0.1318359375, "distill_loss": 0.1181640625, "epoch": 1.9291338582677167, "explain_loss": 1.0625, "loss": 1.65625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.01171875, "step": 4410 }, { "ce_loss": 0.2412109375, "distill_loss": 0.0703125, "epoch": 1.9291338582677167, "explain_loss": 0.7578125, "loss": 1.34375, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.0036163330078125, "step": 4410 }, { "ce_loss": 0.0028533935546875, "distill_loss": 0.10791015625, "epoch": 1.9291338582677167, "explain_loss": 1.1171875, "loss": 1.578125, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0024261474609375, "step": 4410 }, { "epoch": 1.9335083114610674, "grad_norm": 1.622485637664795, "learning_rate": 6.15479625327322e-05, "loss": 1.74501953125, "step": 4420 }, { "ce_loss": 0.130859375, "distill_loss": 0.09375, "epoch": 1.9335083114610674, "explain_loss": 0.84375, "loss": 1.3671875, "ref_ce_loss": 0.236328125, "stage_align_loss": 0.061767578125, "step": 4420 }, { "ce_loss": 0.1357421875, "distill_loss": 0.0869140625, "epoch": 1.9335083114610674, "explain_loss": 1.0859375, "loss": 1.703125, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0108642578125, "step": 4420 }, { "ce_loss": 0.033203125, "distill_loss": 0.1123046875, "epoch": 1.9335083114610674, "explain_loss": 1.09375, "loss": 1.6171875, "ref_ce_loss": 0.375, "stage_align_loss": 0.00408935546875, "step": 4420 }, { "ce_loss": 0.00909423828125, "distill_loss": 0.12158203125, "epoch": 1.9335083114610674, "explain_loss": 0.8359375, "loss": 1.453125, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.0020751953125, "step": 4420 }, { "epoch": 1.937882764654418, "grad_norm": 0.6617094278335571, "learning_rate": 6.110327608024779e-05, "loss": 1.828125, "step": 4430 }, { "ce_loss": 0.051513671875, "distill_loss": 0.1181640625, "epoch": 1.937882764654418, "explain_loss": 0.9375, "loss": 1.53125, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.01055908203125, "step": 4430 }, { "ce_loss": 0.126953125, "distill_loss": 0.09375, "epoch": 1.937882764654418, "explain_loss": 1.171875, "loss": 1.8359375, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.009521484375, "step": 4430 }, { "ce_loss": 0.1455078125, "distill_loss": 0.076171875, "epoch": 1.937882764654418, "explain_loss": 0.828125, "loss": 1.484375, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.003082275390625, "step": 4430 }, { "ce_loss": 0.091796875, "distill_loss": 0.07666015625, "epoch": 1.937882764654418, "explain_loss": 1.03125, "loss": 1.53125, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.01220703125, "step": 4430 }, { "epoch": 1.942257217847769, "grad_norm": 1.0149989128112793, "learning_rate": 6.0659494074691705e-05, "loss": 1.803564453125, "step": 4440 }, { "ce_loss": 0.1435546875, "distill_loss": 0.08544921875, "epoch": 1.942257217847769, "explain_loss": 0.7734375, "loss": 1.234375, "ref_ce_loss": 0.23046875, "stage_align_loss": 0.0025634765625, "step": 4440 }, { "ce_loss": 0.00469970703125, "distill_loss": 0.109375, "epoch": 1.942257217847769, "explain_loss": 1.125, "loss": 1.59375, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.002838134765625, "step": 4440 }, { "ce_loss": 0.0556640625, "distill_loss": 0.103515625, "epoch": 1.942257217847769, "explain_loss": 1.328125, "loss": 1.859375, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.00274658203125, "step": 4440 }, { "ce_loss": 0.287109375, "distill_loss": 0.058837890625, "epoch": 1.942257217847769, "explain_loss": 1.296875, "loss": 1.875, "ref_ce_loss": 0.20703125, "stage_align_loss": 0.021240234375, "step": 4440 }, { "epoch": 1.94663167104112, "grad_norm": 4.375446796417236, "learning_rate": 6.0216626835114685e-05, "loss": 1.80615234375, "step": 4450 }, { "ce_loss": 0.0166015625, "distill_loss": 0.0712890625, "epoch": 1.94663167104112, "explain_loss": 0.75, "loss": 1.234375, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.00250244140625, "step": 4450 }, { "ce_loss": 0.037109375, "distill_loss": 0.064453125, "epoch": 1.94663167104112, "explain_loss": 1.15625, "loss": 1.65625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.00360107421875, "step": 4450 }, { "ce_loss": 0.09765625, "distill_loss": 0.0830078125, "epoch": 1.94663167104112, "explain_loss": 1.125, "loss": 1.7109375, "ref_ce_loss": 0.390625, "stage_align_loss": 0.016357421875, "step": 4450 }, { "ce_loss": 0.1884765625, "distill_loss": 0.107421875, "epoch": 1.94663167104112, "explain_loss": 0.94140625, "loss": 1.6171875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.004364013671875, "step": 4450 }, { "epoch": 1.9510061242344707, "grad_norm": 0.810792863368988, "learning_rate": 5.977468465929703e-05, "loss": 1.70361328125, "step": 4460 }, { "ce_loss": 0.00439453125, "distill_loss": 0.0810546875, "epoch": 1.9510061242344707, "explain_loss": 1.1015625, "loss": 1.71875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0031890869140625, "step": 4460 }, { "ce_loss": 0.1337890625, "distill_loss": 0.0849609375, "epoch": 1.9510061242344707, "explain_loss": 0.9921875, "loss": 1.5078125, "ref_ce_loss": 0.28125, "stage_align_loss": 0.01422119140625, "step": 4460 }, { "ce_loss": 0.1708984375, "distill_loss": 0.0751953125, "epoch": 1.9510061242344707, "explain_loss": 1.0625, "loss": 1.625, "ref_ce_loss": 0.26171875, "stage_align_loss": 0.0517578125, "step": 4460 }, { "ce_loss": 0.279296875, "distill_loss": 0.1650390625, "epoch": 1.9510061242344707, "explain_loss": 1.0234375, "loss": 1.96875, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.051025390625, "step": 4460 }, { "epoch": 1.9553805774278215, "grad_norm": 0.9543859362602234, "learning_rate": 5.933367782350885e-05, "loss": 1.77783203125, "step": 4470 }, { "ce_loss": 1.15625, "distill_loss": 0.080078125, "epoch": 1.9553805774278215, "explain_loss": 1.2734375, "loss": 3.03125, "ref_ce_loss": 0.5, "stage_align_loss": 0.01312255859375, "step": 4470 }, { "ce_loss": 0.057373046875, "distill_loss": 0.08349609375, "epoch": 1.9553805774278215, "explain_loss": 1.2421875, "loss": 1.875, "ref_ce_loss": 0.4921875, "stage_align_loss": 0.0026702880859375, "step": 4470 }, { "ce_loss": 0.1083984375, "distill_loss": 0.076171875, "epoch": 1.9553805774278215, "explain_loss": 0.76953125, "loss": 1.2734375, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.0031280517578125, "step": 4470 }, { "ce_loss": 0.177734375, "distill_loss": 0.08837890625, "epoch": 1.9553805774278215, "explain_loss": 1.203125, "loss": 1.84375, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.00238037109375, "step": 4470 }, { "epoch": 1.9597550306211724, "grad_norm": 0.6674023866653442, "learning_rate": 5.8893616582271324e-05, "loss": 1.74697265625, "step": 4480 }, { "ce_loss": 0.0137939453125, "distill_loss": 0.0869140625, "epoch": 1.9597550306211724, "explain_loss": 1.34375, "loss": 1.9921875, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0035247802734375, "step": 4480 }, { "ce_loss": 0.0026702880859375, "distill_loss": 0.107421875, "epoch": 1.9597550306211724, "explain_loss": 0.9296875, "loss": 1.421875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.0024871826171875, "step": 4480 }, { "ce_loss": 0.006317138671875, "distill_loss": 0.091796875, "epoch": 1.9597550306211724, "explain_loss": 0.91015625, "loss": 1.328125, "ref_ce_loss": 0.3125, "stage_align_loss": 0.004058837890625, "step": 4480 }, { "ce_loss": 0.01171875, "distill_loss": 0.07421875, "epoch": 1.9597550306211724, "explain_loss": 1.0703125, "loss": 1.46875, "ref_ce_loss": 0.3125, "stage_align_loss": 0.00274658203125, "step": 4480 }, { "epoch": 1.9641294838145233, "grad_norm": 0.8517446517944336, "learning_rate": 5.845451116811811e-05, "loss": 1.7388671875, "step": 4490 }, { "ce_loss": 0.00714111328125, "distill_loss": 0.15625, "epoch": 1.9641294838145233, "explain_loss": 1.203125, "loss": 1.734375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0026092529296875, "step": 4490 }, { "ce_loss": 0.01092529296875, "distill_loss": 0.08251953125, "epoch": 1.9641294838145233, "explain_loss": 1.0625, "loss": 1.53125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.00323486328125, "step": 4490 }, { "ce_loss": 0.03857421875, "distill_loss": 0.08251953125, "epoch": 1.9641294838145233, "explain_loss": 1.59375, "loss": 2.296875, "ref_ce_loss": 0.578125, "stage_align_loss": 0.0032806396484375, "step": 4490 }, { "ce_loss": 0.3828125, "distill_loss": 0.083984375, "epoch": 1.9641294838145233, "explain_loss": 1.515625, "loss": 2.421875, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.004425048828125, "step": 4490 }, { "epoch": 1.968503937007874, "grad_norm": 0.9706295132637024, "learning_rate": 5.801637179135757e-05, "loss": 1.734228515625, "step": 4500 }, { "ce_loss": 0.07568359375, "distill_loss": 0.0966796875, "epoch": 1.968503937007874, "explain_loss": 1.15625, "loss": 1.8984375, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0849609375, "step": 4500 }, { "ce_loss": 0.00811767578125, "distill_loss": 0.1015625, "epoch": 1.968503937007874, "explain_loss": 1.2109375, "loss": 1.703125, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.0027008056640625, "step": 4500 }, { "ce_loss": 0.11767578125, "distill_loss": 0.07080078125, "epoch": 1.968503937007874, "explain_loss": 0.9140625, "loss": 1.421875, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.01495361328125, "step": 4500 }, { "ce_loss": 0.6171875, "distill_loss": 0.06787109375, "epoch": 1.968503937007874, "explain_loss": 0.59765625, "loss": 1.8125, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.05224609375, "step": 4500 }, { "epoch": 1.9728783902012248, "grad_norm": 0.9174341559410095, "learning_rate": 5.757920863983509e-05, "loss": 1.690234375, "step": 4510 }, { "ce_loss": 0.0255126953125, "distill_loss": 0.06298828125, "epoch": 1.9728783902012248, "explain_loss": 1.4609375, "loss": 1.9765625, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.00244140625, "step": 4510 }, { "ce_loss": 0.00982666015625, "distill_loss": 0.0791015625, "epoch": 1.9728783902012248, "explain_loss": 0.96875, "loss": 1.5, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.0024871826171875, "step": 4510 }, { "ce_loss": 0.0238037109375, "distill_loss": 0.0732421875, "epoch": 1.9728783902012248, "explain_loss": 1.09375, "loss": 1.453125, "ref_ce_loss": 0.25390625, "stage_align_loss": 0.0042724609375, "step": 4510 }, { "ce_loss": 0.047607421875, "distill_loss": 0.0947265625, "epoch": 1.9728783902012248, "explain_loss": 1.1640625, "loss": 1.609375, "ref_ce_loss": 0.296875, "stage_align_loss": 0.003631591796875, "step": 4510 }, { "epoch": 1.9772528433945755, "grad_norm": 0.9726318717002869, "learning_rate": 5.7143031878696526e-05, "loss": 1.7185546875, "step": 4520 }, { "ce_loss": 0.00078582763671875, "distill_loss": 0.1484375, "epoch": 1.9772528433945755, "explain_loss": 1.109375, "loss": 1.6328125, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.00286865234375, "step": 4520 }, { "ce_loss": 0.26953125, "distill_loss": 0.0693359375, "epoch": 1.9772528433945755, "explain_loss": 1.125, "loss": 1.8515625, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.004180908203125, "step": 4520 }, { "ce_loss": 0.53515625, "distill_loss": 0.0859375, "epoch": 1.9772528433945755, "explain_loss": 1.125, "loss": 2.296875, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.125, "step": 4520 }, { "ce_loss": 0.00079345703125, "distill_loss": 0.09326171875, "epoch": 1.9772528433945755, "explain_loss": 1.53125, "loss": 2.0625, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0026397705078125, "step": 4520 }, { "epoch": 1.9816272965879265, "grad_norm": 0.8462216258049011, "learning_rate": 5.670785165015169e-05, "loss": 1.79375, "step": 4530 }, { "ce_loss": 0.04638671875, "distill_loss": 0.0888671875, "epoch": 1.9816272965879265, "explain_loss": 1.140625, "loss": 1.7734375, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0150146484375, "step": 4530 }, { "ce_loss": 0.0019073486328125, "distill_loss": 0.1103515625, "epoch": 1.9816272965879265, "explain_loss": 1.125, "loss": 1.703125, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.0028228759765625, "step": 4530 }, { "ce_loss": 0.447265625, "distill_loss": 0.10693359375, "epoch": 1.9816272965879265, "explain_loss": 1.359375, "loss": 2.515625, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0654296875, "step": 4530 }, { "ce_loss": 0.09814453125, "distill_loss": 0.09033203125, "epoch": 1.9816272965879265, "explain_loss": 1.1875, "loss": 1.75, "ref_ce_loss": 0.375, "stage_align_loss": 0.0033111572265625, "step": 4530 }, { "epoch": 1.9860017497812774, "grad_norm": 0.7776380777359009, "learning_rate": 5.627367807323826e-05, "loss": 1.76650390625, "step": 4540 }, { "ce_loss": 0.240234375, "distill_loss": 0.083984375, "epoch": 1.9860017497812774, "explain_loss": 1.21875, "loss": 2.03125, "ref_ce_loss": 0.4375, "stage_align_loss": 0.046875, "step": 4540 }, { "ce_loss": 0.451171875, "distill_loss": 0.10986328125, "epoch": 1.9860017497812774, "explain_loss": 1.2109375, "loss": 2.40625, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.016357421875, "step": 4540 }, { "ce_loss": 0.00872802734375, "distill_loss": 0.11279296875, "epoch": 1.9860017497812774, "explain_loss": 1.21875, "loss": 1.7578125, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.004547119140625, "step": 4540 }, { "ce_loss": 0.091796875, "distill_loss": 0.1064453125, "epoch": 1.9860017497812774, "explain_loss": 1.4453125, "loss": 2.375, "ref_ce_loss": 0.68359375, "stage_align_loss": 0.051025390625, "step": 4540 }, { "epoch": 1.9903762029746281, "grad_norm": 0.9909000396728516, "learning_rate": 5.584052124358698e-05, "loss": 1.78984375, "step": 4550 }, { "ce_loss": 0.032470703125, "distill_loss": 0.083984375, "epoch": 1.9903762029746281, "explain_loss": 1.1875, "loss": 1.765625, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.062255859375, "step": 4550 }, { "ce_loss": 0.0185546875, "distill_loss": 0.12451171875, "epoch": 1.9903762029746281, "explain_loss": 1.203125, "loss": 1.765625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00323486328125, "step": 4550 }, { "ce_loss": 0.52734375, "distill_loss": 0.0810546875, "epoch": 1.9903762029746281, "explain_loss": 0.921875, "loss": 1.8125, "ref_ce_loss": 0.2451171875, "stage_align_loss": 0.040283203125, "step": 4550 }, { "ce_loss": 0.1005859375, "distill_loss": 0.0732421875, "epoch": 1.9903762029746281, "explain_loss": 1.484375, "loss": 1.859375, "ref_ce_loss": 0.19921875, "stage_align_loss": 0.00238037109375, "step": 4550 }, { "epoch": 1.9947506561679789, "grad_norm": 0.9932711720466614, "learning_rate": 5.540839123318653e-05, "loss": 1.84658203125, "step": 4560 }, { "ce_loss": 0.02294921875, "distill_loss": 0.08544921875, "epoch": 1.9947506561679789, "explain_loss": 1.2109375, "loss": 1.78125, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.003082275390625, "step": 4560 }, { "ce_loss": 0.02294921875, "distill_loss": 0.103515625, "epoch": 1.9947506561679789, "explain_loss": 1.203125, "loss": 1.6875, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.0021209716796875, "step": 4560 }, { "ce_loss": 0.0086669921875, "distill_loss": 0.1181640625, "epoch": 1.9947506561679789, "explain_loss": 1.09375, "loss": 1.7109375, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0023956298828125, "step": 4560 }, { "ce_loss": 0.01007080078125, "distill_loss": 0.1005859375, "epoch": 1.9947506561679789, "explain_loss": 1.3125, "loss": 1.71875, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0021209716796875, "step": 4560 }, { "epoch": 1.9991251093613298, "grad_norm": 0.6329666972160339, "learning_rate": 5.497729809014949e-05, "loss": 1.72001953125, "step": 4570 }, { "ce_loss": 0.056884765625, "distill_loss": 0.0908203125, "epoch": 1.9991251093613298, "explain_loss": 1.1796875, "loss": 1.6484375, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.00390625, "step": 4570 }, { "ce_loss": 0.053955078125, "distill_loss": 0.09228515625, "epoch": 1.9991251093613298, "explain_loss": 0.984375, "loss": 1.4765625, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.003143310546875, "step": 4570 }, { "ce_loss": 0.046630859375, "distill_loss": 0.0693359375, "epoch": 1.9991251093613298, "explain_loss": 0.78125, "loss": 1.296875, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0030517578125, "step": 4570 }, { "ce_loss": 0.0255126953125, "distill_loss": 0.103515625, "epoch": 1.9991251093613298, "explain_loss": 0.84375, "loss": 1.359375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.00750732421875, "step": 4570 }, { "epoch": 2.0034995625546808, "grad_norm": 0.823514997959137, "learning_rate": 5.45472518384786e-05, "loss": 1.6404296875, "step": 4580 }, { "ce_loss": 0.062255859375, "distill_loss": 0.083984375, "epoch": 2.0034995625546808, "explain_loss": 0.734375, "loss": 1.078125, "ref_ce_loss": 0.1943359375, "stage_align_loss": 0.00396728515625, "step": 4580 }, { "ce_loss": 0.00482177734375, "distill_loss": 0.107421875, "epoch": 2.0034995625546808, "explain_loss": 1.0390625, "loss": 1.40625, "ref_ce_loss": 0.25390625, "stage_align_loss": 0.0022430419921875, "step": 4580 }, { "ce_loss": 1.0703125, "distill_loss": 0.087890625, "epoch": 2.0034995625546808, "explain_loss": 1.109375, "loss": 2.875, "ref_ce_loss": 0.55078125, "stage_align_loss": 0.05908203125, "step": 4580 }, { "ce_loss": 0.002227783203125, "distill_loss": 0.072265625, "epoch": 2.0034995625546808, "explain_loss": 1.171875, "loss": 1.578125, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.002532958984375, "step": 4580 }, { "epoch": 2.0078740157480315, "grad_norm": 0.5488307476043701, "learning_rate": 5.4118262477833846e-05, "loss": 1.6693359375, "step": 4590 }, { "ce_loss": 0.7109375, "distill_loss": 0.0927734375, "epoch": 2.0078740157480315, "explain_loss": 1.0625, "loss": 2.28125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.05712890625, "step": 4590 }, { "ce_loss": 0.018798828125, "distill_loss": 0.0908203125, "epoch": 2.0078740157480315, "explain_loss": 1.046875, "loss": 1.46875, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0028228759765625, "step": 4590 }, { "ce_loss": 0.1572265625, "distill_loss": 0.0693359375, "epoch": 2.0078740157480315, "explain_loss": 1.171875, "loss": 1.796875, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.0031280517578125, "step": 4590 }, { "ce_loss": 0.50390625, "distill_loss": 0.091796875, "epoch": 2.0078740157480315, "explain_loss": 1.546875, "loss": 2.546875, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0556640625, "step": 4590 }, { "epoch": 2.012248468941382, "grad_norm": 0.5521101355552673, "learning_rate": 5.3690339983299755e-05, "loss": 1.67353515625, "step": 4600 }, { "ce_loss": 0.0888671875, "distill_loss": 0.080078125, "epoch": 2.012248468941382, "explain_loss": 1.1015625, "loss": 1.8203125, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.048583984375, "step": 4600 }, { "ce_loss": 0.006134033203125, "distill_loss": 0.109375, "epoch": 2.012248468941382, "explain_loss": 1.109375, "loss": 1.5859375, "ref_ce_loss": 0.359375, "stage_align_loss": 0.003021240234375, "step": 4600 }, { "ce_loss": 0.027099609375, "distill_loss": 0.10791015625, "epoch": 2.012248468941382, "explain_loss": 1.09375, "loss": 1.5859375, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.0029754638671875, "step": 4600 }, { "ce_loss": 0.076171875, "distill_loss": 0.0859375, "epoch": 2.012248468941382, "explain_loss": 1.265625, "loss": 1.78125, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.004608154296875, "step": 4600 }, { "epoch": 2.0166229221347334, "grad_norm": 0.5843865871429443, "learning_rate": 5.326349430515352e-05, "loss": 1.68798828125, "step": 4610 }, { "ce_loss": 0.005706787109375, "distill_loss": 0.076171875, "epoch": 2.0166229221347334, "explain_loss": 1.2578125, "loss": 1.6953125, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0025787353515625, "step": 4610 }, { "ce_loss": 0.00152587890625, "distill_loss": 0.07861328125, "epoch": 2.0166229221347334, "explain_loss": 1.21875, "loss": 1.625, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.002899169921875, "step": 4610 }, { "ce_loss": 0.01806640625, "distill_loss": 0.0830078125, "epoch": 2.0166229221347334, "explain_loss": 0.8046875, "loss": 1.140625, "ref_ce_loss": 0.2353515625, "stage_align_loss": 0.0029754638671875, "step": 4610 }, { "ce_loss": 0.0478515625, "distill_loss": 0.1005859375, "epoch": 2.0166229221347334, "explain_loss": 1.09375, "loss": 1.71875, "ref_ce_loss": 0.466796875, "stage_align_loss": 0.00238037109375, "step": 4610 }, { "epoch": 2.020997375328084, "grad_norm": 0.5991840958595276, "learning_rate": 5.283773536863363e-05, "loss": 1.57236328125, "step": 4620 }, { "ce_loss": 0.05126953125, "distill_loss": 0.07763671875, "epoch": 2.020997375328084, "explain_loss": 0.9921875, "loss": 1.5, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00457763671875, "step": 4620 }, { "ce_loss": 0.06591796875, "distill_loss": 0.0693359375, "epoch": 2.020997375328084, "explain_loss": 0.890625, "loss": 1.328125, "ref_ce_loss": 0.255859375, "stage_align_loss": 0.043212890625, "step": 4620 }, { "ce_loss": 0.01226806640625, "distill_loss": 0.09814453125, "epoch": 2.020997375328084, "explain_loss": 0.625, "loss": 0.99609375, "ref_ce_loss": 0.2578125, "stage_align_loss": 0.004852294921875, "step": 4620 }, { "ce_loss": 0.3359375, "distill_loss": 0.06494140625, "epoch": 2.020997375328084, "explain_loss": 0.765625, "loss": 1.671875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0032806396484375, "step": 4620 }, { "epoch": 2.025371828521435, "grad_norm": 1.273041009902954, "learning_rate": 5.241307307370922e-05, "loss": 1.592822265625, "step": 4630 }, { "ce_loss": 0.006103515625, "distill_loss": 0.064453125, "epoch": 2.025371828521435, "explain_loss": 0.9765625, "loss": 1.25, "ref_ce_loss": 0.1982421875, "stage_align_loss": 0.0032958984375, "step": 4630 }, { "ce_loss": 0.0380859375, "distill_loss": 0.08544921875, "epoch": 2.025371828521435, "explain_loss": 0.90625, "loss": 1.34375, "ref_ce_loss": 0.310546875, "stage_align_loss": 0.00408935546875, "step": 4630 }, { "ce_loss": 0.002685546875, "distill_loss": 0.07080078125, "epoch": 2.025371828521435, "explain_loss": 1.078125, "loss": 1.4453125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0037994384765625, "step": 4630 }, { "ce_loss": 0.05322265625, "distill_loss": 0.08984375, "epoch": 2.025371828521435, "explain_loss": 0.8671875, "loss": 1.40625, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0033416748046875, "step": 4630 }, { "epoch": 2.0297462817147855, "grad_norm": 0.8125293254852295, "learning_rate": 5.1989517294849554e-05, "loss": 1.63056640625, "step": 4640 }, { "ce_loss": 0.05419921875, "distill_loss": 0.0654296875, "epoch": 2.0297462817147855, "explain_loss": 0.80859375, "loss": 1.1875, "ref_ce_loss": 0.26171875, "stage_align_loss": 0.0028228759765625, "step": 4640 }, { "ce_loss": 0.08740234375, "distill_loss": 0.099609375, "epoch": 2.0297462817147855, "explain_loss": 0.98828125, "loss": 1.6875, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0036163330078125, "step": 4640 }, { "ce_loss": 0.0250244140625, "distill_loss": 0.1162109375, "epoch": 2.0297462817147855, "explain_loss": 1.375, "loss": 1.828125, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.0048828125, "step": 4640 }, { "ce_loss": 0.046875, "distill_loss": 0.07421875, "epoch": 2.0297462817147855, "explain_loss": 0.921875, "loss": 1.78125, "ref_ce_loss": 0.66015625, "stage_align_loss": 0.0830078125, "step": 4640 }, { "epoch": 2.0341207349081363, "grad_norm": 0.8536164164543152, "learning_rate": 5.1567077880794725e-05, "loss": 1.5763671875, "step": 4650 }, { "ce_loss": 1.1328125, "distill_loss": 0.138671875, "epoch": 2.0341207349081363, "explain_loss": 1.0546875, "loss": 2.8125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.00286865234375, "step": 4650 }, { "ce_loss": 0.443359375, "distill_loss": 0.087890625, "epoch": 2.0341207349081363, "explain_loss": 0.73828125, "loss": 1.734375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.10205078125, "step": 4650 }, { "ce_loss": 0.0010833740234375, "distill_loss": 0.11669921875, "epoch": 2.0341207349081363, "explain_loss": 0.875, "loss": 1.4453125, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.0032958984375, "step": 4650 }, { "ce_loss": 0.0006866455078125, "distill_loss": 0.095703125, "epoch": 2.0341207349081363, "explain_loss": 1.171875, "loss": 1.578125, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.0034027099609375, "step": 4650 }, { "epoch": 2.0384951881014874, "grad_norm": 0.9378067255020142, "learning_rate": 5.1145764654326456e-05, "loss": 1.61240234375, "step": 4660 }, { "ce_loss": 0.0008544921875, "distill_loss": 0.06298828125, "epoch": 2.0384951881014874, "explain_loss": 0.9921875, "loss": 1.5078125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.00433349609375, "step": 4660 }, { "ce_loss": 0.01068115234375, "distill_loss": 0.1171875, "epoch": 2.0384951881014874, "explain_loss": 1.2265625, "loss": 1.703125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.0036468505859375, "step": 4660 }, { "ce_loss": 0.3515625, "distill_loss": 0.0859375, "epoch": 2.0384951881014874, "explain_loss": 1.25, "loss": 2.1875, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.0859375, "step": 4660 }, { "ce_loss": 0.0054931640625, "distill_loss": 0.07666015625, "epoch": 2.0384951881014874, "explain_loss": 1.4296875, "loss": 1.84375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.00396728515625, "step": 4660 }, { "epoch": 2.042869641294838, "grad_norm": 0.6094887256622314, "learning_rate": 5.0725587412039874e-05, "loss": 1.647314453125, "step": 4670 }, { "ce_loss": 0.0004138946533203125, "distill_loss": 0.087890625, "epoch": 2.042869641294838, "explain_loss": 1.0546875, "loss": 1.5390625, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0028839111328125, "step": 4670 }, { "ce_loss": 0.00099945068359375, "distill_loss": 0.1064453125, "epoch": 2.042869641294838, "explain_loss": 1.125, "loss": 1.578125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.00372314453125, "step": 4670 }, { "ce_loss": 0.5078125, "distill_loss": 0.080078125, "epoch": 2.042869641294838, "explain_loss": 0.96484375, "loss": 2.078125, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.043212890625, "step": 4670 }, { "ce_loss": 0.000888824462890625, "distill_loss": 0.1240234375, "epoch": 2.042869641294838, "explain_loss": 1.203125, "loss": 1.859375, "ref_ce_loss": 0.52734375, "stage_align_loss": 0.0035247802734375, "step": 4670 }, { "epoch": 2.047244094488189, "grad_norm": 0.7096444368362427, "learning_rate": 5.0306555924115514e-05, "loss": 1.62001953125, "step": 4680 }, { "ce_loss": 0.003936767578125, "distill_loss": 0.0830078125, "epoch": 2.047244094488189, "explain_loss": 1.0625, "loss": 1.5, "ref_ce_loss": 0.34375, "stage_align_loss": 0.003631591796875, "step": 4680 }, { "ce_loss": 0.0693359375, "distill_loss": 0.0947265625, "epoch": 2.047244094488189, "explain_loss": 1.0703125, "loss": 1.6328125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0062255859375, "step": 4680 }, { "ce_loss": 0.01263427734375, "distill_loss": 0.10986328125, "epoch": 2.047244094488189, "explain_loss": 1.234375, "loss": 1.8125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.0025787353515625, "step": 4680 }, { "ce_loss": 0.31640625, "distill_loss": 0.103515625, "epoch": 2.047244094488189, "explain_loss": 0.890625, "loss": 1.765625, "ref_ce_loss": 0.390625, "stage_align_loss": 0.06494140625, "step": 4680 }, { "epoch": 2.0516185476815396, "grad_norm": 0.6841128468513489, "learning_rate": 4.988867993409223e-05, "loss": 1.657666015625, "step": 4690 }, { "ce_loss": 0.150390625, "distill_loss": 0.1064453125, "epoch": 2.0516185476815396, "explain_loss": 1.2109375, "loss": 1.8984375, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00677490234375, "step": 4690 }, { "ce_loss": 0.64453125, "distill_loss": 0.078125, "epoch": 2.0516185476815396, "explain_loss": 1.109375, "loss": 2.28125, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.003997802734375, "step": 4690 }, { "ce_loss": 0.0869140625, "distill_loss": 0.154296875, "epoch": 2.0516185476815396, "explain_loss": 0.9453125, "loss": 1.8125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.080078125, "step": 4690 }, { "ce_loss": 0.005828857421875, "distill_loss": 0.0888671875, "epoch": 2.0516185476815396, "explain_loss": 1.390625, "loss": 1.765625, "ref_ce_loss": 0.28125, "stage_align_loss": 0.003814697265625, "step": 4690 }, { "epoch": 2.055993000874891, "grad_norm": 0.7963618636131287, "learning_rate": 4.9471969158640786e-05, "loss": 1.6681640625, "step": 4700 }, { "ce_loss": 0.01123046875, "distill_loss": 0.0751953125, "epoch": 2.055993000874891, "explain_loss": 1.0703125, "loss": 1.4609375, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.003692626953125, "step": 4700 }, { "ce_loss": 0.0010833740234375, "distill_loss": 0.12109375, "epoch": 2.055993000874891, "explain_loss": 1.375, "loss": 1.875, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.0033416748046875, "step": 4700 }, { "ce_loss": 0.0032958984375, "distill_loss": 0.1015625, "epoch": 2.055993000874891, "explain_loss": 0.98046875, "loss": 1.3359375, "ref_ce_loss": 0.2470703125, "stage_align_loss": 0.003631591796875, "step": 4700 }, { "ce_loss": 0.0004444122314453125, "distill_loss": 0.0791015625, "epoch": 2.055993000874891, "explain_loss": 0.91015625, "loss": 1.2890625, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.00372314453125, "step": 4700 }, { "epoch": 2.0603674540682415, "grad_norm": 0.5700620412826538, "learning_rate": 4.9056433287337486e-05, "loss": 1.65986328125, "step": 4710 }, { "ce_loss": 0.000690460205078125, "distill_loss": 0.115234375, "epoch": 2.0603674540682415, "explain_loss": 1.1171875, "loss": 1.734375, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.0029449462890625, "step": 4710 }, { "ce_loss": 0.00072479248046875, "distill_loss": 0.10205078125, "epoch": 2.0603674540682415, "explain_loss": 1.0859375, "loss": 1.4453125, "ref_ce_loss": 0.25390625, "stage_align_loss": 0.00390625, "step": 4710 }, { "ce_loss": 0.029052734375, "distill_loss": 0.0693359375, "epoch": 2.0603674540682415, "explain_loss": 1.09375, "loss": 1.609375, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.0029754638671875, "step": 4710 }, { "ce_loss": 0.00119781494140625, "distill_loss": 0.0849609375, "epoch": 2.0603674540682415, "explain_loss": 0.55859375, "loss": 0.90625, "ref_ce_loss": 0.26171875, "stage_align_loss": 0.002685546875, "step": 4710 }, { "epoch": 2.0647419072615922, "grad_norm": 0.6601372361183167, "learning_rate": 4.864208198243944e-05, "loss": 1.59697265625, "step": 4720 }, { "ce_loss": 0.0030975341796875, "distill_loss": 0.076171875, "epoch": 2.0647419072615922, "explain_loss": 1.3125, "loss": 1.71875, "ref_ce_loss": 0.328125, "stage_align_loss": 0.0030517578125, "step": 4720 }, { "ce_loss": 0.12109375, "distill_loss": 0.0634765625, "epoch": 2.0647419072615922, "explain_loss": 1.390625, "loss": 1.875, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.0037689208984375, "step": 4720 }, { "ce_loss": 0.58984375, "distill_loss": 0.0947265625, "epoch": 2.0647419072615922, "explain_loss": 1.078125, "loss": 2.28125, "ref_ce_loss": 0.4375, "stage_align_loss": 0.06689453125, "step": 4720 }, { "ce_loss": 0.004241943359375, "distill_loss": 0.12109375, "epoch": 2.0647419072615922, "explain_loss": 1.234375, "loss": 1.7109375, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.003143310546875, "step": 4720 }, { "epoch": 2.069116360454943, "grad_norm": 2.87430739402771, "learning_rate": 4.822892487865943e-05, "loss": 1.70361328125, "step": 4730 }, { "ce_loss": 0.08837890625, "distill_loss": 0.130859375, "epoch": 2.069116360454943, "explain_loss": 0.8125, "loss": 1.453125, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.004058837890625, "step": 4730 }, { "ce_loss": 0.1923828125, "distill_loss": 0.1044921875, "epoch": 2.069116360454943, "explain_loss": 0.8046875, "loss": 1.515625, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.07080078125, "step": 4730 }, { "ce_loss": 0.00445556640625, "distill_loss": 0.078125, "epoch": 2.069116360454943, "explain_loss": 0.90625, "loss": 1.34375, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0038299560546875, "step": 4730 }, { "ce_loss": 0.8359375, "distill_loss": 0.123046875, "epoch": 2.069116360454943, "explain_loss": 0.921875, "loss": 2.171875, "ref_ce_loss": 0.1962890625, "stage_align_loss": 0.09375, "step": 4730 }, { "epoch": 2.073490813648294, "grad_norm": 0.796497642993927, "learning_rate": 4.781697158294211e-05, "loss": 1.5919921875, "step": 4740 }, { "ce_loss": 0.01611328125, "distill_loss": 0.1083984375, "epoch": 2.073490813648294, "explain_loss": 1.09375, "loss": 1.4609375, "ref_ce_loss": 0.2392578125, "stage_align_loss": 0.005584716796875, "step": 4740 }, { "ce_loss": 0.00150299072265625, "distill_loss": 0.0693359375, "epoch": 2.073490813648294, "explain_loss": 1.296875, "loss": 1.8359375, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.00360107421875, "step": 4740 }, { "ce_loss": 0.00118255615234375, "distill_loss": 0.1015625, "epoch": 2.073490813648294, "explain_loss": 1.1953125, "loss": 1.671875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.003997802734375, "step": 4740 }, { "ce_loss": 0.00119781494140625, "distill_loss": 0.0732421875, "epoch": 2.073490813648294, "explain_loss": 0.90625, "loss": 1.2421875, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.0031890869140625, "step": 4740 }, { "epoch": 2.077865266841645, "grad_norm": 1.1155695915222168, "learning_rate": 4.740623167424049e-05, "loss": 1.644384765625, "step": 4750 }, { "ce_loss": 0.0003566741943359375, "distill_loss": 0.076171875, "epoch": 2.077865266841645, "explain_loss": 1.1015625, "loss": 1.46875, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.00360107421875, "step": 4750 }, { "ce_loss": 0.00106048583984375, "distill_loss": 0.083984375, "epoch": 2.077865266841645, "explain_loss": 1.0390625, "loss": 1.3828125, "ref_ce_loss": 0.25390625, "stage_align_loss": 0.0036163330078125, "step": 4750 }, { "ce_loss": 0.224609375, "distill_loss": 0.0859375, "epoch": 2.077865266841645, "explain_loss": 1.3671875, "loss": 2.40625, "ref_ce_loss": 0.515625, "stage_align_loss": 0.201171875, "step": 4750 }, { "ce_loss": 0.00128173828125, "distill_loss": 0.07666015625, "epoch": 2.077865266841645, "explain_loss": 1.015625, "loss": 1.484375, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.003814697265625, "step": 4750 }, { "epoch": 2.0822397200349956, "grad_norm": 0.657552182674408, "learning_rate": 4.699671470329342e-05, "loss": 1.60693359375, "step": 4760 }, { "ce_loss": 0.034423828125, "distill_loss": 0.07177734375, "epoch": 2.0822397200349956, "explain_loss": 1.1875, "loss": 1.796875, "ref_ce_loss": 0.494140625, "stage_align_loss": 0.007415771484375, "step": 4760 }, { "ce_loss": 0.71484375, "distill_loss": 0.07373046875, "epoch": 2.0822397200349956, "explain_loss": 1.25, "loss": 2.5, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.0079345703125, "step": 4760 }, { "ce_loss": 0.0003528594970703125, "distill_loss": 0.103515625, "epoch": 2.0822397200349956, "explain_loss": 1.125, "loss": 1.65625, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.003204345703125, "step": 4760 }, { "ce_loss": 0.0086669921875, "distill_loss": 0.09765625, "epoch": 2.0822397200349956, "explain_loss": 1.03125, "loss": 1.484375, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.0031585693359375, "step": 4760 }, { "epoch": 2.0866141732283463, "grad_norm": 1.3408544063568115, "learning_rate": 4.658843019240322e-05, "loss": 1.663671875, "step": 4770 }, { "ce_loss": 0.061279296875, "distill_loss": 0.103515625, "epoch": 2.0866141732283463, "explain_loss": 1.171875, "loss": 1.890625, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0120849609375, "step": 4770 }, { "ce_loss": 0.0654296875, "distill_loss": 0.09375, "epoch": 2.0866141732283463, "explain_loss": 0.8984375, "loss": 1.6796875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0888671875, "step": 4770 }, { "ce_loss": 0.00109100341796875, "distill_loss": 0.091796875, "epoch": 2.0866141732283463, "explain_loss": 0.8515625, "loss": 1.296875, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.002410888671875, "step": 4770 }, { "ce_loss": 0.00078582763671875, "distill_loss": 0.11181640625, "epoch": 2.0866141732283463, "explain_loss": 1.03125, "loss": 1.515625, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.0029296875, "step": 4770 }, { "epoch": 2.0909886264216975, "grad_norm": 0.753288745880127, "learning_rate": 4.618138763521448e-05, "loss": 1.56552734375, "step": 4780 }, { "ce_loss": 0.01513671875, "distill_loss": 0.095703125, "epoch": 2.0909886264216975, "explain_loss": 0.9765625, "loss": 1.59375, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.0038299560546875, "step": 4780 }, { "ce_loss": 0.045166015625, "distill_loss": 0.078125, "epoch": 2.0909886264216975, "explain_loss": 1.0390625, "loss": 1.5078125, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.0032806396484375, "step": 4780 }, { "ce_loss": 0.01025390625, "distill_loss": 0.11669921875, "epoch": 2.0909886264216975, "explain_loss": 1.03125, "loss": 1.484375, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.003997802734375, "step": 4780 }, { "ce_loss": 0.000919342041015625, "distill_loss": 0.091796875, "epoch": 2.0909886264216975, "explain_loss": 0.98046875, "loss": 1.453125, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.00286865234375, "step": 4780 }, { "epoch": 2.095363079615048, "grad_norm": 0.823583722114563, "learning_rate": 4.577559649649316e-05, "loss": 1.65791015625, "step": 4790 }, { "ce_loss": 0.0120849609375, "distill_loss": 0.06884765625, "epoch": 2.095363079615048, "explain_loss": 0.80078125, "loss": 1.171875, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.004486083984375, "step": 4790 }, { "ce_loss": 0.058837890625, "distill_loss": 0.0751953125, "epoch": 2.095363079615048, "explain_loss": 1.078125, "loss": 1.6015625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.005645751953125, "step": 4790 }, { "ce_loss": 0.007659912109375, "distill_loss": 0.1181640625, "epoch": 2.095363079615048, "explain_loss": 1.15625, "loss": 1.6875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.0030364990234375, "step": 4790 }, { "ce_loss": 0.01556396484375, "distill_loss": 0.107421875, "epoch": 2.095363079615048, "explain_loss": 0.921875, "loss": 1.453125, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.005035400390625, "step": 4790 }, { "epoch": 2.099737532808399, "grad_norm": 0.6005744338035583, "learning_rate": 4.537106621190674e-05, "loss": 1.603515625, "step": 4800 }, { "ce_loss": 0.58984375, "distill_loss": 0.08984375, "epoch": 2.099737532808399, "explain_loss": 1.3359375, "loss": 2.40625, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.003387451171875, "step": 4800 }, { "ce_loss": 0.392578125, "distill_loss": 0.0830078125, "epoch": 2.099737532808399, "explain_loss": 1.390625, "loss": 2.359375, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.10107421875, "step": 4800 }, { "ce_loss": 0.08251953125, "distill_loss": 0.07421875, "epoch": 2.099737532808399, "explain_loss": 0.89453125, "loss": 1.453125, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.004364013671875, "step": 4800 }, { "ce_loss": 0.006011962890625, "distill_loss": 0.12451171875, "epoch": 2.099737532808399, "explain_loss": 0.94921875, "loss": 1.4140625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.0026397705078125, "step": 4800 }, { "epoch": 2.1041119860017496, "grad_norm": 0.7981647253036499, "learning_rate": 4.4967806187804426e-05, "loss": 1.622021484375, "step": 4810 }, { "ce_loss": 0.0247802734375, "distill_loss": 0.0966796875, "epoch": 2.1041119860017496, "explain_loss": 0.85546875, "loss": 1.21875, "ref_ce_loss": 0.2412109375, "stage_align_loss": 0.00347900390625, "step": 4810 }, { "ce_loss": 0.00811767578125, "distill_loss": 0.1162109375, "epoch": 2.1041119860017496, "explain_loss": 1.2890625, "loss": 1.75, "ref_ce_loss": 0.328125, "stage_align_loss": 0.00396728515625, "step": 4810 }, { "ce_loss": 0.00958251953125, "distill_loss": 0.08349609375, "epoch": 2.1041119860017496, "explain_loss": 0.875, "loss": 1.328125, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0035552978515625, "step": 4810 }, { "ce_loss": 0.042724609375, "distill_loss": 0.080078125, "epoch": 2.1041119860017496, "explain_loss": 1.0234375, "loss": 1.46875, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.0045166015625, "step": 4810 }, { "epoch": 2.108486439195101, "grad_norm": 0.7543461918830872, "learning_rate": 4.456582580099886e-05, "loss": 1.6009765625, "step": 4820 }, { "ce_loss": 0.004974365234375, "distill_loss": 0.056640625, "epoch": 2.108486439195101, "explain_loss": 1.421875, "loss": 1.8984375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.004913330078125, "step": 4820 }, { "ce_loss": 0.40234375, "distill_loss": 0.07373046875, "epoch": 2.108486439195101, "explain_loss": 1.1015625, "loss": 1.9453125, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0037689208984375, "step": 4820 }, { "ce_loss": 0.0020294189453125, "distill_loss": 0.125, "epoch": 2.108486439195101, "explain_loss": 1.125, "loss": 1.5625, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.0029449462890625, "step": 4820 }, { "ce_loss": 0.001983642578125, "distill_loss": 0.0888671875, "epoch": 2.108486439195101, "explain_loss": 1.15625, "loss": 1.71875, "ref_ce_loss": 0.47265625, "stage_align_loss": 0.002899169921875, "step": 4820 }, { "epoch": 2.1128608923884515, "grad_norm": 0.6934460401535034, "learning_rate": 4.416513439854787e-05, "loss": 1.607568359375, "step": 4830 }, { "ce_loss": 0.140625, "distill_loss": 0.0966796875, "epoch": 2.1128608923884515, "explain_loss": 0.73046875, "loss": 1.3515625, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.03369140625, "step": 4830 }, { "ce_loss": 0.0172119140625, "distill_loss": 0.0751953125, "epoch": 2.1128608923884515, "explain_loss": 1.3515625, "loss": 1.90625, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.006072998046875, "step": 4830 }, { "ce_loss": 0.28125, "distill_loss": 0.0732421875, "epoch": 2.1128608923884515, "explain_loss": 1.140625, "loss": 1.9609375, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.06640625, "step": 4830 }, { "ce_loss": 0.000965118408203125, "distill_loss": 0.1259765625, "epoch": 2.1128608923884515, "explain_loss": 1.0625, "loss": 1.453125, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.0034027099609375, "step": 4830 }, { "epoch": 2.1172353455818023, "grad_norm": 0.8165754675865173, "learning_rate": 4.376574129753705e-05, "loss": 1.58740234375, "step": 4840 }, { "ce_loss": 0.000804901123046875, "distill_loss": 0.11474609375, "epoch": 2.1172353455818023, "explain_loss": 0.890625, "loss": 1.4453125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0036163330078125, "step": 4840 }, { "ce_loss": 0.01031494140625, "distill_loss": 0.087890625, "epoch": 2.1172353455818023, "explain_loss": 0.78125, "loss": 1.265625, "ref_ce_loss": 0.359375, "stage_align_loss": 0.0289306640625, "step": 4840 }, { "ce_loss": 0.0030517578125, "distill_loss": 0.1064453125, "epoch": 2.1172353455818023, "explain_loss": 1.1015625, "loss": 1.5703125, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.00390625, "step": 4840 }, { "ce_loss": 0.0006866455078125, "distill_loss": 0.1015625, "epoch": 2.1172353455818023, "explain_loss": 1.046875, "loss": 1.515625, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.003387451171875, "step": 4840 }, { "epoch": 2.121609798775153, "grad_norm": 2.038672924041748, "learning_rate": 4.336765578486328e-05, "loss": 1.65771484375, "step": 4850 }, { "ce_loss": 0.05126953125, "distill_loss": 0.1083984375, "epoch": 2.121609798775153, "explain_loss": 1.203125, "loss": 1.7578125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.004058837890625, "step": 4850 }, { "ce_loss": 1.0625, "distill_loss": 0.0830078125, "epoch": 2.121609798775153, "explain_loss": 1.140625, "loss": 2.8125, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.091796875, "step": 4850 }, { "ce_loss": 0.423828125, "distill_loss": 0.0810546875, "epoch": 2.121609798775153, "explain_loss": 1.03125, "loss": 2.421875, "ref_ce_loss": 0.69921875, "stage_align_loss": 0.1845703125, "step": 4850 }, { "ce_loss": 0.34765625, "distill_loss": 0.10009765625, "epoch": 2.121609798775153, "explain_loss": 0.92578125, "loss": 1.796875, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.07080078125, "step": 4850 }, { "epoch": 2.1259842519685037, "grad_norm": 0.5623254776000977, "learning_rate": 4.297088711701867e-05, "loss": 1.62412109375, "step": 4860 }, { "ce_loss": 0.0111083984375, "distill_loss": 0.07421875, "epoch": 2.1259842519685037, "explain_loss": 1.0390625, "loss": 1.5, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.00421142578125, "step": 4860 }, { "ce_loss": 0.03466796875, "distill_loss": 0.12109375, "epoch": 2.1259842519685037, "explain_loss": 1.46875, "loss": 2.015625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.005340576171875, "step": 4860 }, { "ce_loss": 0.00188446044921875, "distill_loss": 0.1142578125, "epoch": 2.1259842519685037, "explain_loss": 1.140625, "loss": 1.6171875, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.003997802734375, "step": 4860 }, { "ce_loss": 0.023681640625, "distill_loss": 0.11474609375, "epoch": 2.1259842519685037, "explain_loss": 1.1953125, "loss": 1.5625, "ref_ce_loss": 0.22265625, "stage_align_loss": 0.002716064453125, "step": 4860 }, { "epoch": 2.130358705161855, "grad_norm": 1.1251463890075684, "learning_rate": 4.257544451987546e-05, "loss": 1.573291015625, "step": 4870 }, { "ce_loss": 0.047119140625, "distill_loss": 0.09814453125, "epoch": 2.130358705161855, "explain_loss": 1.234375, "loss": 1.703125, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.004150390625, "step": 4870 }, { "ce_loss": 0.0014190673828125, "distill_loss": 0.0615234375, "epoch": 2.130358705161855, "explain_loss": 1.0, "loss": 1.265625, "ref_ce_loss": 0.197265625, "stage_align_loss": 0.00750732421875, "step": 4870 }, { "ce_loss": 0.00070953369140625, "distill_loss": 0.08544921875, "epoch": 2.130358705161855, "explain_loss": 1.03125, "loss": 1.3046875, "ref_ce_loss": 0.1845703125, "stage_align_loss": 0.00616455078125, "step": 4870 }, { "ce_loss": 0.00445556640625, "distill_loss": 0.0751953125, "epoch": 2.130358705161855, "explain_loss": 1.125, "loss": 1.4765625, "ref_ce_loss": 0.26953125, "stage_align_loss": 0.00457763671875, "step": 4870 }, { "epoch": 2.1347331583552056, "grad_norm": 0.9646345973014832, "learning_rate": 4.218133718847124e-05, "loss": 1.596875, "step": 4880 }, { "ce_loss": 0.130859375, "distill_loss": 0.076171875, "epoch": 2.1347331583552056, "explain_loss": 1.203125, "loss": 1.859375, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.004425048828125, "step": 4880 }, { "ce_loss": 0.251953125, "distill_loss": 0.1064453125, "epoch": 2.1347331583552056, "explain_loss": 1.09375, "loss": 2.125, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.16015625, "step": 4880 }, { "ce_loss": 0.65625, "distill_loss": 0.1240234375, "epoch": 2.1347331583552056, "explain_loss": 0.97265625, "loss": 2.21875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.0654296875, "step": 4880 }, { "ce_loss": 0.00213623046875, "distill_loss": 0.060546875, "epoch": 2.1347331583552056, "explain_loss": 0.93359375, "loss": 1.4375, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0034332275390625, "step": 4880 }, { "epoch": 2.1391076115485563, "grad_norm": 0.7356272339820862, "learning_rate": 4.178857428679548e-05, "loss": 1.589111328125, "step": 4890 }, { "ce_loss": 0.050048828125, "distill_loss": 0.080078125, "epoch": 2.1391076115485563, "explain_loss": 1.28125, "loss": 1.796875, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0035858154296875, "step": 4890 }, { "ce_loss": 0.001373291015625, "distill_loss": 0.095703125, "epoch": 2.1391076115485563, "explain_loss": 1.0, "loss": 1.40625, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.01129150390625, "step": 4890 }, { "ce_loss": 0.0007171630859375, "distill_loss": 0.076171875, "epoch": 2.1391076115485563, "explain_loss": 1.28125, "loss": 1.625, "ref_ce_loss": 0.26171875, "stage_align_loss": 0.0035400390625, "step": 4890 }, { "ce_loss": 0.00176239013671875, "distill_loss": 0.080078125, "epoch": 2.1391076115485563, "explain_loss": 1.375, "loss": 1.890625, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.003204345703125, "step": 4890 }, { "epoch": 2.143482064741907, "grad_norm": 0.6796180605888367, "learning_rate": 4.1397164947576163e-05, "loss": 1.5951171875, "step": 4900 }, { "ce_loss": 0.000553131103515625, "distill_loss": 0.10693359375, "epoch": 2.143482064741907, "explain_loss": 1.2421875, "loss": 1.734375, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.003021240234375, "step": 4900 }, { "ce_loss": 0.00040435791015625, "distill_loss": 0.10009765625, "epoch": 2.143482064741907, "explain_loss": 0.9140625, "loss": 1.2578125, "ref_ce_loss": 0.2373046875, "stage_align_loss": 0.0042724609375, "step": 4900 }, { "ce_loss": 0.000766754150390625, "distill_loss": 0.1005859375, "epoch": 2.143482064741907, "explain_loss": 0.85546875, "loss": 1.1640625, "ref_ce_loss": 0.2041015625, "stage_align_loss": 0.0042724609375, "step": 4900 }, { "ce_loss": 0.0035247802734375, "distill_loss": 0.09716796875, "epoch": 2.143482064741907, "explain_loss": 0.79296875, "loss": 1.203125, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.0037994384765625, "step": 4900 }, { "epoch": 2.1478565179352582, "grad_norm": 0.7609624266624451, "learning_rate": 4.1007118272067543e-05, "loss": 1.61708984375, "step": 4910 }, { "ce_loss": 0.0004138946533203125, "distill_loss": 0.078125, "epoch": 2.1478565179352582, "explain_loss": 1.0546875, "loss": 1.53125, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.003997802734375, "step": 4910 }, { "ce_loss": 0.0419921875, "distill_loss": 0.0576171875, "epoch": 2.1478565179352582, "explain_loss": 1.0859375, "loss": 1.6328125, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.00445556640625, "step": 4910 }, { "ce_loss": 0.0033721923828125, "distill_loss": 0.0810546875, "epoch": 2.1478565179352582, "explain_loss": 0.93359375, "loss": 1.5859375, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.1279296875, "step": 4910 }, { "ce_loss": 0.0047607421875, "distill_loss": 0.0771484375, "epoch": 2.1478565179352582, "explain_loss": 0.9453125, "loss": 1.296875, "ref_ce_loss": 0.265625, "stage_align_loss": 0.00347900390625, "step": 4910 }, { "epoch": 2.152230971128609, "grad_norm": 0.7443755865097046, "learning_rate": 4.0618443329838494e-05, "loss": 1.60361328125, "step": 4920 }, { "ce_loss": 0.9453125, "distill_loss": 0.076171875, "epoch": 2.152230971128609, "explain_loss": 1.1640625, "loss": 2.84375, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.189453125, "step": 4920 }, { "ce_loss": 0.259765625, "distill_loss": 0.08740234375, "epoch": 2.152230971128609, "explain_loss": 0.65234375, "loss": 1.3125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0245361328125, "step": 4920 }, { "ce_loss": 0.0004863739013671875, "distill_loss": 0.1103515625, "epoch": 2.152230971128609, "explain_loss": 0.9296875, "loss": 1.4140625, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.00360107421875, "step": 4920 }, { "ce_loss": 0.0625, "distill_loss": 0.0703125, "epoch": 2.152230971128609, "explain_loss": 1.140625, "loss": 1.6875, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0052490234375, "step": 4920 }, { "epoch": 2.1566054243219597, "grad_norm": 1.1227854490280151, "learning_rate": 4.023114915856173e-05, "loss": 1.6478515625, "step": 4930 }, { "ce_loss": 0.00994873046875, "distill_loss": 0.12109375, "epoch": 2.1566054243219597, "explain_loss": 1.2890625, "loss": 1.625, "ref_ce_loss": 0.1982421875, "stage_align_loss": 0.003875732421875, "step": 4930 }, { "ce_loss": 0.0034332275390625, "distill_loss": 0.111328125, "epoch": 2.1566054243219597, "explain_loss": 0.921875, "loss": 1.375, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.00372314453125, "step": 4930 }, { "ce_loss": 0.003997802734375, "distill_loss": 0.076171875, "epoch": 2.1566054243219597, "explain_loss": 0.9375, "loss": 1.453125, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.00360107421875, "step": 4930 }, { "ce_loss": 0.0078125, "distill_loss": 0.099609375, "epoch": 2.1566054243219597, "explain_loss": 1.15625, "loss": 1.71875, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.004547119140625, "step": 4930 }, { "epoch": 2.1609798775153104, "grad_norm": 0.811127245426178, "learning_rate": 3.9845244763803445e-05, "loss": 1.576416015625, "step": 4940 }, { "ce_loss": 0.5, "distill_loss": 0.0986328125, "epoch": 2.1609798775153104, "explain_loss": 0.93359375, "loss": 1.828125, "ref_ce_loss": 0.201171875, "stage_align_loss": 0.09423828125, "step": 4940 }, { "ce_loss": 0.00799560546875, "distill_loss": 0.107421875, "epoch": 2.1609798775153104, "explain_loss": 1.1171875, "loss": 1.578125, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.031982421875, "step": 4940 }, { "ce_loss": 0.02685546875, "distill_loss": 0.1181640625, "epoch": 2.1609798775153104, "explain_loss": 0.8203125, "loss": 1.2734375, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0198974609375, "step": 4940 }, { "ce_loss": 0.07958984375, "distill_loss": 0.0908203125, "epoch": 2.1609798775153104, "explain_loss": 0.90234375, "loss": 1.359375, "ref_ce_loss": 0.27734375, "stage_align_loss": 0.00811767578125, "step": 4940 }, { "epoch": 2.1653543307086616, "grad_norm": 1.2062304019927979, "learning_rate": 3.9460739118814064e-05, "loss": 1.681005859375, "step": 4950 }, { "ce_loss": 0.19140625, "distill_loss": 0.07666015625, "epoch": 2.1653543307086616, "explain_loss": 1.0, "loss": 1.578125, "ref_ce_loss": 0.3125, "stage_align_loss": 0.005523681640625, "step": 4950 }, { "ce_loss": 0.001251220703125, "distill_loss": 0.08056640625, "epoch": 2.1653543307086616, "explain_loss": 0.77734375, "loss": 1.234375, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.00384521484375, "step": 4950 }, { "ce_loss": 0.1767578125, "distill_loss": 0.09228515625, "epoch": 2.1653543307086616, "explain_loss": 0.9765625, "loss": 1.65625, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0172119140625, "step": 4950 }, { "ce_loss": 0.1162109375, "distill_loss": 0.10205078125, "epoch": 2.1653543307086616, "explain_loss": 1.234375, "loss": 1.7890625, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.0732421875, "step": 4950 }, { "epoch": 2.1697287839020123, "grad_norm": 0.8468434810638428, "learning_rate": 3.9077641164319525e-05, "loss": 1.63681640625, "step": 4960 }, { "ce_loss": 0.27734375, "distill_loss": 0.08740234375, "epoch": 2.1697287839020123, "explain_loss": 1.171875, "loss": 1.9609375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0634765625, "step": 4960 }, { "ce_loss": 0.07080078125, "distill_loss": 0.08349609375, "epoch": 2.1697287839020123, "explain_loss": 0.84375, "loss": 1.21875, "ref_ce_loss": 0.220703125, "stage_align_loss": 0.0036468505859375, "step": 4960 }, { "ce_loss": 0.028076171875, "distill_loss": 0.07763671875, "epoch": 2.1697287839020123, "explain_loss": 1.140625, "loss": 1.53125, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.002899169921875, "step": 4960 }, { "ce_loss": 0.1015625, "distill_loss": 0.10498046875, "epoch": 2.1697287839020123, "explain_loss": 1.296875, "loss": 1.8671875, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0079345703125, "step": 4960 }, { "epoch": 2.174103237095363, "grad_norm": 0.5986310839653015, "learning_rate": 3.86959598083135e-05, "loss": 1.5642578125, "step": 4970 }, { "ce_loss": 0.001220703125, "distill_loss": 0.076171875, "epoch": 2.174103237095363, "explain_loss": 0.85546875, "loss": 1.140625, "ref_ce_loss": 0.203125, "stage_align_loss": 0.003082275390625, "step": 4970 }, { "ce_loss": 0.01007080078125, "distill_loss": 0.08203125, "epoch": 2.174103237095363, "explain_loss": 0.734375, "loss": 1.1015625, "ref_ce_loss": 0.2734375, "stage_align_loss": 0.0029449462890625, "step": 4970 }, { "ce_loss": 0.0419921875, "distill_loss": 0.080078125, "epoch": 2.174103237095363, "explain_loss": 0.9375, "loss": 1.5390625, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.00616455078125, "step": 4970 }, { "ce_loss": 0.07373046875, "distill_loss": 0.0966796875, "epoch": 2.174103237095363, "explain_loss": 0.56640625, "loss": 1.0390625, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.04052734375, "step": 4970 }, { "epoch": 2.1784776902887137, "grad_norm": 0.5547401309013367, "learning_rate": 3.8315703925850066e-05, "loss": 1.543359375, "step": 4980 }, { "ce_loss": 0.0015106201171875, "distill_loss": 0.111328125, "epoch": 2.1784776902887137, "explain_loss": 1.1640625, "loss": 1.5625, "ref_ce_loss": 0.28125, "stage_align_loss": 0.0027618408203125, "step": 4980 }, { "ce_loss": 0.00225830078125, "distill_loss": 0.07666015625, "epoch": 2.1784776902887137, "explain_loss": 0.9453125, "loss": 1.4765625, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.00372314453125, "step": 4980 }, { "ce_loss": 0.197265625, "distill_loss": 0.0869140625, "epoch": 2.1784776902887137, "explain_loss": 0.90625, "loss": 1.578125, "ref_ce_loss": 0.328125, "stage_align_loss": 0.061279296875, "step": 4980 }, { "ce_loss": 0.0014190673828125, "distill_loss": 0.107421875, "epoch": 2.1784776902887137, "explain_loss": 1.28125, "loss": 1.75, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00274658203125, "step": 4980 }, { "epoch": 2.182852143482065, "grad_norm": 0.762994110584259, "learning_rate": 3.793688235883749e-05, "loss": 1.599365234375, "step": 4990 }, { "ce_loss": 0.02685546875, "distill_loss": 0.08837890625, "epoch": 2.182852143482065, "explain_loss": 0.98828125, "loss": 1.3671875, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.004364013671875, "step": 4990 }, { "ce_loss": 0.15625, "distill_loss": 0.1259765625, "epoch": 2.182852143482065, "explain_loss": 0.94921875, "loss": 1.59375, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.006439208984375, "step": 4990 }, { "ce_loss": 0.01458740234375, "distill_loss": 0.080078125, "epoch": 2.182852143482065, "explain_loss": 0.90625, "loss": 1.34375, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.002960205078125, "step": 4990 }, { "ce_loss": 0.20703125, "distill_loss": 0.103515625, "epoch": 2.182852143482065, "explain_loss": 1.5, "loss": 2.1875, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.056640625, "step": 4990 }, { "epoch": 2.1872265966754156, "grad_norm": 0.69655442237854, "learning_rate": 3.755950391583263e-05, "loss": 1.5998046875, "step": 5000 }, { "ce_loss": 0.000579833984375, "distill_loss": 0.0849609375, "epoch": 2.1872265966754156, "explain_loss": 1.265625, "loss": 1.703125, "ref_ce_loss": 0.349609375, "stage_align_loss": 0.003082275390625, "step": 5000 }, { "ce_loss": 0.0145263671875, "distill_loss": 0.0859375, "epoch": 2.1872265966754156, "explain_loss": 1.09375, "loss": 1.546875, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0027618408203125, "step": 5000 }, { "ce_loss": 0.002838134765625, "distill_loss": 0.09228515625, "epoch": 2.1872265966754156, "explain_loss": 0.8046875, "loss": 1.375, "ref_ce_loss": 0.46875, "stage_align_loss": 0.004119873046875, "step": 5000 }, { "ce_loss": 0.103515625, "distill_loss": 0.087890625, "epoch": 2.1872265966754156, "explain_loss": 0.703125, "loss": 1.203125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.01507568359375, "step": 5000 }, { "epoch": 2.1916010498687664, "grad_norm": 0.5789415240287781, "learning_rate": 3.7183577371835996e-05, "loss": 1.623046875, "step": 5010 }, { "ce_loss": 0.3515625, "distill_loss": 0.07421875, "epoch": 2.1916010498687664, "explain_loss": 0.72265625, "loss": 1.484375, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.035400390625, "step": 5010 }, { "ce_loss": 0.0030059814453125, "distill_loss": 0.083984375, "epoch": 2.1916010498687664, "explain_loss": 0.8984375, "loss": 1.328125, "ref_ce_loss": 0.341796875, "stage_align_loss": 0.0032958984375, "step": 5010 }, { "ce_loss": 0.005889892578125, "distill_loss": 0.1015625, "epoch": 2.1916010498687664, "explain_loss": 1.125, "loss": 1.609375, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.003082275390625, "step": 5010 }, { "ce_loss": 0.0135498046875, "distill_loss": 0.111328125, "epoch": 2.1916010498687664, "explain_loss": 0.984375, "loss": 1.59375, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.003021240234375, "step": 5010 }, { "epoch": 2.195975503062117, "grad_norm": 0.6645960211753845, "learning_rate": 3.6809111468087845e-05, "loss": 1.580078125, "step": 5020 }, { "ce_loss": 0.00194549560546875, "distill_loss": 0.07421875, "epoch": 2.195975503062117, "explain_loss": 0.953125, "loss": 1.34375, "ref_ce_loss": 0.310546875, "stage_align_loss": 0.0027923583984375, "step": 5020 }, { "ce_loss": 0.39453125, "distill_loss": 0.06787109375, "epoch": 2.195975503062117, "explain_loss": 0.8515625, "loss": 1.640625, "ref_ce_loss": 0.2490234375, "stage_align_loss": 0.07666015625, "step": 5020 }, { "ce_loss": 0.000934600830078125, "distill_loss": 0.11962890625, "epoch": 2.195975503062117, "explain_loss": 1.03125, "loss": 1.453125, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.002349853515625, "step": 5020 }, { "ce_loss": 0.1630859375, "distill_loss": 0.115234375, "epoch": 2.195975503062117, "explain_loss": 0.9609375, "loss": 1.609375, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.00494384765625, "step": 5020 }, { "epoch": 2.2003499562554683, "grad_norm": 0.7045149803161621, "learning_rate": 3.6436114911864774e-05, "loss": 1.5798828125, "step": 5030 }, { "ce_loss": 0.006072998046875, "distill_loss": 0.1259765625, "epoch": 2.2003499562554683, "explain_loss": 1.2734375, "loss": 1.7421875, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.00347900390625, "step": 5030 }, { "ce_loss": 0.146484375, "distill_loss": 0.0732421875, "epoch": 2.2003499562554683, "explain_loss": 0.75390625, "loss": 1.40625, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.08544921875, "step": 5030 }, { "ce_loss": 0.0021820068359375, "distill_loss": 0.07666015625, "epoch": 2.2003499562554683, "explain_loss": 0.71484375, "loss": 1.078125, "ref_ce_loss": 0.27734375, "stage_align_loss": 0.0032196044921875, "step": 5030 }, { "ce_loss": 0.0286865234375, "distill_loss": 0.10498046875, "epoch": 2.2003499562554683, "explain_loss": 1.21875, "loss": 1.7734375, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.006683349609375, "step": 5030 }, { "epoch": 2.204724409448819, "grad_norm": 0.5636362433433533, "learning_rate": 3.606459637627752e-05, "loss": 1.5810546875, "step": 5040 }, { "ce_loss": 0.26953125, "distill_loss": 0.06640625, "epoch": 2.204724409448819, "explain_loss": 1.28125, "loss": 2.09375, "ref_ce_loss": 0.447265625, "stage_align_loss": 0.024658203125, "step": 5040 }, { "ce_loss": 0.087890625, "distill_loss": 0.11767578125, "epoch": 2.204724409448819, "explain_loss": 1.21875, "loss": 1.796875, "ref_ce_loss": 0.375, "stage_align_loss": 0.005126953125, "step": 5040 }, { "ce_loss": 0.00141143798828125, "distill_loss": 0.08056640625, "epoch": 2.204724409448819, "explain_loss": 0.91796875, "loss": 1.328125, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.002960205078125, "step": 5040 }, { "ce_loss": 0.0673828125, "distill_loss": 0.1123046875, "epoch": 2.204724409448819, "explain_loss": 0.953125, "loss": 1.53125, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.003814697265625, "step": 5040 }, { "epoch": 2.2090988626421697, "grad_norm": 0.6519250869750977, "learning_rate": 3.569456450006888e-05, "loss": 1.61220703125, "step": 5050 }, { "ce_loss": 0.000667572021484375, "distill_loss": 0.095703125, "epoch": 2.2090988626421697, "explain_loss": 0.77734375, "loss": 1.2734375, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.00482177734375, "step": 5050 }, { "ce_loss": 0.001312255859375, "distill_loss": 0.0771484375, "epoch": 2.2090988626421697, "explain_loss": 0.9765625, "loss": 1.390625, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.0038299560546875, "step": 5050 }, { "ce_loss": 0.0113525390625, "distill_loss": 0.08740234375, "epoch": 2.2090988626421697, "explain_loss": 0.828125, "loss": 1.2265625, "ref_ce_loss": 0.296875, "stage_align_loss": 0.003875732421875, "step": 5050 }, { "ce_loss": 0.2353515625, "distill_loss": 0.0927734375, "epoch": 2.2090988626421697, "explain_loss": 0.96875, "loss": 1.7890625, "ref_ce_loss": 0.484375, "stage_align_loss": 0.006439208984375, "step": 5050 }, { "epoch": 2.2134733158355204, "grad_norm": 0.823628306388855, "learning_rate": 3.5326027887413285e-05, "loss": 1.532958984375, "step": 5060 }, { "ce_loss": 0.040771484375, "distill_loss": 0.072265625, "epoch": 2.2134733158355204, "explain_loss": 1.0625, "loss": 1.515625, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.00982666015625, "step": 5060 }, { "ce_loss": 0.057373046875, "distill_loss": 0.080078125, "epoch": 2.2134733158355204, "explain_loss": 0.9609375, "loss": 1.6484375, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.04296875, "step": 5060 }, { "ce_loss": 0.032470703125, "distill_loss": 0.095703125, "epoch": 2.2134733158355204, "explain_loss": 0.9921875, "loss": 1.375, "ref_ce_loss": 0.24609375, "stage_align_loss": 0.00311279296875, "step": 5060 }, { "ce_loss": 0.57421875, "distill_loss": 0.083984375, "epoch": 2.2134733158355204, "explain_loss": 1.1484375, "loss": 2.59375, "ref_ce_loss": 0.61328125, "stage_align_loss": 0.1787109375, "step": 5060 }, { "epoch": 2.2178477690288716, "grad_norm": 0.7710326313972473, "learning_rate": 3.495899510771639e-05, "loss": 1.760302734375, "step": 5070 }, { "ce_loss": 0.455078125, "distill_loss": 0.08251953125, "epoch": 2.2178477690288716, "explain_loss": 1.1875, "loss": 2.3125, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.0230712890625, "step": 5070 }, { "ce_loss": 0.0014495849609375, "distill_loss": 0.1005859375, "epoch": 2.2178477690288716, "explain_loss": 0.88671875, "loss": 1.3203125, "ref_ce_loss": 0.328125, "stage_align_loss": 0.003936767578125, "step": 5070 }, { "ce_loss": 0.005584716796875, "distill_loss": 0.095703125, "epoch": 2.2178477690288716, "explain_loss": 1.015625, "loss": 1.453125, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.004730224609375, "step": 5070 }, { "ce_loss": 0.056884765625, "distill_loss": 0.083984375, "epoch": 2.2178477690288716, "explain_loss": 0.9140625, "loss": 1.25, "ref_ce_loss": 0.19140625, "stage_align_loss": 0.0038909912109375, "step": 5070 }, { "epoch": 2.2222222222222223, "grad_norm": 1.0006499290466309, "learning_rate": 3.459347469541596e-05, "loss": 1.600390625, "step": 5080 }, { "ce_loss": 0.032470703125, "distill_loss": 0.115234375, "epoch": 2.2222222222222223, "explain_loss": 0.85546875, "loss": 1.515625, "ref_ce_loss": 0.50390625, "stage_align_loss": 0.00439453125, "step": 5080 }, { "ce_loss": 0.0255126953125, "distill_loss": 0.06396484375, "epoch": 2.2222222222222223, "explain_loss": 0.6015625, "loss": 1.03125, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.00335693359375, "step": 5080 }, { "ce_loss": 0.00116729736328125, "distill_loss": 0.099609375, "epoch": 2.2222222222222223, "explain_loss": 1.1796875, "loss": 1.703125, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.004119873046875, "step": 5080 }, { "ce_loss": 0.00262451171875, "distill_loss": 0.09814453125, "epoch": 2.2222222222222223, "explain_loss": 1.109375, "loss": 1.5, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.003143310546875, "step": 5080 }, { "epoch": 2.226596675415573, "grad_norm": 1.0145878791809082, "learning_rate": 3.422947514978341e-05, "loss": 1.6048828125, "step": 5090 }, { "ce_loss": 0.0011749267578125, "distill_loss": 0.11669921875, "epoch": 2.226596675415573, "explain_loss": 1.015625, "loss": 1.5390625, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.0031890869140625, "step": 5090 }, { "ce_loss": 0.006317138671875, "distill_loss": 0.09375, "epoch": 2.226596675415573, "explain_loss": 0.9375, "loss": 1.515625, "ref_ce_loss": 0.470703125, "stage_align_loss": 0.003997802734375, "step": 5090 }, { "ce_loss": 0.064453125, "distill_loss": 0.1142578125, "epoch": 2.226596675415573, "explain_loss": 0.7578125, "loss": 1.25, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.004852294921875, "step": 5090 }, { "ce_loss": 0.005401611328125, "distill_loss": 0.09228515625, "epoch": 2.226596675415573, "explain_loss": 0.54296875, "loss": 0.87109375, "ref_ce_loss": 0.2265625, "stage_align_loss": 0.0031280517578125, "step": 5090 }, { "epoch": 2.2309711286089238, "grad_norm": 0.8537073135375977, "learning_rate": 3.386700493472623e-05, "loss": 1.574853515625, "step": 5100 }, { "ce_loss": 1.2421875, "distill_loss": 0.0673828125, "epoch": 2.2309711286089238, "explain_loss": 0.8203125, "loss": 2.625, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.0072021484375, "step": 5100 }, { "ce_loss": 0.0179443359375, "distill_loss": 0.0595703125, "epoch": 2.2309711286089238, "explain_loss": 0.90625, "loss": 1.515625, "ref_ce_loss": 0.53125, "stage_align_loss": 0.0035552978515625, "step": 5100 }, { "ce_loss": 0.0478515625, "distill_loss": 0.072265625, "epoch": 2.2309711286089238, "explain_loss": 1.484375, "loss": 2.046875, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.0078125, "step": 5100 }, { "ce_loss": 0.00125885009765625, "distill_loss": 0.115234375, "epoch": 2.2309711286089238, "explain_loss": 0.58984375, "loss": 0.9921875, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.003387451171875, "step": 5100 }, { "epoch": 2.2353455818022745, "grad_norm": 0.7876828908920288, "learning_rate": 3.350607247859104e-05, "loss": 1.6677734375, "step": 5110 }, { "ce_loss": 0.00101470947265625, "distill_loss": 0.091796875, "epoch": 2.2353455818022745, "explain_loss": 0.59375, "loss": 0.96875, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.003173828125, "step": 5110 }, { "ce_loss": 0.00115203857421875, "distill_loss": 0.06494140625, "epoch": 2.2353455818022745, "explain_loss": 1.25, "loss": 1.7265625, "ref_ce_loss": 0.40625, "stage_align_loss": 0.002838134765625, "step": 5110 }, { "ce_loss": 0.006500244140625, "distill_loss": 0.0830078125, "epoch": 2.2353455818022745, "explain_loss": 1.03125, "loss": 1.578125, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.005218505859375, "step": 5110 }, { "ce_loss": 0.0022125244140625, "distill_loss": 0.0712890625, "epoch": 2.2353455818022745, "explain_loss": 0.87109375, "loss": 1.171875, "ref_ce_loss": 0.2265625, "stage_align_loss": 0.0025482177734375, "step": 5110 }, { "epoch": 2.2397200349956257, "grad_norm": 0.6271067261695862, "learning_rate": 3.3146686173967686e-05, "loss": 1.52744140625, "step": 5120 }, { "ce_loss": 0.1142578125, "distill_loss": 0.087890625, "epoch": 2.2397200349956257, "explain_loss": 1.0, "loss": 1.796875, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.029541015625, "step": 5120 }, { "ce_loss": 0.003570556640625, "distill_loss": 0.099609375, "epoch": 2.2397200349956257, "explain_loss": 1.328125, "loss": 1.65625, "ref_ce_loss": 0.224609375, "stage_align_loss": 0.003143310546875, "step": 5120 }, { "ce_loss": 0.85546875, "distill_loss": 0.10009765625, "epoch": 2.2397200349956257, "explain_loss": 1.0859375, "loss": 2.6875, "ref_ce_loss": 0.56640625, "stage_align_loss": 0.0751953125, "step": 5120 }, { "ce_loss": 0.001068115234375, "distill_loss": 0.0849609375, "epoch": 2.2397200349956257, "explain_loss": 1.21875, "loss": 1.84375, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0034637451171875, "step": 5120 }, { "epoch": 2.2440944881889764, "grad_norm": 0.6751693487167358, "learning_rate": 3.278885437749418e-05, "loss": 1.631201171875, "step": 5130 }, { "ce_loss": 0.52734375, "distill_loss": 0.0966796875, "epoch": 2.2440944881889764, "explain_loss": 1.125, "loss": 2.28125, "ref_ce_loss": 0.44921875, "stage_align_loss": 0.07421875, "step": 5130 }, { "ce_loss": 0.12890625, "distill_loss": 0.078125, "epoch": 2.2440944881889764, "explain_loss": 0.9921875, "loss": 1.5546875, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.00531005859375, "step": 5130 }, { "ce_loss": 0.1943359375, "distill_loss": 0.099609375, "epoch": 2.2440944881889764, "explain_loss": 1.0625, "loss": 1.828125, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.01104736328125, "step": 5130 }, { "ce_loss": 0.453125, "distill_loss": 0.0673828125, "epoch": 2.2440944881889764, "explain_loss": 1.21875, "loss": 2.1875, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.09716796875, "step": 5130 }, { "epoch": 2.248468941382327, "grad_norm": 0.8273366093635559, "learning_rate": 3.24325854096622e-05, "loss": 1.66044921875, "step": 5140 }, { "ce_loss": 0.003814697265625, "distill_loss": 0.126953125, "epoch": 2.248468941382327, "explain_loss": 1.1015625, "loss": 1.640625, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0027618408203125, "step": 5140 }, { "ce_loss": 0.0830078125, "distill_loss": 0.1259765625, "epoch": 2.248468941382327, "explain_loss": 0.7265625, "loss": 1.34375, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0244140625, "step": 5140 }, { "ce_loss": 0.146484375, "distill_loss": 0.09375, "epoch": 2.248468941382327, "explain_loss": 1.0234375, "loss": 1.5703125, "ref_ce_loss": 0.296875, "stage_align_loss": 0.00689697265625, "step": 5140 }, { "ce_loss": 0.0296630859375, "distill_loss": 0.0966796875, "epoch": 2.248468941382327, "explain_loss": 1.109375, "loss": 1.609375, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.003692626953125, "step": 5140 }, { "epoch": 2.2528433945756783, "grad_norm": 0.8326661586761475, "learning_rate": 3.207788755462376e-05, "loss": 1.6017578125, "step": 5150 }, { "ce_loss": 0.005096435546875, "distill_loss": 0.087890625, "epoch": 2.2528433945756783, "explain_loss": 0.75, "loss": 1.171875, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.0033111572265625, "step": 5150 }, { "ce_loss": 0.0301513671875, "distill_loss": 0.08349609375, "epoch": 2.2528433945756783, "explain_loss": 1.1484375, "loss": 1.578125, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0028228759765625, "step": 5150 }, { "ce_loss": 0.04296875, "distill_loss": 0.07373046875, "epoch": 2.2528433945756783, "explain_loss": 1.015625, "loss": 1.5703125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0036773681640625, "step": 5150 }, { "ce_loss": 0.03173828125, "distill_loss": 0.08203125, "epoch": 2.2528433945756783, "explain_loss": 1.046875, "loss": 1.546875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00421142578125, "step": 5150 }, { "epoch": 2.257217847769029, "grad_norm": 0.8494007587432861, "learning_rate": 3.1724769059998495e-05, "loss": 1.5833984375, "step": 5160 }, { "ce_loss": 1.8359375, "distill_loss": 0.0732421875, "epoch": 2.257217847769029, "explain_loss": 1.296875, "loss": 3.578125, "ref_ce_loss": 0.375, "stage_align_loss": 0.003204345703125, "step": 5160 }, { "ce_loss": 0.02001953125, "distill_loss": 0.099609375, "epoch": 2.257217847769029, "explain_loss": 0.8828125, "loss": 1.375, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.00482177734375, "step": 5160 }, { "ce_loss": 0.0020751953125, "distill_loss": 0.087890625, "epoch": 2.257217847769029, "explain_loss": 1.046875, "loss": 1.515625, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.0032501220703125, "step": 5160 }, { "ce_loss": 0.30859375, "distill_loss": 0.087890625, "epoch": 2.257217847769029, "explain_loss": 0.98046875, "loss": 1.8515625, "ref_ce_loss": 0.390625, "stage_align_loss": 0.08203125, "step": 5160 }, { "epoch": 2.2615923009623797, "grad_norm": 0.7439550757408142, "learning_rate": 3.137323813668207e-05, "loss": 1.6408203125, "step": 5170 }, { "ce_loss": 0.0137939453125, "distill_loss": 0.0693359375, "epoch": 2.2615923009623797, "explain_loss": 1.0390625, "loss": 1.53125, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.003753662109375, "step": 5170 }, { "ce_loss": 0.015625, "distill_loss": 0.064453125, "epoch": 2.2615923009623797, "explain_loss": 1.171875, "loss": 1.453125, "ref_ce_loss": 0.193359375, "stage_align_loss": 0.00311279296875, "step": 5170 }, { "ce_loss": 0.0164794921875, "distill_loss": 0.1083984375, "epoch": 2.2615923009623797, "explain_loss": 1.296875, "loss": 1.7734375, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.003204345703125, "step": 5170 }, { "ce_loss": 0.00408935546875, "distill_loss": 0.0986328125, "epoch": 2.2615923009623797, "explain_loss": 1.03125, "loss": 1.4921875, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.002716064453125, "step": 5170 }, { "epoch": 2.2659667541557305, "grad_norm": 1.02117919921875, "learning_rate": 3.102330295865488e-05, "loss": 1.593359375, "step": 5180 }, { "ce_loss": 0.000690460205078125, "distill_loss": 0.099609375, "epoch": 2.2659667541557305, "explain_loss": 0.6484375, "loss": 1.0234375, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.003021240234375, "step": 5180 }, { "ce_loss": 0.01263427734375, "distill_loss": 0.095703125, "epoch": 2.2659667541557305, "explain_loss": 1.25, "loss": 1.671875, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0037384033203125, "step": 5180 }, { "ce_loss": 0.00579833984375, "distill_loss": 0.07373046875, "epoch": 2.2659667541557305, "explain_loss": 0.7421875, "loss": 1.015625, "ref_ce_loss": 0.1884765625, "stage_align_loss": 0.0027618408203125, "step": 5180 }, { "ce_loss": 0.0010223388671875, "distill_loss": 0.09326171875, "epoch": 2.2659667541557305, "explain_loss": 1.0546875, "loss": 1.4296875, "ref_ce_loss": 0.27734375, "stage_align_loss": 0.0029449462890625, "step": 5180 }, { "epoch": 2.270341207349081, "grad_norm": 0.6563661694526672, "learning_rate": 3.067497166279244e-05, "loss": 1.66376953125, "step": 5190 }, { "ce_loss": 0.001708984375, "distill_loss": 0.091796875, "epoch": 2.270341207349081, "explain_loss": 1.1015625, "loss": 1.5390625, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.0038299560546875, "step": 5190 }, { "ce_loss": 0.1044921875, "distill_loss": 0.072265625, "epoch": 2.270341207349081, "explain_loss": 1.3125, "loss": 2.03125, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.0106201171875, "step": 5190 }, { "ce_loss": 0.61328125, "distill_loss": 0.1064453125, "epoch": 2.270341207349081, "explain_loss": 0.765625, "loss": 1.9296875, "ref_ce_loss": 0.359375, "stage_align_loss": 0.08349609375, "step": 5190 }, { "ce_loss": 0.008544921875, "distill_loss": 0.1435546875, "epoch": 2.270341207349081, "explain_loss": 1.046875, "loss": 1.625, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.0037078857421875, "step": 5190 }, { "epoch": 2.2747156605424323, "grad_norm": 0.729371190071106, "learning_rate": 3.03282523486758e-05, "loss": 1.58515625, "step": 5200 }, { "ce_loss": 0.451171875, "distill_loss": 0.072265625, "epoch": 2.2747156605424323, "explain_loss": 1.375, "loss": 2.5, "ref_ce_loss": 0.51953125, "stage_align_loss": 0.0830078125, "step": 5200 }, { "ce_loss": 0.232421875, "distill_loss": 0.0859375, "epoch": 2.2747156605424323, "explain_loss": 1.0546875, "loss": 1.75, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00347900390625, "step": 5200 }, { "ce_loss": 0.000408172607421875, "distill_loss": 0.08349609375, "epoch": 2.2747156605424323, "explain_loss": 0.890625, "loss": 1.1875, "ref_ce_loss": 0.2060546875, "stage_align_loss": 0.0029449462890625, "step": 5200 }, { "ce_loss": 0.0035858154296875, "distill_loss": 0.1142578125, "epoch": 2.2747156605424323, "explain_loss": 0.81640625, "loss": 1.203125, "ref_ce_loss": 0.26953125, "stage_align_loss": 0.003143310546875, "step": 5200 }, { "epoch": 2.279090113735783, "grad_norm": 0.63022381067276, "learning_rate": 2.9983153078403547e-05, "loss": 1.562890625, "step": 5210 }, { "ce_loss": 0.001373291015625, "distill_loss": 0.0771484375, "epoch": 2.279090113735783, "explain_loss": 1.4375, "loss": 1.9375, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.003326416015625, "step": 5210 }, { "ce_loss": 0.0032501220703125, "distill_loss": 0.07666015625, "epoch": 2.279090113735783, "explain_loss": 1.2734375, "loss": 1.734375, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.0034637451171875, "step": 5210 }, { "ce_loss": 0.1259765625, "distill_loss": 0.119140625, "epoch": 2.279090113735783, "explain_loss": 1.1953125, "loss": 1.84375, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.01251220703125, "step": 5210 }, { "ce_loss": 0.01434326171875, "distill_loss": 0.08740234375, "epoch": 2.279090113735783, "explain_loss": 0.99609375, "loss": 1.453125, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.005645751953125, "step": 5210 }, { "epoch": 2.283464566929134, "grad_norm": 0.5925443172454834, "learning_rate": 2.96396818764039e-05, "loss": 1.55625, "step": 5220 }, { "ce_loss": 0.001007080078125, "distill_loss": 0.083984375, "epoch": 2.283464566929134, "explain_loss": 0.95703125, "loss": 1.453125, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.0038909912109375, "step": 5220 }, { "ce_loss": 0.06494140625, "distill_loss": 0.054931640625, "epoch": 2.283464566929134, "explain_loss": 0.86328125, "loss": 1.484375, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0035552978515625, "step": 5220 }, { "ce_loss": 0.046142578125, "distill_loss": 0.09375, "epoch": 2.283464566929134, "explain_loss": 1.3125, "loss": 1.8125, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.00640869140625, "step": 5220 }, { "ce_loss": 0.0032501220703125, "distill_loss": 0.0732421875, "epoch": 2.283464566929134, "explain_loss": 1.171875, "loss": 1.5625, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.002532958984375, "step": 5220 }, { "epoch": 2.2878390201224845, "grad_norm": 0.7881119251251221, "learning_rate": 2.9297846729248657e-05, "loss": 1.59619140625, "step": 5230 }, { "ce_loss": 0.000537872314453125, "distill_loss": 0.087890625, "epoch": 2.2878390201224845, "explain_loss": 0.62890625, "loss": 0.99609375, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.00469970703125, "step": 5230 }, { "ce_loss": 0.0047607421875, "distill_loss": 0.076171875, "epoch": 2.2878390201224845, "explain_loss": 0.9140625, "loss": 1.34375, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.003021240234375, "step": 5230 }, { "ce_loss": 0.0115966796875, "distill_loss": 0.1015625, "epoch": 2.2878390201224845, "explain_loss": 0.89453125, "loss": 1.4296875, "ref_ce_loss": 0.41796875, "stage_align_loss": 0.0030670166015625, "step": 5230 }, { "ce_loss": 0.083984375, "distill_loss": 0.08740234375, "epoch": 2.2878390201224845, "explain_loss": 1.0078125, "loss": 1.4765625, "ref_ce_loss": 0.28125, "stage_align_loss": 0.01385498046875, "step": 5230 }, { "epoch": 2.2922134733158357, "grad_norm": 0.6074493527412415, "learning_rate": 2.8957655585467048e-05, "loss": 1.526904296875, "step": 5240 }, { "ce_loss": 0.00994873046875, "distill_loss": 0.06298828125, "epoch": 2.2922134733158357, "explain_loss": 0.9609375, "loss": 1.296875, "ref_ce_loss": 0.2578125, "stage_align_loss": 0.00323486328125, "step": 5240 }, { "ce_loss": 0.00151824951171875, "distill_loss": 0.1259765625, "epoch": 2.2922134733158357, "explain_loss": 1.609375, "loss": 2.03125, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.003082275390625, "step": 5240 }, { "ce_loss": 0.0037384033203125, "distill_loss": 0.08837890625, "epoch": 2.2922134733158357, "explain_loss": 1.1640625, "loss": 1.5625, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.0030059814453125, "step": 5240 }, { "ce_loss": 0.1767578125, "distill_loss": 0.095703125, "epoch": 2.2922134733158357, "explain_loss": 0.96875, "loss": 1.6875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.043701171875, "step": 5240 }, { "epoch": 2.2965879265091864, "grad_norm": 0.6508122086524963, "learning_rate": 2.861911635536112e-05, "loss": 1.642138671875, "step": 5250 }, { "ce_loss": 0.031494140625, "distill_loss": 0.07763671875, "epoch": 2.2965879265091864, "explain_loss": 1.140625, "loss": 1.671875, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.027587890625, "step": 5250 }, { "ce_loss": 0.044677734375, "distill_loss": 0.07421875, "epoch": 2.2965879265091864, "explain_loss": 0.95703125, "loss": 1.484375, "ref_ce_loss": 0.328125, "stage_align_loss": 0.07958984375, "step": 5250 }, { "ce_loss": 0.28515625, "distill_loss": 0.080078125, "epoch": 2.2965879265091864, "explain_loss": 1.1015625, "loss": 2.03125, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.10498046875, "step": 5250 }, { "ce_loss": 0.059814453125, "distill_loss": 0.10693359375, "epoch": 2.2965879265091864, "explain_loss": 1.21875, "loss": 1.9375, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.08837890625, "step": 5250 }, { "epoch": 2.300962379702537, "grad_norm": 1.2626807689666748, "learning_rate": 2.8282236910821735e-05, "loss": 1.603759765625, "step": 5260 }, { "ce_loss": 0.04052734375, "distill_loss": 0.080078125, "epoch": 2.300962379702537, "explain_loss": 1.0, "loss": 1.5, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.004150390625, "step": 5260 }, { "ce_loss": 0.11962890625, "distill_loss": 0.072265625, "epoch": 2.300962379702537, "explain_loss": 1.0, "loss": 1.6015625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.006927490234375, "step": 5260 }, { "ce_loss": 0.000888824462890625, "distill_loss": 0.087890625, "epoch": 2.300962379702537, "explain_loss": 1.4609375, "loss": 2.078125, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.00408935546875, "step": 5260 }, { "ce_loss": 0.01116943359375, "distill_loss": 0.09033203125, "epoch": 2.300962379702537, "explain_loss": 0.8125, "loss": 1.09375, "ref_ce_loss": 0.171875, "stage_align_loss": 0.0037994384765625, "step": 5260 }, { "epoch": 2.305336832895888, "grad_norm": 0.5719014406204224, "learning_rate": 2.7947025085145663e-05, "loss": 1.5880859375, "step": 5270 }, { "ce_loss": 0.00072479248046875, "distill_loss": 0.072265625, "epoch": 2.305336832895888, "explain_loss": 0.9765625, "loss": 1.34375, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.00433349609375, "step": 5270 }, { "ce_loss": 0.0159912109375, "distill_loss": 0.0771484375, "epoch": 2.305336832895888, "explain_loss": 0.78125, "loss": 1.203125, "ref_ce_loss": 0.328125, "stage_align_loss": 0.00421142578125, "step": 5270 }, { "ce_loss": 0.056884765625, "distill_loss": 0.07275390625, "epoch": 2.305336832895888, "explain_loss": 1.390625, "loss": 1.984375, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.00634765625, "step": 5270 }, { "ce_loss": 0.01141357421875, "distill_loss": 0.0751953125, "epoch": 2.305336832895888, "explain_loss": 1.09375, "loss": 1.4765625, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.003662109375, "step": 5270 }, { "epoch": 2.309711286089239, "grad_norm": 0.7801466584205627, "learning_rate": 2.7613488672853205e-05, "loss": 1.64853515625, "step": 5280 }, { "ce_loss": 0.003936767578125, "distill_loss": 0.07421875, "epoch": 2.309711286089239, "explain_loss": 1.15625, "loss": 1.765625, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.00372314453125, "step": 5280 }, { "ce_loss": 0.005401611328125, "distill_loss": 0.083984375, "epoch": 2.309711286089239, "explain_loss": 1.140625, "loss": 1.515625, "ref_ce_loss": 0.28125, "stage_align_loss": 0.004058837890625, "step": 5280 }, { "ce_loss": 0.10888671875, "distill_loss": 0.10693359375, "epoch": 2.309711286089239, "explain_loss": 0.90625, "loss": 1.484375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.0279541015625, "step": 5280 }, { "ce_loss": 0.00151824951171875, "distill_loss": 0.0712890625, "epoch": 2.309711286089239, "explain_loss": 1.140625, "loss": 1.65625, "ref_ce_loss": 0.4375, "stage_align_loss": 0.003997802734375, "step": 5280 }, { "epoch": 2.3140857392825898, "grad_norm": 1.1268389225006104, "learning_rate": 2.7281635429507103e-05, "loss": 1.64951171875, "step": 5290 }, { "ce_loss": 0.0035858154296875, "distill_loss": 0.10498046875, "epoch": 2.3140857392825898, "explain_loss": 1.28125, "loss": 1.65625, "ref_ce_loss": 0.267578125, "stage_align_loss": 0.0035552978515625, "step": 5290 }, { "ce_loss": 0.1435546875, "distill_loss": 0.07080078125, "epoch": 2.3140857392825898, "explain_loss": 0.7109375, "loss": 1.265625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.006683349609375, "step": 5290 }, { "ce_loss": 0.0067138671875, "distill_loss": 0.111328125, "epoch": 2.3140857392825898, "explain_loss": 1.078125, "loss": 1.59375, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.003997802734375, "step": 5290 }, { "ce_loss": 0.1083984375, "distill_loss": 0.09033203125, "epoch": 2.3140857392825898, "explain_loss": 1.15625, "loss": 1.84375, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.08349609375, "step": 5290 }, { "epoch": 2.3184601924759405, "grad_norm": 0.6886988878250122, "learning_rate": 2.695147307153225e-05, "loss": 1.6255859375, "step": 5300 }, { "ce_loss": 0.001983642578125, "distill_loss": 0.09228515625, "epoch": 2.3184601924759405, "explain_loss": 1.140625, "loss": 1.6640625, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.003631591796875, "step": 5300 }, { "ce_loss": 0.015625, "distill_loss": 0.091796875, "epoch": 2.3184601924759405, "explain_loss": 1.2421875, "loss": 1.6484375, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.0031280517578125, "step": 5300 }, { "ce_loss": 0.01123046875, "distill_loss": 0.09326171875, "epoch": 2.3184601924759405, "explain_loss": 1.25, "loss": 1.796875, "ref_ce_loss": 0.44140625, "stage_align_loss": 0.0030517578125, "step": 5300 }, { "ce_loss": 0.326171875, "distill_loss": 0.08349609375, "epoch": 2.3184601924759405, "explain_loss": 1.1484375, "loss": 2.0625, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0211181640625, "step": 5300 }, { "epoch": 2.322834645669291, "grad_norm": 0.9099979996681213, "learning_rate": 2.662300927603609e-05, "loss": 1.648046875, "step": 5310 }, { "ce_loss": 0.029052734375, "distill_loss": 0.0810546875, "epoch": 2.322834645669291, "explain_loss": 0.7421875, "loss": 1.28125, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0299072265625, "step": 5310 }, { "ce_loss": 0.0101318359375, "distill_loss": 0.07421875, "epoch": 2.322834645669291, "explain_loss": 1.328125, "loss": 1.875, "ref_ce_loss": 0.458984375, "stage_align_loss": 0.0031585693359375, "step": 5310 }, { "ce_loss": 0.2451171875, "distill_loss": 0.07421875, "epoch": 2.322834645669291, "explain_loss": 0.59375, "loss": 1.2265625, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.013671875, "step": 5310 }, { "ce_loss": 0.01397705078125, "distill_loss": 0.09765625, "epoch": 2.322834645669291, "explain_loss": 1.140625, "loss": 1.71875, "ref_ce_loss": 0.3828125, "stage_align_loss": 0.07861328125, "step": 5310 }, { "epoch": 2.3272090988626424, "grad_norm": 0.7462450265884399, "learning_rate": 2.629625168063028e-05, "loss": 1.60517578125, "step": 5320 }, { "ce_loss": 0.1787109375, "distill_loss": 0.10009765625, "epoch": 2.3272090988626424, "explain_loss": 0.9453125, "loss": 1.609375, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.004302978515625, "step": 5320 }, { "ce_loss": 0.016357421875, "distill_loss": 0.1025390625, "epoch": 2.3272090988626424, "explain_loss": 1.09375, "loss": 1.65625, "ref_ce_loss": 0.4375, "stage_align_loss": 0.005706787109375, "step": 5320 }, { "ce_loss": 0.059326171875, "distill_loss": 0.107421875, "epoch": 2.3272090988626424, "explain_loss": 1.546875, "loss": 2.21875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.072265625, "step": 5320 }, { "ce_loss": 0.1083984375, "distill_loss": 0.07177734375, "epoch": 2.3272090988626424, "explain_loss": 0.9609375, "loss": 1.515625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.07080078125, "step": 5320 }, { "epoch": 2.331583552055993, "grad_norm": 0.6302648782730103, "learning_rate": 2.5971207883252946e-05, "loss": 1.63623046875, "step": 5330 }, { "ce_loss": 0.05712890625, "distill_loss": 0.0888671875, "epoch": 2.331583552055993, "explain_loss": 0.8828125, "loss": 1.3984375, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.006500244140625, "step": 5330 }, { "ce_loss": 0.000946044921875, "distill_loss": 0.0703125, "epoch": 2.331583552055993, "explain_loss": 0.9921875, "loss": 1.453125, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0031890869140625, "step": 5330 }, { "ce_loss": 0.001556396484375, "distill_loss": 0.0830078125, "epoch": 2.331583552055993, "explain_loss": 0.8984375, "loss": 1.40625, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.00286865234375, "step": 5330 }, { "ce_loss": 0.00116729736328125, "distill_loss": 0.115234375, "epoch": 2.331583552055993, "explain_loss": 0.80859375, "loss": 1.265625, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.0028076171875, "step": 5330 }, { "epoch": 2.335958005249344, "grad_norm": 0.5605002641677856, "learning_rate": 2.5647885441992225e-05, "loss": 1.6376953125, "step": 5340 }, { "ce_loss": 0.04541015625, "distill_loss": 0.08544921875, "epoch": 2.335958005249344, "explain_loss": 0.9765625, "loss": 1.3515625, "ref_ce_loss": 0.224609375, "stage_align_loss": 0.0220947265625, "step": 5340 }, { "ce_loss": 0.000682830810546875, "distill_loss": 0.1025390625, "epoch": 2.335958005249344, "explain_loss": 0.89453125, "loss": 1.40625, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.005401611328125, "step": 5340 }, { "ce_loss": 0.01080322265625, "distill_loss": 0.06005859375, "epoch": 2.335958005249344, "explain_loss": 1.21875, "loss": 1.75, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.0031585693359375, "step": 5340 }, { "ce_loss": 0.0013885498046875, "distill_loss": 0.1015625, "epoch": 2.335958005249344, "explain_loss": 0.703125, "loss": 1.078125, "ref_ce_loss": 0.267578125, "stage_align_loss": 0.0028076171875, "step": 5340 }, { "epoch": 2.3403324584426946, "grad_norm": 0.5279343128204346, "learning_rate": 2.5326291874910167e-05, "loss": 1.5837890625, "step": 5350 }, { "ce_loss": 0.036376953125, "distill_loss": 0.09814453125, "epoch": 2.3403324584426946, "explain_loss": 0.78125, "loss": 1.34375, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.027587890625, "step": 5350 }, { "ce_loss": 0.000934600830078125, "distill_loss": 0.115234375, "epoch": 2.3403324584426946, "explain_loss": 1.3984375, "loss": 1.875, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0033416748046875, "step": 5350 }, { "ce_loss": 0.023193359375, "distill_loss": 0.087890625, "epoch": 2.3403324584426946, "explain_loss": 1.0, "loss": 1.484375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0057373046875, "step": 5350 }, { "ce_loss": 0.000774383544921875, "distill_loss": 0.115234375, "epoch": 2.3403324584426946, "explain_loss": 0.796875, "loss": 1.296875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00225830078125, "step": 5350 }, { "epoch": 2.3447069116360453, "grad_norm": 0.812629222869873, "learning_rate": 2.500643465986837e-05, "loss": 1.57578125, "step": 5360 }, { "ce_loss": 0.031494140625, "distill_loss": 0.123046875, "epoch": 2.3447069116360453, "explain_loss": 0.90625, "loss": 1.421875, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.0045166015625, "step": 5360 }, { "ce_loss": 0.26953125, "distill_loss": 0.0693359375, "epoch": 2.3447069116360453, "explain_loss": 1.09375, "loss": 1.796875, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.044921875, "step": 5360 }, { "ce_loss": 0.016357421875, "distill_loss": 0.0927734375, "epoch": 2.3447069116360453, "explain_loss": 0.58984375, "loss": 1.0703125, "ref_ce_loss": 0.369140625, "stage_align_loss": 0.00445556640625, "step": 5360 }, { "ce_loss": 0.1669921875, "distill_loss": 0.076171875, "epoch": 2.3447069116360453, "explain_loss": 0.859375, "loss": 1.578125, "ref_ce_loss": 0.37109375, "stage_align_loss": 0.1044921875, "step": 5360 }, { "epoch": 2.3490813648293964, "grad_norm": 0.8581985235214233, "learning_rate": 2.4688321234353708e-05, "loss": 1.58779296875, "step": 5370 }, { "ce_loss": 0.193359375, "distill_loss": 0.0771484375, "epoch": 2.3490813648293964, "explain_loss": 0.8671875, "loss": 1.53125, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.01092529296875, "step": 5370 }, { "ce_loss": 0.000457763671875, "distill_loss": 0.09228515625, "epoch": 2.3490813648293964, "explain_loss": 1.0546875, "loss": 1.421875, "ref_ce_loss": 0.275390625, "stage_align_loss": 0.004119873046875, "step": 5370 }, { "ce_loss": 0.0020751953125, "distill_loss": 0.07763671875, "epoch": 2.3490813648293964, "explain_loss": 0.9296875, "loss": 1.421875, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.002899169921875, "step": 5370 }, { "ce_loss": 0.004119873046875, "distill_loss": 0.080078125, "epoch": 2.3490813648293964, "explain_loss": 0.9375, "loss": 1.421875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.0037994384765625, "step": 5370 }, { "epoch": 2.353455818022747, "grad_norm": 0.6364625096321106, "learning_rate": 2.4371958995305622e-05, "loss": 1.593359375, "step": 5380 }, { "ce_loss": 0.0177001953125, "distill_loss": 0.09033203125, "epoch": 2.353455818022747, "explain_loss": 1.140625, "loss": 1.65625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.007415771484375, "step": 5380 }, { "ce_loss": 0.000438690185546875, "distill_loss": 0.11669921875, "epoch": 2.353455818022747, "explain_loss": 1.0546875, "loss": 1.59375, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.003997802734375, "step": 5380 }, { "ce_loss": 0.00152587890625, "distill_loss": 0.1162109375, "epoch": 2.353455818022747, "explain_loss": 0.875, "loss": 1.28125, "ref_ce_loss": 0.28125, "stage_align_loss": 0.003509521484375, "step": 5380 }, { "ce_loss": 0.009521484375, "distill_loss": 0.0888671875, "epoch": 2.353455818022747, "explain_loss": 1.2109375, "loss": 1.734375, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00286865234375, "step": 5380 }, { "epoch": 2.357830271216098, "grad_norm": 0.5519525408744812, "learning_rate": 2.4057355298943983e-05, "loss": 1.533447265625, "step": 5390 }, { "ce_loss": 0.0006561279296875, "distill_loss": 0.08837890625, "epoch": 2.357830271216098, "explain_loss": 0.9921875, "loss": 1.5, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.00323486328125, "step": 5390 }, { "ce_loss": 0.003509521484375, "distill_loss": 0.087890625, "epoch": 2.357830271216098, "explain_loss": 0.97265625, "loss": 1.4140625, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.0035858154296875, "step": 5390 }, { "ce_loss": 0.000293731689453125, "distill_loss": 0.14453125, "epoch": 2.357830271216098, "explain_loss": 1.1484375, "loss": 1.78125, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.0035247802734375, "step": 5390 }, { "ce_loss": 0.051025390625, "distill_loss": 0.083984375, "epoch": 2.357830271216098, "explain_loss": 1.2265625, "loss": 1.859375, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.09228515625, "step": 5390 }, { "epoch": 2.362204724409449, "grad_norm": 0.5573031306266785, "learning_rate": 2.3744517460598237e-05, "loss": 1.58642578125, "step": 5400 }, { "ce_loss": 0.003387451171875, "distill_loss": 0.053955078125, "epoch": 2.362204724409449, "explain_loss": 0.9453125, "loss": 1.5, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.0035247802734375, "step": 5400 }, { "ce_loss": 0.01409912109375, "distill_loss": 0.07763671875, "epoch": 2.362204724409449, "explain_loss": 1.03125, "loss": 1.5625, "ref_ce_loss": 0.4296875, "stage_align_loss": 0.00421142578125, "step": 5400 }, { "ce_loss": 0.09765625, "distill_loss": 0.0888671875, "epoch": 2.362204724409449, "explain_loss": 0.81640625, "loss": 1.234375, "ref_ce_loss": 0.22265625, "stage_align_loss": 0.0031890869140625, "step": 5400 }, { "ce_loss": 0.0059814453125, "distill_loss": 0.083984375, "epoch": 2.362204724409449, "explain_loss": 1.5234375, "loss": 2.0625, "ref_ce_loss": 0.4375, "stage_align_loss": 0.004486083984375, "step": 5400 }, { "epoch": 2.3665791776028, "grad_norm": 0.7330650091171265, "learning_rate": 2.3433452754537055e-05, "loss": 1.71982421875, "step": 5410 }, { "ce_loss": 0.29296875, "distill_loss": 0.09326171875, "epoch": 2.3665791776028, "explain_loss": 0.9140625, "loss": 1.78125, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.08935546875, "step": 5410 }, { "ce_loss": 0.007659912109375, "distill_loss": 0.0634765625, "epoch": 2.3665791776028, "explain_loss": 1.375, "loss": 2.0625, "ref_ce_loss": 0.62109375, "stage_align_loss": 0.003936767578125, "step": 5410 }, { "ce_loss": 0.005706787109375, "distill_loss": 0.091796875, "epoch": 2.3665791776028, "explain_loss": 1.375, "loss": 2.03125, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.003814697265625, "step": 5410 }, { "ce_loss": 1.0546875, "distill_loss": 0.07861328125, "epoch": 2.3665791776028, "explain_loss": 1.484375, "loss": 3.09375, "ref_ce_loss": 0.48046875, "stage_align_loss": 0.0037384033203125, "step": 5410 }, { "epoch": 2.3709536307961505, "grad_norm": 0.8410758972167969, "learning_rate": 2.312416841379933e-05, "loss": 1.5580078125, "step": 5420 }, { "ce_loss": 0.1611328125, "distill_loss": 0.1201171875, "epoch": 2.3709536307961505, "explain_loss": 1.046875, "loss": 1.8671875, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.1123046875, "step": 5420 }, { "ce_loss": 0.002105712890625, "distill_loss": 0.0751953125, "epoch": 2.3709536307961505, "explain_loss": 0.72265625, "loss": 1.125, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.00274658203125, "step": 5420 }, { "ce_loss": 0.03125, "distill_loss": 0.07666015625, "epoch": 2.3709536307961505, "explain_loss": 1.359375, "loss": 1.6875, "ref_ce_loss": 0.21484375, "stage_align_loss": 0.0034027099609375, "step": 5420 }, { "ce_loss": 0.11376953125, "distill_loss": 0.0751953125, "epoch": 2.3709536307961505, "explain_loss": 0.921875, "loss": 1.5078125, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.00469970703125, "step": 5420 }, { "epoch": 2.3753280839895012, "grad_norm": 0.6198067665100098, "learning_rate": 2.2816671630026066e-05, "loss": 1.5533203125, "step": 5430 }, { "ce_loss": 0.050537109375, "distill_loss": 0.078125, "epoch": 2.3753280839895012, "explain_loss": 0.75390625, "loss": 1.3203125, "ref_ce_loss": 0.380859375, "stage_align_loss": 0.056884765625, "step": 5430 }, { "ce_loss": 1.4140625, "distill_loss": 0.07861328125, "epoch": 2.3753280839895012, "explain_loss": 1.21875, "loss": 3.28125, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.10595703125, "step": 5430 }, { "ce_loss": 0.001983642578125, "distill_loss": 0.072265625, "epoch": 2.3753280839895012, "explain_loss": 0.6328125, "loss": 1.0859375, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.003448486328125, "step": 5430 }, { "ce_loss": 0.0155029296875, "distill_loss": 0.10302734375, "epoch": 2.3753280839895012, "explain_loss": 1.0703125, "loss": 1.671875, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0233154296875, "step": 5430 }, { "epoch": 2.379702537182852, "grad_norm": 0.7355261445045471, "learning_rate": 2.251096955329295e-05, "loss": 1.556884765625, "step": 5440 }, { "ce_loss": 0.0206298828125, "distill_loss": 0.1328125, "epoch": 2.379702537182852, "explain_loss": 0.98046875, "loss": 1.640625, "ref_ce_loss": 0.5078125, "stage_align_loss": 0.00408935546875, "step": 5440 }, { "ce_loss": 0.0174560546875, "distill_loss": 0.115234375, "epoch": 2.379702537182852, "explain_loss": 1.359375, "loss": 1.8046875, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.0031585693359375, "step": 5440 }, { "ce_loss": 0.0262451171875, "distill_loss": 0.119140625, "epoch": 2.379702537182852, "explain_loss": 0.9140625, "loss": 1.3984375, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.00677490234375, "step": 5440 }, { "ce_loss": 0.00518798828125, "distill_loss": 0.08203125, "epoch": 2.379702537182852, "explain_loss": 0.890625, "loss": 1.265625, "ref_ce_loss": 0.279296875, "stage_align_loss": 0.0030670166015625, "step": 5440 }, { "epoch": 2.384076990376203, "grad_norm": 0.6384096741676331, "learning_rate": 2.220706929194424e-05, "loss": 1.58583984375, "step": 5450 }, { "ce_loss": 0.0235595703125, "distill_loss": 0.095703125, "epoch": 2.384076990376203, "explain_loss": 0.953125, "loss": 1.34375, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.0037384033203125, "step": 5450 }, { "ce_loss": 0.020263671875, "distill_loss": 0.08203125, "epoch": 2.384076990376203, "explain_loss": 1.0625, "loss": 1.546875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.003997802734375, "step": 5450 }, { "ce_loss": 0.0771484375, "distill_loss": 0.07373046875, "epoch": 2.384076990376203, "explain_loss": 0.921875, "loss": 1.421875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.005462646484375, "step": 5450 }, { "ce_loss": 0.134765625, "distill_loss": 0.080078125, "epoch": 2.384076990376203, "explain_loss": 0.8828125, "loss": 1.515625, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.10205078125, "step": 5450 }, { "epoch": 2.388451443569554, "grad_norm": 1.473445177078247, "learning_rate": 2.1904977912427416e-05, "loss": 1.6435546875, "step": 5460 }, { "ce_loss": 0.000225067138671875, "distill_loss": 0.0703125, "epoch": 2.388451443569554, "explain_loss": 1.078125, "loss": 1.5625, "ref_ce_loss": 0.41015625, "stage_align_loss": 0.0033721923828125, "step": 5460 }, { "ce_loss": 0.10107421875, "distill_loss": 0.078125, "epoch": 2.388451443569554, "explain_loss": 0.96875, "loss": 1.671875, "ref_ce_loss": 0.45703125, "stage_align_loss": 0.0693359375, "step": 5460 }, { "ce_loss": 0.00131988525390625, "distill_loss": 0.06640625, "epoch": 2.388451443569554, "explain_loss": 1.359375, "loss": 1.8671875, "ref_ce_loss": 0.435546875, "stage_align_loss": 0.00372314453125, "step": 5460 }, { "ce_loss": 0.2197265625, "distill_loss": 0.058837890625, "epoch": 2.388451443569554, "explain_loss": 1.109375, "loss": 1.7265625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.006439208984375, "step": 5460 }, { "epoch": 2.3928258967629046, "grad_norm": 0.6920400857925415, "learning_rate": 2.1604702439128987e-05, "loss": 1.55517578125, "step": 5470 }, { "ce_loss": 0.000858306884765625, "distill_loss": 0.1181640625, "epoch": 2.3928258967629046, "explain_loss": 0.90625, "loss": 1.515625, "ref_ce_loss": 0.484375, "stage_align_loss": 0.005035400390625, "step": 5470 }, { "ce_loss": 0.000530242919921875, "distill_loss": 0.09521484375, "epoch": 2.3928258967629046, "explain_loss": 0.87890625, "loss": 1.234375, "ref_ce_loss": 0.259765625, "stage_align_loss": 0.0030975341796875, "step": 5470 }, { "ce_loss": 0.267578125, "distill_loss": 0.07373046875, "epoch": 2.3928258967629046, "explain_loss": 1.0, "loss": 1.625, "ref_ce_loss": 0.2470703125, "stage_align_loss": 0.039306640625, "step": 5470 }, { "ce_loss": 0.07958984375, "distill_loss": 0.12158203125, "epoch": 2.3928258967629046, "explain_loss": 0.75390625, "loss": 1.3203125, "ref_ce_loss": 0.359375, "stage_align_loss": 0.00396728515625, "step": 5470 }, { "epoch": 2.3972003499562553, "grad_norm": 1.368297815322876, "learning_rate": 2.130624985421098e-05, "loss": 1.59052734375, "step": 5480 }, { "ce_loss": 0.85546875, "distill_loss": 0.06201171875, "epoch": 2.3972003499562553, "explain_loss": 0.9609375, "loss": 2.234375, "ref_ce_loss": 0.34375, "stage_align_loss": 0.004608154296875, "step": 5480 }, { "ce_loss": 0.451171875, "distill_loss": 0.125, "epoch": 2.3972003499562553, "explain_loss": 0.9375, "loss": 2.125, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.09765625, "step": 5480 }, { "ce_loss": 0.0030670166015625, "distill_loss": 0.1220703125, "epoch": 2.3972003499562553, "explain_loss": 1.1875, "loss": 1.859375, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.003082275390625, "step": 5480 }, { "ce_loss": 0.111328125, "distill_loss": 0.1083984375, "epoch": 2.3972003499562553, "explain_loss": 0.8984375, "loss": 1.546875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0030670166015625, "step": 5480 }, { "epoch": 2.4015748031496065, "grad_norm": 0.6924290060997009, "learning_rate": 2.100962709744867e-05, "loss": 1.5921875, "step": 5490 }, { "ce_loss": 0.00113677978515625, "distill_loss": 0.095703125, "epoch": 2.4015748031496065, "explain_loss": 1.03125, "loss": 1.4296875, "ref_ce_loss": 0.296875, "stage_align_loss": 0.0032196044921875, "step": 5490 }, { "ce_loss": 0.00142669677734375, "distill_loss": 0.10009765625, "epoch": 2.4015748031496065, "explain_loss": 1.1875, "loss": 1.53125, "ref_ce_loss": 0.244140625, "stage_align_loss": 0.0027923583984375, "step": 5490 }, { "ce_loss": 0.009521484375, "distill_loss": 0.0791015625, "epoch": 2.4015748031496065, "explain_loss": 0.99609375, "loss": 1.5, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.004730224609375, "step": 5490 }, { "ce_loss": 0.0191650390625, "distill_loss": 0.09326171875, "epoch": 2.4015748031496065, "explain_loss": 0.9453125, "loss": 1.359375, "ref_ce_loss": 0.28515625, "stage_align_loss": 0.017333984375, "step": 5490 }, { "epoch": 2.405949256342957, "grad_norm": 0.9191516041755676, "learning_rate": 2.0714841066069203e-05, "loss": 1.6376953125, "step": 5500 }, { "ce_loss": 0.004974365234375, "distill_loss": 0.111328125, "epoch": 2.405949256342957, "explain_loss": 1.2109375, "loss": 1.75, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.0036163330078125, "step": 5500 }, { "ce_loss": 0.1474609375, "distill_loss": 0.11962890625, "epoch": 2.405949256342957, "explain_loss": 1.2109375, "loss": 1.90625, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.020263671875, "step": 5500 }, { "ce_loss": 0.025146484375, "distill_loss": 0.1005859375, "epoch": 2.405949256342957, "explain_loss": 1.21875, "loss": 1.8671875, "ref_ce_loss": 0.419921875, "stage_align_loss": 0.10009765625, "step": 5500 }, { "ce_loss": 0.0247802734375, "distill_loss": 0.1083984375, "epoch": 2.405949256342957, "explain_loss": 1.1015625, "loss": 1.6171875, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00433349609375, "step": 5500 }, { "epoch": 2.410323709536308, "grad_norm": 1.0503305196762085, "learning_rate": 2.042189861459136e-05, "loss": 1.594091796875, "step": 5510 }, { "ce_loss": 0.047119140625, "distill_loss": 0.087890625, "epoch": 2.410323709536308, "explain_loss": 0.953125, "loss": 1.4921875, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.005401611328125, "step": 5510 }, { "ce_loss": 0.0634765625, "distill_loss": 0.07763671875, "epoch": 2.410323709536308, "explain_loss": 0.94921875, "loss": 1.390625, "ref_ce_loss": 0.283203125, "stage_align_loss": 0.01434326171875, "step": 5510 }, { "ce_loss": 0.0023040771484375, "distill_loss": 0.123046875, "epoch": 2.410323709536308, "explain_loss": 1.3125, "loss": 1.78125, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.0029754638671875, "step": 5510 }, { "ce_loss": 0.06298828125, "distill_loss": 0.0693359375, "epoch": 2.410323709536308, "explain_loss": 0.7890625, "loss": 1.265625, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.00531005859375, "step": 5510 }, { "epoch": 2.4146981627296586, "grad_norm": 0.7208366990089417, "learning_rate": 2.0130806554665804e-05, "loss": 1.5916015625, "step": 5520 }, { "ce_loss": 0.01080322265625, "distill_loss": 0.087890625, "epoch": 2.4146981627296586, "explain_loss": 0.9140625, "loss": 1.3828125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00750732421875, "step": 5520 }, { "ce_loss": 0.000705718994140625, "distill_loss": 0.091796875, "epoch": 2.4146981627296586, "explain_loss": 0.83984375, "loss": 1.2734375, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.00408935546875, "step": 5520 }, { "ce_loss": 0.00142669677734375, "distill_loss": 0.08349609375, "epoch": 2.4146981627296586, "explain_loss": 1.40625, "loss": 1.8125, "ref_ce_loss": 0.314453125, "stage_align_loss": 0.004180908203125, "step": 5520 }, { "ce_loss": 0.1962890625, "distill_loss": 0.10009765625, "epoch": 2.4146981627296586, "explain_loss": 1.296875, "loss": 2.046875, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.0947265625, "step": 5520 }, { "epoch": 2.41907261592301, "grad_norm": 1.2343124151229858, "learning_rate": 1.9841571654917148e-05, "loss": 1.63583984375, "step": 5530 }, { "ce_loss": 0.2265625, "distill_loss": 0.10498046875, "epoch": 2.41907261592301, "explain_loss": 1.4140625, "loss": 2.40625, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.11376953125, "step": 5530 }, { "ce_loss": 0.17578125, "distill_loss": 0.09814453125, "epoch": 2.41907261592301, "explain_loss": 0.8515625, "loss": 1.421875, "ref_ce_loss": 0.2099609375, "stage_align_loss": 0.09033203125, "step": 5530 }, { "ce_loss": 0.275390625, "distill_loss": 0.0751953125, "epoch": 2.41907261592301, "explain_loss": 1.0234375, "loss": 1.8359375, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.09423828125, "step": 5530 }, { "ce_loss": 0.0216064453125, "distill_loss": 0.09716796875, "epoch": 2.41907261592301, "explain_loss": 1.53125, "loss": 2.171875, "ref_ce_loss": 0.515625, "stage_align_loss": 0.0037078857421875, "step": 5530 }, { "epoch": 2.4234470691163605, "grad_norm": 0.9970301389694214, "learning_rate": 1.9554200640786235e-05, "loss": 1.6224609375, "step": 5540 }, { "ce_loss": 0.00555419921875, "distill_loss": 0.09716796875, "epoch": 2.4234470691163605, "explain_loss": 0.890625, "loss": 1.359375, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.002685546875, "step": 5540 }, { "ce_loss": 0.0184326171875, "distill_loss": 0.0927734375, "epoch": 2.4234470691163605, "explain_loss": 1.0, "loss": 1.4453125, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.0030670166015625, "step": 5540 }, { "ce_loss": 0.01153564453125, "distill_loss": 0.0654296875, "epoch": 2.4234470691163605, "explain_loss": 1.078125, "loss": 1.453125, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.003204345703125, "step": 5540 }, { "ce_loss": 0.01171875, "distill_loss": 0.1064453125, "epoch": 2.4234470691163605, "explain_loss": 1.0, "loss": 1.5546875, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.00390625, "step": 5540 }, { "epoch": 2.4278215223097113, "grad_norm": 0.7592344880104065, "learning_rate": 1.926870019437389e-05, "loss": 1.59638671875, "step": 5550 }, { "ce_loss": 0.000545501708984375, "distill_loss": 0.0673828125, "epoch": 2.4278215223097113, "explain_loss": 1.2265625, "loss": 1.65625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.01446533203125, "step": 5550 }, { "ce_loss": 0.0031890869140625, "distill_loss": 0.083984375, "epoch": 2.4278215223097113, "explain_loss": 1.1875, "loss": 1.578125, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.003021240234375, "step": 5550 }, { "ce_loss": 0.006103515625, "distill_loss": 0.09375, "epoch": 2.4278215223097113, "explain_loss": 1.140625, "loss": 1.625, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.004241943359375, "step": 5550 }, { "ce_loss": 0.0036468505859375, "distill_loss": 0.07666015625, "epoch": 2.4278215223097113, "explain_loss": 0.8359375, "loss": 1.328125, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.0031890869140625, "step": 5550 }, { "epoch": 2.432195975503062, "grad_norm": 1.2753320932388306, "learning_rate": 1.8985076954285496e-05, "loss": 1.62529296875, "step": 5560 }, { "ce_loss": 0.00634765625, "distill_loss": 0.09814453125, "epoch": 2.432195975503062, "explain_loss": 1.21875, "loss": 1.7578125, "ref_ce_loss": 0.431640625, "stage_align_loss": 0.0038604736328125, "step": 5560 }, { "ce_loss": 0.0189208984375, "distill_loss": 0.0888671875, "epoch": 2.432195975503062, "explain_loss": 1.171875, "loss": 1.625, "ref_ce_loss": 0.34375, "stage_align_loss": 0.0038299560546875, "step": 5560 }, { "ce_loss": 0.0186767578125, "distill_loss": 0.0849609375, "epoch": 2.432195975503062, "explain_loss": 1.1328125, "loss": 1.8125, "ref_ce_loss": 0.578125, "stage_align_loss": 0.0035552978515625, "step": 5560 }, { "ce_loss": 0.291015625, "distill_loss": 0.0986328125, "epoch": 2.432195975503062, "explain_loss": 1.390625, "loss": 2.3125, "ref_ce_loss": 0.484375, "stage_align_loss": 0.0556640625, "step": 5560 }, { "epoch": 2.436570428696413, "grad_norm": 1.0244991779327393, "learning_rate": 1.8703337515476725e-05, "loss": 1.610693359375, "step": 5570 }, { "ce_loss": 0.00927734375, "distill_loss": 0.07275390625, "epoch": 2.436570428696413, "explain_loss": 0.6796875, "loss": 1.171875, "ref_ce_loss": 0.40625, "stage_align_loss": 0.0034942626953125, "step": 5570 }, { "ce_loss": 0.00531005859375, "distill_loss": 0.083984375, "epoch": 2.436570428696413, "explain_loss": 0.9375, "loss": 1.4296875, "ref_ce_loss": 0.3984375, "stage_align_loss": 0.0037994384765625, "step": 5570 }, { "ce_loss": 0.00135040283203125, "distill_loss": 0.07275390625, "epoch": 2.436570428696413, "explain_loss": 0.8125, "loss": 1.265625, "ref_ce_loss": 0.375, "stage_align_loss": 0.0030670166015625, "step": 5570 }, { "ce_loss": 0.00022792816162109375, "distill_loss": 0.06640625, "epoch": 2.436570428696413, "explain_loss": 1.0390625, "loss": 1.53125, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0038299560546875, "step": 5570 }, { "epoch": 2.440944881889764, "grad_norm": 0.610280454158783, "learning_rate": 1.8423488429100055e-05, "loss": 1.52353515625, "step": 5580 }, { "ce_loss": 0.20703125, "distill_loss": 0.1083984375, "epoch": 2.440944881889764, "explain_loss": 0.84765625, "loss": 1.46875, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.024658203125, "step": 5580 }, { "ce_loss": 0.1337890625, "distill_loss": 0.10546875, "epoch": 2.440944881889764, "explain_loss": 0.74609375, "loss": 1.34375, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.004425048828125, "step": 5580 }, { "ce_loss": 0.00225830078125, "distill_loss": 0.07861328125, "epoch": 2.440944881889764, "explain_loss": 0.8515625, "loss": 1.21875, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.003143310546875, "step": 5580 }, { "ce_loss": 0.00188446044921875, "distill_loss": 0.0673828125, "epoch": 2.440944881889764, "explain_loss": 1.03125, "loss": 1.375, "ref_ce_loss": 0.265625, "stage_align_loss": 0.003326416015625, "step": 5580 }, { "epoch": 2.4453193350831146, "grad_norm": 0.726874053478241, "learning_rate": 1.8145536202352488e-05, "loss": 1.60458984375, "step": 5590 }, { "ce_loss": 0.703125, "distill_loss": 0.10498046875, "epoch": 2.4453193350831146, "explain_loss": 1.0859375, "loss": 2.34375, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.09130859375, "step": 5590 }, { "ce_loss": 0.1201171875, "distill_loss": 0.0693359375, "epoch": 2.4453193350831146, "explain_loss": 0.84375, "loss": 1.4296875, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.006683349609375, "step": 5590 }, { "ce_loss": 0.08984375, "distill_loss": 0.0712890625, "epoch": 2.4453193350831146, "explain_loss": 1.0546875, "loss": 1.546875, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.0047607421875, "step": 5590 }, { "ce_loss": 0.0008392333984375, "distill_loss": 0.091796875, "epoch": 2.4453193350831146, "explain_loss": 0.8203125, "loss": 1.234375, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.00341796875, "step": 5590 }, { "epoch": 2.4496937882764653, "grad_norm": 0.7568503618240356, "learning_rate": 1.786948729832434e-05, "loss": 1.5486328125, "step": 5600 }, { "ce_loss": 0.349609375, "distill_loss": 0.080078125, "epoch": 2.4496937882764653, "explain_loss": 0.9453125, "loss": 2.0, "ref_ce_loss": 0.5234375, "stage_align_loss": 0.1103515625, "step": 5600 }, { "ce_loss": 0.08154296875, "distill_loss": 0.06787109375, "epoch": 2.4496937882764653, "explain_loss": 0.7265625, "loss": 1.21875, "ref_ce_loss": 0.287109375, "stage_align_loss": 0.052001953125, "step": 5600 }, { "ce_loss": 0.00141143798828125, "distill_loss": 0.0830078125, "epoch": 2.4496937882764653, "explain_loss": 0.96875, "loss": 1.4453125, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.00360107421875, "step": 5600 }, { "ce_loss": 0.009033203125, "distill_loss": 0.0615234375, "epoch": 2.4496937882764653, "explain_loss": 1.421875, "loss": 1.859375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.0030059814453125, "step": 5600 }, { "epoch": 2.454068241469816, "grad_norm": 1.2103965282440186, "learning_rate": 1.75953481358488e-05, "loss": 1.56474609375, "step": 5610 }, { "ce_loss": 0.006683349609375, "distill_loss": 0.1044921875, "epoch": 2.454068241469816, "explain_loss": 0.8671875, "loss": 1.3203125, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.004547119140625, "step": 5610 }, { "ce_loss": 0.046630859375, "distill_loss": 0.1142578125, "epoch": 2.454068241469816, "explain_loss": 0.90625, "loss": 1.546875, "ref_ce_loss": 0.4765625, "stage_align_loss": 0.0025634765625, "step": 5610 }, { "ce_loss": 0.000705718994140625, "distill_loss": 0.07861328125, "epoch": 2.454068241469816, "explain_loss": 0.94140625, "loss": 1.4375, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.002471923828125, "step": 5610 }, { "ce_loss": 0.00156402587890625, "distill_loss": 0.07861328125, "epoch": 2.454068241469816, "explain_loss": 0.953125, "loss": 1.328125, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.0038604736328125, "step": 5610 }, { "epoch": 2.4584426946631672, "grad_norm": 0.6746156811714172, "learning_rate": 1.7323125089352776e-05, "loss": 1.6392578125, "step": 5620 }, { "ce_loss": 0.0031890869140625, "distill_loss": 0.107421875, "epoch": 2.4584426946631672, "explain_loss": 1.34375, "loss": 1.796875, "ref_ce_loss": 0.34375, "stage_align_loss": 0.0028533935546875, "step": 5620 }, { "ce_loss": 0.025146484375, "distill_loss": 0.06103515625, "epoch": 2.4584426946631672, "explain_loss": 0.8203125, "loss": 1.1328125, "ref_ce_loss": 0.22265625, "stage_align_loss": 0.0034027099609375, "step": 5620 }, { "ce_loss": 0.00125885009765625, "distill_loss": 0.1103515625, "epoch": 2.4584426946631672, "explain_loss": 1.40625, "loss": 1.78125, "ref_ce_loss": 0.263671875, "stage_align_loss": 0.0029144287109375, "step": 5620 }, { "ce_loss": 0.001220703125, "distill_loss": 0.0751953125, "epoch": 2.4584426946631672, "explain_loss": 0.78515625, "loss": 1.265625, "ref_ce_loss": 0.400390625, "stage_align_loss": 0.0023651123046875, "step": 5620 }, { "epoch": 2.462817147856518, "grad_norm": 1.1218650341033936, "learning_rate": 1.7052824488708618e-05, "loss": 1.61572265625, "step": 5630 }, { "ce_loss": 0.0014801025390625, "distill_loss": 0.0810546875, "epoch": 2.462817147856518, "explain_loss": 0.84375, "loss": 1.2734375, "ref_ce_loss": 0.34375, "stage_align_loss": 0.0026702880859375, "step": 5630 }, { "ce_loss": 0.0012664794921875, "distill_loss": 0.11181640625, "epoch": 2.462817147856518, "explain_loss": 1.0, "loss": 1.5078125, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0036773681640625, "step": 5630 }, { "ce_loss": 1.1875, "distill_loss": 0.0751953125, "epoch": 2.462817147856518, "explain_loss": 1.3125, "loss": 3.0625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.0986328125, "step": 5630 }, { "ce_loss": 0.009521484375, "distill_loss": 0.0869140625, "epoch": 2.462817147856518, "explain_loss": 1.1875, "loss": 1.5234375, "ref_ce_loss": 0.2392578125, "stage_align_loss": 0.0026092529296875, "step": 5630 }, { "epoch": 2.4671916010498687, "grad_norm": 0.7106099724769592, "learning_rate": 1.678445261908702e-05, "loss": 1.5701171875, "step": 5640 }, { "ce_loss": 0.15234375, "distill_loss": 0.0888671875, "epoch": 2.4671916010498687, "explain_loss": 0.9921875, "loss": 1.796875, "ref_ce_loss": 0.498046875, "stage_align_loss": 0.07177734375, "step": 5640 }, { "ce_loss": 0.373046875, "distill_loss": 0.1240234375, "epoch": 2.4671916010498687, "explain_loss": 1.0390625, "loss": 2.03125, "ref_ce_loss": 0.462890625, "stage_align_loss": 0.0277099609375, "step": 5640 }, { "ce_loss": 0.00830078125, "distill_loss": 0.142578125, "epoch": 2.4671916010498687, "explain_loss": 1.3515625, "loss": 1.921875, "ref_ce_loss": 0.4140625, "stage_align_loss": 0.00677490234375, "step": 5640 }, { "ce_loss": 0.0089111328125, "distill_loss": 0.083984375, "epoch": 2.4671916010498687, "explain_loss": 0.74609375, "loss": 1.1640625, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.0252685546875, "step": 5640 }, { "epoch": 2.47156605424322, "grad_norm": 0.8831452131271362, "learning_rate": 1.6518015720810775e-05, "loss": 1.556396484375, "step": 5650 }, { "ce_loss": 0.05615234375, "distill_loss": 0.1259765625, "epoch": 2.47156605424322, "explain_loss": 0.953125, "loss": 1.65625, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.01171875, "step": 5650 }, { "ce_loss": 0.0010528564453125, "distill_loss": 0.07275390625, "epoch": 2.47156605424322, "explain_loss": 1.3671875, "loss": 1.8828125, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.002685546875, "step": 5650 }, { "ce_loss": 0.001068115234375, "distill_loss": 0.083984375, "epoch": 2.47156605424322, "explain_loss": 0.84375, "loss": 1.296875, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.0032196044921875, "step": 5650 }, { "ce_loss": 0.00159454345703125, "distill_loss": 0.0947265625, "epoch": 2.47156605424322, "explain_loss": 0.9140625, "loss": 1.3125, "ref_ce_loss": 0.294921875, "stage_align_loss": 0.0040283203125, "step": 5650 }, { "epoch": 2.4759405074365706, "grad_norm": 1.0877931118011475, "learning_rate": 1.625351998920971e-05, "loss": 1.567431640625, "step": 5660 }, { "ce_loss": 0.000911712646484375, "distill_loss": 0.08251953125, "epoch": 2.4759405074365706, "explain_loss": 0.9375, "loss": 1.4375, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.0036163330078125, "step": 5660 }, { "ce_loss": 0.00244140625, "distill_loss": 0.095703125, "epoch": 2.4759405074365706, "explain_loss": 1.2734375, "loss": 1.765625, "ref_ce_loss": 0.392578125, "stage_align_loss": 0.0023956298828125, "step": 5660 }, { "ce_loss": 0.004608154296875, "distill_loss": 0.09375, "epoch": 2.4759405074365706, "explain_loss": 0.9765625, "loss": 1.5234375, "ref_ce_loss": 0.4453125, "stage_align_loss": 0.003936767578125, "step": 5660 }, { "ce_loss": 0.494140625, "distill_loss": 0.10693359375, "epoch": 2.4759405074365706, "explain_loss": 1.078125, "loss": 2.046875, "ref_ce_loss": 0.25390625, "stage_align_loss": 0.1181640625, "step": 5660 }, { "epoch": 2.4803149606299213, "grad_norm": 0.7851812839508057, "learning_rate": 1.599097157447662e-05, "loss": 1.638671875, "step": 5670 }, { "ce_loss": 0.00131988525390625, "distill_loss": 0.09326171875, "epoch": 2.4803149606299213, "explain_loss": 0.79296875, "loss": 1.1015625, "ref_ce_loss": 0.21484375, "stage_align_loss": 0.002471923828125, "step": 5670 }, { "ce_loss": 0.19921875, "distill_loss": 0.058837890625, "epoch": 2.4803149606299213, "explain_loss": 1.765625, "loss": 2.453125, "ref_ce_loss": 0.421875, "stage_align_loss": 0.004730224609375, "step": 5670 }, { "ce_loss": 0.0027008056640625, "distill_loss": 0.12890625, "epoch": 2.4803149606299213, "explain_loss": 0.69921875, "loss": 1.0859375, "ref_ce_loss": 0.240234375, "stage_align_loss": 0.01385498046875, "step": 5670 }, { "ce_loss": 0.0198974609375, "distill_loss": 0.09814453125, "epoch": 2.4803149606299213, "explain_loss": 1.015625, "loss": 1.546875, "ref_ce_loss": 0.384765625, "stage_align_loss": 0.0242919921875, "step": 5670 }, { "epoch": 2.484689413823272, "grad_norm": 0.9440992474555969, "learning_rate": 1.5730376581524363e-05, "loss": 1.576953125, "step": 5680 }, { "ce_loss": 0.00115203857421875, "distill_loss": 0.09716796875, "epoch": 2.484689413823272, "explain_loss": 0.99609375, "loss": 1.4453125, "ref_ce_loss": 0.345703125, "stage_align_loss": 0.003326416015625, "step": 5680 }, { "ce_loss": 0.0908203125, "distill_loss": 0.08544921875, "epoch": 2.484689413823272, "explain_loss": 1.0078125, "loss": 1.65625, "ref_ce_loss": 0.451171875, "stage_align_loss": 0.0267333984375, "step": 5680 }, { "ce_loss": 0.009521484375, "distill_loss": 0.09326171875, "epoch": 2.484689413823272, "explain_loss": 1.09375, "loss": 1.59375, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.00335693359375, "step": 5680 }, { "ce_loss": 0.03125, "distill_loss": 0.1064453125, "epoch": 2.484689413823272, "explain_loss": 1.484375, "loss": 2.0, "ref_ce_loss": 0.375, "stage_align_loss": 0.0030670166015625, "step": 5680 }, { "epoch": 2.4890638670166227, "grad_norm": 0.9038488268852234, "learning_rate": 1.547174106984365e-05, "loss": 1.55830078125, "step": 5690 }, { "ce_loss": 0.00347900390625, "distill_loss": 0.09033203125, "epoch": 2.4890638670166227, "explain_loss": 1.28125, "loss": 1.7734375, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0030059814453125, "step": 5690 }, { "ce_loss": 0.026123046875, "distill_loss": 0.078125, "epoch": 2.4890638670166227, "explain_loss": 0.7734375, "loss": 1.359375, "ref_ce_loss": 0.40625, "stage_align_loss": 0.06982421875, "step": 5690 }, { "ce_loss": 0.1552734375, "distill_loss": 0.0908203125, "epoch": 2.4890638670166227, "explain_loss": 1.0625, "loss": 1.671875, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.002716064453125, "step": 5690 }, { "ce_loss": 0.000774383544921875, "distill_loss": 0.1123046875, "epoch": 2.4890638670166227, "explain_loss": 1.234375, "loss": 1.78125, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0027313232421875, "step": 5690 }, { "epoch": 2.493438320209974, "grad_norm": 0.8078233003616333, "learning_rate": 1.5215071053362462e-05, "loss": 1.583935546875, "step": 5700 }, { "ce_loss": 0.0038604736328125, "distill_loss": 0.08203125, "epoch": 2.493438320209974, "explain_loss": 1.140625, "loss": 1.5546875, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0031280517578125, "step": 5700 }, { "ce_loss": 0.0201416015625, "distill_loss": 0.0849609375, "epoch": 2.493438320209974, "explain_loss": 1.09375, "loss": 1.5, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.0030975341796875, "step": 5700 }, { "ce_loss": 0.275390625, "distill_loss": 0.1025390625, "epoch": 2.493438320209974, "explain_loss": 0.921875, "loss": 1.8125, "ref_ce_loss": 0.478515625, "stage_align_loss": 0.029052734375, "step": 5700 }, { "ce_loss": 0.00360107421875, "distill_loss": 0.0849609375, "epoch": 2.493438320209974, "explain_loss": 1.0390625, "loss": 1.453125, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0028228759765625, "step": 5700 }, { "epoch": 2.4978127734033246, "grad_norm": 0.5919866561889648, "learning_rate": 1.4960372500305997e-05, "loss": 1.5927734375, "step": 5710 }, { "ce_loss": 0.1376953125, "distill_loss": 0.08056640625, "epoch": 2.4978127734033246, "explain_loss": 1.15625, "loss": 1.9140625, "ref_ce_loss": 0.4609375, "stage_align_loss": 0.07763671875, "step": 5710 }, { "ce_loss": 0.001251220703125, "distill_loss": 0.09375, "epoch": 2.4978127734033246, "explain_loss": 1.078125, "loss": 1.4765625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.0025787353515625, "step": 5710 }, { "ce_loss": 0.0037994384765625, "distill_loss": 0.0751953125, "epoch": 2.4978127734033246, "explain_loss": 1.0, "loss": 1.421875, "ref_ce_loss": 0.33984375, "stage_align_loss": 0.00299072265625, "step": 5710 }, { "ce_loss": 0.02880859375, "distill_loss": 0.115234375, "epoch": 2.4978127734033246, "explain_loss": 0.85546875, "loss": 1.3984375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.00299072265625, "step": 5710 }, { "epoch": 2.5021872265966754, "grad_norm": 0.7495662569999695, "learning_rate": 1.4707651333057926e-05, "loss": 1.5697265625, "step": 5720 }, { "ce_loss": 0.02099609375, "distill_loss": 0.0791015625, "epoch": 2.5021872265966754, "explain_loss": 1.0703125, "loss": 1.65625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.00262451171875, "step": 5720 }, { "ce_loss": 0.220703125, "distill_loss": 0.10498046875, "epoch": 2.5021872265966754, "explain_loss": 0.890625, "loss": 1.78125, "ref_ce_loss": 0.48828125, "stage_align_loss": 0.07470703125, "step": 5720 }, { "ce_loss": 0.09814453125, "distill_loss": 0.08203125, "epoch": 2.5021872265966754, "explain_loss": 1.0234375, "loss": 1.6328125, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.00787353515625, "step": 5720 }, { "ce_loss": 0.032470703125, "distill_loss": 0.080078125, "epoch": 2.5021872265966754, "explain_loss": 1.0, "loss": 1.46875, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.023681640625, "step": 5720 }, { "epoch": 2.506561679790026, "grad_norm": 0.7109233140945435, "learning_rate": 1.445691342802279e-05, "loss": 1.5775390625, "step": 5730 }, { "ce_loss": 0.0007476806640625, "distill_loss": 0.0693359375, "epoch": 2.506561679790026, "explain_loss": 1.28125, "loss": 1.65625, "ref_ce_loss": 0.30078125, "stage_align_loss": 0.002471923828125, "step": 5730 }, { "ce_loss": 0.004608154296875, "distill_loss": 0.07421875, "epoch": 2.506561679790026, "explain_loss": 1.234375, "loss": 1.671875, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.0025787353515625, "step": 5730 }, { "ce_loss": 0.00124359130859375, "distill_loss": 0.1328125, "epoch": 2.506561679790026, "explain_loss": 1.015625, "loss": 1.59375, "ref_ce_loss": 0.443359375, "stage_align_loss": 0.0033111572265625, "step": 5730 }, { "ce_loss": 0.1220703125, "distill_loss": 0.1044921875, "epoch": 2.506561679790026, "explain_loss": 1.0078125, "loss": 1.5625, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.00421142578125, "step": 5730 }, { "epoch": 2.5109361329833773, "grad_norm": 0.7214124202728271, "learning_rate": 1.4208164615489217e-05, "loss": 1.66025390625, "step": 5740 }, { "ce_loss": 0.025146484375, "distill_loss": 0.08251953125, "epoch": 2.5109361329833773, "explain_loss": 1.40625, "loss": 1.8671875, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.0034637451171875, "step": 5740 }, { "ce_loss": 0.0205078125, "distill_loss": 0.083984375, "epoch": 2.5109361329833773, "explain_loss": 0.9921875, "loss": 1.4921875, "ref_ce_loss": 0.390625, "stage_align_loss": 0.0048828125, "step": 5740 }, { "ce_loss": 0.01385498046875, "distill_loss": 0.12109375, "epoch": 2.5109361329833773, "explain_loss": 1.28125, "loss": 1.7265625, "ref_ce_loss": 0.306640625, "stage_align_loss": 0.0030975341796875, "step": 5740 }, { "ce_loss": 0.1787109375, "distill_loss": 0.09521484375, "epoch": 2.5109361329833773, "explain_loss": 1.1875, "loss": 2.078125, "ref_ce_loss": 0.5390625, "stage_align_loss": 0.078125, "step": 5740 }, { "epoch": 2.515310586176728, "grad_norm": 0.9584798812866211, "learning_rate": 1.3961410679494457e-05, "loss": 1.6951171875, "step": 5750 }, { "ce_loss": 0.1748046875, "distill_loss": 0.134765625, "epoch": 2.515310586176728, "explain_loss": 1.0390625, "loss": 1.75, "ref_ce_loss": 0.3359375, "stage_align_loss": 0.07177734375, "step": 5750 }, { "ce_loss": 0.166015625, "distill_loss": 0.107421875, "epoch": 2.515310586176728, "explain_loss": 0.6328125, "loss": 1.203125, "ref_ce_loss": 0.298828125, "stage_align_loss": 0.005706787109375, "step": 5750 }, { "ce_loss": 0.01171875, "distill_loss": 0.07861328125, "epoch": 2.515310586176728, "explain_loss": 0.734375, "loss": 1.1875, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.0028228759765625, "step": 5750 }, { "ce_loss": 0.0032958984375, "distill_loss": 0.0732421875, "epoch": 2.515310586176728, "explain_loss": 0.765625, "loss": 1.0625, "ref_ce_loss": 0.216796875, "stage_align_loss": 0.0026092529296875, "step": 5750 }, { "epoch": 2.5196850393700787, "grad_norm": 0.6906399130821228, "learning_rate": 1.3716657357689789e-05, "loss": 1.56708984375, "step": 5760 }, { "ce_loss": 0.00860595703125, "distill_loss": 0.1181640625, "epoch": 2.5196850393700787, "explain_loss": 0.796875, "loss": 1.1953125, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.002685546875, "step": 5760 }, { "ce_loss": 0.0037841796875, "distill_loss": 0.107421875, "epoch": 2.5196850393700787, "explain_loss": 1.546875, "loss": 2.1875, "ref_ce_loss": 0.53125, "stage_align_loss": 0.004730224609375, "step": 5760 }, { "ce_loss": 0.158203125, "distill_loss": 0.0791015625, "epoch": 2.5196850393700787, "explain_loss": 1.203125, "loss": 1.8125, "ref_ce_loss": 0.35546875, "stage_align_loss": 0.0179443359375, "step": 5760 }, { "ce_loss": 0.00390625, "distill_loss": 0.072265625, "epoch": 2.5196850393700787, "explain_loss": 1.2734375, "loss": 1.703125, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.002838134765625, "step": 5760 }, { "epoch": 2.5240594925634294, "grad_norm": 0.5636615753173828, "learning_rate": 1.3473910341207274e-05, "loss": 1.601806640625, "step": 5770 }, { "ce_loss": 0.0167236328125, "distill_loss": 0.0849609375, "epoch": 2.5240594925634294, "explain_loss": 1.34375, "loss": 1.7734375, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.01190185546875, "step": 5770 }, { "ce_loss": 0.000934600830078125, "distill_loss": 0.138671875, "epoch": 2.5240594925634294, "explain_loss": 1.25, "loss": 1.765625, "ref_ce_loss": 0.3671875, "stage_align_loss": 0.002838134765625, "step": 5770 }, { "ce_loss": 0.03515625, "distill_loss": 0.11669921875, "epoch": 2.5240594925634294, "explain_loss": 0.8984375, "loss": 1.484375, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.0034637451171875, "step": 5770 }, { "ce_loss": 0.0005340576171875, "distill_loss": 0.103515625, "epoch": 2.5240594925634294, "explain_loss": 1.0703125, "loss": 1.546875, "ref_ce_loss": 0.373046875, "stage_align_loss": 0.0037689208984375, "step": 5770 }, { "epoch": 2.52843394575678, "grad_norm": 0.8804066181182861, "learning_rate": 1.3233175274527209e-05, "loss": 1.59697265625, "step": 5780 }, { "ce_loss": 0.033203125, "distill_loss": 0.107421875, "epoch": 2.52843394575678, "explain_loss": 1.140625, "loss": 1.7421875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.039306640625, "step": 5780 }, { "ce_loss": 0.03564453125, "distill_loss": 0.115234375, "epoch": 2.52843394575678, "explain_loss": 1.203125, "loss": 1.734375, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.00531005859375, "step": 5780 }, { "ce_loss": 0.046630859375, "distill_loss": 0.07763671875, "epoch": 2.52843394575678, "explain_loss": 0.84765625, "loss": 1.265625, "ref_ce_loss": 0.251953125, "stage_align_loss": 0.03857421875, "step": 5780 }, { "ce_loss": 0.007110595703125, "distill_loss": 0.111328125, "epoch": 2.52843394575678, "explain_loss": 0.91015625, "loss": 1.4296875, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0029754638671875, "step": 5780 }, { "epoch": 2.5328083989501313, "grad_norm": 0.6295313835144043, "learning_rate": 1.2994457755347045e-05, "loss": 1.55625, "step": 5790 }, { "ce_loss": 0.01446533203125, "distill_loss": 0.09375, "epoch": 2.5328083989501313, "explain_loss": 0.96484375, "loss": 1.59375, "ref_ce_loss": 0.51171875, "stage_align_loss": 0.004119873046875, "step": 5790 }, { "ce_loss": 0.04736328125, "distill_loss": 0.06884765625, "epoch": 2.5328083989501313, "explain_loss": 0.8203125, "loss": 1.375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.0654296875, "step": 5790 }, { "ce_loss": 0.279296875, "distill_loss": 0.087890625, "epoch": 2.5328083989501313, "explain_loss": 0.9296875, "loss": 1.859375, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.087890625, "step": 5790 }, { "ce_loss": 0.08154296875, "distill_loss": 0.107421875, "epoch": 2.5328083989501313, "explain_loss": 1.0625, "loss": 1.59375, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.0184326171875, "step": 5790 }, { "epoch": 2.537182852143482, "grad_norm": 0.5792137384414673, "learning_rate": 1.2757763334451112e-05, "loss": 1.5884765625, "step": 5800 }, { "ce_loss": 0.00066375732421875, "distill_loss": 0.08251953125, "epoch": 2.537182852143482, "explain_loss": 1.25, "loss": 1.59375, "ref_ce_loss": 0.255859375, "stage_align_loss": 0.00286865234375, "step": 5800 }, { "ce_loss": 0.00162506103515625, "distill_loss": 0.07861328125, "epoch": 2.537182852143482, "explain_loss": 0.828125, "loss": 1.3515625, "ref_ce_loss": 0.439453125, "stage_align_loss": 0.00390625, "step": 5800 }, { "ce_loss": 0.373046875, "distill_loss": 0.1201171875, "epoch": 2.537182852143482, "explain_loss": 1.015625, "loss": 1.859375, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.011962890625, "step": 5800 }, { "ce_loss": 0.0011749267578125, "distill_loss": 0.087890625, "epoch": 2.537182852143482, "explain_loss": 1.15625, "loss": 1.5, "ref_ce_loss": 0.255859375, "stage_align_loss": 0.0033416748046875, "step": 5800 }, { "epoch": 2.5415573053368328, "grad_norm": 0.8605143427848816, "learning_rate": 1.2523097515581683e-05, "loss": 1.58701171875, "step": 5810 }, { "ce_loss": 0.03759765625, "distill_loss": 0.0732421875, "epoch": 2.5415573053368328, "explain_loss": 0.96484375, "loss": 1.5625, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.003692626953125, "step": 5810 }, { "ce_loss": 0.11083984375, "distill_loss": 0.083984375, "epoch": 2.5415573053368328, "explain_loss": 1.046875, "loss": 1.578125, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.012451171875, "step": 5810 }, { "ce_loss": 0.01806640625, "distill_loss": 0.1142578125, "epoch": 2.5415573053368328, "explain_loss": 1.0625, "loss": 1.6328125, "ref_ce_loss": 0.43359375, "stage_align_loss": 0.004241943359375, "step": 5810 }, { "ce_loss": 0.4375, "distill_loss": 0.091796875, "epoch": 2.5415573053368328, "explain_loss": 1.265625, "loss": 2.265625, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.1025390625, "step": 5810 }, { "epoch": 2.545931758530184, "grad_norm": 1.2693125009536743, "learning_rate": 1.2290465755310842e-05, "loss": 1.649609375, "step": 5820 }, { "ce_loss": 0.00469970703125, "distill_loss": 0.08056640625, "epoch": 2.545931758530184, "explain_loss": 1.234375, "loss": 1.6171875, "ref_ce_loss": 0.29296875, "stage_align_loss": 0.0025482177734375, "step": 5820 }, { "ce_loss": 0.00848388671875, "distill_loss": 0.06103515625, "epoch": 2.545931758530184, "explain_loss": 0.75, "loss": 1.140625, "ref_ce_loss": 0.322265625, "stage_align_loss": 0.003875732421875, "step": 5820 }, { "ce_loss": 0.0201416015625, "distill_loss": 0.10693359375, "epoch": 2.545931758530184, "explain_loss": 0.984375, "loss": 1.421875, "ref_ce_loss": 0.3046875, "stage_align_loss": 0.0031890869140625, "step": 5820 }, { "ce_loss": 0.302734375, "distill_loss": 0.09521484375, "epoch": 2.545931758530184, "explain_loss": 0.498046875, "loss": 1.3203125, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.09814453125, "step": 5820 }, { "epoch": 2.5503062117235347, "grad_norm": 1.775122046470642, "learning_rate": 1.2059873462913728e-05, "loss": 1.56337890625, "step": 5830 }, { "ce_loss": 0.0028076171875, "distill_loss": 0.08251953125, "epoch": 2.5503062117235347, "explain_loss": 1.1640625, "loss": 1.484375, "ref_ce_loss": 0.228515625, "stage_align_loss": 0.00439453125, "step": 5830 }, { "ce_loss": 0.0010223388671875, "distill_loss": 0.055908203125, "epoch": 2.5503062117235347, "explain_loss": 0.75390625, "loss": 1.234375, "ref_ce_loss": 0.42578125, "stage_align_loss": 0.0027618408203125, "step": 5830 }, { "ce_loss": 0.025634765625, "distill_loss": 0.08056640625, "epoch": 2.5503062117235347, "explain_loss": 0.875, "loss": 1.34375, "ref_ce_loss": 0.3203125, "stage_align_loss": 0.04052734375, "step": 5830 }, { "ce_loss": 0.00103759765625, "distill_loss": 0.1044921875, "epoch": 2.5503062117235347, "explain_loss": 1.0703125, "loss": 1.5078125, "ref_ce_loss": 0.328125, "stage_align_loss": 0.00274658203125, "step": 5830 }, { "epoch": 2.5546806649168854, "grad_norm": 0.7070197463035583, "learning_rate": 1.1831326000242659e-05, "loss": 1.59638671875, "step": 5840 }, { "ce_loss": 0.00665283203125, "distill_loss": 0.08349609375, "epoch": 2.5546806649168854, "explain_loss": 0.8671875, "loss": 1.21875, "ref_ce_loss": 0.26171875, "stage_align_loss": 0.0031890869140625, "step": 5840 }, { "ce_loss": 0.0228271484375, "distill_loss": 0.0615234375, "epoch": 2.5546806649168854, "explain_loss": 1.171875, "loss": 1.5703125, "ref_ce_loss": 0.30859375, "stage_align_loss": 0.0029754638671875, "step": 5840 }, { "ce_loss": 0.55859375, "distill_loss": 0.0712890625, "epoch": 2.5546806649168854, "explain_loss": 1.03125, "loss": 1.953125, "ref_ce_loss": 0.1875, "stage_align_loss": 0.10986328125, "step": 5840 }, { "ce_loss": 0.0032196044921875, "distill_loss": 0.09326171875, "epoch": 2.5546806649168854, "explain_loss": 1.4375, "loss": 1.890625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.00262451171875, "step": 5840 }, { "epoch": 2.559055118110236, "grad_norm": 0.5842452645301819, "learning_rate": 1.1604828681602597e-05, "loss": 1.58828125, "step": 5850 }, { "ce_loss": 0.000675201416015625, "distill_loss": 0.072265625, "epoch": 2.559055118110236, "explain_loss": 1.109375, "loss": 1.515625, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.0030364990234375, "step": 5850 }, { "ce_loss": 0.002227783203125, "distill_loss": 0.0966796875, "epoch": 2.559055118110236, "explain_loss": 1.125, "loss": 1.5, "ref_ce_loss": 0.26953125, "stage_align_loss": 0.002777099609375, "step": 5850 }, { "ce_loss": 0.06884765625, "distill_loss": 0.0888671875, "epoch": 2.559055118110236, "explain_loss": 1.21875, "loss": 1.9375, "ref_ce_loss": 0.5546875, "stage_align_loss": 0.006195068359375, "step": 5850 }, { "ce_loss": 0.0014801025390625, "distill_loss": 0.1181640625, "epoch": 2.559055118110236, "explain_loss": 1.34375, "loss": 1.859375, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0034637451171875, "step": 5850 }, { "epoch": 2.563429571303587, "grad_norm": 0.8402511477470398, "learning_rate": 1.1380386773627338e-05, "loss": 1.576953125, "step": 5860 }, { "ce_loss": 0.0036163330078125, "distill_loss": 0.0751953125, "epoch": 2.563429571303587, "explain_loss": 0.90625, "loss": 1.234375, "ref_ce_loss": 0.251953125, "stage_align_loss": 0.002655029296875, "step": 5860 }, { "ce_loss": 0.34765625, "distill_loss": 0.1015625, "epoch": 2.563429571303587, "explain_loss": 1.0859375, "loss": 1.984375, "ref_ce_loss": 0.39453125, "stage_align_loss": 0.052490234375, "step": 5860 }, { "ce_loss": 0.0013885498046875, "distill_loss": 0.064453125, "epoch": 2.563429571303587, "explain_loss": 1.25, "loss": 1.6640625, "ref_ce_loss": 0.333984375, "stage_align_loss": 0.01336669921875, "step": 5860 }, { "ce_loss": 0.00142669677734375, "distill_loss": 0.109375, "epoch": 2.563429571303587, "explain_loss": 0.96484375, "loss": 1.2890625, "ref_ce_loss": 0.2119140625, "stage_align_loss": 0.0026702880859375, "step": 5860 }, { "epoch": 2.567804024496938, "grad_norm": 0.6033753752708435, "learning_rate": 1.1158005495157342e-05, "loss": 1.610791015625, "step": 5870 }, { "ce_loss": 0.000682830810546875, "distill_loss": 0.09814453125, "epoch": 2.567804024496938, "explain_loss": 1.0703125, "loss": 1.578125, "ref_ce_loss": 0.40625, "stage_align_loss": 0.0028076171875, "step": 5870 }, { "ce_loss": 0.0028533935546875, "distill_loss": 0.0869140625, "epoch": 2.567804024496938, "explain_loss": 0.95703125, "loss": 1.390625, "ref_ce_loss": 0.337890625, "stage_align_loss": 0.0031890869140625, "step": 5870 }, { "ce_loss": 0.000492095947265625, "distill_loss": 0.09814453125, "epoch": 2.567804024496938, "explain_loss": 0.94140625, "loss": 1.2890625, "ref_ce_loss": 0.24609375, "stage_align_loss": 0.0029296875, "step": 5870 }, { "ce_loss": 0.005706787109375, "distill_loss": 0.06298828125, "epoch": 2.567804024496938, "explain_loss": 1.09375, "loss": 1.4765625, "ref_ce_loss": 0.3125, "stage_align_loss": 0.0031585693359375, "step": 5870 }, { "epoch": 2.5721784776902887, "grad_norm": 0.6598145961761475, "learning_rate": 1.0937690017118197e-05, "loss": 1.63095703125, "step": 5880 }, { "ce_loss": 0.0067138671875, "distill_loss": 0.08251953125, "epoch": 2.5721784776902887, "explain_loss": 0.62890625, "loss": 1.171875, "ref_ce_loss": 0.453125, "stage_align_loss": 0.00250244140625, "step": 5880 }, { "ce_loss": 0.95703125, "distill_loss": 0.0810546875, "epoch": 2.5721784776902887, "explain_loss": 0.953125, "loss": 2.53125, "ref_ce_loss": 0.54296875, "stage_align_loss": 0.0025634765625, "step": 5880 }, { "ce_loss": 0.034912109375, "distill_loss": 0.07421875, "epoch": 2.5721784776902887, "explain_loss": 1.4765625, "loss": 2.125, "ref_ce_loss": 0.482421875, "stage_align_loss": 0.057861328125, "step": 5880 }, { "ce_loss": 0.1328125, "distill_loss": 0.07666015625, "epoch": 2.5721784776902887, "explain_loss": 0.91015625, "loss": 1.4765625, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.0026397705078125, "step": 5880 }, { "epoch": 2.5765529308836395, "grad_norm": 1.2137559652328491, "learning_rate": 1.0719445462400323e-05, "loss": 1.59853515625, "step": 5890 }, { "ce_loss": 0.04052734375, "distill_loss": 0.0556640625, "epoch": 2.5765529308836395, "explain_loss": 0.93359375, "loss": 1.3671875, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.003326416015625, "step": 5890 }, { "ce_loss": 0.00250244140625, "distill_loss": 0.078125, "epoch": 2.5765529308836395, "explain_loss": 1.171875, "loss": 1.578125, "ref_ce_loss": 0.326171875, "stage_align_loss": 0.0027923583984375, "step": 5890 }, { "ce_loss": 0.0022430419921875, "distill_loss": 0.076171875, "epoch": 2.5765529308836395, "explain_loss": 1.3203125, "loss": 1.828125, "ref_ce_loss": 0.427734375, "stage_align_loss": 0.0024566650390625, "step": 5890 }, { "ce_loss": 0.0018463134765625, "distill_loss": 0.06884765625, "epoch": 2.5765529308836395, "explain_loss": 1.265625, "loss": 1.7265625, "ref_ce_loss": 0.38671875, "stage_align_loss": 0.004669189453125, "step": 5890 }, { "epoch": 2.5809273840769906, "grad_norm": 0.8912107944488525, "learning_rate": 1.05032769057401e-05, "loss": 1.58203125, "step": 5900 }, { "ce_loss": 0.00064849853515625, "distill_loss": 0.08984375, "epoch": 2.5809273840769906, "explain_loss": 1.3125, "loss": 1.65625, "ref_ce_loss": 0.2470703125, "stage_align_loss": 0.0035247802734375, "step": 5900 }, { "ce_loss": 0.0888671875, "distill_loss": 0.07421875, "epoch": 2.5809273840769906, "explain_loss": 0.80078125, "loss": 1.40625, "ref_ce_loss": 0.4375, "stage_align_loss": 0.006103515625, "step": 5900 }, { "ce_loss": 0.00701904296875, "distill_loss": 0.091796875, "epoch": 2.5809273840769906, "explain_loss": 1.1484375, "loss": 1.5390625, "ref_ce_loss": 0.291015625, "stage_align_loss": 0.00244140625, "step": 5900 }, { "ce_loss": 0.000537872314453125, "distill_loss": 0.1103515625, "epoch": 2.5809273840769906, "explain_loss": 0.8125, "loss": 1.2890625, "ref_ce_loss": 0.365234375, "stage_align_loss": 0.00274658203125, "step": 5900 }, { "epoch": 2.5853018372703414, "grad_norm": 0.7140970826148987, "learning_rate": 1.0289189373601616e-05, "loss": 1.54638671875, "step": 5910 }, { "ce_loss": 0.01409912109375, "distill_loss": 0.09326171875, "epoch": 2.5853018372703414, "explain_loss": 1.0625, "loss": 1.4921875, "ref_ce_loss": 0.318359375, "stage_align_loss": 0.004547119140625, "step": 5910 }, { "ce_loss": 0.0244140625, "distill_loss": 0.0751953125, "epoch": 2.5853018372703414, "explain_loss": 1.0078125, "loss": 1.515625, "ref_ce_loss": 0.404296875, "stage_align_loss": 0.00408935546875, "step": 5910 }, { "ce_loss": 0.17578125, "distill_loss": 0.08203125, "epoch": 2.5853018372703414, "explain_loss": 1.265625, "loss": 1.96875, "ref_ce_loss": 0.40234375, "stage_align_loss": 0.0478515625, "step": 5910 }, { "ce_loss": 0.56640625, "distill_loss": 0.0712890625, "epoch": 2.5853018372703414, "explain_loss": 1.0546875, "loss": 2.078125, "ref_ce_loss": 0.31640625, "stage_align_loss": 0.0712890625, "step": 5910 }, { "epoch": 2.589676290463692, "grad_norm": 0.6852843165397644, "learning_rate": 1.0077187844059921e-05, "loss": 1.58564453125, "step": 5920 }, { "ce_loss": 0.051025390625, "distill_loss": 0.0703125, "epoch": 2.589676290463692, "explain_loss": 0.8359375, "loss": 1.3203125, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.00390625, "step": 5920 }, { "ce_loss": 0.0184326171875, "distill_loss": 0.0908203125, "epoch": 2.589676290463692, "explain_loss": 0.9921875, "loss": 1.484375, "ref_ce_loss": 0.376953125, "stage_align_loss": 0.003021240234375, "step": 5920 }, { "ce_loss": 0.404296875, "distill_loss": 0.0888671875, "epoch": 2.589676290463692, "explain_loss": 0.7734375, "loss": 1.546875, "ref_ce_loss": 0.271484375, "stage_align_loss": 0.0037384033203125, "step": 5920 }, { "ce_loss": 0.046630859375, "distill_loss": 0.08837890625, "epoch": 2.589676290463692, "explain_loss": 1.21875, "loss": 1.6875, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.006927490234375, "step": 5920 }, { "epoch": 2.594050743657043, "grad_norm": 0.6505897045135498, "learning_rate": 9.867277246685225e-06, "loss": 1.584375, "step": 5930 }, { "ce_loss": 0.00168609619140625, "distill_loss": 0.11181640625, "epoch": 2.594050743657043, "explain_loss": 0.828125, "loss": 1.3046875, "ref_ce_loss": 0.361328125, "stage_align_loss": 0.0035858154296875, "step": 5930 }, { "ce_loss": 0.10888671875, "distill_loss": 0.123046875, "epoch": 2.594050743657043, "explain_loss": 0.98828125, "loss": 1.5, "ref_ce_loss": 0.2734375, "stage_align_loss": 0.003997802734375, "step": 5930 }, { "ce_loss": 0.8359375, "distill_loss": 0.08544921875, "epoch": 2.594050743657043, "explain_loss": 1.03125, "loss": 2.34375, "ref_ce_loss": 0.228515625, "stage_align_loss": 0.169921875, "step": 5930 }, { "ce_loss": 0.00762939453125, "distill_loss": 0.1142578125, "epoch": 2.594050743657043, "explain_loss": 1.046875, "loss": 1.5859375, "ref_ce_loss": 0.412109375, "stage_align_loss": 0.0036773681640625, "step": 5930 }, { "epoch": 2.5984251968503935, "grad_norm": 0.6318841576576233, "learning_rate": 9.659462462428348e-06, "loss": 1.598046875, "step": 5940 }, { "ce_loss": 0.0015869140625, "distill_loss": 0.08740234375, "epoch": 2.5984251968503935, "explain_loss": 0.984375, "loss": 1.640625, "ref_ce_loss": 0.484375, "stage_align_loss": 0.08203125, "step": 5940 }, { "ce_loss": 0.00421142578125, "distill_loss": 0.0732421875, "epoch": 2.5984251968503935, "explain_loss": 1.109375, "loss": 1.5234375, "ref_ce_loss": 0.330078125, "stage_align_loss": 0.0029449462890625, "step": 5940 }, { "ce_loss": 0.046142578125, "distill_loss": 0.11474609375, "epoch": 2.5984251968503935, "explain_loss": 1.125, "loss": 1.71875, "ref_ce_loss": 0.423828125, "stage_align_loss": 0.004608154296875, "step": 5940 }, { "ce_loss": 0.130859375, "distill_loss": 0.09521484375, "epoch": 2.5984251968503935, "explain_loss": 0.9453125, "loss": 1.5859375, "ref_ce_loss": 0.36328125, "stage_align_loss": 0.052490234375, "step": 5940 }, { "epoch": 2.6027996500437447, "grad_norm": 0.6254478693008423, "learning_rate": 9.453748323507106e-06, "loss": 1.60810546875, "step": 5950 }, { "ce_loss": 0.12451171875, "distill_loss": 0.083984375, "epoch": 2.6027996500437447, "explain_loss": 1.1875, "loss": 1.71875, "ref_ce_loss": 0.32421875, "stage_align_loss": 0.003753662109375, "step": 5950 }, { "ce_loss": 0.267578125, "distill_loss": 0.06787109375, "epoch": 2.6027996500437447, "explain_loss": 0.9765625, "loss": 1.7265625, "ref_ce_loss": 0.34765625, "stage_align_loss": 0.06689453125, "step": 5950 }, { "ce_loss": 0.04638671875, "distill_loss": 0.0830078125, "epoch": 2.6027996500437447, "explain_loss": 0.9609375, "loss": 1.65625, "ref_ce_loss": 0.486328125, "stage_align_loss": 0.0830078125, "step": 5950 }, { "ce_loss": 0.0005340576171875, "distill_loss": 0.10205078125, "epoch": 2.6027996500437447, "explain_loss": 1.1015625, "loss": 1.625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.00408935546875, "step": 5950 }, { "epoch": 2.6071741032370954, "grad_norm": 0.641730010509491, "learning_rate": 9.25013961329404e-06, "loss": 1.58310546875, "step": 5960 }, { "ce_loss": 0.1806640625, "distill_loss": 0.1083984375, "epoch": 2.6071741032370954, "explain_loss": 1.140625, "loss": 1.8671875, "ref_ce_loss": 0.388671875, "stage_align_loss": 0.0458984375, "step": 5960 }, { "ce_loss": 0.06494140625, "distill_loss": 0.076171875, "epoch": 2.6071741032370954, "explain_loss": 1.03125, "loss": 1.4609375, "ref_ce_loss": 0.2890625, "stage_align_loss": 0.0028076171875, "step": 5960 }, { "ce_loss": 0.10302734375, "distill_loss": 0.07861328125, "epoch": 2.6071741032370954, "explain_loss": 0.9296875, "loss": 1.5390625, "ref_ce_loss": 0.416015625, "stage_align_loss": 0.010498046875, "step": 5960 }, { "ce_loss": 0.00118255615234375, "distill_loss": 0.0888671875, "epoch": 2.6071741032370954, "explain_loss": 0.984375, "loss": 1.546875, "ref_ce_loss": 0.474609375, "stage_align_loss": 0.0035247802734375, "step": 5960 }, { "epoch": 2.611548556430446, "grad_norm": 0.8284358978271484, "learning_rate": 9.048641066205155e-06, "loss": 1.58564453125, "step": 5970 }, { "ce_loss": 0.142578125, "distill_loss": 0.095703125, "epoch": 2.611548556430446, "explain_loss": 0.625, "loss": 1.046875, "ref_ce_loss": 0.1826171875, "stage_align_loss": 0.0029296875, "step": 5970 }, { "ce_loss": 0.357421875, "distill_loss": 0.0908203125, "epoch": 2.611548556430446, "explain_loss": 1.40625, "loss": 2.53125, "ref_ce_loss": 0.57421875, "stage_align_loss": 0.10107421875, "step": 5970 }, { "ce_loss": 0.306640625, "distill_loss": 0.11279296875, "epoch": 2.611548556430446, "explain_loss": 0.8125, "loss": 1.625, "ref_ce_loss": 0.3515625, "stage_align_loss": 0.033447265625, "step": 5970 }, { "ce_loss": 0.126953125, "distill_loss": 0.1044921875, "epoch": 2.611548556430446, "explain_loss": 0.91015625, "loss": 1.5625, "ref_ce_loss": 0.37890625, "stage_align_loss": 0.044189453125, "step": 5970 }, { "epoch": 2.615923009623797, "grad_norm": 0.5685972571372986, "learning_rate": 8.849257367589869e-06, "loss": 1.5759765625, "step": 5980 }, { "ce_loss": 0.0015106201171875, "distill_loss": 0.078125, "epoch": 2.615923009623797, "explain_loss": 1.375, "loss": 1.96875, "ref_ce_loss": 0.515625, "stage_align_loss": 0.00390625, "step": 5980 }, { "ce_loss": 0.00616455078125, "distill_loss": 0.10302734375, "epoch": 2.615923009623797, "explain_loss": 0.90625, "loss": 1.3515625, "ref_ce_loss": 0.33203125, "stage_align_loss": 0.0031280517578125, "step": 5980 }, { "ce_loss": 0.0013275146484375, "distill_loss": 0.080078125, "epoch": 2.615923009623797, "explain_loss": 1.3359375, "loss": 1.78125, "ref_ce_loss": 0.357421875, "stage_align_loss": 0.0027313232421875, "step": 5980 }, { "ce_loss": 0.2119140625, "distill_loss": 0.076171875, "epoch": 2.615923009623797, "explain_loss": 1.109375, "loss": 1.796875, "ref_ce_loss": 0.396484375, "stage_align_loss": 0.0037994384765625, "step": 5980 }, { "epoch": 2.620297462817148, "grad_norm": 0.8899451494216919, "learning_rate": 8.651993153622007e-06, "loss": 1.623193359375, "step": 5990 }, { "ce_loss": 0.490234375, "distill_loss": 0.1064453125, "epoch": 2.620297462817148, "explain_loss": 1.1796875, "loss": 2.4375, "ref_ce_loss": 0.55859375, "stage_align_loss": 0.10400390625, "step": 5990 }, { "ce_loss": 0.00262451171875, "distill_loss": 0.06494140625, "epoch": 2.620297462817148, "explain_loss": 0.7265625, "loss": 1.15625, "ref_ce_loss": 0.353515625, "stage_align_loss": 0.0031280517578125, "step": 5990 }, { "ce_loss": 0.0216064453125, "distill_loss": 0.1025390625, "epoch": 2.620297462817148, "explain_loss": 1.28125, "loss": 1.8203125, "ref_ce_loss": 0.408203125, "stage_align_loss": 0.00616455078125, "step": 5990 }, { "ce_loss": 0.12890625, "distill_loss": 0.0771484375, "epoch": 2.620297462817148, "explain_loss": 1.203125, "loss": 1.8515625, "ref_ce_loss": 0.421875, "stage_align_loss": 0.021240234375, "step": 5990 }, { "epoch": 2.6246719160104988, "grad_norm": 0.7859241366386414, "learning_rate": 8.456853011192045e-06, "loss": 1.53017578125, "step": 6000 } ], "logging_steps": 10, "max_steps": 6858, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }