| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 2000, | |
| "global_step": 21900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00684931506849315, | |
| "grad_norm": 85.37890625, | |
| "learning_rate": 1.9963470319634707e-05, | |
| "loss": 5.693, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0136986301369863, | |
| "grad_norm": 199.43943786621094, | |
| "learning_rate": 1.9918721461187217e-05, | |
| "loss": 4.1718, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02054794520547945, | |
| "grad_norm": 281.33978271484375, | |
| "learning_rate": 1.9873059360730596e-05, | |
| "loss": 3.2414, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 54.39994812011719, | |
| "learning_rate": 1.9827397260273974e-05, | |
| "loss": 2.4046, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03424657534246575, | |
| "grad_norm": 34.490028381347656, | |
| "learning_rate": 1.9781735159817353e-05, | |
| "loss": 1.9848, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0410958904109589, | |
| "grad_norm": 318.8103332519531, | |
| "learning_rate": 1.973607305936073e-05, | |
| "loss": 1.8404, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04794520547945205, | |
| "grad_norm": 78.03404235839844, | |
| "learning_rate": 1.969041095890411e-05, | |
| "loss": 1.6072, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 108.01981353759766, | |
| "learning_rate": 1.9644748858447492e-05, | |
| "loss": 1.4967, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06164383561643835, | |
| "grad_norm": 71.82508087158203, | |
| "learning_rate": 1.9599086757990867e-05, | |
| "loss": 1.5211, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0684931506849315, | |
| "grad_norm": 16.820402145385742, | |
| "learning_rate": 1.955342465753425e-05, | |
| "loss": 1.6117, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.07534246575342465, | |
| "grad_norm": 23.496423721313477, | |
| "learning_rate": 1.9507762557077627e-05, | |
| "loss": 1.5837, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 23.132579803466797, | |
| "learning_rate": 1.9462100456621006e-05, | |
| "loss": 1.4387, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08904109589041095, | |
| "grad_norm": 23.49315643310547, | |
| "learning_rate": 1.9416438356164384e-05, | |
| "loss": 1.3564, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.0958904109589041, | |
| "grad_norm": 29.00982093811035, | |
| "learning_rate": 1.9370776255707763e-05, | |
| "loss": 1.3524, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.10273972602739725, | |
| "grad_norm": 17.096088409423828, | |
| "learning_rate": 1.9325114155251145e-05, | |
| "loss": 1.3412, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 39.139991760253906, | |
| "learning_rate": 1.9279452054794523e-05, | |
| "loss": 1.3968, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.11643835616438356, | |
| "grad_norm": 25.544593811035156, | |
| "learning_rate": 1.9233789954337902e-05, | |
| "loss": 1.3257, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1232876712328767, | |
| "grad_norm": 27.78721046447754, | |
| "learning_rate": 1.918812785388128e-05, | |
| "loss": 1.2974, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.13013698630136986, | |
| "grad_norm": 21.869457244873047, | |
| "learning_rate": 1.914246575342466e-05, | |
| "loss": 1.2354, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 41.35359573364258, | |
| "learning_rate": 1.9096803652968037e-05, | |
| "loss": 1.1967, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14383561643835616, | |
| "grad_norm": 27.000818252563477, | |
| "learning_rate": 1.9051141552511416e-05, | |
| "loss": 1.1838, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1506849315068493, | |
| "grad_norm": 46.611236572265625, | |
| "learning_rate": 1.9005479452054795e-05, | |
| "loss": 1.2294, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.15753424657534246, | |
| "grad_norm": 23.135637283325195, | |
| "learning_rate": 1.8959817351598176e-05, | |
| "loss": 1.1739, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 36.13066482543945, | |
| "learning_rate": 1.8914155251141555e-05, | |
| "loss": 1.1893, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.17123287671232876, | |
| "grad_norm": 26.738330841064453, | |
| "learning_rate": 1.8868493150684933e-05, | |
| "loss": 1.177, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1780821917808219, | |
| "grad_norm": 23.816173553466797, | |
| "learning_rate": 1.8822831050228312e-05, | |
| "loss": 1.1999, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.18493150684931506, | |
| "grad_norm": 41.335750579833984, | |
| "learning_rate": 1.877716894977169e-05, | |
| "loss": 1.2134, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 99.51826477050781, | |
| "learning_rate": 1.8731506849315072e-05, | |
| "loss": 1.3351, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.19863013698630136, | |
| "grad_norm": 31.743043899536133, | |
| "learning_rate": 1.8685844748858448e-05, | |
| "loss": 1.3438, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.2054794520547945, | |
| "grad_norm": 24.562389373779297, | |
| "learning_rate": 1.864018264840183e-05, | |
| "loss": 1.1273, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21232876712328766, | |
| "grad_norm": 44.048587799072266, | |
| "learning_rate": 1.8594520547945208e-05, | |
| "loss": 1.2602, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 29.480670928955078, | |
| "learning_rate": 1.8548858447488587e-05, | |
| "loss": 1.1501, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.22602739726027396, | |
| "grad_norm": 19.74498176574707, | |
| "learning_rate": 1.8503196347031965e-05, | |
| "loss": 1.1797, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.2328767123287671, | |
| "grad_norm": 83.28339385986328, | |
| "learning_rate": 1.8457534246575344e-05, | |
| "loss": 1.1652, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23972602739726026, | |
| "grad_norm": 40.529632568359375, | |
| "learning_rate": 1.8411872146118722e-05, | |
| "loss": 1.1377, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 19.91573715209961, | |
| "learning_rate": 1.8366210045662104e-05, | |
| "loss": 1.1338, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.2534246575342466, | |
| "grad_norm": 15.861178398132324, | |
| "learning_rate": 1.832054794520548e-05, | |
| "loss": 1.0917, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.2602739726027397, | |
| "grad_norm": 18.794872283935547, | |
| "learning_rate": 1.827488584474886e-05, | |
| "loss": 1.1589, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2671232876712329, | |
| "grad_norm": 69.27964782714844, | |
| "learning_rate": 1.822922374429224e-05, | |
| "loss": 1.1292, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 22.189298629760742, | |
| "learning_rate": 1.8183561643835618e-05, | |
| "loss": 0.9727, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "eval_exact_match": 76.61305581835383, | |
| "eval_f1": 85.71545945318721, | |
| "eval_runtime": 413.7588, | |
| "eval_samples_per_second": 25.546, | |
| "eval_steps_per_second": 1.598, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2808219178082192, | |
| "grad_norm": 38.37203598022461, | |
| "learning_rate": 1.8137899543378997e-05, | |
| "loss": 1.2412, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2876712328767123, | |
| "grad_norm": 200.2743682861328, | |
| "learning_rate": 1.8092237442922375e-05, | |
| "loss": 1.2055, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2945205479452055, | |
| "grad_norm": 17.487380981445312, | |
| "learning_rate": 1.8046575342465754e-05, | |
| "loss": 1.1285, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 29.930660247802734, | |
| "learning_rate": 1.8000913242009136e-05, | |
| "loss": 1.074, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3082191780821918, | |
| "grad_norm": 1504.7388916015625, | |
| "learning_rate": 1.7956164383561646e-05, | |
| "loss": 1.0404, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3150684931506849, | |
| "grad_norm": 17.754243850708008, | |
| "learning_rate": 1.7910502283105024e-05, | |
| "loss": 1.1388, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3219178082191781, | |
| "grad_norm": 24.204845428466797, | |
| "learning_rate": 1.7864840182648403e-05, | |
| "loss": 1.003, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 35.900047302246094, | |
| "learning_rate": 1.781917808219178e-05, | |
| "loss": 1.0618, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3356164383561644, | |
| "grad_norm": 112.26127624511719, | |
| "learning_rate": 1.777351598173516e-05, | |
| "loss": 1.1026, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 9.092456817626953, | |
| "learning_rate": 1.7727853881278542e-05, | |
| "loss": 1.1771, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3493150684931507, | |
| "grad_norm": 29.461143493652344, | |
| "learning_rate": 1.768219178082192e-05, | |
| "loss": 1.0458, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 13.496334075927734, | |
| "learning_rate": 1.76365296803653e-05, | |
| "loss": 1.0562, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.363013698630137, | |
| "grad_norm": 13.762040138244629, | |
| "learning_rate": 1.7590867579908677e-05, | |
| "loss": 0.9811, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.3698630136986301, | |
| "grad_norm": 10.256505012512207, | |
| "learning_rate": 1.7545205479452056e-05, | |
| "loss": 0.9995, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3767123287671233, | |
| "grad_norm": 22.89681053161621, | |
| "learning_rate": 1.7499543378995438e-05, | |
| "loss": 0.9709, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 18.46767234802246, | |
| "learning_rate": 1.7453881278538813e-05, | |
| "loss": 1.0183, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.3904109589041096, | |
| "grad_norm": 16.715681076049805, | |
| "learning_rate": 1.740821917808219e-05, | |
| "loss": 0.9484, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3972602739726027, | |
| "grad_norm": 24.110916137695312, | |
| "learning_rate": 1.7362557077625573e-05, | |
| "loss": 0.9787, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4041095890410959, | |
| "grad_norm": 63.249969482421875, | |
| "learning_rate": 1.7316894977168952e-05, | |
| "loss": 0.9529, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 63.814510345458984, | |
| "learning_rate": 1.727123287671233e-05, | |
| "loss": 0.9848, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4178082191780822, | |
| "grad_norm": 37.51655960083008, | |
| "learning_rate": 1.722557077625571e-05, | |
| "loss": 0.9452, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.4246575342465753, | |
| "grad_norm": 54.34503936767578, | |
| "learning_rate": 1.7179908675799087e-05, | |
| "loss": 0.7963, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4315068493150685, | |
| "grad_norm": 82.58960723876953, | |
| "learning_rate": 1.713424657534247e-05, | |
| "loss": 0.8767, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 23.595382690429688, | |
| "learning_rate": 1.7088584474885844e-05, | |
| "loss": 0.9605, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4452054794520548, | |
| "grad_norm": 13.270450592041016, | |
| "learning_rate": 1.7042922374429226e-05, | |
| "loss": 1.0003, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.4520547945205479, | |
| "grad_norm": 50.18392562866211, | |
| "learning_rate": 1.6997260273972605e-05, | |
| "loss": 0.9164, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4589041095890411, | |
| "grad_norm": 27.26555633544922, | |
| "learning_rate": 1.6951598173515983e-05, | |
| "loss": 0.9223, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 15.29233455657959, | |
| "learning_rate": 1.6905936073059362e-05, | |
| "loss": 0.9466, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4726027397260274, | |
| "grad_norm": 32.846405029296875, | |
| "learning_rate": 1.686027397260274e-05, | |
| "loss": 0.9374, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.4794520547945205, | |
| "grad_norm": 35.43315887451172, | |
| "learning_rate": 1.681461187214612e-05, | |
| "loss": 0.8864, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4863013698630137, | |
| "grad_norm": 35.38363265991211, | |
| "learning_rate": 1.67689497716895e-05, | |
| "loss": 0.8921, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 39.07960510253906, | |
| "learning_rate": 1.6723287671232876e-05, | |
| "loss": 0.8609, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 31.375228881835938, | |
| "learning_rate": 1.6677625570776258e-05, | |
| "loss": 0.844, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 23.44841766357422, | |
| "learning_rate": 1.6631963470319636e-05, | |
| "loss": 1.1013, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5136986301369864, | |
| "grad_norm": 32.550357818603516, | |
| "learning_rate": 1.6586301369863015e-05, | |
| "loss": 0.8648, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 25.50990867614746, | |
| "learning_rate": 1.6540639269406393e-05, | |
| "loss": 0.865, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5273972602739726, | |
| "grad_norm": 29.534744262695312, | |
| "learning_rate": 1.6494977168949772e-05, | |
| "loss": 0.9502, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5342465753424658, | |
| "grad_norm": 10.336551666259766, | |
| "learning_rate": 1.6449315068493154e-05, | |
| "loss": 0.9009, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.541095890410959, | |
| "grad_norm": 31.593719482421875, | |
| "learning_rate": 1.6403652968036532e-05, | |
| "loss": 0.9182, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 9.0987548828125, | |
| "learning_rate": 1.6358904109589043e-05, | |
| "loss": 0.9706, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_exact_match": 81.11636707663197, | |
| "eval_f1": 89.25530068369032, | |
| "eval_runtime": 411.9919, | |
| "eval_samples_per_second": 25.656, | |
| "eval_steps_per_second": 1.604, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5547945205479452, | |
| "grad_norm": 9.672983169555664, | |
| "learning_rate": 1.631324200913242e-05, | |
| "loss": 1.0179, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.5616438356164384, | |
| "grad_norm": 12.515457153320312, | |
| "learning_rate": 1.62675799086758e-05, | |
| "loss": 0.9346, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5684931506849316, | |
| "grad_norm": 23.963817596435547, | |
| "learning_rate": 1.6221917808219178e-05, | |
| "loss": 0.9121, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 26.651578903198242, | |
| "learning_rate": 1.6176255707762557e-05, | |
| "loss": 0.9604, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5821917808219178, | |
| "grad_norm": 4.494334697723389, | |
| "learning_rate": 1.613059360730594e-05, | |
| "loss": 1.9663, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.589041095890411, | |
| "grad_norm": 2.634287118911743, | |
| "learning_rate": 1.6084931506849317e-05, | |
| "loss": 4.9987, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5958904109589042, | |
| "grad_norm": 6.1256422996521, | |
| "learning_rate": 1.6039269406392696e-05, | |
| "loss": 5.0254, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 469.1513366699219, | |
| "learning_rate": 1.5993607305936074e-05, | |
| "loss": 4.1133, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6095890410958904, | |
| "grad_norm": 1784.662109375, | |
| "learning_rate": 1.5947945205479453e-05, | |
| "loss": 3.0178, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6164383561643836, | |
| "grad_norm": 29.62358856201172, | |
| "learning_rate": 1.5903196347031963e-05, | |
| "loss": 2.4815, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6232876712328768, | |
| "grad_norm": 17.960285186767578, | |
| "learning_rate": 1.5857534246575345e-05, | |
| "loss": 1.9513, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 34.22579574584961, | |
| "learning_rate": 1.5811872146118723e-05, | |
| "loss": 1.7791, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.636986301369863, | |
| "grad_norm": 14.769521713256836, | |
| "learning_rate": 1.5766210045662102e-05, | |
| "loss": 1.5022, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.6438356164383562, | |
| "grad_norm": 25.712989807128906, | |
| "learning_rate": 1.572054794520548e-05, | |
| "loss": 1.5074, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6506849315068494, | |
| "grad_norm": 89.54208374023438, | |
| "learning_rate": 1.567488584474886e-05, | |
| "loss": 1.2388, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 31.36192512512207, | |
| "learning_rate": 1.5629223744292237e-05, | |
| "loss": 1.1044, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.6643835616438356, | |
| "grad_norm": 25.994415283203125, | |
| "learning_rate": 1.558356164383562e-05, | |
| "loss": 1.2388, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.6712328767123288, | |
| "grad_norm": 36.64569091796875, | |
| "learning_rate": 1.5537899543378994e-05, | |
| "loss": 1.1491, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.678082191780822, | |
| "grad_norm": 30.16401481628418, | |
| "learning_rate": 1.5492237442922376e-05, | |
| "loss": 1.0918, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 41.026405334472656, | |
| "learning_rate": 1.5446575342465755e-05, | |
| "loss": 1.0654, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6917808219178082, | |
| "grad_norm": 19.094816207885742, | |
| "learning_rate": 1.5400913242009133e-05, | |
| "loss": 1.0635, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.6986301369863014, | |
| "grad_norm": 26.169776916503906, | |
| "learning_rate": 1.5355251141552512e-05, | |
| "loss": 1.2011, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7054794520547946, | |
| "grad_norm": 29.09198760986328, | |
| "learning_rate": 1.530958904109589e-05, | |
| "loss": 1.1482, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 13.325521469116211, | |
| "learning_rate": 1.5263926940639272e-05, | |
| "loss": 1.1105, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7191780821917808, | |
| "grad_norm": 26.117162704467773, | |
| "learning_rate": 1.521826484018265e-05, | |
| "loss": 1.1409, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.726027397260274, | |
| "grad_norm": 24.220075607299805, | |
| "learning_rate": 1.5172602739726028e-05, | |
| "loss": 1.0106, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.7328767123287672, | |
| "grad_norm": 21.246910095214844, | |
| "learning_rate": 1.5126940639269408e-05, | |
| "loss": 1.0285, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 122.90814208984375, | |
| "learning_rate": 1.5081278538812786e-05, | |
| "loss": 1.0871, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7465753424657534, | |
| "grad_norm": 48.804752349853516, | |
| "learning_rate": 1.5035616438356167e-05, | |
| "loss": 1.0243, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "grad_norm": 47.379703521728516, | |
| "learning_rate": 1.4989954337899543e-05, | |
| "loss": 1.1038, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7602739726027398, | |
| "grad_norm": 19.76502799987793, | |
| "learning_rate": 1.4944292237442924e-05, | |
| "loss": 0.9939, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 17.515962600708008, | |
| "learning_rate": 1.4898630136986302e-05, | |
| "loss": 0.9739, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.773972602739726, | |
| "grad_norm": 15.223257064819336, | |
| "learning_rate": 1.4852968036529682e-05, | |
| "loss": 0.9963, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.7808219178082192, | |
| "grad_norm": 23.438310623168945, | |
| "learning_rate": 1.480730593607306e-05, | |
| "loss": 1.053, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7876712328767124, | |
| "grad_norm": 20.485931396484375, | |
| "learning_rate": 1.476164383561644e-05, | |
| "loss": 1.0661, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 44.34764099121094, | |
| "learning_rate": 1.471598173515982e-05, | |
| "loss": 1.0459, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.8013698630136986, | |
| "grad_norm": 15.977239608764648, | |
| "learning_rate": 1.4670319634703198e-05, | |
| "loss": 1.072, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.8082191780821918, | |
| "grad_norm": 38.571083068847656, | |
| "learning_rate": 1.4624657534246575e-05, | |
| "loss": 0.9674, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.815068493150685, | |
| "grad_norm": 21.06000518798828, | |
| "learning_rate": 1.4578995433789955e-05, | |
| "loss": 0.9274, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 13.814413070678711, | |
| "learning_rate": 1.4533333333333335e-05, | |
| "loss": 0.9853, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "eval_exact_match": 81.98675496688742, | |
| "eval_f1": 89.93884090786635, | |
| "eval_runtime": 410.8052, | |
| "eval_samples_per_second": 25.73, | |
| "eval_steps_per_second": 1.609, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8287671232876712, | |
| "grad_norm": 24.676000595092773, | |
| "learning_rate": 1.4487671232876714e-05, | |
| "loss": 0.945, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.8356164383561644, | |
| "grad_norm": 12.431649208068848, | |
| "learning_rate": 1.4442009132420092e-05, | |
| "loss": 0.9395, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8424657534246576, | |
| "grad_norm": 31.419618606567383, | |
| "learning_rate": 1.4396347031963471e-05, | |
| "loss": 1.076, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 9.308188438415527, | |
| "learning_rate": 1.4350684931506851e-05, | |
| "loss": 1.0023, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8561643835616438, | |
| "grad_norm": 27.188827514648438, | |
| "learning_rate": 1.430502283105023e-05, | |
| "loss": 0.9309, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.863013698630137, | |
| "grad_norm": 12.23662281036377, | |
| "learning_rate": 1.4259360730593608e-05, | |
| "loss": 0.9248, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8698630136986302, | |
| "grad_norm": 27.70292854309082, | |
| "learning_rate": 1.4213698630136987e-05, | |
| "loss": 1.018, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 34.40644454956055, | |
| "learning_rate": 1.4168036529680367e-05, | |
| "loss": 0.8974, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8835616438356164, | |
| "grad_norm": 16.194690704345703, | |
| "learning_rate": 1.4122374429223747e-05, | |
| "loss": 0.9723, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.8904109589041096, | |
| "grad_norm": 16.42717742919922, | |
| "learning_rate": 1.4076712328767124e-05, | |
| "loss": 0.9421, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8972602739726028, | |
| "grad_norm": 9.460033416748047, | |
| "learning_rate": 1.4031050228310502e-05, | |
| "loss": 0.8857, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 20.79571533203125, | |
| "learning_rate": 1.3985388127853883e-05, | |
| "loss": 0.8813, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.910958904109589, | |
| "grad_norm": 11.8665771484375, | |
| "learning_rate": 1.3939726027397263e-05, | |
| "loss": 0.964, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.9178082191780822, | |
| "grad_norm": 8.933015823364258, | |
| "learning_rate": 1.389406392694064e-05, | |
| "loss": 0.8678, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.9246575342465754, | |
| "grad_norm": 90.67884826660156, | |
| "learning_rate": 1.384840182648402e-05, | |
| "loss": 0.9355, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 50.61793518066406, | |
| "learning_rate": 1.3802739726027398e-05, | |
| "loss": 0.8578, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9383561643835616, | |
| "grad_norm": 19.59413719177246, | |
| "learning_rate": 1.3757077625570779e-05, | |
| "loss": 0.9403, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.9452054794520548, | |
| "grad_norm": 18.14301872253418, | |
| "learning_rate": 1.3711415525114155e-05, | |
| "loss": 0.9549, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.952054794520548, | |
| "grad_norm": 22.918977737426758, | |
| "learning_rate": 1.3665753424657536e-05, | |
| "loss": 0.9292, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 31.385986328125, | |
| "learning_rate": 1.3620091324200914e-05, | |
| "loss": 0.9122, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9657534246575342, | |
| "grad_norm": 17.648303985595703, | |
| "learning_rate": 1.3574429223744294e-05, | |
| "loss": 1.0112, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.9726027397260274, | |
| "grad_norm": 310.2109069824219, | |
| "learning_rate": 1.3528767123287671e-05, | |
| "loss": 0.9799, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.9794520547945206, | |
| "grad_norm": 38.04465103149414, | |
| "learning_rate": 1.3483105022831051e-05, | |
| "loss": 0.9595, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 50.285972595214844, | |
| "learning_rate": 1.343744292237443e-05, | |
| "loss": 0.9403, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.9931506849315068, | |
| "grad_norm": 18.11605453491211, | |
| "learning_rate": 1.339178082191781e-05, | |
| "loss": 0.9119, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 18.265396118164062, | |
| "learning_rate": 1.3346118721461187e-05, | |
| "loss": 0.7718, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.0068493150684932, | |
| "grad_norm": 17.143844604492188, | |
| "learning_rate": 1.3300456621004567e-05, | |
| "loss": 0.7068, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 8.762425422668457, | |
| "learning_rate": 1.3254794520547947e-05, | |
| "loss": 0.7369, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.0205479452054795, | |
| "grad_norm": 8.921041488647461, | |
| "learning_rate": 1.3209132420091326e-05, | |
| "loss": 0.7536, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.0273972602739727, | |
| "grad_norm": 10.755156517028809, | |
| "learning_rate": 1.3163470319634703e-05, | |
| "loss": 0.7538, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0342465753424657, | |
| "grad_norm": 25.93258285522461, | |
| "learning_rate": 1.3117808219178083e-05, | |
| "loss": 0.7839, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 10.306924819946289, | |
| "learning_rate": 1.3072146118721463e-05, | |
| "loss": 0.6808, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.047945205479452, | |
| "grad_norm": 24.816646575927734, | |
| "learning_rate": 1.3026484018264842e-05, | |
| "loss": 0.7554, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.0547945205479452, | |
| "grad_norm": 16.481876373291016, | |
| "learning_rate": 1.298082191780822e-05, | |
| "loss": 0.6706, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.0616438356164384, | |
| "grad_norm": 25.950448989868164, | |
| "learning_rate": 1.2935159817351599e-05, | |
| "loss": 0.7109, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 15.693510055541992, | |
| "learning_rate": 1.2889497716894979e-05, | |
| "loss": 0.6759, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.0753424657534247, | |
| "grad_norm": 24.4326229095459, | |
| "learning_rate": 1.2843835616438358e-05, | |
| "loss": 0.7348, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.0821917808219177, | |
| "grad_norm": 21.493825912475586, | |
| "learning_rate": 1.2798173515981736e-05, | |
| "loss": 0.6992, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.0890410958904109, | |
| "grad_norm": 15.809690475463867, | |
| "learning_rate": 1.2752511415525115e-05, | |
| "loss": 0.8123, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 19.380176544189453, | |
| "learning_rate": 1.2706849315068495e-05, | |
| "loss": 0.6854, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 82.20435193945127, | |
| "eval_f1": 90.42497415492494, | |
| "eval_runtime": 411.3984, | |
| "eval_samples_per_second": 25.693, | |
| "eval_steps_per_second": 1.607, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1027397260273972, | |
| "grad_norm": 23.904098510742188, | |
| "learning_rate": 1.2661187214611875e-05, | |
| "loss": 0.6933, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.1095890410958904, | |
| "grad_norm": 13.447206497192383, | |
| "learning_rate": 1.2615525114155252e-05, | |
| "loss": 0.782, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.1164383561643836, | |
| "grad_norm": 15.532490730285645, | |
| "learning_rate": 1.256986301369863e-05, | |
| "loss": 0.8741, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 21.778491973876953, | |
| "learning_rate": 1.252420091324201e-05, | |
| "loss": 0.7621, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.13013698630137, | |
| "grad_norm": 23.019596099853516, | |
| "learning_rate": 1.247853881278539e-05, | |
| "loss": 0.7105, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.1369863013698631, | |
| "grad_norm": 16.143905639648438, | |
| "learning_rate": 1.2432876712328768e-05, | |
| "loss": 0.724, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.143835616438356, | |
| "grad_norm": 13.059601783752441, | |
| "learning_rate": 1.2387214611872148e-05, | |
| "loss": 0.6949, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 20.782901763916016, | |
| "learning_rate": 1.2341552511415526e-05, | |
| "loss": 0.6602, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.1575342465753424, | |
| "grad_norm": 26.851299285888672, | |
| "learning_rate": 1.2295890410958907e-05, | |
| "loss": 0.7715, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.1643835616438356, | |
| "grad_norm": 13.973319053649902, | |
| "learning_rate": 1.2250228310502283e-05, | |
| "loss": 0.7333, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.1712328767123288, | |
| "grad_norm": 12.021798133850098, | |
| "learning_rate": 1.2204566210045664e-05, | |
| "loss": 0.7025, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 19.536468505859375, | |
| "learning_rate": 1.2158904109589042e-05, | |
| "loss": 0.6907, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.1849315068493151, | |
| "grad_norm": 13.144827842712402, | |
| "learning_rate": 1.2113242009132422e-05, | |
| "loss": 0.7402, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.191780821917808, | |
| "grad_norm": 21.790563583374023, | |
| "learning_rate": 1.2067579908675799e-05, | |
| "loss": 0.7615, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.1986301369863013, | |
| "grad_norm": 27.805265426635742, | |
| "learning_rate": 1.202191780821918e-05, | |
| "loss": 0.6817, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 18.263689041137695, | |
| "learning_rate": 1.1976255707762558e-05, | |
| "loss": 0.7266, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.2123287671232876, | |
| "grad_norm": 15.77778434753418, | |
| "learning_rate": 1.1930593607305938e-05, | |
| "loss": 0.8104, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.2191780821917808, | |
| "grad_norm": 28.699615478515625, | |
| "learning_rate": 1.1884931506849315e-05, | |
| "loss": 0.752, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.226027397260274, | |
| "grad_norm": 43.594669342041016, | |
| "learning_rate": 1.1839269406392695e-05, | |
| "loss": 0.7933, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 29.294349670410156, | |
| "learning_rate": 1.1793607305936075e-05, | |
| "loss": 0.6522, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2397260273972603, | |
| "grad_norm": 27.752662658691406, | |
| "learning_rate": 1.1747945205479454e-05, | |
| "loss": 0.7179, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.2465753424657535, | |
| "grad_norm": 13.734334945678711, | |
| "learning_rate": 1.170228310502283e-05, | |
| "loss": 0.7396, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.2534246575342465, | |
| "grad_norm": 15.601795196533203, | |
| "learning_rate": 1.1656621004566211e-05, | |
| "loss": 0.6691, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 20.292036056518555, | |
| "learning_rate": 1.1610958904109591e-05, | |
| "loss": 0.7035, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.2671232876712328, | |
| "grad_norm": 23.35797882080078, | |
| "learning_rate": 1.156529680365297e-05, | |
| "loss": 0.7263, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.273972602739726, | |
| "grad_norm": 24.030485153198242, | |
| "learning_rate": 1.1519634703196348e-05, | |
| "loss": 0.6525, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.2808219178082192, | |
| "grad_norm": 32.249656677246094, | |
| "learning_rate": 1.1473972602739727e-05, | |
| "loss": 0.7199, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 15.356407165527344, | |
| "learning_rate": 1.1428310502283107e-05, | |
| "loss": 0.7946, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.2945205479452055, | |
| "grad_norm": 7.511958122253418, | |
| "learning_rate": 1.1382648401826485e-05, | |
| "loss": 0.6527, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.3013698630136985, | |
| "grad_norm": 21.929622650146484, | |
| "learning_rate": 1.1336986301369864e-05, | |
| "loss": 0.7163, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.308219178082192, | |
| "grad_norm": 8.035719871520996, | |
| "learning_rate": 1.1291324200913242e-05, | |
| "loss": 0.7448, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 7.959759712219238, | |
| "learning_rate": 1.1245662100456623e-05, | |
| "loss": 0.617, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.321917808219178, | |
| "grad_norm": 51.80229568481445, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.722, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.3287671232876712, | |
| "grad_norm": 25.232810974121094, | |
| "learning_rate": 1.115433789954338e-05, | |
| "loss": 0.6917, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.3356164383561644, | |
| "grad_norm": 31.596223831176758, | |
| "learning_rate": 1.1108675799086758e-05, | |
| "loss": 0.7442, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 21.05303382873535, | |
| "learning_rate": 1.1063013698630138e-05, | |
| "loss": 0.6787, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.3493150684931507, | |
| "grad_norm": 33.4599723815918, | |
| "learning_rate": 1.1017351598173519e-05, | |
| "loss": 0.7315, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.356164383561644, | |
| "grad_norm": 52.5392951965332, | |
| "learning_rate": 1.0971689497716895e-05, | |
| "loss": 0.6484, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.3630136986301369, | |
| "grad_norm": 14.650938987731934, | |
| "learning_rate": 1.0926027397260274e-05, | |
| "loss": 0.6919, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 16.416217803955078, | |
| "learning_rate": 1.0880365296803654e-05, | |
| "loss": 0.6912, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "eval_exact_match": 83.50047303689688, | |
| "eval_f1": 90.81864087489008, | |
| "eval_runtime": 411.8012, | |
| "eval_samples_per_second": 25.668, | |
| "eval_steps_per_second": 1.605, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.3767123287671232, | |
| "grad_norm": 8.23530101776123, | |
| "learning_rate": 1.0834703196347034e-05, | |
| "loss": 0.6905, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.3835616438356164, | |
| "grad_norm": 9.29169750213623, | |
| "learning_rate": 1.0789041095890411e-05, | |
| "loss": 0.6088, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.3904109589041096, | |
| "grad_norm": 18.476957321166992, | |
| "learning_rate": 1.0743378995433791e-05, | |
| "loss": 0.7723, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 16.31971549987793, | |
| "learning_rate": 1.069771689497717e-05, | |
| "loss": 0.6369, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.404109589041096, | |
| "grad_norm": 15.640203475952148, | |
| "learning_rate": 1.065205479452055e-05, | |
| "loss": 0.6221, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.410958904109589, | |
| "grad_norm": 12.571090698242188, | |
| "learning_rate": 1.0606392694063927e-05, | |
| "loss": 0.607, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.4178082191780823, | |
| "grad_norm": 16.55551528930664, | |
| "learning_rate": 1.0560730593607307e-05, | |
| "loss": 0.8085, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 4.7176899909973145, | |
| "learning_rate": 1.0515068493150686e-05, | |
| "loss": 0.7011, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.4315068493150684, | |
| "grad_norm": 24.650453567504883, | |
| "learning_rate": 1.0469406392694066e-05, | |
| "loss": 0.7331, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.4383561643835616, | |
| "grad_norm": 22.386030197143555, | |
| "learning_rate": 1.0423744292237443e-05, | |
| "loss": 0.77, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.4452054794520548, | |
| "grad_norm": 14.527100563049316, | |
| "learning_rate": 1.0378082191780823e-05, | |
| "loss": 0.6114, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 12.87186050415039, | |
| "learning_rate": 1.0332420091324202e-05, | |
| "loss": 0.7005, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.4589041095890412, | |
| "grad_norm": 36.79588317871094, | |
| "learning_rate": 1.028675799086758e-05, | |
| "loss": 0.8049, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.4657534246575343, | |
| "grad_norm": 59.679832458496094, | |
| "learning_rate": 1.0241095890410959e-05, | |
| "loss": 0.7088, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.4726027397260273, | |
| "grad_norm": 24.512659072875977, | |
| "learning_rate": 1.0196347031963472e-05, | |
| "loss": 0.5994, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 19.418733596801758, | |
| "learning_rate": 1.015068493150685e-05, | |
| "loss": 0.7339, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.4863013698630136, | |
| "grad_norm": 21.704984664916992, | |
| "learning_rate": 1.010502283105023e-05, | |
| "loss": 0.6641, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.4931506849315068, | |
| "grad_norm": 10.757671356201172, | |
| "learning_rate": 1.0059360730593608e-05, | |
| "loss": 0.6841, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 12.122688293457031, | |
| "learning_rate": 1.0013698630136988e-05, | |
| "loss": 0.6453, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 11.158235549926758, | |
| "learning_rate": 9.968036529680366e-06, | |
| "loss": 0.7134, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.5136986301369864, | |
| "grad_norm": 21.53203582763672, | |
| "learning_rate": 9.922374429223745e-06, | |
| "loss": 0.7184, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 1.5205479452054793, | |
| "grad_norm": 26.664949417114258, | |
| "learning_rate": 9.876712328767123e-06, | |
| "loss": 0.6593, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.5273972602739727, | |
| "grad_norm": 23.013343811035156, | |
| "learning_rate": 9.831050228310504e-06, | |
| "loss": 0.726, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 20.908411026000977, | |
| "learning_rate": 9.785388127853882e-06, | |
| "loss": 0.6909, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 1.541095890410959, | |
| "grad_norm": 17.533647537231445, | |
| "learning_rate": 9.73972602739726e-06, | |
| "loss": 0.6686, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.547945205479452, | |
| "grad_norm": 9.584930419921875, | |
| "learning_rate": 9.69406392694064e-06, | |
| "loss": 0.7108, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 1.5547945205479452, | |
| "grad_norm": 10.897212028503418, | |
| "learning_rate": 9.64840182648402e-06, | |
| "loss": 0.7319, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 23.082847595214844, | |
| "learning_rate": 9.602739726027398e-06, | |
| "loss": 0.7352, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.5684931506849316, | |
| "grad_norm": 18.70546531677246, | |
| "learning_rate": 9.557077625570778e-06, | |
| "loss": 0.6495, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 1.5753424657534247, | |
| "grad_norm": 13.218006134033203, | |
| "learning_rate": 9.511415525114157e-06, | |
| "loss": 0.612, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.5821917808219177, | |
| "grad_norm": 17.062456130981445, | |
| "learning_rate": 9.465753424657535e-06, | |
| "loss": 0.7507, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 6.609577178955078, | |
| "learning_rate": 9.420091324200914e-06, | |
| "loss": 0.6781, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 1.595890410958904, | |
| "grad_norm": 13.452157020568848, | |
| "learning_rate": 9.374429223744294e-06, | |
| "loss": 0.7017, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 1.6027397260273972, | |
| "grad_norm": 9.165887832641602, | |
| "learning_rate": 9.328767123287673e-06, | |
| "loss": 0.6882, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.6095890410958904, | |
| "grad_norm": 10.622401237487793, | |
| "learning_rate": 9.283105022831051e-06, | |
| "loss": 0.6923, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 13.718300819396973, | |
| "learning_rate": 9.23744292237443e-06, | |
| "loss": 0.7852, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 1.6232876712328768, | |
| "grad_norm": 34.71263122558594, | |
| "learning_rate": 9.19178082191781e-06, | |
| "loss": 0.6129, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 1.6301369863013697, | |
| "grad_norm": 17.547475814819336, | |
| "learning_rate": 9.146118721461188e-06, | |
| "loss": 0.6077, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 1.6369863013698631, | |
| "grad_norm": 9.284356117248535, | |
| "learning_rate": 9.100456621004567e-06, | |
| "loss": 0.6784, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 10.464165687561035, | |
| "learning_rate": 9.054794520547945e-06, | |
| "loss": 0.6866, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_exact_match": 84.74929044465468, | |
| "eval_f1": 91.66661038518258, | |
| "eval_runtime": 410.7296, | |
| "eval_samples_per_second": 25.735, | |
| "eval_steps_per_second": 1.609, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.6506849315068495, | |
| "grad_norm": 14.635241508483887, | |
| "learning_rate": 9.009132420091326e-06, | |
| "loss": 0.6278, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 1.6575342465753424, | |
| "grad_norm": 11.001940727233887, | |
| "learning_rate": 8.963470319634704e-06, | |
| "loss": 0.7224, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 1.6643835616438356, | |
| "grad_norm": 13.27009105682373, | |
| "learning_rate": 8.917808219178083e-06, | |
| "loss": 0.6702, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 47.36380386352539, | |
| "learning_rate": 8.872146118721461e-06, | |
| "loss": 0.6811, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 1.678082191780822, | |
| "grad_norm": 34.635963439941406, | |
| "learning_rate": 8.82648401826484e-06, | |
| "loss": 0.6489, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.6849315068493151, | |
| "grad_norm": 6.181822776794434, | |
| "learning_rate": 8.78082191780822e-06, | |
| "loss": 0.646, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.691780821917808, | |
| "grad_norm": 18.2498722076416, | |
| "learning_rate": 8.735159817351598e-06, | |
| "loss": 0.6963, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 10.06582260131836, | |
| "learning_rate": 8.689497716894979e-06, | |
| "loss": 0.7804, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 1.7054794520547945, | |
| "grad_norm": 8.077252388000488, | |
| "learning_rate": 8.643835616438357e-06, | |
| "loss": 0.7383, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 1.7123287671232876, | |
| "grad_norm": 11.147444725036621, | |
| "learning_rate": 8.598173515981736e-06, | |
| "loss": 0.7016, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.7191780821917808, | |
| "grad_norm": 12.845279693603516, | |
| "learning_rate": 8.552511415525114e-06, | |
| "loss": 0.6312, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 16.080331802368164, | |
| "learning_rate": 8.506849315068494e-06, | |
| "loss": 0.5791, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.7328767123287672, | |
| "grad_norm": 7.987480163574219, | |
| "learning_rate": 8.461187214611873e-06, | |
| "loss": 0.5942, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 1.7397260273972601, | |
| "grad_norm": 15.770763397216797, | |
| "learning_rate": 8.415525114155251e-06, | |
| "loss": 0.5666, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 1.7465753424657535, | |
| "grad_norm": 11.666452407836914, | |
| "learning_rate": 8.36986301369863e-06, | |
| "loss": 0.6239, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 17.016170501708984, | |
| "learning_rate": 8.32420091324201e-06, | |
| "loss": 0.7785, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 1.7602739726027399, | |
| "grad_norm": 6.31413459777832, | |
| "learning_rate": 8.278538812785389e-06, | |
| "loss": 0.5971, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 1.7671232876712328, | |
| "grad_norm": 15.495176315307617, | |
| "learning_rate": 8.232876712328767e-06, | |
| "loss": 0.6564, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.773972602739726, | |
| "grad_norm": 15.647567749023438, | |
| "learning_rate": 8.187214611872146e-06, | |
| "loss": 0.6556, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 15.780739784240723, | |
| "learning_rate": 8.141552511415526e-06, | |
| "loss": 0.736, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7876712328767124, | |
| "grad_norm": 30.215124130249023, | |
| "learning_rate": 8.095890410958904e-06, | |
| "loss": 0.7227, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 1.7945205479452055, | |
| "grad_norm": 13.4503755569458, | |
| "learning_rate": 8.050228310502285e-06, | |
| "loss": 0.6727, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 1.8013698630136985, | |
| "grad_norm": 17.908418655395508, | |
| "learning_rate": 8.004566210045661e-06, | |
| "loss": 0.6764, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 4.531787872314453, | |
| "learning_rate": 7.958904109589042e-06, | |
| "loss": 0.7197, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.8150684931506849, | |
| "grad_norm": 5.361120700836182, | |
| "learning_rate": 7.91324200913242e-06, | |
| "loss": 0.6696, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 1.821917808219178, | |
| "grad_norm": 26.830280303955078, | |
| "learning_rate": 7.8675799086758e-06, | |
| "loss": 0.6393, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 1.8287671232876712, | |
| "grad_norm": 33.405513763427734, | |
| "learning_rate": 7.821917808219179e-06, | |
| "loss": 0.6637, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 33.269290924072266, | |
| "learning_rate": 7.776255707762557e-06, | |
| "loss": 0.6666, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 1.8424657534246576, | |
| "grad_norm": 39.952476501464844, | |
| "learning_rate": 7.730593607305936e-06, | |
| "loss": 0.7, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 1.8493150684931505, | |
| "grad_norm": 22.751543045043945, | |
| "learning_rate": 7.684931506849316e-06, | |
| "loss": 0.7378, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.856164383561644, | |
| "grad_norm": 12.855796813964844, | |
| "learning_rate": 7.639269406392695e-06, | |
| "loss": 0.6685, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 14.077750205993652, | |
| "learning_rate": 7.593607305936074e-06, | |
| "loss": 0.7293, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 1.8698630136986303, | |
| "grad_norm": 27.806459426879883, | |
| "learning_rate": 7.5479452054794526e-06, | |
| "loss": 0.7365, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 1.8767123287671232, | |
| "grad_norm": 27.054359436035156, | |
| "learning_rate": 7.502283105022832e-06, | |
| "loss": 0.7173, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 1.8835616438356164, | |
| "grad_norm": 24.715951919555664, | |
| "learning_rate": 7.4566210045662105e-06, | |
| "loss": 0.6688, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 10.76326847076416, | |
| "learning_rate": 7.41095890410959e-06, | |
| "loss": 0.5983, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.8972602739726028, | |
| "grad_norm": 23.433746337890625, | |
| "learning_rate": 7.365296803652968e-06, | |
| "loss": 0.6396, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 1.904109589041096, | |
| "grad_norm": 7.178381443023682, | |
| "learning_rate": 7.319634703196348e-06, | |
| "loss": 0.6087, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 1.910958904109589, | |
| "grad_norm": 14.573859214782715, | |
| "learning_rate": 7.273972602739726e-06, | |
| "loss": 0.6768, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 9.847871780395508, | |
| "learning_rate": 7.228310502283106e-06, | |
| "loss": 0.6488, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "eval_exact_match": 85.3926206244087, | |
| "eval_f1": 92.12563353412813, | |
| "eval_runtime": 411.6011, | |
| "eval_samples_per_second": 25.68, | |
| "eval_steps_per_second": 1.606, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9246575342465753, | |
| "grad_norm": 35.478633880615234, | |
| "learning_rate": 7.182648401826484e-06, | |
| "loss": 0.6617, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 1.9315068493150684, | |
| "grad_norm": 19.706714630126953, | |
| "learning_rate": 7.1369863013698635e-06, | |
| "loss": 0.7364, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.9383561643835616, | |
| "grad_norm": 24.93119239807129, | |
| "learning_rate": 7.091324200913242e-06, | |
| "loss": 0.6355, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 22.897789001464844, | |
| "learning_rate": 7.045662100456621e-06, | |
| "loss": 0.6816, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 1.952054794520548, | |
| "grad_norm": 17.719194412231445, | |
| "learning_rate": 7e-06, | |
| "loss": 0.7254, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 1.958904109589041, | |
| "grad_norm": 11.11882209777832, | |
| "learning_rate": 6.954337899543379e-06, | |
| "loss": 0.6603, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 1.9657534246575343, | |
| "grad_norm": 25.994152069091797, | |
| "learning_rate": 6.908675799086758e-06, | |
| "loss": 0.7368, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 28.343494415283203, | |
| "learning_rate": 6.863013698630138e-06, | |
| "loss": 0.6582, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.9794520547945207, | |
| "grad_norm": 20.735702514648438, | |
| "learning_rate": 6.8173515981735165e-06, | |
| "loss": 0.5868, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 1.9863013698630136, | |
| "grad_norm": 11.477102279663086, | |
| "learning_rate": 6.771689497716896e-06, | |
| "loss": 0.6739, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.9931506849315068, | |
| "grad_norm": 10.849296569824219, | |
| "learning_rate": 6.726027397260274e-06, | |
| "loss": 0.6721, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 28.718400955200195, | |
| "learning_rate": 6.680365296803654e-06, | |
| "loss": 0.7819, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.006849315068493, | |
| "grad_norm": 19.068960189819336, | |
| "learning_rate": 6.634703196347032e-06, | |
| "loss": 0.4468, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.0136986301369864, | |
| "grad_norm": 12.063799858093262, | |
| "learning_rate": 6.589041095890412e-06, | |
| "loss": 0.4271, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.0205479452054793, | |
| "grad_norm": 13.464853286743164, | |
| "learning_rate": 6.54337899543379e-06, | |
| "loss": 0.4645, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 12.967561721801758, | |
| "learning_rate": 6.4977168949771695e-06, | |
| "loss": 0.438, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.0342465753424657, | |
| "grad_norm": 3.8958585262298584, | |
| "learning_rate": 6.452054794520548e-06, | |
| "loss": 0.4105, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.041095890410959, | |
| "grad_norm": 7.50264835357666, | |
| "learning_rate": 6.4063926940639274e-06, | |
| "loss": 0.4713, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.047945205479452, | |
| "grad_norm": 16.149507522583008, | |
| "learning_rate": 6.360730593607306e-06, | |
| "loss": 0.4644, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 14.894956588745117, | |
| "learning_rate": 6.315068493150685e-06, | |
| "loss": 0.4356, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.0616438356164384, | |
| "grad_norm": 23.671754837036133, | |
| "learning_rate": 6.269406392694064e-06, | |
| "loss": 0.4477, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.0684931506849313, | |
| "grad_norm": 20.480844497680664, | |
| "learning_rate": 6.223744292237443e-06, | |
| "loss": 0.4944, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.0753424657534247, | |
| "grad_norm": 10.967472076416016, | |
| "learning_rate": 6.178082191780822e-06, | |
| "loss": 0.4322, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 7.305009365081787, | |
| "learning_rate": 6.132420091324202e-06, | |
| "loss": 0.4441, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.089041095890411, | |
| "grad_norm": 31.556331634521484, | |
| "learning_rate": 6.08675799086758e-06, | |
| "loss": 0.4751, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.095890410958904, | |
| "grad_norm": 11.134750366210938, | |
| "learning_rate": 6.04109589041096e-06, | |
| "loss": 0.4144, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.1027397260273974, | |
| "grad_norm": 9.854887008666992, | |
| "learning_rate": 5.995433789954338e-06, | |
| "loss": 0.4095, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 24.200942993164062, | |
| "learning_rate": 5.949771689497718e-06, | |
| "loss": 0.4809, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.1164383561643834, | |
| "grad_norm": 13.767411231994629, | |
| "learning_rate": 5.904109589041096e-06, | |
| "loss": 0.4494, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.1232876712328768, | |
| "grad_norm": 32.64931106567383, | |
| "learning_rate": 5.858447488584476e-06, | |
| "loss": 0.3994, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.1301369863013697, | |
| "grad_norm": 24.33043098449707, | |
| "learning_rate": 5.812785388127854e-06, | |
| "loss": 0.3991, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 21.8504581451416, | |
| "learning_rate": 5.7671232876712335e-06, | |
| "loss": 0.4247, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.143835616438356, | |
| "grad_norm": 17.096647262573242, | |
| "learning_rate": 5.721461187214612e-06, | |
| "loss": 0.4862, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.1506849315068495, | |
| "grad_norm": 7.025681972503662, | |
| "learning_rate": 5.675799086757991e-06, | |
| "loss": 0.4183, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.1575342465753424, | |
| "grad_norm": 8.589903831481934, | |
| "learning_rate": 5.63013698630137e-06, | |
| "loss": 0.478, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 10.5584077835083, | |
| "learning_rate": 5.584474885844749e-06, | |
| "loss": 0.4085, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.171232876712329, | |
| "grad_norm": 26.621702194213867, | |
| "learning_rate": 5.538812785388128e-06, | |
| "loss": 0.4524, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.1780821917808217, | |
| "grad_norm": 6.596956729888916, | |
| "learning_rate": 5.493150684931507e-06, | |
| "loss": 0.3702, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.184931506849315, | |
| "grad_norm": 4.0847673416137695, | |
| "learning_rate": 5.447488584474886e-06, | |
| "loss": 0.3534, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 8.085464477539062, | |
| "learning_rate": 5.401826484018266e-06, | |
| "loss": 0.5179, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 84.88174077578051, | |
| "eval_f1": 91.72384047979602, | |
| "eval_runtime": 410.8959, | |
| "eval_samples_per_second": 25.724, | |
| "eval_steps_per_second": 1.609, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.1986301369863015, | |
| "grad_norm": 6.315828323364258, | |
| "learning_rate": 5.3561643835616435e-06, | |
| "loss": 0.4095, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.2054794520547945, | |
| "grad_norm": 9.403338432312012, | |
| "learning_rate": 5.310502283105024e-06, | |
| "loss": 0.4056, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.212328767123288, | |
| "grad_norm": 21.496809005737305, | |
| "learning_rate": 5.264840182648402e-06, | |
| "loss": 0.4552, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 17.62648582458496, | |
| "learning_rate": 5.219178082191782e-06, | |
| "loss": 0.424, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.2260273972602738, | |
| "grad_norm": 16.730125427246094, | |
| "learning_rate": 5.17351598173516e-06, | |
| "loss": 0.4559, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.232876712328767, | |
| "grad_norm": 11.207518577575684, | |
| "learning_rate": 5.1278538812785395e-06, | |
| "loss": 0.4509, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.23972602739726, | |
| "grad_norm": 17.54400062561035, | |
| "learning_rate": 5.082191780821918e-06, | |
| "loss": 0.4697, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 32.376949310302734, | |
| "learning_rate": 5.0365296803652974e-06, | |
| "loss": 0.4096, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 2.2534246575342465, | |
| "grad_norm": 7.05587911605835, | |
| "learning_rate": 4.990867579908677e-06, | |
| "loss": 0.5346, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 2.26027397260274, | |
| "grad_norm": 12.208063125610352, | |
| "learning_rate": 4.945205479452055e-06, | |
| "loss": 0.4386, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.267123287671233, | |
| "grad_norm": 18.238994598388672, | |
| "learning_rate": 4.899543378995435e-06, | |
| "loss": 0.4373, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 18.42422103881836, | |
| "learning_rate": 4.853881278538813e-06, | |
| "loss": 0.4098, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 2.280821917808219, | |
| "grad_norm": 24.260120391845703, | |
| "learning_rate": 4.8082191780821926e-06, | |
| "loss": 0.454, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 2.287671232876712, | |
| "grad_norm": 16.63239860534668, | |
| "learning_rate": 4.762557077625571e-06, | |
| "loss": 0.4605, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 2.2945205479452055, | |
| "grad_norm": 36.93925094604492, | |
| "learning_rate": 4.7168949771689505e-06, | |
| "loss": 0.4821, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 25.88391876220703, | |
| "learning_rate": 4.671232876712329e-06, | |
| "loss": 0.4057, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 2.308219178082192, | |
| "grad_norm": 23.90517807006836, | |
| "learning_rate": 4.625570776255708e-06, | |
| "loss": 0.4876, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 2.315068493150685, | |
| "grad_norm": 11.417045593261719, | |
| "learning_rate": 4.579908675799088e-06, | |
| "loss": 0.5069, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 2.3219178082191783, | |
| "grad_norm": 14.539742469787598, | |
| "learning_rate": 4.534246575342466e-06, | |
| "loss": 0.4765, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 16.09026527404785, | |
| "learning_rate": 4.488584474885846e-06, | |
| "loss": 0.3791, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.3356164383561646, | |
| "grad_norm": 12.618657112121582, | |
| "learning_rate": 4.442922374429224e-06, | |
| "loss": 0.4316, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 2.3424657534246576, | |
| "grad_norm": 26.28303337097168, | |
| "learning_rate": 4.3972602739726035e-06, | |
| "loss": 0.4391, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 2.3493150684931505, | |
| "grad_norm": 13.06725025177002, | |
| "learning_rate": 4.351598173515982e-06, | |
| "loss": 0.4128, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 13.703057289123535, | |
| "learning_rate": 4.305936073059361e-06, | |
| "loss": 0.4001, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 2.363013698630137, | |
| "grad_norm": 17.728981018066406, | |
| "learning_rate": 4.260273972602741e-06, | |
| "loss": 0.4345, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.3698630136986303, | |
| "grad_norm": 3.869643211364746, | |
| "learning_rate": 4.214611872146119e-06, | |
| "loss": 0.4274, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 2.3767123287671232, | |
| "grad_norm": 24.957796096801758, | |
| "learning_rate": 4.168949771689499e-06, | |
| "loss": 0.4225, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 25.18575096130371, | |
| "learning_rate": 4.123287671232877e-06, | |
| "loss": 0.3921, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.3904109589041096, | |
| "grad_norm": 16.92748260498047, | |
| "learning_rate": 4.0776255707762565e-06, | |
| "loss": 0.3984, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 2.3972602739726026, | |
| "grad_norm": 10.364547729492188, | |
| "learning_rate": 4.031963470319635e-06, | |
| "loss": 0.4216, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.404109589041096, | |
| "grad_norm": 7.2683424949646, | |
| "learning_rate": 3.9863013698630135e-06, | |
| "loss": 0.4247, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 13.846619606018066, | |
| "learning_rate": 3.940639269406393e-06, | |
| "loss": 0.4144, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 2.4178082191780823, | |
| "grad_norm": 18.070049285888672, | |
| "learning_rate": 3.8949771689497714e-06, | |
| "loss": 0.486, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 2.4246575342465753, | |
| "grad_norm": 16.51365852355957, | |
| "learning_rate": 3.849315068493151e-06, | |
| "loss": 0.4958, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.4315068493150687, | |
| "grad_norm": 8.21628189086914, | |
| "learning_rate": 3.8036529680365297e-06, | |
| "loss": 0.4248, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 9.8461275100708, | |
| "learning_rate": 3.7579908675799087e-06, | |
| "loss": 0.4332, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 2.4452054794520546, | |
| "grad_norm": 20.94253921508789, | |
| "learning_rate": 3.7123287671232876e-06, | |
| "loss": 0.396, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 2.452054794520548, | |
| "grad_norm": 20.31191635131836, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 0.389, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 2.458904109589041, | |
| "grad_norm": 16.935224533081055, | |
| "learning_rate": 3.6210045662100455e-06, | |
| "loss": 0.4488, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 10.182275772094727, | |
| "learning_rate": 3.575342465753425e-06, | |
| "loss": 0.4497, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "eval_exact_match": 85.15610217596972, | |
| "eval_f1": 92.00858681094952, | |
| "eval_runtime": 411.6531, | |
| "eval_samples_per_second": 25.677, | |
| "eval_steps_per_second": 1.606, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4726027397260273, | |
| "grad_norm": 8.952213287353516, | |
| "learning_rate": 3.529680365296804e-06, | |
| "loss": 0.4872, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 2.4794520547945207, | |
| "grad_norm": 25.8715763092041, | |
| "learning_rate": 3.4840182648401828e-06, | |
| "loss": 0.4102, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 2.4863013698630136, | |
| "grad_norm": 11.87619400024414, | |
| "learning_rate": 3.4383561643835617e-06, | |
| "loss": 0.4416, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 15.143498420715332, | |
| "learning_rate": 3.3926940639269407e-06, | |
| "loss": 0.3819, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 21.528383255004883, | |
| "learning_rate": 3.3470319634703196e-06, | |
| "loss": 0.5546, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.506849315068493, | |
| "grad_norm": 7.487723350524902, | |
| "learning_rate": 3.3013698630136985e-06, | |
| "loss": 0.3982, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.5136986301369864, | |
| "grad_norm": 11.259400367736816, | |
| "learning_rate": 3.2557077625570775e-06, | |
| "loss": 0.4398, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 10.115367889404297, | |
| "learning_rate": 3.210045662100457e-06, | |
| "loss": 0.5101, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 2.5273972602739727, | |
| "grad_norm": 13.469361305236816, | |
| "learning_rate": 3.164383561643836e-06, | |
| "loss": 0.3954, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 2.5342465753424657, | |
| "grad_norm": 17.569135665893555, | |
| "learning_rate": 3.1187214611872147e-06, | |
| "loss": 0.4303, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.541095890410959, | |
| "grad_norm": 31.99173927307129, | |
| "learning_rate": 3.0730593607305937e-06, | |
| "loss": 0.4507, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 20.723060607910156, | |
| "learning_rate": 3.0273972602739726e-06, | |
| "loss": 0.5304, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.5547945205479454, | |
| "grad_norm": 18.118349075317383, | |
| "learning_rate": 2.9817351598173516e-06, | |
| "loss": 0.4309, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 2.5616438356164384, | |
| "grad_norm": 13.327851295471191, | |
| "learning_rate": 2.9360730593607305e-06, | |
| "loss": 0.4185, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 2.5684931506849313, | |
| "grad_norm": 20.624717712402344, | |
| "learning_rate": 2.8904109589041095e-06, | |
| "loss": 0.3756, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 22.176118850708008, | |
| "learning_rate": 2.8447488584474884e-06, | |
| "loss": 0.4031, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 2.5821917808219177, | |
| "grad_norm": 11.76338005065918, | |
| "learning_rate": 2.7990867579908678e-06, | |
| "loss": 0.3909, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 2.589041095890411, | |
| "grad_norm": 23.879566192626953, | |
| "learning_rate": 2.7534246575342467e-06, | |
| "loss": 0.4541, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.595890410958904, | |
| "grad_norm": 11.940811157226562, | |
| "learning_rate": 2.7077625570776257e-06, | |
| "loss": 0.3963, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 18.752647399902344, | |
| "learning_rate": 2.6621004566210046e-06, | |
| "loss": 0.3865, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.6095890410958904, | |
| "grad_norm": 11.951361656188965, | |
| "learning_rate": 2.6164383561643835e-06, | |
| "loss": 0.4571, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 2.616438356164384, | |
| "grad_norm": 11.501028060913086, | |
| "learning_rate": 2.5707762557077625e-06, | |
| "loss": 0.3833, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 2.6232876712328768, | |
| "grad_norm": 12.390273094177246, | |
| "learning_rate": 2.5251141552511414e-06, | |
| "loss": 0.3501, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 22.666540145874023, | |
| "learning_rate": 2.479452054794521e-06, | |
| "loss": 0.4726, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.636986301369863, | |
| "grad_norm": 3.564478874206543, | |
| "learning_rate": 2.4337899543378997e-06, | |
| "loss": 0.4299, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.643835616438356, | |
| "grad_norm": 6.8566365242004395, | |
| "learning_rate": 2.3881278538812787e-06, | |
| "loss": 0.4241, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 2.6506849315068495, | |
| "grad_norm": 24.194774627685547, | |
| "learning_rate": 2.3424657534246576e-06, | |
| "loss": 0.4408, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 25.843177795410156, | |
| "learning_rate": 2.296803652968037e-06, | |
| "loss": 0.4419, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 2.6643835616438354, | |
| "grad_norm": 13.00156307220459, | |
| "learning_rate": 2.251141552511416e-06, | |
| "loss": 0.4332, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 2.671232876712329, | |
| "grad_norm": 11.119675636291504, | |
| "learning_rate": 2.205479452054795e-06, | |
| "loss": 0.5019, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.678082191780822, | |
| "grad_norm": 8.283000946044922, | |
| "learning_rate": 2.159817351598174e-06, | |
| "loss": 0.4413, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 24.362268447875977, | |
| "learning_rate": 2.1141552511415528e-06, | |
| "loss": 0.4593, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 2.691780821917808, | |
| "grad_norm": 11.896214485168457, | |
| "learning_rate": 2.0684931506849317e-06, | |
| "loss": 0.4538, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 2.6986301369863015, | |
| "grad_norm": 27.805452346801758, | |
| "learning_rate": 2.0228310502283106e-06, | |
| "loss": 0.4669, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 2.7054794520547945, | |
| "grad_norm": 3.7341701984405518, | |
| "learning_rate": 1.9771689497716896e-06, | |
| "loss": 0.4041, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 17.065059661865234, | |
| "learning_rate": 1.931506849315069e-06, | |
| "loss": 0.3414, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.719178082191781, | |
| "grad_norm": 13.087998390197754, | |
| "learning_rate": 1.8858447488584477e-06, | |
| "loss": 0.4264, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 2.7260273972602738, | |
| "grad_norm": 16.43888282775879, | |
| "learning_rate": 1.8401826484018268e-06, | |
| "loss": 0.3675, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 2.732876712328767, | |
| "grad_norm": 15.753283500671387, | |
| "learning_rate": 1.7945205479452058e-06, | |
| "loss": 0.3896, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 28.405746459960938, | |
| "learning_rate": 1.7488584474885847e-06, | |
| "loss": 0.4751, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "eval_exact_match": 85.11825922421949, | |
| "eval_f1": 91.98946924096366, | |
| "eval_runtime": 411.5471, | |
| "eval_samples_per_second": 25.684, | |
| "eval_steps_per_second": 1.606, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.7465753424657535, | |
| "grad_norm": 15.57235050201416, | |
| "learning_rate": 1.7031963470319637e-06, | |
| "loss": 0.4558, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 2.7534246575342465, | |
| "grad_norm": 19.889366149902344, | |
| "learning_rate": 1.6575342465753428e-06, | |
| "loss": 0.4837, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.76027397260274, | |
| "grad_norm": 9.437919616699219, | |
| "learning_rate": 1.6118721461187218e-06, | |
| "loss": 0.4391, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 26.015281677246094, | |
| "learning_rate": 1.5662100456621007e-06, | |
| "loss": 0.4338, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 2.7739726027397262, | |
| "grad_norm": 6.268497943878174, | |
| "learning_rate": 1.5205479452054797e-06, | |
| "loss": 0.4388, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 2.780821917808219, | |
| "grad_norm": 16.472084045410156, | |
| "learning_rate": 1.4748858447488584e-06, | |
| "loss": 0.4591, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 2.787671232876712, | |
| "grad_norm": 44.26100540161133, | |
| "learning_rate": 1.4292237442922373e-06, | |
| "loss": 0.4123, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 9.49738597869873, | |
| "learning_rate": 1.3835616438356165e-06, | |
| "loss": 0.4442, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.8013698630136985, | |
| "grad_norm": 10.390971183776855, | |
| "learning_rate": 1.3378995433789954e-06, | |
| "loss": 0.4316, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 2.808219178082192, | |
| "grad_norm": 19.565282821655273, | |
| "learning_rate": 1.2922374429223744e-06, | |
| "loss": 0.4118, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.815068493150685, | |
| "grad_norm": 14.807978630065918, | |
| "learning_rate": 1.2465753424657535e-06, | |
| "loss": 0.4122, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 15.145249366760254, | |
| "learning_rate": 1.2009132420091325e-06, | |
| "loss": 0.4622, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 2.828767123287671, | |
| "grad_norm": 10.529396057128906, | |
| "learning_rate": 1.1552511415525116e-06, | |
| "loss": 0.3714, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 2.8356164383561646, | |
| "grad_norm": 7.163124084472656, | |
| "learning_rate": 1.1095890410958906e-06, | |
| "loss": 0.4101, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.8424657534246576, | |
| "grad_norm": 12.016539573669434, | |
| "learning_rate": 1.0639269406392695e-06, | |
| "loss": 0.4395, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 49.62667465209961, | |
| "learning_rate": 1.0182648401826485e-06, | |
| "loss": 0.4466, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 2.856164383561644, | |
| "grad_norm": 50.784488677978516, | |
| "learning_rate": 9.726027397260274e-07, | |
| "loss": 0.4313, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 2.863013698630137, | |
| "grad_norm": 8.124464988708496, | |
| "learning_rate": 9.269406392694065e-07, | |
| "loss": 0.426, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 2.8698630136986303, | |
| "grad_norm": 42.21592712402344, | |
| "learning_rate": 8.812785388127855e-07, | |
| "loss": 0.3858, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 17.541553497314453, | |
| "learning_rate": 8.356164383561644e-07, | |
| "loss": 0.4726, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.883561643835616, | |
| "grad_norm": 27.25008773803711, | |
| "learning_rate": 7.899543378995435e-07, | |
| "loss": 0.3574, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 2.8904109589041096, | |
| "grad_norm": 18.12972640991211, | |
| "learning_rate": 7.442922374429224e-07, | |
| "loss": 0.4477, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 2.897260273972603, | |
| "grad_norm": 5.854954242706299, | |
| "learning_rate": 6.986301369863015e-07, | |
| "loss": 0.4051, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 7.640255451202393, | |
| "learning_rate": 6.529680365296804e-07, | |
| "loss": 0.3945, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 2.910958904109589, | |
| "grad_norm": 10.086254119873047, | |
| "learning_rate": 6.073059360730594e-07, | |
| "loss": 0.3804, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 2.9178082191780823, | |
| "grad_norm": 10.96794319152832, | |
| "learning_rate": 5.625570776255707e-07, | |
| "loss": 0.388, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.9246575342465753, | |
| "grad_norm": 37.32515335083008, | |
| "learning_rate": 5.168949771689498e-07, | |
| "loss": 0.3637, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 3.6570990085601807, | |
| "learning_rate": 4.712328767123288e-07, | |
| "loss": 0.4405, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 2.9383561643835616, | |
| "grad_norm": 8.131304740905762, | |
| "learning_rate": 4.255707762557078e-07, | |
| "loss": 0.4112, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 2.9452054794520546, | |
| "grad_norm": 9.430889129638672, | |
| "learning_rate": 3.799086757990868e-07, | |
| "loss": 0.3816, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.952054794520548, | |
| "grad_norm": 16.61997413635254, | |
| "learning_rate": 3.342465753424658e-07, | |
| "loss": 0.4186, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 18.28761100769043, | |
| "learning_rate": 2.885844748858448e-07, | |
| "loss": 0.4461, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.9657534246575343, | |
| "grad_norm": 33.7155876159668, | |
| "learning_rate": 2.4292237442922377e-07, | |
| "loss": 0.3709, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 2.9726027397260273, | |
| "grad_norm": 9.754043579101562, | |
| "learning_rate": 1.9726027397260274e-07, | |
| "loss": 0.4302, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 2.9794520547945207, | |
| "grad_norm": 7.2017502784729, | |
| "learning_rate": 1.5159817351598174e-07, | |
| "loss": 0.4406, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 5.698849678039551, | |
| "learning_rate": 1.0593607305936073e-07, | |
| "loss": 0.3658, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 2.993150684931507, | |
| "grad_norm": 33.57563781738281, | |
| "learning_rate": 6.027397260273973e-08, | |
| "loss": 0.3881, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 32.373390197753906, | |
| "learning_rate": 1.4611872146118723e-08, | |
| "loss": 0.4073, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 21900, | |
| "total_flos": 2.0062620320048087e+18, | |
| "train_loss": 0.801641695292573, | |
| "train_runtime": 58377.4578, | |
| "train_samples_per_second": 4.502, | |
| "train_steps_per_second": 0.375 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 21900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0062620320048087e+18, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |