| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.525252525252525, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04208754208754209, | |
| "grad_norm": 44.5, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 9.708373413085937, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08417508417508418, | |
| "grad_norm": 20.5, | |
| "learning_rate": 4.537037037037038e-06, | |
| "loss": 7.022988891601562, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12626262626262627, | |
| "grad_norm": 13.375, | |
| "learning_rate": 4.94212962962963e-06, | |
| "loss": 4.4798516845703125, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 35.75, | |
| "learning_rate": 4.869791666666667e-06, | |
| "loss": 3.579981384277344, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21043771043771045, | |
| "grad_norm": 33.0, | |
| "learning_rate": 4.797453703703704e-06, | |
| "loss": 2.9001919555664064, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.25252525252525254, | |
| "grad_norm": 38.5, | |
| "learning_rate": 4.725115740740741e-06, | |
| "loss": 2.722982177734375, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2946127946127946, | |
| "grad_norm": 98.0, | |
| "learning_rate": 4.652777777777779e-06, | |
| "loss": 2.658799133300781, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 37.5, | |
| "learning_rate": 4.580439814814815e-06, | |
| "loss": 2.5711898803710938, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3787878787878788, | |
| "grad_norm": 148.0, | |
| "learning_rate": 4.508101851851852e-06, | |
| "loss": 2.5608544921875, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4208754208754209, | |
| "grad_norm": 100.0, | |
| "learning_rate": 4.435763888888889e-06, | |
| "loss": 2.5246621704101564, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.46296296296296297, | |
| "grad_norm": 111.0, | |
| "learning_rate": 4.363425925925927e-06, | |
| "loss": 2.504690399169922, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 74.5, | |
| "learning_rate": 4.2910879629629634e-06, | |
| "loss": 2.456527099609375, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5471380471380471, | |
| "grad_norm": 84.0, | |
| "learning_rate": 4.21875e-06, | |
| "loss": 2.476842498779297, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5892255892255892, | |
| "grad_norm": 26.125, | |
| "learning_rate": 4.1464120370370375e-06, | |
| "loss": 2.447265930175781, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6313131313131313, | |
| "grad_norm": 38.5, | |
| "learning_rate": 4.074074074074074e-06, | |
| "loss": 2.3707008361816406, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 23.375, | |
| "learning_rate": 4.001736111111112e-06, | |
| "loss": 2.374063720703125, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7154882154882155, | |
| "grad_norm": 21.875, | |
| "learning_rate": 3.929398148148148e-06, | |
| "loss": 2.3581019592285157, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 21.25, | |
| "learning_rate": 3.857060185185186e-06, | |
| "loss": 2.3752615356445315, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7996632996632996, | |
| "grad_norm": 48.0, | |
| "learning_rate": 3.7847222222222224e-06, | |
| "loss": 2.309081726074219, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8417508417508418, | |
| "grad_norm": 17.375, | |
| "learning_rate": 3.7123842592592594e-06, | |
| "loss": 2.265882568359375, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8838383838383839, | |
| "grad_norm": 38.5, | |
| "learning_rate": 3.640046296296297e-06, | |
| "loss": 2.245345458984375, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 26.875, | |
| "learning_rate": 3.5677083333333335e-06, | |
| "loss": 2.2155528259277344, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.968013468013468, | |
| "grad_norm": 40.0, | |
| "learning_rate": 3.4953703703703706e-06, | |
| "loss": 2.2166319274902344, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 141.0, | |
| "learning_rate": 3.423032407407408e-06, | |
| "loss": 2.2579159545898437, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0521885521885521, | |
| "grad_norm": 19.75, | |
| "learning_rate": 3.3506944444444447e-06, | |
| "loss": 2.1792347717285154, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.0942760942760943, | |
| "grad_norm": 159.0, | |
| "learning_rate": 3.2783564814814817e-06, | |
| "loss": 2.1700653076171874, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 26.125, | |
| "learning_rate": 3.2060185185185188e-06, | |
| "loss": 2.1729795837402346, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.1784511784511784, | |
| "grad_norm": 112.0, | |
| "learning_rate": 3.1336805555555562e-06, | |
| "loss": 2.1509564208984373, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2205387205387206, | |
| "grad_norm": 18.5, | |
| "learning_rate": 3.061342592592593e-06, | |
| "loss": 2.1331724548339843, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.2626262626262625, | |
| "grad_norm": 72.0, | |
| "learning_rate": 2.98900462962963e-06, | |
| "loss": 2.0941813659667967, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.3047138047138047, | |
| "grad_norm": 56.0, | |
| "learning_rate": 2.916666666666667e-06, | |
| "loss": 2.122632598876953, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "grad_norm": 22.75, | |
| "learning_rate": 2.8443287037037036e-06, | |
| "loss": 2.120499572753906, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 27.5, | |
| "learning_rate": 2.771990740740741e-06, | |
| "loss": 2.1208335876464846, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.430976430976431, | |
| "grad_norm": 19.125, | |
| "learning_rate": 2.699652777777778e-06, | |
| "loss": 2.1359043884277344, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.4730639730639732, | |
| "grad_norm": 18.5, | |
| "learning_rate": 2.6273148148148148e-06, | |
| "loss": 2.1154779052734374, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 25.625, | |
| "learning_rate": 2.554976851851852e-06, | |
| "loss": 2.107982635498047, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.557239057239057, | |
| "grad_norm": 17.375, | |
| "learning_rate": 2.4826388888888893e-06, | |
| "loss": 2.086489410400391, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.5993265993265995, | |
| "grad_norm": 23.125, | |
| "learning_rate": 2.410300925925926e-06, | |
| "loss": 2.1238157653808596, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6414141414141414, | |
| "grad_norm": 17.25, | |
| "learning_rate": 2.3379629629629634e-06, | |
| "loss": 2.0812089538574217, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.6835016835016834, | |
| "grad_norm": 19.5, | |
| "learning_rate": 2.265625e-06, | |
| "loss": 2.041307220458984, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7255892255892256, | |
| "grad_norm": 24.125, | |
| "learning_rate": 2.193287037037037e-06, | |
| "loss": 2.044098205566406, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.7676767676767677, | |
| "grad_norm": 21.875, | |
| "learning_rate": 2.120949074074074e-06, | |
| "loss": 2.1019331359863282, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.8097643097643097, | |
| "grad_norm": 21.25, | |
| "learning_rate": 2.048611111111111e-06, | |
| "loss": 2.048153839111328, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 26.0, | |
| "learning_rate": 1.9762731481481482e-06, | |
| "loss": 2.011888427734375, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.893939393939394, | |
| "grad_norm": 17.5, | |
| "learning_rate": 1.9039351851851853e-06, | |
| "loss": 2.103341369628906, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.936026936026936, | |
| "grad_norm": 21.5, | |
| "learning_rate": 1.8315972222222223e-06, | |
| "loss": 2.0502833557128906, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.9781144781144782, | |
| "grad_norm": 18.375, | |
| "learning_rate": 1.7592592592592594e-06, | |
| "loss": 2.0501133728027345, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.0202020202020203, | |
| "grad_norm": 20.0, | |
| "learning_rate": 1.6869212962962964e-06, | |
| "loss": 2.0467076110839844, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0622895622895623, | |
| "grad_norm": 17.75, | |
| "learning_rate": 1.6145833333333335e-06, | |
| "loss": 2.06115478515625, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.1043771043771042, | |
| "grad_norm": 30.75, | |
| "learning_rate": 1.5422453703703705e-06, | |
| "loss": 2.088555908203125, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.1464646464646466, | |
| "grad_norm": 25.25, | |
| "learning_rate": 1.4699074074074074e-06, | |
| "loss": 2.0032412719726564, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.1885521885521886, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.3975694444444446e-06, | |
| "loss": 2.0343482971191404, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.2306397306397305, | |
| "grad_norm": 14.875, | |
| "learning_rate": 1.3252314814814815e-06, | |
| "loss": 2.0316773986816408, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 17.25, | |
| "learning_rate": 1.2528935185185187e-06, | |
| "loss": 2.059365692138672, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.314814814814815, | |
| "grad_norm": 15.625, | |
| "learning_rate": 1.1805555555555556e-06, | |
| "loss": 2.0488226318359377, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.356902356902357, | |
| "grad_norm": 17.0, | |
| "learning_rate": 1.1082175925925926e-06, | |
| "loss": 2.025330810546875, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.398989898989899, | |
| "grad_norm": 17.75, | |
| "learning_rate": 1.0358796296296297e-06, | |
| "loss": 2.031147155761719, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.441077441077441, | |
| "grad_norm": 70.5, | |
| "learning_rate": 9.635416666666667e-07, | |
| "loss": 2.0263748168945312, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.483164983164983, | |
| "grad_norm": 16.5, | |
| "learning_rate": 8.912037037037037e-07, | |
| "loss": 1.9904652404785157, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.525252525252525, | |
| "grad_norm": 20.0, | |
| "learning_rate": 8.188657407407407e-07, | |
| "loss": 1.9979853820800781, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1782, | |
| "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": 1.3778250766441513e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |