| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 0.1373787671327591, | |
| "learning_rate": 9.928057553956835e-05, | |
| "loss": 1.0964, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.23093585669994354, | |
| "learning_rate": 9.784172661870504e-05, | |
| "loss": 0.864, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 0.1213061660528183, | |
| "learning_rate": 9.640287769784174e-05, | |
| "loss": 0.5938, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.0565805584192276, | |
| "learning_rate": 9.496402877697842e-05, | |
| "loss": 0.5041, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.057311397045850754, | |
| "learning_rate": 9.35251798561151e-05, | |
| "loss": 0.443, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.06346569955348969, | |
| "learning_rate": 9.20863309352518e-05, | |
| "loss": 0.533, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.04802313446998596, | |
| "learning_rate": 9.06474820143885e-05, | |
| "loss": 0.4617, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.05171512812376022, | |
| "learning_rate": 8.920863309352519e-05, | |
| "loss": 0.4229, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.06379946321249008, | |
| "learning_rate": 8.776978417266187e-05, | |
| "loss": 0.4794, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.06747135519981384, | |
| "learning_rate": 8.633093525179857e-05, | |
| "loss": 0.4607, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.05382410064339638, | |
| "learning_rate": 8.489208633093527e-05, | |
| "loss": 0.4922, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.05737805739045143, | |
| "learning_rate": 8.345323741007195e-05, | |
| "loss": 0.5137, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.05210845544934273, | |
| "learning_rate": 8.201438848920863e-05, | |
| "loss": 0.4611, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.059042274951934814, | |
| "learning_rate": 8.057553956834533e-05, | |
| "loss": 0.457, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.05036056786775589, | |
| "learning_rate": 7.913669064748202e-05, | |
| "loss": 0.4712, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.06213638186454773, | |
| "learning_rate": 7.769784172661872e-05, | |
| "loss": 0.4253, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.08077629655599594, | |
| "learning_rate": 7.62589928057554e-05, | |
| "loss": 0.4652, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.067039355635643, | |
| "learning_rate": 7.48201438848921e-05, | |
| "loss": 0.4623, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.06331554055213928, | |
| "learning_rate": 7.338129496402878e-05, | |
| "loss": 0.4991, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.07635468244552612, | |
| "learning_rate": 7.194244604316547e-05, | |
| "loss": 0.4185, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.07588763535022736, | |
| "learning_rate": 7.050359712230215e-05, | |
| "loss": 0.4232, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.06745592504739761, | |
| "learning_rate": 6.906474820143886e-05, | |
| "loss": 0.4254, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.07882258296012878, | |
| "learning_rate": 6.762589928057555e-05, | |
| "loss": 0.4437, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.07268062978982925, | |
| "learning_rate": 6.618705035971223e-05, | |
| "loss": 0.4725, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.07357478886842728, | |
| "learning_rate": 6.474820143884892e-05, | |
| "loss": 0.4185, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.07707969099283218, | |
| "learning_rate": 6.330935251798561e-05, | |
| "loss": 0.4349, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.07666852325201035, | |
| "learning_rate": 6.187050359712231e-05, | |
| "loss": 0.5036, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.07082293182611465, | |
| "learning_rate": 6.0431654676258996e-05, | |
| "loss": 0.4336, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.0714285714285716, | |
| "grad_norm": 0.07426683604717255, | |
| "learning_rate": 5.899280575539569e-05, | |
| "loss": 0.431, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.06738661974668503, | |
| "learning_rate": 5.755395683453237e-05, | |
| "loss": 0.416, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.2142857142857144, | |
| "grad_norm": 0.07940518110990524, | |
| "learning_rate": 5.611510791366906e-05, | |
| "loss": 0.4314, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.08810248970985413, | |
| "learning_rate": 5.467625899280576e-05, | |
| "loss": 0.4893, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.357142857142857, | |
| "grad_norm": 0.08641257137060165, | |
| "learning_rate": 5.323741007194245e-05, | |
| "loss": 0.3923, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 0.09313848614692688, | |
| "learning_rate": 5.179856115107914e-05, | |
| "loss": 0.4295, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.08842173963785172, | |
| "learning_rate": 5.035971223021583e-05, | |
| "loss": 0.4141, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.08694295585155487, | |
| "learning_rate": 4.892086330935252e-05, | |
| "loss": 0.4253, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.642857142857143, | |
| "grad_norm": 0.09336841106414795, | |
| "learning_rate": 4.748201438848921e-05, | |
| "loss": 0.4374, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.08530748635530472, | |
| "learning_rate": 4.60431654676259e-05, | |
| "loss": 0.4324, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.7857142857142856, | |
| "grad_norm": 0.08860356360673904, | |
| "learning_rate": 4.460431654676259e-05, | |
| "loss": 0.421, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.10588660836219788, | |
| "learning_rate": 4.3165467625899284e-05, | |
| "loss": 0.4553, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.928571428571429, | |
| "grad_norm": 0.10128331184387207, | |
| "learning_rate": 4.1726618705035975e-05, | |
| "loss": 0.4084, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.1228618174791336, | |
| "learning_rate": 4.0287769784172666e-05, | |
| "loss": 0.4199, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.0714285714285716, | |
| "grad_norm": 0.10508912056684494, | |
| "learning_rate": 3.884892086330936e-05, | |
| "loss": 0.3835, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 0.09007140249013901, | |
| "learning_rate": 3.741007194244605e-05, | |
| "loss": 0.4385, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.2142857142857144, | |
| "grad_norm": 0.09553549438714981, | |
| "learning_rate": 3.597122302158273e-05, | |
| "loss": 0.4194, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 0.11486919969320297, | |
| "learning_rate": 3.453237410071943e-05, | |
| "loss": 0.4205, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.357142857142857, | |
| "grad_norm": 0.12448228150606155, | |
| "learning_rate": 3.3093525179856116e-05, | |
| "loss": 0.408, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.10864632576704025, | |
| "learning_rate": 3.165467625899281e-05, | |
| "loss": 0.4175, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.11113475263118744, | |
| "learning_rate": 3.0215827338129498e-05, | |
| "loss": 0.3946, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.11032601445913315, | |
| "learning_rate": 2.8776978417266186e-05, | |
| "loss": 0.4134, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.642857142857143, | |
| "grad_norm": 0.11257655918598175, | |
| "learning_rate": 2.733812949640288e-05, | |
| "loss": 0.3905, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 0.12174879759550095, | |
| "learning_rate": 2.589928057553957e-05, | |
| "loss": 0.4063, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.7857142857142856, | |
| "grad_norm": 0.12508544325828552, | |
| "learning_rate": 2.446043165467626e-05, | |
| "loss": 0.4375, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.857142857142857, | |
| "grad_norm": 0.1218411847949028, | |
| "learning_rate": 2.302158273381295e-05, | |
| "loss": 0.3975, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.928571428571429, | |
| "grad_norm": 0.12357600778341293, | |
| "learning_rate": 2.1582733812949642e-05, | |
| "loss": 0.4305, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.10841184109449387, | |
| "learning_rate": 2.0143884892086333e-05, | |
| "loss": 0.4134, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.071428571428571, | |
| "grad_norm": 0.12538909912109375, | |
| "learning_rate": 1.8705035971223024e-05, | |
| "loss": 0.399, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.142857142857143, | |
| "grad_norm": 0.15004506707191467, | |
| "learning_rate": 1.7266187050359716e-05, | |
| "loss": 0.394, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.214285714285714, | |
| "grad_norm": 0.1259927898645401, | |
| "learning_rate": 1.5827338129496403e-05, | |
| "loss": 0.4336, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.13113726675510406, | |
| "learning_rate": 1.4388489208633093e-05, | |
| "loss": 0.4203, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.357142857142857, | |
| "grad_norm": 0.13534702360630035, | |
| "learning_rate": 1.2949640287769784e-05, | |
| "loss": 0.3731, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.428571428571429, | |
| "grad_norm": 0.14081259071826935, | |
| "learning_rate": 1.1510791366906475e-05, | |
| "loss": 0.3869, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.13458286225795746, | |
| "learning_rate": 1.0071942446043167e-05, | |
| "loss": 0.4143, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.12921994924545288, | |
| "learning_rate": 8.633093525179858e-06, | |
| "loss": 0.3854, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.642857142857143, | |
| "grad_norm": 0.14026252925395966, | |
| "learning_rate": 7.1942446043165465e-06, | |
| "loss": 0.4295, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.714285714285714, | |
| "grad_norm": 0.14407575130462646, | |
| "learning_rate": 5.755395683453238e-06, | |
| "loss": 0.4228, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.785714285714286, | |
| "grad_norm": 0.11860533058643341, | |
| "learning_rate": 4.316546762589929e-06, | |
| "loss": 0.3834, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 0.1333409547805786, | |
| "learning_rate": 2.877697841726619e-06, | |
| "loss": 0.364, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 4.928571428571429, | |
| "grad_norm": 0.15710106492042542, | |
| "learning_rate": 1.4388489208633094e-06, | |
| "loss": 0.387, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.16918131709098816, | |
| "learning_rate": 0.0, | |
| "loss": 0.355, | |
| "step": 700 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 700, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.7064167003930624e+17, | |
| "train_batch_size": 10, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |