| { | |
| "best_metric": 0.06773073226213455, | |
| "best_model_checkpoint": "./beans_outputs/checkpoint-520", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 650, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 2.198350667953491, | |
| "learning_rate": 1.9692307692307696e-05, | |
| "loss": 1.0245, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.845354437828064, | |
| "learning_rate": 1.9384615384615386e-05, | |
| "loss": 0.9453, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 2.0754780769348145, | |
| "learning_rate": 1.907692307692308e-05, | |
| "loss": 0.8404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.6118457317352295, | |
| "learning_rate": 1.876923076923077e-05, | |
| "loss": 0.6951, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 4.05247688293457, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.661, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.0832912921905518, | |
| "learning_rate": 1.8153846153846155e-05, | |
| "loss": 0.5799, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.8067588806152344, | |
| "learning_rate": 1.784615384615385e-05, | |
| "loss": 0.5194, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.2357735633850098, | |
| "learning_rate": 1.753846153846154e-05, | |
| "loss": 0.3928, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 2.420240640640259, | |
| "learning_rate": 1.7230769230769234e-05, | |
| "loss": 0.3523, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 2.9507646560668945, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.3683, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.958319902420044, | |
| "learning_rate": 1.6615384615384618e-05, | |
| "loss": 0.3313, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 2.633815050125122, | |
| "learning_rate": 1.630769230769231e-05, | |
| "loss": 0.2443, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.158595561981201, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.2825, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.21424901485443115, | |
| "eval_runtime": 0.8674, | |
| "eval_samples_per_second": 153.326, | |
| "eval_steps_per_second": 19.598, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.5242034196853638, | |
| "learning_rate": 1.5692307692307693e-05, | |
| "loss": 0.199, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 3.173902750015259, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2217, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.485329031944275, | |
| "learning_rate": 1.5076923076923078e-05, | |
| "loss": 0.2043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.51258784532547, | |
| "learning_rate": 1.4769230769230772e-05, | |
| "loss": 0.2196, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.592962920665741, | |
| "learning_rate": 1.4461538461538462e-05, | |
| "loss": 0.2125, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 4.033792018890381, | |
| "learning_rate": 1.4153846153846156e-05, | |
| "loss": 0.1623, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.4626278281211853, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.2522, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.6243513822555542, | |
| "learning_rate": 1.353846153846154e-05, | |
| "loss": 0.1803, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 2.1075823307037354, | |
| "learning_rate": 1.3230769230769231e-05, | |
| "loss": 0.2132, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 2.020235776901245, | |
| "learning_rate": 1.2923076923076925e-05, | |
| "loss": 0.1699, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 5.884667873382568, | |
| "learning_rate": 1.2615384615384616e-05, | |
| "loss": 0.2601, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3920741081237793, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.1413, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3517548441886902, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.1331, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.1346566081047058, | |
| "eval_runtime": 0.7649, | |
| "eval_samples_per_second": 173.887, | |
| "eval_steps_per_second": 22.226, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.6536712646484375, | |
| "learning_rate": 1.1692307692307694e-05, | |
| "loss": 0.0914, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.33924439549446106, | |
| "learning_rate": 1.1384615384615385e-05, | |
| "loss": 0.1479, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 9.13126277923584, | |
| "learning_rate": 1.1076923076923079e-05, | |
| "loss": 0.126, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 8.217650413513184, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.1377, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 5.396190166473389, | |
| "learning_rate": 1.0461538461538463e-05, | |
| "loss": 0.1742, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.3621850311756134, | |
| "learning_rate": 1.0153846153846154e-05, | |
| "loss": 0.1501, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 6.045478820800781, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 0.1869, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 1.4705922603607178, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 0.0727, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.34623193740844727, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.141, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 11.963224411010742, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 0.121, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.3321854770183563, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 0.1558, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 1.1333885192871094, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 0.1326, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.29565006494522095, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.1427, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.09987509995698929, | |
| "eval_runtime": 0.8311, | |
| "eval_samples_per_second": 160.026, | |
| "eval_steps_per_second": 20.454, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 2.2414183616638184, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.1349, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 6.092810153961182, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 0.1578, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 8.454133033752441, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 0.1439, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.23263706266880035, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 0.0719, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 3.942017078399658, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 0.1106, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 9.800943374633789, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.0766, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 4.558837413787842, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.0982, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 8.255152702331543, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 0.0813, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.2182413637638092, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 0.0878, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.21229100227355957, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 0.0768, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.9183093905448914, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.137, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.21079778671264648, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.1281, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.30409324169158936, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0808, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9924812030075187, | |
| "eval_loss": 0.06773073226213455, | |
| "eval_runtime": 1.1246, | |
| "eval_samples_per_second": 118.26, | |
| "eval_steps_per_second": 15.116, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 8.730031967163086, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 0.1009, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 1.607771635055542, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 0.1619, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.3554823398590088, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.1553, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 1.2201733589172363, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.0496, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.32220742106437683, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.1368, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.26765453815460205, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 0.0792, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 9.684674263000488, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 0.0723, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 3.574254274368286, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.0882, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 4.952356338500977, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.0877, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.25640788674354553, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 0.0826, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.21630094945430756, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 0.0651, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.23137542605400085, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.0801, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6233453750610352, | |
| "learning_rate": 0.0, | |
| "loss": 0.1156, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.08386705815792084, | |
| "eval_runtime": 0.7441, | |
| "eval_samples_per_second": 178.742, | |
| "eval_steps_per_second": 22.847, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 650, | |
| "total_flos": 4.006371770595533e+17, | |
| "train_loss": 0.2191925185918808, | |
| "train_runtime": 70.8951, | |
| "train_samples_per_second": 72.925, | |
| "train_steps_per_second": 9.168 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 650, | |
| "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": 4.006371770595533e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |