| { | |
| "best_metric": 27.47628083491461, | |
| "best_model_checkpoint": "./checkpoint-1000", | |
| "epoch": 24.04, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.45e-08, | |
| "loss": 1.3486, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 7.2e-08, | |
| "loss": 1.1313, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 1.095e-07, | |
| "loss": 1.2018, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 1.4699999999999998e-07, | |
| "loss": 1.0324, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.845e-07, | |
| "loss": 0.7448, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 2.2199999999999998e-07, | |
| "loss": 0.6496, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 2.5949999999999996e-07, | |
| "loss": 0.4192, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 2.9699999999999997e-07, | |
| "loss": 0.5148, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "eval_loss": 0.4189453125, | |
| "eval_runtime": 732.3243, | |
| "eval_samples_per_second": 3.029, | |
| "eval_steps_per_second": 0.096, | |
| "eval_wer": 39.34535104364326, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 2.997683575414969e-07, | |
| "loss": 0.3818, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 2.9899197858556995e-07, | |
| "loss": 0.3758, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 2.9767194405746293e-07, | |
| "loss": 0.3957, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "learning_rate": 2.958130706200157e-07, | |
| "loss": 0.2837, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "learning_rate": 2.934221411008129e-07, | |
| "loss": 0.4078, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 2.905078797423817e-07, | |
| "loss": 0.27, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 2.870809203683774e-07, | |
| "loss": 0.3376, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 2.831537675819134e-07, | |
| "loss": 0.3041, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_loss": 0.33349609375, | |
| "eval_runtime": 737.3023, | |
| "eval_samples_per_second": 3.008, | |
| "eval_steps_per_second": 0.095, | |
| "eval_wer": 29.573055028462996, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 2.78740751137621e-07, | |
| "loss": 0.259, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "learning_rate": 2.738579736539286e-07, | |
| "loss": 0.334, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "learning_rate": 2.6852325185635355e-07, | |
| "loss": 0.2305, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "learning_rate": 2.6275605156620243e-07, | |
| "loss": 0.3156, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "learning_rate": 2.5657741667190066e-07, | |
| "loss": 0.2408, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "learning_rate": 2.500098923421254e-07, | |
| "loss": 0.2577, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "learning_rate": 2.4307744276092984e-07, | |
| "loss": 0.2818, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "learning_rate": 2.3580536368503515e-07, | |
| "loss": 0.1961, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "eval_loss": 0.318603515625, | |
| "eval_runtime": 746.4809, | |
| "eval_samples_per_second": 2.971, | |
| "eval_steps_per_second": 0.094, | |
| "eval_wer": 27.779886148007588, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "learning_rate": 2.2822019014235677e-07, | |
| "loss": 0.3018, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "learning_rate": 2.2034959960856529e-07, | |
| "loss": 0.2028, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "learning_rate": 2.1222231101497956e-07, | |
| "loss": 0.2588, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "learning_rate": 2.0386797995630034e-07, | |
| "loss": 0.2308, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 1.953170904805618e-07, | |
| "loss": 0.2022, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "learning_rate": 1.8660084385614562e-07, | |
| "loss": 0.2688, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 12.04, | |
| "learning_rate": 1.7775104472173675e-07, | |
| "loss": 0.1752, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "learning_rate": 1.6879998503464563e-07, | |
| "loss": 0.2579, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "eval_loss": 0.316650390625, | |
| "eval_runtime": 737.1707, | |
| "eval_samples_per_second": 3.009, | |
| "eval_steps_per_second": 0.095, | |
| "eval_wer": 27.571157495256166, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "learning_rate": 1.5978032624095823e-07, | |
| "loss": 0.195, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 14.01, | |
| "learning_rate": 1.5072498009746107e-07, | |
| "loss": 0.2044, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 1.416669885802099e-07, | |
| "loss": 0.2308, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "learning_rate": 1.3263940331794248e-07, | |
| "loss": 0.1651, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "learning_rate": 1.2367516499026965e-07, | |
| "loss": 0.2565, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "learning_rate": 1.1480698313070749e-07, | |
| "loss": 0.1693, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 16.01, | |
| "learning_rate": 1.0606721677313713e-07, | |
| "loss": 0.2152, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "learning_rate": 9.748775637719916e-08, | |
| "loss": 0.2034, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "eval_loss": 0.31787109375, | |
| "eval_runtime": 744.5018, | |
| "eval_samples_per_second": 2.979, | |
| "eval_steps_per_second": 0.094, | |
| "eval_wer": 27.47628083491461, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 8.976315836886941e-08, | |
| "loss": 0.1694, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "learning_rate": 8.157864194541929e-08, | |
| "loss": 0.227, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 17.04, | |
| "learning_rate": 7.36437876374443e-08, | |
| "loss": 0.1607, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "learning_rate": 6.598754886780604e-08, | |
| "loss": 0.2253, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "learning_rate": 5.863786242148148e-08, | |
| "loss": 0.1737, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "learning_rate": 5.1621546507264605e-08, | |
| "loss": 0.1891, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "learning_rate": 4.496420290103045e-08, | |
| "loss": 0.2066, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 19.04, | |
| "learning_rate": 3.869012352763169e-08, | |
| "loss": 0.1478, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 19.04, | |
| "eval_loss": 0.3193359375, | |
| "eval_runtime": 742.6411, | |
| "eval_samples_per_second": 2.987, | |
| "eval_steps_per_second": 0.094, | |
| "eval_wer": 27.523719165085385, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "learning_rate": 3.28222018222991e-08, | |
| "loss": 0.2362, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 20.03, | |
| "learning_rate": 2.738184919497899e-08, | |
| "loss": 0.1564, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "learning_rate": 2.2388916902420573e-08, | |
| "loss": 0.2014, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "learning_rate": 1.7861623613092718e-08, | |
| "loss": 0.1918, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "learning_rate": 1.381648892923838e-08, | |
| "loss": 0.1614, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "learning_rate": 1.0268273108637776e-08, | |
| "loss": 0.2187, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 22.04, | |
| "learning_rate": 7.229923206028154e-09, | |
| "loss": 0.1492, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "learning_rate": 4.712525830705338e-09, | |
| "loss": 0.2169, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "eval_loss": 0.31982421875, | |
| "eval_runtime": 739.4797, | |
| "eval_samples_per_second": 2.999, | |
| "eval_steps_per_second": 0.095, | |
| "eval_wer": 27.504743833017077, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 23.03, | |
| "learning_rate": 2.7252666926886157e-09, | |
| "loss": 0.1702, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "learning_rate": 1.2753970850610251e-09, | |
| "loss": 0.1815, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "learning_rate": 3.682074247873257e-10, | |
| "loss": 0.2022, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 24.04, | |
| "learning_rate": 7.0079485561924665e-12, | |
| "loss": 0.1448, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 24.04, | |
| "step": 1500, | |
| "total_flos": 2.0191356490809868e+20, | |
| "train_loss": 0.31971333821614584, | |
| "train_runtime": 22038.1709, | |
| "train_samples_per_second": 4.356, | |
| "train_steps_per_second": 0.068 | |
| } | |
| ], | |
| "max_steps": 1500, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 2.0191356490809868e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |