| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 5750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5.7166666666666664e-06, | |
| "loss": 13.1691, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 1.155e-05, | |
| "loss": 5.7916, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 1.738333333333333e-05, | |
| "loss": 3.9554, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "eval_loss": 3.8166096210479736, | |
| "eval_runtime": 76.9069, | |
| "eval_samples_per_second": 22.105, | |
| "eval_steps_per_second": 1.391, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 2.3216666666666664e-05, | |
| "loss": 3.4177, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 2.9049999999999995e-05, | |
| "loss": 3.1088, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 3.488333333333333e-05, | |
| "loss": 2.9853, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "eval_loss": 2.979321002960205, | |
| "eval_runtime": 76.5995, | |
| "eval_samples_per_second": 22.193, | |
| "eval_steps_per_second": 1.397, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "learning_rate": 4.071666666666666e-05, | |
| "loss": 2.6633, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 4.655e-05, | |
| "loss": 1.3761, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 5.2383333333333324e-05, | |
| "loss": 0.6939, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "eval_loss": 0.972253143787384, | |
| "eval_runtime": 77.2431, | |
| "eval_samples_per_second": 22.008, | |
| "eval_steps_per_second": 1.385, | |
| "eval_wer": 0.9149368374277915, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 5.821666666666666e-05, | |
| "loss": 0.5141, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "learning_rate": 6.405e-05, | |
| "loss": 0.4189, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "learning_rate": 6.988333333333333e-05, | |
| "loss": 0.352, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "eval_loss": 0.6556435823440552, | |
| "eval_runtime": 76.3866, | |
| "eval_samples_per_second": 22.255, | |
| "eval_steps_per_second": 1.401, | |
| "eval_wer": 0.7296387989589285, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "learning_rate": 6.849230769230769e-05, | |
| "loss": 0.307, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.17, | |
| "learning_rate": 6.695384615384615e-05, | |
| "loss": 0.2658, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "learning_rate": 6.541538461538461e-05, | |
| "loss": 0.2369, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "eval_loss": 0.6314664483070374, | |
| "eval_runtime": 76.2318, | |
| "eval_samples_per_second": 22.3, | |
| "eval_steps_per_second": 1.404, | |
| "eval_wer": 0.6927569351869485, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.91, | |
| "learning_rate": 6.387692307692307e-05, | |
| "loss": 0.2084, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "learning_rate": 6.233846153846153e-05, | |
| "loss": 0.1904, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "learning_rate": 6.0799999999999994e-05, | |
| "loss": 0.1757, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "eval_loss": 0.6387060284614563, | |
| "eval_runtime": 76.1573, | |
| "eval_samples_per_second": 22.322, | |
| "eval_steps_per_second": 1.405, | |
| "eval_wer": 0.6723798641528598, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 16.52, | |
| "learning_rate": 5.9261538461538453e-05, | |
| "loss": 0.1693, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "learning_rate": 5.772307692307692e-05, | |
| "loss": 0.1572, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.26, | |
| "learning_rate": 5.618461538461538e-05, | |
| "loss": 0.1517, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.26, | |
| "eval_loss": 0.6607237458229065, | |
| "eval_runtime": 77.027, | |
| "eval_samples_per_second": 22.07, | |
| "eval_steps_per_second": 1.389, | |
| "eval_wer": 0.6265473243191773, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 19.13, | |
| "learning_rate": 5.464615384615384e-05, | |
| "loss": 0.1403, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 5.3107692307692305e-05, | |
| "loss": 0.1382, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 20.87, | |
| "learning_rate": 5.1569230769230765e-05, | |
| "loss": 0.1229, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 20.87, | |
| "eval_loss": 0.6519985198974609, | |
| "eval_runtime": 76.1295, | |
| "eval_samples_per_second": 22.33, | |
| "eval_steps_per_second": 1.405, | |
| "eval_wer": 0.6182949279502317, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "learning_rate": 5.0030769230769225e-05, | |
| "loss": 0.129, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 22.61, | |
| "learning_rate": 4.849230769230769e-05, | |
| "loss": 0.1172, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 23.48, | |
| "learning_rate": 4.695384615384615e-05, | |
| "loss": 0.1201, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 23.48, | |
| "eval_loss": 0.67499840259552, | |
| "eval_runtime": 76.4187, | |
| "eval_samples_per_second": 22.246, | |
| "eval_steps_per_second": 1.4, | |
| "eval_wer": 0.6115660509109376, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 24.35, | |
| "learning_rate": 4.541538461538461e-05, | |
| "loss": 0.1112, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 25.22, | |
| "learning_rate": 4.387692307692307e-05, | |
| "loss": 0.1131, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "learning_rate": 4.2338461538461536e-05, | |
| "loss": 0.1076, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "eval_loss": 0.6698060035705566, | |
| "eval_runtime": 75.169, | |
| "eval_samples_per_second": 22.616, | |
| "eval_steps_per_second": 1.423, | |
| "eval_wer": 0.6006474957151019, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 26.96, | |
| "learning_rate": 4.0799999999999996e-05, | |
| "loss": 0.104, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 27.83, | |
| "learning_rate": 3.9261538461538455e-05, | |
| "loss": 0.1014, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 28.7, | |
| "learning_rate": 3.772307692307692e-05, | |
| "loss": 0.1006, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 28.7, | |
| "eval_loss": 0.680385947227478, | |
| "eval_runtime": 75.8528, | |
| "eval_samples_per_second": 22.412, | |
| "eval_steps_per_second": 1.411, | |
| "eval_wer": 0.5871897416365137, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 29.57, | |
| "learning_rate": 3.618461538461538e-05, | |
| "loss": 0.1013, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 30.43, | |
| "learning_rate": 3.464615384615384e-05, | |
| "loss": 0.0972, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 31.3, | |
| "learning_rate": 3.310769230769231e-05, | |
| "loss": 0.0952, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 31.3, | |
| "eval_loss": 0.7009897828102112, | |
| "eval_runtime": 74.9836, | |
| "eval_samples_per_second": 22.672, | |
| "eval_steps_per_second": 1.427, | |
| "eval_wer": 0.600266615882689, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 32.17, | |
| "learning_rate": 3.156923076923077e-05, | |
| "loss": 0.091, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 33.04, | |
| "learning_rate": 3.0030769230769226e-05, | |
| "loss": 0.0924, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 33.91, | |
| "learning_rate": 2.849230769230769e-05, | |
| "loss": 0.0894, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 33.91, | |
| "eval_loss": 0.7068695425987244, | |
| "eval_runtime": 74.6323, | |
| "eval_samples_per_second": 22.778, | |
| "eval_steps_per_second": 1.434, | |
| "eval_wer": 0.584015743033073, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 34.78, | |
| "learning_rate": 2.695384615384615e-05, | |
| "loss": 0.0869, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 35.65, | |
| "learning_rate": 2.5415384615384612e-05, | |
| "loss": 0.0848, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 36.52, | |
| "learning_rate": 2.3876923076923075e-05, | |
| "loss": 0.0873, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 36.52, | |
| "eval_loss": 0.6764819622039795, | |
| "eval_runtime": 75.5212, | |
| "eval_samples_per_second": 22.51, | |
| "eval_steps_per_second": 1.417, | |
| "eval_wer": 0.5781121056306735, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 37.39, | |
| "learning_rate": 2.2338461538461534e-05, | |
| "loss": 0.0845, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 38.26, | |
| "learning_rate": 2.0799999999999997e-05, | |
| "loss": 0.0813, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 39.13, | |
| "learning_rate": 1.9261538461538457e-05, | |
| "loss": 0.0798, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 39.13, | |
| "eval_loss": 0.6821776032447815, | |
| "eval_runtime": 76.5806, | |
| "eval_samples_per_second": 22.199, | |
| "eval_steps_per_second": 1.397, | |
| "eval_wer": 0.5617342728369199, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 1.772307692307692e-05, | |
| "loss": 0.0783, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 40.87, | |
| "learning_rate": 1.6199999999999997e-05, | |
| "loss": 0.0753, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 41.74, | |
| "learning_rate": 1.466153846153846e-05, | |
| "loss": 0.0767, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 41.74, | |
| "eval_loss": 0.7003459334373474, | |
| "eval_runtime": 74.7077, | |
| "eval_samples_per_second": 22.755, | |
| "eval_steps_per_second": 1.432, | |
| "eval_wer": 0.5637656319431219, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 42.61, | |
| "learning_rate": 1.3123076923076922e-05, | |
| "loss": 0.0739, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "learning_rate": 1.1584615384615385e-05, | |
| "loss": 0.0766, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 44.35, | |
| "learning_rate": 1.0046153846153846e-05, | |
| "loss": 0.0717, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 44.35, | |
| "eval_loss": 0.6969868540763855, | |
| "eval_runtime": 74.6644, | |
| "eval_samples_per_second": 22.769, | |
| "eval_steps_per_second": 1.433, | |
| "eval_wer": 0.5590681140100299, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 45.22, | |
| "learning_rate": 8.507692307692307e-06, | |
| "loss": 0.0732, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 46.09, | |
| "learning_rate": 6.9692307692307684e-06, | |
| "loss": 0.0713, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 46.96, | |
| "learning_rate": 5.4307692307692306e-06, | |
| "loss": 0.0687, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 46.96, | |
| "eval_loss": 0.7072671055793762, | |
| "eval_runtime": 74.5796, | |
| "eval_samples_per_second": 22.794, | |
| "eval_steps_per_second": 1.435, | |
| "eval_wer": 0.560337713451406, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 47.83, | |
| "learning_rate": 3.892307692307692e-06, | |
| "loss": 0.0674, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 48.7, | |
| "learning_rate": 2.3538461538461536e-06, | |
| "loss": 0.0702, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 49.57, | |
| "learning_rate": 8.153846153846153e-07, | |
| "loss": 0.0669, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 49.57, | |
| "eval_loss": 0.7059715390205383, | |
| "eval_runtime": 74.624, | |
| "eval_samples_per_second": 22.781, | |
| "eval_steps_per_second": 1.434, | |
| "eval_wer": 0.5558941154065892, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 5750, | |
| "total_flos": 2.9609940258263142e+19, | |
| "train_loss": 0.7598760600297347, | |
| "train_runtime": 13000.5685, | |
| "train_samples_per_second": 14.138, | |
| "train_steps_per_second": 0.442 | |
| } | |
| ], | |
| "max_steps": 5750, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.9609940258263142e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |