| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 200.0, | |
| "global_step": 3600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 5.82e-05, | |
| "loss": 9.8118, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "learning_rate": 0.0001182, | |
| "loss": 3.4789, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 0.00017819999999999997, | |
| "loss": 3.114, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 22.22, | |
| "learning_rate": 0.0002382, | |
| "loss": 2.721, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 27.78, | |
| "learning_rate": 0.0002982, | |
| "loss": 1.7294, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 27.78, | |
| "eval_loss": 0.8540233373641968, | |
| "eval_runtime": 10.9769, | |
| "eval_samples_per_second": 19.313, | |
| "eval_steps_per_second": 0.638, | |
| "eval_wer": 0.9943609022556391, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 0.0002906129032258064, | |
| "loss": 1.351, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 38.89, | |
| "learning_rate": 0.0002809354838709677, | |
| "loss": 1.1775, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 44.44, | |
| "learning_rate": 0.000271258064516129, | |
| "loss": 1.0539, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.0002615806451612903, | |
| "loss": 0.9587, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 55.56, | |
| "learning_rate": 0.0002519032258064516, | |
| "loss": 0.8863, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 55.56, | |
| "eval_loss": 0.7282317280769348, | |
| "eval_runtime": 11.4744, | |
| "eval_samples_per_second": 18.476, | |
| "eval_steps_per_second": 0.61, | |
| "eval_wer": 0.731203007518797, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 61.11, | |
| "learning_rate": 0.00024222580645161287, | |
| "loss": 0.7985, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 0.00023254838709677416, | |
| "loss": 0.7323, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 72.22, | |
| "learning_rate": 0.00022287096774193548, | |
| "loss": 0.6712, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 77.78, | |
| "learning_rate": 0.00021319354838709678, | |
| "loss": 0.6214, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "learning_rate": 0.00020351612903225804, | |
| "loss": 0.5789, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "eval_loss": 0.817844033241272, | |
| "eval_runtime": 11.023, | |
| "eval_samples_per_second": 19.232, | |
| "eval_steps_per_second": 0.635, | |
| "eval_wer": 0.8101503759398496, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 88.89, | |
| "learning_rate": 0.00019383870967741934, | |
| "loss": 0.5519, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 94.44, | |
| "learning_rate": 0.00018416129032258063, | |
| "loss": 0.4957, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 0.00017448387096774193, | |
| "loss": 0.4599, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 105.56, | |
| "learning_rate": 0.0001648064516129032, | |
| "loss": 0.4272, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 111.11, | |
| "learning_rate": 0.0001551290322580645, | |
| "loss": 0.3899, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 111.11, | |
| "eval_loss": 0.803404688835144, | |
| "eval_runtime": 10.8752, | |
| "eval_samples_per_second": 19.494, | |
| "eval_steps_per_second": 0.644, | |
| "eval_wer": 0.7700501253132832, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 116.67, | |
| "learning_rate": 0.0001454516129032258, | |
| "loss": 0.3725, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 122.22, | |
| "learning_rate": 0.00013577419354838708, | |
| "loss": 0.3373, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 127.78, | |
| "learning_rate": 0.00012609677419354837, | |
| "loss": 0.3134, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 133.33, | |
| "learning_rate": 0.00011641935483870967, | |
| "loss": 0.3057, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 138.89, | |
| "learning_rate": 0.00010674193548387096, | |
| "loss": 0.2869, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 138.89, | |
| "eval_loss": 0.9060825705528259, | |
| "eval_runtime": 10.9301, | |
| "eval_samples_per_second": 19.396, | |
| "eval_steps_per_second": 0.64, | |
| "eval_wer": 0.699874686716792, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 144.44, | |
| "learning_rate": 9.706451612903224e-05, | |
| "loss": 0.2579, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 8.738709677419355e-05, | |
| "loss": 0.2383, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 155.56, | |
| "learning_rate": 7.770967741935483e-05, | |
| "loss": 0.2253, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 161.11, | |
| "learning_rate": 6.803225806451613e-05, | |
| "loss": 0.2144, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "learning_rate": 5.835483870967742e-05, | |
| "loss": 0.1934, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "eval_loss": 0.9400397539138794, | |
| "eval_runtime": 11.0289, | |
| "eval_samples_per_second": 19.222, | |
| "eval_steps_per_second": 0.635, | |
| "eval_wer": 0.7105263157894737, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 172.22, | |
| "learning_rate": 4.8677419354838705e-05, | |
| "loss": 0.1874, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 177.78, | |
| "learning_rate": 3.9e-05, | |
| "loss": 0.1778, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 183.33, | |
| "learning_rate": 2.932258064516129e-05, | |
| "loss": 0.171, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 188.89, | |
| "learning_rate": 1.9645161290322582e-05, | |
| "loss": 0.1568, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 194.44, | |
| "learning_rate": 9.96774193548387e-06, | |
| "loss": 0.1551, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 194.44, | |
| "eval_loss": 0.966656506061554, | |
| "eval_runtime": 11.4149, | |
| "eval_samples_per_second": 18.572, | |
| "eval_steps_per_second": 0.613, | |
| "eval_wer": 0.6954887218045113, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 2.903225806451613e-07, | |
| "loss": 0.145, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "step": 3600, | |
| "total_flos": 2.405940879390409e+19, | |
| "train_loss": 0.9929841698540581, | |
| "train_runtime": 8863.3906, | |
| "train_samples_per_second": 12.501, | |
| "train_steps_per_second": 0.406 | |
| } | |
| ], | |
| "max_steps": 3600, | |
| "num_train_epochs": 200, | |
| "total_flos": 2.405940879390409e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |