| { | |
| "best_metric": 1.0742337703704834, | |
| "best_model_checkpoint": "./checkpoint-200", | |
| "epoch": 71.42857142857143, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.4e-07, | |
| "loss": 3.2633, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 3.4000000000000003e-07, | |
| "loss": 3.235, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 5.4e-07, | |
| "loss": 3.164, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "learning_rate": 7.4e-07, | |
| "loss": 2.9792, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 9.400000000000001e-07, | |
| "loss": 2.7533, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.57, | |
| "learning_rate": 1.14e-06, | |
| "loss": 2.5502, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 1.34e-06, | |
| "loss": 2.3845, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 11.43, | |
| "learning_rate": 1.54e-06, | |
| "loss": 2.161, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 12.86, | |
| "learning_rate": 1.74e-06, | |
| "loss": 1.9855, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 1.94e-06, | |
| "loss": 1.8225, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "eval_loss": 1.760784387588501, | |
| "eval_runtime": 425.4331, | |
| "eval_samples_per_second": 1.203, | |
| "eval_steps_per_second": 0.075, | |
| "eval_wer": 105.31930992736078, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 15.71, | |
| "learning_rate": 2.1400000000000003e-06, | |
| "loss": 1.6373, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 17.14, | |
| "learning_rate": 2.3400000000000005e-06, | |
| "loss": 1.4756, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 18.57, | |
| "learning_rate": 2.5400000000000002e-06, | |
| "loss": 1.3534, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 2.7400000000000004e-06, | |
| "loss": 1.2259, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 21.43, | |
| "learning_rate": 2.9400000000000002e-06, | |
| "loss": 1.1304, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "learning_rate": 3.1400000000000004e-06, | |
| "loss": 1.0195, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 24.29, | |
| "learning_rate": 3.3400000000000006e-06, | |
| "loss": 0.9428, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 25.71, | |
| "learning_rate": 3.54e-06, | |
| "loss": 0.8721, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 27.14, | |
| "learning_rate": 3.74e-06, | |
| "loss": 0.7904, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 3.94e-06, | |
| "loss": 0.7281, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "eval_loss": 1.0742337703704834, | |
| "eval_runtime": 174.027, | |
| "eval_samples_per_second": 2.942, | |
| "eval_steps_per_second": 0.184, | |
| "eval_wer": 69.61259079903148, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 4.14e-06, | |
| "loss": 0.6704, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 31.43, | |
| "learning_rate": 4.34e-06, | |
| "loss": 0.6118, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 32.86, | |
| "learning_rate": 4.540000000000001e-06, | |
| "loss": 0.5494, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 34.29, | |
| "learning_rate": 4.74e-06, | |
| "loss": 0.5024, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 35.71, | |
| "learning_rate": 4.94e-06, | |
| "loss": 0.4511, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 37.14, | |
| "learning_rate": 5.140000000000001e-06, | |
| "loss": 0.4026, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 38.57, | |
| "learning_rate": 5.3400000000000005e-06, | |
| "loss": 0.3507, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 5.540000000000001e-06, | |
| "loss": 0.3108, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 41.43, | |
| "learning_rate": 5.74e-06, | |
| "loss": 0.2687, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "learning_rate": 5.94e-06, | |
| "loss": 0.2329, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "eval_loss": 1.1192402839660645, | |
| "eval_runtime": 175.174, | |
| "eval_samples_per_second": 2.923, | |
| "eval_steps_per_second": 0.183, | |
| "eval_wer": 67.02481840193705, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 44.29, | |
| "learning_rate": 6.1400000000000005e-06, | |
| "loss": 0.1932, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 45.71, | |
| "learning_rate": 6.34e-06, | |
| "loss": 0.1642, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 47.14, | |
| "learning_rate": 6.540000000000001e-06, | |
| "loss": 0.134, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 48.57, | |
| "learning_rate": 6.740000000000001e-06, | |
| "loss": 0.1047, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 6.9400000000000005e-06, | |
| "loss": 0.0869, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 51.43, | |
| "learning_rate": 7.14e-06, | |
| "loss": 0.0648, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 52.86, | |
| "learning_rate": 7.340000000000001e-06, | |
| "loss": 0.0517, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 54.29, | |
| "learning_rate": 7.540000000000001e-06, | |
| "loss": 0.0391, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 55.71, | |
| "learning_rate": 7.74e-06, | |
| "loss": 0.0311, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 57.14, | |
| "learning_rate": 7.94e-06, | |
| "loss": 0.0247, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 57.14, | |
| "eval_loss": 1.3494515419006348, | |
| "eval_runtime": 166.7411, | |
| "eval_samples_per_second": 3.071, | |
| "eval_steps_per_second": 0.192, | |
| "eval_wer": 66.37409200968523, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 58.57, | |
| "learning_rate": 8.14e-06, | |
| "loss": 0.0195, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "learning_rate": 8.34e-06, | |
| "loss": 0.0157, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 61.43, | |
| "learning_rate": 8.540000000000001e-06, | |
| "loss": 0.0129, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 62.86, | |
| "learning_rate": 8.740000000000001e-06, | |
| "loss": 0.0112, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 64.29, | |
| "learning_rate": 8.94e-06, | |
| "loss": 0.0097, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 65.71, | |
| "learning_rate": 9.14e-06, | |
| "loss": 0.0085, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 67.14, | |
| "learning_rate": 9.340000000000002e-06, | |
| "loss": 0.0075, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 68.57, | |
| "learning_rate": 9.54e-06, | |
| "loss": 0.0069, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "learning_rate": 9.74e-06, | |
| "loss": 0.0062, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 71.43, | |
| "learning_rate": 9.940000000000001e-06, | |
| "loss": 0.0057, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 71.43, | |
| "eval_loss": 1.5055396556854248, | |
| "eval_runtime": 189.4887, | |
| "eval_samples_per_second": 2.702, | |
| "eval_steps_per_second": 0.169, | |
| "eval_wer": 67.28964891041163, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 71.43, | |
| "step": 500, | |
| "total_flos": 1.99723386175488e+18, | |
| "train_loss": 0.8564610816463828, | |
| "train_runtime": 2065.7487, | |
| "train_samples_per_second": 15.491, | |
| "train_steps_per_second": 0.242 | |
| } | |
| ], | |
| "max_steps": 500, | |
| "num_train_epochs": 72, | |
| "total_flos": 1.99723386175488e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |