| { | |
| "best_metric": 26.980130911344357, | |
| "best_model_checkpoint": "./checkpoint-200", | |
| "epoch": 1.4492753623188406, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.9955, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.7842, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 7e-06, | |
| "loss": 0.533, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.5079, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.4739, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 8.875e-06, | |
| "loss": 0.4768, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 8.25e-06, | |
| "loss": 0.4344, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 7.625e-06, | |
| "loss": 0.4655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 7e-06, | |
| "loss": 0.4108, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 6.375e-06, | |
| "loss": 0.3908, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 0.5148195028305054, | |
| "eval_runtime": 2783.9171, | |
| "eval_samples_per_second": 1.186, | |
| "eval_steps_per_second": 0.074, | |
| "eval_wer": 30.706069173864208, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5.75e-06, | |
| "loss": 0.3863, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5.125e-06, | |
| "loss": 0.3902, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.3788, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 3.875e-06, | |
| "loss": 0.403, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.3781, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2.6250000000000003e-06, | |
| "loss": 0.4094, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.3732, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.3750000000000002e-06, | |
| "loss": 0.4084, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.3786, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.2500000000000002e-07, | |
| "loss": 0.3519, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_loss": 0.46850404143333435, | |
| "eval_runtime": 2556.1826, | |
| "eval_samples_per_second": 1.292, | |
| "eval_steps_per_second": 0.081, | |
| "eval_wer": 26.980130911344357, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 3.538461538461539e-06, | |
| "loss": 0.3721, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.153846153846154e-06, | |
| "loss": 0.2392, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.2854, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 2.384615384615385e-06, | |
| "loss": 0.2517, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.2506, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.6153846153846157e-06, | |
| "loss": 0.2462, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.2179, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 8.461538461538463e-07, | |
| "loss": 0.2579, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4.615384615384616e-07, | |
| "loss": 0.2446, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 7.692307692307694e-08, | |
| "loss": 0.2426, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_loss": 0.4635893404483795, | |
| "eval_runtime": 2857.1857, | |
| "eval_samples_per_second": 1.156, | |
| "eval_steps_per_second": 0.072, | |
| "eval_wer": 28.502265773267748, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "step": 300, | |
| "total_flos": 9.77943622385664e+18, | |
| "train_loss": 0.08693385044733683, | |
| "train_runtime": 3481.8761, | |
| "train_samples_per_second": 2.757, | |
| "train_steps_per_second": 0.086 | |
| } | |
| ], | |
| "max_steps": 300, | |
| "num_train_epochs": 2, | |
| "total_flos": 9.77943622385664e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |