| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3415, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4842, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4016, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3892, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4543589743589744, | |
| "eval_loss": 3.4055869579315186, | |
| "eval_runtime": 4.1688, | |
| "eval_samples_per_second": 119.938, | |
| "eval_steps_per_second": 15.112, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 0.09711501224106496, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.054, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3608, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3456, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3499, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4576923076923077, | |
| "eval_loss": 3.453113317489624, | |
| "eval_runtime": 4.4031, | |
| "eval_samples_per_second": 113.556, | |
| "eval_steps_per_second": 14.308, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 0.1225593574040313, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.056, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3238, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2486, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2452, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2427, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.4584102564102564, | |
| "eval_loss": 3.671137809753418, | |
| "eval_runtime": 4.4094, | |
| "eval_samples_per_second": 113.393, | |
| "eval_steps_per_second": 14.288, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 0.13165889774889974, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.056, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1507, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1218, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1231, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.45697435897435895, | |
| "eval_loss": 3.7999510765075684, | |
| "eval_runtime": 4.0886, | |
| "eval_samples_per_second": 122.29, | |
| "eval_steps_per_second": 15.409, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 0.15413978034957934, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.054, | |
| "step": 1366 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0823, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 0.0001, | |
| "loss": 0.997, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0024, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 0.0001, | |
| "loss": 0.995, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.4552307692307692, | |
| "eval_loss": 3.953216314315796, | |
| "eval_runtime": 4.9824, | |
| "eval_samples_per_second": 100.354, | |
| "eval_steps_per_second": 12.645, | |
| "step": 1707 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 0.1611344512465791, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.046, | |
| "step": 1707 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8495, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 0.0001, | |
| "loss": 0.854, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8693, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.45261538461538464, | |
| "eval_loss": 4.0766072273254395, | |
| "eval_runtime": 4.8095, | |
| "eval_samples_per_second": 103.961, | |
| "eval_steps_per_second": 13.099, | |
| "step": 2049 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 0.1489904300924264, | |
| "eval_exact_match": 0.0, | |
| "eval_prefix_exact_match": 0.044, | |
| "step": 2049 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7917, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7061, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7302, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.4501025641025641, | |
| "eval_loss": 4.171727180480957, | |
| "eval_runtime": 4.2169, | |
| "eval_samples_per_second": 118.57, | |
| "eval_steps_per_second": 14.94, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7227, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "learning_rate": 0.0001, | |
| "loss": 0.573, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6036, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6033, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.448, | |
| "eval_loss": 4.277780055999756, | |
| "eval_runtime": 4.0947, | |
| "eval_samples_per_second": 122.11, | |
| "eval_steps_per_second": 15.386, | |
| "step": 2732 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 2732 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4945, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4718, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 8.78, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4825, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.4462051282051282, | |
| "eval_loss": 4.34151554107666, | |
| "eval_runtime": 4.8113, | |
| "eval_samples_per_second": 103.921, | |
| "eval_steps_per_second": 13.094, | |
| "step": 3073 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "step": 3073 | |
| }, | |
| { | |
| "epoch": 9.08, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4538, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 9.37, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3609, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 9.66, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3769, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 9.96, | |
| "learning_rate": 0.0001, | |
| "loss": 0.387, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.4463076923076923, | |
| "eval_loss": 4.413115978240967, | |
| "eval_runtime": 4.2874, | |
| "eval_samples_per_second": 116.62, | |
| "eval_steps_per_second": 14.694, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 3415 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 17050, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 6.763068176298148e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |