| { | |
| "best_metric": 0.7109387516975403, | |
| "best_model_checkpoint": "../KevinKibe/nllb-200-distilled-1.3B-finetuned-finetuned/checkpoint-800", | |
| "epoch": 1000.0, | |
| "eval_steps": 100, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 15.290511131286621, | |
| "learning_rate": 1.8060000000000003e-05, | |
| "loss": 6.4789, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_gen_len": 32.5, | |
| "eval_loss": 5.883636951446533, | |
| "eval_rouge": 0.082, | |
| "eval_runtime": 2.1783, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.459, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "grad_norm": 15.802659034729004, | |
| "learning_rate": 1.6060000000000002e-05, | |
| "loss": 4.8961, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "eval_gen_len": 29.0, | |
| "eval_loss": 4.802582263946533, | |
| "eval_rouge": 0.1552, | |
| "eval_runtime": 2.0128, | |
| "eval_samples_per_second": 0.994, | |
| "eval_steps_per_second": 0.497, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "grad_norm": 15.621439933776855, | |
| "learning_rate": 1.4060000000000001e-05, | |
| "loss": 3.636, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "eval_gen_len": 63.0, | |
| "eval_loss": 3.5860304832458496, | |
| "eval_rouge": 0.0292, | |
| "eval_runtime": 5.5559, | |
| "eval_samples_per_second": 0.36, | |
| "eval_steps_per_second": 0.18, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 400.0, | |
| "grad_norm": 14.424906730651855, | |
| "learning_rate": 1.2060000000000001e-05, | |
| "loss": 2.4528, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 400.0, | |
| "eval_gen_len": 28.0, | |
| "eval_loss": 2.453127384185791, | |
| "eval_rouge": 0.2018, | |
| "eval_runtime": 1.6344, | |
| "eval_samples_per_second": 1.224, | |
| "eval_steps_per_second": 0.612, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 500.0, | |
| "grad_norm": 11.111770629882812, | |
| "learning_rate": 1.006e-05, | |
| "loss": 1.4112, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 500.0, | |
| "eval_gen_len": 70.0, | |
| "eval_loss": 1.6681925058364868, | |
| "eval_rouge": 0.1899, | |
| "eval_runtime": 6.0247, | |
| "eval_samples_per_second": 0.332, | |
| "eval_steps_per_second": 0.166, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 600.0, | |
| "grad_norm": 6.4994120597839355, | |
| "learning_rate": 8.06e-06, | |
| "loss": 0.6738, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 600.0, | |
| "eval_gen_len": 68.0, | |
| "eval_loss": 1.1857625246047974, | |
| "eval_rouge": 0.1907, | |
| "eval_runtime": 6.0626, | |
| "eval_samples_per_second": 0.33, | |
| "eval_steps_per_second": 0.165, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 700.0, | |
| "grad_norm": 3.2670979499816895, | |
| "learning_rate": 6.0600000000000004e-06, | |
| "loss": 0.2921, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 700.0, | |
| "eval_gen_len": 28.0, | |
| "eval_loss": 0.8545865416526794, | |
| "eval_rouge": 0.2776, | |
| "eval_runtime": 1.6454, | |
| "eval_samples_per_second": 1.215, | |
| "eval_steps_per_second": 0.608, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 800.0, | |
| "grad_norm": 1.7595362663269043, | |
| "learning_rate": 4.060000000000001e-06, | |
| "loss": 0.1361, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 800.0, | |
| "eval_gen_len": 59.5, | |
| "eval_loss": 0.7109387516975403, | |
| "eval_rouge": 0.3649, | |
| "eval_runtime": 10.8248, | |
| "eval_samples_per_second": 0.185, | |
| "eval_steps_per_second": 0.092, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 900.0, | |
| "grad_norm": 1.1663947105407715, | |
| "learning_rate": 2.06e-06, | |
| "loss": 0.0764, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 900.0, | |
| "eval_gen_len": 54.0, | |
| "eval_loss": 0.7293275594711304, | |
| "eval_rouge": 0.4568, | |
| "eval_runtime": 9.9208, | |
| "eval_samples_per_second": 0.202, | |
| "eval_steps_per_second": 0.101, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1000.0, | |
| "grad_norm": 0.9859239459037781, | |
| "learning_rate": 6.000000000000001e-08, | |
| "loss": 0.0559, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1000.0, | |
| "eval_gen_len": 49.0, | |
| "eval_loss": 0.7133963704109192, | |
| "eval_rouge": 0.467, | |
| "eval_runtime": 9.5903, | |
| "eval_samples_per_second": 0.209, | |
| "eval_steps_per_second": 0.104, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1000, | |
| "save_steps": 100, | |
| "total_flos": 2500287332352000.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |