| { | |
| "best_metric": 9.354, | |
| "best_model_checkpoint": "./results/10k/arn-es/en-es/checkpoint-3750", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 11.143331527709961, | |
| "learning_rate": 0.00026032, | |
| "loss": 4.7539, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 12.624526023864746, | |
| "learning_rate": 0.00022032, | |
| "loss": 4.1504, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_++": 17.872, | |
| "eval_bleu": 4.805, | |
| "eval_chrf": 18.443, | |
| "eval_gen_len": 7.819, | |
| "eval_loss": 3.621842861175537, | |
| "eval_runtime": 53.4319, | |
| "eval_samples_per_second": 23.394, | |
| "eval_sentbleu": 4.805, | |
| "eval_steps_per_second": 2.938, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 9.965263366699219, | |
| "learning_rate": 0.00018031999999999998, | |
| "loss": 3.5231, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 10.572758674621582, | |
| "learning_rate": 0.00014031999999999998, | |
| "loss": 3.0617, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 10.447027206420898, | |
| "learning_rate": 0.00010031999999999999, | |
| "loss": 2.8165, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_++": 25.267, | |
| "eval_bleu": 7.694, | |
| "eval_chrf": 26.004, | |
| "eval_gen_len": 8.928, | |
| "eval_loss": 3.0477733612060547, | |
| "eval_runtime": 54.7364, | |
| "eval_samples_per_second": 22.837, | |
| "eval_sentbleu": 7.694, | |
| "eval_steps_per_second": 2.868, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 10.377391815185547, | |
| "learning_rate": 6.048e-05, | |
| "loss": 2.0591, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 9.189433097839355, | |
| "learning_rate": 2.048e-05, | |
| "loss": 1.983, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_++": 29.133, | |
| "eval_bleu": 9.354, | |
| "eval_chrf": 30.093, | |
| "eval_gen_len": 9.145, | |
| "eval_loss": 2.864931106567383, | |
| "eval_runtime": 59.6201, | |
| "eval_samples_per_second": 20.966, | |
| "eval_sentbleu": 9.354, | |
| "eval_steps_per_second": 2.633, | |
| "step": 3750 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 114087065812992.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |