{ "best_metric": 6.668918132781982, "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_recent/mlm_unmasking/fr_mlm_new/wikipedia_30/checkpoint-8000", "epoch": 9.35125657510228, "eval_steps": 2000, "global_step": 8000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 2.33781414377557, "eval_loss": 6.900431156158447, "eval_runtime": 1.1096, "eval_samples_per_second": 1285.131, "eval_steps_per_second": 81.109, "step": 2000 }, { "epoch": 4.67562828755114, "grad_norm": 0.7734344005584717, "learning_rate": 4e-05, "loss": 7.3814, "step": 4000 }, { "epoch": 4.67562828755114, "eval_loss": 6.925902366638184, "eval_runtime": 1.0887, "eval_samples_per_second": 1309.827, "eval_steps_per_second": 82.668, "step": 4000 }, { "epoch": 7.01344243132671, "eval_loss": 6.909808158874512, "eval_runtime": 1.0952, "eval_samples_per_second": 1302.036, "eval_steps_per_second": 82.176, "step": 6000 }, { "epoch": 9.35125657510228, "grad_norm": 1.7065298557281494, "learning_rate": 8e-05, "loss": 6.845, "step": 8000 }, { "epoch": 9.35125657510228, "eval_loss": 6.668918132781982, "eval_runtime": 1.0886, "eval_samples_per_second": 1309.937, "eval_steps_per_second": 82.675, "step": 8000 } ], "logging_steps": 4000, "max_steps": 400000, "num_input_tokens_seen": 0, "num_train_epochs": 468, "save_steps": 4000, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 3, "early_stopping_threshold": 0.001 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1669131736842240.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }