{ "best_metric": 3.99312686920166, "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_recent/mlm_unmasking/en_mlm_new/childes_mlm_unmasking_sent_13/checkpoint-8000", "epoch": 0.29627983630539045, "eval_steps": 2000, "global_step": 8000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.07406995907634761, "eval_loss": 5.4637370109558105, "eval_runtime": 24.2102, "eval_samples_per_second": 1237.041, "eval_steps_per_second": 77.323, "step": 2000 }, { "epoch": 0.14813991815269523, "grad_norm": 4.785717487335205, "learning_rate": 4e-05, "loss": 6.1476, "step": 4000 }, { "epoch": 0.14813991815269523, "eval_loss": 4.718623161315918, "eval_runtime": 24.2123, "eval_samples_per_second": 1236.934, "eval_steps_per_second": 77.316, "step": 4000 }, { "epoch": 0.22220987722904284, "eval_loss": 4.352034091949463, "eval_runtime": 24.1788, "eval_samples_per_second": 1238.646, "eval_steps_per_second": 77.423, "step": 6000 }, { "epoch": 0.29627983630539045, "grad_norm": 7.057693004608154, "learning_rate": 8e-05, "loss": 4.4471, "step": 8000 }, { "epoch": 0.29627983630539045, "eval_loss": 3.99312686920166, "eval_runtime": 24.1035, "eval_samples_per_second": 1242.518, "eval_steps_per_second": 77.665, "step": 8000 } ], "logging_steps": 4000, "max_steps": 400000, "num_input_tokens_seen": 0, "num_train_epochs": 15, "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": 1669249105920000.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }