| { | |
| "best_metric": 7.183506965637207, | |
| "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/de_mlm/wikipedia_13/checkpoint-8000", | |
| "epoch": 5.862953462806889, | |
| "eval_steps": 2000, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.4657383657017222, | |
| "eval_loss": 8.087221145629883, | |
| "eval_runtime": 2.215, | |
| "eval_samples_per_second": 1035.232, | |
| "eval_steps_per_second": 65.012, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.9314767314034444, | |
| "grad_norm": 0.9454871416091919, | |
| "learning_rate": 1e-05, | |
| "loss": 8.1326, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.9314767314034444, | |
| "eval_loss": 7.397463798522949, | |
| "eval_runtime": 2.2812, | |
| "eval_samples_per_second": 1005.158, | |
| "eval_steps_per_second": 63.124, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.397215097105167, | |
| "eval_loss": 7.2718353271484375, | |
| "eval_runtime": 2.3459, | |
| "eval_samples_per_second": 977.446, | |
| "eval_steps_per_second": 61.383, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 5.862953462806889, | |
| "grad_norm": 1.4085547924041748, | |
| "learning_rate": 2e-05, | |
| "loss": 7.2846, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.862953462806889, | |
| "eval_loss": 7.183506965637207, | |
| "eval_runtime": 2.2787, | |
| "eval_samples_per_second": 1006.271, | |
| "eval_steps_per_second": 63.194, | |
| "step": 8000 | |
| } | |
| ], | |
| "logging_steps": 4000, | |
| "max_steps": 100000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 74, | |
| "save_steps": 4000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2495200998850560.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |