| { | |
| "best_metric": 0.5124370455741882, | |
| "best_model_checkpoint": "/home/ankur/projects/llm_test/Akshay/fine_tuning/llama_3_snippets_v3/checkpoint-2574", | |
| "epoch": 3.9992231501262383, | |
| "eval_steps": 500, | |
| "global_step": 3432, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9998057875315596, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.6301, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.9998057875315596, | |
| "eval_loss": 0.6022469997406006, | |
| "eval_runtime": 796.222, | |
| "eval_samples_per_second": 3.219, | |
| "eval_steps_per_second": 1.61, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.9996115750631192, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5104, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 1.9996115750631192, | |
| "eval_loss": 0.5411043763160706, | |
| "eval_runtime": 805.4973, | |
| "eval_samples_per_second": 3.182, | |
| "eval_steps_per_second": 1.592, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 2.9994173625946785, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3799, | |
| "step": 2574 | |
| }, | |
| { | |
| "epoch": 2.9994173625946785, | |
| "eval_loss": 0.5124370455741882, | |
| "eval_runtime": 806.6908, | |
| "eval_samples_per_second": 3.177, | |
| "eval_steps_per_second": 1.589, | |
| "step": 2574 | |
| }, | |
| { | |
| "epoch": 3.9992231501262383, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.0, | |
| "loss": 0.2803, | |
| "step": 3432 | |
| }, | |
| { | |
| "epoch": 3.9992231501262383, | |
| "eval_loss": 0.5235692858695984, | |
| "eval_runtime": 815.8051, | |
| "eval_samples_per_second": 3.142, | |
| "eval_steps_per_second": 1.571, | |
| "step": 3432 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3432, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 1.0617164129518879e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |