| { |
| "best_global_step": 50, |
| "best_metric": 0.7305803894996643, |
| "best_model_checkpoint": "models/sft-qwen-chat-jsonl/checkpoint-50", |
| "epoch": 2.992248062015504, |
| "eval_steps": 50, |
| "global_step": 50, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.5951878294348716, |
| "epoch": 0.6201550387596899, |
| "grad_norm": 23.375, |
| "learning_rate": 3.6000000000000003e-06, |
| "loss": 2.4601, |
| "mean_token_accuracy": 0.5130733169615269, |
| "num_tokens": 94493.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 1.65908399510057, |
| "epoch": 1.1860465116279069, |
| "grad_norm": 16.25, |
| "learning_rate": 7.600000000000001e-06, |
| "loss": 2.0435, |
| "mean_token_accuracy": 0.5384233569445676, |
| "num_tokens": 179977.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 1.3006820909678936, |
| "epoch": 1.806201550387597, |
| "grad_norm": 4.8125, |
| "learning_rate": 1.16e-05, |
| "loss": 1.1905, |
| "mean_token_accuracy": 0.6729449413716793, |
| "num_tokens": 274486.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.7956327183605874, |
| "epoch": 2.3720930232558137, |
| "grad_norm": 1.90625, |
| "learning_rate": 1.5600000000000003e-05, |
| "loss": 0.7824, |
| "mean_token_accuracy": 0.7548726584813367, |
| "num_tokens": 360114.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.728295186161995, |
| "epoch": 2.992248062015504, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.9600000000000002e-05, |
| "loss": 0.7234, |
| "mean_token_accuracy": 0.7646816596388817, |
| "num_tokens": 454862.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 2.992248062015504, |
| "eval_entropy": 0.7218151768048604, |
| "eval_loss": 0.7305803894996643, |
| "eval_mean_token_accuracy": 0.7592896223068237, |
| "eval_num_tokens": 454862.0, |
| "eval_runtime": 0.8129, |
| "eval_samples_per_second": 70.118, |
| "eval_steps_per_second": 18.452, |
| "step": 50 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 51, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3824577615049728.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
