| { |
| "best_global_step": 102, |
| "best_metric": 0.3486737906932831, |
| "best_model_checkpoint": "/content/models/gemma_qlora_lmh/checkpoint-102", |
| "epoch": 2.0, |
| "eval_steps": 20, |
| "global_step": 102, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 2.861568535119295, |
| "epoch": 0.39408866995073893, |
| "grad_norm": 0.29296875, |
| "learning_rate": 8.137254901960784e-06, |
| "loss": 2.0227262496948244, |
| "mean_token_accuracy": 0.37890625, |
| "num_tokens": 48092.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.39408866995073893, |
| "eval_entropy": 2.772029831128962, |
| "eval_loss": 0.594829261302948, |
| "eval_mean_token_accuracy": 0.8112745098039216, |
| "eval_num_tokens": 48092.0, |
| "eval_runtime": 63.9252, |
| "eval_samples_per_second": 3.176, |
| "eval_steps_per_second": 1.596, |
| "step": 20 |
| }, |
| { |
| "entropy": 2.78899439163506, |
| "epoch": 0.7881773399014779, |
| "grad_norm": 0.1728515625, |
| "learning_rate": 6.176470588235295e-06, |
| "loss": 0.45571513175964357, |
| "mean_token_accuracy": 0.7703125, |
| "num_tokens": 98284.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.7881773399014779, |
| "eval_entropy": 2.7210274104978525, |
| "eval_loss": 0.38465243577957153, |
| "eval_mean_token_accuracy": 0.8186274509803921, |
| "eval_num_tokens": 98284.0, |
| "eval_runtime": 63.7489, |
| "eval_samples_per_second": 3.184, |
| "eval_steps_per_second": 1.6, |
| "step": 40 |
| }, |
| { |
| "entropy": 2.777331072318403, |
| "epoch": 1.1773399014778325, |
| "grad_norm": 0.1806640625, |
| "learning_rate": 4.215686274509805e-06, |
| "loss": 0.40930929183959963, |
| "mean_token_accuracy": 0.7879746835443038, |
| "num_tokens": 147466.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 1.1773399014778325, |
| "eval_entropy": 2.7073173113897737, |
| "eval_loss": 0.3590787351131439, |
| "eval_mean_token_accuracy": 0.8284313725490197, |
| "eval_num_tokens": 147466.0, |
| "eval_runtime": 63.5848, |
| "eval_samples_per_second": 3.193, |
| "eval_steps_per_second": 1.604, |
| "step": 60 |
| }, |
| { |
| "entropy": 2.758040429279208, |
| "epoch": 1.5714285714285714, |
| "grad_norm": 0.1015625, |
| "learning_rate": 2.254901960784314e-06, |
| "loss": 0.367598819732666, |
| "mean_token_accuracy": 0.81640625, |
| "num_tokens": 197917.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.5714285714285714, |
| "eval_entropy": 2.703577122267555, |
| "eval_loss": 0.3489890992641449, |
| "eval_mean_token_accuracy": 0.8455882352941176, |
| "eval_num_tokens": 197917.0, |
| "eval_runtime": 63.4414, |
| "eval_samples_per_second": 3.2, |
| "eval_steps_per_second": 1.608, |
| "step": 80 |
| }, |
| { |
| "entropy": 2.7525506377220155, |
| "epoch": 1.9655172413793105, |
| "grad_norm": 0.061767578125, |
| "learning_rate": 2.9411764705882356e-07, |
| "loss": 0.3567954540252686, |
| "mean_token_accuracy": 0.80546875, |
| "num_tokens": 247245.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.9655172413793105, |
| "eval_entropy": 2.702592772596023, |
| "eval_loss": 0.3506617546081543, |
| "eval_mean_token_accuracy": 0.821078431372549, |
| "eval_num_tokens": 247245.0, |
| "eval_runtime": 63.6354, |
| "eval_samples_per_second": 3.19, |
| "eval_steps_per_second": 1.603, |
| "step": 100 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_entropy": 2.703158517678579, |
| "eval_loss": 0.3486737906932831, |
| "eval_mean_token_accuracy": 0.8235294117647058, |
| "eval_num_tokens": 251508.0, |
| "eval_runtime": 63.7813, |
| "eval_samples_per_second": 3.183, |
| "eval_steps_per_second": 1.599, |
| "step": 102 |
| } |
| ], |
| "logging_steps": 20, |
| "max_steps": 102, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 20, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3713786885514240.0, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
