| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 9915, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.25, |
| "grad_norm": 0.6509245038032532, |
| "learning_rate": 1.9495713565305096e-05, |
| "loss": 3.3172, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 0.3449840843677521, |
| "learning_rate": 1.8991427130610186e-05, |
| "loss": 0.6868, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 0.3152226507663727, |
| "learning_rate": 1.8487140695915284e-05, |
| "loss": 0.556, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 0.40746939182281494, |
| "eval_runtime": 1.4293, |
| "eval_samples_per_second": 139.932, |
| "eval_steps_per_second": 2.799, |
| "step": 1983 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 0.2724578380584717, |
| "learning_rate": 1.7982854261220374e-05, |
| "loss": 0.4964, |
| "step": 2000 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 0.2669610381126404, |
| "learning_rate": 1.747856782652547e-05, |
| "loss": 0.4694, |
| "step": 2500 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 0.30558499693870544, |
| "learning_rate": 1.697428139183056e-05, |
| "loss": 0.4483, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 0.24562540650367737, |
| "learning_rate": 1.6469994957135657e-05, |
| "loss": 0.4312, |
| "step": 3500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_loss": 0.3607844114303589, |
| "eval_runtime": 1.4442, |
| "eval_samples_per_second": 138.488, |
| "eval_steps_per_second": 2.77, |
| "step": 3966 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 0.19550573825836182, |
| "learning_rate": 1.5965708522440747e-05, |
| "loss": 0.4217, |
| "step": 4000 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 0.20502419769763947, |
| "learning_rate": 1.546142208774584e-05, |
| "loss": 0.4102, |
| "step": 4500 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 0.20957614481449127, |
| "learning_rate": 1.4957135653050934e-05, |
| "loss": 0.4002, |
| "step": 5000 |
| }, |
| { |
| "epoch": 2.77, |
| "grad_norm": 0.251958429813385, |
| "learning_rate": 1.4452849218356026e-05, |
| "loss": 0.3971, |
| "step": 5500 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_loss": 0.34329766035079956, |
| "eval_runtime": 1.412, |
| "eval_samples_per_second": 141.644, |
| "eval_steps_per_second": 2.833, |
| "step": 5949 |
| }, |
| { |
| "epoch": 3.03, |
| "grad_norm": 0.24571850895881653, |
| "learning_rate": 1.3948562783661122e-05, |
| "loss": 0.3927, |
| "step": 6000 |
| }, |
| { |
| "epoch": 3.28, |
| "grad_norm": 0.30599531531333923, |
| "learning_rate": 1.3444276348966214e-05, |
| "loss": 0.3817, |
| "step": 6500 |
| }, |
| { |
| "epoch": 3.53, |
| "grad_norm": 0.20877033472061157, |
| "learning_rate": 1.2939989914271307e-05, |
| "loss": 0.3822, |
| "step": 7000 |
| }, |
| { |
| "epoch": 3.78, |
| "grad_norm": 0.23654815554618835, |
| "learning_rate": 1.2435703479576399e-05, |
| "loss": 0.3742, |
| "step": 7500 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_loss": 0.33198702335357666, |
| "eval_runtime": 1.3925, |
| "eval_samples_per_second": 143.629, |
| "eval_steps_per_second": 2.873, |
| "step": 7932 |
| }, |
| { |
| "epoch": 4.03, |
| "grad_norm": 0.19346973299980164, |
| "learning_rate": 1.1931417044881495e-05, |
| "loss": 0.3723, |
| "step": 8000 |
| }, |
| { |
| "epoch": 4.29, |
| "grad_norm": 0.15517400205135345, |
| "learning_rate": 1.1427130610186587e-05, |
| "loss": 0.3634, |
| "step": 8500 |
| }, |
| { |
| "epoch": 4.54, |
| "grad_norm": 0.20213666558265686, |
| "learning_rate": 1.0922844175491681e-05, |
| "loss": 0.3656, |
| "step": 9000 |
| }, |
| { |
| "epoch": 4.79, |
| "grad_norm": 0.20277400314807892, |
| "learning_rate": 1.0418557740796773e-05, |
| "loss": 0.3638, |
| "step": 9500 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_loss": 0.3258039355278015, |
| "eval_runtime": 1.2465, |
| "eval_samples_per_second": 160.451, |
| "eval_steps_per_second": 3.209, |
| "step": 9915 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 19830, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 500, |
| "total_flos": 1.3277425805166182e+17, |
| "train_batch_size": 64, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
