| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 21056, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 7.977410793304443, | |
| "learning_rate": 1.9867686170212766e-05, | |
| "loss": 8.4229, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 3.635096788406372, | |
| "eval_runtime": 12.1685, | |
| "eval_samples_per_second": 416.402, | |
| "eval_steps_per_second": 13.066, | |
| "step": 3008 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 12.781421661376953, | |
| "learning_rate": 1.9734707446808513e-05, | |
| "loss": 4.4356, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 13.996766090393066, | |
| "learning_rate": 1.9601728723404256e-05, | |
| "loss": 2.1407, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.1635645627975464, | |
| "eval_runtime": 12.1736, | |
| "eval_samples_per_second": 416.227, | |
| "eval_steps_per_second": 13.061, | |
| "step": 6016 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 9.880745887756348, | |
| "learning_rate": 1.9468750000000002e-05, | |
| "loss": 1.2343, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.6885451674461365, | |
| "eval_runtime": 12.1613, | |
| "eval_samples_per_second": 416.65, | |
| "eval_steps_per_second": 13.074, | |
| "step": 9024 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 9.76446533203125, | |
| "learning_rate": 1.9335837765957447e-05, | |
| "loss": 0.9126, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 10.145390510559082, | |
| "learning_rate": 1.9202859042553194e-05, | |
| "loss": 0.7534, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5545051097869873, | |
| "eval_runtime": 12.1431, | |
| "eval_samples_per_second": 417.274, | |
| "eval_steps_per_second": 13.094, | |
| "step": 12032 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 9.280902862548828, | |
| "learning_rate": 1.9069880319148937e-05, | |
| "loss": 0.6302, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.48846569657325745, | |
| "eval_runtime": 12.1559, | |
| "eval_samples_per_second": 416.836, | |
| "eval_steps_per_second": 13.08, | |
| "step": 15040 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "grad_norm": 8.525071144104004, | |
| "learning_rate": 1.8936968085106385e-05, | |
| "loss": 0.5579, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 8.482599258422852, | |
| "learning_rate": 1.880405585106383e-05, | |
| "loss": 0.4992, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.4417724311351776, | |
| "eval_runtime": 12.1719, | |
| "eval_samples_per_second": 416.286, | |
| "eval_steps_per_second": 13.063, | |
| "step": 18048 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "grad_norm": 7.430637359619141, | |
| "learning_rate": 1.8671077127659576e-05, | |
| "loss": 0.4421, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.40214258432388306, | |
| "eval_runtime": 12.1628, | |
| "eval_samples_per_second": 416.599, | |
| "eval_steps_per_second": 13.073, | |
| "step": 21056 | |
| } | |
| ], | |
| "logging_steps": 2000, | |
| "max_steps": 300800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 5.002004203009344e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |