| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.25, | |
| "eval_steps": 20, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2199, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.0002, | |
| "loss": 1.5673, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.2322771549224854, | |
| "eval_runtime": 150.2452, | |
| "eval_samples_per_second": 0.532, | |
| "eval_steps_per_second": 0.067, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1048, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9257, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.8084919452667236, | |
| "eval_runtime": 149.4891, | |
| "eval_samples_per_second": 0.535, | |
| "eval_steps_per_second": 0.067, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8219, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.0002, | |
| "loss": 0.745, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.6932044625282288, | |
| "eval_runtime": 149.2813, | |
| "eval_samples_per_second": 0.536, | |
| "eval_steps_per_second": 0.067, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7321, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.654, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6394110321998596, | |
| "eval_runtime": 149.674, | |
| "eval_samples_per_second": 0.534, | |
| "eval_steps_per_second": 0.067, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6249, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5969, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.606323778629303, | |
| "eval_runtime": 149.8781, | |
| "eval_samples_per_second": 0.534, | |
| "eval_steps_per_second": 0.067, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 8771079870873600.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |