| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.7408000000000001, | |
| "eval_steps": 500, | |
| "global_step": 136, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.844448880484722, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7673, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.851370146022714, | |
| "learning_rate": 1.993800990199235e-05, | |
| "loss": 0.6958, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.9773835522231158, | |
| "learning_rate": 1.944669091607919e-05, | |
| "loss": 0.6505, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.613999907750424, | |
| "learning_rate": 1.8488342952846074e-05, | |
| "loss": 0.6289, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5045322898097253, | |
| "learning_rate": 1.711034517110761e-05, | |
| "loss": 0.6127, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.4665742556940108, | |
| "learning_rate": 1.5380823531633727e-05, | |
| "loss": 0.6077, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.46191050014663815, | |
| "learning_rate": 1.3385282760863758e-05, | |
| "loss": 0.602, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.5517829769602872, | |
| "learning_rate": 1.1222379136816347e-05, | |
| "loss": 0.5795, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.4756581891233078, | |
| "learning_rate": 8.999043083759016e-06, | |
| "loss": 0.5125, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5264850481107349, | |
| "learning_rate": 6.8251927065945755e-06, | |
| "loss": 0.509, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.4493840501728769, | |
| "learning_rate": 4.808299619194251e-06, | |
| "loss": 0.5013, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.4061406187346213, | |
| "learning_rate": 3.0480757232535773e-06, | |
| "loss": 0.501, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.4224909726082761, | |
| "learning_rate": 1.6315436146506702e-06, | |
| "loss": 0.5023, | |
| "step": 130 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 156, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 17, | |
| "total_flos": 327624362295296.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |