| { | |
| "best_metric": NaN, | |
| "best_model_checkpoint": "/mnt/bn/qingyi-bn-lq/llama/saved_llama-7b-hf_TQA/checkpoint-200", | |
| "epoch": 0.6990715456034954, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5.1e-05, | |
| "loss": 15278261862.4, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.00010799999999999998, | |
| "loss": 155534963507.2, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.000168, | |
| "loss": 26388629094.4, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.00022799999999999999, | |
| "loss": 1740693708.8, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 353188315136.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.0002936675461741425, | |
| "loss": 15921389568.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 0.00028575197889182057, | |
| "loss": 8279447142.4, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 0.00027783641160949866, | |
| "loss": 31305360998.4, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00026992084432717674, | |
| "loss": 100943947366.4, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.0002620052770448549, | |
| "loss": 1337943859.2, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_loss": NaN, | |
| "eval_runtime": 33.6458, | |
| "eval_samples_per_second": 59.443, | |
| "eval_steps_per_second": 1.872, | |
| "step": 200 | |
| } | |
| ], | |
| "max_steps": 858, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.0791367854467318e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |