| { | |
| "best_metric": 1.08203125, | |
| "best_model_checkpoint": "/home/zhengxinyong/outputs/bloom-7b1_th_continual-pretrain_100000samples_-1vocab_original/checkpoint-25000", | |
| "epoch": 1.2926577042399172, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5461, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2551, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 1.248046875, | |
| "eval_runtime": 23.2835, | |
| "eval_samples_per_second": 66.442, | |
| "eval_steps_per_second": 8.332, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1849, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1396, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_loss": 1.1669921875, | |
| "eval_runtime": 23.3161, | |
| "eval_samples_per_second": 66.349, | |
| "eval_steps_per_second": 8.32, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 0.0001, | |
| "loss": 1.117, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.0001, | |
| "loss": 1.09, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_loss": 1.1279296875, | |
| "eval_runtime": 23.3005, | |
| "eval_samples_per_second": 66.393, | |
| "eval_steps_per_second": 8.326, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0693, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0416, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "eval_loss": 1.099609375, | |
| "eval_runtime": 23.2985, | |
| "eval_samples_per_second": 66.399, | |
| "eval_steps_per_second": 8.327, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9778, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9761, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_loss": 1.08203125, | |
| "eval_runtime": 23.3142, | |
| "eval_samples_per_second": 66.354, | |
| "eval_steps_per_second": 8.321, | |
| "step": 25000 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 2, | |
| "total_flos": 7.423686731586601e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |