| { | |
| "best_metric": 0.5791521668434143, | |
| "best_model_checkpoint": "/home/khalid/Documents/github_rep/bigscience/data/processed/21/bloom-560m_my_continual-pretrain_100000samples_-1vocab_original/checkpoint-25000", | |
| "epoch": 0.39669788679035706, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9e-05, | |
| "loss": 0.8854, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 8e-05, | |
| "loss": 0.7087, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_loss": 0.7005436420440674, | |
| "eval_runtime": 502.012, | |
| "eval_samples_per_second": 9.398, | |
| "eval_steps_per_second": 4.699, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 7e-05, | |
| "loss": 0.6665, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 6e-05, | |
| "loss": 0.6442, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_loss": 0.6457962393760681, | |
| "eval_runtime": 502.0206, | |
| "eval_samples_per_second": 9.398, | |
| "eval_steps_per_second": 4.699, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6229, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6083, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_loss": 0.6155888438224792, | |
| "eval_runtime": 501.1781, | |
| "eval_samples_per_second": 9.414, | |
| "eval_steps_per_second": 4.707, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5988, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5859, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.5931491255760193, | |
| "eval_runtime": 501.6202, | |
| "eval_samples_per_second": 9.406, | |
| "eval_steps_per_second": 4.703, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5767, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.0, | |
| "loss": 0.5705, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.5791521668434143, | |
| "eval_runtime": 501.9862, | |
| "eval_samples_per_second": 9.399, | |
| "eval_steps_per_second": 4.699, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "step": 25000, | |
| "total_flos": 3.714827943936e+17, | |
| "train_loss": 0.6467895458984375, | |
| "train_runtime": 73141.1233, | |
| "train_samples_per_second": 2.734, | |
| "train_steps_per_second": 0.342 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 1, | |
| "total_flos": 3.714827943936e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |