| { | |
| "best_metric": 0.43870946764945984, | |
| "best_model_checkpoint": "./mistral/06-03-24-Weni-pipeline_test_Zeroshot-2_max_steps-3224_batch_64_2024-03-06_ppid_7/checkpoint-200", | |
| "epoch": 0.49597024178549287, | |
| "eval_steps": 100, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.8498313426971436, | |
| "learning_rate": 1.1801242236024846e-05, | |
| "loss": 1.9347, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 8.336210250854492, | |
| "learning_rate": 2.4223602484472053e-05, | |
| "loss": 1.6596, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.0044610500335693, | |
| "learning_rate": 3.6645962732919256e-05, | |
| "loss": 1.3842, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.6825340390205383, | |
| "learning_rate": 4.906832298136646e-05, | |
| "loss": 0.9527, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.24993008375167847, | |
| "learning_rate": 6.149068322981368e-05, | |
| "loss": 0.5546, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.49679049849510193, | |
| "eval_runtime": 179.6977, | |
| "eval_samples_per_second": 15.955, | |
| "eval_steps_per_second": 1.002, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.2225101739168167, | |
| "learning_rate": 7.391304347826086e-05, | |
| "loss": 0.4941, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.20095179975032806, | |
| "learning_rate": 8.633540372670808e-05, | |
| "loss": 0.467, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.25985196232795715, | |
| "learning_rate": 9.875776397515528e-05, | |
| "loss": 0.4542, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.22290097177028656, | |
| "learning_rate": 0.00011118012422360249, | |
| "loss": 0.4569, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.2086612433195114, | |
| "learning_rate": 0.0001236024844720497, | |
| "loss": 0.4381, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.43870946764945984, | |
| "eval_runtime": 179.4882, | |
| "eval_samples_per_second": 15.973, | |
| "eval_steps_per_second": 1.003, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 3224, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 100, | |
| "total_flos": 4.0748291337905766e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |