| { | |
| "best_metric": 1.2369898557662964, | |
| "best_model_checkpoint": "/users/zyong2/data/zyong2/bigscience/data/processed/024/bloom-350m_si_bitfit_100000samples_-1vocab_original-frozen/checkpoint-25000", | |
| "epoch": 0.4656230502034773, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4718, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3617, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_loss": 1.3304363489151, | |
| "eval_runtime": 2281.5601, | |
| "eval_samples_per_second": 9.533, | |
| "eval_steps_per_second": 4.767, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3157, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2935, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 1.2805551290512085, | |
| "eval_runtime": 2274.7206, | |
| "eval_samples_per_second": 9.562, | |
| "eval_steps_per_second": 4.781, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2754, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4e-05, | |
| "loss": 1.263, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_loss": 1.253080129623413, | |
| "eval_runtime": 2282.2955, | |
| "eval_samples_per_second": 9.53, | |
| "eval_steps_per_second": 4.765, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2525, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2492, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_loss": 1.2420716285705566, | |
| "eval_runtime": 2282.8423, | |
| "eval_samples_per_second": 9.528, | |
| "eval_steps_per_second": 4.764, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2439, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 0.0, | |
| "loss": 1.2429, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_loss": 1.2369898557662964, | |
| "eval_runtime": 2284.8813, | |
| "eval_samples_per_second": 9.52, | |
| "eval_steps_per_second": 4.76, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "step": 25000, | |
| "total_flos": 3.714827943936e+17, | |
| "train_loss": 1.296959794921875, | |
| "train_runtime": 61646.1041, | |
| "train_samples_per_second": 3.244, | |
| "train_steps_per_second": 0.406 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 1, | |
| "total_flos": 3.714827943936e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |