| { | |
| "best_metric": 0.7879278063774109, | |
| "best_model_checkpoint": "/home/khalid/Documents/github_rep/bigscience/data/processed/35/bloom-560m_my_sft_100000samples_-1vocab_original-frozen/checkpoint-25000", | |
| "epoch": 0.39631741854686253, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 9e-05, | |
| "loss": 1.0457, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8994, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_loss": 0.8788017630577087, | |
| "eval_runtime": 2668.4849, | |
| "eval_samples_per_second": 9.399, | |
| "eval_steps_per_second": 4.7, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8556, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 6e-05, | |
| "loss": 0.8299, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_loss": 0.8276682496070862, | |
| "eval_runtime": 2661.8009, | |
| "eval_samples_per_second": 9.423, | |
| "eval_steps_per_second": 4.711, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.815, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.8019, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_loss": 0.804674506187439, | |
| "eval_runtime": 2658.9167, | |
| "eval_samples_per_second": 9.433, | |
| "eval_steps_per_second": 4.717, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7965, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7875, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.7925558686256409, | |
| "eval_runtime": 2660.1365, | |
| "eval_samples_per_second": 9.428, | |
| "eval_steps_per_second": 4.714, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7866, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "l1_reg_loss": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.7835, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.7879278063774109, | |
| "eval_runtime": 2660.4205, | |
| "eval_samples_per_second": 9.427, | |
| "eval_steps_per_second": 4.714, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "step": 25000, | |
| "total_flos": 3.714827943936e+17, | |
| "train_loss": 0.8401528125, | |
| "train_runtime": 72520.074, | |
| "train_samples_per_second": 2.758, | |
| "train_steps_per_second": 0.345 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 1, | |
| "total_flos": 3.714827943936e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |