| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.0, | |
| "eval_steps": 500, | |
| "global_step": 2679, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 42.0918083190918, | |
| "learning_rate": 1.929078014184397e-06, | |
| "loss": 10.0975, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 54.439884185791016, | |
| "learning_rate": 1.8581560283687943e-06, | |
| "loss": 5.8147, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 60.155399322509766, | |
| "learning_rate": 1.7872340425531913e-06, | |
| "loss": 4.1279, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 49.89198684692383, | |
| "learning_rate": 1.7163120567375885e-06, | |
| "loss": 2.8192, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 24.662052154541016, | |
| "learning_rate": 1.6453900709219858e-06, | |
| "loss": 2.3071, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 46.319374084472656, | |
| "learning_rate": 1.574468085106383e-06, | |
| "loss": 1.788, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 19.011852264404297, | |
| "learning_rate": 1.50354609929078e-06, | |
| "loss": 1.6626, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "grad_norm": 24.53301239013672, | |
| "learning_rate": 1.4326241134751774e-06, | |
| "loss": 1.5502, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "grad_norm": 29.92524528503418, | |
| "learning_rate": 1.3617021276595744e-06, | |
| "loss": 1.4522, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "grad_norm": 49.022037506103516, | |
| "learning_rate": 1.2907801418439716e-06, | |
| "loss": 1.3627, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 15.343302726745605, | |
| "learning_rate": 1.2198581560283688e-06, | |
| "loss": 1.2928, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.51, | |
| "grad_norm": 20.46482276916504, | |
| "learning_rate": 1.148936170212766e-06, | |
| "loss": 1.2246, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "grad_norm": 26.57284927368164, | |
| "learning_rate": 1.078014184397163e-06, | |
| "loss": 1.153, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.93, | |
| "grad_norm": 15.145112991333008, | |
| "learning_rate": 1.00709219858156e-06, | |
| "loss": 1.1536, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.64, | |
| "grad_norm": 19.93513298034668, | |
| "learning_rate": 9.361702127659575e-07, | |
| "loss": 1.0598, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.35, | |
| "grad_norm": 27.3331298828125, | |
| "learning_rate": 8.652482269503546e-07, | |
| "loss": 1.047, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 12.06, | |
| "grad_norm": 43.462093353271484, | |
| "learning_rate": 7.943262411347518e-07, | |
| "loss": 1.0587, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 12.77, | |
| "grad_norm": 36.50919723510742, | |
| "learning_rate": 7.23404255319149e-07, | |
| "loss": 1.0709, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 13.48, | |
| "grad_norm": 19.904586791992188, | |
| "learning_rate": 6.524822695035461e-07, | |
| "loss": 0.972, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 14.18, | |
| "grad_norm": 19.49388313293457, | |
| "learning_rate": 5.815602836879432e-07, | |
| "loss": 1.0102, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 14.89, | |
| "grad_norm": 16.17923927307129, | |
| "learning_rate": 5.106382978723403e-07, | |
| "loss": 0.9229, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "grad_norm": 20.12999153137207, | |
| "learning_rate": 4.397163120567376e-07, | |
| "loss": 0.9488, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 16.31, | |
| "grad_norm": 26.802228927612305, | |
| "learning_rate": 3.687943262411347e-07, | |
| "loss": 0.9173, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "grad_norm": 30.8055419921875, | |
| "learning_rate": 2.978723404255319e-07, | |
| "loss": 0.9306, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 17.73, | |
| "grad_norm": 15.404391288757324, | |
| "learning_rate": 2.2695035460992907e-07, | |
| "loss": 0.898, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 18.44, | |
| "grad_norm": 11.279256820678711, | |
| "learning_rate": 1.5602836879432623e-07, | |
| "loss": 0.9177, | |
| "step": 2600 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2820, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 7.173452904406272e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |