| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.07056229327453142, | |
| "eval_steps": 5, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0035281146637265713, | |
| "grad_norm": 4.04030704498291, | |
| "learning_rate": 1e-05, | |
| "loss": 4.1831, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0035281146637265713, | |
| "eval_loss": 4.5313615798950195, | |
| "eval_runtime": 56.8526, | |
| "eval_samples_per_second": 8.408, | |
| "eval_steps_per_second": 4.204, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0070562293274531425, | |
| "grad_norm": 3.4920718669891357, | |
| "learning_rate": 2e-05, | |
| "loss": 4.2023, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010584343991179712, | |
| "grad_norm": 4.01487398147583, | |
| "learning_rate": 3e-05, | |
| "loss": 4.4356, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014112458654906285, | |
| "grad_norm": 4.010401725769043, | |
| "learning_rate": 4e-05, | |
| "loss": 4.2765, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.017640573318632856, | |
| "grad_norm": 3.706618070602417, | |
| "learning_rate": 5e-05, | |
| "loss": 4.126, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.017640573318632856, | |
| "eval_loss": 4.215136528015137, | |
| "eval_runtime": 12.1488, | |
| "eval_samples_per_second": 39.346, | |
| "eval_steps_per_second": 19.673, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021168687982359425, | |
| "grad_norm": 3.581737756729126, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9871, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.024696802646085998, | |
| "grad_norm": 3.523745536804199, | |
| "learning_rate": 7e-05, | |
| "loss": 4.2483, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02822491730981257, | |
| "grad_norm": 3.2551581859588623, | |
| "learning_rate": 8e-05, | |
| "loss": 3.7095, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03175303197353914, | |
| "grad_norm": 3.1398942470550537, | |
| "learning_rate": 9e-05, | |
| "loss": 3.5496, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03528114663726571, | |
| "grad_norm": 3.5311527252197266, | |
| "learning_rate": 0.0001, | |
| "loss": 3.3988, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03528114663726571, | |
| "eval_loss": 3.404372215270996, | |
| "eval_runtime": 20.1793, | |
| "eval_samples_per_second": 23.688, | |
| "eval_steps_per_second": 11.844, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.038809261300992284, | |
| "grad_norm": 3.467017650604248, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.9477, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04233737596471885, | |
| "grad_norm": 4.313315391540527, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.7795, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04586549062844542, | |
| "grad_norm": 5.207818508148193, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.2422, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.049393605292171995, | |
| "grad_norm": 4.348426818847656, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.1503, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05292171995589857, | |
| "grad_norm": 5.540859699249268, | |
| "learning_rate": 5e-05, | |
| "loss": 3.0576, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05292171995589857, | |
| "eval_loss": 2.8416812419891357, | |
| "eval_runtime": 21.5414, | |
| "eval_samples_per_second": 22.19, | |
| "eval_steps_per_second": 11.095, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05644983461962514, | |
| "grad_norm": 4.4173736572265625, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.9691, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.059977949283351706, | |
| "grad_norm": 4.002551078796387, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.8753, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06350606394707828, | |
| "grad_norm": 3.4553565979003906, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.7968, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06703417861080485, | |
| "grad_norm": 3.165393590927124, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.6322, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07056229327453142, | |
| "grad_norm": 3.181989908218384, | |
| "learning_rate": 0.0, | |
| "loss": 2.3748, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07056229327453142, | |
| "eval_loss": 2.685425281524658, | |
| "eval_runtime": 12.1887, | |
| "eval_samples_per_second": 39.217, | |
| "eval_steps_per_second": 19.608, | |
| "step": 20 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.52887313301504e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |