{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.02859951738314416, "eval_steps": 25, "global_step": 50, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005719903476628832, "grad_norm": NaN, "learning_rate": 5e-05, "loss": 0.0, "step": 1 }, { "epoch": 0.0005719903476628832, "eval_loss": NaN, "eval_runtime": 3.0953, "eval_samples_per_second": 16.154, "eval_steps_per_second": 4.2, "step": 1 }, { "epoch": 0.0011439806953257663, "grad_norm": NaN, "learning_rate": 0.0001, "loss": 0.0, "step": 2 }, { "epoch": 0.0017159710429886495, "grad_norm": NaN, "learning_rate": 9.990365154573717e-05, "loss": 0.0, "step": 3 }, { "epoch": 0.0022879613906515326, "grad_norm": NaN, "learning_rate": 9.961501876182148e-05, "loss": 0.0, "step": 4 }, { "epoch": 0.002859951738314416, "grad_norm": NaN, "learning_rate": 9.913533761814537e-05, "loss": 0.0, "step": 5 }, { "epoch": 0.003431942085977299, "grad_norm": NaN, "learning_rate": 9.846666218300807e-05, "loss": 0.0, "step": 6 }, { "epoch": 0.004003932433640182, "grad_norm": NaN, "learning_rate": 9.761185582727977e-05, "loss": 0.0, "step": 7 }, { "epoch": 0.004575922781303065, "grad_norm": NaN, "learning_rate": 9.657457896300791e-05, "loss": 0.0, "step": 8 }, { "epoch": 0.005147913128965949, "grad_norm": NaN, "learning_rate": 9.535927336897098e-05, "loss": 0.0, "step": 9 }, { "epoch": 0.005719903476628832, "grad_norm": NaN, "learning_rate": 9.397114317029975e-05, "loss": 0.0, "step": 10 }, { "epoch": 0.006291893824291715, "grad_norm": NaN, "learning_rate": 9.241613255361455e-05, "loss": 0.0, "step": 11 }, { "epoch": 0.006863884171954598, "grad_norm": NaN, "learning_rate": 9.070090031310558e-05, "loss": 0.0, "step": 12 }, { "epoch": 0.007435874519617481, "grad_norm": NaN, "learning_rate": 8.883279133655399e-05, "loss": 0.0, "step": 13 }, { "epoch": 0.008007864867280364, "grad_norm": NaN, "learning_rate": 8.681980515339464e-05, "loss": 0.0, "step": 14 }, { "epoch": 0.008579855214943248, "grad_norm": NaN, "learning_rate": 8.467056167950311e-05, "loss": 0.0, "step": 15 }, { "epoch": 0.00915184556260613, "grad_norm": NaN, "learning_rate": 8.239426430539243e-05, "loss": 0.0, "step": 16 }, { "epoch": 0.009723835910269015, "grad_norm": NaN, "learning_rate": 8.000066048588211e-05, "loss": 0.0, "step": 17 }, { "epoch": 0.010295826257931897, "grad_norm": NaN, "learning_rate": 7.75e-05, "loss": 0.0, "step": 18 }, { "epoch": 0.01086781660559478, "grad_norm": NaN, "learning_rate": 7.490299105985507e-05, "loss": 0.0, "step": 19 }, { "epoch": 0.011439806953257664, "grad_norm": NaN, "learning_rate": 7.222075445642904e-05, "loss": 0.0, "step": 20 }, { "epoch": 0.012011797300920547, "grad_norm": NaN, "learning_rate": 6.946477593864228e-05, "loss": 0.0, "step": 21 }, { "epoch": 0.01258378764858343, "grad_norm": NaN, "learning_rate": 6.664685702961344e-05, "loss": 0.0, "step": 22 }, { "epoch": 0.013155777996246313, "grad_norm": NaN, "learning_rate": 6.377906449072578e-05, "loss": 0.0, "step": 23 }, { "epoch": 0.013727768343909196, "grad_norm": NaN, "learning_rate": 6.087367864990233e-05, "loss": 0.0, "step": 24 }, { "epoch": 0.01429975869157208, "grad_norm": NaN, "learning_rate": 5.794314081535644e-05, "loss": 0.0, "step": 25 }, { "epoch": 0.01429975869157208, "eval_loss": NaN, "eval_runtime": 3.131, "eval_samples_per_second": 15.97, "eval_steps_per_second": 4.152, "step": 25 }, { "epoch": 0.014871749039234963, "grad_norm": NaN, "learning_rate": 5.500000000000001e-05, "loss": 0.0, "step": 26 }, { "epoch": 0.015443739386897847, "grad_norm": NaN, "learning_rate": 5.205685918464356e-05, "loss": 0.0, "step": 27 }, { "epoch": 0.016015729734560728, "grad_norm": NaN, "learning_rate": 4.912632135009769e-05, "loss": 0.0, "step": 28 }, { "epoch": 0.016587720082223614, "grad_norm": NaN, "learning_rate": 4.6220935509274235e-05, "loss": 0.0, "step": 29 }, { "epoch": 0.017159710429886496, "grad_norm": NaN, "learning_rate": 4.3353142970386564e-05, "loss": 0.0, "step": 30 }, { "epoch": 0.01773170077754938, "grad_norm": NaN, "learning_rate": 4.053522406135775e-05, "loss": 0.0, "step": 31 }, { "epoch": 0.01830369112521226, "grad_norm": NaN, "learning_rate": 3.777924554357096e-05, "loss": 0.0, "step": 32 }, { "epoch": 0.018875681472875144, "grad_norm": NaN, "learning_rate": 3.509700894014496e-05, "loss": 0.0, "step": 33 }, { "epoch": 0.01944767182053803, "grad_norm": NaN, "learning_rate": 3.250000000000001e-05, "loss": 0.0, "step": 34 }, { "epoch": 0.020019662168200912, "grad_norm": NaN, "learning_rate": 2.9999339514117912e-05, "loss": 0.0, "step": 35 }, { "epoch": 0.020591652515863795, "grad_norm": NaN, "learning_rate": 2.760573569460757e-05, "loss": 0.0, "step": 36 }, { "epoch": 0.021163642863526677, "grad_norm": NaN, "learning_rate": 2.53294383204969e-05, "loss": 0.0, "step": 37 }, { "epoch": 0.02173563321118956, "grad_norm": NaN, "learning_rate": 2.3180194846605367e-05, "loss": 0.0, "step": 38 }, { "epoch": 0.022307623558852446, "grad_norm": NaN, "learning_rate": 2.1167208663446025e-05, "loss": 0.0, "step": 39 }, { "epoch": 0.022879613906515328, "grad_norm": NaN, "learning_rate": 1.9299099686894423e-05, "loss": 0.0, "step": 40 }, { "epoch": 0.02345160425417821, "grad_norm": NaN, "learning_rate": 1.758386744638546e-05, "loss": 0.0, "step": 41 }, { "epoch": 0.024023594601841093, "grad_norm": NaN, "learning_rate": 1.602885682970026e-05, "loss": 0.0, "step": 42 }, { "epoch": 0.024595584949503976, "grad_norm": NaN, "learning_rate": 1.464072663102903e-05, "loss": 0.0, "step": 43 }, { "epoch": 0.02516757529716686, "grad_norm": NaN, "learning_rate": 1.3425421036992098e-05, "loss": 0.0, "step": 44 }, { "epoch": 0.025739565644829744, "grad_norm": NaN, "learning_rate": 1.2388144172720251e-05, "loss": 0.0, "step": 45 }, { "epoch": 0.026311555992492627, "grad_norm": NaN, "learning_rate": 1.1533337816991932e-05, "loss": 0.0, "step": 46 }, { "epoch": 0.02688354634015551, "grad_norm": NaN, "learning_rate": 1.0864662381854632e-05, "loss": 0.0, "step": 47 }, { "epoch": 0.02745553668781839, "grad_norm": NaN, "learning_rate": 1.0384981238178534e-05, "loss": 0.0, "step": 48 }, { "epoch": 0.028027527035481278, "grad_norm": NaN, "learning_rate": 1.0096348454262845e-05, "loss": 0.0, "step": 49 }, { "epoch": 0.02859951738314416, "grad_norm": NaN, "learning_rate": 1e-05, "loss": 0.0, "step": 50 }, { "epoch": 0.02859951738314416, "eval_loss": NaN, "eval_runtime": 3.1387, "eval_samples_per_second": 15.93, "eval_steps_per_second": 4.142, "step": 50 } ], "logging_steps": 1, "max_steps": 50, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 25, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.658021338284032e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }