| { | |
| "best_metric": 0.9566597938537598, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_60-samples_config-2_full_auto/checkpoint-43", | |
| "epoch": 21.91304347826087, | |
| "eval_steps": 500, | |
| "global_step": 63, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.4215196371078491, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3941, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.398699015378952, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3458, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.3567560911178589, | |
| "eval_runtime": 4.419, | |
| "eval_samples_per_second": 2.716, | |
| "eval_steps_per_second": 2.716, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.4273262321949005, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3656, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.3314852714538574, | |
| "eval_runtime": 4.1881, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.4063573479652405, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3759, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.3449285328388214, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3043, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.2450376749038696, | |
| "eval_runtime": 4.1903, | |
| "eval_samples_per_second": 2.864, | |
| "eval_steps_per_second": 2.864, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.24114809930324554, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2137, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.1759628057479858, | |
| "eval_runtime": 4.1827, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.2410249561071396, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.1801, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.27519962191581726, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.1349, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.1266146898269653, | |
| "eval_runtime": 4.1843, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.23488768935203552, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.0892, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.0785940885543823, | |
| "eval_runtime": 4.187, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 2.866, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.24930627644062042, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.0527, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.33710455894470215, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.0063, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.0362237691879272, | |
| "eval_runtime": 4.1829, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.18661880493164062, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.9866, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.0086289644241333, | |
| "eval_runtime": 4.1811, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.16058190166950226, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.9437, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.9958365559577942, | |
| "eval_runtime": 4.1876, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 2.866, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.16996340453624725, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.9322, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.1740841567516327, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9105, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.9804616570472717, | |
| "eval_runtime": 4.186, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.14579086005687714, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.9086, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.9710781574249268, | |
| "eval_runtime": 4.1836, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.1706189066171646, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.8654, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.13660015165805817, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.884, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.9645407199859619, | |
| "eval_runtime": 4.1832, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.1543744057416916, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.852, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.9601142406463623, | |
| "eval_runtime": 4.1852, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.14513741433620453, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.8525, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.13408692181110382, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.8465, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.9574996829032898, | |
| "eval_runtime": 4.1833, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.13671351969242096, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.8197, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.9566597938537598, | |
| "eval_runtime": 4.1845, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.13350413739681244, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.8164, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.14021657407283783, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.8147, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.9570537209510803, | |
| "eval_runtime": 4.1897, | |
| "eval_samples_per_second": 2.864, | |
| "eval_steps_per_second": 2.864, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.28469404578208923, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.7741, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 0.9576375484466553, | |
| "eval_runtime": 4.1859, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.16427278518676758, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.7843, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 0.9597352147102356, | |
| "eval_runtime": 4.1861, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.1639845371246338, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.7963, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.18433023989200592, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.7714, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 0.9629867672920227, | |
| "eval_runtime": 4.1865, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 2.866, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.17088112235069275, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.7444, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 0.9685425758361816, | |
| "eval_runtime": 4.1854, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.18618661165237427, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.7527, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.18860526382923126, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.7328, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 0.9758319854736328, | |
| "eval_runtime": 4.1832, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.1849859207868576, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.7226, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 0.9808992743492126, | |
| "eval_runtime": 4.1853, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "step": 63, | |
| "total_flos": 2.784926601366733e+16, | |
| "train_loss": 0.9541767854539175, | |
| "train_runtime": 1196.6871, | |
| "train_samples_per_second": 1.922, | |
| "train_steps_per_second": 0.084 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.784926601366733e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |