| { | |
| "best_metric": 0.957007646560669, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_60-samples_config-1_full_auto/checkpoint-51", | |
| "epoch": 16.0, | |
| "eval_steps": 500, | |
| "global_step": 92, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.4200100898742676, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.3872, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.4541209638118744, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.401, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.4271780848503113, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.3456, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 1.3519130945205688, | |
| "eval_runtime": 4.4181, | |
| "eval_samples_per_second": 2.716, | |
| "eval_steps_per_second": 2.716, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.4392004907131195, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.3627, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.41168466210365295, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.3527, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.3292771875858307, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3545, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 1.2756292819976807, | |
| "eval_runtime": 4.186, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.3539218008518219, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.3163, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.24902880191802979, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2073, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.2531520426273346, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.2474, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 1.1764642000198364, | |
| "eval_runtime": 4.1843, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.2633010447025299, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.1407, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.30559298396110535, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1415, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 0.3694708049297333, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.1006, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.0924828052520752, | |
| "eval_runtime": 4.1792, | |
| "eval_samples_per_second": 2.871, | |
| "eval_steps_per_second": 2.871, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.27051129937171936, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0946, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.29818761348724365, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.0402, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.2893217206001282, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.995, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.0252578258514404, | |
| "eval_runtime": 4.1859, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.2374761402606964, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.9436, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.21178118884563446, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.9607, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.1853170096874237, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.9268, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.9853155612945557, | |
| "eval_runtime": 4.1814, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.20083829760551453, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.9079, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.23746357858181, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.9027, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.18566443026065826, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.8574, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.966564953327179, | |
| "eval_runtime": 4.1795, | |
| "eval_samples_per_second": 2.871, | |
| "eval_steps_per_second": 2.871, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.1986018568277359, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.9089, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.24835766851902008, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.8447, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.21141505241394043, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.8398, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9580891132354736, | |
| "eval_runtime": 4.1806, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.21047359704971313, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.8407, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.2132870852947235, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.8249, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.957007646560669, | |
| "eval_runtime": 4.1925, | |
| "eval_samples_per_second": 2.862, | |
| "eval_steps_per_second": 2.862, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.2190542221069336, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.8124, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 0.21587738394737244, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.7676, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.22896969318389893, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.8095, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 0.9601499438285828, | |
| "eval_runtime": 4.1827, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.23069250583648682, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.7966, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.25453636050224304, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.7623, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 0.30019015073776245, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.7215, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 0.9763060212135315, | |
| "eval_runtime": 4.1859, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.3246862292289734, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.7067, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 0.3141219913959503, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.7635, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.3520950675010681, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.6764, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9986979961395264, | |
| "eval_runtime": 4.181, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.36836376786231995, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.6839, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.373884916305542, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.6167, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 0.47338593006134033, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.6692, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.0206706523895264, | |
| "eval_runtime": 4.1863, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 2.866, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.492342084646225, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.6173, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 0.5400599241256714, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.6135, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.6018920540809631, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.5679, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.0890549421310425, | |
| "eval_runtime": 4.1819, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.26086956521739, | |
| "grad_norm": 0.6805917620658875, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.5083, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.728790283203125, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.5092, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "grad_norm": 0.6641848087310791, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.5212, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.144668459892273, | |
| "eval_runtime": 4.1842, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.7231183052062988, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.4767, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 15.652173913043478, | |
| "grad_norm": 0.9120333194732666, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.3785, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.8271422982215881, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.5149, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.2058030366897583, | |
| "eval_runtime": 4.1937, | |
| "eval_samples_per_second": 2.861, | |
| "eval_steps_per_second": 2.861, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "step": 92, | |
| "total_flos": 2.0254011609317376e+16, | |
| "train_loss": 0.8770715257395869, | |
| "train_runtime": 871.7394, | |
| "train_samples_per_second": 2.638, | |
| "train_steps_per_second": 0.287 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "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": 2.0254011609317376e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |