| { | |
| "best_metric": 0.9456735253334045, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-1_60-samples_config-1_full_auto/checkpoint-46", | |
| "epoch": 14.956521739130435, | |
| "eval_steps": 500, | |
| "global_step": 86, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 1.3397960662841797, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.2159, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.3249859809875488, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.3977, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.2380784749984741, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.3284, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 2.1983325481414795, | |
| "eval_runtime": 1.0883, | |
| "eval_samples_per_second": 11.027, | |
| "eval_steps_per_second": 11.027, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 1.2494009733200073, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.3316, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 1.3510078191757202, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.2415, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 1.1784576177597046, | |
| "learning_rate": 4e-05, | |
| "loss": 2.161, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 1.9086942672729492, | |
| "eval_runtime": 1.0867, | |
| "eval_samples_per_second": 11.042, | |
| "eval_steps_per_second": 11.042, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.1469122171401978, | |
| "learning_rate": 4.8e-05, | |
| "loss": 2.0859, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.9704182744026184, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.8599, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.857855498790741, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.6845, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 1.4876232147216797, | |
| "eval_runtime": 1.0835, | |
| "eval_samples_per_second": 11.075, | |
| "eval_steps_per_second": 11.075, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.8267468810081482, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.601, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.969063937664032, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4118, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 1.0307605266571045, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.2931, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.1763328313827515, | |
| "eval_runtime": 1.0558, | |
| "eval_samples_per_second": 11.365, | |
| "eval_steps_per_second": 11.365, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.9416760206222534, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.2083, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.7075971961021423, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.106, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.6510626077651978, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.989, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.0414292812347412, | |
| "eval_runtime": 1.0589, | |
| "eval_samples_per_second": 11.332, | |
| "eval_steps_per_second": 11.332, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.5567235946655273, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.9363, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.6300645470619202, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.9645, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.5417482852935791, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.8809, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.9824466705322266, | |
| "eval_runtime": 1.0598, | |
| "eval_samples_per_second": 11.323, | |
| "eval_steps_per_second": 11.323, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.5442620515823364, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.8284, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.5363752245903015, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.8278, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.5813398957252502, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.7819, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.9495639801025391, | |
| "eval_runtime": 1.0603, | |
| "eval_samples_per_second": 11.317, | |
| "eval_steps_per_second": 11.317, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.5982780456542969, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.7836, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.5553918480873108, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.6956, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6128191351890564, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.7199, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9456735253334045, | |
| "eval_runtime": 1.0675, | |
| "eval_samples_per_second": 11.241, | |
| "eval_steps_per_second": 11.241, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.6114911437034607, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.619, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.7002654075622559, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.654, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.9812335968017578, | |
| "eval_runtime": 1.0883, | |
| "eval_samples_per_second": 11.026, | |
| "eval_steps_per_second": 11.026, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.7416772246360779, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.6269, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 0.7041908502578735, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.5275, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.8489178419113159, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.5351, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 1.0528734922409058, | |
| "eval_runtime": 1.0871, | |
| "eval_samples_per_second": 11.038, | |
| "eval_steps_per_second": 11.038, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.8118213415145874, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.5266, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.824499785900116, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.4026, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 0.9984014630317688, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.4639, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 1.1717332601547241, | |
| "eval_runtime": 1.085, | |
| "eval_samples_per_second": 11.06, | |
| "eval_steps_per_second": 11.06, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.9029918909072876, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.385, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 0.90236896276474, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.308, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 1.171663522720337, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.3218, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.4098140001296997, | |
| "eval_runtime": 1.0809, | |
| "eval_samples_per_second": 11.102, | |
| "eval_steps_per_second": 11.102, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.9012685418128967, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.2679, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.9381864666938782, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.2418, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 1.1532056331634521, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.2169, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.5839310884475708, | |
| "eval_runtime": 1.0854, | |
| "eval_samples_per_second": 11.056, | |
| "eval_steps_per_second": 11.056, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 1.0229898691177368, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.1895, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 0.8955756425857544, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.1564, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 1.2189979553222656, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.1623, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.813218593597412, | |
| "eval_runtime": 1.1052, | |
| "eval_samples_per_second": 10.857, | |
| "eval_steps_per_second": 10.857, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.26086956521739, | |
| "grad_norm": 1.292395830154419, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.1342, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 1.0407696962356567, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.1234, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "grad_norm": 1.2466880083084106, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.1304, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.9696978330612183, | |
| "eval_runtime": 1.085, | |
| "eval_samples_per_second": 11.06, | |
| "eval_steps_per_second": 11.06, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "step": 86, | |
| "total_flos": 2945762073772032.0, | |
| "train_loss": 0.9078863977866117, | |
| "train_runtime": 312.4754, | |
| "train_samples_per_second": 7.361, | |
| "train_steps_per_second": 0.8 | |
| } | |
| ], | |
| "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": 2945762073772032.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |