| { | |
| "best_metric": 0.6944165229797363, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-2_60-samples_config-1/checkpoint-40", | |
| "epoch": 13.91304347826087, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.7895397543907166, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0312, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.9888849258422852, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.0807, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.8398869633674622, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.0003, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "eval_loss": 1.0525766611099243, | |
| "eval_runtime": 5.3262, | |
| "eval_samples_per_second": 2.253, | |
| "eval_steps_per_second": 2.253, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.8245088458061218, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.9811, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.7923673391342163, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9591, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.7004882097244263, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9778, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "eval_loss": 0.914442777633667, | |
| "eval_runtime": 5.0249, | |
| "eval_samples_per_second": 2.388, | |
| "eval_steps_per_second": 2.388, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.6176119446754456, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.849, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.5323354005813599, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.7568, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.4606993794441223, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.8521, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "eval_loss": 0.8022434711456299, | |
| "eval_runtime": 5.0227, | |
| "eval_samples_per_second": 2.389, | |
| "eval_steps_per_second": 2.389, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.130434782608696, | |
| "grad_norm": 0.44734710454940796, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.7085, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.491365522146225, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6665, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "grad_norm": 0.45671552419662476, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.7041, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.7458145022392273, | |
| "eval_runtime": 5.0187, | |
| "eval_samples_per_second": 2.391, | |
| "eval_steps_per_second": 2.391, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.38127702474594116, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.6834, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.521739130434782, | |
| "grad_norm": 0.40372902154922485, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.6341, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.4290044605731964, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.6478, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.7177431583404541, | |
| "eval_runtime": 5.0212, | |
| "eval_samples_per_second": 2.39, | |
| "eval_steps_per_second": 2.39, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.217391304347826, | |
| "grad_norm": 0.35873129963874817, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.5597, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.36796948313713074, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.5757, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "grad_norm": 0.3624632954597473, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.5764, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.6999881267547607, | |
| "eval_runtime": 5.019, | |
| "eval_samples_per_second": 2.391, | |
| "eval_steps_per_second": 2.391, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.35226380825042725, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.4908, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.608695652173913, | |
| "grad_norm": 0.41741761565208435, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.5364, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.41319116950035095, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.4782, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.6944165229797363, | |
| "eval_runtime": 5.0169, | |
| "eval_samples_per_second": 2.392, | |
| "eval_steps_per_second": 2.392, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.304347826086957, | |
| "grad_norm": 0.36885157227516174, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.4893, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.3935936689376831, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.4454, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.43766430020332336, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.4344, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.6962693333625793, | |
| "eval_runtime": 5.0166, | |
| "eval_samples_per_second": 2.392, | |
| "eval_steps_per_second": 2.392, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.4267623722553253, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.4137, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.5441224575042725, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.3571, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.869565217391305, | |
| "eval_loss": 0.7275979518890381, | |
| "eval_runtime": 5.016, | |
| "eval_samples_per_second": 2.392, | |
| "eval_steps_per_second": 2.392, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.5679859519004822, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.3658, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.391304347826088, | |
| "grad_norm": 0.5481728911399841, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.3009, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.6838057041168213, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.2965, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 9.91304347826087, | |
| "eval_loss": 0.8026370406150818, | |
| "eval_runtime": 5.0264, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 10.08695652173913, | |
| "grad_norm": 0.660568356513977, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.2925, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.6920638084411621, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.2164, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "grad_norm": 0.7863826751708984, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.2046, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 10.956521739130435, | |
| "eval_loss": 0.9340725541114807, | |
| "eval_runtime": 5.0224, | |
| "eval_samples_per_second": 2.389, | |
| "eval_steps_per_second": 2.389, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.7646209001541138, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.182, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 11.478260869565217, | |
| "grad_norm": 1.089479923248291, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.1571, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.8993852138519287, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.119, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.0342997312545776, | |
| "eval_runtime": 5.028, | |
| "eval_samples_per_second": 2.387, | |
| "eval_steps_per_second": 2.387, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 12.173913043478262, | |
| "grad_norm": 0.6343057751655579, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.091, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.8913945555686951, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.0739, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "grad_norm": 1.002933144569397, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.064, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.2800191640853882, | |
| "eval_runtime": 5.0216, | |
| "eval_samples_per_second": 2.39, | |
| "eval_steps_per_second": 2.39, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.5404753684997559, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0483, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 13.565217391304348, | |
| "grad_norm": 0.8066350221633911, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.0314, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.749337911605835, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0358, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.414562702178955, | |
| "eval_runtime": 5.0167, | |
| "eval_samples_per_second": 2.392, | |
| "eval_steps_per_second": 2.392, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "step": 80, | |
| "total_flos": 2.1337014435577856e+16, | |
| "train_loss": 0.48295602201251314, | |
| "train_runtime": 925.7582, | |
| "train_samples_per_second": 2.484, | |
| "train_steps_per_second": 0.27 | |
| } | |
| ], | |
| "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": 2.1337014435577856e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |