| { | |
| "best_metric": 1.3267595767974854, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-1_120-samples_config-2/checkpoint-38", | |
| "epoch": 14.0, | |
| "eval_steps": 500, | |
| "global_step": 77, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 2.5245699882507324, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.5052, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.7616875171661377, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.4313, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 2.373145341873169, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.4033, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 2.363821029663086, | |
| "eval_runtime": 2.2328, | |
| "eval_samples_per_second": 10.749, | |
| "eval_steps_per_second": 10.749, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 2.6781208515167236, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.3471, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 2.3583812713623047, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.3318, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 2.3059582710266113, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1644, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.9085012674331665, | |
| "eval_runtime": 2.2159, | |
| "eval_samples_per_second": 10.831, | |
| "eval_steps_per_second": 10.831, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 1.878598690032959, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.9701, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.6238771677017212, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.7284, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 1.4010099172592163, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.5867, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 1.6096315383911133, | |
| "eval_runtime": 2.2189, | |
| "eval_samples_per_second": 10.816, | |
| "eval_steps_per_second": 10.816, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.9943261742591858, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.5703, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 1.0258581638336182, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4806, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9659519791603088, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.3937, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.4499346017837524, | |
| "eval_runtime": 2.191, | |
| "eval_samples_per_second": 10.954, | |
| "eval_steps_per_second": 10.954, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.8161854147911072, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.3292, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.7501208782196045, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.2999, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 1.3671011924743652, | |
| "eval_runtime": 2.2834, | |
| "eval_samples_per_second": 10.511, | |
| "eval_steps_per_second": 10.511, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.7938846349716187, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 1.2026, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.758736252784729, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.1153, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.8454522490501404, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 1.1454, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.329568862915039, | |
| "eval_runtime": 2.1499, | |
| "eval_samples_per_second": 11.163, | |
| "eval_steps_per_second": 11.163, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.806365966796875, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 1.0849, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.8713198900222778, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.9951, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.9592146277427673, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.9351, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 1.3267595767974854, | |
| "eval_runtime": 2.298, | |
| "eval_samples_per_second": 10.444, | |
| "eval_steps_per_second": 10.444, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.8650667667388916, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.8645, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.9249936938285828, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.8049, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.106773853302002, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.8108, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.355177879333496, | |
| "eval_runtime": 2.2733, | |
| "eval_samples_per_second": 10.557, | |
| "eval_steps_per_second": 10.557, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 1.0761725902557373, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.7095, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 1.1069787740707397, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.6191, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 1.4658174514770508, | |
| "eval_runtime": 2.2892, | |
| "eval_samples_per_second": 10.484, | |
| "eval_steps_per_second": 10.484, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 1.5594687461853027, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.6243, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 1.2610054016113281, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.5089, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 1.41365647315979, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.4271, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.7157014608383179, | |
| "eval_runtime": 2.2343, | |
| "eval_samples_per_second": 10.741, | |
| "eval_steps_per_second": 10.741, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 1.3679523468017578, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.4122, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 1.6108626127243042, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.3251, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 1.9596925973892212, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.3163, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 2.0603339672088623, | |
| "eval_runtime": 2.2678, | |
| "eval_samples_per_second": 10.583, | |
| "eval_steps_per_second": 10.583, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 1.2331981658935547, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.1779, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 1.564605474472046, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.1847, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 2.0967092514038086, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.1817, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 2.5517404079437256, | |
| "eval_runtime": 2.2226, | |
| "eval_samples_per_second": 10.798, | |
| "eval_steps_per_second": 10.798, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 1.0268821716308594, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.0953, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 1.8212908506393433, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.104, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 2.906930923461914, | |
| "eval_runtime": 2.268, | |
| "eval_samples_per_second": 10.582, | |
| "eval_steps_per_second": 10.582, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 1.7785996198654175, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.076, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 1.1120202541351318, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.0483, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 1.672891616821289, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0403, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 3.1679420471191406, | |
| "eval_runtime": 2.2313, | |
| "eval_samples_per_second": 10.756, | |
| "eval_steps_per_second": 10.756, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "step": 77, | |
| "total_flos": 4919776673529856.0, | |
| "train_loss": 0.9845529866295976, | |
| "train_runtime": 544.3113, | |
| "train_samples_per_second": 8.084, | |
| "train_steps_per_second": 0.459 | |
| } | |
| ], | |
| "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": 4919776673529856.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |