immortalPi/stack_exc_binary_instr_lm_head
012
1{2 "best_global_step": 100,3 "best_metric": 0.28937000036239624,4 "best_model_checkpoint": "/content/models/gemma_qlora_lmh_inst/checkpoint-100",5 "epoch": 1.9607843137254903,6 "eval_steps": 20,7 "global_step": 100,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "entropy": 2.1317927479743957,14 "epoch": 0.39215686274509803,15 "grad_norm": 4.763621807098389,16 "learning_rate": 8.137254901960784e-06,17 "loss": 0.7536,18 "mean_token_accuracy": 0.76171875,19 "num_tokens": 46761.0,20 "step": 2021 },22 {23 "epoch": 0.39215686274509803,24 "eval_entropy": 2.128274752543523,25 "eval_loss": 0.34577950835227966,26 "eval_mean_token_accuracy": 0.7480332163664011,27 "eval_num_tokens": 46761.0,28 "eval_runtime": 2.4531,29 "eval_samples_per_second": 82.754,30 "eval_steps_per_second": 5.3,31 "step": 2032 },33 {34 "entropy": 2.1670219361782075,35 "epoch": 0.7843137254901961,36 "grad_norm": 5.078281402587891,37 "learning_rate": 6.176470588235295e-06,38 "loss": 0.3072,39 "mean_token_accuracy": 0.8203125,40 "num_tokens": 95614.0,41 "step": 4042 },43 {44 "epoch": 0.7843137254901961,45 "eval_entropy": 2.1048440016233005,46 "eval_loss": 0.35016557574272156,47 "eval_mean_token_accuracy": 0.8236451057287363,48 "eval_num_tokens": 95614.0,49 "eval_runtime": 2.4636,50 "eval_samples_per_second": 82.4,51 "eval_steps_per_second": 5.277,52 "step": 4053 },54 {55 "entropy": 2.1607252150774,56 "epoch": 1.1764705882352942,57 "grad_norm": 10.205085754394531,58 "learning_rate": 4.215686274509805e-06,59 "loss": 0.266,60 "mean_token_accuracy": 0.86328125,61 "num_tokens": 143455.0,62 "step": 6063 },64 {65 "epoch": 1.1764705882352942,66 "eval_entropy": 2.12537236397083,67 "eval_loss": 0.3281736969947815,68 "eval_mean_token_accuracy": 0.835664336497967,69 "eval_num_tokens": 143455.0,70 "eval_runtime": 2.4452,71 "eval_samples_per_second": 83.019,72 "eval_steps_per_second": 5.317,73 "step": 6074 },75 {76 "entropy": 2.1871365696191787,77 "epoch": 1.5686274509803921,78 "grad_norm": 6.319624423980713,79 "learning_rate": 2.254901960784314e-06,80 "loss": 0.2363,81 "mean_token_accuracy": 0.88125,82 "num_tokens": 192572.0,83 "step": 8084 },85 {86 "epoch": 1.5686274509803921,87 "eval_entropy": 2.1334355702767005,88 "eval_loss": 0.29352903366088867,89 "eval_mean_token_accuracy": 0.8465909086740934,90 "eval_num_tokens": 192572.0,91 "eval_runtime": 2.4718,92 "eval_samples_per_second": 82.128,93 "eval_steps_per_second": 5.259,94 "step": 8095 },96 {97 "entropy": 2.1668645977973937,98 "epoch": 1.9607843137254903,99 "grad_norm": 3.9370977878570557,100 "learning_rate": 2.9411764705882356e-07,101 "loss": 0.2172,102 "mean_token_accuracy": 0.8984375,103 "num_tokens": 240555.0,104 "step": 100105 },106 {107 "epoch": 1.9607843137254903,108 "eval_entropy": 2.129720770395719,109 "eval_loss": 0.28937000036239624,110 "eval_mean_token_accuracy": 0.8597027980364286,111 "eval_num_tokens": 240555.0,112 "eval_runtime": 2.4499,113 "eval_samples_per_second": 82.861,114 "eval_steps_per_second": 5.306,115 "step": 100116 }117 ],118 "logging_steps": 20,119 "max_steps": 102,120 "num_input_tokens_seen": 0,121 "num_train_epochs": 2,122 "save_steps": 20,123 "stateful_callbacks": {124 "TrainerControl": {125 "args": {126 "should_epoch_stop": false,127 "should_evaluate": false,128 "should_log": false,129 "should_save": true,130 "should_training_stop": false131 },132 "attributes": {}133 }134 },135 "total_flos": 6060538554167808.0,136 "train_batch_size": 16,137 "trial_name": null,138 "trial_params": null139}140 