| { |
| "best_metric": 0.7266666666666667, |
| "best_model_checkpoint": "./llama_reward_model_1e5-bz32/checkpoint-350", |
| "epoch": 2.482269503546099, |
| "eval_steps": 50, |
| "global_step": 350, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.07092198581560284, |
| "grad_norm": 388.0, |
| "learning_rate": 1.4084507042253523e-06, |
| "loss": 2.1268, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.14184397163120568, |
| "grad_norm": 219.0, |
| "learning_rate": 2.8169014084507046e-06, |
| "loss": 0.9385, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.2127659574468085, |
| "grad_norm": 247.0, |
| "learning_rate": 4.225352112676057e-06, |
| "loss": 0.9245, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.28368794326241137, |
| "grad_norm": 76.5, |
| "learning_rate": 5.633802816901409e-06, |
| "loss": 0.7768, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.3546099290780142, |
| "grad_norm": 152.0, |
| "learning_rate": 7.042253521126761e-06, |
| "loss": 0.8706, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3546099290780142, |
| "eval_accuracy": 0.5166666666666667, |
| "eval_auc": 0.6492222222222223, |
| "eval_f1": 0.6635730858468677, |
| "eval_loss": 1.0949409008026123, |
| "eval_precision": 0.5088967971530249, |
| "eval_recall": 0.9533333333333334, |
| "eval_runtime": 21.6558, |
| "eval_samples_per_second": 13.853, |
| "eval_steps_per_second": 0.231, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.425531914893617, |
| "grad_norm": 149.0, |
| "learning_rate": 8.450704225352114e-06, |
| "loss": 0.866, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.49645390070921985, |
| "grad_norm": 112.5, |
| "learning_rate": 9.859154929577466e-06, |
| "loss": 0.7572, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.5673758865248227, |
| "grad_norm": 32.0, |
| "learning_rate": 9.995028650728335e-06, |
| "loss": 0.8987, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.6382978723404256, |
| "grad_norm": 82.0, |
| "learning_rate": 9.977856431060221e-06, |
| "loss": 0.7988, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.7092198581560284, |
| "grad_norm": 148.0, |
| "learning_rate": 9.948464112207811e-06, |
| "loss": 0.8813, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.7092198581560284, |
| "eval_accuracy": 0.6633333333333333, |
| "eval_auc": 0.7096444444444445, |
| "eval_f1": 0.6007905138339921, |
| "eval_loss": 0.6915330290794373, |
| "eval_precision": 0.7378640776699029, |
| "eval_recall": 0.5066666666666667, |
| "eval_runtime": 21.6662, |
| "eval_samples_per_second": 13.846, |
| "eval_steps_per_second": 0.231, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.7801418439716312, |
| "grad_norm": 36.75, |
| "learning_rate": 9.906923849135118e-06, |
| "loss": 0.7012, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.851063829787234, |
| "grad_norm": 26.0, |
| "learning_rate": 9.853337618695413e-06, |
| "loss": 0.6147, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.9219858156028369, |
| "grad_norm": 22.125, |
| "learning_rate": 9.78783696928909e-06, |
| "loss": 0.5603, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.9929078014184397, |
| "grad_norm": 72.5, |
| "learning_rate": 9.710582697926562e-06, |
| "loss": 0.6269, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.0638297872340425, |
| "grad_norm": 129.0, |
| "learning_rate": 9.62176445548899e-06, |
| "loss": 0.7091, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.0638297872340425, |
| "eval_accuracy": 0.6133333333333333, |
| "eval_auc": 0.7025777777777777, |
| "eval_f1": 0.5126050420168067, |
| "eval_loss": 0.7151511907577515, |
| "eval_precision": 0.6931818181818182, |
| "eval_recall": 0.4066666666666667, |
| "eval_runtime": 21.9923, |
| "eval_samples_per_second": 13.641, |
| "eval_steps_per_second": 0.227, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.1347517730496455, |
| "grad_norm": 73.0, |
| "learning_rate": 9.521600281155894e-06, |
| "loss": 0.5991, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.2056737588652482, |
| "grad_norm": 69.0, |
| "learning_rate": 9.410336067142525e-06, |
| "loss": 0.4835, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.2765957446808511, |
| "grad_norm": 27.25, |
| "learning_rate": 9.28824495506109e-06, |
| "loss": 0.4869, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.3475177304964538, |
| "grad_norm": 71.5, |
| "learning_rate": 9.155626665387625e-06, |
| "loss": 0.4671, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.4184397163120568, |
| "grad_norm": 25.875, |
| "learning_rate": 9.012806761680642e-06, |
| "loss": 0.4129, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.4184397163120568, |
| "eval_accuracy": 0.69, |
| "eval_auc": 0.7723555555555555, |
| "eval_f1": 0.7240356083086054, |
| "eval_loss": 0.6506745219230652, |
| "eval_precision": 0.6524064171122995, |
| "eval_recall": 0.8133333333333334, |
| "eval_runtime": 21.8002, |
| "eval_samples_per_second": 13.761, |
| "eval_steps_per_second": 0.229, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.4893617021276595, |
| "grad_norm": 57.25, |
| "learning_rate": 8.860135851357803e-06, |
| "loss": 0.517, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.5602836879432624, |
| "grad_norm": 30.5, |
| "learning_rate": 8.697988724992633e-06, |
| "loss": 0.4431, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.6312056737588652, |
| "grad_norm": 16.0, |
| "learning_rate": 8.526763436244184e-06, |
| "loss": 0.4721, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.702127659574468, |
| "grad_norm": 19.5, |
| "learning_rate": 8.346880324678359e-06, |
| "loss": 0.4094, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.773049645390071, |
| "grad_norm": 114.0, |
| "learning_rate": 8.158780983879737e-06, |
| "loss": 0.5608, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.773049645390071, |
| "eval_accuracy": 0.71, |
| "eval_auc": 0.7636, |
| "eval_f1": 0.7220447284345048, |
| "eval_loss": 0.6192474365234375, |
| "eval_precision": 0.6932515337423313, |
| "eval_recall": 0.7533333333333333, |
| "eval_runtime": 21.9348, |
| "eval_samples_per_second": 13.677, |
| "eval_steps_per_second": 0.228, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.8439716312056738, |
| "grad_norm": 53.25, |
| "learning_rate": 7.962927177387085e-06, |
| "loss": 0.4805, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.9148936170212765, |
| "grad_norm": 9.625, |
| "learning_rate": 7.759799705113797e-06, |
| "loss": 0.4569, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.9858156028368794, |
| "grad_norm": 23.375, |
| "learning_rate": 7.54989722303612e-06, |
| "loss": 0.5058, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.0567375886524824, |
| "grad_norm": 9.75, |
| "learning_rate": 7.33373501904665e-06, |
| "loss": 0.3455, |
| "step": 290 |
| }, |
| { |
| "epoch": 2.127659574468085, |
| "grad_norm": 31.875, |
| "learning_rate": 7.111843747978296e-06, |
| "loss": 0.3486, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.127659574468085, |
| "eval_accuracy": 0.7066666666666667, |
| "eval_auc": 0.7925333333333333, |
| "eval_f1": 0.7426900584795322, |
| "eval_loss": 0.6051149964332581, |
| "eval_precision": 0.6614583333333334, |
| "eval_recall": 0.8466666666666667, |
| "eval_runtime": 21.6917, |
| "eval_samples_per_second": 13.83, |
| "eval_steps_per_second": 0.231, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.198581560283688, |
| "grad_norm": 50.0, |
| "learning_rate": 6.884768128904038e-06, |
| "loss": 0.33, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.269503546099291, |
| "grad_norm": 29.125, |
| "learning_rate": 6.653065607910535e-06, |
| "loss": 0.356, |
| "step": 320 |
| }, |
| { |
| "epoch": 2.3404255319148937, |
| "grad_norm": 15.25, |
| "learning_rate": 6.4173049896282525e-06, |
| "loss": 0.2589, |
| "step": 330 |
| }, |
| { |
| "epoch": 2.4113475177304964, |
| "grad_norm": 15.375, |
| "learning_rate": 6.1780650408776214e-06, |
| "loss": 0.2613, |
| "step": 340 |
| }, |
| { |
| "epoch": 2.482269503546099, |
| "grad_norm": 35.75, |
| "learning_rate": 5.9359330698590835e-06, |
| "loss": 0.2617, |
| "step": 350 |
| }, |
| { |
| "epoch": 2.482269503546099, |
| "eval_accuracy": 0.7266666666666667, |
| "eval_auc": 0.7925111111111112, |
| "eval_f1": 0.7210884353741497, |
| "eval_loss": 0.6037541627883911, |
| "eval_precision": 0.7361111111111112, |
| "eval_recall": 0.7066666666666667, |
| "eval_runtime": 21.8112, |
| "eval_samples_per_second": 13.754, |
| "eval_steps_per_second": 0.229, |
| "step": 350 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 705, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 7.568021622454272e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
