| { | |
| "best_metric": 0.39467453956604004, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_120-samples_config-2/checkpoint-27", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 66, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 21.41504669189453, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 4.4397, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 17.722919464111328, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 3.8584, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 15.595569610595703, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 3.4696, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 2.7714335918426514, | |
| "eval_runtime": 25.7676, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 12.627395629882812, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.8196, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 9.122146606445312, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.1912, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 5.698470115661621, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5154, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.9233453869819641, | |
| "eval_runtime": 25.7652, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 3.769749879837036, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.911, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.8143624067306519, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.7367, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 1.2752869129180908, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.4441, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 0.5486097931861877, | |
| "eval_runtime": 25.7657, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.5312128663063049, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.3839, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 1.1591869592666626, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4282, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9519180655479431, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.4453, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.4419020116329193, | |
| "eval_runtime": 25.7589, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.7251240015029907, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.3325, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.5680262446403503, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.2943, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.39467453956604004, | |
| "eval_runtime": 25.7651, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.895659327507019, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.2955, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 1.0607842206954956, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.2129, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.5883296132087708, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.209, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.3949660360813141, | |
| "eval_runtime": 25.76, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.47473806142807007, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.2097, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.6922392845153809, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.1934, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.7173542976379395, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.167, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.4693656265735626, | |
| "eval_runtime": 25.7635, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.7706111073493958, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.1081, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.8268452286720276, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.0544, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.9987695217132568, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.0931, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.5509907007217407, | |
| "eval_runtime": 25.7644, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.49683877825737, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.0463, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.6137292385101318, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.0271, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.627763569355011, | |
| "eval_runtime": 25.7665, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.9003904461860657, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0372, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 1.11908757686615, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.0182, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.6121003031730652, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.0121, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.7050547003746033, | |
| "eval_runtime": 25.7565, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.17396920919418335, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.0075, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.3935445249080658, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.0059, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.3881361186504364, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0101, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.7721331715583801, | |
| "eval_runtime": 25.7619, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.932, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.46985000371932983, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.0089, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.5655374526977539, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.0102, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 38.224891662597656, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.0546, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.7857134342193604, | |
| "eval_runtime": 25.7658, | |
| "eval_samples_per_second": 0.931, | |
| "eval_steps_per_second": 0.931, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 66, | |
| "total_flos": 1.0256775657396634e+17, | |
| "train_loss": 0.603098966773938, | |
| "train_runtime": 3818.1022, | |
| "train_samples_per_second": 1.152, | |
| "train_steps_per_second": 0.065 | |
| } | |
| ], | |
| "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": 1.0256775657396634e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |