| { | |
| "best_metric": 0.49241259694099426, | |
| "best_model_checkpoint": "./vit-dropout-0.4/checkpoint-2889", | |
| "epoch": 19.0, | |
| "eval_steps": 500, | |
| "global_step": 6099, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 31.014915466308594, | |
| "learning_rate": 2.6004922067268256e-05, | |
| "loss": 1.6459, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7291955617198336, | |
| "eval_f1": 0.6910099536320229, | |
| "eval_loss": 0.8160464763641357, | |
| "eval_precision": 0.7119244517951253, | |
| "eval_recall": 0.7291955617198336, | |
| "eval_runtime": 23.8108, | |
| "eval_samples_per_second": 121.122, | |
| "eval_steps_per_second": 15.161, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 51.539710998535156, | |
| "learning_rate": 5.233798195242002e-05, | |
| "loss": 1.0644, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7513869625520111, | |
| "eval_f1": 0.7575064499227677, | |
| "eval_loss": 0.6323224306106567, | |
| "eval_precision": 0.7687292978490674, | |
| "eval_recall": 0.7513869625520111, | |
| "eval_runtime": 22.9246, | |
| "eval_samples_per_second": 125.804, | |
| "eval_steps_per_second": 15.747, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.284310340881348, | |
| "learning_rate": 7.867104183757178e-05, | |
| "loss": 1.0213, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7635228848821082, | |
| "eval_f1": 0.7421686424281113, | |
| "eval_loss": 0.6378183960914612, | |
| "eval_precision": 0.771790199319593, | |
| "eval_recall": 0.7635228848821082, | |
| "eval_runtime": 23.1763, | |
| "eval_samples_per_second": 124.438, | |
| "eval_steps_per_second": 15.576, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 26.305831909179688, | |
| "learning_rate": 9.99845966779335e-05, | |
| "loss": 0.9727, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7437586685159501, | |
| "eval_f1": 0.7593875900140131, | |
| "eval_loss": 0.604518473148346, | |
| "eval_precision": 0.7972091673281284, | |
| "eval_recall": 0.7437586685159501, | |
| "eval_runtime": 23.2491, | |
| "eval_samples_per_second": 124.048, | |
| "eval_steps_per_second": 15.527, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 7.570508003234863, | |
| "learning_rate": 9.93971225198763e-05, | |
| "loss": 0.9062, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7902219140083218, | |
| "eval_f1": 0.7801714004286384, | |
| "eval_loss": 0.5953397750854492, | |
| "eval_precision": 0.7964554700693868, | |
| "eval_recall": 0.7902219140083218, | |
| "eval_runtime": 23.2969, | |
| "eval_samples_per_second": 123.793, | |
| "eval_steps_per_second": 15.496, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 14.241063117980957, | |
| "learning_rate": 9.796799913911281e-05, | |
| "loss": 0.8719, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7742718446601942, | |
| "eval_f1": 0.783910785094317, | |
| "eval_loss": 0.6095318794250488, | |
| "eval_precision": 0.8083582615215125, | |
| "eval_recall": 0.7742718446601942, | |
| "eval_runtime": 23.8601, | |
| "eval_samples_per_second": 120.871, | |
| "eval_steps_per_second": 15.13, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 7.0408935546875, | |
| "learning_rate": 9.572157654878572e-05, | |
| "loss": 0.7537, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.763869625520111, | |
| "eval_f1": 0.7777621987888312, | |
| "eval_loss": 0.5970358848571777, | |
| "eval_precision": 0.8124970718532031, | |
| "eval_recall": 0.763869625520111, | |
| "eval_runtime": 23.585, | |
| "eval_samples_per_second": 122.281, | |
| "eval_steps_per_second": 15.306, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.94745135307312, | |
| "learning_rate": 9.26961302542397e-05, | |
| "loss": 0.677, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7073509015256588, | |
| "eval_f1": 0.7301082277170373, | |
| "eval_loss": 0.7107765078544617, | |
| "eval_precision": 0.8147970005525182, | |
| "eval_recall": 0.7073509015256588, | |
| "eval_runtime": 22.7748, | |
| "eval_samples_per_second": 126.631, | |
| "eval_steps_per_second": 15.851, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 11.812759399414062, | |
| "learning_rate": 8.89432090986511e-05, | |
| "loss": 0.6638, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8290568654646324, | |
| "eval_f1": 0.8244010809843568, | |
| "eval_loss": 0.49241259694099426, | |
| "eval_precision": 0.8249534082413604, | |
| "eval_recall": 0.8290568654646324, | |
| "eval_runtime": 23.6529, | |
| "eval_samples_per_second": 121.93, | |
| "eval_steps_per_second": 15.262, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 8.607577323913574, | |
| "learning_rate": 8.45267569518721e-05, | |
| "loss": 0.5787, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8162274618585298, | |
| "eval_f1": 0.82222451204188, | |
| "eval_loss": 0.54153972864151, | |
| "eval_precision": 0.8405647026780846, | |
| "eval_recall": 0.8162274618585298, | |
| "eval_runtime": 23.1538, | |
| "eval_samples_per_second": 124.559, | |
| "eval_steps_per_second": 15.591, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 22.480600357055664, | |
| "learning_rate": 7.952202320752798e-05, | |
| "loss": 0.5373, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8103328710124826, | |
| "eval_f1": 0.8188747354879483, | |
| "eval_loss": 0.5297914147377014, | |
| "eval_precision": 0.8409211427101868, | |
| "eval_recall": 0.8103328710124826, | |
| "eval_runtime": 23.2794, | |
| "eval_samples_per_second": 123.886, | |
| "eval_steps_per_second": 15.507, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 37.149688720703125, | |
| "learning_rate": 7.401428065178325e-05, | |
| "loss": 0.4923, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8117198335644937, | |
| "eval_f1": 0.8212533704879323, | |
| "eval_loss": 0.5427994728088379, | |
| "eval_precision": 0.8444195978517209, | |
| "eval_recall": 0.8117198335644937, | |
| "eval_runtime": 23.1273, | |
| "eval_samples_per_second": 124.701, | |
| "eval_steps_per_second": 15.609, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 7.792755126953125, | |
| "learning_rate": 6.80973725492743e-05, | |
| "loss": 0.3798, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8498613037447988, | |
| "eval_f1": 0.8466618158413786, | |
| "eval_loss": 0.49677935242652893, | |
| "eval_precision": 0.8470322647118588, | |
| "eval_recall": 0.8498613037447988, | |
| "eval_runtime": 22.6968, | |
| "eval_samples_per_second": 127.066, | |
| "eval_steps_per_second": 15.905, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 12.735553741455078, | |
| "learning_rate": 6.187211370157784e-05, | |
| "loss": 0.3912, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8443134535367545, | |
| "eval_f1": 0.8460191446346724, | |
| "eval_loss": 0.5338780879974365, | |
| "eval_precision": 0.8530926029478479, | |
| "eval_recall": 0.8443134535367545, | |
| "eval_runtime": 22.4911, | |
| "eval_samples_per_second": 128.229, | |
| "eval_steps_per_second": 16.051, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.7020573616027832, | |
| "learning_rate": 5.544457272166217e-05, | |
| "loss": 0.3002, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.84500693481276, | |
| "eval_f1": 0.8481459887384893, | |
| "eval_loss": 0.521908164024353, | |
| "eval_precision": 0.8547788879749323, | |
| "eval_recall": 0.84500693481276, | |
| "eval_runtime": 22.516, | |
| "eval_samples_per_second": 128.087, | |
| "eval_steps_per_second": 16.033, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 10.823068618774414, | |
| "learning_rate": 4.894460661440583e-05, | |
| "loss": 0.2744, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8203883495145631, | |
| "eval_f1": 0.827968123001256, | |
| "eval_loss": 0.636923611164093, | |
| "eval_precision": 0.8481543275037586, | |
| "eval_recall": 0.8203883495145631, | |
| "eval_runtime": 22.8235, | |
| "eval_samples_per_second": 126.361, | |
| "eval_steps_per_second": 15.817, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.4310046434402466, | |
| "learning_rate": 4.244239774409037e-05, | |
| "loss": 0.2251, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8571428571428571, | |
| "eval_f1": 0.8555949794732268, | |
| "eval_loss": 0.515616238117218, | |
| "eval_precision": 0.8560689662280283, | |
| "eval_recall": 0.8571428571428571, | |
| "eval_runtime": 22.9434, | |
| "eval_samples_per_second": 125.701, | |
| "eval_steps_per_second": 15.734, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 13.193678855895996, | |
| "learning_rate": 3.606895852147351e-05, | |
| "loss": 0.2187, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8457004160887656, | |
| "eval_f1": 0.849063653899852, | |
| "eval_loss": 0.5825002789497375, | |
| "eval_precision": 0.8549951611862869, | |
| "eval_recall": 0.8457004160887656, | |
| "eval_runtime": 23.1083, | |
| "eval_samples_per_second": 124.804, | |
| "eval_steps_per_second": 15.622, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.2162574678659439, | |
| "learning_rate": 2.9932882319894417e-05, | |
| "loss": 0.1767, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.852635228848821, | |
| "eval_f1": 0.8550909208177824, | |
| "eval_loss": 0.5693491697311401, | |
| "eval_precision": 0.8604835069555719, | |
| "eval_recall": 0.852635228848821, | |
| "eval_runtime": 22.8714, | |
| "eval_samples_per_second": 126.097, | |
| "eval_steps_per_second": 15.784, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "step": 6099, | |
| "total_flos": 7.550537882380222e+18, | |
| "train_loss": 0.6395374234923966, | |
| "train_runtime": 2241.3643, | |
| "train_samples_per_second": 228.789, | |
| "train_steps_per_second": 14.322 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 7.550537882380222e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |