| { | |
| "best_metric": 0.8673913043478261, | |
| "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-teeth_dataset/checkpoint-67", | |
| "epoch": 40.0, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.8, | |
| "eval_accuracy": 0.7913043478260869, | |
| "eval_loss": 1.1515443325042725, | |
| "eval_runtime": 4.209, | |
| "eval_samples_per_second": 109.29, | |
| "eval_steps_per_second": 3.564, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.7630434782608696, | |
| "eval_loss": 1.1545782089233398, | |
| "eval_runtime": 4.4373, | |
| "eval_samples_per_second": 103.665, | |
| "eval_steps_per_second": 3.38, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 12.2432222366333, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.6441, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_accuracy": 0.8065217391304348, | |
| "eval_loss": 1.070844292640686, | |
| "eval_runtime": 4.2607, | |
| "eval_samples_per_second": 107.964, | |
| "eval_steps_per_second": 3.521, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8217391304347826, | |
| "eval_loss": 0.992546796798706, | |
| "eval_runtime": 4.1431, | |
| "eval_samples_per_second": 111.027, | |
| "eval_steps_per_second": 3.62, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_accuracy": 0.8108695652173913, | |
| "eval_loss": 0.99262535572052, | |
| "eval_runtime": 3.9996, | |
| "eval_samples_per_second": 115.012, | |
| "eval_steps_per_second": 3.75, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 12.744916915893555, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.6042, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "eval_accuracy": 0.7934782608695652, | |
| "eval_loss": 1.016530990600586, | |
| "eval_runtime": 4.0517, | |
| "eval_samples_per_second": 113.532, | |
| "eval_steps_per_second": 3.702, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "eval_accuracy": 0.8347826086956521, | |
| "eval_loss": 0.8679854273796082, | |
| "eval_runtime": 3.9943, | |
| "eval_samples_per_second": 115.163, | |
| "eval_steps_per_second": 3.755, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 23.645719528198242, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.4293, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8347826086956521, | |
| "eval_loss": 0.8360960483551025, | |
| "eval_runtime": 4.0619, | |
| "eval_samples_per_second": 113.248, | |
| "eval_steps_per_second": 3.693, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "eval_accuracy": 0.8565217391304348, | |
| "eval_loss": 0.8075701594352722, | |
| "eval_runtime": 4.0291, | |
| "eval_samples_per_second": 114.17, | |
| "eval_steps_per_second": 3.723, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 9.87, | |
| "eval_accuracy": 0.8565217391304348, | |
| "eval_loss": 0.7547765970230103, | |
| "eval_runtime": 4.0783, | |
| "eval_samples_per_second": 112.793, | |
| "eval_steps_per_second": 3.678, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "grad_norm": 20.158720016479492, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.4471, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 10.93, | |
| "eval_accuracy": 0.8369565217391305, | |
| "eval_loss": 0.7901058793067932, | |
| "eval_runtime": 4.2465, | |
| "eval_samples_per_second": 108.325, | |
| "eval_steps_per_second": 3.532, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8543478260869565, | |
| "eval_loss": 0.7326875925064087, | |
| "eval_runtime": 4.3562, | |
| "eval_samples_per_second": 105.597, | |
| "eval_steps_per_second": 3.443, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "eval_accuracy": 0.8478260869565217, | |
| "eval_loss": 0.748528003692627, | |
| "eval_runtime": 4.402, | |
| "eval_samples_per_second": 104.498, | |
| "eval_steps_per_second": 3.408, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "grad_norm": 11.568655967712402, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.3436, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 13.87, | |
| "eval_accuracy": 0.8347826086956521, | |
| "eval_loss": 0.7844414710998535, | |
| "eval_runtime": 4.2227, | |
| "eval_samples_per_second": 108.934, | |
| "eval_steps_per_second": 3.552, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "eval_accuracy": 0.8630434782608696, | |
| "eval_loss": 0.6644619107246399, | |
| "eval_runtime": 4.1242, | |
| "eval_samples_per_second": 111.537, | |
| "eval_steps_per_second": 3.637, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 14.64358139038086, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.2813, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8391304347826087, | |
| "eval_loss": 0.7370373010635376, | |
| "eval_runtime": 4.0228, | |
| "eval_samples_per_second": 114.348, | |
| "eval_steps_per_second": 3.729, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "eval_accuracy": 0.8347826086956521, | |
| "eval_loss": 0.7598132491111755, | |
| "eval_runtime": 3.9928, | |
| "eval_samples_per_second": 115.208, | |
| "eval_steps_per_second": 3.757, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 17.87, | |
| "eval_accuracy": 0.8673913043478261, | |
| "eval_loss": 0.6579979062080383, | |
| "eval_runtime": 4.1194, | |
| "eval_samples_per_second": 111.667, | |
| "eval_steps_per_second": 3.641, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 18.67, | |
| "grad_norm": 16.545446395874023, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.275, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 18.93, | |
| "eval_accuracy": 0.8456521739130435, | |
| "eval_loss": 0.7026421427726746, | |
| "eval_runtime": 4.039, | |
| "eval_samples_per_second": 113.889, | |
| "eval_steps_per_second": 3.714, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8369565217391305, | |
| "eval_loss": 0.7055774927139282, | |
| "eval_runtime": 4.0414, | |
| "eval_samples_per_second": 113.823, | |
| "eval_steps_per_second": 3.712, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "eval_accuracy": 0.8543478260869565, | |
| "eval_loss": 0.6517285108566284, | |
| "eval_runtime": 4.1333, | |
| "eval_samples_per_second": 111.292, | |
| "eval_steps_per_second": 3.629, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 21.33, | |
| "grad_norm": 13.078596115112305, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.2674, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 21.87, | |
| "eval_accuracy": 0.8608695652173913, | |
| "eval_loss": 0.5930755734443665, | |
| "eval_runtime": 4.2994, | |
| "eval_samples_per_second": 106.991, | |
| "eval_steps_per_second": 3.489, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 22.93, | |
| "eval_accuracy": 0.8608695652173913, | |
| "eval_loss": 0.5985938310623169, | |
| "eval_runtime": 4.3353, | |
| "eval_samples_per_second": 106.105, | |
| "eval_steps_per_second": 3.46, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 8.3045015335083, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.2278, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.8608695652173913, | |
| "eval_loss": 0.5982087254524231, | |
| "eval_runtime": 4.3231, | |
| "eval_samples_per_second": 106.404, | |
| "eval_steps_per_second": 3.47, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 24.8, | |
| "eval_accuracy": 0.8543478260869565, | |
| "eval_loss": 0.604324460029602, | |
| "eval_runtime": 4.0948, | |
| "eval_samples_per_second": 112.337, | |
| "eval_steps_per_second": 3.663, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 25.87, | |
| "eval_accuracy": 0.8521739130434782, | |
| "eval_loss": 0.6148393750190735, | |
| "eval_runtime": 4.0061, | |
| "eval_samples_per_second": 114.824, | |
| "eval_steps_per_second": 3.744, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "grad_norm": 15.322602272033691, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.2034, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 26.93, | |
| "eval_accuracy": 0.85, | |
| "eval_loss": 0.6160370111465454, | |
| "eval_runtime": 4.0736, | |
| "eval_samples_per_second": 112.922, | |
| "eval_steps_per_second": 3.682, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.8521739130434782, | |
| "eval_loss": 0.5903608798980713, | |
| "eval_runtime": 4.1333, | |
| "eval_samples_per_second": 111.291, | |
| "eval_steps_per_second": 3.629, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 28.8, | |
| "eval_accuracy": 0.8608695652173913, | |
| "eval_loss": 0.5824848413467407, | |
| "eval_runtime": 4.2951, | |
| "eval_samples_per_second": 107.1, | |
| "eval_steps_per_second": 3.492, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 29.33, | |
| "grad_norm": 12.949268341064453, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.2207, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 29.87, | |
| "eval_accuracy": 0.8478260869565217, | |
| "eval_loss": 0.6132237911224365, | |
| "eval_runtime": 4.382, | |
| "eval_samples_per_second": 104.975, | |
| "eval_steps_per_second": 3.423, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 30.93, | |
| "eval_accuracy": 0.8391304347826087, | |
| "eval_loss": 0.6181351542472839, | |
| "eval_runtime": 4.27, | |
| "eval_samples_per_second": 107.728, | |
| "eval_steps_per_second": 3.513, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 5.803860664367676, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1701, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.85, | |
| "eval_loss": 0.5688785910606384, | |
| "eval_runtime": 4.1338, | |
| "eval_samples_per_second": 111.279, | |
| "eval_steps_per_second": 3.629, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 32.8, | |
| "eval_accuracy": 0.8608695652173913, | |
| "eval_loss": 0.5633876919746399, | |
| "eval_runtime": 4.7704, | |
| "eval_samples_per_second": 96.428, | |
| "eval_steps_per_second": 3.144, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 33.87, | |
| "eval_accuracy": 0.8652173913043478, | |
| "eval_loss": 0.5659446120262146, | |
| "eval_runtime": 4.0029, | |
| "eval_samples_per_second": 114.915, | |
| "eval_steps_per_second": 3.747, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 34.67, | |
| "grad_norm": 18.19119644165039, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1681, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 34.93, | |
| "eval_accuracy": 0.8565217391304348, | |
| "eval_loss": 0.5895132422447205, | |
| "eval_runtime": 4.0203, | |
| "eval_samples_per_second": 114.42, | |
| "eval_steps_per_second": 3.731, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.85, | |
| "eval_loss": 0.6094810366630554, | |
| "eval_runtime": 4.1594, | |
| "eval_samples_per_second": 110.594, | |
| "eval_steps_per_second": 3.606, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 36.8, | |
| "eval_accuracy": 0.8456521739130435, | |
| "eval_loss": 0.614500105381012, | |
| "eval_runtime": 4.3427, | |
| "eval_samples_per_second": 105.924, | |
| "eval_steps_per_second": 3.454, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 37.33, | |
| "grad_norm": 7.945148468017578, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.1435, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 37.87, | |
| "eval_accuracy": 0.8478260869565217, | |
| "eval_loss": 0.6111825704574585, | |
| "eval_runtime": 4.3974, | |
| "eval_samples_per_second": 104.606, | |
| "eval_steps_per_second": 3.411, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 38.93, | |
| "eval_accuracy": 0.8456521739130435, | |
| "eval_loss": 0.599669873714447, | |
| "eval_runtime": 4.2726, | |
| "eval_samples_per_second": 107.662, | |
| "eval_steps_per_second": 3.511, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 11.281535148620605, | |
| "learning_rate": 0.0, | |
| "loss": 0.1586, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.85, | |
| "eval_loss": 0.5965761542320251, | |
| "eval_runtime": 4.0634, | |
| "eval_samples_per_second": 113.205, | |
| "eval_steps_per_second": 3.691, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "step": 150, | |
| "total_flos": 4.585003492737024e+17, | |
| "train_loss": 0.305610773563385, | |
| "train_runtime": 583.5167, | |
| "train_samples_per_second": 39.416, | |
| "train_steps_per_second": 0.257 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 4.585003492737024e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |