| { | |
| "best_metric": 1.0, | |
| "best_model_checkpoint": "vit-base-patch16-224-dmae-va-U5-40/checkpoint-65", | |
| "epoch": 36.92307692307692, | |
| "eval_steps": 500, | |
| "global_step": 240, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.36054421768707484, | |
| "eval_loss": 1.4235920906066895, | |
| "eval_runtime": 2.4199, | |
| "eval_samples_per_second": 60.746, | |
| "eval_steps_per_second": 2.066, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 4.381593227386475, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.3165, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5306122448979592, | |
| "eval_loss": 1.107216715812683, | |
| "eval_runtime": 1.8757, | |
| "eval_samples_per_second": 78.37, | |
| "eval_steps_per_second": 2.666, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "eval_accuracy": 0.6462585034013606, | |
| "eval_loss": 0.9370094537734985, | |
| "eval_runtime": 1.9321, | |
| "eval_samples_per_second": 76.085, | |
| "eval_steps_per_second": 2.588, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 2.5531413555145264, | |
| "learning_rate": 5e-05, | |
| "loss": 0.93, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7687074829931972, | |
| "eval_loss": 0.6869749426841736, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 74.243, | |
| "eval_steps_per_second": 2.525, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "eval_accuracy": 0.8163265306122449, | |
| "eval_loss": 0.47430145740509033, | |
| "eval_runtime": 1.9979, | |
| "eval_samples_per_second": 73.576, | |
| "eval_steps_per_second": 2.503, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 3.2196173667907715, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 0.5368, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9183673469387755, | |
| "eval_loss": 0.2825295627117157, | |
| "eval_runtime": 1.9929, | |
| "eval_samples_per_second": 73.762, | |
| "eval_steps_per_second": 2.509, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "eval_accuracy": 0.9523809523809523, | |
| "eval_loss": 0.20659707486629486, | |
| "eval_runtime": 1.9939, | |
| "eval_samples_per_second": 73.724, | |
| "eval_steps_per_second": 2.508, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 7.38, | |
| "grad_norm": 2.4960756301879883, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.2989, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9727891156462585, | |
| "eval_loss": 0.12244854122400284, | |
| "eval_runtime": 1.9835, | |
| "eval_samples_per_second": 74.113, | |
| "eval_steps_per_second": 2.521, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "eval_accuracy": 0.9591836734693877, | |
| "eval_loss": 0.1452772617340088, | |
| "eval_runtime": 2.011, | |
| "eval_samples_per_second": 73.097, | |
| "eval_steps_per_second": 2.486, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 9.23, | |
| "grad_norm": 2.9144787788391113, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.1746, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.03671738877892494, | |
| "eval_runtime": 2.0115, | |
| "eval_samples_per_second": 73.081, | |
| "eval_steps_per_second": 2.486, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 10.92, | |
| "eval_accuracy": 0.9863945578231292, | |
| "eval_loss": 0.06163691729307175, | |
| "eval_runtime": 1.9765, | |
| "eval_samples_per_second": 74.373, | |
| "eval_steps_per_second": 2.53, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 11.08, | |
| "grad_norm": 1.5245277881622314, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.1596, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.02339293248951435, | |
| "eval_runtime": 1.9968, | |
| "eval_samples_per_second": 73.619, | |
| "eval_steps_per_second": 2.504, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "grad_norm": 2.6396472454071045, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.1094, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.029785707592964172, | |
| "eval_runtime": 1.9947, | |
| "eval_samples_per_second": 73.695, | |
| "eval_steps_per_second": 2.507, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9931972789115646, | |
| "eval_loss": 0.044449541717767715, | |
| "eval_runtime": 1.9933, | |
| "eval_samples_per_second": 73.747, | |
| "eval_steps_per_second": 2.508, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 14.77, | |
| "grad_norm": 1.8975120782852173, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.1123, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 14.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.025111636146903038, | |
| "eval_runtime": 2.0344, | |
| "eval_samples_per_second": 72.256, | |
| "eval_steps_per_second": 2.458, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.018478814512491226, | |
| "eval_runtime": 1.9976, | |
| "eval_samples_per_second": 73.589, | |
| "eval_steps_per_second": 2.503, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 16.62, | |
| "grad_norm": 2.431682825088501, | |
| "learning_rate": 3.055555555555556e-05, | |
| "loss": 0.0761, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 16.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.015857575461268425, | |
| "eval_runtime": 2.0057, | |
| "eval_samples_per_second": 73.291, | |
| "eval_steps_per_second": 2.493, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.01796538569033146, | |
| "eval_runtime": 2.0303, | |
| "eval_samples_per_second": 72.403, | |
| "eval_steps_per_second": 2.463, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 18.46, | |
| "grad_norm": 2.1142959594726562, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0743, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 18.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.011095724068582058, | |
| "eval_runtime": 1.9932, | |
| "eval_samples_per_second": 73.752, | |
| "eval_steps_per_second": 2.509, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.013449358753859997, | |
| "eval_runtime": 1.9948, | |
| "eval_samples_per_second": 73.693, | |
| "eval_steps_per_second": 2.507, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 20.31, | |
| "grad_norm": 1.569281816482544, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.072, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 20.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.012339694425463676, | |
| "eval_runtime": 1.9955, | |
| "eval_samples_per_second": 73.666, | |
| "eval_steps_per_second": 2.506, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.010039811953902245, | |
| "eval_runtime": 1.9958, | |
| "eval_samples_per_second": 73.653, | |
| "eval_steps_per_second": 2.505, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 22.15, | |
| "grad_norm": 2.3920440673828125, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0744, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 22.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.007376910652965307, | |
| "eval_runtime": 1.9896, | |
| "eval_samples_per_second": 73.884, | |
| "eval_steps_per_second": 2.513, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.7604403495788574, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.0688, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.006399971432983875, | |
| "eval_runtime": 1.9944, | |
| "eval_samples_per_second": 73.705, | |
| "eval_steps_per_second": 2.507, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 24.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.006952380761504173, | |
| "eval_runtime": 2.0094, | |
| "eval_samples_per_second": 73.157, | |
| "eval_steps_per_second": 2.488, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 25.85, | |
| "grad_norm": 1.1607767343521118, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0737, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.006380416918545961, | |
| "eval_runtime": 2.0125, | |
| "eval_samples_per_second": 73.044, | |
| "eval_steps_per_second": 2.484, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 26.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.00552676897495985, | |
| "eval_runtime": 2.2042, | |
| "eval_samples_per_second": 66.69, | |
| "eval_steps_per_second": 2.268, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 27.69, | |
| "grad_norm": 2.0216870307922363, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.053, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.007457624655216932, | |
| "eval_runtime": 2.2171, | |
| "eval_samples_per_second": 66.304, | |
| "eval_steps_per_second": 2.255, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 28.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.004635449033230543, | |
| "eval_runtime": 2.1386, | |
| "eval_samples_per_second": 68.735, | |
| "eval_steps_per_second": 2.338, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 29.54, | |
| "grad_norm": 1.227410912513733, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0677, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.004566570743918419, | |
| "eval_runtime": 2.0055, | |
| "eval_samples_per_second": 73.299, | |
| "eval_steps_per_second": 2.493, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 30.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.00975565891712904, | |
| "eval_runtime": 1.9971, | |
| "eval_samples_per_second": 73.607, | |
| "eval_steps_per_second": 2.504, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 31.38, | |
| "grad_norm": 1.6289031505584717, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.055, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.008506382815539837, | |
| "eval_runtime": 1.9954, | |
| "eval_samples_per_second": 73.671, | |
| "eval_steps_per_second": 2.506, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 32.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.0055803339928388596, | |
| "eval_runtime": 2.0106, | |
| "eval_samples_per_second": 73.111, | |
| "eval_steps_per_second": 2.487, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 33.23, | |
| "grad_norm": 1.2801518440246582, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0576, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.005879435688257217, | |
| "eval_runtime": 1.9938, | |
| "eval_samples_per_second": 73.727, | |
| "eval_steps_per_second": 2.508, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 34.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.005405152682214975, | |
| "eval_runtime": 2.0162, | |
| "eval_samples_per_second": 72.911, | |
| "eval_steps_per_second": 2.48, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 35.08, | |
| "grad_norm": 1.9271787405014038, | |
| "learning_rate": 2.777777777777778e-06, | |
| "loss": 0.0697, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.005514008458703756, | |
| "eval_runtime": 1.9717, | |
| "eval_samples_per_second": 74.555, | |
| "eval_steps_per_second": 2.536, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 36.92, | |
| "grad_norm": 0.9594835638999939, | |
| "learning_rate": 0.0, | |
| "loss": 0.0415, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 36.92, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.005595473572611809, | |
| "eval_runtime": 2.0157, | |
| "eval_samples_per_second": 72.927, | |
| "eval_steps_per_second": 2.481, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 36.92, | |
| "step": 240, | |
| "total_flos": 2.366647785100247e+18, | |
| "train_loss": 0.2210820493598779, | |
| "train_runtime": 1366.588, | |
| "train_samples_per_second": 24.206, | |
| "train_steps_per_second": 0.176 | |
| } | |
| ], | |
| "logging_steps": 12, | |
| "max_steps": 240, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 40, | |
| "save_steps": 500, | |
| "total_flos": 2.366647785100247e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |