|
{ |
|
"best_metric": 0.6388888888888888, |
|
"best_model_checkpoint": "deit-base-distilled-patch16-224-hasta-65-fold4/checkpoint-77", |
|
"epoch": 57.142857142857146, |
|
"eval_steps": 500, |
|
"global_step": 100, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.5714285714285714, |
|
"eval_accuracy": 0.2777777777777778, |
|
"eval_loss": 1.2682678699493408, |
|
"eval_runtime": 0.4974, |
|
"eval_samples_per_second": 72.376, |
|
"eval_steps_per_second": 4.021, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 1.7142857142857144, |
|
"eval_accuracy": 0.3055555555555556, |
|
"eval_loss": 1.1169365644454956, |
|
"eval_runtime": 0.5012, |
|
"eval_samples_per_second": 71.823, |
|
"eval_steps_per_second": 3.99, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"eval_accuracy": 0.2777777777777778, |
|
"eval_loss": 1.2962708473205566, |
|
"eval_runtime": 0.5221, |
|
"eval_samples_per_second": 68.953, |
|
"eval_steps_per_second": 3.831, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.2777777777777778, |
|
"eval_loss": 1.2728766202926636, |
|
"eval_runtime": 0.511, |
|
"eval_samples_per_second": 70.456, |
|
"eval_steps_per_second": 3.914, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 4.571428571428571, |
|
"eval_accuracy": 0.2777777777777778, |
|
"eval_loss": 1.1326817274093628, |
|
"eval_runtime": 0.5066, |
|
"eval_samples_per_second": 71.058, |
|
"eval_steps_per_second": 3.948, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"grad_norm": 3.3327667713165283, |
|
"learning_rate": 5e-05, |
|
"loss": 1.1554, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"eval_accuracy": 0.3888888888888889, |
|
"eval_loss": 1.0658352375030518, |
|
"eval_runtime": 0.4944, |
|
"eval_samples_per_second": 72.821, |
|
"eval_steps_per_second": 4.046, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 6.857142857142857, |
|
"eval_accuracy": 0.3611111111111111, |
|
"eval_loss": 1.151309847831726, |
|
"eval_runtime": 0.485, |
|
"eval_samples_per_second": 74.223, |
|
"eval_steps_per_second": 4.124, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.3888888888888889, |
|
"eval_loss": 1.1798954010009766, |
|
"eval_runtime": 0.4875, |
|
"eval_samples_per_second": 73.845, |
|
"eval_steps_per_second": 4.102, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 8.571428571428571, |
|
"eval_accuracy": 0.3611111111111111, |
|
"eval_loss": 1.1289156675338745, |
|
"eval_runtime": 0.4857, |
|
"eval_samples_per_second": 74.117, |
|
"eval_steps_per_second": 4.118, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 9.714285714285714, |
|
"eval_accuracy": 0.4166666666666667, |
|
"eval_loss": 1.0166828632354736, |
|
"eval_runtime": 0.4896, |
|
"eval_samples_per_second": 73.531, |
|
"eval_steps_per_second": 4.085, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 10.857142857142858, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.007408618927002, |
|
"eval_runtime": 0.4925, |
|
"eval_samples_per_second": 73.1, |
|
"eval_steps_per_second": 4.061, |
|
"step": 19 |
|
}, |
|
{ |
|
"epoch": 11.428571428571429, |
|
"grad_norm": 3.3315768241882324, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 0.9967, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 0.9982408881187439, |
|
"eval_runtime": 0.4991, |
|
"eval_samples_per_second": 72.13, |
|
"eval_steps_per_second": 4.007, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 12.571428571428571, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.9706911444664001, |
|
"eval_runtime": 0.4937, |
|
"eval_samples_per_second": 72.921, |
|
"eval_steps_per_second": 4.051, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 13.714285714285714, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.9401130676269531, |
|
"eval_runtime": 0.4894, |
|
"eval_samples_per_second": 73.56, |
|
"eval_steps_per_second": 4.087, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 14.857142857142858, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.0155110359191895, |
|
"eval_runtime": 0.5095, |
|
"eval_samples_per_second": 70.661, |
|
"eval_steps_per_second": 3.926, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.4444444444444444, |
|
"eval_loss": 1.1357495784759521, |
|
"eval_runtime": 0.4879, |
|
"eval_samples_per_second": 73.781, |
|
"eval_steps_per_second": 4.099, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 16.571428571428573, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.1284376382827759, |
|
"eval_runtime": 0.484, |
|
"eval_samples_per_second": 74.381, |
|
"eval_steps_per_second": 4.132, |
|
"step": 29 |
|
}, |
|
{ |
|
"epoch": 17.142857142857142, |
|
"grad_norm": 6.601634979248047, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.8472, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 17.714285714285715, |
|
"eval_accuracy": 0.4166666666666667, |
|
"eval_loss": 1.1839802265167236, |
|
"eval_runtime": 0.4942, |
|
"eval_samples_per_second": 72.842, |
|
"eval_steps_per_second": 4.047, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 18.857142857142858, |
|
"eval_accuracy": 0.4444444444444444, |
|
"eval_loss": 1.3015050888061523, |
|
"eval_runtime": 0.4925, |
|
"eval_samples_per_second": 73.09, |
|
"eval_steps_per_second": 4.061, |
|
"step": 33 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 0.9754727482795715, |
|
"eval_runtime": 0.5073, |
|
"eval_samples_per_second": 70.962, |
|
"eval_steps_per_second": 3.942, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 20.571428571428573, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.9602252244949341, |
|
"eval_runtime": 0.4935, |
|
"eval_samples_per_second": 72.943, |
|
"eval_steps_per_second": 4.052, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 21.714285714285715, |
|
"eval_accuracy": 0.4444444444444444, |
|
"eval_loss": 1.0950214862823486, |
|
"eval_runtime": 0.5017, |
|
"eval_samples_per_second": 71.753, |
|
"eval_steps_per_second": 3.986, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"grad_norm": 7.189661502838135, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.7133, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"eval_accuracy": 0.4722222222222222, |
|
"eval_loss": 1.060685157775879, |
|
"eval_runtime": 0.4875, |
|
"eval_samples_per_second": 73.839, |
|
"eval_steps_per_second": 4.102, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 0.9962939023971558, |
|
"eval_runtime": 0.489, |
|
"eval_samples_per_second": 73.624, |
|
"eval_steps_per_second": 4.09, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 24.571428571428573, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.02350914478302, |
|
"eval_runtime": 0.502, |
|
"eval_samples_per_second": 71.715, |
|
"eval_steps_per_second": 3.984, |
|
"step": 43 |
|
}, |
|
{ |
|
"epoch": 25.714285714285715, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.0871917009353638, |
|
"eval_runtime": 0.4849, |
|
"eval_samples_per_second": 74.239, |
|
"eval_steps_per_second": 4.124, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 26.857142857142858, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.0525965690612793, |
|
"eval_runtime": 0.4947, |
|
"eval_samples_per_second": 72.765, |
|
"eval_steps_per_second": 4.042, |
|
"step": 47 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 1.157909870147705, |
|
"eval_runtime": 0.4887, |
|
"eval_samples_per_second": 73.67, |
|
"eval_steps_per_second": 4.093, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"grad_norm": 2.7332067489624023, |
|
"learning_rate": 2.777777777777778e-05, |
|
"loss": 0.5648, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"eval_accuracy": 0.4444444444444444, |
|
"eval_loss": 1.2433772087097168, |
|
"eval_runtime": 0.4871, |
|
"eval_samples_per_second": 73.904, |
|
"eval_steps_per_second": 4.106, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 29.714285714285715, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.165271520614624, |
|
"eval_runtime": 0.4955, |
|
"eval_samples_per_second": 72.648, |
|
"eval_steps_per_second": 4.036, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 30.857142857142858, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.094743251800537, |
|
"eval_runtime": 0.4971, |
|
"eval_samples_per_second": 72.424, |
|
"eval_steps_per_second": 4.024, |
|
"step": 54 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.1530675888061523, |
|
"eval_runtime": 0.4877, |
|
"eval_samples_per_second": 73.82, |
|
"eval_steps_per_second": 4.101, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 32.57142857142857, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.12678861618042, |
|
"eval_runtime": 0.4983, |
|
"eval_samples_per_second": 72.244, |
|
"eval_steps_per_second": 4.014, |
|
"step": 57 |
|
}, |
|
{ |
|
"epoch": 33.714285714285715, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.0664329528808594, |
|
"eval_runtime": 0.4942, |
|
"eval_samples_per_second": 72.852, |
|
"eval_steps_per_second": 4.047, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 34.285714285714285, |
|
"grad_norm": 5.203519821166992, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.4429, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 34.857142857142854, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.1503134965896606, |
|
"eval_runtime": 0.4982, |
|
"eval_samples_per_second": 72.253, |
|
"eval_steps_per_second": 4.014, |
|
"step": 61 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.3472707271575928, |
|
"eval_runtime": 0.4995, |
|
"eval_samples_per_second": 72.074, |
|
"eval_steps_per_second": 4.004, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 36.57142857142857, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.2785837650299072, |
|
"eval_runtime": 0.5062, |
|
"eval_samples_per_second": 71.114, |
|
"eval_steps_per_second": 3.951, |
|
"step": 64 |
|
}, |
|
{ |
|
"epoch": 37.714285714285715, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.0905375480651855, |
|
"eval_runtime": 0.4967, |
|
"eval_samples_per_second": 72.474, |
|
"eval_steps_per_second": 4.026, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 38.857142857142854, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.0916866064071655, |
|
"eval_runtime": 0.4962, |
|
"eval_samples_per_second": 72.545, |
|
"eval_steps_per_second": 4.03, |
|
"step": 68 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"grad_norm": 7.39570951461792, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.4313, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.207929253578186, |
|
"eval_runtime": 0.498, |
|
"eval_samples_per_second": 72.291, |
|
"eval_steps_per_second": 4.016, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 40.57142857142857, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.2500560283660889, |
|
"eval_runtime": 0.4904, |
|
"eval_samples_per_second": 73.41, |
|
"eval_steps_per_second": 4.078, |
|
"step": 71 |
|
}, |
|
{ |
|
"epoch": 41.714285714285715, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.17886483669281, |
|
"eval_runtime": 0.4909, |
|
"eval_samples_per_second": 73.339, |
|
"eval_steps_per_second": 4.074, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.1125812530517578, |
|
"eval_runtime": 0.4967, |
|
"eval_samples_per_second": 72.474, |
|
"eval_steps_per_second": 4.026, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 1.110924243927002, |
|
"eval_runtime": 0.4911, |
|
"eval_samples_per_second": 73.298, |
|
"eval_steps_per_second": 4.072, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 44.57142857142857, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 1.123566746711731, |
|
"eval_runtime": 0.4872, |
|
"eval_samples_per_second": 73.888, |
|
"eval_steps_per_second": 4.105, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"grad_norm": 2.354403257369995, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.3764, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.221116304397583, |
|
"eval_runtime": 0.4928, |
|
"eval_samples_per_second": 73.053, |
|
"eval_steps_per_second": 4.058, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 46.857142857142854, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 1.3020737171173096, |
|
"eval_runtime": 0.4951, |
|
"eval_samples_per_second": 72.707, |
|
"eval_steps_per_second": 4.039, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.3182320594787598, |
|
"eval_runtime": 0.5085, |
|
"eval_samples_per_second": 70.797, |
|
"eval_steps_per_second": 3.933, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 48.57142857142857, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 1.2830231189727783, |
|
"eval_runtime": 0.495, |
|
"eval_samples_per_second": 72.724, |
|
"eval_steps_per_second": 4.04, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 49.714285714285715, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 1.2502837181091309, |
|
"eval_runtime": 0.4928, |
|
"eval_samples_per_second": 73.047, |
|
"eval_steps_per_second": 4.058, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 50.857142857142854, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 1.1926276683807373, |
|
"eval_runtime": 0.492, |
|
"eval_samples_per_second": 73.165, |
|
"eval_steps_per_second": 4.065, |
|
"step": 89 |
|
}, |
|
{ |
|
"epoch": 51.42857142857143, |
|
"grad_norm": 6.255433082580566, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.3441, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.2091859579086304, |
|
"eval_runtime": 0.4889, |
|
"eval_samples_per_second": 73.638, |
|
"eval_steps_per_second": 4.091, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 52.57142857142857, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 1.2265750169754028, |
|
"eval_runtime": 0.4977, |
|
"eval_samples_per_second": 72.329, |
|
"eval_steps_per_second": 4.018, |
|
"step": 92 |
|
}, |
|
{ |
|
"epoch": 53.714285714285715, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 1.2684531211853027, |
|
"eval_runtime": 0.4955, |
|
"eval_samples_per_second": 72.652, |
|
"eval_steps_per_second": 4.036, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 54.857142857142854, |
|
"eval_accuracy": 0.4722222222222222, |
|
"eval_loss": 1.3097991943359375, |
|
"eval_runtime": 0.4944, |
|
"eval_samples_per_second": 72.816, |
|
"eval_steps_per_second": 4.045, |
|
"step": 96 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"eval_accuracy": 0.4722222222222222, |
|
"eval_loss": 1.3068737983703613, |
|
"eval_runtime": 0.4935, |
|
"eval_samples_per_second": 72.944, |
|
"eval_steps_per_second": 4.052, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 56.57142857142857, |
|
"eval_accuracy": 0.4722222222222222, |
|
"eval_loss": 1.302114486694336, |
|
"eval_runtime": 0.4883, |
|
"eval_samples_per_second": 73.725, |
|
"eval_steps_per_second": 4.096, |
|
"step": 99 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"grad_norm": 3.261929750442505, |
|
"learning_rate": 0.0, |
|
"loss": 0.322, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.3012207746505737, |
|
"eval_runtime": 0.4985, |
|
"eval_samples_per_second": 72.219, |
|
"eval_steps_per_second": 4.012, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"step": 100, |
|
"total_flos": 8.903293865079828e+17, |
|
"train_loss": 0.6194213128089905, |
|
"train_runtime": 554.626, |
|
"train_samples_per_second": 36.241, |
|
"train_steps_per_second": 0.18 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 1.110924243927002, |
|
"eval_runtime": 0.5431, |
|
"eval_samples_per_second": 66.286, |
|
"eval_steps_per_second": 3.683, |
|
"step": 100 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 100, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 100, |
|
"save_steps": 500, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": true |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 8.903293865079828e+17, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|