|
{ |
|
"best_metric": 0.7222222222222222, |
|
"best_model_checkpoint": "beit-base-patch16-224-hasta-55-fold3/checkpoint-42", |
|
"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": 0.3055555555555556, |
|
"eval_loss": 1.2994396686553955, |
|
"eval_runtime": 0.5048, |
|
"eval_samples_per_second": 71.321, |
|
"eval_steps_per_second": 3.962, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 1.7142857142857144, |
|
"eval_accuracy": 0.3611111111111111, |
|
"eval_loss": 1.065199613571167, |
|
"eval_runtime": 0.5019, |
|
"eval_samples_per_second": 71.728, |
|
"eval_steps_per_second": 3.985, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"eval_accuracy": 0.3611111111111111, |
|
"eval_loss": 1.2855467796325684, |
|
"eval_runtime": 0.5044, |
|
"eval_samples_per_second": 71.379, |
|
"eval_steps_per_second": 3.965, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.4166666666666667, |
|
"eval_loss": 1.088363766670227, |
|
"eval_runtime": 0.5018, |
|
"eval_samples_per_second": 71.74, |
|
"eval_steps_per_second": 3.986, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 4.571428571428571, |
|
"eval_accuracy": 0.4166666666666667, |
|
"eval_loss": 1.0787646770477295, |
|
"eval_runtime": 0.4998, |
|
"eval_samples_per_second": 72.035, |
|
"eval_steps_per_second": 4.002, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"grad_norm": 7.267054080963135, |
|
"learning_rate": 5e-05, |
|
"loss": 1.1909, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 1.0455502271652222, |
|
"eval_runtime": 0.5022, |
|
"eval_samples_per_second": 71.689, |
|
"eval_steps_per_second": 3.983, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 6.857142857142857, |
|
"eval_accuracy": 0.5833333333333334, |
|
"eval_loss": 0.9973945617675781, |
|
"eval_runtime": 0.5012, |
|
"eval_samples_per_second": 71.834, |
|
"eval_steps_per_second": 3.991, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 1.0067553520202637, |
|
"eval_runtime": 0.5199, |
|
"eval_samples_per_second": 69.245, |
|
"eval_steps_per_second": 3.847, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 8.571428571428571, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 0.9822875261306763, |
|
"eval_runtime": 0.5198, |
|
"eval_samples_per_second": 69.26, |
|
"eval_steps_per_second": 3.848, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 9.714285714285714, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 0.9753089547157288, |
|
"eval_runtime": 0.5033, |
|
"eval_samples_per_second": 71.533, |
|
"eval_steps_per_second": 3.974, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 10.857142857142858, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.9455310106277466, |
|
"eval_runtime": 0.5007, |
|
"eval_samples_per_second": 71.901, |
|
"eval_steps_per_second": 3.995, |
|
"step": 19 |
|
}, |
|
{ |
|
"epoch": 11.428571428571429, |
|
"grad_norm": 10.073379516601562, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 0.9469, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.4444444444444444, |
|
"eval_loss": 0.9914661049842834, |
|
"eval_runtime": 0.5004, |
|
"eval_samples_per_second": 71.947, |
|
"eval_steps_per_second": 3.997, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 12.571428571428571, |
|
"eval_accuracy": 0.5, |
|
"eval_loss": 0.9366938471794128, |
|
"eval_runtime": 0.5169, |
|
"eval_samples_per_second": 69.644, |
|
"eval_steps_per_second": 3.869, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 13.714285714285714, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.9859954118728638, |
|
"eval_runtime": 0.5076, |
|
"eval_samples_per_second": 70.924, |
|
"eval_steps_per_second": 3.94, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 14.857142857142858, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.8634018898010254, |
|
"eval_runtime": 0.5186, |
|
"eval_samples_per_second": 69.422, |
|
"eval_steps_per_second": 3.857, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.5277777777777778, |
|
"eval_loss": 0.8749646544456482, |
|
"eval_runtime": 0.5144, |
|
"eval_samples_per_second": 69.988, |
|
"eval_steps_per_second": 3.888, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 16.571428571428573, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.8053224086761475, |
|
"eval_runtime": 0.518, |
|
"eval_samples_per_second": 69.498, |
|
"eval_steps_per_second": 3.861, |
|
"step": 29 |
|
}, |
|
{ |
|
"epoch": 17.142857142857142, |
|
"grad_norm": 5.438823699951172, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.9167, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 17.714285714285715, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.8072555065155029, |
|
"eval_runtime": 0.516, |
|
"eval_samples_per_second": 69.761, |
|
"eval_steps_per_second": 3.876, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 18.857142857142858, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 0.9109321236610413, |
|
"eval_runtime": 0.5245, |
|
"eval_samples_per_second": 68.636, |
|
"eval_steps_per_second": 3.813, |
|
"step": 33 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.4722222222222222, |
|
"eval_loss": 0.9607120156288147, |
|
"eval_runtime": 0.5113, |
|
"eval_samples_per_second": 70.405, |
|
"eval_steps_per_second": 3.911, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 20.571428571428573, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 0.8575677871704102, |
|
"eval_runtime": 0.5347, |
|
"eval_samples_per_second": 67.333, |
|
"eval_steps_per_second": 3.741, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 21.714285714285715, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.7514231204986572, |
|
"eval_runtime": 0.5207, |
|
"eval_samples_per_second": 69.143, |
|
"eval_steps_per_second": 3.841, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"grad_norm": 12.949257850646973, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.8799, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.7742569446563721, |
|
"eval_runtime": 0.5163, |
|
"eval_samples_per_second": 69.728, |
|
"eval_steps_per_second": 3.874, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.7222222222222222, |
|
"eval_loss": 0.7104657292366028, |
|
"eval_runtime": 0.52, |
|
"eval_samples_per_second": 69.234, |
|
"eval_steps_per_second": 3.846, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 24.571428571428573, |
|
"eval_accuracy": 0.6944444444444444, |
|
"eval_loss": 0.6894196271896362, |
|
"eval_runtime": 0.5195, |
|
"eval_samples_per_second": 69.291, |
|
"eval_steps_per_second": 3.849, |
|
"step": 43 |
|
}, |
|
{ |
|
"epoch": 25.714285714285715, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6869931221008301, |
|
"eval_runtime": 0.515, |
|
"eval_samples_per_second": 69.907, |
|
"eval_steps_per_second": 3.884, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 26.857142857142858, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.7546054124832153, |
|
"eval_runtime": 0.5185, |
|
"eval_samples_per_second": 69.427, |
|
"eval_steps_per_second": 3.857, |
|
"step": 47 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.649876058101654, |
|
"eval_runtime": 0.5136, |
|
"eval_samples_per_second": 70.097, |
|
"eval_steps_per_second": 3.894, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"grad_norm": 5.872138023376465, |
|
"learning_rate": 2.777777777777778e-05, |
|
"loss": 0.6763, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"eval_accuracy": 0.6944444444444444, |
|
"eval_loss": 0.6386779546737671, |
|
"eval_runtime": 0.5241, |
|
"eval_samples_per_second": 68.695, |
|
"eval_steps_per_second": 3.816, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 29.714285714285715, |
|
"eval_accuracy": 0.7222222222222222, |
|
"eval_loss": 0.729865550994873, |
|
"eval_runtime": 0.52, |
|
"eval_samples_per_second": 69.228, |
|
"eval_steps_per_second": 3.846, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 30.857142857142858, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.6783073544502258, |
|
"eval_runtime": 0.5204, |
|
"eval_samples_per_second": 69.176, |
|
"eval_steps_per_second": 3.843, |
|
"step": 54 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.7893038988113403, |
|
"eval_runtime": 0.5078, |
|
"eval_samples_per_second": 70.888, |
|
"eval_steps_per_second": 3.938, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 32.57142857142857, |
|
"eval_accuracy": 0.5555555555555556, |
|
"eval_loss": 0.8102678060531616, |
|
"eval_runtime": 0.516, |
|
"eval_samples_per_second": 69.77, |
|
"eval_steps_per_second": 3.876, |
|
"step": 57 |
|
}, |
|
{ |
|
"epoch": 33.714285714285715, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.7723207473754883, |
|
"eval_runtime": 0.5205, |
|
"eval_samples_per_second": 69.158, |
|
"eval_steps_per_second": 3.842, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 34.285714285714285, |
|
"grad_norm": 8.18472957611084, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.593, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 34.857142857142854, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.8221275806427002, |
|
"eval_runtime": 0.5146, |
|
"eval_samples_per_second": 69.952, |
|
"eval_steps_per_second": 3.886, |
|
"step": 61 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6759254336357117, |
|
"eval_runtime": 0.5303, |
|
"eval_samples_per_second": 67.889, |
|
"eval_steps_per_second": 3.772, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 36.57142857142857, |
|
"eval_accuracy": 0.6944444444444444, |
|
"eval_loss": 0.6458982825279236, |
|
"eval_runtime": 0.5151, |
|
"eval_samples_per_second": 69.889, |
|
"eval_steps_per_second": 3.883, |
|
"step": 64 |
|
}, |
|
{ |
|
"epoch": 37.714285714285715, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6828778982162476, |
|
"eval_runtime": 0.5166, |
|
"eval_samples_per_second": 69.693, |
|
"eval_steps_per_second": 3.872, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 38.857142857142854, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6855132579803467, |
|
"eval_runtime": 0.5126, |
|
"eval_samples_per_second": 70.228, |
|
"eval_steps_per_second": 3.902, |
|
"step": 68 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"grad_norm": 7.5129570960998535, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.5137, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6837748289108276, |
|
"eval_runtime": 0.5144, |
|
"eval_samples_per_second": 69.983, |
|
"eval_steps_per_second": 3.888, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 40.57142857142857, |
|
"eval_accuracy": 0.7222222222222222, |
|
"eval_loss": 0.7038333415985107, |
|
"eval_runtime": 0.5406, |
|
"eval_samples_per_second": 66.593, |
|
"eval_steps_per_second": 3.7, |
|
"step": 71 |
|
}, |
|
{ |
|
"epoch": 41.714285714285715, |
|
"eval_accuracy": 0.6944444444444444, |
|
"eval_loss": 0.6592175364494324, |
|
"eval_runtime": 0.5127, |
|
"eval_samples_per_second": 70.222, |
|
"eval_steps_per_second": 3.901, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6781519651412964, |
|
"eval_runtime": 0.5219, |
|
"eval_samples_per_second": 68.977, |
|
"eval_steps_per_second": 3.832, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.7684199810028076, |
|
"eval_runtime": 0.5185, |
|
"eval_samples_per_second": 69.429, |
|
"eval_steps_per_second": 3.857, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 44.57142857142857, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.753241777420044, |
|
"eval_runtime": 0.5164, |
|
"eval_samples_per_second": 69.708, |
|
"eval_steps_per_second": 3.873, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"grad_norm": 8.91373348236084, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.4517, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.7316144108772278, |
|
"eval_runtime": 0.5161, |
|
"eval_samples_per_second": 69.758, |
|
"eval_steps_per_second": 3.875, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 46.857142857142854, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6814126372337341, |
|
"eval_runtime": 0.5191, |
|
"eval_samples_per_second": 69.353, |
|
"eval_steps_per_second": 3.853, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.6642955541610718, |
|
"eval_runtime": 0.5218, |
|
"eval_samples_per_second": 68.994, |
|
"eval_steps_per_second": 3.833, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 48.57142857142857, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6613295078277588, |
|
"eval_runtime": 0.518, |
|
"eval_samples_per_second": 69.496, |
|
"eval_steps_per_second": 3.861, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 49.714285714285715, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6539488434791565, |
|
"eval_runtime": 0.5215, |
|
"eval_samples_per_second": 69.028, |
|
"eval_steps_per_second": 3.835, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 50.857142857142854, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6436237692832947, |
|
"eval_runtime": 0.5149, |
|
"eval_samples_per_second": 69.92, |
|
"eval_steps_per_second": 3.884, |
|
"step": 89 |
|
}, |
|
{ |
|
"epoch": 51.42857142857143, |
|
"grad_norm": 8.68427562713623, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.4126, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.6380901336669922, |
|
"eval_runtime": 0.5244, |
|
"eval_samples_per_second": 68.652, |
|
"eval_steps_per_second": 3.814, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 52.57142857142857, |
|
"eval_accuracy": 0.6666666666666666, |
|
"eval_loss": 0.636466383934021, |
|
"eval_runtime": 0.5312, |
|
"eval_samples_per_second": 67.768, |
|
"eval_steps_per_second": 3.765, |
|
"step": 92 |
|
}, |
|
{ |
|
"epoch": 53.714285714285715, |
|
"eval_accuracy": 0.6388888888888888, |
|
"eval_loss": 0.6374828815460205, |
|
"eval_runtime": 0.5171, |
|
"eval_samples_per_second": 69.623, |
|
"eval_steps_per_second": 3.868, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 54.857142857142854, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.6452690958976746, |
|
"eval_runtime": 0.5129, |
|
"eval_samples_per_second": 70.185, |
|
"eval_steps_per_second": 3.899, |
|
"step": 96 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.648621678352356, |
|
"eval_runtime": 0.525, |
|
"eval_samples_per_second": 68.571, |
|
"eval_steps_per_second": 3.809, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 56.57142857142857, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.648608922958374, |
|
"eval_runtime": 0.5121, |
|
"eval_samples_per_second": 70.297, |
|
"eval_steps_per_second": 3.905, |
|
"step": 99 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"grad_norm": 8.446288108825684, |
|
"learning_rate": 0.0, |
|
"loss": 0.3715, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"eval_accuracy": 0.6111111111111112, |
|
"eval_loss": 0.6486251354217529, |
|
"eval_runtime": 0.5197, |
|
"eval_samples_per_second": 69.265, |
|
"eval_steps_per_second": 3.848, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"step": 100, |
|
"total_flos": 8.766873360020644e+17, |
|
"train_loss": 0.6953265452384949, |
|
"train_runtime": 563.1349, |
|
"train_samples_per_second": 35.16, |
|
"train_steps_per_second": 0.178 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"eval_accuracy": 0.7222222222222222, |
|
"eval_loss": 0.7104657292366028, |
|
"eval_runtime": 0.5899, |
|
"eval_samples_per_second": 61.025, |
|
"eval_steps_per_second": 3.39, |
|
"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": 8.766873360020644e+17, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|