| { | |
| "best_metric": 0.7130943536758423, | |
| "best_model_checkpoint": "/media/mldrive/kcardenas/limb_classification_person_crop/beit-large-patch16-384/8_5e-5_1e-4_0.05/checkpoint-1712", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2140, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11682242990654206, | |
| "grad_norm": 1340459.625, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.5852, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2336448598130841, | |
| "grad_norm": 994590.0, | |
| "learning_rate": 5e-06, | |
| "loss": 1.5062, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35046728971962615, | |
| "grad_norm": 835627.5, | |
| "learning_rate": 7.5e-06, | |
| "loss": 1.4702, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4672897196261682, | |
| "grad_norm": 708100.8125, | |
| "learning_rate": 1e-05, | |
| "loss": 1.405, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5841121495327103, | |
| "grad_norm": 19124280.0, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.3983, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7009345794392523, | |
| "grad_norm": 281499.28125, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.3889, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8177570093457944, | |
| "grad_norm": 727612.8125, | |
| "learning_rate": 1.75e-05, | |
| "loss": 1.3625, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9345794392523364, | |
| "grad_norm": 628275.8125, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3649, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.36401326699834163, | |
| "eval_loss": 1.5214860439300537, | |
| "eval_runtime": 15.5708, | |
| "eval_samples_per_second": 77.453, | |
| "eval_steps_per_second": 2.44, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0514018691588785, | |
| "grad_norm": 456936.21875, | |
| "learning_rate": 2.25e-05, | |
| "loss": 1.3042, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.1682242990654206, | |
| "grad_norm": 314467.59375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.275, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2850467289719627, | |
| "grad_norm": 409308.0625, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 1.2176, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.4018691588785046, | |
| "grad_norm": 343987.34375, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2691, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5186915887850467, | |
| "grad_norm": 415037.3125, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 1.2335, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.6355140186915889, | |
| "grad_norm": 369134.59375, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.2058, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.7523364485981308, | |
| "grad_norm": 246889.234375, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.2045, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.8691588785046729, | |
| "grad_norm": 388048.875, | |
| "learning_rate": 4e-05, | |
| "loss": 1.143, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.985981308411215, | |
| "grad_norm": 442224.96875, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.1297, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6003316749585407, | |
| "eval_loss": 1.0013911724090576, | |
| "eval_runtime": 15.593, | |
| "eval_samples_per_second": 77.343, | |
| "eval_steps_per_second": 2.437, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.102803738317757, | |
| "grad_norm": 518648.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.1438, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.2196261682242993, | |
| "grad_norm": 496547.5625, | |
| "learning_rate": 4.75e-05, | |
| "loss": 1.1246, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.336448598130841, | |
| "grad_norm": 306726.9375, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1061, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.453271028037383, | |
| "grad_norm": 327747.03125, | |
| "learning_rate": 4.923780487804878e-05, | |
| "loss": 1.1146, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.5700934579439254, | |
| "grad_norm": 250996.6875, | |
| "learning_rate": 4.847560975609756e-05, | |
| "loss": 1.0954, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.6869158878504673, | |
| "grad_norm": 292352.34375, | |
| "learning_rate": 4.771341463414634e-05, | |
| "loss": 1.129, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.803738317757009, | |
| "grad_norm": 299764.5, | |
| "learning_rate": 4.695121951219512e-05, | |
| "loss": 1.0878, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.9205607476635516, | |
| "grad_norm": 263247.34375, | |
| "learning_rate": 4.618902439024391e-05, | |
| "loss": 1.0881, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6558872305140961, | |
| "eval_loss": 0.9018123745918274, | |
| "eval_runtime": 15.6284, | |
| "eval_samples_per_second": 77.167, | |
| "eval_steps_per_second": 2.431, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0373831775700935, | |
| "grad_norm": 281239.8125, | |
| "learning_rate": 4.542682926829269e-05, | |
| "loss": 1.094, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.1542056074766354, | |
| "grad_norm": 263740.625, | |
| "learning_rate": 4.466463414634147e-05, | |
| "loss": 1.036, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.2710280373831777, | |
| "grad_norm": 345132.46875, | |
| "learning_rate": 4.390243902439025e-05, | |
| "loss": 1.0542, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.3878504672897196, | |
| "grad_norm": 281706.03125, | |
| "learning_rate": 4.314024390243903e-05, | |
| "loss": 1.0712, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 3.5046728971962615, | |
| "grad_norm": 256964.125, | |
| "learning_rate": 4.237804878048781e-05, | |
| "loss": 1.0417, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.621495327102804, | |
| "grad_norm": 426164.21875, | |
| "learning_rate": 4.161585365853659e-05, | |
| "loss": 1.0363, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.7383177570093458, | |
| "grad_norm": 366963.625, | |
| "learning_rate": 4.085365853658537e-05, | |
| "loss": 1.0382, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.8551401869158877, | |
| "grad_norm": 314130.34375, | |
| "learning_rate": 4.0091463414634153e-05, | |
| "loss": 1.0255, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.97196261682243, | |
| "grad_norm": 343246.59375, | |
| "learning_rate": 3.932926829268293e-05, | |
| "loss": 1.0065, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.599502487562189, | |
| "eval_loss": 0.9687954187393188, | |
| "eval_runtime": 15.5027, | |
| "eval_samples_per_second": 77.793, | |
| "eval_steps_per_second": 2.451, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.088785046728972, | |
| "grad_norm": 355268.8125, | |
| "learning_rate": 3.856707317073171e-05, | |
| "loss": 0.9995, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 4.205607476635514, | |
| "grad_norm": 266058.53125, | |
| "learning_rate": 3.780487804878049e-05, | |
| "loss": 0.9972, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.322429906542056, | |
| "grad_norm": 447425.875, | |
| "learning_rate": 3.704268292682927e-05, | |
| "loss": 1.004, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 4.4392523364485985, | |
| "grad_norm": 382122.4375, | |
| "learning_rate": 3.628048780487805e-05, | |
| "loss": 0.967, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.55607476635514, | |
| "grad_norm": 379657.3125, | |
| "learning_rate": 3.551829268292683e-05, | |
| "loss": 0.9769, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 4.672897196261682, | |
| "grad_norm": 281210.5625, | |
| "learning_rate": 3.475609756097561e-05, | |
| "loss": 0.9521, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.789719626168225, | |
| "grad_norm": 337339.375, | |
| "learning_rate": 3.399390243902439e-05, | |
| "loss": 0.9654, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.906542056074766, | |
| "grad_norm": 227255.0625, | |
| "learning_rate": 3.323170731707317e-05, | |
| "loss": 1.0028, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7014925373134329, | |
| "eval_loss": 0.8239747285842896, | |
| "eval_runtime": 15.7023, | |
| "eval_samples_per_second": 76.804, | |
| "eval_steps_per_second": 2.42, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.0233644859813085, | |
| "grad_norm": 396188.9375, | |
| "learning_rate": 3.246951219512195e-05, | |
| "loss": 1.002, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 5.140186915887851, | |
| "grad_norm": 331960.46875, | |
| "learning_rate": 3.170731707317073e-05, | |
| "loss": 0.9421, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.257009345794392, | |
| "grad_norm": 390978.8125, | |
| "learning_rate": 3.094512195121951e-05, | |
| "loss": 0.9338, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 5.373831775700935, | |
| "grad_norm": 321306.375, | |
| "learning_rate": 3.0182926829268294e-05, | |
| "loss": 0.9425, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.490654205607477, | |
| "grad_norm": 363200.09375, | |
| "learning_rate": 2.9420731707317074e-05, | |
| "loss": 0.9054, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 5.607476635514018, | |
| "grad_norm": 315464.1875, | |
| "learning_rate": 2.8658536585365854e-05, | |
| "loss": 0.9164, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.724299065420561, | |
| "grad_norm": 257387.84375, | |
| "learning_rate": 2.7896341463414637e-05, | |
| "loss": 0.9579, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 5.841121495327103, | |
| "grad_norm": 268010.53125, | |
| "learning_rate": 2.7134146341463417e-05, | |
| "loss": 0.9108, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.957943925233645, | |
| "grad_norm": 313754.125, | |
| "learning_rate": 2.6371951219512197e-05, | |
| "loss": 0.9225, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7520729684908789, | |
| "eval_loss": 0.735476553440094, | |
| "eval_runtime": 15.6884, | |
| "eval_samples_per_second": 76.872, | |
| "eval_steps_per_second": 2.422, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 6.074766355140187, | |
| "grad_norm": 319647.28125, | |
| "learning_rate": 2.5609756097560977e-05, | |
| "loss": 0.8836, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.191588785046729, | |
| "grad_norm": 440299.0, | |
| "learning_rate": 2.4847560975609756e-05, | |
| "loss": 0.8731, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 6.308411214953271, | |
| "grad_norm": 294800.8125, | |
| "learning_rate": 2.4085365853658536e-05, | |
| "loss": 0.8843, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 6.425233644859813, | |
| "grad_norm": 390740.28125, | |
| "learning_rate": 2.332317073170732e-05, | |
| "loss": 0.8199, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 6.542056074766355, | |
| "grad_norm": 393681.1875, | |
| "learning_rate": 2.25609756097561e-05, | |
| "loss": 0.8497, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.658878504672897, | |
| "grad_norm": 378513.21875, | |
| "learning_rate": 2.179878048780488e-05, | |
| "loss": 0.8445, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 6.775700934579439, | |
| "grad_norm": 330039.0625, | |
| "learning_rate": 2.103658536585366e-05, | |
| "loss": 0.8562, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 6.892523364485982, | |
| "grad_norm": 318586.625, | |
| "learning_rate": 2.0274390243902442e-05, | |
| "loss": 0.8522, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.746268656716418, | |
| "eval_loss": 0.76929771900177, | |
| "eval_runtime": 15.5157, | |
| "eval_samples_per_second": 77.728, | |
| "eval_steps_per_second": 2.449, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 7.009345794392523, | |
| "grad_norm": 331889.5, | |
| "learning_rate": 1.9512195121951222e-05, | |
| "loss": 0.8529, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.126168224299065, | |
| "grad_norm": 381441.40625, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.7812, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 7.242990654205608, | |
| "grad_norm": 248858.21875, | |
| "learning_rate": 1.798780487804878e-05, | |
| "loss": 0.8792, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 7.359813084112149, | |
| "grad_norm": 339380.375, | |
| "learning_rate": 1.722560975609756e-05, | |
| "loss": 0.8191, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 7.4766355140186915, | |
| "grad_norm": 363421.21875, | |
| "learning_rate": 1.6463414634146345e-05, | |
| "loss": 0.7828, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.593457943925234, | |
| "grad_norm": 332095.78125, | |
| "learning_rate": 1.5701219512195124e-05, | |
| "loss": 0.8328, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 7.710280373831775, | |
| "grad_norm": 267183.625, | |
| "learning_rate": 1.4939024390243902e-05, | |
| "loss": 0.8255, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 7.827102803738318, | |
| "grad_norm": 463742.25, | |
| "learning_rate": 1.4176829268292682e-05, | |
| "loss": 0.792, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 7.94392523364486, | |
| "grad_norm": 254509.796875, | |
| "learning_rate": 1.3414634146341466e-05, | |
| "loss": 0.821, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7678275290215588, | |
| "eval_loss": 0.7130943536758423, | |
| "eval_runtime": 15.4772, | |
| "eval_samples_per_second": 77.921, | |
| "eval_steps_per_second": 2.455, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 8.060747663551401, | |
| "grad_norm": 432679.84375, | |
| "learning_rate": 1.2652439024390245e-05, | |
| "loss": 0.7849, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 8.177570093457945, | |
| "grad_norm": 361022.125, | |
| "learning_rate": 1.1890243902439025e-05, | |
| "loss": 0.7447, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 8.294392523364486, | |
| "grad_norm": 416379.84375, | |
| "learning_rate": 1.1128048780487805e-05, | |
| "loss": 0.7898, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 8.411214953271028, | |
| "grad_norm": 386266.90625, | |
| "learning_rate": 1.0365853658536585e-05, | |
| "loss": 0.742, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.52803738317757, | |
| "grad_norm": 321598.21875, | |
| "learning_rate": 9.603658536585366e-06, | |
| "loss": 0.7681, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 8.644859813084112, | |
| "grad_norm": 386202.53125, | |
| "learning_rate": 8.841463414634146e-06, | |
| "loss": 0.7725, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 8.761682242990654, | |
| "grad_norm": 319017.71875, | |
| "learning_rate": 8.079268292682928e-06, | |
| "loss": 0.7394, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 8.878504672897197, | |
| "grad_norm": 388873.65625, | |
| "learning_rate": 7.317073170731707e-06, | |
| "loss": 0.7435, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.995327102803738, | |
| "grad_norm": 302018.6875, | |
| "learning_rate": 6.554878048780488e-06, | |
| "loss": 0.735, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7761194029850746, | |
| "eval_loss": 0.7315632104873657, | |
| "eval_runtime": 15.8083, | |
| "eval_samples_per_second": 76.289, | |
| "eval_steps_per_second": 2.404, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 9.11214953271028, | |
| "grad_norm": 400649.4375, | |
| "learning_rate": 5.792682926829269e-06, | |
| "loss": 0.6635, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 9.228971962616823, | |
| "grad_norm": 330898.75, | |
| "learning_rate": 5.030487804878049e-06, | |
| "loss": 0.7556, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 9.345794392523365, | |
| "grad_norm": 298238.21875, | |
| "learning_rate": 4.26829268292683e-06, | |
| "loss": 0.6982, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.462616822429906, | |
| "grad_norm": 244765.453125, | |
| "learning_rate": 3.5060975609756102e-06, | |
| "loss": 0.657, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 9.57943925233645, | |
| "grad_norm": 357814.375, | |
| "learning_rate": 2.7439024390243905e-06, | |
| "loss": 0.7641, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 9.69626168224299, | |
| "grad_norm": 500304.03125, | |
| "learning_rate": 1.9817073170731707e-06, | |
| "loss": 0.7561, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 9.813084112149532, | |
| "grad_norm": 457811.4375, | |
| "learning_rate": 1.2195121951219514e-06, | |
| "loss": 0.7183, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.929906542056075, | |
| "grad_norm": 392788.8125, | |
| "learning_rate": 4.573170731707317e-07, | |
| "loss": 0.7123, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 0.7300788760185242, | |
| "eval_runtime": 15.6127, | |
| "eval_samples_per_second": 77.245, | |
| "eval_steps_per_second": 2.434, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 2140, | |
| "total_flos": 5.510586115727032e+19, | |
| "train_loss": 0.9844673860852964, | |
| "train_runtime": 1946.7718, | |
| "train_samples_per_second": 35.084, | |
| "train_steps_per_second": 1.099 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 5.510586115727032e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |