|
{ |
|
"best_metric": 1.0969369411468506, |
|
"best_model_checkpoint": "./vit-base-melSpecImagesCREMA/checkpoint-400", |
|
"epoch": 10.0, |
|
"eval_steps": 100, |
|
"global_step": 1310, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 1.503008246421814, |
|
"learning_rate": 0.0001984732824427481, |
|
"loss": 1.7751, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 1.1629976034164429, |
|
"learning_rate": 0.0001969465648854962, |
|
"loss": 1.6558, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 1.205520749092102, |
|
"learning_rate": 0.00019541984732824428, |
|
"loss": 1.6062, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 1.4008492231369019, |
|
"learning_rate": 0.00019389312977099237, |
|
"loss": 1.6114, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 1.3961032629013062, |
|
"learning_rate": 0.00019236641221374049, |
|
"loss": 1.5886, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 1.31687331199646, |
|
"learning_rate": 0.00019083969465648857, |
|
"loss": 1.544, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 1.3430259227752686, |
|
"learning_rate": 0.00018931297709923666, |
|
"loss": 1.493, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 2.1518867015838623, |
|
"learning_rate": 0.00018778625954198475, |
|
"loss": 1.5494, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 1.066909909248352, |
|
"learning_rate": 0.00018625954198473284, |
|
"loss": 1.4324, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 1.1103651523590088, |
|
"learning_rate": 0.00018473282442748093, |
|
"loss": 1.5606, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"eval_accuracy": 0.40786948176583493, |
|
"eval_loss": 1.4423701763153076, |
|
"eval_runtime": 10.9463, |
|
"eval_samples_per_second": 95.192, |
|
"eval_steps_per_second": 11.967, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 1.20200777053833, |
|
"learning_rate": 0.00018320610687022902, |
|
"loss": 1.429, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 1.1876991987228394, |
|
"learning_rate": 0.0001816793893129771, |
|
"loss": 1.4265, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 1.1246036291122437, |
|
"learning_rate": 0.00018015267175572518, |
|
"loss": 1.4526, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 1.8578764200210571, |
|
"learning_rate": 0.0001786259541984733, |
|
"loss": 1.349, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 1.15, |
|
"grad_norm": 1.8486779928207397, |
|
"learning_rate": 0.00017709923664122138, |
|
"loss": 1.3172, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"grad_norm": 1.775232195854187, |
|
"learning_rate": 0.00017557251908396947, |
|
"loss": 1.2735, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 1.3, |
|
"grad_norm": 1.7306294441223145, |
|
"learning_rate": 0.00017404580152671756, |
|
"loss": 1.3665, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 1.37, |
|
"grad_norm": 1.8813070058822632, |
|
"learning_rate": 0.00017251908396946565, |
|
"loss": 1.322, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 1.45, |
|
"grad_norm": 1.4844669103622437, |
|
"learning_rate": 0.00017099236641221374, |
|
"loss": 1.324, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"grad_norm": 2.489121913909912, |
|
"learning_rate": 0.00016946564885496183, |
|
"loss": 1.2841, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"eval_accuracy": 0.3694817658349328, |
|
"eval_loss": 1.498081088066101, |
|
"eval_runtime": 11.2582, |
|
"eval_samples_per_second": 92.555, |
|
"eval_steps_per_second": 11.636, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"grad_norm": 1.9091380834579468, |
|
"learning_rate": 0.00016793893129770992, |
|
"loss": 1.2919, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 1.68, |
|
"grad_norm": 1.8453233242034912, |
|
"learning_rate": 0.00016641221374045804, |
|
"loss": 1.2998, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 1.76, |
|
"grad_norm": 2.699350595474243, |
|
"learning_rate": 0.00016488549618320613, |
|
"loss": 1.3153, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 1.83, |
|
"grad_norm": 2.409592628479004, |
|
"learning_rate": 0.00016335877862595422, |
|
"loss": 1.1943, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 1.91, |
|
"grad_norm": 1.59845769405365, |
|
"learning_rate": 0.0001618320610687023, |
|
"loss": 1.2302, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.98, |
|
"grad_norm": 2.2508938312530518, |
|
"learning_rate": 0.0001603053435114504, |
|
"loss": 1.4177, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 2.06, |
|
"grad_norm": 1.6891452074050903, |
|
"learning_rate": 0.00015877862595419848, |
|
"loss": 1.0828, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 2.14, |
|
"grad_norm": 3.317949056625366, |
|
"learning_rate": 0.00015725190839694657, |
|
"loss": 1.1312, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 2.21, |
|
"grad_norm": 3.0603318214416504, |
|
"learning_rate": 0.00015572519083969466, |
|
"loss": 1.0713, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 2.29, |
|
"grad_norm": 2.0383269786834717, |
|
"learning_rate": 0.00015419847328244275, |
|
"loss": 1.0159, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 2.29, |
|
"eval_accuracy": 0.5518234165067178, |
|
"eval_loss": 1.1692744493484497, |
|
"eval_runtime": 11.2045, |
|
"eval_samples_per_second": 92.998, |
|
"eval_steps_per_second": 11.692, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 2.37, |
|
"grad_norm": 1.9367152452468872, |
|
"learning_rate": 0.00015267175572519084, |
|
"loss": 0.9986, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"grad_norm": 2.3003132343292236, |
|
"learning_rate": 0.00015114503816793893, |
|
"loss": 1.0508, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 2.52, |
|
"grad_norm": 3.84384822845459, |
|
"learning_rate": 0.00014961832061068702, |
|
"loss": 1.0623, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"grad_norm": 1.7760237455368042, |
|
"learning_rate": 0.0001480916030534351, |
|
"loss": 1.0425, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"grad_norm": 2.6222035884857178, |
|
"learning_rate": 0.0001465648854961832, |
|
"loss": 1.0366, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 2.75, |
|
"grad_norm": 2.4612748622894287, |
|
"learning_rate": 0.0001450381679389313, |
|
"loss": 1.0451, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 2.82, |
|
"grad_norm": 2.3277204036712646, |
|
"learning_rate": 0.00014351145038167938, |
|
"loss": 1.1176, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 2.9, |
|
"grad_norm": 2.187209129333496, |
|
"learning_rate": 0.00014198473282442747, |
|
"loss": 0.9908, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 2.98, |
|
"grad_norm": 3.229308843612671, |
|
"learning_rate": 0.00014045801526717556, |
|
"loss": 1.0181, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 3.05, |
|
"grad_norm": 1.847447156906128, |
|
"learning_rate": 0.00013893129770992368, |
|
"loss": 0.9868, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 3.05, |
|
"eval_accuracy": 0.5930902111324377, |
|
"eval_loss": 1.0969369411468506, |
|
"eval_runtime": 11.1277, |
|
"eval_samples_per_second": 93.64, |
|
"eval_steps_per_second": 11.772, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 3.13, |
|
"grad_norm": 2.6871042251586914, |
|
"learning_rate": 0.00013740458015267177, |
|
"loss": 0.8034, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 3.21, |
|
"grad_norm": 3.8769664764404297, |
|
"learning_rate": 0.00013587786259541986, |
|
"loss": 0.8141, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 3.28, |
|
"grad_norm": 2.8785126209259033, |
|
"learning_rate": 0.00013435114503816795, |
|
"loss": 0.7343, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 3.36, |
|
"grad_norm": 2.4167301654815674, |
|
"learning_rate": 0.00013282442748091604, |
|
"loss": 0.8251, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 3.44, |
|
"grad_norm": 2.921082019805908, |
|
"learning_rate": 0.00013129770992366413, |
|
"loss": 0.7511, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 3.51, |
|
"grad_norm": 2.847332000732422, |
|
"learning_rate": 0.00012977099236641222, |
|
"loss": 0.7392, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 3.59, |
|
"grad_norm": 3.5761501789093018, |
|
"learning_rate": 0.0001282442748091603, |
|
"loss": 0.6463, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 3.66, |
|
"grad_norm": 3.0666024684906006, |
|
"learning_rate": 0.0001267175572519084, |
|
"loss": 0.7027, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 3.74, |
|
"grad_norm": 1.9951330423355103, |
|
"learning_rate": 0.0001251908396946565, |
|
"loss": 0.7666, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 3.82, |
|
"grad_norm": 3.5129895210266113, |
|
"learning_rate": 0.0001236641221374046, |
|
"loss": 0.8477, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 3.82, |
|
"eval_accuracy": 0.5796545105566219, |
|
"eval_loss": 1.1718865633010864, |
|
"eval_runtime": 11.143, |
|
"eval_samples_per_second": 93.511, |
|
"eval_steps_per_second": 11.756, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 3.89, |
|
"grad_norm": 2.490349054336548, |
|
"learning_rate": 0.0001221374045801527, |
|
"loss": 0.6279, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 3.97, |
|
"grad_norm": 4.751981258392334, |
|
"learning_rate": 0.00012061068702290077, |
|
"loss": 0.6861, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"grad_norm": 1.843367338180542, |
|
"learning_rate": 0.00011908396946564886, |
|
"loss": 0.6348, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 4.12, |
|
"grad_norm": 3.823432445526123, |
|
"learning_rate": 0.00011755725190839695, |
|
"loss": 0.4487, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 4.2, |
|
"grad_norm": 1.805336833000183, |
|
"learning_rate": 0.00011603053435114504, |
|
"loss": 0.4393, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 4.27, |
|
"grad_norm": 2.604633331298828, |
|
"learning_rate": 0.00011450381679389313, |
|
"loss": 0.4098, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 4.35, |
|
"grad_norm": 3.3479952812194824, |
|
"learning_rate": 0.00011297709923664124, |
|
"loss": 0.5057, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 4.43, |
|
"grad_norm": 2.961186408996582, |
|
"learning_rate": 0.00011145038167938933, |
|
"loss": 0.5446, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 4.5, |
|
"grad_norm": 4.528743267059326, |
|
"learning_rate": 0.00010992366412213742, |
|
"loss": 0.5032, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 4.58, |
|
"grad_norm": 3.593770742416382, |
|
"learning_rate": 0.0001083969465648855, |
|
"loss": 0.5495, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 4.58, |
|
"eval_accuracy": 0.5806142034548945, |
|
"eval_loss": 1.234790563583374, |
|
"eval_runtime": 11.1508, |
|
"eval_samples_per_second": 93.446, |
|
"eval_steps_per_second": 11.748, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 4.66, |
|
"grad_norm": 2.7948455810546875, |
|
"learning_rate": 0.00010687022900763359, |
|
"loss": 0.519, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 4.73, |
|
"grad_norm": 5.143344879150391, |
|
"learning_rate": 0.00010534351145038168, |
|
"loss": 0.5493, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 4.81, |
|
"grad_norm": 4.292489528656006, |
|
"learning_rate": 0.00010381679389312977, |
|
"loss": 0.5947, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 4.89, |
|
"grad_norm": 3.135993480682373, |
|
"learning_rate": 0.00010229007633587786, |
|
"loss": 0.4873, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 4.96, |
|
"grad_norm": 2.5363755226135254, |
|
"learning_rate": 0.00010076335877862595, |
|
"loss": 0.4813, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 5.04, |
|
"grad_norm": 2.4878172874450684, |
|
"learning_rate": 9.923664122137405e-05, |
|
"loss": 0.4249, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 5.11, |
|
"grad_norm": 2.6882851123809814, |
|
"learning_rate": 9.770992366412214e-05, |
|
"loss": 0.2691, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 5.19, |
|
"grad_norm": 2.534364700317383, |
|
"learning_rate": 9.618320610687024e-05, |
|
"loss": 0.3055, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 5.27, |
|
"grad_norm": 4.340401649475098, |
|
"learning_rate": 9.465648854961833e-05, |
|
"loss": 0.2444, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 5.34, |
|
"grad_norm": 5.553411483764648, |
|
"learning_rate": 9.312977099236642e-05, |
|
"loss": 0.2671, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 5.34, |
|
"eval_accuracy": 0.5854126679462572, |
|
"eval_loss": 1.3457223176956177, |
|
"eval_runtime": 11.154, |
|
"eval_samples_per_second": 93.419, |
|
"eval_steps_per_second": 11.745, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 5.42, |
|
"grad_norm": 2.335097551345825, |
|
"learning_rate": 9.160305343511451e-05, |
|
"loss": 0.3027, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 5.5, |
|
"grad_norm": 6.312303066253662, |
|
"learning_rate": 9.007633587786259e-05, |
|
"loss": 0.3205, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 5.57, |
|
"grad_norm": 7.102092266082764, |
|
"learning_rate": 8.854961832061069e-05, |
|
"loss": 0.2796, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 5.65, |
|
"grad_norm": 4.718852996826172, |
|
"learning_rate": 8.702290076335878e-05, |
|
"loss": 0.3254, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 5.73, |
|
"grad_norm": 5.368557453155518, |
|
"learning_rate": 8.549618320610687e-05, |
|
"loss": 0.329, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 5.8, |
|
"grad_norm": 1.7768850326538086, |
|
"learning_rate": 8.396946564885496e-05, |
|
"loss": 0.2239, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 5.88, |
|
"grad_norm": 1.6067969799041748, |
|
"learning_rate": 8.244274809160306e-05, |
|
"loss": 0.2289, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 5.95, |
|
"grad_norm": 0.8688230514526367, |
|
"learning_rate": 8.091603053435115e-05, |
|
"loss": 0.2715, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 6.03, |
|
"grad_norm": 2.5685417652130127, |
|
"learning_rate": 7.938931297709924e-05, |
|
"loss": 0.2098, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 6.11, |
|
"grad_norm": 1.9992344379425049, |
|
"learning_rate": 7.786259541984733e-05, |
|
"loss": 0.1388, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 6.11, |
|
"eval_accuracy": 0.5786948176583493, |
|
"eval_loss": 1.389073371887207, |
|
"eval_runtime": 11.1211, |
|
"eval_samples_per_second": 93.696, |
|
"eval_steps_per_second": 11.779, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 6.18, |
|
"grad_norm": 2.335876703262329, |
|
"learning_rate": 7.633587786259542e-05, |
|
"loss": 0.1762, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 6.26, |
|
"grad_norm": 4.345489025115967, |
|
"learning_rate": 7.480916030534351e-05, |
|
"loss": 0.1829, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 6.34, |
|
"grad_norm": 1.5818490982055664, |
|
"learning_rate": 7.32824427480916e-05, |
|
"loss": 0.2543, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 6.41, |
|
"grad_norm": 2.3892974853515625, |
|
"learning_rate": 7.175572519083969e-05, |
|
"loss": 0.2035, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 6.49, |
|
"grad_norm": 0.9355500936508179, |
|
"learning_rate": 7.022900763358778e-05, |
|
"loss": 0.2287, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 6.56, |
|
"grad_norm": 3.3484275341033936, |
|
"learning_rate": 6.870229007633588e-05, |
|
"loss": 0.1433, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 6.64, |
|
"grad_norm": 1.1796901226043701, |
|
"learning_rate": 6.717557251908397e-05, |
|
"loss": 0.1943, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 6.72, |
|
"grad_norm": 3.2419047355651855, |
|
"learning_rate": 6.564885496183206e-05, |
|
"loss": 0.2079, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 6.79, |
|
"grad_norm": 5.841382026672363, |
|
"learning_rate": 6.412213740458015e-05, |
|
"loss": 0.2108, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 6.87, |
|
"grad_norm": 1.1594499349594116, |
|
"learning_rate": 6.259541984732826e-05, |
|
"loss": 0.1548, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 6.87, |
|
"eval_accuracy": 0.5978886756238004, |
|
"eval_loss": 1.4216477870941162, |
|
"eval_runtime": 11.0828, |
|
"eval_samples_per_second": 94.019, |
|
"eval_steps_per_second": 11.82, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 6.95, |
|
"grad_norm": 1.131686806678772, |
|
"learning_rate": 6.106870229007635e-05, |
|
"loss": 0.1264, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 7.02, |
|
"grad_norm": 0.8231382966041565, |
|
"learning_rate": 5.954198473282443e-05, |
|
"loss": 0.1726, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 7.1, |
|
"grad_norm": 0.48267778754234314, |
|
"learning_rate": 5.801526717557252e-05, |
|
"loss": 0.1198, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 7.18, |
|
"grad_norm": 2.5278027057647705, |
|
"learning_rate": 5.648854961832062e-05, |
|
"loss": 0.1702, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 7.25, |
|
"grad_norm": 2.183692455291748, |
|
"learning_rate": 5.496183206106871e-05, |
|
"loss": 0.0854, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 7.33, |
|
"grad_norm": 0.40628916025161743, |
|
"learning_rate": 5.3435114503816794e-05, |
|
"loss": 0.1078, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 7.4, |
|
"grad_norm": 0.19366604089736938, |
|
"learning_rate": 5.1908396946564884e-05, |
|
"loss": 0.0979, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 7.48, |
|
"grad_norm": 0.4136432707309723, |
|
"learning_rate": 5.038167938931297e-05, |
|
"loss": 0.0954, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 7.56, |
|
"grad_norm": 0.17071671783924103, |
|
"learning_rate": 4.885496183206107e-05, |
|
"loss": 0.0432, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 7.63, |
|
"grad_norm": 0.9517059326171875, |
|
"learning_rate": 4.7328244274809166e-05, |
|
"loss": 0.0906, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 7.63, |
|
"eval_accuracy": 0.564299424184261, |
|
"eval_loss": 1.640116810798645, |
|
"eval_runtime": 11.0958, |
|
"eval_samples_per_second": 93.91, |
|
"eval_steps_per_second": 11.806, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 7.71, |
|
"grad_norm": 0.1622392237186432, |
|
"learning_rate": 4.5801526717557256e-05, |
|
"loss": 0.1284, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 7.79, |
|
"grad_norm": 6.343606948852539, |
|
"learning_rate": 4.4274809160305345e-05, |
|
"loss": 0.0985, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 7.86, |
|
"grad_norm": 0.7874533534049988, |
|
"learning_rate": 4.2748091603053435e-05, |
|
"loss": 0.1168, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 7.94, |
|
"grad_norm": 0.31581443548202515, |
|
"learning_rate": 4.122137404580153e-05, |
|
"loss": 0.1668, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 8.02, |
|
"grad_norm": 0.3773091435432434, |
|
"learning_rate": 3.969465648854962e-05, |
|
"loss": 0.1027, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 8.09, |
|
"grad_norm": 0.09075737744569778, |
|
"learning_rate": 3.816793893129771e-05, |
|
"loss": 0.0301, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 8.17, |
|
"grad_norm": 0.1646617352962494, |
|
"learning_rate": 3.66412213740458e-05, |
|
"loss": 0.077, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 8.24, |
|
"grad_norm": 1.9873558282852173, |
|
"learning_rate": 3.511450381679389e-05, |
|
"loss": 0.0443, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 8.32, |
|
"grad_norm": 0.13392330706119537, |
|
"learning_rate": 3.358778625954199e-05, |
|
"loss": 0.0823, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"grad_norm": 1.0865620374679565, |
|
"learning_rate": 3.2061068702290076e-05, |
|
"loss": 0.1047, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"eval_accuracy": 0.5873320537428023, |
|
"eval_loss": 1.6780017614364624, |
|
"eval_runtime": 11.1335, |
|
"eval_samples_per_second": 93.591, |
|
"eval_steps_per_second": 11.766, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 8.47, |
|
"grad_norm": 0.10025681555271149, |
|
"learning_rate": 3.053435114503817e-05, |
|
"loss": 0.0592, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 8.55, |
|
"grad_norm": 0.14586412906646729, |
|
"learning_rate": 2.900763358778626e-05, |
|
"loss": 0.0393, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 8.63, |
|
"grad_norm": 0.5182372331619263, |
|
"learning_rate": 2.7480916030534355e-05, |
|
"loss": 0.0802, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 8.7, |
|
"grad_norm": 0.3386911153793335, |
|
"learning_rate": 2.5954198473282442e-05, |
|
"loss": 0.0748, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 8.78, |
|
"grad_norm": 0.10998167097568512, |
|
"learning_rate": 2.4427480916030535e-05, |
|
"loss": 0.0366, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 8.85, |
|
"grad_norm": 0.12017246335744858, |
|
"learning_rate": 2.2900763358778628e-05, |
|
"loss": 0.0461, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 8.93, |
|
"grad_norm": 0.09363219887018204, |
|
"learning_rate": 2.1374045801526718e-05, |
|
"loss": 0.056, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 9.01, |
|
"grad_norm": 0.08617076277732849, |
|
"learning_rate": 1.984732824427481e-05, |
|
"loss": 0.0407, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 9.08, |
|
"grad_norm": 0.07992846518754959, |
|
"learning_rate": 1.83206106870229e-05, |
|
"loss": 0.0225, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 9.16, |
|
"grad_norm": 0.5971510410308838, |
|
"learning_rate": 1.6793893129770993e-05, |
|
"loss": 0.0583, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 9.16, |
|
"eval_accuracy": 0.5767754318618042, |
|
"eval_loss": 1.6794880628585815, |
|
"eval_runtime": 11.1041, |
|
"eval_samples_per_second": 93.839, |
|
"eval_steps_per_second": 11.797, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 9.24, |
|
"grad_norm": 0.06741204857826233, |
|
"learning_rate": 1.5267175572519086e-05, |
|
"loss": 0.0221, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 9.31, |
|
"grad_norm": 0.08024097234010696, |
|
"learning_rate": 1.3740458015267178e-05, |
|
"loss": 0.0357, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 9.39, |
|
"grad_norm": 0.17311328649520874, |
|
"learning_rate": 1.2213740458015267e-05, |
|
"loss": 0.0336, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 9.47, |
|
"grad_norm": 0.09014473855495453, |
|
"learning_rate": 1.0687022900763359e-05, |
|
"loss": 0.0514, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 9.54, |
|
"grad_norm": 0.07394669950008392, |
|
"learning_rate": 9.16030534351145e-06, |
|
"loss": 0.0534, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 9.62, |
|
"grad_norm": 0.11295180022716522, |
|
"learning_rate": 7.633587786259543e-06, |
|
"loss": 0.0254, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 9.69, |
|
"grad_norm": 1.9013243913650513, |
|
"learning_rate": 6.106870229007634e-06, |
|
"loss": 0.0919, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 9.77, |
|
"grad_norm": 0.08759523928165436, |
|
"learning_rate": 4.580152671755725e-06, |
|
"loss": 0.0476, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 9.85, |
|
"grad_norm": 0.07936747372150421, |
|
"learning_rate": 3.053435114503817e-06, |
|
"loss": 0.0365, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 9.92, |
|
"grad_norm": 0.13975408673286438, |
|
"learning_rate": 1.5267175572519084e-06, |
|
"loss": 0.0228, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 9.92, |
|
"eval_accuracy": 0.5882917466410749, |
|
"eval_loss": 1.6925907135009766, |
|
"eval_runtime": 11.1309, |
|
"eval_samples_per_second": 93.613, |
|
"eval_steps_per_second": 11.769, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 0.3255839943885803, |
|
"learning_rate": 0.0, |
|
"loss": 0.0364, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"step": 1310, |
|
"total_flos": 3.229206972532531e+18, |
|
"train_loss": 0.583979975267221, |
|
"train_runtime": 954.0647, |
|
"train_samples_per_second": 43.676, |
|
"train_steps_per_second": 1.373 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 1310, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 10, |
|
"save_steps": 100, |
|
"total_flos": 3.229206972532531e+18, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|