| { | |
| "best_metric": 0.9131455399061033, | |
| "best_model_checkpoint": "PhoWhisper-medium-vispeech-classifier-v1/checkpoint-4407", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4407, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010211027910142955, | |
| "grad_norm": 79976.4921875, | |
| "learning_rate": 6.122448979591837e-08, | |
| "loss": 2.0788, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02042205582028591, | |
| "grad_norm": 49252.46484375, | |
| "learning_rate": 1.2244897959183673e-07, | |
| "loss": 2.0787, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03063308373042886, | |
| "grad_norm": 66237.7890625, | |
| "learning_rate": 1.836734693877551e-07, | |
| "loss": 2.0734, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04084411164057182, | |
| "grad_norm": 66530.140625, | |
| "learning_rate": 2.4489795918367347e-07, | |
| "loss": 2.0745, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05105513955071477, | |
| "grad_norm": 75123.3515625, | |
| "learning_rate": 3.0612244897959183e-07, | |
| "loss": 2.0715, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06126616746085772, | |
| "grad_norm": 91761.9921875, | |
| "learning_rate": 3.673469387755102e-07, | |
| "loss": 2.0702, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07147719537100068, | |
| "grad_norm": 76544.7421875, | |
| "learning_rate": 4.2857142857142857e-07, | |
| "loss": 2.0659, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.08168822328114364, | |
| "grad_norm": 85211.5078125, | |
| "learning_rate": 4.897959183673469e-07, | |
| "loss": 2.0632, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09189925119128659, | |
| "grad_norm": 86731.6171875, | |
| "learning_rate": 5.510204081632653e-07, | |
| "loss": 2.0579, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.10211027910142954, | |
| "grad_norm": 73202.421875, | |
| "learning_rate": 6.122448979591837e-07, | |
| "loss": 2.0466, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1123213070115725, | |
| "grad_norm": 99795.828125, | |
| "learning_rate": 6.73469387755102e-07, | |
| "loss": 2.0266, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.12253233492171545, | |
| "grad_norm": 78665.6015625, | |
| "learning_rate": 7.346938775510204e-07, | |
| "loss": 2.0133, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13274336283185842, | |
| "grad_norm": 109804.5546875, | |
| "learning_rate": 7.959183673469389e-07, | |
| "loss": 2.0033, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.14295439074200136, | |
| "grad_norm": 88988.0234375, | |
| "learning_rate": 8.571428571428571e-07, | |
| "loss": 1.9912, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1531654186521443, | |
| "grad_norm": 163108.234375, | |
| "learning_rate": 9.183673469387755e-07, | |
| "loss": 1.9714, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.16337644656228728, | |
| "grad_norm": 147248.25, | |
| "learning_rate": 9.795918367346939e-07, | |
| "loss": 1.9368, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17358747447243023, | |
| "grad_norm": 161750.640625, | |
| "learning_rate": 1.0408163265306123e-06, | |
| "loss": 1.8955, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.18379850238257317, | |
| "grad_norm": 189184.296875, | |
| "learning_rate": 1.1020408163265306e-06, | |
| "loss": 1.8849, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.19400953029271614, | |
| "grad_norm": 160164.859375, | |
| "learning_rate": 1.163265306122449e-06, | |
| "loss": 1.8496, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2042205582028591, | |
| "grad_norm": 147315.296875, | |
| "learning_rate": 1.2244897959183673e-06, | |
| "loss": 1.8118, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21443158611300203, | |
| "grad_norm": 164890.859375, | |
| "learning_rate": 1.2857142857142856e-06, | |
| "loss": 1.7829, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.224642614023145, | |
| "grad_norm": 195027.578125, | |
| "learning_rate": 1.346938775510204e-06, | |
| "loss": 1.7938, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.23485364193328795, | |
| "grad_norm": 206460.03125, | |
| "learning_rate": 1.4081632653061225e-06, | |
| "loss": 1.7258, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.2450646698434309, | |
| "grad_norm": 247498.296875, | |
| "learning_rate": 1.4693877551020408e-06, | |
| "loss": 1.6498, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.25527569775357384, | |
| "grad_norm": 282144.96875, | |
| "learning_rate": 1.5306122448979593e-06, | |
| "loss": 1.6515, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.26548672566371684, | |
| "grad_norm": 244598.515625, | |
| "learning_rate": 1.5918367346938777e-06, | |
| "loss": 1.6327, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2756977535738598, | |
| "grad_norm": 169616.5, | |
| "learning_rate": 1.6530612244897958e-06, | |
| "loss": 1.576, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2859087814840027, | |
| "grad_norm": 186143.078125, | |
| "learning_rate": 1.7142857142857143e-06, | |
| "loss": 1.6067, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.29611980939414567, | |
| "grad_norm": 239726.171875, | |
| "learning_rate": 1.7755102040816327e-06, | |
| "loss": 1.5676, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.3063308373042886, | |
| "grad_norm": 224399.15625, | |
| "learning_rate": 1.836734693877551e-06, | |
| "loss": 1.4658, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31654186521443156, | |
| "grad_norm": 585805.375, | |
| "learning_rate": 1.8979591836734695e-06, | |
| "loss": 1.4288, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.32675289312457456, | |
| "grad_norm": 204082.328125, | |
| "learning_rate": 1.9591836734693877e-06, | |
| "loss": 1.4079, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3369639210347175, | |
| "grad_norm": 253897.078125, | |
| "learning_rate": 2.020408163265306e-06, | |
| "loss": 1.4658, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.34717494894486045, | |
| "grad_norm": 322156.15625, | |
| "learning_rate": 2.0816326530612247e-06, | |
| "loss": 1.3193, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.3573859768550034, | |
| "grad_norm": 209002.03125, | |
| "learning_rate": 2.142857142857143e-06, | |
| "loss": 1.3258, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.36759700476514634, | |
| "grad_norm": 239622.390625, | |
| "learning_rate": 2.204081632653061e-06, | |
| "loss": 1.3396, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.3778080326752893, | |
| "grad_norm": 302531.125, | |
| "learning_rate": 2.2653061224489797e-06, | |
| "loss": 1.2724, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.3880190605854323, | |
| "grad_norm": 277807.9375, | |
| "learning_rate": 2.326530612244898e-06, | |
| "loss": 1.2506, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.39823008849557523, | |
| "grad_norm": 543051.25, | |
| "learning_rate": 2.387755102040816e-06, | |
| "loss": 1.3114, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.4084411164057182, | |
| "grad_norm": 278458.59375, | |
| "learning_rate": 2.4489795918367347e-06, | |
| "loss": 1.2505, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4186521443158611, | |
| "grad_norm": 227854.078125, | |
| "learning_rate": 2.510204081632653e-06, | |
| "loss": 1.2032, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.42886317222600406, | |
| "grad_norm": 669112.0625, | |
| "learning_rate": 2.571428571428571e-06, | |
| "loss": 1.2144, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.43907420013614706, | |
| "grad_norm": 291158.375, | |
| "learning_rate": 2.6326530612244897e-06, | |
| "loss": 1.1663, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.44928522804629, | |
| "grad_norm": 322840.65625, | |
| "learning_rate": 2.693877551020408e-06, | |
| "loss": 1.1802, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.45949625595643295, | |
| "grad_norm": 399617.4375, | |
| "learning_rate": 2.7551020408163266e-06, | |
| "loss": 1.0181, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.4697072838665759, | |
| "grad_norm": 399545.5, | |
| "learning_rate": 2.816326530612245e-06, | |
| "loss": 1.2468, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.47991831177671884, | |
| "grad_norm": 1116027.875, | |
| "learning_rate": 2.877551020408163e-06, | |
| "loss": 1.2023, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.4901293396868618, | |
| "grad_norm": 300487.4375, | |
| "learning_rate": 2.9387755102040816e-06, | |
| "loss": 1.147, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5003403675970047, | |
| "grad_norm": 413331.65625, | |
| "learning_rate": 3e-06, | |
| "loss": 0.9871, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5105513955071477, | |
| "grad_norm": 715814.125, | |
| "learning_rate": 2.993192133131619e-06, | |
| "loss": 1.1643, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5207624234172906, | |
| "grad_norm": 1614147.0, | |
| "learning_rate": 2.9863842662632373e-06, | |
| "loss": 0.9421, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 639895.0625, | |
| "learning_rate": 2.9795763993948565e-06, | |
| "loss": 1.0573, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5411844792375766, | |
| "grad_norm": 345273.84375, | |
| "learning_rate": 2.9727685325264754e-06, | |
| "loss": 1.0337, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.5513955071477196, | |
| "grad_norm": 611370.375, | |
| "learning_rate": 2.9659606656580938e-06, | |
| "loss": 1.1131, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5616065350578625, | |
| "grad_norm": 239903.25, | |
| "learning_rate": 2.9591527987897126e-06, | |
| "loss": 0.9076, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.5718175629680055, | |
| "grad_norm": 257958.703125, | |
| "learning_rate": 2.9523449319213314e-06, | |
| "loss": 0.9152, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5820285908781484, | |
| "grad_norm": 590921.375, | |
| "learning_rate": 2.9455370650529502e-06, | |
| "loss": 0.8795, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.5922396187882913, | |
| "grad_norm": 363986.1875, | |
| "learning_rate": 2.938729198184569e-06, | |
| "loss": 1.0641, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6024506466984343, | |
| "grad_norm": 1878926.875, | |
| "learning_rate": 2.9319213313161875e-06, | |
| "loss": 0.9414, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.6126616746085772, | |
| "grad_norm": 2941328.0, | |
| "learning_rate": 2.9251134644478063e-06, | |
| "loss": 0.9541, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6228727025187202, | |
| "grad_norm": 1327942.75, | |
| "learning_rate": 2.918305597579425e-06, | |
| "loss": 0.8319, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.6330837304288631, | |
| "grad_norm": 161807.125, | |
| "learning_rate": 2.911497730711044e-06, | |
| "loss": 0.9268, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.6432947583390062, | |
| "grad_norm": 899555.875, | |
| "learning_rate": 2.9046898638426628e-06, | |
| "loss": 0.8278, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.6535057862491491, | |
| "grad_norm": 268360.0625, | |
| "learning_rate": 2.8978819969742816e-06, | |
| "loss": 0.9021, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6637168141592921, | |
| "grad_norm": 555703.0625, | |
| "learning_rate": 2.8910741301059e-06, | |
| "loss": 1.0146, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.673927842069435, | |
| "grad_norm": 222425.40625, | |
| "learning_rate": 2.884266263237519e-06, | |
| "loss": 1.0005, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.684138869979578, | |
| "grad_norm": 681254.4375, | |
| "learning_rate": 2.877458396369138e-06, | |
| "loss": 0.842, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.6943498978897209, | |
| "grad_norm": 1348221.125, | |
| "learning_rate": 2.8706505295007565e-06, | |
| "loss": 0.9394, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7045609257998638, | |
| "grad_norm": 225728.046875, | |
| "learning_rate": 2.8638426626323753e-06, | |
| "loss": 0.8289, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.7147719537100068, | |
| "grad_norm": 413089.65625, | |
| "learning_rate": 2.857034795763994e-06, | |
| "loss": 1.0175, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.7249829816201497, | |
| "grad_norm": 279751.71875, | |
| "learning_rate": 2.8502269288956125e-06, | |
| "loss": 0.9563, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.7351940095302927, | |
| "grad_norm": 422167.125, | |
| "learning_rate": 2.8434190620272318e-06, | |
| "loss": 0.8566, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.7454050374404356, | |
| "grad_norm": 2015877.125, | |
| "learning_rate": 2.83661119515885e-06, | |
| "loss": 0.7012, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.7556160653505786, | |
| "grad_norm": 502859.5, | |
| "learning_rate": 2.829803328290469e-06, | |
| "loss": 0.7563, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.7658270932607216, | |
| "grad_norm": 537785.125, | |
| "learning_rate": 2.822995461422088e-06, | |
| "loss": 0.9854, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.7760381211708646, | |
| "grad_norm": 390231.125, | |
| "learning_rate": 2.8161875945537066e-06, | |
| "loss": 1.0036, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.7862491490810075, | |
| "grad_norm": 370246.0625, | |
| "learning_rate": 2.8093797276853255e-06, | |
| "loss": 0.8186, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.7964601769911505, | |
| "grad_norm": 1279901.375, | |
| "learning_rate": 2.8025718608169443e-06, | |
| "loss": 0.9289, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8066712049012934, | |
| "grad_norm": 316319.25, | |
| "learning_rate": 2.7957639939485627e-06, | |
| "loss": 0.7542, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.8168822328114363, | |
| "grad_norm": 577071.5, | |
| "learning_rate": 2.7889561270801815e-06, | |
| "loss": 0.8579, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8270932607215793, | |
| "grad_norm": 412667.71875, | |
| "learning_rate": 2.7821482602118003e-06, | |
| "loss": 0.8424, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.8373042886317222, | |
| "grad_norm": 555197.125, | |
| "learning_rate": 2.775340393343419e-06, | |
| "loss": 0.8225, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.8475153165418652, | |
| "grad_norm": 429715.0, | |
| "learning_rate": 2.768532526475038e-06, | |
| "loss": 1.1042, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.8577263444520081, | |
| "grad_norm": 1407773.25, | |
| "learning_rate": 2.761724659606657e-06, | |
| "loss": 0.9892, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8679373723621511, | |
| "grad_norm": 328839.6875, | |
| "learning_rate": 2.754916792738275e-06, | |
| "loss": 0.7921, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.8781484002722941, | |
| "grad_norm": 196251.03125, | |
| "learning_rate": 2.748108925869894e-06, | |
| "loss": 0.6631, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.8883594281824371, | |
| "grad_norm": 761081.6875, | |
| "learning_rate": 2.7413010590015133e-06, | |
| "loss": 0.8396, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.89857045609258, | |
| "grad_norm": 905756.375, | |
| "learning_rate": 2.7344931921331317e-06, | |
| "loss": 0.8238, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.908781484002723, | |
| "grad_norm": 1221640.875, | |
| "learning_rate": 2.7276853252647505e-06, | |
| "loss": 0.7228, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.9189925119128659, | |
| "grad_norm": 2633817.0, | |
| "learning_rate": 2.7208774583963693e-06, | |
| "loss": 0.8766, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.9292035398230089, | |
| "grad_norm": 454206.53125, | |
| "learning_rate": 2.7140695915279877e-06, | |
| "loss": 0.922, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.9394145677331518, | |
| "grad_norm": 715028.8125, | |
| "learning_rate": 2.707261724659607e-06, | |
| "loss": 0.8212, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9496255956432947, | |
| "grad_norm": 384052.40625, | |
| "learning_rate": 2.7004538577912254e-06, | |
| "loss": 0.6728, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.9598366235534377, | |
| "grad_norm": 268327.0, | |
| "learning_rate": 2.693645990922844e-06, | |
| "loss": 0.7077, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.9700476514635806, | |
| "grad_norm": 430912.0625, | |
| "learning_rate": 2.686838124054463e-06, | |
| "loss": 0.8267, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.9802586793737236, | |
| "grad_norm": 448801.65625, | |
| "learning_rate": 2.680030257186082e-06, | |
| "loss": 0.6444, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.9904697072838666, | |
| "grad_norm": 625466.3125, | |
| "learning_rate": 2.6732223903177007e-06, | |
| "loss": 0.8255, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8819584171696848, | |
| "eval_loss": 0.5119155645370483, | |
| "eval_runtime": 579.0751, | |
| "eval_samples_per_second": 5.15, | |
| "eval_steps_per_second": 0.644, | |
| "step": 1469 | |
| }, | |
| { | |
| "epoch": 1.0006807351940095, | |
| "grad_norm": 166140.59375, | |
| "learning_rate": 2.6664145234493195e-06, | |
| "loss": 0.8523, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.0108917631041525, | |
| "grad_norm": 1578961.375, | |
| "learning_rate": 2.659606656580938e-06, | |
| "loss": 0.8466, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 1.0211027910142954, | |
| "grad_norm": 3046324.75, | |
| "learning_rate": 2.6527987897125567e-06, | |
| "loss": 0.7421, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0313138189244384, | |
| "grad_norm": 486359.28125, | |
| "learning_rate": 2.6459909228441755e-06, | |
| "loss": 0.586, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 1.0415248468345812, | |
| "grad_norm": 476911.5, | |
| "learning_rate": 2.6391830559757944e-06, | |
| "loss": 0.7123, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.0517358747447243, | |
| "grad_norm": 312268.75, | |
| "learning_rate": 2.632375189107413e-06, | |
| "loss": 0.7437, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 555973.375, | |
| "learning_rate": 2.625567322239032e-06, | |
| "loss": 0.7994, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.0721579305650102, | |
| "grad_norm": 441313.59375, | |
| "learning_rate": 2.6187594553706504e-06, | |
| "loss": 0.7675, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.0823689584751532, | |
| "grad_norm": 2035757.875, | |
| "learning_rate": 2.6119515885022692e-06, | |
| "loss": 0.7534, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.092579986385296, | |
| "grad_norm": 453271.59375, | |
| "learning_rate": 2.605143721633888e-06, | |
| "loss": 0.645, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.1027910142954391, | |
| "grad_norm": 319587.0625, | |
| "learning_rate": 2.598335854765507e-06, | |
| "loss": 0.765, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.113002042205582, | |
| "grad_norm": 732482.8125, | |
| "learning_rate": 2.5915279878971257e-06, | |
| "loss": 0.6155, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.123213070115725, | |
| "grad_norm": 396029.65625, | |
| "learning_rate": 2.5847201210287445e-06, | |
| "loss": 0.7213, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.1334240980258679, | |
| "grad_norm": 326143.03125, | |
| "learning_rate": 2.577912254160363e-06, | |
| "loss": 0.7134, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.143635125936011, | |
| "grad_norm": 1083100.875, | |
| "learning_rate": 2.571104387291982e-06, | |
| "loss": 0.7249, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 905634.25, | |
| "learning_rate": 2.5642965204236006e-06, | |
| "loss": 0.767, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.1640571817562968, | |
| "grad_norm": 656056.9375, | |
| "learning_rate": 2.5574886535552194e-06, | |
| "loss": 0.8304, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.1742682096664399, | |
| "grad_norm": 722782.5625, | |
| "learning_rate": 2.5506807866868382e-06, | |
| "loss": 0.6995, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.1844792375765827, | |
| "grad_norm": 577313.5625, | |
| "learning_rate": 2.5438729198184566e-06, | |
| "loss": 0.9264, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.1946902654867257, | |
| "grad_norm": 2527231.5, | |
| "learning_rate": 2.537065052950076e-06, | |
| "loss": 0.9283, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.2049012933968686, | |
| "grad_norm": 1089097.5, | |
| "learning_rate": 2.5302571860816947e-06, | |
| "loss": 0.6552, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.2151123213070116, | |
| "grad_norm": 747680.25, | |
| "learning_rate": 2.523449319213313e-06, | |
| "loss": 0.7181, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.2253233492171545, | |
| "grad_norm": 143748.9375, | |
| "learning_rate": 2.516641452344932e-06, | |
| "loss": 0.5906, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.2355343771272975, | |
| "grad_norm": 782987.3125, | |
| "learning_rate": 2.5098335854765507e-06, | |
| "loss": 0.7338, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.2457454050374404, | |
| "grad_norm": 199057.203125, | |
| "learning_rate": 2.5030257186081696e-06, | |
| "loss": 0.6712, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.2559564329475834, | |
| "grad_norm": 169465.65625, | |
| "learning_rate": 2.4962178517397884e-06, | |
| "loss": 0.4682, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.2661674608577265, | |
| "grad_norm": 458018.5, | |
| "learning_rate": 2.4894099848714072e-06, | |
| "loss": 0.7659, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.2763784887678693, | |
| "grad_norm": 373506.46875, | |
| "learning_rate": 2.4826021180030256e-06, | |
| "loss": 0.7384, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.2865895166780121, | |
| "grad_norm": 140858.234375, | |
| "learning_rate": 2.4757942511346444e-06, | |
| "loss": 0.723, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.2968005445881552, | |
| "grad_norm": 758790.1875, | |
| "learning_rate": 2.4689863842662633e-06, | |
| "loss": 0.9274, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.3070115724982982, | |
| "grad_norm": 311553.5625, | |
| "learning_rate": 2.462178517397882e-06, | |
| "loss": 0.6113, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.317222600408441, | |
| "grad_norm": 167712.234375, | |
| "learning_rate": 2.455370650529501e-06, | |
| "loss": 0.6976, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.3274336283185841, | |
| "grad_norm": 500111.15625, | |
| "learning_rate": 2.4485627836611193e-06, | |
| "loss": 0.7833, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.337644656228727, | |
| "grad_norm": 517175.0, | |
| "learning_rate": 2.441754916792738e-06, | |
| "loss": 0.6971, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.34785568413887, | |
| "grad_norm": 659138.6875, | |
| "learning_rate": 2.4349470499243574e-06, | |
| "loss": 0.7602, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.3580667120490129, | |
| "grad_norm": 301056.09375, | |
| "learning_rate": 2.4281391830559758e-06, | |
| "loss": 0.6703, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.368277739959156, | |
| "grad_norm": 273580.0625, | |
| "learning_rate": 2.4213313161875946e-06, | |
| "loss": 0.6183, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.3784887678692987, | |
| "grad_norm": 559704.875, | |
| "learning_rate": 2.4145234493192134e-06, | |
| "loss": 0.564, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.3886997957794418, | |
| "grad_norm": 593091.6875, | |
| "learning_rate": 2.407715582450832e-06, | |
| "loss": 0.605, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.3989108236895849, | |
| "grad_norm": 307976.4375, | |
| "learning_rate": 2.400907715582451e-06, | |
| "loss": 0.771, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.4091218515997277, | |
| "grad_norm": 593720.625, | |
| "learning_rate": 2.39409984871407e-06, | |
| "loss": 0.7066, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.4193328795098705, | |
| "grad_norm": 831806.0625, | |
| "learning_rate": 2.3872919818456883e-06, | |
| "loss": 0.5807, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 1.4295439074200136, | |
| "grad_norm": 206191.21875, | |
| "learning_rate": 2.380484114977307e-06, | |
| "loss": 0.6574, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.4397549353301566, | |
| "grad_norm": 258367.828125, | |
| "learning_rate": 2.373676248108926e-06, | |
| "loss": 0.6792, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 1.4499659632402995, | |
| "grad_norm": 1177121.25, | |
| "learning_rate": 2.3668683812405448e-06, | |
| "loss": 0.8608, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.4601769911504425, | |
| "grad_norm": 368604.9375, | |
| "learning_rate": 2.3600605143721636e-06, | |
| "loss": 0.7224, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 1.4703880190605854, | |
| "grad_norm": 570852.125, | |
| "learning_rate": 2.3532526475037824e-06, | |
| "loss": 0.6717, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.4805990469707284, | |
| "grad_norm": 240292.90625, | |
| "learning_rate": 2.346444780635401e-06, | |
| "loss": 0.5415, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.4908100748808715, | |
| "grad_norm": 1409754.875, | |
| "learning_rate": 2.3396369137670197e-06, | |
| "loss": 0.66, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.5010211027910143, | |
| "grad_norm": 1670060.75, | |
| "learning_rate": 2.3328290468986385e-06, | |
| "loss": 0.686, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 1.5112321307011571, | |
| "grad_norm": 530769.5, | |
| "learning_rate": 2.3260211800302573e-06, | |
| "loss": 0.612, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.5214431586113002, | |
| "grad_norm": 1758232.75, | |
| "learning_rate": 2.319213313161876e-06, | |
| "loss": 0.6849, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 1.5316541865214433, | |
| "grad_norm": 1104782.0, | |
| "learning_rate": 2.3124054462934945e-06, | |
| "loss": 0.7704, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.541865214431586, | |
| "grad_norm": 826967.125, | |
| "learning_rate": 2.3055975794251133e-06, | |
| "loss": 0.5943, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 1.552076242341729, | |
| "grad_norm": 1028021.3125, | |
| "learning_rate": 2.2987897125567326e-06, | |
| "loss": 0.6579, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.562287270251872, | |
| "grad_norm": 250268.609375, | |
| "learning_rate": 2.291981845688351e-06, | |
| "loss": 0.671, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 1.572498298162015, | |
| "grad_norm": 608503.5625, | |
| "learning_rate": 2.28517397881997e-06, | |
| "loss": 0.546, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.582709326072158, | |
| "grad_norm": 2906189.5, | |
| "learning_rate": 2.2783661119515886e-06, | |
| "loss": 0.5798, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.592920353982301, | |
| "grad_norm": 1186615.875, | |
| "learning_rate": 2.271558245083207e-06, | |
| "loss": 0.8234, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.6031313818924438, | |
| "grad_norm": 189114.234375, | |
| "learning_rate": 2.2647503782148263e-06, | |
| "loss": 0.7017, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 1.6133424098025868, | |
| "grad_norm": 1086454.375, | |
| "learning_rate": 2.257942511346445e-06, | |
| "loss": 0.6067, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.6235534377127299, | |
| "grad_norm": 438766.8125, | |
| "learning_rate": 2.2511346444780635e-06, | |
| "loss": 0.7277, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 1.6337644656228727, | |
| "grad_norm": 746603.625, | |
| "learning_rate": 2.2443267776096823e-06, | |
| "loss": 0.7049, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.6439754935330155, | |
| "grad_norm": 1114921.375, | |
| "learning_rate": 2.237518910741301e-06, | |
| "loss": 0.6517, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 1.6541865214431586, | |
| "grad_norm": 119118.2421875, | |
| "learning_rate": 2.23071104387292e-06, | |
| "loss": 0.6355, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.6643975493533016, | |
| "grad_norm": 278499.84375, | |
| "learning_rate": 2.223903177004539e-06, | |
| "loss": 0.528, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 1.6746085772634445, | |
| "grad_norm": 2842614.5, | |
| "learning_rate": 2.2170953101361572e-06, | |
| "loss": 0.818, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.6848196051735873, | |
| "grad_norm": 681713.6875, | |
| "learning_rate": 2.210287443267776e-06, | |
| "loss": 0.5879, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.6950306330837304, | |
| "grad_norm": 955702.4375, | |
| "learning_rate": 2.203479576399395e-06, | |
| "loss": 0.6868, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.7052416609938734, | |
| "grad_norm": 293277.28125, | |
| "learning_rate": 2.1966717095310137e-06, | |
| "loss": 0.6262, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 1.7154526889040165, | |
| "grad_norm": 1372940.875, | |
| "learning_rate": 2.1898638426626325e-06, | |
| "loss": 0.7799, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.7256637168141593, | |
| "grad_norm": 277903.0625, | |
| "learning_rate": 2.1830559757942513e-06, | |
| "loss": 0.6204, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 1.7358747447243021, | |
| "grad_norm": 959110.125, | |
| "learning_rate": 2.1762481089258697e-06, | |
| "loss": 0.7283, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.7460857726344452, | |
| "grad_norm": 870804.125, | |
| "learning_rate": 2.1694402420574886e-06, | |
| "loss": 0.6073, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 1.7562968005445883, | |
| "grad_norm": 723633.5, | |
| "learning_rate": 2.162632375189108e-06, | |
| "loss": 0.8315, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.766507828454731, | |
| "grad_norm": 279254.0625, | |
| "learning_rate": 2.155824508320726e-06, | |
| "loss": 0.6998, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 1.776718856364874, | |
| "grad_norm": 816422.0625, | |
| "learning_rate": 2.149016641452345e-06, | |
| "loss": 0.7695, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.786929884275017, | |
| "grad_norm": 277112.28125, | |
| "learning_rate": 2.142208774583964e-06, | |
| "loss": 0.5903, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.79714091218516, | |
| "grad_norm": 1158773.75, | |
| "learning_rate": 2.1354009077155823e-06, | |
| "loss": 0.5811, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.807351940095303, | |
| "grad_norm": 3944267.75, | |
| "learning_rate": 2.1285930408472015e-06, | |
| "loss": 0.5903, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 1.817562968005446, | |
| "grad_norm": 774781.9375, | |
| "learning_rate": 2.12178517397882e-06, | |
| "loss": 0.7741, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.8277739959155888, | |
| "grad_norm": 744016.25, | |
| "learning_rate": 2.1149773071104387e-06, | |
| "loss": 0.7171, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 1.8379850238257318, | |
| "grad_norm": 5070120.5, | |
| "learning_rate": 2.1081694402420575e-06, | |
| "loss": 0.5936, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.8481960517358749, | |
| "grad_norm": 249426.0, | |
| "learning_rate": 2.1013615733736764e-06, | |
| "loss": 0.7609, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 1.8584070796460177, | |
| "grad_norm": 403828.75, | |
| "learning_rate": 2.094553706505295e-06, | |
| "loss": 0.6597, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.8686181075561605, | |
| "grad_norm": 829779.0625, | |
| "learning_rate": 2.087745839636914e-06, | |
| "loss": 0.6788, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 1.8788291354663036, | |
| "grad_norm": 939866.5625, | |
| "learning_rate": 2.0809379727685324e-06, | |
| "loss": 0.5589, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.8890401633764466, | |
| "grad_norm": 1203267.0, | |
| "learning_rate": 2.0741301059001512e-06, | |
| "loss": 0.7759, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.8992511912865895, | |
| "grad_norm": 1656486.5, | |
| "learning_rate": 2.06732223903177e-06, | |
| "loss": 0.4373, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.9094622191967323, | |
| "grad_norm": 1833536.875, | |
| "learning_rate": 2.060514372163389e-06, | |
| "loss": 1.0234, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 1.9196732471068754, | |
| "grad_norm": 429027.65625, | |
| "learning_rate": 2.0537065052950077e-06, | |
| "loss": 0.5515, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.9298842750170184, | |
| "grad_norm": 2480347.0, | |
| "learning_rate": 2.0468986384266265e-06, | |
| "loss": 0.679, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 1.9400953029271615, | |
| "grad_norm": 682525.625, | |
| "learning_rate": 2.040090771558245e-06, | |
| "loss": 0.5025, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.9503063308373043, | |
| "grad_norm": 289022.15625, | |
| "learning_rate": 2.0332829046898638e-06, | |
| "loss": 0.5361, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 1.9605173587474471, | |
| "grad_norm": 633536.75, | |
| "learning_rate": 2.0264750378214826e-06, | |
| "loss": 0.6829, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.9707283866575902, | |
| "grad_norm": 551145.0, | |
| "learning_rate": 2.0196671709531014e-06, | |
| "loss": 0.591, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 1.9809394145677333, | |
| "grad_norm": 951443.5, | |
| "learning_rate": 2.0128593040847202e-06, | |
| "loss": 0.5114, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.991150442477876, | |
| "grad_norm": 1422993.0, | |
| "learning_rate": 2.006051437216339e-06, | |
| "loss": 0.6774, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9124748490945674, | |
| "eval_loss": 0.3880421221256256, | |
| "eval_runtime": 580.6935, | |
| "eval_samples_per_second": 5.135, | |
| "eval_steps_per_second": 0.642, | |
| "step": 2938 | |
| }, | |
| { | |
| "epoch": 2.001361470388019, | |
| "grad_norm": 149413.0625, | |
| "learning_rate": 1.9992435703479575e-06, | |
| "loss": 0.5516, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.011572498298162, | |
| "grad_norm": 92282.625, | |
| "learning_rate": 1.9924357034795767e-06, | |
| "loss": 0.5141, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 2.021783526208305, | |
| "grad_norm": 771094.375, | |
| "learning_rate": 1.985627836611195e-06, | |
| "loss": 0.5162, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.031994554118448, | |
| "grad_norm": 1059304.125, | |
| "learning_rate": 1.978819969742814e-06, | |
| "loss": 0.5794, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 2.0422055820285907, | |
| "grad_norm": 588466.6875, | |
| "learning_rate": 1.9720121028744328e-06, | |
| "loss": 0.6778, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0524166099387338, | |
| "grad_norm": 571242.3125, | |
| "learning_rate": 1.965204236006051e-06, | |
| "loss": 0.6637, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 2.062627637848877, | |
| "grad_norm": 1182853.5, | |
| "learning_rate": 1.9583963691376704e-06, | |
| "loss": 0.4922, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.07283866575902, | |
| "grad_norm": 669133.8125, | |
| "learning_rate": 1.9515885022692892e-06, | |
| "loss": 0.4657, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 2.0830496936691625, | |
| "grad_norm": 323752.25, | |
| "learning_rate": 1.9447806354009076e-06, | |
| "loss": 0.596, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.0932607215793055, | |
| "grad_norm": 1483114.375, | |
| "learning_rate": 1.9379727685325265e-06, | |
| "loss": 0.5028, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 2.1034717494894486, | |
| "grad_norm": 214087.359375, | |
| "learning_rate": 1.9311649016641453e-06, | |
| "loss": 0.763, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.1136827773995917, | |
| "grad_norm": 136251.953125, | |
| "learning_rate": 1.924357034795764e-06, | |
| "loss": 0.4622, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 132665.53125, | |
| "learning_rate": 1.917549167927383e-06, | |
| "loss": 0.5564, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.1341048332198773, | |
| "grad_norm": 741566.0, | |
| "learning_rate": 1.9107413010590018e-06, | |
| "loss": 0.5569, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 2.1443158611300204, | |
| "grad_norm": 612487.625, | |
| "learning_rate": 1.9039334341906204e-06, | |
| "loss": 0.576, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.1545268890401634, | |
| "grad_norm": 953638.5625, | |
| "learning_rate": 1.897125567322239e-06, | |
| "loss": 0.3974, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 2.1647379169503065, | |
| "grad_norm": 663514.3125, | |
| "learning_rate": 1.890317700453858e-06, | |
| "loss": 0.471, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.174948944860449, | |
| "grad_norm": 197147.5, | |
| "learning_rate": 1.8835098335854766e-06, | |
| "loss": 0.6322, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 2.185159972770592, | |
| "grad_norm": 61634.29296875, | |
| "learning_rate": 1.8767019667170954e-06, | |
| "loss": 0.5017, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.195371000680735, | |
| "grad_norm": 309183.3125, | |
| "learning_rate": 1.869894099848714e-06, | |
| "loss": 0.5658, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 2.2055820285908783, | |
| "grad_norm": 124933.015625, | |
| "learning_rate": 1.8630862329803327e-06, | |
| "loss": 0.4846, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.2157930565010213, | |
| "grad_norm": 547002.375, | |
| "learning_rate": 1.8562783661119517e-06, | |
| "loss": 0.4722, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 2.226004084411164, | |
| "grad_norm": 220335.03125, | |
| "learning_rate": 1.8494704992435705e-06, | |
| "loss": 0.4243, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 2.236215112321307, | |
| "grad_norm": 882093.1875, | |
| "learning_rate": 1.8426626323751891e-06, | |
| "loss": 0.5559, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 2.24642614023145, | |
| "grad_norm": 947867.0625, | |
| "learning_rate": 1.835854765506808e-06, | |
| "loss": 0.4759, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.256637168141593, | |
| "grad_norm": 507538.5, | |
| "learning_rate": 1.8290468986384266e-06, | |
| "loss": 0.4976, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 2.2668481960517357, | |
| "grad_norm": 1198014.5, | |
| "learning_rate": 1.8222390317700456e-06, | |
| "loss": 0.4435, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 2.2770592239618788, | |
| "grad_norm": 219044.03125, | |
| "learning_rate": 1.8154311649016642e-06, | |
| "loss": 0.6013, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 2.287270251872022, | |
| "grad_norm": 525360.9375, | |
| "learning_rate": 1.808623298033283e-06, | |
| "loss": 0.5409, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 2.297481279782165, | |
| "grad_norm": 343133.25, | |
| "learning_rate": 1.8018154311649017e-06, | |
| "loss": 0.6779, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 994074.375, | |
| "learning_rate": 1.7950075642965203e-06, | |
| "loss": 0.3746, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 2.3179033356024505, | |
| "grad_norm": 675095.9375, | |
| "learning_rate": 1.7881996974281393e-06, | |
| "loss": 0.6972, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 2.3281143635125936, | |
| "grad_norm": 1367137.125, | |
| "learning_rate": 1.7813918305597581e-06, | |
| "loss": 0.462, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 2.3383253914227367, | |
| "grad_norm": 302348.9375, | |
| "learning_rate": 1.7745839636913767e-06, | |
| "loss": 0.3931, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 2.3485364193328797, | |
| "grad_norm": 727419.4375, | |
| "learning_rate": 1.7677760968229954e-06, | |
| "loss": 0.687, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.3587474472430223, | |
| "grad_norm": 3086808.0, | |
| "learning_rate": 1.7609682299546142e-06, | |
| "loss": 0.4411, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 2.3689584751531654, | |
| "grad_norm": 930683.1875, | |
| "learning_rate": 1.7541603630862332e-06, | |
| "loss": 0.669, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 2.3791695030633084, | |
| "grad_norm": 282012.15625, | |
| "learning_rate": 1.7473524962178518e-06, | |
| "loss": 0.5843, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 2.3893805309734515, | |
| "grad_norm": 106134.8125, | |
| "learning_rate": 1.7405446293494707e-06, | |
| "loss": 0.5899, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 2.399591558883594, | |
| "grad_norm": 56236.19140625, | |
| "learning_rate": 1.7337367624810893e-06, | |
| "loss": 0.4536, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 2.409802586793737, | |
| "grad_norm": 344606.1875, | |
| "learning_rate": 1.7269288956127079e-06, | |
| "loss": 0.443, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 2.42001361470388, | |
| "grad_norm": 771505.9375, | |
| "learning_rate": 1.720121028744327e-06, | |
| "loss": 0.4357, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 2.4302246426140233, | |
| "grad_norm": 80650.0703125, | |
| "learning_rate": 1.7133131618759457e-06, | |
| "loss": 0.5167, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 2.440435670524166, | |
| "grad_norm": 2634711.0, | |
| "learning_rate": 1.7065052950075644e-06, | |
| "loss": 0.3663, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 2.450646698434309, | |
| "grad_norm": 1415573.875, | |
| "learning_rate": 1.699697428139183e-06, | |
| "loss": 0.4403, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.460857726344452, | |
| "grad_norm": 2269082.75, | |
| "learning_rate": 1.6928895612708018e-06, | |
| "loss": 0.573, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 2.471068754254595, | |
| "grad_norm": 1052448.875, | |
| "learning_rate": 1.6860816944024208e-06, | |
| "loss": 0.4889, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 2.481279782164738, | |
| "grad_norm": 770280.5, | |
| "learning_rate": 1.6792738275340394e-06, | |
| "loss": 0.6652, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 2.4914908100748807, | |
| "grad_norm": 1298375.5, | |
| "learning_rate": 1.6724659606656583e-06, | |
| "loss": 0.4182, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 2.5017018379850238, | |
| "grad_norm": 1415979.75, | |
| "learning_rate": 1.6656580937972769e-06, | |
| "loss": 0.4441, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 2.511912865895167, | |
| "grad_norm": 663093.6875, | |
| "learning_rate": 1.6588502269288955e-06, | |
| "loss": 0.4322, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 2.52212389380531, | |
| "grad_norm": 2289919.25, | |
| "learning_rate": 1.6520423600605145e-06, | |
| "loss": 0.4012, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 2.532334921715453, | |
| "grad_norm": 1178189.375, | |
| "learning_rate": 1.6452344931921333e-06, | |
| "loss": 0.5652, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 2.5425459496255955, | |
| "grad_norm": 222618.671875, | |
| "learning_rate": 1.638426626323752e-06, | |
| "loss": 0.3636, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 2.5527569775357386, | |
| "grad_norm": 642872.9375, | |
| "learning_rate": 1.6316187594553706e-06, | |
| "loss": 0.6289, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.5629680054458817, | |
| "grad_norm": 316546.15625, | |
| "learning_rate": 1.6248108925869894e-06, | |
| "loss": 0.6035, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 2.5731790333560243, | |
| "grad_norm": 63267.38671875, | |
| "learning_rate": 1.6180030257186084e-06, | |
| "loss": 0.4131, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 2.5833900612661673, | |
| "grad_norm": 3657979.0, | |
| "learning_rate": 1.611195158850227e-06, | |
| "loss": 0.7483, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 2.5936010891763104, | |
| "grad_norm": 168698.3125, | |
| "learning_rate": 1.6043872919818457e-06, | |
| "loss": 0.4738, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 2.6038121170864534, | |
| "grad_norm": 680610.5625, | |
| "learning_rate": 1.5975794251134645e-06, | |
| "loss": 0.4139, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 2.6140231449965965, | |
| "grad_norm": 326585.28125, | |
| "learning_rate": 1.590771558245083e-06, | |
| "loss": 0.2408, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 2.624234172906739, | |
| "grad_norm": 409793.6875, | |
| "learning_rate": 1.5839636913767021e-06, | |
| "loss": 0.5334, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 2.634445200816882, | |
| "grad_norm": 55384.58984375, | |
| "learning_rate": 1.577155824508321e-06, | |
| "loss": 0.7308, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 2.644656228727025, | |
| "grad_norm": 2711447.5, | |
| "learning_rate": 1.5703479576399396e-06, | |
| "loss": 0.675, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 2.6548672566371683, | |
| "grad_norm": 1392973.0, | |
| "learning_rate": 1.5635400907715582e-06, | |
| "loss": 0.6709, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.6650782845473113, | |
| "grad_norm": 115502.75, | |
| "learning_rate": 1.556732223903177e-06, | |
| "loss": 0.4405, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 2.675289312457454, | |
| "grad_norm": 715057.4375, | |
| "learning_rate": 1.549924357034796e-06, | |
| "loss": 0.3218, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 2.685500340367597, | |
| "grad_norm": 608108.9375, | |
| "learning_rate": 1.5431164901664146e-06, | |
| "loss": 0.4415, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 2.69571136827774, | |
| "grad_norm": 427073.5, | |
| "learning_rate": 1.5363086232980333e-06, | |
| "loss": 0.5927, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 2.7059223961878827, | |
| "grad_norm": 296582.28125, | |
| "learning_rate": 1.529500756429652e-06, | |
| "loss": 0.4821, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 2.7161334240980257, | |
| "grad_norm": 7283655.0, | |
| "learning_rate": 1.5226928895612707e-06, | |
| "loss": 0.6949, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 2.7263444520081688, | |
| "grad_norm": 4511815.5, | |
| "learning_rate": 1.5158850226928897e-06, | |
| "loss": 0.4827, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 2.736555479918312, | |
| "grad_norm": 2752351.5, | |
| "learning_rate": 1.5090771558245083e-06, | |
| "loss": 0.4366, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 2.746766507828455, | |
| "grad_norm": 2160780.75, | |
| "learning_rate": 1.5022692889561272e-06, | |
| "loss": 0.6442, | |
| "step": 4035 | |
| }, | |
| { | |
| "epoch": 2.7569775357385975, | |
| "grad_norm": 842177.375, | |
| "learning_rate": 1.4954614220877458e-06, | |
| "loss": 0.3898, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 2.7671885636487406, | |
| "grad_norm": 488163.53125, | |
| "learning_rate": 1.4886535552193646e-06, | |
| "loss": 0.3194, | |
| "step": 4065 | |
| }, | |
| { | |
| "epoch": 2.7773995915588836, | |
| "grad_norm": 1358681.25, | |
| "learning_rate": 1.4818456883509834e-06, | |
| "loss": 0.8422, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 2.7876106194690267, | |
| "grad_norm": 2055414.25, | |
| "learning_rate": 1.475037821482602e-06, | |
| "loss": 0.7359, | |
| "step": 4095 | |
| }, | |
| { | |
| "epoch": 2.7978216473791697, | |
| "grad_norm": 1410838.125, | |
| "learning_rate": 1.4682299546142209e-06, | |
| "loss": 0.5694, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 2.8080326752893123, | |
| "grad_norm": 485180.3125, | |
| "learning_rate": 1.4614220877458397e-06, | |
| "loss": 0.3498, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 2.8182437031994554, | |
| "grad_norm": 442321.53125, | |
| "learning_rate": 1.4546142208774585e-06, | |
| "loss": 0.3907, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 2.8284547311095984, | |
| "grad_norm": 117394.3359375, | |
| "learning_rate": 1.4478063540090771e-06, | |
| "loss": 0.4456, | |
| "step": 4155 | |
| }, | |
| { | |
| "epoch": 2.838665759019741, | |
| "grad_norm": 1168203.125, | |
| "learning_rate": 1.440998487140696e-06, | |
| "loss": 0.4467, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 2.8488767869298846, | |
| "grad_norm": 1550697.0, | |
| "learning_rate": 1.4341906202723148e-06, | |
| "loss": 0.3427, | |
| "step": 4185 | |
| }, | |
| { | |
| "epoch": 2.859087814840027, | |
| "grad_norm": 865472.9375, | |
| "learning_rate": 1.4273827534039334e-06, | |
| "loss": 0.5203, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.86929884275017, | |
| "grad_norm": 548500.1875, | |
| "learning_rate": 1.4205748865355522e-06, | |
| "loss": 0.6109, | |
| "step": 4215 | |
| }, | |
| { | |
| "epoch": 2.8795098706603133, | |
| "grad_norm": 2530923.75, | |
| "learning_rate": 1.413767019667171e-06, | |
| "loss": 0.6251, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 2.889720898570456, | |
| "grad_norm": 1599674.125, | |
| "learning_rate": 1.4069591527987896e-06, | |
| "loss": 0.3419, | |
| "step": 4245 | |
| }, | |
| { | |
| "epoch": 2.899931926480599, | |
| "grad_norm": 1614385.875, | |
| "learning_rate": 1.4001512859304085e-06, | |
| "loss": 0.7362, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 2.910142954390742, | |
| "grad_norm": 2576167.5, | |
| "learning_rate": 1.3933434190620273e-06, | |
| "loss": 0.5736, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 2.920353982300885, | |
| "grad_norm": 3079889.75, | |
| "learning_rate": 1.3865355521936461e-06, | |
| "loss": 0.4168, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 2.930565010211028, | |
| "grad_norm": 483019.09375, | |
| "learning_rate": 1.3797276853252647e-06, | |
| "loss": 0.5769, | |
| "step": 4305 | |
| }, | |
| { | |
| "epoch": 2.9407760381211707, | |
| "grad_norm": 953947.1875, | |
| "learning_rate": 1.3729198184568836e-06, | |
| "loss": 0.5267, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 2.9509870660313138, | |
| "grad_norm": 951762.1875, | |
| "learning_rate": 1.3661119515885024e-06, | |
| "loss": 0.465, | |
| "step": 4335 | |
| }, | |
| { | |
| "epoch": 2.961198093941457, | |
| "grad_norm": 925738.1875, | |
| "learning_rate": 1.359304084720121e-06, | |
| "loss": 0.4431, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.9714091218516, | |
| "grad_norm": 2421353.0, | |
| "learning_rate": 1.3524962178517398e-06, | |
| "loss": 0.4951, | |
| "step": 4365 | |
| }, | |
| { | |
| "epoch": 2.981620149761743, | |
| "grad_norm": 2792055.5, | |
| "learning_rate": 1.3456883509833586e-06, | |
| "loss": 0.4273, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 2.9918311776718856, | |
| "grad_norm": 78971.71875, | |
| "learning_rate": 1.3388804841149772e-06, | |
| "loss": 0.452, | |
| "step": 4395 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9131455399061033, | |
| "eval_loss": 0.36749738454818726, | |
| "eval_runtime": 574.4362, | |
| "eval_samples_per_second": 5.191, | |
| "eval_steps_per_second": 0.649, | |
| "step": 4407 | |
| } | |
| ], | |
| "logging_steps": 15, | |
| "max_steps": 7345, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.560567924145152e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |