{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.15373489260673523, "learning_rate": 9e-07, "loss": 1.2116, "step": 10 }, { "grad_norm": 0.12143079191446304, "learning_rate": 1.9e-06, "loss": 1.2054, "step": 20 }, { "grad_norm": 0.11240886896848679, "learning_rate": 2.9e-06, "loss": 1.2027, "step": 30 }, { "grad_norm": 0.21381829679012299, "learning_rate": 3.9e-06, "loss": 1.1886, "step": 40 }, { "grad_norm": 0.2458868771791458, "learning_rate": 4.9000000000000005e-06, "loss": 1.1589, "step": 50 }, { "grad_norm": 0.2544420659542084, "learning_rate": 5.9e-06, "loss": 1.1296, "step": 60 }, { "grad_norm": 0.23853255808353424, "learning_rate": 6.900000000000001e-06, "loss": 1.1103, "step": 70 }, { "grad_norm": 0.31489869952201843, "learning_rate": 7.9e-06, "loss": 1.0948, "step": 80 }, { "grad_norm": 0.2596352696418762, "learning_rate": 8.9e-06, "loss": 1.0914, "step": 90 }, { "grad_norm": 0.2258705496788025, "learning_rate": 9.900000000000002e-06, "loss": 1.0912, "step": 100 }, { "grad_norm": 0.5682782530784607, "learning_rate": 1.09e-05, "loss": 1.0893, "step": 110 }, { "grad_norm": 0.3197626769542694, "learning_rate": 1.19e-05, "loss": 1.081, "step": 120 }, { "grad_norm": 0.18166182935237885, "learning_rate": 1.29e-05, "loss": 1.073, "step": 130 }, { "grad_norm": 0.4554118812084198, "learning_rate": 1.3900000000000002e-05, "loss": 1.0634, "step": 140 }, { "grad_norm": 0.42795446515083313, "learning_rate": 1.49e-05, "loss": 1.047, "step": 150 }, { "grad_norm": 0.5283658504486084, "learning_rate": 1.59e-05, "loss": 1.013, "step": 160 }, { "grad_norm": 0.3358186185359955, "learning_rate": 1.69e-05, "loss": 0.9804, "step": 170 }, { "grad_norm": 0.5145778656005859, "learning_rate": 1.79e-05, "loss": 0.936, "step": 180 }, { "grad_norm": 0.5755268335342407, "learning_rate": 1.8900000000000002e-05, "loss": 0.8896, "step": 190 }, { "grad_norm": 0.6188399791717529, "learning_rate": 1.9900000000000003e-05, "loss": 0.8577, "step": 200 }, { "grad_norm": 0.9716882109642029, "learning_rate": 2.09e-05, "loss": 0.8271, "step": 210 }, { "grad_norm": 0.918986976146698, "learning_rate": 2.19e-05, "loss": 0.7906, "step": 220 }, { "grad_norm": 0.8000361919403076, "learning_rate": 2.29e-05, "loss": 0.751, "step": 230 }, { "grad_norm": 1.0514482259750366, "learning_rate": 2.39e-05, "loss": 0.7155, "step": 240 }, { "grad_norm": 0.7911136746406555, "learning_rate": 2.4900000000000002e-05, "loss": 0.6906, "step": 250 }, { "grad_norm": 0.92540043592453, "learning_rate": 2.5900000000000003e-05, "loss": 0.6638, "step": 260 }, { "grad_norm": 1.0331710577011108, "learning_rate": 2.6900000000000003e-05, "loss": 0.6444, "step": 270 }, { "grad_norm": 1.2573391199111938, "learning_rate": 2.7900000000000004e-05, "loss": 0.618, "step": 280 }, { "grad_norm": 1.2672662734985352, "learning_rate": 2.8899999999999998e-05, "loss": 0.6095, "step": 290 }, { "grad_norm": 1.0279985666275024, "learning_rate": 2.9900000000000002e-05, "loss": 0.5921, "step": 300 }, { "grad_norm": 0.9896909594535828, "learning_rate": 3.09e-05, "loss": 0.5765, "step": 310 }, { "grad_norm": 1.0756268501281738, "learning_rate": 3.19e-05, "loss": 0.5679, "step": 320 }, { "grad_norm": 1.3084145784378052, "learning_rate": 3.29e-05, "loss": 0.5569, "step": 330 }, { "grad_norm": 1.2895151376724243, "learning_rate": 3.3900000000000004e-05, "loss": 0.5475, "step": 340 }, { "grad_norm": 1.1118186712265015, "learning_rate": 3.49e-05, "loss": 0.5367, "step": 350 }, { "grad_norm": 0.9377628564834595, "learning_rate": 3.59e-05, "loss": 0.5192, "step": 360 }, { "grad_norm": 1.1136053800582886, "learning_rate": 3.69e-05, "loss": 0.5096, "step": 370 }, { "grad_norm": 0.8958713412284851, "learning_rate": 3.79e-05, "loss": 0.4938, "step": 380 }, { "grad_norm": 1.4132113456726074, "learning_rate": 3.8900000000000004e-05, "loss": 0.4777, "step": 390 }, { "grad_norm": 1.250508189201355, "learning_rate": 3.99e-05, "loss": 0.4575, "step": 400 }, { "grad_norm": 1.2522258758544922, "learning_rate": 4.09e-05, "loss": 0.4386, "step": 410 }, { "grad_norm": 1.4475443363189697, "learning_rate": 4.19e-05, "loss": 0.4295, "step": 420 }, { "grad_norm": 0.925055742263794, "learning_rate": 4.29e-05, "loss": 0.4157, "step": 430 }, { "grad_norm": 1.2969681024551392, "learning_rate": 4.39e-05, "loss": 0.4052, "step": 440 }, { "grad_norm": 1.3583041429519653, "learning_rate": 4.49e-05, "loss": 0.3953, "step": 450 }, { "grad_norm": 1.028383493423462, "learning_rate": 4.5900000000000004e-05, "loss": 0.375, "step": 460 }, { "grad_norm": 1.255179762840271, "learning_rate": 4.69e-05, "loss": 0.3699, "step": 470 }, { "grad_norm": 1.326106309890747, "learning_rate": 4.79e-05, "loss": 0.3579, "step": 480 }, { "grad_norm": 1.2961987257003784, "learning_rate": 4.89e-05, "loss": 0.3478, "step": 490 }, { "grad_norm": 1.2104370594024658, "learning_rate": 4.99e-05, "loss": 0.339, "step": 500 }, { "grad_norm": 1.4508124589920044, "learning_rate": 5.0900000000000004e-05, "loss": 0.3219, "step": 510 }, { "grad_norm": 1.4312357902526855, "learning_rate": 5.19e-05, "loss": 0.3042, "step": 520 }, { "grad_norm": 1.7213497161865234, "learning_rate": 5.2900000000000005e-05, "loss": 0.2887, "step": 530 }, { "grad_norm": 1.7162240743637085, "learning_rate": 5.390000000000001e-05, "loss": 0.2856, "step": 540 }, { "grad_norm": 1.1437758207321167, "learning_rate": 5.4900000000000006e-05, "loss": 0.2642, "step": 550 }, { "grad_norm": 1.4461580514907837, "learning_rate": 5.590000000000001e-05, "loss": 0.2508, "step": 560 }, { "grad_norm": 1.4058153629302979, "learning_rate": 5.69e-05, "loss": 0.2437, "step": 570 }, { "grad_norm": 1.3282971382141113, "learning_rate": 5.79e-05, "loss": 0.2275, "step": 580 }, { "grad_norm": 1.6089696884155273, "learning_rate": 5.89e-05, "loss": 0.2166, "step": 590 }, { "grad_norm": 1.5542049407958984, "learning_rate": 5.99e-05, "loss": 0.2008, "step": 600 }, { "grad_norm": 1.372789740562439, "learning_rate": 6.09e-05, "loss": 0.1888, "step": 610 }, { "grad_norm": 1.3900244235992432, "learning_rate": 6.19e-05, "loss": 0.1825, "step": 620 }, { "grad_norm": 1.1616129875183105, "learning_rate": 6.29e-05, "loss": 0.1673, "step": 630 }, { "grad_norm": 1.3904050588607788, "learning_rate": 6.390000000000001e-05, "loss": 0.1631, "step": 640 }, { "grad_norm": 1.6666728258132935, "learning_rate": 6.49e-05, "loss": 0.1508, "step": 650 }, { "grad_norm": 1.458060383796692, "learning_rate": 6.59e-05, "loss": 0.134, "step": 660 }, { "grad_norm": 1.2632263898849487, "learning_rate": 6.690000000000001e-05, "loss": 0.1435, "step": 670 }, { "grad_norm": 1.3523911237716675, "learning_rate": 6.790000000000001e-05, "loss": 0.1383, "step": 680 }, { "grad_norm": 1.3925788402557373, "learning_rate": 6.89e-05, "loss": 0.1317, "step": 690 }, { "grad_norm": 1.6858913898468018, "learning_rate": 6.99e-05, "loss": 0.1309, "step": 700 }, { "grad_norm": 1.2360647916793823, "learning_rate": 7.09e-05, "loss": 0.1267, "step": 710 }, { "grad_norm": 1.1521002054214478, "learning_rate": 7.19e-05, "loss": 0.1259, "step": 720 }, { "grad_norm": 1.232092022895813, "learning_rate": 7.29e-05, "loss": 0.1333, "step": 730 }, { "grad_norm": 1.4635534286499023, "learning_rate": 7.390000000000001e-05, "loss": 0.129, "step": 740 }, { "grad_norm": 1.108782172203064, "learning_rate": 7.49e-05, "loss": 0.1285, "step": 750 }, { "grad_norm": 1.3770654201507568, "learning_rate": 7.59e-05, "loss": 0.1247, "step": 760 }, { "grad_norm": 1.1716821193695068, "learning_rate": 7.69e-05, "loss": 0.1297, "step": 770 }, { "grad_norm": 1.185289740562439, "learning_rate": 7.790000000000001e-05, "loss": 0.1246, "step": 780 }, { "grad_norm": 1.3426638841629028, "learning_rate": 7.890000000000001e-05, "loss": 0.1209, "step": 790 }, { "grad_norm": 1.0914347171783447, "learning_rate": 7.99e-05, "loss": 0.1226, "step": 800 }, { "grad_norm": 1.47992742061615, "learning_rate": 8.090000000000001e-05, "loss": 0.1116, "step": 810 }, { "grad_norm": 1.189318299293518, "learning_rate": 8.19e-05, "loss": 0.1155, "step": 820 }, { "grad_norm": 1.218897819519043, "learning_rate": 8.29e-05, "loss": 0.1106, "step": 830 }, { "grad_norm": 1.1049423217773438, "learning_rate": 8.39e-05, "loss": 0.1143, "step": 840 }, { "grad_norm": 1.3933051824569702, "learning_rate": 8.49e-05, "loss": 0.1241, "step": 850 }, { "grad_norm": 1.0603047609329224, "learning_rate": 8.59e-05, "loss": 0.1111, "step": 860 }, { "grad_norm": 1.1517871618270874, "learning_rate": 8.69e-05, "loss": 0.1114, "step": 870 }, { "grad_norm": 1.2335821390151978, "learning_rate": 8.790000000000001e-05, "loss": 0.1135, "step": 880 }, { "grad_norm": 1.471954107284546, "learning_rate": 8.89e-05, "loss": 0.1149, "step": 890 }, { "grad_norm": 1.1479299068450928, "learning_rate": 8.99e-05, "loss": 0.1144, "step": 900 }, { "grad_norm": 1.1517722606658936, "learning_rate": 9.090000000000001e-05, "loss": 0.1097, "step": 910 }, { "grad_norm": 1.2328661680221558, "learning_rate": 9.190000000000001e-05, "loss": 0.11, "step": 920 }, { "grad_norm": 1.227306842803955, "learning_rate": 9.290000000000001e-05, "loss": 0.1144, "step": 930 }, { "grad_norm": 1.1643450260162354, "learning_rate": 9.39e-05, "loss": 0.1074, "step": 940 }, { "grad_norm": 1.1138417720794678, "learning_rate": 9.49e-05, "loss": 0.113, "step": 950 }, { "grad_norm": 1.0035343170166016, "learning_rate": 9.59e-05, "loss": 0.1102, "step": 960 }, { "grad_norm": 1.089536190032959, "learning_rate": 9.69e-05, "loss": 0.1089, "step": 970 }, { "grad_norm": 1.1418471336364746, "learning_rate": 9.790000000000001e-05, "loss": 0.1075, "step": 980 }, { "grad_norm": 1.0369817018508911, "learning_rate": 9.89e-05, "loss": 0.1094, "step": 990 }, { "grad_norm": 1.081180214881897, "learning_rate": 9.99e-05, "loss": 0.1069, "step": 1000 }, { "grad_norm": 1.1965157985687256, "learning_rate": 9.999994463727085e-05, "loss": 0.1078, "step": 1010 }, { "grad_norm": 1.0836145877838135, "learning_rate": 9.999975326009292e-05, "loss": 0.1034, "step": 1020 }, { "grad_norm": 1.0701448917388916, "learning_rate": 9.999942518549879e-05, "loss": 0.1063, "step": 1030 }, { "grad_norm": 1.0245521068572998, "learning_rate": 9.999896041438544e-05, "loss": 0.1003, "step": 1040 }, { "grad_norm": 0.9008662104606628, "learning_rate": 9.999835894802353e-05, "loss": 0.1007, "step": 1050 }, { "grad_norm": 1.405444860458374, "learning_rate": 9.999762078805743e-05, "loss": 0.0983, "step": 1060 }, { "grad_norm": 0.9956735968589783, "learning_rate": 9.999674593650526e-05, "loss": 0.1015, "step": 1070 }, { "grad_norm": 0.9828646779060364, "learning_rate": 9.99957343957588e-05, "loss": 0.099, "step": 1080 }, { "grad_norm": 1.3071095943450928, "learning_rate": 9.99945861685836e-05, "loss": 0.0949, "step": 1090 }, { "grad_norm": 0.9832029342651367, "learning_rate": 9.999330125811884e-05, "loss": 0.102, "step": 1100 }, { "grad_norm": 1.0148258209228516, "learning_rate": 9.999187966787744e-05, "loss": 0.103, "step": 1110 }, { "grad_norm": 1.0134457349777222, "learning_rate": 9.999032140174595e-05, "loss": 0.0949, "step": 1120 }, { "grad_norm": 0.9154807329177856, "learning_rate": 9.998862646398464e-05, "loss": 0.0989, "step": 1130 }, { "grad_norm": 1.0683287382125854, "learning_rate": 9.998679485922739e-05, "loss": 0.0937, "step": 1140 }, { "grad_norm": 0.9443909525871277, "learning_rate": 9.998482659248174e-05, "loss": 0.0964, "step": 1150 }, { "grad_norm": 0.8048564195632935, "learning_rate": 9.998272166912883e-05, "loss": 0.0978, "step": 1160 }, { "grad_norm": 1.0226809978485107, "learning_rate": 9.998048009492347e-05, "loss": 0.0979, "step": 1170 }, { "grad_norm": 1.0117919445037842, "learning_rate": 9.997810187599403e-05, "loss": 0.0993, "step": 1180 }, { "grad_norm": 0.8559236526489258, "learning_rate": 9.997558701884249e-05, "loss": 0.095, "step": 1190 }, { "grad_norm": 1.009264349937439, "learning_rate": 9.997293553034433e-05, "loss": 0.0947, "step": 1200 }, { "grad_norm": 1.0090675354003906, "learning_rate": 9.997014741774866e-05, "loss": 0.0991, "step": 1210 }, { "grad_norm": 0.9128223657608032, "learning_rate": 9.996722268867803e-05, "loss": 0.0973, "step": 1220 }, { "grad_norm": 0.8223145008087158, "learning_rate": 9.996416135112858e-05, "loss": 0.0943, "step": 1230 }, { "grad_norm": 0.6892135143280029, "learning_rate": 9.996096341346988e-05, "loss": 0.0914, "step": 1240 }, { "grad_norm": 0.8441371321678162, "learning_rate": 9.995762888444495e-05, "loss": 0.095, "step": 1250 }, { "grad_norm": 0.8410964608192444, "learning_rate": 9.995415777317027e-05, "loss": 0.0905, "step": 1260 }, { "grad_norm": 0.9202972650527954, "learning_rate": 9.995055008913574e-05, "loss": 0.09, "step": 1270 }, { "grad_norm": 0.969850480556488, "learning_rate": 9.994680584220463e-05, "loss": 0.0977, "step": 1280 }, { "grad_norm": 0.9843644499778748, "learning_rate": 9.994292504261355e-05, "loss": 0.0903, "step": 1290 }, { "grad_norm": 0.8046810030937195, "learning_rate": 9.993890770097247e-05, "loss": 0.0885, "step": 1300 }, { "grad_norm": 1.0213329792022705, "learning_rate": 9.993475382826467e-05, "loss": 0.0906, "step": 1310 }, { "grad_norm": 0.9399948120117188, "learning_rate": 9.993046343584664e-05, "loss": 0.0885, "step": 1320 }, { "grad_norm": 0.8615891337394714, "learning_rate": 9.992603653544816e-05, "loss": 0.0898, "step": 1330 }, { "grad_norm": 0.7804251909255981, "learning_rate": 9.992147313917222e-05, "loss": 0.0879, "step": 1340 }, { "grad_norm": 0.7505974173545837, "learning_rate": 9.991677325949497e-05, "loss": 0.0877, "step": 1350 }, { "grad_norm": 0.8564302921295166, "learning_rate": 9.991193690926568e-05, "loss": 0.0853, "step": 1360 }, { "grad_norm": 0.921603798866272, "learning_rate": 9.990696410170678e-05, "loss": 0.0868, "step": 1370 }, { "grad_norm": 0.905284583568573, "learning_rate": 9.990185485041371e-05, "loss": 0.0812, "step": 1380 }, { "grad_norm": 0.8342713713645935, "learning_rate": 9.989660916935498e-05, "loss": 0.0845, "step": 1390 }, { "grad_norm": 0.8646072149276733, "learning_rate": 9.989122707287208e-05, "loss": 0.0863, "step": 1400 }, { "grad_norm": 1.0666006803512573, "learning_rate": 9.988570857567945e-05, "loss": 0.0887, "step": 1410 }, { "grad_norm": 0.8190611004829407, "learning_rate": 9.988005369286446e-05, "loss": 0.0845, "step": 1420 }, { "grad_norm": 0.8214516639709473, "learning_rate": 9.987426243988734e-05, "loss": 0.0883, "step": 1430 }, { "grad_norm": 0.7905929684638977, "learning_rate": 9.986833483258114e-05, "loss": 0.0861, "step": 1440 }, { "grad_norm": 0.8426095247268677, "learning_rate": 9.986227088715173e-05, "loss": 0.0823, "step": 1450 }, { "grad_norm": 0.9610787630081177, "learning_rate": 9.98560706201777e-05, "loss": 0.0835, "step": 1460 }, { "grad_norm": 0.8222280740737915, "learning_rate": 9.984973404861036e-05, "loss": 0.0878, "step": 1470 }, { "grad_norm": 0.840451717376709, "learning_rate": 9.984326118977361e-05, "loss": 0.0812, "step": 1480 }, { "grad_norm": 0.7655553817749023, "learning_rate": 9.983665206136406e-05, "loss": 0.0784, "step": 1490 }, { "grad_norm": 0.9331004023551941, "learning_rate": 9.982990668145075e-05, "loss": 0.0779, "step": 1500 }, { "grad_norm": 0.7345725297927856, "learning_rate": 9.982302506847534e-05, "loss": 0.0792, "step": 1510 }, { "grad_norm": 0.8967569470405579, "learning_rate": 9.981600724125189e-05, "loss": 0.0782, "step": 1520 }, { "grad_norm": 0.7933437824249268, "learning_rate": 9.980885321896685e-05, "loss": 0.0773, "step": 1530 }, { "grad_norm": 0.6864101886749268, "learning_rate": 9.980156302117905e-05, "loss": 0.0773, "step": 1540 }, { "grad_norm": 0.689123272895813, "learning_rate": 9.979413666781963e-05, "loss": 0.0824, "step": 1550 }, { "grad_norm": 0.8207864165306091, "learning_rate": 9.978657417919193e-05, "loss": 0.0803, "step": 1560 }, { "grad_norm": 0.7813171744346619, "learning_rate": 9.977887557597153e-05, "loss": 0.0817, "step": 1570 }, { "grad_norm": 0.695106565952301, "learning_rate": 9.97710408792061e-05, "loss": 0.0783, "step": 1580 }, { "grad_norm": 0.9048652052879333, "learning_rate": 9.976307011031542e-05, "loss": 0.0788, "step": 1590 }, { "grad_norm": 0.8499309420585632, "learning_rate": 9.975496329109126e-05, "loss": 0.081, "step": 1600 }, { "grad_norm": 0.8718334436416626, "learning_rate": 9.974672044369732e-05, "loss": 0.0777, "step": 1610 }, { "grad_norm": 0.6900801658630371, "learning_rate": 9.97383415906693e-05, "loss": 0.077, "step": 1620 }, { "grad_norm": 0.8092792630195618, "learning_rate": 9.97298267549146e-05, "loss": 0.0788, "step": 1630 }, { "grad_norm": 0.8192611932754517, "learning_rate": 9.972117595971249e-05, "loss": 0.0799, "step": 1640 }, { "grad_norm": 0.8722266554832458, "learning_rate": 9.971238922871391e-05, "loss": 0.0819, "step": 1650 }, { "grad_norm": 0.914159893989563, "learning_rate": 9.970346658594142e-05, "loss": 0.0838, "step": 1660 }, { "grad_norm": 0.8131964206695557, "learning_rate": 9.969440805578923e-05, "loss": 0.0797, "step": 1670 }, { "grad_norm": 0.7147591710090637, "learning_rate": 9.968521366302298e-05, "loss": 0.0799, "step": 1680 }, { "grad_norm": 0.8249319791793823, "learning_rate": 9.967588343277981e-05, "loss": 0.0783, "step": 1690 }, { "grad_norm": 0.7327215075492859, "learning_rate": 9.966641739056818e-05, "loss": 0.0787, "step": 1700 }, { "grad_norm": 0.7018980979919434, "learning_rate": 9.965681556226793e-05, "loss": 0.0705, "step": 1710 }, { "grad_norm": 0.8972001671791077, "learning_rate": 9.964707797413006e-05, "loss": 0.0796, "step": 1720 }, { "grad_norm": 0.8128800392150879, "learning_rate": 9.963720465277679e-05, "loss": 0.0818, "step": 1730 }, { "grad_norm": 0.9391877055168152, "learning_rate": 9.96271956252014e-05, "loss": 0.0758, "step": 1740 }, { "grad_norm": 0.8188132047653198, "learning_rate": 9.961705091876816e-05, "loss": 0.0756, "step": 1750 }, { "grad_norm": 0.8037185668945312, "learning_rate": 9.960677056121235e-05, "loss": 0.0783, "step": 1760 }, { "grad_norm": 0.6627317070960999, "learning_rate": 9.959635458064005e-05, "loss": 0.0773, "step": 1770 }, { "grad_norm": 0.6626058220863342, "learning_rate": 9.958580300552815e-05, "loss": 0.0775, "step": 1780 }, { "grad_norm": 0.7204414010047913, "learning_rate": 9.957511586472426e-05, "loss": 0.0762, "step": 1790 }, { "grad_norm": 0.6517882347106934, "learning_rate": 9.956429318744662e-05, "loss": 0.0776, "step": 1800 }, { "grad_norm": 0.7082467675209045, "learning_rate": 9.955333500328404e-05, "loss": 0.0737, "step": 1810 }, { "grad_norm": 0.7636476159095764, "learning_rate": 9.95422413421957e-05, "loss": 0.0733, "step": 1820 }, { "grad_norm": 0.72133868932724, "learning_rate": 9.953101223451133e-05, "loss": 0.0749, "step": 1830 }, { "grad_norm": 0.7455642223358154, "learning_rate": 9.951964771093085e-05, "loss": 0.074, "step": 1840 }, { "grad_norm": 0.6844678521156311, "learning_rate": 9.950814780252442e-05, "loss": 0.0773, "step": 1850 }, { "grad_norm": 0.49062681198120117, "learning_rate": 9.949651254073236e-05, "loss": 0.0732, "step": 1860 }, { "grad_norm": 0.789148211479187, "learning_rate": 9.948474195736504e-05, "loss": 0.0718, "step": 1870 }, { "grad_norm": 0.6764615178108215, "learning_rate": 9.947283608460277e-05, "loss": 0.0742, "step": 1880 }, { "grad_norm": 0.6999238133430481, "learning_rate": 9.946079495499577e-05, "loss": 0.0747, "step": 1890 }, { "grad_norm": 0.7517783045768738, "learning_rate": 9.944861860146401e-05, "loss": 0.0753, "step": 1900 }, { "grad_norm": 0.7183238863945007, "learning_rate": 9.943630705729719e-05, "loss": 0.0697, "step": 1910 }, { "grad_norm": 0.5847482681274414, "learning_rate": 9.942386035615459e-05, "loss": 0.072, "step": 1920 }, { "grad_norm": 0.5976929068565369, "learning_rate": 9.941127853206503e-05, "loss": 0.0694, "step": 1930 }, { "grad_norm": 0.5992315411567688, "learning_rate": 9.939856161942673e-05, "loss": 0.0729, "step": 1940 }, { "grad_norm": 0.6920360922813416, "learning_rate": 9.938570965300724e-05, "loss": 0.0693, "step": 1950 }, { "grad_norm": 0.7227588891983032, "learning_rate": 9.937272266794335e-05, "loss": 0.0779, "step": 1960 }, { "grad_norm": 0.6451845169067383, "learning_rate": 9.935960069974096e-05, "loss": 0.0705, "step": 1970 }, { "grad_norm": 0.7243984341621399, "learning_rate": 9.934634378427506e-05, "loss": 0.0708, "step": 1980 }, { "grad_norm": 0.6210780143737793, "learning_rate": 9.933295195778954e-05, "loss": 0.0772, "step": 1990 }, { "grad_norm": 0.5807631611824036, "learning_rate": 9.931942525689715e-05, "loss": 0.0745, "step": 2000 }, { "grad_norm": 0.6609013676643372, "learning_rate": 9.930576371857936e-05, "loss": 0.0723, "step": 2010 }, { "grad_norm": 0.7919302582740784, "learning_rate": 9.929196738018629e-05, "loss": 0.0702, "step": 2020 }, { "grad_norm": 0.6295292973518372, "learning_rate": 9.927803627943662e-05, "loss": 0.0674, "step": 2030 }, { "grad_norm": 0.756155788898468, "learning_rate": 9.926397045441744e-05, "loss": 0.0687, "step": 2040 }, { "grad_norm": 0.7688308954238892, "learning_rate": 9.924976994358417e-05, "loss": 0.0718, "step": 2050 }, { "grad_norm": 0.8509033918380737, "learning_rate": 9.923543478576048e-05, "loss": 0.0727, "step": 2060 }, { "grad_norm": 0.7590917944908142, "learning_rate": 9.922096502013813e-05, "loss": 0.074, "step": 2070 }, { "grad_norm": 0.8109246492385864, "learning_rate": 9.92063606862769e-05, "loss": 0.0721, "step": 2080 }, { "grad_norm": 0.6400628685951233, "learning_rate": 9.919162182410453e-05, "loss": 0.0707, "step": 2090 }, { "grad_norm": 0.5556660890579224, "learning_rate": 9.917674847391645e-05, "loss": 0.0663, "step": 2100 }, { "grad_norm": 0.7919337749481201, "learning_rate": 9.916174067637584e-05, "loss": 0.0675, "step": 2110 }, { "grad_norm": 0.6571515798568726, "learning_rate": 9.914659847251348e-05, "loss": 0.0683, "step": 2120 }, { "grad_norm": 0.5666832327842712, "learning_rate": 9.913132190372753e-05, "loss": 0.0706, "step": 2130 }, { "grad_norm": 0.6068550944328308, "learning_rate": 9.911591101178359e-05, "loss": 0.067, "step": 2140 }, { "grad_norm": 0.5472414493560791, "learning_rate": 9.910036583881443e-05, "loss": 0.0686, "step": 2150 }, { "grad_norm": 0.6243656277656555, "learning_rate": 9.908468642731995e-05, "loss": 0.0663, "step": 2160 }, { "grad_norm": 0.685313880443573, "learning_rate": 9.906887282016707e-05, "loss": 0.0704, "step": 2170 }, { "grad_norm": 0.6772751212120056, "learning_rate": 9.90529250605896e-05, "loss": 0.0651, "step": 2180 }, { "grad_norm": 0.6815426349639893, "learning_rate": 9.903684319218809e-05, "loss": 0.0728, "step": 2190 }, { "grad_norm": 0.822087287902832, "learning_rate": 9.902062725892976e-05, "loss": 0.0723, "step": 2200 }, { "grad_norm": 0.6740921139717102, "learning_rate": 9.900427730514834e-05, "loss": 0.0722, "step": 2210 }, { "grad_norm": 0.730047345161438, "learning_rate": 9.8987793375544e-05, "loss": 0.0717, "step": 2220 }, { "grad_norm": 0.8769873976707458, "learning_rate": 9.897117551518318e-05, "loss": 0.0672, "step": 2230 }, { "grad_norm": 0.5988404750823975, "learning_rate": 9.895442376949844e-05, "loss": 0.0691, "step": 2240 }, { "grad_norm": 0.7594494223594666, "learning_rate": 9.893753818428845e-05, "loss": 0.0703, "step": 2250 }, { "grad_norm": 0.7655356526374817, "learning_rate": 9.892051880571773e-05, "loss": 0.0643, "step": 2260 }, { "grad_norm": 0.6700913906097412, "learning_rate": 9.890336568031663e-05, "loss": 0.066, "step": 2270 }, { "grad_norm": 0.6776233315467834, "learning_rate": 9.888607885498113e-05, "loss": 0.0713, "step": 2280 }, { "grad_norm": 0.6251364350318909, "learning_rate": 9.886865837697275e-05, "loss": 0.0675, "step": 2290 }, { "grad_norm": 0.6322036981582642, "learning_rate": 9.88511042939184e-05, "loss": 0.07, "step": 2300 }, { "grad_norm": 0.6706236600875854, "learning_rate": 9.883341665381028e-05, "loss": 0.0707, "step": 2310 }, { "grad_norm": 0.7654581069946289, "learning_rate": 9.881559550500575e-05, "loss": 0.0684, "step": 2320 }, { "grad_norm": 0.6615346074104309, "learning_rate": 9.879764089622712e-05, "loss": 0.0664, "step": 2330 }, { "grad_norm": 0.6900296807289124, "learning_rate": 9.87795528765616e-05, "loss": 0.0703, "step": 2340 }, { "grad_norm": 0.6962850689888, "learning_rate": 9.876133149546118e-05, "loss": 0.0712, "step": 2350 }, { "grad_norm": 0.5966289639472961, "learning_rate": 9.874297680274238e-05, "loss": 0.0705, "step": 2360 }, { "grad_norm": 0.5869346261024475, "learning_rate": 9.872448884858624e-05, "loss": 0.0673, "step": 2370 }, { "grad_norm": 0.6729479432106018, "learning_rate": 9.870586768353815e-05, "loss": 0.0698, "step": 2380 }, { "grad_norm": 0.4968750476837158, "learning_rate": 9.868711335850764e-05, "loss": 0.071, "step": 2390 }, { "grad_norm": 0.5750929713249207, "learning_rate": 9.866822592476833e-05, "loss": 0.066, "step": 2400 }, { "grad_norm": 0.6520802974700928, "learning_rate": 9.86492054339577e-05, "loss": 0.0704, "step": 2410 }, { "grad_norm": 0.6432913541793823, "learning_rate": 9.863005193807711e-05, "loss": 0.0669, "step": 2420 }, { "grad_norm": 0.7633799314498901, "learning_rate": 9.861076548949143e-05, "loss": 0.0641, "step": 2430 }, { "grad_norm": 0.727251410484314, "learning_rate": 9.859134614092912e-05, "loss": 0.0666, "step": 2440 }, { "grad_norm": 0.6507017016410828, "learning_rate": 9.857179394548191e-05, "loss": 0.0631, "step": 2450 }, { "grad_norm": 0.802526593208313, "learning_rate": 9.855210895660477e-05, "loss": 0.0695, "step": 2460 }, { "grad_norm": 0.4968661069869995, "learning_rate": 9.853229122811568e-05, "loss": 0.0668, "step": 2470 }, { "grad_norm": 0.5049183964729309, "learning_rate": 9.851234081419559e-05, "loss": 0.0679, "step": 2480 }, { "grad_norm": 0.5562856793403625, "learning_rate": 9.849225776938814e-05, "loss": 0.067, "step": 2490 }, { "grad_norm": 0.578382134437561, "learning_rate": 9.847204214859964e-05, "loss": 0.0641, "step": 2500 }, { "grad_norm": 0.6934925317764282, "learning_rate": 9.845169400709879e-05, "loss": 0.0677, "step": 2510 }, { "grad_norm": 0.729125440120697, "learning_rate": 9.843121340051664e-05, "loss": 0.0637, "step": 2520 }, { "grad_norm": 0.6229711174964905, "learning_rate": 9.841060038484641e-05, "loss": 0.064, "step": 2530 }, { "grad_norm": 0.6090768575668335, "learning_rate": 9.838985501644328e-05, "loss": 0.0623, "step": 2540 }, { "grad_norm": 0.8319109082221985, "learning_rate": 9.83689773520243e-05, "loss": 0.0669, "step": 2550 }, { "grad_norm": 0.6815285086631775, "learning_rate": 9.834796744866819e-05, "loss": 0.0661, "step": 2560 }, { "grad_norm": 0.49073100090026855, "learning_rate": 9.832682536381525e-05, "loss": 0.0686, "step": 2570 }, { "grad_norm": 0.5500509142875671, "learning_rate": 9.830555115526711e-05, "loss": 0.0663, "step": 2580 }, { "grad_norm": 0.5543439388275146, "learning_rate": 9.828414488118667e-05, "loss": 0.0586, "step": 2590 }, { "grad_norm": 0.5412293672561646, "learning_rate": 9.826260660009785e-05, "loss": 0.065, "step": 2600 }, { "grad_norm": 0.46607404947280884, "learning_rate": 9.824093637088547e-05, "loss": 0.0643, "step": 2610 }, { "grad_norm": 0.5089811682701111, "learning_rate": 9.821913425279514e-05, "loss": 0.0627, "step": 2620 }, { "grad_norm": 0.6004795432090759, "learning_rate": 9.8197200305433e-05, "loss": 0.0648, "step": 2630 }, { "grad_norm": 0.5946650505065918, "learning_rate": 9.817513458876564e-05, "loss": 0.0634, "step": 2640 }, { "grad_norm": 0.7084515690803528, "learning_rate": 9.815293716311987e-05, "loss": 0.0653, "step": 2650 }, { "grad_norm": 0.6806345582008362, "learning_rate": 9.813060808918262e-05, "loss": 0.0639, "step": 2660 }, { "grad_norm": 0.6408096551895142, "learning_rate": 9.810814742800069e-05, "loss": 0.0641, "step": 2670 }, { "grad_norm": 0.5025924444198608, "learning_rate": 9.808555524098074e-05, "loss": 0.063, "step": 2680 }, { "grad_norm": 0.7290829420089722, "learning_rate": 9.806283158988887e-05, "loss": 0.0647, "step": 2690 }, { "grad_norm": 0.5606433153152466, "learning_rate": 9.803997653685072e-05, "loss": 0.062, "step": 2700 }, { "grad_norm": 0.5064647793769836, "learning_rate": 9.801699014435112e-05, "loss": 0.0643, "step": 2710 }, { "grad_norm": 0.5947048664093018, "learning_rate": 9.799387247523398e-05, "loss": 0.06, "step": 2720 }, { "grad_norm": 0.6373162269592285, "learning_rate": 9.797062359270215e-05, "loss": 0.0596, "step": 2730 }, { "grad_norm": 0.683765709400177, "learning_rate": 9.794724356031715e-05, "loss": 0.0641, "step": 2740 }, { "grad_norm": 0.6411890983581543, "learning_rate": 9.792373244199913e-05, "loss": 0.0687, "step": 2750 }, { "grad_norm": 0.4775662422180176, "learning_rate": 9.790009030202658e-05, "loss": 0.0622, "step": 2760 }, { "grad_norm": 0.651835560798645, "learning_rate": 9.78763172050362e-05, "loss": 0.0621, "step": 2770 }, { "grad_norm": 0.6854885220527649, "learning_rate": 9.785241321602274e-05, "loss": 0.0667, "step": 2780 }, { "grad_norm": 0.6907970309257507, "learning_rate": 9.782837840033879e-05, "loss": 0.0669, "step": 2790 }, { "grad_norm": 0.5823140740394592, "learning_rate": 9.780421282369461e-05, "loss": 0.0678, "step": 2800 }, { "grad_norm": 0.6364890933036804, "learning_rate": 9.777991655215797e-05, "loss": 0.0663, "step": 2810 }, { "grad_norm": 0.5718041658401489, "learning_rate": 9.775548965215394e-05, "loss": 0.0626, "step": 2820 }, { "grad_norm": 0.4957229793071747, "learning_rate": 9.773093219046474e-05, "loss": 0.0634, "step": 2830 }, { "grad_norm": 0.6113382577896118, "learning_rate": 9.770624423422954e-05, "loss": 0.0665, "step": 2840 }, { "grad_norm": 0.6246403455734253, "learning_rate": 9.768142585094426e-05, "loss": 0.0626, "step": 2850 }, { "grad_norm": 0.6030095219612122, "learning_rate": 9.765647710846142e-05, "loss": 0.0662, "step": 2860 }, { "grad_norm": 0.5451913475990295, "learning_rate": 9.763139807498991e-05, "loss": 0.0692, "step": 2870 }, { "grad_norm": 0.5483195185661316, "learning_rate": 9.760618881909487e-05, "loss": 0.0599, "step": 2880 }, { "grad_norm": 0.47611215710639954, "learning_rate": 9.758084940969744e-05, "loss": 0.0635, "step": 2890 }, { "grad_norm": 0.5544406175613403, "learning_rate": 9.755537991607459e-05, "loss": 0.0592, "step": 2900 }, { "grad_norm": 0.6410639882087708, "learning_rate": 9.752978040785895e-05, "loss": 0.0613, "step": 2910 }, { "grad_norm": 0.6519464254379272, "learning_rate": 9.750405095503859e-05, "loss": 0.0584, "step": 2920 }, { "grad_norm": 0.6715825200080872, "learning_rate": 9.747819162795686e-05, "loss": 0.064, "step": 2930 }, { "grad_norm": 0.5294116139411926, "learning_rate": 9.745220249731217e-05, "loss": 0.0617, "step": 2940 }, { "grad_norm": 0.5013177394866943, "learning_rate": 9.742608363415781e-05, "loss": 0.0579, "step": 2950 }, { "grad_norm": 0.7086690664291382, "learning_rate": 9.739983510990176e-05, "loss": 0.0627, "step": 2960 }, { "grad_norm": 0.6349759697914124, "learning_rate": 9.737345699630647e-05, "loss": 0.0676, "step": 2970 }, { "grad_norm": 0.6648471355438232, "learning_rate": 9.734694936548869e-05, "loss": 0.0642, "step": 2980 }, { "grad_norm": 0.4997161328792572, "learning_rate": 9.732031228991932e-05, "loss": 0.0603, "step": 2990 }, { "grad_norm": 0.4930708408355713, "learning_rate": 9.729354584242302e-05, "loss": 0.0596, "step": 3000 }, { "grad_norm": 0.6939645409584045, "learning_rate": 9.726665009617832e-05, "loss": 0.0627, "step": 3010 }, { "grad_norm": 0.6745036244392395, "learning_rate": 9.723962512471714e-05, "loss": 0.0622, "step": 3020 }, { "grad_norm": 0.6208877563476562, "learning_rate": 9.72124710019247e-05, "loss": 0.0599, "step": 3030 }, { "grad_norm": 0.615725576877594, "learning_rate": 9.718518780203934e-05, "loss": 0.0611, "step": 3040 }, { "grad_norm": 0.5784436464309692, "learning_rate": 9.715777559965228e-05, "loss": 0.0598, "step": 3050 }, { "grad_norm": 0.6067401766777039, "learning_rate": 9.713023446970746e-05, "loss": 0.0647, "step": 3060 }, { "grad_norm": 0.5806649327278137, "learning_rate": 9.710256448750126e-05, "loss": 0.0625, "step": 3070 }, { "grad_norm": 0.6504300236701965, "learning_rate": 9.707476572868235e-05, "loss": 0.058, "step": 3080 }, { "grad_norm": 0.500622570514679, "learning_rate": 9.704683826925149e-05, "loss": 0.0577, "step": 3090 }, { "grad_norm": 0.6325183510780334, "learning_rate": 9.701878218556129e-05, "loss": 0.0603, "step": 3100 }, { "grad_norm": 0.555566132068634, "learning_rate": 9.699059755431598e-05, "loss": 0.0616, "step": 3110 }, { "grad_norm": 0.4982530474662781, "learning_rate": 9.696228445257132e-05, "loss": 0.0594, "step": 3120 }, { "grad_norm": 0.5743557214736938, "learning_rate": 9.693384295773419e-05, "loss": 0.056, "step": 3130 }, { "grad_norm": 0.4599805772304535, "learning_rate": 9.690527314756259e-05, "loss": 0.0551, "step": 3140 }, { "grad_norm": 0.6725175976753235, "learning_rate": 9.687657510016527e-05, "loss": 0.0586, "step": 3150 }, { "grad_norm": 0.5373369455337524, "learning_rate": 9.684774889400161e-05, "loss": 0.0576, "step": 3160 }, { "grad_norm": 0.6253339648246765, "learning_rate": 9.681879460788135e-05, "loss": 0.0586, "step": 3170 }, { "grad_norm": 0.6447578072547913, "learning_rate": 9.67897123209644e-05, "loss": 0.0639, "step": 3180 }, { "grad_norm": 0.44229140877723694, "learning_rate": 9.676050211276062e-05, "loss": 0.0606, "step": 3190 }, { "grad_norm": 0.6652916669845581, "learning_rate": 9.673116406312962e-05, "loss": 0.0629, "step": 3200 }, { "grad_norm": 0.7552599906921387, "learning_rate": 9.67016982522805e-05, "loss": 0.0666, "step": 3210 }, { "grad_norm": 0.6908206939697266, "learning_rate": 9.667210476077164e-05, "loss": 0.0606, "step": 3220 }, { "grad_norm": 0.5687106251716614, "learning_rate": 9.664238366951055e-05, "loss": 0.0599, "step": 3230 }, { "grad_norm": 0.563493549823761, "learning_rate": 9.661253505975355e-05, "loss": 0.0569, "step": 3240 }, { "grad_norm": 0.6353240609169006, "learning_rate": 9.658255901310557e-05, "loss": 0.0593, "step": 3250 }, { "grad_norm": 0.5277099609375, "learning_rate": 9.655245561152e-05, "loss": 0.0608, "step": 3260 }, { "grad_norm": 0.5423051118850708, "learning_rate": 9.65222249372984e-05, "loss": 0.0549, "step": 3270 }, { "grad_norm": 0.474439412355423, "learning_rate": 9.649186707309026e-05, "loss": 0.0592, "step": 3280 }, { "grad_norm": 0.5447307825088501, "learning_rate": 9.646138210189283e-05, "loss": 0.0573, "step": 3290 }, { "grad_norm": 0.46958911418914795, "learning_rate": 9.643077010705087e-05, "loss": 0.0648, "step": 3300 }, { "grad_norm": 0.6247355341911316, "learning_rate": 9.640003117225637e-05, "loss": 0.0649, "step": 3310 }, { "grad_norm": 0.622942328453064, "learning_rate": 9.636916538154846e-05, "loss": 0.0669, "step": 3320 }, { "grad_norm": 0.46045130491256714, "learning_rate": 9.633817281931296e-05, "loss": 0.0622, "step": 3330 }, { "grad_norm": 0.5312515497207642, "learning_rate": 9.630705357028242e-05, "loss": 0.0597, "step": 3340 }, { "grad_norm": 0.4561923146247864, "learning_rate": 9.627580771953563e-05, "loss": 0.0564, "step": 3350 }, { "grad_norm": 0.4700702428817749, "learning_rate": 9.624443535249759e-05, "loss": 0.0633, "step": 3360 }, { "grad_norm": 0.5625237226486206, "learning_rate": 9.621293655493913e-05, "loss": 0.0578, "step": 3370 }, { "grad_norm": 0.5612451434135437, "learning_rate": 9.618131141297675e-05, "loss": 0.0649, "step": 3380 }, { "grad_norm": 0.5531405210494995, "learning_rate": 9.614956001307242e-05, "loss": 0.0636, "step": 3390 }, { "grad_norm": 0.5203084945678711, "learning_rate": 9.611768244203321e-05, "loss": 0.0597, "step": 3400 }, { "grad_norm": 0.48514410853385925, "learning_rate": 9.60856787870112e-05, "loss": 0.0618, "step": 3410 }, { "grad_norm": 0.42436739802360535, "learning_rate": 9.605354913550318e-05, "loss": 0.0594, "step": 3420 }, { "grad_norm": 0.4707908630371094, "learning_rate": 9.602129357535037e-05, "loss": 0.0581, "step": 3430 }, { "grad_norm": 0.6007688045501709, "learning_rate": 9.598891219473825e-05, "loss": 0.0583, "step": 3440 }, { "grad_norm": 0.58519446849823, "learning_rate": 9.595640508219625e-05, "loss": 0.0582, "step": 3450 }, { "grad_norm": 0.6219605207443237, "learning_rate": 9.592377232659761e-05, "loss": 0.0607, "step": 3460 }, { "grad_norm": 0.5018323659896851, "learning_rate": 9.589101401715904e-05, "loss": 0.0551, "step": 3470 }, { "grad_norm": 0.5631371140480042, "learning_rate": 9.585813024344045e-05, "loss": 0.0579, "step": 3480 }, { "grad_norm": 0.7580733299255371, "learning_rate": 9.58251210953449e-05, "loss": 0.0571, "step": 3490 }, { "grad_norm": 0.6364933252334595, "learning_rate": 9.579198666311809e-05, "loss": 0.0545, "step": 3500 }, { "grad_norm": 0.6756724715232849, "learning_rate": 9.575872703734832e-05, "loss": 0.0589, "step": 3510 }, { "grad_norm": 0.4428686499595642, "learning_rate": 9.572534230896611e-05, "loss": 0.0565, "step": 3520 }, { "grad_norm": 0.5265988707542419, "learning_rate": 9.569183256924403e-05, "loss": 0.0531, "step": 3530 }, { "grad_norm": 0.6523367762565613, "learning_rate": 9.565819790979646e-05, "loss": 0.058, "step": 3540 }, { "grad_norm": 0.5738118886947632, "learning_rate": 9.562443842257925e-05, "loss": 0.0607, "step": 3550 }, { "grad_norm": 0.5877414345741272, "learning_rate": 9.559055419988956e-05, "loss": 0.0599, "step": 3560 }, { "grad_norm": 0.6353724598884583, "learning_rate": 9.555654533436557e-05, "loss": 0.0581, "step": 3570 }, { "grad_norm": 0.42666953802108765, "learning_rate": 9.552241191898621e-05, "loss": 0.0556, "step": 3580 }, { "grad_norm": 0.5157475471496582, "learning_rate": 9.548815404707092e-05, "loss": 0.0604, "step": 3590 }, { "grad_norm": 0.7215608954429626, "learning_rate": 9.545377181227942e-05, "loss": 0.0601, "step": 3600 }, { "grad_norm": 0.5162255764007568, "learning_rate": 9.541926530861145e-05, "loss": 0.055, "step": 3610 }, { "grad_norm": 0.4929755628108978, "learning_rate": 9.538463463040645e-05, "loss": 0.0624, "step": 3620 }, { "grad_norm": 0.4630160331726074, "learning_rate": 9.534987987234337e-05, "loss": 0.0531, "step": 3630 }, { "grad_norm": 0.6550790071487427, "learning_rate": 9.53150011294404e-05, "loss": 0.06, "step": 3640 }, { "grad_norm": 0.6224126815795898, "learning_rate": 9.527999849705471e-05, "loss": 0.058, "step": 3650 }, { "grad_norm": 0.5941552519798279, "learning_rate": 9.524487207088213e-05, "loss": 0.0605, "step": 3660 }, { "grad_norm": 0.5191814303398132, "learning_rate": 9.520962194695698e-05, "loss": 0.0554, "step": 3670 }, { "grad_norm": 0.5197587013244629, "learning_rate": 9.517424822165175e-05, "loss": 0.0604, "step": 3680 }, { "grad_norm": 0.6364939212799072, "learning_rate": 9.513875099167685e-05, "loss": 0.0607, "step": 3690 }, { "grad_norm": 0.6344776153564453, "learning_rate": 9.510313035408035e-05, "loss": 0.0591, "step": 3700 }, { "grad_norm": 0.5283899903297424, "learning_rate": 9.506738640624775e-05, "loss": 0.0548, "step": 3710 }, { "grad_norm": 0.5169802308082581, "learning_rate": 9.50315192459016e-05, "loss": 0.0568, "step": 3720 }, { "grad_norm": 0.5627011656761169, "learning_rate": 9.499552897110136e-05, "loss": 0.0575, "step": 3730 }, { "grad_norm": 0.5730915665626526, "learning_rate": 9.495941568024304e-05, "loss": 0.0544, "step": 3740 }, { "grad_norm": 0.5705208778381348, "learning_rate": 9.492317947205904e-05, "loss": 0.0554, "step": 3750 }, { "grad_norm": 0.48615023493766785, "learning_rate": 9.488682044561775e-05, "loss": 0.054, "step": 3760 }, { "grad_norm": 0.5147746205329895, "learning_rate": 9.485033870032335e-05, "loss": 0.0559, "step": 3770 }, { "grad_norm": 0.5804494023323059, "learning_rate": 9.481373433591556e-05, "loss": 0.0593, "step": 3780 }, { "grad_norm": 0.5439186096191406, "learning_rate": 9.47770074524693e-05, "loss": 0.0551, "step": 3790 }, { "grad_norm": 0.529241681098938, "learning_rate": 9.474015815039446e-05, "loss": 0.0579, "step": 3800 }, { "grad_norm": 0.42327943444252014, "learning_rate": 9.470318653043565e-05, "loss": 0.0551, "step": 3810 }, { "grad_norm": 0.4692992568016052, "learning_rate": 9.466609269367185e-05, "loss": 0.0577, "step": 3820 }, { "grad_norm": 0.6473419070243835, "learning_rate": 9.46288767415162e-05, "loss": 0.0609, "step": 3830 }, { "grad_norm": 0.624902069568634, "learning_rate": 9.459153877571567e-05, "loss": 0.0592, "step": 3840 }, { "grad_norm": 0.6024203896522522, "learning_rate": 9.455407889835087e-05, "loss": 0.0575, "step": 3850 }, { "grad_norm": 0.4566102623939514, "learning_rate": 9.451649721183564e-05, "loss": 0.0536, "step": 3860 }, { "grad_norm": 0.5603846907615662, "learning_rate": 9.447879381891692e-05, "loss": 0.0541, "step": 3870 }, { "grad_norm": 0.4703678488731384, "learning_rate": 9.444096882267428e-05, "loss": 0.0544, "step": 3880 }, { "grad_norm": 0.5888656377792358, "learning_rate": 9.440302232651988e-05, "loss": 0.0562, "step": 3890 }, { "grad_norm": 0.541254460811615, "learning_rate": 9.436495443419795e-05, "loss": 0.0531, "step": 3900 }, { "grad_norm": 0.48808667063713074, "learning_rate": 9.432676524978466e-05, "loss": 0.0515, "step": 3910 }, { "grad_norm": 0.6694585084915161, "learning_rate": 9.42884548776878e-05, "loss": 0.0552, "step": 3920 }, { "grad_norm": 0.683584451675415, "learning_rate": 9.425002342264646e-05, "loss": 0.0565, "step": 3930 }, { "grad_norm": 0.5609367489814758, "learning_rate": 9.421147098973077e-05, "loss": 0.0547, "step": 3940 }, { "grad_norm": 0.6370012164115906, "learning_rate": 9.41727976843416e-05, "loss": 0.0616, "step": 3950 }, { "grad_norm": 0.5923592448234558, "learning_rate": 9.413400361221029e-05, "loss": 0.0583, "step": 3960 }, { "grad_norm": 0.6023064255714417, "learning_rate": 9.409508887939835e-05, "loss": 0.0588, "step": 3970 }, { "grad_norm": 0.6249566078186035, "learning_rate": 9.40560535922972e-05, "loss": 0.0587, "step": 3980 }, { "grad_norm": 0.4997934103012085, "learning_rate": 9.40168978576278e-05, "loss": 0.0654, "step": 3990 }, { "grad_norm": 0.5568627119064331, "learning_rate": 9.397762178244043e-05, "loss": 0.055, "step": 4000 }, { "grad_norm": 0.5785378217697144, "learning_rate": 9.393822547411439e-05, "loss": 0.0525, "step": 4010 }, { "grad_norm": 0.5153332352638245, "learning_rate": 9.389870904035769e-05, "loss": 0.0566, "step": 4020 }, { "grad_norm": 0.4392877519130707, "learning_rate": 9.385907258920672e-05, "loss": 0.0532, "step": 4030 }, { "grad_norm": 0.49974969029426575, "learning_rate": 9.381931622902607e-05, "loss": 0.0538, "step": 4040 }, { "grad_norm": 0.6000987887382507, "learning_rate": 9.377944006850807e-05, "loss": 0.0555, "step": 4050 }, { "grad_norm": 0.5420299768447876, "learning_rate": 9.373944421667265e-05, "loss": 0.0532, "step": 4060 }, { "grad_norm": 0.4731873869895935, "learning_rate": 9.369932878286691e-05, "loss": 0.0523, "step": 4070 }, { "grad_norm": 0.4393448829650879, "learning_rate": 9.365909387676494e-05, "loss": 0.0546, "step": 4080 }, { "grad_norm": 0.4877563416957855, "learning_rate": 9.361873960836744e-05, "loss": 0.0541, "step": 4090 }, { "grad_norm": 0.5272069573402405, "learning_rate": 9.357826608800142e-05, "loss": 0.0556, "step": 4100 }, { "grad_norm": 0.5248222351074219, "learning_rate": 9.353767342631994e-05, "loss": 0.0515, "step": 4110 }, { "grad_norm": 0.5554358959197998, "learning_rate": 9.34969617343018e-05, "loss": 0.0522, "step": 4120 }, { "grad_norm": 0.562025785446167, "learning_rate": 9.345613112325122e-05, "loss": 0.0542, "step": 4130 }, { "grad_norm": 0.43112912774086, "learning_rate": 9.34151817047975e-05, "loss": 0.0537, "step": 4140 }, { "grad_norm": 0.4652014374732971, "learning_rate": 9.33741135908948e-05, "loss": 0.0538, "step": 4150 }, { "grad_norm": 0.41591545939445496, "learning_rate": 9.33329268938218e-05, "loss": 0.0526, "step": 4160 }, { "grad_norm": 0.4402368366718292, "learning_rate": 9.329162172618132e-05, "loss": 0.0532, "step": 4170 }, { "grad_norm": 0.4815138280391693, "learning_rate": 9.325019820090013e-05, "loss": 0.0518, "step": 4180 }, { "grad_norm": 0.43706417083740234, "learning_rate": 9.320865643122855e-05, "loss": 0.0496, "step": 4190 }, { "grad_norm": 0.45009341835975647, "learning_rate": 9.316699653074023e-05, "loss": 0.0544, "step": 4200 }, { "grad_norm": 0.4609179198741913, "learning_rate": 9.312521861333172e-05, "loss": 0.0555, "step": 4210 }, { "grad_norm": 0.4796389639377594, "learning_rate": 9.308332279322224e-05, "loss": 0.0517, "step": 4220 }, { "grad_norm": 0.4630071222782135, "learning_rate": 9.304130918495338e-05, "loss": 0.0584, "step": 4230 }, { "grad_norm": 0.5116509795188904, "learning_rate": 9.299917790338874e-05, "loss": 0.0517, "step": 4240 }, { "grad_norm": 0.4714801609516144, "learning_rate": 9.295692906371363e-05, "loss": 0.0566, "step": 4250 }, { "grad_norm": 0.6243820786476135, "learning_rate": 9.291456278143476e-05, "loss": 0.0541, "step": 4260 }, { "grad_norm": 0.627763569355011, "learning_rate": 9.287207917237994e-05, "loss": 0.0567, "step": 4270 }, { "grad_norm": 0.371865451335907, "learning_rate": 9.282947835269773e-05, "loss": 0.0558, "step": 4280 }, { "grad_norm": 0.5337596535682678, "learning_rate": 9.278676043885715e-05, "loss": 0.0594, "step": 4290 }, { "grad_norm": 0.5586329698562622, "learning_rate": 9.274392554764733e-05, "loss": 0.0544, "step": 4300 }, { "grad_norm": 0.5547240972518921, "learning_rate": 9.270097379617723e-05, "loss": 0.054, "step": 4310 }, { "grad_norm": 0.5266514420509338, "learning_rate": 9.26579053018753e-05, "loss": 0.0583, "step": 4320 }, { "grad_norm": 0.5926889777183533, "learning_rate": 9.261472018248918e-05, "loss": 0.0552, "step": 4330 }, { "grad_norm": 0.4509258568286896, "learning_rate": 9.25714185560853e-05, "loss": 0.0589, "step": 4340 }, { "grad_norm": 0.5349770784378052, "learning_rate": 9.252800054104868e-05, "loss": 0.0493, "step": 4350 }, { "grad_norm": 0.5949016213417053, "learning_rate": 9.248446625608252e-05, "loss": 0.0561, "step": 4360 }, { "grad_norm": 0.4947768449783325, "learning_rate": 9.244081582020789e-05, "loss": 0.0544, "step": 4370 }, { "grad_norm": 0.5038443207740784, "learning_rate": 9.239704935276339e-05, "loss": 0.0497, "step": 4380 }, { "grad_norm": 0.5768541693687439, "learning_rate": 9.235316697340489e-05, "loss": 0.0588, "step": 4390 }, { "grad_norm": 0.5308362245559692, "learning_rate": 9.230916880210512e-05, "loss": 0.0537, "step": 4400 }, { "grad_norm": 0.5097739100456238, "learning_rate": 9.226505495915342e-05, "loss": 0.0523, "step": 4410 }, { "grad_norm": 0.5145155191421509, "learning_rate": 9.222082556515536e-05, "loss": 0.057, "step": 4420 }, { "grad_norm": 0.3700404167175293, "learning_rate": 9.217648074103242e-05, "loss": 0.0552, "step": 4430 }, { "grad_norm": 0.42606955766677856, "learning_rate": 9.213202060802161e-05, "loss": 0.0531, "step": 4440 }, { "grad_norm": 0.5095778703689575, "learning_rate": 9.208744528767528e-05, "loss": 0.0565, "step": 4450 }, { "grad_norm": 0.5339433550834656, "learning_rate": 9.204275490186064e-05, "loss": 0.0551, "step": 4460 }, { "grad_norm": 0.46008360385894775, "learning_rate": 9.199794957275949e-05, "loss": 0.0572, "step": 4470 }, { "grad_norm": 0.5331209301948547, "learning_rate": 9.19530294228679e-05, "loss": 0.0524, "step": 4480 }, { "grad_norm": 0.5454930067062378, "learning_rate": 9.190799457499583e-05, "loss": 0.0578, "step": 4490 }, { "grad_norm": 0.6232542395591736, "learning_rate": 9.186284515226686e-05, "loss": 0.057, "step": 4500 }, { "grad_norm": 0.5831615328788757, "learning_rate": 9.181758127811777e-05, "loss": 0.0596, "step": 4510 }, { "grad_norm": 0.6251550912857056, "learning_rate": 9.177220307629825e-05, "loss": 0.0563, "step": 4520 }, { "grad_norm": 0.5718427896499634, "learning_rate": 9.172671067087059e-05, "loss": 0.0545, "step": 4530 }, { "grad_norm": 0.4496645927429199, "learning_rate": 9.16811041862093e-05, "loss": 0.0544, "step": 4540 }, { "grad_norm": 0.6001369953155518, "learning_rate": 9.163538374700076e-05, "loss": 0.0533, "step": 4550 }, { "grad_norm": 0.5739479660987854, "learning_rate": 9.158954947824287e-05, "loss": 0.0525, "step": 4560 }, { "grad_norm": 0.5372488498687744, "learning_rate": 9.154360150524482e-05, "loss": 0.0509, "step": 4570 }, { "grad_norm": 0.42326685786247253, "learning_rate": 9.14975399536266e-05, "loss": 0.0538, "step": 4580 }, { "grad_norm": 0.37366053462028503, "learning_rate": 9.14513649493187e-05, "loss": 0.05, "step": 4590 }, { "grad_norm": 0.42000848054885864, "learning_rate": 9.140507661856187e-05, "loss": 0.055, "step": 4600 }, { "grad_norm": 0.30610954761505127, "learning_rate": 9.135867508790661e-05, "loss": 0.0536, "step": 4610 }, { "grad_norm": 0.4791013300418854, "learning_rate": 9.131216048421291e-05, "loss": 0.0514, "step": 4620 }, { "grad_norm": 0.5309016704559326, "learning_rate": 9.126553293464998e-05, "loss": 0.0539, "step": 4630 }, { "grad_norm": 0.4759531021118164, "learning_rate": 9.121879256669572e-05, "loss": 0.056, "step": 4640 }, { "grad_norm": 0.3590352237224579, "learning_rate": 9.117193950813652e-05, "loss": 0.0536, "step": 4650 }, { "grad_norm": 0.42257943749427795, "learning_rate": 9.112497388706685e-05, "loss": 0.0496, "step": 4660 }, { "grad_norm": 0.5374065041542053, "learning_rate": 9.10778958318889e-05, "loss": 0.0555, "step": 4670 }, { "grad_norm": 0.4869564175605774, "learning_rate": 9.103070547131232e-05, "loss": 0.0558, "step": 4680 }, { "grad_norm": 0.5503962635993958, "learning_rate": 9.098340293435375e-05, "loss": 0.0504, "step": 4690 }, { "grad_norm": 0.4951989948749542, "learning_rate": 9.093598835033649e-05, "loss": 0.0548, "step": 4700 }, { "grad_norm": 0.46331873536109924, "learning_rate": 9.088846184889021e-05, "loss": 0.0537, "step": 4710 }, { "grad_norm": 0.5856289863586426, "learning_rate": 9.084082355995057e-05, "loss": 0.0512, "step": 4720 }, { "grad_norm": 0.5602821111679077, "learning_rate": 9.079307361375882e-05, "loss": 0.0526, "step": 4730 }, { "grad_norm": 0.5624141097068787, "learning_rate": 9.074521214086149e-05, "loss": 0.0541, "step": 4740 }, { "grad_norm": 0.5045981407165527, "learning_rate": 9.069723927211001e-05, "loss": 0.0481, "step": 4750 }, { "grad_norm": 0.5227964520454407, "learning_rate": 9.064915513866037e-05, "loss": 0.0552, "step": 4760 }, { "grad_norm": 0.6533440947532654, "learning_rate": 9.060095987197279e-05, "loss": 0.0503, "step": 4770 }, { "grad_norm": 0.4607369005680084, "learning_rate": 9.055265360381126e-05, "loss": 0.0516, "step": 4780 }, { "grad_norm": 0.5377238988876343, "learning_rate": 9.050423646624326e-05, "loss": 0.0531, "step": 4790 }, { "grad_norm": 0.4877373278141022, "learning_rate": 9.045570859163943e-05, "loss": 0.0534, "step": 4800 }, { "grad_norm": 0.5039026737213135, "learning_rate": 9.04070701126731e-05, "loss": 0.05, "step": 4810 }, { "grad_norm": 0.5420809388160706, "learning_rate": 9.035832116232001e-05, "loss": 0.0519, "step": 4820 }, { "grad_norm": 0.5420294404029846, "learning_rate": 9.030946187385796e-05, "loss": 0.0513, "step": 4830 }, { "grad_norm": 0.5666866302490234, "learning_rate": 9.026049238086635e-05, "loss": 0.0529, "step": 4840 }, { "grad_norm": 0.3657691776752472, "learning_rate": 9.021141281722591e-05, "loss": 0.0521, "step": 4850 }, { "grad_norm": 0.6102502346038818, "learning_rate": 9.01622233171183e-05, "loss": 0.0507, "step": 4860 }, { "grad_norm": 0.5554990768432617, "learning_rate": 9.011292401502574e-05, "loss": 0.0529, "step": 4870 }, { "grad_norm": 0.4869811236858368, "learning_rate": 9.006351504573063e-05, "loss": 0.0514, "step": 4880 }, { "grad_norm": 0.3845446705818176, "learning_rate": 9.001399654431519e-05, "loss": 0.0511, "step": 4890 }, { "grad_norm": 0.43466147780418396, "learning_rate": 8.996436864616116e-05, "loss": 0.05, "step": 4900 }, { "grad_norm": 0.5801776647567749, "learning_rate": 8.991463148694925e-05, "loss": 0.0485, "step": 4910 }, { "grad_norm": 0.6859143376350403, "learning_rate": 8.986478520265902e-05, "loss": 0.0542, "step": 4920 }, { "grad_norm": 0.6430835723876953, "learning_rate": 8.981482992956827e-05, "loss": 0.0543, "step": 4930 }, { "grad_norm": 0.42302078008651733, "learning_rate": 8.976476580425282e-05, "loss": 0.0535, "step": 4940 }, { "grad_norm": 0.4328843057155609, "learning_rate": 8.971459296358606e-05, "loss": 0.0522, "step": 4950 }, { "grad_norm": 0.40373238921165466, "learning_rate": 8.966431154473864e-05, "loss": 0.0541, "step": 4960 }, { "grad_norm": 0.5615489482879639, "learning_rate": 8.961392168517803e-05, "loss": 0.0567, "step": 4970 }, { "grad_norm": 0.4355982542037964, "learning_rate": 8.956342352266821e-05, "loss": 0.0511, "step": 4980 }, { "grad_norm": 0.5548534393310547, "learning_rate": 8.95128171952692e-05, "loss": 0.0514, "step": 4990 }, { "grad_norm": 0.5604633092880249, "learning_rate": 8.946210284133676e-05, "loss": 0.0511, "step": 5000 }, { "grad_norm": 0.5363962054252625, "learning_rate": 8.941128059952201e-05, "loss": 0.0511, "step": 5010 }, { "grad_norm": 0.491742879152298, "learning_rate": 8.936035060877102e-05, "loss": 0.0505, "step": 5020 }, { "grad_norm": 0.4461692273616791, "learning_rate": 8.930931300832443e-05, "loss": 0.0498, "step": 5030 }, { "grad_norm": 0.49962547421455383, "learning_rate": 8.925816793771711e-05, "loss": 0.0543, "step": 5040 }, { "grad_norm": 0.619042158126831, "learning_rate": 8.92069155367777e-05, "loss": 0.0531, "step": 5050 }, { "grad_norm": 0.4674154818058014, "learning_rate": 8.915555594562834e-05, "loss": 0.0521, "step": 5060 }, { "grad_norm": 0.558987021446228, "learning_rate": 8.910408930468416e-05, "loss": 0.0491, "step": 5070 }, { "grad_norm": 0.4956831634044647, "learning_rate": 8.905251575465303e-05, "loss": 0.0474, "step": 5080 }, { "grad_norm": 0.5653302669525146, "learning_rate": 8.900083543653502e-05, "loss": 0.0554, "step": 5090 }, { "grad_norm": 0.4657272696495056, "learning_rate": 8.894904849162218e-05, "loss": 0.0484, "step": 5100 }, { "grad_norm": 0.4607899487018585, "learning_rate": 8.889715506149802e-05, "loss": 0.0519, "step": 5110 }, { "grad_norm": 0.37972721457481384, "learning_rate": 8.884515528803722e-05, "loss": 0.0477, "step": 5120 }, { "grad_norm": 0.4828026592731476, "learning_rate": 8.879304931340517e-05, "loss": 0.0494, "step": 5130 }, { "grad_norm": 0.4062140882015228, "learning_rate": 8.874083728005759e-05, "loss": 0.0519, "step": 5140 }, { "grad_norm": 0.5053769946098328, "learning_rate": 8.868851933074021e-05, "loss": 0.0525, "step": 5150 }, { "grad_norm": 0.4690435528755188, "learning_rate": 8.863609560848829e-05, "loss": 0.0471, "step": 5160 }, { "grad_norm": 0.48639559745788574, "learning_rate": 8.85835662566263e-05, "loss": 0.0513, "step": 5170 }, { "grad_norm": 0.43334561586380005, "learning_rate": 8.853093141876747e-05, "loss": 0.0473, "step": 5180 }, { "grad_norm": 0.4833715856075287, "learning_rate": 8.847819123881343e-05, "loss": 0.0466, "step": 5190 }, { "grad_norm": 0.5589087605476379, "learning_rate": 8.842534586095383e-05, "loss": 0.0529, "step": 5200 }, { "grad_norm": 0.4683074951171875, "learning_rate": 8.837239542966593e-05, "loss": 0.0496, "step": 5210 }, { "grad_norm": 0.5216234922409058, "learning_rate": 8.831934008971417e-05, "loss": 0.0462, "step": 5220 }, { "grad_norm": 0.4678767919540405, "learning_rate": 8.826617998614982e-05, "loss": 0.0487, "step": 5230 }, { "grad_norm": 0.530791699886322, "learning_rate": 8.821291526431056e-05, "loss": 0.0475, "step": 5240 }, { "grad_norm": 0.460438072681427, "learning_rate": 8.815954606982015e-05, "loss": 0.0521, "step": 5250 }, { "grad_norm": 0.4873076379299164, "learning_rate": 8.810607254858789e-05, "loss": 0.0447, "step": 5260 }, { "grad_norm": 0.5115181803703308, "learning_rate": 8.805249484680838e-05, "loss": 0.0519, "step": 5270 }, { "grad_norm": 0.46512988209724426, "learning_rate": 8.799881311096096e-05, "loss": 0.0491, "step": 5280 }, { "grad_norm": 0.5357261896133423, "learning_rate": 8.794502748780949e-05, "loss": 0.0476, "step": 5290 }, { "grad_norm": 0.4545869529247284, "learning_rate": 8.78911381244018e-05, "loss": 0.0476, "step": 5300 }, { "grad_norm": 0.45949622988700867, "learning_rate": 8.783714516806933e-05, "loss": 0.0491, "step": 5310 }, { "grad_norm": 0.4598928391933441, "learning_rate": 8.77830487664268e-05, "loss": 0.0526, "step": 5320 }, { "grad_norm": 0.522425651550293, "learning_rate": 8.772884906737167e-05, "loss": 0.0499, "step": 5330 }, { "grad_norm": 0.5587599873542786, "learning_rate": 8.767454621908387e-05, "loss": 0.0525, "step": 5340 }, { "grad_norm": 0.49348875880241394, "learning_rate": 8.76201403700253e-05, "loss": 0.0519, "step": 5350 }, { "grad_norm": 0.5256590247154236, "learning_rate": 8.756563166893949e-05, "loss": 0.0535, "step": 5360 }, { "grad_norm": 0.4977443814277649, "learning_rate": 8.751102026485113e-05, "loss": 0.0521, "step": 5370 }, { "grad_norm": 0.5837010145187378, "learning_rate": 8.745630630706571e-05, "loss": 0.0513, "step": 5380 }, { "grad_norm": 0.39375320076942444, "learning_rate": 8.740148994516912e-05, "loss": 0.046, "step": 5390 }, { "grad_norm": 0.5269045233726501, "learning_rate": 8.73465713290272e-05, "loss": 0.0534, "step": 5400 }, { "grad_norm": 0.5881283283233643, "learning_rate": 8.729155060878533e-05, "loss": 0.0518, "step": 5410 }, { "grad_norm": 0.6109328269958496, "learning_rate": 8.723642793486809e-05, "loss": 0.0486, "step": 5420 }, { "grad_norm": 0.3970491588115692, "learning_rate": 8.718120345797873e-05, "loss": 0.0502, "step": 5430 }, { "grad_norm": 0.4655765891075134, "learning_rate": 8.712587732909889e-05, "loss": 0.049, "step": 5440 }, { "grad_norm": 0.49529194831848145, "learning_rate": 8.707044969948806e-05, "loss": 0.0473, "step": 5450 }, { "grad_norm": 0.4889856278896332, "learning_rate": 8.701492072068329e-05, "loss": 0.048, "step": 5460 }, { "grad_norm": 0.6724857687950134, "learning_rate": 8.695929054449869e-05, "loss": 0.0505, "step": 5470 }, { "grad_norm": 0.576635479927063, "learning_rate": 8.690355932302501e-05, "loss": 0.0475, "step": 5480 }, { "grad_norm": 0.40675845742225647, "learning_rate": 8.684772720862931e-05, "loss": 0.0517, "step": 5490 }, { "grad_norm": 0.48108670115470886, "learning_rate": 8.679179435395446e-05, "loss": 0.0488, "step": 5500 }, { "grad_norm": 0.3651122450828552, "learning_rate": 8.673576091191874e-05, "loss": 0.0483, "step": 5510 }, { "grad_norm": 0.40905848145484924, "learning_rate": 8.667962703571541e-05, "loss": 0.0457, "step": 5520 }, { "grad_norm": 0.4346217215061188, "learning_rate": 8.662339287881238e-05, "loss": 0.0487, "step": 5530 }, { "grad_norm": 0.5116409659385681, "learning_rate": 8.656705859495169e-05, "loss": 0.0472, "step": 5540 }, { "grad_norm": 0.5556833148002625, "learning_rate": 8.651062433814912e-05, "loss": 0.0467, "step": 5550 }, { "grad_norm": 0.44723305106163025, "learning_rate": 8.645409026269375e-05, "loss": 0.0484, "step": 5560 }, { "grad_norm": 0.5550429224967957, "learning_rate": 8.639745652314759e-05, "loss": 0.0496, "step": 5570 }, { "grad_norm": 0.39598211646080017, "learning_rate": 8.634072327434515e-05, "loss": 0.0471, "step": 5580 }, { "grad_norm": 0.4221607744693756, "learning_rate": 8.628389067139294e-05, "loss": 0.0476, "step": 5590 }, { "grad_norm": 0.3684441149234772, "learning_rate": 8.622695886966911e-05, "loss": 0.0497, "step": 5600 }, { "grad_norm": 0.4303925931453705, "learning_rate": 8.616992802482308e-05, "loss": 0.0476, "step": 5610 }, { "grad_norm": 0.46542948484420776, "learning_rate": 8.611279829277496e-05, "loss": 0.0522, "step": 5620 }, { "grad_norm": 0.4745675325393677, "learning_rate": 8.605556982971528e-05, "loss": 0.0478, "step": 5630 }, { "grad_norm": 0.4970300793647766, "learning_rate": 8.599824279210447e-05, "loss": 0.0462, "step": 5640 }, { "grad_norm": 0.545353353023529, "learning_rate": 8.594081733667243e-05, "loss": 0.0457, "step": 5650 }, { "grad_norm": 0.41005998849868774, "learning_rate": 8.58832936204182e-05, "loss": 0.0457, "step": 5660 }, { "grad_norm": 0.48379749059677124, "learning_rate": 8.582567180060942e-05, "loss": 0.048, "step": 5670 }, { "grad_norm": 0.4568611979484558, "learning_rate": 8.576795203478194e-05, "loss": 0.0494, "step": 5680 }, { "grad_norm": 0.4423889219760895, "learning_rate": 8.571013448073939e-05, "loss": 0.0499, "step": 5690 }, { "grad_norm": 0.40718817710876465, "learning_rate": 8.565221929655275e-05, "loss": 0.0511, "step": 5700 }, { "grad_norm": 0.5141193866729736, "learning_rate": 8.559420664055992e-05, "loss": 0.0462, "step": 5710 }, { "grad_norm": 0.5302907228469849, "learning_rate": 8.553609667136532e-05, "loss": 0.0467, "step": 5720 }, { "grad_norm": 0.4390532672405243, "learning_rate": 8.547788954783936e-05, "loss": 0.05, "step": 5730 }, { "grad_norm": 0.41804423928260803, "learning_rate": 8.541958542911808e-05, "loss": 0.0478, "step": 5740 }, { "grad_norm": 0.43645694851875305, "learning_rate": 8.536118447460275e-05, "loss": 0.0462, "step": 5750 }, { "grad_norm": 0.49388405680656433, "learning_rate": 8.530268684395932e-05, "loss": 0.0484, "step": 5760 }, { "grad_norm": 0.5820779800415039, "learning_rate": 8.524409269711807e-05, "loss": 0.0512, "step": 5770 }, { "grad_norm": 0.5733522772789001, "learning_rate": 8.51854021942732e-05, "loss": 0.0506, "step": 5780 }, { "grad_norm": 0.4883822202682495, "learning_rate": 8.512661549588227e-05, "loss": 0.0475, "step": 5790 }, { "grad_norm": 0.5154078006744385, "learning_rate": 8.506773276266588e-05, "loss": 0.047, "step": 5800 }, { "grad_norm": 0.5151110887527466, "learning_rate": 8.500875415560721e-05, "loss": 0.0538, "step": 5810 }, { "grad_norm": 0.5699576139450073, "learning_rate": 8.494967983595144e-05, "loss": 0.0516, "step": 5820 }, { "grad_norm": 0.6452685594558716, "learning_rate": 8.489050996520558e-05, "loss": 0.05, "step": 5830 }, { "grad_norm": 0.48322704434394836, "learning_rate": 8.483124470513775e-05, "loss": 0.0473, "step": 5840 }, { "grad_norm": 0.471245676279068, "learning_rate": 8.477188421777692e-05, "loss": 0.0492, "step": 5850 }, { "grad_norm": 0.4499649703502655, "learning_rate": 8.47124286654124e-05, "loss": 0.0463, "step": 5860 }, { "grad_norm": 0.4111122786998749, "learning_rate": 8.465287821059341e-05, "loss": 0.0485, "step": 5870 }, { "grad_norm": 0.38970747590065, "learning_rate": 8.45932330161286e-05, "loss": 0.0492, "step": 5880 }, { "grad_norm": 0.4419807493686676, "learning_rate": 8.453349324508567e-05, "loss": 0.0431, "step": 5890 }, { "grad_norm": 0.4473804831504822, "learning_rate": 8.447365906079088e-05, "loss": 0.0478, "step": 5900 }, { "grad_norm": 0.5254375338554382, "learning_rate": 8.441373062682856e-05, "loss": 0.0489, "step": 5910 }, { "grad_norm": 0.4527287185192108, "learning_rate": 8.43537081070408e-05, "loss": 0.0507, "step": 5920 }, { "grad_norm": 0.352102667093277, "learning_rate": 8.429359166552689e-05, "loss": 0.0464, "step": 5930 }, { "grad_norm": 0.4094739258289337, "learning_rate": 8.423338146664284e-05, "loss": 0.0483, "step": 5940 }, { "grad_norm": 0.47909194231033325, "learning_rate": 8.417307767500107e-05, "loss": 0.0462, "step": 5950 }, { "grad_norm": 0.3791649043560028, "learning_rate": 8.411268045546983e-05, "loss": 0.0469, "step": 5960 }, { "grad_norm": 0.34563949704170227, "learning_rate": 8.405218997317281e-05, "loss": 0.0534, "step": 5970 }, { "grad_norm": 0.41109538078308105, "learning_rate": 8.399160639348869e-05, "loss": 0.0486, "step": 5980 }, { "grad_norm": 0.4183802008628845, "learning_rate": 8.393092988205065e-05, "loss": 0.0469, "step": 5990 }, { "grad_norm": 0.3775635063648224, "learning_rate": 8.387016060474597e-05, "loss": 0.0486, "step": 6000 }, { "grad_norm": 0.47170305252075195, "learning_rate": 8.380929872771551e-05, "loss": 0.0485, "step": 6010 }, { "grad_norm": 0.37962549924850464, "learning_rate": 8.374834441735335e-05, "loss": 0.0466, "step": 6020 }, { "grad_norm": 0.33892786502838135, "learning_rate": 8.368729784030622e-05, "loss": 0.0488, "step": 6030 }, { "grad_norm": 0.4295814335346222, "learning_rate": 8.362615916347315e-05, "loss": 0.0466, "step": 6040 }, { "grad_norm": 0.42452430725097656, "learning_rate": 8.356492855400493e-05, "loss": 0.0472, "step": 6050 }, { "grad_norm": 0.424972265958786, "learning_rate": 8.350360617930371e-05, "loss": 0.048, "step": 6060 }, { "grad_norm": 0.4107874631881714, "learning_rate": 8.344219220702255e-05, "loss": 0.0448, "step": 6070 }, { "grad_norm": 0.3743925392627716, "learning_rate": 8.338068680506485e-05, "loss": 0.0474, "step": 6080 }, { "grad_norm": 0.4670070707798004, "learning_rate": 8.33190901415841e-05, "loss": 0.0463, "step": 6090 }, { "grad_norm": 0.49129682779312134, "learning_rate": 8.325740238498317e-05, "loss": 0.0464, "step": 6100 }, { "grad_norm": 0.4617558419704437, "learning_rate": 8.319562370391406e-05, "loss": 0.0456, "step": 6110 }, { "grad_norm": 0.4431152641773224, "learning_rate": 8.31337542672773e-05, "loss": 0.0475, "step": 6120 }, { "grad_norm": 0.3690592050552368, "learning_rate": 8.307179424422158e-05, "loss": 0.0423, "step": 6130 }, { "grad_norm": 0.4119509160518646, "learning_rate": 8.300974380414327e-05, "loss": 0.0466, "step": 6140 }, { "grad_norm": 0.33776113390922546, "learning_rate": 8.294760311668586e-05, "loss": 0.0474, "step": 6150 }, { "grad_norm": 0.45812442898750305, "learning_rate": 8.288537235173961e-05, "loss": 0.0473, "step": 6160 }, { "grad_norm": 0.44699740409851074, "learning_rate": 8.282305167944108e-05, "loss": 0.0458, "step": 6170 }, { "grad_norm": 0.4798637628555298, "learning_rate": 8.276064127017262e-05, "loss": 0.0465, "step": 6180 }, { "grad_norm": 0.5563379526138306, "learning_rate": 8.269814129456189e-05, "loss": 0.0426, "step": 6190 }, { "grad_norm": 0.4753587543964386, "learning_rate": 8.263555192348143e-05, "loss": 0.0443, "step": 6200 }, { "grad_norm": 0.4133961796760559, "learning_rate": 8.257287332804819e-05, "loss": 0.0431, "step": 6210 }, { "grad_norm": 0.4012044072151184, "learning_rate": 8.251010567962307e-05, "loss": 0.0464, "step": 6220 }, { "grad_norm": 0.4883323609828949, "learning_rate": 8.244724914981041e-05, "loss": 0.0437, "step": 6230 }, { "grad_norm": 0.42000195384025574, "learning_rate": 8.238430391045757e-05, "loss": 0.0458, "step": 6240 }, { "grad_norm": 0.4328843951225281, "learning_rate": 8.232127013365445e-05, "loss": 0.0446, "step": 6250 }, { "grad_norm": 0.49242380261421204, "learning_rate": 8.225814799173295e-05, "loss": 0.0463, "step": 6260 }, { "grad_norm": 0.47059646248817444, "learning_rate": 8.219493765726663e-05, "loss": 0.0442, "step": 6270 }, { "grad_norm": 0.39271095395088196, "learning_rate": 8.21316393030701e-05, "loss": 0.0463, "step": 6280 }, { "grad_norm": 0.4726349711418152, "learning_rate": 8.206825310219865e-05, "loss": 0.0471, "step": 6290 }, { "grad_norm": 0.42810365557670593, "learning_rate": 8.200477922794776e-05, "loss": 0.0424, "step": 6300 }, { "grad_norm": 0.4144385755062103, "learning_rate": 8.194121785385256e-05, "loss": 0.0401, "step": 6310 }, { "grad_norm": 0.4185318946838379, "learning_rate": 8.187756915368741e-05, "loss": 0.0464, "step": 6320 }, { "grad_norm": 0.36527198553085327, "learning_rate": 8.181383330146544e-05, "loss": 0.042, "step": 6330 }, { "grad_norm": 0.4455040693283081, "learning_rate": 8.175001047143804e-05, "loss": 0.0447, "step": 6340 }, { "grad_norm": 0.31558549404144287, "learning_rate": 8.168610083809438e-05, "loss": 0.0445, "step": 6350 }, { "grad_norm": 0.39845818281173706, "learning_rate": 8.162210457616095e-05, "loss": 0.0468, "step": 6360 }, { "grad_norm": 0.47270894050598145, "learning_rate": 8.155802186060109e-05, "loss": 0.0459, "step": 6370 }, { "grad_norm": 0.3671899735927582, "learning_rate": 8.149385286661453e-05, "loss": 0.0478, "step": 6380 }, { "grad_norm": 0.45018354058265686, "learning_rate": 8.14295977696368e-05, "loss": 0.0476, "step": 6390 }, { "grad_norm": 0.6423484086990356, "learning_rate": 8.13652567453389e-05, "loss": 0.048, "step": 6400 }, { "grad_norm": 0.49249398708343506, "learning_rate": 8.130082996962676e-05, "loss": 0.0449, "step": 6410 }, { "grad_norm": 0.45846104621887207, "learning_rate": 8.123631761864068e-05, "loss": 0.0409, "step": 6420 }, { "grad_norm": 0.3806885778903961, "learning_rate": 8.1171719868755e-05, "loss": 0.0451, "step": 6430 }, { "grad_norm": 0.35434088110923767, "learning_rate": 8.110703689657748e-05, "loss": 0.0437, "step": 6440 }, { "grad_norm": 0.3482916057109833, "learning_rate": 8.104226887894892e-05, "loss": 0.0448, "step": 6450 }, { "grad_norm": 0.3681597411632538, "learning_rate": 8.097741599294257e-05, "loss": 0.0435, "step": 6460 }, { "grad_norm": 0.46956971287727356, "learning_rate": 8.091247841586378e-05, "loss": 0.0465, "step": 6470 }, { "grad_norm": 0.45849505066871643, "learning_rate": 8.084745632524939e-05, "loss": 0.0462, "step": 6480 }, { "grad_norm": 0.39770910143852234, "learning_rate": 8.07823498988673e-05, "loss": 0.0437, "step": 6490 }, { "grad_norm": 0.46747109293937683, "learning_rate": 8.071715931471602e-05, "loss": 0.0468, "step": 6500 }, { "grad_norm": 0.4514947533607483, "learning_rate": 8.06518847510241e-05, "loss": 0.0456, "step": 6510 }, { "grad_norm": 0.5073133707046509, "learning_rate": 8.058652638624971e-05, "loss": 0.0418, "step": 6520 }, { "grad_norm": 0.5380352139472961, "learning_rate": 8.052108439908013e-05, "loss": 0.0489, "step": 6530 }, { "grad_norm": 0.4667007625102997, "learning_rate": 8.045555896843125e-05, "loss": 0.0475, "step": 6540 }, { "grad_norm": 0.45686766505241394, "learning_rate": 8.03899502734471e-05, "loss": 0.0466, "step": 6550 }, { "grad_norm": 0.3877220153808594, "learning_rate": 8.032425849349931e-05, "loss": 0.044, "step": 6560 }, { "grad_norm": 0.4714347720146179, "learning_rate": 8.025848380818674e-05, "loss": 0.0421, "step": 6570 }, { "grad_norm": 0.3551757037639618, "learning_rate": 8.019262639733487e-05, "loss": 0.0406, "step": 6580 }, { "grad_norm": 0.35637393593788147, "learning_rate": 8.012668644099531e-05, "loss": 0.0447, "step": 6590 }, { "grad_norm": 0.3633667528629303, "learning_rate": 8.006066411944542e-05, "loss": 0.0433, "step": 6600 }, { "grad_norm": 0.38073790073394775, "learning_rate": 7.999455961318769e-05, "loss": 0.043, "step": 6610 }, { "grad_norm": 0.4017251431941986, "learning_rate": 7.992837310294932e-05, "loss": 0.0429, "step": 6620 }, { "grad_norm": 0.43783628940582275, "learning_rate": 7.986210476968167e-05, "loss": 0.0451, "step": 6630 }, { "grad_norm": 0.38079744577407837, "learning_rate": 7.97957547945599e-05, "loss": 0.0409, "step": 6640 }, { "grad_norm": 0.3851889669895172, "learning_rate": 7.972932335898226e-05, "loss": 0.0464, "step": 6650 }, { "grad_norm": 0.3811066150665283, "learning_rate": 7.966281064456975e-05, "loss": 0.0458, "step": 6660 }, { "grad_norm": 0.5243368148803711, "learning_rate": 7.959621683316563e-05, "loss": 0.0444, "step": 6670 }, { "grad_norm": 0.6172375679016113, "learning_rate": 7.952954210683481e-05, "loss": 0.0474, "step": 6680 }, { "grad_norm": 0.3869527578353882, "learning_rate": 7.946278664786345e-05, "loss": 0.0437, "step": 6690 }, { "grad_norm": 0.36588379740715027, "learning_rate": 7.939595063875842e-05, "loss": 0.0455, "step": 6700 }, { "grad_norm": 0.3870604336261749, "learning_rate": 7.932903426224683e-05, "loss": 0.0481, "step": 6710 }, { "grad_norm": 0.4223274290561676, "learning_rate": 7.926203770127552e-05, "loss": 0.0483, "step": 6720 }, { "grad_norm": 0.44375845789909363, "learning_rate": 7.919496113901046e-05, "loss": 0.0433, "step": 6730 }, { "grad_norm": 0.4170505702495575, "learning_rate": 7.912780475883649e-05, "loss": 0.0453, "step": 6740 }, { "grad_norm": 0.41197896003723145, "learning_rate": 7.906056874435652e-05, "loss": 0.0425, "step": 6750 }, { "grad_norm": 0.3659522533416748, "learning_rate": 7.899325327939131e-05, "loss": 0.0424, "step": 6760 }, { "grad_norm": 0.42875581979751587, "learning_rate": 7.892585854797872e-05, "loss": 0.0432, "step": 6770 }, { "grad_norm": 0.46886080503463745, "learning_rate": 7.88583847343734e-05, "loss": 0.0457, "step": 6780 }, { "grad_norm": 0.47526654601097107, "learning_rate": 7.879083202304616e-05, "loss": 0.0452, "step": 6790 }, { "grad_norm": 0.42834460735321045, "learning_rate": 7.872320059868355e-05, "loss": 0.0467, "step": 6800 }, { "grad_norm": 0.3765343725681305, "learning_rate": 7.865549064618729e-05, "loss": 0.0417, "step": 6810 }, { "grad_norm": 0.3671344518661499, "learning_rate": 7.858770235067381e-05, "loss": 0.0433, "step": 6820 }, { "grad_norm": 0.5173855423927307, "learning_rate": 7.851983589747374e-05, "loss": 0.0443, "step": 6830 }, { "grad_norm": 0.48901090025901794, "learning_rate": 7.845189147213133e-05, "loss": 0.0478, "step": 6840 }, { "grad_norm": 0.5155407190322876, "learning_rate": 7.838386926040407e-05, "loss": 0.0465, "step": 6850 }, { "grad_norm": 0.4107778072357178, "learning_rate": 7.83157694482621e-05, "loss": 0.0446, "step": 6860 }, { "grad_norm": 0.5714010000228882, "learning_rate": 7.824759222188768e-05, "loss": 0.0458, "step": 6870 }, { "grad_norm": 0.40384605526924133, "learning_rate": 7.817933776767478e-05, "loss": 0.0488, "step": 6880 }, { "grad_norm": 0.39226093888282776, "learning_rate": 7.811100627222842e-05, "loss": 0.0419, "step": 6890 }, { "grad_norm": 0.4260278642177582, "learning_rate": 7.804259792236435e-05, "loss": 0.041, "step": 6900 }, { "grad_norm": 0.4284556210041046, "learning_rate": 7.797411290510835e-05, "loss": 0.042, "step": 6910 }, { "grad_norm": 0.45743831992149353, "learning_rate": 7.790555140769586e-05, "loss": 0.0467, "step": 6920 }, { "grad_norm": 0.3447015881538391, "learning_rate": 7.78369136175714e-05, "loss": 0.0406, "step": 6930 }, { "grad_norm": 0.3831125497817993, "learning_rate": 7.776819972238806e-05, "loss": 0.0464, "step": 6940 }, { "grad_norm": 0.4599185287952423, "learning_rate": 7.7699409910007e-05, "loss": 0.0427, "step": 6950 }, { "grad_norm": 0.4301486611366272, "learning_rate": 7.763054436849694e-05, "loss": 0.0445, "step": 6960 }, { "grad_norm": 0.3690599501132965, "learning_rate": 7.756160328613364e-05, "loss": 0.0391, "step": 6970 }, { "grad_norm": 0.35586291551589966, "learning_rate": 7.749258685139942e-05, "loss": 0.0459, "step": 6980 }, { "grad_norm": 0.35467594861984253, "learning_rate": 7.742349525298253e-05, "loss": 0.0449, "step": 6990 }, { "grad_norm": 0.4966159760951996, "learning_rate": 7.735432867977679e-05, "loss": 0.0415, "step": 7000 }, { "grad_norm": 0.4681990146636963, "learning_rate": 7.728508732088096e-05, "loss": 0.0455, "step": 7010 }, { "grad_norm": 0.4373011589050293, "learning_rate": 7.721577136559825e-05, "loss": 0.0407, "step": 7020 }, { "grad_norm": 0.5192168951034546, "learning_rate": 7.714638100343588e-05, "loss": 0.045, "step": 7030 }, { "grad_norm": 0.5360577702522278, "learning_rate": 7.707691642410444e-05, "loss": 0.0397, "step": 7040 }, { "grad_norm": 0.40055781602859497, "learning_rate": 7.70073778175174e-05, "loss": 0.0427, "step": 7050 }, { "grad_norm": 0.43129482865333557, "learning_rate": 7.69377653737907e-05, "loss": 0.0445, "step": 7060 }, { "grad_norm": 0.49011117219924927, "learning_rate": 7.686807928324209e-05, "loss": 0.0438, "step": 7070 }, { "grad_norm": 0.5534836649894714, "learning_rate": 7.679831973639065e-05, "loss": 0.0415, "step": 7080 }, { "grad_norm": 0.6506836414337158, "learning_rate": 7.672848692395637e-05, "loss": 0.0441, "step": 7090 }, { "grad_norm": 0.4654604494571686, "learning_rate": 7.665858103685944e-05, "loss": 0.0443, "step": 7100 }, { "grad_norm": 0.5389629006385803, "learning_rate": 7.658860226621991e-05, "loss": 0.0474, "step": 7110 }, { "grad_norm": 0.43550923466682434, "learning_rate": 7.651855080335708e-05, "loss": 0.0468, "step": 7120 }, { "grad_norm": 0.4543823003768921, "learning_rate": 7.644842683978896e-05, "loss": 0.0461, "step": 7130 }, { "grad_norm": 0.5698819756507874, "learning_rate": 7.63782305672318e-05, "loss": 0.0435, "step": 7140 }, { "grad_norm": 0.3969264030456543, "learning_rate": 7.63079621775995e-05, "loss": 0.0439, "step": 7150 }, { "grad_norm": 0.44757407903671265, "learning_rate": 7.623762186300319e-05, "loss": 0.047, "step": 7160 }, { "grad_norm": 0.4573357105255127, "learning_rate": 7.616720981575057e-05, "loss": 0.0456, "step": 7170 }, { "grad_norm": 0.479949951171875, "learning_rate": 7.609672622834552e-05, "loss": 0.0441, "step": 7180 }, { "grad_norm": 0.3876534700393677, "learning_rate": 7.602617129348747e-05, "loss": 0.0425, "step": 7190 }, { "grad_norm": 0.45770972967147827, "learning_rate": 7.595554520407088e-05, "loss": 0.0436, "step": 7200 }, { "grad_norm": 0.4312308430671692, "learning_rate": 7.588484815318484e-05, "loss": 0.0389, "step": 7210 }, { "grad_norm": 0.45063266158103943, "learning_rate": 7.581408033411234e-05, "loss": 0.0433, "step": 7220 }, { "grad_norm": 0.3957497179508209, "learning_rate": 7.574324194032995e-05, "loss": 0.0415, "step": 7230 }, { "grad_norm": 0.3903205990791321, "learning_rate": 7.567233316550705e-05, "loss": 0.0398, "step": 7240 }, { "grad_norm": 0.3856774568557739, "learning_rate": 7.560135420350562e-05, "loss": 0.0419, "step": 7250 }, { "grad_norm": 0.44264569878578186, "learning_rate": 7.553030524837935e-05, "loss": 0.0419, "step": 7260 }, { "grad_norm": 0.35515904426574707, "learning_rate": 7.545918649437341e-05, "loss": 0.0411, "step": 7270 }, { "grad_norm": 0.49978896975517273, "learning_rate": 7.538799813592377e-05, "loss": 0.0456, "step": 7280 }, { "grad_norm": 0.5393272638320923, "learning_rate": 7.531674036765662e-05, "loss": 0.0398, "step": 7290 }, { "grad_norm": 0.5591117739677429, "learning_rate": 7.524541338438807e-05, "loss": 0.0433, "step": 7300 }, { "grad_norm": 0.3824617266654968, "learning_rate": 7.517401738112328e-05, "loss": 0.042, "step": 7310 }, { "grad_norm": 0.44299110770225525, "learning_rate": 7.510255255305628e-05, "loss": 0.0417, "step": 7320 }, { "grad_norm": 0.38524433970451355, "learning_rate": 7.503101909556911e-05, "loss": 0.0393, "step": 7330 }, { "grad_norm": 0.3656904399394989, "learning_rate": 7.495941720423154e-05, "loss": 0.0434, "step": 7340 }, { "grad_norm": 0.34718024730682373, "learning_rate": 7.488774707480042e-05, "loss": 0.0419, "step": 7350 }, { "grad_norm": 0.3500952422618866, "learning_rate": 7.481600890321911e-05, "loss": 0.0401, "step": 7360 }, { "grad_norm": 0.39389342069625854, "learning_rate": 7.474420288561708e-05, "loss": 0.0414, "step": 7370 }, { "grad_norm": 0.444288432598114, "learning_rate": 7.467232921830921e-05, "loss": 0.0455, "step": 7380 }, { "grad_norm": 0.4019412696361542, "learning_rate": 7.460038809779537e-05, "loss": 0.0422, "step": 7390 }, { "grad_norm": 0.4176471531391144, "learning_rate": 7.452837972075983e-05, "loss": 0.0426, "step": 7400 }, { "grad_norm": 0.47806599736213684, "learning_rate": 7.445630428407074e-05, "loss": 0.0399, "step": 7410 }, { "grad_norm": 0.395693302154541, "learning_rate": 7.43841619847796e-05, "loss": 0.042, "step": 7420 }, { "grad_norm": 0.4286665618419647, "learning_rate": 7.431195302012072e-05, "loss": 0.0401, "step": 7430 }, { "grad_norm": 0.4172682464122772, "learning_rate": 7.423967758751061e-05, "loss": 0.0428, "step": 7440 }, { "grad_norm": 0.39035043120384216, "learning_rate": 7.416733588454758e-05, "loss": 0.0425, "step": 7450 }, { "grad_norm": 0.3431130051612854, "learning_rate": 7.409492810901106e-05, "loss": 0.0426, "step": 7460 }, { "grad_norm": 0.3785865008831024, "learning_rate": 7.402245445886116e-05, "loss": 0.0449, "step": 7470 }, { "grad_norm": 0.3927653729915619, "learning_rate": 7.394991513223806e-05, "loss": 0.0399, "step": 7480 }, { "grad_norm": 0.32441505789756775, "learning_rate": 7.38773103274615e-05, "loss": 0.0428, "step": 7490 }, { "grad_norm": 0.3413672149181366, "learning_rate": 7.380464024303028e-05, "loss": 0.0423, "step": 7500 }, { "grad_norm": 0.4154110848903656, "learning_rate": 7.373190507762162e-05, "loss": 0.0443, "step": 7510 }, { "grad_norm": 0.4669798016548157, "learning_rate": 7.365910503009066e-05, "loss": 0.0397, "step": 7520 }, { "grad_norm": 0.39043378829956055, "learning_rate": 7.358624029946996e-05, "loss": 0.0421, "step": 7530 }, { "grad_norm": 0.39156562089920044, "learning_rate": 7.351331108496893e-05, "loss": 0.0477, "step": 7540 }, { "grad_norm": 0.3723638653755188, "learning_rate": 7.344031758597325e-05, "loss": 0.0442, "step": 7550 }, { "grad_norm": 0.3827393054962158, "learning_rate": 7.336726000204435e-05, "loss": 0.0431, "step": 7560 }, { "grad_norm": 0.3807544410228729, "learning_rate": 7.32941385329189e-05, "loss": 0.0412, "step": 7570 }, { "grad_norm": 0.40453916788101196, "learning_rate": 7.322095337850816e-05, "loss": 0.0391, "step": 7580 }, { "grad_norm": 0.4249589145183563, "learning_rate": 7.314770473889758e-05, "loss": 0.0412, "step": 7590 }, { "grad_norm": 0.35291120409965515, "learning_rate": 7.307439281434615e-05, "loss": 0.0433, "step": 7600 }, { "grad_norm": 0.35160666704177856, "learning_rate": 7.300101780528585e-05, "loss": 0.0422, "step": 7610 }, { "grad_norm": 0.3762139081954956, "learning_rate": 7.292757991232117e-05, "loss": 0.0408, "step": 7620 }, { "grad_norm": 0.44464778900146484, "learning_rate": 7.285407933622848e-05, "loss": 0.0402, "step": 7630 }, { "grad_norm": 0.3863670527935028, "learning_rate": 7.278051627795557e-05, "loss": 0.0418, "step": 7640 }, { "grad_norm": 0.42160648107528687, "learning_rate": 7.270689093862105e-05, "loss": 0.0437, "step": 7650 }, { "grad_norm": 0.5077595710754395, "learning_rate": 7.263320351951374e-05, "loss": 0.0362, "step": 7660 }, { "grad_norm": 0.4446108937263489, "learning_rate": 7.255945422209227e-05, "loss": 0.041, "step": 7670 }, { "grad_norm": 0.4235784709453583, "learning_rate": 7.248564324798437e-05, "loss": 0.0421, "step": 7680 }, { "grad_norm": 0.3724041283130646, "learning_rate": 7.241177079898644e-05, "loss": 0.0434, "step": 7690 }, { "grad_norm": 0.34507879614830017, "learning_rate": 7.233783707706295e-05, "loss": 0.0437, "step": 7700 }, { "grad_norm": 0.4122774302959442, "learning_rate": 7.226384228434586e-05, "loss": 0.0408, "step": 7710 }, { "grad_norm": 0.42959263920783997, "learning_rate": 7.21897866231341e-05, "loss": 0.0428, "step": 7720 }, { "grad_norm": 0.39758214354515076, "learning_rate": 7.211567029589303e-05, "loss": 0.0393, "step": 7730 }, { "grad_norm": 0.3889966607093811, "learning_rate": 7.204149350525387e-05, "loss": 0.0399, "step": 7740 }, { "grad_norm": 0.38079938292503357, "learning_rate": 7.196725645401309e-05, "loss": 0.0452, "step": 7750 }, { "grad_norm": 0.4374406039714813, "learning_rate": 7.1892959345132e-05, "loss": 0.0437, "step": 7760 }, { "grad_norm": 0.34991899132728577, "learning_rate": 7.181860238173605e-05, "loss": 0.0415, "step": 7770 }, { "grad_norm": 0.3476291000843048, "learning_rate": 7.174418576711432e-05, "loss": 0.0441, "step": 7780 }, { "grad_norm": 0.45317548513412476, "learning_rate": 7.1669709704719e-05, "loss": 0.0433, "step": 7790 }, { "grad_norm": 0.46312931180000305, "learning_rate": 7.159517439816481e-05, "loss": 0.0385, "step": 7800 }, { "grad_norm": 0.4653787910938263, "learning_rate": 7.152058005122842e-05, "loss": 0.0396, "step": 7810 }, { "grad_norm": 0.48601633310317993, "learning_rate": 7.144592686784793e-05, "loss": 0.0391, "step": 7820 }, { "grad_norm": 0.40843722224235535, "learning_rate": 7.137121505212229e-05, "loss": 0.041, "step": 7830 }, { "grad_norm": 0.4419427216053009, "learning_rate": 7.129644480831077e-05, "loss": 0.0388, "step": 7840 }, { "grad_norm": 0.40701529383659363, "learning_rate": 7.122161634083234e-05, "loss": 0.0396, "step": 7850 }, { "grad_norm": 0.5698205828666687, "learning_rate": 7.114672985426516e-05, "loss": 0.0396, "step": 7860 }, { "grad_norm": 0.3405832350254059, "learning_rate": 7.107178555334606e-05, "loss": 0.0389, "step": 7870 }, { "grad_norm": 0.4127151668071747, "learning_rate": 7.099678364296989e-05, "loss": 0.0434, "step": 7880 }, { "grad_norm": 0.3826993405818939, "learning_rate": 7.0921724328189e-05, "loss": 0.0436, "step": 7890 }, { "grad_norm": 0.34934094548225403, "learning_rate": 7.084660781421268e-05, "loss": 0.0468, "step": 7900 }, { "grad_norm": 0.439170777797699, "learning_rate": 7.077143430640662e-05, "loss": 0.0417, "step": 7910 }, { "grad_norm": 0.46485957503318787, "learning_rate": 7.069620401029232e-05, "loss": 0.0401, "step": 7920 }, { "grad_norm": 0.37791186571121216, "learning_rate": 7.062091713154655e-05, "loss": 0.0466, "step": 7930 }, { "grad_norm": 0.33394289016723633, "learning_rate": 7.054557387600075e-05, "loss": 0.0409, "step": 7940 }, { "grad_norm": 0.32268017530441284, "learning_rate": 7.04701744496405e-05, "loss": 0.0426, "step": 7950 }, { "grad_norm": 0.3536238372325897, "learning_rate": 7.039471905860495e-05, "loss": 0.0414, "step": 7960 }, { "grad_norm": 0.3841630220413208, "learning_rate": 7.031920790918628e-05, "loss": 0.039, "step": 7970 }, { "grad_norm": 0.3383556604385376, "learning_rate": 7.024364120782906e-05, "loss": 0.0403, "step": 7980 }, { "grad_norm": 0.43794724345207214, "learning_rate": 7.016801916112978e-05, "loss": 0.0412, "step": 7990 }, { "grad_norm": 0.4184318482875824, "learning_rate": 7.009234197583623e-05, "loss": 0.0385, "step": 8000 }, { "grad_norm": 0.4434663653373718, "learning_rate": 7.001660985884692e-05, "loss": 0.0389, "step": 8010 }, { "grad_norm": 0.45659393072128296, "learning_rate": 6.994082301721063e-05, "loss": 0.0382, "step": 8020 }, { "grad_norm": 0.3736106753349304, "learning_rate": 6.986498165812563e-05, "loss": 0.0435, "step": 8030 }, { "grad_norm": 0.5281041860580444, "learning_rate": 6.978908598893932e-05, "loss": 0.0412, "step": 8040 }, { "grad_norm": 0.5153106451034546, "learning_rate": 6.971313621714756e-05, "loss": 0.0413, "step": 8050 }, { "grad_norm": 0.4496375322341919, "learning_rate": 6.96371325503941e-05, "loss": 0.0418, "step": 8060 }, { "grad_norm": 0.4705381691455841, "learning_rate": 6.956107519647014e-05, "loss": 0.0394, "step": 8070 }, { "grad_norm": 0.3798416554927826, "learning_rate": 6.94849643633135e-05, "loss": 0.0408, "step": 8080 }, { "grad_norm": 0.374737024307251, "learning_rate": 6.940880025900834e-05, "loss": 0.0429, "step": 8090 }, { "grad_norm": 0.37165018916130066, "learning_rate": 6.933258309178438e-05, "loss": 0.0443, "step": 8100 }, { "grad_norm": 0.3653257191181183, "learning_rate": 6.925631307001646e-05, "loss": 0.0426, "step": 8110 }, { "grad_norm": 0.3856312930583954, "learning_rate": 6.91799904022239e-05, "loss": 0.0364, "step": 8120 }, { "grad_norm": 0.37950006127357483, "learning_rate": 6.910361529706997e-05, "loss": 0.042, "step": 8130 }, { "grad_norm": 0.4022243022918701, "learning_rate": 6.902718796336131e-05, "loss": 0.0431, "step": 8140 }, { "grad_norm": 0.45228278636932373, "learning_rate": 6.895070861004729e-05, "loss": 0.046, "step": 8150 }, { "grad_norm": 0.34973660111427307, "learning_rate": 6.887417744621956e-05, "loss": 0.0405, "step": 8160 }, { "grad_norm": 0.386653333902359, "learning_rate": 6.87975946811114e-05, "loss": 0.0423, "step": 8170 }, { "grad_norm": 0.4350385367870331, "learning_rate": 6.872096052409718e-05, "loss": 0.04, "step": 8180 }, { "grad_norm": 0.3624569773674011, "learning_rate": 6.864427518469174e-05, "loss": 0.0387, "step": 8190 }, { "grad_norm": 0.3344216048717499, "learning_rate": 6.856753887254986e-05, "loss": 0.0405, "step": 8200 }, { "grad_norm": 0.3799411952495575, "learning_rate": 6.849075179746572e-05, "loss": 0.0401, "step": 8210 }, { "grad_norm": 0.42303794622421265, "learning_rate": 6.841391416937221e-05, "loss": 0.039, "step": 8220 }, { "grad_norm": 0.35377931594848633, "learning_rate": 6.833702619834053e-05, "loss": 0.0384, "step": 8230 }, { "grad_norm": 0.3753434717655182, "learning_rate": 6.82600880945794e-05, "loss": 0.0391, "step": 8240 }, { "grad_norm": 0.3280651271343231, "learning_rate": 6.818310006843468e-05, "loss": 0.0394, "step": 8250 }, { "grad_norm": 0.31104692816734314, "learning_rate": 6.810606233038868e-05, "loss": 0.039, "step": 8260 }, { "grad_norm": 0.36656251549720764, "learning_rate": 6.802897509105966e-05, "loss": 0.0376, "step": 8270 }, { "grad_norm": 0.4361065924167633, "learning_rate": 6.79518385612012e-05, "loss": 0.0388, "step": 8280 }, { "grad_norm": 0.4188891053199768, "learning_rate": 6.787465295170157e-05, "loss": 0.0396, "step": 8290 }, { "grad_norm": 0.3269003629684448, "learning_rate": 6.779741847358332e-05, "loss": 0.0418, "step": 8300 }, { "grad_norm": 0.3358224928379059, "learning_rate": 6.772013533800256e-05, "loss": 0.0388, "step": 8310 }, { "grad_norm": 0.37758901715278625, "learning_rate": 6.764280375624843e-05, "loss": 0.038, "step": 8320 }, { "grad_norm": 0.3867717385292053, "learning_rate": 6.756542393974252e-05, "loss": 0.0377, "step": 8330 }, { "grad_norm": 0.4209394156932831, "learning_rate": 6.748799610003828e-05, "loss": 0.0398, "step": 8340 }, { "grad_norm": 0.2888592481613159, "learning_rate": 6.741052044882048e-05, "loss": 0.0413, "step": 8350 }, { "grad_norm": 0.2753848135471344, "learning_rate": 6.73329971979046e-05, "loss": 0.0362, "step": 8360 }, { "grad_norm": 0.3237358629703522, "learning_rate": 6.725542655923625e-05, "loss": 0.0363, "step": 8370 }, { "grad_norm": 0.3385677635669708, "learning_rate": 6.717780874489057e-05, "loss": 0.0389, "step": 8380 }, { "grad_norm": 0.3221762180328369, "learning_rate": 6.710014396707172e-05, "loss": 0.038, "step": 8390 }, { "grad_norm": 0.40469878911972046, "learning_rate": 6.702243243811221e-05, "loss": 0.0394, "step": 8400 }, { "grad_norm": 0.4259459376335144, "learning_rate": 6.694467437047244e-05, "loss": 0.0393, "step": 8410 }, { "grad_norm": 0.5154764652252197, "learning_rate": 6.686686997673997e-05, "loss": 0.0403, "step": 8420 }, { "grad_norm": 0.4254332482814789, "learning_rate": 6.678901946962903e-05, "loss": 0.0389, "step": 8430 }, { "grad_norm": 0.4097931385040283, "learning_rate": 6.671112306197996e-05, "loss": 0.038, "step": 8440 }, { "grad_norm": 0.41378700733184814, "learning_rate": 6.663318096675854e-05, "loss": 0.0398, "step": 8450 }, { "grad_norm": 0.3372202515602112, "learning_rate": 6.655519339705552e-05, "loss": 0.0378, "step": 8460 }, { "grad_norm": 0.38872626423835754, "learning_rate": 6.647716056608588e-05, "loss": 0.0363, "step": 8470 }, { "grad_norm": 0.4180516004562378, "learning_rate": 6.639908268718843e-05, "loss": 0.0375, "step": 8480 }, { "grad_norm": 0.3590203523635864, "learning_rate": 6.632095997382514e-05, "loss": 0.0375, "step": 8490 }, { "grad_norm": 0.6900638937950134, "learning_rate": 6.624279263958047e-05, "loss": 0.0428, "step": 8500 }, { "grad_norm": 0.40207773447036743, "learning_rate": 6.616458089816097e-05, "loss": 0.0409, "step": 8510 }, { "grad_norm": 0.2881513237953186, "learning_rate": 6.608632496339454e-05, "loss": 0.0402, "step": 8520 }, { "grad_norm": 0.34924304485321045, "learning_rate": 6.600802504922988e-05, "loss": 0.0386, "step": 8530 }, { "grad_norm": 0.5409790873527527, "learning_rate": 6.592968136973604e-05, "loss": 0.0381, "step": 8540 }, { "grad_norm": 0.4964514672756195, "learning_rate": 6.585129413910159e-05, "loss": 0.0389, "step": 8550 }, { "grad_norm": 0.5977535247802734, "learning_rate": 6.577286357163424e-05, "loss": 0.0416, "step": 8560 }, { "grad_norm": 0.5585235953330994, "learning_rate": 6.569438988176018e-05, "loss": 0.0414, "step": 8570 }, { "grad_norm": 0.3908786177635193, "learning_rate": 6.561587328402347e-05, "loss": 0.0392, "step": 8580 }, { "grad_norm": 0.29729393124580383, "learning_rate": 6.553731399308549e-05, "loss": 0.0415, "step": 8590 }, { "grad_norm": 0.4183690547943115, "learning_rate": 6.545871222372436e-05, "loss": 0.0388, "step": 8600 }, { "grad_norm": 0.43635568022727966, "learning_rate": 6.538006819083426e-05, "loss": 0.0402, "step": 8610 }, { "grad_norm": 0.4668086767196655, "learning_rate": 6.530138210942505e-05, "loss": 0.0403, "step": 8620 }, { "grad_norm": 0.5347704291343689, "learning_rate": 6.522265419462141e-05, "loss": 0.0395, "step": 8630 }, { "grad_norm": 0.4697801172733307, "learning_rate": 6.514388466166248e-05, "loss": 0.0388, "step": 8640 }, { "grad_norm": 0.498806357383728, "learning_rate": 6.506507372590119e-05, "loss": 0.0385, "step": 8650 }, { "grad_norm": 0.4334677755832672, "learning_rate": 6.498622160280355e-05, "loss": 0.041, "step": 8660 }, { "grad_norm": 0.35923147201538086, "learning_rate": 6.490732850794832e-05, "loss": 0.0364, "step": 8670 }, { "grad_norm": 0.39771413803100586, "learning_rate": 6.482839465702616e-05, "loss": 0.0369, "step": 8680 }, { "grad_norm": 0.44982534646987915, "learning_rate": 6.474942026583923e-05, "loss": 0.0394, "step": 8690 }, { "grad_norm": 0.4596177637577057, "learning_rate": 6.467040555030052e-05, "loss": 0.0411, "step": 8700 }, { "grad_norm": 0.47968462109565735, "learning_rate": 6.459135072643321e-05, "loss": 0.0367, "step": 8710 }, { "grad_norm": 0.40840670466423035, "learning_rate": 6.451225601037019e-05, "loss": 0.034, "step": 8720 }, { "grad_norm": 0.3966841399669647, "learning_rate": 6.443312161835338e-05, "loss": 0.039, "step": 8730 }, { "grad_norm": 0.3349541127681732, "learning_rate": 6.43539477667332e-05, "loss": 0.0372, "step": 8740 }, { "grad_norm": 0.4185710847377777, "learning_rate": 6.427473467196793e-05, "loss": 0.0374, "step": 8750 }, { "grad_norm": 0.40412721037864685, "learning_rate": 6.419548255062315e-05, "loss": 0.0387, "step": 8760 }, { "grad_norm": 0.3526269495487213, "learning_rate": 6.411619161937112e-05, "loss": 0.041, "step": 8770 }, { "grad_norm": 0.3758630156517029, "learning_rate": 6.403686209499022e-05, "loss": 0.0392, "step": 8780 }, { "grad_norm": 0.3683103621006012, "learning_rate": 6.395749419436437e-05, "loss": 0.0399, "step": 8790 }, { "grad_norm": 0.43625909090042114, "learning_rate": 6.387808813448234e-05, "loss": 0.0394, "step": 8800 }, { "grad_norm": 0.4969613254070282, "learning_rate": 6.37986441324373e-05, "loss": 0.0354, "step": 8810 }, { "grad_norm": 0.4200226366519928, "learning_rate": 6.37191624054261e-05, "loss": 0.042, "step": 8820 }, { "grad_norm": 0.40950220823287964, "learning_rate": 6.363964317074872e-05, "loss": 0.0407, "step": 8830 }, { "grad_norm": 0.42140623927116394, "learning_rate": 6.356008664580776e-05, "loss": 0.0365, "step": 8840 }, { "grad_norm": 0.3767819106578827, "learning_rate": 6.348049304810771e-05, "loss": 0.0392, "step": 8850 }, { "grad_norm": 0.41109201312065125, "learning_rate": 6.340086259525442e-05, "loss": 0.0358, "step": 8860 }, { "grad_norm": 0.40867123007774353, "learning_rate": 6.332119550495448e-05, "loss": 0.0359, "step": 8870 }, { "grad_norm": 0.38878798484802246, "learning_rate": 6.324149199501473e-05, "loss": 0.0348, "step": 8880 }, { "grad_norm": 0.43662306666374207, "learning_rate": 6.316175228334146e-05, "loss": 0.0369, "step": 8890 }, { "grad_norm": 0.40835919976234436, "learning_rate": 6.308197658794003e-05, "loss": 0.037, "step": 8900 }, { "grad_norm": 0.40312886238098145, "learning_rate": 6.300216512691417e-05, "loss": 0.0393, "step": 8910 }, { "grad_norm": 0.33371880650520325, "learning_rate": 6.292231811846532e-05, "loss": 0.0388, "step": 8920 }, { "grad_norm": 0.3376159965991974, "learning_rate": 6.284243578089217e-05, "loss": 0.0388, "step": 8930 }, { "grad_norm": 0.3949618637561798, "learning_rate": 6.276251833258999e-05, "loss": 0.0374, "step": 8940 }, { "grad_norm": 0.40749624371528625, "learning_rate": 6.268256599205003e-05, "loss": 0.0366, "step": 8950 }, { "grad_norm": 0.3583541214466095, "learning_rate": 6.260257897785892e-05, "loss": 0.0368, "step": 8960 }, { "grad_norm": 0.3517057001590729, "learning_rate": 6.252255750869811e-05, "loss": 0.0362, "step": 8970 }, { "grad_norm": 0.3645634055137634, "learning_rate": 6.244250180334325e-05, "loss": 0.0356, "step": 8980 }, { "grad_norm": 0.4441187381744385, "learning_rate": 6.236241208066356e-05, "loss": 0.0355, "step": 8990 }, { "grad_norm": 0.3607874810695648, "learning_rate": 6.228228855962133e-05, "loss": 0.0359, "step": 9000 }, { "grad_norm": 0.42591002583503723, "learning_rate": 6.220213145927115e-05, "loss": 0.0363, "step": 9010 }, { "grad_norm": 0.26376354694366455, "learning_rate": 6.212194099875951e-05, "loss": 0.038, "step": 9020 }, { "grad_norm": 0.4516811966896057, "learning_rate": 6.204171739732405e-05, "loss": 0.0361, "step": 9030 }, { "grad_norm": 0.46166104078292847, "learning_rate": 6.196146087429303e-05, "loss": 0.0386, "step": 9040 }, { "grad_norm": 0.4406202435493469, "learning_rate": 6.188117164908474e-05, "loss": 0.0365, "step": 9050 }, { "grad_norm": 0.41246140003204346, "learning_rate": 6.180084994120684e-05, "loss": 0.0355, "step": 9060 }, { "grad_norm": 0.4201594293117523, "learning_rate": 6.17204959702558e-05, "loss": 0.0399, "step": 9070 }, { "grad_norm": 0.38419055938720703, "learning_rate": 6.164010995591635e-05, "loss": 0.0346, "step": 9080 }, { "grad_norm": 0.38071274757385254, "learning_rate": 6.155969211796076e-05, "loss": 0.0388, "step": 9090 }, { "grad_norm": 0.39762747287750244, "learning_rate": 6.147924267624829e-05, "loss": 0.0361, "step": 9100 }, { "grad_norm": 0.31926289200782776, "learning_rate": 6.13987618507247e-05, "loss": 0.0365, "step": 9110 }, { "grad_norm": 0.34313949942588806, "learning_rate": 6.131824986142147e-05, "loss": 0.0358, "step": 9120 }, { "grad_norm": 0.3615455627441406, "learning_rate": 6.123770692845529e-05, "loss": 0.0375, "step": 9130 }, { "grad_norm": 0.3325113356113434, "learning_rate": 6.11571332720275e-05, "loss": 0.0382, "step": 9140 }, { "grad_norm": 0.4277583658695221, "learning_rate": 6.107652911242336e-05, "loss": 0.0393, "step": 9150 }, { "grad_norm": 0.38978931307792664, "learning_rate": 6.0995894670011586e-05, "loss": 0.0383, "step": 9160 }, { "grad_norm": 0.49869653582572937, "learning_rate": 6.091523016524368e-05, "loss": 0.0418, "step": 9170 }, { "grad_norm": 0.4501328766345978, "learning_rate": 6.083453581865328e-05, "loss": 0.0365, "step": 9180 }, { "grad_norm": 0.49607619643211365, "learning_rate": 6.075381185085568e-05, "loss": 0.0412, "step": 9190 }, { "grad_norm": 0.47689756751060486, "learning_rate": 6.067305848254709e-05, "loss": 0.0386, "step": 9200 }, { "grad_norm": 0.40081101655960083, "learning_rate": 6.059227593450418e-05, "loss": 0.0394, "step": 9210 }, { "grad_norm": 0.4014520049095154, "learning_rate": 6.051146442758333e-05, "loss": 0.0363, "step": 9220 }, { "grad_norm": 0.37776657938957214, "learning_rate": 6.043062418272012e-05, "loss": 0.0368, "step": 9230 }, { "grad_norm": 0.3149407207965851, "learning_rate": 6.0349755420928666e-05, "loss": 0.0345, "step": 9240 }, { "grad_norm": 0.3492046594619751, "learning_rate": 6.0268858363301105e-05, "loss": 0.0357, "step": 9250 }, { "grad_norm": 0.3109935224056244, "learning_rate": 6.018793323100689e-05, "loss": 0.0396, "step": 9260 }, { "grad_norm": 0.4470113515853882, "learning_rate": 6.0106980245292255e-05, "loss": 0.038, "step": 9270 }, { "grad_norm": 0.41872069239616394, "learning_rate": 6.002599962747957e-05, "loss": 0.0363, "step": 9280 }, { "grad_norm": 0.3347885310649872, "learning_rate": 5.994499159896673e-05, "loss": 0.0364, "step": 9290 }, { "grad_norm": 0.3755440413951874, "learning_rate": 5.9863956381226607e-05, "loss": 0.0371, "step": 9300 }, { "grad_norm": 0.36123523116111755, "learning_rate": 5.9782894195806394e-05, "loss": 0.0359, "step": 9310 }, { "grad_norm": 0.37258800864219666, "learning_rate": 5.9701805264327004e-05, "loss": 0.0363, "step": 9320 }, { "grad_norm": 0.33783793449401855, "learning_rate": 5.96206898084825e-05, "loss": 0.0344, "step": 9330 }, { "grad_norm": 0.44993335008621216, "learning_rate": 5.953954805003942e-05, "loss": 0.0355, "step": 9340 }, { "grad_norm": 0.4486129581928253, "learning_rate": 5.945838021083623e-05, "loss": 0.0388, "step": 9350 }, { "grad_norm": 0.4506584405899048, "learning_rate": 5.9377186512782714e-05, "loss": 0.0388, "step": 9360 }, { "grad_norm": 0.5251940488815308, "learning_rate": 5.929596717785935e-05, "loss": 0.0386, "step": 9370 }, { "grad_norm": 0.4545130431652069, "learning_rate": 5.921472242811668e-05, "loss": 0.0385, "step": 9380 }, { "grad_norm": 0.4602312743663788, "learning_rate": 5.913345248567475e-05, "loss": 0.0373, "step": 9390 }, { "grad_norm": 0.3565285801887512, "learning_rate": 5.905215757272248e-05, "loss": 0.0379, "step": 9400 }, { "grad_norm": 0.32982316613197327, "learning_rate": 5.897083791151706e-05, "loss": 0.0384, "step": 9410 }, { "grad_norm": 0.4797773063182831, "learning_rate": 5.888949372438336e-05, "loss": 0.0387, "step": 9420 }, { "grad_norm": 0.33019569516181946, "learning_rate": 5.8808125233713255e-05, "loss": 0.0383, "step": 9430 }, { "grad_norm": 0.3605564534664154, "learning_rate": 5.872673266196509e-05, "loss": 0.0357, "step": 9440 }, { "grad_norm": 0.44483819603919983, "learning_rate": 5.864531623166305e-05, "loss": 0.0354, "step": 9450 }, { "grad_norm": 0.3982445001602173, "learning_rate": 5.856387616539656e-05, "loss": 0.0371, "step": 9460 }, { "grad_norm": 0.3757554292678833, "learning_rate": 5.848241268581967e-05, "loss": 0.0346, "step": 9470 }, { "grad_norm": 0.3673747181892395, "learning_rate": 5.840092601565037e-05, "loss": 0.0362, "step": 9480 }, { "grad_norm": 0.31699010729789734, "learning_rate": 5.8319416377670144e-05, "loss": 0.039, "step": 9490 }, { "grad_norm": 0.3722139894962311, "learning_rate": 5.82378839947232e-05, "loss": 0.0369, "step": 9500 }, { "grad_norm": 0.34847187995910645, "learning_rate": 5.815632908971599e-05, "loss": 0.0356, "step": 9510 }, { "grad_norm": 0.344765305519104, "learning_rate": 5.80747518856165e-05, "loss": 0.0409, "step": 9520 }, { "grad_norm": 0.4038139879703522, "learning_rate": 5.799315260545367e-05, "loss": 0.0385, "step": 9530 }, { "grad_norm": 0.35588642954826355, "learning_rate": 5.791153147231686e-05, "loss": 0.037, "step": 9540 }, { "grad_norm": 0.3555145561695099, "learning_rate": 5.782988870935509e-05, "loss": 0.0393, "step": 9550 }, { "grad_norm": 0.3669459819793701, "learning_rate": 5.774822453977657e-05, "loss": 0.0398, "step": 9560 }, { "grad_norm": 0.41855886578559875, "learning_rate": 5.7666539186848036e-05, "loss": 0.0388, "step": 9570 }, { "grad_norm": 0.47370752692222595, "learning_rate": 5.758483287389411e-05, "loss": 0.0375, "step": 9580 }, { "grad_norm": 0.4525252878665924, "learning_rate": 5.7503105824296735e-05, "loss": 0.0354, "step": 9590 }, { "grad_norm": 0.5328614711761475, "learning_rate": 5.742135826149453e-05, "loss": 0.0403, "step": 9600 }, { "grad_norm": 0.5064107775688171, "learning_rate": 5.7339590408982223e-05, "loss": 0.0367, "step": 9610 }, { "grad_norm": 0.526964545249939, "learning_rate": 5.725780249031e-05, "loss": 0.0377, "step": 9620 }, { "grad_norm": 0.34784722328186035, "learning_rate": 5.717599472908292e-05, "loss": 0.0402, "step": 9630 }, { "grad_norm": 0.3463391065597534, "learning_rate": 5.7094167348960237e-05, "loss": 0.0371, "step": 9640 }, { "grad_norm": 0.3569811284542084, "learning_rate": 5.7012320573654945e-05, "loss": 0.0363, "step": 9650 }, { "grad_norm": 0.3847939670085907, "learning_rate": 5.693045462693295e-05, "loss": 0.0374, "step": 9660 }, { "grad_norm": 0.4336046874523163, "learning_rate": 5.684856973261266e-05, "loss": 0.0353, "step": 9670 }, { "grad_norm": 0.3406628370285034, "learning_rate": 5.6766666114564215e-05, "loss": 0.0343, "step": 9680 }, { "grad_norm": 0.42734235525131226, "learning_rate": 5.668474399670899e-05, "loss": 0.037, "step": 9690 }, { "grad_norm": 0.3883262872695923, "learning_rate": 5.660280360301896e-05, "loss": 0.0368, "step": 9700 }, { "grad_norm": 0.4002635180950165, "learning_rate": 5.652084515751599e-05, "loss": 0.0381, "step": 9710 }, { "grad_norm": 0.3676852583885193, "learning_rate": 5.643886888427137e-05, "loss": 0.0382, "step": 9720 }, { "grad_norm": 0.31124910712242126, "learning_rate": 5.6356875007405074e-05, "loss": 0.0328, "step": 9730 }, { "grad_norm": 0.3403521776199341, "learning_rate": 5.627486375108525e-05, "loss": 0.0338, "step": 9740 }, { "grad_norm": 0.2891167104244232, "learning_rate": 5.619283533952754e-05, "loss": 0.0323, "step": 9750 }, { "grad_norm": 0.3233775496482849, "learning_rate": 5.6110789996994474e-05, "loss": 0.0389, "step": 9760 }, { "grad_norm": 0.32156941294670105, "learning_rate": 5.602872794779491e-05, "loss": 0.0359, "step": 9770 }, { "grad_norm": 0.3124691843986511, "learning_rate": 5.594664941628334e-05, "loss": 0.0363, "step": 9780 }, { "grad_norm": 0.32511281967163086, "learning_rate": 5.5864554626859324e-05, "loss": 0.0352, "step": 9790 }, { "grad_norm": 0.35578253865242004, "learning_rate": 5.578244380396691e-05, "loss": 0.0331, "step": 9800 }, { "grad_norm": 0.31595662236213684, "learning_rate": 5.570031717209394e-05, "loss": 0.036, "step": 9810 }, { "grad_norm": 0.28294721245765686, "learning_rate": 5.561817495577147e-05, "loss": 0.0351, "step": 9820 }, { "grad_norm": 0.2633134722709656, "learning_rate": 5.5536017379573215e-05, "loss": 0.0381, "step": 9830 }, { "grad_norm": 0.27349770069122314, "learning_rate": 5.545384466811483e-05, "loss": 0.0375, "step": 9840 }, { "grad_norm": 0.35391056537628174, "learning_rate": 5.5371657046053384e-05, "loss": 0.0327, "step": 9850 }, { "grad_norm": 0.3499240279197693, "learning_rate": 5.528945473808669e-05, "loss": 0.0329, "step": 9860 }, { "grad_norm": 0.31064409017562866, "learning_rate": 5.520723796895272e-05, "loss": 0.034, "step": 9870 }, { "grad_norm": 0.30534276366233826, "learning_rate": 5.512500696342897e-05, "loss": 0.0352, "step": 9880 }, { "grad_norm": 0.33043572306632996, "learning_rate": 5.504276194633188e-05, "loss": 0.0339, "step": 9890 }, { "grad_norm": 0.31157317757606506, "learning_rate": 5.49605031425162e-05, "loss": 0.0379, "step": 9900 }, { "grad_norm": 0.29970693588256836, "learning_rate": 5.487823077687434e-05, "loss": 0.0348, "step": 9910 }, { "grad_norm": 0.29958575963974, "learning_rate": 5.4795945074335806e-05, "loss": 0.0341, "step": 9920 }, { "grad_norm": 0.3550179600715637, "learning_rate": 5.471364625986657e-05, "loss": 0.0343, "step": 9930 }, { "grad_norm": 0.3421706259250641, "learning_rate": 5.463133455846845e-05, "loss": 0.0343, "step": 9940 }, { "grad_norm": 0.31181326508522034, "learning_rate": 5.4549010195178505e-05, "loss": 0.0337, "step": 9950 }, { "grad_norm": 0.35365423560142517, "learning_rate": 5.446667339506838e-05, "loss": 0.0365, "step": 9960 }, { "grad_norm": 0.361100971698761, "learning_rate": 5.4384324383243756e-05, "loss": 0.0372, "step": 9970 }, { "grad_norm": 0.4163217544555664, "learning_rate": 5.430196338484368e-05, "loss": 0.0382, "step": 9980 }, { "grad_norm": 0.4440409541130066, "learning_rate": 5.4219590625039975e-05, "loss": 0.0348, "step": 9990 }, { "grad_norm": 0.3922373354434967, "learning_rate": 5.413720632903664e-05, "loss": 0.0337, "step": 10000 }, { "grad_norm": 0.295221745967865, "learning_rate": 5.405481072206917e-05, "loss": 0.0323, "step": 10010 }, { "grad_norm": 0.3418806791305542, "learning_rate": 5.397240402940402e-05, "loss": 0.0346, "step": 10020 }, { "grad_norm": 0.35579973459243774, "learning_rate": 5.388998647633794e-05, "loss": 0.0303, "step": 10030 }, { "grad_norm": 0.29146939516067505, "learning_rate": 5.380755828819737e-05, "loss": 0.0343, "step": 10040 }, { "grad_norm": 0.36018234491348267, "learning_rate": 5.3725119690337846e-05, "loss": 0.0327, "step": 10050 }, { "grad_norm": 0.3252221643924713, "learning_rate": 5.3642670908143324e-05, "loss": 0.0341, "step": 10060 }, { "grad_norm": 0.4648992121219635, "learning_rate": 5.356021216702562e-05, "loss": 0.0356, "step": 10070 }, { "grad_norm": 0.4105115532875061, "learning_rate": 5.347774369242381e-05, "loss": 0.0357, "step": 10080 }, { "grad_norm": 0.4321806728839874, "learning_rate": 5.3395265709803545e-05, "loss": 0.0351, "step": 10090 }, { "grad_norm": 0.3845973014831543, "learning_rate": 5.331277844465647e-05, "loss": 0.0349, "step": 10100 }, { "grad_norm": 0.3999387323856354, "learning_rate": 5.323028212249963e-05, "loss": 0.0392, "step": 10110 }, { "grad_norm": 0.37703561782836914, "learning_rate": 5.314777696887481e-05, "loss": 0.0369, "step": 10120 }, { "grad_norm": 0.33284875750541687, "learning_rate": 5.306526320934796e-05, "loss": 0.0345, "step": 10130 }, { "grad_norm": 0.32707634568214417, "learning_rate": 5.298274106950854e-05, "loss": 0.0334, "step": 10140 }, { "grad_norm": 0.2574497163295746, "learning_rate": 5.290021077496893e-05, "loss": 0.0343, "step": 10150 }, { "grad_norm": 0.2821744680404663, "learning_rate": 5.2817672551363816e-05, "loss": 0.0322, "step": 10160 }, { "grad_norm": 0.30553850531578064, "learning_rate": 5.273512662434952e-05, "loss": 0.0331, "step": 10170 }, { "grad_norm": 0.361822247505188, "learning_rate": 5.265257321960349e-05, "loss": 0.0359, "step": 10180 }, { "grad_norm": 0.3418872356414795, "learning_rate": 5.257001256282357e-05, "loss": 0.0382, "step": 10190 }, { "grad_norm": 0.3403816819190979, "learning_rate": 5.248744487972742e-05, "loss": 0.0331, "step": 10200 }, { "grad_norm": 0.3123070001602173, "learning_rate": 5.240487039605196e-05, "loss": 0.0328, "step": 10210 }, { "grad_norm": 0.30127012729644775, "learning_rate": 5.232228933755267e-05, "loss": 0.0343, "step": 10220 }, { "grad_norm": 0.31802234053611755, "learning_rate": 5.2239701930003006e-05, "loss": 0.0332, "step": 10230 }, { "grad_norm": 0.36530429124832153, "learning_rate": 5.215710839919379e-05, "loss": 0.0356, "step": 10240 }, { "grad_norm": 0.3415067195892334, "learning_rate": 5.207450897093257e-05, "loss": 0.0334, "step": 10250 }, { "grad_norm": 0.4484879970550537, "learning_rate": 5.1991903871043046e-05, "loss": 0.0326, "step": 10260 }, { "grad_norm": 0.39791139960289, "learning_rate": 5.190929332536439e-05, "loss": 0.0366, "step": 10270 }, { "grad_norm": 0.3716643452644348, "learning_rate": 5.182667755975071e-05, "loss": 0.034, "step": 10280 }, { "grad_norm": 0.3939172327518463, "learning_rate": 5.1744056800070315e-05, "loss": 0.0361, "step": 10290 }, { "grad_norm": 0.3341255187988281, "learning_rate": 5.166143127220524e-05, "loss": 0.0347, "step": 10300 }, { "grad_norm": 0.33152660727500916, "learning_rate": 5.1578801202050485e-05, "loss": 0.0329, "step": 10310 }, { "grad_norm": 0.3467143774032593, "learning_rate": 5.149616681551355e-05, "loss": 0.0338, "step": 10320 }, { "grad_norm": 0.27762672305107117, "learning_rate": 5.141352833851367e-05, "loss": 0.0318, "step": 10330 }, { "grad_norm": 0.2779501974582672, "learning_rate": 5.1330885996981285e-05, "loss": 0.0312, "step": 10340 }, { "grad_norm": 0.3416571319103241, "learning_rate": 5.124824001685741e-05, "loss": 0.0343, "step": 10350 }, { "grad_norm": 0.32800135016441345, "learning_rate": 5.116559062409298e-05, "loss": 0.0348, "step": 10360 }, { "grad_norm": 0.3508056104183197, "learning_rate": 5.10829380446483e-05, "loss": 0.0335, "step": 10370 }, { "grad_norm": 0.40828368067741394, "learning_rate": 5.100028250449235e-05, "loss": 0.0327, "step": 10380 }, { "grad_norm": 0.37461936473846436, "learning_rate": 5.0917624229602234e-05, "loss": 0.0339, "step": 10390 }, { "grad_norm": 0.3638474941253662, "learning_rate": 5.0834963445962524e-05, "loss": 0.031, "step": 10400 }, { "grad_norm": 0.3886479437351227, "learning_rate": 5.075230037956461e-05, "loss": 0.0338, "step": 10410 }, { "grad_norm": 0.35828426480293274, "learning_rate": 5.0669635256406213e-05, "loss": 0.0335, "step": 10420 }, { "grad_norm": 0.37223929166793823, "learning_rate": 5.058696830249058e-05, "loss": 0.0339, "step": 10430 }, { "grad_norm": 0.3858705163002014, "learning_rate": 5.050429974382602e-05, "loss": 0.0341, "step": 10440 }, { "grad_norm": 0.5186141729354858, "learning_rate": 5.042162980642523e-05, "loss": 0.0379, "step": 10450 }, { "grad_norm": 0.4434109032154083, "learning_rate": 5.033895871630462e-05, "loss": 0.0368, "step": 10460 }, { "grad_norm": 0.3636118173599243, "learning_rate": 5.025628669948386e-05, "loss": 0.0334, "step": 10470 }, { "grad_norm": 0.3104057312011719, "learning_rate": 5.017361398198502e-05, "loss": 0.0356, "step": 10480 }, { "grad_norm": 0.3214111328125, "learning_rate": 5.009094078983221e-05, "loss": 0.036, "step": 10490 }, { "grad_norm": 0.4062266945838928, "learning_rate": 5.000826734905073e-05, "loss": 0.0333, "step": 10500 }, { "grad_norm": 0.38136008381843567, "learning_rate": 4.9925593885666645e-05, "loss": 0.0377, "step": 10510 }, { "grad_norm": 0.34270915389060974, "learning_rate": 4.984292062570602e-05, "loss": 0.0339, "step": 10520 }, { "grad_norm": 0.3638343811035156, "learning_rate": 4.976024779519442e-05, "loss": 0.0324, "step": 10530 }, { "grad_norm": 0.344582200050354, "learning_rate": 4.9677575620156194e-05, "loss": 0.0316, "step": 10540 }, { "grad_norm": 0.3976815342903137, "learning_rate": 4.959490432661391e-05, "loss": 0.0337, "step": 10550 }, { "grad_norm": 0.4906829595565796, "learning_rate": 4.9512234140587726e-05, "loss": 0.0318, "step": 10560 }, { "grad_norm": 0.42497268319129944, "learning_rate": 4.942956528809477e-05, "loss": 0.0351, "step": 10570 }, { "grad_norm": 0.38242366909980774, "learning_rate": 4.934689799514854e-05, "loss": 0.0326, "step": 10580 }, { "grad_norm": 0.3553409278392792, "learning_rate": 4.926423248775827e-05, "loss": 0.0341, "step": 10590 }, { "grad_norm": 0.35382357239723206, "learning_rate": 4.918156899192826e-05, "loss": 0.035, "step": 10600 }, { "grad_norm": 0.33643949031829834, "learning_rate": 4.909890773365738e-05, "loss": 0.0336, "step": 10610 }, { "grad_norm": 0.34661176800727844, "learning_rate": 4.9016248938938344e-05, "loss": 0.0366, "step": 10620 }, { "grad_norm": 0.3244113326072693, "learning_rate": 4.8933592833757156e-05, "loss": 0.0297, "step": 10630 }, { "grad_norm": 0.3418128192424774, "learning_rate": 4.8850939644092435e-05, "loss": 0.0326, "step": 10640 }, { "grad_norm": 0.3282473683357239, "learning_rate": 4.876828959591485e-05, "loss": 0.0326, "step": 10650 }, { "grad_norm": 0.3417307436466217, "learning_rate": 4.8685642915186474e-05, "loss": 0.0345, "step": 10660 }, { "grad_norm": 0.33395904302597046, "learning_rate": 4.860299982786018e-05, "loss": 0.0335, "step": 10670 }, { "grad_norm": 0.2940755784511566, "learning_rate": 4.852036055987901e-05, "loss": 0.0337, "step": 10680 }, { "grad_norm": 0.3062250018119812, "learning_rate": 4.843772533717558e-05, "loss": 0.0325, "step": 10690 }, { "grad_norm": 0.30112338066101074, "learning_rate": 4.835509438567142e-05, "loss": 0.0317, "step": 10700 }, { "grad_norm": 0.3053971230983734, "learning_rate": 4.827246793127639e-05, "loss": 0.0339, "step": 10710 }, { "grad_norm": 0.3600434958934784, "learning_rate": 4.818984619988807e-05, "loss": 0.0335, "step": 10720 }, { "grad_norm": 0.34694772958755493, "learning_rate": 4.810722941739115e-05, "loss": 0.0326, "step": 10730 }, { "grad_norm": 0.3914850652217865, "learning_rate": 4.8024617809656684e-05, "loss": 0.0341, "step": 10740 }, { "grad_norm": 0.37521252036094666, "learning_rate": 4.794201160254171e-05, "loss": 0.0323, "step": 10750 }, { "grad_norm": 0.40086889266967773, "learning_rate": 4.785941102188844e-05, "loss": 0.0373, "step": 10760 }, { "grad_norm": 0.4394494593143463, "learning_rate": 4.7776816293523686e-05, "loss": 0.032, "step": 10770 }, { "grad_norm": 0.40844082832336426, "learning_rate": 4.769422764325832e-05, "loss": 0.0333, "step": 10780 }, { "grad_norm": 0.35590860247612, "learning_rate": 4.76116452968865e-05, "loss": 0.0332, "step": 10790 }, { "grad_norm": 0.31394702196121216, "learning_rate": 4.752906948018525e-05, "loss": 0.0306, "step": 10800 }, { "grad_norm": 0.26099473237991333, "learning_rate": 4.7446500418913684e-05, "loss": 0.0332, "step": 10810 }, { "grad_norm": 0.31250280141830444, "learning_rate": 4.736393833881247e-05, "loss": 0.0321, "step": 10820 }, { "grad_norm": 0.375406950712204, "learning_rate": 4.7281383465603194e-05, "loss": 0.0316, "step": 10830 }, { "grad_norm": 0.35306090116500854, "learning_rate": 4.71988360249877e-05, "loss": 0.0338, "step": 10840 }, { "grad_norm": 0.41992902755737305, "learning_rate": 4.7116296242647554e-05, "loss": 0.0346, "step": 10850 }, { "grad_norm": 0.42688849568367004, "learning_rate": 4.703376434424336e-05, "loss": 0.033, "step": 10860 }, { "grad_norm": 0.34574684500694275, "learning_rate": 4.695124055541421e-05, "loss": 0.0336, "step": 10870 }, { "grad_norm": 0.36481809616088867, "learning_rate": 4.6868725101776934e-05, "loss": 0.0341, "step": 10880 }, { "grad_norm": 0.38581812381744385, "learning_rate": 4.678621820892567e-05, "loss": 0.0319, "step": 10890 }, { "grad_norm": 0.3097820580005646, "learning_rate": 4.670372010243111e-05, "loss": 0.0304, "step": 10900 }, { "grad_norm": 0.3022705316543579, "learning_rate": 4.662123100783992e-05, "loss": 0.0333, "step": 10910 }, { "grad_norm": 0.34199461340904236, "learning_rate": 4.653875115067415e-05, "loss": 0.0316, "step": 10920 }, { "grad_norm": 0.389080673456192, "learning_rate": 4.6456280756430545e-05, "loss": 0.0327, "step": 10930 }, { "grad_norm": 0.46568605303764343, "learning_rate": 4.637382005058004e-05, "loss": 0.0345, "step": 10940 }, { "grad_norm": 0.37168508768081665, "learning_rate": 4.629136925856705e-05, "loss": 0.0328, "step": 10950 }, { "grad_norm": 0.34711089730262756, "learning_rate": 4.6208928605808895e-05, "loss": 0.0319, "step": 10960 }, { "grad_norm": 0.2555166184902191, "learning_rate": 4.612649831769519e-05, "loss": 0.0344, "step": 10970 }, { "grad_norm": 0.3039989769458771, "learning_rate": 4.604407861958715e-05, "loss": 0.0303, "step": 10980 }, { "grad_norm": 0.308428019285202, "learning_rate": 4.5961669736817114e-05, "loss": 0.0344, "step": 10990 }, { "grad_norm": 0.31173935532569885, "learning_rate": 4.5879271894687814e-05, "loss": 0.0331, "step": 11000 }, { "grad_norm": 0.33829259872436523, "learning_rate": 4.5796885318471826e-05, "loss": 0.0327, "step": 11010 }, { "grad_norm": 0.4762391448020935, "learning_rate": 4.571451023341086e-05, "loss": 0.0333, "step": 11020 }, { "grad_norm": 0.3283219635486603, "learning_rate": 4.563214686471527e-05, "loss": 0.0316, "step": 11030 }, { "grad_norm": 0.3304908573627472, "learning_rate": 4.5549795437563365e-05, "loss": 0.0322, "step": 11040 }, { "grad_norm": 0.28589898347854614, "learning_rate": 4.546745617710081e-05, "loss": 0.0319, "step": 11050 }, { "grad_norm": 0.2485097348690033, "learning_rate": 4.5385129308440014e-05, "loss": 0.0318, "step": 11060 }, { "grad_norm": 0.23899348080158234, "learning_rate": 4.530281505665944e-05, "loss": 0.0337, "step": 11070 }, { "grad_norm": 0.34242603182792664, "learning_rate": 4.5220513646803134e-05, "loss": 0.0325, "step": 11080 }, { "grad_norm": 0.28031793236732483, "learning_rate": 4.513822530388003e-05, "loss": 0.0333, "step": 11090 }, { "grad_norm": 0.3344396650791168, "learning_rate": 4.5055950252863296e-05, "loss": 0.0321, "step": 11100 }, { "grad_norm": 0.31304892897605896, "learning_rate": 4.4973688718689803e-05, "loss": 0.0329, "step": 11110 }, { "grad_norm": 0.24820049107074738, "learning_rate": 4.4891440926259406e-05, "loss": 0.0305, "step": 11120 }, { "grad_norm": 0.3467671573162079, "learning_rate": 4.480920710043443e-05, "loss": 0.03, "step": 11130 }, { "grad_norm": 0.3268544375896454, "learning_rate": 4.4726987466039044e-05, "loss": 0.0326, "step": 11140 }, { "grad_norm": 0.2365504503250122, "learning_rate": 4.46447822478586e-05, "loss": 0.0313, "step": 11150 }, { "grad_norm": 0.29360121488571167, "learning_rate": 4.4562591670638974e-05, "loss": 0.0308, "step": 11160 }, { "grad_norm": 0.32567086815834045, "learning_rate": 4.4480415959086105e-05, "loss": 0.0331, "step": 11170 }, { "grad_norm": 0.3778060972690582, "learning_rate": 4.439825533786522e-05, "loss": 0.0328, "step": 11180 }, { "grad_norm": 0.33359208703041077, "learning_rate": 4.431611003160035e-05, "loss": 0.0331, "step": 11190 }, { "grad_norm": 0.30053460597991943, "learning_rate": 4.4233980264873636e-05, "loss": 0.0335, "step": 11200 }, { "grad_norm": 0.3604303002357483, "learning_rate": 4.4151866262224684e-05, "loss": 0.0357, "step": 11210 }, { "grad_norm": 0.3518749177455902, "learning_rate": 4.406976824815006e-05, "loss": 0.0328, "step": 11220 }, { "grad_norm": 0.3328659236431122, "learning_rate": 4.3987686447102595e-05, "loss": 0.0364, "step": 11230 }, { "grad_norm": 0.2887894809246063, "learning_rate": 4.3905621083490804e-05, "loss": 0.0302, "step": 11240 }, { "grad_norm": 0.41350215673446655, "learning_rate": 4.3823572381678286e-05, "loss": 0.0301, "step": 11250 }, { "grad_norm": 0.2979221045970917, "learning_rate": 4.374154056598301e-05, "loss": 0.0332, "step": 11260 }, { "grad_norm": 0.3856852352619171, "learning_rate": 4.3659525860676845e-05, "loss": 0.0312, "step": 11270 }, { "grad_norm": 0.3440620005130768, "learning_rate": 4.3577528489984854e-05, "loss": 0.0312, "step": 11280 }, { "grad_norm": 0.36620384454727173, "learning_rate": 4.349554867808476e-05, "loss": 0.0314, "step": 11290 }, { "grad_norm": 0.40967389941215515, "learning_rate": 4.34135866491062e-05, "loss": 0.0311, "step": 11300 }, { "grad_norm": 0.39507225155830383, "learning_rate": 4.333164262713022e-05, "loss": 0.0342, "step": 11310 }, { "grad_norm": 0.4715583622455597, "learning_rate": 4.324971683618868e-05, "loss": 0.0326, "step": 11320 }, { "grad_norm": 0.38295483589172363, "learning_rate": 4.316780950026354e-05, "loss": 0.0342, "step": 11330 }, { "grad_norm": 0.42344504594802856, "learning_rate": 4.308592084328637e-05, "loss": 0.0325, "step": 11340 }, { "grad_norm": 0.3771597445011139, "learning_rate": 4.3004051089137576e-05, "loss": 0.0324, "step": 11350 }, { "grad_norm": 0.33110055327415466, "learning_rate": 4.292220046164597e-05, "loss": 0.0336, "step": 11360 }, { "grad_norm": 0.3312799036502838, "learning_rate": 4.2840369184588035e-05, "loss": 0.0303, "step": 11370 }, { "grad_norm": 0.3566407263278961, "learning_rate": 4.2758557481687345e-05, "loss": 0.0328, "step": 11380 }, { "grad_norm": 0.33346277475357056, "learning_rate": 4.267676557661403e-05, "loss": 0.0306, "step": 11390 }, { "grad_norm": 0.35976019501686096, "learning_rate": 4.2594993692983955e-05, "loss": 0.0289, "step": 11400 }, { "grad_norm": 0.44564294815063477, "learning_rate": 4.251324205435837e-05, "loss": 0.0298, "step": 11410 }, { "grad_norm": 0.376081645488739, "learning_rate": 4.243151088424312e-05, "loss": 0.029, "step": 11420 }, { "grad_norm": 0.3556605577468872, "learning_rate": 4.234980040608813e-05, "loss": 0.0313, "step": 11430 }, { "grad_norm": 0.44315528869628906, "learning_rate": 4.22681108432867e-05, "loss": 0.0341, "step": 11440 }, { "grad_norm": 0.32159557938575745, "learning_rate": 4.2186442419174984e-05, "loss": 0.0296, "step": 11450 }, { "grad_norm": 0.34073030948638916, "learning_rate": 4.210479535703133e-05, "loss": 0.0333, "step": 11460 }, { "grad_norm": 0.3008778691291809, "learning_rate": 4.202316988007567e-05, "loss": 0.0326, "step": 11470 }, { "grad_norm": 0.2522810101509094, "learning_rate": 4.194156621146901e-05, "loss": 0.029, "step": 11480 }, { "grad_norm": 0.30322501063346863, "learning_rate": 4.1859984574312596e-05, "loss": 0.0308, "step": 11490 }, { "grad_norm": 0.3263147473335266, "learning_rate": 4.177842519164752e-05, "loss": 0.0305, "step": 11500 }, { "grad_norm": 0.2685795724391937, "learning_rate": 4.169688828645404e-05, "loss": 0.0291, "step": 11510 }, { "grad_norm": 0.3001214563846588, "learning_rate": 4.161537408165092e-05, "loss": 0.0314, "step": 11520 }, { "grad_norm": 0.3336268961429596, "learning_rate": 4.1533882800094924e-05, "loss": 0.0323, "step": 11530 }, { "grad_norm": 0.30662280321121216, "learning_rate": 4.145241466458005e-05, "loss": 0.0294, "step": 11540 }, { "grad_norm": 0.3222416937351227, "learning_rate": 4.13709698978371e-05, "loss": 0.0319, "step": 11550 }, { "grad_norm": 0.32898375391960144, "learning_rate": 4.1289548722532944e-05, "loss": 0.0327, "step": 11560 }, { "grad_norm": 0.37442901730537415, "learning_rate": 4.120815136126999e-05, "loss": 0.0293, "step": 11570 }, { "grad_norm": 0.3069573640823364, "learning_rate": 4.112677803658548e-05, "loss": 0.0352, "step": 11580 }, { "grad_norm": 0.31572937965393066, "learning_rate": 4.1045428970951e-05, "loss": 0.0312, "step": 11590 }, { "grad_norm": 0.40392741560935974, "learning_rate": 4.0964104386771785e-05, "loss": 0.0324, "step": 11600 }, { "grad_norm": 0.2962772846221924, "learning_rate": 4.0882804506386144e-05, "loss": 0.03, "step": 11610 }, { "grad_norm": 0.27624839544296265, "learning_rate": 4.080152955206485e-05, "loss": 0.0312, "step": 11620 }, { "grad_norm": 0.42355725169181824, "learning_rate": 4.0720279746010505e-05, "loss": 0.0324, "step": 11630 }, { "grad_norm": 0.3529970645904541, "learning_rate": 4.063905531035699e-05, "loss": 0.0311, "step": 11640 }, { "grad_norm": 0.38403409719467163, "learning_rate": 4.055785646716882e-05, "loss": 0.0305, "step": 11650 }, { "grad_norm": 0.46179282665252686, "learning_rate": 4.047668343844051e-05, "loss": 0.0307, "step": 11660 }, { "grad_norm": 0.41956236958503723, "learning_rate": 4.039553644609604e-05, "loss": 0.0335, "step": 11670 }, { "grad_norm": 0.5280377268791199, "learning_rate": 4.0314415711988176e-05, "loss": 0.0314, "step": 11680 }, { "grad_norm": 0.40182310342788696, "learning_rate": 4.023332145789792e-05, "loss": 0.0325, "step": 11690 }, { "grad_norm": 0.3790675103664398, "learning_rate": 4.015225390553385e-05, "loss": 0.0314, "step": 11700 }, { "grad_norm": 0.2706011235713959, "learning_rate": 4.007121327653158e-05, "loss": 0.0332, "step": 11710 }, { "grad_norm": 0.3027481436729431, "learning_rate": 3.9990199792453064e-05, "loss": 0.0313, "step": 11720 }, { "grad_norm": 0.30774548649787903, "learning_rate": 3.9909213674786103e-05, "loss": 0.0292, "step": 11730 }, { "grad_norm": 0.369537353515625, "learning_rate": 3.982825514494363e-05, "loss": 0.0343, "step": 11740 }, { "grad_norm": 0.3897591233253479, "learning_rate": 3.974732442426319e-05, "loss": 0.0295, "step": 11750 }, { "grad_norm": 0.33581236004829407, "learning_rate": 3.966642173400629e-05, "loss": 0.0311, "step": 11760 }, { "grad_norm": 0.3117932677268982, "learning_rate": 3.9585547295357764e-05, "loss": 0.0332, "step": 11770 }, { "grad_norm": 0.33020293712615967, "learning_rate": 3.950470132942526e-05, "loss": 0.0354, "step": 11780 }, { "grad_norm": 0.35057035088539124, "learning_rate": 3.942388405723856e-05, "loss": 0.0333, "step": 11790 }, { "grad_norm": 0.2778347134590149, "learning_rate": 3.9343095699749e-05, "loss": 0.0297, "step": 11800 }, { "grad_norm": 0.28069084882736206, "learning_rate": 3.9262336477828874e-05, "loss": 0.0316, "step": 11810 }, { "grad_norm": 0.29877424240112305, "learning_rate": 3.9181606612270794e-05, "loss": 0.0298, "step": 11820 }, { "grad_norm": 0.3295137584209442, "learning_rate": 3.910090632378713e-05, "loss": 0.0351, "step": 11830 }, { "grad_norm": 0.3852570056915283, "learning_rate": 3.90202358330094e-05, "loss": 0.0296, "step": 11840 }, { "grad_norm": 0.30145591497421265, "learning_rate": 3.8939595360487656e-05, "loss": 0.029, "step": 11850 }, { "grad_norm": 0.3299306631088257, "learning_rate": 3.885898512668984e-05, "loss": 0.0307, "step": 11860 }, { "grad_norm": 0.37984395027160645, "learning_rate": 3.877840535200127e-05, "loss": 0.0307, "step": 11870 }, { "grad_norm": 0.2986476421356201, "learning_rate": 3.869785625672397e-05, "loss": 0.0317, "step": 11880 }, { "grad_norm": 0.36006155610084534, "learning_rate": 3.8617338061076094e-05, "loss": 0.0301, "step": 11890 }, { "grad_norm": 0.3553725779056549, "learning_rate": 3.853685098519132e-05, "loss": 0.0305, "step": 11900 }, { "grad_norm": 0.30955687165260315, "learning_rate": 3.845639524911823e-05, "loss": 0.0324, "step": 11910 }, { "grad_norm": 0.33492401242256165, "learning_rate": 3.837597107281974e-05, "loss": 0.0312, "step": 11920 }, { "grad_norm": 0.3539911210536957, "learning_rate": 3.829557867617247e-05, "loss": 0.0329, "step": 11930 }, { "grad_norm": 0.3339459002017975, "learning_rate": 3.821521827896618e-05, "loss": 0.031, "step": 11940 }, { "grad_norm": 0.35425862669944763, "learning_rate": 3.81348901009031e-05, "loss": 0.0318, "step": 11950 }, { "grad_norm": 0.29651182889938354, "learning_rate": 3.805459436159741e-05, "loss": 0.0291, "step": 11960 }, { "grad_norm": 0.312656432390213, "learning_rate": 3.797433128057461e-05, "loss": 0.0295, "step": 11970 }, { "grad_norm": 0.37365925312042236, "learning_rate": 3.789410107727089e-05, "loss": 0.0327, "step": 11980 }, { "grad_norm": 0.3338954448699951, "learning_rate": 3.781390397103257e-05, "loss": 0.0331, "step": 11990 }, { "grad_norm": 0.27225568890571594, "learning_rate": 3.7733740181115455e-05, "loss": 0.0296, "step": 12000 }, { "grad_norm": 0.2729542553424835, "learning_rate": 3.7653609926684306e-05, "loss": 0.0305, "step": 12010 }, { "grad_norm": 0.3292298913002014, "learning_rate": 3.757351342681217e-05, "loss": 0.028, "step": 12020 }, { "grad_norm": 0.2579280138015747, "learning_rate": 3.749345090047982e-05, "loss": 0.0302, "step": 12030 }, { "grad_norm": 0.21126453578472137, "learning_rate": 3.741342256657515e-05, "loss": 0.0294, "step": 12040 }, { "grad_norm": 0.2892661988735199, "learning_rate": 3.7333428643892567e-05, "loss": 0.0291, "step": 12050 }, { "grad_norm": 0.28784769773483276, "learning_rate": 3.725346935113239e-05, "loss": 0.0276, "step": 12060 }, { "grad_norm": 0.3033449649810791, "learning_rate": 3.717354490690029e-05, "loss": 0.0295, "step": 12070 }, { "grad_norm": 0.3606759011745453, "learning_rate": 3.709365552970664e-05, "loss": 0.0293, "step": 12080 }, { "grad_norm": 0.32346218824386597, "learning_rate": 3.7013801437965945e-05, "loss": 0.032, "step": 12090 }, { "grad_norm": 0.29915186762809753, "learning_rate": 3.693398284999623e-05, "loss": 0.0294, "step": 12100 }, { "grad_norm": 0.29532694816589355, "learning_rate": 3.6854199984018484e-05, "loss": 0.0312, "step": 12110 }, { "grad_norm": 0.2685435116291046, "learning_rate": 3.677445305815601e-05, "loss": 0.0275, "step": 12120 }, { "grad_norm": 0.2817489504814148, "learning_rate": 3.669474229043387e-05, "loss": 0.0298, "step": 12130 }, { "grad_norm": 0.2453491985797882, "learning_rate": 3.6615067898778235e-05, "loss": 0.029, "step": 12140 }, { "grad_norm": 0.2464577704668045, "learning_rate": 3.6535430101015866e-05, "loss": 0.0309, "step": 12150 }, { "grad_norm": 0.2532227337360382, "learning_rate": 3.645582911487345e-05, "loss": 0.0328, "step": 12160 }, { "grad_norm": 0.34933650493621826, "learning_rate": 3.637626515797706e-05, "loss": 0.0307, "step": 12170 }, { "grad_norm": 0.29114821553230286, "learning_rate": 3.629673844785152e-05, "loss": 0.0315, "step": 12180 }, { "grad_norm": 0.25326043367385864, "learning_rate": 3.621724920191979e-05, "loss": 0.0286, "step": 12190 }, { "grad_norm": 0.3382149636745453, "learning_rate": 3.6137797637502444e-05, "loss": 0.0303, "step": 12200 }, { "grad_norm": 0.3487905263900757, "learning_rate": 3.6058383971817035e-05, "loss": 0.0328, "step": 12210 }, { "grad_norm": 0.34212473034858704, "learning_rate": 3.59790084219775e-05, "loss": 0.0327, "step": 12220 }, { "grad_norm": 0.3381357491016388, "learning_rate": 3.589967120499353e-05, "loss": 0.0315, "step": 12230 }, { "grad_norm": 0.3345690071582794, "learning_rate": 3.5820372537770075e-05, "loss": 0.0299, "step": 12240 }, { "grad_norm": 0.2755870819091797, "learning_rate": 3.5741112637106655e-05, "loss": 0.0287, "step": 12250 }, { "grad_norm": 0.28571024537086487, "learning_rate": 3.5661891719696804e-05, "loss": 0.0291, "step": 12260 }, { "grad_norm": 0.2644094228744507, "learning_rate": 3.5582710002127504e-05, "loss": 0.0291, "step": 12270 }, { "grad_norm": 0.23365595936775208, "learning_rate": 3.550356770087853e-05, "loss": 0.0288, "step": 12280 }, { "grad_norm": 0.2516094744205475, "learning_rate": 3.5424465032321914e-05, "loss": 0.0294, "step": 12290 }, { "grad_norm": 0.2984798550605774, "learning_rate": 3.5345402212721335e-05, "loss": 0.0294, "step": 12300 }, { "grad_norm": 0.3395840525627136, "learning_rate": 3.526637945823152e-05, "loss": 0.0278, "step": 12310 }, { "grad_norm": 0.3249552845954895, "learning_rate": 3.518739698489767e-05, "loss": 0.0274, "step": 12320 }, { "grad_norm": 0.31955403089523315, "learning_rate": 3.510845500865485e-05, "loss": 0.0293, "step": 12330 }, { "grad_norm": 0.29379382729530334, "learning_rate": 3.502955374532739e-05, "loss": 0.0268, "step": 12340 }, { "grad_norm": 0.35874009132385254, "learning_rate": 3.495069341062836e-05, "loss": 0.0309, "step": 12350 }, { "grad_norm": 0.30443790555000305, "learning_rate": 3.4871874220158896e-05, "loss": 0.0315, "step": 12360 }, { "grad_norm": 0.27383244037628174, "learning_rate": 3.479309638940762e-05, "loss": 0.0277, "step": 12370 }, { "grad_norm": 0.2251000702381134, "learning_rate": 3.4714360133750146e-05, "loss": 0.0284, "step": 12380 }, { "grad_norm": 0.28867101669311523, "learning_rate": 3.463566566844839e-05, "loss": 0.031, "step": 12390 }, { "grad_norm": 0.2438642382621765, "learning_rate": 3.4557013208650016e-05, "loss": 0.0314, "step": 12400 }, { "grad_norm": 0.27053719758987427, "learning_rate": 3.4478402969387857e-05, "loss": 0.0309, "step": 12410 }, { "grad_norm": 0.27533602714538574, "learning_rate": 3.4399835165579266e-05, "loss": 0.0304, "step": 12420 }, { "grad_norm": 0.2492624670267105, "learning_rate": 3.4321310012025645e-05, "loss": 0.0331, "step": 12430 }, { "grad_norm": 0.2905811667442322, "learning_rate": 3.424282772341176e-05, "loss": 0.0298, "step": 12440 }, { "grad_norm": 0.22408539056777954, "learning_rate": 3.416438851430519e-05, "loss": 0.0285, "step": 12450 }, { "grad_norm": 0.20589610934257507, "learning_rate": 3.408599259915577e-05, "loss": 0.0322, "step": 12460 }, { "grad_norm": 0.2222246676683426, "learning_rate": 3.400764019229487e-05, "loss": 0.0312, "step": 12470 }, { "grad_norm": 0.25970837473869324, "learning_rate": 3.3929331507935035e-05, "loss": 0.0289, "step": 12480 }, { "grad_norm": 0.27809834480285645, "learning_rate": 3.3851066760169196e-05, "loss": 0.029, "step": 12490 }, { "grad_norm": 0.25369513034820557, "learning_rate": 3.377284616297021e-05, "loss": 0.0312, "step": 12500 }, { "grad_norm": 0.2690669596195221, "learning_rate": 3.3694669930190166e-05, "loss": 0.0289, "step": 12510 }, { "grad_norm": 0.2658199667930603, "learning_rate": 3.36165382755599e-05, "loss": 0.0309, "step": 12520 }, { "grad_norm": 0.26479092240333557, "learning_rate": 3.35384514126884e-05, "loss": 0.0301, "step": 12530 }, { "grad_norm": 0.27861544489860535, "learning_rate": 3.3460409555062154e-05, "loss": 0.0293, "step": 12540 }, { "grad_norm": 0.29851022362709045, "learning_rate": 3.3382412916044645e-05, "loss": 0.0309, "step": 12550 }, { "grad_norm": 0.29198166728019714, "learning_rate": 3.330446170887566e-05, "loss": 0.0299, "step": 12560 }, { "grad_norm": 0.23903554677963257, "learning_rate": 3.3226556146670834e-05, "loss": 0.0293, "step": 12570 }, { "grad_norm": 0.2771153450012207, "learning_rate": 3.314869644242102e-05, "loss": 0.0291, "step": 12580 }, { "grad_norm": 0.26589226722717285, "learning_rate": 3.3070882808991674e-05, "loss": 0.0291, "step": 12590 }, { "grad_norm": 0.25851404666900635, "learning_rate": 3.2993115459122305e-05, "loss": 0.0295, "step": 12600 }, { "grad_norm": 0.30803415179252625, "learning_rate": 3.2915394605425835e-05, "loss": 0.0294, "step": 12610 }, { "grad_norm": 0.30770590901374817, "learning_rate": 3.283772046038816e-05, "loss": 0.0283, "step": 12620 }, { "grad_norm": 0.27101626992225647, "learning_rate": 3.276009323636739e-05, "loss": 0.0297, "step": 12630 }, { "grad_norm": 0.33805960416793823, "learning_rate": 3.268251314559344e-05, "loss": 0.0288, "step": 12640 }, { "grad_norm": 0.3255416452884674, "learning_rate": 3.2604980400167254e-05, "loss": 0.0278, "step": 12650 }, { "grad_norm": 0.3049745261669159, "learning_rate": 3.252749521206042e-05, "loss": 0.0295, "step": 12660 }, { "grad_norm": 0.3390018343925476, "learning_rate": 3.2450057793114494e-05, "loss": 0.0297, "step": 12670 }, { "grad_norm": 0.2769322693347931, "learning_rate": 3.2372668355040435e-05, "loss": 0.0301, "step": 12680 }, { "grad_norm": 0.2684820294380188, "learning_rate": 3.2295327109418005e-05, "loss": 0.0281, "step": 12690 }, { "grad_norm": 0.2523753345012665, "learning_rate": 3.221803426769518e-05, "loss": 0.0267, "step": 12700 }, { "grad_norm": 0.23303191363811493, "learning_rate": 3.214079004118768e-05, "loss": 0.0315, "step": 12710 }, { "grad_norm": 0.2656311094760895, "learning_rate": 3.2063594641078234e-05, "loss": 0.0277, "step": 12720 }, { "grad_norm": 0.35646116733551025, "learning_rate": 3.198644827841616e-05, "loss": 0.0298, "step": 12730 }, { "grad_norm": 0.3525397479534149, "learning_rate": 3.1909351164116654e-05, "loss": 0.0317, "step": 12740 }, { "grad_norm": 0.4316791296005249, "learning_rate": 3.183230350896026e-05, "loss": 0.0323, "step": 12750 }, { "grad_norm": 0.3111700713634491, "learning_rate": 3.1755305523592337e-05, "loss": 0.03, "step": 12760 }, { "grad_norm": 0.34381139278411865, "learning_rate": 3.167835741852245e-05, "loss": 0.0308, "step": 12770 }, { "grad_norm": 0.29795366525650024, "learning_rate": 3.160145940412378e-05, "loss": 0.03, "step": 12780 }, { "grad_norm": 0.3273204565048218, "learning_rate": 3.1524611690632545e-05, "loss": 0.0293, "step": 12790 }, { "grad_norm": 0.3016532361507416, "learning_rate": 3.144781448814746e-05, "loss": 0.0322, "step": 12800 }, { "grad_norm": 0.27190399169921875, "learning_rate": 3.1371068006629145e-05, "loss": 0.0294, "step": 12810 }, { "grad_norm": 0.2823968231678009, "learning_rate": 3.129437245589956e-05, "loss": 0.0278, "step": 12820 }, { "grad_norm": 0.2936570644378662, "learning_rate": 3.121772804564143e-05, "loss": 0.0266, "step": 12830 }, { "grad_norm": 0.2808833420276642, "learning_rate": 3.11411349853976e-05, "loss": 0.0292, "step": 12840 }, { "grad_norm": 0.20888058841228485, "learning_rate": 3.10645934845706e-05, "loss": 0.03, "step": 12850 }, { "grad_norm": 0.24918581545352936, "learning_rate": 3.098810375242196e-05, "loss": 0.0298, "step": 12860 }, { "grad_norm": 0.33017584681510925, "learning_rate": 3.0911665998071704e-05, "loss": 0.0279, "step": 12870 }, { "grad_norm": 0.3024176061153412, "learning_rate": 3.083528043049774e-05, "loss": 0.0282, "step": 12880 }, { "grad_norm": 0.3263440430164337, "learning_rate": 3.0758947258535255e-05, "loss": 0.0292, "step": 12890 }, { "grad_norm": 0.27665892243385315, "learning_rate": 3.068266669087625e-05, "loss": 0.0319, "step": 12900 }, { "grad_norm": 0.33148685097694397, "learning_rate": 3.060643893606887e-05, "loss": 0.0276, "step": 12910 }, { "grad_norm": 0.2590135633945465, "learning_rate": 3.053026420251693e-05, "loss": 0.0291, "step": 12920 }, { "grad_norm": 0.2351110279560089, "learning_rate": 3.0454142698479183e-05, "loss": 0.0278, "step": 12930 }, { "grad_norm": 0.2428625226020813, "learning_rate": 3.0378074632068954e-05, "loss": 0.0286, "step": 12940 }, { "grad_norm": 0.22445577383041382, "learning_rate": 3.0302060211253408e-05, "loss": 0.0313, "step": 12950 }, { "grad_norm": 0.2502662241458893, "learning_rate": 3.0226099643853073e-05, "loss": 0.0262, "step": 12960 }, { "grad_norm": 0.31453508138656616, "learning_rate": 3.0150193137541283e-05, "loss": 0.0297, "step": 12970 }, { "grad_norm": 0.33120566606521606, "learning_rate": 3.0074340899843467e-05, "loss": 0.0272, "step": 12980 }, { "grad_norm": 0.2509555518627167, "learning_rate": 2.999854313813677e-05, "loss": 0.0306, "step": 12990 }, { "grad_norm": 0.260007381439209, "learning_rate": 2.9922800059649382e-05, "loss": 0.0304, "step": 13000 }, { "grad_norm": 0.28036147356033325, "learning_rate": 2.9847111871459976e-05, "loss": 0.0283, "step": 13010 }, { "grad_norm": 0.2666558027267456, "learning_rate": 2.977147878049721e-05, "loss": 0.0286, "step": 13020 }, { "grad_norm": 0.2597411572933197, "learning_rate": 2.9695900993539006e-05, "loss": 0.0304, "step": 13030 }, { "grad_norm": 0.23704904317855835, "learning_rate": 2.9620378717212183e-05, "loss": 0.0273, "step": 13040 }, { "grad_norm": 0.2792782485485077, "learning_rate": 2.9544912157991745e-05, "loss": 0.0254, "step": 13050 }, { "grad_norm": 0.30436092615127563, "learning_rate": 2.9469501522200405e-05, "loss": 0.032, "step": 13060 }, { "grad_norm": 0.2784832715988159, "learning_rate": 2.9394147016007946e-05, "loss": 0.026, "step": 13070 }, { "grad_norm": 0.26810792088508606, "learning_rate": 2.9318848845430702e-05, "loss": 0.0278, "step": 13080 }, { "grad_norm": 0.24108867347240448, "learning_rate": 2.9243607216331013e-05, "loss": 0.0271, "step": 13090 }, { "grad_norm": 0.25536903738975525, "learning_rate": 2.916842233441661e-05, "loss": 0.0283, "step": 13100 }, { "grad_norm": 0.25954097509384155, "learning_rate": 2.90932944052401e-05, "loss": 0.0275, "step": 13110 }, { "grad_norm": 0.2944454848766327, "learning_rate": 2.9018223634198354e-05, "loss": 0.0293, "step": 13120 }, { "grad_norm": 0.27303725481033325, "learning_rate": 2.8943210226532025e-05, "loss": 0.0309, "step": 13130 }, { "grad_norm": 0.2559761703014374, "learning_rate": 2.8868254387324857e-05, "loss": 0.0291, "step": 13140 }, { "grad_norm": 0.2437514364719391, "learning_rate": 2.8793356321503306e-05, "loss": 0.026, "step": 13150 }, { "grad_norm": 0.24898093938827515, "learning_rate": 2.87185162338358e-05, "loss": 0.0324, "step": 13160 }, { "grad_norm": 0.31701815128326416, "learning_rate": 2.8643734328932253e-05, "loss": 0.0319, "step": 13170 }, { "grad_norm": 0.33045822381973267, "learning_rate": 2.856901081124359e-05, "loss": 0.0278, "step": 13180 }, { "grad_norm": 0.31905725598335266, "learning_rate": 2.8494345885061002e-05, "loss": 0.0279, "step": 13190 }, { "grad_norm": 0.34471267461776733, "learning_rate": 2.8419739754515616e-05, "loss": 0.0332, "step": 13200 }, { "grad_norm": 0.33136820793151855, "learning_rate": 2.8345192623577666e-05, "loss": 0.0261, "step": 13210 }, { "grad_norm": 0.2150506228208542, "learning_rate": 2.8270704696056193e-05, "loss": 0.0289, "step": 13220 }, { "grad_norm": 0.2412453293800354, "learning_rate": 2.8196276175598367e-05, "loss": 0.0265, "step": 13230 }, { "grad_norm": 0.2808520793914795, "learning_rate": 2.8121907265688884e-05, "loss": 0.03, "step": 13240 }, { "grad_norm": 0.30650612711906433, "learning_rate": 2.804759816964957e-05, "loss": 0.0288, "step": 13250 }, { "grad_norm": 0.3092111349105835, "learning_rate": 2.797334909063857e-05, "loss": 0.0288, "step": 13260 }, { "grad_norm": 0.3186614513397217, "learning_rate": 2.7899160231650056e-05, "loss": 0.0286, "step": 13270 }, { "grad_norm": 0.30622637271881104, "learning_rate": 2.7825031795513585e-05, "loss": 0.0289, "step": 13280 }, { "grad_norm": 0.2965271472930908, "learning_rate": 2.775096398489341e-05, "loss": 0.0318, "step": 13290 }, { "grad_norm": 0.33050239086151123, "learning_rate": 2.7676957002288163e-05, "loss": 0.0292, "step": 13300 }, { "grad_norm": 0.32535937428474426, "learning_rate": 2.760301105003003e-05, "loss": 0.0268, "step": 13310 }, { "grad_norm": 0.28698915243148804, "learning_rate": 2.752912633028446e-05, "loss": 0.0302, "step": 13320 }, { "grad_norm": 0.2791098654270172, "learning_rate": 2.7455303045049474e-05, "loss": 0.0288, "step": 13330 }, { "grad_norm": 0.2642954885959625, "learning_rate": 2.7381541396155098e-05, "loss": 0.0295, "step": 13340 }, { "grad_norm": 0.28750064969062805, "learning_rate": 2.730784158526286e-05, "loss": 0.0292, "step": 13350 }, { "grad_norm": 0.32892757654190063, "learning_rate": 2.723420381386521e-05, "loss": 0.0289, "step": 13360 }, { "grad_norm": 0.26780834794044495, "learning_rate": 2.7160628283285018e-05, "loss": 0.03, "step": 13370 }, { "grad_norm": 0.28775331377983093, "learning_rate": 2.7087115194675007e-05, "loss": 0.0287, "step": 13380 }, { "grad_norm": 0.25106871128082275, "learning_rate": 2.701366474901712e-05, "loss": 0.027, "step": 13390 }, { "grad_norm": 0.25987139344215393, "learning_rate": 2.6940277147122085e-05, "loss": 0.0315, "step": 13400 }, { "grad_norm": 0.2250460386276245, "learning_rate": 2.686695258962878e-05, "loss": 0.0264, "step": 13410 }, { "grad_norm": 0.2559645473957062, "learning_rate": 2.679369127700375e-05, "loss": 0.0269, "step": 13420 }, { "grad_norm": 0.19469673931598663, "learning_rate": 2.672049340954067e-05, "loss": 0.0296, "step": 13430 }, { "grad_norm": 0.2675566077232361, "learning_rate": 2.6647359187359676e-05, "loss": 0.0277, "step": 13440 }, { "grad_norm": 0.30726444721221924, "learning_rate": 2.6574288810406946e-05, "loss": 0.028, "step": 13450 }, { "grad_norm": 0.30544790625572205, "learning_rate": 2.6501282478454083e-05, "loss": 0.0266, "step": 13460 }, { "grad_norm": 0.27595752477645874, "learning_rate": 2.6428340391097618e-05, "loss": 0.0271, "step": 13470 }, { "grad_norm": 0.2430611401796341, "learning_rate": 2.6355462747758485e-05, "loss": 0.0269, "step": 13480 }, { "grad_norm": 0.27391859889030457, "learning_rate": 2.6282649747681304e-05, "loss": 0.0265, "step": 13490 }, { "grad_norm": 0.30007538199424744, "learning_rate": 2.620990158993406e-05, "loss": 0.0301, "step": 13500 }, { "grad_norm": 0.2463942915201187, "learning_rate": 2.6137218473407477e-05, "loss": 0.0284, "step": 13510 }, { "grad_norm": 0.29149022698402405, "learning_rate": 2.606460059681436e-05, "loss": 0.03, "step": 13520 }, { "grad_norm": 0.21014155447483063, "learning_rate": 2.599204815868928e-05, "loss": 0.0289, "step": 13530 }, { "grad_norm": 0.22925542294979095, "learning_rate": 2.5919561357387756e-05, "loss": 0.0311, "step": 13540 }, { "grad_norm": 0.1943470686674118, "learning_rate": 2.5847140391085972e-05, "loss": 0.0258, "step": 13550 }, { "grad_norm": 0.2077137529850006, "learning_rate": 2.5774785457780103e-05, "loss": 0.0262, "step": 13560 }, { "grad_norm": 0.236194908618927, "learning_rate": 2.5702496755285753e-05, "loss": 0.0262, "step": 13570 }, { "grad_norm": 0.23975755274295807, "learning_rate": 2.5630274481237483e-05, "loss": 0.0254, "step": 13580 }, { "grad_norm": 0.3014545738697052, "learning_rate": 2.5558118833088197e-05, "loss": 0.029, "step": 13590 }, { "grad_norm": 0.2689688801765442, "learning_rate": 2.548603000810872e-05, "loss": 0.0268, "step": 13600 }, { "grad_norm": 0.23730401694774628, "learning_rate": 2.5414008203387152e-05, "loss": 0.0295, "step": 13610 }, { "grad_norm": 0.18164879083633423, "learning_rate": 2.534205361582834e-05, "loss": 0.0255, "step": 13620 }, { "grad_norm": 0.2477237433195114, "learning_rate": 2.527016644215338e-05, "loss": 0.029, "step": 13630 }, { "grad_norm": 0.21201635897159576, "learning_rate": 2.519834687889905e-05, "loss": 0.0256, "step": 13640 }, { "grad_norm": 0.21268650889396667, "learning_rate": 2.5126595122417295e-05, "loss": 0.028, "step": 13650 }, { "grad_norm": 0.24123318493366241, "learning_rate": 2.5054911368874713e-05, "loss": 0.0297, "step": 13660 }, { "grad_norm": 0.267495721578598, "learning_rate": 2.4983295814251916e-05, "loss": 0.0283, "step": 13670 }, { "grad_norm": 0.27470314502716064, "learning_rate": 2.4911748654343105e-05, "loss": 0.0264, "step": 13680 }, { "grad_norm": 0.3179301917552948, "learning_rate": 2.4840270084755463e-05, "loss": 0.0269, "step": 13690 }, { "grad_norm": 0.23449891805648804, "learning_rate": 2.4768860300908685e-05, "loss": 0.0264, "step": 13700 }, { "grad_norm": 0.20081140100955963, "learning_rate": 2.469751949803443e-05, "loss": 0.0264, "step": 13710 }, { "grad_norm": 0.23352238535881042, "learning_rate": 2.4626247871175666e-05, "loss": 0.0289, "step": 13720 }, { "grad_norm": 0.20830415189266205, "learning_rate": 2.4555045615186346e-05, "loss": 0.0303, "step": 13730 }, { "grad_norm": 0.24995477497577667, "learning_rate": 2.4483912924730677e-05, "loss": 0.0268, "step": 13740 }, { "grad_norm": 0.21795034408569336, "learning_rate": 2.4412849994282742e-05, "loss": 0.026, "step": 13750 }, { "grad_norm": 0.2797374725341797, "learning_rate": 2.434185701812592e-05, "loss": 0.0306, "step": 13760 }, { "grad_norm": 0.24893301725387573, "learning_rate": 2.4270934190352218e-05, "loss": 0.0267, "step": 13770 }, { "grad_norm": 0.3009544014930725, "learning_rate": 2.4200081704861998e-05, "loss": 0.0283, "step": 13780 }, { "grad_norm": 0.2575647234916687, "learning_rate": 2.412929975536321e-05, "loss": 0.0258, "step": 13790 }, { "grad_norm": 0.2058427482843399, "learning_rate": 2.4058588535371017e-05, "loss": 0.0277, "step": 13800 }, { "grad_norm": 0.26451143622398376, "learning_rate": 2.3987948238207243e-05, "loss": 0.0274, "step": 13810 }, { "grad_norm": 0.21795976161956787, "learning_rate": 2.3917379056999678e-05, "loss": 0.0283, "step": 13820 }, { "grad_norm": 0.2307184785604477, "learning_rate": 2.3846881184681824e-05, "loss": 0.0293, "step": 13830 }, { "grad_norm": 0.2368544340133667, "learning_rate": 2.377645481399214e-05, "loss": 0.0275, "step": 13840 }, { "grad_norm": 0.2084178328514099, "learning_rate": 2.3706100137473667e-05, "loss": 0.0279, "step": 13850 }, { "grad_norm": 0.22903449833393097, "learning_rate": 2.3635817347473394e-05, "loss": 0.0322, "step": 13860 }, { "grad_norm": 0.2475515604019165, "learning_rate": 2.3565606636141757e-05, "loss": 0.0282, "step": 13870 }, { "grad_norm": 0.21406354010105133, "learning_rate": 2.3495468195432203e-05, "loss": 0.028, "step": 13880 }, { "grad_norm": 0.23272515833377838, "learning_rate": 2.3425402217100507e-05, "loss": 0.0279, "step": 13890 }, { "grad_norm": 0.23804205656051636, "learning_rate": 2.3355408892704424e-05, "loss": 0.0304, "step": 13900 }, { "grad_norm": 0.23774747550487518, "learning_rate": 2.3285488413603003e-05, "loss": 0.0278, "step": 13910 }, { "grad_norm": 0.26178932189941406, "learning_rate": 2.321564097095615e-05, "loss": 0.0271, "step": 13920 }, { "grad_norm": 0.2691616714000702, "learning_rate": 2.3145866755724142e-05, "loss": 0.0268, "step": 13930 }, { "grad_norm": 0.2186966836452484, "learning_rate": 2.307616595866699e-05, "loss": 0.029, "step": 13940 }, { "grad_norm": 0.20739421248435974, "learning_rate": 2.3006538770344032e-05, "loss": 0.0295, "step": 13950 }, { "grad_norm": 0.22641143202781677, "learning_rate": 2.293698538111334e-05, "loss": 0.0304, "step": 13960 }, { "grad_norm": 0.23681509494781494, "learning_rate": 2.28675059811312e-05, "loss": 0.0252, "step": 13970 }, { "grad_norm": 0.2549561858177185, "learning_rate": 2.279810076035167e-05, "loss": 0.0285, "step": 13980 }, { "grad_norm": 0.20355308055877686, "learning_rate": 2.272876990852596e-05, "loss": 0.0252, "step": 13990 }, { "grad_norm": 0.2341877520084381, "learning_rate": 2.265951361520195e-05, "loss": 0.0308, "step": 14000 }, { "grad_norm": 0.2410147488117218, "learning_rate": 2.2590332069723748e-05, "loss": 0.0279, "step": 14010 }, { "grad_norm": 0.22342273592948914, "learning_rate": 2.2521225461231004e-05, "loss": 0.0284, "step": 14020 }, { "grad_norm": 0.2040673792362213, "learning_rate": 2.2452193978658597e-05, "loss": 0.0284, "step": 14030 }, { "grad_norm": 0.2338394671678543, "learning_rate": 2.238323781073594e-05, "loss": 0.0278, "step": 14040 }, { "grad_norm": 0.25011929869651794, "learning_rate": 2.2314357145986552e-05, "loss": 0.0293, "step": 14050 }, { "grad_norm": 0.25296780467033386, "learning_rate": 2.224555217272757e-05, "loss": 0.0293, "step": 14060 }, { "grad_norm": 0.24030199646949768, "learning_rate": 2.2176823079069127e-05, "loss": 0.0291, "step": 14070 }, { "grad_norm": 0.23051589727401733, "learning_rate": 2.210817005291398e-05, "loss": 0.0284, "step": 14080 }, { "grad_norm": 0.2510325014591217, "learning_rate": 2.203959328195686e-05, "loss": 0.028, "step": 14090 }, { "grad_norm": 0.21724915504455566, "learning_rate": 2.1971092953684026e-05, "loss": 0.0244, "step": 14100 }, { "grad_norm": 0.2615448236465454, "learning_rate": 2.1902669255372788e-05, "loss": 0.0256, "step": 14110 }, { "grad_norm": 0.18911784887313843, "learning_rate": 2.1834322374090897e-05, "loss": 0.0252, "step": 14120 }, { "grad_norm": 0.20625148713588715, "learning_rate": 2.1766052496696153e-05, "loss": 0.0304, "step": 14130 }, { "grad_norm": 0.23142102360725403, "learning_rate": 2.169785980983577e-05, "loss": 0.0272, "step": 14140 }, { "grad_norm": 0.23305609822273254, "learning_rate": 2.162974449994593e-05, "loss": 0.0277, "step": 14150 }, { "grad_norm": 0.2625384032726288, "learning_rate": 2.1561706753251337e-05, "loss": 0.0284, "step": 14160 }, { "grad_norm": 0.24369961023330688, "learning_rate": 2.1493746755764544e-05, "loss": 0.028, "step": 14170 }, { "grad_norm": 0.224081888794899, "learning_rate": 2.1425864693285635e-05, "loss": 0.0259, "step": 14180 }, { "grad_norm": 0.2389301359653473, "learning_rate": 2.1358060751401547e-05, "loss": 0.0266, "step": 14190 }, { "grad_norm": 0.21444953978061676, "learning_rate": 2.129033511548566e-05, "loss": 0.0262, "step": 14200 }, { "grad_norm": 0.2485407292842865, "learning_rate": 2.1222687970697315e-05, "loss": 0.026, "step": 14210 }, { "grad_norm": 0.1933886706829071, "learning_rate": 2.1155119501981173e-05, "loss": 0.0247, "step": 14220 }, { "grad_norm": 0.19920894503593445, "learning_rate": 2.1087629894066895e-05, "loss": 0.0252, "step": 14230 }, { "grad_norm": 0.20729900896549225, "learning_rate": 2.1020219331468473e-05, "loss": 0.0267, "step": 14240 }, { "grad_norm": 0.22705607116222382, "learning_rate": 2.095288799848379e-05, "loss": 0.027, "step": 14250 }, { "grad_norm": 0.21499715745449066, "learning_rate": 2.088563607919417e-05, "loss": 0.0252, "step": 14260 }, { "grad_norm": 0.18663831055164337, "learning_rate": 2.0818463757463786e-05, "loss": 0.0252, "step": 14270 }, { "grad_norm": 0.20180989801883698, "learning_rate": 2.0751371216939175e-05, "loss": 0.026, "step": 14280 }, { "grad_norm": 0.2210698276758194, "learning_rate": 2.068435864104882e-05, "loss": 0.0284, "step": 14290 }, { "grad_norm": 0.19439390301704407, "learning_rate": 2.0617426213002506e-05, "loss": 0.0254, "step": 14300 }, { "grad_norm": 0.22832122445106506, "learning_rate": 2.055057411579097e-05, "loss": 0.0258, "step": 14310 }, { "grad_norm": 0.23682066798210144, "learning_rate": 2.0483802532185286e-05, "loss": 0.0274, "step": 14320 }, { "grad_norm": 0.2482992261648178, "learning_rate": 2.041711164473638e-05, "loss": 0.026, "step": 14330 }, { "grad_norm": 0.2589288651943207, "learning_rate": 2.0350501635774637e-05, "loss": 0.0273, "step": 14340 }, { "grad_norm": 0.30175599455833435, "learning_rate": 2.0283972687409247e-05, "loss": 0.0274, "step": 14350 }, { "grad_norm": 0.23242254555225372, "learning_rate": 2.021752498152784e-05, "loss": 0.0255, "step": 14360 }, { "grad_norm": 0.20487669110298157, "learning_rate": 2.015115869979589e-05, "loss": 0.0255, "step": 14370 }, { "grad_norm": 0.19777999818325043, "learning_rate": 2.0084874023656265e-05, "loss": 0.0307, "step": 14380 }, { "grad_norm": 0.2506314516067505, "learning_rate": 2.001867113432877e-05, "loss": 0.0278, "step": 14390 }, { "grad_norm": 0.21752817928791046, "learning_rate": 1.995255021280954e-05, "loss": 0.0245, "step": 14400 }, { "grad_norm": 0.2024872750043869, "learning_rate": 1.9886511439870688e-05, "loss": 0.0293, "step": 14410 }, { "grad_norm": 0.2034156769514084, "learning_rate": 1.9820554996059675e-05, "loss": 0.0272, "step": 14420 }, { "grad_norm": 0.225850909948349, "learning_rate": 1.9754681061698893e-05, "loss": 0.028, "step": 14430 }, { "grad_norm": 0.1910875141620636, "learning_rate": 1.9688889816885185e-05, "loss": 0.0301, "step": 14440 }, { "grad_norm": 0.24020196497440338, "learning_rate": 1.962318144148928e-05, "loss": 0.0266, "step": 14450 }, { "grad_norm": 0.18492121994495392, "learning_rate": 1.955755611515539e-05, "loss": 0.028, "step": 14460 }, { "grad_norm": 0.2343754917383194, "learning_rate": 1.9492014017300642e-05, "loss": 0.0269, "step": 14470 }, { "grad_norm": 0.23466475307941437, "learning_rate": 1.942655532711461e-05, "loss": 0.0283, "step": 14480 }, { "grad_norm": 0.25688883662223816, "learning_rate": 1.9361180223558882e-05, "loss": 0.0271, "step": 14490 }, { "grad_norm": 0.24565629661083221, "learning_rate": 1.929588888536647e-05, "loss": 0.0261, "step": 14500 }, { "grad_norm": 0.17586012184619904, "learning_rate": 1.9230681491041425e-05, "loss": 0.0272, "step": 14510 }, { "grad_norm": 0.23417030274868011, "learning_rate": 1.9165558218858264e-05, "loss": 0.0266, "step": 14520 }, { "grad_norm": 0.1913340538740158, "learning_rate": 1.9100519246861505e-05, "loss": 0.0294, "step": 14530 }, { "grad_norm": 0.18265140056610107, "learning_rate": 1.9035564752865248e-05, "loss": 0.026, "step": 14540 }, { "grad_norm": 0.1807863861322403, "learning_rate": 1.897069491445258e-05, "loss": 0.0257, "step": 14550 }, { "grad_norm": 0.17120328545570374, "learning_rate": 1.890590990897515e-05, "loss": 0.0279, "step": 14560 }, { "grad_norm": 0.17552773654460907, "learning_rate": 1.884120991355272e-05, "loss": 0.0263, "step": 14570 }, { "grad_norm": 0.18269352614879608, "learning_rate": 1.8776595105072576e-05, "loss": 0.0288, "step": 14580 }, { "grad_norm": 0.18173980712890625, "learning_rate": 1.8712065660189166e-05, "loss": 0.0274, "step": 14590 }, { "grad_norm": 0.16327178478240967, "learning_rate": 1.8647621755323513e-05, "loss": 0.0232, "step": 14600 }, { "grad_norm": 0.20019380748271942, "learning_rate": 1.858326356666278e-05, "loss": 0.0254, "step": 14610 }, { "grad_norm": 0.18608596920967102, "learning_rate": 1.851899127015983e-05, "loss": 0.0266, "step": 14620 }, { "grad_norm": 0.19959886372089386, "learning_rate": 1.8454805041532626e-05, "loss": 0.0255, "step": 14630 }, { "grad_norm": 0.1950630247592926, "learning_rate": 1.8390705056263906e-05, "loss": 0.0268, "step": 14640 }, { "grad_norm": 0.18073537945747375, "learning_rate": 1.832669148960057e-05, "loss": 0.0259, "step": 14650 }, { "grad_norm": 0.22221310436725616, "learning_rate": 1.8262764516553233e-05, "loss": 0.0278, "step": 14660 }, { "grad_norm": 0.20623879134655, "learning_rate": 1.8198924311895843e-05, "loss": 0.0273, "step": 14670 }, { "grad_norm": 0.23011673986911774, "learning_rate": 1.813517105016505e-05, "loss": 0.0293, "step": 14680 }, { "grad_norm": 0.18137459456920624, "learning_rate": 1.8071504905659888e-05, "loss": 0.0253, "step": 14690 }, { "grad_norm": 0.1807180643081665, "learning_rate": 1.800792605244109e-05, "loss": 0.026, "step": 14700 }, { "grad_norm": 0.17482785880565643, "learning_rate": 1.7944434664330844e-05, "loss": 0.0273, "step": 14710 }, { "grad_norm": 0.15572845935821533, "learning_rate": 1.7881030914912212e-05, "loss": 0.0268, "step": 14720 }, { "grad_norm": 0.20199055969715118, "learning_rate": 1.7817714977528577e-05, "loss": 0.027, "step": 14730 }, { "grad_norm": 0.1913933902978897, "learning_rate": 1.7754487025283332e-05, "loss": 0.026, "step": 14740 }, { "grad_norm": 0.21558964252471924, "learning_rate": 1.7691347231039275e-05, "loss": 0.0268, "step": 14750 }, { "grad_norm": 0.2050299495458603, "learning_rate": 1.7628295767418164e-05, "loss": 0.0275, "step": 14760 }, { "grad_norm": 0.18185485899448395, "learning_rate": 1.7565332806800333e-05, "loss": 0.0266, "step": 14770 }, { "grad_norm": 0.19919078052043915, "learning_rate": 1.750245852132408e-05, "loss": 0.0245, "step": 14780 }, { "grad_norm": 0.187032088637352, "learning_rate": 1.7439673082885323e-05, "loss": 0.028, "step": 14790 }, { "grad_norm": 0.18795423209667206, "learning_rate": 1.7376976663137047e-05, "loss": 0.0267, "step": 14800 }, { "grad_norm": 0.221416175365448, "learning_rate": 1.7314369433488853e-05, "loss": 0.0264, "step": 14810 }, { "grad_norm": 0.17592893540859222, "learning_rate": 1.7251851565106548e-05, "loss": 0.0278, "step": 14820 }, { "grad_norm": 0.1839088499546051, "learning_rate": 1.7189423228911574e-05, "loss": 0.0252, "step": 14830 }, { "grad_norm": 0.17952990531921387, "learning_rate": 1.7127084595580606e-05, "loss": 0.0253, "step": 14840 }, { "grad_norm": 0.20094038546085358, "learning_rate": 1.706483583554513e-05, "loss": 0.0256, "step": 14850 }, { "grad_norm": 0.25491389632225037, "learning_rate": 1.700267711899083e-05, "loss": 0.0258, "step": 14860 }, { "grad_norm": 0.2540971040725708, "learning_rate": 1.69406086158573e-05, "loss": 0.0255, "step": 14870 }, { "grad_norm": 0.20938199758529663, "learning_rate": 1.6878630495837455e-05, "loss": 0.0276, "step": 14880 }, { "grad_norm": 0.25789573788642883, "learning_rate": 1.681674292837707e-05, "loss": 0.0282, "step": 14890 }, { "grad_norm": 0.21144695580005646, "learning_rate": 1.6754946082674444e-05, "loss": 0.0286, "step": 14900 }, { "grad_norm": 0.21453383564949036, "learning_rate": 1.6693240127679748e-05, "loss": 0.0258, "step": 14910 }, { "grad_norm": 0.20267248153686523, "learning_rate": 1.663162523209475e-05, "loss": 0.0253, "step": 14920 }, { "grad_norm": 0.17632077634334564, "learning_rate": 1.6570101564372193e-05, "loss": 0.0274, "step": 14930 }, { "grad_norm": 0.18181011080741882, "learning_rate": 1.650866929271543e-05, "loss": 0.0259, "step": 14940 }, { "grad_norm": 0.149251788854599, "learning_rate": 1.644732858507797e-05, "loss": 0.0277, "step": 14950 }, { "grad_norm": 0.1831086426973343, "learning_rate": 1.6386079609162943e-05, "loss": 0.0274, "step": 14960 }, { "grad_norm": 0.19382578134536743, "learning_rate": 1.6324922532422742e-05, "loss": 0.027, "step": 14970 }, { "grad_norm": 0.18145793676376343, "learning_rate": 1.6263857522058434e-05, "loss": 0.0248, "step": 14980 }, { "grad_norm": 0.16226986050605774, "learning_rate": 1.6202884745019443e-05, "loss": 0.026, "step": 14990 }, { "grad_norm": 0.17216363549232483, "learning_rate": 1.614200436800304e-05, "loss": 0.0236, "step": 15000 }, { "grad_norm": 0.1891501098871231, "learning_rate": 1.6081216557453814e-05, "loss": 0.0277, "step": 15010 }, { "grad_norm": 0.17754697799682617, "learning_rate": 1.6020521479563367e-05, "loss": 0.0268, "step": 15020 }, { "grad_norm": 0.16175393760204315, "learning_rate": 1.5959919300269654e-05, "loss": 0.0263, "step": 15030 }, { "grad_norm": 0.16272148489952087, "learning_rate": 1.5899410185256764e-05, "loss": 0.0269, "step": 15040 }, { "grad_norm": 0.18611407279968262, "learning_rate": 1.583899429995431e-05, "loss": 0.0271, "step": 15050 }, { "grad_norm": 0.1717323213815689, "learning_rate": 1.5778671809536993e-05, "loss": 0.0255, "step": 15060 }, { "grad_norm": 0.22149649262428284, "learning_rate": 1.5718442878924246e-05, "loss": 0.0286, "step": 15070 }, { "grad_norm": 0.19988371431827545, "learning_rate": 1.5658307672779593e-05, "loss": 0.0258, "step": 15080 }, { "grad_norm": 0.18709450960159302, "learning_rate": 1.5598266355510427e-05, "loss": 0.0258, "step": 15090 }, { "grad_norm": 0.2044028788805008, "learning_rate": 1.553831909126744e-05, "loss": 0.0248, "step": 15100 }, { "grad_norm": 0.20206883549690247, "learning_rate": 1.5478466043944135e-05, "loss": 0.0262, "step": 15110 }, { "grad_norm": 0.1678619533777237, "learning_rate": 1.5418707377176468e-05, "loss": 0.026, "step": 15120 }, { "grad_norm": 0.18956395983695984, "learning_rate": 1.535904325434233e-05, "loss": 0.0261, "step": 15130 }, { "grad_norm": 0.2123091071844101, "learning_rate": 1.529947383856118e-05, "loss": 0.0254, "step": 15140 }, { "grad_norm": 0.17061492800712585, "learning_rate": 1.5239999292693524e-05, "loss": 0.0255, "step": 15150 }, { "grad_norm": 0.1714896559715271, "learning_rate": 1.5180619779340505e-05, "loss": 0.0241, "step": 15160 }, { "grad_norm": 0.19768406450748444, "learning_rate": 1.5121335460843428e-05, "loss": 0.0261, "step": 15170 }, { "grad_norm": 0.15599793195724487, "learning_rate": 1.5062146499283347e-05, "loss": 0.0263, "step": 15180 }, { "grad_norm": 0.16431331634521484, "learning_rate": 1.5003053056480643e-05, "loss": 0.0238, "step": 15190 }, { "grad_norm": 0.17185670137405396, "learning_rate": 1.4944055293994551e-05, "loss": 0.0273, "step": 15200 }, { "grad_norm": 0.18507447838783264, "learning_rate": 1.4885153373122656e-05, "loss": 0.0267, "step": 15210 }, { "grad_norm": 0.16483792662620544, "learning_rate": 1.482634745490059e-05, "loss": 0.0225, "step": 15220 }, { "grad_norm": 0.1572796255350113, "learning_rate": 1.4767637700101466e-05, "loss": 0.0255, "step": 15230 }, { "grad_norm": 0.16017326712608337, "learning_rate": 1.4709024269235528e-05, "loss": 0.023, "step": 15240 }, { "grad_norm": 0.14422522485256195, "learning_rate": 1.4650507322549684e-05, "loss": 0.0224, "step": 15250 }, { "grad_norm": 0.1607881784439087, "learning_rate": 1.4592087020026972e-05, "loss": 0.0248, "step": 15260 }, { "grad_norm": 0.18437881767749786, "learning_rate": 1.4533763521386318e-05, "loss": 0.0258, "step": 15270 }, { "grad_norm": 0.17402410507202148, "learning_rate": 1.44755369860819e-05, "loss": 0.0244, "step": 15280 }, { "grad_norm": 0.19908024370670319, "learning_rate": 1.441740757330287e-05, "loss": 0.0238, "step": 15290 }, { "grad_norm": 0.17220504581928253, "learning_rate": 1.4359375441972844e-05, "loss": 0.0263, "step": 15300 }, { "grad_norm": 0.14140546321868896, "learning_rate": 1.4301440750749395e-05, "loss": 0.0255, "step": 15310 }, { "grad_norm": 0.155573770403862, "learning_rate": 1.4243603658023808e-05, "loss": 0.0239, "step": 15320 }, { "grad_norm": 0.17696604132652283, "learning_rate": 1.4185864321920444e-05, "loss": 0.0232, "step": 15330 }, { "grad_norm": 0.16915282607078552, "learning_rate": 1.4128222900296485e-05, "loss": 0.0265, "step": 15340 }, { "grad_norm": 0.16938745975494385, "learning_rate": 1.407067955074135e-05, "loss": 0.0254, "step": 15350 }, { "grad_norm": 0.15487852692604065, "learning_rate": 1.4013234430576356e-05, "loss": 0.0275, "step": 15360 }, { "grad_norm": 0.2000579833984375, "learning_rate": 1.3955887696854286e-05, "loss": 0.0261, "step": 15370 }, { "grad_norm": 0.1404605358839035, "learning_rate": 1.38986395063589e-05, "loss": 0.0271, "step": 15380 }, { "grad_norm": 0.17215444147586823, "learning_rate": 1.3841490015604597e-05, "loss": 0.0253, "step": 15390 }, { "grad_norm": 0.19491508603096008, "learning_rate": 1.3784439380835879e-05, "loss": 0.0239, "step": 15400 }, { "grad_norm": 0.14231058955192566, "learning_rate": 1.3727487758026986e-05, "loss": 0.0235, "step": 15410 }, { "grad_norm": 0.16652671992778778, "learning_rate": 1.3670635302881525e-05, "loss": 0.024, "step": 15420 }, { "grad_norm": 0.19951625168323517, "learning_rate": 1.3613882170831888e-05, "loss": 0.0243, "step": 15430 }, { "grad_norm": 0.16555453836917877, "learning_rate": 1.355722851703901e-05, "loss": 0.0239, "step": 15440 }, { "grad_norm": 0.16589972376823425, "learning_rate": 1.3500674496391814e-05, "loss": 0.028, "step": 15450 }, { "grad_norm": 0.1437523066997528, "learning_rate": 1.3444220263506795e-05, "loss": 0.0219, "step": 15460 }, { "grad_norm": 0.14988291263580322, "learning_rate": 1.3387865972727714e-05, "loss": 0.0238, "step": 15470 }, { "grad_norm": 0.16319052875041962, "learning_rate": 1.3331611778125036e-05, "loss": 0.0272, "step": 15480 }, { "grad_norm": 0.19352303445339203, "learning_rate": 1.3275457833495564e-05, "loss": 0.0231, "step": 15490 }, { "grad_norm": 0.17037786543369293, "learning_rate": 1.3219404292362065e-05, "loss": 0.0257, "step": 15500 }, { "grad_norm": 0.16728441417217255, "learning_rate": 1.3163451307972751e-05, "loss": 0.026, "step": 15510 }, { "grad_norm": 0.16533498466014862, "learning_rate": 1.3107599033300977e-05, "loss": 0.0251, "step": 15520 }, { "grad_norm": 0.18903371691703796, "learning_rate": 1.305184762104471e-05, "loss": 0.0274, "step": 15530 }, { "grad_norm": 0.18545706570148468, "learning_rate": 1.2996197223626178e-05, "loss": 0.0252, "step": 15540 }, { "grad_norm": 0.17952699959278107, "learning_rate": 1.2940647993191457e-05, "loss": 0.0248, "step": 15550 }, { "grad_norm": 0.15502162277698517, "learning_rate": 1.2885200081610005e-05, "loss": 0.0275, "step": 15560 }, { "grad_norm": 0.16711102426052094, "learning_rate": 1.2829853640474316e-05, "loss": 0.0241, "step": 15570 }, { "grad_norm": 0.12957648932933807, "learning_rate": 1.2774608821099438e-05, "loss": 0.0232, "step": 15580 }, { "grad_norm": 0.14523953199386597, "learning_rate": 1.2719465774522577e-05, "loss": 0.0266, "step": 15590 }, { "grad_norm": 0.1757688671350479, "learning_rate": 1.2664424651502755e-05, "loss": 0.0269, "step": 15600 }, { "grad_norm": 0.17657236754894257, "learning_rate": 1.260948560252026e-05, "loss": 0.0254, "step": 15610 }, { "grad_norm": 0.16933436691761017, "learning_rate": 1.2554648777776396e-05, "loss": 0.0226, "step": 15620 }, { "grad_norm": 0.15576335787773132, "learning_rate": 1.2499914327192919e-05, "loss": 0.0253, "step": 15630 }, { "grad_norm": 0.19012698531150818, "learning_rate": 1.2445282400411722e-05, "loss": 0.0254, "step": 15640 }, { "grad_norm": 0.17535921931266785, "learning_rate": 1.2390753146794437e-05, "loss": 0.026, "step": 15650 }, { "grad_norm": 0.16591106355190277, "learning_rate": 1.2336326715421925e-05, "loss": 0.0264, "step": 15660 }, { "grad_norm": 0.17381109297275543, "learning_rate": 1.2282003255094005e-05, "loss": 0.0281, "step": 15670 }, { "grad_norm": 0.15091031789779663, "learning_rate": 1.2227782914328928e-05, "loss": 0.0238, "step": 15680 }, { "grad_norm": 0.1373746395111084, "learning_rate": 1.2173665841363018e-05, "loss": 0.0241, "step": 15690 }, { "grad_norm": 0.13374920189380646, "learning_rate": 1.211965218415032e-05, "loss": 0.0265, "step": 15700 }, { "grad_norm": 0.15238770842552185, "learning_rate": 1.2065742090362082e-05, "loss": 0.0258, "step": 15710 }, { "grad_norm": 0.1686936467885971, "learning_rate": 1.2011935707386457e-05, "loss": 0.0247, "step": 15720 }, { "grad_norm": 0.15188641846179962, "learning_rate": 1.1958233182328044e-05, "loss": 0.0227, "step": 15730 }, { "grad_norm": 0.14769868552684784, "learning_rate": 1.1904634662007474e-05, "loss": 0.0254, "step": 15740 }, { "grad_norm": 0.17211928963661194, "learning_rate": 1.1851140292961088e-05, "loss": 0.0232, "step": 15750 }, { "grad_norm": 0.17551949620246887, "learning_rate": 1.1797750221440424e-05, "loss": 0.0259, "step": 15760 }, { "grad_norm": 0.17025800049304962, "learning_rate": 1.1744464593411897e-05, "loss": 0.0227, "step": 15770 }, { "grad_norm": 0.16701391339302063, "learning_rate": 1.1691283554556399e-05, "loss": 0.026, "step": 15780 }, { "grad_norm": 0.14495499432086945, "learning_rate": 1.1638207250268834e-05, "loss": 0.0239, "step": 15790 }, { "grad_norm": 0.15502488613128662, "learning_rate": 1.158523582565782e-05, "loss": 0.0232, "step": 15800 }, { "grad_norm": 0.15616542100906372, "learning_rate": 1.1532369425545192e-05, "loss": 0.0249, "step": 15810 }, { "grad_norm": 0.1773899346590042, "learning_rate": 1.1479608194465662e-05, "loss": 0.0244, "step": 15820 }, { "grad_norm": 0.1428750455379486, "learning_rate": 1.1426952276666442e-05, "loss": 0.0242, "step": 15830 }, { "grad_norm": 0.15367811918258667, "learning_rate": 1.1374401816106778e-05, "loss": 0.0246, "step": 15840 }, { "grad_norm": 0.14885669946670532, "learning_rate": 1.1321956956457646e-05, "loss": 0.0247, "step": 15850 }, { "grad_norm": 0.16445371508598328, "learning_rate": 1.1269617841101277e-05, "loss": 0.0241, "step": 15860 }, { "grad_norm": 0.14984504878520966, "learning_rate": 1.1217384613130804e-05, "loss": 0.0247, "step": 15870 }, { "grad_norm": 0.1326625496149063, "learning_rate": 1.11652574153499e-05, "loss": 0.0242, "step": 15880 }, { "grad_norm": 0.15226830542087555, "learning_rate": 1.1113236390272303e-05, "loss": 0.0252, "step": 15890 }, { "grad_norm": 0.19900281727313995, "learning_rate": 1.106132168012155e-05, "loss": 0.0263, "step": 15900 }, { "grad_norm": 0.18125948309898376, "learning_rate": 1.1009513426830448e-05, "loss": 0.0263, "step": 15910 }, { "grad_norm": 0.1417820304632187, "learning_rate": 1.0957811772040777e-05, "loss": 0.026, "step": 15920 }, { "grad_norm": 0.16321755945682526, "learning_rate": 1.0906216857102913e-05, "loss": 0.0279, "step": 15930 }, { "grad_norm": 0.16897064447402954, "learning_rate": 1.0854728823075355e-05, "loss": 0.0247, "step": 15940 }, { "grad_norm": 0.15518981218338013, "learning_rate": 1.0803347810724452e-05, "loss": 0.0266, "step": 15950 }, { "grad_norm": 0.14046376943588257, "learning_rate": 1.0752073960523911e-05, "loss": 0.0235, "step": 15960 }, { "grad_norm": 0.13833487033843994, "learning_rate": 1.070090741265447e-05, "loss": 0.0248, "step": 15970 }, { "grad_norm": 0.14494533836841583, "learning_rate": 1.0649848307003547e-05, "loss": 0.026, "step": 15980 }, { "grad_norm": 0.1644425243139267, "learning_rate": 1.0598896783164757e-05, "loss": 0.0272, "step": 15990 }, { "grad_norm": 0.1618363857269287, "learning_rate": 1.0548052980437645e-05, "loss": 0.0231, "step": 16000 }, { "grad_norm": 0.135115846991539, "learning_rate": 1.049731703782722e-05, "loss": 0.0264, "step": 16010 }, { "grad_norm": 0.14955632388591766, "learning_rate": 1.0446689094043587e-05, "loss": 0.0263, "step": 16020 }, { "grad_norm": 0.13370957970619202, "learning_rate": 1.039616928750165e-05, "loss": 0.0243, "step": 16030 }, { "grad_norm": 0.1464678943157196, "learning_rate": 1.0345757756320612e-05, "loss": 0.0233, "step": 16040 }, { "grad_norm": 0.14538177847862244, "learning_rate": 1.0295454638323666e-05, "loss": 0.0267, "step": 16050 }, { "grad_norm": 0.1585129350423813, "learning_rate": 1.0245260071037632e-05, "loss": 0.0262, "step": 16060 }, { "grad_norm": 0.18208138644695282, "learning_rate": 1.0195174191692518e-05, "loss": 0.0254, "step": 16070 }, { "grad_norm": 0.16572202742099762, "learning_rate": 1.014519713722124e-05, "loss": 0.0249, "step": 16080 }, { "grad_norm": 0.13396842777729034, "learning_rate": 1.0095329044259132e-05, "loss": 0.0233, "step": 16090 }, { "grad_norm": 0.12757429480552673, "learning_rate": 1.004557004914365e-05, "loss": 0.0255, "step": 16100 }, { "grad_norm": 0.1495392620563507, "learning_rate": 9.995920287914007e-06, "loss": 0.0255, "step": 16110 }, { "grad_norm": 0.142767533659935, "learning_rate": 9.946379896310737e-06, "loss": 0.0248, "step": 16120 }, { "grad_norm": 0.1585933268070221, "learning_rate": 9.896949009775396e-06, "loss": 0.0245, "step": 16130 }, { "grad_norm": 0.1262839138507843, "learning_rate": 9.847627763450134e-06, "loss": 0.0239, "step": 16140 }, { "grad_norm": 0.144893616437912, "learning_rate": 9.798416292177337e-06, "loss": 0.0251, "step": 16150 }, { "grad_norm": 0.13876210153102875, "learning_rate": 9.74931473049932e-06, "loss": 0.027, "step": 16160 }, { "grad_norm": 0.15736502408981323, "learning_rate": 9.700323212657847e-06, "loss": 0.0245, "step": 16170 }, { "grad_norm": 0.15896163880825043, "learning_rate": 9.65144187259388e-06, "loss": 0.025, "step": 16180 }, { "grad_norm": 0.16438022255897522, "learning_rate": 9.602670843947132e-06, "loss": 0.0254, "step": 16190 }, { "grad_norm": 0.14859507977962494, "learning_rate": 9.554010260055713e-06, "loss": 0.0247, "step": 16200 }, { "grad_norm": 0.1846451312303543, "learning_rate": 9.505460253955834e-06, "loss": 0.0281, "step": 16210 }, { "grad_norm": 0.18279597163200378, "learning_rate": 9.457020958381324e-06, "loss": 0.0264, "step": 16220 }, { "grad_norm": 0.133096843957901, "learning_rate": 9.408692505763395e-06, "loss": 0.0243, "step": 16230 }, { "grad_norm": 0.14880084991455078, "learning_rate": 9.360475028230181e-06, "loss": 0.0248, "step": 16240 }, { "grad_norm": 0.13168004155158997, "learning_rate": 9.312368657606412e-06, "loss": 0.0262, "step": 16250 }, { "grad_norm": 0.14390091598033905, "learning_rate": 9.264373525413096e-06, "loss": 0.0255, "step": 16260 }, { "grad_norm": 0.17718757688999176, "learning_rate": 9.216489762867058e-06, "loss": 0.0251, "step": 16270 }, { "grad_norm": 0.13048505783081055, "learning_rate": 9.168717500880708e-06, "loss": 0.0246, "step": 16280 }, { "grad_norm": 0.14658060669898987, "learning_rate": 9.121056870061574e-06, "loss": 0.0267, "step": 16290 }, { "grad_norm": 0.14411421120166779, "learning_rate": 9.073508000711983e-06, "loss": 0.023, "step": 16300 }, { "grad_norm": 0.13382606208324432, "learning_rate": 9.026071022828758e-06, "loss": 0.0258, "step": 16310 }, { "grad_norm": 0.16291281580924988, "learning_rate": 8.978746066102771e-06, "loss": 0.0242, "step": 16320 }, { "grad_norm": 0.14279551804065704, "learning_rate": 8.931533259918634e-06, "loss": 0.0238, "step": 16330 }, { "grad_norm": 0.14823980629444122, "learning_rate": 8.884432733354382e-06, "loss": 0.0246, "step": 16340 }, { "grad_norm": 0.13507714867591858, "learning_rate": 8.837444615181029e-06, "loss": 0.026, "step": 16350 }, { "grad_norm": 0.1421726793050766, "learning_rate": 8.790569033862323e-06, "loss": 0.0235, "step": 16360 }, { "grad_norm": 0.1545550376176834, "learning_rate": 8.7438061175543e-06, "loss": 0.0247, "step": 16370 }, { "grad_norm": 0.16946233808994293, "learning_rate": 8.697155994104978e-06, "loss": 0.026, "step": 16380 }, { "grad_norm": 0.10995147377252579, "learning_rate": 8.650618791054033e-06, "loss": 0.0283, "step": 16390 }, { "grad_norm": 0.15010538697242737, "learning_rate": 8.604194635632373e-06, "loss": 0.0242, "step": 16400 }, { "grad_norm": 0.12514448165893555, "learning_rate": 8.557883654761906e-06, "loss": 0.0241, "step": 16410 }, { "grad_norm": 0.14020079374313354, "learning_rate": 8.511685975055061e-06, "loss": 0.0224, "step": 16420 }, { "grad_norm": 0.14358924329280853, "learning_rate": 8.46560172281452e-06, "loss": 0.0251, "step": 16430 }, { "grad_norm": 0.12778478860855103, "learning_rate": 8.419631024032893e-06, "loss": 0.0244, "step": 16440 }, { "grad_norm": 0.15085013210773468, "learning_rate": 8.373774004392293e-06, "loss": 0.0249, "step": 16450 }, { "grad_norm": 0.15054288506507874, "learning_rate": 8.32803078926409e-06, "loss": 0.0229, "step": 16460 }, { "grad_norm": 0.12330876290798187, "learning_rate": 8.282401503708454e-06, "loss": 0.0229, "step": 16470 }, { "grad_norm": 0.1296664923429489, "learning_rate": 8.23688627247412e-06, "loss": 0.0242, "step": 16480 }, { "grad_norm": 0.132670059800148, "learning_rate": 8.191485219998007e-06, "loss": 0.0238, "step": 16490 }, { "grad_norm": 0.1523398607969284, "learning_rate": 8.146198470404843e-06, "loss": 0.0238, "step": 16500 }, { "grad_norm": 0.12279612571001053, "learning_rate": 8.101026147506897e-06, "loss": 0.0239, "step": 16510 }, { "grad_norm": 0.14855366945266724, "learning_rate": 8.05596837480353e-06, "loss": 0.0235, "step": 16520 }, { "grad_norm": 0.12448320537805557, "learning_rate": 8.011025275480998e-06, "loss": 0.0233, "step": 16530 }, { "grad_norm": 0.12857545912265778, "learning_rate": 7.966196972412027e-06, "loss": 0.0234, "step": 16540 }, { "grad_norm": 0.1400635689496994, "learning_rate": 7.92148358815547e-06, "loss": 0.0242, "step": 16550 }, { "grad_norm": 0.1419910192489624, "learning_rate": 7.87688524495604e-06, "loss": 0.0211, "step": 16560 }, { "grad_norm": 0.11800920963287354, "learning_rate": 7.83240206474386e-06, "loss": 0.0226, "step": 16570 }, { "grad_norm": 0.12714478373527527, "learning_rate": 7.788034169134272e-06, "loss": 0.0251, "step": 16580 }, { "grad_norm": 0.14429329335689545, "learning_rate": 7.743781679427414e-06, "loss": 0.0236, "step": 16590 }, { "grad_norm": 0.1541374772787094, "learning_rate": 7.699644716607895e-06, "loss": 0.0238, "step": 16600 }, { "grad_norm": 0.13760797679424286, "learning_rate": 7.655623401344486e-06, "loss": 0.0254, "step": 16610 }, { "grad_norm": 0.17412935197353363, "learning_rate": 7.611717853989775e-06, "loss": 0.0276, "step": 16620 }, { "grad_norm": 0.10819465667009354, "learning_rate": 7.567928194579854e-06, "loss": 0.0252, "step": 16630 }, { "grad_norm": 0.12968656420707703, "learning_rate": 7.524254542833997e-06, "loss": 0.0233, "step": 16640 }, { "grad_norm": 0.1914411336183548, "learning_rate": 7.480697018154286e-06, "loss": 0.0248, "step": 16650 }, { "grad_norm": 0.12658068537712097, "learning_rate": 7.437255739625332e-06, "loss": 0.0251, "step": 16660 }, { "grad_norm": 0.15159441530704498, "learning_rate": 7.393930826013923e-06, "loss": 0.0247, "step": 16670 }, { "grad_norm": 0.12754197418689728, "learning_rate": 7.350722395768722e-06, "loss": 0.0257, "step": 16680 }, { "grad_norm": 0.12192249298095703, "learning_rate": 7.307630567019963e-06, "loss": 0.0224, "step": 16690 }, { "grad_norm": 0.10991044342517853, "learning_rate": 7.264655457579e-06, "loss": 0.0229, "step": 16700 }, { "grad_norm": 0.1220596507191658, "learning_rate": 7.221797184938184e-06, "loss": 0.0237, "step": 16710 }, { "grad_norm": 0.1740283966064453, "learning_rate": 7.179055866270373e-06, "loss": 0.0264, "step": 16720 }, { "grad_norm": 0.12445994466543198, "learning_rate": 7.136431618428707e-06, "loss": 0.0252, "step": 16730 }, { "grad_norm": 0.14821989834308624, "learning_rate": 7.09392455794628e-06, "loss": 0.0249, "step": 16740 }, { "grad_norm": 0.14826400578022003, "learning_rate": 7.051534801035725e-06, "loss": 0.0248, "step": 16750 }, { "grad_norm": 0.12139523029327393, "learning_rate": 7.00926246358905e-06, "loss": 0.0241, "step": 16760 }, { "grad_norm": 0.0979883223772049, "learning_rate": 6.967107661177191e-06, "loss": 0.0253, "step": 16770 }, { "grad_norm": 0.14146219193935394, "learning_rate": 6.925070509049786e-06, "loss": 0.0253, "step": 16780 }, { "grad_norm": 0.13401368260383606, "learning_rate": 6.883151122134812e-06, "loss": 0.0248, "step": 16790 }, { "grad_norm": 0.11482026427984238, "learning_rate": 6.8413496150382394e-06, "loss": 0.0235, "step": 16800 }, { "grad_norm": 0.13530462980270386, "learning_rate": 6.7996661020438165e-06, "loss": 0.0237, "step": 16810 }, { "grad_norm": 0.11568176001310349, "learning_rate": 6.758100697112662e-06, "loss": 0.0255, "step": 16820 }, { "grad_norm": 0.14525540173053741, "learning_rate": 6.716653513883026e-06, "loss": 0.0241, "step": 16830 }, { "grad_norm": 0.14404776692390442, "learning_rate": 6.675324665669913e-06, "loss": 0.0239, "step": 16840 }, { "grad_norm": 0.12877868115901947, "learning_rate": 6.634114265464803e-06, "loss": 0.025, "step": 16850 }, { "grad_norm": 0.1437789499759674, "learning_rate": 6.59302242593538e-06, "loss": 0.0237, "step": 16860 }, { "grad_norm": 0.12646815180778503, "learning_rate": 6.552049259425141e-06, "loss": 0.0229, "step": 16870 }, { "grad_norm": 0.1402336210012436, "learning_rate": 6.511194877953181e-06, "loss": 0.024, "step": 16880 }, { "grad_norm": 0.1390192061662674, "learning_rate": 6.470459393213813e-06, "loss": 0.0241, "step": 16890 }, { "grad_norm": 0.13139183819293976, "learning_rate": 6.429842916576279e-06, "loss": 0.0242, "step": 16900 }, { "grad_norm": 0.14657972753047943, "learning_rate": 6.389345559084503e-06, "loss": 0.0232, "step": 16910 }, { "grad_norm": 0.1229761391878128, "learning_rate": 6.348967431456682e-06, "loss": 0.0241, "step": 16920 }, { "grad_norm": 0.1160895898938179, "learning_rate": 6.30870864408511e-06, "loss": 0.0239, "step": 16930 }, { "grad_norm": 0.1294250637292862, "learning_rate": 6.268569307035754e-06, "loss": 0.026, "step": 16940 }, { "grad_norm": 0.1368834376335144, "learning_rate": 6.228549530048022e-06, "loss": 0.0275, "step": 16950 }, { "grad_norm": 0.1470441222190857, "learning_rate": 6.1886494225344814e-06, "loss": 0.022, "step": 16960 }, { "grad_norm": 0.12819932401180267, "learning_rate": 6.148869093580479e-06, "loss": 0.025, "step": 16970 }, { "grad_norm": 0.12555848062038422, "learning_rate": 6.109208651943921e-06, "loss": 0.025, "step": 16980 }, { "grad_norm": 0.11611313372850418, "learning_rate": 6.069668206054946e-06, "loss": 0.0237, "step": 16990 }, { "grad_norm": 0.11743858456611633, "learning_rate": 6.0302478640156145e-06, "loss": 0.0225, "step": 17000 }, { "grad_norm": 0.11387670040130615, "learning_rate": 5.990947733599644e-06, "loss": 0.0227, "step": 17010 }, { "grad_norm": 0.12867337465286255, "learning_rate": 5.951767922252105e-06, "loss": 0.0258, "step": 17020 }, { "grad_norm": 0.1321384310722351, "learning_rate": 5.912708537089068e-06, "loss": 0.0252, "step": 17030 }, { "grad_norm": 0.11824095249176025, "learning_rate": 5.873769684897434e-06, "loss": 0.0209, "step": 17040 }, { "grad_norm": 0.13404354453086853, "learning_rate": 5.834951472134514e-06, "loss": 0.024, "step": 17050 }, { "grad_norm": 0.11989074945449829, "learning_rate": 5.796254004927832e-06, "loss": 0.0225, "step": 17060 }, { "grad_norm": 0.135044127702713, "learning_rate": 5.757677389074806e-06, "loss": 0.026, "step": 17070 }, { "grad_norm": 0.12645506858825684, "learning_rate": 5.719221730042385e-06, "loss": 0.0245, "step": 17080 }, { "grad_norm": 0.12745541334152222, "learning_rate": 5.680887132966911e-06, "loss": 0.0238, "step": 17090 }, { "grad_norm": 0.11851977556943893, "learning_rate": 5.642673702653683e-06, "loss": 0.0262, "step": 17100 }, { "grad_norm": 0.12033791840076447, "learning_rate": 5.604581543576781e-06, "loss": 0.0234, "step": 17110 }, { "grad_norm": 0.11608454585075378, "learning_rate": 5.566610759878704e-06, "loss": 0.0229, "step": 17120 }, { "grad_norm": 0.11721862852573395, "learning_rate": 5.528761455370119e-06, "loss": 0.024, "step": 17130 }, { "grad_norm": 0.13248895108699799, "learning_rate": 5.491033733529594e-06, "loss": 0.0252, "step": 17140 }, { "grad_norm": 0.12607073783874512, "learning_rate": 5.453427697503255e-06, "loss": 0.0235, "step": 17150 }, { "grad_norm": 0.1425388753414154, "learning_rate": 5.415943450104599e-06, "loss": 0.0286, "step": 17160 }, { "grad_norm": 0.13556380569934845, "learning_rate": 5.378581093814111e-06, "loss": 0.029, "step": 17170 }, { "grad_norm": 0.12600865960121155, "learning_rate": 5.3413407307790375e-06, "loss": 0.0264, "step": 17180 }, { "grad_norm": 0.1356276422739029, "learning_rate": 5.30422246281313e-06, "loss": 0.0236, "step": 17190 }, { "grad_norm": 0.1046186313033104, "learning_rate": 5.267226391396296e-06, "loss": 0.0242, "step": 17200 }, { "grad_norm": 0.10092248767614365, "learning_rate": 5.2303526176744e-06, "loss": 0.0259, "step": 17210 }, { "grad_norm": 0.12372326105833054, "learning_rate": 5.193601242458929e-06, "loss": 0.0215, "step": 17220 }, { "grad_norm": 0.15749198198318481, "learning_rate": 5.156972366226714e-06, "loss": 0.0232, "step": 17230 }, { "grad_norm": 0.12315038591623306, "learning_rate": 5.120466089119735e-06, "loss": 0.0241, "step": 17240 }, { "grad_norm": 0.14535780251026154, "learning_rate": 5.084082510944749e-06, "loss": 0.0251, "step": 17250 }, { "grad_norm": 0.10578538477420807, "learning_rate": 5.047821731173058e-06, "loss": 0.023, "step": 17260 }, { "grad_norm": 0.11672276258468628, "learning_rate": 5.011683848940274e-06, "loss": 0.0276, "step": 17270 }, { "grad_norm": 0.13364452123641968, "learning_rate": 4.975668963045954e-06, "loss": 0.0252, "step": 17280 }, { "grad_norm": 0.12619805335998535, "learning_rate": 4.9397771719534525e-06, "loss": 0.0227, "step": 17290 }, { "grad_norm": 0.11984161287546158, "learning_rate": 4.904008573789548e-06, "loss": 0.028, "step": 17300 }, { "grad_norm": 0.13856521248817444, "learning_rate": 4.8683632663442005e-06, "loss": 0.0232, "step": 17310 }, { "grad_norm": 0.12743894755840302, "learning_rate": 4.832841347070343e-06, "loss": 0.0243, "step": 17320 }, { "grad_norm": 0.11950954049825668, "learning_rate": 4.797442913083539e-06, "loss": 0.0244, "step": 17330 }, { "grad_norm": 0.11553730815649033, "learning_rate": 4.7621680611617596e-06, "loss": 0.023, "step": 17340 }, { "grad_norm": 0.10659778863191605, "learning_rate": 4.727016887745095e-06, "loss": 0.0242, "step": 17350 }, { "grad_norm": 0.10067514330148697, "learning_rate": 4.691989488935511e-06, "loss": 0.022, "step": 17360 }, { "grad_norm": 0.11429009586572647, "learning_rate": 4.657085960496588e-06, "loss": 0.0248, "step": 17370 }, { "grad_norm": 0.12288064509630203, "learning_rate": 4.6223063978532265e-06, "loss": 0.0275, "step": 17380 }, { "grad_norm": 0.09415844827890396, "learning_rate": 4.587650896091439e-06, "loss": 0.0235, "step": 17390 }, { "grad_norm": 0.13084301352500916, "learning_rate": 4.553119549958035e-06, "loss": 0.0244, "step": 17400 }, { "grad_norm": 0.10119587182998657, "learning_rate": 4.518712453860385e-06, "loss": 0.0251, "step": 17410 }, { "grad_norm": 0.12192751467227936, "learning_rate": 4.484429701866205e-06, "loss": 0.0227, "step": 17420 }, { "grad_norm": 0.1199282556772232, "learning_rate": 4.4502713877031975e-06, "loss": 0.0244, "step": 17430 }, { "grad_norm": 0.1184970960021019, "learning_rate": 4.416237604758911e-06, "loss": 0.0226, "step": 17440 }, { "grad_norm": 0.11002273857593536, "learning_rate": 4.3823284460804025e-06, "loss": 0.0238, "step": 17450 }, { "grad_norm": 0.10795886069536209, "learning_rate": 4.348544004374011e-06, "loss": 0.0244, "step": 17460 }, { "grad_norm": 0.11268334835767746, "learning_rate": 4.314884372005123e-06, "loss": 0.025, "step": 17470 }, { "grad_norm": 0.10237261652946472, "learning_rate": 4.281349640997867e-06, "loss": 0.0238, "step": 17480 }, { "grad_norm": 0.11249697953462601, "learning_rate": 4.247939903034942e-06, "loss": 0.0229, "step": 17490 }, { "grad_norm": 0.12462349236011505, "learning_rate": 4.214655249457284e-06, "loss": 0.0238, "step": 17500 }, { "grad_norm": 0.10992822796106339, "learning_rate": 4.181495771263855e-06, "loss": 0.0228, "step": 17510 }, { "grad_norm": 0.12625648081302643, "learning_rate": 4.148461559111427e-06, "loss": 0.0231, "step": 17520 }, { "grad_norm": 0.10650869458913803, "learning_rate": 4.115552703314252e-06, "loss": 0.0235, "step": 17530 }, { "grad_norm": 0.10854001343250275, "learning_rate": 4.082769293843886e-06, "loss": 0.0241, "step": 17540 }, { "grad_norm": 0.1260230541229248, "learning_rate": 4.050111420328939e-06, "loss": 0.0237, "step": 17550 }, { "grad_norm": 0.11088349670171738, "learning_rate": 4.017579172054764e-06, "loss": 0.0234, "step": 17560 }, { "grad_norm": 0.10440419614315033, "learning_rate": 3.985172637963308e-06, "loss": 0.0201, "step": 17570 }, { "grad_norm": 0.11542070657014847, "learning_rate": 3.952891906652784e-06, "loss": 0.0235, "step": 17580 }, { "grad_norm": 0.13438129425048828, "learning_rate": 3.920737066377478e-06, "loss": 0.0213, "step": 17590 }, { "grad_norm": 0.09839281439781189, "learning_rate": 3.888708205047509e-06, "loss": 0.022, "step": 17600 }, { "grad_norm": 0.10244724899530411, "learning_rate": 3.856805410228542e-06, "loss": 0.0254, "step": 17610 }, { "grad_norm": 0.15502825379371643, "learning_rate": 3.82502876914162e-06, "loss": 0.0259, "step": 17620 }, { "grad_norm": 0.11315834522247314, "learning_rate": 3.7933783686628586e-06, "loss": 0.0216, "step": 17630 }, { "grad_norm": 0.10236119478940964, "learning_rate": 3.7618542953232306e-06, "loss": 0.0236, "step": 17640 }, { "grad_norm": 0.10119117051362991, "learning_rate": 3.7304566353083658e-06, "loss": 0.0236, "step": 17650 }, { "grad_norm": 0.10944440960884094, "learning_rate": 3.6991854744582555e-06, "loss": 0.023, "step": 17660 }, { "grad_norm": 0.09412858635187149, "learning_rate": 3.6680408982670777e-06, "loss": 0.0251, "step": 17670 }, { "grad_norm": 0.09549499303102493, "learning_rate": 3.637022991882899e-06, "loss": 0.024, "step": 17680 }, { "grad_norm": 0.11111559718847275, "learning_rate": 3.606131840107485e-06, "loss": 0.0247, "step": 17690 }, { "grad_norm": 0.1009516492486, "learning_rate": 3.575367527396084e-06, "loss": 0.0228, "step": 17700 }, { "grad_norm": 0.1122208908200264, "learning_rate": 3.5447301378571386e-06, "loss": 0.0257, "step": 17710 }, { "grad_norm": 0.1249672919511795, "learning_rate": 3.514219755252113e-06, "loss": 0.0241, "step": 17720 }, { "grad_norm": 0.08895356208086014, "learning_rate": 3.4838364629952213e-06, "loss": 0.0225, "step": 17730 }, { "grad_norm": 0.10254279524087906, "learning_rate": 3.4535803441532123e-06, "loss": 0.0231, "step": 17740 }, { "grad_norm": 0.10215990990400314, "learning_rate": 3.4234514814451836e-06, "loss": 0.0241, "step": 17750 }, { "grad_norm": 0.13270002603530884, "learning_rate": 3.393449957242273e-06, "loss": 0.0247, "step": 17760 }, { "grad_norm": 0.12597966194152832, "learning_rate": 3.363575853567524e-06, "loss": 0.0247, "step": 17770 }, { "grad_norm": 0.11643832176923752, "learning_rate": 3.3338292520955826e-06, "loss": 0.0247, "step": 17780 }, { "grad_norm": 0.10814177244901657, "learning_rate": 3.304210234152516e-06, "loss": 0.0231, "step": 17790 }, { "grad_norm": 0.09975577890872955, "learning_rate": 3.2747188807155993e-06, "loss": 0.0231, "step": 17800 }, { "grad_norm": 0.0897221490740776, "learning_rate": 3.2453552724130643e-06, "loss": 0.0235, "step": 17810 }, { "grad_norm": 0.15383723378181458, "learning_rate": 3.216119489523889e-06, "loss": 0.0226, "step": 17820 }, { "grad_norm": 0.10056844353675842, "learning_rate": 3.1870116119775917e-06, "loss": 0.0262, "step": 17830 }, { "grad_norm": 0.09615091234445572, "learning_rate": 3.158031719353999e-06, "loss": 0.0228, "step": 17840 }, { "grad_norm": 0.11905992776155472, "learning_rate": 3.1291798908830273e-06, "loss": 0.0229, "step": 17850 }, { "grad_norm": 0.09205687791109085, "learning_rate": 3.1004562054444853e-06, "loss": 0.0224, "step": 17860 }, { "grad_norm": 0.07834277302026749, "learning_rate": 3.071860741567806e-06, "loss": 0.022, "step": 17870 }, { "grad_norm": 0.10026639699935913, "learning_rate": 3.04339357743193e-06, "loss": 0.0232, "step": 17880 }, { "grad_norm": 0.1038300171494484, "learning_rate": 3.0150547908649628e-06, "loss": 0.0229, "step": 17890 }, { "grad_norm": 0.11887870728969574, "learning_rate": 2.9868444593440957e-06, "loss": 0.0224, "step": 17900 }, { "grad_norm": 0.12054941803216934, "learning_rate": 2.9587626599952846e-06, "loss": 0.0233, "step": 17910 }, { "grad_norm": 0.17431654036045074, "learning_rate": 2.930809469593082e-06, "loss": 0.0223, "step": 17920 }, { "grad_norm": 0.07807499915361404, "learning_rate": 2.9029849645604733e-06, "loss": 0.0219, "step": 17930 }, { "grad_norm": 0.1143551841378212, "learning_rate": 2.8752892209685632e-06, "loss": 0.023, "step": 17940 }, { "grad_norm": 0.12804624438285828, "learning_rate": 2.847722314536483e-06, "loss": 0.023, "step": 17950 }, { "grad_norm": 0.11565615981817245, "learning_rate": 2.820284320631078e-06, "loss": 0.024, "step": 17960 }, { "grad_norm": 0.10989917069673538, "learning_rate": 2.792975314266788e-06, "loss": 0.0213, "step": 17970 }, { "grad_norm": 0.11816196888685226, "learning_rate": 2.7657953701054007e-06, "loss": 0.0219, "step": 17980 }, { "grad_norm": 0.11401066929101944, "learning_rate": 2.7387445624558306e-06, "loss": 0.0255, "step": 17990 }, { "grad_norm": 0.1037629097700119, "learning_rate": 2.7118229652739747e-06, "loss": 0.0251, "step": 18000 }, { "grad_norm": 0.10773545503616333, "learning_rate": 2.6850306521624236e-06, "loss": 0.0211, "step": 18010 }, { "grad_norm": 0.09961125999689102, "learning_rate": 2.6583676963703507e-06, "loss": 0.0228, "step": 18020 }, { "grad_norm": 0.13344767689704895, "learning_rate": 2.631834170793268e-06, "loss": 0.0226, "step": 18030 }, { "grad_norm": 0.10772242397069931, "learning_rate": 2.6054301479728036e-06, "loss": 0.0239, "step": 18040 }, { "grad_norm": 0.09602147340774536, "learning_rate": 2.579155700096575e-06, "loss": 0.0233, "step": 18050 }, { "grad_norm": 0.10497970879077911, "learning_rate": 2.5530108989978873e-06, "loss": 0.0227, "step": 18060 }, { "grad_norm": 0.10348208993673325, "learning_rate": 2.5269958161556416e-06, "loss": 0.0216, "step": 18070 }, { "grad_norm": 0.1250126212835312, "learning_rate": 2.5011105226940888e-06, "loss": 0.023, "step": 18080 }, { "grad_norm": 0.09561929106712341, "learning_rate": 2.4753550893826248e-06, "loss": 0.025, "step": 18090 }, { "grad_norm": 0.09662984311580658, "learning_rate": 2.4497295866356296e-06, "loss": 0.0234, "step": 18100 }, { "grad_norm": 0.09393294155597687, "learning_rate": 2.424234084512228e-06, "loss": 0.0239, "step": 18110 }, { "grad_norm": 0.10085354745388031, "learning_rate": 2.3988686527161687e-06, "loss": 0.0259, "step": 18120 }, { "grad_norm": 0.11587812751531601, "learning_rate": 2.373633360595573e-06, "loss": 0.0235, "step": 18130 }, { "grad_norm": 0.1004292219877243, "learning_rate": 2.3485282771427585e-06, "loss": 0.0253, "step": 18140 }, { "grad_norm": 0.08596368879079819, "learning_rate": 2.3235534709940665e-06, "loss": 0.0263, "step": 18150 }, { "grad_norm": 0.11063624173402786, "learning_rate": 2.2987090104296617e-06, "loss": 0.0229, "step": 18160 }, { "grad_norm": 0.09148791432380676, "learning_rate": 2.273994963373355e-06, "loss": 0.0237, "step": 18170 }, { "grad_norm": 0.09805736690759659, "learning_rate": 2.249411397392409e-06, "loss": 0.0235, "step": 18180 }, { "grad_norm": 0.09009591490030289, "learning_rate": 2.2249583796973506e-06, "loss": 0.0256, "step": 18190 }, { "grad_norm": 0.09204314649105072, "learning_rate": 2.200635977141796e-06, "loss": 0.0276, "step": 18200 }, { "grad_norm": 0.09285461157560349, "learning_rate": 2.17644425622226e-06, "loss": 0.0253, "step": 18210 }, { "grad_norm": 0.10498350113630295, "learning_rate": 2.152383283077991e-06, "loss": 0.023, "step": 18220 }, { "grad_norm": 0.12797679007053375, "learning_rate": 2.128453123490781e-06, "loss": 0.024, "step": 18230 }, { "grad_norm": 0.1216844841837883, "learning_rate": 2.1046538428847462e-06, "loss": 0.0243, "step": 18240 }, { "grad_norm": 0.10825832188129425, "learning_rate": 2.0809855063262273e-06, "loss": 0.0249, "step": 18250 }, { "grad_norm": 0.0917518138885498, "learning_rate": 2.057448178523558e-06, "loss": 0.0225, "step": 18260 }, { "grad_norm": 0.12411748617887497, "learning_rate": 2.034041923826885e-06, "loss": 0.0231, "step": 18270 }, { "grad_norm": 0.07514967769384384, "learning_rate": 2.0107668062280204e-06, "loss": 0.0231, "step": 18280 }, { "grad_norm": 0.09314003586769104, "learning_rate": 1.9876228893602357e-06, "loss": 0.0235, "step": 18290 }, { "grad_norm": 0.12532272934913635, "learning_rate": 1.9646102364981266e-06, "loss": 0.0245, "step": 18300 }, { "grad_norm": 0.09739029407501221, "learning_rate": 1.9417289105574053e-06, "loss": 0.0221, "step": 18310 }, { "grad_norm": 0.10282210260629654, "learning_rate": 1.9189789740947427e-06, "loss": 0.0276, "step": 18320 }, { "grad_norm": 0.12987682223320007, "learning_rate": 1.896360489307597e-06, "loss": 0.0233, "step": 18330 }, { "grad_norm": 0.09377358108758926, "learning_rate": 1.8738735180340362e-06, "loss": 0.0219, "step": 18340 }, { "grad_norm": 0.09026184678077698, "learning_rate": 1.8515181217525824e-06, "loss": 0.0225, "step": 18350 }, { "grad_norm": 0.08504945784807205, "learning_rate": 1.8292943615820457e-06, "loss": 0.0245, "step": 18360 }, { "grad_norm": 0.0785132348537445, "learning_rate": 1.8072022982813296e-06, "loss": 0.0234, "step": 18370 }, { "grad_norm": 0.07971486449241638, "learning_rate": 1.7852419922492925e-06, "loss": 0.022, "step": 18380 }, { "grad_norm": 0.09572304040193558, "learning_rate": 1.763413503524569e-06, "loss": 0.0236, "step": 18390 }, { "grad_norm": 0.09432708472013474, "learning_rate": 1.7417168917854165e-06, "loss": 0.0215, "step": 18400 }, { "grad_norm": 0.08415650576353073, "learning_rate": 1.720152216349552e-06, "loss": 0.0242, "step": 18410 }, { "grad_norm": 0.09118799865245819, "learning_rate": 1.6987195361739595e-06, "loss": 0.0221, "step": 18420 }, { "grad_norm": 0.08558237552642822, "learning_rate": 1.6774189098547832e-06, "loss": 0.0236, "step": 18430 }, { "grad_norm": 0.10001794248819351, "learning_rate": 1.6562503956271069e-06, "loss": 0.0246, "step": 18440 }, { "grad_norm": 0.08565393835306168, "learning_rate": 1.6352140513648417e-06, "loss": 0.023, "step": 18450 }, { "grad_norm": 0.08789020776748657, "learning_rate": 1.6143099345805712e-06, "loss": 0.0235, "step": 18460 }, { "grad_norm": 0.13098320364952087, "learning_rate": 1.5935381024253293e-06, "loss": 0.0231, "step": 18470 }, { "grad_norm": 0.11493830382823944, "learning_rate": 1.572898611688517e-06, "loss": 0.0227, "step": 18480 }, { "grad_norm": 0.11890452355146408, "learning_rate": 1.5523915187977133e-06, "loss": 0.0233, "step": 18490 }, { "grad_norm": 0.08426733314990997, "learning_rate": 1.532016879818532e-06, "loss": 0.0221, "step": 18500 }, { "grad_norm": 0.08867798745632172, "learning_rate": 1.51177475045447e-06, "loss": 0.0225, "step": 18510 }, { "grad_norm": 0.10221709311008453, "learning_rate": 1.4916651860467035e-06, "loss": 0.026, "step": 18520 }, { "grad_norm": 0.14561931788921356, "learning_rate": 1.471688241574043e-06, "loss": 0.0255, "step": 18530 }, { "grad_norm": 0.1005135029554367, "learning_rate": 1.451843971652672e-06, "loss": 0.0249, "step": 18540 }, { "grad_norm": 0.121422179043293, "learning_rate": 1.432132430536076e-06, "loss": 0.023, "step": 18550 }, { "grad_norm": 0.10070405900478363, "learning_rate": 1.412553672114869e-06, "loss": 0.0228, "step": 18560 }, { "grad_norm": 0.07680639624595642, "learning_rate": 1.3931077499166056e-06, "loss": 0.0224, "step": 18570 }, { "grad_norm": 0.0856013298034668, "learning_rate": 1.3737947171057085e-06, "loss": 0.0236, "step": 18580 }, { "grad_norm": 0.11457344889640808, "learning_rate": 1.3546146264832582e-06, "loss": 0.0237, "step": 18590 }, { "grad_norm": 0.10266296565532684, "learning_rate": 1.3355675304869086e-06, "loss": 0.0246, "step": 18600 }, { "grad_norm": 0.09215828776359558, "learning_rate": 1.3166534811906827e-06, "loss": 0.0225, "step": 18610 }, { "grad_norm": 0.07224027067422867, "learning_rate": 1.2978725303048666e-06, "loss": 0.024, "step": 18620 }, { "grad_norm": 0.08439242094755173, "learning_rate": 1.2792247291758762e-06, "loss": 0.0242, "step": 18630 }, { "grad_norm": 0.09039051830768585, "learning_rate": 1.2607101287860635e-06, "loss": 0.0237, "step": 18640 }, { "grad_norm": 0.09681685268878937, "learning_rate": 1.2423287797536654e-06, "loss": 0.0253, "step": 18650 }, { "grad_norm": 0.06782346218824387, "learning_rate": 1.2240807323325776e-06, "loss": 0.0238, "step": 18660 }, { "grad_norm": 0.08303060382604599, "learning_rate": 1.205966036412254e-06, "loss": 0.0227, "step": 18670 }, { "grad_norm": 0.10353781282901764, "learning_rate": 1.1879847415175949e-06, "loss": 0.0221, "step": 18680 }, { "grad_norm": 0.09322594106197357, "learning_rate": 1.1701368968087712e-06, "loss": 0.0238, "step": 18690 }, { "grad_norm": 0.08827011287212372, "learning_rate": 1.1524225510811116e-06, "loss": 0.0232, "step": 18700 }, { "grad_norm": 0.10493210703134537, "learning_rate": 1.1348417527649535e-06, "loss": 0.0247, "step": 18710 }, { "grad_norm": 0.10401958972215652, "learning_rate": 1.1173945499255268e-06, "loss": 0.0228, "step": 18720 }, { "grad_norm": 0.07403160631656647, "learning_rate": 1.1000809902628307e-06, "loss": 0.0265, "step": 18730 }, { "grad_norm": 0.07509875297546387, "learning_rate": 1.082901121111468e-06, "loss": 0.0223, "step": 18740 }, { "grad_norm": 0.10791140049695969, "learning_rate": 1.0658549894405456e-06, "loss": 0.0251, "step": 18750 }, { "grad_norm": 0.08866315335035324, "learning_rate": 1.0489426418535342e-06, "loss": 0.0266, "step": 18760 }, { "grad_norm": 0.0937662124633789, "learning_rate": 1.0321641245881474e-06, "loss": 0.0249, "step": 18770 }, { "grad_norm": 0.1152704507112503, "learning_rate": 1.015519483516214e-06, "loss": 0.0234, "step": 18780 }, { "grad_norm": 0.08053148537874222, "learning_rate": 9.990087641435443e-07, "loss": 0.023, "step": 18790 }, { "grad_norm": 0.10238146781921387, "learning_rate": 9.826320116098132e-07, "loss": 0.0282, "step": 18800 }, { "grad_norm": 0.09985664486885071, "learning_rate": 9.663892706884447e-07, "loss": 0.0226, "step": 18810 }, { "grad_norm": 0.08032656461000443, "learning_rate": 9.502805857864616e-07, "loss": 0.0232, "step": 18820 }, { "grad_norm": 0.08727730065584183, "learning_rate": 9.34306000944396e-07, "loss": 0.0234, "step": 18830 }, { "grad_norm": 0.08486419171094894, "learning_rate": 9.184655598361624e-07, "loss": 0.0217, "step": 18840 }, { "grad_norm": 0.09468334913253784, "learning_rate": 9.027593057689076e-07, "loss": 0.024, "step": 18850 }, { "grad_norm": 0.0847364291548729, "learning_rate": 8.871872816829441e-07, "loss": 0.0234, "step": 18860 }, { "grad_norm": 0.0880277007818222, "learning_rate": 8.717495301515777e-07, "loss": 0.0251, "step": 18870 }, { "grad_norm": 0.09313037991523743, "learning_rate": 8.564460933810415e-07, "loss": 0.0243, "step": 18880 }, { "grad_norm": 0.07175064831972122, "learning_rate": 8.412770132103453e-07, "loss": 0.0233, "step": 18890 }, { "grad_norm": 0.10527171939611435, "learning_rate": 8.262423311111711e-07, "loss": 0.0236, "step": 18900 }, { "grad_norm": 0.07161501049995422, "learning_rate": 8.113420881877665e-07, "loss": 0.0226, "step": 18910 }, { "grad_norm": 0.08372929692268372, "learning_rate": 7.965763251768288e-07, "loss": 0.0235, "step": 18920 }, { "grad_norm": 0.138661727309227, "learning_rate": 7.819450824473995e-07, "loss": 0.0233, "step": 18930 }, { "grad_norm": 0.09790371358394623, "learning_rate": 7.674484000007198e-07, "loss": 0.0263, "step": 18940 }, { "grad_norm": 0.08232130110263824, "learning_rate": 7.530863174701752e-07, "loss": 0.022, "step": 18950 }, { "grad_norm": 0.09777482599020004, "learning_rate": 7.38858874121151e-07, "loss": 0.024, "step": 18960 }, { "grad_norm": 0.07351008802652359, "learning_rate": 7.247661088509328e-07, "loss": 0.0236, "step": 18970 }, { "grad_norm": 0.09060505777597427, "learning_rate": 7.108080601886002e-07, "loss": 0.0238, "step": 18980 }, { "grad_norm": 0.06474795937538147, "learning_rate": 6.969847662949336e-07, "loss": 0.0241, "step": 18990 }, { "grad_norm": 0.07816425710916519, "learning_rate": 6.832962649622798e-07, "loss": 0.0217, "step": 19000 }, { "grad_norm": 0.08738447725772858, "learning_rate": 6.697425936144863e-07, "loss": 0.0216, "step": 19010 }, { "grad_norm": 0.07450190931558609, "learning_rate": 6.563237893067731e-07, "loss": 0.0225, "step": 19020 }, { "grad_norm": 0.08284712582826614, "learning_rate": 6.430398887256328e-07, "loss": 0.0246, "step": 19030 }, { "grad_norm": 0.0863705575466156, "learning_rate": 6.298909281887478e-07, "loss": 0.0258, "step": 19040 }, { "grad_norm": 0.0696210041642189, "learning_rate": 6.168769436448673e-07, "loss": 0.0225, "step": 19050 }, { "grad_norm": 0.059944313019514084, "learning_rate": 6.03997970673742e-07, "loss": 0.0228, "step": 19060 }, { "grad_norm": 0.07013697177171707, "learning_rate": 5.912540444859782e-07, "loss": 0.0227, "step": 19070 }, { "grad_norm": 0.095664843916893, "learning_rate": 5.786451999229837e-07, "loss": 0.0258, "step": 19080 }, { "grad_norm": 0.07465808838605881, "learning_rate": 5.661714714568722e-07, "loss": 0.0229, "step": 19090 }, { "grad_norm": 0.08802983164787292, "learning_rate": 5.538328931903259e-07, "loss": 0.0229, "step": 19100 }, { "grad_norm": 0.08487019687891006, "learning_rate": 5.416294988565551e-07, "loss": 0.0228, "step": 19110 }, { "grad_norm": 0.07133125513792038, "learning_rate": 5.29561321819172e-07, "loss": 0.0203, "step": 19120 }, { "grad_norm": 0.084416463971138, "learning_rate": 5.176283950721061e-07, "loss": 0.0218, "step": 19130 }, { "grad_norm": 0.07942307740449905, "learning_rate": 5.058307512395332e-07, "loss": 0.0233, "step": 19140 }, { "grad_norm": 0.09991355240345001, "learning_rate": 4.941684225757526e-07, "loss": 0.0246, "step": 19150 }, { "grad_norm": 0.08930649608373642, "learning_rate": 4.826414409651314e-07, "loss": 0.0224, "step": 19160 }, { "grad_norm": 0.08521349728107452, "learning_rate": 4.712498379219943e-07, "loss": 0.0244, "step": 19170 }, { "grad_norm": 0.07980799674987793, "learning_rate": 4.599936445905506e-07, "loss": 0.0224, "step": 19180 }, { "grad_norm": 0.06415129452943802, "learning_rate": 4.4887289174480594e-07, "loss": 0.0205, "step": 19190 }, { "grad_norm": 0.0766814798116684, "learning_rate": 4.378876097884621e-07, "loss": 0.0245, "step": 19200 }, { "grad_norm": 0.082685686647892, "learning_rate": 4.2703782875487264e-07, "loss": 0.0217, "step": 19210 }, { "grad_norm": 0.08632442355155945, "learning_rate": 4.163235783069208e-07, "loss": 0.0228, "step": 19220 }, { "grad_norm": 0.06868918240070343, "learning_rate": 4.057448877369585e-07, "loss": 0.0252, "step": 19230 }, { "grad_norm": 0.08034586161375046, "learning_rate": 3.9530178596672295e-07, "loss": 0.0226, "step": 19240 }, { "grad_norm": 0.07758335769176483, "learning_rate": 3.849943015472479e-07, "loss": 0.0231, "step": 19250 }, { "grad_norm": 0.11591295152902603, "learning_rate": 3.748224626588137e-07, "loss": 0.0242, "step": 19260 }, { "grad_norm": 0.10653503239154816, "learning_rate": 3.647862971108307e-07, "loss": 0.0242, "step": 19270 }, { "grad_norm": 0.19091148674488068, "learning_rate": 3.5488583234179473e-07, "loss": 0.0234, "step": 19280 }, { "grad_norm": 0.07622084766626358, "learning_rate": 3.4512109541920413e-07, "loss": 0.0229, "step": 19290 }, { "grad_norm": 0.09629981219768524, "learning_rate": 3.354921130394706e-07, "loss": 0.0245, "step": 19300 }, { "grad_norm": 0.0808778777718544, "learning_rate": 3.259989115278639e-07, "loss": 0.0247, "step": 19310 }, { "grad_norm": 0.1767803132534027, "learning_rate": 3.1664151683843403e-07, "loss": 0.0218, "step": 19320 }, { "grad_norm": 0.15766870975494385, "learning_rate": 3.074199545539447e-07, "loss": 0.0237, "step": 19330 }, { "grad_norm": 0.06397061049938202, "learning_rate": 2.983342498857955e-07, "loss": 0.0242, "step": 19340 }, { "grad_norm": 0.07286729663610458, "learning_rate": 2.893844276739499e-07, "loss": 0.0236, "step": 19350 }, { "grad_norm": 0.08929965645074844, "learning_rate": 2.8057051238688514e-07, "loss": 0.0214, "step": 19360 }, { "grad_norm": 0.07660838961601257, "learning_rate": 2.71892528121509e-07, "loss": 0.0242, "step": 19370 }, { "grad_norm": 0.06558053940534592, "learning_rate": 2.633504986030988e-07, "loss": 0.0245, "step": 19380 }, { "grad_norm": 0.12542490661144257, "learning_rate": 2.549444471852347e-07, "loss": 0.0219, "step": 19390 }, { "grad_norm": 0.072273388504982, "learning_rate": 2.4667439684974423e-07, "loss": 0.0219, "step": 19400 }, { "grad_norm": 0.06846167892217636, "learning_rate": 2.3854037020662467e-07, "loss": 0.0229, "step": 19410 }, { "grad_norm": 0.09123676270246506, "learning_rate": 2.3054238949399288e-07, "loss": 0.0245, "step": 19420 }, { "grad_norm": 0.10553047060966492, "learning_rate": 2.2268047657802993e-07, "loss": 0.0216, "step": 19430 }, { "grad_norm": 0.06332673132419586, "learning_rate": 2.149546529529034e-07, "loss": 0.0242, "step": 19440 }, { "grad_norm": 0.08394449949264526, "learning_rate": 2.0736493974071736e-07, "loss": 0.0224, "step": 19450 }, { "grad_norm": 0.06405133754014969, "learning_rate": 1.9991135769145686e-07, "loss": 0.0226, "step": 19460 }, { "grad_norm": 0.06550749391317368, "learning_rate": 1.9259392718293245e-07, "loss": 0.0232, "step": 19470 }, { "grad_norm": 0.08512809872627258, "learning_rate": 1.8541266822072467e-07, "loss": 0.0219, "step": 19480 }, { "grad_norm": 0.06339780986309052, "learning_rate": 1.7836760043811184e-07, "loss": 0.0202, "step": 19490 }, { "grad_norm": 0.06605102121829987, "learning_rate": 1.7145874309604792e-07, "loss": 0.0252, "step": 19500 }, { "grad_norm": 0.07945529371500015, "learning_rate": 1.6468611508308474e-07, "loss": 0.0227, "step": 19510 }, { "grad_norm": 0.12295973300933838, "learning_rate": 1.5804973491532204e-07, "loss": 0.0235, "step": 19520 }, { "grad_norm": 0.060020361095666885, "learning_rate": 1.5154962073637424e-07, "loss": 0.0218, "step": 19530 }, { "grad_norm": 0.11446372419595718, "learning_rate": 1.4518579031730372e-07, "loss": 0.0219, "step": 19540 }, { "grad_norm": 0.06633087247610092, "learning_rate": 1.389582610565876e-07, "loss": 0.0215, "step": 19550 }, { "grad_norm": 0.07763561606407166, "learning_rate": 1.3286704998003995e-07, "loss": 0.0222, "step": 19560 }, { "grad_norm": 0.09211281687021255, "learning_rate": 1.2691217374080632e-07, "loss": 0.0226, "step": 19570 }, { "grad_norm": 0.08439243584871292, "learning_rate": 1.2109364861929705e-07, "loss": 0.0232, "step": 19580 }, { "grad_norm": 0.10579697787761688, "learning_rate": 1.1541149052312628e-07, "loss": 0.025, "step": 19590 }, { "grad_norm": 0.10882402956485748, "learning_rate": 1.0986571498710074e-07, "loss": 0.0238, "step": 19600 }, { "grad_norm": 0.10885309427976608, "learning_rate": 1.0445633717316438e-07, "loss": 0.0249, "step": 19610 }, { "grad_norm": 0.16508835554122925, "learning_rate": 9.918337187034277e-08, "loss": 0.0236, "step": 19620 }, { "grad_norm": 0.08304081112146378, "learning_rate": 9.404683349472643e-08, "loss": 0.0249, "step": 19630 }, { "grad_norm": 0.09056566655635834, "learning_rate": 8.904673608940983e-08, "loss": 0.0244, "step": 19640 }, { "grad_norm": 0.0901278555393219, "learning_rate": 8.418309332447471e-08, "loss": 0.0209, "step": 19650 }, { "grad_norm": 0.10605024546384811, "learning_rate": 7.945591849692902e-08, "loss": 0.0272, "step": 19660 }, { "grad_norm": 0.08081994205713272, "learning_rate": 7.486522453069578e-08, "loss": 0.0249, "step": 19670 }, { "grad_norm": 0.06961613893508911, "learning_rate": 7.041102397655208e-08, "loss": 0.0244, "step": 19680 }, { "grad_norm": 0.06830307096242905, "learning_rate": 6.609332901210685e-08, "loss": 0.0247, "step": 19690 }, { "grad_norm": 0.06158261373639107, "learning_rate": 6.191215144178419e-08, "loss": 0.0227, "step": 19700 }, { "grad_norm": 0.09181398153305054, "learning_rate": 5.786750269675678e-08, "loss": 0.0262, "step": 19710 }, { "grad_norm": 0.07624372094869614, "learning_rate": 5.395939383494031e-08, "loss": 0.0228, "step": 19720 }, { "grad_norm": 0.06973942369222641, "learning_rate": 5.018783554095463e-08, "loss": 0.0246, "step": 19730 }, { "grad_norm": 0.08752191811800003, "learning_rate": 4.655283812610156e-08, "loss": 0.0218, "step": 19740 }, { "grad_norm": 0.07653931528329849, "learning_rate": 4.305441152831491e-08, "loss": 0.0197, "step": 19750 }, { "grad_norm": 0.06155616417527199, "learning_rate": 3.9692565312171584e-08, "loss": 0.0222, "step": 19760 }, { "grad_norm": 0.08457348495721817, "learning_rate": 3.6467308668824975e-08, "loss": 0.0241, "step": 19770 }, { "grad_norm": 0.06177319586277008, "learning_rate": 3.3378650416004964e-08, "loss": 0.0219, "step": 19780 }, { "grad_norm": 0.07499980926513672, "learning_rate": 3.042659899797906e-08, "loss": 0.0261, "step": 19790 }, { "grad_norm": 0.0788775235414505, "learning_rate": 2.76111624855524e-08, "loss": 0.0222, "step": 19800 }, { "grad_norm": 0.06092666834592819, "learning_rate": 2.4932348576017784e-08, "loss": 0.024, "step": 19810 }, { "grad_norm": 0.07280290126800537, "learning_rate": 2.239016459314458e-08, "loss": 0.0222, "step": 19820 }, { "grad_norm": 0.07072023302316666, "learning_rate": 1.9984617487173174e-08, "loss": 0.0244, "step": 19830 }, { "grad_norm": 0.056819792836904526, "learning_rate": 1.7715713834776105e-08, "loss": 0.0218, "step": 19840 }, { "grad_norm": 0.08554646372795105, "learning_rate": 1.5583459839046964e-08, "loss": 0.027, "step": 19850 }, { "grad_norm": 0.06952627748250961, "learning_rate": 1.3587861329489304e-08, "loss": 0.0224, "step": 19860 }, { "grad_norm": 0.12072184681892395, "learning_rate": 1.1728923761994415e-08, "loss": 0.0239, "step": 19870 }, { "grad_norm": 0.07015801221132278, "learning_rate": 1.0006652218819135e-08, "loss": 0.0219, "step": 19880 }, { "grad_norm": 0.07342763990163803, "learning_rate": 8.421051408596947e-09, "loss": 0.0233, "step": 19890 }, { "grad_norm": 0.09631600230932236, "learning_rate": 6.972125666299123e-09, "loss": 0.0251, "step": 19900 }, { "grad_norm": 0.08711550384759903, "learning_rate": 5.659878953229169e-09, "loss": 0.0241, "step": 19910 }, { "grad_norm": 0.0909322202205658, "learning_rate": 4.48431485701728e-09, "loss": 0.0232, "step": 19920 }, { "grad_norm": 0.08080588281154633, "learning_rate": 3.4454365916203322e-09, "loss": 0.0236, "step": 19930 }, { "grad_norm": 0.06317497044801712, "learning_rate": 2.5432469972830332e-09, "loss": 0.025, "step": 19940 }, { "grad_norm": 0.08788690716028214, "learning_rate": 1.7777485405601203e-09, "loss": 0.0236, "step": 19950 }, { "grad_norm": 0.054821401834487915, "learning_rate": 1.1489433142941597e-09, "loss": 0.0242, "step": 19960 }, { "grad_norm": 0.07889028638601303, "learning_rate": 6.568330376210963e-10, "loss": 0.0217, "step": 19970 }, { "grad_norm": 0.06963111460208893, "learning_rate": 3.0141905594249787e-10, "loss": 0.0221, "step": 19980 }, { "grad_norm": 0.06058225408196449, "learning_rate": 8.270234094776008e-11, "loss": 0.0237, "step": 19990 }, { "grad_norm": 0.08581823855638504, "learning_rate": 6.834906085551041e-13, "loss": 0.0252, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 4000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }