{ "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": 0.9539716839790344, "learning_rate": 9e-07, "loss": 1.2399, "step": 10 }, { "grad_norm": 1.2878984212875366, "learning_rate": 1.9e-06, "loss": 1.2235, "step": 20 }, { "grad_norm": 0.36264777183532715, "learning_rate": 2.9e-06, "loss": 1.22, "step": 30 }, { "grad_norm": 0.5637423396110535, "learning_rate": 3.9e-06, "loss": 1.1656, "step": 40 }, { "grad_norm": 0.3330078125, "learning_rate": 4.9000000000000005e-06, "loss": 1.2042, "step": 50 }, { "grad_norm": 0.2072010189294815, "learning_rate": 5.9e-06, "loss": 1.2139, "step": 60 }, { "grad_norm": 0.8359606266021729, "learning_rate": 6.900000000000001e-06, "loss": 1.1648, "step": 70 }, { "grad_norm": 1.3386600017547607, "learning_rate": 7.9e-06, "loss": 1.0778, "step": 80 }, { "grad_norm": 0.891655445098877, "learning_rate": 8.9e-06, "loss": 1.0763, "step": 90 }, { "grad_norm": 1.3295084238052368, "learning_rate": 9.900000000000002e-06, "loss": 1.0603, "step": 100 }, { "grad_norm": 1.0589511394500732, "learning_rate": 1.09e-05, "loss": 1.062, "step": 110 }, { "grad_norm": 0.7706884741783142, "learning_rate": 1.19e-05, "loss": 1.0382, "step": 120 }, { "grad_norm": 3.3254613876342773, "learning_rate": 1.29e-05, "loss": 0.9469, "step": 130 }, { "grad_norm": 3.730990171432495, "learning_rate": 1.3900000000000002e-05, "loss": 0.7897, "step": 140 }, { "grad_norm": 2.3984339237213135, "learning_rate": 1.49e-05, "loss": 0.5844, "step": 150 }, { "grad_norm": 4.606634140014648, "learning_rate": 1.59e-05, "loss": 0.4534, "step": 160 }, { "grad_norm": 3.5254385471343994, "learning_rate": 1.69e-05, "loss": 0.3486, "step": 170 }, { "grad_norm": 3.0136377811431885, "learning_rate": 1.79e-05, "loss": 0.2602, "step": 180 }, { "grad_norm": 2.514943838119507, "learning_rate": 1.8900000000000002e-05, "loss": 0.2281, "step": 190 }, { "grad_norm": 2.7475783824920654, "learning_rate": 1.9900000000000003e-05, "loss": 0.1841, "step": 200 }, { "grad_norm": 3.1382977962493896, "learning_rate": 2.09e-05, "loss": 0.1734, "step": 210 }, { "grad_norm": 2.4567947387695312, "learning_rate": 2.19e-05, "loss": 0.1583, "step": 220 }, { "grad_norm": 2.9212474822998047, "learning_rate": 2.29e-05, "loss": 0.1511, "step": 230 }, { "grad_norm": 1.8118640184402466, "learning_rate": 2.39e-05, "loss": 0.1378, "step": 240 }, { "grad_norm": 2.218918561935425, "learning_rate": 2.4900000000000002e-05, "loss": 0.1334, "step": 250 }, { "grad_norm": 1.8789929151535034, "learning_rate": 2.5900000000000003e-05, "loss": 0.1266, "step": 260 }, { "grad_norm": 2.095660924911499, "learning_rate": 2.6900000000000003e-05, "loss": 0.1245, "step": 270 }, { "grad_norm": 1.8253940343856812, "learning_rate": 2.7900000000000004e-05, "loss": 0.1251, "step": 280 }, { "grad_norm": 1.2256525754928589, "learning_rate": 2.8899999999999998e-05, "loss": 0.13, "step": 290 }, { "grad_norm": 2.1088387966156006, "learning_rate": 2.9900000000000002e-05, "loss": 0.1281, "step": 300 }, { "grad_norm": 1.6173534393310547, "learning_rate": 3.09e-05, "loss": 0.1218, "step": 310 }, { "grad_norm": 1.9908654689788818, "learning_rate": 3.19e-05, "loss": 0.1106, "step": 320 }, { "grad_norm": 1.8997098207473755, "learning_rate": 3.29e-05, "loss": 0.1166, "step": 330 }, { "grad_norm": 2.013174295425415, "learning_rate": 3.3900000000000004e-05, "loss": 0.1161, "step": 340 }, { "grad_norm": 1.4469252824783325, "learning_rate": 3.49e-05, "loss": 0.1086, "step": 350 }, { "grad_norm": 1.395389437675476, "learning_rate": 3.59e-05, "loss": 0.1087, "step": 360 }, { "grad_norm": 2.150270938873291, "learning_rate": 3.69e-05, "loss": 0.1018, "step": 370 }, { "grad_norm": 1.6811128854751587, "learning_rate": 3.79e-05, "loss": 0.1022, "step": 380 }, { "grad_norm": 1.2427788972854614, "learning_rate": 3.8900000000000004e-05, "loss": 0.1044, "step": 390 }, { "grad_norm": 1.766543984413147, "learning_rate": 3.99e-05, "loss": 0.106, "step": 400 }, { "grad_norm": 1.5145665407180786, "learning_rate": 4.09e-05, "loss": 0.1044, "step": 410 }, { "grad_norm": 1.704944372177124, "learning_rate": 4.19e-05, "loss": 0.0967, "step": 420 }, { "grad_norm": 1.7721291780471802, "learning_rate": 4.29e-05, "loss": 0.0984, "step": 430 }, { "grad_norm": 1.0307819843292236, "learning_rate": 4.39e-05, "loss": 0.0934, "step": 440 }, { "grad_norm": 1.3351352214813232, "learning_rate": 4.49e-05, "loss": 0.0894, "step": 450 }, { "grad_norm": 0.901075005531311, "learning_rate": 4.5900000000000004e-05, "loss": 0.0813, "step": 460 }, { "grad_norm": 0.967912495136261, "learning_rate": 4.69e-05, "loss": 0.0855, "step": 470 }, { "grad_norm": 1.1338616609573364, "learning_rate": 4.79e-05, "loss": 0.0869, "step": 480 }, { "grad_norm": 1.4671498537063599, "learning_rate": 4.89e-05, "loss": 0.0843, "step": 490 }, { "grad_norm": 0.911688506603241, "learning_rate": 4.99e-05, "loss": 0.0911, "step": 500 }, { "grad_norm": 1.407004714012146, "learning_rate": 5.0900000000000004e-05, "loss": 0.087, "step": 510 }, { "grad_norm": 0.9018402099609375, "learning_rate": 5.19e-05, "loss": 0.083, "step": 520 }, { "grad_norm": 1.5423208475112915, "learning_rate": 5.2900000000000005e-05, "loss": 0.0903, "step": 530 }, { "grad_norm": 1.228566288948059, "learning_rate": 5.390000000000001e-05, "loss": 0.0856, "step": 540 }, { "grad_norm": 1.0500738620758057, "learning_rate": 5.4900000000000006e-05, "loss": 0.0768, "step": 550 }, { "grad_norm": 1.0046727657318115, "learning_rate": 5.590000000000001e-05, "loss": 0.0854, "step": 560 }, { "grad_norm": 0.8127021789550781, "learning_rate": 5.69e-05, "loss": 0.0826, "step": 570 }, { "grad_norm": 1.7604862451553345, "learning_rate": 5.79e-05, "loss": 0.0817, "step": 580 }, { "grad_norm": 0.819057047367096, "learning_rate": 5.89e-05, "loss": 0.0783, "step": 590 }, { "grad_norm": 0.6968788504600525, "learning_rate": 5.99e-05, "loss": 0.0743, "step": 600 }, { "grad_norm": 0.826816201210022, "learning_rate": 6.09e-05, "loss": 0.0771, "step": 610 }, { "grad_norm": 1.1997026205062866, "learning_rate": 6.19e-05, "loss": 0.0778, "step": 620 }, { "grad_norm": 1.1628166437149048, "learning_rate": 6.29e-05, "loss": 0.0762, "step": 630 }, { "grad_norm": 1.3362361192703247, "learning_rate": 6.390000000000001e-05, "loss": 0.0777, "step": 640 }, { "grad_norm": 0.6102016568183899, "learning_rate": 6.49e-05, "loss": 0.0794, "step": 650 }, { "grad_norm": 1.0114165544509888, "learning_rate": 6.59e-05, "loss": 0.0729, "step": 660 }, { "grad_norm": 1.007080316543579, "learning_rate": 6.690000000000001e-05, "loss": 0.0706, "step": 670 }, { "grad_norm": 0.8234179019927979, "learning_rate": 6.790000000000001e-05, "loss": 0.0713, "step": 680 }, { "grad_norm": 1.3303985595703125, "learning_rate": 6.89e-05, "loss": 0.0711, "step": 690 }, { "grad_norm": 0.8679249286651611, "learning_rate": 6.99e-05, "loss": 0.0721, "step": 700 }, { "grad_norm": 0.9954879879951477, "learning_rate": 7.09e-05, "loss": 0.0752, "step": 710 }, { "grad_norm": 0.7408309578895569, "learning_rate": 7.19e-05, "loss": 0.0738, "step": 720 }, { "grad_norm": 0.7630719542503357, "learning_rate": 7.29e-05, "loss": 0.0745, "step": 730 }, { "grad_norm": 1.0726048946380615, "learning_rate": 7.390000000000001e-05, "loss": 0.0845, "step": 740 }, { "grad_norm": 0.5797054171562195, "learning_rate": 7.49e-05, "loss": 0.0806, "step": 750 }, { "grad_norm": 0.7958113551139832, "learning_rate": 7.59e-05, "loss": 0.0743, "step": 760 }, { "grad_norm": 0.7454976439476013, "learning_rate": 7.69e-05, "loss": 0.0772, "step": 770 }, { "grad_norm": 0.6472405791282654, "learning_rate": 7.790000000000001e-05, "loss": 0.0769, "step": 780 }, { "grad_norm": 0.6711792945861816, "learning_rate": 7.890000000000001e-05, "loss": 0.0705, "step": 790 }, { "grad_norm": 0.8817596435546875, "learning_rate": 7.99e-05, "loss": 0.0651, "step": 800 }, { "grad_norm": 0.5537639260292053, "learning_rate": 8.090000000000001e-05, "loss": 0.0628, "step": 810 }, { "grad_norm": 0.6646618843078613, "learning_rate": 8.19e-05, "loss": 0.0688, "step": 820 }, { "grad_norm": 0.4053765535354614, "learning_rate": 8.29e-05, "loss": 0.0671, "step": 830 }, { "grad_norm": 1.0079501867294312, "learning_rate": 8.39e-05, "loss": 0.069, "step": 840 }, { "grad_norm": 0.6765267848968506, "learning_rate": 8.49e-05, "loss": 0.0608, "step": 850 }, { "grad_norm": 0.6772356629371643, "learning_rate": 8.59e-05, "loss": 0.0623, "step": 860 }, { "grad_norm": 0.5116077661514282, "learning_rate": 8.69e-05, "loss": 0.0636, "step": 870 }, { "grad_norm": 0.49881091713905334, "learning_rate": 8.790000000000001e-05, "loss": 0.0689, "step": 880 }, { "grad_norm": 0.6379665732383728, "learning_rate": 8.89e-05, "loss": 0.0651, "step": 890 }, { "grad_norm": 0.7225998640060425, "learning_rate": 8.99e-05, "loss": 0.0712, "step": 900 }, { "grad_norm": 0.7093841433525085, "learning_rate": 9.090000000000001e-05, "loss": 0.0631, "step": 910 }, { "grad_norm": 0.9616966247558594, "learning_rate": 9.190000000000001e-05, "loss": 0.0683, "step": 920 }, { "grad_norm": 0.46344366669654846, "learning_rate": 9.290000000000001e-05, "loss": 0.0666, "step": 930 }, { "grad_norm": 0.6637529730796814, "learning_rate": 9.39e-05, "loss": 0.0622, "step": 940 }, { "grad_norm": 0.5119460225105286, "learning_rate": 9.49e-05, "loss": 0.0637, "step": 950 }, { "grad_norm": 0.6212114691734314, "learning_rate": 9.59e-05, "loss": 0.066, "step": 960 }, { "grad_norm": 0.714377224445343, "learning_rate": 9.69e-05, "loss": 0.0639, "step": 970 }, { "grad_norm": 0.5530987977981567, "learning_rate": 9.790000000000001e-05, "loss": 0.0688, "step": 980 }, { "grad_norm": 0.7494563460350037, "learning_rate": 9.89e-05, "loss": 0.0707, "step": 990 }, { "grad_norm": 0.7097394466400146, "learning_rate": 9.99e-05, "loss": 0.0612, "step": 1000 }, { "grad_norm": 0.5310402512550354, "learning_rate": 9.999994463727085e-05, "loss": 0.0557, "step": 1010 }, { "grad_norm": 0.6620810031890869, "learning_rate": 9.999975326009292e-05, "loss": 0.0635, "step": 1020 }, { "grad_norm": 0.6903439164161682, "learning_rate": 9.999942518549879e-05, "loss": 0.0526, "step": 1030 }, { "grad_norm": 0.4423273205757141, "learning_rate": 9.999896041438544e-05, "loss": 0.0516, "step": 1040 }, { "grad_norm": 0.5590900778770447, "learning_rate": 9.999835894802353e-05, "loss": 0.0562, "step": 1050 }, { "grad_norm": 0.6998360753059387, "learning_rate": 9.999762078805743e-05, "loss": 0.0591, "step": 1060 }, { "grad_norm": 0.6909574866294861, "learning_rate": 9.999674593650526e-05, "loss": 0.0533, "step": 1070 }, { "grad_norm": 0.6840800642967224, "learning_rate": 9.99957343957588e-05, "loss": 0.0588, "step": 1080 }, { "grad_norm": 0.9030687212944031, "learning_rate": 9.99945861685836e-05, "loss": 0.052, "step": 1090 }, { "grad_norm": 0.6525619029998779, "learning_rate": 9.999330125811884e-05, "loss": 0.0562, "step": 1100 }, { "grad_norm": 0.8846177458763123, "learning_rate": 9.999187966787744e-05, "loss": 0.0533, "step": 1110 }, { "grad_norm": 0.6119235157966614, "learning_rate": 9.999032140174595e-05, "loss": 0.049, "step": 1120 }, { "grad_norm": 0.4912249743938446, "learning_rate": 9.998862646398464e-05, "loss": 0.0546, "step": 1130 }, { "grad_norm": 0.5428407192230225, "learning_rate": 9.998679485922739e-05, "loss": 0.0544, "step": 1140 }, { "grad_norm": 0.8638484477996826, "learning_rate": 9.998482659248174e-05, "loss": 0.0504, "step": 1150 }, { "grad_norm": 0.4278607964515686, "learning_rate": 9.998272166912883e-05, "loss": 0.0531, "step": 1160 }, { "grad_norm": 0.7931933403015137, "learning_rate": 9.998048009492347e-05, "loss": 0.0602, "step": 1170 }, { "grad_norm": 0.4066477417945862, "learning_rate": 9.997810187599403e-05, "loss": 0.0574, "step": 1180 }, { "grad_norm": 0.6033307313919067, "learning_rate": 9.997558701884249e-05, "loss": 0.0486, "step": 1190 }, { "grad_norm": 0.7459061145782471, "learning_rate": 9.997293553034433e-05, "loss": 0.0547, "step": 1200 }, { "grad_norm": 0.6617655754089355, "learning_rate": 9.997014741774866e-05, "loss": 0.0531, "step": 1210 }, { "grad_norm": 0.4468367397785187, "learning_rate": 9.996722268867803e-05, "loss": 0.0559, "step": 1220 }, { "grad_norm": 0.3588188588619232, "learning_rate": 9.996416135112858e-05, "loss": 0.0533, "step": 1230 }, { "grad_norm": 0.5053278803825378, "learning_rate": 9.996096341346988e-05, "loss": 0.0476, "step": 1240 }, { "grad_norm": 0.613800585269928, "learning_rate": 9.995762888444495e-05, "loss": 0.0499, "step": 1250 }, { "grad_norm": 0.3598322570323944, "learning_rate": 9.995415777317027e-05, "loss": 0.0449, "step": 1260 }, { "grad_norm": 0.5255052447319031, "learning_rate": 9.995055008913574e-05, "loss": 0.0466, "step": 1270 }, { "grad_norm": 0.4071582555770874, "learning_rate": 9.994680584220463e-05, "loss": 0.0476, "step": 1280 }, { "grad_norm": 0.3095243573188782, "learning_rate": 9.994292504261355e-05, "loss": 0.0491, "step": 1290 }, { "grad_norm": 0.5441323518753052, "learning_rate": 9.993890770097247e-05, "loss": 0.0539, "step": 1300 }, { "grad_norm": 0.7022833228111267, "learning_rate": 9.993475382826467e-05, "loss": 0.0523, "step": 1310 }, { "grad_norm": 0.5823666453361511, "learning_rate": 9.993046343584664e-05, "loss": 0.0458, "step": 1320 }, { "grad_norm": 0.529151439666748, "learning_rate": 9.992603653544816e-05, "loss": 0.0541, "step": 1330 }, { "grad_norm": 0.5635100603103638, "learning_rate": 9.992147313917222e-05, "loss": 0.0517, "step": 1340 }, { "grad_norm": 0.5174548625946045, "learning_rate": 9.991677325949497e-05, "loss": 0.0521, "step": 1350 }, { "grad_norm": 0.6938949227333069, "learning_rate": 9.991193690926568e-05, "loss": 0.0492, "step": 1360 }, { "grad_norm": 0.4387352466583252, "learning_rate": 9.990696410170678e-05, "loss": 0.0494, "step": 1370 }, { "grad_norm": 0.5664272308349609, "learning_rate": 9.990185485041371e-05, "loss": 0.0463, "step": 1380 }, { "grad_norm": 0.3515314757823944, "learning_rate": 9.989660916935498e-05, "loss": 0.0441, "step": 1390 }, { "grad_norm": 0.46024447679519653, "learning_rate": 9.989122707287208e-05, "loss": 0.0439, "step": 1400 }, { "grad_norm": 0.5430362224578857, "learning_rate": 9.988570857567945e-05, "loss": 0.0404, "step": 1410 }, { "grad_norm": 0.4762286841869354, "learning_rate": 9.988005369286446e-05, "loss": 0.0478, "step": 1420 }, { "grad_norm": 0.5672691464424133, "learning_rate": 9.987426243988734e-05, "loss": 0.05, "step": 1430 }, { "grad_norm": 0.5396414995193481, "learning_rate": 9.986833483258114e-05, "loss": 0.0585, "step": 1440 }, { "grad_norm": 0.35184526443481445, "learning_rate": 9.986227088715173e-05, "loss": 0.0479, "step": 1450 }, { "grad_norm": 0.46804168820381165, "learning_rate": 9.98560706201777e-05, "loss": 0.0531, "step": 1460 }, { "grad_norm": 0.4966842532157898, "learning_rate": 9.984973404861036e-05, "loss": 0.0499, "step": 1470 }, { "grad_norm": 0.5503523349761963, "learning_rate": 9.984326118977361e-05, "loss": 0.0453, "step": 1480 }, { "grad_norm": 0.39526286721229553, "learning_rate": 9.983665206136406e-05, "loss": 0.0451, "step": 1490 }, { "grad_norm": 0.7096656560897827, "learning_rate": 9.982990668145075e-05, "loss": 0.043, "step": 1500 }, { "grad_norm": 0.8284592032432556, "learning_rate": 9.982302506847534e-05, "loss": 0.0495, "step": 1510 }, { "grad_norm": 0.7668936848640442, "learning_rate": 9.981600724125189e-05, "loss": 0.0474, "step": 1520 }, { "grad_norm": 0.4689315855503082, "learning_rate": 9.980885321896685e-05, "loss": 0.0475, "step": 1530 }, { "grad_norm": 0.5250937938690186, "learning_rate": 9.980156302117905e-05, "loss": 0.0459, "step": 1540 }, { "grad_norm": 0.3923928141593933, "learning_rate": 9.979413666781963e-05, "loss": 0.0453, "step": 1550 }, { "grad_norm": 0.4199478328227997, "learning_rate": 9.978657417919193e-05, "loss": 0.0489, "step": 1560 }, { "grad_norm": 0.6422154307365417, "learning_rate": 9.977887557597153e-05, "loss": 0.0441, "step": 1570 }, { "grad_norm": 0.37972068786621094, "learning_rate": 9.97710408792061e-05, "loss": 0.0414, "step": 1580 }, { "grad_norm": 0.3207695484161377, "learning_rate": 9.976307011031542e-05, "loss": 0.0394, "step": 1590 }, { "grad_norm": 0.6480545997619629, "learning_rate": 9.975496329109126e-05, "loss": 0.0457, "step": 1600 }, { "grad_norm": 0.5391230583190918, "learning_rate": 9.974672044369732e-05, "loss": 0.0423, "step": 1610 }, { "grad_norm": 0.33618906140327454, "learning_rate": 9.97383415906693e-05, "loss": 0.0423, "step": 1620 }, { "grad_norm": 0.3441476821899414, "learning_rate": 9.97298267549146e-05, "loss": 0.0466, "step": 1630 }, { "grad_norm": 0.49657508730888367, "learning_rate": 9.972117595971249e-05, "loss": 0.0429, "step": 1640 }, { "grad_norm": 0.30471065640449524, "learning_rate": 9.971238922871391e-05, "loss": 0.0385, "step": 1650 }, { "grad_norm": 0.26801300048828125, "learning_rate": 9.970346658594142e-05, "loss": 0.0365, "step": 1660 }, { "grad_norm": 0.5888593196868896, "learning_rate": 9.969440805578923e-05, "loss": 0.0357, "step": 1670 }, { "grad_norm": 0.33331847190856934, "learning_rate": 9.968521366302298e-05, "loss": 0.0422, "step": 1680 }, { "grad_norm": 0.4909774959087372, "learning_rate": 9.967588343277981e-05, "loss": 0.0496, "step": 1690 }, { "grad_norm": 0.41679391264915466, "learning_rate": 9.966641739056818e-05, "loss": 0.0444, "step": 1700 }, { "grad_norm": 0.39262354373931885, "learning_rate": 9.965681556226793e-05, "loss": 0.0422, "step": 1710 }, { "grad_norm": 0.480535626411438, "learning_rate": 9.964707797413006e-05, "loss": 0.0421, "step": 1720 }, { "grad_norm": 0.43621885776519775, "learning_rate": 9.963720465277679e-05, "loss": 0.0407, "step": 1730 }, { "grad_norm": 1.2241843938827515, "learning_rate": 9.96271956252014e-05, "loss": 0.0386, "step": 1740 }, { "grad_norm": 0.4801706373691559, "learning_rate": 9.961705091876816e-05, "loss": 0.0493, "step": 1750 }, { "grad_norm": 0.3784044086933136, "learning_rate": 9.960677056121235e-05, "loss": 0.042, "step": 1760 }, { "grad_norm": 0.630437970161438, "learning_rate": 9.959635458064005e-05, "loss": 0.045, "step": 1770 }, { "grad_norm": 0.47157248854637146, "learning_rate": 9.958580300552815e-05, "loss": 0.0488, "step": 1780 }, { "grad_norm": 0.43877923488616943, "learning_rate": 9.957511586472426e-05, "loss": 0.0439, "step": 1790 }, { "grad_norm": 0.3307915925979614, "learning_rate": 9.956429318744662e-05, "loss": 0.0399, "step": 1800 }, { "grad_norm": 0.5732173919677734, "learning_rate": 9.955333500328404e-05, "loss": 0.0388, "step": 1810 }, { "grad_norm": 0.48091012239456177, "learning_rate": 9.95422413421957e-05, "loss": 0.0392, "step": 1820 }, { "grad_norm": 0.5842894911766052, "learning_rate": 9.953101223451133e-05, "loss": 0.0385, "step": 1830 }, { "grad_norm": 0.44385597109794617, "learning_rate": 9.951964771093085e-05, "loss": 0.0399, "step": 1840 }, { "grad_norm": 0.49984610080718994, "learning_rate": 9.950814780252442e-05, "loss": 0.0406, "step": 1850 }, { "grad_norm": 0.7289251089096069, "learning_rate": 9.949651254073236e-05, "loss": 0.0396, "step": 1860 }, { "grad_norm": 0.4685819447040558, "learning_rate": 9.948474195736504e-05, "loss": 0.0398, "step": 1870 }, { "grad_norm": 0.37194064259529114, "learning_rate": 9.947283608460277e-05, "loss": 0.0419, "step": 1880 }, { "grad_norm": 0.4372093975543976, "learning_rate": 9.946079495499577e-05, "loss": 0.0399, "step": 1890 }, { "grad_norm": 0.49088940024375916, "learning_rate": 9.944861860146401e-05, "loss": 0.0374, "step": 1900 }, { "grad_norm": 0.47523224353790283, "learning_rate": 9.943630705729719e-05, "loss": 0.0354, "step": 1910 }, { "grad_norm": 0.3566550016403198, "learning_rate": 9.942386035615459e-05, "loss": 0.0347, "step": 1920 }, { "grad_norm": 0.6321758031845093, "learning_rate": 9.941127853206503e-05, "loss": 0.0356, "step": 1930 }, { "grad_norm": 0.3624858260154724, "learning_rate": 9.939856161942673e-05, "loss": 0.0341, "step": 1940 }, { "grad_norm": 0.49837809801101685, "learning_rate": 9.938570965300724e-05, "loss": 0.038, "step": 1950 }, { "grad_norm": 0.6436309814453125, "learning_rate": 9.937272266794335e-05, "loss": 0.0407, "step": 1960 }, { "grad_norm": 0.37623298168182373, "learning_rate": 9.935960069974096e-05, "loss": 0.0471, "step": 1970 }, { "grad_norm": 0.40575772523880005, "learning_rate": 9.934634378427506e-05, "loss": 0.0401, "step": 1980 }, { "grad_norm": 0.3604365587234497, "learning_rate": 9.933295195778954e-05, "loss": 0.036, "step": 1990 }, { "grad_norm": 0.389886736869812, "learning_rate": 9.931942525689715e-05, "loss": 0.0356, "step": 2000 }, { "grad_norm": 0.27841874957084656, "learning_rate": 9.930576371857936e-05, "loss": 0.0328, "step": 2010 }, { "grad_norm": 0.4331666827201843, "learning_rate": 9.929196738018629e-05, "loss": 0.0373, "step": 2020 }, { "grad_norm": 0.5234594941139221, "learning_rate": 9.927803627943662e-05, "loss": 0.034, "step": 2030 }, { "grad_norm": 0.5398783087730408, "learning_rate": 9.926397045441744e-05, "loss": 0.0342, "step": 2040 }, { "grad_norm": 0.5524482131004333, "learning_rate": 9.924976994358417e-05, "loss": 0.0341, "step": 2050 }, { "grad_norm": 0.5834764242172241, "learning_rate": 9.923543478576048e-05, "loss": 0.0333, "step": 2060 }, { "grad_norm": 0.32027482986450195, "learning_rate": 9.922096502013813e-05, "loss": 0.0336, "step": 2070 }, { "grad_norm": 0.4293094277381897, "learning_rate": 9.92063606862769e-05, "loss": 0.0358, "step": 2080 }, { "grad_norm": 0.5508143901824951, "learning_rate": 9.919162182410453e-05, "loss": 0.0387, "step": 2090 }, { "grad_norm": 0.29482051730155945, "learning_rate": 9.917674847391645e-05, "loss": 0.0374, "step": 2100 }, { "grad_norm": 0.3904075622558594, "learning_rate": 9.916174067637584e-05, "loss": 0.0381, "step": 2110 }, { "grad_norm": 0.33555909991264343, "learning_rate": 9.914659847251348e-05, "loss": 0.0396, "step": 2120 }, { "grad_norm": 0.29747244715690613, "learning_rate": 9.913132190372753e-05, "loss": 0.0371, "step": 2130 }, { "grad_norm": 0.4222758710384369, "learning_rate": 9.911591101178359e-05, "loss": 0.0387, "step": 2140 }, { "grad_norm": 0.3572886288166046, "learning_rate": 9.910036583881443e-05, "loss": 0.0347, "step": 2150 }, { "grad_norm": 0.32537350058555603, "learning_rate": 9.908468642731995e-05, "loss": 0.0334, "step": 2160 }, { "grad_norm": 0.2405768483877182, "learning_rate": 9.906887282016707e-05, "loss": 0.033, "step": 2170 }, { "grad_norm": 0.29916712641716003, "learning_rate": 9.90529250605896e-05, "loss": 0.0372, "step": 2180 }, { "grad_norm": 0.2440979778766632, "learning_rate": 9.903684319218809e-05, "loss": 0.0301, "step": 2190 }, { "grad_norm": 0.29854586720466614, "learning_rate": 9.902062725892976e-05, "loss": 0.0368, "step": 2200 }, { "grad_norm": 0.36209118366241455, "learning_rate": 9.900427730514834e-05, "loss": 0.0342, "step": 2210 }, { "grad_norm": 0.5953738689422607, "learning_rate": 9.8987793375544e-05, "loss": 0.0387, "step": 2220 }, { "grad_norm": 0.43017303943634033, "learning_rate": 9.897117551518318e-05, "loss": 0.033, "step": 2230 }, { "grad_norm": 0.39768266677856445, "learning_rate": 9.895442376949844e-05, "loss": 0.0319, "step": 2240 }, { "grad_norm": 0.40631070733070374, "learning_rate": 9.893753818428845e-05, "loss": 0.0312, "step": 2250 }, { "grad_norm": 0.2722180187702179, "learning_rate": 9.892051880571773e-05, "loss": 0.0352, "step": 2260 }, { "grad_norm": 0.4332314133644104, "learning_rate": 9.890336568031663e-05, "loss": 0.0339, "step": 2270 }, { "grad_norm": 0.47213590145111084, "learning_rate": 9.888607885498113e-05, "loss": 0.0303, "step": 2280 }, { "grad_norm": 0.24052350223064423, "learning_rate": 9.886865837697275e-05, "loss": 0.0306, "step": 2290 }, { "grad_norm": 0.3359028995037079, "learning_rate": 9.88511042939184e-05, "loss": 0.0297, "step": 2300 }, { "grad_norm": 0.4439032971858978, "learning_rate": 9.883341665381028e-05, "loss": 0.0326, "step": 2310 }, { "grad_norm": 0.4808836579322815, "learning_rate": 9.881559550500575e-05, "loss": 0.0356, "step": 2320 }, { "grad_norm": 0.4477006196975708, "learning_rate": 9.879764089622712e-05, "loss": 0.04, "step": 2330 }, { "grad_norm": 0.31514281034469604, "learning_rate": 9.87795528765616e-05, "loss": 0.0325, "step": 2340 }, { "grad_norm": 0.5117734670639038, "learning_rate": 9.876133149546118e-05, "loss": 0.038, "step": 2350 }, { "grad_norm": 0.48117560148239136, "learning_rate": 9.874297680274238e-05, "loss": 0.032, "step": 2360 }, { "grad_norm": 0.3276621997356415, "learning_rate": 9.872448884858624e-05, "loss": 0.0315, "step": 2370 }, { "grad_norm": 0.43945470452308655, "learning_rate": 9.870586768353815e-05, "loss": 0.0326, "step": 2380 }, { "grad_norm": 0.32929229736328125, "learning_rate": 9.868711335850764e-05, "loss": 0.0345, "step": 2390 }, { "grad_norm": 0.5516417622566223, "learning_rate": 9.866822592476833e-05, "loss": 0.0357, "step": 2400 }, { "grad_norm": 0.40468400716781616, "learning_rate": 9.86492054339577e-05, "loss": 0.0319, "step": 2410 }, { "grad_norm": 0.25286388397216797, "learning_rate": 9.863005193807711e-05, "loss": 0.0345, "step": 2420 }, { "grad_norm": 0.32567211985588074, "learning_rate": 9.861076548949143e-05, "loss": 0.0323, "step": 2430 }, { "grad_norm": 0.2829413115978241, "learning_rate": 9.859134614092912e-05, "loss": 0.0308, "step": 2440 }, { "grad_norm": 0.3048607110977173, "learning_rate": 9.857179394548191e-05, "loss": 0.0299, "step": 2450 }, { "grad_norm": 0.5933793187141418, "learning_rate": 9.855210895660477e-05, "loss": 0.0302, "step": 2460 }, { "grad_norm": 0.44336336851119995, "learning_rate": 9.853229122811568e-05, "loss": 0.034, "step": 2470 }, { "grad_norm": 0.29292410612106323, "learning_rate": 9.851234081419559e-05, "loss": 0.0335, "step": 2480 }, { "grad_norm": 0.2657058835029602, "learning_rate": 9.849225776938814e-05, "loss": 0.0328, "step": 2490 }, { "grad_norm": 0.34056296944618225, "learning_rate": 9.847204214859964e-05, "loss": 0.0296, "step": 2500 }, { "grad_norm": 0.4818284511566162, "learning_rate": 9.845169400709879e-05, "loss": 0.0325, "step": 2510 }, { "grad_norm": 0.31733641028404236, "learning_rate": 9.843121340051664e-05, "loss": 0.0307, "step": 2520 }, { "grad_norm": 0.34739330410957336, "learning_rate": 9.841060038484641e-05, "loss": 0.034, "step": 2530 }, { "grad_norm": 0.28935590386390686, "learning_rate": 9.838985501644328e-05, "loss": 0.0317, "step": 2540 }, { "grad_norm": 0.34936049580574036, "learning_rate": 9.83689773520243e-05, "loss": 0.0334, "step": 2550 }, { "grad_norm": 0.2873668670654297, "learning_rate": 9.834796744866819e-05, "loss": 0.0342, "step": 2560 }, { "grad_norm": 0.2874118685722351, "learning_rate": 9.832682536381525e-05, "loss": 0.0333, "step": 2570 }, { "grad_norm": 0.2726486623287201, "learning_rate": 9.830555115526711e-05, "loss": 0.0314, "step": 2580 }, { "grad_norm": 0.25717464089393616, "learning_rate": 9.828414488118667e-05, "loss": 0.031, "step": 2590 }, { "grad_norm": 0.2881430983543396, "learning_rate": 9.826260660009785e-05, "loss": 0.0371, "step": 2600 }, { "grad_norm": 0.33533430099487305, "learning_rate": 9.824093637088547e-05, "loss": 0.0322, "step": 2610 }, { "grad_norm": 0.4128776788711548, "learning_rate": 9.821913425279514e-05, "loss": 0.0334, "step": 2620 }, { "grad_norm": 0.37056928873062134, "learning_rate": 9.8197200305433e-05, "loss": 0.033, "step": 2630 }, { "grad_norm": 0.5027377605438232, "learning_rate": 9.817513458876564e-05, "loss": 0.0308, "step": 2640 }, { "grad_norm": 0.30660566687583923, "learning_rate": 9.815293716311987e-05, "loss": 0.0307, "step": 2650 }, { "grad_norm": 0.4116220474243164, "learning_rate": 9.813060808918262e-05, "loss": 0.0342, "step": 2660 }, { "grad_norm": 0.26979532837867737, "learning_rate": 9.810814742800069e-05, "loss": 0.0284, "step": 2670 }, { "grad_norm": 0.36868610978126526, "learning_rate": 9.808555524098074e-05, "loss": 0.0281, "step": 2680 }, { "grad_norm": 0.3522500991821289, "learning_rate": 9.806283158988887e-05, "loss": 0.0315, "step": 2690 }, { "grad_norm": 0.31462499499320984, "learning_rate": 9.803997653685072e-05, "loss": 0.0284, "step": 2700 }, { "grad_norm": 0.29388993978500366, "learning_rate": 9.801699014435112e-05, "loss": 0.0287, "step": 2710 }, { "grad_norm": 0.34596309065818787, "learning_rate": 9.799387247523398e-05, "loss": 0.0317, "step": 2720 }, { "grad_norm": 0.26501259207725525, "learning_rate": 9.797062359270215e-05, "loss": 0.0326, "step": 2730 }, { "grad_norm": 0.3088298439979553, "learning_rate": 9.794724356031715e-05, "loss": 0.0307, "step": 2740 }, { "grad_norm": 0.4290851652622223, "learning_rate": 9.792373244199913e-05, "loss": 0.028, "step": 2750 }, { "grad_norm": 0.3980506956577301, "learning_rate": 9.790009030202658e-05, "loss": 0.0281, "step": 2760 }, { "grad_norm": 0.3892729878425598, "learning_rate": 9.78763172050362e-05, "loss": 0.0309, "step": 2770 }, { "grad_norm": 0.4693174958229065, "learning_rate": 9.785241321602274e-05, "loss": 0.0288, "step": 2780 }, { "grad_norm": 0.3513578176498413, "learning_rate": 9.782837840033879e-05, "loss": 0.0267, "step": 2790 }, { "grad_norm": 0.26329755783081055, "learning_rate": 9.780421282369461e-05, "loss": 0.0302, "step": 2800 }, { "grad_norm": 0.2563781440258026, "learning_rate": 9.777991655215797e-05, "loss": 0.033, "step": 2810 }, { "grad_norm": 0.4989384114742279, "learning_rate": 9.775548965215394e-05, "loss": 0.0336, "step": 2820 }, { "grad_norm": 0.484646201133728, "learning_rate": 9.773093219046474e-05, "loss": 0.0279, "step": 2830 }, { "grad_norm": 0.4296814799308777, "learning_rate": 9.770624423422954e-05, "loss": 0.0352, "step": 2840 }, { "grad_norm": 0.3954630494117737, "learning_rate": 9.768142585094426e-05, "loss": 0.0322, "step": 2850 }, { "grad_norm": 0.4004117250442505, "learning_rate": 9.765647710846142e-05, "loss": 0.0384, "step": 2860 }, { "grad_norm": 0.2551356554031372, "learning_rate": 9.763139807498991e-05, "loss": 0.0367, "step": 2870 }, { "grad_norm": 0.35151585936546326, "learning_rate": 9.760618881909487e-05, "loss": 0.0313, "step": 2880 }, { "grad_norm": 0.40542301535606384, "learning_rate": 9.758084940969744e-05, "loss": 0.0345, "step": 2890 }, { "grad_norm": 0.2810773551464081, "learning_rate": 9.755537991607459e-05, "loss": 0.0313, "step": 2900 }, { "grad_norm": 0.3146008849143982, "learning_rate": 9.752978040785895e-05, "loss": 0.027, "step": 2910 }, { "grad_norm": 0.3968224823474884, "learning_rate": 9.750405095503859e-05, "loss": 0.0345, "step": 2920 }, { "grad_norm": 0.42338794469833374, "learning_rate": 9.747819162795686e-05, "loss": 0.0344, "step": 2930 }, { "grad_norm": 0.46489864587783813, "learning_rate": 9.745220249731217e-05, "loss": 0.0369, "step": 2940 }, { "grad_norm": 0.5021902918815613, "learning_rate": 9.742608363415781e-05, "loss": 0.0321, "step": 2950 }, { "grad_norm": 0.3437994718551636, "learning_rate": 9.739983510990176e-05, "loss": 0.0319, "step": 2960 }, { "grad_norm": 0.3038621246814728, "learning_rate": 9.737345699630647e-05, "loss": 0.0288, "step": 2970 }, { "grad_norm": 0.21653184294700623, "learning_rate": 9.734694936548869e-05, "loss": 0.0288, "step": 2980 }, { "grad_norm": 0.4039348363876343, "learning_rate": 9.732031228991932e-05, "loss": 0.0254, "step": 2990 }, { "grad_norm": 0.38072285056114197, "learning_rate": 9.729354584242302e-05, "loss": 0.027, "step": 3000 }, { "grad_norm": 0.3224892020225525, "learning_rate": 9.726665009617832e-05, "loss": 0.0269, "step": 3010 }, { "grad_norm": 0.43210142850875854, "learning_rate": 9.723962512471714e-05, "loss": 0.0263, "step": 3020 }, { "grad_norm": 0.34511932730674744, "learning_rate": 9.72124710019247e-05, "loss": 0.0259, "step": 3030 }, { "grad_norm": 0.4183204174041748, "learning_rate": 9.718518780203934e-05, "loss": 0.0324, "step": 3040 }, { "grad_norm": 0.6096486449241638, "learning_rate": 9.715777559965228e-05, "loss": 0.0287, "step": 3050 }, { "grad_norm": 0.2685447633266449, "learning_rate": 9.713023446970746e-05, "loss": 0.0302, "step": 3060 }, { "grad_norm": 0.31860867142677307, "learning_rate": 9.710256448750126e-05, "loss": 0.0318, "step": 3070 }, { "grad_norm": 0.39600130915641785, "learning_rate": 9.707476572868235e-05, "loss": 0.0291, "step": 3080 }, { "grad_norm": 0.25682029128074646, "learning_rate": 9.704683826925149e-05, "loss": 0.0273, "step": 3090 }, { "grad_norm": 0.5371003746986389, "learning_rate": 9.701878218556129e-05, "loss": 0.0301, "step": 3100 }, { "grad_norm": 0.22642365097999573, "learning_rate": 9.699059755431598e-05, "loss": 0.0278, "step": 3110 }, { "grad_norm": 0.3493439257144928, "learning_rate": 9.696228445257132e-05, "loss": 0.0292, "step": 3120 }, { "grad_norm": 0.25111737847328186, "learning_rate": 9.693384295773419e-05, "loss": 0.0278, "step": 3130 }, { "grad_norm": 0.295900821685791, "learning_rate": 9.690527314756259e-05, "loss": 0.0265, "step": 3140 }, { "grad_norm": 0.29801145195961, "learning_rate": 9.687657510016527e-05, "loss": 0.0292, "step": 3150 }, { "grad_norm": 0.37923622131347656, "learning_rate": 9.684774889400161e-05, "loss": 0.0302, "step": 3160 }, { "grad_norm": 0.3373178541660309, "learning_rate": 9.681879460788135e-05, "loss": 0.0261, "step": 3170 }, { "grad_norm": 0.265952467918396, "learning_rate": 9.67897123209644e-05, "loss": 0.0257, "step": 3180 }, { "grad_norm": 0.26399895548820496, "learning_rate": 9.676050211276062e-05, "loss": 0.0257, "step": 3190 }, { "grad_norm": 0.3100784420967102, "learning_rate": 9.673116406312962e-05, "loss": 0.028, "step": 3200 }, { "grad_norm": 0.32248544692993164, "learning_rate": 9.67016982522805e-05, "loss": 0.0306, "step": 3210 }, { "grad_norm": 0.30626142024993896, "learning_rate": 9.667210476077164e-05, "loss": 0.0277, "step": 3220 }, { "grad_norm": 0.3038720190525055, "learning_rate": 9.664238366951055e-05, "loss": 0.0241, "step": 3230 }, { "grad_norm": 0.2561943829059601, "learning_rate": 9.661253505975355e-05, "loss": 0.0257, "step": 3240 }, { "grad_norm": 0.2544701099395752, "learning_rate": 9.658255901310557e-05, "loss": 0.0285, "step": 3250 }, { "grad_norm": 0.41683459281921387, "learning_rate": 9.655245561152e-05, "loss": 0.0312, "step": 3260 }, { "grad_norm": 0.33521905541419983, "learning_rate": 9.65222249372984e-05, "loss": 0.0309, "step": 3270 }, { "grad_norm": 0.24427078664302826, "learning_rate": 9.649186707309026e-05, "loss": 0.0274, "step": 3280 }, { "grad_norm": 0.3226320743560791, "learning_rate": 9.646138210189283e-05, "loss": 0.0328, "step": 3290 }, { "grad_norm": 0.3597772419452667, "learning_rate": 9.643077010705087e-05, "loss": 0.0263, "step": 3300 }, { "grad_norm": 0.4194526672363281, "learning_rate": 9.640003117225637e-05, "loss": 0.0307, "step": 3310 }, { "grad_norm": 0.6525530815124512, "learning_rate": 9.636916538154846e-05, "loss": 0.0297, "step": 3320 }, { "grad_norm": 0.37794530391693115, "learning_rate": 9.633817281931296e-05, "loss": 0.0352, "step": 3330 }, { "grad_norm": 0.3521789312362671, "learning_rate": 9.630705357028242e-05, "loss": 0.0372, "step": 3340 }, { "grad_norm": 0.36322444677352905, "learning_rate": 9.627580771953563e-05, "loss": 0.0321, "step": 3350 }, { "grad_norm": 0.4812558591365814, "learning_rate": 9.624443535249759e-05, "loss": 0.0304, "step": 3360 }, { "grad_norm": 0.2628926932811737, "learning_rate": 9.621293655493913e-05, "loss": 0.0337, "step": 3370 }, { "grad_norm": 0.40409624576568604, "learning_rate": 9.618131141297675e-05, "loss": 0.0289, "step": 3380 }, { "grad_norm": 0.3117417097091675, "learning_rate": 9.614956001307242e-05, "loss": 0.0267, "step": 3390 }, { "grad_norm": 0.37534040212631226, "learning_rate": 9.611768244203321e-05, "loss": 0.0278, "step": 3400 }, { "grad_norm": 0.37183597683906555, "learning_rate": 9.60856787870112e-05, "loss": 0.0283, "step": 3410 }, { "grad_norm": 0.2955438494682312, "learning_rate": 9.605354913550318e-05, "loss": 0.0235, "step": 3420 }, { "grad_norm": 0.278422087430954, "learning_rate": 9.602129357535037e-05, "loss": 0.0276, "step": 3430 }, { "grad_norm": 0.2792695164680481, "learning_rate": 9.598891219473825e-05, "loss": 0.0291, "step": 3440 }, { "grad_norm": 0.3486826419830322, "learning_rate": 9.595640508219625e-05, "loss": 0.0238, "step": 3450 }, { "grad_norm": 0.28768232464790344, "learning_rate": 9.592377232659761e-05, "loss": 0.0264, "step": 3460 }, { "grad_norm": 0.26629701256752014, "learning_rate": 9.589101401715904e-05, "loss": 0.024, "step": 3470 }, { "grad_norm": 0.32577693462371826, "learning_rate": 9.585813024344045e-05, "loss": 0.0286, "step": 3480 }, { "grad_norm": 0.2613057792186737, "learning_rate": 9.58251210953449e-05, "loss": 0.0251, "step": 3490 }, { "grad_norm": 0.30816107988357544, "learning_rate": 9.579198666311809e-05, "loss": 0.0225, "step": 3500 }, { "grad_norm": 0.2704595625400543, "learning_rate": 9.575872703734832e-05, "loss": 0.0238, "step": 3510 }, { "grad_norm": 0.2988549768924713, "learning_rate": 9.572534230896611e-05, "loss": 0.0258, "step": 3520 }, { "grad_norm": 0.3485180139541626, "learning_rate": 9.569183256924403e-05, "loss": 0.026, "step": 3530 }, { "grad_norm": 0.2783620357513428, "learning_rate": 9.565819790979646e-05, "loss": 0.0251, "step": 3540 }, { "grad_norm": 0.395157128572464, "learning_rate": 9.562443842257925e-05, "loss": 0.0218, "step": 3550 }, { "grad_norm": 0.3738989531993866, "learning_rate": 9.559055419988956e-05, "loss": 0.0247, "step": 3560 }, { "grad_norm": 0.2959824204444885, "learning_rate": 9.555654533436557e-05, "loss": 0.0246, "step": 3570 }, { "grad_norm": 0.39156705141067505, "learning_rate": 9.552241191898621e-05, "loss": 0.0295, "step": 3580 }, { "grad_norm": 0.30096226930618286, "learning_rate": 9.548815404707092e-05, "loss": 0.0233, "step": 3590 }, { "grad_norm": 0.26329532265663147, "learning_rate": 9.545377181227942e-05, "loss": 0.0324, "step": 3600 }, { "grad_norm": 0.2940422594547272, "learning_rate": 9.541926530861145e-05, "loss": 0.0248, "step": 3610 }, { "grad_norm": 0.4254361391067505, "learning_rate": 9.538463463040645e-05, "loss": 0.0261, "step": 3620 }, { "grad_norm": 0.32532799243927, "learning_rate": 9.534987987234337e-05, "loss": 0.0285, "step": 3630 }, { "grad_norm": 0.29015398025512695, "learning_rate": 9.53150011294404e-05, "loss": 0.0261, "step": 3640 }, { "grad_norm": 0.2412039339542389, "learning_rate": 9.527999849705471e-05, "loss": 0.0252, "step": 3650 }, { "grad_norm": 0.29864776134490967, "learning_rate": 9.524487207088213e-05, "loss": 0.0311, "step": 3660 }, { "grad_norm": 0.25623393058776855, "learning_rate": 9.520962194695698e-05, "loss": 0.0329, "step": 3670 }, { "grad_norm": 0.2749664783477783, "learning_rate": 9.517424822165175e-05, "loss": 0.0272, "step": 3680 }, { "grad_norm": 0.2094661444425583, "learning_rate": 9.513875099167685e-05, "loss": 0.0304, "step": 3690 }, { "grad_norm": 0.2849757671356201, "learning_rate": 9.510313035408035e-05, "loss": 0.0271, "step": 3700 }, { "grad_norm": 0.32192251086235046, "learning_rate": 9.506738640624775e-05, "loss": 0.027, "step": 3710 }, { "grad_norm": 0.37695351243019104, "learning_rate": 9.50315192459016e-05, "loss": 0.0254, "step": 3720 }, { "grad_norm": 0.3017573952674866, "learning_rate": 9.499552897110136e-05, "loss": 0.0274, "step": 3730 }, { "grad_norm": 0.3239597678184509, "learning_rate": 9.495941568024304e-05, "loss": 0.0291, "step": 3740 }, { "grad_norm": 0.35338523983955383, "learning_rate": 9.492317947205904e-05, "loss": 0.0301, "step": 3750 }, { "grad_norm": 0.4384407699108124, "learning_rate": 9.488682044561775e-05, "loss": 0.0323, "step": 3760 }, { "grad_norm": 0.33729779720306396, "learning_rate": 9.485033870032335e-05, "loss": 0.0277, "step": 3770 }, { "grad_norm": 0.366960734128952, "learning_rate": 9.481373433591556e-05, "loss": 0.0268, "step": 3780 }, { "grad_norm": 0.35313597321510315, "learning_rate": 9.47770074524693e-05, "loss": 0.0293, "step": 3790 }, { "grad_norm": 0.2847875654697418, "learning_rate": 9.474015815039446e-05, "loss": 0.0255, "step": 3800 }, { "grad_norm": 0.2263738214969635, "learning_rate": 9.470318653043565e-05, "loss": 0.024, "step": 3810 }, { "grad_norm": 0.27012160420417786, "learning_rate": 9.466609269367185e-05, "loss": 0.0266, "step": 3820 }, { "grad_norm": 0.32216185331344604, "learning_rate": 9.46288767415162e-05, "loss": 0.0277, "step": 3830 }, { "grad_norm": 0.2832684814929962, "learning_rate": 9.459153877571567e-05, "loss": 0.0245, "step": 3840 }, { "grad_norm": 0.25901156663894653, "learning_rate": 9.455407889835087e-05, "loss": 0.0263, "step": 3850 }, { "grad_norm": 0.39876577258110046, "learning_rate": 9.451649721183564e-05, "loss": 0.0247, "step": 3860 }, { "grad_norm": 0.2584032714366913, "learning_rate": 9.447879381891692e-05, "loss": 0.0276, "step": 3870 }, { "grad_norm": 0.27052536606788635, "learning_rate": 9.444096882267428e-05, "loss": 0.0252, "step": 3880 }, { "grad_norm": 0.3236894905567169, "learning_rate": 9.440302232651988e-05, "loss": 0.0244, "step": 3890 }, { "grad_norm": 0.42329999804496765, "learning_rate": 9.436495443419795e-05, "loss": 0.0258, "step": 3900 }, { "grad_norm": 0.29471972584724426, "learning_rate": 9.432676524978466e-05, "loss": 0.0275, "step": 3910 }, { "grad_norm": 0.39826878905296326, "learning_rate": 9.42884548776878e-05, "loss": 0.0321, "step": 3920 }, { "grad_norm": 0.26865747570991516, "learning_rate": 9.425002342264646e-05, "loss": 0.0252, "step": 3930 }, { "grad_norm": 0.3103424310684204, "learning_rate": 9.421147098973077e-05, "loss": 0.0253, "step": 3940 }, { "grad_norm": 0.36165323853492737, "learning_rate": 9.41727976843416e-05, "loss": 0.0234, "step": 3950 }, { "grad_norm": 0.34988442063331604, "learning_rate": 9.413400361221029e-05, "loss": 0.0259, "step": 3960 }, { "grad_norm": 0.3354589343070984, "learning_rate": 9.409508887939835e-05, "loss": 0.0239, "step": 3970 }, { "grad_norm": 0.304311603307724, "learning_rate": 9.40560535922972e-05, "loss": 0.0322, "step": 3980 }, { "grad_norm": 0.28600895404815674, "learning_rate": 9.40168978576278e-05, "loss": 0.0325, "step": 3990 }, { "grad_norm": 0.3256906569004059, "learning_rate": 9.397762178244043e-05, "loss": 0.0274, "step": 4000 }, { "grad_norm": 0.2898476719856262, "learning_rate": 9.393822547411439e-05, "loss": 0.0251, "step": 4010 }, { "grad_norm": 0.23647788166999817, "learning_rate": 9.389870904035769e-05, "loss": 0.0268, "step": 4020 }, { "grad_norm": 0.3351585268974304, "learning_rate": 9.385907258920672e-05, "loss": 0.0234, "step": 4030 }, { "grad_norm": 0.382069855928421, "learning_rate": 9.381931622902607e-05, "loss": 0.0239, "step": 4040 }, { "grad_norm": 0.350198358297348, "learning_rate": 9.377944006850807e-05, "loss": 0.0272, "step": 4050 }, { "grad_norm": 0.3371732532978058, "learning_rate": 9.373944421667265e-05, "loss": 0.0261, "step": 4060 }, { "grad_norm": 0.43648192286491394, "learning_rate": 9.369932878286691e-05, "loss": 0.0256, "step": 4070 }, { "grad_norm": 0.37950098514556885, "learning_rate": 9.365909387676494e-05, "loss": 0.0291, "step": 4080 }, { "grad_norm": 0.2791661322116852, "learning_rate": 9.361873960836744e-05, "loss": 0.0305, "step": 4090 }, { "grad_norm": 0.255513995885849, "learning_rate": 9.357826608800142e-05, "loss": 0.0278, "step": 4100 }, { "grad_norm": 0.29305359721183777, "learning_rate": 9.353767342631994e-05, "loss": 0.0253, "step": 4110 }, { "grad_norm": 0.5364317893981934, "learning_rate": 9.34969617343018e-05, "loss": 0.0244, "step": 4120 }, { "grad_norm": 0.2864401340484619, "learning_rate": 9.345613112325122e-05, "loss": 0.0233, "step": 4130 }, { "grad_norm": 0.27019721269607544, "learning_rate": 9.34151817047975e-05, "loss": 0.0294, "step": 4140 }, { "grad_norm": 0.41067349910736084, "learning_rate": 9.33741135908948e-05, "loss": 0.0252, "step": 4150 }, { "grad_norm": 0.3420601487159729, "learning_rate": 9.33329268938218e-05, "loss": 0.0295, "step": 4160 }, { "grad_norm": 0.23789533972740173, "learning_rate": 9.329162172618132e-05, "loss": 0.0264, "step": 4170 }, { "grad_norm": 0.26067236065864563, "learning_rate": 9.325019820090013e-05, "loss": 0.0271, "step": 4180 }, { "grad_norm": 0.32032111287117004, "learning_rate": 9.320865643122855e-05, "loss": 0.0222, "step": 4190 }, { "grad_norm": 0.3144882321357727, "learning_rate": 9.316699653074023e-05, "loss": 0.0257, "step": 4200 }, { "grad_norm": 0.36000847816467285, "learning_rate": 9.312521861333172e-05, "loss": 0.0248, "step": 4210 }, { "grad_norm": 0.32531729340553284, "learning_rate": 9.308332279322224e-05, "loss": 0.0229, "step": 4220 }, { "grad_norm": 0.3027832806110382, "learning_rate": 9.304130918495338e-05, "loss": 0.0219, "step": 4230 }, { "grad_norm": 0.38661858439445496, "learning_rate": 9.299917790338874e-05, "loss": 0.0278, "step": 4240 }, { "grad_norm": 0.2941456735134125, "learning_rate": 9.295692906371363e-05, "loss": 0.0236, "step": 4250 }, { "grad_norm": 0.34513744711875916, "learning_rate": 9.291456278143476e-05, "loss": 0.0235, "step": 4260 }, { "grad_norm": 0.19920577108860016, "learning_rate": 9.287207917237994e-05, "loss": 0.0248, "step": 4270 }, { "grad_norm": 0.2991308569908142, "learning_rate": 9.282947835269773e-05, "loss": 0.0259, "step": 4280 }, { "grad_norm": 0.3709297776222229, "learning_rate": 9.278676043885715e-05, "loss": 0.0261, "step": 4290 }, { "grad_norm": 0.22525301575660706, "learning_rate": 9.274392554764733e-05, "loss": 0.0231, "step": 4300 }, { "grad_norm": 0.24396969377994537, "learning_rate": 9.270097379617723e-05, "loss": 0.0219, "step": 4310 }, { "grad_norm": 0.2269345074892044, "learning_rate": 9.26579053018753e-05, "loss": 0.0218, "step": 4320 }, { "grad_norm": 0.31672370433807373, "learning_rate": 9.261472018248918e-05, "loss": 0.0242, "step": 4330 }, { "grad_norm": 0.4205857217311859, "learning_rate": 9.25714185560853e-05, "loss": 0.0244, "step": 4340 }, { "grad_norm": 0.2906699478626251, "learning_rate": 9.252800054104868e-05, "loss": 0.0226, "step": 4350 }, { "grad_norm": 0.28578364849090576, "learning_rate": 9.248446625608252e-05, "loss": 0.0194, "step": 4360 }, { "grad_norm": 0.24371084570884705, "learning_rate": 9.244081582020789e-05, "loss": 0.0221, "step": 4370 }, { "grad_norm": 0.29041123390197754, "learning_rate": 9.239704935276339e-05, "loss": 0.0211, "step": 4380 }, { "grad_norm": 0.29150763154029846, "learning_rate": 9.235316697340489e-05, "loss": 0.0233, "step": 4390 }, { "grad_norm": 0.1873895823955536, "learning_rate": 9.230916880210512e-05, "loss": 0.0226, "step": 4400 }, { "grad_norm": 0.25301605463027954, "learning_rate": 9.226505495915342e-05, "loss": 0.0231, "step": 4410 }, { "grad_norm": 0.354509562253952, "learning_rate": 9.222082556515536e-05, "loss": 0.0206, "step": 4420 }, { "grad_norm": 0.27436500787734985, "learning_rate": 9.217648074103242e-05, "loss": 0.0237, "step": 4430 }, { "grad_norm": 0.23855647444725037, "learning_rate": 9.213202060802161e-05, "loss": 0.0213, "step": 4440 }, { "grad_norm": 0.37332093715667725, "learning_rate": 9.208744528767528e-05, "loss": 0.0219, "step": 4450 }, { "grad_norm": 0.17822399735450745, "learning_rate": 9.204275490186064e-05, "loss": 0.0226, "step": 4460 }, { "grad_norm": 0.2742353677749634, "learning_rate": 9.199794957275949e-05, "loss": 0.0207, "step": 4470 }, { "grad_norm": 0.3224351406097412, "learning_rate": 9.19530294228679e-05, "loss": 0.0262, "step": 4480 }, { "grad_norm": 0.3522495925426483, "learning_rate": 9.190799457499583e-05, "loss": 0.0225, "step": 4490 }, { "grad_norm": 0.39795807003974915, "learning_rate": 9.186284515226686e-05, "loss": 0.0249, "step": 4500 }, { "grad_norm": 0.3742760419845581, "learning_rate": 9.181758127811777e-05, "loss": 0.0238, "step": 4510 }, { "grad_norm": 0.2735960781574249, "learning_rate": 9.177220307629825e-05, "loss": 0.0227, "step": 4520 }, { "grad_norm": 0.3303253948688507, "learning_rate": 9.172671067087059e-05, "loss": 0.0241, "step": 4530 }, { "grad_norm": 0.21396523714065552, "learning_rate": 9.16811041862093e-05, "loss": 0.0231, "step": 4540 }, { "grad_norm": 0.3396661877632141, "learning_rate": 9.163538374700076e-05, "loss": 0.0248, "step": 4550 }, { "grad_norm": 0.31576722860336304, "learning_rate": 9.158954947824287e-05, "loss": 0.0271, "step": 4560 }, { "grad_norm": 0.38072559237480164, "learning_rate": 9.154360150524482e-05, "loss": 0.0253, "step": 4570 }, { "grad_norm": 0.3530745506286621, "learning_rate": 9.14975399536266e-05, "loss": 0.0249, "step": 4580 }, { "grad_norm": 0.2674061357975006, "learning_rate": 9.14513649493187e-05, "loss": 0.0255, "step": 4590 }, { "grad_norm": 0.2291128933429718, "learning_rate": 9.140507661856187e-05, "loss": 0.0208, "step": 4600 }, { "grad_norm": 0.3165527284145355, "learning_rate": 9.135867508790661e-05, "loss": 0.0225, "step": 4610 }, { "grad_norm": 0.32957351207733154, "learning_rate": 9.131216048421291e-05, "loss": 0.0246, "step": 4620 }, { "grad_norm": 0.38280531764030457, "learning_rate": 9.126553293464998e-05, "loss": 0.0268, "step": 4630 }, { "grad_norm": 0.33500155806541443, "learning_rate": 9.121879256669572e-05, "loss": 0.0255, "step": 4640 }, { "grad_norm": 0.3858092725276947, "learning_rate": 9.117193950813652e-05, "loss": 0.0276, "step": 4650 }, { "grad_norm": 0.24267694354057312, "learning_rate": 9.112497388706685e-05, "loss": 0.0229, "step": 4660 }, { "grad_norm": 0.18580585718154907, "learning_rate": 9.10778958318889e-05, "loss": 0.0247, "step": 4670 }, { "grad_norm": 0.25818824768066406, "learning_rate": 9.103070547131232e-05, "loss": 0.0265, "step": 4680 }, { "grad_norm": 0.3480759859085083, "learning_rate": 9.098340293435375e-05, "loss": 0.0272, "step": 4690 }, { "grad_norm": 0.33650335669517517, "learning_rate": 9.093598835033649e-05, "loss": 0.0234, "step": 4700 }, { "grad_norm": 0.22793474793434143, "learning_rate": 9.088846184889021e-05, "loss": 0.0227, "step": 4710 }, { "grad_norm": 0.39595478773117065, "learning_rate": 9.084082355995057e-05, "loss": 0.0234, "step": 4720 }, { "grad_norm": 0.32657697796821594, "learning_rate": 9.079307361375882e-05, "loss": 0.0219, "step": 4730 }, { "grad_norm": 0.31126877665519714, "learning_rate": 9.074521214086149e-05, "loss": 0.0243, "step": 4740 }, { "grad_norm": 0.36304011940956116, "learning_rate": 9.069723927211001e-05, "loss": 0.0264, "step": 4750 }, { "grad_norm": 0.32150155305862427, "learning_rate": 9.064915513866037e-05, "loss": 0.0262, "step": 4760 }, { "grad_norm": 0.2674364149570465, "learning_rate": 9.060095987197279e-05, "loss": 0.0267, "step": 4770 }, { "grad_norm": 0.25562605261802673, "learning_rate": 9.055265360381126e-05, "loss": 0.024, "step": 4780 }, { "grad_norm": 0.29030928015708923, "learning_rate": 9.050423646624326e-05, "loss": 0.0237, "step": 4790 }, { "grad_norm": 0.27083271741867065, "learning_rate": 9.045570859163943e-05, "loss": 0.0202, "step": 4800 }, { "grad_norm": 0.16025730967521667, "learning_rate": 9.04070701126731e-05, "loss": 0.0252, "step": 4810 }, { "grad_norm": 0.2560851275920868, "learning_rate": 9.035832116232001e-05, "loss": 0.0216, "step": 4820 }, { "grad_norm": 0.25820013880729675, "learning_rate": 9.030946187385796e-05, "loss": 0.0242, "step": 4830 }, { "grad_norm": 0.23408178985118866, "learning_rate": 9.026049238086635e-05, "loss": 0.021, "step": 4840 }, { "grad_norm": 0.23758606612682343, "learning_rate": 9.021141281722591e-05, "loss": 0.0213, "step": 4850 }, { "grad_norm": 0.283035546541214, "learning_rate": 9.01622233171183e-05, "loss": 0.0225, "step": 4860 }, { "grad_norm": 0.32862725853919983, "learning_rate": 9.011292401502574e-05, "loss": 0.0195, "step": 4870 }, { "grad_norm": 0.3817649483680725, "learning_rate": 9.006351504573063e-05, "loss": 0.0217, "step": 4880 }, { "grad_norm": 0.2556794285774231, "learning_rate": 9.001399654431519e-05, "loss": 0.0222, "step": 4890 }, { "grad_norm": 0.25241684913635254, "learning_rate": 8.996436864616116e-05, "loss": 0.0236, "step": 4900 }, { "grad_norm": 0.34253990650177, "learning_rate": 8.991463148694925e-05, "loss": 0.0223, "step": 4910 }, { "grad_norm": 0.3526601791381836, "learning_rate": 8.986478520265902e-05, "loss": 0.0251, "step": 4920 }, { "grad_norm": 0.2376994490623474, "learning_rate": 8.981482992956827e-05, "loss": 0.0226, "step": 4930 }, { "grad_norm": 0.3224436044692993, "learning_rate": 8.976476580425282e-05, "loss": 0.024, "step": 4940 }, { "grad_norm": 0.22784872353076935, "learning_rate": 8.971459296358606e-05, "loss": 0.0227, "step": 4950 }, { "grad_norm": 0.24046412110328674, "learning_rate": 8.966431154473864e-05, "loss": 0.0266, "step": 4960 }, { "grad_norm": 0.24184997379779816, "learning_rate": 8.961392168517803e-05, "loss": 0.0265, "step": 4970 }, { "grad_norm": 0.23005761206150055, "learning_rate": 8.956342352266821e-05, "loss": 0.0259, "step": 4980 }, { "grad_norm": 0.23201654851436615, "learning_rate": 8.95128171952692e-05, "loss": 0.0205, "step": 4990 }, { "grad_norm": 0.2478942573070526, "learning_rate": 8.946210284133676e-05, "loss": 0.0235, "step": 5000 }, { "grad_norm": 0.2591395378112793, "learning_rate": 8.941128059952201e-05, "loss": 0.0213, "step": 5010 }, { "grad_norm": 0.26887205243110657, "learning_rate": 8.936035060877102e-05, "loss": 0.0235, "step": 5020 }, { "grad_norm": 0.2109590470790863, "learning_rate": 8.930931300832443e-05, "loss": 0.0222, "step": 5030 }, { "grad_norm": 0.3316001892089844, "learning_rate": 8.925816793771711e-05, "loss": 0.0222, "step": 5040 }, { "grad_norm": 0.2748890817165375, "learning_rate": 8.92069155367777e-05, "loss": 0.0226, "step": 5050 }, { "grad_norm": 0.4072771668434143, "learning_rate": 8.915555594562834e-05, "loss": 0.022, "step": 5060 }, { "grad_norm": 0.37765249609947205, "learning_rate": 8.910408930468416e-05, "loss": 0.0227, "step": 5070 }, { "grad_norm": 0.4228350520133972, "learning_rate": 8.905251575465303e-05, "loss": 0.0227, "step": 5080 }, { "grad_norm": 0.2961061894893646, "learning_rate": 8.900083543653502e-05, "loss": 0.0242, "step": 5090 }, { "grad_norm": 0.28316530585289, "learning_rate": 8.894904849162218e-05, "loss": 0.0241, "step": 5100 }, { "grad_norm": 0.2788536846637726, "learning_rate": 8.889715506149802e-05, "loss": 0.0226, "step": 5110 }, { "grad_norm": 0.20152918994426727, "learning_rate": 8.884515528803722e-05, "loss": 0.0233, "step": 5120 }, { "grad_norm": 0.2372935265302658, "learning_rate": 8.879304931340517e-05, "loss": 0.0224, "step": 5130 }, { "grad_norm": 0.3998088836669922, "learning_rate": 8.874083728005759e-05, "loss": 0.0264, "step": 5140 }, { "grad_norm": 0.491787314414978, "learning_rate": 8.868851933074021e-05, "loss": 0.0268, "step": 5150 }, { "grad_norm": 0.3271983563899994, "learning_rate": 8.863609560848829e-05, "loss": 0.025, "step": 5160 }, { "grad_norm": 0.33869922161102295, "learning_rate": 8.85835662566263e-05, "loss": 0.0216, "step": 5170 }, { "grad_norm": 0.2963501513004303, "learning_rate": 8.853093141876747e-05, "loss": 0.0213, "step": 5180 }, { "grad_norm": 0.36952269077301025, "learning_rate": 8.847819123881343e-05, "loss": 0.0228, "step": 5190 }, { "grad_norm": 0.22745965421199799, "learning_rate": 8.842534586095383e-05, "loss": 0.022, "step": 5200 }, { "grad_norm": 0.249125137925148, "learning_rate": 8.837239542966593e-05, "loss": 0.0214, "step": 5210 }, { "grad_norm": 0.2186688929796219, "learning_rate": 8.831934008971417e-05, "loss": 0.0205, "step": 5220 }, { "grad_norm": 0.22503463923931122, "learning_rate": 8.826617998614982e-05, "loss": 0.0204, "step": 5230 }, { "grad_norm": 0.1794584095478058, "learning_rate": 8.821291526431056e-05, "loss": 0.02, "step": 5240 }, { "grad_norm": 0.27132466435432434, "learning_rate": 8.815954606982015e-05, "loss": 0.0226, "step": 5250 }, { "grad_norm": 0.1912720501422882, "learning_rate": 8.810607254858789e-05, "loss": 0.0252, "step": 5260 }, { "grad_norm": 0.1879526525735855, "learning_rate": 8.805249484680838e-05, "loss": 0.0213, "step": 5270 }, { "grad_norm": 0.2038632333278656, "learning_rate": 8.799881311096096e-05, "loss": 0.0178, "step": 5280 }, { "grad_norm": 0.2604949176311493, "learning_rate": 8.794502748780949e-05, "loss": 0.0191, "step": 5290 }, { "grad_norm": 0.18916107714176178, "learning_rate": 8.78911381244018e-05, "loss": 0.0187, "step": 5300 }, { "grad_norm": 0.2976568639278412, "learning_rate": 8.783714516806933e-05, "loss": 0.0199, "step": 5310 }, { "grad_norm": 0.3691166341304779, "learning_rate": 8.77830487664268e-05, "loss": 0.0204, "step": 5320 }, { "grad_norm": 0.21564500033855438, "learning_rate": 8.772884906737167e-05, "loss": 0.0205, "step": 5330 }, { "grad_norm": 0.24967233836650848, "learning_rate": 8.767454621908387e-05, "loss": 0.0224, "step": 5340 }, { "grad_norm": 0.23411712050437927, "learning_rate": 8.76201403700253e-05, "loss": 0.0221, "step": 5350 }, { "grad_norm": 0.21195891499519348, "learning_rate": 8.756563166893949e-05, "loss": 0.0202, "step": 5360 }, { "grad_norm": 0.1819506585597992, "learning_rate": 8.751102026485113e-05, "loss": 0.0207, "step": 5370 }, { "grad_norm": 0.41229695081710815, "learning_rate": 8.745630630706571e-05, "loss": 0.0201, "step": 5380 }, { "grad_norm": 0.2500326335430145, "learning_rate": 8.740148994516912e-05, "loss": 0.0213, "step": 5390 }, { "grad_norm": 0.2843233048915863, "learning_rate": 8.73465713290272e-05, "loss": 0.0244, "step": 5400 }, { "grad_norm": 0.3113899528980255, "learning_rate": 8.729155060878533e-05, "loss": 0.0204, "step": 5410 }, { "grad_norm": 0.2570742666721344, "learning_rate": 8.723642793486809e-05, "loss": 0.0232, "step": 5420 }, { "grad_norm": 0.29251629114151, "learning_rate": 8.718120345797873e-05, "loss": 0.0197, "step": 5430 }, { "grad_norm": 0.33219635486602783, "learning_rate": 8.712587732909889e-05, "loss": 0.0195, "step": 5440 }, { "grad_norm": 0.30283066630363464, "learning_rate": 8.707044969948806e-05, "loss": 0.0201, "step": 5450 }, { "grad_norm": 0.20181602239608765, "learning_rate": 8.701492072068329e-05, "loss": 0.0213, "step": 5460 }, { "grad_norm": 0.24962152540683746, "learning_rate": 8.695929054449869e-05, "loss": 0.022, "step": 5470 }, { "grad_norm": 0.2875845730304718, "learning_rate": 8.690355932302501e-05, "loss": 0.0195, "step": 5480 }, { "grad_norm": 0.39994218945503235, "learning_rate": 8.684772720862931e-05, "loss": 0.021, "step": 5490 }, { "grad_norm": 0.16599278151988983, "learning_rate": 8.679179435395446e-05, "loss": 0.0241, "step": 5500 }, { "grad_norm": 0.24829648435115814, "learning_rate": 8.673576091191874e-05, "loss": 0.0225, "step": 5510 }, { "grad_norm": 0.23723724484443665, "learning_rate": 8.667962703571541e-05, "loss": 0.0237, "step": 5520 }, { "grad_norm": 0.3509126603603363, "learning_rate": 8.662339287881238e-05, "loss": 0.0244, "step": 5530 }, { "grad_norm": 0.38592252135276794, "learning_rate": 8.656705859495169e-05, "loss": 0.0268, "step": 5540 }, { "grad_norm": 0.16808390617370605, "learning_rate": 8.651062433814912e-05, "loss": 0.0225, "step": 5550 }, { "grad_norm": 0.2451239675283432, "learning_rate": 8.645409026269375e-05, "loss": 0.0228, "step": 5560 }, { "grad_norm": 0.2739429771900177, "learning_rate": 8.639745652314759e-05, "loss": 0.0201, "step": 5570 }, { "grad_norm": 0.2701444923877716, "learning_rate": 8.634072327434515e-05, "loss": 0.0252, "step": 5580 }, { "grad_norm": 0.2244923859834671, "learning_rate": 8.628389067139294e-05, "loss": 0.0233, "step": 5590 }, { "grad_norm": 0.1947420835494995, "learning_rate": 8.622695886966911e-05, "loss": 0.0223, "step": 5600 }, { "grad_norm": 0.2751563787460327, "learning_rate": 8.616992802482308e-05, "loss": 0.0199, "step": 5610 }, { "grad_norm": 0.256033718585968, "learning_rate": 8.611279829277496e-05, "loss": 0.0234, "step": 5620 }, { "grad_norm": 0.24640978872776031, "learning_rate": 8.605556982971528e-05, "loss": 0.022, "step": 5630 }, { "grad_norm": 0.260158509016037, "learning_rate": 8.599824279210447e-05, "loss": 0.0204, "step": 5640 }, { "grad_norm": 0.28527361154556274, "learning_rate": 8.594081733667243e-05, "loss": 0.0208, "step": 5650 }, { "grad_norm": 0.3622000515460968, "learning_rate": 8.58832936204182e-05, "loss": 0.0202, "step": 5660 }, { "grad_norm": 0.2987412214279175, "learning_rate": 8.582567180060942e-05, "loss": 0.0219, "step": 5670 }, { "grad_norm": 0.1687716543674469, "learning_rate": 8.576795203478194e-05, "loss": 0.0185, "step": 5680 }, { "grad_norm": 0.20221880078315735, "learning_rate": 8.571013448073939e-05, "loss": 0.0212, "step": 5690 }, { "grad_norm": 0.2531888484954834, "learning_rate": 8.565221929655275e-05, "loss": 0.016, "step": 5700 }, { "grad_norm": 0.22998863458633423, "learning_rate": 8.559420664055992e-05, "loss": 0.02, "step": 5710 }, { "grad_norm": 0.321410208940506, "learning_rate": 8.553609667136532e-05, "loss": 0.0202, "step": 5720 }, { "grad_norm": 0.35016027092933655, "learning_rate": 8.547788954783936e-05, "loss": 0.0238, "step": 5730 }, { "grad_norm": 0.37484341859817505, "learning_rate": 8.541958542911808e-05, "loss": 0.0235, "step": 5740 }, { "grad_norm": 0.17683212459087372, "learning_rate": 8.536118447460275e-05, "loss": 0.0192, "step": 5750 }, { "grad_norm": 0.28222477436065674, "learning_rate": 8.530268684395932e-05, "loss": 0.0225, "step": 5760 }, { "grad_norm": 0.21024705469608307, "learning_rate": 8.524409269711807e-05, "loss": 0.0212, "step": 5770 }, { "grad_norm": 0.24258175492286682, "learning_rate": 8.51854021942732e-05, "loss": 0.0198, "step": 5780 }, { "grad_norm": 0.16889381408691406, "learning_rate": 8.512661549588227e-05, "loss": 0.0196, "step": 5790 }, { "grad_norm": 0.1552414745092392, "learning_rate": 8.506773276266588e-05, "loss": 0.02, "step": 5800 }, { "grad_norm": 0.2113981992006302, "learning_rate": 8.500875415560721e-05, "loss": 0.0184, "step": 5810 }, { "grad_norm": 0.2851541042327881, "learning_rate": 8.494967983595144e-05, "loss": 0.0179, "step": 5820 }, { "grad_norm": 0.25244030356407166, "learning_rate": 8.489050996520558e-05, "loss": 0.0192, "step": 5830 }, { "grad_norm": 0.21890607476234436, "learning_rate": 8.483124470513775e-05, "loss": 0.0188, "step": 5840 }, { "grad_norm": 0.22464267909526825, "learning_rate": 8.477188421777692e-05, "loss": 0.0221, "step": 5850 }, { "grad_norm": 0.25398364663124084, "learning_rate": 8.47124286654124e-05, "loss": 0.0217, "step": 5860 }, { "grad_norm": 0.2727900445461273, "learning_rate": 8.465287821059341e-05, "loss": 0.02, "step": 5870 }, { "grad_norm": 0.21523083746433258, "learning_rate": 8.45932330161286e-05, "loss": 0.0204, "step": 5880 }, { "grad_norm": 0.40077197551727295, "learning_rate": 8.453349324508567e-05, "loss": 0.02, "step": 5890 }, { "grad_norm": 0.2907058894634247, "learning_rate": 8.447365906079088e-05, "loss": 0.0214, "step": 5900 }, { "grad_norm": 0.29472312331199646, "learning_rate": 8.441373062682856e-05, "loss": 0.0232, "step": 5910 }, { "grad_norm": 0.2412874698638916, "learning_rate": 8.43537081070408e-05, "loss": 0.0206, "step": 5920 }, { "grad_norm": 0.3658469021320343, "learning_rate": 8.429359166552689e-05, "loss": 0.0223, "step": 5930 }, { "grad_norm": 0.33984073996543884, "learning_rate": 8.423338146664284e-05, "loss": 0.0246, "step": 5940 }, { "grad_norm": 0.34227797389030457, "learning_rate": 8.417307767500107e-05, "loss": 0.0255, "step": 5950 }, { "grad_norm": 0.2985921800136566, "learning_rate": 8.411268045546983e-05, "loss": 0.0226, "step": 5960 }, { "grad_norm": 0.2959548830986023, "learning_rate": 8.405218997317281e-05, "loss": 0.0212, "step": 5970 }, { "grad_norm": 0.26979583501815796, "learning_rate": 8.399160639348869e-05, "loss": 0.018, "step": 5980 }, { "grad_norm": 0.3153114318847656, "learning_rate": 8.393092988205065e-05, "loss": 0.0196, "step": 5990 }, { "grad_norm": 0.2691565155982971, "learning_rate": 8.387016060474597e-05, "loss": 0.0211, "step": 6000 }, { "grad_norm": 0.3701017498970032, "learning_rate": 8.380929872771551e-05, "loss": 0.0228, "step": 6010 }, { "grad_norm": 0.22233252227306366, "learning_rate": 8.374834441735335e-05, "loss": 0.0213, "step": 6020 }, { "grad_norm": 0.26406341791152954, "learning_rate": 8.368729784030622e-05, "loss": 0.0208, "step": 6030 }, { "grad_norm": 0.2812860310077667, "learning_rate": 8.362615916347315e-05, "loss": 0.0198, "step": 6040 }, { "grad_norm": 0.25451943278312683, "learning_rate": 8.356492855400493e-05, "loss": 0.0188, "step": 6050 }, { "grad_norm": 0.31670141220092773, "learning_rate": 8.350360617930371e-05, "loss": 0.0211, "step": 6060 }, { "grad_norm": 0.21673224866390228, "learning_rate": 8.344219220702255e-05, "loss": 0.0203, "step": 6070 }, { "grad_norm": 0.25056251883506775, "learning_rate": 8.338068680506485e-05, "loss": 0.0259, "step": 6080 }, { "grad_norm": 0.26861751079559326, "learning_rate": 8.33190901415841e-05, "loss": 0.0232, "step": 6090 }, { "grad_norm": 0.21489325165748596, "learning_rate": 8.325740238498317e-05, "loss": 0.0207, "step": 6100 }, { "grad_norm": 0.32015421986579895, "learning_rate": 8.319562370391406e-05, "loss": 0.0202, "step": 6110 }, { "grad_norm": 0.2570993900299072, "learning_rate": 8.31337542672773e-05, "loss": 0.0218, "step": 6120 }, { "grad_norm": 0.18765386939048767, "learning_rate": 8.307179424422158e-05, "loss": 0.0205, "step": 6130 }, { "grad_norm": 0.5417460799217224, "learning_rate": 8.300974380414327e-05, "loss": 0.0215, "step": 6140 }, { "grad_norm": 0.23118659853935242, "learning_rate": 8.294760311668586e-05, "loss": 0.0218, "step": 6150 }, { "grad_norm": 0.2574794292449951, "learning_rate": 8.288537235173961e-05, "loss": 0.0227, "step": 6160 }, { "grad_norm": 0.23618067800998688, "learning_rate": 8.282305167944108e-05, "loss": 0.0219, "step": 6170 }, { "grad_norm": 0.3178730607032776, "learning_rate": 8.276064127017262e-05, "loss": 0.0207, "step": 6180 }, { "grad_norm": 0.2722536623477936, "learning_rate": 8.269814129456189e-05, "loss": 0.0198, "step": 6190 }, { "grad_norm": 0.2670479416847229, "learning_rate": 8.263555192348143e-05, "loss": 0.0205, "step": 6200 }, { "grad_norm": 0.28584906458854675, "learning_rate": 8.257287332804819e-05, "loss": 0.0208, "step": 6210 }, { "grad_norm": 0.31292831897735596, "learning_rate": 8.251010567962307e-05, "loss": 0.0197, "step": 6220 }, { "grad_norm": 0.23862189054489136, "learning_rate": 8.244724914981041e-05, "loss": 0.0175, "step": 6230 }, { "grad_norm": 0.28549885749816895, "learning_rate": 8.238430391045757e-05, "loss": 0.0178, "step": 6240 }, { "grad_norm": 0.2776296138763428, "learning_rate": 8.232127013365445e-05, "loss": 0.0204, "step": 6250 }, { "grad_norm": 0.3639026880264282, "learning_rate": 8.225814799173295e-05, "loss": 0.017, "step": 6260 }, { "grad_norm": 0.2942827045917511, "learning_rate": 8.219493765726663e-05, "loss": 0.0189, "step": 6270 }, { "grad_norm": 0.3902848958969116, "learning_rate": 8.21316393030701e-05, "loss": 0.0203, "step": 6280 }, { "grad_norm": 0.29440492391586304, "learning_rate": 8.206825310219865e-05, "loss": 0.0253, "step": 6290 }, { "grad_norm": 0.18431399762630463, "learning_rate": 8.200477922794776e-05, "loss": 0.019, "step": 6300 }, { "grad_norm": 0.22909630835056305, "learning_rate": 8.194121785385256e-05, "loss": 0.0187, "step": 6310 }, { "grad_norm": 0.2831685245037079, "learning_rate": 8.187756915368741e-05, "loss": 0.0191, "step": 6320 }, { "grad_norm": 0.28863203525543213, "learning_rate": 8.181383330146544e-05, "loss": 0.0179, "step": 6330 }, { "grad_norm": 0.2010720670223236, "learning_rate": 8.175001047143804e-05, "loss": 0.0185, "step": 6340 }, { "grad_norm": 0.1787133365869522, "learning_rate": 8.168610083809438e-05, "loss": 0.0205, "step": 6350 }, { "grad_norm": 0.18968579173088074, "learning_rate": 8.162210457616095e-05, "loss": 0.0189, "step": 6360 }, { "grad_norm": 0.30687886476516724, "learning_rate": 8.155802186060109e-05, "loss": 0.0182, "step": 6370 }, { "grad_norm": 0.21296708285808563, "learning_rate": 8.149385286661453e-05, "loss": 0.0175, "step": 6380 }, { "grad_norm": 0.22786547243595123, "learning_rate": 8.14295977696368e-05, "loss": 0.0192, "step": 6390 }, { "grad_norm": 0.38357463479042053, "learning_rate": 8.13652567453389e-05, "loss": 0.0224, "step": 6400 }, { "grad_norm": 0.2020588219165802, "learning_rate": 8.130082996962676e-05, "loss": 0.0221, "step": 6410 }, { "grad_norm": 0.2801573872566223, "learning_rate": 8.123631761864068e-05, "loss": 0.0217, "step": 6420 }, { "grad_norm": 0.24846212565898895, "learning_rate": 8.1171719868755e-05, "loss": 0.0203, "step": 6430 }, { "grad_norm": 0.1739795207977295, "learning_rate": 8.110703689657748e-05, "loss": 0.0192, "step": 6440 }, { "grad_norm": 0.17746397852897644, "learning_rate": 8.104226887894892e-05, "loss": 0.0195, "step": 6450 }, { "grad_norm": 0.41194823384284973, "learning_rate": 8.097741599294257e-05, "loss": 0.0216, "step": 6460 }, { "grad_norm": 0.19666379690170288, "learning_rate": 8.091247841586378e-05, "loss": 0.0195, "step": 6470 }, { "grad_norm": 0.23032647371292114, "learning_rate": 8.084745632524939e-05, "loss": 0.0168, "step": 6480 }, { "grad_norm": 0.21140994131565094, "learning_rate": 8.07823498988673e-05, "loss": 0.0214, "step": 6490 }, { "grad_norm": 0.1878626048564911, "learning_rate": 8.071715931471602e-05, "loss": 0.0197, "step": 6500 }, { "grad_norm": 0.3530450463294983, "learning_rate": 8.06518847510241e-05, "loss": 0.0193, "step": 6510 }, { "grad_norm": 0.2196698635816574, "learning_rate": 8.058652638624971e-05, "loss": 0.0169, "step": 6520 }, { "grad_norm": 0.4630400538444519, "learning_rate": 8.052108439908013e-05, "loss": 0.0178, "step": 6530 }, { "grad_norm": 0.2753053605556488, "learning_rate": 8.045555896843125e-05, "loss": 0.0162, "step": 6540 }, { "grad_norm": 0.23965266346931458, "learning_rate": 8.03899502734471e-05, "loss": 0.0205, "step": 6550 }, { "grad_norm": 0.1771155595779419, "learning_rate": 8.032425849349931e-05, "loss": 0.0181, "step": 6560 }, { "grad_norm": 0.20107156038284302, "learning_rate": 8.025848380818674e-05, "loss": 0.02, "step": 6570 }, { "grad_norm": 0.2622062861919403, "learning_rate": 8.019262639733487e-05, "loss": 0.0169, "step": 6580 }, { "grad_norm": 0.33602598309516907, "learning_rate": 8.012668644099531e-05, "loss": 0.0184, "step": 6590 }, { "grad_norm": 0.2180503010749817, "learning_rate": 8.006066411944542e-05, "loss": 0.0188, "step": 6600 }, { "grad_norm": 0.2111474871635437, "learning_rate": 7.999455961318769e-05, "loss": 0.0207, "step": 6610 }, { "grad_norm": 0.2788291871547699, "learning_rate": 7.992837310294932e-05, "loss": 0.0226, "step": 6620 }, { "grad_norm": 0.26197880506515503, "learning_rate": 7.986210476968167e-05, "loss": 0.021, "step": 6630 }, { "grad_norm": 0.16272090375423431, "learning_rate": 7.97957547945599e-05, "loss": 0.0211, "step": 6640 }, { "grad_norm": 0.24087786674499512, "learning_rate": 7.972932335898226e-05, "loss": 0.0226, "step": 6650 }, { "grad_norm": 0.25072023272514343, "learning_rate": 7.966281064456975e-05, "loss": 0.0196, "step": 6660 }, { "grad_norm": 0.31276509165763855, "learning_rate": 7.959621683316563e-05, "loss": 0.0185, "step": 6670 }, { "grad_norm": 0.22264394164085388, "learning_rate": 7.952954210683481e-05, "loss": 0.0179, "step": 6680 }, { "grad_norm": 0.19706203043460846, "learning_rate": 7.946278664786345e-05, "loss": 0.02, "step": 6690 }, { "grad_norm": 0.20434200763702393, "learning_rate": 7.939595063875842e-05, "loss": 0.0193, "step": 6700 }, { "grad_norm": 0.2938847839832306, "learning_rate": 7.932903426224683e-05, "loss": 0.0181, "step": 6710 }, { "grad_norm": 0.3453494608402252, "learning_rate": 7.926203770127552e-05, "loss": 0.025, "step": 6720 }, { "grad_norm": 0.2594270408153534, "learning_rate": 7.919496113901046e-05, "loss": 0.0205, "step": 6730 }, { "grad_norm": 0.24612809717655182, "learning_rate": 7.912780475883649e-05, "loss": 0.0175, "step": 6740 }, { "grad_norm": 0.3781329393386841, "learning_rate": 7.906056874435652e-05, "loss": 0.0225, "step": 6750 }, { "grad_norm": 0.18152102828025818, "learning_rate": 7.899325327939131e-05, "loss": 0.02, "step": 6760 }, { "grad_norm": 0.36237430572509766, "learning_rate": 7.892585854797872e-05, "loss": 0.0198, "step": 6770 }, { "grad_norm": 0.2064998745918274, "learning_rate": 7.88583847343734e-05, "loss": 0.0217, "step": 6780 }, { "grad_norm": 0.32214224338531494, "learning_rate": 7.879083202304616e-05, "loss": 0.0173, "step": 6790 }, { "grad_norm": 0.3429054021835327, "learning_rate": 7.872320059868355e-05, "loss": 0.022, "step": 6800 }, { "grad_norm": 0.17543268203735352, "learning_rate": 7.865549064618729e-05, "loss": 0.021, "step": 6810 }, { "grad_norm": 0.27797698974609375, "learning_rate": 7.858770235067381e-05, "loss": 0.0212, "step": 6820 }, { "grad_norm": 0.2212258130311966, "learning_rate": 7.851983589747374e-05, "loss": 0.0189, "step": 6830 }, { "grad_norm": 0.27862241864204407, "learning_rate": 7.845189147213133e-05, "loss": 0.0195, "step": 6840 }, { "grad_norm": 0.3431626558303833, "learning_rate": 7.838386926040407e-05, "loss": 0.019, "step": 6850 }, { "grad_norm": 0.23172588646411896, "learning_rate": 7.83157694482621e-05, "loss": 0.0196, "step": 6860 }, { "grad_norm": 0.31196215748786926, "learning_rate": 7.824759222188768e-05, "loss": 0.0173, "step": 6870 }, { "grad_norm": 0.13588187098503113, "learning_rate": 7.817933776767478e-05, "loss": 0.018, "step": 6880 }, { "grad_norm": 0.18465150892734528, "learning_rate": 7.811100627222842e-05, "loss": 0.016, "step": 6890 }, { "grad_norm": 0.18009082973003387, "learning_rate": 7.804259792236435e-05, "loss": 0.0164, "step": 6900 }, { "grad_norm": 0.28484469652175903, "learning_rate": 7.797411290510835e-05, "loss": 0.0186, "step": 6910 }, { "grad_norm": 0.30705904960632324, "learning_rate": 7.790555140769586e-05, "loss": 0.0191, "step": 6920 }, { "grad_norm": 0.22255361080169678, "learning_rate": 7.78369136175714e-05, "loss": 0.0189, "step": 6930 }, { "grad_norm": 0.24771253764629364, "learning_rate": 7.776819972238806e-05, "loss": 0.0163, "step": 6940 }, { "grad_norm": 0.16050483286380768, "learning_rate": 7.7699409910007e-05, "loss": 0.0185, "step": 6950 }, { "grad_norm": 0.32632383704185486, "learning_rate": 7.763054436849694e-05, "loss": 0.0186, "step": 6960 }, { "grad_norm": 0.5328749418258667, "learning_rate": 7.756160328613364e-05, "loss": 0.021, "step": 6970 }, { "grad_norm": 0.2242063283920288, "learning_rate": 7.749258685139942e-05, "loss": 0.0206, "step": 6980 }, { "grad_norm": 0.18545366823673248, "learning_rate": 7.742349525298253e-05, "loss": 0.0181, "step": 6990 }, { "grad_norm": 0.17437046766281128, "learning_rate": 7.735432867977679e-05, "loss": 0.0185, "step": 7000 }, { "grad_norm": 0.19849321246147156, "learning_rate": 7.728508732088096e-05, "loss": 0.0183, "step": 7010 }, { "grad_norm": 0.2999904453754425, "learning_rate": 7.721577136559825e-05, "loss": 0.0171, "step": 7020 }, { "grad_norm": 0.22566469013690948, "learning_rate": 7.714638100343588e-05, "loss": 0.0157, "step": 7030 }, { "grad_norm": 0.34885144233703613, "learning_rate": 7.707691642410444e-05, "loss": 0.0155, "step": 7040 }, { "grad_norm": 0.26660025119781494, "learning_rate": 7.70073778175174e-05, "loss": 0.0165, "step": 7050 }, { "grad_norm": 0.29459723830223083, "learning_rate": 7.69377653737907e-05, "loss": 0.019, "step": 7060 }, { "grad_norm": 0.314586341381073, "learning_rate": 7.686807928324209e-05, "loss": 0.0192, "step": 7070 }, { "grad_norm": 0.2616482079029083, "learning_rate": 7.679831973639065e-05, "loss": 0.016, "step": 7080 }, { "grad_norm": 0.32726597785949707, "learning_rate": 7.672848692395637e-05, "loss": 0.0181, "step": 7090 }, { "grad_norm": 0.16973477602005005, "learning_rate": 7.665858103685944e-05, "loss": 0.0168, "step": 7100 }, { "grad_norm": 0.18675169348716736, "learning_rate": 7.658860226621991e-05, "loss": 0.017, "step": 7110 }, { "grad_norm": 0.18446040153503418, "learning_rate": 7.651855080335708e-05, "loss": 0.0174, "step": 7120 }, { "grad_norm": 0.24896429479122162, "learning_rate": 7.644842683978896e-05, "loss": 0.0191, "step": 7130 }, { "grad_norm": 0.25105035305023193, "learning_rate": 7.63782305672318e-05, "loss": 0.0181, "step": 7140 }, { "grad_norm": 0.2629919648170471, "learning_rate": 7.63079621775995e-05, "loss": 0.0185, "step": 7150 }, { "grad_norm": 0.24058060348033905, "learning_rate": 7.623762186300319e-05, "loss": 0.0168, "step": 7160 }, { "grad_norm": 0.23369988799095154, "learning_rate": 7.616720981575057e-05, "loss": 0.0168, "step": 7170 }, { "grad_norm": 0.28397685289382935, "learning_rate": 7.609672622834552e-05, "loss": 0.0187, "step": 7180 }, { "grad_norm": 0.19403743743896484, "learning_rate": 7.602617129348747e-05, "loss": 0.0182, "step": 7190 }, { "grad_norm": 0.31797435879707336, "learning_rate": 7.595554520407088e-05, "loss": 0.0167, "step": 7200 }, { "grad_norm": 0.17207518219947815, "learning_rate": 7.588484815318484e-05, "loss": 0.0166, "step": 7210 }, { "grad_norm": 0.3094694912433624, "learning_rate": 7.581408033411234e-05, "loss": 0.0171, "step": 7220 }, { "grad_norm": 0.16179588437080383, "learning_rate": 7.574324194032995e-05, "loss": 0.0189, "step": 7230 }, { "grad_norm": 0.25959447026252747, "learning_rate": 7.567233316550705e-05, "loss": 0.0214, "step": 7240 }, { "grad_norm": 0.16207729279994965, "learning_rate": 7.560135420350562e-05, "loss": 0.016, "step": 7250 }, { "grad_norm": 0.24039432406425476, "learning_rate": 7.553030524837935e-05, "loss": 0.0166, "step": 7260 }, { "grad_norm": 0.360493004322052, "learning_rate": 7.545918649437341e-05, "loss": 0.0187, "step": 7270 }, { "grad_norm": 0.26531821489334106, "learning_rate": 7.538799813592377e-05, "loss": 0.0181, "step": 7280 }, { "grad_norm": 0.2173588126897812, "learning_rate": 7.531674036765662e-05, "loss": 0.017, "step": 7290 }, { "grad_norm": 0.1440780758857727, "learning_rate": 7.524541338438807e-05, "loss": 0.0167, "step": 7300 }, { "grad_norm": 0.23497842252254486, "learning_rate": 7.517401738112328e-05, "loss": 0.0164, "step": 7310 }, { "grad_norm": 0.25431498885154724, "learning_rate": 7.510255255305628e-05, "loss": 0.0183, "step": 7320 }, { "grad_norm": 0.20436953008174896, "learning_rate": 7.503101909556911e-05, "loss": 0.0166, "step": 7330 }, { "grad_norm": 0.20386800169944763, "learning_rate": 7.495941720423154e-05, "loss": 0.0161, "step": 7340 }, { "grad_norm": 0.18978512287139893, "learning_rate": 7.488774707480042e-05, "loss": 0.0173, "step": 7350 }, { "grad_norm": 0.2931479811668396, "learning_rate": 7.481600890321911e-05, "loss": 0.0231, "step": 7360 }, { "grad_norm": 0.3574753403663635, "learning_rate": 7.474420288561708e-05, "loss": 0.0207, "step": 7370 }, { "grad_norm": 0.31382858753204346, "learning_rate": 7.467232921830921e-05, "loss": 0.0181, "step": 7380 }, { "grad_norm": 0.24046793580055237, "learning_rate": 7.460038809779537e-05, "loss": 0.0208, "step": 7390 }, { "grad_norm": 0.30976778268814087, "learning_rate": 7.452837972075983e-05, "loss": 0.0185, "step": 7400 }, { "grad_norm": 0.2859417200088501, "learning_rate": 7.445630428407074e-05, "loss": 0.0168, "step": 7410 }, { "grad_norm": 0.23785966634750366, "learning_rate": 7.43841619847796e-05, "loss": 0.0171, "step": 7420 }, { "grad_norm": 0.241751030087471, "learning_rate": 7.431195302012072e-05, "loss": 0.0176, "step": 7430 }, { "grad_norm": 0.34365054965019226, "learning_rate": 7.423967758751061e-05, "loss": 0.0185, "step": 7440 }, { "grad_norm": 0.23825693130493164, "learning_rate": 7.416733588454758e-05, "loss": 0.0163, "step": 7450 }, { "grad_norm": 0.31214067339897156, "learning_rate": 7.409492810901106e-05, "loss": 0.0189, "step": 7460 }, { "grad_norm": 0.24603687226772308, "learning_rate": 7.402245445886116e-05, "loss": 0.0215, "step": 7470 }, { "grad_norm": 0.2149009108543396, "learning_rate": 7.394991513223806e-05, "loss": 0.017, "step": 7480 }, { "grad_norm": 0.21110311150550842, "learning_rate": 7.38773103274615e-05, "loss": 0.0156, "step": 7490 }, { "grad_norm": 0.27563247084617615, "learning_rate": 7.380464024303028e-05, "loss": 0.019, "step": 7500 }, { "grad_norm": 0.22046314179897308, "learning_rate": 7.373190507762162e-05, "loss": 0.0162, "step": 7510 }, { "grad_norm": 0.24888190627098083, "learning_rate": 7.365910503009066e-05, "loss": 0.0199, "step": 7520 }, { "grad_norm": 0.20471400022506714, "learning_rate": 7.358624029946996e-05, "loss": 0.0187, "step": 7530 }, { "grad_norm": 0.27773335576057434, "learning_rate": 7.351331108496893e-05, "loss": 0.0149, "step": 7540 }, { "grad_norm": 0.34374943375587463, "learning_rate": 7.344031758597325e-05, "loss": 0.0169, "step": 7550 }, { "grad_norm": 0.18476596474647522, "learning_rate": 7.336726000204435e-05, "loss": 0.0184, "step": 7560 }, { "grad_norm": 0.2686666250228882, "learning_rate": 7.32941385329189e-05, "loss": 0.0168, "step": 7570 }, { "grad_norm": 0.1734338104724884, "learning_rate": 7.322095337850816e-05, "loss": 0.0163, "step": 7580 }, { "grad_norm": 0.18169257044792175, "learning_rate": 7.314770473889758e-05, "loss": 0.0169, "step": 7590 }, { "grad_norm": 0.28833094239234924, "learning_rate": 7.307439281434615e-05, "loss": 0.0168, "step": 7600 }, { "grad_norm": 0.2249266803264618, "learning_rate": 7.300101780528585e-05, "loss": 0.0175, "step": 7610 }, { "grad_norm": 0.27109813690185547, "learning_rate": 7.292757991232117e-05, "loss": 0.0174, "step": 7620 }, { "grad_norm": 0.2606053352355957, "learning_rate": 7.285407933622848e-05, "loss": 0.0198, "step": 7630 }, { "grad_norm": 0.24064379930496216, "learning_rate": 7.278051627795557e-05, "loss": 0.0169, "step": 7640 }, { "grad_norm": 0.18343688547611237, "learning_rate": 7.270689093862105e-05, "loss": 0.0206, "step": 7650 }, { "grad_norm": 0.2549031674861908, "learning_rate": 7.263320351951374e-05, "loss": 0.0224, "step": 7660 }, { "grad_norm": 0.2148849070072174, "learning_rate": 7.255945422209227e-05, "loss": 0.0195, "step": 7670 }, { "grad_norm": 0.21542178094387054, "learning_rate": 7.248564324798437e-05, "loss": 0.0157, "step": 7680 }, { "grad_norm": 0.19588719308376312, "learning_rate": 7.241177079898644e-05, "loss": 0.0146, "step": 7690 }, { "grad_norm": 0.21754485368728638, "learning_rate": 7.233783707706295e-05, "loss": 0.0209, "step": 7700 }, { "grad_norm": 0.4288097321987152, "learning_rate": 7.226384228434586e-05, "loss": 0.0171, "step": 7710 }, { "grad_norm": 0.23700541257858276, "learning_rate": 7.21897866231341e-05, "loss": 0.0158, "step": 7720 }, { "grad_norm": 0.22373467683792114, "learning_rate": 7.211567029589303e-05, "loss": 0.0161, "step": 7730 }, { "grad_norm": 0.19316595792770386, "learning_rate": 7.204149350525387e-05, "loss": 0.0167, "step": 7740 }, { "grad_norm": 0.24723002314567566, "learning_rate": 7.196725645401309e-05, "loss": 0.0174, "step": 7750 }, { "grad_norm": 0.2575382888317108, "learning_rate": 7.1892959345132e-05, "loss": 0.0167, "step": 7760 }, { "grad_norm": 0.5586070418357849, "learning_rate": 7.181860238173605e-05, "loss": 0.0182, "step": 7770 }, { "grad_norm": 0.3310788869857788, "learning_rate": 7.174418576711432e-05, "loss": 0.0162, "step": 7780 }, { "grad_norm": 0.21224679052829742, "learning_rate": 7.1669709704719e-05, "loss": 0.0165, "step": 7790 }, { "grad_norm": 0.1979682892560959, "learning_rate": 7.159517439816481e-05, "loss": 0.0169, "step": 7800 }, { "grad_norm": 0.2028234452009201, "learning_rate": 7.152058005122842e-05, "loss": 0.0161, "step": 7810 }, { "grad_norm": 0.2093898057937622, "learning_rate": 7.144592686784793e-05, "loss": 0.0134, "step": 7820 }, { "grad_norm": 0.24137136340141296, "learning_rate": 7.137121505212229e-05, "loss": 0.0174, "step": 7830 }, { "grad_norm": 0.15440595149993896, "learning_rate": 7.129644480831077e-05, "loss": 0.0161, "step": 7840 }, { "grad_norm": 0.23603911697864532, "learning_rate": 7.122161634083234e-05, "loss": 0.0162, "step": 7850 }, { "grad_norm": 0.3225100636482239, "learning_rate": 7.114672985426516e-05, "loss": 0.0166, "step": 7860 }, { "grad_norm": 0.21036957204341888, "learning_rate": 7.107178555334606e-05, "loss": 0.0154, "step": 7870 }, { "grad_norm": 0.3252936601638794, "learning_rate": 7.099678364296989e-05, "loss": 0.0169, "step": 7880 }, { "grad_norm": 0.19355446100234985, "learning_rate": 7.0921724328189e-05, "loss": 0.0172, "step": 7890 }, { "grad_norm": 0.2391633242368698, "learning_rate": 7.084660781421268e-05, "loss": 0.0159, "step": 7900 }, { "grad_norm": 0.2408432960510254, "learning_rate": 7.077143430640662e-05, "loss": 0.0167, "step": 7910 }, { "grad_norm": 0.18131977319717407, "learning_rate": 7.069620401029232e-05, "loss": 0.0147, "step": 7920 }, { "grad_norm": 0.2135278731584549, "learning_rate": 7.062091713154655e-05, "loss": 0.018, "step": 7930 }, { "grad_norm": 0.2346886694431305, "learning_rate": 7.054557387600075e-05, "loss": 0.0181, "step": 7940 }, { "grad_norm": 0.25148677825927734, "learning_rate": 7.04701744496405e-05, "loss": 0.0155, "step": 7950 }, { "grad_norm": 0.14919376373291016, "learning_rate": 7.039471905860495e-05, "loss": 0.0169, "step": 7960 }, { "grad_norm": 0.17660413682460785, "learning_rate": 7.031920790918628e-05, "loss": 0.0183, "step": 7970 }, { "grad_norm": 0.23157574236392975, "learning_rate": 7.024364120782906e-05, "loss": 0.0162, "step": 7980 }, { "grad_norm": 0.17774859070777893, "learning_rate": 7.016801916112978e-05, "loss": 0.0167, "step": 7990 }, { "grad_norm": 0.15804070234298706, "learning_rate": 7.009234197583623e-05, "loss": 0.0139, "step": 8000 }, { "grad_norm": 0.15397830307483673, "learning_rate": 7.001660985884692e-05, "loss": 0.0128, "step": 8010 }, { "grad_norm": 0.17988398671150208, "learning_rate": 6.994082301721063e-05, "loss": 0.0151, "step": 8020 }, { "grad_norm": 0.2010420560836792, "learning_rate": 6.986498165812563e-05, "loss": 0.0138, "step": 8030 }, { "grad_norm": 0.14748282730579376, "learning_rate": 6.978908598893932e-05, "loss": 0.0145, "step": 8040 }, { "grad_norm": 0.22533001005649567, "learning_rate": 6.971313621714756e-05, "loss": 0.0148, "step": 8050 }, { "grad_norm": 0.22773291170597076, "learning_rate": 6.96371325503941e-05, "loss": 0.0163, "step": 8060 }, { "grad_norm": 0.2609845995903015, "learning_rate": 6.956107519647014e-05, "loss": 0.0156, "step": 8070 }, { "grad_norm": 0.2427939921617508, "learning_rate": 6.94849643633135e-05, "loss": 0.0184, "step": 8080 }, { "grad_norm": 0.29630377888679504, "learning_rate": 6.940880025900834e-05, "loss": 0.0193, "step": 8090 }, { "grad_norm": 0.20500099658966064, "learning_rate": 6.933258309178438e-05, "loss": 0.0161, "step": 8100 }, { "grad_norm": 0.17892125248908997, "learning_rate": 6.925631307001646e-05, "loss": 0.0142, "step": 8110 }, { "grad_norm": 0.169921413064003, "learning_rate": 6.91799904022239e-05, "loss": 0.0162, "step": 8120 }, { "grad_norm": 0.2468617707490921, "learning_rate": 6.910361529706997e-05, "loss": 0.0159, "step": 8130 }, { "grad_norm": 0.20839247107505798, "learning_rate": 6.902718796336131e-05, "loss": 0.0174, "step": 8140 }, { "grad_norm": 0.19165746867656708, "learning_rate": 6.895070861004729e-05, "loss": 0.0192, "step": 8150 }, { "grad_norm": 0.21431583166122437, "learning_rate": 6.887417744621956e-05, "loss": 0.0158, "step": 8160 }, { "grad_norm": 0.2116219401359558, "learning_rate": 6.87975946811114e-05, "loss": 0.0136, "step": 8170 }, { "grad_norm": 0.23228122293949127, "learning_rate": 6.872096052409718e-05, "loss": 0.0161, "step": 8180 }, { "grad_norm": 0.21339166164398193, "learning_rate": 6.864427518469174e-05, "loss": 0.0169, "step": 8190 }, { "grad_norm": 0.20013035833835602, "learning_rate": 6.856753887254986e-05, "loss": 0.0133, "step": 8200 }, { "grad_norm": 0.169180765748024, "learning_rate": 6.849075179746572e-05, "loss": 0.0137, "step": 8210 }, { "grad_norm": 0.23158656060695648, "learning_rate": 6.841391416937221e-05, "loss": 0.0129, "step": 8220 }, { "grad_norm": 0.29065436124801636, "learning_rate": 6.833702619834053e-05, "loss": 0.0138, "step": 8230 }, { "grad_norm": 0.18044963479042053, "learning_rate": 6.82600880945794e-05, "loss": 0.0148, "step": 8240 }, { "grad_norm": 0.1538570672273636, "learning_rate": 6.818310006843468e-05, "loss": 0.0152, "step": 8250 }, { "grad_norm": 0.20727285742759705, "learning_rate": 6.810606233038868e-05, "loss": 0.0148, "step": 8260 }, { "grad_norm": 0.22833110392093658, "learning_rate": 6.802897509105966e-05, "loss": 0.0146, "step": 8270 }, { "grad_norm": 0.23356859385967255, "learning_rate": 6.79518385612012e-05, "loss": 0.0128, "step": 8280 }, { "grad_norm": 0.25582438707351685, "learning_rate": 6.787465295170157e-05, "loss": 0.0162, "step": 8290 }, { "grad_norm": 0.19632817804813385, "learning_rate": 6.779741847358332e-05, "loss": 0.0164, "step": 8300 }, { "grad_norm": 0.3011960983276367, "learning_rate": 6.772013533800256e-05, "loss": 0.0164, "step": 8310 }, { "grad_norm": 0.1842845231294632, "learning_rate": 6.764280375624843e-05, "loss": 0.0132, "step": 8320 }, { "grad_norm": 0.14166337251663208, "learning_rate": 6.756542393974252e-05, "loss": 0.0137, "step": 8330 }, { "grad_norm": 0.25268304347991943, "learning_rate": 6.748799610003828e-05, "loss": 0.0141, "step": 8340 }, { "grad_norm": 0.3172411024570465, "learning_rate": 6.741052044882048e-05, "loss": 0.0168, "step": 8350 }, { "grad_norm": 0.1464390754699707, "learning_rate": 6.73329971979046e-05, "loss": 0.0147, "step": 8360 }, { "grad_norm": 0.18174967169761658, "learning_rate": 6.725542655923625e-05, "loss": 0.0119, "step": 8370 }, { "grad_norm": 0.16886651515960693, "learning_rate": 6.717780874489057e-05, "loss": 0.0115, "step": 8380 }, { "grad_norm": 0.24811872839927673, "learning_rate": 6.710014396707172e-05, "loss": 0.0147, "step": 8390 }, { "grad_norm": 0.22273807227611542, "learning_rate": 6.702243243811221e-05, "loss": 0.016, "step": 8400 }, { "grad_norm": 0.22023224830627441, "learning_rate": 6.694467437047244e-05, "loss": 0.0163, "step": 8410 }, { "grad_norm": 0.35881307721138, "learning_rate": 6.686686997673997e-05, "loss": 0.0139, "step": 8420 }, { "grad_norm": 0.1732298880815506, "learning_rate": 6.678901946962903e-05, "loss": 0.0137, "step": 8430 }, { "grad_norm": 0.15345846116542816, "learning_rate": 6.671112306197996e-05, "loss": 0.014, "step": 8440 }, { "grad_norm": 0.2203318327665329, "learning_rate": 6.663318096675854e-05, "loss": 0.0177, "step": 8450 }, { "grad_norm": 0.22170431911945343, "learning_rate": 6.655519339705552e-05, "loss": 0.0144, "step": 8460 }, { "grad_norm": 0.2769310176372528, "learning_rate": 6.647716056608588e-05, "loss": 0.0141, "step": 8470 }, { "grad_norm": 0.2043648064136505, "learning_rate": 6.639908268718843e-05, "loss": 0.0148, "step": 8480 }, { "grad_norm": 0.1574820578098297, "learning_rate": 6.632095997382514e-05, "loss": 0.0137, "step": 8490 }, { "grad_norm": 0.2324148714542389, "learning_rate": 6.624279263958047e-05, "loss": 0.016, "step": 8500 }, { "grad_norm": 0.30512377619743347, "learning_rate": 6.616458089816097e-05, "loss": 0.0151, "step": 8510 }, { "grad_norm": 0.12082749605178833, "learning_rate": 6.608632496339454e-05, "loss": 0.0156, "step": 8520 }, { "grad_norm": 0.21001307666301727, "learning_rate": 6.600802504922988e-05, "loss": 0.0147, "step": 8530 }, { "grad_norm": 0.16640181839466095, "learning_rate": 6.592968136973604e-05, "loss": 0.0159, "step": 8540 }, { "grad_norm": 0.17555391788482666, "learning_rate": 6.585129413910159e-05, "loss": 0.0119, "step": 8550 }, { "grad_norm": 0.23169393837451935, "learning_rate": 6.577286357163424e-05, "loss": 0.0147, "step": 8560 }, { "grad_norm": 0.38672158122062683, "learning_rate": 6.569438988176018e-05, "loss": 0.0171, "step": 8570 }, { "grad_norm": 0.2521936893463135, "learning_rate": 6.561587328402347e-05, "loss": 0.0142, "step": 8580 }, { "grad_norm": 0.3228203058242798, "learning_rate": 6.553731399308549e-05, "loss": 0.0132, "step": 8590 }, { "grad_norm": 0.2778136432170868, "learning_rate": 6.545871222372436e-05, "loss": 0.0137, "step": 8600 }, { "grad_norm": 0.19902174174785614, "learning_rate": 6.538006819083426e-05, "loss": 0.0175, "step": 8610 }, { "grad_norm": 0.15141315758228302, "learning_rate": 6.530138210942505e-05, "loss": 0.016, "step": 8620 }, { "grad_norm": 0.1684686243534088, "learning_rate": 6.522265419462141e-05, "loss": 0.0157, "step": 8630 }, { "grad_norm": 0.18699170649051666, "learning_rate": 6.514388466166248e-05, "loss": 0.0141, "step": 8640 }, { "grad_norm": 0.23442110419273376, "learning_rate": 6.506507372590119e-05, "loss": 0.0158, "step": 8650 }, { "grad_norm": 0.5508381724357605, "learning_rate": 6.498622160280355e-05, "loss": 0.0171, "step": 8660 }, { "grad_norm": 0.27397122979164124, "learning_rate": 6.490732850794832e-05, "loss": 0.0167, "step": 8670 }, { "grad_norm": 0.20226314663887024, "learning_rate": 6.482839465702616e-05, "loss": 0.0175, "step": 8680 }, { "grad_norm": 0.22117167711257935, "learning_rate": 6.474942026583923e-05, "loss": 0.0138, "step": 8690 }, { "grad_norm": 0.20024767518043518, "learning_rate": 6.467040555030052e-05, "loss": 0.0126, "step": 8700 }, { "grad_norm": 0.20151349902153015, "learning_rate": 6.459135072643321e-05, "loss": 0.0146, "step": 8710 }, { "grad_norm": 0.17252632975578308, "learning_rate": 6.451225601037019e-05, "loss": 0.0143, "step": 8720 }, { "grad_norm": 0.15924711525440216, "learning_rate": 6.443312161835338e-05, "loss": 0.0143, "step": 8730 }, { "grad_norm": 0.13117161393165588, "learning_rate": 6.43539477667332e-05, "loss": 0.0144, "step": 8740 }, { "grad_norm": 0.19753259420394897, "learning_rate": 6.427473467196793e-05, "loss": 0.0123, "step": 8750 }, { "grad_norm": 0.21458126604557037, "learning_rate": 6.419548255062315e-05, "loss": 0.0128, "step": 8760 }, { "grad_norm": 0.25365519523620605, "learning_rate": 6.411619161937112e-05, "loss": 0.0159, "step": 8770 }, { "grad_norm": 0.22291269898414612, "learning_rate": 6.403686209499022e-05, "loss": 0.0165, "step": 8780 }, { "grad_norm": 0.16859787702560425, "learning_rate": 6.395749419436437e-05, "loss": 0.0156, "step": 8790 }, { "grad_norm": 0.15040329098701477, "learning_rate": 6.387808813448234e-05, "loss": 0.0117, "step": 8800 }, { "grad_norm": 0.2395641803741455, "learning_rate": 6.37986441324373e-05, "loss": 0.0118, "step": 8810 }, { "grad_norm": 0.14066743850708008, "learning_rate": 6.37191624054261e-05, "loss": 0.0139, "step": 8820 }, { "grad_norm": 0.21710516512393951, "learning_rate": 6.363964317074872e-05, "loss": 0.0156, "step": 8830 }, { "grad_norm": 0.2219642996788025, "learning_rate": 6.356008664580776e-05, "loss": 0.0137, "step": 8840 }, { "grad_norm": 0.17102472484111786, "learning_rate": 6.348049304810771e-05, "loss": 0.0144, "step": 8850 }, { "grad_norm": 0.19437281787395477, "learning_rate": 6.340086259525442e-05, "loss": 0.0136, "step": 8860 }, { "grad_norm": 0.19022820889949799, "learning_rate": 6.332119550495448e-05, "loss": 0.0142, "step": 8870 }, { "grad_norm": 0.20764730870723724, "learning_rate": 6.324149199501473e-05, "loss": 0.0136, "step": 8880 }, { "grad_norm": 0.1867000311613083, "learning_rate": 6.316175228334146e-05, "loss": 0.0149, "step": 8890 }, { "grad_norm": 0.28641214966773987, "learning_rate": 6.308197658794003e-05, "loss": 0.0146, "step": 8900 }, { "grad_norm": 0.300482839345932, "learning_rate": 6.300216512691417e-05, "loss": 0.0153, "step": 8910 }, { "grad_norm": 0.18399882316589355, "learning_rate": 6.292231811846532e-05, "loss": 0.0142, "step": 8920 }, { "grad_norm": 0.23361952602863312, "learning_rate": 6.284243578089217e-05, "loss": 0.0171, "step": 8930 }, { "grad_norm": 0.31907209753990173, "learning_rate": 6.276251833258999e-05, "loss": 0.014, "step": 8940 }, { "grad_norm": 0.228986456990242, "learning_rate": 6.268256599205003e-05, "loss": 0.0137, "step": 8950 }, { "grad_norm": 0.26814061403274536, "learning_rate": 6.260257897785892e-05, "loss": 0.0137, "step": 8960 }, { "grad_norm": 0.23505839705467224, "learning_rate": 6.252255750869811e-05, "loss": 0.0149, "step": 8970 }, { "grad_norm": 0.2591574788093567, "learning_rate": 6.244250180334325e-05, "loss": 0.014, "step": 8980 }, { "grad_norm": 0.2925415337085724, "learning_rate": 6.236241208066356e-05, "loss": 0.0146, "step": 8990 }, { "grad_norm": 0.28943586349487305, "learning_rate": 6.228228855962133e-05, "loss": 0.0171, "step": 9000 }, { "grad_norm": 0.1855161190032959, "learning_rate": 6.220213145927115e-05, "loss": 0.0153, "step": 9010 }, { "grad_norm": 0.25652560591697693, "learning_rate": 6.212194099875951e-05, "loss": 0.0169, "step": 9020 }, { "grad_norm": 0.3078072667121887, "learning_rate": 6.204171739732405e-05, "loss": 0.0158, "step": 9030 }, { "grad_norm": 0.15821434557437897, "learning_rate": 6.196146087429303e-05, "loss": 0.0139, "step": 9040 }, { "grad_norm": 0.1915762573480606, "learning_rate": 6.188117164908474e-05, "loss": 0.0148, "step": 9050 }, { "grad_norm": 0.22909589111804962, "learning_rate": 6.180084994120684e-05, "loss": 0.0144, "step": 9060 }, { "grad_norm": 0.2742500901222229, "learning_rate": 6.17204959702558e-05, "loss": 0.0141, "step": 9070 }, { "grad_norm": 0.3247205913066864, "learning_rate": 6.164010995591635e-05, "loss": 0.0146, "step": 9080 }, { "grad_norm": 0.27257001399993896, "learning_rate": 6.155969211796076e-05, "loss": 0.0134, "step": 9090 }, { "grad_norm": 0.20286352932453156, "learning_rate": 6.147924267624829e-05, "loss": 0.0132, "step": 9100 }, { "grad_norm": 0.23123455047607422, "learning_rate": 6.13987618507247e-05, "loss": 0.0125, "step": 9110 }, { "grad_norm": 0.1758718341588974, "learning_rate": 6.131824986142147e-05, "loss": 0.0145, "step": 9120 }, { "grad_norm": 0.1888793706893921, "learning_rate": 6.123770692845529e-05, "loss": 0.0138, "step": 9130 }, { "grad_norm": 0.21240395307540894, "learning_rate": 6.11571332720275e-05, "loss": 0.0129, "step": 9140 }, { "grad_norm": 0.2523875832557678, "learning_rate": 6.107652911242336e-05, "loss": 0.0146, "step": 9150 }, { "grad_norm": 0.1263609230518341, "learning_rate": 6.0995894670011586e-05, "loss": 0.0142, "step": 9160 }, { "grad_norm": 0.17078685760498047, "learning_rate": 6.091523016524368e-05, "loss": 0.0142, "step": 9170 }, { "grad_norm": 0.13385668396949768, "learning_rate": 6.083453581865328e-05, "loss": 0.0123, "step": 9180 }, { "grad_norm": 0.22579160332679749, "learning_rate": 6.075381185085568e-05, "loss": 0.0148, "step": 9190 }, { "grad_norm": 0.2314026653766632, "learning_rate": 6.067305848254709e-05, "loss": 0.0148, "step": 9200 }, { "grad_norm": 0.19935566186904907, "learning_rate": 6.059227593450418e-05, "loss": 0.0171, "step": 9210 }, { "grad_norm": 0.16658079624176025, "learning_rate": 6.051146442758333e-05, "loss": 0.0147, "step": 9220 }, { "grad_norm": 0.14751195907592773, "learning_rate": 6.043062418272012e-05, "loss": 0.0134, "step": 9230 }, { "grad_norm": 0.13917523622512817, "learning_rate": 6.0349755420928666e-05, "loss": 0.0128, "step": 9240 }, { "grad_norm": 0.1302308440208435, "learning_rate": 6.0268858363301105e-05, "loss": 0.0147, "step": 9250 }, { "grad_norm": 0.20324723422527313, "learning_rate": 6.018793323100689e-05, "loss": 0.015, "step": 9260 }, { "grad_norm": 0.1756460815668106, "learning_rate": 6.0106980245292255e-05, "loss": 0.0147, "step": 9270 }, { "grad_norm": 0.17870934307575226, "learning_rate": 6.002599962747957e-05, "loss": 0.0143, "step": 9280 }, { "grad_norm": 0.14137393236160278, "learning_rate": 5.994499159896673e-05, "loss": 0.0139, "step": 9290 }, { "grad_norm": 0.40131819248199463, "learning_rate": 5.9863956381226607e-05, "loss": 0.0138, "step": 9300 }, { "grad_norm": 0.18956096470355988, "learning_rate": 5.9782894195806394e-05, "loss": 0.0121, "step": 9310 }, { "grad_norm": 0.15222765505313873, "learning_rate": 5.9701805264327004e-05, "loss": 0.0116, "step": 9320 }, { "grad_norm": 0.23643285036087036, "learning_rate": 5.96206898084825e-05, "loss": 0.0115, "step": 9330 }, { "grad_norm": 0.10720936954021454, "learning_rate": 5.953954805003942e-05, "loss": 0.017, "step": 9340 }, { "grad_norm": 0.2760174572467804, "learning_rate": 5.945838021083623e-05, "loss": 0.0127, "step": 9350 }, { "grad_norm": 0.17752394080162048, "learning_rate": 5.9377186512782714e-05, "loss": 0.0124, "step": 9360 }, { "grad_norm": 0.19948691129684448, "learning_rate": 5.929596717785935e-05, "loss": 0.012, "step": 9370 }, { "grad_norm": 0.16890588402748108, "learning_rate": 5.921472242811668e-05, "loss": 0.0118, "step": 9380 }, { "grad_norm": 0.17586396634578705, "learning_rate": 5.913345248567475e-05, "loss": 0.0128, "step": 9390 }, { "grad_norm": 0.20363761484622955, "learning_rate": 5.905215757272248e-05, "loss": 0.0126, "step": 9400 }, { "grad_norm": 0.19073963165283203, "learning_rate": 5.897083791151706e-05, "loss": 0.013, "step": 9410 }, { "grad_norm": 0.18090808391571045, "learning_rate": 5.888949372438336e-05, "loss": 0.0113, "step": 9420 }, { "grad_norm": 0.18554657697677612, "learning_rate": 5.8808125233713255e-05, "loss": 0.0124, "step": 9430 }, { "grad_norm": 0.24118909239768982, "learning_rate": 5.872673266196509e-05, "loss": 0.013, "step": 9440 }, { "grad_norm": 0.2634093165397644, "learning_rate": 5.864531623166305e-05, "loss": 0.012, "step": 9450 }, { "grad_norm": 0.23046427965164185, "learning_rate": 5.856387616539656e-05, "loss": 0.0152, "step": 9460 }, { "grad_norm": 0.26411688327789307, "learning_rate": 5.848241268581967e-05, "loss": 0.0122, "step": 9470 }, { "grad_norm": 0.1820603311061859, "learning_rate": 5.840092601565037e-05, "loss": 0.0119, "step": 9480 }, { "grad_norm": 0.21226665377616882, "learning_rate": 5.8319416377670144e-05, "loss": 0.0136, "step": 9490 }, { "grad_norm": 0.13919393718242645, "learning_rate": 5.82378839947232e-05, "loss": 0.0097, "step": 9500 }, { "grad_norm": 0.22376754879951477, "learning_rate": 5.815632908971599e-05, "loss": 0.0142, "step": 9510 }, { "grad_norm": 0.1878610998392105, "learning_rate": 5.80747518856165e-05, "loss": 0.0141, "step": 9520 }, { "grad_norm": 0.1755683273077011, "learning_rate": 5.799315260545367e-05, "loss": 0.0104, "step": 9530 }, { "grad_norm": 0.17256873846054077, "learning_rate": 5.791153147231686e-05, "loss": 0.012, "step": 9540 }, { "grad_norm": 0.17082209885120392, "learning_rate": 5.782988870935509e-05, "loss": 0.0117, "step": 9550 }, { "grad_norm": 0.24998190999031067, "learning_rate": 5.774822453977657e-05, "loss": 0.015, "step": 9560 }, { "grad_norm": 0.16367016732692719, "learning_rate": 5.7666539186848036e-05, "loss": 0.0138, "step": 9570 }, { "grad_norm": 0.2711407244205475, "learning_rate": 5.758483287389411e-05, "loss": 0.0135, "step": 9580 }, { "grad_norm": 0.24749813973903656, "learning_rate": 5.7503105824296735e-05, "loss": 0.0132, "step": 9590 }, { "grad_norm": 0.33168014883995056, "learning_rate": 5.742135826149453e-05, "loss": 0.0131, "step": 9600 }, { "grad_norm": 0.17513227462768555, "learning_rate": 5.7339590408982223e-05, "loss": 0.0122, "step": 9610 }, { "grad_norm": 0.1499035805463791, "learning_rate": 5.725780249031e-05, "loss": 0.0134, "step": 9620 }, { "grad_norm": 0.11795676499605179, "learning_rate": 5.717599472908292e-05, "loss": 0.0131, "step": 9630 }, { "grad_norm": 0.18526910245418549, "learning_rate": 5.7094167348960237e-05, "loss": 0.0129, "step": 9640 }, { "grad_norm": 0.18724125623703003, "learning_rate": 5.7012320573654945e-05, "loss": 0.0135, "step": 9650 }, { "grad_norm": 0.19242244958877563, "learning_rate": 5.693045462693295e-05, "loss": 0.0122, "step": 9660 }, { "grad_norm": 0.17911255359649658, "learning_rate": 5.684856973261266e-05, "loss": 0.0117, "step": 9670 }, { "grad_norm": 0.22174498438835144, "learning_rate": 5.6766666114564215e-05, "loss": 0.0109, "step": 9680 }, { "grad_norm": 0.13627007603645325, "learning_rate": 5.668474399670899e-05, "loss": 0.0122, "step": 9690 }, { "grad_norm": 0.1978825330734253, "learning_rate": 5.660280360301896e-05, "loss": 0.0119, "step": 9700 }, { "grad_norm": 0.15797913074493408, "learning_rate": 5.652084515751599e-05, "loss": 0.0124, "step": 9710 }, { "grad_norm": 0.2503998279571533, "learning_rate": 5.643886888427137e-05, "loss": 0.0123, "step": 9720 }, { "grad_norm": 0.34509292244911194, "learning_rate": 5.6356875007405074e-05, "loss": 0.011, "step": 9730 }, { "grad_norm": 0.2570599317550659, "learning_rate": 5.627486375108525e-05, "loss": 0.0151, "step": 9740 }, { "grad_norm": 0.20971865952014923, "learning_rate": 5.619283533952754e-05, "loss": 0.0139, "step": 9750 }, { "grad_norm": 0.23567263782024384, "learning_rate": 5.6110789996994474e-05, "loss": 0.013, "step": 9760 }, { "grad_norm": 0.170766219496727, "learning_rate": 5.602872794779491e-05, "loss": 0.0127, "step": 9770 }, { "grad_norm": 0.21201038360595703, "learning_rate": 5.594664941628334e-05, "loss": 0.0148, "step": 9780 }, { "grad_norm": 0.19467711448669434, "learning_rate": 5.5864554626859324e-05, "loss": 0.0139, "step": 9790 }, { "grad_norm": 0.18591301143169403, "learning_rate": 5.578244380396691e-05, "loss": 0.0103, "step": 9800 }, { "grad_norm": 0.15956811606884003, "learning_rate": 5.570031717209394e-05, "loss": 0.0106, "step": 9810 }, { "grad_norm": 0.14005182683467865, "learning_rate": 5.561817495577147e-05, "loss": 0.0119, "step": 9820 }, { "grad_norm": 0.15536241233348846, "learning_rate": 5.5536017379573215e-05, "loss": 0.0122, "step": 9830 }, { "grad_norm": 0.44114235043525696, "learning_rate": 5.545384466811483e-05, "loss": 0.0127, "step": 9840 }, { "grad_norm": 0.20538516342639923, "learning_rate": 5.5371657046053384e-05, "loss": 0.0135, "step": 9850 }, { "grad_norm": 0.14697428047657013, "learning_rate": 5.528945473808669e-05, "loss": 0.0131, "step": 9860 }, { "grad_norm": 0.17225128412246704, "learning_rate": 5.520723796895272e-05, "loss": 0.0129, "step": 9870 }, { "grad_norm": 0.2053793966770172, "learning_rate": 5.512500696342897e-05, "loss": 0.0127, "step": 9880 }, { "grad_norm": 0.1962839961051941, "learning_rate": 5.504276194633188e-05, "loss": 0.0137, "step": 9890 }, { "grad_norm": 0.25017449259757996, "learning_rate": 5.49605031425162e-05, "loss": 0.013, "step": 9900 }, { "grad_norm": 0.24997183680534363, "learning_rate": 5.487823077687434e-05, "loss": 0.0113, "step": 9910 }, { "grad_norm": 0.14494562149047852, "learning_rate": 5.4795945074335806e-05, "loss": 0.0141, "step": 9920 }, { "grad_norm": 0.17810435593128204, "learning_rate": 5.471364625986657e-05, "loss": 0.0146, "step": 9930 }, { "grad_norm": 0.14798429608345032, "learning_rate": 5.463133455846845e-05, "loss": 0.0127, "step": 9940 }, { "grad_norm": 0.1260615438222885, "learning_rate": 5.4549010195178505e-05, "loss": 0.0099, "step": 9950 }, { "grad_norm": 0.1621381938457489, "learning_rate": 5.446667339506838e-05, "loss": 0.011, "step": 9960 }, { "grad_norm": 0.23774299025535583, "learning_rate": 5.4384324383243756e-05, "loss": 0.0109, "step": 9970 }, { "grad_norm": 0.2302779108285904, "learning_rate": 5.430196338484368e-05, "loss": 0.0121, "step": 9980 }, { "grad_norm": 0.15244658291339874, "learning_rate": 5.4219590625039975e-05, "loss": 0.0131, "step": 9990 }, { "grad_norm": 0.1994250863790512, "learning_rate": 5.413720632903664e-05, "loss": 0.0116, "step": 10000 }, { "grad_norm": 0.14933815598487854, "learning_rate": 5.405481072206917e-05, "loss": 0.0104, "step": 10010 }, { "grad_norm": 0.15649552643299103, "learning_rate": 5.397240402940402e-05, "loss": 0.0129, "step": 10020 }, { "grad_norm": 0.26181524991989136, "learning_rate": 5.388998647633794e-05, "loss": 0.0114, "step": 10030 }, { "grad_norm": 0.17449145019054413, "learning_rate": 5.380755828819737e-05, "loss": 0.0106, "step": 10040 }, { "grad_norm": 0.22064201533794403, "learning_rate": 5.3725119690337846e-05, "loss": 0.0119, "step": 10050 }, { "grad_norm": 0.27196648716926575, "learning_rate": 5.3642670908143324e-05, "loss": 0.0118, "step": 10060 }, { "grad_norm": 0.18223531544208527, "learning_rate": 5.356021216702562e-05, "loss": 0.0118, "step": 10070 }, { "grad_norm": 0.21584288775920868, "learning_rate": 5.347774369242381e-05, "loss": 0.0124, "step": 10080 }, { "grad_norm": 0.14566057920455933, "learning_rate": 5.3395265709803545e-05, "loss": 0.0121, "step": 10090 }, { "grad_norm": 0.1772134006023407, "learning_rate": 5.331277844465647e-05, "loss": 0.0122, "step": 10100 }, { "grad_norm": 0.11947708576917648, "learning_rate": 5.323028212249963e-05, "loss": 0.0109, "step": 10110 }, { "grad_norm": 0.2737846374511719, "learning_rate": 5.314777696887481e-05, "loss": 0.0102, "step": 10120 }, { "grad_norm": 0.2330601066350937, "learning_rate": 5.306526320934796e-05, "loss": 0.0098, "step": 10130 }, { "grad_norm": 0.2621106505393982, "learning_rate": 5.298274106950854e-05, "loss": 0.0113, "step": 10140 }, { "grad_norm": 0.2515595257282257, "learning_rate": 5.290021077496893e-05, "loss": 0.0118, "step": 10150 }, { "grad_norm": 0.14646384119987488, "learning_rate": 5.2817672551363816e-05, "loss": 0.01, "step": 10160 }, { "grad_norm": 0.2486342489719391, "learning_rate": 5.273512662434952e-05, "loss": 0.0126, "step": 10170 }, { "grad_norm": 0.2677910029888153, "learning_rate": 5.265257321960349e-05, "loss": 0.012, "step": 10180 }, { "grad_norm": 0.19955258071422577, "learning_rate": 5.257001256282357e-05, "loss": 0.0096, "step": 10190 }, { "grad_norm": 0.15984804928302765, "learning_rate": 5.248744487972742e-05, "loss": 0.011, "step": 10200 }, { "grad_norm": 0.2594794034957886, "learning_rate": 5.240487039605196e-05, "loss": 0.0126, "step": 10210 }, { "grad_norm": 0.21173891425132751, "learning_rate": 5.232228933755267e-05, "loss": 0.0136, "step": 10220 }, { "grad_norm": 0.15130645036697388, "learning_rate": 5.2239701930003006e-05, "loss": 0.0094, "step": 10230 }, { "grad_norm": 0.1467563658952713, "learning_rate": 5.215710839919379e-05, "loss": 0.0114, "step": 10240 }, { "grad_norm": 0.16668064892292023, "learning_rate": 5.207450897093257e-05, "loss": 0.0121, "step": 10250 }, { "grad_norm": 0.18487215042114258, "learning_rate": 5.1991903871043046e-05, "loss": 0.0112, "step": 10260 }, { "grad_norm": 0.215226948261261, "learning_rate": 5.190929332536439e-05, "loss": 0.0115, "step": 10270 }, { "grad_norm": 0.2345028966665268, "learning_rate": 5.182667755975071e-05, "loss": 0.0117, "step": 10280 }, { "grad_norm": 0.1739572137594223, "learning_rate": 5.1744056800070315e-05, "loss": 0.0119, "step": 10290 }, { "grad_norm": 0.22923319041728973, "learning_rate": 5.166143127220524e-05, "loss": 0.0133, "step": 10300 }, { "grad_norm": 0.44578859210014343, "learning_rate": 5.1578801202050485e-05, "loss": 0.0109, "step": 10310 }, { "grad_norm": 0.15892961621284485, "learning_rate": 5.149616681551355e-05, "loss": 0.0125, "step": 10320 }, { "grad_norm": 0.24809853732585907, "learning_rate": 5.141352833851367e-05, "loss": 0.0125, "step": 10330 }, { "grad_norm": 0.25976279377937317, "learning_rate": 5.1330885996981285e-05, "loss": 0.0114, "step": 10340 }, { "grad_norm": 0.24694496393203735, "learning_rate": 5.124824001685741e-05, "loss": 0.0145, "step": 10350 }, { "grad_norm": 0.1293613165616989, "learning_rate": 5.116559062409298e-05, "loss": 0.0114, "step": 10360 }, { "grad_norm": 0.1778794527053833, "learning_rate": 5.10829380446483e-05, "loss": 0.0113, "step": 10370 }, { "grad_norm": 0.26746320724487305, "learning_rate": 5.100028250449235e-05, "loss": 0.0106, "step": 10380 }, { "grad_norm": 0.25291013717651367, "learning_rate": 5.0917624229602234e-05, "loss": 0.0121, "step": 10390 }, { "grad_norm": 0.20073992013931274, "learning_rate": 5.0834963445962524e-05, "loss": 0.0115, "step": 10400 }, { "grad_norm": 0.40504828095436096, "learning_rate": 5.075230037956461e-05, "loss": 0.01, "step": 10410 }, { "grad_norm": 0.19573934376239777, "learning_rate": 5.0669635256406213e-05, "loss": 0.0116, "step": 10420 }, { "grad_norm": 0.21186774969100952, "learning_rate": 5.058696830249058e-05, "loss": 0.0112, "step": 10430 }, { "grad_norm": 0.15050864219665527, "learning_rate": 5.050429974382602e-05, "loss": 0.0099, "step": 10440 }, { "grad_norm": 0.15649381279945374, "learning_rate": 5.042162980642523e-05, "loss": 0.0119, "step": 10450 }, { "grad_norm": 0.1823073923587799, "learning_rate": 5.033895871630462e-05, "loss": 0.0106, "step": 10460 }, { "grad_norm": 0.16103525459766388, "learning_rate": 5.025628669948386e-05, "loss": 0.0119, "step": 10470 }, { "grad_norm": 0.3252846896648407, "learning_rate": 5.017361398198502e-05, "loss": 0.0111, "step": 10480 }, { "grad_norm": 0.1312829703092575, "learning_rate": 5.009094078983221e-05, "loss": 0.013, "step": 10490 }, { "grad_norm": 0.16311180591583252, "learning_rate": 5.000826734905073e-05, "loss": 0.0149, "step": 10500 }, { "grad_norm": 0.2581283450126648, "learning_rate": 4.9925593885666645e-05, "loss": 0.0115, "step": 10510 }, { "grad_norm": 0.12821823358535767, "learning_rate": 4.984292062570602e-05, "loss": 0.0121, "step": 10520 }, { "grad_norm": 0.12288004159927368, "learning_rate": 4.976024779519442e-05, "loss": 0.0097, "step": 10530 }, { "grad_norm": 0.18156060576438904, "learning_rate": 4.9677575620156194e-05, "loss": 0.0114, "step": 10540 }, { "grad_norm": 0.17647844552993774, "learning_rate": 4.959490432661391e-05, "loss": 0.0112, "step": 10550 }, { "grad_norm": 0.11945334821939468, "learning_rate": 4.9512234140587726e-05, "loss": 0.0108, "step": 10560 }, { "grad_norm": 0.1570923626422882, "learning_rate": 4.942956528809477e-05, "loss": 0.0109, "step": 10570 }, { "grad_norm": 0.13661563396453857, "learning_rate": 4.934689799514854e-05, "loss": 0.0107, "step": 10580 }, { "grad_norm": 0.6402286887168884, "learning_rate": 4.926423248775827e-05, "loss": 0.0113, "step": 10590 }, { "grad_norm": 0.14856669306755066, "learning_rate": 4.918156899192826e-05, "loss": 0.0095, "step": 10600 }, { "grad_norm": 0.2545925974845886, "learning_rate": 4.909890773365738e-05, "loss": 0.0101, "step": 10610 }, { "grad_norm": 0.243443101644516, "learning_rate": 4.9016248938938344e-05, "loss": 0.0122, "step": 10620 }, { "grad_norm": 0.22133122384548187, "learning_rate": 4.8933592833757156e-05, "loss": 0.0134, "step": 10630 }, { "grad_norm": 0.2558547556400299, "learning_rate": 4.8850939644092435e-05, "loss": 0.0128, "step": 10640 }, { "grad_norm": 0.21341577172279358, "learning_rate": 4.876828959591485e-05, "loss": 0.0119, "step": 10650 }, { "grad_norm": 0.2044287770986557, "learning_rate": 4.8685642915186474e-05, "loss": 0.0091, "step": 10660 }, { "grad_norm": 0.155178040266037, "learning_rate": 4.860299982786018e-05, "loss": 0.0099, "step": 10670 }, { "grad_norm": 0.19052208960056305, "learning_rate": 4.852036055987901e-05, "loss": 0.0123, "step": 10680 }, { "grad_norm": 0.19072329998016357, "learning_rate": 4.843772533717558e-05, "loss": 0.0116, "step": 10690 }, { "grad_norm": 0.2663835883140564, "learning_rate": 4.835509438567142e-05, "loss": 0.0116, "step": 10700 }, { "grad_norm": 0.17421448230743408, "learning_rate": 4.827246793127639e-05, "loss": 0.0102, "step": 10710 }, { "grad_norm": 0.1984112560749054, "learning_rate": 4.818984619988807e-05, "loss": 0.0119, "step": 10720 }, { "grad_norm": 0.1903335452079773, "learning_rate": 4.810722941739115e-05, "loss": 0.0115, "step": 10730 }, { "grad_norm": 0.4528978168964386, "learning_rate": 4.8024617809656684e-05, "loss": 0.0115, "step": 10740 }, { "grad_norm": 0.2015117108821869, "learning_rate": 4.794201160254171e-05, "loss": 0.0106, "step": 10750 }, { "grad_norm": 0.25602856278419495, "learning_rate": 4.785941102188844e-05, "loss": 0.0106, "step": 10760 }, { "grad_norm": 0.2576582431793213, "learning_rate": 4.7776816293523686e-05, "loss": 0.0105, "step": 10770 }, { "grad_norm": 0.17176516354084015, "learning_rate": 4.769422764325832e-05, "loss": 0.0097, "step": 10780 }, { "grad_norm": 0.19414331018924713, "learning_rate": 4.76116452968865e-05, "loss": 0.0118, "step": 10790 }, { "grad_norm": 0.16774576902389526, "learning_rate": 4.752906948018525e-05, "loss": 0.0113, "step": 10800 }, { "grad_norm": 0.21350081264972687, "learning_rate": 4.7446500418913684e-05, "loss": 0.0123, "step": 10810 }, { "grad_norm": 0.16081300377845764, "learning_rate": 4.736393833881247e-05, "loss": 0.0116, "step": 10820 }, { "grad_norm": 0.19511790573596954, "learning_rate": 4.7281383465603194e-05, "loss": 0.0104, "step": 10830 }, { "grad_norm": 0.17713306844234467, "learning_rate": 4.71988360249877e-05, "loss": 0.0108, "step": 10840 }, { "grad_norm": 0.20568551123142242, "learning_rate": 4.7116296242647554e-05, "loss": 0.014, "step": 10850 }, { "grad_norm": 0.1911221593618393, "learning_rate": 4.703376434424336e-05, "loss": 0.0113, "step": 10860 }, { "grad_norm": 0.1673281490802765, "learning_rate": 4.695124055541421e-05, "loss": 0.0108, "step": 10870 }, { "grad_norm": 0.14203645288944244, "learning_rate": 4.6868725101776934e-05, "loss": 0.0111, "step": 10880 }, { "grad_norm": 0.15682800114154816, "learning_rate": 4.678621820892567e-05, "loss": 0.0105, "step": 10890 }, { "grad_norm": 0.19690567255020142, "learning_rate": 4.670372010243111e-05, "loss": 0.0107, "step": 10900 }, { "grad_norm": 0.20088303089141846, "learning_rate": 4.662123100783992e-05, "loss": 0.0091, "step": 10910 }, { "grad_norm": 0.16964007914066315, "learning_rate": 4.653875115067415e-05, "loss": 0.0112, "step": 10920 }, { "grad_norm": 0.15116742253303528, "learning_rate": 4.6456280756430545e-05, "loss": 0.0112, "step": 10930 }, { "grad_norm": 0.24841725826263428, "learning_rate": 4.637382005058004e-05, "loss": 0.011, "step": 10940 }, { "grad_norm": 0.2837328612804413, "learning_rate": 4.629136925856705e-05, "loss": 0.0093, "step": 10950 }, { "grad_norm": 0.1390158087015152, "learning_rate": 4.6208928605808895e-05, "loss": 0.0101, "step": 10960 }, { "grad_norm": 0.2120196521282196, "learning_rate": 4.612649831769519e-05, "loss": 0.0126, "step": 10970 }, { "grad_norm": 0.11088117957115173, "learning_rate": 4.604407861958715e-05, "loss": 0.0094, "step": 10980 }, { "grad_norm": 0.1297864019870758, "learning_rate": 4.5961669736817114e-05, "loss": 0.0108, "step": 10990 }, { "grad_norm": 0.1555159091949463, "learning_rate": 4.5879271894687814e-05, "loss": 0.0099, "step": 11000 }, { "grad_norm": 0.22824782133102417, "learning_rate": 4.5796885318471826e-05, "loss": 0.0108, "step": 11010 }, { "grad_norm": 0.21547405421733856, "learning_rate": 4.571451023341086e-05, "loss": 0.0128, "step": 11020 }, { "grad_norm": 0.13591019809246063, "learning_rate": 4.563214686471527e-05, "loss": 0.011, "step": 11030 }, { "grad_norm": 0.121599480509758, "learning_rate": 4.5549795437563365e-05, "loss": 0.009, "step": 11040 }, { "grad_norm": 0.1390773504972458, "learning_rate": 4.546745617710081e-05, "loss": 0.0116, "step": 11050 }, { "grad_norm": 0.19202475249767303, "learning_rate": 4.5385129308440014e-05, "loss": 0.0113, "step": 11060 }, { "grad_norm": 0.22333449125289917, "learning_rate": 4.530281505665944e-05, "loss": 0.0112, "step": 11070 }, { "grad_norm": 0.13204337656497955, "learning_rate": 4.5220513646803134e-05, "loss": 0.0108, "step": 11080 }, { "grad_norm": 0.1430213451385498, "learning_rate": 4.513822530388003e-05, "loss": 0.0099, "step": 11090 }, { "grad_norm": 0.10505180060863495, "learning_rate": 4.5055950252863296e-05, "loss": 0.0097, "step": 11100 }, { "grad_norm": 0.1555517315864563, "learning_rate": 4.4973688718689803e-05, "loss": 0.0105, "step": 11110 }, { "grad_norm": 0.18975730240345, "learning_rate": 4.4891440926259406e-05, "loss": 0.0112, "step": 11120 }, { "grad_norm": 0.18583598732948303, "learning_rate": 4.480920710043443e-05, "loss": 0.0116, "step": 11130 }, { "grad_norm": 0.2092936486005783, "learning_rate": 4.4726987466039044e-05, "loss": 0.0104, "step": 11140 }, { "grad_norm": 0.12778379023075104, "learning_rate": 4.46447822478586e-05, "loss": 0.0088, "step": 11150 }, { "grad_norm": 0.213963121175766, "learning_rate": 4.4562591670638974e-05, "loss": 0.0114, "step": 11160 }, { "grad_norm": 0.1884763389825821, "learning_rate": 4.4480415959086105e-05, "loss": 0.0106, "step": 11170 }, { "grad_norm": 0.11079375445842743, "learning_rate": 4.439825533786522e-05, "loss": 0.0087, "step": 11180 }, { "grad_norm": 0.1653805822134018, "learning_rate": 4.431611003160035e-05, "loss": 0.0092, "step": 11190 }, { "grad_norm": 0.18368390202522278, "learning_rate": 4.4233980264873636e-05, "loss": 0.0109, "step": 11200 }, { "grad_norm": 0.20007635653018951, "learning_rate": 4.4151866262224684e-05, "loss": 0.0098, "step": 11210 }, { "grad_norm": 0.1305672824382782, "learning_rate": 4.406976824815006e-05, "loss": 0.0089, "step": 11220 }, { "grad_norm": 0.1783028542995453, "learning_rate": 4.3987686447102595e-05, "loss": 0.014, "step": 11230 }, { "grad_norm": 0.12125745415687561, "learning_rate": 4.3905621083490804e-05, "loss": 0.0086, "step": 11240 }, { "grad_norm": 0.14725035429000854, "learning_rate": 4.3823572381678286e-05, "loss": 0.0112, "step": 11250 }, { "grad_norm": 0.2169550210237503, "learning_rate": 4.374154056598301e-05, "loss": 0.0114, "step": 11260 }, { "grad_norm": 0.18034595251083374, "learning_rate": 4.3659525860676845e-05, "loss": 0.0106, "step": 11270 }, { "grad_norm": 0.1690545678138733, "learning_rate": 4.3577528489984854e-05, "loss": 0.0091, "step": 11280 }, { "grad_norm": 0.1983601599931717, "learning_rate": 4.349554867808476e-05, "loss": 0.0102, "step": 11290 }, { "grad_norm": 0.14944985508918762, "learning_rate": 4.34135866491062e-05, "loss": 0.0083, "step": 11300 }, { "grad_norm": 0.15976741909980774, "learning_rate": 4.333164262713022e-05, "loss": 0.0097, "step": 11310 }, { "grad_norm": 0.10774119198322296, "learning_rate": 4.324971683618868e-05, "loss": 0.0097, "step": 11320 }, { "grad_norm": 0.1354939192533493, "learning_rate": 4.316780950026354e-05, "loss": 0.0091, "step": 11330 }, { "grad_norm": 0.19260857999324799, "learning_rate": 4.308592084328637e-05, "loss": 0.0105, "step": 11340 }, { "grad_norm": 0.23097765445709229, "learning_rate": 4.3004051089137576e-05, "loss": 0.0131, "step": 11350 }, { "grad_norm": 0.1865207999944687, "learning_rate": 4.292220046164597e-05, "loss": 0.0116, "step": 11360 }, { "grad_norm": 0.21810993552207947, "learning_rate": 4.2840369184588035e-05, "loss": 0.0106, "step": 11370 }, { "grad_norm": 0.21561044454574585, "learning_rate": 4.2758557481687345e-05, "loss": 0.011, "step": 11380 }, { "grad_norm": 0.12443456798791885, "learning_rate": 4.267676557661403e-05, "loss": 0.0096, "step": 11390 }, { "grad_norm": 0.17131224274635315, "learning_rate": 4.2594993692983955e-05, "loss": 0.0102, "step": 11400 }, { "grad_norm": 0.13312192261219025, "learning_rate": 4.251324205435837e-05, "loss": 0.011, "step": 11410 }, { "grad_norm": 0.18005920946598053, "learning_rate": 4.243151088424312e-05, "loss": 0.0096, "step": 11420 }, { "grad_norm": 0.12228751927614212, "learning_rate": 4.234980040608813e-05, "loss": 0.0098, "step": 11430 }, { "grad_norm": 0.13147979974746704, "learning_rate": 4.22681108432867e-05, "loss": 0.0084, "step": 11440 }, { "grad_norm": 0.47965574264526367, "learning_rate": 4.2186442419174984e-05, "loss": 0.0104, "step": 11450 }, { "grad_norm": 0.24131140112876892, "learning_rate": 4.210479535703133e-05, "loss": 0.0109, "step": 11460 }, { "grad_norm": 0.23223792016506195, "learning_rate": 4.202316988007567e-05, "loss": 0.0092, "step": 11470 }, { "grad_norm": 0.16638493537902832, "learning_rate": 4.194156621146901e-05, "loss": 0.0092, "step": 11480 }, { "grad_norm": 0.16978468000888824, "learning_rate": 4.1859984574312596e-05, "loss": 0.0102, "step": 11490 }, { "grad_norm": 0.1903512179851532, "learning_rate": 4.177842519164752e-05, "loss": 0.0092, "step": 11500 }, { "grad_norm": 0.210654154419899, "learning_rate": 4.169688828645404e-05, "loss": 0.0098, "step": 11510 }, { "grad_norm": 0.16281193494796753, "learning_rate": 4.161537408165092e-05, "loss": 0.0106, "step": 11520 }, { "grad_norm": 0.1550668478012085, "learning_rate": 4.1533882800094924e-05, "loss": 0.0114, "step": 11530 }, { "grad_norm": 0.10089075565338135, "learning_rate": 4.145241466458005e-05, "loss": 0.0115, "step": 11540 }, { "grad_norm": 0.15693098306655884, "learning_rate": 4.13709698978371e-05, "loss": 0.0106, "step": 11550 }, { "grad_norm": 0.11359676718711853, "learning_rate": 4.1289548722532944e-05, "loss": 0.01, "step": 11560 }, { "grad_norm": 0.18243944644927979, "learning_rate": 4.120815136126999e-05, "loss": 0.0096, "step": 11570 }, { "grad_norm": 0.12796063721179962, "learning_rate": 4.112677803658548e-05, "loss": 0.0099, "step": 11580 }, { "grad_norm": 0.13827471435070038, "learning_rate": 4.1045428970951e-05, "loss": 0.01, "step": 11590 }, { "grad_norm": 0.1367706060409546, "learning_rate": 4.0964104386771785e-05, "loss": 0.0098, "step": 11600 }, { "grad_norm": 0.1550777405500412, "learning_rate": 4.0882804506386144e-05, "loss": 0.0133, "step": 11610 }, { "grad_norm": 0.20738302171230316, "learning_rate": 4.080152955206485e-05, "loss": 0.0128, "step": 11620 }, { "grad_norm": 0.24252954125404358, "learning_rate": 4.0720279746010505e-05, "loss": 0.0147, "step": 11630 }, { "grad_norm": 0.2178487628698349, "learning_rate": 4.063905531035699e-05, "loss": 0.0104, "step": 11640 }, { "grad_norm": 0.21983154118061066, "learning_rate": 4.055785646716882e-05, "loss": 0.0101, "step": 11650 }, { "grad_norm": 0.11781840771436691, "learning_rate": 4.047668343844051e-05, "loss": 0.0094, "step": 11660 }, { "grad_norm": 0.1457848995923996, "learning_rate": 4.039553644609604e-05, "loss": 0.0114, "step": 11670 }, { "grad_norm": 0.1829780638217926, "learning_rate": 4.0314415711988176e-05, "loss": 0.0086, "step": 11680 }, { "grad_norm": 0.149851992726326, "learning_rate": 4.023332145789792e-05, "loss": 0.0088, "step": 11690 }, { "grad_norm": 0.14567367732524872, "learning_rate": 4.015225390553385e-05, "loss": 0.0097, "step": 11700 }, { "grad_norm": 0.17327073216438293, "learning_rate": 4.007121327653158e-05, "loss": 0.0094, "step": 11710 }, { "grad_norm": 0.17579294741153717, "learning_rate": 3.9990199792453064e-05, "loss": 0.0098, "step": 11720 }, { "grad_norm": 0.1801060438156128, "learning_rate": 3.9909213674786103e-05, "loss": 0.0091, "step": 11730 }, { "grad_norm": 0.11638766527175903, "learning_rate": 3.982825514494363e-05, "loss": 0.0099, "step": 11740 }, { "grad_norm": 0.48739859461784363, "learning_rate": 3.974732442426319e-05, "loss": 0.0098, "step": 11750 }, { "grad_norm": 0.22315716743469238, "learning_rate": 3.966642173400629e-05, "loss": 0.01, "step": 11760 }, { "grad_norm": 0.10296326130628586, "learning_rate": 3.9585547295357764e-05, "loss": 0.0087, "step": 11770 }, { "grad_norm": 0.13268917798995972, "learning_rate": 3.950470132942526e-05, "loss": 0.0072, "step": 11780 }, { "grad_norm": 0.10527493059635162, "learning_rate": 3.942388405723856e-05, "loss": 0.0102, "step": 11790 }, { "grad_norm": 0.1391136795282364, "learning_rate": 3.9343095699749e-05, "loss": 0.0104, "step": 11800 }, { "grad_norm": 0.13424727320671082, "learning_rate": 3.9262336477828874e-05, "loss": 0.0096, "step": 11810 }, { "grad_norm": 0.15437060594558716, "learning_rate": 3.9181606612270794e-05, "loss": 0.0086, "step": 11820 }, { "grad_norm": 0.17718632519245148, "learning_rate": 3.910090632378713e-05, "loss": 0.0109, "step": 11830 }, { "grad_norm": 0.15048770606517792, "learning_rate": 3.90202358330094e-05, "loss": 0.0119, "step": 11840 }, { "grad_norm": 0.1139109656214714, "learning_rate": 3.8939595360487656e-05, "loss": 0.0101, "step": 11850 }, { "grad_norm": 0.10814749449491501, "learning_rate": 3.885898512668984e-05, "loss": 0.01, "step": 11860 }, { "grad_norm": 0.11100966483354568, "learning_rate": 3.877840535200127e-05, "loss": 0.0112, "step": 11870 }, { "grad_norm": 0.12319933623075485, "learning_rate": 3.869785625672397e-05, "loss": 0.0094, "step": 11880 }, { "grad_norm": 0.15187878906726837, "learning_rate": 3.8617338061076094e-05, "loss": 0.0084, "step": 11890 }, { "grad_norm": 0.18130457401275635, "learning_rate": 3.853685098519132e-05, "loss": 0.009, "step": 11900 }, { "grad_norm": 0.1765391230583191, "learning_rate": 3.845639524911823e-05, "loss": 0.0084, "step": 11910 }, { "grad_norm": 0.14383216202259064, "learning_rate": 3.837597107281974e-05, "loss": 0.009, "step": 11920 }, { "grad_norm": 0.1829141229391098, "learning_rate": 3.829557867617247e-05, "loss": 0.0111, "step": 11930 }, { "grad_norm": 0.2771480977535248, "learning_rate": 3.821521827896618e-05, "loss": 0.0117, "step": 11940 }, { "grad_norm": 0.1759459674358368, "learning_rate": 3.81348901009031e-05, "loss": 0.01, "step": 11950 }, { "grad_norm": 0.2309315800666809, "learning_rate": 3.805459436159741e-05, "loss": 0.0117, "step": 11960 }, { "grad_norm": 0.13144825398921967, "learning_rate": 3.797433128057461e-05, "loss": 0.0094, "step": 11970 }, { "grad_norm": 0.17214351892471313, "learning_rate": 3.789410107727089e-05, "loss": 0.0106, "step": 11980 }, { "grad_norm": 0.16214017570018768, "learning_rate": 3.781390397103257e-05, "loss": 0.0089, "step": 11990 }, { "grad_norm": 0.13959148526191711, "learning_rate": 3.7733740181115455e-05, "loss": 0.01, "step": 12000 }, { "grad_norm": 0.1522863358259201, "learning_rate": 3.7653609926684306e-05, "loss": 0.0107, "step": 12010 }, { "grad_norm": 0.24375051259994507, "learning_rate": 3.757351342681217e-05, "loss": 0.0086, "step": 12020 }, { "grad_norm": 0.25720569491386414, "learning_rate": 3.749345090047982e-05, "loss": 0.008, "step": 12030 }, { "grad_norm": 0.31324502825737, "learning_rate": 3.741342256657515e-05, "loss": 0.0092, "step": 12040 }, { "grad_norm": 0.12476992607116699, "learning_rate": 3.7333428643892567e-05, "loss": 0.0092, "step": 12050 }, { "grad_norm": 0.420350044965744, "learning_rate": 3.725346935113239e-05, "loss": 0.0088, "step": 12060 }, { "grad_norm": 0.16135434806346893, "learning_rate": 3.717354490690029e-05, "loss": 0.0079, "step": 12070 }, { "grad_norm": 0.1343766748905182, "learning_rate": 3.709365552970664e-05, "loss": 0.0085, "step": 12080 }, { "grad_norm": 0.18249738216400146, "learning_rate": 3.7013801437965945e-05, "loss": 0.0095, "step": 12090 }, { "grad_norm": 0.16720671951770782, "learning_rate": 3.693398284999623e-05, "loss": 0.0099, "step": 12100 }, { "grad_norm": 0.17868733406066895, "learning_rate": 3.6854199984018484e-05, "loss": 0.0092, "step": 12110 }, { "grad_norm": 0.1737927943468094, "learning_rate": 3.677445305815601e-05, "loss": 0.01, "step": 12120 }, { "grad_norm": 0.17534655332565308, "learning_rate": 3.669474229043387e-05, "loss": 0.009, "step": 12130 }, { "grad_norm": 0.1454739272594452, "learning_rate": 3.6615067898778235e-05, "loss": 0.0089, "step": 12140 }, { "grad_norm": 0.15352599322795868, "learning_rate": 3.6535430101015866e-05, "loss": 0.0071, "step": 12150 }, { "grad_norm": 0.14931641519069672, "learning_rate": 3.645582911487345e-05, "loss": 0.0084, "step": 12160 }, { "grad_norm": 0.1675536185503006, "learning_rate": 3.637626515797706e-05, "loss": 0.0087, "step": 12170 }, { "grad_norm": 0.21787935495376587, "learning_rate": 3.629673844785152e-05, "loss": 0.0103, "step": 12180 }, { "grad_norm": 0.12941792607307434, "learning_rate": 3.621724920191979e-05, "loss": 0.0088, "step": 12190 }, { "grad_norm": 0.18592920899391174, "learning_rate": 3.6137797637502444e-05, "loss": 0.0102, "step": 12200 }, { "grad_norm": 0.17619739472866058, "learning_rate": 3.6058383971817035e-05, "loss": 0.0105, "step": 12210 }, { "grad_norm": 0.12265519052743912, "learning_rate": 3.59790084219775e-05, "loss": 0.0077, "step": 12220 }, { "grad_norm": 0.11367000639438629, "learning_rate": 3.589967120499353e-05, "loss": 0.0096, "step": 12230 }, { "grad_norm": 0.14443044364452362, "learning_rate": 3.5820372537770075e-05, "loss": 0.0108, "step": 12240 }, { "grad_norm": 0.24261073768138885, "learning_rate": 3.5741112637106655e-05, "loss": 0.0087, "step": 12250 }, { "grad_norm": 0.27764225006103516, "learning_rate": 3.5661891719696804e-05, "loss": 0.0081, "step": 12260 }, { "grad_norm": 0.17809917032718658, "learning_rate": 3.5582710002127504e-05, "loss": 0.0101, "step": 12270 }, { "grad_norm": 0.148143008351326, "learning_rate": 3.550356770087853e-05, "loss": 0.0083, "step": 12280 }, { "grad_norm": 0.14444215595722198, "learning_rate": 3.5424465032321914e-05, "loss": 0.0082, "step": 12290 }, { "grad_norm": 0.09975260496139526, "learning_rate": 3.5345402212721335e-05, "loss": 0.0097, "step": 12300 }, { "grad_norm": 0.11628086119890213, "learning_rate": 3.526637945823152e-05, "loss": 0.0093, "step": 12310 }, { "grad_norm": 0.2141520231962204, "learning_rate": 3.518739698489767e-05, "loss": 0.0091, "step": 12320 }, { "grad_norm": 0.18678916990756989, "learning_rate": 3.510845500865485e-05, "loss": 0.008, "step": 12330 }, { "grad_norm": 0.1692144274711609, "learning_rate": 3.502955374532739e-05, "loss": 0.0081, "step": 12340 }, { "grad_norm": 0.13698938488960266, "learning_rate": 3.495069341062836e-05, "loss": 0.0108, "step": 12350 }, { "grad_norm": 0.08223753422498703, "learning_rate": 3.4871874220158896e-05, "loss": 0.0081, "step": 12360 }, { "grad_norm": 0.14901715517044067, "learning_rate": 3.479309638940762e-05, "loss": 0.0083, "step": 12370 }, { "grad_norm": 0.08682109415531158, "learning_rate": 3.4714360133750146e-05, "loss": 0.0077, "step": 12380 }, { "grad_norm": 0.11437799036502838, "learning_rate": 3.463566566844839e-05, "loss": 0.009, "step": 12390 }, { "grad_norm": 0.15462744235992432, "learning_rate": 3.4557013208650016e-05, "loss": 0.0091, "step": 12400 }, { "grad_norm": 0.1422770619392395, "learning_rate": 3.4478402969387857e-05, "loss": 0.007, "step": 12410 }, { "grad_norm": 0.13215522468090057, "learning_rate": 3.4399835165579266e-05, "loss": 0.0081, "step": 12420 }, { "grad_norm": 0.14801208674907684, "learning_rate": 3.4321310012025645e-05, "loss": 0.0103, "step": 12430 }, { "grad_norm": 0.11809186637401581, "learning_rate": 3.424282772341176e-05, "loss": 0.0081, "step": 12440 }, { "grad_norm": 0.11880609393119812, "learning_rate": 3.416438851430519e-05, "loss": 0.008, "step": 12450 }, { "grad_norm": 0.18348689377307892, "learning_rate": 3.408599259915577e-05, "loss": 0.0082, "step": 12460 }, { "grad_norm": 0.14303068816661835, "learning_rate": 3.400764019229487e-05, "loss": 0.0085, "step": 12470 }, { "grad_norm": 0.26667627692222595, "learning_rate": 3.3929331507935035e-05, "loss": 0.0084, "step": 12480 }, { "grad_norm": 0.1350974291563034, "learning_rate": 3.3851066760169196e-05, "loss": 0.0103, "step": 12490 }, { "grad_norm": 0.1454838663339615, "learning_rate": 3.377284616297021e-05, "loss": 0.0079, "step": 12500 }, { "grad_norm": 0.1497240513563156, "learning_rate": 3.3694669930190166e-05, "loss": 0.0085, "step": 12510 }, { "grad_norm": 0.15258042514324188, "learning_rate": 3.36165382755599e-05, "loss": 0.0079, "step": 12520 }, { "grad_norm": 0.20647813379764557, "learning_rate": 3.35384514126884e-05, "loss": 0.0099, "step": 12530 }, { "grad_norm": 0.16483457386493683, "learning_rate": 3.3460409555062154e-05, "loss": 0.0072, "step": 12540 }, { "grad_norm": 0.12539590895175934, "learning_rate": 3.3382412916044645e-05, "loss": 0.0094, "step": 12550 }, { "grad_norm": 0.11970549076795578, "learning_rate": 3.330446170887566e-05, "loss": 0.0099, "step": 12560 }, { "grad_norm": 0.24044790863990784, "learning_rate": 3.3226556146670834e-05, "loss": 0.0088, "step": 12570 }, { "grad_norm": 0.13254645466804504, "learning_rate": 3.314869644242102e-05, "loss": 0.011, "step": 12580 }, { "grad_norm": 0.14270584285259247, "learning_rate": 3.3070882808991674e-05, "loss": 0.0082, "step": 12590 }, { "grad_norm": 0.11893715709447861, "learning_rate": 3.2993115459122305e-05, "loss": 0.0075, "step": 12600 }, { "grad_norm": 0.12498091906309128, "learning_rate": 3.2915394605425835e-05, "loss": 0.0066, "step": 12610 }, { "grad_norm": 0.11811614036560059, "learning_rate": 3.283772046038816e-05, "loss": 0.007, "step": 12620 }, { "grad_norm": 0.20748299360275269, "learning_rate": 3.276009323636739e-05, "loss": 0.0077, "step": 12630 }, { "grad_norm": 0.17796072363853455, "learning_rate": 3.268251314559344e-05, "loss": 0.0086, "step": 12640 }, { "grad_norm": 0.11839158087968826, "learning_rate": 3.2604980400167254e-05, "loss": 0.007, "step": 12650 }, { "grad_norm": 0.15314362943172455, "learning_rate": 3.252749521206042e-05, "loss": 0.008, "step": 12660 }, { "grad_norm": 0.1412954181432724, "learning_rate": 3.2450057793114494e-05, "loss": 0.0079, "step": 12670 }, { "grad_norm": 0.18570758402347565, "learning_rate": 3.2372668355040435e-05, "loss": 0.0086, "step": 12680 }, { "grad_norm": 0.14956063032150269, "learning_rate": 3.2295327109418005e-05, "loss": 0.0111, "step": 12690 }, { "grad_norm": 0.26317936182022095, "learning_rate": 3.221803426769518e-05, "loss": 0.0082, "step": 12700 }, { "grad_norm": 0.26742488145828247, "learning_rate": 3.214079004118768e-05, "loss": 0.0083, "step": 12710 }, { "grad_norm": 0.24738709628582, "learning_rate": 3.2063594641078234e-05, "loss": 0.0094, "step": 12720 }, { "grad_norm": 0.14898015558719635, "learning_rate": 3.198644827841616e-05, "loss": 0.0093, "step": 12730 }, { "grad_norm": 0.10274504125118256, "learning_rate": 3.1909351164116654e-05, "loss": 0.007, "step": 12740 }, { "grad_norm": 0.1233067438006401, "learning_rate": 3.183230350896026e-05, "loss": 0.0083, "step": 12750 }, { "grad_norm": 0.0923084244132042, "learning_rate": 3.1755305523592337e-05, "loss": 0.0105, "step": 12760 }, { "grad_norm": 0.18828091025352478, "learning_rate": 3.167835741852245e-05, "loss": 0.0102, "step": 12770 }, { "grad_norm": 0.15861301124095917, "learning_rate": 3.160145940412378e-05, "loss": 0.0087, "step": 12780 }, { "grad_norm": 0.15530025959014893, "learning_rate": 3.1524611690632545e-05, "loss": 0.0092, "step": 12790 }, { "grad_norm": 0.1585802584886551, "learning_rate": 3.144781448814746e-05, "loss": 0.0078, "step": 12800 }, { "grad_norm": 0.15195177495479584, "learning_rate": 3.1371068006629145e-05, "loss": 0.0076, "step": 12810 }, { "grad_norm": 0.09938367456197739, "learning_rate": 3.129437245589956e-05, "loss": 0.0084, "step": 12820 }, { "grad_norm": 0.11793087422847748, "learning_rate": 3.121772804564143e-05, "loss": 0.0075, "step": 12830 }, { "grad_norm": 0.11879447102546692, "learning_rate": 3.11411349853976e-05, "loss": 0.0079, "step": 12840 }, { "grad_norm": 0.20238202810287476, "learning_rate": 3.10645934845706e-05, "loss": 0.0093, "step": 12850 }, { "grad_norm": 0.19098562002182007, "learning_rate": 3.098810375242196e-05, "loss": 0.0075, "step": 12860 }, { "grad_norm": 0.1366349160671234, "learning_rate": 3.0911665998071704e-05, "loss": 0.0084, "step": 12870 }, { "grad_norm": 0.2373533844947815, "learning_rate": 3.083528043049774e-05, "loss": 0.0077, "step": 12880 }, { "grad_norm": 0.23982878029346466, "learning_rate": 3.0758947258535255e-05, "loss": 0.0091, "step": 12890 }, { "grad_norm": 0.15512998402118683, "learning_rate": 3.068266669087625e-05, "loss": 0.008, "step": 12900 }, { "grad_norm": 0.19569528102874756, "learning_rate": 3.060643893606887e-05, "loss": 0.0077, "step": 12910 }, { "grad_norm": 0.09887193888425827, "learning_rate": 3.053026420251693e-05, "loss": 0.0081, "step": 12920 }, { "grad_norm": 0.15418772399425507, "learning_rate": 3.0454142698479183e-05, "loss": 0.0087, "step": 12930 }, { "grad_norm": 0.09930013120174408, "learning_rate": 3.0378074632068954e-05, "loss": 0.0078, "step": 12940 }, { "grad_norm": 0.1349790245294571, "learning_rate": 3.0302060211253408e-05, "loss": 0.0083, "step": 12950 }, { "grad_norm": 0.18979017436504364, "learning_rate": 3.0226099643853073e-05, "loss": 0.0078, "step": 12960 }, { "grad_norm": 0.14786294102668762, "learning_rate": 3.0150193137541283e-05, "loss": 0.0066, "step": 12970 }, { "grad_norm": 0.13491657376289368, "learning_rate": 3.0074340899843467e-05, "loss": 0.0076, "step": 12980 }, { "grad_norm": 0.13889476656913757, "learning_rate": 2.999854313813677e-05, "loss": 0.0084, "step": 12990 }, { "grad_norm": 0.167078897356987, "learning_rate": 2.9922800059649382e-05, "loss": 0.0092, "step": 13000 }, { "grad_norm": 0.15625059604644775, "learning_rate": 2.9847111871459976e-05, "loss": 0.0088, "step": 13010 }, { "grad_norm": 0.12252073734998703, "learning_rate": 2.977147878049721e-05, "loss": 0.008, "step": 13020 }, { "grad_norm": 0.19133198261260986, "learning_rate": 2.9695900993539006e-05, "loss": 0.0081, "step": 13030 }, { "grad_norm": 0.15134230256080627, "learning_rate": 2.9620378717212183e-05, "loss": 0.0085, "step": 13040 }, { "grad_norm": 0.16135051846504211, "learning_rate": 2.9544912157991745e-05, "loss": 0.0089, "step": 13050 }, { "grad_norm": 0.10704461485147476, "learning_rate": 2.9469501522200405e-05, "loss": 0.0074, "step": 13060 }, { "grad_norm": 0.11292983591556549, "learning_rate": 2.9394147016007946e-05, "loss": 0.0057, "step": 13070 }, { "grad_norm": 0.20408639311790466, "learning_rate": 2.9318848845430702e-05, "loss": 0.0087, "step": 13080 }, { "grad_norm": 0.20489825308322906, "learning_rate": 2.9243607216331013e-05, "loss": 0.0092, "step": 13090 }, { "grad_norm": 0.13939593732357025, "learning_rate": 2.916842233441661e-05, "loss": 0.0076, "step": 13100 }, { "grad_norm": 0.18661564588546753, "learning_rate": 2.90932944052401e-05, "loss": 0.0083, "step": 13110 }, { "grad_norm": 0.17595498263835907, "learning_rate": 2.9018223634198354e-05, "loss": 0.0071, "step": 13120 }, { "grad_norm": 0.12035000324249268, "learning_rate": 2.8943210226532025e-05, "loss": 0.0073, "step": 13130 }, { "grad_norm": 0.07949049025774002, "learning_rate": 2.8868254387324857e-05, "loss": 0.0072, "step": 13140 }, { "grad_norm": 0.1403672844171524, "learning_rate": 2.8793356321503306e-05, "loss": 0.0085, "step": 13150 }, { "grad_norm": 0.253409206867218, "learning_rate": 2.87185162338358e-05, "loss": 0.0079, "step": 13160 }, { "grad_norm": 0.1536179482936859, "learning_rate": 2.8643734328932253e-05, "loss": 0.0075, "step": 13170 }, { "grad_norm": 0.135816290974617, "learning_rate": 2.856901081124359e-05, "loss": 0.0091, "step": 13180 }, { "grad_norm": 0.20334555208683014, "learning_rate": 2.8494345885061002e-05, "loss": 0.0099, "step": 13190 }, { "grad_norm": 0.2641476094722748, "learning_rate": 2.8419739754515616e-05, "loss": 0.0072, "step": 13200 }, { "grad_norm": 0.144565612077713, "learning_rate": 2.8345192623577666e-05, "loss": 0.0085, "step": 13210 }, { "grad_norm": 0.1478653848171234, "learning_rate": 2.8270704696056193e-05, "loss": 0.0079, "step": 13220 }, { "grad_norm": 0.12270165234804153, "learning_rate": 2.8196276175598367e-05, "loss": 0.0085, "step": 13230 }, { "grad_norm": 0.13047835230827332, "learning_rate": 2.8121907265688884e-05, "loss": 0.0067, "step": 13240 }, { "grad_norm": 0.16233249008655548, "learning_rate": 2.804759816964957e-05, "loss": 0.0082, "step": 13250 }, { "grad_norm": 0.17996032536029816, "learning_rate": 2.797334909063857e-05, "loss": 0.0084, "step": 13260 }, { "grad_norm": 0.11844580620527267, "learning_rate": 2.7899160231650056e-05, "loss": 0.0077, "step": 13270 }, { "grad_norm": 0.08626241236925125, "learning_rate": 2.7825031795513585e-05, "loss": 0.0076, "step": 13280 }, { "grad_norm": 0.10358838737010956, "learning_rate": 2.775096398489341e-05, "loss": 0.0096, "step": 13290 }, { "grad_norm": 0.14505915343761444, "learning_rate": 2.7676957002288163e-05, "loss": 0.0074, "step": 13300 }, { "grad_norm": 0.09684775769710541, "learning_rate": 2.760301105003003e-05, "loss": 0.0065, "step": 13310 }, { "grad_norm": 0.2556889057159424, "learning_rate": 2.752912633028446e-05, "loss": 0.0089, "step": 13320 }, { "grad_norm": 0.12605854868888855, "learning_rate": 2.7455303045049474e-05, "loss": 0.0082, "step": 13330 }, { "grad_norm": 0.13645316660404205, "learning_rate": 2.7381541396155098e-05, "loss": 0.0082, "step": 13340 }, { "grad_norm": 0.17580971121788025, "learning_rate": 2.730784158526286e-05, "loss": 0.0067, "step": 13350 }, { "grad_norm": 0.1490408182144165, "learning_rate": 2.723420381386521e-05, "loss": 0.0092, "step": 13360 }, { "grad_norm": 0.0799587294459343, "learning_rate": 2.7160628283285018e-05, "loss": 0.0065, "step": 13370 }, { "grad_norm": 0.15931019186973572, "learning_rate": 2.7087115194675007e-05, "loss": 0.0073, "step": 13380 }, { "grad_norm": 0.564987301826477, "learning_rate": 2.701366474901712e-05, "loss": 0.0083, "step": 13390 }, { "grad_norm": 0.15791669487953186, "learning_rate": 2.6940277147122085e-05, "loss": 0.0086, "step": 13400 }, { "grad_norm": 0.1255493462085724, "learning_rate": 2.686695258962878e-05, "loss": 0.0077, "step": 13410 }, { "grad_norm": 0.2558315396308899, "learning_rate": 2.679369127700375e-05, "loss": 0.0082, "step": 13420 }, { "grad_norm": 0.2315502017736435, "learning_rate": 2.672049340954067e-05, "loss": 0.0093, "step": 13430 }, { "grad_norm": 0.1591860055923462, "learning_rate": 2.6647359187359676e-05, "loss": 0.0086, "step": 13440 }, { "grad_norm": 0.1351492553949356, "learning_rate": 2.6574288810406946e-05, "loss": 0.0082, "step": 13450 }, { "grad_norm": 0.10353130847215652, "learning_rate": 2.6501282478454083e-05, "loss": 0.0075, "step": 13460 }, { "grad_norm": 0.11632611602544785, "learning_rate": 2.6428340391097618e-05, "loss": 0.0078, "step": 13470 }, { "grad_norm": 0.12431365996599197, "learning_rate": 2.6355462747758485e-05, "loss": 0.008, "step": 13480 }, { "grad_norm": 0.16089442372322083, "learning_rate": 2.6282649747681304e-05, "loss": 0.0072, "step": 13490 }, { "grad_norm": 0.10761647671461105, "learning_rate": 2.620990158993406e-05, "loss": 0.008, "step": 13500 }, { "grad_norm": 0.14507047832012177, "learning_rate": 2.6137218473407477e-05, "loss": 0.0076, "step": 13510 }, { "grad_norm": 0.18290990591049194, "learning_rate": 2.606460059681436e-05, "loss": 0.0076, "step": 13520 }, { "grad_norm": 0.14335665106773376, "learning_rate": 2.599204815868928e-05, "loss": 0.0076, "step": 13530 }, { "grad_norm": 0.13241387903690338, "learning_rate": 2.5919561357387756e-05, "loss": 0.0075, "step": 13540 }, { "grad_norm": 0.07921767234802246, "learning_rate": 2.5847140391085972e-05, "loss": 0.0087, "step": 13550 }, { "grad_norm": 0.14680445194244385, "learning_rate": 2.5774785457780103e-05, "loss": 0.0075, "step": 13560 }, { "grad_norm": 0.10443998128175735, "learning_rate": 2.5702496755285753e-05, "loss": 0.0065, "step": 13570 }, { "grad_norm": 0.12179769575595856, "learning_rate": 2.5630274481237483e-05, "loss": 0.0066, "step": 13580 }, { "grad_norm": 0.13228487968444824, "learning_rate": 2.5558118833088197e-05, "loss": 0.0073, "step": 13590 }, { "grad_norm": 0.1697179228067398, "learning_rate": 2.548603000810872e-05, "loss": 0.0065, "step": 13600 }, { "grad_norm": 0.17730006575584412, "learning_rate": 2.5414008203387152e-05, "loss": 0.0061, "step": 13610 }, { "grad_norm": 0.13608282804489136, "learning_rate": 2.534205361582834e-05, "loss": 0.006, "step": 13620 }, { "grad_norm": 0.12047142535448074, "learning_rate": 2.527016644215338e-05, "loss": 0.0062, "step": 13630 }, { "grad_norm": 0.1495821624994278, "learning_rate": 2.519834687889905e-05, "loss": 0.0061, "step": 13640 }, { "grad_norm": 0.10948540270328522, "learning_rate": 2.5126595122417295e-05, "loss": 0.0066, "step": 13650 }, { "grad_norm": 0.13396184146404266, "learning_rate": 2.5054911368874713e-05, "loss": 0.0067, "step": 13660 }, { "grad_norm": 0.16184811294078827, "learning_rate": 2.4983295814251916e-05, "loss": 0.0078, "step": 13670 }, { "grad_norm": 0.14357778429985046, "learning_rate": 2.4911748654343105e-05, "loss": 0.0089, "step": 13680 }, { "grad_norm": 0.12215641140937805, "learning_rate": 2.4840270084755463e-05, "loss": 0.0075, "step": 13690 }, { "grad_norm": 0.1100117489695549, "learning_rate": 2.4768860300908685e-05, "loss": 0.0075, "step": 13700 }, { "grad_norm": 0.13753214478492737, "learning_rate": 2.469751949803443e-05, "loss": 0.008, "step": 13710 }, { "grad_norm": 0.08801797777414322, "learning_rate": 2.4626247871175666e-05, "loss": 0.0068, "step": 13720 }, { "grad_norm": 0.1276213675737381, "learning_rate": 2.4555045615186346e-05, "loss": 0.0084, "step": 13730 }, { "grad_norm": 0.11146289855241776, "learning_rate": 2.4483912924730677e-05, "loss": 0.0068, "step": 13740 }, { "grad_norm": 0.29949143528938293, "learning_rate": 2.4412849994282742e-05, "loss": 0.0089, "step": 13750 }, { "grad_norm": 0.23834139108657837, "learning_rate": 2.434185701812592e-05, "loss": 0.0076, "step": 13760 }, { "grad_norm": 0.12105236947536469, "learning_rate": 2.4270934190352218e-05, "loss": 0.0061, "step": 13770 }, { "grad_norm": 0.22550742328166962, "learning_rate": 2.4200081704861998e-05, "loss": 0.007, "step": 13780 }, { "grad_norm": 0.11833983659744263, "learning_rate": 2.412929975536321e-05, "loss": 0.0068, "step": 13790 }, { "grad_norm": 0.4742639660835266, "learning_rate": 2.4058588535371017e-05, "loss": 0.007, "step": 13800 }, { "grad_norm": 0.13116216659545898, "learning_rate": 2.3987948238207243e-05, "loss": 0.0074, "step": 13810 }, { "grad_norm": 0.12723734974861145, "learning_rate": 2.3917379056999678e-05, "loss": 0.0083, "step": 13820 }, { "grad_norm": 0.2062266319990158, "learning_rate": 2.3846881184681824e-05, "loss": 0.0082, "step": 13830 }, { "grad_norm": 0.18303969502449036, "learning_rate": 2.377645481399214e-05, "loss": 0.0073, "step": 13840 }, { "grad_norm": 0.15242987871170044, "learning_rate": 2.3706100137473667e-05, "loss": 0.0074, "step": 13850 }, { "grad_norm": 0.12032338231801987, "learning_rate": 2.3635817347473394e-05, "loss": 0.006, "step": 13860 }, { "grad_norm": 0.11630221456289291, "learning_rate": 2.3565606636141757e-05, "loss": 0.0085, "step": 13870 }, { "grad_norm": 0.10775570571422577, "learning_rate": 2.3495468195432203e-05, "loss": 0.0065, "step": 13880 }, { "grad_norm": 0.15366296470165253, "learning_rate": 2.3425402217100507e-05, "loss": 0.0069, "step": 13890 }, { "grad_norm": 0.13231560587882996, "learning_rate": 2.3355408892704424e-05, "loss": 0.0072, "step": 13900 }, { "grad_norm": 0.09862794727087021, "learning_rate": 2.3285488413603003e-05, "loss": 0.0071, "step": 13910 }, { "grad_norm": 0.14172619581222534, "learning_rate": 2.321564097095615e-05, "loss": 0.0075, "step": 13920 }, { "grad_norm": 0.11843056976795197, "learning_rate": 2.3145866755724142e-05, "loss": 0.0064, "step": 13930 }, { "grad_norm": 0.0757211372256279, "learning_rate": 2.307616595866699e-05, "loss": 0.0057, "step": 13940 }, { "grad_norm": 0.10985478013753891, "learning_rate": 2.3006538770344032e-05, "loss": 0.0068, "step": 13950 }, { "grad_norm": 0.14996905624866486, "learning_rate": 2.293698538111334e-05, "loss": 0.0072, "step": 13960 }, { "grad_norm": 0.19500389695167542, "learning_rate": 2.28675059811312e-05, "loss": 0.0077, "step": 13970 }, { "grad_norm": 0.12211684137582779, "learning_rate": 2.279810076035167e-05, "loss": 0.0075, "step": 13980 }, { "grad_norm": 0.1440882533788681, "learning_rate": 2.272876990852596e-05, "loss": 0.008, "step": 13990 }, { "grad_norm": 0.10326642543077469, "learning_rate": 2.265951361520195e-05, "loss": 0.0052, "step": 14000 }, { "grad_norm": 0.17940270900726318, "learning_rate": 2.2590332069723748e-05, "loss": 0.0076, "step": 14010 }, { "grad_norm": 0.1167125329375267, "learning_rate": 2.2521225461231004e-05, "loss": 0.0071, "step": 14020 }, { "grad_norm": 0.12672792375087738, "learning_rate": 2.2452193978658597e-05, "loss": 0.0052, "step": 14030 }, { "grad_norm": 0.11368203908205032, "learning_rate": 2.238323781073594e-05, "loss": 0.0059, "step": 14040 }, { "grad_norm": 0.1316523402929306, "learning_rate": 2.2314357145986552e-05, "loss": 0.0066, "step": 14050 }, { "grad_norm": 0.1306915581226349, "learning_rate": 2.224555217272757e-05, "loss": 0.0072, "step": 14060 }, { "grad_norm": 0.1668822318315506, "learning_rate": 2.2176823079069127e-05, "loss": 0.0052, "step": 14070 }, { "grad_norm": 0.1152377501130104, "learning_rate": 2.210817005291398e-05, "loss": 0.0061, "step": 14080 }, { "grad_norm": 0.15531480312347412, "learning_rate": 2.203959328195686e-05, "loss": 0.0071, "step": 14090 }, { "grad_norm": 0.09332505613565445, "learning_rate": 2.1971092953684026e-05, "loss": 0.0071, "step": 14100 }, { "grad_norm": 0.1171257421374321, "learning_rate": 2.1902669255372788e-05, "loss": 0.007, "step": 14110 }, { "grad_norm": 0.09076493978500366, "learning_rate": 2.1834322374090897e-05, "loss": 0.0063, "step": 14120 }, { "grad_norm": 0.20567858219146729, "learning_rate": 2.1766052496696153e-05, "loss": 0.0077, "step": 14130 }, { "grad_norm": 0.16322502493858337, "learning_rate": 2.169785980983577e-05, "loss": 0.0068, "step": 14140 }, { "grad_norm": 0.12871262431144714, "learning_rate": 2.162974449994593e-05, "loss": 0.0053, "step": 14150 }, { "grad_norm": 0.11759986728429794, "learning_rate": 2.1561706753251337e-05, "loss": 0.0072, "step": 14160 }, { "grad_norm": 0.16359606385231018, "learning_rate": 2.1493746755764544e-05, "loss": 0.0072, "step": 14170 }, { "grad_norm": 0.1419595181941986, "learning_rate": 2.1425864693285635e-05, "loss": 0.0075, "step": 14180 }, { "grad_norm": 0.12449288368225098, "learning_rate": 2.1358060751401547e-05, "loss": 0.0074, "step": 14190 }, { "grad_norm": 0.09465201199054718, "learning_rate": 2.129033511548566e-05, "loss": 0.0061, "step": 14200 }, { "grad_norm": 0.13305670022964478, "learning_rate": 2.1222687970697315e-05, "loss": 0.0069, "step": 14210 }, { "grad_norm": 0.09377169609069824, "learning_rate": 2.1155119501981173e-05, "loss": 0.0071, "step": 14220 }, { "grad_norm": 0.172748863697052, "learning_rate": 2.1087629894066895e-05, "loss": 0.0071, "step": 14230 }, { "grad_norm": 0.1658395230770111, "learning_rate": 2.1020219331468473e-05, "loss": 0.0062, "step": 14240 }, { "grad_norm": 0.10715547949075699, "learning_rate": 2.095288799848379e-05, "loss": 0.0072, "step": 14250 }, { "grad_norm": 0.10814423859119415, "learning_rate": 2.088563607919417e-05, "loss": 0.007, "step": 14260 }, { "grad_norm": 0.14961855113506317, "learning_rate": 2.0818463757463786e-05, "loss": 0.0078, "step": 14270 }, { "grad_norm": 0.16352008283138275, "learning_rate": 2.0751371216939175e-05, "loss": 0.0065, "step": 14280 }, { "grad_norm": 0.19634564220905304, "learning_rate": 2.068435864104882e-05, "loss": 0.0061, "step": 14290 }, { "grad_norm": 0.16410329937934875, "learning_rate": 2.0617426213002506e-05, "loss": 0.0074, "step": 14300 }, { "grad_norm": 0.08360293507575989, "learning_rate": 2.055057411579097e-05, "loss": 0.0081, "step": 14310 }, { "grad_norm": 0.11911585927009583, "learning_rate": 2.0483802532185286e-05, "loss": 0.0076, "step": 14320 }, { "grad_norm": 0.32665789127349854, "learning_rate": 2.041711164473638e-05, "loss": 0.006, "step": 14330 }, { "grad_norm": 0.12266720086336136, "learning_rate": 2.0350501635774637e-05, "loss": 0.0064, "step": 14340 }, { "grad_norm": 0.08855657279491425, "learning_rate": 2.0283972687409247e-05, "loss": 0.0069, "step": 14350 }, { "grad_norm": 0.13959892094135284, "learning_rate": 2.021752498152784e-05, "loss": 0.0064, "step": 14360 }, { "grad_norm": 0.14273284375667572, "learning_rate": 2.015115869979589e-05, "loss": 0.0054, "step": 14370 }, { "grad_norm": 0.1526416838169098, "learning_rate": 2.0084874023656265e-05, "loss": 0.007, "step": 14380 }, { "grad_norm": 0.22072456777095795, "learning_rate": 2.001867113432877e-05, "loss": 0.0082, "step": 14390 }, { "grad_norm": 0.11803358793258667, "learning_rate": 1.995255021280954e-05, "loss": 0.0063, "step": 14400 }, { "grad_norm": 0.16811753809452057, "learning_rate": 1.9886511439870688e-05, "loss": 0.0061, "step": 14410 }, { "grad_norm": 0.1087285503745079, "learning_rate": 1.9820554996059675e-05, "loss": 0.0053, "step": 14420 }, { "grad_norm": 0.09446712583303452, "learning_rate": 1.9754681061698893e-05, "loss": 0.0058, "step": 14430 }, { "grad_norm": 0.14372065663337708, "learning_rate": 1.9688889816885185e-05, "loss": 0.0071, "step": 14440 }, { "grad_norm": 0.11408334970474243, "learning_rate": 1.962318144148928e-05, "loss": 0.0054, "step": 14450 }, { "grad_norm": 0.1018875241279602, "learning_rate": 1.955755611515539e-05, "loss": 0.0061, "step": 14460 }, { "grad_norm": 0.13650502264499664, "learning_rate": 1.9492014017300642e-05, "loss": 0.0073, "step": 14470 }, { "grad_norm": 0.14343740046024323, "learning_rate": 1.942655532711461e-05, "loss": 0.0056, "step": 14480 }, { "grad_norm": 0.27338704466819763, "learning_rate": 1.9361180223558882e-05, "loss": 0.0055, "step": 14490 }, { "grad_norm": 0.19755274057388306, "learning_rate": 1.929588888536647e-05, "loss": 0.0059, "step": 14500 }, { "grad_norm": 0.13208508491516113, "learning_rate": 1.9230681491041425e-05, "loss": 0.0073, "step": 14510 }, { "grad_norm": 0.1566578894853592, "learning_rate": 1.9165558218858264e-05, "loss": 0.0066, "step": 14520 }, { "grad_norm": 0.17309707403182983, "learning_rate": 1.9100519246861505e-05, "loss": 0.006, "step": 14530 }, { "grad_norm": 0.10424192994832993, "learning_rate": 1.9035564752865248e-05, "loss": 0.005, "step": 14540 }, { "grad_norm": 0.09956467151641846, "learning_rate": 1.897069491445258e-05, "loss": 0.0077, "step": 14550 }, { "grad_norm": 0.12068826705217361, "learning_rate": 1.890590990897515e-05, "loss": 0.0064, "step": 14560 }, { "grad_norm": 0.11269792914390564, "learning_rate": 1.884120991355272e-05, "loss": 0.0061, "step": 14570 }, { "grad_norm": 0.15886837244033813, "learning_rate": 1.8776595105072576e-05, "loss": 0.0068, "step": 14580 }, { "grad_norm": 0.11482223123311996, "learning_rate": 1.8712065660189166e-05, "loss": 0.0066, "step": 14590 }, { "grad_norm": 0.10462388396263123, "learning_rate": 1.8647621755323513e-05, "loss": 0.0063, "step": 14600 }, { "grad_norm": 0.10871457308530807, "learning_rate": 1.858326356666278e-05, "loss": 0.0067, "step": 14610 }, { "grad_norm": 0.13479308784008026, "learning_rate": 1.851899127015983e-05, "loss": 0.0062, "step": 14620 }, { "grad_norm": 0.1286613941192627, "learning_rate": 1.8454805041532626e-05, "loss": 0.0064, "step": 14630 }, { "grad_norm": 0.13269053399562836, "learning_rate": 1.8390705056263906e-05, "loss": 0.0063, "step": 14640 }, { "grad_norm": 0.10598579049110413, "learning_rate": 1.832669148960057e-05, "loss": 0.005, "step": 14650 }, { "grad_norm": 0.08851759880781174, "learning_rate": 1.8262764516553233e-05, "loss": 0.0065, "step": 14660 }, { "grad_norm": 0.11064887046813965, "learning_rate": 1.8198924311895843e-05, "loss": 0.0064, "step": 14670 }, { "grad_norm": 0.10063286870718002, "learning_rate": 1.813517105016505e-05, "loss": 0.0064, "step": 14680 }, { "grad_norm": 0.15070444345474243, "learning_rate": 1.8071504905659888e-05, "loss": 0.0054, "step": 14690 }, { "grad_norm": 0.13329288363456726, "learning_rate": 1.800792605244109e-05, "loss": 0.0061, "step": 14700 }, { "grad_norm": 0.11602115631103516, "learning_rate": 1.7944434664330844e-05, "loss": 0.0055, "step": 14710 }, { "grad_norm": 0.08986308425664902, "learning_rate": 1.7881030914912212e-05, "loss": 0.0066, "step": 14720 }, { "grad_norm": 0.21984991431236267, "learning_rate": 1.7817714977528577e-05, "loss": 0.0064, "step": 14730 }, { "grad_norm": 0.16804246604442596, "learning_rate": 1.7754487025283332e-05, "loss": 0.0052, "step": 14740 }, { "grad_norm": 0.16058765351772308, "learning_rate": 1.7691347231039275e-05, "loss": 0.0056, "step": 14750 }, { "grad_norm": 0.10759300738573074, "learning_rate": 1.7628295767418164e-05, "loss": 0.0064, "step": 14760 }, { "grad_norm": 0.1578758805990219, "learning_rate": 1.7565332806800333e-05, "loss": 0.0059, "step": 14770 }, { "grad_norm": 0.19713455438613892, "learning_rate": 1.750245852132408e-05, "loss": 0.0064, "step": 14780 }, { "grad_norm": 0.15213815867900848, "learning_rate": 1.7439673082885323e-05, "loss": 0.0056, "step": 14790 }, { "grad_norm": 0.23376938700675964, "learning_rate": 1.7376976663137047e-05, "loss": 0.0066, "step": 14800 }, { "grad_norm": 0.4487079679965973, "learning_rate": 1.7314369433488853e-05, "loss": 0.0061, "step": 14810 }, { "grad_norm": 0.13894130289554596, "learning_rate": 1.7251851565106548e-05, "loss": 0.0063, "step": 14820 }, { "grad_norm": 0.12897264957427979, "learning_rate": 1.7189423228911574e-05, "loss": 0.006, "step": 14830 }, { "grad_norm": 0.1501796841621399, "learning_rate": 1.7127084595580606e-05, "loss": 0.0059, "step": 14840 }, { "grad_norm": 0.091891810297966, "learning_rate": 1.706483583554513e-05, "loss": 0.0054, "step": 14850 }, { "grad_norm": 0.2075316458940506, "learning_rate": 1.700267711899083e-05, "loss": 0.0067, "step": 14860 }, { "grad_norm": 0.15214751660823822, "learning_rate": 1.69406086158573e-05, "loss": 0.0072, "step": 14870 }, { "grad_norm": 0.13683956861495972, "learning_rate": 1.6878630495837455e-05, "loss": 0.0059, "step": 14880 }, { "grad_norm": 0.17635053396224976, "learning_rate": 1.681674292837707e-05, "loss": 0.0061, "step": 14890 }, { "grad_norm": 0.2253178209066391, "learning_rate": 1.6754946082674444e-05, "loss": 0.0068, "step": 14900 }, { "grad_norm": 0.29448312520980835, "learning_rate": 1.6693240127679748e-05, "loss": 0.0077, "step": 14910 }, { "grad_norm": 0.11132287234067917, "learning_rate": 1.663162523209475e-05, "loss": 0.0065, "step": 14920 }, { "grad_norm": 0.2733027935028076, "learning_rate": 1.6570101564372193e-05, "loss": 0.0067, "step": 14930 }, { "grad_norm": 0.14367184042930603, "learning_rate": 1.650866929271543e-05, "loss": 0.0067, "step": 14940 }, { "grad_norm": 0.13276757299900055, "learning_rate": 1.644732858507797e-05, "loss": 0.0066, "step": 14950 }, { "grad_norm": 0.14067184925079346, "learning_rate": 1.6386079609162943e-05, "loss": 0.0063, "step": 14960 }, { "grad_norm": 0.11195389181375504, "learning_rate": 1.6324922532422742e-05, "loss": 0.0062, "step": 14970 }, { "grad_norm": 0.12702465057373047, "learning_rate": 1.6263857522058434e-05, "loss": 0.0068, "step": 14980 }, { "grad_norm": 0.1812618523836136, "learning_rate": 1.6202884745019443e-05, "loss": 0.0053, "step": 14990 }, { "grad_norm": 0.11995403468608856, "learning_rate": 1.614200436800304e-05, "loss": 0.0063, "step": 15000 }, { "grad_norm": 0.18218998610973358, "learning_rate": 1.6081216557453814e-05, "loss": 0.0074, "step": 15010 }, { "grad_norm": 0.11837029457092285, "learning_rate": 1.6020521479563367e-05, "loss": 0.0065, "step": 15020 }, { "grad_norm": 0.1076301708817482, "learning_rate": 1.5959919300269654e-05, "loss": 0.0066, "step": 15030 }, { "grad_norm": 0.13307075202465057, "learning_rate": 1.5899410185256764e-05, "loss": 0.0053, "step": 15040 }, { "grad_norm": 0.09704990684986115, "learning_rate": 1.583899429995431e-05, "loss": 0.006, "step": 15050 }, { "grad_norm": 0.14444661140441895, "learning_rate": 1.5778671809536993e-05, "loss": 0.0056, "step": 15060 }, { "grad_norm": 0.10517879575490952, "learning_rate": 1.5718442878924246e-05, "loss": 0.0069, "step": 15070 }, { "grad_norm": 0.19705863296985626, "learning_rate": 1.5658307672779593e-05, "loss": 0.0076, "step": 15080 }, { "grad_norm": 0.1281328648328781, "learning_rate": 1.5598266355510427e-05, "loss": 0.0082, "step": 15090 }, { "grad_norm": 0.15537413954734802, "learning_rate": 1.553831909126744e-05, "loss": 0.007, "step": 15100 }, { "grad_norm": 0.10809335112571716, "learning_rate": 1.5478466043944135e-05, "loss": 0.0059, "step": 15110 }, { "grad_norm": 0.1599857062101364, "learning_rate": 1.5418707377176468e-05, "loss": 0.0066, "step": 15120 }, { "grad_norm": 0.071555495262146, "learning_rate": 1.535904325434233e-05, "loss": 0.0048, "step": 15130 }, { "grad_norm": 0.16743166744709015, "learning_rate": 1.529947383856118e-05, "loss": 0.0064, "step": 15140 }, { "grad_norm": 0.149888277053833, "learning_rate": 1.5239999292693524e-05, "loss": 0.0054, "step": 15150 }, { "grad_norm": 0.1532762199640274, "learning_rate": 1.5180619779340505e-05, "loss": 0.0058, "step": 15160 }, { "grad_norm": 0.12081128358840942, "learning_rate": 1.5121335460843428e-05, "loss": 0.0056, "step": 15170 }, { "grad_norm": 0.14688536524772644, "learning_rate": 1.5062146499283347e-05, "loss": 0.0055, "step": 15180 }, { "grad_norm": 0.09615129232406616, "learning_rate": 1.5003053056480643e-05, "loss": 0.0069, "step": 15190 }, { "grad_norm": 0.10955497622489929, "learning_rate": 1.4944055293994551e-05, "loss": 0.0065, "step": 15200 }, { "grad_norm": 0.1229109987616539, "learning_rate": 1.4885153373122656e-05, "loss": 0.006, "step": 15210 }, { "grad_norm": 0.13104379177093506, "learning_rate": 1.482634745490059e-05, "loss": 0.0066, "step": 15220 }, { "grad_norm": 0.1774502694606781, "learning_rate": 1.4767637700101466e-05, "loss": 0.0062, "step": 15230 }, { "grad_norm": 0.16169649362564087, "learning_rate": 1.4709024269235528e-05, "loss": 0.0058, "step": 15240 }, { "grad_norm": 0.16058167815208435, "learning_rate": 1.4650507322549684e-05, "loss": 0.0059, "step": 15250 }, { "grad_norm": 0.1655225306749344, "learning_rate": 1.4592087020026972e-05, "loss": 0.0055, "step": 15260 }, { "grad_norm": 0.13236191868782043, "learning_rate": 1.4533763521386318e-05, "loss": 0.0078, "step": 15270 }, { "grad_norm": 0.15737107396125793, "learning_rate": 1.44755369860819e-05, "loss": 0.0065, "step": 15280 }, { "grad_norm": 0.24444864690303802, "learning_rate": 1.441740757330287e-05, "loss": 0.0071, "step": 15290 }, { "grad_norm": 0.11665450036525726, "learning_rate": 1.4359375441972844e-05, "loss": 0.0063, "step": 15300 }, { "grad_norm": 0.10856233537197113, "learning_rate": 1.4301440750749395e-05, "loss": 0.0058, "step": 15310 }, { "grad_norm": 0.1393858939409256, "learning_rate": 1.4243603658023808e-05, "loss": 0.0039, "step": 15320 }, { "grad_norm": 0.09225007891654968, "learning_rate": 1.4185864321920444e-05, "loss": 0.0054, "step": 15330 }, { "grad_norm": 0.13283637166023254, "learning_rate": 1.4128222900296485e-05, "loss": 0.0047, "step": 15340 }, { "grad_norm": 0.29579898715019226, "learning_rate": 1.407067955074135e-05, "loss": 0.0054, "step": 15350 }, { "grad_norm": 0.11594395339488983, "learning_rate": 1.4013234430576356e-05, "loss": 0.0063, "step": 15360 }, { "grad_norm": 0.3565695285797119, "learning_rate": 1.3955887696854286e-05, "loss": 0.0072, "step": 15370 }, { "grad_norm": 0.09114328771829605, "learning_rate": 1.38986395063589e-05, "loss": 0.0057, "step": 15380 }, { "grad_norm": 0.14908339083194733, "learning_rate": 1.3841490015604597e-05, "loss": 0.0061, "step": 15390 }, { "grad_norm": 0.09309130162000656, "learning_rate": 1.3784439380835879e-05, "loss": 0.0066, "step": 15400 }, { "grad_norm": 0.12462113052606583, "learning_rate": 1.3727487758026986e-05, "loss": 0.0045, "step": 15410 }, { "grad_norm": 0.15490370988845825, "learning_rate": 1.3670635302881525e-05, "loss": 0.008, "step": 15420 }, { "grad_norm": 0.09742511808872223, "learning_rate": 1.3613882170831888e-05, "loss": 0.0051, "step": 15430 }, { "grad_norm": 0.1563878357410431, "learning_rate": 1.355722851703901e-05, "loss": 0.0057, "step": 15440 }, { "grad_norm": 0.1549827605485916, "learning_rate": 1.3500674496391814e-05, "loss": 0.0046, "step": 15450 }, { "grad_norm": 0.06776552647352219, "learning_rate": 1.3444220263506795e-05, "loss": 0.0052, "step": 15460 }, { "grad_norm": 0.07800491154193878, "learning_rate": 1.3387865972727714e-05, "loss": 0.004, "step": 15470 }, { "grad_norm": 0.12390141934156418, "learning_rate": 1.3331611778125036e-05, "loss": 0.0065, "step": 15480 }, { "grad_norm": 0.14546164870262146, "learning_rate": 1.3275457833495564e-05, "loss": 0.0069, "step": 15490 }, { "grad_norm": 0.09419216960668564, "learning_rate": 1.3219404292362065e-05, "loss": 0.0062, "step": 15500 }, { "grad_norm": 0.14071811735630035, "learning_rate": 1.3163451307972751e-05, "loss": 0.0058, "step": 15510 }, { "grad_norm": 0.08080799877643585, "learning_rate": 1.3107599033300977e-05, "loss": 0.0065, "step": 15520 }, { "grad_norm": 0.15930302441120148, "learning_rate": 1.305184762104471e-05, "loss": 0.0052, "step": 15530 }, { "grad_norm": 0.11846110224723816, "learning_rate": 1.2996197223626178e-05, "loss": 0.0059, "step": 15540 }, { "grad_norm": 0.12355508655309677, "learning_rate": 1.2940647993191457e-05, "loss": 0.0055, "step": 15550 }, { "grad_norm": 0.07697183638811111, "learning_rate": 1.2885200081610005e-05, "loss": 0.0069, "step": 15560 }, { "grad_norm": 0.11336750537157059, "learning_rate": 1.2829853640474316e-05, "loss": 0.0058, "step": 15570 }, { "grad_norm": 0.1930548995733261, "learning_rate": 1.2774608821099438e-05, "loss": 0.0059, "step": 15580 }, { "grad_norm": 0.09620604664087296, "learning_rate": 1.2719465774522577e-05, "loss": 0.0054, "step": 15590 }, { "grad_norm": 0.0698646828532219, "learning_rate": 1.2664424651502755e-05, "loss": 0.0045, "step": 15600 }, { "grad_norm": 0.10103446245193481, "learning_rate": 1.260948560252026e-05, "loss": 0.0054, "step": 15610 }, { "grad_norm": 0.17563627660274506, "learning_rate": 1.2554648777776396e-05, "loss": 0.0053, "step": 15620 }, { "grad_norm": 0.09682541340589523, "learning_rate": 1.2499914327192919e-05, "loss": 0.0045, "step": 15630 }, { "grad_norm": 0.09335658699274063, "learning_rate": 1.2445282400411722e-05, "loss": 0.0055, "step": 15640 }, { "grad_norm": 0.11927438527345657, "learning_rate": 1.2390753146794437e-05, "loss": 0.005, "step": 15650 }, { "grad_norm": 0.19313247501850128, "learning_rate": 1.2336326715421925e-05, "loss": 0.0059, "step": 15660 }, { "grad_norm": 0.07288102060556412, "learning_rate": 1.2282003255094005e-05, "loss": 0.0041, "step": 15670 }, { "grad_norm": 0.10079466551542282, "learning_rate": 1.2227782914328928e-05, "loss": 0.0051, "step": 15680 }, { "grad_norm": 0.1444721519947052, "learning_rate": 1.2173665841363018e-05, "loss": 0.0057, "step": 15690 }, { "grad_norm": 0.1291830986738205, "learning_rate": 1.211965218415032e-05, "loss": 0.0054, "step": 15700 }, { "grad_norm": 0.10040965676307678, "learning_rate": 1.2065742090362082e-05, "loss": 0.0073, "step": 15710 }, { "grad_norm": 0.09511888027191162, "learning_rate": 1.2011935707386457e-05, "loss": 0.0049, "step": 15720 }, { "grad_norm": 0.09394008666276932, "learning_rate": 1.1958233182328044e-05, "loss": 0.0068, "step": 15730 }, { "grad_norm": 0.110257588326931, "learning_rate": 1.1904634662007474e-05, "loss": 0.0051, "step": 15740 }, { "grad_norm": 0.10997658967971802, "learning_rate": 1.1851140292961088e-05, "loss": 0.007, "step": 15750 }, { "grad_norm": 0.13037024438381195, "learning_rate": 1.1797750221440424e-05, "loss": 0.0047, "step": 15760 }, { "grad_norm": 0.1182718276977539, "learning_rate": 1.1744464593411897e-05, "loss": 0.0061, "step": 15770 }, { "grad_norm": 0.10341058671474457, "learning_rate": 1.1691283554556399e-05, "loss": 0.005, "step": 15780 }, { "grad_norm": 0.12444820255041122, "learning_rate": 1.1638207250268834e-05, "loss": 0.0087, "step": 15790 }, { "grad_norm": 0.13136543333530426, "learning_rate": 1.158523582565782e-05, "loss": 0.006, "step": 15800 }, { "grad_norm": 0.08554934710264206, "learning_rate": 1.1532369425545192e-05, "loss": 0.0054, "step": 15810 }, { "grad_norm": 0.1257864385843277, "learning_rate": 1.1479608194465662e-05, "loss": 0.0054, "step": 15820 }, { "grad_norm": 0.14481514692306519, "learning_rate": 1.1426952276666442e-05, "loss": 0.006, "step": 15830 }, { "grad_norm": 0.15901052951812744, "learning_rate": 1.1374401816106778e-05, "loss": 0.0057, "step": 15840 }, { "grad_norm": 0.08751460164785385, "learning_rate": 1.1321956956457646e-05, "loss": 0.0058, "step": 15850 }, { "grad_norm": 0.10504160821437836, "learning_rate": 1.1269617841101277e-05, "loss": 0.0066, "step": 15860 }, { "grad_norm": 0.08571986109018326, "learning_rate": 1.1217384613130804e-05, "loss": 0.0046, "step": 15870 }, { "grad_norm": 0.11240265518426895, "learning_rate": 1.11652574153499e-05, "loss": 0.0052, "step": 15880 }, { "grad_norm": 0.14849026501178741, "learning_rate": 1.1113236390272303e-05, "loss": 0.0069, "step": 15890 }, { "grad_norm": 0.1427428424358368, "learning_rate": 1.106132168012155e-05, "loss": 0.005, "step": 15900 }, { "grad_norm": 0.140606090426445, "learning_rate": 1.1009513426830448e-05, "loss": 0.0059, "step": 15910 }, { "grad_norm": 0.15393853187561035, "learning_rate": 1.0957811772040777e-05, "loss": 0.0047, "step": 15920 }, { "grad_norm": 0.093324214220047, "learning_rate": 1.0906216857102913e-05, "loss": 0.005, "step": 15930 }, { "grad_norm": 0.12069685012102127, "learning_rate": 1.0854728823075355e-05, "loss": 0.0048, "step": 15940 }, { "grad_norm": 0.5936830639839172, "learning_rate": 1.0803347810724452e-05, "loss": 0.0058, "step": 15950 }, { "grad_norm": 0.20460212230682373, "learning_rate": 1.0752073960523911e-05, "loss": 0.0051, "step": 15960 }, { "grad_norm": 0.16782912611961365, "learning_rate": 1.070090741265447e-05, "loss": 0.0052, "step": 15970 }, { "grad_norm": 0.2064383178949356, "learning_rate": 1.0649848307003547e-05, "loss": 0.0062, "step": 15980 }, { "grad_norm": 0.12357888370752335, "learning_rate": 1.0598896783164757e-05, "loss": 0.006, "step": 15990 }, { "grad_norm": 0.09672047942876816, "learning_rate": 1.0548052980437645e-05, "loss": 0.0066, "step": 16000 }, { "grad_norm": 0.10122738033533096, "learning_rate": 1.049731703782722e-05, "loss": 0.0055, "step": 16010 }, { "grad_norm": 0.14516356587409973, "learning_rate": 1.0446689094043587e-05, "loss": 0.005, "step": 16020 }, { "grad_norm": 0.2363949567079544, "learning_rate": 1.039616928750165e-05, "loss": 0.0046, "step": 16030 }, { "grad_norm": 0.10833202302455902, "learning_rate": 1.0345757756320612e-05, "loss": 0.0051, "step": 16040 }, { "grad_norm": 0.14556190371513367, "learning_rate": 1.0295454638323666e-05, "loss": 0.0055, "step": 16050 }, { "grad_norm": 0.16646552085876465, "learning_rate": 1.0245260071037632e-05, "loss": 0.0054, "step": 16060 }, { "grad_norm": 0.1648065149784088, "learning_rate": 1.0195174191692518e-05, "loss": 0.005, "step": 16070 }, { "grad_norm": 0.3625747859477997, "learning_rate": 1.014519713722124e-05, "loss": 0.0053, "step": 16080 }, { "grad_norm": 0.15022151172161102, "learning_rate": 1.0095329044259132e-05, "loss": 0.008, "step": 16090 }, { "grad_norm": 0.10850266367197037, "learning_rate": 1.004557004914365e-05, "loss": 0.0037, "step": 16100 }, { "grad_norm": 0.1304628998041153, "learning_rate": 9.995920287914007e-06, "loss": 0.0049, "step": 16110 }, { "grad_norm": 0.09549619257450104, "learning_rate": 9.946379896310737e-06, "loss": 0.0049, "step": 16120 }, { "grad_norm": 0.10411006212234497, "learning_rate": 9.896949009775396e-06, "loss": 0.0056, "step": 16130 }, { "grad_norm": 0.1455530971288681, "learning_rate": 9.847627763450134e-06, "loss": 0.0054, "step": 16140 }, { "grad_norm": 0.14030906558036804, "learning_rate": 9.798416292177337e-06, "loss": 0.005, "step": 16150 }, { "grad_norm": 0.08618499338626862, "learning_rate": 9.74931473049932e-06, "loss": 0.0046, "step": 16160 }, { "grad_norm": 0.16993960738182068, "learning_rate": 9.700323212657847e-06, "loss": 0.0045, "step": 16170 }, { "grad_norm": 0.11044935137033463, "learning_rate": 9.65144187259388e-06, "loss": 0.0053, "step": 16180 }, { "grad_norm": 0.17418694496154785, "learning_rate": 9.602670843947132e-06, "loss": 0.005, "step": 16190 }, { "grad_norm": 0.055452197790145874, "learning_rate": 9.554010260055713e-06, "loss": 0.0067, "step": 16200 }, { "grad_norm": 0.11019381880760193, "learning_rate": 9.505460253955834e-06, "loss": 0.0053, "step": 16210 }, { "grad_norm": 0.1988607794046402, "learning_rate": 9.457020958381324e-06, "loss": 0.0054, "step": 16220 }, { "grad_norm": 0.11723852157592773, "learning_rate": 9.408692505763395e-06, "loss": 0.0049, "step": 16230 }, { "grad_norm": 0.13030314445495605, "learning_rate": 9.360475028230181e-06, "loss": 0.0047, "step": 16240 }, { "grad_norm": 0.1287561058998108, "learning_rate": 9.312368657606412e-06, "loss": 0.0059, "step": 16250 }, { "grad_norm": 0.11027056723833084, "learning_rate": 9.264373525413096e-06, "loss": 0.0052, "step": 16260 }, { "grad_norm": 0.14443908631801605, "learning_rate": 9.216489762867058e-06, "loss": 0.0063, "step": 16270 }, { "grad_norm": 0.10529366135597229, "learning_rate": 9.168717500880708e-06, "loss": 0.0046, "step": 16280 }, { "grad_norm": 0.1178966760635376, "learning_rate": 9.121056870061574e-06, "loss": 0.0068, "step": 16290 }, { "grad_norm": 0.10544776171445847, "learning_rate": 9.073508000711983e-06, "loss": 0.0056, "step": 16300 }, { "grad_norm": 0.07063297182321548, "learning_rate": 9.026071022828758e-06, "loss": 0.0063, "step": 16310 }, { "grad_norm": 0.13934491574764252, "learning_rate": 8.978746066102771e-06, "loss": 0.0052, "step": 16320 }, { "grad_norm": 0.10978896915912628, "learning_rate": 8.931533259918634e-06, "loss": 0.0048, "step": 16330 }, { "grad_norm": 0.11833730340003967, "learning_rate": 8.884432733354382e-06, "loss": 0.0055, "step": 16340 }, { "grad_norm": 0.14003394544124603, "learning_rate": 8.837444615181029e-06, "loss": 0.0043, "step": 16350 }, { "grad_norm": 0.08632051944732666, "learning_rate": 8.790569033862323e-06, "loss": 0.0046, "step": 16360 }, { "grad_norm": 0.13919036090373993, "learning_rate": 8.7438061175543e-06, "loss": 0.0044, "step": 16370 }, { "grad_norm": 0.11670856922864914, "learning_rate": 8.697155994104978e-06, "loss": 0.0053, "step": 16380 }, { "grad_norm": 0.09226898103952408, "learning_rate": 8.650618791054033e-06, "loss": 0.0055, "step": 16390 }, { "grad_norm": 0.18644751608371735, "learning_rate": 8.604194635632373e-06, "loss": 0.0047, "step": 16400 }, { "grad_norm": 0.1540301889181137, "learning_rate": 8.557883654761906e-06, "loss": 0.0048, "step": 16410 }, { "grad_norm": 0.14981454610824585, "learning_rate": 8.511685975055061e-06, "loss": 0.0051, "step": 16420 }, { "grad_norm": 0.10548312962055206, "learning_rate": 8.46560172281452e-06, "loss": 0.0054, "step": 16430 }, { "grad_norm": 0.18674735724925995, "learning_rate": 8.419631024032893e-06, "loss": 0.0054, "step": 16440 }, { "grad_norm": 0.11619240790605545, "learning_rate": 8.373774004392293e-06, "loss": 0.0047, "step": 16450 }, { "grad_norm": 0.09176240861415863, "learning_rate": 8.32803078926409e-06, "loss": 0.004, "step": 16460 }, { "grad_norm": 0.1235954612493515, "learning_rate": 8.282401503708454e-06, "loss": 0.0042, "step": 16470 }, { "grad_norm": 0.12665116786956787, "learning_rate": 8.23688627247412e-06, "loss": 0.0051, "step": 16480 }, { "grad_norm": 0.13140486180782318, "learning_rate": 8.191485219998007e-06, "loss": 0.0058, "step": 16490 }, { "grad_norm": 0.10530330240726471, "learning_rate": 8.146198470404843e-06, "loss": 0.0039, "step": 16500 }, { "grad_norm": 0.14705541729927063, "learning_rate": 8.101026147506897e-06, "loss": 0.0045, "step": 16510 }, { "grad_norm": 0.11801851540803909, "learning_rate": 8.05596837480353e-06, "loss": 0.0043, "step": 16520 }, { "grad_norm": 0.21611303091049194, "learning_rate": 8.011025275480998e-06, "loss": 0.005, "step": 16530 }, { "grad_norm": 0.15860483050346375, "learning_rate": 7.966196972412027e-06, "loss": 0.0053, "step": 16540 }, { "grad_norm": 0.08368507772684097, "learning_rate": 7.92148358815547e-06, "loss": 0.0042, "step": 16550 }, { "grad_norm": 0.08468005806207657, "learning_rate": 7.87688524495604e-06, "loss": 0.0046, "step": 16560 }, { "grad_norm": 0.1131843775510788, "learning_rate": 7.83240206474386e-06, "loss": 0.0045, "step": 16570 }, { "grad_norm": 0.09109779447317123, "learning_rate": 7.788034169134272e-06, "loss": 0.0052, "step": 16580 }, { "grad_norm": 0.1052711084485054, "learning_rate": 7.743781679427414e-06, "loss": 0.0049, "step": 16590 }, { "grad_norm": 0.06818226724863052, "learning_rate": 7.699644716607895e-06, "loss": 0.0063, "step": 16600 }, { "grad_norm": 0.09830304235219955, "learning_rate": 7.655623401344486e-06, "loss": 0.0045, "step": 16610 }, { "grad_norm": 0.11249382048845291, "learning_rate": 7.611717853989775e-06, "loss": 0.0042, "step": 16620 }, { "grad_norm": 0.10611110180616379, "learning_rate": 7.567928194579854e-06, "loss": 0.0051, "step": 16630 }, { "grad_norm": 0.07424785941839218, "learning_rate": 7.524254542833997e-06, "loss": 0.0048, "step": 16640 }, { "grad_norm": 0.10688845068216324, "learning_rate": 7.480697018154286e-06, "loss": 0.0048, "step": 16650 }, { "grad_norm": 0.10375754535198212, "learning_rate": 7.437255739625332e-06, "loss": 0.0049, "step": 16660 }, { "grad_norm": 0.11682040989398956, "learning_rate": 7.393930826013923e-06, "loss": 0.0057, "step": 16670 }, { "grad_norm": 0.14548049867153168, "learning_rate": 7.350722395768722e-06, "loss": 0.0052, "step": 16680 }, { "grad_norm": 0.11602824181318283, "learning_rate": 7.307630567019963e-06, "loss": 0.0072, "step": 16690 }, { "grad_norm": 0.15168939530849457, "learning_rate": 7.264655457579e-06, "loss": 0.0062, "step": 16700 }, { "grad_norm": 0.10517384111881256, "learning_rate": 7.221797184938184e-06, "loss": 0.0043, "step": 16710 }, { "grad_norm": 0.12812364101409912, "learning_rate": 7.179055866270373e-06, "loss": 0.0047, "step": 16720 }, { "grad_norm": 0.09526520222425461, "learning_rate": 7.136431618428707e-06, "loss": 0.0043, "step": 16730 }, { "grad_norm": 0.14298106729984283, "learning_rate": 7.09392455794628e-06, "loss": 0.0044, "step": 16740 }, { "grad_norm": 0.12755927443504333, "learning_rate": 7.051534801035725e-06, "loss": 0.0049, "step": 16750 }, { "grad_norm": 0.10625048726797104, "learning_rate": 7.00926246358905e-06, "loss": 0.0049, "step": 16760 }, { "grad_norm": 0.05810859426856041, "learning_rate": 6.967107661177191e-06, "loss": 0.0045, "step": 16770 }, { "grad_norm": 0.11782581359148026, "learning_rate": 6.925070509049786e-06, "loss": 0.0045, "step": 16780 }, { "grad_norm": 0.06086992472410202, "learning_rate": 6.883151122134812e-06, "loss": 0.0044, "step": 16790 }, { "grad_norm": 0.0968971997499466, "learning_rate": 6.8413496150382394e-06, "loss": 0.0054, "step": 16800 }, { "grad_norm": 0.14773517847061157, "learning_rate": 6.7996661020438165e-06, "loss": 0.0055, "step": 16810 }, { "grad_norm": 0.13275542855262756, "learning_rate": 6.758100697112662e-06, "loss": 0.0047, "step": 16820 }, { "grad_norm": 0.09497711062431335, "learning_rate": 6.716653513883026e-06, "loss": 0.0057, "step": 16830 }, { "grad_norm": 0.11659703403711319, "learning_rate": 6.675324665669913e-06, "loss": 0.005, "step": 16840 }, { "grad_norm": 0.17388087511062622, "learning_rate": 6.634114265464803e-06, "loss": 0.0041, "step": 16850 }, { "grad_norm": 0.11791954934597015, "learning_rate": 6.59302242593538e-06, "loss": 0.0044, "step": 16860 }, { "grad_norm": 0.17051368951797485, "learning_rate": 6.552049259425141e-06, "loss": 0.0052, "step": 16870 }, { "grad_norm": 0.08162733167409897, "learning_rate": 6.511194877953181e-06, "loss": 0.0046, "step": 16880 }, { "grad_norm": 0.09056273847818375, "learning_rate": 6.470459393213813e-06, "loss": 0.0043, "step": 16890 }, { "grad_norm": 0.14687728881835938, "learning_rate": 6.429842916576279e-06, "loss": 0.0045, "step": 16900 }, { "grad_norm": 0.11188974231481552, "learning_rate": 6.389345559084503e-06, "loss": 0.0047, "step": 16910 }, { "grad_norm": 0.10093161463737488, "learning_rate": 6.348967431456682e-06, "loss": 0.0041, "step": 16920 }, { "grad_norm": 0.10220538079738617, "learning_rate": 6.30870864408511e-06, "loss": 0.0039, "step": 16930 }, { "grad_norm": 0.09332132339477539, "learning_rate": 6.268569307035754e-06, "loss": 0.0039, "step": 16940 }, { "grad_norm": 0.11174672096967697, "learning_rate": 6.228549530048022e-06, "loss": 0.0039, "step": 16950 }, { "grad_norm": 0.07925552874803543, "learning_rate": 6.1886494225344814e-06, "loss": 0.004, "step": 16960 }, { "grad_norm": 0.10965389758348465, "learning_rate": 6.148869093580479e-06, "loss": 0.0029, "step": 16970 }, { "grad_norm": 0.11813022196292877, "learning_rate": 6.109208651943921e-06, "loss": 0.0039, "step": 16980 }, { "grad_norm": 0.18598681688308716, "learning_rate": 6.069668206054946e-06, "loss": 0.0063, "step": 16990 }, { "grad_norm": 0.14955657720565796, "learning_rate": 6.0302478640156145e-06, "loss": 0.0048, "step": 17000 }, { "grad_norm": 0.08954836428165436, "learning_rate": 5.990947733599644e-06, "loss": 0.0054, "step": 17010 }, { "grad_norm": 0.18158337473869324, "learning_rate": 5.951767922252105e-06, "loss": 0.0049, "step": 17020 }, { "grad_norm": 0.23328961431980133, "learning_rate": 5.912708537089068e-06, "loss": 0.0041, "step": 17030 }, { "grad_norm": 0.14590199291706085, "learning_rate": 5.873769684897434e-06, "loss": 0.0057, "step": 17040 }, { "grad_norm": 0.10576408356428146, "learning_rate": 5.834951472134514e-06, "loss": 0.0037, "step": 17050 }, { "grad_norm": 0.05856657773256302, "learning_rate": 5.796254004927832e-06, "loss": 0.0062, "step": 17060 }, { "grad_norm": 0.1027015820145607, "learning_rate": 5.757677389074806e-06, "loss": 0.0047, "step": 17070 }, { "grad_norm": 0.0811288058757782, "learning_rate": 5.719221730042385e-06, "loss": 0.005, "step": 17080 }, { "grad_norm": 0.10462873429059982, "learning_rate": 5.680887132966911e-06, "loss": 0.0044, "step": 17090 }, { "grad_norm": 0.11901343613862991, "learning_rate": 5.642673702653683e-06, "loss": 0.005, "step": 17100 }, { "grad_norm": 0.09503764659166336, "learning_rate": 5.604581543576781e-06, "loss": 0.0055, "step": 17110 }, { "grad_norm": 0.07215610891580582, "learning_rate": 5.566610759878704e-06, "loss": 0.0039, "step": 17120 }, { "grad_norm": 0.07807325571775436, "learning_rate": 5.528761455370119e-06, "loss": 0.0051, "step": 17130 }, { "grad_norm": 0.1379242241382599, "learning_rate": 5.491033733529594e-06, "loss": 0.0047, "step": 17140 }, { "grad_norm": 0.22390474379062653, "learning_rate": 5.453427697503255e-06, "loss": 0.0048, "step": 17150 }, { "grad_norm": 0.10660412162542343, "learning_rate": 5.415943450104599e-06, "loss": 0.0045, "step": 17160 }, { "grad_norm": 0.10713186860084534, "learning_rate": 5.378581093814111e-06, "loss": 0.0037, "step": 17170 }, { "grad_norm": 0.19166870415210724, "learning_rate": 5.3413407307790375e-06, "loss": 0.0047, "step": 17180 }, { "grad_norm": 0.14736299216747284, "learning_rate": 5.30422246281313e-06, "loss": 0.0045, "step": 17190 }, { "grad_norm": 0.08934909105300903, "learning_rate": 5.267226391396296e-06, "loss": 0.0056, "step": 17200 }, { "grad_norm": 0.11917974054813385, "learning_rate": 5.2303526176744e-06, "loss": 0.0041, "step": 17210 }, { "grad_norm": 0.3687226176261902, "learning_rate": 5.193601242458929e-06, "loss": 0.0053, "step": 17220 }, { "grad_norm": 0.14266933500766754, "learning_rate": 5.156972366226714e-06, "loss": 0.0047, "step": 17230 }, { "grad_norm": 0.0723942369222641, "learning_rate": 5.120466089119735e-06, "loss": 0.0034, "step": 17240 }, { "grad_norm": 0.09861062467098236, "learning_rate": 5.084082510944749e-06, "loss": 0.0043, "step": 17250 }, { "grad_norm": 0.09976638853549957, "learning_rate": 5.047821731173058e-06, "loss": 0.0049, "step": 17260 }, { "grad_norm": 0.0760389044880867, "learning_rate": 5.011683848940274e-06, "loss": 0.0035, "step": 17270 }, { "grad_norm": 0.08932527154684067, "learning_rate": 4.975668963045954e-06, "loss": 0.0044, "step": 17280 }, { "grad_norm": 0.07628341764211655, "learning_rate": 4.9397771719534525e-06, "loss": 0.0043, "step": 17290 }, { "grad_norm": 0.16545511782169342, "learning_rate": 4.904008573789548e-06, "loss": 0.0047, "step": 17300 }, { "grad_norm": 0.15013083815574646, "learning_rate": 4.8683632663442005e-06, "loss": 0.0053, "step": 17310 }, { "grad_norm": 0.1142745390534401, "learning_rate": 4.832841347070343e-06, "loss": 0.0045, "step": 17320 }, { "grad_norm": 0.09129402786493301, "learning_rate": 4.797442913083539e-06, "loss": 0.0043, "step": 17330 }, { "grad_norm": 0.11391004920005798, "learning_rate": 4.7621680611617596e-06, "loss": 0.0037, "step": 17340 }, { "grad_norm": 0.09616543352603912, "learning_rate": 4.727016887745095e-06, "loss": 0.0042, "step": 17350 }, { "grad_norm": 0.2256506234407425, "learning_rate": 4.691989488935511e-06, "loss": 0.0043, "step": 17360 }, { "grad_norm": 0.11314600706100464, "learning_rate": 4.657085960496588e-06, "loss": 0.0048, "step": 17370 }, { "grad_norm": 0.09878886491060257, "learning_rate": 4.6223063978532265e-06, "loss": 0.004, "step": 17380 }, { "grad_norm": 0.15160433948040009, "learning_rate": 4.587650896091439e-06, "loss": 0.0049, "step": 17390 }, { "grad_norm": 0.15633244812488556, "learning_rate": 4.553119549958035e-06, "loss": 0.0059, "step": 17400 }, { "grad_norm": 0.23170021176338196, "learning_rate": 4.518712453860385e-06, "loss": 0.0038, "step": 17410 }, { "grad_norm": 0.06972615420818329, "learning_rate": 4.484429701866205e-06, "loss": 0.0046, "step": 17420 }, { "grad_norm": 0.07664463669061661, "learning_rate": 4.4502713877031975e-06, "loss": 0.0043, "step": 17430 }, { "grad_norm": 0.0626700296998024, "learning_rate": 4.416237604758911e-06, "loss": 0.0033, "step": 17440 }, { "grad_norm": 0.09504947811365128, "learning_rate": 4.3823284460804025e-06, "loss": 0.004, "step": 17450 }, { "grad_norm": 0.27570420503616333, "learning_rate": 4.348544004374011e-06, "loss": 0.005, "step": 17460 }, { "grad_norm": 0.13640616834163666, "learning_rate": 4.314884372005123e-06, "loss": 0.0051, "step": 17470 }, { "grad_norm": 0.12549041211605072, "learning_rate": 4.281349640997867e-06, "loss": 0.0043, "step": 17480 }, { "grad_norm": 0.07296334207057953, "learning_rate": 4.247939903034942e-06, "loss": 0.0049, "step": 17490 }, { "grad_norm": 0.08326546102762222, "learning_rate": 4.214655249457284e-06, "loss": 0.0036, "step": 17500 }, { "grad_norm": 0.11816396564245224, "learning_rate": 4.181495771263855e-06, "loss": 0.0036, "step": 17510 }, { "grad_norm": 0.12310322374105453, "learning_rate": 4.148461559111427e-06, "loss": 0.005, "step": 17520 }, { "grad_norm": 0.0948910191655159, "learning_rate": 4.115552703314252e-06, "loss": 0.006, "step": 17530 }, { "grad_norm": 0.11463034152984619, "learning_rate": 4.082769293843886e-06, "loss": 0.0038, "step": 17540 }, { "grad_norm": 0.17851631343364716, "learning_rate": 4.050111420328939e-06, "loss": 0.0043, "step": 17550 }, { "grad_norm": 0.08955517411231995, "learning_rate": 4.017579172054764e-06, "loss": 0.0042, "step": 17560 }, { "grad_norm": 0.06377570331096649, "learning_rate": 3.985172637963308e-06, "loss": 0.0054, "step": 17570 }, { "grad_norm": 0.12516333162784576, "learning_rate": 3.952891906652784e-06, "loss": 0.0039, "step": 17580 }, { "grad_norm": 0.12787026166915894, "learning_rate": 3.920737066377478e-06, "loss": 0.0046, "step": 17590 }, { "grad_norm": 0.10668189823627472, "learning_rate": 3.888708205047509e-06, "loss": 0.0047, "step": 17600 }, { "grad_norm": 0.24125619232654572, "learning_rate": 3.856805410228542e-06, "loss": 0.0038, "step": 17610 }, { "grad_norm": 0.11872827261686325, "learning_rate": 3.82502876914162e-06, "loss": 0.0037, "step": 17620 }, { "grad_norm": 0.16281035542488098, "learning_rate": 3.7933783686628586e-06, "loss": 0.0047, "step": 17630 }, { "grad_norm": 0.25093579292297363, "learning_rate": 3.7618542953232306e-06, "loss": 0.0053, "step": 17640 }, { "grad_norm": 0.07927124202251434, "learning_rate": 3.7304566353083658e-06, "loss": 0.0044, "step": 17650 }, { "grad_norm": 0.12427982687950134, "learning_rate": 3.6991854744582555e-06, "loss": 0.0048, "step": 17660 }, { "grad_norm": 0.4300476908683777, "learning_rate": 3.6680408982670777e-06, "loss": 0.0032, "step": 17670 }, { "grad_norm": 0.09834052622318268, "learning_rate": 3.637022991882899e-06, "loss": 0.004, "step": 17680 }, { "grad_norm": 0.1160050556063652, "learning_rate": 3.606131840107485e-06, "loss": 0.0055, "step": 17690 }, { "grad_norm": 0.11207425594329834, "learning_rate": 3.575367527396084e-06, "loss": 0.004, "step": 17700 }, { "grad_norm": 0.054858822375535965, "learning_rate": 3.5447301378571386e-06, "loss": 0.0039, "step": 17710 }, { "grad_norm": 0.1281631588935852, "learning_rate": 3.514219755252113e-06, "loss": 0.0043, "step": 17720 }, { "grad_norm": 0.20224082469940186, "learning_rate": 3.4838364629952213e-06, "loss": 0.0044, "step": 17730 }, { "grad_norm": 0.11759786307811737, "learning_rate": 3.4535803441532123e-06, "loss": 0.004, "step": 17740 }, { "grad_norm": 0.059311628341674805, "learning_rate": 3.4234514814451836e-06, "loss": 0.0045, "step": 17750 }, { "grad_norm": 0.18201330304145813, "learning_rate": 3.393449957242273e-06, "loss": 0.0037, "step": 17760 }, { "grad_norm": 0.10715977847576141, "learning_rate": 3.363575853567524e-06, "loss": 0.0037, "step": 17770 }, { "grad_norm": 0.11222989112138748, "learning_rate": 3.3338292520955826e-06, "loss": 0.0045, "step": 17780 }, { "grad_norm": 0.11800965666770935, "learning_rate": 3.304210234152516e-06, "loss": 0.0045, "step": 17790 }, { "grad_norm": 0.1032290831208229, "learning_rate": 3.2747188807155993e-06, "loss": 0.004, "step": 17800 }, { "grad_norm": 0.17009277641773224, "learning_rate": 3.2453552724130643e-06, "loss": 0.0049, "step": 17810 }, { "grad_norm": 0.09233362227678299, "learning_rate": 3.216119489523889e-06, "loss": 0.0038, "step": 17820 }, { "grad_norm": 0.06266705691814423, "learning_rate": 3.1870116119775917e-06, "loss": 0.0037, "step": 17830 }, { "grad_norm": 0.08733876049518585, "learning_rate": 3.158031719353999e-06, "loss": 0.0054, "step": 17840 }, { "grad_norm": 0.1207728311419487, "learning_rate": 3.1291798908830273e-06, "loss": 0.0046, "step": 17850 }, { "grad_norm": 0.3035689890384674, "learning_rate": 3.1004562054444853e-06, "loss": 0.0041, "step": 17860 }, { "grad_norm": 0.11321306973695755, "learning_rate": 3.071860741567806e-06, "loss": 0.005, "step": 17870 }, { "grad_norm": 0.1885700523853302, "learning_rate": 3.04339357743193e-06, "loss": 0.0044, "step": 17880 }, { "grad_norm": 0.21838459372520447, "learning_rate": 3.0150547908649628e-06, "loss": 0.0037, "step": 17890 }, { "grad_norm": 0.0833037793636322, "learning_rate": 2.9868444593440957e-06, "loss": 0.005, "step": 17900 }, { "grad_norm": 0.151641383767128, "learning_rate": 2.9587626599952846e-06, "loss": 0.0047, "step": 17910 }, { "grad_norm": 0.1258433759212494, "learning_rate": 2.930809469593082e-06, "loss": 0.0042, "step": 17920 }, { "grad_norm": 0.13322289288043976, "learning_rate": 2.9029849645604733e-06, "loss": 0.0043, "step": 17930 }, { "grad_norm": 0.06377483159303665, "learning_rate": 2.8752892209685632e-06, "loss": 0.0036, "step": 17940 }, { "grad_norm": 0.11490548402070999, "learning_rate": 2.847722314536483e-06, "loss": 0.0046, "step": 17950 }, { "grad_norm": 0.31529927253723145, "learning_rate": 2.820284320631078e-06, "loss": 0.0036, "step": 17960 }, { "grad_norm": 0.08499549329280853, "learning_rate": 2.792975314266788e-06, "loss": 0.0042, "step": 17970 }, { "grad_norm": 0.05816160887479782, "learning_rate": 2.7657953701054007e-06, "loss": 0.0035, "step": 17980 }, { "grad_norm": 0.15403470396995544, "learning_rate": 2.7387445624558306e-06, "loss": 0.0038, "step": 17990 }, { "grad_norm": 0.16821523010730743, "learning_rate": 2.7118229652739747e-06, "loss": 0.0043, "step": 18000 }, { "grad_norm": 0.10773248225450516, "learning_rate": 2.6850306521624236e-06, "loss": 0.0047, "step": 18010 }, { "grad_norm": 0.06794001162052155, "learning_rate": 2.6583676963703507e-06, "loss": 0.0036, "step": 18020 }, { "grad_norm": 0.13314583897590637, "learning_rate": 2.631834170793268e-06, "loss": 0.0041, "step": 18030 }, { "grad_norm": 0.084385946393013, "learning_rate": 2.6054301479728036e-06, "loss": 0.0038, "step": 18040 }, { "grad_norm": 0.10536587238311768, "learning_rate": 2.579155700096575e-06, "loss": 0.004, "step": 18050 }, { "grad_norm": 0.1555139720439911, "learning_rate": 2.5530108989978873e-06, "loss": 0.0033, "step": 18060 }, { "grad_norm": 0.19492387771606445, "learning_rate": 2.5269958161556416e-06, "loss": 0.0043, "step": 18070 }, { "grad_norm": 0.08726193010807037, "learning_rate": 2.5011105226940888e-06, "loss": 0.0037, "step": 18080 }, { "grad_norm": 0.17334680259227753, "learning_rate": 2.4753550893826248e-06, "loss": 0.005, "step": 18090 }, { "grad_norm": 0.0890972837805748, "learning_rate": 2.4497295866356296e-06, "loss": 0.0037, "step": 18100 }, { "grad_norm": 0.10847669839859009, "learning_rate": 2.424234084512228e-06, "loss": 0.0039, "step": 18110 }, { "grad_norm": 0.10054522007703781, "learning_rate": 2.3988686527161687e-06, "loss": 0.0041, "step": 18120 }, { "grad_norm": 0.09032320231199265, "learning_rate": 2.373633360595573e-06, "loss": 0.0039, "step": 18130 }, { "grad_norm": 0.0800553485751152, "learning_rate": 2.3485282771427585e-06, "loss": 0.0054, "step": 18140 }, { "grad_norm": 0.052749328315258026, "learning_rate": 2.3235534709940665e-06, "loss": 0.0032, "step": 18150 }, { "grad_norm": 0.05785638466477394, "learning_rate": 2.2987090104296617e-06, "loss": 0.0036, "step": 18160 }, { "grad_norm": 0.11513590067625046, "learning_rate": 2.273994963373355e-06, "loss": 0.0049, "step": 18170 }, { "grad_norm": 0.08320346474647522, "learning_rate": 2.249411397392409e-06, "loss": 0.004, "step": 18180 }, { "grad_norm": 0.07805438339710236, "learning_rate": 2.2249583796973506e-06, "loss": 0.0035, "step": 18190 }, { "grad_norm": 0.11261925846338272, "learning_rate": 2.200635977141796e-06, "loss": 0.0034, "step": 18200 }, { "grad_norm": 0.09928590804338455, "learning_rate": 2.17644425622226e-06, "loss": 0.0041, "step": 18210 }, { "grad_norm": 0.1457936316728592, "learning_rate": 2.152383283077991e-06, "loss": 0.0054, "step": 18220 }, { "grad_norm": 0.11142061650753021, "learning_rate": 2.128453123490781e-06, "loss": 0.0048, "step": 18230 }, { "grad_norm": 0.1675397753715515, "learning_rate": 2.1046538428847462e-06, "loss": 0.0045, "step": 18240 }, { "grad_norm": 0.16817127168178558, "learning_rate": 2.0809855063262273e-06, "loss": 0.0041, "step": 18250 }, { "grad_norm": 0.08730722963809967, "learning_rate": 2.057448178523558e-06, "loss": 0.0042, "step": 18260 }, { "grad_norm": 0.115006223320961, "learning_rate": 2.034041923826885e-06, "loss": 0.0041, "step": 18270 }, { "grad_norm": 0.08479628711938858, "learning_rate": 2.0107668062280204e-06, "loss": 0.0041, "step": 18280 }, { "grad_norm": 0.06060807406902313, "learning_rate": 1.9876228893602357e-06, "loss": 0.0039, "step": 18290 }, { "grad_norm": 0.08128322660923004, "learning_rate": 1.9646102364981266e-06, "loss": 0.0037, "step": 18300 }, { "grad_norm": 0.10805995762348175, "learning_rate": 1.9417289105574053e-06, "loss": 0.005, "step": 18310 }, { "grad_norm": 0.06586245447397232, "learning_rate": 1.9189789740947427e-06, "loss": 0.0034, "step": 18320 }, { "grad_norm": 0.1522088497877121, "learning_rate": 1.896360489307597e-06, "loss": 0.0044, "step": 18330 }, { "grad_norm": 0.09820322692394257, "learning_rate": 1.8738735180340362e-06, "loss": 0.0032, "step": 18340 }, { "grad_norm": 0.09084205329418182, "learning_rate": 1.8515181217525824e-06, "loss": 0.0032, "step": 18350 }, { "grad_norm": 0.07543554902076721, "learning_rate": 1.8292943615820457e-06, "loss": 0.0037, "step": 18360 }, { "grad_norm": 0.12446766346693039, "learning_rate": 1.8072022982813296e-06, "loss": 0.0038, "step": 18370 }, { "grad_norm": 0.10695412009954453, "learning_rate": 1.7852419922492925e-06, "loss": 0.0054, "step": 18380 }, { "grad_norm": 0.11815108358860016, "learning_rate": 1.763413503524569e-06, "loss": 0.0037, "step": 18390 }, { "grad_norm": 0.08739940077066422, "learning_rate": 1.7417168917854165e-06, "loss": 0.0034, "step": 18400 }, { "grad_norm": 0.07336952537298203, "learning_rate": 1.720152216349552e-06, "loss": 0.005, "step": 18410 }, { "grad_norm": 0.08150866627693176, "learning_rate": 1.6987195361739595e-06, "loss": 0.0038, "step": 18420 }, { "grad_norm": 0.08547519892454147, "learning_rate": 1.6774189098547832e-06, "loss": 0.005, "step": 18430 }, { "grad_norm": 0.0670129805803299, "learning_rate": 1.6562503956271069e-06, "loss": 0.0037, "step": 18440 }, { "grad_norm": 0.10758806765079498, "learning_rate": 1.6352140513648417e-06, "loss": 0.0045, "step": 18450 }, { "grad_norm": 0.08748574554920197, "learning_rate": 1.6143099345805712e-06, "loss": 0.0038, "step": 18460 }, { "grad_norm": 0.06216035410761833, "learning_rate": 1.5935381024253293e-06, "loss": 0.0035, "step": 18470 }, { "grad_norm": 0.09668684005737305, "learning_rate": 1.572898611688517e-06, "loss": 0.0033, "step": 18480 }, { "grad_norm": 0.07534801214933395, "learning_rate": 1.5523915187977133e-06, "loss": 0.0034, "step": 18490 }, { "grad_norm": 0.07793285697698593, "learning_rate": 1.532016879818532e-06, "loss": 0.0036, "step": 18500 }, { "grad_norm": 0.055785562843084335, "learning_rate": 1.51177475045447e-06, "loss": 0.0045, "step": 18510 }, { "grad_norm": 0.0543552003800869, "learning_rate": 1.4916651860467035e-06, "loss": 0.0048, "step": 18520 }, { "grad_norm": 0.08041830360889435, "learning_rate": 1.471688241574043e-06, "loss": 0.0029, "step": 18530 }, { "grad_norm": 0.13896146416664124, "learning_rate": 1.451843971652672e-06, "loss": 0.0043, "step": 18540 }, { "grad_norm": 0.10504639148712158, "learning_rate": 1.432132430536076e-06, "loss": 0.0037, "step": 18550 }, { "grad_norm": 0.18970540165901184, "learning_rate": 1.412553672114869e-06, "loss": 0.004, "step": 18560 }, { "grad_norm": 0.12213291972875595, "learning_rate": 1.3931077499166056e-06, "loss": 0.0056, "step": 18570 }, { "grad_norm": 0.06228169798851013, "learning_rate": 1.3737947171057085e-06, "loss": 0.0034, "step": 18580 }, { "grad_norm": 0.078560009598732, "learning_rate": 1.3546146264832582e-06, "loss": 0.004, "step": 18590 }, { "grad_norm": 0.11034126579761505, "learning_rate": 1.3355675304869086e-06, "loss": 0.0041, "step": 18600 }, { "grad_norm": 0.06909634917974472, "learning_rate": 1.3166534811906827e-06, "loss": 0.0048, "step": 18610 }, { "grad_norm": 0.13800327479839325, "learning_rate": 1.2978725303048666e-06, "loss": 0.003, "step": 18620 }, { "grad_norm": 0.06910396367311478, "learning_rate": 1.2792247291758762e-06, "loss": 0.0041, "step": 18630 }, { "grad_norm": 0.09044212847948074, "learning_rate": 1.2607101287860635e-06, "loss": 0.0038, "step": 18640 }, { "grad_norm": 0.10872463881969452, "learning_rate": 1.2423287797536654e-06, "loss": 0.0033, "step": 18650 }, { "grad_norm": 0.04557178169488907, "learning_rate": 1.2240807323325776e-06, "loss": 0.004, "step": 18660 }, { "grad_norm": 0.09198397397994995, "learning_rate": 1.205966036412254e-06, "loss": 0.0027, "step": 18670 }, { "grad_norm": 0.06418166309595108, "learning_rate": 1.1879847415175949e-06, "loss": 0.004, "step": 18680 }, { "grad_norm": 0.08102408051490784, "learning_rate": 1.1701368968087712e-06, "loss": 0.0035, "step": 18690 }, { "grad_norm": 0.11330026388168335, "learning_rate": 1.1524225510811116e-06, "loss": 0.0035, "step": 18700 }, { "grad_norm": 0.10285083949565887, "learning_rate": 1.1348417527649535e-06, "loss": 0.0038, "step": 18710 }, { "grad_norm": 0.0938316360116005, "learning_rate": 1.1173945499255268e-06, "loss": 0.0037, "step": 18720 }, { "grad_norm": 0.16485825181007385, "learning_rate": 1.1000809902628307e-06, "loss": 0.0032, "step": 18730 }, { "grad_norm": 0.05527719110250473, "learning_rate": 1.082901121111468e-06, "loss": 0.0041, "step": 18740 }, { "grad_norm": 0.18675890564918518, "learning_rate": 1.0658549894405456e-06, "loss": 0.0041, "step": 18750 }, { "grad_norm": 0.18073934316635132, "learning_rate": 1.0489426418535342e-06, "loss": 0.0037, "step": 18760 }, { "grad_norm": 0.14849156141281128, "learning_rate": 1.0321641245881474e-06, "loss": 0.0037, "step": 18770 }, { "grad_norm": 0.07752328366041183, "learning_rate": 1.015519483516214e-06, "loss": 0.0037, "step": 18780 }, { "grad_norm": 0.07465663552284241, "learning_rate": 9.990087641435443e-07, "loss": 0.0045, "step": 18790 }, { "grad_norm": 0.07782334089279175, "learning_rate": 9.826320116098132e-07, "loss": 0.0045, "step": 18800 }, { "grad_norm": 0.09546821564435959, "learning_rate": 9.663892706884447e-07, "loss": 0.0046, "step": 18810 }, { "grad_norm": 0.4725317060947418, "learning_rate": 9.502805857864616e-07, "loss": 0.0045, "step": 18820 }, { "grad_norm": 0.17174983024597168, "learning_rate": 9.34306000944396e-07, "loss": 0.0035, "step": 18830 }, { "grad_norm": 0.06762441247701645, "learning_rate": 9.184655598361624e-07, "loss": 0.0044, "step": 18840 }, { "grad_norm": 0.11659327149391174, "learning_rate": 9.027593057689076e-07, "loss": 0.004, "step": 18850 }, { "grad_norm": 0.21454183757305145, "learning_rate": 8.871872816829441e-07, "loss": 0.0045, "step": 18860 }, { "grad_norm": 0.11002448201179504, "learning_rate": 8.717495301515777e-07, "loss": 0.0036, "step": 18870 }, { "grad_norm": 0.11095000058412552, "learning_rate": 8.564460933810415e-07, "loss": 0.0039, "step": 18880 }, { "grad_norm": 0.11998914927244186, "learning_rate": 8.412770132103453e-07, "loss": 0.0038, "step": 18890 }, { "grad_norm": 0.09635243564844131, "learning_rate": 8.262423311111711e-07, "loss": 0.0035, "step": 18900 }, { "grad_norm": 0.08217710256576538, "learning_rate": 8.113420881877665e-07, "loss": 0.0034, "step": 18910 }, { "grad_norm": 0.36349567770957947, "learning_rate": 7.965763251768288e-07, "loss": 0.0043, "step": 18920 }, { "grad_norm": 0.14502088725566864, "learning_rate": 7.819450824473995e-07, "loss": 0.0033, "step": 18930 }, { "grad_norm": 0.12831059098243713, "learning_rate": 7.674484000007198e-07, "loss": 0.0032, "step": 18940 }, { "grad_norm": 0.20514126121997833, "learning_rate": 7.530863174701752e-07, "loss": 0.0047, "step": 18950 }, { "grad_norm": 0.18135295808315277, "learning_rate": 7.38858874121151e-07, "loss": 0.0033, "step": 18960 }, { "grad_norm": 0.10481374710798264, "learning_rate": 7.247661088509328e-07, "loss": 0.0046, "step": 18970 }, { "grad_norm": 0.11239234358072281, "learning_rate": 7.108080601886002e-07, "loss": 0.004, "step": 18980 }, { "grad_norm": 0.2265988290309906, "learning_rate": 6.969847662949336e-07, "loss": 0.0041, "step": 18990 }, { "grad_norm": 0.11195097118616104, "learning_rate": 6.832962649622798e-07, "loss": 0.004, "step": 19000 }, { "grad_norm": 0.15957877039909363, "learning_rate": 6.697425936144863e-07, "loss": 0.003, "step": 19010 }, { "grad_norm": 0.08707629889249802, "learning_rate": 6.563237893067731e-07, "loss": 0.0033, "step": 19020 }, { "grad_norm": 0.11115802824497223, "learning_rate": 6.430398887256328e-07, "loss": 0.0041, "step": 19030 }, { "grad_norm": 0.21788974106311798, "learning_rate": 6.298909281887478e-07, "loss": 0.0041, "step": 19040 }, { "grad_norm": 0.3119131028652191, "learning_rate": 6.168769436448673e-07, "loss": 0.0043, "step": 19050 }, { "grad_norm": 0.09872477501630783, "learning_rate": 6.03997970673742e-07, "loss": 0.0049, "step": 19060 }, { "grad_norm": 0.09719311445951462, "learning_rate": 5.912540444859782e-07, "loss": 0.0036, "step": 19070 }, { "grad_norm": 0.08987168967723846, "learning_rate": 5.786451999229837e-07, "loss": 0.0042, "step": 19080 }, { "grad_norm": 0.1413099467754364, "learning_rate": 5.661714714568722e-07, "loss": 0.0041, "step": 19090 }, { "grad_norm": 0.37126514315605164, "learning_rate": 5.538328931903259e-07, "loss": 0.0048, "step": 19100 }, { "grad_norm": 0.10437651723623276, "learning_rate": 5.416294988565551e-07, "loss": 0.0055, "step": 19110 }, { "grad_norm": 0.1415119767189026, "learning_rate": 5.29561321819172e-07, "loss": 0.0045, "step": 19120 }, { "grad_norm": 0.25738152861595154, "learning_rate": 5.176283950721061e-07, "loss": 0.0038, "step": 19130 }, { "grad_norm": 0.1541583240032196, "learning_rate": 5.058307512395332e-07, "loss": 0.0043, "step": 19140 }, { "grad_norm": 0.3246905207633972, "learning_rate": 4.941684225757526e-07, "loss": 0.0043, "step": 19150 }, { "grad_norm": 0.12075149267911911, "learning_rate": 4.826414409651314e-07, "loss": 0.0047, "step": 19160 }, { "grad_norm": 0.10267707705497742, "learning_rate": 4.712498379219943e-07, "loss": 0.0035, "step": 19170 }, { "grad_norm": 0.09861812740564346, "learning_rate": 4.599936445905506e-07, "loss": 0.004, "step": 19180 }, { "grad_norm": 0.09127381443977356, "learning_rate": 4.4887289174480594e-07, "loss": 0.0041, "step": 19190 }, { "grad_norm": 0.08192761242389679, "learning_rate": 4.378876097884621e-07, "loss": 0.0041, "step": 19200 }, { "grad_norm": 0.1953418105840683, "learning_rate": 4.2703782875487264e-07, "loss": 0.0036, "step": 19210 }, { "grad_norm": 0.10277601331472397, "learning_rate": 4.163235783069208e-07, "loss": 0.0045, "step": 19220 }, { "grad_norm": 0.09032998979091644, "learning_rate": 4.057448877369585e-07, "loss": 0.0037, "step": 19230 }, { "grad_norm": 0.07254410535097122, "learning_rate": 3.9530178596672295e-07, "loss": 0.0034, "step": 19240 }, { "grad_norm": 0.09380718320608139, "learning_rate": 3.849943015472479e-07, "loss": 0.0056, "step": 19250 }, { "grad_norm": 0.0810515433549881, "learning_rate": 3.748224626588137e-07, "loss": 0.0042, "step": 19260 }, { "grad_norm": 0.10923217982053757, "learning_rate": 3.647862971108307e-07, "loss": 0.003, "step": 19270 }, { "grad_norm": 0.213851198554039, "learning_rate": 3.5488583234179473e-07, "loss": 0.005, "step": 19280 }, { "grad_norm": 0.06757736951112747, "learning_rate": 3.4512109541920413e-07, "loss": 0.0036, "step": 19290 }, { "grad_norm": 0.061954066157341, "learning_rate": 3.354921130394706e-07, "loss": 0.0041, "step": 19300 }, { "grad_norm": 0.06342040747404099, "learning_rate": 3.259989115278639e-07, "loss": 0.0033, "step": 19310 }, { "grad_norm": 0.1331111639738083, "learning_rate": 3.1664151683843403e-07, "loss": 0.0043, "step": 19320 }, { "grad_norm": 0.09285904467105865, "learning_rate": 3.074199545539447e-07, "loss": 0.0036, "step": 19330 }, { "grad_norm": 0.17465020716190338, "learning_rate": 2.983342498857955e-07, "loss": 0.0042, "step": 19340 }, { "grad_norm": 0.07131197303533554, "learning_rate": 2.893844276739499e-07, "loss": 0.0032, "step": 19350 }, { "grad_norm": 0.16769234836101532, "learning_rate": 2.8057051238688514e-07, "loss": 0.0041, "step": 19360 }, { "grad_norm": 0.11477795988321304, "learning_rate": 2.71892528121509e-07, "loss": 0.0045, "step": 19370 }, { "grad_norm": 0.112590491771698, "learning_rate": 2.633504986030988e-07, "loss": 0.0039, "step": 19380 }, { "grad_norm": 0.13309821486473083, "learning_rate": 2.549444471852347e-07, "loss": 0.0045, "step": 19390 }, { "grad_norm": 0.05306132510304451, "learning_rate": 2.4667439684974423e-07, "loss": 0.0046, "step": 19400 }, { "grad_norm": 0.12931303679943085, "learning_rate": 2.3854037020662467e-07, "loss": 0.005, "step": 19410 }, { "grad_norm": 0.07420578598976135, "learning_rate": 2.3054238949399288e-07, "loss": 0.0043, "step": 19420 }, { "grad_norm": 0.10231036692857742, "learning_rate": 2.2268047657802993e-07, "loss": 0.0037, "step": 19430 }, { "grad_norm": 0.23958081007003784, "learning_rate": 2.149546529529034e-07, "loss": 0.0045, "step": 19440 }, { "grad_norm": 0.048716939985752106, "learning_rate": 2.0736493974071736e-07, "loss": 0.0027, "step": 19450 }, { "grad_norm": 0.09505472332239151, "learning_rate": 1.9991135769145686e-07, "loss": 0.0042, "step": 19460 }, { "grad_norm": 0.231720432639122, "learning_rate": 1.9259392718293245e-07, "loss": 0.0039, "step": 19470 }, { "grad_norm": 0.06609362363815308, "learning_rate": 1.8541266822072467e-07, "loss": 0.0036, "step": 19480 }, { "grad_norm": 0.06773880869150162, "learning_rate": 1.7836760043811184e-07, "loss": 0.0029, "step": 19490 }, { "grad_norm": 0.0752679854631424, "learning_rate": 1.7145874309604792e-07, "loss": 0.0033, "step": 19500 }, { "grad_norm": 0.17139793932437897, "learning_rate": 1.6468611508308474e-07, "loss": 0.0048, "step": 19510 }, { "grad_norm": 0.15455612540245056, "learning_rate": 1.5804973491532204e-07, "loss": 0.004, "step": 19520 }, { "grad_norm": 0.06510385125875473, "learning_rate": 1.5154962073637424e-07, "loss": 0.0036, "step": 19530 }, { "grad_norm": 0.08411870151758194, "learning_rate": 1.4518579031730372e-07, "loss": 0.0035, "step": 19540 }, { "grad_norm": 0.054938413202762604, "learning_rate": 1.389582610565876e-07, "loss": 0.0037, "step": 19550 }, { "grad_norm": 0.07002374529838562, "learning_rate": 1.3286704998003995e-07, "loss": 0.0044, "step": 19560 }, { "grad_norm": 0.09384164214134216, "learning_rate": 1.2691217374080632e-07, "loss": 0.0043, "step": 19570 }, { "grad_norm": 0.11545012891292572, "learning_rate": 1.2109364861929705e-07, "loss": 0.0031, "step": 19580 }, { "grad_norm": 0.07397060096263885, "learning_rate": 1.1541149052312628e-07, "loss": 0.0037, "step": 19590 }, { "grad_norm": 0.13322952389717102, "learning_rate": 1.0986571498710074e-07, "loss": 0.0045, "step": 19600 }, { "grad_norm": 0.06104133278131485, "learning_rate": 1.0445633717316438e-07, "loss": 0.0035, "step": 19610 }, { "grad_norm": 0.05543619021773338, "learning_rate": 9.918337187034277e-08, "loss": 0.0034, "step": 19620 }, { "grad_norm": 0.09925747662782669, "learning_rate": 9.404683349472643e-08, "loss": 0.0055, "step": 19630 }, { "grad_norm": 0.09441298991441727, "learning_rate": 8.904673608940983e-08, "loss": 0.004, "step": 19640 }, { "grad_norm": 0.16413694620132446, "learning_rate": 8.418309332447471e-08, "loss": 0.0043, "step": 19650 }, { "grad_norm": 0.13242879509925842, "learning_rate": 7.945591849692902e-08, "loss": 0.0045, "step": 19660 }, { "grad_norm": 0.09196300804615021, "learning_rate": 7.486522453069578e-08, "loss": 0.0052, "step": 19670 }, { "grad_norm": 0.09193015843629837, "learning_rate": 7.041102397655208e-08, "loss": 0.0044, "step": 19680 }, { "grad_norm": 0.10367585718631744, "learning_rate": 6.609332901210685e-08, "loss": 0.0036, "step": 19690 }, { "grad_norm": 0.1656600683927536, "learning_rate": 6.191215144178419e-08, "loss": 0.0034, "step": 19700 }, { "grad_norm": 0.09132969379425049, "learning_rate": 5.786750269675678e-08, "loss": 0.0028, "step": 19710 }, { "grad_norm": 0.10719091445207596, "learning_rate": 5.395939383494031e-08, "loss": 0.0031, "step": 19720 }, { "grad_norm": 0.09140028804540634, "learning_rate": 5.018783554095463e-08, "loss": 0.0037, "step": 19730 }, { "grad_norm": 0.08317473530769348, "learning_rate": 4.655283812610156e-08, "loss": 0.0034, "step": 19740 }, { "grad_norm": 0.09623251855373383, "learning_rate": 4.305441152831491e-08, "loss": 0.0029, "step": 19750 }, { "grad_norm": 0.05224360153079033, "learning_rate": 3.9692565312171584e-08, "loss": 0.006, "step": 19760 }, { "grad_norm": 0.120881587266922, "learning_rate": 3.6467308668824975e-08, "loss": 0.0035, "step": 19770 }, { "grad_norm": 0.11368340998888016, "learning_rate": 3.3378650416004964e-08, "loss": 0.0031, "step": 19780 }, { "grad_norm": 0.13341814279556274, "learning_rate": 3.042659899797906e-08, "loss": 0.0042, "step": 19790 }, { "grad_norm": 0.07534883916378021, "learning_rate": 2.76111624855524e-08, "loss": 0.0048, "step": 19800 }, { "grad_norm": 0.13459919393062592, "learning_rate": 2.4932348576017784e-08, "loss": 0.0038, "step": 19810 }, { "grad_norm": 0.3475951850414276, "learning_rate": 2.239016459314458e-08, "loss": 0.0039, "step": 19820 }, { "grad_norm": 0.07803675532341003, "learning_rate": 1.9984617487173174e-08, "loss": 0.0037, "step": 19830 }, { "grad_norm": 0.12165316939353943, "learning_rate": 1.7715713834776105e-08, "loss": 0.0036, "step": 19840 }, { "grad_norm": 0.06452445685863495, "learning_rate": 1.5583459839046964e-08, "loss": 0.0041, "step": 19850 }, { "grad_norm": 0.1213124468922615, "learning_rate": 1.3587861329489304e-08, "loss": 0.0038, "step": 19860 }, { "grad_norm": 0.06894344091415405, "learning_rate": 1.1728923761994415e-08, "loss": 0.0037, "step": 19870 }, { "grad_norm": 0.11265594512224197, "learning_rate": 1.0006652218819135e-08, "loss": 0.0034, "step": 19880 }, { "grad_norm": 0.1181468814611435, "learning_rate": 8.421051408596947e-09, "loss": 0.0036, "step": 19890 }, { "grad_norm": 0.0781790241599083, "learning_rate": 6.972125666299123e-09, "loss": 0.0037, "step": 19900 }, { "grad_norm": 0.09450574219226837, "learning_rate": 5.659878953229169e-09, "loss": 0.0035, "step": 19910 }, { "grad_norm": 0.09104134887456894, "learning_rate": 4.48431485701728e-09, "loss": 0.0047, "step": 19920 }, { "grad_norm": 0.14649967849254608, "learning_rate": 3.4454365916203322e-09, "loss": 0.0047, "step": 19930 }, { "grad_norm": 0.07070212811231613, "learning_rate": 2.5432469972830332e-09, "loss": 0.0038, "step": 19940 }, { "grad_norm": 0.18107648193836212, "learning_rate": 1.7777485405601203e-09, "loss": 0.0039, "step": 19950 }, { "grad_norm": 0.13908658921718597, "learning_rate": 1.1489433142941597e-09, "loss": 0.0059, "step": 19960 }, { "grad_norm": 0.10341087728738785, "learning_rate": 6.568330376210963e-10, "loss": 0.0036, "step": 19970 }, { "grad_norm": 0.07556610554456711, "learning_rate": 3.0141905594249787e-10, "loss": 0.0045, "step": 19980 }, { "grad_norm": 0.08986358344554901, "learning_rate": 8.270234094776008e-11, "loss": 0.0026, "step": 19990 }, { "grad_norm": 0.1137193962931633, "learning_rate": 6.834906085551041e-13, "loss": 0.0043, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "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": 64, "trial_name": null, "trial_params": null }