{ "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": 0.6680845618247986, "learning_rate": 9e-07, "loss": 1.3768, "step": 10 }, { "grad_norm": 0.6591176986694336, "learning_rate": 1.9e-06, "loss": 1.3777, "step": 20 }, { "grad_norm": 0.6956794261932373, "learning_rate": 2.9e-06, "loss": 1.3582, "step": 30 }, { "grad_norm": 0.9620835185050964, "learning_rate": 3.9e-06, "loss": 1.2779, "step": 40 }, { "grad_norm": 0.6989454627037048, "learning_rate": 4.9000000000000005e-06, "loss": 1.2158, "step": 50 }, { "grad_norm": 0.6227983832359314, "learning_rate": 5.9e-06, "loss": 1.1852, "step": 60 }, { "grad_norm": 0.7214789986610413, "learning_rate": 6.900000000000001e-06, "loss": 1.1787, "step": 70 }, { "grad_norm": 0.547904908657074, "learning_rate": 7.9e-06, "loss": 1.1441, "step": 80 }, { "grad_norm": 0.6629834771156311, "learning_rate": 8.9e-06, "loss": 1.1361, "step": 90 }, { "grad_norm": 0.720084011554718, "learning_rate": 9.900000000000002e-06, "loss": 1.1318, "step": 100 }, { "grad_norm": 0.6548635959625244, "learning_rate": 1.09e-05, "loss": 1.1131, "step": 110 }, { "grad_norm": 0.6230436563491821, "learning_rate": 1.19e-05, "loss": 1.1051, "step": 120 }, { "grad_norm": 0.601715087890625, "learning_rate": 1.29e-05, "loss": 1.0957, "step": 130 }, { "grad_norm": 0.5525340437889099, "learning_rate": 1.3900000000000002e-05, "loss": 1.0869, "step": 140 }, { "grad_norm": 0.5316364169120789, "learning_rate": 1.49e-05, "loss": 1.0629, "step": 150 }, { "grad_norm": 0.7396922707557678, "learning_rate": 1.59e-05, "loss": 1.0543, "step": 160 }, { "grad_norm": 0.8624541759490967, "learning_rate": 1.69e-05, "loss": 1.0439, "step": 170 }, { "grad_norm": 0.9352287650108337, "learning_rate": 1.79e-05, "loss": 1.0332, "step": 180 }, { "grad_norm": 1.045367956161499, "learning_rate": 1.8900000000000002e-05, "loss": 1.0056, "step": 190 }, { "grad_norm": 1.0384280681610107, "learning_rate": 1.9900000000000003e-05, "loss": 0.9617, "step": 200 }, { "grad_norm": 1.1651086807250977, "learning_rate": 2.09e-05, "loss": 0.9231, "step": 210 }, { "grad_norm": 0.9465230107307434, "learning_rate": 2.19e-05, "loss": 0.8721, "step": 220 }, { "grad_norm": 1.1878538131713867, "learning_rate": 2.29e-05, "loss": 0.8217, "step": 230 }, { "grad_norm": 1.1953779458999634, "learning_rate": 2.39e-05, "loss": 0.7622, "step": 240 }, { "grad_norm": 1.2242176532745361, "learning_rate": 2.4900000000000002e-05, "loss": 0.7008, "step": 250 }, { "grad_norm": 1.484410285949707, "learning_rate": 2.5900000000000003e-05, "loss": 0.6456, "step": 260 }, { "grad_norm": 1.4062706232070923, "learning_rate": 2.6900000000000003e-05, "loss": 0.6219, "step": 270 }, { "grad_norm": 1.375077724456787, "learning_rate": 2.7900000000000004e-05, "loss": 0.5778, "step": 280 }, { "grad_norm": 1.5725032091140747, "learning_rate": 2.8899999999999998e-05, "loss": 0.5561, "step": 290 }, { "grad_norm": 1.6575546264648438, "learning_rate": 2.9900000000000002e-05, "loss": 0.5031, "step": 300 }, { "grad_norm": 1.644376277923584, "learning_rate": 3.09e-05, "loss": 0.4791, "step": 310 }, { "grad_norm": 1.8037251234054565, "learning_rate": 3.19e-05, "loss": 0.4438, "step": 320 }, { "grad_norm": 1.9494669437408447, "learning_rate": 3.29e-05, "loss": 0.4197, "step": 330 }, { "grad_norm": 1.9074978828430176, "learning_rate": 3.3900000000000004e-05, "loss": 0.3947, "step": 340 }, { "grad_norm": 1.701619029045105, "learning_rate": 3.49e-05, "loss": 0.3597, "step": 350 }, { "grad_norm": 2.000351667404175, "learning_rate": 3.59e-05, "loss": 0.3368, "step": 360 }, { "grad_norm": 1.503682255744934, "learning_rate": 3.69e-05, "loss": 0.3302, "step": 370 }, { "grad_norm": 1.5455348491668701, "learning_rate": 3.79e-05, "loss": 0.296, "step": 380 }, { "grad_norm": 1.5771034955978394, "learning_rate": 3.8900000000000004e-05, "loss": 0.2849, "step": 390 }, { "grad_norm": 1.8001818656921387, "learning_rate": 3.99e-05, "loss": 0.2538, "step": 400 }, { "grad_norm": 1.8554712533950806, "learning_rate": 4.09e-05, "loss": 0.2418, "step": 410 }, { "grad_norm": 1.5319920778274536, "learning_rate": 4.19e-05, "loss": 0.2187, "step": 420 }, { "grad_norm": 1.4891407489776611, "learning_rate": 4.29e-05, "loss": 0.2055, "step": 430 }, { "grad_norm": 1.2718569040298462, "learning_rate": 4.39e-05, "loss": 0.1904, "step": 440 }, { "grad_norm": 1.8918226957321167, "learning_rate": 4.49e-05, "loss": 0.1848, "step": 450 }, { "grad_norm": 1.9343944787979126, "learning_rate": 4.5900000000000004e-05, "loss": 0.1756, "step": 460 }, { "grad_norm": 1.4376416206359863, "learning_rate": 4.69e-05, "loss": 0.1758, "step": 470 }, { "grad_norm": 1.9403059482574463, "learning_rate": 4.79e-05, "loss": 0.1585, "step": 480 }, { "grad_norm": 1.4809025526046753, "learning_rate": 4.89e-05, "loss": 0.1381, "step": 490 }, { "grad_norm": 1.5962443351745605, "learning_rate": 4.99e-05, "loss": 0.1315, "step": 500 }, { "grad_norm": 1.3945517539978027, "learning_rate": 5.0900000000000004e-05, "loss": 0.135, "step": 510 }, { "grad_norm": 1.8225041627883911, "learning_rate": 5.19e-05, "loss": 0.1252, "step": 520 }, { "grad_norm": 1.3881994485855103, "learning_rate": 5.2900000000000005e-05, "loss": 0.1349, "step": 530 }, { "grad_norm": 1.6404107809066772, "learning_rate": 5.390000000000001e-05, "loss": 0.1226, "step": 540 }, { "grad_norm": 1.6567991971969604, "learning_rate": 5.4900000000000006e-05, "loss": 0.1221, "step": 550 }, { "grad_norm": 1.4174875020980835, "learning_rate": 5.590000000000001e-05, "loss": 0.1144, "step": 560 }, { "grad_norm": 1.37897527217865, "learning_rate": 5.69e-05, "loss": 0.1098, "step": 570 }, { "grad_norm": 1.2954742908477783, "learning_rate": 5.79e-05, "loss": 0.1134, "step": 580 }, { "grad_norm": 1.5248891115188599, "learning_rate": 5.89e-05, "loss": 0.1141, "step": 590 }, { "grad_norm": 1.3169476985931396, "learning_rate": 5.99e-05, "loss": 0.1208, "step": 600 }, { "grad_norm": 1.3748068809509277, "learning_rate": 6.09e-05, "loss": 0.1276, "step": 610 }, { "grad_norm": 1.6218098402023315, "learning_rate": 6.19e-05, "loss": 0.1102, "step": 620 }, { "grad_norm": 1.4759454727172852, "learning_rate": 6.29e-05, "loss": 0.1028, "step": 630 }, { "grad_norm": 1.3745266199111938, "learning_rate": 6.390000000000001e-05, "loss": 0.1063, "step": 640 }, { "grad_norm": 1.491376519203186, "learning_rate": 6.49e-05, "loss": 0.1084, "step": 650 }, { "grad_norm": 1.409429907798767, "learning_rate": 6.59e-05, "loss": 0.1019, "step": 660 }, { "grad_norm": 1.3525744676589966, "learning_rate": 6.690000000000001e-05, "loss": 0.1194, "step": 670 }, { "grad_norm": 1.5082148313522339, "learning_rate": 6.790000000000001e-05, "loss": 0.1067, "step": 680 }, { "grad_norm": 1.3496453762054443, "learning_rate": 6.89e-05, "loss": 0.1054, "step": 690 }, { "grad_norm": 1.1628518104553223, "learning_rate": 6.99e-05, "loss": 0.106, "step": 700 }, { "grad_norm": 1.176711082458496, "learning_rate": 7.09e-05, "loss": 0.1179, "step": 710 }, { "grad_norm": 1.0932114124298096, "learning_rate": 7.19e-05, "loss": 0.0981, "step": 720 }, { "grad_norm": 0.9976407289505005, "learning_rate": 7.29e-05, "loss": 0.1024, "step": 730 }, { "grad_norm": 1.138742446899414, "learning_rate": 7.390000000000001e-05, "loss": 0.0857, "step": 740 }, { "grad_norm": 1.1312651634216309, "learning_rate": 7.49e-05, "loss": 0.0979, "step": 750 }, { "grad_norm": 1.189576506614685, "learning_rate": 7.59e-05, "loss": 0.0888, "step": 760 }, { "grad_norm": 1.002389907836914, "learning_rate": 7.69e-05, "loss": 0.0927, "step": 770 }, { "grad_norm": 1.2123193740844727, "learning_rate": 7.790000000000001e-05, "loss": 0.0939, "step": 780 }, { "grad_norm": 1.3161468505859375, "learning_rate": 7.890000000000001e-05, "loss": 0.0902, "step": 790 }, { "grad_norm": 1.3190264701843262, "learning_rate": 7.99e-05, "loss": 0.0949, "step": 800 }, { "grad_norm": 1.041558861732483, "learning_rate": 8.090000000000001e-05, "loss": 0.0968, "step": 810 }, { "grad_norm": 1.058013677597046, "learning_rate": 8.19e-05, "loss": 0.0931, "step": 820 }, { "grad_norm": 1.310149073600769, "learning_rate": 8.29e-05, "loss": 0.0978, "step": 830 }, { "grad_norm": 1.0980340242385864, "learning_rate": 8.39e-05, "loss": 0.0887, "step": 840 }, { "grad_norm": 0.9291770458221436, "learning_rate": 8.49e-05, "loss": 0.101, "step": 850 }, { "grad_norm": 1.1781959533691406, "learning_rate": 8.59e-05, "loss": 0.0888, "step": 860 }, { "grad_norm": 1.2317875623703003, "learning_rate": 8.69e-05, "loss": 0.0931, "step": 870 }, { "grad_norm": 1.065956711769104, "learning_rate": 8.790000000000001e-05, "loss": 0.0859, "step": 880 }, { "grad_norm": 1.242351770401001, "learning_rate": 8.89e-05, "loss": 0.0888, "step": 890 }, { "grad_norm": 0.7749603390693665, "learning_rate": 8.99e-05, "loss": 0.0968, "step": 900 }, { "grad_norm": 1.1918364763259888, "learning_rate": 9.090000000000001e-05, "loss": 0.0833, "step": 910 }, { "grad_norm": 1.0837980508804321, "learning_rate": 9.190000000000001e-05, "loss": 0.0912, "step": 920 }, { "grad_norm": 1.1501940488815308, "learning_rate": 9.290000000000001e-05, "loss": 0.0828, "step": 930 }, { "grad_norm": 1.0251370668411255, "learning_rate": 9.39e-05, "loss": 0.0794, "step": 940 }, { "grad_norm": 0.9658505320549011, "learning_rate": 9.49e-05, "loss": 0.0909, "step": 950 }, { "grad_norm": 0.9038907885551453, "learning_rate": 9.59e-05, "loss": 0.0724, "step": 960 }, { "grad_norm": 1.13161039352417, "learning_rate": 9.69e-05, "loss": 0.0784, "step": 970 }, { "grad_norm": 1.0953474044799805, "learning_rate": 9.790000000000001e-05, "loss": 0.0847, "step": 980 }, { "grad_norm": 1.0212229490280151, "learning_rate": 9.89e-05, "loss": 0.0882, "step": 990 }, { "grad_norm": 1.0233827829360962, "learning_rate": 9.99e-05, "loss": 0.0817, "step": 1000 }, { "grad_norm": 1.0430985689163208, "learning_rate": 9.999994463727085e-05, "loss": 0.0808, "step": 1010 }, { "grad_norm": 0.8222044110298157, "learning_rate": 9.999975326009292e-05, "loss": 0.0798, "step": 1020 }, { "grad_norm": 0.980501651763916, "learning_rate": 9.999942518549879e-05, "loss": 0.0788, "step": 1030 }, { "grad_norm": 1.0080885887145996, "learning_rate": 9.999896041438544e-05, "loss": 0.0801, "step": 1040 }, { "grad_norm": 0.8249175548553467, "learning_rate": 9.999835894802353e-05, "loss": 0.0713, "step": 1050 }, { "grad_norm": 0.8810263872146606, "learning_rate": 9.999762078805743e-05, "loss": 0.0864, "step": 1060 }, { "grad_norm": 0.7997922897338867, "learning_rate": 9.999674593650526e-05, "loss": 0.0802, "step": 1070 }, { "grad_norm": 1.0426141023635864, "learning_rate": 9.99957343957588e-05, "loss": 0.0818, "step": 1080 }, { "grad_norm": 0.9908887147903442, "learning_rate": 9.99945861685836e-05, "loss": 0.0654, "step": 1090 }, { "grad_norm": 0.983393132686615, "learning_rate": 9.999330125811884e-05, "loss": 0.0745, "step": 1100 }, { "grad_norm": 1.1195029020309448, "learning_rate": 9.999187966787744e-05, "loss": 0.0807, "step": 1110 }, { "grad_norm": 0.7650071978569031, "learning_rate": 9.999032140174595e-05, "loss": 0.0776, "step": 1120 }, { "grad_norm": 0.7270625829696655, "learning_rate": 9.998862646398464e-05, "loss": 0.0834, "step": 1130 }, { "grad_norm": 1.1447160243988037, "learning_rate": 9.998679485922739e-05, "loss": 0.0794, "step": 1140 }, { "grad_norm": 0.7775225639343262, "learning_rate": 9.998482659248174e-05, "loss": 0.0763, "step": 1150 }, { "grad_norm": 0.9670314192771912, "learning_rate": 9.998272166912883e-05, "loss": 0.0747, "step": 1160 }, { "grad_norm": 0.9335726499557495, "learning_rate": 9.998048009492347e-05, "loss": 0.0701, "step": 1170 }, { "grad_norm": 0.8706818222999573, "learning_rate": 9.997810187599403e-05, "loss": 0.0714, "step": 1180 }, { "grad_norm": 0.8713046908378601, "learning_rate": 9.997558701884249e-05, "loss": 0.0725, "step": 1190 }, { "grad_norm": 0.6847192049026489, "learning_rate": 9.997293553034433e-05, "loss": 0.0729, "step": 1200 }, { "grad_norm": 0.8429000377655029, "learning_rate": 9.997014741774866e-05, "loss": 0.0696, "step": 1210 }, { "grad_norm": 0.7561994194984436, "learning_rate": 9.996722268867803e-05, "loss": 0.0647, "step": 1220 }, { "grad_norm": 0.7966417670249939, "learning_rate": 9.996416135112858e-05, "loss": 0.0666, "step": 1230 }, { "grad_norm": 0.8023107051849365, "learning_rate": 9.996096341346988e-05, "loss": 0.0623, "step": 1240 }, { "grad_norm": 0.7967690229415894, "learning_rate": 9.995762888444495e-05, "loss": 0.0678, "step": 1250 }, { "grad_norm": 0.8207305073738098, "learning_rate": 9.995415777317027e-05, "loss": 0.0627, "step": 1260 }, { "grad_norm": 0.69288569688797, "learning_rate": 9.995055008913574e-05, "loss": 0.0674, "step": 1270 }, { "grad_norm": 0.8063837885856628, "learning_rate": 9.994680584220463e-05, "loss": 0.0614, "step": 1280 }, { "grad_norm": 1.0185527801513672, "learning_rate": 9.994292504261355e-05, "loss": 0.0649, "step": 1290 }, { "grad_norm": 0.8515709638595581, "learning_rate": 9.993890770097247e-05, "loss": 0.0621, "step": 1300 }, { "grad_norm": 0.6488317251205444, "learning_rate": 9.993475382826467e-05, "loss": 0.059, "step": 1310 }, { "grad_norm": 0.8927388191223145, "learning_rate": 9.993046343584664e-05, "loss": 0.0649, "step": 1320 }, { "grad_norm": 0.6591321229934692, "learning_rate": 9.992603653544816e-05, "loss": 0.0567, "step": 1330 }, { "grad_norm": 0.7598251104354858, "learning_rate": 9.992147313917222e-05, "loss": 0.0587, "step": 1340 }, { "grad_norm": 0.8597753643989563, "learning_rate": 9.991677325949497e-05, "loss": 0.0522, "step": 1350 }, { "grad_norm": 0.6892462372779846, "learning_rate": 9.991193690926568e-05, "loss": 0.0554, "step": 1360 }, { "grad_norm": 0.6661257743835449, "learning_rate": 9.990696410170678e-05, "loss": 0.049, "step": 1370 }, { "grad_norm": 0.6577138304710388, "learning_rate": 9.990185485041371e-05, "loss": 0.0528, "step": 1380 }, { "grad_norm": 0.8383041024208069, "learning_rate": 9.989660916935498e-05, "loss": 0.0496, "step": 1390 }, { "grad_norm": 0.7992843389511108, "learning_rate": 9.989122707287208e-05, "loss": 0.0588, "step": 1400 }, { "grad_norm": 0.9479343295097351, "learning_rate": 9.988570857567945e-05, "loss": 0.0529, "step": 1410 }, { "grad_norm": 0.6767778992652893, "learning_rate": 9.988005369286446e-05, "loss": 0.0595, "step": 1420 }, { "grad_norm": 0.6775434613227844, "learning_rate": 9.987426243988734e-05, "loss": 0.0552, "step": 1430 }, { "grad_norm": 0.8384919762611389, "learning_rate": 9.986833483258114e-05, "loss": 0.0562, "step": 1440 }, { "grad_norm": 0.7093543410301208, "learning_rate": 9.986227088715173e-05, "loss": 0.0569, "step": 1450 }, { "grad_norm": 0.6243401765823364, "learning_rate": 9.98560706201777e-05, "loss": 0.0535, "step": 1460 }, { "grad_norm": 0.7216861844062805, "learning_rate": 9.984973404861036e-05, "loss": 0.0528, "step": 1470 }, { "grad_norm": 0.7486140131950378, "learning_rate": 9.984326118977361e-05, "loss": 0.0582, "step": 1480 }, { "grad_norm": 0.5842824578285217, "learning_rate": 9.983665206136406e-05, "loss": 0.048, "step": 1490 }, { "grad_norm": 0.6758549213409424, "learning_rate": 9.982990668145075e-05, "loss": 0.0614, "step": 1500 }, { "grad_norm": 0.7478466629981995, "learning_rate": 9.982302506847534e-05, "loss": 0.0513, "step": 1510 }, { "grad_norm": 0.6645094752311707, "learning_rate": 9.981600724125189e-05, "loss": 0.0513, "step": 1520 }, { "grad_norm": 0.8030089139938354, "learning_rate": 9.980885321896685e-05, "loss": 0.0531, "step": 1530 }, { "grad_norm": 0.8659258484840393, "learning_rate": 9.980156302117905e-05, "loss": 0.06, "step": 1540 }, { "grad_norm": 0.7310059666633606, "learning_rate": 9.979413666781963e-05, "loss": 0.0558, "step": 1550 }, { "grad_norm": 0.6846129894256592, "learning_rate": 9.978657417919193e-05, "loss": 0.0518, "step": 1560 }, { "grad_norm": 0.8213050365447998, "learning_rate": 9.977887557597153e-05, "loss": 0.0498, "step": 1570 }, { "grad_norm": 0.6869653463363647, "learning_rate": 9.97710408792061e-05, "loss": 0.0496, "step": 1580 }, { "grad_norm": 0.7208738923072815, "learning_rate": 9.976307011031542e-05, "loss": 0.052, "step": 1590 }, { "grad_norm": 0.7028236389160156, "learning_rate": 9.975496329109126e-05, "loss": 0.0507, "step": 1600 }, { "grad_norm": 0.5279938578605652, "learning_rate": 9.974672044369732e-05, "loss": 0.0448, "step": 1610 }, { "grad_norm": 0.6937764286994934, "learning_rate": 9.97383415906693e-05, "loss": 0.043, "step": 1620 }, { "grad_norm": 0.7590876221656799, "learning_rate": 9.97298267549146e-05, "loss": 0.0444, "step": 1630 }, { "grad_norm": 0.9302288293838501, "learning_rate": 9.972117595971249e-05, "loss": 0.0487, "step": 1640 }, { "grad_norm": 0.707331120967865, "learning_rate": 9.971238922871391e-05, "loss": 0.0535, "step": 1650 }, { "grad_norm": 0.5543134212493896, "learning_rate": 9.970346658594142e-05, "loss": 0.0403, "step": 1660 }, { "grad_norm": 0.6895405650138855, "learning_rate": 9.969440805578923e-05, "loss": 0.0474, "step": 1670 }, { "grad_norm": 0.5335018634796143, "learning_rate": 9.968521366302298e-05, "loss": 0.0499, "step": 1680 }, { "grad_norm": 0.7291369438171387, "learning_rate": 9.967588343277981e-05, "loss": 0.0517, "step": 1690 }, { "grad_norm": 0.7558074593544006, "learning_rate": 9.966641739056818e-05, "loss": 0.0515, "step": 1700 }, { "grad_norm": 0.5659112334251404, "learning_rate": 9.965681556226793e-05, "loss": 0.0484, "step": 1710 }, { "grad_norm": 0.6762385964393616, "learning_rate": 9.964707797413006e-05, "loss": 0.0508, "step": 1720 }, { "grad_norm": 0.7822985053062439, "learning_rate": 9.963720465277679e-05, "loss": 0.0462, "step": 1730 }, { "grad_norm": 0.7491819262504578, "learning_rate": 9.96271956252014e-05, "loss": 0.0427, "step": 1740 }, { "grad_norm": 0.8984391689300537, "learning_rate": 9.961705091876816e-05, "loss": 0.048, "step": 1750 }, { "grad_norm": 0.7463229894638062, "learning_rate": 9.960677056121235e-05, "loss": 0.0505, "step": 1760 }, { "grad_norm": 0.6638537049293518, "learning_rate": 9.959635458064005e-05, "loss": 0.0467, "step": 1770 }, { "grad_norm": 0.5115174055099487, "learning_rate": 9.958580300552815e-05, "loss": 0.0447, "step": 1780 }, { "grad_norm": 0.6805025935173035, "learning_rate": 9.957511586472426e-05, "loss": 0.0512, "step": 1790 }, { "grad_norm": 0.483161985874176, "learning_rate": 9.956429318744662e-05, "loss": 0.0451, "step": 1800 }, { "grad_norm": 0.6332815885543823, "learning_rate": 9.955333500328404e-05, "loss": 0.0441, "step": 1810 }, { "grad_norm": 0.7583509087562561, "learning_rate": 9.95422413421957e-05, "loss": 0.0412, "step": 1820 }, { "grad_norm": 0.5967929363250732, "learning_rate": 9.953101223451133e-05, "loss": 0.0414, "step": 1830 }, { "grad_norm": 0.5968555212020874, "learning_rate": 9.951964771093085e-05, "loss": 0.041, "step": 1840 }, { "grad_norm": 0.61930251121521, "learning_rate": 9.950814780252442e-05, "loss": 0.0416, "step": 1850 }, { "grad_norm": 0.6521567106246948, "learning_rate": 9.949651254073236e-05, "loss": 0.0466, "step": 1860 }, { "grad_norm": 0.6867579221725464, "learning_rate": 9.948474195736504e-05, "loss": 0.0481, "step": 1870 }, { "grad_norm": 0.6284758448600769, "learning_rate": 9.947283608460277e-05, "loss": 0.0399, "step": 1880 }, { "grad_norm": 0.6692090630531311, "learning_rate": 9.946079495499577e-05, "loss": 0.0421, "step": 1890 }, { "grad_norm": 0.6486384272575378, "learning_rate": 9.944861860146401e-05, "loss": 0.0518, "step": 1900 }, { "grad_norm": 0.5604891777038574, "learning_rate": 9.943630705729719e-05, "loss": 0.0465, "step": 1910 }, { "grad_norm": 0.5939029455184937, "learning_rate": 9.942386035615459e-05, "loss": 0.0409, "step": 1920 }, { "grad_norm": 0.5414155721664429, "learning_rate": 9.941127853206503e-05, "loss": 0.0434, "step": 1930 }, { "grad_norm": 0.7509816288948059, "learning_rate": 9.939856161942673e-05, "loss": 0.0439, "step": 1940 }, { "grad_norm": 0.656455397605896, "learning_rate": 9.938570965300724e-05, "loss": 0.0421, "step": 1950 }, { "grad_norm": 0.492107629776001, "learning_rate": 9.937272266794335e-05, "loss": 0.0347, "step": 1960 }, { "grad_norm": 0.5308566689491272, "learning_rate": 9.935960069974096e-05, "loss": 0.0408, "step": 1970 }, { "grad_norm": 0.8537633419036865, "learning_rate": 9.934634378427506e-05, "loss": 0.0482, "step": 1980 }, { "grad_norm": 0.7671170234680176, "learning_rate": 9.933295195778954e-05, "loss": 0.0377, "step": 1990 }, { "grad_norm": 0.6063556671142578, "learning_rate": 9.931942525689715e-05, "loss": 0.0346, "step": 2000 }, { "grad_norm": 0.6594319343566895, "learning_rate": 9.930576371857936e-05, "loss": 0.0339, "step": 2010 }, { "grad_norm": 0.6467959880828857, "learning_rate": 9.929196738018629e-05, "loss": 0.0394, "step": 2020 }, { "grad_norm": 0.5596520304679871, "learning_rate": 9.927803627943662e-05, "loss": 0.0403, "step": 2030 }, { "grad_norm": 0.6540936231613159, "learning_rate": 9.926397045441744e-05, "loss": 0.0431, "step": 2040 }, { "grad_norm": 0.6736294627189636, "learning_rate": 9.924976994358417e-05, "loss": 0.0392, "step": 2050 }, { "grad_norm": 0.564605176448822, "learning_rate": 9.923543478576048e-05, "loss": 0.0339, "step": 2060 }, { "grad_norm": 0.6072659492492676, "learning_rate": 9.922096502013813e-05, "loss": 0.0338, "step": 2070 }, { "grad_norm": 0.613389253616333, "learning_rate": 9.92063606862769e-05, "loss": 0.0305, "step": 2080 }, { "grad_norm": 0.685646116733551, "learning_rate": 9.919162182410453e-05, "loss": 0.0332, "step": 2090 }, { "grad_norm": 0.5550647974014282, "learning_rate": 9.917674847391645e-05, "loss": 0.033, "step": 2100 }, { "grad_norm": 0.5961691737174988, "learning_rate": 9.916174067637584e-05, "loss": 0.0338, "step": 2110 }, { "grad_norm": 0.5193253755569458, "learning_rate": 9.914659847251348e-05, "loss": 0.0321, "step": 2120 }, { "grad_norm": 0.601878821849823, "learning_rate": 9.913132190372753e-05, "loss": 0.0343, "step": 2130 }, { "grad_norm": 0.6159898042678833, "learning_rate": 9.911591101178359e-05, "loss": 0.0346, "step": 2140 }, { "grad_norm": 0.5584116578102112, "learning_rate": 9.910036583881443e-05, "loss": 0.0403, "step": 2150 }, { "grad_norm": 0.6862971186637878, "learning_rate": 9.908468642731995e-05, "loss": 0.0358, "step": 2160 }, { "grad_norm": 0.46330341696739197, "learning_rate": 9.906887282016707e-05, "loss": 0.0325, "step": 2170 }, { "grad_norm": 0.6166136264801025, "learning_rate": 9.90529250605896e-05, "loss": 0.0381, "step": 2180 }, { "grad_norm": 0.7882437109947205, "learning_rate": 9.903684319218809e-05, "loss": 0.0338, "step": 2190 }, { "grad_norm": 0.4674419164657593, "learning_rate": 9.902062725892976e-05, "loss": 0.0341, "step": 2200 }, { "grad_norm": 0.5291664004325867, "learning_rate": 9.900427730514834e-05, "loss": 0.0307, "step": 2210 }, { "grad_norm": 0.7311906218528748, "learning_rate": 9.8987793375544e-05, "loss": 0.0311, "step": 2220 }, { "grad_norm": 0.6623303294181824, "learning_rate": 9.897117551518318e-05, "loss": 0.036, "step": 2230 }, { "grad_norm": 0.7440094351768494, "learning_rate": 9.895442376949844e-05, "loss": 0.0313, "step": 2240 }, { "grad_norm": 0.5417069792747498, "learning_rate": 9.893753818428845e-05, "loss": 0.0381, "step": 2250 }, { "grad_norm": 0.7058647871017456, "learning_rate": 9.892051880571773e-05, "loss": 0.0353, "step": 2260 }, { "grad_norm": 0.4503210783004761, "learning_rate": 9.890336568031663e-05, "loss": 0.031, "step": 2270 }, { "grad_norm": 0.686693012714386, "learning_rate": 9.888607885498113e-05, "loss": 0.0376, "step": 2280 }, { "grad_norm": 0.5677979588508606, "learning_rate": 9.886865837697275e-05, "loss": 0.0339, "step": 2290 }, { "grad_norm": 0.5067443251609802, "learning_rate": 9.88511042939184e-05, "loss": 0.0291, "step": 2300 }, { "grad_norm": 0.4356783926486969, "learning_rate": 9.883341665381028e-05, "loss": 0.039, "step": 2310 }, { "grad_norm": 0.5999183654785156, "learning_rate": 9.881559550500575e-05, "loss": 0.0305, "step": 2320 }, { "grad_norm": 0.6807839274406433, "learning_rate": 9.879764089622712e-05, "loss": 0.0406, "step": 2330 }, { "grad_norm": 0.6369580030441284, "learning_rate": 9.87795528765616e-05, "loss": 0.0323, "step": 2340 }, { "grad_norm": 0.555894136428833, "learning_rate": 9.876133149546118e-05, "loss": 0.0381, "step": 2350 }, { "grad_norm": 0.6021659970283508, "learning_rate": 9.874297680274238e-05, "loss": 0.0329, "step": 2360 }, { "grad_norm": 0.6177252531051636, "learning_rate": 9.872448884858624e-05, "loss": 0.0331, "step": 2370 }, { "grad_norm": 0.6552705764770508, "learning_rate": 9.870586768353815e-05, "loss": 0.0304, "step": 2380 }, { "grad_norm": 0.5255332589149475, "learning_rate": 9.868711335850764e-05, "loss": 0.035, "step": 2390 }, { "grad_norm": 0.5219599008560181, "learning_rate": 9.866822592476833e-05, "loss": 0.0348, "step": 2400 }, { "grad_norm": 0.6282548308372498, "learning_rate": 9.86492054339577e-05, "loss": 0.0299, "step": 2410 }, { "grad_norm": 0.6167646050453186, "learning_rate": 9.863005193807711e-05, "loss": 0.0305, "step": 2420 }, { "grad_norm": 0.48692786693573, "learning_rate": 9.861076548949143e-05, "loss": 0.0285, "step": 2430 }, { "grad_norm": 0.7296221852302551, "learning_rate": 9.859134614092912e-05, "loss": 0.0334, "step": 2440 }, { "grad_norm": 0.5842213034629822, "learning_rate": 9.857179394548191e-05, "loss": 0.0315, "step": 2450 }, { "grad_norm": 0.5974501967430115, "learning_rate": 9.855210895660477e-05, "loss": 0.0357, "step": 2460 }, { "grad_norm": 0.6318084597587585, "learning_rate": 9.853229122811568e-05, "loss": 0.0332, "step": 2470 }, { "grad_norm": 0.4725353419780731, "learning_rate": 9.851234081419559e-05, "loss": 0.0315, "step": 2480 }, { "grad_norm": 0.6043561100959778, "learning_rate": 9.849225776938814e-05, "loss": 0.0259, "step": 2490 }, { "grad_norm": 0.5101690292358398, "learning_rate": 9.847204214859964e-05, "loss": 0.0333, "step": 2500 }, { "grad_norm": 0.5351490378379822, "learning_rate": 9.845169400709879e-05, "loss": 0.0309, "step": 2510 }, { "grad_norm": 0.613457202911377, "learning_rate": 9.843121340051664e-05, "loss": 0.0312, "step": 2520 }, { "grad_norm": 0.5415329337120056, "learning_rate": 9.841060038484641e-05, "loss": 0.0283, "step": 2530 }, { "grad_norm": 0.5314369201660156, "learning_rate": 9.838985501644328e-05, "loss": 0.032, "step": 2540 }, { "grad_norm": 0.6180905699729919, "learning_rate": 9.83689773520243e-05, "loss": 0.0367, "step": 2550 }, { "grad_norm": 0.6391300559043884, "learning_rate": 9.834796744866819e-05, "loss": 0.0326, "step": 2560 }, { "grad_norm": 0.604874849319458, "learning_rate": 9.832682536381525e-05, "loss": 0.0304, "step": 2570 }, { "grad_norm": 0.6227478981018066, "learning_rate": 9.830555115526711e-05, "loss": 0.0323, "step": 2580 }, { "grad_norm": 0.49382513761520386, "learning_rate": 9.828414488118667e-05, "loss": 0.0309, "step": 2590 }, { "grad_norm": 0.4658672511577606, "learning_rate": 9.826260660009785e-05, "loss": 0.0282, "step": 2600 }, { "grad_norm": 0.5284295678138733, "learning_rate": 9.824093637088547e-05, "loss": 0.0275, "step": 2610 }, { "grad_norm": 0.40948909521102905, "learning_rate": 9.821913425279514e-05, "loss": 0.0307, "step": 2620 }, { "grad_norm": 0.5642905831336975, "learning_rate": 9.8197200305433e-05, "loss": 0.0282, "step": 2630 }, { "grad_norm": 0.5310893058776855, "learning_rate": 9.817513458876564e-05, "loss": 0.027, "step": 2640 }, { "grad_norm": 0.6105229258537292, "learning_rate": 9.815293716311987e-05, "loss": 0.0322, "step": 2650 }, { "grad_norm": 0.548760712146759, "learning_rate": 9.813060808918262e-05, "loss": 0.0398, "step": 2660 }, { "grad_norm": 0.5801066756248474, "learning_rate": 9.810814742800069e-05, "loss": 0.0326, "step": 2670 }, { "grad_norm": 0.6136864423751831, "learning_rate": 9.808555524098074e-05, "loss": 0.0366, "step": 2680 }, { "grad_norm": 0.5054121017456055, "learning_rate": 9.806283158988887e-05, "loss": 0.0311, "step": 2690 }, { "grad_norm": 0.6326132416725159, "learning_rate": 9.803997653685072e-05, "loss": 0.0353, "step": 2700 }, { "grad_norm": 0.494823157787323, "learning_rate": 9.801699014435112e-05, "loss": 0.0336, "step": 2710 }, { "grad_norm": 0.5357293486595154, "learning_rate": 9.799387247523398e-05, "loss": 0.0307, "step": 2720 }, { "grad_norm": 0.577335000038147, "learning_rate": 9.797062359270215e-05, "loss": 0.0318, "step": 2730 }, { "grad_norm": 0.5836052298545837, "learning_rate": 9.794724356031715e-05, "loss": 0.0331, "step": 2740 }, { "grad_norm": 0.45126667618751526, "learning_rate": 9.792373244199913e-05, "loss": 0.0271, "step": 2750 }, { "grad_norm": 0.5190860033035278, "learning_rate": 9.790009030202658e-05, "loss": 0.0278, "step": 2760 }, { "grad_norm": 0.5671499371528625, "learning_rate": 9.78763172050362e-05, "loss": 0.0291, "step": 2770 }, { "grad_norm": 0.4337978661060333, "learning_rate": 9.785241321602274e-05, "loss": 0.0307, "step": 2780 }, { "grad_norm": 0.5242369174957275, "learning_rate": 9.782837840033879e-05, "loss": 0.0271, "step": 2790 }, { "grad_norm": 0.6008498072624207, "learning_rate": 9.780421282369461e-05, "loss": 0.0287, "step": 2800 }, { "grad_norm": 0.49873828887939453, "learning_rate": 9.777991655215797e-05, "loss": 0.0306, "step": 2810 }, { "grad_norm": 0.6602286100387573, "learning_rate": 9.775548965215394e-05, "loss": 0.0265, "step": 2820 }, { "grad_norm": 0.5901004076004028, "learning_rate": 9.773093219046474e-05, "loss": 0.0266, "step": 2830 }, { "grad_norm": 0.597578227519989, "learning_rate": 9.770624423422954e-05, "loss": 0.0279, "step": 2840 }, { "grad_norm": 0.502971351146698, "learning_rate": 9.768142585094426e-05, "loss": 0.0273, "step": 2850 }, { "grad_norm": 0.5371848940849304, "learning_rate": 9.765647710846142e-05, "loss": 0.0308, "step": 2860 }, { "grad_norm": 0.6622079014778137, "learning_rate": 9.763139807498991e-05, "loss": 0.0278, "step": 2870 }, { "grad_norm": 0.5037965774536133, "learning_rate": 9.760618881909487e-05, "loss": 0.0258, "step": 2880 }, { "grad_norm": 0.5630285143852234, "learning_rate": 9.758084940969744e-05, "loss": 0.0254, "step": 2890 }, { "grad_norm": 0.5627454519271851, "learning_rate": 9.755537991607459e-05, "loss": 0.0272, "step": 2900 }, { "grad_norm": 0.513208270072937, "learning_rate": 9.752978040785895e-05, "loss": 0.0366, "step": 2910 }, { "grad_norm": 0.41890227794647217, "learning_rate": 9.750405095503859e-05, "loss": 0.0264, "step": 2920 }, { "grad_norm": 0.5011561512947083, "learning_rate": 9.747819162795686e-05, "loss": 0.0253, "step": 2930 }, { "grad_norm": 0.5106188654899597, "learning_rate": 9.745220249731217e-05, "loss": 0.0252, "step": 2940 }, { "grad_norm": 0.40530043840408325, "learning_rate": 9.742608363415781e-05, "loss": 0.0232, "step": 2950 }, { "grad_norm": 0.5470148921012878, "learning_rate": 9.739983510990176e-05, "loss": 0.0307, "step": 2960 }, { "grad_norm": 0.49785226583480835, "learning_rate": 9.737345699630647e-05, "loss": 0.0299, "step": 2970 }, { "grad_norm": 0.446857213973999, "learning_rate": 9.734694936548869e-05, "loss": 0.0232, "step": 2980 }, { "grad_norm": 0.48182350397109985, "learning_rate": 9.732031228991932e-05, "loss": 0.0222, "step": 2990 }, { "grad_norm": 0.5331111550331116, "learning_rate": 9.729354584242302e-05, "loss": 0.0283, "step": 3000 }, { "grad_norm": 0.5684876441955566, "learning_rate": 9.726665009617832e-05, "loss": 0.0253, "step": 3010 }, { "grad_norm": 0.5371915102005005, "learning_rate": 9.723962512471714e-05, "loss": 0.0307, "step": 3020 }, { "grad_norm": 0.4998340606689453, "learning_rate": 9.72124710019247e-05, "loss": 0.0244, "step": 3030 }, { "grad_norm": 0.42797619104385376, "learning_rate": 9.718518780203934e-05, "loss": 0.0226, "step": 3040 }, { "grad_norm": 0.489924818277359, "learning_rate": 9.715777559965228e-05, "loss": 0.0276, "step": 3050 }, { "grad_norm": 0.400907427072525, "learning_rate": 9.713023446970746e-05, "loss": 0.0275, "step": 3060 }, { "grad_norm": 0.7238937020301819, "learning_rate": 9.710256448750126e-05, "loss": 0.0283, "step": 3070 }, { "grad_norm": 0.4171030521392822, "learning_rate": 9.707476572868235e-05, "loss": 0.0253, "step": 3080 }, { "grad_norm": 0.46262308955192566, "learning_rate": 9.704683826925149e-05, "loss": 0.0243, "step": 3090 }, { "grad_norm": 0.4571302533149719, "learning_rate": 9.701878218556129e-05, "loss": 0.0259, "step": 3100 }, { "grad_norm": 0.6798497438430786, "learning_rate": 9.699059755431598e-05, "loss": 0.031, "step": 3110 }, { "grad_norm": 0.3899567723274231, "learning_rate": 9.696228445257132e-05, "loss": 0.0253, "step": 3120 }, { "grad_norm": 0.5749596357345581, "learning_rate": 9.693384295773419e-05, "loss": 0.0284, "step": 3130 }, { "grad_norm": 0.6523962020874023, "learning_rate": 9.690527314756259e-05, "loss": 0.0278, "step": 3140 }, { "grad_norm": 0.555970311164856, "learning_rate": 9.687657510016527e-05, "loss": 0.0265, "step": 3150 }, { "grad_norm": 0.5093320608139038, "learning_rate": 9.684774889400161e-05, "loss": 0.0237, "step": 3160 }, { "grad_norm": 0.5063197612762451, "learning_rate": 9.681879460788135e-05, "loss": 0.0289, "step": 3170 }, { "grad_norm": 0.3915996551513672, "learning_rate": 9.67897123209644e-05, "loss": 0.0256, "step": 3180 }, { "grad_norm": 0.514234721660614, "learning_rate": 9.676050211276062e-05, "loss": 0.0252, "step": 3190 }, { "grad_norm": 0.5912958383560181, "learning_rate": 9.673116406312962e-05, "loss": 0.0288, "step": 3200 }, { "grad_norm": 0.6297309994697571, "learning_rate": 9.67016982522805e-05, "loss": 0.0278, "step": 3210 }, { "grad_norm": 0.5655257105827332, "learning_rate": 9.667210476077164e-05, "loss": 0.024, "step": 3220 }, { "grad_norm": 0.5640935301780701, "learning_rate": 9.664238366951055e-05, "loss": 0.0257, "step": 3230 }, { "grad_norm": 0.4491579532623291, "learning_rate": 9.661253505975355e-05, "loss": 0.0215, "step": 3240 }, { "grad_norm": 0.5641351342201233, "learning_rate": 9.658255901310557e-05, "loss": 0.0214, "step": 3250 }, { "grad_norm": 0.5178804993629456, "learning_rate": 9.655245561152e-05, "loss": 0.0325, "step": 3260 }, { "grad_norm": 0.5302542448043823, "learning_rate": 9.65222249372984e-05, "loss": 0.0266, "step": 3270 }, { "grad_norm": 0.345249205827713, "learning_rate": 9.649186707309026e-05, "loss": 0.0245, "step": 3280 }, { "grad_norm": 0.4994320571422577, "learning_rate": 9.646138210189283e-05, "loss": 0.0263, "step": 3290 }, { "grad_norm": 0.46781864762306213, "learning_rate": 9.643077010705087e-05, "loss": 0.0246, "step": 3300 }, { "grad_norm": 0.4072483479976654, "learning_rate": 9.640003117225637e-05, "loss": 0.0224, "step": 3310 }, { "grad_norm": 0.428013414144516, "learning_rate": 9.636916538154846e-05, "loss": 0.0221, "step": 3320 }, { "grad_norm": 0.46030426025390625, "learning_rate": 9.633817281931296e-05, "loss": 0.0246, "step": 3330 }, { "grad_norm": 0.5344061851501465, "learning_rate": 9.630705357028242e-05, "loss": 0.0227, "step": 3340 }, { "grad_norm": 0.45514267683029175, "learning_rate": 9.627580771953563e-05, "loss": 0.028, "step": 3350 }, { "grad_norm": 0.5697163939476013, "learning_rate": 9.624443535249759e-05, "loss": 0.0236, "step": 3360 }, { "grad_norm": 0.44105857610702515, "learning_rate": 9.621293655493913e-05, "loss": 0.0323, "step": 3370 }, { "grad_norm": 0.48319315910339355, "learning_rate": 9.618131141297675e-05, "loss": 0.0272, "step": 3380 }, { "grad_norm": 0.4819811284542084, "learning_rate": 9.614956001307242e-05, "loss": 0.0285, "step": 3390 }, { "grad_norm": 0.48486095666885376, "learning_rate": 9.611768244203321e-05, "loss": 0.0313, "step": 3400 }, { "grad_norm": 0.4497077167034149, "learning_rate": 9.60856787870112e-05, "loss": 0.0262, "step": 3410 }, { "grad_norm": 0.4462644159793854, "learning_rate": 9.605354913550318e-05, "loss": 0.026, "step": 3420 }, { "grad_norm": 0.4085383415222168, "learning_rate": 9.602129357535037e-05, "loss": 0.0221, "step": 3430 }, { "grad_norm": 0.45945093035697937, "learning_rate": 9.598891219473825e-05, "loss": 0.0246, "step": 3440 }, { "grad_norm": 0.4276796579360962, "learning_rate": 9.595640508219625e-05, "loss": 0.0202, "step": 3450 }, { "grad_norm": 0.45382067561149597, "learning_rate": 9.592377232659761e-05, "loss": 0.0261, "step": 3460 }, { "grad_norm": 0.5346838235855103, "learning_rate": 9.589101401715904e-05, "loss": 0.02, "step": 3470 }, { "grad_norm": 0.5592041611671448, "learning_rate": 9.585813024344045e-05, "loss": 0.0248, "step": 3480 }, { "grad_norm": 0.5535897612571716, "learning_rate": 9.58251210953449e-05, "loss": 0.0241, "step": 3490 }, { "grad_norm": 0.5072527527809143, "learning_rate": 9.579198666311809e-05, "loss": 0.0314, "step": 3500 }, { "grad_norm": 0.528148353099823, "learning_rate": 9.575872703734832e-05, "loss": 0.0248, "step": 3510 }, { "grad_norm": 0.4021073281764984, "learning_rate": 9.572534230896611e-05, "loss": 0.0226, "step": 3520 }, { "grad_norm": 0.5062975287437439, "learning_rate": 9.569183256924403e-05, "loss": 0.0224, "step": 3530 }, { "grad_norm": 0.49278247356414795, "learning_rate": 9.565819790979646e-05, "loss": 0.0278, "step": 3540 }, { "grad_norm": 0.5090126991271973, "learning_rate": 9.562443842257925e-05, "loss": 0.0242, "step": 3550 }, { "grad_norm": 0.40777304768562317, "learning_rate": 9.559055419988956e-05, "loss": 0.025, "step": 3560 }, { "grad_norm": 0.5558459162712097, "learning_rate": 9.555654533436557e-05, "loss": 0.0274, "step": 3570 }, { "grad_norm": 0.5879586935043335, "learning_rate": 9.552241191898621e-05, "loss": 0.0265, "step": 3580 }, { "grad_norm": 0.4956682324409485, "learning_rate": 9.548815404707092e-05, "loss": 0.0242, "step": 3590 }, { "grad_norm": 0.5432218909263611, "learning_rate": 9.545377181227942e-05, "loss": 0.0254, "step": 3600 }, { "grad_norm": 0.34619632363319397, "learning_rate": 9.541926530861145e-05, "loss": 0.0208, "step": 3610 }, { "grad_norm": 0.3456198275089264, "learning_rate": 9.538463463040645e-05, "loss": 0.0216, "step": 3620 }, { "grad_norm": 0.46307194232940674, "learning_rate": 9.534987987234337e-05, "loss": 0.0231, "step": 3630 }, { "grad_norm": 0.40911322832107544, "learning_rate": 9.53150011294404e-05, "loss": 0.0247, "step": 3640 }, { "grad_norm": 0.4260674715042114, "learning_rate": 9.527999849705471e-05, "loss": 0.0259, "step": 3650 }, { "grad_norm": 0.43744611740112305, "learning_rate": 9.524487207088213e-05, "loss": 0.0196, "step": 3660 }, { "grad_norm": 0.47888562083244324, "learning_rate": 9.520962194695698e-05, "loss": 0.0221, "step": 3670 }, { "grad_norm": 0.5920397639274597, "learning_rate": 9.517424822165175e-05, "loss": 0.0268, "step": 3680 }, { "grad_norm": 0.5179389119148254, "learning_rate": 9.513875099167685e-05, "loss": 0.0266, "step": 3690 }, { "grad_norm": 0.43874409794807434, "learning_rate": 9.510313035408035e-05, "loss": 0.0222, "step": 3700 }, { "grad_norm": 0.4436259865760803, "learning_rate": 9.506738640624775e-05, "loss": 0.0292, "step": 3710 }, { "grad_norm": 0.3324022591114044, "learning_rate": 9.50315192459016e-05, "loss": 0.0265, "step": 3720 }, { "grad_norm": 0.40740567445755005, "learning_rate": 9.499552897110136e-05, "loss": 0.0207, "step": 3730 }, { "grad_norm": 0.4901602268218994, "learning_rate": 9.495941568024304e-05, "loss": 0.0271, "step": 3740 }, { "grad_norm": 0.541377604007721, "learning_rate": 9.492317947205904e-05, "loss": 0.0241, "step": 3750 }, { "grad_norm": 0.6429196000099182, "learning_rate": 9.488682044561775e-05, "loss": 0.0234, "step": 3760 }, { "grad_norm": 0.4320107102394104, "learning_rate": 9.485033870032335e-05, "loss": 0.0283, "step": 3770 }, { "grad_norm": 0.5156699419021606, "learning_rate": 9.481373433591556e-05, "loss": 0.0232, "step": 3780 }, { "grad_norm": 0.4935128688812256, "learning_rate": 9.47770074524693e-05, "loss": 0.0215, "step": 3790 }, { "grad_norm": 0.4156445562839508, "learning_rate": 9.474015815039446e-05, "loss": 0.0248, "step": 3800 }, { "grad_norm": 0.5076112747192383, "learning_rate": 9.470318653043565e-05, "loss": 0.0194, "step": 3810 }, { "grad_norm": 0.4178719222545624, "learning_rate": 9.466609269367185e-05, "loss": 0.0254, "step": 3820 }, { "grad_norm": 0.4669092893600464, "learning_rate": 9.46288767415162e-05, "loss": 0.0268, "step": 3830 }, { "grad_norm": 0.3840748965740204, "learning_rate": 9.459153877571567e-05, "loss": 0.0212, "step": 3840 }, { "grad_norm": 0.4355137348175049, "learning_rate": 9.455407889835087e-05, "loss": 0.0218, "step": 3850 }, { "grad_norm": 0.5335889458656311, "learning_rate": 9.451649721183564e-05, "loss": 0.0244, "step": 3860 }, { "grad_norm": 0.47289928793907166, "learning_rate": 9.447879381891692e-05, "loss": 0.0231, "step": 3870 }, { "grad_norm": 0.5082388520240784, "learning_rate": 9.444096882267428e-05, "loss": 0.0223, "step": 3880 }, { "grad_norm": 0.3259719908237457, "learning_rate": 9.440302232651988e-05, "loss": 0.0212, "step": 3890 }, { "grad_norm": 0.4621064066886902, "learning_rate": 9.436495443419795e-05, "loss": 0.0237, "step": 3900 }, { "grad_norm": 0.4587567150592804, "learning_rate": 9.432676524978466e-05, "loss": 0.0231, "step": 3910 }, { "grad_norm": 0.518619179725647, "learning_rate": 9.42884548776878e-05, "loss": 0.0217, "step": 3920 }, { "grad_norm": 0.5491647124290466, "learning_rate": 9.425002342264646e-05, "loss": 0.0214, "step": 3930 }, { "grad_norm": 0.5315170288085938, "learning_rate": 9.421147098973077e-05, "loss": 0.0242, "step": 3940 }, { "grad_norm": 0.47827431559562683, "learning_rate": 9.41727976843416e-05, "loss": 0.0247, "step": 3950 }, { "grad_norm": 0.4667820334434509, "learning_rate": 9.413400361221029e-05, "loss": 0.0196, "step": 3960 }, { "grad_norm": 0.5719646215438843, "learning_rate": 9.409508887939835e-05, "loss": 0.0234, "step": 3970 }, { "grad_norm": 0.4673570394515991, "learning_rate": 9.40560535922972e-05, "loss": 0.0202, "step": 3980 }, { "grad_norm": 0.39664575457572937, "learning_rate": 9.40168978576278e-05, "loss": 0.0202, "step": 3990 }, { "grad_norm": 0.41226640343666077, "learning_rate": 9.397762178244043e-05, "loss": 0.023, "step": 4000 }, { "grad_norm": 0.3983164131641388, "learning_rate": 9.393822547411439e-05, "loss": 0.0208, "step": 4010 }, { "grad_norm": 0.4979904294013977, "learning_rate": 9.389870904035769e-05, "loss": 0.0193, "step": 4020 }, { "grad_norm": 0.39191460609436035, "learning_rate": 9.385907258920672e-05, "loss": 0.0236, "step": 4030 }, { "grad_norm": 0.35063233971595764, "learning_rate": 9.381931622902607e-05, "loss": 0.0237, "step": 4040 }, { "grad_norm": 0.392373263835907, "learning_rate": 9.377944006850807e-05, "loss": 0.021, "step": 4050 }, { "grad_norm": 0.5558865666389465, "learning_rate": 9.373944421667265e-05, "loss": 0.0277, "step": 4060 }, { "grad_norm": 0.5507460236549377, "learning_rate": 9.369932878286691e-05, "loss": 0.0251, "step": 4070 }, { "grad_norm": 0.48909345269203186, "learning_rate": 9.365909387676494e-05, "loss": 0.0211, "step": 4080 }, { "grad_norm": 0.5330439805984497, "learning_rate": 9.361873960836744e-05, "loss": 0.0251, "step": 4090 }, { "grad_norm": 0.45844435691833496, "learning_rate": 9.357826608800142e-05, "loss": 0.0281, "step": 4100 }, { "grad_norm": 0.4987274706363678, "learning_rate": 9.353767342631994e-05, "loss": 0.0298, "step": 4110 }, { "grad_norm": 0.3550601601600647, "learning_rate": 9.34969617343018e-05, "loss": 0.0276, "step": 4120 }, { "grad_norm": 0.40923011302948, "learning_rate": 9.345613112325122e-05, "loss": 0.027, "step": 4130 }, { "grad_norm": 0.5412095189094543, "learning_rate": 9.34151817047975e-05, "loss": 0.0296, "step": 4140 }, { "grad_norm": 0.4865875840187073, "learning_rate": 9.33741135908948e-05, "loss": 0.0271, "step": 4150 }, { "grad_norm": 0.3915463387966156, "learning_rate": 9.33329268938218e-05, "loss": 0.0287, "step": 4160 }, { "grad_norm": 0.5590789914131165, "learning_rate": 9.329162172618132e-05, "loss": 0.0325, "step": 4170 }, { "grad_norm": 0.48749634623527527, "learning_rate": 9.325019820090013e-05, "loss": 0.0307, "step": 4180 }, { "grad_norm": 0.4697338342666626, "learning_rate": 9.320865643122855e-05, "loss": 0.0303, "step": 4190 }, { "grad_norm": 0.5346032381057739, "learning_rate": 9.316699653074023e-05, "loss": 0.0261, "step": 4200 }, { "grad_norm": 0.45071813464164734, "learning_rate": 9.312521861333172e-05, "loss": 0.0264, "step": 4210 }, { "grad_norm": 0.5847572088241577, "learning_rate": 9.308332279322224e-05, "loss": 0.0273, "step": 4220 }, { "grad_norm": 0.5333300232887268, "learning_rate": 9.304130918495338e-05, "loss": 0.0243, "step": 4230 }, { "grad_norm": 0.5429439544677734, "learning_rate": 9.299917790338874e-05, "loss": 0.03, "step": 4240 }, { "grad_norm": 0.5822819471359253, "learning_rate": 9.295692906371363e-05, "loss": 0.0312, "step": 4250 }, { "grad_norm": 0.538055419921875, "learning_rate": 9.291456278143476e-05, "loss": 0.0316, "step": 4260 }, { "grad_norm": 0.5319395065307617, "learning_rate": 9.287207917237994e-05, "loss": 0.0273, "step": 4270 }, { "grad_norm": 0.5766119360923767, "learning_rate": 9.282947835269773e-05, "loss": 0.0242, "step": 4280 }, { "grad_norm": 0.5136768817901611, "learning_rate": 9.278676043885715e-05, "loss": 0.0245, "step": 4290 }, { "grad_norm": 0.4315184950828552, "learning_rate": 9.274392554764733e-05, "loss": 0.0256, "step": 4300 }, { "grad_norm": 0.596505343914032, "learning_rate": 9.270097379617723e-05, "loss": 0.0226, "step": 4310 }, { "grad_norm": 0.4706619679927826, "learning_rate": 9.26579053018753e-05, "loss": 0.0258, "step": 4320 }, { "grad_norm": 0.3769397735595703, "learning_rate": 9.261472018248918e-05, "loss": 0.0221, "step": 4330 }, { "grad_norm": 0.5097721815109253, "learning_rate": 9.25714185560853e-05, "loss": 0.0217, "step": 4340 }, { "grad_norm": 0.48693984746932983, "learning_rate": 9.252800054104868e-05, "loss": 0.0218, "step": 4350 }, { "grad_norm": 0.4414394795894623, "learning_rate": 9.248446625608252e-05, "loss": 0.0203, "step": 4360 }, { "grad_norm": 0.5023688673973083, "learning_rate": 9.244081582020789e-05, "loss": 0.0216, "step": 4370 }, { "grad_norm": 0.5579922795295715, "learning_rate": 9.239704935276339e-05, "loss": 0.0239, "step": 4380 }, { "grad_norm": 0.5292399525642395, "learning_rate": 9.235316697340489e-05, "loss": 0.0288, "step": 4390 }, { "grad_norm": 0.5150898098945618, "learning_rate": 9.230916880210512e-05, "loss": 0.0297, "step": 4400 }, { "grad_norm": 0.576615035533905, "learning_rate": 9.226505495915342e-05, "loss": 0.0239, "step": 4410 }, { "grad_norm": 0.5010238289833069, "learning_rate": 9.222082556515536e-05, "loss": 0.027, "step": 4420 }, { "grad_norm": 0.5346187353134155, "learning_rate": 9.217648074103242e-05, "loss": 0.0265, "step": 4430 }, { "grad_norm": 0.4911963939666748, "learning_rate": 9.213202060802161e-05, "loss": 0.03, "step": 4440 }, { "grad_norm": 0.46768343448638916, "learning_rate": 9.208744528767528e-05, "loss": 0.0208, "step": 4450 }, { "grad_norm": 0.4228092133998871, "learning_rate": 9.204275490186064e-05, "loss": 0.0232, "step": 4460 }, { "grad_norm": 0.4640195369720459, "learning_rate": 9.199794957275949e-05, "loss": 0.0211, "step": 4470 }, { "grad_norm": 0.5089555382728577, "learning_rate": 9.19530294228679e-05, "loss": 0.0226, "step": 4480 }, { "grad_norm": 0.5100897550582886, "learning_rate": 9.190799457499583e-05, "loss": 0.0245, "step": 4490 }, { "grad_norm": 0.39724409580230713, "learning_rate": 9.186284515226686e-05, "loss": 0.0227, "step": 4500 }, { "grad_norm": 0.5432016253471375, "learning_rate": 9.181758127811777e-05, "loss": 0.0278, "step": 4510 }, { "grad_norm": 0.40130162239074707, "learning_rate": 9.177220307629825e-05, "loss": 0.021, "step": 4520 }, { "grad_norm": 0.47915610671043396, "learning_rate": 9.172671067087059e-05, "loss": 0.0221, "step": 4530 }, { "grad_norm": 0.5432549715042114, "learning_rate": 9.16811041862093e-05, "loss": 0.0211, "step": 4540 }, { "grad_norm": 0.4278182089328766, "learning_rate": 9.163538374700076e-05, "loss": 0.0197, "step": 4550 }, { "grad_norm": 0.289884477853775, "learning_rate": 9.158954947824287e-05, "loss": 0.0185, "step": 4560 }, { "grad_norm": 0.4499456584453583, "learning_rate": 9.154360150524482e-05, "loss": 0.0222, "step": 4570 }, { "grad_norm": 0.38757213950157166, "learning_rate": 9.14975399536266e-05, "loss": 0.0217, "step": 4580 }, { "grad_norm": 0.3739558160305023, "learning_rate": 9.14513649493187e-05, "loss": 0.0189, "step": 4590 }, { "grad_norm": 0.4203878343105316, "learning_rate": 9.140507661856187e-05, "loss": 0.0195, "step": 4600 }, { "grad_norm": 0.4116365909576416, "learning_rate": 9.135867508790661e-05, "loss": 0.0199, "step": 4610 }, { "grad_norm": 0.4453210234642029, "learning_rate": 9.131216048421291e-05, "loss": 0.0239, "step": 4620 }, { "grad_norm": 0.4263717830181122, "learning_rate": 9.126553293464998e-05, "loss": 0.0213, "step": 4630 }, { "grad_norm": 0.5204765200614929, "learning_rate": 9.121879256669572e-05, "loss": 0.0228, "step": 4640 }, { "grad_norm": 0.43253791332244873, "learning_rate": 9.117193950813652e-05, "loss": 0.0229, "step": 4650 }, { "grad_norm": 0.4306985139846802, "learning_rate": 9.112497388706685e-05, "loss": 0.0303, "step": 4660 }, { "grad_norm": 0.33264538645744324, "learning_rate": 9.10778958318889e-05, "loss": 0.0227, "step": 4670 }, { "grad_norm": 0.3860510289669037, "learning_rate": 9.103070547131232e-05, "loss": 0.0176, "step": 4680 }, { "grad_norm": 0.3967917263507843, "learning_rate": 9.098340293435375e-05, "loss": 0.0239, "step": 4690 }, { "grad_norm": 0.4579527974128723, "learning_rate": 9.093598835033649e-05, "loss": 0.027, "step": 4700 }, { "grad_norm": 0.5328963994979858, "learning_rate": 9.088846184889021e-05, "loss": 0.0195, "step": 4710 }, { "grad_norm": 0.3771918714046478, "learning_rate": 9.084082355995057e-05, "loss": 0.0224, "step": 4720 }, { "grad_norm": 0.4708661735057831, "learning_rate": 9.079307361375882e-05, "loss": 0.0218, "step": 4730 }, { "grad_norm": 0.42308884859085083, "learning_rate": 9.074521214086149e-05, "loss": 0.0201, "step": 4740 }, { "grad_norm": 0.43889549374580383, "learning_rate": 9.069723927211001e-05, "loss": 0.0206, "step": 4750 }, { "grad_norm": 0.49833813309669495, "learning_rate": 9.064915513866037e-05, "loss": 0.0214, "step": 4760 }, { "grad_norm": 0.6608289480209351, "learning_rate": 9.060095987197279e-05, "loss": 0.0222, "step": 4770 }, { "grad_norm": 0.39463120698928833, "learning_rate": 9.055265360381126e-05, "loss": 0.026, "step": 4780 }, { "grad_norm": 0.4461647868156433, "learning_rate": 9.050423646624326e-05, "loss": 0.0222, "step": 4790 }, { "grad_norm": 0.4858885109424591, "learning_rate": 9.045570859163943e-05, "loss": 0.0233, "step": 4800 }, { "grad_norm": 0.43125656247138977, "learning_rate": 9.04070701126731e-05, "loss": 0.0189, "step": 4810 }, { "grad_norm": 0.4540945291519165, "learning_rate": 9.035832116232001e-05, "loss": 0.0184, "step": 4820 }, { "grad_norm": 0.41744503378868103, "learning_rate": 9.030946187385796e-05, "loss": 0.0243, "step": 4830 }, { "grad_norm": 0.39860087633132935, "learning_rate": 9.026049238086635e-05, "loss": 0.029, "step": 4840 }, { "grad_norm": 0.35753244161605835, "learning_rate": 9.021141281722591e-05, "loss": 0.0214, "step": 4850 }, { "grad_norm": 0.34049132466316223, "learning_rate": 9.01622233171183e-05, "loss": 0.0198, "step": 4860 }, { "grad_norm": 0.34213677048683167, "learning_rate": 9.011292401502574e-05, "loss": 0.02, "step": 4870 }, { "grad_norm": 0.31533804535865784, "learning_rate": 9.006351504573063e-05, "loss": 0.025, "step": 4880 }, { "grad_norm": 0.46508583426475525, "learning_rate": 9.001399654431519e-05, "loss": 0.0219, "step": 4890 }, { "grad_norm": 0.4160831570625305, "learning_rate": 8.996436864616116e-05, "loss": 0.0211, "step": 4900 }, { "grad_norm": 0.5274258255958557, "learning_rate": 8.991463148694925e-05, "loss": 0.0202, "step": 4910 }, { "grad_norm": 0.4051196575164795, "learning_rate": 8.986478520265902e-05, "loss": 0.0179, "step": 4920 }, { "grad_norm": 0.4379321336746216, "learning_rate": 8.981482992956827e-05, "loss": 0.0213, "step": 4930 }, { "grad_norm": 0.5809224247932434, "learning_rate": 8.976476580425282e-05, "loss": 0.018, "step": 4940 }, { "grad_norm": 0.47466161847114563, "learning_rate": 8.971459296358606e-05, "loss": 0.0201, "step": 4950 }, { "grad_norm": 0.48958829045295715, "learning_rate": 8.966431154473864e-05, "loss": 0.0177, "step": 4960 }, { "grad_norm": 0.42177024483680725, "learning_rate": 8.961392168517803e-05, "loss": 0.0255, "step": 4970 }, { "grad_norm": 0.5376948714256287, "learning_rate": 8.956342352266821e-05, "loss": 0.0241, "step": 4980 }, { "grad_norm": 0.34932762384414673, "learning_rate": 8.95128171952692e-05, "loss": 0.0188, "step": 4990 }, { "grad_norm": 0.5328025817871094, "learning_rate": 8.946210284133676e-05, "loss": 0.0203, "step": 5000 }, { "grad_norm": 0.4846362769603729, "learning_rate": 8.941128059952201e-05, "loss": 0.0206, "step": 5010 }, { "grad_norm": 0.4178476333618164, "learning_rate": 8.936035060877102e-05, "loss": 0.0206, "step": 5020 }, { "grad_norm": 0.42340797185897827, "learning_rate": 8.930931300832443e-05, "loss": 0.0186, "step": 5030 }, { "grad_norm": 0.3460429608821869, "learning_rate": 8.925816793771711e-05, "loss": 0.0175, "step": 5040 }, { "grad_norm": 0.46100160479545593, "learning_rate": 8.92069155367777e-05, "loss": 0.0196, "step": 5050 }, { "grad_norm": 0.39093226194381714, "learning_rate": 8.915555594562834e-05, "loss": 0.0197, "step": 5060 }, { "grad_norm": 0.408971905708313, "learning_rate": 8.910408930468416e-05, "loss": 0.0156, "step": 5070 }, { "grad_norm": 0.33198070526123047, "learning_rate": 8.905251575465303e-05, "loss": 0.02, "step": 5080 }, { "grad_norm": 0.5037837028503418, "learning_rate": 8.900083543653502e-05, "loss": 0.0206, "step": 5090 }, { "grad_norm": 0.4338358938694, "learning_rate": 8.894904849162218e-05, "loss": 0.0223, "step": 5100 }, { "grad_norm": 0.45139795541763306, "learning_rate": 8.889715506149802e-05, "loss": 0.0188, "step": 5110 }, { "grad_norm": 0.33429694175720215, "learning_rate": 8.884515528803722e-05, "loss": 0.0257, "step": 5120 }, { "grad_norm": 0.42427387833595276, "learning_rate": 8.879304931340517e-05, "loss": 0.0171, "step": 5130 }, { "grad_norm": 0.37521177530288696, "learning_rate": 8.874083728005759e-05, "loss": 0.0187, "step": 5140 }, { "grad_norm": 0.43861252069473267, "learning_rate": 8.868851933074021e-05, "loss": 0.0174, "step": 5150 }, { "grad_norm": 0.45748278498649597, "learning_rate": 8.863609560848829e-05, "loss": 0.0194, "step": 5160 }, { "grad_norm": 0.5359494686126709, "learning_rate": 8.85835662566263e-05, "loss": 0.0145, "step": 5170 }, { "grad_norm": 0.35847169160842896, "learning_rate": 8.853093141876747e-05, "loss": 0.0226, "step": 5180 }, { "grad_norm": 0.47711053490638733, "learning_rate": 8.847819123881343e-05, "loss": 0.0236, "step": 5190 }, { "grad_norm": 0.4167478382587433, "learning_rate": 8.842534586095383e-05, "loss": 0.0255, "step": 5200 }, { "grad_norm": 0.5768662095069885, "learning_rate": 8.837239542966593e-05, "loss": 0.0216, "step": 5210 }, { "grad_norm": 0.36879026889801025, "learning_rate": 8.831934008971417e-05, "loss": 0.0208, "step": 5220 }, { "grad_norm": 0.3993513286113739, "learning_rate": 8.826617998614982e-05, "loss": 0.0204, "step": 5230 }, { "grad_norm": 0.7548524141311646, "learning_rate": 8.821291526431056e-05, "loss": 0.0188, "step": 5240 }, { "grad_norm": 0.4723445177078247, "learning_rate": 8.815954606982015e-05, "loss": 0.0194, "step": 5250 }, { "grad_norm": 0.4291101098060608, "learning_rate": 8.810607254858789e-05, "loss": 0.0248, "step": 5260 }, { "grad_norm": 0.44897890090942383, "learning_rate": 8.805249484680838e-05, "loss": 0.0225, "step": 5270 }, { "grad_norm": 0.44614699482917786, "learning_rate": 8.799881311096096e-05, "loss": 0.0213, "step": 5280 }, { "grad_norm": 0.34205174446105957, "learning_rate": 8.794502748780949e-05, "loss": 0.0212, "step": 5290 }, { "grad_norm": 0.39593151211738586, "learning_rate": 8.78911381244018e-05, "loss": 0.0209, "step": 5300 }, { "grad_norm": 0.36930403113365173, "learning_rate": 8.783714516806933e-05, "loss": 0.019, "step": 5310 }, { "grad_norm": 0.5041482448577881, "learning_rate": 8.77830487664268e-05, "loss": 0.0222, "step": 5320 }, { "grad_norm": 0.3954504430294037, "learning_rate": 8.772884906737167e-05, "loss": 0.0181, "step": 5330 }, { "grad_norm": 0.3742181956768036, "learning_rate": 8.767454621908387e-05, "loss": 0.0227, "step": 5340 }, { "grad_norm": 0.4373776316642761, "learning_rate": 8.76201403700253e-05, "loss": 0.0205, "step": 5350 }, { "grad_norm": 0.4728417694568634, "learning_rate": 8.756563166893949e-05, "loss": 0.0223, "step": 5360 }, { "grad_norm": 0.49849778413772583, "learning_rate": 8.751102026485113e-05, "loss": 0.0235, "step": 5370 }, { "grad_norm": 0.45782575011253357, "learning_rate": 8.745630630706571e-05, "loss": 0.0211, "step": 5380 }, { "grad_norm": 0.4546658396720886, "learning_rate": 8.740148994516912e-05, "loss": 0.0191, "step": 5390 }, { "grad_norm": 0.4528074264526367, "learning_rate": 8.73465713290272e-05, "loss": 0.0207, "step": 5400 }, { "grad_norm": 0.5178644061088562, "learning_rate": 8.729155060878533e-05, "loss": 0.0211, "step": 5410 }, { "grad_norm": 0.4988405108451843, "learning_rate": 8.723642793486809e-05, "loss": 0.0184, "step": 5420 }, { "grad_norm": 0.4913215935230255, "learning_rate": 8.718120345797873e-05, "loss": 0.0208, "step": 5430 }, { "grad_norm": 0.3928888142108917, "learning_rate": 8.712587732909889e-05, "loss": 0.0218, "step": 5440 }, { "grad_norm": 0.5482167601585388, "learning_rate": 8.707044969948806e-05, "loss": 0.0247, "step": 5450 }, { "grad_norm": 0.5294741988182068, "learning_rate": 8.701492072068329e-05, "loss": 0.0224, "step": 5460 }, { "grad_norm": 0.3789893686771393, "learning_rate": 8.695929054449869e-05, "loss": 0.0229, "step": 5470 }, { "grad_norm": 0.3447975218296051, "learning_rate": 8.690355932302501e-05, "loss": 0.0218, "step": 5480 }, { "grad_norm": 0.32427018880844116, "learning_rate": 8.684772720862931e-05, "loss": 0.0168, "step": 5490 }, { "grad_norm": 0.3327474892139435, "learning_rate": 8.679179435395446e-05, "loss": 0.0173, "step": 5500 }, { "grad_norm": 0.3477976620197296, "learning_rate": 8.673576091191874e-05, "loss": 0.0193, "step": 5510 }, { "grad_norm": 0.3345457911491394, "learning_rate": 8.667962703571541e-05, "loss": 0.0153, "step": 5520 }, { "grad_norm": 0.36662566661834717, "learning_rate": 8.662339287881238e-05, "loss": 0.0234, "step": 5530 }, { "grad_norm": 0.28898152709007263, "learning_rate": 8.656705859495169e-05, "loss": 0.0165, "step": 5540 }, { "grad_norm": 0.36867937445640564, "learning_rate": 8.651062433814912e-05, "loss": 0.0155, "step": 5550 }, { "grad_norm": 0.5341900587081909, "learning_rate": 8.645409026269375e-05, "loss": 0.0166, "step": 5560 }, { "grad_norm": 0.4450433850288391, "learning_rate": 8.639745652314759e-05, "loss": 0.0156, "step": 5570 }, { "grad_norm": 0.4063837230205536, "learning_rate": 8.634072327434515e-05, "loss": 0.02, "step": 5580 }, { "grad_norm": 0.3526541292667389, "learning_rate": 8.628389067139294e-05, "loss": 0.0184, "step": 5590 }, { "grad_norm": 0.49016597867012024, "learning_rate": 8.622695886966911e-05, "loss": 0.0217, "step": 5600 }, { "grad_norm": 0.5720815062522888, "learning_rate": 8.616992802482308e-05, "loss": 0.0219, "step": 5610 }, { "grad_norm": 0.4507187604904175, "learning_rate": 8.611279829277496e-05, "loss": 0.02, "step": 5620 }, { "grad_norm": 0.43047383427619934, "learning_rate": 8.605556982971528e-05, "loss": 0.0172, "step": 5630 }, { "grad_norm": 0.3531322479248047, "learning_rate": 8.599824279210447e-05, "loss": 0.0227, "step": 5640 }, { "grad_norm": 0.3927733898162842, "learning_rate": 8.594081733667243e-05, "loss": 0.0144, "step": 5650 }, { "grad_norm": 0.39618992805480957, "learning_rate": 8.58832936204182e-05, "loss": 0.0186, "step": 5660 }, { "grad_norm": 0.37251290678977966, "learning_rate": 8.582567180060942e-05, "loss": 0.0182, "step": 5670 }, { "grad_norm": 0.34259548783302307, "learning_rate": 8.576795203478194e-05, "loss": 0.018, "step": 5680 }, { "grad_norm": 0.36769115924835205, "learning_rate": 8.571013448073939e-05, "loss": 0.0215, "step": 5690 }, { "grad_norm": 0.38593974709510803, "learning_rate": 8.565221929655275e-05, "loss": 0.0146, "step": 5700 }, { "grad_norm": 0.4075409173965454, "learning_rate": 8.559420664055992e-05, "loss": 0.0204, "step": 5710 }, { "grad_norm": 0.3751410245895386, "learning_rate": 8.553609667136532e-05, "loss": 0.0215, "step": 5720 }, { "grad_norm": 0.3994200825691223, "learning_rate": 8.547788954783936e-05, "loss": 0.0221, "step": 5730 }, { "grad_norm": 0.3578428030014038, "learning_rate": 8.541958542911808e-05, "loss": 0.0185, "step": 5740 }, { "grad_norm": 0.4283462464809418, "learning_rate": 8.536118447460275e-05, "loss": 0.0183, "step": 5750 }, { "grad_norm": 0.4787594974040985, "learning_rate": 8.530268684395932e-05, "loss": 0.0207, "step": 5760 }, { "grad_norm": 0.36684277653694153, "learning_rate": 8.524409269711807e-05, "loss": 0.0208, "step": 5770 }, { "grad_norm": 0.3540728688240051, "learning_rate": 8.51854021942732e-05, "loss": 0.0212, "step": 5780 }, { "grad_norm": 0.2948935329914093, "learning_rate": 8.512661549588227e-05, "loss": 0.0183, "step": 5790 }, { "grad_norm": 0.41187894344329834, "learning_rate": 8.506773276266588e-05, "loss": 0.016, "step": 5800 }, { "grad_norm": 0.352934867143631, "learning_rate": 8.500875415560721e-05, "loss": 0.0153, "step": 5810 }, { "grad_norm": 0.38501155376434326, "learning_rate": 8.494967983595144e-05, "loss": 0.0161, "step": 5820 }, { "grad_norm": 0.51210618019104, "learning_rate": 8.489050996520558e-05, "loss": 0.021, "step": 5830 }, { "grad_norm": 0.4871633052825928, "learning_rate": 8.483124470513775e-05, "loss": 0.0183, "step": 5840 }, { "grad_norm": 0.3729078769683838, "learning_rate": 8.477188421777692e-05, "loss": 0.0251, "step": 5850 }, { "grad_norm": 0.3691781461238861, "learning_rate": 8.47124286654124e-05, "loss": 0.0175, "step": 5860 }, { "grad_norm": 0.36910364031791687, "learning_rate": 8.465287821059341e-05, "loss": 0.0167, "step": 5870 }, { "grad_norm": 0.4670969545841217, "learning_rate": 8.45932330161286e-05, "loss": 0.0253, "step": 5880 }, { "grad_norm": 0.45545393228530884, "learning_rate": 8.453349324508567e-05, "loss": 0.0159, "step": 5890 }, { "grad_norm": 0.37817850708961487, "learning_rate": 8.447365906079088e-05, "loss": 0.0202, "step": 5900 }, { "grad_norm": 0.3597967326641083, "learning_rate": 8.441373062682856e-05, "loss": 0.0195, "step": 5910 }, { "grad_norm": 0.3996320962905884, "learning_rate": 8.43537081070408e-05, "loss": 0.0184, "step": 5920 }, { "grad_norm": 0.33463937044143677, "learning_rate": 8.429359166552689e-05, "loss": 0.0162, "step": 5930 }, { "grad_norm": 0.31746619939804077, "learning_rate": 8.423338146664284e-05, "loss": 0.0169, "step": 5940 }, { "grad_norm": 0.3840077221393585, "learning_rate": 8.417307767500107e-05, "loss": 0.0152, "step": 5950 }, { "grad_norm": 0.4162236452102661, "learning_rate": 8.411268045546983e-05, "loss": 0.0153, "step": 5960 }, { "grad_norm": 0.3609810471534729, "learning_rate": 8.405218997317281e-05, "loss": 0.015, "step": 5970 }, { "grad_norm": 0.43075743317604065, "learning_rate": 8.399160639348869e-05, "loss": 0.018, "step": 5980 }, { "grad_norm": 0.3736489415168762, "learning_rate": 8.393092988205065e-05, "loss": 0.0176, "step": 5990 }, { "grad_norm": 0.3479081094264984, "learning_rate": 8.387016060474597e-05, "loss": 0.0174, "step": 6000 }, { "grad_norm": 0.3594217598438263, "learning_rate": 8.380929872771551e-05, "loss": 0.0146, "step": 6010 }, { "grad_norm": 0.5351815223693848, "learning_rate": 8.374834441735335e-05, "loss": 0.0165, "step": 6020 }, { "grad_norm": 0.3614652454853058, "learning_rate": 8.368729784030622e-05, "loss": 0.0143, "step": 6030 }, { "grad_norm": 0.4107961356639862, "learning_rate": 8.362615916347315e-05, "loss": 0.0195, "step": 6040 }, { "grad_norm": 0.4212474226951599, "learning_rate": 8.356492855400493e-05, "loss": 0.014, "step": 6050 }, { "grad_norm": 0.3367299735546112, "learning_rate": 8.350360617930371e-05, "loss": 0.0167, "step": 6060 }, { "grad_norm": 0.48478373885154724, "learning_rate": 8.344219220702255e-05, "loss": 0.0156, "step": 6070 }, { "grad_norm": 0.374034583568573, "learning_rate": 8.338068680506485e-05, "loss": 0.019, "step": 6080 }, { "grad_norm": 0.27915093302726746, "learning_rate": 8.33190901415841e-05, "loss": 0.0167, "step": 6090 }, { "grad_norm": 0.3451411724090576, "learning_rate": 8.325740238498317e-05, "loss": 0.0141, "step": 6100 }, { "grad_norm": 0.35112038254737854, "learning_rate": 8.319562370391406e-05, "loss": 0.0145, "step": 6110 }, { "grad_norm": 0.5322951674461365, "learning_rate": 8.31337542672773e-05, "loss": 0.0184, "step": 6120 }, { "grad_norm": 0.5702270269393921, "learning_rate": 8.307179424422158e-05, "loss": 0.0179, "step": 6130 }, { "grad_norm": 0.3600696623325348, "learning_rate": 8.300974380414327e-05, "loss": 0.0154, "step": 6140 }, { "grad_norm": 0.30546438694000244, "learning_rate": 8.294760311668586e-05, "loss": 0.0154, "step": 6150 }, { "grad_norm": 0.30942267179489136, "learning_rate": 8.288537235173961e-05, "loss": 0.0191, "step": 6160 }, { "grad_norm": 0.4043451249599457, "learning_rate": 8.282305167944108e-05, "loss": 0.0182, "step": 6170 }, { "grad_norm": 0.42044222354888916, "learning_rate": 8.276064127017262e-05, "loss": 0.0181, "step": 6180 }, { "grad_norm": 0.3470224142074585, "learning_rate": 8.269814129456189e-05, "loss": 0.0199, "step": 6190 }, { "grad_norm": 0.4206221401691437, "learning_rate": 8.263555192348143e-05, "loss": 0.017, "step": 6200 }, { "grad_norm": 0.34851309657096863, "learning_rate": 8.257287332804819e-05, "loss": 0.0161, "step": 6210 }, { "grad_norm": 0.3413292169570923, "learning_rate": 8.251010567962307e-05, "loss": 0.0156, "step": 6220 }, { "grad_norm": 0.36740970611572266, "learning_rate": 8.244724914981041e-05, "loss": 0.0158, "step": 6230 }, { "grad_norm": 0.3461947739124298, "learning_rate": 8.238430391045757e-05, "loss": 0.0178, "step": 6240 }, { "grad_norm": 0.28556302189826965, "learning_rate": 8.232127013365445e-05, "loss": 0.0194, "step": 6250 }, { "grad_norm": 0.3151070475578308, "learning_rate": 8.225814799173295e-05, "loss": 0.0198, "step": 6260 }, { "grad_norm": 0.49902215600013733, "learning_rate": 8.219493765726663e-05, "loss": 0.0164, "step": 6270 }, { "grad_norm": 0.3842081129550934, "learning_rate": 8.21316393030701e-05, "loss": 0.0177, "step": 6280 }, { "grad_norm": 0.569420576095581, "learning_rate": 8.206825310219865e-05, "loss": 0.0192, "step": 6290 }, { "grad_norm": 0.39977946877479553, "learning_rate": 8.200477922794776e-05, "loss": 0.0151, "step": 6300 }, { "grad_norm": 0.361404150724411, "learning_rate": 8.194121785385256e-05, "loss": 0.0173, "step": 6310 }, { "grad_norm": 0.4068446457386017, "learning_rate": 8.187756915368741e-05, "loss": 0.0158, "step": 6320 }, { "grad_norm": 0.4394710063934326, "learning_rate": 8.181383330146544e-05, "loss": 0.0152, "step": 6330 }, { "grad_norm": 0.411819189786911, "learning_rate": 8.175001047143804e-05, "loss": 0.0168, "step": 6340 }, { "grad_norm": 0.3534773588180542, "learning_rate": 8.168610083809438e-05, "loss": 0.0141, "step": 6350 }, { "grad_norm": 0.3123517632484436, "learning_rate": 8.162210457616095e-05, "loss": 0.0157, "step": 6360 }, { "grad_norm": 0.39107418060302734, "learning_rate": 8.155802186060109e-05, "loss": 0.018, "step": 6370 }, { "grad_norm": 0.37755516171455383, "learning_rate": 8.149385286661453e-05, "loss": 0.0151, "step": 6380 }, { "grad_norm": 0.38543444871902466, "learning_rate": 8.14295977696368e-05, "loss": 0.0177, "step": 6390 }, { "grad_norm": 0.3679066598415375, "learning_rate": 8.13652567453389e-05, "loss": 0.0164, "step": 6400 }, { "grad_norm": 0.41568827629089355, "learning_rate": 8.130082996962676e-05, "loss": 0.0163, "step": 6410 }, { "grad_norm": 0.3882181644439697, "learning_rate": 8.123631761864068e-05, "loss": 0.0167, "step": 6420 }, { "grad_norm": 0.3251507878303528, "learning_rate": 8.1171719868755e-05, "loss": 0.0131, "step": 6430 }, { "grad_norm": 0.27662867307662964, "learning_rate": 8.110703689657748e-05, "loss": 0.0149, "step": 6440 }, { "grad_norm": 0.39593058824539185, "learning_rate": 8.104226887894892e-05, "loss": 0.0129, "step": 6450 }, { "grad_norm": 0.4692983627319336, "learning_rate": 8.097741599294257e-05, "loss": 0.0174, "step": 6460 }, { "grad_norm": 0.36407753825187683, "learning_rate": 8.091247841586378e-05, "loss": 0.0191, "step": 6470 }, { "grad_norm": 0.35336682200431824, "learning_rate": 8.084745632524939e-05, "loss": 0.0184, "step": 6480 }, { "grad_norm": 0.33163946866989136, "learning_rate": 8.07823498988673e-05, "loss": 0.0165, "step": 6490 }, { "grad_norm": 0.39307963848114014, "learning_rate": 8.071715931471602e-05, "loss": 0.0157, "step": 6500 }, { "grad_norm": 0.4258623421192169, "learning_rate": 8.06518847510241e-05, "loss": 0.0194, "step": 6510 }, { "grad_norm": 0.3046746551990509, "learning_rate": 8.058652638624971e-05, "loss": 0.0202, "step": 6520 }, { "grad_norm": 0.361548513174057, "learning_rate": 8.052108439908013e-05, "loss": 0.0174, "step": 6530 }, { "grad_norm": 0.3740829527378082, "learning_rate": 8.045555896843125e-05, "loss": 0.0181, "step": 6540 }, { "grad_norm": 0.3932648301124573, "learning_rate": 8.03899502734471e-05, "loss": 0.0169, "step": 6550 }, { "grad_norm": 0.38999900221824646, "learning_rate": 8.032425849349931e-05, "loss": 0.018, "step": 6560 }, { "grad_norm": 0.31013891100883484, "learning_rate": 8.025848380818674e-05, "loss": 0.0217, "step": 6570 }, { "grad_norm": 0.37987402081489563, "learning_rate": 8.019262639733487e-05, "loss": 0.0223, "step": 6580 }, { "grad_norm": 0.398289293050766, "learning_rate": 8.012668644099531e-05, "loss": 0.0179, "step": 6590 }, { "grad_norm": 0.34927797317504883, "learning_rate": 8.006066411944542e-05, "loss": 0.0162, "step": 6600 }, { "grad_norm": 0.3365561068058014, "learning_rate": 7.999455961318769e-05, "loss": 0.0164, "step": 6610 }, { "grad_norm": 0.34584563970565796, "learning_rate": 7.992837310294932e-05, "loss": 0.0155, "step": 6620 }, { "grad_norm": 0.32173582911491394, "learning_rate": 7.986210476968167e-05, "loss": 0.0191, "step": 6630 }, { "grad_norm": 0.4247036576271057, "learning_rate": 7.97957547945599e-05, "loss": 0.017, "step": 6640 }, { "grad_norm": 0.3082517683506012, "learning_rate": 7.972932335898226e-05, "loss": 0.0149, "step": 6650 }, { "grad_norm": 0.4710291922092438, "learning_rate": 7.966281064456975e-05, "loss": 0.0155, "step": 6660 }, { "grad_norm": 0.49894455075263977, "learning_rate": 7.959621683316563e-05, "loss": 0.015, "step": 6670 }, { "grad_norm": 0.38014379143714905, "learning_rate": 7.952954210683481e-05, "loss": 0.0224, "step": 6680 }, { "grad_norm": 0.4054047167301178, "learning_rate": 7.946278664786345e-05, "loss": 0.0168, "step": 6690 }, { "grad_norm": 0.3950726091861725, "learning_rate": 7.939595063875842e-05, "loss": 0.0193, "step": 6700 }, { "grad_norm": 0.5603879690170288, "learning_rate": 7.932903426224683e-05, "loss": 0.0193, "step": 6710 }, { "grad_norm": 0.4113398492336273, "learning_rate": 7.926203770127552e-05, "loss": 0.0188, "step": 6720 }, { "grad_norm": 0.322177529335022, "learning_rate": 7.919496113901046e-05, "loss": 0.0191, "step": 6730 }, { "grad_norm": 0.4656381607055664, "learning_rate": 7.912780475883649e-05, "loss": 0.0204, "step": 6740 }, { "grad_norm": 0.3852991759777069, "learning_rate": 7.906056874435652e-05, "loss": 0.0174, "step": 6750 }, { "grad_norm": 0.4209986627101898, "learning_rate": 7.899325327939131e-05, "loss": 0.016, "step": 6760 }, { "grad_norm": 0.29678499698638916, "learning_rate": 7.892585854797872e-05, "loss": 0.0143, "step": 6770 }, { "grad_norm": 0.31080397963523865, "learning_rate": 7.88583847343734e-05, "loss": 0.0141, "step": 6780 }, { "grad_norm": 0.4389200210571289, "learning_rate": 7.879083202304616e-05, "loss": 0.0149, "step": 6790 }, { "grad_norm": 0.3473112881183624, "learning_rate": 7.872320059868355e-05, "loss": 0.0201, "step": 6800 }, { "grad_norm": 0.3027634620666504, "learning_rate": 7.865549064618729e-05, "loss": 0.0132, "step": 6810 }, { "grad_norm": 0.43607717752456665, "learning_rate": 7.858770235067381e-05, "loss": 0.0149, "step": 6820 }, { "grad_norm": 0.3470240533351898, "learning_rate": 7.851983589747374e-05, "loss": 0.0151, "step": 6830 }, { "grad_norm": 0.3219432234764099, "learning_rate": 7.845189147213133e-05, "loss": 0.0163, "step": 6840 }, { "grad_norm": 0.3376360833644867, "learning_rate": 7.838386926040407e-05, "loss": 0.0147, "step": 6850 }, { "grad_norm": 0.45921677350997925, "learning_rate": 7.83157694482621e-05, "loss": 0.0184, "step": 6860 }, { "grad_norm": 0.23526109755039215, "learning_rate": 7.824759222188768e-05, "loss": 0.0181, "step": 6870 }, { "grad_norm": 0.48352450132369995, "learning_rate": 7.817933776767478e-05, "loss": 0.0216, "step": 6880 }, { "grad_norm": 0.4205549955368042, "learning_rate": 7.811100627222842e-05, "loss": 0.0171, "step": 6890 }, { "grad_norm": 0.43240806460380554, "learning_rate": 7.804259792236435e-05, "loss": 0.0157, "step": 6900 }, { "grad_norm": 0.3053477704524994, "learning_rate": 7.797411290510835e-05, "loss": 0.016, "step": 6910 }, { "grad_norm": 0.3972766399383545, "learning_rate": 7.790555140769586e-05, "loss": 0.0206, "step": 6920 }, { "grad_norm": 0.3287155032157898, "learning_rate": 7.78369136175714e-05, "loss": 0.0131, "step": 6930 }, { "grad_norm": 0.3477262854576111, "learning_rate": 7.776819972238806e-05, "loss": 0.0126, "step": 6940 }, { "grad_norm": 0.3573828637599945, "learning_rate": 7.7699409910007e-05, "loss": 0.0172, "step": 6950 }, { "grad_norm": 0.39164838194847107, "learning_rate": 7.763054436849694e-05, "loss": 0.0156, "step": 6960 }, { "grad_norm": 0.4324200451374054, "learning_rate": 7.756160328613364e-05, "loss": 0.0165, "step": 6970 }, { "grad_norm": 0.3232486844062805, "learning_rate": 7.749258685139942e-05, "loss": 0.0159, "step": 6980 }, { "grad_norm": 0.4205050766468048, "learning_rate": 7.742349525298253e-05, "loss": 0.0185, "step": 6990 }, { "grad_norm": 0.4342349171638489, "learning_rate": 7.735432867977679e-05, "loss": 0.0153, "step": 7000 }, { "grad_norm": 0.3709491491317749, "learning_rate": 7.728508732088096e-05, "loss": 0.0153, "step": 7010 }, { "grad_norm": 0.37684348225593567, "learning_rate": 7.721577136559825e-05, "loss": 0.0121, "step": 7020 }, { "grad_norm": 0.4780879616737366, "learning_rate": 7.714638100343588e-05, "loss": 0.0157, "step": 7030 }, { "grad_norm": 0.399712473154068, "learning_rate": 7.707691642410444e-05, "loss": 0.0186, "step": 7040 }, { "grad_norm": 0.4575657546520233, "learning_rate": 7.70073778175174e-05, "loss": 0.023, "step": 7050 }, { "grad_norm": 0.3211320638656616, "learning_rate": 7.69377653737907e-05, "loss": 0.0157, "step": 7060 }, { "grad_norm": 0.372444212436676, "learning_rate": 7.686807928324209e-05, "loss": 0.0181, "step": 7070 }, { "grad_norm": 0.37389302253723145, "learning_rate": 7.679831973639065e-05, "loss": 0.0141, "step": 7080 }, { "grad_norm": 0.33742061257362366, "learning_rate": 7.672848692395637e-05, "loss": 0.0144, "step": 7090 }, { "grad_norm": 0.352120578289032, "learning_rate": 7.665858103685944e-05, "loss": 0.0164, "step": 7100 }, { "grad_norm": 0.42304760217666626, "learning_rate": 7.658860226621991e-05, "loss": 0.0162, "step": 7110 }, { "grad_norm": 0.4286142587661743, "learning_rate": 7.651855080335708e-05, "loss": 0.0164, "step": 7120 }, { "grad_norm": 0.29648739099502563, "learning_rate": 7.644842683978896e-05, "loss": 0.0139, "step": 7130 }, { "grad_norm": 0.463482528924942, "learning_rate": 7.63782305672318e-05, "loss": 0.0174, "step": 7140 }, { "grad_norm": 0.36419156193733215, "learning_rate": 7.63079621775995e-05, "loss": 0.0182, "step": 7150 }, { "grad_norm": 0.3546525239944458, "learning_rate": 7.623762186300319e-05, "loss": 0.0183, "step": 7160 }, { "grad_norm": 0.5350743532180786, "learning_rate": 7.616720981575057e-05, "loss": 0.0171, "step": 7170 }, { "grad_norm": 0.24242767691612244, "learning_rate": 7.609672622834552e-05, "loss": 0.0176, "step": 7180 }, { "grad_norm": 0.30482882261276245, "learning_rate": 7.602617129348747e-05, "loss": 0.0218, "step": 7190 }, { "grad_norm": 0.37188151478767395, "learning_rate": 7.595554520407088e-05, "loss": 0.0161, "step": 7200 }, { "grad_norm": 0.34896478056907654, "learning_rate": 7.588484815318484e-05, "loss": 0.0125, "step": 7210 }, { "grad_norm": 0.4123700261116028, "learning_rate": 7.581408033411234e-05, "loss": 0.0182, "step": 7220 }, { "grad_norm": 0.30237212777137756, "learning_rate": 7.574324194032995e-05, "loss": 0.0133, "step": 7230 }, { "grad_norm": 0.28881415724754333, "learning_rate": 7.567233316550705e-05, "loss": 0.0134, "step": 7240 }, { "grad_norm": 0.4022943675518036, "learning_rate": 7.560135420350562e-05, "loss": 0.0132, "step": 7250 }, { "grad_norm": 0.34321826696395874, "learning_rate": 7.553030524837935e-05, "loss": 0.0148, "step": 7260 }, { "grad_norm": 0.37172824144363403, "learning_rate": 7.545918649437341e-05, "loss": 0.0138, "step": 7270 }, { "grad_norm": 0.38254666328430176, "learning_rate": 7.538799813592377e-05, "loss": 0.013, "step": 7280 }, { "grad_norm": 0.3662221133708954, "learning_rate": 7.531674036765662e-05, "loss": 0.0136, "step": 7290 }, { "grad_norm": 0.48948177695274353, "learning_rate": 7.524541338438807e-05, "loss": 0.0158, "step": 7300 }, { "grad_norm": 0.31744667887687683, "learning_rate": 7.517401738112328e-05, "loss": 0.0177, "step": 7310 }, { "grad_norm": 0.35522255301475525, "learning_rate": 7.510255255305628e-05, "loss": 0.0135, "step": 7320 }, { "grad_norm": 0.38696733117103577, "learning_rate": 7.503101909556911e-05, "loss": 0.016, "step": 7330 }, { "grad_norm": 0.3476850390434265, "learning_rate": 7.495941720423154e-05, "loss": 0.0148, "step": 7340 }, { "grad_norm": 0.30933910608291626, "learning_rate": 7.488774707480042e-05, "loss": 0.0131, "step": 7350 }, { "grad_norm": 0.38087910413742065, "learning_rate": 7.481600890321911e-05, "loss": 0.0139, "step": 7360 }, { "grad_norm": 0.3373841941356659, "learning_rate": 7.474420288561708e-05, "loss": 0.015, "step": 7370 }, { "grad_norm": 0.35937246680259705, "learning_rate": 7.467232921830921e-05, "loss": 0.0166, "step": 7380 }, { "grad_norm": 0.2769072949886322, "learning_rate": 7.460038809779537e-05, "loss": 0.0179, "step": 7390 }, { "grad_norm": 0.35545891523361206, "learning_rate": 7.452837972075983e-05, "loss": 0.0153, "step": 7400 }, { "grad_norm": 0.32261455059051514, "learning_rate": 7.445630428407074e-05, "loss": 0.0141, "step": 7410 }, { "grad_norm": 0.43360084295272827, "learning_rate": 7.43841619847796e-05, "loss": 0.0171, "step": 7420 }, { "grad_norm": 0.41580498218536377, "learning_rate": 7.431195302012072e-05, "loss": 0.0156, "step": 7430 }, { "grad_norm": 0.3012297451496124, "learning_rate": 7.423967758751061e-05, "loss": 0.0174, "step": 7440 }, { "grad_norm": 0.38392868638038635, "learning_rate": 7.416733588454758e-05, "loss": 0.0141, "step": 7450 }, { "grad_norm": 0.3797113001346588, "learning_rate": 7.409492810901106e-05, "loss": 0.0167, "step": 7460 }, { "grad_norm": 0.34098389744758606, "learning_rate": 7.402245445886116e-05, "loss": 0.0148, "step": 7470 }, { "grad_norm": 0.24612897634506226, "learning_rate": 7.394991513223806e-05, "loss": 0.0147, "step": 7480 }, { "grad_norm": 0.3127574920654297, "learning_rate": 7.38773103274615e-05, "loss": 0.0166, "step": 7490 }, { "grad_norm": 0.3971444368362427, "learning_rate": 7.380464024303028e-05, "loss": 0.0136, "step": 7500 }, { "grad_norm": 0.38821256160736084, "learning_rate": 7.373190507762162e-05, "loss": 0.0116, "step": 7510 }, { "grad_norm": 0.3289741575717926, "learning_rate": 7.365910503009066e-05, "loss": 0.0168, "step": 7520 }, { "grad_norm": 0.5070462226867676, "learning_rate": 7.358624029946996e-05, "loss": 0.0156, "step": 7530 }, { "grad_norm": 0.35847413539886475, "learning_rate": 7.351331108496893e-05, "loss": 0.0126, "step": 7540 }, { "grad_norm": 0.29011738300323486, "learning_rate": 7.344031758597325e-05, "loss": 0.0122, "step": 7550 }, { "grad_norm": 0.40054792165756226, "learning_rate": 7.336726000204435e-05, "loss": 0.0122, "step": 7560 }, { "grad_norm": 0.3142523467540741, "learning_rate": 7.32941385329189e-05, "loss": 0.0155, "step": 7570 }, { "grad_norm": 0.44921424984931946, "learning_rate": 7.322095337850816e-05, "loss": 0.0161, "step": 7580 }, { "grad_norm": 0.4189692735671997, "learning_rate": 7.314770473889758e-05, "loss": 0.014, "step": 7590 }, { "grad_norm": 0.31764888763427734, "learning_rate": 7.307439281434615e-05, "loss": 0.0161, "step": 7600 }, { "grad_norm": 0.3387962281703949, "learning_rate": 7.300101780528585e-05, "loss": 0.0126, "step": 7610 }, { "grad_norm": 0.3220970034599304, "learning_rate": 7.292757991232117e-05, "loss": 0.0135, "step": 7620 }, { "grad_norm": 0.29096081852912903, "learning_rate": 7.285407933622848e-05, "loss": 0.0135, "step": 7630 }, { "grad_norm": 0.40973615646362305, "learning_rate": 7.278051627795557e-05, "loss": 0.0128, "step": 7640 }, { "grad_norm": 0.3823937177658081, "learning_rate": 7.270689093862105e-05, "loss": 0.0158, "step": 7650 }, { "grad_norm": 0.29811376333236694, "learning_rate": 7.263320351951374e-05, "loss": 0.0123, "step": 7660 }, { "grad_norm": 0.3254123330116272, "learning_rate": 7.255945422209227e-05, "loss": 0.0135, "step": 7670 }, { "grad_norm": 0.3211536109447479, "learning_rate": 7.248564324798437e-05, "loss": 0.0115, "step": 7680 }, { "grad_norm": 0.4133204519748688, "learning_rate": 7.241177079898644e-05, "loss": 0.013, "step": 7690 }, { "grad_norm": 0.42851558327674866, "learning_rate": 7.233783707706295e-05, "loss": 0.0132, "step": 7700 }, { "grad_norm": 0.3776419460773468, "learning_rate": 7.226384228434586e-05, "loss": 0.0147, "step": 7710 }, { "grad_norm": 0.3443582057952881, "learning_rate": 7.21897866231341e-05, "loss": 0.0124, "step": 7720 }, { "grad_norm": 0.4206068217754364, "learning_rate": 7.211567029589303e-05, "loss": 0.016, "step": 7730 }, { "grad_norm": 0.40361082553863525, "learning_rate": 7.204149350525387e-05, "loss": 0.0125, "step": 7740 }, { "grad_norm": 0.4060281217098236, "learning_rate": 7.196725645401309e-05, "loss": 0.0136, "step": 7750 }, { "grad_norm": 0.40346235036849976, "learning_rate": 7.1892959345132e-05, "loss": 0.0132, "step": 7760 }, { "grad_norm": 0.36330267786979675, "learning_rate": 7.181860238173605e-05, "loss": 0.0113, "step": 7770 }, { "grad_norm": 0.4041442275047302, "learning_rate": 7.174418576711432e-05, "loss": 0.0148, "step": 7780 }, { "grad_norm": 0.37861278653144836, "learning_rate": 7.1669709704719e-05, "loss": 0.014, "step": 7790 }, { "grad_norm": 0.34117135405540466, "learning_rate": 7.159517439816481e-05, "loss": 0.0148, "step": 7800 }, { "grad_norm": 0.35832077264785767, "learning_rate": 7.152058005122842e-05, "loss": 0.0154, "step": 7810 }, { "grad_norm": 0.3983708918094635, "learning_rate": 7.144592686784793e-05, "loss": 0.0182, "step": 7820 }, { "grad_norm": 0.3107507824897766, "learning_rate": 7.137121505212229e-05, "loss": 0.0204, "step": 7830 }, { "grad_norm": 0.44682419300079346, "learning_rate": 7.129644480831077e-05, "loss": 0.0152, "step": 7840 }, { "grad_norm": 0.3947400450706482, "learning_rate": 7.122161634083234e-05, "loss": 0.0131, "step": 7850 }, { "grad_norm": 0.3726661801338196, "learning_rate": 7.114672985426516e-05, "loss": 0.0119, "step": 7860 }, { "grad_norm": 0.2852039039134979, "learning_rate": 7.107178555334606e-05, "loss": 0.0126, "step": 7870 }, { "grad_norm": 0.3068068325519562, "learning_rate": 7.099678364296989e-05, "loss": 0.0133, "step": 7880 }, { "grad_norm": 0.34647849202156067, "learning_rate": 7.0921724328189e-05, "loss": 0.0125, "step": 7890 }, { "grad_norm": 0.3441995680332184, "learning_rate": 7.084660781421268e-05, "loss": 0.0126, "step": 7900 }, { "grad_norm": 0.32749423384666443, "learning_rate": 7.077143430640662e-05, "loss": 0.0163, "step": 7910 }, { "grad_norm": 0.3520297110080719, "learning_rate": 7.069620401029232e-05, "loss": 0.0132, "step": 7920 }, { "grad_norm": 0.28060904145240784, "learning_rate": 7.062091713154655e-05, "loss": 0.0142, "step": 7930 }, { "grad_norm": 0.34220680594444275, "learning_rate": 7.054557387600075e-05, "loss": 0.0133, "step": 7940 }, { "grad_norm": 0.30184972286224365, "learning_rate": 7.04701744496405e-05, "loss": 0.0187, "step": 7950 }, { "grad_norm": 0.4050672948360443, "learning_rate": 7.039471905860495e-05, "loss": 0.0208, "step": 7960 }, { "grad_norm": 0.403519868850708, "learning_rate": 7.031920790918628e-05, "loss": 0.0149, "step": 7970 }, { "grad_norm": 0.327185720205307, "learning_rate": 7.024364120782906e-05, "loss": 0.0219, "step": 7980 }, { "grad_norm": 0.29921942949295044, "learning_rate": 7.016801916112978e-05, "loss": 0.0152, "step": 7990 }, { "grad_norm": 0.2583206593990326, "learning_rate": 7.009234197583623e-05, "loss": 0.0137, "step": 8000 }, { "grad_norm": 0.24494603276252747, "learning_rate": 7.001660985884692e-05, "loss": 0.0156, "step": 8010 }, { "grad_norm": 0.34610432386398315, "learning_rate": 6.994082301721063e-05, "loss": 0.0137, "step": 8020 }, { "grad_norm": 0.31391584873199463, "learning_rate": 6.986498165812563e-05, "loss": 0.0143, "step": 8030 }, { "grad_norm": 0.2702752351760864, "learning_rate": 6.978908598893932e-05, "loss": 0.0123, "step": 8040 }, { "grad_norm": 0.3130239248275757, "learning_rate": 6.971313621714756e-05, "loss": 0.0108, "step": 8050 }, { "grad_norm": 0.3876408338546753, "learning_rate": 6.96371325503941e-05, "loss": 0.0132, "step": 8060 }, { "grad_norm": 0.2897747755050659, "learning_rate": 6.956107519647014e-05, "loss": 0.0115, "step": 8070 }, { "grad_norm": 0.276584655046463, "learning_rate": 6.94849643633135e-05, "loss": 0.0122, "step": 8080 }, { "grad_norm": 0.3668515384197235, "learning_rate": 6.940880025900834e-05, "loss": 0.0111, "step": 8090 }, { "grad_norm": 0.2898213267326355, "learning_rate": 6.933258309178438e-05, "loss": 0.011, "step": 8100 }, { "grad_norm": 0.22109735012054443, "learning_rate": 6.925631307001646e-05, "loss": 0.0124, "step": 8110 }, { "grad_norm": 0.2808288037776947, "learning_rate": 6.91799904022239e-05, "loss": 0.0131, "step": 8120 }, { "grad_norm": 0.30975374579429626, "learning_rate": 6.910361529706997e-05, "loss": 0.0106, "step": 8130 }, { "grad_norm": 0.34880274534225464, "learning_rate": 6.902718796336131e-05, "loss": 0.0138, "step": 8140 }, { "grad_norm": 0.35838931798934937, "learning_rate": 6.895070861004729e-05, "loss": 0.0122, "step": 8150 }, { "grad_norm": 0.28413712978363037, "learning_rate": 6.887417744621956e-05, "loss": 0.012, "step": 8160 }, { "grad_norm": 0.2813417315483093, "learning_rate": 6.87975946811114e-05, "loss": 0.0123, "step": 8170 }, { "grad_norm": 0.42694875597953796, "learning_rate": 6.872096052409718e-05, "loss": 0.015, "step": 8180 }, { "grad_norm": 0.4038955867290497, "learning_rate": 6.864427518469174e-05, "loss": 0.0128, "step": 8190 }, { "grad_norm": 0.3710439205169678, "learning_rate": 6.856753887254986e-05, "loss": 0.0123, "step": 8200 }, { "grad_norm": 0.2244020253419876, "learning_rate": 6.849075179746572e-05, "loss": 0.0129, "step": 8210 }, { "grad_norm": 0.23305191099643707, "learning_rate": 6.841391416937221e-05, "loss": 0.0133, "step": 8220 }, { "grad_norm": 0.45240503549575806, "learning_rate": 6.833702619834053e-05, "loss": 0.0142, "step": 8230 }, { "grad_norm": 0.2898019254207611, "learning_rate": 6.82600880945794e-05, "loss": 0.0114, "step": 8240 }, { "grad_norm": 0.2994280457496643, "learning_rate": 6.818310006843468e-05, "loss": 0.0109, "step": 8250 }, { "grad_norm": 0.2994014620780945, "learning_rate": 6.810606233038868e-05, "loss": 0.0117, "step": 8260 }, { "grad_norm": 0.40484005212783813, "learning_rate": 6.802897509105966e-05, "loss": 0.0183, "step": 8270 }, { "grad_norm": 0.3237367272377014, "learning_rate": 6.79518385612012e-05, "loss": 0.0138, "step": 8280 }, { "grad_norm": 0.3805217444896698, "learning_rate": 6.787465295170157e-05, "loss": 0.0141, "step": 8290 }, { "grad_norm": 0.38457584381103516, "learning_rate": 6.779741847358332e-05, "loss": 0.012, "step": 8300 }, { "grad_norm": 0.3109392523765564, "learning_rate": 6.772013533800256e-05, "loss": 0.0169, "step": 8310 }, { "grad_norm": 0.38964930176734924, "learning_rate": 6.764280375624843e-05, "loss": 0.0131, "step": 8320 }, { "grad_norm": 0.3863414525985718, "learning_rate": 6.756542393974252e-05, "loss": 0.0129, "step": 8330 }, { "grad_norm": 0.3213522136211395, "learning_rate": 6.748799610003828e-05, "loss": 0.0125, "step": 8340 }, { "grad_norm": 0.3722427189350128, "learning_rate": 6.741052044882048e-05, "loss": 0.0139, "step": 8350 }, { "grad_norm": 0.4136035442352295, "learning_rate": 6.73329971979046e-05, "loss": 0.0121, "step": 8360 }, { "grad_norm": 0.3353002369403839, "learning_rate": 6.725542655923625e-05, "loss": 0.01, "step": 8370 }, { "grad_norm": 0.24848094582557678, "learning_rate": 6.717780874489057e-05, "loss": 0.0125, "step": 8380 }, { "grad_norm": 0.3231047987937927, "learning_rate": 6.710014396707172e-05, "loss": 0.013, "step": 8390 }, { "grad_norm": 0.3344343602657318, "learning_rate": 6.702243243811221e-05, "loss": 0.0104, "step": 8400 }, { "grad_norm": 0.3377920687198639, "learning_rate": 6.694467437047244e-05, "loss": 0.0115, "step": 8410 }, { "grad_norm": 0.36382901668548584, "learning_rate": 6.686686997673997e-05, "loss": 0.0122, "step": 8420 }, { "grad_norm": 0.35179221630096436, "learning_rate": 6.678901946962903e-05, "loss": 0.0124, "step": 8430 }, { "grad_norm": 0.31546005606651306, "learning_rate": 6.671112306197996e-05, "loss": 0.0114, "step": 8440 }, { "grad_norm": 0.3519761860370636, "learning_rate": 6.663318096675854e-05, "loss": 0.0109, "step": 8450 }, { "grad_norm": 0.27498510479927063, "learning_rate": 6.655519339705552e-05, "loss": 0.0097, "step": 8460 }, { "grad_norm": 0.2949582636356354, "learning_rate": 6.647716056608588e-05, "loss": 0.0102, "step": 8470 }, { "grad_norm": 0.3293176293373108, "learning_rate": 6.639908268718843e-05, "loss": 0.0165, "step": 8480 }, { "grad_norm": 0.42313507199287415, "learning_rate": 6.632095997382514e-05, "loss": 0.0139, "step": 8490 }, { "grad_norm": 0.39244991540908813, "learning_rate": 6.624279263958047e-05, "loss": 0.011, "step": 8500 }, { "grad_norm": 0.3285278379917145, "learning_rate": 6.616458089816097e-05, "loss": 0.0132, "step": 8510 }, { "grad_norm": 0.3824060261249542, "learning_rate": 6.608632496339454e-05, "loss": 0.0115, "step": 8520 }, { "grad_norm": 0.32687434554100037, "learning_rate": 6.600802504922988e-05, "loss": 0.0108, "step": 8530 }, { "grad_norm": 0.2568399906158447, "learning_rate": 6.592968136973604e-05, "loss": 0.0106, "step": 8540 }, { "grad_norm": 0.3117976188659668, "learning_rate": 6.585129413910159e-05, "loss": 0.0122, "step": 8550 }, { "grad_norm": 0.26663267612457275, "learning_rate": 6.577286357163424e-05, "loss": 0.0107, "step": 8560 }, { "grad_norm": 0.380484402179718, "learning_rate": 6.569438988176018e-05, "loss": 0.0125, "step": 8570 }, { "grad_norm": 0.35131096839904785, "learning_rate": 6.561587328402347e-05, "loss": 0.011, "step": 8580 }, { "grad_norm": 0.26992419362068176, "learning_rate": 6.553731399308549e-05, "loss": 0.0101, "step": 8590 }, { "grad_norm": 0.38081663846969604, "learning_rate": 6.545871222372436e-05, "loss": 0.0135, "step": 8600 }, { "grad_norm": 0.33266904950141907, "learning_rate": 6.538006819083426e-05, "loss": 0.0132, "step": 8610 }, { "grad_norm": 0.28045380115509033, "learning_rate": 6.530138210942505e-05, "loss": 0.0128, "step": 8620 }, { "grad_norm": 0.34392043948173523, "learning_rate": 6.522265419462141e-05, "loss": 0.0115, "step": 8630 }, { "grad_norm": 0.248205304145813, "learning_rate": 6.514388466166248e-05, "loss": 0.0106, "step": 8640 }, { "grad_norm": 0.23977991938591003, "learning_rate": 6.506507372590119e-05, "loss": 0.0145, "step": 8650 }, { "grad_norm": 0.41651660203933716, "learning_rate": 6.498622160280355e-05, "loss": 0.0116, "step": 8660 }, { "grad_norm": 0.3240855932235718, "learning_rate": 6.490732850794832e-05, "loss": 0.0119, "step": 8670 }, { "grad_norm": 0.3398783504962921, "learning_rate": 6.482839465702616e-05, "loss": 0.0149, "step": 8680 }, { "grad_norm": 0.43641889095306396, "learning_rate": 6.474942026583923e-05, "loss": 0.0114, "step": 8690 }, { "grad_norm": 0.3041832447052002, "learning_rate": 6.467040555030052e-05, "loss": 0.0129, "step": 8700 }, { "grad_norm": 0.28747713565826416, "learning_rate": 6.459135072643321e-05, "loss": 0.0089, "step": 8710 }, { "grad_norm": 0.31746140122413635, "learning_rate": 6.451225601037019e-05, "loss": 0.0111, "step": 8720 }, { "grad_norm": 0.24119898676872253, "learning_rate": 6.443312161835338e-05, "loss": 0.0122, "step": 8730 }, { "grad_norm": 0.3357892632484436, "learning_rate": 6.43539477667332e-05, "loss": 0.0112, "step": 8740 }, { "grad_norm": 0.4408676028251648, "learning_rate": 6.427473467196793e-05, "loss": 0.0127, "step": 8750 }, { "grad_norm": 0.4194582402706146, "learning_rate": 6.419548255062315e-05, "loss": 0.0128, "step": 8760 }, { "grad_norm": 0.2998194396495819, "learning_rate": 6.411619161937112e-05, "loss": 0.0118, "step": 8770 }, { "grad_norm": 0.3880668580532074, "learning_rate": 6.403686209499022e-05, "loss": 0.0138, "step": 8780 }, { "grad_norm": 0.41864970326423645, "learning_rate": 6.395749419436437e-05, "loss": 0.0145, "step": 8790 }, { "grad_norm": 0.31571438908576965, "learning_rate": 6.387808813448234e-05, "loss": 0.0124, "step": 8800 }, { "grad_norm": 0.2976234555244446, "learning_rate": 6.37986441324373e-05, "loss": 0.0152, "step": 8810 }, { "grad_norm": 0.38963109254837036, "learning_rate": 6.37191624054261e-05, "loss": 0.013, "step": 8820 }, { "grad_norm": 0.35120075941085815, "learning_rate": 6.363964317074872e-05, "loss": 0.0118, "step": 8830 }, { "grad_norm": 0.3403359651565552, "learning_rate": 6.356008664580776e-05, "loss": 0.0097, "step": 8840 }, { "grad_norm": 0.3545701801776886, "learning_rate": 6.348049304810771e-05, "loss": 0.0129, "step": 8850 }, { "grad_norm": 0.3559361696243286, "learning_rate": 6.340086259525442e-05, "loss": 0.0134, "step": 8860 }, { "grad_norm": 0.30377861857414246, "learning_rate": 6.332119550495448e-05, "loss": 0.0106, "step": 8870 }, { "grad_norm": 0.4019133150577545, "learning_rate": 6.324149199501473e-05, "loss": 0.0126, "step": 8880 }, { "grad_norm": 0.3313150703907013, "learning_rate": 6.316175228334146e-05, "loss": 0.0109, "step": 8890 }, { "grad_norm": 0.28782913088798523, "learning_rate": 6.308197658794003e-05, "loss": 0.0152, "step": 8900 }, { "grad_norm": 0.2732824683189392, "learning_rate": 6.300216512691417e-05, "loss": 0.01, "step": 8910 }, { "grad_norm": 0.3199782967567444, "learning_rate": 6.292231811846532e-05, "loss": 0.0106, "step": 8920 }, { "grad_norm": 0.21855194866657257, "learning_rate": 6.284243578089217e-05, "loss": 0.0104, "step": 8930 }, { "grad_norm": 0.3417510688304901, "learning_rate": 6.276251833258999e-05, "loss": 0.0113, "step": 8940 }, { "grad_norm": 0.28877514600753784, "learning_rate": 6.268256599205003e-05, "loss": 0.0134, "step": 8950 }, { "grad_norm": 0.34601452946662903, "learning_rate": 6.260257897785892e-05, "loss": 0.0131, "step": 8960 }, { "grad_norm": 0.3036784529685974, "learning_rate": 6.252255750869811e-05, "loss": 0.0099, "step": 8970 }, { "grad_norm": 0.21555252373218536, "learning_rate": 6.244250180334325e-05, "loss": 0.0142, "step": 8980 }, { "grad_norm": 0.38473767042160034, "learning_rate": 6.236241208066356e-05, "loss": 0.0129, "step": 8990 }, { "grad_norm": 0.4134522080421448, "learning_rate": 6.228228855962133e-05, "loss": 0.0157, "step": 9000 }, { "grad_norm": 0.3341606855392456, "learning_rate": 6.220213145927115e-05, "loss": 0.0116, "step": 9010 }, { "grad_norm": 0.2900019586086273, "learning_rate": 6.212194099875951e-05, "loss": 0.0102, "step": 9020 }, { "grad_norm": 0.3076825439929962, "learning_rate": 6.204171739732405e-05, "loss": 0.0105, "step": 9030 }, { "grad_norm": 0.27437257766723633, "learning_rate": 6.196146087429303e-05, "loss": 0.01, "step": 9040 }, { "grad_norm": 0.2546038329601288, "learning_rate": 6.188117164908474e-05, "loss": 0.0113, "step": 9050 }, { "grad_norm": 0.25115546584129333, "learning_rate": 6.180084994120684e-05, "loss": 0.0119, "step": 9060 }, { "grad_norm": 0.39839228987693787, "learning_rate": 6.17204959702558e-05, "loss": 0.0118, "step": 9070 }, { "grad_norm": 0.3079279065132141, "learning_rate": 6.164010995591635e-05, "loss": 0.0121, "step": 9080 }, { "grad_norm": 0.29462292790412903, "learning_rate": 6.155969211796076e-05, "loss": 0.011, "step": 9090 }, { "grad_norm": 0.32893216609954834, "learning_rate": 6.147924267624829e-05, "loss": 0.0129, "step": 9100 }, { "grad_norm": 0.29549315571784973, "learning_rate": 6.13987618507247e-05, "loss": 0.0137, "step": 9110 }, { "grad_norm": 0.4640045166015625, "learning_rate": 6.131824986142147e-05, "loss": 0.0125, "step": 9120 }, { "grad_norm": 0.2915726900100708, "learning_rate": 6.123770692845529e-05, "loss": 0.0091, "step": 9130 }, { "grad_norm": 0.2656380534172058, "learning_rate": 6.11571332720275e-05, "loss": 0.0116, "step": 9140 }, { "grad_norm": 0.3688591718673706, "learning_rate": 6.107652911242336e-05, "loss": 0.015, "step": 9150 }, { "grad_norm": 0.29144078493118286, "learning_rate": 6.0995894670011586e-05, "loss": 0.0134, "step": 9160 }, { "grad_norm": 0.3975711464881897, "learning_rate": 6.091523016524368e-05, "loss": 0.0141, "step": 9170 }, { "grad_norm": 0.34893518686294556, "learning_rate": 6.083453581865328e-05, "loss": 0.0132, "step": 9180 }, { "grad_norm": 0.283613920211792, "learning_rate": 6.075381185085568e-05, "loss": 0.0113, "step": 9190 }, { "grad_norm": 0.33413824439048767, "learning_rate": 6.067305848254709e-05, "loss": 0.0143, "step": 9200 }, { "grad_norm": 0.32717165350914, "learning_rate": 6.059227593450418e-05, "loss": 0.0115, "step": 9210 }, { "grad_norm": 0.27025970816612244, "learning_rate": 6.051146442758333e-05, "loss": 0.0151, "step": 9220 }, { "grad_norm": 0.33592984080314636, "learning_rate": 6.043062418272012e-05, "loss": 0.0122, "step": 9230 }, { "grad_norm": 0.3122248351573944, "learning_rate": 6.0349755420928666e-05, "loss": 0.0111, "step": 9240 }, { "grad_norm": 0.39123398065567017, "learning_rate": 6.0268858363301105e-05, "loss": 0.0132, "step": 9250 }, { "grad_norm": 0.2760530114173889, "learning_rate": 6.018793323100689e-05, "loss": 0.0112, "step": 9260 }, { "grad_norm": 0.28451481461524963, "learning_rate": 6.0106980245292255e-05, "loss": 0.0117, "step": 9270 }, { "grad_norm": 0.3141275942325592, "learning_rate": 6.002599962747957e-05, "loss": 0.0088, "step": 9280 }, { "grad_norm": 0.3256250321865082, "learning_rate": 5.994499159896673e-05, "loss": 0.011, "step": 9290 }, { "grad_norm": 0.34313273429870605, "learning_rate": 5.9863956381226607e-05, "loss": 0.0133, "step": 9300 }, { "grad_norm": 0.2991064190864563, "learning_rate": 5.9782894195806394e-05, "loss": 0.012, "step": 9310 }, { "grad_norm": 0.2878324091434479, "learning_rate": 5.9701805264327004e-05, "loss": 0.0121, "step": 9320 }, { "grad_norm": 0.34685784578323364, "learning_rate": 5.96206898084825e-05, "loss": 0.0123, "step": 9330 }, { "grad_norm": 0.3606438934803009, "learning_rate": 5.953954805003942e-05, "loss": 0.0127, "step": 9340 }, { "grad_norm": 0.29037320613861084, "learning_rate": 5.945838021083623e-05, "loss": 0.0105, "step": 9350 }, { "grad_norm": 0.3555743992328644, "learning_rate": 5.9377186512782714e-05, "loss": 0.0154, "step": 9360 }, { "grad_norm": 0.24244317412376404, "learning_rate": 5.929596717785935e-05, "loss": 0.0163, "step": 9370 }, { "grad_norm": 0.2734943628311157, "learning_rate": 5.921472242811668e-05, "loss": 0.0141, "step": 9380 }, { "grad_norm": 0.3811601400375366, "learning_rate": 5.913345248567475e-05, "loss": 0.0138, "step": 9390 }, { "grad_norm": 0.42668387293815613, "learning_rate": 5.905215757272248e-05, "loss": 0.0104, "step": 9400 }, { "grad_norm": 0.37520602345466614, "learning_rate": 5.897083791151706e-05, "loss": 0.0114, "step": 9410 }, { "grad_norm": 0.26095592975616455, "learning_rate": 5.888949372438336e-05, "loss": 0.0124, "step": 9420 }, { "grad_norm": 0.3601202368736267, "learning_rate": 5.8808125233713255e-05, "loss": 0.0108, "step": 9430 }, { "grad_norm": 0.3249405324459076, "learning_rate": 5.872673266196509e-05, "loss": 0.0144, "step": 9440 }, { "grad_norm": 0.352830708026886, "learning_rate": 5.864531623166305e-05, "loss": 0.0115, "step": 9450 }, { "grad_norm": 0.3386164903640747, "learning_rate": 5.856387616539656e-05, "loss": 0.0135, "step": 9460 }, { "grad_norm": 0.37539342045783997, "learning_rate": 5.848241268581967e-05, "loss": 0.0125, "step": 9470 }, { "grad_norm": 0.33391866087913513, "learning_rate": 5.840092601565037e-05, "loss": 0.0104, "step": 9480 }, { "grad_norm": 0.31910452246665955, "learning_rate": 5.8319416377670144e-05, "loss": 0.0134, "step": 9490 }, { "grad_norm": 0.3447106182575226, "learning_rate": 5.82378839947232e-05, "loss": 0.0146, "step": 9500 }, { "grad_norm": 0.30419379472732544, "learning_rate": 5.815632908971599e-05, "loss": 0.0128, "step": 9510 }, { "grad_norm": 0.34695306420326233, "learning_rate": 5.80747518856165e-05, "loss": 0.0149, "step": 9520 }, { "grad_norm": 0.32170993089675903, "learning_rate": 5.799315260545367e-05, "loss": 0.011, "step": 9530 }, { "grad_norm": 0.3228655457496643, "learning_rate": 5.791153147231686e-05, "loss": 0.0137, "step": 9540 }, { "grad_norm": 0.4017237722873688, "learning_rate": 5.782988870935509e-05, "loss": 0.0133, "step": 9550 }, { "grad_norm": 0.47601988911628723, "learning_rate": 5.774822453977657e-05, "loss": 0.0131, "step": 9560 }, { "grad_norm": 0.3700137138366699, "learning_rate": 5.7666539186848036e-05, "loss": 0.0131, "step": 9570 }, { "grad_norm": 0.32791244983673096, "learning_rate": 5.758483287389411e-05, "loss": 0.0137, "step": 9580 }, { "grad_norm": 0.2311929613351822, "learning_rate": 5.7503105824296735e-05, "loss": 0.0091, "step": 9590 }, { "grad_norm": 0.30388620495796204, "learning_rate": 5.742135826149453e-05, "loss": 0.0112, "step": 9600 }, { "grad_norm": 0.2672225534915924, "learning_rate": 5.7339590408982223e-05, "loss": 0.0133, "step": 9610 }, { "grad_norm": 0.3468799293041229, "learning_rate": 5.725780249031e-05, "loss": 0.0119, "step": 9620 }, { "grad_norm": 0.38424280285835266, "learning_rate": 5.717599472908292e-05, "loss": 0.0137, "step": 9630 }, { "grad_norm": 0.2739831507205963, "learning_rate": 5.7094167348960237e-05, "loss": 0.0098, "step": 9640 }, { "grad_norm": 0.37700366973876953, "learning_rate": 5.7012320573654945e-05, "loss": 0.0139, "step": 9650 }, { "grad_norm": 0.3473300337791443, "learning_rate": 5.693045462693295e-05, "loss": 0.011, "step": 9660 }, { "grad_norm": 0.28770017623901367, "learning_rate": 5.684856973261266e-05, "loss": 0.0106, "step": 9670 }, { "grad_norm": 0.34632372856140137, "learning_rate": 5.6766666114564215e-05, "loss": 0.0109, "step": 9680 }, { "grad_norm": 0.2771446406841278, "learning_rate": 5.668474399670899e-05, "loss": 0.0094, "step": 9690 }, { "grad_norm": 0.23416583240032196, "learning_rate": 5.660280360301896e-05, "loss": 0.009, "step": 9700 }, { "grad_norm": 0.36180397868156433, "learning_rate": 5.652084515751599e-05, "loss": 0.0112, "step": 9710 }, { "grad_norm": 0.37159043550491333, "learning_rate": 5.643886888427137e-05, "loss": 0.0131, "step": 9720 }, { "grad_norm": 0.27281299233436584, "learning_rate": 5.6356875007405074e-05, "loss": 0.0119, "step": 9730 }, { "grad_norm": 0.2391415238380432, "learning_rate": 5.627486375108525e-05, "loss": 0.0098, "step": 9740 }, { "grad_norm": 0.30104291439056396, "learning_rate": 5.619283533952754e-05, "loss": 0.0102, "step": 9750 }, { "grad_norm": 0.28203320503234863, "learning_rate": 5.6110789996994474e-05, "loss": 0.0089, "step": 9760 }, { "grad_norm": 0.33479997515678406, "learning_rate": 5.602872794779491e-05, "loss": 0.009, "step": 9770 }, { "grad_norm": 0.2430616170167923, "learning_rate": 5.594664941628334e-05, "loss": 0.0094, "step": 9780 }, { "grad_norm": 0.23133553564548492, "learning_rate": 5.5864554626859324e-05, "loss": 0.0112, "step": 9790 }, { "grad_norm": 0.28496891260147095, "learning_rate": 5.578244380396691e-05, "loss": 0.0082, "step": 9800 }, { "grad_norm": 0.3224879205226898, "learning_rate": 5.570031717209394e-05, "loss": 0.0112, "step": 9810 }, { "grad_norm": 0.4145388603210449, "learning_rate": 5.561817495577147e-05, "loss": 0.0082, "step": 9820 }, { "grad_norm": 0.3475486934185028, "learning_rate": 5.5536017379573215e-05, "loss": 0.0133, "step": 9830 }, { "grad_norm": 0.37255167961120605, "learning_rate": 5.545384466811483e-05, "loss": 0.0132, "step": 9840 }, { "grad_norm": 0.3716874420642853, "learning_rate": 5.5371657046053384e-05, "loss": 0.0141, "step": 9850 }, { "grad_norm": 0.36983785033226013, "learning_rate": 5.528945473808669e-05, "loss": 0.0134, "step": 9860 }, { "grad_norm": 0.27859365940093994, "learning_rate": 5.520723796895272e-05, "loss": 0.0102, "step": 9870 }, { "grad_norm": 0.27919843792915344, "learning_rate": 5.512500696342897e-05, "loss": 0.0129, "step": 9880 }, { "grad_norm": 0.3171057105064392, "learning_rate": 5.504276194633188e-05, "loss": 0.0146, "step": 9890 }, { "grad_norm": 0.34983888268470764, "learning_rate": 5.49605031425162e-05, "loss": 0.0135, "step": 9900 }, { "grad_norm": 0.38844263553619385, "learning_rate": 5.487823077687434e-05, "loss": 0.0147, "step": 9910 }, { "grad_norm": 0.2128477394580841, "learning_rate": 5.4795945074335806e-05, "loss": 0.0153, "step": 9920 }, { "grad_norm": 0.28172314167022705, "learning_rate": 5.471364625986657e-05, "loss": 0.0089, "step": 9930 }, { "grad_norm": 0.24061772227287292, "learning_rate": 5.463133455846845e-05, "loss": 0.0136, "step": 9940 }, { "grad_norm": 0.3619382679462433, "learning_rate": 5.4549010195178505e-05, "loss": 0.0132, "step": 9950 }, { "grad_norm": 0.32074296474456787, "learning_rate": 5.446667339506838e-05, "loss": 0.0118, "step": 9960 }, { "grad_norm": 0.30590108036994934, "learning_rate": 5.4384324383243756e-05, "loss": 0.0107, "step": 9970 }, { "grad_norm": 0.3535405099391937, "learning_rate": 5.430196338484368e-05, "loss": 0.0111, "step": 9980 }, { "grad_norm": 0.38065770268440247, "learning_rate": 5.4219590625039975e-05, "loss": 0.012, "step": 9990 }, { "grad_norm": 0.37209638953208923, "learning_rate": 5.413720632903664e-05, "loss": 0.0139, "step": 10000 }, { "grad_norm": 0.2975228428840637, "learning_rate": 5.405481072206917e-05, "loss": 0.0129, "step": 10010 }, { "grad_norm": 0.2783951461315155, "learning_rate": 5.397240402940402e-05, "loss": 0.011, "step": 10020 }, { "grad_norm": 0.31021326780319214, "learning_rate": 5.388998647633794e-05, "loss": 0.0111, "step": 10030 }, { "grad_norm": 0.3945060074329376, "learning_rate": 5.380755828819737e-05, "loss": 0.0131, "step": 10040 }, { "grad_norm": 0.28576409816741943, "learning_rate": 5.3725119690337846e-05, "loss": 0.0096, "step": 10050 }, { "grad_norm": 0.36083370447158813, "learning_rate": 5.3642670908143324e-05, "loss": 0.0129, "step": 10060 }, { "grad_norm": 0.3521101772785187, "learning_rate": 5.356021216702562e-05, "loss": 0.0133, "step": 10070 }, { "grad_norm": 0.2402394413948059, "learning_rate": 5.347774369242381e-05, "loss": 0.0115, "step": 10080 }, { "grad_norm": 0.23048235476016998, "learning_rate": 5.3395265709803545e-05, "loss": 0.01, "step": 10090 }, { "grad_norm": 0.2619010806083679, "learning_rate": 5.331277844465647e-05, "loss": 0.0103, "step": 10100 }, { "grad_norm": 0.2945280075073242, "learning_rate": 5.323028212249963e-05, "loss": 0.0106, "step": 10110 }, { "grad_norm": 0.39543095231056213, "learning_rate": 5.314777696887481e-05, "loss": 0.0123, "step": 10120 }, { "grad_norm": 0.3344680964946747, "learning_rate": 5.306526320934796e-05, "loss": 0.0111, "step": 10130 }, { "grad_norm": 0.2581799626350403, "learning_rate": 5.298274106950854e-05, "loss": 0.0095, "step": 10140 }, { "grad_norm": 0.2766934931278229, "learning_rate": 5.290021077496893e-05, "loss": 0.0096, "step": 10150 }, { "grad_norm": 0.27712833881378174, "learning_rate": 5.2817672551363816e-05, "loss": 0.009, "step": 10160 }, { "grad_norm": 0.2934103012084961, "learning_rate": 5.273512662434952e-05, "loss": 0.0117, "step": 10170 }, { "grad_norm": 0.3500294089317322, "learning_rate": 5.265257321960349e-05, "loss": 0.0099, "step": 10180 }, { "grad_norm": 0.312491238117218, "learning_rate": 5.257001256282357e-05, "loss": 0.0106, "step": 10190 }, { "grad_norm": 0.3397667407989502, "learning_rate": 5.248744487972742e-05, "loss": 0.0106, "step": 10200 }, { "grad_norm": 0.24076667428016663, "learning_rate": 5.240487039605196e-05, "loss": 0.0115, "step": 10210 }, { "grad_norm": 0.2931477427482605, "learning_rate": 5.232228933755267e-05, "loss": 0.0108, "step": 10220 }, { "grad_norm": 0.2667030394077301, "learning_rate": 5.2239701930003006e-05, "loss": 0.0106, "step": 10230 }, { "grad_norm": 0.33440306782722473, "learning_rate": 5.215710839919379e-05, "loss": 0.0091, "step": 10240 }, { "grad_norm": 0.39789000153541565, "learning_rate": 5.207450897093257e-05, "loss": 0.01, "step": 10250 }, { "grad_norm": 0.27144238352775574, "learning_rate": 5.1991903871043046e-05, "loss": 0.0122, "step": 10260 }, { "grad_norm": 0.2558620572090149, "learning_rate": 5.190929332536439e-05, "loss": 0.0098, "step": 10270 }, { "grad_norm": 0.30597659945487976, "learning_rate": 5.182667755975071e-05, "loss": 0.0149, "step": 10280 }, { "grad_norm": 0.31705403327941895, "learning_rate": 5.1744056800070315e-05, "loss": 0.0119, "step": 10290 }, { "grad_norm": 0.2605944871902466, "learning_rate": 5.166143127220524e-05, "loss": 0.0093, "step": 10300 }, { "grad_norm": 0.2419065237045288, "learning_rate": 5.1578801202050485e-05, "loss": 0.01, "step": 10310 }, { "grad_norm": 0.28380969166755676, "learning_rate": 5.149616681551355e-05, "loss": 0.0128, "step": 10320 }, { "grad_norm": 0.2639677822589874, "learning_rate": 5.141352833851367e-05, "loss": 0.0092, "step": 10330 }, { "grad_norm": 0.27700352668762207, "learning_rate": 5.1330885996981285e-05, "loss": 0.009, "step": 10340 }, { "grad_norm": 0.3262895941734314, "learning_rate": 5.124824001685741e-05, "loss": 0.0093, "step": 10350 }, { "grad_norm": 0.1890796571969986, "learning_rate": 5.116559062409298e-05, "loss": 0.0068, "step": 10360 }, { "grad_norm": 0.35798323154449463, "learning_rate": 5.10829380446483e-05, "loss": 0.011, "step": 10370 }, { "grad_norm": 0.45386284589767456, "learning_rate": 5.100028250449235e-05, "loss": 0.01, "step": 10380 }, { "grad_norm": 0.29239875078201294, "learning_rate": 5.0917624229602234e-05, "loss": 0.0092, "step": 10390 }, { "grad_norm": 0.2647804319858551, "learning_rate": 5.0834963445962524e-05, "loss": 0.0127, "step": 10400 }, { "grad_norm": 0.210999995470047, "learning_rate": 5.075230037956461e-05, "loss": 0.0102, "step": 10410 }, { "grad_norm": 0.2932009696960449, "learning_rate": 5.0669635256406213e-05, "loss": 0.0106, "step": 10420 }, { "grad_norm": 0.29658085107803345, "learning_rate": 5.058696830249058e-05, "loss": 0.0101, "step": 10430 }, { "grad_norm": 0.2942424714565277, "learning_rate": 5.050429974382602e-05, "loss": 0.0109, "step": 10440 }, { "grad_norm": 0.26298585534095764, "learning_rate": 5.042162980642523e-05, "loss": 0.01, "step": 10450 }, { "grad_norm": 0.27998000383377075, "learning_rate": 5.033895871630462e-05, "loss": 0.0085, "step": 10460 }, { "grad_norm": 0.2950718402862549, "learning_rate": 5.025628669948386e-05, "loss": 0.0107, "step": 10470 }, { "grad_norm": 0.23349003493785858, "learning_rate": 5.017361398198502e-05, "loss": 0.0091, "step": 10480 }, { "grad_norm": 0.29305344820022583, "learning_rate": 5.009094078983221e-05, "loss": 0.0107, "step": 10490 }, { "grad_norm": 0.22409258782863617, "learning_rate": 5.000826734905073e-05, "loss": 0.0095, "step": 10500 }, { "grad_norm": 0.2693020701408386, "learning_rate": 4.9925593885666645e-05, "loss": 0.0121, "step": 10510 }, { "grad_norm": 0.31766942143440247, "learning_rate": 4.984292062570602e-05, "loss": 0.0112, "step": 10520 }, { "grad_norm": 0.3130606710910797, "learning_rate": 4.976024779519442e-05, "loss": 0.0117, "step": 10530 }, { "grad_norm": 0.32871875166893005, "learning_rate": 4.9677575620156194e-05, "loss": 0.0086, "step": 10540 }, { "grad_norm": 0.29740282893180847, "learning_rate": 4.959490432661391e-05, "loss": 0.0077, "step": 10550 }, { "grad_norm": 0.27452680468559265, "learning_rate": 4.9512234140587726e-05, "loss": 0.0095, "step": 10560 }, { "grad_norm": 0.30771225690841675, "learning_rate": 4.942956528809477e-05, "loss": 0.0091, "step": 10570 }, { "grad_norm": 0.37801772356033325, "learning_rate": 4.934689799514854e-05, "loss": 0.0081, "step": 10580 }, { "grad_norm": 0.33152902126312256, "learning_rate": 4.926423248775827e-05, "loss": 0.0105, "step": 10590 }, { "grad_norm": 0.20458362996578217, "learning_rate": 4.918156899192826e-05, "loss": 0.0088, "step": 10600 }, { "grad_norm": 0.2844529151916504, "learning_rate": 4.909890773365738e-05, "loss": 0.009, "step": 10610 }, { "grad_norm": 0.22535687685012817, "learning_rate": 4.9016248938938344e-05, "loss": 0.0085, "step": 10620 }, { "grad_norm": 0.34423062205314636, "learning_rate": 4.8933592833757156e-05, "loss": 0.0107, "step": 10630 }, { "grad_norm": 0.26987671852111816, "learning_rate": 4.8850939644092435e-05, "loss": 0.0084, "step": 10640 }, { "grad_norm": 0.2494860142469406, "learning_rate": 4.876828959591485e-05, "loss": 0.0088, "step": 10650 }, { "grad_norm": 0.21101397275924683, "learning_rate": 4.8685642915186474e-05, "loss": 0.0108, "step": 10660 }, { "grad_norm": 0.25187385082244873, "learning_rate": 4.860299982786018e-05, "loss": 0.0091, "step": 10670 }, { "grad_norm": 0.3421100378036499, "learning_rate": 4.852036055987901e-05, "loss": 0.0133, "step": 10680 }, { "grad_norm": 0.3339287340641022, "learning_rate": 4.843772533717558e-05, "loss": 0.0089, "step": 10690 }, { "grad_norm": 0.277946412563324, "learning_rate": 4.835509438567142e-05, "loss": 0.0093, "step": 10700 }, { "grad_norm": 0.41331660747528076, "learning_rate": 4.827246793127639e-05, "loss": 0.0128, "step": 10710 }, { "grad_norm": 0.33479300141334534, "learning_rate": 4.818984619988807e-05, "loss": 0.0085, "step": 10720 }, { "grad_norm": 0.26021715998649597, "learning_rate": 4.810722941739115e-05, "loss": 0.0085, "step": 10730 }, { "grad_norm": 0.26176100969314575, "learning_rate": 4.8024617809656684e-05, "loss": 0.0113, "step": 10740 }, { "grad_norm": 0.26029080152511597, "learning_rate": 4.794201160254171e-05, "loss": 0.0091, "step": 10750 }, { "grad_norm": 0.31911373138427734, "learning_rate": 4.785941102188844e-05, "loss": 0.0094, "step": 10760 }, { "grad_norm": 0.294737309217453, "learning_rate": 4.7776816293523686e-05, "loss": 0.0091, "step": 10770 }, { "grad_norm": 0.28514543175697327, "learning_rate": 4.769422764325832e-05, "loss": 0.0083, "step": 10780 }, { "grad_norm": 0.4033351540565491, "learning_rate": 4.76116452968865e-05, "loss": 0.012, "step": 10790 }, { "grad_norm": 0.33696362376213074, "learning_rate": 4.752906948018525e-05, "loss": 0.0094, "step": 10800 }, { "grad_norm": 0.2613891065120697, "learning_rate": 4.7446500418913684e-05, "loss": 0.0083, "step": 10810 }, { "grad_norm": 0.27233195304870605, "learning_rate": 4.736393833881247e-05, "loss": 0.0115, "step": 10820 }, { "grad_norm": 0.37831413745880127, "learning_rate": 4.7281383465603194e-05, "loss": 0.0103, "step": 10830 }, { "grad_norm": 0.32709556818008423, "learning_rate": 4.71988360249877e-05, "loss": 0.0094, "step": 10840 }, { "grad_norm": 0.19801153242588043, "learning_rate": 4.7116296242647554e-05, "loss": 0.009, "step": 10850 }, { "grad_norm": 0.2524295449256897, "learning_rate": 4.703376434424336e-05, "loss": 0.0085, "step": 10860 }, { "grad_norm": 0.2975400686264038, "learning_rate": 4.695124055541421e-05, "loss": 0.0082, "step": 10870 }, { "grad_norm": 0.3047940135002136, "learning_rate": 4.6868725101776934e-05, "loss": 0.009, "step": 10880 }, { "grad_norm": 0.39293280243873596, "learning_rate": 4.678621820892567e-05, "loss": 0.0082, "step": 10890 }, { "grad_norm": 0.27886682748794556, "learning_rate": 4.670372010243111e-05, "loss": 0.0092, "step": 10900 }, { "grad_norm": 0.3043139576911926, "learning_rate": 4.662123100783992e-05, "loss": 0.0088, "step": 10910 }, { "grad_norm": 0.2170037478208542, "learning_rate": 4.653875115067415e-05, "loss": 0.0083, "step": 10920 }, { "grad_norm": 0.2518364489078522, "learning_rate": 4.6456280756430545e-05, "loss": 0.0104, "step": 10930 }, { "grad_norm": 0.29944872856140137, "learning_rate": 4.637382005058004e-05, "loss": 0.0077, "step": 10940 }, { "grad_norm": 0.28369656205177307, "learning_rate": 4.629136925856705e-05, "loss": 0.0095, "step": 10950 }, { "grad_norm": 0.31047528982162476, "learning_rate": 4.6208928605808895e-05, "loss": 0.0079, "step": 10960 }, { "grad_norm": 0.21829445660114288, "learning_rate": 4.612649831769519e-05, "loss": 0.0084, "step": 10970 }, { "grad_norm": 0.3248969614505768, "learning_rate": 4.604407861958715e-05, "loss": 0.0096, "step": 10980 }, { "grad_norm": 0.30876240134239197, "learning_rate": 4.5961669736817114e-05, "loss": 0.0119, "step": 10990 }, { "grad_norm": 0.3592730760574341, "learning_rate": 4.5879271894687814e-05, "loss": 0.0136, "step": 11000 }, { "grad_norm": 0.23764269053936005, "learning_rate": 4.5796885318471826e-05, "loss": 0.0089, "step": 11010 }, { "grad_norm": 0.2311507910490036, "learning_rate": 4.571451023341086e-05, "loss": 0.0088, "step": 11020 }, { "grad_norm": 0.21218560636043549, "learning_rate": 4.563214686471527e-05, "loss": 0.0106, "step": 11030 }, { "grad_norm": 0.27804577350616455, "learning_rate": 4.5549795437563365e-05, "loss": 0.0082, "step": 11040 }, { "grad_norm": 0.26429250836372375, "learning_rate": 4.546745617710081e-05, "loss": 0.0092, "step": 11050 }, { "grad_norm": 0.25713181495666504, "learning_rate": 4.5385129308440014e-05, "loss": 0.0071, "step": 11060 }, { "grad_norm": 0.2942948341369629, "learning_rate": 4.530281505665944e-05, "loss": 0.011, "step": 11070 }, { "grad_norm": 0.2995373010635376, "learning_rate": 4.5220513646803134e-05, "loss": 0.0109, "step": 11080 }, { "grad_norm": 0.3345450162887573, "learning_rate": 4.513822530388003e-05, "loss": 0.0112, "step": 11090 }, { "grad_norm": 0.40226829051971436, "learning_rate": 4.5055950252863296e-05, "loss": 0.0104, "step": 11100 }, { "grad_norm": 0.3961140215396881, "learning_rate": 4.4973688718689803e-05, "loss": 0.0136, "step": 11110 }, { "grad_norm": 0.418488085269928, "learning_rate": 4.4891440926259406e-05, "loss": 0.0108, "step": 11120 }, { "grad_norm": 0.2921876311302185, "learning_rate": 4.480920710043443e-05, "loss": 0.0093, "step": 11130 }, { "grad_norm": 0.26075279712677, "learning_rate": 4.4726987466039044e-05, "loss": 0.0126, "step": 11140 }, { "grad_norm": 0.22061918675899506, "learning_rate": 4.46447822478586e-05, "loss": 0.0149, "step": 11150 }, { "grad_norm": 0.35924506187438965, "learning_rate": 4.4562591670638974e-05, "loss": 0.0134, "step": 11160 }, { "grad_norm": 0.4507744312286377, "learning_rate": 4.4480415959086105e-05, "loss": 0.0118, "step": 11170 }, { "grad_norm": 0.33580678701400757, "learning_rate": 4.439825533786522e-05, "loss": 0.0101, "step": 11180 }, { "grad_norm": 0.36705753207206726, "learning_rate": 4.431611003160035e-05, "loss": 0.0141, "step": 11190 }, { "grad_norm": 0.3142641484737396, "learning_rate": 4.4233980264873636e-05, "loss": 0.0138, "step": 11200 }, { "grad_norm": 0.3167910873889923, "learning_rate": 4.4151866262224684e-05, "loss": 0.0132, "step": 11210 }, { "grad_norm": 0.28718140721321106, "learning_rate": 4.406976824815006e-05, "loss": 0.0088, "step": 11220 }, { "grad_norm": 0.2741796374320984, "learning_rate": 4.3987686447102595e-05, "loss": 0.0111, "step": 11230 }, { "grad_norm": 0.24285542964935303, "learning_rate": 4.3905621083490804e-05, "loss": 0.0083, "step": 11240 }, { "grad_norm": 0.3054063022136688, "learning_rate": 4.3823572381678286e-05, "loss": 0.0122, "step": 11250 }, { "grad_norm": 0.32679957151412964, "learning_rate": 4.374154056598301e-05, "loss": 0.008, "step": 11260 }, { "grad_norm": 0.28160157799720764, "learning_rate": 4.3659525860676845e-05, "loss": 0.0084, "step": 11270 }, { "grad_norm": 0.19414947926998138, "learning_rate": 4.3577528489984854e-05, "loss": 0.0083, "step": 11280 }, { "grad_norm": 0.31703388690948486, "learning_rate": 4.349554867808476e-05, "loss": 0.0097, "step": 11290 }, { "grad_norm": 0.23897582292556763, "learning_rate": 4.34135866491062e-05, "loss": 0.0093, "step": 11300 }, { "grad_norm": 0.23693439364433289, "learning_rate": 4.333164262713022e-05, "loss": 0.0091, "step": 11310 }, { "grad_norm": 0.1998605728149414, "learning_rate": 4.324971683618868e-05, "loss": 0.0074, "step": 11320 }, { "grad_norm": 0.2250213921070099, "learning_rate": 4.316780950026354e-05, "loss": 0.0113, "step": 11330 }, { "grad_norm": 0.3492794930934906, "learning_rate": 4.308592084328637e-05, "loss": 0.0111, "step": 11340 }, { "grad_norm": 0.30541759729385376, "learning_rate": 4.3004051089137576e-05, "loss": 0.0105, "step": 11350 }, { "grad_norm": 0.27780118584632874, "learning_rate": 4.292220046164597e-05, "loss": 0.0088, "step": 11360 }, { "grad_norm": 0.34747496247291565, "learning_rate": 4.2840369184588035e-05, "loss": 0.0132, "step": 11370 }, { "grad_norm": 0.3092314302921295, "learning_rate": 4.2758557481687345e-05, "loss": 0.0087, "step": 11380 }, { "grad_norm": 0.24002385139465332, "learning_rate": 4.267676557661403e-05, "loss": 0.0081, "step": 11390 }, { "grad_norm": 0.26428428292274475, "learning_rate": 4.2594993692983955e-05, "loss": 0.0081, "step": 11400 }, { "grad_norm": 0.3460390865802765, "learning_rate": 4.251324205435837e-05, "loss": 0.0102, "step": 11410 }, { "grad_norm": 0.27683359384536743, "learning_rate": 4.243151088424312e-05, "loss": 0.0121, "step": 11420 }, { "grad_norm": 0.2775872051715851, "learning_rate": 4.234980040608813e-05, "loss": 0.0104, "step": 11430 }, { "grad_norm": 0.3487337529659271, "learning_rate": 4.22681108432867e-05, "loss": 0.0091, "step": 11440 }, { "grad_norm": 0.21562695503234863, "learning_rate": 4.2186442419174984e-05, "loss": 0.0099, "step": 11450 }, { "grad_norm": 0.3062504827976227, "learning_rate": 4.210479535703133e-05, "loss": 0.0131, "step": 11460 }, { "grad_norm": 0.23322391510009766, "learning_rate": 4.202316988007567e-05, "loss": 0.0103, "step": 11470 }, { "grad_norm": 0.2752143442630768, "learning_rate": 4.194156621146901e-05, "loss": 0.0102, "step": 11480 }, { "grad_norm": 0.16598841547966003, "learning_rate": 4.1859984574312596e-05, "loss": 0.0072, "step": 11490 }, { "grad_norm": 0.2305120825767517, "learning_rate": 4.177842519164752e-05, "loss": 0.0108, "step": 11500 }, { "grad_norm": 0.28615257143974304, "learning_rate": 4.169688828645404e-05, "loss": 0.0082, "step": 11510 }, { "grad_norm": 0.21288816630840302, "learning_rate": 4.161537408165092e-05, "loss": 0.0098, "step": 11520 }, { "grad_norm": 0.2908499240875244, "learning_rate": 4.1533882800094924e-05, "loss": 0.0102, "step": 11530 }, { "grad_norm": 0.34949150681495667, "learning_rate": 4.145241466458005e-05, "loss": 0.0071, "step": 11540 }, { "grad_norm": 0.3808799088001251, "learning_rate": 4.13709698978371e-05, "loss": 0.009, "step": 11550 }, { "grad_norm": 0.26844173669815063, "learning_rate": 4.1289548722532944e-05, "loss": 0.0095, "step": 11560 }, { "grad_norm": 0.25560107827186584, "learning_rate": 4.120815136126999e-05, "loss": 0.0089, "step": 11570 }, { "grad_norm": 0.2132156640291214, "learning_rate": 4.112677803658548e-05, "loss": 0.0065, "step": 11580 }, { "grad_norm": 0.2574012279510498, "learning_rate": 4.1045428970951e-05, "loss": 0.0096, "step": 11590 }, { "grad_norm": 0.3037318289279938, "learning_rate": 4.0964104386771785e-05, "loss": 0.0077, "step": 11600 }, { "grad_norm": 0.2965768873691559, "learning_rate": 4.0882804506386144e-05, "loss": 0.0086, "step": 11610 }, { "grad_norm": 0.2514228820800781, "learning_rate": 4.080152955206485e-05, "loss": 0.0095, "step": 11620 }, { "grad_norm": 0.2959059774875641, "learning_rate": 4.0720279746010505e-05, "loss": 0.0081, "step": 11630 }, { "grad_norm": 0.28880995512008667, "learning_rate": 4.063905531035699e-05, "loss": 0.0114, "step": 11640 }, { "grad_norm": 0.2837924361228943, "learning_rate": 4.055785646716882e-05, "loss": 0.0079, "step": 11650 }, { "grad_norm": 0.2794390618801117, "learning_rate": 4.047668343844051e-05, "loss": 0.0076, "step": 11660 }, { "grad_norm": 0.32419878244400024, "learning_rate": 4.039553644609604e-05, "loss": 0.01, "step": 11670 }, { "grad_norm": 0.27437660098075867, "learning_rate": 4.0314415711988176e-05, "loss": 0.0069, "step": 11680 }, { "grad_norm": 0.19072072207927704, "learning_rate": 4.023332145789792e-05, "loss": 0.0084, "step": 11690 }, { "grad_norm": 0.30108916759490967, "learning_rate": 4.015225390553385e-05, "loss": 0.0101, "step": 11700 }, { "grad_norm": 0.27425938844680786, "learning_rate": 4.007121327653158e-05, "loss": 0.0123, "step": 11710 }, { "grad_norm": 0.19379594922065735, "learning_rate": 3.9990199792453064e-05, "loss": 0.0074, "step": 11720 }, { "grad_norm": 0.2703425884246826, "learning_rate": 3.9909213674786103e-05, "loss": 0.0085, "step": 11730 }, { "grad_norm": 0.3462633788585663, "learning_rate": 3.982825514494363e-05, "loss": 0.0077, "step": 11740 }, { "grad_norm": 0.3126048147678375, "learning_rate": 3.974732442426319e-05, "loss": 0.0078, "step": 11750 }, { "grad_norm": 0.25300177931785583, "learning_rate": 3.966642173400629e-05, "loss": 0.0091, "step": 11760 }, { "grad_norm": 0.2746008336544037, "learning_rate": 3.9585547295357764e-05, "loss": 0.0061, "step": 11770 }, { "grad_norm": 0.2269902527332306, "learning_rate": 3.950470132942526e-05, "loss": 0.0063, "step": 11780 }, { "grad_norm": 0.23431739211082458, "learning_rate": 3.942388405723856e-05, "loss": 0.0081, "step": 11790 }, { "grad_norm": 0.1898859292268753, "learning_rate": 3.9343095699749e-05, "loss": 0.0073, "step": 11800 }, { "grad_norm": 0.28582820296287537, "learning_rate": 3.9262336477828874e-05, "loss": 0.0072, "step": 11810 }, { "grad_norm": 0.2513090670108795, "learning_rate": 3.9181606612270794e-05, "loss": 0.01, "step": 11820 }, { "grad_norm": 0.27929598093032837, "learning_rate": 3.910090632378713e-05, "loss": 0.0078, "step": 11830 }, { "grad_norm": 0.29875457286834717, "learning_rate": 3.90202358330094e-05, "loss": 0.0067, "step": 11840 }, { "grad_norm": 0.1761026531457901, "learning_rate": 3.8939595360487656e-05, "loss": 0.0107, "step": 11850 }, { "grad_norm": 0.2306380569934845, "learning_rate": 3.885898512668984e-05, "loss": 0.0106, "step": 11860 }, { "grad_norm": 0.21686899662017822, "learning_rate": 3.877840535200127e-05, "loss": 0.0114, "step": 11870 }, { "grad_norm": 0.34274980425834656, "learning_rate": 3.869785625672397e-05, "loss": 0.0094, "step": 11880 }, { "grad_norm": 0.4146466851234436, "learning_rate": 3.8617338061076094e-05, "loss": 0.0112, "step": 11890 }, { "grad_norm": 0.2690848410129547, "learning_rate": 3.853685098519132e-05, "loss": 0.0122, "step": 11900 }, { "grad_norm": 0.3507445156574249, "learning_rate": 3.845639524911823e-05, "loss": 0.0099, "step": 11910 }, { "grad_norm": 0.27721738815307617, "learning_rate": 3.837597107281974e-05, "loss": 0.0083, "step": 11920 }, { "grad_norm": 0.23185427486896515, "learning_rate": 3.829557867617247e-05, "loss": 0.0073, "step": 11930 }, { "grad_norm": 0.22188237309455872, "learning_rate": 3.821521827896618e-05, "loss": 0.0091, "step": 11940 }, { "grad_norm": 0.3477785885334015, "learning_rate": 3.81348901009031e-05, "loss": 0.0086, "step": 11950 }, { "grad_norm": 0.21982355415821075, "learning_rate": 3.805459436159741e-05, "loss": 0.0096, "step": 11960 }, { "grad_norm": 0.264242023229599, "learning_rate": 3.797433128057461e-05, "loss": 0.0107, "step": 11970 }, { "grad_norm": 0.31478801369667053, "learning_rate": 3.789410107727089e-05, "loss": 0.0074, "step": 11980 }, { "grad_norm": 0.24611283838748932, "learning_rate": 3.781390397103257e-05, "loss": 0.0082, "step": 11990 }, { "grad_norm": 0.1611613631248474, "learning_rate": 3.7733740181115455e-05, "loss": 0.0069, "step": 12000 }, { "grad_norm": 0.28428375720977783, "learning_rate": 3.7653609926684306e-05, "loss": 0.0089, "step": 12010 }, { "grad_norm": 0.33680009841918945, "learning_rate": 3.757351342681217e-05, "loss": 0.0108, "step": 12020 }, { "grad_norm": 0.3011215925216675, "learning_rate": 3.749345090047982e-05, "loss": 0.0096, "step": 12030 }, { "grad_norm": 0.20513713359832764, "learning_rate": 3.741342256657515e-05, "loss": 0.0075, "step": 12040 }, { "grad_norm": 0.2366105169057846, "learning_rate": 3.7333428643892567e-05, "loss": 0.0094, "step": 12050 }, { "grad_norm": 0.2522220015525818, "learning_rate": 3.725346935113239e-05, "loss": 0.0105, "step": 12060 }, { "grad_norm": 0.34309864044189453, "learning_rate": 3.717354490690029e-05, "loss": 0.0081, "step": 12070 }, { "grad_norm": 0.31223011016845703, "learning_rate": 3.709365552970664e-05, "loss": 0.0073, "step": 12080 }, { "grad_norm": 0.2302858531475067, "learning_rate": 3.7013801437965945e-05, "loss": 0.0066, "step": 12090 }, { "grad_norm": 0.3095393776893616, "learning_rate": 3.693398284999623e-05, "loss": 0.0096, "step": 12100 }, { "grad_norm": 0.293329656124115, "learning_rate": 3.6854199984018484e-05, "loss": 0.0078, "step": 12110 }, { "grad_norm": 0.3112761080265045, "learning_rate": 3.677445305815601e-05, "loss": 0.0088, "step": 12120 }, { "grad_norm": 0.28105172514915466, "learning_rate": 3.669474229043387e-05, "loss": 0.0073, "step": 12130 }, { "grad_norm": 0.21714749932289124, "learning_rate": 3.6615067898778235e-05, "loss": 0.0081, "step": 12140 }, { "grad_norm": 0.2943991720676422, "learning_rate": 3.6535430101015866e-05, "loss": 0.0084, "step": 12150 }, { "grad_norm": 0.28031906485557556, "learning_rate": 3.645582911487345e-05, "loss": 0.0141, "step": 12160 }, { "grad_norm": 0.2581537663936615, "learning_rate": 3.637626515797706e-05, "loss": 0.0072, "step": 12170 }, { "grad_norm": 0.1806012988090515, "learning_rate": 3.629673844785152e-05, "loss": 0.0099, "step": 12180 }, { "grad_norm": 0.2714550197124481, "learning_rate": 3.621724920191979e-05, "loss": 0.0102, "step": 12190 }, { "grad_norm": 0.29836222529411316, "learning_rate": 3.6137797637502444e-05, "loss": 0.0059, "step": 12200 }, { "grad_norm": 0.4245578646659851, "learning_rate": 3.6058383971817035e-05, "loss": 0.0099, "step": 12210 }, { "grad_norm": 0.22129003703594208, "learning_rate": 3.59790084219775e-05, "loss": 0.0082, "step": 12220 }, { "grad_norm": 0.35264456272125244, "learning_rate": 3.589967120499353e-05, "loss": 0.0106, "step": 12230 }, { "grad_norm": 0.27610617876052856, "learning_rate": 3.5820372537770075e-05, "loss": 0.0074, "step": 12240 }, { "grad_norm": 0.24030449986457825, "learning_rate": 3.5741112637106655e-05, "loss": 0.008, "step": 12250 }, { "grad_norm": 0.34359195828437805, "learning_rate": 3.5661891719696804e-05, "loss": 0.0091, "step": 12260 }, { "grad_norm": 0.2971857488155365, "learning_rate": 3.5582710002127504e-05, "loss": 0.0089, "step": 12270 }, { "grad_norm": 0.3562963604927063, "learning_rate": 3.550356770087853e-05, "loss": 0.0096, "step": 12280 }, { "grad_norm": 0.2590996325016022, "learning_rate": 3.5424465032321914e-05, "loss": 0.0071, "step": 12290 }, { "grad_norm": 0.2519053816795349, "learning_rate": 3.5345402212721335e-05, "loss": 0.0066, "step": 12300 }, { "grad_norm": 0.2250411957502365, "learning_rate": 3.526637945823152e-05, "loss": 0.0076, "step": 12310 }, { "grad_norm": 0.2911670207977295, "learning_rate": 3.518739698489767e-05, "loss": 0.0073, "step": 12320 }, { "grad_norm": 0.22276492416858673, "learning_rate": 3.510845500865485e-05, "loss": 0.0065, "step": 12330 }, { "grad_norm": 0.28191274404525757, "learning_rate": 3.502955374532739e-05, "loss": 0.0067, "step": 12340 }, { "grad_norm": 0.210043802857399, "learning_rate": 3.495069341062836e-05, "loss": 0.0075, "step": 12350 }, { "grad_norm": 0.19946876168251038, "learning_rate": 3.4871874220158896e-05, "loss": 0.0061, "step": 12360 }, { "grad_norm": 0.23715545237064362, "learning_rate": 3.479309638940762e-05, "loss": 0.0054, "step": 12370 }, { "grad_norm": 0.26544538140296936, "learning_rate": 3.4714360133750146e-05, "loss": 0.01, "step": 12380 }, { "grad_norm": 0.16757625341415405, "learning_rate": 3.463566566844839e-05, "loss": 0.0067, "step": 12390 }, { "grad_norm": 0.32589516043663025, "learning_rate": 3.4557013208650016e-05, "loss": 0.0091, "step": 12400 }, { "grad_norm": 0.3588258922100067, "learning_rate": 3.4478402969387857e-05, "loss": 0.0088, "step": 12410 }, { "grad_norm": 0.2870354950428009, "learning_rate": 3.4399835165579266e-05, "loss": 0.0089, "step": 12420 }, { "grad_norm": 0.27898696064949036, "learning_rate": 3.4321310012025645e-05, "loss": 0.0091, "step": 12430 }, { "grad_norm": 0.2062765657901764, "learning_rate": 3.424282772341176e-05, "loss": 0.0072, "step": 12440 }, { "grad_norm": 0.20485492050647736, "learning_rate": 3.416438851430519e-05, "loss": 0.0059, "step": 12450 }, { "grad_norm": 0.33573809266090393, "learning_rate": 3.408599259915577e-05, "loss": 0.0083, "step": 12460 }, { "grad_norm": 0.2708123028278351, "learning_rate": 3.400764019229487e-05, "loss": 0.0071, "step": 12470 }, { "grad_norm": 0.34195250272750854, "learning_rate": 3.3929331507935035e-05, "loss": 0.0097, "step": 12480 }, { "grad_norm": 0.24106758832931519, "learning_rate": 3.3851066760169196e-05, "loss": 0.0081, "step": 12490 }, { "grad_norm": 0.3639850914478302, "learning_rate": 3.377284616297021e-05, "loss": 0.007, "step": 12500 }, { "grad_norm": 0.21595130860805511, "learning_rate": 3.3694669930190166e-05, "loss": 0.0063, "step": 12510 }, { "grad_norm": 0.2641875743865967, "learning_rate": 3.36165382755599e-05, "loss": 0.0068, "step": 12520 }, { "grad_norm": 0.1990833431482315, "learning_rate": 3.35384514126884e-05, "loss": 0.007, "step": 12530 }, { "grad_norm": 0.24879232048988342, "learning_rate": 3.3460409555062154e-05, "loss": 0.0062, "step": 12540 }, { "grad_norm": 0.24407650530338287, "learning_rate": 3.3382412916044645e-05, "loss": 0.0083, "step": 12550 }, { "grad_norm": 0.20692312717437744, "learning_rate": 3.330446170887566e-05, "loss": 0.0056, "step": 12560 }, { "grad_norm": 0.31624099612236023, "learning_rate": 3.3226556146670834e-05, "loss": 0.009, "step": 12570 }, { "grad_norm": 0.2939915060997009, "learning_rate": 3.314869644242102e-05, "loss": 0.007, "step": 12580 }, { "grad_norm": 0.30532944202423096, "learning_rate": 3.3070882808991674e-05, "loss": 0.0067, "step": 12590 }, { "grad_norm": 0.3181285858154297, "learning_rate": 3.2993115459122305e-05, "loss": 0.0075, "step": 12600 }, { "grad_norm": 0.2092517912387848, "learning_rate": 3.2915394605425835e-05, "loss": 0.0072, "step": 12610 }, { "grad_norm": 0.20576176047325134, "learning_rate": 3.283772046038816e-05, "loss": 0.0068, "step": 12620 }, { "grad_norm": 0.36213627457618713, "learning_rate": 3.276009323636739e-05, "loss": 0.0089, "step": 12630 }, { "grad_norm": 0.2462421953678131, "learning_rate": 3.268251314559344e-05, "loss": 0.0063, "step": 12640 }, { "grad_norm": 0.26896142959594727, "learning_rate": 3.2604980400167254e-05, "loss": 0.0067, "step": 12650 }, { "grad_norm": 0.2979550361633301, "learning_rate": 3.252749521206042e-05, "loss": 0.0079, "step": 12660 }, { "grad_norm": 0.35195228457450867, "learning_rate": 3.2450057793114494e-05, "loss": 0.0087, "step": 12670 }, { "grad_norm": 0.26797011494636536, "learning_rate": 3.2372668355040435e-05, "loss": 0.0057, "step": 12680 }, { "grad_norm": 0.2521347105503082, "learning_rate": 3.2295327109418005e-05, "loss": 0.0099, "step": 12690 }, { "grad_norm": 0.379652738571167, "learning_rate": 3.221803426769518e-05, "loss": 0.0108, "step": 12700 }, { "grad_norm": 0.23067383468151093, "learning_rate": 3.214079004118768e-05, "loss": 0.0057, "step": 12710 }, { "grad_norm": 0.2105378657579422, "learning_rate": 3.2063594641078234e-05, "loss": 0.0077, "step": 12720 }, { "grad_norm": 0.20356619358062744, "learning_rate": 3.198644827841616e-05, "loss": 0.0065, "step": 12730 }, { "grad_norm": 0.17580711841583252, "learning_rate": 3.1909351164116654e-05, "loss": 0.0075, "step": 12740 }, { "grad_norm": 0.2302006334066391, "learning_rate": 3.183230350896026e-05, "loss": 0.0067, "step": 12750 }, { "grad_norm": 0.19220423698425293, "learning_rate": 3.1755305523592337e-05, "loss": 0.0077, "step": 12760 }, { "grad_norm": 0.21260245144367218, "learning_rate": 3.167835741852245e-05, "loss": 0.0066, "step": 12770 }, { "grad_norm": 0.31843772530555725, "learning_rate": 3.160145940412378e-05, "loss": 0.0111, "step": 12780 }, { "grad_norm": 0.2688577175140381, "learning_rate": 3.1524611690632545e-05, "loss": 0.0099, "step": 12790 }, { "grad_norm": 0.2487054467201233, "learning_rate": 3.144781448814746e-05, "loss": 0.0094, "step": 12800 }, { "grad_norm": 0.23169320821762085, "learning_rate": 3.1371068006629145e-05, "loss": 0.0092, "step": 12810 }, { "grad_norm": 0.27694079279899597, "learning_rate": 3.129437245589956e-05, "loss": 0.0082, "step": 12820 }, { "grad_norm": 0.2721058428287506, "learning_rate": 3.121772804564143e-05, "loss": 0.0077, "step": 12830 }, { "grad_norm": 0.22766993939876556, "learning_rate": 3.11411349853976e-05, "loss": 0.0073, "step": 12840 }, { "grad_norm": 0.26015397906303406, "learning_rate": 3.10645934845706e-05, "loss": 0.0058, "step": 12850 }, { "grad_norm": 0.18540212512016296, "learning_rate": 3.098810375242196e-05, "loss": 0.0066, "step": 12860 }, { "grad_norm": 0.31042733788490295, "learning_rate": 3.0911665998071704e-05, "loss": 0.0079, "step": 12870 }, { "grad_norm": 0.2469450831413269, "learning_rate": 3.083528043049774e-05, "loss": 0.0082, "step": 12880 }, { "grad_norm": 0.22905239462852478, "learning_rate": 3.0758947258535255e-05, "loss": 0.0067, "step": 12890 }, { "grad_norm": 0.20388999581336975, "learning_rate": 3.068266669087625e-05, "loss": 0.0068, "step": 12900 }, { "grad_norm": 0.27276334166526794, "learning_rate": 3.060643893606887e-05, "loss": 0.0066, "step": 12910 }, { "grad_norm": 0.28945600986480713, "learning_rate": 3.053026420251693e-05, "loss": 0.0081, "step": 12920 }, { "grad_norm": 0.2451259344816208, "learning_rate": 3.0454142698479183e-05, "loss": 0.0068, "step": 12930 }, { "grad_norm": 0.21257218718528748, "learning_rate": 3.0378074632068954e-05, "loss": 0.0051, "step": 12940 }, { "grad_norm": 0.192915141582489, "learning_rate": 3.0302060211253408e-05, "loss": 0.0063, "step": 12950 }, { "grad_norm": 0.2408614456653595, "learning_rate": 3.0226099643853073e-05, "loss": 0.007, "step": 12960 }, { "grad_norm": 0.2703664302825928, "learning_rate": 3.0150193137541283e-05, "loss": 0.0077, "step": 12970 }, { "grad_norm": 0.260464072227478, "learning_rate": 3.0074340899843467e-05, "loss": 0.0088, "step": 12980 }, { "grad_norm": 0.2323172241449356, "learning_rate": 2.999854313813677e-05, "loss": 0.0077, "step": 12990 }, { "grad_norm": 0.28334948420524597, "learning_rate": 2.9922800059649382e-05, "loss": 0.0076, "step": 13000 }, { "grad_norm": 0.24106933176517487, "learning_rate": 2.9847111871459976e-05, "loss": 0.0104, "step": 13010 }, { "grad_norm": 0.36025455594062805, "learning_rate": 2.977147878049721e-05, "loss": 0.0081, "step": 13020 }, { "grad_norm": 0.2709331214427948, "learning_rate": 2.9695900993539006e-05, "loss": 0.0073, "step": 13030 }, { "grad_norm": 0.24559737741947174, "learning_rate": 2.9620378717212183e-05, "loss": 0.007, "step": 13040 }, { "grad_norm": 0.18448466062545776, "learning_rate": 2.9544912157991745e-05, "loss": 0.0069, "step": 13050 }, { "grad_norm": 0.21023514866828918, "learning_rate": 2.9469501522200405e-05, "loss": 0.0074, "step": 13060 }, { "grad_norm": 0.18496470153331757, "learning_rate": 2.9394147016007946e-05, "loss": 0.0046, "step": 13070 }, { "grad_norm": 0.2527117133140564, "learning_rate": 2.9318848845430702e-05, "loss": 0.0055, "step": 13080 }, { "grad_norm": 0.1908368170261383, "learning_rate": 2.9243607216331013e-05, "loss": 0.008, "step": 13090 }, { "grad_norm": 0.35127463936805725, "learning_rate": 2.916842233441661e-05, "loss": 0.0076, "step": 13100 }, { "grad_norm": 0.28152093291282654, "learning_rate": 2.90932944052401e-05, "loss": 0.0065, "step": 13110 }, { "grad_norm": 0.4291995167732239, "learning_rate": 2.9018223634198354e-05, "loss": 0.0083, "step": 13120 }, { "grad_norm": 0.21902692317962646, "learning_rate": 2.8943210226532025e-05, "loss": 0.0097, "step": 13130 }, { "grad_norm": 0.20828410983085632, "learning_rate": 2.8868254387324857e-05, "loss": 0.0064, "step": 13140 }, { "grad_norm": 0.2285706102848053, "learning_rate": 2.8793356321503306e-05, "loss": 0.0063, "step": 13150 }, { "grad_norm": 0.16433380544185638, "learning_rate": 2.87185162338358e-05, "loss": 0.005, "step": 13160 }, { "grad_norm": 0.24990630149841309, "learning_rate": 2.8643734328932253e-05, "loss": 0.0078, "step": 13170 }, { "grad_norm": 0.22981254756450653, "learning_rate": 2.856901081124359e-05, "loss": 0.0062, "step": 13180 }, { "grad_norm": 0.2027965933084488, "learning_rate": 2.8494345885061002e-05, "loss": 0.0054, "step": 13190 }, { "grad_norm": 0.24758584797382355, "learning_rate": 2.8419739754515616e-05, "loss": 0.01, "step": 13200 }, { "grad_norm": 0.2455262988805771, "learning_rate": 2.8345192623577666e-05, "loss": 0.0072, "step": 13210 }, { "grad_norm": 0.26795145869255066, "learning_rate": 2.8270704696056193e-05, "loss": 0.0077, "step": 13220 }, { "grad_norm": 0.27453330159187317, "learning_rate": 2.8196276175598367e-05, "loss": 0.0078, "step": 13230 }, { "grad_norm": 0.1935349404811859, "learning_rate": 2.8121907265688884e-05, "loss": 0.0103, "step": 13240 }, { "grad_norm": 0.19487586617469788, "learning_rate": 2.804759816964957e-05, "loss": 0.0089, "step": 13250 }, { "grad_norm": 0.28974649310112, "learning_rate": 2.797334909063857e-05, "loss": 0.01, "step": 13260 }, { "grad_norm": 0.2818593680858612, "learning_rate": 2.7899160231650056e-05, "loss": 0.0089, "step": 13270 }, { "grad_norm": 0.30783990025520325, "learning_rate": 2.7825031795513585e-05, "loss": 0.0076, "step": 13280 }, { "grad_norm": 0.19885502755641937, "learning_rate": 2.775096398489341e-05, "loss": 0.0087, "step": 13290 }, { "grad_norm": 0.28991419076919556, "learning_rate": 2.7676957002288163e-05, "loss": 0.0089, "step": 13300 }, { "grad_norm": 0.20957441627979279, "learning_rate": 2.760301105003003e-05, "loss": 0.0059, "step": 13310 }, { "grad_norm": 0.1530667245388031, "learning_rate": 2.752912633028446e-05, "loss": 0.0068, "step": 13320 }, { "grad_norm": 0.23960092663764954, "learning_rate": 2.7455303045049474e-05, "loss": 0.0082, "step": 13330 }, { "grad_norm": 0.237693652510643, "learning_rate": 2.7381541396155098e-05, "loss": 0.0067, "step": 13340 }, { "grad_norm": 0.18768084049224854, "learning_rate": 2.730784158526286e-05, "loss": 0.005, "step": 13350 }, { "grad_norm": 0.17086821794509888, "learning_rate": 2.723420381386521e-05, "loss": 0.0073, "step": 13360 }, { "grad_norm": 0.20543023943901062, "learning_rate": 2.7160628283285018e-05, "loss": 0.0064, "step": 13370 }, { "grad_norm": 0.2849728763103485, "learning_rate": 2.7087115194675007e-05, "loss": 0.007, "step": 13380 }, { "grad_norm": 0.2168434113264084, "learning_rate": 2.701366474901712e-05, "loss": 0.0085, "step": 13390 }, { "grad_norm": 0.16423042118549347, "learning_rate": 2.6940277147122085e-05, "loss": 0.0054, "step": 13400 }, { "grad_norm": 0.22179411351680756, "learning_rate": 2.686695258962878e-05, "loss": 0.006, "step": 13410 }, { "grad_norm": 0.31073468923568726, "learning_rate": 2.679369127700375e-05, "loss": 0.007, "step": 13420 }, { "grad_norm": 0.19968819618225098, "learning_rate": 2.672049340954067e-05, "loss": 0.0073, "step": 13430 }, { "grad_norm": 0.1817862093448639, "learning_rate": 2.6647359187359676e-05, "loss": 0.0105, "step": 13440 }, { "grad_norm": 0.2022973746061325, "learning_rate": 2.6574288810406946e-05, "loss": 0.0064, "step": 13450 }, { "grad_norm": 0.26418814063072205, "learning_rate": 2.6501282478454083e-05, "loss": 0.0085, "step": 13460 }, { "grad_norm": 0.1963866949081421, "learning_rate": 2.6428340391097618e-05, "loss": 0.0078, "step": 13470 }, { "grad_norm": 0.1557660698890686, "learning_rate": 2.6355462747758485e-05, "loss": 0.0056, "step": 13480 }, { "grad_norm": 0.3091756999492645, "learning_rate": 2.6282649747681304e-05, "loss": 0.0115, "step": 13490 }, { "grad_norm": 0.29082000255584717, "learning_rate": 2.620990158993406e-05, "loss": 0.0094, "step": 13500 }, { "grad_norm": 0.30466699600219727, "learning_rate": 2.6137218473407477e-05, "loss": 0.0103, "step": 13510 }, { "grad_norm": 0.19289511442184448, "learning_rate": 2.606460059681436e-05, "loss": 0.0055, "step": 13520 }, { "grad_norm": 0.17380407452583313, "learning_rate": 2.599204815868928e-05, "loss": 0.0059, "step": 13530 }, { "grad_norm": 0.23569989204406738, "learning_rate": 2.5919561357387756e-05, "loss": 0.0104, "step": 13540 }, { "grad_norm": 0.3003978431224823, "learning_rate": 2.5847140391085972e-05, "loss": 0.0081, "step": 13550 }, { "grad_norm": 0.22126668691635132, "learning_rate": 2.5774785457780103e-05, "loss": 0.0084, "step": 13560 }, { "grad_norm": 0.13515742123126984, "learning_rate": 2.5702496755285753e-05, "loss": 0.0047, "step": 13570 }, { "grad_norm": 0.26540789008140564, "learning_rate": 2.5630274481237483e-05, "loss": 0.0096, "step": 13580 }, { "grad_norm": 0.15105336904525757, "learning_rate": 2.5558118833088197e-05, "loss": 0.0066, "step": 13590 }, { "grad_norm": 0.1857864409685135, "learning_rate": 2.548603000810872e-05, "loss": 0.0077, "step": 13600 }, { "grad_norm": 0.2596299946308136, "learning_rate": 2.5414008203387152e-05, "loss": 0.0074, "step": 13610 }, { "grad_norm": 0.26205721497535706, "learning_rate": 2.534205361582834e-05, "loss": 0.007, "step": 13620 }, { "grad_norm": 0.33126112818717957, "learning_rate": 2.527016644215338e-05, "loss": 0.0072, "step": 13630 }, { "grad_norm": 0.22200165688991547, "learning_rate": 2.519834687889905e-05, "loss": 0.0093, "step": 13640 }, { "grad_norm": 0.23359975218772888, "learning_rate": 2.5126595122417295e-05, "loss": 0.006, "step": 13650 }, { "grad_norm": 0.22043204307556152, "learning_rate": 2.5054911368874713e-05, "loss": 0.0082, "step": 13660 }, { "grad_norm": 0.19226586818695068, "learning_rate": 2.4983295814251916e-05, "loss": 0.0052, "step": 13670 }, { "grad_norm": 0.2468976527452469, "learning_rate": 2.4911748654343105e-05, "loss": 0.0095, "step": 13680 }, { "grad_norm": 0.23663662374019623, "learning_rate": 2.4840270084755463e-05, "loss": 0.0059, "step": 13690 }, { "grad_norm": 0.23496931791305542, "learning_rate": 2.4768860300908685e-05, "loss": 0.0058, "step": 13700 }, { "grad_norm": 0.26951465010643005, "learning_rate": 2.469751949803443e-05, "loss": 0.0079, "step": 13710 }, { "grad_norm": 0.317673921585083, "learning_rate": 2.4626247871175666e-05, "loss": 0.0053, "step": 13720 }, { "grad_norm": 0.30217137932777405, "learning_rate": 2.4555045615186346e-05, "loss": 0.007, "step": 13730 }, { "grad_norm": 0.2663007378578186, "learning_rate": 2.4483912924730677e-05, "loss": 0.006, "step": 13740 }, { "grad_norm": 0.22336773574352264, "learning_rate": 2.4412849994282742e-05, "loss": 0.009, "step": 13750 }, { "grad_norm": 0.2654953896999359, "learning_rate": 2.434185701812592e-05, "loss": 0.0061, "step": 13760 }, { "grad_norm": 0.32645687460899353, "learning_rate": 2.4270934190352218e-05, "loss": 0.0057, "step": 13770 }, { "grad_norm": 0.2892066240310669, "learning_rate": 2.4200081704861998e-05, "loss": 0.0059, "step": 13780 }, { "grad_norm": 0.36011695861816406, "learning_rate": 2.412929975536321e-05, "loss": 0.0069, "step": 13790 }, { "grad_norm": 0.34654074907302856, "learning_rate": 2.4058588535371017e-05, "loss": 0.0086, "step": 13800 }, { "grad_norm": 0.2702867388725281, "learning_rate": 2.3987948238207243e-05, "loss": 0.0092, "step": 13810 }, { "grad_norm": 0.24130818247795105, "learning_rate": 2.3917379056999678e-05, "loss": 0.0062, "step": 13820 }, { "grad_norm": 0.3321293890476227, "learning_rate": 2.3846881184681824e-05, "loss": 0.0089, "step": 13830 }, { "grad_norm": 0.2750306725502014, "learning_rate": 2.377645481399214e-05, "loss": 0.0066, "step": 13840 }, { "grad_norm": 0.17475612461566925, "learning_rate": 2.3706100137473667e-05, "loss": 0.0053, "step": 13850 }, { "grad_norm": 0.26268622279167175, "learning_rate": 2.3635817347473394e-05, "loss": 0.0089, "step": 13860 }, { "grad_norm": 0.24029071629047394, "learning_rate": 2.3565606636141757e-05, "loss": 0.005, "step": 13870 }, { "grad_norm": 0.1715182065963745, "learning_rate": 2.3495468195432203e-05, "loss": 0.0062, "step": 13880 }, { "grad_norm": 0.23351286351680756, "learning_rate": 2.3425402217100507e-05, "loss": 0.0059, "step": 13890 }, { "grad_norm": 0.24628567695617676, "learning_rate": 2.3355408892704424e-05, "loss": 0.0079, "step": 13900 }, { "grad_norm": 0.20386043190956116, "learning_rate": 2.3285488413603003e-05, "loss": 0.0056, "step": 13910 }, { "grad_norm": 0.21232497692108154, "learning_rate": 2.321564097095615e-05, "loss": 0.0073, "step": 13920 }, { "grad_norm": 0.24907775223255157, "learning_rate": 2.3145866755724142e-05, "loss": 0.0054, "step": 13930 }, { "grad_norm": 0.15889552235603333, "learning_rate": 2.307616595866699e-05, "loss": 0.0064, "step": 13940 }, { "grad_norm": 0.1938760131597519, "learning_rate": 2.3006538770344032e-05, "loss": 0.0081, "step": 13950 }, { "grad_norm": 0.15316195785999298, "learning_rate": 2.293698538111334e-05, "loss": 0.0061, "step": 13960 }, { "grad_norm": 0.176402747631073, "learning_rate": 2.28675059811312e-05, "loss": 0.005, "step": 13970 }, { "grad_norm": 0.1642218977212906, "learning_rate": 2.279810076035167e-05, "loss": 0.0055, "step": 13980 }, { "grad_norm": 0.20960569381713867, "learning_rate": 2.272876990852596e-05, "loss": 0.0062, "step": 13990 }, { "grad_norm": 0.16663093864917755, "learning_rate": 2.265951361520195e-05, "loss": 0.0053, "step": 14000 }, { "grad_norm": 0.20839343965053558, "learning_rate": 2.2590332069723748e-05, "loss": 0.0067, "step": 14010 }, { "grad_norm": 0.3321441113948822, "learning_rate": 2.2521225461231004e-05, "loss": 0.0073, "step": 14020 }, { "grad_norm": 0.18815380334854126, "learning_rate": 2.2452193978658597e-05, "loss": 0.0059, "step": 14030 }, { "grad_norm": 0.2880175709724426, "learning_rate": 2.238323781073594e-05, "loss": 0.0063, "step": 14040 }, { "grad_norm": 0.1992231011390686, "learning_rate": 2.2314357145986552e-05, "loss": 0.0053, "step": 14050 }, { "grad_norm": 0.2528548240661621, "learning_rate": 2.224555217272757e-05, "loss": 0.006, "step": 14060 }, { "grad_norm": 0.1951238065958023, "learning_rate": 2.2176823079069127e-05, "loss": 0.0063, "step": 14070 }, { "grad_norm": 0.1805834025144577, "learning_rate": 2.210817005291398e-05, "loss": 0.0056, "step": 14080 }, { "grad_norm": 0.3065294623374939, "learning_rate": 2.203959328195686e-05, "loss": 0.0055, "step": 14090 }, { "grad_norm": 0.31868666410446167, "learning_rate": 2.1971092953684026e-05, "loss": 0.0067, "step": 14100 }, { "grad_norm": 0.21438057720661163, "learning_rate": 2.1902669255372788e-05, "loss": 0.0047, "step": 14110 }, { "grad_norm": 0.22993038594722748, "learning_rate": 2.1834322374090897e-05, "loss": 0.0054, "step": 14120 }, { "grad_norm": 0.22806033492088318, "learning_rate": 2.1766052496696153e-05, "loss": 0.008, "step": 14130 }, { "grad_norm": 0.2617322504520416, "learning_rate": 2.169785980983577e-05, "loss": 0.0082, "step": 14140 }, { "grad_norm": 0.23122186958789825, "learning_rate": 2.162974449994593e-05, "loss": 0.006, "step": 14150 }, { "grad_norm": 0.22069738805294037, "learning_rate": 2.1561706753251337e-05, "loss": 0.0072, "step": 14160 }, { "grad_norm": 0.20628024637699127, "learning_rate": 2.1493746755764544e-05, "loss": 0.0093, "step": 14170 }, { "grad_norm": 0.2738070487976074, "learning_rate": 2.1425864693285635e-05, "loss": 0.0073, "step": 14180 }, { "grad_norm": 0.21884934604167938, "learning_rate": 2.1358060751401547e-05, "loss": 0.0074, "step": 14190 }, { "grad_norm": 0.23741525411605835, "learning_rate": 2.129033511548566e-05, "loss": 0.0061, "step": 14200 }, { "grad_norm": 0.23324857652187347, "learning_rate": 2.1222687970697315e-05, "loss": 0.0069, "step": 14210 }, { "grad_norm": 0.19664190709590912, "learning_rate": 2.1155119501981173e-05, "loss": 0.0071, "step": 14220 }, { "grad_norm": 0.2240017056465149, "learning_rate": 2.1087629894066895e-05, "loss": 0.0066, "step": 14230 }, { "grad_norm": 0.23721759021282196, "learning_rate": 2.1020219331468473e-05, "loss": 0.0053, "step": 14240 }, { "grad_norm": 0.1736283302307129, "learning_rate": 2.095288799848379e-05, "loss": 0.0064, "step": 14250 }, { "grad_norm": 0.3125157058238983, "learning_rate": 2.088563607919417e-05, "loss": 0.008, "step": 14260 }, { "grad_norm": 0.20235560834407806, "learning_rate": 2.0818463757463786e-05, "loss": 0.0054, "step": 14270 }, { "grad_norm": 0.25843682885169983, "learning_rate": 2.0751371216939175e-05, "loss": 0.0058, "step": 14280 }, { "grad_norm": 0.1329217255115509, "learning_rate": 2.068435864104882e-05, "loss": 0.0041, "step": 14290 }, { "grad_norm": 0.14160753786563873, "learning_rate": 2.0617426213002506e-05, "loss": 0.006, "step": 14300 }, { "grad_norm": 0.19008612632751465, "learning_rate": 2.055057411579097e-05, "loss": 0.0053, "step": 14310 }, { "grad_norm": 0.18673664331436157, "learning_rate": 2.0483802532185286e-05, "loss": 0.0081, "step": 14320 }, { "grad_norm": 0.2571374177932739, "learning_rate": 2.041711164473638e-05, "loss": 0.0073, "step": 14330 }, { "grad_norm": 0.29293274879455566, "learning_rate": 2.0350501635774637e-05, "loss": 0.0067, "step": 14340 }, { "grad_norm": 0.26819995045661926, "learning_rate": 2.0283972687409247e-05, "loss": 0.0059, "step": 14350 }, { "grad_norm": 0.18850700557231903, "learning_rate": 2.021752498152784e-05, "loss": 0.0083, "step": 14360 }, { "grad_norm": 0.1649707555770874, "learning_rate": 2.015115869979589e-05, "loss": 0.0073, "step": 14370 }, { "grad_norm": 0.19669297337532043, "learning_rate": 2.0084874023656265e-05, "loss": 0.0075, "step": 14380 }, { "grad_norm": 0.20495162904262543, "learning_rate": 2.001867113432877e-05, "loss": 0.007, "step": 14390 }, { "grad_norm": 0.18790069222450256, "learning_rate": 1.995255021280954e-05, "loss": 0.0059, "step": 14400 }, { "grad_norm": 0.2360209822654724, "learning_rate": 1.9886511439870688e-05, "loss": 0.0069, "step": 14410 }, { "grad_norm": 0.1849478781223297, "learning_rate": 1.9820554996059675e-05, "loss": 0.006, "step": 14420 }, { "grad_norm": 0.2239524871110916, "learning_rate": 1.9754681061698893e-05, "loss": 0.0045, "step": 14430 }, { "grad_norm": 0.23751626908779144, "learning_rate": 1.9688889816885185e-05, "loss": 0.0048, "step": 14440 }, { "grad_norm": 0.226037859916687, "learning_rate": 1.962318144148928e-05, "loss": 0.0097, "step": 14450 }, { "grad_norm": 0.23243887722492218, "learning_rate": 1.955755611515539e-05, "loss": 0.005, "step": 14460 }, { "grad_norm": 0.23626203835010529, "learning_rate": 1.9492014017300642e-05, "loss": 0.0077, "step": 14470 }, { "grad_norm": 0.22626632452011108, "learning_rate": 1.942655532711461e-05, "loss": 0.0094, "step": 14480 }, { "grad_norm": 0.25426167249679565, "learning_rate": 1.9361180223558882e-05, "loss": 0.0058, "step": 14490 }, { "grad_norm": 0.16956761479377747, "learning_rate": 1.929588888536647e-05, "loss": 0.0042, "step": 14500 }, { "grad_norm": 0.1914738118648529, "learning_rate": 1.9230681491041425e-05, "loss": 0.0058, "step": 14510 }, { "grad_norm": 0.21050724387168884, "learning_rate": 1.9165558218858264e-05, "loss": 0.0049, "step": 14520 }, { "grad_norm": 0.18956254422664642, "learning_rate": 1.9100519246861505e-05, "loss": 0.0071, "step": 14530 }, { "grad_norm": 0.3159593343734741, "learning_rate": 1.9035564752865248e-05, "loss": 0.0062, "step": 14540 }, { "grad_norm": 0.26023414731025696, "learning_rate": 1.897069491445258e-05, "loss": 0.0073, "step": 14550 }, { "grad_norm": 0.20540374517440796, "learning_rate": 1.890590990897515e-05, "loss": 0.0057, "step": 14560 }, { "grad_norm": 0.3301018476486206, "learning_rate": 1.884120991355272e-05, "loss": 0.0066, "step": 14570 }, { "grad_norm": 0.2460355907678604, "learning_rate": 1.8776595105072576e-05, "loss": 0.0065, "step": 14580 }, { "grad_norm": 0.2898869514465332, "learning_rate": 1.8712065660189166e-05, "loss": 0.0063, "step": 14590 }, { "grad_norm": 0.17953884601593018, "learning_rate": 1.8647621755323513e-05, "loss": 0.0056, "step": 14600 }, { "grad_norm": 0.14495480060577393, "learning_rate": 1.858326356666278e-05, "loss": 0.0061, "step": 14610 }, { "grad_norm": 0.2407841682434082, "learning_rate": 1.851899127015983e-05, "loss": 0.0051, "step": 14620 }, { "grad_norm": 0.22782306373119354, "learning_rate": 1.8454805041532626e-05, "loss": 0.0088, "step": 14630 }, { "grad_norm": 0.2950453460216522, "learning_rate": 1.8390705056263906e-05, "loss": 0.006, "step": 14640 }, { "grad_norm": 0.1985420435667038, "learning_rate": 1.832669148960057e-05, "loss": 0.0056, "step": 14650 }, { "grad_norm": 0.14298784732818604, "learning_rate": 1.8262764516553233e-05, "loss": 0.0047, "step": 14660 }, { "grad_norm": 0.17360757291316986, "learning_rate": 1.8198924311895843e-05, "loss": 0.0043, "step": 14670 }, { "grad_norm": 0.1732722371816635, "learning_rate": 1.813517105016505e-05, "loss": 0.006, "step": 14680 }, { "grad_norm": 0.19590047001838684, "learning_rate": 1.8071504905659888e-05, "loss": 0.0049, "step": 14690 }, { "grad_norm": 0.17283903062343597, "learning_rate": 1.800792605244109e-05, "loss": 0.0061, "step": 14700 }, { "grad_norm": 0.1889088898897171, "learning_rate": 1.7944434664330844e-05, "loss": 0.0147, "step": 14710 }, { "grad_norm": 0.2242112010717392, "learning_rate": 1.7881030914912212e-05, "loss": 0.0064, "step": 14720 }, { "grad_norm": 0.2090156078338623, "learning_rate": 1.7817714977528577e-05, "loss": 0.0062, "step": 14730 }, { "grad_norm": 0.3117174506187439, "learning_rate": 1.7754487025283332e-05, "loss": 0.0092, "step": 14740 }, { "grad_norm": 0.3212093114852905, "learning_rate": 1.7691347231039275e-05, "loss": 0.0067, "step": 14750 }, { "grad_norm": 0.1588798463344574, "learning_rate": 1.7628295767418164e-05, "loss": 0.0075, "step": 14760 }, { "grad_norm": 0.36109018325805664, "learning_rate": 1.7565332806800333e-05, "loss": 0.0074, "step": 14770 }, { "grad_norm": 0.21150946617126465, "learning_rate": 1.750245852132408e-05, "loss": 0.0053, "step": 14780 }, { "grad_norm": 0.23343341052532196, "learning_rate": 1.7439673082885323e-05, "loss": 0.007, "step": 14790 }, { "grad_norm": 0.1730482578277588, "learning_rate": 1.7376976663137047e-05, "loss": 0.0041, "step": 14800 }, { "grad_norm": 0.19543083012104034, "learning_rate": 1.7314369433488853e-05, "loss": 0.005, "step": 14810 }, { "grad_norm": 0.16112789511680603, "learning_rate": 1.7251851565106548e-05, "loss": 0.0048, "step": 14820 }, { "grad_norm": 0.2860219478607178, "learning_rate": 1.7189423228911574e-05, "loss": 0.0056, "step": 14830 }, { "grad_norm": 0.15021967887878418, "learning_rate": 1.7127084595580606e-05, "loss": 0.0046, "step": 14840 }, { "grad_norm": 0.18744181096553802, "learning_rate": 1.706483583554513e-05, "loss": 0.0048, "step": 14850 }, { "grad_norm": 0.10445283353328705, "learning_rate": 1.700267711899083e-05, "loss": 0.0053, "step": 14860 }, { "grad_norm": 0.18256716430187225, "learning_rate": 1.69406086158573e-05, "loss": 0.0053, "step": 14870 }, { "grad_norm": 0.2134125828742981, "learning_rate": 1.6878630495837455e-05, "loss": 0.0047, "step": 14880 }, { "grad_norm": 0.13954190909862518, "learning_rate": 1.681674292837707e-05, "loss": 0.006, "step": 14890 }, { "grad_norm": 0.24265289306640625, "learning_rate": 1.6754946082674444e-05, "loss": 0.0052, "step": 14900 }, { "grad_norm": 0.22468554973602295, "learning_rate": 1.6693240127679748e-05, "loss": 0.0048, "step": 14910 }, { "grad_norm": 0.25264444947242737, "learning_rate": 1.663162523209475e-05, "loss": 0.0094, "step": 14920 }, { "grad_norm": 0.1693328320980072, "learning_rate": 1.6570101564372193e-05, "loss": 0.0058, "step": 14930 }, { "grad_norm": 0.24644319713115692, "learning_rate": 1.650866929271543e-05, "loss": 0.0054, "step": 14940 }, { "grad_norm": 0.19081056118011475, "learning_rate": 1.644732858507797e-05, "loss": 0.0062, "step": 14950 }, { "grad_norm": 0.13013608753681183, "learning_rate": 1.6386079609162943e-05, "loss": 0.0043, "step": 14960 }, { "grad_norm": 0.14223334193229675, "learning_rate": 1.6324922532422742e-05, "loss": 0.004, "step": 14970 }, { "grad_norm": 0.17659057676792145, "learning_rate": 1.6263857522058434e-05, "loss": 0.0046, "step": 14980 }, { "grad_norm": 0.25959113240242004, "learning_rate": 1.6202884745019443e-05, "loss": 0.0068, "step": 14990 }, { "grad_norm": 0.24221943318843842, "learning_rate": 1.614200436800304e-05, "loss": 0.0046, "step": 15000 }, { "grad_norm": 0.20682360231876373, "learning_rate": 1.6081216557453814e-05, "loss": 0.0048, "step": 15010 }, { "grad_norm": 0.12759315967559814, "learning_rate": 1.6020521479563367e-05, "loss": 0.0042, "step": 15020 }, { "grad_norm": 0.1808764487504959, "learning_rate": 1.5959919300269654e-05, "loss": 0.0067, "step": 15030 }, { "grad_norm": 0.2767375111579895, "learning_rate": 1.5899410185256764e-05, "loss": 0.0068, "step": 15040 }, { "grad_norm": 0.165812149643898, "learning_rate": 1.583899429995431e-05, "loss": 0.0053, "step": 15050 }, { "grad_norm": 0.14400404691696167, "learning_rate": 1.5778671809536993e-05, "loss": 0.0051, "step": 15060 }, { "grad_norm": 0.19589701294898987, "learning_rate": 1.5718442878924246e-05, "loss": 0.0057, "step": 15070 }, { "grad_norm": 0.21216771006584167, "learning_rate": 1.5658307672779593e-05, "loss": 0.0043, "step": 15080 }, { "grad_norm": 0.1631125807762146, "learning_rate": 1.5598266355510427e-05, "loss": 0.0054, "step": 15090 }, { "grad_norm": 0.2034980058670044, "learning_rate": 1.553831909126744e-05, "loss": 0.0044, "step": 15100 }, { "grad_norm": 0.23151086270809174, "learning_rate": 1.5478466043944135e-05, "loss": 0.0078, "step": 15110 }, { "grad_norm": 0.2196432501077652, "learning_rate": 1.5418707377176468e-05, "loss": 0.0091, "step": 15120 }, { "grad_norm": 0.2634916305541992, "learning_rate": 1.535904325434233e-05, "loss": 0.0059, "step": 15130 }, { "grad_norm": 0.17393429577350616, "learning_rate": 1.529947383856118e-05, "loss": 0.0042, "step": 15140 }, { "grad_norm": 0.1786508411169052, "learning_rate": 1.5239999292693524e-05, "loss": 0.0054, "step": 15150 }, { "grad_norm": 0.15929776430130005, "learning_rate": 1.5180619779340505e-05, "loss": 0.0042, "step": 15160 }, { "grad_norm": 0.18316993117332458, "learning_rate": 1.5121335460843428e-05, "loss": 0.005, "step": 15170 }, { "grad_norm": 0.17363984882831573, "learning_rate": 1.5062146499283347e-05, "loss": 0.0041, "step": 15180 }, { "grad_norm": 0.15891006588935852, "learning_rate": 1.5003053056480643e-05, "loss": 0.0064, "step": 15190 }, { "grad_norm": 0.16135641932487488, "learning_rate": 1.4944055293994551e-05, "loss": 0.0059, "step": 15200 }, { "grad_norm": 0.19979242980480194, "learning_rate": 1.4885153373122656e-05, "loss": 0.0043, "step": 15210 }, { "grad_norm": 0.17464110255241394, "learning_rate": 1.482634745490059e-05, "loss": 0.005, "step": 15220 }, { "grad_norm": 0.1621282547712326, "learning_rate": 1.4767637700101466e-05, "loss": 0.0047, "step": 15230 }, { "grad_norm": 0.2143547683954239, "learning_rate": 1.4709024269235528e-05, "loss": 0.0058, "step": 15240 }, { "grad_norm": 0.1770196408033371, "learning_rate": 1.4650507322549684e-05, "loss": 0.0078, "step": 15250 }, { "grad_norm": 0.2484661042690277, "learning_rate": 1.4592087020026972e-05, "loss": 0.005, "step": 15260 }, { "grad_norm": 0.17090244591236115, "learning_rate": 1.4533763521386318e-05, "loss": 0.0063, "step": 15270 }, { "grad_norm": 0.16400618851184845, "learning_rate": 1.44755369860819e-05, "loss": 0.0043, "step": 15280 }, { "grad_norm": 0.18535998463630676, "learning_rate": 1.441740757330287e-05, "loss": 0.0069, "step": 15290 }, { "grad_norm": 0.19011028110980988, "learning_rate": 1.4359375441972844e-05, "loss": 0.005, "step": 15300 }, { "grad_norm": 0.1798458695411682, "learning_rate": 1.4301440750749395e-05, "loss": 0.0062, "step": 15310 }, { "grad_norm": 0.389713317155838, "learning_rate": 1.4243603658023808e-05, "loss": 0.0082, "step": 15320 }, { "grad_norm": 0.2927606999874115, "learning_rate": 1.4185864321920444e-05, "loss": 0.0065, "step": 15330 }, { "grad_norm": 0.223611980676651, "learning_rate": 1.4128222900296485e-05, "loss": 0.0053, "step": 15340 }, { "grad_norm": 0.18469679355621338, "learning_rate": 1.407067955074135e-05, "loss": 0.005, "step": 15350 }, { "grad_norm": 0.16220906376838684, "learning_rate": 1.4013234430576356e-05, "loss": 0.0048, "step": 15360 }, { "grad_norm": 0.2546613812446594, "learning_rate": 1.3955887696854286e-05, "loss": 0.0044, "step": 15370 }, { "grad_norm": 0.18166832625865936, "learning_rate": 1.38986395063589e-05, "loss": 0.0051, "step": 15380 }, { "grad_norm": 0.15993843972682953, "learning_rate": 1.3841490015604597e-05, "loss": 0.0078, "step": 15390 }, { "grad_norm": 0.2392805814743042, "learning_rate": 1.3784439380835879e-05, "loss": 0.0065, "step": 15400 }, { "grad_norm": 0.18197618424892426, "learning_rate": 1.3727487758026986e-05, "loss": 0.0049, "step": 15410 }, { "grad_norm": 0.2519821226596832, "learning_rate": 1.3670635302881525e-05, "loss": 0.0077, "step": 15420 }, { "grad_norm": 0.149385467171669, "learning_rate": 1.3613882170831888e-05, "loss": 0.0044, "step": 15430 }, { "grad_norm": 0.1491548866033554, "learning_rate": 1.355722851703901e-05, "loss": 0.0046, "step": 15440 }, { "grad_norm": 0.18058425188064575, "learning_rate": 1.3500674496391814e-05, "loss": 0.0066, "step": 15450 }, { "grad_norm": 0.15405023097991943, "learning_rate": 1.3444220263506795e-05, "loss": 0.0049, "step": 15460 }, { "grad_norm": 0.1475231945514679, "learning_rate": 1.3387865972727714e-05, "loss": 0.0049, "step": 15470 }, { "grad_norm": 0.20055174827575684, "learning_rate": 1.3331611778125036e-05, "loss": 0.0041, "step": 15480 }, { "grad_norm": 0.23059506714344025, "learning_rate": 1.3275457833495564e-05, "loss": 0.0066, "step": 15490 }, { "grad_norm": 0.23637868463993073, "learning_rate": 1.3219404292362065e-05, "loss": 0.0045, "step": 15500 }, { "grad_norm": 0.21013008058071136, "learning_rate": 1.3163451307972751e-05, "loss": 0.0045, "step": 15510 }, { "grad_norm": 0.18651491403579712, "learning_rate": 1.3107599033300977e-05, "loss": 0.0052, "step": 15520 }, { "grad_norm": 0.22967149317264557, "learning_rate": 1.305184762104471e-05, "loss": 0.0107, "step": 15530 }, { "grad_norm": 0.24542079865932465, "learning_rate": 1.2996197223626178e-05, "loss": 0.0072, "step": 15540 }, { "grad_norm": 0.1853976845741272, "learning_rate": 1.2940647993191457e-05, "loss": 0.007, "step": 15550 }, { "grad_norm": 0.17506596446037292, "learning_rate": 1.2885200081610005e-05, "loss": 0.0049, "step": 15560 }, { "grad_norm": 0.21625709533691406, "learning_rate": 1.2829853640474316e-05, "loss": 0.0039, "step": 15570 }, { "grad_norm": 0.18540851771831512, "learning_rate": 1.2774608821099438e-05, "loss": 0.0047, "step": 15580 }, { "grad_norm": 0.21105524897575378, "learning_rate": 1.2719465774522577e-05, "loss": 0.0053, "step": 15590 }, { "grad_norm": 0.15021876990795135, "learning_rate": 1.2664424651502755e-05, "loss": 0.0047, "step": 15600 }, { "grad_norm": 0.15401819348335266, "learning_rate": 1.260948560252026e-05, "loss": 0.0049, "step": 15610 }, { "grad_norm": 0.18564783036708832, "learning_rate": 1.2554648777776396e-05, "loss": 0.0046, "step": 15620 }, { "grad_norm": 0.12305077910423279, "learning_rate": 1.2499914327192919e-05, "loss": 0.0034, "step": 15630 }, { "grad_norm": 0.166657492518425, "learning_rate": 1.2445282400411722e-05, "loss": 0.0044, "step": 15640 }, { "grad_norm": 0.16441041231155396, "learning_rate": 1.2390753146794437e-05, "loss": 0.0082, "step": 15650 }, { "grad_norm": 0.1849135309457779, "learning_rate": 1.2336326715421925e-05, "loss": 0.0074, "step": 15660 }, { "grad_norm": 0.1206001341342926, "learning_rate": 1.2282003255094005e-05, "loss": 0.0047, "step": 15670 }, { "grad_norm": 0.2652372121810913, "learning_rate": 1.2227782914328928e-05, "loss": 0.0045, "step": 15680 }, { "grad_norm": 0.20016765594482422, "learning_rate": 1.2173665841363018e-05, "loss": 0.0059, "step": 15690 }, { "grad_norm": 0.2042270004749298, "learning_rate": 1.211965218415032e-05, "loss": 0.0045, "step": 15700 }, { "grad_norm": 0.2085312008857727, "learning_rate": 1.2065742090362082e-05, "loss": 0.0058, "step": 15710 }, { "grad_norm": 0.13603872060775757, "learning_rate": 1.2011935707386457e-05, "loss": 0.0058, "step": 15720 }, { "grad_norm": 0.1352131962776184, "learning_rate": 1.1958233182328044e-05, "loss": 0.0042, "step": 15730 }, { "grad_norm": 0.217603862285614, "learning_rate": 1.1904634662007474e-05, "loss": 0.0053, "step": 15740 }, { "grad_norm": 0.2739396393299103, "learning_rate": 1.1851140292961088e-05, "loss": 0.0042, "step": 15750 }, { "grad_norm": 0.21274851262569427, "learning_rate": 1.1797750221440424e-05, "loss": 0.0054, "step": 15760 }, { "grad_norm": 0.18774281442165375, "learning_rate": 1.1744464593411897e-05, "loss": 0.0056, "step": 15770 }, { "grad_norm": 0.18023785948753357, "learning_rate": 1.1691283554556399e-05, "loss": 0.0057, "step": 15780 }, { "grad_norm": 0.18424303829669952, "learning_rate": 1.1638207250268834e-05, "loss": 0.0051, "step": 15790 }, { "grad_norm": 0.22490055859088898, "learning_rate": 1.158523582565782e-05, "loss": 0.0051, "step": 15800 }, { "grad_norm": 0.16953471302986145, "learning_rate": 1.1532369425545192e-05, "loss": 0.0044, "step": 15810 }, { "grad_norm": 0.17409905791282654, "learning_rate": 1.1479608194465662e-05, "loss": 0.0067, "step": 15820 }, { "grad_norm": 0.18748362362384796, "learning_rate": 1.1426952276666442e-05, "loss": 0.0051, "step": 15830 }, { "grad_norm": 0.15942522883415222, "learning_rate": 1.1374401816106778e-05, "loss": 0.0044, "step": 15840 }, { "grad_norm": 0.14493024349212646, "learning_rate": 1.1321956956457646e-05, "loss": 0.0073, "step": 15850 }, { "grad_norm": 0.10824569314718246, "learning_rate": 1.1269617841101277e-05, "loss": 0.0042, "step": 15860 }, { "grad_norm": 0.12193453311920166, "learning_rate": 1.1217384613130804e-05, "loss": 0.0042, "step": 15870 }, { "grad_norm": 0.16226844489574432, "learning_rate": 1.11652574153499e-05, "loss": 0.0077, "step": 15880 }, { "grad_norm": 0.22241581976413727, "learning_rate": 1.1113236390272303e-05, "loss": 0.0045, "step": 15890 }, { "grad_norm": 0.20633941888809204, "learning_rate": 1.106132168012155e-05, "loss": 0.0062, "step": 15900 }, { "grad_norm": 0.1466503143310547, "learning_rate": 1.1009513426830448e-05, "loss": 0.0071, "step": 15910 }, { "grad_norm": 0.19736774265766144, "learning_rate": 1.0957811772040777e-05, "loss": 0.0053, "step": 15920 }, { "grad_norm": 0.19769582152366638, "learning_rate": 1.0906216857102913e-05, "loss": 0.009, "step": 15930 }, { "grad_norm": 0.21258661150932312, "learning_rate": 1.0854728823075355e-05, "loss": 0.0043, "step": 15940 }, { "grad_norm": 0.19439363479614258, "learning_rate": 1.0803347810724452e-05, "loss": 0.0051, "step": 15950 }, { "grad_norm": 0.14848843216896057, "learning_rate": 1.0752073960523911e-05, "loss": 0.0072, "step": 15960 }, { "grad_norm": 0.17548304796218872, "learning_rate": 1.070090741265447e-05, "loss": 0.0043, "step": 15970 }, { "grad_norm": 0.18810178339481354, "learning_rate": 1.0649848307003547e-05, "loss": 0.0049, "step": 15980 }, { "grad_norm": 0.15141376852989197, "learning_rate": 1.0598896783164757e-05, "loss": 0.0045, "step": 15990 }, { "grad_norm": 0.14855995774269104, "learning_rate": 1.0548052980437645e-05, "loss": 0.004, "step": 16000 }, { "grad_norm": 0.20158180594444275, "learning_rate": 1.049731703782722e-05, "loss": 0.0041, "step": 16010 }, { "grad_norm": 0.16021481156349182, "learning_rate": 1.0446689094043587e-05, "loss": 0.0032, "step": 16020 }, { "grad_norm": 0.10641957819461823, "learning_rate": 1.039616928750165e-05, "loss": 0.0038, "step": 16030 }, { "grad_norm": 0.16106703877449036, "learning_rate": 1.0345757756320612e-05, "loss": 0.0056, "step": 16040 }, { "grad_norm": 0.1110115721821785, "learning_rate": 1.0295454638323666e-05, "loss": 0.005, "step": 16050 }, { "grad_norm": 0.13663776218891144, "learning_rate": 1.0245260071037632e-05, "loss": 0.0038, "step": 16060 }, { "grad_norm": 0.29903626441955566, "learning_rate": 1.0195174191692518e-05, "loss": 0.005, "step": 16070 }, { "grad_norm": 0.1696806699037552, "learning_rate": 1.014519713722124e-05, "loss": 0.0056, "step": 16080 }, { "grad_norm": 0.1798056811094284, "learning_rate": 1.0095329044259132e-05, "loss": 0.0049, "step": 16090 }, { "grad_norm": 0.13110291957855225, "learning_rate": 1.004557004914365e-05, "loss": 0.005, "step": 16100 }, { "grad_norm": 0.13511288166046143, "learning_rate": 9.995920287914007e-06, "loss": 0.0061, "step": 16110 }, { "grad_norm": 0.16194112598896027, "learning_rate": 9.946379896310737e-06, "loss": 0.0042, "step": 16120 }, { "grad_norm": 0.1300184428691864, "learning_rate": 9.896949009775396e-06, "loss": 0.0039, "step": 16130 }, { "grad_norm": 0.1603885293006897, "learning_rate": 9.847627763450134e-06, "loss": 0.0055, "step": 16140 }, { "grad_norm": 0.15740537643432617, "learning_rate": 9.798416292177337e-06, "loss": 0.0047, "step": 16150 }, { "grad_norm": 0.1694498360157013, "learning_rate": 9.74931473049932e-06, "loss": 0.0073, "step": 16160 }, { "grad_norm": 0.21871136128902435, "learning_rate": 9.700323212657847e-06, "loss": 0.004, "step": 16170 }, { "grad_norm": 0.13767218589782715, "learning_rate": 9.65144187259388e-06, "loss": 0.0069, "step": 16180 }, { "grad_norm": 0.16583232581615448, "learning_rate": 9.602670843947132e-06, "loss": 0.0048, "step": 16190 }, { "grad_norm": 0.20373769104480743, "learning_rate": 9.554010260055713e-06, "loss": 0.0065, "step": 16200 }, { "grad_norm": 0.1713123768568039, "learning_rate": 9.505460253955834e-06, "loss": 0.0038, "step": 16210 }, { "grad_norm": 0.21792767941951752, "learning_rate": 9.457020958381324e-06, "loss": 0.0048, "step": 16220 }, { "grad_norm": 0.19638560712337494, "learning_rate": 9.408692505763395e-06, "loss": 0.005, "step": 16230 }, { "grad_norm": 0.14792053401470184, "learning_rate": 9.360475028230181e-06, "loss": 0.005, "step": 16240 }, { "grad_norm": 0.18360894918441772, "learning_rate": 9.312368657606412e-06, "loss": 0.0044, "step": 16250 }, { "grad_norm": 0.2135966718196869, "learning_rate": 9.264373525413096e-06, "loss": 0.0052, "step": 16260 }, { "grad_norm": 0.1251610815525055, "learning_rate": 9.216489762867058e-06, "loss": 0.0042, "step": 16270 }, { "grad_norm": 0.18885968625545502, "learning_rate": 9.168717500880708e-06, "loss": 0.004, "step": 16280 }, { "grad_norm": 0.1281227469444275, "learning_rate": 9.121056870061574e-06, "loss": 0.006, "step": 16290 }, { "grad_norm": 0.1496238261461258, "learning_rate": 9.073508000711983e-06, "loss": 0.0042, "step": 16300 }, { "grad_norm": 0.17822037637233734, "learning_rate": 9.026071022828758e-06, "loss": 0.006, "step": 16310 }, { "grad_norm": 0.14292031526565552, "learning_rate": 8.978746066102771e-06, "loss": 0.004, "step": 16320 }, { "grad_norm": 0.1662091165781021, "learning_rate": 8.931533259918634e-06, "loss": 0.0059, "step": 16330 }, { "grad_norm": 0.12459095567464828, "learning_rate": 8.884432733354382e-06, "loss": 0.0038, "step": 16340 }, { "grad_norm": 0.150796040892601, "learning_rate": 8.837444615181029e-06, "loss": 0.0048, "step": 16350 }, { "grad_norm": 0.1963094025850296, "learning_rate": 8.790569033862323e-06, "loss": 0.0058, "step": 16360 }, { "grad_norm": 0.20088692009449005, "learning_rate": 8.7438061175543e-06, "loss": 0.0047, "step": 16370 }, { "grad_norm": 0.15101876854896545, "learning_rate": 8.697155994104978e-06, "loss": 0.006, "step": 16380 }, { "grad_norm": 0.14375591278076172, "learning_rate": 8.650618791054033e-06, "loss": 0.0049, "step": 16390 }, { "grad_norm": 0.1365630179643631, "learning_rate": 8.604194635632373e-06, "loss": 0.0045, "step": 16400 }, { "grad_norm": 0.16900146007537842, "learning_rate": 8.557883654761906e-06, "loss": 0.0057, "step": 16410 }, { "grad_norm": 0.1234845370054245, "learning_rate": 8.511685975055061e-06, "loss": 0.0034, "step": 16420 }, { "grad_norm": 0.09210969507694244, "learning_rate": 8.46560172281452e-06, "loss": 0.0034, "step": 16430 }, { "grad_norm": 0.11992689222097397, "learning_rate": 8.419631024032893e-06, "loss": 0.0039, "step": 16440 }, { "grad_norm": 0.14271646738052368, "learning_rate": 8.373774004392293e-06, "loss": 0.0034, "step": 16450 }, { "grad_norm": 0.1353248953819275, "learning_rate": 8.32803078926409e-06, "loss": 0.0047, "step": 16460 }, { "grad_norm": 0.14115722477436066, "learning_rate": 8.282401503708454e-06, "loss": 0.0047, "step": 16470 }, { "grad_norm": 0.1739596277475357, "learning_rate": 8.23688627247412e-06, "loss": 0.0053, "step": 16480 }, { "grad_norm": 0.19525675475597382, "learning_rate": 8.191485219998007e-06, "loss": 0.0062, "step": 16490 }, { "grad_norm": 0.17563992738723755, "learning_rate": 8.146198470404843e-06, "loss": 0.0047, "step": 16500 }, { "grad_norm": 0.17101417481899261, "learning_rate": 8.101026147506897e-06, "loss": 0.0038, "step": 16510 }, { "grad_norm": 0.18212486803531647, "learning_rate": 8.05596837480353e-06, "loss": 0.0038, "step": 16520 }, { "grad_norm": 0.13793456554412842, "learning_rate": 8.011025275480998e-06, "loss": 0.0041, "step": 16530 }, { "grad_norm": 0.12852394580841064, "learning_rate": 7.966196972412027e-06, "loss": 0.0041, "step": 16540 }, { "grad_norm": 0.11012130230665207, "learning_rate": 7.92148358815547e-06, "loss": 0.0046, "step": 16550 }, { "grad_norm": 0.1766318529844284, "learning_rate": 7.87688524495604e-06, "loss": 0.0041, "step": 16560 }, { "grad_norm": 0.16893427073955536, "learning_rate": 7.83240206474386e-06, "loss": 0.0044, "step": 16570 }, { "grad_norm": 0.25989559292793274, "learning_rate": 7.788034169134272e-06, "loss": 0.0053, "step": 16580 }, { "grad_norm": 0.1792820245027542, "learning_rate": 7.743781679427414e-06, "loss": 0.004, "step": 16590 }, { "grad_norm": 0.09085817635059357, "learning_rate": 7.699644716607895e-06, "loss": 0.004, "step": 16600 }, { "grad_norm": 0.18499186635017395, "learning_rate": 7.655623401344486e-06, "loss": 0.0062, "step": 16610 }, { "grad_norm": 0.11358114331960678, "learning_rate": 7.611717853989775e-06, "loss": 0.0048, "step": 16620 }, { "grad_norm": 0.1566750407218933, "learning_rate": 7.567928194579854e-06, "loss": 0.004, "step": 16630 }, { "grad_norm": 0.12232224643230438, "learning_rate": 7.524254542833997e-06, "loss": 0.0039, "step": 16640 }, { "grad_norm": 0.15875451266765594, "learning_rate": 7.480697018154286e-06, "loss": 0.0049, "step": 16650 }, { "grad_norm": 0.14348337054252625, "learning_rate": 7.437255739625332e-06, "loss": 0.0041, "step": 16660 }, { "grad_norm": 0.19628210365772247, "learning_rate": 7.393930826013923e-06, "loss": 0.0041, "step": 16670 }, { "grad_norm": 0.13403742015361786, "learning_rate": 7.350722395768722e-06, "loss": 0.0035, "step": 16680 }, { "grad_norm": 0.25174054503440857, "learning_rate": 7.307630567019963e-06, "loss": 0.0046, "step": 16690 }, { "grad_norm": 0.20793183147907257, "learning_rate": 7.264655457579e-06, "loss": 0.0069, "step": 16700 }, { "grad_norm": 0.12223444879055023, "learning_rate": 7.221797184938184e-06, "loss": 0.0033, "step": 16710 }, { "grad_norm": 0.13930709660053253, "learning_rate": 7.179055866270373e-06, "loss": 0.0072, "step": 16720 }, { "grad_norm": 0.08039425313472748, "learning_rate": 7.136431618428707e-06, "loss": 0.0038, "step": 16730 }, { "grad_norm": 0.07301311939954758, "learning_rate": 7.09392455794628e-06, "loss": 0.0031, "step": 16740 }, { "grad_norm": 0.10197556018829346, "learning_rate": 7.051534801035725e-06, "loss": 0.0035, "step": 16750 }, { "grad_norm": 0.21718019247055054, "learning_rate": 7.00926246358905e-06, "loss": 0.0045, "step": 16760 }, { "grad_norm": 0.14700235426425934, "learning_rate": 6.967107661177191e-06, "loss": 0.0053, "step": 16770 }, { "grad_norm": 0.12779207527637482, "learning_rate": 6.925070509049786e-06, "loss": 0.004, "step": 16780 }, { "grad_norm": 0.14491210877895355, "learning_rate": 6.883151122134812e-06, "loss": 0.0042, "step": 16790 }, { "grad_norm": 0.08013974130153656, "learning_rate": 6.8413496150382394e-06, "loss": 0.0053, "step": 16800 }, { "grad_norm": 0.18335309624671936, "learning_rate": 6.7996661020438165e-06, "loss": 0.0073, "step": 16810 }, { "grad_norm": 0.12072831392288208, "learning_rate": 6.758100697112662e-06, "loss": 0.0042, "step": 16820 }, { "grad_norm": 0.15185323357582092, "learning_rate": 6.716653513883026e-06, "loss": 0.0038, "step": 16830 }, { "grad_norm": 0.17190273106098175, "learning_rate": 6.675324665669913e-06, "loss": 0.0047, "step": 16840 }, { "grad_norm": 0.1471686214208603, "learning_rate": 6.634114265464803e-06, "loss": 0.0039, "step": 16850 }, { "grad_norm": 0.10595371574163437, "learning_rate": 6.59302242593538e-06, "loss": 0.0035, "step": 16860 }, { "grad_norm": 0.11482971906661987, "learning_rate": 6.552049259425141e-06, "loss": 0.0037, "step": 16870 }, { "grad_norm": 0.13058196008205414, "learning_rate": 6.511194877953181e-06, "loss": 0.0059, "step": 16880 }, { "grad_norm": 0.15087415277957916, "learning_rate": 6.470459393213813e-06, "loss": 0.0067, "step": 16890 }, { "grad_norm": 0.21332256495952606, "learning_rate": 6.429842916576279e-06, "loss": 0.0074, "step": 16900 }, { "grad_norm": 0.1266644150018692, "learning_rate": 6.389345559084503e-06, "loss": 0.0038, "step": 16910 }, { "grad_norm": 0.11508467048406601, "learning_rate": 6.348967431456682e-06, "loss": 0.0042, "step": 16920 }, { "grad_norm": 0.119036465883255, "learning_rate": 6.30870864408511e-06, "loss": 0.0034, "step": 16930 }, { "grad_norm": 0.1802026331424713, "learning_rate": 6.268569307035754e-06, "loss": 0.0043, "step": 16940 }, { "grad_norm": 0.14222128689289093, "learning_rate": 6.228549530048022e-06, "loss": 0.0034, "step": 16950 }, { "grad_norm": 0.11104612797498703, "learning_rate": 6.1886494225344814e-06, "loss": 0.0056, "step": 16960 }, { "grad_norm": 0.13774815201759338, "learning_rate": 6.148869093580479e-06, "loss": 0.006, "step": 16970 }, { "grad_norm": 0.19442835450172424, "learning_rate": 6.109208651943921e-06, "loss": 0.0034, "step": 16980 }, { "grad_norm": 0.13840056955814362, "learning_rate": 6.069668206054946e-06, "loss": 0.0052, "step": 16990 }, { "grad_norm": 0.2756652534008026, "learning_rate": 6.0302478640156145e-06, "loss": 0.0061, "step": 17000 }, { "grad_norm": 0.15658795833587646, "learning_rate": 5.990947733599644e-06, "loss": 0.0039, "step": 17010 }, { "grad_norm": 0.10911890864372253, "learning_rate": 5.951767922252105e-06, "loss": 0.0049, "step": 17020 }, { "grad_norm": 0.09619476646184921, "learning_rate": 5.912708537089068e-06, "loss": 0.004, "step": 17030 }, { "grad_norm": 0.1371019035577774, "learning_rate": 5.873769684897434e-06, "loss": 0.0035, "step": 17040 }, { "grad_norm": 0.1407761126756668, "learning_rate": 5.834951472134514e-06, "loss": 0.004, "step": 17050 }, { "grad_norm": 0.12553346157073975, "learning_rate": 5.796254004927832e-06, "loss": 0.0038, "step": 17060 }, { "grad_norm": 0.180441752076149, "learning_rate": 5.757677389074806e-06, "loss": 0.0066, "step": 17070 }, { "grad_norm": 0.16166085004806519, "learning_rate": 5.719221730042385e-06, "loss": 0.0046, "step": 17080 }, { "grad_norm": 0.07763278484344482, "learning_rate": 5.680887132966911e-06, "loss": 0.0028, "step": 17090 }, { "grad_norm": 0.08134908229112625, "learning_rate": 5.642673702653683e-06, "loss": 0.0058, "step": 17100 }, { "grad_norm": 0.26258209347724915, "learning_rate": 5.604581543576781e-06, "loss": 0.0053, "step": 17110 }, { "grad_norm": 0.18005384504795074, "learning_rate": 5.566610759878704e-06, "loss": 0.0045, "step": 17120 }, { "grad_norm": 0.13104481995105743, "learning_rate": 5.528761455370119e-06, "loss": 0.0065, "step": 17130 }, { "grad_norm": 0.12947246432304382, "learning_rate": 5.491033733529594e-06, "loss": 0.0079, "step": 17140 }, { "grad_norm": 0.1431690901517868, "learning_rate": 5.453427697503255e-06, "loss": 0.0036, "step": 17150 }, { "grad_norm": 0.1604069620370865, "learning_rate": 5.415943450104599e-06, "loss": 0.0045, "step": 17160 }, { "grad_norm": 0.14525003731250763, "learning_rate": 5.378581093814111e-06, "loss": 0.0039, "step": 17170 }, { "grad_norm": 0.14420951902866364, "learning_rate": 5.3413407307790375e-06, "loss": 0.004, "step": 17180 }, { "grad_norm": 0.12466453015804291, "learning_rate": 5.30422246281313e-06, "loss": 0.0036, "step": 17190 }, { "grad_norm": 0.20653589069843292, "learning_rate": 5.267226391396296e-06, "loss": 0.0041, "step": 17200 }, { "grad_norm": 0.09900219738483429, "learning_rate": 5.2303526176744e-06, "loss": 0.0063, "step": 17210 }, { "grad_norm": 0.08034467697143555, "learning_rate": 5.193601242458929e-06, "loss": 0.0035, "step": 17220 }, { "grad_norm": 0.09568826854228973, "learning_rate": 5.156972366226714e-06, "loss": 0.0036, "step": 17230 }, { "grad_norm": 0.11780333518981934, "learning_rate": 5.120466089119735e-06, "loss": 0.0044, "step": 17240 }, { "grad_norm": 0.11058741807937622, "learning_rate": 5.084082510944749e-06, "loss": 0.0076, "step": 17250 }, { "grad_norm": 0.202459454536438, "learning_rate": 5.047821731173058e-06, "loss": 0.0043, "step": 17260 }, { "grad_norm": 0.1387709677219391, "learning_rate": 5.011683848940274e-06, "loss": 0.0041, "step": 17270 }, { "grad_norm": 0.1834157258272171, "learning_rate": 4.975668963045954e-06, "loss": 0.0052, "step": 17280 }, { "grad_norm": 0.16564951837062836, "learning_rate": 4.9397771719534525e-06, "loss": 0.004, "step": 17290 }, { "grad_norm": 0.16223926842212677, "learning_rate": 4.904008573789548e-06, "loss": 0.0051, "step": 17300 }, { "grad_norm": 0.20775602757930756, "learning_rate": 4.8683632663442005e-06, "loss": 0.0061, "step": 17310 }, { "grad_norm": 0.09628815203905106, "learning_rate": 4.832841347070343e-06, "loss": 0.0029, "step": 17320 }, { "grad_norm": 0.12429912388324738, "learning_rate": 4.797442913083539e-06, "loss": 0.0043, "step": 17330 }, { "grad_norm": 0.1724526584148407, "learning_rate": 4.7621680611617596e-06, "loss": 0.0046, "step": 17340 }, { "grad_norm": 0.17094425857067108, "learning_rate": 4.727016887745095e-06, "loss": 0.005, "step": 17350 }, { "grad_norm": 0.1515076756477356, "learning_rate": 4.691989488935511e-06, "loss": 0.0046, "step": 17360 }, { "grad_norm": 0.15153151750564575, "learning_rate": 4.657085960496588e-06, "loss": 0.0039, "step": 17370 }, { "grad_norm": 0.0940147116780281, "learning_rate": 4.6223063978532265e-06, "loss": 0.0034, "step": 17380 }, { "grad_norm": 0.12322202324867249, "learning_rate": 4.587650896091439e-06, "loss": 0.0034, "step": 17390 }, { "grad_norm": 0.20023894309997559, "learning_rate": 4.553119549958035e-06, "loss": 0.0036, "step": 17400 }, { "grad_norm": 0.0966777428984642, "learning_rate": 4.518712453860385e-06, "loss": 0.0043, "step": 17410 }, { "grad_norm": 0.1916528344154358, "learning_rate": 4.484429701866205e-06, "loss": 0.0044, "step": 17420 }, { "grad_norm": 0.1058620885014534, "learning_rate": 4.4502713877031975e-06, "loss": 0.0068, "step": 17430 }, { "grad_norm": 0.09409364312887192, "learning_rate": 4.416237604758911e-06, "loss": 0.0031, "step": 17440 }, { "grad_norm": 0.10363601893186569, "learning_rate": 4.3823284460804025e-06, "loss": 0.0035, "step": 17450 }, { "grad_norm": 0.11878269910812378, "learning_rate": 4.348544004374011e-06, "loss": 0.0046, "step": 17460 }, { "grad_norm": 0.1685004085302353, "learning_rate": 4.314884372005123e-06, "loss": 0.004, "step": 17470 }, { "grad_norm": 0.11799660325050354, "learning_rate": 4.281349640997867e-06, "loss": 0.0043, "step": 17480 }, { "grad_norm": 0.10389797389507294, "learning_rate": 4.247939903034942e-06, "loss": 0.0036, "step": 17490 }, { "grad_norm": 0.11342811584472656, "learning_rate": 4.214655249457284e-06, "loss": 0.0027, "step": 17500 }, { "grad_norm": 0.08605656772851944, "learning_rate": 4.181495771263855e-06, "loss": 0.0033, "step": 17510 }, { "grad_norm": 0.07319090515375137, "learning_rate": 4.148461559111427e-06, "loss": 0.0036, "step": 17520 }, { "grad_norm": 0.14404094219207764, "learning_rate": 4.115552703314252e-06, "loss": 0.0045, "step": 17530 }, { "grad_norm": 0.10764741897583008, "learning_rate": 4.082769293843886e-06, "loss": 0.0057, "step": 17540 }, { "grad_norm": 0.15143528580665588, "learning_rate": 4.050111420328939e-06, "loss": 0.0059, "step": 17550 }, { "grad_norm": 0.18116922676563263, "learning_rate": 4.017579172054764e-06, "loss": 0.0048, "step": 17560 }, { "grad_norm": 0.11765165627002716, "learning_rate": 3.985172637963308e-06, "loss": 0.0038, "step": 17570 }, { "grad_norm": 0.15436911582946777, "learning_rate": 3.952891906652784e-06, "loss": 0.0051, "step": 17580 }, { "grad_norm": 0.09385086596012115, "learning_rate": 3.920737066377478e-06, "loss": 0.0031, "step": 17590 }, { "grad_norm": 0.05163466930389404, "learning_rate": 3.888708205047509e-06, "loss": 0.0026, "step": 17600 }, { "grad_norm": 0.06750795245170593, "learning_rate": 3.856805410228542e-06, "loss": 0.0054, "step": 17610 }, { "grad_norm": 0.08441822230815887, "learning_rate": 3.82502876914162e-06, "loss": 0.0035, "step": 17620 }, { "grad_norm": 0.19725516438484192, "learning_rate": 3.7933783686628586e-06, "loss": 0.0037, "step": 17630 }, { "grad_norm": 0.1849585324525833, "learning_rate": 3.7618542953232306e-06, "loss": 0.0043, "step": 17640 }, { "grad_norm": 0.08998493105173111, "learning_rate": 3.7304566353083658e-06, "loss": 0.0034, "step": 17650 }, { "grad_norm": 0.12049571424722672, "learning_rate": 3.6991854744582555e-06, "loss": 0.0046, "step": 17660 }, { "grad_norm": 0.17697420716285706, "learning_rate": 3.6680408982670777e-06, "loss": 0.0069, "step": 17670 }, { "grad_norm": 0.1326104998588562, "learning_rate": 3.637022991882899e-06, "loss": 0.0036, "step": 17680 }, { "grad_norm": 0.17959579825401306, "learning_rate": 3.606131840107485e-06, "loss": 0.0031, "step": 17690 }, { "grad_norm": 0.08260238915681839, "learning_rate": 3.575367527396084e-06, "loss": 0.0073, "step": 17700 }, { "grad_norm": 0.07719326764345169, "learning_rate": 3.5447301378571386e-06, "loss": 0.0039, "step": 17710 }, { "grad_norm": 0.19096577167510986, "learning_rate": 3.514219755252113e-06, "loss": 0.0041, "step": 17720 }, { "grad_norm": 0.14167147874832153, "learning_rate": 3.4838364629952213e-06, "loss": 0.0051, "step": 17730 }, { "grad_norm": 0.11633794009685516, "learning_rate": 3.4535803441532123e-06, "loss": 0.0049, "step": 17740 }, { "grad_norm": 0.11027508974075317, "learning_rate": 3.4234514814451836e-06, "loss": 0.0034, "step": 17750 }, { "grad_norm": 0.1757984161376953, "learning_rate": 3.393449957242273e-06, "loss": 0.0043, "step": 17760 }, { "grad_norm": 0.12752801179885864, "learning_rate": 3.363575853567524e-06, "loss": 0.0062, "step": 17770 }, { "grad_norm": 0.1456461250782013, "learning_rate": 3.3338292520955826e-06, "loss": 0.0046, "step": 17780 }, { "grad_norm": 0.15725676715373993, "learning_rate": 3.304210234152516e-06, "loss": 0.0041, "step": 17790 }, { "grad_norm": 0.09031637758016586, "learning_rate": 3.2747188807155993e-06, "loss": 0.0065, "step": 17800 }, { "grad_norm": 0.12988491356372833, "learning_rate": 3.2453552724130643e-06, "loss": 0.0041, "step": 17810 }, { "grad_norm": 0.2224876880645752, "learning_rate": 3.216119489523889e-06, "loss": 0.0035, "step": 17820 }, { "grad_norm": 0.09608910232782364, "learning_rate": 3.1870116119775917e-06, "loss": 0.0039, "step": 17830 }, { "grad_norm": 0.18000908195972443, "learning_rate": 3.158031719353999e-06, "loss": 0.0054, "step": 17840 }, { "grad_norm": 0.1523892730474472, "learning_rate": 3.1291798908830273e-06, "loss": 0.0098, "step": 17850 }, { "grad_norm": 0.13003237545490265, "learning_rate": 3.1004562054444853e-06, "loss": 0.0035, "step": 17860 }, { "grad_norm": 0.10257936269044876, "learning_rate": 3.071860741567806e-06, "loss": 0.0033, "step": 17870 }, { "grad_norm": 0.1925061047077179, "learning_rate": 3.04339357743193e-06, "loss": 0.005, "step": 17880 }, { "grad_norm": 0.0762476995587349, "learning_rate": 3.0150547908649628e-06, "loss": 0.0029, "step": 17890 }, { "grad_norm": 0.09263581782579422, "learning_rate": 2.9868444593440957e-06, "loss": 0.005, "step": 17900 }, { "grad_norm": 0.11614616215229034, "learning_rate": 2.9587626599952846e-06, "loss": 0.0032, "step": 17910 }, { "grad_norm": 0.10844535380601883, "learning_rate": 2.930809469593082e-06, "loss": 0.0073, "step": 17920 }, { "grad_norm": 0.19387802481651306, "learning_rate": 2.9029849645604733e-06, "loss": 0.0044, "step": 17930 }, { "grad_norm": 0.20903635025024414, "learning_rate": 2.8752892209685632e-06, "loss": 0.0058, "step": 17940 }, { "grad_norm": 0.0903766006231308, "learning_rate": 2.847722314536483e-06, "loss": 0.0031, "step": 17950 }, { "grad_norm": 0.20239296555519104, "learning_rate": 2.820284320631078e-06, "loss": 0.0045, "step": 17960 }, { "grad_norm": 0.11584734171628952, "learning_rate": 2.792975314266788e-06, "loss": 0.0074, "step": 17970 }, { "grad_norm": 0.10949089378118515, "learning_rate": 2.7657953701054007e-06, "loss": 0.0035, "step": 17980 }, { "grad_norm": 0.10295064747333527, "learning_rate": 2.7387445624558306e-06, "loss": 0.0049, "step": 17990 }, { "grad_norm": 0.17109937965869904, "learning_rate": 2.7118229652739747e-06, "loss": 0.0047, "step": 18000 }, { "grad_norm": 0.14163026213645935, "learning_rate": 2.6850306521624236e-06, "loss": 0.004, "step": 18010 }, { "grad_norm": 0.09707952290773392, "learning_rate": 2.6583676963703507e-06, "loss": 0.006, "step": 18020 }, { "grad_norm": 0.13564935326576233, "learning_rate": 2.631834170793268e-06, "loss": 0.0059, "step": 18030 }, { "grad_norm": 0.10735955834388733, "learning_rate": 2.6054301479728036e-06, "loss": 0.0055, "step": 18040 }, { "grad_norm": 0.0932333767414093, "learning_rate": 2.579155700096575e-06, "loss": 0.0067, "step": 18050 }, { "grad_norm": 0.09174794703722, "learning_rate": 2.5530108989978873e-06, "loss": 0.0051, "step": 18060 }, { "grad_norm": 0.08465638756752014, "learning_rate": 2.5269958161556416e-06, "loss": 0.0038, "step": 18070 }, { "grad_norm": 0.10658536851406097, "learning_rate": 2.5011105226940888e-06, "loss": 0.0035, "step": 18080 }, { "grad_norm": 0.07544179260730743, "learning_rate": 2.4753550893826248e-06, "loss": 0.0034, "step": 18090 }, { "grad_norm": 0.08290700614452362, "learning_rate": 2.4497295866356296e-06, "loss": 0.0041, "step": 18100 }, { "grad_norm": 0.08353956043720245, "learning_rate": 2.424234084512228e-06, "loss": 0.0039, "step": 18110 }, { "grad_norm": 0.08789733052253723, "learning_rate": 2.3988686527161687e-06, "loss": 0.0037, "step": 18120 }, { "grad_norm": 0.08249858021736145, "learning_rate": 2.373633360595573e-06, "loss": 0.0027, "step": 18130 }, { "grad_norm": 0.06505706161260605, "learning_rate": 2.3485282771427585e-06, "loss": 0.0051, "step": 18140 }, { "grad_norm": 0.09220980852842331, "learning_rate": 2.3235534709940665e-06, "loss": 0.0028, "step": 18150 }, { "grad_norm": 0.0920565277338028, "learning_rate": 2.2987090104296617e-06, "loss": 0.0049, "step": 18160 }, { "grad_norm": 0.08199271559715271, "learning_rate": 2.273994963373355e-06, "loss": 0.0031, "step": 18170 }, { "grad_norm": 0.18866243958473206, "learning_rate": 2.249411397392409e-06, "loss": 0.0054, "step": 18180 }, { "grad_norm": 0.15900681912899017, "learning_rate": 2.2249583796973506e-06, "loss": 0.0046, "step": 18190 }, { "grad_norm": 0.08085703104734421, "learning_rate": 2.200635977141796e-06, "loss": 0.0028, "step": 18200 }, { "grad_norm": 0.09156780689954758, "learning_rate": 2.17644425622226e-06, "loss": 0.0027, "step": 18210 }, { "grad_norm": 0.12192914634943008, "learning_rate": 2.152383283077991e-06, "loss": 0.004, "step": 18220 }, { "grad_norm": 0.06557256728410721, "learning_rate": 2.128453123490781e-06, "loss": 0.005, "step": 18230 }, { "grad_norm": 0.15613125264644623, "learning_rate": 2.1046538428847462e-06, "loss": 0.0038, "step": 18240 }, { "grad_norm": 0.21774674952030182, "learning_rate": 2.0809855063262273e-06, "loss": 0.0047, "step": 18250 }, { "grad_norm": 0.05827488377690315, "learning_rate": 2.057448178523558e-06, "loss": 0.0036, "step": 18260 }, { "grad_norm": 0.08559198677539825, "learning_rate": 2.034041923826885e-06, "loss": 0.0039, "step": 18270 }, { "grad_norm": 0.08992449194192886, "learning_rate": 2.0107668062280204e-06, "loss": 0.0032, "step": 18280 }, { "grad_norm": 0.12441548705101013, "learning_rate": 1.9876228893602357e-06, "loss": 0.0051, "step": 18290 }, { "grad_norm": 0.18422020971775055, "learning_rate": 1.9646102364981266e-06, "loss": 0.0046, "step": 18300 }, { "grad_norm": 0.06744041293859482, "learning_rate": 1.9417289105574053e-06, "loss": 0.0037, "step": 18310 }, { "grad_norm": 0.0947098433971405, "learning_rate": 1.9189789740947427e-06, "loss": 0.0025, "step": 18320 }, { "grad_norm": 0.1519598811864853, "learning_rate": 1.896360489307597e-06, "loss": 0.0052, "step": 18330 }, { "grad_norm": 0.09345407038927078, "learning_rate": 1.8738735180340362e-06, "loss": 0.0033, "step": 18340 }, { "grad_norm": 0.04639227315783501, "learning_rate": 1.8515181217525824e-06, "loss": 0.003, "step": 18350 }, { "grad_norm": 0.08520929515361786, "learning_rate": 1.8292943615820457e-06, "loss": 0.0044, "step": 18360 }, { "grad_norm": 0.27233782410621643, "learning_rate": 1.8072022982813296e-06, "loss": 0.0056, "step": 18370 }, { "grad_norm": 0.21458378434181213, "learning_rate": 1.7852419922492925e-06, "loss": 0.0036, "step": 18380 }, { "grad_norm": 0.07499372959136963, "learning_rate": 1.763413503524569e-06, "loss": 0.0035, "step": 18390 }, { "grad_norm": 0.06435967236757278, "learning_rate": 1.7417168917854165e-06, "loss": 0.0049, "step": 18400 }, { "grad_norm": 0.08679071813821793, "learning_rate": 1.720152216349552e-06, "loss": 0.0066, "step": 18410 }, { "grad_norm": 0.08633027225732803, "learning_rate": 1.6987195361739595e-06, "loss": 0.004, "step": 18420 }, { "grad_norm": 0.09986413270235062, "learning_rate": 1.6774189098547832e-06, "loss": 0.0041, "step": 18430 }, { "grad_norm": 0.18124908208847046, "learning_rate": 1.6562503956271069e-06, "loss": 0.0043, "step": 18440 }, { "grad_norm": 0.07088620960712433, "learning_rate": 1.6352140513648417e-06, "loss": 0.0034, "step": 18450 }, { "grad_norm": 0.11781052500009537, "learning_rate": 1.6143099345805712e-06, "loss": 0.0049, "step": 18460 }, { "grad_norm": 0.09301770478487015, "learning_rate": 1.5935381024253293e-06, "loss": 0.0031, "step": 18470 }, { "grad_norm": 0.24067574739456177, "learning_rate": 1.572898611688517e-06, "loss": 0.0032, "step": 18480 }, { "grad_norm": 0.12292911857366562, "learning_rate": 1.5523915187977133e-06, "loss": 0.0039, "step": 18490 }, { "grad_norm": 0.16176709532737732, "learning_rate": 1.532016879818532e-06, "loss": 0.005, "step": 18500 }, { "grad_norm": 0.17161191999912262, "learning_rate": 1.51177475045447e-06, "loss": 0.0074, "step": 18510 }, { "grad_norm": 0.08660425245761871, "learning_rate": 1.4916651860467035e-06, "loss": 0.0044, "step": 18520 }, { "grad_norm": 0.11082902550697327, "learning_rate": 1.471688241574043e-06, "loss": 0.0043, "step": 18530 }, { "grad_norm": 0.13112390041351318, "learning_rate": 1.451843971652672e-06, "loss": 0.007, "step": 18540 }, { "grad_norm": 0.06571222096681595, "learning_rate": 1.432132430536076e-06, "loss": 0.0033, "step": 18550 }, { "grad_norm": 0.09628504514694214, "learning_rate": 1.412553672114869e-06, "loss": 0.006, "step": 18560 }, { "grad_norm": 0.07170583307743073, "learning_rate": 1.3931077499166056e-06, "loss": 0.0039, "step": 18570 }, { "grad_norm": 0.06113622337579727, "learning_rate": 1.3737947171057085e-06, "loss": 0.0036, "step": 18580 }, { "grad_norm": 0.1881685107946396, "learning_rate": 1.3546146264832582e-06, "loss": 0.0046, "step": 18590 }, { "grad_norm": 0.16483014822006226, "learning_rate": 1.3355675304869086e-06, "loss": 0.0042, "step": 18600 }, { "grad_norm": 0.11587239056825638, "learning_rate": 1.3166534811906827e-06, "loss": 0.0042, "step": 18610 }, { "grad_norm": 0.10228906571865082, "learning_rate": 1.2978725303048666e-06, "loss": 0.0029, "step": 18620 }, { "grad_norm": 0.06450185179710388, "learning_rate": 1.2792247291758762e-06, "loss": 0.0043, "step": 18630 }, { "grad_norm": 0.04407062381505966, "learning_rate": 1.2607101287860635e-06, "loss": 0.0026, "step": 18640 }, { "grad_norm": 0.06498788297176361, "learning_rate": 1.2423287797536654e-06, "loss": 0.0032, "step": 18650 }, { "grad_norm": 0.11582018435001373, "learning_rate": 1.2240807323325776e-06, "loss": 0.004, "step": 18660 }, { "grad_norm": 0.13144077360630035, "learning_rate": 1.205966036412254e-06, "loss": 0.0095, "step": 18670 }, { "grad_norm": 0.08476530760526657, "learning_rate": 1.1879847415175949e-06, "loss": 0.0048, "step": 18680 }, { "grad_norm": 0.08811729401350021, "learning_rate": 1.1701368968087712e-06, "loss": 0.004, "step": 18690 }, { "grad_norm": 0.06819802522659302, "learning_rate": 1.1524225510811116e-06, "loss": 0.0057, "step": 18700 }, { "grad_norm": 0.051221612840890884, "learning_rate": 1.1348417527649535e-06, "loss": 0.0041, "step": 18710 }, { "grad_norm": 0.15694133937358856, "learning_rate": 1.1173945499255268e-06, "loss": 0.003, "step": 18720 }, { "grad_norm": 0.043205779045820236, "learning_rate": 1.1000809902628307e-06, "loss": 0.0032, "step": 18730 }, { "grad_norm": 0.16620707511901855, "learning_rate": 1.082901121111468e-06, "loss": 0.004, "step": 18740 }, { "grad_norm": 0.06932929158210754, "learning_rate": 1.0658549894405456e-06, "loss": 0.003, "step": 18750 }, { "grad_norm": 0.06264208257198334, "learning_rate": 1.0489426418535342e-06, "loss": 0.0034, "step": 18760 }, { "grad_norm": 0.1080767884850502, "learning_rate": 1.0321641245881474e-06, "loss": 0.0032, "step": 18770 }, { "grad_norm": 0.12831155955791473, "learning_rate": 1.015519483516214e-06, "loss": 0.0041, "step": 18780 }, { "grad_norm": 0.1368539184331894, "learning_rate": 9.990087641435443e-07, "loss": 0.0046, "step": 18790 }, { "grad_norm": 0.11672380566596985, "learning_rate": 9.826320116098132e-07, "loss": 0.0039, "step": 18800 }, { "grad_norm": 0.08051031827926636, "learning_rate": 9.663892706884447e-07, "loss": 0.0038, "step": 18810 }, { "grad_norm": 0.06871426105499268, "learning_rate": 9.502805857864616e-07, "loss": 0.004, "step": 18820 }, { "grad_norm": 0.13891440629959106, "learning_rate": 9.34306000944396e-07, "loss": 0.003, "step": 18830 }, { "grad_norm": 0.056919511407613754, "learning_rate": 9.184655598361624e-07, "loss": 0.0029, "step": 18840 }, { "grad_norm": 0.0948851928114891, "learning_rate": 9.027593057689076e-07, "loss": 0.0038, "step": 18850 }, { "grad_norm": 0.09866155683994293, "learning_rate": 8.871872816829441e-07, "loss": 0.0026, "step": 18860 }, { "grad_norm": 0.06725290417671204, "learning_rate": 8.717495301515777e-07, "loss": 0.0032, "step": 18870 }, { "grad_norm": 0.07159546762704849, "learning_rate": 8.564460933810415e-07, "loss": 0.0041, "step": 18880 }, { "grad_norm": 0.06492818892002106, "learning_rate": 8.412770132103453e-07, "loss": 0.0029, "step": 18890 }, { "grad_norm": 0.04782559350132942, "learning_rate": 8.262423311111711e-07, "loss": 0.0046, "step": 18900 }, { "grad_norm": 0.19077003002166748, "learning_rate": 8.113420881877665e-07, "loss": 0.0033, "step": 18910 }, { "grad_norm": 0.1281125396490097, "learning_rate": 7.965763251768288e-07, "loss": 0.0051, "step": 18920 }, { "grad_norm": 0.05809188634157181, "learning_rate": 7.819450824473995e-07, "loss": 0.0038, "step": 18930 }, { "grad_norm": 0.06381773203611374, "learning_rate": 7.674484000007198e-07, "loss": 0.0032, "step": 18940 }, { "grad_norm": 0.06802604347467422, "learning_rate": 7.530863174701752e-07, "loss": 0.0028, "step": 18950 }, { "grad_norm": 0.07965709269046783, "learning_rate": 7.38858874121151e-07, "loss": 0.0038, "step": 18960 }, { "grad_norm": 0.05135740339756012, "learning_rate": 7.247661088509328e-07, "loss": 0.0035, "step": 18970 }, { "grad_norm": 0.04405200481414795, "learning_rate": 7.108080601886002e-07, "loss": 0.0031, "step": 18980 }, { "grad_norm": 0.08332010358572006, "learning_rate": 6.969847662949336e-07, "loss": 0.0023, "step": 18990 }, { "grad_norm": 0.05207156017422676, "learning_rate": 6.832962649622798e-07, "loss": 0.0027, "step": 19000 }, { "grad_norm": 0.05658917501568794, "learning_rate": 6.697425936144863e-07, "loss": 0.0031, "step": 19010 }, { "grad_norm": 0.10787039995193481, "learning_rate": 6.563237893067731e-07, "loss": 0.0045, "step": 19020 }, { "grad_norm": 0.11102147400379181, "learning_rate": 6.430398887256328e-07, "loss": 0.005, "step": 19030 }, { "grad_norm": 0.08387942612171173, "learning_rate": 6.298909281887478e-07, "loss": 0.003, "step": 19040 }, { "grad_norm": 0.047253288328647614, "learning_rate": 6.168769436448673e-07, "loss": 0.003, "step": 19050 }, { "grad_norm": 0.08331707119941711, "learning_rate": 6.03997970673742e-07, "loss": 0.003, "step": 19060 }, { "grad_norm": 0.05192408338189125, "learning_rate": 5.912540444859782e-07, "loss": 0.0037, "step": 19070 }, { "grad_norm": 0.11672984063625336, "learning_rate": 5.786451999229837e-07, "loss": 0.0035, "step": 19080 }, { "grad_norm": 0.085425466299057, "learning_rate": 5.661714714568722e-07, "loss": 0.0031, "step": 19090 }, { "grad_norm": 0.11592360585927963, "learning_rate": 5.538328931903259e-07, "loss": 0.0034, "step": 19100 }, { "grad_norm": 0.049993764609098434, "learning_rate": 5.416294988565551e-07, "loss": 0.003, "step": 19110 }, { "grad_norm": 0.06833920627832413, "learning_rate": 5.29561321819172e-07, "loss": 0.0035, "step": 19120 }, { "grad_norm": 0.06917779892683029, "learning_rate": 5.176283950721061e-07, "loss": 0.0049, "step": 19130 }, { "grad_norm": 0.089486263692379, "learning_rate": 5.058307512395332e-07, "loss": 0.004, "step": 19140 }, { "grad_norm": 0.07048797607421875, "learning_rate": 4.941684225757526e-07, "loss": 0.0038, "step": 19150 }, { "grad_norm": 0.1524529904127121, "learning_rate": 4.826414409651314e-07, "loss": 0.0044, "step": 19160 }, { "grad_norm": 0.12049482017755508, "learning_rate": 4.712498379219943e-07, "loss": 0.006, "step": 19170 }, { "grad_norm": 0.10811007767915726, "learning_rate": 4.599936445905506e-07, "loss": 0.0033, "step": 19180 }, { "grad_norm": 0.07990206032991409, "learning_rate": 4.4887289174480594e-07, "loss": 0.0034, "step": 19190 }, { "grad_norm": 0.11032144725322723, "learning_rate": 4.378876097884621e-07, "loss": 0.0036, "step": 19200 }, { "grad_norm": 0.05706889182329178, "learning_rate": 4.2703782875487264e-07, "loss": 0.0027, "step": 19210 }, { "grad_norm": 0.09622795134782791, "learning_rate": 4.163235783069208e-07, "loss": 0.0029, "step": 19220 }, { "grad_norm": 0.10857948660850525, "learning_rate": 4.057448877369585e-07, "loss": 0.0046, "step": 19230 }, { "grad_norm": 0.08431418985128403, "learning_rate": 3.9530178596672295e-07, "loss": 0.0034, "step": 19240 }, { "grad_norm": 0.06911571323871613, "learning_rate": 3.849943015472479e-07, "loss": 0.004, "step": 19250 }, { "grad_norm": 0.042932331562042236, "learning_rate": 3.748224626588137e-07, "loss": 0.0032, "step": 19260 }, { "grad_norm": 0.0972372442483902, "learning_rate": 3.647862971108307e-07, "loss": 0.0029, "step": 19270 }, { "grad_norm": 0.1620805710554123, "learning_rate": 3.5488583234179473e-07, "loss": 0.005, "step": 19280 }, { "grad_norm": 0.18250656127929688, "learning_rate": 3.4512109541920413e-07, "loss": 0.0036, "step": 19290 }, { "grad_norm": 0.0328088216483593, "learning_rate": 3.354921130394706e-07, "loss": 0.0029, "step": 19300 }, { "grad_norm": 0.06825669854879379, "learning_rate": 3.259989115278639e-07, "loss": 0.0032, "step": 19310 }, { "grad_norm": 0.037347208708524704, "learning_rate": 3.1664151683843403e-07, "loss": 0.0029, "step": 19320 }, { "grad_norm": 0.06825114786624908, "learning_rate": 3.074199545539447e-07, "loss": 0.0037, "step": 19330 }, { "grad_norm": 0.048928894102573395, "learning_rate": 2.983342498857955e-07, "loss": 0.004, "step": 19340 }, { "grad_norm": 0.12561197578907013, "learning_rate": 2.893844276739499e-07, "loss": 0.0048, "step": 19350 }, { "grad_norm": 0.042123641818761826, "learning_rate": 2.8057051238688514e-07, "loss": 0.0035, "step": 19360 }, { "grad_norm": 0.1048402264714241, "learning_rate": 2.71892528121509e-07, "loss": 0.0032, "step": 19370 }, { "grad_norm": 0.07271413505077362, "learning_rate": 2.633504986030988e-07, "loss": 0.0035, "step": 19380 }, { "grad_norm": 0.13717639446258545, "learning_rate": 2.549444471852347e-07, "loss": 0.0042, "step": 19390 }, { "grad_norm": 0.05676209554076195, "learning_rate": 2.4667439684974423e-07, "loss": 0.003, "step": 19400 }, { "grad_norm": 0.13559632003307343, "learning_rate": 2.3854037020662467e-07, "loss": 0.0056, "step": 19410 }, { "grad_norm": 0.10882473737001419, "learning_rate": 2.3054238949399288e-07, "loss": 0.003, "step": 19420 }, { "grad_norm": 0.04371321573853493, "learning_rate": 2.2268047657802993e-07, "loss": 0.0031, "step": 19430 }, { "grad_norm": 0.09959394484758377, "learning_rate": 2.149546529529034e-07, "loss": 0.0043, "step": 19440 }, { "grad_norm": 0.06530318409204483, "learning_rate": 2.0736493974071736e-07, "loss": 0.003, "step": 19450 }, { "grad_norm": 0.07622455805540085, "learning_rate": 1.9991135769145686e-07, "loss": 0.003, "step": 19460 }, { "grad_norm": 0.04550927132368088, "learning_rate": 1.9259392718293245e-07, "loss": 0.0044, "step": 19470 }, { "grad_norm": 0.06660426408052444, "learning_rate": 1.8541266822072467e-07, "loss": 0.0058, "step": 19480 }, { "grad_norm": 0.061952270567417145, "learning_rate": 1.7836760043811184e-07, "loss": 0.0056, "step": 19490 }, { "grad_norm": 0.12157411873340607, "learning_rate": 1.7145874309604792e-07, "loss": 0.0034, "step": 19500 }, { "grad_norm": 0.06773949414491653, "learning_rate": 1.6468611508308474e-07, "loss": 0.0047, "step": 19510 }, { "grad_norm": 0.0389065183699131, "learning_rate": 1.5804973491532204e-07, "loss": 0.003, "step": 19520 }, { "grad_norm": 0.0565793402493, "learning_rate": 1.5154962073637424e-07, "loss": 0.004, "step": 19530 }, { "grad_norm": 0.058748792856931686, "learning_rate": 1.4518579031730372e-07, "loss": 0.0034, "step": 19540 }, { "grad_norm": 0.10500644892454147, "learning_rate": 1.389582610565876e-07, "loss": 0.0077, "step": 19550 }, { "grad_norm": 0.1324421912431717, "learning_rate": 1.3286704998003995e-07, "loss": 0.005, "step": 19560 }, { "grad_norm": 0.11811163276433945, "learning_rate": 1.2691217374080632e-07, "loss": 0.005, "step": 19570 }, { "grad_norm": 0.03684741258621216, "learning_rate": 1.2109364861929705e-07, "loss": 0.0043, "step": 19580 }, { "grad_norm": 0.03240930289030075, "learning_rate": 1.1541149052312628e-07, "loss": 0.0031, "step": 19590 }, { "grad_norm": 0.043710801750421524, "learning_rate": 1.0986571498710074e-07, "loss": 0.0032, "step": 19600 }, { "grad_norm": 0.05616920813918114, "learning_rate": 1.0445633717316438e-07, "loss": 0.0024, "step": 19610 }, { "grad_norm": 0.06716179102659225, "learning_rate": 9.918337187034277e-08, "loss": 0.0034, "step": 19620 }, { "grad_norm": 0.06252413988113403, "learning_rate": 9.404683349472643e-08, "loss": 0.0034, "step": 19630 }, { "grad_norm": 0.05004303902387619, "learning_rate": 8.904673608940983e-08, "loss": 0.0027, "step": 19640 }, { "grad_norm": 0.045323677361011505, "learning_rate": 8.418309332447471e-08, "loss": 0.0025, "step": 19650 }, { "grad_norm": 0.06611169874668121, "learning_rate": 7.945591849692902e-08, "loss": 0.0041, "step": 19660 }, { "grad_norm": 0.07558487355709076, "learning_rate": 7.486522453069578e-08, "loss": 0.0044, "step": 19670 }, { "grad_norm": 0.04437953606247902, "learning_rate": 7.041102397655208e-08, "loss": 0.0028, "step": 19680 }, { "grad_norm": 0.0719234049320221, "learning_rate": 6.609332901210685e-08, "loss": 0.0031, "step": 19690 }, { "grad_norm": 0.1436450481414795, "learning_rate": 6.191215144178419e-08, "loss": 0.0106, "step": 19700 }, { "grad_norm": 0.04914439842104912, "learning_rate": 5.786750269675678e-08, "loss": 0.0033, "step": 19710 }, { "grad_norm": 0.06302205473184586, "learning_rate": 5.395939383494031e-08, "loss": 0.0037, "step": 19720 }, { "grad_norm": 0.05430988967418671, "learning_rate": 5.018783554095463e-08, "loss": 0.0047, "step": 19730 }, { "grad_norm": 0.12179956585168839, "learning_rate": 4.655283812610156e-08, "loss": 0.0063, "step": 19740 }, { "grad_norm": 0.04501037672162056, "learning_rate": 4.305441152831491e-08, "loss": 0.0042, "step": 19750 }, { "grad_norm": 0.09513222426176071, "learning_rate": 3.9692565312171584e-08, "loss": 0.0052, "step": 19760 }, { "grad_norm": 0.04066900163888931, "learning_rate": 3.6467308668824975e-08, "loss": 0.0035, "step": 19770 }, { "grad_norm": 0.06860118359327316, "learning_rate": 3.3378650416004964e-08, "loss": 0.0064, "step": 19780 }, { "grad_norm": 0.1298064887523651, "learning_rate": 3.042659899797906e-08, "loss": 0.0063, "step": 19790 }, { "grad_norm": 0.10037373751401901, "learning_rate": 2.76111624855524e-08, "loss": 0.0034, "step": 19800 }, { "grad_norm": 0.044566810131073, "learning_rate": 2.4932348576017784e-08, "loss": 0.0037, "step": 19810 }, { "grad_norm": 0.14785318076610565, "learning_rate": 2.239016459314458e-08, "loss": 0.0036, "step": 19820 }, { "grad_norm": 0.03427545353770256, "learning_rate": 1.9984617487173174e-08, "loss": 0.005, "step": 19830 }, { "grad_norm": 0.05439184606075287, "learning_rate": 1.7715713834776105e-08, "loss": 0.0065, "step": 19840 }, { "grad_norm": 0.19172483682632446, "learning_rate": 1.5583459839046964e-08, "loss": 0.005, "step": 19850 }, { "grad_norm": 0.09647603332996368, "learning_rate": 1.3587861329489304e-08, "loss": 0.0046, "step": 19860 }, { "grad_norm": 0.0617683045566082, "learning_rate": 1.1728923761994415e-08, "loss": 0.0039, "step": 19870 }, { "grad_norm": 0.06156707555055618, "learning_rate": 1.0006652218819135e-08, "loss": 0.0032, "step": 19880 }, { "grad_norm": 0.03269469738006592, "learning_rate": 8.421051408596947e-09, "loss": 0.0041, "step": 19890 }, { "grad_norm": 0.13029080629348755, "learning_rate": 6.972125666299123e-09, "loss": 0.0036, "step": 19900 }, { "grad_norm": 0.046240948140621185, "learning_rate": 5.659878953229169e-09, "loss": 0.0033, "step": 19910 }, { "grad_norm": 0.052675407379865646, "learning_rate": 4.48431485701728e-09, "loss": 0.0032, "step": 19920 }, { "grad_norm": 0.03239002451300621, "learning_rate": 3.4454365916203322e-09, "loss": 0.0027, "step": 19930 }, { "grad_norm": 0.08773598074913025, "learning_rate": 2.5432469972830332e-09, "loss": 0.0031, "step": 19940 }, { "grad_norm": 0.05771934613585472, "learning_rate": 1.7777485405601203e-09, "loss": 0.0044, "step": 19950 }, { "grad_norm": 0.07388681173324585, "learning_rate": 1.1489433142941597e-09, "loss": 0.004, "step": 19960 }, { "grad_norm": 0.08763538300991058, "learning_rate": 6.568330376210963e-10, "loss": 0.0045, "step": 19970 }, { "grad_norm": 0.12037558108568192, "learning_rate": 3.0141905594249787e-10, "loss": 0.0029, "step": 19980 }, { "grad_norm": 0.09306905418634415, "learning_rate": 8.270234094776008e-11, "loss": 0.0039, "step": 19990 }, { "grad_norm": 0.059976037591695786, "learning_rate": 6.834906085551041e-13, "loss": 0.0027, "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": 32, "trial_name": null, "trial_params": null }