{ "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": 5.806057929992676, "learning_rate": 9e-07, "loss": 1.4801, "step": 10 }, { "grad_norm": 3.5745508670806885, "learning_rate": 1.9e-06, "loss": 1.4363, "step": 20 }, { "grad_norm": 1.5968669652938843, "learning_rate": 2.9e-06, "loss": 1.2855, "step": 30 }, { "grad_norm": 0.7521708011627197, "learning_rate": 3.9e-06, "loss": 1.1672, "step": 40 }, { "grad_norm": 0.5106793642044067, "learning_rate": 4.9000000000000005e-06, "loss": 1.1121, "step": 50 }, { "grad_norm": 0.43003425002098083, "learning_rate": 5.9e-06, "loss": 1.1043, "step": 60 }, { "grad_norm": 0.4448665380477905, "learning_rate": 6.900000000000001e-06, "loss": 1.0824, "step": 70 }, { "grad_norm": 0.4131520092487335, "learning_rate": 7.9e-06, "loss": 1.0637, "step": 80 }, { "grad_norm": 0.3829840421676636, "learning_rate": 8.9e-06, "loss": 1.057, "step": 90 }, { "grad_norm": 0.4175017476081848, "learning_rate": 9.900000000000002e-06, "loss": 1.0418, "step": 100 }, { "grad_norm": 0.4737490117549896, "learning_rate": 1.09e-05, "loss": 1.0402, "step": 110 }, { "grad_norm": 0.5096083283424377, "learning_rate": 1.19e-05, "loss": 1.0383, "step": 120 }, { "grad_norm": 0.8121296167373657, "learning_rate": 1.29e-05, "loss": 1.0246, "step": 130 }, { "grad_norm": 0.7205566167831421, "learning_rate": 1.3900000000000002e-05, "loss": 1.0129, "step": 140 }, { "grad_norm": 0.9506844282150269, "learning_rate": 1.49e-05, "loss": 0.9668, "step": 150 }, { "grad_norm": 1.1657111644744873, "learning_rate": 1.59e-05, "loss": 0.875, "step": 160 }, { "grad_norm": 1.3662290573120117, "learning_rate": 1.69e-05, "loss": 0.7477, "step": 170 }, { "grad_norm": 1.5999897718429565, "learning_rate": 1.79e-05, "loss": 0.6338, "step": 180 }, { "grad_norm": 1.6561886072158813, "learning_rate": 1.8900000000000002e-05, "loss": 0.5431, "step": 190 }, { "grad_norm": 1.7963553667068481, "learning_rate": 1.9900000000000003e-05, "loss": 0.4662, "step": 200 }, { "grad_norm": 2.200051784515381, "learning_rate": 2.09e-05, "loss": 0.4148, "step": 210 }, { "grad_norm": 1.6306260824203491, "learning_rate": 2.19e-05, "loss": 0.3606, "step": 220 }, { "grad_norm": 2.1751301288604736, "learning_rate": 2.29e-05, "loss": 0.3147, "step": 230 }, { "grad_norm": 2.2360777854919434, "learning_rate": 2.39e-05, "loss": 0.2774, "step": 240 }, { "grad_norm": 2.1259782314300537, "learning_rate": 2.4900000000000002e-05, "loss": 0.2327, "step": 250 }, { "grad_norm": 2.060389995574951, "learning_rate": 2.5900000000000003e-05, "loss": 0.2066, "step": 260 }, { "grad_norm": 1.9270410537719727, "learning_rate": 2.6900000000000003e-05, "loss": 0.1862, "step": 270 }, { "grad_norm": 2.5995590686798096, "learning_rate": 2.7900000000000004e-05, "loss": 0.1659, "step": 280 }, { "grad_norm": 2.2844016551971436, "learning_rate": 2.8899999999999998e-05, "loss": 0.1506, "step": 290 }, { "grad_norm": 1.8172041177749634, "learning_rate": 2.9900000000000002e-05, "loss": 0.1577, "step": 300 }, { "grad_norm": 2.2584943771362305, "learning_rate": 3.09e-05, "loss": 0.146, "step": 310 }, { "grad_norm": 2.3862850666046143, "learning_rate": 3.19e-05, "loss": 0.1277, "step": 320 }, { "grad_norm": 2.074338912963867, "learning_rate": 3.29e-05, "loss": 0.131, "step": 330 }, { "grad_norm": 2.1384787559509277, "learning_rate": 3.3900000000000004e-05, "loss": 0.139, "step": 340 }, { "grad_norm": 2.0320332050323486, "learning_rate": 3.49e-05, "loss": 0.1261, "step": 350 }, { "grad_norm": 1.8671631813049316, "learning_rate": 3.59e-05, "loss": 0.1312, "step": 360 }, { "grad_norm": 1.8936989307403564, "learning_rate": 3.69e-05, "loss": 0.1137, "step": 370 }, { "grad_norm": 2.6163225173950195, "learning_rate": 3.79e-05, "loss": 0.1199, "step": 380 }, { "grad_norm": 1.9205894470214844, "learning_rate": 3.8900000000000004e-05, "loss": 0.1162, "step": 390 }, { "grad_norm": 2.034376859664917, "learning_rate": 3.99e-05, "loss": 0.112, "step": 400 }, { "grad_norm": 1.919226884841919, "learning_rate": 4.09e-05, "loss": 0.1151, "step": 410 }, { "grad_norm": 2.080712080001831, "learning_rate": 4.19e-05, "loss": 0.1084, "step": 420 }, { "grad_norm": 1.7925103902816772, "learning_rate": 4.29e-05, "loss": 0.1123, "step": 430 }, { "grad_norm": 2.3330483436584473, "learning_rate": 4.39e-05, "loss": 0.1036, "step": 440 }, { "grad_norm": 1.6252362728118896, "learning_rate": 4.49e-05, "loss": 0.1057, "step": 450 }, { "grad_norm": 2.0447468757629395, "learning_rate": 4.5900000000000004e-05, "loss": 0.1001, "step": 460 }, { "grad_norm": 2.2220237255096436, "learning_rate": 4.69e-05, "loss": 0.0964, "step": 470 }, { "grad_norm": 1.8954991102218628, "learning_rate": 4.79e-05, "loss": 0.1017, "step": 480 }, { "grad_norm": 2.2161989212036133, "learning_rate": 4.89e-05, "loss": 0.0985, "step": 490 }, { "grad_norm": 1.7932627201080322, "learning_rate": 4.99e-05, "loss": 0.1007, "step": 500 }, { "grad_norm": 1.863224744796753, "learning_rate": 5.0900000000000004e-05, "loss": 0.0941, "step": 510 }, { "grad_norm": 2.241349697113037, "learning_rate": 5.19e-05, "loss": 0.0973, "step": 520 }, { "grad_norm": 1.8485618829727173, "learning_rate": 5.2900000000000005e-05, "loss": 0.1004, "step": 530 }, { "grad_norm": 1.4937787055969238, "learning_rate": 5.390000000000001e-05, "loss": 0.0981, "step": 540 }, { "grad_norm": 2.147996664047241, "learning_rate": 5.4900000000000006e-05, "loss": 0.0881, "step": 550 }, { "grad_norm": 1.749125361442566, "learning_rate": 5.590000000000001e-05, "loss": 0.0952, "step": 560 }, { "grad_norm": 1.5299872159957886, "learning_rate": 5.69e-05, "loss": 0.1019, "step": 570 }, { "grad_norm": 1.8565928936004639, "learning_rate": 5.79e-05, "loss": 0.0958, "step": 580 }, { "grad_norm": 1.3000495433807373, "learning_rate": 5.89e-05, "loss": 0.0893, "step": 590 }, { "grad_norm": 1.8269810676574707, "learning_rate": 5.99e-05, "loss": 0.0933, "step": 600 }, { "grad_norm": 1.6384656429290771, "learning_rate": 6.09e-05, "loss": 0.0876, "step": 610 }, { "grad_norm": 1.4812359809875488, "learning_rate": 6.19e-05, "loss": 0.0893, "step": 620 }, { "grad_norm": 1.5381309986114502, "learning_rate": 6.29e-05, "loss": 0.0903, "step": 630 }, { "grad_norm": 1.9036744832992554, "learning_rate": 6.390000000000001e-05, "loss": 0.0952, "step": 640 }, { "grad_norm": 1.2610657215118408, "learning_rate": 6.49e-05, "loss": 0.0914, "step": 650 }, { "grad_norm": 1.6459393501281738, "learning_rate": 6.59e-05, "loss": 0.0938, "step": 660 }, { "grad_norm": 1.5960263013839722, "learning_rate": 6.690000000000001e-05, "loss": 0.0862, "step": 670 }, { "grad_norm": 1.428434133529663, "learning_rate": 6.790000000000001e-05, "loss": 0.0833, "step": 680 }, { "grad_norm": 1.392797589302063, "learning_rate": 6.89e-05, "loss": 0.0781, "step": 690 }, { "grad_norm": 1.4707822799682617, "learning_rate": 6.99e-05, "loss": 0.0826, "step": 700 }, { "grad_norm": 1.5200772285461426, "learning_rate": 7.09e-05, "loss": 0.0814, "step": 710 }, { "grad_norm": 1.6483746767044067, "learning_rate": 7.19e-05, "loss": 0.0859, "step": 720 }, { "grad_norm": 1.3560274839401245, "learning_rate": 7.29e-05, "loss": 0.0885, "step": 730 }, { "grad_norm": 1.5266467332839966, "learning_rate": 7.390000000000001e-05, "loss": 0.089, "step": 740 }, { "grad_norm": 1.1018054485321045, "learning_rate": 7.49e-05, "loss": 0.087, "step": 750 }, { "grad_norm": 1.4310693740844727, "learning_rate": 7.59e-05, "loss": 0.0817, "step": 760 }, { "grad_norm": 1.223397135734558, "learning_rate": 7.69e-05, "loss": 0.0879, "step": 770 }, { "grad_norm": 1.4443864822387695, "learning_rate": 7.790000000000001e-05, "loss": 0.0856, "step": 780 }, { "grad_norm": 1.5365742444992065, "learning_rate": 7.890000000000001e-05, "loss": 0.0789, "step": 790 }, { "grad_norm": 1.3867871761322021, "learning_rate": 7.99e-05, "loss": 0.0803, "step": 800 }, { "grad_norm": 1.4013606309890747, "learning_rate": 8.090000000000001e-05, "loss": 0.0731, "step": 810 }, { "grad_norm": 1.263858437538147, "learning_rate": 8.19e-05, "loss": 0.0782, "step": 820 }, { "grad_norm": 1.1962145566940308, "learning_rate": 8.29e-05, "loss": 0.0776, "step": 830 }, { "grad_norm": 1.8160847425460815, "learning_rate": 8.39e-05, "loss": 0.0859, "step": 840 }, { "grad_norm": 1.1883392333984375, "learning_rate": 8.49e-05, "loss": 0.0799, "step": 850 }, { "grad_norm": 1.6796576976776123, "learning_rate": 8.59e-05, "loss": 0.081, "step": 860 }, { "grad_norm": 1.3270323276519775, "learning_rate": 8.69e-05, "loss": 0.0763, "step": 870 }, { "grad_norm": 1.4315977096557617, "learning_rate": 8.790000000000001e-05, "loss": 0.0799, "step": 880 }, { "grad_norm": 1.3804460763931274, "learning_rate": 8.89e-05, "loss": 0.0777, "step": 890 }, { "grad_norm": 1.2989784479141235, "learning_rate": 8.99e-05, "loss": 0.0804, "step": 900 }, { "grad_norm": 1.049817442893982, "learning_rate": 9.090000000000001e-05, "loss": 0.0749, "step": 910 }, { "grad_norm": 0.9954274892807007, "learning_rate": 9.190000000000001e-05, "loss": 0.0734, "step": 920 }, { "grad_norm": 1.098987102508545, "learning_rate": 9.290000000000001e-05, "loss": 0.079, "step": 930 }, { "grad_norm": 1.392482042312622, "learning_rate": 9.39e-05, "loss": 0.073, "step": 940 }, { "grad_norm": 1.2326686382293701, "learning_rate": 9.49e-05, "loss": 0.075, "step": 950 }, { "grad_norm": 1.3740872144699097, "learning_rate": 9.59e-05, "loss": 0.0734, "step": 960 }, { "grad_norm": 1.1349748373031616, "learning_rate": 9.69e-05, "loss": 0.0765, "step": 970 }, { "grad_norm": 1.3485573530197144, "learning_rate": 9.790000000000001e-05, "loss": 0.0723, "step": 980 }, { "grad_norm": 1.1115580797195435, "learning_rate": 9.89e-05, "loss": 0.081, "step": 990 }, { "grad_norm": 1.0948477983474731, "learning_rate": 9.99e-05, "loss": 0.0707, "step": 1000 }, { "grad_norm": 0.891538679599762, "learning_rate": 9.999994463727085e-05, "loss": 0.0708, "step": 1010 }, { "grad_norm": 1.060524821281433, "learning_rate": 9.999975326009292e-05, "loss": 0.0756, "step": 1020 }, { "grad_norm": 1.0088512897491455, "learning_rate": 9.999942518549879e-05, "loss": 0.0693, "step": 1030 }, { "grad_norm": 0.9776875972747803, "learning_rate": 9.999896041438544e-05, "loss": 0.0657, "step": 1040 }, { "grad_norm": 0.9027684330940247, "learning_rate": 9.999835894802353e-05, "loss": 0.0637, "step": 1050 }, { "grad_norm": 1.152626395225525, "learning_rate": 9.999762078805743e-05, "loss": 0.0654, "step": 1060 }, { "grad_norm": 1.255368709564209, "learning_rate": 9.999674593650526e-05, "loss": 0.0687, "step": 1070 }, { "grad_norm": 0.8104651570320129, "learning_rate": 9.99957343957588e-05, "loss": 0.0728, "step": 1080 }, { "grad_norm": 1.3187296390533447, "learning_rate": 9.99945861685836e-05, "loss": 0.0673, "step": 1090 }, { "grad_norm": 1.1457287073135376, "learning_rate": 9.999330125811884e-05, "loss": 0.0652, "step": 1100 }, { "grad_norm": 1.0516674518585205, "learning_rate": 9.999187966787744e-05, "loss": 0.0637, "step": 1110 }, { "grad_norm": 1.0536901950836182, "learning_rate": 9.999032140174595e-05, "loss": 0.0655, "step": 1120 }, { "grad_norm": 0.898169994354248, "learning_rate": 9.998862646398464e-05, "loss": 0.0703, "step": 1130 }, { "grad_norm": 1.029990553855896, "learning_rate": 9.998679485922739e-05, "loss": 0.0676, "step": 1140 }, { "grad_norm": 0.7401315569877625, "learning_rate": 9.998482659248174e-05, "loss": 0.061, "step": 1150 }, { "grad_norm": 1.0644394159317017, "learning_rate": 9.998272166912883e-05, "loss": 0.0664, "step": 1160 }, { "grad_norm": 1.0024806261062622, "learning_rate": 9.998048009492347e-05, "loss": 0.068, "step": 1170 }, { "grad_norm": 0.896041750907898, "learning_rate": 9.997810187599403e-05, "loss": 0.0629, "step": 1180 }, { "grad_norm": 0.8876444101333618, "learning_rate": 9.997558701884249e-05, "loss": 0.0598, "step": 1190 }, { "grad_norm": 0.9452154636383057, "learning_rate": 9.997293553034433e-05, "loss": 0.0665, "step": 1200 }, { "grad_norm": 0.6464991569519043, "learning_rate": 9.997014741774866e-05, "loss": 0.0598, "step": 1210 }, { "grad_norm": 0.8307116627693176, "learning_rate": 9.996722268867803e-05, "loss": 0.0598, "step": 1220 }, { "grad_norm": 0.9347904920578003, "learning_rate": 9.996416135112858e-05, "loss": 0.0647, "step": 1230 }, { "grad_norm": 0.8860670924186707, "learning_rate": 9.996096341346988e-05, "loss": 0.0579, "step": 1240 }, { "grad_norm": 0.890996515750885, "learning_rate": 9.995762888444495e-05, "loss": 0.0561, "step": 1250 }, { "grad_norm": 1.04674232006073, "learning_rate": 9.995415777317027e-05, "loss": 0.0603, "step": 1260 }, { "grad_norm": 0.8565642237663269, "learning_rate": 9.995055008913574e-05, "loss": 0.0581, "step": 1270 }, { "grad_norm": 0.7723894119262695, "learning_rate": 9.994680584220463e-05, "loss": 0.0636, "step": 1280 }, { "grad_norm": 1.0010874271392822, "learning_rate": 9.994292504261355e-05, "loss": 0.0605, "step": 1290 }, { "grad_norm": 0.7321175932884216, "learning_rate": 9.993890770097247e-05, "loss": 0.0581, "step": 1300 }, { "grad_norm": 0.9652481079101562, "learning_rate": 9.993475382826467e-05, "loss": 0.0561, "step": 1310 }, { "grad_norm": 0.7752932906150818, "learning_rate": 9.993046343584664e-05, "loss": 0.0531, "step": 1320 }, { "grad_norm": 1.0333092212677002, "learning_rate": 9.992603653544816e-05, "loss": 0.0641, "step": 1330 }, { "grad_norm": 0.8251863121986389, "learning_rate": 9.992147313917222e-05, "loss": 0.0614, "step": 1340 }, { "grad_norm": 0.771634042263031, "learning_rate": 9.991677325949497e-05, "loss": 0.0585, "step": 1350 }, { "grad_norm": 0.7842410802841187, "learning_rate": 9.991193690926568e-05, "loss": 0.0577, "step": 1360 }, { "grad_norm": 0.7830896377563477, "learning_rate": 9.990696410170678e-05, "loss": 0.0576, "step": 1370 }, { "grad_norm": 0.7047857642173767, "learning_rate": 9.990185485041371e-05, "loss": 0.0523, "step": 1380 }, { "grad_norm": 0.8119860887527466, "learning_rate": 9.989660916935498e-05, "loss": 0.0548, "step": 1390 }, { "grad_norm": 0.8157444000244141, "learning_rate": 9.989122707287208e-05, "loss": 0.0538, "step": 1400 }, { "grad_norm": 0.8182013630867004, "learning_rate": 9.988570857567945e-05, "loss": 0.0531, "step": 1410 }, { "grad_norm": 0.8910542130470276, "learning_rate": 9.988005369286446e-05, "loss": 0.0542, "step": 1420 }, { "grad_norm": 0.8687204718589783, "learning_rate": 9.987426243988734e-05, "loss": 0.0544, "step": 1430 }, { "grad_norm": 0.6190545558929443, "learning_rate": 9.986833483258114e-05, "loss": 0.0512, "step": 1440 }, { "grad_norm": 0.7352811098098755, "learning_rate": 9.986227088715173e-05, "loss": 0.0527, "step": 1450 }, { "grad_norm": 0.7671709656715393, "learning_rate": 9.98560706201777e-05, "loss": 0.0608, "step": 1460 }, { "grad_norm": 0.7920308113098145, "learning_rate": 9.984973404861036e-05, "loss": 0.0526, "step": 1470 }, { "grad_norm": 0.7265468835830688, "learning_rate": 9.984326118977361e-05, "loss": 0.0534, "step": 1480 }, { "grad_norm": 0.7387204170227051, "learning_rate": 9.983665206136406e-05, "loss": 0.0541, "step": 1490 }, { "grad_norm": 0.7927471399307251, "learning_rate": 9.982990668145075e-05, "loss": 0.0515, "step": 1500 }, { "grad_norm": 0.8133869767189026, "learning_rate": 9.982302506847534e-05, "loss": 0.0547, "step": 1510 }, { "grad_norm": 0.6622949838638306, "learning_rate": 9.981600724125189e-05, "loss": 0.0484, "step": 1520 }, { "grad_norm": 0.9121345281600952, "learning_rate": 9.980885321896685e-05, "loss": 0.054, "step": 1530 }, { "grad_norm": 0.8064318895339966, "learning_rate": 9.980156302117905e-05, "loss": 0.0551, "step": 1540 }, { "grad_norm": 0.8083546161651611, "learning_rate": 9.979413666781963e-05, "loss": 0.0541, "step": 1550 }, { "grad_norm": 0.8815498352050781, "learning_rate": 9.978657417919193e-05, "loss": 0.0532, "step": 1560 }, { "grad_norm": 0.9235912561416626, "learning_rate": 9.977887557597153e-05, "loss": 0.0545, "step": 1570 }, { "grad_norm": 0.7402552366256714, "learning_rate": 9.97710408792061e-05, "loss": 0.052, "step": 1580 }, { "grad_norm": 0.8354697823524475, "learning_rate": 9.976307011031542e-05, "loss": 0.0512, "step": 1590 }, { "grad_norm": 0.8086682558059692, "learning_rate": 9.975496329109126e-05, "loss": 0.054, "step": 1600 }, { "grad_norm": 0.5833176374435425, "learning_rate": 9.974672044369732e-05, "loss": 0.0502, "step": 1610 }, { "grad_norm": 0.5650737285614014, "learning_rate": 9.97383415906693e-05, "loss": 0.0455, "step": 1620 }, { "grad_norm": 0.7207682728767395, "learning_rate": 9.97298267549146e-05, "loss": 0.0505, "step": 1630 }, { "grad_norm": 0.717327892780304, "learning_rate": 9.972117595971249e-05, "loss": 0.0475, "step": 1640 }, { "grad_norm": 0.7371031045913696, "learning_rate": 9.971238922871391e-05, "loss": 0.0478, "step": 1650 }, { "grad_norm": 0.8945282101631165, "learning_rate": 9.970346658594142e-05, "loss": 0.0463, "step": 1660 }, { "grad_norm": 0.8559687733650208, "learning_rate": 9.969440805578923e-05, "loss": 0.0488, "step": 1670 }, { "grad_norm": 0.5055457353591919, "learning_rate": 9.968521366302298e-05, "loss": 0.0508, "step": 1680 }, { "grad_norm": 0.8108320832252502, "learning_rate": 9.967588343277981e-05, "loss": 0.0549, "step": 1690 }, { "grad_norm": 0.7972011566162109, "learning_rate": 9.966641739056818e-05, "loss": 0.0561, "step": 1700 }, { "grad_norm": 0.8416505455970764, "learning_rate": 9.965681556226793e-05, "loss": 0.0508, "step": 1710 }, { "grad_norm": 0.8390809893608093, "learning_rate": 9.964707797413006e-05, "loss": 0.0486, "step": 1720 }, { "grad_norm": 0.7652764320373535, "learning_rate": 9.963720465277679e-05, "loss": 0.0495, "step": 1730 }, { "grad_norm": 0.7617776393890381, "learning_rate": 9.96271956252014e-05, "loss": 0.0475, "step": 1740 }, { "grad_norm": 0.6340845227241516, "learning_rate": 9.961705091876816e-05, "loss": 0.0519, "step": 1750 }, { "grad_norm": 0.6515089869499207, "learning_rate": 9.960677056121235e-05, "loss": 0.0488, "step": 1760 }, { "grad_norm": 0.703163743019104, "learning_rate": 9.959635458064005e-05, "loss": 0.0508, "step": 1770 }, { "grad_norm": 0.6431951522827148, "learning_rate": 9.958580300552815e-05, "loss": 0.0479, "step": 1780 }, { "grad_norm": 0.7388184070587158, "learning_rate": 9.957511586472426e-05, "loss": 0.0499, "step": 1790 }, { "grad_norm": 0.7050029635429382, "learning_rate": 9.956429318744662e-05, "loss": 0.0478, "step": 1800 }, { "grad_norm": 0.4888589084148407, "learning_rate": 9.955333500328404e-05, "loss": 0.0445, "step": 1810 }, { "grad_norm": 0.6852853298187256, "learning_rate": 9.95422413421957e-05, "loss": 0.0466, "step": 1820 }, { "grad_norm": 0.7017318606376648, "learning_rate": 9.953101223451133e-05, "loss": 0.0459, "step": 1830 }, { "grad_norm": 0.6769763827323914, "learning_rate": 9.951964771093085e-05, "loss": 0.046, "step": 1840 }, { "grad_norm": 0.7448799014091492, "learning_rate": 9.950814780252442e-05, "loss": 0.0463, "step": 1850 }, { "grad_norm": 0.8778994679450989, "learning_rate": 9.949651254073236e-05, "loss": 0.0492, "step": 1860 }, { "grad_norm": 0.7002668380737305, "learning_rate": 9.948474195736504e-05, "loss": 0.0479, "step": 1870 }, { "grad_norm": 0.6542785167694092, "learning_rate": 9.947283608460277e-05, "loss": 0.0457, "step": 1880 }, { "grad_norm": 0.7363675236701965, "learning_rate": 9.946079495499577e-05, "loss": 0.0504, "step": 1890 }, { "grad_norm": 0.7688233256340027, "learning_rate": 9.944861860146401e-05, "loss": 0.0455, "step": 1900 }, { "grad_norm": 0.552820086479187, "learning_rate": 9.943630705729719e-05, "loss": 0.0438, "step": 1910 }, { "grad_norm": 0.730654776096344, "learning_rate": 9.942386035615459e-05, "loss": 0.0467, "step": 1920 }, { "grad_norm": 0.7699289321899414, "learning_rate": 9.941127853206503e-05, "loss": 0.0468, "step": 1930 }, { "grad_norm": 0.8171266317367554, "learning_rate": 9.939856161942673e-05, "loss": 0.0475, "step": 1940 }, { "grad_norm": 0.6612003445625305, "learning_rate": 9.938570965300724e-05, "loss": 0.0456, "step": 1950 }, { "grad_norm": 0.7253115177154541, "learning_rate": 9.937272266794335e-05, "loss": 0.0464, "step": 1960 }, { "grad_norm": 0.7715169191360474, "learning_rate": 9.935960069974096e-05, "loss": 0.0477, "step": 1970 }, { "grad_norm": 0.7735244631767273, "learning_rate": 9.934634378427506e-05, "loss": 0.0462, "step": 1980 }, { "grad_norm": 0.7821980118751526, "learning_rate": 9.933295195778954e-05, "loss": 0.0473, "step": 1990 }, { "grad_norm": 0.6536567807197571, "learning_rate": 9.931942525689715e-05, "loss": 0.0456, "step": 2000 }, { "grad_norm": 0.8191483616828918, "learning_rate": 9.930576371857936e-05, "loss": 0.0439, "step": 2010 }, { "grad_norm": 0.8047411441802979, "learning_rate": 9.929196738018629e-05, "loss": 0.0487, "step": 2020 }, { "grad_norm": 0.7432626485824585, "learning_rate": 9.927803627943662e-05, "loss": 0.0476, "step": 2030 }, { "grad_norm": 0.7016537189483643, "learning_rate": 9.926397045441744e-05, "loss": 0.0454, "step": 2040 }, { "grad_norm": 0.42724964022636414, "learning_rate": 9.924976994358417e-05, "loss": 0.046, "step": 2050 }, { "grad_norm": 0.742233395576477, "learning_rate": 9.923543478576048e-05, "loss": 0.0446, "step": 2060 }, { "grad_norm": 0.7966566681861877, "learning_rate": 9.922096502013813e-05, "loss": 0.0469, "step": 2070 }, { "grad_norm": 0.5727256536483765, "learning_rate": 9.92063606862769e-05, "loss": 0.0422, "step": 2080 }, { "grad_norm": 0.6520189046859741, "learning_rate": 9.919162182410453e-05, "loss": 0.0474, "step": 2090 }, { "grad_norm": 0.7384130954742432, "learning_rate": 9.917674847391645e-05, "loss": 0.043, "step": 2100 }, { "grad_norm": 0.6171939373016357, "learning_rate": 9.916174067637584e-05, "loss": 0.0388, "step": 2110 }, { "grad_norm": 0.7047433257102966, "learning_rate": 9.914659847251348e-05, "loss": 0.0419, "step": 2120 }, { "grad_norm": 0.6102860569953918, "learning_rate": 9.913132190372753e-05, "loss": 0.0457, "step": 2130 }, { "grad_norm": 0.5510637164115906, "learning_rate": 9.911591101178359e-05, "loss": 0.0427, "step": 2140 }, { "grad_norm": 0.5301130414009094, "learning_rate": 9.910036583881443e-05, "loss": 0.0423, "step": 2150 }, { "grad_norm": 0.5777102708816528, "learning_rate": 9.908468642731995e-05, "loss": 0.0403, "step": 2160 }, { "grad_norm": 0.48912176489830017, "learning_rate": 9.906887282016707e-05, "loss": 0.0389, "step": 2170 }, { "grad_norm": 0.6603399515151978, "learning_rate": 9.90529250605896e-05, "loss": 0.0395, "step": 2180 }, { "grad_norm": 0.5303890109062195, "learning_rate": 9.903684319218809e-05, "loss": 0.0404, "step": 2190 }, { "grad_norm": 0.542991578578949, "learning_rate": 9.902062725892976e-05, "loss": 0.0485, "step": 2200 }, { "grad_norm": 0.7258626222610474, "learning_rate": 9.900427730514834e-05, "loss": 0.0433, "step": 2210 }, { "grad_norm": 0.6577914357185364, "learning_rate": 9.8987793375544e-05, "loss": 0.0466, "step": 2220 }, { "grad_norm": 0.5233342051506042, "learning_rate": 9.897117551518318e-05, "loss": 0.0408, "step": 2230 }, { "grad_norm": 0.68238765001297, "learning_rate": 9.895442376949844e-05, "loss": 0.0411, "step": 2240 }, { "grad_norm": 0.6401824355125427, "learning_rate": 9.893753818428845e-05, "loss": 0.0416, "step": 2250 }, { "grad_norm": 0.5401156544685364, "learning_rate": 9.892051880571773e-05, "loss": 0.0446, "step": 2260 }, { "grad_norm": 0.5518479943275452, "learning_rate": 9.890336568031663e-05, "loss": 0.0408, "step": 2270 }, { "grad_norm": 0.5495585799217224, "learning_rate": 9.888607885498113e-05, "loss": 0.037, "step": 2280 }, { "grad_norm": 0.5683923363685608, "learning_rate": 9.886865837697275e-05, "loss": 0.0402, "step": 2290 }, { "grad_norm": 0.5670053362846375, "learning_rate": 9.88511042939184e-05, "loss": 0.0403, "step": 2300 }, { "grad_norm": 0.5962883830070496, "learning_rate": 9.883341665381028e-05, "loss": 0.0425, "step": 2310 }, { "grad_norm": 0.6567088961601257, "learning_rate": 9.881559550500575e-05, "loss": 0.0415, "step": 2320 }, { "grad_norm": 0.6355565190315247, "learning_rate": 9.879764089622712e-05, "loss": 0.0458, "step": 2330 }, { "grad_norm": 0.6942632794380188, "learning_rate": 9.87795528765616e-05, "loss": 0.0387, "step": 2340 }, { "grad_norm": 0.6719608306884766, "learning_rate": 9.876133149546118e-05, "loss": 0.0469, "step": 2350 }, { "grad_norm": 0.7083789706230164, "learning_rate": 9.874297680274238e-05, "loss": 0.0405, "step": 2360 }, { "grad_norm": 0.588875949382782, "learning_rate": 9.872448884858624e-05, "loss": 0.0398, "step": 2370 }, { "grad_norm": 0.5804881453514099, "learning_rate": 9.870586768353815e-05, "loss": 0.0414, "step": 2380 }, { "grad_norm": 0.6627275347709656, "learning_rate": 9.868711335850764e-05, "loss": 0.0425, "step": 2390 }, { "grad_norm": 0.6153960227966309, "learning_rate": 9.866822592476833e-05, "loss": 0.0417, "step": 2400 }, { "grad_norm": 0.5534101128578186, "learning_rate": 9.86492054339577e-05, "loss": 0.036, "step": 2410 }, { "grad_norm": 0.7058823108673096, "learning_rate": 9.863005193807711e-05, "loss": 0.0459, "step": 2420 }, { "grad_norm": 0.6398679614067078, "learning_rate": 9.861076548949143e-05, "loss": 0.0398, "step": 2430 }, { "grad_norm": 0.576840877532959, "learning_rate": 9.859134614092912e-05, "loss": 0.0415, "step": 2440 }, { "grad_norm": 0.575426459312439, "learning_rate": 9.857179394548191e-05, "loss": 0.0413, "step": 2450 }, { "grad_norm": 0.5618106722831726, "learning_rate": 9.855210895660477e-05, "loss": 0.0379, "step": 2460 }, { "grad_norm": 0.523193359375, "learning_rate": 9.853229122811568e-05, "loss": 0.0399, "step": 2470 }, { "grad_norm": 0.5672025680541992, "learning_rate": 9.851234081419559e-05, "loss": 0.0372, "step": 2480 }, { "grad_norm": 0.6958249807357788, "learning_rate": 9.849225776938814e-05, "loss": 0.0458, "step": 2490 }, { "grad_norm": 0.7104291915893555, "learning_rate": 9.847204214859964e-05, "loss": 0.0393, "step": 2500 }, { "grad_norm": 0.6773826479911804, "learning_rate": 9.845169400709879e-05, "loss": 0.0398, "step": 2510 }, { "grad_norm": 0.5029601454734802, "learning_rate": 9.843121340051664e-05, "loss": 0.0383, "step": 2520 }, { "grad_norm": 0.7401909232139587, "learning_rate": 9.841060038484641e-05, "loss": 0.0388, "step": 2530 }, { "grad_norm": 0.6366117596626282, "learning_rate": 9.838985501644328e-05, "loss": 0.0364, "step": 2540 }, { "grad_norm": 0.6136524081230164, "learning_rate": 9.83689773520243e-05, "loss": 0.0405, "step": 2550 }, { "grad_norm": 0.682521402835846, "learning_rate": 9.834796744866819e-05, "loss": 0.0442, "step": 2560 }, { "grad_norm": 0.6501883864402771, "learning_rate": 9.832682536381525e-05, "loss": 0.0371, "step": 2570 }, { "grad_norm": 0.6832765340805054, "learning_rate": 9.830555115526711e-05, "loss": 0.0394, "step": 2580 }, { "grad_norm": 0.6529347896575928, "learning_rate": 9.828414488118667e-05, "loss": 0.0361, "step": 2590 }, { "grad_norm": 0.5274178981781006, "learning_rate": 9.826260660009785e-05, "loss": 0.0421, "step": 2600 }, { "grad_norm": 0.6173925399780273, "learning_rate": 9.824093637088547e-05, "loss": 0.0396, "step": 2610 }, { "grad_norm": 0.7607269287109375, "learning_rate": 9.821913425279514e-05, "loss": 0.0394, "step": 2620 }, { "grad_norm": 0.6358497738838196, "learning_rate": 9.8197200305433e-05, "loss": 0.0368, "step": 2630 }, { "grad_norm": 0.5698424577713013, "learning_rate": 9.817513458876564e-05, "loss": 0.0397, "step": 2640 }, { "grad_norm": 0.6986652612686157, "learning_rate": 9.815293716311987e-05, "loss": 0.0388, "step": 2650 }, { "grad_norm": 0.5638768672943115, "learning_rate": 9.813060808918262e-05, "loss": 0.0368, "step": 2660 }, { "grad_norm": 0.636012077331543, "learning_rate": 9.810814742800069e-05, "loss": 0.0386, "step": 2670 }, { "grad_norm": 0.7297207117080688, "learning_rate": 9.808555524098074e-05, "loss": 0.0384, "step": 2680 }, { "grad_norm": 0.6475580930709839, "learning_rate": 9.806283158988887e-05, "loss": 0.0407, "step": 2690 }, { "grad_norm": 0.8668317794799805, "learning_rate": 9.803997653685072e-05, "loss": 0.0369, "step": 2700 }, { "grad_norm": 0.5016562342643738, "learning_rate": 9.801699014435112e-05, "loss": 0.0361, "step": 2710 }, { "grad_norm": 0.6411369442939758, "learning_rate": 9.799387247523398e-05, "loss": 0.0366, "step": 2720 }, { "grad_norm": 0.6267861127853394, "learning_rate": 9.797062359270215e-05, "loss": 0.0375, "step": 2730 }, { "grad_norm": 0.542231023311615, "learning_rate": 9.794724356031715e-05, "loss": 0.0346, "step": 2740 }, { "grad_norm": 0.5228654146194458, "learning_rate": 9.792373244199913e-05, "loss": 0.032, "step": 2750 }, { "grad_norm": 0.6026210784912109, "learning_rate": 9.790009030202658e-05, "loss": 0.0339, "step": 2760 }, { "grad_norm": 0.5474340319633484, "learning_rate": 9.78763172050362e-05, "loss": 0.0388, "step": 2770 }, { "grad_norm": 0.6412479281425476, "learning_rate": 9.785241321602274e-05, "loss": 0.0335, "step": 2780 }, { "grad_norm": 0.5595812797546387, "learning_rate": 9.782837840033879e-05, "loss": 0.0353, "step": 2790 }, { "grad_norm": 0.46222779154777527, "learning_rate": 9.780421282369461e-05, "loss": 0.0375, "step": 2800 }, { "grad_norm": 0.5727255344390869, "learning_rate": 9.777991655215797e-05, "loss": 0.0384, "step": 2810 }, { "grad_norm": 0.4622480273246765, "learning_rate": 9.775548965215394e-05, "loss": 0.0369, "step": 2820 }, { "grad_norm": 0.5925295948982239, "learning_rate": 9.773093219046474e-05, "loss": 0.0339, "step": 2830 }, { "grad_norm": 0.6303073763847351, "learning_rate": 9.770624423422954e-05, "loss": 0.0349, "step": 2840 }, { "grad_norm": 0.649684727191925, "learning_rate": 9.768142585094426e-05, "loss": 0.0351, "step": 2850 }, { "grad_norm": 0.564365565776825, "learning_rate": 9.765647710846142e-05, "loss": 0.0393, "step": 2860 }, { "grad_norm": 0.5238656997680664, "learning_rate": 9.763139807498991e-05, "loss": 0.0374, "step": 2870 }, { "grad_norm": 0.6247703433036804, "learning_rate": 9.760618881909487e-05, "loss": 0.0339, "step": 2880 }, { "grad_norm": 0.7072002291679382, "learning_rate": 9.758084940969744e-05, "loss": 0.0364, "step": 2890 }, { "grad_norm": 0.5810469388961792, "learning_rate": 9.755537991607459e-05, "loss": 0.0344, "step": 2900 }, { "grad_norm": 0.6288248300552368, "learning_rate": 9.752978040785895e-05, "loss": 0.0328, "step": 2910 }, { "grad_norm": 0.6866697072982788, "learning_rate": 9.750405095503859e-05, "loss": 0.0386, "step": 2920 }, { "grad_norm": 0.5289666056632996, "learning_rate": 9.747819162795686e-05, "loss": 0.0393, "step": 2930 }, { "grad_norm": 0.6614279747009277, "learning_rate": 9.745220249731217e-05, "loss": 0.0412, "step": 2940 }, { "grad_norm": 0.5238449573516846, "learning_rate": 9.742608363415781e-05, "loss": 0.0349, "step": 2950 }, { "grad_norm": 0.5450305342674255, "learning_rate": 9.739983510990176e-05, "loss": 0.036, "step": 2960 }, { "grad_norm": 0.504685640335083, "learning_rate": 9.737345699630647e-05, "loss": 0.0324, "step": 2970 }, { "grad_norm": 0.5616886615753174, "learning_rate": 9.734694936548869e-05, "loss": 0.0317, "step": 2980 }, { "grad_norm": 0.627938985824585, "learning_rate": 9.732031228991932e-05, "loss": 0.0331, "step": 2990 }, { "grad_norm": 0.48203128576278687, "learning_rate": 9.729354584242302e-05, "loss": 0.0329, "step": 3000 }, { "grad_norm": 0.560325562953949, "learning_rate": 9.726665009617832e-05, "loss": 0.0309, "step": 3010 }, { "grad_norm": 0.5883328914642334, "learning_rate": 9.723962512471714e-05, "loss": 0.0332, "step": 3020 }, { "grad_norm": 0.4695000946521759, "learning_rate": 9.72124710019247e-05, "loss": 0.0304, "step": 3030 }, { "grad_norm": 0.7158396244049072, "learning_rate": 9.718518780203934e-05, "loss": 0.0339, "step": 3040 }, { "grad_norm": 0.5638872981071472, "learning_rate": 9.715777559965228e-05, "loss": 0.0317, "step": 3050 }, { "grad_norm": 0.5769653916358948, "learning_rate": 9.713023446970746e-05, "loss": 0.0319, "step": 3060 }, { "grad_norm": 0.5889293551445007, "learning_rate": 9.710256448750126e-05, "loss": 0.0313, "step": 3070 }, { "grad_norm": 0.5529425740242004, "learning_rate": 9.707476572868235e-05, "loss": 0.0327, "step": 3080 }, { "grad_norm": 0.47855859994888306, "learning_rate": 9.704683826925149e-05, "loss": 0.0328, "step": 3090 }, { "grad_norm": 0.5483555197715759, "learning_rate": 9.701878218556129e-05, "loss": 0.0333, "step": 3100 }, { "grad_norm": 0.5374535918235779, "learning_rate": 9.699059755431598e-05, "loss": 0.0317, "step": 3110 }, { "grad_norm": 0.5370588898658752, "learning_rate": 9.696228445257132e-05, "loss": 0.0318, "step": 3120 }, { "grad_norm": 0.4525238573551178, "learning_rate": 9.693384295773419e-05, "loss": 0.0337, "step": 3130 }, { "grad_norm": 0.6433699131011963, "learning_rate": 9.690527314756259e-05, "loss": 0.0318, "step": 3140 }, { "grad_norm": 0.5980566740036011, "learning_rate": 9.687657510016527e-05, "loss": 0.0312, "step": 3150 }, { "grad_norm": 0.6685139536857605, "learning_rate": 9.684774889400161e-05, "loss": 0.0299, "step": 3160 }, { "grad_norm": 0.5671114325523376, "learning_rate": 9.681879460788135e-05, "loss": 0.0308, "step": 3170 }, { "grad_norm": 0.5917689800262451, "learning_rate": 9.67897123209644e-05, "loss": 0.0303, "step": 3180 }, { "grad_norm": 0.6136913299560547, "learning_rate": 9.676050211276062e-05, "loss": 0.0313, "step": 3190 }, { "grad_norm": 0.5971332788467407, "learning_rate": 9.673116406312962e-05, "loss": 0.0317, "step": 3200 }, { "grad_norm": 0.5821442604064941, "learning_rate": 9.67016982522805e-05, "loss": 0.0319, "step": 3210 }, { "grad_norm": 0.6876721382141113, "learning_rate": 9.667210476077164e-05, "loss": 0.0293, "step": 3220 }, { "grad_norm": 0.6031916737556458, "learning_rate": 9.664238366951055e-05, "loss": 0.0285, "step": 3230 }, { "grad_norm": 0.5616409182548523, "learning_rate": 9.661253505975355e-05, "loss": 0.0315, "step": 3240 }, { "grad_norm": 0.7697345018386841, "learning_rate": 9.658255901310557e-05, "loss": 0.0293, "step": 3250 }, { "grad_norm": 0.5797312259674072, "learning_rate": 9.655245561152e-05, "loss": 0.0335, "step": 3260 }, { "grad_norm": 0.5256711840629578, "learning_rate": 9.65222249372984e-05, "loss": 0.0307, "step": 3270 }, { "grad_norm": 0.6105689406394958, "learning_rate": 9.649186707309026e-05, "loss": 0.0312, "step": 3280 }, { "grad_norm": 0.6437607407569885, "learning_rate": 9.646138210189283e-05, "loss": 0.0365, "step": 3290 }, { "grad_norm": 0.525951087474823, "learning_rate": 9.643077010705087e-05, "loss": 0.0312, "step": 3300 }, { "grad_norm": 0.4761003255844116, "learning_rate": 9.640003117225637e-05, "loss": 0.0286, "step": 3310 }, { "grad_norm": 0.710482120513916, "learning_rate": 9.636916538154846e-05, "loss": 0.0318, "step": 3320 }, { "grad_norm": 0.6405505537986755, "learning_rate": 9.633817281931296e-05, "loss": 0.0274, "step": 3330 }, { "grad_norm": 0.48160025477409363, "learning_rate": 9.630705357028242e-05, "loss": 0.031, "step": 3340 }, { "grad_norm": 0.6867256760597229, "learning_rate": 9.627580771953563e-05, "loss": 0.0307, "step": 3350 }, { "grad_norm": 0.49422842264175415, "learning_rate": 9.624443535249759e-05, "loss": 0.0289, "step": 3360 }, { "grad_norm": 0.5430372953414917, "learning_rate": 9.621293655493913e-05, "loss": 0.0316, "step": 3370 }, { "grad_norm": 0.5640618205070496, "learning_rate": 9.618131141297675e-05, "loss": 0.0277, "step": 3380 }, { "grad_norm": 0.47579532861709595, "learning_rate": 9.614956001307242e-05, "loss": 0.0253, "step": 3390 }, { "grad_norm": 0.5587700605392456, "learning_rate": 9.611768244203321e-05, "loss": 0.0305, "step": 3400 }, { "grad_norm": 0.5596683621406555, "learning_rate": 9.60856787870112e-05, "loss": 0.0254, "step": 3410 }, { "grad_norm": 0.5264420509338379, "learning_rate": 9.605354913550318e-05, "loss": 0.0261, "step": 3420 }, { "grad_norm": 0.7247219681739807, "learning_rate": 9.602129357535037e-05, "loss": 0.0278, "step": 3430 }, { "grad_norm": 0.5680612921714783, "learning_rate": 9.598891219473825e-05, "loss": 0.0319, "step": 3440 }, { "grad_norm": 0.6597133874893188, "learning_rate": 9.595640508219625e-05, "loss": 0.0297, "step": 3450 }, { "grad_norm": 0.5399404764175415, "learning_rate": 9.592377232659761e-05, "loss": 0.0328, "step": 3460 }, { "grad_norm": 0.5794700980186462, "learning_rate": 9.589101401715904e-05, "loss": 0.0244, "step": 3470 }, { "grad_norm": 0.5431540012359619, "learning_rate": 9.585813024344045e-05, "loss": 0.0298, "step": 3480 }, { "grad_norm": 0.5857890844345093, "learning_rate": 9.58251210953449e-05, "loss": 0.0266, "step": 3490 }, { "grad_norm": 0.659814715385437, "learning_rate": 9.579198666311809e-05, "loss": 0.0281, "step": 3500 }, { "grad_norm": 0.4802360534667969, "learning_rate": 9.575872703734832e-05, "loss": 0.0251, "step": 3510 }, { "grad_norm": 0.6888503432273865, "learning_rate": 9.572534230896611e-05, "loss": 0.0291, "step": 3520 }, { "grad_norm": 0.5367239117622375, "learning_rate": 9.569183256924403e-05, "loss": 0.0301, "step": 3530 }, { "grad_norm": 0.6264481544494629, "learning_rate": 9.565819790979646e-05, "loss": 0.0269, "step": 3540 }, { "grad_norm": 0.644710123538971, "learning_rate": 9.562443842257925e-05, "loss": 0.0258, "step": 3550 }, { "grad_norm": 0.49636292457580566, "learning_rate": 9.559055419988956e-05, "loss": 0.0272, "step": 3560 }, { "grad_norm": 0.5108775496482849, "learning_rate": 9.555654533436557e-05, "loss": 0.0267, "step": 3570 }, { "grad_norm": 0.6691751480102539, "learning_rate": 9.552241191898621e-05, "loss": 0.0302, "step": 3580 }, { "grad_norm": 0.5360530018806458, "learning_rate": 9.548815404707092e-05, "loss": 0.0273, "step": 3590 }, { "grad_norm": 0.5134739279747009, "learning_rate": 9.545377181227942e-05, "loss": 0.0296, "step": 3600 }, { "grad_norm": 0.5842033624649048, "learning_rate": 9.541926530861145e-05, "loss": 0.026, "step": 3610 }, { "grad_norm": 0.4610866606235504, "learning_rate": 9.538463463040645e-05, "loss": 0.0247, "step": 3620 }, { "grad_norm": 0.5100011229515076, "learning_rate": 9.534987987234337e-05, "loss": 0.0244, "step": 3630 }, { "grad_norm": 0.4213336110115051, "learning_rate": 9.53150011294404e-05, "loss": 0.0246, "step": 3640 }, { "grad_norm": 0.5512347221374512, "learning_rate": 9.527999849705471e-05, "loss": 0.0257, "step": 3650 }, { "grad_norm": 0.5232389569282532, "learning_rate": 9.524487207088213e-05, "loss": 0.0267, "step": 3660 }, { "grad_norm": 0.4374615550041199, "learning_rate": 9.520962194695698e-05, "loss": 0.0266, "step": 3670 }, { "grad_norm": 0.4799204468727112, "learning_rate": 9.517424822165175e-05, "loss": 0.0237, "step": 3680 }, { "grad_norm": 0.7146117687225342, "learning_rate": 9.513875099167685e-05, "loss": 0.0254, "step": 3690 }, { "grad_norm": 0.4083298444747925, "learning_rate": 9.510313035408035e-05, "loss": 0.0265, "step": 3700 }, { "grad_norm": 0.5064089894294739, "learning_rate": 9.506738640624775e-05, "loss": 0.0249, "step": 3710 }, { "grad_norm": 0.6063523292541504, "learning_rate": 9.50315192459016e-05, "loss": 0.0262, "step": 3720 }, { "grad_norm": 0.40898430347442627, "learning_rate": 9.499552897110136e-05, "loss": 0.0256, "step": 3730 }, { "grad_norm": 0.4674573838710785, "learning_rate": 9.495941568024304e-05, "loss": 0.0243, "step": 3740 }, { "grad_norm": 0.4625905156135559, "learning_rate": 9.492317947205904e-05, "loss": 0.0261, "step": 3750 }, { "grad_norm": 0.41100046038627625, "learning_rate": 9.488682044561775e-05, "loss": 0.025, "step": 3760 }, { "grad_norm": 0.7739120721817017, "learning_rate": 9.485033870032335e-05, "loss": 0.0251, "step": 3770 }, { "grad_norm": 0.44929733872413635, "learning_rate": 9.481373433591556e-05, "loss": 0.0243, "step": 3780 }, { "grad_norm": 0.5701534152030945, "learning_rate": 9.47770074524693e-05, "loss": 0.0272, "step": 3790 }, { "grad_norm": 0.46684274077415466, "learning_rate": 9.474015815039446e-05, "loss": 0.0266, "step": 3800 }, { "grad_norm": 0.4567999839782715, "learning_rate": 9.470318653043565e-05, "loss": 0.0216, "step": 3810 }, { "grad_norm": 0.5459327101707458, "learning_rate": 9.466609269367185e-05, "loss": 0.0251, "step": 3820 }, { "grad_norm": 0.45713010430336, "learning_rate": 9.46288767415162e-05, "loss": 0.0273, "step": 3830 }, { "grad_norm": 0.48132744431495667, "learning_rate": 9.459153877571567e-05, "loss": 0.0245, "step": 3840 }, { "grad_norm": 0.5496471524238586, "learning_rate": 9.455407889835087e-05, "loss": 0.0248, "step": 3850 }, { "grad_norm": 0.6081171035766602, "learning_rate": 9.451649721183564e-05, "loss": 0.0251, "step": 3860 }, { "grad_norm": 0.5593532919883728, "learning_rate": 9.447879381891692e-05, "loss": 0.0254, "step": 3870 }, { "grad_norm": 0.4086363613605499, "learning_rate": 9.444096882267428e-05, "loss": 0.0232, "step": 3880 }, { "grad_norm": 0.5858079791069031, "learning_rate": 9.440302232651988e-05, "loss": 0.0249, "step": 3890 }, { "grad_norm": 0.4531865119934082, "learning_rate": 9.436495443419795e-05, "loss": 0.0252, "step": 3900 }, { "grad_norm": 0.4340672791004181, "learning_rate": 9.432676524978466e-05, "loss": 0.0291, "step": 3910 }, { "grad_norm": 0.5526710748672485, "learning_rate": 9.42884548776878e-05, "loss": 0.0312, "step": 3920 }, { "grad_norm": 0.5022578835487366, "learning_rate": 9.425002342264646e-05, "loss": 0.023, "step": 3930 }, { "grad_norm": 0.4777950048446655, "learning_rate": 9.421147098973077e-05, "loss": 0.0251, "step": 3940 }, { "grad_norm": 0.4220910370349884, "learning_rate": 9.41727976843416e-05, "loss": 0.0236, "step": 3950 }, { "grad_norm": 0.3944031298160553, "learning_rate": 9.413400361221029e-05, "loss": 0.0252, "step": 3960 }, { "grad_norm": 0.48305076360702515, "learning_rate": 9.409508887939835e-05, "loss": 0.0246, "step": 3970 }, { "grad_norm": 0.5534611344337463, "learning_rate": 9.40560535922972e-05, "loss": 0.0309, "step": 3980 }, { "grad_norm": 0.5242286324501038, "learning_rate": 9.40168978576278e-05, "loss": 0.0298, "step": 3990 }, { "grad_norm": 0.5534637570381165, "learning_rate": 9.397762178244043e-05, "loss": 0.0242, "step": 4000 }, { "grad_norm": 0.4518067538738251, "learning_rate": 9.393822547411439e-05, "loss": 0.0241, "step": 4010 }, { "grad_norm": 0.5187405347824097, "learning_rate": 9.389870904035769e-05, "loss": 0.0238, "step": 4020 }, { "grad_norm": 0.5443305969238281, "learning_rate": 9.385907258920672e-05, "loss": 0.0223, "step": 4030 }, { "grad_norm": 0.45573845505714417, "learning_rate": 9.381931622902607e-05, "loss": 0.0225, "step": 4040 }, { "grad_norm": 0.5460073351860046, "learning_rate": 9.377944006850807e-05, "loss": 0.0205, "step": 4050 }, { "grad_norm": 0.5589492321014404, "learning_rate": 9.373944421667265e-05, "loss": 0.0238, "step": 4060 }, { "grad_norm": 0.6063708066940308, "learning_rate": 9.369932878286691e-05, "loss": 0.0226, "step": 4070 }, { "grad_norm": 0.5444350242614746, "learning_rate": 9.365909387676494e-05, "loss": 0.023, "step": 4080 }, { "grad_norm": 0.4418700635433197, "learning_rate": 9.361873960836744e-05, "loss": 0.0233, "step": 4090 }, { "grad_norm": 0.4823611378669739, "learning_rate": 9.357826608800142e-05, "loss": 0.0224, "step": 4100 }, { "grad_norm": 0.43311965465545654, "learning_rate": 9.353767342631994e-05, "loss": 0.0237, "step": 4110 }, { "grad_norm": 0.5318973660469055, "learning_rate": 9.34969617343018e-05, "loss": 0.0216, "step": 4120 }, { "grad_norm": 0.44471144676208496, "learning_rate": 9.345613112325122e-05, "loss": 0.0207, "step": 4130 }, { "grad_norm": 0.48535484075546265, "learning_rate": 9.34151817047975e-05, "loss": 0.0257, "step": 4140 }, { "grad_norm": 0.4556364417076111, "learning_rate": 9.33741135908948e-05, "loss": 0.0211, "step": 4150 }, { "grad_norm": 0.3791288435459137, "learning_rate": 9.33329268938218e-05, "loss": 0.0254, "step": 4160 }, { "grad_norm": 0.4834822416305542, "learning_rate": 9.329162172618132e-05, "loss": 0.0232, "step": 4170 }, { "grad_norm": 0.3859091103076935, "learning_rate": 9.325019820090013e-05, "loss": 0.0227, "step": 4180 }, { "grad_norm": 0.4519246220588684, "learning_rate": 9.320865643122855e-05, "loss": 0.0212, "step": 4190 }, { "grad_norm": 0.37780097126960754, "learning_rate": 9.316699653074023e-05, "loss": 0.0208, "step": 4200 }, { "grad_norm": 0.37577447295188904, "learning_rate": 9.312521861333172e-05, "loss": 0.0209, "step": 4210 }, { "grad_norm": 0.44460707902908325, "learning_rate": 9.308332279322224e-05, "loss": 0.0207, "step": 4220 }, { "grad_norm": 0.4449606239795685, "learning_rate": 9.304130918495338e-05, "loss": 0.019, "step": 4230 }, { "grad_norm": 0.4963322579860687, "learning_rate": 9.299917790338874e-05, "loss": 0.0241, "step": 4240 }, { "grad_norm": 0.3826870918273926, "learning_rate": 9.295692906371363e-05, "loss": 0.0232, "step": 4250 }, { "grad_norm": 0.5094498991966248, "learning_rate": 9.291456278143476e-05, "loss": 0.0238, "step": 4260 }, { "grad_norm": 0.41625168919563293, "learning_rate": 9.287207917237994e-05, "loss": 0.0215, "step": 4270 }, { "grad_norm": 0.39011603593826294, "learning_rate": 9.282947835269773e-05, "loss": 0.0211, "step": 4280 }, { "grad_norm": 0.47361427545547485, "learning_rate": 9.278676043885715e-05, "loss": 0.0216, "step": 4290 }, { "grad_norm": 0.44294679164886475, "learning_rate": 9.274392554764733e-05, "loss": 0.0226, "step": 4300 }, { "grad_norm": 0.4601147174835205, "learning_rate": 9.270097379617723e-05, "loss": 0.0235, "step": 4310 }, { "grad_norm": 0.35954734683036804, "learning_rate": 9.26579053018753e-05, "loss": 0.0204, "step": 4320 }, { "grad_norm": 0.4399864375591278, "learning_rate": 9.261472018248918e-05, "loss": 0.022, "step": 4330 }, { "grad_norm": 0.4609746038913727, "learning_rate": 9.25714185560853e-05, "loss": 0.0237, "step": 4340 }, { "grad_norm": 0.4951179623603821, "learning_rate": 9.252800054104868e-05, "loss": 0.0219, "step": 4350 }, { "grad_norm": 0.48376208543777466, "learning_rate": 9.248446625608252e-05, "loss": 0.0198, "step": 4360 }, { "grad_norm": 0.4229479134082794, "learning_rate": 9.244081582020789e-05, "loss": 0.0225, "step": 4370 }, { "grad_norm": 0.42398297786712646, "learning_rate": 9.239704935276339e-05, "loss": 0.0236, "step": 4380 }, { "grad_norm": 0.5964459776878357, "learning_rate": 9.235316697340489e-05, "loss": 0.0224, "step": 4390 }, { "grad_norm": 0.5380576848983765, "learning_rate": 9.230916880210512e-05, "loss": 0.0222, "step": 4400 }, { "grad_norm": 0.5750419497489929, "learning_rate": 9.226505495915342e-05, "loss": 0.0246, "step": 4410 }, { "grad_norm": 0.5668116211891174, "learning_rate": 9.222082556515536e-05, "loss": 0.0213, "step": 4420 }, { "grad_norm": 0.47798052430152893, "learning_rate": 9.217648074103242e-05, "loss": 0.0234, "step": 4430 }, { "grad_norm": 0.41830500960350037, "learning_rate": 9.213202060802161e-05, "loss": 0.0215, "step": 4440 }, { "grad_norm": 0.5164875388145447, "learning_rate": 9.208744528767528e-05, "loss": 0.0236, "step": 4450 }, { "grad_norm": 0.34764987230300903, "learning_rate": 9.204275490186064e-05, "loss": 0.0192, "step": 4460 }, { "grad_norm": 0.45307204127311707, "learning_rate": 9.199794957275949e-05, "loss": 0.0193, "step": 4470 }, { "grad_norm": 0.4712885916233063, "learning_rate": 9.19530294228679e-05, "loss": 0.0232, "step": 4480 }, { "grad_norm": 0.5202063322067261, "learning_rate": 9.190799457499583e-05, "loss": 0.0202, "step": 4490 }, { "grad_norm": 0.4464549124240875, "learning_rate": 9.186284515226686e-05, "loss": 0.0234, "step": 4500 }, { "grad_norm": 0.451564759016037, "learning_rate": 9.181758127811777e-05, "loss": 0.0194, "step": 4510 }, { "grad_norm": 0.44847092032432556, "learning_rate": 9.177220307629825e-05, "loss": 0.0197, "step": 4520 }, { "grad_norm": 0.366699755191803, "learning_rate": 9.172671067087059e-05, "loss": 0.0206, "step": 4530 }, { "grad_norm": 0.41190415620803833, "learning_rate": 9.16811041862093e-05, "loss": 0.02, "step": 4540 }, { "grad_norm": 0.42311564087867737, "learning_rate": 9.163538374700076e-05, "loss": 0.0213, "step": 4550 }, { "grad_norm": 0.4553145468235016, "learning_rate": 9.158954947824287e-05, "loss": 0.0215, "step": 4560 }, { "grad_norm": 0.4500235319137573, "learning_rate": 9.154360150524482e-05, "loss": 0.0187, "step": 4570 }, { "grad_norm": 0.43962424993515015, "learning_rate": 9.14975399536266e-05, "loss": 0.0238, "step": 4580 }, { "grad_norm": 0.5361406207084656, "learning_rate": 9.14513649493187e-05, "loss": 0.0236, "step": 4590 }, { "grad_norm": 0.3396156132221222, "learning_rate": 9.140507661856187e-05, "loss": 0.0211, "step": 4600 }, { "grad_norm": 0.41475018858909607, "learning_rate": 9.135867508790661e-05, "loss": 0.0223, "step": 4610 }, { "grad_norm": 0.5147063136100769, "learning_rate": 9.131216048421291e-05, "loss": 0.0206, "step": 4620 }, { "grad_norm": 0.3312698006629944, "learning_rate": 9.126553293464998e-05, "loss": 0.0212, "step": 4630 }, { "grad_norm": 0.3740409016609192, "learning_rate": 9.121879256669572e-05, "loss": 0.0194, "step": 4640 }, { "grad_norm": 0.3619629740715027, "learning_rate": 9.117193950813652e-05, "loss": 0.0206, "step": 4650 }, { "grad_norm": 0.28074416518211365, "learning_rate": 9.112497388706685e-05, "loss": 0.0205, "step": 4660 }, { "grad_norm": 0.45012614130973816, "learning_rate": 9.10778958318889e-05, "loss": 0.0216, "step": 4670 }, { "grad_norm": 0.4578162133693695, "learning_rate": 9.103070547131232e-05, "loss": 0.0198, "step": 4680 }, { "grad_norm": 0.5081265568733215, "learning_rate": 9.098340293435375e-05, "loss": 0.0249, "step": 4690 }, { "grad_norm": 0.4535750448703766, "learning_rate": 9.093598835033649e-05, "loss": 0.0205, "step": 4700 }, { "grad_norm": 0.46924296021461487, "learning_rate": 9.088846184889021e-05, "loss": 0.0209, "step": 4710 }, { "grad_norm": 0.452144980430603, "learning_rate": 9.084082355995057e-05, "loss": 0.0224, "step": 4720 }, { "grad_norm": 0.4810640513896942, "learning_rate": 9.079307361375882e-05, "loss": 0.0214, "step": 4730 }, { "grad_norm": 0.3681242763996124, "learning_rate": 9.074521214086149e-05, "loss": 0.0209, "step": 4740 }, { "grad_norm": 0.5337221026420593, "learning_rate": 9.069723927211001e-05, "loss": 0.0233, "step": 4750 }, { "grad_norm": 0.4345974624156952, "learning_rate": 9.064915513866037e-05, "loss": 0.0226, "step": 4760 }, { "grad_norm": 0.4884747862815857, "learning_rate": 9.060095987197279e-05, "loss": 0.0218, "step": 4770 }, { "grad_norm": 0.42217907309532166, "learning_rate": 9.055265360381126e-05, "loss": 0.0227, "step": 4780 }, { "grad_norm": 0.4298636019229889, "learning_rate": 9.050423646624326e-05, "loss": 0.0223, "step": 4790 }, { "grad_norm": 0.48070669174194336, "learning_rate": 9.045570859163943e-05, "loss": 0.0198, "step": 4800 }, { "grad_norm": 0.4453526735305786, "learning_rate": 9.04070701126731e-05, "loss": 0.0252, "step": 4810 }, { "grad_norm": 0.4198642075061798, "learning_rate": 9.035832116232001e-05, "loss": 0.0208, "step": 4820 }, { "grad_norm": 0.4447406530380249, "learning_rate": 9.030946187385796e-05, "loss": 0.022, "step": 4830 }, { "grad_norm": 0.4191599190235138, "learning_rate": 9.026049238086635e-05, "loss": 0.0215, "step": 4840 }, { "grad_norm": 0.3787648677825928, "learning_rate": 9.021141281722591e-05, "loss": 0.0193, "step": 4850 }, { "grad_norm": 0.5228450298309326, "learning_rate": 9.01622233171183e-05, "loss": 0.0229, "step": 4860 }, { "grad_norm": 0.48083633184432983, "learning_rate": 9.011292401502574e-05, "loss": 0.0217, "step": 4870 }, { "grad_norm": 0.4551771283149719, "learning_rate": 9.006351504573063e-05, "loss": 0.0199, "step": 4880 }, { "grad_norm": 0.36222052574157715, "learning_rate": 9.001399654431519e-05, "loss": 0.0191, "step": 4890 }, { "grad_norm": 0.3971017599105835, "learning_rate": 8.996436864616116e-05, "loss": 0.0194, "step": 4900 }, { "grad_norm": 0.45447686314582825, "learning_rate": 8.991463148694925e-05, "loss": 0.0197, "step": 4910 }, { "grad_norm": 0.5166178345680237, "learning_rate": 8.986478520265902e-05, "loss": 0.0242, "step": 4920 }, { "grad_norm": 0.4508470892906189, "learning_rate": 8.981482992956827e-05, "loss": 0.023, "step": 4930 }, { "grad_norm": 0.4044328033924103, "learning_rate": 8.976476580425282e-05, "loss": 0.02, "step": 4940 }, { "grad_norm": 0.35950005054473877, "learning_rate": 8.971459296358606e-05, "loss": 0.0229, "step": 4950 }, { "grad_norm": 0.4009755551815033, "learning_rate": 8.966431154473864e-05, "loss": 0.0226, "step": 4960 }, { "grad_norm": 0.37669211626052856, "learning_rate": 8.961392168517803e-05, "loss": 0.0206, "step": 4970 }, { "grad_norm": 0.4629116654396057, "learning_rate": 8.956342352266821e-05, "loss": 0.0221, "step": 4980 }, { "grad_norm": 0.43860289454460144, "learning_rate": 8.95128171952692e-05, "loss": 0.0209, "step": 4990 }, { "grad_norm": 0.5497449040412903, "learning_rate": 8.946210284133676e-05, "loss": 0.023, "step": 5000 }, { "grad_norm": 0.5355510711669922, "learning_rate": 8.941128059952201e-05, "loss": 0.0209, "step": 5010 }, { "grad_norm": 0.40008142590522766, "learning_rate": 8.936035060877102e-05, "loss": 0.0222, "step": 5020 }, { "grad_norm": 0.5445119142532349, "learning_rate": 8.930931300832443e-05, "loss": 0.02, "step": 5030 }, { "grad_norm": 0.4147629737854004, "learning_rate": 8.925816793771711e-05, "loss": 0.0204, "step": 5040 }, { "grad_norm": 0.34871765971183777, "learning_rate": 8.92069155367777e-05, "loss": 0.0184, "step": 5050 }, { "grad_norm": 0.47016510367393494, "learning_rate": 8.915555594562834e-05, "loss": 0.0196, "step": 5060 }, { "grad_norm": 0.4153212606906891, "learning_rate": 8.910408930468416e-05, "loss": 0.018, "step": 5070 }, { "grad_norm": 0.4597940444946289, "learning_rate": 8.905251575465303e-05, "loss": 0.018, "step": 5080 }, { "grad_norm": 0.501157283782959, "learning_rate": 8.900083543653502e-05, "loss": 0.0195, "step": 5090 }, { "grad_norm": 0.4347209334373474, "learning_rate": 8.894904849162218e-05, "loss": 0.0194, "step": 5100 }, { "grad_norm": 0.4628390967845917, "learning_rate": 8.889715506149802e-05, "loss": 0.0202, "step": 5110 }, { "grad_norm": 0.4560784697532654, "learning_rate": 8.884515528803722e-05, "loss": 0.0193, "step": 5120 }, { "grad_norm": 0.533446192741394, "learning_rate": 8.879304931340517e-05, "loss": 0.0204, "step": 5130 }, { "grad_norm": 0.42130136489868164, "learning_rate": 8.874083728005759e-05, "loss": 0.0223, "step": 5140 }, { "grad_norm": 0.5781335830688477, "learning_rate": 8.868851933074021e-05, "loss": 0.0202, "step": 5150 }, { "grad_norm": 0.37779104709625244, "learning_rate": 8.863609560848829e-05, "loss": 0.0204, "step": 5160 }, { "grad_norm": 0.36015090346336365, "learning_rate": 8.85835662566263e-05, "loss": 0.0171, "step": 5170 }, { "grad_norm": 0.30681851506233215, "learning_rate": 8.853093141876747e-05, "loss": 0.0173, "step": 5180 }, { "grad_norm": 0.3919036090373993, "learning_rate": 8.847819123881343e-05, "loss": 0.0212, "step": 5190 }, { "grad_norm": 0.3171193301677704, "learning_rate": 8.842534586095383e-05, "loss": 0.0192, "step": 5200 }, { "grad_norm": 0.4242173433303833, "learning_rate": 8.837239542966593e-05, "loss": 0.0193, "step": 5210 }, { "grad_norm": 0.4665341079235077, "learning_rate": 8.831934008971417e-05, "loss": 0.0196, "step": 5220 }, { "grad_norm": 0.48564082384109497, "learning_rate": 8.826617998614982e-05, "loss": 0.0207, "step": 5230 }, { "grad_norm": 0.3576866090297699, "learning_rate": 8.821291526431056e-05, "loss": 0.0209, "step": 5240 }, { "grad_norm": 0.4278208017349243, "learning_rate": 8.815954606982015e-05, "loss": 0.0192, "step": 5250 }, { "grad_norm": 0.44816070795059204, "learning_rate": 8.810607254858789e-05, "loss": 0.0219, "step": 5260 }, { "grad_norm": 0.41479766368865967, "learning_rate": 8.805249484680838e-05, "loss": 0.0202, "step": 5270 }, { "grad_norm": 0.46126794815063477, "learning_rate": 8.799881311096096e-05, "loss": 0.0197, "step": 5280 }, { "grad_norm": 0.3611409664154053, "learning_rate": 8.794502748780949e-05, "loss": 0.0183, "step": 5290 }, { "grad_norm": 0.4593311548233032, "learning_rate": 8.78911381244018e-05, "loss": 0.0197, "step": 5300 }, { "grad_norm": 0.39980199933052063, "learning_rate": 8.783714516806933e-05, "loss": 0.0182, "step": 5310 }, { "grad_norm": 0.36190369725227356, "learning_rate": 8.77830487664268e-05, "loss": 0.0188, "step": 5320 }, { "grad_norm": 0.383963942527771, "learning_rate": 8.772884906737167e-05, "loss": 0.0182, "step": 5330 }, { "grad_norm": 0.3983629643917084, "learning_rate": 8.767454621908387e-05, "loss": 0.0195, "step": 5340 }, { "grad_norm": 0.5490370392799377, "learning_rate": 8.76201403700253e-05, "loss": 0.0188, "step": 5350 }, { "grad_norm": 0.4522782862186432, "learning_rate": 8.756563166893949e-05, "loss": 0.02, "step": 5360 }, { "grad_norm": 0.4271126389503479, "learning_rate": 8.751102026485113e-05, "loss": 0.0192, "step": 5370 }, { "grad_norm": 0.4281386435031891, "learning_rate": 8.745630630706571e-05, "loss": 0.0185, "step": 5380 }, { "grad_norm": 0.4630153477191925, "learning_rate": 8.740148994516912e-05, "loss": 0.0181, "step": 5390 }, { "grad_norm": 0.37567976117134094, "learning_rate": 8.73465713290272e-05, "loss": 0.0185, "step": 5400 }, { "grad_norm": 0.38618242740631104, "learning_rate": 8.729155060878533e-05, "loss": 0.0182, "step": 5410 }, { "grad_norm": 0.44467517733573914, "learning_rate": 8.723642793486809e-05, "loss": 0.0202, "step": 5420 }, { "grad_norm": 0.448213130235672, "learning_rate": 8.718120345797873e-05, "loss": 0.0189, "step": 5430 }, { "grad_norm": 0.3939504325389862, "learning_rate": 8.712587732909889e-05, "loss": 0.0187, "step": 5440 }, { "grad_norm": 0.5161504149436951, "learning_rate": 8.707044969948806e-05, "loss": 0.0186, "step": 5450 }, { "grad_norm": 0.3795504570007324, "learning_rate": 8.701492072068329e-05, "loss": 0.0175, "step": 5460 }, { "grad_norm": 0.40509411692619324, "learning_rate": 8.695929054449869e-05, "loss": 0.0192, "step": 5470 }, { "grad_norm": 0.315085232257843, "learning_rate": 8.690355932302501e-05, "loss": 0.0181, "step": 5480 }, { "grad_norm": 0.27766484022140503, "learning_rate": 8.684772720862931e-05, "loss": 0.0171, "step": 5490 }, { "grad_norm": 0.3611393868923187, "learning_rate": 8.679179435395446e-05, "loss": 0.0212, "step": 5500 }, { "grad_norm": 0.4436500370502472, "learning_rate": 8.673576091191874e-05, "loss": 0.0172, "step": 5510 }, { "grad_norm": 0.4194873571395874, "learning_rate": 8.667962703571541e-05, "loss": 0.018, "step": 5520 }, { "grad_norm": 0.4791281223297119, "learning_rate": 8.662339287881238e-05, "loss": 0.0168, "step": 5530 }, { "grad_norm": 0.5352644324302673, "learning_rate": 8.656705859495169e-05, "loss": 0.0196, "step": 5540 }, { "grad_norm": 0.37074315547943115, "learning_rate": 8.651062433814912e-05, "loss": 0.0189, "step": 5550 }, { "grad_norm": 0.3646714389324188, "learning_rate": 8.645409026269375e-05, "loss": 0.0214, "step": 5560 }, { "grad_norm": 0.3759939670562744, "learning_rate": 8.639745652314759e-05, "loss": 0.019, "step": 5570 }, { "grad_norm": 0.3688820004463196, "learning_rate": 8.634072327434515e-05, "loss": 0.0229, "step": 5580 }, { "grad_norm": 0.3983011245727539, "learning_rate": 8.628389067139294e-05, "loss": 0.0189, "step": 5590 }, { "grad_norm": 0.3327280879020691, "learning_rate": 8.622695886966911e-05, "loss": 0.0241, "step": 5600 }, { "grad_norm": 0.4210691452026367, "learning_rate": 8.616992802482308e-05, "loss": 0.0215, "step": 5610 }, { "grad_norm": 0.3366101086139679, "learning_rate": 8.611279829277496e-05, "loss": 0.0194, "step": 5620 }, { "grad_norm": 0.3807818591594696, "learning_rate": 8.605556982971528e-05, "loss": 0.0208, "step": 5630 }, { "grad_norm": 0.3806864619255066, "learning_rate": 8.599824279210447e-05, "loss": 0.017, "step": 5640 }, { "grad_norm": 0.35536840558052063, "learning_rate": 8.594081733667243e-05, "loss": 0.0192, "step": 5650 }, { "grad_norm": 0.40165260434150696, "learning_rate": 8.58832936204182e-05, "loss": 0.0179, "step": 5660 }, { "grad_norm": 0.3451436161994934, "learning_rate": 8.582567180060942e-05, "loss": 0.0216, "step": 5670 }, { "grad_norm": 0.3748839199542999, "learning_rate": 8.576795203478194e-05, "loss": 0.017, "step": 5680 }, { "grad_norm": 0.335220068693161, "learning_rate": 8.571013448073939e-05, "loss": 0.0188, "step": 5690 }, { "grad_norm": 0.4378359913825989, "learning_rate": 8.565221929655275e-05, "loss": 0.0158, "step": 5700 }, { "grad_norm": 0.4868551194667816, "learning_rate": 8.559420664055992e-05, "loss": 0.0197, "step": 5710 }, { "grad_norm": 0.32548487186431885, "learning_rate": 8.553609667136532e-05, "loss": 0.0168, "step": 5720 }, { "grad_norm": 0.46054816246032715, "learning_rate": 8.547788954783936e-05, "loss": 0.0212, "step": 5730 }, { "grad_norm": 0.390742689371109, "learning_rate": 8.541958542911808e-05, "loss": 0.0196, "step": 5740 }, { "grad_norm": 0.3806167542934418, "learning_rate": 8.536118447460275e-05, "loss": 0.0183, "step": 5750 }, { "grad_norm": 0.49352869391441345, "learning_rate": 8.530268684395932e-05, "loss": 0.0183, "step": 5760 }, { "grad_norm": 0.3853674829006195, "learning_rate": 8.524409269711807e-05, "loss": 0.019, "step": 5770 }, { "grad_norm": 0.4234221577644348, "learning_rate": 8.51854021942732e-05, "loss": 0.0182, "step": 5780 }, { "grad_norm": 0.37306585907936096, "learning_rate": 8.512661549588227e-05, "loss": 0.0169, "step": 5790 }, { "grad_norm": 0.34233710169792175, "learning_rate": 8.506773276266588e-05, "loss": 0.0181, "step": 5800 }, { "grad_norm": 0.41399529576301575, "learning_rate": 8.500875415560721e-05, "loss": 0.0183, "step": 5810 }, { "grad_norm": 0.32870060205459595, "learning_rate": 8.494967983595144e-05, "loss": 0.0173, "step": 5820 }, { "grad_norm": 0.3568831980228424, "learning_rate": 8.489050996520558e-05, "loss": 0.0161, "step": 5830 }, { "grad_norm": 0.4269106090068817, "learning_rate": 8.483124470513775e-05, "loss": 0.0174, "step": 5840 }, { "grad_norm": 0.424483984708786, "learning_rate": 8.477188421777692e-05, "loss": 0.0188, "step": 5850 }, { "grad_norm": 0.35122954845428467, "learning_rate": 8.47124286654124e-05, "loss": 0.0188, "step": 5860 }, { "grad_norm": 0.3396788239479065, "learning_rate": 8.465287821059341e-05, "loss": 0.0165, "step": 5870 }, { "grad_norm": 0.4248601496219635, "learning_rate": 8.45932330161286e-05, "loss": 0.0175, "step": 5880 }, { "grad_norm": 0.41610410809516907, "learning_rate": 8.453349324508567e-05, "loss": 0.015, "step": 5890 }, { "grad_norm": 0.43096163868904114, "learning_rate": 8.447365906079088e-05, "loss": 0.018, "step": 5900 }, { "grad_norm": 0.4176206588745117, "learning_rate": 8.441373062682856e-05, "loss": 0.0179, "step": 5910 }, { "grad_norm": 0.433650940656662, "learning_rate": 8.43537081070408e-05, "loss": 0.0168, "step": 5920 }, { "grad_norm": 0.3554627001285553, "learning_rate": 8.429359166552689e-05, "loss": 0.0198, "step": 5930 }, { "grad_norm": 0.3301181197166443, "learning_rate": 8.423338146664284e-05, "loss": 0.0177, "step": 5940 }, { "grad_norm": 0.38023489713668823, "learning_rate": 8.417307767500107e-05, "loss": 0.0199, "step": 5950 }, { "grad_norm": 0.4711596667766571, "learning_rate": 8.411268045546983e-05, "loss": 0.0174, "step": 5960 }, { "grad_norm": 0.40072959661483765, "learning_rate": 8.405218997317281e-05, "loss": 0.0166, "step": 5970 }, { "grad_norm": 0.3313545882701874, "learning_rate": 8.399160639348869e-05, "loss": 0.0165, "step": 5980 }, { "grad_norm": 0.392516165971756, "learning_rate": 8.393092988205065e-05, "loss": 0.0183, "step": 5990 }, { "grad_norm": 0.33452412486076355, "learning_rate": 8.387016060474597e-05, "loss": 0.0156, "step": 6000 }, { "grad_norm": 0.39347368478775024, "learning_rate": 8.380929872771551e-05, "loss": 0.0193, "step": 6010 }, { "grad_norm": 0.4057534337043762, "learning_rate": 8.374834441735335e-05, "loss": 0.0181, "step": 6020 }, { "grad_norm": 0.3926507830619812, "learning_rate": 8.368729784030622e-05, "loss": 0.0204, "step": 6030 }, { "grad_norm": 0.4500643312931061, "learning_rate": 8.362615916347315e-05, "loss": 0.0166, "step": 6040 }, { "grad_norm": 0.38587507605552673, "learning_rate": 8.356492855400493e-05, "loss": 0.0172, "step": 6050 }, { "grad_norm": 0.3147795498371124, "learning_rate": 8.350360617930371e-05, "loss": 0.018, "step": 6060 }, { "grad_norm": 0.38834115862846375, "learning_rate": 8.344219220702255e-05, "loss": 0.0168, "step": 6070 }, { "grad_norm": 0.46887388825416565, "learning_rate": 8.338068680506485e-05, "loss": 0.0227, "step": 6080 }, { "grad_norm": 0.5128589272499084, "learning_rate": 8.33190901415841e-05, "loss": 0.0186, "step": 6090 }, { "grad_norm": 0.3226715624332428, "learning_rate": 8.325740238498317e-05, "loss": 0.0174, "step": 6100 }, { "grad_norm": 0.39391645789146423, "learning_rate": 8.319562370391406e-05, "loss": 0.019, "step": 6110 }, { "grad_norm": 0.4222080409526825, "learning_rate": 8.31337542672773e-05, "loss": 0.0179, "step": 6120 }, { "grad_norm": 0.4449750781059265, "learning_rate": 8.307179424422158e-05, "loss": 0.0192, "step": 6130 }, { "grad_norm": 0.36552441120147705, "learning_rate": 8.300974380414327e-05, "loss": 0.0189, "step": 6140 }, { "grad_norm": 0.35965070128440857, "learning_rate": 8.294760311668586e-05, "loss": 0.0157, "step": 6150 }, { "grad_norm": 0.3176078498363495, "learning_rate": 8.288537235173961e-05, "loss": 0.0159, "step": 6160 }, { "grad_norm": 0.35276660323143005, "learning_rate": 8.282305167944108e-05, "loss": 0.0155, "step": 6170 }, { "grad_norm": 0.40630823373794556, "learning_rate": 8.276064127017262e-05, "loss": 0.0153, "step": 6180 }, { "grad_norm": 0.43017467856407166, "learning_rate": 8.269814129456189e-05, "loss": 0.0167, "step": 6190 }, { "grad_norm": 0.32223403453826904, "learning_rate": 8.263555192348143e-05, "loss": 0.0172, "step": 6200 }, { "grad_norm": 0.38439658284187317, "learning_rate": 8.257287332804819e-05, "loss": 0.0185, "step": 6210 }, { "grad_norm": 0.4889625608921051, "learning_rate": 8.251010567962307e-05, "loss": 0.0175, "step": 6220 }, { "grad_norm": 0.29807227849960327, "learning_rate": 8.244724914981041e-05, "loss": 0.0171, "step": 6230 }, { "grad_norm": 0.4694109559059143, "learning_rate": 8.238430391045757e-05, "loss": 0.0169, "step": 6240 }, { "grad_norm": 0.3182734251022339, "learning_rate": 8.232127013365445e-05, "loss": 0.0164, "step": 6250 }, { "grad_norm": 0.350523442029953, "learning_rate": 8.225814799173295e-05, "loss": 0.0149, "step": 6260 }, { "grad_norm": 0.3217054009437561, "learning_rate": 8.219493765726663e-05, "loss": 0.0148, "step": 6270 }, { "grad_norm": 0.3635313808917999, "learning_rate": 8.21316393030701e-05, "loss": 0.0148, "step": 6280 }, { "grad_norm": 0.3793472945690155, "learning_rate": 8.206825310219865e-05, "loss": 0.0175, "step": 6290 }, { "grad_norm": 0.3133487105369568, "learning_rate": 8.200477922794776e-05, "loss": 0.0161, "step": 6300 }, { "grad_norm": 0.39939531683921814, "learning_rate": 8.194121785385256e-05, "loss": 0.016, "step": 6310 }, { "grad_norm": 0.41280031204223633, "learning_rate": 8.187756915368741e-05, "loss": 0.0166, "step": 6320 }, { "grad_norm": 0.31678634881973267, "learning_rate": 8.181383330146544e-05, "loss": 0.017, "step": 6330 }, { "grad_norm": 0.3112114667892456, "learning_rate": 8.175001047143804e-05, "loss": 0.0159, "step": 6340 }, { "grad_norm": 0.2574307322502136, "learning_rate": 8.168610083809438e-05, "loss": 0.0199, "step": 6350 }, { "grad_norm": 0.3815240263938904, "learning_rate": 8.162210457616095e-05, "loss": 0.0148, "step": 6360 }, { "grad_norm": 0.33287325501441956, "learning_rate": 8.155802186060109e-05, "loss": 0.0164, "step": 6370 }, { "grad_norm": 0.43456634879112244, "learning_rate": 8.149385286661453e-05, "loss": 0.0161, "step": 6380 }, { "grad_norm": 0.4100537598133087, "learning_rate": 8.14295977696368e-05, "loss": 0.018, "step": 6390 }, { "grad_norm": 0.3589387536048889, "learning_rate": 8.13652567453389e-05, "loss": 0.0202, "step": 6400 }, { "grad_norm": 0.36775296926498413, "learning_rate": 8.130082996962676e-05, "loss": 0.0191, "step": 6410 }, { "grad_norm": 0.3752509653568268, "learning_rate": 8.123631761864068e-05, "loss": 0.0178, "step": 6420 }, { "grad_norm": 0.354131281375885, "learning_rate": 8.1171719868755e-05, "loss": 0.0149, "step": 6430 }, { "grad_norm": 0.3606273829936981, "learning_rate": 8.110703689657748e-05, "loss": 0.0165, "step": 6440 }, { "grad_norm": 0.3572201430797577, "learning_rate": 8.104226887894892e-05, "loss": 0.0157, "step": 6450 }, { "grad_norm": 0.3627186715602875, "learning_rate": 8.097741599294257e-05, "loss": 0.0168, "step": 6460 }, { "grad_norm": 0.4589333236217499, "learning_rate": 8.091247841586378e-05, "loss": 0.0153, "step": 6470 }, { "grad_norm": 0.38536688685417175, "learning_rate": 8.084745632524939e-05, "loss": 0.0165, "step": 6480 }, { "grad_norm": 0.38711464405059814, "learning_rate": 8.07823498988673e-05, "loss": 0.0176, "step": 6490 }, { "grad_norm": 0.3633858561515808, "learning_rate": 8.071715931471602e-05, "loss": 0.0172, "step": 6500 }, { "grad_norm": 0.33100709319114685, "learning_rate": 8.06518847510241e-05, "loss": 0.0196, "step": 6510 }, { "grad_norm": 0.40418097376823425, "learning_rate": 8.058652638624971e-05, "loss": 0.0158, "step": 6520 }, { "grad_norm": 0.4954093396663666, "learning_rate": 8.052108439908013e-05, "loss": 0.0183, "step": 6530 }, { "grad_norm": 0.34467601776123047, "learning_rate": 8.045555896843125e-05, "loss": 0.0136, "step": 6540 }, { "grad_norm": 0.3729366362094879, "learning_rate": 8.03899502734471e-05, "loss": 0.0182, "step": 6550 }, { "grad_norm": 0.3742976784706116, "learning_rate": 8.032425849349931e-05, "loss": 0.0173, "step": 6560 }, { "grad_norm": 0.33351370692253113, "learning_rate": 8.025848380818674e-05, "loss": 0.0166, "step": 6570 }, { "grad_norm": 0.28320348262786865, "learning_rate": 8.019262639733487e-05, "loss": 0.0137, "step": 6580 }, { "grad_norm": 0.31575536727905273, "learning_rate": 8.012668644099531e-05, "loss": 0.0158, "step": 6590 }, { "grad_norm": 0.34597280621528625, "learning_rate": 8.006066411944542e-05, "loss": 0.0165, "step": 6600 }, { "grad_norm": 0.34442538022994995, "learning_rate": 7.999455961318769e-05, "loss": 0.0159, "step": 6610 }, { "grad_norm": 0.42213618755340576, "learning_rate": 7.992837310294932e-05, "loss": 0.0188, "step": 6620 }, { "grad_norm": 0.3660734295845032, "learning_rate": 7.986210476968167e-05, "loss": 0.0184, "step": 6630 }, { "grad_norm": 0.2793416976928711, "learning_rate": 7.97957547945599e-05, "loss": 0.018, "step": 6640 }, { "grad_norm": 0.3526296317577362, "learning_rate": 7.972932335898226e-05, "loss": 0.0192, "step": 6650 }, { "grad_norm": 0.368076354265213, "learning_rate": 7.966281064456975e-05, "loss": 0.015, "step": 6660 }, { "grad_norm": 0.35590529441833496, "learning_rate": 7.959621683316563e-05, "loss": 0.0174, "step": 6670 }, { "grad_norm": 0.30636563897132874, "learning_rate": 7.952954210683481e-05, "loss": 0.0151, "step": 6680 }, { "grad_norm": 0.2883080840110779, "learning_rate": 7.946278664786345e-05, "loss": 0.0161, "step": 6690 }, { "grad_norm": 0.3183375895023346, "learning_rate": 7.939595063875842e-05, "loss": 0.0143, "step": 6700 }, { "grad_norm": 0.5150044560432434, "learning_rate": 7.932903426224683e-05, "loss": 0.0159, "step": 6710 }, { "grad_norm": 0.43665438890457153, "learning_rate": 7.926203770127552e-05, "loss": 0.0191, "step": 6720 }, { "grad_norm": 0.35014694929122925, "learning_rate": 7.919496113901046e-05, "loss": 0.017, "step": 6730 }, { "grad_norm": 0.28918254375457764, "learning_rate": 7.912780475883649e-05, "loss": 0.0156, "step": 6740 }, { "grad_norm": 0.32302325963974, "learning_rate": 7.906056874435652e-05, "loss": 0.0207, "step": 6750 }, { "grad_norm": 0.3602825701236725, "learning_rate": 7.899325327939131e-05, "loss": 0.0177, "step": 6760 }, { "grad_norm": 0.3301911950111389, "learning_rate": 7.892585854797872e-05, "loss": 0.018, "step": 6770 }, { "grad_norm": 0.48308515548706055, "learning_rate": 7.88583847343734e-05, "loss": 0.0184, "step": 6780 }, { "grad_norm": 0.3158496916294098, "learning_rate": 7.879083202304616e-05, "loss": 0.0165, "step": 6790 }, { "grad_norm": 0.2676745355129242, "learning_rate": 7.872320059868355e-05, "loss": 0.0166, "step": 6800 }, { "grad_norm": 0.422525018453598, "learning_rate": 7.865549064618729e-05, "loss": 0.0161, "step": 6810 }, { "grad_norm": 0.4001980721950531, "learning_rate": 7.858770235067381e-05, "loss": 0.0185, "step": 6820 }, { "grad_norm": 0.38706573843955994, "learning_rate": 7.851983589747374e-05, "loss": 0.0169, "step": 6830 }, { "grad_norm": 0.31454819440841675, "learning_rate": 7.845189147213133e-05, "loss": 0.0159, "step": 6840 }, { "grad_norm": 0.3571377396583557, "learning_rate": 7.838386926040407e-05, "loss": 0.0152, "step": 6850 }, { "grad_norm": 0.35372400283813477, "learning_rate": 7.83157694482621e-05, "loss": 0.0154, "step": 6860 }, { "grad_norm": 0.37327447533607483, "learning_rate": 7.824759222188768e-05, "loss": 0.0157, "step": 6870 }, { "grad_norm": 0.3270111083984375, "learning_rate": 7.817933776767478e-05, "loss": 0.0163, "step": 6880 }, { "grad_norm": 0.3582060933113098, "learning_rate": 7.811100627222842e-05, "loss": 0.0145, "step": 6890 }, { "grad_norm": 0.3434228003025055, "learning_rate": 7.804259792236435e-05, "loss": 0.0171, "step": 6900 }, { "grad_norm": 0.37895652651786804, "learning_rate": 7.797411290510835e-05, "loss": 0.0175, "step": 6910 }, { "grad_norm": 0.3012599050998688, "learning_rate": 7.790555140769586e-05, "loss": 0.0175, "step": 6920 }, { "grad_norm": 0.2983647882938385, "learning_rate": 7.78369136175714e-05, "loss": 0.0167, "step": 6930 }, { "grad_norm": 0.3346477150917053, "learning_rate": 7.776819972238806e-05, "loss": 0.0142, "step": 6940 }, { "grad_norm": 0.35186803340911865, "learning_rate": 7.7699409910007e-05, "loss": 0.0163, "step": 6950 }, { "grad_norm": 0.29901954531669617, "learning_rate": 7.763054436849694e-05, "loss": 0.0151, "step": 6960 }, { "grad_norm": 0.37545618414878845, "learning_rate": 7.756160328613364e-05, "loss": 0.0154, "step": 6970 }, { "grad_norm": 0.30803343653678894, "learning_rate": 7.749258685139942e-05, "loss": 0.0175, "step": 6980 }, { "grad_norm": 0.3543394207954407, "learning_rate": 7.742349525298253e-05, "loss": 0.0145, "step": 6990 }, { "grad_norm": 0.30735963582992554, "learning_rate": 7.735432867977679e-05, "loss": 0.0143, "step": 7000 }, { "grad_norm": 0.32498881220817566, "learning_rate": 7.728508732088096e-05, "loss": 0.0161, "step": 7010 }, { "grad_norm": 0.24017657339572906, "learning_rate": 7.721577136559825e-05, "loss": 0.0145, "step": 7020 }, { "grad_norm": 0.3195246160030365, "learning_rate": 7.714638100343588e-05, "loss": 0.0139, "step": 7030 }, { "grad_norm": 0.33438125252723694, "learning_rate": 7.707691642410444e-05, "loss": 0.0151, "step": 7040 }, { "grad_norm": 0.3537907004356384, "learning_rate": 7.70073778175174e-05, "loss": 0.0149, "step": 7050 }, { "grad_norm": 0.43645310401916504, "learning_rate": 7.69377653737907e-05, "loss": 0.0138, "step": 7060 }, { "grad_norm": 0.37180522084236145, "learning_rate": 7.686807928324209e-05, "loss": 0.018, "step": 7070 }, { "grad_norm": 0.3348531424999237, "learning_rate": 7.679831973639065e-05, "loss": 0.0146, "step": 7080 }, { "grad_norm": 0.3049207627773285, "learning_rate": 7.672848692395637e-05, "loss": 0.016, "step": 7090 }, { "grad_norm": 0.3360995948314667, "learning_rate": 7.665858103685944e-05, "loss": 0.0147, "step": 7100 }, { "grad_norm": 0.3614061176776886, "learning_rate": 7.658860226621991e-05, "loss": 0.0161, "step": 7110 }, { "grad_norm": 0.3797241747379303, "learning_rate": 7.651855080335708e-05, "loss": 0.0153, "step": 7120 }, { "grad_norm": 0.32377055287361145, "learning_rate": 7.644842683978896e-05, "loss": 0.0169, "step": 7130 }, { "grad_norm": 0.31434696912765503, "learning_rate": 7.63782305672318e-05, "loss": 0.0144, "step": 7140 }, { "grad_norm": 0.3028211295604706, "learning_rate": 7.63079621775995e-05, "loss": 0.0136, "step": 7150 }, { "grad_norm": 0.3660510182380676, "learning_rate": 7.623762186300319e-05, "loss": 0.0162, "step": 7160 }, { "grad_norm": 0.3252999186515808, "learning_rate": 7.616720981575057e-05, "loss": 0.017, "step": 7170 }, { "grad_norm": 0.34896740317344666, "learning_rate": 7.609672622834552e-05, "loss": 0.0168, "step": 7180 }, { "grad_norm": 0.3260767459869385, "learning_rate": 7.602617129348747e-05, "loss": 0.0172, "step": 7190 }, { "grad_norm": 0.2719978988170624, "learning_rate": 7.595554520407088e-05, "loss": 0.015, "step": 7200 }, { "grad_norm": 0.30716848373413086, "learning_rate": 7.588484815318484e-05, "loss": 0.015, "step": 7210 }, { "grad_norm": 0.3577578067779541, "learning_rate": 7.581408033411234e-05, "loss": 0.0135, "step": 7220 }, { "grad_norm": 0.4308195114135742, "learning_rate": 7.574324194032995e-05, "loss": 0.0192, "step": 7230 }, { "grad_norm": 0.2891302704811096, "learning_rate": 7.567233316550705e-05, "loss": 0.0194, "step": 7240 }, { "grad_norm": 0.2902106046676636, "learning_rate": 7.560135420350562e-05, "loss": 0.015, "step": 7250 }, { "grad_norm": 0.2827083468437195, "learning_rate": 7.553030524837935e-05, "loss": 0.0146, "step": 7260 }, { "grad_norm": 0.43613916635513306, "learning_rate": 7.545918649437341e-05, "loss": 0.0176, "step": 7270 }, { "grad_norm": 0.39818841218948364, "learning_rate": 7.538799813592377e-05, "loss": 0.0163, "step": 7280 }, { "grad_norm": 0.33063554763793945, "learning_rate": 7.531674036765662e-05, "loss": 0.0166, "step": 7290 }, { "grad_norm": 0.3117780089378357, "learning_rate": 7.524541338438807e-05, "loss": 0.014, "step": 7300 }, { "grad_norm": 0.3826318085193634, "learning_rate": 7.517401738112328e-05, "loss": 0.0152, "step": 7310 }, { "grad_norm": 0.36848127841949463, "learning_rate": 7.510255255305628e-05, "loss": 0.0183, "step": 7320 }, { "grad_norm": 0.2968866527080536, "learning_rate": 7.503101909556911e-05, "loss": 0.0152, "step": 7330 }, { "grad_norm": 0.31580352783203125, "learning_rate": 7.495941720423154e-05, "loss": 0.0141, "step": 7340 }, { "grad_norm": 0.32644447684288025, "learning_rate": 7.488774707480042e-05, "loss": 0.0141, "step": 7350 }, { "grad_norm": 0.3435586988925934, "learning_rate": 7.481600890321911e-05, "loss": 0.0207, "step": 7360 }, { "grad_norm": 0.44137218594551086, "learning_rate": 7.474420288561708e-05, "loss": 0.0195, "step": 7370 }, { "grad_norm": 0.38762012124061584, "learning_rate": 7.467232921830921e-05, "loss": 0.0168, "step": 7380 }, { "grad_norm": 0.40698912739753723, "learning_rate": 7.460038809779537e-05, "loss": 0.0175, "step": 7390 }, { "grad_norm": 0.30646368861198425, "learning_rate": 7.452837972075983e-05, "loss": 0.0144, "step": 7400 }, { "grad_norm": 0.37220999598503113, "learning_rate": 7.445630428407074e-05, "loss": 0.0139, "step": 7410 }, { "grad_norm": 0.3270203173160553, "learning_rate": 7.43841619847796e-05, "loss": 0.0142, "step": 7420 }, { "grad_norm": 0.2739071846008301, "learning_rate": 7.431195302012072e-05, "loss": 0.0144, "step": 7430 }, { "grad_norm": 0.387885719537735, "learning_rate": 7.423967758751061e-05, "loss": 0.0143, "step": 7440 }, { "grad_norm": 0.3448900282382965, "learning_rate": 7.416733588454758e-05, "loss": 0.0144, "step": 7450 }, { "grad_norm": 0.289900004863739, "learning_rate": 7.409492810901106e-05, "loss": 0.017, "step": 7460 }, { "grad_norm": 0.36896154284477234, "learning_rate": 7.402245445886116e-05, "loss": 0.0153, "step": 7470 }, { "grad_norm": 0.35660406947135925, "learning_rate": 7.394991513223806e-05, "loss": 0.0134, "step": 7480 }, { "grad_norm": 0.3479897081851959, "learning_rate": 7.38773103274615e-05, "loss": 0.0129, "step": 7490 }, { "grad_norm": 0.2974018454551697, "learning_rate": 7.380464024303028e-05, "loss": 0.0153, "step": 7500 }, { "grad_norm": 0.3199508786201477, "learning_rate": 7.373190507762162e-05, "loss": 0.0127, "step": 7510 }, { "grad_norm": 0.36683499813079834, "learning_rate": 7.365910503009066e-05, "loss": 0.0158, "step": 7520 }, { "grad_norm": 0.33760377764701843, "learning_rate": 7.358624029946996e-05, "loss": 0.0146, "step": 7530 }, { "grad_norm": 0.2612580358982086, "learning_rate": 7.351331108496893e-05, "loss": 0.0137, "step": 7540 }, { "grad_norm": 0.29707035422325134, "learning_rate": 7.344031758597325e-05, "loss": 0.0146, "step": 7550 }, { "grad_norm": 0.39799124002456665, "learning_rate": 7.336726000204435e-05, "loss": 0.0184, "step": 7560 }, { "grad_norm": 0.31683585047721863, "learning_rate": 7.32941385329189e-05, "loss": 0.0148, "step": 7570 }, { "grad_norm": 0.40371984243392944, "learning_rate": 7.322095337850816e-05, "loss": 0.0148, "step": 7580 }, { "grad_norm": 0.33899393677711487, "learning_rate": 7.314770473889758e-05, "loss": 0.0138, "step": 7590 }, { "grad_norm": 0.2862969636917114, "learning_rate": 7.307439281434615e-05, "loss": 0.0162, "step": 7600 }, { "grad_norm": 0.37267830967903137, "learning_rate": 7.300101780528585e-05, "loss": 0.0143, "step": 7610 }, { "grad_norm": 0.370268851518631, "learning_rate": 7.292757991232117e-05, "loss": 0.0146, "step": 7620 }, { "grad_norm": 0.37717458605766296, "learning_rate": 7.285407933622848e-05, "loss": 0.0157, "step": 7630 }, { "grad_norm": 0.39152792096138, "learning_rate": 7.278051627795557e-05, "loss": 0.0149, "step": 7640 }, { "grad_norm": 0.44765567779541016, "learning_rate": 7.270689093862105e-05, "loss": 0.0157, "step": 7650 }, { "grad_norm": 0.3343313932418823, "learning_rate": 7.263320351951374e-05, "loss": 0.0191, "step": 7660 }, { "grad_norm": 0.3631482422351837, "learning_rate": 7.255945422209227e-05, "loss": 0.0157, "step": 7670 }, { "grad_norm": 0.33056262135505676, "learning_rate": 7.248564324798437e-05, "loss": 0.0138, "step": 7680 }, { "grad_norm": 0.32769566774368286, "learning_rate": 7.241177079898644e-05, "loss": 0.0136, "step": 7690 }, { "grad_norm": 0.3817163407802582, "learning_rate": 7.233783707706295e-05, "loss": 0.0176, "step": 7700 }, { "grad_norm": 0.40625038743019104, "learning_rate": 7.226384228434586e-05, "loss": 0.0162, "step": 7710 }, { "grad_norm": 0.27253443002700806, "learning_rate": 7.21897866231341e-05, "loss": 0.0138, "step": 7720 }, { "grad_norm": 0.3633173406124115, "learning_rate": 7.211567029589303e-05, "loss": 0.0159, "step": 7730 }, { "grad_norm": 0.378581166267395, "learning_rate": 7.204149350525387e-05, "loss": 0.0139, "step": 7740 }, { "grad_norm": 0.3962623178958893, "learning_rate": 7.196725645401309e-05, "loss": 0.0153, "step": 7750 }, { "grad_norm": 0.29194292426109314, "learning_rate": 7.1892959345132e-05, "loss": 0.014, "step": 7760 }, { "grad_norm": 0.3924247920513153, "learning_rate": 7.181860238173605e-05, "loss": 0.0195, "step": 7770 }, { "grad_norm": 0.3192293643951416, "learning_rate": 7.174418576711432e-05, "loss": 0.0146, "step": 7780 }, { "grad_norm": 0.30789491534233093, "learning_rate": 7.1669709704719e-05, "loss": 0.0128, "step": 7790 }, { "grad_norm": 0.21125775575637817, "learning_rate": 7.159517439816481e-05, "loss": 0.0138, "step": 7800 }, { "grad_norm": 0.3250451385974884, "learning_rate": 7.152058005122842e-05, "loss": 0.0141, "step": 7810 }, { "grad_norm": 0.35397419333457947, "learning_rate": 7.144592686784793e-05, "loss": 0.0126, "step": 7820 }, { "grad_norm": 0.2984922528266907, "learning_rate": 7.137121505212229e-05, "loss": 0.0164, "step": 7830 }, { "grad_norm": 0.36077097058296204, "learning_rate": 7.129644480831077e-05, "loss": 0.0133, "step": 7840 }, { "grad_norm": 0.3865697979927063, "learning_rate": 7.122161634083234e-05, "loss": 0.0141, "step": 7850 }, { "grad_norm": 0.2705821990966797, "learning_rate": 7.114672985426516e-05, "loss": 0.0153, "step": 7860 }, { "grad_norm": 0.2906481623649597, "learning_rate": 7.107178555334606e-05, "loss": 0.0134, "step": 7870 }, { "grad_norm": 0.3103125989437103, "learning_rate": 7.099678364296989e-05, "loss": 0.0144, "step": 7880 }, { "grad_norm": 0.33912110328674316, "learning_rate": 7.0921724328189e-05, "loss": 0.0156, "step": 7890 }, { "grad_norm": 0.3316667973995209, "learning_rate": 7.084660781421268e-05, "loss": 0.0154, "step": 7900 }, { "grad_norm": 0.30535224080085754, "learning_rate": 7.077143430640662e-05, "loss": 0.0147, "step": 7910 }, { "grad_norm": 0.2779540717601776, "learning_rate": 7.069620401029232e-05, "loss": 0.0134, "step": 7920 }, { "grad_norm": 0.24611879885196686, "learning_rate": 7.062091713154655e-05, "loss": 0.0141, "step": 7930 }, { "grad_norm": 0.37212198972702026, "learning_rate": 7.054557387600075e-05, "loss": 0.016, "step": 7940 }, { "grad_norm": 0.3506477475166321, "learning_rate": 7.04701744496405e-05, "loss": 0.0143, "step": 7950 }, { "grad_norm": 0.307402104139328, "learning_rate": 7.039471905860495e-05, "loss": 0.0163, "step": 7960 }, { "grad_norm": 0.3445793092250824, "learning_rate": 7.031920790918628e-05, "loss": 0.0173, "step": 7970 }, { "grad_norm": 0.33717432618141174, "learning_rate": 7.024364120782906e-05, "loss": 0.0169, "step": 7980 }, { "grad_norm": 0.334218293428421, "learning_rate": 7.016801916112978e-05, "loss": 0.0152, "step": 7990 }, { "grad_norm": 0.3217158615589142, "learning_rate": 7.009234197583623e-05, "loss": 0.0134, "step": 8000 }, { "grad_norm": 0.3041670322418213, "learning_rate": 7.001660985884692e-05, "loss": 0.0121, "step": 8010 }, { "grad_norm": 0.391548216342926, "learning_rate": 6.994082301721063e-05, "loss": 0.0137, "step": 8020 }, { "grad_norm": 0.3116053342819214, "learning_rate": 6.986498165812563e-05, "loss": 0.0128, "step": 8030 }, { "grad_norm": 0.3296670913696289, "learning_rate": 6.978908598893932e-05, "loss": 0.0146, "step": 8040 }, { "grad_norm": 0.2785910665988922, "learning_rate": 6.971313621714756e-05, "loss": 0.0137, "step": 8050 }, { "grad_norm": 0.2963089942932129, "learning_rate": 6.96371325503941e-05, "loss": 0.0138, "step": 8060 }, { "grad_norm": 0.3166458010673523, "learning_rate": 6.956107519647014e-05, "loss": 0.0137, "step": 8070 }, { "grad_norm": 0.2915117144584656, "learning_rate": 6.94849643633135e-05, "loss": 0.0151, "step": 8080 }, { "grad_norm": 0.3173351287841797, "learning_rate": 6.940880025900834e-05, "loss": 0.0152, "step": 8090 }, { "grad_norm": 0.36650222539901733, "learning_rate": 6.933258309178438e-05, "loss": 0.0153, "step": 8100 }, { "grad_norm": 0.241787388920784, "learning_rate": 6.925631307001646e-05, "loss": 0.0129, "step": 8110 }, { "grad_norm": 0.36973029375076294, "learning_rate": 6.91799904022239e-05, "loss": 0.0145, "step": 8120 }, { "grad_norm": 0.3560163974761963, "learning_rate": 6.910361529706997e-05, "loss": 0.0145, "step": 8130 }, { "grad_norm": 0.31722480058670044, "learning_rate": 6.902718796336131e-05, "loss": 0.0129, "step": 8140 }, { "grad_norm": 0.2820875346660614, "learning_rate": 6.895070861004729e-05, "loss": 0.0145, "step": 8150 }, { "grad_norm": 0.2585252523422241, "learning_rate": 6.887417744621956e-05, "loss": 0.014, "step": 8160 }, { "grad_norm": 0.3343718647956848, "learning_rate": 6.87975946811114e-05, "loss": 0.0139, "step": 8170 }, { "grad_norm": 0.2586306929588318, "learning_rate": 6.872096052409718e-05, "loss": 0.0144, "step": 8180 }, { "grad_norm": 0.3091847002506256, "learning_rate": 6.864427518469174e-05, "loss": 0.0147, "step": 8190 }, { "grad_norm": 0.35684236884117126, "learning_rate": 6.856753887254986e-05, "loss": 0.0138, "step": 8200 }, { "grad_norm": 0.3107839524745941, "learning_rate": 6.849075179746572e-05, "loss": 0.0135, "step": 8210 }, { "grad_norm": 0.4236902594566345, "learning_rate": 6.841391416937221e-05, "loss": 0.0121, "step": 8220 }, { "grad_norm": 0.30133530497550964, "learning_rate": 6.833702619834053e-05, "loss": 0.0126, "step": 8230 }, { "grad_norm": 0.2861315608024597, "learning_rate": 6.82600880945794e-05, "loss": 0.0144, "step": 8240 }, { "grad_norm": 0.38043755292892456, "learning_rate": 6.818310006843468e-05, "loss": 0.0161, "step": 8250 }, { "grad_norm": 0.3650735318660736, "learning_rate": 6.810606233038868e-05, "loss": 0.0127, "step": 8260 }, { "grad_norm": 0.31282851099967957, "learning_rate": 6.802897509105966e-05, "loss": 0.0128, "step": 8270 }, { "grad_norm": 0.2905808985233307, "learning_rate": 6.79518385612012e-05, "loss": 0.0135, "step": 8280 }, { "grad_norm": 0.3251124620437622, "learning_rate": 6.787465295170157e-05, "loss": 0.0149, "step": 8290 }, { "grad_norm": 0.35777050256729126, "learning_rate": 6.779741847358332e-05, "loss": 0.013, "step": 8300 }, { "grad_norm": 0.2992149293422699, "learning_rate": 6.772013533800256e-05, "loss": 0.015, "step": 8310 }, { "grad_norm": 0.3211064636707306, "learning_rate": 6.764280375624843e-05, "loss": 0.0127, "step": 8320 }, { "grad_norm": 0.2964476943016052, "learning_rate": 6.756542393974252e-05, "loss": 0.0135, "step": 8330 }, { "grad_norm": 0.3587735891342163, "learning_rate": 6.748799610003828e-05, "loss": 0.0129, "step": 8340 }, { "grad_norm": 0.32010990381240845, "learning_rate": 6.741052044882048e-05, "loss": 0.0139, "step": 8350 }, { "grad_norm": 0.27652424573898315, "learning_rate": 6.73329971979046e-05, "loss": 0.0125, "step": 8360 }, { "grad_norm": 0.32056814432144165, "learning_rate": 6.725542655923625e-05, "loss": 0.0113, "step": 8370 }, { "grad_norm": 0.2358877956867218, "learning_rate": 6.717780874489057e-05, "loss": 0.012, "step": 8380 }, { "grad_norm": 0.34469664096832275, "learning_rate": 6.710014396707172e-05, "loss": 0.0133, "step": 8390 }, { "grad_norm": 0.26492443680763245, "learning_rate": 6.702243243811221e-05, "loss": 0.0135, "step": 8400 }, { "grad_norm": 0.3154253363609314, "learning_rate": 6.694467437047244e-05, "loss": 0.0155, "step": 8410 }, { "grad_norm": 0.31697356700897217, "learning_rate": 6.686686997673997e-05, "loss": 0.0129, "step": 8420 }, { "grad_norm": 0.3026600480079651, "learning_rate": 6.678901946962903e-05, "loss": 0.0129, "step": 8430 }, { "grad_norm": 0.32163363695144653, "learning_rate": 6.671112306197996e-05, "loss": 0.0139, "step": 8440 }, { "grad_norm": 0.2811448574066162, "learning_rate": 6.663318096675854e-05, "loss": 0.0174, "step": 8450 }, { "grad_norm": 0.3415367007255554, "learning_rate": 6.655519339705552e-05, "loss": 0.0126, "step": 8460 }, { "grad_norm": 0.3847247362136841, "learning_rate": 6.647716056608588e-05, "loss": 0.0127, "step": 8470 }, { "grad_norm": 0.36795440316200256, "learning_rate": 6.639908268718843e-05, "loss": 0.0131, "step": 8480 }, { "grad_norm": 0.40341344475746155, "learning_rate": 6.632095997382514e-05, "loss": 0.0137, "step": 8490 }, { "grad_norm": 0.37306714057922363, "learning_rate": 6.624279263958047e-05, "loss": 0.0164, "step": 8500 }, { "grad_norm": 0.42083632946014404, "learning_rate": 6.616458089816097e-05, "loss": 0.0138, "step": 8510 }, { "grad_norm": 0.46660590171813965, "learning_rate": 6.608632496339454e-05, "loss": 0.0134, "step": 8520 }, { "grad_norm": 0.28021371364593506, "learning_rate": 6.600802504922988e-05, "loss": 0.0131, "step": 8530 }, { "grad_norm": 0.3621174395084381, "learning_rate": 6.592968136973604e-05, "loss": 0.0161, "step": 8540 }, { "grad_norm": 0.2531496286392212, "learning_rate": 6.585129413910159e-05, "loss": 0.0122, "step": 8550 }, { "grad_norm": 0.4560607373714447, "learning_rate": 6.577286357163424e-05, "loss": 0.0142, "step": 8560 }, { "grad_norm": 0.28561127185821533, "learning_rate": 6.569438988176018e-05, "loss": 0.018, "step": 8570 }, { "grad_norm": 0.38550880551338196, "learning_rate": 6.561587328402347e-05, "loss": 0.0141, "step": 8580 }, { "grad_norm": 0.4112071096897125, "learning_rate": 6.553731399308549e-05, "loss": 0.0132, "step": 8590 }, { "grad_norm": 0.3395620584487915, "learning_rate": 6.545871222372436e-05, "loss": 0.0126, "step": 8600 }, { "grad_norm": 0.2981310188770294, "learning_rate": 6.538006819083426e-05, "loss": 0.014, "step": 8610 }, { "grad_norm": 0.3370702862739563, "learning_rate": 6.530138210942505e-05, "loss": 0.0127, "step": 8620 }, { "grad_norm": 0.298734188079834, "learning_rate": 6.522265419462141e-05, "loss": 0.0148, "step": 8630 }, { "grad_norm": 0.35672906041145325, "learning_rate": 6.514388466166248e-05, "loss": 0.0144, "step": 8640 }, { "grad_norm": 0.35721224546432495, "learning_rate": 6.506507372590119e-05, "loss": 0.0144, "step": 8650 }, { "grad_norm": 0.3300948143005371, "learning_rate": 6.498622160280355e-05, "loss": 0.0132, "step": 8660 }, { "grad_norm": 0.37679141759872437, "learning_rate": 6.490732850794832e-05, "loss": 0.0144, "step": 8670 }, { "grad_norm": 0.3516535460948944, "learning_rate": 6.482839465702616e-05, "loss": 0.0149, "step": 8680 }, { "grad_norm": 0.2637430727481842, "learning_rate": 6.474942026583923e-05, "loss": 0.0113, "step": 8690 }, { "grad_norm": 0.4087792932987213, "learning_rate": 6.467040555030052e-05, "loss": 0.0113, "step": 8700 }, { "grad_norm": 0.3166436553001404, "learning_rate": 6.459135072643321e-05, "loss": 0.0144, "step": 8710 }, { "grad_norm": 0.2865603566169739, "learning_rate": 6.451225601037019e-05, "loss": 0.0136, "step": 8720 }, { "grad_norm": 0.38709285855293274, "learning_rate": 6.443312161835338e-05, "loss": 0.0133, "step": 8730 }, { "grad_norm": 0.29368987679481506, "learning_rate": 6.43539477667332e-05, "loss": 0.0132, "step": 8740 }, { "grad_norm": 0.18769951164722443, "learning_rate": 6.427473467196793e-05, "loss": 0.0113, "step": 8750 }, { "grad_norm": 0.3281852602958679, "learning_rate": 6.419548255062315e-05, "loss": 0.0113, "step": 8760 }, { "grad_norm": 0.31929275393486023, "learning_rate": 6.411619161937112e-05, "loss": 0.0152, "step": 8770 }, { "grad_norm": 0.329874724149704, "learning_rate": 6.403686209499022e-05, "loss": 0.0151, "step": 8780 }, { "grad_norm": 0.35192087292671204, "learning_rate": 6.395749419436437e-05, "loss": 0.0139, "step": 8790 }, { "grad_norm": 0.2578299939632416, "learning_rate": 6.387808813448234e-05, "loss": 0.011, "step": 8800 }, { "grad_norm": 0.24708716571331024, "learning_rate": 6.37986441324373e-05, "loss": 0.0123, "step": 8810 }, { "grad_norm": 0.30916261672973633, "learning_rate": 6.37191624054261e-05, "loss": 0.0135, "step": 8820 }, { "grad_norm": 0.3506395220756531, "learning_rate": 6.363964317074872e-05, "loss": 0.0125, "step": 8830 }, { "grad_norm": 0.2862837314605713, "learning_rate": 6.356008664580776e-05, "loss": 0.0145, "step": 8840 }, { "grad_norm": 0.30628344416618347, "learning_rate": 6.348049304810771e-05, "loss": 0.0153, "step": 8850 }, { "grad_norm": 0.3329446613788605, "learning_rate": 6.340086259525442e-05, "loss": 0.0128, "step": 8860 }, { "grad_norm": 0.37223020195961, "learning_rate": 6.332119550495448e-05, "loss": 0.0144, "step": 8870 }, { "grad_norm": 0.29098615050315857, "learning_rate": 6.324149199501473e-05, "loss": 0.0125, "step": 8880 }, { "grad_norm": 0.26124271750450134, "learning_rate": 6.316175228334146e-05, "loss": 0.0154, "step": 8890 }, { "grad_norm": 0.25448232889175415, "learning_rate": 6.308197658794003e-05, "loss": 0.0137, "step": 8900 }, { "grad_norm": 0.4169330894947052, "learning_rate": 6.300216512691417e-05, "loss": 0.0135, "step": 8910 }, { "grad_norm": 0.2796359360218048, "learning_rate": 6.292231811846532e-05, "loss": 0.0145, "step": 8920 }, { "grad_norm": 0.31257712841033936, "learning_rate": 6.284243578089217e-05, "loss": 0.0168, "step": 8930 }, { "grad_norm": 0.35514628887176514, "learning_rate": 6.276251833258999e-05, "loss": 0.013, "step": 8940 }, { "grad_norm": 0.27406173944473267, "learning_rate": 6.268256599205003e-05, "loss": 0.0133, "step": 8950 }, { "grad_norm": 0.36449480056762695, "learning_rate": 6.260257897785892e-05, "loss": 0.0143, "step": 8960 }, { "grad_norm": 0.28533563017845154, "learning_rate": 6.252255750869811e-05, "loss": 0.0147, "step": 8970 }, { "grad_norm": 0.3759564757347107, "learning_rate": 6.244250180334325e-05, "loss": 0.0144, "step": 8980 }, { "grad_norm": 0.3371324837207794, "learning_rate": 6.236241208066356e-05, "loss": 0.016, "step": 8990 }, { "grad_norm": 0.33379560708999634, "learning_rate": 6.228228855962133e-05, "loss": 0.0153, "step": 9000 }, { "grad_norm": 0.3548969328403473, "learning_rate": 6.220213145927115e-05, "loss": 0.0154, "step": 9010 }, { "grad_norm": 0.27993297576904297, "learning_rate": 6.212194099875951e-05, "loss": 0.0156, "step": 9020 }, { "grad_norm": 0.2835732400417328, "learning_rate": 6.204171739732405e-05, "loss": 0.0161, "step": 9030 }, { "grad_norm": 0.36675626039505005, "learning_rate": 6.196146087429303e-05, "loss": 0.0126, "step": 9040 }, { "grad_norm": 0.2749966084957123, "learning_rate": 6.188117164908474e-05, "loss": 0.012, "step": 9050 }, { "grad_norm": 0.32398658990859985, "learning_rate": 6.180084994120684e-05, "loss": 0.0135, "step": 9060 }, { "grad_norm": 0.38084620237350464, "learning_rate": 6.17204959702558e-05, "loss": 0.0125, "step": 9070 }, { "grad_norm": 0.3909030258655548, "learning_rate": 6.164010995591635e-05, "loss": 0.0125, "step": 9080 }, { "grad_norm": 0.236577108502388, "learning_rate": 6.155969211796076e-05, "loss": 0.0125, "step": 9090 }, { "grad_norm": 0.27379488945007324, "learning_rate": 6.147924267624829e-05, "loss": 0.0122, "step": 9100 }, { "grad_norm": 0.27393123507499695, "learning_rate": 6.13987618507247e-05, "loss": 0.0105, "step": 9110 }, { "grad_norm": 0.3020084500312805, "learning_rate": 6.131824986142147e-05, "loss": 0.0136, "step": 9120 }, { "grad_norm": 0.2824421525001526, "learning_rate": 6.123770692845529e-05, "loss": 0.0127, "step": 9130 }, { "grad_norm": 0.2659713327884674, "learning_rate": 6.11571332720275e-05, "loss": 0.0116, "step": 9140 }, { "grad_norm": 0.29965081810951233, "learning_rate": 6.107652911242336e-05, "loss": 0.0147, "step": 9150 }, { "grad_norm": 0.33463791012763977, "learning_rate": 6.0995894670011586e-05, "loss": 0.0113, "step": 9160 }, { "grad_norm": 0.3048359453678131, "learning_rate": 6.091523016524368e-05, "loss": 0.0136, "step": 9170 }, { "grad_norm": 0.21435023844242096, "learning_rate": 6.083453581865328e-05, "loss": 0.0123, "step": 9180 }, { "grad_norm": 0.34093180298805237, "learning_rate": 6.075381185085568e-05, "loss": 0.0148, "step": 9190 }, { "grad_norm": 0.3163021504878998, "learning_rate": 6.067305848254709e-05, "loss": 0.0136, "step": 9200 }, { "grad_norm": 0.30066946148872375, "learning_rate": 6.059227593450418e-05, "loss": 0.0141, "step": 9210 }, { "grad_norm": 0.32790908217430115, "learning_rate": 6.051146442758333e-05, "loss": 0.0133, "step": 9220 }, { "grad_norm": 0.33160141110420227, "learning_rate": 6.043062418272012e-05, "loss": 0.0135, "step": 9230 }, { "grad_norm": 0.33114877343177795, "learning_rate": 6.0349755420928666e-05, "loss": 0.0112, "step": 9240 }, { "grad_norm": 0.314362496137619, "learning_rate": 6.0268858363301105e-05, "loss": 0.0134, "step": 9250 }, { "grad_norm": 0.25598523020744324, "learning_rate": 6.018793323100689e-05, "loss": 0.0148, "step": 9260 }, { "grad_norm": 0.2698698341846466, "learning_rate": 6.0106980245292255e-05, "loss": 0.0152, "step": 9270 }, { "grad_norm": 0.40565288066864014, "learning_rate": 6.002599962747957e-05, "loss": 0.0143, "step": 9280 }, { "grad_norm": 0.35545414686203003, "learning_rate": 5.994499159896673e-05, "loss": 0.013, "step": 9290 }, { "grad_norm": 0.3111059069633484, "learning_rate": 5.9863956381226607e-05, "loss": 0.0145, "step": 9300 }, { "grad_norm": 0.30533847212791443, "learning_rate": 5.9782894195806394e-05, "loss": 0.0126, "step": 9310 }, { "grad_norm": 0.2851952910423279, "learning_rate": 5.9701805264327004e-05, "loss": 0.0117, "step": 9320 }, { "grad_norm": 0.2506934106349945, "learning_rate": 5.96206898084825e-05, "loss": 0.0113, "step": 9330 }, { "grad_norm": 0.3302782475948334, "learning_rate": 5.953954805003942e-05, "loss": 0.0163, "step": 9340 }, { "grad_norm": 0.29572638869285583, "learning_rate": 5.945838021083623e-05, "loss": 0.0146, "step": 9350 }, { "grad_norm": 0.3404829204082489, "learning_rate": 5.9377186512782714e-05, "loss": 0.0134, "step": 9360 }, { "grad_norm": 0.2756536602973938, "learning_rate": 5.929596717785935e-05, "loss": 0.0122, "step": 9370 }, { "grad_norm": 0.23491211235523224, "learning_rate": 5.921472242811668e-05, "loss": 0.0106, "step": 9380 }, { "grad_norm": 0.29871824383735657, "learning_rate": 5.913345248567475e-05, "loss": 0.0126, "step": 9390 }, { "grad_norm": 0.2928222119808197, "learning_rate": 5.905215757272248e-05, "loss": 0.0133, "step": 9400 }, { "grad_norm": 0.3399744927883148, "learning_rate": 5.897083791151706e-05, "loss": 0.0131, "step": 9410 }, { "grad_norm": 0.31033551692962646, "learning_rate": 5.888949372438336e-05, "loss": 0.0113, "step": 9420 }, { "grad_norm": 0.40233132243156433, "learning_rate": 5.8808125233713255e-05, "loss": 0.0117, "step": 9430 }, { "grad_norm": 0.4090713858604431, "learning_rate": 5.872673266196509e-05, "loss": 0.0125, "step": 9440 }, { "grad_norm": 0.3115910589694977, "learning_rate": 5.864531623166305e-05, "loss": 0.0122, "step": 9450 }, { "grad_norm": 0.3200939893722534, "learning_rate": 5.856387616539656e-05, "loss": 0.0133, "step": 9460 }, { "grad_norm": 0.2775849997997284, "learning_rate": 5.848241268581967e-05, "loss": 0.0116, "step": 9470 }, { "grad_norm": 0.30292901396751404, "learning_rate": 5.840092601565037e-05, "loss": 0.0114, "step": 9480 }, { "grad_norm": 0.29233402013778687, "learning_rate": 5.8319416377670144e-05, "loss": 0.011, "step": 9490 }, { "grad_norm": 0.2505704462528229, "learning_rate": 5.82378839947232e-05, "loss": 0.0093, "step": 9500 }, { "grad_norm": 0.31264427304267883, "learning_rate": 5.815632908971599e-05, "loss": 0.0122, "step": 9510 }, { "grad_norm": 0.3411703407764435, "learning_rate": 5.80747518856165e-05, "loss": 0.0132, "step": 9520 }, { "grad_norm": 0.2732933759689331, "learning_rate": 5.799315260545367e-05, "loss": 0.0102, "step": 9530 }, { "grad_norm": 0.337495356798172, "learning_rate": 5.791153147231686e-05, "loss": 0.012, "step": 9540 }, { "grad_norm": 0.22131115198135376, "learning_rate": 5.782988870935509e-05, "loss": 0.0116, "step": 9550 }, { "grad_norm": 0.29564183950424194, "learning_rate": 5.774822453977657e-05, "loss": 0.01, "step": 9560 }, { "grad_norm": 0.28418228030204773, "learning_rate": 5.7666539186848036e-05, "loss": 0.0123, "step": 9570 }, { "grad_norm": 0.2695867419242859, "learning_rate": 5.758483287389411e-05, "loss": 0.0129, "step": 9580 }, { "grad_norm": 0.2940092980861664, "learning_rate": 5.7503105824296735e-05, "loss": 0.0128, "step": 9590 }, { "grad_norm": 0.3368409276008606, "learning_rate": 5.742135826149453e-05, "loss": 0.0114, "step": 9600 }, { "grad_norm": 0.36215800046920776, "learning_rate": 5.7339590408982223e-05, "loss": 0.0104, "step": 9610 }, { "grad_norm": 0.2585611939430237, "learning_rate": 5.725780249031e-05, "loss": 0.0121, "step": 9620 }, { "grad_norm": 0.3266613483428955, "learning_rate": 5.717599472908292e-05, "loss": 0.0114, "step": 9630 }, { "grad_norm": 0.31014779210090637, "learning_rate": 5.7094167348960237e-05, "loss": 0.0124, "step": 9640 }, { "grad_norm": 0.3297708332538605, "learning_rate": 5.7012320573654945e-05, "loss": 0.0122, "step": 9650 }, { "grad_norm": 0.2845945358276367, "learning_rate": 5.693045462693295e-05, "loss": 0.01, "step": 9660 }, { "grad_norm": 0.3458256721496582, "learning_rate": 5.684856973261266e-05, "loss": 0.0125, "step": 9670 }, { "grad_norm": 0.2767352759838104, "learning_rate": 5.6766666114564215e-05, "loss": 0.0132, "step": 9680 }, { "grad_norm": 0.25337672233581543, "learning_rate": 5.668474399670899e-05, "loss": 0.0116, "step": 9690 }, { "grad_norm": 0.31617945432662964, "learning_rate": 5.660280360301896e-05, "loss": 0.0131, "step": 9700 }, { "grad_norm": 0.2635818123817444, "learning_rate": 5.652084515751599e-05, "loss": 0.0131, "step": 9710 }, { "grad_norm": 0.29256612062454224, "learning_rate": 5.643886888427137e-05, "loss": 0.012, "step": 9720 }, { "grad_norm": 0.2743358016014099, "learning_rate": 5.6356875007405074e-05, "loss": 0.0109, "step": 9730 }, { "grad_norm": 0.35154983401298523, "learning_rate": 5.627486375108525e-05, "loss": 0.0145, "step": 9740 }, { "grad_norm": 0.3335588276386261, "learning_rate": 5.619283533952754e-05, "loss": 0.0143, "step": 9750 }, { "grad_norm": 0.2750737965106964, "learning_rate": 5.6110789996994474e-05, "loss": 0.0118, "step": 9760 }, { "grad_norm": 0.2875451147556305, "learning_rate": 5.602872794779491e-05, "loss": 0.0122, "step": 9770 }, { "grad_norm": 0.27061742544174194, "learning_rate": 5.594664941628334e-05, "loss": 0.015, "step": 9780 }, { "grad_norm": 0.3677661120891571, "learning_rate": 5.5864554626859324e-05, "loss": 0.0136, "step": 9790 }, { "grad_norm": 0.3061023950576782, "learning_rate": 5.578244380396691e-05, "loss": 0.0111, "step": 9800 }, { "grad_norm": 0.32556185126304626, "learning_rate": 5.570031717209394e-05, "loss": 0.0103, "step": 9810 }, { "grad_norm": 0.32993027567863464, "learning_rate": 5.561817495577147e-05, "loss": 0.0105, "step": 9820 }, { "grad_norm": 0.3360203504562378, "learning_rate": 5.5536017379573215e-05, "loss": 0.0112, "step": 9830 }, { "grad_norm": 0.2788389027118683, "learning_rate": 5.545384466811483e-05, "loss": 0.0123, "step": 9840 }, { "grad_norm": 0.33716967701911926, "learning_rate": 5.5371657046053384e-05, "loss": 0.0149, "step": 9850 }, { "grad_norm": 0.44690340757369995, "learning_rate": 5.528945473808669e-05, "loss": 0.0109, "step": 9860 }, { "grad_norm": 0.334193617105484, "learning_rate": 5.520723796895272e-05, "loss": 0.0131, "step": 9870 }, { "grad_norm": 0.20678123831748962, "learning_rate": 5.512500696342897e-05, "loss": 0.0131, "step": 9880 }, { "grad_norm": 0.3405758738517761, "learning_rate": 5.504276194633188e-05, "loss": 0.0139, "step": 9890 }, { "grad_norm": 0.21453185379505157, "learning_rate": 5.49605031425162e-05, "loss": 0.0129, "step": 9900 }, { "grad_norm": 0.26784011721611023, "learning_rate": 5.487823077687434e-05, "loss": 0.0121, "step": 9910 }, { "grad_norm": 0.22109808027744293, "learning_rate": 5.4795945074335806e-05, "loss": 0.014, "step": 9920 }, { "grad_norm": 0.2858145534992218, "learning_rate": 5.471364625986657e-05, "loss": 0.0149, "step": 9930 }, { "grad_norm": 0.29672735929489136, "learning_rate": 5.463133455846845e-05, "loss": 0.012, "step": 9940 }, { "grad_norm": 0.25313398241996765, "learning_rate": 5.4549010195178505e-05, "loss": 0.0121, "step": 9950 }, { "grad_norm": 0.312531441450119, "learning_rate": 5.446667339506838e-05, "loss": 0.0108, "step": 9960 }, { "grad_norm": 0.27393147349357605, "learning_rate": 5.4384324383243756e-05, "loss": 0.0108, "step": 9970 }, { "grad_norm": 0.28694894909858704, "learning_rate": 5.430196338484368e-05, "loss": 0.0121, "step": 9980 }, { "grad_norm": 0.3046279847621918, "learning_rate": 5.4219590625039975e-05, "loss": 0.0113, "step": 9990 }, { "grad_norm": 0.2781636714935303, "learning_rate": 5.413720632903664e-05, "loss": 0.011, "step": 10000 }, { "grad_norm": 0.27266985177993774, "learning_rate": 5.405481072206917e-05, "loss": 0.0102, "step": 10010 }, { "grad_norm": 0.3278231918811798, "learning_rate": 5.397240402940402e-05, "loss": 0.0112, "step": 10020 }, { "grad_norm": 0.31201794743537903, "learning_rate": 5.388998647633794e-05, "loss": 0.0113, "step": 10030 }, { "grad_norm": 0.21082890033721924, "learning_rate": 5.380755828819737e-05, "loss": 0.0105, "step": 10040 }, { "grad_norm": 0.28620654344558716, "learning_rate": 5.3725119690337846e-05, "loss": 0.0104, "step": 10050 }, { "grad_norm": 0.2828350365161896, "learning_rate": 5.3642670908143324e-05, "loss": 0.0107, "step": 10060 }, { "grad_norm": 0.26399290561676025, "learning_rate": 5.356021216702562e-05, "loss": 0.0112, "step": 10070 }, { "grad_norm": 0.352477490901947, "learning_rate": 5.347774369242381e-05, "loss": 0.0137, "step": 10080 }, { "grad_norm": 0.2641969621181488, "learning_rate": 5.3395265709803545e-05, "loss": 0.0129, "step": 10090 }, { "grad_norm": 0.3277811110019684, "learning_rate": 5.331277844465647e-05, "loss": 0.0136, "step": 10100 }, { "grad_norm": 0.34632810950279236, "learning_rate": 5.323028212249963e-05, "loss": 0.0118, "step": 10110 }, { "grad_norm": 0.35658159852027893, "learning_rate": 5.314777696887481e-05, "loss": 0.0111, "step": 10120 }, { "grad_norm": 0.28721654415130615, "learning_rate": 5.306526320934796e-05, "loss": 0.0105, "step": 10130 }, { "grad_norm": 0.2973387837409973, "learning_rate": 5.298274106950854e-05, "loss": 0.0108, "step": 10140 }, { "grad_norm": 0.25076013803482056, "learning_rate": 5.290021077496893e-05, "loss": 0.0098, "step": 10150 }, { "grad_norm": 0.2708686590194702, "learning_rate": 5.2817672551363816e-05, "loss": 0.0102, "step": 10160 }, { "grad_norm": 0.3272770047187805, "learning_rate": 5.273512662434952e-05, "loss": 0.0138, "step": 10170 }, { "grad_norm": 0.3223007917404175, "learning_rate": 5.265257321960349e-05, "loss": 0.0112, "step": 10180 }, { "grad_norm": 0.34330347180366516, "learning_rate": 5.257001256282357e-05, "loss": 0.01, "step": 10190 }, { "grad_norm": 0.23593580722808838, "learning_rate": 5.248744487972742e-05, "loss": 0.0108, "step": 10200 }, { "grad_norm": 0.21458671987056732, "learning_rate": 5.240487039605196e-05, "loss": 0.0113, "step": 10210 }, { "grad_norm": 0.3227942883968353, "learning_rate": 5.232228933755267e-05, "loss": 0.0123, "step": 10220 }, { "grad_norm": 0.24782851338386536, "learning_rate": 5.2239701930003006e-05, "loss": 0.0104, "step": 10230 }, { "grad_norm": 0.2505333125591278, "learning_rate": 5.215710839919379e-05, "loss": 0.0112, "step": 10240 }, { "grad_norm": 0.30845287442207336, "learning_rate": 5.207450897093257e-05, "loss": 0.0099, "step": 10250 }, { "grad_norm": 0.29096677899360657, "learning_rate": 5.1991903871043046e-05, "loss": 0.0125, "step": 10260 }, { "grad_norm": 0.237869992852211, "learning_rate": 5.190929332536439e-05, "loss": 0.011, "step": 10270 }, { "grad_norm": 0.28825893998146057, "learning_rate": 5.182667755975071e-05, "loss": 0.012, "step": 10280 }, { "grad_norm": 0.3486955761909485, "learning_rate": 5.1744056800070315e-05, "loss": 0.0121, "step": 10290 }, { "grad_norm": 0.3456575572490692, "learning_rate": 5.166143127220524e-05, "loss": 0.0112, "step": 10300 }, { "grad_norm": 0.20626604557037354, "learning_rate": 5.1578801202050485e-05, "loss": 0.0108, "step": 10310 }, { "grad_norm": 0.2725940942764282, "learning_rate": 5.149616681551355e-05, "loss": 0.0092, "step": 10320 }, { "grad_norm": 0.345093697309494, "learning_rate": 5.141352833851367e-05, "loss": 0.0125, "step": 10330 }, { "grad_norm": 0.2966163456439972, "learning_rate": 5.1330885996981285e-05, "loss": 0.0105, "step": 10340 }, { "grad_norm": 0.34058234095573425, "learning_rate": 5.124824001685741e-05, "loss": 0.0111, "step": 10350 }, { "grad_norm": 0.21292267739772797, "learning_rate": 5.116559062409298e-05, "loss": 0.0102, "step": 10360 }, { "grad_norm": 0.2639862298965454, "learning_rate": 5.10829380446483e-05, "loss": 0.0105, "step": 10370 }, { "grad_norm": 0.2518553137779236, "learning_rate": 5.100028250449235e-05, "loss": 0.0091, "step": 10380 }, { "grad_norm": 0.2629271447658539, "learning_rate": 5.0917624229602234e-05, "loss": 0.0126, "step": 10390 }, { "grad_norm": 0.2112114280462265, "learning_rate": 5.0834963445962524e-05, "loss": 0.013, "step": 10400 }, { "grad_norm": 0.27500322461128235, "learning_rate": 5.075230037956461e-05, "loss": 0.0112, "step": 10410 }, { "grad_norm": 0.26655086874961853, "learning_rate": 5.0669635256406213e-05, "loss": 0.0104, "step": 10420 }, { "grad_norm": 0.2902207672595978, "learning_rate": 5.058696830249058e-05, "loss": 0.0097, "step": 10430 }, { "grad_norm": 0.27600282430648804, "learning_rate": 5.050429974382602e-05, "loss": 0.01, "step": 10440 }, { "grad_norm": 0.270409494638443, "learning_rate": 5.042162980642523e-05, "loss": 0.0102, "step": 10450 }, { "grad_norm": 0.24618802964687347, "learning_rate": 5.033895871630462e-05, "loss": 0.0122, "step": 10460 }, { "grad_norm": 0.41161370277404785, "learning_rate": 5.025628669948386e-05, "loss": 0.0116, "step": 10470 }, { "grad_norm": 0.2142290472984314, "learning_rate": 5.017361398198502e-05, "loss": 0.0102, "step": 10480 }, { "grad_norm": 0.3058607578277588, "learning_rate": 5.009094078983221e-05, "loss": 0.0124, "step": 10490 }, { "grad_norm": 0.27186456322669983, "learning_rate": 5.000826734905073e-05, "loss": 0.0155, "step": 10500 }, { "grad_norm": 0.2254273146390915, "learning_rate": 4.9925593885666645e-05, "loss": 0.0089, "step": 10510 }, { "grad_norm": 0.21741361916065216, "learning_rate": 4.984292062570602e-05, "loss": 0.0106, "step": 10520 }, { "grad_norm": 0.3418532907962799, "learning_rate": 4.976024779519442e-05, "loss": 0.0117, "step": 10530 }, { "grad_norm": 0.3184923231601715, "learning_rate": 4.9677575620156194e-05, "loss": 0.0128, "step": 10540 }, { "grad_norm": 0.27130362391471863, "learning_rate": 4.959490432661391e-05, "loss": 0.0107, "step": 10550 }, { "grad_norm": 0.19377490878105164, "learning_rate": 4.9512234140587726e-05, "loss": 0.0101, "step": 10560 }, { "grad_norm": 0.3222799003124237, "learning_rate": 4.942956528809477e-05, "loss": 0.0109, "step": 10570 }, { "grad_norm": 0.21442875266075134, "learning_rate": 4.934689799514854e-05, "loss": 0.0098, "step": 10580 }, { "grad_norm": 0.26878002285957336, "learning_rate": 4.926423248775827e-05, "loss": 0.0105, "step": 10590 }, { "grad_norm": 0.2647032141685486, "learning_rate": 4.918156899192826e-05, "loss": 0.0101, "step": 10600 }, { "grad_norm": 0.29381993412971497, "learning_rate": 4.909890773365738e-05, "loss": 0.0096, "step": 10610 }, { "grad_norm": 0.23105354607105255, "learning_rate": 4.9016248938938344e-05, "loss": 0.0122, "step": 10620 }, { "grad_norm": 0.27361103892326355, "learning_rate": 4.8933592833757156e-05, "loss": 0.0114, "step": 10630 }, { "grad_norm": 0.27087342739105225, "learning_rate": 4.8850939644092435e-05, "loss": 0.0134, "step": 10640 }, { "grad_norm": 0.3496844172477722, "learning_rate": 4.876828959591485e-05, "loss": 0.0124, "step": 10650 }, { "grad_norm": 0.28266310691833496, "learning_rate": 4.8685642915186474e-05, "loss": 0.0091, "step": 10660 }, { "grad_norm": 0.2780430018901825, "learning_rate": 4.860299982786018e-05, "loss": 0.0105, "step": 10670 }, { "grad_norm": 0.2705066204071045, "learning_rate": 4.852036055987901e-05, "loss": 0.012, "step": 10680 }, { "grad_norm": 0.3544326722621918, "learning_rate": 4.843772533717558e-05, "loss": 0.0116, "step": 10690 }, { "grad_norm": 0.3298768699169159, "learning_rate": 4.835509438567142e-05, "loss": 0.011, "step": 10700 }, { "grad_norm": 0.29528871178627014, "learning_rate": 4.827246793127639e-05, "loss": 0.0114, "step": 10710 }, { "grad_norm": 0.32441943883895874, "learning_rate": 4.818984619988807e-05, "loss": 0.0118, "step": 10720 }, { "grad_norm": 0.28215181827545166, "learning_rate": 4.810722941739115e-05, "loss": 0.012, "step": 10730 }, { "grad_norm": 0.26859140396118164, "learning_rate": 4.8024617809656684e-05, "loss": 0.0121, "step": 10740 }, { "grad_norm": 0.34395527839660645, "learning_rate": 4.794201160254171e-05, "loss": 0.011, "step": 10750 }, { "grad_norm": 0.23144233226776123, "learning_rate": 4.785941102188844e-05, "loss": 0.0109, "step": 10760 }, { "grad_norm": 0.21479493379592896, "learning_rate": 4.7776816293523686e-05, "loss": 0.0101, "step": 10770 }, { "grad_norm": 0.27038565278053284, "learning_rate": 4.769422764325832e-05, "loss": 0.0092, "step": 10780 }, { "grad_norm": 0.30681028962135315, "learning_rate": 4.76116452968865e-05, "loss": 0.0109, "step": 10790 }, { "grad_norm": 0.2905797064304352, "learning_rate": 4.752906948018525e-05, "loss": 0.0115, "step": 10800 }, { "grad_norm": 0.32362100481987, "learning_rate": 4.7446500418913684e-05, "loss": 0.0122, "step": 10810 }, { "grad_norm": 0.3390662670135498, "learning_rate": 4.736393833881247e-05, "loss": 0.0112, "step": 10820 }, { "grad_norm": 0.28670960664749146, "learning_rate": 4.7281383465603194e-05, "loss": 0.009, "step": 10830 }, { "grad_norm": 0.2820775508880615, "learning_rate": 4.71988360249877e-05, "loss": 0.0096, "step": 10840 }, { "grad_norm": 0.3001626133918762, "learning_rate": 4.7116296242647554e-05, "loss": 0.0116, "step": 10850 }, { "grad_norm": 0.29720932245254517, "learning_rate": 4.703376434424336e-05, "loss": 0.0101, "step": 10860 }, { "grad_norm": 0.26660263538360596, "learning_rate": 4.695124055541421e-05, "loss": 0.0105, "step": 10870 }, { "grad_norm": 0.2760893702507019, "learning_rate": 4.6868725101776934e-05, "loss": 0.0111, "step": 10880 }, { "grad_norm": 0.2770101726055145, "learning_rate": 4.678621820892567e-05, "loss": 0.0104, "step": 10890 }, { "grad_norm": 0.2585640847682953, "learning_rate": 4.670372010243111e-05, "loss": 0.01, "step": 10900 }, { "grad_norm": 0.26153644919395447, "learning_rate": 4.662123100783992e-05, "loss": 0.0109, "step": 10910 }, { "grad_norm": 0.23419438302516937, "learning_rate": 4.653875115067415e-05, "loss": 0.0105, "step": 10920 }, { "grad_norm": 0.36212146282196045, "learning_rate": 4.6456280756430545e-05, "loss": 0.0113, "step": 10930 }, { "grad_norm": 0.3199692666530609, "learning_rate": 4.637382005058004e-05, "loss": 0.0106, "step": 10940 }, { "grad_norm": 0.32606041431427, "learning_rate": 4.629136925856705e-05, "loss": 0.0102, "step": 10950 }, { "grad_norm": 0.2290044128894806, "learning_rate": 4.6208928605808895e-05, "loss": 0.0098, "step": 10960 }, { "grad_norm": 0.3161752223968506, "learning_rate": 4.612649831769519e-05, "loss": 0.0124, "step": 10970 }, { "grad_norm": 0.22953271865844727, "learning_rate": 4.604407861958715e-05, "loss": 0.0098, "step": 10980 }, { "grad_norm": 0.2739090919494629, "learning_rate": 4.5961669736817114e-05, "loss": 0.0115, "step": 10990 }, { "grad_norm": 0.31810733675956726, "learning_rate": 4.5879271894687814e-05, "loss": 0.0103, "step": 11000 }, { "grad_norm": 0.17931731045246124, "learning_rate": 4.5796885318471826e-05, "loss": 0.0098, "step": 11010 }, { "grad_norm": 0.2645072937011719, "learning_rate": 4.571451023341086e-05, "loss": 0.0101, "step": 11020 }, { "grad_norm": 0.26546722650527954, "learning_rate": 4.563214686471527e-05, "loss": 0.0091, "step": 11030 }, { "grad_norm": 0.28363990783691406, "learning_rate": 4.5549795437563365e-05, "loss": 0.0094, "step": 11040 }, { "grad_norm": 0.2589702904224396, "learning_rate": 4.546745617710081e-05, "loss": 0.0105, "step": 11050 }, { "grad_norm": 0.2975773513317108, "learning_rate": 4.5385129308440014e-05, "loss": 0.0117, "step": 11060 }, { "grad_norm": 0.25610610842704773, "learning_rate": 4.530281505665944e-05, "loss": 0.0092, "step": 11070 }, { "grad_norm": 0.2994934618473053, "learning_rate": 4.5220513646803134e-05, "loss": 0.0107, "step": 11080 }, { "grad_norm": 0.2841961979866028, "learning_rate": 4.513822530388003e-05, "loss": 0.0093, "step": 11090 }, { "grad_norm": 0.25211066007614136, "learning_rate": 4.5055950252863296e-05, "loss": 0.0104, "step": 11100 }, { "grad_norm": 0.3067310154438019, "learning_rate": 4.4973688718689803e-05, "loss": 0.0101, "step": 11110 }, { "grad_norm": 0.3465525805950165, "learning_rate": 4.4891440926259406e-05, "loss": 0.0119, "step": 11120 }, { "grad_norm": 0.28742724657058716, "learning_rate": 4.480920710043443e-05, "loss": 0.0115, "step": 11130 }, { "grad_norm": 0.3307540714740753, "learning_rate": 4.4726987466039044e-05, "loss": 0.0098, "step": 11140 }, { "grad_norm": 0.283673495054245, "learning_rate": 4.46447822478586e-05, "loss": 0.0088, "step": 11150 }, { "grad_norm": 0.24630238115787506, "learning_rate": 4.4562591670638974e-05, "loss": 0.0101, "step": 11160 }, { "grad_norm": 0.31368690729141235, "learning_rate": 4.4480415959086105e-05, "loss": 0.0105, "step": 11170 }, { "grad_norm": 0.3112127482891083, "learning_rate": 4.439825533786522e-05, "loss": 0.0093, "step": 11180 }, { "grad_norm": 0.2516249418258667, "learning_rate": 4.431611003160035e-05, "loss": 0.0118, "step": 11190 }, { "grad_norm": 0.233255535364151, "learning_rate": 4.4233980264873636e-05, "loss": 0.0111, "step": 11200 }, { "grad_norm": 0.23719282448291779, "learning_rate": 4.4151866262224684e-05, "loss": 0.0109, "step": 11210 }, { "grad_norm": 0.26012346148490906, "learning_rate": 4.406976824815006e-05, "loss": 0.0088, "step": 11220 }, { "grad_norm": 0.2692776024341583, "learning_rate": 4.3987686447102595e-05, "loss": 0.0145, "step": 11230 }, { "grad_norm": 0.3251066505908966, "learning_rate": 4.3905621083490804e-05, "loss": 0.0096, "step": 11240 }, { "grad_norm": 0.2645580470561981, "learning_rate": 4.3823572381678286e-05, "loss": 0.0104, "step": 11250 }, { "grad_norm": 0.277410626411438, "learning_rate": 4.374154056598301e-05, "loss": 0.0097, "step": 11260 }, { "grad_norm": 0.2509636878967285, "learning_rate": 4.3659525860676845e-05, "loss": 0.0097, "step": 11270 }, { "grad_norm": 0.2591349184513092, "learning_rate": 4.3577528489984854e-05, "loss": 0.0081, "step": 11280 }, { "grad_norm": 0.25828084349632263, "learning_rate": 4.349554867808476e-05, "loss": 0.0089, "step": 11290 }, { "grad_norm": 0.2765969932079315, "learning_rate": 4.34135866491062e-05, "loss": 0.0085, "step": 11300 }, { "grad_norm": 0.2651883661746979, "learning_rate": 4.333164262713022e-05, "loss": 0.0097, "step": 11310 }, { "grad_norm": 0.23360306024551392, "learning_rate": 4.324971683618868e-05, "loss": 0.0091, "step": 11320 }, { "grad_norm": 0.28962427377700806, "learning_rate": 4.316780950026354e-05, "loss": 0.0096, "step": 11330 }, { "grad_norm": 0.24416466057300568, "learning_rate": 4.308592084328637e-05, "loss": 0.0107, "step": 11340 }, { "grad_norm": 0.33153170347213745, "learning_rate": 4.3004051089137576e-05, "loss": 0.0118, "step": 11350 }, { "grad_norm": 0.2842811942100525, "learning_rate": 4.292220046164597e-05, "loss": 0.0104, "step": 11360 }, { "grad_norm": 0.3026377856731415, "learning_rate": 4.2840369184588035e-05, "loss": 0.0118, "step": 11370 }, { "grad_norm": 0.33202430605888367, "learning_rate": 4.2758557481687345e-05, "loss": 0.0115, "step": 11380 }, { "grad_norm": 0.2605847418308258, "learning_rate": 4.267676557661403e-05, "loss": 0.011, "step": 11390 }, { "grad_norm": 0.32446685433387756, "learning_rate": 4.2594993692983955e-05, "loss": 0.0126, "step": 11400 }, { "grad_norm": 0.2502276599407196, "learning_rate": 4.251324205435837e-05, "loss": 0.0111, "step": 11410 }, { "grad_norm": 0.3160892426967621, "learning_rate": 4.243151088424312e-05, "loss": 0.0095, "step": 11420 }, { "grad_norm": 0.31053662300109863, "learning_rate": 4.234980040608813e-05, "loss": 0.0096, "step": 11430 }, { "grad_norm": 0.2545183300971985, "learning_rate": 4.22681108432867e-05, "loss": 0.0087, "step": 11440 }, { "grad_norm": 0.21548305451869965, "learning_rate": 4.2186442419174984e-05, "loss": 0.0107, "step": 11450 }, { "grad_norm": 0.30552878975868225, "learning_rate": 4.210479535703133e-05, "loss": 0.0108, "step": 11460 }, { "grad_norm": 0.22063080966472626, "learning_rate": 4.202316988007567e-05, "loss": 0.0094, "step": 11470 }, { "grad_norm": 0.27372005581855774, "learning_rate": 4.194156621146901e-05, "loss": 0.0095, "step": 11480 }, { "grad_norm": 0.25082293152809143, "learning_rate": 4.1859984574312596e-05, "loss": 0.0098, "step": 11490 }, { "grad_norm": 0.23153097927570343, "learning_rate": 4.177842519164752e-05, "loss": 0.01, "step": 11500 }, { "grad_norm": 0.24062694609165192, "learning_rate": 4.169688828645404e-05, "loss": 0.0098, "step": 11510 }, { "grad_norm": 0.27024033665657043, "learning_rate": 4.161537408165092e-05, "loss": 0.0097, "step": 11520 }, { "grad_norm": 0.2536773085594177, "learning_rate": 4.1533882800094924e-05, "loss": 0.0123, "step": 11530 }, { "grad_norm": 0.34290581941604614, "learning_rate": 4.145241466458005e-05, "loss": 0.0124, "step": 11540 }, { "grad_norm": 0.3443569540977478, "learning_rate": 4.13709698978371e-05, "loss": 0.0111, "step": 11550 }, { "grad_norm": 0.2333383411169052, "learning_rate": 4.1289548722532944e-05, "loss": 0.0112, "step": 11560 }, { "grad_norm": 0.21718323230743408, "learning_rate": 4.120815136126999e-05, "loss": 0.0096, "step": 11570 }, { "grad_norm": 0.23510585725307465, "learning_rate": 4.112677803658548e-05, "loss": 0.0121, "step": 11580 }, { "grad_norm": 0.23313622176647186, "learning_rate": 4.1045428970951e-05, "loss": 0.0107, "step": 11590 }, { "grad_norm": 0.18950779736042023, "learning_rate": 4.0964104386771785e-05, "loss": 0.0098, "step": 11600 }, { "grad_norm": 0.2325366884469986, "learning_rate": 4.0882804506386144e-05, "loss": 0.0118, "step": 11610 }, { "grad_norm": 0.3247773051261902, "learning_rate": 4.080152955206485e-05, "loss": 0.0145, "step": 11620 }, { "grad_norm": 0.38361653685569763, "learning_rate": 4.0720279746010505e-05, "loss": 0.0136, "step": 11630 }, { "grad_norm": 0.28736957907676697, "learning_rate": 4.063905531035699e-05, "loss": 0.0102, "step": 11640 }, { "grad_norm": 0.27127301692962646, "learning_rate": 4.055785646716882e-05, "loss": 0.0101, "step": 11650 }, { "grad_norm": 0.2827381491661072, "learning_rate": 4.047668343844051e-05, "loss": 0.0093, "step": 11660 }, { "grad_norm": 0.2901775538921356, "learning_rate": 4.039553644609604e-05, "loss": 0.0102, "step": 11670 }, { "grad_norm": 0.24694688618183136, "learning_rate": 4.0314415711988176e-05, "loss": 0.0081, "step": 11680 }, { "grad_norm": 0.19248171150684357, "learning_rate": 4.023332145789792e-05, "loss": 0.0102, "step": 11690 }, { "grad_norm": 0.24262814223766327, "learning_rate": 4.015225390553385e-05, "loss": 0.0109, "step": 11700 }, { "grad_norm": 0.23979149758815765, "learning_rate": 4.007121327653158e-05, "loss": 0.0093, "step": 11710 }, { "grad_norm": 0.328031063079834, "learning_rate": 3.9990199792453064e-05, "loss": 0.0109, "step": 11720 }, { "grad_norm": 0.2189372181892395, "learning_rate": 3.9909213674786103e-05, "loss": 0.0107, "step": 11730 }, { "grad_norm": 0.23874934017658234, "learning_rate": 3.982825514494363e-05, "loss": 0.011, "step": 11740 }, { "grad_norm": 0.21877633035182953, "learning_rate": 3.974732442426319e-05, "loss": 0.0095, "step": 11750 }, { "grad_norm": 0.2521791458129883, "learning_rate": 3.966642173400629e-05, "loss": 0.0093, "step": 11760 }, { "grad_norm": 0.18499670922756195, "learning_rate": 3.9585547295357764e-05, "loss": 0.0096, "step": 11770 }, { "grad_norm": 0.22345219552516937, "learning_rate": 3.950470132942526e-05, "loss": 0.0081, "step": 11780 }, { "grad_norm": 0.27067831158638, "learning_rate": 3.942388405723856e-05, "loss": 0.0098, "step": 11790 }, { "grad_norm": 0.3026236295700073, "learning_rate": 3.9343095699749e-05, "loss": 0.0108, "step": 11800 }, { "grad_norm": 0.26777273416519165, "learning_rate": 3.9262336477828874e-05, "loss": 0.0102, "step": 11810 }, { "grad_norm": 0.25054633617401123, "learning_rate": 3.9181606612270794e-05, "loss": 0.0115, "step": 11820 }, { "grad_norm": 0.24228033423423767, "learning_rate": 3.910090632378713e-05, "loss": 0.0105, "step": 11830 }, { "grad_norm": 0.33287522196769714, "learning_rate": 3.90202358330094e-05, "loss": 0.0107, "step": 11840 }, { "grad_norm": 0.27432605624198914, "learning_rate": 3.8939595360487656e-05, "loss": 0.0094, "step": 11850 }, { "grad_norm": 0.2535651624202728, "learning_rate": 3.885898512668984e-05, "loss": 0.0087, "step": 11860 }, { "grad_norm": 0.23170016705989838, "learning_rate": 3.877840535200127e-05, "loss": 0.0116, "step": 11870 }, { "grad_norm": 0.22862641513347626, "learning_rate": 3.869785625672397e-05, "loss": 0.0106, "step": 11880 }, { "grad_norm": 0.24671798944473267, "learning_rate": 3.8617338061076094e-05, "loss": 0.0105, "step": 11890 }, { "grad_norm": 0.2194424569606781, "learning_rate": 3.853685098519132e-05, "loss": 0.01, "step": 11900 }, { "grad_norm": 0.23390281200408936, "learning_rate": 3.845639524911823e-05, "loss": 0.0099, "step": 11910 }, { "grad_norm": 0.2205561399459839, "learning_rate": 3.837597107281974e-05, "loss": 0.01, "step": 11920 }, { "grad_norm": 0.3232882618904114, "learning_rate": 3.829557867617247e-05, "loss": 0.0121, "step": 11930 }, { "grad_norm": 0.3102942109107971, "learning_rate": 3.821521827896618e-05, "loss": 0.0121, "step": 11940 }, { "grad_norm": 0.2581358253955841, "learning_rate": 3.81348901009031e-05, "loss": 0.0088, "step": 11950 }, { "grad_norm": 0.30217039585113525, "learning_rate": 3.805459436159741e-05, "loss": 0.012, "step": 11960 }, { "grad_norm": 0.2423533797264099, "learning_rate": 3.797433128057461e-05, "loss": 0.0101, "step": 11970 }, { "grad_norm": 0.2862881124019623, "learning_rate": 3.789410107727089e-05, "loss": 0.0093, "step": 11980 }, { "grad_norm": 0.2324875295162201, "learning_rate": 3.781390397103257e-05, "loss": 0.0089, "step": 11990 }, { "grad_norm": 0.24767933785915375, "learning_rate": 3.7733740181115455e-05, "loss": 0.0111, "step": 12000 }, { "grad_norm": 0.2591205835342407, "learning_rate": 3.7653609926684306e-05, "loss": 0.0112, "step": 12010 }, { "grad_norm": 0.2178564965724945, "learning_rate": 3.757351342681217e-05, "loss": 0.0095, "step": 12020 }, { "grad_norm": 0.20686309039592743, "learning_rate": 3.749345090047982e-05, "loss": 0.0099, "step": 12030 }, { "grad_norm": 0.2328433096408844, "learning_rate": 3.741342256657515e-05, "loss": 0.009, "step": 12040 }, { "grad_norm": 0.21766003966331482, "learning_rate": 3.7333428643892567e-05, "loss": 0.0094, "step": 12050 }, { "grad_norm": 0.26858773827552795, "learning_rate": 3.725346935113239e-05, "loss": 0.0101, "step": 12060 }, { "grad_norm": 0.2604144215583801, "learning_rate": 3.717354490690029e-05, "loss": 0.0086, "step": 12070 }, { "grad_norm": 0.2502373158931732, "learning_rate": 3.709365552970664e-05, "loss": 0.0093, "step": 12080 }, { "grad_norm": 0.23591120541095734, "learning_rate": 3.7013801437965945e-05, "loss": 0.0087, "step": 12090 }, { "grad_norm": 0.23928320407867432, "learning_rate": 3.693398284999623e-05, "loss": 0.0092, "step": 12100 }, { "grad_norm": 0.3035443425178528, "learning_rate": 3.6854199984018484e-05, "loss": 0.0093, "step": 12110 }, { "grad_norm": 0.26433151960372925, "learning_rate": 3.677445305815601e-05, "loss": 0.0089, "step": 12120 }, { "grad_norm": 0.2083326280117035, "learning_rate": 3.669474229043387e-05, "loss": 0.0096, "step": 12130 }, { "grad_norm": 0.2654914855957031, "learning_rate": 3.6615067898778235e-05, "loss": 0.0095, "step": 12140 }, { "grad_norm": 0.23549270629882812, "learning_rate": 3.6535430101015866e-05, "loss": 0.0086, "step": 12150 }, { "grad_norm": 0.26526257395744324, "learning_rate": 3.645582911487345e-05, "loss": 0.0101, "step": 12160 }, { "grad_norm": 0.2421717345714569, "learning_rate": 3.637626515797706e-05, "loss": 0.0083, "step": 12170 }, { "grad_norm": 0.27901777625083923, "learning_rate": 3.629673844785152e-05, "loss": 0.0097, "step": 12180 }, { "grad_norm": 0.21836718916893005, "learning_rate": 3.621724920191979e-05, "loss": 0.0078, "step": 12190 }, { "grad_norm": 0.25917720794677734, "learning_rate": 3.6137797637502444e-05, "loss": 0.0108, "step": 12200 }, { "grad_norm": 0.3016909956932068, "learning_rate": 3.6058383971817035e-05, "loss": 0.0112, "step": 12210 }, { "grad_norm": 0.2145087569952011, "learning_rate": 3.59790084219775e-05, "loss": 0.0077, "step": 12220 }, { "grad_norm": 0.21351005136966705, "learning_rate": 3.589967120499353e-05, "loss": 0.0095, "step": 12230 }, { "grad_norm": 0.25227054953575134, "learning_rate": 3.5820372537770075e-05, "loss": 0.0102, "step": 12240 }, { "grad_norm": 0.2252119481563568, "learning_rate": 3.5741112637106655e-05, "loss": 0.0077, "step": 12250 }, { "grad_norm": 0.22010651230812073, "learning_rate": 3.5661891719696804e-05, "loss": 0.0085, "step": 12260 }, { "grad_norm": 0.2561614513397217, "learning_rate": 3.5582710002127504e-05, "loss": 0.0115, "step": 12270 }, { "grad_norm": 0.202015683054924, "learning_rate": 3.550356770087853e-05, "loss": 0.0077, "step": 12280 }, { "grad_norm": 0.23294034600257874, "learning_rate": 3.5424465032321914e-05, "loss": 0.0067, "step": 12290 }, { "grad_norm": 0.15683414041996002, "learning_rate": 3.5345402212721335e-05, "loss": 0.0116, "step": 12300 }, { "grad_norm": 0.26399266719818115, "learning_rate": 3.526637945823152e-05, "loss": 0.01, "step": 12310 }, { "grad_norm": 0.18970103561878204, "learning_rate": 3.518739698489767e-05, "loss": 0.0085, "step": 12320 }, { "grad_norm": 0.21932461857795715, "learning_rate": 3.510845500865485e-05, "loss": 0.0083, "step": 12330 }, { "grad_norm": 0.20167824625968933, "learning_rate": 3.502955374532739e-05, "loss": 0.009, "step": 12340 }, { "grad_norm": 0.2470921128988266, "learning_rate": 3.495069341062836e-05, "loss": 0.0107, "step": 12350 }, { "grad_norm": 0.25685083866119385, "learning_rate": 3.4871874220158896e-05, "loss": 0.0083, "step": 12360 }, { "grad_norm": 0.23612642288208008, "learning_rate": 3.479309638940762e-05, "loss": 0.0082, "step": 12370 }, { "grad_norm": 0.22509972751140594, "learning_rate": 3.4714360133750146e-05, "loss": 0.0082, "step": 12380 }, { "grad_norm": 0.21028248965740204, "learning_rate": 3.463566566844839e-05, "loss": 0.0084, "step": 12390 }, { "grad_norm": 0.13853280246257782, "learning_rate": 3.4557013208650016e-05, "loss": 0.0089, "step": 12400 }, { "grad_norm": 0.22347675263881683, "learning_rate": 3.4478402969387857e-05, "loss": 0.0072, "step": 12410 }, { "grad_norm": 0.20229947566986084, "learning_rate": 3.4399835165579266e-05, "loss": 0.0091, "step": 12420 }, { "grad_norm": 0.17870646715164185, "learning_rate": 3.4321310012025645e-05, "loss": 0.0109, "step": 12430 }, { "grad_norm": 0.24232256412506104, "learning_rate": 3.424282772341176e-05, "loss": 0.0077, "step": 12440 }, { "grad_norm": 0.20649486780166626, "learning_rate": 3.416438851430519e-05, "loss": 0.0081, "step": 12450 }, { "grad_norm": 0.16799978911876678, "learning_rate": 3.408599259915577e-05, "loss": 0.0075, "step": 12460 }, { "grad_norm": 0.21411868929862976, "learning_rate": 3.400764019229487e-05, "loss": 0.01, "step": 12470 }, { "grad_norm": 0.24117551743984222, "learning_rate": 3.3929331507935035e-05, "loss": 0.0085, "step": 12480 }, { "grad_norm": 0.25985953211784363, "learning_rate": 3.3851066760169196e-05, "loss": 0.0119, "step": 12490 }, { "grad_norm": 0.2836676836013794, "learning_rate": 3.377284616297021e-05, "loss": 0.0083, "step": 12500 }, { "grad_norm": 0.27042141556739807, "learning_rate": 3.3694669930190166e-05, "loss": 0.0106, "step": 12510 }, { "grad_norm": 0.21987470984458923, "learning_rate": 3.36165382755599e-05, "loss": 0.0075, "step": 12520 }, { "grad_norm": 0.2589154839515686, "learning_rate": 3.35384514126884e-05, "loss": 0.0099, "step": 12530 }, { "grad_norm": 0.2140827476978302, "learning_rate": 3.3460409555062154e-05, "loss": 0.0076, "step": 12540 }, { "grad_norm": 0.23703579604625702, "learning_rate": 3.3382412916044645e-05, "loss": 0.0092, "step": 12550 }, { "grad_norm": 0.27550041675567627, "learning_rate": 3.330446170887566e-05, "loss": 0.0109, "step": 12560 }, { "grad_norm": 0.28704145550727844, "learning_rate": 3.3226556146670834e-05, "loss": 0.0093, "step": 12570 }, { "grad_norm": 0.2982449531555176, "learning_rate": 3.314869644242102e-05, "loss": 0.0106, "step": 12580 }, { "grad_norm": 0.236547589302063, "learning_rate": 3.3070882808991674e-05, "loss": 0.0081, "step": 12590 }, { "grad_norm": 0.2499636858701706, "learning_rate": 3.2993115459122305e-05, "loss": 0.0087, "step": 12600 }, { "grad_norm": 0.26351433992385864, "learning_rate": 3.2915394605425835e-05, "loss": 0.0066, "step": 12610 }, { "grad_norm": 0.13744564354419708, "learning_rate": 3.283772046038816e-05, "loss": 0.0077, "step": 12620 }, { "grad_norm": 0.18071873486042023, "learning_rate": 3.276009323636739e-05, "loss": 0.0094, "step": 12630 }, { "grad_norm": 0.29874253273010254, "learning_rate": 3.268251314559344e-05, "loss": 0.0104, "step": 12640 }, { "grad_norm": 0.21857863664627075, "learning_rate": 3.2604980400167254e-05, "loss": 0.0081, "step": 12650 }, { "grad_norm": 0.2512739896774292, "learning_rate": 3.252749521206042e-05, "loss": 0.0087, "step": 12660 }, { "grad_norm": 0.23679611086845398, "learning_rate": 3.2450057793114494e-05, "loss": 0.0088, "step": 12670 }, { "grad_norm": 0.23160910606384277, "learning_rate": 3.2372668355040435e-05, "loss": 0.0084, "step": 12680 }, { "grad_norm": 0.25161051750183105, "learning_rate": 3.2295327109418005e-05, "loss": 0.0108, "step": 12690 }, { "grad_norm": 0.22988596558570862, "learning_rate": 3.221803426769518e-05, "loss": 0.0075, "step": 12700 }, { "grad_norm": 0.20260567963123322, "learning_rate": 3.214079004118768e-05, "loss": 0.0087, "step": 12710 }, { "grad_norm": 0.2881484031677246, "learning_rate": 3.2063594641078234e-05, "loss": 0.0114, "step": 12720 }, { "grad_norm": 0.30858102440834045, "learning_rate": 3.198644827841616e-05, "loss": 0.011, "step": 12730 }, { "grad_norm": 0.2536660134792328, "learning_rate": 3.1909351164116654e-05, "loss": 0.0076, "step": 12740 }, { "grad_norm": 0.24544335901737213, "learning_rate": 3.183230350896026e-05, "loss": 0.0097, "step": 12750 }, { "grad_norm": 0.23002316057682037, "learning_rate": 3.1755305523592337e-05, "loss": 0.0116, "step": 12760 }, { "grad_norm": 0.31295904517173767, "learning_rate": 3.167835741852245e-05, "loss": 0.0102, "step": 12770 }, { "grad_norm": 0.3014332056045532, "learning_rate": 3.160145940412378e-05, "loss": 0.0091, "step": 12780 }, { "grad_norm": 0.24293272197246552, "learning_rate": 3.1524611690632545e-05, "loss": 0.0102, "step": 12790 }, { "grad_norm": 0.21639424562454224, "learning_rate": 3.144781448814746e-05, "loss": 0.0098, "step": 12800 }, { "grad_norm": 0.19848895072937012, "learning_rate": 3.1371068006629145e-05, "loss": 0.008, "step": 12810 }, { "grad_norm": 0.22924159467220306, "learning_rate": 3.129437245589956e-05, "loss": 0.0096, "step": 12820 }, { "grad_norm": 0.24061842262744904, "learning_rate": 3.121772804564143e-05, "loss": 0.0085, "step": 12830 }, { "grad_norm": 0.216056227684021, "learning_rate": 3.11411349853976e-05, "loss": 0.0083, "step": 12840 }, { "grad_norm": 0.2402229756116867, "learning_rate": 3.10645934845706e-05, "loss": 0.0104, "step": 12850 }, { "grad_norm": 0.23421300947666168, "learning_rate": 3.098810375242196e-05, "loss": 0.0089, "step": 12860 }, { "grad_norm": 0.18824493885040283, "learning_rate": 3.0911665998071704e-05, "loss": 0.0094, "step": 12870 }, { "grad_norm": 0.23529373109340668, "learning_rate": 3.083528043049774e-05, "loss": 0.0079, "step": 12880 }, { "grad_norm": 0.29571592807769775, "learning_rate": 3.0758947258535255e-05, "loss": 0.0095, "step": 12890 }, { "grad_norm": 0.2197941243648529, "learning_rate": 3.068266669087625e-05, "loss": 0.0073, "step": 12900 }, { "grad_norm": 0.23564009368419647, "learning_rate": 3.060643893606887e-05, "loss": 0.0069, "step": 12910 }, { "grad_norm": 0.2670494318008423, "learning_rate": 3.053026420251693e-05, "loss": 0.0099, "step": 12920 }, { "grad_norm": 0.22617891430854797, "learning_rate": 3.0454142698479183e-05, "loss": 0.0104, "step": 12930 }, { "grad_norm": 0.25962555408477783, "learning_rate": 3.0378074632068954e-05, "loss": 0.008, "step": 12940 }, { "grad_norm": 0.20958614349365234, "learning_rate": 3.0302060211253408e-05, "loss": 0.009, "step": 12950 }, { "grad_norm": 0.1768726110458374, "learning_rate": 3.0226099643853073e-05, "loss": 0.0079, "step": 12960 }, { "grad_norm": 0.2227836698293686, "learning_rate": 3.0150193137541283e-05, "loss": 0.0072, "step": 12970 }, { "grad_norm": 0.20849312841892242, "learning_rate": 3.0074340899843467e-05, "loss": 0.0084, "step": 12980 }, { "grad_norm": 0.22601276636123657, "learning_rate": 2.999854313813677e-05, "loss": 0.0081, "step": 12990 }, { "grad_norm": 0.2885317802429199, "learning_rate": 2.9922800059649382e-05, "loss": 0.0105, "step": 13000 }, { "grad_norm": 0.3363044559955597, "learning_rate": 2.9847111871459976e-05, "loss": 0.0085, "step": 13010 }, { "grad_norm": 0.23627261817455292, "learning_rate": 2.977147878049721e-05, "loss": 0.0092, "step": 13020 }, { "grad_norm": 0.2035563737154007, "learning_rate": 2.9695900993539006e-05, "loss": 0.0091, "step": 13030 }, { "grad_norm": 0.214701309800148, "learning_rate": 2.9620378717212183e-05, "loss": 0.0075, "step": 13040 }, { "grad_norm": 0.2293005734682083, "learning_rate": 2.9544912157991745e-05, "loss": 0.0099, "step": 13050 }, { "grad_norm": 0.2266826182603836, "learning_rate": 2.9469501522200405e-05, "loss": 0.0083, "step": 13060 }, { "grad_norm": 0.17154182493686676, "learning_rate": 2.9394147016007946e-05, "loss": 0.0069, "step": 13070 }, { "grad_norm": 0.23514717817306519, "learning_rate": 2.9318848845430702e-05, "loss": 0.0097, "step": 13080 }, { "grad_norm": 0.30718061327934265, "learning_rate": 2.9243607216331013e-05, "loss": 0.0091, "step": 13090 }, { "grad_norm": 0.34604424238204956, "learning_rate": 2.916842233441661e-05, "loss": 0.0082, "step": 13100 }, { "grad_norm": 0.21868540346622467, "learning_rate": 2.90932944052401e-05, "loss": 0.0087, "step": 13110 }, { "grad_norm": 0.17228583991527557, "learning_rate": 2.9018223634198354e-05, "loss": 0.0082, "step": 13120 }, { "grad_norm": 0.2711525857448578, "learning_rate": 2.8943210226532025e-05, "loss": 0.0084, "step": 13130 }, { "grad_norm": 0.21707332134246826, "learning_rate": 2.8868254387324857e-05, "loss": 0.0074, "step": 13140 }, { "grad_norm": 0.19769783318042755, "learning_rate": 2.8793356321503306e-05, "loss": 0.0084, "step": 13150 }, { "grad_norm": 0.17765411734580994, "learning_rate": 2.87185162338358e-05, "loss": 0.008, "step": 13160 }, { "grad_norm": 0.21457405388355255, "learning_rate": 2.8643734328932253e-05, "loss": 0.0082, "step": 13170 }, { "grad_norm": 0.24327674508094788, "learning_rate": 2.856901081124359e-05, "loss": 0.0084, "step": 13180 }, { "grad_norm": 0.30767297744750977, "learning_rate": 2.8494345885061002e-05, "loss": 0.01, "step": 13190 }, { "grad_norm": 0.24764621257781982, "learning_rate": 2.8419739754515616e-05, "loss": 0.0067, "step": 13200 }, { "grad_norm": 0.25052186846733093, "learning_rate": 2.8345192623577666e-05, "loss": 0.0085, "step": 13210 }, { "grad_norm": 0.22043953835964203, "learning_rate": 2.8270704696056193e-05, "loss": 0.0084, "step": 13220 }, { "grad_norm": 0.1629994958639145, "learning_rate": 2.8196276175598367e-05, "loss": 0.0095, "step": 13230 }, { "grad_norm": 0.20400097966194153, "learning_rate": 2.8121907265688884e-05, "loss": 0.0069, "step": 13240 }, { "grad_norm": 0.19919005036354065, "learning_rate": 2.804759816964957e-05, "loss": 0.0076, "step": 13250 }, { "grad_norm": 0.2581764757633209, "learning_rate": 2.797334909063857e-05, "loss": 0.0098, "step": 13260 }, { "grad_norm": 0.15397116541862488, "learning_rate": 2.7899160231650056e-05, "loss": 0.0086, "step": 13270 }, { "grad_norm": 0.2158154994249344, "learning_rate": 2.7825031795513585e-05, "loss": 0.0081, "step": 13280 }, { "grad_norm": 0.2348286658525467, "learning_rate": 2.775096398489341e-05, "loss": 0.0096, "step": 13290 }, { "grad_norm": 0.21137693524360657, "learning_rate": 2.7676957002288163e-05, "loss": 0.0092, "step": 13300 }, { "grad_norm": 0.23455867171287537, "learning_rate": 2.760301105003003e-05, "loss": 0.0079, "step": 13310 }, { "grad_norm": 0.24594523012638092, "learning_rate": 2.752912633028446e-05, "loss": 0.0106, "step": 13320 }, { "grad_norm": 0.21590381860733032, "learning_rate": 2.7455303045049474e-05, "loss": 0.0091, "step": 13330 }, { "grad_norm": 0.20414042472839355, "learning_rate": 2.7381541396155098e-05, "loss": 0.0083, "step": 13340 }, { "grad_norm": 0.20674766600131989, "learning_rate": 2.730784158526286e-05, "loss": 0.0085, "step": 13350 }, { "grad_norm": 0.29816320538520813, "learning_rate": 2.723420381386521e-05, "loss": 0.0101, "step": 13360 }, { "grad_norm": 0.17939582467079163, "learning_rate": 2.7160628283285018e-05, "loss": 0.0078, "step": 13370 }, { "grad_norm": 0.22700299322605133, "learning_rate": 2.7087115194675007e-05, "loss": 0.0082, "step": 13380 }, { "grad_norm": 0.28414255380630493, "learning_rate": 2.701366474901712e-05, "loss": 0.0077, "step": 13390 }, { "grad_norm": 0.19984044134616852, "learning_rate": 2.6940277147122085e-05, "loss": 0.0092, "step": 13400 }, { "grad_norm": 0.29019051790237427, "learning_rate": 2.686695258962878e-05, "loss": 0.0089, "step": 13410 }, { "grad_norm": 0.3070339262485504, "learning_rate": 2.679369127700375e-05, "loss": 0.0091, "step": 13420 }, { "grad_norm": 0.29468387365341187, "learning_rate": 2.672049340954067e-05, "loss": 0.0106, "step": 13430 }, { "grad_norm": 0.21320025622844696, "learning_rate": 2.6647359187359676e-05, "loss": 0.0095, "step": 13440 }, { "grad_norm": 0.2156619280576706, "learning_rate": 2.6574288810406946e-05, "loss": 0.0078, "step": 13450 }, { "grad_norm": 0.17435118556022644, "learning_rate": 2.6501282478454083e-05, "loss": 0.0083, "step": 13460 }, { "grad_norm": 0.23973551392555237, "learning_rate": 2.6428340391097618e-05, "loss": 0.0099, "step": 13470 }, { "grad_norm": 0.2153310775756836, "learning_rate": 2.6355462747758485e-05, "loss": 0.0087, "step": 13480 }, { "grad_norm": 0.16609278321266174, "learning_rate": 2.6282649747681304e-05, "loss": 0.0073, "step": 13490 }, { "grad_norm": 0.1961102932691574, "learning_rate": 2.620990158993406e-05, "loss": 0.0089, "step": 13500 }, { "grad_norm": 0.2184671014547348, "learning_rate": 2.6137218473407477e-05, "loss": 0.0084, "step": 13510 }, { "grad_norm": 0.2727261781692505, "learning_rate": 2.606460059681436e-05, "loss": 0.0079, "step": 13520 }, { "grad_norm": 0.16959179937839508, "learning_rate": 2.599204815868928e-05, "loss": 0.0072, "step": 13530 }, { "grad_norm": 0.19170914590358734, "learning_rate": 2.5919561357387756e-05, "loss": 0.0087, "step": 13540 }, { "grad_norm": 0.20225897431373596, "learning_rate": 2.5847140391085972e-05, "loss": 0.0104, "step": 13550 }, { "grad_norm": 0.19184699654579163, "learning_rate": 2.5774785457780103e-05, "loss": 0.0082, "step": 13560 }, { "grad_norm": 0.20423159003257751, "learning_rate": 2.5702496755285753e-05, "loss": 0.0088, "step": 13570 }, { "grad_norm": 0.19636958837509155, "learning_rate": 2.5630274481237483e-05, "loss": 0.0085, "step": 13580 }, { "grad_norm": 0.20022667944431305, "learning_rate": 2.5558118833088197e-05, "loss": 0.011, "step": 13590 }, { "grad_norm": 0.22879184782505035, "learning_rate": 2.548603000810872e-05, "loss": 0.009, "step": 13600 }, { "grad_norm": 0.22294972836971283, "learning_rate": 2.5414008203387152e-05, "loss": 0.0073, "step": 13610 }, { "grad_norm": 0.15397511422634125, "learning_rate": 2.534205361582834e-05, "loss": 0.0076, "step": 13620 }, { "grad_norm": 0.2075217366218567, "learning_rate": 2.527016644215338e-05, "loss": 0.0082, "step": 13630 }, { "grad_norm": 0.17194764316082, "learning_rate": 2.519834687889905e-05, "loss": 0.0077, "step": 13640 }, { "grad_norm": 0.224113330245018, "learning_rate": 2.5126595122417295e-05, "loss": 0.0097, "step": 13650 }, { "grad_norm": 0.27634575963020325, "learning_rate": 2.5054911368874713e-05, "loss": 0.0083, "step": 13660 }, { "grad_norm": 0.2299531102180481, "learning_rate": 2.4983295814251916e-05, "loss": 0.008, "step": 13670 }, { "grad_norm": 0.23409384489059448, "learning_rate": 2.4911748654343105e-05, "loss": 0.0096, "step": 13680 }, { "grad_norm": 0.2994837462902069, "learning_rate": 2.4840270084755463e-05, "loss": 0.0085, "step": 13690 }, { "grad_norm": 0.18947498500347137, "learning_rate": 2.4768860300908685e-05, "loss": 0.0101, "step": 13700 }, { "grad_norm": 0.18877004086971283, "learning_rate": 2.469751949803443e-05, "loss": 0.0087, "step": 13710 }, { "grad_norm": 0.18352939188480377, "learning_rate": 2.4626247871175666e-05, "loss": 0.0066, "step": 13720 }, { "grad_norm": 0.2318023443222046, "learning_rate": 2.4555045615186346e-05, "loss": 0.0112, "step": 13730 }, { "grad_norm": 0.2080456018447876, "learning_rate": 2.4483912924730677e-05, "loss": 0.0072, "step": 13740 }, { "grad_norm": 0.2666807472705841, "learning_rate": 2.4412849994282742e-05, "loss": 0.0105, "step": 13750 }, { "grad_norm": 0.2613977789878845, "learning_rate": 2.434185701812592e-05, "loss": 0.009, "step": 13760 }, { "grad_norm": 0.2180042266845703, "learning_rate": 2.4270934190352218e-05, "loss": 0.0067, "step": 13770 }, { "grad_norm": 0.23998405039310455, "learning_rate": 2.4200081704861998e-05, "loss": 0.0067, "step": 13780 }, { "grad_norm": 0.21981388330459595, "learning_rate": 2.412929975536321e-05, "loss": 0.0077, "step": 13790 }, { "grad_norm": 0.1397922933101654, "learning_rate": 2.4058588535371017e-05, "loss": 0.0066, "step": 13800 }, { "grad_norm": 0.18262873589992523, "learning_rate": 2.3987948238207243e-05, "loss": 0.0072, "step": 13810 }, { "grad_norm": 0.22597359120845795, "learning_rate": 2.3917379056999678e-05, "loss": 0.0082, "step": 13820 }, { "grad_norm": 0.18720413744449615, "learning_rate": 2.3846881184681824e-05, "loss": 0.009, "step": 13830 }, { "grad_norm": 0.22752739489078522, "learning_rate": 2.377645481399214e-05, "loss": 0.0083, "step": 13840 }, { "grad_norm": 0.22498559951782227, "learning_rate": 2.3706100137473667e-05, "loss": 0.0068, "step": 13850 }, { "grad_norm": 0.18567043542861938, "learning_rate": 2.3635817347473394e-05, "loss": 0.0064, "step": 13860 }, { "grad_norm": 0.15326108038425446, "learning_rate": 2.3565606636141757e-05, "loss": 0.0097, "step": 13870 }, { "grad_norm": 0.22536373138427734, "learning_rate": 2.3495468195432203e-05, "loss": 0.0078, "step": 13880 }, { "grad_norm": 0.21939432621002197, "learning_rate": 2.3425402217100507e-05, "loss": 0.0081, "step": 13890 }, { "grad_norm": 0.22812145948410034, "learning_rate": 2.3355408892704424e-05, "loss": 0.0065, "step": 13900 }, { "grad_norm": 0.2459571212530136, "learning_rate": 2.3285488413603003e-05, "loss": 0.0094, "step": 13910 }, { "grad_norm": 0.20481348037719727, "learning_rate": 2.321564097095615e-05, "loss": 0.0078, "step": 13920 }, { "grad_norm": 0.187086284160614, "learning_rate": 2.3145866755724142e-05, "loss": 0.0067, "step": 13930 }, { "grad_norm": 0.18507111072540283, "learning_rate": 2.307616595866699e-05, "loss": 0.0072, "step": 13940 }, { "grad_norm": 0.2029513716697693, "learning_rate": 2.3006538770344032e-05, "loss": 0.0074, "step": 13950 }, { "grad_norm": 0.1743609607219696, "learning_rate": 2.293698538111334e-05, "loss": 0.008, "step": 13960 }, { "grad_norm": 0.2573196291923523, "learning_rate": 2.28675059811312e-05, "loss": 0.0096, "step": 13970 }, { "grad_norm": 0.1670626848936081, "learning_rate": 2.279810076035167e-05, "loss": 0.0076, "step": 13980 }, { "grad_norm": 0.24466916918754578, "learning_rate": 2.272876990852596e-05, "loss": 0.0074, "step": 13990 }, { "grad_norm": 0.20826183259487152, "learning_rate": 2.265951361520195e-05, "loss": 0.0062, "step": 14000 }, { "grad_norm": 0.19271427392959595, "learning_rate": 2.2590332069723748e-05, "loss": 0.0083, "step": 14010 }, { "grad_norm": 0.2048574984073639, "learning_rate": 2.2521225461231004e-05, "loss": 0.0083, "step": 14020 }, { "grad_norm": 0.19147121906280518, "learning_rate": 2.2452193978658597e-05, "loss": 0.0074, "step": 14030 }, { "grad_norm": 0.1696682721376419, "learning_rate": 2.238323781073594e-05, "loss": 0.0068, "step": 14040 }, { "grad_norm": 0.15960067510604858, "learning_rate": 2.2314357145986552e-05, "loss": 0.0089, "step": 14050 }, { "grad_norm": 0.19080790877342224, "learning_rate": 2.224555217272757e-05, "loss": 0.007, "step": 14060 }, { "grad_norm": 0.1976376175880432, "learning_rate": 2.2176823079069127e-05, "loss": 0.0054, "step": 14070 }, { "grad_norm": 0.2040068507194519, "learning_rate": 2.210817005291398e-05, "loss": 0.0065, "step": 14080 }, { "grad_norm": 0.17031896114349365, "learning_rate": 2.203959328195686e-05, "loss": 0.0065, "step": 14090 }, { "grad_norm": 0.2634528577327728, "learning_rate": 2.1971092953684026e-05, "loss": 0.0073, "step": 14100 }, { "grad_norm": 0.1954677551984787, "learning_rate": 2.1902669255372788e-05, "loss": 0.0075, "step": 14110 }, { "grad_norm": 0.19398218393325806, "learning_rate": 2.1834322374090897e-05, "loss": 0.0079, "step": 14120 }, { "grad_norm": 0.22374853491783142, "learning_rate": 2.1766052496696153e-05, "loss": 0.0076, "step": 14130 }, { "grad_norm": 0.21147312223911285, "learning_rate": 2.169785980983577e-05, "loss": 0.0072, "step": 14140 }, { "grad_norm": 0.17338190972805023, "learning_rate": 2.162974449994593e-05, "loss": 0.0058, "step": 14150 }, { "grad_norm": 0.21454408764839172, "learning_rate": 2.1561706753251337e-05, "loss": 0.0077, "step": 14160 }, { "grad_norm": 0.22508595883846283, "learning_rate": 2.1493746755764544e-05, "loss": 0.0088, "step": 14170 }, { "grad_norm": 0.1822136789560318, "learning_rate": 2.1425864693285635e-05, "loss": 0.0069, "step": 14180 }, { "grad_norm": 0.23661668598651886, "learning_rate": 2.1358060751401547e-05, "loss": 0.0086, "step": 14190 }, { "grad_norm": 0.19174724817276, "learning_rate": 2.129033511548566e-05, "loss": 0.0075, "step": 14200 }, { "grad_norm": 0.27132779359817505, "learning_rate": 2.1222687970697315e-05, "loss": 0.0088, "step": 14210 }, { "grad_norm": 0.23561926186084747, "learning_rate": 2.1155119501981173e-05, "loss": 0.0083, "step": 14220 }, { "grad_norm": 0.2456965446472168, "learning_rate": 2.1087629894066895e-05, "loss": 0.0085, "step": 14230 }, { "grad_norm": 0.2051001787185669, "learning_rate": 2.1020219331468473e-05, "loss": 0.0072, "step": 14240 }, { "grad_norm": 0.23078320920467377, "learning_rate": 2.095288799848379e-05, "loss": 0.0081, "step": 14250 }, { "grad_norm": 0.23119160532951355, "learning_rate": 2.088563607919417e-05, "loss": 0.0077, "step": 14260 }, { "grad_norm": 0.24694542586803436, "learning_rate": 2.0818463757463786e-05, "loss": 0.0078, "step": 14270 }, { "grad_norm": 0.19064529240131378, "learning_rate": 2.0751371216939175e-05, "loss": 0.0064, "step": 14280 }, { "grad_norm": 0.23178653419017792, "learning_rate": 2.068435864104882e-05, "loss": 0.0067, "step": 14290 }, { "grad_norm": 0.20428496599197388, "learning_rate": 2.0617426213002506e-05, "loss": 0.0076, "step": 14300 }, { "grad_norm": 0.19072294235229492, "learning_rate": 2.055057411579097e-05, "loss": 0.01, "step": 14310 }, { "grad_norm": 0.2197023630142212, "learning_rate": 2.0483802532185286e-05, "loss": 0.0077, "step": 14320 }, { "grad_norm": 0.13945351541042328, "learning_rate": 2.041711164473638e-05, "loss": 0.006, "step": 14330 }, { "grad_norm": 0.2202615588903427, "learning_rate": 2.0350501635774637e-05, "loss": 0.0078, "step": 14340 }, { "grad_norm": 0.1682051718235016, "learning_rate": 2.0283972687409247e-05, "loss": 0.0066, "step": 14350 }, { "grad_norm": 0.21179743111133575, "learning_rate": 2.021752498152784e-05, "loss": 0.007, "step": 14360 }, { "grad_norm": 0.2066740244626999, "learning_rate": 2.015115869979589e-05, "loss": 0.0063, "step": 14370 }, { "grad_norm": 0.1520823836326599, "learning_rate": 2.0084874023656265e-05, "loss": 0.0077, "step": 14380 }, { "grad_norm": 0.22145669162273407, "learning_rate": 2.001867113432877e-05, "loss": 0.0082, "step": 14390 }, { "grad_norm": 0.2755450904369354, "learning_rate": 1.995255021280954e-05, "loss": 0.0067, "step": 14400 }, { "grad_norm": 0.18435436487197876, "learning_rate": 1.9886511439870688e-05, "loss": 0.0058, "step": 14410 }, { "grad_norm": 0.125247523188591, "learning_rate": 1.9820554996059675e-05, "loss": 0.0057, "step": 14420 }, { "grad_norm": 0.15834739804267883, "learning_rate": 1.9754681061698893e-05, "loss": 0.0067, "step": 14430 }, { "grad_norm": 0.2082749903202057, "learning_rate": 1.9688889816885185e-05, "loss": 0.008, "step": 14440 }, { "grad_norm": 0.1739068478345871, "learning_rate": 1.962318144148928e-05, "loss": 0.0065, "step": 14450 }, { "grad_norm": 0.1826048493385315, "learning_rate": 1.955755611515539e-05, "loss": 0.0078, "step": 14460 }, { "grad_norm": 0.15158823132514954, "learning_rate": 1.9492014017300642e-05, "loss": 0.0071, "step": 14470 }, { "grad_norm": 0.20962245762348175, "learning_rate": 1.942655532711461e-05, "loss": 0.0061, "step": 14480 }, { "grad_norm": 0.16392742097377777, "learning_rate": 1.9361180223558882e-05, "loss": 0.0072, "step": 14490 }, { "grad_norm": 0.23526209592819214, "learning_rate": 1.929588888536647e-05, "loss": 0.0084, "step": 14500 }, { "grad_norm": 0.21417514979839325, "learning_rate": 1.9230681491041425e-05, "loss": 0.0079, "step": 14510 }, { "grad_norm": 0.3308451175689697, "learning_rate": 1.9165558218858264e-05, "loss": 0.0068, "step": 14520 }, { "grad_norm": 0.19294549524784088, "learning_rate": 1.9100519246861505e-05, "loss": 0.0072, "step": 14530 }, { "grad_norm": 0.2418549507856369, "learning_rate": 1.9035564752865248e-05, "loss": 0.0068, "step": 14540 }, { "grad_norm": 0.30061328411102295, "learning_rate": 1.897069491445258e-05, "loss": 0.0087, "step": 14550 }, { "grad_norm": 0.22146408259868622, "learning_rate": 1.890590990897515e-05, "loss": 0.0081, "step": 14560 }, { "grad_norm": 0.1591458022594452, "learning_rate": 1.884120991355272e-05, "loss": 0.0061, "step": 14570 }, { "grad_norm": 0.17493392527103424, "learning_rate": 1.8776595105072576e-05, "loss": 0.0065, "step": 14580 }, { "grad_norm": 0.18869565427303314, "learning_rate": 1.8712065660189166e-05, "loss": 0.0075, "step": 14590 }, { "grad_norm": 0.1741727888584137, "learning_rate": 1.8647621755323513e-05, "loss": 0.0073, "step": 14600 }, { "grad_norm": 0.19196191430091858, "learning_rate": 1.858326356666278e-05, "loss": 0.0081, "step": 14610 }, { "grad_norm": 0.1934737116098404, "learning_rate": 1.851899127015983e-05, "loss": 0.0062, "step": 14620 }, { "grad_norm": 0.1917276382446289, "learning_rate": 1.8454805041532626e-05, "loss": 0.0071, "step": 14630 }, { "grad_norm": 0.24973466992378235, "learning_rate": 1.8390705056263906e-05, "loss": 0.0072, "step": 14640 }, { "grad_norm": 0.19900476932525635, "learning_rate": 1.832669148960057e-05, "loss": 0.0068, "step": 14650 }, { "grad_norm": 0.1635986715555191, "learning_rate": 1.8262764516553233e-05, "loss": 0.0066, "step": 14660 }, { "grad_norm": 0.19996638596057892, "learning_rate": 1.8198924311895843e-05, "loss": 0.0075, "step": 14670 }, { "grad_norm": 0.21615362167358398, "learning_rate": 1.813517105016505e-05, "loss": 0.0071, "step": 14680 }, { "grad_norm": 0.2334795892238617, "learning_rate": 1.8071504905659888e-05, "loss": 0.0065, "step": 14690 }, { "grad_norm": 0.2185830920934677, "learning_rate": 1.800792605244109e-05, "loss": 0.0079, "step": 14700 }, { "grad_norm": 0.22458897531032562, "learning_rate": 1.7944434664330844e-05, "loss": 0.0061, "step": 14710 }, { "grad_norm": 0.15321502089500427, "learning_rate": 1.7881030914912212e-05, "loss": 0.0065, "step": 14720 }, { "grad_norm": 0.25509488582611084, "learning_rate": 1.7817714977528577e-05, "loss": 0.0082, "step": 14730 }, { "grad_norm": 0.1694638580083847, "learning_rate": 1.7754487025283332e-05, "loss": 0.0068, "step": 14740 }, { "grad_norm": 0.1656860113143921, "learning_rate": 1.7691347231039275e-05, "loss": 0.006, "step": 14750 }, { "grad_norm": 0.20742027461528778, "learning_rate": 1.7628295767418164e-05, "loss": 0.0089, "step": 14760 }, { "grad_norm": 0.17415648698806763, "learning_rate": 1.7565332806800333e-05, "loss": 0.0061, "step": 14770 }, { "grad_norm": 0.1499238908290863, "learning_rate": 1.750245852132408e-05, "loss": 0.0068, "step": 14780 }, { "grad_norm": 0.1698511689901352, "learning_rate": 1.7439673082885323e-05, "loss": 0.0066, "step": 14790 }, { "grad_norm": 0.1638394147157669, "learning_rate": 1.7376976663137047e-05, "loss": 0.0091, "step": 14800 }, { "grad_norm": 0.1550319641828537, "learning_rate": 1.7314369433488853e-05, "loss": 0.0062, "step": 14810 }, { "grad_norm": 0.15720082819461823, "learning_rate": 1.7251851565106548e-05, "loss": 0.0062, "step": 14820 }, { "grad_norm": 0.17503388226032257, "learning_rate": 1.7189423228911574e-05, "loss": 0.0076, "step": 14830 }, { "grad_norm": 0.17515377700328827, "learning_rate": 1.7127084595580606e-05, "loss": 0.0068, "step": 14840 }, { "grad_norm": 0.16075149178504944, "learning_rate": 1.706483583554513e-05, "loss": 0.0055, "step": 14850 }, { "grad_norm": 0.18190525472164154, "learning_rate": 1.700267711899083e-05, "loss": 0.0073, "step": 14860 }, { "grad_norm": 0.21981409192085266, "learning_rate": 1.69406086158573e-05, "loss": 0.0072, "step": 14870 }, { "grad_norm": 0.21673648059368134, "learning_rate": 1.6878630495837455e-05, "loss": 0.0073, "step": 14880 }, { "grad_norm": 0.19914159178733826, "learning_rate": 1.681674292837707e-05, "loss": 0.0076, "step": 14890 }, { "grad_norm": 0.23613770306110382, "learning_rate": 1.6754946082674444e-05, "loss": 0.0081, "step": 14900 }, { "grad_norm": 0.2050044983625412, "learning_rate": 1.6693240127679748e-05, "loss": 0.0078, "step": 14910 }, { "grad_norm": 0.12192806601524353, "learning_rate": 1.663162523209475e-05, "loss": 0.0076, "step": 14920 }, { "grad_norm": 0.18714375793933868, "learning_rate": 1.6570101564372193e-05, "loss": 0.0083, "step": 14930 }, { "grad_norm": 0.14735259115695953, "learning_rate": 1.650866929271543e-05, "loss": 0.007, "step": 14940 }, { "grad_norm": 0.2103862166404724, "learning_rate": 1.644732858507797e-05, "loss": 0.0067, "step": 14950 }, { "grad_norm": 0.1522349864244461, "learning_rate": 1.6386079609162943e-05, "loss": 0.0059, "step": 14960 }, { "grad_norm": 0.1726202815771103, "learning_rate": 1.6324922532422742e-05, "loss": 0.0069, "step": 14970 }, { "grad_norm": 0.18499019742012024, "learning_rate": 1.6263857522058434e-05, "loss": 0.0072, "step": 14980 }, { "grad_norm": 0.2429019808769226, "learning_rate": 1.6202884745019443e-05, "loss": 0.0065, "step": 14990 }, { "grad_norm": 0.18389727175235748, "learning_rate": 1.614200436800304e-05, "loss": 0.0066, "step": 15000 }, { "grad_norm": 0.1807696670293808, "learning_rate": 1.6081216557453814e-05, "loss": 0.0091, "step": 15010 }, { "grad_norm": 0.18630445003509521, "learning_rate": 1.6020521479563367e-05, "loss": 0.0079, "step": 15020 }, { "grad_norm": 0.20189686119556427, "learning_rate": 1.5959919300269654e-05, "loss": 0.006, "step": 15030 }, { "grad_norm": 0.16778138279914856, "learning_rate": 1.5899410185256764e-05, "loss": 0.0055, "step": 15040 }, { "grad_norm": 0.1611793488264084, "learning_rate": 1.583899429995431e-05, "loss": 0.0081, "step": 15050 }, { "grad_norm": 0.16563250124454498, "learning_rate": 1.5778671809536993e-05, "loss": 0.0072, "step": 15060 }, { "grad_norm": 0.15945349633693695, "learning_rate": 1.5718442878924246e-05, "loss": 0.0084, "step": 15070 }, { "grad_norm": 0.17074576020240784, "learning_rate": 1.5658307672779593e-05, "loss": 0.0092, "step": 15080 }, { "grad_norm": 0.2316267490386963, "learning_rate": 1.5598266355510427e-05, "loss": 0.0091, "step": 15090 }, { "grad_norm": 0.18924427032470703, "learning_rate": 1.553831909126744e-05, "loss": 0.0078, "step": 15100 }, { "grad_norm": 0.1796412467956543, "learning_rate": 1.5478466043944135e-05, "loss": 0.0079, "step": 15110 }, { "grad_norm": 0.24395161867141724, "learning_rate": 1.5418707377176468e-05, "loss": 0.007, "step": 15120 }, { "grad_norm": 0.1799534559249878, "learning_rate": 1.535904325434233e-05, "loss": 0.0055, "step": 15130 }, { "grad_norm": 0.23490634560585022, "learning_rate": 1.529947383856118e-05, "loss": 0.0072, "step": 15140 }, { "grad_norm": 0.17504674196243286, "learning_rate": 1.5239999292693524e-05, "loss": 0.006, "step": 15150 }, { "grad_norm": 0.20216858386993408, "learning_rate": 1.5180619779340505e-05, "loss": 0.0067, "step": 15160 }, { "grad_norm": 0.1365063488483429, "learning_rate": 1.5121335460843428e-05, "loss": 0.007, "step": 15170 }, { "grad_norm": 0.18992237746715546, "learning_rate": 1.5062146499283347e-05, "loss": 0.0058, "step": 15180 }, { "grad_norm": 0.20612695813179016, "learning_rate": 1.5003053056480643e-05, "loss": 0.0087, "step": 15190 }, { "grad_norm": 0.17117105424404144, "learning_rate": 1.4944055293994551e-05, "loss": 0.0071, "step": 15200 }, { "grad_norm": 0.22734136879444122, "learning_rate": 1.4885153373122656e-05, "loss": 0.0062, "step": 15210 }, { "grad_norm": 0.2334449291229248, "learning_rate": 1.482634745490059e-05, "loss": 0.0059, "step": 15220 }, { "grad_norm": 0.23524287343025208, "learning_rate": 1.4767637700101466e-05, "loss": 0.0079, "step": 15230 }, { "grad_norm": 0.19524085521697998, "learning_rate": 1.4709024269235528e-05, "loss": 0.0072, "step": 15240 }, { "grad_norm": 0.21651960909366608, "learning_rate": 1.4650507322549684e-05, "loss": 0.0066, "step": 15250 }, { "grad_norm": 0.18719661235809326, "learning_rate": 1.4592087020026972e-05, "loss": 0.0067, "step": 15260 }, { "grad_norm": 0.20163607597351074, "learning_rate": 1.4533763521386318e-05, "loss": 0.0084, "step": 15270 }, { "grad_norm": 0.20613111555576324, "learning_rate": 1.44755369860819e-05, "loss": 0.0078, "step": 15280 }, { "grad_norm": 0.21348059177398682, "learning_rate": 1.441740757330287e-05, "loss": 0.0081, "step": 15290 }, { "grad_norm": 0.15436752140522003, "learning_rate": 1.4359375441972844e-05, "loss": 0.0086, "step": 15300 }, { "grad_norm": 0.15415067970752716, "learning_rate": 1.4301440750749395e-05, "loss": 0.0062, "step": 15310 }, { "grad_norm": 0.14078067243099213, "learning_rate": 1.4243603658023808e-05, "loss": 0.0044, "step": 15320 }, { "grad_norm": 0.1595512181520462, "learning_rate": 1.4185864321920444e-05, "loss": 0.0071, "step": 15330 }, { "grad_norm": 0.1645655632019043, "learning_rate": 1.4128222900296485e-05, "loss": 0.0059, "step": 15340 }, { "grad_norm": 0.1521766483783722, "learning_rate": 1.407067955074135e-05, "loss": 0.0061, "step": 15350 }, { "grad_norm": 0.1605157107114792, "learning_rate": 1.4013234430576356e-05, "loss": 0.0075, "step": 15360 }, { "grad_norm": 0.1705978959798813, "learning_rate": 1.3955887696854286e-05, "loss": 0.0078, "step": 15370 }, { "grad_norm": 0.20459316670894623, "learning_rate": 1.38986395063589e-05, "loss": 0.0065, "step": 15380 }, { "grad_norm": 0.15137852728366852, "learning_rate": 1.3841490015604597e-05, "loss": 0.0091, "step": 15390 }, { "grad_norm": 0.1907092034816742, "learning_rate": 1.3784439380835879e-05, "loss": 0.0081, "step": 15400 }, { "grad_norm": 0.18163183331489563, "learning_rate": 1.3727487758026986e-05, "loss": 0.005, "step": 15410 }, { "grad_norm": 0.1449543684720993, "learning_rate": 1.3670635302881525e-05, "loss": 0.0093, "step": 15420 }, { "grad_norm": 0.14413036406040192, "learning_rate": 1.3613882170831888e-05, "loss": 0.0062, "step": 15430 }, { "grad_norm": 0.20081110298633575, "learning_rate": 1.355722851703901e-05, "loss": 0.0082, "step": 15440 }, { "grad_norm": 0.17026251554489136, "learning_rate": 1.3500674496391814e-05, "loss": 0.0055, "step": 15450 }, { "grad_norm": 0.12738275527954102, "learning_rate": 1.3444220263506795e-05, "loss": 0.0057, "step": 15460 }, { "grad_norm": 0.17119310796260834, "learning_rate": 1.3387865972727714e-05, "loss": 0.0054, "step": 15470 }, { "grad_norm": 0.20374588668346405, "learning_rate": 1.3331611778125036e-05, "loss": 0.0086, "step": 15480 }, { "grad_norm": 0.14917443692684174, "learning_rate": 1.3275457833495564e-05, "loss": 0.0072, "step": 15490 }, { "grad_norm": 0.14239677786827087, "learning_rate": 1.3219404292362065e-05, "loss": 0.0087, "step": 15500 }, { "grad_norm": 0.18588335812091827, "learning_rate": 1.3163451307972751e-05, "loss": 0.0067, "step": 15510 }, { "grad_norm": 0.21023252606391907, "learning_rate": 1.3107599033300977e-05, "loss": 0.0079, "step": 15520 }, { "grad_norm": 0.14899829030036926, "learning_rate": 1.305184762104471e-05, "loss": 0.0064, "step": 15530 }, { "grad_norm": 0.14481469988822937, "learning_rate": 1.2996197223626178e-05, "loss": 0.0068, "step": 15540 }, { "grad_norm": 0.15361857414245605, "learning_rate": 1.2940647993191457e-05, "loss": 0.0073, "step": 15550 }, { "grad_norm": 0.157449409365654, "learning_rate": 1.2885200081610005e-05, "loss": 0.0087, "step": 15560 }, { "grad_norm": 0.15409617125988007, "learning_rate": 1.2829853640474316e-05, "loss": 0.0074, "step": 15570 }, { "grad_norm": 0.17350949347019196, "learning_rate": 1.2774608821099438e-05, "loss": 0.0079, "step": 15580 }, { "grad_norm": 0.3029111623764038, "learning_rate": 1.2719465774522577e-05, "loss": 0.0055, "step": 15590 }, { "grad_norm": 0.14163382351398468, "learning_rate": 1.2664424651502755e-05, "loss": 0.0067, "step": 15600 }, { "grad_norm": 0.1825382262468338, "learning_rate": 1.260948560252026e-05, "loss": 0.006, "step": 15610 }, { "grad_norm": 0.18032106757164001, "learning_rate": 1.2554648777776396e-05, "loss": 0.0064, "step": 15620 }, { "grad_norm": 0.13015508651733398, "learning_rate": 1.2499914327192919e-05, "loss": 0.0056, "step": 15630 }, { "grad_norm": 0.2085597962141037, "learning_rate": 1.2445282400411722e-05, "loss": 0.0063, "step": 15640 }, { "grad_norm": 0.19599343836307526, "learning_rate": 1.2390753146794437e-05, "loss": 0.0064, "step": 15650 }, { "grad_norm": 0.22105611860752106, "learning_rate": 1.2336326715421925e-05, "loss": 0.0067, "step": 15660 }, { "grad_norm": 0.17722660303115845, "learning_rate": 1.2282003255094005e-05, "loss": 0.0058, "step": 15670 }, { "grad_norm": 0.16634654998779297, "learning_rate": 1.2227782914328928e-05, "loss": 0.0063, "step": 15680 }, { "grad_norm": 0.1348385512828827, "learning_rate": 1.2173665841363018e-05, "loss": 0.0059, "step": 15690 }, { "grad_norm": 0.15744681656360626, "learning_rate": 1.211965218415032e-05, "loss": 0.0057, "step": 15700 }, { "grad_norm": 0.15957064926624298, "learning_rate": 1.2065742090362082e-05, "loss": 0.0077, "step": 15710 }, { "grad_norm": 0.1617710292339325, "learning_rate": 1.2011935707386457e-05, "loss": 0.0056, "step": 15720 }, { "grad_norm": 0.13023534417152405, "learning_rate": 1.1958233182328044e-05, "loss": 0.0069, "step": 15730 }, { "grad_norm": 0.12357404083013535, "learning_rate": 1.1904634662007474e-05, "loss": 0.0062, "step": 15740 }, { "grad_norm": 0.14402985572814941, "learning_rate": 1.1851140292961088e-05, "loss": 0.008, "step": 15750 }, { "grad_norm": 0.16984955966472626, "learning_rate": 1.1797750221440424e-05, "loss": 0.0064, "step": 15760 }, { "grad_norm": 0.20491139590740204, "learning_rate": 1.1744464593411897e-05, "loss": 0.0071, "step": 15770 }, { "grad_norm": 0.195168137550354, "learning_rate": 1.1691283554556399e-05, "loss": 0.0063, "step": 15780 }, { "grad_norm": 0.17448197305202484, "learning_rate": 1.1638207250268834e-05, "loss": 0.0073, "step": 15790 }, { "grad_norm": 0.21772824227809906, "learning_rate": 1.158523582565782e-05, "loss": 0.0076, "step": 15800 }, { "grad_norm": 0.22596101462841034, "learning_rate": 1.1532369425545192e-05, "loss": 0.0064, "step": 15810 }, { "grad_norm": 0.15587890148162842, "learning_rate": 1.1479608194465662e-05, "loss": 0.0067, "step": 15820 }, { "grad_norm": 0.2001703977584839, "learning_rate": 1.1426952276666442e-05, "loss": 0.0068, "step": 15830 }, { "grad_norm": 0.16261820495128632, "learning_rate": 1.1374401816106778e-05, "loss": 0.0062, "step": 15840 }, { "grad_norm": 0.20567862689495087, "learning_rate": 1.1321956956457646e-05, "loss": 0.0067, "step": 15850 }, { "grad_norm": 0.22944943606853485, "learning_rate": 1.1269617841101277e-05, "loss": 0.0081, "step": 15860 }, { "grad_norm": 0.1390569657087326, "learning_rate": 1.1217384613130804e-05, "loss": 0.006, "step": 15870 }, { "grad_norm": 0.20961983501911163, "learning_rate": 1.11652574153499e-05, "loss": 0.0074, "step": 15880 }, { "grad_norm": 0.2552478313446045, "learning_rate": 1.1113236390272303e-05, "loss": 0.0076, "step": 15890 }, { "grad_norm": 0.2062332034111023, "learning_rate": 1.106132168012155e-05, "loss": 0.0071, "step": 15900 }, { "grad_norm": 0.2061680555343628, "learning_rate": 1.1009513426830448e-05, "loss": 0.0057, "step": 15910 }, { "grad_norm": 0.1733689308166504, "learning_rate": 1.0957811772040777e-05, "loss": 0.0063, "step": 15920 }, { "grad_norm": 0.2033698409795761, "learning_rate": 1.0906216857102913e-05, "loss": 0.0074, "step": 15930 }, { "grad_norm": 0.15552908182144165, "learning_rate": 1.0854728823075355e-05, "loss": 0.0059, "step": 15940 }, { "grad_norm": 0.20099680125713348, "learning_rate": 1.0803347810724452e-05, "loss": 0.007, "step": 15950 }, { "grad_norm": 0.19517113268375397, "learning_rate": 1.0752073960523911e-05, "loss": 0.0058, "step": 15960 }, { "grad_norm": 0.1645195186138153, "learning_rate": 1.070090741265447e-05, "loss": 0.0066, "step": 15970 }, { "grad_norm": 0.20157958567142487, "learning_rate": 1.0649848307003547e-05, "loss": 0.007, "step": 15980 }, { "grad_norm": 0.1886875033378601, "learning_rate": 1.0598896783164757e-05, "loss": 0.007, "step": 15990 }, { "grad_norm": 0.1652027815580368, "learning_rate": 1.0548052980437645e-05, "loss": 0.0061, "step": 16000 }, { "grad_norm": 0.13825348019599915, "learning_rate": 1.049731703782722e-05, "loss": 0.0058, "step": 16010 }, { "grad_norm": 0.1408551037311554, "learning_rate": 1.0446689094043587e-05, "loss": 0.0073, "step": 16020 }, { "grad_norm": 0.1619911789894104, "learning_rate": 1.039616928750165e-05, "loss": 0.0058, "step": 16030 }, { "grad_norm": 0.14013567566871643, "learning_rate": 1.0345757756320612e-05, "loss": 0.0064, "step": 16040 }, { "grad_norm": 0.168819859623909, "learning_rate": 1.0295454638323666e-05, "loss": 0.0062, "step": 16050 }, { "grad_norm": 0.16616371273994446, "learning_rate": 1.0245260071037632e-05, "loss": 0.0078, "step": 16060 }, { "grad_norm": 0.13828977942466736, "learning_rate": 1.0195174191692518e-05, "loss": 0.0052, "step": 16070 }, { "grad_norm": 0.1465461105108261, "learning_rate": 1.014519713722124e-05, "loss": 0.0065, "step": 16080 }, { "grad_norm": 0.23143143951892853, "learning_rate": 1.0095329044259132e-05, "loss": 0.0087, "step": 16090 }, { "grad_norm": 0.12432654201984406, "learning_rate": 1.004557004914365e-05, "loss": 0.0043, "step": 16100 }, { "grad_norm": 0.12660931050777435, "learning_rate": 9.995920287914007e-06, "loss": 0.0059, "step": 16110 }, { "grad_norm": 0.15112225711345673, "learning_rate": 9.946379896310737e-06, "loss": 0.0062, "step": 16120 }, { "grad_norm": 0.13577619194984436, "learning_rate": 9.896949009775396e-06, "loss": 0.0066, "step": 16130 }, { "grad_norm": 0.14603663980960846, "learning_rate": 9.847627763450134e-06, "loss": 0.006, "step": 16140 }, { "grad_norm": 0.16121412813663483, "learning_rate": 9.798416292177337e-06, "loss": 0.0065, "step": 16150 }, { "grad_norm": 0.14652404189109802, "learning_rate": 9.74931473049932e-06, "loss": 0.0059, "step": 16160 }, { "grad_norm": 0.15207968652248383, "learning_rate": 9.700323212657847e-06, "loss": 0.0061, "step": 16170 }, { "grad_norm": 0.19286802411079407, "learning_rate": 9.65144187259388e-06, "loss": 0.0058, "step": 16180 }, { "grad_norm": 0.2087399959564209, "learning_rate": 9.602670843947132e-06, "loss": 0.0065, "step": 16190 }, { "grad_norm": 0.13611094653606415, "learning_rate": 9.554010260055713e-06, "loss": 0.0082, "step": 16200 }, { "grad_norm": 0.20372994244098663, "learning_rate": 9.505460253955834e-06, "loss": 0.0064, "step": 16210 }, { "grad_norm": 0.20064783096313477, "learning_rate": 9.457020958381324e-06, "loss": 0.0073, "step": 16220 }, { "grad_norm": 0.0998687669634819, "learning_rate": 9.408692505763395e-06, "loss": 0.0059, "step": 16230 }, { "grad_norm": 0.14523406326770782, "learning_rate": 9.360475028230181e-06, "loss": 0.0049, "step": 16240 }, { "grad_norm": 0.11863675713539124, "learning_rate": 9.312368657606412e-06, "loss": 0.0071, "step": 16250 }, { "grad_norm": 0.1829414963722229, "learning_rate": 9.264373525413096e-06, "loss": 0.0066, "step": 16260 }, { "grad_norm": 0.21367043256759644, "learning_rate": 9.216489762867058e-06, "loss": 0.0072, "step": 16270 }, { "grad_norm": 0.2600301206111908, "learning_rate": 9.168717500880708e-06, "loss": 0.0061, "step": 16280 }, { "grad_norm": 0.24217285215854645, "learning_rate": 9.121056870061574e-06, "loss": 0.0096, "step": 16290 }, { "grad_norm": 0.20866048336029053, "learning_rate": 9.073508000711983e-06, "loss": 0.0077, "step": 16300 }, { "grad_norm": 0.18939869105815887, "learning_rate": 9.026071022828758e-06, "loss": 0.0059, "step": 16310 }, { "grad_norm": 0.14091739058494568, "learning_rate": 8.978746066102771e-06, "loss": 0.0065, "step": 16320 }, { "grad_norm": 0.10054482519626617, "learning_rate": 8.931533259918634e-06, "loss": 0.0069, "step": 16330 }, { "grad_norm": 0.1474171280860901, "learning_rate": 8.884432733354382e-06, "loss": 0.0063, "step": 16340 }, { "grad_norm": 0.15554077923297882, "learning_rate": 8.837444615181029e-06, "loss": 0.005, "step": 16350 }, { "grad_norm": 0.14641925692558289, "learning_rate": 8.790569033862323e-06, "loss": 0.0061, "step": 16360 }, { "grad_norm": 0.1225331500172615, "learning_rate": 8.7438061175543e-06, "loss": 0.0051, "step": 16370 }, { "grad_norm": 0.14073370397090912, "learning_rate": 8.697155994104978e-06, "loss": 0.0055, "step": 16380 }, { "grad_norm": 0.12099365890026093, "learning_rate": 8.650618791054033e-06, "loss": 0.0081, "step": 16390 }, { "grad_norm": 0.1618109792470932, "learning_rate": 8.604194635632373e-06, "loss": 0.0058, "step": 16400 }, { "grad_norm": 0.13691937923431396, "learning_rate": 8.557883654761906e-06, "loss": 0.0057, "step": 16410 }, { "grad_norm": 0.14606285095214844, "learning_rate": 8.511685975055061e-06, "loss": 0.0055, "step": 16420 }, { "grad_norm": 0.16242413222789764, "learning_rate": 8.46560172281452e-06, "loss": 0.007, "step": 16430 }, { "grad_norm": 0.16387271881103516, "learning_rate": 8.419631024032893e-06, "loss": 0.0056, "step": 16440 }, { "grad_norm": 0.12340812385082245, "learning_rate": 8.373774004392293e-06, "loss": 0.0062, "step": 16450 }, { "grad_norm": 0.14751890301704407, "learning_rate": 8.32803078926409e-06, "loss": 0.0051, "step": 16460 }, { "grad_norm": 0.14238260686397552, "learning_rate": 8.282401503708454e-06, "loss": 0.0051, "step": 16470 }, { "grad_norm": 0.13176187872886658, "learning_rate": 8.23688627247412e-06, "loss": 0.0077, "step": 16480 }, { "grad_norm": 0.1804213672876358, "learning_rate": 8.191485219998007e-06, "loss": 0.0073, "step": 16490 }, { "grad_norm": 0.16777563095092773, "learning_rate": 8.146198470404843e-06, "loss": 0.0049, "step": 16500 }, { "grad_norm": 0.14023834466934204, "learning_rate": 8.101026147506897e-06, "loss": 0.0059, "step": 16510 }, { "grad_norm": 0.15105780959129333, "learning_rate": 8.05596837480353e-06, "loss": 0.0054, "step": 16520 }, { "grad_norm": 0.174442857503891, "learning_rate": 8.011025275480998e-06, "loss": 0.0059, "step": 16530 }, { "grad_norm": 0.12707863748073578, "learning_rate": 7.966196972412027e-06, "loss": 0.005, "step": 16540 }, { "grad_norm": 0.1012113019824028, "learning_rate": 7.92148358815547e-06, "loss": 0.0057, "step": 16550 }, { "grad_norm": 0.12641534209251404, "learning_rate": 7.87688524495604e-06, "loss": 0.0053, "step": 16560 }, { "grad_norm": 0.15847855806350708, "learning_rate": 7.83240206474386e-06, "loss": 0.0057, "step": 16570 }, { "grad_norm": 0.16235165297985077, "learning_rate": 7.788034169134272e-06, "loss": 0.007, "step": 16580 }, { "grad_norm": 0.14874836802482605, "learning_rate": 7.743781679427414e-06, "loss": 0.0059, "step": 16590 }, { "grad_norm": 0.1387220025062561, "learning_rate": 7.699644716607895e-06, "loss": 0.0082, "step": 16600 }, { "grad_norm": 0.14721742272377014, "learning_rate": 7.655623401344486e-06, "loss": 0.0057, "step": 16610 }, { "grad_norm": 0.2262549251317978, "learning_rate": 7.611717853989775e-06, "loss": 0.0054, "step": 16620 }, { "grad_norm": 0.11670374125242233, "learning_rate": 7.567928194579854e-06, "loss": 0.0057, "step": 16630 }, { "grad_norm": 0.15404009819030762, "learning_rate": 7.524254542833997e-06, "loss": 0.0077, "step": 16640 }, { "grad_norm": 0.2097034752368927, "learning_rate": 7.480697018154286e-06, "loss": 0.0055, "step": 16650 }, { "grad_norm": 0.2310960739850998, "learning_rate": 7.437255739625332e-06, "loss": 0.006, "step": 16660 }, { "grad_norm": 0.13493993878364563, "learning_rate": 7.393930826013923e-06, "loss": 0.0065, "step": 16670 }, { "grad_norm": 0.11167895048856735, "learning_rate": 7.350722395768722e-06, "loss": 0.0066, "step": 16680 }, { "grad_norm": 0.1722927987575531, "learning_rate": 7.307630567019963e-06, "loss": 0.0095, "step": 16690 }, { "grad_norm": 0.11166276037693024, "learning_rate": 7.264655457579e-06, "loss": 0.0062, "step": 16700 }, { "grad_norm": 0.149301216006279, "learning_rate": 7.221797184938184e-06, "loss": 0.005, "step": 16710 }, { "grad_norm": 0.13837826251983643, "learning_rate": 7.179055866270373e-06, "loss": 0.0062, "step": 16720 }, { "grad_norm": 0.1699582040309906, "learning_rate": 7.136431618428707e-06, "loss": 0.0052, "step": 16730 }, { "grad_norm": 0.10907237231731415, "learning_rate": 7.09392455794628e-06, "loss": 0.0053, "step": 16740 }, { "grad_norm": 0.16739453375339508, "learning_rate": 7.051534801035725e-06, "loss": 0.0079, "step": 16750 }, { "grad_norm": 0.11083260923624039, "learning_rate": 7.00926246358905e-06, "loss": 0.0057, "step": 16760 }, { "grad_norm": 0.10749899595975876, "learning_rate": 6.967107661177191e-06, "loss": 0.0059, "step": 16770 }, { "grad_norm": 0.12547160685062408, "learning_rate": 6.925070509049786e-06, "loss": 0.0055, "step": 16780 }, { "grad_norm": 0.11779487133026123, "learning_rate": 6.883151122134812e-06, "loss": 0.006, "step": 16790 }, { "grad_norm": 0.16693837940692902, "learning_rate": 6.8413496150382394e-06, "loss": 0.007, "step": 16800 }, { "grad_norm": 0.12908636033535004, "learning_rate": 6.7996661020438165e-06, "loss": 0.0073, "step": 16810 }, { "grad_norm": 0.12982450425624847, "learning_rate": 6.758100697112662e-06, "loss": 0.0071, "step": 16820 }, { "grad_norm": 0.1596035659313202, "learning_rate": 6.716653513883026e-06, "loss": 0.0067, "step": 16830 }, { "grad_norm": 0.17830030620098114, "learning_rate": 6.675324665669913e-06, "loss": 0.0052, "step": 16840 }, { "grad_norm": 0.14436714351177216, "learning_rate": 6.634114265464803e-06, "loss": 0.0056, "step": 16850 }, { "grad_norm": 0.11855025589466095, "learning_rate": 6.59302242593538e-06, "loss": 0.0052, "step": 16860 }, { "grad_norm": 0.16071195900440216, "learning_rate": 6.552049259425141e-06, "loss": 0.0067, "step": 16870 }, { "grad_norm": 0.10466548055410385, "learning_rate": 6.511194877953181e-06, "loss": 0.0048, "step": 16880 }, { "grad_norm": 0.11581787467002869, "learning_rate": 6.470459393213813e-06, "loss": 0.0056, "step": 16890 }, { "grad_norm": 0.15448977053165436, "learning_rate": 6.429842916576279e-06, "loss": 0.0073, "step": 16900 }, { "grad_norm": 0.12454946339130402, "learning_rate": 6.389345559084503e-06, "loss": 0.005, "step": 16910 }, { "grad_norm": 0.17361870408058167, "learning_rate": 6.348967431456682e-06, "loss": 0.0055, "step": 16920 }, { "grad_norm": 0.11961667239665985, "learning_rate": 6.30870864408511e-06, "loss": 0.0056, "step": 16930 }, { "grad_norm": 0.1253196895122528, "learning_rate": 6.268569307035754e-06, "loss": 0.0049, "step": 16940 }, { "grad_norm": 0.1187659502029419, "learning_rate": 6.228549530048022e-06, "loss": 0.0053, "step": 16950 }, { "grad_norm": 0.14393658936023712, "learning_rate": 6.1886494225344814e-06, "loss": 0.0056, "step": 16960 }, { "grad_norm": 0.11983685195446014, "learning_rate": 6.148869093580479e-06, "loss": 0.0042, "step": 16970 }, { "grad_norm": 0.17619433999061584, "learning_rate": 6.109208651943921e-06, "loss": 0.0064, "step": 16980 }, { "grad_norm": 0.14167249202728271, "learning_rate": 6.069668206054946e-06, "loss": 0.0077, "step": 16990 }, { "grad_norm": 0.19161079823970795, "learning_rate": 6.0302478640156145e-06, "loss": 0.0076, "step": 17000 }, { "grad_norm": 0.16364793479442596, "learning_rate": 5.990947733599644e-06, "loss": 0.0063, "step": 17010 }, { "grad_norm": 0.1505390703678131, "learning_rate": 5.951767922252105e-06, "loss": 0.0069, "step": 17020 }, { "grad_norm": 0.14058342576026917, "learning_rate": 5.912708537089068e-06, "loss": 0.005, "step": 17030 }, { "grad_norm": 0.138787180185318, "learning_rate": 5.873769684897434e-06, "loss": 0.0062, "step": 17040 }, { "grad_norm": 0.14517366886138916, "learning_rate": 5.834951472134514e-06, "loss": 0.0048, "step": 17050 }, { "grad_norm": 0.1447465419769287, "learning_rate": 5.796254004927832e-06, "loss": 0.0069, "step": 17060 }, { "grad_norm": 0.14510542154312134, "learning_rate": 5.757677389074806e-06, "loss": 0.0053, "step": 17070 }, { "grad_norm": 0.13588163256645203, "learning_rate": 5.719221730042385e-06, "loss": 0.006, "step": 17080 }, { "grad_norm": 0.1612691432237625, "learning_rate": 5.680887132966911e-06, "loss": 0.0059, "step": 17090 }, { "grad_norm": 0.1703179031610489, "learning_rate": 5.642673702653683e-06, "loss": 0.0059, "step": 17100 }, { "grad_norm": 0.15151354670524597, "learning_rate": 5.604581543576781e-06, "loss": 0.0068, "step": 17110 }, { "grad_norm": 0.15857680141925812, "learning_rate": 5.566610759878704e-06, "loss": 0.0049, "step": 17120 }, { "grad_norm": 0.12346552312374115, "learning_rate": 5.528761455370119e-06, "loss": 0.0056, "step": 17130 }, { "grad_norm": 0.12027148902416229, "learning_rate": 5.491033733529594e-06, "loss": 0.0064, "step": 17140 }, { "grad_norm": 0.16653959453105927, "learning_rate": 5.453427697503255e-06, "loss": 0.0052, "step": 17150 }, { "grad_norm": 0.1118609607219696, "learning_rate": 5.415943450104599e-06, "loss": 0.0067, "step": 17160 }, { "grad_norm": 0.1377202272415161, "learning_rate": 5.378581093814111e-06, "loss": 0.0052, "step": 17170 }, { "grad_norm": 0.12130595743656158, "learning_rate": 5.3413407307790375e-06, "loss": 0.0065, "step": 17180 }, { "grad_norm": 0.16329962015151978, "learning_rate": 5.30422246281313e-06, "loss": 0.0057, "step": 17190 }, { "grad_norm": 0.1540021002292633, "learning_rate": 5.267226391396296e-06, "loss": 0.0051, "step": 17200 }, { "grad_norm": 0.11072094738483429, "learning_rate": 5.2303526176744e-06, "loss": 0.0057, "step": 17210 }, { "grad_norm": 0.1866774559020996, "learning_rate": 5.193601242458929e-06, "loss": 0.0073, "step": 17220 }, { "grad_norm": 0.16121235489845276, "learning_rate": 5.156972366226714e-06, "loss": 0.0059, "step": 17230 }, { "grad_norm": 0.18929295241832733, "learning_rate": 5.120466089119735e-06, "loss": 0.005, "step": 17240 }, { "grad_norm": 0.13211718201637268, "learning_rate": 5.084082510944749e-06, "loss": 0.0058, "step": 17250 }, { "grad_norm": 0.13629141449928284, "learning_rate": 5.047821731173058e-06, "loss": 0.0056, "step": 17260 }, { "grad_norm": 0.10753912478685379, "learning_rate": 5.011683848940274e-06, "loss": 0.0046, "step": 17270 }, { "grad_norm": 0.10351518541574478, "learning_rate": 4.975668963045954e-06, "loss": 0.0057, "step": 17280 }, { "grad_norm": 0.12005636096000671, "learning_rate": 4.9397771719534525e-06, "loss": 0.0047, "step": 17290 }, { "grad_norm": 0.12120399624109268, "learning_rate": 4.904008573789548e-06, "loss": 0.0072, "step": 17300 }, { "grad_norm": 0.13392853736877441, "learning_rate": 4.8683632663442005e-06, "loss": 0.0072, "step": 17310 }, { "grad_norm": 0.13340626657009125, "learning_rate": 4.832841347070343e-06, "loss": 0.0056, "step": 17320 }, { "grad_norm": 0.13785862922668457, "learning_rate": 4.797442913083539e-06, "loss": 0.006, "step": 17330 }, { "grad_norm": 0.11659137159585953, "learning_rate": 4.7621680611617596e-06, "loss": 0.0046, "step": 17340 }, { "grad_norm": 0.12524235248565674, "learning_rate": 4.727016887745095e-06, "loss": 0.0053, "step": 17350 }, { "grad_norm": 0.15582780539989471, "learning_rate": 4.691989488935511e-06, "loss": 0.0052, "step": 17360 }, { "grad_norm": 0.1551336795091629, "learning_rate": 4.657085960496588e-06, "loss": 0.0059, "step": 17370 }, { "grad_norm": 0.16865605115890503, "learning_rate": 4.6223063978532265e-06, "loss": 0.0055, "step": 17380 }, { "grad_norm": 0.14389252662658691, "learning_rate": 4.587650896091439e-06, "loss": 0.0062, "step": 17390 }, { "grad_norm": 0.13224855065345764, "learning_rate": 4.553119549958035e-06, "loss": 0.0088, "step": 17400 }, { "grad_norm": 0.14510995149612427, "learning_rate": 4.518712453860385e-06, "loss": 0.0051, "step": 17410 }, { "grad_norm": 0.12474753707647324, "learning_rate": 4.484429701866205e-06, "loss": 0.0052, "step": 17420 }, { "grad_norm": 0.14201487600803375, "learning_rate": 4.4502713877031975e-06, "loss": 0.0057, "step": 17430 }, { "grad_norm": 0.10890962928533554, "learning_rate": 4.416237604758911e-06, "loss": 0.0041, "step": 17440 }, { "grad_norm": 0.10517103970050812, "learning_rate": 4.3823284460804025e-06, "loss": 0.0046, "step": 17450 }, { "grad_norm": 0.17719407379627228, "learning_rate": 4.348544004374011e-06, "loss": 0.0068, "step": 17460 }, { "grad_norm": 0.1262749880552292, "learning_rate": 4.314884372005123e-06, "loss": 0.0053, "step": 17470 }, { "grad_norm": 0.15107092261314392, "learning_rate": 4.281349640997867e-06, "loss": 0.0053, "step": 17480 }, { "grad_norm": 0.11313611268997192, "learning_rate": 4.247939903034942e-06, "loss": 0.0066, "step": 17490 }, { "grad_norm": 0.12144047021865845, "learning_rate": 4.214655249457284e-06, "loss": 0.005, "step": 17500 }, { "grad_norm": 0.13662070035934448, "learning_rate": 4.181495771263855e-06, "loss": 0.0049, "step": 17510 }, { "grad_norm": 0.11010880023241043, "learning_rate": 4.148461559111427e-06, "loss": 0.0066, "step": 17520 }, { "grad_norm": 0.10328162461519241, "learning_rate": 4.115552703314252e-06, "loss": 0.0066, "step": 17530 }, { "grad_norm": 0.15623381733894348, "learning_rate": 4.082769293843886e-06, "loss": 0.006, "step": 17540 }, { "grad_norm": 0.10762166231870651, "learning_rate": 4.050111420328939e-06, "loss": 0.0053, "step": 17550 }, { "grad_norm": 0.07675281167030334, "learning_rate": 4.017579172054764e-06, "loss": 0.0061, "step": 17560 }, { "grad_norm": 0.19645680487155914, "learning_rate": 3.985172637963308e-06, "loss": 0.0073, "step": 17570 }, { "grad_norm": 0.12864311039447784, "learning_rate": 3.952891906652784e-06, "loss": 0.0058, "step": 17580 }, { "grad_norm": 0.1760128289461136, "learning_rate": 3.920737066377478e-06, "loss": 0.0061, "step": 17590 }, { "grad_norm": 0.12532979249954224, "learning_rate": 3.888708205047509e-06, "loss": 0.0063, "step": 17600 }, { "grad_norm": 0.10734823346138, "learning_rate": 3.856805410228542e-06, "loss": 0.0049, "step": 17610 }, { "grad_norm": 0.1140839084982872, "learning_rate": 3.82502876914162e-06, "loss": 0.0039, "step": 17620 }, { "grad_norm": 0.13573074340820312, "learning_rate": 3.7933783686628586e-06, "loss": 0.0067, "step": 17630 }, { "grad_norm": 0.10941028594970703, "learning_rate": 3.7618542953232306e-06, "loss": 0.0059, "step": 17640 }, { "grad_norm": 0.11974181979894638, "learning_rate": 3.7304566353083658e-06, "loss": 0.0061, "step": 17650 }, { "grad_norm": 0.14694556593894958, "learning_rate": 3.6991854744582555e-06, "loss": 0.0063, "step": 17660 }, { "grad_norm": 0.11917652189731598, "learning_rate": 3.6680408982670777e-06, "loss": 0.0047, "step": 17670 }, { "grad_norm": 0.11840870976448059, "learning_rate": 3.637022991882899e-06, "loss": 0.0048, "step": 17680 }, { "grad_norm": 0.21175576746463776, "learning_rate": 3.606131840107485e-06, "loss": 0.0081, "step": 17690 }, { "grad_norm": 0.1199529618024826, "learning_rate": 3.575367527396084e-06, "loss": 0.005, "step": 17700 }, { "grad_norm": 0.09004362672567368, "learning_rate": 3.5447301378571386e-06, "loss": 0.0046, "step": 17710 }, { "grad_norm": 0.11391471326351166, "learning_rate": 3.514219755252113e-06, "loss": 0.0045, "step": 17720 }, { "grad_norm": 0.12353096157312393, "learning_rate": 3.4838364629952213e-06, "loss": 0.0051, "step": 17730 }, { "grad_norm": 0.09329228848218918, "learning_rate": 3.4535803441532123e-06, "loss": 0.0048, "step": 17740 }, { "grad_norm": 0.0791904628276825, "learning_rate": 3.4234514814451836e-06, "loss": 0.0048, "step": 17750 }, { "grad_norm": 0.08211734890937805, "learning_rate": 3.393449957242273e-06, "loss": 0.0048, "step": 17760 }, { "grad_norm": 0.11121908575296402, "learning_rate": 3.363575853567524e-06, "loss": 0.0049, "step": 17770 }, { "grad_norm": 0.129592627286911, "learning_rate": 3.3338292520955826e-06, "loss": 0.005, "step": 17780 }, { "grad_norm": 0.13581551611423492, "learning_rate": 3.304210234152516e-06, "loss": 0.0057, "step": 17790 }, { "grad_norm": 0.10036849230527878, "learning_rate": 3.2747188807155993e-06, "loss": 0.0051, "step": 17800 }, { "grad_norm": 0.1078474149107933, "learning_rate": 3.2453552724130643e-06, "loss": 0.0064, "step": 17810 }, { "grad_norm": 0.1062614917755127, "learning_rate": 3.216119489523889e-06, "loss": 0.0061, "step": 17820 }, { "grad_norm": 0.1100551038980484, "learning_rate": 3.1870116119775917e-06, "loss": 0.0048, "step": 17830 }, { "grad_norm": 0.11397500336170197, "learning_rate": 3.158031719353999e-06, "loss": 0.0057, "step": 17840 }, { "grad_norm": 0.1346493363380432, "learning_rate": 3.1291798908830273e-06, "loss": 0.0059, "step": 17850 }, { "grad_norm": 0.11316579580307007, "learning_rate": 3.1004562054444853e-06, "loss": 0.0054, "step": 17860 }, { "grad_norm": 0.29134660959243774, "learning_rate": 3.071860741567806e-06, "loss": 0.006, "step": 17870 }, { "grad_norm": 0.10469093173742294, "learning_rate": 3.04339357743193e-06, "loss": 0.0056, "step": 17880 }, { "grad_norm": 0.08655456453561783, "learning_rate": 3.0150547908649628e-06, "loss": 0.0049, "step": 17890 }, { "grad_norm": 0.09187447279691696, "learning_rate": 2.9868444593440957e-06, "loss": 0.005, "step": 17900 }, { "grad_norm": 0.1007750928401947, "learning_rate": 2.9587626599952846e-06, "loss": 0.0052, "step": 17910 }, { "grad_norm": 0.15991482138633728, "learning_rate": 2.930809469593082e-06, "loss": 0.0061, "step": 17920 }, { "grad_norm": 0.09056466817855835, "learning_rate": 2.9029849645604733e-06, "loss": 0.005, "step": 17930 }, { "grad_norm": 0.11904217302799225, "learning_rate": 2.8752892209685632e-06, "loss": 0.0051, "step": 17940 }, { "grad_norm": 0.07194598019123077, "learning_rate": 2.847722314536483e-06, "loss": 0.0049, "step": 17950 }, { "grad_norm": 0.12528035044670105, "learning_rate": 2.820284320631078e-06, "loss": 0.005, "step": 17960 }, { "grad_norm": 0.13251729309558868, "learning_rate": 2.792975314266788e-06, "loss": 0.0065, "step": 17970 }, { "grad_norm": 0.10227187722921371, "learning_rate": 2.7657953701054007e-06, "loss": 0.0042, "step": 17980 }, { "grad_norm": 0.15362229943275452, "learning_rate": 2.7387445624558306e-06, "loss": 0.0047, "step": 17990 }, { "grad_norm": 0.12829068303108215, "learning_rate": 2.7118229652739747e-06, "loss": 0.0052, "step": 18000 }, { "grad_norm": 0.10153088718652725, "learning_rate": 2.6850306521624236e-06, "loss": 0.0059, "step": 18010 }, { "grad_norm": 0.101234070956707, "learning_rate": 2.6583676963703507e-06, "loss": 0.0048, "step": 18020 }, { "grad_norm": 0.15708227455615997, "learning_rate": 2.631834170793268e-06, "loss": 0.0057, "step": 18030 }, { "grad_norm": 0.11492820829153061, "learning_rate": 2.6054301479728036e-06, "loss": 0.0058, "step": 18040 }, { "grad_norm": 0.08885297924280167, "learning_rate": 2.579155700096575e-06, "loss": 0.0066, "step": 18050 }, { "grad_norm": 0.0978974997997284, "learning_rate": 2.5530108989978873e-06, "loss": 0.0045, "step": 18060 }, { "grad_norm": 0.11558495461940765, "learning_rate": 2.5269958161556416e-06, "loss": 0.0058, "step": 18070 }, { "grad_norm": 0.11425362527370453, "learning_rate": 2.5011105226940888e-06, "loss": 0.0051, "step": 18080 }, { "grad_norm": 0.12994791567325592, "learning_rate": 2.4753550893826248e-06, "loss": 0.006, "step": 18090 }, { "grad_norm": 0.09137760102748871, "learning_rate": 2.4497295866356296e-06, "loss": 0.0047, "step": 18100 }, { "grad_norm": 0.1834350973367691, "learning_rate": 2.424234084512228e-06, "loss": 0.0056, "step": 18110 }, { "grad_norm": 0.11132363975048065, "learning_rate": 2.3988686527161687e-06, "loss": 0.0053, "step": 18120 }, { "grad_norm": 0.12850946187973022, "learning_rate": 2.373633360595573e-06, "loss": 0.0045, "step": 18130 }, { "grad_norm": 0.12878695130348206, "learning_rate": 2.3485282771427585e-06, "loss": 0.0076, "step": 18140 }, { "grad_norm": 0.08848581463098526, "learning_rate": 2.3235534709940665e-06, "loss": 0.004, "step": 18150 }, { "grad_norm": 0.10086081176996231, "learning_rate": 2.2987090104296617e-06, "loss": 0.0048, "step": 18160 }, { "grad_norm": 0.12925221025943756, "learning_rate": 2.273994963373355e-06, "loss": 0.0079, "step": 18170 }, { "grad_norm": 0.11536352336406708, "learning_rate": 2.249411397392409e-06, "loss": 0.0054, "step": 18180 }, { "grad_norm": 0.13330219686031342, "learning_rate": 2.2249583796973506e-06, "loss": 0.0057, "step": 18190 }, { "grad_norm": 0.08231659233570099, "learning_rate": 2.200635977141796e-06, "loss": 0.0046, "step": 18200 }, { "grad_norm": 0.08151145279407501, "learning_rate": 2.17644425622226e-06, "loss": 0.0054, "step": 18210 }, { "grad_norm": 0.12449768930673599, "learning_rate": 2.152383283077991e-06, "loss": 0.0064, "step": 18220 }, { "grad_norm": 0.09090007841587067, "learning_rate": 2.128453123490781e-06, "loss": 0.0056, "step": 18230 }, { "grad_norm": 0.11433246731758118, "learning_rate": 2.1046538428847462e-06, "loss": 0.0054, "step": 18240 }, { "grad_norm": 0.1320870816707611, "learning_rate": 2.0809855063262273e-06, "loss": 0.005, "step": 18250 }, { "grad_norm": 0.09928063303232193, "learning_rate": 2.057448178523558e-06, "loss": 0.005, "step": 18260 }, { "grad_norm": 0.10505295544862747, "learning_rate": 2.034041923826885e-06, "loss": 0.0056, "step": 18270 }, { "grad_norm": 0.1426680088043213, "learning_rate": 2.0107668062280204e-06, "loss": 0.0062, "step": 18280 }, { "grad_norm": 0.08686286956071854, "learning_rate": 1.9876228893602357e-06, "loss": 0.0051, "step": 18290 }, { "grad_norm": 0.07270726561546326, "learning_rate": 1.9646102364981266e-06, "loss": 0.0053, "step": 18300 }, { "grad_norm": 0.104503333568573, "learning_rate": 1.9417289105574053e-06, "loss": 0.0062, "step": 18310 }, { "grad_norm": 0.11189525574445724, "learning_rate": 1.9189789740947427e-06, "loss": 0.005, "step": 18320 }, { "grad_norm": 0.10227441787719727, "learning_rate": 1.896360489307597e-06, "loss": 0.0066, "step": 18330 }, { "grad_norm": 0.10397855937480927, "learning_rate": 1.8738735180340362e-06, "loss": 0.0053, "step": 18340 }, { "grad_norm": 0.09096582233905792, "learning_rate": 1.8515181217525824e-06, "loss": 0.0047, "step": 18350 }, { "grad_norm": 0.0782511904835701, "learning_rate": 1.8292943615820457e-06, "loss": 0.0051, "step": 18360 }, { "grad_norm": 0.09423886239528656, "learning_rate": 1.8072022982813296e-06, "loss": 0.0057, "step": 18370 }, { "grad_norm": 0.09022817760705948, "learning_rate": 1.7852419922492925e-06, "loss": 0.0068, "step": 18380 }, { "grad_norm": 0.11074832826852798, "learning_rate": 1.763413503524569e-06, "loss": 0.0062, "step": 18390 }, { "grad_norm": 0.1083219051361084, "learning_rate": 1.7417168917854165e-06, "loss": 0.006, "step": 18400 }, { "grad_norm": 0.1038384959101677, "learning_rate": 1.720152216349552e-06, "loss": 0.0076, "step": 18410 }, { "grad_norm": 0.08994381129741669, "learning_rate": 1.6987195361739595e-06, "loss": 0.005, "step": 18420 }, { "grad_norm": 0.15875884890556335, "learning_rate": 1.6774189098547832e-06, "loss": 0.0073, "step": 18430 }, { "grad_norm": 0.10981348156929016, "learning_rate": 1.6562503956271069e-06, "loss": 0.0041, "step": 18440 }, { "grad_norm": 0.12093932181596756, "learning_rate": 1.6352140513648417e-06, "loss": 0.0057, "step": 18450 }, { "grad_norm": 0.07950583845376968, "learning_rate": 1.6143099345805712e-06, "loss": 0.0054, "step": 18460 }, { "grad_norm": 0.08926916867494583, "learning_rate": 1.5935381024253293e-06, "loss": 0.0052, "step": 18470 }, { "grad_norm": 0.10398837178945541, "learning_rate": 1.572898611688517e-06, "loss": 0.0053, "step": 18480 }, { "grad_norm": 0.10671838372945786, "learning_rate": 1.5523915187977133e-06, "loss": 0.0051, "step": 18490 }, { "grad_norm": 0.10671882331371307, "learning_rate": 1.532016879818532e-06, "loss": 0.0052, "step": 18500 }, { "grad_norm": 0.10309796780347824, "learning_rate": 1.51177475045447e-06, "loss": 0.0061, "step": 18510 }, { "grad_norm": 0.07866723090410233, "learning_rate": 1.4916651860467035e-06, "loss": 0.0068, "step": 18520 }, { "grad_norm": 0.09305112063884735, "learning_rate": 1.471688241574043e-06, "loss": 0.0042, "step": 18530 }, { "grad_norm": 0.13106580078601837, "learning_rate": 1.451843971652672e-06, "loss": 0.0065, "step": 18540 }, { "grad_norm": 0.1263704001903534, "learning_rate": 1.432132430536076e-06, "loss": 0.0049, "step": 18550 }, { "grad_norm": 0.07306749373674393, "learning_rate": 1.412553672114869e-06, "loss": 0.0051, "step": 18560 }, { "grad_norm": 0.10874597728252411, "learning_rate": 1.3931077499166056e-06, "loss": 0.0071, "step": 18570 }, { "grad_norm": 0.06817004829645157, "learning_rate": 1.3737947171057085e-06, "loss": 0.0043, "step": 18580 }, { "grad_norm": 0.08748575299978256, "learning_rate": 1.3546146264832582e-06, "loss": 0.0058, "step": 18590 }, { "grad_norm": 0.1245136559009552, "learning_rate": 1.3355675304869086e-06, "loss": 0.0055, "step": 18600 }, { "grad_norm": 0.0832233875989914, "learning_rate": 1.3166534811906827e-06, "loss": 0.0065, "step": 18610 }, { "grad_norm": 0.10253541171550751, "learning_rate": 1.2978725303048666e-06, "loss": 0.0047, "step": 18620 }, { "grad_norm": 0.07595419883728027, "learning_rate": 1.2792247291758762e-06, "loss": 0.0049, "step": 18630 }, { "grad_norm": 0.11781004816293716, "learning_rate": 1.2607101287860635e-06, "loss": 0.0055, "step": 18640 }, { "grad_norm": 0.0805324986577034, "learning_rate": 1.2423287797536654e-06, "loss": 0.0042, "step": 18650 }, { "grad_norm": 0.08511217683553696, "learning_rate": 1.2240807323325776e-06, "loss": 0.0047, "step": 18660 }, { "grad_norm": 0.08469349890947342, "learning_rate": 1.205966036412254e-06, "loss": 0.0039, "step": 18670 }, { "grad_norm": 0.08725052326917648, "learning_rate": 1.1879847415175949e-06, "loss": 0.006, "step": 18680 }, { "grad_norm": 0.14567002654075623, "learning_rate": 1.1701368968087712e-06, "loss": 0.0064, "step": 18690 }, { "grad_norm": 0.13138769567012787, "learning_rate": 1.1524225510811116e-06, "loss": 0.0052, "step": 18700 }, { "grad_norm": 0.09690315276384354, "learning_rate": 1.1348417527649535e-06, "loss": 0.0045, "step": 18710 }, { "grad_norm": 0.11779256165027618, "learning_rate": 1.1173945499255268e-06, "loss": 0.0053, "step": 18720 }, { "grad_norm": 0.0843549445271492, "learning_rate": 1.1000809902628307e-06, "loss": 0.0038, "step": 18730 }, { "grad_norm": 0.10155168920755386, "learning_rate": 1.082901121111468e-06, "loss": 0.0068, "step": 18740 }, { "grad_norm": 0.08227349817752838, "learning_rate": 1.0658549894405456e-06, "loss": 0.0057, "step": 18750 }, { "grad_norm": 0.07986979931592941, "learning_rate": 1.0489426418535342e-06, "loss": 0.0045, "step": 18760 }, { "grad_norm": 0.09837976843118668, "learning_rate": 1.0321641245881474e-06, "loss": 0.0052, "step": 18770 }, { "grad_norm": 0.08113233000040054, "learning_rate": 1.015519483516214e-06, "loss": 0.0054, "step": 18780 }, { "grad_norm": 0.0933399498462677, "learning_rate": 9.990087641435443e-07, "loss": 0.0059, "step": 18790 }, { "grad_norm": 0.09679632633924484, "learning_rate": 9.826320116098132e-07, "loss": 0.0063, "step": 18800 }, { "grad_norm": 0.06762275099754333, "learning_rate": 9.663892706884447e-07, "loss": 0.0057, "step": 18810 }, { "grad_norm": 0.08988959342241287, "learning_rate": 9.502805857864616e-07, "loss": 0.0049, "step": 18820 }, { "grad_norm": 0.12832339107990265, "learning_rate": 9.34306000944396e-07, "loss": 0.0048, "step": 18830 }, { "grad_norm": 0.091166652739048, "learning_rate": 9.184655598361624e-07, "loss": 0.0055, "step": 18840 }, { "grad_norm": 0.11329247057437897, "learning_rate": 9.027593057689076e-07, "loss": 0.0059, "step": 18850 }, { "grad_norm": 0.0701126977801323, "learning_rate": 8.871872816829441e-07, "loss": 0.0052, "step": 18860 }, { "grad_norm": 0.07603112608194351, "learning_rate": 8.717495301515777e-07, "loss": 0.0045, "step": 18870 }, { "grad_norm": 0.07715100795030594, "learning_rate": 8.564460933810415e-07, "loss": 0.005, "step": 18880 }, { "grad_norm": 0.09751439839601517, "learning_rate": 8.412770132103453e-07, "loss": 0.0049, "step": 18890 }, { "grad_norm": 0.09091783314943314, "learning_rate": 8.262423311111711e-07, "loss": 0.0051, "step": 18900 }, { "grad_norm": 0.09288947284221649, "learning_rate": 8.113420881877665e-07, "loss": 0.0047, "step": 18910 }, { "grad_norm": 0.11184137314558029, "learning_rate": 7.965763251768288e-07, "loss": 0.0064, "step": 18920 }, { "grad_norm": 0.11003249138593674, "learning_rate": 7.819450824473995e-07, "loss": 0.0052, "step": 18930 }, { "grad_norm": 0.07448048144578934, "learning_rate": 7.674484000007198e-07, "loss": 0.0042, "step": 18940 }, { "grad_norm": 0.07449383288621902, "learning_rate": 7.530863174701752e-07, "loss": 0.0059, "step": 18950 }, { "grad_norm": 0.08693499863147736, "learning_rate": 7.38858874121151e-07, "loss": 0.0046, "step": 18960 }, { "grad_norm": 0.06182488054037094, "learning_rate": 7.247661088509328e-07, "loss": 0.0066, "step": 18970 }, { "grad_norm": 0.07086401432752609, "learning_rate": 7.108080601886002e-07, "loss": 0.0059, "step": 18980 }, { "grad_norm": 0.1523330956697464, "learning_rate": 6.969847662949336e-07, "loss": 0.0064, "step": 18990 }, { "grad_norm": 0.07839422672986984, "learning_rate": 6.832962649622798e-07, "loss": 0.0051, "step": 19000 }, { "grad_norm": 0.08144526183605194, "learning_rate": 6.697425936144863e-07, "loss": 0.0042, "step": 19010 }, { "grad_norm": 0.0743093341588974, "learning_rate": 6.563237893067731e-07, "loss": 0.0045, "step": 19020 }, { "grad_norm": 0.06004352867603302, "learning_rate": 6.430398887256328e-07, "loss": 0.0061, "step": 19030 }, { "grad_norm": 0.08406111598014832, "learning_rate": 6.298909281887478e-07, "loss": 0.0065, "step": 19040 }, { "grad_norm": 0.0815540999174118, "learning_rate": 6.168769436448673e-07, "loss": 0.0055, "step": 19050 }, { "grad_norm": 0.09734529256820679, "learning_rate": 6.03997970673742e-07, "loss": 0.0082, "step": 19060 }, { "grad_norm": 0.11770673096179962, "learning_rate": 5.912540444859782e-07, "loss": 0.006, "step": 19070 }, { "grad_norm": 0.07455963641405106, "learning_rate": 5.786451999229837e-07, "loss": 0.006, "step": 19080 }, { "grad_norm": 0.09497006237506866, "learning_rate": 5.661714714568722e-07, "loss": 0.0059, "step": 19090 }, { "grad_norm": 0.10224205255508423, "learning_rate": 5.538328931903259e-07, "loss": 0.0064, "step": 19100 }, { "grad_norm": 0.06576225906610489, "learning_rate": 5.416294988565551e-07, "loss": 0.008, "step": 19110 }, { "grad_norm": 0.09745963662862778, "learning_rate": 5.29561321819172e-07, "loss": 0.0056, "step": 19120 }, { "grad_norm": 0.11956269294023514, "learning_rate": 5.176283950721061e-07, "loss": 0.006, "step": 19130 }, { "grad_norm": 0.11937618255615234, "learning_rate": 5.058307512395332e-07, "loss": 0.0065, "step": 19140 }, { "grad_norm": 0.0796266570687294, "learning_rate": 4.941684225757526e-07, "loss": 0.0059, "step": 19150 }, { "grad_norm": 0.05752451717853546, "learning_rate": 4.826414409651314e-07, "loss": 0.005, "step": 19160 }, { "grad_norm": 0.047165073454380035, "learning_rate": 4.712498379219943e-07, "loss": 0.0042, "step": 19170 }, { "grad_norm": 0.13842470943927765, "learning_rate": 4.599936445905506e-07, "loss": 0.0052, "step": 19180 }, { "grad_norm": 0.10091246664524078, "learning_rate": 4.4887289174480594e-07, "loss": 0.0063, "step": 19190 }, { "grad_norm": 0.09250704199075699, "learning_rate": 4.378876097884621e-07, "loss": 0.0064, "step": 19200 }, { "grad_norm": 0.06610969454050064, "learning_rate": 4.2703782875487264e-07, "loss": 0.005, "step": 19210 }, { "grad_norm": 0.10367883741855621, "learning_rate": 4.163235783069208e-07, "loss": 0.0055, "step": 19220 }, { "grad_norm": 0.09097083657979965, "learning_rate": 4.057448877369585e-07, "loss": 0.005, "step": 19230 }, { "grad_norm": 0.07897879928350449, "learning_rate": 3.9530178596672295e-07, "loss": 0.0052, "step": 19240 }, { "grad_norm": 0.14147546887397766, "learning_rate": 3.849943015472479e-07, "loss": 0.0076, "step": 19250 }, { "grad_norm": 0.09322396665811539, "learning_rate": 3.748224626588137e-07, "loss": 0.0056, "step": 19260 }, { "grad_norm": 0.10565441846847534, "learning_rate": 3.647862971108307e-07, "loss": 0.0043, "step": 19270 }, { "grad_norm": 0.1601743847131729, "learning_rate": 3.5488583234179473e-07, "loss": 0.0069, "step": 19280 }, { "grad_norm": 0.07183156162500381, "learning_rate": 3.4512109541920413e-07, "loss": 0.0046, "step": 19290 }, { "grad_norm": 0.07974271476268768, "learning_rate": 3.354921130394706e-07, "loss": 0.0056, "step": 19300 }, { "grad_norm": 0.045639269053936005, "learning_rate": 3.259989115278639e-07, "loss": 0.0052, "step": 19310 }, { "grad_norm": 0.14597158133983612, "learning_rate": 3.1664151683843403e-07, "loss": 0.0069, "step": 19320 }, { "grad_norm": 0.056339800357818604, "learning_rate": 3.074199545539447e-07, "loss": 0.0047, "step": 19330 }, { "grad_norm": 0.07108741998672485, "learning_rate": 2.983342498857955e-07, "loss": 0.0053, "step": 19340 }, { "grad_norm": 0.0384625606238842, "learning_rate": 2.893844276739499e-07, "loss": 0.0048, "step": 19350 }, { "grad_norm": 0.050127141177654266, "learning_rate": 2.8057051238688514e-07, "loss": 0.0048, "step": 19360 }, { "grad_norm": 0.12492375075817108, "learning_rate": 2.71892528121509e-07, "loss": 0.0082, "step": 19370 }, { "grad_norm": 0.0854385644197464, "learning_rate": 2.633504986030988e-07, "loss": 0.0052, "step": 19380 }, { "grad_norm": 0.07452654838562012, "learning_rate": 2.549444471852347e-07, "loss": 0.0067, "step": 19390 }, { "grad_norm": 0.05686022713780403, "learning_rate": 2.4667439684974423e-07, "loss": 0.0057, "step": 19400 }, { "grad_norm": 0.07921334356069565, "learning_rate": 2.3854037020662467e-07, "loss": 0.0068, "step": 19410 }, { "grad_norm": 0.05087551474571228, "learning_rate": 2.3054238949399288e-07, "loss": 0.0054, "step": 19420 }, { "grad_norm": 0.1222381740808487, "learning_rate": 2.2268047657802993e-07, "loss": 0.005, "step": 19430 }, { "grad_norm": 0.07431309670209885, "learning_rate": 2.149546529529034e-07, "loss": 0.0067, "step": 19440 }, { "grad_norm": 0.055628616362810135, "learning_rate": 2.0736493974071736e-07, "loss": 0.0037, "step": 19450 }, { "grad_norm": 0.10360267013311386, "learning_rate": 1.9991135769145686e-07, "loss": 0.0052, "step": 19460 }, { "grad_norm": 0.0853598341345787, "learning_rate": 1.9259392718293245e-07, "loss": 0.0047, "step": 19470 }, { "grad_norm": 0.052843816578388214, "learning_rate": 1.8541266822072467e-07, "loss": 0.0045, "step": 19480 }, { "grad_norm": 0.056647755205631256, "learning_rate": 1.7836760043811184e-07, "loss": 0.0042, "step": 19490 }, { "grad_norm": 0.06782002747058868, "learning_rate": 1.7145874309604792e-07, "loss": 0.0042, "step": 19500 }, { "grad_norm": 0.05878906324505806, "learning_rate": 1.6468611508308474e-07, "loss": 0.0075, "step": 19510 }, { "grad_norm": 0.11906403303146362, "learning_rate": 1.5804973491532204e-07, "loss": 0.0049, "step": 19520 }, { "grad_norm": 0.07525583356618881, "learning_rate": 1.5154962073637424e-07, "loss": 0.0057, "step": 19530 }, { "grad_norm": 0.0704353004693985, "learning_rate": 1.4518579031730372e-07, "loss": 0.004, "step": 19540 }, { "grad_norm": 0.07189474999904633, "learning_rate": 1.389582610565876e-07, "loss": 0.0056, "step": 19550 }, { "grad_norm": 0.05611153692007065, "learning_rate": 1.3286704998003995e-07, "loss": 0.0063, "step": 19560 }, { "grad_norm": 0.0708836019039154, "learning_rate": 1.2691217374080632e-07, "loss": 0.006, "step": 19570 }, { "grad_norm": 0.09524258971214294, "learning_rate": 1.2109364861929705e-07, "loss": 0.0052, "step": 19580 }, { "grad_norm": 0.09185714274644852, "learning_rate": 1.1541149052312628e-07, "loss": 0.0041, "step": 19590 }, { "grad_norm": 0.08231845498085022, "learning_rate": 1.0986571498710074e-07, "loss": 0.0048, "step": 19600 }, { "grad_norm": 0.06013771519064903, "learning_rate": 1.0445633717316438e-07, "loss": 0.0058, "step": 19610 }, { "grad_norm": 0.1281096488237381, "learning_rate": 9.918337187034277e-08, "loss": 0.0047, "step": 19620 }, { "grad_norm": 0.07719490677118301, "learning_rate": 9.404683349472643e-08, "loss": 0.0061, "step": 19630 }, { "grad_norm": 0.0726354643702507, "learning_rate": 8.904673608940983e-08, "loss": 0.004, "step": 19640 }, { "grad_norm": 0.0537855438888073, "learning_rate": 8.418309332447471e-08, "loss": 0.0064, "step": 19650 }, { "grad_norm": 0.07978323101997375, "learning_rate": 7.945591849692902e-08, "loss": 0.0069, "step": 19660 }, { "grad_norm": 0.11475620418787003, "learning_rate": 7.486522453069578e-08, "loss": 0.007, "step": 19670 }, { "grad_norm": 0.08882953226566315, "learning_rate": 7.041102397655208e-08, "loss": 0.0045, "step": 19680 }, { "grad_norm": 0.0823129415512085, "learning_rate": 6.609332901210685e-08, "loss": 0.0046, "step": 19690 }, { "grad_norm": 0.1146220788359642, "learning_rate": 6.191215144178419e-08, "loss": 0.0046, "step": 19700 }, { "grad_norm": 0.04314541444182396, "learning_rate": 5.786750269675678e-08, "loss": 0.0042, "step": 19710 }, { "grad_norm": 0.0934709757566452, "learning_rate": 5.395939383494031e-08, "loss": 0.0049, "step": 19720 }, { "grad_norm": 0.0507025346159935, "learning_rate": 5.018783554095463e-08, "loss": 0.0047, "step": 19730 }, { "grad_norm": 0.10285497456789017, "learning_rate": 4.655283812610156e-08, "loss": 0.0053, "step": 19740 }, { "grad_norm": 0.07106712460517883, "learning_rate": 4.305441152831491e-08, "loss": 0.0054, "step": 19750 }, { "grad_norm": 0.07566630095243454, "learning_rate": 3.9692565312171584e-08, "loss": 0.0079, "step": 19760 }, { "grad_norm": 0.098568394780159, "learning_rate": 3.6467308668824975e-08, "loss": 0.0053, "step": 19770 }, { "grad_norm": 0.0953674167394638, "learning_rate": 3.3378650416004964e-08, "loss": 0.0043, "step": 19780 }, { "grad_norm": 0.08401796221733093, "learning_rate": 3.042659899797906e-08, "loss": 0.0061, "step": 19790 }, { "grad_norm": 0.05664513260126114, "learning_rate": 2.76111624855524e-08, "loss": 0.0058, "step": 19800 }, { "grad_norm": 0.06266386061906815, "learning_rate": 2.4932348576017784e-08, "loss": 0.0058, "step": 19810 }, { "grad_norm": 0.09404028952121735, "learning_rate": 2.239016459314458e-08, "loss": 0.0047, "step": 19820 }, { "grad_norm": 0.07271256297826767, "learning_rate": 1.9984617487173174e-08, "loss": 0.0057, "step": 19830 }, { "grad_norm": 0.07685194164514542, "learning_rate": 1.7715713834776105e-08, "loss": 0.0052, "step": 19840 }, { "grad_norm": 0.039293911308050156, "learning_rate": 1.5583459839046964e-08, "loss": 0.0047, "step": 19850 }, { "grad_norm": 0.1079830750823021, "learning_rate": 1.3587861329489304e-08, "loss": 0.0054, "step": 19860 }, { "grad_norm": 0.06120683252811432, "learning_rate": 1.1728923761994415e-08, "loss": 0.0053, "step": 19870 }, { "grad_norm": 0.08583294600248337, "learning_rate": 1.0006652218819135e-08, "loss": 0.0053, "step": 19880 }, { "grad_norm": 0.10686157643795013, "learning_rate": 8.421051408596947e-09, "loss": 0.0052, "step": 19890 }, { "grad_norm": 0.060749560594558716, "learning_rate": 6.972125666299123e-09, "loss": 0.0049, "step": 19900 }, { "grad_norm": 0.060505788773298264, "learning_rate": 5.659878953229169e-09, "loss": 0.0065, "step": 19910 }, { "grad_norm": 0.09156258404254913, "learning_rate": 4.48431485701728e-09, "loss": 0.0057, "step": 19920 }, { "grad_norm": 0.12245126813650131, "learning_rate": 3.4454365916203322e-09, "loss": 0.0075, "step": 19930 }, { "grad_norm": 0.06031413748860359, "learning_rate": 2.5432469972830332e-09, "loss": 0.0047, "step": 19940 }, { "grad_norm": 0.06300681084394455, "learning_rate": 1.7777485405601203e-09, "loss": 0.0049, "step": 19950 }, { "grad_norm": 0.1259831041097641, "learning_rate": 1.1489433142941597e-09, "loss": 0.0079, "step": 19960 }, { "grad_norm": 0.09711910039186478, "learning_rate": 6.568330376210963e-10, "loss": 0.0043, "step": 19970 }, { "grad_norm": 0.09780178964138031, "learning_rate": 3.0141905594249787e-10, "loss": 0.0053, "step": 19980 }, { "grad_norm": 0.05600352585315704, "learning_rate": 8.270234094776008e-11, "loss": 0.0045, "step": 19990 }, { "grad_norm": 0.06647201627492905, "learning_rate": 6.834906085551041e-13, "loss": 0.0057, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }