{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.4338237047195435, "learning_rate": 4.5e-07, "loss": 1.3123, "step": 10 }, { "grad_norm": 1.6511415243148804, "learning_rate": 9.5e-07, "loss": 1.3129, "step": 20 }, { "grad_norm": 1.306490182876587, "learning_rate": 1.45e-06, "loss": 1.2929, "step": 30 }, { "grad_norm": 0.8660150170326233, "learning_rate": 1.95e-06, "loss": 1.2857, "step": 40 }, { "grad_norm": 0.6951407194137573, "learning_rate": 2.4500000000000003e-06, "loss": 1.2488, "step": 50 }, { "grad_norm": 0.40770986676216125, "learning_rate": 2.95e-06, "loss": 1.2168, "step": 60 }, { "grad_norm": 0.3530285954475403, "learning_rate": 3.4500000000000004e-06, "loss": 1.1988, "step": 70 }, { "grad_norm": 0.3432512879371643, "learning_rate": 3.95e-06, "loss": 1.193, "step": 80 }, { "grad_norm": 0.5069421529769897, "learning_rate": 4.45e-06, "loss": 1.1749, "step": 90 }, { "grad_norm": 0.5906297564506531, "learning_rate": 4.950000000000001e-06, "loss": 1.1457, "step": 100 }, { "grad_norm": 0.4744424819946289, "learning_rate": 5.45e-06, "loss": 1.0997, "step": 110 }, { "grad_norm": 0.6645118594169617, "learning_rate": 5.95e-06, "loss": 1.0868, "step": 120 }, { "grad_norm": 0.4118809998035431, "learning_rate": 6.45e-06, "loss": 1.0821, "step": 130 }, { "grad_norm": 0.39242810010910034, "learning_rate": 6.950000000000001e-06, "loss": 1.0752, "step": 140 }, { "grad_norm": 0.4555911123752594, "learning_rate": 7.45e-06, "loss": 1.0632, "step": 150 }, { "grad_norm": 0.5800536870956421, "learning_rate": 7.95e-06, "loss": 1.0557, "step": 160 }, { "grad_norm": 0.47157812118530273, "learning_rate": 8.45e-06, "loss": 1.0481, "step": 170 }, { "grad_norm": 0.5194073915481567, "learning_rate": 8.95e-06, "loss": 1.049, "step": 180 }, { "grad_norm": 0.48847419023513794, "learning_rate": 9.450000000000001e-06, "loss": 1.0428, "step": 190 }, { "grad_norm": 0.4875481426715851, "learning_rate": 9.950000000000001e-06, "loss": 1.0415, "step": 200 }, { "grad_norm": 0.45302116870880127, "learning_rate": 1.045e-05, "loss": 1.038, "step": 210 }, { "grad_norm": 0.46819838881492615, "learning_rate": 1.095e-05, "loss": 1.0383, "step": 220 }, { "grad_norm": 0.5408715009689331, "learning_rate": 1.145e-05, "loss": 1.0277, "step": 230 }, { "grad_norm": 0.52568519115448, "learning_rate": 1.195e-05, "loss": 1.0301, "step": 240 }, { "grad_norm": 0.501237154006958, "learning_rate": 1.2450000000000001e-05, "loss": 1.0189, "step": 250 }, { "grad_norm": 1.000428318977356, "learning_rate": 1.2950000000000001e-05, "loss": 1.0109, "step": 260 }, { "grad_norm": 0.7719165086746216, "learning_rate": 1.3450000000000002e-05, "loss": 0.987, "step": 270 }, { "grad_norm": 1.130938172340393, "learning_rate": 1.3950000000000002e-05, "loss": 0.9546, "step": 280 }, { "grad_norm": 1.0383124351501465, "learning_rate": 1.4449999999999999e-05, "loss": 0.9145, "step": 290 }, { "grad_norm": 1.2015764713287354, "learning_rate": 1.4950000000000001e-05, "loss": 0.8873, "step": 300 }, { "grad_norm": 1.3447803258895874, "learning_rate": 1.545e-05, "loss": 0.8662, "step": 310 }, { "grad_norm": 1.29972505569458, "learning_rate": 1.595e-05, "loss": 0.8195, "step": 320 }, { "grad_norm": 1.6097290515899658, "learning_rate": 1.645e-05, "loss": 0.7704, "step": 330 }, { "grad_norm": 1.1583374738693237, "learning_rate": 1.6950000000000002e-05, "loss": 0.7272, "step": 340 }, { "grad_norm": 1.6486802101135254, "learning_rate": 1.745e-05, "loss": 0.693, "step": 350 }, { "grad_norm": 1.6061633825302124, "learning_rate": 1.795e-05, "loss": 0.6625, "step": 360 }, { "grad_norm": 1.1828113794326782, "learning_rate": 1.845e-05, "loss": 0.6288, "step": 370 }, { "grad_norm": 1.5731960535049438, "learning_rate": 1.895e-05, "loss": 0.6125, "step": 380 }, { "grad_norm": 1.477593183517456, "learning_rate": 1.9450000000000002e-05, "loss": 0.5889, "step": 390 }, { "grad_norm": 1.5948655605316162, "learning_rate": 1.995e-05, "loss": 0.5482, "step": 400 }, { "grad_norm": 1.3191596269607544, "learning_rate": 2.045e-05, "loss": 0.5218, "step": 410 }, { "grad_norm": 1.8621004819869995, "learning_rate": 2.095e-05, "loss": 0.4725, "step": 420 }, { "grad_norm": 1.6861158609390259, "learning_rate": 2.145e-05, "loss": 0.451, "step": 430 }, { "grad_norm": 1.8634982109069824, "learning_rate": 2.195e-05, "loss": 0.4281, "step": 440 }, { "grad_norm": 1.88406240940094, "learning_rate": 2.245e-05, "loss": 0.4081, "step": 450 }, { "grad_norm": 1.9895440340042114, "learning_rate": 2.2950000000000002e-05, "loss": 0.3689, "step": 460 }, { "grad_norm": 1.8112493753433228, "learning_rate": 2.345e-05, "loss": 0.3474, "step": 470 }, { "grad_norm": 1.6414830684661865, "learning_rate": 2.395e-05, "loss": 0.306, "step": 480 }, { "grad_norm": 1.948697566986084, "learning_rate": 2.445e-05, "loss": 0.2784, "step": 490 }, { "grad_norm": 1.7593988180160522, "learning_rate": 2.495e-05, "loss": 0.2674, "step": 500 }, { "grad_norm": 1.4780184030532837, "learning_rate": 2.5450000000000002e-05, "loss": 0.2503, "step": 510 }, { "grad_norm": 1.4023900032043457, "learning_rate": 2.595e-05, "loss": 0.2288, "step": 520 }, { "grad_norm": 2.032135486602783, "learning_rate": 2.6450000000000003e-05, "loss": 0.2115, "step": 530 }, { "grad_norm": 1.2554078102111816, "learning_rate": 2.6950000000000005e-05, "loss": 0.2006, "step": 540 }, { "grad_norm": 1.6227610111236572, "learning_rate": 2.7450000000000003e-05, "loss": 0.1748, "step": 550 }, { "grad_norm": 1.5409941673278809, "learning_rate": 2.7950000000000005e-05, "loss": 0.158, "step": 560 }, { "grad_norm": 1.6654449701309204, "learning_rate": 2.845e-05, "loss": 0.1563, "step": 570 }, { "grad_norm": 2.0191874504089355, "learning_rate": 2.895e-05, "loss": 0.1438, "step": 580 }, { "grad_norm": 1.5469263792037964, "learning_rate": 2.945e-05, "loss": 0.1307, "step": 590 }, { "grad_norm": 1.8088122606277466, "learning_rate": 2.995e-05, "loss": 0.1323, "step": 600 }, { "grad_norm": 1.821864128112793, "learning_rate": 3.045e-05, "loss": 0.1194, "step": 610 }, { "grad_norm": 1.493886947631836, "learning_rate": 3.095e-05, "loss": 0.1272, "step": 620 }, { "grad_norm": 1.4056199789047241, "learning_rate": 3.145e-05, "loss": 0.1228, "step": 630 }, { "grad_norm": 1.8555517196655273, "learning_rate": 3.1950000000000004e-05, "loss": 0.1135, "step": 640 }, { "grad_norm": 1.459626317024231, "learning_rate": 3.245e-05, "loss": 0.1139, "step": 650 }, { "grad_norm": 1.8944323062896729, "learning_rate": 3.295e-05, "loss": 0.1198, "step": 660 }, { "grad_norm": 1.667932152748108, "learning_rate": 3.345000000000001e-05, "loss": 0.1131, "step": 670 }, { "grad_norm": 1.3105109930038452, "learning_rate": 3.3950000000000005e-05, "loss": 0.105, "step": 680 }, { "grad_norm": 1.4199398756027222, "learning_rate": 3.445e-05, "loss": 0.1053, "step": 690 }, { "grad_norm": 1.2526229619979858, "learning_rate": 3.495e-05, "loss": 0.1016, "step": 700 }, { "grad_norm": 2.087236166000366, "learning_rate": 3.545e-05, "loss": 0.1169, "step": 710 }, { "grad_norm": 1.685860276222229, "learning_rate": 3.595e-05, "loss": 0.1064, "step": 720 }, { "grad_norm": 1.3472065925598145, "learning_rate": 3.645e-05, "loss": 0.0936, "step": 730 }, { "grad_norm": 1.359641432762146, "learning_rate": 3.6950000000000004e-05, "loss": 0.1041, "step": 740 }, { "grad_norm": 1.50807785987854, "learning_rate": 3.745e-05, "loss": 0.105, "step": 750 }, { "grad_norm": 1.144506812095642, "learning_rate": 3.795e-05, "loss": 0.1111, "step": 760 }, { "grad_norm": 1.2602143287658691, "learning_rate": 3.845e-05, "loss": 0.1016, "step": 770 }, { "grad_norm": 1.5277796983718872, "learning_rate": 3.8950000000000005e-05, "loss": 0.1037, "step": 780 }, { "grad_norm": 0.8165663480758667, "learning_rate": 3.9450000000000003e-05, "loss": 0.1006, "step": 790 }, { "grad_norm": 1.0910130739212036, "learning_rate": 3.995e-05, "loss": 0.1002, "step": 800 }, { "grad_norm": 1.1227126121520996, "learning_rate": 4.045000000000001e-05, "loss": 0.0968, "step": 810 }, { "grad_norm": 1.3342103958129883, "learning_rate": 4.095e-05, "loss": 0.0943, "step": 820 }, { "grad_norm": 1.2951674461364746, "learning_rate": 4.145e-05, "loss": 0.0989, "step": 830 }, { "grad_norm": 1.1592192649841309, "learning_rate": 4.195e-05, "loss": 0.0907, "step": 840 }, { "grad_norm": 1.114880084991455, "learning_rate": 4.245e-05, "loss": 0.1002, "step": 850 }, { "grad_norm": 1.5469603538513184, "learning_rate": 4.295e-05, "loss": 0.1019, "step": 860 }, { "grad_norm": 1.2737501859664917, "learning_rate": 4.345e-05, "loss": 0.0908, "step": 870 }, { "grad_norm": 1.0775599479675293, "learning_rate": 4.3950000000000004e-05, "loss": 0.0963, "step": 880 }, { "grad_norm": 1.2390283346176147, "learning_rate": 4.445e-05, "loss": 0.0916, "step": 890 }, { "grad_norm": 1.2179304361343384, "learning_rate": 4.495e-05, "loss": 0.0843, "step": 900 }, { "grad_norm": 1.0226386785507202, "learning_rate": 4.545000000000001e-05, "loss": 0.0829, "step": 910 }, { "grad_norm": 0.9212608337402344, "learning_rate": 4.5950000000000006e-05, "loss": 0.08, "step": 920 }, { "grad_norm": 1.1385608911514282, "learning_rate": 4.6450000000000004e-05, "loss": 0.0829, "step": 930 }, { "grad_norm": 1.0507900714874268, "learning_rate": 4.695e-05, "loss": 0.0814, "step": 940 }, { "grad_norm": 1.26546311378479, "learning_rate": 4.745e-05, "loss": 0.0887, "step": 950 }, { "grad_norm": 1.0216257572174072, "learning_rate": 4.795e-05, "loss": 0.0851, "step": 960 }, { "grad_norm": 1.0690704584121704, "learning_rate": 4.845e-05, "loss": 0.0856, "step": 970 }, { "grad_norm": 1.2248351573944092, "learning_rate": 4.8950000000000004e-05, "loss": 0.0881, "step": 980 }, { "grad_norm": 1.0149790048599243, "learning_rate": 4.945e-05, "loss": 0.0886, "step": 990 }, { "grad_norm": 1.0083993673324585, "learning_rate": 4.995e-05, "loss": 0.0944, "step": 1000 }, { "grad_norm": 1.0354951620101929, "learning_rate": 5.045e-05, "loss": 0.0813, "step": 1010 }, { "grad_norm": 1.0324630737304688, "learning_rate": 5.095e-05, "loss": 0.0886, "step": 1020 }, { "grad_norm": 1.1402848958969116, "learning_rate": 5.145e-05, "loss": 0.0899, "step": 1030 }, { "grad_norm": 1.0474698543548584, "learning_rate": 5.1949999999999996e-05, "loss": 0.0818, "step": 1040 }, { "grad_norm": 0.8638880252838135, "learning_rate": 5.245e-05, "loss": 0.0763, "step": 1050 }, { "grad_norm": 1.1923514604568481, "learning_rate": 5.295e-05, "loss": 0.0755, "step": 1060 }, { "grad_norm": 0.8120508193969727, "learning_rate": 5.345e-05, "loss": 0.0766, "step": 1070 }, { "grad_norm": 1.164006233215332, "learning_rate": 5.3950000000000004e-05, "loss": 0.0705, "step": 1080 }, { "grad_norm": 1.044539213180542, "learning_rate": 5.445e-05, "loss": 0.0779, "step": 1090 }, { "grad_norm": 1.0373808145523071, "learning_rate": 5.495e-05, "loss": 0.08, "step": 1100 }, { "grad_norm": 0.7515385150909424, "learning_rate": 5.545e-05, "loss": 0.0705, "step": 1110 }, { "grad_norm": 1.07625150680542, "learning_rate": 5.5950000000000005e-05, "loss": 0.0792, "step": 1120 }, { "grad_norm": 1.1466848850250244, "learning_rate": 5.645e-05, "loss": 0.0829, "step": 1130 }, { "grad_norm": 1.101448893547058, "learning_rate": 5.695e-05, "loss": 0.0795, "step": 1140 }, { "grad_norm": 0.9182584881782532, "learning_rate": 5.745e-05, "loss": 0.0829, "step": 1150 }, { "grad_norm": 1.0032544136047363, "learning_rate": 5.7950000000000006e-05, "loss": 0.0747, "step": 1160 }, { "grad_norm": 1.01189386844635, "learning_rate": 5.8450000000000005e-05, "loss": 0.0783, "step": 1170 }, { "grad_norm": 1.0717337131500244, "learning_rate": 5.895e-05, "loss": 0.0797, "step": 1180 }, { "grad_norm": 0.8767481446266174, "learning_rate": 5.945000000000001e-05, "loss": 0.0748, "step": 1190 }, { "grad_norm": 0.8342227935791016, "learning_rate": 5.995000000000001e-05, "loss": 0.0814, "step": 1200 }, { "grad_norm": 0.8623939752578735, "learning_rate": 6.0450000000000006e-05, "loss": 0.0715, "step": 1210 }, { "grad_norm": 0.8532181978225708, "learning_rate": 6.0950000000000004e-05, "loss": 0.0659, "step": 1220 }, { "grad_norm": 0.7885238528251648, "learning_rate": 6.145e-05, "loss": 0.0669, "step": 1230 }, { "grad_norm": 1.0225868225097656, "learning_rate": 6.195e-05, "loss": 0.0744, "step": 1240 }, { "grad_norm": 0.8491858243942261, "learning_rate": 6.245000000000001e-05, "loss": 0.0717, "step": 1250 }, { "grad_norm": 0.9173391461372375, "learning_rate": 6.295e-05, "loss": 0.0607, "step": 1260 }, { "grad_norm": 0.9839532971382141, "learning_rate": 6.345e-05, "loss": 0.0767, "step": 1270 }, { "grad_norm": 1.1901814937591553, "learning_rate": 6.395e-05, "loss": 0.0698, "step": 1280 }, { "grad_norm": 0.8837777376174927, "learning_rate": 6.445e-05, "loss": 0.0745, "step": 1290 }, { "grad_norm": 0.9592897295951843, "learning_rate": 6.494999999999999e-05, "loss": 0.0739, "step": 1300 }, { "grad_norm": 0.9955523014068604, "learning_rate": 6.545e-05, "loss": 0.0648, "step": 1310 }, { "grad_norm": 0.7312012314796448, "learning_rate": 6.595e-05, "loss": 0.0807, "step": 1320 }, { "grad_norm": 0.7515588402748108, "learning_rate": 6.645e-05, "loss": 0.067, "step": 1330 }, { "grad_norm": 0.7312848567962646, "learning_rate": 6.695e-05, "loss": 0.0627, "step": 1340 }, { "grad_norm": 1.0297894477844238, "learning_rate": 6.745e-05, "loss": 0.0723, "step": 1350 }, { "grad_norm": 0.8895304799079895, "learning_rate": 6.795e-05, "loss": 0.0738, "step": 1360 }, { "grad_norm": 0.9083397388458252, "learning_rate": 6.845e-05, "loss": 0.0671, "step": 1370 }, { "grad_norm": 0.7077662944793701, "learning_rate": 6.895000000000001e-05, "loss": 0.0647, "step": 1380 }, { "grad_norm": 0.6716082692146301, "learning_rate": 6.945000000000001e-05, "loss": 0.0593, "step": 1390 }, { "grad_norm": 0.7517297863960266, "learning_rate": 6.995e-05, "loss": 0.0615, "step": 1400 }, { "grad_norm": 0.7263442277908325, "learning_rate": 7.045e-05, "loss": 0.0634, "step": 1410 }, { "grad_norm": 0.8439946174621582, "learning_rate": 7.095e-05, "loss": 0.061, "step": 1420 }, { "grad_norm": 0.8031980991363525, "learning_rate": 7.145e-05, "loss": 0.0747, "step": 1430 }, { "grad_norm": 0.8101364970207214, "learning_rate": 7.195e-05, "loss": 0.065, "step": 1440 }, { "grad_norm": 0.744645357131958, "learning_rate": 7.245000000000001e-05, "loss": 0.0633, "step": 1450 }, { "grad_norm": 0.7511458396911621, "learning_rate": 7.295000000000001e-05, "loss": 0.0668, "step": 1460 }, { "grad_norm": 0.8914756774902344, "learning_rate": 7.345000000000001e-05, "loss": 0.0614, "step": 1470 }, { "grad_norm": 0.8601694703102112, "learning_rate": 7.395000000000001e-05, "loss": 0.0613, "step": 1480 }, { "grad_norm": 0.7909899353981018, "learning_rate": 7.445000000000001e-05, "loss": 0.0652, "step": 1490 }, { "grad_norm": 0.5729796290397644, "learning_rate": 7.495e-05, "loss": 0.0616, "step": 1500 }, { "grad_norm": 0.7638779878616333, "learning_rate": 7.545e-05, "loss": 0.064, "step": 1510 }, { "grad_norm": 0.4940090775489807, "learning_rate": 7.595e-05, "loss": 0.0623, "step": 1520 }, { "grad_norm": 0.613258421421051, "learning_rate": 7.645e-05, "loss": 0.0595, "step": 1530 }, { "grad_norm": 0.7410160303115845, "learning_rate": 7.695e-05, "loss": 0.0637, "step": 1540 }, { "grad_norm": 0.8912451267242432, "learning_rate": 7.745e-05, "loss": 0.0623, "step": 1550 }, { "grad_norm": 0.6318579316139221, "learning_rate": 7.795e-05, "loss": 0.0618, "step": 1560 }, { "grad_norm": 0.7374716997146606, "learning_rate": 7.845e-05, "loss": 0.062, "step": 1570 }, { "grad_norm": 0.7169421911239624, "learning_rate": 7.895000000000001e-05, "loss": 0.0592, "step": 1580 }, { "grad_norm": 0.6151158809661865, "learning_rate": 7.945e-05, "loss": 0.0599, "step": 1590 }, { "grad_norm": 0.5616509318351746, "learning_rate": 7.995e-05, "loss": 0.0643, "step": 1600 }, { "grad_norm": 0.5412330627441406, "learning_rate": 8.045e-05, "loss": 0.0558, "step": 1610 }, { "grad_norm": 0.7328203916549683, "learning_rate": 8.095e-05, "loss": 0.0541, "step": 1620 }, { "grad_norm": 0.6231189966201782, "learning_rate": 8.145e-05, "loss": 0.0592, "step": 1630 }, { "grad_norm": 0.6689634919166565, "learning_rate": 8.195e-05, "loss": 0.0616, "step": 1640 }, { "grad_norm": 0.7098618745803833, "learning_rate": 8.245e-05, "loss": 0.0558, "step": 1650 }, { "grad_norm": 0.7709417939186096, "learning_rate": 8.295000000000001e-05, "loss": 0.0538, "step": 1660 }, { "grad_norm": 0.7209882140159607, "learning_rate": 8.345000000000001e-05, "loss": 0.0593, "step": 1670 }, { "grad_norm": 0.60904860496521, "learning_rate": 8.395000000000001e-05, "loss": 0.0574, "step": 1680 }, { "grad_norm": 0.5733396410942078, "learning_rate": 8.445e-05, "loss": 0.0556, "step": 1690 }, { "grad_norm": 0.7508549690246582, "learning_rate": 8.495e-05, "loss": 0.0649, "step": 1700 }, { "grad_norm": 0.6308852434158325, "learning_rate": 8.545e-05, "loss": 0.0579, "step": 1710 }, { "grad_norm": 0.852810800075531, "learning_rate": 8.595e-05, "loss": 0.0632, "step": 1720 }, { "grad_norm": 0.7620308995246887, "learning_rate": 8.645000000000001e-05, "loss": 0.0622, "step": 1730 }, { "grad_norm": 0.7086771130561829, "learning_rate": 8.695000000000001e-05, "loss": 0.0538, "step": 1740 }, { "grad_norm": 0.6035969257354736, "learning_rate": 8.745000000000001e-05, "loss": 0.0567, "step": 1750 }, { "grad_norm": 0.5534874796867371, "learning_rate": 8.795e-05, "loss": 0.0521, "step": 1760 }, { "grad_norm": 0.8147499561309814, "learning_rate": 8.845e-05, "loss": 0.0674, "step": 1770 }, { "grad_norm": 0.6611758470535278, "learning_rate": 8.895e-05, "loss": 0.062, "step": 1780 }, { "grad_norm": 0.5858861207962036, "learning_rate": 8.945e-05, "loss": 0.0587, "step": 1790 }, { "grad_norm": 0.660905659198761, "learning_rate": 8.995e-05, "loss": 0.0641, "step": 1800 }, { "grad_norm": 0.6903167366981506, "learning_rate": 9.045e-05, "loss": 0.0555, "step": 1810 }, { "grad_norm": 0.6571730375289917, "learning_rate": 9.095e-05, "loss": 0.0586, "step": 1820 }, { "grad_norm": 0.5872697830200195, "learning_rate": 9.145e-05, "loss": 0.0552, "step": 1830 }, { "grad_norm": 0.8540509343147278, "learning_rate": 9.195e-05, "loss": 0.0532, "step": 1840 }, { "grad_norm": 0.6255236864089966, "learning_rate": 9.245e-05, "loss": 0.0572, "step": 1850 }, { "grad_norm": 0.5606691241264343, "learning_rate": 9.295000000000001e-05, "loss": 0.0557, "step": 1860 }, { "grad_norm": 0.7602635622024536, "learning_rate": 9.345000000000001e-05, "loss": 0.0578, "step": 1870 }, { "grad_norm": 0.616365373134613, "learning_rate": 9.395000000000001e-05, "loss": 0.0576, "step": 1880 }, { "grad_norm": 0.5941797494888306, "learning_rate": 9.445e-05, "loss": 0.0511, "step": 1890 }, { "grad_norm": 0.7084371447563171, "learning_rate": 9.495e-05, "loss": 0.0566, "step": 1900 }, { "grad_norm": 0.6298006176948547, "learning_rate": 9.545e-05, "loss": 0.0548, "step": 1910 }, { "grad_norm": 0.7448440790176392, "learning_rate": 9.595e-05, "loss": 0.052, "step": 1920 }, { "grad_norm": 0.6331716179847717, "learning_rate": 9.645000000000001e-05, "loss": 0.0566, "step": 1930 }, { "grad_norm": 0.5065562129020691, "learning_rate": 9.695000000000001e-05, "loss": 0.0646, "step": 1940 }, { "grad_norm": 0.5008827447891235, "learning_rate": 9.745000000000001e-05, "loss": 0.0531, "step": 1950 }, { "grad_norm": 0.4982711374759674, "learning_rate": 9.795000000000001e-05, "loss": 0.052, "step": 1960 }, { "grad_norm": 0.7280076146125793, "learning_rate": 9.845000000000001e-05, "loss": 0.0547, "step": 1970 }, { "grad_norm": 0.6335408091545105, "learning_rate": 9.895e-05, "loss": 0.0624, "step": 1980 }, { "grad_norm": 0.6593841314315796, "learning_rate": 9.945e-05, "loss": 0.055, "step": 1990 }, { "grad_norm": 0.5683601498603821, "learning_rate": 9.995e-05, "loss": 0.0544, "step": 2000 }, { "grad_norm": 0.5583152770996094, "learning_rate": 9.99999861593158e-05, "loss": 0.0542, "step": 2010 }, { "grad_norm": 0.6122778654098511, "learning_rate": 9.999993831498517e-05, "loss": 0.0572, "step": 2020 }, { "grad_norm": 0.5194892287254333, "learning_rate": 9.99998562961682e-05, "loss": 0.0507, "step": 2030 }, { "grad_norm": 0.5109156966209412, "learning_rate": 9.99997401029209e-05, "loss": 0.0554, "step": 2040 }, { "grad_norm": 0.6974542140960693, "learning_rate": 9.999958973532271e-05, "loss": 0.0566, "step": 2050 }, { "grad_norm": 0.6491975784301758, "learning_rate": 9.999940519347642e-05, "loss": 0.0543, "step": 2060 }, { "grad_norm": 0.6925873160362244, "learning_rate": 9.999918647750812e-05, "loss": 0.0544, "step": 2070 }, { "grad_norm": 0.5563998818397522, "learning_rate": 9.999893358756736e-05, "loss": 0.0556, "step": 2080 }, { "grad_norm": 0.5706394910812378, "learning_rate": 9.999864652382692e-05, "loss": 0.0558, "step": 2090 }, { "grad_norm": 0.6373856663703918, "learning_rate": 9.999832528648307e-05, "loss": 0.0533, "step": 2100 }, { "grad_norm": 0.4866330325603485, "learning_rate": 9.999796987575532e-05, "loss": 0.0548, "step": 2110 }, { "grad_norm": 0.5110390186309814, "learning_rate": 9.999758029188662e-05, "loss": 0.0534, "step": 2120 }, { "grad_norm": 0.6468943357467651, "learning_rate": 9.999715653514324e-05, "loss": 0.0579, "step": 2130 }, { "grad_norm": 0.4656778573989868, "learning_rate": 9.999669860581482e-05, "loss": 0.0529, "step": 2140 }, { "grad_norm": 0.6737771034240723, "learning_rate": 9.999620650421433e-05, "loss": 0.0529, "step": 2150 }, { "grad_norm": 0.5764867067337036, "learning_rate": 9.999568023067813e-05, "loss": 0.0454, "step": 2160 }, { "grad_norm": 0.5284596681594849, "learning_rate": 9.999511978556595e-05, "loss": 0.0495, "step": 2170 }, { "grad_norm": 0.6407092213630676, "learning_rate": 9.99945251692608e-05, "loss": 0.0531, "step": 2180 }, { "grad_norm": 0.496069997549057, "learning_rate": 9.999389638216912e-05, "loss": 0.0528, "step": 2190 }, { "grad_norm": 0.47385409474372864, "learning_rate": 9.999323342472068e-05, "loss": 0.0468, "step": 2200 }, { "grad_norm": 0.4711074233055115, "learning_rate": 9.99925362973686e-05, "loss": 0.0453, "step": 2210 }, { "grad_norm": 0.6398211121559143, "learning_rate": 9.999180500058935e-05, "loss": 0.0467, "step": 2220 }, { "grad_norm": 0.5827088952064514, "learning_rate": 9.99910395348828e-05, "loss": 0.0513, "step": 2230 }, { "grad_norm": 0.45031294226646423, "learning_rate": 9.99902399007721e-05, "loss": 0.0525, "step": 2240 }, { "grad_norm": 0.465216726064682, "learning_rate": 9.99894060988038e-05, "loss": 0.0574, "step": 2250 }, { "grad_norm": 0.5349944233894348, "learning_rate": 9.998853812954783e-05, "loss": 0.0561, "step": 2260 }, { "grad_norm": 0.393606036901474, "learning_rate": 9.998763599359739e-05, "loss": 0.056, "step": 2270 }, { "grad_norm": 0.4193604290485382, "learning_rate": 9.998669969156912e-05, "loss": 0.0488, "step": 2280 }, { "grad_norm": 0.45064371824264526, "learning_rate": 9.998572922410294e-05, "loss": 0.0439, "step": 2290 }, { "grad_norm": 0.6237956285476685, "learning_rate": 9.998472459186219e-05, "loss": 0.0492, "step": 2300 }, { "grad_norm": 0.34804174304008484, "learning_rate": 9.99836857955335e-05, "loss": 0.0517, "step": 2310 }, { "grad_norm": 0.6358661651611328, "learning_rate": 9.998261283582688e-05, "loss": 0.0471, "step": 2320 }, { "grad_norm": 0.4673178195953369, "learning_rate": 9.998150571347571e-05, "loss": 0.0535, "step": 2330 }, { "grad_norm": 0.574245810508728, "learning_rate": 9.998036442923667e-05, "loss": 0.0517, "step": 2340 }, { "grad_norm": 0.5474517345428467, "learning_rate": 9.997918898388983e-05, "loss": 0.0467, "step": 2350 }, { "grad_norm": 0.6009110808372498, "learning_rate": 9.997797937823859e-05, "loss": 0.0531, "step": 2360 }, { "grad_norm": 0.49925968050956726, "learning_rate": 9.997673561310971e-05, "loss": 0.0516, "step": 2370 }, { "grad_norm": 0.5288291573524475, "learning_rate": 9.997545768935331e-05, "loss": 0.0522, "step": 2380 }, { "grad_norm": 0.4994864761829376, "learning_rate": 9.997414560784282e-05, "loss": 0.0505, "step": 2390 }, { "grad_norm": 0.5867973566055298, "learning_rate": 9.997279936947502e-05, "loss": 0.0507, "step": 2400 }, { "grad_norm": 0.5132080316543579, "learning_rate": 9.997141897517006e-05, "loss": 0.0464, "step": 2410 }, { "grad_norm": 0.4675925374031067, "learning_rate": 9.997000442587145e-05, "loss": 0.047, "step": 2420 }, { "grad_norm": 0.5084108710289001, "learning_rate": 9.9968555722546e-05, "loss": 0.0479, "step": 2430 }, { "grad_norm": 0.55105060338974, "learning_rate": 9.996707286618387e-05, "loss": 0.0471, "step": 2440 }, { "grad_norm": 0.5624989867210388, "learning_rate": 9.996555585779861e-05, "loss": 0.0476, "step": 2450 }, { "grad_norm": 0.38945266604423523, "learning_rate": 9.996400469842708e-05, "loss": 0.0514, "step": 2460 }, { "grad_norm": 0.42426618933677673, "learning_rate": 9.996241938912945e-05, "loss": 0.0438, "step": 2470 }, { "grad_norm": 0.47934967279434204, "learning_rate": 9.99607999309893e-05, "loss": 0.0521, "step": 2480 }, { "grad_norm": 0.6640028357505798, "learning_rate": 9.99591463251135e-05, "loss": 0.0479, "step": 2490 }, { "grad_norm": 0.5095248222351074, "learning_rate": 9.995745857263226e-05, "loss": 0.0504, "step": 2500 }, { "grad_norm": 0.5072497129440308, "learning_rate": 9.995573667469918e-05, "loss": 0.0503, "step": 2510 }, { "grad_norm": 0.5451476573944092, "learning_rate": 9.995398063249111e-05, "loss": 0.0482, "step": 2520 }, { "grad_norm": 0.44531258940696716, "learning_rate": 9.995219044720833e-05, "loss": 0.0452, "step": 2530 }, { "grad_norm": 0.45352044701576233, "learning_rate": 9.99503661200744e-05, "loss": 0.0462, "step": 2540 }, { "grad_norm": 0.5263320803642273, "learning_rate": 9.994850765233624e-05, "loss": 0.0519, "step": 2550 }, { "grad_norm": 0.4902527630329132, "learning_rate": 9.994661504526406e-05, "loss": 0.0465, "step": 2560 }, { "grad_norm": 0.45006585121154785, "learning_rate": 9.994468830015148e-05, "loss": 0.0448, "step": 2570 }, { "grad_norm": 0.5068536400794983, "learning_rate": 9.99427274183154e-05, "loss": 0.0497, "step": 2580 }, { "grad_norm": 0.4662571847438812, "learning_rate": 9.994073240109606e-05, "loss": 0.0451, "step": 2590 }, { "grad_norm": 0.5326297283172607, "learning_rate": 9.993870324985703e-05, "loss": 0.0481, "step": 2600 }, { "grad_norm": 0.5748021602630615, "learning_rate": 9.993663996598523e-05, "loss": 0.0482, "step": 2610 }, { "grad_norm": 0.5121206641197205, "learning_rate": 9.993454255089089e-05, "loss": 0.0444, "step": 2620 }, { "grad_norm": 0.5121270418167114, "learning_rate": 9.993241100600757e-05, "loss": 0.0514, "step": 2630 }, { "grad_norm": 0.5291215181350708, "learning_rate": 9.993024533279215e-05, "loss": 0.0486, "step": 2640 }, { "grad_norm": 0.44423821568489075, "learning_rate": 9.992804553272487e-05, "loss": 0.0486, "step": 2650 }, { "grad_norm": 0.4990832209587097, "learning_rate": 9.992581160730926e-05, "loss": 0.0472, "step": 2660 }, { "grad_norm": 0.36550381779670715, "learning_rate": 9.992354355807218e-05, "loss": 0.0491, "step": 2670 }, { "grad_norm": 0.4329356849193573, "learning_rate": 9.992124138656385e-05, "loss": 0.0436, "step": 2680 }, { "grad_norm": 0.3941628932952881, "learning_rate": 9.991890509435774e-05, "loss": 0.0472, "step": 2690 }, { "grad_norm": 0.4875953197479248, "learning_rate": 9.991653468305071e-05, "loss": 0.0462, "step": 2700 }, { "grad_norm": 0.46691417694091797, "learning_rate": 9.991413015426292e-05, "loss": 0.0508, "step": 2710 }, { "grad_norm": 0.4404756724834442, "learning_rate": 9.991169150963781e-05, "loss": 0.0439, "step": 2720 }, { "grad_norm": 0.6306264996528625, "learning_rate": 9.990921875084221e-05, "loss": 0.0476, "step": 2730 }, { "grad_norm": 0.49788156151771545, "learning_rate": 9.990671187956622e-05, "loss": 0.0493, "step": 2740 }, { "grad_norm": 0.4824599623680115, "learning_rate": 9.990417089752324e-05, "loss": 0.0486, "step": 2750 }, { "grad_norm": 0.5095981955528259, "learning_rate": 9.990159580645001e-05, "loss": 0.0477, "step": 2760 }, { "grad_norm": 0.5280710458755493, "learning_rate": 9.98989866081066e-05, "loss": 0.0404, "step": 2770 }, { "grad_norm": 0.4050522744655609, "learning_rate": 9.989634330427636e-05, "loss": 0.0461, "step": 2780 }, { "grad_norm": 0.3820000886917114, "learning_rate": 9.989366589676596e-05, "loss": 0.0467, "step": 2790 }, { "grad_norm": 0.5190500617027283, "learning_rate": 9.98909543874054e-05, "loss": 0.0435, "step": 2800 }, { "grad_norm": 0.342467725276947, "learning_rate": 9.988820877804795e-05, "loss": 0.0411, "step": 2810 }, { "grad_norm": 0.3889336585998535, "learning_rate": 9.988542907057022e-05, "loss": 0.0392, "step": 2820 }, { "grad_norm": 0.4009574353694916, "learning_rate": 9.988261526687212e-05, "loss": 0.0438, "step": 2830 }, { "grad_norm": 0.3832969665527344, "learning_rate": 9.987976736887685e-05, "loss": 0.0478, "step": 2840 }, { "grad_norm": 0.2997523844242096, "learning_rate": 9.987688537853091e-05, "loss": 0.0419, "step": 2850 }, { "grad_norm": 0.46297094225883484, "learning_rate": 9.987396929780413e-05, "loss": 0.0411, "step": 2860 }, { "grad_norm": 0.4701792895793915, "learning_rate": 9.987101912868962e-05, "loss": 0.0426, "step": 2870 }, { "grad_norm": 0.39719727635383606, "learning_rate": 9.98680348732038e-05, "loss": 0.0406, "step": 2880 }, { "grad_norm": 0.3401271104812622, "learning_rate": 9.986501653338636e-05, "loss": 0.0393, "step": 2890 }, { "grad_norm": 0.5679183006286621, "learning_rate": 9.986196411130031e-05, "loss": 0.0429, "step": 2900 }, { "grad_norm": 0.41431841254234314, "learning_rate": 9.985887760903197e-05, "loss": 0.0408, "step": 2910 }, { "grad_norm": 0.43026959896087646, "learning_rate": 9.985575702869093e-05, "loss": 0.0403, "step": 2920 }, { "grad_norm": 0.47557464241981506, "learning_rate": 9.985260237241008e-05, "loss": 0.0477, "step": 2930 }, { "grad_norm": 0.5165030360221863, "learning_rate": 9.984941364234557e-05, "loss": 0.0489, "step": 2940 }, { "grad_norm": 0.3893532156944275, "learning_rate": 9.98461908406769e-05, "loss": 0.047, "step": 2950 }, { "grad_norm": 0.4213593602180481, "learning_rate": 9.98429339696068e-05, "loss": 0.0502, "step": 2960 }, { "grad_norm": 0.478645920753479, "learning_rate": 9.983964303136133e-05, "loss": 0.044, "step": 2970 }, { "grad_norm": 0.4124567210674286, "learning_rate": 9.983631802818981e-05, "loss": 0.0423, "step": 2980 }, { "grad_norm": 0.33617475628852844, "learning_rate": 9.983295896236484e-05, "loss": 0.0416, "step": 2990 }, { "grad_norm": 0.32510945200920105, "learning_rate": 9.982956583618232e-05, "loss": 0.0426, "step": 3000 }, { "grad_norm": 0.5893015265464783, "learning_rate": 9.982613865196142e-05, "loss": 0.0493, "step": 3010 }, { "grad_norm": 0.3509642779827118, "learning_rate": 9.982267741204458e-05, "loss": 0.0426, "step": 3020 }, { "grad_norm": 0.3222624659538269, "learning_rate": 9.981918211879753e-05, "loss": 0.041, "step": 3030 }, { "grad_norm": 0.4265778064727783, "learning_rate": 9.981565277460927e-05, "loss": 0.0389, "step": 3040 }, { "grad_norm": 0.438810259103775, "learning_rate": 9.981208938189206e-05, "loss": 0.0415, "step": 3050 }, { "grad_norm": 0.574134111404419, "learning_rate": 9.980849194308149e-05, "loss": 0.0398, "step": 3060 }, { "grad_norm": 0.427533358335495, "learning_rate": 9.98048604606363e-05, "loss": 0.0423, "step": 3070 }, { "grad_norm": 0.3630805015563965, "learning_rate": 9.980119493703864e-05, "loss": 0.0417, "step": 3080 }, { "grad_norm": 0.3512152433395386, "learning_rate": 9.979749537479383e-05, "loss": 0.0439, "step": 3090 }, { "grad_norm": 0.44711706042289734, "learning_rate": 9.979376177643051e-05, "loss": 0.0472, "step": 3100 }, { "grad_norm": 0.4862109124660492, "learning_rate": 9.978999414450052e-05, "loss": 0.0488, "step": 3110 }, { "grad_norm": 0.405716210603714, "learning_rate": 9.978619248157904e-05, "loss": 0.0389, "step": 3120 }, { "grad_norm": 0.54551762342453, "learning_rate": 9.978235679026445e-05, "loss": 0.0488, "step": 3130 }, { "grad_norm": 0.45173701643943787, "learning_rate": 9.97784870731784e-05, "loss": 0.0409, "step": 3140 }, { "grad_norm": 0.4855295419692993, "learning_rate": 9.977458333296584e-05, "loss": 0.0473, "step": 3150 }, { "grad_norm": 0.4203748106956482, "learning_rate": 9.977064557229492e-05, "loss": 0.0376, "step": 3160 }, { "grad_norm": 0.3028692603111267, "learning_rate": 9.976667379385705e-05, "loss": 0.0417, "step": 3170 }, { "grad_norm": 0.3870784342288971, "learning_rate": 9.97626680003669e-05, "loss": 0.0429, "step": 3180 }, { "grad_norm": 0.5136834383010864, "learning_rate": 9.975862819456242e-05, "loss": 0.0456, "step": 3190 }, { "grad_norm": 0.44241195917129517, "learning_rate": 9.975455437920477e-05, "loss": 0.0406, "step": 3200 }, { "grad_norm": 0.3893604874610901, "learning_rate": 9.975044655707834e-05, "loss": 0.0382, "step": 3210 }, { "grad_norm": 0.41208815574645996, "learning_rate": 9.974630473099082e-05, "loss": 0.0443, "step": 3220 }, { "grad_norm": 0.4270910322666168, "learning_rate": 9.974212890377311e-05, "loss": 0.0486, "step": 3230 }, { "grad_norm": 0.33294859528541565, "learning_rate": 9.973791907827931e-05, "loss": 0.0415, "step": 3240 }, { "grad_norm": 0.4029678404331207, "learning_rate": 9.973367525738683e-05, "loss": 0.042, "step": 3250 }, { "grad_norm": 0.4246334731578827, "learning_rate": 9.972939744399627e-05, "loss": 0.0435, "step": 3260 }, { "grad_norm": 0.4796226918697357, "learning_rate": 9.97250856410315e-05, "loss": 0.0431, "step": 3270 }, { "grad_norm": 0.456011027097702, "learning_rate": 9.972073985143955e-05, "loss": 0.0374, "step": 3280 }, { "grad_norm": 0.4488547146320343, "learning_rate": 9.971636007819074e-05, "loss": 0.0425, "step": 3290 }, { "grad_norm": 0.47248077392578125, "learning_rate": 9.971194632427863e-05, "loss": 0.0444, "step": 3300 }, { "grad_norm": 0.435719758272171, "learning_rate": 9.970749859271997e-05, "loss": 0.0458, "step": 3310 }, { "grad_norm": 0.35676801204681396, "learning_rate": 9.970301688655473e-05, "loss": 0.0463, "step": 3320 }, { "grad_norm": 0.4043083190917969, "learning_rate": 9.969850120884612e-05, "loss": 0.0428, "step": 3330 }, { "grad_norm": 0.4107603430747986, "learning_rate": 9.969395156268055e-05, "loss": 0.0436, "step": 3340 }, { "grad_norm": 0.4701997935771942, "learning_rate": 9.968936795116768e-05, "loss": 0.0439, "step": 3350 }, { "grad_norm": 0.36342141032218933, "learning_rate": 9.968475037744036e-05, "loss": 0.0393, "step": 3360 }, { "grad_norm": 0.29374411702156067, "learning_rate": 9.968009884465465e-05, "loss": 0.0398, "step": 3370 }, { "grad_norm": 0.43677929043769836, "learning_rate": 9.967541335598984e-05, "loss": 0.0392, "step": 3380 }, { "grad_norm": 0.39345496892929077, "learning_rate": 9.967069391464841e-05, "loss": 0.0432, "step": 3390 }, { "grad_norm": 0.3908386528491974, "learning_rate": 9.966594052385608e-05, "loss": 0.0446, "step": 3400 }, { "grad_norm": 0.43464964628219604, "learning_rate": 9.96611531868617e-05, "loss": 0.0457, "step": 3410 }, { "grad_norm": 0.4960820972919464, "learning_rate": 9.96563319069374e-05, "loss": 0.0402, "step": 3420 }, { "grad_norm": 0.42703381180763245, "learning_rate": 9.965147668737847e-05, "loss": 0.0386, "step": 3430 }, { "grad_norm": 0.5438716411590576, "learning_rate": 9.96465875315034e-05, "loss": 0.0447, "step": 3440 }, { "grad_norm": 0.3769811987876892, "learning_rate": 9.96416644426539e-05, "loss": 0.036, "step": 3450 }, { "grad_norm": 0.4758771061897278, "learning_rate": 9.963670742419485e-05, "loss": 0.0422, "step": 3460 }, { "grad_norm": 0.35315006971359253, "learning_rate": 9.96317164795143e-05, "loss": 0.0454, "step": 3470 }, { "grad_norm": 0.42954620718955994, "learning_rate": 9.962669161202356e-05, "loss": 0.0411, "step": 3480 }, { "grad_norm": 0.3790585994720459, "learning_rate": 9.962163282515705e-05, "loss": 0.0418, "step": 3490 }, { "grad_norm": 0.4889375865459442, "learning_rate": 9.961654012237241e-05, "loss": 0.0388, "step": 3500 }, { "grad_norm": 0.43765440583229065, "learning_rate": 9.961141350715044e-05, "loss": 0.0448, "step": 3510 }, { "grad_norm": 0.45030760765075684, "learning_rate": 9.960625298299516e-05, "loss": 0.0419, "step": 3520 }, { "grad_norm": 0.3697926998138428, "learning_rate": 9.960105855343372e-05, "loss": 0.0413, "step": 3530 }, { "grad_norm": 0.37472888827323914, "learning_rate": 9.959583022201647e-05, "loss": 0.0401, "step": 3540 }, { "grad_norm": 0.3540658950805664, "learning_rate": 9.959056799231692e-05, "loss": 0.0391, "step": 3550 }, { "grad_norm": 0.5391460657119751, "learning_rate": 9.958527186793176e-05, "loss": 0.0408, "step": 3560 }, { "grad_norm": 0.3354576528072357, "learning_rate": 9.957994185248086e-05, "loss": 0.0405, "step": 3570 }, { "grad_norm": 0.41552233695983887, "learning_rate": 9.957457794960718e-05, "loss": 0.0409, "step": 3580 }, { "grad_norm": 0.3863587975502014, "learning_rate": 9.956918016297694e-05, "loss": 0.0382, "step": 3590 }, { "grad_norm": 0.5188900828361511, "learning_rate": 9.956374849627948e-05, "loss": 0.0391, "step": 3600 }, { "grad_norm": 0.329557329416275, "learning_rate": 9.955828295322728e-05, "loss": 0.0413, "step": 3610 }, { "grad_norm": 0.3678283989429474, "learning_rate": 9.955278353755598e-05, "loss": 0.0414, "step": 3620 }, { "grad_norm": 0.4326399564743042, "learning_rate": 9.954725025302439e-05, "loss": 0.0443, "step": 3630 }, { "grad_norm": 0.4455494284629822, "learning_rate": 9.954168310341445e-05, "loss": 0.039, "step": 3640 }, { "grad_norm": 0.35780996084213257, "learning_rate": 9.953608209253126e-05, "loss": 0.0367, "step": 3650 }, { "grad_norm": 0.4393651783466339, "learning_rate": 9.953044722420307e-05, "loss": 0.037, "step": 3660 }, { "grad_norm": 0.3539571762084961, "learning_rate": 9.952477850228124e-05, "loss": 0.0368, "step": 3670 }, { "grad_norm": 0.3123975694179535, "learning_rate": 9.95190759306403e-05, "loss": 0.0374, "step": 3680 }, { "grad_norm": 0.4090181589126587, "learning_rate": 9.951333951317789e-05, "loss": 0.045, "step": 3690 }, { "grad_norm": 0.5658549666404724, "learning_rate": 9.950756925381479e-05, "loss": 0.0361, "step": 3700 }, { "grad_norm": 0.4106267988681793, "learning_rate": 9.950176515649496e-05, "loss": 0.0388, "step": 3710 }, { "grad_norm": 0.35073405504226685, "learning_rate": 9.949592722518542e-05, "loss": 0.0399, "step": 3720 }, { "grad_norm": 0.34277796745300293, "learning_rate": 9.949005546387631e-05, "loss": 0.0405, "step": 3730 }, { "grad_norm": 0.33378249406814575, "learning_rate": 9.948414987658098e-05, "loss": 0.0415, "step": 3740 }, { "grad_norm": 0.29598870873451233, "learning_rate": 9.94782104673358e-05, "loss": 0.039, "step": 3750 }, { "grad_norm": 0.39936327934265137, "learning_rate": 9.947223724020034e-05, "loss": 0.0423, "step": 3760 }, { "grad_norm": 0.3253559470176697, "learning_rate": 9.94662301992572e-05, "loss": 0.0324, "step": 3770 }, { "grad_norm": 0.4331583082675934, "learning_rate": 9.946018934861216e-05, "loss": 0.0403, "step": 3780 }, { "grad_norm": 0.3425239622592926, "learning_rate": 9.945411469239409e-05, "loss": 0.0374, "step": 3790 }, { "grad_norm": 0.3541868031024933, "learning_rate": 9.944800623475497e-05, "loss": 0.0411, "step": 3800 }, { "grad_norm": 0.4586838185787201, "learning_rate": 9.944186397986984e-05, "loss": 0.0432, "step": 3810 }, { "grad_norm": 0.35010725259780884, "learning_rate": 9.943568793193687e-05, "loss": 0.0467, "step": 3820 }, { "grad_norm": 0.2609131634235382, "learning_rate": 9.942947809517737e-05, "loss": 0.0356, "step": 3830 }, { "grad_norm": 0.3281535804271698, "learning_rate": 9.942323447383568e-05, "loss": 0.0365, "step": 3840 }, { "grad_norm": 0.41413575410842896, "learning_rate": 9.941695707217925e-05, "loss": 0.0394, "step": 3850 }, { "grad_norm": 0.42927926778793335, "learning_rate": 9.941064589449865e-05, "loss": 0.0409, "step": 3860 }, { "grad_norm": 0.42291101813316345, "learning_rate": 9.940430094510748e-05, "loss": 0.0361, "step": 3870 }, { "grad_norm": 0.36126959323883057, "learning_rate": 9.939792222834248e-05, "loss": 0.0389, "step": 3880 }, { "grad_norm": 0.4104115962982178, "learning_rate": 9.939150974856343e-05, "loss": 0.043, "step": 3890 }, { "grad_norm": 0.29402396082878113, "learning_rate": 9.93850635101532e-05, "loss": 0.0447, "step": 3900 }, { "grad_norm": 0.4395533502101898, "learning_rate": 9.937858351751775e-05, "loss": 0.0394, "step": 3910 }, { "grad_norm": 0.4372180998325348, "learning_rate": 9.937206977508604e-05, "loss": 0.0467, "step": 3920 }, { "grad_norm": 0.38096633553504944, "learning_rate": 9.936552228731022e-05, "loss": 0.0361, "step": 3930 }, { "grad_norm": 0.3380042314529419, "learning_rate": 9.93589410586654e-05, "loss": 0.0401, "step": 3940 }, { "grad_norm": 0.4042423963546753, "learning_rate": 9.93523260936498e-05, "loss": 0.0398, "step": 3950 }, { "grad_norm": 0.3645128607749939, "learning_rate": 9.934567739678467e-05, "loss": 0.0386, "step": 3960 }, { "grad_norm": 0.3448069095611572, "learning_rate": 9.933899497261433e-05, "loss": 0.0355, "step": 3970 }, { "grad_norm": 0.3040964901447296, "learning_rate": 9.933227882570617e-05, "loss": 0.0342, "step": 3980 }, { "grad_norm": 0.37062138319015503, "learning_rate": 9.932552896065062e-05, "loss": 0.0365, "step": 3990 }, { "grad_norm": 0.42263221740722656, "learning_rate": 9.931874538206114e-05, "loss": 0.0396, "step": 4000 }, { "grad_norm": 0.40672457218170166, "learning_rate": 9.931192809457423e-05, "loss": 0.0353, "step": 4010 }, { "grad_norm": 0.3535928726196289, "learning_rate": 9.930507710284944e-05, "loss": 0.0354, "step": 4020 }, { "grad_norm": 0.3491552174091339, "learning_rate": 9.92981924115694e-05, "loss": 0.0386, "step": 4030 }, { "grad_norm": 0.33314934372901917, "learning_rate": 9.929127402543968e-05, "loss": 0.0362, "step": 4040 }, { "grad_norm": 0.4575192928314209, "learning_rate": 9.928432194918895e-05, "loss": 0.0387, "step": 4050 }, { "grad_norm": 0.35964757204055786, "learning_rate": 9.927733618756889e-05, "loss": 0.0373, "step": 4060 }, { "grad_norm": 0.4931086599826813, "learning_rate": 9.92703167453542e-05, "loss": 0.0337, "step": 4070 }, { "grad_norm": 0.3923843502998352, "learning_rate": 9.926326362734263e-05, "loss": 0.0389, "step": 4080 }, { "grad_norm": 0.30370762944221497, "learning_rate": 9.925617683835489e-05, "loss": 0.0357, "step": 4090 }, { "grad_norm": 0.26706552505493164, "learning_rate": 9.924905638323472e-05, "loss": 0.0336, "step": 4100 }, { "grad_norm": 0.3393687903881073, "learning_rate": 9.92419022668489e-05, "loss": 0.037, "step": 4110 }, { "grad_norm": 0.3864586055278778, "learning_rate": 9.923471449408723e-05, "loss": 0.0394, "step": 4120 }, { "grad_norm": 0.36737120151519775, "learning_rate": 9.922749306986244e-05, "loss": 0.0405, "step": 4130 }, { "grad_norm": 0.33752137422561646, "learning_rate": 9.922023799911033e-05, "loss": 0.0344, "step": 4140 }, { "grad_norm": 0.3143002986907959, "learning_rate": 9.921294928678968e-05, "loss": 0.0381, "step": 4150 }, { "grad_norm": 0.334157258272171, "learning_rate": 9.920562693788222e-05, "loss": 0.0346, "step": 4160 }, { "grad_norm": 0.3777976930141449, "learning_rate": 9.919827095739273e-05, "loss": 0.0388, "step": 4170 }, { "grad_norm": 0.5512045621871948, "learning_rate": 9.919088135034898e-05, "loss": 0.0376, "step": 4180 }, { "grad_norm": 0.3319365382194519, "learning_rate": 9.918345812180165e-05, "loss": 0.0441, "step": 4190 }, { "grad_norm": 0.36799919605255127, "learning_rate": 9.917600127682446e-05, "loss": 0.0379, "step": 4200 }, { "grad_norm": 0.3646402657032013, "learning_rate": 9.91685108205141e-05, "loss": 0.0369, "step": 4210 }, { "grad_norm": 0.38940131664276123, "learning_rate": 9.916098675799024e-05, "loss": 0.0366, "step": 4220 }, { "grad_norm": 0.32485100626945496, "learning_rate": 9.915342909439548e-05, "loss": 0.0365, "step": 4230 }, { "grad_norm": 0.43473953008651733, "learning_rate": 9.914583783489543e-05, "loss": 0.0366, "step": 4240 }, { "grad_norm": 0.3093690276145935, "learning_rate": 9.913821298467863e-05, "loss": 0.037, "step": 4250 }, { "grad_norm": 0.43008124828338623, "learning_rate": 9.913055454895661e-05, "loss": 0.0377, "step": 4260 }, { "grad_norm": 0.4347173273563385, "learning_rate": 9.912286253296381e-05, "loss": 0.039, "step": 4270 }, { "grad_norm": 0.3867764174938202, "learning_rate": 9.911513694195768e-05, "loss": 0.0379, "step": 4280 }, { "grad_norm": 0.2747771143913269, "learning_rate": 9.910737778121859e-05, "loss": 0.0336, "step": 4290 }, { "grad_norm": 0.4684167802333832, "learning_rate": 9.909958505604984e-05, "loss": 0.0341, "step": 4300 }, { "grad_norm": 0.43003231287002563, "learning_rate": 9.909175877177767e-05, "loss": 0.0362, "step": 4310 }, { "grad_norm": 0.3280821740627289, "learning_rate": 9.908389893375129e-05, "loss": 0.0381, "step": 4320 }, { "grad_norm": 0.29483285546302795, "learning_rate": 9.907600554734283e-05, "loss": 0.0358, "step": 4330 }, { "grad_norm": 0.2873283922672272, "learning_rate": 9.906807861794734e-05, "loss": 0.0385, "step": 4340 }, { "grad_norm": 0.38026294112205505, "learning_rate": 9.906011815098279e-05, "loss": 0.0403, "step": 4350 }, { "grad_norm": 0.38237181305885315, "learning_rate": 9.90521241518901e-05, "loss": 0.0343, "step": 4360 }, { "grad_norm": 0.3906157910823822, "learning_rate": 9.904409662613308e-05, "loss": 0.037, "step": 4370 }, { "grad_norm": 0.3297280967235565, "learning_rate": 9.903603557919849e-05, "loss": 0.0386, "step": 4380 }, { "grad_norm": 0.3545493185520172, "learning_rate": 9.902794101659594e-05, "loss": 0.0419, "step": 4390 }, { "grad_norm": 0.3200553357601166, "learning_rate": 9.901981294385803e-05, "loss": 0.0355, "step": 4400 }, { "grad_norm": 0.4041753113269806, "learning_rate": 9.901165136654018e-05, "loss": 0.0377, "step": 4410 }, { "grad_norm": 0.37479498982429504, "learning_rate": 9.900345629022079e-05, "loss": 0.0363, "step": 4420 }, { "grad_norm": 0.3027077317237854, "learning_rate": 9.89952277205011e-05, "loss": 0.0337, "step": 4430 }, { "grad_norm": 0.3918471932411194, "learning_rate": 9.898696566300527e-05, "loss": 0.0362, "step": 4440 }, { "grad_norm": 0.35986316204071045, "learning_rate": 9.897867012338032e-05, "loss": 0.0375, "step": 4450 }, { "grad_norm": 0.3680419623851776, "learning_rate": 9.897034110729617e-05, "loss": 0.0354, "step": 4460 }, { "grad_norm": 0.3550345301628113, "learning_rate": 9.896197862044564e-05, "loss": 0.0335, "step": 4470 }, { "grad_norm": 0.40165647864341736, "learning_rate": 9.895358266854442e-05, "loss": 0.0361, "step": 4480 }, { "grad_norm": 0.4391635060310364, "learning_rate": 9.894515325733103e-05, "loss": 0.0418, "step": 4490 }, { "grad_norm": 0.3594600260257721, "learning_rate": 9.893669039256693e-05, "loss": 0.0362, "step": 4500 }, { "grad_norm": 0.3988608121871948, "learning_rate": 9.89281940800364e-05, "loss": 0.0326, "step": 4510 }, { "grad_norm": 0.31035083532333374, "learning_rate": 9.891966432554655e-05, "loss": 0.0332, "step": 4520 }, { "grad_norm": 0.34515008330345154, "learning_rate": 9.891110113492745e-05, "loss": 0.0362, "step": 4530 }, { "grad_norm": 0.3963836133480072, "learning_rate": 9.89025045140319e-05, "loss": 0.0358, "step": 4540 }, { "grad_norm": 0.3310410976409912, "learning_rate": 9.889387446873567e-05, "loss": 0.0318, "step": 4550 }, { "grad_norm": 0.3400961458683014, "learning_rate": 9.888521100493726e-05, "loss": 0.0339, "step": 4560 }, { "grad_norm": 0.2290128469467163, "learning_rate": 9.887651412855809e-05, "loss": 0.0325, "step": 4570 }, { "grad_norm": 0.327698290348053, "learning_rate": 9.88677838455424e-05, "loss": 0.0383, "step": 4580 }, { "grad_norm": 0.4216516315937042, "learning_rate": 9.885902016185725e-05, "loss": 0.0386, "step": 4590 }, { "grad_norm": 0.42152732610702515, "learning_rate": 9.885022308349252e-05, "loss": 0.0347, "step": 4600 }, { "grad_norm": 0.5130208134651184, "learning_rate": 9.884139261646095e-05, "loss": 0.0401, "step": 4610 }, { "grad_norm": 0.3245272636413574, "learning_rate": 9.883252876679807e-05, "loss": 0.0348, "step": 4620 }, { "grad_norm": 0.46825337409973145, "learning_rate": 9.882363154056225e-05, "loss": 0.0362, "step": 4630 }, { "grad_norm": 0.2964421510696411, "learning_rate": 9.881470094383465e-05, "loss": 0.0371, "step": 4640 }, { "grad_norm": 0.32779255509376526, "learning_rate": 9.880573698271924e-05, "loss": 0.0353, "step": 4650 }, { "grad_norm": 0.31194373965263367, "learning_rate": 9.879673966334282e-05, "loss": 0.0367, "step": 4660 }, { "grad_norm": 0.4152418076992035, "learning_rate": 9.878770899185496e-05, "loss": 0.035, "step": 4670 }, { "grad_norm": 0.37284529209136963, "learning_rate": 9.877864497442804e-05, "loss": 0.0316, "step": 4680 }, { "grad_norm": 0.35191890597343445, "learning_rate": 9.876954761725723e-05, "loss": 0.0351, "step": 4690 }, { "grad_norm": 0.3765595555305481, "learning_rate": 9.876041692656052e-05, "loss": 0.0439, "step": 4700 }, { "grad_norm": 0.3873654305934906, "learning_rate": 9.87512529085786e-05, "loss": 0.0366, "step": 4710 }, { "grad_norm": 0.3539074957370758, "learning_rate": 9.874205556957503e-05, "loss": 0.0325, "step": 4720 }, { "grad_norm": 0.3480886220932007, "learning_rate": 9.873282491583608e-05, "loss": 0.0338, "step": 4730 }, { "grad_norm": 0.39521682262420654, "learning_rate": 9.872356095367084e-05, "loss": 0.0352, "step": 4740 }, { "grad_norm": 0.28991377353668213, "learning_rate": 9.87142636894111e-05, "loss": 0.0365, "step": 4750 }, { "grad_norm": 0.38692039251327515, "learning_rate": 9.870493312941148e-05, "loss": 0.0374, "step": 4760 }, { "grad_norm": 0.32139909267425537, "learning_rate": 9.869556928004933e-05, "loss": 0.0387, "step": 4770 }, { "grad_norm": 0.33928897976875305, "learning_rate": 9.868617214772476e-05, "loss": 0.0346, "step": 4780 }, { "grad_norm": 0.366608589887619, "learning_rate": 9.867674173886059e-05, "loss": 0.0378, "step": 4790 }, { "grad_norm": 0.26452842354774475, "learning_rate": 9.866727805990245e-05, "loss": 0.0361, "step": 4800 }, { "grad_norm": 0.4257228970527649, "learning_rate": 9.865778111731865e-05, "loss": 0.0312, "step": 4810 }, { "grad_norm": 0.29310479760169983, "learning_rate": 9.864825091760028e-05, "loss": 0.0341, "step": 4820 }, { "grad_norm": 0.3838813304901123, "learning_rate": 9.863868746726112e-05, "loss": 0.0321, "step": 4830 }, { "grad_norm": 0.30367690324783325, "learning_rate": 9.862909077283773e-05, "loss": 0.0338, "step": 4840 }, { "grad_norm": 0.27524229884147644, "learning_rate": 9.861946084088933e-05, "loss": 0.0338, "step": 4850 }, { "grad_norm": 0.34939301013946533, "learning_rate": 9.860979767799792e-05, "loss": 0.0317, "step": 4860 }, { "grad_norm": 0.34509697556495667, "learning_rate": 9.860010129076813e-05, "loss": 0.039, "step": 4870 }, { "grad_norm": 0.34553882479667664, "learning_rate": 9.859037168582741e-05, "loss": 0.0343, "step": 4880 }, { "grad_norm": 0.3702239394187927, "learning_rate": 9.85806088698258e-05, "loss": 0.0343, "step": 4890 }, { "grad_norm": 0.30264168977737427, "learning_rate": 9.857081284943612e-05, "loss": 0.0348, "step": 4900 }, { "grad_norm": 0.20421327650547028, "learning_rate": 9.856098363135385e-05, "loss": 0.0305, "step": 4910 }, { "grad_norm": 0.34878841042518616, "learning_rate": 9.855112122229717e-05, "loss": 0.0353, "step": 4920 }, { "grad_norm": 0.3703024089336395, "learning_rate": 9.854122562900696e-05, "loss": 0.0356, "step": 4930 }, { "grad_norm": 0.3724825978279114, "learning_rate": 9.853129685824673e-05, "loss": 0.0302, "step": 4940 }, { "grad_norm": 0.28980815410614014, "learning_rate": 9.852133491680273e-05, "loss": 0.0346, "step": 4950 }, { "grad_norm": 0.3057869076728821, "learning_rate": 9.851133981148385e-05, "loss": 0.0355, "step": 4960 }, { "grad_norm": 0.32194775342941284, "learning_rate": 9.850131154912164e-05, "loss": 0.0308, "step": 4970 }, { "grad_norm": 0.29713332653045654, "learning_rate": 9.849125013657031e-05, "loss": 0.0384, "step": 4980 }, { "grad_norm": 0.3187062740325928, "learning_rate": 9.848115558070676e-05, "loss": 0.0348, "step": 4990 }, { "grad_norm": 0.32051244378089905, "learning_rate": 9.847102788843055e-05, "loss": 0.0329, "step": 5000 }, { "grad_norm": 0.2737869918346405, "learning_rate": 9.84608670666638e-05, "loss": 0.0336, "step": 5010 }, { "grad_norm": 0.35246574878692627, "learning_rate": 9.845067312235138e-05, "loss": 0.0365, "step": 5020 }, { "grad_norm": 0.2579406499862671, "learning_rate": 9.844044606246074e-05, "loss": 0.0307, "step": 5030 }, { "grad_norm": 0.28639641404151917, "learning_rate": 9.843018589398199e-05, "loss": 0.031, "step": 5040 }, { "grad_norm": 0.3307488262653351, "learning_rate": 9.841989262392785e-05, "loss": 0.0351, "step": 5050 }, { "grad_norm": 0.3859657645225525, "learning_rate": 9.840956625933367e-05, "loss": 0.0363, "step": 5060 }, { "grad_norm": 0.3440208435058594, "learning_rate": 9.839920680725743e-05, "loss": 0.0333, "step": 5070 }, { "grad_norm": 0.28082558512687683, "learning_rate": 9.83888142747797e-05, "loss": 0.0313, "step": 5080 }, { "grad_norm": 0.3393252193927765, "learning_rate": 9.83783886690037e-05, "loss": 0.0347, "step": 5090 }, { "grad_norm": 0.3462052047252655, "learning_rate": 9.836792999705524e-05, "loss": 0.0357, "step": 5100 }, { "grad_norm": 0.33803021907806396, "learning_rate": 9.835743826608271e-05, "loss": 0.0336, "step": 5110 }, { "grad_norm": 0.3681130111217499, "learning_rate": 9.834691348325709e-05, "loss": 0.0306, "step": 5120 }, { "grad_norm": 0.3571389317512512, "learning_rate": 9.833635565577199e-05, "loss": 0.0317, "step": 5130 }, { "grad_norm": 0.3850052058696747, "learning_rate": 9.832576479084359e-05, "loss": 0.0349, "step": 5140 }, { "grad_norm": 0.31258249282836914, "learning_rate": 9.831514089571064e-05, "loss": 0.0332, "step": 5150 }, { "grad_norm": 0.23021695017814636, "learning_rate": 9.830448397763447e-05, "loss": 0.0337, "step": 5160 }, { "grad_norm": 0.3072669804096222, "learning_rate": 9.8293794043899e-05, "loss": 0.0296, "step": 5170 }, { "grad_norm": 0.3260138928890228, "learning_rate": 9.828307110181066e-05, "loss": 0.0285, "step": 5180 }, { "grad_norm": 0.3707156777381897, "learning_rate": 9.827231515869852e-05, "loss": 0.0338, "step": 5190 }, { "grad_norm": 0.327912300825119, "learning_rate": 9.826152622191413e-05, "loss": 0.0329, "step": 5200 }, { "grad_norm": 0.4181981682777405, "learning_rate": 9.825070429883167e-05, "loss": 0.0372, "step": 5210 }, { "grad_norm": 0.34173649549484253, "learning_rate": 9.82398493968478e-05, "loss": 0.0389, "step": 5220 }, { "grad_norm": 0.31742584705352783, "learning_rate": 9.822896152338173e-05, "loss": 0.0354, "step": 5230 }, { "grad_norm": 0.2796247601509094, "learning_rate": 9.821804068587523e-05, "loss": 0.0295, "step": 5240 }, { "grad_norm": 0.34349510073661804, "learning_rate": 9.820708689179259e-05, "loss": 0.0347, "step": 5250 }, { "grad_norm": 0.3799993097782135, "learning_rate": 9.819610014862063e-05, "loss": 0.0349, "step": 5260 }, { "grad_norm": 0.3046055734157562, "learning_rate": 9.818508046386868e-05, "loss": 0.0334, "step": 5270 }, { "grad_norm": 0.3102630078792572, "learning_rate": 9.817402784506859e-05, "loss": 0.0354, "step": 5280 }, { "grad_norm": 0.3256523311138153, "learning_rate": 9.816294229977472e-05, "loss": 0.0336, "step": 5290 }, { "grad_norm": 0.3114723265171051, "learning_rate": 9.815182383556394e-05, "loss": 0.0376, "step": 5300 }, { "grad_norm": 0.43364790081977844, "learning_rate": 9.81406724600356e-05, "loss": 0.0411, "step": 5310 }, { "grad_norm": 0.2520511746406555, "learning_rate": 9.812948818081161e-05, "loss": 0.0412, "step": 5320 }, { "grad_norm": 0.35789668560028076, "learning_rate": 9.811827100553623e-05, "loss": 0.0318, "step": 5330 }, { "grad_norm": 0.3829971253871918, "learning_rate": 9.810702094187638e-05, "loss": 0.0343, "step": 5340 }, { "grad_norm": 0.3212273418903351, "learning_rate": 9.809573799752135e-05, "loss": 0.0345, "step": 5350 }, { "grad_norm": 0.39963144063949585, "learning_rate": 9.808442218018288e-05, "loss": 0.0307, "step": 5360 }, { "grad_norm": 0.2841002643108368, "learning_rate": 9.807307349759527e-05, "loss": 0.0314, "step": 5370 }, { "grad_norm": 0.2998031675815582, "learning_rate": 9.806169195751525e-05, "loss": 0.0326, "step": 5380 }, { "grad_norm": 0.26443615555763245, "learning_rate": 9.805027756772194e-05, "loss": 0.0303, "step": 5390 }, { "grad_norm": 0.33047863841056824, "learning_rate": 9.803883033601702e-05, "loss": 0.0289, "step": 5400 }, { "grad_norm": 0.25059646368026733, "learning_rate": 9.802735027022453e-05, "loss": 0.0322, "step": 5410 }, { "grad_norm": 0.2384401261806488, "learning_rate": 9.8015837378191e-05, "loss": 0.0313, "step": 5420 }, { "grad_norm": 0.30502861738204956, "learning_rate": 9.80042916677854e-05, "loss": 0.0352, "step": 5430 }, { "grad_norm": 0.3657150864601135, "learning_rate": 9.799271314689908e-05, "loss": 0.0331, "step": 5440 }, { "grad_norm": 0.2849947214126587, "learning_rate": 9.798110182344588e-05, "loss": 0.0337, "step": 5450 }, { "grad_norm": 0.2707001566886902, "learning_rate": 9.796945770536204e-05, "loss": 0.0331, "step": 5460 }, { "grad_norm": 0.36579960584640503, "learning_rate": 9.795778080060615e-05, "loss": 0.0325, "step": 5470 }, { "grad_norm": 0.2983980178833008, "learning_rate": 9.79460711171593e-05, "loss": 0.0339, "step": 5480 }, { "grad_norm": 0.27584290504455566, "learning_rate": 9.793432866302497e-05, "loss": 0.0323, "step": 5490 }, { "grad_norm": 0.3393056392669678, "learning_rate": 9.792255344622897e-05, "loss": 0.0311, "step": 5500 }, { "grad_norm": 0.3579215109348297, "learning_rate": 9.791074547481957e-05, "loss": 0.0348, "step": 5510 }, { "grad_norm": 0.34067314863204956, "learning_rate": 9.789890475686742e-05, "loss": 0.0346, "step": 5520 }, { "grad_norm": 0.3707297444343567, "learning_rate": 9.788703130046552e-05, "loss": 0.0329, "step": 5530 }, { "grad_norm": 0.36883386969566345, "learning_rate": 9.78751251137293e-05, "loss": 0.0316, "step": 5540 }, { "grad_norm": 0.3442963659763336, "learning_rate": 9.786318620479646e-05, "loss": 0.0319, "step": 5550 }, { "grad_norm": 0.387145072221756, "learning_rate": 9.78512145818272e-05, "loss": 0.0322, "step": 5560 }, { "grad_norm": 0.3218199908733368, "learning_rate": 9.783921025300398e-05, "loss": 0.0323, "step": 5570 }, { "grad_norm": 0.3681231141090393, "learning_rate": 9.782717322653164e-05, "loss": 0.0283, "step": 5580 }, { "grad_norm": 0.26863422989845276, "learning_rate": 9.781510351063738e-05, "loss": 0.0314, "step": 5590 }, { "grad_norm": 0.3015608489513397, "learning_rate": 9.780300111357075e-05, "loss": 0.0312, "step": 5600 }, { "grad_norm": 0.2962975800037384, "learning_rate": 9.779086604360361e-05, "loss": 0.0303, "step": 5610 }, { "grad_norm": 0.3451528251171112, "learning_rate": 9.777869830903019e-05, "loss": 0.0307, "step": 5620 }, { "grad_norm": 0.34626367688179016, "learning_rate": 9.776649791816698e-05, "loss": 0.0291, "step": 5630 }, { "grad_norm": 0.30702656507492065, "learning_rate": 9.775426487935285e-05, "loss": 0.0294, "step": 5640 }, { "grad_norm": 0.25669342279434204, "learning_rate": 9.774199920094898e-05, "loss": 0.0285, "step": 5650 }, { "grad_norm": 0.3448568880558014, "learning_rate": 9.772970089133884e-05, "loss": 0.0376, "step": 5660 }, { "grad_norm": 0.3759368360042572, "learning_rate": 9.77173699589282e-05, "loss": 0.0297, "step": 5670 }, { "grad_norm": 0.36348339915275574, "learning_rate": 9.770500641214513e-05, "loss": 0.0313, "step": 5680 }, { "grad_norm": 0.27063438296318054, "learning_rate": 9.769261025944003e-05, "loss": 0.0285, "step": 5690 }, { "grad_norm": 0.2572079598903656, "learning_rate": 9.768018150928552e-05, "loss": 0.0314, "step": 5700 }, { "grad_norm": 0.34673404693603516, "learning_rate": 9.766772017017654e-05, "loss": 0.0284, "step": 5710 }, { "grad_norm": 0.36465510725975037, "learning_rate": 9.765522625063032e-05, "loss": 0.0321, "step": 5720 }, { "grad_norm": 0.323629230260849, "learning_rate": 9.76426997591863e-05, "loss": 0.0334, "step": 5730 }, { "grad_norm": 0.30021965503692627, "learning_rate": 9.763014070440625e-05, "loss": 0.0353, "step": 5740 }, { "grad_norm": 0.28831347823143005, "learning_rate": 9.761754909487415e-05, "loss": 0.0332, "step": 5750 }, { "grad_norm": 0.29339149594306946, "learning_rate": 9.760492493919626e-05, "loss": 0.0296, "step": 5760 }, { "grad_norm": 0.2769908905029297, "learning_rate": 9.759226824600105e-05, "loss": 0.0323, "step": 5770 }, { "grad_norm": 0.27332574129104614, "learning_rate": 9.757957902393928e-05, "loss": 0.0311, "step": 5780 }, { "grad_norm": 0.3913270831108093, "learning_rate": 9.756685728168387e-05, "loss": 0.0345, "step": 5790 }, { "grad_norm": 0.3499385714530945, "learning_rate": 9.755410302793004e-05, "loss": 0.0315, "step": 5800 }, { "grad_norm": 0.3098371922969818, "learning_rate": 9.754131627139522e-05, "loss": 0.0305, "step": 5810 }, { "grad_norm": 0.3495834469795227, "learning_rate": 9.752849702081901e-05, "loss": 0.0336, "step": 5820 }, { "grad_norm": 0.3704340159893036, "learning_rate": 9.751564528496324e-05, "loss": 0.034, "step": 5830 }, { "grad_norm": 0.2951725423336029, "learning_rate": 9.750276107261197e-05, "loss": 0.0339, "step": 5840 }, { "grad_norm": 0.21020959317684174, "learning_rate": 9.748984439257142e-05, "loss": 0.0327, "step": 5850 }, { "grad_norm": 0.24040254950523376, "learning_rate": 9.747689525367005e-05, "loss": 0.034, "step": 5860 }, { "grad_norm": 0.31322363018989563, "learning_rate": 9.746391366475845e-05, "loss": 0.0342, "step": 5870 }, { "grad_norm": 0.3510359525680542, "learning_rate": 9.745089963470942e-05, "loss": 0.0319, "step": 5880 }, { "grad_norm": 0.29859283566474915, "learning_rate": 9.743785317241791e-05, "loss": 0.0297, "step": 5890 }, { "grad_norm": 0.3504839539527893, "learning_rate": 9.742477428680108e-05, "loss": 0.031, "step": 5900 }, { "grad_norm": 0.32250672578811646, "learning_rate": 9.741166298679821e-05, "loss": 0.0315, "step": 5910 }, { "grad_norm": 0.44023045897483826, "learning_rate": 9.739851928137076e-05, "loss": 0.0313, "step": 5920 }, { "grad_norm": 0.3380875289440155, "learning_rate": 9.738534317950232e-05, "loss": 0.0272, "step": 5930 }, { "grad_norm": 0.25726962089538574, "learning_rate": 9.737213469019864e-05, "loss": 0.0331, "step": 5940 }, { "grad_norm": 0.2557287812232971, "learning_rate": 9.735889382248757e-05, "loss": 0.0271, "step": 5950 }, { "grad_norm": 0.4016687870025635, "learning_rate": 9.734562058541916e-05, "loss": 0.0284, "step": 5960 }, { "grad_norm": 0.3491223454475403, "learning_rate": 9.733231498806552e-05, "loss": 0.0316, "step": 5970 }, { "grad_norm": 0.28919312357902527, "learning_rate": 9.731897703952088e-05, "loss": 0.0288, "step": 5980 }, { "grad_norm": 0.29605379700660706, "learning_rate": 9.730560674890165e-05, "loss": 0.0307, "step": 5990 }, { "grad_norm": 0.33510705828666687, "learning_rate": 9.729220412534628e-05, "loss": 0.0341, "step": 6000 }, { "grad_norm": 0.3583182394504547, "learning_rate": 9.72787691780153e-05, "loss": 0.0285, "step": 6010 }, { "grad_norm": 0.3129652738571167, "learning_rate": 9.726530191609142e-05, "loss": 0.0283, "step": 6020 }, { "grad_norm": 0.34205812215805054, "learning_rate": 9.725180234877937e-05, "loss": 0.0281, "step": 6030 }, { "grad_norm": 0.2633786201477051, "learning_rate": 9.723827048530595e-05, "loss": 0.0307, "step": 6040 }, { "grad_norm": 0.2981113791465759, "learning_rate": 9.722470633492011e-05, "loss": 0.0288, "step": 6050 }, { "grad_norm": 0.3252401649951935, "learning_rate": 9.721110990689278e-05, "loss": 0.027, "step": 6060 }, { "grad_norm": 0.29060742259025574, "learning_rate": 9.719748121051699e-05, "loss": 0.0315, "step": 6070 }, { "grad_norm": 0.34483039379119873, "learning_rate": 9.718382025510785e-05, "loss": 0.0339, "step": 6080 }, { "grad_norm": 0.387736052274704, "learning_rate": 9.717012705000248e-05, "loss": 0.0262, "step": 6090 }, { "grad_norm": 0.33958616852760315, "learning_rate": 9.715640160456006e-05, "loss": 0.0339, "step": 6100 }, { "grad_norm": 0.2881743013858795, "learning_rate": 9.714264392816181e-05, "loss": 0.0293, "step": 6110 }, { "grad_norm": 0.288390189409256, "learning_rate": 9.712885403021095e-05, "loss": 0.0336, "step": 6120 }, { "grad_norm": 0.2961685061454773, "learning_rate": 9.711503192013276e-05, "loss": 0.0292, "step": 6130 }, { "grad_norm": 0.3371663987636566, "learning_rate": 9.710117760737454e-05, "loss": 0.0294, "step": 6140 }, { "grad_norm": 0.32758694887161255, "learning_rate": 9.708729110140554e-05, "loss": 0.0334, "step": 6150 }, { "grad_norm": 0.2888786792755127, "learning_rate": 9.707337241171709e-05, "loss": 0.0293, "step": 6160 }, { "grad_norm": 0.2870728671550751, "learning_rate": 9.705942154782245e-05, "loss": 0.0314, "step": 6170 }, { "grad_norm": 0.3253091275691986, "learning_rate": 9.704543851925696e-05, "loss": 0.0297, "step": 6180 }, { "grad_norm": 0.3644893169403076, "learning_rate": 9.703142333557784e-05, "loss": 0.0279, "step": 6190 }, { "grad_norm": 0.3325464725494385, "learning_rate": 9.701737600636436e-05, "loss": 0.0306, "step": 6200 }, { "grad_norm": 0.32552286982536316, "learning_rate": 9.700329654121771e-05, "loss": 0.0318, "step": 6210 }, { "grad_norm": 0.32679203152656555, "learning_rate": 9.698918494976109e-05, "loss": 0.0348, "step": 6220 }, { "grad_norm": 0.2939686179161072, "learning_rate": 9.697504124163965e-05, "loss": 0.0317, "step": 6230 }, { "grad_norm": 0.32028719782829285, "learning_rate": 9.696086542652045e-05, "loss": 0.03, "step": 6240 }, { "grad_norm": 0.28813475370407104, "learning_rate": 9.694665751409256e-05, "loss": 0.0342, "step": 6250 }, { "grad_norm": 0.24833524227142334, "learning_rate": 9.693241751406694e-05, "loss": 0.0341, "step": 6260 }, { "grad_norm": 0.25627240538597107, "learning_rate": 9.69181454361765e-05, "loss": 0.0281, "step": 6270 }, { "grad_norm": 0.35726574063301086, "learning_rate": 9.690384129017605e-05, "loss": 0.0323, "step": 6280 }, { "grad_norm": 0.3148500621318817, "learning_rate": 9.688950508584237e-05, "loss": 0.027, "step": 6290 }, { "grad_norm": 0.3099932372570038, "learning_rate": 9.68751368329741e-05, "loss": 0.0364, "step": 6300 }, { "grad_norm": 0.3557538688182831, "learning_rate": 9.68607365413918e-05, "loss": 0.0326, "step": 6310 }, { "grad_norm": 0.28950875997543335, "learning_rate": 9.684630422093797e-05, "loss": 0.0314, "step": 6320 }, { "grad_norm": 0.25334855914115906, "learning_rate": 9.683183988147693e-05, "loss": 0.0316, "step": 6330 }, { "grad_norm": 0.2737284004688263, "learning_rate": 9.681734353289491e-05, "loss": 0.032, "step": 6340 }, { "grad_norm": 0.3514362871646881, "learning_rate": 9.680281518510006e-05, "loss": 0.0328, "step": 6350 }, { "grad_norm": 0.2936137616634369, "learning_rate": 9.678825484802236e-05, "loss": 0.0346, "step": 6360 }, { "grad_norm": 0.3055417537689209, "learning_rate": 9.677366253161365e-05, "loss": 0.0326, "step": 6370 }, { "grad_norm": 0.2818107306957245, "learning_rate": 9.675903824584765e-05, "loss": 0.0344, "step": 6380 }, { "grad_norm": 0.3474093973636627, "learning_rate": 9.674438200071991e-05, "loss": 0.0328, "step": 6390 }, { "grad_norm": 0.3185848295688629, "learning_rate": 9.672969380624786e-05, "loss": 0.0287, "step": 6400 }, { "grad_norm": 0.2904534637928009, "learning_rate": 9.671497367247073e-05, "loss": 0.0342, "step": 6410 }, { "grad_norm": 0.37376561760902405, "learning_rate": 9.670022160944959e-05, "loss": 0.0343, "step": 6420 }, { "grad_norm": 0.2735671401023865, "learning_rate": 9.668543762726734e-05, "loss": 0.0292, "step": 6430 }, { "grad_norm": 0.35097837448120117, "learning_rate": 9.667062173602868e-05, "loss": 0.0307, "step": 6440 }, { "grad_norm": 0.3375859558582306, "learning_rate": 9.665577394586013e-05, "loss": 0.0305, "step": 6450 }, { "grad_norm": 0.36888834834098816, "learning_rate": 9.664089426691006e-05, "loss": 0.0301, "step": 6460 }, { "grad_norm": 0.291920006275177, "learning_rate": 9.662598270934852e-05, "loss": 0.0298, "step": 6470 }, { "grad_norm": 0.29324302077293396, "learning_rate": 9.661103928336748e-05, "loss": 0.0337, "step": 6480 }, { "grad_norm": 0.3170052468776703, "learning_rate": 9.659606399918058e-05, "loss": 0.0315, "step": 6490 }, { "grad_norm": 0.2448112517595291, "learning_rate": 9.658105686702334e-05, "loss": 0.0276, "step": 6500 }, { "grad_norm": 0.3895646929740906, "learning_rate": 9.656601789715294e-05, "loss": 0.0291, "step": 6510 }, { "grad_norm": 0.2716320753097534, "learning_rate": 9.655094709984842e-05, "loss": 0.0285, "step": 6520 }, { "grad_norm": 0.2708210051059723, "learning_rate": 9.653584448541048e-05, "loss": 0.0297, "step": 6530 }, { "grad_norm": 0.3218449652194977, "learning_rate": 9.652071006416166e-05, "loss": 0.0294, "step": 6540 }, { "grad_norm": 0.2763862609863281, "learning_rate": 9.650554384644617e-05, "loss": 0.03, "step": 6550 }, { "grad_norm": 0.34127867221832275, "learning_rate": 9.649034584263e-05, "loss": 0.0316, "step": 6560 }, { "grad_norm": 0.2997555732727051, "learning_rate": 9.64751160631008e-05, "loss": 0.0331, "step": 6570 }, { "grad_norm": 0.3066577911376953, "learning_rate": 9.645985451826803e-05, "loss": 0.0277, "step": 6580 }, { "grad_norm": 0.31131234765052795, "learning_rate": 9.644456121856275e-05, "loss": 0.0299, "step": 6590 }, { "grad_norm": 0.3347361385822296, "learning_rate": 9.642923617443784e-05, "loss": 0.0343, "step": 6600 }, { "grad_norm": 0.3467046916484833, "learning_rate": 9.641387939636782e-05, "loss": 0.0342, "step": 6610 }, { "grad_norm": 0.30912765860557556, "learning_rate": 9.639849089484888e-05, "loss": 0.0258, "step": 6620 }, { "grad_norm": 0.3087506890296936, "learning_rate": 9.638307068039893e-05, "loss": 0.033, "step": 6630 }, { "grad_norm": 0.2909829616546631, "learning_rate": 9.636761876355753e-05, "loss": 0.0304, "step": 6640 }, { "grad_norm": 0.2978293001651764, "learning_rate": 9.635213515488593e-05, "loss": 0.0277, "step": 6650 }, { "grad_norm": 0.25319603085517883, "learning_rate": 9.633661986496702e-05, "loss": 0.0318, "step": 6660 }, { "grad_norm": 0.22676987946033478, "learning_rate": 9.632107290440535e-05, "loss": 0.0306, "step": 6670 }, { "grad_norm": 0.30086416006088257, "learning_rate": 9.630549428382715e-05, "loss": 0.0306, "step": 6680 }, { "grad_norm": 0.33707430958747864, "learning_rate": 9.628988401388024e-05, "loss": 0.0304, "step": 6690 }, { "grad_norm": 0.2530555725097656, "learning_rate": 9.627424210523407e-05, "loss": 0.0303, "step": 6700 }, { "grad_norm": 0.28887492418289185, "learning_rate": 9.62585685685798e-05, "loss": 0.0309, "step": 6710 }, { "grad_norm": 0.24906565248966217, "learning_rate": 9.624286341463008e-05, "loss": 0.029, "step": 6720 }, { "grad_norm": 0.3479105830192566, "learning_rate": 9.622712665411927e-05, "loss": 0.0264, "step": 6730 }, { "grad_norm": 0.31323543190956116, "learning_rate": 9.621135829780328e-05, "loss": 0.0292, "step": 6740 }, { "grad_norm": 0.31687647104263306, "learning_rate": 9.619555835645964e-05, "loss": 0.0281, "step": 6750 }, { "grad_norm": 0.26080283522605896, "learning_rate": 9.617972684088747e-05, "loss": 0.0286, "step": 6760 }, { "grad_norm": 0.2761741876602173, "learning_rate": 9.616386376190745e-05, "loss": 0.0278, "step": 6770 }, { "grad_norm": 0.25777819752693176, "learning_rate": 9.614796913036184e-05, "loss": 0.0289, "step": 6780 }, { "grad_norm": 0.22107671201229095, "learning_rate": 9.61320429571145e-05, "loss": 0.0274, "step": 6790 }, { "grad_norm": 0.23758363723754883, "learning_rate": 9.611608525305078e-05, "loss": 0.0284, "step": 6800 }, { "grad_norm": 0.2618251144886017, "learning_rate": 9.610009602907764e-05, "loss": 0.0294, "step": 6810 }, { "grad_norm": 0.3467927873134613, "learning_rate": 9.608407529612357e-05, "loss": 0.0314, "step": 6820 }, { "grad_norm": 0.32582366466522217, "learning_rate": 9.606802306513857e-05, "loss": 0.0276, "step": 6830 }, { "grad_norm": 0.3082886040210724, "learning_rate": 9.605193934709421e-05, "loss": 0.0296, "step": 6840 }, { "grad_norm": 0.2651321291923523, "learning_rate": 9.603582415298354e-05, "loss": 0.0305, "step": 6850 }, { "grad_norm": 0.2824176847934723, "learning_rate": 9.601967749382117e-05, "loss": 0.0326, "step": 6860 }, { "grad_norm": 0.35174012184143066, "learning_rate": 9.600349938064316e-05, "loss": 0.0266, "step": 6870 }, { "grad_norm": 0.2591400742530823, "learning_rate": 9.598728982450711e-05, "loss": 0.0314, "step": 6880 }, { "grad_norm": 0.3088349997997284, "learning_rate": 9.597104883649211e-05, "loss": 0.0253, "step": 6890 }, { "grad_norm": 0.3184434771537781, "learning_rate": 9.59547764276987e-05, "loss": 0.0256, "step": 6900 }, { "grad_norm": 0.37358859181404114, "learning_rate": 9.593847260924892e-05, "loss": 0.0257, "step": 6910 }, { "grad_norm": 0.4406052827835083, "learning_rate": 9.59221373922863e-05, "loss": 0.0286, "step": 6920 }, { "grad_norm": 0.34767359495162964, "learning_rate": 9.59057707879758e-05, "loss": 0.0303, "step": 6930 }, { "grad_norm": 0.32391783595085144, "learning_rate": 9.588937280750382e-05, "loss": 0.028, "step": 6940 }, { "grad_norm": 0.3464968800544739, "learning_rate": 9.587294346207824e-05, "loss": 0.0266, "step": 6950 }, { "grad_norm": 0.2730143368244171, "learning_rate": 9.585648276292836e-05, "loss": 0.0293, "step": 6960 }, { "grad_norm": 0.2093573659658432, "learning_rate": 9.583999072130492e-05, "loss": 0.0258, "step": 6970 }, { "grad_norm": 0.21776670217514038, "learning_rate": 9.582346734848005e-05, "loss": 0.028, "step": 6980 }, { "grad_norm": 0.31314799189567566, "learning_rate": 9.580691265574735e-05, "loss": 0.0261, "step": 6990 }, { "grad_norm": 0.3506203293800354, "learning_rate": 9.579032665442178e-05, "loss": 0.0279, "step": 7000 }, { "grad_norm": 0.27647268772125244, "learning_rate": 9.577370935583971e-05, "loss": 0.0304, "step": 7010 }, { "grad_norm": 0.27982640266418457, "learning_rate": 9.575706077135893e-05, "loss": 0.0311, "step": 7020 }, { "grad_norm": 0.2258869856595993, "learning_rate": 9.574038091235857e-05, "loss": 0.0266, "step": 7030 }, { "grad_norm": 0.3093338906764984, "learning_rate": 9.572366979023916e-05, "loss": 0.0287, "step": 7040 }, { "grad_norm": 0.2341330647468567, "learning_rate": 9.57069274164226e-05, "loss": 0.0269, "step": 7050 }, { "grad_norm": 0.2977883815765381, "learning_rate": 9.569015380235213e-05, "loss": 0.0288, "step": 7060 }, { "grad_norm": 0.2912971079349518, "learning_rate": 9.567334895949238e-05, "loss": 0.0257, "step": 7070 }, { "grad_norm": 0.35004955530166626, "learning_rate": 9.565651289932928e-05, "loss": 0.0276, "step": 7080 }, { "grad_norm": 0.22597353160381317, "learning_rate": 9.563964563337015e-05, "loss": 0.0274, "step": 7090 }, { "grad_norm": 0.2863006889820099, "learning_rate": 9.562274717314357e-05, "loss": 0.0241, "step": 7100 }, { "grad_norm": 0.23226788640022278, "learning_rate": 9.56058175301995e-05, "loss": 0.0274, "step": 7110 }, { "grad_norm": 0.32920441031455994, "learning_rate": 9.558885671610918e-05, "loss": 0.0285, "step": 7120 }, { "grad_norm": 0.31631603837013245, "learning_rate": 9.557186474246517e-05, "loss": 0.0248, "step": 7130 }, { "grad_norm": 0.364551305770874, "learning_rate": 9.555484162088133e-05, "loss": 0.0295, "step": 7140 }, { "grad_norm": 0.2645432949066162, "learning_rate": 9.553778736299279e-05, "loss": 0.0265, "step": 7150 }, { "grad_norm": 0.3499490022659302, "learning_rate": 9.552070198045599e-05, "loss": 0.0276, "step": 7160 }, { "grad_norm": 0.29266706109046936, "learning_rate": 9.550358548494863e-05, "loss": 0.026, "step": 7170 }, { "grad_norm": 0.3052130937576294, "learning_rate": 9.548643788816965e-05, "loss": 0.0283, "step": 7180 }, { "grad_norm": 0.2929697036743164, "learning_rate": 9.546925920183929e-05, "loss": 0.0276, "step": 7190 }, { "grad_norm": 0.3946594297885895, "learning_rate": 9.545204943769902e-05, "loss": 0.0287, "step": 7200 }, { "grad_norm": 0.3214876353740692, "learning_rate": 9.543480860751155e-05, "loss": 0.0295, "step": 7210 }, { "grad_norm": 0.39759424328804016, "learning_rate": 9.541753672306081e-05, "loss": 0.0283, "step": 7220 }, { "grad_norm": 0.28790152072906494, "learning_rate": 9.540023379615198e-05, "loss": 0.0275, "step": 7230 }, { "grad_norm": 0.24587273597717285, "learning_rate": 9.538289983861146e-05, "loss": 0.0244, "step": 7240 }, { "grad_norm": 0.37207937240600586, "learning_rate": 9.536553486228683e-05, "loss": 0.0257, "step": 7250 }, { "grad_norm": 0.3136146664619446, "learning_rate": 9.534813887904689e-05, "loss": 0.0257, "step": 7260 }, { "grad_norm": 0.3080083131790161, "learning_rate": 9.533071190078163e-05, "loss": 0.0261, "step": 7270 }, { "grad_norm": 0.23103229701519012, "learning_rate": 9.531325393940225e-05, "loss": 0.0247, "step": 7280 }, { "grad_norm": 0.25581589341163635, "learning_rate": 9.529576500684105e-05, "loss": 0.0269, "step": 7290 }, { "grad_norm": 0.3035750687122345, "learning_rate": 9.52782451150516e-05, "loss": 0.0264, "step": 7300 }, { "grad_norm": 0.3017367422580719, "learning_rate": 9.526069427600857e-05, "loss": 0.0248, "step": 7310 }, { "grad_norm": 0.33162030577659607, "learning_rate": 9.524311250170776e-05, "loss": 0.0269, "step": 7320 }, { "grad_norm": 0.24668215215206146, "learning_rate": 9.522549980416619e-05, "loss": 0.0303, "step": 7330 }, { "grad_norm": 0.23611865937709808, "learning_rate": 9.520785619542196e-05, "loss": 0.0275, "step": 7340 }, { "grad_norm": 0.3501213788986206, "learning_rate": 9.519018168753428e-05, "loss": 0.0284, "step": 7350 }, { "grad_norm": 0.25408846139907837, "learning_rate": 9.517247629258357e-05, "loss": 0.0245, "step": 7360 }, { "grad_norm": 0.2909450829029083, "learning_rate": 9.515474002267124e-05, "loss": 0.0254, "step": 7370 }, { "grad_norm": 0.3275704085826874, "learning_rate": 9.513697288991989e-05, "loss": 0.0264, "step": 7380 }, { "grad_norm": 0.32102933526039124, "learning_rate": 9.51191749064732e-05, "loss": 0.0284, "step": 7390 }, { "grad_norm": 0.3350803554058075, "learning_rate": 9.51013460844959e-05, "loss": 0.0287, "step": 7400 }, { "grad_norm": 0.36927127838134766, "learning_rate": 9.508348643617382e-05, "loss": 0.0314, "step": 7410 }, { "grad_norm": 0.2913030982017517, "learning_rate": 9.50655959737139e-05, "loss": 0.0282, "step": 7420 }, { "grad_norm": 0.2936769723892212, "learning_rate": 9.504767470934405e-05, "loss": 0.0357, "step": 7430 }, { "grad_norm": 0.23415061831474304, "learning_rate": 9.502972265531332e-05, "loss": 0.0299, "step": 7440 }, { "grad_norm": 0.27525657415390015, "learning_rate": 9.501173982389175e-05, "loss": 0.0279, "step": 7450 }, { "grad_norm": 0.2529526352882385, "learning_rate": 9.499372622737047e-05, "loss": 0.0287, "step": 7460 }, { "grad_norm": 0.34796956181526184, "learning_rate": 9.497568187806156e-05, "loss": 0.0333, "step": 7470 }, { "grad_norm": 0.4228200912475586, "learning_rate": 9.495760678829819e-05, "loss": 0.0268, "step": 7480 }, { "grad_norm": 0.27667564153671265, "learning_rate": 9.49395009704345e-05, "loss": 0.0298, "step": 7490 }, { "grad_norm": 0.3212486505508423, "learning_rate": 9.492136443684565e-05, "loss": 0.0296, "step": 7500 }, { "grad_norm": 0.3442087173461914, "learning_rate": 9.49031971999278e-05, "loss": 0.0259, "step": 7510 }, { "grad_norm": 0.2198597937822342, "learning_rate": 9.488499927209806e-05, "loss": 0.0235, "step": 7520 }, { "grad_norm": 0.262505441904068, "learning_rate": 9.486677066579456e-05, "loss": 0.0281, "step": 7530 }, { "grad_norm": 0.29834192991256714, "learning_rate": 9.484851139347639e-05, "loss": 0.028, "step": 7540 }, { "grad_norm": 0.2845780551433563, "learning_rate": 9.483022146762358e-05, "loss": 0.027, "step": 7550 }, { "grad_norm": 0.29360055923461914, "learning_rate": 9.48119009007371e-05, "loss": 0.0289, "step": 7560 }, { "grad_norm": 0.2781851589679718, "learning_rate": 9.479354970533892e-05, "loss": 0.0235, "step": 7570 }, { "grad_norm": 0.2674378454685211, "learning_rate": 9.477516789397189e-05, "loss": 0.0278, "step": 7580 }, { "grad_norm": 0.25910258293151855, "learning_rate": 9.475675547919983e-05, "loss": 0.0222, "step": 7590 }, { "grad_norm": 0.2658877968788147, "learning_rate": 9.47383124736074e-05, "loss": 0.0234, "step": 7600 }, { "grad_norm": 0.3179932236671448, "learning_rate": 9.471983888980028e-05, "loss": 0.0271, "step": 7610 }, { "grad_norm": 0.3225870132446289, "learning_rate": 9.470133474040494e-05, "loss": 0.0262, "step": 7620 }, { "grad_norm": 0.25730034708976746, "learning_rate": 9.468280003806882e-05, "loss": 0.0253, "step": 7630 }, { "grad_norm": 0.31448206305503845, "learning_rate": 9.46642347954602e-05, "loss": 0.0305, "step": 7640 }, { "grad_norm": 0.2568877339363098, "learning_rate": 9.464563902526826e-05, "loss": 0.0282, "step": 7650 }, { "grad_norm": 0.3970494568347931, "learning_rate": 9.462701274020303e-05, "loss": 0.0221, "step": 7660 }, { "grad_norm": 0.3699084520339966, "learning_rate": 9.46083559529954e-05, "loss": 0.0261, "step": 7670 }, { "grad_norm": 0.2663901448249817, "learning_rate": 9.458966867639712e-05, "loss": 0.027, "step": 7680 }, { "grad_norm": 0.21546752750873566, "learning_rate": 9.457095092318074e-05, "loss": 0.0264, "step": 7690 }, { "grad_norm": 0.29680266976356506, "learning_rate": 9.455220270613969e-05, "loss": 0.0247, "step": 7700 }, { "grad_norm": 0.2643856704235077, "learning_rate": 9.45334240380882e-05, "loss": 0.0255, "step": 7710 }, { "grad_norm": 0.27475506067276, "learning_rate": 9.451461493186129e-05, "loss": 0.0221, "step": 7720 }, { "grad_norm": 0.22947971522808075, "learning_rate": 9.449577540031482e-05, "loss": 0.0255, "step": 7730 }, { "grad_norm": 0.2972739040851593, "learning_rate": 9.447690545632544e-05, "loss": 0.0257, "step": 7740 }, { "grad_norm": 0.27710360288619995, "learning_rate": 9.445800511279058e-05, "loss": 0.0249, "step": 7750 }, { "grad_norm": 0.2977507412433624, "learning_rate": 9.443907438262844e-05, "loss": 0.0257, "step": 7760 }, { "grad_norm": 0.3082138001918793, "learning_rate": 9.442011327877798e-05, "loss": 0.0231, "step": 7770 }, { "grad_norm": 0.2509134113788605, "learning_rate": 9.440112181419895e-05, "loss": 0.0254, "step": 7780 }, { "grad_norm": 0.27079811692237854, "learning_rate": 9.438210000187186e-05, "loss": 0.0219, "step": 7790 }, { "grad_norm": 0.3026971220970154, "learning_rate": 9.43630478547979e-05, "loss": 0.0293, "step": 7800 }, { "grad_norm": 0.32309842109680176, "learning_rate": 9.434396538599906e-05, "loss": 0.028, "step": 7810 }, { "grad_norm": 0.23756740987300873, "learning_rate": 9.4324852608518e-05, "loss": 0.0257, "step": 7820 }, { "grad_norm": 0.2650049924850464, "learning_rate": 9.430570953541816e-05, "loss": 0.0242, "step": 7830 }, { "grad_norm": 0.40826505422592163, "learning_rate": 9.428653617978362e-05, "loss": 0.0254, "step": 7840 }, { "grad_norm": 0.293452650308609, "learning_rate": 9.42673325547192e-05, "loss": 0.0218, "step": 7850 }, { "grad_norm": 0.2603070139884949, "learning_rate": 9.42480986733504e-05, "loss": 0.0243, "step": 7860 }, { "grad_norm": 0.2677017152309418, "learning_rate": 9.422883454882338e-05, "loss": 0.022, "step": 7870 }, { "grad_norm": 0.32471850514411926, "learning_rate": 9.4209540194305e-05, "loss": 0.028, "step": 7880 }, { "grad_norm": 0.29776236414909363, "learning_rate": 9.419021562298278e-05, "loss": 0.0241, "step": 7890 }, { "grad_norm": 0.20936940610408783, "learning_rate": 9.417086084806486e-05, "loss": 0.0244, "step": 7900 }, { "grad_norm": 0.3175731599330902, "learning_rate": 9.415147588278005e-05, "loss": 0.0292, "step": 7910 }, { "grad_norm": 0.20533034205436707, "learning_rate": 9.413206074037781e-05, "loss": 0.0261, "step": 7920 }, { "grad_norm": 0.3036215603351593, "learning_rate": 9.411261543412819e-05, "loss": 0.0283, "step": 7930 }, { "grad_norm": 0.2161635011434555, "learning_rate": 9.409313997732188e-05, "loss": 0.0276, "step": 7940 }, { "grad_norm": 0.23634673655033112, "learning_rate": 9.407363438327018e-05, "loss": 0.0225, "step": 7950 }, { "grad_norm": 0.25447598099708557, "learning_rate": 9.405409866530496e-05, "loss": 0.0246, "step": 7960 }, { "grad_norm": 0.30755704641342163, "learning_rate": 9.40345328367787e-05, "loss": 0.0256, "step": 7970 }, { "grad_norm": 0.24388138949871063, "learning_rate": 9.401493691106446e-05, "loss": 0.0262, "step": 7980 }, { "grad_norm": 0.28241968154907227, "learning_rate": 9.39953109015559e-05, "loss": 0.023, "step": 7990 }, { "grad_norm": 0.26056113839149475, "learning_rate": 9.397565482166718e-05, "loss": 0.0231, "step": 8000 }, { "grad_norm": 0.2579169273376465, "learning_rate": 9.395596868483305e-05, "loss": 0.0239, "step": 8010 }, { "grad_norm": 0.24183133244514465, "learning_rate": 9.393625250450881e-05, "loss": 0.0279, "step": 8020 }, { "grad_norm": 0.24629710614681244, "learning_rate": 9.391650629417028e-05, "loss": 0.0224, "step": 8030 }, { "grad_norm": 0.3177964389324188, "learning_rate": 9.38967300673138e-05, "loss": 0.0271, "step": 8040 }, { "grad_norm": 0.233359232544899, "learning_rate": 9.387692383745624e-05, "loss": 0.0232, "step": 8050 }, { "grad_norm": 0.2218725085258484, "learning_rate": 9.385708761813496e-05, "loss": 0.0238, "step": 8060 }, { "grad_norm": 0.2714308500289917, "learning_rate": 9.383722142290787e-05, "loss": 0.0263, "step": 8070 }, { "grad_norm": 0.3039526343345642, "learning_rate": 9.381732526535328e-05, "loss": 0.0264, "step": 8080 }, { "grad_norm": 0.24046148359775543, "learning_rate": 9.379739915907004e-05, "loss": 0.024, "step": 8090 }, { "grad_norm": 0.2542411684989929, "learning_rate": 9.377744311767746e-05, "loss": 0.0249, "step": 8100 }, { "grad_norm": 0.20919090509414673, "learning_rate": 9.375745715481532e-05, "loss": 0.0291, "step": 8110 }, { "grad_norm": 0.27349570393562317, "learning_rate": 9.373744128414382e-05, "loss": 0.0271, "step": 8120 }, { "grad_norm": 0.2615624964237213, "learning_rate": 9.371739551934359e-05, "loss": 0.0253, "step": 8130 }, { "grad_norm": 0.2964158058166504, "learning_rate": 9.36973198741158e-05, "loss": 0.0262, "step": 8140 }, { "grad_norm": 0.27479565143585205, "learning_rate": 9.367721436218189e-05, "loss": 0.0257, "step": 8150 }, { "grad_norm": 0.2923128008842468, "learning_rate": 9.365707899728383e-05, "loss": 0.024, "step": 8160 }, { "grad_norm": 0.23255978524684906, "learning_rate": 9.363691379318391e-05, "loss": 0.023, "step": 8170 }, { "grad_norm": 0.27319592237472534, "learning_rate": 9.361671876366491e-05, "loss": 0.0245, "step": 8180 }, { "grad_norm": 0.2343263179063797, "learning_rate": 9.35964939225299e-05, "loss": 0.0247, "step": 8190 }, { "grad_norm": 0.2389761209487915, "learning_rate": 9.35762392836024e-05, "loss": 0.0206, "step": 8200 }, { "grad_norm": 0.3541310131549835, "learning_rate": 9.355595486072624e-05, "loss": 0.022, "step": 8210 }, { "grad_norm": 0.22738701105117798, "learning_rate": 9.353564066776563e-05, "loss": 0.0212, "step": 8220 }, { "grad_norm": 0.34077563881874084, "learning_rate": 9.351529671860516e-05, "loss": 0.0231, "step": 8230 }, { "grad_norm": 0.27848365902900696, "learning_rate": 9.349492302714969e-05, "loss": 0.0217, "step": 8240 }, { "grad_norm": 0.24166624248027802, "learning_rate": 9.347451960732447e-05, "loss": 0.0223, "step": 8250 }, { "grad_norm": 0.18881413340568542, "learning_rate": 9.345408647307506e-05, "loss": 0.0243, "step": 8260 }, { "grad_norm": 0.1954502910375595, "learning_rate": 9.343362363836726e-05, "loss": 0.0261, "step": 8270 }, { "grad_norm": 0.2565886974334717, "learning_rate": 9.341313111718728e-05, "loss": 0.0266, "step": 8280 }, { "grad_norm": 0.2674219012260437, "learning_rate": 9.339260892354153e-05, "loss": 0.0257, "step": 8290 }, { "grad_norm": 0.3212849199771881, "learning_rate": 9.337205707145676e-05, "loss": 0.023, "step": 8300 }, { "grad_norm": 0.24065330624580383, "learning_rate": 9.335147557497994e-05, "loss": 0.0252, "step": 8310 }, { "grad_norm": 0.3518981635570526, "learning_rate": 9.333086444817835e-05, "loss": 0.0247, "step": 8320 }, { "grad_norm": 0.251309335231781, "learning_rate": 9.331022370513951e-05, "loss": 0.0227, "step": 8330 }, { "grad_norm": 0.32785359025001526, "learning_rate": 9.328955335997113e-05, "loss": 0.0241, "step": 8340 }, { "grad_norm": 0.19077974557876587, "learning_rate": 9.326885342680125e-05, "loss": 0.0219, "step": 8350 }, { "grad_norm": 0.28100287914276123, "learning_rate": 9.324812391977806e-05, "loss": 0.022, "step": 8360 }, { "grad_norm": 0.3541359603404999, "learning_rate": 9.322736485306995e-05, "loss": 0.0241, "step": 8370 }, { "grad_norm": 0.2711920440196991, "learning_rate": 9.32065762408656e-05, "loss": 0.0251, "step": 8380 }, { "grad_norm": 0.2917647361755371, "learning_rate": 9.318575809737377e-05, "loss": 0.0247, "step": 8390 }, { "grad_norm": 0.2709057927131653, "learning_rate": 9.316491043682351e-05, "loss": 0.0232, "step": 8400 }, { "grad_norm": 0.3021368086338043, "learning_rate": 9.3144033273464e-05, "loss": 0.0228, "step": 8410 }, { "grad_norm": 0.2945931851863861, "learning_rate": 9.312312662156457e-05, "loss": 0.0236, "step": 8420 }, { "grad_norm": 0.27581465244293213, "learning_rate": 9.310219049541471e-05, "loss": 0.0214, "step": 8430 }, { "grad_norm": 0.2515217065811157, "learning_rate": 9.308122490932409e-05, "loss": 0.0243, "step": 8440 }, { "grad_norm": 0.22129814326763153, "learning_rate": 9.306022987762246e-05, "loss": 0.0243, "step": 8450 }, { "grad_norm": 0.2523355185985565, "learning_rate": 9.303920541465974e-05, "loss": 0.0265, "step": 8460 }, { "grad_norm": 0.2515203356742859, "learning_rate": 9.301815153480598e-05, "loss": 0.0249, "step": 8470 }, { "grad_norm": 0.2818777859210968, "learning_rate": 9.299706825245126e-05, "loss": 0.0221, "step": 8480 }, { "grad_norm": 0.27821144461631775, "learning_rate": 9.297595558200583e-05, "loss": 0.0212, "step": 8490 }, { "grad_norm": 0.2365865707397461, "learning_rate": 9.295481353790001e-05, "loss": 0.023, "step": 8500 }, { "grad_norm": 0.2957240045070648, "learning_rate": 9.293364213458417e-05, "loss": 0.0254, "step": 8510 }, { "grad_norm": 0.27481934428215027, "learning_rate": 9.291244138652877e-05, "loss": 0.0259, "step": 8520 }, { "grad_norm": 0.2442121058702469, "learning_rate": 9.289121130822431e-05, "loss": 0.0267, "step": 8530 }, { "grad_norm": 0.24839699268341064, "learning_rate": 9.286995191418137e-05, "loss": 0.0259, "step": 8540 }, { "grad_norm": 0.22057054936885834, "learning_rate": 9.284866321893056e-05, "loss": 0.0239, "step": 8550 }, { "grad_norm": 0.3274008631706238, "learning_rate": 9.282734523702244e-05, "loss": 0.0254, "step": 8560 }, { "grad_norm": 0.24115698039531708, "learning_rate": 9.280599798302771e-05, "loss": 0.0208, "step": 8570 }, { "grad_norm": 0.21825836598873138, "learning_rate": 9.278462147153699e-05, "loss": 0.024, "step": 8580 }, { "grad_norm": 0.20918643474578857, "learning_rate": 9.276321571716094e-05, "loss": 0.0229, "step": 8590 }, { "grad_norm": 0.240987166762352, "learning_rate": 9.274178073453017e-05, "loss": 0.0217, "step": 8600 }, { "grad_norm": 0.2889709174633026, "learning_rate": 9.27203165382953e-05, "loss": 0.0235, "step": 8610 }, { "grad_norm": 0.24812336266040802, "learning_rate": 9.26988231431269e-05, "loss": 0.0232, "step": 8620 }, { "grad_norm": 0.2584074139595032, "learning_rate": 9.267730056371551e-05, "loss": 0.0202, "step": 8630 }, { "grad_norm": 0.2119694948196411, "learning_rate": 9.265574881477161e-05, "loss": 0.0227, "step": 8640 }, { "grad_norm": 0.21296511590480804, "learning_rate": 9.263416791102561e-05, "loss": 0.0235, "step": 8650 }, { "grad_norm": 0.2301672399044037, "learning_rate": 9.261255786722786e-05, "loss": 0.0244, "step": 8660 }, { "grad_norm": 0.23107638955116272, "learning_rate": 9.259091869814864e-05, "loss": 0.026, "step": 8670 }, { "grad_norm": 0.22840340435504913, "learning_rate": 9.256925041857807e-05, "loss": 0.0257, "step": 8680 }, { "grad_norm": 0.2845821976661682, "learning_rate": 9.254755304332626e-05, "loss": 0.0234, "step": 8690 }, { "grad_norm": 0.2454874962568283, "learning_rate": 9.252582658722313e-05, "loss": 0.0219, "step": 8700 }, { "grad_norm": 0.23145116865634918, "learning_rate": 9.250407106511853e-05, "loss": 0.0221, "step": 8710 }, { "grad_norm": 0.2921099066734314, "learning_rate": 9.248228649188215e-05, "loss": 0.0196, "step": 8720 }, { "grad_norm": 0.27122461795806885, "learning_rate": 9.246047288240354e-05, "loss": 0.0244, "step": 8730 }, { "grad_norm": 0.26507461071014404, "learning_rate": 9.243863025159208e-05, "loss": 0.0233, "step": 8740 }, { "grad_norm": 0.2172965258359909, "learning_rate": 9.241675861437704e-05, "loss": 0.0265, "step": 8750 }, { "grad_norm": 0.24536068737506866, "learning_rate": 9.239485798570742e-05, "loss": 0.0218, "step": 8760 }, { "grad_norm": 0.21922439336776733, "learning_rate": 9.237292838055215e-05, "loss": 0.0198, "step": 8770 }, { "grad_norm": 0.20164522528648376, "learning_rate": 9.235096981389988e-05, "loss": 0.0205, "step": 8780 }, { "grad_norm": 0.2469058334827423, "learning_rate": 9.232898230075909e-05, "loss": 0.021, "step": 8790 }, { "grad_norm": 0.2168002873659134, "learning_rate": 9.230696585615801e-05, "loss": 0.0258, "step": 8800 }, { "grad_norm": 0.2594526708126068, "learning_rate": 9.228492049514473e-05, "loss": 0.0236, "step": 8810 }, { "grad_norm": 0.2451883852481842, "learning_rate": 9.2262846232787e-05, "loss": 0.0243, "step": 8820 }, { "grad_norm": 0.3122040033340454, "learning_rate": 9.224074308417238e-05, "loss": 0.0275, "step": 8830 }, { "grad_norm": 0.28484851121902466, "learning_rate": 9.221861106440818e-05, "loss": 0.0241, "step": 8840 }, { "grad_norm": 0.2622624337673187, "learning_rate": 9.219645018862142e-05, "loss": 0.0266, "step": 8850 }, { "grad_norm": 0.26287025213241577, "learning_rate": 9.217426047195882e-05, "loss": 0.0243, "step": 8860 }, { "grad_norm": 0.2617987394332886, "learning_rate": 9.21520419295869e-05, "loss": 0.0244, "step": 8870 }, { "grad_norm": 0.28980693221092224, "learning_rate": 9.212979457669178e-05, "loss": 0.0223, "step": 8880 }, { "grad_norm": 0.24466194212436676, "learning_rate": 9.210751842847931e-05, "loss": 0.0261, "step": 8890 }, { "grad_norm": 0.3030514121055603, "learning_rate": 9.208521350017506e-05, "loss": 0.0196, "step": 8900 }, { "grad_norm": 0.24446405470371246, "learning_rate": 9.206287980702422e-05, "loss": 0.0216, "step": 8910 }, { "grad_norm": 0.3540600538253784, "learning_rate": 9.204051736429166e-05, "loss": 0.0212, "step": 8920 }, { "grad_norm": 0.2210170030593872, "learning_rate": 9.20181261872619e-05, "loss": 0.0207, "step": 8930 }, { "grad_norm": 0.26153144240379333, "learning_rate": 9.19957062912391e-05, "loss": 0.0205, "step": 8940 }, { "grad_norm": 0.33273324370384216, "learning_rate": 9.197325769154704e-05, "loss": 0.0235, "step": 8950 }, { "grad_norm": 0.2623891532421112, "learning_rate": 9.195078040352914e-05, "loss": 0.0258, "step": 8960 }, { "grad_norm": 0.2649778723716736, "learning_rate": 9.192827444254839e-05, "loss": 0.0231, "step": 8970 }, { "grad_norm": 0.23149016499519348, "learning_rate": 9.190573982398744e-05, "loss": 0.0252, "step": 8980 }, { "grad_norm": 0.21975590288639069, "learning_rate": 9.188317656324845e-05, "loss": 0.0225, "step": 8990 }, { "grad_norm": 0.21691535413265228, "learning_rate": 9.186058467575322e-05, "loss": 0.0254, "step": 9000 }, { "grad_norm": 0.24422022700309753, "learning_rate": 9.183796417694309e-05, "loss": 0.0252, "step": 9010 }, { "grad_norm": 0.22867073118686676, "learning_rate": 9.181531508227897e-05, "loss": 0.0205, "step": 9020 }, { "grad_norm": 0.2107008397579193, "learning_rate": 9.179263740724125e-05, "loss": 0.025, "step": 9030 }, { "grad_norm": 0.3064573109149933, "learning_rate": 9.176993116732996e-05, "loss": 0.024, "step": 9040 }, { "grad_norm": 0.22909101843833923, "learning_rate": 9.17471963780646e-05, "loss": 0.0203, "step": 9050 }, { "grad_norm": 0.21902616322040558, "learning_rate": 9.172443305498414e-05, "loss": 0.0211, "step": 9060 }, { "grad_norm": 0.23765935003757477, "learning_rate": 9.170164121364714e-05, "loss": 0.0225, "step": 9070 }, { "grad_norm": 0.2750506103038788, "learning_rate": 9.167882086963158e-05, "loss": 0.0241, "step": 9080 }, { "grad_norm": 0.2481062263250351, "learning_rate": 9.165597203853499e-05, "loss": 0.0226, "step": 9090 }, { "grad_norm": 0.23933981359004974, "learning_rate": 9.163309473597428e-05, "loss": 0.0224, "step": 9100 }, { "grad_norm": 0.2476133406162262, "learning_rate": 9.161018897758591e-05, "loss": 0.0223, "step": 9110 }, { "grad_norm": 0.234623983502388, "learning_rate": 9.158725477902572e-05, "loss": 0.0237, "step": 9120 }, { "grad_norm": 0.27361729741096497, "learning_rate": 9.156429215596905e-05, "loss": 0.0224, "step": 9130 }, { "grad_norm": 0.28362837433815, "learning_rate": 9.154130112411061e-05, "loss": 0.0232, "step": 9140 }, { "grad_norm": 0.2722865343093872, "learning_rate": 9.151828169916455e-05, "loss": 0.0208, "step": 9150 }, { "grad_norm": 0.2745051980018616, "learning_rate": 9.149523389686445e-05, "loss": 0.0214, "step": 9160 }, { "grad_norm": 0.2825072109699249, "learning_rate": 9.147215773296327e-05, "loss": 0.0246, "step": 9170 }, { "grad_norm": 0.23496505618095398, "learning_rate": 9.144905322323334e-05, "loss": 0.0222, "step": 9180 }, { "grad_norm": 0.2969590723514557, "learning_rate": 9.142592038346635e-05, "loss": 0.0229, "step": 9190 }, { "grad_norm": 0.22086234390735626, "learning_rate": 9.140275922947343e-05, "loss": 0.0224, "step": 9200 }, { "grad_norm": 0.29896020889282227, "learning_rate": 9.137956977708496e-05, "loss": 0.0242, "step": 9210 }, { "grad_norm": 0.3109434247016907, "learning_rate": 9.135635204215075e-05, "loss": 0.0261, "step": 9220 }, { "grad_norm": 0.24603396654129028, "learning_rate": 9.133310604053986e-05, "loss": 0.0218, "step": 9230 }, { "grad_norm": 0.26985934376716614, "learning_rate": 9.130983178814077e-05, "loss": 0.0228, "step": 9240 }, { "grad_norm": 0.26301923394203186, "learning_rate": 9.128652930086116e-05, "loss": 0.0202, "step": 9250 }, { "grad_norm": 0.2593895494937897, "learning_rate": 9.126319859462808e-05, "loss": 0.0194, "step": 9260 }, { "grad_norm": 0.16785892844200134, "learning_rate": 9.123983968538787e-05, "loss": 0.0218, "step": 9270 }, { "grad_norm": 0.246376171708107, "learning_rate": 9.121645258910609e-05, "loss": 0.0222, "step": 9280 }, { "grad_norm": 0.21530038118362427, "learning_rate": 9.119303732176762e-05, "loss": 0.0215, "step": 9290 }, { "grad_norm": 0.26150277256965637, "learning_rate": 9.116959389937656e-05, "loss": 0.0227, "step": 9300 }, { "grad_norm": 0.2950771749019623, "learning_rate": 9.114612233795628e-05, "loss": 0.0199, "step": 9310 }, { "grad_norm": 0.3002603054046631, "learning_rate": 9.112262265354937e-05, "loss": 0.0193, "step": 9320 }, { "grad_norm": 0.31651565432548523, "learning_rate": 9.109909486221763e-05, "loss": 0.021, "step": 9330 }, { "grad_norm": 0.2795189321041107, "learning_rate": 9.107553898004208e-05, "loss": 0.0203, "step": 9340 }, { "grad_norm": 0.25990426540374756, "learning_rate": 9.105195502312298e-05, "loss": 0.0228, "step": 9350 }, { "grad_norm": 0.22616752982139587, "learning_rate": 9.102834300757972e-05, "loss": 0.0212, "step": 9360 }, { "grad_norm": 0.23084451258182526, "learning_rate": 9.100470294955087e-05, "loss": 0.019, "step": 9370 }, { "grad_norm": 0.24284148216247559, "learning_rate": 9.098103486519423e-05, "loss": 0.0206, "step": 9380 }, { "grad_norm": 0.23376473784446716, "learning_rate": 9.095733877068666e-05, "loss": 0.0192, "step": 9390 }, { "grad_norm": 0.2669183909893036, "learning_rate": 9.093361468222426e-05, "loss": 0.0222, "step": 9400 }, { "grad_norm": 0.2746864855289459, "learning_rate": 9.090986261602221e-05, "loss": 0.0194, "step": 9410 }, { "grad_norm": 0.26738181710243225, "learning_rate": 9.088608258831482e-05, "loss": 0.0224, "step": 9420 }, { "grad_norm": 0.22514821588993073, "learning_rate": 9.08622746153555e-05, "loss": 0.0238, "step": 9430 }, { "grad_norm": 0.27767127752304077, "learning_rate": 9.083843871341679e-05, "loss": 0.0204, "step": 9440 }, { "grad_norm": 0.2640741467475891, "learning_rate": 9.08145748987903e-05, "loss": 0.022, "step": 9450 }, { "grad_norm": 0.20908592641353607, "learning_rate": 9.079068318778674e-05, "loss": 0.0195, "step": 9460 }, { "grad_norm": 0.30638599395751953, "learning_rate": 9.076676359673586e-05, "loss": 0.0207, "step": 9470 }, { "grad_norm": 0.2941993176937103, "learning_rate": 9.074281614198645e-05, "loss": 0.0241, "step": 9480 }, { "grad_norm": 0.21782274544239044, "learning_rate": 9.071884083990638e-05, "loss": 0.0213, "step": 9490 }, { "grad_norm": 0.24893096089363098, "learning_rate": 9.069483770688259e-05, "loss": 0.0226, "step": 9500 }, { "grad_norm": 0.2450924664735794, "learning_rate": 9.067080675932093e-05, "loss": 0.0222, "step": 9510 }, { "grad_norm": 0.18195796012878418, "learning_rate": 9.064674801364634e-05, "loss": 0.0186, "step": 9520 }, { "grad_norm": 0.2247251272201538, "learning_rate": 9.062266148630279e-05, "loss": 0.0165, "step": 9530 }, { "grad_norm": 0.18547512590885162, "learning_rate": 9.059854719375313e-05, "loss": 0.0199, "step": 9540 }, { "grad_norm": 0.24902354180812836, "learning_rate": 9.057440515247932e-05, "loss": 0.0254, "step": 9550 }, { "grad_norm": 0.31469258666038513, "learning_rate": 9.055023537898215e-05, "loss": 0.0219, "step": 9560 }, { "grad_norm": 0.31413158774375916, "learning_rate": 9.052603788978149e-05, "loss": 0.0208, "step": 9570 }, { "grad_norm": 0.3116361200809479, "learning_rate": 9.050181270141606e-05, "loss": 0.0235, "step": 9580 }, { "grad_norm": 0.2721191644668579, "learning_rate": 9.047755983044355e-05, "loss": 0.0241, "step": 9590 }, { "grad_norm": 0.21902664005756378, "learning_rate": 9.045327929344058e-05, "loss": 0.0204, "step": 9600 }, { "grad_norm": 0.21708272397518158, "learning_rate": 9.042897110700268e-05, "loss": 0.0231, "step": 9610 }, { "grad_norm": 0.2067185938358307, "learning_rate": 9.040463528774423e-05, "loss": 0.0188, "step": 9620 }, { "grad_norm": 0.2706296145915985, "learning_rate": 9.038027185229856e-05, "loss": 0.0199, "step": 9630 }, { "grad_norm": 0.249213308095932, "learning_rate": 9.035588081731784e-05, "loss": 0.0239, "step": 9640 }, { "grad_norm": 0.25187763571739197, "learning_rate": 9.033146219947313e-05, "loss": 0.0234, "step": 9650 }, { "grad_norm": 0.203270822763443, "learning_rate": 9.030701601545429e-05, "loss": 0.0223, "step": 9660 }, { "grad_norm": 0.2530041038990021, "learning_rate": 9.028254228197009e-05, "loss": 0.0179, "step": 9670 }, { "grad_norm": 0.2137671858072281, "learning_rate": 9.025804101574808e-05, "loss": 0.0185, "step": 9680 }, { "grad_norm": 0.23890291154384613, "learning_rate": 9.023351223353463e-05, "loss": 0.0235, "step": 9690 }, { "grad_norm": 0.2893233299255371, "learning_rate": 9.020895595209498e-05, "loss": 0.0228, "step": 9700 }, { "grad_norm": 0.2888612449169159, "learning_rate": 9.018437218821305e-05, "loss": 0.0235, "step": 9710 }, { "grad_norm": 0.22156023979187012, "learning_rate": 9.015976095869166e-05, "loss": 0.0225, "step": 9720 }, { "grad_norm": 0.2601009011268616, "learning_rate": 9.013512228035235e-05, "loss": 0.0211, "step": 9730 }, { "grad_norm": 0.17970360815525055, "learning_rate": 9.01104561700354e-05, "loss": 0.0212, "step": 9740 }, { "grad_norm": 0.28424832224845886, "learning_rate": 9.008576264459987e-05, "loss": 0.0234, "step": 9750 }, { "grad_norm": 0.28393563628196716, "learning_rate": 9.006104172092357e-05, "loss": 0.0209, "step": 9760 }, { "grad_norm": 0.28482145071029663, "learning_rate": 9.0036293415903e-05, "loss": 0.0192, "step": 9770 }, { "grad_norm": 0.20902645587921143, "learning_rate": 9.00115177464534e-05, "loss": 0.0181, "step": 9780 }, { "grad_norm": 0.2787724733352661, "learning_rate": 8.998671472950872e-05, "loss": 0.0202, "step": 9790 }, { "grad_norm": 0.18188750743865967, "learning_rate": 8.996188438202156e-05, "loss": 0.0204, "step": 9800 }, { "grad_norm": 0.30744469165802, "learning_rate": 8.993702672096324e-05, "loss": 0.0279, "step": 9810 }, { "grad_norm": 0.31435006856918335, "learning_rate": 8.991214176332377e-05, "loss": 0.0234, "step": 9820 }, { "grad_norm": 0.24308079481124878, "learning_rate": 8.988722952611171e-05, "loss": 0.0248, "step": 9830 }, { "grad_norm": 0.21374386548995972, "learning_rate": 8.986229002635441e-05, "loss": 0.023, "step": 9840 }, { "grad_norm": 0.2762964069843292, "learning_rate": 8.983732328109775e-05, "loss": 0.0242, "step": 9850 }, { "grad_norm": 0.26984378695487976, "learning_rate": 8.981232930740627e-05, "loss": 0.0212, "step": 9860 }, { "grad_norm": 0.22955334186553955, "learning_rate": 8.978730812236313e-05, "loss": 0.0183, "step": 9870 }, { "grad_norm": 0.3325570821762085, "learning_rate": 8.976225974307004e-05, "loss": 0.0193, "step": 9880 }, { "grad_norm": 0.22263091802597046, "learning_rate": 8.973718418664736e-05, "loss": 0.0203, "step": 9890 }, { "grad_norm": 0.21340137720108032, "learning_rate": 8.971208147023397e-05, "loss": 0.0192, "step": 9900 }, { "grad_norm": 0.16173513233661652, "learning_rate": 8.968695161098739e-05, "loss": 0.0199, "step": 9910 }, { "grad_norm": 0.3063925504684448, "learning_rate": 8.96617946260836e-05, "loss": 0.019, "step": 9920 }, { "grad_norm": 0.2667577266693115, "learning_rate": 8.963661053271716e-05, "loss": 0.0195, "step": 9930 }, { "grad_norm": 0.2725638151168823, "learning_rate": 8.961139934810117e-05, "loss": 0.018, "step": 9940 }, { "grad_norm": 0.2795158624649048, "learning_rate": 8.958616108946725e-05, "loss": 0.0253, "step": 9950 }, { "grad_norm": 0.24791176617145538, "learning_rate": 8.95608957740655e-05, "loss": 0.0208, "step": 9960 }, { "grad_norm": 0.29863521456718445, "learning_rate": 8.953560341916452e-05, "loss": 0.0198, "step": 9970 }, { "grad_norm": 0.24745269119739532, "learning_rate": 8.95102840420514e-05, "loss": 0.017, "step": 9980 }, { "grad_norm": 0.30091291666030884, "learning_rate": 8.94849376600317e-05, "loss": 0.0195, "step": 9990 }, { "grad_norm": 0.26105543971061707, "learning_rate": 8.945956429042943e-05, "loss": 0.0205, "step": 10000 }, { "grad_norm": 0.22215516865253448, "learning_rate": 8.943416395058705e-05, "loss": 0.0181, "step": 10010 }, { "grad_norm": 0.25117915868759155, "learning_rate": 8.940873665786544e-05, "loss": 0.021, "step": 10020 }, { "grad_norm": 0.34157833456993103, "learning_rate": 8.938328242964394e-05, "loss": 0.0223, "step": 10030 }, { "grad_norm": 0.23542296886444092, "learning_rate": 8.935780128332026e-05, "loss": 0.0184, "step": 10040 }, { "grad_norm": 0.24906638264656067, "learning_rate": 8.933229323631052e-05, "loss": 0.0191, "step": 10050 }, { "grad_norm": 0.1712479442358017, "learning_rate": 8.930675830604925e-05, "loss": 0.0181, "step": 10060 }, { "grad_norm": 0.18746250867843628, "learning_rate": 8.92811965099893e-05, "loss": 0.0174, "step": 10070 }, { "grad_norm": 0.2352558821439743, "learning_rate": 8.925560786560194e-05, "loss": 0.0265, "step": 10080 }, { "grad_norm": 0.3187348544597626, "learning_rate": 8.922999239037677e-05, "loss": 0.02, "step": 10090 }, { "grad_norm": 0.22223259508609772, "learning_rate": 8.920435010182171e-05, "loss": 0.0185, "step": 10100 }, { "grad_norm": 0.23617500066757202, "learning_rate": 8.917868101746302e-05, "loss": 0.0174, "step": 10110 }, { "grad_norm": 0.22665368020534515, "learning_rate": 8.91529851548453e-05, "loss": 0.0195, "step": 10120 }, { "grad_norm": 0.24984513223171234, "learning_rate": 8.912726253153142e-05, "loss": 0.0211, "step": 10130 }, { "grad_norm": 0.19397790729999542, "learning_rate": 8.910151316510255e-05, "loss": 0.0177, "step": 10140 }, { "grad_norm": 0.2587132155895233, "learning_rate": 8.907573707315813e-05, "loss": 0.0172, "step": 10150 }, { "grad_norm": 0.25530922412872314, "learning_rate": 8.904993427331588e-05, "loss": 0.02, "step": 10160 }, { "grad_norm": 0.20462194085121155, "learning_rate": 8.902410478321176e-05, "loss": 0.0211, "step": 10170 }, { "grad_norm": 0.2912362217903137, "learning_rate": 8.899824862050002e-05, "loss": 0.0216, "step": 10180 }, { "grad_norm": 0.18036779761314392, "learning_rate": 8.897236580285308e-05, "loss": 0.02, "step": 10190 }, { "grad_norm": 0.22071509063243866, "learning_rate": 8.894645634796159e-05, "loss": 0.0197, "step": 10200 }, { "grad_norm": 0.2693204879760742, "learning_rate": 8.892052027353444e-05, "loss": 0.0224, "step": 10210 }, { "grad_norm": 0.22684092819690704, "learning_rate": 8.889455759729866e-05, "loss": 0.0232, "step": 10220 }, { "grad_norm": 0.2073947787284851, "learning_rate": 8.886856833699955e-05, "loss": 0.0202, "step": 10230 }, { "grad_norm": 0.20239147543907166, "learning_rate": 8.884255251040046e-05, "loss": 0.0191, "step": 10240 }, { "grad_norm": 0.22937780618667603, "learning_rate": 8.8816510135283e-05, "loss": 0.0169, "step": 10250 }, { "grad_norm": 0.22636641561985016, "learning_rate": 8.879044122944688e-05, "loss": 0.0227, "step": 10260 }, { "grad_norm": 0.21027106046676636, "learning_rate": 8.876434581070996e-05, "loss": 0.0218, "step": 10270 }, { "grad_norm": 0.276741623878479, "learning_rate": 8.87382238969082e-05, "loss": 0.022, "step": 10280 }, { "grad_norm": 0.24520862102508545, "learning_rate": 8.871207550589568e-05, "loss": 0.018, "step": 10290 }, { "grad_norm": 0.23584702610969543, "learning_rate": 8.868590065554458e-05, "loss": 0.0201, "step": 10300 }, { "grad_norm": 0.265521764755249, "learning_rate": 8.865969936374519e-05, "loss": 0.0201, "step": 10310 }, { "grad_norm": 0.3441570997238159, "learning_rate": 8.863347164840581e-05, "loss": 0.0179, "step": 10320 }, { "grad_norm": 0.2561526298522949, "learning_rate": 8.860721752745285e-05, "loss": 0.0223, "step": 10330 }, { "grad_norm": 0.2728956341743469, "learning_rate": 8.858093701883077e-05, "loss": 0.019, "step": 10340 }, { "grad_norm": 0.3359625041484833, "learning_rate": 8.8554630140502e-05, "loss": 0.025, "step": 10350 }, { "grad_norm": 0.2656126320362091, "learning_rate": 8.85282969104471e-05, "loss": 0.0219, "step": 10360 }, { "grad_norm": 0.23968838155269623, "learning_rate": 8.850193734666456e-05, "loss": 0.0188, "step": 10370 }, { "grad_norm": 0.25299686193466187, "learning_rate": 8.84755514671709e-05, "loss": 0.0184, "step": 10380 }, { "grad_norm": 0.3337089419364929, "learning_rate": 8.84491392900006e-05, "loss": 0.0195, "step": 10390 }, { "grad_norm": 0.24130408465862274, "learning_rate": 8.842270083320617e-05, "loss": 0.0184, "step": 10400 }, { "grad_norm": 0.2500981092453003, "learning_rate": 8.839623611485801e-05, "loss": 0.018, "step": 10410 }, { "grad_norm": 0.20847031474113464, "learning_rate": 8.836974515304453e-05, "loss": 0.0158, "step": 10420 }, { "grad_norm": 0.22296738624572754, "learning_rate": 8.834322796587204e-05, "loss": 0.0198, "step": 10430 }, { "grad_norm": 0.2207707315683365, "learning_rate": 8.831668457146478e-05, "loss": 0.0193, "step": 10440 }, { "grad_norm": 0.27005651593208313, "learning_rate": 8.829011498796493e-05, "loss": 0.0181, "step": 10450 }, { "grad_norm": 0.30843138694763184, "learning_rate": 8.826351923353253e-05, "loss": 0.0197, "step": 10460 }, { "grad_norm": 0.25124669075012207, "learning_rate": 8.823689732634555e-05, "loss": 0.0196, "step": 10470 }, { "grad_norm": 0.2143675535917282, "learning_rate": 8.82102492845998e-05, "loss": 0.0186, "step": 10480 }, { "grad_norm": 0.29483094811439514, "learning_rate": 8.818357512650896e-05, "loss": 0.0183, "step": 10490 }, { "grad_norm": 0.2770272195339203, "learning_rate": 8.815687487030458e-05, "loss": 0.0185, "step": 10500 }, { "grad_norm": 0.2650271952152252, "learning_rate": 8.8130148534236e-05, "loss": 0.0211, "step": 10510 }, { "grad_norm": 0.3367460072040558, "learning_rate": 8.810339613657047e-05, "loss": 0.0226, "step": 10520 }, { "grad_norm": 0.22345970571041107, "learning_rate": 8.807661769559295e-05, "loss": 0.021, "step": 10530 }, { "grad_norm": 0.24243248999118805, "learning_rate": 8.804981322960628e-05, "loss": 0.02, "step": 10540 }, { "grad_norm": 0.3006793260574341, "learning_rate": 8.802298275693106e-05, "loss": 0.0227, "step": 10550 }, { "grad_norm": 0.2038179486989975, "learning_rate": 8.799612629590568e-05, "loss": 0.0181, "step": 10560 }, { "grad_norm": 0.25536635518074036, "learning_rate": 8.796924386488624e-05, "loss": 0.0209, "step": 10570 }, { "grad_norm": 0.21317420899868011, "learning_rate": 8.794233548224666e-05, "loss": 0.0179, "step": 10580 }, { "grad_norm": 0.18500062823295593, "learning_rate": 8.791540116637853e-05, "loss": 0.0236, "step": 10590 }, { "grad_norm": 0.17616601288318634, "learning_rate": 8.788844093569124e-05, "loss": 0.0204, "step": 10600 }, { "grad_norm": 0.19113516807556152, "learning_rate": 8.786145480861184e-05, "loss": 0.0232, "step": 10610 }, { "grad_norm": 0.2439746856689453, "learning_rate": 8.783444280358507e-05, "loss": 0.0236, "step": 10620 }, { "grad_norm": 0.25458112359046936, "learning_rate": 8.780740493907342e-05, "loss": 0.0186, "step": 10630 }, { "grad_norm": 0.25391292572021484, "learning_rate": 8.778034123355698e-05, "loss": 0.0219, "step": 10640 }, { "grad_norm": 0.1900772750377655, "learning_rate": 8.775325170553357e-05, "loss": 0.0188, "step": 10650 }, { "grad_norm": 0.2907557487487793, "learning_rate": 8.77261363735186e-05, "loss": 0.0208, "step": 10660 }, { "grad_norm": 0.26870766282081604, "learning_rate": 8.769899525604517e-05, "loss": 0.0184, "step": 10670 }, { "grad_norm": 0.25862255692481995, "learning_rate": 8.767182837166397e-05, "loss": 0.0164, "step": 10680 }, { "grad_norm": 0.2176693230867386, "learning_rate": 8.764463573894328e-05, "loss": 0.0176, "step": 10690 }, { "grad_norm": 0.19560718536376953, "learning_rate": 8.761741737646902e-05, "loss": 0.0193, "step": 10700 }, { "grad_norm": 0.25090163946151733, "learning_rate": 8.759017330284471e-05, "loss": 0.0176, "step": 10710 }, { "grad_norm": 0.30748772621154785, "learning_rate": 8.756290353669142e-05, "loss": 0.0209, "step": 10720 }, { "grad_norm": 0.24951443076133728, "learning_rate": 8.753560809664774e-05, "loss": 0.0213, "step": 10730 }, { "grad_norm": 0.22543860971927643, "learning_rate": 8.750828700136986e-05, "loss": 0.0196, "step": 10740 }, { "grad_norm": 0.24240624904632568, "learning_rate": 8.74809402695315e-05, "loss": 0.0184, "step": 10750 }, { "grad_norm": 0.28887274861335754, "learning_rate": 8.745356791982391e-05, "loss": 0.0191, "step": 10760 }, { "grad_norm": 0.23907332122325897, "learning_rate": 8.742616997095578e-05, "loss": 0.0203, "step": 10770 }, { "grad_norm": 0.31905224919319153, "learning_rate": 8.739874644165341e-05, "loss": 0.0189, "step": 10780 }, { "grad_norm": 0.3664916455745697, "learning_rate": 8.737129735066048e-05, "loss": 0.0233, "step": 10790 }, { "grad_norm": 0.25929102301597595, "learning_rate": 8.734382271673821e-05, "loss": 0.0211, "step": 10800 }, { "grad_norm": 0.17529219388961792, "learning_rate": 8.731632255866525e-05, "loss": 0.0153, "step": 10810 }, { "grad_norm": 0.2659747004508972, "learning_rate": 8.728879689523767e-05, "loss": 0.0209, "step": 10820 }, { "grad_norm": 0.2383040487766266, "learning_rate": 8.726124574526905e-05, "loss": 0.0208, "step": 10830 }, { "grad_norm": 0.2467503398656845, "learning_rate": 8.72336691275903e-05, "loss": 0.0196, "step": 10840 }, { "grad_norm": 0.23609012365341187, "learning_rate": 8.720606706104979e-05, "loss": 0.0169, "step": 10850 }, { "grad_norm": 0.24395468831062317, "learning_rate": 8.71784395645133e-05, "loss": 0.0216, "step": 10860 }, { "grad_norm": 0.22184745967388153, "learning_rate": 8.715078665686392e-05, "loss": 0.0197, "step": 10870 }, { "grad_norm": 0.22744403779506683, "learning_rate": 8.712310835700218e-05, "loss": 0.0182, "step": 10880 }, { "grad_norm": 0.2295120358467102, "learning_rate": 8.709540468384591e-05, "loss": 0.0202, "step": 10890 }, { "grad_norm": 0.18560217320919037, "learning_rate": 8.706767565633033e-05, "loss": 0.0161, "step": 10900 }, { "grad_norm": 0.23202499747276306, "learning_rate": 8.7039921293408e-05, "loss": 0.0192, "step": 10910 }, { "grad_norm": 0.15415595471858978, "learning_rate": 8.70121416140487e-05, "loss": 0.0194, "step": 10920 }, { "grad_norm": 0.22386115789413452, "learning_rate": 8.698433663723962e-05, "loss": 0.0184, "step": 10930 }, { "grad_norm": 0.20881429314613342, "learning_rate": 8.695650638198518e-05, "loss": 0.0187, "step": 10940 }, { "grad_norm": 0.18674205243587494, "learning_rate": 8.692865086730713e-05, "loss": 0.016, "step": 10950 }, { "grad_norm": 0.15843376517295837, "learning_rate": 8.69007701122444e-05, "loss": 0.0161, "step": 10960 }, { "grad_norm": 0.17951293289661407, "learning_rate": 8.687286413585328e-05, "loss": 0.015, "step": 10970 }, { "grad_norm": 0.18391259014606476, "learning_rate": 8.684493295720719e-05, "loss": 0.0188, "step": 10980 }, { "grad_norm": 0.20475603640079498, "learning_rate": 8.681697659539685e-05, "loss": 0.0206, "step": 10990 }, { "grad_norm": 0.30364346504211426, "learning_rate": 8.678899506953019e-05, "loss": 0.0194, "step": 11000 }, { "grad_norm": 0.24425257742404938, "learning_rate": 8.676098839873227e-05, "loss": 0.0184, "step": 11010 }, { "grad_norm": 0.24779610335826874, "learning_rate": 8.673295660214545e-05, "loss": 0.0178, "step": 11020 }, { "grad_norm": 0.2540641725063324, "learning_rate": 8.670489969892914e-05, "loss": 0.0188, "step": 11030 }, { "grad_norm": 0.23731890320777893, "learning_rate": 8.667681770826e-05, "loss": 0.0183, "step": 11040 }, { "grad_norm": 0.6130825877189636, "learning_rate": 8.66487106493318e-05, "loss": 0.0209, "step": 11050 }, { "grad_norm": 0.30291688442230225, "learning_rate": 8.662057854135544e-05, "loss": 0.0203, "step": 11060 }, { "grad_norm": 0.25104424357414246, "learning_rate": 8.659242140355897e-05, "loss": 0.0199, "step": 11070 }, { "grad_norm": 0.27191582322120667, "learning_rate": 8.65642392551875e-05, "loss": 0.0207, "step": 11080 }, { "grad_norm": 0.27227821946144104, "learning_rate": 8.65360321155033e-05, "loss": 0.0175, "step": 11090 }, { "grad_norm": 0.21981225907802582, "learning_rate": 8.650780000378566e-05, "loss": 0.0185, "step": 11100 }, { "grad_norm": 0.26831111311912537, "learning_rate": 8.647954293933096e-05, "loss": 0.0177, "step": 11110 }, { "grad_norm": 0.229702889919281, "learning_rate": 8.645126094145264e-05, "loss": 0.0178, "step": 11120 }, { "grad_norm": 0.1753263771533966, "learning_rate": 8.642295402948117e-05, "loss": 0.0194, "step": 11130 }, { "grad_norm": 0.24257521331310272, "learning_rate": 8.639462222276409e-05, "loss": 0.0191, "step": 11140 }, { "grad_norm": 0.2198186069726944, "learning_rate": 8.636626554066589e-05, "loss": 0.0214, "step": 11150 }, { "grad_norm": 0.2623521089553833, "learning_rate": 8.633788400256811e-05, "loss": 0.0151, "step": 11160 }, { "grad_norm": 0.20003876090049744, "learning_rate": 8.630947762786927e-05, "loss": 0.017, "step": 11170 }, { "grad_norm": 0.19312936067581177, "learning_rate": 8.628104643598483e-05, "loss": 0.0195, "step": 11180 }, { "grad_norm": 0.25728538632392883, "learning_rate": 8.625259044634726e-05, "loss": 0.0164, "step": 11190 }, { "grad_norm": 0.19003871083259583, "learning_rate": 8.622410967840597e-05, "loss": 0.0176, "step": 11200 }, { "grad_norm": 0.2172747552394867, "learning_rate": 8.619560415162731e-05, "loss": 0.021, "step": 11210 }, { "grad_norm": 0.37889447808265686, "learning_rate": 8.616707388549447e-05, "loss": 0.0173, "step": 11220 }, { "grad_norm": 0.20873455703258514, "learning_rate": 8.613851889950771e-05, "loss": 0.0175, "step": 11230 }, { "grad_norm": 0.1924503594636917, "learning_rate": 8.610993921318402e-05, "loss": 0.0167, "step": 11240 }, { "grad_norm": 0.21296125650405884, "learning_rate": 8.608133484605738e-05, "loss": 0.0181, "step": 11250 }, { "grad_norm": 0.2885970175266266, "learning_rate": 8.605270581767859e-05, "loss": 0.0205, "step": 11260 }, { "grad_norm": 0.3101246654987335, "learning_rate": 8.602405214761536e-05, "loss": 0.0231, "step": 11270 }, { "grad_norm": 0.20907266438007355, "learning_rate": 8.599537385545215e-05, "loss": 0.0174, "step": 11280 }, { "grad_norm": 0.19936591386795044, "learning_rate": 8.596667096079032e-05, "loss": 0.0153, "step": 11290 }, { "grad_norm": 0.21146397292613983, "learning_rate": 8.593794348324806e-05, "loss": 0.017, "step": 11300 }, { "grad_norm": 0.23022054135799408, "learning_rate": 8.590919144246028e-05, "loss": 0.0208, "step": 11310 }, { "grad_norm": 0.2226577252149582, "learning_rate": 8.588041485807876e-05, "loss": 0.0154, "step": 11320 }, { "grad_norm": 0.15944652259349823, "learning_rate": 8.585161374977202e-05, "loss": 0.0191, "step": 11330 }, { "grad_norm": 0.22995106875896454, "learning_rate": 8.582278813722533e-05, "loss": 0.0206, "step": 11340 }, { "grad_norm": 0.23276503384113312, "learning_rate": 8.579393804014076e-05, "loss": 0.0202, "step": 11350 }, { "grad_norm": 0.21582365036010742, "learning_rate": 8.576506347823703e-05, "loss": 0.0155, "step": 11360 }, { "grad_norm": 0.28353002667427063, "learning_rate": 8.573616447124968e-05, "loss": 0.0187, "step": 11370 }, { "grad_norm": 0.22872577607631683, "learning_rate": 8.570724103893086e-05, "loss": 0.0204, "step": 11380 }, { "grad_norm": 0.2888127267360687, "learning_rate": 8.567829320104951e-05, "loss": 0.0198, "step": 11390 }, { "grad_norm": 0.2502804696559906, "learning_rate": 8.564932097739118e-05, "loss": 0.0183, "step": 11400 }, { "grad_norm": 0.22016920149326324, "learning_rate": 8.562032438775811e-05, "loss": 0.021, "step": 11410 }, { "grad_norm": 0.22050458192825317, "learning_rate": 8.559130345196921e-05, "loss": 0.0216, "step": 11420 }, { "grad_norm": 0.2636343538761139, "learning_rate": 8.556225818986e-05, "loss": 0.0212, "step": 11430 }, { "grad_norm": 0.24654430150985718, "learning_rate": 8.553318862128265e-05, "loss": 0.0202, "step": 11440 }, { "grad_norm": 0.27679264545440674, "learning_rate": 8.550409476610593e-05, "loss": 0.0171, "step": 11450 }, { "grad_norm": 0.19404137134552002, "learning_rate": 8.547497664421522e-05, "loss": 0.0189, "step": 11460 }, { "grad_norm": 0.19811709225177765, "learning_rate": 8.544583427551249e-05, "loss": 0.0191, "step": 11470 }, { "grad_norm": 0.21521785855293274, "learning_rate": 8.541666767991628e-05, "loss": 0.0186, "step": 11480 }, { "grad_norm": 0.2123854011297226, "learning_rate": 8.538747687736166e-05, "loss": 0.0166, "step": 11490 }, { "grad_norm": 0.22218434512615204, "learning_rate": 8.535826188780026e-05, "loss": 0.021, "step": 11500 }, { "grad_norm": 0.23296226561069489, "learning_rate": 8.532902273120029e-05, "loss": 0.0187, "step": 11510 }, { "grad_norm": 0.36751121282577515, "learning_rate": 8.52997594275464e-05, "loss": 0.0209, "step": 11520 }, { "grad_norm": 0.30906546115875244, "learning_rate": 8.52704719968398e-05, "loss": 0.0202, "step": 11530 }, { "grad_norm": 0.2917117476463318, "learning_rate": 8.524116045909818e-05, "loss": 0.0209, "step": 11540 }, { "grad_norm": 0.3181881606578827, "learning_rate": 8.521182483435569e-05, "loss": 0.0162, "step": 11550 }, { "grad_norm": 0.22398002445697784, "learning_rate": 8.518246514266295e-05, "loss": 0.022, "step": 11560 }, { "grad_norm": 0.27261483669281006, "learning_rate": 8.515308140408703e-05, "loss": 0.0197, "step": 11570 }, { "grad_norm": 0.2367214411497116, "learning_rate": 8.512367363871145e-05, "loss": 0.0175, "step": 11580 }, { "grad_norm": 0.21022246778011322, "learning_rate": 8.509424186663614e-05, "loss": 0.0154, "step": 11590 }, { "grad_norm": 0.22319594025611877, "learning_rate": 8.506478610797743e-05, "loss": 0.0149, "step": 11600 }, { "grad_norm": 0.24129194021224976, "learning_rate": 8.503530638286805e-05, "loss": 0.0188, "step": 11610 }, { "grad_norm": 0.20496773719787598, "learning_rate": 8.500580271145712e-05, "loss": 0.0204, "step": 11620 }, { "grad_norm": 0.21973097324371338, "learning_rate": 8.497627511391014e-05, "loss": 0.0155, "step": 11630 }, { "grad_norm": 0.1868925243616104, "learning_rate": 8.494672361040891e-05, "loss": 0.0165, "step": 11640 }, { "grad_norm": 0.23213548958301544, "learning_rate": 8.491714822115162e-05, "loss": 0.0172, "step": 11650 }, { "grad_norm": 0.2302439957857132, "learning_rate": 8.488754896635277e-05, "loss": 0.0162, "step": 11660 }, { "grad_norm": 0.3314053416252136, "learning_rate": 8.485792586624317e-05, "loss": 0.0145, "step": 11670 }, { "grad_norm": 0.24914032220840454, "learning_rate": 8.482827894106993e-05, "loss": 0.0166, "step": 11680 }, { "grad_norm": 0.24039843678474426, "learning_rate": 8.479860821109646e-05, "loss": 0.0209, "step": 11690 }, { "grad_norm": 0.2458333671092987, "learning_rate": 8.476891369660239e-05, "loss": 0.0194, "step": 11700 }, { "grad_norm": 0.2372313141822815, "learning_rate": 8.473919541788366e-05, "loss": 0.0174, "step": 11710 }, { "grad_norm": 0.2367023378610611, "learning_rate": 8.470945339525245e-05, "loss": 0.0187, "step": 11720 }, { "grad_norm": 0.17580176889896393, "learning_rate": 8.467968764903713e-05, "loss": 0.0147, "step": 11730 }, { "grad_norm": 0.17165589332580566, "learning_rate": 8.46498981995823e-05, "loss": 0.0156, "step": 11740 }, { "grad_norm": 0.2137993723154068, "learning_rate": 8.462008506724879e-05, "loss": 0.0164, "step": 11750 }, { "grad_norm": 0.2562062740325928, "learning_rate": 8.459024827241359e-05, "loss": 0.0182, "step": 11760 }, { "grad_norm": 0.29746121168136597, "learning_rate": 8.456038783546985e-05, "loss": 0.0163, "step": 11770 }, { "grad_norm": 0.21679754555225372, "learning_rate": 8.453050377682691e-05, "loss": 0.0214, "step": 11780 }, { "grad_norm": 0.1729767620563507, "learning_rate": 8.450059611691026e-05, "loss": 0.0161, "step": 11790 }, { "grad_norm": 0.3102567493915558, "learning_rate": 8.447066487616146e-05, "loss": 0.0226, "step": 11800 }, { "grad_norm": 0.2407054752111435, "learning_rate": 8.444071007503826e-05, "loss": 0.0182, "step": 11810 }, { "grad_norm": 0.2239379733800888, "learning_rate": 8.441073173401449e-05, "loss": 0.0215, "step": 11820 }, { "grad_norm": 0.20733599364757538, "learning_rate": 8.438072987358006e-05, "loss": 0.0182, "step": 11830 }, { "grad_norm": 0.2502686679363251, "learning_rate": 8.435070451424094e-05, "loss": 0.0165, "step": 11840 }, { "grad_norm": 0.25518798828125, "learning_rate": 8.432065567651919e-05, "loss": 0.0187, "step": 11850 }, { "grad_norm": 0.2016567438840866, "learning_rate": 8.429058338095291e-05, "loss": 0.0185, "step": 11860 }, { "grad_norm": 0.22486667335033417, "learning_rate": 8.426048764809624e-05, "loss": 0.0199, "step": 11870 }, { "grad_norm": 0.2117517590522766, "learning_rate": 8.423036849851932e-05, "loss": 0.0178, "step": 11880 }, { "grad_norm": 0.22635532915592194, "learning_rate": 8.42002259528083e-05, "loss": 0.0178, "step": 11890 }, { "grad_norm": 0.3198435604572296, "learning_rate": 8.417006003156532e-05, "loss": 0.0259, "step": 11900 }, { "grad_norm": 0.24303726851940155, "learning_rate": 8.413987075540852e-05, "loss": 0.0166, "step": 11910 }, { "grad_norm": 0.30342817306518555, "learning_rate": 8.4109658144972e-05, "loss": 0.0185, "step": 11920 }, { "grad_norm": 0.2860047519207001, "learning_rate": 8.407942222090573e-05, "loss": 0.0187, "step": 11930 }, { "grad_norm": 0.25140270590782166, "learning_rate": 8.404916300387576e-05, "loss": 0.0165, "step": 11940 }, { "grad_norm": 0.22377729415893555, "learning_rate": 8.401888051456391e-05, "loss": 0.0209, "step": 11950 }, { "grad_norm": 0.35688838362693787, "learning_rate": 8.398857477366803e-05, "loss": 0.0201, "step": 11960 }, { "grad_norm": 0.17974823713302612, "learning_rate": 8.395824580190178e-05, "loss": 0.0173, "step": 11970 }, { "grad_norm": 0.21865910291671753, "learning_rate": 8.392789361999473e-05, "loss": 0.0224, "step": 11980 }, { "grad_norm": 0.23724694550037384, "learning_rate": 8.38975182486923e-05, "loss": 0.0184, "step": 11990 }, { "grad_norm": 0.180343896150589, "learning_rate": 8.386711970875581e-05, "loss": 0.0206, "step": 12000 }, { "grad_norm": 0.2187199592590332, "learning_rate": 8.383669802096232e-05, "loss": 0.0211, "step": 12010 }, { "grad_norm": 0.2076634168624878, "learning_rate": 8.38062532061048e-05, "loss": 0.0185, "step": 12020 }, { "grad_norm": 0.2549191415309906, "learning_rate": 8.3775785284992e-05, "loss": 0.0193, "step": 12030 }, { "grad_norm": 0.17996518313884735, "learning_rate": 8.374529427844843e-05, "loss": 0.02, "step": 12040 }, { "grad_norm": 0.23560988903045654, "learning_rate": 8.371478020731442e-05, "loss": 0.0162, "step": 12050 }, { "grad_norm": 0.2114965319633484, "learning_rate": 8.368424309244607e-05, "loss": 0.0166, "step": 12060 }, { "grad_norm": 0.21519646048545837, "learning_rate": 8.365368295471517e-05, "loss": 0.0175, "step": 12070 }, { "grad_norm": 0.21451829373836517, "learning_rate": 8.362309981500931e-05, "loss": 0.0174, "step": 12080 }, { "grad_norm": 0.2086758315563202, "learning_rate": 8.359249369423177e-05, "loss": 0.016, "step": 12090 }, { "grad_norm": 0.25311732292175293, "learning_rate": 8.356186461330155e-05, "loss": 0.0179, "step": 12100 }, { "grad_norm": 0.17742426693439484, "learning_rate": 8.353121259315334e-05, "loss": 0.0166, "step": 12110 }, { "grad_norm": 0.2890436053276062, "learning_rate": 8.350053765473751e-05, "loss": 0.0208, "step": 12120 }, { "grad_norm": 0.1746264398097992, "learning_rate": 8.346983981902005e-05, "loss": 0.0175, "step": 12130 }, { "grad_norm": 0.22429224848747253, "learning_rate": 8.343911910698271e-05, "loss": 0.0153, "step": 12140 }, { "grad_norm": 0.247451514005661, "learning_rate": 8.340837553962278e-05, "loss": 0.0179, "step": 12150 }, { "grad_norm": 0.23079366981983185, "learning_rate": 8.337760913795316e-05, "loss": 0.0186, "step": 12160 }, { "grad_norm": 0.30063241720199585, "learning_rate": 8.334681992300244e-05, "loss": 0.0191, "step": 12170 }, { "grad_norm": 0.21775972843170166, "learning_rate": 8.331600791581475e-05, "loss": 0.0166, "step": 12180 }, { "grad_norm": 0.2115747630596161, "learning_rate": 8.328517313744978e-05, "loss": 0.0167, "step": 12190 }, { "grad_norm": 0.28884294629096985, "learning_rate": 8.325431560898286e-05, "loss": 0.0196, "step": 12200 }, { "grad_norm": 0.24262720346450806, "learning_rate": 8.322343535150478e-05, "loss": 0.0162, "step": 12210 }, { "grad_norm": 0.19995225965976715, "learning_rate": 8.319253238612191e-05, "loss": 0.0173, "step": 12220 }, { "grad_norm": 0.21896888315677643, "learning_rate": 8.316160673395614e-05, "loss": 0.0152, "step": 12230 }, { "grad_norm": 0.1877744197845459, "learning_rate": 8.313065841614487e-05, "loss": 0.0165, "step": 12240 }, { "grad_norm": 0.20550663769245148, "learning_rate": 8.309968745384096e-05, "loss": 0.0186, "step": 12250 }, { "grad_norm": 0.2203931361436844, "learning_rate": 8.306869386821282e-05, "loss": 0.0178, "step": 12260 }, { "grad_norm": 0.22008444368839264, "learning_rate": 8.30376776804442e-05, "loss": 0.0207, "step": 12270 }, { "grad_norm": 0.24439984560012817, "learning_rate": 8.300663891173443e-05, "loss": 0.0185, "step": 12280 }, { "grad_norm": 0.24773767590522766, "learning_rate": 8.297557758329822e-05, "loss": 0.0179, "step": 12290 }, { "grad_norm": 0.2381221204996109, "learning_rate": 8.294449371636564e-05, "loss": 0.0188, "step": 12300 }, { "grad_norm": 0.3012796938419342, "learning_rate": 8.291338733218226e-05, "loss": 0.0199, "step": 12310 }, { "grad_norm": 0.2523251473903656, "learning_rate": 8.2882258452009e-05, "loss": 0.0155, "step": 12320 }, { "grad_norm": 0.2409708946943283, "learning_rate": 8.285110709712214e-05, "loss": 0.0188, "step": 12330 }, { "grad_norm": 0.2073945850133896, "learning_rate": 8.281993328881337e-05, "loss": 0.0217, "step": 12340 }, { "grad_norm": 0.20019684731960297, "learning_rate": 8.278873704838964e-05, "loss": 0.0187, "step": 12350 }, { "grad_norm": 0.26508891582489014, "learning_rate": 8.275751839717334e-05, "loss": 0.0243, "step": 12360 }, { "grad_norm": 0.2632002830505371, "learning_rate": 8.272627735650208e-05, "loss": 0.0191, "step": 12370 }, { "grad_norm": 0.2109309434890747, "learning_rate": 8.269501394772884e-05, "loss": 0.0179, "step": 12380 }, { "grad_norm": 0.20536452531814575, "learning_rate": 8.266372819222189e-05, "loss": 0.0169, "step": 12390 }, { "grad_norm": 0.2250053435564041, "learning_rate": 8.26324201113647e-05, "loss": 0.0204, "step": 12400 }, { "grad_norm": 0.21860669553279877, "learning_rate": 8.260108972655606e-05, "loss": 0.0172, "step": 12410 }, { "grad_norm": 0.1929463893175125, "learning_rate": 8.256973705921e-05, "loss": 0.0165, "step": 12420 }, { "grad_norm": 0.1888851821422577, "learning_rate": 8.25383621307558e-05, "loss": 0.0148, "step": 12430 }, { "grad_norm": 0.19118458032608032, "learning_rate": 8.25069649626379e-05, "loss": 0.0169, "step": 12440 }, { "grad_norm": 0.16539081931114197, "learning_rate": 8.247554557631596e-05, "loss": 0.0238, "step": 12450 }, { "grad_norm": 0.2881024479866028, "learning_rate": 8.244410399326483e-05, "loss": 0.0172, "step": 12460 }, { "grad_norm": 0.2498202621936798, "learning_rate": 8.241264023497457e-05, "loss": 0.0217, "step": 12470 }, { "grad_norm": 0.1856665313243866, "learning_rate": 8.238115432295034e-05, "loss": 0.0157, "step": 12480 }, { "grad_norm": 0.23858493566513062, "learning_rate": 8.234964627871247e-05, "loss": 0.0163, "step": 12490 }, { "grad_norm": 0.21332816779613495, "learning_rate": 8.231811612379639e-05, "loss": 0.0164, "step": 12500 }, { "grad_norm": 0.16411803662776947, "learning_rate": 8.228656387975268e-05, "loss": 0.0147, "step": 12510 }, { "grad_norm": 0.2099577635526657, "learning_rate": 8.225498956814702e-05, "loss": 0.0164, "step": 12520 }, { "grad_norm": 0.22355705499649048, "learning_rate": 8.222339321056014e-05, "loss": 0.0191, "step": 12530 }, { "grad_norm": 0.2081795334815979, "learning_rate": 8.219177482858785e-05, "loss": 0.0175, "step": 12540 }, { "grad_norm": 0.27255335450172424, "learning_rate": 8.216013444384099e-05, "loss": 0.0183, "step": 12550 }, { "grad_norm": 0.27526572346687317, "learning_rate": 8.21284720779455e-05, "loss": 0.0165, "step": 12560 }, { "grad_norm": 0.24578741192817688, "learning_rate": 8.209678775254231e-05, "loss": 0.0166, "step": 12570 }, { "grad_norm": 0.20395155251026154, "learning_rate": 8.206508148928733e-05, "loss": 0.0162, "step": 12580 }, { "grad_norm": 0.26296132802963257, "learning_rate": 8.203335330985151e-05, "loss": 0.0166, "step": 12590 }, { "grad_norm": 0.17880253493785858, "learning_rate": 8.200160323592076e-05, "loss": 0.0171, "step": 12600 }, { "grad_norm": 0.19592520594596863, "learning_rate": 8.196983128919598e-05, "loss": 0.0174, "step": 12610 }, { "grad_norm": 0.19919532537460327, "learning_rate": 8.193803749139295e-05, "loss": 0.0186, "step": 12620 }, { "grad_norm": 0.21953949332237244, "learning_rate": 8.190622186424244e-05, "loss": 0.0153, "step": 12630 }, { "grad_norm": 0.3240562677383423, "learning_rate": 8.187438442949016e-05, "loss": 0.0162, "step": 12640 }, { "grad_norm": 0.2559094727039337, "learning_rate": 8.184252520889668e-05, "loss": 0.0179, "step": 12650 }, { "grad_norm": 0.21664273738861084, "learning_rate": 8.181064422423748e-05, "loss": 0.0145, "step": 12660 }, { "grad_norm": 0.22593152523040771, "learning_rate": 8.177874149730289e-05, "loss": 0.0133, "step": 12670 }, { "grad_norm": 0.2576131224632263, "learning_rate": 8.174681704989816e-05, "loss": 0.0175, "step": 12680 }, { "grad_norm": 0.20763225853443146, "learning_rate": 8.171487090384333e-05, "loss": 0.0175, "step": 12690 }, { "grad_norm": 0.23718459904193878, "learning_rate": 8.168290308097328e-05, "loss": 0.0144, "step": 12700 }, { "grad_norm": 0.17051364481449127, "learning_rate": 8.165091360313774e-05, "loss": 0.0149, "step": 12710 }, { "grad_norm": 0.21540166437625885, "learning_rate": 8.161890249220119e-05, "loss": 0.0158, "step": 12720 }, { "grad_norm": 0.19107170403003693, "learning_rate": 8.158686977004295e-05, "loss": 0.0152, "step": 12730 }, { "grad_norm": 0.22146956622600555, "learning_rate": 8.155481545855706e-05, "loss": 0.0147, "step": 12740 }, { "grad_norm": 0.20645342767238617, "learning_rate": 8.152273957965233e-05, "loss": 0.0159, "step": 12750 }, { "grad_norm": 0.23416759073734283, "learning_rate": 8.149064215525237e-05, "loss": 0.0161, "step": 12760 }, { "grad_norm": 0.16799239814281464, "learning_rate": 8.14585232072954e-05, "loss": 0.0169, "step": 12770 }, { "grad_norm": 0.21013560891151428, "learning_rate": 8.142638275773449e-05, "loss": 0.0191, "step": 12780 }, { "grad_norm": 0.18844157457351685, "learning_rate": 8.139422082853729e-05, "loss": 0.0151, "step": 12790 }, { "grad_norm": 0.25486886501312256, "learning_rate": 8.136203744168618e-05, "loss": 0.0166, "step": 12800 }, { "grad_norm": 0.24058876931667328, "learning_rate": 8.132983261917819e-05, "loss": 0.014, "step": 12810 }, { "grad_norm": 0.2764141857624054, "learning_rate": 8.129760638302504e-05, "loss": 0.0169, "step": 12820 }, { "grad_norm": 0.21541941165924072, "learning_rate": 8.126535875525305e-05, "loss": 0.0157, "step": 12830 }, { "grad_norm": 0.24408547580242157, "learning_rate": 8.123308975790316e-05, "loss": 0.016, "step": 12840 }, { "grad_norm": 0.27642789483070374, "learning_rate": 8.120079941303094e-05, "loss": 0.016, "step": 12850 }, { "grad_norm": 0.2073434591293335, "learning_rate": 8.116848774270651e-05, "loss": 0.0174, "step": 12860 }, { "grad_norm": 0.24771548807621002, "learning_rate": 8.113615476901461e-05, "loss": 0.0156, "step": 12870 }, { "grad_norm": 0.20121346414089203, "learning_rate": 8.110380051405454e-05, "loss": 0.0149, "step": 12880 }, { "grad_norm": 0.22831960022449493, "learning_rate": 8.107142499994009e-05, "loss": 0.0189, "step": 12890 }, { "grad_norm": 0.19623863697052002, "learning_rate": 8.103902824879966e-05, "loss": 0.0169, "step": 12900 }, { "grad_norm": 0.20377130806446075, "learning_rate": 8.10066102827761e-05, "loss": 0.0147, "step": 12910 }, { "grad_norm": 0.17880605161190033, "learning_rate": 8.097417112402676e-05, "loss": 0.017, "step": 12920 }, { "grad_norm": 0.2137020230293274, "learning_rate": 8.094171079472355e-05, "loss": 0.0173, "step": 12930 }, { "grad_norm": 0.24629808962345123, "learning_rate": 8.090922931705277e-05, "loss": 0.0187, "step": 12940 }, { "grad_norm": 0.24034066498279572, "learning_rate": 8.08767267132152e-05, "loss": 0.0156, "step": 12950 }, { "grad_norm": 0.2262631207704544, "learning_rate": 8.084420300542608e-05, "loss": 0.0167, "step": 12960 }, { "grad_norm": 0.1559891253709793, "learning_rate": 8.081165821591505e-05, "loss": 0.014, "step": 12970 }, { "grad_norm": 0.25349661707878113, "learning_rate": 8.077909236692615e-05, "loss": 0.017, "step": 12980 }, { "grad_norm": 0.25078877806663513, "learning_rate": 8.074650548071787e-05, "loss": 0.0227, "step": 12990 }, { "grad_norm": 0.1988590508699417, "learning_rate": 8.071389757956301e-05, "loss": 0.015, "step": 13000 }, { "grad_norm": 0.16543497145175934, "learning_rate": 8.068126868574876e-05, "loss": 0.016, "step": 13010 }, { "grad_norm": 0.20197251439094543, "learning_rate": 8.064861882157668e-05, "loss": 0.0157, "step": 13020 }, { "grad_norm": 0.17019738256931305, "learning_rate": 8.061594800936263e-05, "loss": 0.0154, "step": 13030 }, { "grad_norm": 0.19683925807476044, "learning_rate": 8.058325627143681e-05, "loss": 0.0207, "step": 13040 }, { "grad_norm": 0.2684325873851776, "learning_rate": 8.055054363014372e-05, "loss": 0.0192, "step": 13050 }, { "grad_norm": 0.21842093765735626, "learning_rate": 8.051781010784211e-05, "loss": 0.0154, "step": 13060 }, { "grad_norm": 0.24398529529571533, "learning_rate": 8.048505572690506e-05, "loss": 0.0154, "step": 13070 }, { "grad_norm": 0.21975106000900269, "learning_rate": 8.045228050971988e-05, "loss": 0.015, "step": 13080 }, { "grad_norm": 0.2433367371559143, "learning_rate": 8.041948447868814e-05, "loss": 0.0149, "step": 13090 }, { "grad_norm": 0.21382109820842743, "learning_rate": 8.038666765622558e-05, "loss": 0.0132, "step": 13100 }, { "grad_norm": 0.2516253888607025, "learning_rate": 8.03538300647622e-05, "loss": 0.0184, "step": 13110 }, { "grad_norm": 0.19520582258701324, "learning_rate": 8.03209717267422e-05, "loss": 0.0154, "step": 13120 }, { "grad_norm": 0.2110031098127365, "learning_rate": 8.028809266462395e-05, "loss": 0.018, "step": 13130 }, { "grad_norm": 0.23579959571361542, "learning_rate": 8.025519290087994e-05, "loss": 0.0173, "step": 13140 }, { "grad_norm": 0.22254222631454468, "learning_rate": 8.022227245799688e-05, "loss": 0.0189, "step": 13150 }, { "grad_norm": 0.2254694253206253, "learning_rate": 8.018933135847557e-05, "loss": 0.0159, "step": 13160 }, { "grad_norm": 0.20948080718517303, "learning_rate": 8.015636962483096e-05, "loss": 0.0192, "step": 13170 }, { "grad_norm": 0.19733671844005585, "learning_rate": 8.012338727959205e-05, "loss": 0.0153, "step": 13180 }, { "grad_norm": 0.21104082465171814, "learning_rate": 8.009038434530198e-05, "loss": 0.0168, "step": 13190 }, { "grad_norm": 0.213162362575531, "learning_rate": 8.005736084451796e-05, "loss": 0.0176, "step": 13200 }, { "grad_norm": 0.19361552596092224, "learning_rate": 8.002431679981122e-05, "loss": 0.0156, "step": 13210 }, { "grad_norm": 0.20205773413181305, "learning_rate": 7.999125223376706e-05, "loss": 0.0178, "step": 13220 }, { "grad_norm": 0.2094661444425583, "learning_rate": 7.99581671689848e-05, "loss": 0.0185, "step": 13230 }, { "grad_norm": 0.2515265643596649, "learning_rate": 7.992506162807775e-05, "loss": 0.0193, "step": 13240 }, { "grad_norm": 0.18742847442626953, "learning_rate": 7.989193563367328e-05, "loss": 0.0155, "step": 13250 }, { "grad_norm": 0.26928284764289856, "learning_rate": 7.985878920841266e-05, "loss": 0.0163, "step": 13260 }, { "grad_norm": 0.17018422484397888, "learning_rate": 7.982562237495117e-05, "loss": 0.014, "step": 13270 }, { "grad_norm": 0.18019500374794006, "learning_rate": 7.979243515595802e-05, "loss": 0.0151, "step": 13280 }, { "grad_norm": 0.19194470345973969, "learning_rate": 7.975922757411636e-05, "loss": 0.0132, "step": 13290 }, { "grad_norm": 0.24023868143558502, "learning_rate": 7.972599965212329e-05, "loss": 0.0168, "step": 13300 }, { "grad_norm": 0.20531193912029266, "learning_rate": 7.969275141268973e-05, "loss": 0.0152, "step": 13310 }, { "grad_norm": 0.21063114702701569, "learning_rate": 7.96594828785406e-05, "loss": 0.0146, "step": 13320 }, { "grad_norm": 0.17180630564689636, "learning_rate": 7.962619407241456e-05, "loss": 0.0161, "step": 13330 }, { "grad_norm": 0.1993652582168579, "learning_rate": 7.959288501706424e-05, "loss": 0.0144, "step": 13340 }, { "grad_norm": 0.18854546546936035, "learning_rate": 7.955955573525605e-05, "loss": 0.0141, "step": 13350 }, { "grad_norm": 0.20158731937408447, "learning_rate": 7.952620624977026e-05, "loss": 0.0153, "step": 13360 }, { "grad_norm": 0.25222715735435486, "learning_rate": 7.949283658340089e-05, "loss": 0.0155, "step": 13370 }, { "grad_norm": 0.25968554615974426, "learning_rate": 7.945944675895585e-05, "loss": 0.0145, "step": 13380 }, { "grad_norm": 0.19824601709842682, "learning_rate": 7.942603679925671e-05, "loss": 0.0132, "step": 13390 }, { "grad_norm": 0.16940373182296753, "learning_rate": 7.93926067271389e-05, "loss": 0.0157, "step": 13400 }, { "grad_norm": 0.1545998752117157, "learning_rate": 7.935915656545155e-05, "loss": 0.0142, "step": 13410 }, { "grad_norm": 0.17300759255886078, "learning_rate": 7.932568633705752e-05, "loss": 0.0148, "step": 13420 }, { "grad_norm": 0.20782044529914856, "learning_rate": 7.929219606483341e-05, "loss": 0.0172, "step": 13430 }, { "grad_norm": 0.24752721190452576, "learning_rate": 7.925868577166948e-05, "loss": 0.0187, "step": 13440 }, { "grad_norm": 0.19158002734184265, "learning_rate": 7.922515548046974e-05, "loss": 0.0166, "step": 13450 }, { "grad_norm": 0.24212509393692017, "learning_rate": 7.919160521415179e-05, "loss": 0.0172, "step": 13460 }, { "grad_norm": 0.20972943305969238, "learning_rate": 7.915803499564694e-05, "loss": 0.0204, "step": 13470 }, { "grad_norm": 0.27237847447395325, "learning_rate": 7.912444484790013e-05, "loss": 0.0179, "step": 13480 }, { "grad_norm": 0.3035251200199127, "learning_rate": 7.909083479386987e-05, "loss": 0.0151, "step": 13490 }, { "grad_norm": 0.15957576036453247, "learning_rate": 7.905720485652836e-05, "loss": 0.0143, "step": 13500 }, { "grad_norm": 0.18580631911754608, "learning_rate": 7.902355505886132e-05, "loss": 0.0168, "step": 13510 }, { "grad_norm": 0.19364473223686218, "learning_rate": 7.898988542386805e-05, "loss": 0.0132, "step": 13520 }, { "grad_norm": 0.21756140887737274, "learning_rate": 7.895619597456147e-05, "loss": 0.0137, "step": 13530 }, { "grad_norm": 0.21654999256134033, "learning_rate": 7.892248673396798e-05, "loss": 0.0139, "step": 13540 }, { "grad_norm": 0.23169736564159393, "learning_rate": 7.888875772512754e-05, "loss": 0.0146, "step": 13550 }, { "grad_norm": 0.2139684408903122, "learning_rate": 7.885500897109359e-05, "loss": 0.0196, "step": 13560 }, { "grad_norm": 0.26361969113349915, "learning_rate": 7.882124049493309e-05, "loss": 0.016, "step": 13570 }, { "grad_norm": 0.23658619821071625, "learning_rate": 7.878745231972649e-05, "loss": 0.015, "step": 13580 }, { "grad_norm": 0.20777083933353424, "learning_rate": 7.875364446856766e-05, "loss": 0.0169, "step": 13590 }, { "grad_norm": 0.22314894199371338, "learning_rate": 7.871981696456398e-05, "loss": 0.014, "step": 13600 }, { "grad_norm": 0.23980151116847992, "learning_rate": 7.868596983083623e-05, "loss": 0.0162, "step": 13610 }, { "grad_norm": 0.23887999355793, "learning_rate": 7.865210309051858e-05, "loss": 0.0156, "step": 13620 }, { "grad_norm": 0.13620759546756744, "learning_rate": 7.861821676675863e-05, "loss": 0.0159, "step": 13630 }, { "grad_norm": 0.19841620326042175, "learning_rate": 7.858431088271739e-05, "loss": 0.0156, "step": 13640 }, { "grad_norm": 0.20556743443012238, "learning_rate": 7.855038546156918e-05, "loss": 0.015, "step": 13650 }, { "grad_norm": 0.2289557009935379, "learning_rate": 7.851644052650173e-05, "loss": 0.0125, "step": 13660 }, { "grad_norm": 0.26681771874427795, "learning_rate": 7.848247610071609e-05, "loss": 0.0161, "step": 13670 }, { "grad_norm": 0.23265118896961212, "learning_rate": 7.844849220742658e-05, "loss": 0.0151, "step": 13680 }, { "grad_norm": 0.2238052785396576, "learning_rate": 7.841448886986092e-05, "loss": 0.0165, "step": 13690 }, { "grad_norm": 0.23702161014080048, "learning_rate": 7.838046611126004e-05, "loss": 0.0149, "step": 13700 }, { "grad_norm": 0.2031276524066925, "learning_rate": 7.834642395487819e-05, "loss": 0.016, "step": 13710 }, { "grad_norm": 0.22214768826961517, "learning_rate": 7.831236242398285e-05, "loss": 0.0177, "step": 13720 }, { "grad_norm": 0.23577041923999786, "learning_rate": 7.827828154185477e-05, "loss": 0.0157, "step": 13730 }, { "grad_norm": 0.2551276683807373, "learning_rate": 7.82441813317879e-05, "loss": 0.0174, "step": 13740 }, { "grad_norm": 0.2384129911661148, "learning_rate": 7.821006181708944e-05, "loss": 0.0129, "step": 13750 }, { "grad_norm": 0.18845956027507782, "learning_rate": 7.81759230210797e-05, "loss": 0.0145, "step": 13760 }, { "grad_norm": 0.18680448830127716, "learning_rate": 7.814176496709227e-05, "loss": 0.0158, "step": 13770 }, { "grad_norm": 0.1993766874074936, "learning_rate": 7.810758767847385e-05, "loss": 0.0183, "step": 13780 }, { "grad_norm": 0.26006099581718445, "learning_rate": 7.807339117858427e-05, "loss": 0.016, "step": 13790 }, { "grad_norm": 0.20304886996746063, "learning_rate": 7.803917549079655e-05, "loss": 0.0152, "step": 13800 }, { "grad_norm": 0.19951952993869781, "learning_rate": 7.800494063849679e-05, "loss": 0.0143, "step": 13810 }, { "grad_norm": 0.17860150337219238, "learning_rate": 7.797068664508416e-05, "loss": 0.0153, "step": 13820 }, { "grad_norm": 0.16636033356189728, "learning_rate": 7.793641353397096e-05, "loss": 0.0132, "step": 13830 }, { "grad_norm": 0.1854110211133957, "learning_rate": 7.790212132858253e-05, "loss": 0.0137, "step": 13840 }, { "grad_norm": 0.20957574248313904, "learning_rate": 7.786781005235728e-05, "loss": 0.0161, "step": 13850 }, { "grad_norm": 0.22982902824878693, "learning_rate": 7.783347972874662e-05, "loss": 0.0173, "step": 13860 }, { "grad_norm": 0.1615365445613861, "learning_rate": 7.779913038121504e-05, "loss": 0.0163, "step": 13870 }, { "grad_norm": 0.2018740177154541, "learning_rate": 7.776476203323997e-05, "loss": 0.0156, "step": 13880 }, { "grad_norm": 0.19926461577415466, "learning_rate": 7.773037470831185e-05, "loss": 0.0147, "step": 13890 }, { "grad_norm": 0.14621637761592865, "learning_rate": 7.76959684299341e-05, "loss": 0.0163, "step": 13900 }, { "grad_norm": 0.20612511038780212, "learning_rate": 7.76615432216231e-05, "loss": 0.0126, "step": 13910 }, { "grad_norm": 0.23188261687755585, "learning_rate": 7.762709910690811e-05, "loss": 0.014, "step": 13920 }, { "grad_norm": 0.2202957272529602, "learning_rate": 7.759263610933141e-05, "loss": 0.0151, "step": 13930 }, { "grad_norm": 0.21434450149536133, "learning_rate": 7.755815425244811e-05, "loss": 0.0154, "step": 13940 }, { "grad_norm": 0.21211746335029602, "learning_rate": 7.752365355982624e-05, "loss": 0.0162, "step": 13950 }, { "grad_norm": 0.2918245792388916, "learning_rate": 7.748913405504668e-05, "loss": 0.0162, "step": 13960 }, { "grad_norm": 0.23693518340587616, "learning_rate": 7.745459576170322e-05, "loss": 0.0163, "step": 13970 }, { "grad_norm": 0.19045722484588623, "learning_rate": 7.742003870340242e-05, "loss": 0.0177, "step": 13980 }, { "grad_norm": 0.22638161480426788, "learning_rate": 7.738546290376373e-05, "loss": 0.018, "step": 13990 }, { "grad_norm": 0.1802339404821396, "learning_rate": 7.735086838641937e-05, "loss": 0.0203, "step": 14000 }, { "grad_norm": 0.18877999484539032, "learning_rate": 7.731625517501437e-05, "loss": 0.0128, "step": 14010 }, { "grad_norm": 0.17057697474956512, "learning_rate": 7.728162329320655e-05, "loss": 0.0134, "step": 14020 }, { "grad_norm": 0.22819986939430237, "learning_rate": 7.724697276466645e-05, "loss": 0.0125, "step": 14030 }, { "grad_norm": 0.3224698305130005, "learning_rate": 7.721230361307738e-05, "loss": 0.0159, "step": 14040 }, { "grad_norm": 0.21606358885765076, "learning_rate": 7.71776158621354e-05, "loss": 0.0167, "step": 14050 }, { "grad_norm": 0.26058927178382874, "learning_rate": 7.714290953554925e-05, "loss": 0.0168, "step": 14060 }, { "grad_norm": 0.18929091095924377, "learning_rate": 7.710818465704037e-05, "loss": 0.0163, "step": 14070 }, { "grad_norm": 0.22681069374084473, "learning_rate": 7.707344125034288e-05, "loss": 0.0145, "step": 14080 }, { "grad_norm": 0.22353090345859528, "learning_rate": 7.703867933920359e-05, "loss": 0.0158, "step": 14090 }, { "grad_norm": 0.16793595254421234, "learning_rate": 7.700389894738194e-05, "loss": 0.0145, "step": 14100 }, { "grad_norm": 0.23906242847442627, "learning_rate": 7.696910009864999e-05, "loss": 0.0139, "step": 14110 }, { "grad_norm": 0.18768469989299774, "learning_rate": 7.693428281679241e-05, "loss": 0.0139, "step": 14120 }, { "grad_norm": 0.25872692465782166, "learning_rate": 7.689944712560652e-05, "loss": 0.0148, "step": 14130 }, { "grad_norm": 0.20493973791599274, "learning_rate": 7.686459304890214e-05, "loss": 0.0122, "step": 14140 }, { "grad_norm": 0.24843093752861023, "learning_rate": 7.682972061050175e-05, "loss": 0.0164, "step": 14150 }, { "grad_norm": 0.25636807084083557, "learning_rate": 7.679482983424032e-05, "loss": 0.0168, "step": 14160 }, { "grad_norm": 0.305909663438797, "learning_rate": 7.675992074396534e-05, "loss": 0.016, "step": 14170 }, { "grad_norm": 0.19563986361026764, "learning_rate": 7.672499336353687e-05, "loss": 0.0193, "step": 14180 }, { "grad_norm": 0.19544082880020142, "learning_rate": 7.669004771682744e-05, "loss": 0.0171, "step": 14190 }, { "grad_norm": 0.2992183268070221, "learning_rate": 7.665508382772206e-05, "loss": 0.0173, "step": 14200 }, { "grad_norm": 0.22140900790691376, "learning_rate": 7.662010172011824e-05, "loss": 0.0121, "step": 14210 }, { "grad_norm": 0.2126418650150299, "learning_rate": 7.658510141792588e-05, "loss": 0.0142, "step": 14220 }, { "grad_norm": 0.1652589738368988, "learning_rate": 7.65500829450674e-05, "loss": 0.0139, "step": 14230 }, { "grad_norm": 0.2208600640296936, "learning_rate": 7.651504632547759e-05, "loss": 0.0175, "step": 14240 }, { "grad_norm": 0.2575294077396393, "learning_rate": 7.647999158310364e-05, "loss": 0.016, "step": 14250 }, { "grad_norm": 0.23885753750801086, "learning_rate": 7.644491874190512e-05, "loss": 0.0137, "step": 14260 }, { "grad_norm": 0.18331624567508698, "learning_rate": 7.6409827825854e-05, "loss": 0.0155, "step": 14270 }, { "grad_norm": 0.17421506345272064, "learning_rate": 7.637471885893459e-05, "loss": 0.0146, "step": 14280 }, { "grad_norm": 0.20470581948757172, "learning_rate": 7.633959186514354e-05, "loss": 0.0155, "step": 14290 }, { "grad_norm": 0.22740599513053894, "learning_rate": 7.630444686848984e-05, "loss": 0.0188, "step": 14300 }, { "grad_norm": 0.17559725046157837, "learning_rate": 7.626928389299471e-05, "loss": 0.0147, "step": 14310 }, { "grad_norm": 0.22878043353557587, "learning_rate": 7.623410296269175e-05, "loss": 0.0159, "step": 14320 }, { "grad_norm": 0.1737985610961914, "learning_rate": 7.61989041016268e-05, "loss": 0.016, "step": 14330 }, { "grad_norm": 0.24595926702022552, "learning_rate": 7.616368733385793e-05, "loss": 0.0137, "step": 14340 }, { "grad_norm": 0.2111639827489853, "learning_rate": 7.612845268345547e-05, "loss": 0.0152, "step": 14350 }, { "grad_norm": 0.19031749665737152, "learning_rate": 7.609320017450199e-05, "loss": 0.0153, "step": 14360 }, { "grad_norm": 0.13511539995670319, "learning_rate": 7.605792983109222e-05, "loss": 0.0123, "step": 14370 }, { "grad_norm": 0.18074005842208862, "learning_rate": 7.602264167733313e-05, "loss": 0.0128, "step": 14380 }, { "grad_norm": 0.22212286293506622, "learning_rate": 7.598733573734384e-05, "loss": 0.0173, "step": 14390 }, { "grad_norm": 0.23578713834285736, "learning_rate": 7.595201203525561e-05, "loss": 0.0148, "step": 14400 }, { "grad_norm": 0.19092415273189545, "learning_rate": 7.591667059521187e-05, "loss": 0.0165, "step": 14410 }, { "grad_norm": 0.194881409406662, "learning_rate": 7.588131144136815e-05, "loss": 0.0153, "step": 14420 }, { "grad_norm": 0.2663712799549103, "learning_rate": 7.584593459789212e-05, "loss": 0.0134, "step": 14430 }, { "grad_norm": 0.18820077180862427, "learning_rate": 7.58105400889635e-05, "loss": 0.0153, "step": 14440 }, { "grad_norm": 0.20907944440841675, "learning_rate": 7.57751279387741e-05, "loss": 0.0152, "step": 14450 }, { "grad_norm": 0.21999479830265045, "learning_rate": 7.573969817152782e-05, "loss": 0.0142, "step": 14460 }, { "grad_norm": 0.2039901167154312, "learning_rate": 7.570425081144052e-05, "loss": 0.0168, "step": 14470 }, { "grad_norm": 0.15253888070583344, "learning_rate": 7.56687858827402e-05, "loss": 0.0136, "step": 14480 }, { "grad_norm": 0.15254393219947815, "learning_rate": 7.563330340966675e-05, "loss": 0.0145, "step": 14490 }, { "grad_norm": 0.2090948522090912, "learning_rate": 7.559780341647212e-05, "loss": 0.0141, "step": 14500 }, { "grad_norm": 0.22554334998130798, "learning_rate": 7.556228592742026e-05, "loss": 0.0163, "step": 14510 }, { "grad_norm": 0.19287480413913727, "learning_rate": 7.552675096678696e-05, "loss": 0.0144, "step": 14520 }, { "grad_norm": 0.25899502635002136, "learning_rate": 7.549119855886012e-05, "loss": 0.0152, "step": 14530 }, { "grad_norm": 0.2110418975353241, "learning_rate": 7.545562872793941e-05, "loss": 0.0177, "step": 14540 }, { "grad_norm": 0.1995914727449417, "learning_rate": 7.542004149833648e-05, "loss": 0.0146, "step": 14550 }, { "grad_norm": 0.18074378371238708, "learning_rate": 7.538443689437492e-05, "loss": 0.0138, "step": 14560 }, { "grad_norm": 0.2020174264907837, "learning_rate": 7.53488149403901e-05, "loss": 0.0164, "step": 14570 }, { "grad_norm": 0.2126014083623886, "learning_rate": 7.531317566072929e-05, "loss": 0.0152, "step": 14580 }, { "grad_norm": 0.23360122740268707, "learning_rate": 7.527751907975158e-05, "loss": 0.0135, "step": 14590 }, { "grad_norm": 0.2195478081703186, "learning_rate": 7.524184522182793e-05, "loss": 0.0134, "step": 14600 }, { "grad_norm": 0.15868331491947174, "learning_rate": 7.520615411134112e-05, "loss": 0.0151, "step": 14610 }, { "grad_norm": 0.2748250365257263, "learning_rate": 7.517044577268564e-05, "loss": 0.0164, "step": 14620 }, { "grad_norm": 0.23611576855182648, "learning_rate": 7.513472023026782e-05, "loss": 0.0166, "step": 14630 }, { "grad_norm": 0.2137519121170044, "learning_rate": 7.509897750850572e-05, "loss": 0.0147, "step": 14640 }, { "grad_norm": 0.2249249964952469, "learning_rate": 7.506321763182918e-05, "loss": 0.0129, "step": 14650 }, { "grad_norm": 0.18097388744354248, "learning_rate": 7.50274406246797e-05, "loss": 0.0177, "step": 14660 }, { "grad_norm": 0.1975650042295456, "learning_rate": 7.499164651151056e-05, "loss": 0.0136, "step": 14670 }, { "grad_norm": 0.15203598141670227, "learning_rate": 7.495583531678669e-05, "loss": 0.0147, "step": 14680 }, { "grad_norm": 0.24795982241630554, "learning_rate": 7.492000706498469e-05, "loss": 0.0143, "step": 14690 }, { "grad_norm": 0.2108859419822693, "learning_rate": 7.488416178059284e-05, "loss": 0.015, "step": 14700 }, { "grad_norm": 0.19385579228401184, "learning_rate": 7.484829948811107e-05, "loss": 0.0114, "step": 14710 }, { "grad_norm": 0.17478223145008087, "learning_rate": 7.48124202120509e-05, "loss": 0.0148, "step": 14720 }, { "grad_norm": 0.2175598293542862, "learning_rate": 7.477652397693549e-05, "loss": 0.0142, "step": 14730 }, { "grad_norm": 0.2114301174879074, "learning_rate": 7.474061080729955e-05, "loss": 0.0175, "step": 14740 }, { "grad_norm": 0.22834338247776031, "learning_rate": 7.470468072768941e-05, "loss": 0.014, "step": 14750 }, { "grad_norm": 0.17605258524417877, "learning_rate": 7.466873376266297e-05, "loss": 0.0124, "step": 14760 }, { "grad_norm": 0.4835730791091919, "learning_rate": 7.46327699367896e-05, "loss": 0.0152, "step": 14770 }, { "grad_norm": 0.2071414291858673, "learning_rate": 7.459678927465026e-05, "loss": 0.0143, "step": 14780 }, { "grad_norm": 0.2545996308326721, "learning_rate": 7.456079180083737e-05, "loss": 0.0155, "step": 14790 }, { "grad_norm": 0.22576677799224854, "learning_rate": 7.452477753995489e-05, "loss": 0.0137, "step": 14800 }, { "grad_norm": 0.2258751094341278, "learning_rate": 7.448874651661823e-05, "loss": 0.0133, "step": 14810 }, { "grad_norm": 0.27315032482147217, "learning_rate": 7.445269875545423e-05, "loss": 0.0163, "step": 14820 }, { "grad_norm": 0.22070413827896118, "learning_rate": 7.44166342811012e-05, "loss": 0.0149, "step": 14830 }, { "grad_norm": 0.1733708530664444, "learning_rate": 7.438055311820886e-05, "loss": 0.0156, "step": 14840 }, { "grad_norm": 0.1697515845298767, "learning_rate": 7.434445529143837e-05, "loss": 0.0141, "step": 14850 }, { "grad_norm": 0.14587059617042542, "learning_rate": 7.430834082546225e-05, "loss": 0.0115, "step": 14860 }, { "grad_norm": 0.16969414055347443, "learning_rate": 7.427220974496438e-05, "loss": 0.0114, "step": 14870 }, { "grad_norm": 0.22311609983444214, "learning_rate": 7.423606207464005e-05, "loss": 0.0147, "step": 14880 }, { "grad_norm": 0.18424567580223083, "learning_rate": 7.419989783919578e-05, "loss": 0.0148, "step": 14890 }, { "grad_norm": 0.17235246300697327, "learning_rate": 7.416371706334956e-05, "loss": 0.0139, "step": 14900 }, { "grad_norm": 0.18654458224773407, "learning_rate": 7.412751977183056e-05, "loss": 0.0128, "step": 14910 }, { "grad_norm": 0.20223529636859894, "learning_rate": 7.409130598937932e-05, "loss": 0.0169, "step": 14920 }, { "grad_norm": 0.22764131426811218, "learning_rate": 7.40550757407476e-05, "loss": 0.016, "step": 14930 }, { "grad_norm": 0.20072123408317566, "learning_rate": 7.401882905069843e-05, "loss": 0.0168, "step": 14940 }, { "grad_norm": 0.2678380012512207, "learning_rate": 7.39825659440061e-05, "loss": 0.0134, "step": 14950 }, { "grad_norm": 0.2489381730556488, "learning_rate": 7.394628644545609e-05, "loss": 0.0139, "step": 14960 }, { "grad_norm": 0.17469072341918945, "learning_rate": 7.390999057984507e-05, "loss": 0.0196, "step": 14970 }, { "grad_norm": 0.16393601894378662, "learning_rate": 7.387367837198097e-05, "loss": 0.0146, "step": 14980 }, { "grad_norm": 0.2066475749015808, "learning_rate": 7.383734984668281e-05, "loss": 0.0132, "step": 14990 }, { "grad_norm": 0.22261108458042145, "learning_rate": 7.38010050287808e-05, "loss": 0.0162, "step": 15000 }, { "grad_norm": 0.18373098969459534, "learning_rate": 7.376464394311628e-05, "loss": 0.0131, "step": 15010 }, { "grad_norm": 0.21188947558403015, "learning_rate": 7.372826661454172e-05, "loss": 0.0146, "step": 15020 }, { "grad_norm": 0.2252521812915802, "learning_rate": 7.369187306792068e-05, "loss": 0.0131, "step": 15030 }, { "grad_norm": 0.18310433626174927, "learning_rate": 7.365546332812779e-05, "loss": 0.0122, "step": 15040 }, { "grad_norm": 0.22318659722805023, "learning_rate": 7.361903742004876e-05, "loss": 0.0142, "step": 15050 }, { "grad_norm": 0.2491396963596344, "learning_rate": 7.358259536858039e-05, "loss": 0.0123, "step": 15060 }, { "grad_norm": 0.22390137612819672, "learning_rate": 7.354613719863044e-05, "loss": 0.0136, "step": 15070 }, { "grad_norm": 0.19221873581409454, "learning_rate": 7.350966293511776e-05, "loss": 0.014, "step": 15080 }, { "grad_norm": 0.22375787794589996, "learning_rate": 7.347317260297212e-05, "loss": 0.0132, "step": 15090 }, { "grad_norm": 0.21130616962909698, "learning_rate": 7.343666622713437e-05, "loss": 0.0136, "step": 15100 }, { "grad_norm": 0.19023920595645905, "learning_rate": 7.340014383255624e-05, "loss": 0.0144, "step": 15110 }, { "grad_norm": 0.17112010717391968, "learning_rate": 7.336360544420044e-05, "loss": 0.0138, "step": 15120 }, { "grad_norm": 0.2038303017616272, "learning_rate": 7.332705108704064e-05, "loss": 0.0146, "step": 15130 }, { "grad_norm": 0.25066861510276794, "learning_rate": 7.329048078606138e-05, "loss": 0.0186, "step": 15140 }, { "grad_norm": 0.16552118957042694, "learning_rate": 7.32538945662581e-05, "loss": 0.0176, "step": 15150 }, { "grad_norm": 0.25236818194389343, "learning_rate": 7.321729245263718e-05, "loss": 0.0182, "step": 15160 }, { "grad_norm": 0.2223636358976364, "learning_rate": 7.318067447021578e-05, "loss": 0.0151, "step": 15170 }, { "grad_norm": 0.2950967252254486, "learning_rate": 7.314404064402198e-05, "loss": 0.0155, "step": 15180 }, { "grad_norm": 0.1724252700805664, "learning_rate": 7.310739099909461e-05, "loss": 0.0174, "step": 15190 }, { "grad_norm": 0.22409798204898834, "learning_rate": 7.307072556048339e-05, "loss": 0.0138, "step": 15200 }, { "grad_norm": 0.16798049211502075, "learning_rate": 7.30340443532488e-05, "loss": 0.0155, "step": 15210 }, { "grad_norm": 0.21218204498291016, "learning_rate": 7.299734740246208e-05, "loss": 0.014, "step": 15220 }, { "grad_norm": 0.18590475618839264, "learning_rate": 7.296063473320528e-05, "loss": 0.0142, "step": 15230 }, { "grad_norm": 0.20204861462116241, "learning_rate": 7.292390637057113e-05, "loss": 0.0143, "step": 15240 }, { "grad_norm": 0.23921939730644226, "learning_rate": 7.288716233966314e-05, "loss": 0.0141, "step": 15250 }, { "grad_norm": 0.2131955772638321, "learning_rate": 7.285040266559551e-05, "loss": 0.0144, "step": 15260 }, { "grad_norm": 0.18642555177211761, "learning_rate": 7.281362737349312e-05, "loss": 0.0142, "step": 15270 }, { "grad_norm": 0.20207498967647552, "learning_rate": 7.277683648849153e-05, "loss": 0.0149, "step": 15280 }, { "grad_norm": 0.1917656660079956, "learning_rate": 7.2740030035737e-05, "loss": 0.0137, "step": 15290 }, { "grad_norm": 0.18923576176166534, "learning_rate": 7.270320804038634e-05, "loss": 0.0137, "step": 15300 }, { "grad_norm": 0.17901380360126495, "learning_rate": 7.266637052760708e-05, "loss": 0.0121, "step": 15310 }, { "grad_norm": 0.21175146102905273, "learning_rate": 7.262951752257728e-05, "loss": 0.0149, "step": 15320 }, { "grad_norm": 0.16930100321769714, "learning_rate": 7.259264905048564e-05, "loss": 0.013, "step": 15330 }, { "grad_norm": 0.23241311311721802, "learning_rate": 7.255576513653142e-05, "loss": 0.0146, "step": 15340 }, { "grad_norm": 0.2048899531364441, "learning_rate": 7.251886580592439e-05, "loss": 0.0147, "step": 15350 }, { "grad_norm": 0.21799498796463013, "learning_rate": 7.248195108388496e-05, "loss": 0.0143, "step": 15360 }, { "grad_norm": 0.19162389636039734, "learning_rate": 7.244502099564395e-05, "loss": 0.0171, "step": 15370 }, { "grad_norm": 0.28924089670181274, "learning_rate": 7.240807556644271e-05, "loss": 0.0169, "step": 15380 }, { "grad_norm": 0.1777907758951187, "learning_rate": 7.237111482153314e-05, "loss": 0.0141, "step": 15390 }, { "grad_norm": 0.16986696422100067, "learning_rate": 7.233413878617751e-05, "loss": 0.0163, "step": 15400 }, { "grad_norm": 0.20273634791374207, "learning_rate": 7.229714748564864e-05, "loss": 0.0151, "step": 15410 }, { "grad_norm": 0.23767301440238953, "learning_rate": 7.22601409452297e-05, "loss": 0.0143, "step": 15420 }, { "grad_norm": 0.13322456181049347, "learning_rate": 7.222311919021433e-05, "loss": 0.0148, "step": 15430 }, { "grad_norm": 0.18652357161045074, "learning_rate": 7.218608224590655e-05, "loss": 0.0164, "step": 15440 }, { "grad_norm": 0.23669257760047913, "learning_rate": 7.214903013762074e-05, "loss": 0.0137, "step": 15450 }, { "grad_norm": 0.16099762916564941, "learning_rate": 7.21119628906817e-05, "loss": 0.0143, "step": 15460 }, { "grad_norm": 0.19298362731933594, "learning_rate": 7.207488053042454e-05, "loss": 0.0164, "step": 15470 }, { "grad_norm": 0.2749035656452179, "learning_rate": 7.203778308219467e-05, "loss": 0.0169, "step": 15480 }, { "grad_norm": 0.19163298606872559, "learning_rate": 7.200067057134787e-05, "loss": 0.0217, "step": 15490 }, { "grad_norm": 0.22860394418239594, "learning_rate": 7.196354302325019e-05, "loss": 0.0181, "step": 15500 }, { "grad_norm": 0.1966446489095688, "learning_rate": 7.192640046327795e-05, "loss": 0.0166, "step": 15510 }, { "grad_norm": 0.15251299738883972, "learning_rate": 7.188924291681777e-05, "loss": 0.0149, "step": 15520 }, { "grad_norm": 0.20426134765148163, "learning_rate": 7.185207040926643e-05, "loss": 0.0145, "step": 15530 }, { "grad_norm": 0.1337195783853531, "learning_rate": 7.181488296603103e-05, "loss": 0.0128, "step": 15540 }, { "grad_norm": 0.17039400339126587, "learning_rate": 7.177768061252885e-05, "loss": 0.0151, "step": 15550 }, { "grad_norm": 0.19700536131858826, "learning_rate": 7.174046337418729e-05, "loss": 0.0136, "step": 15560 }, { "grad_norm": 0.15868952870368958, "learning_rate": 7.170323127644403e-05, "loss": 0.0127, "step": 15570 }, { "grad_norm": 0.1963389664888382, "learning_rate": 7.166598434474683e-05, "loss": 0.012, "step": 15580 }, { "grad_norm": 0.1975003331899643, "learning_rate": 7.162872260455364e-05, "loss": 0.0148, "step": 15590 }, { "grad_norm": 0.17942415177822113, "learning_rate": 7.159144608133248e-05, "loss": 0.0154, "step": 15600 }, { "grad_norm": 0.23926803469657898, "learning_rate": 7.155415480056153e-05, "loss": 0.0152, "step": 15610 }, { "grad_norm": 0.16099141538143158, "learning_rate": 7.151684878772902e-05, "loss": 0.0122, "step": 15620 }, { "grad_norm": 0.15148788690567017, "learning_rate": 7.147952806833324e-05, "loss": 0.014, "step": 15630 }, { "grad_norm": 0.20320595800876617, "learning_rate": 7.14421926678826e-05, "loss": 0.0145, "step": 15640 }, { "grad_norm": 0.24275369942188263, "learning_rate": 7.140484261189543e-05, "loss": 0.0145, "step": 15650 }, { "grad_norm": 0.2387792319059372, "learning_rate": 7.136747792590017e-05, "loss": 0.0157, "step": 15660 }, { "grad_norm": 0.22089146077632904, "learning_rate": 7.133009863543524e-05, "loss": 0.0149, "step": 15670 }, { "grad_norm": 0.22362515330314636, "learning_rate": 7.129270476604901e-05, "loss": 0.0145, "step": 15680 }, { "grad_norm": 0.1777709275484085, "learning_rate": 7.125529634329988e-05, "loss": 0.014, "step": 15690 }, { "grad_norm": 0.1857052743434906, "learning_rate": 7.12178733927561e-05, "loss": 0.0114, "step": 15700 }, { "grad_norm": 0.19088907539844513, "learning_rate": 7.118043593999593e-05, "loss": 0.0122, "step": 15710 }, { "grad_norm": 0.18900533020496368, "learning_rate": 7.114298401060752e-05, "loss": 0.0137, "step": 15720 }, { "grad_norm": 0.2339150458574295, "learning_rate": 7.11055176301889e-05, "loss": 0.0156, "step": 15730 }, { "grad_norm": 0.20475392043590546, "learning_rate": 7.1068036824348e-05, "loss": 0.0139, "step": 15740 }, { "grad_norm": 0.18147680163383484, "learning_rate": 7.10305416187026e-05, "loss": 0.0132, "step": 15750 }, { "grad_norm": 0.2367795705795288, "learning_rate": 7.099303203888029e-05, "loss": 0.0156, "step": 15760 }, { "grad_norm": 0.14591412246227264, "learning_rate": 7.095550811051855e-05, "loss": 0.0147, "step": 15770 }, { "grad_norm": 0.19061103463172913, "learning_rate": 7.09179698592646e-05, "loss": 0.0139, "step": 15780 }, { "grad_norm": 0.19827157258987427, "learning_rate": 7.088041731077551e-05, "loss": 0.0152, "step": 15790 }, { "grad_norm": 0.255199670791626, "learning_rate": 7.084285049071806e-05, "loss": 0.0141, "step": 15800 }, { "grad_norm": 0.17121218144893646, "learning_rate": 7.080526942476886e-05, "loss": 0.013, "step": 15810 }, { "grad_norm": 0.22335653007030487, "learning_rate": 7.076767413861418e-05, "loss": 0.0175, "step": 15820 }, { "grad_norm": 0.16917935013771057, "learning_rate": 7.073006465795005e-05, "loss": 0.0119, "step": 15830 }, { "grad_norm": 0.1905422806739807, "learning_rate": 7.06924410084822e-05, "loss": 0.0179, "step": 15840 }, { "grad_norm": 0.1386592537164688, "learning_rate": 7.065480321592604e-05, "loss": 0.0139, "step": 15850 }, { "grad_norm": 0.20134294033050537, "learning_rate": 7.061715130600663e-05, "loss": 0.0142, "step": 15860 }, { "grad_norm": 0.16774122416973114, "learning_rate": 7.057948530445873e-05, "loss": 0.013, "step": 15870 }, { "grad_norm": 0.2175653874874115, "learning_rate": 7.054180523702668e-05, "loss": 0.0155, "step": 15880 }, { "grad_norm": 0.20883457362651825, "learning_rate": 7.050411112946442e-05, "loss": 0.0144, "step": 15890 }, { "grad_norm": 0.2596830725669861, "learning_rate": 7.046640300753557e-05, "loss": 0.0186, "step": 15900 }, { "grad_norm": 0.22651982307434082, "learning_rate": 7.042868089701325e-05, "loss": 0.0162, "step": 15910 }, { "grad_norm": 0.19964386522769928, "learning_rate": 7.039094482368016e-05, "loss": 0.0194, "step": 15920 }, { "grad_norm": 0.22633415460586548, "learning_rate": 7.035319481332858e-05, "loss": 0.0147, "step": 15930 }, { "grad_norm": 0.29360589385032654, "learning_rate": 7.031543089176023e-05, "loss": 0.0142, "step": 15940 }, { "grad_norm": 0.184280127286911, "learning_rate": 7.027765308478644e-05, "loss": 0.0167, "step": 15950 }, { "grad_norm": 0.19031444191932678, "learning_rate": 7.023986141822798e-05, "loss": 0.0138, "step": 15960 }, { "grad_norm": 0.18177564442157745, "learning_rate": 7.02020559179151e-05, "loss": 0.0147, "step": 15970 }, { "grad_norm": 0.20401884615421295, "learning_rate": 7.016423660968748e-05, "loss": 0.0139, "step": 15980 }, { "grad_norm": 0.17849060893058777, "learning_rate": 7.012640351939428e-05, "loss": 0.0139, "step": 15990 }, { "grad_norm": 0.199040949344635, "learning_rate": 7.008855667289404e-05, "loss": 0.0167, "step": 16000 }, { "grad_norm": 0.23821081221103668, "learning_rate": 7.005069609605476e-05, "loss": 0.0146, "step": 16010 }, { "grad_norm": 0.19873157143592834, "learning_rate": 7.001282181475377e-05, "loss": 0.016, "step": 16020 }, { "grad_norm": 0.1951589733362198, "learning_rate": 6.997493385487775e-05, "loss": 0.0131, "step": 16030 }, { "grad_norm": 0.18429474532604218, "learning_rate": 6.99370322423228e-05, "loss": 0.0141, "step": 16040 }, { "grad_norm": 0.1910611242055893, "learning_rate": 6.989911700299433e-05, "loss": 0.0137, "step": 16050 }, { "grad_norm": 0.25028324127197266, "learning_rate": 6.9861188162807e-05, "loss": 0.0152, "step": 16060 }, { "grad_norm": 0.15786105394363403, "learning_rate": 6.982324574768487e-05, "loss": 0.0122, "step": 16070 }, { "grad_norm": 0.24203747510910034, "learning_rate": 6.978528978356117e-05, "loss": 0.0145, "step": 16080 }, { "grad_norm": 0.19641628861427307, "learning_rate": 6.974732029637846e-05, "loss": 0.0123, "step": 16090 }, { "grad_norm": 0.20186905562877655, "learning_rate": 6.970933731208855e-05, "loss": 0.0131, "step": 16100 }, { "grad_norm": 0.26690346002578735, "learning_rate": 6.967134085665244e-05, "loss": 0.0135, "step": 16110 }, { "grad_norm": 0.18945716321468353, "learning_rate": 6.963333095604034e-05, "loss": 0.0131, "step": 16120 }, { "grad_norm": 0.19374299049377441, "learning_rate": 6.959530763623166e-05, "loss": 0.014, "step": 16130 }, { "grad_norm": 0.20484396815299988, "learning_rate": 6.955727092321497e-05, "loss": 0.0136, "step": 16140 }, { "grad_norm": 0.20674289762973785, "learning_rate": 6.951922084298803e-05, "loss": 0.0139, "step": 16150 }, { "grad_norm": 0.23511150479316711, "learning_rate": 6.948115742155769e-05, "loss": 0.0148, "step": 16160 }, { "grad_norm": 0.1585136353969574, "learning_rate": 6.944308068493996e-05, "loss": 0.0123, "step": 16170 }, { "grad_norm": 0.2932646572589874, "learning_rate": 6.940499065915992e-05, "loss": 0.0151, "step": 16180 }, { "grad_norm": 0.2037735879421234, "learning_rate": 6.936688737025173e-05, "loss": 0.014, "step": 16190 }, { "grad_norm": 0.28493982553482056, "learning_rate": 6.932877084425867e-05, "loss": 0.0136, "step": 16200 }, { "grad_norm": 0.1951891928911209, "learning_rate": 6.929064110723297e-05, "loss": 0.0133, "step": 16210 }, { "grad_norm": 0.18224433064460754, "learning_rate": 6.925249818523598e-05, "loss": 0.0146, "step": 16220 }, { "grad_norm": 0.17020758986473083, "learning_rate": 6.921434210433801e-05, "loss": 0.0143, "step": 16230 }, { "grad_norm": 0.22730571031570435, "learning_rate": 6.917617289061841e-05, "loss": 0.0146, "step": 16240 }, { "grad_norm": 0.20444399118423462, "learning_rate": 6.913799057016547e-05, "loss": 0.0115, "step": 16250 }, { "grad_norm": 0.22195719182491302, "learning_rate": 6.909979516907641e-05, "loss": 0.0166, "step": 16260 }, { "grad_norm": 0.16880939900875092, "learning_rate": 6.906158671345746e-05, "loss": 0.0141, "step": 16270 }, { "grad_norm": 0.19597245752811432, "learning_rate": 6.902336522942374e-05, "loss": 0.0164, "step": 16280 }, { "grad_norm": 0.16364359855651855, "learning_rate": 6.898513074309924e-05, "loss": 0.0114, "step": 16290 }, { "grad_norm": 0.23767156898975372, "learning_rate": 6.894688328061693e-05, "loss": 0.0136, "step": 16300 }, { "grad_norm": 0.1759963482618332, "learning_rate": 6.890862286811853e-05, "loss": 0.0148, "step": 16310 }, { "grad_norm": 0.21437881886959076, "learning_rate": 6.88703495317547e-05, "loss": 0.0132, "step": 16320 }, { "grad_norm": 0.20532171428203583, "learning_rate": 6.883206329768492e-05, "loss": 0.0118, "step": 16330 }, { "grad_norm": 0.14962106943130493, "learning_rate": 6.879376419207743e-05, "loss": 0.0114, "step": 16340 }, { "grad_norm": 0.21749365329742432, "learning_rate": 6.875545224110935e-05, "loss": 0.0142, "step": 16350 }, { "grad_norm": 0.251735121011734, "learning_rate": 6.871712747096651e-05, "loss": 0.0168, "step": 16360 }, { "grad_norm": 0.18739481270313263, "learning_rate": 6.867878990784353e-05, "loss": 0.0119, "step": 16370 }, { "grad_norm": 0.16180357336997986, "learning_rate": 6.864043957794377e-05, "loss": 0.0115, "step": 16380 }, { "grad_norm": 0.22688469290733337, "learning_rate": 6.860207650747934e-05, "loss": 0.0128, "step": 16390 }, { "grad_norm": 0.15962931513786316, "learning_rate": 6.856370072267104e-05, "loss": 0.0171, "step": 16400 }, { "grad_norm": 0.25713393092155457, "learning_rate": 6.852531224974831e-05, "loss": 0.0154, "step": 16410 }, { "grad_norm": 0.2280074656009674, "learning_rate": 6.848691111494936e-05, "loss": 0.0151, "step": 16420 }, { "grad_norm": 0.19119304418563843, "learning_rate": 6.844849734452097e-05, "loss": 0.0115, "step": 16430 }, { "grad_norm": 0.1938064843416214, "learning_rate": 6.841007096471862e-05, "loss": 0.012, "step": 16440 }, { "grad_norm": 0.17371268570423126, "learning_rate": 6.837163200180636e-05, "loss": 0.0152, "step": 16450 }, { "grad_norm": 0.25815272331237793, "learning_rate": 6.833318048205684e-05, "loss": 0.0168, "step": 16460 }, { "grad_norm": 0.1749637871980667, "learning_rate": 6.829471643175136e-05, "loss": 0.0131, "step": 16470 }, { "grad_norm": 0.14513781666755676, "learning_rate": 6.825623987717969e-05, "loss": 0.0138, "step": 16480 }, { "grad_norm": 0.1549557000398636, "learning_rate": 6.821775084464022e-05, "loss": 0.0123, "step": 16490 }, { "grad_norm": 0.16850869357585907, "learning_rate": 6.817924936043982e-05, "loss": 0.0124, "step": 16500 }, { "grad_norm": 0.21592403948307037, "learning_rate": 6.81407354508939e-05, "loss": 0.0099, "step": 16510 }, { "grad_norm": 0.1620928943157196, "learning_rate": 6.810220914232636e-05, "loss": 0.0132, "step": 16520 }, { "grad_norm": 0.20565897226333618, "learning_rate": 6.806367046106959e-05, "loss": 0.0148, "step": 16530 }, { "grad_norm": 0.2001412808895111, "learning_rate": 6.802511943346435e-05, "loss": 0.0128, "step": 16540 }, { "grad_norm": 0.12873278558254242, "learning_rate": 6.798655608585997e-05, "loss": 0.01, "step": 16550 }, { "grad_norm": 0.22490514814853668, "learning_rate": 6.79479804446141e-05, "loss": 0.0138, "step": 16560 }, { "grad_norm": 0.16655392944812775, "learning_rate": 6.790939253609284e-05, "loss": 0.014, "step": 16570 }, { "grad_norm": 0.16443310678005219, "learning_rate": 6.787079238667065e-05, "loss": 0.012, "step": 16580 }, { "grad_norm": 0.17688758671283722, "learning_rate": 6.783218002273039e-05, "loss": 0.0125, "step": 16590 }, { "grad_norm": 0.20129631459712982, "learning_rate": 6.779355547066322e-05, "loss": 0.013, "step": 16600 }, { "grad_norm": 0.20195087790489197, "learning_rate": 6.775491875686865e-05, "loss": 0.0106, "step": 16610 }, { "grad_norm": 0.23164449632167816, "learning_rate": 6.771626990775457e-05, "loss": 0.0132, "step": 16620 }, { "grad_norm": 0.1742621809244156, "learning_rate": 6.767760894973704e-05, "loss": 0.0157, "step": 16630 }, { "grad_norm": 0.26081007719039917, "learning_rate": 6.763893590924048e-05, "loss": 0.0128, "step": 16640 }, { "grad_norm": 0.21369799971580505, "learning_rate": 6.760025081269756e-05, "loss": 0.0177, "step": 16650 }, { "grad_norm": 0.2116093635559082, "learning_rate": 6.756155368654915e-05, "loss": 0.012, "step": 16660 }, { "grad_norm": 0.2046116590499878, "learning_rate": 6.752284455724442e-05, "loss": 0.0148, "step": 16670 }, { "grad_norm": 0.17935846745967865, "learning_rate": 6.748412345124065e-05, "loss": 0.0119, "step": 16680 }, { "grad_norm": 0.18672868609428406, "learning_rate": 6.744539039500335e-05, "loss": 0.011, "step": 16690 }, { "grad_norm": 0.23851880431175232, "learning_rate": 6.740664541500625e-05, "loss": 0.0119, "step": 16700 }, { "grad_norm": 0.15583275258541107, "learning_rate": 6.736788853773112e-05, "loss": 0.0137, "step": 16710 }, { "grad_norm": 0.16656583547592163, "learning_rate": 6.732911978966796e-05, "loss": 0.0117, "step": 16720 }, { "grad_norm": 0.23405347764492035, "learning_rate": 6.729033919731482e-05, "loss": 0.0126, "step": 16730 }, { "grad_norm": 0.15477243065834045, "learning_rate": 6.725154678717787e-05, "loss": 0.0157, "step": 16740 }, { "grad_norm": 0.20349405705928802, "learning_rate": 6.721274258577138e-05, "loss": 0.0132, "step": 16750 }, { "grad_norm": 0.1810123324394226, "learning_rate": 6.717392661961763e-05, "loss": 0.0134, "step": 16760 }, { "grad_norm": 0.2245379090309143, "learning_rate": 6.713509891524697e-05, "loss": 0.0135, "step": 16770 }, { "grad_norm": 0.23547479510307312, "learning_rate": 6.709625949919777e-05, "loss": 0.0143, "step": 16780 }, { "grad_norm": 0.18447837233543396, "learning_rate": 6.705740839801642e-05, "loss": 0.0129, "step": 16790 }, { "grad_norm": 0.20924361050128937, "learning_rate": 6.701854563825727e-05, "loss": 0.0127, "step": 16800 }, { "grad_norm": 0.20447467267513275, "learning_rate": 6.697967124648266e-05, "loss": 0.0092, "step": 16810 }, { "grad_norm": 0.19971248507499695, "learning_rate": 6.694078524926285e-05, "loss": 0.0122, "step": 16820 }, { "grad_norm": 0.2025611251592636, "learning_rate": 6.690188767317607e-05, "loss": 0.0133, "step": 16830 }, { "grad_norm": 0.15331822633743286, "learning_rate": 6.686297854480843e-05, "loss": 0.0113, "step": 16840 }, { "grad_norm": 0.17967984080314636, "learning_rate": 6.682405789075398e-05, "loss": 0.0147, "step": 16850 }, { "grad_norm": 0.2112436294555664, "learning_rate": 6.67851257376146e-05, "loss": 0.0122, "step": 16860 }, { "grad_norm": 0.16308879852294922, "learning_rate": 6.674618211200004e-05, "loss": 0.0123, "step": 16870 }, { "grad_norm": 0.18752634525299072, "learning_rate": 6.670722704052792e-05, "loss": 0.0134, "step": 16880 }, { "grad_norm": 0.1522303968667984, "learning_rate": 6.666826054982365e-05, "loss": 0.0128, "step": 16890 }, { "grad_norm": 0.191489577293396, "learning_rate": 6.662928266652048e-05, "loss": 0.0138, "step": 16900 }, { "grad_norm": 0.19635926187038422, "learning_rate": 6.659029341725941e-05, "loss": 0.0118, "step": 16910 }, { "grad_norm": 0.2180328220129013, "learning_rate": 6.655129282868923e-05, "loss": 0.0171, "step": 16920 }, { "grad_norm": 0.18950976431369781, "learning_rate": 6.651228092746646e-05, "loss": 0.0117, "step": 16930 }, { "grad_norm": 0.1628253310918808, "learning_rate": 6.647325774025539e-05, "loss": 0.0121, "step": 16940 }, { "grad_norm": 0.18169161677360535, "learning_rate": 6.643422329372798e-05, "loss": 0.0144, "step": 16950 }, { "grad_norm": 0.16343605518341064, "learning_rate": 6.639517761456392e-05, "loss": 0.0144, "step": 16960 }, { "grad_norm": 0.18663516640663147, "learning_rate": 6.635612072945054e-05, "loss": 0.0148, "step": 16970 }, { "grad_norm": 0.14159619808197021, "learning_rate": 6.631705266508289e-05, "loss": 0.0111, "step": 16980 }, { "grad_norm": 0.22337856888771057, "learning_rate": 6.62779734481636e-05, "loss": 0.0115, "step": 16990 }, { "grad_norm": 0.20475469529628754, "learning_rate": 6.623888310540294e-05, "loss": 0.0137, "step": 17000 }, { "grad_norm": 0.21310046315193176, "learning_rate": 6.619978166351882e-05, "loss": 0.0181, "step": 17010 }, { "grad_norm": 0.2212187796831131, "learning_rate": 6.616066914923666e-05, "loss": 0.0141, "step": 17020 }, { "grad_norm": 0.22463467717170715, "learning_rate": 6.612154558928955e-05, "loss": 0.0119, "step": 17030 }, { "grad_norm": 0.18792827427387238, "learning_rate": 6.608241101041804e-05, "loss": 0.0144, "step": 17040 }, { "grad_norm": 0.20782551169395447, "learning_rate": 6.604326543937025e-05, "loss": 0.0141, "step": 17050 }, { "grad_norm": 0.18617777526378632, "learning_rate": 6.60041089029018e-05, "loss": 0.0152, "step": 17060 }, { "grad_norm": 0.21797718107700348, "learning_rate": 6.596494142777583e-05, "loss": 0.0153, "step": 17070 }, { "grad_norm": 0.20221228897571564, "learning_rate": 6.592576304076294e-05, "loss": 0.0122, "step": 17080 }, { "grad_norm": 0.19352906942367554, "learning_rate": 6.588657376864119e-05, "loss": 0.0148, "step": 17090 }, { "grad_norm": 0.15120241045951843, "learning_rate": 6.584737363819605e-05, "loss": 0.0116, "step": 17100 }, { "grad_norm": 0.2440505027770996, "learning_rate": 6.580816267622048e-05, "loss": 0.0108, "step": 17110 }, { "grad_norm": 0.24981343746185303, "learning_rate": 6.576894090951478e-05, "loss": 0.0136, "step": 17120 }, { "grad_norm": 0.1805567592382431, "learning_rate": 6.572970836488665e-05, "loss": 0.0128, "step": 17130 }, { "grad_norm": 0.2809458076953888, "learning_rate": 6.569046506915119e-05, "loss": 0.0148, "step": 17140 }, { "grad_norm": 0.19754429161548615, "learning_rate": 6.56512110491308e-05, "loss": 0.0133, "step": 17150 }, { "grad_norm": 0.1600060909986496, "learning_rate": 6.561194633165523e-05, "loss": 0.0134, "step": 17160 }, { "grad_norm": 0.1367950588464737, "learning_rate": 6.557267094356155e-05, "loss": 0.0112, "step": 17170 }, { "grad_norm": 0.15937110781669617, "learning_rate": 6.553338491169414e-05, "loss": 0.0121, "step": 17180 }, { "grad_norm": 0.20227207243442535, "learning_rate": 6.54940882629046e-05, "loss": 0.0168, "step": 17190 }, { "grad_norm": 0.20732130110263824, "learning_rate": 6.545478102405184e-05, "loss": 0.0114, "step": 17200 }, { "grad_norm": 0.17297974228858948, "learning_rate": 6.541546322200199e-05, "loss": 0.0102, "step": 17210 }, { "grad_norm": 0.17328256368637085, "learning_rate": 6.537613488362837e-05, "loss": 0.0127, "step": 17220 }, { "grad_norm": 0.16467073559761047, "learning_rate": 6.533679603581155e-05, "loss": 0.0143, "step": 17230 }, { "grad_norm": 0.15183262526988983, "learning_rate": 6.529744670543926e-05, "loss": 0.0146, "step": 17240 }, { "grad_norm": 0.19755159318447113, "learning_rate": 6.52580869194064e-05, "loss": 0.0117, "step": 17250 }, { "grad_norm": 0.18123239278793335, "learning_rate": 6.521871670461499e-05, "loss": 0.0131, "step": 17260 }, { "grad_norm": 0.21872568130493164, "learning_rate": 6.517933608797422e-05, "loss": 0.013, "step": 17270 }, { "grad_norm": 0.16498662531375885, "learning_rate": 6.513994509640038e-05, "loss": 0.0121, "step": 17280 }, { "grad_norm": 0.1936172991991043, "learning_rate": 6.510054375681682e-05, "loss": 0.0109, "step": 17290 }, { "grad_norm": 0.18151947855949402, "learning_rate": 6.506113209615398e-05, "loss": 0.0141, "step": 17300 }, { "grad_norm": 0.17706139385700226, "learning_rate": 6.502171014134938e-05, "loss": 0.011, "step": 17310 }, { "grad_norm": 0.23204928636550903, "learning_rate": 6.498227791934755e-05, "loss": 0.0119, "step": 17320 }, { "grad_norm": 0.2321646511554718, "learning_rate": 6.494283545710003e-05, "loss": 0.0149, "step": 17330 }, { "grad_norm": 0.2419634908437729, "learning_rate": 6.490338278156538e-05, "loss": 0.0125, "step": 17340 }, { "grad_norm": 0.19513872265815735, "learning_rate": 6.486391991970913e-05, "loss": 0.0143, "step": 17350 }, { "grad_norm": 0.1488119661808014, "learning_rate": 6.482444689850377e-05, "loss": 0.0121, "step": 17360 }, { "grad_norm": 0.2215634137392044, "learning_rate": 6.478496374492875e-05, "loss": 0.0114, "step": 17370 }, { "grad_norm": 0.1828773468732834, "learning_rate": 6.474547048597042e-05, "loss": 0.0127, "step": 17380 }, { "grad_norm": 0.1709953248500824, "learning_rate": 6.470596714862205e-05, "loss": 0.0131, "step": 17390 }, { "grad_norm": 0.19277705252170563, "learning_rate": 6.46664537598838e-05, "loss": 0.0152, "step": 17400 }, { "grad_norm": 0.16604217886924744, "learning_rate": 6.462693034676271e-05, "loss": 0.0125, "step": 17410 }, { "grad_norm": 0.20147918164730072, "learning_rate": 6.458739693627265e-05, "loss": 0.0117, "step": 17420 }, { "grad_norm": 0.24300579726696014, "learning_rate": 6.454785355543432e-05, "loss": 0.0118, "step": 17430 }, { "grad_norm": 0.17147789895534515, "learning_rate": 6.450830023127528e-05, "loss": 0.013, "step": 17440 }, { "grad_norm": 0.23333947360515594, "learning_rate": 6.446873699082982e-05, "loss": 0.0139, "step": 17450 }, { "grad_norm": 0.12689687311649323, "learning_rate": 6.44291638611391e-05, "loss": 0.011, "step": 17460 }, { "grad_norm": 0.16544483602046967, "learning_rate": 6.43895808692509e-05, "loss": 0.0131, "step": 17470 }, { "grad_norm": 0.17457318305969238, "learning_rate": 6.434998804221986e-05, "loss": 0.0129, "step": 17480 }, { "grad_norm": 0.15833328664302826, "learning_rate": 6.431038540710732e-05, "loss": 0.0119, "step": 17490 }, { "grad_norm": 0.16058148443698883, "learning_rate": 6.427077299098129e-05, "loss": 0.0127, "step": 17500 }, { "grad_norm": 0.2213699221611023, "learning_rate": 6.423115082091651e-05, "loss": 0.0133, "step": 17510 }, { "grad_norm": 0.21699084341526031, "learning_rate": 6.419151892399429e-05, "loss": 0.01, "step": 17520 }, { "grad_norm": 0.22166000306606293, "learning_rate": 6.415187732730273e-05, "loss": 0.014, "step": 17530 }, { "grad_norm": 0.16790813207626343, "learning_rate": 6.411222605793645e-05, "loss": 0.0125, "step": 17540 }, { "grad_norm": 0.18560069799423218, "learning_rate": 6.407256514299674e-05, "loss": 0.0117, "step": 17550 }, { "grad_norm": 0.15461671352386475, "learning_rate": 6.403289460959147e-05, "loss": 0.0121, "step": 17560 }, { "grad_norm": 0.1984506994485855, "learning_rate": 6.399321448483501e-05, "loss": 0.0136, "step": 17570 }, { "grad_norm": 0.15585866570472717, "learning_rate": 6.395352479584844e-05, "loss": 0.0126, "step": 17580 }, { "grad_norm": 0.15272316336631775, "learning_rate": 6.391382556975923e-05, "loss": 0.0128, "step": 17590 }, { "grad_norm": 0.18078581988811493, "learning_rate": 6.387411683370144e-05, "loss": 0.0106, "step": 17600 }, { "grad_norm": 0.2174571305513382, "learning_rate": 6.383439861481562e-05, "loss": 0.0126, "step": 17610 }, { "grad_norm": 0.18889717757701874, "learning_rate": 6.379467094024879e-05, "loss": 0.0114, "step": 17620 }, { "grad_norm": 0.19497686624526978, "learning_rate": 6.375493383715445e-05, "loss": 0.0135, "step": 17630 }, { "grad_norm": 0.15387728810310364, "learning_rate": 6.371518733269254e-05, "loss": 0.0124, "step": 17640 }, { "grad_norm": 0.17589683830738068, "learning_rate": 6.367543145402942e-05, "loss": 0.0093, "step": 17650 }, { "grad_norm": 0.13694395124912262, "learning_rate": 6.363566622833785e-05, "loss": 0.0098, "step": 17660 }, { "grad_norm": 0.17408420145511627, "learning_rate": 6.359589168279698e-05, "loss": 0.0123, "step": 17670 }, { "grad_norm": 0.1754775196313858, "learning_rate": 6.355610784459235e-05, "loss": 0.0151, "step": 17680 }, { "grad_norm": 0.20781032741069794, "learning_rate": 6.351631474091585e-05, "loss": 0.0125, "step": 17690 }, { "grad_norm": 0.1928866058588028, "learning_rate": 6.347651239896566e-05, "loss": 0.0149, "step": 17700 }, { "grad_norm": 0.17636334896087646, "learning_rate": 6.343670084594633e-05, "loss": 0.0117, "step": 17710 }, { "grad_norm": 0.15786302089691162, "learning_rate": 6.339688010906866e-05, "loss": 0.0139, "step": 17720 }, { "grad_norm": 0.15959495306015015, "learning_rate": 6.335705021554975e-05, "loss": 0.0115, "step": 17730 }, { "grad_norm": 0.18510645627975464, "learning_rate": 6.3317211192613e-05, "loss": 0.0165, "step": 17740 }, { "grad_norm": 0.27180424332618713, "learning_rate": 6.327736306748795e-05, "loss": 0.0117, "step": 17750 }, { "grad_norm": 0.19801484048366547, "learning_rate": 6.323750586741047e-05, "loss": 0.0182, "step": 17760 }, { "grad_norm": 0.17422282695770264, "learning_rate": 6.319763961962252e-05, "loss": 0.0135, "step": 17770 }, { "grad_norm": 0.18428108096122742, "learning_rate": 6.315776435137233e-05, "loss": 0.0126, "step": 17780 }, { "grad_norm": 0.14954443275928497, "learning_rate": 6.311788008991432e-05, "loss": 0.0132, "step": 17790 }, { "grad_norm": 0.1650993376970291, "learning_rate": 6.307798686250891e-05, "loss": 0.0113, "step": 17800 }, { "grad_norm": 0.16485588252544403, "learning_rate": 6.303808469642284e-05, "loss": 0.0144, "step": 17810 }, { "grad_norm": 0.15303447842597961, "learning_rate": 6.29981736189288e-05, "loss": 0.0125, "step": 17820 }, { "grad_norm": 0.1781272143125534, "learning_rate": 6.295825365730567e-05, "loss": 0.0097, "step": 17830 }, { "grad_norm": 0.17429302632808685, "learning_rate": 6.291832483883835e-05, "loss": 0.013, "step": 17840 }, { "grad_norm": 0.1768922656774521, "learning_rate": 6.28783871908178e-05, "loss": 0.0118, "step": 17850 }, { "grad_norm": 0.2127104252576828, "learning_rate": 6.283844074054107e-05, "loss": 0.0128, "step": 17860 }, { "grad_norm": 0.2583281993865967, "learning_rate": 6.279848551531112e-05, "loss": 0.0144, "step": 17870 }, { "grad_norm": 0.27096232771873474, "learning_rate": 6.275852154243702e-05, "loss": 0.0112, "step": 17880 }, { "grad_norm": 0.21550710499286652, "learning_rate": 6.271854884923377e-05, "loss": 0.0144, "step": 17890 }, { "grad_norm": 0.20501084625720978, "learning_rate": 6.267856746302228e-05, "loss": 0.0123, "step": 17900 }, { "grad_norm": 0.23835186660289764, "learning_rate": 6.263857741112948e-05, "loss": 0.0138, "step": 17910 }, { "grad_norm": 0.19410638511180878, "learning_rate": 6.259857872088821e-05, "loss": 0.0146, "step": 17920 }, { "grad_norm": 0.17886847257614136, "learning_rate": 6.255857141963719e-05, "loss": 0.012, "step": 17930 }, { "grad_norm": 0.20626698434352875, "learning_rate": 6.251855553472101e-05, "loss": 0.014, "step": 17940 }, { "grad_norm": 0.25731325149536133, "learning_rate": 6.247853109349016e-05, "loss": 0.0106, "step": 17950 }, { "grad_norm": 0.2145681530237198, "learning_rate": 6.243849812330098e-05, "loss": 0.0148, "step": 17960 }, { "grad_norm": 0.1754208207130432, "learning_rate": 6.239845665151563e-05, "loss": 0.0129, "step": 17970 }, { "grad_norm": 0.2137141227722168, "learning_rate": 6.235840670550204e-05, "loss": 0.0122, "step": 17980 }, { "grad_norm": 0.15854111313819885, "learning_rate": 6.231834831263403e-05, "loss": 0.0135, "step": 17990 }, { "grad_norm": 0.1963963806629181, "learning_rate": 6.22782815002911e-05, "loss": 0.0116, "step": 18000 }, { "grad_norm": 0.22625461220741272, "learning_rate": 6.223820629585852e-05, "loss": 0.0137, "step": 18010 }, { "grad_norm": 0.14777448773384094, "learning_rate": 6.219812272672737e-05, "loss": 0.0107, "step": 18020 }, { "grad_norm": 0.1647416055202484, "learning_rate": 6.215803082029434e-05, "loss": 0.0104, "step": 18030 }, { "grad_norm": 0.1810571402311325, "learning_rate": 6.211793060396188e-05, "loss": 0.0113, "step": 18040 }, { "grad_norm": 0.17750869691371918, "learning_rate": 6.207782210513811e-05, "loss": 0.0123, "step": 18050 }, { "grad_norm": 0.23846927285194397, "learning_rate": 6.203770535123683e-05, "loss": 0.0141, "step": 18060 }, { "grad_norm": 0.21702279150485992, "learning_rate": 6.199758036967747e-05, "loss": 0.0143, "step": 18070 }, { "grad_norm": 0.20278654992580414, "learning_rate": 6.195744718788503e-05, "loss": 0.0126, "step": 18080 }, { "grad_norm": 0.21116149425506592, "learning_rate": 6.191730583329021e-05, "loss": 0.0136, "step": 18090 }, { "grad_norm": 0.16216342151165009, "learning_rate": 6.187715633332921e-05, "loss": 0.0137, "step": 18100 }, { "grad_norm": 0.12944698333740234, "learning_rate": 6.183699871544386e-05, "loss": 0.011, "step": 18110 }, { "grad_norm": 0.21081271767616272, "learning_rate": 6.179683300708152e-05, "loss": 0.0171, "step": 18120 }, { "grad_norm": 0.17281553149223328, "learning_rate": 6.175665923569503e-05, "loss": 0.0138, "step": 18130 }, { "grad_norm": 0.20265480875968933, "learning_rate": 6.171647742874281e-05, "loss": 0.0105, "step": 18140 }, { "grad_norm": 0.163585364818573, "learning_rate": 6.167628761368875e-05, "loss": 0.0103, "step": 18150 }, { "grad_norm": 0.23344099521636963, "learning_rate": 6.163608981800222e-05, "loss": 0.0126, "step": 18160 }, { "grad_norm": 0.2004968523979187, "learning_rate": 6.159588406915803e-05, "loss": 0.0119, "step": 18170 }, { "grad_norm": 0.19531548023223877, "learning_rate": 6.155567039463639e-05, "loss": 0.012, "step": 18180 }, { "grad_norm": 0.1526426523923874, "learning_rate": 6.151544882192302e-05, "loss": 0.0129, "step": 18190 }, { "grad_norm": 0.1977100521326065, "learning_rate": 6.147521937850895e-05, "loss": 0.0134, "step": 18200 }, { "grad_norm": 0.1756574809551239, "learning_rate": 6.143498209189066e-05, "loss": 0.0115, "step": 18210 }, { "grad_norm": 0.1687043011188507, "learning_rate": 6.139473698956993e-05, "loss": 0.0125, "step": 18220 }, { "grad_norm": 0.2415481060743332, "learning_rate": 6.13544840990539e-05, "loss": 0.0139, "step": 18230 }, { "grad_norm": 0.1589270383119583, "learning_rate": 6.131422344785507e-05, "loss": 0.013, "step": 18240 }, { "grad_norm": 0.16078968346118927, "learning_rate": 6.127395506349119e-05, "loss": 0.0122, "step": 18250 }, { "grad_norm": 0.1998097449541092, "learning_rate": 6.123367897348533e-05, "loss": 0.0122, "step": 18260 }, { "grad_norm": 0.16920232772827148, "learning_rate": 6.119339520536584e-05, "loss": 0.0129, "step": 18270 }, { "grad_norm": 0.17591418325901031, "learning_rate": 6.115310378666625e-05, "loss": 0.0129, "step": 18280 }, { "grad_norm": 0.17789077758789062, "learning_rate": 6.11128047449254e-05, "loss": 0.0105, "step": 18290 }, { "grad_norm": 0.20240525901317596, "learning_rate": 6.107249810768729e-05, "loss": 0.0111, "step": 18300 }, { "grad_norm": 0.1385553628206253, "learning_rate": 6.1032183902501125e-05, "loss": 0.0161, "step": 18310 }, { "grad_norm": 0.2014526128768921, "learning_rate": 6.099186215692131e-05, "loss": 0.0108, "step": 18320 }, { "grad_norm": 0.1997677981853485, "learning_rate": 6.095153289850734e-05, "loss": 0.0158, "step": 18330 }, { "grad_norm": 0.2064720243215561, "learning_rate": 6.0911196154823904e-05, "loss": 0.0133, "step": 18340 }, { "grad_norm": 0.163669154047966, "learning_rate": 6.087085195344079e-05, "loss": 0.0124, "step": 18350 }, { "grad_norm": 0.24186915159225464, "learning_rate": 6.083050032193286e-05, "loss": 0.0116, "step": 18360 }, { "grad_norm": 0.16124999523162842, "learning_rate": 6.0790141287880097e-05, "loss": 0.0105, "step": 18370 }, { "grad_norm": 0.179641991853714, "learning_rate": 6.0749774878867496e-05, "loss": 0.0125, "step": 18380 }, { "grad_norm": 0.21914787590503693, "learning_rate": 6.0709401122485146e-05, "loss": 0.0098, "step": 18390 }, { "grad_norm": 0.18180200457572937, "learning_rate": 6.066902004632811e-05, "loss": 0.0105, "step": 18400 }, { "grad_norm": 0.24980273842811584, "learning_rate": 6.062863167799646e-05, "loss": 0.0117, "step": 18410 }, { "grad_norm": 0.1986217200756073, "learning_rate": 6.058823604509529e-05, "loss": 0.0115, "step": 18420 }, { "grad_norm": 0.18876828253269196, "learning_rate": 6.054783317523462e-05, "loss": 0.0134, "step": 18430 }, { "grad_norm": 0.1733626127243042, "learning_rate": 6.050742309602944e-05, "loss": 0.0143, "step": 18440 }, { "grad_norm": 0.20131638646125793, "learning_rate": 6.046700583509965e-05, "loss": 0.0132, "step": 18450 }, { "grad_norm": 0.2164369821548462, "learning_rate": 6.042658142007007e-05, "loss": 0.0105, "step": 18460 }, { "grad_norm": 0.1560552418231964, "learning_rate": 6.038614987857041e-05, "loss": 0.0107, "step": 18470 }, { "grad_norm": 0.14458568394184113, "learning_rate": 6.0345711238235224e-05, "loss": 0.0102, "step": 18480 }, { "grad_norm": 0.20590871572494507, "learning_rate": 6.030526552670399e-05, "loss": 0.0142, "step": 18490 }, { "grad_norm": 0.19156810641288757, "learning_rate": 6.0264812771620925e-05, "loss": 0.0117, "step": 18500 }, { "grad_norm": 0.2155051827430725, "learning_rate": 6.022435300063512e-05, "loss": 0.012, "step": 18510 }, { "grad_norm": 0.17904746532440186, "learning_rate": 6.0183886241400466e-05, "loss": 0.0119, "step": 18520 }, { "grad_norm": 0.1678176075220108, "learning_rate": 6.0143412521575584e-05, "loss": 0.0106, "step": 18530 }, { "grad_norm": 0.20904132723808289, "learning_rate": 6.010293186882389e-05, "loss": 0.0122, "step": 18540 }, { "grad_norm": 0.2097642719745636, "learning_rate": 6.0062444310813525e-05, "loss": 0.0135, "step": 18550 }, { "grad_norm": 0.14805833995342255, "learning_rate": 6.0021949875217355e-05, "loss": 0.0118, "step": 18560 }, { "grad_norm": 0.16481482982635498, "learning_rate": 5.998144858971295e-05, "loss": 0.0109, "step": 18570 }, { "grad_norm": 0.1431080549955368, "learning_rate": 5.994094048198257e-05, "loss": 0.0092, "step": 18580 }, { "grad_norm": 0.15488454699516296, "learning_rate": 5.990042557971307e-05, "loss": 0.0114, "step": 18590 }, { "grad_norm": 0.1472177356481552, "learning_rate": 5.985990391059607e-05, "loss": 0.0104, "step": 18600 }, { "grad_norm": 0.15136666595935822, "learning_rate": 5.981937550232771e-05, "loss": 0.0123, "step": 18610 }, { "grad_norm": 0.24515578150749207, "learning_rate": 5.9778840382608794e-05, "loss": 0.0136, "step": 18620 }, { "grad_norm": 0.20459699630737305, "learning_rate": 5.9738298579144695e-05, "loss": 0.0134, "step": 18630 }, { "grad_norm": 0.11956202238798141, "learning_rate": 5.9697750119645314e-05, "loss": 0.0105, "step": 18640 }, { "grad_norm": 0.1594531089067459, "learning_rate": 5.96571950318252e-05, "loss": 0.0113, "step": 18650 }, { "grad_norm": 0.17172035574913025, "learning_rate": 5.9616633343403316e-05, "loss": 0.0115, "step": 18660 }, { "grad_norm": 0.18449990451335907, "learning_rate": 5.957606508210324e-05, "loss": 0.0106, "step": 18670 }, { "grad_norm": 0.19983114302158356, "learning_rate": 5.953549027565297e-05, "loss": 0.0135, "step": 18680 }, { "grad_norm": 0.2143043577671051, "learning_rate": 5.949490895178501e-05, "loss": 0.013, "step": 18690 }, { "grad_norm": 0.18456369638442993, "learning_rate": 5.945432113823632e-05, "loss": 0.0112, "step": 18700 }, { "grad_norm": 0.17830118536949158, "learning_rate": 5.9413726862748276e-05, "loss": 0.0106, "step": 18710 }, { "grad_norm": 0.15156260132789612, "learning_rate": 5.9373126153066694e-05, "loss": 0.0103, "step": 18720 }, { "grad_norm": 0.15723346173763275, "learning_rate": 5.933251903694177e-05, "loss": 0.0141, "step": 18730 }, { "grad_norm": 0.19947652518749237, "learning_rate": 5.929190554212807e-05, "loss": 0.0104, "step": 18740 }, { "grad_norm": 0.21102474629878998, "learning_rate": 5.9251285696384565e-05, "loss": 0.0113, "step": 18750 }, { "grad_norm": 0.13110241293907166, "learning_rate": 5.921065952747451e-05, "loss": 0.0084, "step": 18760 }, { "grad_norm": 0.19377794861793518, "learning_rate": 5.917002706316552e-05, "loss": 0.01, "step": 18770 }, { "grad_norm": 0.20052604377269745, "learning_rate": 5.912938833122952e-05, "loss": 0.0107, "step": 18780 }, { "grad_norm": 0.1650138944387436, "learning_rate": 5.908874335944265e-05, "loss": 0.0099, "step": 18790 }, { "grad_norm": 0.22189657390117645, "learning_rate": 5.904809217558542e-05, "loss": 0.0117, "step": 18800 }, { "grad_norm": 0.15318329632282257, "learning_rate": 5.90074348074425e-05, "loss": 0.0105, "step": 18810 }, { "grad_norm": 0.1767650693655014, "learning_rate": 5.8966771282802814e-05, "loss": 0.0139, "step": 18820 }, { "grad_norm": 0.25628626346588135, "learning_rate": 5.892610162945952e-05, "loss": 0.012, "step": 18830 }, { "grad_norm": 0.24415789544582367, "learning_rate": 5.8885425875209924e-05, "loss": 0.0122, "step": 18840 }, { "grad_norm": 0.20591045916080475, "learning_rate": 5.884474404785553e-05, "loss": 0.0113, "step": 18850 }, { "grad_norm": 0.24011316895484924, "learning_rate": 5.8804056175201983e-05, "loss": 0.0099, "step": 18860 }, { "grad_norm": 0.17545545101165771, "learning_rate": 5.876336228505904e-05, "loss": 0.0098, "step": 18870 }, { "grad_norm": 0.17654675245285034, "learning_rate": 5.872266240524062e-05, "loss": 0.0127, "step": 18880 }, { "grad_norm": 0.15076716244220734, "learning_rate": 5.86819565635647e-05, "loss": 0.0135, "step": 18890 }, { "grad_norm": 0.1767057329416275, "learning_rate": 5.8641244787853334e-05, "loss": 0.0105, "step": 18900 }, { "grad_norm": 0.15378950536251068, "learning_rate": 5.860052710593265e-05, "loss": 0.012, "step": 18910 }, { "grad_norm": 0.19073374569416046, "learning_rate": 5.855980354563276e-05, "loss": 0.0126, "step": 18920 }, { "grad_norm": 0.20253092050552368, "learning_rate": 5.8519074134787874e-05, "loss": 0.0126, "step": 18930 }, { "grad_norm": 0.19710157811641693, "learning_rate": 5.847833890123614e-05, "loss": 0.011, "step": 18940 }, { "grad_norm": 0.16406312584877014, "learning_rate": 5.8437597872819737e-05, "loss": 0.0124, "step": 18950 }, { "grad_norm": 0.11620765924453735, "learning_rate": 5.839685107738473e-05, "loss": 0.0125, "step": 18960 }, { "grad_norm": 0.1673503965139389, "learning_rate": 5.835609854278118e-05, "loss": 0.0114, "step": 18970 }, { "grad_norm": 0.2224600911140442, "learning_rate": 5.831534029686308e-05, "loss": 0.0121, "step": 18980 }, { "grad_norm": 0.16549180448055267, "learning_rate": 5.82745763674883e-05, "loss": 0.0125, "step": 18990 }, { "grad_norm": 0.24940556287765503, "learning_rate": 5.823380678251861e-05, "loss": 0.013, "step": 19000 }, { "grad_norm": 0.1904170960187912, "learning_rate": 5.81930315698196e-05, "loss": 0.0114, "step": 19010 }, { "grad_norm": 0.15991990268230438, "learning_rate": 5.815225075726076e-05, "loss": 0.0114, "step": 19020 }, { "grad_norm": 0.12930265069007874, "learning_rate": 5.811146437271543e-05, "loss": 0.0128, "step": 19030 }, { "grad_norm": 0.1730014979839325, "learning_rate": 5.807067244406066e-05, "loss": 0.0104, "step": 19040 }, { "grad_norm": 0.17663109302520752, "learning_rate": 5.8029874999177405e-05, "loss": 0.0103, "step": 19050 }, { "grad_norm": 0.1553577035665512, "learning_rate": 5.798907206595029e-05, "loss": 0.0102, "step": 19060 }, { "grad_norm": 0.15353430807590485, "learning_rate": 5.794826367226773e-05, "loss": 0.0133, "step": 19070 }, { "grad_norm": 0.21315713226795197, "learning_rate": 5.790744984602193e-05, "loss": 0.015, "step": 19080 }, { "grad_norm": 0.14688564836978912, "learning_rate": 5.786663061510872e-05, "loss": 0.0112, "step": 19090 }, { "grad_norm": 0.242301344871521, "learning_rate": 5.782580600742765e-05, "loss": 0.0113, "step": 19100 }, { "grad_norm": 0.17422018945217133, "learning_rate": 5.7784976050881965e-05, "loss": 0.0113, "step": 19110 }, { "grad_norm": 0.14065511524677277, "learning_rate": 5.774414077337855e-05, "loss": 0.0126, "step": 19120 }, { "grad_norm": 0.22403790056705475, "learning_rate": 5.770330020282796e-05, "loss": 0.01, "step": 19130 }, { "grad_norm": 0.19208909571170807, "learning_rate": 5.7662454367144317e-05, "loss": 0.0106, "step": 19140 }, { "grad_norm": 0.17297302186489105, "learning_rate": 5.762160329424536e-05, "loss": 0.0118, "step": 19150 }, { "grad_norm": 0.3436042070388794, "learning_rate": 5.7580747012052416e-05, "loss": 0.0096, "step": 19160 }, { "grad_norm": 0.2186223864555359, "learning_rate": 5.753988554849037e-05, "loss": 0.0117, "step": 19170 }, { "grad_norm": 0.2787349224090576, "learning_rate": 5.749901893148766e-05, "loss": 0.0136, "step": 19180 }, { "grad_norm": 0.14235682785511017, "learning_rate": 5.745814718897621e-05, "loss": 0.0127, "step": 19190 }, { "grad_norm": 0.17340363562107086, "learning_rate": 5.74172703488915e-05, "loss": 0.0106, "step": 19200 }, { "grad_norm": 0.12868739664554596, "learning_rate": 5.737638843917242e-05, "loss": 0.0098, "step": 19210 }, { "grad_norm": 0.1532471776008606, "learning_rate": 5.73355014877614e-05, "loss": 0.009, "step": 19220 }, { "grad_norm": 0.18966786563396454, "learning_rate": 5.7294609522604316e-05, "loss": 0.0112, "step": 19230 }, { "grad_norm": 0.22150318324565887, "learning_rate": 5.7253712571650376e-05, "loss": 0.0104, "step": 19240 }, { "grad_norm": 0.13151231408119202, "learning_rate": 5.721281066285229e-05, "loss": 0.0133, "step": 19250 }, { "grad_norm": 0.16553260385990143, "learning_rate": 5.717190382416615e-05, "loss": 0.0099, "step": 19260 }, { "grad_norm": 0.17138563096523285, "learning_rate": 5.713099208355135e-05, "loss": 0.0112, "step": 19270 }, { "grad_norm": 0.21826452016830444, "learning_rate": 5.709007546897074e-05, "loss": 0.0127, "step": 19280 }, { "grad_norm": 0.14899246394634247, "learning_rate": 5.704915400839037e-05, "loss": 0.0118, "step": 19290 }, { "grad_norm": 0.158478245139122, "learning_rate": 5.700822772977971e-05, "loss": 0.0094, "step": 19300 }, { "grad_norm": 0.1192425861954689, "learning_rate": 5.696729666111148e-05, "loss": 0.0091, "step": 19310 }, { "grad_norm": 0.18083564937114716, "learning_rate": 5.692636083036168e-05, "loss": 0.0112, "step": 19320 }, { "grad_norm": 0.1482033133506775, "learning_rate": 5.688542026550958e-05, "loss": 0.0127, "step": 19330 }, { "grad_norm": 0.16074611246585846, "learning_rate": 5.684447499453763e-05, "loss": 0.0121, "step": 19340 }, { "grad_norm": 0.14054329693317413, "learning_rate": 5.680352504543156e-05, "loss": 0.0099, "step": 19350 }, { "grad_norm": 0.2083030343055725, "learning_rate": 5.67625704461803e-05, "loss": 0.0105, "step": 19360 }, { "grad_norm": 0.1870032399892807, "learning_rate": 5.672161122477589e-05, "loss": 0.01, "step": 19370 }, { "grad_norm": 0.1699010133743286, "learning_rate": 5.668064740921359e-05, "loss": 0.0108, "step": 19380 }, { "grad_norm": 0.20328998565673828, "learning_rate": 5.663967902749179e-05, "loss": 0.0126, "step": 19390 }, { "grad_norm": 0.2108919471502304, "learning_rate": 5.6598706107611965e-05, "loss": 0.0111, "step": 19400 }, { "grad_norm": 0.20684179663658142, "learning_rate": 5.655772867757876e-05, "loss": 0.0128, "step": 19410 }, { "grad_norm": 0.1321077197790146, "learning_rate": 5.651674676539982e-05, "loss": 0.0098, "step": 19420 }, { "grad_norm": 0.16804124414920807, "learning_rate": 5.647576039908593e-05, "loss": 0.0099, "step": 19430 }, { "grad_norm": 0.13769981265068054, "learning_rate": 5.6434769606650864e-05, "loss": 0.0104, "step": 19440 }, { "grad_norm": 0.22308778762817383, "learning_rate": 5.639377441611143e-05, "loss": 0.0116, "step": 19450 }, { "grad_norm": 0.2626425623893738, "learning_rate": 5.635277485548751e-05, "loss": 0.0099, "step": 19460 }, { "grad_norm": 0.20041291415691376, "learning_rate": 5.631177095280186e-05, "loss": 0.0127, "step": 19470 }, { "grad_norm": 0.18701273202896118, "learning_rate": 5.627076273608027e-05, "loss": 0.0128, "step": 19480 }, { "grad_norm": 0.17116715013980865, "learning_rate": 5.622975023335148e-05, "loss": 0.01, "step": 19490 }, { "grad_norm": 0.17630213499069214, "learning_rate": 5.618873347264716e-05, "loss": 0.0093, "step": 19500 }, { "grad_norm": 0.1976991593837738, "learning_rate": 5.614771248200188e-05, "loss": 0.0133, "step": 19510 }, { "grad_norm": 0.21130220592021942, "learning_rate": 5.6106687289453066e-05, "loss": 0.01, "step": 19520 }, { "grad_norm": 0.18598775565624237, "learning_rate": 5.606565792304108e-05, "loss": 0.0113, "step": 19530 }, { "grad_norm": 0.15803469717502594, "learning_rate": 5.602462441080909e-05, "loss": 0.0105, "step": 19540 }, { "grad_norm": 0.26154857873916626, "learning_rate": 5.5983586780803135e-05, "loss": 0.0134, "step": 19550 }, { "grad_norm": 0.20463597774505615, "learning_rate": 5.594254506107205e-05, "loss": 0.0114, "step": 19560 }, { "grad_norm": 0.17139588296413422, "learning_rate": 5.590149927966743e-05, "loss": 0.0124, "step": 19570 }, { "grad_norm": 0.16595961153507233, "learning_rate": 5.58604494646437e-05, "loss": 0.0083, "step": 19580 }, { "grad_norm": 0.17736771702766418, "learning_rate": 5.5819395644058025e-05, "loss": 0.012, "step": 19590 }, { "grad_norm": 0.15832625329494476, "learning_rate": 5.577833784597031e-05, "loss": 0.0128, "step": 19600 }, { "grad_norm": 0.13744552433490753, "learning_rate": 5.573727609844316e-05, "loss": 0.0098, "step": 19610 }, { "grad_norm": 0.1575765609741211, "learning_rate": 5.5696210429541884e-05, "loss": 0.0111, "step": 19620 }, { "grad_norm": 0.17395706474781036, "learning_rate": 5.565514086733451e-05, "loss": 0.0095, "step": 19630 }, { "grad_norm": 0.18265791237354279, "learning_rate": 5.5614067439891657e-05, "loss": 0.0126, "step": 19640 }, { "grad_norm": 0.14385151863098145, "learning_rate": 5.557299017528666e-05, "loss": 0.0098, "step": 19650 }, { "grad_norm": 0.23068904876708984, "learning_rate": 5.5531909101595436e-05, "loss": 0.0101, "step": 19660 }, { "grad_norm": 0.1650722324848175, "learning_rate": 5.549082424689649e-05, "loss": 0.0105, "step": 19670 }, { "grad_norm": 0.18441665172576904, "learning_rate": 5.544973563927095e-05, "loss": 0.0119, "step": 19680 }, { "grad_norm": 0.22624541819095612, "learning_rate": 5.540864330680249e-05, "loss": 0.0105, "step": 19690 }, { "grad_norm": 0.15144185721874237, "learning_rate": 5.536754727757733e-05, "loss": 0.0084, "step": 19700 }, { "grad_norm": 0.16923312842845917, "learning_rate": 5.532644757968422e-05, "loss": 0.012, "step": 19710 }, { "grad_norm": 0.18090881407260895, "learning_rate": 5.528534424121441e-05, "loss": 0.0101, "step": 19720 }, { "grad_norm": 0.21753372251987457, "learning_rate": 5.524423729026165e-05, "loss": 0.0145, "step": 19730 }, { "grad_norm": 0.22991928458213806, "learning_rate": 5.5203126754922164e-05, "loss": 0.0108, "step": 19740 }, { "grad_norm": 0.18655601143836975, "learning_rate": 5.5162012663294585e-05, "loss": 0.0102, "step": 19750 }, { "grad_norm": 0.2103322297334671, "learning_rate": 5.512089504348003e-05, "loss": 0.0097, "step": 19760 }, { "grad_norm": 0.23478513956069946, "learning_rate": 5.5079773923582e-05, "loss": 0.0137, "step": 19770 }, { "grad_norm": 0.1896728128194809, "learning_rate": 5.50386493317064e-05, "loss": 0.0101, "step": 19780 }, { "grad_norm": 0.16997915506362915, "learning_rate": 5.49975212959615e-05, "loss": 0.0112, "step": 19790 }, { "grad_norm": 0.15039899945259094, "learning_rate": 5.4956389844457904e-05, "loss": 0.0089, "step": 19800 }, { "grad_norm": 0.19073167443275452, "learning_rate": 5.491525500530859e-05, "loss": 0.0112, "step": 19810 }, { "grad_norm": 0.15496446192264557, "learning_rate": 5.487411680662882e-05, "loss": 0.0116, "step": 19820 }, { "grad_norm": 0.30736249685287476, "learning_rate": 5.483297527653618e-05, "loss": 0.0129, "step": 19830 }, { "grad_norm": 0.20753253996372223, "learning_rate": 5.4791830443150516e-05, "loss": 0.0122, "step": 19840 }, { "grad_norm": 0.1487683206796646, "learning_rate": 5.475068233459392e-05, "loss": 0.0104, "step": 19850 }, { "grad_norm": 0.21190504729747772, "learning_rate": 5.470953097899075e-05, "loss": 0.0138, "step": 19860 }, { "grad_norm": 0.18093030154705048, "learning_rate": 5.466837640446756e-05, "loss": 0.0129, "step": 19870 }, { "grad_norm": 0.1480749547481537, "learning_rate": 5.462721863915312e-05, "loss": 0.0131, "step": 19880 }, { "grad_norm": 0.15297353267669678, "learning_rate": 5.4586057711178374e-05, "loss": 0.0122, "step": 19890 }, { "grad_norm": 0.18061943352222443, "learning_rate": 5.454489364867642e-05, "loss": 0.0124, "step": 19900 }, { "grad_norm": 0.16061453521251678, "learning_rate": 5.4503726479782523e-05, "loss": 0.0139, "step": 19910 }, { "grad_norm": 0.16561543941497803, "learning_rate": 5.446255623263403e-05, "loss": 0.01, "step": 19920 }, { "grad_norm": 0.20964108407497406, "learning_rate": 5.4421382935370445e-05, "loss": 0.0109, "step": 19930 }, { "grad_norm": 0.14122013747692108, "learning_rate": 5.438020661613331e-05, "loss": 0.011, "step": 19940 }, { "grad_norm": 0.146755650639534, "learning_rate": 5.433902730306625e-05, "loss": 0.0098, "step": 19950 }, { "grad_norm": 0.22255884110927582, "learning_rate": 5.429784502431495e-05, "loss": 0.0124, "step": 19960 }, { "grad_norm": 0.23333603143692017, "learning_rate": 5.42566598080271e-05, "loss": 0.0139, "step": 19970 }, { "grad_norm": 0.17668990790843964, "learning_rate": 5.421547168235241e-05, "loss": 0.0102, "step": 19980 }, { "grad_norm": 0.15050621330738068, "learning_rate": 5.417428067544258e-05, "loss": 0.0102, "step": 19990 }, { "grad_norm": 0.20039522647857666, "learning_rate": 5.413308681545126e-05, "loss": 0.0129, "step": 20000 }, { "grad_norm": 0.20523694157600403, "learning_rate": 5.409189013053408e-05, "loss": 0.0143, "step": 20010 }, { "grad_norm": 0.1750054508447647, "learning_rate": 5.4050690648848576e-05, "loss": 0.0111, "step": 20020 }, { "grad_norm": 0.1853441447019577, "learning_rate": 5.400948839855421e-05, "loss": 0.0109, "step": 20030 }, { "grad_norm": 0.1614084243774414, "learning_rate": 5.396828340781234e-05, "loss": 0.01, "step": 20040 }, { "grad_norm": 0.14277350902557373, "learning_rate": 5.392707570478617e-05, "loss": 0.0098, "step": 20050 }, { "grad_norm": 0.1237936019897461, "learning_rate": 5.388586531764078e-05, "loss": 0.011, "step": 20060 }, { "grad_norm": 0.18935835361480713, "learning_rate": 5.384465227454311e-05, "loss": 0.0121, "step": 20070 }, { "grad_norm": 0.18697305023670197, "learning_rate": 5.380343660366184e-05, "loss": 0.0102, "step": 20080 }, { "grad_norm": 0.19770930707454681, "learning_rate": 5.376221833316752e-05, "loss": 0.0121, "step": 20090 }, { "grad_norm": 0.2053295075893402, "learning_rate": 5.3720997491232436e-05, "loss": 0.0157, "step": 20100 }, { "grad_norm": 0.17481350898742676, "learning_rate": 5.367977410603068e-05, "loss": 0.0131, "step": 20110 }, { "grad_norm": 0.19119560718536377, "learning_rate": 5.3638548205738004e-05, "loss": 0.014, "step": 20120 }, { "grad_norm": 0.22180065512657166, "learning_rate": 5.359731981853194e-05, "loss": 0.0126, "step": 20130 }, { "grad_norm": 0.1893962323665619, "learning_rate": 5.35560889725917e-05, "loss": 0.0118, "step": 20140 }, { "grad_norm": 0.24221593141555786, "learning_rate": 5.3514855696098176e-05, "loss": 0.0143, "step": 20150 }, { "grad_norm": 0.18735502660274506, "learning_rate": 5.347362001723394e-05, "loss": 0.0102, "step": 20160 }, { "grad_norm": 0.22477152943611145, "learning_rate": 5.3432381964183176e-05, "loss": 0.0095, "step": 20170 }, { "grad_norm": 0.1299329698085785, "learning_rate": 5.3391141565131685e-05, "loss": 0.01, "step": 20180 }, { "grad_norm": 0.13806502521038055, "learning_rate": 5.3349898848266935e-05, "loss": 0.0107, "step": 20190 }, { "grad_norm": 0.16806964576244354, "learning_rate": 5.330865384177789e-05, "loss": 0.0092, "step": 20200 }, { "grad_norm": 0.16564123332500458, "learning_rate": 5.326740657385515e-05, "loss": 0.0108, "step": 20210 }, { "grad_norm": 0.23138833045959473, "learning_rate": 5.322615707269083e-05, "loss": 0.0119, "step": 20220 }, { "grad_norm": 0.20094098150730133, "learning_rate": 5.318490536647856e-05, "loss": 0.0088, "step": 20230 }, { "grad_norm": 0.2327081263065338, "learning_rate": 5.3143651483413524e-05, "loss": 0.0093, "step": 20240 }, { "grad_norm": 0.21148844063282013, "learning_rate": 5.310239545169232e-05, "loss": 0.0118, "step": 20250 }, { "grad_norm": 0.237352654337883, "learning_rate": 5.30611372995131e-05, "loss": 0.0127, "step": 20260 }, { "grad_norm": 0.2625342309474945, "learning_rate": 5.30198770550754e-05, "loss": 0.01, "step": 20270 }, { "grad_norm": 0.23618794977664948, "learning_rate": 5.297861474658019e-05, "loss": 0.0108, "step": 20280 }, { "grad_norm": 0.17728564143180847, "learning_rate": 5.29373504022299e-05, "loss": 0.0106, "step": 20290 }, { "grad_norm": 0.17865923047065735, "learning_rate": 5.28960840502283e-05, "loss": 0.0097, "step": 20300 }, { "grad_norm": 0.14775164425373077, "learning_rate": 5.285481571878056e-05, "loss": 0.0144, "step": 20310 }, { "grad_norm": 0.2389378696680069, "learning_rate": 5.281354543609321e-05, "loss": 0.0106, "step": 20320 }, { "grad_norm": 0.2123534381389618, "learning_rate": 5.277227323037406e-05, "loss": 0.0111, "step": 20330 }, { "grad_norm": 0.1665085107088089, "learning_rate": 5.273099912983233e-05, "loss": 0.0113, "step": 20340 }, { "grad_norm": 0.18871785700321198, "learning_rate": 5.268972316267843e-05, "loss": 0.0128, "step": 20350 }, { "grad_norm": 0.16967450082302094, "learning_rate": 5.26484453571241e-05, "loss": 0.0107, "step": 20360 }, { "grad_norm": 0.2591538727283478, "learning_rate": 5.260716574138235e-05, "loss": 0.0109, "step": 20370 }, { "grad_norm": 0.2094735950231552, "learning_rate": 5.256588434366739e-05, "loss": 0.0119, "step": 20380 }, { "grad_norm": 0.15748779475688934, "learning_rate": 5.25246011921947e-05, "loss": 0.0127, "step": 20390 }, { "grad_norm": 0.13478122651576996, "learning_rate": 5.248331631518089e-05, "loss": 0.0118, "step": 20400 }, { "grad_norm": 0.1454327255487442, "learning_rate": 5.244202974084379e-05, "loss": 0.0108, "step": 20410 }, { "grad_norm": 0.17928260564804077, "learning_rate": 5.240074149740239e-05, "loss": 0.0076, "step": 20420 }, { "grad_norm": 0.1207803264260292, "learning_rate": 5.2359451613076814e-05, "loss": 0.0082, "step": 20430 }, { "grad_norm": 0.3778174817562103, "learning_rate": 5.231816011608832e-05, "loss": 0.0122, "step": 20440 }, { "grad_norm": 0.2374403327703476, "learning_rate": 5.227686703465924e-05, "loss": 0.0102, "step": 20450 }, { "grad_norm": 0.16592712700366974, "learning_rate": 5.2235572397013e-05, "loss": 0.0112, "step": 20460 }, { "grad_norm": 0.195661723613739, "learning_rate": 5.2194276231374114e-05, "loss": 0.0106, "step": 20470 }, { "grad_norm": 0.1774659901857376, "learning_rate": 5.21529785659681e-05, "loss": 0.0119, "step": 20480 }, { "grad_norm": 0.2414444535970688, "learning_rate": 5.2111679429021565e-05, "loss": 0.0137, "step": 20490 }, { "grad_norm": 0.22908265888690948, "learning_rate": 5.207037884876205e-05, "loss": 0.0118, "step": 20500 }, { "grad_norm": 0.2839491665363312, "learning_rate": 5.202907685341809e-05, "loss": 0.0101, "step": 20510 }, { "grad_norm": 0.16354557871818542, "learning_rate": 5.198777347121926e-05, "loss": 0.0147, "step": 20520 }, { "grad_norm": 0.14166414737701416, "learning_rate": 5.194646873039598e-05, "loss": 0.0126, "step": 20530 }, { "grad_norm": 0.13617683947086334, "learning_rate": 5.1905162659179696e-05, "loss": 0.0087, "step": 20540 }, { "grad_norm": 0.2175985872745514, "learning_rate": 5.18638552858027e-05, "loss": 0.012, "step": 20550 }, { "grad_norm": 0.14866045117378235, "learning_rate": 5.182254663849818e-05, "loss": 0.0128, "step": 20560 }, { "grad_norm": 0.1468195915222168, "learning_rate": 5.178123674550023e-05, "loss": 0.0106, "step": 20570 }, { "grad_norm": 0.14988432824611664, "learning_rate": 5.173992563504375e-05, "loss": 0.0111, "step": 20580 }, { "grad_norm": 0.14899785816669464, "learning_rate": 5.169861333536451e-05, "loss": 0.0101, "step": 20590 }, { "grad_norm": 0.1146642342209816, "learning_rate": 5.165729987469907e-05, "loss": 0.0091, "step": 20600 }, { "grad_norm": 0.1257292628288269, "learning_rate": 5.161598528128478e-05, "loss": 0.0109, "step": 20610 }, { "grad_norm": 0.1888774037361145, "learning_rate": 5.157466958335981e-05, "loss": 0.0083, "step": 20620 }, { "grad_norm": 0.18485897779464722, "learning_rate": 5.1533352809163025e-05, "loss": 0.012, "step": 20630 }, { "grad_norm": 0.19455192983150482, "learning_rate": 5.1492034986934046e-05, "loss": 0.0115, "step": 20640 }, { "grad_norm": 0.1765831708908081, "learning_rate": 5.1450716144913225e-05, "loss": 0.0129, "step": 20650 }, { "grad_norm": 0.18618805706501007, "learning_rate": 5.1409396311341595e-05, "loss": 0.0126, "step": 20660 }, { "grad_norm": 0.16662266850471497, "learning_rate": 5.136807551446089e-05, "loss": 0.0119, "step": 20670 }, { "grad_norm": 0.1739405393600464, "learning_rate": 5.132675378251346e-05, "loss": 0.0106, "step": 20680 }, { "grad_norm": 0.1702536791563034, "learning_rate": 5.1285431143742325e-05, "loss": 0.0107, "step": 20690 }, { "grad_norm": 0.1559237539768219, "learning_rate": 5.1244107626391136e-05, "loss": 0.0114, "step": 20700 }, { "grad_norm": 0.23261064291000366, "learning_rate": 5.12027832587041e-05, "loss": 0.0108, "step": 20710 }, { "grad_norm": 0.17230984568595886, "learning_rate": 5.116145806892607e-05, "loss": 0.0102, "step": 20720 }, { "grad_norm": 0.1809496134519577, "learning_rate": 5.1120132085302384e-05, "loss": 0.0125, "step": 20730 }, { "grad_norm": 0.2441035360097885, "learning_rate": 5.107880533607898e-05, "loss": 0.0132, "step": 20740 }, { "grad_norm": 0.17388148605823517, "learning_rate": 5.103747784950231e-05, "loss": 0.0106, "step": 20750 }, { "grad_norm": 0.24142691493034363, "learning_rate": 5.09961496538193e-05, "loss": 0.0119, "step": 20760 }, { "grad_norm": 0.2722826302051544, "learning_rate": 5.095482077727742e-05, "loss": 0.0141, "step": 20770 }, { "grad_norm": 0.13314403593540192, "learning_rate": 5.091349124812452e-05, "loss": 0.0088, "step": 20780 }, { "grad_norm": 0.17796207964420319, "learning_rate": 5.087216109460897e-05, "loss": 0.0101, "step": 20790 }, { "grad_norm": 0.19711577892303467, "learning_rate": 5.083083034497954e-05, "loss": 0.0102, "step": 20800 }, { "grad_norm": 0.14978978037834167, "learning_rate": 5.07894990274854e-05, "loss": 0.0082, "step": 20810 }, { "grad_norm": 0.21021412312984467, "learning_rate": 5.074816717037614e-05, "loss": 0.0105, "step": 20820 }, { "grad_norm": 0.21884457767009735, "learning_rate": 5.070683480190165e-05, "loss": 0.0112, "step": 20830 }, { "grad_norm": 0.16607099771499634, "learning_rate": 5.066550195031223e-05, "loss": 0.0096, "step": 20840 }, { "grad_norm": 0.20561252534389496, "learning_rate": 5.062416864385852e-05, "loss": 0.0099, "step": 20850 }, { "grad_norm": 0.1832585632801056, "learning_rate": 5.058283491079142e-05, "loss": 0.0108, "step": 20860 }, { "grad_norm": 0.16747991740703583, "learning_rate": 5.054150077936216e-05, "loss": 0.0089, "step": 20870 }, { "grad_norm": 0.20517121255397797, "learning_rate": 5.0500166277822214e-05, "loss": 0.01, "step": 20880 }, { "grad_norm": 0.10699398815631866, "learning_rate": 5.0458831434423334e-05, "loss": 0.0087, "step": 20890 }, { "grad_norm": 0.17235346138477325, "learning_rate": 5.0417496277417506e-05, "loss": 0.0082, "step": 20900 }, { "grad_norm": 0.18732629716396332, "learning_rate": 5.037616083505691e-05, "loss": 0.0114, "step": 20910 }, { "grad_norm": 0.16035906970500946, "learning_rate": 5.0334825135593935e-05, "loss": 0.0101, "step": 20920 }, { "grad_norm": 0.1590651422739029, "learning_rate": 5.029348920728111e-05, "loss": 0.0102, "step": 20930 }, { "grad_norm": 0.20482806861400604, "learning_rate": 5.0252153078371186e-05, "loss": 0.0132, "step": 20940 }, { "grad_norm": 0.1797042042016983, "learning_rate": 5.021081677711704e-05, "loss": 0.0095, "step": 20950 }, { "grad_norm": 0.1742580384016037, "learning_rate": 5.016948033177159e-05, "loss": 0.0104, "step": 20960 }, { "grad_norm": 0.1494574099779129, "learning_rate": 5.012814377058793e-05, "loss": 0.0103, "step": 20970 }, { "grad_norm": 0.15364673733711243, "learning_rate": 5.008680712181921e-05, "loss": 0.0112, "step": 20980 }, { "grad_norm": 0.2193729728460312, "learning_rate": 5.0045470413718645e-05, "loss": 0.0093, "step": 20990 }, { "grad_norm": 0.1419246941804886, "learning_rate": 5.00041336745395e-05, "loss": 0.0133, "step": 21000 }, { "grad_norm": 0.16727520525455475, "learning_rate": 4.996279693253499e-05, "loss": 0.0104, "step": 21010 }, { "grad_norm": 0.23682697117328644, "learning_rate": 4.992146021595847e-05, "loss": 0.012, "step": 21020 }, { "grad_norm": 0.13317497074604034, "learning_rate": 4.988012355306313e-05, "loss": 0.0088, "step": 21030 }, { "grad_norm": 0.16880793869495392, "learning_rate": 4.98387869721022e-05, "loss": 0.0109, "step": 21040 }, { "grad_norm": 0.22881287336349487, "learning_rate": 4.9797450501328866e-05, "loss": 0.0131, "step": 21050 }, { "grad_norm": 0.2007744014263153, "learning_rate": 4.97561141689962e-05, "loss": 0.0093, "step": 21060 }, { "grad_norm": 0.1454089879989624, "learning_rate": 4.971477800335721e-05, "loss": 0.0126, "step": 21070 }, { "grad_norm": 0.1622297167778015, "learning_rate": 4.967344203266475e-05, "loss": 0.0105, "step": 21080 }, { "grad_norm": 0.17552435398101807, "learning_rate": 4.9632106285171584e-05, "loss": 0.0105, "step": 21090 }, { "grad_norm": 0.17016011476516724, "learning_rate": 4.959077078913031e-05, "loss": 0.0104, "step": 21100 }, { "grad_norm": 0.20775283873081207, "learning_rate": 4.954943557279333e-05, "loss": 0.0121, "step": 21110 }, { "grad_norm": 0.15504397451877594, "learning_rate": 4.9508100664412916e-05, "loss": 0.0095, "step": 21120 }, { "grad_norm": 0.1806158572435379, "learning_rate": 4.946676609224105e-05, "loss": 0.011, "step": 21130 }, { "grad_norm": 0.21461908519268036, "learning_rate": 4.942543188452952e-05, "loss": 0.0114, "step": 21140 }, { "grad_norm": 0.1682797223329544, "learning_rate": 4.938409806952988e-05, "loss": 0.0121, "step": 21150 }, { "grad_norm": 0.16431841254234314, "learning_rate": 4.93427646754934e-05, "loss": 0.0103, "step": 21160 }, { "grad_norm": 0.16803325712680817, "learning_rate": 4.930143173067108e-05, "loss": 0.0113, "step": 21170 }, { "grad_norm": 0.14259310066699982, "learning_rate": 4.9260099263313565e-05, "loss": 0.0112, "step": 21180 }, { "grad_norm": 0.19304826855659485, "learning_rate": 4.921876730167123e-05, "loss": 0.0131, "step": 21190 }, { "grad_norm": 0.17570267617702484, "learning_rate": 4.917743587399409e-05, "loss": 0.012, "step": 21200 }, { "grad_norm": 0.23490867018699646, "learning_rate": 4.913610500853178e-05, "loss": 0.0098, "step": 21210 }, { "grad_norm": 0.23300299048423767, "learning_rate": 4.909477473353354e-05, "loss": 0.0117, "step": 21220 }, { "grad_norm": 0.1581895351409912, "learning_rate": 4.9053445077248236e-05, "loss": 0.0112, "step": 21230 }, { "grad_norm": 0.1785595417022705, "learning_rate": 4.901211606792429e-05, "loss": 0.0099, "step": 21240 }, { "grad_norm": 0.17192251980304718, "learning_rate": 4.89707877338097e-05, "loss": 0.0091, "step": 21250 }, { "grad_norm": 0.15306618809700012, "learning_rate": 4.892946010315199e-05, "loss": 0.0095, "step": 21260 }, { "grad_norm": 0.13022740185260773, "learning_rate": 4.8888133204198204e-05, "loss": 0.012, "step": 21270 }, { "grad_norm": 0.19151437282562256, "learning_rate": 4.8846807065194886e-05, "loss": 0.0102, "step": 21280 }, { "grad_norm": 0.16872170567512512, "learning_rate": 4.880548171438806e-05, "loss": 0.0121, "step": 21290 }, { "grad_norm": 0.20959411561489105, "learning_rate": 4.8764157180023245e-05, "loss": 0.0101, "step": 21300 }, { "grad_norm": 0.2436714917421341, "learning_rate": 4.872283349034533e-05, "loss": 0.0126, "step": 21310 }, { "grad_norm": 0.13195396959781647, "learning_rate": 4.8681510673598674e-05, "loss": 0.0106, "step": 21320 }, { "grad_norm": 0.129905104637146, "learning_rate": 4.8640188758027046e-05, "loss": 0.0069, "step": 21330 }, { "grad_norm": 0.1285659819841385, "learning_rate": 4.859886777187357e-05, "loss": 0.0114, "step": 21340 }, { "grad_norm": 0.19101060926914215, "learning_rate": 4.855754774338077e-05, "loss": 0.0099, "step": 21350 }, { "grad_norm": 0.19425636529922485, "learning_rate": 4.851622870079048e-05, "loss": 0.0096, "step": 21360 }, { "grad_norm": 0.16221320629119873, "learning_rate": 4.847491067234389e-05, "loss": 0.0087, "step": 21370 }, { "grad_norm": 0.14020879566669464, "learning_rate": 4.843359368628146e-05, "loss": 0.0081, "step": 21380 }, { "grad_norm": 0.1648164987564087, "learning_rate": 4.8392277770842975e-05, "loss": 0.0106, "step": 21390 }, { "grad_norm": 0.2297167032957077, "learning_rate": 4.83509629542675e-05, "loss": 0.0102, "step": 21400 }, { "grad_norm": 0.14849717915058136, "learning_rate": 4.830964926479329e-05, "loss": 0.0104, "step": 21410 }, { "grad_norm": 0.14987900853157043, "learning_rate": 4.826833673065785e-05, "loss": 0.0105, "step": 21420 }, { "grad_norm": 0.1834854632616043, "learning_rate": 4.822702538009794e-05, "loss": 0.0104, "step": 21430 }, { "grad_norm": 0.1803760975599289, "learning_rate": 4.818571524134945e-05, "loss": 0.0111, "step": 21440 }, { "grad_norm": 0.16374047100543976, "learning_rate": 4.8144406342647496e-05, "loss": 0.0141, "step": 21450 }, { "grad_norm": 0.14251740276813507, "learning_rate": 4.81030987122263e-05, "loss": 0.0087, "step": 21460 }, { "grad_norm": 0.15806162357330322, "learning_rate": 4.806179237831926e-05, "loss": 0.0077, "step": 21470 }, { "grad_norm": 0.14271025359630585, "learning_rate": 4.802048736915884e-05, "loss": 0.0092, "step": 21480 }, { "grad_norm": 0.14372320473194122, "learning_rate": 4.797918371297666e-05, "loss": 0.0095, "step": 21490 }, { "grad_norm": 0.11639206856489182, "learning_rate": 4.793788143800334e-05, "loss": 0.0116, "step": 21500 }, { "grad_norm": 0.21395504474639893, "learning_rate": 4.789658057246862e-05, "loss": 0.0091, "step": 21510 }, { "grad_norm": 0.15659451484680176, "learning_rate": 4.7855281144601227e-05, "loss": 0.0107, "step": 21520 }, { "grad_norm": 0.14283038675785065, "learning_rate": 4.781398318262897e-05, "loss": 0.01, "step": 21530 }, { "grad_norm": 0.12540075182914734, "learning_rate": 4.777268671477858e-05, "loss": 0.0115, "step": 21540 }, { "grad_norm": 0.16363205015659332, "learning_rate": 4.773139176927582e-05, "loss": 0.0095, "step": 21550 }, { "grad_norm": 0.20567408204078674, "learning_rate": 4.769009837434539e-05, "loss": 0.0098, "step": 21560 }, { "grad_norm": 0.18492895364761353, "learning_rate": 4.764880655821095e-05, "loss": 0.0089, "step": 21570 }, { "grad_norm": 0.12200620770454407, "learning_rate": 4.760751634909508e-05, "loss": 0.0119, "step": 21580 }, { "grad_norm": 0.18552199006080627, "learning_rate": 4.756622777521919e-05, "loss": 0.011, "step": 21590 }, { "grad_norm": 0.20269040763378143, "learning_rate": 4.752494086480368e-05, "loss": 0.0131, "step": 21600 }, { "grad_norm": 0.22734388709068298, "learning_rate": 4.7483655646067744e-05, "loss": 0.0102, "step": 21610 }, { "grad_norm": 0.1622183918952942, "learning_rate": 4.744237214722944e-05, "loss": 0.0109, "step": 21620 }, { "grad_norm": 0.18221184611320496, "learning_rate": 4.740109039650567e-05, "loss": 0.01, "step": 21630 }, { "grad_norm": 0.18223989009857178, "learning_rate": 4.73598104221121e-05, "loss": 0.0105, "step": 21640 }, { "grad_norm": 0.11712048202753067, "learning_rate": 4.731853225226322e-05, "loss": 0.009, "step": 21650 }, { "grad_norm": 0.18499518930912018, "learning_rate": 4.727725591517225e-05, "loss": 0.0104, "step": 21660 }, { "grad_norm": 0.16023044288158417, "learning_rate": 4.723598143905119e-05, "loss": 0.0107, "step": 21670 }, { "grad_norm": 0.15208090841770172, "learning_rate": 4.719470885211077e-05, "loss": 0.0102, "step": 21680 }, { "grad_norm": 0.19876700639724731, "learning_rate": 4.7153438182560387e-05, "loss": 0.0108, "step": 21690 }, { "grad_norm": 0.22922465205192566, "learning_rate": 4.711216945860815e-05, "loss": 0.0107, "step": 21700 }, { "grad_norm": 0.15035590529441833, "learning_rate": 4.707090270846088e-05, "loss": 0.01, "step": 21710 }, { "grad_norm": 0.23081527650356293, "learning_rate": 4.702963796032397e-05, "loss": 0.0149, "step": 21720 }, { "grad_norm": 0.14511607587337494, "learning_rate": 4.6988375242401514e-05, "loss": 0.0123, "step": 21730 }, { "grad_norm": 0.14302873611450195, "learning_rate": 4.694711458289618e-05, "loss": 0.0089, "step": 21740 }, { "grad_norm": 0.1312163919210434, "learning_rate": 4.690585601000925e-05, "loss": 0.0096, "step": 21750 }, { "grad_norm": 0.14784011244773865, "learning_rate": 4.686459955194055e-05, "loss": 0.0102, "step": 21760 }, { "grad_norm": 0.1467597484588623, "learning_rate": 4.6823345236888504e-05, "loss": 0.0122, "step": 21770 }, { "grad_norm": 0.22661152482032776, "learning_rate": 4.678209309305002e-05, "loss": 0.0096, "step": 21780 }, { "grad_norm": 0.18839740753173828, "learning_rate": 4.674084314862057e-05, "loss": 0.0095, "step": 21790 }, { "grad_norm": 0.16920584440231323, "learning_rate": 4.669959543179409e-05, "loss": 0.0087, "step": 21800 }, { "grad_norm": 0.13554799556732178, "learning_rate": 4.665834997076303e-05, "loss": 0.0101, "step": 21810 }, { "grad_norm": 0.13395676016807556, "learning_rate": 4.661710679371823e-05, "loss": 0.0082, "step": 21820 }, { "grad_norm": 0.1595124751329422, "learning_rate": 4.657586592884905e-05, "loss": 0.01, "step": 21830 }, { "grad_norm": 0.17244470119476318, "learning_rate": 4.653462740434322e-05, "loss": 0.0091, "step": 21840 }, { "grad_norm": 0.19302242994308472, "learning_rate": 4.649339124838689e-05, "loss": 0.0116, "step": 21850 }, { "grad_norm": 0.14715814590454102, "learning_rate": 4.6452157489164574e-05, "loss": 0.0096, "step": 21860 }, { "grad_norm": 0.17265015840530396, "learning_rate": 4.6410926154859155e-05, "loss": 0.0106, "step": 21870 }, { "grad_norm": 0.13441070914268494, "learning_rate": 4.636969727365186e-05, "loss": 0.0086, "step": 21880 }, { "grad_norm": 0.14167942106723785, "learning_rate": 4.632847087372226e-05, "loss": 0.01, "step": 21890 }, { "grad_norm": 0.1761697232723236, "learning_rate": 4.628724698324818e-05, "loss": 0.0103, "step": 21900 }, { "grad_norm": 0.18022553622722626, "learning_rate": 4.6246025630405795e-05, "loss": 0.0096, "step": 21910 }, { "grad_norm": 0.16583110392093658, "learning_rate": 4.6204806843369474e-05, "loss": 0.0094, "step": 21920 }, { "grad_norm": 0.12877622246742249, "learning_rate": 4.616359065031191e-05, "loss": 0.0097, "step": 21930 }, { "grad_norm": 0.11992545425891876, "learning_rate": 4.6122377079403946e-05, "loss": 0.0103, "step": 21940 }, { "grad_norm": 0.16360853612422943, "learning_rate": 4.6081166158814695e-05, "loss": 0.0075, "step": 21950 }, { "grad_norm": 0.1503250151872635, "learning_rate": 4.603995791671144e-05, "loss": 0.0093, "step": 21960 }, { "grad_norm": 0.13304147124290466, "learning_rate": 4.599875238125957e-05, "loss": 0.0111, "step": 21970 }, { "grad_norm": 0.16184620559215546, "learning_rate": 4.595754958062273e-05, "loss": 0.0112, "step": 21980 }, { "grad_norm": 0.23343272507190704, "learning_rate": 4.591634954296265e-05, "loss": 0.0101, "step": 21990 }, { "grad_norm": 0.2071828991174698, "learning_rate": 4.587515229643913e-05, "loss": 0.0082, "step": 22000 }, { "grad_norm": 0.1145549789071083, "learning_rate": 4.583395786921013e-05, "loss": 0.0091, "step": 22010 }, { "grad_norm": 0.1452537477016449, "learning_rate": 4.579276628943164e-05, "loss": 0.0095, "step": 22020 }, { "grad_norm": 0.13260865211486816, "learning_rate": 4.575157758525772e-05, "loss": 0.0086, "step": 22030 }, { "grad_norm": 0.21231764554977417, "learning_rate": 4.571039178484046e-05, "loss": 0.0102, "step": 22040 }, { "grad_norm": 0.1722312569618225, "learning_rate": 4.566920891632998e-05, "loss": 0.0092, "step": 22050 }, { "grad_norm": 0.16838915646076202, "learning_rate": 4.562802900787436e-05, "loss": 0.0082, "step": 22060 }, { "grad_norm": 0.20401868224143982, "learning_rate": 4.558685208761968e-05, "loss": 0.0095, "step": 22070 }, { "grad_norm": 0.12529121339321136, "learning_rate": 4.554567818370998e-05, "loss": 0.008, "step": 22080 }, { "grad_norm": 0.21601779758930206, "learning_rate": 4.550450732428726e-05, "loss": 0.0094, "step": 22090 }, { "grad_norm": 0.1525724232196808, "learning_rate": 4.546333953749137e-05, "loss": 0.012, "step": 22100 }, { "grad_norm": 0.18266528844833374, "learning_rate": 4.5422174851460154e-05, "loss": 0.0083, "step": 22110 }, { "grad_norm": 0.12643033266067505, "learning_rate": 4.538101329432924e-05, "loss": 0.0091, "step": 22120 }, { "grad_norm": 0.19065485894680023, "learning_rate": 4.5339854894232195e-05, "loss": 0.0128, "step": 22130 }, { "grad_norm": 0.11954587697982788, "learning_rate": 4.52986996793004e-05, "loss": 0.0096, "step": 22140 }, { "grad_norm": 0.163618266582489, "learning_rate": 4.5257547677663024e-05, "loss": 0.0091, "step": 22150 }, { "grad_norm": 0.16041971743106842, "learning_rate": 4.52163989174471e-05, "loss": 0.0088, "step": 22160 }, { "grad_norm": 0.14155414700508118, "learning_rate": 4.51752534267774e-05, "loss": 0.0094, "step": 22170 }, { "grad_norm": 0.13387058675289154, "learning_rate": 4.513411123377649e-05, "loss": 0.01, "step": 22180 }, { "grad_norm": 0.16608300805091858, "learning_rate": 4.5092972366564675e-05, "loss": 0.0087, "step": 22190 }, { "grad_norm": 0.18263712525367737, "learning_rate": 4.505183685325997e-05, "loss": 0.0092, "step": 22200 }, { "grad_norm": 0.1600879281759262, "learning_rate": 4.5010704721978125e-05, "loss": 0.01, "step": 22210 }, { "grad_norm": 0.20194707810878754, "learning_rate": 4.496957600083255e-05, "loss": 0.0112, "step": 22220 }, { "grad_norm": 0.15174196660518646, "learning_rate": 4.4928450717934343e-05, "loss": 0.0096, "step": 22230 }, { "grad_norm": 0.16890500485897064, "learning_rate": 4.488732890139227e-05, "loss": 0.0101, "step": 22240 }, { "grad_norm": 0.14241938292980194, "learning_rate": 4.4846210579312665e-05, "loss": 0.0142, "step": 22250 }, { "grad_norm": 0.20596544444561005, "learning_rate": 4.480509577979953e-05, "loss": 0.0115, "step": 22260 }, { "grad_norm": 0.13545775413513184, "learning_rate": 4.476398453095445e-05, "loss": 0.0086, "step": 22270 }, { "grad_norm": 0.13788026571273804, "learning_rate": 4.472287686087656e-05, "loss": 0.0089, "step": 22280 }, { "grad_norm": 0.17648287117481232, "learning_rate": 4.468177279766259e-05, "loss": 0.0095, "step": 22290 }, { "grad_norm": 0.1654939204454422, "learning_rate": 4.4640672369406746e-05, "loss": 0.0084, "step": 22300 }, { "grad_norm": 0.1282089650630951, "learning_rate": 4.459957560420082e-05, "loss": 0.0097, "step": 22310 }, { "grad_norm": 0.18305744230747223, "learning_rate": 4.455848253013403e-05, "loss": 0.0091, "step": 22320 }, { "grad_norm": 0.20834819972515106, "learning_rate": 4.4517393175293146e-05, "loss": 0.013, "step": 22330 }, { "grad_norm": 0.13699379563331604, "learning_rate": 4.447630756776232e-05, "loss": 0.0097, "step": 22340 }, { "grad_norm": 0.14017580449581146, "learning_rate": 4.443522573562318e-05, "loss": 0.0087, "step": 22350 }, { "grad_norm": 0.15667623281478882, "learning_rate": 4.4394147706954776e-05, "loss": 0.0092, "step": 22360 }, { "grad_norm": 0.14636287093162537, "learning_rate": 4.435307350983355e-05, "loss": 0.0093, "step": 22370 }, { "grad_norm": 0.15820379555225372, "learning_rate": 4.4312003172333326e-05, "loss": 0.0079, "step": 22380 }, { "grad_norm": 0.24704648554325104, "learning_rate": 4.427093672252531e-05, "loss": 0.0118, "step": 22390 }, { "grad_norm": 0.15999315679073334, "learning_rate": 4.422987418847802e-05, "loss": 0.0116, "step": 22400 }, { "grad_norm": 0.1596713364124298, "learning_rate": 4.4188815598257325e-05, "loss": 0.0092, "step": 22410 }, { "grad_norm": 0.11308643966913223, "learning_rate": 4.414776097992638e-05, "loss": 0.0113, "step": 22420 }, { "grad_norm": 0.25657227635383606, "learning_rate": 4.4106710361545595e-05, "loss": 0.0117, "step": 22430 }, { "grad_norm": 0.1810186207294464, "learning_rate": 4.406566377117272e-05, "loss": 0.0118, "step": 22440 }, { "grad_norm": 0.20708689093589783, "learning_rate": 4.40246212368627e-05, "loss": 0.0097, "step": 22450 }, { "grad_norm": 0.19386142492294312, "learning_rate": 4.3983582786667715e-05, "loss": 0.0091, "step": 22460 }, { "grad_norm": 0.13830900192260742, "learning_rate": 4.394254844863716e-05, "loss": 0.0098, "step": 22470 }, { "grad_norm": 0.18090800940990448, "learning_rate": 4.390151825081762e-05, "loss": 0.0089, "step": 22480 }, { "grad_norm": 0.14789050817489624, "learning_rate": 4.386049222125286e-05, "loss": 0.009, "step": 22490 }, { "grad_norm": 0.12487675249576569, "learning_rate": 4.3819470387983774e-05, "loss": 0.0097, "step": 22500 }, { "grad_norm": 0.12554627656936646, "learning_rate": 4.377845277904841e-05, "loss": 0.0087, "step": 22510 }, { "grad_norm": 0.11496446281671524, "learning_rate": 4.37374394224819e-05, "loss": 0.0088, "step": 22520 }, { "grad_norm": 0.12767930328845978, "learning_rate": 4.369643034631648e-05, "loss": 0.0088, "step": 22530 }, { "grad_norm": 0.16610124707221985, "learning_rate": 4.365542557858149e-05, "loss": 0.0107, "step": 22540 }, { "grad_norm": 0.11620879918336868, "learning_rate": 4.361442514730329e-05, "loss": 0.0086, "step": 22550 }, { "grad_norm": 0.19057996571063995, "learning_rate": 4.357342908050528e-05, "loss": 0.0095, "step": 22560 }, { "grad_norm": 0.1616400182247162, "learning_rate": 4.3532437406207895e-05, "loss": 0.0091, "step": 22570 }, { "grad_norm": 0.1083824634552002, "learning_rate": 4.349145015242856e-05, "loss": 0.0078, "step": 22580 }, { "grad_norm": 0.1289709210395813, "learning_rate": 4.345046734718168e-05, "loss": 0.0079, "step": 22590 }, { "grad_norm": 0.1672676056623459, "learning_rate": 4.34094890184786e-05, "loss": 0.009, "step": 22600 }, { "grad_norm": 0.229971781373024, "learning_rate": 4.336851519432765e-05, "loss": 0.0091, "step": 22610 }, { "grad_norm": 0.17204613983631134, "learning_rate": 4.332754590273403e-05, "loss": 0.0073, "step": 22620 }, { "grad_norm": 0.19403840601444244, "learning_rate": 4.3286581171699855e-05, "loss": 0.0092, "step": 22630 }, { "grad_norm": 0.17460408806800842, "learning_rate": 4.324562102922416e-05, "loss": 0.0091, "step": 22640 }, { "grad_norm": 0.20740288496017456, "learning_rate": 4.320466550330278e-05, "loss": 0.0118, "step": 22650 }, { "grad_norm": 0.12084619700908661, "learning_rate": 4.3163714621928466e-05, "loss": 0.0096, "step": 22660 }, { "grad_norm": 0.233347088098526, "learning_rate": 4.312276841309074e-05, "loss": 0.0148, "step": 22670 }, { "grad_norm": 0.2966384291648865, "learning_rate": 4.3081826904775945e-05, "loss": 0.011, "step": 22680 }, { "grad_norm": 0.17335480451583862, "learning_rate": 4.3040890124967246e-05, "loss": 0.0103, "step": 22690 }, { "grad_norm": 0.17352226376533508, "learning_rate": 4.2999958101644537e-05, "loss": 0.0096, "step": 22700 }, { "grad_norm": 0.14994242787361145, "learning_rate": 4.2959030862784435e-05, "loss": 0.011, "step": 22710 }, { "grad_norm": 0.129708394408226, "learning_rate": 4.291810843636036e-05, "loss": 0.008, "step": 22720 }, { "grad_norm": 0.1322478950023651, "learning_rate": 4.2877190850342375e-05, "loss": 0.0086, "step": 22730 }, { "grad_norm": 0.1686919927597046, "learning_rate": 4.2836278132697294e-05, "loss": 0.0101, "step": 22740 }, { "grad_norm": 0.1283806413412094, "learning_rate": 4.279537031138855e-05, "loss": 0.0095, "step": 22750 }, { "grad_norm": 0.15914484858512878, "learning_rate": 4.275446741437625e-05, "loss": 0.0079, "step": 22760 }, { "grad_norm": 0.1662466824054718, "learning_rate": 4.2713569469617176e-05, "loss": 0.0107, "step": 22770 }, { "grad_norm": 0.17468325793743134, "learning_rate": 4.267267650506465e-05, "loss": 0.0096, "step": 22780 }, { "grad_norm": 0.16344209015369415, "learning_rate": 4.263178854866866e-05, "loss": 0.0103, "step": 22790 }, { "grad_norm": 0.1923576295375824, "learning_rate": 4.259090562837571e-05, "loss": 0.0096, "step": 22800 }, { "grad_norm": 0.23143842816352844, "learning_rate": 4.255002777212888e-05, "loss": 0.0128, "step": 22810 }, { "grad_norm": 0.1681886613368988, "learning_rate": 4.250915500786783e-05, "loss": 0.007, "step": 22820 }, { "grad_norm": 0.19338561594486237, "learning_rate": 4.24682873635287e-05, "loss": 0.0101, "step": 22830 }, { "grad_norm": 0.15600846707820892, "learning_rate": 4.242742486704414e-05, "loss": 0.0087, "step": 22840 }, { "grad_norm": 0.20738162100315094, "learning_rate": 4.238656754634327e-05, "loss": 0.0107, "step": 22850 }, { "grad_norm": 0.141790971159935, "learning_rate": 4.234571542935168e-05, "loss": 0.0096, "step": 22860 }, { "grad_norm": 0.19732415676116943, "learning_rate": 4.230486854399144e-05, "loss": 0.01, "step": 22870 }, { "grad_norm": 0.13993710279464722, "learning_rate": 4.226402691818098e-05, "loss": 0.0101, "step": 22880 }, { "grad_norm": 0.21668848395347595, "learning_rate": 4.2223190579835196e-05, "loss": 0.0087, "step": 22890 }, { "grad_norm": 0.19783815741539001, "learning_rate": 4.218235955686531e-05, "loss": 0.0115, "step": 22900 }, { "grad_norm": 0.13472768664360046, "learning_rate": 4.214153387717894e-05, "loss": 0.0093, "step": 22910 }, { "grad_norm": 0.1653233766555786, "learning_rate": 4.210071356868007e-05, "loss": 0.0099, "step": 22920 }, { "grad_norm": 0.1708710938692093, "learning_rate": 4.205989865926898e-05, "loss": 0.0085, "step": 22930 }, { "grad_norm": 0.17477566003799438, "learning_rate": 4.2019089176842294e-05, "loss": 0.0091, "step": 22940 }, { "grad_norm": 0.1437450349330902, "learning_rate": 4.1978285149292894e-05, "loss": 0.01, "step": 22950 }, { "grad_norm": 0.21089521050453186, "learning_rate": 4.193748660450996e-05, "loss": 0.0102, "step": 22960 }, { "grad_norm": 0.17764651775360107, "learning_rate": 4.189669357037891e-05, "loss": 0.0097, "step": 22970 }, { "grad_norm": 0.13604795932769775, "learning_rate": 4.1855906074781405e-05, "loss": 0.0079, "step": 22980 }, { "grad_norm": 0.1459254026412964, "learning_rate": 4.1815124145595285e-05, "loss": 0.0071, "step": 22990 }, { "grad_norm": 0.14720682799816132, "learning_rate": 4.1774347810694644e-05, "loss": 0.0098, "step": 23000 }, { "grad_norm": 0.18865816295146942, "learning_rate": 4.17335770979497e-05, "loss": 0.0087, "step": 23010 }, { "grad_norm": 0.14644157886505127, "learning_rate": 4.169281203522687e-05, "loss": 0.0089, "step": 23020 }, { "grad_norm": 0.14442762732505798, "learning_rate": 4.1652052650388674e-05, "loss": 0.0074, "step": 23030 }, { "grad_norm": 0.1673809289932251, "learning_rate": 4.1611298971293786e-05, "loss": 0.0093, "step": 23040 }, { "grad_norm": 0.18080320954322815, "learning_rate": 4.1570551025796935e-05, "loss": 0.008, "step": 23050 }, { "grad_norm": 0.18846020102500916, "learning_rate": 4.152980884174897e-05, "loss": 0.0105, "step": 23060 }, { "grad_norm": 0.1685812622308731, "learning_rate": 4.148907244699682e-05, "loss": 0.0085, "step": 23070 }, { "grad_norm": 0.17945100367069244, "learning_rate": 4.1448341869383395e-05, "loss": 0.0113, "step": 23080 }, { "grad_norm": 0.19639286398887634, "learning_rate": 4.140761713674765e-05, "loss": 0.0095, "step": 23090 }, { "grad_norm": 0.1855611652135849, "learning_rate": 4.1366898276924574e-05, "loss": 0.0091, "step": 23100 }, { "grad_norm": 0.17852698266506195, "learning_rate": 4.132618531774512e-05, "loss": 0.0085, "step": 23110 }, { "grad_norm": 0.19451630115509033, "learning_rate": 4.128547828703622e-05, "loss": 0.0075, "step": 23120 }, { "grad_norm": 0.16148187220096588, "learning_rate": 4.1244777212620725e-05, "loss": 0.0086, "step": 23130 }, { "grad_norm": 0.17725040018558502, "learning_rate": 4.120408212231746e-05, "loss": 0.0082, "step": 23140 }, { "grad_norm": 0.21566087007522583, "learning_rate": 4.116339304394111e-05, "loss": 0.0098, "step": 23150 }, { "grad_norm": 0.154118612408638, "learning_rate": 4.112271000530229e-05, "loss": 0.0098, "step": 23160 }, { "grad_norm": 0.1068568006157875, "learning_rate": 4.10820330342075e-05, "loss": 0.0082, "step": 23170 }, { "grad_norm": 0.12351270020008087, "learning_rate": 4.1041362158459027e-05, "loss": 0.0079, "step": 23180 }, { "grad_norm": 0.1610109657049179, "learning_rate": 4.1000697405855024e-05, "loss": 0.0113, "step": 23190 }, { "grad_norm": 0.17196112871170044, "learning_rate": 4.096003880418951e-05, "loss": 0.0134, "step": 23200 }, { "grad_norm": 0.13771601021289825, "learning_rate": 4.0919386381252215e-05, "loss": 0.0088, "step": 23210 }, { "grad_norm": 0.18080484867095947, "learning_rate": 4.087874016482872e-05, "loss": 0.0083, "step": 23220 }, { "grad_norm": 0.26712629199028015, "learning_rate": 4.0838100182700295e-05, "loss": 0.0097, "step": 23230 }, { "grad_norm": 0.17121641337871552, "learning_rate": 4.079746646264402e-05, "loss": 0.0097, "step": 23240 }, { "grad_norm": 0.15015770494937897, "learning_rate": 4.075683903243262e-05, "loss": 0.0079, "step": 23250 }, { "grad_norm": 0.1223372370004654, "learning_rate": 4.071621791983462e-05, "loss": 0.0077, "step": 23260 }, { "grad_norm": 0.14583788812160492, "learning_rate": 4.06756031526141e-05, "loss": 0.0086, "step": 23270 }, { "grad_norm": 0.2303861528635025, "learning_rate": 4.063499475853092e-05, "loss": 0.011, "step": 23280 }, { "grad_norm": 0.16934718191623688, "learning_rate": 4.0594392765340506e-05, "loss": 0.0087, "step": 23290 }, { "grad_norm": 0.1568983644247055, "learning_rate": 4.0553797200793954e-05, "loss": 0.0083, "step": 23300 }, { "grad_norm": 0.15806810557842255, "learning_rate": 4.0513208092637926e-05, "loss": 0.0087, "step": 23310 }, { "grad_norm": 0.1385456770658493, "learning_rate": 4.0472625468614735e-05, "loss": 0.0105, "step": 23320 }, { "grad_norm": 0.16203418374061584, "learning_rate": 4.043204935646218e-05, "loss": 0.0079, "step": 23330 }, { "grad_norm": 0.166556715965271, "learning_rate": 4.0391479783913675e-05, "loss": 0.007, "step": 23340 }, { "grad_norm": 0.17983418703079224, "learning_rate": 4.0350916778698155e-05, "loss": 0.0086, "step": 23350 }, { "grad_norm": 0.11440864950418472, "learning_rate": 4.031036036854001e-05, "loss": 0.0099, "step": 23360 }, { "grad_norm": 0.0858214721083641, "learning_rate": 4.026981058115918e-05, "loss": 0.0094, "step": 23370 }, { "grad_norm": 0.1568388193845749, "learning_rate": 4.022926744427108e-05, "loss": 0.0121, "step": 23380 }, { "grad_norm": 0.11547496914863586, "learning_rate": 4.018873098558654e-05, "loss": 0.0084, "step": 23390 }, { "grad_norm": 0.14330632984638214, "learning_rate": 4.014820123281186e-05, "loss": 0.009, "step": 23400 }, { "grad_norm": 0.14270655810832977, "learning_rate": 4.0107678213648735e-05, "loss": 0.0094, "step": 23410 }, { "grad_norm": 0.1546497493982315, "learning_rate": 4.006716195579428e-05, "loss": 0.0122, "step": 23420 }, { "grad_norm": 0.1924239546060562, "learning_rate": 4.002665248694096e-05, "loss": 0.0118, "step": 23430 }, { "grad_norm": 0.15750901401042938, "learning_rate": 3.998614983477664e-05, "loss": 0.009, "step": 23440 }, { "grad_norm": 0.13286122679710388, "learning_rate": 3.994565402698448e-05, "loss": 0.0103, "step": 23450 }, { "grad_norm": 0.1648426502943039, "learning_rate": 3.9905165091242975e-05, "loss": 0.0103, "step": 23460 }, { "grad_norm": 0.16883952915668488, "learning_rate": 3.9864683055225936e-05, "loss": 0.0096, "step": 23470 }, { "grad_norm": 0.15587382018566132, "learning_rate": 3.982420794660247e-05, "loss": 0.0078, "step": 23480 }, { "grad_norm": 0.1625128984451294, "learning_rate": 3.978373979303691e-05, "loss": 0.0062, "step": 23490 }, { "grad_norm": 0.1778608113527298, "learning_rate": 3.974327862218888e-05, "loss": 0.0096, "step": 23500 }, { "grad_norm": 0.16072118282318115, "learning_rate": 3.970282446171318e-05, "loss": 0.0104, "step": 23510 }, { "grad_norm": 0.25754478573799133, "learning_rate": 3.966237733925988e-05, "loss": 0.0113, "step": 23520 }, { "grad_norm": 0.17876853048801422, "learning_rate": 3.962193728247418e-05, "loss": 0.0105, "step": 23530 }, { "grad_norm": 0.21364620327949524, "learning_rate": 3.958150431899651e-05, "loss": 0.0111, "step": 23540 }, { "grad_norm": 0.15404334664344788, "learning_rate": 3.954107847646238e-05, "loss": 0.0084, "step": 23550 }, { "grad_norm": 0.14927391707897186, "learning_rate": 3.950065978250249e-05, "loss": 0.0075, "step": 23560 }, { "grad_norm": 0.1591189205646515, "learning_rate": 3.9460248264742624e-05, "loss": 0.0075, "step": 23570 }, { "grad_norm": 0.172512486577034, "learning_rate": 3.941984395080371e-05, "loss": 0.0117, "step": 23580 }, { "grad_norm": 0.15499533712863922, "learning_rate": 3.937944686830167e-05, "loss": 0.0085, "step": 23590 }, { "grad_norm": 0.17680305242538452, "learning_rate": 3.933905704484756e-05, "loss": 0.0091, "step": 23600 }, { "grad_norm": 0.12941232323646545, "learning_rate": 3.929867450804743e-05, "loss": 0.0077, "step": 23610 }, { "grad_norm": 0.10729007422924042, "learning_rate": 3.925829928550237e-05, "loss": 0.007, "step": 23620 }, { "grad_norm": 0.15455341339111328, "learning_rate": 3.921793140480847e-05, "loss": 0.0081, "step": 23630 }, { "grad_norm": 0.16511620581150055, "learning_rate": 3.917757089355677e-05, "loss": 0.0083, "step": 23640 }, { "grad_norm": 0.18628345429897308, "learning_rate": 3.9137217779333326e-05, "loss": 0.0075, "step": 23650 }, { "grad_norm": 0.13313372433185577, "learning_rate": 3.9096872089719083e-05, "loss": 0.0078, "step": 23660 }, { "grad_norm": 0.1834368109703064, "learning_rate": 3.905653385228996e-05, "loss": 0.0087, "step": 23670 }, { "grad_norm": 0.13535164296627045, "learning_rate": 3.901620309461677e-05, "loss": 0.0067, "step": 23680 }, { "grad_norm": 0.24694545567035675, "learning_rate": 3.897587984426518e-05, "loss": 0.0081, "step": 23690 }, { "grad_norm": 0.14264222979545593, "learning_rate": 3.893556412879577e-05, "loss": 0.0081, "step": 23700 }, { "grad_norm": 0.1608908325433731, "learning_rate": 3.889525597576395e-05, "loss": 0.0078, "step": 23710 }, { "grad_norm": 0.1503826230764389, "learning_rate": 3.8854955412719965e-05, "loss": 0.0101, "step": 23720 }, { "grad_norm": 0.14537063241004944, "learning_rate": 3.881466246720887e-05, "loss": 0.0083, "step": 23730 }, { "grad_norm": 0.11648821085691452, "learning_rate": 3.8774377166770484e-05, "loss": 0.0103, "step": 23740 }, { "grad_norm": 0.18481002748012543, "learning_rate": 3.8734099538939474e-05, "loss": 0.009, "step": 23750 }, { "grad_norm": 0.1297985315322876, "learning_rate": 3.869382961124518e-05, "loss": 0.0093, "step": 23760 }, { "grad_norm": 0.17708592116832733, "learning_rate": 3.8653567411211736e-05, "loss": 0.0078, "step": 23770 }, { "grad_norm": 0.1576770842075348, "learning_rate": 3.8613312966357987e-05, "loss": 0.0091, "step": 23780 }, { "grad_norm": 0.22544685006141663, "learning_rate": 3.857306630419745e-05, "loss": 0.009, "step": 23790 }, { "grad_norm": 0.17928434908390045, "learning_rate": 3.853282745223834e-05, "loss": 0.0087, "step": 23800 }, { "grad_norm": 0.2195926457643509, "learning_rate": 3.8492596437983546e-05, "loss": 0.0081, "step": 23810 }, { "grad_norm": 0.24687062203884125, "learning_rate": 3.8452373288930586e-05, "loss": 0.0101, "step": 23820 }, { "grad_norm": 0.13419346511363983, "learning_rate": 3.841215803257159e-05, "loss": 0.0071, "step": 23830 }, { "grad_norm": 0.09928826987743378, "learning_rate": 3.83719506963933e-05, "loss": 0.0079, "step": 23840 }, { "grad_norm": 0.14878590404987335, "learning_rate": 3.8331751307877087e-05, "loss": 0.0089, "step": 23850 }, { "grad_norm": 0.1784316450357437, "learning_rate": 3.82915598944988e-05, "loss": 0.0079, "step": 23860 }, { "grad_norm": 0.1523984968662262, "learning_rate": 3.825137648372893e-05, "loss": 0.0085, "step": 23870 }, { "grad_norm": 0.1558789312839508, "learning_rate": 3.8211201103032465e-05, "loss": 0.0089, "step": 23880 }, { "grad_norm": 0.15293699502944946, "learning_rate": 3.817103377986887e-05, "loss": 0.0105, "step": 23890 }, { "grad_norm": 0.17686718702316284, "learning_rate": 3.813087454169215e-05, "loss": 0.0091, "step": 23900 }, { "grad_norm": 0.14316199719905853, "learning_rate": 3.809072341595078e-05, "loss": 0.0084, "step": 23910 }, { "grad_norm": 0.19222398102283478, "learning_rate": 3.8050580430087636e-05, "loss": 0.0078, "step": 23920 }, { "grad_norm": 0.2018560916185379, "learning_rate": 3.8010445611540096e-05, "loss": 0.0073, "step": 23930 }, { "grad_norm": 0.18132203817367554, "learning_rate": 3.797031898773992e-05, "loss": 0.0078, "step": 23940 }, { "grad_norm": 0.2611755430698395, "learning_rate": 3.793020058611329e-05, "loss": 0.0085, "step": 23950 }, { "grad_norm": 0.16950997710227966, "learning_rate": 3.789009043408074e-05, "loss": 0.0102, "step": 23960 }, { "grad_norm": 0.13657249510288239, "learning_rate": 3.7849988559057194e-05, "loss": 0.0067, "step": 23970 }, { "grad_norm": 0.1345566064119339, "learning_rate": 3.78098949884519e-05, "loss": 0.007, "step": 23980 }, { "grad_norm": 0.13491509854793549, "learning_rate": 3.776980974966843e-05, "loss": 0.0086, "step": 23990 }, { "grad_norm": 0.14773620665073395, "learning_rate": 3.772973287010468e-05, "loss": 0.0088, "step": 24000 }, { "grad_norm": 0.12668238580226898, "learning_rate": 3.768966437715283e-05, "loss": 0.0084, "step": 24010 }, { "grad_norm": 0.1016845852136612, "learning_rate": 3.7649604298199274e-05, "loss": 0.0098, "step": 24020 }, { "grad_norm": 0.16371393203735352, "learning_rate": 3.760955266062473e-05, "loss": 0.0081, "step": 24030 }, { "grad_norm": 0.1545923948287964, "learning_rate": 3.75695094918041e-05, "loss": 0.0097, "step": 24040 }, { "grad_norm": 0.11532596498727798, "learning_rate": 3.752947481910652e-05, "loss": 0.0089, "step": 24050 }, { "grad_norm": 0.12634488940238953, "learning_rate": 3.7489448669895324e-05, "loss": 0.0108, "step": 24060 }, { "grad_norm": 0.16113051772117615, "learning_rate": 3.744943107152798e-05, "loss": 0.0085, "step": 24070 }, { "grad_norm": 0.1678769439458847, "learning_rate": 3.7409422051356165e-05, "loss": 0.0074, "step": 24080 }, { "grad_norm": 0.21010474860668182, "learning_rate": 3.736942163672564e-05, "loss": 0.0105, "step": 24090 }, { "grad_norm": 0.17276184260845184, "learning_rate": 3.732942985497636e-05, "loss": 0.0085, "step": 24100 }, { "grad_norm": 0.16002923250198364, "learning_rate": 3.728944673344228e-05, "loss": 0.0079, "step": 24110 }, { "grad_norm": 0.1909472793340683, "learning_rate": 3.72494722994515e-05, "loss": 0.0095, "step": 24120 }, { "grad_norm": 0.13829447329044342, "learning_rate": 3.720950658032617e-05, "loss": 0.0078, "step": 24130 }, { "grad_norm": 0.17658403515815735, "learning_rate": 3.716954960338249e-05, "loss": 0.0089, "step": 24140 }, { "grad_norm": 0.16481846570968628, "learning_rate": 3.712960139593066e-05, "loss": 0.0077, "step": 24150 }, { "grad_norm": 0.1370721459388733, "learning_rate": 3.708966198527493e-05, "loss": 0.0092, "step": 24160 }, { "grad_norm": 0.20559240877628326, "learning_rate": 3.704973139871349e-05, "loss": 0.0072, "step": 24170 }, { "grad_norm": 0.16593150794506073, "learning_rate": 3.700980966353853e-05, "loss": 0.0109, "step": 24180 }, { "grad_norm": 0.14594201743602753, "learning_rate": 3.696989680703619e-05, "loss": 0.0106, "step": 24190 }, { "grad_norm": 0.15142013132572174, "learning_rate": 3.69299928564865e-05, "loss": 0.0079, "step": 24200 }, { "grad_norm": 0.14694710075855255, "learning_rate": 3.689009783916345e-05, "loss": 0.0078, "step": 24210 }, { "grad_norm": 0.15246056020259857, "learning_rate": 3.6850211782334895e-05, "loss": 0.0117, "step": 24220 }, { "grad_norm": 0.16140730679035187, "learning_rate": 3.681033471326261e-05, "loss": 0.0098, "step": 24230 }, { "grad_norm": 0.1276780366897583, "learning_rate": 3.677046665920216e-05, "loss": 0.0088, "step": 24240 }, { "grad_norm": 0.19647186994552612, "learning_rate": 3.6730607647403005e-05, "loss": 0.0091, "step": 24250 }, { "grad_norm": 0.13342730700969696, "learning_rate": 3.6690757705108416e-05, "loss": 0.0084, "step": 24260 }, { "grad_norm": 0.12762877345085144, "learning_rate": 3.665091685955542e-05, "loss": 0.0076, "step": 24270 }, { "grad_norm": 0.10752277076244354, "learning_rate": 3.6611085137974896e-05, "loss": 0.0085, "step": 24280 }, { "grad_norm": 0.17256738245487213, "learning_rate": 3.657126256759143e-05, "loss": 0.0109, "step": 24290 }, { "grad_norm": 0.14392928779125214, "learning_rate": 3.653144917562335e-05, "loss": 0.0097, "step": 24300 }, { "grad_norm": 0.13581931591033936, "learning_rate": 3.649164498928277e-05, "loss": 0.0097, "step": 24310 }, { "grad_norm": 0.12099150568246841, "learning_rate": 3.645185003577546e-05, "loss": 0.0079, "step": 24320 }, { "grad_norm": 0.12267202883958817, "learning_rate": 3.6412064342300906e-05, "loss": 0.008, "step": 24330 }, { "grad_norm": 0.1363220065832138, "learning_rate": 3.637228793605224e-05, "loss": 0.0087, "step": 24340 }, { "grad_norm": 0.16230922937393188, "learning_rate": 3.6332520844216264e-05, "loss": 0.0077, "step": 24350 }, { "grad_norm": 0.18120385706424713, "learning_rate": 3.6292763093973425e-05, "loss": 0.0096, "step": 24360 }, { "grad_norm": 0.1811811774969101, "learning_rate": 3.6253014712497754e-05, "loss": 0.0096, "step": 24370 }, { "grad_norm": 0.139017716050148, "learning_rate": 3.621327572695692e-05, "loss": 0.0085, "step": 24380 }, { "grad_norm": 0.1937042623758316, "learning_rate": 3.617354616451211e-05, "loss": 0.0083, "step": 24390 }, { "grad_norm": 0.09292682260274887, "learning_rate": 3.6133826052318116e-05, "loss": 0.008, "step": 24400 }, { "grad_norm": 0.17125628888607025, "learning_rate": 3.609411541752327e-05, "loss": 0.0089, "step": 24410 }, { "grad_norm": 0.1350843906402588, "learning_rate": 3.6054414287269405e-05, "loss": 0.0072, "step": 24420 }, { "grad_norm": 0.1548476666212082, "learning_rate": 3.601472268869188e-05, "loss": 0.0096, "step": 24430 }, { "grad_norm": 0.16504782438278198, "learning_rate": 3.597504064891952e-05, "loss": 0.0103, "step": 24440 }, { "grad_norm": 0.16333124041557312, "learning_rate": 3.5935368195074636e-05, "loss": 0.0085, "step": 24450 }, { "grad_norm": 0.15777702629566193, "learning_rate": 3.589570535427297e-05, "loss": 0.0069, "step": 24460 }, { "grad_norm": 0.18699952960014343, "learning_rate": 3.585605215362371e-05, "loss": 0.0075, "step": 24470 }, { "grad_norm": 0.1381383091211319, "learning_rate": 3.581640862022941e-05, "loss": 0.0083, "step": 24480 }, { "grad_norm": 0.21325956284999847, "learning_rate": 3.57767747811861e-05, "loss": 0.0102, "step": 24490 }, { "grad_norm": 0.23153972625732422, "learning_rate": 3.573715066358308e-05, "loss": 0.0111, "step": 24500 }, { "grad_norm": 0.14684639871120453, "learning_rate": 3.569753629450311e-05, "loss": 0.0081, "step": 24510 }, { "grad_norm": 0.19924578070640564, "learning_rate": 3.565793170102221e-05, "loss": 0.0093, "step": 24520 }, { "grad_norm": 0.0957271084189415, "learning_rate": 3.561833691020976e-05, "loss": 0.0087, "step": 24530 }, { "grad_norm": 0.16303791105747223, "learning_rate": 3.5578751949128415e-05, "loss": 0.0121, "step": 24540 }, { "grad_norm": 0.13646213710308075, "learning_rate": 3.5539176844834125e-05, "loss": 0.0089, "step": 24550 }, { "grad_norm": 0.1639191210269928, "learning_rate": 3.5499611624376125e-05, "loss": 0.0093, "step": 24560 }, { "grad_norm": 0.11365716904401779, "learning_rate": 3.546005631479684e-05, "loss": 0.0103, "step": 24570 }, { "grad_norm": 0.15267601609230042, "learning_rate": 3.542051094313196e-05, "loss": 0.0073, "step": 24580 }, { "grad_norm": 0.14562954008579254, "learning_rate": 3.5380975536410364e-05, "loss": 0.0075, "step": 24590 }, { "grad_norm": 0.25600287318229675, "learning_rate": 3.534145012165415e-05, "loss": 0.0105, "step": 24600 }, { "grad_norm": 0.151582732796669, "learning_rate": 3.5301934725878546e-05, "loss": 0.0088, "step": 24610 }, { "grad_norm": 0.16861668229103088, "learning_rate": 3.526242937609197e-05, "loss": 0.01, "step": 24620 }, { "grad_norm": 0.16508004069328308, "learning_rate": 3.522293409929595e-05, "loss": 0.008, "step": 24630 }, { "grad_norm": 0.15587319433689117, "learning_rate": 3.518344892248513e-05, "loss": 0.0071, "step": 24640 }, { "grad_norm": 0.11843466758728027, "learning_rate": 3.514397387264725e-05, "loss": 0.008, "step": 24650 }, { "grad_norm": 0.18179957568645477, "learning_rate": 3.5104508976763176e-05, "loss": 0.0086, "step": 24660 }, { "grad_norm": 0.14662505686283112, "learning_rate": 3.506505426180674e-05, "loss": 0.0087, "step": 24670 }, { "grad_norm": 0.21444036066532135, "learning_rate": 3.502560975474488e-05, "loss": 0.0096, "step": 24680 }, { "grad_norm": 0.16713008284568787, "learning_rate": 3.4986175482537566e-05, "loss": 0.0097, "step": 24690 }, { "grad_norm": 0.14835524559020996, "learning_rate": 3.4946751472137725e-05, "loss": 0.0084, "step": 24700 }, { "grad_norm": 0.15169331431388855, "learning_rate": 3.490733775049132e-05, "loss": 0.0068, "step": 24710 }, { "grad_norm": 0.1676727682352066, "learning_rate": 3.4867934344537236e-05, "loss": 0.0109, "step": 24720 }, { "grad_norm": 0.20205558836460114, "learning_rate": 3.482854128120735e-05, "loss": 0.0078, "step": 24730 }, { "grad_norm": 0.1609681248664856, "learning_rate": 3.478915858742643e-05, "loss": 0.0087, "step": 24740 }, { "grad_norm": 0.15392738580703735, "learning_rate": 3.4749786290112205e-05, "loss": 0.0084, "step": 24750 }, { "grad_norm": 0.14958222210407257, "learning_rate": 3.471042441617524e-05, "loss": 0.0106, "step": 24760 }, { "grad_norm": 0.12645673751831055, "learning_rate": 3.467107299251902e-05, "loss": 0.0071, "step": 24770 }, { "grad_norm": 0.11025139689445496, "learning_rate": 3.463173204603984e-05, "loss": 0.0077, "step": 24780 }, { "grad_norm": 0.1596340388059616, "learning_rate": 3.4592401603626924e-05, "loss": 0.0096, "step": 24790 }, { "grad_norm": 0.18013525009155273, "learning_rate": 3.45530816921622e-05, "loss": 0.0076, "step": 24800 }, { "grad_norm": 0.21763008832931519, "learning_rate": 3.451377233852051e-05, "loss": 0.0098, "step": 24810 }, { "grad_norm": 0.14522802829742432, "learning_rate": 3.4474473569569385e-05, "loss": 0.0075, "step": 24820 }, { "grad_norm": 0.18349409103393555, "learning_rate": 3.443518541216918e-05, "loss": 0.0081, "step": 24830 }, { "grad_norm": 0.1575125753879547, "learning_rate": 3.439590789317299e-05, "loss": 0.0069, "step": 24840 }, { "grad_norm": 0.13233239948749542, "learning_rate": 3.4356641039426607e-05, "loss": 0.0084, "step": 24850 }, { "grad_norm": 0.12208603322505951, "learning_rate": 3.431738487776857e-05, "loss": 0.0063, "step": 24860 }, { "grad_norm": 0.17462295293807983, "learning_rate": 3.4278139435030084e-05, "loss": 0.0072, "step": 24870 }, { "grad_norm": 0.12425214797258377, "learning_rate": 3.423890473803504e-05, "loss": 0.0077, "step": 24880 }, { "grad_norm": 0.15344862639904022, "learning_rate": 3.41996808136e-05, "loss": 0.0057, "step": 24890 }, { "grad_norm": 0.14947815239429474, "learning_rate": 3.416046768853413e-05, "loss": 0.0077, "step": 24900 }, { "grad_norm": 0.13253572583198547, "learning_rate": 3.412126538963925e-05, "loss": 0.0069, "step": 24910 }, { "grad_norm": 0.185004323720932, "learning_rate": 3.4082073943709727e-05, "loss": 0.0103, "step": 24920 }, { "grad_norm": 0.2201031595468521, "learning_rate": 3.404289337753258e-05, "loss": 0.0081, "step": 24930 }, { "grad_norm": 0.1884375959634781, "learning_rate": 3.400372371788736e-05, "loss": 0.0079, "step": 24940 }, { "grad_norm": 0.15046851336956024, "learning_rate": 3.3964564991546124e-05, "loss": 0.0074, "step": 24950 }, { "grad_norm": 0.14706042408943176, "learning_rate": 3.392541722527351e-05, "loss": 0.0087, "step": 24960 }, { "grad_norm": 0.11536967009305954, "learning_rate": 3.3886280445826644e-05, "loss": 0.0069, "step": 24970 }, { "grad_norm": 0.15914735198020935, "learning_rate": 3.3847154679955154e-05, "loss": 0.01, "step": 24980 }, { "grad_norm": 0.1345221847295761, "learning_rate": 3.380803995440113e-05, "loss": 0.0097, "step": 24990 }, { "grad_norm": 0.1956009715795517, "learning_rate": 3.3768936295899115e-05, "loss": 0.009, "step": 25000 }, { "grad_norm": 0.12389553338289261, "learning_rate": 3.3729843731176094e-05, "loss": 0.0075, "step": 25010 }, { "grad_norm": 0.15220966935157776, "learning_rate": 3.369076228695146e-05, "loss": 0.0078, "step": 25020 }, { "grad_norm": 0.16488762199878693, "learning_rate": 3.365169198993703e-05, "loss": 0.0086, "step": 25030 }, { "grad_norm": 0.1393457055091858, "learning_rate": 3.361263286683697e-05, "loss": 0.0068, "step": 25040 }, { "grad_norm": 0.10202004015445709, "learning_rate": 3.35735849443478e-05, "loss": 0.0089, "step": 25050 }, { "grad_norm": 0.12953998148441315, "learning_rate": 3.3534548249158435e-05, "loss": 0.009, "step": 25060 }, { "grad_norm": 0.13128970563411713, "learning_rate": 3.3495522807950086e-05, "loss": 0.0103, "step": 25070 }, { "grad_norm": 0.17047180235385895, "learning_rate": 3.345650864739627e-05, "loss": 0.0084, "step": 25080 }, { "grad_norm": 0.15138480067253113, "learning_rate": 3.3417505794162794e-05, "loss": 0.0074, "step": 25090 }, { "grad_norm": 0.11858129501342773, "learning_rate": 3.3378514274907745e-05, "loss": 0.009, "step": 25100 }, { "grad_norm": 0.1552431881427765, "learning_rate": 3.333953411628147e-05, "loss": 0.0075, "step": 25110 }, { "grad_norm": 0.1854693740606308, "learning_rate": 3.330056534492653e-05, "loss": 0.0094, "step": 25120 }, { "grad_norm": 0.1552249938249588, "learning_rate": 3.32616079874777e-05, "loss": 0.0081, "step": 25130 }, { "grad_norm": 0.14548230171203613, "learning_rate": 3.322266207056197e-05, "loss": 0.0086, "step": 25140 }, { "grad_norm": 0.17947082221508026, "learning_rate": 3.318372762079852e-05, "loss": 0.0078, "step": 25150 }, { "grad_norm": 0.1709752380847931, "learning_rate": 3.3144804664798666e-05, "loss": 0.0074, "step": 25160 }, { "grad_norm": 0.12860015034675598, "learning_rate": 3.3105893229165894e-05, "loss": 0.0078, "step": 25170 }, { "grad_norm": 0.14953996241092682, "learning_rate": 3.30669933404958e-05, "loss": 0.0085, "step": 25180 }, { "grad_norm": 0.16945916414260864, "learning_rate": 3.302810502537609e-05, "loss": 0.0081, "step": 25190 }, { "grad_norm": 0.12970419228076935, "learning_rate": 3.298922831038655e-05, "loss": 0.0061, "step": 25200 }, { "grad_norm": 0.16198530793190002, "learning_rate": 3.2950363222099073e-05, "loss": 0.0066, "step": 25210 }, { "grad_norm": 0.15791025757789612, "learning_rate": 3.291150978707758e-05, "loss": 0.0105, "step": 25220 }, { "grad_norm": 0.1777058243751526, "learning_rate": 3.287266803187798e-05, "loss": 0.0093, "step": 25230 }, { "grad_norm": 0.14590415358543396, "learning_rate": 3.283383798304829e-05, "loss": 0.0078, "step": 25240 }, { "grad_norm": 0.14229905605316162, "learning_rate": 3.279501966712847e-05, "loss": 0.0066, "step": 25250 }, { "grad_norm": 0.1475653499364853, "learning_rate": 3.275621311065047e-05, "loss": 0.0074, "step": 25260 }, { "grad_norm": 0.1679075062274933, "learning_rate": 3.271741834013822e-05, "loss": 0.009, "step": 25270 }, { "grad_norm": 0.1839408278465271, "learning_rate": 3.267863538210756e-05, "loss": 0.0079, "step": 25280 }, { "grad_norm": 0.15039697289466858, "learning_rate": 3.2639864263066296e-05, "loss": 0.0101, "step": 25290 }, { "grad_norm": 0.16613252460956573, "learning_rate": 3.26011050095141e-05, "loss": 0.0089, "step": 25300 }, { "grad_norm": 0.16627982258796692, "learning_rate": 3.256235764794259e-05, "loss": 0.0091, "step": 25310 }, { "grad_norm": 0.1435709446668625, "learning_rate": 3.2523622204835194e-05, "loss": 0.0095, "step": 25320 }, { "grad_norm": 0.12191150337457657, "learning_rate": 3.2484898706667214e-05, "loss": 0.0072, "step": 25330 }, { "grad_norm": 0.16257436573505402, "learning_rate": 3.2446187179905806e-05, "loss": 0.0075, "step": 25340 }, { "grad_norm": 0.13009889423847198, "learning_rate": 3.240748765100995e-05, "loss": 0.0066, "step": 25350 }, { "grad_norm": 0.21320869028568268, "learning_rate": 3.236880014643039e-05, "loss": 0.0072, "step": 25360 }, { "grad_norm": 0.22211192548274994, "learning_rate": 3.233012469260969e-05, "loss": 0.0092, "step": 25370 }, { "grad_norm": 0.1266310214996338, "learning_rate": 3.229146131598213e-05, "loss": 0.0092, "step": 25380 }, { "grad_norm": 0.14239194989204407, "learning_rate": 3.2252810042973794e-05, "loss": 0.0089, "step": 25390 }, { "grad_norm": 0.15983302891254425, "learning_rate": 3.2214170900002456e-05, "loss": 0.0091, "step": 25400 }, { "grad_norm": 0.13232725858688354, "learning_rate": 3.217554391347758e-05, "loss": 0.0112, "step": 25410 }, { "grad_norm": 0.09986216574907303, "learning_rate": 3.213692910980037e-05, "loss": 0.0098, "step": 25420 }, { "grad_norm": 0.16614672541618347, "learning_rate": 3.2098326515363666e-05, "loss": 0.0075, "step": 25430 }, { "grad_norm": 0.17848896980285645, "learning_rate": 3.205973615655199e-05, "loss": 0.0101, "step": 25440 }, { "grad_norm": 0.15017391741275787, "learning_rate": 3.202115805974149e-05, "loss": 0.006, "step": 25450 }, { "grad_norm": 0.2119695246219635, "learning_rate": 3.1982592251299916e-05, "loss": 0.0091, "step": 25460 }, { "grad_norm": 0.17046424746513367, "learning_rate": 3.1944038757586656e-05, "loss": 0.0124, "step": 25470 }, { "grad_norm": 0.17653916776180267, "learning_rate": 3.190549760495263e-05, "loss": 0.0104, "step": 25480 }, { "grad_norm": 0.13518168032169342, "learning_rate": 3.186696881974039e-05, "loss": 0.0096, "step": 25490 }, { "grad_norm": 0.21431665122509003, "learning_rate": 3.1828452428283986e-05, "loss": 0.0084, "step": 25500 }, { "grad_norm": 0.1148846447467804, "learning_rate": 3.178994845690898e-05, "loss": 0.0079, "step": 25510 }, { "grad_norm": 0.1641210913658142, "learning_rate": 3.17514569319325e-05, "loss": 0.0099, "step": 25520 }, { "grad_norm": 0.1573648750782013, "learning_rate": 3.171297787966312e-05, "loss": 0.0093, "step": 25530 }, { "grad_norm": 0.15985198318958282, "learning_rate": 3.167451132640093e-05, "loss": 0.0086, "step": 25540 }, { "grad_norm": 0.11128171533346176, "learning_rate": 3.163605729843746e-05, "loss": 0.0061, "step": 25550 }, { "grad_norm": 0.10359389334917068, "learning_rate": 3.159761582205565e-05, "loss": 0.0069, "step": 25560 }, { "grad_norm": 0.131092369556427, "learning_rate": 3.155918692352992e-05, "loss": 0.008, "step": 25570 }, { "grad_norm": 0.16517488658428192, "learning_rate": 3.152077062912602e-05, "loss": 0.0094, "step": 25580 }, { "grad_norm": 0.18804457783699036, "learning_rate": 3.148236696510117e-05, "loss": 0.0101, "step": 25590 }, { "grad_norm": 0.13479119539260864, "learning_rate": 3.144397595770388e-05, "loss": 0.0097, "step": 25600 }, { "grad_norm": 0.11441842466592789, "learning_rate": 3.1405597633174036e-05, "loss": 0.0072, "step": 25610 }, { "grad_norm": 0.12186238914728165, "learning_rate": 3.136723201774289e-05, "loss": 0.0081, "step": 25620 }, { "grad_norm": 0.1770278811454773, "learning_rate": 3.132887913763295e-05, "loss": 0.0088, "step": 25630 }, { "grad_norm": 0.12471599876880646, "learning_rate": 3.129053901905806e-05, "loss": 0.0072, "step": 25640 }, { "grad_norm": 0.13311338424682617, "learning_rate": 3.125221168822335e-05, "loss": 0.008, "step": 25650 }, { "grad_norm": 0.14354348182678223, "learning_rate": 3.1213897171325154e-05, "loss": 0.0075, "step": 25660 }, { "grad_norm": 0.14453960955142975, "learning_rate": 3.1175595494551116e-05, "loss": 0.0112, "step": 25670 }, { "grad_norm": 0.10320918262004852, "learning_rate": 3.1137306684080045e-05, "loss": 0.0087, "step": 25680 }, { "grad_norm": 0.10985134541988373, "learning_rate": 3.1099030766081985e-05, "loss": 0.0061, "step": 25690 }, { "grad_norm": 0.14319749176502228, "learning_rate": 3.106076776671818e-05, "loss": 0.0097, "step": 25700 }, { "grad_norm": 0.1460457146167755, "learning_rate": 3.102251771214101e-05, "loss": 0.0093, "step": 25710 }, { "grad_norm": 0.11897866427898407, "learning_rate": 3.098428062849404e-05, "loss": 0.0067, "step": 25720 }, { "grad_norm": 0.10589851438999176, "learning_rate": 3.094605654191195e-05, "loss": 0.0091, "step": 25730 }, { "grad_norm": 0.19099842011928558, "learning_rate": 3.090784547852055e-05, "loss": 0.0101, "step": 25740 }, { "grad_norm": 0.16906221210956573, "learning_rate": 3.0869647464436746e-05, "loss": 0.0087, "step": 25750 }, { "grad_norm": 0.2432193160057068, "learning_rate": 3.0831462525768496e-05, "loss": 0.0086, "step": 25760 }, { "grad_norm": 0.10995883494615555, "learning_rate": 3.079329068861488e-05, "loss": 0.0072, "step": 25770 }, { "grad_norm": 0.18140366673469543, "learning_rate": 3.075513197906597e-05, "loss": 0.009, "step": 25780 }, { "grad_norm": 0.1598038524389267, "learning_rate": 3.071698642320286e-05, "loss": 0.0087, "step": 25790 }, { "grad_norm": 0.119683176279068, "learning_rate": 3.067885404709772e-05, "loss": 0.0067, "step": 25800 }, { "grad_norm": 0.11622342467308044, "learning_rate": 3.0640734876813636e-05, "loss": 0.0076, "step": 25810 }, { "grad_norm": 0.1692061722278595, "learning_rate": 3.060262893840473e-05, "loss": 0.0087, "step": 25820 }, { "grad_norm": 0.10214614868164062, "learning_rate": 3.056453625791603e-05, "loss": 0.0096, "step": 25830 }, { "grad_norm": 0.1269676834344864, "learning_rate": 3.052645686138353e-05, "loss": 0.0082, "step": 25840 }, { "grad_norm": 0.11190276592969894, "learning_rate": 3.0488390774834153e-05, "loss": 0.0087, "step": 25850 }, { "grad_norm": 0.17829684913158417, "learning_rate": 3.0450338024285684e-05, "loss": 0.0111, "step": 25860 }, { "grad_norm": 0.13553091883659363, "learning_rate": 3.0412298635746855e-05, "loss": 0.0069, "step": 25870 }, { "grad_norm": 0.13884611427783966, "learning_rate": 3.03742726352172e-05, "loss": 0.0111, "step": 25880 }, { "grad_norm": 0.2188357561826706, "learning_rate": 3.0336260048687125e-05, "loss": 0.0081, "step": 25890 }, { "grad_norm": 0.15370725095272064, "learning_rate": 3.0298260902137897e-05, "loss": 0.0082, "step": 25900 }, { "grad_norm": 0.1789119690656662, "learning_rate": 3.0260275221541566e-05, "loss": 0.0096, "step": 25910 }, { "grad_norm": 0.1440209001302719, "learning_rate": 3.0222303032860987e-05, "loss": 0.0093, "step": 25920 }, { "grad_norm": 0.10535832494497299, "learning_rate": 3.018434436204979e-05, "loss": 0.0065, "step": 25930 }, { "grad_norm": 0.09712862223386765, "learning_rate": 3.014639923505237e-05, "loss": 0.0063, "step": 25940 }, { "grad_norm": 0.24213559925556183, "learning_rate": 3.0108467677803863e-05, "loss": 0.0077, "step": 25950 }, { "grad_norm": 0.1313050389289856, "learning_rate": 3.0070549716230156e-05, "loss": 0.0066, "step": 25960 }, { "grad_norm": 0.1516273319721222, "learning_rate": 3.003264537624777e-05, "loss": 0.0058, "step": 25970 }, { "grad_norm": 0.1446671187877655, "learning_rate": 2.9994754683764e-05, "loss": 0.0073, "step": 25980 }, { "grad_norm": 0.1514895111322403, "learning_rate": 2.9956877664676754e-05, "loss": 0.0066, "step": 25990 }, { "grad_norm": 0.1753484606742859, "learning_rate": 2.9919014344874636e-05, "loss": 0.0063, "step": 26000 }, { "grad_norm": 0.20395950973033905, "learning_rate": 2.9881164750236857e-05, "loss": 0.0114, "step": 26010 }, { "grad_norm": 0.143011212348938, "learning_rate": 2.984332890663326e-05, "loss": 0.0071, "step": 26020 }, { "grad_norm": 0.16319026052951813, "learning_rate": 2.9805506839924292e-05, "loss": 0.01, "step": 26030 }, { "grad_norm": 0.1518692821264267, "learning_rate": 2.9767698575960968e-05, "loss": 0.0069, "step": 26040 }, { "grad_norm": 0.12229111045598984, "learning_rate": 2.9729904140584913e-05, "loss": 0.0063, "step": 26050 }, { "grad_norm": 0.16395439207553864, "learning_rate": 2.9692123559628234e-05, "loss": 0.0076, "step": 26060 }, { "grad_norm": 0.1415119469165802, "learning_rate": 2.9654356858913596e-05, "loss": 0.0083, "step": 26070 }, { "grad_norm": 0.1340964436531067, "learning_rate": 2.9616604064254206e-05, "loss": 0.0082, "step": 26080 }, { "grad_norm": 0.20312657952308655, "learning_rate": 2.9578865201453732e-05, "loss": 0.0089, "step": 26090 }, { "grad_norm": 0.15762169659137726, "learning_rate": 2.9541140296306335e-05, "loss": 0.0083, "step": 26100 }, { "grad_norm": 0.10215964168310165, "learning_rate": 2.9503429374596627e-05, "loss": 0.0073, "step": 26110 }, { "grad_norm": 0.10677684843540192, "learning_rate": 2.946573246209967e-05, "loss": 0.0062, "step": 26120 }, { "grad_norm": 0.17978116869926453, "learning_rate": 2.942804958458094e-05, "loss": 0.0076, "step": 26130 }, { "grad_norm": 0.15096570551395416, "learning_rate": 2.9390380767796343e-05, "loss": 0.0072, "step": 26140 }, { "grad_norm": 0.12808433175086975, "learning_rate": 2.9352726037492174e-05, "loss": 0.0094, "step": 26150 }, { "grad_norm": 0.15655720233917236, "learning_rate": 2.9315085419405052e-05, "loss": 0.0067, "step": 26160 }, { "grad_norm": 0.18058565258979797, "learning_rate": 2.927745893926199e-05, "loss": 0.0093, "step": 26170 }, { "grad_norm": 0.14617003500461578, "learning_rate": 2.9239846622780358e-05, "loss": 0.0059, "step": 26180 }, { "grad_norm": 0.17261363565921783, "learning_rate": 2.9202248495667788e-05, "loss": 0.0064, "step": 26190 }, { "grad_norm": 0.1294960230588913, "learning_rate": 2.916466458362227e-05, "loss": 0.0088, "step": 26200 }, { "grad_norm": 0.12282274663448334, "learning_rate": 2.9127094912332033e-05, "loss": 0.0103, "step": 26210 }, { "grad_norm": 0.13551951944828033, "learning_rate": 2.9089539507475606e-05, "loss": 0.0079, "step": 26220 }, { "grad_norm": 0.1531704068183899, "learning_rate": 2.9051998394721748e-05, "loss": 0.0104, "step": 26230 }, { "grad_norm": 0.17266125977039337, "learning_rate": 2.901447159972948e-05, "loss": 0.007, "step": 26240 }, { "grad_norm": 0.1518491506576538, "learning_rate": 2.8976959148148e-05, "loss": 0.0064, "step": 26250 }, { "grad_norm": 0.1149701252579689, "learning_rate": 2.8939461065616674e-05, "loss": 0.0079, "step": 26260 }, { "grad_norm": 0.16099999845027924, "learning_rate": 2.8901977377765127e-05, "loss": 0.0071, "step": 26270 }, { "grad_norm": 0.1294792890548706, "learning_rate": 2.8864508110213094e-05, "loss": 0.0051, "step": 26280 }, { "grad_norm": 0.145633265376091, "learning_rate": 2.8827053288570503e-05, "loss": 0.0065, "step": 26290 }, { "grad_norm": 0.12999427318572998, "learning_rate": 2.8789612938437315e-05, "loss": 0.0084, "step": 26300 }, { "grad_norm": 0.1352524310350418, "learning_rate": 2.8752187085403683e-05, "loss": 0.0063, "step": 26310 }, { "grad_norm": 0.1380235105752945, "learning_rate": 2.8714775755049818e-05, "loss": 0.0081, "step": 26320 }, { "grad_norm": 0.12467517703771591, "learning_rate": 2.867737897294604e-05, "loss": 0.0084, "step": 26330 }, { "grad_norm": 0.1665138453245163, "learning_rate": 2.8639996764652653e-05, "loss": 0.0085, "step": 26340 }, { "grad_norm": 0.15835003554821014, "learning_rate": 2.8602629155720084e-05, "loss": 0.0106, "step": 26350 }, { "grad_norm": 0.18309833109378815, "learning_rate": 2.8565276171688703e-05, "loss": 0.0078, "step": 26360 }, { "grad_norm": 0.14383599162101746, "learning_rate": 2.8527937838088943e-05, "loss": 0.0069, "step": 26370 }, { "grad_norm": 0.1773087978363037, "learning_rate": 2.84906141804412e-05, "loss": 0.0093, "step": 26380 }, { "grad_norm": 0.16376902163028717, "learning_rate": 2.8453305224255867e-05, "loss": 0.0092, "step": 26390 }, { "grad_norm": 0.09986706078052521, "learning_rate": 2.8416010995033216e-05, "loss": 0.0071, "step": 26400 }, { "grad_norm": 0.2097344547510147, "learning_rate": 2.8378731518263524e-05, "loss": 0.0078, "step": 26410 }, { "grad_norm": 0.1497352570295334, "learning_rate": 2.834146681942696e-05, "loss": 0.0082, "step": 26420 }, { "grad_norm": 0.10717622935771942, "learning_rate": 2.8304216923993622e-05, "loss": 0.008, "step": 26430 }, { "grad_norm": 0.14458578824996948, "learning_rate": 2.8266981857423413e-05, "loss": 0.0085, "step": 26440 }, { "grad_norm": 0.12267035245895386, "learning_rate": 2.8229761645166197e-05, "loss": 0.0071, "step": 26450 }, { "grad_norm": 0.17202484607696533, "learning_rate": 2.81925563126616e-05, "loss": 0.0076, "step": 26460 }, { "grad_norm": 0.1659736931324005, "learning_rate": 2.8155365885339124e-05, "loss": 0.0076, "step": 26470 }, { "grad_norm": 0.13343046605587006, "learning_rate": 2.8118190388618093e-05, "loss": 0.0079, "step": 26480 }, { "grad_norm": 0.17349931597709656, "learning_rate": 2.8081029847907607e-05, "loss": 0.0085, "step": 26490 }, { "grad_norm": 0.1411243975162506, "learning_rate": 2.8043884288606525e-05, "loss": 0.0067, "step": 26500 }, { "grad_norm": 0.10938723385334015, "learning_rate": 2.8006753736103496e-05, "loss": 0.0074, "step": 26510 }, { "grad_norm": 0.15188850462436676, "learning_rate": 2.7969638215776918e-05, "loss": 0.0081, "step": 26520 }, { "grad_norm": 0.12997688353061676, "learning_rate": 2.793253775299487e-05, "loss": 0.0079, "step": 26530 }, { "grad_norm": 0.1718772053718567, "learning_rate": 2.7895452373115184e-05, "loss": 0.0094, "step": 26540 }, { "grad_norm": 0.16357283294200897, "learning_rate": 2.785838210148539e-05, "loss": 0.0079, "step": 26550 }, { "grad_norm": 0.12049569189548492, "learning_rate": 2.782132696344263e-05, "loss": 0.0083, "step": 26560 }, { "grad_norm": 0.11089755594730377, "learning_rate": 2.7784286984313745e-05, "loss": 0.0091, "step": 26570 }, { "grad_norm": 0.14696292579174042, "learning_rate": 2.7747262189415236e-05, "loss": 0.0093, "step": 26580 }, { "grad_norm": 0.1389995515346527, "learning_rate": 2.7710252604053205e-05, "loss": 0.0065, "step": 26590 }, { "grad_norm": 0.09691738337278366, "learning_rate": 2.767325825352332e-05, "loss": 0.0068, "step": 26600 }, { "grad_norm": 0.11322275549173355, "learning_rate": 2.7636279163110913e-05, "loss": 0.0092, "step": 26610 }, { "grad_norm": 0.15019719302654266, "learning_rate": 2.7599315358090795e-05, "loss": 0.0091, "step": 26620 }, { "grad_norm": 0.19183951616287231, "learning_rate": 2.7562366863727407e-05, "loss": 0.0069, "step": 26630 }, { "grad_norm": 0.11343201249837875, "learning_rate": 2.7525433705274695e-05, "loss": 0.007, "step": 26640 }, { "grad_norm": 0.08377944678068161, "learning_rate": 2.748851590797614e-05, "loss": 0.0069, "step": 26650 }, { "grad_norm": 0.12627474963665009, "learning_rate": 2.7451613497064675e-05, "loss": 0.0083, "step": 26660 }, { "grad_norm": 0.15875983238220215, "learning_rate": 2.7414726497762765e-05, "loss": 0.0073, "step": 26670 }, { "grad_norm": 0.08690188080072403, "learning_rate": 2.737785493528232e-05, "loss": 0.0061, "step": 26680 }, { "grad_norm": 0.12628929316997528, "learning_rate": 2.7340998834824745e-05, "loss": 0.0069, "step": 26690 }, { "grad_norm": 0.15277786552906036, "learning_rate": 2.7304158221580777e-05, "loss": 0.0076, "step": 26700 }, { "grad_norm": 0.1646767109632492, "learning_rate": 2.7267333120730675e-05, "loss": 0.0082, "step": 26710 }, { "grad_norm": 0.23027439415454865, "learning_rate": 2.7230523557444017e-05, "loss": 0.0076, "step": 26720 }, { "grad_norm": 0.1179300919175148, "learning_rate": 2.7193729556879798e-05, "loss": 0.0065, "step": 26730 }, { "grad_norm": 0.14870844781398773, "learning_rate": 2.715695114418637e-05, "loss": 0.0056, "step": 26740 }, { "grad_norm": 0.15971717238426208, "learning_rate": 2.7120188344501475e-05, "loss": 0.0072, "step": 26750 }, { "grad_norm": 0.09164348989725113, "learning_rate": 2.7083441182952067e-05, "loss": 0.0085, "step": 26760 }, { "grad_norm": 0.15129053592681885, "learning_rate": 2.7046709684654527e-05, "loss": 0.0067, "step": 26770 }, { "grad_norm": 0.1485140025615692, "learning_rate": 2.700999387471448e-05, "loss": 0.01, "step": 26780 }, { "grad_norm": 0.11961715668439865, "learning_rate": 2.6973293778226854e-05, "loss": 0.0054, "step": 26790 }, { "grad_norm": 0.15755122900009155, "learning_rate": 2.6936609420275804e-05, "loss": 0.0082, "step": 26800 }, { "grad_norm": 0.10219099372625351, "learning_rate": 2.689994082593472e-05, "loss": 0.007, "step": 26810 }, { "grad_norm": 0.12256032228469849, "learning_rate": 2.6863288020266264e-05, "loss": 0.0076, "step": 26820 }, { "grad_norm": 0.1774783432483673, "learning_rate": 2.682665102832228e-05, "loss": 0.0089, "step": 26830 }, { "grad_norm": 0.12114390730857849, "learning_rate": 2.67900298751438e-05, "loss": 0.0075, "step": 26840 }, { "grad_norm": 0.1066722497344017, "learning_rate": 2.6753424585761067e-05, "loss": 0.0067, "step": 26850 }, { "grad_norm": 0.13626335561275482, "learning_rate": 2.671683518519341e-05, "loss": 0.0064, "step": 26860 }, { "grad_norm": 0.22602806985378265, "learning_rate": 2.668026169844936e-05, "loss": 0.0086, "step": 26870 }, { "grad_norm": 0.193572536110878, "learning_rate": 2.6643704150526538e-05, "loss": 0.0079, "step": 26880 }, { "grad_norm": 0.13853846490383148, "learning_rate": 2.6607162566411716e-05, "loss": 0.009, "step": 26890 }, { "grad_norm": 0.1205768808722496, "learning_rate": 2.6570636971080697e-05, "loss": 0.0074, "step": 26900 }, { "grad_norm": 0.12979018688201904, "learning_rate": 2.6534127389498364e-05, "loss": 0.0061, "step": 26910 }, { "grad_norm": 0.10542088001966476, "learning_rate": 2.6497633846618696e-05, "loss": 0.008, "step": 26920 }, { "grad_norm": 0.16684691607952118, "learning_rate": 2.6461156367384677e-05, "loss": 0.0071, "step": 26930 }, { "grad_norm": 0.10393866896629333, "learning_rate": 2.6424694976728316e-05, "loss": 0.007, "step": 26940 }, { "grad_norm": 0.19123908877372742, "learning_rate": 2.6388249699570667e-05, "loss": 0.0095, "step": 26950 }, { "grad_norm": 0.1581849902868271, "learning_rate": 2.6351820560821672e-05, "loss": 0.0084, "step": 26960 }, { "grad_norm": 0.1468169242143631, "learning_rate": 2.631540758538034e-05, "loss": 0.0074, "step": 26970 }, { "grad_norm": 0.16935475170612335, "learning_rate": 2.6279010798134597e-05, "loss": 0.0068, "step": 26980 }, { "grad_norm": 0.15238891541957855, "learning_rate": 2.6242630223961305e-05, "loss": 0.008, "step": 26990 }, { "grad_norm": 0.1520589292049408, "learning_rate": 2.6206265887726244e-05, "loss": 0.0066, "step": 27000 }, { "grad_norm": 0.12673696875572205, "learning_rate": 2.6169917814284066e-05, "loss": 0.0082, "step": 27010 }, { "grad_norm": 0.14724650979042053, "learning_rate": 2.6133586028478364e-05, "loss": 0.0074, "step": 27020 }, { "grad_norm": 0.12997213006019592, "learning_rate": 2.609727055514155e-05, "loss": 0.0078, "step": 27030 }, { "grad_norm": 0.15525896847248077, "learning_rate": 2.606097141909494e-05, "loss": 0.0069, "step": 27040 }, { "grad_norm": 0.12690947949886322, "learning_rate": 2.6024688645148644e-05, "loss": 0.0064, "step": 27050 }, { "grad_norm": 0.17244873940944672, "learning_rate": 2.5988422258101564e-05, "loss": 0.0087, "step": 27060 }, { "grad_norm": 0.15211284160614014, "learning_rate": 2.5952172282741453e-05, "loss": 0.0081, "step": 27070 }, { "grad_norm": 0.22306673228740692, "learning_rate": 2.5915938743844853e-05, "loss": 0.0085, "step": 27080 }, { "grad_norm": 0.1610042005777359, "learning_rate": 2.5879721666177003e-05, "loss": 0.0057, "step": 27090 }, { "grad_norm": 0.12102359533309937, "learning_rate": 2.5843521074491972e-05, "loss": 0.0062, "step": 27100 }, { "grad_norm": 0.14063388109207153, "learning_rate": 2.5807336993532487e-05, "loss": 0.0061, "step": 27110 }, { "grad_norm": 0.11418506503105164, "learning_rate": 2.577116944803004e-05, "loss": 0.0071, "step": 27120 }, { "grad_norm": 0.14081886410713196, "learning_rate": 2.5735018462704818e-05, "loss": 0.007, "step": 27130 }, { "grad_norm": 0.11232335865497589, "learning_rate": 2.5698884062265665e-05, "loss": 0.0061, "step": 27140 }, { "grad_norm": 0.19594040513038635, "learning_rate": 2.5662766271410134e-05, "loss": 0.0088, "step": 27150 }, { "grad_norm": 0.12291170656681061, "learning_rate": 2.5626665114824343e-05, "loss": 0.0092, "step": 27160 }, { "grad_norm": 0.15751105546951294, "learning_rate": 2.5590580617183148e-05, "loss": 0.0097, "step": 27170 }, { "grad_norm": 0.09365588426589966, "learning_rate": 2.5554512803149912e-05, "loss": 0.0082, "step": 27180 }, { "grad_norm": 0.1333504021167755, "learning_rate": 2.5518461697376662e-05, "loss": 0.0094, "step": 27190 }, { "grad_norm": 0.1446709930896759, "learning_rate": 2.548242732450402e-05, "loss": 0.0066, "step": 27200 }, { "grad_norm": 0.12730516493320465, "learning_rate": 2.5446409709161095e-05, "loss": 0.0071, "step": 27210 }, { "grad_norm": 0.13860462605953217, "learning_rate": 2.541040887596561e-05, "loss": 0.0058, "step": 27220 }, { "grad_norm": 0.15814316272735596, "learning_rate": 2.537442484952378e-05, "loss": 0.0063, "step": 27230 }, { "grad_norm": 0.13788916170597076, "learning_rate": 2.533845765443037e-05, "loss": 0.0065, "step": 27240 }, { "grad_norm": 0.13341425359249115, "learning_rate": 2.530250731526863e-05, "loss": 0.0073, "step": 27250 }, { "grad_norm": 0.13996562361717224, "learning_rate": 2.5266573856610253e-05, "loss": 0.0077, "step": 27260 }, { "grad_norm": 0.12856422364711761, "learning_rate": 2.5230657303015403e-05, "loss": 0.0067, "step": 27270 }, { "grad_norm": 0.09313631057739258, "learning_rate": 2.5194757679032728e-05, "loss": 0.0067, "step": 27280 }, { "grad_norm": 0.11902915686368942, "learning_rate": 2.5158875009199278e-05, "loss": 0.0072, "step": 27290 }, { "grad_norm": 0.14328430593013763, "learning_rate": 2.5123009318040537e-05, "loss": 0.01, "step": 27300 }, { "grad_norm": 0.09830839186906815, "learning_rate": 2.508716063007034e-05, "loss": 0.0079, "step": 27310 }, { "grad_norm": 0.1514248102903366, "learning_rate": 2.5051328969790934e-05, "loss": 0.0065, "step": 27320 }, { "grad_norm": 0.09868398308753967, "learning_rate": 2.501551436169292e-05, "loss": 0.0058, "step": 27330 }, { "grad_norm": 0.10593242943286896, "learning_rate": 2.4979716830255255e-05, "loss": 0.0063, "step": 27340 }, { "grad_norm": 0.16287602484226227, "learning_rate": 2.4943936399945233e-05, "loss": 0.0063, "step": 27350 }, { "grad_norm": 0.1289108544588089, "learning_rate": 2.4908173095218412e-05, "loss": 0.006, "step": 27360 }, { "grad_norm": 0.13860870897769928, "learning_rate": 2.4872426940518663e-05, "loss": 0.0082, "step": 27370 }, { "grad_norm": 0.16843174397945404, "learning_rate": 2.4836697960278156e-05, "loss": 0.0082, "step": 27380 }, { "grad_norm": 0.12652814388275146, "learning_rate": 2.480098617891732e-05, "loss": 0.0071, "step": 27390 }, { "grad_norm": 0.1211424246430397, "learning_rate": 2.4765291620844837e-05, "loss": 0.0063, "step": 27400 }, { "grad_norm": 0.14776793122291565, "learning_rate": 2.472961431045756e-05, "loss": 0.0073, "step": 27410 }, { "grad_norm": 0.1259397268295288, "learning_rate": 2.4693954272140622e-05, "loss": 0.0072, "step": 27420 }, { "grad_norm": 0.14144489169120789, "learning_rate": 2.4658311530267315e-05, "loss": 0.0088, "step": 27430 }, { "grad_norm": 0.22458809614181519, "learning_rate": 2.4622686109199124e-05, "loss": 0.0075, "step": 27440 }, { "grad_norm": 0.19313476979732513, "learning_rate": 2.4587078033285695e-05, "loss": 0.0093, "step": 27450 }, { "grad_norm": 0.14744263887405396, "learning_rate": 2.45514873268648e-05, "loss": 0.0064, "step": 27460 }, { "grad_norm": 0.1446637511253357, "learning_rate": 2.4515914014262336e-05, "loss": 0.008, "step": 27470 }, { "grad_norm": 0.1510876566171646, "learning_rate": 2.4480358119792345e-05, "loss": 0.0078, "step": 27480 }, { "grad_norm": 0.11728136241436005, "learning_rate": 2.4444819667756942e-05, "loss": 0.0061, "step": 27490 }, { "grad_norm": 0.12760253250598907, "learning_rate": 2.4409298682446346e-05, "loss": 0.0077, "step": 27500 }, { "grad_norm": 0.11963516473770142, "learning_rate": 2.437379518813877e-05, "loss": 0.0064, "step": 27510 }, { "grad_norm": 0.17074169218540192, "learning_rate": 2.4338309209100547e-05, "loss": 0.0071, "step": 27520 }, { "grad_norm": 0.15698575973510742, "learning_rate": 2.4302840769586004e-05, "loss": 0.0054, "step": 27530 }, { "grad_norm": 0.17687779664993286, "learning_rate": 2.42673898938375e-05, "loss": 0.0071, "step": 27540 }, { "grad_norm": 0.120387002825737, "learning_rate": 2.4231956606085343e-05, "loss": 0.0071, "step": 27550 }, { "grad_norm": 0.13495662808418274, "learning_rate": 2.419654093054789e-05, "loss": 0.0059, "step": 27560 }, { "grad_norm": 0.1421823799610138, "learning_rate": 2.4161142891431375e-05, "loss": 0.0079, "step": 27570 }, { "grad_norm": 0.2050546258687973, "learning_rate": 2.412576251293005e-05, "loss": 0.0074, "step": 27580 }, { "grad_norm": 0.1197027936577797, "learning_rate": 2.4090399819226068e-05, "loss": 0.0088, "step": 27590 }, { "grad_norm": 0.13720612227916718, "learning_rate": 2.4055054834489514e-05, "loss": 0.0065, "step": 27600 }, { "grad_norm": 0.07994584739208221, "learning_rate": 2.401972758287832e-05, "loss": 0.0082, "step": 27610 }, { "grad_norm": 0.18150989711284637, "learning_rate": 2.398441808853834e-05, "loss": 0.0104, "step": 27620 }, { "grad_norm": 0.17763429880142212, "learning_rate": 2.3949126375603288e-05, "loss": 0.0087, "step": 27630 }, { "grad_norm": 0.22455428540706635, "learning_rate": 2.3913852468194724e-05, "loss": 0.0068, "step": 27640 }, { "grad_norm": 0.12060553580522537, "learning_rate": 2.387859639042201e-05, "loss": 0.0071, "step": 27650 }, { "grad_norm": 0.14243461191654205, "learning_rate": 2.3843358166382368e-05, "loss": 0.0065, "step": 27660 }, { "grad_norm": 0.1966603696346283, "learning_rate": 2.3808137820160757e-05, "loss": 0.0062, "step": 27670 }, { "grad_norm": 0.19788949191570282, "learning_rate": 2.3772935375829975e-05, "loss": 0.006, "step": 27680 }, { "grad_norm": 0.18057449162006378, "learning_rate": 2.3737750857450553e-05, "loss": 0.0079, "step": 27690 }, { "grad_norm": 0.13662752509117126, "learning_rate": 2.3702584289070805e-05, "loss": 0.0082, "step": 27700 }, { "grad_norm": 0.17580145597457886, "learning_rate": 2.36674356947267e-05, "loss": 0.0063, "step": 27710 }, { "grad_norm": 0.15738317370414734, "learning_rate": 2.3632305098442004e-05, "loss": 0.0051, "step": 27720 }, { "grad_norm": 0.13928933441638947, "learning_rate": 2.3597192524228156e-05, "loss": 0.0056, "step": 27730 }, { "grad_norm": 0.1445755660533905, "learning_rate": 2.356209799608424e-05, "loss": 0.0065, "step": 27740 }, { "grad_norm": 0.11152718216180801, "learning_rate": 2.352702153799704e-05, "loss": 0.0078, "step": 27750 }, { "grad_norm": 0.13815364241600037, "learning_rate": 2.3491963173941018e-05, "loss": 0.0065, "step": 27760 }, { "grad_norm": 0.12941738963127136, "learning_rate": 2.3456922927878196e-05, "loss": 0.009, "step": 27770 }, { "grad_norm": 0.1120859682559967, "learning_rate": 2.3421900823758257e-05, "loss": 0.0054, "step": 27780 }, { "grad_norm": 0.10906805098056793, "learning_rate": 2.3386896885518496e-05, "loss": 0.0074, "step": 27790 }, { "grad_norm": 0.19087918102741241, "learning_rate": 2.335191113708378e-05, "loss": 0.0071, "step": 27800 }, { "grad_norm": 0.1750304102897644, "learning_rate": 2.331694360236651e-05, "loss": 0.0076, "step": 27810 }, { "grad_norm": 0.18799926340579987, "learning_rate": 2.3281994305266702e-05, "loss": 0.0083, "step": 27820 }, { "grad_norm": 0.08921878039836884, "learning_rate": 2.3247063269671826e-05, "loss": 0.006, "step": 27830 }, { "grad_norm": 0.10225962847471237, "learning_rate": 2.321215051945695e-05, "loss": 0.0057, "step": 27840 }, { "grad_norm": 0.13991224765777588, "learning_rate": 2.3177256078484588e-05, "loss": 0.0064, "step": 27850 }, { "grad_norm": 0.12454655021429062, "learning_rate": 2.3142379970604798e-05, "loss": 0.0069, "step": 27860 }, { "grad_norm": 0.1797337532043457, "learning_rate": 2.3107522219655025e-05, "loss": 0.0078, "step": 27870 }, { "grad_norm": 0.10656938701868057, "learning_rate": 2.3072682849460236e-05, "loss": 0.0075, "step": 27880 }, { "grad_norm": 0.11477630585432053, "learning_rate": 2.303786188383281e-05, "loss": 0.0083, "step": 27890 }, { "grad_norm": 0.14427627623081207, "learning_rate": 2.300305934657257e-05, "loss": 0.0071, "step": 27900 }, { "grad_norm": 0.16111087799072266, "learning_rate": 2.2968275261466677e-05, "loss": 0.0095, "step": 27910 }, { "grad_norm": 0.18116077780723572, "learning_rate": 2.293350965228977e-05, "loss": 0.0058, "step": 27920 }, { "grad_norm": 0.11855107545852661, "learning_rate": 2.2898762542803776e-05, "loss": 0.0081, "step": 27930 }, { "grad_norm": 0.1620890349149704, "learning_rate": 2.286403395675803e-05, "loss": 0.0063, "step": 27940 }, { "grad_norm": 0.16216430068016052, "learning_rate": 2.28293239178892e-05, "loss": 0.0092, "step": 27950 }, { "grad_norm": 0.1683921068906784, "learning_rate": 2.2794632449921287e-05, "loss": 0.0062, "step": 27960 }, { "grad_norm": 0.1095983237028122, "learning_rate": 2.275995957656555e-05, "loss": 0.0078, "step": 27970 }, { "grad_norm": 0.1619522124528885, "learning_rate": 2.272530532152058e-05, "loss": 0.0065, "step": 27980 }, { "grad_norm": 0.14135414361953735, "learning_rate": 2.2690669708472233e-05, "loss": 0.008, "step": 27990 }, { "grad_norm": 0.16348187625408173, "learning_rate": 2.2656052761093655e-05, "loss": 0.0064, "step": 28000 }, { "grad_norm": 0.12772002816200256, "learning_rate": 2.262145450304517e-05, "loss": 0.0077, "step": 28010 }, { "grad_norm": 0.14168108999729156, "learning_rate": 2.2586874957974352e-05, "loss": 0.0062, "step": 28020 }, { "grad_norm": 0.18816840648651123, "learning_rate": 2.2552314149516012e-05, "loss": 0.0071, "step": 28030 }, { "grad_norm": 0.08397357910871506, "learning_rate": 2.2517772101292133e-05, "loss": 0.0065, "step": 28040 }, { "grad_norm": 0.14248625934123993, "learning_rate": 2.248324883691188e-05, "loss": 0.0095, "step": 28050 }, { "grad_norm": 0.16543562710285187, "learning_rate": 2.24487443799716e-05, "loss": 0.0059, "step": 28060 }, { "grad_norm": 0.16896271705627441, "learning_rate": 2.241425875405472e-05, "loss": 0.0063, "step": 28070 }, { "grad_norm": 0.16215600073337555, "learning_rate": 2.2379791982731868e-05, "loss": 0.006, "step": 28080 }, { "grad_norm": 0.11228254437446594, "learning_rate": 2.2345344089560756e-05, "loss": 0.0047, "step": 28090 }, { "grad_norm": 0.14409536123275757, "learning_rate": 2.2310915098086206e-05, "loss": 0.0083, "step": 28100 }, { "grad_norm": 0.16392533481121063, "learning_rate": 2.227650503184009e-05, "loss": 0.0059, "step": 28110 }, { "grad_norm": 0.13319462537765503, "learning_rate": 2.2242113914341357e-05, "loss": 0.0061, "step": 28120 }, { "grad_norm": 0.11986992508172989, "learning_rate": 2.220774176909602e-05, "loss": 0.0061, "step": 28130 }, { "grad_norm": 0.1578139066696167, "learning_rate": 2.2173388619597114e-05, "loss": 0.0069, "step": 28140 }, { "grad_norm": 0.09773050993680954, "learning_rate": 2.21390544893247e-05, "loss": 0.0047, "step": 28150 }, { "grad_norm": 0.07401615381240845, "learning_rate": 2.210473940174585e-05, "loss": 0.0069, "step": 28160 }, { "grad_norm": 0.14858345687389374, "learning_rate": 2.207044338031456e-05, "loss": 0.0076, "step": 28170 }, { "grad_norm": 0.10424849390983582, "learning_rate": 2.203616644847186e-05, "loss": 0.0055, "step": 28180 }, { "grad_norm": 0.11156938970088959, "learning_rate": 2.200190862964571e-05, "loss": 0.008, "step": 28190 }, { "grad_norm": 0.08860824257135391, "learning_rate": 2.1967669947251024e-05, "loss": 0.0058, "step": 28200 }, { "grad_norm": 0.08248220384120941, "learning_rate": 2.1933450424689583e-05, "loss": 0.007, "step": 28210 }, { "grad_norm": 0.1294381469488144, "learning_rate": 2.1899250085350142e-05, "loss": 0.0071, "step": 28220 }, { "grad_norm": 0.1269058883190155, "learning_rate": 2.1865068952608277e-05, "loss": 0.006, "step": 28230 }, { "grad_norm": 0.1430111527442932, "learning_rate": 2.1830907049826487e-05, "loss": 0.0071, "step": 28240 }, { "grad_norm": 0.14895103871822357, "learning_rate": 2.179676440035411e-05, "loss": 0.0065, "step": 28250 }, { "grad_norm": 0.142942875623703, "learning_rate": 2.1762641027527337e-05, "loss": 0.0073, "step": 28260 }, { "grad_norm": 0.11262538284063339, "learning_rate": 2.1728536954669143e-05, "loss": 0.0051, "step": 28270 }, { "grad_norm": 0.10071741789579391, "learning_rate": 2.169445220508936e-05, "loss": 0.0066, "step": 28280 }, { "grad_norm": 0.12885308265686035, "learning_rate": 2.166038680208461e-05, "loss": 0.006, "step": 28290 }, { "grad_norm": 0.14690975844860077, "learning_rate": 2.162634076893823e-05, "loss": 0.0066, "step": 28300 }, { "grad_norm": 0.08385677635669708, "learning_rate": 2.1592314128920388e-05, "loss": 0.0077, "step": 28310 }, { "grad_norm": 0.13699769973754883, "learning_rate": 2.155830690528799e-05, "loss": 0.0083, "step": 28320 }, { "grad_norm": 0.09747090935707092, "learning_rate": 2.1524319121284613e-05, "loss": 0.0054, "step": 28330 }, { "grad_norm": 0.13318748772144318, "learning_rate": 2.1490350800140607e-05, "loss": 0.0055, "step": 28340 }, { "grad_norm": 0.12750586867332458, "learning_rate": 2.1456401965073002e-05, "loss": 0.0053, "step": 28350 }, { "grad_norm": 0.11928776651620865, "learning_rate": 2.1422472639285524e-05, "loss": 0.0049, "step": 28360 }, { "grad_norm": 0.1091371476650238, "learning_rate": 2.13885628459685e-05, "loss": 0.0053, "step": 28370 }, { "grad_norm": 0.11134473234415054, "learning_rate": 2.135467260829901e-05, "loss": 0.0064, "step": 28380 }, { "grad_norm": 0.20268698036670685, "learning_rate": 2.1320801949440654e-05, "loss": 0.007, "step": 28390 }, { "grad_norm": 0.14401227235794067, "learning_rate": 2.1286950892543744e-05, "loss": 0.0053, "step": 28400 }, { "grad_norm": 0.20482487976551056, "learning_rate": 2.125311946074515e-05, "loss": 0.0097, "step": 28410 }, { "grad_norm": 0.1432790607213974, "learning_rate": 2.1219307677168355e-05, "loss": 0.006, "step": 28420 }, { "grad_norm": 0.1675974428653717, "learning_rate": 2.118551556492336e-05, "loss": 0.0074, "step": 28430 }, { "grad_norm": 0.0972568467259407, "learning_rate": 2.1151743147106774e-05, "loss": 0.0064, "step": 28440 }, { "grad_norm": 0.17925412952899933, "learning_rate": 2.111799044680172e-05, "loss": 0.0072, "step": 28450 }, { "grad_norm": 0.21059511601924896, "learning_rate": 2.1084257487077873e-05, "loss": 0.0071, "step": 28460 }, { "grad_norm": 0.1325257420539856, "learning_rate": 2.1050544290991357e-05, "loss": 0.0078, "step": 28470 }, { "grad_norm": 0.18641379475593567, "learning_rate": 2.101685088158486e-05, "loss": 0.0074, "step": 28480 }, { "grad_norm": 0.0935942530632019, "learning_rate": 2.0983177281887472e-05, "loss": 0.0057, "step": 28490 }, { "grad_norm": 0.14647889137268066, "learning_rate": 2.0949523514914798e-05, "loss": 0.0064, "step": 28500 }, { "grad_norm": 0.146108940243721, "learning_rate": 2.0915889603668876e-05, "loss": 0.0052, "step": 28510 }, { "grad_norm": 0.15425634384155273, "learning_rate": 2.0882275571138175e-05, "loss": 0.0055, "step": 28520 }, { "grad_norm": 0.10658007115125656, "learning_rate": 2.0848681440297545e-05, "loss": 0.0072, "step": 28530 }, { "grad_norm": 0.16401641070842743, "learning_rate": 2.081510723410827e-05, "loss": 0.0069, "step": 28540 }, { "grad_norm": 0.14512862265110016, "learning_rate": 2.0781552975518003e-05, "loss": 0.0072, "step": 28550 }, { "grad_norm": 0.15989923477172852, "learning_rate": 2.074801868746078e-05, "loss": 0.0073, "step": 28560 }, { "grad_norm": 0.13685506582260132, "learning_rate": 2.0714504392856955e-05, "loss": 0.0069, "step": 28570 }, { "grad_norm": 0.12135326862335205, "learning_rate": 2.0681010114613215e-05, "loss": 0.006, "step": 28580 }, { "grad_norm": 0.14720332622528076, "learning_rate": 2.0647535875622597e-05, "loss": 0.0061, "step": 28590 }, { "grad_norm": 0.12961378693580627, "learning_rate": 2.0614081698764432e-05, "loss": 0.0056, "step": 28600 }, { "grad_norm": 0.1848914623260498, "learning_rate": 2.0580647606904334e-05, "loss": 0.0063, "step": 28610 }, { "grad_norm": 0.1159122884273529, "learning_rate": 2.0547233622894208e-05, "loss": 0.0063, "step": 28620 }, { "grad_norm": 0.18063253164291382, "learning_rate": 2.0513839769572157e-05, "loss": 0.0089, "step": 28630 }, { "grad_norm": 0.1435202807188034, "learning_rate": 2.0480466069762584e-05, "loss": 0.0059, "step": 28640 }, { "grad_norm": 0.16253963112831116, "learning_rate": 2.0447112546276104e-05, "loss": 0.007, "step": 28650 }, { "grad_norm": 0.11024479568004608, "learning_rate": 2.0413779221909547e-05, "loss": 0.0053, "step": 28660 }, { "grad_norm": 0.07309540361166, "learning_rate": 2.0380466119445912e-05, "loss": 0.005, "step": 28670 }, { "grad_norm": 0.18523390591144562, "learning_rate": 2.0347173261654373e-05, "loss": 0.0071, "step": 28680 }, { "grad_norm": 0.13684837520122528, "learning_rate": 2.03139006712903e-05, "loss": 0.0064, "step": 28690 }, { "grad_norm": 0.12702544033527374, "learning_rate": 2.028064837109519e-05, "loss": 0.0073, "step": 28700 }, { "grad_norm": 0.1360674500465393, "learning_rate": 2.0247416383796685e-05, "loss": 0.0065, "step": 28710 }, { "grad_norm": 0.13239669799804688, "learning_rate": 2.0214204732108548e-05, "loss": 0.0049, "step": 28720 }, { "grad_norm": 0.10178260505199432, "learning_rate": 2.0181013438730596e-05, "loss": 0.0065, "step": 28730 }, { "grad_norm": 0.11083299666643143, "learning_rate": 2.0147842526348783e-05, "loss": 0.0063, "step": 28740 }, { "grad_norm": 0.138462632894516, "learning_rate": 2.011469201763511e-05, "loss": 0.0076, "step": 28750 }, { "grad_norm": 0.12616848945617676, "learning_rate": 2.0081561935247665e-05, "loss": 0.0058, "step": 28760 }, { "grad_norm": 0.14320510625839233, "learning_rate": 2.0048452301830523e-05, "loss": 0.0071, "step": 28770 }, { "grad_norm": 0.153036966919899, "learning_rate": 2.0015363140013788e-05, "loss": 0.0072, "step": 28780 }, { "grad_norm": 0.13205987215042114, "learning_rate": 1.9982294472413606e-05, "loss": 0.0056, "step": 28790 }, { "grad_norm": 0.11674950271844864, "learning_rate": 1.9949246321632103e-05, "loss": 0.0057, "step": 28800 }, { "grad_norm": 0.14822660386562347, "learning_rate": 1.9916218710257377e-05, "loss": 0.0059, "step": 28810 }, { "grad_norm": 0.15033309161663055, "learning_rate": 1.988321166086351e-05, "loss": 0.0083, "step": 28820 }, { "grad_norm": 0.09904676675796509, "learning_rate": 1.9850225196010468e-05, "loss": 0.0056, "step": 28830 }, { "grad_norm": 0.14472198486328125, "learning_rate": 1.981725933824421e-05, "loss": 0.0064, "step": 28840 }, { "grad_norm": 0.16641177237033844, "learning_rate": 1.978431411009661e-05, "loss": 0.0058, "step": 28850 }, { "grad_norm": 0.07146923989057541, "learning_rate": 1.9751389534085375e-05, "loss": 0.0055, "step": 28860 }, { "grad_norm": 0.0797407403588295, "learning_rate": 1.9718485632714184e-05, "loss": 0.0059, "step": 28870 }, { "grad_norm": 0.1104724109172821, "learning_rate": 1.968560242847251e-05, "loss": 0.0066, "step": 28880 }, { "grad_norm": 0.11099220812320709, "learning_rate": 1.965273994383573e-05, "loss": 0.0061, "step": 28890 }, { "grad_norm": 0.12381775677204132, "learning_rate": 1.961989820126504e-05, "loss": 0.0065, "step": 28900 }, { "grad_norm": 0.1572398990392685, "learning_rate": 1.958707722320746e-05, "loss": 0.006, "step": 28910 }, { "grad_norm": 0.11231724917888641, "learning_rate": 1.955427703209584e-05, "loss": 0.0049, "step": 28920 }, { "grad_norm": 0.15277545154094696, "learning_rate": 1.9521497650348764e-05, "loss": 0.0088, "step": 28930 }, { "grad_norm": 0.0811394602060318, "learning_rate": 1.948873910037067e-05, "loss": 0.0052, "step": 28940 }, { "grad_norm": 0.1741042286157608, "learning_rate": 1.9456001404551678e-05, "loss": 0.0061, "step": 28950 }, { "grad_norm": 0.12892012298107147, "learning_rate": 1.942328458526771e-05, "loss": 0.0085, "step": 28960 }, { "grad_norm": 0.14054277539253235, "learning_rate": 1.9390588664880427e-05, "loss": 0.0067, "step": 28970 }, { "grad_norm": 0.10928968340158463, "learning_rate": 1.9357913665737145e-05, "loss": 0.0065, "step": 28980 }, { "grad_norm": 0.17732040584087372, "learning_rate": 1.932525961017093e-05, "loss": 0.0075, "step": 28990 }, { "grad_norm": 0.15083461999893188, "learning_rate": 1.9292626520500533e-05, "loss": 0.007, "step": 29000 }, { "grad_norm": 0.14603601396083832, "learning_rate": 1.9260014419030354e-05, "loss": 0.0062, "step": 29010 }, { "grad_norm": 0.10582101345062256, "learning_rate": 1.9227423328050475e-05, "loss": 0.0064, "step": 29020 }, { "grad_norm": 0.1370336413383484, "learning_rate": 1.9194853269836582e-05, "loss": 0.0056, "step": 29030 }, { "grad_norm": 0.12118593603372574, "learning_rate": 1.916230426664999e-05, "loss": 0.0088, "step": 29040 }, { "grad_norm": 0.09654794633388519, "learning_rate": 1.912977634073765e-05, "loss": 0.0043, "step": 29050 }, { "grad_norm": 0.11005935817956924, "learning_rate": 1.9097269514332083e-05, "loss": 0.0061, "step": 29060 }, { "grad_norm": 0.17882677912712097, "learning_rate": 1.9064783809651433e-05, "loss": 0.0067, "step": 29070 }, { "grad_norm": 0.13912445306777954, "learning_rate": 1.9032319248899333e-05, "loss": 0.007, "step": 29080 }, { "grad_norm": 0.20840422809123993, "learning_rate": 1.8999875854265015e-05, "loss": 0.0068, "step": 29090 }, { "grad_norm": 0.14682365953922272, "learning_rate": 1.8967453647923232e-05, "loss": 0.0067, "step": 29100 }, { "grad_norm": 0.1940857470035553, "learning_rate": 1.893505265203427e-05, "loss": 0.0055, "step": 29110 }, { "grad_norm": 0.18322297930717468, "learning_rate": 1.8902672888743907e-05, "loss": 0.0067, "step": 29120 }, { "grad_norm": 0.11656998842954636, "learning_rate": 1.8870314380183396e-05, "loss": 0.0067, "step": 29130 }, { "grad_norm": 0.10188497602939606, "learning_rate": 1.8837977148469448e-05, "loss": 0.008, "step": 29140 }, { "grad_norm": 0.15629106760025024, "learning_rate": 1.880566121570429e-05, "loss": 0.0073, "step": 29150 }, { "grad_norm": 0.19168686866760254, "learning_rate": 1.877336660397554e-05, "loss": 0.0068, "step": 29160 }, { "grad_norm": 0.09402637928724289, "learning_rate": 1.874109333535628e-05, "loss": 0.0066, "step": 29170 }, { "grad_norm": 0.12461306899785995, "learning_rate": 1.870884143190496e-05, "loss": 0.0072, "step": 29180 }, { "grad_norm": 0.1064060628414154, "learning_rate": 1.867661091566546e-05, "loss": 0.0061, "step": 29190 }, { "grad_norm": 0.08527438342571259, "learning_rate": 1.864440180866704e-05, "loss": 0.0051, "step": 29200 }, { "grad_norm": 0.1426096111536026, "learning_rate": 1.8612214132924317e-05, "loss": 0.0061, "step": 29210 }, { "grad_norm": 0.11739985644817352, "learning_rate": 1.858004791043728e-05, "loss": 0.007, "step": 29220 }, { "grad_norm": 0.17511652410030365, "learning_rate": 1.854790316319123e-05, "loss": 0.0065, "step": 29230 }, { "grad_norm": 0.16174334287643433, "learning_rate": 1.8515779913156766e-05, "loss": 0.0085, "step": 29240 }, { "grad_norm": 0.12870438396930695, "learning_rate": 1.848367818228986e-05, "loss": 0.0067, "step": 29250 }, { "grad_norm": 0.09646489471197128, "learning_rate": 1.8451597992531733e-05, "loss": 0.0054, "step": 29260 }, { "grad_norm": 0.0843532383441925, "learning_rate": 1.8419539365808914e-05, "loss": 0.0057, "step": 29270 }, { "grad_norm": 0.12195718288421631, "learning_rate": 1.838750232403313e-05, "loss": 0.006, "step": 29280 }, { "grad_norm": 0.12812437117099762, "learning_rate": 1.835548688910142e-05, "loss": 0.007, "step": 29290 }, { "grad_norm": 0.1412561684846878, "learning_rate": 1.8323493082896037e-05, "loss": 0.0077, "step": 29300 }, { "grad_norm": 0.1281844824552536, "learning_rate": 1.8291520927284454e-05, "loss": 0.0056, "step": 29310 }, { "grad_norm": 0.12863405048847198, "learning_rate": 1.8259570444119305e-05, "loss": 0.0059, "step": 29320 }, { "grad_norm": 0.11471179872751236, "learning_rate": 1.8227641655238488e-05, "loss": 0.0068, "step": 29330 }, { "grad_norm": 0.0727989450097084, "learning_rate": 1.819573458246498e-05, "loss": 0.005, "step": 29340 }, { "grad_norm": 0.14026591181755066, "learning_rate": 1.816384924760699e-05, "loss": 0.0067, "step": 29350 }, { "grad_norm": 0.15314441919326782, "learning_rate": 1.813198567245784e-05, "loss": 0.0067, "step": 29360 }, { "grad_norm": 0.12418721616268158, "learning_rate": 1.8100143878796006e-05, "loss": 0.0061, "step": 29370 }, { "grad_norm": 0.1616864800453186, "learning_rate": 1.8068323888385015e-05, "loss": 0.0054, "step": 29380 }, { "grad_norm": 0.1168210580945015, "learning_rate": 1.803652572297355e-05, "loss": 0.0058, "step": 29390 }, { "grad_norm": 0.13153989613056183, "learning_rate": 1.8004749404295353e-05, "loss": 0.0055, "step": 29400 }, { "grad_norm": 0.13995331525802612, "learning_rate": 1.797299495406926e-05, "loss": 0.0052, "step": 29410 }, { "grad_norm": 0.13381081819534302, "learning_rate": 1.7941262393999103e-05, "loss": 0.005, "step": 29420 }, { "grad_norm": 0.15105411410331726, "learning_rate": 1.7909551745773816e-05, "loss": 0.0077, "step": 29430 }, { "grad_norm": 0.11873067170381546, "learning_rate": 1.7877863031067304e-05, "loss": 0.0065, "step": 29440 }, { "grad_norm": 0.13266591727733612, "learning_rate": 1.7846196271538516e-05, "loss": 0.0063, "step": 29450 }, { "grad_norm": 0.12659795582294464, "learning_rate": 1.7814551488831384e-05, "loss": 0.0062, "step": 29460 }, { "grad_norm": 0.11817212402820587, "learning_rate": 1.7782928704574835e-05, "loss": 0.0045, "step": 29470 }, { "grad_norm": 0.19131849706172943, "learning_rate": 1.775132794038271e-05, "loss": 0.007, "step": 29480 }, { "grad_norm": 0.12139537930488586, "learning_rate": 1.7719749217853855e-05, "loss": 0.0048, "step": 29490 }, { "grad_norm": 0.13702994585037231, "learning_rate": 1.7688192558572038e-05, "loss": 0.0064, "step": 29500 }, { "grad_norm": 0.09550647437572479, "learning_rate": 1.7656657984105906e-05, "loss": 0.0069, "step": 29510 }, { "grad_norm": 0.2139042317867279, "learning_rate": 1.7625145516009068e-05, "loss": 0.0068, "step": 29520 }, { "grad_norm": 0.1607227772474289, "learning_rate": 1.7593655175820005e-05, "loss": 0.0047, "step": 29530 }, { "grad_norm": 0.1276678442955017, "learning_rate": 1.7562186985062046e-05, "loss": 0.0085, "step": 29540 }, { "grad_norm": 0.09901469945907593, "learning_rate": 1.7530740965243403e-05, "loss": 0.0058, "step": 29550 }, { "grad_norm": 0.1663932204246521, "learning_rate": 1.7499317137857153e-05, "loss": 0.0073, "step": 29560 }, { "grad_norm": 0.1282488852739334, "learning_rate": 1.7467915524381184e-05, "loss": 0.006, "step": 29570 }, { "grad_norm": 0.08631231635808945, "learning_rate": 1.7436536146278182e-05, "loss": 0.0055, "step": 29580 }, { "grad_norm": 0.12432780861854553, "learning_rate": 1.7405179024995688e-05, "loss": 0.0055, "step": 29590 }, { "grad_norm": 0.18494825065135956, "learning_rate": 1.737384418196596e-05, "loss": 0.0072, "step": 29600 }, { "grad_norm": 0.22440285980701447, "learning_rate": 1.734253163860609e-05, "loss": 0.0081, "step": 29610 }, { "grad_norm": 0.16571803390979767, "learning_rate": 1.7311241416317896e-05, "loss": 0.0077, "step": 29620 }, { "grad_norm": 0.10627832263708115, "learning_rate": 1.7279973536487982e-05, "loss": 0.0051, "step": 29630 }, { "grad_norm": 0.13515885174274445, "learning_rate": 1.724872802048761e-05, "loss": 0.0054, "step": 29640 }, { "grad_norm": 0.0935160294175148, "learning_rate": 1.7217504889672803e-05, "loss": 0.0058, "step": 29650 }, { "grad_norm": 0.1547461748123169, "learning_rate": 1.7186304165384287e-05, "loss": 0.0054, "step": 29660 }, { "grad_norm": 0.09525209665298462, "learning_rate": 1.7155125868947475e-05, "loss": 0.0047, "step": 29670 }, { "grad_norm": 0.12091971188783646, "learning_rate": 1.7123970021672404e-05, "loss": 0.0047, "step": 29680 }, { "grad_norm": 0.11565007269382477, "learning_rate": 1.709283664485384e-05, "loss": 0.0057, "step": 29690 }, { "grad_norm": 0.14164480566978455, "learning_rate": 1.7061725759771113e-05, "loss": 0.0061, "step": 29700 }, { "grad_norm": 0.13164623081684113, "learning_rate": 1.7030637387688248e-05, "loss": 0.0071, "step": 29710 }, { "grad_norm": 0.13755781948566437, "learning_rate": 1.6999571549853836e-05, "loss": 0.0063, "step": 29720 }, { "grad_norm": 0.14421577751636505, "learning_rate": 1.696852826750112e-05, "loss": 0.0095, "step": 29730 }, { "grad_norm": 0.12077150493860245, "learning_rate": 1.6937507561847844e-05, "loss": 0.0054, "step": 29740 }, { "grad_norm": 0.13537010550498962, "learning_rate": 1.6906509454096385e-05, "loss": 0.0072, "step": 29750 }, { "grad_norm": 0.15828560292720795, "learning_rate": 1.687553396543367e-05, "loss": 0.0058, "step": 29760 }, { "grad_norm": 0.18641845881938934, "learning_rate": 1.6844581117031154e-05, "loss": 0.0087, "step": 29770 }, { "grad_norm": 0.09605365246534348, "learning_rate": 1.681365093004481e-05, "loss": 0.0057, "step": 29780 }, { "grad_norm": 0.1445549577474594, "learning_rate": 1.678274342561511e-05, "loss": 0.0064, "step": 29790 }, { "grad_norm": 0.2047271430492401, "learning_rate": 1.675185862486706e-05, "loss": 0.0095, "step": 29800 }, { "grad_norm": 0.13187849521636963, "learning_rate": 1.6720996548910127e-05, "loss": 0.0046, "step": 29810 }, { "grad_norm": 0.12203526496887207, "learning_rate": 1.6690157218838247e-05, "loss": 0.0057, "step": 29820 }, { "grad_norm": 0.1675829142332077, "learning_rate": 1.665934065572984e-05, "loss": 0.007, "step": 29830 }, { "grad_norm": 0.2303021401166916, "learning_rate": 1.6628546880647688e-05, "loss": 0.0061, "step": 29840 }, { "grad_norm": 0.19502632319927216, "learning_rate": 1.6597775914639076e-05, "loss": 0.0073, "step": 29850 }, { "grad_norm": 0.17047223448753357, "learning_rate": 1.6567027778735654e-05, "loss": 0.0074, "step": 29860 }, { "grad_norm": 0.1996438205242157, "learning_rate": 1.653630249395351e-05, "loss": 0.0064, "step": 29870 }, { "grad_norm": 0.14820928871631622, "learning_rate": 1.6505600081293072e-05, "loss": 0.0077, "step": 29880 }, { "grad_norm": 0.15607987344264984, "learning_rate": 1.647492056173912e-05, "loss": 0.0081, "step": 29890 }, { "grad_norm": 0.1899360567331314, "learning_rate": 1.644426395626085e-05, "loss": 0.0064, "step": 29900 }, { "grad_norm": 0.1774318516254425, "learning_rate": 1.641363028581175e-05, "loss": 0.0081, "step": 29910 }, { "grad_norm": 0.09163875132799149, "learning_rate": 1.638301957132965e-05, "loss": 0.0065, "step": 29920 }, { "grad_norm": 0.12935590744018555, "learning_rate": 1.6352431833736703e-05, "loss": 0.0083, "step": 29930 }, { "grad_norm": 0.08560208976268768, "learning_rate": 1.6321867093939298e-05, "loss": 0.0076, "step": 29940 }, { "grad_norm": 0.16435033082962036, "learning_rate": 1.629132537282817e-05, "loss": 0.0061, "step": 29950 }, { "grad_norm": 0.12006856501102448, "learning_rate": 1.62608066912783e-05, "loss": 0.0045, "step": 29960 }, { "grad_norm": 0.11996504664421082, "learning_rate": 1.623031107014893e-05, "loss": 0.0055, "step": 29970 }, { "grad_norm": 0.11580832302570343, "learning_rate": 1.619983853028351e-05, "loss": 0.0044, "step": 29980 }, { "grad_norm": 0.18905122578144073, "learning_rate": 1.6169389092509724e-05, "loss": 0.0053, "step": 29990 }, { "grad_norm": 0.1104143038392067, "learning_rate": 1.6138962777639494e-05, "loss": 0.0056, "step": 30000 }, { "grad_norm": 0.08946605026721954, "learning_rate": 1.610855960646891e-05, "loss": 0.0054, "step": 30010 }, { "grad_norm": 0.10010963678359985, "learning_rate": 1.607817959977826e-05, "loss": 0.0087, "step": 30020 }, { "grad_norm": 0.14470995962619781, "learning_rate": 1.6047822778332005e-05, "loss": 0.0067, "step": 30030 }, { "grad_norm": 0.1828659474849701, "learning_rate": 1.6017489162878713e-05, "loss": 0.0046, "step": 30040 }, { "grad_norm": 0.10107016563415527, "learning_rate": 1.5987178774151147e-05, "loss": 0.0079, "step": 30050 }, { "grad_norm": 0.12064748257398605, "learning_rate": 1.5956891632866184e-05, "loss": 0.0077, "step": 30060 }, { "grad_norm": 0.12297606468200684, "learning_rate": 1.592662775972476e-05, "loss": 0.0054, "step": 30070 }, { "grad_norm": 0.10537903755903244, "learning_rate": 1.5896387175411986e-05, "loss": 0.0053, "step": 30080 }, { "grad_norm": 0.11334668099880219, "learning_rate": 1.586616990059699e-05, "loss": 0.007, "step": 30090 }, { "grad_norm": 0.14097601175308228, "learning_rate": 1.5835975955933002e-05, "loss": 0.0064, "step": 30100 }, { "grad_norm": 0.1700064241886139, "learning_rate": 1.5805805362057295e-05, "loss": 0.0062, "step": 30110 }, { "grad_norm": 0.11068346351385117, "learning_rate": 1.57756581395912e-05, "loss": 0.0063, "step": 30120 }, { "grad_norm": 0.1357174515724182, "learning_rate": 1.574553430914006e-05, "loss": 0.0058, "step": 30130 }, { "grad_norm": 0.14661920070648193, "learning_rate": 1.5715433891293206e-05, "loss": 0.0104, "step": 30140 }, { "grad_norm": 0.1279469132423401, "learning_rate": 1.5685356906624016e-05, "loss": 0.0068, "step": 30150 }, { "grad_norm": 0.13177557289600372, "learning_rate": 1.56553033756898e-05, "loss": 0.0061, "step": 30160 }, { "grad_norm": 0.11312828958034515, "learning_rate": 1.5625273319031873e-05, "loss": 0.0078, "step": 30170 }, { "grad_norm": 0.13303685188293457, "learning_rate": 1.5595266757175518e-05, "loss": 0.0059, "step": 30180 }, { "grad_norm": 0.13588793575763702, "learning_rate": 1.55652837106299e-05, "loss": 0.0056, "step": 30190 }, { "grad_norm": 0.09021973609924316, "learning_rate": 1.5535324199888164e-05, "loss": 0.0074, "step": 30200 }, { "grad_norm": 0.13278000056743622, "learning_rate": 1.5505388245427355e-05, "loss": 0.0069, "step": 30210 }, { "grad_norm": 0.14342714846134186, "learning_rate": 1.5475475867708418e-05, "loss": 0.0083, "step": 30220 }, { "grad_norm": 0.12162662297487259, "learning_rate": 1.5445587087176195e-05, "loss": 0.0047, "step": 30230 }, { "grad_norm": 0.10748577117919922, "learning_rate": 1.5415721924259346e-05, "loss": 0.0053, "step": 30240 }, { "grad_norm": 0.13043799996376038, "learning_rate": 1.538588039937047e-05, "loss": 0.0094, "step": 30250 }, { "grad_norm": 0.11475728452205658, "learning_rate": 1.535606253290594e-05, "loss": 0.0051, "step": 30260 }, { "grad_norm": 0.11237353086471558, "learning_rate": 1.5326268345245993e-05, "loss": 0.0064, "step": 30270 }, { "grad_norm": 0.1722007542848587, "learning_rate": 1.5296497856754693e-05, "loss": 0.0081, "step": 30280 }, { "grad_norm": 0.14846311509609222, "learning_rate": 1.5266751087779858e-05, "loss": 0.0064, "step": 30290 }, { "grad_norm": 0.14646834135055542, "learning_rate": 1.5237028058653142e-05, "loss": 0.0057, "step": 30300 }, { "grad_norm": 0.08200599998235703, "learning_rate": 1.5207328789689946e-05, "loss": 0.0052, "step": 30310 }, { "grad_norm": 0.12058521807193756, "learning_rate": 1.5177653301189454e-05, "loss": 0.0069, "step": 30320 }, { "grad_norm": 0.10534191876649857, "learning_rate": 1.5148001613434587e-05, "loss": 0.0066, "step": 30330 }, { "grad_norm": 0.18368010222911835, "learning_rate": 1.5118373746691983e-05, "loss": 0.0071, "step": 30340 }, { "grad_norm": 0.14834241569042206, "learning_rate": 1.5088769721211992e-05, "loss": 0.0062, "step": 30350 }, { "grad_norm": 0.17970947921276093, "learning_rate": 1.505918955722871e-05, "loss": 0.0071, "step": 30360 }, { "grad_norm": 0.14011497795581818, "learning_rate": 1.5029633274959886e-05, "loss": 0.0074, "step": 30370 }, { "grad_norm": 0.1535424441099167, "learning_rate": 1.5000100894606989e-05, "loss": 0.0056, "step": 30380 }, { "grad_norm": 0.14109553396701813, "learning_rate": 1.4970592436355079e-05, "loss": 0.0064, "step": 30390 }, { "grad_norm": 0.11650156229734421, "learning_rate": 1.4941107920372938e-05, "loss": 0.0049, "step": 30400 }, { "grad_norm": 0.12518583238124847, "learning_rate": 1.4911647366812947e-05, "loss": 0.006, "step": 30410 }, { "grad_norm": 0.132012277841568, "learning_rate": 1.4882210795811114e-05, "loss": 0.0066, "step": 30420 }, { "grad_norm": 0.12809637188911438, "learning_rate": 1.485279822748708e-05, "loss": 0.0037, "step": 30430 }, { "grad_norm": 0.08472009748220444, "learning_rate": 1.4823409681944034e-05, "loss": 0.0046, "step": 30440 }, { "grad_norm": 0.17841586470603943, "learning_rate": 1.4794045179268763e-05, "loss": 0.0049, "step": 30450 }, { "grad_norm": 0.09317311644554138, "learning_rate": 1.4764704739531643e-05, "loss": 0.0048, "step": 30460 }, { "grad_norm": 0.13254566490650177, "learning_rate": 1.473538838278658e-05, "loss": 0.0055, "step": 30470 }, { "grad_norm": 0.10683121532201767, "learning_rate": 1.4706096129071045e-05, "loss": 0.0048, "step": 30480 }, { "grad_norm": 0.19398149847984314, "learning_rate": 1.4676827998405996e-05, "loss": 0.0058, "step": 30490 }, { "grad_norm": 0.133403480052948, "learning_rate": 1.4647584010795928e-05, "loss": 0.0062, "step": 30500 }, { "grad_norm": 0.1364758014678955, "learning_rate": 1.4618364186228834e-05, "loss": 0.0064, "step": 30510 }, { "grad_norm": 0.1385732889175415, "learning_rate": 1.4589168544676207e-05, "loss": 0.0063, "step": 30520 }, { "grad_norm": 0.11824866384267807, "learning_rate": 1.455999710609296e-05, "loss": 0.0072, "step": 30530 }, { "grad_norm": 0.09334558248519897, "learning_rate": 1.453084989041753e-05, "loss": 0.0055, "step": 30540 }, { "grad_norm": 0.15726879239082336, "learning_rate": 1.4501726917571728e-05, "loss": 0.0052, "step": 30550 }, { "grad_norm": 0.10642249882221222, "learning_rate": 1.4472628207460848e-05, "loss": 0.0071, "step": 30560 }, { "grad_norm": 0.1535877287387848, "learning_rate": 1.4443553779973584e-05, "loss": 0.0089, "step": 30570 }, { "grad_norm": 0.154709592461586, "learning_rate": 1.4414503654982037e-05, "loss": 0.0072, "step": 30580 }, { "grad_norm": 0.0990745946764946, "learning_rate": 1.4385477852341706e-05, "loss": 0.0048, "step": 30590 }, { "grad_norm": 0.15254198014736176, "learning_rate": 1.4356476391891421e-05, "loss": 0.0054, "step": 30600 }, { "grad_norm": 0.08952105045318604, "learning_rate": 1.4327499293453423e-05, "loss": 0.0067, "step": 30610 }, { "grad_norm": 0.11926613748073578, "learning_rate": 1.4298546576833304e-05, "loss": 0.0044, "step": 30620 }, { "grad_norm": 0.10912346094846725, "learning_rate": 1.426961826181994e-05, "loss": 0.0071, "step": 30630 }, { "grad_norm": 0.10697457194328308, "learning_rate": 1.4240714368185592e-05, "loss": 0.0062, "step": 30640 }, { "grad_norm": 0.08679800480604172, "learning_rate": 1.4211834915685767e-05, "loss": 0.0051, "step": 30650 }, { "grad_norm": 0.1258361041545868, "learning_rate": 1.4182979924059313e-05, "loss": 0.0068, "step": 30660 }, { "grad_norm": 0.08748875558376312, "learning_rate": 1.415414941302835e-05, "loss": 0.0047, "step": 30670 }, { "grad_norm": 0.12677817046642303, "learning_rate": 1.4125343402298257e-05, "loss": 0.0054, "step": 30680 }, { "grad_norm": 0.12208279967308044, "learning_rate": 1.4096561911557688e-05, "loss": 0.006, "step": 30690 }, { "grad_norm": 0.1390322744846344, "learning_rate": 1.4067804960478481e-05, "loss": 0.0057, "step": 30700 }, { "grad_norm": 0.12143436074256897, "learning_rate": 1.4039072568715784e-05, "loss": 0.0059, "step": 30710 }, { "grad_norm": 0.09880764782428741, "learning_rate": 1.401036475590788e-05, "loss": 0.0062, "step": 30720 }, { "grad_norm": 0.08566445857286453, "learning_rate": 1.3981681541676301e-05, "loss": 0.0066, "step": 30730 }, { "grad_norm": 0.13970623910427094, "learning_rate": 1.3953022945625772e-05, "loss": 0.0057, "step": 30740 }, { "grad_norm": 0.0930621549487114, "learning_rate": 1.3924388987344138e-05, "loss": 0.0043, "step": 30750 }, { "grad_norm": 0.11741407215595245, "learning_rate": 1.3895779686402465e-05, "loss": 0.0058, "step": 30760 }, { "grad_norm": 0.1303301602602005, "learning_rate": 1.3867195062354931e-05, "loss": 0.0066, "step": 30770 }, { "grad_norm": 0.10337809473276138, "learning_rate": 1.3838635134738858e-05, "loss": 0.0072, "step": 30780 }, { "grad_norm": 0.1979617178440094, "learning_rate": 1.3810099923074704e-05, "loss": 0.0043, "step": 30790 }, { "grad_norm": 0.15251818299293518, "learning_rate": 1.3781589446866006e-05, "loss": 0.0073, "step": 30800 }, { "grad_norm": 0.11815519630908966, "learning_rate": 1.3753103725599387e-05, "loss": 0.0067, "step": 30810 }, { "grad_norm": 0.16803507506847382, "learning_rate": 1.3724642778744579e-05, "loss": 0.0064, "step": 30820 }, { "grad_norm": 0.11711857467889786, "learning_rate": 1.3696206625754376e-05, "loss": 0.0071, "step": 30830 }, { "grad_norm": 0.1375533640384674, "learning_rate": 1.3667795286064639e-05, "loss": 0.0045, "step": 30840 }, { "grad_norm": 0.13057318329811096, "learning_rate": 1.3639408779094215e-05, "loss": 0.0058, "step": 30850 }, { "grad_norm": 0.1596667319536209, "learning_rate": 1.361104712424503e-05, "loss": 0.0066, "step": 30860 }, { "grad_norm": 0.17914336919784546, "learning_rate": 1.358271034090201e-05, "loss": 0.0089, "step": 30870 }, { "grad_norm": 0.15612053871154785, "learning_rate": 1.3554398448433076e-05, "loss": 0.0072, "step": 30880 }, { "grad_norm": 0.20718792080879211, "learning_rate": 1.3526111466189156e-05, "loss": 0.0051, "step": 30890 }, { "grad_norm": 0.15652456879615784, "learning_rate": 1.3497849413504121e-05, "loss": 0.006, "step": 30900 }, { "grad_norm": 0.20362335443496704, "learning_rate": 1.3469612309694802e-05, "loss": 0.0056, "step": 30910 }, { "grad_norm": 0.1255751997232437, "learning_rate": 1.3441400174061008e-05, "loss": 0.0061, "step": 30920 }, { "grad_norm": 0.12249413132667542, "learning_rate": 1.3413213025885469e-05, "loss": 0.0055, "step": 30930 }, { "grad_norm": 0.13658414781093597, "learning_rate": 1.3385050884433853e-05, "loss": 0.0045, "step": 30940 }, { "grad_norm": 0.08953391760587692, "learning_rate": 1.3356913768954676e-05, "loss": 0.0054, "step": 30950 }, { "grad_norm": 0.15743906795978546, "learning_rate": 1.3328801698679415e-05, "loss": 0.0063, "step": 30960 }, { "grad_norm": 0.0722465068101883, "learning_rate": 1.3300714692822413e-05, "loss": 0.0052, "step": 30970 }, { "grad_norm": 0.09136733412742615, "learning_rate": 1.3272652770580857e-05, "loss": 0.0053, "step": 30980 }, { "grad_norm": 0.09239678829908371, "learning_rate": 1.3244615951134831e-05, "loss": 0.0072, "step": 30990 }, { "grad_norm": 0.08317113667726517, "learning_rate": 1.3216604253647219e-05, "loss": 0.0048, "step": 31000 }, { "grad_norm": 0.11575011163949966, "learning_rate": 1.3188617697263738e-05, "loss": 0.0059, "step": 31010 }, { "grad_norm": 0.16311568021774292, "learning_rate": 1.3160656301112956e-05, "loss": 0.006, "step": 31020 }, { "grad_norm": 0.13330861926078796, "learning_rate": 1.3132720084306211e-05, "loss": 0.0049, "step": 31030 }, { "grad_norm": 0.19598229229450226, "learning_rate": 1.3104809065937668e-05, "loss": 0.0081, "step": 31040 }, { "grad_norm": 0.12164322286844254, "learning_rate": 1.3076923265084212e-05, "loss": 0.0054, "step": 31050 }, { "grad_norm": 0.09925414621829987, "learning_rate": 1.3049062700805543e-05, "loss": 0.005, "step": 31060 }, { "grad_norm": 0.09200949966907501, "learning_rate": 1.3021227392144092e-05, "loss": 0.0053, "step": 31070 }, { "grad_norm": 0.0812779888510704, "learning_rate": 1.2993417358125037e-05, "loss": 0.0068, "step": 31080 }, { "grad_norm": 0.11830601096153259, "learning_rate": 1.2965632617756257e-05, "loss": 0.005, "step": 31090 }, { "grad_norm": 0.1496943235397339, "learning_rate": 1.2937873190028387e-05, "loss": 0.0066, "step": 31100 }, { "grad_norm": 0.18493543565273285, "learning_rate": 1.2910139093914697e-05, "loss": 0.0053, "step": 31110 }, { "grad_norm": 0.1350683867931366, "learning_rate": 1.2882430348371205e-05, "loss": 0.0064, "step": 31120 }, { "grad_norm": 0.1424187272787094, "learning_rate": 1.2854746972336567e-05, "loss": 0.0046, "step": 31130 }, { "grad_norm": 0.11483755707740784, "learning_rate": 1.2827088984732132e-05, "loss": 0.0058, "step": 31140 }, { "grad_norm": 0.18764007091522217, "learning_rate": 1.279945640446184e-05, "loss": 0.0056, "step": 31150 }, { "grad_norm": 0.1463969200849533, "learning_rate": 1.2771849250412315e-05, "loss": 0.0063, "step": 31160 }, { "grad_norm": 0.16679304838180542, "learning_rate": 1.2744267541452792e-05, "loss": 0.0055, "step": 31170 }, { "grad_norm": 0.12967640161514282, "learning_rate": 1.2716711296435112e-05, "loss": 0.006, "step": 31180 }, { "grad_norm": 0.10612040758132935, "learning_rate": 1.268918053419369e-05, "loss": 0.0051, "step": 31190 }, { "grad_norm": 0.14121782779693604, "learning_rate": 1.2661675273545564e-05, "loss": 0.0054, "step": 31200 }, { "grad_norm": 0.11863677948713303, "learning_rate": 1.26341955332903e-05, "loss": 0.0059, "step": 31210 }, { "grad_norm": 0.15872490406036377, "learning_rate": 1.2606741332210048e-05, "loss": 0.0076, "step": 31220 }, { "grad_norm": 0.12775424122810364, "learning_rate": 1.2579312689069495e-05, "loss": 0.0066, "step": 31230 }, { "grad_norm": 0.09721533954143524, "learning_rate": 1.2551909622615881e-05, "loss": 0.0081, "step": 31240 }, { "grad_norm": 0.1357613205909729, "learning_rate": 1.252453215157891e-05, "loss": 0.006, "step": 31250 }, { "grad_norm": 0.10686536878347397, "learning_rate": 1.2497180294670841e-05, "loss": 0.0045, "step": 31260 }, { "grad_norm": 0.11656283587217331, "learning_rate": 1.2469854070586423e-05, "loss": 0.006, "step": 31270 }, { "grad_norm": 0.08382941782474518, "learning_rate": 1.2442553498002845e-05, "loss": 0.008, "step": 31280 }, { "grad_norm": 0.16029256582260132, "learning_rate": 1.241527859557981e-05, "loss": 0.0073, "step": 31290 }, { "grad_norm": 0.1837281435728073, "learning_rate": 1.238802938195946e-05, "loss": 0.0075, "step": 31300 }, { "grad_norm": 0.09366752207279205, "learning_rate": 1.2360805875766357e-05, "loss": 0.005, "step": 31310 }, { "grad_norm": 0.08309940993785858, "learning_rate": 1.2333608095607524e-05, "loss": 0.0049, "step": 31320 }, { "grad_norm": 0.08554528653621674, "learning_rate": 1.230643606007238e-05, "loss": 0.0055, "step": 31330 }, { "grad_norm": 0.1198941096663475, "learning_rate": 1.2279289787732778e-05, "loss": 0.0058, "step": 31340 }, { "grad_norm": 0.13368400931358337, "learning_rate": 1.2252169297142906e-05, "loss": 0.0067, "step": 31350 }, { "grad_norm": 0.115336574614048, "learning_rate": 1.2225074606839392e-05, "loss": 0.0072, "step": 31360 }, { "grad_norm": 0.09297240525484085, "learning_rate": 1.2198005735341173e-05, "loss": 0.0065, "step": 31370 }, { "grad_norm": 0.1262674182653427, "learning_rate": 1.2170962701149584e-05, "loss": 0.0062, "step": 31380 }, { "grad_norm": 0.11310553550720215, "learning_rate": 1.2143945522748285e-05, "loss": 0.0088, "step": 31390 }, { "grad_norm": 0.08725433051586151, "learning_rate": 1.2116954218603266e-05, "loss": 0.0059, "step": 31400 }, { "grad_norm": 0.10730426758527756, "learning_rate": 1.2089988807162817e-05, "loss": 0.007, "step": 31410 }, { "grad_norm": 0.14505863189697266, "learning_rate": 1.2063049306857543e-05, "loss": 0.0078, "step": 31420 }, { "grad_norm": 0.11650381982326508, "learning_rate": 1.2036135736100346e-05, "loss": 0.0059, "step": 31430 }, { "grad_norm": 0.14519859850406647, "learning_rate": 1.200924811328641e-05, "loss": 0.0063, "step": 31440 }, { "grad_norm": 0.13687238097190857, "learning_rate": 1.1982386456793143e-05, "loss": 0.0052, "step": 31450 }, { "grad_norm": 0.11201611906290054, "learning_rate": 1.1955550784980257e-05, "loss": 0.0072, "step": 31460 }, { "grad_norm": 0.1236787661910057, "learning_rate": 1.1928741116189663e-05, "loss": 0.0058, "step": 31470 }, { "grad_norm": 0.1502746045589447, "learning_rate": 1.1901957468745529e-05, "loss": 0.0053, "step": 31480 }, { "grad_norm": 0.10219243913888931, "learning_rate": 1.1875199860954217e-05, "loss": 0.006, "step": 31490 }, { "grad_norm": 0.13118495047092438, "learning_rate": 1.1848468311104316e-05, "loss": 0.0061, "step": 31500 }, { "grad_norm": 0.11719825118780136, "learning_rate": 1.1821762837466566e-05, "loss": 0.0056, "step": 31510 }, { "grad_norm": 0.0923442468047142, "learning_rate": 1.179508345829392e-05, "loss": 0.0054, "step": 31520 }, { "grad_norm": 0.13272875547409058, "learning_rate": 1.1768430191821478e-05, "loss": 0.0046, "step": 31530 }, { "grad_norm": 0.1051948070526123, "learning_rate": 1.174180305626651e-05, "loss": 0.0079, "step": 31540 }, { "grad_norm": 0.10814747959375381, "learning_rate": 1.1715202069828401e-05, "loss": 0.0084, "step": 31550 }, { "grad_norm": 0.12318321317434311, "learning_rate": 1.1688627250688661e-05, "loss": 0.0047, "step": 31560 }, { "grad_norm": 0.09410551190376282, "learning_rate": 1.1662078617010935e-05, "loss": 0.0044, "step": 31570 }, { "grad_norm": 0.11497769504785538, "learning_rate": 1.1635556186940972e-05, "loss": 0.004, "step": 31580 }, { "grad_norm": 0.10995882749557495, "learning_rate": 1.1609059978606602e-05, "loss": 0.0049, "step": 31590 }, { "grad_norm": 0.14610172808170319, "learning_rate": 1.1582590010117745e-05, "loss": 0.0058, "step": 31600 }, { "grad_norm": 0.14146395027637482, "learning_rate": 1.1556146299566345e-05, "loss": 0.0054, "step": 31610 }, { "grad_norm": 0.19382086396217346, "learning_rate": 1.1529728865026452e-05, "loss": 0.0081, "step": 31620 }, { "grad_norm": 0.13294313848018646, "learning_rate": 1.150333772455412e-05, "loss": 0.0039, "step": 31630 }, { "grad_norm": 0.14665858447551727, "learning_rate": 1.1476972896187466e-05, "loss": 0.0073, "step": 31640 }, { "grad_norm": 0.1336429864168167, "learning_rate": 1.1450634397946592e-05, "loss": 0.006, "step": 31650 }, { "grad_norm": 0.13245058059692383, "learning_rate": 1.142432224783359e-05, "loss": 0.0072, "step": 31660 }, { "grad_norm": 0.14209356904029846, "learning_rate": 1.1398036463832584e-05, "loss": 0.007, "step": 31670 }, { "grad_norm": 0.11865252256393433, "learning_rate": 1.1371777063909661e-05, "loss": 0.0059, "step": 31680 }, { "grad_norm": 0.10404808074235916, "learning_rate": 1.1345544066012874e-05, "loss": 0.0045, "step": 31690 }, { "grad_norm": 0.14657647907733917, "learning_rate": 1.131933748807224e-05, "loss": 0.006, "step": 31700 }, { "grad_norm": 0.14532265067100525, "learning_rate": 1.1293157347999683e-05, "loss": 0.0053, "step": 31710 }, { "grad_norm": 0.11712052673101425, "learning_rate": 1.1267003663689097e-05, "loss": 0.0048, "step": 31720 }, { "grad_norm": 0.15655732154846191, "learning_rate": 1.1240876453016275e-05, "loss": 0.0055, "step": 31730 }, { "grad_norm": 0.09156293421983719, "learning_rate": 1.1214775733838933e-05, "loss": 0.0055, "step": 31740 }, { "grad_norm": 0.13773000240325928, "learning_rate": 1.118870152399666e-05, "loss": 0.0065, "step": 31750 }, { "grad_norm": 0.1344355195760727, "learning_rate": 1.1162653841310905e-05, "loss": 0.0052, "step": 31760 }, { "grad_norm": 0.17028498649597168, "learning_rate": 1.1136632703585043e-05, "loss": 0.0063, "step": 31770 }, { "grad_norm": 0.12043128907680511, "learning_rate": 1.111063812860426e-05, "loss": 0.0053, "step": 31780 }, { "grad_norm": 0.18833370506763458, "learning_rate": 1.1084670134135616e-05, "loss": 0.006, "step": 31790 }, { "grad_norm": 0.20924395322799683, "learning_rate": 1.1058728737927998e-05, "loss": 0.0061, "step": 31800 }, { "grad_norm": 0.0866171345114708, "learning_rate": 1.1032813957712086e-05, "loss": 0.0058, "step": 31810 }, { "grad_norm": 0.10136119276285172, "learning_rate": 1.1006925811200385e-05, "loss": 0.0053, "step": 31820 }, { "grad_norm": 0.08675794303417206, "learning_rate": 1.0981064316087231e-05, "loss": 0.0062, "step": 31830 }, { "grad_norm": 0.12455251067876816, "learning_rate": 1.0955229490048675e-05, "loss": 0.005, "step": 31840 }, { "grad_norm": 0.1331498920917511, "learning_rate": 1.0929421350742603e-05, "loss": 0.0053, "step": 31850 }, { "grad_norm": 0.1318119913339615, "learning_rate": 1.090363991580861e-05, "loss": 0.0052, "step": 31860 }, { "grad_norm": 0.13069255650043488, "learning_rate": 1.0877885202868083e-05, "loss": 0.0051, "step": 31870 }, { "grad_norm": 0.09980485588312149, "learning_rate": 1.0852157229524118e-05, "loss": 0.0046, "step": 31880 }, { "grad_norm": 0.08976244926452637, "learning_rate": 1.0826456013361542e-05, "loss": 0.0043, "step": 31890 }, { "grad_norm": 0.19860054552555084, "learning_rate": 1.080078157194691e-05, "loss": 0.0049, "step": 31900 }, { "grad_norm": 0.10480981320142746, "learning_rate": 1.077513392282844e-05, "loss": 0.0059, "step": 31910 }, { "grad_norm": 0.12972302734851837, "learning_rate": 1.0749513083536073e-05, "loss": 0.0054, "step": 31920 }, { "grad_norm": 0.1179666817188263, "learning_rate": 1.072391907158139e-05, "loss": 0.005, "step": 31930 }, { "grad_norm": 0.12914207577705383, "learning_rate": 1.0698351904457676e-05, "loss": 0.0083, "step": 31940 }, { "grad_norm": 0.13989941775798798, "learning_rate": 1.0672811599639849e-05, "loss": 0.0056, "step": 31950 }, { "grad_norm": 0.1804496794939041, "learning_rate": 1.0647298174584453e-05, "loss": 0.0088, "step": 31960 }, { "grad_norm": 0.0969996526837349, "learning_rate": 1.0621811646729673e-05, "loss": 0.0047, "step": 31970 }, { "grad_norm": 0.09886324405670166, "learning_rate": 1.0596352033495328e-05, "loss": 0.0068, "step": 31980 }, { "grad_norm": 0.11406455188989639, "learning_rate": 1.057091935228281e-05, "loss": 0.0046, "step": 31990 }, { "grad_norm": 0.12893877923488617, "learning_rate": 1.0545513620475134e-05, "loss": 0.005, "step": 32000 }, { "grad_norm": 0.0978609248995781, "learning_rate": 1.0520134855436863e-05, "loss": 0.0066, "step": 32010 }, { "grad_norm": 0.14193041622638702, "learning_rate": 1.049478307451413e-05, "loss": 0.0049, "step": 32020 }, { "grad_norm": 0.14668647944927216, "learning_rate": 1.0469458295034656e-05, "loss": 0.005, "step": 32030 }, { "grad_norm": 0.0939999595284462, "learning_rate": 1.044416053430769e-05, "loss": 0.0047, "step": 32040 }, { "grad_norm": 0.10883685946464539, "learning_rate": 1.0418889809624021e-05, "loss": 0.0044, "step": 32050 }, { "grad_norm": 0.11119500547647476, "learning_rate": 1.0393646138255925e-05, "loss": 0.0059, "step": 32060 }, { "grad_norm": 0.11453124135732651, "learning_rate": 1.036842953745723e-05, "loss": 0.0039, "step": 32070 }, { "grad_norm": 0.18495404720306396, "learning_rate": 1.034324002446324e-05, "loss": 0.0061, "step": 32080 }, { "grad_norm": 0.10323978215456009, "learning_rate": 1.0318077616490751e-05, "loss": 0.0058, "step": 32090 }, { "grad_norm": 0.10241413861513138, "learning_rate": 1.0292942330738054e-05, "loss": 0.0043, "step": 32100 }, { "grad_norm": 0.07880427688360214, "learning_rate": 1.026783418438485e-05, "loss": 0.0045, "step": 32110 }, { "grad_norm": 0.12482021749019623, "learning_rate": 1.0242753194592319e-05, "loss": 0.0055, "step": 32120 }, { "grad_norm": 0.17071546614170074, "learning_rate": 1.0217699378503092e-05, "loss": 0.0052, "step": 32130 }, { "grad_norm": 0.10948595404624939, "learning_rate": 1.0192672753241218e-05, "loss": 0.0062, "step": 32140 }, { "grad_norm": 0.09446168690919876, "learning_rate": 1.0167673335912164e-05, "loss": 0.007, "step": 32150 }, { "grad_norm": 0.14266367256641388, "learning_rate": 1.0142701143602779e-05, "loss": 0.0045, "step": 32160 }, { "grad_norm": 0.20837317407131195, "learning_rate": 1.0117756193381328e-05, "loss": 0.005, "step": 32170 }, { "grad_norm": 0.1476965695619583, "learning_rate": 1.0092838502297453e-05, "loss": 0.0053, "step": 32180 }, { "grad_norm": 0.08046279102563858, "learning_rate": 1.0067948087382162e-05, "loss": 0.0048, "step": 32190 }, { "grad_norm": 0.08161529898643494, "learning_rate": 1.0043084965647826e-05, "loss": 0.0046, "step": 32200 }, { "grad_norm": 0.11386086791753769, "learning_rate": 1.001824915408815e-05, "loss": 0.0096, "step": 32210 }, { "grad_norm": 0.13158529996871948, "learning_rate": 9.99344066967815e-06, "loss": 0.0051, "step": 32220 }, { "grad_norm": 0.11530893296003342, "learning_rate": 9.968659529374219e-06, "loss": 0.0061, "step": 32230 }, { "grad_norm": 0.19272443652153015, "learning_rate": 9.943905750114025e-06, "loss": 0.006, "step": 32240 }, { "grad_norm": 0.16914881765842438, "learning_rate": 9.919179348816559e-06, "loss": 0.0043, "step": 32250 }, { "grad_norm": 0.11883225291967392, "learning_rate": 9.894480342382056e-06, "loss": 0.0077, "step": 32260 }, { "grad_norm": 0.1337100863456726, "learning_rate": 9.869808747692061e-06, "loss": 0.0054, "step": 32270 }, { "grad_norm": 0.18100041151046753, "learning_rate": 9.84516458160939e-06, "loss": 0.0063, "step": 32280 }, { "grad_norm": 0.12995818257331848, "learning_rate": 9.820547860978097e-06, "loss": 0.0061, "step": 32290 }, { "grad_norm": 0.12851528823375702, "learning_rate": 9.795958602623461e-06, "loss": 0.0055, "step": 32300 }, { "grad_norm": 0.1674843281507492, "learning_rate": 9.77139682335203e-06, "loss": 0.0045, "step": 32310 }, { "grad_norm": 0.09862326830625534, "learning_rate": 9.74686253995153e-06, "loss": 0.0039, "step": 32320 }, { "grad_norm": 0.09865310788154602, "learning_rate": 9.722355769190911e-06, "loss": 0.0054, "step": 32330 }, { "grad_norm": 0.10320007801055908, "learning_rate": 9.697876527820333e-06, "loss": 0.0051, "step": 32340 }, { "grad_norm": 0.08999566733837128, "learning_rate": 9.673424832571132e-06, "loss": 0.0052, "step": 32350 }, { "grad_norm": 0.07442521303892136, "learning_rate": 9.649000700155791e-06, "loss": 0.0056, "step": 32360 }, { "grad_norm": 0.09472926706075668, "learning_rate": 9.624604147267984e-06, "loss": 0.0049, "step": 32370 }, { "grad_norm": 0.07592618465423584, "learning_rate": 9.600235190582524e-06, "loss": 0.0045, "step": 32380 }, { "grad_norm": 0.18560846149921417, "learning_rate": 9.57589384675538e-06, "loss": 0.0077, "step": 32390 }, { "grad_norm": 0.17536893486976624, "learning_rate": 9.551580132423598e-06, "loss": 0.0062, "step": 32400 }, { "grad_norm": 0.1146709993481636, "learning_rate": 9.527294064205405e-06, "loss": 0.0084, "step": 32410 }, { "grad_norm": 0.1886138767004013, "learning_rate": 9.503035658700072e-06, "loss": 0.0061, "step": 32420 }, { "grad_norm": 0.15129122138023376, "learning_rate": 9.478804932488e-06, "loss": 0.0047, "step": 32430 }, { "grad_norm": 0.08488570898771286, "learning_rate": 9.454601902130667e-06, "loss": 0.0045, "step": 32440 }, { "grad_norm": 0.07380606979131699, "learning_rate": 9.430426584170626e-06, "loss": 0.0057, "step": 32450 }, { "grad_norm": 0.11336857080459595, "learning_rate": 9.40627899513145e-06, "loss": 0.006, "step": 32460 }, { "grad_norm": 0.12657026946544647, "learning_rate": 9.38215915151781e-06, "loss": 0.0066, "step": 32470 }, { "grad_norm": 0.15018609166145325, "learning_rate": 9.358067069815396e-06, "loss": 0.0064, "step": 32480 }, { "grad_norm": 0.11540603637695312, "learning_rate": 9.334002766490901e-06, "loss": 0.0049, "step": 32490 }, { "grad_norm": 0.09140694886445999, "learning_rate": 9.309966257992054e-06, "loss": 0.0043, "step": 32500 }, { "grad_norm": 0.108395516872406, "learning_rate": 9.2859575607476e-06, "loss": 0.0051, "step": 32510 }, { "grad_norm": 0.1643579751253128, "learning_rate": 9.26197669116723e-06, "loss": 0.0063, "step": 32520 }, { "grad_norm": 0.14760349690914154, "learning_rate": 9.238023665641655e-06, "loss": 0.0051, "step": 32530 }, { "grad_norm": 0.17843948304653168, "learning_rate": 9.21409850054254e-06, "loss": 0.0095, "step": 32540 }, { "grad_norm": 0.13395318388938904, "learning_rate": 9.190201212222526e-06, "loss": 0.0049, "step": 32550 }, { "grad_norm": 0.11136334389448166, "learning_rate": 9.166331817015155e-06, "loss": 0.0044, "step": 32560 }, { "grad_norm": 0.20288673043251038, "learning_rate": 9.142490331234965e-06, "loss": 0.0055, "step": 32570 }, { "grad_norm": 0.17430752515792847, "learning_rate": 9.11867677117736e-06, "loss": 0.0065, "step": 32580 }, { "grad_norm": 0.09647944569587708, "learning_rate": 9.09489115311869e-06, "loss": 0.0076, "step": 32590 }, { "grad_norm": 0.08177340030670166, "learning_rate": 9.071133493316202e-06, "loss": 0.0085, "step": 32600 }, { "grad_norm": 0.10607489943504333, "learning_rate": 9.047403808008053e-06, "loss": 0.0039, "step": 32610 }, { "grad_norm": 0.135772705078125, "learning_rate": 9.023702113413224e-06, "loss": 0.0036, "step": 32620 }, { "grad_norm": 0.10212912410497665, "learning_rate": 9.00002842573162e-06, "loss": 0.0061, "step": 32630 }, { "grad_norm": 0.08919470757246017, "learning_rate": 8.976382761143981e-06, "loss": 0.0047, "step": 32640 }, { "grad_norm": 0.19446420669555664, "learning_rate": 8.952765135811909e-06, "loss": 0.004, "step": 32650 }, { "grad_norm": 0.17990098893642426, "learning_rate": 8.929175565877795e-06, "loss": 0.0062, "step": 32660 }, { "grad_norm": 0.074907585978508, "learning_rate": 8.905614067464924e-06, "loss": 0.005, "step": 32670 }, { "grad_norm": 0.0983472540974617, "learning_rate": 8.882080656677333e-06, "loss": 0.0077, "step": 32680 }, { "grad_norm": 0.12635299563407898, "learning_rate": 8.85857534959989e-06, "loss": 0.006, "step": 32690 }, { "grad_norm": 0.15076972544193268, "learning_rate": 8.835098162298256e-06, "loss": 0.0054, "step": 32700 }, { "grad_norm": 0.07241670042276382, "learning_rate": 8.811649110818875e-06, "loss": 0.0038, "step": 32710 }, { "grad_norm": 0.10839580744504929, "learning_rate": 8.78822821118893e-06, "loss": 0.0045, "step": 32720 }, { "grad_norm": 0.14555032551288605, "learning_rate": 8.764835479416405e-06, "loss": 0.0068, "step": 32730 }, { "grad_norm": 0.10116106271743774, "learning_rate": 8.741470931490003e-06, "loss": 0.0035, "step": 32740 }, { "grad_norm": 0.1201654002070427, "learning_rate": 8.718134583379189e-06, "loss": 0.0058, "step": 32750 }, { "grad_norm": 0.11342626065015793, "learning_rate": 8.69482645103412e-06, "loss": 0.0047, "step": 32760 }, { "grad_norm": 0.14663302898406982, "learning_rate": 8.671546550385679e-06, "loss": 0.0062, "step": 32770 }, { "grad_norm": 0.1225062757730484, "learning_rate": 8.648294897345466e-06, "loss": 0.0051, "step": 32780 }, { "grad_norm": 0.14066335558891296, "learning_rate": 8.62507150780576e-06, "loss": 0.0044, "step": 32790 }, { "grad_norm": 0.15104442834854126, "learning_rate": 8.601876397639542e-06, "loss": 0.0047, "step": 32800 }, { "grad_norm": 0.16385579109191895, "learning_rate": 8.578709582700456e-06, "loss": 0.0046, "step": 32810 }, { "grad_norm": 0.09102800488471985, "learning_rate": 8.555571078822783e-06, "loss": 0.0038, "step": 32820 }, { "grad_norm": 0.14797019958496094, "learning_rate": 8.532460901821477e-06, "loss": 0.0079, "step": 32830 }, { "grad_norm": 0.12482073158025742, "learning_rate": 8.509379067492128e-06, "loss": 0.0043, "step": 32840 }, { "grad_norm": 0.09622306376695633, "learning_rate": 8.486325591610967e-06, "loss": 0.0048, "step": 32850 }, { "grad_norm": 0.1692706197500229, "learning_rate": 8.463300489934811e-06, "loss": 0.0045, "step": 32860 }, { "grad_norm": 0.06146732717752457, "learning_rate": 8.440303778201093e-06, "loss": 0.0049, "step": 32870 }, { "grad_norm": 0.08559597283601761, "learning_rate": 8.417335472127858e-06, "loss": 0.0066, "step": 32880 }, { "grad_norm": 0.07387171685695648, "learning_rate": 8.394395587413722e-06, "loss": 0.0043, "step": 32890 }, { "grad_norm": 0.11065037548542023, "learning_rate": 8.371484139737889e-06, "loss": 0.006, "step": 32900 }, { "grad_norm": 0.07115335762500763, "learning_rate": 8.348601144760127e-06, "loss": 0.0034, "step": 32910 }, { "grad_norm": 0.14341780543327332, "learning_rate": 8.325746618120716e-06, "loss": 0.0059, "step": 32920 }, { "grad_norm": 0.14130601286888123, "learning_rate": 8.302920575440531e-06, "loss": 0.0048, "step": 32930 }, { "grad_norm": 0.14879052340984344, "learning_rate": 8.280123032320953e-06, "loss": 0.0079, "step": 32940 }, { "grad_norm": 0.10638098418712616, "learning_rate": 8.257354004343892e-06, "loss": 0.0059, "step": 32950 }, { "grad_norm": 0.08909853547811508, "learning_rate": 8.234613507071754e-06, "loss": 0.0043, "step": 32960 }, { "grad_norm": 0.1732315868139267, "learning_rate": 8.211901556047475e-06, "loss": 0.0043, "step": 32970 }, { "grad_norm": 0.16210900247097015, "learning_rate": 8.189218166794427e-06, "loss": 0.0051, "step": 32980 }, { "grad_norm": 0.0876123383641243, "learning_rate": 8.166563354816509e-06, "loss": 0.0061, "step": 32990 }, { "grad_norm": 0.06685898452997208, "learning_rate": 8.143937135598074e-06, "loss": 0.0048, "step": 33000 }, { "grad_norm": 0.09608610719442368, "learning_rate": 8.121339524603943e-06, "loss": 0.0045, "step": 33010 }, { "grad_norm": 0.11319822072982788, "learning_rate": 8.098770537279338e-06, "loss": 0.0051, "step": 33020 }, { "grad_norm": 0.07577566057443619, "learning_rate": 8.076230189049966e-06, "loss": 0.0072, "step": 33030 }, { "grad_norm": 0.07686548680067062, "learning_rate": 8.05371849532196e-06, "loss": 0.0057, "step": 33040 }, { "grad_norm": 0.14452698826789856, "learning_rate": 8.031235471481813e-06, "loss": 0.0062, "step": 33050 }, { "grad_norm": 0.09420034289360046, "learning_rate": 8.008781132896486e-06, "loss": 0.0053, "step": 33060 }, { "grad_norm": 0.1153821274638176, "learning_rate": 7.986355494913312e-06, "loss": 0.0051, "step": 33070 }, { "grad_norm": 0.08388783037662506, "learning_rate": 7.963958572859975e-06, "loss": 0.007, "step": 33080 }, { "grad_norm": 0.12108151614665985, "learning_rate": 7.941590382044583e-06, "loss": 0.0076, "step": 33090 }, { "grad_norm": 0.13555370271205902, "learning_rate": 7.919250937755574e-06, "loss": 0.0058, "step": 33100 }, { "grad_norm": 0.10200910270214081, "learning_rate": 7.896940255261764e-06, "loss": 0.005, "step": 33110 }, { "grad_norm": 0.08333548158407211, "learning_rate": 7.874658349812269e-06, "loss": 0.005, "step": 33120 }, { "grad_norm": 0.12891116738319397, "learning_rate": 7.85240523663659e-06, "loss": 0.0049, "step": 33130 }, { "grad_norm": 0.15497110784053802, "learning_rate": 7.830180930944487e-06, "loss": 0.0079, "step": 33140 }, { "grad_norm": 0.11120659857988358, "learning_rate": 7.80798544792608e-06, "loss": 0.0055, "step": 33150 }, { "grad_norm": 0.0997796356678009, "learning_rate": 7.785818802751776e-06, "loss": 0.0056, "step": 33160 }, { "grad_norm": 0.09689056873321533, "learning_rate": 7.763681010572277e-06, "loss": 0.0076, "step": 33170 }, { "grad_norm": 0.13008198142051697, "learning_rate": 7.741572086518528e-06, "loss": 0.0063, "step": 33180 }, { "grad_norm": 0.12687397003173828, "learning_rate": 7.719492045701782e-06, "loss": 0.0041, "step": 33190 }, { "grad_norm": 0.1231890544295311, "learning_rate": 7.697440903213549e-06, "loss": 0.0045, "step": 33200 }, { "grad_norm": 0.15663586556911469, "learning_rate": 7.675418674125573e-06, "loss": 0.0073, "step": 33210 }, { "grad_norm": 0.09568531066179276, "learning_rate": 7.65342537348983e-06, "loss": 0.0051, "step": 33220 }, { "grad_norm": 0.14049135148525238, "learning_rate": 7.631461016338549e-06, "loss": 0.0053, "step": 33230 }, { "grad_norm": 0.11985337734222412, "learning_rate": 7.609525617684138e-06, "loss": 0.0049, "step": 33240 }, { "grad_norm": 0.07338908314704895, "learning_rate": 7.587619192519252e-06, "loss": 0.0053, "step": 33250 }, { "grad_norm": 0.11353054642677307, "learning_rate": 7.565741755816725e-06, "loss": 0.0082, "step": 33260 }, { "grad_norm": 0.11354842782020569, "learning_rate": 7.543893322529588e-06, "loss": 0.0059, "step": 33270 }, { "grad_norm": 0.09981539100408554, "learning_rate": 7.522073907591015e-06, "loss": 0.0061, "step": 33280 }, { "grad_norm": 0.07476329803466797, "learning_rate": 7.500283525914381e-06, "loss": 0.0056, "step": 33290 }, { "grad_norm": 0.11073248088359833, "learning_rate": 7.478522192393211e-06, "loss": 0.0044, "step": 33300 }, { "grad_norm": 0.14894990622997284, "learning_rate": 7.456789921901181e-06, "loss": 0.0052, "step": 33310 }, { "grad_norm": 0.08509424328804016, "learning_rate": 7.435086729292079e-06, "loss": 0.0039, "step": 33320 }, { "grad_norm": 0.15809080004692078, "learning_rate": 7.413412629399824e-06, "loss": 0.0088, "step": 33330 }, { "grad_norm": 0.10662905871868134, "learning_rate": 7.391767637038466e-06, "loss": 0.0055, "step": 33340 }, { "grad_norm": 0.11962850391864777, "learning_rate": 7.370151767002159e-06, "loss": 0.0041, "step": 33350 }, { "grad_norm": 0.14126715064048767, "learning_rate": 7.348565034065147e-06, "loss": 0.0055, "step": 33360 }, { "grad_norm": 0.09332563728094101, "learning_rate": 7.327007452981765e-06, "loss": 0.005, "step": 33370 }, { "grad_norm": 0.16744552552700043, "learning_rate": 7.305479038486396e-06, "loss": 0.0055, "step": 33380 }, { "grad_norm": 0.09975653886795044, "learning_rate": 7.283979805293517e-06, "loss": 0.0055, "step": 33390 }, { "grad_norm": 0.15195417404174805, "learning_rate": 7.262509768097653e-06, "loss": 0.0051, "step": 33400 }, { "grad_norm": 0.1907191276550293, "learning_rate": 7.241068941573381e-06, "loss": 0.0057, "step": 33410 }, { "grad_norm": 0.12024426460266113, "learning_rate": 7.219657340375296e-06, "loss": 0.0039, "step": 33420 }, { "grad_norm": 0.13212329149246216, "learning_rate": 7.198274979138014e-06, "loss": 0.0049, "step": 33430 }, { "grad_norm": 0.12311124801635742, "learning_rate": 7.176921872476189e-06, "loss": 0.0086, "step": 33440 }, { "grad_norm": 0.09952236711978912, "learning_rate": 7.1555980349844674e-06, "loss": 0.0046, "step": 33450 }, { "grad_norm": 0.10583987087011337, "learning_rate": 7.134303481237492e-06, "loss": 0.0036, "step": 33460 }, { "grad_norm": 0.13010269403457642, "learning_rate": 7.113038225789892e-06, "loss": 0.0071, "step": 33470 }, { "grad_norm": 0.10951512306928635, "learning_rate": 7.091802283176258e-06, "loss": 0.0072, "step": 33480 }, { "grad_norm": 0.08425619453191757, "learning_rate": 7.070595667911162e-06, "loss": 0.0047, "step": 33490 }, { "grad_norm": 0.051724616438150406, "learning_rate": 7.049418394489127e-06, "loss": 0.0042, "step": 33500 }, { "grad_norm": 0.12098324298858643, "learning_rate": 7.028270477384636e-06, "loss": 0.0077, "step": 33510 }, { "grad_norm": 0.08211419731378555, "learning_rate": 7.0071519310520695e-06, "loss": 0.0038, "step": 33520 }, { "grad_norm": 0.10132355988025665, "learning_rate": 6.98606276992575e-06, "loss": 0.0066, "step": 33530 }, { "grad_norm": 0.08260331302881241, "learning_rate": 6.965003008419929e-06, "loss": 0.0038, "step": 33540 }, { "grad_norm": 0.11803293228149414, "learning_rate": 6.943972660928755e-06, "loss": 0.0059, "step": 33550 }, { "grad_norm": 0.06129005551338196, "learning_rate": 6.922971741826273e-06, "loss": 0.0045, "step": 33560 }, { "grad_norm": 0.15276959538459778, "learning_rate": 6.902000265466424e-06, "loss": 0.0048, "step": 33570 }, { "grad_norm": 0.09222974628210068, "learning_rate": 6.881058246182998e-06, "loss": 0.0056, "step": 33580 }, { "grad_norm": 0.15936525166034698, "learning_rate": 6.860145698289666e-06, "loss": 0.0051, "step": 33590 }, { "grad_norm": 0.07040811330080032, "learning_rate": 6.839262636079985e-06, "loss": 0.004, "step": 33600 }, { "grad_norm": 0.11020719259977341, "learning_rate": 6.818409073827292e-06, "loss": 0.0066, "step": 33610 }, { "grad_norm": 0.07589146494865417, "learning_rate": 6.79758502578483e-06, "loss": 0.0049, "step": 33620 }, { "grad_norm": 0.10017842054367065, "learning_rate": 6.776790506185621e-06, "loss": 0.0057, "step": 33630 }, { "grad_norm": 0.1024414524435997, "learning_rate": 6.756025529242533e-06, "loss": 0.0044, "step": 33640 }, { "grad_norm": 0.15209269523620605, "learning_rate": 6.735290109148223e-06, "loss": 0.0064, "step": 33650 }, { "grad_norm": 0.09647950530052185, "learning_rate": 6.714584260075163e-06, "loss": 0.005, "step": 33660 }, { "grad_norm": 0.11329124867916107, "learning_rate": 6.693907996175619e-06, "loss": 0.0047, "step": 33670 }, { "grad_norm": 0.1069544106721878, "learning_rate": 6.673261331581593e-06, "loss": 0.004, "step": 33680 }, { "grad_norm": 0.0967564731836319, "learning_rate": 6.652644280404912e-06, "loss": 0.0067, "step": 33690 }, { "grad_norm": 0.06596735119819641, "learning_rate": 6.6320568567371175e-06, "loss": 0.0037, "step": 33700 }, { "grad_norm": 0.08382901549339294, "learning_rate": 6.611499074649518e-06, "loss": 0.0057, "step": 33710 }, { "grad_norm": 0.07875029742717743, "learning_rate": 6.590970948193187e-06, "loss": 0.0051, "step": 33720 }, { "grad_norm": 0.14188607037067413, "learning_rate": 6.5704724913988825e-06, "loss": 0.0048, "step": 33730 }, { "grad_norm": 0.10099786520004272, "learning_rate": 6.550003718277109e-06, "loss": 0.0037, "step": 33740 }, { "grad_norm": 0.06612001359462738, "learning_rate": 6.529564642818087e-06, "loss": 0.0049, "step": 33750 }, { "grad_norm": 0.06802347302436829, "learning_rate": 6.50915527899173e-06, "loss": 0.0043, "step": 33760 }, { "grad_norm": 0.06743888556957245, "learning_rate": 6.4887756407476575e-06, "loss": 0.0055, "step": 33770 }, { "grad_norm": 0.10289239883422852, "learning_rate": 6.468425742015149e-06, "loss": 0.0049, "step": 33780 }, { "grad_norm": 0.10280665755271912, "learning_rate": 6.448105596703152e-06, "loss": 0.0055, "step": 33790 }, { "grad_norm": 0.10086096823215485, "learning_rate": 6.427815218700317e-06, "loss": 0.0064, "step": 33800 }, { "grad_norm": 0.1066555380821228, "learning_rate": 6.407554621874911e-06, "loss": 0.0051, "step": 33810 }, { "grad_norm": 0.09929776191711426, "learning_rate": 6.3873238200748864e-06, "loss": 0.0034, "step": 33820 }, { "grad_norm": 0.140473410487175, "learning_rate": 6.367122827127775e-06, "loss": 0.0063, "step": 33830 }, { "grad_norm": 0.13274450600147247, "learning_rate": 6.346951656840777e-06, "loss": 0.0045, "step": 33840 }, { "grad_norm": 0.14835694432258606, "learning_rate": 6.326810323000698e-06, "loss": 0.0055, "step": 33850 }, { "grad_norm": 0.06616168469190598, "learning_rate": 6.306698839373948e-06, "loss": 0.0038, "step": 33860 }, { "grad_norm": 0.06888452917337418, "learning_rate": 6.286617219706547e-06, "loss": 0.0042, "step": 33870 }, { "grad_norm": 0.12533091008663177, "learning_rate": 6.26656547772409e-06, "loss": 0.0049, "step": 33880 }, { "grad_norm": 0.16111081838607788, "learning_rate": 6.246543627131734e-06, "loss": 0.0055, "step": 33890 }, { "grad_norm": 0.14020688831806183, "learning_rate": 6.22655168161424e-06, "loss": 0.0068, "step": 33900 }, { "grad_norm": 0.10487232357263565, "learning_rate": 6.20658965483591e-06, "loss": 0.0036, "step": 33910 }, { "grad_norm": 0.09656932204961777, "learning_rate": 6.186657560440612e-06, "loss": 0.0057, "step": 33920 }, { "grad_norm": 0.19734227657318115, "learning_rate": 6.166755412051734e-06, "loss": 0.0071, "step": 33930 }, { "grad_norm": 0.09266480803489685, "learning_rate": 6.1468832232721985e-06, "loss": 0.0048, "step": 33940 }, { "grad_norm": 0.11634895205497742, "learning_rate": 6.127041007684476e-06, "loss": 0.0043, "step": 33950 }, { "grad_norm": 0.09962182492017746, "learning_rate": 6.107228778850527e-06, "loss": 0.0058, "step": 33960 }, { "grad_norm": 0.16108159720897675, "learning_rate": 6.087446550311837e-06, "loss": 0.006, "step": 33970 }, { "grad_norm": 0.18780864775180817, "learning_rate": 6.067694335589353e-06, "loss": 0.0066, "step": 33980 }, { "grad_norm": 0.15757259726524353, "learning_rate": 6.047972148183528e-06, "loss": 0.0047, "step": 33990 }, { "grad_norm": 0.11747189611196518, "learning_rate": 6.0282800015742936e-06, "loss": 0.0061, "step": 34000 }, { "grad_norm": 0.07562675327062607, "learning_rate": 6.008617909221054e-06, "loss": 0.0038, "step": 34010 }, { "grad_norm": 0.12842826545238495, "learning_rate": 5.988985884562664e-06, "loss": 0.0052, "step": 34020 }, { "grad_norm": 0.09134622663259506, "learning_rate": 5.969383941017414e-06, "loss": 0.0069, "step": 34030 }, { "grad_norm": 0.10233454406261444, "learning_rate": 5.949812091983059e-06, "loss": 0.0048, "step": 34040 }, { "grad_norm": 0.13609090447425842, "learning_rate": 5.9302703508367695e-06, "loss": 0.0047, "step": 34050 }, { "grad_norm": 0.11776399612426758, "learning_rate": 5.910758730935157e-06, "loss": 0.0077, "step": 34060 }, { "grad_norm": 0.22085826098918915, "learning_rate": 5.8912772456142016e-06, "loss": 0.0062, "step": 34070 }, { "grad_norm": 0.11197703331708908, "learning_rate": 5.871825908189338e-06, "loss": 0.0056, "step": 34080 }, { "grad_norm": 0.11022434383630753, "learning_rate": 5.852404731955352e-06, "loss": 0.0057, "step": 34090 }, { "grad_norm": 0.08063757419586182, "learning_rate": 5.833013730186443e-06, "loss": 0.0046, "step": 34100 }, { "grad_norm": 0.1551254540681839, "learning_rate": 5.813652916136181e-06, "loss": 0.0069, "step": 34110 }, { "grad_norm": 0.10377125442028046, "learning_rate": 5.794322303037508e-06, "loss": 0.0058, "step": 34120 }, { "grad_norm": 0.15515172481536865, "learning_rate": 5.775021904102701e-06, "loss": 0.0068, "step": 34130 }, { "grad_norm": 0.12066097557544708, "learning_rate": 5.755751732523401e-06, "loss": 0.0038, "step": 34140 }, { "grad_norm": 0.09500934928655624, "learning_rate": 5.736511801470595e-06, "loss": 0.005, "step": 34150 }, { "grad_norm": 0.13727621734142303, "learning_rate": 5.717302124094603e-06, "loss": 0.0039, "step": 34160 }, { "grad_norm": 0.07410220801830292, "learning_rate": 5.6981227135250396e-06, "loss": 0.0042, "step": 34170 }, { "grad_norm": 0.11832693219184875, "learning_rate": 5.678973582870867e-06, "loss": 0.0059, "step": 34180 }, { "grad_norm": 0.11233044415712357, "learning_rate": 5.659854745220317e-06, "loss": 0.0049, "step": 34190 }, { "grad_norm": 0.19351589679718018, "learning_rate": 5.640766213640947e-06, "loss": 0.0042, "step": 34200 }, { "grad_norm": 0.11501949280500412, "learning_rate": 5.6217080011795934e-06, "loss": 0.0035, "step": 34210 }, { "grad_norm": 0.1201326847076416, "learning_rate": 5.602680120862364e-06, "loss": 0.0044, "step": 34220 }, { "grad_norm": 0.1467662900686264, "learning_rate": 5.583682585694616e-06, "loss": 0.0048, "step": 34230 }, { "grad_norm": 0.14024563133716583, "learning_rate": 5.564715408661009e-06, "loss": 0.0057, "step": 34240 }, { "grad_norm": 0.06532196700572968, "learning_rate": 5.5457786027254265e-06, "loss": 0.0049, "step": 34250 }, { "grad_norm": 0.1413997858762741, "learning_rate": 5.526872180830977e-06, "loss": 0.0074, "step": 34260 }, { "grad_norm": 0.15122584998607635, "learning_rate": 5.507996155900041e-06, "loss": 0.0049, "step": 34270 }, { "grad_norm": 0.13332870602607727, "learning_rate": 5.489150540834203e-06, "loss": 0.0062, "step": 34280 }, { "grad_norm": 0.15213054418563843, "learning_rate": 5.47033534851425e-06, "loss": 0.0054, "step": 34290 }, { "grad_norm": 0.10172738134860992, "learning_rate": 5.451550591800198e-06, "loss": 0.0044, "step": 34300 }, { "grad_norm": 0.14203979074954987, "learning_rate": 5.43279628353125e-06, "loss": 0.0038, "step": 34310 }, { "grad_norm": 0.13386023044586182, "learning_rate": 5.4140724365258075e-06, "loss": 0.0058, "step": 34320 }, { "grad_norm": 0.055415213108062744, "learning_rate": 5.395379063581424e-06, "loss": 0.0047, "step": 34330 }, { "grad_norm": 0.09513407200574875, "learning_rate": 5.376716177474866e-06, "loss": 0.0042, "step": 34340 }, { "grad_norm": 0.12310820817947388, "learning_rate": 5.358083790962015e-06, "loss": 0.0054, "step": 34350 }, { "grad_norm": 0.1496109813451767, "learning_rate": 5.33948191677795e-06, "loss": 0.0053, "step": 34360 }, { "grad_norm": 0.137562096118927, "learning_rate": 5.320910567636866e-06, "loss": 0.0045, "step": 34370 }, { "grad_norm": 0.13121691346168518, "learning_rate": 5.302369756232123e-06, "loss": 0.0048, "step": 34380 }, { "grad_norm": 0.24936576187610626, "learning_rate": 5.283859495236165e-06, "loss": 0.0079, "step": 34390 }, { "grad_norm": 0.13865643739700317, "learning_rate": 5.265379797300602e-06, "loss": 0.0039, "step": 34400 }, { "grad_norm": 0.08741313219070435, "learning_rate": 5.246930675056117e-06, "loss": 0.0043, "step": 34410 }, { "grad_norm": 0.11013752222061157, "learning_rate": 5.228512141112535e-06, "loss": 0.0042, "step": 34420 }, { "grad_norm": 0.1367962509393692, "learning_rate": 5.21012420805872e-06, "loss": 0.0048, "step": 34430 }, { "grad_norm": 0.10910593718290329, "learning_rate": 5.191766888462679e-06, "loss": 0.0037, "step": 34440 }, { "grad_norm": 0.16480113565921783, "learning_rate": 5.173440194871437e-06, "loss": 0.0059, "step": 34450 }, { "grad_norm": 0.11023473739624023, "learning_rate": 5.155144139811141e-06, "loss": 0.0047, "step": 34460 }, { "grad_norm": 0.057319168001413345, "learning_rate": 5.13687873578696e-06, "loss": 0.0041, "step": 34470 }, { "grad_norm": 0.10883167386054993, "learning_rate": 5.11864399528314e-06, "loss": 0.0055, "step": 34480 }, { "grad_norm": 0.13612736761569977, "learning_rate": 5.1004399307629346e-06, "loss": 0.0048, "step": 34490 }, { "grad_norm": 0.07907218486070633, "learning_rate": 5.082266554668652e-06, "loss": 0.005, "step": 34500 }, { "grad_norm": 0.12124504894018173, "learning_rate": 5.06412387942164e-06, "loss": 0.0057, "step": 34510 }, { "grad_norm": 0.14113007485866547, "learning_rate": 5.046011917422239e-06, "loss": 0.0059, "step": 34520 }, { "grad_norm": 0.1228959932923317, "learning_rate": 5.0279306810498026e-06, "loss": 0.0048, "step": 34530 }, { "grad_norm": 0.09710095822811127, "learning_rate": 5.009880182662679e-06, "loss": 0.0045, "step": 34540 }, { "grad_norm": 0.07618299126625061, "learning_rate": 4.991860434598217e-06, "loss": 0.0048, "step": 34550 }, { "grad_norm": 0.1101863831281662, "learning_rate": 4.973871449172746e-06, "loss": 0.0045, "step": 34560 }, { "grad_norm": 0.07157514244318008, "learning_rate": 4.9559132386815695e-06, "loss": 0.0056, "step": 34570 }, { "grad_norm": 0.1043379157781601, "learning_rate": 4.93798581539896e-06, "loss": 0.0076, "step": 34580 }, { "grad_norm": 0.11566462367773056, "learning_rate": 4.920089191578131e-06, "loss": 0.0042, "step": 34590 }, { "grad_norm": 0.15026305615901947, "learning_rate": 4.90222337945126e-06, "loss": 0.0033, "step": 34600 }, { "grad_norm": 0.06468398123979568, "learning_rate": 4.8843883912294605e-06, "loss": 0.0054, "step": 34610 }, { "grad_norm": 0.12884563207626343, "learning_rate": 4.866584239102795e-06, "loss": 0.0052, "step": 34620 }, { "grad_norm": 0.08339788764715195, "learning_rate": 4.8488109352402164e-06, "loss": 0.0062, "step": 34630 }, { "grad_norm": 0.07716327905654907, "learning_rate": 4.831068491789603e-06, "loss": 0.0055, "step": 34640 }, { "grad_norm": 0.19545142352581024, "learning_rate": 4.81335692087776e-06, "loss": 0.0078, "step": 34650 }, { "grad_norm": 0.09609583020210266, "learning_rate": 4.795676234610374e-06, "loss": 0.0038, "step": 34660 }, { "grad_norm": 0.12717682123184204, "learning_rate": 4.778026445072037e-06, "loss": 0.0043, "step": 34670 }, { "grad_norm": 0.10957561433315277, "learning_rate": 4.760407564326208e-06, "loss": 0.0058, "step": 34680 }, { "grad_norm": 0.08948840945959091, "learning_rate": 4.742819604415222e-06, "loss": 0.0045, "step": 34690 }, { "grad_norm": 0.13764864206314087, "learning_rate": 4.725262577360284e-06, "loss": 0.0058, "step": 34700 }, { "grad_norm": 0.13622836768627167, "learning_rate": 4.707736495161458e-06, "loss": 0.0052, "step": 34710 }, { "grad_norm": 0.20179606974124908, "learning_rate": 4.690241369797671e-06, "loss": 0.0052, "step": 34720 }, { "grad_norm": 0.15211491286754608, "learning_rate": 4.672777213226659e-06, "loss": 0.004, "step": 34730 }, { "grad_norm": 0.13392961025238037, "learning_rate": 4.655344037385001e-06, "loss": 0.0064, "step": 34740 }, { "grad_norm": 0.08985096216201782, "learning_rate": 4.6379418541881215e-06, "loss": 0.0057, "step": 34750 }, { "grad_norm": 0.07436027377843857, "learning_rate": 4.620570675530245e-06, "loss": 0.0043, "step": 34760 }, { "grad_norm": 0.14742395281791687, "learning_rate": 4.603230513284407e-06, "loss": 0.0042, "step": 34770 }, { "grad_norm": 0.07614313066005707, "learning_rate": 4.58592137930246e-06, "loss": 0.0049, "step": 34780 }, { "grad_norm": 0.11488420516252518, "learning_rate": 4.568643285415009e-06, "loss": 0.0044, "step": 34790 }, { "grad_norm": 0.10107734054327011, "learning_rate": 4.551396243431483e-06, "loss": 0.0042, "step": 34800 }, { "grad_norm": 0.16408540308475494, "learning_rate": 4.534180265140075e-06, "loss": 0.0059, "step": 34810 }, { "grad_norm": 0.16414599120616913, "learning_rate": 4.5169953623077356e-06, "loss": 0.004, "step": 34820 }, { "grad_norm": 0.08657010644674301, "learning_rate": 4.499841546680189e-06, "loss": 0.0057, "step": 34830 }, { "grad_norm": 0.13889996707439423, "learning_rate": 4.4827188299819025e-06, "loss": 0.004, "step": 34840 }, { "grad_norm": 0.09171462059020996, "learning_rate": 4.465627223916086e-06, "loss": 0.0041, "step": 34850 }, { "grad_norm": 0.10151714086532593, "learning_rate": 4.448566740164706e-06, "loss": 0.0042, "step": 34860 }, { "grad_norm": 0.0848136842250824, "learning_rate": 4.431537390388429e-06, "loss": 0.007, "step": 34870 }, { "grad_norm": 0.056932829320430756, "learning_rate": 4.414539186226674e-06, "loss": 0.0046, "step": 34880 }, { "grad_norm": 0.09101065248250961, "learning_rate": 4.397572139297529e-06, "loss": 0.006, "step": 34890 }, { "grad_norm": 0.0880657285451889, "learning_rate": 4.380636261197829e-06, "loss": 0.0038, "step": 34900 }, { "grad_norm": 0.12138768285512924, "learning_rate": 4.36373156350307e-06, "loss": 0.004, "step": 34910 }, { "grad_norm": 0.09896478056907654, "learning_rate": 4.3468580577674665e-06, "loss": 0.0042, "step": 34920 }, { "grad_norm": 0.08914708346128464, "learning_rate": 4.330015755523908e-06, "loss": 0.0047, "step": 34930 }, { "grad_norm": 0.09909263998270035, "learning_rate": 4.313204668283926e-06, "loss": 0.0058, "step": 34940 }, { "grad_norm": 0.13940992951393127, "learning_rate": 4.296424807537752e-06, "loss": 0.0043, "step": 34950 }, { "grad_norm": 0.12182299047708511, "learning_rate": 4.279676184754267e-06, "loss": 0.0043, "step": 34960 }, { "grad_norm": 0.06790286302566528, "learning_rate": 4.262958811380985e-06, "loss": 0.004, "step": 34970 }, { "grad_norm": 0.16336150467395782, "learning_rate": 4.246272698844095e-06, "loss": 0.0074, "step": 34980 }, { "grad_norm": 0.10392053425312042, "learning_rate": 4.229617858548374e-06, "loss": 0.0052, "step": 34990 }, { "grad_norm": 0.13284200429916382, "learning_rate": 4.212994301877271e-06, "loss": 0.0052, "step": 35000 }, { "grad_norm": 0.16456374526023865, "learning_rate": 4.196402040192804e-06, "loss": 0.009, "step": 35010 }, { "grad_norm": 0.1495591253042221, "learning_rate": 4.179841084835639e-06, "loss": 0.005, "step": 35020 }, { "grad_norm": 0.10048934817314148, "learning_rate": 4.163311447125046e-06, "loss": 0.0056, "step": 35030 }, { "grad_norm": 0.10619179159402847, "learning_rate": 4.146813138358851e-06, "loss": 0.005, "step": 35040 }, { "grad_norm": 0.13780441880226135, "learning_rate": 4.130346169813509e-06, "loss": 0.0045, "step": 35050 }, { "grad_norm": 0.07254817336797714, "learning_rate": 4.113910552744038e-06, "loss": 0.0038, "step": 35060 }, { "grad_norm": 0.16326658427715302, "learning_rate": 4.097506298384024e-06, "loss": 0.0057, "step": 35070 }, { "grad_norm": 0.11760199069976807, "learning_rate": 4.081133417945626e-06, "loss": 0.0085, "step": 35080 }, { "grad_norm": 0.15527774393558502, "learning_rate": 4.0647919226195484e-06, "loss": 0.0056, "step": 35090 }, { "grad_norm": 0.19906750321388245, "learning_rate": 4.04848182357504e-06, "loss": 0.0052, "step": 35100 }, { "grad_norm": 0.10753054916858673, "learning_rate": 4.032203131959911e-06, "loss": 0.0081, "step": 35110 }, { "grad_norm": 0.16797561943531036, "learning_rate": 4.015955858900483e-06, "loss": 0.0053, "step": 35120 }, { "grad_norm": 0.11972041428089142, "learning_rate": 3.999740015501641e-06, "loss": 0.0052, "step": 35130 }, { "grad_norm": 0.10878206044435501, "learning_rate": 3.9835556128467236e-06, "loss": 0.0044, "step": 35140 }, { "grad_norm": 0.1209234967827797, "learning_rate": 3.967402661997644e-06, "loss": 0.0051, "step": 35150 }, { "grad_norm": 0.1392289251089096, "learning_rate": 3.951281173994781e-06, "loss": 0.0042, "step": 35160 }, { "grad_norm": 0.12393234670162201, "learning_rate": 3.935191159857016e-06, "loss": 0.0042, "step": 35170 }, { "grad_norm": 0.08644958585500717, "learning_rate": 3.919132630581746e-06, "loss": 0.0068, "step": 35180 }, { "grad_norm": 0.1250041276216507, "learning_rate": 3.903105597144802e-06, "loss": 0.0064, "step": 35190 }, { "grad_norm": 0.10416384041309357, "learning_rate": 3.887110070500505e-06, "loss": 0.0078, "step": 35200 }, { "grad_norm": 0.07262086868286133, "learning_rate": 3.871146061581654e-06, "loss": 0.0046, "step": 35210 }, { "grad_norm": 0.1684284806251526, "learning_rate": 3.855213581299505e-06, "loss": 0.0047, "step": 35220 }, { "grad_norm": 0.1434158980846405, "learning_rate": 3.839312640543763e-06, "loss": 0.0048, "step": 35230 }, { "grad_norm": 0.15877065062522888, "learning_rate": 3.823443250182551e-06, "loss": 0.0055, "step": 35240 }, { "grad_norm": 0.0753139853477478, "learning_rate": 3.80760542106246e-06, "loss": 0.0051, "step": 35250 }, { "grad_norm": 0.07152330875396729, "learning_rate": 3.7917991640084972e-06, "loss": 0.0053, "step": 35260 }, { "grad_norm": 0.13680584728717804, "learning_rate": 3.7760244898240925e-06, "loss": 0.0043, "step": 35270 }, { "grad_norm": 0.051757827401161194, "learning_rate": 3.760281409291078e-06, "loss": 0.0035, "step": 35280 }, { "grad_norm": 0.06112319231033325, "learning_rate": 3.7445699331697113e-06, "loss": 0.0056, "step": 35290 }, { "grad_norm": 0.07356466352939606, "learning_rate": 3.7288900721986262e-06, "loss": 0.0056, "step": 35300 }, { "grad_norm": 0.13062018156051636, "learning_rate": 3.7132418370948663e-06, "loss": 0.0048, "step": 35310 }, { "grad_norm": 0.13838928937911987, "learning_rate": 3.6976252385538535e-06, "loss": 0.0057, "step": 35320 }, { "grad_norm": 0.15963995456695557, "learning_rate": 3.6820402872493832e-06, "loss": 0.0055, "step": 35330 }, { "grad_norm": 0.06626967340707779, "learning_rate": 3.6664869938336343e-06, "loss": 0.0054, "step": 35340 }, { "grad_norm": 0.09415558725595474, "learning_rate": 3.650965368937115e-06, "loss": 0.0053, "step": 35350 }, { "grad_norm": 0.10585330426692963, "learning_rate": 3.6354754231687214e-06, "loss": 0.0041, "step": 35360 }, { "grad_norm": 0.08337967097759247, "learning_rate": 3.620017167115697e-06, "loss": 0.008, "step": 35370 }, { "grad_norm": 0.09175460040569305, "learning_rate": 3.604590611343589e-06, "loss": 0.0036, "step": 35380 }, { "grad_norm": 0.12061246484518051, "learning_rate": 3.589195766396325e-06, "loss": 0.0057, "step": 35390 }, { "grad_norm": 0.17275260388851166, "learning_rate": 3.573832642796121e-06, "loss": 0.0051, "step": 35400 }, { "grad_norm": 0.07199397683143616, "learning_rate": 3.5585012510435336e-06, "loss": 0.0041, "step": 35410 }, { "grad_norm": 0.14206917583942413, "learning_rate": 3.5432016016174184e-06, "loss": 0.0061, "step": 35420 }, { "grad_norm": 0.06618551909923553, "learning_rate": 3.5279337049749506e-06, "loss": 0.005, "step": 35430 }, { "grad_norm": 0.15268369019031525, "learning_rate": 3.5126975715515985e-06, "loss": 0.0075, "step": 35440 }, { "grad_norm": 0.11530357599258423, "learning_rate": 3.4974932117610994e-06, "loss": 0.0052, "step": 35450 }, { "grad_norm": 0.06574276089668274, "learning_rate": 3.4823206359955065e-06, "loss": 0.0059, "step": 35460 }, { "grad_norm": 0.11411336064338684, "learning_rate": 3.4671798546251133e-06, "loss": 0.0056, "step": 35470 }, { "grad_norm": 0.09874624758958817, "learning_rate": 3.4520708779985088e-06, "loss": 0.005, "step": 35480 }, { "grad_norm": 0.11763589084148407, "learning_rate": 3.4369937164425493e-06, "loss": 0.0032, "step": 35490 }, { "grad_norm": 0.06172323226928711, "learning_rate": 3.421948380262313e-06, "loss": 0.0064, "step": 35500 }, { "grad_norm": 0.13009874522686005, "learning_rate": 3.406934879741158e-06, "loss": 0.0076, "step": 35510 }, { "grad_norm": 0.0747419074177742, "learning_rate": 3.391953225140665e-06, "loss": 0.0055, "step": 35520 }, { "grad_norm": 0.11572299897670746, "learning_rate": 3.3770034267006533e-06, "loss": 0.0052, "step": 35530 }, { "grad_norm": 0.16373084485530853, "learning_rate": 3.362085494639178e-06, "loss": 0.0052, "step": 35540 }, { "grad_norm": 0.16812238097190857, "learning_rate": 3.3471994391524986e-06, "loss": 0.0042, "step": 35550 }, { "grad_norm": 0.0668003112077713, "learning_rate": 3.3323452704150825e-06, "loss": 0.0053, "step": 35560 }, { "grad_norm": 0.11536645889282227, "learning_rate": 3.3175229985796253e-06, "loss": 0.0053, "step": 35570 }, { "grad_norm": 0.1350553035736084, "learning_rate": 3.3027326337770115e-06, "loss": 0.0041, "step": 35580 }, { "grad_norm": 0.15079349279403687, "learning_rate": 3.2879741861163216e-06, "loss": 0.005, "step": 35590 }, { "grad_norm": 0.09646932035684586, "learning_rate": 3.2732476656847976e-06, "loss": 0.0048, "step": 35600 }, { "grad_norm": 0.15102384984493256, "learning_rate": 3.2585530825478883e-06, "loss": 0.0054, "step": 35610 }, { "grad_norm": 0.1256898045539856, "learning_rate": 3.243890446749204e-06, "loss": 0.006, "step": 35620 }, { "grad_norm": 0.13537977635860443, "learning_rate": 3.2292597683105107e-06, "loss": 0.0069, "step": 35630 }, { "grad_norm": 0.19902817904949188, "learning_rate": 3.214661057231755e-06, "loss": 0.0061, "step": 35640 }, { "grad_norm": 0.1850621998310089, "learning_rate": 3.2000943234910097e-06, "loss": 0.0045, "step": 35650 }, { "grad_norm": 0.0723680704832077, "learning_rate": 3.1855595770444835e-06, "loss": 0.0043, "step": 35660 }, { "grad_norm": 0.06776819378137589, "learning_rate": 3.1710568278265574e-06, "loss": 0.0045, "step": 35670 }, { "grad_norm": 0.11554140597581863, "learning_rate": 3.156586085749719e-06, "loss": 0.0053, "step": 35680 }, { "grad_norm": 0.18066905438899994, "learning_rate": 3.1421473607045893e-06, "loss": 0.0063, "step": 35690 }, { "grad_norm": 0.09255938977003098, "learning_rate": 3.127740662559886e-06, "loss": 0.0054, "step": 35700 }, { "grad_norm": 0.11772732436656952, "learning_rate": 3.1133660011624655e-06, "loss": 0.0044, "step": 35710 }, { "grad_norm": 0.10229182988405228, "learning_rate": 3.099023386337263e-06, "loss": 0.0073, "step": 35720 }, { "grad_norm": 0.10076887905597687, "learning_rate": 3.0847128278873317e-06, "loss": 0.0058, "step": 35730 }, { "grad_norm": 0.11874886602163315, "learning_rate": 3.0704343355938027e-06, "loss": 0.0052, "step": 35740 }, { "grad_norm": 0.0979604572057724, "learning_rate": 3.056187919215886e-06, "loss": 0.0046, "step": 35750 }, { "grad_norm": 0.08389444649219513, "learning_rate": 3.041973588490865e-06, "loss": 0.0043, "step": 35760 }, { "grad_norm": 0.12013111263513565, "learning_rate": 3.0277913531341128e-06, "loss": 0.006, "step": 35770 }, { "grad_norm": 0.08439846336841583, "learning_rate": 3.0136412228390476e-06, "loss": 0.0034, "step": 35780 }, { "grad_norm": 0.08326462656259537, "learning_rate": 2.999523207277155e-06, "loss": 0.0034, "step": 35790 }, { "grad_norm": 0.1675816923379898, "learning_rate": 2.9854373160979552e-06, "loss": 0.0056, "step": 35800 }, { "grad_norm": 0.12616175413131714, "learning_rate": 2.9713835589290362e-06, "loss": 0.0066, "step": 35810 }, { "grad_norm": 0.112790547311306, "learning_rate": 2.9573619453759913e-06, "loss": 0.0037, "step": 35820 }, { "grad_norm": 0.09907086193561554, "learning_rate": 2.943372485022489e-06, "loss": 0.0049, "step": 35830 }, { "grad_norm": 0.13827526569366455, "learning_rate": 2.9294151874301623e-06, "loss": 0.0047, "step": 35840 }, { "grad_norm": 0.05799366533756256, "learning_rate": 2.9154900621387093e-06, "loss": 0.0043, "step": 35850 }, { "grad_norm": 0.09102669358253479, "learning_rate": 2.9015971186658164e-06, "loss": 0.0031, "step": 35860 }, { "grad_norm": 0.07409216463565826, "learning_rate": 2.887736366507182e-06, "loss": 0.0068, "step": 35870 }, { "grad_norm": 0.12324732542037964, "learning_rate": 2.873907815136501e-06, "loss": 0.0035, "step": 35880 }, { "grad_norm": 0.11728942394256592, "learning_rate": 2.8601114740054556e-06, "loss": 0.0045, "step": 35890 }, { "grad_norm": 0.1584140658378601, "learning_rate": 2.8463473525437155e-06, "loss": 0.0043, "step": 35900 }, { "grad_norm": 0.16613620519638062, "learning_rate": 2.8326154601589216e-06, "loss": 0.0065, "step": 35910 }, { "grad_norm": 0.11805953085422516, "learning_rate": 2.818915806236705e-06, "loss": 0.0033, "step": 35920 }, { "grad_norm": 0.10587016493082047, "learning_rate": 2.8052484001406478e-06, "loss": 0.0047, "step": 35930 }, { "grad_norm": 0.14422613382339478, "learning_rate": 2.791613251212277e-06, "loss": 0.004, "step": 35940 }, { "grad_norm": 0.06426966935396194, "learning_rate": 2.7780103687711166e-06, "loss": 0.0045, "step": 35950 }, { "grad_norm": 0.07248946279287338, "learning_rate": 2.764439762114579e-06, "loss": 0.004, "step": 35960 }, { "grad_norm": 0.18750639259815216, "learning_rate": 2.750901440518061e-06, "loss": 0.0054, "step": 35970 }, { "grad_norm": 0.11788904666900635, "learning_rate": 2.7373954132348776e-06, "loss": 0.005, "step": 35980 }, { "grad_norm": 0.1413707733154297, "learning_rate": 2.7239216894962784e-06, "loss": 0.0057, "step": 35990 }, { "grad_norm": 0.08658997714519501, "learning_rate": 2.710480278511407e-06, "loss": 0.0056, "step": 36000 }, { "grad_norm": 0.08752218633890152, "learning_rate": 2.6970711894673483e-06, "loss": 0.0051, "step": 36010 }, { "grad_norm": 0.1793542057275772, "learning_rate": 2.6836944315291047e-06, "loss": 0.0045, "step": 36020 }, { "grad_norm": 0.1065191775560379, "learning_rate": 2.67035001383954e-06, "loss": 0.007, "step": 36030 }, { "grad_norm": 0.07144483178853989, "learning_rate": 2.6570379455194527e-06, "loss": 0.0057, "step": 36040 }, { "grad_norm": 0.06627168506383896, "learning_rate": 2.6437582356675206e-06, "loss": 0.0043, "step": 36050 }, { "grad_norm": 0.055746011435985565, "learning_rate": 2.6305108933602775e-06, "loss": 0.0057, "step": 36060 }, { "grad_norm": 0.11330091208219528, "learning_rate": 2.6172959276521803e-06, "loss": 0.0047, "step": 36070 }, { "grad_norm": 0.14501334726810455, "learning_rate": 2.6041133475755153e-06, "loss": 0.0049, "step": 36080 }, { "grad_norm": 0.12262345105409622, "learning_rate": 2.590963162140475e-06, "loss": 0.0056, "step": 36090 }, { "grad_norm": 0.1351185441017151, "learning_rate": 2.5778453803350587e-06, "loss": 0.0071, "step": 36100 }, { "grad_norm": 0.04950525239109993, "learning_rate": 2.5647600111251723e-06, "loss": 0.0045, "step": 36110 }, { "grad_norm": 0.07128050178289413, "learning_rate": 2.551707063454528e-06, "loss": 0.0051, "step": 36120 }, { "grad_norm": 0.11979761719703674, "learning_rate": 2.5386865462446897e-06, "loss": 0.0045, "step": 36130 }, { "grad_norm": 0.08780655264854431, "learning_rate": 2.52569846839506e-06, "loss": 0.0033, "step": 36140 }, { "grad_norm": 0.09631843864917755, "learning_rate": 2.512742838782883e-06, "loss": 0.0046, "step": 36150 }, { "grad_norm": 0.04333241283893585, "learning_rate": 2.4998196662631924e-06, "loss": 0.0037, "step": 36160 }, { "grad_norm": 0.09079335629940033, "learning_rate": 2.4869289596688604e-06, "loss": 0.0066, "step": 36170 }, { "grad_norm": 0.09292557090520859, "learning_rate": 2.4740707278105567e-06, "loss": 0.0041, "step": 36180 }, { "grad_norm": 0.12682722508907318, "learning_rate": 2.461244979476779e-06, "loss": 0.0043, "step": 36190 }, { "grad_norm": 0.10848325490951538, "learning_rate": 2.4484517234337866e-06, "loss": 0.0036, "step": 36200 }, { "grad_norm": 0.07283230870962143, "learning_rate": 2.4356909684256634e-06, "loss": 0.0035, "step": 36210 }, { "grad_norm": 0.08900574594736099, "learning_rate": 2.4229627231742545e-06, "loss": 0.0036, "step": 36220 }, { "grad_norm": 0.13052313029766083, "learning_rate": 2.4102669963791956e-06, "loss": 0.0069, "step": 36230 }, { "grad_norm": 0.184688001871109, "learning_rate": 2.3976037967179065e-06, "loss": 0.0052, "step": 36240 }, { "grad_norm": 0.054574426263570786, "learning_rate": 2.3849731328455582e-06, "loss": 0.0045, "step": 36250 }, { "grad_norm": 0.12634873390197754, "learning_rate": 2.3723750133950894e-06, "loss": 0.005, "step": 36260 }, { "grad_norm": 0.09607597440481186, "learning_rate": 2.3598094469771957e-06, "loss": 0.0051, "step": 36270 }, { "grad_norm": 0.1427236646413803, "learning_rate": 2.347276442180324e-06, "loss": 0.0053, "step": 36280 }, { "grad_norm": 0.12609142065048218, "learning_rate": 2.3347760075706715e-06, "loss": 0.0052, "step": 36290 }, { "grad_norm": 0.12678201496601105, "learning_rate": 2.3223081516921595e-06, "loss": 0.0043, "step": 36300 }, { "grad_norm": 0.1163204237818718, "learning_rate": 2.3098728830664494e-06, "loss": 0.0045, "step": 36310 }, { "grad_norm": 0.07335802167654037, "learning_rate": 2.2974702101929314e-06, "loss": 0.0037, "step": 36320 }, { "grad_norm": 0.14115102589130402, "learning_rate": 2.2851001415487185e-06, "loss": 0.0061, "step": 36330 }, { "grad_norm": 0.15069939196109772, "learning_rate": 2.272762685588631e-06, "loss": 0.0056, "step": 36340 }, { "grad_norm": 0.11287935078144073, "learning_rate": 2.260457850745218e-06, "loss": 0.0044, "step": 36350 }, { "grad_norm": 0.0713861733675003, "learning_rate": 2.2481856454286963e-06, "loss": 0.0033, "step": 36360 }, { "grad_norm": 0.11263541132211685, "learning_rate": 2.235946078027018e-06, "loss": 0.0044, "step": 36370 }, { "grad_norm": 0.19844330847263336, "learning_rate": 2.223739156905802e-06, "loss": 0.0055, "step": 36380 }, { "grad_norm": 0.11400946974754333, "learning_rate": 2.2115648904083796e-06, "loss": 0.0068, "step": 36390 }, { "grad_norm": 0.09543442726135254, "learning_rate": 2.199423286855734e-06, "loss": 0.0037, "step": 36400 }, { "grad_norm": 0.1135159283876419, "learning_rate": 2.1873143545465323e-06, "loss": 0.0042, "step": 36410 }, { "grad_norm": 0.11577396094799042, "learning_rate": 2.1752381017571267e-06, "loss": 0.0036, "step": 36420 }, { "grad_norm": 0.08206740021705627, "learning_rate": 2.1631945367415142e-06, "loss": 0.005, "step": 36430 }, { "grad_norm": 0.08351889252662659, "learning_rate": 2.15118366773136e-06, "loss": 0.0042, "step": 36440 }, { "grad_norm": 0.20890340209007263, "learning_rate": 2.1392055029359926e-06, "loss": 0.0043, "step": 36450 }, { "grad_norm": 0.09160646796226501, "learning_rate": 2.127260050542357e-06, "loss": 0.004, "step": 36460 }, { "grad_norm": 0.1382857710123062, "learning_rate": 2.1153473187150608e-06, "loss": 0.0046, "step": 36470 }, { "grad_norm": 0.05705099180340767, "learning_rate": 2.1034673155963467e-06, "loss": 0.0044, "step": 36480 }, { "grad_norm": 0.16409339010715485, "learning_rate": 2.0916200493060976e-06, "loss": 0.0044, "step": 36490 }, { "grad_norm": 0.14142824709415436, "learning_rate": 2.0798055279417916e-06, "loss": 0.0038, "step": 36500 }, { "grad_norm": 0.09487280249595642, "learning_rate": 2.0680237595785356e-06, "loss": 0.0035, "step": 36510 }, { "grad_norm": 0.04371973127126694, "learning_rate": 2.0562747522690774e-06, "loss": 0.003, "step": 36520 }, { "grad_norm": 0.04857010766863823, "learning_rate": 2.0445585140437373e-06, "loss": 0.0032, "step": 36530 }, { "grad_norm": 0.1327725648880005, "learning_rate": 2.0328750529104602e-06, "loss": 0.0054, "step": 36540 }, { "grad_norm": 0.10299406945705414, "learning_rate": 2.021224376854791e-06, "loss": 0.0042, "step": 36550 }, { "grad_norm": 0.08017078787088394, "learning_rate": 2.0096064938398386e-06, "loss": 0.0037, "step": 36560 }, { "grad_norm": 0.12762725353240967, "learning_rate": 1.998021411806328e-06, "loss": 0.0036, "step": 36570 }, { "grad_norm": 0.09893865883350372, "learning_rate": 1.9864691386725586e-06, "loss": 0.0065, "step": 36580 }, { "grad_norm": 0.12481576949357986, "learning_rate": 1.9749496823343814e-06, "loss": 0.0059, "step": 36590 }, { "grad_norm": 0.11012348532676697, "learning_rate": 1.963463050665254e-06, "loss": 0.0046, "step": 36600 }, { "grad_norm": 0.12605594098567963, "learning_rate": 1.9520092515161682e-06, "loss": 0.0059, "step": 36610 }, { "grad_norm": 0.05240476876497269, "learning_rate": 1.940588292715695e-06, "loss": 0.0053, "step": 36620 }, { "grad_norm": 0.15603280067443848, "learning_rate": 1.929200182069951e-06, "loss": 0.0044, "step": 36630 }, { "grad_norm": 0.12732763588428497, "learning_rate": 1.9178449273625997e-06, "loss": 0.0042, "step": 36640 }, { "grad_norm": 0.1203857883810997, "learning_rate": 1.90652253635486e-06, "loss": 0.006, "step": 36650 }, { "grad_norm": 0.10490109026432037, "learning_rate": 1.8952330167854693e-06, "loss": 0.0036, "step": 36660 }, { "grad_norm": 0.10223348438739777, "learning_rate": 1.8839763763707107e-06, "loss": 0.0042, "step": 36670 }, { "grad_norm": 0.0719238817691803, "learning_rate": 1.8727526228043912e-06, "loss": 0.0044, "step": 36680 }, { "grad_norm": 0.09345142543315887, "learning_rate": 1.8615617637578408e-06, "loss": 0.0058, "step": 36690 }, { "grad_norm": 0.06527514010667801, "learning_rate": 1.850403806879908e-06, "loss": 0.0045, "step": 36700 }, { "grad_norm": 0.09157374501228333, "learning_rate": 1.8392787597969475e-06, "loss": 0.0043, "step": 36710 }, { "grad_norm": 0.16117049753665924, "learning_rate": 1.828186630112827e-06, "loss": 0.0055, "step": 36720 }, { "grad_norm": 0.06795558333396912, "learning_rate": 1.8171274254089098e-06, "loss": 0.004, "step": 36730 }, { "grad_norm": 0.1321815848350525, "learning_rate": 1.806101153244061e-06, "loss": 0.0045, "step": 36740 }, { "grad_norm": 0.06804265081882477, "learning_rate": 1.7951078211546401e-06, "loss": 0.0065, "step": 36750 }, { "grad_norm": 0.07975699007511139, "learning_rate": 1.7841474366544764e-06, "loss": 0.0053, "step": 36760 }, { "grad_norm": 0.06685475260019302, "learning_rate": 1.7732200072348825e-06, "loss": 0.0035, "step": 36770 }, { "grad_norm": 0.07955482602119446, "learning_rate": 1.7623255403646677e-06, "loss": 0.0041, "step": 36780 }, { "grad_norm": 0.10426214337348938, "learning_rate": 1.7514640434900865e-06, "loss": 0.0045, "step": 36790 }, { "grad_norm": 0.11445353180170059, "learning_rate": 1.7406355240348838e-06, "loss": 0.0051, "step": 36800 }, { "grad_norm": 0.05952947959303856, "learning_rate": 1.7298399894002392e-06, "loss": 0.0035, "step": 36810 }, { "grad_norm": 0.15550152957439423, "learning_rate": 1.7190774469647952e-06, "loss": 0.0061, "step": 36820 }, { "grad_norm": 0.12756778299808502, "learning_rate": 1.7083479040846617e-06, "loss": 0.0067, "step": 36830 }, { "grad_norm": 0.13145995140075684, "learning_rate": 1.697651368093367e-06, "loss": 0.0041, "step": 36840 }, { "grad_norm": 0.09605395048856735, "learning_rate": 1.6869878463019017e-06, "loss": 0.0049, "step": 36850 }, { "grad_norm": 0.1165844276547432, "learning_rate": 1.6763573459986858e-06, "loss": 0.0053, "step": 36860 }, { "grad_norm": 0.09455583989620209, "learning_rate": 1.6657598744495516e-06, "loss": 0.0047, "step": 36870 }, { "grad_norm": 0.040522731840610504, "learning_rate": 1.6551954388977774e-06, "loss": 0.0046, "step": 36880 }, { "grad_norm": 0.11605706065893173, "learning_rate": 1.644664046564054e-06, "loss": 0.0036, "step": 36890 }, { "grad_norm": 0.12718476355075836, "learning_rate": 1.634165704646501e-06, "loss": 0.0041, "step": 36900 }, { "grad_norm": 0.15324144065380096, "learning_rate": 1.6237004203206175e-06, "loss": 0.0078, "step": 36910 }, { "grad_norm": 0.0671081468462944, "learning_rate": 1.6132682007393373e-06, "loss": 0.0041, "step": 36920 }, { "grad_norm": 0.056839097291231155, "learning_rate": 1.6028690530329847e-06, "loss": 0.0033, "step": 36930 }, { "grad_norm": 0.1653580218553543, "learning_rate": 1.5925029843092742e-06, "loss": 0.0056, "step": 36940 }, { "grad_norm": 0.15963490307331085, "learning_rate": 1.5821700016533214e-06, "loss": 0.0078, "step": 36950 }, { "grad_norm": 0.07231568545103073, "learning_rate": 1.5718701121276268e-06, "loss": 0.008, "step": 36960 }, { "grad_norm": 0.12657949328422546, "learning_rate": 1.5616033227720483e-06, "loss": 0.0051, "step": 36970 }, { "grad_norm": 0.07309355586767197, "learning_rate": 1.5513696406038558e-06, "loss": 0.0042, "step": 36980 }, { "grad_norm": 0.1831165999174118, "learning_rate": 1.5411690726176653e-06, "loss": 0.0041, "step": 36990 }, { "grad_norm": 0.05630592256784439, "learning_rate": 1.5310016257854832e-06, "loss": 0.0066, "step": 37000 }, { "grad_norm": 0.17086449265480042, "learning_rate": 1.5208673070566504e-06, "loss": 0.0052, "step": 37010 }, { "grad_norm": 0.09837187081575394, "learning_rate": 1.5107661233578818e-06, "loss": 0.0041, "step": 37020 }, { "grad_norm": 0.14766815304756165, "learning_rate": 1.5006980815932437e-06, "loss": 0.0042, "step": 37030 }, { "grad_norm": 0.07596653699874878, "learning_rate": 1.490663188644148e-06, "loss": 0.0071, "step": 37040 }, { "grad_norm": 0.10908591747283936, "learning_rate": 1.4806614513693418e-06, "loss": 0.0052, "step": 37050 }, { "grad_norm": 0.10071621090173721, "learning_rate": 1.4706928766049289e-06, "loss": 0.0059, "step": 37060 }, { "grad_norm": 0.07684821635484695, "learning_rate": 1.4607574711643257e-06, "loss": 0.004, "step": 37070 }, { "grad_norm": 0.08697441220283508, "learning_rate": 1.4508552418382949e-06, "loss": 0.0042, "step": 37080 }, { "grad_norm": 0.12051893025636673, "learning_rate": 1.4409861953949112e-06, "loss": 0.0065, "step": 37090 }, { "grad_norm": 0.11560344696044922, "learning_rate": 1.43115033857959e-06, "loss": 0.0051, "step": 37100 }, { "grad_norm": 0.13750262558460236, "learning_rate": 1.4213476781150258e-06, "loss": 0.0035, "step": 37110 }, { "grad_norm": 0.17384716868400574, "learning_rate": 1.411578220701254e-06, "loss": 0.0034, "step": 37120 }, { "grad_norm": 0.06287524849176407, "learning_rate": 1.4018419730156108e-06, "loss": 0.0045, "step": 37130 }, { "grad_norm": 0.07363676279783249, "learning_rate": 1.3921389417127283e-06, "loss": 0.0034, "step": 37140 }, { "grad_norm": 0.07629311829805374, "learning_rate": 1.3824691334245298e-06, "loss": 0.0046, "step": 37150 }, { "grad_norm": 0.12066977471113205, "learning_rate": 1.3728325547602505e-06, "loss": 0.0041, "step": 37160 }, { "grad_norm": 0.0733271986246109, "learning_rate": 1.3632292123063828e-06, "loss": 0.0057, "step": 37170 }, { "grad_norm": 0.12549471855163574, "learning_rate": 1.3536591126267317e-06, "loss": 0.0054, "step": 37180 }, { "grad_norm": 0.16679319739341736, "learning_rate": 1.3441222622623761e-06, "loss": 0.0077, "step": 37190 }, { "grad_norm": 0.06072230637073517, "learning_rate": 1.3346186677316575e-06, "loss": 0.0053, "step": 37200 }, { "grad_norm": 0.10501459240913391, "learning_rate": 1.3251483355301852e-06, "loss": 0.0072, "step": 37210 }, { "grad_norm": 0.08039948344230652, "learning_rate": 1.315711272130854e-06, "loss": 0.0051, "step": 37220 }, { "grad_norm": 0.08853708207607269, "learning_rate": 1.306307483983804e-06, "loss": 0.0061, "step": 37230 }, { "grad_norm": 0.16509504616260529, "learning_rate": 1.2969369775164275e-06, "loss": 0.0034, "step": 37240 }, { "grad_norm": 0.14516223967075348, "learning_rate": 1.28759975913339e-06, "loss": 0.0045, "step": 37250 }, { "grad_norm": 0.06776583194732666, "learning_rate": 1.2782958352165863e-06, "loss": 0.004, "step": 37260 }, { "grad_norm": 0.14542214572429657, "learning_rate": 1.2690252121251633e-06, "loss": 0.004, "step": 37270 }, { "grad_norm": 0.17441938817501068, "learning_rate": 1.2597878961954968e-06, "loss": 0.0059, "step": 37280 }, { "grad_norm": 0.10462455451488495, "learning_rate": 1.250583893741214e-06, "loss": 0.0029, "step": 37290 }, { "grad_norm": 0.14332644641399384, "learning_rate": 1.2414132110531607e-06, "loss": 0.0048, "step": 37300 }, { "grad_norm": 0.09782975167036057, "learning_rate": 1.2322758543994062e-06, "loss": 0.0041, "step": 37310 }, { "grad_norm": 0.10027400404214859, "learning_rate": 1.2231718300252603e-06, "loss": 0.0076, "step": 37320 }, { "grad_norm": 0.14967361092567444, "learning_rate": 1.214101144153229e-06, "loss": 0.0062, "step": 37330 }, { "grad_norm": 0.1448131650686264, "learning_rate": 1.205063802983042e-06, "loss": 0.0053, "step": 37340 }, { "grad_norm": 0.08935777097940445, "learning_rate": 1.1960598126916301e-06, "loss": 0.0065, "step": 37350 }, { "grad_norm": 0.07183230668306351, "learning_rate": 1.187089179433154e-06, "loss": 0.0055, "step": 37360 }, { "grad_norm": 0.08562599867582321, "learning_rate": 1.178151909338937e-06, "loss": 0.0045, "step": 37370 }, { "grad_norm": 0.11301567405462265, "learning_rate": 1.1692480085175317e-06, "loss": 0.0043, "step": 37380 }, { "grad_norm": 0.04916642606258392, "learning_rate": 1.1603774830546587e-06, "loss": 0.0059, "step": 37390 }, { "grad_norm": 0.1000727191567421, "learning_rate": 1.1515403390132573e-06, "loss": 0.0053, "step": 37400 }, { "grad_norm": 0.06591156870126724, "learning_rate": 1.1427365824334124e-06, "loss": 0.0065, "step": 37410 }, { "grad_norm": 0.11545737832784653, "learning_rate": 1.133966219332422e-06, "loss": 0.0045, "step": 37420 }, { "grad_norm": 0.13622866570949554, "learning_rate": 1.1252292557047405e-06, "loss": 0.0043, "step": 37430 }, { "grad_norm": 0.08197493851184845, "learning_rate": 1.1165256975219974e-06, "loss": 0.0069, "step": 37440 }, { "grad_norm": 0.10730148106813431, "learning_rate": 1.1078555507330002e-06, "loss": 0.0044, "step": 37450 }, { "grad_norm": 0.08890367299318314, "learning_rate": 1.0992188212637089e-06, "loss": 0.006, "step": 37460 }, { "grad_norm": 0.09844283759593964, "learning_rate": 1.0906155150172458e-06, "loss": 0.006, "step": 37470 }, { "grad_norm": 0.13079802691936493, "learning_rate": 1.0820456378738852e-06, "loss": 0.0046, "step": 37480 }, { "grad_norm": 0.09300777316093445, "learning_rate": 1.0735091956910692e-06, "loss": 0.0053, "step": 37490 }, { "grad_norm": 0.13889186084270477, "learning_rate": 1.0650061943033695e-06, "loss": 0.0052, "step": 37500 }, { "grad_norm": 0.09942030906677246, "learning_rate": 1.0565366395225041e-06, "loss": 0.0047, "step": 37510 }, { "grad_norm": 0.10892189294099808, "learning_rate": 1.0481005371373366e-06, "loss": 0.0059, "step": 37520 }, { "grad_norm": 0.18046662211418152, "learning_rate": 1.0396978929138602e-06, "loss": 0.0055, "step": 37530 }, { "grad_norm": 0.11524263024330139, "learning_rate": 1.0313287125952032e-06, "loss": 0.0049, "step": 37540 }, { "grad_norm": 0.15253093838691711, "learning_rate": 1.0229930019016288e-06, "loss": 0.0045, "step": 37550 }, { "grad_norm": 0.06181686744093895, "learning_rate": 1.0146907665305183e-06, "loss": 0.0045, "step": 37560 }, { "grad_norm": 0.06902436167001724, "learning_rate": 1.0064220121563551e-06, "loss": 0.0066, "step": 37570 }, { "grad_norm": 0.1392659693956375, "learning_rate": 9.98186744430768e-07, "loss": 0.0039, "step": 37580 }, { "grad_norm": 0.15346872806549072, "learning_rate": 9.899849689824826e-07, "loss": 0.0043, "step": 37590 }, { "grad_norm": 0.09278157353401184, "learning_rate": 9.818166914173365e-07, "loss": 0.0031, "step": 37600 }, { "grad_norm": 0.12912170588970184, "learning_rate": 9.736819173182642e-07, "loss": 0.0045, "step": 37610 }, { "grad_norm": 0.14030902087688446, "learning_rate": 9.655806522453126e-07, "loss": 0.0034, "step": 37620 }, { "grad_norm": 0.1008242815732956, "learning_rate": 9.575129017356187e-07, "loss": 0.0037, "step": 37630 }, { "grad_norm": 0.11448147147893906, "learning_rate": 9.494786713034109e-07, "loss": 0.0038, "step": 37640 }, { "grad_norm": 0.1329721212387085, "learning_rate": 9.414779664400075e-07, "loss": 0.0055, "step": 37650 }, { "grad_norm": 0.05629802867770195, "learning_rate": 9.335107926138342e-07, "loss": 0.0054, "step": 37660 }, { "grad_norm": 0.13449041545391083, "learning_rate": 9.255771552703519e-07, "loss": 0.0061, "step": 37670 }, { "grad_norm": 0.13001896440982819, "learning_rate": 9.176770598321449e-07, "loss": 0.0069, "step": 37680 }, { "grad_norm": 0.08041563630104065, "learning_rate": 9.098105116988498e-07, "loss": 0.0057, "step": 37690 }, { "grad_norm": 0.05103453993797302, "learning_rate": 9.019775162471767e-07, "loss": 0.0036, "step": 37700 }, { "grad_norm": 0.10697662085294724, "learning_rate": 8.94178078830904e-07, "loss": 0.0054, "step": 37710 }, { "grad_norm": 0.09371161460876465, "learning_rate": 8.864122047808731e-07, "loss": 0.0044, "step": 37720 }, { "grad_norm": 0.05627840757369995, "learning_rate": 8.786798994049882e-07, "loss": 0.0033, "step": 37730 }, { "grad_norm": 0.144275963306427, "learning_rate": 8.709811679882052e-07, "loss": 0.0047, "step": 37740 }, { "grad_norm": 0.06670618057250977, "learning_rate": 8.633160157925313e-07, "loss": 0.0043, "step": 37750 }, { "grad_norm": 0.04802268370985985, "learning_rate": 8.556844480570369e-07, "loss": 0.0038, "step": 37760 }, { "grad_norm": 0.11127651482820511, "learning_rate": 8.480864699978164e-07, "loss": 0.0034, "step": 37770 }, { "grad_norm": 0.09163660556077957, "learning_rate": 8.405220868080154e-07, "loss": 0.0065, "step": 37780 }, { "grad_norm": 0.1003507748246193, "learning_rate": 8.329913036578263e-07, "loss": 0.0035, "step": 37790 }, { "grad_norm": 0.12035197764635086, "learning_rate": 8.254941256944593e-07, "loss": 0.0038, "step": 37800 }, { "grad_norm": 0.07154583185911179, "learning_rate": 8.18030558042171e-07, "loss": 0.0052, "step": 37810 }, { "grad_norm": 0.17069223523139954, "learning_rate": 8.106006058022419e-07, "loss": 0.005, "step": 37820 }, { "grad_norm": 0.18576490879058838, "learning_rate": 8.03204274052971e-07, "loss": 0.0034, "step": 37830 }, { "grad_norm": 0.18310721218585968, "learning_rate": 7.958415678496756e-07, "loss": 0.0048, "step": 37840 }, { "grad_norm": 0.06250732392072678, "learning_rate": 7.885124922247022e-07, "loss": 0.0059, "step": 37850 }, { "grad_norm": 0.06669541448354721, "learning_rate": 7.812170521874051e-07, "loss": 0.0038, "step": 37860 }, { "grad_norm": 0.08704227209091187, "learning_rate": 7.739552527241456e-07, "loss": 0.0034, "step": 37870 }, { "grad_norm": 0.12860026955604553, "learning_rate": 7.667270987983033e-07, "loss": 0.0043, "step": 37880 }, { "grad_norm": 0.06707409024238586, "learning_rate": 7.595325953502375e-07, "loss": 0.0039, "step": 37890 }, { "grad_norm": 0.11465486139059067, "learning_rate": 7.523717472973368e-07, "loss": 0.0033, "step": 37900 }, { "grad_norm": 0.07244153320789337, "learning_rate": 7.452445595339696e-07, "loss": 0.004, "step": 37910 }, { "grad_norm": 0.0917779877781868, "learning_rate": 7.381510369315059e-07, "loss": 0.0037, "step": 37920 }, { "grad_norm": 0.10387419164180756, "learning_rate": 7.310911843382895e-07, "loss": 0.0041, "step": 37930 }, { "grad_norm": 0.09623298794031143, "learning_rate": 7.240650065796716e-07, "loss": 0.004, "step": 37940 }, { "grad_norm": 0.08646437525749207, "learning_rate": 7.170725084579777e-07, "loss": 0.0036, "step": 37950 }, { "grad_norm": 0.051972150802612305, "learning_rate": 7.101136947525233e-07, "loss": 0.0043, "step": 37960 }, { "grad_norm": 0.1109439954161644, "learning_rate": 7.031885702195707e-07, "loss": 0.0048, "step": 37970 }, { "grad_norm": 0.10316566377878189, "learning_rate": 6.962971395924e-07, "loss": 0.0042, "step": 37980 }, { "grad_norm": 0.052693430334329605, "learning_rate": 6.894394075812217e-07, "loss": 0.0033, "step": 37990 }, { "grad_norm": 0.1480342000722885, "learning_rate": 6.826153788732359e-07, "loss": 0.0062, "step": 38000 }, { "grad_norm": 0.11307413130998611, "learning_rate": 6.758250581326009e-07, "loss": 0.0045, "step": 38010 }, { "grad_norm": 0.14146630465984344, "learning_rate": 6.690684500004485e-07, "loss": 0.0044, "step": 38020 }, { "grad_norm": 0.07313809543848038, "learning_rate": 6.623455590948458e-07, "loss": 0.0034, "step": 38030 }, { "grad_norm": 0.04393701255321503, "learning_rate": 6.556563900108226e-07, "loss": 0.0038, "step": 38040 }, { "grad_norm": 0.15218675136566162, "learning_rate": 6.490009473203662e-07, "loss": 0.005, "step": 38050 }, { "grad_norm": 0.09782898426055908, "learning_rate": 6.423792355724156e-07, "loss": 0.0039, "step": 38060 }, { "grad_norm": 0.07427822053432465, "learning_rate": 6.35791259292845e-07, "loss": 0.0035, "step": 38070 }, { "grad_norm": 0.21878774464130402, "learning_rate": 6.292370229844691e-07, "loss": 0.0047, "step": 38080 }, { "grad_norm": 0.10453519970178604, "learning_rate": 6.227165311270488e-07, "loss": 0.0061, "step": 38090 }, { "grad_norm": 0.16385459899902344, "learning_rate": 6.16229788177275e-07, "loss": 0.0073, "step": 38100 }, { "grad_norm": 0.06453841179609299, "learning_rate": 6.097767985687841e-07, "loss": 0.0045, "step": 38110 }, { "grad_norm": 0.12464076280593872, "learning_rate": 6.033575667121317e-07, "loss": 0.0049, "step": 38120 }, { "grad_norm": 0.14569154381752014, "learning_rate": 5.969720969947967e-07, "loss": 0.0066, "step": 38130 }, { "grad_norm": 0.08878745138645172, "learning_rate": 5.906203937811883e-07, "loss": 0.0032, "step": 38140 }, { "grad_norm": 0.12274862080812454, "learning_rate": 5.843024614126447e-07, "loss": 0.0039, "step": 38150 }, { "grad_norm": 0.14598612487316132, "learning_rate": 5.780183042074061e-07, "loss": 0.0049, "step": 38160 }, { "grad_norm": 0.13052771985530853, "learning_rate": 5.717679264606312e-07, "loss": 0.0064, "step": 38170 }, { "grad_norm": 0.06942648440599442, "learning_rate": 5.655513324444028e-07, "loss": 0.0032, "step": 38180 }, { "grad_norm": 0.12742944061756134, "learning_rate": 5.593685264076942e-07, "loss": 0.005, "step": 38190 }, { "grad_norm": 0.09220876544713974, "learning_rate": 5.532195125764029e-07, "loss": 0.0043, "step": 38200 }, { "grad_norm": 0.08730637282133102, "learning_rate": 5.471042951533167e-07, "loss": 0.0042, "step": 38210 }, { "grad_norm": 0.10186808556318283, "learning_rate": 5.410228783181371e-07, "loss": 0.0041, "step": 38220 }, { "grad_norm": 0.07239686697721481, "learning_rate": 5.349752662274444e-07, "loss": 0.0057, "step": 38230 }, { "grad_norm": 0.08030302077531815, "learning_rate": 5.289614630147321e-07, "loss": 0.004, "step": 38240 }, { "grad_norm": 0.15143562853336334, "learning_rate": 5.229814727903737e-07, "loss": 0.0051, "step": 38250 }, { "grad_norm": 0.077704519033432, "learning_rate": 5.170352996416439e-07, "loss": 0.0049, "step": 38260 }, { "grad_norm": 0.09765303879976273, "learning_rate": 5.111229476326918e-07, "loss": 0.0037, "step": 38270 }, { "grad_norm": 0.1494513750076294, "learning_rate": 5.052444208045459e-07, "loss": 0.0043, "step": 38280 }, { "grad_norm": 0.15120817720890045, "learning_rate": 4.993997231751312e-07, "loss": 0.0051, "step": 38290 }, { "grad_norm": 0.059100840240716934, "learning_rate": 4.935888587392468e-07, "loss": 0.0046, "step": 38300 }, { "grad_norm": 0.11483117192983627, "learning_rate": 4.878118314685598e-07, "loss": 0.0054, "step": 38310 }, { "grad_norm": 0.06479094922542572, "learning_rate": 4.820686453116119e-07, "loss": 0.0042, "step": 38320 }, { "grad_norm": 0.08826127648353577, "learning_rate": 4.763593041938186e-07, "loss": 0.0057, "step": 38330 }, { "grad_norm": 0.07818050682544708, "learning_rate": 4.706838120174584e-07, "loss": 0.0074, "step": 38340 }, { "grad_norm": 0.11020657420158386, "learning_rate": 4.650421726616838e-07, "loss": 0.0048, "step": 38350 }, { "grad_norm": 0.1382933109998703, "learning_rate": 4.594343899824993e-07, "loss": 0.0041, "step": 38360 }, { "grad_norm": 0.25844326615333557, "learning_rate": 4.538604678127667e-07, "loss": 0.006, "step": 38370 }, { "grad_norm": 0.1733020544052124, "learning_rate": 4.483204099622107e-07, "loss": 0.0049, "step": 38380 }, { "grad_norm": 0.1291975975036621, "learning_rate": 4.4281422021740793e-07, "loss": 0.0037, "step": 38390 }, { "grad_norm": 0.09293898940086365, "learning_rate": 4.373419023417924e-07, "loss": 0.004, "step": 38400 }, { "grad_norm": 0.16561207175254822, "learning_rate": 4.319034600756389e-07, "loss": 0.007, "step": 38410 }, { "grad_norm": 0.09433763474225998, "learning_rate": 4.264988971360739e-07, "loss": 0.0052, "step": 38420 }, { "grad_norm": 0.10956765711307526, "learning_rate": 4.211282172170594e-07, "loss": 0.0048, "step": 38430 }, { "grad_norm": 0.12391354143619537, "learning_rate": 4.1579142398940894e-07, "loss": 0.0038, "step": 38440 }, { "grad_norm": 0.11673881858587265, "learning_rate": 4.1048852110076585e-07, "loss": 0.0041, "step": 38450 }, { "grad_norm": 0.06280696392059326, "learning_rate": 4.052195121756197e-07, "loss": 0.0057, "step": 38460 }, { "grad_norm": 0.10222101956605911, "learning_rate": 3.9998440081528974e-07, "loss": 0.0061, "step": 38470 }, { "grad_norm": 0.07853025943040848, "learning_rate": 3.9478319059791936e-07, "loss": 0.005, "step": 38480 }, { "grad_norm": 0.06969985365867615, "learning_rate": 3.8961588507848143e-07, "loss": 0.0052, "step": 38490 }, { "grad_norm": 0.05352512374520302, "learning_rate": 3.844824877887954e-07, "loss": 0.0063, "step": 38500 }, { "grad_norm": 0.07113464176654816, "learning_rate": 3.7938300223748224e-07, "loss": 0.0059, "step": 38510 }, { "grad_norm": 0.06925992667675018, "learning_rate": 3.743174319099929e-07, "loss": 0.0036, "step": 38520 }, { "grad_norm": 0.07472389936447144, "learning_rate": 3.6928578026859673e-07, "loss": 0.0041, "step": 38530 }, { "grad_norm": 0.06146614998579025, "learning_rate": 3.642880507523816e-07, "loss": 0.0059, "step": 38540 }, { "grad_norm": 0.08368668705224991, "learning_rate": 3.5932424677724285e-07, "loss": 0.0062, "step": 38550 }, { "grad_norm": 0.06703100353479385, "learning_rate": 3.5439437173589996e-07, "loss": 0.0044, "step": 38560 }, { "grad_norm": 0.07250954210758209, "learning_rate": 3.4949842899787423e-07, "loss": 0.0049, "step": 38570 }, { "grad_norm": 0.12768878042697906, "learning_rate": 3.446364219094944e-07, "loss": 0.0042, "step": 38580 }, { "grad_norm": 0.17198292911052704, "learning_rate": 3.3980835379389675e-07, "loss": 0.0061, "step": 38590 }, { "grad_norm": 0.042867209762334824, "learning_rate": 3.350142279510249e-07, "loss": 0.005, "step": 38600 }, { "grad_norm": 0.13114264607429504, "learning_rate": 3.3025404765761324e-07, "loss": 0.0049, "step": 38610 }, { "grad_norm": 0.14205686748027802, "learning_rate": 3.255278161672037e-07, "loss": 0.0072, "step": 38620 }, { "grad_norm": 0.04521928355097771, "learning_rate": 3.20835536710129e-07, "loss": 0.0042, "step": 38630 }, { "grad_norm": 0.13331238925457, "learning_rate": 3.1617721249351804e-07, "loss": 0.0049, "step": 38640 }, { "grad_norm": 0.12579849362373352, "learning_rate": 3.115528467012963e-07, "loss": 0.0057, "step": 38650 }, { "grad_norm": 0.11712220311164856, "learning_rate": 3.0696244249416305e-07, "loss": 0.0052, "step": 38660 }, { "grad_norm": 0.11660759150981903, "learning_rate": 3.024060030096254e-07, "loss": 0.0038, "step": 38670 }, { "grad_norm": 0.08539452403783798, "learning_rate": 2.9788353136196435e-07, "loss": 0.005, "step": 38680 }, { "grad_norm": 0.10948626697063446, "learning_rate": 2.9339503064224615e-07, "loss": 0.0044, "step": 38690 }, { "grad_norm": 0.05147339031100273, "learning_rate": 2.889405039183224e-07, "loss": 0.0035, "step": 38700 }, { "grad_norm": 0.07114948332309723, "learning_rate": 2.8451995423481315e-07, "loss": 0.0044, "step": 38710 }, { "grad_norm": 0.06160040199756622, "learning_rate": 2.801333846131293e-07, "loss": 0.005, "step": 38720 }, { "grad_norm": 0.20016276836395264, "learning_rate": 2.757807980514504e-07, "loss": 0.0055, "step": 38730 }, { "grad_norm": 0.12213889509439468, "learning_rate": 2.714621975247189e-07, "loss": 0.0048, "step": 38740 }, { "grad_norm": 0.17171381413936615, "learning_rate": 2.671775859846626e-07, "loss": 0.0059, "step": 38750 }, { "grad_norm": 0.1101907417178154, "learning_rate": 2.629269663597722e-07, "loss": 0.0036, "step": 38760 }, { "grad_norm": 0.1570807546377182, "learning_rate": 2.587103415553127e-07, "loss": 0.0059, "step": 38770 }, { "grad_norm": 0.07622716575860977, "learning_rate": 2.545277144533009e-07, "loss": 0.0042, "step": 38780 }, { "grad_norm": 0.09914133697748184, "learning_rate": 2.5037908791252227e-07, "loss": 0.0038, "step": 38790 }, { "grad_norm": 0.06945658475160599, "learning_rate": 2.462644647685308e-07, "loss": 0.0052, "step": 38800 }, { "grad_norm": 0.1451258808374405, "learning_rate": 2.421838478336269e-07, "loss": 0.0053, "step": 38810 }, { "grad_norm": 0.06995885819196701, "learning_rate": 2.3813723989687398e-07, "loss": 0.0051, "step": 38820 }, { "grad_norm": 0.10373453050851822, "learning_rate": 2.3412464372409293e-07, "loss": 0.0044, "step": 38830 }, { "grad_norm": 0.06068344786763191, "learning_rate": 2.301460620578455e-07, "loss": 0.0053, "step": 38840 }, { "grad_norm": 0.09163659065961838, "learning_rate": 2.262014976174731e-07, "loss": 0.0051, "step": 38850 }, { "grad_norm": 0.06749215722084045, "learning_rate": 2.2229095309903025e-07, "loss": 0.0072, "step": 38860 }, { "grad_norm": 0.19452662765979767, "learning_rate": 2.184144311753511e-07, "loss": 0.0049, "step": 38870 }, { "grad_norm": 0.17791436612606049, "learning_rate": 2.14571934495994e-07, "loss": 0.0045, "step": 38880 }, { "grad_norm": 0.11395599693059921, "learning_rate": 2.1076346568726923e-07, "loss": 0.0045, "step": 38890 }, { "grad_norm": 0.07033458352088928, "learning_rate": 2.0698902735223346e-07, "loss": 0.0057, "step": 38900 }, { "grad_norm": 0.06204380840063095, "learning_rate": 2.0324862207067863e-07, "loss": 0.0046, "step": 38910 }, { "grad_norm": 0.09343241155147552, "learning_rate": 1.9954225239913195e-07, "loss": 0.0043, "step": 38920 }, { "grad_norm": 0.053086016327142715, "learning_rate": 1.9586992087087252e-07, "loss": 0.0033, "step": 38930 }, { "grad_norm": 0.09956888109445572, "learning_rate": 1.9223162999588705e-07, "loss": 0.0047, "step": 38940 }, { "grad_norm": 0.17263935506343842, "learning_rate": 1.886273822609308e-07, "loss": 0.0061, "step": 38950 }, { "grad_norm": 0.1042691171169281, "learning_rate": 1.8505718012946094e-07, "loss": 0.0046, "step": 38960 }, { "grad_norm": 0.10886425524950027, "learning_rate": 1.8152102604168664e-07, "loss": 0.0044, "step": 38970 }, { "grad_norm": 0.10518214106559753, "learning_rate": 1.7801892241452457e-07, "loss": 0.0045, "step": 38980 }, { "grad_norm": 0.16414302587509155, "learning_rate": 1.745508716416433e-07, "loss": 0.0059, "step": 38990 }, { "grad_norm": 0.11705225706100464, "learning_rate": 1.7111687609341342e-07, "loss": 0.0045, "step": 39000 }, { "grad_norm": 0.18048527836799622, "learning_rate": 1.6771693811693523e-07, "loss": 0.0053, "step": 39010 }, { "grad_norm": 0.06619539111852646, "learning_rate": 1.643510600360443e-07, "loss": 0.004, "step": 39020 }, { "grad_norm": 0.0648217499256134, "learning_rate": 1.610192441512781e-07, "loss": 0.0053, "step": 39030 }, { "grad_norm": 0.1721707433462143, "learning_rate": 1.5772149273990956e-07, "loss": 0.006, "step": 39040 }, { "grad_norm": 0.11328821629285812, "learning_rate": 1.5445780805591337e-07, "loss": 0.0036, "step": 39050 }, { "grad_norm": 0.1263105720281601, "learning_rate": 1.5122819232998852e-07, "loss": 0.006, "step": 39060 }, { "grad_norm": 0.12520481646060944, "learning_rate": 1.4803264776955262e-07, "loss": 0.0061, "step": 39070 }, { "grad_norm": 0.13132013380527496, "learning_rate": 1.448711765587196e-07, "loss": 0.0036, "step": 39080 }, { "grad_norm": 0.08489525318145752, "learning_rate": 1.4174378085833885e-07, "loss": 0.0045, "step": 39090 }, { "grad_norm": 0.06805674731731415, "learning_rate": 1.3865046280594486e-07, "loss": 0.006, "step": 39100 }, { "grad_norm": 0.15010708570480347, "learning_rate": 1.3559122451579643e-07, "loss": 0.0034, "step": 39110 }, { "grad_norm": 0.13335278630256653, "learning_rate": 1.3256606807884874e-07, "loss": 0.007, "step": 39120 }, { "grad_norm": 0.07548757642507553, "learning_rate": 1.2957499556277563e-07, "loss": 0.0061, "step": 39130 }, { "grad_norm": 0.09721417725086212, "learning_rate": 1.266180090119473e-07, "loss": 0.0056, "step": 39140 }, { "grad_norm": 0.049359727650880814, "learning_rate": 1.2369511044742488e-07, "loss": 0.0032, "step": 39150 }, { "grad_norm": 0.17918796837329865, "learning_rate": 1.2080630186698805e-07, "loss": 0.0039, "step": 39160 }, { "grad_norm": 0.04256189614534378, "learning_rate": 1.1795158524511851e-07, "loss": 0.0035, "step": 39170 }, { "grad_norm": 0.036973271518945694, "learning_rate": 1.1513096253297773e-07, "loss": 0.0035, "step": 39180 }, { "grad_norm": 0.0813162624835968, "learning_rate": 1.1234443565843466e-07, "loss": 0.0053, "step": 39190 }, { "grad_norm": 0.10540996491909027, "learning_rate": 1.0959200652606028e-07, "loss": 0.0039, "step": 39200 }, { "grad_norm": 0.11816899478435516, "learning_rate": 1.0687367701711082e-07, "loss": 0.0046, "step": 39210 }, { "grad_norm": 0.08696374297142029, "learning_rate": 1.0418944898953898e-07, "loss": 0.0046, "step": 39220 }, { "grad_norm": 0.09089510887861252, "learning_rate": 1.0153932427798829e-07, "loss": 0.0042, "step": 39230 }, { "grad_norm": 0.2553768455982208, "learning_rate": 9.892330469379874e-08, "loss": 0.0084, "step": 39240 }, { "grad_norm": 0.0685281828045845, "learning_rate": 9.634139202498449e-08, "loss": 0.0029, "step": 39250 }, { "grad_norm": 0.0841868668794632, "learning_rate": 9.379358803627281e-08, "loss": 0.0033, "step": 39260 }, { "grad_norm": 0.15823249518871307, "learning_rate": 9.12798944690485e-08, "loss": 0.004, "step": 39270 }, { "grad_norm": 0.1694367378950119, "learning_rate": 8.880031304141501e-08, "loss": 0.0066, "step": 39280 }, { "grad_norm": 0.11669868230819702, "learning_rate": 8.635484544812777e-08, "loss": 0.007, "step": 39290 }, { "grad_norm": 0.15660767257213593, "learning_rate": 8.394349336064422e-08, "loss": 0.0039, "step": 39300 }, { "grad_norm": 0.12022961676120758, "learning_rate": 8.156625842710153e-08, "loss": 0.0059, "step": 39310 }, { "grad_norm": 0.14309529960155487, "learning_rate": 7.92231422723222e-08, "loss": 0.0047, "step": 39320 }, { "grad_norm": 0.05034496635198593, "learning_rate": 7.691414649780293e-08, "loss": 0.0059, "step": 39330 }, { "grad_norm": 0.07587256282567978, "learning_rate": 7.463927268171467e-08, "loss": 0.0033, "step": 39340 }, { "grad_norm": 0.13210664689540863, "learning_rate": 7.239852237891365e-08, "loss": 0.0057, "step": 39350 }, { "grad_norm": 0.11146321892738342, "learning_rate": 7.019189712093588e-08, "loss": 0.0054, "step": 39360 }, { "grad_norm": 0.07236701995134354, "learning_rate": 6.801939841598603e-08, "loss": 0.004, "step": 39370 }, { "grad_norm": 0.11097169667482376, "learning_rate": 6.588102774894856e-08, "loss": 0.0047, "step": 39380 }, { "grad_norm": 0.09818994253873825, "learning_rate": 6.377678658137653e-08, "loss": 0.0039, "step": 39390 }, { "grad_norm": 0.10067541152238846, "learning_rate": 6.17066763514973e-08, "loss": 0.0031, "step": 39400 }, { "grad_norm": 0.1535744071006775, "learning_rate": 5.967069847421792e-08, "loss": 0.0039, "step": 39410 }, { "grad_norm": 0.07015911489725113, "learning_rate": 5.7668854341108583e-08, "loss": 0.0044, "step": 39420 }, { "grad_norm": 0.1405186504125595, "learning_rate": 5.570114532040815e-08, "loss": 0.0055, "step": 39430 }, { "grad_norm": 0.05974655970931053, "learning_rate": 5.3767572757029706e-08, "loss": 0.0059, "step": 39440 }, { "grad_norm": 0.05183366313576698, "learning_rate": 5.186813797254386e-08, "loss": 0.0051, "step": 39450 }, { "grad_norm": 0.19786740839481354, "learning_rate": 5.000284226521213e-08, "loss": 0.0046, "step": 39460 }, { "grad_norm": 0.07087104767560959, "learning_rate": 4.8171686909931346e-08, "loss": 0.0041, "step": 39470 }, { "grad_norm": 0.0781642496585846, "learning_rate": 4.637467315828925e-08, "loss": 0.0043, "step": 39480 }, { "grad_norm": 0.14296001195907593, "learning_rate": 4.461180223852002e-08, "loss": 0.0049, "step": 39490 }, { "grad_norm": 0.1525079309940338, "learning_rate": 4.288307535553204e-08, "loss": 0.0043, "step": 39500 }, { "grad_norm": 0.15494118630886078, "learning_rate": 4.1188493690896836e-08, "loss": 0.0086, "step": 39510 }, { "grad_norm": 0.13947105407714844, "learning_rate": 3.952805840284346e-08, "loss": 0.0055, "step": 39520 }, { "grad_norm": 0.05333811044692993, "learning_rate": 3.790177062625855e-08, "loss": 0.0033, "step": 39530 }, { "grad_norm": 0.07750646770000458, "learning_rate": 3.6309631472702946e-08, "loss": 0.0043, "step": 39540 }, { "grad_norm": 0.09263765066862106, "learning_rate": 3.475164203038395e-08, "loss": 0.0039, "step": 39550 }, { "grad_norm": 0.062074385583400726, "learning_rate": 3.3227803364171974e-08, "loss": 0.0036, "step": 39560 }, { "grad_norm": 0.06641267985105515, "learning_rate": 3.173811651560055e-08, "loss": 0.0038, "step": 39570 }, { "grad_norm": 0.10558293759822845, "learning_rate": 3.0282582502855204e-08, "loss": 0.0063, "step": 39580 }, { "grad_norm": 0.08824598789215088, "learning_rate": 2.8861202320773496e-08, "loss": 0.0053, "step": 39590 }, { "grad_norm": 0.13672702014446259, "learning_rate": 2.7473976940867175e-08, "loss": 0.0039, "step": 39600 }, { "grad_norm": 0.07490906119346619, "learning_rate": 2.6120907311277808e-08, "loss": 0.0053, "step": 39610 }, { "grad_norm": 0.13172200322151184, "learning_rate": 2.480199435682673e-08, "loss": 0.0035, "step": 39620 }, { "grad_norm": 0.08086824417114258, "learning_rate": 2.3517238978976174e-08, "loss": 0.004, "step": 39630 }, { "grad_norm": 0.0692644864320755, "learning_rate": 2.2266642055834842e-08, "loss": 0.0037, "step": 39640 }, { "grad_norm": 0.09321790188550949, "learning_rate": 2.105020444219119e-08, "loss": 0.0035, "step": 39650 }, { "grad_norm": 0.12687189877033234, "learning_rate": 1.9867926969452388e-08, "loss": 0.0039, "step": 39660 }, { "grad_norm": 0.12195374071598053, "learning_rate": 1.8719810445705366e-08, "loss": 0.0046, "step": 39670 }, { "grad_norm": 0.0875532329082489, "learning_rate": 1.760585565567241e-08, "loss": 0.0046, "step": 39680 }, { "grad_norm": 0.05253427475690842, "learning_rate": 1.6526063360727816e-08, "loss": 0.0042, "step": 39690 }, { "grad_norm": 0.1241232380270958, "learning_rate": 1.5480434298908997e-08, "loss": 0.0048, "step": 39700 }, { "grad_norm": 0.09781590104103088, "learning_rate": 1.4468969184883164e-08, "loss": 0.0056, "step": 39710 }, { "grad_norm": 0.1360105723142624, "learning_rate": 1.3491668709980643e-08, "loss": 0.0038, "step": 39720 }, { "grad_norm": 0.18042081594467163, "learning_rate": 1.254853354217822e-08, "loss": 0.0048, "step": 39730 }, { "grad_norm": 0.08090737462043762, "learning_rate": 1.1639564326099139e-08, "loss": 0.0071, "step": 39740 }, { "grad_norm": 0.07036187499761581, "learning_rate": 1.0764761683018653e-08, "loss": 0.0035, "step": 39750 }, { "grad_norm": 0.13990077376365662, "learning_rate": 9.924126210852925e-09, "loss": 0.0045, "step": 39760 }, { "grad_norm": 0.09336527436971664, "learning_rate": 9.117658484170122e-09, "loss": 0.0045, "step": 39770 }, { "grad_norm": 0.06625070422887802, "learning_rate": 8.345359054179324e-09, "loss": 0.0049, "step": 39780 }, { "grad_norm": 0.11704856902360916, "learning_rate": 7.60722844874162e-09, "loss": 0.0062, "step": 39790 }, { "grad_norm": 0.07892632484436035, "learning_rate": 6.903267172364558e-09, "loss": 0.0057, "step": 39800 }, { "grad_norm": 0.2853752076625824, "learning_rate": 6.23347570619659e-09, "loss": 0.0046, "step": 39810 }, { "grad_norm": 0.09412170946598053, "learning_rate": 5.59785450803818e-09, "loss": 0.0055, "step": 39820 }, { "grad_norm": 0.11525558680295944, "learning_rate": 4.996404012325151e-09, "loss": 0.0043, "step": 39830 }, { "grad_norm": 0.17790645360946655, "learning_rate": 4.429124630145331e-09, "loss": 0.0047, "step": 39840 }, { "grad_norm": 0.12091731280088425, "learning_rate": 3.89601674922746e-09, "loss": 0.0058, "step": 39850 }, { "grad_norm": 0.11416323482990265, "learning_rate": 3.3970807339467336e-09, "loss": 0.0043, "step": 39860 }, { "grad_norm": 0.12327949702739716, "learning_rate": 2.9323169253248073e-09, "loss": 0.0041, "step": 39870 }, { "grad_norm": 0.09279265254735947, "learning_rate": 2.501725641018693e-09, "loss": 0.0041, "step": 39880 }, { "grad_norm": 0.18975351750850677, "learning_rate": 2.105307175331861e-09, "loss": 0.007, "step": 39890 }, { "grad_norm": 0.07955949008464813, "learning_rate": 1.7430617992197918e-09, "loss": 0.0058, "step": 39900 }, { "grad_norm": 0.13656866550445557, "learning_rate": 1.4149897602677708e-09, "loss": 0.0046, "step": 39910 }, { "grad_norm": 0.05144687741994858, "learning_rate": 1.1210912827130937e-09, "loss": 0.0054, "step": 39920 }, { "grad_norm": 0.12052619457244873, "learning_rate": 8.613665674284121e-10, "loss": 0.0065, "step": 39930 }, { "grad_norm": 0.09748044610023499, "learning_rate": 6.358157919383878e-10, "loss": 0.0037, "step": 39940 }, { "grad_norm": 0.060764655470848083, "learning_rate": 4.4443911040303877e-10, "loss": 0.003, "step": 39950 }, { "grad_norm": 0.1668044477701187, "learning_rate": 2.8723665362329066e-10, "loss": 0.0056, "step": 39960 }, { "grad_norm": 0.10378217697143555, "learning_rate": 1.6420852905207895e-10, "loss": 0.0034, "step": 39970 }, { "grad_norm": 0.05499101057648659, "learning_rate": 7.535482076659328e-11, "loss": 0.0044, "step": 39980 }, { "grad_norm": 0.14028501510620117, "learning_rate": 2.0675589512686443e-11, "loss": 0.0032, "step": 39990 }, { "grad_norm": 0.09810902178287506, "learning_rate": 1.7087265491433358e-13, "loss": 0.0037, "step": 40000 } ], "logging_steps": 10, "max_steps": 40000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "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": 48, "trial_name": null, "trial_params": null }