{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 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 } ], "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }