{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.8302667737007141, "learning_rate": 1.8e-07, "loss": 1.4972, "step": 10 }, { "grad_norm": 0.9987985491752625, "learning_rate": 3.8e-07, "loss": 1.5292, "step": 20 }, { "grad_norm": 1.0243146419525146, "learning_rate": 5.8e-07, "loss": 1.5136, "step": 30 }, { "grad_norm": 0.8147416114807129, "learning_rate": 7.8e-07, "loss": 1.5034, "step": 40 }, { "grad_norm": 0.8266986012458801, "learning_rate": 9.8e-07, "loss": 1.5043, "step": 50 }, { "grad_norm": 1.0818147659301758, "learning_rate": 1.18e-06, "loss": 1.4872, "step": 60 }, { "grad_norm": 2.1023054122924805, "learning_rate": 1.3800000000000001e-06, "loss": 1.4102, "step": 70 }, { "grad_norm": 2.5768675804138184, "learning_rate": 1.5800000000000003e-06, "loss": 1.3446, "step": 80 }, { "grad_norm": 1.571958065032959, "learning_rate": 1.7800000000000001e-06, "loss": 1.3076, "step": 90 }, { "grad_norm": 1.5125161409378052, "learning_rate": 1.98e-06, "loss": 1.2864, "step": 100 }, { "grad_norm": 1.7268950939178467, "learning_rate": 2.1800000000000003e-06, "loss": 1.2424, "step": 110 }, { "grad_norm": 3.208794355392456, "learning_rate": 2.38e-06, "loss": 1.221, "step": 120 }, { "grad_norm": 1.791698694229126, "learning_rate": 2.5800000000000003e-06, "loss": 1.2131, "step": 130 }, { "grad_norm": 1.7060472965240479, "learning_rate": 2.78e-06, "loss": 1.1834, "step": 140 }, { "grad_norm": 1.6625208854675293, "learning_rate": 2.9800000000000003e-06, "loss": 1.1745, "step": 150 }, { "grad_norm": 2.4024059772491455, "learning_rate": 3.1800000000000005e-06, "loss": 1.1323, "step": 160 }, { "grad_norm": 1.363789677619934, "learning_rate": 3.38e-06, "loss": 1.2006, "step": 170 }, { "grad_norm": 1.7876535654067993, "learning_rate": 3.58e-06, "loss": 1.1798, "step": 180 }, { "grad_norm": 1.6704213619232178, "learning_rate": 3.7800000000000002e-06, "loss": 1.1921, "step": 190 }, { "grad_norm": 1.7402814626693726, "learning_rate": 3.98e-06, "loss": 1.1593, "step": 200 }, { "grad_norm": 1.6397954225540161, "learning_rate": 4.18e-06, "loss": 1.1694, "step": 210 }, { "grad_norm": 1.383514642715454, "learning_rate": 4.38e-06, "loss": 1.1539, "step": 220 }, { "grad_norm": 1.8937309980392456, "learning_rate": 4.58e-06, "loss": 1.1631, "step": 230 }, { "grad_norm": 1.6143232583999634, "learning_rate": 4.780000000000001e-06, "loss": 1.1362, "step": 240 }, { "grad_norm": 1.43906831741333, "learning_rate": 4.98e-06, "loss": 1.0993, "step": 250 }, { "grad_norm": 1.1479823589324951, "learning_rate": 5.18e-06, "loss": 1.1046, "step": 260 }, { "grad_norm": 3.779651403427124, "learning_rate": 5.38e-06, "loss": 1.1281, "step": 270 }, { "grad_norm": 1.5249624252319336, "learning_rate": 5.580000000000001e-06, "loss": 1.1272, "step": 280 }, { "grad_norm": 2.0277793407440186, "learning_rate": 5.78e-06, "loss": 1.111, "step": 290 }, { "grad_norm": 1.4434971809387207, "learning_rate": 5.98e-06, "loss": 1.1036, "step": 300 }, { "grad_norm": 2.2934181690216064, "learning_rate": 6.18e-06, "loss": 1.1004, "step": 310 }, { "grad_norm": 2.3238894939422607, "learning_rate": 6.38e-06, "loss": 1.1119, "step": 320 }, { "grad_norm": 1.8201934099197388, "learning_rate": 6.58e-06, "loss": 1.0951, "step": 330 }, { "grad_norm": 1.681266188621521, "learning_rate": 6.78e-06, "loss": 1.0914, "step": 340 }, { "grad_norm": 1.60459566116333, "learning_rate": 6.98e-06, "loss": 1.0872, "step": 350 }, { "grad_norm": 1.7755944728851318, "learning_rate": 7.180000000000001e-06, "loss": 1.1018, "step": 360 }, { "grad_norm": 1.9850140810012817, "learning_rate": 7.3800000000000005e-06, "loss": 1.0873, "step": 370 }, { "grad_norm": 1.5803022384643555, "learning_rate": 7.580000000000001e-06, "loss": 1.0848, "step": 380 }, { "grad_norm": 2.4670138359069824, "learning_rate": 7.78e-06, "loss": 1.0639, "step": 390 }, { "grad_norm": 1.3110121488571167, "learning_rate": 7.98e-06, "loss": 1.0937, "step": 400 }, { "grad_norm": 2.062417984008789, "learning_rate": 8.18e-06, "loss": 1.1012, "step": 410 }, { "grad_norm": 1.2316129207611084, "learning_rate": 8.380000000000001e-06, "loss": 1.0918, "step": 420 }, { "grad_norm": 1.8419245481491089, "learning_rate": 8.580000000000001e-06, "loss": 1.0794, "step": 430 }, { "grad_norm": 1.3531033992767334, "learning_rate": 8.78e-06, "loss": 1.0815, "step": 440 }, { "grad_norm": 1.4888677597045898, "learning_rate": 8.98e-06, "loss": 1.0653, "step": 450 }, { "grad_norm": 1.7598991394042969, "learning_rate": 9.180000000000002e-06, "loss": 1.0691, "step": 460 }, { "grad_norm": 1.9838061332702637, "learning_rate": 9.38e-06, "loss": 1.0796, "step": 470 }, { "grad_norm": 2.894461154937744, "learning_rate": 9.58e-06, "loss": 1.0877, "step": 480 }, { "grad_norm": 1.4334461688995361, "learning_rate": 9.78e-06, "loss": 1.0779, "step": 490 }, { "grad_norm": 1.6248480081558228, "learning_rate": 9.980000000000001e-06, "loss": 1.0688, "step": 500 }, { "grad_norm": 1.33085298538208, "learning_rate": 1.018e-05, "loss": 1.0665, "step": 510 }, { "grad_norm": 1.4935133457183838, "learning_rate": 1.038e-05, "loss": 1.0634, "step": 520 }, { "grad_norm": 1.791306495666504, "learning_rate": 1.058e-05, "loss": 1.0505, "step": 530 }, { "grad_norm": 1.8696728944778442, "learning_rate": 1.0780000000000002e-05, "loss": 1.0284, "step": 540 }, { "grad_norm": 1.558366060256958, "learning_rate": 1.098e-05, "loss": 1.0334, "step": 550 }, { "grad_norm": 2.090794086456299, "learning_rate": 1.118e-05, "loss": 1.0168, "step": 560 }, { "grad_norm": 1.8543426990509033, "learning_rate": 1.1380000000000001e-05, "loss": 1.0073, "step": 570 }, { "grad_norm": 2.5221805572509766, "learning_rate": 1.1580000000000001e-05, "loss": 0.9296, "step": 580 }, { "grad_norm": 2.5150296688079834, "learning_rate": 1.178e-05, "loss": 0.9096, "step": 590 }, { "grad_norm": 2.3156580924987793, "learning_rate": 1.198e-05, "loss": 0.8703, "step": 600 }, { "grad_norm": 2.4866106510162354, "learning_rate": 1.2180000000000002e-05, "loss": 0.8483, "step": 610 }, { "grad_norm": 4.273382186889648, "learning_rate": 1.238e-05, "loss": 0.8299, "step": 620 }, { "grad_norm": 2.5873730182647705, "learning_rate": 1.258e-05, "loss": 0.7874, "step": 630 }, { "grad_norm": 2.5607151985168457, "learning_rate": 1.278e-05, "loss": 0.7577, "step": 640 }, { "grad_norm": 2.1048667430877686, "learning_rate": 1.2980000000000001e-05, "loss": 0.7228, "step": 650 }, { "grad_norm": 2.356624126434326, "learning_rate": 1.3180000000000001e-05, "loss": 0.7032, "step": 660 }, { "grad_norm": 2.1268818378448486, "learning_rate": 1.338e-05, "loss": 0.6657, "step": 670 }, { "grad_norm": 2.912905216217041, "learning_rate": 1.358e-05, "loss": 0.663, "step": 680 }, { "grad_norm": 2.4040870666503906, "learning_rate": 1.3780000000000002e-05, "loss": 0.6251, "step": 690 }, { "grad_norm": 3.115990161895752, "learning_rate": 1.3980000000000002e-05, "loss": 0.6194, "step": 700 }, { "grad_norm": 2.801574230194092, "learning_rate": 1.4180000000000001e-05, "loss": 0.5353, "step": 710 }, { "grad_norm": 2.7314226627349854, "learning_rate": 1.4380000000000001e-05, "loss": 0.5799, "step": 720 }, { "grad_norm": 2.5567100048065186, "learning_rate": 1.4580000000000003e-05, "loss": 0.504, "step": 730 }, { "grad_norm": 2.3663487434387207, "learning_rate": 1.4779999999999999e-05, "loss": 0.461, "step": 740 }, { "grad_norm": 2.5341503620147705, "learning_rate": 1.4979999999999999e-05, "loss": 0.4687, "step": 750 }, { "grad_norm": 2.3281168937683105, "learning_rate": 1.518e-05, "loss": 0.451, "step": 760 }, { "grad_norm": 2.169271230697632, "learning_rate": 1.538e-05, "loss": 0.4277, "step": 770 }, { "grad_norm": 3.338282585144043, "learning_rate": 1.558e-05, "loss": 0.4347, "step": 780 }, { "grad_norm": 2.3700883388519287, "learning_rate": 1.578e-05, "loss": 0.3793, "step": 790 }, { "grad_norm": 2.5786471366882324, "learning_rate": 1.598e-05, "loss": 0.3396, "step": 800 }, { "grad_norm": 2.8967769145965576, "learning_rate": 1.618e-05, "loss": 0.3491, "step": 810 }, { "grad_norm": 2.14170241355896, "learning_rate": 1.6380000000000002e-05, "loss": 0.3329, "step": 820 }, { "grad_norm": 2.857529401779175, "learning_rate": 1.658e-05, "loss": 0.3513, "step": 830 }, { "grad_norm": 3.0106070041656494, "learning_rate": 1.6780000000000002e-05, "loss": 0.3352, "step": 840 }, { "grad_norm": 2.012646436691284, "learning_rate": 1.698e-05, "loss": 0.317, "step": 850 }, { "grad_norm": 3.160024404525757, "learning_rate": 1.718e-05, "loss": 0.2827, "step": 860 }, { "grad_norm": 2.093728542327881, "learning_rate": 1.7380000000000003e-05, "loss": 0.3075, "step": 870 }, { "grad_norm": 2.673781156539917, "learning_rate": 1.758e-05, "loss": 0.2653, "step": 880 }, { "grad_norm": 2.533823013305664, "learning_rate": 1.7780000000000003e-05, "loss": 0.2627, "step": 890 }, { "grad_norm": 2.415320873260498, "learning_rate": 1.798e-05, "loss": 0.2753, "step": 900 }, { "grad_norm": 2.4683878421783447, "learning_rate": 1.818e-05, "loss": 0.2502, "step": 910 }, { "grad_norm": 2.3129923343658447, "learning_rate": 1.838e-05, "loss": 0.2232, "step": 920 }, { "grad_norm": 2.9956793785095215, "learning_rate": 1.858e-05, "loss": 0.2355, "step": 930 }, { "grad_norm": 3.2365026473999023, "learning_rate": 1.878e-05, "loss": 0.2198, "step": 940 }, { "grad_norm": 2.586899995803833, "learning_rate": 1.898e-05, "loss": 0.2031, "step": 950 }, { "grad_norm": 2.6202635765075684, "learning_rate": 1.918e-05, "loss": 0.2123, "step": 960 }, { "grad_norm": 2.3446850776672363, "learning_rate": 1.938e-05, "loss": 0.2483, "step": 970 }, { "grad_norm": 2.2668423652648926, "learning_rate": 1.9580000000000002e-05, "loss": 0.1957, "step": 980 }, { "grad_norm": 2.4811458587646484, "learning_rate": 1.978e-05, "loss": 0.172, "step": 990 }, { "grad_norm": 2.7890713214874268, "learning_rate": 1.9980000000000002e-05, "loss": 0.1874, "step": 1000 }, { "grad_norm": 2.1010892391204834, "learning_rate": 2.0180000000000003e-05, "loss": 0.1856, "step": 1010 }, { "grad_norm": 2.3264524936676025, "learning_rate": 2.038e-05, "loss": 0.2253, "step": 1020 }, { "grad_norm": 2.63651442527771, "learning_rate": 2.0580000000000003e-05, "loss": 0.1753, "step": 1030 }, { "grad_norm": 2.5058341026306152, "learning_rate": 2.078e-05, "loss": 0.1916, "step": 1040 }, { "grad_norm": 2.6587650775909424, "learning_rate": 2.098e-05, "loss": 0.1972, "step": 1050 }, { "grad_norm": 2.0875351428985596, "learning_rate": 2.118e-05, "loss": 0.211, "step": 1060 }, { "grad_norm": 2.0941810607910156, "learning_rate": 2.138e-05, "loss": 0.2248, "step": 1070 }, { "grad_norm": 3.1594479084014893, "learning_rate": 2.158e-05, "loss": 0.2166, "step": 1080 }, { "grad_norm": 1.7430800199508667, "learning_rate": 2.178e-05, "loss": 0.1985, "step": 1090 }, { "grad_norm": 1.9269602298736572, "learning_rate": 2.198e-05, "loss": 0.1458, "step": 1100 }, { "grad_norm": 2.525444507598877, "learning_rate": 2.218e-05, "loss": 0.2292, "step": 1110 }, { "grad_norm": 1.8316104412078857, "learning_rate": 2.2380000000000003e-05, "loss": 0.1514, "step": 1120 }, { "grad_norm": 1.9662659168243408, "learning_rate": 2.258e-05, "loss": 0.1792, "step": 1130 }, { "grad_norm": 2.476553201675415, "learning_rate": 2.2780000000000002e-05, "loss": 0.2318, "step": 1140 }, { "grad_norm": 2.557565689086914, "learning_rate": 2.298e-05, "loss": 0.2124, "step": 1150 }, { "grad_norm": 2.1349127292633057, "learning_rate": 2.318e-05, "loss": 0.1843, "step": 1160 }, { "grad_norm": 1.8948332071304321, "learning_rate": 2.3380000000000003e-05, "loss": 0.1915, "step": 1170 }, { "grad_norm": 2.0268115997314453, "learning_rate": 2.358e-05, "loss": 0.154, "step": 1180 }, { "grad_norm": 2.024138927459717, "learning_rate": 2.3780000000000003e-05, "loss": 0.1729, "step": 1190 }, { "grad_norm": 3.220212697982788, "learning_rate": 2.398e-05, "loss": 0.1877, "step": 1200 }, { "grad_norm": 3.8860487937927246, "learning_rate": 2.418e-05, "loss": 0.2103, "step": 1210 }, { "grad_norm": 1.9243223667144775, "learning_rate": 2.438e-05, "loss": 0.1833, "step": 1220 }, { "grad_norm": 2.2182934284210205, "learning_rate": 2.4580000000000002e-05, "loss": 0.1831, "step": 1230 }, { "grad_norm": 1.953171968460083, "learning_rate": 2.478e-05, "loss": 0.1886, "step": 1240 }, { "grad_norm": 2.327004909515381, "learning_rate": 2.498e-05, "loss": 0.1672, "step": 1250 }, { "grad_norm": 2.3172519207000732, "learning_rate": 2.5180000000000003e-05, "loss": 0.1779, "step": 1260 }, { "grad_norm": 2.0873301029205322, "learning_rate": 2.5380000000000004e-05, "loss": 0.1518, "step": 1270 }, { "grad_norm": 1.4226692914962769, "learning_rate": 2.5580000000000002e-05, "loss": 0.171, "step": 1280 }, { "grad_norm": 1.731871247291565, "learning_rate": 2.5779999999999997e-05, "loss": 0.1808, "step": 1290 }, { "grad_norm": 1.844114899635315, "learning_rate": 2.598e-05, "loss": 0.1615, "step": 1300 }, { "grad_norm": 2.1101784706115723, "learning_rate": 2.618e-05, "loss": 0.1565, "step": 1310 }, { "grad_norm": 1.9808052778244019, "learning_rate": 2.6379999999999998e-05, "loss": 0.1649, "step": 1320 }, { "grad_norm": 1.8760545253753662, "learning_rate": 2.658e-05, "loss": 0.1416, "step": 1330 }, { "grad_norm": 2.2632837295532227, "learning_rate": 2.678e-05, "loss": 0.1442, "step": 1340 }, { "grad_norm": 2.345886468887329, "learning_rate": 2.698e-05, "loss": 0.1804, "step": 1350 }, { "grad_norm": 2.3369407653808594, "learning_rate": 2.718e-05, "loss": 0.1653, "step": 1360 }, { "grad_norm": 2.395482301712036, "learning_rate": 2.738e-05, "loss": 0.1401, "step": 1370 }, { "grad_norm": 1.7876805067062378, "learning_rate": 2.758e-05, "loss": 0.1438, "step": 1380 }, { "grad_norm": 2.405266046524048, "learning_rate": 2.778e-05, "loss": 0.1541, "step": 1390 }, { "grad_norm": 2.2028887271881104, "learning_rate": 2.798e-05, "loss": 0.1462, "step": 1400 }, { "grad_norm": 1.7188876867294312, "learning_rate": 2.818e-05, "loss": 0.1575, "step": 1410 }, { "grad_norm": 2.1337013244628906, "learning_rate": 2.8380000000000003e-05, "loss": 0.15, "step": 1420 }, { "grad_norm": 1.9807653427124023, "learning_rate": 2.858e-05, "loss": 0.1292, "step": 1430 }, { "grad_norm": 1.7444109916687012, "learning_rate": 2.8780000000000002e-05, "loss": 0.1609, "step": 1440 }, { "grad_norm": 1.635120153427124, "learning_rate": 2.898e-05, "loss": 0.1695, "step": 1450 }, { "grad_norm": 1.4914056062698364, "learning_rate": 2.9180000000000002e-05, "loss": 0.1586, "step": 1460 }, { "grad_norm": 1.5270581245422363, "learning_rate": 2.9380000000000003e-05, "loss": 0.1502, "step": 1470 }, { "grad_norm": 1.6483925580978394, "learning_rate": 2.958e-05, "loss": 0.1404, "step": 1480 }, { "grad_norm": 1.6925885677337646, "learning_rate": 2.9780000000000003e-05, "loss": 0.1481, "step": 1490 }, { "grad_norm": 1.0536781549453735, "learning_rate": 2.998e-05, "loss": 0.1546, "step": 1500 }, { "grad_norm": 1.405958652496338, "learning_rate": 3.0180000000000002e-05, "loss": 0.1248, "step": 1510 }, { "grad_norm": 1.9017162322998047, "learning_rate": 3.0380000000000004e-05, "loss": 0.141, "step": 1520 }, { "grad_norm": 1.4677672386169434, "learning_rate": 3.058e-05, "loss": 0.1489, "step": 1530 }, { "grad_norm": 1.7075588703155518, "learning_rate": 3.078e-05, "loss": 0.1355, "step": 1540 }, { "grad_norm": 1.7219197750091553, "learning_rate": 3.0980000000000005e-05, "loss": 0.1373, "step": 1550 }, { "grad_norm": 1.4989590644836426, "learning_rate": 3.118e-05, "loss": 0.1564, "step": 1560 }, { "grad_norm": 1.3752597570419312, "learning_rate": 3.138e-05, "loss": 0.1194, "step": 1570 }, { "grad_norm": 1.6655077934265137, "learning_rate": 3.1580000000000006e-05, "loss": 0.1352, "step": 1580 }, { "grad_norm": 1.8171226978302002, "learning_rate": 3.1780000000000004e-05, "loss": 0.1417, "step": 1590 }, { "grad_norm": 1.4073342084884644, "learning_rate": 3.198e-05, "loss": 0.1252, "step": 1600 }, { "grad_norm": 1.3610535860061646, "learning_rate": 3.218e-05, "loss": 0.1312, "step": 1610 }, { "grad_norm": 1.7262972593307495, "learning_rate": 3.238e-05, "loss": 0.1515, "step": 1620 }, { "grad_norm": 2.1511173248291016, "learning_rate": 3.2579999999999996e-05, "loss": 0.1882, "step": 1630 }, { "grad_norm": 1.2590535879135132, "learning_rate": 3.278e-05, "loss": 0.146, "step": 1640 }, { "grad_norm": 1.7426694631576538, "learning_rate": 3.298e-05, "loss": 0.1599, "step": 1650 }, { "grad_norm": 1.6521662473678589, "learning_rate": 3.318e-05, "loss": 0.149, "step": 1660 }, { "grad_norm": 1.214168906211853, "learning_rate": 3.338e-05, "loss": 0.1235, "step": 1670 }, { "grad_norm": 1.9543861150741577, "learning_rate": 3.358e-05, "loss": 0.1389, "step": 1680 }, { "grad_norm": 1.6471354961395264, "learning_rate": 3.378e-05, "loss": 0.1418, "step": 1690 }, { "grad_norm": 1.35808265209198, "learning_rate": 3.398e-05, "loss": 0.1525, "step": 1700 }, { "grad_norm": 1.7626267671585083, "learning_rate": 3.418e-05, "loss": 0.1829, "step": 1710 }, { "grad_norm": 1.5621792078018188, "learning_rate": 3.438e-05, "loss": 0.1535, "step": 1720 }, { "grad_norm": 1.1318283081054688, "learning_rate": 3.4580000000000004e-05, "loss": 0.1392, "step": 1730 }, { "grad_norm": 1.680284857749939, "learning_rate": 3.478e-05, "loss": 0.1294, "step": 1740 }, { "grad_norm": 1.6954269409179688, "learning_rate": 3.498e-05, "loss": 0.1412, "step": 1750 }, { "grad_norm": 1.8102375268936157, "learning_rate": 3.518e-05, "loss": 0.1358, "step": 1760 }, { "grad_norm": 1.121153473854065, "learning_rate": 3.5380000000000003e-05, "loss": 0.1256, "step": 1770 }, { "grad_norm": 1.5703797340393066, "learning_rate": 3.558e-05, "loss": 0.1514, "step": 1780 }, { "grad_norm": 1.4536999464035034, "learning_rate": 3.578e-05, "loss": 0.1335, "step": 1790 }, { "grad_norm": 1.4852492809295654, "learning_rate": 3.5980000000000004e-05, "loss": 0.1447, "step": 1800 }, { "grad_norm": 1.0509697198867798, "learning_rate": 3.618e-05, "loss": 0.1401, "step": 1810 }, { "grad_norm": 1.9936548471450806, "learning_rate": 3.638e-05, "loss": 0.1416, "step": 1820 }, { "grad_norm": 2.0499610900878906, "learning_rate": 3.6580000000000006e-05, "loss": 0.1373, "step": 1830 }, { "grad_norm": 1.5279027223587036, "learning_rate": 3.6780000000000004e-05, "loss": 0.1498, "step": 1840 }, { "grad_norm": 1.3684357404708862, "learning_rate": 3.698e-05, "loss": 0.1684, "step": 1850 }, { "grad_norm": 1.8808075189590454, "learning_rate": 3.7180000000000007e-05, "loss": 0.17, "step": 1860 }, { "grad_norm": 1.3157434463500977, "learning_rate": 3.7380000000000005e-05, "loss": 0.1459, "step": 1870 }, { "grad_norm": 1.1820274591445923, "learning_rate": 3.758e-05, "loss": 0.1128, "step": 1880 }, { "grad_norm": 1.5040606260299683, "learning_rate": 3.778000000000001e-05, "loss": 0.1332, "step": 1890 }, { "grad_norm": 1.3454407453536987, "learning_rate": 3.7980000000000006e-05, "loss": 0.1249, "step": 1900 }, { "grad_norm": 1.8486157655715942, "learning_rate": 3.818e-05, "loss": 0.1723, "step": 1910 }, { "grad_norm": 1.371432900428772, "learning_rate": 3.838e-05, "loss": 0.1455, "step": 1920 }, { "grad_norm": 1.1388105154037476, "learning_rate": 3.858e-05, "loss": 0.1181, "step": 1930 }, { "grad_norm": 1.612823247909546, "learning_rate": 3.878e-05, "loss": 0.1379, "step": 1940 }, { "grad_norm": 1.19147789478302, "learning_rate": 3.898e-05, "loss": 0.1105, "step": 1950 }, { "grad_norm": 1.821384310722351, "learning_rate": 3.918e-05, "loss": 0.1524, "step": 1960 }, { "grad_norm": 1.404465913772583, "learning_rate": 3.938e-05, "loss": 0.1065, "step": 1970 }, { "grad_norm": 1.4029079675674438, "learning_rate": 3.958e-05, "loss": 0.1569, "step": 1980 }, { "grad_norm": 1.4289780855178833, "learning_rate": 3.978e-05, "loss": 0.1146, "step": 1990 }, { "grad_norm": 1.3185495138168335, "learning_rate": 3.998e-05, "loss": 0.1375, "step": 2000 }, { "grad_norm": 1.4377394914627075, "learning_rate": 4.018e-05, "loss": 0.1705, "step": 2010 }, { "grad_norm": 1.3897649049758911, "learning_rate": 4.038e-05, "loss": 0.1345, "step": 2020 }, { "grad_norm": 1.4261077642440796, "learning_rate": 4.058e-05, "loss": 0.1254, "step": 2030 }, { "grad_norm": 1.6259567737579346, "learning_rate": 4.078e-05, "loss": 0.138, "step": 2040 }, { "grad_norm": 1.4632620811462402, "learning_rate": 4.0980000000000004e-05, "loss": 0.1398, "step": 2050 }, { "grad_norm": 1.0062205791473389, "learning_rate": 4.118e-05, "loss": 0.1219, "step": 2060 }, { "grad_norm": 1.2301710844039917, "learning_rate": 4.138e-05, "loss": 0.119, "step": 2070 }, { "grad_norm": 1.5787880420684814, "learning_rate": 4.1580000000000005e-05, "loss": 0.1281, "step": 2080 }, { "grad_norm": 1.4043161869049072, "learning_rate": 4.178e-05, "loss": 0.1366, "step": 2090 }, { "grad_norm": 1.332322120666504, "learning_rate": 4.198e-05, "loss": 0.1317, "step": 2100 }, { "grad_norm": 1.2323939800262451, "learning_rate": 4.2180000000000006e-05, "loss": 0.1534, "step": 2110 }, { "grad_norm": 1.289387583732605, "learning_rate": 4.2380000000000004e-05, "loss": 0.1279, "step": 2120 }, { "grad_norm": 1.1738481521606445, "learning_rate": 4.258e-05, "loss": 0.1173, "step": 2130 }, { "grad_norm": 1.2140624523162842, "learning_rate": 4.278e-05, "loss": 0.1185, "step": 2140 }, { "grad_norm": 1.0176326036453247, "learning_rate": 4.2980000000000005e-05, "loss": 0.1589, "step": 2150 }, { "grad_norm": 1.1877814531326294, "learning_rate": 4.318e-05, "loss": 0.1323, "step": 2160 }, { "grad_norm": 1.3711165189743042, "learning_rate": 4.338e-05, "loss": 0.1214, "step": 2170 }, { "grad_norm": 1.5072064399719238, "learning_rate": 4.3580000000000006e-05, "loss": 0.1484, "step": 2180 }, { "grad_norm": 0.9885004162788391, "learning_rate": 4.3780000000000004e-05, "loss": 0.1254, "step": 2190 }, { "grad_norm": 1.6488767862319946, "learning_rate": 4.398e-05, "loss": 0.1223, "step": 2200 }, { "grad_norm": 1.1923884153366089, "learning_rate": 4.418000000000001e-05, "loss": 0.1276, "step": 2210 }, { "grad_norm": 1.5209479331970215, "learning_rate": 4.438e-05, "loss": 0.1267, "step": 2220 }, { "grad_norm": 1.3945579528808594, "learning_rate": 4.458e-05, "loss": 0.1179, "step": 2230 }, { "grad_norm": 1.3731368780136108, "learning_rate": 4.478e-05, "loss": 0.1215, "step": 2240 }, { "grad_norm": 1.462674856185913, "learning_rate": 4.498e-05, "loss": 0.1359, "step": 2250 }, { "grad_norm": 1.2331199645996094, "learning_rate": 4.518e-05, "loss": 0.1283, "step": 2260 }, { "grad_norm": 1.2018945217132568, "learning_rate": 4.538e-05, "loss": 0.1187, "step": 2270 }, { "grad_norm": 1.147979497909546, "learning_rate": 4.558e-05, "loss": 0.1211, "step": 2280 }, { "grad_norm": 1.5541942119598389, "learning_rate": 4.578e-05, "loss": 0.151, "step": 2290 }, { "grad_norm": 1.4305075407028198, "learning_rate": 4.5980000000000004e-05, "loss": 0.1226, "step": 2300 }, { "grad_norm": 1.3872853517532349, "learning_rate": 4.618e-05, "loss": 0.1227, "step": 2310 }, { "grad_norm": 0.9022251963615417, "learning_rate": 4.638e-05, "loss": 0.1128, "step": 2320 }, { "grad_norm": 1.3897486925125122, "learning_rate": 4.6580000000000005e-05, "loss": 0.1473, "step": 2330 }, { "grad_norm": 1.0895963907241821, "learning_rate": 4.678e-05, "loss": 0.1238, "step": 2340 }, { "grad_norm": 1.0530273914337158, "learning_rate": 4.698e-05, "loss": 0.129, "step": 2350 }, { "grad_norm": 1.0517337322235107, "learning_rate": 4.718e-05, "loss": 0.1307, "step": 2360 }, { "grad_norm": 1.0449447631835938, "learning_rate": 4.7380000000000004e-05, "loss": 0.1254, "step": 2370 }, { "grad_norm": 1.759291410446167, "learning_rate": 4.758e-05, "loss": 0.1297, "step": 2380 }, { "grad_norm": 1.0336594581604004, "learning_rate": 4.778e-05, "loss": 0.1405, "step": 2390 }, { "grad_norm": 0.9543883800506592, "learning_rate": 4.7980000000000005e-05, "loss": 0.1427, "step": 2400 }, { "grad_norm": 1.3976881504058838, "learning_rate": 4.818e-05, "loss": 0.1364, "step": 2410 }, { "grad_norm": 1.168747067451477, "learning_rate": 4.838e-05, "loss": 0.1648, "step": 2420 }, { "grad_norm": 1.1677969694137573, "learning_rate": 4.8580000000000006e-05, "loss": 0.1218, "step": 2430 }, { "grad_norm": 1.3498605489730835, "learning_rate": 4.8780000000000004e-05, "loss": 0.1477, "step": 2440 }, { "grad_norm": 1.3831976652145386, "learning_rate": 4.898e-05, "loss": 0.1142, "step": 2450 }, { "grad_norm": 0.8940454125404358, "learning_rate": 4.918000000000001e-05, "loss": 0.1147, "step": 2460 }, { "grad_norm": 1.2059282064437866, "learning_rate": 4.9380000000000005e-05, "loss": 0.1218, "step": 2470 }, { "grad_norm": 1.2371861934661865, "learning_rate": 4.958e-05, "loss": 0.1523, "step": 2480 }, { "grad_norm": 1.4224227666854858, "learning_rate": 4.978e-05, "loss": 0.1258, "step": 2490 }, { "grad_norm": 1.0336079597473145, "learning_rate": 4.9980000000000006e-05, "loss": 0.1147, "step": 2500 }, { "grad_norm": 0.910637378692627, "learning_rate": 5.0180000000000004e-05, "loss": 0.1135, "step": 2510 }, { "grad_norm": 0.8720124363899231, "learning_rate": 5.038e-05, "loss": 0.1206, "step": 2520 }, { "grad_norm": 1.3865721225738525, "learning_rate": 5.058000000000001e-05, "loss": 0.1124, "step": 2530 }, { "grad_norm": 0.8348996043205261, "learning_rate": 5.0780000000000005e-05, "loss": 0.1197, "step": 2540 }, { "grad_norm": 1.196441650390625, "learning_rate": 5.098e-05, "loss": 0.1245, "step": 2550 }, { "grad_norm": 1.2119029760360718, "learning_rate": 5.118000000000001e-05, "loss": 0.1417, "step": 2560 }, { "grad_norm": 1.2241936922073364, "learning_rate": 5.1380000000000006e-05, "loss": 0.1208, "step": 2570 }, { "grad_norm": 1.0189845561981201, "learning_rate": 5.1580000000000004e-05, "loss": 0.1299, "step": 2580 }, { "grad_norm": 0.8876668214797974, "learning_rate": 5.178000000000001e-05, "loss": 0.1204, "step": 2590 }, { "grad_norm": 0.832863450050354, "learning_rate": 5.198000000000001e-05, "loss": 0.1376, "step": 2600 }, { "grad_norm": 0.9820502400398254, "learning_rate": 5.2180000000000005e-05, "loss": 0.1178, "step": 2610 }, { "grad_norm": 0.8422629237174988, "learning_rate": 5.238000000000001e-05, "loss": 0.1089, "step": 2620 }, { "grad_norm": 1.0422873497009277, "learning_rate": 5.258000000000001e-05, "loss": 0.108, "step": 2630 }, { "grad_norm": 1.176374912261963, "learning_rate": 5.2780000000000006e-05, "loss": 0.1377, "step": 2640 }, { "grad_norm": 0.9370591044425964, "learning_rate": 5.2980000000000004e-05, "loss": 0.1276, "step": 2650 }, { "grad_norm": 1.2153502702713013, "learning_rate": 5.318000000000001e-05, "loss": 0.1279, "step": 2660 }, { "grad_norm": 1.0334415435791016, "learning_rate": 5.338000000000001e-05, "loss": 0.1399, "step": 2670 }, { "grad_norm": 1.1552765369415283, "learning_rate": 5.3580000000000005e-05, "loss": 0.1545, "step": 2680 }, { "grad_norm": 1.0800364017486572, "learning_rate": 5.378e-05, "loss": 0.1238, "step": 2690 }, { "grad_norm": 1.0979945659637451, "learning_rate": 5.3979999999999995e-05, "loss": 0.1048, "step": 2700 }, { "grad_norm": 0.9427095055580139, "learning_rate": 5.418e-05, "loss": 0.115, "step": 2710 }, { "grad_norm": 0.8426902890205383, "learning_rate": 5.438e-05, "loss": 0.1417, "step": 2720 }, { "grad_norm": 1.2854052782058716, "learning_rate": 5.4579999999999996e-05, "loss": 0.1158, "step": 2730 }, { "grad_norm": 0.7644017338752747, "learning_rate": 5.478e-05, "loss": 0.1173, "step": 2740 }, { "grad_norm": 0.9505190849304199, "learning_rate": 5.498e-05, "loss": 0.122, "step": 2750 }, { "grad_norm": 1.1244020462036133, "learning_rate": 5.518e-05, "loss": 0.1143, "step": 2760 }, { "grad_norm": 0.9162235856056213, "learning_rate": 5.538e-05, "loss": 0.1274, "step": 2770 }, { "grad_norm": 1.006614327430725, "learning_rate": 5.558e-05, "loss": 0.1286, "step": 2780 }, { "grad_norm": 0.8882136344909668, "learning_rate": 5.578e-05, "loss": 0.1179, "step": 2790 }, { "grad_norm": 0.9471865296363831, "learning_rate": 5.5979999999999996e-05, "loss": 0.101, "step": 2800 }, { "grad_norm": 1.2137210369110107, "learning_rate": 5.618e-05, "loss": 0.1506, "step": 2810 }, { "grad_norm": 1.3932740688323975, "learning_rate": 5.638e-05, "loss": 0.1203, "step": 2820 }, { "grad_norm": 0.7280333638191223, "learning_rate": 5.658e-05, "loss": 0.1112, "step": 2830 }, { "grad_norm": 0.9369848966598511, "learning_rate": 5.678e-05, "loss": 0.1392, "step": 2840 }, { "grad_norm": 1.1433457136154175, "learning_rate": 5.698e-05, "loss": 0.1237, "step": 2850 }, { "grad_norm": 1.0469534397125244, "learning_rate": 5.718e-05, "loss": 0.1167, "step": 2860 }, { "grad_norm": 0.8998421430587769, "learning_rate": 5.738e-05, "loss": 0.0931, "step": 2870 }, { "grad_norm": 0.9915281534194946, "learning_rate": 5.758e-05, "loss": 0.1034, "step": 2880 }, { "grad_norm": 1.1016395092010498, "learning_rate": 5.778e-05, "loss": 0.1118, "step": 2890 }, { "grad_norm": 0.9247371554374695, "learning_rate": 5.7980000000000004e-05, "loss": 0.1219, "step": 2900 }, { "grad_norm": 1.0784991979599, "learning_rate": 5.818e-05, "loss": 0.1138, "step": 2910 }, { "grad_norm": 0.8966009020805359, "learning_rate": 5.838e-05, "loss": 0.1191, "step": 2920 }, { "grad_norm": 1.0337107181549072, "learning_rate": 5.858e-05, "loss": 0.0932, "step": 2930 }, { "grad_norm": 1.082211971282959, "learning_rate": 5.878e-05, "loss": 0.1129, "step": 2940 }, { "grad_norm": 1.1634011268615723, "learning_rate": 5.898e-05, "loss": 0.1046, "step": 2950 }, { "grad_norm": 0.6972677111625671, "learning_rate": 5.918e-05, "loss": 0.1052, "step": 2960 }, { "grad_norm": 0.9229133129119873, "learning_rate": 5.9380000000000004e-05, "loss": 0.0957, "step": 2970 }, { "grad_norm": 0.9163077473640442, "learning_rate": 5.958e-05, "loss": 0.1154, "step": 2980 }, { "grad_norm": 0.9731354713439941, "learning_rate": 5.978e-05, "loss": 0.1218, "step": 2990 }, { "grad_norm": 0.900286853313446, "learning_rate": 5.9980000000000005e-05, "loss": 0.1202, "step": 3000 }, { "grad_norm": 0.8344378471374512, "learning_rate": 6.018e-05, "loss": 0.0978, "step": 3010 }, { "grad_norm": 0.8164976239204407, "learning_rate": 6.038e-05, "loss": 0.106, "step": 3020 }, { "grad_norm": 0.9403643012046814, "learning_rate": 6.0580000000000006e-05, "loss": 0.1058, "step": 3030 }, { "grad_norm": 0.98274165391922, "learning_rate": 6.0780000000000004e-05, "loss": 0.109, "step": 3040 }, { "grad_norm": 1.0599340200424194, "learning_rate": 6.098e-05, "loss": 0.1219, "step": 3050 }, { "grad_norm": 0.6672210097312927, "learning_rate": 6.118000000000001e-05, "loss": 0.1148, "step": 3060 }, { "grad_norm": 0.8277584314346313, "learning_rate": 6.138e-05, "loss": 0.1008, "step": 3070 }, { "grad_norm": 0.8409461975097656, "learning_rate": 6.158e-05, "loss": 0.1144, "step": 3080 }, { "grad_norm": 1.078061580657959, "learning_rate": 6.178000000000001e-05, "loss": 0.125, "step": 3090 }, { "grad_norm": 1.0554660558700562, "learning_rate": 6.198e-05, "loss": 0.1092, "step": 3100 }, { "grad_norm": 0.7911130785942078, "learning_rate": 6.218e-05, "loss": 0.0926, "step": 3110 }, { "grad_norm": 0.9909573793411255, "learning_rate": 6.238000000000001e-05, "loss": 0.1002, "step": 3120 }, { "grad_norm": 0.7699849009513855, "learning_rate": 6.258e-05, "loss": 0.093, "step": 3130 }, { "grad_norm": 0.8656036257743835, "learning_rate": 6.278e-05, "loss": 0.1163, "step": 3140 }, { "grad_norm": 0.9035831689834595, "learning_rate": 6.298000000000001e-05, "loss": 0.1068, "step": 3150 }, { "grad_norm": 0.8729832172393799, "learning_rate": 6.318e-05, "loss": 0.1092, "step": 3160 }, { "grad_norm": 0.7984445095062256, "learning_rate": 6.338e-05, "loss": 0.0934, "step": 3170 }, { "grad_norm": 0.7719618678092957, "learning_rate": 6.358000000000001e-05, "loss": 0.1173, "step": 3180 }, { "grad_norm": 0.8543338179588318, "learning_rate": 6.378e-05, "loss": 0.1106, "step": 3190 }, { "grad_norm": 0.8806470036506653, "learning_rate": 6.398000000000001e-05, "loss": 0.1195, "step": 3200 }, { "grad_norm": 0.873964250087738, "learning_rate": 6.418000000000001e-05, "loss": 0.0951, "step": 3210 }, { "grad_norm": 1.279923677444458, "learning_rate": 6.438e-05, "loss": 0.1386, "step": 3220 }, { "grad_norm": 0.802855372428894, "learning_rate": 6.458000000000001e-05, "loss": 0.1148, "step": 3230 }, { "grad_norm": 0.5675228834152222, "learning_rate": 6.478000000000001e-05, "loss": 0.0911, "step": 3240 }, { "grad_norm": 0.9144189357757568, "learning_rate": 6.498e-05, "loss": 0.1325, "step": 3250 }, { "grad_norm": 0.9465150833129883, "learning_rate": 6.518000000000001e-05, "loss": 0.1283, "step": 3260 }, { "grad_norm": 1.0044598579406738, "learning_rate": 6.538000000000001e-05, "loss": 0.1124, "step": 3270 }, { "grad_norm": 0.9608498215675354, "learning_rate": 6.558e-05, "loss": 0.0979, "step": 3280 }, { "grad_norm": 0.7287391424179077, "learning_rate": 6.578000000000001e-05, "loss": 0.0989, "step": 3290 }, { "grad_norm": 0.7430444955825806, "learning_rate": 6.598e-05, "loss": 0.098, "step": 3300 }, { "grad_norm": 1.057220458984375, "learning_rate": 6.618e-05, "loss": 0.1154, "step": 3310 }, { "grad_norm": 0.6688080430030823, "learning_rate": 6.638e-05, "loss": 0.13, "step": 3320 }, { "grad_norm": 0.6194174885749817, "learning_rate": 6.658e-05, "loss": 0.1101, "step": 3330 }, { "grad_norm": 1.0397124290466309, "learning_rate": 6.678e-05, "loss": 0.1157, "step": 3340 }, { "grad_norm": 0.7526581287384033, "learning_rate": 6.698e-05, "loss": 0.1063, "step": 3350 }, { "grad_norm": 0.9489445090293884, "learning_rate": 6.718e-05, "loss": 0.1029, "step": 3360 }, { "grad_norm": 0.8876970410346985, "learning_rate": 6.738e-05, "loss": 0.1204, "step": 3370 }, { "grad_norm": 0.9390643835067749, "learning_rate": 6.758e-05, "loss": 0.1601, "step": 3380 }, { "grad_norm": 0.9390602111816406, "learning_rate": 6.778e-05, "loss": 0.0972, "step": 3390 }, { "grad_norm": 0.846636950969696, "learning_rate": 6.798e-05, "loss": 0.1082, "step": 3400 }, { "grad_norm": 1.3277223110198975, "learning_rate": 6.818e-05, "loss": 0.1206, "step": 3410 }, { "grad_norm": 0.9754562973976135, "learning_rate": 6.838e-05, "loss": 0.1036, "step": 3420 }, { "grad_norm": 0.7597177028656006, "learning_rate": 6.858e-05, "loss": 0.1098, "step": 3430 }, { "grad_norm": 0.9757480621337891, "learning_rate": 6.878e-05, "loss": 0.1253, "step": 3440 }, { "grad_norm": 0.9932282567024231, "learning_rate": 6.898e-05, "loss": 0.1123, "step": 3450 }, { "grad_norm": 0.8488131761550903, "learning_rate": 6.918e-05, "loss": 0.1013, "step": 3460 }, { "grad_norm": 0.7056899070739746, "learning_rate": 6.938e-05, "loss": 0.131, "step": 3470 }, { "grad_norm": 0.8639413714408875, "learning_rate": 6.958e-05, "loss": 0.119, "step": 3480 }, { "grad_norm": 0.8861178159713745, "learning_rate": 6.978e-05, "loss": 0.1203, "step": 3490 }, { "grad_norm": 0.9692727327346802, "learning_rate": 6.998e-05, "loss": 0.0912, "step": 3500 }, { "grad_norm": 0.9439420104026794, "learning_rate": 7.018e-05, "loss": 0.1063, "step": 3510 }, { "grad_norm": 0.6855533719062805, "learning_rate": 7.038e-05, "loss": 0.1055, "step": 3520 }, { "grad_norm": 0.8220735192298889, "learning_rate": 7.058e-05, "loss": 0.1056, "step": 3530 }, { "grad_norm": 0.8382101655006409, "learning_rate": 7.078e-05, "loss": 0.1109, "step": 3540 }, { "grad_norm": 1.2593344449996948, "learning_rate": 7.098e-05, "loss": 0.1216, "step": 3550 }, { "grad_norm": 0.6104736924171448, "learning_rate": 7.118e-05, "loss": 0.0921, "step": 3560 }, { "grad_norm": 0.9526126384735107, "learning_rate": 7.138e-05, "loss": 0.1172, "step": 3570 }, { "grad_norm": 0.9669556617736816, "learning_rate": 7.158e-05, "loss": 0.1217, "step": 3580 }, { "grad_norm": 0.9169387817382812, "learning_rate": 7.178000000000001e-05, "loss": 0.1164, "step": 3590 }, { "grad_norm": 0.7232251167297363, "learning_rate": 7.198e-05, "loss": 0.1081, "step": 3600 }, { "grad_norm": 0.8397467732429504, "learning_rate": 7.218e-05, "loss": 0.0996, "step": 3610 }, { "grad_norm": 0.8312655687332153, "learning_rate": 7.238000000000001e-05, "loss": 0.106, "step": 3620 }, { "grad_norm": 0.7039699554443359, "learning_rate": 7.258e-05, "loss": 0.0903, "step": 3630 }, { "grad_norm": 0.7526912093162537, "learning_rate": 7.278e-05, "loss": 0.1013, "step": 3640 }, { "grad_norm": 0.9088369607925415, "learning_rate": 7.298000000000001e-05, "loss": 0.0979, "step": 3650 }, { "grad_norm": 0.8803685307502747, "learning_rate": 7.318e-05, "loss": 0.1019, "step": 3660 }, { "grad_norm": 1.1012334823608398, "learning_rate": 7.338e-05, "loss": 0.1207, "step": 3670 }, { "grad_norm": 0.7379539608955383, "learning_rate": 7.358000000000001e-05, "loss": 0.1099, "step": 3680 }, { "grad_norm": 0.9585040807723999, "learning_rate": 7.378e-05, "loss": 0.1105, "step": 3690 }, { "grad_norm": 0.7948680520057678, "learning_rate": 7.398e-05, "loss": 0.1091, "step": 3700 }, { "grad_norm": 0.80158931016922, "learning_rate": 7.418000000000001e-05, "loss": 0.1035, "step": 3710 }, { "grad_norm": 0.6940940022468567, "learning_rate": 7.438e-05, "loss": 0.1126, "step": 3720 }, { "grad_norm": 0.8023843765258789, "learning_rate": 7.458000000000001e-05, "loss": 0.0937, "step": 3730 }, { "grad_norm": 0.7796899676322937, "learning_rate": 7.478e-05, "loss": 0.1069, "step": 3740 }, { "grad_norm": 0.7765525579452515, "learning_rate": 7.498e-05, "loss": 0.1305, "step": 3750 }, { "grad_norm": 0.9817521572113037, "learning_rate": 7.518000000000001e-05, "loss": 0.1017, "step": 3760 }, { "grad_norm": 0.7326626181602478, "learning_rate": 7.538e-05, "loss": 0.1136, "step": 3770 }, { "grad_norm": 0.8028500080108643, "learning_rate": 7.558e-05, "loss": 0.1092, "step": 3780 }, { "grad_norm": 0.6716075539588928, "learning_rate": 7.578000000000001e-05, "loss": 0.0994, "step": 3790 }, { "grad_norm": 0.681649386882782, "learning_rate": 7.598e-05, "loss": 0.1335, "step": 3800 }, { "grad_norm": 0.7243669629096985, "learning_rate": 7.618e-05, "loss": 0.0888, "step": 3810 }, { "grad_norm": 0.8934809565544128, "learning_rate": 7.638000000000001e-05, "loss": 0.1112, "step": 3820 }, { "grad_norm": 0.8964493274688721, "learning_rate": 7.658e-05, "loss": 0.0984, "step": 3830 }, { "grad_norm": 0.6626037955284119, "learning_rate": 7.678000000000001e-05, "loss": 0.0972, "step": 3840 }, { "grad_norm": 0.770707368850708, "learning_rate": 7.698000000000001e-05, "loss": 0.1161, "step": 3850 }, { "grad_norm": 0.8772009015083313, "learning_rate": 7.718e-05, "loss": 0.0947, "step": 3860 }, { "grad_norm": 1.0501813888549805, "learning_rate": 7.738000000000001e-05, "loss": 0.0934, "step": 3870 }, { "grad_norm": 0.7093860507011414, "learning_rate": 7.758000000000001e-05, "loss": 0.1037, "step": 3880 }, { "grad_norm": 0.8624283075332642, "learning_rate": 7.778e-05, "loss": 0.103, "step": 3890 }, { "grad_norm": 0.933300793170929, "learning_rate": 7.798000000000001e-05, "loss": 0.113, "step": 3900 }, { "grad_norm": 0.7580389976501465, "learning_rate": 7.818000000000001e-05, "loss": 0.1141, "step": 3910 }, { "grad_norm": 0.9311630129814148, "learning_rate": 7.838e-05, "loss": 0.1159, "step": 3920 }, { "grad_norm": 0.7202032208442688, "learning_rate": 7.858000000000001e-05, "loss": 0.0915, "step": 3930 }, { "grad_norm": 0.8528405427932739, "learning_rate": 7.878e-05, "loss": 0.1155, "step": 3940 }, { "grad_norm": 0.7166099548339844, "learning_rate": 7.897999999999999e-05, "loss": 0.0994, "step": 3950 }, { "grad_norm": 0.9760552048683167, "learning_rate": 7.918e-05, "loss": 0.1223, "step": 3960 }, { "grad_norm": 0.8483225703239441, "learning_rate": 7.938e-05, "loss": 0.1012, "step": 3970 }, { "grad_norm": 0.7632738947868347, "learning_rate": 7.958e-05, "loss": 0.1073, "step": 3980 }, { "grad_norm": 0.9263426065444946, "learning_rate": 7.978e-05, "loss": 0.1121, "step": 3990 }, { "grad_norm": 1.0140142440795898, "learning_rate": 7.998e-05, "loss": 0.1027, "step": 4000 }, { "grad_norm": 0.8083109855651855, "learning_rate": 8.018e-05, "loss": 0.1069, "step": 4010 }, { "grad_norm": 0.5014306306838989, "learning_rate": 8.038e-05, "loss": 0.1012, "step": 4020 }, { "grad_norm": 1.1413568258285522, "learning_rate": 8.058e-05, "loss": 0.1147, "step": 4030 }, { "grad_norm": 0.5702605843544006, "learning_rate": 8.078e-05, "loss": 0.0804, "step": 4040 }, { "grad_norm": 0.8700293898582458, "learning_rate": 8.098e-05, "loss": 0.1032, "step": 4050 }, { "grad_norm": 0.7615049481391907, "learning_rate": 8.118e-05, "loss": 0.0961, "step": 4060 }, { "grad_norm": 0.745336651802063, "learning_rate": 8.138e-05, "loss": 0.0952, "step": 4070 }, { "grad_norm": 0.6941158771514893, "learning_rate": 8.158e-05, "loss": 0.1091, "step": 4080 }, { "grad_norm": 0.8891447186470032, "learning_rate": 8.178e-05, "loss": 0.1049, "step": 4090 }, { "grad_norm": 0.8761116862297058, "learning_rate": 8.198e-05, "loss": 0.1015, "step": 4100 }, { "grad_norm": 0.6348588466644287, "learning_rate": 8.218e-05, "loss": 0.1099, "step": 4110 }, { "grad_norm": 0.7438332438468933, "learning_rate": 8.238000000000001e-05, "loss": 0.1101, "step": 4120 }, { "grad_norm": 0.6999824047088623, "learning_rate": 8.258e-05, "loss": 0.0989, "step": 4130 }, { "grad_norm": 0.7480869293212891, "learning_rate": 8.278e-05, "loss": 0.111, "step": 4140 }, { "grad_norm": 0.6638745069503784, "learning_rate": 8.298000000000001e-05, "loss": 0.116, "step": 4150 }, { "grad_norm": 0.8120664954185486, "learning_rate": 8.318e-05, "loss": 0.1019, "step": 4160 }, { "grad_norm": 0.7075103521347046, "learning_rate": 8.338e-05, "loss": 0.0939, "step": 4170 }, { "grad_norm": 0.6434063911437988, "learning_rate": 8.358e-05, "loss": 0.0881, "step": 4180 }, { "grad_norm": 0.7025808095932007, "learning_rate": 8.378e-05, "loss": 0.0921, "step": 4190 }, { "grad_norm": 0.83742356300354, "learning_rate": 8.398e-05, "loss": 0.0967, "step": 4200 }, { "grad_norm": 0.8349822759628296, "learning_rate": 8.418e-05, "loss": 0.0936, "step": 4210 }, { "grad_norm": 0.8864701390266418, "learning_rate": 8.438e-05, "loss": 0.1182, "step": 4220 }, { "grad_norm": 0.7498098015785217, "learning_rate": 8.458e-05, "loss": 0.1222, "step": 4230 }, { "grad_norm": 0.7999185919761658, "learning_rate": 8.478e-05, "loss": 0.1149, "step": 4240 }, { "grad_norm": 0.765588104724884, "learning_rate": 8.498e-05, "loss": 0.1085, "step": 4250 }, { "grad_norm": 0.7711727619171143, "learning_rate": 8.518000000000001e-05, "loss": 0.1224, "step": 4260 }, { "grad_norm": 0.7174469232559204, "learning_rate": 8.538e-05, "loss": 0.1064, "step": 4270 }, { "grad_norm": 0.745898425579071, "learning_rate": 8.558e-05, "loss": 0.0871, "step": 4280 }, { "grad_norm": 0.7534245252609253, "learning_rate": 8.578000000000001e-05, "loss": 0.0968, "step": 4290 }, { "grad_norm": 0.6369635462760925, "learning_rate": 8.598e-05, "loss": 0.0868, "step": 4300 }, { "grad_norm": 0.5700401067733765, "learning_rate": 8.618e-05, "loss": 0.103, "step": 4310 }, { "grad_norm": 1.1487196683883667, "learning_rate": 8.638000000000001e-05, "loss": 0.1357, "step": 4320 }, { "grad_norm": 0.7154725790023804, "learning_rate": 8.658e-05, "loss": 0.0958, "step": 4330 }, { "grad_norm": 0.5614255666732788, "learning_rate": 8.678e-05, "loss": 0.0762, "step": 4340 }, { "grad_norm": 0.6083617210388184, "learning_rate": 8.698000000000001e-05, "loss": 0.0819, "step": 4350 }, { "grad_norm": 0.8340036273002625, "learning_rate": 8.718e-05, "loss": 0.0907, "step": 4360 }, { "grad_norm": 0.7896839380264282, "learning_rate": 8.738000000000001e-05, "loss": 0.0966, "step": 4370 }, { "grad_norm": 0.6865161657333374, "learning_rate": 8.758000000000001e-05, "loss": 0.1062, "step": 4380 }, { "grad_norm": 0.9634537100791931, "learning_rate": 8.778e-05, "loss": 0.0943, "step": 4390 }, { "grad_norm": 0.8004398345947266, "learning_rate": 8.798000000000001e-05, "loss": 0.12, "step": 4400 }, { "grad_norm": 0.8572943210601807, "learning_rate": 8.818000000000001e-05, "loss": 0.1194, "step": 4410 }, { "grad_norm": 0.6038874387741089, "learning_rate": 8.838e-05, "loss": 0.1233, "step": 4420 }, { "grad_norm": 0.7817746996879578, "learning_rate": 8.858000000000001e-05, "loss": 0.1092, "step": 4430 }, { "grad_norm": 0.797577977180481, "learning_rate": 8.878000000000001e-05, "loss": 0.1117, "step": 4440 }, { "grad_norm": 0.759456217288971, "learning_rate": 8.898e-05, "loss": 0.1119, "step": 4450 }, { "grad_norm": 0.6757457256317139, "learning_rate": 8.918000000000001e-05, "loss": 0.1006, "step": 4460 }, { "grad_norm": 0.7314025163650513, "learning_rate": 8.938e-05, "loss": 0.0958, "step": 4470 }, { "grad_norm": 0.6493166089057922, "learning_rate": 8.958e-05, "loss": 0.1128, "step": 4480 }, { "grad_norm": 0.7058660984039307, "learning_rate": 8.978000000000001e-05, "loss": 0.1005, "step": 4490 }, { "grad_norm": 0.7296248078346252, "learning_rate": 8.998e-05, "loss": 0.1049, "step": 4500 }, { "grad_norm": 0.9141833782196045, "learning_rate": 9.018000000000001e-05, "loss": 0.1305, "step": 4510 }, { "grad_norm": 0.9337164163589478, "learning_rate": 9.038000000000001e-05, "loss": 0.1074, "step": 4520 }, { "grad_norm": 0.7935041189193726, "learning_rate": 9.058e-05, "loss": 0.11, "step": 4530 }, { "grad_norm": 0.8724528551101685, "learning_rate": 9.078000000000001e-05, "loss": 0.1202, "step": 4540 }, { "grad_norm": 0.8767907619476318, "learning_rate": 9.098000000000001e-05, "loss": 0.1227, "step": 4550 }, { "grad_norm": 0.7133527398109436, "learning_rate": 9.118e-05, "loss": 0.1043, "step": 4560 }, { "grad_norm": 0.9057533740997314, "learning_rate": 9.138e-05, "loss": 0.144, "step": 4570 }, { "grad_norm": 1.0965826511383057, "learning_rate": 9.158e-05, "loss": 0.1205, "step": 4580 }, { "grad_norm": 0.8079681396484375, "learning_rate": 9.178e-05, "loss": 0.1108, "step": 4590 }, { "grad_norm": 0.866610586643219, "learning_rate": 9.198e-05, "loss": 0.1065, "step": 4600 }, { "grad_norm": 0.7064515948295593, "learning_rate": 9.218e-05, "loss": 0.1018, "step": 4610 }, { "grad_norm": 0.7826623916625977, "learning_rate": 9.238e-05, "loss": 0.093, "step": 4620 }, { "grad_norm": 0.8937559723854065, "learning_rate": 9.258e-05, "loss": 0.1007, "step": 4630 }, { "grad_norm": 0.5940081477165222, "learning_rate": 9.278e-05, "loss": 0.1152, "step": 4640 }, { "grad_norm": 0.749444842338562, "learning_rate": 9.298e-05, "loss": 0.0995, "step": 4650 }, { "grad_norm": 0.9059867262840271, "learning_rate": 9.318e-05, "loss": 0.1165, "step": 4660 }, { "grad_norm": 0.891306459903717, "learning_rate": 9.338e-05, "loss": 0.1001, "step": 4670 }, { "grad_norm": 1.1904098987579346, "learning_rate": 9.358e-05, "loss": 0.1012, "step": 4680 }, { "grad_norm": 0.6840314865112305, "learning_rate": 9.378e-05, "loss": 0.108, "step": 4690 }, { "grad_norm": 0.8090885877609253, "learning_rate": 9.398e-05, "loss": 0.1065, "step": 4700 }, { "grad_norm": 0.8332207202911377, "learning_rate": 9.418e-05, "loss": 0.1019, "step": 4710 }, { "grad_norm": 0.640914261341095, "learning_rate": 9.438e-05, "loss": 0.0886, "step": 4720 }, { "grad_norm": 0.5018924474716187, "learning_rate": 9.458e-05, "loss": 0.1137, "step": 4730 }, { "grad_norm": 0.5441907048225403, "learning_rate": 9.478e-05, "loss": 0.1043, "step": 4740 }, { "grad_norm": 1.0358927249908447, "learning_rate": 9.498e-05, "loss": 0.0921, "step": 4750 }, { "grad_norm": 0.9127360582351685, "learning_rate": 9.518000000000001e-05, "loss": 0.1157, "step": 4760 }, { "grad_norm": 0.8484509587287903, "learning_rate": 9.538e-05, "loss": 0.1105, "step": 4770 }, { "grad_norm": 0.7806420922279358, "learning_rate": 9.558e-05, "loss": 0.1107, "step": 4780 }, { "grad_norm": 0.6586634516716003, "learning_rate": 9.578000000000001e-05, "loss": 0.081, "step": 4790 }, { "grad_norm": 0.7671559453010559, "learning_rate": 9.598e-05, "loss": 0.0922, "step": 4800 }, { "grad_norm": 0.7722575664520264, "learning_rate": 9.618e-05, "loss": 0.0988, "step": 4810 }, { "grad_norm": 0.6382466554641724, "learning_rate": 9.638000000000001e-05, "loss": 0.1081, "step": 4820 }, { "grad_norm": 0.5385229587554932, "learning_rate": 9.658e-05, "loss": 0.107, "step": 4830 }, { "grad_norm": 0.9645103216171265, "learning_rate": 9.678e-05, "loss": 0.1037, "step": 4840 }, { "grad_norm": 0.9290231466293335, "learning_rate": 9.698000000000001e-05, "loss": 0.1089, "step": 4850 }, { "grad_norm": 0.7405545115470886, "learning_rate": 9.718e-05, "loss": 0.0957, "step": 4860 }, { "grad_norm": 0.5864213109016418, "learning_rate": 9.738e-05, "loss": 0.0912, "step": 4870 }, { "grad_norm": 0.841910183429718, "learning_rate": 9.758000000000001e-05, "loss": 0.0955, "step": 4880 }, { "grad_norm": 0.7559314370155334, "learning_rate": 9.778e-05, "loss": 0.1204, "step": 4890 }, { "grad_norm": 0.6172708868980408, "learning_rate": 9.798000000000001e-05, "loss": 0.093, "step": 4900 }, { "grad_norm": 0.7110695838928223, "learning_rate": 9.818000000000001e-05, "loss": 0.1029, "step": 4910 }, { "grad_norm": 0.686400830745697, "learning_rate": 9.838e-05, "loss": 0.1046, "step": 4920 }, { "grad_norm": 0.8068904876708984, "learning_rate": 9.858000000000001e-05, "loss": 0.1033, "step": 4930 }, { "grad_norm": 0.6980186700820923, "learning_rate": 9.878e-05, "loss": 0.0978, "step": 4940 }, { "grad_norm": 0.6255984902381897, "learning_rate": 9.898e-05, "loss": 0.1258, "step": 4950 }, { "grad_norm": 0.7020201086997986, "learning_rate": 9.918000000000001e-05, "loss": 0.0961, "step": 4960 }, { "grad_norm": 0.6979461312294006, "learning_rate": 9.938e-05, "loss": 0.1114, "step": 4970 }, { "grad_norm": 0.8358283042907715, "learning_rate": 9.958e-05, "loss": 0.1109, "step": 4980 }, { "grad_norm": 0.744779646396637, "learning_rate": 9.978000000000001e-05, "loss": 0.1129, "step": 4990 }, { "grad_norm": 0.7723989486694336, "learning_rate": 9.998e-05, "loss": 0.1146, "step": 5000 }, { "grad_norm": 1.0969884395599365, "learning_rate": 9.999999778549045e-05, "loss": 0.111, "step": 5010 }, { "grad_norm": 0.7702661752700806, "learning_rate": 9.999999013039593e-05, "loss": 0.1165, "step": 5020 }, { "grad_norm": 0.7742420434951782, "learning_rate": 9.999997700737766e-05, "loss": 0.0975, "step": 5030 }, { "grad_norm": 0.7293052673339844, "learning_rate": 9.999995841643709e-05, "loss": 0.1105, "step": 5040 }, { "grad_norm": 0.6846978068351746, "learning_rate": 9.999993435757623e-05, "loss": 0.1205, "step": 5050 }, { "grad_norm": 0.6268492937088013, "learning_rate": 9.999990483079773e-05, "loss": 0.0931, "step": 5060 }, { "grad_norm": 0.6220290064811707, "learning_rate": 9.999986983610481e-05, "loss": 0.1218, "step": 5070 }, { "grad_norm": 0.6787563562393188, "learning_rate": 9.99998293735013e-05, "loss": 0.1245, "step": 5080 }, { "grad_norm": 0.6503967642784119, "learning_rate": 9.999978344299161e-05, "loss": 0.1009, "step": 5090 }, { "grad_norm": 0.6545968651771545, "learning_rate": 9.99997320445808e-05, "loss": 0.098, "step": 5100 }, { "grad_norm": 0.6110095381736755, "learning_rate": 9.999967517827444e-05, "loss": 0.1019, "step": 5110 }, { "grad_norm": 0.8851807117462158, "learning_rate": 9.999961284407879e-05, "loss": 0.1275, "step": 5120 }, { "grad_norm": 0.6862677931785583, "learning_rate": 9.999954504200067e-05, "loss": 0.1134, "step": 5130 }, { "grad_norm": 0.5891072750091553, "learning_rate": 9.999947177204744e-05, "loss": 0.1089, "step": 5140 }, { "grad_norm": 0.5688311457633972, "learning_rate": 9.999939303422718e-05, "loss": 0.0881, "step": 5150 }, { "grad_norm": 0.7086858749389648, "learning_rate": 9.999930882854847e-05, "loss": 0.1204, "step": 5160 }, { "grad_norm": 0.7990338802337646, "learning_rate": 9.999921915502051e-05, "loss": 0.1095, "step": 5170 }, { "grad_norm": 0.6947263479232788, "learning_rate": 9.99991240136531e-05, "loss": 0.0926, "step": 5180 }, { "grad_norm": 0.9889630079269409, "learning_rate": 9.999902340445668e-05, "loss": 0.13, "step": 5190 }, { "grad_norm": 0.5963397026062012, "learning_rate": 9.999891732744224e-05, "loss": 0.0941, "step": 5200 }, { "grad_norm": 0.603400468826294, "learning_rate": 9.999880578262135e-05, "loss": 0.0961, "step": 5210 }, { "grad_norm": 0.6816743612289429, "learning_rate": 9.999868877000624e-05, "loss": 0.096, "step": 5220 }, { "grad_norm": 0.5581279993057251, "learning_rate": 9.99985662896097e-05, "loss": 0.0877, "step": 5230 }, { "grad_norm": 0.6174178719520569, "learning_rate": 9.999843834144513e-05, "loss": 0.0965, "step": 5240 }, { "grad_norm": 0.7064819931983948, "learning_rate": 9.99983049255265e-05, "loss": 0.1109, "step": 5250 }, { "grad_norm": 0.7581620812416077, "learning_rate": 9.999816604186843e-05, "loss": 0.0965, "step": 5260 }, { "grad_norm": 0.6834184527397156, "learning_rate": 9.999802169048609e-05, "loss": 0.1004, "step": 5270 }, { "grad_norm": 0.6279036402702332, "learning_rate": 9.999787187139527e-05, "loss": 0.1029, "step": 5280 }, { "grad_norm": 0.6425046920776367, "learning_rate": 9.999771658461234e-05, "loss": 0.1137, "step": 5290 }, { "grad_norm": 0.6432327032089233, "learning_rate": 9.999755583015431e-05, "loss": 0.0867, "step": 5300 }, { "grad_norm": 0.6955997943878174, "learning_rate": 9.999738960803874e-05, "loss": 0.1069, "step": 5310 }, { "grad_norm": 0.7088593244552612, "learning_rate": 9.99972179182838e-05, "loss": 0.1109, "step": 5320 }, { "grad_norm": 0.7400577664375305, "learning_rate": 9.99970407609083e-05, "loss": 0.1275, "step": 5330 }, { "grad_norm": 0.7548560500144958, "learning_rate": 9.999685813593159e-05, "loss": 0.1014, "step": 5340 }, { "grad_norm": 0.6466402411460876, "learning_rate": 9.999667004337362e-05, "loss": 0.0946, "step": 5350 }, { "grad_norm": 0.6082696914672852, "learning_rate": 9.9996476483255e-05, "loss": 0.0995, "step": 5360 }, { "grad_norm": 0.6394361853599548, "learning_rate": 9.999627745559688e-05, "loss": 0.0933, "step": 5370 }, { "grad_norm": 0.6196509599685669, "learning_rate": 9.999607296042101e-05, "loss": 0.0814, "step": 5380 }, { "grad_norm": 0.9338778257369995, "learning_rate": 9.99958629977498e-05, "loss": 0.0889, "step": 5390 }, { "grad_norm": 0.5860507488250732, "learning_rate": 9.999564756760615e-05, "loss": 0.0995, "step": 5400 }, { "grad_norm": 0.6433256268501282, "learning_rate": 9.999542667001366e-05, "loss": 0.086, "step": 5410 }, { "grad_norm": 0.7201621532440186, "learning_rate": 9.999520030499647e-05, "loss": 0.0858, "step": 5420 }, { "grad_norm": 0.8324630856513977, "learning_rate": 9.999496847257936e-05, "loss": 0.0937, "step": 5430 }, { "grad_norm": 0.8667988181114197, "learning_rate": 9.999473117278764e-05, "loss": 0.12, "step": 5440 }, { "grad_norm": 0.5630154609680176, "learning_rate": 9.999448840564731e-05, "loss": 0.0783, "step": 5450 }, { "grad_norm": 0.6285483241081238, "learning_rate": 9.999424017118488e-05, "loss": 0.08, "step": 5460 }, { "grad_norm": 0.8230447769165039, "learning_rate": 9.999398646942751e-05, "loss": 0.0812, "step": 5470 }, { "grad_norm": 0.6264998316764832, "learning_rate": 9.999372730040296e-05, "loss": 0.0895, "step": 5480 }, { "grad_norm": 1.0053995847702026, "learning_rate": 9.999346266413953e-05, "loss": 0.1049, "step": 5490 }, { "grad_norm": 0.5845614075660706, "learning_rate": 9.99931925606662e-05, "loss": 0.1045, "step": 5500 }, { "grad_norm": 0.7581496238708496, "learning_rate": 9.99929169900125e-05, "loss": 0.0978, "step": 5510 }, { "grad_norm": 0.507310152053833, "learning_rate": 9.999263595220855e-05, "loss": 0.0887, "step": 5520 }, { "grad_norm": 0.4970761835575104, "learning_rate": 9.99923494472851e-05, "loss": 0.0909, "step": 5530 }, { "grad_norm": 0.6096319556236267, "learning_rate": 9.999205747527348e-05, "loss": 0.0821, "step": 5540 }, { "grad_norm": 0.7704127430915833, "learning_rate": 9.999176003620561e-05, "loss": 0.0881, "step": 5550 }, { "grad_norm": 0.5863221287727356, "learning_rate": 9.999145713011405e-05, "loss": 0.0855, "step": 5560 }, { "grad_norm": 0.7784407734870911, "learning_rate": 9.999114875703186e-05, "loss": 0.0949, "step": 5570 }, { "grad_norm": 0.8343446850776672, "learning_rate": 9.999083491699281e-05, "loss": 0.1165, "step": 5580 }, { "grad_norm": 0.647004246711731, "learning_rate": 9.999051561003123e-05, "loss": 0.0943, "step": 5590 }, { "grad_norm": 0.859916090965271, "learning_rate": 9.999019083618202e-05, "loss": 0.1086, "step": 5600 }, { "grad_norm": 0.6064797639846802, "learning_rate": 9.99898605954807e-05, "loss": 0.0956, "step": 5610 }, { "grad_norm": 0.689315915107727, "learning_rate": 9.998952488796338e-05, "loss": 0.0897, "step": 5620 }, { "grad_norm": 0.5052559971809387, "learning_rate": 9.998918371366676e-05, "loss": 0.0814, "step": 5630 }, { "grad_norm": 0.8035668730735779, "learning_rate": 9.99888370726282e-05, "loss": 0.0951, "step": 5640 }, { "grad_norm": 0.7660919427871704, "learning_rate": 9.998848496488556e-05, "loss": 0.1195, "step": 5650 }, { "grad_norm": 0.6206017136573792, "learning_rate": 9.998812739047736e-05, "loss": 0.086, "step": 5660 }, { "grad_norm": 0.6914339661598206, "learning_rate": 9.99877643494427e-05, "loss": 0.0859, "step": 5670 }, { "grad_norm": 0.7295384407043457, "learning_rate": 9.998739584182128e-05, "loss": 0.1001, "step": 5680 }, { "grad_norm": 0.9374474287033081, "learning_rate": 9.998702186765342e-05, "loss": 0.0904, "step": 5690 }, { "grad_norm": 0.6883256435394287, "learning_rate": 9.998664242698e-05, "loss": 0.0926, "step": 5700 }, { "grad_norm": 0.6491462588310242, "learning_rate": 9.998625751984251e-05, "loss": 0.0914, "step": 5710 }, { "grad_norm": 0.5398384928703308, "learning_rate": 9.998586714628307e-05, "loss": 0.0891, "step": 5720 }, { "grad_norm": 0.6655113697052002, "learning_rate": 9.998547130634432e-05, "loss": 0.0869, "step": 5730 }, { "grad_norm": 0.654100239276886, "learning_rate": 9.99850700000696e-05, "loss": 0.0941, "step": 5740 }, { "grad_norm": 0.776863694190979, "learning_rate": 9.998466322750278e-05, "loss": 0.0834, "step": 5750 }, { "grad_norm": 0.5587904453277588, "learning_rate": 9.998425098868834e-05, "loss": 0.0969, "step": 5760 }, { "grad_norm": 0.6140153408050537, "learning_rate": 9.998383328367136e-05, "loss": 0.0902, "step": 5770 }, { "grad_norm": 0.6267032027244568, "learning_rate": 9.99834101124975e-05, "loss": 0.1035, "step": 5780 }, { "grad_norm": 0.714777946472168, "learning_rate": 9.998298147521309e-05, "loss": 0.0841, "step": 5790 }, { "grad_norm": 0.653083860874176, "learning_rate": 9.998254737186496e-05, "loss": 0.1034, "step": 5800 }, { "grad_norm": 0.5717499256134033, "learning_rate": 9.99821078025006e-05, "loss": 0.0873, "step": 5810 }, { "grad_norm": 0.5630428194999695, "learning_rate": 9.998166276716807e-05, "loss": 0.0911, "step": 5820 }, { "grad_norm": 0.6872549653053284, "learning_rate": 9.998121226591606e-05, "loss": 0.1022, "step": 5830 }, { "grad_norm": 0.5406180024147034, "learning_rate": 9.998075629879382e-05, "loss": 0.0886, "step": 5840 }, { "grad_norm": 0.5570785403251648, "learning_rate": 9.99802948658512e-05, "loss": 0.0747, "step": 5850 }, { "grad_norm": 0.7159712314605713, "learning_rate": 9.99798279671387e-05, "loss": 0.1023, "step": 5860 }, { "grad_norm": 0.48266568779945374, "learning_rate": 9.997935560270734e-05, "loss": 0.0887, "step": 5870 }, { "grad_norm": 0.7281983494758606, "learning_rate": 9.997887777260879e-05, "loss": 0.0968, "step": 5880 }, { "grad_norm": 0.7974919080734253, "learning_rate": 9.997839447689532e-05, "loss": 0.0962, "step": 5890 }, { "grad_norm": 0.6899892091751099, "learning_rate": 9.997790571561978e-05, "loss": 0.1042, "step": 5900 }, { "grad_norm": 0.6175925135612488, "learning_rate": 9.99774114888356e-05, "loss": 0.094, "step": 5910 }, { "grad_norm": 0.7564460039138794, "learning_rate": 9.997691179659684e-05, "loss": 0.0868, "step": 5920 }, { "grad_norm": 0.6369931101799011, "learning_rate": 9.997640663895815e-05, "loss": 0.0799, "step": 5930 }, { "grad_norm": 0.4946756362915039, "learning_rate": 9.997589601597477e-05, "loss": 0.0886, "step": 5940 }, { "grad_norm": 0.6445648670196533, "learning_rate": 9.997537992770252e-05, "loss": 0.0832, "step": 5950 }, { "grad_norm": 0.688501238822937, "learning_rate": 9.997485837419788e-05, "loss": 0.0853, "step": 5960 }, { "grad_norm": 0.7742745280265808, "learning_rate": 9.997433135551786e-05, "loss": 0.0909, "step": 5970 }, { "grad_norm": 0.597673773765564, "learning_rate": 9.997379887172009e-05, "loss": 0.0936, "step": 5980 }, { "grad_norm": 0.7447348237037659, "learning_rate": 9.997326092286281e-05, "loss": 0.0988, "step": 5990 }, { "grad_norm": 0.5331181883811951, "learning_rate": 9.997271750900486e-05, "loss": 0.1152, "step": 6000 }, { "grad_norm": 0.5434675216674805, "learning_rate": 9.997216863020565e-05, "loss": 0.0988, "step": 6010 }, { "grad_norm": 0.6622223854064941, "learning_rate": 9.99716142865252e-05, "loss": 0.0878, "step": 6020 }, { "grad_norm": 0.5879945158958435, "learning_rate": 9.997105447802415e-05, "loss": 0.0848, "step": 6030 }, { "grad_norm": 0.8725520968437195, "learning_rate": 9.997048920476373e-05, "loss": 0.0909, "step": 6040 }, { "grad_norm": 0.6954485774040222, "learning_rate": 9.996991846680572e-05, "loss": 0.0989, "step": 6050 }, { "grad_norm": 0.6651812195777893, "learning_rate": 9.996934226421257e-05, "loss": 0.0917, "step": 6060 }, { "grad_norm": 0.637582004070282, "learning_rate": 9.996876059704726e-05, "loss": 0.0833, "step": 6070 }, { "grad_norm": 0.4903852343559265, "learning_rate": 9.996817346537343e-05, "loss": 0.0912, "step": 6080 }, { "grad_norm": 0.6546767354011536, "learning_rate": 9.996758086925526e-05, "loss": 0.0932, "step": 6090 }, { "grad_norm": 0.6635270714759827, "learning_rate": 9.996698280875759e-05, "loss": 0.1125, "step": 6100 }, { "grad_norm": 0.5080387592315674, "learning_rate": 9.99663792839458e-05, "loss": 0.0799, "step": 6110 }, { "grad_norm": 0.4989612102508545, "learning_rate": 9.99657702948859e-05, "loss": 0.0929, "step": 6120 }, { "grad_norm": 0.6858088374137878, "learning_rate": 9.996515584164448e-05, "loss": 0.0744, "step": 6130 }, { "grad_norm": 0.8967225551605225, "learning_rate": 9.996453592428873e-05, "loss": 0.0813, "step": 6140 }, { "grad_norm": 0.6703090667724609, "learning_rate": 9.996391054288646e-05, "loss": 0.1203, "step": 6150 }, { "grad_norm": 0.6445309519767761, "learning_rate": 9.996327969750605e-05, "loss": 0.1108, "step": 6160 }, { "grad_norm": 0.7047435641288757, "learning_rate": 9.996264338821649e-05, "loss": 0.095, "step": 6170 }, { "grad_norm": 0.545658528804779, "learning_rate": 9.996200161508735e-05, "loss": 0.0817, "step": 6180 }, { "grad_norm": 0.8607694506645203, "learning_rate": 9.996135437818885e-05, "loss": 0.0968, "step": 6190 }, { "grad_norm": 0.5383684635162354, "learning_rate": 9.996070167759175e-05, "loss": 0.0989, "step": 6200 }, { "grad_norm": 0.7374556064605713, "learning_rate": 9.996004351336743e-05, "loss": 0.0943, "step": 6210 }, { "grad_norm": 0.6601603031158447, "learning_rate": 9.995937988558785e-05, "loss": 0.0827, "step": 6220 }, { "grad_norm": 0.5659236311912537, "learning_rate": 9.995871079432561e-05, "loss": 0.0928, "step": 6230 }, { "grad_norm": 0.5971928238868713, "learning_rate": 9.995803623965389e-05, "loss": 0.0949, "step": 6240 }, { "grad_norm": 0.5563113689422607, "learning_rate": 9.995735622164641e-05, "loss": 0.0871, "step": 6250 }, { "grad_norm": 0.5013421177864075, "learning_rate": 9.995667074037758e-05, "loss": 0.0741, "step": 6260 }, { "grad_norm": 0.6382880210876465, "learning_rate": 9.995597979592232e-05, "loss": 0.0917, "step": 6270 }, { "grad_norm": 0.5669400691986084, "learning_rate": 9.995528338835625e-05, "loss": 0.0931, "step": 6280 }, { "grad_norm": 0.7387056946754456, "learning_rate": 9.995458151775547e-05, "loss": 0.0901, "step": 6290 }, { "grad_norm": 0.501754105091095, "learning_rate": 9.995387418419677e-05, "loss": 0.0887, "step": 6300 }, { "grad_norm": 0.8165490627288818, "learning_rate": 9.99531613877575e-05, "loss": 0.0964, "step": 6310 }, { "grad_norm": 0.4989464581012726, "learning_rate": 9.995244312851559e-05, "loss": 0.0835, "step": 6320 }, { "grad_norm": 0.7125663161277771, "learning_rate": 9.995171940654961e-05, "loss": 0.0803, "step": 6330 }, { "grad_norm": 0.7469081282615662, "learning_rate": 9.995099022193871e-05, "loss": 0.1096, "step": 6340 }, { "grad_norm": 0.5375286340713501, "learning_rate": 9.995025557476261e-05, "loss": 0.081, "step": 6350 }, { "grad_norm": 0.7097728252410889, "learning_rate": 9.994951546510165e-05, "loss": 0.0955, "step": 6360 }, { "grad_norm": 0.5166842937469482, "learning_rate": 9.994876989303679e-05, "loss": 0.0897, "step": 6370 }, { "grad_norm": 0.6370508670806885, "learning_rate": 9.994801885864955e-05, "loss": 0.0903, "step": 6380 }, { "grad_norm": 0.899745762348175, "learning_rate": 9.994726236202205e-05, "loss": 0.0963, "step": 6390 }, { "grad_norm": 0.6706723570823669, "learning_rate": 9.994650040323704e-05, "loss": 0.0882, "step": 6400 }, { "grad_norm": 0.5132657885551453, "learning_rate": 9.994573298237784e-05, "loss": 0.0704, "step": 6410 }, { "grad_norm": 0.5712428092956543, "learning_rate": 9.994496009952837e-05, "loss": 0.0764, "step": 6420 }, { "grad_norm": 0.47313395142555237, "learning_rate": 9.994418175477316e-05, "loss": 0.08, "step": 6430 }, { "grad_norm": 0.7608187198638916, "learning_rate": 9.994339794819733e-05, "loss": 0.0866, "step": 6440 }, { "grad_norm": 0.5491443276405334, "learning_rate": 9.994260867988658e-05, "loss": 0.0775, "step": 6450 }, { "grad_norm": 0.49030622839927673, "learning_rate": 9.994181394992723e-05, "loss": 0.0801, "step": 6460 }, { "grad_norm": 0.5359430909156799, "learning_rate": 9.994101375840618e-05, "loss": 0.0801, "step": 6470 }, { "grad_norm": 0.7991141080856323, "learning_rate": 9.994020810541098e-05, "loss": 0.0749, "step": 6480 }, { "grad_norm": 0.6057160496711731, "learning_rate": 9.99393969910297e-05, "loss": 0.1018, "step": 6490 }, { "grad_norm": 0.5197307467460632, "learning_rate": 9.993858041535104e-05, "loss": 0.0767, "step": 6500 }, { "grad_norm": 0.635005533695221, "learning_rate": 9.99377583784643e-05, "loss": 0.0712, "step": 6510 }, { "grad_norm": 0.5681917071342468, "learning_rate": 9.993693088045939e-05, "loss": 0.0729, "step": 6520 }, { "grad_norm": 0.5682770609855652, "learning_rate": 9.99360979214268e-05, "loss": 0.0748, "step": 6530 }, { "grad_norm": 0.6086547374725342, "learning_rate": 9.99352595014576e-05, "loss": 0.0846, "step": 6540 }, { "grad_norm": 0.6423312425613403, "learning_rate": 9.993441562064354e-05, "loss": 0.0781, "step": 6550 }, { "grad_norm": 0.5491427779197693, "learning_rate": 9.993356627907685e-05, "loss": 0.0835, "step": 6560 }, { "grad_norm": 0.6036350727081299, "learning_rate": 9.99327114768504e-05, "loss": 0.0811, "step": 6570 }, { "grad_norm": 0.5532122254371643, "learning_rate": 9.99318512140577e-05, "loss": 0.0725, "step": 6580 }, { "grad_norm": 0.7557259202003479, "learning_rate": 9.993098549079284e-05, "loss": 0.0854, "step": 6590 }, { "grad_norm": 0.450575590133667, "learning_rate": 9.993011430715047e-05, "loss": 0.0967, "step": 6600 }, { "grad_norm": 0.5597912669181824, "learning_rate": 9.992923766322586e-05, "loss": 0.0708, "step": 6610 }, { "grad_norm": 0.6135711669921875, "learning_rate": 9.99283555591149e-05, "loss": 0.0704, "step": 6620 }, { "grad_norm": 0.5433968305587769, "learning_rate": 9.992746799491404e-05, "loss": 0.0754, "step": 6630 }, { "grad_norm": 0.5281304717063904, "learning_rate": 9.992657497072033e-05, "loss": 0.0666, "step": 6640 }, { "grad_norm": 0.6087095737457275, "learning_rate": 9.992567648663147e-05, "loss": 0.086, "step": 6650 }, { "grad_norm": 0.5236532092094421, "learning_rate": 9.992477254274568e-05, "loss": 0.0718, "step": 6660 }, { "grad_norm": 0.5042014718055725, "learning_rate": 9.992386313916183e-05, "loss": 0.0715, "step": 6670 }, { "grad_norm": 0.572210967540741, "learning_rate": 9.992294827597934e-05, "loss": 0.0897, "step": 6680 }, { "grad_norm": 0.6205742955207825, "learning_rate": 9.992202795329831e-05, "loss": 0.0889, "step": 6690 }, { "grad_norm": 0.5751516819000244, "learning_rate": 9.992110217121936e-05, "loss": 0.0914, "step": 6700 }, { "grad_norm": 0.7443505525588989, "learning_rate": 9.992017092984372e-05, "loss": 0.088, "step": 6710 }, { "grad_norm": 0.5888525247573853, "learning_rate": 9.991923422927326e-05, "loss": 0.1024, "step": 6720 }, { "grad_norm": 0.5905163884162903, "learning_rate": 9.991829206961037e-05, "loss": 0.0871, "step": 6730 }, { "grad_norm": 0.5188890099525452, "learning_rate": 9.991734445095813e-05, "loss": 0.1104, "step": 6740 }, { "grad_norm": 0.5104064345359802, "learning_rate": 9.991639137342015e-05, "loss": 0.0769, "step": 6750 }, { "grad_norm": 0.6302967667579651, "learning_rate": 9.991543283710064e-05, "loss": 0.0879, "step": 6760 }, { "grad_norm": 0.8476054668426514, "learning_rate": 9.991446884210445e-05, "loss": 0.07, "step": 6770 }, { "grad_norm": 0.7826663255691528, "learning_rate": 9.9913499388537e-05, "loss": 0.0876, "step": 6780 }, { "grad_norm": 0.7101547718048096, "learning_rate": 9.99125244765043e-05, "loss": 0.0804, "step": 6790 }, { "grad_norm": 0.6556239128112793, "learning_rate": 9.991154410611296e-05, "loss": 0.0944, "step": 6800 }, { "grad_norm": 0.6072325110435486, "learning_rate": 9.99105582774702e-05, "loss": 0.0835, "step": 6810 }, { "grad_norm": 0.5888451933860779, "learning_rate": 9.990956699068384e-05, "loss": 0.0943, "step": 6820 }, { "grad_norm": 0.735262930393219, "learning_rate": 9.990857024586224e-05, "loss": 0.089, "step": 6830 }, { "grad_norm": 0.6220729947090149, "learning_rate": 9.990756804311446e-05, "loss": 0.105, "step": 6840 }, { "grad_norm": 0.9339881539344788, "learning_rate": 9.990656038255006e-05, "loss": 0.0932, "step": 6850 }, { "grad_norm": 0.5935235619544983, "learning_rate": 9.990554726427926e-05, "loss": 0.0884, "step": 6860 }, { "grad_norm": 0.7304555177688599, "learning_rate": 9.990452868841284e-05, "loss": 0.1007, "step": 6870 }, { "grad_norm": 0.6005241274833679, "learning_rate": 9.99035046550622e-05, "loss": 0.0987, "step": 6880 }, { "grad_norm": 0.6708545088768005, "learning_rate": 9.99024751643393e-05, "loss": 0.0901, "step": 6890 }, { "grad_norm": 0.7408072352409363, "learning_rate": 9.990144021635677e-05, "loss": 0.092, "step": 6900 }, { "grad_norm": 0.5130012035369873, "learning_rate": 9.990039981122775e-05, "loss": 0.0703, "step": 6910 }, { "grad_norm": 0.6219084858894348, "learning_rate": 9.989935394906602e-05, "loss": 0.0899, "step": 6920 }, { "grad_norm": 0.5259164571762085, "learning_rate": 9.989830262998598e-05, "loss": 0.0855, "step": 6930 }, { "grad_norm": 0.5287942290306091, "learning_rate": 9.989724585410259e-05, "loss": 0.071, "step": 6940 }, { "grad_norm": 0.5967610478401184, "learning_rate": 9.989618362153139e-05, "loss": 0.0743, "step": 6950 }, { "grad_norm": 0.5877750515937805, "learning_rate": 9.989511593238859e-05, "loss": 0.0856, "step": 6960 }, { "grad_norm": 0.5828722715377808, "learning_rate": 9.98940427867909e-05, "loss": 0.0991, "step": 6970 }, { "grad_norm": 0.762936532497406, "learning_rate": 9.989296418485573e-05, "loss": 0.0999, "step": 6980 }, { "grad_norm": 0.6709808111190796, "learning_rate": 9.989188012670101e-05, "loss": 0.0813, "step": 6990 }, { "grad_norm": 0.5593067407608032, "learning_rate": 9.989079061244528e-05, "loss": 0.0829, "step": 7000 }, { "grad_norm": 0.5976501107215881, "learning_rate": 9.988969564220769e-05, "loss": 0.0861, "step": 7010 }, { "grad_norm": 0.5968590378761292, "learning_rate": 9.988859521610801e-05, "loss": 0.0751, "step": 7020 }, { "grad_norm": 0.6048555970191956, "learning_rate": 9.988748933426656e-05, "loss": 0.0952, "step": 7030 }, { "grad_norm": 0.5295006632804871, "learning_rate": 9.988637799680428e-05, "loss": 0.0883, "step": 7040 }, { "grad_norm": 0.5443347692489624, "learning_rate": 9.98852612038427e-05, "loss": 0.0703, "step": 7050 }, { "grad_norm": 0.5214030146598816, "learning_rate": 9.988413895550397e-05, "loss": 0.0723, "step": 7060 }, { "grad_norm": 0.7560358643531799, "learning_rate": 9.98830112519108e-05, "loss": 0.078, "step": 7070 }, { "grad_norm": 0.691066563129425, "learning_rate": 9.98818780931865e-05, "loss": 0.0988, "step": 7080 }, { "grad_norm": 0.6991472244262695, "learning_rate": 9.988073947945502e-05, "loss": 0.0922, "step": 7090 }, { "grad_norm": 0.6737743020057678, "learning_rate": 9.987959541084087e-05, "loss": 0.0774, "step": 7100 }, { "grad_norm": 0.6806911826133728, "learning_rate": 9.987844588746915e-05, "loss": 0.0926, "step": 7110 }, { "grad_norm": 0.5647685527801514, "learning_rate": 9.987729090946558e-05, "loss": 0.07, "step": 7120 }, { "grad_norm": 0.5898720622062683, "learning_rate": 9.987613047695647e-05, "loss": 0.0761, "step": 7130 }, { "grad_norm": 0.5626857876777649, "learning_rate": 9.987496459006871e-05, "loss": 0.0999, "step": 7140 }, { "grad_norm": 0.5414758324623108, "learning_rate": 9.987379324892982e-05, "loss": 0.0896, "step": 7150 }, { "grad_norm": 0.7046065330505371, "learning_rate": 9.987261645366788e-05, "loss": 0.0811, "step": 7160 }, { "grad_norm": 0.5400976538658142, "learning_rate": 9.987143420441158e-05, "loss": 0.0888, "step": 7170 }, { "grad_norm": 0.6162227392196655, "learning_rate": 9.987024650129022e-05, "loss": 0.0677, "step": 7180 }, { "grad_norm": 0.506099283695221, "learning_rate": 9.986905334443368e-05, "loss": 0.0845, "step": 7190 }, { "grad_norm": 0.7970136404037476, "learning_rate": 9.986785473397245e-05, "loss": 0.078, "step": 7200 }, { "grad_norm": 0.4954667389392853, "learning_rate": 9.98666506700376e-05, "loss": 0.0812, "step": 7210 }, { "grad_norm": 0.7198119163513184, "learning_rate": 9.986544115276081e-05, "loss": 0.0761, "step": 7220 }, { "grad_norm": 0.7344428300857544, "learning_rate": 9.986422618227433e-05, "loss": 0.089, "step": 7230 }, { "grad_norm": 0.6065945625305176, "learning_rate": 9.986300575871106e-05, "loss": 0.085, "step": 7240 }, { "grad_norm": 0.4951917827129364, "learning_rate": 9.986177988220444e-05, "loss": 0.068, "step": 7250 }, { "grad_norm": 0.6875990033149719, "learning_rate": 9.986054855288856e-05, "loss": 0.0813, "step": 7260 }, { "grad_norm": 0.5887101292610168, "learning_rate": 9.985931177089802e-05, "loss": 0.0737, "step": 7270 }, { "grad_norm": 0.5300500988960266, "learning_rate": 9.985806953636814e-05, "loss": 0.0737, "step": 7280 }, { "grad_norm": 0.5233163833618164, "learning_rate": 9.985682184943471e-05, "loss": 0.0792, "step": 7290 }, { "grad_norm": 0.7241396903991699, "learning_rate": 9.98555687102342e-05, "loss": 0.0809, "step": 7300 }, { "grad_norm": 0.5255311727523804, "learning_rate": 9.985431011890367e-05, "loss": 0.0891, "step": 7310 }, { "grad_norm": 0.723617434501648, "learning_rate": 9.985304607558075e-05, "loss": 0.0993, "step": 7320 }, { "grad_norm": 0.6238755583763123, "learning_rate": 9.985177658040364e-05, "loss": 0.075, "step": 7330 }, { "grad_norm": 0.5016515254974365, "learning_rate": 9.985050163351119e-05, "loss": 0.0895, "step": 7340 }, { "grad_norm": 0.6797443628311157, "learning_rate": 9.984922123504286e-05, "loss": 0.0874, "step": 7350 }, { "grad_norm": 0.5685251951217651, "learning_rate": 9.984793538513862e-05, "loss": 0.0847, "step": 7360 }, { "grad_norm": 0.5131505727767944, "learning_rate": 9.984664408393912e-05, "loss": 0.0938, "step": 7370 }, { "grad_norm": 0.5174160003662109, "learning_rate": 9.984534733158556e-05, "loss": 0.0892, "step": 7380 }, { "grad_norm": 0.5673221349716187, "learning_rate": 9.984404512821977e-05, "loss": 0.0889, "step": 7390 }, { "grad_norm": 0.48005884885787964, "learning_rate": 9.984273747398411e-05, "loss": 0.0781, "step": 7400 }, { "grad_norm": 0.842741847038269, "learning_rate": 9.984142436902165e-05, "loss": 0.086, "step": 7410 }, { "grad_norm": 0.554317057132721, "learning_rate": 9.984010581347596e-05, "loss": 0.1011, "step": 7420 }, { "grad_norm": 0.6788955330848694, "learning_rate": 9.983878180749121e-05, "loss": 0.0861, "step": 7430 }, { "grad_norm": 0.5215774178504944, "learning_rate": 9.983745235121222e-05, "loss": 0.0825, "step": 7440 }, { "grad_norm": 0.5176992416381836, "learning_rate": 9.983611744478438e-05, "loss": 0.0722, "step": 7450 }, { "grad_norm": 0.7404122352600098, "learning_rate": 9.983477708835365e-05, "loss": 0.0759, "step": 7460 }, { "grad_norm": 0.507444441318512, "learning_rate": 9.983343128206664e-05, "loss": 0.0853, "step": 7470 }, { "grad_norm": 0.4677247107028961, "learning_rate": 9.983208002607049e-05, "loss": 0.0858, "step": 7480 }, { "grad_norm": 0.6000056266784668, "learning_rate": 9.9830723320513e-05, "loss": 0.083, "step": 7490 }, { "grad_norm": 0.605668306350708, "learning_rate": 9.982936116554254e-05, "loss": 0.076, "step": 7500 }, { "grad_norm": 0.538241446018219, "learning_rate": 9.982799356130803e-05, "loss": 0.0672, "step": 7510 }, { "grad_norm": 0.5737293362617493, "learning_rate": 9.982662050795908e-05, "loss": 0.0794, "step": 7520 }, { "grad_norm": 0.5424553155899048, "learning_rate": 9.982524200564583e-05, "loss": 0.0794, "step": 7530 }, { "grad_norm": 0.5767374038696289, "learning_rate": 9.982385805451901e-05, "loss": 0.0856, "step": 7540 }, { "grad_norm": 0.4381382167339325, "learning_rate": 9.982246865472998e-05, "loss": 0.0625, "step": 7550 }, { "grad_norm": 0.5553086400032043, "learning_rate": 9.982107380643069e-05, "loss": 0.0737, "step": 7560 }, { "grad_norm": 0.7008491158485413, "learning_rate": 9.981967350977368e-05, "loss": 0.0883, "step": 7570 }, { "grad_norm": 0.6696354150772095, "learning_rate": 9.981826776491208e-05, "loss": 0.0926, "step": 7580 }, { "grad_norm": 0.424562007188797, "learning_rate": 9.98168565719996e-05, "loss": 0.0777, "step": 7590 }, { "grad_norm": 0.5327266454696655, "learning_rate": 9.98154399311906e-05, "loss": 0.0986, "step": 7600 }, { "grad_norm": 0.37984535098075867, "learning_rate": 9.981401784263997e-05, "loss": 0.0768, "step": 7610 }, { "grad_norm": 0.49232423305511475, "learning_rate": 9.981259030650326e-05, "loss": 0.0802, "step": 7620 }, { "grad_norm": 0.5742623805999756, "learning_rate": 9.981115732293655e-05, "loss": 0.0791, "step": 7630 }, { "grad_norm": 0.6060947775840759, "learning_rate": 9.980971889209659e-05, "loss": 0.0694, "step": 7640 }, { "grad_norm": 0.539577305316925, "learning_rate": 9.980827501414064e-05, "loss": 0.0648, "step": 7650 }, { "grad_norm": 0.47415393590927124, "learning_rate": 9.980682568922663e-05, "loss": 0.0616, "step": 7660 }, { "grad_norm": 0.6655924916267395, "learning_rate": 9.980537091751304e-05, "loss": 0.0741, "step": 7670 }, { "grad_norm": 0.5337203145027161, "learning_rate": 9.980391069915897e-05, "loss": 0.0672, "step": 7680 }, { "grad_norm": 0.6300398707389832, "learning_rate": 9.98024450343241e-05, "loss": 0.0743, "step": 7690 }, { "grad_norm": 0.5481188297271729, "learning_rate": 9.980097392316872e-05, "loss": 0.0628, "step": 7700 }, { "grad_norm": 0.5754105448722839, "learning_rate": 9.97994973658537e-05, "loss": 0.0872, "step": 7710 }, { "grad_norm": 0.5984548330307007, "learning_rate": 9.979801536254054e-05, "loss": 0.0819, "step": 7720 }, { "grad_norm": 0.5429808497428894, "learning_rate": 9.979652791339127e-05, "loss": 0.072, "step": 7730 }, { "grad_norm": 0.6211358308792114, "learning_rate": 9.97950350185686e-05, "loss": 0.0732, "step": 7740 }, { "grad_norm": 0.5669873356819153, "learning_rate": 9.979353667823574e-05, "loss": 0.0751, "step": 7750 }, { "grad_norm": 0.6075378656387329, "learning_rate": 9.979203289255658e-05, "loss": 0.0807, "step": 7760 }, { "grad_norm": 0.4440210461616516, "learning_rate": 9.979052366169557e-05, "loss": 0.0831, "step": 7770 }, { "grad_norm": 0.6072961091995239, "learning_rate": 9.978900898581775e-05, "loss": 0.084, "step": 7780 }, { "grad_norm": 0.636816680431366, "learning_rate": 9.978748886508875e-05, "loss": 0.0891, "step": 7790 }, { "grad_norm": 0.638425886631012, "learning_rate": 9.978596329967484e-05, "loss": 0.0879, "step": 7800 }, { "grad_norm": 0.6206619143486023, "learning_rate": 9.978443228974284e-05, "loss": 0.0849, "step": 7810 }, { "grad_norm": 0.5339764952659607, "learning_rate": 9.978289583546015e-05, "loss": 0.0649, "step": 7820 }, { "grad_norm": 0.5658050775527954, "learning_rate": 9.978135393699484e-05, "loss": 0.0795, "step": 7830 }, { "grad_norm": 0.6056549549102783, "learning_rate": 9.977980659451548e-05, "loss": 0.0761, "step": 7840 }, { "grad_norm": 0.5789365172386169, "learning_rate": 9.977825380819135e-05, "loss": 0.0933, "step": 7850 }, { "grad_norm": 0.6000226140022278, "learning_rate": 9.97766955781922e-05, "loss": 0.0743, "step": 7860 }, { "grad_norm": 0.5152534246444702, "learning_rate": 9.977513190468848e-05, "loss": 0.0955, "step": 7870 }, { "grad_norm": 0.5139080882072449, "learning_rate": 9.977356278785116e-05, "loss": 0.0871, "step": 7880 }, { "grad_norm": 0.4606151878833771, "learning_rate": 9.977198822785184e-05, "loss": 0.0825, "step": 7890 }, { "grad_norm": 0.4928241968154907, "learning_rate": 9.977040822486273e-05, "loss": 0.0917, "step": 7900 }, { "grad_norm": 0.46495410799980164, "learning_rate": 9.97688227790566e-05, "loss": 0.0909, "step": 7910 }, { "grad_norm": 0.5229103565216064, "learning_rate": 9.976723189060684e-05, "loss": 0.1057, "step": 7920 }, { "grad_norm": 0.7143286466598511, "learning_rate": 9.976563555968742e-05, "loss": 0.0785, "step": 7930 }, { "grad_norm": 0.6671535968780518, "learning_rate": 9.976403378647292e-05, "loss": 0.0756, "step": 7940 }, { "grad_norm": 0.5655751824378967, "learning_rate": 9.97624265711385e-05, "loss": 0.0715, "step": 7950 }, { "grad_norm": 0.494257390499115, "learning_rate": 9.976081391385993e-05, "loss": 0.0689, "step": 7960 }, { "grad_norm": 0.5389819741249084, "learning_rate": 9.975919581481356e-05, "loss": 0.0736, "step": 7970 }, { "grad_norm": 0.48318612575531006, "learning_rate": 9.975757227417634e-05, "loss": 0.0696, "step": 7980 }, { "grad_norm": 0.8126080632209778, "learning_rate": 9.975594329212586e-05, "loss": 0.0783, "step": 7990 }, { "grad_norm": 0.5875886082649231, "learning_rate": 9.97543088688402e-05, "loss": 0.0701, "step": 8000 }, { "grad_norm": 0.6604712605476379, "learning_rate": 9.975266900449814e-05, "loss": 0.0698, "step": 8010 }, { "grad_norm": 0.5317050814628601, "learning_rate": 9.975102369927898e-05, "loss": 0.0644, "step": 8020 }, { "grad_norm": 0.4315575361251831, "learning_rate": 9.974937295336269e-05, "loss": 0.0696, "step": 8030 }, { "grad_norm": 0.7758764028549194, "learning_rate": 9.974771676692975e-05, "loss": 0.0749, "step": 8040 }, { "grad_norm": 0.5910906195640564, "learning_rate": 9.974605514016131e-05, "loss": 0.077, "step": 8050 }, { "grad_norm": 0.6343529224395752, "learning_rate": 9.974438807323907e-05, "loss": 0.0709, "step": 8060 }, { "grad_norm": 0.4706456959247589, "learning_rate": 9.974271556634535e-05, "loss": 0.0695, "step": 8070 }, { "grad_norm": 0.7131829857826233, "learning_rate": 9.974103761966302e-05, "loss": 0.0929, "step": 8080 }, { "grad_norm": 0.5653907656669617, "learning_rate": 9.973935423337563e-05, "loss": 0.0837, "step": 8090 }, { "grad_norm": 0.5583134889602661, "learning_rate": 9.973766540766722e-05, "loss": 0.0892, "step": 8100 }, { "grad_norm": 0.679557740688324, "learning_rate": 9.97359711427225e-05, "loss": 0.0846, "step": 8110 }, { "grad_norm": 0.537447988986969, "learning_rate": 9.973427143872677e-05, "loss": 0.0782, "step": 8120 }, { "grad_norm": 0.602020263671875, "learning_rate": 9.973256629586589e-05, "loss": 0.0819, "step": 8130 }, { "grad_norm": 0.5854724645614624, "learning_rate": 9.973085571432632e-05, "loss": 0.0862, "step": 8140 }, { "grad_norm": 0.5479910373687744, "learning_rate": 9.972913969429513e-05, "loss": 0.0782, "step": 8150 }, { "grad_norm": 0.7425801157951355, "learning_rate": 9.972741823596e-05, "loss": 0.0841, "step": 8160 }, { "grad_norm": 0.5818095803260803, "learning_rate": 9.972569133950917e-05, "loss": 0.0744, "step": 8170 }, { "grad_norm": 0.527822732925415, "learning_rate": 9.972395900513151e-05, "loss": 0.0798, "step": 8180 }, { "grad_norm": 0.556321918964386, "learning_rate": 9.972222123301645e-05, "loss": 0.0782, "step": 8190 }, { "grad_norm": 0.5682145357131958, "learning_rate": 9.972047802335403e-05, "loss": 0.0715, "step": 8200 }, { "grad_norm": 0.31567952036857605, "learning_rate": 9.971872937633488e-05, "loss": 0.0609, "step": 8210 }, { "grad_norm": 0.4605236351490021, "learning_rate": 9.971697529215024e-05, "loss": 0.0708, "step": 8220 }, { "grad_norm": 0.5106729865074158, "learning_rate": 9.971521577099192e-05, "loss": 0.083, "step": 8230 }, { "grad_norm": 0.4273492991924286, "learning_rate": 9.971345081305236e-05, "loss": 0.0778, "step": 8240 }, { "grad_norm": 0.6392154097557068, "learning_rate": 9.971168041852456e-05, "loss": 0.0722, "step": 8250 }, { "grad_norm": 0.6187363862991333, "learning_rate": 9.970990458760215e-05, "loss": 0.0672, "step": 8260 }, { "grad_norm": 0.46928301453590393, "learning_rate": 9.970812332047929e-05, "loss": 0.0611, "step": 8270 }, { "grad_norm": 0.5898728370666504, "learning_rate": 9.97063366173508e-05, "loss": 0.0837, "step": 8280 }, { "grad_norm": 0.47435978055000305, "learning_rate": 9.970454447841207e-05, "loss": 0.0627, "step": 8290 }, { "grad_norm": 0.4028644561767578, "learning_rate": 9.970274690385909e-05, "loss": 0.0693, "step": 8300 }, { "grad_norm": 0.734452486038208, "learning_rate": 9.970094389388844e-05, "loss": 0.0737, "step": 8310 }, { "grad_norm": 0.4057399332523346, "learning_rate": 9.969913544869728e-05, "loss": 0.0615, "step": 8320 }, { "grad_norm": 0.5930889844894409, "learning_rate": 9.96973215684834e-05, "loss": 0.0645, "step": 8330 }, { "grad_norm": 0.6111104488372803, "learning_rate": 9.969550225344513e-05, "loss": 0.0644, "step": 8340 }, { "grad_norm": 0.5194043517112732, "learning_rate": 9.969367750378147e-05, "loss": 0.0826, "step": 8350 }, { "grad_norm": 0.587995707988739, "learning_rate": 9.969184731969194e-05, "loss": 0.0654, "step": 8360 }, { "grad_norm": 0.4914413094520569, "learning_rate": 9.96900117013767e-05, "loss": 0.0624, "step": 8370 }, { "grad_norm": 0.500763475894928, "learning_rate": 9.96881706490365e-05, "loss": 0.0703, "step": 8380 }, { "grad_norm": 0.6117989420890808, "learning_rate": 9.968632416287265e-05, "loss": 0.0607, "step": 8390 }, { "grad_norm": 0.5655341148376465, "learning_rate": 9.96844722430871e-05, "loss": 0.0639, "step": 8400 }, { "grad_norm": 0.7360720634460449, "learning_rate": 9.968261488988235e-05, "loss": 0.0662, "step": 8410 }, { "grad_norm": 0.4385564625263214, "learning_rate": 9.968075210346155e-05, "loss": 0.0611, "step": 8420 }, { "grad_norm": 0.6366820335388184, "learning_rate": 9.967888388402839e-05, "loss": 0.0657, "step": 8430 }, { "grad_norm": 0.38514646887779236, "learning_rate": 9.967701023178717e-05, "loss": 0.066, "step": 8440 }, { "grad_norm": 0.7555549144744873, "learning_rate": 9.967513114694282e-05, "loss": 0.0713, "step": 8450 }, { "grad_norm": 0.5278430581092834, "learning_rate": 9.967324662970079e-05, "loss": 0.072, "step": 8460 }, { "grad_norm": 0.5933712124824524, "learning_rate": 9.96713566802672e-05, "loss": 0.074, "step": 8470 }, { "grad_norm": 0.5582693815231323, "learning_rate": 9.966946129884873e-05, "loss": 0.0662, "step": 8480 }, { "grad_norm": 0.6119797229766846, "learning_rate": 9.966756048565265e-05, "loss": 0.0688, "step": 8490 }, { "grad_norm": 0.6300646662712097, "learning_rate": 9.966565424088681e-05, "loss": 0.0779, "step": 8500 }, { "grad_norm": 0.5865777730941772, "learning_rate": 9.96637425647597e-05, "loss": 0.0646, "step": 8510 }, { "grad_norm": 0.5758484601974487, "learning_rate": 9.966182545748038e-05, "loss": 0.0598, "step": 8520 }, { "grad_norm": 0.48947417736053467, "learning_rate": 9.96599029192585e-05, "loss": 0.0726, "step": 8530 }, { "grad_norm": 0.4478510320186615, "learning_rate": 9.965797495030428e-05, "loss": 0.0676, "step": 8540 }, { "grad_norm": 0.5491752624511719, "learning_rate": 9.96560415508286e-05, "loss": 0.0691, "step": 8550 }, { "grad_norm": 0.46545547246932983, "learning_rate": 9.965410272104286e-05, "loss": 0.0689, "step": 8560 }, { "grad_norm": 0.6065381169319153, "learning_rate": 9.96521584611591e-05, "loss": 0.0986, "step": 8570 }, { "grad_norm": 0.4643268883228302, "learning_rate": 9.965020877138994e-05, "loss": 0.0871, "step": 8580 }, { "grad_norm": 0.5737265348434448, "learning_rate": 9.964825365194861e-05, "loss": 0.0822, "step": 8590 }, { "grad_norm": 0.45844173431396484, "learning_rate": 9.96462931030489e-05, "loss": 0.0604, "step": 8600 }, { "grad_norm": 0.7400155067443848, "learning_rate": 9.96443271249052e-05, "loss": 0.0643, "step": 8610 }, { "grad_norm": 0.6466207504272461, "learning_rate": 9.964235571773255e-05, "loss": 0.0922, "step": 8620 }, { "grad_norm": 0.4878484904766083, "learning_rate": 9.96403788817465e-05, "loss": 0.0685, "step": 8630 }, { "grad_norm": 0.5092241764068604, "learning_rate": 9.963839661716325e-05, "loss": 0.0775, "step": 8640 }, { "grad_norm": 0.6762095093727112, "learning_rate": 9.963640892419958e-05, "loss": 0.078, "step": 8650 }, { "grad_norm": 0.7666207551956177, "learning_rate": 9.963441580307286e-05, "loss": 0.0699, "step": 8660 }, { "grad_norm": 0.49609217047691345, "learning_rate": 9.963241725400104e-05, "loss": 0.0614, "step": 8670 }, { "grad_norm": 0.42047858238220215, "learning_rate": 9.963041327720271e-05, "loss": 0.0689, "step": 8680 }, { "grad_norm": 0.46867635846138, "learning_rate": 9.962840387289697e-05, "loss": 0.07, "step": 8690 }, { "grad_norm": 0.4683285057544708, "learning_rate": 9.962638904130363e-05, "loss": 0.0723, "step": 8700 }, { "grad_norm": 0.41943925619125366, "learning_rate": 9.962436878264298e-05, "loss": 0.0749, "step": 8710 }, { "grad_norm": 0.8212710618972778, "learning_rate": 9.962234309713598e-05, "loss": 0.0945, "step": 8720 }, { "grad_norm": 0.7091567516326904, "learning_rate": 9.962031198500414e-05, "loss": 0.0733, "step": 8730 }, { "grad_norm": 0.6367358565330505, "learning_rate": 9.961827544646958e-05, "loss": 0.0723, "step": 8740 }, { "grad_norm": 0.7950409650802612, "learning_rate": 9.961623348175501e-05, "loss": 0.0832, "step": 8750 }, { "grad_norm": 0.5562450885772705, "learning_rate": 9.961418609108377e-05, "loss": 0.074, "step": 8760 }, { "grad_norm": 0.4393463134765625, "learning_rate": 9.961213327467971e-05, "loss": 0.0708, "step": 8770 }, { "grad_norm": 0.7001277804374695, "learning_rate": 9.961007503276736e-05, "loss": 0.0705, "step": 8780 }, { "grad_norm": 0.48276782035827637, "learning_rate": 9.960801136557179e-05, "loss": 0.0613, "step": 8790 }, { "grad_norm": 0.5920506715774536, "learning_rate": 9.960594227331866e-05, "loss": 0.0572, "step": 8800 }, { "grad_norm": 0.3733966648578644, "learning_rate": 9.960386775623429e-05, "loss": 0.059, "step": 8810 }, { "grad_norm": 0.5434300899505615, "learning_rate": 9.96017878145455e-05, "loss": 0.0759, "step": 8820 }, { "grad_norm": 0.5564976334571838, "learning_rate": 9.959970244847977e-05, "loss": 0.0593, "step": 8830 }, { "grad_norm": 0.7337603569030762, "learning_rate": 9.959761165826518e-05, "loss": 0.0704, "step": 8840 }, { "grad_norm": 0.4815351068973541, "learning_rate": 9.959551544413033e-05, "loss": 0.0679, "step": 8850 }, { "grad_norm": 0.5773460268974304, "learning_rate": 9.959341380630448e-05, "loss": 0.0627, "step": 8860 }, { "grad_norm": 0.8505122661590576, "learning_rate": 9.959130674501746e-05, "loss": 0.0774, "step": 8870 }, { "grad_norm": 0.672034740447998, "learning_rate": 9.958919426049968e-05, "loss": 0.0844, "step": 8880 }, { "grad_norm": 0.5275601744651794, "learning_rate": 9.958707635298219e-05, "loss": 0.0739, "step": 8890 }, { "grad_norm": 0.5341920256614685, "learning_rate": 9.958495302269657e-05, "loss": 0.0677, "step": 8900 }, { "grad_norm": 0.5825551152229309, "learning_rate": 9.958282426987503e-05, "loss": 0.0666, "step": 8910 }, { "grad_norm": 0.6330316662788391, "learning_rate": 9.95806900947504e-05, "loss": 0.0866, "step": 8920 }, { "grad_norm": 0.5340069532394409, "learning_rate": 9.957855049755604e-05, "loss": 0.0553, "step": 8930 }, { "grad_norm": 0.4144245982170105, "learning_rate": 9.957640547852593e-05, "loss": 0.0727, "step": 8940 }, { "grad_norm": 0.5117513537406921, "learning_rate": 9.957425503789466e-05, "loss": 0.0663, "step": 8950 }, { "grad_norm": 0.42442259192466736, "learning_rate": 9.957209917589738e-05, "loss": 0.0565, "step": 8960 }, { "grad_norm": 0.4796377420425415, "learning_rate": 9.956993789276987e-05, "loss": 0.0559, "step": 8970 }, { "grad_norm": 0.45709773898124695, "learning_rate": 9.956777118874847e-05, "loss": 0.0649, "step": 8980 }, { "grad_norm": 0.5919724106788635, "learning_rate": 9.956559906407016e-05, "loss": 0.0903, "step": 8990 }, { "grad_norm": 0.5151185393333435, "learning_rate": 9.956342151897245e-05, "loss": 0.0687, "step": 9000 }, { "grad_norm": 0.44834956526756287, "learning_rate": 9.956123855369346e-05, "loss": 0.055, "step": 9010 }, { "grad_norm": 0.6140286326408386, "learning_rate": 9.955905016847196e-05, "loss": 0.0676, "step": 9020 }, { "grad_norm": 0.6001137495040894, "learning_rate": 9.955685636354723e-05, "loss": 0.083, "step": 9030 }, { "grad_norm": 0.6963826417922974, "learning_rate": 9.95546571391592e-05, "loss": 0.0754, "step": 9040 }, { "grad_norm": 0.6051949858665466, "learning_rate": 9.955245249554837e-05, "loss": 0.0689, "step": 9050 }, { "grad_norm": 0.42326822876930237, "learning_rate": 9.955024243295582e-05, "loss": 0.081, "step": 9060 }, { "grad_norm": 0.4602459967136383, "learning_rate": 9.954802695162328e-05, "loss": 0.0622, "step": 9070 }, { "grad_norm": 0.4471937417984009, "learning_rate": 9.954580605179302e-05, "loss": 0.0732, "step": 9080 }, { "grad_norm": 0.6017486453056335, "learning_rate": 9.954357973370788e-05, "loss": 0.0764, "step": 9090 }, { "grad_norm": 0.5077396631240845, "learning_rate": 9.954134799761135e-05, "loss": 0.0896, "step": 9100 }, { "grad_norm": 0.7578203082084656, "learning_rate": 9.953911084374748e-05, "loss": 0.0815, "step": 9110 }, { "grad_norm": 0.5236756801605225, "learning_rate": 9.953686827236093e-05, "loss": 0.0694, "step": 9120 }, { "grad_norm": 0.5760639309883118, "learning_rate": 9.953462028369695e-05, "loss": 0.0803, "step": 9130 }, { "grad_norm": 0.544281542301178, "learning_rate": 9.953236687800136e-05, "loss": 0.0664, "step": 9140 }, { "grad_norm": 0.5471056699752808, "learning_rate": 9.95301080555206e-05, "loss": 0.0775, "step": 9150 }, { "grad_norm": 0.4940040409564972, "learning_rate": 9.952784381650171e-05, "loss": 0.0582, "step": 9160 }, { "grad_norm": 0.5298790335655212, "learning_rate": 9.952557416119226e-05, "loss": 0.0607, "step": 9170 }, { "grad_norm": 0.5002674460411072, "learning_rate": 9.95232990898405e-05, "loss": 0.068, "step": 9180 }, { "grad_norm": 0.795024573802948, "learning_rate": 9.95210186026952e-05, "loss": 0.0608, "step": 9190 }, { "grad_norm": 0.6215788125991821, "learning_rate": 9.951873270000576e-05, "loss": 0.0739, "step": 9200 }, { "grad_norm": 0.49203065037727356, "learning_rate": 9.951644138202216e-05, "loss": 0.0881, "step": 9210 }, { "grad_norm": 0.6179652214050293, "learning_rate": 9.951414464899498e-05, "loss": 0.085, "step": 9220 }, { "grad_norm": 0.54267817735672, "learning_rate": 9.951184250117538e-05, "loss": 0.0698, "step": 9230 }, { "grad_norm": 0.44302013516426086, "learning_rate": 9.950953493881513e-05, "loss": 0.0804, "step": 9240 }, { "grad_norm": 0.49731820821762085, "learning_rate": 9.950722196216658e-05, "loss": 0.0614, "step": 9250 }, { "grad_norm": 0.45681750774383545, "learning_rate": 9.950490357148265e-05, "loss": 0.0739, "step": 9260 }, { "grad_norm": 0.5822360515594482, "learning_rate": 9.950257976701692e-05, "loss": 0.0718, "step": 9270 }, { "grad_norm": 0.46070459485054016, "learning_rate": 9.950025054902348e-05, "loss": 0.0759, "step": 9280 }, { "grad_norm": 0.6228479743003845, "learning_rate": 9.949791591775706e-05, "loss": 0.0875, "step": 9290 }, { "grad_norm": 0.5189799666404724, "learning_rate": 9.949557587347298e-05, "loss": 0.082, "step": 9300 }, { "grad_norm": 0.41877567768096924, "learning_rate": 9.949323041642713e-05, "loss": 0.0533, "step": 9310 }, { "grad_norm": 0.5399551391601562, "learning_rate": 9.949087954687602e-05, "loss": 0.0845, "step": 9320 }, { "grad_norm": 0.46039336919784546, "learning_rate": 9.948852326507672e-05, "loss": 0.0691, "step": 9330 }, { "grad_norm": 0.5112992525100708, "learning_rate": 9.948616157128694e-05, "loss": 0.0689, "step": 9340 }, { "grad_norm": 0.41885897517204285, "learning_rate": 9.948379446576493e-05, "loss": 0.0605, "step": 9350 }, { "grad_norm": 0.4358230233192444, "learning_rate": 9.948142194876952e-05, "loss": 0.0572, "step": 9360 }, { "grad_norm": 0.5822429060935974, "learning_rate": 9.947904402056024e-05, "loss": 0.0781, "step": 9370 }, { "grad_norm": 0.47000187635421753, "learning_rate": 9.947666068139708e-05, "loss": 0.0761, "step": 9380 }, { "grad_norm": 0.6083246469497681, "learning_rate": 9.947427193154071e-05, "loss": 0.0712, "step": 9390 }, { "grad_norm": 0.846312403678894, "learning_rate": 9.947187777125233e-05, "loss": 0.0767, "step": 9400 }, { "grad_norm": 0.5790772438049316, "learning_rate": 9.946947820079377e-05, "loss": 0.0759, "step": 9410 }, { "grad_norm": 0.5443416833877563, "learning_rate": 9.946707322042747e-05, "loss": 0.0682, "step": 9420 }, { "grad_norm": 0.5344129204750061, "learning_rate": 9.94646628304164e-05, "loss": 0.0577, "step": 9430 }, { "grad_norm": 0.5333513617515564, "learning_rate": 9.946224703102418e-05, "loss": 0.072, "step": 9440 }, { "grad_norm": 0.46049708127975464, "learning_rate": 9.945982582251498e-05, "loss": 0.0707, "step": 9450 }, { "grad_norm": 0.4391016662120819, "learning_rate": 9.94573992051536e-05, "loss": 0.0665, "step": 9460 }, { "grad_norm": 0.7260778546333313, "learning_rate": 9.94549671792054e-05, "loss": 0.0926, "step": 9470 }, { "grad_norm": 0.5757972002029419, "learning_rate": 9.945252974493635e-05, "loss": 0.0691, "step": 9480 }, { "grad_norm": 0.4971448481082916, "learning_rate": 9.9450086902613e-05, "loss": 0.0701, "step": 9490 }, { "grad_norm": 0.46362704038619995, "learning_rate": 9.944763865250248e-05, "loss": 0.0654, "step": 9500 }, { "grad_norm": 0.5497282147407532, "learning_rate": 9.944518499487254e-05, "loss": 0.0895, "step": 9510 }, { "grad_norm": 0.5075653791427612, "learning_rate": 9.944272592999151e-05, "loss": 0.0793, "step": 9520 }, { "grad_norm": 0.5864380598068237, "learning_rate": 9.94402614581283e-05, "loss": 0.0712, "step": 9530 }, { "grad_norm": 0.6968353390693665, "learning_rate": 9.943779157955244e-05, "loss": 0.0772, "step": 9540 }, { "grad_norm": 0.4290141761302948, "learning_rate": 9.943531629453403e-05, "loss": 0.0682, "step": 9550 }, { "grad_norm": 0.5545057654380798, "learning_rate": 9.943283560334375e-05, "loss": 0.0775, "step": 9560 }, { "grad_norm": 0.5889592170715332, "learning_rate": 9.943034950625288e-05, "loss": 0.0838, "step": 9570 }, { "grad_norm": 0.5599530339241028, "learning_rate": 9.942785800353332e-05, "loss": 0.0722, "step": 9580 }, { "grad_norm": 0.565834641456604, "learning_rate": 9.942536109545751e-05, "loss": 0.0733, "step": 9590 }, { "grad_norm": 0.50532066822052, "learning_rate": 9.942285878229853e-05, "loss": 0.0625, "step": 9600 }, { "grad_norm": 0.4696681499481201, "learning_rate": 9.942035106433001e-05, "loss": 0.0651, "step": 9610 }, { "grad_norm": 0.5366938710212708, "learning_rate": 9.94178379418262e-05, "loss": 0.0748, "step": 9620 }, { "grad_norm": 0.5753177404403687, "learning_rate": 9.941531941506194e-05, "loss": 0.0731, "step": 9630 }, { "grad_norm": 0.46394431591033936, "learning_rate": 9.941279548431263e-05, "loss": 0.0683, "step": 9640 }, { "grad_norm": 0.44657066464424133, "learning_rate": 9.941026614985431e-05, "loss": 0.0628, "step": 9650 }, { "grad_norm": 0.4845183491706848, "learning_rate": 9.940773141196357e-05, "loss": 0.0743, "step": 9660 }, { "grad_norm": 0.5427964925765991, "learning_rate": 9.94051912709176e-05, "loss": 0.0778, "step": 9670 }, { "grad_norm": 0.41374364495277405, "learning_rate": 9.940264572699421e-05, "loss": 0.0686, "step": 9680 }, { "grad_norm": 0.5261087417602539, "learning_rate": 9.940009478047174e-05, "loss": 0.07, "step": 9690 }, { "grad_norm": 0.7709228992462158, "learning_rate": 9.939753843162918e-05, "loss": 0.0694, "step": 9700 }, { "grad_norm": 0.5393001437187195, "learning_rate": 9.939497668074609e-05, "loss": 0.0807, "step": 9710 }, { "grad_norm": 0.5077633261680603, "learning_rate": 9.93924095281026e-05, "loss": 0.0694, "step": 9720 }, { "grad_norm": 0.6084526181221008, "learning_rate": 9.938983697397948e-05, "loss": 0.088, "step": 9730 }, { "grad_norm": 0.4047349989414215, "learning_rate": 9.938725901865805e-05, "loss": 0.075, "step": 9740 }, { "grad_norm": 0.6904947757720947, "learning_rate": 9.93846756624202e-05, "loss": 0.0696, "step": 9750 }, { "grad_norm": 0.7292225956916809, "learning_rate": 9.938208690554849e-05, "loss": 0.0717, "step": 9760 }, { "grad_norm": 0.3754922151565552, "learning_rate": 9.9379492748326e-05, "loss": 0.0671, "step": 9770 }, { "grad_norm": 0.5759061574935913, "learning_rate": 9.937689319103641e-05, "loss": 0.066, "step": 9780 }, { "grad_norm": 0.8141961097717285, "learning_rate": 9.937428823396404e-05, "loss": 0.0731, "step": 9790 }, { "grad_norm": 0.40335434675216675, "learning_rate": 9.937167787739372e-05, "loss": 0.0592, "step": 9800 }, { "grad_norm": 0.5599032640457153, "learning_rate": 9.936906212161095e-05, "loss": 0.0737, "step": 9810 }, { "grad_norm": 0.43140941858291626, "learning_rate": 9.936644096690176e-05, "loss": 0.0776, "step": 9820 }, { "grad_norm": 0.5108827948570251, "learning_rate": 9.936381441355282e-05, "loss": 0.078, "step": 9830 }, { "grad_norm": 0.5077419281005859, "learning_rate": 9.936118246185136e-05, "loss": 0.0791, "step": 9840 }, { "grad_norm": 0.5110995769500732, "learning_rate": 9.935854511208518e-05, "loss": 0.0802, "step": 9850 }, { "grad_norm": 0.37884846329689026, "learning_rate": 9.935590236454272e-05, "loss": 0.0625, "step": 9860 }, { "grad_norm": 0.5069420337677002, "learning_rate": 9.935325421951298e-05, "loss": 0.0759, "step": 9870 }, { "grad_norm": 0.5308162569999695, "learning_rate": 9.935060067728557e-05, "loss": 0.0753, "step": 9880 }, { "grad_norm": 0.527463436126709, "learning_rate": 9.934794173815067e-05, "loss": 0.0594, "step": 9890 }, { "grad_norm": 0.7364681959152222, "learning_rate": 9.934527740239906e-05, "loss": 0.0833, "step": 9900 }, { "grad_norm": 0.4647808074951172, "learning_rate": 9.934260767032209e-05, "loss": 0.0763, "step": 9910 }, { "grad_norm": 0.5936029553413391, "learning_rate": 9.933993254221172e-05, "loss": 0.083, "step": 9920 }, { "grad_norm": 0.6684466004371643, "learning_rate": 9.933725201836053e-05, "loss": 0.0676, "step": 9930 }, { "grad_norm": 0.5362864136695862, "learning_rate": 9.933456609906162e-05, "loss": 0.0693, "step": 9940 }, { "grad_norm": 0.3591383993625641, "learning_rate": 9.933187478460875e-05, "loss": 0.0616, "step": 9950 }, { "grad_norm": 0.40863946080207825, "learning_rate": 9.93291780752962e-05, "loss": 0.0544, "step": 9960 }, { "grad_norm": 0.5624338388442993, "learning_rate": 9.932647597141893e-05, "loss": 0.0704, "step": 9970 }, { "grad_norm": 0.5784547328948975, "learning_rate": 9.932376847327239e-05, "loss": 0.0687, "step": 9980 }, { "grad_norm": 0.6711048483848572, "learning_rate": 9.932105558115268e-05, "loss": 0.0685, "step": 9990 }, { "grad_norm": 0.6260005831718445, "learning_rate": 9.931833729535651e-05, "loss": 0.0745, "step": 10000 }, { "grad_norm": 0.3647710084915161, "learning_rate": 9.931561361618111e-05, "loss": 0.0689, "step": 10010 }, { "grad_norm": 0.43965262174606323, "learning_rate": 9.931288454392435e-05, "loss": 0.0711, "step": 10020 }, { "grad_norm": 0.5485474467277527, "learning_rate": 9.931015007888467e-05, "loss": 0.0554, "step": 10030 }, { "grad_norm": 0.6030620336532593, "learning_rate": 9.930741022136112e-05, "loss": 0.0745, "step": 10040 }, { "grad_norm": 0.5762953162193298, "learning_rate": 9.930466497165333e-05, "loss": 0.0743, "step": 10050 }, { "grad_norm": 0.5440739393234253, "learning_rate": 9.93019143300615e-05, "loss": 0.0692, "step": 10060 }, { "grad_norm": 0.526276707649231, "learning_rate": 9.929915829688644e-05, "loss": 0.0567, "step": 10070 }, { "grad_norm": 0.43959176540374756, "learning_rate": 9.929639687242955e-05, "loss": 0.0721, "step": 10080 }, { "grad_norm": 0.5366008281707764, "learning_rate": 9.929363005699281e-05, "loss": 0.0624, "step": 10090 }, { "grad_norm": 0.5298662781715393, "learning_rate": 9.92908578508788e-05, "loss": 0.0684, "step": 10100 }, { "grad_norm": 0.3985108733177185, "learning_rate": 9.928808025439069e-05, "loss": 0.0583, "step": 10110 }, { "grad_norm": 0.49016451835632324, "learning_rate": 9.928529726783223e-05, "loss": 0.0618, "step": 10120 }, { "grad_norm": 0.6703681945800781, "learning_rate": 9.928250889150774e-05, "loss": 0.0831, "step": 10130 }, { "grad_norm": 0.5053903460502625, "learning_rate": 9.92797151257222e-05, "loss": 0.0705, "step": 10140 }, { "grad_norm": 0.48274296522140503, "learning_rate": 9.927691597078108e-05, "loss": 0.0811, "step": 10150 }, { "grad_norm": 0.5673996210098267, "learning_rate": 9.927411142699053e-05, "loss": 0.0791, "step": 10160 }, { "grad_norm": 0.444818913936615, "learning_rate": 9.927130149465725e-05, "loss": 0.0686, "step": 10170 }, { "grad_norm": 0.46843844652175903, "learning_rate": 9.92684861740885e-05, "loss": 0.0805, "step": 10180 }, { "grad_norm": 0.5889626145362854, "learning_rate": 9.926566546559217e-05, "loss": 0.0887, "step": 10190 }, { "grad_norm": 0.5544450879096985, "learning_rate": 9.926283936947673e-05, "loss": 0.0742, "step": 10200 }, { "grad_norm": 0.5182235240936279, "learning_rate": 9.926000788605126e-05, "loss": 0.0641, "step": 10210 }, { "grad_norm": 0.4725569486618042, "learning_rate": 9.92571710156254e-05, "loss": 0.0672, "step": 10220 }, { "grad_norm": 0.7945911288261414, "learning_rate": 9.925432875850936e-05, "loss": 0.0685, "step": 10230 }, { "grad_norm": 0.6034489274024963, "learning_rate": 9.925148111501396e-05, "loss": 0.0713, "step": 10240 }, { "grad_norm": 0.632463812828064, "learning_rate": 9.924862808545066e-05, "loss": 0.0595, "step": 10250 }, { "grad_norm": 0.4677138030529022, "learning_rate": 9.924576967013141e-05, "loss": 0.0554, "step": 10260 }, { "grad_norm": 0.5190359950065613, "learning_rate": 9.924290586936887e-05, "loss": 0.0632, "step": 10270 }, { "grad_norm": 0.4912009835243225, "learning_rate": 9.924003668347614e-05, "loss": 0.0745, "step": 10280 }, { "grad_norm": 0.5283300280570984, "learning_rate": 9.923716211276704e-05, "loss": 0.0814, "step": 10290 }, { "grad_norm": 0.6527639031410217, "learning_rate": 9.923428215755594e-05, "loss": 0.0673, "step": 10300 }, { "grad_norm": 0.5653072595596313, "learning_rate": 9.923139681815775e-05, "loss": 0.0628, "step": 10310 }, { "grad_norm": 0.4940369725227356, "learning_rate": 9.922850609488801e-05, "loss": 0.0651, "step": 10320 }, { "grad_norm": 0.40875086188316345, "learning_rate": 9.922560998806287e-05, "loss": 0.0622, "step": 10330 }, { "grad_norm": 0.4526914060115814, "learning_rate": 9.922270849799905e-05, "loss": 0.0642, "step": 10340 }, { "grad_norm": 0.6213496923446655, "learning_rate": 9.92198016250138e-05, "loss": 0.0771, "step": 10350 }, { "grad_norm": 0.4350055754184723, "learning_rate": 9.921688936942506e-05, "loss": 0.0575, "step": 10360 }, { "grad_norm": 0.4692324995994568, "learning_rate": 9.921397173155129e-05, "loss": 0.0536, "step": 10370 }, { "grad_norm": 0.6659978628158569, "learning_rate": 9.921104871171157e-05, "loss": 0.0572, "step": 10380 }, { "grad_norm": 0.5070147514343262, "learning_rate": 9.920812031022554e-05, "loss": 0.0664, "step": 10390 }, { "grad_norm": 0.3765231966972351, "learning_rate": 9.920518652741348e-05, "loss": 0.068, "step": 10400 }, { "grad_norm": 0.4025801718235016, "learning_rate": 9.920224736359618e-05, "loss": 0.0615, "step": 10410 }, { "grad_norm": 0.6406278014183044, "learning_rate": 9.91993028190951e-05, "loss": 0.0596, "step": 10420 }, { "grad_norm": 0.4914718270301819, "learning_rate": 9.919635289423222e-05, "loss": 0.06, "step": 10430 }, { "grad_norm": 0.43930378556251526, "learning_rate": 9.919339758933015e-05, "loss": 0.0506, "step": 10440 }, { "grad_norm": 0.5053418874740601, "learning_rate": 9.919043690471209e-05, "loss": 0.067, "step": 10450 }, { "grad_norm": 0.4871526062488556, "learning_rate": 9.91874708407018e-05, "loss": 0.0645, "step": 10460 }, { "grad_norm": 0.38584524393081665, "learning_rate": 9.918449939762367e-05, "loss": 0.0622, "step": 10470 }, { "grad_norm": 0.5244264006614685, "learning_rate": 9.91815225758026e-05, "loss": 0.0596, "step": 10480 }, { "grad_norm": 0.453357458114624, "learning_rate": 9.917854037556419e-05, "loss": 0.0632, "step": 10490 }, { "grad_norm": 0.44908782839775085, "learning_rate": 9.917555279723454e-05, "loss": 0.0623, "step": 10500 }, { "grad_norm": 0.48256540298461914, "learning_rate": 9.917255984114036e-05, "loss": 0.0749, "step": 10510 }, { "grad_norm": 0.4797414541244507, "learning_rate": 9.916956150760896e-05, "loss": 0.0512, "step": 10520 }, { "grad_norm": 0.46572914719581604, "learning_rate": 9.916655779696826e-05, "loss": 0.0739, "step": 10530 }, { "grad_norm": 0.505089521408081, "learning_rate": 9.916354870954671e-05, "loss": 0.0618, "step": 10540 }, { "grad_norm": 0.46154698729515076, "learning_rate": 9.91605342456734e-05, "loss": 0.0557, "step": 10550 }, { "grad_norm": 0.4973379671573639, "learning_rate": 9.915751440567795e-05, "loss": 0.0552, "step": 10560 }, { "grad_norm": 0.5006741285324097, "learning_rate": 9.915448918989066e-05, "loss": 0.0518, "step": 10570 }, { "grad_norm": 0.41764017939567566, "learning_rate": 9.915145859864232e-05, "loss": 0.0628, "step": 10580 }, { "grad_norm": 0.4928714632987976, "learning_rate": 9.914842263226437e-05, "loss": 0.0731, "step": 10590 }, { "grad_norm": 0.7076020240783691, "learning_rate": 9.914538129108882e-05, "loss": 0.0755, "step": 10600 }, { "grad_norm": 0.6257174015045166, "learning_rate": 9.914233457544825e-05, "loss": 0.0953, "step": 10610 }, { "grad_norm": 0.96288001537323, "learning_rate": 9.913928248567586e-05, "loss": 0.0862, "step": 10620 }, { "grad_norm": 0.49566665291786194, "learning_rate": 9.913622502210542e-05, "loss": 0.0667, "step": 10630 }, { "grad_norm": 0.5608927011489868, "learning_rate": 9.913316218507128e-05, "loss": 0.0701, "step": 10640 }, { "grad_norm": 0.5759821534156799, "learning_rate": 9.91300939749084e-05, "loss": 0.0719, "step": 10650 }, { "grad_norm": 0.4384663701057434, "learning_rate": 9.91270203919523e-05, "loss": 0.0648, "step": 10660 }, { "grad_norm": 0.5536145567893982, "learning_rate": 9.912394143653912e-05, "loss": 0.0624, "step": 10670 }, { "grad_norm": 0.7683601379394531, "learning_rate": 9.912085710900555e-05, "loss": 0.0523, "step": 10680 }, { "grad_norm": 0.4828062355518341, "learning_rate": 9.911776740968892e-05, "loss": 0.0594, "step": 10690 }, { "grad_norm": 0.6510915756225586, "learning_rate": 9.911467233892709e-05, "loss": 0.0868, "step": 10700 }, { "grad_norm": 0.4586699306964874, "learning_rate": 9.911157189705853e-05, "loss": 0.0721, "step": 10710 }, { "grad_norm": 0.5031728148460388, "learning_rate": 9.910846608442229e-05, "loss": 0.0644, "step": 10720 }, { "grad_norm": 0.4309678077697754, "learning_rate": 9.910535490135805e-05, "loss": 0.0561, "step": 10730 }, { "grad_norm": 0.590345561504364, "learning_rate": 9.910223834820603e-05, "loss": 0.0598, "step": 10740 }, { "grad_norm": 0.40635836124420166, "learning_rate": 9.909911642530703e-05, "loss": 0.0696, "step": 10750 }, { "grad_norm": 0.4951041340827942, "learning_rate": 9.909598913300249e-05, "loss": 0.0686, "step": 10760 }, { "grad_norm": 0.6921159625053406, "learning_rate": 9.909285647163438e-05, "loss": 0.0706, "step": 10770 }, { "grad_norm": 0.6823692321777344, "learning_rate": 9.908971844154531e-05, "loss": 0.071, "step": 10780 }, { "grad_norm": 0.44117164611816406, "learning_rate": 9.908657504307843e-05, "loss": 0.0689, "step": 10790 }, { "grad_norm": 0.5374041199684143, "learning_rate": 9.908342627657751e-05, "loss": 0.0734, "step": 10800 }, { "grad_norm": 0.6156074404716492, "learning_rate": 9.908027214238689e-05, "loss": 0.0727, "step": 10810 }, { "grad_norm": 0.6092743277549744, "learning_rate": 9.90771126408515e-05, "loss": 0.0704, "step": 10820 }, { "grad_norm": 0.45443403720855713, "learning_rate": 9.907394777231685e-05, "loss": 0.0607, "step": 10830 }, { "grad_norm": 0.45613494515419006, "learning_rate": 9.907077753712905e-05, "loss": 0.0729, "step": 10840 }, { "grad_norm": 0.6068106889724731, "learning_rate": 9.906760193563482e-05, "loss": 0.0679, "step": 10850 }, { "grad_norm": 0.5615791082382202, "learning_rate": 9.906442096818139e-05, "loss": 0.0589, "step": 10860 }, { "grad_norm": 0.5852482318878174, "learning_rate": 9.906123463511665e-05, "loss": 0.0668, "step": 10870 }, { "grad_norm": 0.38402891159057617, "learning_rate": 9.905804293678907e-05, "loss": 0.0645, "step": 10880 }, { "grad_norm": 0.45175230503082275, "learning_rate": 9.905484587354766e-05, "loss": 0.0587, "step": 10890 }, { "grad_norm": 0.4661758542060852, "learning_rate": 9.905164344574205e-05, "loss": 0.0623, "step": 10900 }, { "grad_norm": 0.46486595273017883, "learning_rate": 9.904843565372248e-05, "loss": 0.0573, "step": 10910 }, { "grad_norm": 0.6193951964378357, "learning_rate": 9.904522249783972e-05, "loss": 0.0634, "step": 10920 }, { "grad_norm": 0.45707619190216064, "learning_rate": 9.904200397844517e-05, "loss": 0.0802, "step": 10930 }, { "grad_norm": 0.5824089050292969, "learning_rate": 9.903878009589078e-05, "loss": 0.0702, "step": 10940 }, { "grad_norm": 0.6092053651809692, "learning_rate": 9.903555085052915e-05, "loss": 0.0786, "step": 10950 }, { "grad_norm": 0.3537718653678894, "learning_rate": 9.903231624271338e-05, "loss": 0.0659, "step": 10960 }, { "grad_norm": 0.4525872468948364, "learning_rate": 9.902907627279724e-05, "loss": 0.0682, "step": 10970 }, { "grad_norm": 0.6279293894767761, "learning_rate": 9.902583094113504e-05, "loss": 0.0667, "step": 10980 }, { "grad_norm": 0.4571482837200165, "learning_rate": 9.902258024808168e-05, "loss": 0.064, "step": 10990 }, { "grad_norm": 0.6257559061050415, "learning_rate": 9.901932419399264e-05, "loss": 0.0602, "step": 11000 }, { "grad_norm": 0.6805198788642883, "learning_rate": 9.9016062779224e-05, "loss": 0.0607, "step": 11010 }, { "grad_norm": 0.5670079588890076, "learning_rate": 9.901279600413242e-05, "loss": 0.0612, "step": 11020 }, { "grad_norm": 0.6296303868293762, "learning_rate": 9.900952386907518e-05, "loss": 0.0708, "step": 11030 }, { "grad_norm": 0.7550753355026245, "learning_rate": 9.90062463744101e-05, "loss": 0.0674, "step": 11040 }, { "grad_norm": 0.529077410697937, "learning_rate": 9.900296352049558e-05, "loss": 0.0597, "step": 11050 }, { "grad_norm": 0.47122836112976074, "learning_rate": 9.899967530769065e-05, "loss": 0.0771, "step": 11060 }, { "grad_norm": 0.4562954008579254, "learning_rate": 9.899638173635489e-05, "loss": 0.0533, "step": 11070 }, { "grad_norm": 0.59267657995224, "learning_rate": 9.899308280684849e-05, "loss": 0.0629, "step": 11080 }, { "grad_norm": 0.4524705708026886, "learning_rate": 9.898977851953222e-05, "loss": 0.0586, "step": 11090 }, { "grad_norm": 0.38104477524757385, "learning_rate": 9.898646887476741e-05, "loss": 0.0685, "step": 11100 }, { "grad_norm": 0.5135459899902344, "learning_rate": 9.898315387291603e-05, "loss": 0.0654, "step": 11110 }, { "grad_norm": 0.5733463764190674, "learning_rate": 9.89798335143406e-05, "loss": 0.0628, "step": 11120 }, { "grad_norm": 0.4083269536495209, "learning_rate": 9.897650779940419e-05, "loss": 0.0611, "step": 11130 }, { "grad_norm": 0.41903701424598694, "learning_rate": 9.897317672847054e-05, "loss": 0.0592, "step": 11140 }, { "grad_norm": 0.36626213788986206, "learning_rate": 9.89698403019039e-05, "loss": 0.051, "step": 11150 }, { "grad_norm": 0.5287185907363892, "learning_rate": 9.896649852006917e-05, "loss": 0.0599, "step": 11160 }, { "grad_norm": 0.40561598539352417, "learning_rate": 9.896315138333177e-05, "loss": 0.0635, "step": 11170 }, { "grad_norm": 0.669829249382019, "learning_rate": 9.895979889205774e-05, "loss": 0.0613, "step": 11180 }, { "grad_norm": 0.44537243247032166, "learning_rate": 9.895644104661372e-05, "loss": 0.0669, "step": 11190 }, { "grad_norm": 0.3487634062767029, "learning_rate": 9.895307784736691e-05, "loss": 0.0586, "step": 11200 }, { "grad_norm": 0.48868846893310547, "learning_rate": 9.894970929468512e-05, "loss": 0.0626, "step": 11210 }, { "grad_norm": 0.5064771771430969, "learning_rate": 9.89463353889367e-05, "loss": 0.0591, "step": 11220 }, { "grad_norm": 0.5429352521896362, "learning_rate": 9.894295613049065e-05, "loss": 0.0734, "step": 11230 }, { "grad_norm": 0.5730993151664734, "learning_rate": 9.893957151971649e-05, "loss": 0.0597, "step": 11240 }, { "grad_norm": 0.479113906621933, "learning_rate": 9.893618155698436e-05, "loss": 0.0472, "step": 11250 }, { "grad_norm": 0.5566200017929077, "learning_rate": 9.8932786242665e-05, "loss": 0.0498, "step": 11260 }, { "grad_norm": 0.37744879722595215, "learning_rate": 9.89293855771297e-05, "loss": 0.0553, "step": 11270 }, { "grad_norm": 0.6170980930328369, "learning_rate": 9.892597956075036e-05, "loss": 0.0731, "step": 11280 }, { "grad_norm": 0.5219377279281616, "learning_rate": 9.892256819389947e-05, "loss": 0.0656, "step": 11290 }, { "grad_norm": 0.47125643491744995, "learning_rate": 9.891915147695006e-05, "loss": 0.0643, "step": 11300 }, { "grad_norm": 0.5459251403808594, "learning_rate": 9.891572941027577e-05, "loss": 0.0639, "step": 11310 }, { "grad_norm": 0.34247350692749023, "learning_rate": 9.89123019942509e-05, "loss": 0.0545, "step": 11320 }, { "grad_norm": 0.4107954502105713, "learning_rate": 9.89088692292502e-05, "loss": 0.0565, "step": 11330 }, { "grad_norm": 0.5032163262367249, "learning_rate": 9.89054311156491e-05, "loss": 0.0594, "step": 11340 }, { "grad_norm": 0.47529539465904236, "learning_rate": 9.890198765382357e-05, "loss": 0.0834, "step": 11350 }, { "grad_norm": 0.46190059185028076, "learning_rate": 9.889853884415021e-05, "loss": 0.0618, "step": 11360 }, { "grad_norm": 0.600127637386322, "learning_rate": 9.889508468700614e-05, "loss": 0.0538, "step": 11370 }, { "grad_norm": 0.5349459648132324, "learning_rate": 9.889162518276915e-05, "loss": 0.0594, "step": 11380 }, { "grad_norm": 0.5594726800918579, "learning_rate": 9.888816033181752e-05, "loss": 0.0638, "step": 11390 }, { "grad_norm": 0.4981619417667389, "learning_rate": 9.888469013453018e-05, "loss": 0.0712, "step": 11400 }, { "grad_norm": 0.6819640398025513, "learning_rate": 9.888121459128663e-05, "loss": 0.059, "step": 11410 }, { "grad_norm": 0.4173343777656555, "learning_rate": 9.887773370246693e-05, "loss": 0.0773, "step": 11420 }, { "grad_norm": 0.4559093713760376, "learning_rate": 9.887424746845177e-05, "loss": 0.0664, "step": 11430 }, { "grad_norm": 0.5317911505699158, "learning_rate": 9.887075588962239e-05, "loss": 0.056, "step": 11440 }, { "grad_norm": 0.6312896013259888, "learning_rate": 9.88672589663606e-05, "loss": 0.0579, "step": 11450 }, { "grad_norm": 0.5002418756484985, "learning_rate": 9.886375669904886e-05, "loss": 0.0684, "step": 11460 }, { "grad_norm": 0.42580828070640564, "learning_rate": 9.886024908807014e-05, "loss": 0.0598, "step": 11470 }, { "grad_norm": 0.7934322953224182, "learning_rate": 9.885673613380806e-05, "loss": 0.0652, "step": 11480 }, { "grad_norm": 0.48178187012672424, "learning_rate": 9.885321783664676e-05, "loss": 0.0718, "step": 11490 }, { "grad_norm": 0.4668395221233368, "learning_rate": 9.884969419697101e-05, "loss": 0.0649, "step": 11500 }, { "grad_norm": 0.4255613684654236, "learning_rate": 9.884616521516614e-05, "loss": 0.0654, "step": 11510 }, { "grad_norm": 0.362962543964386, "learning_rate": 9.88426308916181e-05, "loss": 0.0621, "step": 11520 }, { "grad_norm": 0.4858482778072357, "learning_rate": 9.883909122671335e-05, "loss": 0.0582, "step": 11530 }, { "grad_norm": 0.3240518569946289, "learning_rate": 9.883554622083904e-05, "loss": 0.0563, "step": 11540 }, { "grad_norm": 0.47110164165496826, "learning_rate": 9.88319958743828e-05, "loss": 0.0611, "step": 11550 }, { "grad_norm": 0.48791953921318054, "learning_rate": 9.882844018773291e-05, "loss": 0.0604, "step": 11560 }, { "grad_norm": 0.42533740401268005, "learning_rate": 9.882487916127823e-05, "loss": 0.0605, "step": 11570 }, { "grad_norm": 0.5700827240943909, "learning_rate": 9.882131279540815e-05, "loss": 0.0689, "step": 11580 }, { "grad_norm": 0.47537925839424133, "learning_rate": 9.881774109051271e-05, "loss": 0.0633, "step": 11590 }, { "grad_norm": 0.5357412099838257, "learning_rate": 9.881416404698252e-05, "loss": 0.0544, "step": 11600 }, { "grad_norm": 0.4146280884742737, "learning_rate": 9.881058166520873e-05, "loss": 0.0714, "step": 11610 }, { "grad_norm": 0.5990660786628723, "learning_rate": 9.880699394558311e-05, "loss": 0.0741, "step": 11620 }, { "grad_norm": 0.9357838034629822, "learning_rate": 9.880340088849801e-05, "loss": 0.0846, "step": 11630 }, { "grad_norm": 0.6477710008621216, "learning_rate": 9.879980249434637e-05, "loss": 0.0618, "step": 11640 }, { "grad_norm": 0.6040326952934265, "learning_rate": 9.879619876352168e-05, "loss": 0.0704, "step": 11650 }, { "grad_norm": 1.1634881496429443, "learning_rate": 9.879258969641809e-05, "loss": 0.0743, "step": 11660 }, { "grad_norm": 0.44512107968330383, "learning_rate": 9.878897529343023e-05, "loss": 0.0603, "step": 11670 }, { "grad_norm": 0.5746411085128784, "learning_rate": 9.878535555495338e-05, "loss": 0.0783, "step": 11680 }, { "grad_norm": 0.6375528573989868, "learning_rate": 9.87817304813834e-05, "loss": 0.0686, "step": 11690 }, { "grad_norm": 0.532456636428833, "learning_rate": 9.877810007311671e-05, "loss": 0.0702, "step": 11700 }, { "grad_norm": 0.49254581332206726, "learning_rate": 9.877446433055035e-05, "loss": 0.0655, "step": 11710 }, { "grad_norm": 0.48113659024238586, "learning_rate": 9.877082325408191e-05, "loss": 0.0585, "step": 11720 }, { "grad_norm": 0.4118494391441345, "learning_rate": 9.876717684410954e-05, "loss": 0.0557, "step": 11730 }, { "grad_norm": 0.5874334573745728, "learning_rate": 9.876352510103204e-05, "loss": 0.0657, "step": 11740 }, { "grad_norm": 0.5344106554985046, "learning_rate": 9.875986802524875e-05, "loss": 0.0687, "step": 11750 }, { "grad_norm": 0.4630470871925354, "learning_rate": 9.87562056171596e-05, "loss": 0.0569, "step": 11760 }, { "grad_norm": 0.5127207040786743, "learning_rate": 9.875253787716511e-05, "loss": 0.0597, "step": 11770 }, { "grad_norm": 0.5797615647315979, "learning_rate": 9.874886480566637e-05, "loss": 0.0509, "step": 11780 }, { "grad_norm": 0.3274081349372864, "learning_rate": 9.874518640306507e-05, "loss": 0.05, "step": 11790 }, { "grad_norm": 0.3172381520271301, "learning_rate": 9.874150266976347e-05, "loss": 0.0529, "step": 11800 }, { "grad_norm": 0.4034438133239746, "learning_rate": 9.873781360616443e-05, "loss": 0.068, "step": 11810 }, { "grad_norm": 0.3570849597454071, "learning_rate": 9.873411921267137e-05, "loss": 0.0575, "step": 11820 }, { "grad_norm": 0.6039263606071472, "learning_rate": 9.873041948968829e-05, "loss": 0.0786, "step": 11830 }, { "grad_norm": 0.46031326055526733, "learning_rate": 9.872671443761981e-05, "loss": 0.0532, "step": 11840 }, { "grad_norm": 0.37493470311164856, "learning_rate": 9.872300405687109e-05, "loss": 0.0598, "step": 11850 }, { "grad_norm": 0.415332168340683, "learning_rate": 9.871928834784792e-05, "loss": 0.0648, "step": 11860 }, { "grad_norm": 0.6076071262359619, "learning_rate": 9.871556731095661e-05, "loss": 0.0595, "step": 11870 }, { "grad_norm": 0.4070660471916199, "learning_rate": 9.871184094660411e-05, "loss": 0.0706, "step": 11880 }, { "grad_norm": 0.5459979772567749, "learning_rate": 9.870810925519791e-05, "loss": 0.0581, "step": 11890 }, { "grad_norm": 0.5444684028625488, "learning_rate": 9.870437223714612e-05, "loss": 0.0688, "step": 11900 }, { "grad_norm": 0.5101686120033264, "learning_rate": 9.87006298928574e-05, "loss": 0.0714, "step": 11910 }, { "grad_norm": 0.39151689410209656, "learning_rate": 9.869688222274103e-05, "loss": 0.055, "step": 11920 }, { "grad_norm": 0.39157241582870483, "learning_rate": 9.869312922720681e-05, "loss": 0.0568, "step": 11930 }, { "grad_norm": 0.44698259234428406, "learning_rate": 9.868937090666521e-05, "loss": 0.0508, "step": 11940 }, { "grad_norm": 0.4412986934185028, "learning_rate": 9.86856072615272e-05, "loss": 0.0422, "step": 11950 }, { "grad_norm": 0.4562819004058838, "learning_rate": 9.868183829220438e-05, "loss": 0.0475, "step": 11960 }, { "grad_norm": 0.6331861615180969, "learning_rate": 9.867806399910893e-05, "loss": 0.066, "step": 11970 }, { "grad_norm": 0.5605736374855042, "learning_rate": 9.867428438265356e-05, "loss": 0.0584, "step": 11980 }, { "grad_norm": 0.4144655168056488, "learning_rate": 9.867049944325165e-05, "loss": 0.058, "step": 11990 }, { "grad_norm": 0.4931674897670746, "learning_rate": 9.86667091813171e-05, "loss": 0.0673, "step": 12000 }, { "grad_norm": 0.4928418695926666, "learning_rate": 9.866291359726438e-05, "loss": 0.0601, "step": 12010 }, { "grad_norm": 0.4733254909515381, "learning_rate": 9.865911269150861e-05, "loss": 0.0622, "step": 12020 }, { "grad_norm": 0.38755002617836, "learning_rate": 9.865530646446544e-05, "loss": 0.059, "step": 12030 }, { "grad_norm": 0.5382923483848572, "learning_rate": 9.86514949165511e-05, "loss": 0.0684, "step": 12040 }, { "grad_norm": 0.544581949710846, "learning_rate": 9.864767804818243e-05, "loss": 0.0644, "step": 12050 }, { "grad_norm": 0.512154757976532, "learning_rate": 9.86438558597768e-05, "loss": 0.0628, "step": 12060 }, { "grad_norm": 0.38323748111724854, "learning_rate": 9.864002835175225e-05, "loss": 0.0817, "step": 12070 }, { "grad_norm": 0.5873264670372009, "learning_rate": 9.863619552452734e-05, "loss": 0.0607, "step": 12080 }, { "grad_norm": 0.4607926607131958, "learning_rate": 9.863235737852119e-05, "loss": 0.0599, "step": 12090 }, { "grad_norm": 0.44374480843544006, "learning_rate": 9.862851391415356e-05, "loss": 0.069, "step": 12100 }, { "grad_norm": 0.47468382120132446, "learning_rate": 9.862466513184477e-05, "loss": 0.0631, "step": 12110 }, { "grad_norm": 0.5557988286018372, "learning_rate": 9.86208110320157e-05, "loss": 0.052, "step": 12120 }, { "grad_norm": 0.43850353360176086, "learning_rate": 9.861695161508784e-05, "loss": 0.0581, "step": 12130 }, { "grad_norm": 0.6327385306358337, "learning_rate": 9.861308688148324e-05, "loss": 0.059, "step": 12140 }, { "grad_norm": 0.5102418065071106, "learning_rate": 9.860921683162455e-05, "loss": 0.0573, "step": 12150 }, { "grad_norm": 0.3516898453235626, "learning_rate": 9.860534146593499e-05, "loss": 0.0474, "step": 12160 }, { "grad_norm": 0.4659164249897003, "learning_rate": 9.860146078483836e-05, "loss": 0.0592, "step": 12170 }, { "grad_norm": 0.5904263854026794, "learning_rate": 9.859757478875905e-05, "loss": 0.0686, "step": 12180 }, { "grad_norm": 0.4387100040912628, "learning_rate": 9.859368347812204e-05, "loss": 0.0603, "step": 12190 }, { "grad_norm": 0.35866162180900574, "learning_rate": 9.858978685335285e-05, "loss": 0.0657, "step": 12200 }, { "grad_norm": 0.4266168773174286, "learning_rate": 9.858588491487763e-05, "loss": 0.0767, "step": 12210 }, { "grad_norm": 0.6909045577049255, "learning_rate": 9.858197766312308e-05, "loss": 0.0471, "step": 12220 }, { "grad_norm": 0.5447998642921448, "learning_rate": 9.857806509851649e-05, "loss": 0.0709, "step": 12230 }, { "grad_norm": 0.3588263988494873, "learning_rate": 9.857414722148574e-05, "loss": 0.0556, "step": 12240 }, { "grad_norm": 0.4411187171936035, "learning_rate": 9.857022403245928e-05, "loss": 0.0725, "step": 12250 }, { "grad_norm": 0.6218165159225464, "learning_rate": 9.856629553186615e-05, "loss": 0.0813, "step": 12260 }, { "grad_norm": 0.6165293455123901, "learning_rate": 9.856236172013595e-05, "loss": 0.0663, "step": 12270 }, { "grad_norm": 0.49153196811676025, "learning_rate": 9.85584225976989e-05, "loss": 0.0566, "step": 12280 }, { "grad_norm": 0.4001125991344452, "learning_rate": 9.855447816498575e-05, "loss": 0.0627, "step": 12290 }, { "grad_norm": 0.5242616534233093, "learning_rate": 9.855052842242787e-05, "loss": 0.0502, "step": 12300 }, { "grad_norm": 0.7651994824409485, "learning_rate": 9.85465733704572e-05, "loss": 0.0591, "step": 12310 }, { "grad_norm": 0.7386987805366516, "learning_rate": 9.854261300950624e-05, "loss": 0.0693, "step": 12320 }, { "grad_norm": 0.45058149099349976, "learning_rate": 9.853864734000813e-05, "loss": 0.061, "step": 12330 }, { "grad_norm": 0.556979775428772, "learning_rate": 9.85346763623965e-05, "loss": 0.0724, "step": 12340 }, { "grad_norm": 0.8280488848686218, "learning_rate": 9.853070007710564e-05, "loss": 0.0684, "step": 12350 }, { "grad_norm": 0.44771450757980347, "learning_rate": 9.85267184845704e-05, "loss": 0.0709, "step": 12360 }, { "grad_norm": 0.4334341883659363, "learning_rate": 9.852273158522616e-05, "loss": 0.0525, "step": 12370 }, { "grad_norm": 0.4514278769493103, "learning_rate": 9.851873937950896e-05, "loss": 0.0565, "step": 12380 }, { "grad_norm": 0.6842033267021179, "learning_rate": 9.851474186785537e-05, "loss": 0.0599, "step": 12390 }, { "grad_norm": 0.6051158905029297, "learning_rate": 9.851073905070254e-05, "loss": 0.0682, "step": 12400 }, { "grad_norm": 0.5243008732795715, "learning_rate": 9.850673092848824e-05, "loss": 0.0741, "step": 12410 }, { "grad_norm": 0.5516024231910706, "learning_rate": 9.850271750165077e-05, "loss": 0.0661, "step": 12420 }, { "grad_norm": 1.2966617345809937, "learning_rate": 9.849869877062902e-05, "loss": 0.0857, "step": 12430 }, { "grad_norm": 0.4299858510494232, "learning_rate": 9.849467473586252e-05, "loss": 0.0717, "step": 12440 }, { "grad_norm": 0.5860083699226379, "learning_rate": 9.849064539779127e-05, "loss": 0.0591, "step": 12450 }, { "grad_norm": 0.42706602811813354, "learning_rate": 9.848661075685594e-05, "loss": 0.0622, "step": 12460 }, { "grad_norm": 0.4192958474159241, "learning_rate": 9.848257081349778e-05, "loss": 0.0673, "step": 12470 }, { "grad_norm": 0.7423099875450134, "learning_rate": 9.847852556815856e-05, "loss": 0.0609, "step": 12480 }, { "grad_norm": 0.4339127838611603, "learning_rate": 9.847447502128067e-05, "loss": 0.0581, "step": 12490 }, { "grad_norm": 0.6438903212547302, "learning_rate": 9.847041917330708e-05, "loss": 0.0663, "step": 12500 }, { "grad_norm": 0.555747389793396, "learning_rate": 9.846635802468132e-05, "loss": 0.0765, "step": 12510 }, { "grad_norm": 0.4798680543899536, "learning_rate": 9.84622915758475e-05, "loss": 0.0584, "step": 12520 }, { "grad_norm": 0.3603353798389435, "learning_rate": 9.845821982725034e-05, "loss": 0.0585, "step": 12530 }, { "grad_norm": 0.4311923682689667, "learning_rate": 9.845414277933514e-05, "loss": 0.0635, "step": 12540 }, { "grad_norm": 0.5527379512786865, "learning_rate": 9.845006043254771e-05, "loss": 0.0598, "step": 12550 }, { "grad_norm": 0.49782586097717285, "learning_rate": 9.844597278733451e-05, "loss": 0.0531, "step": 12560 }, { "grad_norm": 0.5030670762062073, "learning_rate": 9.844187984414259e-05, "loss": 0.0603, "step": 12570 }, { "grad_norm": 0.31546613574028015, "learning_rate": 9.84377816034195e-05, "loss": 0.0544, "step": 12580 }, { "grad_norm": 0.44687893986701965, "learning_rate": 9.843367806561345e-05, "loss": 0.0604, "step": 12590 }, { "grad_norm": 0.4051026701927185, "learning_rate": 9.842956923117317e-05, "loss": 0.0575, "step": 12600 }, { "grad_norm": 0.40675216913223267, "learning_rate": 9.842545510054802e-05, "loss": 0.0535, "step": 12610 }, { "grad_norm": 0.519950807094574, "learning_rate": 9.842133567418792e-05, "loss": 0.0543, "step": 12620 }, { "grad_norm": 0.5334843993186951, "learning_rate": 9.841721095254333e-05, "loss": 0.0534, "step": 12630 }, { "grad_norm": 0.44413232803344727, "learning_rate": 9.841308093606537e-05, "loss": 0.073, "step": 12640 }, { "grad_norm": 0.6175841689109802, "learning_rate": 9.840894562520565e-05, "loss": 0.0576, "step": 12650 }, { "grad_norm": 0.4418312609195709, "learning_rate": 9.840480502041642e-05, "loss": 0.0625, "step": 12660 }, { "grad_norm": 0.4252232015132904, "learning_rate": 9.840065912215049e-05, "loss": 0.0513, "step": 12670 }, { "grad_norm": 0.5116055607795715, "learning_rate": 9.839650793086124e-05, "loss": 0.0635, "step": 12680 }, { "grad_norm": 0.5057145953178406, "learning_rate": 9.839235144700265e-05, "loss": 0.057, "step": 12690 }, { "grad_norm": 0.5065198540687561, "learning_rate": 9.838818967102926e-05, "loss": 0.0656, "step": 12700 }, { "grad_norm": 0.7442583441734314, "learning_rate": 9.83840226033962e-05, "loss": 0.0535, "step": 12710 }, { "grad_norm": 0.4830484986305237, "learning_rate": 9.837985024455918e-05, "loss": 0.058, "step": 12720 }, { "grad_norm": 0.6278693675994873, "learning_rate": 9.837567259497447e-05, "loss": 0.0534, "step": 12730 }, { "grad_norm": 0.33731454610824585, "learning_rate": 9.837148965509894e-05, "loss": 0.0541, "step": 12740 }, { "grad_norm": 0.43717217445373535, "learning_rate": 9.836730142539001e-05, "loss": 0.0522, "step": 12750 }, { "grad_norm": 0.5771032571792603, "learning_rate": 9.836310790630574e-05, "loss": 0.0629, "step": 12760 }, { "grad_norm": 0.6096678972244263, "learning_rate": 9.83589090983047e-05, "loss": 0.07, "step": 12770 }, { "grad_norm": 0.5766184329986572, "learning_rate": 9.835470500184605e-05, "loss": 0.0697, "step": 12780 }, { "grad_norm": 0.6161572337150574, "learning_rate": 9.835049561738957e-05, "loss": 0.0623, "step": 12790 }, { "grad_norm": 0.39830470085144043, "learning_rate": 9.834628094539558e-05, "loss": 0.057, "step": 12800 }, { "grad_norm": 0.5689314603805542, "learning_rate": 9.834206098632499e-05, "loss": 0.0549, "step": 12810 }, { "grad_norm": 0.7192450165748596, "learning_rate": 9.833783574063931e-05, "loss": 0.078, "step": 12820 }, { "grad_norm": 0.5981547236442566, "learning_rate": 9.833360520880058e-05, "loss": 0.0799, "step": 12830 }, { "grad_norm": 0.4303728938102722, "learning_rate": 9.832936939127144e-05, "loss": 0.0511, "step": 12840 }, { "grad_norm": 0.4161112606525421, "learning_rate": 9.832512828851515e-05, "loss": 0.0584, "step": 12850 }, { "grad_norm": 0.48438552021980286, "learning_rate": 9.832088190099546e-05, "loss": 0.0568, "step": 12860 }, { "grad_norm": 0.537100613117218, "learning_rate": 9.831663022917679e-05, "loss": 0.0557, "step": 12870 }, { "grad_norm": 0.4993099868297577, "learning_rate": 9.831237327352407e-05, "loss": 0.0544, "step": 12880 }, { "grad_norm": 0.4245164096355438, "learning_rate": 9.830811103450286e-05, "loss": 0.0515, "step": 12890 }, { "grad_norm": 0.4116903245449066, "learning_rate": 9.830384351257924e-05, "loss": 0.0661, "step": 12900 }, { "grad_norm": 0.6689679622650146, "learning_rate": 9.829957070821993e-05, "loss": 0.0557, "step": 12910 }, { "grad_norm": 0.4670776128768921, "learning_rate": 9.829529262189218e-05, "loss": 0.0456, "step": 12920 }, { "grad_norm": 0.48890504240989685, "learning_rate": 9.829100925406385e-05, "loss": 0.0658, "step": 12930 }, { "grad_norm": 0.5924577713012695, "learning_rate": 9.828672060520333e-05, "loss": 0.0631, "step": 12940 }, { "grad_norm": 0.40141868591308594, "learning_rate": 9.828242667577966e-05, "loss": 0.054, "step": 12950 }, { "grad_norm": 0.44179508090019226, "learning_rate": 9.82781274662624e-05, "loss": 0.0553, "step": 12960 }, { "grad_norm": 0.49851194024086, "learning_rate": 9.82738229771217e-05, "loss": 0.0532, "step": 12970 }, { "grad_norm": 0.30699682235717773, "learning_rate": 9.826951320882829e-05, "loss": 0.0445, "step": 12980 }, { "grad_norm": 0.555006742477417, "learning_rate": 9.826519816185351e-05, "loss": 0.0686, "step": 12990 }, { "grad_norm": 0.5306680798530579, "learning_rate": 9.826087783666921e-05, "loss": 0.0524, "step": 13000 }, { "grad_norm": 0.450318843126297, "learning_rate": 9.825655223374787e-05, "loss": 0.0729, "step": 13010 }, { "grad_norm": 0.5783497095108032, "learning_rate": 9.825222135356253e-05, "loss": 0.0719, "step": 13020 }, { "grad_norm": 0.38217201828956604, "learning_rate": 9.82478851965868e-05, "loss": 0.0633, "step": 13030 }, { "grad_norm": 0.6007168292999268, "learning_rate": 9.82435437632949e-05, "loss": 0.0605, "step": 13040 }, { "grad_norm": 0.37855467200279236, "learning_rate": 9.823919705416158e-05, "loss": 0.0434, "step": 13050 }, { "grad_norm": 0.5884948372840881, "learning_rate": 9.82348450696622e-05, "loss": 0.052, "step": 13060 }, { "grad_norm": 0.667320966720581, "learning_rate": 9.823048781027268e-05, "loss": 0.0619, "step": 13070 }, { "grad_norm": 0.40825244784355164, "learning_rate": 9.822612527646953e-05, "loss": 0.0536, "step": 13080 }, { "grad_norm": 0.5304440855979919, "learning_rate": 9.822175746872984e-05, "loss": 0.0604, "step": 13090 }, { "grad_norm": 0.5104284882545471, "learning_rate": 9.821738438753123e-05, "loss": 0.0618, "step": 13100 }, { "grad_norm": 0.452016144990921, "learning_rate": 9.821300603335196e-05, "loss": 0.0552, "step": 13110 }, { "grad_norm": 0.6425236463546753, "learning_rate": 9.820862240667085e-05, "loss": 0.057, "step": 13120 }, { "grad_norm": 0.5054212808609009, "learning_rate": 9.820423350796726e-05, "loss": 0.0705, "step": 13130 }, { "grad_norm": 0.45461922883987427, "learning_rate": 9.819983933772118e-05, "loss": 0.065, "step": 13140 }, { "grad_norm": 0.6331923007965088, "learning_rate": 9.819543989641314e-05, "loss": 0.0602, "step": 13150 }, { "grad_norm": 0.7741827368736267, "learning_rate": 9.819103518452423e-05, "loss": 0.0536, "step": 13160 }, { "grad_norm": 0.43601858615875244, "learning_rate": 9.818662520253618e-05, "loss": 0.0563, "step": 13170 }, { "grad_norm": 0.3842916786670685, "learning_rate": 9.818220995093126e-05, "loss": 0.0639, "step": 13180 }, { "grad_norm": 0.43437543511390686, "learning_rate": 9.817778943019228e-05, "loss": 0.0514, "step": 13190 }, { "grad_norm": 0.516656756401062, "learning_rate": 9.81733636408027e-05, "loss": 0.0717, "step": 13200 }, { "grad_norm": 0.3723599910736084, "learning_rate": 9.816893258324649e-05, "loss": 0.0535, "step": 13210 }, { "grad_norm": 0.36868754029273987, "learning_rate": 9.816449625800823e-05, "loss": 0.0534, "step": 13220 }, { "grad_norm": 0.46379438042640686, "learning_rate": 9.816005466557308e-05, "loss": 0.0601, "step": 13230 }, { "grad_norm": 0.49410590529441833, "learning_rate": 9.815560780642674e-05, "loss": 0.0658, "step": 13240 }, { "grad_norm": 0.36824285984039307, "learning_rate": 9.815115568105555e-05, "loss": 0.0507, "step": 13250 }, { "grad_norm": 0.5761622786521912, "learning_rate": 9.814669828994638e-05, "loss": 0.054, "step": 13260 }, { "grad_norm": 0.4923170804977417, "learning_rate": 9.814223563358665e-05, "loss": 0.0503, "step": 13270 }, { "grad_norm": 0.44520848989486694, "learning_rate": 9.813776771246443e-05, "loss": 0.0586, "step": 13280 }, { "grad_norm": 0.6218951344490051, "learning_rate": 9.813329452706829e-05, "loss": 0.0653, "step": 13290 }, { "grad_norm": 0.815660834312439, "learning_rate": 9.812881607788744e-05, "loss": 0.0658, "step": 13300 }, { "grad_norm": 0.515994131565094, "learning_rate": 9.812433236541163e-05, "loss": 0.0675, "step": 13310 }, { "grad_norm": 0.5080538392066956, "learning_rate": 9.811984339013116e-05, "loss": 0.0947, "step": 13320 }, { "grad_norm": 0.5543578267097473, "learning_rate": 9.811534915253698e-05, "loss": 0.0696, "step": 13330 }, { "grad_norm": 1.0149269104003906, "learning_rate": 9.811084965312056e-05, "loss": 0.0633, "step": 13340 }, { "grad_norm": 0.49413758516311646, "learning_rate": 9.810634489237396e-05, "loss": 0.0579, "step": 13350 }, { "grad_norm": 0.6333500742912292, "learning_rate": 9.81018348707898e-05, "loss": 0.0624, "step": 13360 }, { "grad_norm": 0.44167235493659973, "learning_rate": 9.809731958886131e-05, "loss": 0.0704, "step": 13370 }, { "grad_norm": 0.687048077583313, "learning_rate": 9.809279904708224e-05, "loss": 0.0695, "step": 13380 }, { "grad_norm": 0.4980011582374573, "learning_rate": 9.808827324594699e-05, "loss": 0.0645, "step": 13390 }, { "grad_norm": 0.5268141031265259, "learning_rate": 9.808374218595046e-05, "loss": 0.0711, "step": 13400 }, { "grad_norm": 0.48069363832473755, "learning_rate": 9.80792058675882e-05, "loss": 0.0576, "step": 13410 }, { "grad_norm": 0.5572584271430969, "learning_rate": 9.807466429135627e-05, "loss": 0.0594, "step": 13420 }, { "grad_norm": 0.6814654469490051, "learning_rate": 9.807011745775132e-05, "loss": 0.069, "step": 13430 }, { "grad_norm": 0.6601583957672119, "learning_rate": 9.806556536727061e-05, "loss": 0.069, "step": 13440 }, { "grad_norm": 0.5102643966674805, "learning_rate": 9.806100802041193e-05, "loss": 0.0764, "step": 13450 }, { "grad_norm": 0.6156317591667175, "learning_rate": 9.805644541767368e-05, "loss": 0.0735, "step": 13460 }, { "grad_norm": 0.5193184614181519, "learning_rate": 9.805187755955478e-05, "loss": 0.0784, "step": 13470 }, { "grad_norm": 0.4193895757198334, "learning_rate": 9.804730444655483e-05, "loss": 0.0651, "step": 13480 }, { "grad_norm": 0.41101372241973877, "learning_rate": 9.804272607917388e-05, "loss": 0.064, "step": 13490 }, { "grad_norm": 0.6647838950157166, "learning_rate": 9.803814245791265e-05, "loss": 0.0669, "step": 13500 }, { "grad_norm": 0.624788761138916, "learning_rate": 9.803355358327239e-05, "loss": 0.0692, "step": 13510 }, { "grad_norm": 0.43444985151290894, "learning_rate": 9.802895945575492e-05, "loss": 0.0746, "step": 13520 }, { "grad_norm": 0.6215178370475769, "learning_rate": 9.802436007586266e-05, "loss": 0.0776, "step": 13530 }, { "grad_norm": 0.45474809408187866, "learning_rate": 9.801975544409858e-05, "loss": 0.0658, "step": 13540 }, { "grad_norm": 0.5608227252960205, "learning_rate": 9.801514556096625e-05, "loss": 0.0638, "step": 13550 }, { "grad_norm": 0.5439756512641907, "learning_rate": 9.801053042696977e-05, "loss": 0.0743, "step": 13560 }, { "grad_norm": 0.5421087145805359, "learning_rate": 9.800591004261388e-05, "loss": 0.0625, "step": 13570 }, { "grad_norm": 0.558663010597229, "learning_rate": 9.800128440840385e-05, "loss": 0.0672, "step": 13580 }, { "grad_norm": 0.6652270555496216, "learning_rate": 9.799665352484552e-05, "loss": 0.0804, "step": 13590 }, { "grad_norm": 0.5205460786819458, "learning_rate": 9.799201739244532e-05, "loss": 0.0906, "step": 13600 }, { "grad_norm": 0.449349582195282, "learning_rate": 9.798737601171025e-05, "loss": 0.0642, "step": 13610 }, { "grad_norm": 0.4965846538543701, "learning_rate": 9.79827293831479e-05, "loss": 0.0806, "step": 13620 }, { "grad_norm": 0.4537792205810547, "learning_rate": 9.797807750726638e-05, "loss": 0.0585, "step": 13630 }, { "grad_norm": 0.38896435499191284, "learning_rate": 9.797342038457446e-05, "loss": 0.0599, "step": 13640 }, { "grad_norm": 0.5584247708320618, "learning_rate": 9.796875801558141e-05, "loss": 0.0663, "step": 13650 }, { "grad_norm": 0.5512246489524841, "learning_rate": 9.79640904007971e-05, "loss": 0.0738, "step": 13660 }, { "grad_norm": 0.4483742415904999, "learning_rate": 9.795941754073199e-05, "loss": 0.065, "step": 13670 }, { "grad_norm": 0.4184880256652832, "learning_rate": 9.795473943589705e-05, "loss": 0.0616, "step": 13680 }, { "grad_norm": 0.6615806221961975, "learning_rate": 9.795005608680394e-05, "loss": 0.0559, "step": 13690 }, { "grad_norm": 0.49505701661109924, "learning_rate": 9.794536749396477e-05, "loss": 0.0626, "step": 13700 }, { "grad_norm": 0.4426066279411316, "learning_rate": 9.79406736578923e-05, "loss": 0.0506, "step": 13710 }, { "grad_norm": 0.5799975395202637, "learning_rate": 9.793597457909984e-05, "loss": 0.0516, "step": 13720 }, { "grad_norm": 0.4311286211013794, "learning_rate": 9.793127025810127e-05, "loss": 0.0514, "step": 13730 }, { "grad_norm": 0.42999354004859924, "learning_rate": 9.792656069541104e-05, "loss": 0.053, "step": 13740 }, { "grad_norm": 0.4882931411266327, "learning_rate": 9.79218458915442e-05, "loss": 0.0576, "step": 13750 }, { "grad_norm": 0.4422712028026581, "learning_rate": 9.791712584701634e-05, "loss": 0.075, "step": 13760 }, { "grad_norm": 0.41429415345191956, "learning_rate": 9.791240056234364e-05, "loss": 0.055, "step": 13770 }, { "grad_norm": 0.576885998249054, "learning_rate": 9.790767003804283e-05, "loss": 0.0491, "step": 13780 }, { "grad_norm": 0.5374026894569397, "learning_rate": 9.790293427463126e-05, "loss": 0.0499, "step": 13790 }, { "grad_norm": 0.5260573625564575, "learning_rate": 9.789819327262684e-05, "loss": 0.049, "step": 13800 }, { "grad_norm": 0.456820011138916, "learning_rate": 9.7893447032548e-05, "loss": 0.0532, "step": 13810 }, { "grad_norm": 0.45736315846443176, "learning_rate": 9.78886955549138e-05, "loss": 0.0584, "step": 13820 }, { "grad_norm": 0.4258723258972168, "learning_rate": 9.788393884024387e-05, "loss": 0.0606, "step": 13830 }, { "grad_norm": 0.49217861890792847, "learning_rate": 9.787917688905836e-05, "loss": 0.0676, "step": 13840 }, { "grad_norm": 0.5655595064163208, "learning_rate": 9.787440970187807e-05, "loss": 0.0553, "step": 13850 }, { "grad_norm": 0.45779451727867126, "learning_rate": 9.786963727922429e-05, "loss": 0.0648, "step": 13860 }, { "grad_norm": 0.3897847533226013, "learning_rate": 9.786485962161897e-05, "loss": 0.0681, "step": 13870 }, { "grad_norm": 0.37480658292770386, "learning_rate": 9.786007672958455e-05, "loss": 0.0547, "step": 13880 }, { "grad_norm": 0.6460938453674316, "learning_rate": 9.78552886036441e-05, "loss": 0.0855, "step": 13890 }, { "grad_norm": 0.6109288930892944, "learning_rate": 9.785049524432124e-05, "loss": 0.0524, "step": 13900 }, { "grad_norm": 0.5202329158782959, "learning_rate": 9.784569665214016e-05, "loss": 0.0558, "step": 13910 }, { "grad_norm": 0.5188629031181335, "learning_rate": 9.784089282762563e-05, "loss": 0.0652, "step": 13920 }, { "grad_norm": 0.4714964032173157, "learning_rate": 9.7836083771303e-05, "loss": 0.0546, "step": 13930 }, { "grad_norm": 0.5204988718032837, "learning_rate": 9.783126948369817e-05, "loss": 0.065, "step": 13940 }, { "grad_norm": 0.6833059787750244, "learning_rate": 9.78264499653376e-05, "loss": 0.0564, "step": 13950 }, { "grad_norm": 0.36089828610420227, "learning_rate": 9.782162521674838e-05, "loss": 0.0504, "step": 13960 }, { "grad_norm": 0.627432644367218, "learning_rate": 9.781679523845812e-05, "loss": 0.0858, "step": 13970 }, { "grad_norm": 0.5081884264945984, "learning_rate": 9.781196003099502e-05, "loss": 0.0724, "step": 13980 }, { "grad_norm": 0.4460401237010956, "learning_rate": 9.780711959488786e-05, "loss": 0.073, "step": 13990 }, { "grad_norm": 0.5149438381195068, "learning_rate": 9.780227393066599e-05, "loss": 0.0634, "step": 14000 }, { "grad_norm": 0.5698991417884827, "learning_rate": 9.77974230388593e-05, "loss": 0.0533, "step": 14010 }, { "grad_norm": 0.4414459764957428, "learning_rate": 9.779256691999829e-05, "loss": 0.0719, "step": 14020 }, { "grad_norm": 0.41464340686798096, "learning_rate": 9.778770557461403e-05, "loss": 0.0629, "step": 14030 }, { "grad_norm": 0.5727075934410095, "learning_rate": 9.778283900323812e-05, "loss": 0.0513, "step": 14040 }, { "grad_norm": 0.5166894197463989, "learning_rate": 9.777796720640277e-05, "loss": 0.0515, "step": 14050 }, { "grad_norm": 0.45349910855293274, "learning_rate": 9.777309018464078e-05, "loss": 0.0607, "step": 14060 }, { "grad_norm": 0.5321807265281677, "learning_rate": 9.776820793848547e-05, "loss": 0.0549, "step": 14070 }, { "grad_norm": 0.6038118600845337, "learning_rate": 9.776332046847075e-05, "loss": 0.0581, "step": 14080 }, { "grad_norm": 0.4944215714931488, "learning_rate": 9.775842777513111e-05, "loss": 0.0618, "step": 14090 }, { "grad_norm": 0.6073762774467468, "learning_rate": 9.775352985900163e-05, "loss": 0.0649, "step": 14100 }, { "grad_norm": 0.49578630924224854, "learning_rate": 9.774862672061791e-05, "loss": 0.0718, "step": 14110 }, { "grad_norm": 0.5219376683235168, "learning_rate": 9.774371836051616e-05, "loss": 0.0638, "step": 14120 }, { "grad_norm": 0.41541463136672974, "learning_rate": 9.773880477923315e-05, "loss": 0.0656, "step": 14130 }, { "grad_norm": 0.4086114466190338, "learning_rate": 9.773388597730623e-05, "loss": 0.0635, "step": 14140 }, { "grad_norm": 0.5783771872520447, "learning_rate": 9.77289619552733e-05, "loss": 0.0506, "step": 14150 }, { "grad_norm": 0.6343988180160522, "learning_rate": 9.772403271367285e-05, "loss": 0.0758, "step": 14160 }, { "grad_norm": 0.5075154900550842, "learning_rate": 9.771909825304396e-05, "loss": 0.0615, "step": 14170 }, { "grad_norm": 0.5554609894752502, "learning_rate": 9.771415857392619e-05, "loss": 0.0612, "step": 14180 }, { "grad_norm": 0.676476001739502, "learning_rate": 9.770921367685978e-05, "loss": 0.0644, "step": 14190 }, { "grad_norm": 0.76664799451828, "learning_rate": 9.770426356238551e-05, "loss": 0.0588, "step": 14200 }, { "grad_norm": 0.49374648928642273, "learning_rate": 9.769930823104469e-05, "loss": 0.0619, "step": 14210 }, { "grad_norm": 0.4102998673915863, "learning_rate": 9.769434768337926e-05, "loss": 0.0703, "step": 14220 }, { "grad_norm": 0.4703174829483032, "learning_rate": 9.768938191993164e-05, "loss": 0.0531, "step": 14230 }, { "grad_norm": 0.5504488945007324, "learning_rate": 9.768441094124494e-05, "loss": 0.0565, "step": 14240 }, { "grad_norm": 0.3571348786354065, "learning_rate": 9.767943474786275e-05, "loss": 0.0573, "step": 14250 }, { "grad_norm": 0.4304978549480438, "learning_rate": 9.767445334032923e-05, "loss": 0.0532, "step": 14260 }, { "grad_norm": 0.4199924170970917, "learning_rate": 9.766946671918919e-05, "loss": 0.0556, "step": 14270 }, { "grad_norm": 0.5158932209014893, "learning_rate": 9.766447488498796e-05, "loss": 0.0487, "step": 14280 }, { "grad_norm": 0.544889509677887, "learning_rate": 9.765947783827139e-05, "loss": 0.0537, "step": 14290 }, { "grad_norm": 0.5351892709732056, "learning_rate": 9.765447557958599e-05, "loss": 0.0651, "step": 14300 }, { "grad_norm": 0.5126252770423889, "learning_rate": 9.764946810947879e-05, "loss": 0.0531, "step": 14310 }, { "grad_norm": 0.3279014229774475, "learning_rate": 9.764445542849738e-05, "loss": 0.0622, "step": 14320 }, { "grad_norm": 0.4189080595970154, "learning_rate": 9.763943753718998e-05, "loss": 0.0527, "step": 14330 }, { "grad_norm": 0.43340885639190674, "learning_rate": 9.76344144361053e-05, "loss": 0.0609, "step": 14340 }, { "grad_norm": 0.6282625198364258, "learning_rate": 9.762938612579269e-05, "loss": 0.0628, "step": 14350 }, { "grad_norm": 0.4176422953605652, "learning_rate": 9.762435260680202e-05, "loss": 0.0541, "step": 14360 }, { "grad_norm": 0.5403051972389221, "learning_rate": 9.761931387968373e-05, "loss": 0.061, "step": 14370 }, { "grad_norm": 0.35760414600372314, "learning_rate": 9.76142699449889e-05, "loss": 0.0582, "step": 14380 }, { "grad_norm": 0.35673749446868896, "learning_rate": 9.760922080326908e-05, "loss": 0.0665, "step": 14390 }, { "grad_norm": 0.4715428948402405, "learning_rate": 9.760416645507644e-05, "loss": 0.0553, "step": 14400 }, { "grad_norm": 0.4630015194416046, "learning_rate": 9.759910690096375e-05, "loss": 0.0441, "step": 14410 }, { "grad_norm": 0.6096611618995667, "learning_rate": 9.759404214148429e-05, "loss": 0.052, "step": 14420 }, { "grad_norm": 0.426929235458374, "learning_rate": 9.758897217719191e-05, "loss": 0.0539, "step": 14430 }, { "grad_norm": 0.4467141330242157, "learning_rate": 9.758389700864113e-05, "loss": 0.0484, "step": 14440 }, { "grad_norm": 0.5006381869316101, "learning_rate": 9.757881663638688e-05, "loss": 0.0575, "step": 14450 }, { "grad_norm": 0.8300326466560364, "learning_rate": 9.757373106098478e-05, "loss": 0.0605, "step": 14460 }, { "grad_norm": 0.5521590113639832, "learning_rate": 9.756864028299097e-05, "loss": 0.0644, "step": 14470 }, { "grad_norm": 0.5573165416717529, "learning_rate": 9.75635443029622e-05, "loss": 0.0509, "step": 14480 }, { "grad_norm": 0.4846029579639435, "learning_rate": 9.755844312145572e-05, "loss": 0.0643, "step": 14490 }, { "grad_norm": 0.5558650493621826, "learning_rate": 9.755333673902941e-05, "loss": 0.0635, "step": 14500 }, { "grad_norm": 0.4227449893951416, "learning_rate": 9.75482251562417e-05, "loss": 0.0599, "step": 14510 }, { "grad_norm": 0.45287448167800903, "learning_rate": 9.754310837365155e-05, "loss": 0.0528, "step": 14520 }, { "grad_norm": 0.4667123258113861, "learning_rate": 9.753798639181856e-05, "loss": 0.0751, "step": 14530 }, { "grad_norm": 0.4292019307613373, "learning_rate": 9.753285921130286e-05, "loss": 0.0602, "step": 14540 }, { "grad_norm": 0.3766583204269409, "learning_rate": 9.752772683266512e-05, "loss": 0.0563, "step": 14550 }, { "grad_norm": 0.531678318977356, "learning_rate": 9.752258925646665e-05, "loss": 0.0561, "step": 14560 }, { "grad_norm": 0.5158089399337769, "learning_rate": 9.751744648326926e-05, "loss": 0.0519, "step": 14570 }, { "grad_norm": 0.7689098715782166, "learning_rate": 9.751229851363536e-05, "loss": 0.0688, "step": 14580 }, { "grad_norm": 0.47684815526008606, "learning_rate": 9.750714534812793e-05, "loss": 0.0638, "step": 14590 }, { "grad_norm": 0.3733442425727844, "learning_rate": 9.750198698731053e-05, "loss": 0.0508, "step": 14600 }, { "grad_norm": 0.6939488053321838, "learning_rate": 9.749682343174722e-05, "loss": 0.0543, "step": 14610 }, { "grad_norm": 0.3350786864757538, "learning_rate": 9.749165468200272e-05, "loss": 0.0511, "step": 14620 }, { "grad_norm": 0.5218477249145508, "learning_rate": 9.748648073864229e-05, "loss": 0.052, "step": 14630 }, { "grad_norm": 0.596549928188324, "learning_rate": 9.748130160223168e-05, "loss": 0.0554, "step": 14640 }, { "grad_norm": 0.4391253888607025, "learning_rate": 9.747611727333734e-05, "loss": 0.0574, "step": 14650 }, { "grad_norm": 0.32476329803466797, "learning_rate": 9.74709277525262e-05, "loss": 0.0461, "step": 14660 }, { "grad_norm": 0.6511375308036804, "learning_rate": 9.746573304036576e-05, "loss": 0.0634, "step": 14670 }, { "grad_norm": 0.517327606678009, "learning_rate": 9.746053313742412e-05, "loss": 0.049, "step": 14680 }, { "grad_norm": 0.4776824414730072, "learning_rate": 9.745532804426994e-05, "loss": 0.0493, "step": 14690 }, { "grad_norm": 0.6030625104904175, "learning_rate": 9.745011776147242e-05, "loss": 0.0688, "step": 14700 }, { "grad_norm": 0.5312092304229736, "learning_rate": 9.744490228960138e-05, "loss": 0.0661, "step": 14710 }, { "grad_norm": 0.5202786326408386, "learning_rate": 9.743968162922713e-05, "loss": 0.06, "step": 14720 }, { "grad_norm": 0.37363097071647644, "learning_rate": 9.743445578092064e-05, "loss": 0.0557, "step": 14730 }, { "grad_norm": 0.3449888825416565, "learning_rate": 9.742922474525338e-05, "loss": 0.0695, "step": 14740 }, { "grad_norm": 0.47505277395248413, "learning_rate": 9.742398852279741e-05, "loss": 0.0702, "step": 14750 }, { "grad_norm": 0.536616325378418, "learning_rate": 9.741874711412535e-05, "loss": 0.0639, "step": 14760 }, { "grad_norm": 0.5065904259681702, "learning_rate": 9.741350051981042e-05, "loss": 0.0608, "step": 14770 }, { "grad_norm": 0.5024431943893433, "learning_rate": 9.740824874042633e-05, "loss": 0.0543, "step": 14780 }, { "grad_norm": 0.4811791181564331, "learning_rate": 9.740299177654746e-05, "loss": 0.0489, "step": 14790 }, { "grad_norm": 0.6231053471565247, "learning_rate": 9.739772962874867e-05, "loss": 0.0616, "step": 14800 }, { "grad_norm": 0.4790512025356293, "learning_rate": 9.739246229760541e-05, "loss": 0.0563, "step": 14810 }, { "grad_norm": 0.40083566308021545, "learning_rate": 9.738718978369376e-05, "loss": 0.0549, "step": 14820 }, { "grad_norm": 0.4676298201084137, "learning_rate": 9.738191208759025e-05, "loss": 0.0688, "step": 14830 }, { "grad_norm": 0.520118236541748, "learning_rate": 9.73766292098721e-05, "loss": 0.0712, "step": 14840 }, { "grad_norm": 0.4446108341217041, "learning_rate": 9.737134115111699e-05, "loss": 0.0685, "step": 14850 }, { "grad_norm": 0.39124560356140137, "learning_rate": 9.736604791190323e-05, "loss": 0.0505, "step": 14860 }, { "grad_norm": 0.7226113677024841, "learning_rate": 9.73607494928097e-05, "loss": 0.0554, "step": 14870 }, { "grad_norm": 0.5850375294685364, "learning_rate": 9.735544589441581e-05, "loss": 0.0624, "step": 14880 }, { "grad_norm": 0.3619387149810791, "learning_rate": 9.735013711730154e-05, "loss": 0.0543, "step": 14890 }, { "grad_norm": 0.501615047454834, "learning_rate": 9.734482316204747e-05, "loss": 0.0526, "step": 14900 }, { "grad_norm": 0.5216788053512573, "learning_rate": 9.733950402923473e-05, "loss": 0.0595, "step": 14910 }, { "grad_norm": 0.5619503855705261, "learning_rate": 9.7334179719445e-05, "loss": 0.0573, "step": 14920 }, { "grad_norm": 0.5116930603981018, "learning_rate": 9.732885023326053e-05, "loss": 0.0524, "step": 14930 }, { "grad_norm": 0.3908592164516449, "learning_rate": 9.732351557126418e-05, "loss": 0.0578, "step": 14940 }, { "grad_norm": 0.5567407608032227, "learning_rate": 9.731817573403929e-05, "loss": 0.0587, "step": 14950 }, { "grad_norm": 0.41993242502212524, "learning_rate": 9.731283072216985e-05, "loss": 0.0517, "step": 14960 }, { "grad_norm": 0.5039221048355103, "learning_rate": 9.730748053624039e-05, "loss": 0.0537, "step": 14970 }, { "grad_norm": 0.4185601770877838, "learning_rate": 9.730212517683598e-05, "loss": 0.0512, "step": 14980 }, { "grad_norm": 0.46147316694259644, "learning_rate": 9.729676464454228e-05, "loss": 0.054, "step": 14990 }, { "grad_norm": 0.5322306752204895, "learning_rate": 9.72913989399455e-05, "loss": 0.0586, "step": 15000 }, { "grad_norm": 0.479797899723053, "learning_rate": 9.728602806363242e-05, "loss": 0.0591, "step": 15010 }, { "grad_norm": 0.5874482989311218, "learning_rate": 9.728065201619043e-05, "loss": 0.0527, "step": 15020 }, { "grad_norm": 0.5617862343788147, "learning_rate": 9.727527079820742e-05, "loss": 0.0582, "step": 15030 }, { "grad_norm": 0.3540712594985962, "learning_rate": 9.726988441027186e-05, "loss": 0.0637, "step": 15040 }, { "grad_norm": 0.5656391382217407, "learning_rate": 9.726449285297281e-05, "loss": 0.0646, "step": 15050 }, { "grad_norm": 0.49865251779556274, "learning_rate": 9.72590961268999e-05, "loss": 0.0653, "step": 15060 }, { "grad_norm": 0.45605137944221497, "learning_rate": 9.725369423264328e-05, "loss": 0.0616, "step": 15070 }, { "grad_norm": 0.543462336063385, "learning_rate": 9.72482871707937e-05, "loss": 0.0718, "step": 15080 }, { "grad_norm": 0.48832401633262634, "learning_rate": 9.724287494194247e-05, "loss": 0.0622, "step": 15090 }, { "grad_norm": 0.4055849313735962, "learning_rate": 9.723745754668147e-05, "loss": 0.0507, "step": 15100 }, { "grad_norm": 0.4635656774044037, "learning_rate": 9.723203498560313e-05, "loss": 0.0555, "step": 15110 }, { "grad_norm": 0.4668550193309784, "learning_rate": 9.722660725930046e-05, "loss": 0.0506, "step": 15120 }, { "grad_norm": 0.5507598519325256, "learning_rate": 9.722117436836702e-05, "loss": 0.0498, "step": 15130 }, { "grad_norm": 0.3510218560695648, "learning_rate": 9.721573631339696e-05, "loss": 0.0414, "step": 15140 }, { "grad_norm": 0.5176173448562622, "learning_rate": 9.721029309498494e-05, "loss": 0.0569, "step": 15150 }, { "grad_norm": 0.4883304536342621, "learning_rate": 9.720484471372627e-05, "loss": 0.0534, "step": 15160 }, { "grad_norm": 0.3811901807785034, "learning_rate": 9.719939117021673e-05, "loss": 0.0608, "step": 15170 }, { "grad_norm": 0.47683608531951904, "learning_rate": 9.719393246505275e-05, "loss": 0.056, "step": 15180 }, { "grad_norm": 0.3774224817752838, "learning_rate": 9.718846859883128e-05, "loss": 0.0515, "step": 15190 }, { "grad_norm": 0.5665720105171204, "learning_rate": 9.718299957214982e-05, "loss": 0.0639, "step": 15200 }, { "grad_norm": 0.29998502135276794, "learning_rate": 9.717752538560646e-05, "loss": 0.0535, "step": 15210 }, { "grad_norm": 0.5332859754562378, "learning_rate": 9.717204603979986e-05, "loss": 0.0604, "step": 15220 }, { "grad_norm": 0.5018644332885742, "learning_rate": 9.716656153532922e-05, "loss": 0.0485, "step": 15230 }, { "grad_norm": 0.4498130977153778, "learning_rate": 9.716107187279434e-05, "loss": 0.0514, "step": 15240 }, { "grad_norm": 0.5545697212219238, "learning_rate": 9.715557705279555e-05, "loss": 0.0662, "step": 15250 }, { "grad_norm": 0.40118980407714844, "learning_rate": 9.715007707593372e-05, "loss": 0.0526, "step": 15260 }, { "grad_norm": 0.8915557861328125, "learning_rate": 9.714457194281036e-05, "loss": 0.0579, "step": 15270 }, { "grad_norm": 0.401775985956192, "learning_rate": 9.713906165402751e-05, "loss": 0.0527, "step": 15280 }, { "grad_norm": 0.4405940771102905, "learning_rate": 9.713354621018774e-05, "loss": 0.0735, "step": 15290 }, { "grad_norm": 0.704228401184082, "learning_rate": 9.712802561189422e-05, "loss": 0.0798, "step": 15300 }, { "grad_norm": 0.37943053245544434, "learning_rate": 9.712249985975069e-05, "loss": 0.0718, "step": 15310 }, { "grad_norm": 0.4971451461315155, "learning_rate": 9.71169689543614e-05, "loss": 0.0444, "step": 15320 }, { "grad_norm": 0.5037989616394043, "learning_rate": 9.711143289633123e-05, "loss": 0.049, "step": 15330 }, { "grad_norm": 0.43082529306411743, "learning_rate": 9.710589168626561e-05, "loss": 0.0565, "step": 15340 }, { "grad_norm": 0.835048496723175, "learning_rate": 9.710034532477048e-05, "loss": 0.0797, "step": 15350 }, { "grad_norm": 0.5009061098098755, "learning_rate": 9.709479381245239e-05, "loss": 0.0539, "step": 15360 }, { "grad_norm": 0.4911727011203766, "learning_rate": 9.708923714991847e-05, "loss": 0.0564, "step": 15370 }, { "grad_norm": 0.45911142230033875, "learning_rate": 9.708367533777638e-05, "loss": 0.0554, "step": 15380 }, { "grad_norm": 0.3721823990345001, "learning_rate": 9.707810837663431e-05, "loss": 0.0509, "step": 15390 }, { "grad_norm": 0.45021945238113403, "learning_rate": 9.707253626710113e-05, "loss": 0.0552, "step": 15400 }, { "grad_norm": 0.37468209862709045, "learning_rate": 9.706695900978613e-05, "loss": 0.0452, "step": 15410 }, { "grad_norm": 0.5536936521530151, "learning_rate": 9.706137660529926e-05, "loss": 0.0562, "step": 15420 }, { "grad_norm": 0.7142711877822876, "learning_rate": 9.705578905425101e-05, "loss": 0.0702, "step": 15430 }, { "grad_norm": 0.4470938444137573, "learning_rate": 9.705019635725241e-05, "loss": 0.0474, "step": 15440 }, { "grad_norm": 0.3698354661464691, "learning_rate": 9.704459851491508e-05, "loss": 0.0588, "step": 15450 }, { "grad_norm": 0.44482243061065674, "learning_rate": 9.703899552785118e-05, "loss": 0.0526, "step": 15460 }, { "grad_norm": 0.47786983847618103, "learning_rate": 9.703338739667346e-05, "loss": 0.0505, "step": 15470 }, { "grad_norm": 0.4967902600765228, "learning_rate": 9.70277741219952e-05, "loss": 0.0569, "step": 15480 }, { "grad_norm": 0.43416109681129456, "learning_rate": 9.702215570443027e-05, "loss": 0.0551, "step": 15490 }, { "grad_norm": 0.37822964787483215, "learning_rate": 9.701653214459309e-05, "loss": 0.0563, "step": 15500 }, { "grad_norm": 0.46175163984298706, "learning_rate": 9.701090344309865e-05, "loss": 0.055, "step": 15510 }, { "grad_norm": 0.47372400760650635, "learning_rate": 9.700526960056247e-05, "loss": 0.047, "step": 15520 }, { "grad_norm": 0.420662522315979, "learning_rate": 9.699963061760068e-05, "loss": 0.0486, "step": 15530 }, { "grad_norm": 0.4035399258136749, "learning_rate": 9.699398649482997e-05, "loss": 0.0598, "step": 15540 }, { "grad_norm": 0.341560959815979, "learning_rate": 9.698833723286753e-05, "loss": 0.0505, "step": 15550 }, { "grad_norm": 0.5441876649856567, "learning_rate": 9.698268283233118e-05, "loss": 0.0628, "step": 15560 }, { "grad_norm": 0.48995235562324524, "learning_rate": 9.697702329383929e-05, "loss": 0.0568, "step": 15570 }, { "grad_norm": 0.43653541803359985, "learning_rate": 9.697135861801074e-05, "loss": 0.0651, "step": 15580 }, { "grad_norm": 0.45626112818717957, "learning_rate": 9.696568880546505e-05, "loss": 0.0481, "step": 15590 }, { "grad_norm": 0.421928733587265, "learning_rate": 9.696001385682223e-05, "loss": 0.0519, "step": 15600 }, { "grad_norm": 0.35606375336647034, "learning_rate": 9.695433377270291e-05, "loss": 0.0555, "step": 15610 }, { "grad_norm": 0.37143781781196594, "learning_rate": 9.694864855372824e-05, "loss": 0.0459, "step": 15620 }, { "grad_norm": 0.5387545228004456, "learning_rate": 9.694295820051995e-05, "loss": 0.0543, "step": 15630 }, { "grad_norm": 0.5677148103713989, "learning_rate": 9.693726271370032e-05, "loss": 0.0531, "step": 15640 }, { "grad_norm": 0.4865818917751312, "learning_rate": 9.693156209389221e-05, "loss": 0.051, "step": 15650 }, { "grad_norm": 0.5513033270835876, "learning_rate": 9.692585634171905e-05, "loss": 0.0627, "step": 15660 }, { "grad_norm": 0.5604664087295532, "learning_rate": 9.692014545780476e-05, "loss": 0.0542, "step": 15670 }, { "grad_norm": 0.3935316801071167, "learning_rate": 9.691442944277393e-05, "loss": 0.0419, "step": 15680 }, { "grad_norm": 0.33651256561279297, "learning_rate": 9.690870829725162e-05, "loss": 0.0489, "step": 15690 }, { "grad_norm": 0.5309171676635742, "learning_rate": 9.69029820218635e-05, "loss": 0.0536, "step": 15700 }, { "grad_norm": 0.5105275511741638, "learning_rate": 9.689725061723579e-05, "loss": 0.0507, "step": 15710 }, { "grad_norm": 0.5775290727615356, "learning_rate": 9.689151408399527e-05, "loss": 0.046, "step": 15720 }, { "grad_norm": 0.545551061630249, "learning_rate": 9.688577242276924e-05, "loss": 0.065, "step": 15730 }, { "grad_norm": 0.48393872380256653, "learning_rate": 9.688002563418566e-05, "loss": 0.0544, "step": 15740 }, { "grad_norm": 0.46930456161499023, "learning_rate": 9.687427371887293e-05, "loss": 0.0629, "step": 15750 }, { "grad_norm": 0.4699174761772156, "learning_rate": 9.686851667746012e-05, "loss": 0.0574, "step": 15760 }, { "grad_norm": 0.5330303311347961, "learning_rate": 9.686275451057677e-05, "loss": 0.0557, "step": 15770 }, { "grad_norm": 0.5249598622322083, "learning_rate": 9.685698721885308e-05, "loss": 0.0523, "step": 15780 }, { "grad_norm": 0.5626161098480225, "learning_rate": 9.68512148029197e-05, "loss": 0.0652, "step": 15790 }, { "grad_norm": 0.5140831470489502, "learning_rate": 9.684543726340791e-05, "loss": 0.0674, "step": 15800 }, { "grad_norm": 0.5341334939002991, "learning_rate": 9.683965460094952e-05, "loss": 0.0605, "step": 15810 }, { "grad_norm": 0.4353168308734894, "learning_rate": 9.683386681617694e-05, "loss": 0.0591, "step": 15820 }, { "grad_norm": 0.5060287714004517, "learning_rate": 9.68280739097231e-05, "loss": 0.0531, "step": 15830 }, { "grad_norm": 0.4353206753730774, "learning_rate": 9.682227588222148e-05, "loss": 0.0579, "step": 15840 }, { "grad_norm": 0.5131102800369263, "learning_rate": 9.681647273430618e-05, "loss": 0.0592, "step": 15850 }, { "grad_norm": 0.47092702984809875, "learning_rate": 9.681066446661182e-05, "loss": 0.0538, "step": 15860 }, { "grad_norm": 0.5278776288032532, "learning_rate": 9.680485107977357e-05, "loss": 0.0603, "step": 15870 }, { "grad_norm": 0.36701929569244385, "learning_rate": 9.679903257442716e-05, "loss": 0.0584, "step": 15880 }, { "grad_norm": 0.5531419515609741, "learning_rate": 9.679320895120891e-05, "loss": 0.06, "step": 15890 }, { "grad_norm": 0.7618100047111511, "learning_rate": 9.67873802107557e-05, "loss": 0.0582, "step": 15900 }, { "grad_norm": 0.5380387902259827, "learning_rate": 9.67815463537049e-05, "loss": 0.0551, "step": 15910 }, { "grad_norm": 0.5276015400886536, "learning_rate": 9.677570738069457e-05, "loss": 0.0529, "step": 15920 }, { "grad_norm": 0.45077425241470337, "learning_rate": 9.676986329236318e-05, "loss": 0.0538, "step": 15930 }, { "grad_norm": 0.43780291080474854, "learning_rate": 9.676401408934987e-05, "loss": 0.0496, "step": 15940 }, { "grad_norm": 0.4776512682437897, "learning_rate": 9.675815977229428e-05, "loss": 0.05, "step": 15950 }, { "grad_norm": 0.49525436758995056, "learning_rate": 9.675230034183664e-05, "loss": 0.0449, "step": 15960 }, { "grad_norm": 0.5709330439567566, "learning_rate": 9.674643579861773e-05, "loss": 0.0634, "step": 15970 }, { "grad_norm": 0.6312949061393738, "learning_rate": 9.674056614327886e-05, "loss": 0.0556, "step": 15980 }, { "grad_norm": 0.4562731385231018, "learning_rate": 9.673469137646198e-05, "loss": 0.0572, "step": 15990 }, { "grad_norm": 0.5136928558349609, "learning_rate": 9.67288114988095e-05, "loss": 0.0525, "step": 16000 }, { "grad_norm": 0.5096824765205383, "learning_rate": 9.672292651096447e-05, "loss": 0.0498, "step": 16010 }, { "grad_norm": 0.35645592212677, "learning_rate": 9.671703641357042e-05, "loss": 0.055, "step": 16020 }, { "grad_norm": 0.3807695508003235, "learning_rate": 9.67111412072715e-05, "loss": 0.0535, "step": 16030 }, { "grad_norm": 0.7834240794181824, "learning_rate": 9.670524089271242e-05, "loss": 0.0564, "step": 16040 }, { "grad_norm": 0.5337350964546204, "learning_rate": 9.669933547053842e-05, "loss": 0.0493, "step": 16050 }, { "grad_norm": 0.4865390956401825, "learning_rate": 9.669342494139531e-05, "loss": 0.0526, "step": 16060 }, { "grad_norm": 0.5216208100318909, "learning_rate": 9.668750930592943e-05, "loss": 0.0579, "step": 16070 }, { "grad_norm": 0.6496542096138, "learning_rate": 9.668158856478775e-05, "loss": 0.058, "step": 16080 }, { "grad_norm": 0.47118040919303894, "learning_rate": 9.66756627186177e-05, "loss": 0.0586, "step": 16090 }, { "grad_norm": 0.47749635577201843, "learning_rate": 9.666973176806737e-05, "loss": 0.0603, "step": 16100 }, { "grad_norm": 0.5039518475532532, "learning_rate": 9.666379571378534e-05, "loss": 0.0587, "step": 16110 }, { "grad_norm": 0.6871272325515747, "learning_rate": 9.665785455642076e-05, "loss": 0.0672, "step": 16120 }, { "grad_norm": 0.5961273312568665, "learning_rate": 9.665190829662337e-05, "loss": 0.0704, "step": 16130 }, { "grad_norm": 0.5132696628570557, "learning_rate": 9.664595693504342e-05, "loss": 0.0582, "step": 16140 }, { "grad_norm": 0.3944062292575836, "learning_rate": 9.664000047233175e-05, "loss": 0.0575, "step": 16150 }, { "grad_norm": 0.35127633810043335, "learning_rate": 9.663403890913976e-05, "loss": 0.0607, "step": 16160 }, { "grad_norm": 0.5441190600395203, "learning_rate": 9.662807224611938e-05, "loss": 0.0437, "step": 16170 }, { "grad_norm": 0.3751135766506195, "learning_rate": 9.662210048392311e-05, "loss": 0.0423, "step": 16180 }, { "grad_norm": 0.46570509672164917, "learning_rate": 9.661612362320405e-05, "loss": 0.0429, "step": 16190 }, { "grad_norm": 0.37866511940956116, "learning_rate": 9.661014166461579e-05, "loss": 0.0531, "step": 16200 }, { "grad_norm": 0.3297208547592163, "learning_rate": 9.66041546088125e-05, "loss": 0.0577, "step": 16210 }, { "grad_norm": 0.38912826776504517, "learning_rate": 9.659816245644895e-05, "loss": 0.0457, "step": 16220 }, { "grad_norm": 0.5150530934333801, "learning_rate": 9.65921652081804e-05, "loss": 0.0604, "step": 16230 }, { "grad_norm": 0.6522942781448364, "learning_rate": 9.658616286466271e-05, "loss": 0.0552, "step": 16240 }, { "grad_norm": 0.6989579200744629, "learning_rate": 9.65801554265523e-05, "loss": 0.0609, "step": 16250 }, { "grad_norm": 0.5285415649414062, "learning_rate": 9.657414289450612e-05, "loss": 0.0662, "step": 16260 }, { "grad_norm": 0.48878180980682373, "learning_rate": 9.656812526918171e-05, "loss": 0.0552, "step": 16270 }, { "grad_norm": 0.4778570532798767, "learning_rate": 9.656210255123712e-05, "loss": 0.0547, "step": 16280 }, { "grad_norm": 0.4281465709209442, "learning_rate": 9.6556074741331e-05, "loss": 0.0525, "step": 16290 }, { "grad_norm": 0.5934127569198608, "learning_rate": 9.655004184012256e-05, "loss": 0.0457, "step": 16300 }, { "grad_norm": 0.49648937582969666, "learning_rate": 9.654400384827152e-05, "loss": 0.0569, "step": 16310 }, { "grad_norm": 0.48003098368644714, "learning_rate": 9.653796076643818e-05, "loss": 0.0463, "step": 16320 }, { "grad_norm": 0.40976187586784363, "learning_rate": 9.653191259528344e-05, "loss": 0.0544, "step": 16330 }, { "grad_norm": 0.44529885053634644, "learning_rate": 9.65258593354687e-05, "loss": 0.0441, "step": 16340 }, { "grad_norm": 0.423195481300354, "learning_rate": 9.651980098765591e-05, "loss": 0.0487, "step": 16350 }, { "grad_norm": 0.41272062063217163, "learning_rate": 9.651373755250765e-05, "loss": 0.048, "step": 16360 }, { "grad_norm": 0.42574343085289, "learning_rate": 9.650766903068697e-05, "loss": 0.0527, "step": 16370 }, { "grad_norm": 0.39472490549087524, "learning_rate": 9.650159542285753e-05, "loss": 0.0446, "step": 16380 }, { "grad_norm": 0.46059706807136536, "learning_rate": 9.649551672968353e-05, "loss": 0.0597, "step": 16390 }, { "grad_norm": 0.5601722598075867, "learning_rate": 9.648943295182973e-05, "loss": 0.06, "step": 16400 }, { "grad_norm": 0.5062825083732605, "learning_rate": 9.648334408996144e-05, "loss": 0.0619, "step": 16410 }, { "grad_norm": 0.49004611372947693, "learning_rate": 9.647725014474452e-05, "loss": 0.0547, "step": 16420 }, { "grad_norm": 0.49669328331947327, "learning_rate": 9.64711511168454e-05, "loss": 0.0549, "step": 16430 }, { "grad_norm": 0.45992618799209595, "learning_rate": 9.646504700693108e-05, "loss": 0.0491, "step": 16440 }, { "grad_norm": 0.5546311140060425, "learning_rate": 9.645893781566907e-05, "loss": 0.0522, "step": 16450 }, { "grad_norm": 0.43990474939346313, "learning_rate": 9.645282354372744e-05, "loss": 0.0505, "step": 16460 }, { "grad_norm": 0.9790668487548828, "learning_rate": 9.644670419177491e-05, "loss": 0.0564, "step": 16470 }, { "grad_norm": 0.4959232807159424, "learning_rate": 9.644057976048062e-05, "loss": 0.0633, "step": 16480 }, { "grad_norm": 0.5027780532836914, "learning_rate": 9.643445025051435e-05, "loss": 0.0627, "step": 16490 }, { "grad_norm": 0.57069993019104, "learning_rate": 9.642831566254641e-05, "loss": 0.0528, "step": 16500 }, { "grad_norm": 0.5197017788887024, "learning_rate": 9.642217599724769e-05, "loss": 0.0583, "step": 16510 }, { "grad_norm": 0.46527206897735596, "learning_rate": 9.64160312552896e-05, "loss": 0.0623, "step": 16520 }, { "grad_norm": 0.5390945076942444, "learning_rate": 9.64098814373441e-05, "loss": 0.0555, "step": 16530 }, { "grad_norm": 0.45175352692604065, "learning_rate": 9.640372654408374e-05, "loss": 0.0473, "step": 16540 }, { "grad_norm": 0.5383028388023376, "learning_rate": 9.639756657618162e-05, "loss": 0.058, "step": 16550 }, { "grad_norm": 0.4901192784309387, "learning_rate": 9.639140153431138e-05, "loss": 0.0531, "step": 16560 }, { "grad_norm": 0.4651188850402832, "learning_rate": 9.638523141914721e-05, "loss": 0.0504, "step": 16570 }, { "grad_norm": 0.39090728759765625, "learning_rate": 9.637905623136388e-05, "loss": 0.0615, "step": 16580 }, { "grad_norm": 0.5892738699913025, "learning_rate": 9.637287597163669e-05, "loss": 0.0457, "step": 16590 }, { "grad_norm": 0.38806766271591187, "learning_rate": 9.63666906406415e-05, "loss": 0.0426, "step": 16600 }, { "grad_norm": 0.3629564344882965, "learning_rate": 9.636050023905473e-05, "loss": 0.0425, "step": 16610 }, { "grad_norm": 0.3250882625579834, "learning_rate": 9.635430476755336e-05, "loss": 0.0419, "step": 16620 }, { "grad_norm": 0.5316322445869446, "learning_rate": 9.63481042268149e-05, "loss": 0.0438, "step": 16630 }, { "grad_norm": 0.5790039896965027, "learning_rate": 9.634189861751745e-05, "loss": 0.0544, "step": 16640 }, { "grad_norm": 0.45966318249702454, "learning_rate": 9.633568794033967e-05, "loss": 0.0575, "step": 16650 }, { "grad_norm": 0.46197509765625, "learning_rate": 9.63294721959607e-05, "loss": 0.0502, "step": 16660 }, { "grad_norm": 0.45251190662384033, "learning_rate": 9.63232513850603e-05, "loss": 0.0672, "step": 16670 }, { "grad_norm": 0.4312516450881958, "learning_rate": 9.631702550831878e-05, "loss": 0.0584, "step": 16680 }, { "grad_norm": 0.535463809967041, "learning_rate": 9.631079456641698e-05, "loss": 0.0637, "step": 16690 }, { "grad_norm": 0.5152284502983093, "learning_rate": 9.630455856003632e-05, "loss": 0.0532, "step": 16700 }, { "grad_norm": 0.7263575196266174, "learning_rate": 9.629831748985876e-05, "loss": 0.0535, "step": 16710 }, { "grad_norm": 0.4399901330471039, "learning_rate": 9.629207135656679e-05, "loss": 0.0685, "step": 16720 }, { "grad_norm": 0.5808063745498657, "learning_rate": 9.628582016084353e-05, "loss": 0.0618, "step": 16730 }, { "grad_norm": 0.3814023435115814, "learning_rate": 9.627956390337254e-05, "loss": 0.0502, "step": 16740 }, { "grad_norm": 0.8519272804260254, "learning_rate": 9.627330258483802e-05, "loss": 0.0702, "step": 16750 }, { "grad_norm": 0.3446887135505676, "learning_rate": 9.62670362059247e-05, "loss": 0.0449, "step": 16760 }, { "grad_norm": 0.6060264110565186, "learning_rate": 9.626076476731786e-05, "loss": 0.0494, "step": 16770 }, { "grad_norm": 0.4762723445892334, "learning_rate": 9.625448826970336e-05, "loss": 0.0519, "step": 16780 }, { "grad_norm": 0.42434391379356384, "learning_rate": 9.624820671376755e-05, "loss": 0.0505, "step": 16790 }, { "grad_norm": 0.6665031909942627, "learning_rate": 9.62419201001974e-05, "loss": 0.0595, "step": 16800 }, { "grad_norm": 0.4490540027618408, "learning_rate": 9.623562842968037e-05, "loss": 0.0501, "step": 16810 }, { "grad_norm": 0.43281158804893494, "learning_rate": 9.622933170290454e-05, "loss": 0.0448, "step": 16820 }, { "grad_norm": 0.46976783871650696, "learning_rate": 9.622302992055849e-05, "loss": 0.0545, "step": 16830 }, { "grad_norm": 0.4607047736644745, "learning_rate": 9.62167230833314e-05, "loss": 0.0506, "step": 16840 }, { "grad_norm": 0.4036690890789032, "learning_rate": 9.621041119191295e-05, "loss": 0.0505, "step": 16850 }, { "grad_norm": 0.39909884333610535, "learning_rate": 9.620409424699342e-05, "loss": 0.0583, "step": 16860 }, { "grad_norm": 0.4867842495441437, "learning_rate": 9.619777224926359e-05, "loss": 0.0456, "step": 16870 }, { "grad_norm": 0.4921748638153076, "learning_rate": 9.619144519941485e-05, "loss": 0.0531, "step": 16880 }, { "grad_norm": 0.429684579372406, "learning_rate": 9.618511309813912e-05, "loss": 0.0522, "step": 16890 }, { "grad_norm": 0.3432295322418213, "learning_rate": 9.617877594612886e-05, "loss": 0.0456, "step": 16900 }, { "grad_norm": 0.530676543712616, "learning_rate": 9.617243374407707e-05, "loss": 0.0476, "step": 16910 }, { "grad_norm": 0.5147649049758911, "learning_rate": 9.616608649267736e-05, "loss": 0.0563, "step": 16920 }, { "grad_norm": 0.5046831965446472, "learning_rate": 9.615973419262385e-05, "loss": 0.0618, "step": 16930 }, { "grad_norm": 0.4814158082008362, "learning_rate": 9.615337684461119e-05, "loss": 0.05, "step": 16940 }, { "grad_norm": 0.49667319655418396, "learning_rate": 9.614701444933465e-05, "loss": 0.076, "step": 16950 }, { "grad_norm": 0.5207512378692627, "learning_rate": 9.614064700748997e-05, "loss": 0.0577, "step": 16960 }, { "grad_norm": 0.4731619954109192, "learning_rate": 9.613427451977352e-05, "loss": 0.0622, "step": 16970 }, { "grad_norm": 0.4266653060913086, "learning_rate": 9.612789698688216e-05, "loss": 0.0548, "step": 16980 }, { "grad_norm": 0.5706650614738464, "learning_rate": 9.612151440951334e-05, "loss": 0.0584, "step": 16990 }, { "grad_norm": 0.4770324230194092, "learning_rate": 9.611512678836506e-05, "loss": 0.0469, "step": 17000 }, { "grad_norm": 0.4916643798351288, "learning_rate": 9.610873412413584e-05, "loss": 0.0554, "step": 17010 }, { "grad_norm": 0.5390084385871887, "learning_rate": 9.610233641752476e-05, "loss": 0.0592, "step": 17020 }, { "grad_norm": 0.36651870608329773, "learning_rate": 9.609593366923151e-05, "loss": 0.0469, "step": 17030 }, { "grad_norm": 0.5225926637649536, "learning_rate": 9.608952587995625e-05, "loss": 0.0502, "step": 17040 }, { "grad_norm": 0.49819663166999817, "learning_rate": 9.608311305039972e-05, "loss": 0.0498, "step": 17050 }, { "grad_norm": 0.6823115348815918, "learning_rate": 9.607669518126326e-05, "loss": 0.0711, "step": 17060 }, { "grad_norm": 0.557523250579834, "learning_rate": 9.607027227324866e-05, "loss": 0.0502, "step": 17070 }, { "grad_norm": 0.4236869513988495, "learning_rate": 9.606384432705837e-05, "loss": 0.0467, "step": 17080 }, { "grad_norm": 0.5894594192504883, "learning_rate": 9.60574113433953e-05, "loss": 0.0522, "step": 17090 }, { "grad_norm": 0.542546808719635, "learning_rate": 9.6050973322963e-05, "loss": 0.0584, "step": 17100 }, { "grad_norm": 0.578495979309082, "learning_rate": 9.604453026646547e-05, "loss": 0.0664, "step": 17110 }, { "grad_norm": 0.49073025584220886, "learning_rate": 9.603808217460735e-05, "loss": 0.0541, "step": 17120 }, { "grad_norm": 0.4572761654853821, "learning_rate": 9.603162904809377e-05, "loss": 0.0549, "step": 17130 }, { "grad_norm": 0.3896098732948303, "learning_rate": 9.602517088763045e-05, "loss": 0.0425, "step": 17140 }, { "grad_norm": 0.5402583479881287, "learning_rate": 9.601870769392365e-05, "loss": 0.0564, "step": 17150 }, { "grad_norm": 0.5155171155929565, "learning_rate": 9.601223946768017e-05, "loss": 0.0555, "step": 17160 }, { "grad_norm": 0.5837932825088501, "learning_rate": 9.600576620960734e-05, "loss": 0.0546, "step": 17170 }, { "grad_norm": 0.4787892997264862, "learning_rate": 9.599928792041308e-05, "loss": 0.0521, "step": 17180 }, { "grad_norm": 0.5671133399009705, "learning_rate": 9.599280460080587e-05, "loss": 0.0557, "step": 17190 }, { "grad_norm": 0.48857638239860535, "learning_rate": 9.59863162514947e-05, "loss": 0.0495, "step": 17200 }, { "grad_norm": 0.6331575512886047, "learning_rate": 9.597982287318911e-05, "loss": 0.0492, "step": 17210 }, { "grad_norm": 0.4579910337924957, "learning_rate": 9.597332446659923e-05, "loss": 0.0596, "step": 17220 }, { "grad_norm": 0.45088639855384827, "learning_rate": 9.59668210324357e-05, "loss": 0.0546, "step": 17230 }, { "grad_norm": 0.6643266081809998, "learning_rate": 9.596031257140974e-05, "loss": 0.0517, "step": 17240 }, { "grad_norm": 0.6406622529029846, "learning_rate": 9.59537990842331e-05, "loss": 0.0685, "step": 17250 }, { "grad_norm": 0.4264448583126068, "learning_rate": 9.594728057161806e-05, "loss": 0.053, "step": 17260 }, { "grad_norm": 0.3780568540096283, "learning_rate": 9.594075703427752e-05, "loss": 0.0509, "step": 17270 }, { "grad_norm": 0.40138235688209534, "learning_rate": 9.593422847292486e-05, "loss": 0.0493, "step": 17280 }, { "grad_norm": 0.314018577337265, "learning_rate": 9.592769488827402e-05, "loss": 0.0574, "step": 17290 }, { "grad_norm": 0.4737204909324646, "learning_rate": 9.592115628103952e-05, "loss": 0.0476, "step": 17300 }, { "grad_norm": 0.5289526581764221, "learning_rate": 9.591461265193643e-05, "loss": 0.0474, "step": 17310 }, { "grad_norm": 0.4931846857070923, "learning_rate": 9.590806400168032e-05, "loss": 0.0573, "step": 17320 }, { "grad_norm": 0.49984514713287354, "learning_rate": 9.590151033098735e-05, "loss": 0.0516, "step": 17330 }, { "grad_norm": 0.3393009901046753, "learning_rate": 9.589495164057423e-05, "loss": 0.0476, "step": 17340 }, { "grad_norm": 0.34499090909957886, "learning_rate": 9.58883879311582e-05, "loss": 0.0432, "step": 17350 }, { "grad_norm": 0.4852612614631653, "learning_rate": 9.588181920345705e-05, "loss": 0.0602, "step": 17360 }, { "grad_norm": 0.4542110562324524, "learning_rate": 9.587524545818913e-05, "loss": 0.0521, "step": 17370 }, { "grad_norm": 0.5624490976333618, "learning_rate": 9.586866669607335e-05, "loss": 0.0584, "step": 17380 }, { "grad_norm": 0.4986457824707031, "learning_rate": 9.586208291782915e-05, "loss": 0.0521, "step": 17390 }, { "grad_norm": 0.42315080761909485, "learning_rate": 9.58554941241765e-05, "loss": 0.054, "step": 17400 }, { "grad_norm": 0.5169162750244141, "learning_rate": 9.584890031583596e-05, "loss": 0.0462, "step": 17410 }, { "grad_norm": 0.4815276563167572, "learning_rate": 9.584230149352861e-05, "loss": 0.0538, "step": 17420 }, { "grad_norm": 0.41394758224487305, "learning_rate": 9.58356976579761e-05, "loss": 0.0439, "step": 17430 }, { "grad_norm": 0.4836953282356262, "learning_rate": 9.58290888099006e-05, "loss": 0.0527, "step": 17440 }, { "grad_norm": 0.49499139189720154, "learning_rate": 9.582247495002486e-05, "loss": 0.0516, "step": 17450 }, { "grad_norm": 0.4318191111087799, "learning_rate": 9.581585607907214e-05, "loss": 0.0517, "step": 17460 }, { "grad_norm": 0.3545446991920471, "learning_rate": 9.580923219776628e-05, "loss": 0.0459, "step": 17470 }, { "grad_norm": 0.38144078850746155, "learning_rate": 9.580260330683167e-05, "loss": 0.0552, "step": 17480 }, { "grad_norm": 0.3762991428375244, "learning_rate": 9.579596940699322e-05, "loss": 0.0436, "step": 17490 }, { "grad_norm": 0.48859840631484985, "learning_rate": 9.578933049897643e-05, "loss": 0.0451, "step": 17500 }, { "grad_norm": 0.4659026861190796, "learning_rate": 9.578268658350728e-05, "loss": 0.0513, "step": 17510 }, { "grad_norm": 0.5584519505500793, "learning_rate": 9.577603766131235e-05, "loss": 0.0594, "step": 17520 }, { "grad_norm": 0.49278050661087036, "learning_rate": 9.576938373311878e-05, "loss": 0.0592, "step": 17530 }, { "grad_norm": 0.554408848285675, "learning_rate": 9.576272479965421e-05, "loss": 0.0572, "step": 17540 }, { "grad_norm": 0.4264205992221832, "learning_rate": 9.575606086164687e-05, "loss": 0.0499, "step": 17550 }, { "grad_norm": 0.5834912657737732, "learning_rate": 9.57493919198255e-05, "loss": 0.0667, "step": 17560 }, { "grad_norm": 0.532640814781189, "learning_rate": 9.57427179749194e-05, "loss": 0.0571, "step": 17570 }, { "grad_norm": 0.41860198974609375, "learning_rate": 9.573603902765846e-05, "loss": 0.0574, "step": 17580 }, { "grad_norm": 0.5015078783035278, "learning_rate": 9.572935507877304e-05, "loss": 0.0619, "step": 17590 }, { "grad_norm": 0.4586377441883087, "learning_rate": 9.57226661289941e-05, "loss": 0.0635, "step": 17600 }, { "grad_norm": 0.5120628476142883, "learning_rate": 9.571597217905315e-05, "loss": 0.0626, "step": 17610 }, { "grad_norm": 0.4035997986793518, "learning_rate": 9.57092732296822e-05, "loss": 0.0582, "step": 17620 }, { "grad_norm": 0.43431001901626587, "learning_rate": 9.570256928161385e-05, "loss": 0.047, "step": 17630 }, { "grad_norm": 0.37977689504623413, "learning_rate": 9.569586033558126e-05, "loss": 0.0449, "step": 17640 }, { "grad_norm": 0.4435212314128876, "learning_rate": 9.568914639231807e-05, "loss": 0.048, "step": 17650 }, { "grad_norm": 0.5325537919998169, "learning_rate": 9.568242745255852e-05, "loss": 0.0572, "step": 17660 }, { "grad_norm": 1.028307318687439, "learning_rate": 9.567570351703739e-05, "loss": 0.0579, "step": 17670 }, { "grad_norm": 0.4939020276069641, "learning_rate": 9.566897458649001e-05, "loss": 0.0514, "step": 17680 }, { "grad_norm": 0.429159015417099, "learning_rate": 9.566224066165221e-05, "loss": 0.0521, "step": 17690 }, { "grad_norm": 0.3830351233482361, "learning_rate": 9.565550174326043e-05, "loss": 0.056, "step": 17700 }, { "grad_norm": 0.560553252696991, "learning_rate": 9.564875783205162e-05, "loss": 0.0571, "step": 17710 }, { "grad_norm": 0.44282278418540955, "learning_rate": 9.564200892876328e-05, "loss": 0.0581, "step": 17720 }, { "grad_norm": 0.6810593008995056, "learning_rate": 9.563525503413348e-05, "loss": 0.0506, "step": 17730 }, { "grad_norm": 0.4169962704181671, "learning_rate": 9.562849614890079e-05, "loss": 0.0533, "step": 17740 }, { "grad_norm": 0.4596188962459564, "learning_rate": 9.562173227380436e-05, "loss": 0.0695, "step": 17750 }, { "grad_norm": 0.5048893094062805, "learning_rate": 9.561496340958389e-05, "loss": 0.0539, "step": 17760 }, { "grad_norm": 0.37306344509124756, "learning_rate": 9.560818955697959e-05, "loss": 0.0432, "step": 17770 }, { "grad_norm": 0.40725988149642944, "learning_rate": 9.560141071673228e-05, "loss": 0.0556, "step": 17780 }, { "grad_norm": 0.45954322814941406, "learning_rate": 9.559462688958323e-05, "loss": 0.0465, "step": 17790 }, { "grad_norm": 0.37783974409103394, "learning_rate": 9.558783807627434e-05, "loss": 0.0413, "step": 17800 }, { "grad_norm": 0.5726175904273987, "learning_rate": 9.558104427754801e-05, "loss": 0.0545, "step": 17810 }, { "grad_norm": 0.5572339296340942, "learning_rate": 9.557424549414722e-05, "loss": 0.0534, "step": 17820 }, { "grad_norm": 0.3267563283443451, "learning_rate": 9.556744172681546e-05, "loss": 0.0484, "step": 17830 }, { "grad_norm": 0.41638508439064026, "learning_rate": 9.556063297629677e-05, "loss": 0.0439, "step": 17840 }, { "grad_norm": 0.34977155923843384, "learning_rate": 9.555381924333578e-05, "loss": 0.0441, "step": 17850 }, { "grad_norm": 0.56178218126297, "learning_rate": 9.554700052867758e-05, "loss": 0.0467, "step": 17860 }, { "grad_norm": 0.4815523028373718, "learning_rate": 9.554017683306789e-05, "loss": 0.0468, "step": 17870 }, { "grad_norm": 0.3461894690990448, "learning_rate": 9.553334815725294e-05, "loss": 0.0402, "step": 17880 }, { "grad_norm": 0.6812952756881714, "learning_rate": 9.552651450197949e-05, "loss": 0.0476, "step": 17890 }, { "grad_norm": 0.5933959484100342, "learning_rate": 9.551967586799486e-05, "loss": 0.0455, "step": 17900 }, { "grad_norm": 0.5403727889060974, "learning_rate": 9.551283225604692e-05, "loss": 0.0619, "step": 17910 }, { "grad_norm": 0.4359886050224304, "learning_rate": 9.550598366688406e-05, "loss": 0.0472, "step": 17920 }, { "grad_norm": 0.5500603914260864, "learning_rate": 9.549913010125526e-05, "loss": 0.0507, "step": 17930 }, { "grad_norm": 0.4292227327823639, "learning_rate": 9.549227155990999e-05, "loss": 0.0482, "step": 17940 }, { "grad_norm": 0.41011855006217957, "learning_rate": 9.548540804359828e-05, "loss": 0.0677, "step": 17950 }, { "grad_norm": 0.4694884121417999, "learning_rate": 9.547853955307077e-05, "loss": 0.0553, "step": 17960 }, { "grad_norm": 1.0155253410339355, "learning_rate": 9.547166608907853e-05, "loss": 0.0452, "step": 17970 }, { "grad_norm": 0.49897500872612, "learning_rate": 9.546478765237326e-05, "loss": 0.064, "step": 17980 }, { "grad_norm": 0.4937687814235687, "learning_rate": 9.545790424370715e-05, "loss": 0.0511, "step": 17990 }, { "grad_norm": 0.4335954785346985, "learning_rate": 9.5451015863833e-05, "loss": 0.0477, "step": 18000 }, { "grad_norm": 0.4556643068790436, "learning_rate": 9.544412251350408e-05, "loss": 0.0603, "step": 18010 }, { "grad_norm": 0.4161335527896881, "learning_rate": 9.543722419347422e-05, "loss": 0.0532, "step": 18020 }, { "grad_norm": 0.43512558937072754, "learning_rate": 9.543032090449788e-05, "loss": 0.0563, "step": 18030 }, { "grad_norm": 0.4103081226348877, "learning_rate": 9.542341264732992e-05, "loss": 0.0546, "step": 18040 }, { "grad_norm": 0.4656267762184143, "learning_rate": 9.541649942272585e-05, "loss": 0.0526, "step": 18050 }, { "grad_norm": 0.8354902267456055, "learning_rate": 9.54095812314417e-05, "loss": 0.054, "step": 18060 }, { "grad_norm": 0.4946683943271637, "learning_rate": 9.540265807423401e-05, "loss": 0.0465, "step": 18070 }, { "grad_norm": 0.35816141963005066, "learning_rate": 9.53957299518599e-05, "loss": 0.0516, "step": 18080 }, { "grad_norm": 0.3874970078468323, "learning_rate": 9.5388796865077e-05, "loss": 0.0467, "step": 18090 }, { "grad_norm": 0.5838252305984497, "learning_rate": 9.538185881464353e-05, "loss": 0.0553, "step": 18100 }, { "grad_norm": 0.5301269292831421, "learning_rate": 9.537491580131821e-05, "loss": 0.0556, "step": 18110 }, { "grad_norm": 0.46191859245300293, "learning_rate": 9.53679678258603e-05, "loss": 0.0535, "step": 18120 }, { "grad_norm": 0.5544596910476685, "learning_rate": 9.536101488902966e-05, "loss": 0.0555, "step": 18130 }, { "grad_norm": 0.6304711699485779, "learning_rate": 9.535405699158663e-05, "loss": 0.0434, "step": 18140 }, { "grad_norm": 0.40842264890670776, "learning_rate": 9.53470941342921e-05, "loss": 0.05, "step": 18150 }, { "grad_norm": 0.41566556692123413, "learning_rate": 9.534012631790756e-05, "loss": 0.0499, "step": 18160 }, { "grad_norm": 0.4665977656841278, "learning_rate": 9.533315354319494e-05, "loss": 0.0514, "step": 18170 }, { "grad_norm": 0.4109445810317993, "learning_rate": 9.532617581091682e-05, "loss": 0.0408, "step": 18180 }, { "grad_norm": 0.3342849314212799, "learning_rate": 9.531919312183629e-05, "loss": 0.0493, "step": 18190 }, { "grad_norm": 0.4079100489616394, "learning_rate": 9.531220547671688e-05, "loss": 0.0374, "step": 18200 }, { "grad_norm": 0.40601348876953125, "learning_rate": 9.530521287632285e-05, "loss": 0.0528, "step": 18210 }, { "grad_norm": 0.5953576564788818, "learning_rate": 9.529821532141884e-05, "loss": 0.0455, "step": 18220 }, { "grad_norm": 0.41998592019081116, "learning_rate": 9.52912128127701e-05, "loss": 0.0575, "step": 18230 }, { "grad_norm": 0.9423258304595947, "learning_rate": 9.528420535114244e-05, "loss": 0.0545, "step": 18240 }, { "grad_norm": 0.6182305216789246, "learning_rate": 9.527719293730215e-05, "loss": 0.0596, "step": 18250 }, { "grad_norm": 0.38291752338409424, "learning_rate": 9.527017557201611e-05, "loss": 0.0495, "step": 18260 }, { "grad_norm": 0.7648221850395203, "learning_rate": 9.526315325605176e-05, "loss": 0.0582, "step": 18270 }, { "grad_norm": 0.4178129732608795, "learning_rate": 9.525612599017699e-05, "loss": 0.0441, "step": 18280 }, { "grad_norm": 0.4363729655742645, "learning_rate": 9.524909377516033e-05, "loss": 0.0616, "step": 18290 }, { "grad_norm": 0.41527512669563293, "learning_rate": 9.524205661177081e-05, "loss": 0.0452, "step": 18300 }, { "grad_norm": 0.5096110701560974, "learning_rate": 9.523501450077801e-05, "loss": 0.0533, "step": 18310 }, { "grad_norm": 0.5434151291847229, "learning_rate": 9.522796744295202e-05, "loss": 0.0529, "step": 18320 }, { "grad_norm": 0.3911150097846985, "learning_rate": 9.522091543906352e-05, "loss": 0.0518, "step": 18330 }, { "grad_norm": 0.5184406638145447, "learning_rate": 9.521385848988369e-05, "loss": 0.0493, "step": 18340 }, { "grad_norm": 0.5152271389961243, "learning_rate": 9.520679659618428e-05, "loss": 0.0539, "step": 18350 }, { "grad_norm": 0.39708849787712097, "learning_rate": 9.519972975873754e-05, "loss": 0.0493, "step": 18360 }, { "grad_norm": 0.47536420822143555, "learning_rate": 9.519265797831633e-05, "loss": 0.0519, "step": 18370 }, { "grad_norm": 0.5992770791053772, "learning_rate": 9.518558125569399e-05, "loss": 0.0584, "step": 18380 }, { "grad_norm": 0.3323584794998169, "learning_rate": 9.517849959164442e-05, "loss": 0.0441, "step": 18390 }, { "grad_norm": 0.4890178442001343, "learning_rate": 9.517141298694205e-05, "loss": 0.0553, "step": 18400 }, { "grad_norm": 0.5212576389312744, "learning_rate": 9.516432144236188e-05, "loss": 0.0606, "step": 18410 }, { "grad_norm": 0.6995443105697632, "learning_rate": 9.515722495867941e-05, "loss": 0.0503, "step": 18420 }, { "grad_norm": 0.38860809803009033, "learning_rate": 9.515012353667072e-05, "loss": 0.0497, "step": 18430 }, { "grad_norm": 0.43991321325302124, "learning_rate": 9.51430171771124e-05, "loss": 0.0493, "step": 18440 }, { "grad_norm": 0.41841840744018555, "learning_rate": 9.513590588078159e-05, "loss": 0.0564, "step": 18450 }, { "grad_norm": 0.5685324668884277, "learning_rate": 9.512878964845597e-05, "loss": 0.0529, "step": 18460 }, { "grad_norm": 0.564215898513794, "learning_rate": 9.512166848091377e-05, "loss": 0.0514, "step": 18470 }, { "grad_norm": 0.3437270224094391, "learning_rate": 9.511454237893376e-05, "loss": 0.0505, "step": 18480 }, { "grad_norm": 0.45332056283950806, "learning_rate": 9.51074113432952e-05, "loss": 0.0463, "step": 18490 }, { "grad_norm": 0.5617083311080933, "learning_rate": 9.510027537477797e-05, "loss": 0.0617, "step": 18500 }, { "grad_norm": 0.41985028982162476, "learning_rate": 9.509313447416242e-05, "loss": 0.0533, "step": 18510 }, { "grad_norm": 0.5881697535514832, "learning_rate": 9.508598864222949e-05, "loss": 0.0549, "step": 18520 }, { "grad_norm": 0.42272672057151794, "learning_rate": 9.507883787976062e-05, "loss": 0.0548, "step": 18530 }, { "grad_norm": 0.4129076302051544, "learning_rate": 9.507168218753781e-05, "loss": 0.043, "step": 18540 }, { "grad_norm": 0.38545092940330505, "learning_rate": 9.506452156634362e-05, "loss": 0.0531, "step": 18550 }, { "grad_norm": 0.34904026985168457, "learning_rate": 9.505735601696109e-05, "loss": 0.0614, "step": 18560 }, { "grad_norm": 0.4430590271949768, "learning_rate": 9.505018554017385e-05, "loss": 0.0522, "step": 18570 }, { "grad_norm": 0.5983860492706299, "learning_rate": 9.504301013676604e-05, "loss": 0.0514, "step": 18580 }, { "grad_norm": 0.4436846077442169, "learning_rate": 9.503582980752238e-05, "loss": 0.0476, "step": 18590 }, { "grad_norm": 0.5360317826271057, "learning_rate": 9.502864455322809e-05, "loss": 0.0671, "step": 18600 }, { "grad_norm": 0.35972562432289124, "learning_rate": 9.502145437466891e-05, "loss": 0.0459, "step": 18610 }, { "grad_norm": 0.2608446180820465, "learning_rate": 9.501425927263116e-05, "loss": 0.0508, "step": 18620 }, { "grad_norm": 0.5772404074668884, "learning_rate": 9.500705924790172e-05, "loss": 0.0745, "step": 18630 }, { "grad_norm": 0.441289484500885, "learning_rate": 9.499985430126794e-05, "loss": 0.0567, "step": 18640 }, { "grad_norm": 0.6827414035797119, "learning_rate": 9.499264443351775e-05, "loss": 0.0541, "step": 18650 }, { "grad_norm": 0.4930499792098999, "learning_rate": 9.498542964543961e-05, "loss": 0.0521, "step": 18660 }, { "grad_norm": 0.35196420550346375, "learning_rate": 9.497820993782252e-05, "loss": 0.0543, "step": 18670 }, { "grad_norm": 0.3704168498516083, "learning_rate": 9.497098531145601e-05, "loss": 0.0506, "step": 18680 }, { "grad_norm": 0.37277188897132874, "learning_rate": 9.496375576713017e-05, "loss": 0.05, "step": 18690 }, { "grad_norm": 0.3905121684074402, "learning_rate": 9.49565213056356e-05, "loss": 0.0479, "step": 18700 }, { "grad_norm": 0.2760460376739502, "learning_rate": 9.494928192776342e-05, "loss": 0.0417, "step": 18710 }, { "grad_norm": 0.3346012830734253, "learning_rate": 9.494203763430538e-05, "loss": 0.0349, "step": 18720 }, { "grad_norm": 0.488567978143692, "learning_rate": 9.493478842605366e-05, "loss": 0.0858, "step": 18730 }, { "grad_norm": 0.443128377199173, "learning_rate": 9.492753430380105e-05, "loss": 0.042, "step": 18740 }, { "grad_norm": 0.408766508102417, "learning_rate": 9.492027526834083e-05, "loss": 0.0437, "step": 18750 }, { "grad_norm": 0.4734814763069153, "learning_rate": 9.491301132046684e-05, "loss": 0.0408, "step": 18760 }, { "grad_norm": 0.5212728381156921, "learning_rate": 9.490574246097345e-05, "loss": 0.0663, "step": 18770 }, { "grad_norm": 0.46734732389450073, "learning_rate": 9.48984686906556e-05, "loss": 0.0596, "step": 18780 }, { "grad_norm": 0.6256488561630249, "learning_rate": 9.489119001030871e-05, "loss": 0.0498, "step": 18790 }, { "grad_norm": 0.47730720043182373, "learning_rate": 9.488390642072878e-05, "loss": 0.0507, "step": 18800 }, { "grad_norm": 0.3604285418987274, "learning_rate": 9.48766179227123e-05, "loss": 0.0495, "step": 18810 }, { "grad_norm": 0.46395742893218994, "learning_rate": 9.486932451705636e-05, "loss": 0.0528, "step": 18820 }, { "grad_norm": 0.46459633111953735, "learning_rate": 9.486202620455857e-05, "loss": 0.0518, "step": 18830 }, { "grad_norm": 0.4282236397266388, "learning_rate": 9.485472298601704e-05, "loss": 0.0468, "step": 18840 }, { "grad_norm": 0.563555896282196, "learning_rate": 9.484741486223043e-05, "loss": 0.0508, "step": 18850 }, { "grad_norm": 0.4238136112689972, "learning_rate": 9.484010183399797e-05, "loss": 0.0545, "step": 18860 }, { "grad_norm": 0.5954776406288147, "learning_rate": 9.483278390211938e-05, "loss": 0.0613, "step": 18870 }, { "grad_norm": 0.499843031167984, "learning_rate": 9.482546106739496e-05, "loss": 0.0481, "step": 18880 }, { "grad_norm": 0.5201693773269653, "learning_rate": 9.48181333306255e-05, "loss": 0.0547, "step": 18890 }, { "grad_norm": 0.37553730607032776, "learning_rate": 9.481080069261237e-05, "loss": 0.0622, "step": 18900 }, { "grad_norm": 0.32534557580947876, "learning_rate": 9.480346315415745e-05, "loss": 0.0422, "step": 18910 }, { "grad_norm": 1.214626431465149, "learning_rate": 9.479612071606314e-05, "loss": 0.0534, "step": 18920 }, { "grad_norm": 0.5202715396881104, "learning_rate": 9.478877337913244e-05, "loss": 0.065, "step": 18930 }, { "grad_norm": 0.4024371802806854, "learning_rate": 9.478142114416881e-05, "loss": 0.0481, "step": 18940 }, { "grad_norm": 0.5619843006134033, "learning_rate": 9.47740640119763e-05, "loss": 0.0481, "step": 18950 }, { "grad_norm": 0.39228150248527527, "learning_rate": 9.476670198335947e-05, "loss": 0.0518, "step": 18960 }, { "grad_norm": 0.4684112071990967, "learning_rate": 9.47593350591234e-05, "loss": 0.0575, "step": 18970 }, { "grad_norm": 0.4603588581085205, "learning_rate": 9.475196324007376e-05, "loss": 0.0479, "step": 18980 }, { "grad_norm": 0.55254065990448, "learning_rate": 9.474458652701669e-05, "loss": 0.0482, "step": 18990 }, { "grad_norm": 0.45118990540504456, "learning_rate": 9.473720492075892e-05, "loss": 0.0528, "step": 19000 }, { "grad_norm": 0.5634634494781494, "learning_rate": 9.472981842210768e-05, "loss": 0.0546, "step": 19010 }, { "grad_norm": 0.5608974695205688, "learning_rate": 9.472242703187074e-05, "loss": 0.0532, "step": 19020 }, { "grad_norm": 0.3735717236995697, "learning_rate": 9.471503075085643e-05, "loss": 0.0418, "step": 19030 }, { "grad_norm": 0.5848118662834167, "learning_rate": 9.470762957987359e-05, "loss": 0.0561, "step": 19040 }, { "grad_norm": 0.35019904375076294, "learning_rate": 9.470022351973158e-05, "loss": 0.0474, "step": 19050 }, { "grad_norm": 0.34888848662376404, "learning_rate": 9.469281257124034e-05, "loss": 0.045, "step": 19060 }, { "grad_norm": 0.3954950273036957, "learning_rate": 9.46853967352103e-05, "loss": 0.0594, "step": 19070 }, { "grad_norm": 0.49146124720573425, "learning_rate": 9.467797601245246e-05, "loss": 0.0443, "step": 19080 }, { "grad_norm": 0.35714396834373474, "learning_rate": 9.467055040377834e-05, "loss": 0.0533, "step": 19090 }, { "grad_norm": 0.336662232875824, "learning_rate": 9.466311990999999e-05, "loss": 0.0523, "step": 19100 }, { "grad_norm": 0.36017847061157227, "learning_rate": 9.465568453193e-05, "loss": 0.0449, "step": 19110 }, { "grad_norm": 0.6214938163757324, "learning_rate": 9.464824427038148e-05, "loss": 0.0473, "step": 19120 }, { "grad_norm": 0.28906959295272827, "learning_rate": 9.46407991261681e-05, "loss": 0.0622, "step": 19130 }, { "grad_norm": 0.3698815405368805, "learning_rate": 9.463334910010404e-05, "loss": 0.0436, "step": 19140 }, { "grad_norm": 0.3816661536693573, "learning_rate": 9.462589419300403e-05, "loss": 0.041, "step": 19150 }, { "grad_norm": 0.5958940982818604, "learning_rate": 9.461843440568333e-05, "loss": 0.0563, "step": 19160 }, { "grad_norm": 0.3513500988483429, "learning_rate": 9.461096973895773e-05, "loss": 0.0522, "step": 19170 }, { "grad_norm": 0.43551138043403625, "learning_rate": 9.460350019364355e-05, "loss": 0.0591, "step": 19180 }, { "grad_norm": 0.53420490026474, "learning_rate": 9.459602577055764e-05, "loss": 0.0638, "step": 19190 }, { "grad_norm": 0.5119250416755676, "learning_rate": 9.45885464705174e-05, "loss": 0.0578, "step": 19200 }, { "grad_norm": 0.41861164569854736, "learning_rate": 9.458106229434076e-05, "loss": 0.0485, "step": 19210 }, { "grad_norm": 0.4541553854942322, "learning_rate": 9.457357324284617e-05, "loss": 0.0424, "step": 19220 }, { "grad_norm": 0.4012836515903473, "learning_rate": 9.456607931685262e-05, "loss": 0.0496, "step": 19230 }, { "grad_norm": 0.3608763813972473, "learning_rate": 9.455858051717965e-05, "loss": 0.0458, "step": 19240 }, { "grad_norm": 0.4652370810508728, "learning_rate": 9.45510768446473e-05, "loss": 0.0504, "step": 19250 }, { "grad_norm": 0.4176819324493408, "learning_rate": 9.454356830007618e-05, "loss": 0.0445, "step": 19260 }, { "grad_norm": 0.49406829476356506, "learning_rate": 9.45360548842874e-05, "loss": 0.0391, "step": 19270 }, { "grad_norm": 0.5042688250541687, "learning_rate": 9.452853659810261e-05, "loss": 0.0454, "step": 19280 }, { "grad_norm": 0.453578382730484, "learning_rate": 9.452101344234401e-05, "loss": 0.0642, "step": 19290 }, { "grad_norm": 0.43492478132247925, "learning_rate": 9.451348541783431e-05, "loss": 0.0545, "step": 19300 }, { "grad_norm": 0.4474110007286072, "learning_rate": 9.450595252539678e-05, "loss": 0.0437, "step": 19310 }, { "grad_norm": 0.5030648708343506, "learning_rate": 9.449841476585518e-05, "loss": 0.0642, "step": 19320 }, { "grad_norm": 0.5522092580795288, "learning_rate": 9.449087214003384e-05, "loss": 0.0645, "step": 19330 }, { "grad_norm": 0.5922317504882812, "learning_rate": 9.448332464875765e-05, "loss": 0.0751, "step": 19340 }, { "grad_norm": 0.43538013100624084, "learning_rate": 9.447577229285192e-05, "loss": 0.0635, "step": 19350 }, { "grad_norm": 0.7096980810165405, "learning_rate": 9.446821507314261e-05, "loss": 0.0492, "step": 19360 }, { "grad_norm": 0.5329936742782593, "learning_rate": 9.446065299045617e-05, "loss": 0.0571, "step": 19370 }, { "grad_norm": 0.5018047094345093, "learning_rate": 9.445308604561955e-05, "loss": 0.0592, "step": 19380 }, { "grad_norm": 0.46258294582366943, "learning_rate": 9.444551423946028e-05, "loss": 0.062, "step": 19390 }, { "grad_norm": 0.6224740147590637, "learning_rate": 9.443793757280638e-05, "loss": 0.0577, "step": 19400 }, { "grad_norm": 0.5011950135231018, "learning_rate": 9.443035604648646e-05, "loss": 0.055, "step": 19410 }, { "grad_norm": 0.5104503631591797, "learning_rate": 9.44227696613296e-05, "loss": 0.0452, "step": 19420 }, { "grad_norm": 0.6167073845863342, "learning_rate": 9.441517841816542e-05, "loss": 0.074, "step": 19430 }, { "grad_norm": 0.4909762442111969, "learning_rate": 9.440758231782413e-05, "loss": 0.05, "step": 19440 }, { "grad_norm": 0.5624883770942688, "learning_rate": 9.439998136113639e-05, "loss": 0.0554, "step": 19450 }, { "grad_norm": 0.5781270265579224, "learning_rate": 9.439237554893344e-05, "loss": 0.0507, "step": 19460 }, { "grad_norm": 0.4352567791938782, "learning_rate": 9.438476488204705e-05, "loss": 0.0573, "step": 19470 }, { "grad_norm": 0.3130776584148407, "learning_rate": 9.43771493613095e-05, "loss": 0.0472, "step": 19480 }, { "grad_norm": 0.43715888261795044, "learning_rate": 9.436952898755362e-05, "loss": 0.059, "step": 19490 }, { "grad_norm": 0.4472009241580963, "learning_rate": 9.436190376161276e-05, "loss": 0.0555, "step": 19500 }, { "grad_norm": 0.43309593200683594, "learning_rate": 9.43542736843208e-05, "loss": 0.0607, "step": 19510 }, { "grad_norm": 0.5550023317337036, "learning_rate": 9.434663875651216e-05, "loss": 0.0456, "step": 19520 }, { "grad_norm": 0.45039719343185425, "learning_rate": 9.433899897902177e-05, "loss": 0.0455, "step": 19530 }, { "grad_norm": 0.43744710087776184, "learning_rate": 9.433135435268511e-05, "loss": 0.0483, "step": 19540 }, { "grad_norm": 0.4130777418613434, "learning_rate": 9.432370487833819e-05, "loss": 0.0571, "step": 19550 }, { "grad_norm": 0.589715301990509, "learning_rate": 9.431605055681756e-05, "loss": 0.0615, "step": 19560 }, { "grad_norm": 0.3534713089466095, "learning_rate": 9.430839138896026e-05, "loss": 0.0504, "step": 19570 }, { "grad_norm": 0.33483073115348816, "learning_rate": 9.43007273756039e-05, "loss": 0.0553, "step": 19580 }, { "grad_norm": 0.49315840005874634, "learning_rate": 9.429305851758658e-05, "loss": 0.045, "step": 19590 }, { "grad_norm": 0.4757048785686493, "learning_rate": 9.428538481574699e-05, "loss": 0.0548, "step": 19600 }, { "grad_norm": 0.3556511104106903, "learning_rate": 9.42777062709243e-05, "loss": 0.041, "step": 19610 }, { "grad_norm": 0.39622533321380615, "learning_rate": 9.427002288395821e-05, "loss": 0.0442, "step": 19620 }, { "grad_norm": 0.36131003499031067, "learning_rate": 9.426233465568898e-05, "loss": 0.0479, "step": 19630 }, { "grad_norm": 0.52576744556427, "learning_rate": 9.42546415869574e-05, "loss": 0.0392, "step": 19640 }, { "grad_norm": 0.45540139079093933, "learning_rate": 9.424694367860473e-05, "loss": 0.0505, "step": 19650 }, { "grad_norm": 0.4476318955421448, "learning_rate": 9.423924093147284e-05, "loss": 0.0424, "step": 19660 }, { "grad_norm": 0.5436407923698425, "learning_rate": 9.423153334640407e-05, "loss": 0.051, "step": 19670 }, { "grad_norm": 0.3308407962322235, "learning_rate": 9.42238209242413e-05, "loss": 0.0423, "step": 19680 }, { "grad_norm": 0.3675555884838104, "learning_rate": 9.421610366582798e-05, "loss": 0.048, "step": 19690 }, { "grad_norm": 0.39194756746292114, "learning_rate": 9.420838157200803e-05, "loss": 0.0491, "step": 19700 }, { "grad_norm": 0.34728214144706726, "learning_rate": 9.420065464362594e-05, "loss": 0.0521, "step": 19710 }, { "grad_norm": 0.5276155471801758, "learning_rate": 9.419292288152673e-05, "loss": 0.0511, "step": 19720 }, { "grad_norm": 0.5759470462799072, "learning_rate": 9.418518628655588e-05, "loss": 0.0443, "step": 19730 }, { "grad_norm": 0.45818567276000977, "learning_rate": 9.417744485955951e-05, "loss": 0.0495, "step": 19740 }, { "grad_norm": 0.5994304418563843, "learning_rate": 9.41696986013842e-05, "loss": 0.0483, "step": 19750 }, { "grad_norm": 0.4518221318721771, "learning_rate": 9.416194751287705e-05, "loss": 0.0653, "step": 19760 }, { "grad_norm": 0.5194219350814819, "learning_rate": 9.415419159488572e-05, "loss": 0.0486, "step": 19770 }, { "grad_norm": 0.543285608291626, "learning_rate": 9.414643084825837e-05, "loss": 0.0614, "step": 19780 }, { "grad_norm": 0.5639927983283997, "learning_rate": 9.413866527384372e-05, "loss": 0.0571, "step": 19790 }, { "grad_norm": 0.3769744038581848, "learning_rate": 9.4130894872491e-05, "loss": 0.0426, "step": 19800 }, { "grad_norm": 0.4614863991737366, "learning_rate": 9.412311964504998e-05, "loss": 0.0515, "step": 19810 }, { "grad_norm": 0.9039638042449951, "learning_rate": 9.411533959237091e-05, "loss": 0.0544, "step": 19820 }, { "grad_norm": 0.3087993860244751, "learning_rate": 9.410755471530464e-05, "loss": 0.0446, "step": 19830 }, { "grad_norm": 0.44593656063079834, "learning_rate": 9.40997650147025e-05, "loss": 0.0437, "step": 19840 }, { "grad_norm": 0.5324743986129761, "learning_rate": 9.409197049141637e-05, "loss": 0.0629, "step": 19850 }, { "grad_norm": 0.4294448792934418, "learning_rate": 9.408417114629863e-05, "loss": 0.0541, "step": 19860 }, { "grad_norm": 0.5168408155441284, "learning_rate": 9.40763669802022e-05, "loss": 0.0479, "step": 19870 }, { "grad_norm": 0.4492853283882141, "learning_rate": 9.406855799398056e-05, "loss": 0.0595, "step": 19880 }, { "grad_norm": 0.4165083169937134, "learning_rate": 9.406074418848767e-05, "loss": 0.0567, "step": 19890 }, { "grad_norm": 0.6215134263038635, "learning_rate": 9.405292556457805e-05, "loss": 0.0738, "step": 19900 }, { "grad_norm": 0.43781784176826477, "learning_rate": 9.404510212310671e-05, "loss": 0.0557, "step": 19910 }, { "grad_norm": 0.39610788226127625, "learning_rate": 9.403727386492924e-05, "loss": 0.0464, "step": 19920 }, { "grad_norm": 0.36856409907341003, "learning_rate": 9.40294407909017e-05, "loss": 0.0516, "step": 19930 }, { "grad_norm": 0.388029545545578, "learning_rate": 9.40216029018807e-05, "loss": 0.0476, "step": 19940 }, { "grad_norm": 0.4598864018917084, "learning_rate": 9.401376019872338e-05, "loss": 0.0485, "step": 19950 }, { "grad_norm": 0.6261946558952332, "learning_rate": 9.400591268228746e-05, "loss": 0.0522, "step": 19960 }, { "grad_norm": 0.5442284345626831, "learning_rate": 9.399806035343106e-05, "loss": 0.0469, "step": 19970 }, { "grad_norm": 0.5108778476715088, "learning_rate": 9.399020321301294e-05, "loss": 0.0475, "step": 19980 }, { "grad_norm": 0.5008295178413391, "learning_rate": 9.398234126189234e-05, "loss": 0.0551, "step": 19990 }, { "grad_norm": 0.5583169460296631, "learning_rate": 9.397447450092902e-05, "loss": 0.0474, "step": 20000 }, { "grad_norm": 0.37918099761009216, "learning_rate": 9.39666029309833e-05, "loss": 0.0445, "step": 20010 }, { "grad_norm": 0.7195885181427002, "learning_rate": 9.395872655291596e-05, "loss": 0.0509, "step": 20020 }, { "grad_norm": 0.7403667569160461, "learning_rate": 9.395084536758838e-05, "loss": 0.0471, "step": 20030 }, { "grad_norm": 0.4762256145477295, "learning_rate": 9.394295937586243e-05, "loss": 0.0612, "step": 20040 }, { "grad_norm": 0.49062323570251465, "learning_rate": 9.393506857860052e-05, "loss": 0.0471, "step": 20050 }, { "grad_norm": 0.39373356103897095, "learning_rate": 9.392717297666555e-05, "loss": 0.039, "step": 20060 }, { "grad_norm": 0.5704625248908997, "learning_rate": 9.391927257092101e-05, "loss": 0.0523, "step": 20070 }, { "grad_norm": 0.561783492565155, "learning_rate": 9.391136736223085e-05, "loss": 0.0503, "step": 20080 }, { "grad_norm": 0.4289267063140869, "learning_rate": 9.390345735145956e-05, "loss": 0.055, "step": 20090 }, { "grad_norm": 0.5720263719558716, "learning_rate": 9.389554253947219e-05, "loss": 0.0593, "step": 20100 }, { "grad_norm": 0.3623042702674866, "learning_rate": 9.388762292713428e-05, "loss": 0.0474, "step": 20110 }, { "grad_norm": 0.4625292420387268, "learning_rate": 9.38796985153119e-05, "loss": 0.0501, "step": 20120 }, { "grad_norm": 0.5069807171821594, "learning_rate": 9.387176930487169e-05, "loss": 0.0495, "step": 20130 }, { "grad_norm": 0.4808501601219177, "learning_rate": 9.386383529668072e-05, "loss": 0.0448, "step": 20140 }, { "grad_norm": 0.45843014121055603, "learning_rate": 9.385589649160669e-05, "loss": 0.0465, "step": 20150 }, { "grad_norm": 0.34826725721359253, "learning_rate": 9.384795289051775e-05, "loss": 0.0534, "step": 20160 }, { "grad_norm": 0.4126121997833252, "learning_rate": 9.384000449428261e-05, "loss": 0.0534, "step": 20170 }, { "grad_norm": 0.41357648372650146, "learning_rate": 9.383205130377048e-05, "loss": 0.0584, "step": 20180 }, { "grad_norm": 0.47083455324172974, "learning_rate": 9.382409331985114e-05, "loss": 0.0431, "step": 20190 }, { "grad_norm": 0.3769761621952057, "learning_rate": 9.381613054339482e-05, "loss": 0.0464, "step": 20200 }, { "grad_norm": 0.3253265917301178, "learning_rate": 9.380816297527235e-05, "loss": 0.0433, "step": 20210 }, { "grad_norm": 0.38069361448287964, "learning_rate": 9.380019061635506e-05, "loss": 0.0508, "step": 20220 }, { "grad_norm": 0.5287773013114929, "learning_rate": 9.379221346751474e-05, "loss": 0.0594, "step": 20230 }, { "grad_norm": 0.47126781940460205, "learning_rate": 9.378423152962382e-05, "loss": 0.0442, "step": 20240 }, { "grad_norm": 0.40707165002822876, "learning_rate": 9.377624480355517e-05, "loss": 0.0526, "step": 20250 }, { "grad_norm": 0.42669478058815, "learning_rate": 9.376825329018219e-05, "loss": 0.0399, "step": 20260 }, { "grad_norm": 0.3529042601585388, "learning_rate": 9.376025699037884e-05, "loss": 0.0436, "step": 20270 }, { "grad_norm": 0.5150648355484009, "learning_rate": 9.37522559050196e-05, "loss": 0.0405, "step": 20280 }, { "grad_norm": 0.4544627368450165, "learning_rate": 9.37442500349794e-05, "loss": 0.0577, "step": 20290 }, { "grad_norm": 0.3402476906776428, "learning_rate": 9.373623938113381e-05, "loss": 0.045, "step": 20300 }, { "grad_norm": 0.46287858486175537, "learning_rate": 9.372822394435883e-05, "loss": 0.0637, "step": 20310 }, { "grad_norm": 0.46894073486328125, "learning_rate": 9.372020372553102e-05, "loss": 0.0504, "step": 20320 }, { "grad_norm": 0.47845548391342163, "learning_rate": 9.371217872552746e-05, "loss": 0.0453, "step": 20330 }, { "grad_norm": 0.36570116877555847, "learning_rate": 9.370414894522576e-05, "loss": 0.05, "step": 20340 }, { "grad_norm": 0.426578551530838, "learning_rate": 9.369611438550406e-05, "loss": 0.0584, "step": 20350 }, { "grad_norm": 0.3988212049007416, "learning_rate": 9.368807504724095e-05, "loss": 0.0507, "step": 20360 }, { "grad_norm": 0.5009049773216248, "learning_rate": 9.368003093131565e-05, "loss": 0.0506, "step": 20370 }, { "grad_norm": 0.48177430033683777, "learning_rate": 9.367198203860785e-05, "loss": 0.0586, "step": 20380 }, { "grad_norm": 0.4547085762023926, "learning_rate": 9.366392836999774e-05, "loss": 0.0514, "step": 20390 }, { "grad_norm": 0.41502636671066284, "learning_rate": 9.365586992636607e-05, "loss": 0.041, "step": 20400 }, { "grad_norm": 0.3784855008125305, "learning_rate": 9.364780670859412e-05, "loss": 0.0403, "step": 20410 }, { "grad_norm": 0.6096104383468628, "learning_rate": 9.363973871756364e-05, "loss": 0.0479, "step": 20420 }, { "grad_norm": 0.3918662667274475, "learning_rate": 9.363166595415696e-05, "loss": 0.0435, "step": 20430 }, { "grad_norm": 0.5260177254676819, "learning_rate": 9.362358841925686e-05, "loss": 0.0482, "step": 20440 }, { "grad_norm": 0.3881835639476776, "learning_rate": 9.361550611374674e-05, "loss": 0.0518, "step": 20450 }, { "grad_norm": 0.3653308153152466, "learning_rate": 9.360741903851043e-05, "loss": 0.0694, "step": 20460 }, { "grad_norm": 0.3841651380062103, "learning_rate": 9.359932719443236e-05, "loss": 0.0618, "step": 20470 }, { "grad_norm": 0.37106356024742126, "learning_rate": 9.35912305823974e-05, "loss": 0.0501, "step": 20480 }, { "grad_norm": 0.396149605512619, "learning_rate": 9.358312920329101e-05, "loss": 0.0395, "step": 20490 }, { "grad_norm": 0.3419133722782135, "learning_rate": 9.357502305799914e-05, "loss": 0.0419, "step": 20500 }, { "grad_norm": 0.4836396276950836, "learning_rate": 9.356691214740824e-05, "loss": 0.0487, "step": 20510 }, { "grad_norm": 0.35890573263168335, "learning_rate": 9.355879647240535e-05, "loss": 0.0454, "step": 20520 }, { "grad_norm": 0.41055652499198914, "learning_rate": 9.355067603387798e-05, "loss": 0.0539, "step": 20530 }, { "grad_norm": 0.46419262886047363, "learning_rate": 9.354255083271412e-05, "loss": 0.0503, "step": 20540 }, { "grad_norm": 0.588219165802002, "learning_rate": 9.353442086980239e-05, "loss": 0.0534, "step": 20550 }, { "grad_norm": 0.478013277053833, "learning_rate": 9.352628614603185e-05, "loss": 0.0444, "step": 20560 }, { "grad_norm": 0.5083893537521362, "learning_rate": 9.351814666229209e-05, "loss": 0.0446, "step": 20570 }, { "grad_norm": 0.4110487103462219, "learning_rate": 9.351000241947324e-05, "loss": 0.0464, "step": 20580 }, { "grad_norm": 0.3859608769416809, "learning_rate": 9.350185341846594e-05, "loss": 0.05, "step": 20590 }, { "grad_norm": 0.3718419075012207, "learning_rate": 9.349369966016134e-05, "loss": 0.0481, "step": 20600 }, { "grad_norm": 0.4989255964756012, "learning_rate": 9.348554114545117e-05, "loss": 0.0631, "step": 20610 }, { "grad_norm": 0.27065208554267883, "learning_rate": 9.347737787522758e-05, "loss": 0.0421, "step": 20620 }, { "grad_norm": 0.48379072546958923, "learning_rate": 9.346920985038332e-05, "loss": 0.0528, "step": 20630 }, { "grad_norm": 0.587813138961792, "learning_rate": 9.346103707181162e-05, "loss": 0.0542, "step": 20640 }, { "grad_norm": 0.36018869280815125, "learning_rate": 9.345285954040626e-05, "loss": 0.0386, "step": 20650 }, { "grad_norm": 0.9736115336418152, "learning_rate": 9.34446772570615e-05, "loss": 0.0454, "step": 20660 }, { "grad_norm": 0.3372672498226166, "learning_rate": 9.343649022267214e-05, "loss": 0.0458, "step": 20670 }, { "grad_norm": 0.4863828122615814, "learning_rate": 9.342829843813353e-05, "loss": 0.0439, "step": 20680 }, { "grad_norm": 0.47860535979270935, "learning_rate": 9.342010190434149e-05, "loss": 0.0476, "step": 20690 }, { "grad_norm": 0.6675377488136292, "learning_rate": 9.34119006221924e-05, "loss": 0.0487, "step": 20700 }, { "grad_norm": 0.36214518547058105, "learning_rate": 9.340369459258313e-05, "loss": 0.0475, "step": 20710 }, { "grad_norm": 0.47863101959228516, "learning_rate": 9.339548381641106e-05, "loss": 0.0567, "step": 20720 }, { "grad_norm": 0.6918916702270508, "learning_rate": 9.338726829457413e-05, "loss": 0.0621, "step": 20730 }, { "grad_norm": 0.393008291721344, "learning_rate": 9.337904802797078e-05, "loss": 0.0391, "step": 20740 }, { "grad_norm": 0.6294636726379395, "learning_rate": 9.337082301749993e-05, "loss": 0.0586, "step": 20750 }, { "grad_norm": 0.32825416326522827, "learning_rate": 9.336259326406109e-05, "loss": 0.0538, "step": 20760 }, { "grad_norm": 0.4708007574081421, "learning_rate": 9.335435876855427e-05, "loss": 0.0638, "step": 20770 }, { "grad_norm": 0.5837438106536865, "learning_rate": 9.334611953187994e-05, "loss": 0.0697, "step": 20780 }, { "grad_norm": 0.52474045753479, "learning_rate": 9.333787555493914e-05, "loss": 0.0481, "step": 20790 }, { "grad_norm": 0.45729419589042664, "learning_rate": 9.332962683863345e-05, "loss": 0.0588, "step": 20800 }, { "grad_norm": 0.3704743981361389, "learning_rate": 9.332137338386489e-05, "loss": 0.0458, "step": 20810 }, { "grad_norm": 0.3564785420894623, "learning_rate": 9.33131151915361e-05, "loss": 0.0572, "step": 20820 }, { "grad_norm": 0.42946454882621765, "learning_rate": 9.330485226255012e-05, "loss": 0.043, "step": 20830 }, { "grad_norm": 0.3461100459098816, "learning_rate": 9.329658459781061e-05, "loss": 0.0484, "step": 20840 }, { "grad_norm": 0.433942973613739, "learning_rate": 9.328831219822172e-05, "loss": 0.0463, "step": 20850 }, { "grad_norm": 0.3289070129394531, "learning_rate": 9.328003506468808e-05, "loss": 0.0513, "step": 20860 }, { "grad_norm": 0.457592248916626, "learning_rate": 9.327175319811488e-05, "loss": 0.0471, "step": 20870 }, { "grad_norm": 0.4159301221370697, "learning_rate": 9.326346659940781e-05, "loss": 0.053, "step": 20880 }, { "grad_norm": 0.3596833348274231, "learning_rate": 9.325517526947308e-05, "loss": 0.0501, "step": 20890 }, { "grad_norm": 0.5996096730232239, "learning_rate": 9.32468792092174e-05, "loss": 0.0448, "step": 20900 }, { "grad_norm": 0.3998353183269501, "learning_rate": 9.323857841954803e-05, "loss": 0.0426, "step": 20910 }, { "grad_norm": 0.3137867748737335, "learning_rate": 9.323027290137276e-05, "loss": 0.0422, "step": 20920 }, { "grad_norm": 0.4122415781021118, "learning_rate": 9.322196265559981e-05, "loss": 0.041, "step": 20930 }, { "grad_norm": 0.4009649157524109, "learning_rate": 9.321364768313803e-05, "loss": 0.0429, "step": 20940 }, { "grad_norm": 0.42358991503715515, "learning_rate": 9.32053279848967e-05, "loss": 0.0503, "step": 20950 }, { "grad_norm": 0.450593501329422, "learning_rate": 9.319700356178567e-05, "loss": 0.0482, "step": 20960 }, { "grad_norm": 0.5467767715454102, "learning_rate": 9.318867441471527e-05, "loss": 0.0493, "step": 20970 }, { "grad_norm": 0.5483495593070984, "learning_rate": 9.318034054459637e-05, "loss": 0.0465, "step": 20980 }, { "grad_norm": 0.6007266640663147, "learning_rate": 9.317200195234034e-05, "loss": 0.0532, "step": 20990 }, { "grad_norm": 0.5331292152404785, "learning_rate": 9.316365863885909e-05, "loss": 0.0527, "step": 21000 }, { "grad_norm": 0.7207885384559631, "learning_rate": 9.315531060506502e-05, "loss": 0.0613, "step": 21010 }, { "grad_norm": 0.39674001932144165, "learning_rate": 9.314695785187108e-05, "loss": 0.0457, "step": 21020 }, { "grad_norm": 0.48050472140312195, "learning_rate": 9.313860038019069e-05, "loss": 0.0536, "step": 21030 }, { "grad_norm": 0.4295349419116974, "learning_rate": 9.313023819093782e-05, "loss": 0.0419, "step": 21040 }, { "grad_norm": 0.6108434796333313, "learning_rate": 9.312187128502695e-05, "loss": 0.0468, "step": 21050 }, { "grad_norm": 0.44691047072410583, "learning_rate": 9.311349966337307e-05, "loss": 0.0502, "step": 21060 }, { "grad_norm": 0.3747505843639374, "learning_rate": 9.310512332689169e-05, "loss": 0.0471, "step": 21070 }, { "grad_norm": 0.4624175727367401, "learning_rate": 9.309674227649883e-05, "loss": 0.0473, "step": 21080 }, { "grad_norm": 0.3947164714336395, "learning_rate": 9.308835651311103e-05, "loss": 0.0477, "step": 21090 }, { "grad_norm": 0.48381513357162476, "learning_rate": 9.307996603764533e-05, "loss": 0.0547, "step": 21100 }, { "grad_norm": 0.5311170220375061, "learning_rate": 9.307157085101932e-05, "loss": 0.0454, "step": 21110 }, { "grad_norm": 0.4818868041038513, "learning_rate": 9.306317095415109e-05, "loss": 0.0436, "step": 21120 }, { "grad_norm": 0.35905721783638, "learning_rate": 9.305476634795922e-05, "loss": 0.0495, "step": 21130 }, { "grad_norm": 0.5609683394432068, "learning_rate": 9.304635703336284e-05, "loss": 0.05, "step": 21140 }, { "grad_norm": 0.4674769937992096, "learning_rate": 9.303794301128157e-05, "loss": 0.0401, "step": 21150 }, { "grad_norm": 0.7054286599159241, "learning_rate": 9.302952428263555e-05, "loss": 0.0529, "step": 21160 }, { "grad_norm": 0.35793566703796387, "learning_rate": 9.302110084834545e-05, "loss": 0.0506, "step": 21170 }, { "grad_norm": 0.4478374123573303, "learning_rate": 9.301267270933245e-05, "loss": 0.0494, "step": 21180 }, { "grad_norm": 0.42232799530029297, "learning_rate": 9.300423986651823e-05, "loss": 0.0536, "step": 21190 }, { "grad_norm": 0.6516313552856445, "learning_rate": 9.299580232082501e-05, "loss": 0.0521, "step": 21200 }, { "grad_norm": 0.5821661949157715, "learning_rate": 9.298736007317547e-05, "loss": 0.0646, "step": 21210 }, { "grad_norm": 0.4292038083076477, "learning_rate": 9.297891312449288e-05, "loss": 0.0567, "step": 21220 }, { "grad_norm": 0.44992560148239136, "learning_rate": 9.297046147570094e-05, "loss": 0.0548, "step": 21230 }, { "grad_norm": 0.4843088984489441, "learning_rate": 9.296200512772396e-05, "loss": 0.0459, "step": 21240 }, { "grad_norm": 0.48049408197402954, "learning_rate": 9.295354408148668e-05, "loss": 0.0462, "step": 21250 }, { "grad_norm": 0.45727235078811646, "learning_rate": 9.294507833791441e-05, "loss": 0.0526, "step": 21260 }, { "grad_norm": 0.2817414402961731, "learning_rate": 9.293660789793295e-05, "loss": 0.044, "step": 21270 }, { "grad_norm": 0.3683687448501587, "learning_rate": 9.292813276246858e-05, "loss": 0.0486, "step": 21280 }, { "grad_norm": 0.36798444390296936, "learning_rate": 9.291965293244816e-05, "loss": 0.0428, "step": 21290 }, { "grad_norm": 0.9944269061088562, "learning_rate": 9.291116840879904e-05, "loss": 0.0455, "step": 21300 }, { "grad_norm": 0.4577493369579315, "learning_rate": 9.290267919244904e-05, "loss": 0.0433, "step": 21310 }, { "grad_norm": 0.41733384132385254, "learning_rate": 9.289418528432655e-05, "loss": 0.0479, "step": 21320 }, { "grad_norm": 0.5392957925796509, "learning_rate": 9.288568668536045e-05, "loss": 0.0612, "step": 21330 }, { "grad_norm": 0.6929922699928284, "learning_rate": 9.287718339648013e-05, "loss": 0.046, "step": 21340 }, { "grad_norm": 0.4131379723548889, "learning_rate": 9.28686754186155e-05, "loss": 0.0527, "step": 21350 }, { "grad_norm": 0.8040526509284973, "learning_rate": 9.286016275269698e-05, "loss": 0.0506, "step": 21360 }, { "grad_norm": 0.653273344039917, "learning_rate": 9.285164539965551e-05, "loss": 0.0471, "step": 21370 }, { "grad_norm": 0.6961740255355835, "learning_rate": 9.284312336042251e-05, "loss": 0.054, "step": 21380 }, { "grad_norm": 0.3623643219470978, "learning_rate": 9.283459663592996e-05, "loss": 0.0572, "step": 21390 }, { "grad_norm": 0.536969006061554, "learning_rate": 9.282606522711033e-05, "loss": 0.0537, "step": 21400 }, { "grad_norm": 0.32423102855682373, "learning_rate": 9.281752913489657e-05, "loss": 0.0532, "step": 21410 }, { "grad_norm": 0.42631012201309204, "learning_rate": 9.280898836022222e-05, "loss": 0.0498, "step": 21420 }, { "grad_norm": 0.4487069547176361, "learning_rate": 9.280044290402126e-05, "loss": 0.0455, "step": 21430 }, { "grad_norm": 0.3736741244792938, "learning_rate": 9.279189276722821e-05, "loss": 0.0486, "step": 21440 }, { "grad_norm": 0.4834948480129242, "learning_rate": 9.278333795077812e-05, "loss": 0.0568, "step": 21450 }, { "grad_norm": 0.39164799451828003, "learning_rate": 9.27747784556065e-05, "loss": 0.0466, "step": 21460 }, { "grad_norm": 0.48690298199653625, "learning_rate": 9.276621428264942e-05, "loss": 0.0586, "step": 21470 }, { "grad_norm": 0.3374686539173126, "learning_rate": 9.275764543284345e-05, "loss": 0.0433, "step": 21480 }, { "grad_norm": 0.4473632872104645, "learning_rate": 9.274907190712566e-05, "loss": 0.0418, "step": 21490 }, { "grad_norm": 0.43966802954673767, "learning_rate": 9.274049370643363e-05, "loss": 0.0466, "step": 21500 }, { "grad_norm": 0.37001293897628784, "learning_rate": 9.273191083170547e-05, "loss": 0.0447, "step": 21510 }, { "grad_norm": 0.7063493132591248, "learning_rate": 9.27233232838798e-05, "loss": 0.0432, "step": 21520 }, { "grad_norm": 0.4520166218280792, "learning_rate": 9.27147310638957e-05, "loss": 0.0417, "step": 21530 }, { "grad_norm": 0.3849700093269348, "learning_rate": 9.270613417269286e-05, "loss": 0.0404, "step": 21540 }, { "grad_norm": 0.572388231754303, "learning_rate": 9.269753261121138e-05, "loss": 0.0464, "step": 21550 }, { "grad_norm": 0.41948553919792175, "learning_rate": 9.268892638039194e-05, "loss": 0.0498, "step": 21560 }, { "grad_norm": 0.4176383316516876, "learning_rate": 9.268031548117569e-05, "loss": 0.0459, "step": 21570 }, { "grad_norm": 0.41928717494010925, "learning_rate": 9.26716999145043e-05, "loss": 0.0518, "step": 21580 }, { "grad_norm": 0.4723687171936035, "learning_rate": 9.266307968131998e-05, "loss": 0.0599, "step": 21590 }, { "grad_norm": 0.3726470470428467, "learning_rate": 9.26544547825654e-05, "loss": 0.0445, "step": 21600 }, { "grad_norm": 0.3742199242115021, "learning_rate": 9.264582521918376e-05, "loss": 0.0532, "step": 21610 }, { "grad_norm": 0.305387407541275, "learning_rate": 9.263719099211881e-05, "loss": 0.0449, "step": 21620 }, { "grad_norm": 0.4554564952850342, "learning_rate": 9.262855210231476e-05, "loss": 0.0479, "step": 21630 }, { "grad_norm": 0.7271896600723267, "learning_rate": 9.261990855071633e-05, "loss": 0.045, "step": 21640 }, { "grad_norm": 0.48718470335006714, "learning_rate": 9.261126033826878e-05, "loss": 0.0441, "step": 21650 }, { "grad_norm": 0.779308021068573, "learning_rate": 9.260260746591786e-05, "loss": 0.0466, "step": 21660 }, { "grad_norm": 0.4308623671531677, "learning_rate": 9.259394993460985e-05, "loss": 0.0463, "step": 21670 }, { "grad_norm": 0.7349737286567688, "learning_rate": 9.258528774529151e-05, "loss": 0.0434, "step": 21680 }, { "grad_norm": 0.6614342331886292, "learning_rate": 9.257662089891013e-05, "loss": 0.0512, "step": 21690 }, { "grad_norm": 0.45094001293182373, "learning_rate": 9.25679493964135e-05, "loss": 0.0511, "step": 21700 }, { "grad_norm": 0.5619688034057617, "learning_rate": 9.255927323874994e-05, "loss": 0.0473, "step": 21710 }, { "grad_norm": 0.5150106549263, "learning_rate": 9.255059242686822e-05, "loss": 0.0578, "step": 21720 }, { "grad_norm": 0.47668203711509705, "learning_rate": 9.254190696171769e-05, "loss": 0.0397, "step": 21730 }, { "grad_norm": 0.39688608050346375, "learning_rate": 9.25332168442482e-05, "loss": 0.053, "step": 21740 }, { "grad_norm": 0.932736337184906, "learning_rate": 9.252452207541004e-05, "loss": 0.0528, "step": 21750 }, { "grad_norm": 0.6423521041870117, "learning_rate": 9.251582265615409e-05, "loss": 0.0727, "step": 21760 }, { "grad_norm": 0.4738970696926117, "learning_rate": 9.250711858743169e-05, "loss": 0.0627, "step": 21770 }, { "grad_norm": 0.41675105690956116, "learning_rate": 9.24984098701947e-05, "loss": 0.0597, "step": 21780 }, { "grad_norm": 0.5536840558052063, "learning_rate": 9.248969650539552e-05, "loss": 0.0508, "step": 21790 }, { "grad_norm": 0.6140878200531006, "learning_rate": 9.2480978493987e-05, "loss": 0.0496, "step": 21800 }, { "grad_norm": 0.4051900804042816, "learning_rate": 9.247225583692256e-05, "loss": 0.0423, "step": 21810 }, { "grad_norm": 0.4089197516441345, "learning_rate": 9.246352853515607e-05, "loss": 0.0478, "step": 21820 }, { "grad_norm": 0.5150741338729858, "learning_rate": 9.245479658964194e-05, "loss": 0.053, "step": 21830 }, { "grad_norm": 0.4412766993045807, "learning_rate": 9.244606000133507e-05, "loss": 0.0507, "step": 21840 }, { "grad_norm": 0.42773672938346863, "learning_rate": 9.24373187711909e-05, "loss": 0.0442, "step": 21850 }, { "grad_norm": 0.48029476404190063, "learning_rate": 9.242857290016537e-05, "loss": 0.0548, "step": 21860 }, { "grad_norm": 0.40799587965011597, "learning_rate": 9.241982238921488e-05, "loss": 0.0515, "step": 21870 }, { "grad_norm": 0.5847422480583191, "learning_rate": 9.24110672392964e-05, "loss": 0.0456, "step": 21880 }, { "grad_norm": 0.38403815031051636, "learning_rate": 9.240230745136737e-05, "loss": 0.0389, "step": 21890 }, { "grad_norm": 0.3778294324874878, "learning_rate": 9.239354302638575e-05, "loss": 0.0363, "step": 21900 }, { "grad_norm": 0.727206826210022, "learning_rate": 9.238477396531e-05, "loss": 0.0487, "step": 21910 }, { "grad_norm": 0.47260817885398865, "learning_rate": 9.23760002690991e-05, "loss": 0.0516, "step": 21920 }, { "grad_norm": 0.5300275683403015, "learning_rate": 9.236722193871252e-05, "loss": 0.0425, "step": 21930 }, { "grad_norm": 0.4349202811717987, "learning_rate": 9.235843897511023e-05, "loss": 0.0449, "step": 21940 }, { "grad_norm": 0.3752073347568512, "learning_rate": 9.234965137925276e-05, "loss": 0.0463, "step": 21950 }, { "grad_norm": 0.48752138018608093, "learning_rate": 9.234085915210108e-05, "loss": 0.0423, "step": 21960 }, { "grad_norm": 0.4037015736103058, "learning_rate": 9.23320622946167e-05, "loss": 0.0376, "step": 21970 }, { "grad_norm": 0.3965141475200653, "learning_rate": 9.232326080776163e-05, "loss": 0.0425, "step": 21980 }, { "grad_norm": 0.40717393159866333, "learning_rate": 9.23144546924984e-05, "loss": 0.0396, "step": 21990 }, { "grad_norm": 0.3740255832672119, "learning_rate": 9.230564394979e-05, "loss": 0.0393, "step": 22000 }, { "grad_norm": 0.4673015773296356, "learning_rate": 9.22968285806e-05, "loss": 0.0416, "step": 22010 }, { "grad_norm": 0.4443141520023346, "learning_rate": 9.228800858589242e-05, "loss": 0.0501, "step": 22020 }, { "grad_norm": 0.40311914682388306, "learning_rate": 9.227918396663179e-05, "loss": 0.0508, "step": 22030 }, { "grad_norm": 0.42407798767089844, "learning_rate": 9.227035472378319e-05, "loss": 0.0436, "step": 22040 }, { "grad_norm": 0.3584294021129608, "learning_rate": 9.226152085831213e-05, "loss": 0.043, "step": 22050 }, { "grad_norm": 0.5179896950721741, "learning_rate": 9.22526823711847e-05, "loss": 0.0491, "step": 22060 }, { "grad_norm": 0.37852713465690613, "learning_rate": 9.224383926336745e-05, "loss": 0.0399, "step": 22070 }, { "grad_norm": 0.3889404833316803, "learning_rate": 9.223499153582744e-05, "loss": 0.041, "step": 22080 }, { "grad_norm": 0.49469414353370667, "learning_rate": 9.222613918953226e-05, "loss": 0.0542, "step": 22090 }, { "grad_norm": 0.53215092420578, "learning_rate": 9.221728222544999e-05, "loss": 0.0508, "step": 22100 }, { "grad_norm": 0.5453378558158875, "learning_rate": 9.22084206445492e-05, "loss": 0.0581, "step": 22110 }, { "grad_norm": 0.4223133325576782, "learning_rate": 9.2199554447799e-05, "loss": 0.0603, "step": 22120 }, { "grad_norm": 0.3518027067184448, "learning_rate": 9.219068363616897e-05, "loss": 0.0476, "step": 22130 }, { "grad_norm": 0.3750208914279938, "learning_rate": 9.218180821062919e-05, "loss": 0.0454, "step": 22140 }, { "grad_norm": 0.41633397340774536, "learning_rate": 9.21729281721503e-05, "loss": 0.0504, "step": 22150 }, { "grad_norm": 0.4346683919429779, "learning_rate": 9.216404352170339e-05, "loss": 0.0485, "step": 22160 }, { "grad_norm": 0.3386438488960266, "learning_rate": 9.215515426026007e-05, "loss": 0.0445, "step": 22170 }, { "grad_norm": 0.34335410594940186, "learning_rate": 9.214626038879246e-05, "loss": 0.045, "step": 22180 }, { "grad_norm": 0.3145151138305664, "learning_rate": 9.21373619082732e-05, "loss": 0.0463, "step": 22190 }, { "grad_norm": 0.48254311084747314, "learning_rate": 9.212845881967535e-05, "loss": 0.0462, "step": 22200 }, { "grad_norm": 0.608079195022583, "learning_rate": 9.211955112397262e-05, "loss": 0.0469, "step": 22210 }, { "grad_norm": 0.4791242480278015, "learning_rate": 9.211063882213909e-05, "loss": 0.041, "step": 22220 }, { "grad_norm": 0.41469770669937134, "learning_rate": 9.210172191514942e-05, "loss": 0.0395, "step": 22230 }, { "grad_norm": 0.4857258200645447, "learning_rate": 9.209280040397874e-05, "loss": 0.047, "step": 22240 }, { "grad_norm": 0.36778587102890015, "learning_rate": 9.208387428960268e-05, "loss": 0.0341, "step": 22250 }, { "grad_norm": 0.36729753017425537, "learning_rate": 9.20749435729974e-05, "loss": 0.0442, "step": 22260 }, { "grad_norm": 0.455267995595932, "learning_rate": 9.206600825513957e-05, "loss": 0.0381, "step": 22270 }, { "grad_norm": 0.3939463794231415, "learning_rate": 9.20570683370063e-05, "loss": 0.0496, "step": 22280 }, { "grad_norm": 0.5125149488449097, "learning_rate": 9.204812381957528e-05, "loss": 0.0433, "step": 22290 }, { "grad_norm": 0.4800189435482025, "learning_rate": 9.203917470382465e-05, "loss": 0.0456, "step": 22300 }, { "grad_norm": 0.4456491470336914, "learning_rate": 9.203022099073309e-05, "loss": 0.0451, "step": 22310 }, { "grad_norm": 0.5392191410064697, "learning_rate": 9.202126268127976e-05, "loss": 0.0518, "step": 22320 }, { "grad_norm": 0.4129745066165924, "learning_rate": 9.20122997764443e-05, "loss": 0.0463, "step": 22330 }, { "grad_norm": 0.30676382780075073, "learning_rate": 9.200333227720692e-05, "loss": 0.0541, "step": 22340 }, { "grad_norm": 0.3874857723712921, "learning_rate": 9.199436018454826e-05, "loss": 0.0477, "step": 22350 }, { "grad_norm": 0.3157590925693512, "learning_rate": 9.198538349944952e-05, "loss": 0.0452, "step": 22360 }, { "grad_norm": 0.4997654855251312, "learning_rate": 9.197640222289234e-05, "loss": 0.0662, "step": 22370 }, { "grad_norm": 0.5258808135986328, "learning_rate": 9.196741635585895e-05, "loss": 0.0524, "step": 22380 }, { "grad_norm": 0.35435736179351807, "learning_rate": 9.195842589933199e-05, "loss": 0.0444, "step": 22390 }, { "grad_norm": 0.37457701563835144, "learning_rate": 9.194943085429466e-05, "loss": 0.0496, "step": 22400 }, { "grad_norm": 0.3882056176662445, "learning_rate": 9.194043122173065e-05, "loss": 0.0612, "step": 22410 }, { "grad_norm": 0.5631015300750732, "learning_rate": 9.193142700262413e-05, "loss": 0.0503, "step": 22420 }, { "grad_norm": 0.4808194935321808, "learning_rate": 9.192241819795979e-05, "loss": 0.0395, "step": 22430 }, { "grad_norm": 0.5536881685256958, "learning_rate": 9.191340480872284e-05, "loss": 0.0552, "step": 22440 }, { "grad_norm": 0.43403360247612, "learning_rate": 9.190438683589895e-05, "loss": 0.0465, "step": 22450 }, { "grad_norm": 0.43481406569480896, "learning_rate": 9.189536428047432e-05, "loss": 0.0504, "step": 22460 }, { "grad_norm": 0.44854533672332764, "learning_rate": 9.188633714343564e-05, "loss": 0.0597, "step": 22470 }, { "grad_norm": 0.6313342452049255, "learning_rate": 9.18773054257701e-05, "loss": 0.0535, "step": 22480 }, { "grad_norm": 0.5838398933410645, "learning_rate": 9.18682691284654e-05, "loss": 0.0526, "step": 22490 }, { "grad_norm": 0.5221354961395264, "learning_rate": 9.185922825250974e-05, "loss": 0.046, "step": 22500 }, { "grad_norm": 0.43017712235450745, "learning_rate": 9.185018279889181e-05, "loss": 0.0407, "step": 22510 }, { "grad_norm": 0.7190511226654053, "learning_rate": 9.184113276860082e-05, "loss": 0.0505, "step": 22520 }, { "grad_norm": 0.3660466969013214, "learning_rate": 9.183207816262645e-05, "loss": 0.0485, "step": 22530 }, { "grad_norm": 0.3590860366821289, "learning_rate": 9.182301898195891e-05, "loss": 0.0337, "step": 22540 }, { "grad_norm": 0.23143820464611053, "learning_rate": 9.181395522758889e-05, "loss": 0.0494, "step": 22550 }, { "grad_norm": 0.5156787633895874, "learning_rate": 9.180488690050759e-05, "loss": 0.0531, "step": 22560 }, { "grad_norm": 0.3549990952014923, "learning_rate": 9.179581400170671e-05, "loss": 0.0409, "step": 22570 }, { "grad_norm": 0.33043521642684937, "learning_rate": 9.178673653217845e-05, "loss": 0.0414, "step": 22580 }, { "grad_norm": 0.3960343599319458, "learning_rate": 9.177765449291551e-05, "loss": 0.0384, "step": 22590 }, { "grad_norm": 0.3247123062610626, "learning_rate": 9.176856788491109e-05, "loss": 0.0424, "step": 22600 }, { "grad_norm": 0.3595263361930847, "learning_rate": 9.175947670915887e-05, "loss": 0.0434, "step": 22610 }, { "grad_norm": 0.4006330668926239, "learning_rate": 9.175038096665309e-05, "loss": 0.039, "step": 22620 }, { "grad_norm": 0.4250519275665283, "learning_rate": 9.17412806583884e-05, "loss": 0.0436, "step": 22630 }, { "grad_norm": 0.45626333355903625, "learning_rate": 9.173217578536002e-05, "loss": 0.0574, "step": 22640 }, { "grad_norm": 0.42525091767311096, "learning_rate": 9.172306634856362e-05, "loss": 0.0425, "step": 22650 }, { "grad_norm": 0.550288200378418, "learning_rate": 9.171395234899545e-05, "loss": 0.0452, "step": 22660 }, { "grad_norm": 0.35432589054107666, "learning_rate": 9.170483378765214e-05, "loss": 0.0436, "step": 22670 }, { "grad_norm": 0.35754454135894775, "learning_rate": 9.169571066553091e-05, "loss": 0.0422, "step": 22680 }, { "grad_norm": 0.420322984457016, "learning_rate": 9.168658298362946e-05, "loss": 0.0409, "step": 22690 }, { "grad_norm": 0.5084356069564819, "learning_rate": 9.167745074294598e-05, "loss": 0.0392, "step": 22700 }, { "grad_norm": 0.3951452374458313, "learning_rate": 9.166831394447913e-05, "loss": 0.0512, "step": 22710 }, { "grad_norm": 0.4032508432865143, "learning_rate": 9.165917258922812e-05, "loss": 0.0375, "step": 22720 }, { "grad_norm": 0.6906077265739441, "learning_rate": 9.165002667819262e-05, "loss": 0.0536, "step": 22730 }, { "grad_norm": 0.3653271794319153, "learning_rate": 9.164087621237282e-05, "loss": 0.048, "step": 22740 }, { "grad_norm": 0.4401998817920685, "learning_rate": 9.163172119276942e-05, "loss": 0.0508, "step": 22750 }, { "grad_norm": 0.39135316014289856, "learning_rate": 9.162256162038358e-05, "loss": 0.0563, "step": 22760 }, { "grad_norm": 0.41740870475769043, "learning_rate": 9.161339749621698e-05, "loss": 0.0421, "step": 22770 }, { "grad_norm": 0.4431464970111847, "learning_rate": 9.160422882127177e-05, "loss": 0.0432, "step": 22780 }, { "grad_norm": 0.4982251524925232, "learning_rate": 9.159505559655069e-05, "loss": 0.0391, "step": 22790 }, { "grad_norm": 0.4444688558578491, "learning_rate": 9.158587782305684e-05, "loss": 0.046, "step": 22800 }, { "grad_norm": 0.3037234842777252, "learning_rate": 9.157669550179391e-05, "loss": 0.0474, "step": 22810 }, { "grad_norm": 0.41825708746910095, "learning_rate": 9.156750863376609e-05, "loss": 0.0531, "step": 22820 }, { "grad_norm": 0.4855029881000519, "learning_rate": 9.155831721997801e-05, "loss": 0.0364, "step": 22830 }, { "grad_norm": 0.5072030425071716, "learning_rate": 9.154912126143484e-05, "loss": 0.0569, "step": 22840 }, { "grad_norm": 0.512035608291626, "learning_rate": 9.153992075914224e-05, "loss": 0.0584, "step": 22850 }, { "grad_norm": 0.3283541798591614, "learning_rate": 9.153071571410635e-05, "loss": 0.0359, "step": 22860 }, { "grad_norm": 0.4648641049861908, "learning_rate": 9.152150612733384e-05, "loss": 0.04, "step": 22870 }, { "grad_norm": 0.42712193727493286, "learning_rate": 9.151229199983184e-05, "loss": 0.0539, "step": 22880 }, { "grad_norm": 0.3695993721485138, "learning_rate": 9.150307333260802e-05, "loss": 0.0424, "step": 22890 }, { "grad_norm": 0.47884711623191833, "learning_rate": 9.149385012667048e-05, "loss": 0.0647, "step": 22900 }, { "grad_norm": 0.6188709139823914, "learning_rate": 9.148462238302788e-05, "loss": 0.0547, "step": 22910 }, { "grad_norm": 0.36513468623161316, "learning_rate": 9.147539010268936e-05, "loss": 0.0486, "step": 22920 }, { "grad_norm": 0.3860553205013275, "learning_rate": 9.14661532866645e-05, "loss": 0.0427, "step": 22930 }, { "grad_norm": 0.5156596899032593, "learning_rate": 9.145691193596348e-05, "loss": 0.0411, "step": 22940 }, { "grad_norm": 0.43236708641052246, "learning_rate": 9.144766605159691e-05, "loss": 0.0537, "step": 22950 }, { "grad_norm": 0.384407639503479, "learning_rate": 9.14384156345759e-05, "loss": 0.0393, "step": 22960 }, { "grad_norm": 0.43411901593208313, "learning_rate": 9.142916068591204e-05, "loss": 0.0359, "step": 22970 }, { "grad_norm": 0.41026222705841064, "learning_rate": 9.141990120661746e-05, "loss": 0.0456, "step": 22980 }, { "grad_norm": 0.4847211241722107, "learning_rate": 9.141063719770475e-05, "loss": 0.0461, "step": 22990 }, { "grad_norm": 0.4472845196723938, "learning_rate": 9.140136866018704e-05, "loss": 0.0408, "step": 23000 }, { "grad_norm": 0.23093435168266296, "learning_rate": 9.139209559507788e-05, "loss": 0.0387, "step": 23010 }, { "grad_norm": 0.3909669816493988, "learning_rate": 9.13828180033914e-05, "loss": 0.0409, "step": 23020 }, { "grad_norm": 0.36558258533477783, "learning_rate": 9.137353588614212e-05, "loss": 0.0477, "step": 23030 }, { "grad_norm": 0.422938734292984, "learning_rate": 9.136424924434519e-05, "loss": 0.0386, "step": 23040 }, { "grad_norm": 0.41022181510925293, "learning_rate": 9.135495807901615e-05, "loss": 0.0478, "step": 23050 }, { "grad_norm": 0.37920257449150085, "learning_rate": 9.134566239117108e-05, "loss": 0.039, "step": 23060 }, { "grad_norm": 0.4991760551929474, "learning_rate": 9.13363621818265e-05, "loss": 0.0403, "step": 23070 }, { "grad_norm": 0.4265248775482178, "learning_rate": 9.132705745199953e-05, "loss": 0.0378, "step": 23080 }, { "grad_norm": 0.355305939912796, "learning_rate": 9.131774820270768e-05, "loss": 0.0366, "step": 23090 }, { "grad_norm": 0.41440269351005554, "learning_rate": 9.130843443496901e-05, "loss": 0.0406, "step": 23100 }, { "grad_norm": 0.2623721659183502, "learning_rate": 9.129911614980206e-05, "loss": 0.054, "step": 23110 }, { "grad_norm": 0.482211172580719, "learning_rate": 9.128979334822584e-05, "loss": 0.0398, "step": 23120 }, { "grad_norm": 0.30777519941329956, "learning_rate": 9.128046603125992e-05, "loss": 0.0338, "step": 23130 }, { "grad_norm": 0.33705949783325195, "learning_rate": 9.12711341999243e-05, "loss": 0.035, "step": 23140 }, { "grad_norm": 0.46903082728385925, "learning_rate": 9.12617978552395e-05, "loss": 0.059, "step": 23150 }, { "grad_norm": 0.3423120379447937, "learning_rate": 9.12524569982265e-05, "loss": 0.0426, "step": 23160 }, { "grad_norm": 0.5067928433418274, "learning_rate": 9.124311162990684e-05, "loss": 0.0429, "step": 23170 }, { "grad_norm": 0.6044545769691467, "learning_rate": 9.12337617513025e-05, "loss": 0.0397, "step": 23180 }, { "grad_norm": 0.3404444754123688, "learning_rate": 9.122440736343596e-05, "loss": 0.0394, "step": 23190 }, { "grad_norm": 0.3936929702758789, "learning_rate": 9.12150484673302e-05, "loss": 0.0421, "step": 23200 }, { "grad_norm": 0.4340112805366516, "learning_rate": 9.120568506400873e-05, "loss": 0.0363, "step": 23210 }, { "grad_norm": 0.5177788138389587, "learning_rate": 9.119631715449548e-05, "loss": 0.0364, "step": 23220 }, { "grad_norm": 0.40514039993286133, "learning_rate": 9.118694473981493e-05, "loss": 0.0409, "step": 23230 }, { "grad_norm": 0.4836953282356262, "learning_rate": 9.117756782099203e-05, "loss": 0.0434, "step": 23240 }, { "grad_norm": 0.38632532954216003, "learning_rate": 9.11681863990522e-05, "loss": 0.051, "step": 23250 }, { "grad_norm": 0.5758042335510254, "learning_rate": 9.115880047502142e-05, "loss": 0.0436, "step": 23260 }, { "grad_norm": 0.6220723390579224, "learning_rate": 9.114941004992609e-05, "loss": 0.0507, "step": 23270 }, { "grad_norm": 0.5320327281951904, "learning_rate": 9.114001512479317e-05, "loss": 0.0515, "step": 23280 }, { "grad_norm": 0.4520535469055176, "learning_rate": 9.113061570065003e-05, "loss": 0.0658, "step": 23290 }, { "grad_norm": 0.5256597399711609, "learning_rate": 9.112121177852459e-05, "loss": 0.0518, "step": 23300 }, { "grad_norm": 0.4984566867351532, "learning_rate": 9.111180335944527e-05, "loss": 0.0489, "step": 23310 }, { "grad_norm": 0.5993790030479431, "learning_rate": 9.110239044444093e-05, "loss": 0.0449, "step": 23320 }, { "grad_norm": 0.301161527633667, "learning_rate": 9.109297303454099e-05, "loss": 0.0416, "step": 23330 }, { "grad_norm": 0.37548860907554626, "learning_rate": 9.108355113077526e-05, "loss": 0.0465, "step": 23340 }, { "grad_norm": 0.35403671860694885, "learning_rate": 9.107412473417419e-05, "loss": 0.0429, "step": 23350 }, { "grad_norm": 0.45469555258750916, "learning_rate": 9.106469384576858e-05, "loss": 0.0398, "step": 23360 }, { "grad_norm": 0.3373114764690399, "learning_rate": 9.105525846658978e-05, "loss": 0.045, "step": 23370 }, { "grad_norm": 0.4421926736831665, "learning_rate": 9.104581859766965e-05, "loss": 0.0501, "step": 23380 }, { "grad_norm": 0.6202095746994019, "learning_rate": 9.10363742400405e-05, "loss": 0.0554, "step": 23390 }, { "grad_norm": 0.3096199035644531, "learning_rate": 9.102692539473518e-05, "loss": 0.0435, "step": 23400 }, { "grad_norm": 0.4001771807670593, "learning_rate": 9.101747206278697e-05, "loss": 0.0512, "step": 23410 }, { "grad_norm": 0.4310392141342163, "learning_rate": 9.100801424522968e-05, "loss": 0.0357, "step": 23420 }, { "grad_norm": 0.6682950258255005, "learning_rate": 9.099855194309762e-05, "loss": 0.0484, "step": 23430 }, { "grad_norm": 0.4341863989830017, "learning_rate": 9.098908515742554e-05, "loss": 0.0364, "step": 23440 }, { "grad_norm": 0.4230785071849823, "learning_rate": 9.097961388924873e-05, "loss": 0.0408, "step": 23450 }, { "grad_norm": 0.818611204624176, "learning_rate": 9.097013813960298e-05, "loss": 0.0412, "step": 23460 }, { "grad_norm": 0.4314897656440735, "learning_rate": 9.09606579095245e-05, "loss": 0.0451, "step": 23470 }, { "grad_norm": 0.3302779495716095, "learning_rate": 9.095117320005008e-05, "loss": 0.0397, "step": 23480 }, { "grad_norm": 0.45899730920791626, "learning_rate": 9.094168401221691e-05, "loss": 0.0527, "step": 23490 }, { "grad_norm": 0.4204159677028656, "learning_rate": 9.093219034706273e-05, "loss": 0.0595, "step": 23500 }, { "grad_norm": 0.4751824140548706, "learning_rate": 9.092269220562577e-05, "loss": 0.0482, "step": 23510 }, { "grad_norm": 0.47840234637260437, "learning_rate": 9.09131895889447e-05, "loss": 0.0631, "step": 23520 }, { "grad_norm": 0.40165969729423523, "learning_rate": 9.090368249805873e-05, "loss": 0.0434, "step": 23530 }, { "grad_norm": 0.3326484262943268, "learning_rate": 9.089417093400754e-05, "loss": 0.0456, "step": 23540 }, { "grad_norm": 0.3955256938934326, "learning_rate": 9.088465489783131e-05, "loss": 0.0426, "step": 23550 }, { "grad_norm": 0.38724657893180847, "learning_rate": 9.087513439057068e-05, "loss": 0.0441, "step": 23560 }, { "grad_norm": 0.31611818075180054, "learning_rate": 9.08656094132668e-05, "loss": 0.0429, "step": 23570 }, { "grad_norm": 0.4221184551715851, "learning_rate": 9.085607996696134e-05, "loss": 0.0348, "step": 23580 }, { "grad_norm": 0.39806488156318665, "learning_rate": 9.084654605269639e-05, "loss": 0.0536, "step": 23590 }, { "grad_norm": 0.4511125981807709, "learning_rate": 9.083700767151457e-05, "loss": 0.0429, "step": 23600 }, { "grad_norm": 0.43291646242141724, "learning_rate": 9.082746482445898e-05, "loss": 0.0524, "step": 23610 }, { "grad_norm": 0.343378484249115, "learning_rate": 9.081791751257325e-05, "loss": 0.0412, "step": 23620 }, { "grad_norm": 0.3675134479999542, "learning_rate": 9.080836573690142e-05, "loss": 0.0404, "step": 23630 }, { "grad_norm": 0.47053253650665283, "learning_rate": 9.079880949848805e-05, "loss": 0.0526, "step": 23640 }, { "grad_norm": 0.7887464165687561, "learning_rate": 9.078924879837822e-05, "loss": 0.0517, "step": 23650 }, { "grad_norm": 0.3767990171909332, "learning_rate": 9.077968363761747e-05, "loss": 0.0583, "step": 23660 }, { "grad_norm": 0.38673877716064453, "learning_rate": 9.077011401725182e-05, "loss": 0.0448, "step": 23670 }, { "grad_norm": 0.4160268306732178, "learning_rate": 9.07605399383278e-05, "loss": 0.0572, "step": 23680 }, { "grad_norm": 0.32587915658950806, "learning_rate": 9.075096140189243e-05, "loss": 0.0402, "step": 23690 }, { "grad_norm": 0.37719473242759705, "learning_rate": 9.074137840899318e-05, "loss": 0.0468, "step": 23700 }, { "grad_norm": 0.30495116114616394, "learning_rate": 9.073179096067804e-05, "loss": 0.043, "step": 23710 }, { "grad_norm": 0.3662429749965668, "learning_rate": 9.072219905799549e-05, "loss": 0.0383, "step": 23720 }, { "grad_norm": 0.4581556022167206, "learning_rate": 9.071260270199447e-05, "loss": 0.0505, "step": 23730 }, { "grad_norm": 0.9440050721168518, "learning_rate": 9.070300189372441e-05, "loss": 0.0433, "step": 23740 }, { "grad_norm": 0.33936089277267456, "learning_rate": 9.069339663423528e-05, "loss": 0.0479, "step": 23750 }, { "grad_norm": 0.4114910364151001, "learning_rate": 9.068378692457747e-05, "loss": 0.0558, "step": 23760 }, { "grad_norm": 0.41933587193489075, "learning_rate": 9.067417276580189e-05, "loss": 0.0434, "step": 23770 }, { "grad_norm": 0.6116280555725098, "learning_rate": 9.066455415895993e-05, "loss": 0.0485, "step": 23780 }, { "grad_norm": 0.4308302104473114, "learning_rate": 9.065493110510346e-05, "loss": 0.0471, "step": 23790 }, { "grad_norm": 0.4487057328224182, "learning_rate": 9.064530360528484e-05, "loss": 0.0433, "step": 23800 }, { "grad_norm": 0.5166841149330139, "learning_rate": 9.063567166055695e-05, "loss": 0.0383, "step": 23810 }, { "grad_norm": 0.43306443095207214, "learning_rate": 9.062603527197308e-05, "loss": 0.0401, "step": 23820 }, { "grad_norm": 0.47215983271598816, "learning_rate": 9.06163944405871e-05, "loss": 0.0461, "step": 23830 }, { "grad_norm": 0.4741976261138916, "learning_rate": 9.060674916745327e-05, "loss": 0.0539, "step": 23840 }, { "grad_norm": 0.5313030481338501, "learning_rate": 9.05970994536264e-05, "loss": 0.0494, "step": 23850 }, { "grad_norm": 0.452019602060318, "learning_rate": 9.05874453001618e-05, "loss": 0.047, "step": 23860 }, { "grad_norm": 0.5696617364883423, "learning_rate": 9.057778670811517e-05, "loss": 0.0578, "step": 23870 }, { "grad_norm": 0.4211365878582001, "learning_rate": 9.056812367854281e-05, "loss": 0.0414, "step": 23880 }, { "grad_norm": 0.42561715841293335, "learning_rate": 9.055845621250143e-05, "loss": 0.0557, "step": 23890 }, { "grad_norm": 0.3519609570503235, "learning_rate": 9.054878431104825e-05, "loss": 0.0382, "step": 23900 }, { "grad_norm": 0.42458945512771606, "learning_rate": 9.0539107975241e-05, "loss": 0.0455, "step": 23910 }, { "grad_norm": 0.48449036478996277, "learning_rate": 9.052942720613784e-05, "loss": 0.0531, "step": 23920 }, { "grad_norm": 0.2839517593383789, "learning_rate": 9.051974200479745e-05, "loss": 0.0413, "step": 23930 }, { "grad_norm": 0.3974880874156952, "learning_rate": 9.051005237227901e-05, "loss": 0.0393, "step": 23940 }, { "grad_norm": 0.4070758819580078, "learning_rate": 9.050035830964215e-05, "loss": 0.035, "step": 23950 }, { "grad_norm": 0.8090533018112183, "learning_rate": 9.049065981794698e-05, "loss": 0.0557, "step": 23960 }, { "grad_norm": 0.3291832506656647, "learning_rate": 9.048095689825414e-05, "loss": 0.0404, "step": 23970 }, { "grad_norm": 0.43539613485336304, "learning_rate": 9.047124955162472e-05, "loss": 0.047, "step": 23980 }, { "grad_norm": 0.4037327468395233, "learning_rate": 9.046153777912028e-05, "loss": 0.047, "step": 23990 }, { "grad_norm": 0.39197537302970886, "learning_rate": 9.045182158180292e-05, "loss": 0.0428, "step": 24000 }, { "grad_norm": 0.3671626150608063, "learning_rate": 9.044210096073516e-05, "loss": 0.0352, "step": 24010 }, { "grad_norm": 0.3287736177444458, "learning_rate": 9.043237591698004e-05, "loss": 0.034, "step": 24020 }, { "grad_norm": 0.44656237959861755, "learning_rate": 9.04226464516011e-05, "loss": 0.0386, "step": 24030 }, { "grad_norm": 0.5393972396850586, "learning_rate": 9.041291256566229e-05, "loss": 0.0488, "step": 24040 }, { "grad_norm": 0.5115112662315369, "learning_rate": 9.040317426022814e-05, "loss": 0.0508, "step": 24050 }, { "grad_norm": 0.2644052803516388, "learning_rate": 9.03934315363636e-05, "loss": 0.0435, "step": 24060 }, { "grad_norm": 0.28250980377197266, "learning_rate": 9.038368439513409e-05, "loss": 0.0334, "step": 24070 }, { "grad_norm": 0.641821026802063, "learning_rate": 9.03739328376056e-05, "loss": 0.0502, "step": 24080 }, { "grad_norm": 0.5378654599189758, "learning_rate": 9.036417686484451e-05, "loss": 0.0484, "step": 24090 }, { "grad_norm": 0.41471272706985474, "learning_rate": 9.035441647791773e-05, "loss": 0.0504, "step": 24100 }, { "grad_norm": 0.46547457575798035, "learning_rate": 9.034465167789263e-05, "loss": 0.041, "step": 24110 }, { "grad_norm": 0.5075207948684692, "learning_rate": 9.033488246583706e-05, "loss": 0.042, "step": 24120 }, { "grad_norm": 0.3566848039627075, "learning_rate": 9.032510884281941e-05, "loss": 0.0567, "step": 24130 }, { "grad_norm": 0.43982675671577454, "learning_rate": 9.031533080990848e-05, "loss": 0.0474, "step": 24140 }, { "grad_norm": 0.3721524775028229, "learning_rate": 9.030554836817358e-05, "loss": 0.0403, "step": 24150 }, { "grad_norm": 0.6227020025253296, "learning_rate": 9.029576151868451e-05, "loss": 0.0528, "step": 24160 }, { "grad_norm": 0.3685283958911896, "learning_rate": 9.028597026251155e-05, "loss": 0.0374, "step": 24170 }, { "grad_norm": 0.39254096150398254, "learning_rate": 9.027617460072547e-05, "loss": 0.0461, "step": 24180 }, { "grad_norm": 0.4000247120857239, "learning_rate": 9.026637453439745e-05, "loss": 0.0382, "step": 24190 }, { "grad_norm": 0.35176077485084534, "learning_rate": 9.025657006459927e-05, "loss": 0.0376, "step": 24200 }, { "grad_norm": 0.5436599254608154, "learning_rate": 9.024676119240311e-05, "loss": 0.0377, "step": 24210 }, { "grad_norm": 0.3851430416107178, "learning_rate": 9.023694791888166e-05, "loss": 0.0476, "step": 24220 }, { "grad_norm": 0.3624439239501953, "learning_rate": 9.022713024510808e-05, "loss": 0.0457, "step": 24230 }, { "grad_norm": 0.3161971867084503, "learning_rate": 9.021730817215601e-05, "loss": 0.041, "step": 24240 }, { "grad_norm": 0.40661555528640747, "learning_rate": 9.02074817010996e-05, "loss": 0.0406, "step": 24250 }, { "grad_norm": 0.37299537658691406, "learning_rate": 9.019765083301342e-05, "loss": 0.0302, "step": 24260 }, { "grad_norm": 0.458796888589859, "learning_rate": 9.01878155689726e-05, "loss": 0.0443, "step": 24270 }, { "grad_norm": 0.37663546204566956, "learning_rate": 9.017797591005268e-05, "loss": 0.0375, "step": 24280 }, { "grad_norm": 0.4842454195022583, "learning_rate": 9.016813185732972e-05, "loss": 0.0548, "step": 24290 }, { "grad_norm": 0.3668421506881714, "learning_rate": 9.015828341188027e-05, "loss": 0.0439, "step": 24300 }, { "grad_norm": 0.5597903728485107, "learning_rate": 9.01484305747813e-05, "loss": 0.044, "step": 24310 }, { "grad_norm": 0.5482326149940491, "learning_rate": 9.013857334711033e-05, "loss": 0.0408, "step": 24320 }, { "grad_norm": 0.36669692397117615, "learning_rate": 9.012871172994534e-05, "loss": 0.0314, "step": 24330 }, { "grad_norm": 0.28551995754241943, "learning_rate": 9.011884572436476e-05, "loss": 0.0444, "step": 24340 }, { "grad_norm": 0.409342497587204, "learning_rate": 9.010897533144754e-05, "loss": 0.0521, "step": 24350 }, { "grad_norm": 0.5537530183792114, "learning_rate": 9.009910055227306e-05, "loss": 0.0493, "step": 24360 }, { "grad_norm": 0.33565008640289307, "learning_rate": 9.008922138792124e-05, "loss": 0.0393, "step": 24370 }, { "grad_norm": 0.7969650030136108, "learning_rate": 9.007933783947244e-05, "loss": 0.0443, "step": 24380 }, { "grad_norm": 0.3826882541179657, "learning_rate": 9.006944990800752e-05, "loss": 0.0395, "step": 24390 }, { "grad_norm": 0.46345382928848267, "learning_rate": 9.005955759460779e-05, "loss": 0.0498, "step": 24400 }, { "grad_norm": 0.4957829713821411, "learning_rate": 9.004966090035508e-05, "loss": 0.0465, "step": 24410 }, { "grad_norm": 0.40217798948287964, "learning_rate": 9.003975982633166e-05, "loss": 0.041, "step": 24420 }, { "grad_norm": 0.39646807312965393, "learning_rate": 9.00298543736203e-05, "loss": 0.0416, "step": 24430 }, { "grad_norm": 0.33727240562438965, "learning_rate": 9.001994454330427e-05, "loss": 0.0371, "step": 24440 }, { "grad_norm": 0.942106306552887, "learning_rate": 9.001003033646727e-05, "loss": 0.046, "step": 24450 }, { "grad_norm": 0.4414464235305786, "learning_rate": 9.00001117541935e-05, "loss": 0.0461, "step": 24460 }, { "grad_norm": 0.49247241020202637, "learning_rate": 8.999018879756764e-05, "loss": 0.0517, "step": 24470 }, { "grad_norm": 0.4733559191226959, "learning_rate": 8.998026146767487e-05, "loss": 0.0457, "step": 24480 }, { "grad_norm": 0.42911237478256226, "learning_rate": 8.99703297656008e-05, "loss": 0.047, "step": 24490 }, { "grad_norm": 0.5687592029571533, "learning_rate": 8.996039369243156e-05, "loss": 0.0538, "step": 24500 }, { "grad_norm": 0.35348042845726013, "learning_rate": 8.995045324925378e-05, "loss": 0.0361, "step": 24510 }, { "grad_norm": 0.620530903339386, "learning_rate": 8.994050843715448e-05, "loss": 0.0553, "step": 24520 }, { "grad_norm": 0.4303775727748871, "learning_rate": 8.993055925722121e-05, "loss": 0.0384, "step": 24530 }, { "grad_norm": 0.37951284646987915, "learning_rate": 8.992060571054202e-05, "loss": 0.0334, "step": 24540 }, { "grad_norm": 0.4638354182243347, "learning_rate": 8.991064779820542e-05, "loss": 0.0484, "step": 24550 }, { "grad_norm": 0.34324637055397034, "learning_rate": 8.990068552130036e-05, "loss": 0.0355, "step": 24560 }, { "grad_norm": 0.5035029649734497, "learning_rate": 8.989071888091634e-05, "loss": 0.0423, "step": 24570 }, { "grad_norm": 0.5085214376449585, "learning_rate": 8.988074787814329e-05, "loss": 0.0472, "step": 24580 }, { "grad_norm": 0.45427197217941284, "learning_rate": 8.987077251407158e-05, "loss": 0.0368, "step": 24590 }, { "grad_norm": 0.4599609673023224, "learning_rate": 8.986079278979216e-05, "loss": 0.0392, "step": 24600 }, { "grad_norm": 0.3000871241092682, "learning_rate": 8.985080870639635e-05, "loss": 0.0472, "step": 24610 }, { "grad_norm": 0.44585129618644714, "learning_rate": 8.984082026497603e-05, "loss": 0.0399, "step": 24620 }, { "grad_norm": 0.3910902142524719, "learning_rate": 8.98308274666235e-05, "loss": 0.0598, "step": 24630 }, { "grad_norm": 0.4142967760562897, "learning_rate": 8.982083031243155e-05, "loss": 0.0462, "step": 24640 }, { "grad_norm": 0.3518674075603485, "learning_rate": 8.98108288034935e-05, "loss": 0.0422, "step": 24650 }, { "grad_norm": 0.5595763325691223, "learning_rate": 8.980082294090305e-05, "loss": 0.0488, "step": 24660 }, { "grad_norm": 0.3040735423564911, "learning_rate": 8.979081272575443e-05, "loss": 0.0486, "step": 24670 }, { "grad_norm": 0.4348416030406952, "learning_rate": 8.978079815914236e-05, "loss": 0.0447, "step": 24680 }, { "grad_norm": 0.44924068450927734, "learning_rate": 8.977077924216202e-05, "loss": 0.0389, "step": 24690 }, { "grad_norm": 0.39525070786476135, "learning_rate": 8.976075597590905e-05, "loss": 0.0418, "step": 24700 }, { "grad_norm": 0.5538489818572998, "learning_rate": 8.975072836147958e-05, "loss": 0.0495, "step": 24710 }, { "grad_norm": 0.5947695970535278, "learning_rate": 8.974069639997025e-05, "loss": 0.0376, "step": 24720 }, { "grad_norm": 0.7028986215591431, "learning_rate": 8.973066009247808e-05, "loss": 0.0405, "step": 24730 }, { "grad_norm": 0.43056362867355347, "learning_rate": 8.972061944010066e-05, "loss": 0.0504, "step": 24740 }, { "grad_norm": 0.2736057937145233, "learning_rate": 8.971057444393603e-05, "loss": 0.0429, "step": 24750 }, { "grad_norm": 0.35542020201683044, "learning_rate": 8.970052510508268e-05, "loss": 0.0538, "step": 24760 }, { "grad_norm": 0.38899827003479004, "learning_rate": 8.969047142463959e-05, "loss": 0.0466, "step": 24770 }, { "grad_norm": 0.3555007874965668, "learning_rate": 8.968041340370621e-05, "loss": 0.0428, "step": 24780 }, { "grad_norm": 0.8242965340614319, "learning_rate": 8.96703510433825e-05, "loss": 0.0452, "step": 24790 }, { "grad_norm": 0.33439844846725464, "learning_rate": 8.966028434476883e-05, "loss": 0.0355, "step": 24800 }, { "grad_norm": 0.4030182361602783, "learning_rate": 8.96502133089661e-05, "loss": 0.037, "step": 24810 }, { "grad_norm": 0.31297609210014343, "learning_rate": 8.964013793707564e-05, "loss": 0.0388, "step": 24820 }, { "grad_norm": 0.37857919931411743, "learning_rate": 8.963005823019932e-05, "loss": 0.0423, "step": 24830 }, { "grad_norm": 0.7103894948959351, "learning_rate": 8.961997418943939e-05, "loss": 0.0477, "step": 24840 }, { "grad_norm": 0.6372089982032776, "learning_rate": 8.960988581589865e-05, "loss": 0.0438, "step": 24850 }, { "grad_norm": 0.3956005871295929, "learning_rate": 8.959979311068037e-05, "loss": 0.0362, "step": 24860 }, { "grad_norm": 0.6288209557533264, "learning_rate": 8.958969607488823e-05, "loss": 0.059, "step": 24870 }, { "grad_norm": 0.45459240674972534, "learning_rate": 8.957959470962647e-05, "loss": 0.0392, "step": 24880 }, { "grad_norm": 0.5798158049583435, "learning_rate": 8.956948901599971e-05, "loss": 0.0401, "step": 24890 }, { "grad_norm": 0.6375553011894226, "learning_rate": 8.955937899511315e-05, "loss": 0.0461, "step": 24900 }, { "grad_norm": 0.5751432776451111, "learning_rate": 8.954926464807238e-05, "loss": 0.0479, "step": 24910 }, { "grad_norm": 0.4301179349422455, "learning_rate": 8.953914597598347e-05, "loss": 0.0395, "step": 24920 }, { "grad_norm": 0.3338702321052551, "learning_rate": 8.952902297995303e-05, "loss": 0.0413, "step": 24930 }, { "grad_norm": 0.5591727495193481, "learning_rate": 8.951889566108804e-05, "loss": 0.0484, "step": 24940 }, { "grad_norm": 0.38781917095184326, "learning_rate": 8.950876402049606e-05, "loss": 0.0419, "step": 24950 }, { "grad_norm": 0.40226393938064575, "learning_rate": 8.949862805928504e-05, "loss": 0.0432, "step": 24960 }, { "grad_norm": 0.5588801503181458, "learning_rate": 8.948848777856343e-05, "loss": 0.042, "step": 24970 }, { "grad_norm": 0.49063682556152344, "learning_rate": 8.947834317944017e-05, "loss": 0.0473, "step": 24980 }, { "grad_norm": 0.6741628646850586, "learning_rate": 8.946819426302466e-05, "loss": 0.0617, "step": 24990 }, { "grad_norm": 0.5930131077766418, "learning_rate": 8.945804103042676e-05, "loss": 0.0504, "step": 25000 }, { "grad_norm": 0.32255488634109497, "learning_rate": 8.944788348275681e-05, "loss": 0.0399, "step": 25010 }, { "grad_norm": 0.44118568301200867, "learning_rate": 8.943772162112565e-05, "loss": 0.0434, "step": 25020 }, { "grad_norm": 0.5280837416648865, "learning_rate": 8.942755544664454e-05, "loss": 0.0534, "step": 25030 }, { "grad_norm": 0.41719454526901245, "learning_rate": 8.941738496042525e-05, "loss": 0.0435, "step": 25040 }, { "grad_norm": 0.3020166754722595, "learning_rate": 8.940721016357999e-05, "loss": 0.0428, "step": 25050 }, { "grad_norm": 0.36082586646080017, "learning_rate": 8.939703105722148e-05, "loss": 0.0456, "step": 25060 }, { "grad_norm": 0.37813839316368103, "learning_rate": 8.93868476424629e-05, "loss": 0.0391, "step": 25070 }, { "grad_norm": 0.3423895835876465, "learning_rate": 8.937665992041786e-05, "loss": 0.0447, "step": 25080 }, { "grad_norm": 0.5435647964477539, "learning_rate": 8.93664678922005e-05, "loss": 0.0518, "step": 25090 }, { "grad_norm": 0.3369680941104889, "learning_rate": 8.93562715589254e-05, "loss": 0.0428, "step": 25100 }, { "grad_norm": 0.43208420276641846, "learning_rate": 8.934607092170762e-05, "loss": 0.0531, "step": 25110 }, { "grad_norm": 0.4635172188282013, "learning_rate": 8.933586598166266e-05, "loss": 0.0507, "step": 25120 }, { "grad_norm": 0.5305559039115906, "learning_rate": 8.932565673990655e-05, "loss": 0.0549, "step": 25130 }, { "grad_norm": 0.3983321785926819, "learning_rate": 8.931544319755574e-05, "loss": 0.0435, "step": 25140 }, { "grad_norm": 0.3827718198299408, "learning_rate": 8.930522535572718e-05, "loss": 0.041, "step": 25150 }, { "grad_norm": 0.33410125970840454, "learning_rate": 8.929500321553826e-05, "loss": 0.0432, "step": 25160 }, { "grad_norm": 0.36991551518440247, "learning_rate": 8.928477677810686e-05, "loss": 0.0362, "step": 25170 }, { "grad_norm": 0.33397406339645386, "learning_rate": 8.927454604455137e-05, "loss": 0.0382, "step": 25180 }, { "grad_norm": 1.095096468925476, "learning_rate": 8.926431101599053e-05, "loss": 0.0459, "step": 25190 }, { "grad_norm": 0.3634420335292816, "learning_rate": 8.925407169354369e-05, "loss": 0.0371, "step": 25200 }, { "grad_norm": 0.41213908791542053, "learning_rate": 8.92438280783306e-05, "loss": 0.0438, "step": 25210 }, { "grad_norm": 0.38099539279937744, "learning_rate": 8.923358017147146e-05, "loss": 0.0381, "step": 25220 }, { "grad_norm": 0.43922144174575806, "learning_rate": 8.922332797408697e-05, "loss": 0.0459, "step": 25230 }, { "grad_norm": 0.3027651011943817, "learning_rate": 8.921307148729831e-05, "loss": 0.0338, "step": 25240 }, { "grad_norm": 0.5762755870819092, "learning_rate": 8.920281071222712e-05, "loss": 0.0388, "step": 25250 }, { "grad_norm": 0.379039466381073, "learning_rate": 8.919254564999548e-05, "loss": 0.0416, "step": 25260 }, { "grad_norm": 0.5808419585227966, "learning_rate": 8.918227630172598e-05, "loss": 0.0512, "step": 25270 }, { "grad_norm": 0.3176271319389343, "learning_rate": 8.917200266854165e-05, "loss": 0.0399, "step": 25280 }, { "grad_norm": 0.48705950379371643, "learning_rate": 8.9161724751566e-05, "loss": 0.0482, "step": 25290 }, { "grad_norm": 0.4563620984554291, "learning_rate": 8.915144255192302e-05, "loss": 0.0534, "step": 25300 }, { "grad_norm": 0.34443947672843933, "learning_rate": 8.914115607073714e-05, "loss": 0.0491, "step": 25310 }, { "grad_norm": 0.5367611646652222, "learning_rate": 8.913086530913327e-05, "loss": 0.0382, "step": 25320 }, { "grad_norm": 0.37550824880599976, "learning_rate": 8.912057026823681e-05, "loss": 0.0499, "step": 25330 }, { "grad_norm": 0.5445125102996826, "learning_rate": 8.91102709491736e-05, "loss": 0.0391, "step": 25340 }, { "grad_norm": 0.5203923583030701, "learning_rate": 8.909996735306996e-05, "loss": 0.0424, "step": 25350 }, { "grad_norm": 0.4579969644546509, "learning_rate": 8.908965948105268e-05, "loss": 0.0512, "step": 25360 }, { "grad_norm": 0.47962677478790283, "learning_rate": 8.907934733424901e-05, "loss": 0.0452, "step": 25370 }, { "grad_norm": 0.5764922499656677, "learning_rate": 8.906903091378666e-05, "loss": 0.0524, "step": 25380 }, { "grad_norm": 0.4028795659542084, "learning_rate": 8.905871022079384e-05, "loss": 0.0407, "step": 25390 }, { "grad_norm": 0.36347997188568115, "learning_rate": 8.90483852563992e-05, "loss": 0.0391, "step": 25400 }, { "grad_norm": 0.4150656759738922, "learning_rate": 8.903805602173185e-05, "loss": 0.049, "step": 25410 }, { "grad_norm": 0.5162582993507385, "learning_rate": 8.902772251792137e-05, "loss": 0.0439, "step": 25420 }, { "grad_norm": 0.5413376688957214, "learning_rate": 8.901738474609786e-05, "loss": 0.0529, "step": 25430 }, { "grad_norm": 0.4074924886226654, "learning_rate": 8.900704270739179e-05, "loss": 0.0323, "step": 25440 }, { "grad_norm": 0.3160432279109955, "learning_rate": 8.89966964029342e-05, "loss": 0.0478, "step": 25450 }, { "grad_norm": 0.42403870820999146, "learning_rate": 8.898634583385652e-05, "loss": 0.0403, "step": 25460 }, { "grad_norm": 0.3653183579444885, "learning_rate": 8.897599100129065e-05, "loss": 0.0422, "step": 25470 }, { "grad_norm": 0.47801727056503296, "learning_rate": 8.896563190636903e-05, "loss": 0.036, "step": 25480 }, { "grad_norm": 0.5384616851806641, "learning_rate": 8.895526855022448e-05, "loss": 0.035, "step": 25490 }, { "grad_norm": 0.45273301005363464, "learning_rate": 8.894490093399033e-05, "loss": 0.0491, "step": 25500 }, { "grad_norm": 0.48018911480903625, "learning_rate": 8.893452905880035e-05, "loss": 0.0441, "step": 25510 }, { "grad_norm": 0.6321214437484741, "learning_rate": 8.892415292578883e-05, "loss": 0.0397, "step": 25520 }, { "grad_norm": 0.45256632566452026, "learning_rate": 8.891377253609046e-05, "loss": 0.0448, "step": 25530 }, { "grad_norm": 0.5265847444534302, "learning_rate": 8.890338789084043e-05, "loss": 0.0423, "step": 25540 }, { "grad_norm": 0.4555388391017914, "learning_rate": 8.88929989911744e-05, "loss": 0.0435, "step": 25550 }, { "grad_norm": 0.45402851700782776, "learning_rate": 8.888260583822847e-05, "loss": 0.0418, "step": 25560 }, { "grad_norm": 0.4587850868701935, "learning_rate": 8.887220843313921e-05, "loss": 0.0438, "step": 25570 }, { "grad_norm": 0.5283277630805969, "learning_rate": 8.88618067770437e-05, "loss": 0.056, "step": 25580 }, { "grad_norm": 0.41148659586906433, "learning_rate": 8.885140087107942e-05, "loss": 0.0454, "step": 25590 }, { "grad_norm": 0.49949824810028076, "learning_rate": 8.884099071638436e-05, "loss": 0.0447, "step": 25600 }, { "grad_norm": 0.46889030933380127, "learning_rate": 8.883057631409695e-05, "loss": 0.0463, "step": 25610 }, { "grad_norm": 0.512624204158783, "learning_rate": 8.882015766535608e-05, "loss": 0.0455, "step": 25620 }, { "grad_norm": 0.4255642890930176, "learning_rate": 8.880973477130115e-05, "loss": 0.0469, "step": 25630 }, { "grad_norm": 0.4895669221878052, "learning_rate": 8.879930763307197e-05, "loss": 0.0454, "step": 25640 }, { "grad_norm": 0.7983567714691162, "learning_rate": 8.878887625180884e-05, "loss": 0.0473, "step": 25650 }, { "grad_norm": 0.3190539479255676, "learning_rate": 8.877844062865253e-05, "loss": 0.0323, "step": 25660 }, { "grad_norm": 0.3377387225627899, "learning_rate": 8.876800076474424e-05, "loss": 0.0512, "step": 25670 }, { "grad_norm": 0.47088107466697693, "learning_rate": 8.875755666122568e-05, "loss": 0.0345, "step": 25680 }, { "grad_norm": 0.3623557388782501, "learning_rate": 8.8747108319239e-05, "loss": 0.0436, "step": 25690 }, { "grad_norm": 0.5009939670562744, "learning_rate": 8.87366557399268e-05, "loss": 0.0389, "step": 25700 }, { "grad_norm": 0.584897518157959, "learning_rate": 8.872619892443217e-05, "loss": 0.0415, "step": 25710 }, { "grad_norm": 0.39479392766952515, "learning_rate": 8.871573787389865e-05, "loss": 0.0376, "step": 25720 }, { "grad_norm": 0.4522082507610321, "learning_rate": 8.870527258947024e-05, "loss": 0.0514, "step": 25730 }, { "grad_norm": 0.4458223879337311, "learning_rate": 8.869480307229143e-05, "loss": 0.046, "step": 25740 }, { "grad_norm": 0.8467602133750916, "learning_rate": 8.868432932350712e-05, "loss": 0.0509, "step": 25750 }, { "grad_norm": 0.4664522707462311, "learning_rate": 8.867385134426272e-05, "loss": 0.0465, "step": 25760 }, { "grad_norm": 0.526492178440094, "learning_rate": 8.866336913570407e-05, "loss": 0.047, "step": 25770 }, { "grad_norm": 0.40883728861808777, "learning_rate": 8.865288269897751e-05, "loss": 0.0348, "step": 25780 }, { "grad_norm": 0.36290261149406433, "learning_rate": 8.864239203522981e-05, "loss": 0.0382, "step": 25790 }, { "grad_norm": 0.2770801782608032, "learning_rate": 8.863189714560822e-05, "loss": 0.039, "step": 25800 }, { "grad_norm": 0.4512210190296173, "learning_rate": 8.862139803126043e-05, "loss": 0.0393, "step": 25810 }, { "grad_norm": 0.6549025774002075, "learning_rate": 8.861089469333463e-05, "loss": 0.0408, "step": 25820 }, { "grad_norm": 0.6661602258682251, "learning_rate": 8.860038713297944e-05, "loss": 0.0453, "step": 25830 }, { "grad_norm": 0.3467043936252594, "learning_rate": 8.858987535134394e-05, "loss": 0.0419, "step": 25840 }, { "grad_norm": 0.4730727970600128, "learning_rate": 8.857935934957769e-05, "loss": 0.0429, "step": 25850 }, { "grad_norm": 0.3434668481349945, "learning_rate": 8.856883912883071e-05, "loss": 0.0399, "step": 25860 }, { "grad_norm": 0.3965180814266205, "learning_rate": 8.855831469025346e-05, "loss": 0.0435, "step": 25870 }, { "grad_norm": 0.5544431209564209, "learning_rate": 8.854778603499689e-05, "loss": 0.0544, "step": 25880 }, { "grad_norm": 0.3418215811252594, "learning_rate": 8.85372531642124e-05, "loss": 0.0427, "step": 25890 }, { "grad_norm": 0.32871055603027344, "learning_rate": 8.852671607905185e-05, "loss": 0.0391, "step": 25900 }, { "grad_norm": 0.3865250051021576, "learning_rate": 8.851617478066754e-05, "loss": 0.0431, "step": 25910 }, { "grad_norm": 0.5153000950813293, "learning_rate": 8.850562927021227e-05, "loss": 0.0387, "step": 25920 }, { "grad_norm": 0.5168173313140869, "learning_rate": 8.849507954883928e-05, "loss": 0.05, "step": 25930 }, { "grad_norm": 0.4372154474258423, "learning_rate": 8.848452561770226e-05, "loss": 0.039, "step": 25940 }, { "grad_norm": 0.3491305112838745, "learning_rate": 8.847396747795538e-05, "loss": 0.0481, "step": 25950 }, { "grad_norm": 1.0288118124008179, "learning_rate": 8.846340513075327e-05, "loss": 0.0429, "step": 25960 }, { "grad_norm": 0.5354883670806885, "learning_rate": 8.845283857725099e-05, "loss": 0.0448, "step": 25970 }, { "grad_norm": 0.4651745855808258, "learning_rate": 8.844226781860409e-05, "loss": 0.044, "step": 25980 }, { "grad_norm": 0.4392666816711426, "learning_rate": 8.84316928559686e-05, "loss": 0.0424, "step": 25990 }, { "grad_norm": 0.5446395874023438, "learning_rate": 8.842111369050094e-05, "loss": 0.0396, "step": 26000 }, { "grad_norm": 0.4022338390350342, "learning_rate": 8.841053032335808e-05, "loss": 0.0456, "step": 26010 }, { "grad_norm": 0.4133147597312927, "learning_rate": 8.839994275569735e-05, "loss": 0.0475, "step": 26020 }, { "grad_norm": 0.5744976997375488, "learning_rate": 8.838935098867662e-05, "loss": 0.0416, "step": 26030 }, { "grad_norm": 0.3958437442779541, "learning_rate": 8.837875502345418e-05, "loss": 0.0458, "step": 26040 }, { "grad_norm": 0.37865957617759705, "learning_rate": 8.83681548611888e-05, "loss": 0.0425, "step": 26050 }, { "grad_norm": 0.4456114172935486, "learning_rate": 8.835755050303969e-05, "loss": 0.0425, "step": 26060 }, { "grad_norm": 0.27202627062797546, "learning_rate": 8.834694195016653e-05, "loss": 0.0431, "step": 26070 }, { "grad_norm": 0.4586668312549591, "learning_rate": 8.833632920372942e-05, "loss": 0.0552, "step": 26080 }, { "grad_norm": 0.4935496151447296, "learning_rate": 8.832571226488903e-05, "loss": 0.0465, "step": 26090 }, { "grad_norm": 1.201377272605896, "learning_rate": 8.831509113480634e-05, "loss": 0.0384, "step": 26100 }, { "grad_norm": 0.3279551565647125, "learning_rate": 8.83044658146429e-05, "loss": 0.0338, "step": 26110 }, { "grad_norm": 0.5507262349128723, "learning_rate": 8.829383630556067e-05, "loss": 0.0515, "step": 26120 }, { "grad_norm": 0.29891958832740784, "learning_rate": 8.828320260872207e-05, "loss": 0.0364, "step": 26130 }, { "grad_norm": 0.43122416734695435, "learning_rate": 8.827256472529e-05, "loss": 0.0401, "step": 26140 }, { "grad_norm": 0.4232427775859833, "learning_rate": 8.826192265642778e-05, "loss": 0.0437, "step": 26150 }, { "grad_norm": 0.41208553314208984, "learning_rate": 8.825127640329923e-05, "loss": 0.0356, "step": 26160 }, { "grad_norm": 0.35397377610206604, "learning_rate": 8.824062596706861e-05, "loss": 0.0426, "step": 26170 }, { "grad_norm": 0.4898267984390259, "learning_rate": 8.822997134890062e-05, "loss": 0.0386, "step": 26180 }, { "grad_norm": 0.4617324769496918, "learning_rate": 8.821931254996044e-05, "loss": 0.0376, "step": 26190 }, { "grad_norm": 0.5792317986488342, "learning_rate": 8.82086495714137e-05, "loss": 0.0521, "step": 26200 }, { "grad_norm": 0.5124416947364807, "learning_rate": 8.81979824144265e-05, "loss": 0.0522, "step": 26210 }, { "grad_norm": 0.4423063099384308, "learning_rate": 8.818731108016536e-05, "loss": 0.0441, "step": 26220 }, { "grad_norm": 0.46214941143989563, "learning_rate": 8.81766355697973e-05, "loss": 0.0473, "step": 26230 }, { "grad_norm": 0.36694344878196716, "learning_rate": 8.816595588448977e-05, "loss": 0.0358, "step": 26240 }, { "grad_norm": 0.5000228881835938, "learning_rate": 8.81552720254107e-05, "loss": 0.0442, "step": 26250 }, { "grad_norm": 0.38776475191116333, "learning_rate": 8.814458399372842e-05, "loss": 0.0453, "step": 26260 }, { "grad_norm": 0.47128307819366455, "learning_rate": 8.813389179061181e-05, "loss": 0.0345, "step": 26270 }, { "grad_norm": 0.4865896999835968, "learning_rate": 8.812319541723012e-05, "loss": 0.0521, "step": 26280 }, { "grad_norm": 0.34932249784469604, "learning_rate": 8.811249487475309e-05, "loss": 0.0485, "step": 26290 }, { "grad_norm": 0.40960049629211426, "learning_rate": 8.810179016435092e-05, "loss": 0.0474, "step": 26300 }, { "grad_norm": 0.35078829526901245, "learning_rate": 8.809108128719428e-05, "loss": 0.05, "step": 26310 }, { "grad_norm": 0.359087735414505, "learning_rate": 8.808036824445424e-05, "loss": 0.0427, "step": 26320 }, { "grad_norm": 0.40220800042152405, "learning_rate": 8.806965103730238e-05, "loss": 0.0416, "step": 26330 }, { "grad_norm": 0.3588424623012543, "learning_rate": 8.805892966691074e-05, "loss": 0.0369, "step": 26340 }, { "grad_norm": 0.25587397813796997, "learning_rate": 8.804820413445175e-05, "loss": 0.0343, "step": 26350 }, { "grad_norm": 0.3452758193016052, "learning_rate": 8.803747444109837e-05, "loss": 0.0373, "step": 26360 }, { "grad_norm": 0.3497077524662018, "learning_rate": 8.802674058802399e-05, "loss": 0.0367, "step": 26370 }, { "grad_norm": 0.3804231584072113, "learning_rate": 8.801600257640241e-05, "loss": 0.0377, "step": 26380 }, { "grad_norm": 0.38475969433784485, "learning_rate": 8.800526040740795e-05, "loss": 0.0435, "step": 26390 }, { "grad_norm": 0.42205312848091125, "learning_rate": 8.799451408221535e-05, "loss": 0.039, "step": 26400 }, { "grad_norm": 0.38593095541000366, "learning_rate": 8.798376360199982e-05, "loss": 0.0404, "step": 26410 }, { "grad_norm": 0.3181508481502533, "learning_rate": 8.797300896793701e-05, "loss": 0.0347, "step": 26420 }, { "grad_norm": 0.32394397258758545, "learning_rate": 8.796225018120302e-05, "loss": 0.0465, "step": 26430 }, { "grad_norm": 0.4767957031726837, "learning_rate": 8.795148724297444e-05, "loss": 0.0472, "step": 26440 }, { "grad_norm": 0.3987996280193329, "learning_rate": 8.794072015442825e-05, "loss": 0.0484, "step": 26450 }, { "grad_norm": 0.4776589274406433, "learning_rate": 8.792994891674198e-05, "loss": 0.0451, "step": 26460 }, { "grad_norm": 0.4300503730773926, "learning_rate": 8.79191735310935e-05, "loss": 0.033, "step": 26470 }, { "grad_norm": 0.4574558138847351, "learning_rate": 8.790839399866122e-05, "loss": 0.0375, "step": 26480 }, { "grad_norm": 0.3861891031265259, "learning_rate": 8.789761032062397e-05, "loss": 0.0406, "step": 26490 }, { "grad_norm": 0.4948534667491913, "learning_rate": 8.788682249816103e-05, "loss": 0.0442, "step": 26500 }, { "grad_norm": 0.4311133623123169, "learning_rate": 8.787603053245215e-05, "loss": 0.0447, "step": 26510 }, { "grad_norm": 0.36428770422935486, "learning_rate": 8.78652344246775e-05, "loss": 0.0422, "step": 26520 }, { "grad_norm": 0.44582855701446533, "learning_rate": 8.785443417601776e-05, "loss": 0.0403, "step": 26530 }, { "grad_norm": 0.4196881055831909, "learning_rate": 8.784362978765401e-05, "loss": 0.0375, "step": 26540 }, { "grad_norm": 0.48339903354644775, "learning_rate": 8.783282126076779e-05, "loss": 0.0395, "step": 26550 }, { "grad_norm": 0.3728184700012207, "learning_rate": 8.782200859654112e-05, "loss": 0.0411, "step": 26560 }, { "grad_norm": 0.27443036437034607, "learning_rate": 8.781119179615646e-05, "loss": 0.0342, "step": 26570 }, { "grad_norm": 0.5333198308944702, "learning_rate": 8.780037086079674e-05, "loss": 0.0388, "step": 26580 }, { "grad_norm": 0.47728604078292847, "learning_rate": 8.778954579164527e-05, "loss": 0.0387, "step": 26590 }, { "grad_norm": 0.5798746347427368, "learning_rate": 8.777871658988588e-05, "loss": 0.0432, "step": 26600 }, { "grad_norm": 0.5835195779800415, "learning_rate": 8.776788325670285e-05, "loss": 0.0443, "step": 26610 }, { "grad_norm": 0.7114003896713257, "learning_rate": 8.775704579328089e-05, "loss": 0.0451, "step": 26620 }, { "grad_norm": 0.4506852924823761, "learning_rate": 8.774620420080517e-05, "loss": 0.0477, "step": 26630 }, { "grad_norm": 0.2902609705924988, "learning_rate": 8.773535848046131e-05, "loss": 0.0404, "step": 26640 }, { "grad_norm": 0.3375096917152405, "learning_rate": 8.772450863343538e-05, "loss": 0.0347, "step": 26650 }, { "grad_norm": 0.46537643671035767, "learning_rate": 8.77136546609139e-05, "loss": 0.0447, "step": 26660 }, { "grad_norm": 0.5662434101104736, "learning_rate": 8.770279656408385e-05, "loss": 0.0373, "step": 26670 }, { "grad_norm": 0.4459085762500763, "learning_rate": 8.769193434413265e-05, "loss": 0.0443, "step": 26680 }, { "grad_norm": 0.7168691754341125, "learning_rate": 8.76810680022482e-05, "loss": 0.0373, "step": 26690 }, { "grad_norm": 0.4587478041648865, "learning_rate": 8.767019753961878e-05, "loss": 0.0517, "step": 26700 }, { "grad_norm": 0.3841419219970703, "learning_rate": 8.765932295743321e-05, "loss": 0.0479, "step": 26710 }, { "grad_norm": 0.3573674261569977, "learning_rate": 8.764844425688068e-05, "loss": 0.0403, "step": 26720 }, { "grad_norm": 0.4173823297023773, "learning_rate": 8.763756143915092e-05, "loss": 0.0312, "step": 26730 }, { "grad_norm": 0.664905309677124, "learning_rate": 8.7626674505434e-05, "loss": 0.039, "step": 26740 }, { "grad_norm": 0.4078095555305481, "learning_rate": 8.761578345692053e-05, "loss": 0.0536, "step": 26750 }, { "grad_norm": 0.7917824387550354, "learning_rate": 8.760488829480156e-05, "loss": 0.0425, "step": 26760 }, { "grad_norm": 0.6116124987602234, "learning_rate": 8.759398902026854e-05, "loss": 0.0436, "step": 26770 }, { "grad_norm": 0.515657365322113, "learning_rate": 8.758308563451339e-05, "loss": 0.049, "step": 26780 }, { "grad_norm": 0.4585931599140167, "learning_rate": 8.75721781387285e-05, "loss": 0.048, "step": 26790 }, { "grad_norm": 0.3950536847114563, "learning_rate": 8.75612665341067e-05, "loss": 0.0459, "step": 26800 }, { "grad_norm": 0.4091532528400421, "learning_rate": 8.755035082184126e-05, "loss": 0.0404, "step": 26810 }, { "grad_norm": 0.3551386594772339, "learning_rate": 8.753943100312592e-05, "loss": 0.0461, "step": 26820 }, { "grad_norm": 0.35209837555885315, "learning_rate": 8.752850707915484e-05, "loss": 0.0404, "step": 26830 }, { "grad_norm": 0.5246790051460266, "learning_rate": 8.751757905112264e-05, "loss": 0.0456, "step": 26840 }, { "grad_norm": 0.544792890548706, "learning_rate": 8.75066469202244e-05, "loss": 0.0445, "step": 26850 }, { "grad_norm": 0.40395042300224304, "learning_rate": 8.749571068765567e-05, "loss": 0.0603, "step": 26860 }, { "grad_norm": 0.41531792283058167, "learning_rate": 8.748477035461238e-05, "loss": 0.0461, "step": 26870 }, { "grad_norm": 0.4452408254146576, "learning_rate": 8.747382592229095e-05, "loss": 0.0507, "step": 26880 }, { "grad_norm": 0.33741283416748047, "learning_rate": 8.746287739188828e-05, "loss": 0.0361, "step": 26890 }, { "grad_norm": 0.4159316122531891, "learning_rate": 8.745192476460165e-05, "loss": 0.0393, "step": 26900 }, { "grad_norm": 0.37819337844848633, "learning_rate": 8.744096804162882e-05, "loss": 0.0458, "step": 26910 }, { "grad_norm": 0.4809938669204712, "learning_rate": 8.743000722416804e-05, "loss": 0.0444, "step": 26920 }, { "grad_norm": 0.46821045875549316, "learning_rate": 8.741904231341793e-05, "loss": 0.0434, "step": 26930 }, { "grad_norm": 0.6989850997924805, "learning_rate": 8.740807331057762e-05, "loss": 0.0577, "step": 26940 }, { "grad_norm": 0.33950334787368774, "learning_rate": 8.739710021684667e-05, "loss": 0.0507, "step": 26950 }, { "grad_norm": 0.4687308073043823, "learning_rate": 8.738612303342503e-05, "loss": 0.0375, "step": 26960 }, { "grad_norm": 0.3226050138473511, "learning_rate": 8.73751417615132e-05, "loss": 0.0387, "step": 26970 }, { "grad_norm": 0.36620259284973145, "learning_rate": 8.736415640231208e-05, "loss": 0.0332, "step": 26980 }, { "grad_norm": 0.354463130235672, "learning_rate": 8.735316695702297e-05, "loss": 0.0371, "step": 26990 }, { "grad_norm": 0.6578396558761597, "learning_rate": 8.734217342684769e-05, "loss": 0.034, "step": 27000 }, { "grad_norm": 0.34767693281173706, "learning_rate": 8.733117581298847e-05, "loss": 0.0349, "step": 27010 }, { "grad_norm": 0.4055081307888031, "learning_rate": 8.732017411664796e-05, "loss": 0.0487, "step": 27020 }, { "grad_norm": 0.4420470595359802, "learning_rate": 8.730916833902936e-05, "loss": 0.0412, "step": 27030 }, { "grad_norm": 0.6351407766342163, "learning_rate": 8.729815848133618e-05, "loss": 0.0559, "step": 27040 }, { "grad_norm": 0.4270167350769043, "learning_rate": 8.728714454477247e-05, "loss": 0.04, "step": 27050 }, { "grad_norm": 0.6282505393028259, "learning_rate": 8.727612653054269e-05, "loss": 0.0344, "step": 27060 }, { "grad_norm": 0.3616270124912262, "learning_rate": 8.726510443985176e-05, "loss": 0.035, "step": 27070 }, { "grad_norm": 0.4142947793006897, "learning_rate": 8.725407827390503e-05, "loss": 0.0337, "step": 27080 }, { "grad_norm": 0.3357011377811432, "learning_rate": 8.724304803390833e-05, "loss": 0.0373, "step": 27090 }, { "grad_norm": 0.4164617359638214, "learning_rate": 8.723201372106788e-05, "loss": 0.039, "step": 27100 }, { "grad_norm": 0.518020510673523, "learning_rate": 8.722097533659038e-05, "loss": 0.0418, "step": 27110 }, { "grad_norm": 0.5627051591873169, "learning_rate": 8.720993288168299e-05, "loss": 0.0432, "step": 27120 }, { "grad_norm": 0.4286799430847168, "learning_rate": 8.719888635755327e-05, "loss": 0.0436, "step": 27130 }, { "grad_norm": 0.42723190784454346, "learning_rate": 8.718783576540928e-05, "loss": 0.0481, "step": 27140 }, { "grad_norm": 0.4323354959487915, "learning_rate": 8.717678110645948e-05, "loss": 0.0421, "step": 27150 }, { "grad_norm": 0.21702590584754944, "learning_rate": 8.716572238191279e-05, "loss": 0.0377, "step": 27160 }, { "grad_norm": 0.4046148657798767, "learning_rate": 8.715465959297857e-05, "loss": 0.0537, "step": 27170 }, { "grad_norm": 0.5289581418037415, "learning_rate": 8.714359274086665e-05, "loss": 0.0432, "step": 27180 }, { "grad_norm": 0.7099794745445251, "learning_rate": 8.713252182678726e-05, "loss": 0.0471, "step": 27190 }, { "grad_norm": 0.3874168395996094, "learning_rate": 8.712144685195112e-05, "loss": 0.0456, "step": 27200 }, { "grad_norm": 0.5511634945869446, "learning_rate": 8.711036781756936e-05, "loss": 0.0499, "step": 27210 }, { "grad_norm": 0.5159416794776917, "learning_rate": 8.709928472485357e-05, "loss": 0.0423, "step": 27220 }, { "grad_norm": 0.42860084772109985, "learning_rate": 8.708819757501579e-05, "loss": 0.045, "step": 27230 }, { "grad_norm": 0.4804626405239105, "learning_rate": 8.707710636926846e-05, "loss": 0.0419, "step": 27240 }, { "grad_norm": 0.47899749875068665, "learning_rate": 8.706601110882455e-05, "loss": 0.0447, "step": 27250 }, { "grad_norm": 0.5532068610191345, "learning_rate": 8.705491179489738e-05, "loss": 0.0414, "step": 27260 }, { "grad_norm": 0.5941108465194702, "learning_rate": 8.704380842870077e-05, "loss": 0.0382, "step": 27270 }, { "grad_norm": 0.3183443546295166, "learning_rate": 8.703270101144895e-05, "loss": 0.0425, "step": 27280 }, { "grad_norm": 0.30341508984565735, "learning_rate": 8.702158954435664e-05, "loss": 0.0331, "step": 27290 }, { "grad_norm": 0.34908318519592285, "learning_rate": 8.701047402863896e-05, "loss": 0.0393, "step": 27300 }, { "grad_norm": 0.5653524398803711, "learning_rate": 8.699935446551148e-05, "loss": 0.0492, "step": 27310 }, { "grad_norm": 0.3454056680202484, "learning_rate": 8.698823085619022e-05, "loss": 0.0379, "step": 27320 }, { "grad_norm": 0.3742310702800751, "learning_rate": 8.697710320189166e-05, "loss": 0.0343, "step": 27330 }, { "grad_norm": 0.554734468460083, "learning_rate": 8.696597150383268e-05, "loss": 0.0437, "step": 27340 }, { "grad_norm": 0.26611730456352234, "learning_rate": 8.695483576323063e-05, "loss": 0.0398, "step": 27350 }, { "grad_norm": 0.7548735737800598, "learning_rate": 8.69436959813033e-05, "loss": 0.0471, "step": 27360 }, { "grad_norm": 0.29660072922706604, "learning_rate": 8.693255215926892e-05, "loss": 0.0388, "step": 27370 }, { "grad_norm": 0.43487420678138733, "learning_rate": 8.692140429834617e-05, "loss": 0.0402, "step": 27380 }, { "grad_norm": 0.3282899260520935, "learning_rate": 8.691025239975415e-05, "loss": 0.0459, "step": 27390 }, { "grad_norm": 0.6127834320068359, "learning_rate": 8.689909646471243e-05, "loss": 0.0362, "step": 27400 }, { "grad_norm": 0.34457772970199585, "learning_rate": 8.688793649444099e-05, "loss": 0.0485, "step": 27410 }, { "grad_norm": 0.40193071961402893, "learning_rate": 8.687677249016029e-05, "loss": 0.038, "step": 27420 }, { "grad_norm": 0.43490567803382874, "learning_rate": 8.686560445309118e-05, "loss": 0.0444, "step": 27430 }, { "grad_norm": 0.44704264402389526, "learning_rate": 8.685443238445499e-05, "loss": 0.0432, "step": 27440 }, { "grad_norm": 0.4943084120750427, "learning_rate": 8.68432562854735e-05, "loss": 0.0416, "step": 27450 }, { "grad_norm": 0.5002661347389221, "learning_rate": 8.683207615736887e-05, "loss": 0.0453, "step": 27460 }, { "grad_norm": 0.37799832224845886, "learning_rate": 8.682089200136379e-05, "loss": 0.0573, "step": 27470 }, { "grad_norm": 0.5004562735557556, "learning_rate": 8.680970381868132e-05, "loss": 0.0422, "step": 27480 }, { "grad_norm": 0.46034446358680725, "learning_rate": 8.679851161054498e-05, "loss": 0.0406, "step": 27490 }, { "grad_norm": 0.5101285576820374, "learning_rate": 8.678731537817873e-05, "loss": 0.0476, "step": 27500 }, { "grad_norm": 0.467402845621109, "learning_rate": 8.677611512280697e-05, "loss": 0.0393, "step": 27510 }, { "grad_norm": 0.32367458939552307, "learning_rate": 8.676491084565457e-05, "loss": 0.0502, "step": 27520 }, { "grad_norm": 0.5532536506652832, "learning_rate": 8.675370254794678e-05, "loss": 0.0516, "step": 27530 }, { "grad_norm": 0.42675942182540894, "learning_rate": 8.674249023090935e-05, "loss": 0.0366, "step": 27540 }, { "grad_norm": 0.49549490213394165, "learning_rate": 8.673127389576843e-05, "loss": 0.0424, "step": 27550 }, { "grad_norm": 0.30105501413345337, "learning_rate": 8.67200535437506e-05, "loss": 0.0363, "step": 27560 }, { "grad_norm": 0.38459253311157227, "learning_rate": 8.670882917608296e-05, "loss": 0.0442, "step": 27570 }, { "grad_norm": 0.41298937797546387, "learning_rate": 8.669760079399292e-05, "loss": 0.0373, "step": 27580 }, { "grad_norm": 0.6051907539367676, "learning_rate": 8.668636839870845e-05, "loss": 0.0416, "step": 27590 }, { "grad_norm": 0.45719242095947266, "learning_rate": 8.667513199145789e-05, "loss": 0.0431, "step": 27600 }, { "grad_norm": 0.5505886673927307, "learning_rate": 8.666389157347002e-05, "loss": 0.0482, "step": 27610 }, { "grad_norm": 0.36048296093940735, "learning_rate": 8.66526471459741e-05, "loss": 0.0375, "step": 27620 }, { "grad_norm": 0.3353897035121918, "learning_rate": 8.66413987101998e-05, "loss": 0.0358, "step": 27630 }, { "grad_norm": 0.3509518504142761, "learning_rate": 8.663014626737723e-05, "loss": 0.0421, "step": 27640 }, { "grad_norm": 0.5933994650840759, "learning_rate": 8.661888981873691e-05, "loss": 0.0455, "step": 27650 }, { "grad_norm": 0.38194048404693604, "learning_rate": 8.660762936550988e-05, "loss": 0.0508, "step": 27660 }, { "grad_norm": 0.3924555480480194, "learning_rate": 8.659636490892753e-05, "loss": 0.0524, "step": 27670 }, { "grad_norm": 0.6086640954017639, "learning_rate": 8.658509645022174e-05, "loss": 0.047, "step": 27680 }, { "grad_norm": 0.4280306100845337, "learning_rate": 8.657382399062481e-05, "loss": 0.0408, "step": 27690 }, { "grad_norm": 0.34953993558883667, "learning_rate": 8.656254753136946e-05, "loss": 0.0421, "step": 27700 }, { "grad_norm": 0.38972997665405273, "learning_rate": 8.655126707368891e-05, "loss": 0.0449, "step": 27710 }, { "grad_norm": 0.4030477702617645, "learning_rate": 8.653998261881672e-05, "loss": 0.0417, "step": 27720 }, { "grad_norm": 0.3876950442790985, "learning_rate": 8.652869416798699e-05, "loss": 0.0331, "step": 27730 }, { "grad_norm": 0.5587475299835205, "learning_rate": 8.651740172243417e-05, "loss": 0.0435, "step": 27740 }, { "grad_norm": 0.5042400360107422, "learning_rate": 8.65061052833932e-05, "loss": 0.0452, "step": 27750 }, { "grad_norm": 0.3393658697605133, "learning_rate": 8.649480485209945e-05, "loss": 0.0385, "step": 27760 }, { "grad_norm": 0.7158474326133728, "learning_rate": 8.64835004297887e-05, "loss": 0.0536, "step": 27770 }, { "grad_norm": 0.3871077001094818, "learning_rate": 8.64721920176972e-05, "loss": 0.0359, "step": 27780 }, { "grad_norm": 0.4968283772468567, "learning_rate": 8.646087961706164e-05, "loss": 0.0473, "step": 27790 }, { "grad_norm": 0.4229758083820343, "learning_rate": 8.644956322911908e-05, "loss": 0.0398, "step": 27800 }, { "grad_norm": 0.5320143103599548, "learning_rate": 8.643824285510709e-05, "loss": 0.0411, "step": 27810 }, { "grad_norm": 0.3862229585647583, "learning_rate": 8.642691849626364e-05, "loss": 0.0423, "step": 27820 }, { "grad_norm": 0.36695873737335205, "learning_rate": 8.641559015382717e-05, "loss": 0.0426, "step": 27830 }, { "grad_norm": 0.4633735120296478, "learning_rate": 8.640425782903649e-05, "loss": 0.0515, "step": 27840 }, { "grad_norm": 0.45210379362106323, "learning_rate": 8.639292152313091e-05, "loss": 0.0343, "step": 27850 }, { "grad_norm": 0.3451235592365265, "learning_rate": 8.638158123735015e-05, "loss": 0.033, "step": 27860 }, { "grad_norm": 0.5168785452842712, "learning_rate": 8.637023697293436e-05, "loss": 0.0414, "step": 27870 }, { "grad_norm": 0.4936569929122925, "learning_rate": 8.635888873112414e-05, "loss": 0.0484, "step": 27880 }, { "grad_norm": 0.3602091670036316, "learning_rate": 8.634753651316052e-05, "loss": 0.046, "step": 27890 }, { "grad_norm": 0.30948740243911743, "learning_rate": 8.633618032028496e-05, "loss": 0.0339, "step": 27900 }, { "grad_norm": 0.8186019659042358, "learning_rate": 8.632482015373934e-05, "loss": 0.0413, "step": 27910 }, { "grad_norm": 0.3481990098953247, "learning_rate": 8.6313456014766e-05, "loss": 0.0421, "step": 27920 }, { "grad_norm": 0.35803312063217163, "learning_rate": 8.630208790460771e-05, "loss": 0.0446, "step": 27930 }, { "grad_norm": 0.3222566843032837, "learning_rate": 8.629071582450768e-05, "loss": 0.0438, "step": 27940 }, { "grad_norm": 0.4958183765411377, "learning_rate": 8.62793397757095e-05, "loss": 0.0438, "step": 27950 }, { "grad_norm": 0.5685730576515198, "learning_rate": 8.626795975945729e-05, "loss": 0.0498, "step": 27960 }, { "grad_norm": 0.4051649272441864, "learning_rate": 8.625657577699551e-05, "loss": 0.0394, "step": 27970 }, { "grad_norm": 0.32399675250053406, "learning_rate": 8.624518782956914e-05, "loss": 0.0358, "step": 27980 }, { "grad_norm": 0.6234937906265259, "learning_rate": 8.62337959184235e-05, "loss": 0.0375, "step": 27990 }, { "grad_norm": 0.4143308103084564, "learning_rate": 8.622240004480441e-05, "loss": 0.0462, "step": 28000 }, { "grad_norm": 0.3837812840938568, "learning_rate": 8.621100020995814e-05, "loss": 0.0362, "step": 28010 }, { "grad_norm": 0.46880054473876953, "learning_rate": 8.619959641513132e-05, "loss": 0.0461, "step": 28020 }, { "grad_norm": 0.5662624835968018, "learning_rate": 8.618818866157105e-05, "loss": 0.0466, "step": 28030 }, { "grad_norm": 0.7808968424797058, "learning_rate": 8.617677695052487e-05, "loss": 0.0437, "step": 28040 }, { "grad_norm": 0.530727744102478, "learning_rate": 8.616536128324078e-05, "loss": 0.035, "step": 28050 }, { "grad_norm": 0.3699967861175537, "learning_rate": 8.615394166096712e-05, "loss": 0.0386, "step": 28060 }, { "grad_norm": 0.28322774171829224, "learning_rate": 8.614251808495279e-05, "loss": 0.0326, "step": 28070 }, { "grad_norm": 0.34199804067611694, "learning_rate": 8.6131090556447e-05, "loss": 0.0301, "step": 28080 }, { "grad_norm": 0.5082517862319946, "learning_rate": 8.611965907669947e-05, "loss": 0.0427, "step": 28090 }, { "grad_norm": 0.30672526359558105, "learning_rate": 8.610822364696034e-05, "loss": 0.0379, "step": 28100 }, { "grad_norm": 0.38457271456718445, "learning_rate": 8.609678426848015e-05, "loss": 0.0359, "step": 28110 }, { "grad_norm": 0.42827317118644714, "learning_rate": 8.60853409425099e-05, "loss": 0.0383, "step": 28120 }, { "grad_norm": 0.41301921010017395, "learning_rate": 8.607389367030104e-05, "loss": 0.0351, "step": 28130 }, { "grad_norm": 0.45384514331817627, "learning_rate": 8.606244245310538e-05, "loss": 0.0377, "step": 28140 }, { "grad_norm": 0.6409202814102173, "learning_rate": 8.605098729217525e-05, "loss": 0.0414, "step": 28150 }, { "grad_norm": 0.39627107977867126, "learning_rate": 8.603952818876335e-05, "loss": 0.0401, "step": 28160 }, { "grad_norm": 0.3360016942024231, "learning_rate": 8.602806514412281e-05, "loss": 0.0335, "step": 28170 }, { "grad_norm": 0.5594989657402039, "learning_rate": 8.601659815950726e-05, "loss": 0.0473, "step": 28180 }, { "grad_norm": 0.5977881550788879, "learning_rate": 8.600512723617067e-05, "loss": 0.0619, "step": 28190 }, { "grad_norm": 0.5433416366577148, "learning_rate": 8.59936523753675e-05, "loss": 0.0489, "step": 28200 }, { "grad_norm": 0.4460686147212982, "learning_rate": 8.598217357835264e-05, "loss": 0.0403, "step": 28210 }, { "grad_norm": 0.3207935690879822, "learning_rate": 8.597069084638135e-05, "loss": 0.0353, "step": 28220 }, { "grad_norm": 0.4291301369667053, "learning_rate": 8.595920418070939e-05, "loss": 0.0371, "step": 28230 }, { "grad_norm": 0.3640329837799072, "learning_rate": 8.594771358259295e-05, "loss": 0.0378, "step": 28240 }, { "grad_norm": 0.5630723237991333, "learning_rate": 8.593621905328858e-05, "loss": 0.0476, "step": 28250 }, { "grad_norm": 0.4464929401874542, "learning_rate": 8.592472059405333e-05, "loss": 0.0398, "step": 28260 }, { "grad_norm": 0.40259552001953125, "learning_rate": 8.591321820614464e-05, "loss": 0.0305, "step": 28270 }, { "grad_norm": 0.48708927631378174, "learning_rate": 8.590171189082041e-05, "loss": 0.0391, "step": 28280 }, { "grad_norm": 0.48405250906944275, "learning_rate": 8.589020164933894e-05, "loss": 0.0392, "step": 28290 }, { "grad_norm": 0.4355095326900482, "learning_rate": 8.587868748295898e-05, "loss": 0.0442, "step": 28300 }, { "grad_norm": 0.46366482973098755, "learning_rate": 8.586716939293971e-05, "loss": 0.0466, "step": 28310 }, { "grad_norm": 0.40815985202789307, "learning_rate": 8.58556473805407e-05, "loss": 0.0365, "step": 28320 }, { "grad_norm": 0.5693199038505554, "learning_rate": 8.584412144702202e-05, "loss": 0.044, "step": 28330 }, { "grad_norm": 0.3840924799442291, "learning_rate": 8.58325915936441e-05, "loss": 0.041, "step": 28340 }, { "grad_norm": 0.445156991481781, "learning_rate": 8.582105782166783e-05, "loss": 0.0411, "step": 28350 }, { "grad_norm": 0.43034058809280396, "learning_rate": 8.580952013235455e-05, "loss": 0.0355, "step": 28360 }, { "grad_norm": 0.647331714630127, "learning_rate": 8.579797852696596e-05, "loss": 0.0441, "step": 28370 }, { "grad_norm": 0.41754627227783203, "learning_rate": 8.578643300676428e-05, "loss": 0.0371, "step": 28380 }, { "grad_norm": 0.3755742013454437, "learning_rate": 8.577488357301209e-05, "loss": 0.0437, "step": 28390 }, { "grad_norm": 0.35740455985069275, "learning_rate": 8.576333022697242e-05, "loss": 0.0392, "step": 28400 }, { "grad_norm": 0.5364767909049988, "learning_rate": 8.575177296990873e-05, "loss": 0.0411, "step": 28410 }, { "grad_norm": 0.4025956988334656, "learning_rate": 8.574021180308489e-05, "loss": 0.0392, "step": 28420 }, { "grad_norm": 0.4472423195838928, "learning_rate": 8.572864672776523e-05, "loss": 0.0412, "step": 28430 }, { "grad_norm": 0.3836451470851898, "learning_rate": 8.571707774521447e-05, "loss": 0.0381, "step": 28440 }, { "grad_norm": 0.40236327052116394, "learning_rate": 8.57055048566978e-05, "loss": 0.0336, "step": 28450 }, { "grad_norm": 0.44101130962371826, "learning_rate": 8.569392806348078e-05, "loss": 0.0425, "step": 28460 }, { "grad_norm": 0.3818504810333252, "learning_rate": 8.568234736682947e-05, "loss": 0.0363, "step": 28470 }, { "grad_norm": 0.3906177580356598, "learning_rate": 8.567076276801029e-05, "loss": 0.0435, "step": 28480 }, { "grad_norm": 0.39955011010169983, "learning_rate": 8.565917426829013e-05, "loss": 0.0347, "step": 28490 }, { "grad_norm": 0.32652777433395386, "learning_rate": 8.564758186893628e-05, "loss": 0.0436, "step": 28500 }, { "grad_norm": 0.34258219599723816, "learning_rate": 8.563598557121649e-05, "loss": 0.0318, "step": 28510 }, { "grad_norm": 0.570051372051239, "learning_rate": 8.562438537639888e-05, "loss": 0.043, "step": 28520 }, { "grad_norm": 0.4777708947658539, "learning_rate": 8.561278128575206e-05, "loss": 0.0448, "step": 28530 }, { "grad_norm": 0.3247937858104706, "learning_rate": 8.5601173300545e-05, "loss": 0.0375, "step": 28540 }, { "grad_norm": 0.47280406951904297, "learning_rate": 8.558956142204717e-05, "loss": 0.049, "step": 28550 }, { "grad_norm": 0.4400039613246918, "learning_rate": 8.55779456515284e-05, "loss": 0.0515, "step": 28560 }, { "grad_norm": 0.575627863407135, "learning_rate": 8.556632599025898e-05, "loss": 0.0545, "step": 28570 }, { "grad_norm": 0.3048432767391205, "learning_rate": 8.555470243950964e-05, "loss": 0.0443, "step": 28580 }, { "grad_norm": 0.5746194124221802, "learning_rate": 8.554307500055148e-05, "loss": 0.0391, "step": 28590 }, { "grad_norm": 0.6404807567596436, "learning_rate": 8.553144367465609e-05, "loss": 0.0379, "step": 28600 }, { "grad_norm": 0.4372609257698059, "learning_rate": 8.551980846309544e-05, "loss": 0.0497, "step": 28610 }, { "grad_norm": 0.3960036039352417, "learning_rate": 8.550816936714193e-05, "loss": 0.0451, "step": 28620 }, { "grad_norm": 0.4366016685962677, "learning_rate": 8.549652638806841e-05, "loss": 0.0448, "step": 28630 }, { "grad_norm": 0.40980687737464905, "learning_rate": 8.548487952714812e-05, "loss": 0.0444, "step": 28640 }, { "grad_norm": 0.38885584473609924, "learning_rate": 8.547322878565478e-05, "loss": 0.0461, "step": 28650 }, { "grad_norm": 0.47864967584609985, "learning_rate": 8.546157416486245e-05, "loss": 0.0437, "step": 28660 }, { "grad_norm": 0.40237101912498474, "learning_rate": 8.54499156660457e-05, "loss": 0.0418, "step": 28670 }, { "grad_norm": 0.6526597738265991, "learning_rate": 8.543825329047947e-05, "loss": 0.0455, "step": 28680 }, { "grad_norm": 0.4875834286212921, "learning_rate": 8.542658703943913e-05, "loss": 0.0481, "step": 28690 }, { "grad_norm": 0.31118902564048767, "learning_rate": 8.541491691420051e-05, "loss": 0.0454, "step": 28700 }, { "grad_norm": 0.5087518692016602, "learning_rate": 8.54032429160398e-05, "loss": 0.0389, "step": 28710 }, { "grad_norm": 0.5667688250541687, "learning_rate": 8.539156504623369e-05, "loss": 0.0465, "step": 28720 }, { "grad_norm": 0.54139643907547, "learning_rate": 8.537988330605923e-05, "loss": 0.0591, "step": 28730 }, { "grad_norm": 0.33387818932533264, "learning_rate": 8.536819769679393e-05, "loss": 0.0387, "step": 28740 }, { "grad_norm": 0.40777671337127686, "learning_rate": 8.53565082197157e-05, "loss": 0.0382, "step": 28750 }, { "grad_norm": 0.336847186088562, "learning_rate": 8.534481487610289e-05, "loss": 0.0537, "step": 28760 }, { "grad_norm": 0.6235640048980713, "learning_rate": 8.533311766723428e-05, "loss": 0.0381, "step": 28770 }, { "grad_norm": 0.6261714696884155, "learning_rate": 8.532141659438901e-05, "loss": 0.0397, "step": 28780 }, { "grad_norm": 0.4668026268482208, "learning_rate": 8.530971165884675e-05, "loss": 0.0387, "step": 28790 }, { "grad_norm": 0.6621936559677124, "learning_rate": 8.529800286188752e-05, "loss": 0.0413, "step": 28800 }, { "grad_norm": 0.7007450461387634, "learning_rate": 8.528629020479175e-05, "loss": 0.0347, "step": 28810 }, { "grad_norm": 0.38899824023246765, "learning_rate": 8.527457368884033e-05, "loss": 0.0406, "step": 28820 }, { "grad_norm": 0.3087553381919861, "learning_rate": 8.526285331531458e-05, "loss": 0.0381, "step": 28830 }, { "grad_norm": 0.4584328830242157, "learning_rate": 8.525112908549621e-05, "loss": 0.034, "step": 28840 }, { "grad_norm": 0.4679088592529297, "learning_rate": 8.523940100066735e-05, "loss": 0.0437, "step": 28850 }, { "grad_norm": 0.4206295311450958, "learning_rate": 8.52276690621106e-05, "loss": 0.0427, "step": 28860 }, { "grad_norm": 0.4419923424720764, "learning_rate": 8.521593327110889e-05, "loss": 0.0362, "step": 28870 }, { "grad_norm": 0.3952674865722656, "learning_rate": 8.520419362894569e-05, "loss": 0.0499, "step": 28880 }, { "grad_norm": 0.4357181191444397, "learning_rate": 8.51924501369048e-05, "loss": 0.0394, "step": 28890 }, { "grad_norm": 0.6675794720649719, "learning_rate": 8.518070279627047e-05, "loss": 0.0408, "step": 28900 }, { "grad_norm": 0.34381765127182007, "learning_rate": 8.516895160832737e-05, "loss": 0.0487, "step": 28910 }, { "grad_norm": 0.3669087886810303, "learning_rate": 8.515719657436061e-05, "loss": 0.0396, "step": 28920 }, { "grad_norm": 0.34679991006851196, "learning_rate": 8.514543769565568e-05, "loss": 0.0338, "step": 28930 }, { "grad_norm": 0.3926030397415161, "learning_rate": 8.513367497349853e-05, "loss": 0.044, "step": 28940 }, { "grad_norm": 0.3071904480457306, "learning_rate": 8.51219084091755e-05, "loss": 0.0399, "step": 28950 }, { "grad_norm": 0.30242720246315, "learning_rate": 8.511013800397338e-05, "loss": 0.0316, "step": 28960 }, { "grad_norm": 0.47555050253868103, "learning_rate": 8.509836375917937e-05, "loss": 0.0405, "step": 28970 }, { "grad_norm": 0.36115923523902893, "learning_rate": 8.508658567608104e-05, "loss": 0.037, "step": 28980 }, { "grad_norm": 0.4811100661754608, "learning_rate": 8.507480375596647e-05, "loss": 0.0448, "step": 28990 }, { "grad_norm": 0.4258691072463989, "learning_rate": 8.506301800012408e-05, "loss": 0.0452, "step": 29000 }, { "grad_norm": 0.2994866669178009, "learning_rate": 8.505122840984278e-05, "loss": 0.0377, "step": 29010 }, { "grad_norm": 0.3882859945297241, "learning_rate": 8.503943498641182e-05, "loss": 0.0401, "step": 29020 }, { "grad_norm": 0.4273071885108948, "learning_rate": 8.502763773112095e-05, "loss": 0.0457, "step": 29030 }, { "grad_norm": 0.5075469017028809, "learning_rate": 8.501583664526026e-05, "loss": 0.0447, "step": 29040 }, { "grad_norm": 0.8283064365386963, "learning_rate": 8.500403173012032e-05, "loss": 0.047, "step": 29050 }, { "grad_norm": 0.4380524158477783, "learning_rate": 8.499222298699211e-05, "loss": 0.0396, "step": 29060 }, { "grad_norm": 0.3679298758506775, "learning_rate": 8.498041041716701e-05, "loss": 0.0414, "step": 29070 }, { "grad_norm": 0.3801822364330292, "learning_rate": 8.496859402193681e-05, "loss": 0.0391, "step": 29080 }, { "grad_norm": 0.40052109956741333, "learning_rate": 8.495677380259374e-05, "loss": 0.0402, "step": 29090 }, { "grad_norm": 0.58137047290802, "learning_rate": 8.494494976043045e-05, "loss": 0.0492, "step": 29100 }, { "grad_norm": 0.3392321765422821, "learning_rate": 8.493312189673998e-05, "loss": 0.0346, "step": 29110 }, { "grad_norm": 0.4876959025859833, "learning_rate": 8.492129021281584e-05, "loss": 0.0389, "step": 29120 }, { "grad_norm": 0.47013145685195923, "learning_rate": 8.490945470995188e-05, "loss": 0.0324, "step": 29130 }, { "grad_norm": 0.2996153235435486, "learning_rate": 8.489761538944247e-05, "loss": 0.0328, "step": 29140 }, { "grad_norm": 0.4617217183113098, "learning_rate": 8.48857722525823e-05, "loss": 0.0388, "step": 29150 }, { "grad_norm": 0.46155908703804016, "learning_rate": 8.487392530066652e-05, "loss": 0.0428, "step": 29160 }, { "grad_norm": 0.44753336906433105, "learning_rate": 8.486207453499069e-05, "loss": 0.0578, "step": 29170 }, { "grad_norm": 0.3917921483516693, "learning_rate": 8.485021995685082e-05, "loss": 0.0461, "step": 29180 }, { "grad_norm": 0.45321860909461975, "learning_rate": 8.483836156754328e-05, "loss": 0.0407, "step": 29190 }, { "grad_norm": 0.3152829110622406, "learning_rate": 8.482649936836491e-05, "loss": 0.0455, "step": 29200 }, { "grad_norm": 0.4854075312614441, "learning_rate": 8.481463336061293e-05, "loss": 0.0436, "step": 29210 }, { "grad_norm": 0.33725517988204956, "learning_rate": 8.480276354558496e-05, "loss": 0.0522, "step": 29220 }, { "grad_norm": 0.6475211977958679, "learning_rate": 8.479088992457913e-05, "loss": 0.0436, "step": 29230 }, { "grad_norm": 0.5965729355812073, "learning_rate": 8.477901249889387e-05, "loss": 0.0487, "step": 29240 }, { "grad_norm": 0.47590187191963196, "learning_rate": 8.47671312698281e-05, "loss": 0.0368, "step": 29250 }, { "grad_norm": 0.34345152974128723, "learning_rate": 8.475524623868112e-05, "loss": 0.0482, "step": 29260 }, { "grad_norm": 0.4729747474193573, "learning_rate": 8.474335740675266e-05, "loss": 0.041, "step": 29270 }, { "grad_norm": 0.39505961537361145, "learning_rate": 8.473146477534289e-05, "loss": 0.0412, "step": 29280 }, { "grad_norm": 0.41080424189567566, "learning_rate": 8.471956834575232e-05, "loss": 0.0468, "step": 29290 }, { "grad_norm": 0.4967886507511139, "learning_rate": 8.470766811928197e-05, "loss": 0.0552, "step": 29300 }, { "grad_norm": 0.2838013172149658, "learning_rate": 8.469576409723323e-05, "loss": 0.0487, "step": 29310 }, { "grad_norm": 0.39699238538742065, "learning_rate": 8.468385628090788e-05, "loss": 0.0381, "step": 29320 }, { "grad_norm": 0.4352569580078125, "learning_rate": 8.467194467160815e-05, "loss": 0.0434, "step": 29330 }, { "grad_norm": 0.5723453164100647, "learning_rate": 8.466002927063667e-05, "loss": 0.0363, "step": 29340 }, { "grad_norm": 0.4402070641517639, "learning_rate": 8.464811007929651e-05, "loss": 0.0402, "step": 29350 }, { "grad_norm": 0.338105171918869, "learning_rate": 8.463618709889114e-05, "loss": 0.0383, "step": 29360 }, { "grad_norm": 0.4515955150127411, "learning_rate": 8.462426033072442e-05, "loss": 0.0402, "step": 29370 }, { "grad_norm": 0.4422183930873871, "learning_rate": 8.461232977610061e-05, "loss": 0.0398, "step": 29380 }, { "grad_norm": 0.9128710031509399, "learning_rate": 8.46003954363245e-05, "loss": 0.0372, "step": 29390 }, { "grad_norm": 0.32454806566238403, "learning_rate": 8.458845731270115e-05, "loss": 0.0463, "step": 29400 }, { "grad_norm": 0.4386192262172699, "learning_rate": 8.45765154065361e-05, "loss": 0.0397, "step": 29410 }, { "grad_norm": 0.3899981677532196, "learning_rate": 8.456456971913532e-05, "loss": 0.033, "step": 29420 }, { "grad_norm": 0.4569338262081146, "learning_rate": 8.455262025180517e-05, "loss": 0.0332, "step": 29430 }, { "grad_norm": 0.4499913156032562, "learning_rate": 8.454066700585242e-05, "loss": 0.0426, "step": 29440 }, { "grad_norm": 0.26219847798347473, "learning_rate": 8.452870998258423e-05, "loss": 0.0327, "step": 29450 }, { "grad_norm": 0.3987182378768921, "learning_rate": 8.451674918330825e-05, "loss": 0.035, "step": 29460 }, { "grad_norm": 0.452518492937088, "learning_rate": 8.450478460933246e-05, "loss": 0.0407, "step": 29470 }, { "grad_norm": 0.3632400929927826, "learning_rate": 8.449281626196532e-05, "loss": 0.0408, "step": 29480 }, { "grad_norm": 0.3995096981525421, "learning_rate": 8.448084414251564e-05, "loss": 0.0361, "step": 29490 }, { "grad_norm": 0.3573055863380432, "learning_rate": 8.446886825229271e-05, "loss": 0.0476, "step": 29500 }, { "grad_norm": 0.3911934792995453, "learning_rate": 8.445688859260615e-05, "loss": 0.0387, "step": 29510 }, { "grad_norm": 0.3342858850955963, "learning_rate": 8.444490516476606e-05, "loss": 0.0432, "step": 29520 }, { "grad_norm": 0.2829572856426239, "learning_rate": 8.443291797008293e-05, "loss": 0.0405, "step": 29530 }, { "grad_norm": 0.27298054099082947, "learning_rate": 8.442092700986765e-05, "loss": 0.0473, "step": 29540 }, { "grad_norm": 0.40472614765167236, "learning_rate": 8.440893228543156e-05, "loss": 0.0421, "step": 29550 }, { "grad_norm": 0.3549295961856842, "learning_rate": 8.439693379808638e-05, "loss": 0.0349, "step": 29560 }, { "grad_norm": 0.38845041394233704, "learning_rate": 8.43849315491442e-05, "loss": 0.0512, "step": 29570 }, { "grad_norm": 0.4336089491844177, "learning_rate": 8.437292553991763e-05, "loss": 0.0461, "step": 29580 }, { "grad_norm": 0.42919620871543884, "learning_rate": 8.436091577171959e-05, "loss": 0.038, "step": 29590 }, { "grad_norm": 0.40851709246635437, "learning_rate": 8.434890224586347e-05, "loss": 0.0381, "step": 29600 }, { "grad_norm": 0.4589933156967163, "learning_rate": 8.433688496366303e-05, "loss": 0.031, "step": 29610 }, { "grad_norm": 0.3942793011665344, "learning_rate": 8.432486392643248e-05, "loss": 0.037, "step": 29620 }, { "grad_norm": 0.3870985209941864, "learning_rate": 8.431283913548643e-05, "loss": 0.0392, "step": 29630 }, { "grad_norm": 0.34324467182159424, "learning_rate": 8.430081059213985e-05, "loss": 0.0365, "step": 29640 }, { "grad_norm": 0.28771263360977173, "learning_rate": 8.428877829770823e-05, "loss": 0.0462, "step": 29650 }, { "grad_norm": 0.3703765273094177, "learning_rate": 8.427674225350735e-05, "loss": 0.0524, "step": 29660 }, { "grad_norm": 0.4823734164237976, "learning_rate": 8.426470246085347e-05, "loss": 0.0441, "step": 29670 }, { "grad_norm": 0.41162705421447754, "learning_rate": 8.425265892106324e-05, "loss": 0.0386, "step": 29680 }, { "grad_norm": 0.4787221848964691, "learning_rate": 8.424061163545374e-05, "loss": 0.0367, "step": 29690 }, { "grad_norm": 0.5279578566551208, "learning_rate": 8.422856060534243e-05, "loss": 0.0446, "step": 29700 }, { "grad_norm": 0.565682590007782, "learning_rate": 8.421650583204718e-05, "loss": 0.0476, "step": 29710 }, { "grad_norm": 0.45311233401298523, "learning_rate": 8.420444731688633e-05, "loss": 0.0386, "step": 29720 }, { "grad_norm": 0.3258633315563202, "learning_rate": 8.419238506117852e-05, "loss": 0.0515, "step": 29730 }, { "grad_norm": 0.5925043225288391, "learning_rate": 8.418031906624289e-05, "loss": 0.0441, "step": 29740 }, { "grad_norm": 0.3510085642337799, "learning_rate": 8.416824933339898e-05, "loss": 0.039, "step": 29750 }, { "grad_norm": 0.4387134909629822, "learning_rate": 8.415617586396667e-05, "loss": 0.0338, "step": 29760 }, { "grad_norm": 0.265630304813385, "learning_rate": 8.414409865926632e-05, "loss": 0.0426, "step": 29770 }, { "grad_norm": 0.38773149251937866, "learning_rate": 8.413201772061867e-05, "loss": 0.0364, "step": 29780 }, { "grad_norm": 0.29500117897987366, "learning_rate": 8.411993304934488e-05, "loss": 0.0332, "step": 29790 }, { "grad_norm": 0.27095627784729004, "learning_rate": 8.410784464676654e-05, "loss": 0.0332, "step": 29800 }, { "grad_norm": 0.352933794260025, "learning_rate": 8.409575251420556e-05, "loss": 0.0363, "step": 29810 }, { "grad_norm": 0.3046327829360962, "learning_rate": 8.408365665298435e-05, "loss": 0.0337, "step": 29820 }, { "grad_norm": 0.7063451409339905, "learning_rate": 8.40715570644257e-05, "loss": 0.0544, "step": 29830 }, { "grad_norm": 0.4335009753704071, "learning_rate": 8.40594537498528e-05, "loss": 0.0515, "step": 29840 }, { "grad_norm": 0.48187392950057983, "learning_rate": 8.404734671058924e-05, "loss": 0.0494, "step": 29850 }, { "grad_norm": 0.6141590476036072, "learning_rate": 8.403523594795902e-05, "loss": 0.0448, "step": 29860 }, { "grad_norm": 0.44082769751548767, "learning_rate": 8.402312146328659e-05, "loss": 0.0377, "step": 29870 }, { "grad_norm": 0.3841117024421692, "learning_rate": 8.401100325789675e-05, "loss": 0.0412, "step": 29880 }, { "grad_norm": 0.6744722723960876, "learning_rate": 8.399888133311472e-05, "loss": 0.0364, "step": 29890 }, { "grad_norm": 0.44797950983047485, "learning_rate": 8.398675569026613e-05, "loss": 0.0506, "step": 29900 }, { "grad_norm": 0.5374785661697388, "learning_rate": 8.397462633067705e-05, "loss": 0.0446, "step": 29910 }, { "grad_norm": 0.36389443278312683, "learning_rate": 8.396249325567392e-05, "loss": 0.0349, "step": 29920 }, { "grad_norm": 0.4750891327857971, "learning_rate": 8.395035646658357e-05, "loss": 0.0552, "step": 29930 }, { "grad_norm": 0.35906049609184265, "learning_rate": 8.39382159647333e-05, "loss": 0.0407, "step": 29940 }, { "grad_norm": 0.2880008816719055, "learning_rate": 8.392607175145075e-05, "loss": 0.0398, "step": 29950 }, { "grad_norm": 0.4093531668186188, "learning_rate": 8.3913923828064e-05, "loss": 0.0436, "step": 29960 }, { "grad_norm": 0.7816265821456909, "learning_rate": 8.390177219590152e-05, "loss": 0.0594, "step": 29970 }, { "grad_norm": 0.5686847567558289, "learning_rate": 8.388961685629222e-05, "loss": 0.0467, "step": 29980 }, { "grad_norm": 0.41749849915504456, "learning_rate": 8.387745781056536e-05, "loss": 0.0465, "step": 29990 }, { "grad_norm": 0.45141369104385376, "learning_rate": 8.386529506005065e-05, "loss": 0.0393, "step": 30000 }, { "grad_norm": 1.0110069513320923, "learning_rate": 8.38531286060782e-05, "loss": 0.0377, "step": 30010 }, { "grad_norm": 0.22698523104190826, "learning_rate": 8.384095844997849e-05, "loss": 0.0335, "step": 30020 }, { "grad_norm": 0.41765326261520386, "learning_rate": 8.382878459308245e-05, "loss": 0.0294, "step": 30030 }, { "grad_norm": 0.5721693634986877, "learning_rate": 8.381660703672138e-05, "loss": 0.042, "step": 30040 }, { "grad_norm": 0.6115756630897522, "learning_rate": 8.380442578222702e-05, "loss": 0.0578, "step": 30050 }, { "grad_norm": 0.36464518308639526, "learning_rate": 8.379224083093146e-05, "loss": 0.0426, "step": 30060 }, { "grad_norm": 0.4641943573951721, "learning_rate": 8.378005218416727e-05, "loss": 0.0485, "step": 30070 }, { "grad_norm": 0.5211101770401001, "learning_rate": 8.376785984326735e-05, "loss": 0.043, "step": 30080 }, { "grad_norm": 0.5023161172866821, "learning_rate": 8.375566380956506e-05, "loss": 0.0354, "step": 30090 }, { "grad_norm": 0.37885281443595886, "learning_rate": 8.374346408439411e-05, "loss": 0.0344, "step": 30100 }, { "grad_norm": 0.33286750316619873, "learning_rate": 8.373126066908868e-05, "loss": 0.041, "step": 30110 }, { "grad_norm": 0.3648279011249542, "learning_rate": 8.371905356498326e-05, "loss": 0.0408, "step": 30120 }, { "grad_norm": 0.7778763771057129, "learning_rate": 8.370684277341288e-05, "loss": 0.0383, "step": 30130 }, { "grad_norm": 0.5571849346160889, "learning_rate": 8.369462829571282e-05, "loss": 0.0397, "step": 30140 }, { "grad_norm": 0.5519721508026123, "learning_rate": 8.36824101332189e-05, "loss": 0.0395, "step": 30150 }, { "grad_norm": 0.35831162333488464, "learning_rate": 8.367018828726721e-05, "loss": 0.0321, "step": 30160 }, { "grad_norm": 0.4341462552547455, "learning_rate": 8.365796275919438e-05, "loss": 0.0378, "step": 30170 }, { "grad_norm": 0.25174275040626526, "learning_rate": 8.364573355033734e-05, "loss": 0.0286, "step": 30180 }, { "grad_norm": 0.37102600932121277, "learning_rate": 8.363350066203346e-05, "loss": 0.0302, "step": 30190 }, { "grad_norm": 0.4526326656341553, "learning_rate": 8.362126409562053e-05, "loss": 0.0459, "step": 30200 }, { "grad_norm": 0.37888625264167786, "learning_rate": 8.360902385243667e-05, "loss": 0.0385, "step": 30210 }, { "grad_norm": 0.3692619502544403, "learning_rate": 8.359677993382052e-05, "loss": 0.0319, "step": 30220 }, { "grad_norm": 0.9599718451499939, "learning_rate": 8.358453234111103e-05, "loss": 0.0357, "step": 30230 }, { "grad_norm": 0.3954201936721802, "learning_rate": 8.357228107564756e-05, "loss": 0.049, "step": 30240 }, { "grad_norm": 0.2803926467895508, "learning_rate": 8.356002613876993e-05, "loss": 0.0356, "step": 30250 }, { "grad_norm": 0.40922725200653076, "learning_rate": 8.35477675318183e-05, "loss": 0.0379, "step": 30260 }, { "grad_norm": 0.5009543299674988, "learning_rate": 8.353550525613323e-05, "loss": 0.0376, "step": 30270 }, { "grad_norm": 0.41552600264549255, "learning_rate": 8.352323931305572e-05, "loss": 0.0566, "step": 30280 }, { "grad_norm": 0.3967638909816742, "learning_rate": 8.351096970392717e-05, "loss": 0.0388, "step": 30290 }, { "grad_norm": 0.43515217304229736, "learning_rate": 8.349869643008937e-05, "loss": 0.0384, "step": 30300 }, { "grad_norm": 0.4311642348766327, "learning_rate": 8.348641949288449e-05, "loss": 0.047, "step": 30310 }, { "grad_norm": 0.42438992857933044, "learning_rate": 8.34741388936551e-05, "loss": 0.0372, "step": 30320 }, { "grad_norm": 0.5261125564575195, "learning_rate": 8.346185463374423e-05, "loss": 0.0419, "step": 30330 }, { "grad_norm": 0.38545241951942444, "learning_rate": 8.344956671449524e-05, "loss": 0.0358, "step": 30340 }, { "grad_norm": 0.3030538558959961, "learning_rate": 8.343727513725192e-05, "loss": 0.0286, "step": 30350 }, { "grad_norm": 0.41504237055778503, "learning_rate": 8.342497990335847e-05, "loss": 0.0415, "step": 30360 }, { "grad_norm": 0.439455509185791, "learning_rate": 8.341268101415946e-05, "loss": 0.046, "step": 30370 }, { "grad_norm": 0.3128485381603241, "learning_rate": 8.34003784709999e-05, "loss": 0.0372, "step": 30380 }, { "grad_norm": 0.6834003925323486, "learning_rate": 8.338807227522516e-05, "loss": 0.0469, "step": 30390 }, { "grad_norm": 0.4230023920536041, "learning_rate": 8.337576242818103e-05, "loss": 0.0394, "step": 30400 }, { "grad_norm": 0.508165180683136, "learning_rate": 8.336344893121372e-05, "loss": 0.0385, "step": 30410 }, { "grad_norm": 0.43569642305374146, "learning_rate": 8.335113178566977e-05, "loss": 0.0395, "step": 30420 }, { "grad_norm": 0.298811674118042, "learning_rate": 8.33388109928962e-05, "loss": 0.0422, "step": 30430 }, { "grad_norm": 0.5103329420089722, "learning_rate": 8.33264865542404e-05, "loss": 0.0385, "step": 30440 }, { "grad_norm": 0.37840497493743896, "learning_rate": 8.331415847105013e-05, "loss": 0.0404, "step": 30450 }, { "grad_norm": 0.39820629358291626, "learning_rate": 8.330182674467357e-05, "loss": 0.0486, "step": 30460 }, { "grad_norm": 0.5529724955558777, "learning_rate": 8.32894913764593e-05, "loss": 0.0299, "step": 30470 }, { "grad_norm": 0.748134195804596, "learning_rate": 8.327715236775633e-05, "loss": 0.0435, "step": 30480 }, { "grad_norm": 0.34334972500801086, "learning_rate": 8.326480971991398e-05, "loss": 0.0401, "step": 30490 }, { "grad_norm": 0.32133135199546814, "learning_rate": 8.325246343428206e-05, "loss": 0.0406, "step": 30500 }, { "grad_norm": 0.4576665163040161, "learning_rate": 8.324011351221072e-05, "loss": 0.0371, "step": 30510 }, { "grad_norm": 0.6996537446975708, "learning_rate": 8.322775995505057e-05, "loss": 0.0372, "step": 30520 }, { "grad_norm": 0.3333892822265625, "learning_rate": 8.321540276415254e-05, "loss": 0.0406, "step": 30530 }, { "grad_norm": 0.4753395915031433, "learning_rate": 8.320304194086798e-05, "loss": 0.0434, "step": 30540 }, { "grad_norm": 0.3465922176837921, "learning_rate": 8.31906774865487e-05, "loss": 0.0359, "step": 30550 }, { "grad_norm": 0.3570142686367035, "learning_rate": 8.317830940254682e-05, "loss": 0.0328, "step": 30560 }, { "grad_norm": 0.6208869814872742, "learning_rate": 8.316593769021491e-05, "loss": 0.0433, "step": 30570 }, { "grad_norm": 0.4546583890914917, "learning_rate": 8.315356235090592e-05, "loss": 0.0488, "step": 30580 }, { "grad_norm": 0.3940684199333191, "learning_rate": 8.314118338597319e-05, "loss": 0.0369, "step": 30590 }, { "grad_norm": 0.3895796537399292, "learning_rate": 8.312880079677048e-05, "loss": 0.0443, "step": 30600 }, { "grad_norm": 0.4831746816635132, "learning_rate": 8.311641458465191e-05, "loss": 0.0348, "step": 30610 }, { "grad_norm": 0.42534059286117554, "learning_rate": 8.310402475097205e-05, "loss": 0.0445, "step": 30620 }, { "grad_norm": 0.4031883180141449, "learning_rate": 8.309163129708581e-05, "loss": 0.0359, "step": 30630 }, { "grad_norm": 0.4392654597759247, "learning_rate": 8.307923422434852e-05, "loss": 0.0369, "step": 30640 }, { "grad_norm": 0.4791911244392395, "learning_rate": 8.30668335341159e-05, "loss": 0.0415, "step": 30650 }, { "grad_norm": 0.38122549653053284, "learning_rate": 8.305442922774408e-05, "loss": 0.0383, "step": 30660 }, { "grad_norm": 0.2969455420970917, "learning_rate": 8.304202130658959e-05, "loss": 0.0328, "step": 30670 }, { "grad_norm": 0.3187934160232544, "learning_rate": 8.302960977200931e-05, "loss": 0.0373, "step": 30680 }, { "grad_norm": 0.42549505829811096, "learning_rate": 8.301719462536058e-05, "loss": 0.0344, "step": 30690 }, { "grad_norm": 0.3804838955402374, "learning_rate": 8.300477586800108e-05, "loss": 0.0442, "step": 30700 }, { "grad_norm": 0.44789522886276245, "learning_rate": 8.299235350128892e-05, "loss": 0.0377, "step": 30710 }, { "grad_norm": 0.4714815020561218, "learning_rate": 8.297992752658258e-05, "loss": 0.041, "step": 30720 }, { "grad_norm": 0.35994017124176025, "learning_rate": 8.296749794524094e-05, "loss": 0.0483, "step": 30730 }, { "grad_norm": 0.29765585064888, "learning_rate": 8.295506475862332e-05, "loss": 0.0381, "step": 30740 }, { "grad_norm": 0.33993643522262573, "learning_rate": 8.294262796808933e-05, "loss": 0.0374, "step": 30750 }, { "grad_norm": 0.5907772779464722, "learning_rate": 8.293018757499909e-05, "loss": 0.0344, "step": 30760 }, { "grad_norm": 0.31572848558425903, "learning_rate": 8.291774358071305e-05, "loss": 0.0303, "step": 30770 }, { "grad_norm": 0.4461974799633026, "learning_rate": 8.290529598659205e-05, "loss": 0.0373, "step": 30780 }, { "grad_norm": 0.4549625813961029, "learning_rate": 8.289284479399738e-05, "loss": 0.0373, "step": 30790 }, { "grad_norm": 0.35472381114959717, "learning_rate": 8.288039000429064e-05, "loss": 0.0372, "step": 30800 }, { "grad_norm": 0.3583495318889618, "learning_rate": 8.286793161883388e-05, "loss": 0.0341, "step": 30810 }, { "grad_norm": 0.43016085028648376, "learning_rate": 8.285546963898954e-05, "loss": 0.0364, "step": 30820 }, { "grad_norm": 0.35730892419815063, "learning_rate": 8.284300406612044e-05, "loss": 0.0401, "step": 30830 }, { "grad_norm": 0.3534577488899231, "learning_rate": 8.283053490158978e-05, "loss": 0.0365, "step": 30840 }, { "grad_norm": 0.2956555485725403, "learning_rate": 8.28180621467612e-05, "loss": 0.0294, "step": 30850 }, { "grad_norm": 0.36217620968818665, "learning_rate": 8.280558580299867e-05, "loss": 0.0431, "step": 30860 }, { "grad_norm": 0.3212762475013733, "learning_rate": 8.279310587166661e-05, "loss": 0.0308, "step": 30870 }, { "grad_norm": 0.9783467054367065, "learning_rate": 8.278062235412978e-05, "loss": 0.0314, "step": 30880 }, { "grad_norm": 0.42449459433555603, "learning_rate": 8.276813525175339e-05, "loss": 0.0402, "step": 30890 }, { "grad_norm": 0.4191407263278961, "learning_rate": 8.2755644565903e-05, "loss": 0.0465, "step": 30900 }, { "grad_norm": 0.5128486752510071, "learning_rate": 8.274315029794454e-05, "loss": 0.0449, "step": 30910 }, { "grad_norm": 0.635525107383728, "learning_rate": 8.273065244924443e-05, "loss": 0.0454, "step": 30920 }, { "grad_norm": 0.5189266204833984, "learning_rate": 8.271815102116936e-05, "loss": 0.0443, "step": 30930 }, { "grad_norm": 0.44658488035202026, "learning_rate": 8.270564601508648e-05, "loss": 0.0382, "step": 30940 }, { "grad_norm": 0.5159681439399719, "learning_rate": 8.269313743236333e-05, "loss": 0.0461, "step": 30950 }, { "grad_norm": 0.44649434089660645, "learning_rate": 8.268062527436783e-05, "loss": 0.038, "step": 30960 }, { "grad_norm": 0.38115087151527405, "learning_rate": 8.266810954246828e-05, "loss": 0.0408, "step": 30970 }, { "grad_norm": 0.3659704625606537, "learning_rate": 8.265559023803338e-05, "loss": 0.035, "step": 30980 }, { "grad_norm": 0.495536744594574, "learning_rate": 8.264306736243223e-05, "loss": 0.0423, "step": 30990 }, { "grad_norm": 0.4974904954433441, "learning_rate": 8.263054091703432e-05, "loss": 0.0428, "step": 31000 }, { "grad_norm": 0.3938705325126648, "learning_rate": 8.26180109032095e-05, "loss": 0.0417, "step": 31010 }, { "grad_norm": 0.4565717279911041, "learning_rate": 8.260547732232807e-05, "loss": 0.0308, "step": 31020 }, { "grad_norm": 0.440483957529068, "learning_rate": 8.259294017576063e-05, "loss": 0.0363, "step": 31030 }, { "grad_norm": 0.5071736574172974, "learning_rate": 8.258039946487828e-05, "loss": 0.0435, "step": 31040 }, { "grad_norm": 0.37694722414016724, "learning_rate": 8.256785519105241e-05, "loss": 0.0327, "step": 31050 }, { "grad_norm": 0.25072354078292847, "learning_rate": 8.255530735565488e-05, "loss": 0.0322, "step": 31060 }, { "grad_norm": 0.5464881658554077, "learning_rate": 8.254275596005787e-05, "loss": 0.0371, "step": 31070 }, { "grad_norm": 0.4184740483760834, "learning_rate": 8.2530201005634e-05, "loss": 0.031, "step": 31080 }, { "grad_norm": 0.3811286389827728, "learning_rate": 8.251764249375626e-05, "loss": 0.0417, "step": 31090 }, { "grad_norm": 0.3304554522037506, "learning_rate": 8.250508042579803e-05, "loss": 0.0343, "step": 31100 }, { "grad_norm": 0.47010356187820435, "learning_rate": 8.249251480313307e-05, "loss": 0.0318, "step": 31110 }, { "grad_norm": 0.6181210279464722, "learning_rate": 8.247994562713555e-05, "loss": 0.0362, "step": 31120 }, { "grad_norm": 0.3853655457496643, "learning_rate": 8.246737289917998e-05, "loss": 0.0351, "step": 31130 }, { "grad_norm": 0.35245683789253235, "learning_rate": 8.245479662064135e-05, "loss": 0.0358, "step": 31140 }, { "grad_norm": 0.32111474871635437, "learning_rate": 8.244221679289496e-05, "loss": 0.0354, "step": 31150 }, { "grad_norm": 0.3407772183418274, "learning_rate": 8.242963341731652e-05, "loss": 0.0404, "step": 31160 }, { "grad_norm": 0.33514052629470825, "learning_rate": 8.24170464952821e-05, "loss": 0.0364, "step": 31170 }, { "grad_norm": 0.38444551825523376, "learning_rate": 8.240445602816824e-05, "loss": 0.0502, "step": 31180 }, { "grad_norm": 0.43970462679862976, "learning_rate": 8.239186201735179e-05, "loss": 0.0392, "step": 31190 }, { "grad_norm": 0.37771961092948914, "learning_rate": 8.237926446420998e-05, "loss": 0.034, "step": 31200 }, { "grad_norm": 0.4106801748275757, "learning_rate": 8.236666337012052e-05, "loss": 0.0413, "step": 31210 }, { "grad_norm": 0.2961646020412445, "learning_rate": 8.23540587364614e-05, "loss": 0.0339, "step": 31220 }, { "grad_norm": 0.4191780984401703, "learning_rate": 8.234145056461107e-05, "loss": 0.0424, "step": 31230 }, { "grad_norm": 0.3669520616531372, "learning_rate": 8.232883885594832e-05, "loss": 0.0409, "step": 31240 }, { "grad_norm": 0.35520923137664795, "learning_rate": 8.231622361185236e-05, "loss": 0.029, "step": 31250 }, { "grad_norm": 0.550460159778595, "learning_rate": 8.230360483370278e-05, "loss": 0.0598, "step": 31260 }, { "grad_norm": 0.25890737771987915, "learning_rate": 8.229098252287953e-05, "loss": 0.0271, "step": 31270 }, { "grad_norm": 0.5570688247680664, "learning_rate": 8.2278356680763e-05, "loss": 0.0427, "step": 31280 }, { "grad_norm": 0.5075281858444214, "learning_rate": 8.22657273087339e-05, "loss": 0.0425, "step": 31290 }, { "grad_norm": 0.5322659611701965, "learning_rate": 8.225309440817336e-05, "loss": 0.0504, "step": 31300 }, { "grad_norm": 0.28257226943969727, "learning_rate": 8.224045798046293e-05, "loss": 0.0545, "step": 31310 }, { "grad_norm": 0.4499441981315613, "learning_rate": 8.22278180269845e-05, "loss": 0.0333, "step": 31320 }, { "grad_norm": 0.3433478772640228, "learning_rate": 8.221517454912031e-05, "loss": 0.0321, "step": 31330 }, { "grad_norm": 0.47266656160354614, "learning_rate": 8.220252754825308e-05, "loss": 0.0323, "step": 31340 }, { "grad_norm": 0.5623086094856262, "learning_rate": 8.218987702576586e-05, "loss": 0.0388, "step": 31350 }, { "grad_norm": 0.8127062320709229, "learning_rate": 8.217722298304207e-05, "loss": 0.0447, "step": 31360 }, { "grad_norm": 0.711675763130188, "learning_rate": 8.216456542146557e-05, "loss": 0.0492, "step": 31370 }, { "grad_norm": 0.4870883524417877, "learning_rate": 8.215190434242055e-05, "loss": 0.0418, "step": 31380 }, { "grad_norm": 0.4187343716621399, "learning_rate": 8.213923974729161e-05, "loss": 0.0437, "step": 31390 }, { "grad_norm": 0.4346155524253845, "learning_rate": 8.212657163746373e-05, "loss": 0.043, "step": 31400 }, { "grad_norm": 0.43418243527412415, "learning_rate": 8.211390001432227e-05, "loss": 0.0387, "step": 31410 }, { "grad_norm": 1.1856324672698975, "learning_rate": 8.210122487925297e-05, "loss": 0.0444, "step": 31420 }, { "grad_norm": 0.4458838701248169, "learning_rate": 8.208854623364202e-05, "loss": 0.0391, "step": 31430 }, { "grad_norm": 0.3216419219970703, "learning_rate": 8.207586407887589e-05, "loss": 0.0375, "step": 31440 }, { "grad_norm": 0.32373350858688354, "learning_rate": 8.206317841634148e-05, "loss": 0.0287, "step": 31450 }, { "grad_norm": 0.3923482596874237, "learning_rate": 8.205048924742609e-05, "loss": 0.038, "step": 31460 }, { "grad_norm": 0.3589133620262146, "learning_rate": 8.203779657351738e-05, "loss": 0.0302, "step": 31470 }, { "grad_norm": 0.7489906549453735, "learning_rate": 8.20251003960034e-05, "loss": 0.0391, "step": 31480 }, { "grad_norm": 0.3329634666442871, "learning_rate": 8.201240071627258e-05, "loss": 0.0391, "step": 31490 }, { "grad_norm": 0.32206544280052185, "learning_rate": 8.199969753571377e-05, "loss": 0.0421, "step": 31500 }, { "grad_norm": 0.36096781492233276, "learning_rate": 8.198699085571615e-05, "loss": 0.0382, "step": 31510 }, { "grad_norm": 0.43940773606300354, "learning_rate": 8.197428067766928e-05, "loss": 0.0327, "step": 31520 }, { "grad_norm": 0.4482037425041199, "learning_rate": 8.196156700296316e-05, "loss": 0.0454, "step": 31530 }, { "grad_norm": 0.9005807042121887, "learning_rate": 8.194884983298814e-05, "loss": 0.0369, "step": 31540 }, { "grad_norm": 0.4788748025894165, "learning_rate": 8.193612916913491e-05, "loss": 0.0353, "step": 31550 }, { "grad_norm": 0.3473060131072998, "learning_rate": 8.192340501279463e-05, "loss": 0.0353, "step": 31560 }, { "grad_norm": 0.40709182620048523, "learning_rate": 8.191067736535876e-05, "loss": 0.0371, "step": 31570 }, { "grad_norm": 0.507091224193573, "learning_rate": 8.18979462282192e-05, "loss": 0.0338, "step": 31580 }, { "grad_norm": 0.5063950419425964, "learning_rate": 8.188521160276819e-05, "loss": 0.0354, "step": 31590 }, { "grad_norm": 0.9413139224052429, "learning_rate": 8.187247349039837e-05, "loss": 0.0319, "step": 31600 }, { "grad_norm": 0.4619961380958557, "learning_rate": 8.185973189250278e-05, "loss": 0.044, "step": 31610 }, { "grad_norm": 0.45162731409072876, "learning_rate": 8.184698681047482e-05, "loss": 0.0427, "step": 31620 }, { "grad_norm": 0.38111838698387146, "learning_rate": 8.183423824570827e-05, "loss": 0.0479, "step": 31630 }, { "grad_norm": 0.2873111367225647, "learning_rate": 8.182148619959725e-05, "loss": 0.049, "step": 31640 }, { "grad_norm": 0.28476208448410034, "learning_rate": 8.180873067353636e-05, "loss": 0.0375, "step": 31650 }, { "grad_norm": 0.5411402583122253, "learning_rate": 8.179597166892052e-05, "loss": 0.0392, "step": 31660 }, { "grad_norm": 0.5499514937400818, "learning_rate": 8.178320918714501e-05, "loss": 0.0365, "step": 31670 }, { "grad_norm": 0.35267916321754456, "learning_rate": 8.177044322960554e-05, "loss": 0.0368, "step": 31680 }, { "grad_norm": 0.3547380268573761, "learning_rate": 8.175767379769816e-05, "loss": 0.0306, "step": 31690 }, { "grad_norm": 0.38063955307006836, "learning_rate": 8.174490089281932e-05, "loss": 0.039, "step": 31700 }, { "grad_norm": 0.36628976464271545, "learning_rate": 8.173212451636584e-05, "loss": 0.0353, "step": 31710 }, { "grad_norm": 0.5600707530975342, "learning_rate": 8.171934466973493e-05, "loss": 0.0393, "step": 31720 }, { "grad_norm": 0.39839285612106323, "learning_rate": 8.170656135432418e-05, "loss": 0.0387, "step": 31730 }, { "grad_norm": 0.5093210339546204, "learning_rate": 8.169377457153155e-05, "loss": 0.0439, "step": 31740 }, { "grad_norm": 0.40373048186302185, "learning_rate": 8.168098432275539e-05, "loss": 0.0383, "step": 31750 }, { "grad_norm": 0.4282967448234558, "learning_rate": 8.166819060939442e-05, "loss": 0.0507, "step": 31760 }, { "grad_norm": 0.3478568196296692, "learning_rate": 8.165539343284772e-05, "loss": 0.0413, "step": 31770 }, { "grad_norm": 0.37167060375213623, "learning_rate": 8.16425927945148e-05, "loss": 0.0388, "step": 31780 }, { "grad_norm": 0.34920018911361694, "learning_rate": 8.162978869579551e-05, "loss": 0.0501, "step": 31790 }, { "grad_norm": 0.42806771397590637, "learning_rate": 8.161698113809007e-05, "loss": 0.0376, "step": 31800 }, { "grad_norm": 0.3086654841899872, "learning_rate": 8.160417012279911e-05, "loss": 0.0475, "step": 31810 }, { "grad_norm": 0.38384562730789185, "learning_rate": 8.159135565132363e-05, "loss": 0.0395, "step": 31820 }, { "grad_norm": 0.4375254213809967, "learning_rate": 8.157853772506498e-05, "loss": 0.0436, "step": 31830 }, { "grad_norm": 0.5635280013084412, "learning_rate": 8.156571634542494e-05, "loss": 0.0485, "step": 31840 }, { "grad_norm": 0.39574700593948364, "learning_rate": 8.15528915138056e-05, "loss": 0.0375, "step": 31850 }, { "grad_norm": 0.37972167134284973, "learning_rate": 8.154006323160949e-05, "loss": 0.038, "step": 31860 }, { "grad_norm": 0.6039433479309082, "learning_rate": 8.152723150023949e-05, "loss": 0.0424, "step": 31870 }, { "grad_norm": 0.4918915331363678, "learning_rate": 8.151439632109886e-05, "loss": 0.0437, "step": 31880 }, { "grad_norm": 0.43093347549438477, "learning_rate": 8.150155769559122e-05, "loss": 0.0513, "step": 31890 }, { "grad_norm": 0.4367399215698242, "learning_rate": 8.148871562512058e-05, "loss": 0.0335, "step": 31900 }, { "grad_norm": 0.29789361357688904, "learning_rate": 8.147587011109136e-05, "loss": 0.0403, "step": 31910 }, { "grad_norm": 0.3735589385032654, "learning_rate": 8.14630211549083e-05, "loss": 0.0532, "step": 31920 }, { "grad_norm": 0.38972699642181396, "learning_rate": 8.145016875797655e-05, "loss": 0.0318, "step": 31930 }, { "grad_norm": 0.30550581216812134, "learning_rate": 8.143731292170164e-05, "loss": 0.0313, "step": 31940 }, { "grad_norm": 0.4236713945865631, "learning_rate": 8.142445364748944e-05, "loss": 0.0463, "step": 31950 }, { "grad_norm": 0.5403287410736084, "learning_rate": 8.141159093674624e-05, "loss": 0.0397, "step": 31960 }, { "grad_norm": 0.256056010723114, "learning_rate": 8.139872479087869e-05, "loss": 0.032, "step": 31970 }, { "grad_norm": 0.35338452458381653, "learning_rate": 8.13858552112938e-05, "loss": 0.0282, "step": 31980 }, { "grad_norm": 0.37005123496055603, "learning_rate": 8.137298219939895e-05, "loss": 0.0484, "step": 31990 }, { "grad_norm": 0.3093550205230713, "learning_rate": 8.136010575660196e-05, "loss": 0.0294, "step": 32000 }, { "grad_norm": 0.361117959022522, "learning_rate": 8.134722588431095e-05, "loss": 0.0414, "step": 32010 }, { "grad_norm": 0.74563068151474, "learning_rate": 8.133434258393444e-05, "loss": 0.0348, "step": 32020 }, { "grad_norm": 0.4969033896923065, "learning_rate": 8.132145585688134e-05, "loss": 0.0436, "step": 32030 }, { "grad_norm": 0.421575665473938, "learning_rate": 8.130856570456093e-05, "loss": 0.0427, "step": 32040 }, { "grad_norm": 0.7171282768249512, "learning_rate": 8.129567212838283e-05, "loss": 0.047, "step": 32050 }, { "grad_norm": 0.5567159652709961, "learning_rate": 8.128277512975708e-05, "loss": 0.0439, "step": 32060 }, { "grad_norm": 0.3332096040248871, "learning_rate": 8.126987471009408e-05, "loss": 0.0424, "step": 32070 }, { "grad_norm": 0.45624494552612305, "learning_rate": 8.125697087080459e-05, "loss": 0.0303, "step": 32080 }, { "grad_norm": 0.44824036955833435, "learning_rate": 8.124406361329976e-05, "loss": 0.0527, "step": 32090 }, { "grad_norm": 0.5129711627960205, "learning_rate": 8.123115293899111e-05, "loss": 0.0431, "step": 32100 }, { "grad_norm": 0.5314806699752808, "learning_rate": 8.121823884929055e-05, "loss": 0.0466, "step": 32110 }, { "grad_norm": 0.40068426728248596, "learning_rate": 8.12053213456103e-05, "loss": 0.0348, "step": 32120 }, { "grad_norm": 0.5327275991439819, "learning_rate": 8.119240042936303e-05, "loss": 0.0429, "step": 32130 }, { "grad_norm": 0.4204707145690918, "learning_rate": 8.117947610196175e-05, "loss": 0.0437, "step": 32140 }, { "grad_norm": 0.42767468094825745, "learning_rate": 8.116654836481982e-05, "loss": 0.0412, "step": 32150 }, { "grad_norm": 0.4772699773311615, "learning_rate": 8.115361721935106e-05, "loss": 0.0344, "step": 32160 }, { "grad_norm": 0.5143534541130066, "learning_rate": 8.114068266696953e-05, "loss": 0.0334, "step": 32170 }, { "grad_norm": 0.4155735969543457, "learning_rate": 8.112774470908978e-05, "loss": 0.0305, "step": 32180 }, { "grad_norm": 0.34921225905418396, "learning_rate": 8.111480334712665e-05, "loss": 0.0353, "step": 32190 }, { "grad_norm": 0.3998219668865204, "learning_rate": 8.110185858249542e-05, "loss": 0.0381, "step": 32200 }, { "grad_norm": 0.4891396164894104, "learning_rate": 8.108891041661168e-05, "loss": 0.0366, "step": 32210 }, { "grad_norm": 0.5003045797348022, "learning_rate": 8.107595885089146e-05, "loss": 0.0345, "step": 32220 }, { "grad_norm": 0.7054330706596375, "learning_rate": 8.106300388675108e-05, "loss": 0.036, "step": 32230 }, { "grad_norm": 0.37893790006637573, "learning_rate": 8.105004552560731e-05, "loss": 0.0432, "step": 32240 }, { "grad_norm": 0.49114665389060974, "learning_rate": 8.103708376887724e-05, "loss": 0.0431, "step": 32250 }, { "grad_norm": 0.5382001399993896, "learning_rate": 8.102411861797836e-05, "loss": 0.0416, "step": 32260 }, { "grad_norm": 0.5552709698677063, "learning_rate": 8.101115007432851e-05, "loss": 0.0417, "step": 32270 }, { "grad_norm": 0.36770567297935486, "learning_rate": 8.09981781393459e-05, "loss": 0.0329, "step": 32280 }, { "grad_norm": 0.46392565965652466, "learning_rate": 8.098520281444915e-05, "loss": 0.0389, "step": 32290 }, { "grad_norm": 0.4020409882068634, "learning_rate": 8.09722241010572e-05, "loss": 0.0339, "step": 32300 }, { "grad_norm": 0.36610469222068787, "learning_rate": 8.095924200058939e-05, "loss": 0.0435, "step": 32310 }, { "grad_norm": 0.2916010916233063, "learning_rate": 8.094625651446541e-05, "loss": 0.0277, "step": 32320 }, { "grad_norm": 0.3683401644229889, "learning_rate": 8.093326764410536e-05, "loss": 0.0298, "step": 32330 }, { "grad_norm": 0.7328314185142517, "learning_rate": 8.092027539092966e-05, "loss": 0.0343, "step": 32340 }, { "grad_norm": 0.5936660766601562, "learning_rate": 8.090727975635913e-05, "loss": 0.0373, "step": 32350 }, { "grad_norm": 0.3787858784198761, "learning_rate": 8.089428074181497e-05, "loss": 0.0385, "step": 32360 }, { "grad_norm": 0.38363584876060486, "learning_rate": 8.088127834871871e-05, "loss": 0.0314, "step": 32370 }, { "grad_norm": 0.3409225046634674, "learning_rate": 8.086827257849226e-05, "loss": 0.0365, "step": 32380 }, { "grad_norm": 0.4186847507953644, "learning_rate": 8.085526343255795e-05, "loss": 0.0314, "step": 32390 }, { "grad_norm": 0.49223965406417847, "learning_rate": 8.084225091233842e-05, "loss": 0.0393, "step": 32400 }, { "grad_norm": 0.4698355793952942, "learning_rate": 8.082923501925668e-05, "loss": 0.0424, "step": 32410 }, { "grad_norm": 0.43973493576049805, "learning_rate": 8.081621575473617e-05, "loss": 0.0349, "step": 32420 }, { "grad_norm": 0.3940054774284363, "learning_rate": 8.080319312020064e-05, "loss": 0.0418, "step": 32430 }, { "grad_norm": 0.26051241159439087, "learning_rate": 8.079016711707421e-05, "loss": 0.0315, "step": 32440 }, { "grad_norm": 0.2817424237728119, "learning_rate": 8.077713774678139e-05, "loss": 0.0353, "step": 32450 }, { "grad_norm": 0.3639436364173889, "learning_rate": 8.076410501074707e-05, "loss": 0.038, "step": 32460 }, { "grad_norm": 0.35372111201286316, "learning_rate": 8.075106891039647e-05, "loss": 0.036, "step": 32470 }, { "grad_norm": 0.3554564118385315, "learning_rate": 8.073802944715523e-05, "loss": 0.0379, "step": 32480 }, { "grad_norm": 0.4338301420211792, "learning_rate": 8.072498662244929e-05, "loss": 0.0401, "step": 32490 }, { "grad_norm": 0.34555160999298096, "learning_rate": 8.0711940437705e-05, "loss": 0.0367, "step": 32500 }, { "grad_norm": 0.5884939432144165, "learning_rate": 8.06988908943491e-05, "loss": 0.0326, "step": 32510 }, { "grad_norm": 0.4802248477935791, "learning_rate": 8.068583799380863e-05, "loss": 0.035, "step": 32520 }, { "grad_norm": 0.37903285026550293, "learning_rate": 8.067278173751104e-05, "loss": 0.0343, "step": 32530 }, { "grad_norm": 0.4750746488571167, "learning_rate": 8.065972212688417e-05, "loss": 0.036, "step": 32540 }, { "grad_norm": 0.5045725107192993, "learning_rate": 8.064665916335618e-05, "loss": 0.0372, "step": 32550 }, { "grad_norm": 0.3780130445957184, "learning_rate": 8.063359284835564e-05, "loss": 0.0394, "step": 32560 }, { "grad_norm": 0.32282349467277527, "learning_rate": 8.062052318331142e-05, "loss": 0.0363, "step": 32570 }, { "grad_norm": 0.36641377210617065, "learning_rate": 8.060745016965283e-05, "loss": 0.0419, "step": 32580 }, { "grad_norm": 0.8132416605949402, "learning_rate": 8.059437380880952e-05, "loss": 0.0395, "step": 32590 }, { "grad_norm": 0.44554921984672546, "learning_rate": 8.058129410221146e-05, "loss": 0.0397, "step": 32600 }, { "grad_norm": 0.3437708616256714, "learning_rate": 8.056821105128908e-05, "loss": 0.041, "step": 32610 }, { "grad_norm": 0.39122435450553894, "learning_rate": 8.05551246574731e-05, "loss": 0.0401, "step": 32620 }, { "grad_norm": 0.44662487506866455, "learning_rate": 8.05420349221946e-05, "loss": 0.0376, "step": 32630 }, { "grad_norm": 0.2691766619682312, "learning_rate": 8.05289418468851e-05, "loss": 0.0385, "step": 32640 }, { "grad_norm": 0.32403847575187683, "learning_rate": 8.051584543297642e-05, "loss": 0.0306, "step": 32650 }, { "grad_norm": 0.47253352403640747, "learning_rate": 8.050274568190074e-05, "loss": 0.0413, "step": 32660 }, { "grad_norm": 0.34952059388160706, "learning_rate": 8.048964259509067e-05, "loss": 0.0509, "step": 32670 }, { "grad_norm": 0.4984720051288605, "learning_rate": 8.047653617397914e-05, "loss": 0.0373, "step": 32680 }, { "grad_norm": 0.7019910216331482, "learning_rate": 8.046342641999941e-05, "loss": 0.0436, "step": 32690 }, { "grad_norm": 0.44891074299812317, "learning_rate": 8.045031333458517e-05, "loss": 0.0373, "step": 32700 }, { "grad_norm": 0.4778524935245514, "learning_rate": 8.043719691917047e-05, "loss": 0.0456, "step": 32710 }, { "grad_norm": 0.24816522002220154, "learning_rate": 8.042407717518966e-05, "loss": 0.038, "step": 32720 }, { "grad_norm": 0.3046504557132721, "learning_rate": 8.041095410407751e-05, "loss": 0.0341, "step": 32730 }, { "grad_norm": 0.500836193561554, "learning_rate": 8.039782770726913e-05, "loss": 0.0429, "step": 32740 }, { "grad_norm": 0.4912228286266327, "learning_rate": 8.038469798620004e-05, "loss": 0.035, "step": 32750 }, { "grad_norm": 0.48662328720092773, "learning_rate": 8.037156494230604e-05, "loss": 0.0379, "step": 32760 }, { "grad_norm": 0.3535315990447998, "learning_rate": 8.035842857702338e-05, "loss": 0.0388, "step": 32770 }, { "grad_norm": 0.5355250835418701, "learning_rate": 8.03452888917886e-05, "loss": 0.0422, "step": 32780 }, { "grad_norm": 0.32527750730514526, "learning_rate": 8.033214588803866e-05, "loss": 0.0346, "step": 32790 }, { "grad_norm": 0.47862708568573, "learning_rate": 8.031899956721083e-05, "loss": 0.0331, "step": 32800 }, { "grad_norm": 0.38441139459609985, "learning_rate": 8.030584993074282e-05, "loss": 0.0338, "step": 32810 }, { "grad_norm": 0.442984938621521, "learning_rate": 8.02926969800726e-05, "loss": 0.0309, "step": 32820 }, { "grad_norm": 0.25958263874053955, "learning_rate": 8.027954071663859e-05, "loss": 0.0418, "step": 32830 }, { "grad_norm": 0.42237311601638794, "learning_rate": 8.026638114187954e-05, "loss": 0.0322, "step": 32840 }, { "grad_norm": 0.5933108329772949, "learning_rate": 8.025321825723456e-05, "loss": 0.0401, "step": 32850 }, { "grad_norm": 0.41674405336380005, "learning_rate": 8.02400520641431e-05, "loss": 0.0413, "step": 32860 }, { "grad_norm": 0.3249538242816925, "learning_rate": 8.022688256404501e-05, "loss": 0.0311, "step": 32870 }, { "grad_norm": 0.3141922950744629, "learning_rate": 8.02137097583805e-05, "loss": 0.032, "step": 32880 }, { "grad_norm": 0.31322920322418213, "learning_rate": 8.02005336485901e-05, "loss": 0.0296, "step": 32890 }, { "grad_norm": 0.4165521264076233, "learning_rate": 8.018735423611476e-05, "loss": 0.044, "step": 32900 }, { "grad_norm": 0.33656591176986694, "learning_rate": 8.017417152239574e-05, "loss": 0.0382, "step": 32910 }, { "grad_norm": 0.36195820569992065, "learning_rate": 8.01609855088747e-05, "loss": 0.0373, "step": 32920 }, { "grad_norm": 0.31453439593315125, "learning_rate": 8.014779619699362e-05, "loss": 0.0299, "step": 32930 }, { "grad_norm": 0.30872827768325806, "learning_rate": 8.013460358819489e-05, "loss": 0.0288, "step": 32940 }, { "grad_norm": 0.34489011764526367, "learning_rate": 8.01214076839212e-05, "loss": 0.0397, "step": 32950 }, { "grad_norm": 0.2804045081138611, "learning_rate": 8.010820848561565e-05, "loss": 0.0318, "step": 32960 }, { "grad_norm": 0.3463045358657837, "learning_rate": 8.009500599472171e-05, "loss": 0.0403, "step": 32970 }, { "grad_norm": 0.6896330118179321, "learning_rate": 8.008180021268314e-05, "loss": 0.0333, "step": 32980 }, { "grad_norm": 0.44040241837501526, "learning_rate": 8.006859114094414e-05, "loss": 0.0289, "step": 32990 }, { "grad_norm": 0.41509807109832764, "learning_rate": 8.005537878094921e-05, "loss": 0.0402, "step": 33000 }, { "grad_norm": 0.36682242155075073, "learning_rate": 8.004216313414323e-05, "loss": 0.04, "step": 33010 }, { "grad_norm": 0.4415767192840576, "learning_rate": 8.002894420197149e-05, "loss": 0.0393, "step": 33020 }, { "grad_norm": 0.4540393054485321, "learning_rate": 8.001572198587954e-05, "loss": 0.0391, "step": 33030 }, { "grad_norm": 0.3485429584980011, "learning_rate": 8.000249648731338e-05, "loss": 0.0364, "step": 33040 }, { "grad_norm": 0.5116022825241089, "learning_rate": 7.998926770771928e-05, "loss": 0.0495, "step": 33050 }, { "grad_norm": 0.4370284974575043, "learning_rate": 7.997603564854397e-05, "loss": 0.0391, "step": 33060 }, { "grad_norm": 0.44696781039237976, "learning_rate": 7.996280031123448e-05, "loss": 0.0413, "step": 33070 }, { "grad_norm": 0.36751723289489746, "learning_rate": 7.994956169723818e-05, "loss": 0.0514, "step": 33080 }, { "grad_norm": 0.5662136077880859, "learning_rate": 7.993631980800285e-05, "loss": 0.0364, "step": 33090 }, { "grad_norm": 0.7219932079315186, "learning_rate": 7.992307464497659e-05, "loss": 0.0409, "step": 33100 }, { "grad_norm": 0.2552243173122406, "learning_rate": 7.990982620960787e-05, "loss": 0.0299, "step": 33110 }, { "grad_norm": 0.2974933087825775, "learning_rate": 7.989657450334554e-05, "loss": 0.0361, "step": 33120 }, { "grad_norm": 0.6611621975898743, "learning_rate": 7.988331952763877e-05, "loss": 0.0448, "step": 33130 }, { "grad_norm": 0.3850564658641815, "learning_rate": 7.987006128393709e-05, "loss": 0.0479, "step": 33140 }, { "grad_norm": 0.40495565533638, "learning_rate": 7.985679977369043e-05, "loss": 0.0365, "step": 33150 }, { "grad_norm": 0.6914870738983154, "learning_rate": 7.984353499834902e-05, "loss": 0.0441, "step": 33160 }, { "grad_norm": 0.4481216073036194, "learning_rate": 7.983026695936351e-05, "loss": 0.0463, "step": 33170 }, { "grad_norm": 0.5305141806602478, "learning_rate": 7.981699565818486e-05, "loss": 0.0368, "step": 33180 }, { "grad_norm": 0.6778359413146973, "learning_rate": 7.980372109626437e-05, "loss": 0.0339, "step": 33190 }, { "grad_norm": 0.3657277524471283, "learning_rate": 7.979044327505375e-05, "loss": 0.0352, "step": 33200 }, { "grad_norm": 0.47257646918296814, "learning_rate": 7.977716219600506e-05, "loss": 0.0401, "step": 33210 }, { "grad_norm": 0.3125545084476471, "learning_rate": 7.97638778605707e-05, "loss": 0.0392, "step": 33220 }, { "grad_norm": 0.48767587542533875, "learning_rate": 7.975059027020338e-05, "loss": 0.0511, "step": 33230 }, { "grad_norm": 0.4158107340335846, "learning_rate": 7.973729942635623e-05, "loss": 0.0394, "step": 33240 }, { "grad_norm": 0.4851755201816559, "learning_rate": 7.972400533048273e-05, "loss": 0.0469, "step": 33250 }, { "grad_norm": 0.47675397992134094, "learning_rate": 7.97107079840367e-05, "loss": 0.045, "step": 33260 }, { "grad_norm": 0.5172604918479919, "learning_rate": 7.969740738847231e-05, "loss": 0.0382, "step": 33270 }, { "grad_norm": 0.42030805349349976, "learning_rate": 7.968410354524411e-05, "loss": 0.0374, "step": 33280 }, { "grad_norm": 0.3208051919937134, "learning_rate": 7.967079645580697e-05, "loss": 0.0332, "step": 33290 }, { "grad_norm": 0.5861573815345764, "learning_rate": 7.965748612161612e-05, "loss": 0.0385, "step": 33300 }, { "grad_norm": 0.5218836665153503, "learning_rate": 7.96441725441272e-05, "loss": 0.0365, "step": 33310 }, { "grad_norm": 0.37032392621040344, "learning_rate": 7.963085572479614e-05, "loss": 0.0358, "step": 33320 }, { "grad_norm": 0.36280786991119385, "learning_rate": 7.961753566507924e-05, "loss": 0.0357, "step": 33330 }, { "grad_norm": 0.3548358380794525, "learning_rate": 7.960421236643316e-05, "loss": 0.0431, "step": 33340 }, { "grad_norm": 0.37651440501213074, "learning_rate": 7.959088583031496e-05, "loss": 0.0349, "step": 33350 }, { "grad_norm": 0.37225934863090515, "learning_rate": 7.957755605818194e-05, "loss": 0.0274, "step": 33360 }, { "grad_norm": 0.3079921007156372, "learning_rate": 7.956422305149185e-05, "loss": 0.0415, "step": 33370 }, { "grad_norm": 0.46742546558380127, "learning_rate": 7.95508868117028e-05, "loss": 0.0439, "step": 33380 }, { "grad_norm": 0.32684168219566345, "learning_rate": 7.953754734027318e-05, "loss": 0.0403, "step": 33390 }, { "grad_norm": 0.3799895644187927, "learning_rate": 7.952420463866182e-05, "loss": 0.0285, "step": 33400 }, { "grad_norm": 0.38007864356040955, "learning_rate": 7.951085870832782e-05, "loss": 0.0414, "step": 33410 }, { "grad_norm": 0.433430552482605, "learning_rate": 7.949750955073067e-05, "loss": 0.0361, "step": 33420 }, { "grad_norm": 0.3238533139228821, "learning_rate": 7.948415716733022e-05, "loss": 0.0388, "step": 33430 }, { "grad_norm": 0.2724098265171051, "learning_rate": 7.947080155958669e-05, "loss": 0.0372, "step": 33440 }, { "grad_norm": 0.5144158005714417, "learning_rate": 7.94574427289606e-05, "loss": 0.0349, "step": 33450 }, { "grad_norm": 0.5048820972442627, "learning_rate": 7.944408067691284e-05, "loss": 0.0408, "step": 33460 }, { "grad_norm": 0.36530590057373047, "learning_rate": 7.943071540490473e-05, "loss": 0.0353, "step": 33470 }, { "grad_norm": 0.36721140146255493, "learning_rate": 7.94173469143978e-05, "loss": 0.038, "step": 33480 }, { "grad_norm": 0.3531203866004944, "learning_rate": 7.940397520685406e-05, "loss": 0.0345, "step": 33490 }, { "grad_norm": 0.47691136598587036, "learning_rate": 7.939060028373577e-05, "loss": 0.0368, "step": 33500 }, { "grad_norm": 0.37304913997650146, "learning_rate": 7.937722214650565e-05, "loss": 0.0446, "step": 33510 }, { "grad_norm": 0.35891973972320557, "learning_rate": 7.936384079662666e-05, "loss": 0.0308, "step": 33520 }, { "grad_norm": 0.3474597632884979, "learning_rate": 7.93504562355622e-05, "loss": 0.0403, "step": 33530 }, { "grad_norm": 0.2842729687690735, "learning_rate": 7.933706846477599e-05, "loss": 0.0351, "step": 33540 }, { "grad_norm": 0.5548635721206665, "learning_rate": 7.932367748573206e-05, "loss": 0.0404, "step": 33550 }, { "grad_norm": 0.4931834042072296, "learning_rate": 7.931028329989485e-05, "loss": 0.0354, "step": 33560 }, { "grad_norm": 0.6210326552391052, "learning_rate": 7.929688590872913e-05, "loss": 0.0446, "step": 33570 }, { "grad_norm": 0.5090672969818115, "learning_rate": 7.928348531370003e-05, "loss": 0.0328, "step": 33580 }, { "grad_norm": 0.22821953892707825, "learning_rate": 7.927008151627297e-05, "loss": 0.0265, "step": 33590 }, { "grad_norm": 0.41076138615608215, "learning_rate": 7.925667451791383e-05, "loss": 0.0352, "step": 33600 }, { "grad_norm": 0.5045672655105591, "learning_rate": 7.924326432008874e-05, "loss": 0.0335, "step": 33610 }, { "grad_norm": 0.4214628040790558, "learning_rate": 7.922985092426422e-05, "loss": 0.0333, "step": 33620 }, { "grad_norm": 0.44570720195770264, "learning_rate": 7.921643433190717e-05, "loss": 0.0361, "step": 33630 }, { "grad_norm": 0.4065210223197937, "learning_rate": 7.920301454448478e-05, "loss": 0.0415, "step": 33640 }, { "grad_norm": 0.28662264347076416, "learning_rate": 7.918959156346461e-05, "loss": 0.0308, "step": 33650 }, { "grad_norm": 0.7600328326225281, "learning_rate": 7.91761653903146e-05, "loss": 0.0377, "step": 33660 }, { "grad_norm": 0.6880724430084229, "learning_rate": 7.916273602650302e-05, "loss": 0.0429, "step": 33670 }, { "grad_norm": 0.3700917661190033, "learning_rate": 7.914930347349847e-05, "loss": 0.0346, "step": 33680 }, { "grad_norm": 0.258969247341156, "learning_rate": 7.913586773276992e-05, "loss": 0.0418, "step": 33690 }, { "grad_norm": 0.3429364860057831, "learning_rate": 7.912242880578667e-05, "loss": 0.0303, "step": 33700 }, { "grad_norm": 0.3575563132762909, "learning_rate": 7.910898669401839e-05, "loss": 0.0378, "step": 33710 }, { "grad_norm": 0.36275404691696167, "learning_rate": 7.909554139893511e-05, "loss": 0.0384, "step": 33720 }, { "grad_norm": 0.3878870904445648, "learning_rate": 7.908209292200715e-05, "loss": 0.0341, "step": 33730 }, { "grad_norm": 0.36553964018821716, "learning_rate": 7.906864126470523e-05, "loss": 0.0293, "step": 33740 }, { "grad_norm": 0.46669870615005493, "learning_rate": 7.905518642850041e-05, "loss": 0.0353, "step": 33750 }, { "grad_norm": 0.4002051055431366, "learning_rate": 7.904172841486409e-05, "loss": 0.041, "step": 33760 }, { "grad_norm": 0.3950576186180115, "learning_rate": 7.902826722526801e-05, "loss": 0.04, "step": 33770 }, { "grad_norm": 0.3177647888660431, "learning_rate": 7.901480286118427e-05, "loss": 0.0323, "step": 33780 }, { "grad_norm": 0.41624581813812256, "learning_rate": 7.900133532408531e-05, "loss": 0.0344, "step": 33790 }, { "grad_norm": 0.5012678503990173, "learning_rate": 7.898786461544395e-05, "loss": 0.034, "step": 33800 }, { "grad_norm": 0.3145221769809723, "learning_rate": 7.897439073673325e-05, "loss": 0.039, "step": 33810 }, { "grad_norm": 0.4818723499774933, "learning_rate": 7.896091368942677e-05, "loss": 0.0336, "step": 33820 }, { "grad_norm": 0.2697109878063202, "learning_rate": 7.894743347499832e-05, "loss": 0.03, "step": 33830 }, { "grad_norm": 0.3751284182071686, "learning_rate": 7.893395009492203e-05, "loss": 0.029, "step": 33840 }, { "grad_norm": 0.3318728804588318, "learning_rate": 7.892046355067248e-05, "loss": 0.0355, "step": 33850 }, { "grad_norm": 0.3625089228153229, "learning_rate": 7.890697384372451e-05, "loss": 0.0313, "step": 33860 }, { "grad_norm": 0.3613758981227875, "learning_rate": 7.889348097555336e-05, "loss": 0.0295, "step": 33870 }, { "grad_norm": 0.5350304841995239, "learning_rate": 7.887998494763455e-05, "loss": 0.0339, "step": 33880 }, { "grad_norm": 0.3585984706878662, "learning_rate": 7.886648576144404e-05, "loss": 0.0291, "step": 33890 }, { "grad_norm": 0.5682631731033325, "learning_rate": 7.885298341845802e-05, "loss": 0.0351, "step": 33900 }, { "grad_norm": 0.39679595828056335, "learning_rate": 7.883947792015311e-05, "loss": 0.0399, "step": 33910 }, { "grad_norm": 0.5256622433662415, "learning_rate": 7.882596926800628e-05, "loss": 0.0568, "step": 33920 }, { "grad_norm": 0.45660775899887085, "learning_rate": 7.881245746349477e-05, "loss": 0.0423, "step": 33930 }, { "grad_norm": 0.3930354416370392, "learning_rate": 7.879894250809623e-05, "loss": 0.0423, "step": 33940 }, { "grad_norm": 0.5551088452339172, "learning_rate": 7.878542440328865e-05, "loss": 0.0537, "step": 33950 }, { "grad_norm": 0.40456724166870117, "learning_rate": 7.877190315055031e-05, "loss": 0.0403, "step": 33960 }, { "grad_norm": 0.31452101469039917, "learning_rate": 7.875837875135991e-05, "loss": 0.032, "step": 33970 }, { "grad_norm": 0.5158787369728088, "learning_rate": 7.874485120719646e-05, "loss": 0.0317, "step": 33980 }, { "grad_norm": 0.3679302930831909, "learning_rate": 7.873132051953928e-05, "loss": 0.0352, "step": 33990 }, { "grad_norm": 0.657673180103302, "learning_rate": 7.87177866898681e-05, "loss": 0.0483, "step": 34000 }, { "grad_norm": 0.3875456750392914, "learning_rate": 7.870424971966294e-05, "loss": 0.0411, "step": 34010 }, { "grad_norm": 0.27986839413642883, "learning_rate": 7.869070961040419e-05, "loss": 0.0388, "step": 34020 }, { "grad_norm": 0.3524799942970276, "learning_rate": 7.867716636357257e-05, "loss": 0.0385, "step": 34030 }, { "grad_norm": 0.24097704887390137, "learning_rate": 7.866361998064915e-05, "loss": 0.0292, "step": 34040 }, { "grad_norm": 0.27898722887039185, "learning_rate": 7.865007046311534e-05, "loss": 0.0344, "step": 34050 }, { "grad_norm": 0.4949083924293518, "learning_rate": 7.86365178124529e-05, "loss": 0.0286, "step": 34060 }, { "grad_norm": 0.4164694547653198, "learning_rate": 7.862296203014394e-05, "loss": 0.0351, "step": 34070 }, { "grad_norm": 0.4743354916572571, "learning_rate": 7.860940311767088e-05, "loss": 0.0312, "step": 34080 }, { "grad_norm": 0.4165121912956238, "learning_rate": 7.85958410765165e-05, "loss": 0.0354, "step": 34090 }, { "grad_norm": 0.32436269521713257, "learning_rate": 7.858227590816394e-05, "loss": 0.0317, "step": 34100 }, { "grad_norm": 0.3410772681236267, "learning_rate": 7.856870761409664e-05, "loss": 0.0357, "step": 34110 }, { "grad_norm": 1.071916103363037, "learning_rate": 7.855513619579846e-05, "loss": 0.0372, "step": 34120 }, { "grad_norm": 0.36465194821357727, "learning_rate": 7.85415616547535e-05, "loss": 0.0311, "step": 34130 }, { "grad_norm": 0.43163999915122986, "learning_rate": 7.852798399244627e-05, "loss": 0.0345, "step": 34140 }, { "grad_norm": 0.48429667949676514, "learning_rate": 7.851440321036161e-05, "loss": 0.034, "step": 34150 }, { "grad_norm": 0.39781972765922546, "learning_rate": 7.850081930998468e-05, "loss": 0.0411, "step": 34160 }, { "grad_norm": 0.43037888407707214, "learning_rate": 7.8487232292801e-05, "loss": 0.0392, "step": 34170 }, { "grad_norm": 0.41497084498405457, "learning_rate": 7.847364216029642e-05, "loss": 0.0365, "step": 34180 }, { "grad_norm": 0.566306471824646, "learning_rate": 7.846004891395716e-05, "loss": 0.0374, "step": 34190 }, { "grad_norm": 0.3463467061519623, "learning_rate": 7.844645255526972e-05, "loss": 0.0347, "step": 34200 }, { "grad_norm": 0.41219690442085266, "learning_rate": 7.843285308572101e-05, "loss": 0.0327, "step": 34210 }, { "grad_norm": 0.7275335192680359, "learning_rate": 7.841925050679823e-05, "loss": 0.0334, "step": 34220 }, { "grad_norm": 0.33932042121887207, "learning_rate": 7.840564481998895e-05, "loss": 0.0334, "step": 34230 }, { "grad_norm": 0.41455239057540894, "learning_rate": 7.839203602678105e-05, "loss": 0.0346, "step": 34240 }, { "grad_norm": 0.29795587062835693, "learning_rate": 7.837842412866279e-05, "loss": 0.036, "step": 34250 }, { "grad_norm": 0.3526628315448761, "learning_rate": 7.836480912712272e-05, "loss": 0.0385, "step": 34260 }, { "grad_norm": 0.5163272619247437, "learning_rate": 7.835119102364976e-05, "loss": 0.0449, "step": 34270 }, { "grad_norm": 0.34462520480155945, "learning_rate": 7.83375698197332e-05, "loss": 0.0319, "step": 34280 }, { "grad_norm": 0.3316279351711273, "learning_rate": 7.83239455168626e-05, "loss": 0.0332, "step": 34290 }, { "grad_norm": 0.44018790125846863, "learning_rate": 7.83103181165279e-05, "loss": 0.0296, "step": 34300 }, { "grad_norm": 0.3385857343673706, "learning_rate": 7.829668762021937e-05, "loss": 0.0413, "step": 34310 }, { "grad_norm": 0.5509911179542542, "learning_rate": 7.828305402942764e-05, "loss": 0.0365, "step": 34320 }, { "grad_norm": 0.37006625533103943, "learning_rate": 7.826941734564363e-05, "loss": 0.0316, "step": 34330 }, { "grad_norm": 0.2700747549533844, "learning_rate": 7.825577757035865e-05, "loss": 0.0335, "step": 34340 }, { "grad_norm": 0.20909032225608826, "learning_rate": 7.824213470506431e-05, "loss": 0.0407, "step": 34350 }, { "grad_norm": 0.677606463432312, "learning_rate": 7.822848875125257e-05, "loss": 0.0324, "step": 34360 }, { "grad_norm": 0.39093855023384094, "learning_rate": 7.821483971041576e-05, "loss": 0.0461, "step": 34370 }, { "grad_norm": 0.35027527809143066, "learning_rate": 7.820118758404649e-05, "loss": 0.0318, "step": 34380 }, { "grad_norm": 0.34686923027038574, "learning_rate": 7.818753237363776e-05, "loss": 0.0342, "step": 34390 }, { "grad_norm": 0.5089148283004761, "learning_rate": 7.817387408068286e-05, "loss": 0.035, "step": 34400 }, { "grad_norm": 0.32206523418426514, "learning_rate": 7.816021270667544e-05, "loss": 0.0332, "step": 34410 }, { "grad_norm": 0.29187217354774475, "learning_rate": 7.81465482531095e-05, "loss": 0.0309, "step": 34420 }, { "grad_norm": 0.39489299058914185, "learning_rate": 7.813288072147938e-05, "loss": 0.0343, "step": 34430 }, { "grad_norm": 0.39422178268432617, "learning_rate": 7.811921011327972e-05, "loss": 0.0426, "step": 34440 }, { "grad_norm": 0.3564178943634033, "learning_rate": 7.810553643000549e-05, "loss": 0.0419, "step": 34450 }, { "grad_norm": 0.2923002243041992, "learning_rate": 7.809185967315206e-05, "loss": 0.0322, "step": 34460 }, { "grad_norm": 0.43144604563713074, "learning_rate": 7.80781798442151e-05, "loss": 0.0389, "step": 34470 }, { "grad_norm": 0.38916918635368347, "learning_rate": 7.806449694469061e-05, "loss": 0.0315, "step": 34480 }, { "grad_norm": 0.5118296146392822, "learning_rate": 7.805081097607492e-05, "loss": 0.0389, "step": 34490 }, { "grad_norm": 0.29950299859046936, "learning_rate": 7.803712193986474e-05, "loss": 0.0288, "step": 34500 }, { "grad_norm": 0.4771307110786438, "learning_rate": 7.802342983755702e-05, "loss": 0.0334, "step": 34510 }, { "grad_norm": 0.32827359437942505, "learning_rate": 7.800973467064918e-05, "loss": 0.0342, "step": 34520 }, { "grad_norm": 0.5150448083877563, "learning_rate": 7.799603644063884e-05, "loss": 0.0357, "step": 34530 }, { "grad_norm": 0.4400753676891327, "learning_rate": 7.798233514902405e-05, "loss": 0.037, "step": 34540 }, { "grad_norm": 0.36746329069137573, "learning_rate": 7.796863079730318e-05, "loss": 0.0377, "step": 34550 }, { "grad_norm": 0.37700292468070984, "learning_rate": 7.795492338697488e-05, "loss": 0.0298, "step": 34560 }, { "grad_norm": 0.3259398937225342, "learning_rate": 7.794121291953819e-05, "loss": 0.0393, "step": 34570 }, { "grad_norm": 0.36425885558128357, "learning_rate": 7.792749939649246e-05, "loss": 0.0372, "step": 34580 }, { "grad_norm": 0.407473623752594, "learning_rate": 7.79137828193374e-05, "loss": 0.0379, "step": 34590 }, { "grad_norm": 0.39047491550445557, "learning_rate": 7.790006318957301e-05, "loss": 0.0442, "step": 34600 }, { "grad_norm": 0.46072250604629517, "learning_rate": 7.788634050869965e-05, "loss": 0.0365, "step": 34610 }, { "grad_norm": 0.6196455955505371, "learning_rate": 7.787261477821803e-05, "loss": 0.0305, "step": 34620 }, { "grad_norm": 1.0065144300460815, "learning_rate": 7.785888599962916e-05, "loss": 0.0337, "step": 34630 }, { "grad_norm": 0.5735689997673035, "learning_rate": 7.784515417443439e-05, "loss": 0.0469, "step": 34640 }, { "grad_norm": 1.4141054153442383, "learning_rate": 7.783141930413545e-05, "loss": 0.0402, "step": 34650 }, { "grad_norm": 0.4175620675086975, "learning_rate": 7.78176813902343e-05, "loss": 0.0311, "step": 34660 }, { "grad_norm": 0.40835142135620117, "learning_rate": 7.780394043423336e-05, "loss": 0.0362, "step": 34670 }, { "grad_norm": 0.6722082495689392, "learning_rate": 7.77901964376353e-05, "loss": 0.0392, "step": 34680 }, { "grad_norm": 0.7671869397163391, "learning_rate": 7.777644940194316e-05, "loss": 0.0404, "step": 34690 }, { "grad_norm": 0.3977257311344147, "learning_rate": 7.776269932866023e-05, "loss": 0.0362, "step": 34700 }, { "grad_norm": 0.7127764225006104, "learning_rate": 7.774894621929026e-05, "loss": 0.0445, "step": 34710 }, { "grad_norm": 0.2830311059951782, "learning_rate": 7.773519007533725e-05, "loss": 0.0372, "step": 34720 }, { "grad_norm": 0.22408738732337952, "learning_rate": 7.772143089830556e-05, "loss": 0.0344, "step": 34730 }, { "grad_norm": 0.478155255317688, "learning_rate": 7.770766868969985e-05, "loss": 0.0366, "step": 34740 }, { "grad_norm": 0.36957117915153503, "learning_rate": 7.769390345102518e-05, "loss": 0.0437, "step": 34750 }, { "grad_norm": 0.38293543457984924, "learning_rate": 7.768013518378683e-05, "loss": 0.0472, "step": 34760 }, { "grad_norm": 0.4829690754413605, "learning_rate": 7.766636388949053e-05, "loss": 0.0347, "step": 34770 }, { "grad_norm": 0.34034889936447144, "learning_rate": 7.765258956964229e-05, "loss": 0.0546, "step": 34780 }, { "grad_norm": 0.37725794315338135, "learning_rate": 7.76388122257484e-05, "loss": 0.0312, "step": 34790 }, { "grad_norm": 0.2789585292339325, "learning_rate": 7.762503185931558e-05, "loss": 0.0336, "step": 34800 }, { "grad_norm": 0.21677224338054657, "learning_rate": 7.76112484718508e-05, "loss": 0.0316, "step": 34810 }, { "grad_norm": 0.34158065915107727, "learning_rate": 7.75974620648614e-05, "loss": 0.0321, "step": 34820 }, { "grad_norm": 0.47451552748680115, "learning_rate": 7.758367263985503e-05, "loss": 0.0321, "step": 34830 }, { "grad_norm": 0.3743998408317566, "learning_rate": 7.75698801983397e-05, "loss": 0.0373, "step": 34840 }, { "grad_norm": 0.3908839821815491, "learning_rate": 7.755608474182372e-05, "loss": 0.0333, "step": 34850 }, { "grad_norm": 0.3592486083507538, "learning_rate": 7.754228627181574e-05, "loss": 0.039, "step": 34860 }, { "grad_norm": 0.33109042048454285, "learning_rate": 7.752848478982476e-05, "loss": 0.0332, "step": 34870 }, { "grad_norm": 0.4604174792766571, "learning_rate": 7.751468029736006e-05, "loss": 0.0344, "step": 34880 }, { "grad_norm": 0.34336453676223755, "learning_rate": 7.750087279593129e-05, "loss": 0.041, "step": 34890 }, { "grad_norm": 0.5948544144630432, "learning_rate": 7.748706228704843e-05, "loss": 0.0286, "step": 34900 }, { "grad_norm": 0.3694213330745697, "learning_rate": 7.747324877222176e-05, "loss": 0.034, "step": 34910 }, { "grad_norm": 0.3735342025756836, "learning_rate": 7.745943225296188e-05, "loss": 0.0395, "step": 34920 }, { "grad_norm": 0.4306105971336365, "learning_rate": 7.744561273077981e-05, "loss": 0.0261, "step": 34930 }, { "grad_norm": 0.3103395700454712, "learning_rate": 7.743179020718678e-05, "loss": 0.0223, "step": 34940 }, { "grad_norm": 0.3330991268157959, "learning_rate": 7.741796468369443e-05, "loss": 0.0348, "step": 34950 }, { "grad_norm": 0.4173959791660309, "learning_rate": 7.740413616181466e-05, "loss": 0.0332, "step": 34960 }, { "grad_norm": 0.40071889758110046, "learning_rate": 7.739030464305978e-05, "loss": 0.0351, "step": 34970 }, { "grad_norm": 0.4676373600959778, "learning_rate": 7.737647012894235e-05, "loss": 0.0371, "step": 34980 }, { "grad_norm": 0.33700576424598694, "learning_rate": 7.736263262097532e-05, "loss": 0.0376, "step": 34990 }, { "grad_norm": 0.5638986229896545, "learning_rate": 7.734879212067192e-05, "loss": 0.0309, "step": 35000 }, { "grad_norm": 0.4696741998195648, "learning_rate": 7.733494862954573e-05, "loss": 0.0343, "step": 35010 }, { "grad_norm": 0.4184906780719757, "learning_rate": 7.732110214911066e-05, "loss": 0.0298, "step": 35020 }, { "grad_norm": 0.4900858402252197, "learning_rate": 7.730725268088094e-05, "loss": 0.0444, "step": 35030 }, { "grad_norm": 0.48305222392082214, "learning_rate": 7.729340022637112e-05, "loss": 0.0351, "step": 35040 }, { "grad_norm": 0.5676153302192688, "learning_rate": 7.727954478709607e-05, "loss": 0.033, "step": 35050 }, { "grad_norm": 0.3140619397163391, "learning_rate": 7.726568636457103e-05, "loss": 0.0241, "step": 35060 }, { "grad_norm": 0.5288676023483276, "learning_rate": 7.725182496031153e-05, "loss": 0.0326, "step": 35070 }, { "grad_norm": 0.63810133934021, "learning_rate": 7.723796057583341e-05, "loss": 0.0392, "step": 35080 }, { "grad_norm": 0.3623145520687103, "learning_rate": 7.722409321265287e-05, "loss": 0.0305, "step": 35090 }, { "grad_norm": 0.34404176473617554, "learning_rate": 7.721022287228645e-05, "loss": 0.0483, "step": 35100 }, { "grad_norm": 0.46414661407470703, "learning_rate": 7.719634955625094e-05, "loss": 0.0353, "step": 35110 }, { "grad_norm": 0.42896753549575806, "learning_rate": 7.718247326606355e-05, "loss": 0.0395, "step": 35120 }, { "grad_norm": 0.3134956657886505, "learning_rate": 7.716859400324175e-05, "loss": 0.0354, "step": 35130 }, { "grad_norm": 0.4573233127593994, "learning_rate": 7.715471176930335e-05, "loss": 0.053, "step": 35140 }, { "grad_norm": 0.5204421281814575, "learning_rate": 7.714082656576651e-05, "loss": 0.0313, "step": 35150 }, { "grad_norm": 0.44929298758506775, "learning_rate": 7.712693839414968e-05, "loss": 0.028, "step": 35160 }, { "grad_norm": 0.334073007106781, "learning_rate": 7.711304725597164e-05, "loss": 0.036, "step": 35170 }, { "grad_norm": 0.6381245851516724, "learning_rate": 7.709915315275151e-05, "loss": 0.0308, "step": 35180 }, { "grad_norm": 0.39811593294143677, "learning_rate": 7.708525608600876e-05, "loss": 0.0352, "step": 35190 }, { "grad_norm": 0.4969957768917084, "learning_rate": 7.707135605726311e-05, "loss": 0.0388, "step": 35200 }, { "grad_norm": 0.39059555530548096, "learning_rate": 7.705745306803466e-05, "loss": 0.0573, "step": 35210 }, { "grad_norm": 0.5018898248672485, "learning_rate": 7.704354711984383e-05, "loss": 0.0281, "step": 35220 }, { "grad_norm": 0.3304699957370758, "learning_rate": 7.702963821421133e-05, "loss": 0.0337, "step": 35230 }, { "grad_norm": 0.31921425461769104, "learning_rate": 7.701572635265826e-05, "loss": 0.0345, "step": 35240 }, { "grad_norm": 0.37528249621391296, "learning_rate": 7.700181153670596e-05, "loss": 0.0288, "step": 35250 }, { "grad_norm": 0.6820152401924133, "learning_rate": 7.698789376787616e-05, "loss": 0.031, "step": 35260 }, { "grad_norm": 0.3040010631084442, "learning_rate": 7.697397304769085e-05, "loss": 0.0325, "step": 35270 }, { "grad_norm": 0.489877313375473, "learning_rate": 7.696004937767241e-05, "loss": 0.0451, "step": 35280 }, { "grad_norm": 0.27997905015945435, "learning_rate": 7.694612275934352e-05, "loss": 0.0337, "step": 35290 }, { "grad_norm": 0.5517727732658386, "learning_rate": 7.693219319422714e-05, "loss": 0.038, "step": 35300 }, { "grad_norm": 0.4163852632045746, "learning_rate": 7.69182606838466e-05, "loss": 0.0336, "step": 35310 }, { "grad_norm": 0.39998722076416016, "learning_rate": 7.690432522972558e-05, "loss": 0.0368, "step": 35320 }, { "grad_norm": 0.5135403275489807, "learning_rate": 7.689038683338796e-05, "loss": 0.0387, "step": 35330 }, { "grad_norm": 0.4278431832790375, "learning_rate": 7.687644549635808e-05, "loss": 0.032, "step": 35340 }, { "grad_norm": 0.3361615836620331, "learning_rate": 7.686250122016053e-05, "loss": 0.0342, "step": 35350 }, { "grad_norm": 0.4426538944244385, "learning_rate": 7.684855400632022e-05, "loss": 0.0401, "step": 35360 }, { "grad_norm": 0.448365181684494, "learning_rate": 7.683460385636243e-05, "loss": 0.0323, "step": 35370 }, { "grad_norm": 0.48531243205070496, "learning_rate": 7.68206507718127e-05, "loss": 0.0387, "step": 35380 }, { "grad_norm": 0.4300621449947357, "learning_rate": 7.680669475419693e-05, "loss": 0.0381, "step": 35390 }, { "grad_norm": 0.8532135486602783, "learning_rate": 7.679273580504132e-05, "loss": 0.0433, "step": 35400 }, { "grad_norm": 0.4603659212589264, "learning_rate": 7.67787739258724e-05, "loss": 0.0329, "step": 35410 }, { "grad_norm": 0.36529719829559326, "learning_rate": 7.676480911821705e-05, "loss": 0.0407, "step": 35420 }, { "grad_norm": 0.4332524538040161, "learning_rate": 7.675084138360238e-05, "loss": 0.038, "step": 35430 }, { "grad_norm": 0.31468984484672546, "learning_rate": 7.673687072355592e-05, "loss": 0.0428, "step": 35440 }, { "grad_norm": 0.43004944920539856, "learning_rate": 7.67228971396055e-05, "loss": 0.0327, "step": 35450 }, { "grad_norm": 0.41429030895233154, "learning_rate": 7.670892063327922e-05, "loss": 0.0372, "step": 35460 }, { "grad_norm": 0.4060523808002472, "learning_rate": 7.669494120610552e-05, "loss": 0.0364, "step": 35470 }, { "grad_norm": 0.6070162057876587, "learning_rate": 7.66809588596132e-05, "loss": 0.0362, "step": 35480 }, { "grad_norm": 0.7022320032119751, "learning_rate": 7.666697359533132e-05, "loss": 0.042, "step": 35490 }, { "grad_norm": 0.4233672320842743, "learning_rate": 7.665298541478932e-05, "loss": 0.0544, "step": 35500 }, { "grad_norm": 0.4486980438232422, "learning_rate": 7.66389943195169e-05, "loss": 0.035, "step": 35510 }, { "grad_norm": 0.30739063024520874, "learning_rate": 7.66250003110441e-05, "loss": 0.0381, "step": 35520 }, { "grad_norm": 0.6979390978813171, "learning_rate": 7.66110033909013e-05, "loss": 0.033, "step": 35530 }, { "grad_norm": 0.3535533845424652, "learning_rate": 7.659700356061918e-05, "loss": 0.0289, "step": 35540 }, { "grad_norm": 0.5426275730133057, "learning_rate": 7.658300082172875e-05, "loss": 0.033, "step": 35550 }, { "grad_norm": 0.5155020356178284, "learning_rate": 7.65689951757613e-05, "loss": 0.0339, "step": 35560 }, { "grad_norm": 0.30891889333724976, "learning_rate": 7.655498662424849e-05, "loss": 0.0306, "step": 35570 }, { "grad_norm": 0.3638150095939636, "learning_rate": 7.654097516872227e-05, "loss": 0.03, "step": 35580 }, { "grad_norm": 0.46442800760269165, "learning_rate": 7.65269608107149e-05, "loss": 0.0362, "step": 35590 }, { "grad_norm": 0.40937888622283936, "learning_rate": 7.6512943551759e-05, "loss": 0.0439, "step": 35600 }, { "grad_norm": 0.7297816276550293, "learning_rate": 7.649892339338743e-05, "loss": 0.0351, "step": 35610 }, { "grad_norm": 0.48273560404777527, "learning_rate": 7.648490033713344e-05, "loss": 0.0327, "step": 35620 }, { "grad_norm": 0.2842174768447876, "learning_rate": 7.647087438453058e-05, "loss": 0.0296, "step": 35630 }, { "grad_norm": 0.45871320366859436, "learning_rate": 7.64568455371127e-05, "loss": 0.0364, "step": 35640 }, { "grad_norm": 0.2535514533519745, "learning_rate": 7.644281379641396e-05, "loss": 0.027, "step": 35650 }, { "grad_norm": 0.3720169961452484, "learning_rate": 7.642877916396887e-05, "loss": 0.0298, "step": 35660 }, { "grad_norm": 0.3313932716846466, "learning_rate": 7.641474164131221e-05, "loss": 0.0327, "step": 35670 }, { "grad_norm": 0.2803688645362854, "learning_rate": 7.640070122997913e-05, "loss": 0.0243, "step": 35680 }, { "grad_norm": 0.48066264390945435, "learning_rate": 7.638665793150505e-05, "loss": 0.0311, "step": 35690 }, { "grad_norm": 0.3936574459075928, "learning_rate": 7.637261174742574e-05, "loss": 0.0371, "step": 35700 }, { "grad_norm": 0.8964102268218994, "learning_rate": 7.635856267927725e-05, "loss": 0.0376, "step": 35710 }, { "grad_norm": 0.3982510566711426, "learning_rate": 7.634451072859597e-05, "loss": 0.0301, "step": 35720 }, { "grad_norm": 0.3219672441482544, "learning_rate": 7.633045589691863e-05, "loss": 0.0463, "step": 35730 }, { "grad_norm": 0.5227219462394714, "learning_rate": 7.63163981857822e-05, "loss": 0.0363, "step": 35740 }, { "grad_norm": 0.41313230991363525, "learning_rate": 7.630233759672403e-05, "loss": 0.0383, "step": 35750 }, { "grad_norm": 0.3376103937625885, "learning_rate": 7.628827413128177e-05, "loss": 0.0408, "step": 35760 }, { "grad_norm": 0.41491490602493286, "learning_rate": 7.627420779099338e-05, "loss": 0.0393, "step": 35770 }, { "grad_norm": 0.41457122564315796, "learning_rate": 7.626013857739711e-05, "loss": 0.0339, "step": 35780 }, { "grad_norm": 0.3666101396083832, "learning_rate": 7.624606649203158e-05, "loss": 0.0431, "step": 35790 }, { "grad_norm": 0.3440391421318054, "learning_rate": 7.623199153643569e-05, "loss": 0.0437, "step": 35800 }, { "grad_norm": 0.27952805161476135, "learning_rate": 7.621791371214863e-05, "loss": 0.0364, "step": 35810 }, { "grad_norm": 0.41399842500686646, "learning_rate": 7.620383302070995e-05, "loss": 0.0371, "step": 35820 }, { "grad_norm": 0.4804005026817322, "learning_rate": 7.61897494636595e-05, "loss": 0.0327, "step": 35830 }, { "grad_norm": 0.516589343547821, "learning_rate": 7.617566304253739e-05, "loss": 0.0457, "step": 35840 }, { "grad_norm": 0.4432026743888855, "learning_rate": 7.616157375888416e-05, "loss": 0.0338, "step": 35850 }, { "grad_norm": 0.38363102078437805, "learning_rate": 7.614748161424053e-05, "loss": 0.0294, "step": 35860 }, { "grad_norm": 0.2764432430267334, "learning_rate": 7.613338661014763e-05, "loss": 0.03, "step": 35870 }, { "grad_norm": 0.3351620137691498, "learning_rate": 7.611928874814686e-05, "loss": 0.0336, "step": 35880 }, { "grad_norm": 0.41588473320007324, "learning_rate": 7.610518802977996e-05, "loss": 0.0291, "step": 35890 }, { "grad_norm": 0.31838345527648926, "learning_rate": 7.609108445658893e-05, "loss": 0.0335, "step": 35900 }, { "grad_norm": 0.34818947315216064, "learning_rate": 7.607697803011612e-05, "loss": 0.0485, "step": 35910 }, { "grad_norm": 0.39875057339668274, "learning_rate": 7.606286875190421e-05, "loss": 0.0338, "step": 35920 }, { "grad_norm": 0.3953499495983124, "learning_rate": 7.604875662349617e-05, "loss": 0.0366, "step": 35930 }, { "grad_norm": 0.44039759039878845, "learning_rate": 7.603464164643525e-05, "loss": 0.0466, "step": 35940 }, { "grad_norm": 0.32206836342811584, "learning_rate": 7.602052382226507e-05, "loss": 0.0356, "step": 35950 }, { "grad_norm": 0.4590328633785248, "learning_rate": 7.600640315252954e-05, "loss": 0.0335, "step": 35960 }, { "grad_norm": 0.7660925984382629, "learning_rate": 7.599227963877284e-05, "loss": 0.0429, "step": 35970 }, { "grad_norm": 0.38649940490722656, "learning_rate": 7.597815328253954e-05, "loss": 0.0407, "step": 35980 }, { "grad_norm": 0.453749418258667, "learning_rate": 7.596402408537443e-05, "loss": 0.0349, "step": 35990 }, { "grad_norm": 0.29593709111213684, "learning_rate": 7.59498920488227e-05, "loss": 0.0379, "step": 36000 }, { "grad_norm": 0.3571634888648987, "learning_rate": 7.593575717442979e-05, "loss": 0.0332, "step": 36010 }, { "grad_norm": 0.38736775517463684, "learning_rate": 7.592161946374147e-05, "loss": 0.0362, "step": 36020 }, { "grad_norm": 0.9244265556335449, "learning_rate": 7.590747891830381e-05, "loss": 0.0313, "step": 36030 }, { "grad_norm": 0.43388888239860535, "learning_rate": 7.58933355396632e-05, "loss": 0.0379, "step": 36040 }, { "grad_norm": 0.3761547803878784, "learning_rate": 7.587918932936636e-05, "loss": 0.0331, "step": 36050 }, { "grad_norm": 0.3601093888282776, "learning_rate": 7.586504028896028e-05, "loss": 0.0302, "step": 36060 }, { "grad_norm": 0.35097697377204895, "learning_rate": 7.585088841999228e-05, "loss": 0.0332, "step": 36070 }, { "grad_norm": 0.27397027611732483, "learning_rate": 7.583673372400999e-05, "loss": 0.0275, "step": 36080 }, { "grad_norm": 0.3908909261226654, "learning_rate": 7.582257620256134e-05, "loss": 0.0305, "step": 36090 }, { "grad_norm": 0.3916338384151459, "learning_rate": 7.580841585719458e-05, "loss": 0.0377, "step": 36100 }, { "grad_norm": 0.3675448000431061, "learning_rate": 7.579425268945825e-05, "loss": 0.0262, "step": 36110 }, { "grad_norm": 0.3556070029735565, "learning_rate": 7.578008670090127e-05, "loss": 0.0273, "step": 36120 }, { "grad_norm": 0.367576539516449, "learning_rate": 7.576591789307272e-05, "loss": 0.0287, "step": 36130 }, { "grad_norm": 0.4168151021003723, "learning_rate": 7.575174626752216e-05, "loss": 0.0303, "step": 36140 }, { "grad_norm": 0.46420979499816895, "learning_rate": 7.573757182579934e-05, "loss": 0.0319, "step": 36150 }, { "grad_norm": 0.32051026821136475, "learning_rate": 7.572339456945435e-05, "loss": 0.0299, "step": 36160 }, { "grad_norm": 0.5261945128440857, "learning_rate": 7.570921450003762e-05, "loss": 0.0352, "step": 36170 }, { "grad_norm": 0.5027444958686829, "learning_rate": 7.569503161909983e-05, "loss": 0.0318, "step": 36180 }, { "grad_norm": 0.335843563079834, "learning_rate": 7.568084592819202e-05, "loss": 0.0346, "step": 36190 }, { "grad_norm": 0.4207949638366699, "learning_rate": 7.566665742886551e-05, "loss": 0.0411, "step": 36200 }, { "grad_norm": 0.46502983570098877, "learning_rate": 7.565246612267194e-05, "loss": 0.0382, "step": 36210 }, { "grad_norm": 0.43307065963745117, "learning_rate": 7.563827201116325e-05, "loss": 0.0376, "step": 36220 }, { "grad_norm": 0.37243470549583435, "learning_rate": 7.562407509589164e-05, "loss": 0.0294, "step": 36230 }, { "grad_norm": 0.35283878445625305, "learning_rate": 7.560987537840973e-05, "loss": 0.0272, "step": 36240 }, { "grad_norm": 0.7814509868621826, "learning_rate": 7.559567286027036e-05, "loss": 0.0323, "step": 36250 }, { "grad_norm": 0.5282678604125977, "learning_rate": 7.558146754302668e-05, "loss": 0.0352, "step": 36260 }, { "grad_norm": 0.33551859855651855, "learning_rate": 7.556725942823217e-05, "loss": 0.0318, "step": 36270 }, { "grad_norm": 0.4209841191768646, "learning_rate": 7.555304851744061e-05, "loss": 0.037, "step": 36280 }, { "grad_norm": 0.3414405286312103, "learning_rate": 7.553883481220608e-05, "loss": 0.0428, "step": 36290 }, { "grad_norm": 0.4244738221168518, "learning_rate": 7.552461831408298e-05, "loss": 0.0286, "step": 36300 }, { "grad_norm": 0.39917829632759094, "learning_rate": 7.551039902462599e-05, "loss": 0.0405, "step": 36310 }, { "grad_norm": 0.49714404344558716, "learning_rate": 7.549617694539011e-05, "loss": 0.033, "step": 36320 }, { "grad_norm": 0.5585789680480957, "learning_rate": 7.548195207793066e-05, "loss": 0.0435, "step": 36330 }, { "grad_norm": 0.33688884973526, "learning_rate": 7.546772442380323e-05, "loss": 0.0269, "step": 36340 }, { "grad_norm": 0.28297820687294006, "learning_rate": 7.545349398456376e-05, "loss": 0.036, "step": 36350 }, { "grad_norm": 0.44731640815734863, "learning_rate": 7.543926076176845e-05, "loss": 0.0401, "step": 36360 }, { "grad_norm": 0.6325067281723022, "learning_rate": 7.542502475697385e-05, "loss": 0.0381, "step": 36370 }, { "grad_norm": 0.3705543279647827, "learning_rate": 7.541078597173675e-05, "loss": 0.039, "step": 36380 }, { "grad_norm": 0.3462686836719513, "learning_rate": 7.539654440761431e-05, "loss": 0.0319, "step": 36390 }, { "grad_norm": 0.3126090466976166, "learning_rate": 7.538230006616395e-05, "loss": 0.0338, "step": 36400 }, { "grad_norm": 0.2766929566860199, "learning_rate": 7.536805294894342e-05, "loss": 0.0357, "step": 36410 }, { "grad_norm": 0.3059716820716858, "learning_rate": 7.535380305751076e-05, "loss": 0.0358, "step": 36420 }, { "grad_norm": 0.5052812695503235, "learning_rate": 7.533955039342431e-05, "loss": 0.0366, "step": 36430 }, { "grad_norm": 0.23201502859592438, "learning_rate": 7.532529495824274e-05, "loss": 0.0359, "step": 36440 }, { "grad_norm": 0.34865614771842957, "learning_rate": 7.531103675352497e-05, "loss": 0.0347, "step": 36450 }, { "grad_norm": 0.566303551197052, "learning_rate": 7.529677578083029e-05, "loss": 0.0381, "step": 36460 }, { "grad_norm": 0.32444965839385986, "learning_rate": 7.528251204171823e-05, "loss": 0.0294, "step": 36470 }, { "grad_norm": 0.3910459876060486, "learning_rate": 7.526824553774867e-05, "loss": 0.0348, "step": 36480 }, { "grad_norm": 0.37500160932540894, "learning_rate": 7.525397627048178e-05, "loss": 0.0366, "step": 36490 }, { "grad_norm": 0.2657581567764282, "learning_rate": 7.523970424147802e-05, "loss": 0.0281, "step": 36500 }, { "grad_norm": 0.3216339945793152, "learning_rate": 7.522542945229813e-05, "loss": 0.031, "step": 36510 }, { "grad_norm": 0.29176899790763855, "learning_rate": 7.521115190450321e-05, "loss": 0.0343, "step": 36520 }, { "grad_norm": 0.49798738956451416, "learning_rate": 7.519687159965462e-05, "loss": 0.027, "step": 36530 }, { "grad_norm": 0.4333919584751129, "learning_rate": 7.518258853931403e-05, "loss": 0.0323, "step": 36540 }, { "grad_norm": 0.3612612187862396, "learning_rate": 7.516830272504342e-05, "loss": 0.0252, "step": 36550 }, { "grad_norm": 0.4863929748535156, "learning_rate": 7.515401415840509e-05, "loss": 0.0334, "step": 36560 }, { "grad_norm": 0.479093998670578, "learning_rate": 7.513972284096155e-05, "loss": 0.0297, "step": 36570 }, { "grad_norm": 1.672336459159851, "learning_rate": 7.512542877427574e-05, "loss": 0.044, "step": 36580 }, { "grad_norm": 0.41664743423461914, "learning_rate": 7.51111319599108e-05, "loss": 0.0461, "step": 36590 }, { "grad_norm": 0.38667169213294983, "learning_rate": 7.509683239943024e-05, "loss": 0.0357, "step": 36600 }, { "grad_norm": 0.38081806898117065, "learning_rate": 7.50825300943978e-05, "loss": 0.0352, "step": 36610 }, { "grad_norm": 0.5032275319099426, "learning_rate": 7.50682250463776e-05, "loss": 0.0351, "step": 36620 }, { "grad_norm": 0.7206006050109863, "learning_rate": 7.5053917256934e-05, "loss": 0.0454, "step": 36630 }, { "grad_norm": 0.3553006649017334, "learning_rate": 7.503960672763166e-05, "loss": 0.0305, "step": 36640 }, { "grad_norm": 0.2786256968975067, "learning_rate": 7.502529346003559e-05, "loss": 0.0345, "step": 36650 }, { "grad_norm": 0.37933608889579773, "learning_rate": 7.501097745571106e-05, "loss": 0.0359, "step": 36660 }, { "grad_norm": 0.34397202730178833, "learning_rate": 7.499665871622361e-05, "loss": 0.0423, "step": 36670 }, { "grad_norm": 0.49508798122406006, "learning_rate": 7.498233724313917e-05, "loss": 0.0475, "step": 36680 }, { "grad_norm": 0.3513322174549103, "learning_rate": 7.496801303802387e-05, "loss": 0.041, "step": 36690 }, { "grad_norm": 0.26800209283828735, "learning_rate": 7.495368610244423e-05, "loss": 0.0317, "step": 36700 }, { "grad_norm": 0.3226998448371887, "learning_rate": 7.493935643796697e-05, "loss": 0.0305, "step": 36710 }, { "grad_norm": 0.4445085823535919, "learning_rate": 7.49250240461592e-05, "loss": 0.0376, "step": 36720 }, { "grad_norm": 0.7733994722366333, "learning_rate": 7.491068892858826e-05, "loss": 0.0333, "step": 36730 }, { "grad_norm": 1.1837133169174194, "learning_rate": 7.489635108682184e-05, "loss": 0.0327, "step": 36740 }, { "grad_norm": 0.45174771547317505, "learning_rate": 7.48820105224279e-05, "loss": 0.0411, "step": 36750 }, { "grad_norm": 0.3762781620025635, "learning_rate": 7.48676672369747e-05, "loss": 0.0349, "step": 36760 }, { "grad_norm": 0.5901522040367126, "learning_rate": 7.485332123203079e-05, "loss": 0.0391, "step": 36770 }, { "grad_norm": 0.30190175771713257, "learning_rate": 7.483897250916504e-05, "loss": 0.0362, "step": 36780 }, { "grad_norm": 0.3117730915546417, "learning_rate": 7.48246210699466e-05, "loss": 0.0399, "step": 36790 }, { "grad_norm": 0.36965370178222656, "learning_rate": 7.481026691594492e-05, "loss": 0.0398, "step": 36800 }, { "grad_norm": 0.34307318925857544, "learning_rate": 7.479591004872975e-05, "loss": 0.0342, "step": 36810 }, { "grad_norm": 0.34156107902526855, "learning_rate": 7.478155046987114e-05, "loss": 0.0389, "step": 36820 }, { "grad_norm": 0.35975128412246704, "learning_rate": 7.476718818093944e-05, "loss": 0.0349, "step": 36830 }, { "grad_norm": 0.35818594694137573, "learning_rate": 7.475282318350527e-05, "loss": 0.0329, "step": 36840 }, { "grad_norm": 0.3986935317516327, "learning_rate": 7.473845547913957e-05, "loss": 0.0335, "step": 36850 }, { "grad_norm": 0.44747450947761536, "learning_rate": 7.472408506941357e-05, "loss": 0.0414, "step": 36860 }, { "grad_norm": 0.33341044187545776, "learning_rate": 7.470971195589881e-05, "loss": 0.0343, "step": 36870 }, { "grad_norm": 0.3325587511062622, "learning_rate": 7.46953361401671e-05, "loss": 0.0418, "step": 36880 }, { "grad_norm": 0.3558543920516968, "learning_rate": 7.468095762379055e-05, "loss": 0.0358, "step": 36890 }, { "grad_norm": 0.25098559260368347, "learning_rate": 7.466657640834158e-05, "loss": 0.03, "step": 36900 }, { "grad_norm": 0.2683430314064026, "learning_rate": 7.46521924953929e-05, "loss": 0.03, "step": 36910 }, { "grad_norm": 0.44149357080459595, "learning_rate": 7.463780588651752e-05, "loss": 0.0328, "step": 36920 }, { "grad_norm": 0.6886152625083923, "learning_rate": 7.462341658328873e-05, "loss": 0.0316, "step": 36930 }, { "grad_norm": 0.5458773970603943, "learning_rate": 7.460902458728012e-05, "loss": 0.0316, "step": 36940 }, { "grad_norm": 0.31595921516418457, "learning_rate": 7.459462990006558e-05, "loss": 0.0323, "step": 36950 }, { "grad_norm": 0.45243388414382935, "learning_rate": 7.458023252321929e-05, "loss": 0.0356, "step": 36960 }, { "grad_norm": 0.3928511440753937, "learning_rate": 7.456583245831573e-05, "loss": 0.0294, "step": 36970 }, { "grad_norm": 0.3888256847858429, "learning_rate": 7.455142970692968e-05, "loss": 0.0515, "step": 36980 }, { "grad_norm": 0.3545634150505066, "learning_rate": 7.453702427063615e-05, "loss": 0.0329, "step": 36990 }, { "grad_norm": 0.4799550175666809, "learning_rate": 7.452261615101057e-05, "loss": 0.0306, "step": 37000 }, { "grad_norm": 0.31047797203063965, "learning_rate": 7.450820534962856e-05, "loss": 0.0344, "step": 37010 }, { "grad_norm": 0.462444543838501, "learning_rate": 7.449379186806604e-05, "loss": 0.0335, "step": 37020 }, { "grad_norm": 0.3777731955051422, "learning_rate": 7.44793757078993e-05, "loss": 0.0345, "step": 37030 }, { "grad_norm": 0.4522009789943695, "learning_rate": 7.44649568707048e-05, "loss": 0.034, "step": 37040 }, { "grad_norm": 0.28832969069480896, "learning_rate": 7.445053535805942e-05, "loss": 0.0319, "step": 37050 }, { "grad_norm": 0.24592164158821106, "learning_rate": 7.443611117154024e-05, "loss": 0.0297, "step": 37060 }, { "grad_norm": 0.2930467128753662, "learning_rate": 7.442168431272469e-05, "loss": 0.0373, "step": 37070 }, { "grad_norm": 0.5238826870918274, "learning_rate": 7.440725478319044e-05, "loss": 0.0281, "step": 37080 }, { "grad_norm": 0.3595291078090668, "learning_rate": 7.439282258451553e-05, "loss": 0.0255, "step": 37090 }, { "grad_norm": 0.347515344619751, "learning_rate": 7.43783877182782e-05, "loss": 0.0319, "step": 37100 }, { "grad_norm": 0.3836391866207123, "learning_rate": 7.436395018605707e-05, "loss": 0.0285, "step": 37110 }, { "grad_norm": 0.2839165925979614, "learning_rate": 7.434950998943094e-05, "loss": 0.0257, "step": 37120 }, { "grad_norm": 0.45522305369377136, "learning_rate": 7.433506712997902e-05, "loss": 0.0426, "step": 37130 }, { "grad_norm": 0.2505109906196594, "learning_rate": 7.432062160928077e-05, "loss": 0.0282, "step": 37140 }, { "grad_norm": 0.2549554705619812, "learning_rate": 7.430617342891588e-05, "loss": 0.0237, "step": 37150 }, { "grad_norm": 0.3305193781852722, "learning_rate": 7.429172259046442e-05, "loss": 0.0387, "step": 37160 }, { "grad_norm": 0.3834589719772339, "learning_rate": 7.42772690955067e-05, "loss": 0.0235, "step": 37170 }, { "grad_norm": 0.6257899403572083, "learning_rate": 7.426281294562332e-05, "loss": 0.0454, "step": 37180 }, { "grad_norm": 0.3651088774204254, "learning_rate": 7.424835414239519e-05, "loss": 0.0322, "step": 37190 }, { "grad_norm": 0.43358513712882996, "learning_rate": 7.423389268740352e-05, "loss": 0.0312, "step": 37200 }, { "grad_norm": 0.5763490796089172, "learning_rate": 7.42194285822298e-05, "loss": 0.0358, "step": 37210 }, { "grad_norm": 1.4209873676300049, "learning_rate": 7.420496182845576e-05, "loss": 0.0411, "step": 37220 }, { "grad_norm": 0.49850621819496155, "learning_rate": 7.419049242766348e-05, "loss": 0.0435, "step": 37230 }, { "grad_norm": 0.4305216372013092, "learning_rate": 7.417602038143533e-05, "loss": 0.0374, "step": 37240 }, { "grad_norm": 0.3976294696331024, "learning_rate": 7.416154569135393e-05, "loss": 0.0366, "step": 37250 }, { "grad_norm": 0.3637573719024658, "learning_rate": 7.414706835900222e-05, "loss": 0.0362, "step": 37260 }, { "grad_norm": 0.4079177677631378, "learning_rate": 7.413258838596345e-05, "loss": 0.0355, "step": 37270 }, { "grad_norm": 0.36131301522254944, "learning_rate": 7.411810577382106e-05, "loss": 0.0332, "step": 37280 }, { "grad_norm": 0.3763576149940491, "learning_rate": 7.410362052415889e-05, "loss": 0.0389, "step": 37290 }, { "grad_norm": 0.33563804626464844, "learning_rate": 7.408913263856102e-05, "loss": 0.0348, "step": 37300 }, { "grad_norm": 0.4052284359931946, "learning_rate": 7.407464211861183e-05, "loss": 0.0399, "step": 37310 }, { "grad_norm": 0.4246795177459717, "learning_rate": 7.406014896589597e-05, "loss": 0.0437, "step": 37320 }, { "grad_norm": 0.46494060754776, "learning_rate": 7.404565318199838e-05, "loss": 0.0335, "step": 37330 }, { "grad_norm": 0.3684324324131012, "learning_rate": 7.403115476850432e-05, "loss": 0.0389, "step": 37340 }, { "grad_norm": 0.3264562785625458, "learning_rate": 7.401665372699932e-05, "loss": 0.0334, "step": 37350 }, { "grad_norm": 0.3039493262767792, "learning_rate": 7.400215005906916e-05, "loss": 0.0323, "step": 37360 }, { "grad_norm": 0.27945634722709656, "learning_rate": 7.398764376629996e-05, "loss": 0.0316, "step": 37370 }, { "grad_norm": 0.2967582046985626, "learning_rate": 7.39731348502781e-05, "loss": 0.032, "step": 37380 }, { "grad_norm": 0.2812114655971527, "learning_rate": 7.395862331259028e-05, "loss": 0.0291, "step": 37390 }, { "grad_norm": 0.5482059717178345, "learning_rate": 7.394410915482341e-05, "loss": 0.0354, "step": 37400 }, { "grad_norm": 0.32382330298423767, "learning_rate": 7.392959237856476e-05, "loss": 0.0341, "step": 37410 }, { "grad_norm": 0.4144544303417206, "learning_rate": 7.391507298540187e-05, "loss": 0.0333, "step": 37420 }, { "grad_norm": 0.4344902038574219, "learning_rate": 7.390055097692257e-05, "loss": 0.0348, "step": 37430 }, { "grad_norm": 0.46590712666511536, "learning_rate": 7.388602635471494e-05, "loss": 0.0514, "step": 37440 }, { "grad_norm": 0.42548131942749023, "learning_rate": 7.387149912036738e-05, "loss": 0.0562, "step": 37450 }, { "grad_norm": 1.095685601234436, "learning_rate": 7.385696927546854e-05, "loss": 0.0387, "step": 37460 }, { "grad_norm": 0.5022603869438171, "learning_rate": 7.384243682160743e-05, "loss": 0.0466, "step": 37470 }, { "grad_norm": 0.8829654455184937, "learning_rate": 7.382790176037329e-05, "loss": 0.0421, "step": 37480 }, { "grad_norm": 0.483011931180954, "learning_rate": 7.381336409335562e-05, "loss": 0.0364, "step": 37490 }, { "grad_norm": 0.35461392998695374, "learning_rate": 7.379882382214426e-05, "loss": 0.0438, "step": 37500 }, { "grad_norm": 0.3752925395965576, "learning_rate": 7.378428094832931e-05, "loss": 0.037, "step": 37510 }, { "grad_norm": 0.4204323887825012, "learning_rate": 7.376973547350114e-05, "loss": 0.0423, "step": 37520 }, { "grad_norm": 0.45617496967315674, "learning_rate": 7.375518739925046e-05, "loss": 0.0384, "step": 37530 }, { "grad_norm": 0.6230558156967163, "learning_rate": 7.374063672716818e-05, "loss": 0.0369, "step": 37540 }, { "grad_norm": 0.4463900625705719, "learning_rate": 7.372608345884558e-05, "loss": 0.0386, "step": 37550 }, { "grad_norm": 0.5737513303756714, "learning_rate": 7.371152759587415e-05, "loss": 0.0497, "step": 37560 }, { "grad_norm": 0.3625333905220032, "learning_rate": 7.369696913984572e-05, "loss": 0.0361, "step": 37570 }, { "grad_norm": 0.3612857758998871, "learning_rate": 7.368240809235237e-05, "loss": 0.0277, "step": 37580 }, { "grad_norm": 0.4314194619655609, "learning_rate": 7.366784445498649e-05, "loss": 0.0327, "step": 37590 }, { "grad_norm": 0.39169883728027344, "learning_rate": 7.36532782293407e-05, "loss": 0.0302, "step": 37600 }, { "grad_norm": 0.40239739418029785, "learning_rate": 7.363870941700797e-05, "loss": 0.0318, "step": 37610 }, { "grad_norm": 0.3233392834663391, "learning_rate": 7.362413801958152e-05, "loss": 0.0279, "step": 37620 }, { "grad_norm": 0.4268714189529419, "learning_rate": 7.360956403865486e-05, "loss": 0.0367, "step": 37630 }, { "grad_norm": 0.3625760078430176, "learning_rate": 7.359498747582177e-05, "loss": 0.0285, "step": 37640 }, { "grad_norm": 0.27269503474235535, "learning_rate": 7.358040833267634e-05, "loss": 0.0316, "step": 37650 }, { "grad_norm": 0.341107577085495, "learning_rate": 7.356582661081288e-05, "loss": 0.0314, "step": 37660 }, { "grad_norm": 0.2910209000110626, "learning_rate": 7.355124231182607e-05, "loss": 0.0301, "step": 37670 }, { "grad_norm": 0.42070603370666504, "learning_rate": 7.353665543731083e-05, "loss": 0.0433, "step": 37680 }, { "grad_norm": 0.39727702736854553, "learning_rate": 7.352206598886229e-05, "loss": 0.0302, "step": 37690 }, { "grad_norm": 0.39162343740463257, "learning_rate": 7.350747396807601e-05, "loss": 0.0336, "step": 37700 }, { "grad_norm": 0.4423729479312897, "learning_rate": 7.349287937654772e-05, "loss": 0.0344, "step": 37710 }, { "grad_norm": 0.3715255558490753, "learning_rate": 7.347828221587345e-05, "loss": 0.0375, "step": 37720 }, { "grad_norm": 1.6314496994018555, "learning_rate": 7.346368248764955e-05, "loss": 0.033, "step": 37730 }, { "grad_norm": 0.3126991391181946, "learning_rate": 7.34490801934726e-05, "loss": 0.0307, "step": 37740 }, { "grad_norm": 0.45471444725990295, "learning_rate": 7.343447533493947e-05, "loss": 0.0395, "step": 37750 }, { "grad_norm": 0.3697821795940399, "learning_rate": 7.341986791364739e-05, "loss": 0.0537, "step": 37760 }, { "grad_norm": 0.2917368710041046, "learning_rate": 7.340525793119374e-05, "loss": 0.0346, "step": 37770 }, { "grad_norm": 0.4329269528388977, "learning_rate": 7.33906453891763e-05, "loss": 0.0316, "step": 37780 }, { "grad_norm": 0.3859824240207672, "learning_rate": 7.337603028919303e-05, "loss": 0.028, "step": 37790 }, { "grad_norm": 0.747897207736969, "learning_rate": 7.336141263284225e-05, "loss": 0.0308, "step": 37800 }, { "grad_norm": 0.7128632068634033, "learning_rate": 7.334679242172249e-05, "loss": 0.045, "step": 37810 }, { "grad_norm": 0.36332184076309204, "learning_rate": 7.33321696574326e-05, "loss": 0.0356, "step": 37820 }, { "grad_norm": 0.47825145721435547, "learning_rate": 7.331754434157174e-05, "loss": 0.0361, "step": 37830 }, { "grad_norm": 0.42884257435798645, "learning_rate": 7.330291647573928e-05, "loss": 0.0352, "step": 37840 }, { "grad_norm": 0.3948606848716736, "learning_rate": 7.32882860615349e-05, "loss": 0.0351, "step": 37850 }, { "grad_norm": 0.3890925347805023, "learning_rate": 7.327365310055858e-05, "loss": 0.0381, "step": 37860 }, { "grad_norm": 0.31239160895347595, "learning_rate": 7.325901759441055e-05, "loss": 0.0324, "step": 37870 }, { "grad_norm": 0.596714437007904, "learning_rate": 7.324437954469132e-05, "loss": 0.0474, "step": 37880 }, { "grad_norm": 0.3609071969985962, "learning_rate": 7.32297389530017e-05, "loss": 0.0342, "step": 37890 }, { "grad_norm": 0.22665344178676605, "learning_rate": 7.321509582094274e-05, "loss": 0.0287, "step": 37900 }, { "grad_norm": 0.32634198665618896, "learning_rate": 7.320045015011582e-05, "loss": 0.0337, "step": 37910 }, { "grad_norm": 0.4626005291938782, "learning_rate": 7.318580194212253e-05, "loss": 0.0319, "step": 37920 }, { "grad_norm": 0.3399508595466614, "learning_rate": 7.317115119856483e-05, "loss": 0.0335, "step": 37930 }, { "grad_norm": 0.7013007402420044, "learning_rate": 7.315649792104485e-05, "loss": 0.0352, "step": 37940 }, { "grad_norm": 0.4278114438056946, "learning_rate": 7.314184211116507e-05, "loss": 0.0497, "step": 37950 }, { "grad_norm": 0.3955235183238983, "learning_rate": 7.312718377052823e-05, "loss": 0.0379, "step": 37960 }, { "grad_norm": 0.299600213766098, "learning_rate": 7.311252290073737e-05, "loss": 0.0302, "step": 37970 }, { "grad_norm": 0.3866153359413147, "learning_rate": 7.309785950339572e-05, "loss": 0.0299, "step": 37980 }, { "grad_norm": 0.553183913230896, "learning_rate": 7.308319358010692e-05, "loss": 0.0289, "step": 37990 }, { "grad_norm": 0.44689908623695374, "learning_rate": 7.306852513247474e-05, "loss": 0.045, "step": 38000 }, { "grad_norm": 0.28694045543670654, "learning_rate": 7.305385416210336e-05, "loss": 0.0436, "step": 38010 }, { "grad_norm": 0.3254426121711731, "learning_rate": 7.303918067059711e-05, "loss": 0.0296, "step": 38020 }, { "grad_norm": 0.3380172848701477, "learning_rate": 7.302450465956075e-05, "loss": 0.0374, "step": 38030 }, { "grad_norm": 0.300925612449646, "learning_rate": 7.300982613059914e-05, "loss": 0.0249, "step": 38040 }, { "grad_norm": 0.4696289598941803, "learning_rate": 7.299514508531757e-05, "loss": 0.0404, "step": 38050 }, { "grad_norm": 0.6852820515632629, "learning_rate": 7.298046152532148e-05, "loss": 0.0353, "step": 38060 }, { "grad_norm": 0.242818683385849, "learning_rate": 7.296577545221668e-05, "loss": 0.0282, "step": 38070 }, { "grad_norm": 0.3466356098651886, "learning_rate": 7.29510868676092e-05, "loss": 0.0371, "step": 38080 }, { "grad_norm": 0.42707571387290955, "learning_rate": 7.293639577310538e-05, "loss": 0.0348, "step": 38090 }, { "grad_norm": 0.43076205253601074, "learning_rate": 7.292170217031179e-05, "loss": 0.0299, "step": 38100 }, { "grad_norm": 0.3118334114551544, "learning_rate": 7.290700606083532e-05, "loss": 0.0324, "step": 38110 }, { "grad_norm": 0.4278811812400818, "learning_rate": 7.28923074462831e-05, "loss": 0.0406, "step": 38120 }, { "grad_norm": 0.24916879832744598, "learning_rate": 7.287760632826255e-05, "loss": 0.0409, "step": 38130 }, { "grad_norm": 0.5258107781410217, "learning_rate": 7.286290270838139e-05, "loss": 0.0432, "step": 38140 }, { "grad_norm": 0.35634100437164307, "learning_rate": 7.284819658824756e-05, "loss": 0.0245, "step": 38150 }, { "grad_norm": 0.36132997274398804, "learning_rate": 7.283348796946929e-05, "loss": 0.0332, "step": 38160 }, { "grad_norm": 0.39576256275177, "learning_rate": 7.28187768536551e-05, "loss": 0.03, "step": 38170 }, { "grad_norm": 0.511805534362793, "learning_rate": 7.280406324241379e-05, "loss": 0.0384, "step": 38180 }, { "grad_norm": 0.39751988649368286, "learning_rate": 7.278934713735441e-05, "loss": 0.0416, "step": 38190 }, { "grad_norm": 0.4079326391220093, "learning_rate": 7.277462854008629e-05, "loss": 0.0381, "step": 38200 }, { "grad_norm": 0.46684813499450684, "learning_rate": 7.275990745221904e-05, "loss": 0.0343, "step": 38210 }, { "grad_norm": 0.6035264730453491, "learning_rate": 7.274518387536252e-05, "loss": 0.0408, "step": 38220 }, { "grad_norm": 0.4078914523124695, "learning_rate": 7.273045781112689e-05, "loss": 0.043, "step": 38230 }, { "grad_norm": 0.35404208302497864, "learning_rate": 7.271572926112257e-05, "loss": 0.0352, "step": 38240 }, { "grad_norm": 0.4124438166618347, "learning_rate": 7.270099822696024e-05, "loss": 0.035, "step": 38250 }, { "grad_norm": 0.2997977137565613, "learning_rate": 7.26862647102509e-05, "loss": 0.0293, "step": 38260 }, { "grad_norm": 0.34222137928009033, "learning_rate": 7.267152871260573e-05, "loss": 0.037, "step": 38270 }, { "grad_norm": 0.4983866214752197, "learning_rate": 7.265679023563629e-05, "loss": 0.0377, "step": 38280 }, { "grad_norm": 0.327869176864624, "learning_rate": 7.264204928095433e-05, "loss": 0.026, "step": 38290 }, { "grad_norm": 0.3370918035507202, "learning_rate": 7.262730585017188e-05, "loss": 0.0369, "step": 38300 }, { "grad_norm": 0.3112172782421112, "learning_rate": 7.261255994490131e-05, "loss": 0.0333, "step": 38310 }, { "grad_norm": 0.32996323704719543, "learning_rate": 7.259781156675517e-05, "loss": 0.0393, "step": 38320 }, { "grad_norm": 0.33786171674728394, "learning_rate": 7.258306071734633e-05, "loss": 0.0314, "step": 38330 }, { "grad_norm": 0.40045779943466187, "learning_rate": 7.256830739828792e-05, "loss": 0.0304, "step": 38340 }, { "grad_norm": 0.42191460728645325, "learning_rate": 7.255355161119336e-05, "loss": 0.0304, "step": 38350 }, { "grad_norm": 0.38679707050323486, "learning_rate": 7.253879335767628e-05, "loss": 0.0256, "step": 38360 }, { "grad_norm": 0.2745577394962311, "learning_rate": 7.252403263935067e-05, "loss": 0.0243, "step": 38370 }, { "grad_norm": 0.4147779941558838, "learning_rate": 7.25092694578307e-05, "loss": 0.0367, "step": 38380 }, { "grad_norm": 0.5121501088142395, "learning_rate": 7.249450381473087e-05, "loss": 0.0297, "step": 38390 }, { "grad_norm": 0.28851377964019775, "learning_rate": 7.247973571166593e-05, "loss": 0.0378, "step": 38400 }, { "grad_norm": 0.4228912889957428, "learning_rate": 7.246496515025089e-05, "loss": 0.0349, "step": 38410 }, { "grad_norm": 0.6711790561676025, "learning_rate": 7.245019213210101e-05, "loss": 0.0363, "step": 38420 }, { "grad_norm": 0.32227352261543274, "learning_rate": 7.24354166588319e-05, "loss": 0.0399, "step": 38430 }, { "grad_norm": 0.4158068597316742, "learning_rate": 7.242063873205935e-05, "loss": 0.0413, "step": 38440 }, { "grad_norm": 0.419958233833313, "learning_rate": 7.240585835339946e-05, "loss": 0.0464, "step": 38450 }, { "grad_norm": 0.3168424367904663, "learning_rate": 7.239107552446857e-05, "loss": 0.0379, "step": 38460 }, { "grad_norm": 0.8011845350265503, "learning_rate": 7.237629024688334e-05, "loss": 0.0382, "step": 38470 }, { "grad_norm": 0.35168907046318054, "learning_rate": 7.236150252226064e-05, "loss": 0.0413, "step": 38480 }, { "grad_norm": 0.7655084729194641, "learning_rate": 7.234671235221765e-05, "loss": 0.0443, "step": 38490 }, { "grad_norm": 0.317910373210907, "learning_rate": 7.233191973837179e-05, "loss": 0.032, "step": 38500 }, { "grad_norm": 0.2861861288547516, "learning_rate": 7.231712468234078e-05, "loss": 0.0308, "step": 38510 }, { "grad_norm": 0.777887761592865, "learning_rate": 7.230232718574254e-05, "loss": 0.0388, "step": 38520 }, { "grad_norm": 0.39668846130371094, "learning_rate": 7.228752725019535e-05, "loss": 0.0401, "step": 38530 }, { "grad_norm": 0.4009661078453064, "learning_rate": 7.227272487731769e-05, "loss": 0.0349, "step": 38540 }, { "grad_norm": 0.43675366044044495, "learning_rate": 7.225792006872831e-05, "loss": 0.0275, "step": 38550 }, { "grad_norm": 0.3416602909564972, "learning_rate": 7.224311282604628e-05, "loss": 0.0266, "step": 38560 }, { "grad_norm": 0.46488118171691895, "learning_rate": 7.222830315089085e-05, "loss": 0.0409, "step": 38570 }, { "grad_norm": 0.3493855893611908, "learning_rate": 7.22134910448816e-05, "loss": 0.0291, "step": 38580 }, { "grad_norm": 0.4559085965156555, "learning_rate": 7.219867650963839e-05, "loss": 0.0345, "step": 38590 }, { "grad_norm": 0.3719398081302643, "learning_rate": 7.21838595467813e-05, "loss": 0.0323, "step": 38600 }, { "grad_norm": 0.2756878733634949, "learning_rate": 7.216904015793067e-05, "loss": 0.0286, "step": 38610 }, { "grad_norm": 0.39015454053878784, "learning_rate": 7.215421834470713e-05, "loss": 0.029, "step": 38620 }, { "grad_norm": 0.3721538782119751, "learning_rate": 7.213939410873158e-05, "loss": 0.0381, "step": 38630 }, { "grad_norm": 0.4879457950592041, "learning_rate": 7.212456745162519e-05, "loss": 0.0346, "step": 38640 }, { "grad_norm": 0.3415151536464691, "learning_rate": 7.210973837500937e-05, "loss": 0.033, "step": 38650 }, { "grad_norm": 0.2930328845977783, "learning_rate": 7.209490688050578e-05, "loss": 0.0289, "step": 38660 }, { "grad_norm": 0.3739001452922821, "learning_rate": 7.208007296973641e-05, "loss": 0.0294, "step": 38670 }, { "grad_norm": 0.3465576767921448, "learning_rate": 7.206523664432345e-05, "loss": 0.0371, "step": 38680 }, { "grad_norm": 0.4848270118236542, "learning_rate": 7.205039790588939e-05, "loss": 0.0286, "step": 38690 }, { "grad_norm": 0.4596346616744995, "learning_rate": 7.203555675605697e-05, "loss": 0.0403, "step": 38700 }, { "grad_norm": 0.5387033820152283, "learning_rate": 7.202071319644917e-05, "loss": 0.0414, "step": 38710 }, { "grad_norm": 0.3850954473018646, "learning_rate": 7.200586722868932e-05, "loss": 0.0355, "step": 38720 }, { "grad_norm": 0.4671645164489746, "learning_rate": 7.199101885440088e-05, "loss": 0.032, "step": 38730 }, { "grad_norm": 0.40952590107917786, "learning_rate": 7.197616807520771e-05, "loss": 0.0314, "step": 38740 }, { "grad_norm": 0.34738588333129883, "learning_rate": 7.196131489273381e-05, "loss": 0.0269, "step": 38750 }, { "grad_norm": 0.41156020760536194, "learning_rate": 7.194645930860355e-05, "loss": 0.0344, "step": 38760 }, { "grad_norm": 0.314512699842453, "learning_rate": 7.19316013244415e-05, "loss": 0.0328, "step": 38770 }, { "grad_norm": 0.3845374882221222, "learning_rate": 7.191674094187248e-05, "loss": 0.0299, "step": 38780 }, { "grad_norm": 0.45991185307502747, "learning_rate": 7.190187816252165e-05, "loss": 0.0286, "step": 38790 }, { "grad_norm": 0.27565380930900574, "learning_rate": 7.188701298801435e-05, "loss": 0.0254, "step": 38800 }, { "grad_norm": 0.2918136715888977, "learning_rate": 7.18721454199762e-05, "loss": 0.0378, "step": 38810 }, { "grad_norm": 0.39087486267089844, "learning_rate": 7.185727546003312e-05, "loss": 0.0307, "step": 38820 }, { "grad_norm": 0.41502419114112854, "learning_rate": 7.184240310981126e-05, "loss": 0.0285, "step": 38830 }, { "grad_norm": 0.37725597620010376, "learning_rate": 7.182752837093704e-05, "loss": 0.0343, "step": 38840 }, { "grad_norm": 0.20871955156326294, "learning_rate": 7.181265124503711e-05, "loss": 0.0426, "step": 38850 }, { "grad_norm": 0.24032527208328247, "learning_rate": 7.179777173373847e-05, "loss": 0.0254, "step": 38860 }, { "grad_norm": 0.321571946144104, "learning_rate": 7.178288983866826e-05, "loss": 0.026, "step": 38870 }, { "grad_norm": 0.3704671561717987, "learning_rate": 7.176800556145397e-05, "loss": 0.0324, "step": 38880 }, { "grad_norm": 0.3313432037830353, "learning_rate": 7.175311890372334e-05, "loss": 0.0314, "step": 38890 }, { "grad_norm": 0.4674322009086609, "learning_rate": 7.17382298671043e-05, "loss": 0.0309, "step": 38900 }, { "grad_norm": 0.5864459276199341, "learning_rate": 7.172333845322515e-05, "loss": 0.0338, "step": 38910 }, { "grad_norm": 0.36005815863609314, "learning_rate": 7.170844466371436e-05, "loss": 0.0344, "step": 38920 }, { "grad_norm": 0.43475741147994995, "learning_rate": 7.16935485002007e-05, "loss": 0.0275, "step": 38930 }, { "grad_norm": 0.2673378586769104, "learning_rate": 7.167864996431319e-05, "loss": 0.0381, "step": 38940 }, { "grad_norm": 0.5821629762649536, "learning_rate": 7.166374905768111e-05, "loss": 0.03, "step": 38950 }, { "grad_norm": 0.6757636070251465, "learning_rate": 7.164884578193401e-05, "loss": 0.0359, "step": 38960 }, { "grad_norm": 0.35800033807754517, "learning_rate": 7.163394013870168e-05, "loss": 0.0305, "step": 38970 }, { "grad_norm": 0.5632593035697937, "learning_rate": 7.161903212961419e-05, "loss": 0.046, "step": 38980 }, { "grad_norm": 0.3062051236629486, "learning_rate": 7.160412175630183e-05, "loss": 0.0315, "step": 38990 }, { "grad_norm": 0.34385937452316284, "learning_rate": 7.158920902039521e-05, "loss": 0.0251, "step": 39000 }, { "grad_norm": 0.303658127784729, "learning_rate": 7.157429392352514e-05, "loss": 0.0271, "step": 39010 }, { "grad_norm": 0.49563711881637573, "learning_rate": 7.155937646732275e-05, "loss": 0.0324, "step": 39020 }, { "grad_norm": 0.3752002716064453, "learning_rate": 7.154445665341933e-05, "loss": 0.0317, "step": 39030 }, { "grad_norm": 0.4912131428718567, "learning_rate": 7.152953448344654e-05, "loss": 0.0367, "step": 39040 }, { "grad_norm": 0.4558269679546356, "learning_rate": 7.151460995903624e-05, "loss": 0.0311, "step": 39050 }, { "grad_norm": 0.5608192086219788, "learning_rate": 7.149968308182052e-05, "loss": 0.0292, "step": 39060 }, { "grad_norm": 0.2709311842918396, "learning_rate": 7.14847538534318e-05, "loss": 0.0348, "step": 39070 }, { "grad_norm": 0.35317370295524597, "learning_rate": 7.14698222755027e-05, "loss": 0.0296, "step": 39080 }, { "grad_norm": 0.6553910374641418, "learning_rate": 7.14548883496661e-05, "loss": 0.0387, "step": 39090 }, { "grad_norm": 0.343115895986557, "learning_rate": 7.143995207755517e-05, "loss": 0.0343, "step": 39100 }, { "grad_norm": 0.35141751170158386, "learning_rate": 7.142501346080333e-05, "loss": 0.0374, "step": 39110 }, { "grad_norm": 0.3578009605407715, "learning_rate": 7.141007250104421e-05, "loss": 0.0264, "step": 39120 }, { "grad_norm": 0.39499878883361816, "learning_rate": 7.139512919991176e-05, "loss": 0.0365, "step": 39130 }, { "grad_norm": 0.2910304367542267, "learning_rate": 7.138018355904014e-05, "loss": 0.0285, "step": 39140 }, { "grad_norm": 0.3770562410354614, "learning_rate": 7.13652355800638e-05, "loss": 0.0397, "step": 39150 }, { "grad_norm": 0.24286983907222748, "learning_rate": 7.135028526461743e-05, "loss": 0.0309, "step": 39160 }, { "grad_norm": 0.35746926069259644, "learning_rate": 7.133533261433594e-05, "loss": 0.0319, "step": 39170 }, { "grad_norm": 0.5638810396194458, "learning_rate": 7.132037763085457e-05, "loss": 0.0339, "step": 39180 }, { "grad_norm": 0.38869571685791016, "learning_rate": 7.130542031580875e-05, "loss": 0.0324, "step": 39190 }, { "grad_norm": 0.4530583620071411, "learning_rate": 7.12904606708342e-05, "loss": 0.0324, "step": 39200 }, { "grad_norm": 0.3818117082118988, "learning_rate": 7.127549869756687e-05, "loss": 0.0403, "step": 39210 }, { "grad_norm": 0.43443092703819275, "learning_rate": 7.126053439764299e-05, "loss": 0.0286, "step": 39220 }, { "grad_norm": 0.32962602376937866, "learning_rate": 7.124556777269904e-05, "loss": 0.0331, "step": 39230 }, { "grad_norm": 0.373309463262558, "learning_rate": 7.123059882437174e-05, "loss": 0.0387, "step": 39240 }, { "grad_norm": 0.37855595350265503, "learning_rate": 7.121562755429807e-05, "loss": 0.025, "step": 39250 }, { "grad_norm": 0.32278236746788025, "learning_rate": 7.120065396411528e-05, "loss": 0.0306, "step": 39260 }, { "grad_norm": 0.5021184086799622, "learning_rate": 7.118567805546084e-05, "loss": 0.026, "step": 39270 }, { "grad_norm": 0.46171027421951294, "learning_rate": 7.117069982997248e-05, "loss": 0.0347, "step": 39280 }, { "grad_norm": 0.36879947781562805, "learning_rate": 7.115571928928825e-05, "loss": 0.0472, "step": 39290 }, { "grad_norm": 0.3904656171798706, "learning_rate": 7.114073643504635e-05, "loss": 0.0372, "step": 39300 }, { "grad_norm": 0.34995347261428833, "learning_rate": 7.11257512688853e-05, "loss": 0.0289, "step": 39310 }, { "grad_norm": 0.8646196126937866, "learning_rate": 7.111076379244384e-05, "loss": 0.0316, "step": 39320 }, { "grad_norm": 0.32883453369140625, "learning_rate": 7.109577400736101e-05, "loss": 0.0304, "step": 39330 }, { "grad_norm": 0.3211022615432739, "learning_rate": 7.108078191527605e-05, "loss": 0.0314, "step": 39340 }, { "grad_norm": 0.44220611453056335, "learning_rate": 7.106578751782847e-05, "loss": 0.0414, "step": 39350 }, { "grad_norm": 0.5912030339241028, "learning_rate": 7.105079081665803e-05, "loss": 0.036, "step": 39360 }, { "grad_norm": 0.3633304834365845, "learning_rate": 7.103579181340476e-05, "loss": 0.04, "step": 39370 }, { "grad_norm": 0.39487671852111816, "learning_rate": 7.102079050970893e-05, "loss": 0.0388, "step": 39380 }, { "grad_norm": 0.2917153835296631, "learning_rate": 7.100578690721105e-05, "loss": 0.0395, "step": 39390 }, { "grad_norm": 0.7014195322990417, "learning_rate": 7.09907810075519e-05, "loss": 0.043, "step": 39400 }, { "grad_norm": 0.410545289516449, "learning_rate": 7.097577281237249e-05, "loss": 0.0457, "step": 39410 }, { "grad_norm": 0.34835460782051086, "learning_rate": 7.09607623233141e-05, "loss": 0.0345, "step": 39420 }, { "grad_norm": 0.4152776896953583, "learning_rate": 7.094574954201827e-05, "loss": 0.0321, "step": 39430 }, { "grad_norm": 0.367695689201355, "learning_rate": 7.093073447012675e-05, "loss": 0.0345, "step": 39440 }, { "grad_norm": 0.3848372995853424, "learning_rate": 7.09157171092816e-05, "loss": 0.028, "step": 39450 }, { "grad_norm": 0.2697213888168335, "learning_rate": 7.090069746112504e-05, "loss": 0.0271, "step": 39460 }, { "grad_norm": 0.3939807415008545, "learning_rate": 7.088567552729965e-05, "loss": 0.0297, "step": 39470 }, { "grad_norm": 0.35676178336143494, "learning_rate": 7.087065130944818e-05, "loss": 0.0475, "step": 39480 }, { "grad_norm": 0.4803412854671478, "learning_rate": 7.085562480921366e-05, "loss": 0.0429, "step": 39490 }, { "grad_norm": 0.29513540863990784, "learning_rate": 7.084059602823937e-05, "loss": 0.0346, "step": 39500 }, { "grad_norm": 0.44223758578300476, "learning_rate": 7.082556496816882e-05, "loss": 0.0351, "step": 39510 }, { "grad_norm": 0.4721978008747101, "learning_rate": 7.081053163064582e-05, "loss": 0.0367, "step": 39520 }, { "grad_norm": 0.4426345229148865, "learning_rate": 7.079549601731434e-05, "loss": 0.0376, "step": 39530 }, { "grad_norm": 0.38959765434265137, "learning_rate": 7.07804581298187e-05, "loss": 0.0364, "step": 39540 }, { "grad_norm": 0.3569568395614624, "learning_rate": 7.07654179698034e-05, "loss": 0.0531, "step": 39550 }, { "grad_norm": 0.4960210621356964, "learning_rate": 7.075037553891321e-05, "loss": 0.0393, "step": 39560 }, { "grad_norm": 0.29803407192230225, "learning_rate": 7.073533083879315e-05, "loss": 0.0315, "step": 39570 }, { "grad_norm": 0.6166455745697021, "learning_rate": 7.072028387108849e-05, "loss": 0.0339, "step": 39580 }, { "grad_norm": 0.33675819635391235, "learning_rate": 7.070523463744472e-05, "loss": 0.04, "step": 39590 }, { "grad_norm": 0.33757731318473816, "learning_rate": 7.069018313950763e-05, "loss": 0.0279, "step": 39600 }, { "grad_norm": 0.46514013409614563, "learning_rate": 7.067512937892322e-05, "loss": 0.031, "step": 39610 }, { "grad_norm": 0.47257745265960693, "learning_rate": 7.066007335733775e-05, "loss": 0.0338, "step": 39620 }, { "grad_norm": 0.21683768928050995, "learning_rate": 7.064501507639772e-05, "loss": 0.0392, "step": 39630 }, { "grad_norm": 0.5612828135490417, "learning_rate": 7.062995453774987e-05, "loss": 0.0448, "step": 39640 }, { "grad_norm": 0.35276705026626587, "learning_rate": 7.061489174304121e-05, "loss": 0.0558, "step": 39650 }, { "grad_norm": 0.32893821597099304, "learning_rate": 7.0599826693919e-05, "loss": 0.0322, "step": 39660 }, { "grad_norm": 0.46356335282325745, "learning_rate": 7.05847593920307e-05, "loss": 0.0383, "step": 39670 }, { "grad_norm": 0.44601932168006897, "learning_rate": 7.056968983902406e-05, "loss": 0.0383, "step": 39680 }, { "grad_norm": 0.3493087589740753, "learning_rate": 7.055461803654706e-05, "loss": 0.0359, "step": 39690 }, { "grad_norm": 0.38629284501075745, "learning_rate": 7.053954398624794e-05, "loss": 0.0285, "step": 39700 }, { "grad_norm": 0.34022027254104614, "learning_rate": 7.052446768977518e-05, "loss": 0.0351, "step": 39710 }, { "grad_norm": 0.38568007946014404, "learning_rate": 7.050938914877748e-05, "loss": 0.0274, "step": 39720 }, { "grad_norm": 0.38168758153915405, "learning_rate": 7.049430836490382e-05, "loss": 0.0457, "step": 39730 }, { "grad_norm": 0.23455028235912323, "learning_rate": 7.047922533980342e-05, "loss": 0.0298, "step": 39740 }, { "grad_norm": 0.49673497676849365, "learning_rate": 7.046414007512571e-05, "loss": 0.0353, "step": 39750 }, { "grad_norm": 0.31259289383888245, "learning_rate": 7.044905257252042e-05, "loss": 0.0292, "step": 39760 }, { "grad_norm": 0.3331989049911499, "learning_rate": 7.04339628336375e-05, "loss": 0.0373, "step": 39770 }, { "grad_norm": 0.3513185977935791, "learning_rate": 7.041887086012711e-05, "loss": 0.0368, "step": 39780 }, { "grad_norm": 0.313947468996048, "learning_rate": 7.040377665363969e-05, "loss": 0.0314, "step": 39790 }, { "grad_norm": 0.6661657094955444, "learning_rate": 7.038868021582594e-05, "loss": 0.0365, "step": 39800 }, { "grad_norm": 0.29075703024864197, "learning_rate": 7.037358154833679e-05, "loss": 0.0309, "step": 39810 }, { "grad_norm": 0.44279542565345764, "learning_rate": 7.035848065282339e-05, "loss": 0.0377, "step": 39820 }, { "grad_norm": 0.38055193424224854, "learning_rate": 7.034337753093714e-05, "loss": 0.026, "step": 39830 }, { "grad_norm": 1.0421638488769531, "learning_rate": 7.032827218432972e-05, "loss": 0.0345, "step": 39840 }, { "grad_norm": 0.33807215094566345, "learning_rate": 7.031316461465302e-05, "loss": 0.0256, "step": 39850 }, { "grad_norm": 0.34814000129699707, "learning_rate": 7.029805482355915e-05, "loss": 0.0279, "step": 39860 }, { "grad_norm": 0.3715426027774811, "learning_rate": 7.028294281270055e-05, "loss": 0.0338, "step": 39870 }, { "grad_norm": 0.40227630734443665, "learning_rate": 7.026782858372982e-05, "loss": 0.0406, "step": 39880 }, { "grad_norm": 0.3355448246002197, "learning_rate": 7.025271213829982e-05, "loss": 0.0293, "step": 39890 }, { "grad_norm": 0.3735125660896301, "learning_rate": 7.023759347806366e-05, "loss": 0.0226, "step": 39900 }, { "grad_norm": 0.3135559558868408, "learning_rate": 7.022247260467472e-05, "loss": 0.0297, "step": 39910 }, { "grad_norm": 0.38459327816963196, "learning_rate": 7.020734951978658e-05, "loss": 0.0367, "step": 39920 }, { "grad_norm": 0.45735177397727966, "learning_rate": 7.019222422505307e-05, "loss": 0.0366, "step": 39930 }, { "grad_norm": 0.39783310890197754, "learning_rate": 7.017709672212827e-05, "loss": 0.0332, "step": 39940 }, { "grad_norm": 0.38756492733955383, "learning_rate": 7.016196701266652e-05, "loss": 0.0262, "step": 39950 }, { "grad_norm": 0.31356602907180786, "learning_rate": 7.014683509832238e-05, "loss": 0.0325, "step": 39960 }, { "grad_norm": 0.33500030636787415, "learning_rate": 7.013170098075063e-05, "loss": 0.0372, "step": 39970 }, { "grad_norm": 0.4705820083618164, "learning_rate": 7.011656466160632e-05, "loss": 0.0338, "step": 39980 }, { "grad_norm": 0.20225487649440765, "learning_rate": 7.010142614254475e-05, "loss": 0.0267, "step": 39990 }, { "grad_norm": 0.45466065406799316, "learning_rate": 7.008628542522147e-05, "loss": 0.0323, "step": 40000 }, { "grad_norm": 0.5559428930282593, "learning_rate": 7.007114251129218e-05, "loss": 0.0384, "step": 40010 }, { "grad_norm": 0.4148022532463074, "learning_rate": 7.005599740241293e-05, "loss": 0.0286, "step": 40020 }, { "grad_norm": 0.44890329241752625, "learning_rate": 7.004085010023996e-05, "loss": 0.0379, "step": 40030 }, { "grad_norm": 0.4612416923046112, "learning_rate": 7.002570060642976e-05, "loss": 0.0341, "step": 40040 }, { "grad_norm": 0.4844229519367218, "learning_rate": 7.001054892263903e-05, "loss": 0.0372, "step": 40050 }, { "grad_norm": 0.3682408630847931, "learning_rate": 6.999539505052477e-05, "loss": 0.0358, "step": 40060 }, { "grad_norm": 0.29659393429756165, "learning_rate": 6.998023899174418e-05, "loss": 0.0294, "step": 40070 }, { "grad_norm": 0.3131110370159149, "learning_rate": 6.996508074795467e-05, "loss": 0.0233, "step": 40080 }, { "grad_norm": 0.4535856544971466, "learning_rate": 6.994992032081396e-05, "loss": 0.0345, "step": 40090 }, { "grad_norm": 0.3317393362522125, "learning_rate": 6.993475771197995e-05, "loss": 0.0307, "step": 40100 }, { "grad_norm": 0.5729494690895081, "learning_rate": 6.991959292311082e-05, "loss": 0.0352, "step": 40110 }, { "grad_norm": 0.34789854288101196, "learning_rate": 6.990442595586495e-05, "loss": 0.0347, "step": 40120 }, { "grad_norm": 0.357714980840683, "learning_rate": 6.988925681190098e-05, "loss": 0.0375, "step": 40130 }, { "grad_norm": 0.4810512363910675, "learning_rate": 6.987408549287778e-05, "loss": 0.0345, "step": 40140 }, { "grad_norm": 0.47418954968452454, "learning_rate": 6.985891200045449e-05, "loss": 0.0356, "step": 40150 }, { "grad_norm": 0.3444499373435974, "learning_rate": 6.984373633629045e-05, "loss": 0.0432, "step": 40160 }, { "grad_norm": 0.3929014205932617, "learning_rate": 6.98285585020452e-05, "loss": 0.0271, "step": 40170 }, { "grad_norm": 0.3155328631401062, "learning_rate": 6.981337849937864e-05, "loss": 0.0339, "step": 40180 }, { "grad_norm": 0.38162943720817566, "learning_rate": 6.979819632995078e-05, "loss": 0.0323, "step": 40190 }, { "grad_norm": 0.2723025977611542, "learning_rate": 6.978301199542193e-05, "loss": 0.0279, "step": 40200 }, { "grad_norm": 0.4064401388168335, "learning_rate": 6.976782549745263e-05, "loss": 0.0292, "step": 40210 }, { "grad_norm": 0.4232162833213806, "learning_rate": 6.975263683770367e-05, "loss": 0.0355, "step": 40220 }, { "grad_norm": 0.33981239795684814, "learning_rate": 6.973744601783604e-05, "loss": 0.0365, "step": 40230 }, { "grad_norm": 0.2920960783958435, "learning_rate": 6.972225303951097e-05, "loss": 0.0303, "step": 40240 }, { "grad_norm": 0.3094092309474945, "learning_rate": 6.970705790438998e-05, "loss": 0.0268, "step": 40250 }, { "grad_norm": 0.6252551078796387, "learning_rate": 6.969186061413477e-05, "loss": 0.0378, "step": 40260 }, { "grad_norm": 0.3535996973514557, "learning_rate": 6.967666117040727e-05, "loss": 0.0376, "step": 40270 }, { "grad_norm": 0.8091721534729004, "learning_rate": 6.966145957486972e-05, "loss": 0.0356, "step": 40280 }, { "grad_norm": 0.9091442227363586, "learning_rate": 6.964625582918449e-05, "loss": 0.0335, "step": 40290 }, { "grad_norm": 0.349285751581192, "learning_rate": 6.963104993501425e-05, "loss": 0.0318, "step": 40300 }, { "grad_norm": 0.5055934190750122, "learning_rate": 6.961584189402192e-05, "loss": 0.0346, "step": 40310 }, { "grad_norm": 0.37933921813964844, "learning_rate": 6.96006317078706e-05, "loss": 0.0332, "step": 40320 }, { "grad_norm": 0.38420355319976807, "learning_rate": 6.958541937822367e-05, "loss": 0.0381, "step": 40330 }, { "grad_norm": 0.3312539756298065, "learning_rate": 6.957020490674473e-05, "loss": 0.0355, "step": 40340 }, { "grad_norm": 0.5344234108924866, "learning_rate": 6.95549882950976e-05, "loss": 0.0332, "step": 40350 }, { "grad_norm": 0.5899257063865662, "learning_rate": 6.953976954494635e-05, "loss": 0.0319, "step": 40360 }, { "grad_norm": 0.2838801145553589, "learning_rate": 6.952454865795528e-05, "loss": 0.0295, "step": 40370 }, { "grad_norm": 0.3274043798446655, "learning_rate": 6.950932563578892e-05, "loss": 0.0252, "step": 40380 }, { "grad_norm": 0.3291611969470978, "learning_rate": 6.949410048011206e-05, "loss": 0.0236, "step": 40390 }, { "grad_norm": 0.610046923160553, "learning_rate": 6.947887319258966e-05, "loss": 0.0349, "step": 40400 }, { "grad_norm": 0.2851368188858032, "learning_rate": 6.946364377488696e-05, "loss": 0.0306, "step": 40410 }, { "grad_norm": 0.36691874265670776, "learning_rate": 6.944841222866947e-05, "loss": 0.0384, "step": 40420 }, { "grad_norm": 0.37364181876182556, "learning_rate": 6.943317855560284e-05, "loss": 0.0412, "step": 40430 }, { "grad_norm": 0.4003218114376068, "learning_rate": 6.941794275735302e-05, "loss": 0.0295, "step": 40440 }, { "grad_norm": 0.45377397537231445, "learning_rate": 6.94027048355862e-05, "loss": 0.0381, "step": 40450 }, { "grad_norm": 0.43181177973747253, "learning_rate": 6.938746479196871e-05, "loss": 0.039, "step": 40460 }, { "grad_norm": 0.33403533697128296, "learning_rate": 6.937222262816724e-05, "loss": 0.0316, "step": 40470 }, { "grad_norm": 0.3808250427246094, "learning_rate": 6.935697834584865e-05, "loss": 0.0377, "step": 40480 }, { "grad_norm": 0.29123643040657043, "learning_rate": 6.934173194667998e-05, "loss": 0.0272, "step": 40490 }, { "grad_norm": 0.822131335735321, "learning_rate": 6.93264834323286e-05, "loss": 0.0317, "step": 40500 }, { "grad_norm": 0.4178829789161682, "learning_rate": 6.931123280446206e-05, "loss": 0.0395, "step": 40510 }, { "grad_norm": 0.3515895903110504, "learning_rate": 6.929598006474811e-05, "loss": 0.0292, "step": 40520 }, { "grad_norm": 0.3925331234931946, "learning_rate": 6.92807252148548e-05, "loss": 0.0349, "step": 40530 }, { "grad_norm": 0.3911266326904297, "learning_rate": 6.926546825645037e-05, "loss": 0.0314, "step": 40540 }, { "grad_norm": 0.3785173296928406, "learning_rate": 6.92502091912033e-05, "loss": 0.0306, "step": 40550 }, { "grad_norm": 0.3819195330142975, "learning_rate": 6.923494802078229e-05, "loss": 0.0305, "step": 40560 }, { "grad_norm": 0.4213222563266754, "learning_rate": 6.921968474685628e-05, "loss": 0.0316, "step": 40570 }, { "grad_norm": 0.41940200328826904, "learning_rate": 6.920441937109445e-05, "loss": 0.0402, "step": 40580 }, { "grad_norm": 0.3335915505886078, "learning_rate": 6.91891518951662e-05, "loss": 0.033, "step": 40590 }, { "grad_norm": 0.25074368715286255, "learning_rate": 6.917388232074114e-05, "loss": 0.0272, "step": 40600 }, { "grad_norm": 0.39473894238471985, "learning_rate": 6.915861064948914e-05, "loss": 0.0347, "step": 40610 }, { "grad_norm": 0.5633216500282288, "learning_rate": 6.914333688308025e-05, "loss": 0.0325, "step": 40620 }, { "grad_norm": 0.4145261347293854, "learning_rate": 6.912806102318487e-05, "loss": 0.032, "step": 40630 }, { "grad_norm": 0.35976818203926086, "learning_rate": 6.911278307147347e-05, "loss": 0.03, "step": 40640 }, { "grad_norm": 0.3164820969104767, "learning_rate": 6.909750302961684e-05, "loss": 0.0349, "step": 40650 }, { "grad_norm": 0.5604326725006104, "learning_rate": 6.9082220899286e-05, "loss": 0.0423, "step": 40660 }, { "grad_norm": 0.5829684734344482, "learning_rate": 6.906693668215219e-05, "loss": 0.0414, "step": 40670 }, { "grad_norm": 0.3412320911884308, "learning_rate": 6.905165037988683e-05, "loss": 0.0309, "step": 40680 }, { "grad_norm": 0.2907833158969879, "learning_rate": 6.903636199416164e-05, "loss": 0.0291, "step": 40690 }, { "grad_norm": 0.38837987184524536, "learning_rate": 6.902107152664851e-05, "loss": 0.0283, "step": 40700 }, { "grad_norm": 0.376569539308548, "learning_rate": 6.90057789790196e-05, "loss": 0.029, "step": 40710 }, { "grad_norm": 0.371259480714798, "learning_rate": 6.899048435294728e-05, "loss": 0.035, "step": 40720 }, { "grad_norm": 0.4786057770252228, "learning_rate": 6.897518765010415e-05, "loss": 0.0338, "step": 40730 }, { "grad_norm": 0.2908773720264435, "learning_rate": 6.895988887216302e-05, "loss": 0.0324, "step": 40740 }, { "grad_norm": 0.45743659138679504, "learning_rate": 6.894458802079694e-05, "loss": 0.0357, "step": 40750 }, { "grad_norm": 0.35745707154273987, "learning_rate": 6.892928509767922e-05, "loss": 0.0307, "step": 40760 }, { "grad_norm": 0.2627881169319153, "learning_rate": 6.891398010448333e-05, "loss": 0.0335, "step": 40770 }, { "grad_norm": 0.48053038120269775, "learning_rate": 6.8898673042883e-05, "loss": 0.0331, "step": 40780 }, { "grad_norm": 0.46055424213409424, "learning_rate": 6.888336391455222e-05, "loss": 0.0287, "step": 40790 }, { "grad_norm": 0.41919761896133423, "learning_rate": 6.886805272116513e-05, "loss": 0.0305, "step": 40800 }, { "grad_norm": 0.4130299389362335, "learning_rate": 6.885273946439617e-05, "loss": 0.041, "step": 40810 }, { "grad_norm": 0.3451286256313324, "learning_rate": 6.883742414591998e-05, "loss": 0.0362, "step": 40820 }, { "grad_norm": 0.3018101453781128, "learning_rate": 6.88221067674114e-05, "loss": 0.0267, "step": 40830 }, { "grad_norm": 0.4025309681892395, "learning_rate": 6.88067873305455e-05, "loss": 0.0333, "step": 40840 }, { "grad_norm": 0.2720445692539215, "learning_rate": 6.879146583699765e-05, "loss": 0.0241, "step": 40850 }, { "grad_norm": 0.3534170091152191, "learning_rate": 6.877614228844334e-05, "loss": 0.0358, "step": 40860 }, { "grad_norm": 0.8556013703346252, "learning_rate": 6.876081668655832e-05, "loss": 0.0383, "step": 40870 }, { "grad_norm": 0.5571630001068115, "learning_rate": 6.874548903301861e-05, "loss": 0.0404, "step": 40880 }, { "grad_norm": 0.37329405546188354, "learning_rate": 6.873015932950039e-05, "loss": 0.0358, "step": 40890 }, { "grad_norm": 0.2755001485347748, "learning_rate": 6.871482757768012e-05, "loss": 0.0386, "step": 40900 }, { "grad_norm": 0.3875579833984375, "learning_rate": 6.869949377923442e-05, "loss": 0.0374, "step": 40910 }, { "grad_norm": 0.5449829697608948, "learning_rate": 6.868415793584022e-05, "loss": 0.0375, "step": 40920 }, { "grad_norm": 0.4528901278972626, "learning_rate": 6.86688200491746e-05, "loss": 0.0307, "step": 40930 }, { "grad_norm": 1.950386643409729, "learning_rate": 6.865348012091488e-05, "loss": 0.036, "step": 40940 }, { "grad_norm": 0.3985953629016876, "learning_rate": 6.86381381527386e-05, "loss": 0.0339, "step": 40950 }, { "grad_norm": 1.0607725381851196, "learning_rate": 6.862279414632359e-05, "loss": 0.028, "step": 40960 }, { "grad_norm": 0.4392555356025696, "learning_rate": 6.860744810334778e-05, "loss": 0.0299, "step": 40970 }, { "grad_norm": 0.3003736436367035, "learning_rate": 6.859210002548943e-05, "loss": 0.0279, "step": 40980 }, { "grad_norm": 0.28090640902519226, "learning_rate": 6.8576749914427e-05, "loss": 0.0244, "step": 40990 }, { "grad_norm": 0.39618033170700073, "learning_rate": 6.85613977718391e-05, "loss": 0.0364, "step": 41000 }, { "grad_norm": 0.4040965139865875, "learning_rate": 6.854604359940464e-05, "loss": 0.0241, "step": 41010 }, { "grad_norm": 0.3471435606479645, "learning_rate": 6.853068739880276e-05, "loss": 0.0266, "step": 41020 }, { "grad_norm": 0.4223526120185852, "learning_rate": 6.851532917171276e-05, "loss": 0.0381, "step": 41030 }, { "grad_norm": 0.30049747228622437, "learning_rate": 6.849996891981418e-05, "loss": 0.042, "step": 41040 }, { "grad_norm": 0.3142336905002594, "learning_rate": 6.848460664478684e-05, "loss": 0.0342, "step": 41050 }, { "grad_norm": 0.33923494815826416, "learning_rate": 6.846924234831069e-05, "loss": 0.0321, "step": 41060 }, { "grad_norm": 0.3726920485496521, "learning_rate": 6.845387603206597e-05, "loss": 0.0307, "step": 41070 }, { "grad_norm": 0.3168773353099823, "learning_rate": 6.843850769773312e-05, "loss": 0.0253, "step": 41080 }, { "grad_norm": 0.46291255950927734, "learning_rate": 6.842313734699278e-05, "loss": 0.0391, "step": 41090 }, { "grad_norm": 0.3010317087173462, "learning_rate": 6.840776498152584e-05, "loss": 0.0338, "step": 41100 }, { "grad_norm": 0.4699678421020508, "learning_rate": 6.83923906030134e-05, "loss": 0.0318, "step": 41110 }, { "grad_norm": 0.3066607713699341, "learning_rate": 6.837701421313677e-05, "loss": 0.0326, "step": 41120 }, { "grad_norm": 0.46762368083000183, "learning_rate": 6.83616358135775e-05, "loss": 0.0271, "step": 41130 }, { "grad_norm": 0.5330150723457336, "learning_rate": 6.834625540601734e-05, "loss": 0.0399, "step": 41140 }, { "grad_norm": 0.39066582918167114, "learning_rate": 6.833087299213829e-05, "loss": 0.0287, "step": 41150 }, { "grad_norm": 0.37624049186706543, "learning_rate": 6.83154885736225e-05, "loss": 0.0371, "step": 41160 }, { "grad_norm": 0.6174910068511963, "learning_rate": 6.830010215215244e-05, "loss": 0.0338, "step": 41170 }, { "grad_norm": 0.32352492213249207, "learning_rate": 6.828471372941072e-05, "loss": 0.0329, "step": 41180 }, { "grad_norm": 0.44209569692611694, "learning_rate": 6.826932330708018e-05, "loss": 0.0327, "step": 41190 }, { "grad_norm": 0.3291322886943817, "learning_rate": 6.825393088684393e-05, "loss": 0.0231, "step": 41200 }, { "grad_norm": 0.3084726333618164, "learning_rate": 6.823853647038524e-05, "loss": 0.0214, "step": 41210 }, { "grad_norm": 0.3354801833629608, "learning_rate": 6.822314005938762e-05, "loss": 0.0253, "step": 41220 }, { "grad_norm": 0.6570128798484802, "learning_rate": 6.820774165553479e-05, "loss": 0.0363, "step": 41230 }, { "grad_norm": 0.4711529612541199, "learning_rate": 6.819234126051072e-05, "loss": 0.03, "step": 41240 }, { "grad_norm": 0.4715864658355713, "learning_rate": 6.817693887599956e-05, "loss": 0.0335, "step": 41250 }, { "grad_norm": 0.31370240449905396, "learning_rate": 6.816153450368568e-05, "loss": 0.0356, "step": 41260 }, { "grad_norm": 0.5893088579177856, "learning_rate": 6.814612814525372e-05, "loss": 0.0413, "step": 41270 }, { "grad_norm": 0.41660788655281067, "learning_rate": 6.813071980238846e-05, "loss": 0.0262, "step": 41280 }, { "grad_norm": 0.40437179803848267, "learning_rate": 6.811530947677492e-05, "loss": 0.0304, "step": 41290 }, { "grad_norm": 0.327059268951416, "learning_rate": 6.809989717009839e-05, "loss": 0.0239, "step": 41300 }, { "grad_norm": 0.41017431020736694, "learning_rate": 6.808448288404431e-05, "loss": 0.031, "step": 41310 }, { "grad_norm": 0.27538129687309265, "learning_rate": 6.806906662029838e-05, "loss": 0.0301, "step": 41320 }, { "grad_norm": 0.425750195980072, "learning_rate": 6.805364838054647e-05, "loss": 0.0326, "step": 41330 }, { "grad_norm": 0.6485950946807861, "learning_rate": 6.803822816647474e-05, "loss": 0.0295, "step": 41340 }, { "grad_norm": 0.3352968394756317, "learning_rate": 6.802280597976949e-05, "loss": 0.0306, "step": 41350 }, { "grad_norm": 0.4592716097831726, "learning_rate": 6.800738182211727e-05, "loss": 0.0279, "step": 41360 }, { "grad_norm": 0.38600581884384155, "learning_rate": 6.799195569520485e-05, "loss": 0.0355, "step": 41370 }, { "grad_norm": 0.3646111786365509, "learning_rate": 6.79765276007192e-05, "loss": 0.0238, "step": 41380 }, { "grad_norm": 0.4885454773902893, "learning_rate": 6.796109754034753e-05, "loss": 0.0304, "step": 41390 }, { "grad_norm": 0.3859036862850189, "learning_rate": 6.794566551577724e-05, "loss": 0.0302, "step": 41400 }, { "grad_norm": 0.4782105088233948, "learning_rate": 6.793023152869594e-05, "loss": 0.0288, "step": 41410 }, { "grad_norm": 0.3515993058681488, "learning_rate": 6.791479558079147e-05, "loss": 0.027, "step": 41420 }, { "grad_norm": 0.3148062825202942, "learning_rate": 6.789935767375191e-05, "loss": 0.0301, "step": 41430 }, { "grad_norm": 0.26090356707572937, "learning_rate": 6.78839178092655e-05, "loss": 0.0239, "step": 41440 }, { "grad_norm": 0.4636612832546234, "learning_rate": 6.786847598902072e-05, "loss": 0.0298, "step": 41450 }, { "grad_norm": 0.35856595635414124, "learning_rate": 6.785303221470627e-05, "loss": 0.03, "step": 41460 }, { "grad_norm": 0.40960150957107544, "learning_rate": 6.783758648801108e-05, "loss": 0.0295, "step": 41470 }, { "grad_norm": 0.448203444480896, "learning_rate": 6.782213881062422e-05, "loss": 0.043, "step": 41480 }, { "grad_norm": 0.4458475112915039, "learning_rate": 6.780668918423508e-05, "loss": 0.031, "step": 41490 }, { "grad_norm": 0.3354734480381012, "learning_rate": 6.779123761053317e-05, "loss": 0.029, "step": 41500 }, { "grad_norm": 0.31069356203079224, "learning_rate": 6.777578409120826e-05, "loss": 0.0327, "step": 41510 }, { "grad_norm": 0.35576707124710083, "learning_rate": 6.776032862795033e-05, "loss": 0.0391, "step": 41520 }, { "grad_norm": 0.35903576016426086, "learning_rate": 6.774487122244956e-05, "loss": 0.0295, "step": 41530 }, { "grad_norm": 0.3464463949203491, "learning_rate": 6.772941187639637e-05, "loss": 0.033, "step": 41540 }, { "grad_norm": 0.37313970923423767, "learning_rate": 6.771395059148134e-05, "loss": 0.0302, "step": 41550 }, { "grad_norm": 0.32742801308631897, "learning_rate": 6.76984873693953e-05, "loss": 0.0295, "step": 41560 }, { "grad_norm": 0.45121997594833374, "learning_rate": 6.768302221182931e-05, "loss": 0.0325, "step": 41570 }, { "grad_norm": 0.4939795434474945, "learning_rate": 6.766755512047457e-05, "loss": 0.0411, "step": 41580 }, { "grad_norm": 0.3566770553588867, "learning_rate": 6.765208609702259e-05, "loss": 0.0332, "step": 41590 }, { "grad_norm": 0.29916003346443176, "learning_rate": 6.763661514316499e-05, "loss": 0.0255, "step": 41600 }, { "grad_norm": 0.6226722598075867, "learning_rate": 6.76211422605937e-05, "loss": 0.0369, "step": 41610 }, { "grad_norm": 0.36010000109672546, "learning_rate": 6.760566745100076e-05, "loss": 0.0292, "step": 41620 }, { "grad_norm": 0.4015406668186188, "learning_rate": 6.75901907160785e-05, "loss": 0.0291, "step": 41630 }, { "grad_norm": 0.46689289808273315, "learning_rate": 6.757471205751943e-05, "loss": 0.0343, "step": 41640 }, { "grad_norm": 0.5231136679649353, "learning_rate": 6.75592314770163e-05, "loss": 0.0314, "step": 41650 }, { "grad_norm": 0.31040793657302856, "learning_rate": 6.7543748976262e-05, "loss": 0.0393, "step": 41660 }, { "grad_norm": 0.38179630041122437, "learning_rate": 6.752826455694968e-05, "loss": 0.0235, "step": 41670 }, { "grad_norm": 0.40267056226730347, "learning_rate": 6.751277822077271e-05, "loss": 0.038, "step": 41680 }, { "grad_norm": 0.44890183210372925, "learning_rate": 6.749728996942465e-05, "loss": 0.028, "step": 41690 }, { "grad_norm": 0.6404359340667725, "learning_rate": 6.748179980459924e-05, "loss": 0.0231, "step": 41700 }, { "grad_norm": 0.3611139953136444, "learning_rate": 6.746630772799052e-05, "loss": 0.033, "step": 41710 }, { "grad_norm": 0.38332659006118774, "learning_rate": 6.745081374129262e-05, "loss": 0.0268, "step": 41720 }, { "grad_norm": 0.5161888599395752, "learning_rate": 6.743531784619998e-05, "loss": 0.0422, "step": 41730 }, { "grad_norm": 0.44143885374069214, "learning_rate": 6.741982004440719e-05, "loss": 0.0322, "step": 41740 }, { "grad_norm": 0.6718986630439758, "learning_rate": 6.740432033760907e-05, "loss": 0.0342, "step": 41750 }, { "grad_norm": 0.34543102979660034, "learning_rate": 6.738881872750066e-05, "loss": 0.028, "step": 41760 }, { "grad_norm": 0.332364559173584, "learning_rate": 6.737331521577715e-05, "loss": 0.0302, "step": 41770 }, { "grad_norm": 0.844933032989502, "learning_rate": 6.735780980413403e-05, "loss": 0.0343, "step": 41780 }, { "grad_norm": 0.43585216999053955, "learning_rate": 6.734230249426692e-05, "loss": 0.0374, "step": 41790 }, { "grad_norm": 0.36142149567604065, "learning_rate": 6.732679328787168e-05, "loss": 0.0296, "step": 41800 }, { "grad_norm": 0.2605609595775604, "learning_rate": 6.731128218664438e-05, "loss": 0.0257, "step": 41810 }, { "grad_norm": 0.31341880559921265, "learning_rate": 6.729576919228129e-05, "loss": 0.0233, "step": 41820 }, { "grad_norm": 0.3666255474090576, "learning_rate": 6.728025430647888e-05, "loss": 0.0298, "step": 41830 }, { "grad_norm": 0.33515018224716187, "learning_rate": 6.726473753093383e-05, "loss": 0.0305, "step": 41840 }, { "grad_norm": 0.32108843326568604, "learning_rate": 6.724921886734305e-05, "loss": 0.0308, "step": 41850 }, { "grad_norm": 0.3572220504283905, "learning_rate": 6.723369831740363e-05, "loss": 0.0528, "step": 41860 }, { "grad_norm": 0.3380197584629059, "learning_rate": 6.721817588281288e-05, "loss": 0.0264, "step": 41870 }, { "grad_norm": 0.4171501398086548, "learning_rate": 6.720265156526828e-05, "loss": 0.0287, "step": 41880 }, { "grad_norm": 0.2859269678592682, "learning_rate": 6.718712536646758e-05, "loss": 0.0276, "step": 41890 }, { "grad_norm": 0.37834638357162476, "learning_rate": 6.71715972881087e-05, "loss": 0.0347, "step": 41900 }, { "grad_norm": 0.3565898835659027, "learning_rate": 6.715606733188974e-05, "loss": 0.0449, "step": 41910 }, { "grad_norm": 0.3805696666240692, "learning_rate": 6.714053549950909e-05, "loss": 0.0286, "step": 41920 }, { "grad_norm": 0.4813825488090515, "learning_rate": 6.71250017926652e-05, "loss": 0.0356, "step": 41930 }, { "grad_norm": 0.38605064153671265, "learning_rate": 6.71094662130569e-05, "loss": 0.0323, "step": 41940 }, { "grad_norm": 0.24135424196720123, "learning_rate": 6.709392876238307e-05, "loss": 0.0379, "step": 41950 }, { "grad_norm": 0.4276356101036072, "learning_rate": 6.707838944234289e-05, "loss": 0.0338, "step": 41960 }, { "grad_norm": 0.3499841094017029, "learning_rate": 6.706284825463574e-05, "loss": 0.0341, "step": 41970 }, { "grad_norm": 0.553979218006134, "learning_rate": 6.704730520096114e-05, "loss": 0.0354, "step": 41980 }, { "grad_norm": 0.571088433265686, "learning_rate": 6.703176028301888e-05, "loss": 0.0334, "step": 41990 }, { "grad_norm": 0.25768396258354187, "learning_rate": 6.701621350250892e-05, "loss": 0.029, "step": 42000 }, { "grad_norm": 0.37647104263305664, "learning_rate": 6.700066486113144e-05, "loss": 0.0294, "step": 42010 }, { "grad_norm": 0.3670293092727661, "learning_rate": 6.69851143605868e-05, "loss": 0.0277, "step": 42020 }, { "grad_norm": 0.40162450075149536, "learning_rate": 6.696956200257561e-05, "loss": 0.0332, "step": 42030 }, { "grad_norm": 0.24011139571666718, "learning_rate": 6.695400778879863e-05, "loss": 0.0388, "step": 42040 }, { "grad_norm": 0.42470383644104004, "learning_rate": 6.693845172095683e-05, "loss": 0.0395, "step": 42050 }, { "grad_norm": 0.5239251852035522, "learning_rate": 6.692289380075142e-05, "loss": 0.0361, "step": 42060 }, { "grad_norm": 0.33579859137535095, "learning_rate": 6.690733402988379e-05, "loss": 0.028, "step": 42070 }, { "grad_norm": 0.4232112169265747, "learning_rate": 6.689177241005553e-05, "loss": 0.0359, "step": 42080 }, { "grad_norm": 0.4357791244983673, "learning_rate": 6.687620894296844e-05, "loss": 0.0316, "step": 42090 }, { "grad_norm": 0.5062881112098694, "learning_rate": 6.686064363032451e-05, "loss": 0.0289, "step": 42100 }, { "grad_norm": 0.2286771684885025, "learning_rate": 6.684507647382596e-05, "loss": 0.0286, "step": 42110 }, { "grad_norm": 0.5148400068283081, "learning_rate": 6.682950747517513e-05, "loss": 0.038, "step": 42120 }, { "grad_norm": 0.4049733579158783, "learning_rate": 6.68139366360747e-05, "loss": 0.0297, "step": 42130 }, { "grad_norm": 0.2842022776603699, "learning_rate": 6.679836395822744e-05, "loss": 0.0222, "step": 42140 }, { "grad_norm": 0.3564382791519165, "learning_rate": 6.678278944333633e-05, "loss": 0.0312, "step": 42150 }, { "grad_norm": 0.4104882776737213, "learning_rate": 6.676721309310462e-05, "loss": 0.0298, "step": 42160 }, { "grad_norm": 0.24058222770690918, "learning_rate": 6.67516349092357e-05, "loss": 0.0236, "step": 42170 }, { "grad_norm": 0.2829664647579193, "learning_rate": 6.673605489343315e-05, "loss": 0.0272, "step": 42180 }, { "grad_norm": 0.5428150296211243, "learning_rate": 6.672047304740081e-05, "loss": 0.0304, "step": 42190 }, { "grad_norm": 0.2277570217847824, "learning_rate": 6.670488937284268e-05, "loss": 0.0287, "step": 42200 }, { "grad_norm": 0.44396066665649414, "learning_rate": 6.668930387146294e-05, "loss": 0.0331, "step": 42210 }, { "grad_norm": 0.6341966390609741, "learning_rate": 6.667371654496605e-05, "loss": 0.0246, "step": 42220 }, { "grad_norm": 0.33902254700660706, "learning_rate": 6.665812739505659e-05, "loss": 0.0312, "step": 42230 }, { "grad_norm": 0.30095556378364563, "learning_rate": 6.664253642343935e-05, "loss": 0.0321, "step": 42240 }, { "grad_norm": 0.3956132233142853, "learning_rate": 6.662694363181935e-05, "loss": 0.0352, "step": 42250 }, { "grad_norm": 0.3299776315689087, "learning_rate": 6.66113490219018e-05, "loss": 0.0316, "step": 42260 }, { "grad_norm": 0.2904233932495117, "learning_rate": 6.659575259539207e-05, "loss": 0.0284, "step": 42270 }, { "grad_norm": 0.41652897000312805, "learning_rate": 6.65801543539958e-05, "loss": 0.0297, "step": 42280 }, { "grad_norm": 0.3723911941051483, "learning_rate": 6.65645542994188e-05, "loss": 0.0318, "step": 42290 }, { "grad_norm": 0.34894487261772156, "learning_rate": 6.654895243336702e-05, "loss": 0.0207, "step": 42300 }, { "grad_norm": 0.45382457971572876, "learning_rate": 6.65333487575467e-05, "loss": 0.028, "step": 42310 }, { "grad_norm": 0.25408631563186646, "learning_rate": 6.651774327366422e-05, "loss": 0.0243, "step": 42320 }, { "grad_norm": 0.39164552092552185, "learning_rate": 6.650213598342613e-05, "loss": 0.0227, "step": 42330 }, { "grad_norm": 0.33392345905303955, "learning_rate": 6.648652688853931e-05, "loss": 0.0378, "step": 42340 }, { "grad_norm": 0.5484644770622253, "learning_rate": 6.647091599071066e-05, "loss": 0.0332, "step": 42350 }, { "grad_norm": 0.3969813883304596, "learning_rate": 6.645530329164743e-05, "loss": 0.0321, "step": 42360 }, { "grad_norm": 0.38268959522247314, "learning_rate": 6.643968879305694e-05, "loss": 0.036, "step": 42370 }, { "grad_norm": 0.49608010053634644, "learning_rate": 6.642407249664683e-05, "loss": 0.0303, "step": 42380 }, { "grad_norm": 0.30994799733161926, "learning_rate": 6.640845440412483e-05, "loss": 0.0343, "step": 42390 }, { "grad_norm": 0.35290002822875977, "learning_rate": 6.639283451719893e-05, "loss": 0.0311, "step": 42400 }, { "grad_norm": 0.2501683533191681, "learning_rate": 6.637721283757729e-05, "loss": 0.03, "step": 42410 }, { "grad_norm": 0.3545818030834198, "learning_rate": 6.636158936696828e-05, "loss": 0.0303, "step": 42420 }, { "grad_norm": 0.33266976475715637, "learning_rate": 6.634596410708047e-05, "loss": 0.0311, "step": 42430 }, { "grad_norm": 0.2673719823360443, "learning_rate": 6.633033705962257e-05, "loss": 0.0297, "step": 42440 }, { "grad_norm": 0.3470180034637451, "learning_rate": 6.631470822630359e-05, "loss": 0.0294, "step": 42450 }, { "grad_norm": 0.24795475602149963, "learning_rate": 6.629907760883263e-05, "loss": 0.0304, "step": 42460 }, { "grad_norm": 0.3464612364768982, "learning_rate": 6.628344520891905e-05, "loss": 0.0289, "step": 42470 }, { "grad_norm": 0.4297286868095398, "learning_rate": 6.62678110282724e-05, "loss": 0.0264, "step": 42480 }, { "grad_norm": 0.3818489611148834, "learning_rate": 6.625217506860237e-05, "loss": 0.0386, "step": 42490 }, { "grad_norm": 0.39280831813812256, "learning_rate": 6.623653733161892e-05, "loss": 0.0311, "step": 42500 }, { "grad_norm": 0.42458203434944153, "learning_rate": 6.622089781903217e-05, "loss": 0.0325, "step": 42510 }, { "grad_norm": 0.42980819940567017, "learning_rate": 6.620525653255241e-05, "loss": 0.0304, "step": 42520 }, { "grad_norm": 0.380545973777771, "learning_rate": 6.618961347389015e-05, "loss": 0.0251, "step": 42530 }, { "grad_norm": 0.5747089385986328, "learning_rate": 6.617396864475613e-05, "loss": 0.0313, "step": 42540 }, { "grad_norm": 0.28165990114212036, "learning_rate": 6.61583220468612e-05, "loss": 0.0296, "step": 42550 }, { "grad_norm": 0.34675484895706177, "learning_rate": 6.614267368191645e-05, "loss": 0.0287, "step": 42560 }, { "grad_norm": 0.3771248757839203, "learning_rate": 6.61270235516332e-05, "loss": 0.0276, "step": 42570 }, { "grad_norm": 0.3451880216598511, "learning_rate": 6.611137165772287e-05, "loss": 0.0228, "step": 42580 }, { "grad_norm": 0.533363938331604, "learning_rate": 6.60957180018972e-05, "loss": 0.0313, "step": 42590 }, { "grad_norm": 0.3997398614883423, "learning_rate": 6.608006258586797e-05, "loss": 0.0282, "step": 42600 }, { "grad_norm": 0.29553574323654175, "learning_rate": 6.606440541134731e-05, "loss": 0.0272, "step": 42610 }, { "grad_norm": 0.4995110034942627, "learning_rate": 6.60487464800474e-05, "loss": 0.0362, "step": 42620 }, { "grad_norm": 0.4612564146518707, "learning_rate": 6.603308579368071e-05, "loss": 0.028, "step": 42630 }, { "grad_norm": 0.38021495938301086, "learning_rate": 6.601742335395987e-05, "loss": 0.0322, "step": 42640 }, { "grad_norm": 0.3892028033733368, "learning_rate": 6.600175916259769e-05, "loss": 0.0332, "step": 42650 }, { "grad_norm": 0.35894373059272766, "learning_rate": 6.598609322130718e-05, "loss": 0.036, "step": 42660 }, { "grad_norm": 0.3409455716609955, "learning_rate": 6.597042553180155e-05, "loss": 0.0241, "step": 42670 }, { "grad_norm": 0.3177652955055237, "learning_rate": 6.59547560957942e-05, "loss": 0.0338, "step": 42680 }, { "grad_norm": 0.36112692952156067, "learning_rate": 6.59390849149987e-05, "loss": 0.0388, "step": 42690 }, { "grad_norm": 0.3605443835258484, "learning_rate": 6.592341199112886e-05, "loss": 0.0278, "step": 42700 }, { "grad_norm": 0.25624924898147583, "learning_rate": 6.590773732589861e-05, "loss": 0.0315, "step": 42710 }, { "grad_norm": 0.36095744371414185, "learning_rate": 6.589206092102213e-05, "loss": 0.0405, "step": 42720 }, { "grad_norm": 0.3600037097930908, "learning_rate": 6.587638277821375e-05, "loss": 0.0288, "step": 42730 }, { "grad_norm": 0.330282986164093, "learning_rate": 6.586070289918803e-05, "loss": 0.0281, "step": 42740 }, { "grad_norm": 0.4188723862171173, "learning_rate": 6.584502128565968e-05, "loss": 0.0335, "step": 42750 }, { "grad_norm": 0.2969134747982025, "learning_rate": 6.582933793934363e-05, "loss": 0.0288, "step": 42760 }, { "grad_norm": 0.23935234546661377, "learning_rate": 6.581365286195498e-05, "loss": 0.0259, "step": 42770 }, { "grad_norm": 0.30016541481018066, "learning_rate": 6.579796605520903e-05, "loss": 0.0277, "step": 42780 }, { "grad_norm": 0.5789029598236084, "learning_rate": 6.578227752082127e-05, "loss": 0.0389, "step": 42790 }, { "grad_norm": 0.3361278176307678, "learning_rate": 6.576658726050735e-05, "loss": 0.0269, "step": 42800 }, { "grad_norm": 0.30789926648139954, "learning_rate": 6.575089527598316e-05, "loss": 0.0257, "step": 42810 }, { "grad_norm": 0.4747728109359741, "learning_rate": 6.573520156896476e-05, "loss": 0.0272, "step": 42820 }, { "grad_norm": 0.32187944650650024, "learning_rate": 6.571950614116835e-05, "loss": 0.0306, "step": 42830 }, { "grad_norm": 0.3488163352012634, "learning_rate": 6.570380899431039e-05, "loss": 0.0461, "step": 42840 }, { "grad_norm": 0.39403441548347473, "learning_rate": 6.568811013010749e-05, "loss": 0.0265, "step": 42850 }, { "grad_norm": 0.37560606002807617, "learning_rate": 6.567240955027645e-05, "loss": 0.0302, "step": 42860 }, { "grad_norm": 0.39025697112083435, "learning_rate": 6.565670725653427e-05, "loss": 0.0269, "step": 42870 }, { "grad_norm": 0.7870173454284668, "learning_rate": 6.564100325059812e-05, "loss": 0.032, "step": 42880 }, { "grad_norm": 0.5164692997932434, "learning_rate": 6.562529753418538e-05, "loss": 0.034, "step": 42890 }, { "grad_norm": 0.24327336251735687, "learning_rate": 6.56095901090136e-05, "loss": 0.0258, "step": 42900 }, { "grad_norm": 0.4252649247646332, "learning_rate": 6.55938809768005e-05, "loss": 0.0256, "step": 42910 }, { "grad_norm": 0.6185693144798279, "learning_rate": 6.557817013926402e-05, "loss": 0.0275, "step": 42920 }, { "grad_norm": 0.3621532917022705, "learning_rate": 6.556245759812227e-05, "loss": 0.04, "step": 42930 }, { "grad_norm": 0.561265230178833, "learning_rate": 6.554674335509357e-05, "loss": 0.0255, "step": 42940 }, { "grad_norm": 0.31221380829811096, "learning_rate": 6.553102741189638e-05, "loss": 0.0224, "step": 42950 }, { "grad_norm": 0.2818465828895569, "learning_rate": 6.55153097702494e-05, "loss": 0.0304, "step": 42960 }, { "grad_norm": 0.38781023025512695, "learning_rate": 6.549959043187146e-05, "loss": 0.0325, "step": 42970 }, { "grad_norm": 0.30986708402633667, "learning_rate": 6.54838693984816e-05, "loss": 0.0215, "step": 42980 }, { "grad_norm": 0.3432770371437073, "learning_rate": 6.546814667179909e-05, "loss": 0.0244, "step": 42990 }, { "grad_norm": 0.43528029322624207, "learning_rate": 6.545242225354328e-05, "loss": 0.0268, "step": 43000 }, { "grad_norm": 0.5088915228843689, "learning_rate": 6.543669614543383e-05, "loss": 0.0262, "step": 43010 }, { "grad_norm": 0.39226436614990234, "learning_rate": 6.542096834919049e-05, "loss": 0.0296, "step": 43020 }, { "grad_norm": 0.3127358555793762, "learning_rate": 6.540523886653324e-05, "loss": 0.0279, "step": 43030 }, { "grad_norm": 0.3511599898338318, "learning_rate": 6.538950769918222e-05, "loss": 0.0328, "step": 43040 }, { "grad_norm": 0.40615516901016235, "learning_rate": 6.537377484885779e-05, "loss": 0.0329, "step": 43050 }, { "grad_norm": 0.2722136080265045, "learning_rate": 6.535804031728045e-05, "loss": 0.0312, "step": 43060 }, { "grad_norm": 0.3833118975162506, "learning_rate": 6.534230410617092e-05, "loss": 0.0293, "step": 43070 }, { "grad_norm": 0.48216575384140015, "learning_rate": 6.532656621725007e-05, "loss": 0.0293, "step": 43080 }, { "grad_norm": 0.2601541578769684, "learning_rate": 6.531082665223898e-05, "loss": 0.029, "step": 43090 }, { "grad_norm": 0.5579996109008789, "learning_rate": 6.529508541285889e-05, "loss": 0.0316, "step": 43100 }, { "grad_norm": 0.5738544464111328, "learning_rate": 6.527934250083128e-05, "loss": 0.0316, "step": 43110 }, { "grad_norm": 0.319212406873703, "learning_rate": 6.526359791787775e-05, "loss": 0.0243, "step": 43120 }, { "grad_norm": 0.4301707148551941, "learning_rate": 6.524785166572009e-05, "loss": 0.0245, "step": 43130 }, { "grad_norm": 0.28558772802352905, "learning_rate": 6.523210374608029e-05, "loss": 0.0314, "step": 43140 }, { "grad_norm": 0.32634368538856506, "learning_rate": 6.521635416068054e-05, "loss": 0.0273, "step": 43150 }, { "grad_norm": 0.40069442987442017, "learning_rate": 6.520060291124317e-05, "loss": 0.0452, "step": 43160 }, { "grad_norm": 0.9778966307640076, "learning_rate": 6.518484999949072e-05, "loss": 0.0306, "step": 43170 }, { "grad_norm": 0.4153963327407837, "learning_rate": 6.516909542714591e-05, "loss": 0.0343, "step": 43180 }, { "grad_norm": 0.4678776264190674, "learning_rate": 6.515333919593161e-05, "loss": 0.0267, "step": 43190 }, { "grad_norm": 0.29146113991737366, "learning_rate": 6.513758130757094e-05, "loss": 0.0314, "step": 43200 }, { "grad_norm": 0.36745160818099976, "learning_rate": 6.512182176378713e-05, "loss": 0.0397, "step": 43210 }, { "grad_norm": 0.8481568098068237, "learning_rate": 6.510606056630362e-05, "loss": 0.0255, "step": 43220 }, { "grad_norm": 0.3426607549190521, "learning_rate": 6.509029771684405e-05, "loss": 0.0348, "step": 43230 }, { "grad_norm": 0.37837809324264526, "learning_rate": 6.50745332171322e-05, "loss": 0.0336, "step": 43240 }, { "grad_norm": 0.30363065004348755, "learning_rate": 6.505876706889207e-05, "loss": 0.0313, "step": 43250 }, { "grad_norm": 0.2730315327644348, "learning_rate": 6.504299927384781e-05, "loss": 0.0373, "step": 43260 }, { "grad_norm": 0.33608973026275635, "learning_rate": 6.502722983372378e-05, "loss": 0.0297, "step": 43270 }, { "grad_norm": 0.38515734672546387, "learning_rate": 6.501145875024446e-05, "loss": 0.0303, "step": 43280 }, { "grad_norm": 0.39125362038612366, "learning_rate": 6.499568602513459e-05, "loss": 0.0375, "step": 43290 }, { "grad_norm": 0.6304488778114319, "learning_rate": 6.497991166011903e-05, "loss": 0.0288, "step": 43300 }, { "grad_norm": 0.3251555562019348, "learning_rate": 6.49641356569229e-05, "loss": 0.0309, "step": 43310 }, { "grad_norm": 0.34351035952568054, "learning_rate": 6.494835801727135e-05, "loss": 0.0332, "step": 43320 }, { "grad_norm": 0.8738058805465698, "learning_rate": 6.493257874288986e-05, "loss": 0.0312, "step": 43330 }, { "grad_norm": 0.5840418338775635, "learning_rate": 6.491679783550399e-05, "loss": 0.0349, "step": 43340 }, { "grad_norm": 0.6141979098320007, "learning_rate": 6.490101529683957e-05, "loss": 0.0387, "step": 43350 }, { "grad_norm": 0.3558095693588257, "learning_rate": 6.488523112862248e-05, "loss": 0.0394, "step": 43360 }, { "grad_norm": 0.34022921323776245, "learning_rate": 6.486944533257891e-05, "loss": 0.0344, "step": 43370 }, { "grad_norm": 0.3004843592643738, "learning_rate": 6.485365791043517e-05, "loss": 0.0376, "step": 43380 }, { "grad_norm": 0.480924516916275, "learning_rate": 6.483786886391772e-05, "loss": 0.031, "step": 43390 }, { "grad_norm": 0.33466634154319763, "learning_rate": 6.482207819475323e-05, "loss": 0.0257, "step": 43400 }, { "grad_norm": 0.7095600962638855, "learning_rate": 6.480628590466857e-05, "loss": 0.0286, "step": 43410 }, { "grad_norm": 0.35411256551742554, "learning_rate": 6.479049199539074e-05, "loss": 0.0245, "step": 43420 }, { "grad_norm": 0.6237808465957642, "learning_rate": 6.477469646864693e-05, "loss": 0.0335, "step": 43430 }, { "grad_norm": 0.3716028928756714, "learning_rate": 6.475889932616454e-05, "loss": 0.0276, "step": 43440 }, { "grad_norm": 0.3341417908668518, "learning_rate": 6.474310056967111e-05, "loss": 0.0292, "step": 43450 }, { "grad_norm": 0.44544801115989685, "learning_rate": 6.472730020089437e-05, "loss": 0.0326, "step": 43460 }, { "grad_norm": 0.2251034826040268, "learning_rate": 6.471149822156222e-05, "loss": 0.025, "step": 43470 }, { "grad_norm": 0.41211825609207153, "learning_rate": 6.469569463340275e-05, "loss": 0.0237, "step": 43480 }, { "grad_norm": 0.252054899930954, "learning_rate": 6.467988943814418e-05, "loss": 0.0282, "step": 43490 }, { "grad_norm": 0.5216546654701233, "learning_rate": 6.4664082637515e-05, "loss": 0.0305, "step": 43500 }, { "grad_norm": 0.42700326442718506, "learning_rate": 6.464827423324377e-05, "loss": 0.0304, "step": 43510 }, { "grad_norm": 0.37771689891815186, "learning_rate": 6.463246422705931e-05, "loss": 0.0294, "step": 43520 }, { "grad_norm": 0.6301009654998779, "learning_rate": 6.461665262069055e-05, "loss": 0.0312, "step": 43530 }, { "grad_norm": 0.3315533995628357, "learning_rate": 6.460083941586665e-05, "loss": 0.03, "step": 43540 }, { "grad_norm": 1.3092786073684692, "learning_rate": 6.45850246143169e-05, "loss": 0.0284, "step": 43550 }, { "grad_norm": 0.343951016664505, "learning_rate": 6.456920821777077e-05, "loss": 0.0307, "step": 43560 }, { "grad_norm": 0.4269546866416931, "learning_rate": 6.455339022795796e-05, "loss": 0.0365, "step": 43570 }, { "grad_norm": 0.2652205526828766, "learning_rate": 6.453757064660825e-05, "loss": 0.0299, "step": 43580 }, { "grad_norm": 0.38063058257102966, "learning_rate": 6.452174947545169e-05, "loss": 0.0237, "step": 43590 }, { "grad_norm": 0.34473201632499695, "learning_rate": 6.450592671621842e-05, "loss": 0.0274, "step": 43600 }, { "grad_norm": 0.46135538816452026, "learning_rate": 6.449010237063882e-05, "loss": 0.0353, "step": 43610 }, { "grad_norm": 0.43028736114501953, "learning_rate": 6.447427644044343e-05, "loss": 0.0294, "step": 43620 }, { "grad_norm": 0.35216280817985535, "learning_rate": 6.44584489273629e-05, "loss": 0.0329, "step": 43630 }, { "grad_norm": 0.34700721502304077, "learning_rate": 6.444261983312817e-05, "loss": 0.029, "step": 43640 }, { "grad_norm": 0.3707125186920166, "learning_rate": 6.442678915947023e-05, "loss": 0.0273, "step": 43650 }, { "grad_norm": 0.3162952661514282, "learning_rate": 6.441095690812032e-05, "loss": 0.0326, "step": 43660 }, { "grad_norm": 0.31080570816993713, "learning_rate": 6.439512308080982e-05, "loss": 0.0327, "step": 43670 }, { "grad_norm": 0.32795196771621704, "learning_rate": 6.43792876792703e-05, "loss": 0.0249, "step": 43680 }, { "grad_norm": 0.36931416392326355, "learning_rate": 6.43634507052335e-05, "loss": 0.0311, "step": 43690 }, { "grad_norm": 0.5222204327583313, "learning_rate": 6.434761216043135e-05, "loss": 0.0366, "step": 43700 }, { "grad_norm": 0.4764382839202881, "learning_rate": 6.433177204659589e-05, "loss": 0.0338, "step": 43710 }, { "grad_norm": 0.2800212502479553, "learning_rate": 6.431593036545938e-05, "loss": 0.0258, "step": 43720 }, { "grad_norm": 0.32805708050727844, "learning_rate": 6.430008711875426e-05, "loss": 0.0224, "step": 43730 }, { "grad_norm": 0.19429798424243927, "learning_rate": 6.42842423082131e-05, "loss": 0.0324, "step": 43740 }, { "grad_norm": 0.34208229184150696, "learning_rate": 6.426839593556868e-05, "loss": 0.0357, "step": 43750 }, { "grad_norm": 0.42175719141960144, "learning_rate": 6.425254800255392e-05, "loss": 0.0262, "step": 43760 }, { "grad_norm": 0.4972241222858429, "learning_rate": 6.423669851090196e-05, "loss": 0.0353, "step": 43770 }, { "grad_norm": 0.23291945457458496, "learning_rate": 6.422084746234604e-05, "loss": 0.03, "step": 43780 }, { "grad_norm": 0.3164263069629669, "learning_rate": 6.420499485861965e-05, "loss": 0.0278, "step": 43790 }, { "grad_norm": 0.4077319800853729, "learning_rate": 6.418914070145636e-05, "loss": 0.0322, "step": 43800 }, { "grad_norm": 0.3643242120742798, "learning_rate": 6.417328499258996e-05, "loss": 0.0264, "step": 43810 }, { "grad_norm": 0.31321945786476135, "learning_rate": 6.415742773375446e-05, "loss": 0.0247, "step": 43820 }, { "grad_norm": 0.301272988319397, "learning_rate": 6.414156892668393e-05, "loss": 0.0279, "step": 43830 }, { "grad_norm": 0.3159199655056, "learning_rate": 6.412570857311267e-05, "loss": 0.0319, "step": 43840 }, { "grad_norm": 0.3748486042022705, "learning_rate": 6.410984667477518e-05, "loss": 0.0265, "step": 43850 }, { "grad_norm": 0.2820354402065277, "learning_rate": 6.409398323340607e-05, "loss": 0.0262, "step": 43860 }, { "grad_norm": 0.277033269405365, "learning_rate": 6.407811825074012e-05, "loss": 0.0297, "step": 43870 }, { "grad_norm": 0.4591265618801117, "learning_rate": 6.406225172851234e-05, "loss": 0.0464, "step": 43880 }, { "grad_norm": 0.4893956482410431, "learning_rate": 6.404638366845786e-05, "loss": 0.0386, "step": 43890 }, { "grad_norm": 0.5255329608917236, "learning_rate": 6.403051407231196e-05, "loss": 0.038, "step": 43900 }, { "grad_norm": 0.48607686161994934, "learning_rate": 6.401464294181016e-05, "loss": 0.0382, "step": 43910 }, { "grad_norm": 0.3225981593132019, "learning_rate": 6.399877027868808e-05, "loss": 0.035, "step": 43920 }, { "grad_norm": 0.4206812381744385, "learning_rate": 6.39828960846815e-05, "loss": 0.0338, "step": 43930 }, { "grad_norm": 0.43123164772987366, "learning_rate": 6.396702036152644e-05, "loss": 0.036, "step": 43940 }, { "grad_norm": 0.5737733840942383, "learning_rate": 6.395114311095905e-05, "loss": 0.0373, "step": 43950 }, { "grad_norm": 0.3810689449310303, "learning_rate": 6.393526433471559e-05, "loss": 0.0348, "step": 43960 }, { "grad_norm": 0.31171754002571106, "learning_rate": 6.391938403453262e-05, "loss": 0.0334, "step": 43970 }, { "grad_norm": 0.4672233760356903, "learning_rate": 6.390350221214671e-05, "loss": 0.0411, "step": 43980 }, { "grad_norm": 0.3773093521595001, "learning_rate": 6.38876188692947e-05, "loss": 0.0335, "step": 43990 }, { "grad_norm": 0.3325183093547821, "learning_rate": 6.387173400771358e-05, "loss": 0.0326, "step": 44000 }, { "grad_norm": 0.2505892515182495, "learning_rate": 6.385584762914047e-05, "loss": 0.03, "step": 44010 }, { "grad_norm": 0.7029669880867004, "learning_rate": 6.38399597353127e-05, "loss": 0.0319, "step": 44020 }, { "grad_norm": 0.48822611570358276, "learning_rate": 6.382407032796775e-05, "loss": 0.0285, "step": 44030 }, { "grad_norm": 0.35527512431144714, "learning_rate": 6.380817940884325e-05, "loss": 0.029, "step": 44040 }, { "grad_norm": 0.3294227123260498, "learning_rate": 6.379228697967702e-05, "loss": 0.0314, "step": 44050 }, { "grad_norm": 0.4480861723423004, "learning_rate": 6.3776393042207e-05, "loss": 0.031, "step": 44060 }, { "grad_norm": 0.33666086196899414, "learning_rate": 6.376049759817137e-05, "loss": 0.0267, "step": 44070 }, { "grad_norm": 0.2620634138584137, "learning_rate": 6.37446006493084e-05, "loss": 0.0345, "step": 44080 }, { "grad_norm": 0.3504845201969147, "learning_rate": 6.372870219735656e-05, "loss": 0.037, "step": 44090 }, { "grad_norm": 0.3115255534648895, "learning_rate": 6.371280224405452e-05, "loss": 0.038, "step": 44100 }, { "grad_norm": 0.395326167345047, "learning_rate": 6.369690079114101e-05, "loss": 0.0343, "step": 44110 }, { "grad_norm": 0.35368984937667847, "learning_rate": 6.368099784035504e-05, "loss": 0.0349, "step": 44120 }, { "grad_norm": 0.5310772657394409, "learning_rate": 6.366509339343572e-05, "loss": 0.0324, "step": 44130 }, { "grad_norm": 0.35958802700042725, "learning_rate": 6.364918745212232e-05, "loss": 0.0313, "step": 44140 }, { "grad_norm": 0.3534683585166931, "learning_rate": 6.363328001815431e-05, "loss": 0.0345, "step": 44150 }, { "grad_norm": 0.38216909766197205, "learning_rate": 6.361737109327128e-05, "loss": 0.0267, "step": 44160 }, { "grad_norm": 0.31933119893074036, "learning_rate": 6.360146067921301e-05, "loss": 0.0272, "step": 44170 }, { "grad_norm": 0.24873773753643036, "learning_rate": 6.35855487777195e-05, "loss": 0.0258, "step": 44180 }, { "grad_norm": 0.30586251616477966, "learning_rate": 6.356963539053076e-05, "loss": 0.0382, "step": 44190 }, { "grad_norm": 0.3000909686088562, "learning_rate": 6.355372051938711e-05, "loss": 0.0383, "step": 44200 }, { "grad_norm": 0.3763538897037506, "learning_rate": 6.353780416602894e-05, "loss": 0.0352, "step": 44210 }, { "grad_norm": 0.9244351387023926, "learning_rate": 6.352188633219689e-05, "loss": 0.029, "step": 44220 }, { "grad_norm": 0.4519997239112854, "learning_rate": 6.350596701963166e-05, "loss": 0.0249, "step": 44230 }, { "grad_norm": 0.45055440068244934, "learning_rate": 6.349004623007419e-05, "loss": 0.0261, "step": 44240 }, { "grad_norm": 0.9224011898040771, "learning_rate": 6.347412396526555e-05, "loss": 0.0412, "step": 44250 }, { "grad_norm": 0.7099422216415405, "learning_rate": 6.345820022694696e-05, "loss": 0.0392, "step": 44260 }, { "grad_norm": 0.31442031264305115, "learning_rate": 6.344227501685984e-05, "loss": 0.03, "step": 44270 }, { "grad_norm": 0.4705718159675598, "learning_rate": 6.342634833674572e-05, "loss": 0.0343, "step": 44280 }, { "grad_norm": 0.42320317029953003, "learning_rate": 6.341042018834635e-05, "loss": 0.0353, "step": 44290 }, { "grad_norm": 0.33380985260009766, "learning_rate": 6.339449057340359e-05, "loss": 0.0325, "step": 44300 }, { "grad_norm": 0.3841760754585266, "learning_rate": 6.337855949365945e-05, "loss": 0.0315, "step": 44310 }, { "grad_norm": 0.3332182466983795, "learning_rate": 6.336262695085619e-05, "loss": 0.0349, "step": 44320 }, { "grad_norm": 0.36277028918266296, "learning_rate": 6.334669294673612e-05, "loss": 0.0291, "step": 44330 }, { "grad_norm": 0.2547627389431, "learning_rate": 6.333075748304179e-05, "loss": 0.0377, "step": 44340 }, { "grad_norm": 0.46864596009254456, "learning_rate": 6.331482056151585e-05, "loss": 0.0348, "step": 44350 }, { "grad_norm": 0.2953910231590271, "learning_rate": 6.329888218390117e-05, "loss": 0.0281, "step": 44360 }, { "grad_norm": 0.5269335508346558, "learning_rate": 6.328294235194072e-05, "loss": 0.0293, "step": 44370 }, { "grad_norm": 0.4309115707874298, "learning_rate": 6.326700106737765e-05, "loss": 0.0347, "step": 44380 }, { "grad_norm": 0.2913498878479004, "learning_rate": 6.325105833195531e-05, "loss": 0.0285, "step": 44390 }, { "grad_norm": 0.47707319259643555, "learning_rate": 6.323511414741715e-05, "loss": 0.032, "step": 44400 }, { "grad_norm": 0.3646247684955597, "learning_rate": 6.321916851550678e-05, "loss": 0.0399, "step": 44410 }, { "grad_norm": 0.290944904088974, "learning_rate": 6.320322143796806e-05, "loss": 0.0363, "step": 44420 }, { "grad_norm": 0.4745144844055176, "learning_rate": 6.318727291654488e-05, "loss": 0.0275, "step": 44430 }, { "grad_norm": 0.3855837285518646, "learning_rate": 6.317132295298134e-05, "loss": 0.0258, "step": 44440 }, { "grad_norm": 0.3417644202709198, "learning_rate": 6.315537154902173e-05, "loss": 0.0239, "step": 44450 }, { "grad_norm": 0.33508700132369995, "learning_rate": 6.313941870641048e-05, "loss": 0.0277, "step": 44460 }, { "grad_norm": 0.36395561695098877, "learning_rate": 6.312346442689214e-05, "loss": 0.032, "step": 44470 }, { "grad_norm": 0.32930612564086914, "learning_rate": 6.310750871221147e-05, "loss": 0.0296, "step": 44480 }, { "grad_norm": 0.3824504315853119, "learning_rate": 6.309155156411335e-05, "loss": 0.0268, "step": 44490 }, { "grad_norm": 0.31607869267463684, "learning_rate": 6.307559298434284e-05, "loss": 0.0313, "step": 44500 }, { "grad_norm": 0.3032940924167633, "learning_rate": 6.305963297464512e-05, "loss": 0.0244, "step": 44510 }, { "grad_norm": 0.4553558826446533, "learning_rate": 6.304367153676561e-05, "loss": 0.026, "step": 44520 }, { "grad_norm": 0.5650200843811035, "learning_rate": 6.302770867244978e-05, "loss": 0.0244, "step": 44530 }, { "grad_norm": 0.3137357831001282, "learning_rate": 6.301174438344328e-05, "loss": 0.0341, "step": 44540 }, { "grad_norm": 0.299672394990921, "learning_rate": 6.299577867149202e-05, "loss": 0.0403, "step": 44550 }, { "grad_norm": 0.3391458988189697, "learning_rate": 6.297981153834192e-05, "loss": 0.0296, "step": 44560 }, { "grad_norm": 0.2784331142902374, "learning_rate": 6.296384298573916e-05, "loss": 0.0395, "step": 44570 }, { "grad_norm": 0.4064558744430542, "learning_rate": 6.294787301543001e-05, "loss": 0.0479, "step": 44580 }, { "grad_norm": 0.2897711992263794, "learning_rate": 6.293190162916095e-05, "loss": 0.0256, "step": 44590 }, { "grad_norm": 0.3616340160369873, "learning_rate": 6.291592882867855e-05, "loss": 0.0296, "step": 44600 }, { "grad_norm": 0.42658746242523193, "learning_rate": 6.28999546157296e-05, "loss": 0.0329, "step": 44610 }, { "grad_norm": 0.2575434744358063, "learning_rate": 6.288397899206102e-05, "loss": 0.026, "step": 44620 }, { "grad_norm": 0.45219576358795166, "learning_rate": 6.286800195941984e-05, "loss": 0.0434, "step": 44630 }, { "grad_norm": 0.32298094034194946, "learning_rate": 6.285202351955334e-05, "loss": 0.0333, "step": 44640 }, { "grad_norm": 0.3574686050415039, "learning_rate": 6.283604367420887e-05, "loss": 0.0379, "step": 44650 }, { "grad_norm": 0.3473719656467438, "learning_rate": 6.282006242513394e-05, "loss": 0.0278, "step": 44660 }, { "grad_norm": 0.2671148478984833, "learning_rate": 6.280407977407628e-05, "loss": 0.0384, "step": 44670 }, { "grad_norm": 0.3630234897136688, "learning_rate": 6.27880957227837e-05, "loss": 0.0327, "step": 44680 }, { "grad_norm": 0.3278215527534485, "learning_rate": 6.27721102730042e-05, "loss": 0.0336, "step": 44690 }, { "grad_norm": 0.4747059941291809, "learning_rate": 6.275612342648592e-05, "loss": 0.0341, "step": 44700 }, { "grad_norm": 0.4252495765686035, "learning_rate": 6.274013518497716e-05, "loss": 0.0305, "step": 44710 }, { "grad_norm": 0.3231126070022583, "learning_rate": 6.272414555022636e-05, "loss": 0.0205, "step": 44720 }, { "grad_norm": 0.38506513833999634, "learning_rate": 6.270815452398215e-05, "loss": 0.0244, "step": 44730 }, { "grad_norm": 0.2838740348815918, "learning_rate": 6.269216210799326e-05, "loss": 0.0345, "step": 44740 }, { "grad_norm": 0.33387207984924316, "learning_rate": 6.26761683040086e-05, "loss": 0.0371, "step": 44750 }, { "grad_norm": 0.42774951457977295, "learning_rate": 6.266017311377723e-05, "loss": 0.0342, "step": 44760 }, { "grad_norm": 0.32189950346946716, "learning_rate": 6.264417653904839e-05, "loss": 0.0342, "step": 44770 }, { "grad_norm": 0.38909241557121277, "learning_rate": 6.262817858157139e-05, "loss": 0.0319, "step": 44780 }, { "grad_norm": 0.4692456126213074, "learning_rate": 6.261217924309576e-05, "loss": 0.0434, "step": 44790 }, { "grad_norm": 0.3866775631904602, "learning_rate": 6.259617852537118e-05, "loss": 0.0264, "step": 44800 }, { "grad_norm": 0.42718252539634705, "learning_rate": 6.258017643014747e-05, "loss": 0.0329, "step": 44810 }, { "grad_norm": 0.27771544456481934, "learning_rate": 6.256417295917456e-05, "loss": 0.0387, "step": 44820 }, { "grad_norm": 0.2838616669178009, "learning_rate": 6.254816811420258e-05, "loss": 0.0282, "step": 44830 }, { "grad_norm": 0.32195013761520386, "learning_rate": 6.253216189698183e-05, "loss": 0.0335, "step": 44840 }, { "grad_norm": 0.5457612872123718, "learning_rate": 6.251615430926267e-05, "loss": 0.028, "step": 44850 }, { "grad_norm": 0.38547399640083313, "learning_rate": 6.25001453527957e-05, "loss": 0.0323, "step": 44860 }, { "grad_norm": 0.3355644643306732, "learning_rate": 6.248413502933164e-05, "loss": 0.0296, "step": 44870 }, { "grad_norm": 0.31503742933273315, "learning_rate": 6.246812334062133e-05, "loss": 0.0303, "step": 44880 }, { "grad_norm": 0.3816992938518524, "learning_rate": 6.245211028841579e-05, "loss": 0.0294, "step": 44890 }, { "grad_norm": 0.37140583992004395, "learning_rate": 6.24360958744662e-05, "loss": 0.025, "step": 44900 }, { "grad_norm": 0.2650041878223419, "learning_rate": 6.242008010052387e-05, "loss": 0.0264, "step": 44910 }, { "grad_norm": 0.39211660623550415, "learning_rate": 6.240406296834024e-05, "loss": 0.0353, "step": 44920 }, { "grad_norm": 0.4303019642829895, "learning_rate": 6.238804447966694e-05, "loss": 0.022, "step": 44930 }, { "grad_norm": 1.0969135761260986, "learning_rate": 6.237202463625573e-05, "loss": 0.0306, "step": 44940 }, { "grad_norm": 0.5384639501571655, "learning_rate": 6.235600343985848e-05, "loss": 0.0308, "step": 44950 }, { "grad_norm": 0.22744205594062805, "learning_rate": 6.233998089222729e-05, "loss": 0.0269, "step": 44960 }, { "grad_norm": 0.23820184171199799, "learning_rate": 6.232395699511433e-05, "loss": 0.0351, "step": 44970 }, { "grad_norm": 0.2466188222169876, "learning_rate": 6.230793175027199e-05, "loss": 0.0367, "step": 44980 }, { "grad_norm": 0.3167627155780792, "learning_rate": 6.22919051594527e-05, "loss": 0.0254, "step": 44990 }, { "grad_norm": 0.4030609428882599, "learning_rate": 6.227587722440917e-05, "loss": 0.0293, "step": 45000 }, { "grad_norm": 0.378968745470047, "learning_rate": 6.225984794689414e-05, "loss": 0.0304, "step": 45010 }, { "grad_norm": 0.3604598343372345, "learning_rate": 6.22438173286606e-05, "loss": 0.0297, "step": 45020 }, { "grad_norm": 1.1176559925079346, "learning_rate": 6.222778537146159e-05, "loss": 0.029, "step": 45030 }, { "grad_norm": 0.5686715841293335, "learning_rate": 6.221175207705037e-05, "loss": 0.0309, "step": 45040 }, { "grad_norm": 0.3932269513607025, "learning_rate": 6.219571744718027e-05, "loss": 0.027, "step": 45050 }, { "grad_norm": 0.3455570340156555, "learning_rate": 6.217968148360487e-05, "loss": 0.0349, "step": 45060 }, { "grad_norm": 0.3120822012424469, "learning_rate": 6.216364418807782e-05, "loss": 0.0279, "step": 45070 }, { "grad_norm": 0.3191896975040436, "learning_rate": 6.214760556235292e-05, "loss": 0.0259, "step": 45080 }, { "grad_norm": 0.3438718020915985, "learning_rate": 6.213156560818414e-05, "loss": 0.028, "step": 45090 }, { "grad_norm": 0.29856938123703003, "learning_rate": 6.211552432732559e-05, "loss": 0.0301, "step": 45100 }, { "grad_norm": 0.23977887630462646, "learning_rate": 6.20994817215315e-05, "loss": 0.0308, "step": 45110 }, { "grad_norm": 0.3286810517311096, "learning_rate": 6.208343779255628e-05, "loss": 0.0314, "step": 45120 }, { "grad_norm": 0.43828698992729187, "learning_rate": 6.206739254215449e-05, "loss": 0.0324, "step": 45130 }, { "grad_norm": 0.3240658640861511, "learning_rate": 6.205134597208077e-05, "loss": 0.0297, "step": 45140 }, { "grad_norm": 0.3210139274597168, "learning_rate": 6.203529808409e-05, "loss": 0.0223, "step": 45150 }, { "grad_norm": 0.41865894198417664, "learning_rate": 6.201924887993712e-05, "loss": 0.0406, "step": 45160 }, { "grad_norm": 0.4459247291088104, "learning_rate": 6.200319836137725e-05, "loss": 0.0331, "step": 45170 }, { "grad_norm": 0.2773452401161194, "learning_rate": 6.198714653016565e-05, "loss": 0.0319, "step": 45180 }, { "grad_norm": 0.6178678870201111, "learning_rate": 6.197109338805774e-05, "loss": 0.0296, "step": 45190 }, { "grad_norm": 0.36691316962242126, "learning_rate": 6.195503893680903e-05, "loss": 0.0297, "step": 45200 }, { "grad_norm": 0.3336731493473053, "learning_rate": 6.193898317817524e-05, "loss": 0.0244, "step": 45210 }, { "grad_norm": 0.3162611424922943, "learning_rate": 6.192292611391221e-05, "loss": 0.0196, "step": 45220 }, { "grad_norm": 0.35285982489585876, "learning_rate": 6.190686774577591e-05, "loss": 0.0289, "step": 45230 }, { "grad_norm": 0.34867197275161743, "learning_rate": 6.189080807552245e-05, "loss": 0.0318, "step": 45240 }, { "grad_norm": 0.279326856136322, "learning_rate": 6.187474710490809e-05, "loss": 0.0345, "step": 45250 }, { "grad_norm": 0.3360591232776642, "learning_rate": 6.185868483568926e-05, "loss": 0.0285, "step": 45260 }, { "grad_norm": 0.5606098175048828, "learning_rate": 6.184262126962245e-05, "loss": 0.0348, "step": 45270 }, { "grad_norm": 0.42880454659461975, "learning_rate": 6.182655640846442e-05, "loss": 0.026, "step": 45280 }, { "grad_norm": 0.3859589695930481, "learning_rate": 6.181049025397196e-05, "loss": 0.033, "step": 45290 }, { "grad_norm": 0.3109838664531708, "learning_rate": 6.179442280790202e-05, "loss": 0.0393, "step": 45300 }, { "grad_norm": 0.23537205159664154, "learning_rate": 6.177835407201174e-05, "loss": 0.031, "step": 45310 }, { "grad_norm": 0.21115082502365112, "learning_rate": 6.176228404805839e-05, "loss": 0.025, "step": 45320 }, { "grad_norm": 0.4193246066570282, "learning_rate": 6.174621273779932e-05, "loss": 0.0266, "step": 45330 }, { "grad_norm": 0.32221952080726624, "learning_rate": 6.17301401429921e-05, "loss": 0.0293, "step": 45340 }, { "grad_norm": 0.2453061193227768, "learning_rate": 6.171406626539436e-05, "loss": 0.0283, "step": 45350 }, { "grad_norm": 0.46757304668426514, "learning_rate": 6.169799110676398e-05, "loss": 0.048, "step": 45360 }, { "grad_norm": 0.31631651520729065, "learning_rate": 6.168191466885885e-05, "loss": 0.0341, "step": 45370 }, { "grad_norm": 0.5891126990318298, "learning_rate": 6.166583695343711e-05, "loss": 0.0262, "step": 45380 }, { "grad_norm": 0.3278929591178894, "learning_rate": 6.164975796225698e-05, "loss": 0.0291, "step": 45390 }, { "grad_norm": 0.25471892952919006, "learning_rate": 6.163367769707683e-05, "loss": 0.0276, "step": 45400 }, { "grad_norm": 0.2518846094608307, "learning_rate": 6.161759615965519e-05, "loss": 0.0354, "step": 45410 }, { "grad_norm": 0.26759278774261475, "learning_rate": 6.16015133517507e-05, "loss": 0.0256, "step": 45420 }, { "grad_norm": 0.3315194845199585, "learning_rate": 6.158542927512214e-05, "loss": 0.0319, "step": 45430 }, { "grad_norm": 0.2937929332256317, "learning_rate": 6.156934393152846e-05, "loss": 0.0285, "step": 45440 }, { "grad_norm": 0.3329724073410034, "learning_rate": 6.15532573227287e-05, "loss": 0.0361, "step": 45450 }, { "grad_norm": 0.4742008149623871, "learning_rate": 6.153716945048212e-05, "loss": 0.0343, "step": 45460 }, { "grad_norm": 0.4948384165763855, "learning_rate": 6.152108031654802e-05, "loss": 0.0342, "step": 45470 }, { "grad_norm": 0.41260427236557007, "learning_rate": 6.15049899226859e-05, "loss": 0.0327, "step": 45480 }, { "grad_norm": 0.6720632910728455, "learning_rate": 6.148889827065537e-05, "loss": 0.0285, "step": 45490 }, { "grad_norm": 0.5373414754867554, "learning_rate": 6.147280536221622e-05, "loss": 0.025, "step": 45500 }, { "grad_norm": 0.3739789128303528, "learning_rate": 6.145671119912832e-05, "loss": 0.0281, "step": 45510 }, { "grad_norm": 0.27118560671806335, "learning_rate": 6.144061578315169e-05, "loss": 0.027, "step": 45520 }, { "grad_norm": 0.5012187957763672, "learning_rate": 6.142451911604654e-05, "loss": 0.0222, "step": 45530 }, { "grad_norm": 0.23581817746162415, "learning_rate": 6.140842119957315e-05, "loss": 0.0201, "step": 45540 }, { "grad_norm": 0.22599728405475616, "learning_rate": 6.139232203549197e-05, "loss": 0.0263, "step": 45550 }, { "grad_norm": 0.2367623895406723, "learning_rate": 6.137622162556357e-05, "loss": 0.0259, "step": 45560 }, { "grad_norm": 0.359423965215683, "learning_rate": 6.13601199715487e-05, "loss": 0.0308, "step": 45570 }, { "grad_norm": 0.3081298768520355, "learning_rate": 6.134401707520816e-05, "loss": 0.0221, "step": 45580 }, { "grad_norm": 0.25426042079925537, "learning_rate": 6.132791293830298e-05, "loss": 0.0302, "step": 45590 }, { "grad_norm": 0.2949584126472473, "learning_rate": 6.131180756259428e-05, "loss": 0.0254, "step": 45600 }, { "grad_norm": 0.21322403848171234, "learning_rate": 6.129570094984331e-05, "loss": 0.026, "step": 45610 }, { "grad_norm": 0.3225023150444031, "learning_rate": 6.127959310181145e-05, "loss": 0.0351, "step": 45620 }, { "grad_norm": 0.3882937729358673, "learning_rate": 6.126348402026026e-05, "loss": 0.0297, "step": 45630 }, { "grad_norm": 0.21693196892738342, "learning_rate": 6.12473737069514e-05, "loss": 0.0255, "step": 45640 }, { "grad_norm": 0.666485071182251, "learning_rate": 6.123126216364665e-05, "loss": 0.0255, "step": 45650 }, { "grad_norm": 0.5597348213195801, "learning_rate": 6.121514939210797e-05, "loss": 0.0241, "step": 45660 }, { "grad_norm": 0.3791372776031494, "learning_rate": 6.119903539409741e-05, "loss": 0.0303, "step": 45670 }, { "grad_norm": 0.2222471535205841, "learning_rate": 6.118292017137716e-05, "loss": 0.0306, "step": 45680 }, { "grad_norm": 0.34140491485595703, "learning_rate": 6.116680372570959e-05, "loss": 0.0262, "step": 45690 }, { "grad_norm": 0.2890647351741791, "learning_rate": 6.115068605885713e-05, "loss": 0.0317, "step": 45700 }, { "grad_norm": 0.4412383437156677, "learning_rate": 6.113456717258243e-05, "loss": 0.0357, "step": 45710 }, { "grad_norm": 0.32476046681404114, "learning_rate": 6.11184470686482e-05, "loss": 0.028, "step": 45720 }, { "grad_norm": 0.36672332882881165, "learning_rate": 6.110232574881731e-05, "loss": 0.0239, "step": 45730 }, { "grad_norm": 0.47930556535720825, "learning_rate": 6.108620321485277e-05, "loss": 0.0302, "step": 45740 }, { "grad_norm": 0.29683732986450195, "learning_rate": 6.107007946851773e-05, "loss": 0.0237, "step": 45750 }, { "grad_norm": 0.3212144076824188, "learning_rate": 6.105395451157542e-05, "loss": 0.0341, "step": 45760 }, { "grad_norm": 0.4088457226753235, "learning_rate": 6.103782834578928e-05, "loss": 0.0265, "step": 45770 }, { "grad_norm": 0.7454924583435059, "learning_rate": 6.102170097292281e-05, "loss": 0.0314, "step": 45780 }, { "grad_norm": 0.3638675808906555, "learning_rate": 6.1005572394739716e-05, "loss": 0.0319, "step": 45790 }, { "grad_norm": 0.2804865539073944, "learning_rate": 6.0989442613003765e-05, "loss": 0.0293, "step": 45800 }, { "grad_norm": 0.4023095369338989, "learning_rate": 6.097331162947888e-05, "loss": 0.0338, "step": 45810 }, { "grad_norm": 0.32142210006713867, "learning_rate": 6.095717944592914e-05, "loss": 0.0235, "step": 45820 }, { "grad_norm": 0.5365260243415833, "learning_rate": 6.094104606411873e-05, "loss": 0.0409, "step": 45830 }, { "grad_norm": 0.3407062292098999, "learning_rate": 6.0924911485811966e-05, "loss": 0.0452, "step": 45840 }, { "grad_norm": 0.6122772097587585, "learning_rate": 6.090877571277331e-05, "loss": 0.0326, "step": 45850 }, { "grad_norm": 1.3919132947921753, "learning_rate": 6.089263874676736e-05, "loss": 0.033, "step": 45860 }, { "grad_norm": 0.3924405574798584, "learning_rate": 6.0876500589558796e-05, "loss": 0.0313, "step": 45870 }, { "grad_norm": 0.32279807329177856, "learning_rate": 6.086036124291248e-05, "loss": 0.0392, "step": 45880 }, { "grad_norm": 0.32206782698631287, "learning_rate": 6.084422070859339e-05, "loss": 0.0418, "step": 45890 }, { "grad_norm": 0.38288599252700806, "learning_rate": 6.082807898836663e-05, "loss": 0.0272, "step": 45900 }, { "grad_norm": 0.19774165749549866, "learning_rate": 6.081193608399742e-05, "loss": 0.0267, "step": 45910 }, { "grad_norm": 0.4031032919883728, "learning_rate": 6.0795791997251164e-05, "loss": 0.0296, "step": 45920 }, { "grad_norm": 0.3227434456348419, "learning_rate": 6.0779646729893294e-05, "loss": 0.0262, "step": 45930 }, { "grad_norm": 0.31508636474609375, "learning_rate": 6.0763500283689476e-05, "loss": 0.0259, "step": 45940 }, { "grad_norm": 0.4269659221172333, "learning_rate": 6.0747352660405455e-05, "loss": 0.0326, "step": 45950 }, { "grad_norm": 0.5277454257011414, "learning_rate": 6.073120386180709e-05, "loss": 0.0319, "step": 45960 }, { "grad_norm": 0.26899465918540955, "learning_rate": 6.0715053889660425e-05, "loss": 0.0259, "step": 45970 }, { "grad_norm": 0.24334122240543365, "learning_rate": 6.069890274573157e-05, "loss": 0.0269, "step": 45980 }, { "grad_norm": 0.3332505226135254, "learning_rate": 6.068275043178679e-05, "loss": 0.045, "step": 45990 }, { "grad_norm": 0.5129149556159973, "learning_rate": 6.066659694959248e-05, "loss": 0.0265, "step": 46000 }, { "grad_norm": 0.31974831223487854, "learning_rate": 6.0650442300915176e-05, "loss": 0.0253, "step": 46010 }, { "grad_norm": 0.30957648158073425, "learning_rate": 6.063428648752152e-05, "loss": 0.0227, "step": 46020 }, { "grad_norm": 0.38079962134361267, "learning_rate": 6.0618129511178266e-05, "loss": 0.0309, "step": 46030 }, { "grad_norm": 0.2557632029056549, "learning_rate": 6.060197137365234e-05, "loss": 0.0249, "step": 46040 }, { "grad_norm": 0.2731754183769226, "learning_rate": 6.058581207671077e-05, "loss": 0.0257, "step": 46050 }, { "grad_norm": 0.28124597668647766, "learning_rate": 6.056965162212072e-05, "loss": 0.0317, "step": 46060 }, { "grad_norm": 0.311191201210022, "learning_rate": 6.055349001164943e-05, "loss": 0.0362, "step": 46070 }, { "grad_norm": 0.26613593101501465, "learning_rate": 6.0537327247064347e-05, "loss": 0.0275, "step": 46080 }, { "grad_norm": 0.5250742435455322, "learning_rate": 6.0521163330133014e-05, "loss": 0.0357, "step": 46090 }, { "grad_norm": 0.4400927722454071, "learning_rate": 6.050499826262306e-05, "loss": 0.0265, "step": 46100 }, { "grad_norm": 0.3797573745250702, "learning_rate": 6.0488832046302294e-05, "loss": 0.0367, "step": 46110 }, { "grad_norm": 0.19286152720451355, "learning_rate": 6.0472664682938626e-05, "loss": 0.0254, "step": 46120 }, { "grad_norm": 0.2710513770580292, "learning_rate": 6.045649617430009e-05, "loss": 0.0241, "step": 46130 }, { "grad_norm": 0.3214162290096283, "learning_rate": 6.0440326522154866e-05, "loss": 0.0252, "step": 46140 }, { "grad_norm": 0.3560466170310974, "learning_rate": 6.0424155728271224e-05, "loss": 0.0313, "step": 46150 }, { "grad_norm": 0.3115043342113495, "learning_rate": 6.040798379441758e-05, "loss": 0.0246, "step": 46160 }, { "grad_norm": 0.38773801922798157, "learning_rate": 6.0391810722362485e-05, "loss": 0.0271, "step": 46170 }, { "grad_norm": 0.4580755829811096, "learning_rate": 6.037563651387458e-05, "loss": 0.0312, "step": 46180 }, { "grad_norm": 0.3792092502117157, "learning_rate": 6.0359461170722666e-05, "loss": 0.024, "step": 46190 }, { "grad_norm": 0.26809391379356384, "learning_rate": 6.034328469467566e-05, "loss": 0.0292, "step": 46200 }, { "grad_norm": 0.8247509598731995, "learning_rate": 6.0327107087502596e-05, "loss": 0.0276, "step": 46210 }, { "grad_norm": 0.2738238573074341, "learning_rate": 6.031092835097262e-05, "loss": 0.0257, "step": 46220 }, { "grad_norm": 0.2197989672422409, "learning_rate": 6.0294748486855024e-05, "loss": 0.0265, "step": 46230 }, { "grad_norm": 0.21232569217681885, "learning_rate": 6.0278567496919216e-05, "loss": 0.0256, "step": 46240 }, { "grad_norm": 0.3790738880634308, "learning_rate": 6.026238538293472e-05, "loss": 0.0256, "step": 46250 }, { "grad_norm": 0.3075115978717804, "learning_rate": 6.024620214667118e-05, "loss": 0.0301, "step": 46260 }, { "grad_norm": 0.36695873737335205, "learning_rate": 6.0230017789898384e-05, "loss": 0.02, "step": 46270 }, { "grad_norm": 0.19445808231830597, "learning_rate": 6.021383231438622e-05, "loss": 0.0247, "step": 46280 }, { "grad_norm": 0.3829520642757416, "learning_rate": 6.0197645721904704e-05, "loss": 0.0279, "step": 46290 }, { "grad_norm": 0.386905699968338, "learning_rate": 6.0181458014224e-05, "loss": 0.0348, "step": 46300 }, { "grad_norm": 0.25665292143821716, "learning_rate": 6.016526919311435e-05, "loss": 0.0279, "step": 46310 }, { "grad_norm": 0.40369459986686707, "learning_rate": 6.014907926034613e-05, "loss": 0.0264, "step": 46320 }, { "grad_norm": 0.329685777425766, "learning_rate": 6.0132888217689875e-05, "loss": 0.0304, "step": 46330 }, { "grad_norm": 0.5104641914367676, "learning_rate": 6.01166960669162e-05, "loss": 0.0409, "step": 46340 }, { "grad_norm": 0.294525146484375, "learning_rate": 6.0100502809795845e-05, "loss": 0.0309, "step": 46350 }, { "grad_norm": 0.3066065013408661, "learning_rate": 6.00843084480997e-05, "loss": 0.0362, "step": 46360 }, { "grad_norm": 0.22667470574378967, "learning_rate": 6.0068112983598736e-05, "loss": 0.028, "step": 46370 }, { "grad_norm": 0.3195526599884033, "learning_rate": 6.005191641806407e-05, "loss": 0.0285, "step": 46380 }, { "grad_norm": 0.37107354402542114, "learning_rate": 6.003571875326694e-05, "loss": 0.0289, "step": 46390 }, { "grad_norm": 0.2298482060432434, "learning_rate": 6.00195199909787e-05, "loss": 0.0272, "step": 46400 }, { "grad_norm": 0.37696021795272827, "learning_rate": 6.000332013297082e-05, "loss": 0.0312, "step": 46410 }, { "grad_norm": 0.3755608797073364, "learning_rate": 5.998711918101487e-05, "loss": 0.0282, "step": 46420 }, { "grad_norm": 0.28239378333091736, "learning_rate": 5.997091713688261e-05, "loss": 0.0266, "step": 46430 }, { "grad_norm": 0.3845959007740021, "learning_rate": 5.9954714002345836e-05, "loss": 0.0253, "step": 46440 }, { "grad_norm": 0.3186633884906769, "learning_rate": 5.993850977917649e-05, "loss": 0.0269, "step": 46450 }, { "grad_norm": 0.37208905816078186, "learning_rate": 5.992230446914667e-05, "loss": 0.0254, "step": 46460 }, { "grad_norm": 0.34703347086906433, "learning_rate": 5.990609807402855e-05, "loss": 0.0291, "step": 46470 }, { "grad_norm": 0.3245585262775421, "learning_rate": 5.988989059559443e-05, "loss": 0.0327, "step": 46480 }, { "grad_norm": 0.5169079899787903, "learning_rate": 5.987368203561675e-05, "loss": 0.0276, "step": 46490 }, { "grad_norm": 0.3504091203212738, "learning_rate": 5.9857472395868055e-05, "loss": 0.0261, "step": 46500 }, { "grad_norm": 0.3205493986606598, "learning_rate": 5.9841261678120983e-05, "loss": 0.033, "step": 46510 }, { "grad_norm": 0.3037155568599701, "learning_rate": 5.982504988414834e-05, "loss": 0.022, "step": 46520 }, { "grad_norm": 0.39379608631134033, "learning_rate": 5.9808837015723015e-05, "loss": 0.0407, "step": 46530 }, { "grad_norm": 0.34789353609085083, "learning_rate": 5.9792623074618016e-05, "loss": 0.0271, "step": 46540 }, { "grad_norm": 0.5027474164962769, "learning_rate": 5.977640806260648e-05, "loss": 0.0243, "step": 46550 }, { "grad_norm": 0.21963004767894745, "learning_rate": 5.976019198146166e-05, "loss": 0.0206, "step": 46560 }, { "grad_norm": 0.18762095272541046, "learning_rate": 5.974397483295692e-05, "loss": 0.0284, "step": 46570 }, { "grad_norm": 0.3971823751926422, "learning_rate": 5.972775661886575e-05, "loss": 0.0314, "step": 46580 }, { "grad_norm": 0.3045651912689209, "learning_rate": 5.971153734096173e-05, "loss": 0.0222, "step": 46590 }, { "grad_norm": 0.2809770703315735, "learning_rate": 5.969531700101859e-05, "loss": 0.0242, "step": 46600 }, { "grad_norm": 0.2815982401371002, "learning_rate": 5.9679095600810155e-05, "loss": 0.0339, "step": 46610 }, { "grad_norm": 0.3795882761478424, "learning_rate": 5.9662873142110384e-05, "loss": 0.024, "step": 46620 }, { "grad_norm": 0.4695439040660858, "learning_rate": 5.964664962669333e-05, "loss": 0.0252, "step": 46630 }, { "grad_norm": 0.31649214029312134, "learning_rate": 5.9630425056333186e-05, "loss": 0.0239, "step": 46640 }, { "grad_norm": 0.3462187945842743, "learning_rate": 5.961419943280422e-05, "loss": 0.0315, "step": 46650 }, { "grad_norm": 0.3231242299079895, "learning_rate": 5.959797275788087e-05, "loss": 0.0264, "step": 46660 }, { "grad_norm": 0.27924689650535583, "learning_rate": 5.958174503333765e-05, "loss": 0.0317, "step": 46670 }, { "grad_norm": 0.266418993473053, "learning_rate": 5.9565516260949195e-05, "loss": 0.0299, "step": 46680 }, { "grad_norm": 0.3307729959487915, "learning_rate": 5.9549286442490273e-05, "loss": 0.0247, "step": 46690 }, { "grad_norm": 1.4230401515960693, "learning_rate": 5.953305557973572e-05, "loss": 0.0336, "step": 46700 }, { "grad_norm": 0.5997231602668762, "learning_rate": 5.9516823674460564e-05, "loss": 0.0231, "step": 46710 }, { "grad_norm": 0.34724465012550354, "learning_rate": 5.9500590728439875e-05, "loss": 0.0306, "step": 46720 }, { "grad_norm": 0.3224658668041229, "learning_rate": 5.948435674344886e-05, "loss": 0.0346, "step": 46730 }, { "grad_norm": 0.31806033849716187, "learning_rate": 5.946812172126285e-05, "loss": 0.0247, "step": 46740 }, { "grad_norm": 0.29989367723464966, "learning_rate": 5.945188566365729e-05, "loss": 0.0281, "step": 46750 }, { "grad_norm": 0.3251211643218994, "learning_rate": 5.943564857240773e-05, "loss": 0.0336, "step": 46760 }, { "grad_norm": 0.35329872369766235, "learning_rate": 5.941941044928981e-05, "loss": 0.03, "step": 46770 }, { "grad_norm": 0.2787095010280609, "learning_rate": 5.940317129607935e-05, "loss": 0.0305, "step": 46780 }, { "grad_norm": 0.27157509326934814, "learning_rate": 5.9386931114552204e-05, "loss": 0.0276, "step": 46790 }, { "grad_norm": 0.5114303231239319, "learning_rate": 5.937068990648438e-05, "loss": 0.0343, "step": 46800 }, { "grad_norm": 0.35631921887397766, "learning_rate": 5.935444767365199e-05, "loss": 0.0421, "step": 46810 }, { "grad_norm": 0.3559451103210449, "learning_rate": 5.933820441783129e-05, "loss": 0.0322, "step": 46820 }, { "grad_norm": 0.28996700048446655, "learning_rate": 5.932196014079857e-05, "loss": 0.0321, "step": 46830 }, { "grad_norm": 0.3030775487422943, "learning_rate": 5.930571484433032e-05, "loss": 0.0256, "step": 46840 }, { "grad_norm": 0.3559708595275879, "learning_rate": 5.928946853020309e-05, "loss": 0.0213, "step": 46850 }, { "grad_norm": 0.2596319615840912, "learning_rate": 5.927322120019352e-05, "loss": 0.0247, "step": 46860 }, { "grad_norm": 0.25431564450263977, "learning_rate": 5.925697285607844e-05, "loss": 0.0311, "step": 46870 }, { "grad_norm": 0.20941774547100067, "learning_rate": 5.924072349963472e-05, "loss": 0.029, "step": 46880 }, { "grad_norm": 0.43137750029563904, "learning_rate": 5.922447313263938e-05, "loss": 0.0315, "step": 46890 }, { "grad_norm": 0.35746073722839355, "learning_rate": 5.920822175686952e-05, "loss": 0.0256, "step": 46900 }, { "grad_norm": 0.3584725260734558, "learning_rate": 5.919196937410239e-05, "loss": 0.0322, "step": 46910 }, { "grad_norm": 0.46811074018478394, "learning_rate": 5.917571598611529e-05, "loss": 0.0291, "step": 46920 }, { "grad_norm": 0.6469347476959229, "learning_rate": 5.9159461594685686e-05, "loss": 0.0258, "step": 46930 }, { "grad_norm": 0.34404730796813965, "learning_rate": 5.914320620159114e-05, "loss": 0.0324, "step": 46940 }, { "grad_norm": 0.3313755691051483, "learning_rate": 5.912694980860931e-05, "loss": 0.0273, "step": 46950 }, { "grad_norm": 0.32044386863708496, "learning_rate": 5.9110692417517964e-05, "loss": 0.0295, "step": 46960 }, { "grad_norm": 0.30964937806129456, "learning_rate": 5.9094434030095e-05, "loss": 0.0319, "step": 46970 }, { "grad_norm": 0.21099288761615753, "learning_rate": 5.907817464811841e-05, "loss": 0.0301, "step": 46980 }, { "grad_norm": 0.33858606219291687, "learning_rate": 5.906191427336627e-05, "loss": 0.025, "step": 46990 }, { "grad_norm": 0.4310038685798645, "learning_rate": 5.9045652907616835e-05, "loss": 0.0283, "step": 47000 }, { "grad_norm": 0.4159489572048187, "learning_rate": 5.902939055264838e-05, "loss": 0.0295, "step": 47010 }, { "grad_norm": 0.2814152240753174, "learning_rate": 5.901312721023935e-05, "loss": 0.0275, "step": 47020 }, { "grad_norm": 0.4403281509876251, "learning_rate": 5.8996862882168294e-05, "loss": 0.0389, "step": 47030 }, { "grad_norm": 0.3959895074367523, "learning_rate": 5.8980597570213826e-05, "loss": 0.0396, "step": 47040 }, { "grad_norm": 0.45760732889175415, "learning_rate": 5.896433127615471e-05, "loss": 0.0277, "step": 47050 }, { "grad_norm": 0.24857577681541443, "learning_rate": 5.894806400176981e-05, "loss": 0.0318, "step": 47060 }, { "grad_norm": 0.32329264283180237, "learning_rate": 5.893179574883808e-05, "loss": 0.0353, "step": 47070 }, { "grad_norm": 0.3303690254688263, "learning_rate": 5.8915526519138585e-05, "loss": 0.0285, "step": 47080 }, { "grad_norm": 0.30412283539772034, "learning_rate": 5.889925631445053e-05, "loss": 0.0253, "step": 47090 }, { "grad_norm": 0.5430341362953186, "learning_rate": 5.888298513655318e-05, "loss": 0.0312, "step": 47100 }, { "grad_norm": 0.3544090986251831, "learning_rate": 5.886671298722592e-05, "loss": 0.0198, "step": 47110 }, { "grad_norm": 0.3333195447921753, "learning_rate": 5.885043986824828e-05, "loss": 0.0251, "step": 47120 }, { "grad_norm": 0.32372355461120605, "learning_rate": 5.883416578139982e-05, "loss": 0.0239, "step": 47130 }, { "grad_norm": 0.26792991161346436, "learning_rate": 5.881789072846029e-05, "loss": 0.0306, "step": 47140 }, { "grad_norm": 0.30151647329330444, "learning_rate": 5.8801614711209474e-05, "loss": 0.0215, "step": 47150 }, { "grad_norm": 0.4010801315307617, "learning_rate": 5.878533773142732e-05, "loss": 0.0309, "step": 47160 }, { "grad_norm": 0.3867208957672119, "learning_rate": 5.8769059790893834e-05, "loss": 0.0291, "step": 47170 }, { "grad_norm": 0.30748000741004944, "learning_rate": 5.8752780891389145e-05, "loss": 0.0257, "step": 47180 }, { "grad_norm": 0.4385201036930084, "learning_rate": 5.873650103469351e-05, "loss": 0.0285, "step": 47190 }, { "grad_norm": 0.3566100001335144, "learning_rate": 5.872022022258725e-05, "loss": 0.0234, "step": 47200 }, { "grad_norm": 0.4177057147026062, "learning_rate": 5.8703938456850814e-05, "loss": 0.0237, "step": 47210 }, { "grad_norm": 0.5804566740989685, "learning_rate": 5.868765573926477e-05, "loss": 0.0306, "step": 47220 }, { "grad_norm": 0.4223233461380005, "learning_rate": 5.867137207160974e-05, "loss": 0.0305, "step": 47230 }, { "grad_norm": 0.24041183292865753, "learning_rate": 5.86550874556665e-05, "loss": 0.0258, "step": 47240 }, { "grad_norm": 0.23566363751888275, "learning_rate": 5.863880189321592e-05, "loss": 0.0286, "step": 47250 }, { "grad_norm": 0.5600816011428833, "learning_rate": 5.862251538603894e-05, "loss": 0.0397, "step": 47260 }, { "grad_norm": 0.40784287452697754, "learning_rate": 5.8606227935916656e-05, "loss": 0.0296, "step": 47270 }, { "grad_norm": 0.23981280624866486, "learning_rate": 5.858993954463021e-05, "loss": 0.0237, "step": 47280 }, { "grad_norm": 0.2590555250644684, "learning_rate": 5.8573650213960896e-05, "loss": 0.0311, "step": 47290 }, { "grad_norm": 0.29698893427848816, "learning_rate": 5.8557359945690084e-05, "loss": 0.0242, "step": 47300 }, { "grad_norm": 0.3713221251964569, "learning_rate": 5.8541068741599246e-05, "loss": 0.0286, "step": 47310 }, { "grad_norm": 0.293058305978775, "learning_rate": 5.8524776603469985e-05, "loss": 0.0283, "step": 47320 }, { "grad_norm": 0.23486390709877014, "learning_rate": 5.850848353308397e-05, "loss": 0.0272, "step": 47330 }, { "grad_norm": 0.36319154500961304, "learning_rate": 5.849218953222297e-05, "loss": 0.0238, "step": 47340 }, { "grad_norm": 0.27163365483283997, "learning_rate": 5.847589460266891e-05, "loss": 0.0263, "step": 47350 }, { "grad_norm": 0.606987714767456, "learning_rate": 5.845959874620375e-05, "loss": 0.0275, "step": 47360 }, { "grad_norm": 0.3161807358264923, "learning_rate": 5.84433019646096e-05, "loss": 0.0318, "step": 47370 }, { "grad_norm": 0.27373209595680237, "learning_rate": 5.842700425966863e-05, "loss": 0.0277, "step": 47380 }, { "grad_norm": 0.31942319869995117, "learning_rate": 5.841070563316315e-05, "loss": 0.0258, "step": 47390 }, { "grad_norm": 0.4238526225090027, "learning_rate": 5.839440608687554e-05, "loss": 0.0301, "step": 47400 }, { "grad_norm": 0.34334102272987366, "learning_rate": 5.837810562258831e-05, "loss": 0.0309, "step": 47410 }, { "grad_norm": 0.30216923356056213, "learning_rate": 5.836180424208405e-05, "loss": 0.0319, "step": 47420 }, { "grad_norm": 0.34151723980903625, "learning_rate": 5.834550194714543e-05, "loss": 0.0259, "step": 47430 }, { "grad_norm": 0.3633962869644165, "learning_rate": 5.832919873955528e-05, "loss": 0.0364, "step": 47440 }, { "grad_norm": 0.22693060338497162, "learning_rate": 5.831289462109648e-05, "loss": 0.0278, "step": 47450 }, { "grad_norm": 0.26981109380722046, "learning_rate": 5.829658959355202e-05, "loss": 0.0422, "step": 47460 }, { "grad_norm": 0.4031575620174408, "learning_rate": 5.8280283658704995e-05, "loss": 0.0315, "step": 47470 }, { "grad_norm": 0.4168596863746643, "learning_rate": 5.826397681833859e-05, "loss": 0.0265, "step": 47480 }, { "grad_norm": 0.4835536479949951, "learning_rate": 5.824766907423612e-05, "loss": 0.0246, "step": 47490 }, { "grad_norm": 0.2737116515636444, "learning_rate": 5.8231360428180945e-05, "loss": 0.027, "step": 47500 }, { "grad_norm": 0.30986467003822327, "learning_rate": 5.821505088195658e-05, "loss": 0.0283, "step": 47510 }, { "grad_norm": 0.43354201316833496, "learning_rate": 5.819874043734661e-05, "loss": 0.0297, "step": 47520 }, { "grad_norm": 0.5326836109161377, "learning_rate": 5.8182429096134695e-05, "loss": 0.0259, "step": 47530 }, { "grad_norm": 0.17477348446846008, "learning_rate": 5.816611686010465e-05, "loss": 0.0233, "step": 47540 }, { "grad_norm": 0.40172213315963745, "learning_rate": 5.814980373104033e-05, "loss": 0.0351, "step": 47550 }, { "grad_norm": 0.475716769695282, "learning_rate": 5.813348971072572e-05, "loss": 0.0301, "step": 47560 }, { "grad_norm": 0.2726936638355255, "learning_rate": 5.811717480094492e-05, "loss": 0.0269, "step": 47570 }, { "grad_norm": 0.4962204694747925, "learning_rate": 5.810085900348209e-05, "loss": 0.0293, "step": 47580 }, { "grad_norm": 0.36777064204216003, "learning_rate": 5.8084542320121483e-05, "loss": 0.0293, "step": 47590 }, { "grad_norm": 0.3583383858203888, "learning_rate": 5.8068224752647497e-05, "loss": 0.045, "step": 47600 }, { "grad_norm": 0.4158218502998352, "learning_rate": 5.805190630284457e-05, "loss": 0.0243, "step": 47610 }, { "grad_norm": 0.4127182960510254, "learning_rate": 5.803558697249729e-05, "loss": 0.0282, "step": 47620 }, { "grad_norm": 0.33928820490837097, "learning_rate": 5.8019266763390276e-05, "loss": 0.0355, "step": 47630 }, { "grad_norm": 0.9608336091041565, "learning_rate": 5.800294567730833e-05, "loss": 0.0278, "step": 47640 }, { "grad_norm": 0.2517803907394409, "learning_rate": 5.798662371603628e-05, "loss": 0.0289, "step": 47650 }, { "grad_norm": 0.3334357440471649, "learning_rate": 5.797030088135904e-05, "loss": 0.027, "step": 47660 }, { "grad_norm": 0.35605546832084656, "learning_rate": 5.7953977175061705e-05, "loss": 0.0231, "step": 47670 }, { "grad_norm": 0.2964875400066376, "learning_rate": 5.793765259892938e-05, "loss": 0.0306, "step": 47680 }, { "grad_norm": 0.3866051137447357, "learning_rate": 5.792132715474729e-05, "loss": 0.0238, "step": 47690 }, { "grad_norm": 0.9031987190246582, "learning_rate": 5.790500084430078e-05, "loss": 0.0222, "step": 47700 }, { "grad_norm": 0.3892650306224823, "learning_rate": 5.7888673669375274e-05, "loss": 0.0264, "step": 47710 }, { "grad_norm": 0.3367617130279541, "learning_rate": 5.787234563175625e-05, "loss": 0.0247, "step": 47720 }, { "grad_norm": 1.4003863334655762, "learning_rate": 5.7856016733229355e-05, "loss": 0.0332, "step": 47730 }, { "grad_norm": 0.3207932710647583, "learning_rate": 5.7839686975580297e-05, "loss": 0.0346, "step": 47740 }, { "grad_norm": 0.3629719913005829, "learning_rate": 5.782335636059484e-05, "loss": 0.0244, "step": 47750 }, { "grad_norm": 0.36981621384620667, "learning_rate": 5.780702489005889e-05, "loss": 0.0313, "step": 47760 }, { "grad_norm": 0.3233462870121002, "learning_rate": 5.779069256575845e-05, "loss": 0.0301, "step": 47770 }, { "grad_norm": 0.3768742084503174, "learning_rate": 5.7774359389479574e-05, "loss": 0.0397, "step": 47780 }, { "grad_norm": 0.31739065051078796, "learning_rate": 5.775802536300845e-05, "loss": 0.0284, "step": 47790 }, { "grad_norm": 0.25507333874702454, "learning_rate": 5.774169048813134e-05, "loss": 0.0267, "step": 47800 }, { "grad_norm": 0.3287622332572937, "learning_rate": 5.77253547666346e-05, "loss": 0.0258, "step": 47810 }, { "grad_norm": 0.4410226345062256, "learning_rate": 5.770901820030465e-05, "loss": 0.0322, "step": 47820 }, { "grad_norm": 0.33664271235466003, "learning_rate": 5.769268079092809e-05, "loss": 0.0286, "step": 47830 }, { "grad_norm": 0.19494113326072693, "learning_rate": 5.767634254029151e-05, "loss": 0.019, "step": 47840 }, { "grad_norm": 0.47791966795921326, "learning_rate": 5.7660003450181655e-05, "loss": 0.048, "step": 47850 }, { "grad_norm": 0.24857576191425323, "learning_rate": 5.764366352238534e-05, "loss": 0.0189, "step": 47860 }, { "grad_norm": 0.24915291368961334, "learning_rate": 5.7627322758689474e-05, "loss": 0.0203, "step": 47870 }, { "grad_norm": 0.37560445070266724, "learning_rate": 5.761098116088105e-05, "loss": 0.029, "step": 47880 }, { "grad_norm": 0.28162115812301636, "learning_rate": 5.759463873074717e-05, "loss": 0.0295, "step": 47890 }, { "grad_norm": 0.4389714002609253, "learning_rate": 5.757829547007504e-05, "loss": 0.0345, "step": 47900 }, { "grad_norm": 0.25146326422691345, "learning_rate": 5.756195138065189e-05, "loss": 0.0207, "step": 47910 }, { "grad_norm": 0.35016778111457825, "learning_rate": 5.754560646426511e-05, "loss": 0.0341, "step": 47920 }, { "grad_norm": 0.3263314664363861, "learning_rate": 5.752926072270216e-05, "loss": 0.0261, "step": 47930 }, { "grad_norm": 0.5469518899917603, "learning_rate": 5.7512914157750563e-05, "loss": 0.0245, "step": 47940 }, { "grad_norm": 0.5279032588005066, "learning_rate": 5.749656677119798e-05, "loss": 0.0243, "step": 47950 }, { "grad_norm": 0.3465985655784607, "learning_rate": 5.7480218564832125e-05, "loss": 0.0311, "step": 47960 }, { "grad_norm": 0.45032593607902527, "learning_rate": 5.746386954044082e-05, "loss": 0.0297, "step": 47970 }, { "grad_norm": 0.30733051896095276, "learning_rate": 5.744751969981195e-05, "loss": 0.0218, "step": 47980 }, { "grad_norm": 0.36903053522109985, "learning_rate": 5.7431169044733526e-05, "loss": 0.0331, "step": 47990 }, { "grad_norm": 0.2600531578063965, "learning_rate": 5.741481757699364e-05, "loss": 0.0233, "step": 48000 }, { "grad_norm": 0.49359044432640076, "learning_rate": 5.7398465298380434e-05, "loss": 0.0301, "step": 48010 }, { "grad_norm": 0.6576289534568787, "learning_rate": 5.7382112210682193e-05, "loss": 0.0388, "step": 48020 }, { "grad_norm": 0.8464311361312866, "learning_rate": 5.7365758315687266e-05, "loss": 0.0365, "step": 48030 }, { "grad_norm": 0.7317018508911133, "learning_rate": 5.734940361518407e-05, "loss": 0.024, "step": 48040 }, { "grad_norm": 1.816085696220398, "learning_rate": 5.733304811096116e-05, "loss": 0.0271, "step": 48050 }, { "grad_norm": 0.5104748010635376, "learning_rate": 5.731669180480713e-05, "loss": 0.0282, "step": 48060 }, { "grad_norm": 0.35353368520736694, "learning_rate": 5.730033469851067e-05, "loss": 0.0367, "step": 48070 }, { "grad_norm": 0.29453182220458984, "learning_rate": 5.7283976793860607e-05, "loss": 0.0286, "step": 48080 }, { "grad_norm": 0.2615646719932556, "learning_rate": 5.726761809264577e-05, "loss": 0.0256, "step": 48090 }, { "grad_norm": 0.2637993395328522, "learning_rate": 5.7251258596655155e-05, "loss": 0.0282, "step": 48100 }, { "grad_norm": 0.28480401635169983, "learning_rate": 5.72348983076778e-05, "loss": 0.0274, "step": 48110 }, { "grad_norm": 0.3646724224090576, "learning_rate": 5.7218537227502854e-05, "loss": 0.0313, "step": 48120 }, { "grad_norm": 0.3044414222240448, "learning_rate": 5.720217535791951e-05, "loss": 0.0224, "step": 48130 }, { "grad_norm": 0.3337861895561218, "learning_rate": 5.718581270071711e-05, "loss": 0.0255, "step": 48140 }, { "grad_norm": 0.21306388080120087, "learning_rate": 5.716944925768505e-05, "loss": 0.0276, "step": 48150 }, { "grad_norm": 0.5314265489578247, "learning_rate": 5.7153085030612786e-05, "loss": 0.0325, "step": 48160 }, { "grad_norm": 0.28849339485168457, "learning_rate": 5.713672002128989e-05, "loss": 0.0269, "step": 48170 }, { "grad_norm": 0.3252162039279938, "learning_rate": 5.712035423150602e-05, "loss": 0.026, "step": 48180 }, { "grad_norm": 0.3073900640010834, "learning_rate": 5.710398766305094e-05, "loss": 0.0288, "step": 48190 }, { "grad_norm": 0.37995442748069763, "learning_rate": 5.708762031771442e-05, "loss": 0.0284, "step": 48200 }, { "grad_norm": 0.2972094416618347, "learning_rate": 5.707125219728642e-05, "loss": 0.0244, "step": 48210 }, { "grad_norm": 0.5256661176681519, "learning_rate": 5.7054883303556905e-05, "loss": 0.0247, "step": 48220 }, { "grad_norm": 0.28028157353401184, "learning_rate": 5.703851363831595e-05, "loss": 0.0317, "step": 48230 }, { "grad_norm": 0.3334098756313324, "learning_rate": 5.702214320335374e-05, "loss": 0.0356, "step": 48240 }, { "grad_norm": 0.24417045712471008, "learning_rate": 5.7005772000460514e-05, "loss": 0.0305, "step": 48250 }, { "grad_norm": 0.26551833748817444, "learning_rate": 5.698940003142656e-05, "loss": 0.0238, "step": 48260 }, { "grad_norm": 0.2775651812553406, "learning_rate": 5.697302729804236e-05, "loss": 0.0242, "step": 48270 }, { "grad_norm": 0.4209510385990143, "learning_rate": 5.695665380209837e-05, "loss": 0.0248, "step": 48280 }, { "grad_norm": 0.27064815163612366, "learning_rate": 5.6940279545385165e-05, "loss": 0.0236, "step": 48290 }, { "grad_norm": 0.3169494867324829, "learning_rate": 5.692390452969344e-05, "loss": 0.0236, "step": 48300 }, { "grad_norm": 0.2278173416852951, "learning_rate": 5.690752875681392e-05, "loss": 0.0265, "step": 48310 }, { "grad_norm": 0.385976105928421, "learning_rate": 5.6891152228537435e-05, "loss": 0.0317, "step": 48320 }, { "grad_norm": 0.3190283179283142, "learning_rate": 5.687477494665492e-05, "loss": 0.0244, "step": 48330 }, { "grad_norm": 0.25226831436157227, "learning_rate": 5.685839691295733e-05, "loss": 0.0275, "step": 48340 }, { "grad_norm": 0.33672377467155457, "learning_rate": 5.6842018129235786e-05, "loss": 0.0281, "step": 48350 }, { "grad_norm": 0.32927432656288147, "learning_rate": 5.6825638597281404e-05, "loss": 0.0209, "step": 48360 }, { "grad_norm": 0.33698180317878723, "learning_rate": 5.680925831888546e-05, "loss": 0.0267, "step": 48370 }, { "grad_norm": 0.29266729950904846, "learning_rate": 5.6792877295839274e-05, "loss": 0.0271, "step": 48380 }, { "grad_norm": 0.29220351576805115, "learning_rate": 5.6776495529934224e-05, "loss": 0.0293, "step": 48390 }, { "grad_norm": 0.23947273194789886, "learning_rate": 5.6760113022961824e-05, "loss": 0.0228, "step": 48400 }, { "grad_norm": 0.24748477339744568, "learning_rate": 5.6743729776713617e-05, "loss": 0.0214, "step": 48410 }, { "grad_norm": 0.5952112674713135, "learning_rate": 5.672734579298126e-05, "loss": 0.0274, "step": 48420 }, { "grad_norm": 0.35612642765045166, "learning_rate": 5.671096107355649e-05, "loss": 0.0265, "step": 48430 }, { "grad_norm": 0.3727259933948517, "learning_rate": 5.66945756202311e-05, "loss": 0.0215, "step": 48440 }, { "grad_norm": 0.5643673539161682, "learning_rate": 5.667818943479699e-05, "loss": 0.0262, "step": 48450 }, { "grad_norm": 0.2691890299320221, "learning_rate": 5.666180251904612e-05, "loss": 0.0417, "step": 48460 }, { "grad_norm": 0.37802112102508545, "learning_rate": 5.6645414874770555e-05, "loss": 0.0289, "step": 48470 }, { "grad_norm": 0.45874810218811035, "learning_rate": 5.66290265037624e-05, "loss": 0.0297, "step": 48480 }, { "grad_norm": 0.4239053428173065, "learning_rate": 5.661263740781386e-05, "loss": 0.0331, "step": 48490 }, { "grad_norm": 0.4823640286922455, "learning_rate": 5.6596247588717254e-05, "loss": 0.0231, "step": 48500 }, { "grad_norm": 0.65486079454422, "learning_rate": 5.6579857048264926e-05, "loss": 0.0288, "step": 48510 }, { "grad_norm": 0.48620304465293884, "learning_rate": 5.6563465788249314e-05, "loss": 0.0247, "step": 48520 }, { "grad_norm": 0.6737030744552612, "learning_rate": 5.6547073810462956e-05, "loss": 0.0275, "step": 48530 }, { "grad_norm": 0.5955802798271179, "learning_rate": 5.653068111669846e-05, "loss": 0.0268, "step": 48540 }, { "grad_norm": 0.4359583556652069, "learning_rate": 5.651428770874848e-05, "loss": 0.0363, "step": 48550 }, { "grad_norm": 0.48106086254119873, "learning_rate": 5.64978935884058e-05, "loss": 0.0318, "step": 48560 }, { "grad_norm": 0.3505423963069916, "learning_rate": 5.6481498757463244e-05, "loss": 0.0408, "step": 48570 }, { "grad_norm": 0.3809976279735565, "learning_rate": 5.646510321771373e-05, "loss": 0.0367, "step": 48580 }, { "grad_norm": 0.39682504534721375, "learning_rate": 5.644870697095024e-05, "loss": 0.0273, "step": 48590 }, { "grad_norm": 0.3113829493522644, "learning_rate": 5.643231001896586e-05, "loss": 0.0246, "step": 48600 }, { "grad_norm": 0.2750124931335449, "learning_rate": 5.6415912363553726e-05, "loss": 0.0556, "step": 48610 }, { "grad_norm": 0.3219100832939148, "learning_rate": 5.639951400650706e-05, "loss": 0.024, "step": 48620 }, { "grad_norm": 0.3772505521774292, "learning_rate": 5.6383114949619165e-05, "loss": 0.0284, "step": 48630 }, { "grad_norm": 0.4814695119857788, "learning_rate": 5.636671519468342e-05, "loss": 0.027, "step": 48640 }, { "grad_norm": 0.48426559567451477, "learning_rate": 5.635031474349327e-05, "loss": 0.0295, "step": 48650 }, { "grad_norm": 0.3735158145427704, "learning_rate": 5.6333913597842246e-05, "loss": 0.0208, "step": 48660 }, { "grad_norm": 0.2641488015651703, "learning_rate": 5.6317511759523955e-05, "loss": 0.0291, "step": 48670 }, { "grad_norm": 0.316197007894516, "learning_rate": 5.630110923033207e-05, "loss": 0.0283, "step": 48680 }, { "grad_norm": 0.31686556339263916, "learning_rate": 5.628470601206036e-05, "loss": 0.0308, "step": 48690 }, { "grad_norm": 0.37333154678344727, "learning_rate": 5.626830210650266e-05, "loss": 0.0311, "step": 48700 }, { "grad_norm": 0.2211167961359024, "learning_rate": 5.625189751545285e-05, "loss": 0.0294, "step": 48710 }, { "grad_norm": 0.3273548185825348, "learning_rate": 5.6235492240704936e-05, "loss": 0.0279, "step": 48720 }, { "grad_norm": 0.4037574827671051, "learning_rate": 5.621908628405296e-05, "loss": 0.0281, "step": 48730 }, { "grad_norm": 0.30240321159362793, "learning_rate": 5.620267964729106e-05, "loss": 0.0278, "step": 48740 }, { "grad_norm": 0.3345634639263153, "learning_rate": 5.618627233221344e-05, "loss": 0.0223, "step": 48750 }, { "grad_norm": 0.3742862343788147, "learning_rate": 5.616986434061438e-05, "loss": 0.0312, "step": 48760 }, { "grad_norm": 0.49688708782196045, "learning_rate": 5.615345567428822e-05, "loss": 0.0238, "step": 48770 }, { "grad_norm": 0.3231690227985382, "learning_rate": 5.613704633502941e-05, "loss": 0.0244, "step": 48780 }, { "grad_norm": 0.3108839690685272, "learning_rate": 5.612063632463245e-05, "loss": 0.0209, "step": 48790 }, { "grad_norm": 0.4367067813873291, "learning_rate": 5.610422564489188e-05, "loss": 0.0336, "step": 48800 }, { "grad_norm": 0.40129828453063965, "learning_rate": 5.608781429760239e-05, "loss": 0.0326, "step": 48810 }, { "grad_norm": 0.28126055002212524, "learning_rate": 5.607140228455866e-05, "loss": 0.0351, "step": 48820 }, { "grad_norm": 0.2369423359632492, "learning_rate": 5.605498960755553e-05, "loss": 0.0344, "step": 48830 }, { "grad_norm": 0.32991084456443787, "learning_rate": 5.603857626838782e-05, "loss": 0.03, "step": 48840 }, { "grad_norm": 0.31402158737182617, "learning_rate": 5.60221622688505e-05, "loss": 0.0266, "step": 48850 }, { "grad_norm": 0.3245849013328552, "learning_rate": 5.6005747610738565e-05, "loss": 0.026, "step": 48860 }, { "grad_norm": 0.37210068106651306, "learning_rate": 5.59893322958471e-05, "loss": 0.0274, "step": 48870 }, { "grad_norm": 0.37837496399879456, "learning_rate": 5.5972916325971256e-05, "loss": 0.0295, "step": 48880 }, { "grad_norm": 0.3456900417804718, "learning_rate": 5.595649970290628e-05, "loss": 0.0301, "step": 48890 }, { "grad_norm": 0.4260154962539673, "learning_rate": 5.5940082428447426e-05, "loss": 0.0343, "step": 48900 }, { "grad_norm": 0.3559516966342926, "learning_rate": 5.592366450439012e-05, "loss": 0.039, "step": 48910 }, { "grad_norm": 0.3472227156162262, "learning_rate": 5.590724593252975e-05, "loss": 0.0342, "step": 48920 }, { "grad_norm": 0.4103039503097534, "learning_rate": 5.589082671466184e-05, "loss": 0.0241, "step": 48930 }, { "grad_norm": 0.424938827753067, "learning_rate": 5.587440685258199e-05, "loss": 0.0257, "step": 48940 }, { "grad_norm": 0.5050029754638672, "learning_rate": 5.585798634808583e-05, "loss": 0.0371, "step": 48950 }, { "grad_norm": 0.3685312867164612, "learning_rate": 5.584156520296909e-05, "loss": 0.0311, "step": 48960 }, { "grad_norm": 0.4150295555591583, "learning_rate": 5.582514341902757e-05, "loss": 0.0275, "step": 48970 }, { "grad_norm": 0.35215088725090027, "learning_rate": 5.580872099805713e-05, "loss": 0.0281, "step": 48980 }, { "grad_norm": 0.30763566493988037, "learning_rate": 5.5792297941853674e-05, "loss": 0.0251, "step": 48990 }, { "grad_norm": 0.3636856973171234, "learning_rate": 5.5775874252213247e-05, "loss": 0.0272, "step": 49000 }, { "grad_norm": 0.40577808022499084, "learning_rate": 5.575944993093189e-05, "loss": 0.0367, "step": 49010 }, { "grad_norm": 0.25833794474601746, "learning_rate": 5.574302497980574e-05, "loss": 0.0237, "step": 49020 }, { "grad_norm": 0.4066746234893799, "learning_rate": 5.5726599400631005e-05, "loss": 0.0263, "step": 49030 }, { "grad_norm": 0.35070401430130005, "learning_rate": 5.571017319520399e-05, "loss": 0.0245, "step": 49040 }, { "grad_norm": 0.29725682735443115, "learning_rate": 5.5693746365320984e-05, "loss": 0.0306, "step": 49050 }, { "grad_norm": 0.3174603283405304, "learning_rate": 5.567731891277848e-05, "loss": 0.0305, "step": 49060 }, { "grad_norm": 0.287069171667099, "learning_rate": 5.566089083937288e-05, "loss": 0.0312, "step": 49070 }, { "grad_norm": 0.21735170483589172, "learning_rate": 5.564446214690079e-05, "loss": 0.0304, "step": 49080 }, { "grad_norm": 0.2946692407131195, "learning_rate": 5.56280328371588e-05, "loss": 0.0267, "step": 49090 }, { "grad_norm": 0.35330212116241455, "learning_rate": 5.5611602911943596e-05, "loss": 0.0314, "step": 49100 }, { "grad_norm": 0.6530991196632385, "learning_rate": 5.559517237305194e-05, "loss": 0.0348, "step": 49110 }, { "grad_norm": 0.3566119372844696, "learning_rate": 5.557874122228064e-05, "loss": 0.0336, "step": 49120 }, { "grad_norm": 0.3455730676651001, "learning_rate": 5.55623094614266e-05, "loss": 0.026, "step": 49130 }, { "grad_norm": 0.36253640055656433, "learning_rate": 5.5545877092286743e-05, "loss": 0.0245, "step": 49140 }, { "grad_norm": 0.31978553533554077, "learning_rate": 5.552944411665811e-05, "loss": 0.0256, "step": 49150 }, { "grad_norm": 0.4385864734649658, "learning_rate": 5.5513010536337795e-05, "loss": 0.0303, "step": 49160 }, { "grad_norm": 0.4211294651031494, "learning_rate": 5.5496576353122934e-05, "loss": 0.0297, "step": 49170 }, { "grad_norm": 0.34190815687179565, "learning_rate": 5.548014156881074e-05, "loss": 0.0288, "step": 49180 }, { "grad_norm": 0.49341443181037903, "learning_rate": 5.5463706185198494e-05, "loss": 0.0241, "step": 49190 }, { "grad_norm": 0.30068090558052063, "learning_rate": 5.544727020408358e-05, "loss": 0.0239, "step": 49200 }, { "grad_norm": 0.2762329876422882, "learning_rate": 5.5430833627263366e-05, "loss": 0.026, "step": 49210 }, { "grad_norm": 0.48208075761795044, "learning_rate": 5.5414396456535354e-05, "loss": 0.0275, "step": 49220 }, { "grad_norm": 0.4407501518726349, "learning_rate": 5.539795869369709e-05, "loss": 0.0254, "step": 49230 }, { "grad_norm": 0.32549968361854553, "learning_rate": 5.538152034054618e-05, "loss": 0.0238, "step": 49240 }, { "grad_norm": 0.3418225944042206, "learning_rate": 5.536508139888028e-05, "loss": 0.0288, "step": 49250 }, { "grad_norm": 0.3253632187843323, "learning_rate": 5.534864187049716e-05, "loss": 0.0217, "step": 49260 }, { "grad_norm": 0.2816062271595001, "learning_rate": 5.5332201757194615e-05, "loss": 0.0249, "step": 49270 }, { "grad_norm": 0.5573532581329346, "learning_rate": 5.531576106077049e-05, "loss": 0.0224, "step": 49280 }, { "grad_norm": 0.2720547318458557, "learning_rate": 5.529931978302272e-05, "loss": 0.0264, "step": 49290 }, { "grad_norm": 0.2346397042274475, "learning_rate": 5.528287792574932e-05, "loss": 0.0244, "step": 49300 }, { "grad_norm": 0.2883451282978058, "learning_rate": 5.5266435490748324e-05, "loss": 0.0216, "step": 49310 }, { "grad_norm": 0.33495229482650757, "learning_rate": 5.524999247981787e-05, "loss": 0.0227, "step": 49320 }, { "grad_norm": 0.2963239848613739, "learning_rate": 5.523354889475613e-05, "loss": 0.0206, "step": 49330 }, { "grad_norm": 0.4423016309738159, "learning_rate": 5.521710473736134e-05, "loss": 0.03, "step": 49340 }, { "grad_norm": 0.20664887130260468, "learning_rate": 5.5200660009431826e-05, "loss": 0.0265, "step": 49350 }, { "grad_norm": 0.29076963663101196, "learning_rate": 5.518421471276596e-05, "loss": 0.0299, "step": 49360 }, { "grad_norm": 0.2603667974472046, "learning_rate": 5.5167768849162174e-05, "loss": 0.0262, "step": 49370 }, { "grad_norm": 0.3793829381465912, "learning_rate": 5.515132242041893e-05, "loss": 0.0345, "step": 49380 }, { "grad_norm": 0.3623183071613312, "learning_rate": 5.513487542833483e-05, "loss": 0.0359, "step": 49390 }, { "grad_norm": 0.24663184583187103, "learning_rate": 5.5118427874708466e-05, "loss": 0.0244, "step": 49400 }, { "grad_norm": 0.2774827778339386, "learning_rate": 5.510197976133853e-05, "loss": 0.0215, "step": 49410 }, { "grad_norm": 0.4123910665512085, "learning_rate": 5.508553109002376e-05, "loss": 0.0258, "step": 49420 }, { "grad_norm": 0.31674519181251526, "learning_rate": 5.5069081862562957e-05, "loss": 0.0225, "step": 49430 }, { "grad_norm": 0.5438292026519775, "learning_rate": 5.5052632080754965e-05, "loss": 0.0292, "step": 49440 }, { "grad_norm": 0.29434487223625183, "learning_rate": 5.503618174639874e-05, "loss": 0.0221, "step": 49450 }, { "grad_norm": 0.31916937232017517, "learning_rate": 5.501973086129325e-05, "loss": 0.0177, "step": 49460 }, { "grad_norm": 0.418999046087265, "learning_rate": 5.500327942723753e-05, "loss": 0.0229, "step": 49470 }, { "grad_norm": 0.47412875294685364, "learning_rate": 5.498682744603071e-05, "loss": 0.0225, "step": 49480 }, { "grad_norm": 0.25896772742271423, "learning_rate": 5.497037491947194e-05, "loss": 0.0245, "step": 49490 }, { "grad_norm": 0.2696077823638916, "learning_rate": 5.4953921849360424e-05, "loss": 0.0304, "step": 49500 }, { "grad_norm": 0.4153038561344147, "learning_rate": 5.493746823749547e-05, "loss": 0.0345, "step": 49510 }, { "grad_norm": 0.3498018980026245, "learning_rate": 5.49210140856764e-05, "loss": 0.0279, "step": 49520 }, { "grad_norm": 0.29369235038757324, "learning_rate": 5.4904559395702647e-05, "loss": 0.0227, "step": 49530 }, { "grad_norm": 0.2948577404022217, "learning_rate": 5.488810416937364e-05, "loss": 0.024, "step": 49540 }, { "grad_norm": 0.3332669734954834, "learning_rate": 5.4871648408488926e-05, "loss": 0.0274, "step": 49550 }, { "grad_norm": 0.3599075973033905, "learning_rate": 5.485519211484807e-05, "loss": 0.0233, "step": 49560 }, { "grad_norm": 0.29586976766586304, "learning_rate": 5.4838735290250676e-05, "loss": 0.0289, "step": 49570 }, { "grad_norm": 0.37934261560440063, "learning_rate": 5.482227793649649e-05, "loss": 0.0265, "step": 49580 }, { "grad_norm": 0.4154565930366516, "learning_rate": 5.480582005538524e-05, "loss": 0.0315, "step": 49590 }, { "grad_norm": 1.021849513053894, "learning_rate": 5.478936164871671e-05, "loss": 0.0351, "step": 49600 }, { "grad_norm": 0.4577310383319855, "learning_rate": 5.477290271829081e-05, "loss": 0.0432, "step": 49610 }, { "grad_norm": 0.260276734828949, "learning_rate": 5.475644326590744e-05, "loss": 0.029, "step": 49620 }, { "grad_norm": 0.6396231651306152, "learning_rate": 5.473998329336658e-05, "loss": 0.0292, "step": 49630 }, { "grad_norm": 0.7239484786987305, "learning_rate": 5.4723522802468286e-05, "loss": 0.0289, "step": 49640 }, { "grad_norm": 0.3986476957798004, "learning_rate": 5.470706179501264e-05, "loss": 0.0269, "step": 49650 }, { "grad_norm": 0.2927871346473694, "learning_rate": 5.4690600272799776e-05, "loss": 0.0247, "step": 49660 }, { "grad_norm": 0.6034978032112122, "learning_rate": 5.467413823762993e-05, "loss": 0.0279, "step": 49670 }, { "grad_norm": 0.4362974166870117, "learning_rate": 5.465767569130335e-05, "loss": 0.0308, "step": 49680 }, { "grad_norm": 0.3885929584503174, "learning_rate": 5.464121263562036e-05, "loss": 0.0264, "step": 49690 }, { "grad_norm": 0.6832392811775208, "learning_rate": 5.4624749072381343e-05, "loss": 0.0239, "step": 49700 }, { "grad_norm": 0.2996215224266052, "learning_rate": 5.460828500338672e-05, "loss": 0.0243, "step": 49710 }, { "grad_norm": 0.5431159734725952, "learning_rate": 5.459182043043698e-05, "loss": 0.0331, "step": 49720 }, { "grad_norm": 0.22261148691177368, "learning_rate": 5.457535535533265e-05, "loss": 0.0253, "step": 49730 }, { "grad_norm": 0.37384819984436035, "learning_rate": 5.4558889779874334e-05, "loss": 0.0237, "step": 49740 }, { "grad_norm": 0.28960883617401123, "learning_rate": 5.454242370586269e-05, "loss": 0.0294, "step": 49750 }, { "grad_norm": 0.2328379899263382, "learning_rate": 5.452595713509843e-05, "loss": 0.0265, "step": 49760 }, { "grad_norm": 0.3937903344631195, "learning_rate": 5.4509490069382274e-05, "loss": 0.0258, "step": 49770 }, { "grad_norm": 0.22865669429302216, "learning_rate": 5.449302251051509e-05, "loss": 0.0248, "step": 49780 }, { "grad_norm": 0.6629926562309265, "learning_rate": 5.44765544602977e-05, "loss": 0.0225, "step": 49790 }, { "grad_norm": 0.32815852761268616, "learning_rate": 5.446008592053107e-05, "loss": 0.0298, "step": 49800 }, { "grad_norm": 0.2796162962913513, "learning_rate": 5.4443616893016135e-05, "loss": 0.0394, "step": 49810 }, { "grad_norm": 0.27457043528556824, "learning_rate": 5.4427147379553934e-05, "loss": 0.0276, "step": 49820 }, { "grad_norm": 0.3089441955089569, "learning_rate": 5.441067738194556e-05, "loss": 0.0237, "step": 49830 }, { "grad_norm": 0.33483031392097473, "learning_rate": 5.439420690199214e-05, "loss": 0.0336, "step": 49840 }, { "grad_norm": 0.2582971751689911, "learning_rate": 5.437773594149484e-05, "loss": 0.0278, "step": 49850 }, { "grad_norm": 0.2581215500831604, "learning_rate": 5.436126450225495e-05, "loss": 0.0226, "step": 49860 }, { "grad_norm": 0.4120672345161438, "learning_rate": 5.434479258607371e-05, "loss": 0.0229, "step": 49870 }, { "grad_norm": 0.4536017179489136, "learning_rate": 5.43283201947525e-05, "loss": 0.0284, "step": 49880 }, { "grad_norm": 0.32732120156288147, "learning_rate": 5.431184733009268e-05, "loss": 0.0386, "step": 49890 }, { "grad_norm": 0.3388591408729553, "learning_rate": 5.4295373993895736e-05, "loss": 0.0208, "step": 49900 }, { "grad_norm": 0.298916220664978, "learning_rate": 5.4278900187963157e-05, "loss": 0.0203, "step": 49910 }, { "grad_norm": 0.38596010208129883, "learning_rate": 5.426242591409646e-05, "loss": 0.0295, "step": 49920 }, { "grad_norm": 0.5779551863670349, "learning_rate": 5.42459511740973e-05, "loss": 0.0268, "step": 49930 }, { "grad_norm": 0.3869728744029999, "learning_rate": 5.422947596976729e-05, "loss": 0.0258, "step": 49940 }, { "grad_norm": 0.37947922945022583, "learning_rate": 5.4213000302908134e-05, "loss": 0.0311, "step": 49950 }, { "grad_norm": 0.3786855638027191, "learning_rate": 5.419652417532162e-05, "loss": 0.024, "step": 49960 }, { "grad_norm": 0.3014862835407257, "learning_rate": 5.4180047588809534e-05, "loss": 0.0268, "step": 49970 }, { "grad_norm": 0.45988503098487854, "learning_rate": 5.4163570545173704e-05, "loss": 0.0254, "step": 49980 }, { "grad_norm": 0.3437362611293793, "learning_rate": 5.414709304621608e-05, "loss": 0.0258, "step": 49990 }, { "grad_norm": 0.3104787766933441, "learning_rate": 5.413061509373858e-05, "loss": 0.0293, "step": 50000 }, { "grad_norm": 0.2785079777240753, "learning_rate": 5.411413668954324e-05, "loss": 0.028, "step": 50010 }, { "grad_norm": 0.2382562756538391, "learning_rate": 5.4097657835432083e-05, "loss": 0.0231, "step": 50020 }, { "grad_norm": 0.39866915345191956, "learning_rate": 5.408117853320723e-05, "loss": 0.0224, "step": 50030 }, { "grad_norm": 0.42016366124153137, "learning_rate": 5.406469878467084e-05, "loss": 0.0322, "step": 50040 }, { "grad_norm": 0.37699657678604126, "learning_rate": 5.404821859162509e-05, "loss": 0.0321, "step": 50050 }, { "grad_norm": 0.30610501766204834, "learning_rate": 5.403173795587225e-05, "loss": 0.0271, "step": 50060 }, { "grad_norm": 0.3588050305843353, "learning_rate": 5.4015256879214606e-05, "loss": 0.023, "step": 50070 }, { "grad_norm": 0.21381215751171112, "learning_rate": 5.39987753634545e-05, "loss": 0.023, "step": 50080 }, { "grad_norm": 0.24647514522075653, "learning_rate": 5.3982293410394335e-05, "loss": 0.0204, "step": 50090 }, { "grad_norm": 0.35366371273994446, "learning_rate": 5.396581102183655e-05, "loss": 0.0266, "step": 50100 }, { "grad_norm": 0.3289124369621277, "learning_rate": 5.3949328199583634e-05, "loss": 0.0218, "step": 50110 }, { "grad_norm": 0.3896147310733795, "learning_rate": 5.393284494543813e-05, "loss": 0.0272, "step": 50120 }, { "grad_norm": 0.2918842136859894, "learning_rate": 5.391636126120262e-05, "loss": 0.0275, "step": 50130 }, { "grad_norm": 1.0038567781448364, "learning_rate": 5.389987714867971e-05, "loss": 0.0249, "step": 50140 }, { "grad_norm": 0.3536776900291443, "learning_rate": 5.3883392609672114e-05, "loss": 0.0202, "step": 50150 }, { "grad_norm": 0.580048680305481, "learning_rate": 5.3866907645982545e-05, "loss": 0.0242, "step": 50160 }, { "grad_norm": 0.4206967353820801, "learning_rate": 5.385042225941375e-05, "loss": 0.0268, "step": 50170 }, { "grad_norm": 0.41950634121894836, "learning_rate": 5.3833936451768585e-05, "loss": 0.0346, "step": 50180 }, { "grad_norm": 0.45851293206214905, "learning_rate": 5.3817450224849884e-05, "loss": 0.0309, "step": 50190 }, { "grad_norm": 0.34236618876457214, "learning_rate": 5.380096358046056e-05, "loss": 0.0367, "step": 50200 }, { "grad_norm": 0.3125709891319275, "learning_rate": 5.378447652040359e-05, "loss": 0.0239, "step": 50210 }, { "grad_norm": 0.4447803199291229, "learning_rate": 5.376798904648195e-05, "loss": 0.0239, "step": 50220 }, { "grad_norm": 0.27416279911994934, "learning_rate": 5.375150116049869e-05, "loss": 0.024, "step": 50230 }, { "grad_norm": 0.3362833261489868, "learning_rate": 5.373501286425691e-05, "loss": 0.0276, "step": 50240 }, { "grad_norm": 0.31606167554855347, "learning_rate": 5.3718524159559726e-05, "loss": 0.0262, "step": 50250 }, { "grad_norm": 0.43930351734161377, "learning_rate": 5.370203504821034e-05, "loss": 0.0349, "step": 50260 }, { "grad_norm": 0.25958824157714844, "learning_rate": 5.368554553201196e-05, "loss": 0.0342, "step": 50270 }, { "grad_norm": 0.5440941452980042, "learning_rate": 5.3669055612767874e-05, "loss": 0.0384, "step": 50280 }, { "grad_norm": 0.37728452682495117, "learning_rate": 5.3652565292281376e-05, "loss": 0.0227, "step": 50290 }, { "grad_norm": 0.4424213469028473, "learning_rate": 5.3636074572355824e-05, "loss": 0.0226, "step": 50300 }, { "grad_norm": 0.3306540250778198, "learning_rate": 5.361958345479463e-05, "loss": 0.0228, "step": 50310 }, { "grad_norm": 0.2800436317920685, "learning_rate": 5.360309194140123e-05, "loss": 0.0283, "step": 50320 }, { "grad_norm": 0.47668978571891785, "learning_rate": 5.3586600033979106e-05, "loss": 0.0263, "step": 50330 }, { "grad_norm": 0.28502705693244934, "learning_rate": 5.357010773433181e-05, "loss": 0.0212, "step": 50340 }, { "grad_norm": 0.3153529465198517, "learning_rate": 5.3553615044262895e-05, "loss": 0.0234, "step": 50350 }, { "grad_norm": 0.5198949575424194, "learning_rate": 5.3537121965575964e-05, "loss": 0.0266, "step": 50360 }, { "grad_norm": 0.3060752749443054, "learning_rate": 5.352062850007471e-05, "loss": 0.0263, "step": 50370 }, { "grad_norm": 0.35296252369880676, "learning_rate": 5.350413464956282e-05, "loss": 0.0292, "step": 50380 }, { "grad_norm": 0.39482662081718445, "learning_rate": 5.348764041584403e-05, "loss": 0.0302, "step": 50390 }, { "grad_norm": 0.31892016530036926, "learning_rate": 5.347114580072212e-05, "loss": 0.0272, "step": 50400 }, { "grad_norm": 0.3715527057647705, "learning_rate": 5.345465080600093e-05, "loss": 0.0249, "step": 50410 }, { "grad_norm": 0.3788905143737793, "learning_rate": 5.3438155433484326e-05, "loss": 0.022, "step": 50420 }, { "grad_norm": 1.18340003490448, "learning_rate": 5.3421659684976197e-05, "loss": 0.0241, "step": 50430 }, { "grad_norm": 0.47053226828575134, "learning_rate": 5.340516356228052e-05, "loss": 0.0316, "step": 50440 }, { "grad_norm": 0.5762728452682495, "learning_rate": 5.338866706720128e-05, "loss": 0.0296, "step": 50450 }, { "grad_norm": 0.394687294960022, "learning_rate": 5.337217020154249e-05, "loss": 0.0267, "step": 50460 }, { "grad_norm": 0.3394773602485657, "learning_rate": 5.335567296710825e-05, "loss": 0.0296, "step": 50470 }, { "grad_norm": 0.3485845625400543, "learning_rate": 5.333917536570265e-05, "loss": 0.0282, "step": 50480 }, { "grad_norm": 0.5260578989982605, "learning_rate": 5.332267739912986e-05, "loss": 0.0251, "step": 50490 }, { "grad_norm": 0.2786434292793274, "learning_rate": 5.330617906919405e-05, "loss": 0.0348, "step": 50500 }, { "grad_norm": 0.27491727471351624, "learning_rate": 5.328968037769949e-05, "loss": 0.0256, "step": 50510 }, { "grad_norm": 0.2684950530529022, "learning_rate": 5.327318132645042e-05, "loss": 0.0291, "step": 50520 }, { "grad_norm": 0.3335348665714264, "learning_rate": 5.325668191725116e-05, "loss": 0.0233, "step": 50530 }, { "grad_norm": 0.3498164117336273, "learning_rate": 5.324018215190606e-05, "loss": 0.0212, "step": 50540 }, { "grad_norm": 0.401980459690094, "learning_rate": 5.3223682032219515e-05, "loss": 0.0249, "step": 50550 }, { "grad_norm": 0.33003684878349304, "learning_rate": 5.320718155999595e-05, "loss": 0.0317, "step": 50560 }, { "grad_norm": 0.3461483418941498, "learning_rate": 5.3190680737039835e-05, "loss": 0.0349, "step": 50570 }, { "grad_norm": 0.21342000365257263, "learning_rate": 5.317417956515567e-05, "loss": 0.0236, "step": 50580 }, { "grad_norm": 0.5502384305000305, "learning_rate": 5.315767804614798e-05, "loss": 0.0318, "step": 50590 }, { "grad_norm": 0.39263463020324707, "learning_rate": 5.3141176181821395e-05, "loss": 0.0308, "step": 50600 }, { "grad_norm": 0.690116822719574, "learning_rate": 5.31246739739805e-05, "loss": 0.0258, "step": 50610 }, { "grad_norm": 0.31412073969841003, "learning_rate": 5.310817142442994e-05, "loss": 0.0337, "step": 50620 }, { "grad_norm": 0.48733437061309814, "learning_rate": 5.309166853497445e-05, "loss": 0.0289, "step": 50630 }, { "grad_norm": 0.23981356620788574, "learning_rate": 5.307516530741873e-05, "loss": 0.0226, "step": 50640 }, { "grad_norm": 0.3772335350513458, "learning_rate": 5.305866174356754e-05, "loss": 0.0306, "step": 50650 }, { "grad_norm": 0.4104599058628082, "learning_rate": 5.304215784522571e-05, "loss": 0.0228, "step": 50660 }, { "grad_norm": 0.5445385575294495, "learning_rate": 5.302565361419808e-05, "loss": 0.0353, "step": 50670 }, { "grad_norm": 0.26704978942871094, "learning_rate": 5.3009149052289507e-05, "loss": 0.0366, "step": 50680 }, { "grad_norm": 0.2268119901418686, "learning_rate": 5.299264416130493e-05, "loss": 0.0225, "step": 50690 }, { "grad_norm": 0.3336280584335327, "learning_rate": 5.297613894304928e-05, "loss": 0.0257, "step": 50700 }, { "grad_norm": 0.6312664747238159, "learning_rate": 5.2959633399327534e-05, "loss": 0.0258, "step": 50710 }, { "grad_norm": 0.3740352988243103, "learning_rate": 5.294312753194476e-05, "loss": 0.0306, "step": 50720 }, { "grad_norm": 0.34035807847976685, "learning_rate": 5.292662134270596e-05, "loss": 0.0293, "step": 50730 }, { "grad_norm": 0.2951470911502838, "learning_rate": 5.291011483341626e-05, "loss": 0.0325, "step": 50740 }, { "grad_norm": 0.48580408096313477, "learning_rate": 5.2893608005880767e-05, "loss": 0.0368, "step": 50750 }, { "grad_norm": 0.3655747175216675, "learning_rate": 5.287710086190467e-05, "loss": 0.0301, "step": 50760 }, { "grad_norm": 0.3234226703643799, "learning_rate": 5.286059340329314e-05, "loss": 0.0248, "step": 50770 }, { "grad_norm": 0.4440953731536865, "learning_rate": 5.284408563185141e-05, "loss": 0.0372, "step": 50780 }, { "grad_norm": 0.9651662707328796, "learning_rate": 5.282757754938476e-05, "loss": 0.0249, "step": 50790 }, { "grad_norm": 0.29189303517341614, "learning_rate": 5.281106915769849e-05, "loss": 0.028, "step": 50800 }, { "grad_norm": 0.2980719208717346, "learning_rate": 5.2794560458597897e-05, "loss": 0.0222, "step": 50810 }, { "grad_norm": 0.37169942259788513, "learning_rate": 5.27780514538884e-05, "loss": 0.0253, "step": 50820 }, { "grad_norm": 0.2317368984222412, "learning_rate": 5.2761542145375365e-05, "loss": 0.0295, "step": 50830 }, { "grad_norm": 0.3847435712814331, "learning_rate": 5.274503253486421e-05, "loss": 0.0291, "step": 50840 }, { "grad_norm": 0.4809771478176117, "learning_rate": 5.272852262416046e-05, "loss": 0.0359, "step": 50850 }, { "grad_norm": 0.23737213015556335, "learning_rate": 5.2712012415069555e-05, "loss": 0.0272, "step": 50860 }, { "grad_norm": 0.5338777303695679, "learning_rate": 5.269550190939705e-05, "loss": 0.0249, "step": 50870 }, { "grad_norm": 0.5287941098213196, "learning_rate": 5.267899110894852e-05, "loss": 0.0264, "step": 50880 }, { "grad_norm": 0.24727696180343628, "learning_rate": 5.266248001552955e-05, "loss": 0.0219, "step": 50890 }, { "grad_norm": 0.28060483932495117, "learning_rate": 5.264596863094575e-05, "loss": 0.0244, "step": 50900 }, { "grad_norm": 0.2913452088832855, "learning_rate": 5.262945695700282e-05, "loss": 0.0284, "step": 50910 }, { "grad_norm": 0.33009248971939087, "learning_rate": 5.261294499550643e-05, "loss": 0.0316, "step": 50920 }, { "grad_norm": 0.2955050468444824, "learning_rate": 5.25964327482623e-05, "loss": 0.0203, "step": 50930 }, { "grad_norm": 0.37353959679603577, "learning_rate": 5.257992021707617e-05, "loss": 0.0239, "step": 50940 }, { "grad_norm": 0.39512670040130615, "learning_rate": 5.256340740375387e-05, "loss": 0.0324, "step": 50950 }, { "grad_norm": 0.43793052434921265, "learning_rate": 5.254689431010117e-05, "loss": 0.0271, "step": 50960 }, { "grad_norm": 0.18707731366157532, "learning_rate": 5.253038093792395e-05, "loss": 0.0243, "step": 50970 }, { "grad_norm": 0.6353442668914795, "learning_rate": 5.251386728902806e-05, "loss": 0.0273, "step": 50980 }, { "grad_norm": 0.3294086158275604, "learning_rate": 5.2497353365219446e-05, "loss": 0.0223, "step": 50990 }, { "grad_norm": 0.379666805267334, "learning_rate": 5.2480839168304e-05, "loss": 0.0248, "step": 51000 }, { "grad_norm": 0.3550436198711395, "learning_rate": 5.246432470008773e-05, "loss": 0.0247, "step": 51010 }, { "grad_norm": 0.23041439056396484, "learning_rate": 5.2447809962376616e-05, "loss": 0.023, "step": 51020 }, { "grad_norm": 0.28111696243286133, "learning_rate": 5.243129495697666e-05, "loss": 0.023, "step": 51030 }, { "grad_norm": 0.24238063395023346, "learning_rate": 5.241477968569397e-05, "loss": 0.0261, "step": 51040 }, { "grad_norm": 0.46386590600013733, "learning_rate": 5.23982641503346e-05, "loss": 0.0216, "step": 51050 }, { "grad_norm": 0.41115573048591614, "learning_rate": 5.238174835270465e-05, "loss": 0.0268, "step": 51060 }, { "grad_norm": 0.2177087813615799, "learning_rate": 5.23652322946103e-05, "loss": 0.0198, "step": 51070 }, { "grad_norm": 0.45555171370506287, "learning_rate": 5.234871597785771e-05, "loss": 0.0239, "step": 51080 }, { "grad_norm": 0.38111311197280884, "learning_rate": 5.233219940425306e-05, "loss": 0.0257, "step": 51090 }, { "grad_norm": 0.4348794221878052, "learning_rate": 5.2315682575602576e-05, "loss": 0.0391, "step": 51100 }, { "grad_norm": 0.4931522011756897, "learning_rate": 5.229916549371254e-05, "loss": 0.02, "step": 51110 }, { "grad_norm": 0.35068291425704956, "learning_rate": 5.228264816038924e-05, "loss": 0.0244, "step": 51120 }, { "grad_norm": 0.3473489582538605, "learning_rate": 5.2266130577438935e-05, "loss": 0.0239, "step": 51130 }, { "grad_norm": 0.34840646386146545, "learning_rate": 5.224961274666802e-05, "loss": 0.0283, "step": 51140 }, { "grad_norm": 0.2739931046962738, "learning_rate": 5.223309466988283e-05, "loss": 0.0338, "step": 51150 }, { "grad_norm": 0.25220105051994324, "learning_rate": 5.221657634888976e-05, "loss": 0.0274, "step": 51160 }, { "grad_norm": 0.3065818250179291, "learning_rate": 5.2200057785495235e-05, "loss": 0.0333, "step": 51170 }, { "grad_norm": 0.39087000489234924, "learning_rate": 5.2183538981505706e-05, "loss": 0.0252, "step": 51180 }, { "grad_norm": 0.5201908946037292, "learning_rate": 5.216701993872762e-05, "loss": 0.0266, "step": 51190 }, { "grad_norm": 0.2919620871543884, "learning_rate": 5.215050065896752e-05, "loss": 0.0316, "step": 51200 }, { "grad_norm": 0.220989391207695, "learning_rate": 5.2133981144031875e-05, "loss": 0.0373, "step": 51210 }, { "grad_norm": 0.3094140589237213, "learning_rate": 5.211746139572727e-05, "loss": 0.0234, "step": 51220 }, { "grad_norm": 0.27222368121147156, "learning_rate": 5.210094141586026e-05, "loss": 0.0266, "step": 51230 }, { "grad_norm": 0.4199901819229126, "learning_rate": 5.2084421206237476e-05, "loss": 0.0274, "step": 51240 }, { "grad_norm": 0.30950284004211426, "learning_rate": 5.20679007686655e-05, "loss": 0.0323, "step": 51250 }, { "grad_norm": 0.1974448561668396, "learning_rate": 5.205138010495102e-05, "loss": 0.0253, "step": 51260 }, { "grad_norm": 0.35085541009902954, "learning_rate": 5.20348592169007e-05, "loss": 0.0267, "step": 51270 }, { "grad_norm": 0.5320694446563721, "learning_rate": 5.201833810632123e-05, "loss": 0.0251, "step": 51280 }, { "grad_norm": 0.34078794717788696, "learning_rate": 5.200181677501933e-05, "loss": 0.0218, "step": 51290 }, { "grad_norm": 0.35723593831062317, "learning_rate": 5.1985295224801777e-05, "loss": 0.0227, "step": 51300 }, { "grad_norm": 0.4518958032131195, "learning_rate": 5.196877345747531e-05, "loss": 0.0294, "step": 51310 }, { "grad_norm": 0.31144407391548157, "learning_rate": 5.195225147484674e-05, "loss": 0.0274, "step": 51320 }, { "grad_norm": 0.27707988023757935, "learning_rate": 5.193572927872288e-05, "loss": 0.0221, "step": 51330 }, { "grad_norm": 0.2943558990955353, "learning_rate": 5.191920687091059e-05, "loss": 0.0296, "step": 51340 }, { "grad_norm": 0.41156983375549316, "learning_rate": 5.1902684253216695e-05, "loss": 0.0317, "step": 51350 }, { "grad_norm": 0.3349514305591583, "learning_rate": 5.1886161427448134e-05, "loss": 0.0277, "step": 51360 }, { "grad_norm": 0.2566646635532379, "learning_rate": 5.186963839541179e-05, "loss": 0.0249, "step": 51370 }, { "grad_norm": 0.30159902572631836, "learning_rate": 5.185311515891459e-05, "loss": 0.0246, "step": 51380 }, { "grad_norm": 0.4417831599712372, "learning_rate": 5.1836591719763516e-05, "loss": 0.0314, "step": 51390 }, { "grad_norm": 0.30360984802246094, "learning_rate": 5.182006807976554e-05, "loss": 0.0248, "step": 51400 }, { "grad_norm": 0.1990003138780594, "learning_rate": 5.180354424072763e-05, "loss": 0.0199, "step": 51410 }, { "grad_norm": 0.42778798937797546, "learning_rate": 5.178702020445686e-05, "loss": 0.0316, "step": 51420 }, { "grad_norm": 0.36433014273643494, "learning_rate": 5.177049597276024e-05, "loss": 0.0226, "step": 51430 }, { "grad_norm": 0.45377328991889954, "learning_rate": 5.175397154744482e-05, "loss": 0.0275, "step": 51440 }, { "grad_norm": 0.36777836084365845, "learning_rate": 5.1737446930317736e-05, "loss": 0.0222, "step": 51450 }, { "grad_norm": 0.23878063261508942, "learning_rate": 5.172092212318604e-05, "loss": 0.0237, "step": 51460 }, { "grad_norm": 0.20568954944610596, "learning_rate": 5.170439712785692e-05, "loss": 0.0247, "step": 51470 }, { "grad_norm": 0.3987996280193329, "learning_rate": 5.1687871946137467e-05, "loss": 0.027, "step": 51480 }, { "grad_norm": 0.3683275878429413, "learning_rate": 5.16713465798349e-05, "loss": 0.0237, "step": 51490 }, { "grad_norm": 0.3407362401485443, "learning_rate": 5.165482103075637e-05, "loss": 0.0253, "step": 51500 }, { "grad_norm": 0.6062538027763367, "learning_rate": 5.163829530070909e-05, "loss": 0.0317, "step": 51510 }, { "grad_norm": 0.3728979229927063, "learning_rate": 5.162176939150032e-05, "loss": 0.0228, "step": 51520 }, { "grad_norm": 0.39501678943634033, "learning_rate": 5.160524330493728e-05, "loss": 0.0269, "step": 51530 }, { "grad_norm": 0.267396479845047, "learning_rate": 5.158871704282724e-05, "loss": 0.0217, "step": 51540 }, { "grad_norm": 0.2198103964328766, "learning_rate": 5.15721906069775e-05, "loss": 0.0298, "step": 51550 }, { "grad_norm": 0.3336361050605774, "learning_rate": 5.155566399919536e-05, "loss": 0.0228, "step": 51560 }, { "grad_norm": 0.4494572579860687, "learning_rate": 5.153913722128813e-05, "loss": 0.0353, "step": 51570 }, { "grad_norm": 0.37217381596565247, "learning_rate": 5.1522610275063176e-05, "loss": 0.0249, "step": 51580 }, { "grad_norm": 0.28991711139678955, "learning_rate": 5.150608316232786e-05, "loss": 0.0233, "step": 51590 }, { "grad_norm": 0.2560898959636688, "learning_rate": 5.148955588488955e-05, "loss": 0.0249, "step": 51600 }, { "grad_norm": 0.271054208278656, "learning_rate": 5.147302844455564e-05, "loss": 0.0258, "step": 51610 }, { "grad_norm": 0.37545084953308105, "learning_rate": 5.145650084313356e-05, "loss": 0.0278, "step": 51620 }, { "grad_norm": 0.40506303310394287, "learning_rate": 5.1439973082430734e-05, "loss": 0.0262, "step": 51630 }, { "grad_norm": 0.24904903769493103, "learning_rate": 5.142344516425461e-05, "loss": 0.0254, "step": 51640 }, { "grad_norm": 0.3971808850765228, "learning_rate": 5.1406917090412673e-05, "loss": 0.0348, "step": 51650 }, { "grad_norm": 0.46383097767829895, "learning_rate": 5.1390388862712404e-05, "loss": 0.0344, "step": 51660 }, { "grad_norm": 0.31967979669570923, "learning_rate": 5.137386048296129e-05, "loss": 0.0236, "step": 51670 }, { "grad_norm": 0.29969123005867004, "learning_rate": 5.135733195296686e-05, "loss": 0.0288, "step": 51680 }, { "grad_norm": 0.7603276968002319, "learning_rate": 5.134080327453664e-05, "loss": 0.0311, "step": 51690 }, { "grad_norm": 0.2803696393966675, "learning_rate": 5.1324274449478204e-05, "loss": 0.0235, "step": 51700 }, { "grad_norm": 0.6402662396430969, "learning_rate": 5.130774547959909e-05, "loss": 0.0272, "step": 51710 }, { "grad_norm": 0.5763791799545288, "learning_rate": 5.129121636670692e-05, "loss": 0.026, "step": 51720 }, { "grad_norm": 0.22254104912281036, "learning_rate": 5.1274687112609246e-05, "loss": 0.0225, "step": 51730 }, { "grad_norm": 0.28731828927993774, "learning_rate": 5.125815771911372e-05, "loss": 0.0204, "step": 51740 }, { "grad_norm": 0.48605790734291077, "learning_rate": 5.1241628188027956e-05, "loss": 0.0227, "step": 51750 }, { "grad_norm": 0.3364018499851227, "learning_rate": 5.1225098521159585e-05, "loss": 0.0203, "step": 51760 }, { "grad_norm": 0.39783942699432373, "learning_rate": 5.12085687203163e-05, "loss": 0.0236, "step": 51770 }, { "grad_norm": 0.3935631811618805, "learning_rate": 5.119203878730575e-05, "loss": 0.0277, "step": 51780 }, { "grad_norm": 0.3730373978614807, "learning_rate": 5.117550872393564e-05, "loss": 0.0327, "step": 51790 }, { "grad_norm": 0.345777690410614, "learning_rate": 5.1158978532013645e-05, "loss": 0.0266, "step": 51800 }, { "grad_norm": 0.40569615364074707, "learning_rate": 5.114244821334751e-05, "loss": 0.0276, "step": 51810 }, { "grad_norm": 0.7238178849220276, "learning_rate": 5.112591776974497e-05, "loss": 0.0259, "step": 51820 }, { "grad_norm": 0.38048025965690613, "learning_rate": 5.110938720301374e-05, "loss": 0.0275, "step": 51830 }, { "grad_norm": 0.35373350977897644, "learning_rate": 5.1092856514961595e-05, "loss": 0.0335, "step": 51840 }, { "grad_norm": 0.3830605745315552, "learning_rate": 5.107632570739631e-05, "loss": 0.0265, "step": 51850 }, { "grad_norm": 0.29835617542266846, "learning_rate": 5.1059794782125645e-05, "loss": 0.0296, "step": 51860 }, { "grad_norm": 0.48286697268486023, "learning_rate": 5.104326374095744e-05, "loss": 0.0378, "step": 51870 }, { "grad_norm": 0.34554439783096313, "learning_rate": 5.102673258569947e-05, "loss": 0.0297, "step": 51880 }, { "grad_norm": 0.416663259267807, "learning_rate": 5.101020131815956e-05, "loss": 0.0341, "step": 51890 }, { "grad_norm": 0.2885551154613495, "learning_rate": 5.099366994014556e-05, "loss": 0.0282, "step": 51900 }, { "grad_norm": 0.2843458950519562, "learning_rate": 5.097713845346532e-05, "loss": 0.0298, "step": 51910 }, { "grad_norm": 0.24171669781208038, "learning_rate": 5.096060685992668e-05, "loss": 0.0243, "step": 51920 }, { "grad_norm": 0.4739999771118164, "learning_rate": 5.0944075161337514e-05, "loss": 0.0285, "step": 51930 }, { "grad_norm": 0.31739673018455505, "learning_rate": 5.09275433595057e-05, "loss": 0.0263, "step": 51940 }, { "grad_norm": 0.32267996668815613, "learning_rate": 5.0911011456239157e-05, "loss": 0.0287, "step": 51950 }, { "grad_norm": 0.7985670566558838, "learning_rate": 5.089447945334574e-05, "loss": 0.0268, "step": 51960 }, { "grad_norm": 0.3168126344680786, "learning_rate": 5.087794735263343e-05, "loss": 0.024, "step": 51970 }, { "grad_norm": 0.20485326647758484, "learning_rate": 5.086141515591011e-05, "loss": 0.0224, "step": 51980 }, { "grad_norm": 0.337522953748703, "learning_rate": 5.084488286498371e-05, "loss": 0.0246, "step": 51990 }, { "grad_norm": 0.34817224740982056, "learning_rate": 5.08283504816622e-05, "loss": 0.0292, "step": 52000 }, { "grad_norm": 0.4955455958843231, "learning_rate": 5.081181800775353e-05, "loss": 0.034, "step": 52010 }, { "grad_norm": 0.3444499373435974, "learning_rate": 5.079528544506565e-05, "loss": 0.0292, "step": 52020 }, { "grad_norm": 0.40006738901138306, "learning_rate": 5.077875279540657e-05, "loss": 0.0269, "step": 52030 }, { "grad_norm": 0.2498311996459961, "learning_rate": 5.076222006058424e-05, "loss": 0.0336, "step": 52040 }, { "grad_norm": 0.4811253845691681, "learning_rate": 5.074568724240667e-05, "loss": 0.0294, "step": 52050 }, { "grad_norm": 0.3301822543144226, "learning_rate": 5.072915434268187e-05, "loss": 0.0343, "step": 52060 }, { "grad_norm": 0.3629744052886963, "learning_rate": 5.071262136321785e-05, "loss": 0.0275, "step": 52070 }, { "grad_norm": 0.424809068441391, "learning_rate": 5.069608830582262e-05, "loss": 0.0282, "step": 52080 }, { "grad_norm": 0.26942625641822815, "learning_rate": 5.067955517230423e-05, "loss": 0.0268, "step": 52090 }, { "grad_norm": 0.24590638279914856, "learning_rate": 5.06630219644707e-05, "loss": 0.0302, "step": 52100 }, { "grad_norm": 0.3399603068828583, "learning_rate": 5.064648868413008e-05, "loss": 0.0262, "step": 52110 }, { "grad_norm": 0.27620887756347656, "learning_rate": 5.062995533309044e-05, "loss": 0.0205, "step": 52120 }, { "grad_norm": 0.3113323748111725, "learning_rate": 5.061342191315983e-05, "loss": 0.0277, "step": 52130 }, { "grad_norm": 0.3162737190723419, "learning_rate": 5.0596888426146325e-05, "loss": 0.0265, "step": 52140 }, { "grad_norm": 1.4853339195251465, "learning_rate": 5.058035487385798e-05, "loss": 0.0325, "step": 52150 }, { "grad_norm": 0.27283382415771484, "learning_rate": 5.0563821258102915e-05, "loss": 0.0209, "step": 52160 }, { "grad_norm": 0.318310022354126, "learning_rate": 5.05472875806892e-05, "loss": 0.021, "step": 52170 }, { "grad_norm": 0.34699568152427673, "learning_rate": 5.053075384342494e-05, "loss": 0.0279, "step": 52180 }, { "grad_norm": 0.3039321303367615, "learning_rate": 5.0514220048118234e-05, "loss": 0.0258, "step": 52190 }, { "grad_norm": 0.3340569734573364, "learning_rate": 5.049768619657721e-05, "loss": 0.0273, "step": 52200 }, { "grad_norm": 0.28190091252326965, "learning_rate": 5.048115229060996e-05, "loss": 0.0279, "step": 52210 }, { "grad_norm": 0.36721259355545044, "learning_rate": 5.046461833202462e-05, "loss": 0.0314, "step": 52220 }, { "grad_norm": 0.2405906617641449, "learning_rate": 5.044808432262932e-05, "loss": 0.0332, "step": 52230 }, { "grad_norm": 0.3751195967197418, "learning_rate": 5.043155026423219e-05, "loss": 0.0229, "step": 52240 }, { "grad_norm": 0.3691813051700592, "learning_rate": 5.041501615864137e-05, "loss": 0.0157, "step": 52250 }, { "grad_norm": 0.32129940390586853, "learning_rate": 5.039848200766502e-05, "loss": 0.0244, "step": 52260 }, { "grad_norm": 0.5178605914115906, "learning_rate": 5.038194781311125e-05, "loss": 0.0243, "step": 52270 }, { "grad_norm": 0.4704601466655731, "learning_rate": 5.0365413576788264e-05, "loss": 0.0271, "step": 52280 }, { "grad_norm": 0.3924577534198761, "learning_rate": 5.0348879300504194e-05, "loss": 0.0252, "step": 52290 }, { "grad_norm": 0.24355971813201904, "learning_rate": 5.033234498606721e-05, "loss": 0.0199, "step": 52300 }, { "grad_norm": 0.42153653502464294, "learning_rate": 5.0315810635285476e-05, "loss": 0.0223, "step": 52310 }, { "grad_norm": 0.3322833776473999, "learning_rate": 5.0299276249967164e-05, "loss": 0.0237, "step": 52320 }, { "grad_norm": 0.28406232595443726, "learning_rate": 5.0282741831920454e-05, "loss": 0.034, "step": 52330 }, { "grad_norm": 0.5287095904350281, "learning_rate": 5.026620738295351e-05, "loss": 0.0251, "step": 52340 }, { "grad_norm": 0.6877027153968811, "learning_rate": 5.024967290487455e-05, "loss": 0.0257, "step": 52350 }, { "grad_norm": 0.6479992866516113, "learning_rate": 5.023313839949172e-05, "loss": 0.0299, "step": 52360 }, { "grad_norm": 0.4442969858646393, "learning_rate": 5.0216603868613224e-05, "loss": 0.0268, "step": 52370 }, { "grad_norm": 0.36319735646247864, "learning_rate": 5.020006931404726e-05, "loss": 0.027, "step": 52380 }, { "grad_norm": 0.3254499137401581, "learning_rate": 5.0183534737602014e-05, "loss": 0.0225, "step": 52390 }, { "grad_norm": 0.3199078142642975, "learning_rate": 5.0167000141085686e-05, "loss": 0.0301, "step": 52400 }, { "grad_norm": 0.27772006392478943, "learning_rate": 5.015046552630647e-05, "loss": 0.0238, "step": 52410 }, { "grad_norm": 0.35680943727493286, "learning_rate": 5.013393089507258e-05, "loss": 0.0309, "step": 52420 }, { "grad_norm": 0.2828739881515503, "learning_rate": 5.011739624919219e-05, "loss": 0.0238, "step": 52430 }, { "grad_norm": 0.45060503482818604, "learning_rate": 5.0100861590473545e-05, "loss": 0.0202, "step": 52440 }, { "grad_norm": 0.32015103101730347, "learning_rate": 5.008432692072482e-05, "loss": 0.021, "step": 52450 }, { "grad_norm": 0.41613873839378357, "learning_rate": 5.006779224175423e-05, "loss": 0.031, "step": 52460 }, { "grad_norm": 0.4189659357070923, "learning_rate": 5.005125755536998e-05, "loss": 0.0295, "step": 52470 }, { "grad_norm": 0.24009360373020172, "learning_rate": 5.003472286338029e-05, "loss": 0.0292, "step": 52480 }, { "grad_norm": 0.2936019003391266, "learning_rate": 5.001818816759335e-05, "loss": 0.033, "step": 52490 }, { "grad_norm": 0.2290768176317215, "learning_rate": 5.000165346981738e-05, "loss": 0.0226, "step": 52500 }, { "grad_norm": 0.33766594529151917, "learning_rate": 4.9985118771860595e-05, "loss": 0.0267, "step": 52510 }, { "grad_norm": 0.37401023507118225, "learning_rate": 4.996858407553119e-05, "loss": 0.0243, "step": 52520 }, { "grad_norm": 0.25281503796577454, "learning_rate": 4.99520493826374e-05, "loss": 0.0235, "step": 52530 }, { "grad_norm": 0.35406771302223206, "learning_rate": 4.9935514694987384e-05, "loss": 0.0267, "step": 52540 }, { "grad_norm": 0.266284316778183, "learning_rate": 4.991898001438941e-05, "loss": 0.0203, "step": 52550 }, { "grad_norm": 0.40152138471603394, "learning_rate": 4.990244534265166e-05, "loss": 0.0313, "step": 52560 }, { "grad_norm": 0.358469694852829, "learning_rate": 4.988591068158234e-05, "loss": 0.0266, "step": 52570 }, { "grad_norm": 0.36678051948547363, "learning_rate": 4.986937603298965e-05, "loss": 0.0288, "step": 52580 }, { "grad_norm": 0.37214910984039307, "learning_rate": 4.985284139868181e-05, "loss": 0.0206, "step": 52590 }, { "grad_norm": 0.48207029700279236, "learning_rate": 4.9836306780467e-05, "loss": 0.0261, "step": 52600 }, { "grad_norm": 0.2955208420753479, "learning_rate": 4.981977218015343e-05, "loss": 0.0221, "step": 52610 }, { "grad_norm": 0.2993270456790924, "learning_rate": 4.980323759954932e-05, "loss": 0.0277, "step": 52620 }, { "grad_norm": 0.23627084493637085, "learning_rate": 4.978670304046284e-05, "loss": 0.022, "step": 52630 }, { "grad_norm": 0.22453533113002777, "learning_rate": 4.9770168504702195e-05, "loss": 0.0249, "step": 52640 }, { "grad_norm": 0.3717082142829895, "learning_rate": 4.975363399407559e-05, "loss": 0.0237, "step": 52650 }, { "grad_norm": 0.5127458572387695, "learning_rate": 4.9737099510391194e-05, "loss": 0.0282, "step": 52660 }, { "grad_norm": 0.297745019197464, "learning_rate": 4.972056505545719e-05, "loss": 0.0335, "step": 52670 }, { "grad_norm": 0.2944374084472656, "learning_rate": 4.9704030631081775e-05, "loss": 0.024, "step": 52680 }, { "grad_norm": 0.24629628658294678, "learning_rate": 4.968749623907313e-05, "loss": 0.0232, "step": 52690 }, { "grad_norm": 0.2830260992050171, "learning_rate": 4.967096188123943e-05, "loss": 0.0328, "step": 52700 }, { "grad_norm": 0.37658512592315674, "learning_rate": 4.965442755938884e-05, "loss": 0.0258, "step": 52710 }, { "grad_norm": 0.3139815032482147, "learning_rate": 4.963789327532954e-05, "loss": 0.0219, "step": 52720 }, { "grad_norm": 0.36284339427948, "learning_rate": 4.962135903086968e-05, "loss": 0.0221, "step": 52730 }, { "grad_norm": 0.3142892122268677, "learning_rate": 4.960482482781741e-05, "loss": 0.0204, "step": 52740 }, { "grad_norm": 0.32988324761390686, "learning_rate": 4.958829066798092e-05, "loss": 0.0213, "step": 52750 }, { "grad_norm": 0.30579930543899536, "learning_rate": 4.957175655316835e-05, "loss": 0.0222, "step": 52760 }, { "grad_norm": 0.38975459337234497, "learning_rate": 4.9555222485187834e-05, "loss": 0.0284, "step": 52770 }, { "grad_norm": 0.46459972858428955, "learning_rate": 4.953868846584752e-05, "loss": 0.0249, "step": 52780 }, { "grad_norm": 0.2834029793739319, "learning_rate": 4.9522154496955544e-05, "loss": 0.0337, "step": 52790 }, { "grad_norm": 0.6456732153892517, "learning_rate": 4.950562058032001e-05, "loss": 0.0228, "step": 52800 }, { "grad_norm": 0.22701025009155273, "learning_rate": 4.948908671774909e-05, "loss": 0.0271, "step": 52810 }, { "grad_norm": 0.28435564041137695, "learning_rate": 4.947255291105087e-05, "loss": 0.0246, "step": 52820 }, { "grad_norm": 0.31865328550338745, "learning_rate": 4.945601916203348e-05, "loss": 0.0239, "step": 52830 }, { "grad_norm": 0.3891160488128662, "learning_rate": 4.9439485472505016e-05, "loss": 0.0259, "step": 52840 }, { "grad_norm": 0.6074011921882629, "learning_rate": 4.942295184427356e-05, "loss": 0.0308, "step": 52850 }, { "grad_norm": 0.2856079936027527, "learning_rate": 4.9406418279147234e-05, "loss": 0.0234, "step": 52860 }, { "grad_norm": 0.4550679624080658, "learning_rate": 4.9389884778934116e-05, "loss": 0.0357, "step": 52870 }, { "grad_norm": 0.22132474184036255, "learning_rate": 4.937335134544228e-05, "loss": 0.0234, "step": 52880 }, { "grad_norm": 0.48809152841567993, "learning_rate": 4.935681798047979e-05, "loss": 0.0233, "step": 52890 }, { "grad_norm": 0.4550634026527405, "learning_rate": 4.934028468585473e-05, "loss": 0.0286, "step": 52900 }, { "grad_norm": 0.444908082485199, "learning_rate": 4.932375146337514e-05, "loss": 0.0288, "step": 52910 }, { "grad_norm": 0.24940276145935059, "learning_rate": 4.930721831484908e-05, "loss": 0.0197, "step": 52920 }, { "grad_norm": 0.451821506023407, "learning_rate": 4.929068524208456e-05, "loss": 0.0273, "step": 52930 }, { "grad_norm": 0.32536840438842773, "learning_rate": 4.927415224688965e-05, "loss": 0.0328, "step": 52940 }, { "grad_norm": 0.3192013204097748, "learning_rate": 4.9257619331072375e-05, "loss": 0.0395, "step": 52950 }, { "grad_norm": 0.3662243187427521, "learning_rate": 4.924108649644072e-05, "loss": 0.0206, "step": 52960 }, { "grad_norm": 0.35329025983810425, "learning_rate": 4.922455374480271e-05, "loss": 0.0379, "step": 52970 }, { "grad_norm": 0.3873736262321472, "learning_rate": 4.920802107796634e-05, "loss": 0.0275, "step": 52980 }, { "grad_norm": 0.46317151188850403, "learning_rate": 4.9191488497739564e-05, "loss": 0.024, "step": 52990 }, { "grad_norm": 0.26252877712249756, "learning_rate": 4.917495600593042e-05, "loss": 0.0241, "step": 53000 }, { "grad_norm": 0.40470612049102783, "learning_rate": 4.915842360434684e-05, "loss": 0.0257, "step": 53010 }, { "grad_norm": 0.4231685698032379, "learning_rate": 4.914189129479679e-05, "loss": 0.038, "step": 53020 }, { "grad_norm": 0.27554377913475037, "learning_rate": 4.912535907908822e-05, "loss": 0.0256, "step": 53030 }, { "grad_norm": 0.3684856593608856, "learning_rate": 4.910882695902907e-05, "loss": 0.0328, "step": 53040 }, { "grad_norm": 0.335812509059906, "learning_rate": 4.9092294936427234e-05, "loss": 0.0252, "step": 53050 }, { "grad_norm": 0.35200628638267517, "learning_rate": 4.9075763013090685e-05, "loss": 0.0247, "step": 53060 }, { "grad_norm": 0.3872368037700653, "learning_rate": 4.9059231190827305e-05, "loss": 0.0331, "step": 53070 }, { "grad_norm": 0.3410489559173584, "learning_rate": 4.904269947144499e-05, "loss": 0.0413, "step": 53080 }, { "grad_norm": 0.2924044728279114, "learning_rate": 4.902616785675161e-05, "loss": 0.0243, "step": 53090 }, { "grad_norm": 0.36584553122520447, "learning_rate": 4.900963634855505e-05, "loss": 0.029, "step": 53100 }, { "grad_norm": 0.24743179976940155, "learning_rate": 4.8993104948663173e-05, "loss": 0.0219, "step": 53110 }, { "grad_norm": 0.38570407032966614, "learning_rate": 4.8976573658883823e-05, "loss": 0.0279, "step": 53120 }, { "grad_norm": 0.35630717873573303, "learning_rate": 4.896004248102484e-05, "loss": 0.0271, "step": 53130 }, { "grad_norm": 0.3700225353240967, "learning_rate": 4.8943511416894045e-05, "loss": 0.0258, "step": 53140 }, { "grad_norm": 0.21583618223667145, "learning_rate": 4.892698046829926e-05, "loss": 0.0238, "step": 53150 }, { "grad_norm": 0.5045012831687927, "learning_rate": 4.8910449637048275e-05, "loss": 0.034, "step": 53160 }, { "grad_norm": 0.47152870893478394, "learning_rate": 4.8893918924948884e-05, "loss": 0.0322, "step": 53170 }, { "grad_norm": 0.3201390206813812, "learning_rate": 4.8877388333808835e-05, "loss": 0.0265, "step": 53180 }, { "grad_norm": 0.37489908933639526, "learning_rate": 4.8860857865435925e-05, "loss": 0.0281, "step": 53190 }, { "grad_norm": 0.2602739930152893, "learning_rate": 4.884432752163789e-05, "loss": 0.021, "step": 53200 }, { "grad_norm": 0.3070337176322937, "learning_rate": 4.8827797304222466e-05, "loss": 0.0266, "step": 53210 }, { "grad_norm": 0.4815758168697357, "learning_rate": 4.8811267214997366e-05, "loss": 0.0253, "step": 53220 }, { "grad_norm": 0.4924395680427551, "learning_rate": 4.879473725577029e-05, "loss": 0.0279, "step": 53230 }, { "grad_norm": 0.3852648437023163, "learning_rate": 4.877820742834891e-05, "loss": 0.0207, "step": 53240 }, { "grad_norm": 0.39310553669929504, "learning_rate": 4.8761677734540956e-05, "loss": 0.0268, "step": 53250 }, { "grad_norm": 0.23396308720111847, "learning_rate": 4.874514817615407e-05, "loss": 0.0282, "step": 53260 }, { "grad_norm": 0.2517567276954651, "learning_rate": 4.872861875499588e-05, "loss": 0.0302, "step": 53270 }, { "grad_norm": 0.3980911374092102, "learning_rate": 4.871208947287404e-05, "loss": 0.0309, "step": 53280 }, { "grad_norm": 0.5926041603088379, "learning_rate": 4.869556033159615e-05, "loss": 0.0248, "step": 53290 }, { "grad_norm": 0.22627997398376465, "learning_rate": 4.8679031332969835e-05, "loss": 0.0213, "step": 53300 }, { "grad_norm": 0.27874886989593506, "learning_rate": 4.8662502478802635e-05, "loss": 0.0359, "step": 53310 }, { "grad_norm": 0.3713403046131134, "learning_rate": 4.8645973770902176e-05, "loss": 0.0321, "step": 53320 }, { "grad_norm": 0.37238961458206177, "learning_rate": 4.862944521107599e-05, "loss": 0.024, "step": 53330 }, { "grad_norm": 0.4508076608181, "learning_rate": 4.861291680113162e-05, "loss": 0.028, "step": 53340 }, { "grad_norm": 0.2610829770565033, "learning_rate": 4.8596388542876563e-05, "loss": 0.0196, "step": 53350 }, { "grad_norm": 0.2557763159275055, "learning_rate": 4.857986043811836e-05, "loss": 0.023, "step": 53360 }, { "grad_norm": 0.332273006439209, "learning_rate": 4.856333248866448e-05, "loss": 0.0224, "step": 53370 }, { "grad_norm": 0.3078494369983673, "learning_rate": 4.85468046963224e-05, "loss": 0.0197, "step": 53380 }, { "grad_norm": 0.3815392851829529, "learning_rate": 4.853027706289957e-05, "loss": 0.0231, "step": 53390 }, { "grad_norm": 0.4621836841106415, "learning_rate": 4.851374959020344e-05, "loss": 0.0223, "step": 53400 }, { "grad_norm": 0.4275457561016083, "learning_rate": 4.849722228004141e-05, "loss": 0.0339, "step": 53410 }, { "grad_norm": 0.22655700147151947, "learning_rate": 4.848069513422091e-05, "loss": 0.024, "step": 53420 }, { "grad_norm": 0.4503273665904999, "learning_rate": 4.8464168154549294e-05, "loss": 0.0223, "step": 53430 }, { "grad_norm": 0.44118791818618774, "learning_rate": 4.844764134283392e-05, "loss": 0.0263, "step": 53440 }, { "grad_norm": 0.28018760681152344, "learning_rate": 4.843111470088218e-05, "loss": 0.0254, "step": 53450 }, { "grad_norm": 0.4273603558540344, "learning_rate": 4.841458823050137e-05, "loss": 0.024, "step": 53460 }, { "grad_norm": 0.2470264881849289, "learning_rate": 4.8398061933498816e-05, "loss": 0.0228, "step": 53470 }, { "grad_norm": 0.4877064824104309, "learning_rate": 4.83815358116818e-05, "loss": 0.0241, "step": 53480 }, { "grad_norm": 0.3673839569091797, "learning_rate": 4.8365009866857584e-05, "loss": 0.0319, "step": 53490 }, { "grad_norm": 0.3614731729030609, "learning_rate": 4.834848410083342e-05, "loss": 0.0205, "step": 53500 }, { "grad_norm": 0.38437825441360474, "learning_rate": 4.833195851541657e-05, "loss": 0.0303, "step": 53510 }, { "grad_norm": 0.5262593030929565, "learning_rate": 4.831543311241423e-05, "loss": 0.0281, "step": 53520 }, { "grad_norm": 0.50107741355896, "learning_rate": 4.829890789363359e-05, "loss": 0.0311, "step": 53530 }, { "grad_norm": 0.40178170800209045, "learning_rate": 4.828238286088182e-05, "loss": 0.02, "step": 53540 }, { "grad_norm": 0.3469194173812866, "learning_rate": 4.826585801596609e-05, "loss": 0.0277, "step": 53550 }, { "grad_norm": 0.306540310382843, "learning_rate": 4.8249333360693486e-05, "loss": 0.0244, "step": 53560 }, { "grad_norm": 0.30425551533699036, "learning_rate": 4.8232808896871185e-05, "loss": 0.0227, "step": 53570 }, { "grad_norm": 0.32934147119522095, "learning_rate": 4.821628462630624e-05, "loss": 0.0202, "step": 53580 }, { "grad_norm": 0.414747029542923, "learning_rate": 4.8199760550805735e-05, "loss": 0.0251, "step": 53590 }, { "grad_norm": 0.4082730710506439, "learning_rate": 4.818323667217669e-05, "loss": 0.0208, "step": 53600 }, { "grad_norm": 0.3766939342021942, "learning_rate": 4.8166712992226165e-05, "loss": 0.025, "step": 53610 }, { "grad_norm": 0.37595126032829285, "learning_rate": 4.8150189512761154e-05, "loss": 0.0226, "step": 53620 }, { "grad_norm": 0.21350769698619843, "learning_rate": 4.813366623558862e-05, "loss": 0.0245, "step": 53630 }, { "grad_norm": 0.26632291078567505, "learning_rate": 4.811714316251554e-05, "loss": 0.0248, "step": 53640 }, { "grad_norm": 0.34883391857147217, "learning_rate": 4.810062029534886e-05, "loss": 0.0259, "step": 53650 }, { "grad_norm": 0.3261337876319885, "learning_rate": 4.8084097635895505e-05, "loss": 0.0283, "step": 53660 }, { "grad_norm": 0.2662467956542969, "learning_rate": 4.8067575185962335e-05, "loss": 0.0255, "step": 53670 }, { "grad_norm": 0.6889213919639587, "learning_rate": 4.805105294735625e-05, "loss": 0.0253, "step": 53680 }, { "grad_norm": 0.3842369019985199, "learning_rate": 4.803453092188406e-05, "loss": 0.0292, "step": 53690 }, { "grad_norm": 0.4196058213710785, "learning_rate": 4.801800911135263e-05, "loss": 0.0298, "step": 53700 }, { "grad_norm": 0.2575852870941162, "learning_rate": 4.8001487517568744e-05, "loss": 0.0197, "step": 53710 }, { "grad_norm": 0.6109157800674438, "learning_rate": 4.798496614233918e-05, "loss": 0.0333, "step": 53720 }, { "grad_norm": 0.22297149896621704, "learning_rate": 4.796844498747069e-05, "loss": 0.0329, "step": 53730 }, { "grad_norm": 0.39293134212493896, "learning_rate": 4.795192405477e-05, "loss": 0.0235, "step": 53740 }, { "grad_norm": 0.4828093945980072, "learning_rate": 4.79354033460438e-05, "loss": 0.0276, "step": 53750 }, { "grad_norm": 0.4636825919151306, "learning_rate": 4.79188828630988e-05, "loss": 0.0331, "step": 53760 }, { "grad_norm": 0.4550035297870636, "learning_rate": 4.790236260774165e-05, "loss": 0.0228, "step": 53770 }, { "grad_norm": 0.2732059359550476, "learning_rate": 4.788584258177896e-05, "loss": 0.0245, "step": 53780 }, { "grad_norm": 0.5055434107780457, "learning_rate": 4.7869322787017356e-05, "loss": 0.0289, "step": 53790 }, { "grad_norm": 0.3758983314037323, "learning_rate": 4.78528032252634e-05, "loss": 0.0227, "step": 53800 }, { "grad_norm": 0.31109878420829773, "learning_rate": 4.783628389832366e-05, "loss": 0.0269, "step": 53810 }, { "grad_norm": 0.2859196960926056, "learning_rate": 4.7819764808004644e-05, "loss": 0.0255, "step": 53820 }, { "grad_norm": 0.44458091259002686, "learning_rate": 4.780324595611289e-05, "loss": 0.0267, "step": 53830 }, { "grad_norm": 0.28021860122680664, "learning_rate": 4.7786727344454854e-05, "loss": 0.0284, "step": 53840 }, { "grad_norm": 0.35042375326156616, "learning_rate": 4.777020897483697e-05, "loss": 0.0274, "step": 53850 }, { "grad_norm": 0.294562429189682, "learning_rate": 4.775369084906569e-05, "loss": 0.0238, "step": 53860 }, { "grad_norm": 0.2747413218021393, "learning_rate": 4.773717296894742e-05, "loss": 0.0307, "step": 53870 }, { "grad_norm": 0.3149530589580536, "learning_rate": 4.772065533628848e-05, "loss": 0.0257, "step": 53880 }, { "grad_norm": 0.36379480361938477, "learning_rate": 4.7704137952895265e-05, "loss": 0.0195, "step": 53890 }, { "grad_norm": 0.6774382591247559, "learning_rate": 4.768762082057407e-05, "loss": 0.0314, "step": 53900 }, { "grad_norm": 0.45317184925079346, "learning_rate": 4.7671103941131186e-05, "loss": 0.0285, "step": 53910 }, { "grad_norm": 0.3267820477485657, "learning_rate": 4.765458731637287e-05, "loss": 0.0233, "step": 53920 }, { "grad_norm": 0.28370201587677, "learning_rate": 4.763807094810537e-05, "loss": 0.0278, "step": 53930 }, { "grad_norm": 0.29249441623687744, "learning_rate": 4.762155483813487e-05, "loss": 0.0259, "step": 53940 }, { "grad_norm": 0.56343674659729, "learning_rate": 4.760503898826754e-05, "loss": 0.0331, "step": 53950 }, { "grad_norm": 0.3650006353855133, "learning_rate": 4.758852340030957e-05, "loss": 0.0342, "step": 53960 }, { "grad_norm": 0.32942208647727966, "learning_rate": 4.757200807606705e-05, "loss": 0.0281, "step": 53970 }, { "grad_norm": 0.2714601457118988, "learning_rate": 4.755549301734607e-05, "loss": 0.0254, "step": 53980 }, { "grad_norm": 0.27696385979652405, "learning_rate": 4.753897822595271e-05, "loss": 0.0369, "step": 53990 }, { "grad_norm": 0.31076711416244507, "learning_rate": 4.7522463703692994e-05, "loss": 0.027, "step": 54000 }, { "grad_norm": 0.3204643428325653, "learning_rate": 4.75059494523729e-05, "loss": 0.0282, "step": 54010 }, { "grad_norm": 0.24070575833320618, "learning_rate": 4.748943547379843e-05, "loss": 0.0262, "step": 54020 }, { "grad_norm": 0.290511816740036, "learning_rate": 4.7472921769775535e-05, "loss": 0.0321, "step": 54030 }, { "grad_norm": 0.4547467529773712, "learning_rate": 4.745640834211011e-05, "loss": 0.0353, "step": 54040 }, { "grad_norm": 0.36717039346694946, "learning_rate": 4.7439895192608054e-05, "loss": 0.0269, "step": 54050 }, { "grad_norm": 0.38608992099761963, "learning_rate": 4.7423382323075186e-05, "loss": 0.0274, "step": 54060 }, { "grad_norm": 0.29342320561408997, "learning_rate": 4.740686973531737e-05, "loss": 0.0181, "step": 54070 }, { "grad_norm": 0.5261562466621399, "learning_rate": 4.7390357431140375e-05, "loss": 0.0261, "step": 54080 }, { "grad_norm": 0.35806992650032043, "learning_rate": 4.737384541234998e-05, "loss": 0.0269, "step": 54090 }, { "grad_norm": 0.3645831346511841, "learning_rate": 4.735733368075189e-05, "loss": 0.021, "step": 54100 }, { "grad_norm": 0.3108954429626465, "learning_rate": 4.734082223815182e-05, "loss": 0.0264, "step": 54110 }, { "grad_norm": 0.31784260272979736, "learning_rate": 4.7324311086355436e-05, "loss": 0.0191, "step": 54120 }, { "grad_norm": 0.36374375224113464, "learning_rate": 4.730780022716836e-05, "loss": 0.0266, "step": 54130 }, { "grad_norm": 0.345673143863678, "learning_rate": 4.729128966239619e-05, "loss": 0.021, "step": 54140 }, { "grad_norm": 0.42874327301979065, "learning_rate": 4.7274779393844536e-05, "loss": 0.0312, "step": 54150 }, { "grad_norm": 0.630115807056427, "learning_rate": 4.725826942331891e-05, "loss": 0.0471, "step": 54160 }, { "grad_norm": 0.2641415596008301, "learning_rate": 4.724175975262481e-05, "loss": 0.0232, "step": 54170 }, { "grad_norm": 0.3746200501918793, "learning_rate": 4.722525038356772e-05, "loss": 0.0236, "step": 54180 }, { "grad_norm": 0.2785765528678894, "learning_rate": 4.7208741317953075e-05, "loss": 0.0253, "step": 54190 }, { "grad_norm": 0.35812056064605713, "learning_rate": 4.7192232557586264e-05, "loss": 0.0412, "step": 54200 }, { "grad_norm": 0.31106114387512207, "learning_rate": 4.717572410427269e-05, "loss": 0.0243, "step": 54210 }, { "grad_norm": 0.317735880613327, "learning_rate": 4.71592159598177e-05, "loss": 0.0265, "step": 54220 }, { "grad_norm": 0.6512383818626404, "learning_rate": 4.714270812602657e-05, "loss": 0.0283, "step": 54230 }, { "grad_norm": 0.26177290081977844, "learning_rate": 4.712620060470458e-05, "loss": 0.026, "step": 54240 }, { "grad_norm": 0.3854628801345825, "learning_rate": 4.7109693397656986e-05, "loss": 0.0278, "step": 54250 }, { "grad_norm": 0.17729303240776062, "learning_rate": 4.709318650668894e-05, "loss": 0.0188, "step": 54260 }, { "grad_norm": 0.2543083131313324, "learning_rate": 4.7076679933605676e-05, "loss": 0.0227, "step": 54270 }, { "grad_norm": 0.25773563981056213, "learning_rate": 4.70601736802123e-05, "loss": 0.0246, "step": 54280 }, { "grad_norm": 0.4041895270347595, "learning_rate": 4.704366774831391e-05, "loss": 0.0403, "step": 54290 }, { "grad_norm": 0.21884016692638397, "learning_rate": 4.702716213971557e-05, "loss": 0.026, "step": 54300 }, { "grad_norm": 0.3190443217754364, "learning_rate": 4.70106568562223e-05, "loss": 0.0227, "step": 54310 }, { "grad_norm": 0.24683961272239685, "learning_rate": 4.6994151899639105e-05, "loss": 0.0257, "step": 54320 }, { "grad_norm": 0.21914485096931458, "learning_rate": 4.697764727177093e-05, "loss": 0.0282, "step": 54330 }, { "grad_norm": 0.42606183886528015, "learning_rate": 4.696114297442272e-05, "loss": 0.022, "step": 54340 }, { "grad_norm": 0.24898916482925415, "learning_rate": 4.694463900939933e-05, "loss": 0.025, "step": 54350 }, { "grad_norm": 0.28515318036079407, "learning_rate": 4.692813537850564e-05, "loss": 0.0242, "step": 54360 }, { "grad_norm": 0.22478091716766357, "learning_rate": 4.6911632083546445e-05, "loss": 0.017, "step": 54370 }, { "grad_norm": 0.42541465163230896, "learning_rate": 4.6895129126326525e-05, "loss": 0.0312, "step": 54380 }, { "grad_norm": 0.2573666274547577, "learning_rate": 4.687862650865059e-05, "loss": 0.0217, "step": 54390 }, { "grad_norm": 1.738332986831665, "learning_rate": 4.686212423232339e-05, "loss": 0.0327, "step": 54400 }, { "grad_norm": 0.39069831371307373, "learning_rate": 4.6845622299149565e-05, "loss": 0.0253, "step": 54410 }, { "grad_norm": 0.2768397033214569, "learning_rate": 4.682912071093374e-05, "loss": 0.0293, "step": 54420 }, { "grad_norm": 0.262989342212677, "learning_rate": 4.681261946948052e-05, "loss": 0.0217, "step": 54430 }, { "grad_norm": 1.2953764200210571, "learning_rate": 4.679611857659443e-05, "loss": 0.0359, "step": 54440 }, { "grad_norm": 0.31084755063056946, "learning_rate": 4.6779618034079975e-05, "loss": 0.0249, "step": 54450 }, { "grad_norm": 0.35014858841896057, "learning_rate": 4.6763117843741675e-05, "loss": 0.0302, "step": 54460 }, { "grad_norm": 0.3478705883026123, "learning_rate": 4.674661800738394e-05, "loss": 0.0227, "step": 54470 }, { "grad_norm": 0.4130021929740906, "learning_rate": 4.673011852681117e-05, "loss": 0.022, "step": 54480 }, { "grad_norm": 0.29023319482803345, "learning_rate": 4.671361940382772e-05, "loss": 0.0298, "step": 54490 }, { "grad_norm": 0.27891218662261963, "learning_rate": 4.669712064023791e-05, "loss": 0.0285, "step": 54500 }, { "grad_norm": 0.3205404579639435, "learning_rate": 4.668062223784602e-05, "loss": 0.0223, "step": 54510 }, { "grad_norm": 0.3555539548397064, "learning_rate": 4.6664124198456274e-05, "loss": 0.026, "step": 54520 }, { "grad_norm": 0.35239043831825256, "learning_rate": 4.6647626523872905e-05, "loss": 0.0275, "step": 54530 }, { "grad_norm": 0.2383500635623932, "learning_rate": 4.6631129215900066e-05, "loss": 0.0254, "step": 54540 }, { "grad_norm": 0.22047623991966248, "learning_rate": 4.661463227634187e-05, "loss": 0.0359, "step": 54550 }, { "grad_norm": 0.4337112605571747, "learning_rate": 4.659813570700238e-05, "loss": 0.0192, "step": 54560 }, { "grad_norm": 0.2697330713272095, "learning_rate": 4.658163950968568e-05, "loss": 0.0195, "step": 54570 }, { "grad_norm": 0.2431943565607071, "learning_rate": 4.6565143686195726e-05, "loss": 0.0213, "step": 54580 }, { "grad_norm": 0.28467798233032227, "learning_rate": 4.654864823833649e-05, "loss": 0.0237, "step": 54590 }, { "grad_norm": 0.28460216522216797, "learning_rate": 4.65321531679119e-05, "loss": 0.022, "step": 54600 }, { "grad_norm": 0.2593911588191986, "learning_rate": 4.6515658476725834e-05, "loss": 0.0358, "step": 54610 }, { "grad_norm": 2.4530458450317383, "learning_rate": 4.649916416658212e-05, "loss": 0.0358, "step": 54620 }, { "grad_norm": 0.3354771137237549, "learning_rate": 4.648267023928455e-05, "loss": 0.0248, "step": 54630 }, { "grad_norm": 0.5148501992225647, "learning_rate": 4.646617669663687e-05, "loss": 0.0225, "step": 54640 }, { "grad_norm": 0.6311168074607849, "learning_rate": 4.644968354044277e-05, "loss": 0.0224, "step": 54650 }, { "grad_norm": 0.3987295627593994, "learning_rate": 4.643319077250596e-05, "loss": 0.0258, "step": 54660 }, { "grad_norm": 0.24553564190864563, "learning_rate": 4.6416698394630055e-05, "loss": 0.0249, "step": 54670 }, { "grad_norm": 0.3426618278026581, "learning_rate": 4.640020640861862e-05, "loss": 0.0231, "step": 54680 }, { "grad_norm": 0.32459762692451477, "learning_rate": 4.6383714816275195e-05, "loss": 0.0275, "step": 54690 }, { "grad_norm": 0.34082743525505066, "learning_rate": 4.636722361940329e-05, "loss": 0.0231, "step": 54700 }, { "grad_norm": 0.36755624413490295, "learning_rate": 4.635073281980632e-05, "loss": 0.0298, "step": 54710 }, { "grad_norm": 0.3296370208263397, "learning_rate": 4.633424241928775e-05, "loss": 0.0254, "step": 54720 }, { "grad_norm": 0.2861193120479584, "learning_rate": 4.6317752419650914e-05, "loss": 0.0186, "step": 54730 }, { "grad_norm": 0.25659382343292236, "learning_rate": 4.630126282269914e-05, "loss": 0.0261, "step": 54740 }, { "grad_norm": 0.5205438733100891, "learning_rate": 4.628477363023569e-05, "loss": 0.023, "step": 54750 }, { "grad_norm": 0.3951779007911682, "learning_rate": 4.626828484406383e-05, "loss": 0.0271, "step": 54760 }, { "grad_norm": 0.3197830021381378, "learning_rate": 4.625179646598669e-05, "loss": 0.0269, "step": 54770 }, { "grad_norm": 0.4989904463291168, "learning_rate": 4.6235308497807487e-05, "loss": 0.0319, "step": 54780 }, { "grad_norm": 0.3141191303730011, "learning_rate": 4.6218820941329276e-05, "loss": 0.0241, "step": 54790 }, { "grad_norm": 0.23714715242385864, "learning_rate": 4.620233379835513e-05, "loss": 0.0213, "step": 54800 }, { "grad_norm": 0.23553171753883362, "learning_rate": 4.618584707068804e-05, "loss": 0.0235, "step": 54810 }, { "grad_norm": 0.6666156649589539, "learning_rate": 4.616936076013099e-05, "loss": 0.0272, "step": 54820 }, { "grad_norm": 0.280750036239624, "learning_rate": 4.615287486848688e-05, "loss": 0.0328, "step": 54830 }, { "grad_norm": 0.402616024017334, "learning_rate": 4.6136389397558584e-05, "loss": 0.0221, "step": 54840 }, { "grad_norm": 0.29793664813041687, "learning_rate": 4.6119904349148935e-05, "loss": 0.0198, "step": 54850 }, { "grad_norm": 1.0763823986053467, "learning_rate": 4.6103419725060726e-05, "loss": 0.0285, "step": 54860 }, { "grad_norm": 0.2391456514596939, "learning_rate": 4.608693552709667e-05, "loss": 0.0196, "step": 54870 }, { "grad_norm": 0.5041871070861816, "learning_rate": 4.607045175705947e-05, "loss": 0.0255, "step": 54880 }, { "grad_norm": 0.3538303077220917, "learning_rate": 4.605396841675176e-05, "loss": 0.0259, "step": 54890 }, { "grad_norm": 0.638930082321167, "learning_rate": 4.6037485507976094e-05, "loss": 0.0272, "step": 54900 }, { "grad_norm": 1.1942377090454102, "learning_rate": 4.602100303253509e-05, "loss": 0.0227, "step": 54910 }, { "grad_norm": 0.3008846640586853, "learning_rate": 4.600452099223122e-05, "loss": 0.0275, "step": 54920 }, { "grad_norm": 0.3877730071544647, "learning_rate": 4.598803938886691e-05, "loss": 0.0199, "step": 54930 }, { "grad_norm": 0.22068382799625397, "learning_rate": 4.597155822424459e-05, "loss": 0.0234, "step": 54940 }, { "grad_norm": 0.6335996389389038, "learning_rate": 4.59550775001666e-05, "loss": 0.0248, "step": 54950 }, { "grad_norm": 0.35007551312446594, "learning_rate": 4.5938597218435234e-05, "loss": 0.0254, "step": 54960 }, { "grad_norm": 0.2871439456939697, "learning_rate": 4.592211738085279e-05, "loss": 0.0317, "step": 54970 }, { "grad_norm": 0.23730874061584473, "learning_rate": 4.590563798922146e-05, "loss": 0.0202, "step": 54980 }, { "grad_norm": 0.2859085202217102, "learning_rate": 4.5889159045343404e-05, "loss": 0.0268, "step": 54990 }, { "grad_norm": 0.39400187134742737, "learning_rate": 4.5872680551020734e-05, "loss": 0.0254, "step": 55000 }, { "grad_norm": 0.47128546237945557, "learning_rate": 4.585620250805552e-05, "loss": 0.0209, "step": 55010 }, { "grad_norm": 0.4615882337093353, "learning_rate": 4.583972491824975e-05, "loss": 0.0244, "step": 55020 }, { "grad_norm": 0.2894386053085327, "learning_rate": 4.582324778340541e-05, "loss": 0.0234, "step": 55030 }, { "grad_norm": 0.27132928371429443, "learning_rate": 4.5806771105324424e-05, "loss": 0.0258, "step": 55040 }, { "grad_norm": 0.21872016787528992, "learning_rate": 4.579029488580865e-05, "loss": 0.0248, "step": 55050 }, { "grad_norm": 0.24664205312728882, "learning_rate": 4.5773819126659876e-05, "loss": 0.0186, "step": 55060 }, { "grad_norm": 0.32056450843811035, "learning_rate": 4.575734382967991e-05, "loss": 0.0196, "step": 55070 }, { "grad_norm": 0.3271632492542267, "learning_rate": 4.574086899667044e-05, "loss": 0.023, "step": 55080 }, { "grad_norm": 0.2299843430519104, "learning_rate": 4.5724394629433106e-05, "loss": 0.0269, "step": 55090 }, { "grad_norm": 0.22770629823207855, "learning_rate": 4.570792072976958e-05, "loss": 0.0269, "step": 55100 }, { "grad_norm": 0.26049306988716125, "learning_rate": 4.569144729948138e-05, "loss": 0.0268, "step": 55110 }, { "grad_norm": 0.2817751467227936, "learning_rate": 4.5674974340370033e-05, "loss": 0.0193, "step": 55120 }, { "grad_norm": 0.264346718788147, "learning_rate": 4.565850185423699e-05, "loss": 0.0213, "step": 55130 }, { "grad_norm": 0.250666081905365, "learning_rate": 4.564202984288365e-05, "loss": 0.0275, "step": 55140 }, { "grad_norm": 0.3256707489490509, "learning_rate": 4.5625558308111356e-05, "loss": 0.0216, "step": 55150 }, { "grad_norm": 0.3717881441116333, "learning_rate": 4.560908725172145e-05, "loss": 0.0198, "step": 55160 }, { "grad_norm": 0.5091449618339539, "learning_rate": 4.559261667551516e-05, "loss": 0.0238, "step": 55170 }, { "grad_norm": 0.2213786095380783, "learning_rate": 4.5576146581293685e-05, "loss": 0.026, "step": 55180 }, { "grad_norm": 0.3604004681110382, "learning_rate": 4.5559676970858164e-05, "loss": 0.0198, "step": 55190 }, { "grad_norm": 0.5107541084289551, "learning_rate": 4.5543207846009705e-05, "loss": 0.0311, "step": 55200 }, { "grad_norm": 0.2676074206829071, "learning_rate": 4.552673920854933e-05, "loss": 0.0292, "step": 55210 }, { "grad_norm": 0.7687284350395203, "learning_rate": 4.5510271060278006e-05, "loss": 0.0234, "step": 55220 }, { "grad_norm": 0.3207551836967468, "learning_rate": 4.5493803402996704e-05, "loss": 0.0331, "step": 55230 }, { "grad_norm": 0.32689517736434937, "learning_rate": 4.5477336238506295e-05, "loss": 0.0253, "step": 55240 }, { "grad_norm": 0.39077028632164, "learning_rate": 4.546086956860759e-05, "loss": 0.0193, "step": 55250 }, { "grad_norm": 0.24957925081253052, "learning_rate": 4.5444403395101374e-05, "loss": 0.0198, "step": 55260 }, { "grad_norm": 0.3743310868740082, "learning_rate": 4.542793771978834e-05, "loss": 0.0294, "step": 55270 }, { "grad_norm": 0.3808307647705078, "learning_rate": 4.541147254446916e-05, "loss": 0.0289, "step": 55280 }, { "grad_norm": 0.22474437952041626, "learning_rate": 4.539500787094445e-05, "loss": 0.0333, "step": 55290 }, { "grad_norm": 0.3191453516483307, "learning_rate": 4.537854370101475e-05, "loss": 0.0279, "step": 55300 }, { "grad_norm": 0.7938529253005981, "learning_rate": 4.5362080036480566e-05, "loss": 0.0234, "step": 55310 }, { "grad_norm": 0.30070364475250244, "learning_rate": 4.534561687914235e-05, "loss": 0.022, "step": 55320 }, { "grad_norm": 0.6484445333480835, "learning_rate": 4.5329154230800466e-05, "loss": 0.0222, "step": 55330 }, { "grad_norm": 0.30145782232284546, "learning_rate": 4.531269209325525e-05, "loss": 0.0211, "step": 55340 }, { "grad_norm": 0.3846066892147064, "learning_rate": 4.529623046830696e-05, "loss": 0.0202, "step": 55350 }, { "grad_norm": 0.4198438227176666, "learning_rate": 4.5279769357755866e-05, "loss": 0.0269, "step": 55360 }, { "grad_norm": 0.30700185894966125, "learning_rate": 4.5263308763402084e-05, "loss": 0.0153, "step": 55370 }, { "grad_norm": 0.24031572043895721, "learning_rate": 4.524684868704574e-05, "loss": 0.0264, "step": 55380 }, { "grad_norm": 0.27038824558258057, "learning_rate": 4.523038913048688e-05, "loss": 0.0212, "step": 55390 }, { "grad_norm": 0.2743456959724426, "learning_rate": 4.521393009552549e-05, "loss": 0.0174, "step": 55400 }, { "grad_norm": 0.3138200044631958, "learning_rate": 4.5197471583961484e-05, "loss": 0.0272, "step": 55410 }, { "grad_norm": 0.2484888732433319, "learning_rate": 4.51810135975948e-05, "loss": 0.0313, "step": 55420 }, { "grad_norm": 0.31170204281806946, "learning_rate": 4.516455613822521e-05, "loss": 0.0251, "step": 55430 }, { "grad_norm": 0.3317287266254425, "learning_rate": 4.514809920765249e-05, "loss": 0.0225, "step": 55440 }, { "grad_norm": 0.41674232482910156, "learning_rate": 4.5131642807676356e-05, "loss": 0.0289, "step": 55450 }, { "grad_norm": 0.2592191994190216, "learning_rate": 4.511518694009644e-05, "loss": 0.0227, "step": 55460 }, { "grad_norm": 0.31003740429878235, "learning_rate": 4.509873160671231e-05, "loss": 0.0239, "step": 55470 }, { "grad_norm": 0.34518420696258545, "learning_rate": 4.508227680932356e-05, "loss": 0.0234, "step": 55480 }, { "grad_norm": 0.5957788825035095, "learning_rate": 4.5065822549729605e-05, "loss": 0.0315, "step": 55490 }, { "grad_norm": 0.5520884990692139, "learning_rate": 4.504936882972989e-05, "loss": 0.0258, "step": 55500 }, { "grad_norm": 0.5117770433425903, "learning_rate": 4.503291565112374e-05, "loss": 0.0218, "step": 55510 }, { "grad_norm": 0.45874181389808655, "learning_rate": 4.501646301571048e-05, "loss": 0.0232, "step": 55520 }, { "grad_norm": 0.32044491171836853, "learning_rate": 4.500001092528933e-05, "loss": 0.0316, "step": 55530 }, { "grad_norm": 0.33653122186660767, "learning_rate": 4.4983559381659455e-05, "loss": 0.0322, "step": 55540 }, { "grad_norm": 0.4029584527015686, "learning_rate": 4.496710838661999e-05, "loss": 0.0227, "step": 55550 }, { "grad_norm": 0.2179957777261734, "learning_rate": 4.495065794196999e-05, "loss": 0.0273, "step": 55560 }, { "grad_norm": 0.270740807056427, "learning_rate": 4.4934208049508455e-05, "loss": 0.023, "step": 55570 }, { "grad_norm": 0.46196871995925903, "learning_rate": 4.4917758711034303e-05, "loss": 0.031, "step": 55580 }, { "grad_norm": 0.2384748011827469, "learning_rate": 4.490130992834642e-05, "loss": 0.022, "step": 55590 }, { "grad_norm": 0.439532071352005, "learning_rate": 4.488486170324359e-05, "loss": 0.0321, "step": 55600 }, { "grad_norm": 0.39399105310440063, "learning_rate": 4.486841403752463e-05, "loss": 0.0219, "step": 55610 }, { "grad_norm": 0.24127303063869476, "learning_rate": 4.4851966932988174e-05, "loss": 0.0178, "step": 55620 }, { "grad_norm": 0.30441123247146606, "learning_rate": 4.4835520391432894e-05, "loss": 0.026, "step": 55630 }, { "grad_norm": 0.3346301019191742, "learning_rate": 4.481907441465733e-05, "loss": 0.0271, "step": 55640 }, { "grad_norm": 0.47515976428985596, "learning_rate": 4.480262900446e-05, "loss": 0.0354, "step": 55650 }, { "grad_norm": 0.4916629493236542, "learning_rate": 4.478618416263933e-05, "loss": 0.0252, "step": 55660 }, { "grad_norm": 0.32625389099121094, "learning_rate": 4.476973989099373e-05, "loss": 0.0233, "step": 55670 }, { "grad_norm": 0.47165095806121826, "learning_rate": 4.475329619132153e-05, "loss": 0.0301, "step": 55680 }, { "grad_norm": 0.26875054836273193, "learning_rate": 4.4736853065420956e-05, "loss": 0.0269, "step": 55690 }, { "grad_norm": 0.27405959367752075, "learning_rate": 4.472041051509023e-05, "loss": 0.0227, "step": 55700 }, { "grad_norm": 0.2992287576198578, "learning_rate": 4.470396854212746e-05, "loss": 0.0221, "step": 55710 }, { "grad_norm": 0.2618907392024994, "learning_rate": 4.468752714833072e-05, "loss": 0.0216, "step": 55720 }, { "grad_norm": 0.8705679178237915, "learning_rate": 4.467108633549803e-05, "loss": 0.0216, "step": 55730 }, { "grad_norm": 0.26105061173439026, "learning_rate": 4.4654646105427335e-05, "loss": 0.0275, "step": 55740 }, { "grad_norm": 0.3716176152229309, "learning_rate": 4.463820645991651e-05, "loss": 0.0217, "step": 55750 }, { "grad_norm": 0.36600199341773987, "learning_rate": 4.462176740076334e-05, "loss": 0.0247, "step": 55760 }, { "grad_norm": 0.3062003552913666, "learning_rate": 4.4605328929765624e-05, "loss": 0.0306, "step": 55770 }, { "grad_norm": 0.36793550848960876, "learning_rate": 4.458889104872102e-05, "loss": 0.0232, "step": 55780 }, { "grad_norm": 0.43895816802978516, "learning_rate": 4.457245375942715e-05, "loss": 0.0283, "step": 55790 }, { "grad_norm": 0.41224899888038635, "learning_rate": 4.455601706368157e-05, "loss": 0.0249, "step": 55800 }, { "grad_norm": 0.42173510789871216, "learning_rate": 4.453958096328179e-05, "loss": 0.0238, "step": 55810 }, { "grad_norm": 0.29960381984710693, "learning_rate": 4.452314546002523e-05, "loss": 0.0282, "step": 55820 }, { "grad_norm": 0.2893340289592743, "learning_rate": 4.4506710555709246e-05, "loss": 0.0248, "step": 55830 }, { "grad_norm": 0.2539817988872528, "learning_rate": 4.449027625213114e-05, "loss": 0.0253, "step": 55840 }, { "grad_norm": 0.25728556513786316, "learning_rate": 4.4473842551088135e-05, "loss": 0.022, "step": 55850 }, { "grad_norm": 0.2627381384372711, "learning_rate": 4.445740945437738e-05, "loss": 0.0221, "step": 55860 }, { "grad_norm": 0.2755461633205414, "learning_rate": 4.444097696379602e-05, "loss": 0.0239, "step": 55870 }, { "grad_norm": 0.32155388593673706, "learning_rate": 4.442454508114106e-05, "loss": 0.0232, "step": 55880 }, { "grad_norm": 0.2603565454483032, "learning_rate": 4.4408113808209465e-05, "loss": 0.0223, "step": 55890 }, { "grad_norm": 0.2415756732225418, "learning_rate": 4.439168314679813e-05, "loss": 0.0261, "step": 55900 }, { "grad_norm": 0.2614223062992096, "learning_rate": 4.43752530987039e-05, "loss": 0.0209, "step": 55910 }, { "grad_norm": 0.36158064007759094, "learning_rate": 4.4358823665723515e-05, "loss": 0.0209, "step": 55920 }, { "grad_norm": 0.43309271335601807, "learning_rate": 4.434239484965371e-05, "loss": 0.0228, "step": 55930 }, { "grad_norm": 0.23561306297779083, "learning_rate": 4.4325966652291103e-05, "loss": 0.0214, "step": 55940 }, { "grad_norm": 1.2029904127120972, "learning_rate": 4.430953907543225e-05, "loss": 0.0221, "step": 55950 }, { "grad_norm": 0.3412073850631714, "learning_rate": 4.4293112120873645e-05, "loss": 0.0234, "step": 55960 }, { "grad_norm": 0.3237887918949127, "learning_rate": 4.427668579041172e-05, "loss": 0.0204, "step": 55970 }, { "grad_norm": 0.3169238269329071, "learning_rate": 4.4260260085842816e-05, "loss": 0.0249, "step": 55980 }, { "grad_norm": 0.21078339219093323, "learning_rate": 4.4243835008963256e-05, "loss": 0.0203, "step": 55990 }, { "grad_norm": 0.4635917842388153, "learning_rate": 4.422741056156925e-05, "loss": 0.0184, "step": 56000 }, { "grad_norm": 0.2065586894750595, "learning_rate": 4.421098674545693e-05, "loss": 0.0248, "step": 56010 }, { "grad_norm": 0.42351034283638, "learning_rate": 4.419456356242241e-05, "loss": 0.0267, "step": 56020 }, { "grad_norm": 0.277812123298645, "learning_rate": 4.4178141014261685e-05, "loss": 0.0306, "step": 56030 }, { "grad_norm": 0.2689896821975708, "learning_rate": 4.416171910277071e-05, "loss": 0.0241, "step": 56040 }, { "grad_norm": 0.41976335644721985, "learning_rate": 4.414529782974533e-05, "loss": 0.0253, "step": 56050 }, { "grad_norm": 0.3019268810749054, "learning_rate": 4.41288771969814e-05, "loss": 0.0275, "step": 56060 }, { "grad_norm": 0.4422752559185028, "learning_rate": 4.411245720627464e-05, "loss": 0.0287, "step": 56070 }, { "grad_norm": 0.2598419785499573, "learning_rate": 4.40960378594207e-05, "loss": 0.0276, "step": 56080 }, { "grad_norm": 0.2899826169013977, "learning_rate": 4.407961915821519e-05, "loss": 0.0229, "step": 56090 }, { "grad_norm": 0.49211961030960083, "learning_rate": 4.406320110445362e-05, "loss": 0.0266, "step": 56100 }, { "grad_norm": 0.46177762746810913, "learning_rate": 4.404678369993144e-05, "loss": 0.0239, "step": 56110 }, { "grad_norm": 0.27707356214523315, "learning_rate": 4.403036694644406e-05, "loss": 0.0223, "step": 56120 }, { "grad_norm": 0.3585452437400818, "learning_rate": 4.4013950845786764e-05, "loss": 0.0228, "step": 56130 }, { "grad_norm": 0.4929976761341095, "learning_rate": 4.399753539975482e-05, "loss": 0.0279, "step": 56140 }, { "grad_norm": 0.40176498889923096, "learning_rate": 4.398112061014337e-05, "loss": 0.0212, "step": 56150 }, { "grad_norm": 1.3844999074935913, "learning_rate": 4.396470647874753e-05, "loss": 0.0284, "step": 56160 }, { "grad_norm": 0.3499948978424072, "learning_rate": 4.394829300736229e-05, "loss": 0.0234, "step": 56170 }, { "grad_norm": 0.2916189730167389, "learning_rate": 4.393188019778265e-05, "loss": 0.0241, "step": 56180 }, { "grad_norm": 1.1133203506469727, "learning_rate": 4.391546805180347e-05, "loss": 0.0248, "step": 56190 }, { "grad_norm": 0.3274831473827362, "learning_rate": 4.389905657121955e-05, "loss": 0.0214, "step": 56200 }, { "grad_norm": 0.3656029999256134, "learning_rate": 4.3882645757825635e-05, "loss": 0.0262, "step": 56210 }, { "grad_norm": 0.2965990900993347, "learning_rate": 4.386623561341637e-05, "loss": 0.0237, "step": 56220 }, { "grad_norm": 0.5144520998001099, "learning_rate": 4.384982613978637e-05, "loss": 0.0269, "step": 56230 }, { "grad_norm": 0.4465181827545166, "learning_rate": 4.383341733873012e-05, "loss": 0.0229, "step": 56240 }, { "grad_norm": 0.5565459728240967, "learning_rate": 4.381700921204209e-05, "loss": 0.0298, "step": 56250 }, { "grad_norm": 0.4945765435695648, "learning_rate": 4.380060176151663e-05, "loss": 0.0248, "step": 56260 }, { "grad_norm": 0.41088753938674927, "learning_rate": 4.378419498894805e-05, "loss": 0.0267, "step": 56270 }, { "grad_norm": 0.43583613634109497, "learning_rate": 4.376778889613056e-05, "loss": 0.0244, "step": 56280 }, { "grad_norm": 0.3358362913131714, "learning_rate": 4.375138348485831e-05, "loss": 0.0263, "step": 56290 }, { "grad_norm": 0.2978234589099884, "learning_rate": 4.3734978756925346e-05, "loss": 0.0178, "step": 56300 }, { "grad_norm": 0.16819125413894653, "learning_rate": 4.3718574714125706e-05, "loss": 0.0219, "step": 56310 }, { "grad_norm": 0.5069205164909363, "learning_rate": 4.370217135825329e-05, "loss": 0.0276, "step": 56320 }, { "grad_norm": 0.24214990437030792, "learning_rate": 4.368576869110194e-05, "loss": 0.019, "step": 56330 }, { "grad_norm": 0.23217277228832245, "learning_rate": 4.366936671446544e-05, "loss": 0.0213, "step": 56340 }, { "grad_norm": 0.31942299008369446, "learning_rate": 4.3652965430137474e-05, "loss": 0.0218, "step": 56350 }, { "grad_norm": 0.30580782890319824, "learning_rate": 4.3636564839911646e-05, "loss": 0.0258, "step": 56360 }, { "grad_norm": 0.2057456076145172, "learning_rate": 4.3620164945581545e-05, "loss": 0.0162, "step": 56370 }, { "grad_norm": 0.313071608543396, "learning_rate": 4.360376574894061e-05, "loss": 0.0243, "step": 56380 }, { "grad_norm": 0.2748842239379883, "learning_rate": 4.358736725178224e-05, "loss": 0.021, "step": 56390 }, { "grad_norm": 0.38704368472099304, "learning_rate": 4.357096945589974e-05, "loss": 0.0256, "step": 56400 }, { "grad_norm": 0.291404128074646, "learning_rate": 4.355457236308636e-05, "loss": 0.023, "step": 56410 }, { "grad_norm": 0.24966645240783691, "learning_rate": 4.353817597513526e-05, "loss": 0.0172, "step": 56420 }, { "grad_norm": 0.2702038288116455, "learning_rate": 4.352178029383948e-05, "loss": 0.017, "step": 56430 }, { "grad_norm": 0.41211873292922974, "learning_rate": 4.3505385320992105e-05, "loss": 0.025, "step": 56440 }, { "grad_norm": 0.25245201587677, "learning_rate": 4.348899105838602e-05, "loss": 0.0176, "step": 56450 }, { "grad_norm": 0.2674057185649872, "learning_rate": 4.3472597507814087e-05, "loss": 0.023, "step": 56460 }, { "grad_norm": 0.43717673420906067, "learning_rate": 4.3456204671069066e-05, "loss": 0.0258, "step": 56470 }, { "grad_norm": 0.3226078748703003, "learning_rate": 4.343981254994367e-05, "loss": 0.0243, "step": 56480 }, { "grad_norm": 0.3246747553348541, "learning_rate": 4.34234211462305e-05, "loss": 0.0268, "step": 56490 }, { "grad_norm": 0.3234120011329651, "learning_rate": 4.340703046172213e-05, "loss": 0.0279, "step": 56500 }, { "grad_norm": 1.7237439155578613, "learning_rate": 4.339064049821097e-05, "loss": 0.0248, "step": 56510 }, { "grad_norm": 0.3313162326812744, "learning_rate": 4.3374251257489446e-05, "loss": 0.0238, "step": 56520 }, { "grad_norm": 0.29852399230003357, "learning_rate": 4.335786274134984e-05, "loss": 0.0227, "step": 56530 }, { "grad_norm": 0.3723946213722229, "learning_rate": 4.334147495158438e-05, "loss": 0.0285, "step": 56540 }, { "grad_norm": 0.36328136920928955, "learning_rate": 4.33250878899852e-05, "loss": 0.0238, "step": 56550 }, { "grad_norm": 0.8273773193359375, "learning_rate": 4.330870155834437e-05, "loss": 0.0227, "step": 56560 }, { "grad_norm": 0.3214995265007019, "learning_rate": 4.3292315958453884e-05, "loss": 0.0226, "step": 56570 }, { "grad_norm": 0.34824442863464355, "learning_rate": 4.3275931092105646e-05, "loss": 0.0172, "step": 56580 }, { "grad_norm": 0.2687237858772278, "learning_rate": 4.325954696109147e-05, "loss": 0.0193, "step": 56590 }, { "grad_norm": 0.21995729207992554, "learning_rate": 4.32431635672031e-05, "loss": 0.0239, "step": 56600 }, { "grad_norm": 0.31850525736808777, "learning_rate": 4.3226780912232214e-05, "loss": 0.0298, "step": 56610 }, { "grad_norm": 0.34007763862609863, "learning_rate": 4.321039899797036e-05, "loss": 0.0195, "step": 56620 }, { "grad_norm": 0.34603652358055115, "learning_rate": 4.319401782620908e-05, "loss": 0.0181, "step": 56630 }, { "grad_norm": 0.286312073469162, "learning_rate": 4.317763739873978e-05, "loss": 0.0255, "step": 56640 }, { "grad_norm": 0.8331287503242493, "learning_rate": 4.31612577173538e-05, "loss": 0.0249, "step": 56650 }, { "grad_norm": 0.328950971364975, "learning_rate": 4.314487878384239e-05, "loss": 0.0314, "step": 56660 }, { "grad_norm": 0.41459351778030396, "learning_rate": 4.3128500599996736e-05, "loss": 0.0259, "step": 56670 }, { "grad_norm": 0.2546883523464203, "learning_rate": 4.31121231676079e-05, "loss": 0.0184, "step": 56680 }, { "grad_norm": 0.4083893597126007, "learning_rate": 4.309574648846694e-05, "loss": 0.0267, "step": 56690 }, { "grad_norm": 0.3386150300502777, "learning_rate": 4.3079370564364755e-05, "loss": 0.0293, "step": 56700 }, { "grad_norm": 0.38121357560157776, "learning_rate": 4.306299539709221e-05, "loss": 0.0249, "step": 56710 }, { "grad_norm": 0.26876285672187805, "learning_rate": 4.3046620988440046e-05, "loss": 0.0264, "step": 56720 }, { "grad_norm": 0.35951194167137146, "learning_rate": 4.303024734019897e-05, "loss": 0.0168, "step": 56730 }, { "grad_norm": 0.20854972302913666, "learning_rate": 4.3013874454159565e-05, "loss": 0.0215, "step": 56740 }, { "grad_norm": 0.2584090232849121, "learning_rate": 4.299750233211233e-05, "loss": 0.0284, "step": 56750 }, { "grad_norm": 0.3258535861968994, "learning_rate": 4.2981130975847715e-05, "loss": 0.0207, "step": 56760 }, { "grad_norm": 0.3748093843460083, "learning_rate": 4.2964760387156076e-05, "loss": 0.0259, "step": 56770 }, { "grad_norm": 0.25939682126045227, "learning_rate": 4.294839056782767e-05, "loss": 0.0226, "step": 56780 }, { "grad_norm": 0.25396984815597534, "learning_rate": 4.293202151965266e-05, "loss": 0.0238, "step": 56790 }, { "grad_norm": 0.38508328795433044, "learning_rate": 4.291565324442115e-05, "loss": 0.0248, "step": 56800 }, { "grad_norm": 0.3491985499858856, "learning_rate": 4.2899285743923135e-05, "loss": 0.0255, "step": 56810 }, { "grad_norm": 0.29364848136901855, "learning_rate": 4.288291901994857e-05, "loss": 0.0222, "step": 56820 }, { "grad_norm": 0.36539018154144287, "learning_rate": 4.286655307428729e-05, "loss": 0.0221, "step": 56830 }, { "grad_norm": 0.353546679019928, "learning_rate": 4.285018790872903e-05, "loss": 0.0252, "step": 56840 }, { "grad_norm": 0.3679234981536865, "learning_rate": 4.283382352506347e-05, "loss": 0.0269, "step": 56850 }, { "grad_norm": 0.3622516393661499, "learning_rate": 4.2817459925080205e-05, "loss": 0.0291, "step": 56860 }, { "grad_norm": 0.3505434989929199, "learning_rate": 4.28010971105687e-05, "loss": 0.0265, "step": 56870 }, { "grad_norm": 0.3827276825904846, "learning_rate": 4.278473508331841e-05, "loss": 0.0202, "step": 56880 }, { "grad_norm": 0.39708411693573, "learning_rate": 4.276837384511864e-05, "loss": 0.029, "step": 56890 }, { "grad_norm": 0.45438089966773987, "learning_rate": 4.275201339775864e-05, "loss": 0.0167, "step": 56900 }, { "grad_norm": 0.2611086368560791, "learning_rate": 4.273565374302756e-05, "loss": 0.0239, "step": 56910 }, { "grad_norm": 0.25942638516426086, "learning_rate": 4.2719294882714454e-05, "loss": 0.0199, "step": 56920 }, { "grad_norm": 0.3141327500343323, "learning_rate": 4.270293681860831e-05, "loss": 0.024, "step": 56930 }, { "grad_norm": 0.20196372270584106, "learning_rate": 4.268657955249802e-05, "loss": 0.0182, "step": 56940 }, { "grad_norm": 0.2780138850212097, "learning_rate": 4.267022308617241e-05, "loss": 0.0205, "step": 56950 }, { "grad_norm": 0.3808530271053314, "learning_rate": 4.265386742142018e-05, "loss": 0.0215, "step": 56960 }, { "grad_norm": 0.44532424211502075, "learning_rate": 4.263751256002996e-05, "loss": 0.0257, "step": 56970 }, { "grad_norm": 0.3639109432697296, "learning_rate": 4.26211585037903e-05, "loss": 0.0274, "step": 56980 }, { "grad_norm": 0.39170515537261963, "learning_rate": 4.2604805254489656e-05, "loss": 0.0268, "step": 56990 }, { "grad_norm": 0.3140137493610382, "learning_rate": 4.258845281391638e-05, "loss": 0.0234, "step": 57000 }, { "grad_norm": 0.27715805172920227, "learning_rate": 4.257210118385876e-05, "loss": 0.0379, "step": 57010 }, { "grad_norm": 0.30938565731048584, "learning_rate": 4.2555750366104994e-05, "loss": 0.0228, "step": 57020 }, { "grad_norm": 0.1978202909231186, "learning_rate": 4.2539400362443184e-05, "loss": 0.0182, "step": 57030 }, { "grad_norm": 0.31405770778656006, "learning_rate": 4.252305117466133e-05, "loss": 0.0258, "step": 57040 }, { "grad_norm": 0.4516547620296478, "learning_rate": 4.2506702804547356e-05, "loss": 0.027, "step": 57050 }, { "grad_norm": 0.363720178604126, "learning_rate": 4.249035525388907e-05, "loss": 0.027, "step": 57060 }, { "grad_norm": 0.3113609850406647, "learning_rate": 4.247400852447427e-05, "loss": 0.0185, "step": 57070 }, { "grad_norm": 0.38662517070770264, "learning_rate": 4.245766261809059e-05, "loss": 0.0205, "step": 57080 }, { "grad_norm": 0.20851491391658783, "learning_rate": 4.244131753652558e-05, "loss": 0.022, "step": 57090 }, { "grad_norm": 0.25543564558029175, "learning_rate": 4.242497328156672e-05, "loss": 0.0203, "step": 57100 }, { "grad_norm": 0.33178675174713135, "learning_rate": 4.240862985500139e-05, "loss": 0.0235, "step": 57110 }, { "grad_norm": 0.2883821129798889, "learning_rate": 4.2392287258616894e-05, "loss": 0.013, "step": 57120 }, { "grad_norm": 0.2640477418899536, "learning_rate": 4.2375945494200396e-05, "loss": 0.0173, "step": 57130 }, { "grad_norm": 0.31990140676498413, "learning_rate": 4.2359604563539054e-05, "loss": 0.0192, "step": 57140 }, { "grad_norm": 0.38367170095443726, "learning_rate": 4.2343264468419885e-05, "loss": 0.0234, "step": 57150 }, { "grad_norm": 0.43541860580444336, "learning_rate": 4.232692521062979e-05, "loss": 0.0201, "step": 57160 }, { "grad_norm": 0.3641285300254822, "learning_rate": 4.231058679195562e-05, "loss": 0.0227, "step": 57170 }, { "grad_norm": 0.29092004895210266, "learning_rate": 4.229424921418411e-05, "loss": 0.0182, "step": 57180 }, { "grad_norm": 0.2919851839542389, "learning_rate": 4.2277912479101916e-05, "loss": 0.0319, "step": 57190 }, { "grad_norm": 0.2605898082256317, "learning_rate": 4.226157658849562e-05, "loss": 0.0225, "step": 57200 }, { "grad_norm": 0.3906589448451996, "learning_rate": 4.2245241544151664e-05, "loss": 0.0296, "step": 57210 }, { "grad_norm": 0.3594992160797119, "learning_rate": 4.2228907347856425e-05, "loss": 0.0217, "step": 57220 }, { "grad_norm": 0.3673996031284332, "learning_rate": 4.2212574001396214e-05, "loss": 0.025, "step": 57230 }, { "grad_norm": 0.30190953612327576, "learning_rate": 4.21962415065572e-05, "loss": 0.0216, "step": 57240 }, { "grad_norm": 0.27872368693351746, "learning_rate": 4.217990986512548e-05, "loss": 0.022, "step": 57250 }, { "grad_norm": 0.30706700682640076, "learning_rate": 4.216357907888704e-05, "loss": 0.0337, "step": 57260 }, { "grad_norm": 0.36053192615509033, "learning_rate": 4.2147249149627824e-05, "loss": 0.0267, "step": 57270 }, { "grad_norm": 0.23561587929725647, "learning_rate": 4.213092007913364e-05, "loss": 0.0277, "step": 57280 }, { "grad_norm": 0.3001936078071594, "learning_rate": 4.2114591869190205e-05, "loss": 0.024, "step": 57290 }, { "grad_norm": 0.3203820586204529, "learning_rate": 4.209826452158315e-05, "loss": 0.0246, "step": 57300 }, { "grad_norm": 0.391361266374588, "learning_rate": 4.2081938038098e-05, "loss": 0.0233, "step": 57310 }, { "grad_norm": 0.2877313494682312, "learning_rate": 4.2065612420520185e-05, "loss": 0.023, "step": 57320 }, { "grad_norm": 0.3108369708061218, "learning_rate": 4.204928767063509e-05, "loss": 0.0208, "step": 57330 }, { "grad_norm": 0.36017248034477234, "learning_rate": 4.2032963790227944e-05, "loss": 0.0227, "step": 57340 }, { "grad_norm": 0.32740795612335205, "learning_rate": 4.2016640781083887e-05, "loss": 0.0259, "step": 57350 }, { "grad_norm": 0.25197121500968933, "learning_rate": 4.2000318644988004e-05, "loss": 0.0331, "step": 57360 }, { "grad_norm": 0.27110254764556885, "learning_rate": 4.198399738372524e-05, "loss": 0.0246, "step": 57370 }, { "grad_norm": 0.21172484755516052, "learning_rate": 4.196767699908045e-05, "loss": 0.0215, "step": 57380 }, { "grad_norm": 0.2938379645347595, "learning_rate": 4.195135749283845e-05, "loss": 0.0201, "step": 57390 }, { "grad_norm": 0.7094495296478271, "learning_rate": 4.1935038866783894e-05, "loss": 0.0222, "step": 57400 }, { "grad_norm": 0.31036412715911865, "learning_rate": 4.1918721122701355e-05, "loss": 0.024, "step": 57410 }, { "grad_norm": 0.3543757200241089, "learning_rate": 4.190240426237533e-05, "loss": 0.0258, "step": 57420 }, { "grad_norm": 1.2722479104995728, "learning_rate": 4.188608828759019e-05, "loss": 0.0196, "step": 57430 }, { "grad_norm": 0.3231036365032196, "learning_rate": 4.1869773200130246e-05, "loss": 0.0329, "step": 57440 }, { "grad_norm": 0.37336471676826477, "learning_rate": 4.185345900177967e-05, "loss": 0.0378, "step": 57450 }, { "grad_norm": 0.2714139521121979, "learning_rate": 4.183714569432258e-05, "loss": 0.0304, "step": 57460 }, { "grad_norm": 0.2782882750034332, "learning_rate": 4.182083327954295e-05, "loss": 0.023, "step": 57470 }, { "grad_norm": 0.24039047956466675, "learning_rate": 4.1804521759224714e-05, "loss": 0.0272, "step": 57480 }, { "grad_norm": 0.40445414185523987, "learning_rate": 4.178821113515165e-05, "loss": 0.0245, "step": 57490 }, { "grad_norm": 0.31544551253318787, "learning_rate": 4.1771901409107476e-05, "loss": 0.0275, "step": 57500 }, { "grad_norm": 0.3812853991985321, "learning_rate": 4.175559258287577e-05, "loss": 0.0242, "step": 57510 }, { "grad_norm": 0.21467141807079315, "learning_rate": 4.173928465824008e-05, "loss": 0.0222, "step": 57520 }, { "grad_norm": 0.24618765711784363, "learning_rate": 4.17229776369838e-05, "loss": 0.0235, "step": 57530 }, { "grad_norm": 0.38248851895332336, "learning_rate": 4.170667152089025e-05, "loss": 0.0334, "step": 57540 }, { "grad_norm": 0.21213547885417938, "learning_rate": 4.1690366311742626e-05, "loss": 0.0186, "step": 57550 }, { "grad_norm": 0.3960159122943878, "learning_rate": 4.1674062011324055e-05, "loss": 0.0193, "step": 57560 }, { "grad_norm": 0.2982766926288605, "learning_rate": 4.165775862141752e-05, "loss": 0.0226, "step": 57570 }, { "grad_norm": 0.42174771428108215, "learning_rate": 4.164145614380598e-05, "loss": 0.0242, "step": 57580 }, { "grad_norm": 0.46027982234954834, "learning_rate": 4.162515458027223e-05, "loss": 0.0247, "step": 57590 }, { "grad_norm": 0.5530217885971069, "learning_rate": 4.160885393259899e-05, "loss": 0.0243, "step": 57600 }, { "grad_norm": 0.4862438440322876, "learning_rate": 4.159255420256886e-05, "loss": 0.0236, "step": 57610 }, { "grad_norm": 0.3274024426937103, "learning_rate": 4.157625539196436e-05, "loss": 0.0265, "step": 57620 }, { "grad_norm": 0.19630642235279083, "learning_rate": 4.155995750256791e-05, "loss": 0.0328, "step": 57630 }, { "grad_norm": 0.21740731596946716, "learning_rate": 4.15436605361618e-05, "loss": 0.017, "step": 57640 }, { "grad_norm": 0.30989012122154236, "learning_rate": 4.152736449452827e-05, "loss": 0.0265, "step": 57650 }, { "grad_norm": 0.4948665201663971, "learning_rate": 4.151106937944943e-05, "loss": 0.0226, "step": 57660 }, { "grad_norm": 0.21110090613365173, "learning_rate": 4.149477519270727e-05, "loss": 0.0223, "step": 57670 }, { "grad_norm": 0.2755574584007263, "learning_rate": 4.147848193608371e-05, "loss": 0.0197, "step": 57680 }, { "grad_norm": 0.3062743842601776, "learning_rate": 4.146218961136056e-05, "loss": 0.0267, "step": 57690 }, { "grad_norm": 0.3482271730899811, "learning_rate": 4.14458982203195e-05, "loss": 0.0289, "step": 57700 }, { "grad_norm": 0.3279834985733032, "learning_rate": 4.1429607764742165e-05, "loss": 0.0231, "step": 57710 }, { "grad_norm": 0.3566415309906006, "learning_rate": 4.141331824641003e-05, "loss": 0.0236, "step": 57720 }, { "grad_norm": 0.26573190093040466, "learning_rate": 4.139702966710452e-05, "loss": 0.0191, "step": 57730 }, { "grad_norm": 0.3440326452255249, "learning_rate": 4.13807420286069e-05, "loss": 0.0225, "step": 57740 }, { "grad_norm": 0.33242812752723694, "learning_rate": 4.136445533269837e-05, "loss": 0.0246, "step": 57750 }, { "grad_norm": 0.2720787823200226, "learning_rate": 4.134816958116003e-05, "loss": 0.0189, "step": 57760 }, { "grad_norm": 0.3608412742614746, "learning_rate": 4.133188477577283e-05, "loss": 0.0206, "step": 57770 }, { "grad_norm": 0.2635359764099121, "learning_rate": 4.131560091831771e-05, "loss": 0.0241, "step": 57780 }, { "grad_norm": 0.31475746631622314, "learning_rate": 4.1299318010575405e-05, "loss": 0.0246, "step": 57790 }, { "grad_norm": 0.41738560795783997, "learning_rate": 4.128303605432661e-05, "loss": 0.0486, "step": 57800 }, { "grad_norm": 0.3896327614784241, "learning_rate": 4.126675505135188e-05, "loss": 0.0189, "step": 57810 }, { "grad_norm": 0.6399316191673279, "learning_rate": 4.125047500343169e-05, "loss": 0.0254, "step": 57820 }, { "grad_norm": 0.27950718998908997, "learning_rate": 4.1234195912346376e-05, "loss": 0.024, "step": 57830 }, { "grad_norm": 0.4110959768295288, "learning_rate": 4.1217917779876235e-05, "loss": 0.0209, "step": 57840 }, { "grad_norm": 0.32771041989326477, "learning_rate": 4.120164060780141e-05, "loss": 0.0259, "step": 57850 }, { "grad_norm": 0.5870562195777893, "learning_rate": 4.1185364397901935e-05, "loss": 0.035, "step": 57860 }, { "grad_norm": 0.23599256575107574, "learning_rate": 4.116908915195776e-05, "loss": 0.0165, "step": 57870 }, { "grad_norm": 0.44447681307792664, "learning_rate": 4.115281487174872e-05, "loss": 0.0305, "step": 57880 }, { "grad_norm": 0.33515945076942444, "learning_rate": 4.113654155905452e-05, "loss": 0.0224, "step": 57890 }, { "grad_norm": 0.23279450833797455, "learning_rate": 4.112026921565483e-05, "loss": 0.0217, "step": 57900 }, { "grad_norm": 0.25436216592788696, "learning_rate": 4.1103997843329147e-05, "loss": 0.023, "step": 57910 }, { "grad_norm": 0.4400067627429962, "learning_rate": 4.108772744385689e-05, "loss": 0.0371, "step": 57920 }, { "grad_norm": 0.2926851809024811, "learning_rate": 4.1071458019017346e-05, "loss": 0.0229, "step": 57930 }, { "grad_norm": 0.24286982417106628, "learning_rate": 4.1055189570589734e-05, "loss": 0.0189, "step": 57940 }, { "grad_norm": 0.32097750902175903, "learning_rate": 4.103892210035315e-05, "loss": 0.0208, "step": 57950 }, { "grad_norm": 0.3248918950557709, "learning_rate": 4.102265561008656e-05, "loss": 0.0265, "step": 57960 }, { "grad_norm": 0.28504472970962524, "learning_rate": 4.100639010156885e-05, "loss": 0.0189, "step": 57970 }, { "grad_norm": 0.350178986787796, "learning_rate": 4.099012557657882e-05, "loss": 0.0208, "step": 57980 }, { "grad_norm": 0.3950600326061249, "learning_rate": 4.09738620368951e-05, "loss": 0.0221, "step": 57990 }, { "grad_norm": 0.2831409275531769, "learning_rate": 4.095759948429626e-05, "loss": 0.0216, "step": 58000 }, { "grad_norm": 0.17794150114059448, "learning_rate": 4.094133792056075e-05, "loss": 0.0185, "step": 58010 }, { "grad_norm": 0.34912317991256714, "learning_rate": 4.092507734746688e-05, "loss": 0.0217, "step": 58020 }, { "grad_norm": 0.30581918358802795, "learning_rate": 4.090881776679293e-05, "loss": 0.0194, "step": 58030 }, { "grad_norm": 0.37937313318252563, "learning_rate": 4.0892559180316994e-05, "loss": 0.0206, "step": 58040 }, { "grad_norm": 0.20879504084587097, "learning_rate": 4.08763015898171e-05, "loss": 0.0157, "step": 58050 }, { "grad_norm": 0.4316106140613556, "learning_rate": 4.0860044997071145e-05, "loss": 0.0227, "step": 58060 }, { "grad_norm": 0.336307168006897, "learning_rate": 4.084378940385693e-05, "loss": 0.022, "step": 58070 }, { "grad_norm": 0.37343937158584595, "learning_rate": 4.082753481195211e-05, "loss": 0.0223, "step": 58080 }, { "grad_norm": 0.29705119132995605, "learning_rate": 4.0811281223134324e-05, "loss": 0.0223, "step": 58090 }, { "grad_norm": 0.7573809027671814, "learning_rate": 4.0795028639181e-05, "loss": 0.0288, "step": 58100 }, { "grad_norm": 0.36081385612487793, "learning_rate": 4.077877706186951e-05, "loss": 0.0236, "step": 58110 }, { "grad_norm": 0.37355145812034607, "learning_rate": 4.076252649297709e-05, "loss": 0.0309, "step": 58120 }, { "grad_norm": 0.39538946747779846, "learning_rate": 4.07462769342809e-05, "loss": 0.0258, "step": 58130 }, { "grad_norm": 0.308226078748703, "learning_rate": 4.073002838755793e-05, "loss": 0.03, "step": 58140 }, { "grad_norm": 0.3499045670032501, "learning_rate": 4.0713780854585115e-05, "loss": 0.0269, "step": 58150 }, { "grad_norm": 0.3984062075614929, "learning_rate": 4.069753433713927e-05, "loss": 0.0264, "step": 58160 }, { "grad_norm": 0.4322405755519867, "learning_rate": 4.068128883699709e-05, "loss": 0.0276, "step": 58170 }, { "grad_norm": 0.22298134863376617, "learning_rate": 4.0665044355935155e-05, "loss": 0.0235, "step": 58180 }, { "grad_norm": 0.3002922236919403, "learning_rate": 4.064880089572993e-05, "loss": 0.0198, "step": 58190 }, { "grad_norm": 0.308009535074234, "learning_rate": 4.063255845815779e-05, "loss": 0.0363, "step": 58200 }, { "grad_norm": 0.3697109520435333, "learning_rate": 4.0616317044994945e-05, "loss": 0.0212, "step": 58210 }, { "grad_norm": 0.2330932319164276, "learning_rate": 4.0600076658017585e-05, "loss": 0.022, "step": 58220 }, { "grad_norm": 0.17105326056480408, "learning_rate": 4.0583837299001706e-05, "loss": 0.024, "step": 58230 }, { "grad_norm": 0.4352327883243561, "learning_rate": 4.056759896972322e-05, "loss": 0.0258, "step": 58240 }, { "grad_norm": 0.35476136207580566, "learning_rate": 4.055136167195793e-05, "loss": 0.0226, "step": 58250 }, { "grad_norm": 0.2947114109992981, "learning_rate": 4.053512540748152e-05, "loss": 0.0241, "step": 58260 }, { "grad_norm": 0.2954957187175751, "learning_rate": 4.051889017806955e-05, "loss": 0.0265, "step": 58270 }, { "grad_norm": 0.41310185194015503, "learning_rate": 4.0502655985497516e-05, "loss": 0.0255, "step": 58280 }, { "grad_norm": 0.3000931739807129, "learning_rate": 4.048642283154074e-05, "loss": 0.024, "step": 58290 }, { "grad_norm": 0.33137765526771545, "learning_rate": 4.0470190717974464e-05, "loss": 0.0243, "step": 58300 }, { "grad_norm": 0.28527960181236267, "learning_rate": 4.045395964657381e-05, "loss": 0.0238, "step": 58310 }, { "grad_norm": 0.6655330657958984, "learning_rate": 4.043772961911376e-05, "loss": 0.0175, "step": 58320 }, { "grad_norm": 0.3030645251274109, "learning_rate": 4.042150063736924e-05, "loss": 0.0255, "step": 58330 }, { "grad_norm": 0.3977842628955841, "learning_rate": 4.0405272703114976e-05, "loss": 0.0215, "step": 58340 }, { "grad_norm": 0.2823117673397064, "learning_rate": 4.038904581812569e-05, "loss": 0.0159, "step": 58350 }, { "grad_norm": 0.32782182097435, "learning_rate": 4.03728199841759e-05, "loss": 0.0212, "step": 58360 }, { "grad_norm": 0.3203345239162445, "learning_rate": 4.0356595203040044e-05, "loss": 0.023, "step": 58370 }, { "grad_norm": 0.24575616419315338, "learning_rate": 4.0340371476492426e-05, "loss": 0.0197, "step": 58380 }, { "grad_norm": 0.45622190833091736, "learning_rate": 4.032414880630727e-05, "loss": 0.0249, "step": 58390 }, { "grad_norm": 0.20813560485839844, "learning_rate": 4.0307927194258646e-05, "loss": 0.0202, "step": 58400 }, { "grad_norm": 0.5913639664649963, "learning_rate": 4.029170664212054e-05, "loss": 0.0221, "step": 58410 }, { "grad_norm": 0.5702532529830933, "learning_rate": 4.0275487151666794e-05, "loss": 0.0218, "step": 58420 }, { "grad_norm": 0.3649020195007324, "learning_rate": 4.025926872467116e-05, "loss": 0.0267, "step": 58430 }, { "grad_norm": 0.28292179107666016, "learning_rate": 4.024305136290725e-05, "loss": 0.0224, "step": 58440 }, { "grad_norm": 0.22291803359985352, "learning_rate": 4.022683506814858e-05, "loss": 0.0275, "step": 58450 }, { "grad_norm": 0.3173506557941437, "learning_rate": 4.021061984216853e-05, "loss": 0.0204, "step": 58460 }, { "grad_norm": 0.3143600821495056, "learning_rate": 4.019440568674035e-05, "loss": 0.0233, "step": 58470 }, { "grad_norm": 0.4129575788974762, "learning_rate": 4.017819260363724e-05, "loss": 0.0242, "step": 58480 }, { "grad_norm": 0.21592183411121368, "learning_rate": 4.016198059463223e-05, "loss": 0.0304, "step": 58490 }, { "grad_norm": 0.30855071544647217, "learning_rate": 4.0145769661498233e-05, "loss": 0.0246, "step": 58500 }, { "grad_norm": 0.21409174799919128, "learning_rate": 4.012955980600804e-05, "loss": 0.0259, "step": 58510 }, { "grad_norm": 0.406318724155426, "learning_rate": 4.0113351029934355e-05, "loss": 0.0247, "step": 58520 }, { "grad_norm": 0.2813192903995514, "learning_rate": 4.0097143335049714e-05, "loss": 0.0228, "step": 58530 }, { "grad_norm": 0.3446716070175171, "learning_rate": 4.0080936723126606e-05, "loss": 0.024, "step": 58540 }, { "grad_norm": 0.3486798107624054, "learning_rate": 4.006473119593735e-05, "loss": 0.0228, "step": 58550 }, { "grad_norm": 0.3454342186450958, "learning_rate": 4.004852675525415e-05, "loss": 0.0217, "step": 58560 }, { "grad_norm": 0.1629638969898224, "learning_rate": 4.00323234028491e-05, "loss": 0.0224, "step": 58570 }, { "grad_norm": 0.307523638010025, "learning_rate": 4.0016121140494186e-05, "loss": 0.0237, "step": 58580 }, { "grad_norm": 0.30266502499580383, "learning_rate": 3.999991996996122e-05, "loss": 0.0216, "step": 58590 }, { "grad_norm": 0.21147802472114563, "learning_rate": 3.9983719893022e-05, "loss": 0.0177, "step": 58600 }, { "grad_norm": 0.2994958162307739, "learning_rate": 3.996752091144812e-05, "loss": 0.0227, "step": 58610 }, { "grad_norm": 0.3361584544181824, "learning_rate": 3.9951323027011055e-05, "loss": 0.0217, "step": 58620 }, { "grad_norm": 0.2586078643798828, "learning_rate": 3.993512624148219e-05, "loss": 0.0222, "step": 58630 }, { "grad_norm": 0.29559609293937683, "learning_rate": 3.99189305566328e-05, "loss": 0.0242, "step": 58640 }, { "grad_norm": 0.3623407781124115, "learning_rate": 3.990273597423401e-05, "loss": 0.0212, "step": 58650 }, { "grad_norm": 0.22178295254707336, "learning_rate": 3.988654249605681e-05, "loss": 0.0181, "step": 58660 }, { "grad_norm": 0.33468568325042725, "learning_rate": 3.987035012387213e-05, "loss": 0.0232, "step": 58670 }, { "grad_norm": 0.29867953062057495, "learning_rate": 3.9854158859450735e-05, "loss": 0.0223, "step": 58680 }, { "grad_norm": 0.3390352129936218, "learning_rate": 3.983796870456328e-05, "loss": 0.0241, "step": 58690 }, { "grad_norm": 0.21781101822853088, "learning_rate": 3.982177966098028e-05, "loss": 0.0205, "step": 58700 }, { "grad_norm": 0.2783890962600708, "learning_rate": 3.980559173047216e-05, "loss": 0.0273, "step": 58710 }, { "grad_norm": 0.35669583082199097, "learning_rate": 3.9789404914809176e-05, "loss": 0.0208, "step": 58720 }, { "grad_norm": 0.1798323094844818, "learning_rate": 3.9773219215761544e-05, "loss": 0.0173, "step": 58730 }, { "grad_norm": 0.3293796479701996, "learning_rate": 3.975703463509928e-05, "loss": 0.0216, "step": 58740 }, { "grad_norm": 0.2356281280517578, "learning_rate": 3.974085117459231e-05, "loss": 0.0221, "step": 58750 }, { "grad_norm": 0.3317803740501404, "learning_rate": 3.972466883601043e-05, "loss": 0.0301, "step": 58760 }, { "grad_norm": 0.32972681522369385, "learning_rate": 3.9708487621123316e-05, "loss": 0.0246, "step": 58770 }, { "grad_norm": 0.5530526041984558, "learning_rate": 3.969230753170051e-05, "loss": 0.0262, "step": 58780 }, { "grad_norm": 0.1444019079208374, "learning_rate": 3.967612856951146e-05, "loss": 0.0217, "step": 58790 }, { "grad_norm": 0.27150097489356995, "learning_rate": 3.9659950736325476e-05, "loss": 0.0209, "step": 58800 }, { "grad_norm": 0.2909390330314636, "learning_rate": 3.964377403391174e-05, "loss": 0.0411, "step": 58810 }, { "grad_norm": 0.3843088150024414, "learning_rate": 3.9627598464039303e-05, "loss": 0.0418, "step": 58820 }, { "grad_norm": 0.31938010454177856, "learning_rate": 3.9611424028477096e-05, "loss": 0.0215, "step": 58830 }, { "grad_norm": 0.26052117347717285, "learning_rate": 3.959525072899394e-05, "loss": 0.0291, "step": 58840 }, { "grad_norm": 0.3173447549343109, "learning_rate": 3.9579078567358506e-05, "loss": 0.0215, "step": 58850 }, { "grad_norm": 0.3527000844478607, "learning_rate": 3.956290754533939e-05, "loss": 0.0242, "step": 58860 }, { "grad_norm": 0.34681010246276855, "learning_rate": 3.954673766470501e-05, "loss": 0.031, "step": 58870 }, { "grad_norm": 0.30843859910964966, "learning_rate": 3.9530568927223674e-05, "loss": 0.0217, "step": 58880 }, { "grad_norm": 0.25307050347328186, "learning_rate": 3.95144013346636e-05, "loss": 0.0248, "step": 58890 }, { "grad_norm": 0.33980467915534973, "learning_rate": 3.949823488879282e-05, "loss": 0.0149, "step": 58900 }, { "grad_norm": 0.37156715989112854, "learning_rate": 3.9482069591379276e-05, "loss": 0.0318, "step": 58910 }, { "grad_norm": 0.25865623354911804, "learning_rate": 3.9465905444190784e-05, "loss": 0.0211, "step": 58920 }, { "grad_norm": 0.3018776774406433, "learning_rate": 3.944974244899505e-05, "loss": 0.0218, "step": 58930 }, { "grad_norm": 0.23187823593616486, "learning_rate": 3.9433580607559616e-05, "loss": 0.0251, "step": 58940 }, { "grad_norm": 0.2599678933620453, "learning_rate": 3.941741992165193e-05, "loss": 0.0309, "step": 58950 }, { "grad_norm": 0.30340468883514404, "learning_rate": 3.940126039303928e-05, "loss": 0.0264, "step": 58960 }, { "grad_norm": 0.6564708352088928, "learning_rate": 3.938510202348886e-05, "loss": 0.0231, "step": 58970 }, { "grad_norm": 0.2976830303668976, "learning_rate": 3.93689448147677e-05, "loss": 0.021, "step": 58980 }, { "grad_norm": 0.2903265357017517, "learning_rate": 3.935278876864278e-05, "loss": 0.0189, "step": 58990 }, { "grad_norm": 0.3070239722728729, "learning_rate": 3.933663388688087e-05, "loss": 0.0359, "step": 59000 }, { "grad_norm": 0.5173068642616272, "learning_rate": 3.9320480171248645e-05, "loss": 0.019, "step": 59010 }, { "grad_norm": 0.48546457290649414, "learning_rate": 3.930432762351265e-05, "loss": 0.0265, "step": 59020 }, { "grad_norm": 0.3140539526939392, "learning_rate": 3.928817624543931e-05, "loss": 0.026, "step": 59030 }, { "grad_norm": 0.31529587507247925, "learning_rate": 3.9272026038794885e-05, "loss": 0.0176, "step": 59040 }, { "grad_norm": 0.2559486925601959, "learning_rate": 3.9255877005345586e-05, "loss": 0.0264, "step": 59050 }, { "grad_norm": 0.2342655211687088, "learning_rate": 3.923972914685743e-05, "loss": 0.0275, "step": 59060 }, { "grad_norm": 0.33795541524887085, "learning_rate": 3.922358246509631e-05, "loss": 0.0221, "step": 59070 }, { "grad_norm": 0.43627190589904785, "learning_rate": 3.920743696182801e-05, "loss": 0.0295, "step": 59080 }, { "grad_norm": 0.26142579317092896, "learning_rate": 3.9191292638818164e-05, "loss": 0.0265, "step": 59090 }, { "grad_norm": 0.15650013089179993, "learning_rate": 3.9175149497832316e-05, "loss": 0.0242, "step": 59100 }, { "grad_norm": 0.20749053359031677, "learning_rate": 3.915900754063584e-05, "loss": 0.022, "step": 59110 }, { "grad_norm": 0.22800293564796448, "learning_rate": 3.9142866768994014e-05, "loss": 0.0207, "step": 59120 }, { "grad_norm": 0.3587101399898529, "learning_rate": 3.912672718467193e-05, "loss": 0.0238, "step": 59130 }, { "grad_norm": 0.22366994619369507, "learning_rate": 3.911058878943463e-05, "loss": 0.0261, "step": 59140 }, { "grad_norm": 0.1962929368019104, "learning_rate": 3.9094451585046974e-05, "loss": 0.0236, "step": 59150 }, { "grad_norm": 0.2726054787635803, "learning_rate": 3.90783155732737e-05, "loss": 0.0322, "step": 59160 }, { "grad_norm": 0.339091420173645, "learning_rate": 3.9062180755879373e-05, "loss": 0.026, "step": 59170 }, { "grad_norm": 0.2412397265434265, "learning_rate": 3.904604713462855e-05, "loss": 0.0203, "step": 59180 }, { "grad_norm": 0.23410208523273468, "learning_rate": 3.902991471128554e-05, "loss": 0.0181, "step": 59190 }, { "grad_norm": 0.199601411819458, "learning_rate": 3.901378348761456e-05, "loss": 0.0189, "step": 59200 }, { "grad_norm": 0.29845497012138367, "learning_rate": 3.899765346537971e-05, "loss": 0.0231, "step": 59210 }, { "grad_norm": 0.25971630215644836, "learning_rate": 3.898152464634493e-05, "loss": 0.0175, "step": 59220 }, { "grad_norm": 1.3747190237045288, "learning_rate": 3.896539703227402e-05, "loss": 0.0237, "step": 59230 }, { "grad_norm": 0.23290272057056427, "learning_rate": 3.894927062493073e-05, "loss": 0.0205, "step": 59240 }, { "grad_norm": 0.27094975113868713, "learning_rate": 3.893314542607859e-05, "loss": 0.0183, "step": 59250 }, { "grad_norm": 0.3268333971500397, "learning_rate": 3.891702143748103e-05, "loss": 0.024, "step": 59260 }, { "grad_norm": 0.40093737840652466, "learning_rate": 3.890089866090133e-05, "loss": 0.0203, "step": 59270 }, { "grad_norm": 0.3391086161136627, "learning_rate": 3.888477709810267e-05, "loss": 0.0234, "step": 59280 }, { "grad_norm": 0.2397446483373642, "learning_rate": 3.886865675084806e-05, "loss": 0.0201, "step": 59290 }, { "grad_norm": 0.3395921289920807, "learning_rate": 3.885253762090043e-05, "loss": 0.02, "step": 59300 }, { "grad_norm": 0.303077757358551, "learning_rate": 3.8836419710022534e-05, "loss": 0.0273, "step": 59310 }, { "grad_norm": 0.33940601348876953, "learning_rate": 3.882030301997698e-05, "loss": 0.0237, "step": 59320 }, { "grad_norm": 0.21582303941249847, "learning_rate": 3.880418755252629e-05, "loss": 0.019, "step": 59330 }, { "grad_norm": 0.2598712742328644, "learning_rate": 3.87880733094328e-05, "loss": 0.023, "step": 59340 }, { "grad_norm": 0.26965954899787903, "learning_rate": 3.877196029245877e-05, "loss": 0.02, "step": 59350 }, { "grad_norm": 0.22260554134845734, "learning_rate": 3.875584850336627e-05, "loss": 0.0218, "step": 59360 }, { "grad_norm": 0.2672021687030792, "learning_rate": 3.873973794391728e-05, "loss": 0.0236, "step": 59370 }, { "grad_norm": 0.22385269403457642, "learning_rate": 3.872362861587361e-05, "loss": 0.0211, "step": 59380 }, { "grad_norm": 0.2870047390460968, "learning_rate": 3.870752052099698e-05, "loss": 0.0223, "step": 59390 }, { "grad_norm": 0.22332775592803955, "learning_rate": 3.8691413661048916e-05, "loss": 0.0218, "step": 59400 }, { "grad_norm": 0.2936776876449585, "learning_rate": 3.867530803779085e-05, "loss": 0.0185, "step": 59410 }, { "grad_norm": 0.28877657651901245, "learning_rate": 3.8659203652984054e-05, "loss": 0.0246, "step": 59420 }, { "grad_norm": 0.32499635219573975, "learning_rate": 3.864310050838972e-05, "loss": 0.0282, "step": 59430 }, { "grad_norm": 0.46054989099502563, "learning_rate": 3.8626998605768835e-05, "loss": 0.0259, "step": 59440 }, { "grad_norm": 0.3010917603969574, "learning_rate": 3.8610897946882294e-05, "loss": 0.0252, "step": 59450 }, { "grad_norm": 0.2798038125038147, "learning_rate": 3.859479853349082e-05, "loss": 0.0207, "step": 59460 }, { "grad_norm": 0.4458991587162018, "learning_rate": 3.857870036735503e-05, "loss": 0.0206, "step": 59470 }, { "grad_norm": 0.34746742248535156, "learning_rate": 3.856260345023538e-05, "loss": 0.0307, "step": 59480 }, { "grad_norm": 0.3481885492801666, "learning_rate": 3.854650778389224e-05, "loss": 0.0226, "step": 59490 }, { "grad_norm": 0.2844467759132385, "learning_rate": 3.8530413370085786e-05, "loss": 0.0241, "step": 59500 }, { "grad_norm": 0.2848433554172516, "learning_rate": 3.851432021057608e-05, "loss": 0.0193, "step": 59510 }, { "grad_norm": 0.21693900227546692, "learning_rate": 3.849822830712306e-05, "loss": 0.0222, "step": 59520 }, { "grad_norm": 1.081724762916565, "learning_rate": 3.848213766148649e-05, "loss": 0.024, "step": 59530 }, { "grad_norm": 0.2693028450012207, "learning_rate": 3.846604827542602e-05, "loss": 0.0192, "step": 59540 }, { "grad_norm": 0.3041355013847351, "learning_rate": 3.844996015070117e-05, "loss": 0.0248, "step": 59550 }, { "grad_norm": 0.6746068000793457, "learning_rate": 3.843387328907132e-05, "loss": 0.0174, "step": 59560 }, { "grad_norm": 0.723900556564331, "learning_rate": 3.8417787692295696e-05, "loss": 0.0279, "step": 59570 }, { "grad_norm": 0.3553825914859772, "learning_rate": 3.84017033621334e-05, "loss": 0.0181, "step": 59580 }, { "grad_norm": 0.30628135800361633, "learning_rate": 3.838562030034338e-05, "loss": 0.0233, "step": 59590 }, { "grad_norm": 0.5717845559120178, "learning_rate": 3.836953850868447e-05, "loss": 0.0219, "step": 59600 }, { "grad_norm": 0.3170301616191864, "learning_rate": 3.835345798891533e-05, "loss": 0.0187, "step": 59610 }, { "grad_norm": 0.4347691535949707, "learning_rate": 3.833737874279452e-05, "loss": 0.0261, "step": 59620 }, { "grad_norm": 0.38992562890052795, "learning_rate": 3.8321300772080433e-05, "loss": 0.02, "step": 59630 }, { "grad_norm": 0.3360040485858917, "learning_rate": 3.8305224078531345e-05, "loss": 0.0188, "step": 59640 }, { "grad_norm": 0.3512086570262909, "learning_rate": 3.828914866390537e-05, "loss": 0.0229, "step": 59650 }, { "grad_norm": 0.3519008457660675, "learning_rate": 3.827307452996048e-05, "loss": 0.0265, "step": 59660 }, { "grad_norm": 0.33631274104118347, "learning_rate": 3.825700167845454e-05, "loss": 0.0322, "step": 59670 }, { "grad_norm": 0.533619225025177, "learning_rate": 3.8240930111145214e-05, "loss": 0.0233, "step": 59680 }, { "grad_norm": 0.18727755546569824, "learning_rate": 3.8224859829790116e-05, "loss": 0.015, "step": 59690 }, { "grad_norm": 0.29677489399909973, "learning_rate": 3.820879083614664e-05, "loss": 0.0252, "step": 59700 }, { "grad_norm": 0.3909876346588135, "learning_rate": 3.819272313197208e-05, "loss": 0.0227, "step": 59710 }, { "grad_norm": 0.3948250114917755, "learning_rate": 3.817665671902355e-05, "loss": 0.0175, "step": 59720 }, { "grad_norm": 0.27152809500694275, "learning_rate": 3.816059159905808e-05, "loss": 0.0265, "step": 59730 }, { "grad_norm": 0.25655397772789, "learning_rate": 3.8144527773832476e-05, "loss": 0.0252, "step": 59740 }, { "grad_norm": 0.39598962664604187, "learning_rate": 3.812846524510352e-05, "loss": 0.0226, "step": 59750 }, { "grad_norm": 0.34660565853118896, "learning_rate": 3.811240401462775e-05, "loss": 0.0191, "step": 59760 }, { "grad_norm": 0.47783634066581726, "learning_rate": 3.809634408416162e-05, "loss": 0.0378, "step": 59770 }, { "grad_norm": 0.2721399962902069, "learning_rate": 3.8080285455461394e-05, "loss": 0.0249, "step": 59780 }, { "grad_norm": 0.266946941614151, "learning_rate": 3.806422813028323e-05, "loss": 0.0236, "step": 59790 }, { "grad_norm": 0.4218495488166809, "learning_rate": 3.8048172110383114e-05, "loss": 0.0237, "step": 59800 }, { "grad_norm": 0.32122868299484253, "learning_rate": 3.8032117397516944e-05, "loss": 0.0199, "step": 59810 }, { "grad_norm": 0.31909361481666565, "learning_rate": 3.801606399344042e-05, "loss": 0.0197, "step": 59820 }, { "grad_norm": 0.6051260828971863, "learning_rate": 3.800001189990913e-05, "loss": 0.0212, "step": 59830 }, { "grad_norm": 0.35211580991744995, "learning_rate": 3.798396111867847e-05, "loss": 0.0227, "step": 59840 }, { "grad_norm": 0.20994678139686584, "learning_rate": 3.796791165150378e-05, "loss": 0.0173, "step": 59850 }, { "grad_norm": 0.42233866453170776, "learning_rate": 3.795186350014018e-05, "loss": 0.0238, "step": 59860 }, { "grad_norm": 0.3755556046962738, "learning_rate": 3.793581666634266e-05, "loss": 0.0213, "step": 59870 }, { "grad_norm": 0.25156170129776, "learning_rate": 3.791977115186609e-05, "loss": 0.0191, "step": 59880 }, { "grad_norm": 0.43680909276008606, "learning_rate": 3.79037269584652e-05, "loss": 0.0272, "step": 59890 }, { "grad_norm": 0.2950243651866913, "learning_rate": 3.7887684087894545e-05, "loss": 0.0294, "step": 59900 }, { "grad_norm": 0.2334231734275818, "learning_rate": 3.787164254190854e-05, "loss": 0.0174, "step": 59910 }, { "grad_norm": 0.3656892776489258, "learning_rate": 3.7855602322261473e-05, "loss": 0.0204, "step": 59920 }, { "grad_norm": 0.4269828498363495, "learning_rate": 3.783956343070746e-05, "loss": 0.0304, "step": 59930 }, { "grad_norm": 0.341735303401947, "learning_rate": 3.7823525869000526e-05, "loss": 0.0198, "step": 59940 }, { "grad_norm": 0.3695630133152008, "learning_rate": 3.78074896388945e-05, "loss": 0.0282, "step": 59950 }, { "grad_norm": 0.26335060596466064, "learning_rate": 3.7791454742143075e-05, "loss": 0.0243, "step": 59960 }, { "grad_norm": 0.25561633706092834, "learning_rate": 3.7775421180499795e-05, "loss": 0.0196, "step": 59970 }, { "grad_norm": 0.31911635398864746, "learning_rate": 3.7759388955718086e-05, "loss": 0.02, "step": 59980 }, { "grad_norm": 0.20486444234848022, "learning_rate": 3.774335806955117e-05, "loss": 0.0236, "step": 59990 }, { "grad_norm": 0.3897337317466736, "learning_rate": 3.7727328523752215e-05, "loss": 0.0227, "step": 60000 }, { "grad_norm": 0.2610844075679779, "learning_rate": 3.771130032007415e-05, "loss": 0.0225, "step": 60010 }, { "grad_norm": 0.20674745738506317, "learning_rate": 3.769527346026982e-05, "loss": 0.0182, "step": 60020 }, { "grad_norm": 0.29189595580101013, "learning_rate": 3.767924794609188e-05, "loss": 0.0214, "step": 60030 }, { "grad_norm": 0.2829357981681824, "learning_rate": 3.7663223779292854e-05, "loss": 0.0225, "step": 60040 }, { "grad_norm": 0.21794869005680084, "learning_rate": 3.764720096162512e-05, "loss": 0.0184, "step": 60050 }, { "grad_norm": 0.39627397060394287, "learning_rate": 3.763117949484092e-05, "loss": 0.0233, "step": 60060 }, { "grad_norm": 0.40365421772003174, "learning_rate": 3.7615159380692344e-05, "loss": 0.0225, "step": 60070 }, { "grad_norm": 0.22846315801143646, "learning_rate": 3.759914062093132e-05, "loss": 0.0249, "step": 60080 }, { "grad_norm": 0.21863485872745514, "learning_rate": 3.7583123217309615e-05, "loss": 0.0232, "step": 60090 }, { "grad_norm": 0.41201525926589966, "learning_rate": 3.7567107171578904e-05, "loss": 0.0272, "step": 60100 }, { "grad_norm": 0.29837244749069214, "learning_rate": 3.755109248549066e-05, "loss": 0.0241, "step": 60110 }, { "grad_norm": 0.25226014852523804, "learning_rate": 3.75350791607962e-05, "loss": 0.021, "step": 60120 }, { "grad_norm": 0.24943356215953827, "learning_rate": 3.751906719924676e-05, "loss": 0.0187, "step": 60130 }, { "grad_norm": 0.3400140702724457, "learning_rate": 3.750305660259337e-05, "loss": 0.0273, "step": 60140 }, { "grad_norm": 0.2792806625366211, "learning_rate": 3.748704737258693e-05, "loss": 0.0228, "step": 60150 }, { "grad_norm": 0.2734217941761017, "learning_rate": 3.747103951097816e-05, "loss": 0.0301, "step": 60160 }, { "grad_norm": 0.2379414439201355, "learning_rate": 3.745503301951768e-05, "loss": 0.0223, "step": 60170 }, { "grad_norm": 0.4073469340801239, "learning_rate": 3.743902789995592e-05, "loss": 0.0237, "step": 60180 }, { "grad_norm": 0.29994839429855347, "learning_rate": 3.7423024154043183e-05, "loss": 0.0221, "step": 60190 }, { "grad_norm": 0.4120652675628662, "learning_rate": 3.740702178352963e-05, "loss": 0.0236, "step": 60200 }, { "grad_norm": 0.36842992901802063, "learning_rate": 3.739102079016523e-05, "loss": 0.0208, "step": 60210 }, { "grad_norm": 0.29718902707099915, "learning_rate": 3.7375021175699846e-05, "loss": 0.021, "step": 60220 }, { "grad_norm": 0.2919233739376068, "learning_rate": 3.735902294188317e-05, "loss": 0.0251, "step": 60230 }, { "grad_norm": 0.19218754768371582, "learning_rate": 3.734302609046473e-05, "loss": 0.0174, "step": 60240 }, { "grad_norm": 0.4385252296924591, "learning_rate": 3.732703062319391e-05, "loss": 0.0217, "step": 60250 }, { "grad_norm": 0.4111931025981903, "learning_rate": 3.731103654181999e-05, "loss": 0.0237, "step": 60260 }, { "grad_norm": 0.37071770429611206, "learning_rate": 3.729504384809203e-05, "loss": 0.0245, "step": 60270 }, { "grad_norm": 0.19649451971054077, "learning_rate": 3.727905254375898e-05, "loss": 0.0257, "step": 60280 }, { "grad_norm": 0.2300558090209961, "learning_rate": 3.7263062630569614e-05, "loss": 0.0203, "step": 60290 }, { "grad_norm": 0.20614957809448242, "learning_rate": 3.724707411027256e-05, "loss": 0.0177, "step": 60300 }, { "grad_norm": 0.3921784460544586, "learning_rate": 3.723108698461631e-05, "loss": 0.0298, "step": 60310 }, { "grad_norm": 0.2259017527103424, "learning_rate": 3.72151012553492e-05, "loss": 0.0205, "step": 60320 }, { "grad_norm": 0.30574774742126465, "learning_rate": 3.719911692421939e-05, "loss": 0.0202, "step": 60330 }, { "grad_norm": 0.4266822934150696, "learning_rate": 3.71831339929749e-05, "loss": 0.0286, "step": 60340 }, { "grad_norm": 0.28137943148612976, "learning_rate": 3.716715246336362e-05, "loss": 0.0265, "step": 60350 }, { "grad_norm": 0.20904113352298737, "learning_rate": 3.715117233713324e-05, "loss": 0.0154, "step": 60360 }, { "grad_norm": 0.40535280108451843, "learning_rate": 3.713519361603135e-05, "loss": 0.0218, "step": 60370 }, { "grad_norm": 0.4558654725551605, "learning_rate": 3.711921630180532e-05, "loss": 0.0252, "step": 60380 }, { "grad_norm": 0.2392357587814331, "learning_rate": 3.710324039620245e-05, "loss": 0.0187, "step": 60390 }, { "grad_norm": 0.43325167894363403, "learning_rate": 3.708726590096982e-05, "loss": 0.0219, "step": 60400 }, { "grad_norm": 0.24989068508148193, "learning_rate": 3.707129281785437e-05, "loss": 0.0203, "step": 60410 }, { "grad_norm": 0.3537572920322418, "learning_rate": 3.705532114860291e-05, "loss": 0.0222, "step": 60420 }, { "grad_norm": 0.5031082630157471, "learning_rate": 3.703935089496207e-05, "loss": 0.0306, "step": 60430 }, { "grad_norm": 0.26080918312072754, "learning_rate": 3.702338205867829e-05, "loss": 0.0182, "step": 60440 }, { "grad_norm": 0.30955278873443604, "learning_rate": 3.7007414641497976e-05, "loss": 0.0289, "step": 60450 }, { "grad_norm": 0.25846734642982483, "learning_rate": 3.699144864516726e-05, "loss": 0.033, "step": 60460 }, { "grad_norm": 0.3328574597835541, "learning_rate": 3.6975484071432153e-05, "loss": 0.0243, "step": 60470 }, { "grad_norm": 0.1992875188589096, "learning_rate": 3.695952092203854e-05, "loss": 0.0181, "step": 60480 }, { "grad_norm": 0.35678204894065857, "learning_rate": 3.69435591987321e-05, "loss": 0.0329, "step": 60490 }, { "grad_norm": 0.22795704007148743, "learning_rate": 3.6927598903258374e-05, "loss": 0.0201, "step": 60500 }, { "grad_norm": 0.33846843242645264, "learning_rate": 3.691164003736281e-05, "loss": 0.0197, "step": 60510 }, { "grad_norm": 0.3560529351234436, "learning_rate": 3.689568260279061e-05, "loss": 0.0265, "step": 60520 }, { "grad_norm": 0.29640254378318787, "learning_rate": 3.687972660128686e-05, "loss": 0.0196, "step": 60530 }, { "grad_norm": 0.3326544463634491, "learning_rate": 3.686377203459648e-05, "loss": 0.0218, "step": 60540 }, { "grad_norm": 0.39297202229499817, "learning_rate": 3.6847818904464226e-05, "loss": 0.0199, "step": 60550 }, { "grad_norm": 0.29835429787635803, "learning_rate": 3.683186721263474e-05, "loss": 0.0216, "step": 60560 }, { "grad_norm": 0.4943266808986664, "learning_rate": 3.6815916960852435e-05, "loss": 0.0228, "step": 60570 }, { "grad_norm": 0.21234072744846344, "learning_rate": 3.679996815086165e-05, "loss": 0.0222, "step": 60580 }, { "grad_norm": 0.1983702927827835, "learning_rate": 3.678402078440649e-05, "loss": 0.0203, "step": 60590 }, { "grad_norm": 0.5923258662223816, "learning_rate": 3.676807486323096e-05, "loss": 0.0208, "step": 60600 }, { "grad_norm": 0.2547662556171417, "learning_rate": 3.6752130389078864e-05, "loss": 0.0233, "step": 60610 }, { "grad_norm": 0.24119554460048676, "learning_rate": 3.673618736369388e-05, "loss": 0.0215, "step": 60620 }, { "grad_norm": 0.2880396842956543, "learning_rate": 3.6720245788819474e-05, "loss": 0.021, "step": 60630 }, { "grad_norm": 0.3316795825958252, "learning_rate": 3.6704305666199044e-05, "loss": 0.0208, "step": 60640 }, { "grad_norm": 0.20834980905056, "learning_rate": 3.668836699757576e-05, "loss": 0.0173, "step": 60650 }, { "grad_norm": 0.3354819416999817, "learning_rate": 3.667242978469265e-05, "loss": 0.0227, "step": 60660 }, { "grad_norm": 0.29784613847732544, "learning_rate": 3.6656494029292575e-05, "loss": 0.028, "step": 60670 }, { "grad_norm": 0.21222761273384094, "learning_rate": 3.664055973311825e-05, "loss": 0.0251, "step": 60680 }, { "grad_norm": 0.23964416980743408, "learning_rate": 3.662462689791221e-05, "loss": 0.0189, "step": 60690 }, { "grad_norm": 0.3245505690574646, "learning_rate": 3.660869552541689e-05, "loss": 0.0173, "step": 60700 }, { "grad_norm": 0.26491257548332214, "learning_rate": 3.659276561737448e-05, "loss": 0.0184, "step": 60710 }, { "grad_norm": 0.24545343220233917, "learning_rate": 3.657683717552708e-05, "loss": 0.0232, "step": 60720 }, { "grad_norm": 0.2876802980899811, "learning_rate": 3.6560910201616574e-05, "loss": 0.018, "step": 60730 }, { "grad_norm": 0.33289802074432373, "learning_rate": 3.654498469738472e-05, "loss": 0.0249, "step": 60740 }, { "grad_norm": 0.5150061845779419, "learning_rate": 3.6529060664573126e-05, "loss": 0.029, "step": 60750 }, { "grad_norm": 0.3139900863170624, "learning_rate": 3.6513138104923176e-05, "loss": 0.0243, "step": 60760 }, { "grad_norm": 0.5389301776885986, "learning_rate": 3.6497217020176176e-05, "loss": 0.0221, "step": 60770 }, { "grad_norm": 0.3073153495788574, "learning_rate": 3.648129741207323e-05, "loss": 0.0264, "step": 60780 }, { "grad_norm": 0.23134367167949677, "learning_rate": 3.646537928235527e-05, "loss": 0.0235, "step": 60790 }, { "grad_norm": 0.2822435796260834, "learning_rate": 3.6449462632763075e-05, "loss": 0.0241, "step": 60800 }, { "grad_norm": 0.4409770965576172, "learning_rate": 3.643354746503729e-05, "loss": 0.027, "step": 60810 }, { "grad_norm": 0.5794135332107544, "learning_rate": 3.6417633780918345e-05, "loss": 0.0228, "step": 60820 }, { "grad_norm": 0.322964608669281, "learning_rate": 3.640172158214655e-05, "loss": 0.0178, "step": 60830 }, { "grad_norm": 0.5626555681228638, "learning_rate": 3.6385810870462035e-05, "loss": 0.0238, "step": 60840 }, { "grad_norm": 0.38017353415489197, "learning_rate": 3.636990164760479e-05, "loss": 0.0238, "step": 60850 }, { "grad_norm": 0.44159895181655884, "learning_rate": 3.6353993915314596e-05, "loss": 0.022, "step": 60860 }, { "grad_norm": 0.3261745274066925, "learning_rate": 3.6338087675331124e-05, "loss": 0.0185, "step": 60870 }, { "grad_norm": 0.26338863372802734, "learning_rate": 3.632218292939383e-05, "loss": 0.0212, "step": 60880 }, { "grad_norm": 0.3881741464138031, "learning_rate": 3.6306279679242033e-05, "loss": 0.0216, "step": 60890 }, { "grad_norm": 0.2258579283952713, "learning_rate": 3.629037792661492e-05, "loss": 0.0175, "step": 60900 }, { "grad_norm": 0.25452175736427307, "learning_rate": 3.627447767325147e-05, "loss": 0.0238, "step": 60910 }, { "grad_norm": 0.26030203700065613, "learning_rate": 3.62585789208905e-05, "loss": 0.016, "step": 60920 }, { "grad_norm": 1.106208086013794, "learning_rate": 3.624268167127068e-05, "loss": 0.0259, "step": 60930 }, { "grad_norm": 0.24855315685272217, "learning_rate": 3.622678592613051e-05, "loss": 0.0193, "step": 60940 }, { "grad_norm": 0.3322596848011017, "learning_rate": 3.6210891687208304e-05, "loss": 0.0233, "step": 60950 }, { "grad_norm": 0.23094625771045685, "learning_rate": 3.619499895624228e-05, "loss": 0.0234, "step": 60960 }, { "grad_norm": 0.2662110924720764, "learning_rate": 3.61791077349704e-05, "loss": 0.0213, "step": 60970 }, { "grad_norm": 0.3020460903644562, "learning_rate": 3.616321802513053e-05, "loss": 0.0235, "step": 60980 }, { "grad_norm": 0.3826046884059906, "learning_rate": 3.6147329828460325e-05, "loss": 0.0178, "step": 60990 }, { "grad_norm": 0.32819244265556335, "learning_rate": 3.6131443146697306e-05, "loss": 0.0186, "step": 61000 }, { "grad_norm": 0.18500101566314697, "learning_rate": 3.611555798157879e-05, "loss": 0.019, "step": 61010 }, { "grad_norm": 0.2661599814891815, "learning_rate": 3.6099674334842e-05, "loss": 0.0197, "step": 61020 }, { "grad_norm": 0.20617736876010895, "learning_rate": 3.6083792208223934e-05, "loss": 0.0161, "step": 61030 }, { "grad_norm": 0.33735793828964233, "learning_rate": 3.606791160346142e-05, "loss": 0.0245, "step": 61040 }, { "grad_norm": 0.31100836396217346, "learning_rate": 3.6052032522291134e-05, "loss": 0.0222, "step": 61050 }, { "grad_norm": 0.523947536945343, "learning_rate": 3.603615496644962e-05, "loss": 0.0307, "step": 61060 }, { "grad_norm": 0.3357214033603668, "learning_rate": 3.60202789376732e-05, "loss": 0.019, "step": 61070 }, { "grad_norm": 0.30050942301750183, "learning_rate": 3.6004404437698036e-05, "loss": 0.0172, "step": 61080 }, { "grad_norm": 0.45512855052948, "learning_rate": 3.5988531468260184e-05, "loss": 0.0189, "step": 61090 }, { "grad_norm": 0.2906760573387146, "learning_rate": 3.597266003109546e-05, "loss": 0.0186, "step": 61100 }, { "grad_norm": 0.33541181683540344, "learning_rate": 3.5956790127939557e-05, "loss": 0.0224, "step": 61110 }, { "grad_norm": 1.0501139163970947, "learning_rate": 3.594092176052796e-05, "loss": 0.0242, "step": 61120 }, { "grad_norm": 0.3717952072620392, "learning_rate": 3.592505493059603e-05, "loss": 0.0294, "step": 61130 }, { "grad_norm": 0.36260321736335754, "learning_rate": 3.5909189639878905e-05, "loss": 0.0269, "step": 61140 }, { "grad_norm": 0.3927435278892517, "learning_rate": 3.5893325890111626e-05, "loss": 0.0172, "step": 61150 }, { "grad_norm": 0.18158592283725739, "learning_rate": 3.5877463683029034e-05, "loss": 0.0281, "step": 61160 }, { "grad_norm": 0.24211382865905762, "learning_rate": 3.586160302036578e-05, "loss": 0.027, "step": 61170 }, { "grad_norm": 0.27097854018211365, "learning_rate": 3.584574390385636e-05, "loss": 0.0186, "step": 61180 }, { "grad_norm": 0.4373067617416382, "learning_rate": 3.582988633523511e-05, "loss": 0.0191, "step": 61190 }, { "grad_norm": 0.5646230578422546, "learning_rate": 3.581403031623616e-05, "loss": 0.0216, "step": 61200 }, { "grad_norm": 0.2609955072402954, "learning_rate": 3.579817584859354e-05, "loss": 0.0192, "step": 61210 }, { "grad_norm": 0.4449188709259033, "learning_rate": 3.578232293404107e-05, "loss": 0.0246, "step": 61220 }, { "grad_norm": 0.3395151197910309, "learning_rate": 3.5766471574312385e-05, "loss": 0.0198, "step": 61230 }, { "grad_norm": 0.3818552792072296, "learning_rate": 3.5750621771140964e-05, "loss": 0.0291, "step": 61240 }, { "grad_norm": 0.38145190477371216, "learning_rate": 3.5734773526260115e-05, "loss": 0.0261, "step": 61250 }, { "grad_norm": 0.21610529720783234, "learning_rate": 3.5718926841402996e-05, "loss": 0.0256, "step": 61260 }, { "grad_norm": 0.33433806896209717, "learning_rate": 3.5703081718302555e-05, "loss": 0.0268, "step": 61270 }, { "grad_norm": 0.3142957091331482, "learning_rate": 3.568723815869162e-05, "loss": 0.0299, "step": 61280 }, { "grad_norm": 0.3241858184337616, "learning_rate": 3.5671396164302777e-05, "loss": 0.0273, "step": 61290 }, { "grad_norm": 0.2538089156150818, "learning_rate": 3.565555573686853e-05, "loss": 0.0208, "step": 61300 }, { "grad_norm": 0.36585772037506104, "learning_rate": 3.563971687812113e-05, "loss": 0.0215, "step": 61310 }, { "grad_norm": 0.2520529627799988, "learning_rate": 3.5623879589792706e-05, "loss": 0.0217, "step": 61320 }, { "grad_norm": 0.39631685614585876, "learning_rate": 3.5608043873615185e-05, "loss": 0.0276, "step": 61330 }, { "grad_norm": 0.1917552798986435, "learning_rate": 3.559220973132035e-05, "loss": 0.0202, "step": 61340 }, { "grad_norm": 0.21950988471508026, "learning_rate": 3.557637716463981e-05, "loss": 0.0215, "step": 61350 }, { "grad_norm": 0.303570419549942, "learning_rate": 3.556054617530497e-05, "loss": 0.0219, "step": 61360 }, { "grad_norm": 0.33893632888793945, "learning_rate": 3.5544716765047106e-05, "loss": 0.0196, "step": 61370 }, { "grad_norm": 0.583791196346283, "learning_rate": 3.5528888935597274e-05, "loss": 0.0304, "step": 61380 }, { "grad_norm": 0.27511584758758545, "learning_rate": 3.551306268868636e-05, "loss": 0.0195, "step": 61390 }, { "grad_norm": 0.7043638229370117, "learning_rate": 3.5497238026045166e-05, "loss": 0.0234, "step": 61400 }, { "grad_norm": 0.473695307970047, "learning_rate": 3.548141494940421e-05, "loss": 0.0258, "step": 61410 }, { "grad_norm": 0.28414732217788696, "learning_rate": 3.546559346049388e-05, "loss": 0.0204, "step": 61420 }, { "grad_norm": 0.5233159065246582, "learning_rate": 3.544977356104441e-05, "loss": 0.0163, "step": 61430 }, { "grad_norm": 0.3910604417324066, "learning_rate": 3.543395525278582e-05, "loss": 0.0313, "step": 61440 }, { "grad_norm": 0.2975237965583801, "learning_rate": 3.5418138537447985e-05, "loss": 0.0171, "step": 61450 }, { "grad_norm": 0.5639474391937256, "learning_rate": 3.5402323416760574e-05, "loss": 0.022, "step": 61460 }, { "grad_norm": 0.5618035793304443, "learning_rate": 3.538650989245316e-05, "loss": 0.0261, "step": 61470 }, { "grad_norm": 0.3620342016220093, "learning_rate": 3.537069796625504e-05, "loss": 0.0216, "step": 61480 }, { "grad_norm": 0.2466646283864975, "learning_rate": 3.53548876398954e-05, "loss": 0.0156, "step": 61490 }, { "grad_norm": 0.24594317376613617, "learning_rate": 3.5339078915103214e-05, "loss": 0.0273, "step": 61500 }, { "grad_norm": 0.3118450343608856, "learning_rate": 3.532327179360733e-05, "loss": 0.0205, "step": 61510 }, { "grad_norm": 0.5030657052993774, "learning_rate": 3.530746627713636e-05, "loss": 0.0211, "step": 61520 }, { "grad_norm": 0.26283198595046997, "learning_rate": 3.5291662367418805e-05, "loss": 0.0161, "step": 61530 }, { "grad_norm": 1.317893624305725, "learning_rate": 3.527586006618293e-05, "loss": 0.0244, "step": 61540 }, { "grad_norm": 0.25565555691719055, "learning_rate": 3.526005937515687e-05, "loss": 0.0248, "step": 61550 }, { "grad_norm": 0.2282995581626892, "learning_rate": 3.524426029606856e-05, "loss": 0.0331, "step": 61560 }, { "grad_norm": 0.2594924867153168, "learning_rate": 3.5228462830645746e-05, "loss": 0.0243, "step": 61570 }, { "grad_norm": 0.30316346883773804, "learning_rate": 3.5212666980616036e-05, "loss": 0.0197, "step": 61580 }, { "grad_norm": 0.33628469705581665, "learning_rate": 3.519687274770681e-05, "loss": 0.0261, "step": 61590 }, { "grad_norm": 0.3415064215660095, "learning_rate": 3.518108013364535e-05, "loss": 0.0314, "step": 61600 }, { "grad_norm": 0.24167579412460327, "learning_rate": 3.516528914015868e-05, "loss": 0.0245, "step": 61610 }, { "grad_norm": 0.24379822611808777, "learning_rate": 3.5149499768973694e-05, "loss": 0.0202, "step": 61620 }, { "grad_norm": 0.35411784052848816, "learning_rate": 3.513371202181708e-05, "loss": 0.019, "step": 61630 }, { "grad_norm": 0.8558757305145264, "learning_rate": 3.511792590041537e-05, "loss": 0.017, "step": 61640 }, { "grad_norm": 0.3247331976890564, "learning_rate": 3.510214140649488e-05, "loss": 0.0293, "step": 61650 }, { "grad_norm": 0.326466828584671, "learning_rate": 3.508635854178184e-05, "loss": 0.0318, "step": 61660 }, { "grad_norm": 0.3800937235355377, "learning_rate": 3.50705773080022e-05, "loss": 0.0287, "step": 61670 }, { "grad_norm": 0.29977601766586304, "learning_rate": 3.505479770688178e-05, "loss": 0.0212, "step": 61680 }, { "grad_norm": 0.5343005657196045, "learning_rate": 3.503901974014622e-05, "loss": 0.0242, "step": 61690 }, { "grad_norm": 0.24589712917804718, "learning_rate": 3.502324340952095e-05, "loss": 0.0234, "step": 61700 }, { "grad_norm": 0.3009549677371979, "learning_rate": 3.500746871673126e-05, "loss": 0.0214, "step": 61710 }, { "grad_norm": 0.2782612442970276, "learning_rate": 3.499169566350227e-05, "loss": 0.0195, "step": 61720 }, { "grad_norm": 0.28183817863464355, "learning_rate": 3.4975924251558876e-05, "loss": 0.0193, "step": 61730 }, { "grad_norm": 0.3422013819217682, "learning_rate": 3.496015448262582e-05, "loss": 0.0234, "step": 61740 }, { "grad_norm": 0.4054357707500458, "learning_rate": 3.494438635842764e-05, "loss": 0.0194, "step": 61750 }, { "grad_norm": 0.22984807193279266, "learning_rate": 3.492861988068874e-05, "loss": 0.0176, "step": 61760 }, { "grad_norm": 0.5647718906402588, "learning_rate": 3.491285505113332e-05, "loss": 0.0274, "step": 61770 }, { "grad_norm": 0.1825115829706192, "learning_rate": 3.4897091871485376e-05, "loss": 0.0217, "step": 61780 }, { "grad_norm": 0.19012495875358582, "learning_rate": 3.488133034346875e-05, "loss": 0.0249, "step": 61790 }, { "grad_norm": 0.375952810049057, "learning_rate": 3.486557046880713e-05, "loss": 0.0205, "step": 61800 }, { "grad_norm": 0.36694809794425964, "learning_rate": 3.484981224922395e-05, "loss": 0.0252, "step": 61810 }, { "grad_norm": 0.8514930009841919, "learning_rate": 3.4834055686442544e-05, "loss": 0.0295, "step": 61820 }, { "grad_norm": 0.4111427366733551, "learning_rate": 3.4818300782186e-05, "loss": 0.0217, "step": 61830 }, { "grad_norm": 0.32807672023773193, "learning_rate": 3.4802547538177236e-05, "loss": 0.0229, "step": 61840 }, { "grad_norm": 0.2731539011001587, "learning_rate": 3.478679595613905e-05, "loss": 0.0172, "step": 61850 }, { "grad_norm": 0.31610509753227234, "learning_rate": 3.4771046037793975e-05, "loss": 0.0274, "step": 61860 }, { "grad_norm": 0.4053959250450134, "learning_rate": 3.4755297784864416e-05, "loss": 0.0229, "step": 61870 }, { "grad_norm": 0.4230175316333771, "learning_rate": 3.473955119907257e-05, "loss": 0.0206, "step": 61880 }, { "grad_norm": 0.8272391557693481, "learning_rate": 3.472380628214046e-05, "loss": 0.0213, "step": 61890 }, { "grad_norm": 0.23181968927383423, "learning_rate": 3.4708063035789906e-05, "loss": 0.0163, "step": 61900 }, { "grad_norm": 0.299106240272522, "learning_rate": 3.4692321461742614e-05, "loss": 0.0213, "step": 61910 }, { "grad_norm": 0.31922489404678345, "learning_rate": 3.4676581561720023e-05, "loss": 0.0184, "step": 61920 }, { "grad_norm": 0.31128573417663574, "learning_rate": 3.466084333744344e-05, "loss": 0.0236, "step": 61930 }, { "grad_norm": 0.29981932044029236, "learning_rate": 3.4645106790633964e-05, "loss": 0.0257, "step": 61940 }, { "grad_norm": 0.18008354306221008, "learning_rate": 3.4629371923012526e-05, "loss": 0.0221, "step": 61950 }, { "grad_norm": 0.3278111517429352, "learning_rate": 3.4613638736299856e-05, "loss": 0.0186, "step": 61960 }, { "grad_norm": 0.2241639345884323, "learning_rate": 3.459790723221652e-05, "loss": 0.0165, "step": 61970 }, { "grad_norm": 0.2298232614994049, "learning_rate": 3.45821774124829e-05, "loss": 0.0231, "step": 61980 }, { "grad_norm": 0.3040926456451416, "learning_rate": 3.456644927881918e-05, "loss": 0.0169, "step": 61990 }, { "grad_norm": 0.2523754835128784, "learning_rate": 3.4550722832945354e-05, "loss": 0.0184, "step": 62000 }, { "grad_norm": 0.863375186920166, "learning_rate": 3.453499807658126e-05, "loss": 0.0187, "step": 62010 }, { "grad_norm": 0.17761635780334473, "learning_rate": 3.451927501144653e-05, "loss": 0.0222, "step": 62020 }, { "grad_norm": 0.3605680465698242, "learning_rate": 3.450355363926059e-05, "loss": 0.0152, "step": 62030 }, { "grad_norm": 0.2399369180202484, "learning_rate": 3.448783396174273e-05, "loss": 0.0216, "step": 62040 }, { "grad_norm": 0.42978256940841675, "learning_rate": 3.4472115980612034e-05, "loss": 0.0244, "step": 62050 }, { "grad_norm": 0.23447026312351227, "learning_rate": 3.4456399697587396e-05, "loss": 0.0179, "step": 62060 }, { "grad_norm": 0.32709938287734985, "learning_rate": 3.4440685114387506e-05, "loss": 0.0181, "step": 62070 }, { "grad_norm": 0.2718568742275238, "learning_rate": 3.442497223273091e-05, "loss": 0.0206, "step": 62080 }, { "grad_norm": 0.2929788827896118, "learning_rate": 3.4409261054335926e-05, "loss": 0.0319, "step": 62090 }, { "grad_norm": 0.42829906940460205, "learning_rate": 3.439355158092069e-05, "loss": 0.0272, "step": 62100 }, { "grad_norm": 0.27947551012039185, "learning_rate": 3.437784381420322e-05, "loss": 0.0236, "step": 62110 }, { "grad_norm": 0.26772215962409973, "learning_rate": 3.436213775590126e-05, "loss": 0.0154, "step": 62120 }, { "grad_norm": 0.28447282314300537, "learning_rate": 3.4346433407732405e-05, "loss": 0.0235, "step": 62130 }, { "grad_norm": 0.21694117784500122, "learning_rate": 3.433073077141405e-05, "loss": 0.0171, "step": 62140 }, { "grad_norm": 0.3184451460838318, "learning_rate": 3.431502984866343e-05, "loss": 0.0235, "step": 62150 }, { "grad_norm": 0.3793109655380249, "learning_rate": 3.429933064119753e-05, "loss": 0.0344, "step": 62160 }, { "grad_norm": 0.35668280720710754, "learning_rate": 3.4283633150733255e-05, "loss": 0.0192, "step": 62170 }, { "grad_norm": 0.4418492019176483, "learning_rate": 3.4267937378987225e-05, "loss": 0.0296, "step": 62180 }, { "grad_norm": 0.20723453164100647, "learning_rate": 3.425224332767592e-05, "loss": 0.0165, "step": 62190 }, { "grad_norm": 0.23349781334400177, "learning_rate": 3.4236550998515595e-05, "loss": 0.0232, "step": 62200 }, { "grad_norm": 0.3083488643169403, "learning_rate": 3.422086039322235e-05, "loss": 0.0208, "step": 62210 }, { "grad_norm": 0.3077223598957062, "learning_rate": 3.420517151351209e-05, "loss": 0.0184, "step": 62220 }, { "grad_norm": 0.2774481177330017, "learning_rate": 3.418948436110053e-05, "loss": 0.0185, "step": 62230 }, { "grad_norm": 0.2502760589122772, "learning_rate": 3.417379893770319e-05, "loss": 0.0208, "step": 62240 }, { "grad_norm": 0.2303209900856018, "learning_rate": 3.41581152450354e-05, "loss": 0.0217, "step": 62250 }, { "grad_norm": 0.24815037846565247, "learning_rate": 3.414243328481231e-05, "loss": 0.0201, "step": 62260 }, { "grad_norm": 0.20213107764720917, "learning_rate": 3.412675305874888e-05, "loss": 0.0173, "step": 62270 }, { "grad_norm": 0.2697150707244873, "learning_rate": 3.4111074568559876e-05, "loss": 0.0202, "step": 62280 }, { "grad_norm": 0.2651040256023407, "learning_rate": 3.4095397815959846e-05, "loss": 0.0246, "step": 62290 }, { "grad_norm": 0.1868945062160492, "learning_rate": 3.407972280266322e-05, "loss": 0.0205, "step": 62300 }, { "grad_norm": 0.29502248764038086, "learning_rate": 3.4064049530384166e-05, "loss": 0.0194, "step": 62310 }, { "grad_norm": 0.2203853577375412, "learning_rate": 3.40483780008367e-05, "loss": 0.0249, "step": 62320 }, { "grad_norm": 0.2475065141916275, "learning_rate": 3.4032708215734635e-05, "loss": 0.0229, "step": 62330 }, { "grad_norm": 0.2762399911880493, "learning_rate": 3.4017040176791584e-05, "loss": 0.0236, "step": 62340 }, { "grad_norm": 0.4016628563404083, "learning_rate": 3.4001373885720976e-05, "loss": 0.0229, "step": 62350 }, { "grad_norm": 0.3234446346759796, "learning_rate": 3.398570934423608e-05, "loss": 0.025, "step": 62360 }, { "grad_norm": 0.2489803433418274, "learning_rate": 3.397004655404994e-05, "loss": 0.0168, "step": 62370 }, { "grad_norm": 0.38504907488822937, "learning_rate": 3.395438551687541e-05, "loss": 0.0258, "step": 62380 }, { "grad_norm": 0.3731243908405304, "learning_rate": 3.393872623442515e-05, "loss": 0.0231, "step": 62390 }, { "grad_norm": 0.3163797855377197, "learning_rate": 3.392306870841164e-05, "loss": 0.018, "step": 62400 }, { "grad_norm": 0.37868666648864746, "learning_rate": 3.390741294054715e-05, "loss": 0.0228, "step": 62410 }, { "grad_norm": 0.26079514622688293, "learning_rate": 3.38917589325438e-05, "loss": 0.0145, "step": 62420 }, { "grad_norm": 0.24368777871131897, "learning_rate": 3.387610668611349e-05, "loss": 0.02, "step": 62430 }, { "grad_norm": 0.1904771476984024, "learning_rate": 3.38604562029679e-05, "loss": 0.0209, "step": 62440 }, { "grad_norm": 0.17002560198307037, "learning_rate": 3.384480748481855e-05, "loss": 0.0234, "step": 62450 }, { "grad_norm": 0.2564562261104584, "learning_rate": 3.382916053337676e-05, "loss": 0.0187, "step": 62460 }, { "grad_norm": 0.3029238283634186, "learning_rate": 3.381351535035366e-05, "loss": 0.0204, "step": 62470 }, { "grad_norm": 0.2936004102230072, "learning_rate": 3.379787193746019e-05, "loss": 0.0212, "step": 62480 }, { "grad_norm": 0.27704331278800964, "learning_rate": 3.378223029640709e-05, "loss": 0.0155, "step": 62490 }, { "grad_norm": 0.33848118782043457, "learning_rate": 3.376659042890489e-05, "loss": 0.0166, "step": 62500 }, { "grad_norm": 0.29626989364624023, "learning_rate": 3.375095233666397e-05, "loss": 0.0193, "step": 62510 }, { "grad_norm": 0.3367207646369934, "learning_rate": 3.373531602139447e-05, "loss": 0.0235, "step": 62520 }, { "grad_norm": 0.3860393166542053, "learning_rate": 3.371968148480635e-05, "loss": 0.0172, "step": 62530 }, { "grad_norm": 0.3286169469356537, "learning_rate": 3.370404872860937e-05, "loss": 0.018, "step": 62540 }, { "grad_norm": 0.4695817232131958, "learning_rate": 3.3688417754513145e-05, "loss": 0.0229, "step": 62550 }, { "grad_norm": 0.21691814064979553, "learning_rate": 3.367278856422703e-05, "loss": 0.0225, "step": 62560 }, { "grad_norm": 0.4186796545982361, "learning_rate": 3.365716115946021e-05, "loss": 0.0196, "step": 62570 }, { "grad_norm": 0.33478233218193054, "learning_rate": 3.3641535541921666e-05, "loss": 0.0338, "step": 62580 }, { "grad_norm": 0.3086870610713959, "learning_rate": 3.36259117133202e-05, "loss": 0.0246, "step": 62590 }, { "grad_norm": 0.2921932637691498, "learning_rate": 3.36102896753644e-05, "loss": 0.0221, "step": 62600 }, { "grad_norm": 0.2668130099773407, "learning_rate": 3.35946694297627e-05, "loss": 0.0194, "step": 62610 }, { "grad_norm": 0.2677568793296814, "learning_rate": 3.357905097822328e-05, "loss": 0.0194, "step": 62620 }, { "grad_norm": 0.24871619045734406, "learning_rate": 3.356343432245416e-05, "loss": 0.0189, "step": 62630 }, { "grad_norm": 0.3818661570549011, "learning_rate": 3.354781946416315e-05, "loss": 0.0319, "step": 62640 }, { "grad_norm": 0.3638986647129059, "learning_rate": 3.3532206405057864e-05, "loss": 0.0202, "step": 62650 }, { "grad_norm": 0.4380282759666443, "learning_rate": 3.351659514684573e-05, "loss": 0.0207, "step": 62660 }, { "grad_norm": 0.35374847054481506, "learning_rate": 3.350098569123395e-05, "loss": 0.0237, "step": 62670 }, { "grad_norm": 0.312061071395874, "learning_rate": 3.348537803992958e-05, "loss": 0.0202, "step": 62680 }, { "grad_norm": 0.16870610415935516, "learning_rate": 3.346977219463945e-05, "loss": 0.0234, "step": 62690 }, { "grad_norm": 0.22146880626678467, "learning_rate": 3.345416815707017e-05, "loss": 0.0249, "step": 62700 }, { "grad_norm": 0.6142004132270813, "learning_rate": 3.3438565928928185e-05, "loss": 0.0234, "step": 62710 }, { "grad_norm": 0.2000357061624527, "learning_rate": 3.3422965511919736e-05, "loss": 0.021, "step": 62720 }, { "grad_norm": 0.2340356409549713, "learning_rate": 3.3407366907750846e-05, "loss": 0.0198, "step": 62730 }, { "grad_norm": 0.2588747441768646, "learning_rate": 3.339177011812737e-05, "loss": 0.0218, "step": 62740 }, { "grad_norm": 0.39763346314430237, "learning_rate": 3.3376175144754935e-05, "loss": 0.0247, "step": 62750 }, { "grad_norm": 0.430552214384079, "learning_rate": 3.3360581989339004e-05, "loss": 0.018, "step": 62760 }, { "grad_norm": 0.27330178022384644, "learning_rate": 3.3344990653584804e-05, "loss": 0.0192, "step": 62770 }, { "grad_norm": 0.38275405764579773, "learning_rate": 3.332940113919739e-05, "loss": 0.019, "step": 62780 }, { "grad_norm": 0.15225906670093536, "learning_rate": 3.331381344788161e-05, "loss": 0.0204, "step": 62790 }, { "grad_norm": 0.47864922881126404, "learning_rate": 3.3298227581342076e-05, "loss": 0.0205, "step": 62800 }, { "grad_norm": 0.21074925363063812, "learning_rate": 3.328264354128329e-05, "loss": 0.0298, "step": 62810 }, { "grad_norm": 0.29183050990104675, "learning_rate": 3.3267061329409465e-05, "loss": 0.0254, "step": 62820 }, { "grad_norm": 0.22504085302352905, "learning_rate": 3.325148094742466e-05, "loss": 0.0205, "step": 62830 }, { "grad_norm": 0.1713186353445053, "learning_rate": 3.323590239703271e-05, "loss": 0.0168, "step": 62840 }, { "grad_norm": 0.2834845185279846, "learning_rate": 3.322032567993728e-05, "loss": 0.0217, "step": 62850 }, { "grad_norm": 0.29956546425819397, "learning_rate": 3.3204750797841774e-05, "loss": 0.0209, "step": 62860 }, { "grad_norm": 0.2949327230453491, "learning_rate": 3.31891777524495e-05, "loss": 0.0187, "step": 62870 }, { "grad_norm": 0.2302764356136322, "learning_rate": 3.317360654546347e-05, "loss": 0.0213, "step": 62880 }, { "grad_norm": 0.18059833347797394, "learning_rate": 3.3158037178586534e-05, "loss": 0.0176, "step": 62890 }, { "grad_norm": 0.25055161118507385, "learning_rate": 3.3142469653521316e-05, "loss": 0.0271, "step": 62900 }, { "grad_norm": 0.2737274765968323, "learning_rate": 3.3126903971970284e-05, "loss": 0.022, "step": 62910 }, { "grad_norm": 0.384716659784317, "learning_rate": 3.311134013563564e-05, "loss": 0.0226, "step": 62920 }, { "grad_norm": 0.2020663470029831, "learning_rate": 3.309577814621947e-05, "loss": 0.0191, "step": 62930 }, { "grad_norm": 0.3700943887233734, "learning_rate": 3.308021800542359e-05, "loss": 0.0224, "step": 62940 }, { "grad_norm": 0.9236152768135071, "learning_rate": 3.306465971494963e-05, "loss": 0.0264, "step": 62950 }, { "grad_norm": 0.305908203125, "learning_rate": 3.304910327649901e-05, "loss": 0.0238, "step": 62960 }, { "grad_norm": 0.7183271050453186, "learning_rate": 3.303354869177297e-05, "loss": 0.0185, "step": 62970 }, { "grad_norm": 0.5833094120025635, "learning_rate": 3.301799596247255e-05, "loss": 0.0197, "step": 62980 }, { "grad_norm": 0.32311752438545227, "learning_rate": 3.300244509029854e-05, "loss": 0.0181, "step": 62990 }, { "grad_norm": 0.31390002369880676, "learning_rate": 3.2986896076951585e-05, "loss": 0.0167, "step": 63000 }, { "grad_norm": 0.3870585262775421, "learning_rate": 3.297134892413211e-05, "loss": 0.0222, "step": 63010 }, { "grad_norm": 0.3328283131122589, "learning_rate": 3.295580363354031e-05, "loss": 0.0179, "step": 63020 }, { "grad_norm": 0.4383186101913452, "learning_rate": 3.2940260206876194e-05, "loss": 0.0233, "step": 63030 }, { "grad_norm": 0.28817811608314514, "learning_rate": 3.292471864583958e-05, "loss": 0.0196, "step": 63040 }, { "grad_norm": 0.2587966024875641, "learning_rate": 3.290917895213004e-05, "loss": 0.024, "step": 63050 }, { "grad_norm": 0.3619876801967621, "learning_rate": 3.289364112744701e-05, "loss": 0.0294, "step": 63060 }, { "grad_norm": 0.3884512186050415, "learning_rate": 3.287810517348967e-05, "loss": 0.0283, "step": 63070 }, { "grad_norm": 0.38710448145866394, "learning_rate": 3.2862571091957014e-05, "loss": 0.0263, "step": 63080 }, { "grad_norm": 0.38073140382766724, "learning_rate": 3.284703888454781e-05, "loss": 0.0187, "step": 63090 }, { "grad_norm": 0.31148943305015564, "learning_rate": 3.2831508552960646e-05, "loss": 0.0255, "step": 63100 }, { "grad_norm": 0.23249982297420502, "learning_rate": 3.281598009889387e-05, "loss": 0.0234, "step": 63110 }, { "grad_norm": 0.3207933008670807, "learning_rate": 3.28004535240457e-05, "loss": 0.029, "step": 63120 }, { "grad_norm": 0.4935820400714874, "learning_rate": 3.278492883011407e-05, "loss": 0.0237, "step": 63130 }, { "grad_norm": 0.430475115776062, "learning_rate": 3.276940601879675e-05, "loss": 0.0236, "step": 63140 }, { "grad_norm": 0.5603750348091125, "learning_rate": 3.275388509179127e-05, "loss": 0.0259, "step": 63150 }, { "grad_norm": 0.2736569046974182, "learning_rate": 3.2738366050794985e-05, "loss": 0.0266, "step": 63160 }, { "grad_norm": 0.2686327397823334, "learning_rate": 3.2722848897505034e-05, "loss": 0.023, "step": 63170 }, { "grad_norm": 0.25308847427368164, "learning_rate": 3.2707333633618356e-05, "loss": 0.0155, "step": 63180 }, { "grad_norm": 0.20921990275382996, "learning_rate": 3.269182026083167e-05, "loss": 0.0172, "step": 63190 }, { "grad_norm": 0.47732052206993103, "learning_rate": 3.2676308780841516e-05, "loss": 0.0258, "step": 63200 }, { "grad_norm": 0.46742138266563416, "learning_rate": 3.266079919534416e-05, "loss": 0.0213, "step": 63210 }, { "grad_norm": 0.38317909836769104, "learning_rate": 3.264529150603575e-05, "loss": 0.0227, "step": 63220 }, { "grad_norm": 0.2902190685272217, "learning_rate": 3.262978571461217e-05, "loss": 0.0162, "step": 63230 }, { "grad_norm": 0.21137204766273499, "learning_rate": 3.26142818227691e-05, "loss": 0.0199, "step": 63240 }, { "grad_norm": 0.3527778387069702, "learning_rate": 3.259877983220203e-05, "loss": 0.0248, "step": 63250 }, { "grad_norm": 0.16495996713638306, "learning_rate": 3.2583279744606246e-05, "loss": 0.0274, "step": 63260 }, { "grad_norm": 0.3175505995750427, "learning_rate": 3.25677815616768e-05, "loss": 0.0197, "step": 63270 }, { "grad_norm": 0.2891598343849182, "learning_rate": 3.255228528510856e-05, "loss": 0.0227, "step": 63280 }, { "grad_norm": 0.17041024565696716, "learning_rate": 3.253679091659617e-05, "loss": 0.0195, "step": 63290 }, { "grad_norm": 0.2568424940109253, "learning_rate": 3.252129845783404e-05, "loss": 0.0243, "step": 63300 }, { "grad_norm": 0.2901822328567505, "learning_rate": 3.2505807910516456e-05, "loss": 0.0191, "step": 63310 }, { "grad_norm": 0.28885164856910706, "learning_rate": 3.249031927633742e-05, "loss": 0.0211, "step": 63320 }, { "grad_norm": 0.3053242266178131, "learning_rate": 3.2474832556990737e-05, "loss": 0.0188, "step": 63330 }, { "grad_norm": 0.539941668510437, "learning_rate": 3.245934775417002e-05, "loss": 0.0206, "step": 63340 }, { "grad_norm": 0.25149279832839966, "learning_rate": 3.244386486956866e-05, "loss": 0.0226, "step": 63350 }, { "grad_norm": 0.28848761320114136, "learning_rate": 3.2428383904879846e-05, "loss": 0.0164, "step": 63360 }, { "grad_norm": 0.3016796112060547, "learning_rate": 3.241290486179652e-05, "loss": 0.0356, "step": 63370 }, { "grad_norm": 0.2929230332374573, "learning_rate": 3.239742774201151e-05, "loss": 0.0232, "step": 63380 }, { "grad_norm": 0.367464542388916, "learning_rate": 3.238195254721733e-05, "loss": 0.0221, "step": 63390 }, { "grad_norm": 0.3207036852836609, "learning_rate": 3.236647927910633e-05, "loss": 0.0203, "step": 63400 }, { "grad_norm": 0.5607892870903015, "learning_rate": 3.2351007939370646e-05, "loss": 0.0191, "step": 63410 }, { "grad_norm": 0.3032178282737732, "learning_rate": 3.2335538529702184e-05, "loss": 0.0198, "step": 63420 }, { "grad_norm": 0.31638461351394653, "learning_rate": 3.232007105179268e-05, "loss": 0.0211, "step": 63430 }, { "grad_norm": 0.29817068576812744, "learning_rate": 3.230460550733363e-05, "loss": 0.0181, "step": 63440 }, { "grad_norm": 0.24132747948169708, "learning_rate": 3.2289141898016315e-05, "loss": 0.0213, "step": 63450 }, { "grad_norm": 0.42492568492889404, "learning_rate": 3.227368022553181e-05, "loss": 0.0272, "step": 63460 }, { "grad_norm": 0.5000033974647522, "learning_rate": 3.225822049157099e-05, "loss": 0.0306, "step": 63470 }, { "grad_norm": 0.26338863372802734, "learning_rate": 3.2242762697824494e-05, "loss": 0.0184, "step": 63480 }, { "grad_norm": 0.3898211121559143, "learning_rate": 3.222730684598279e-05, "loss": 0.0317, "step": 63490 }, { "grad_norm": 0.29810988903045654, "learning_rate": 3.221185293773605e-05, "loss": 0.0258, "step": 63500 }, { "grad_norm": 0.32528969645500183, "learning_rate": 3.2196400974774354e-05, "loss": 0.0205, "step": 63510 }, { "grad_norm": 0.2664644122123718, "learning_rate": 3.2180950958787485e-05, "loss": 0.027, "step": 63520 }, { "grad_norm": 0.20599612593650818, "learning_rate": 3.2165502891465024e-05, "loss": 0.0213, "step": 63530 }, { "grad_norm": 0.2916008234024048, "learning_rate": 3.215005677449636e-05, "loss": 0.0223, "step": 63540 }, { "grad_norm": 0.3282778859138489, "learning_rate": 3.2134612609570644e-05, "loss": 0.0227, "step": 63550 }, { "grad_norm": 0.2180691510438919, "learning_rate": 3.211917039837682e-05, "loss": 0.0271, "step": 63560 }, { "grad_norm": 0.6725775003433228, "learning_rate": 3.210373014260365e-05, "loss": 0.0258, "step": 63570 }, { "grad_norm": 0.297587126493454, "learning_rate": 3.208829184393965e-05, "loss": 0.0154, "step": 63580 }, { "grad_norm": 0.2835824489593506, "learning_rate": 3.207285550407312e-05, "loss": 0.0283, "step": 63590 }, { "grad_norm": 0.3446255028247833, "learning_rate": 3.205742112469217e-05, "loss": 0.0176, "step": 63600 }, { "grad_norm": 0.3687252998352051, "learning_rate": 3.204198870748465e-05, "loss": 0.0265, "step": 63610 }, { "grad_norm": 0.34859463572502136, "learning_rate": 3.202655825413824e-05, "loss": 0.0264, "step": 63620 }, { "grad_norm": 0.2861262559890747, "learning_rate": 3.2011129766340416e-05, "loss": 0.0303, "step": 63630 }, { "grad_norm": 0.28950709104537964, "learning_rate": 3.1995703245778384e-05, "loss": 0.0182, "step": 63640 }, { "grad_norm": 0.2464434653520584, "learning_rate": 3.1980278694139186e-05, "loss": 0.019, "step": 63650 }, { "grad_norm": 0.24572430551052094, "learning_rate": 3.196485611310962e-05, "loss": 0.0247, "step": 63660 }, { "grad_norm": 0.25321969389915466, "learning_rate": 3.1949435504376257e-05, "loss": 0.0221, "step": 63670 }, { "grad_norm": 0.21818412840366364, "learning_rate": 3.1934016869625505e-05, "loss": 0.0229, "step": 63680 }, { "grad_norm": 0.4413715898990631, "learning_rate": 3.191860021054349e-05, "loss": 0.0308, "step": 63690 }, { "grad_norm": 0.2391268014907837, "learning_rate": 3.190318552881618e-05, "loss": 0.0301, "step": 63700 }, { "grad_norm": 0.27621784806251526, "learning_rate": 3.1887772826129285e-05, "loss": 0.0191, "step": 63710 }, { "grad_norm": 0.2465272694826126, "learning_rate": 3.1872362104168335e-05, "loss": 0.0203, "step": 63720 }, { "grad_norm": 0.2594130039215088, "learning_rate": 3.185695336461861e-05, "loss": 0.0185, "step": 63730 }, { "grad_norm": 0.30713000893592834, "learning_rate": 3.184154660916519e-05, "loss": 0.0194, "step": 63740 }, { "grad_norm": 0.33381369709968567, "learning_rate": 3.1826141839492906e-05, "loss": 0.0201, "step": 63750 }, { "grad_norm": 0.2629621624946594, "learning_rate": 3.181073905728644e-05, "loss": 0.0201, "step": 63760 }, { "grad_norm": 0.40232640504837036, "learning_rate": 3.1795338264230214e-05, "loss": 0.0162, "step": 63770 }, { "grad_norm": 0.4276188015937805, "learning_rate": 3.1779939462008426e-05, "loss": 0.0234, "step": 63780 }, { "grad_norm": 0.28213196992874146, "learning_rate": 3.1764542652305066e-05, "loss": 0.0193, "step": 63790 }, { "grad_norm": 0.3698970377445221, "learning_rate": 3.17491478368039e-05, "loss": 0.0211, "step": 63800 }, { "grad_norm": 0.34863194823265076, "learning_rate": 3.173375501718847e-05, "loss": 0.0201, "step": 63810 }, { "grad_norm": 0.2623846232891083, "learning_rate": 3.171836419514215e-05, "loss": 0.025, "step": 63820 }, { "grad_norm": 0.30177250504493713, "learning_rate": 3.170297537234803e-05, "loss": 0.0244, "step": 63830 }, { "grad_norm": 0.27516838908195496, "learning_rate": 3.168758855048901e-05, "loss": 0.0212, "step": 63840 }, { "grad_norm": 0.3078688681125641, "learning_rate": 3.1672203731247785e-05, "loss": 0.0226, "step": 63850 }, { "grad_norm": 0.26791536808013916, "learning_rate": 3.16568209163068e-05, "loss": 0.0185, "step": 63860 }, { "grad_norm": 0.2867092192173004, "learning_rate": 3.1641440107348295e-05, "loss": 0.0195, "step": 63870 }, { "grad_norm": 0.36764249205589294, "learning_rate": 3.162606130605429e-05, "loss": 0.028, "step": 63880 }, { "grad_norm": 0.2580952048301697, "learning_rate": 3.16106845141066e-05, "loss": 0.0222, "step": 63890 }, { "grad_norm": 0.3017500936985016, "learning_rate": 3.159530973318682e-05, "loss": 0.0174, "step": 63900 }, { "grad_norm": 0.29397138953208923, "learning_rate": 3.1579936964976286e-05, "loss": 0.0232, "step": 63910 }, { "grad_norm": 0.2781546115875244, "learning_rate": 3.156456621115615e-05, "loss": 0.0219, "step": 63920 }, { "grad_norm": 0.23561052978038788, "learning_rate": 3.154919747340734e-05, "loss": 0.019, "step": 63930 }, { "grad_norm": 0.3241729140281677, "learning_rate": 3.153383075341054e-05, "loss": 0.0213, "step": 63940 }, { "grad_norm": 0.32158201932907104, "learning_rate": 3.1518466052846264e-05, "loss": 0.0208, "step": 63950 }, { "grad_norm": 0.30874037742614746, "learning_rate": 3.150310337339474e-05, "loss": 0.0227, "step": 63960 }, { "grad_norm": 0.2761445641517639, "learning_rate": 3.148774271673603e-05, "loss": 0.0199, "step": 63970 }, { "grad_norm": 0.31560322642326355, "learning_rate": 3.147238408454995e-05, "loss": 0.0253, "step": 63980 }, { "grad_norm": 0.26578062772750854, "learning_rate": 3.1457027478516086e-05, "loss": 0.0217, "step": 63990 }, { "grad_norm": 0.3609391450881958, "learning_rate": 3.144167290031382e-05, "loss": 0.0209, "step": 64000 }, { "grad_norm": 0.23739120364189148, "learning_rate": 3.142632035162227e-05, "loss": 0.0188, "step": 64010 }, { "grad_norm": 0.28339990973472595, "learning_rate": 3.141096983412044e-05, "loss": 0.0171, "step": 64020 }, { "grad_norm": 0.16803865134716034, "learning_rate": 3.139562134948699e-05, "loss": 0.0283, "step": 64030 }, { "grad_norm": 0.15482427179813385, "learning_rate": 3.1380274899400424e-05, "loss": 0.0235, "step": 64040 }, { "grad_norm": 0.4207342267036438, "learning_rate": 3.1364930485539e-05, "loss": 0.0185, "step": 64050 }, { "grad_norm": 0.39616331458091736, "learning_rate": 3.134958810958076e-05, "loss": 0.0196, "step": 64060 }, { "grad_norm": 0.3789306581020355, "learning_rate": 3.13342477732035e-05, "loss": 0.0179, "step": 64070 }, { "grad_norm": 0.48078033328056335, "learning_rate": 3.131890947808487e-05, "loss": 0.0243, "step": 64080 }, { "grad_norm": 0.2552056312561035, "learning_rate": 3.13035732259022e-05, "loss": 0.0195, "step": 64090 }, { "grad_norm": 0.19205062091350555, "learning_rate": 3.1288239018332656e-05, "loss": 0.0205, "step": 64100 }, { "grad_norm": 0.5394971370697021, "learning_rate": 3.1272906857053164e-05, "loss": 0.0283, "step": 64110 }, { "grad_norm": 0.22200259566307068, "learning_rate": 3.125757674374042e-05, "loss": 0.0249, "step": 64120 }, { "grad_norm": 0.23080036044120789, "learning_rate": 3.124224868007088e-05, "loss": 0.0229, "step": 64130 }, { "grad_norm": 0.36492493748664856, "learning_rate": 3.1226922667720854e-05, "loss": 0.0197, "step": 64140 }, { "grad_norm": 0.44002312421798706, "learning_rate": 3.121159870836633e-05, "loss": 0.0249, "step": 64150 }, { "grad_norm": 0.2977204918861389, "learning_rate": 3.119627680368312e-05, "loss": 0.0208, "step": 64160 }, { "grad_norm": 0.36963146924972534, "learning_rate": 3.1180956955346816e-05, "loss": 0.0211, "step": 64170 }, { "grad_norm": 0.3497215807437897, "learning_rate": 3.116563916503276e-05, "loss": 0.0243, "step": 64180 }, { "grad_norm": 0.30901235342025757, "learning_rate": 3.1150323434416095e-05, "loss": 0.0203, "step": 64190 }, { "grad_norm": 0.26634320616722107, "learning_rate": 3.1135009765171696e-05, "loss": 0.0159, "step": 64200 }, { "grad_norm": 0.2713852822780609, "learning_rate": 3.111969815897429e-05, "loss": 0.018, "step": 64210 }, { "grad_norm": 0.3028816282749176, "learning_rate": 3.1104388617498295e-05, "loss": 0.0221, "step": 64220 }, { "grad_norm": 0.4263601005077362, "learning_rate": 3.1089081142417966e-05, "loss": 0.0214, "step": 64230 }, { "grad_norm": 0.20725609362125397, "learning_rate": 3.1073775735407285e-05, "loss": 0.0145, "step": 64240 }, { "grad_norm": 0.26898226141929626, "learning_rate": 3.1058472398140035e-05, "loss": 0.0216, "step": 64250 }, { "grad_norm": 0.38489556312561035, "learning_rate": 3.1043171132289736e-05, "loss": 0.0204, "step": 64260 }, { "grad_norm": 0.3340916931629181, "learning_rate": 3.1027871939529774e-05, "loss": 0.0187, "step": 64270 }, { "grad_norm": 0.33771762251853943, "learning_rate": 3.1012574821533214e-05, "loss": 0.0296, "step": 64280 }, { "grad_norm": 0.2932793200016022, "learning_rate": 3.099727977997291e-05, "loss": 0.0193, "step": 64290 }, { "grad_norm": 0.2767978310585022, "learning_rate": 3.098198681652154e-05, "loss": 0.0231, "step": 64300 }, { "grad_norm": 0.27925100922584534, "learning_rate": 3.0966695932851475e-05, "loss": 0.0158, "step": 64310 }, { "grad_norm": 0.24567320942878723, "learning_rate": 3.0951407130634926e-05, "loss": 0.0229, "step": 64320 }, { "grad_norm": 0.31627339124679565, "learning_rate": 3.0936120411543864e-05, "loss": 0.0228, "step": 64330 }, { "grad_norm": 0.3355688452720642, "learning_rate": 3.0920835777250015e-05, "loss": 0.0209, "step": 64340 }, { "grad_norm": 0.2535845935344696, "learning_rate": 3.090555322942488e-05, "loss": 0.017, "step": 64350 }, { "grad_norm": 0.33289003372192383, "learning_rate": 3.089027276973973e-05, "loss": 0.0205, "step": 64360 }, { "grad_norm": 0.4676477015018463, "learning_rate": 3.0874994399865615e-05, "loss": 0.0282, "step": 64370 }, { "grad_norm": 0.40821561217308044, "learning_rate": 3.085971812147337e-05, "loss": 0.0277, "step": 64380 }, { "grad_norm": 0.26264119148254395, "learning_rate": 3.0844443936233557e-05, "loss": 0.0236, "step": 64390 }, { "grad_norm": 0.2814640402793884, "learning_rate": 3.082917184581657e-05, "loss": 0.0173, "step": 64400 }, { "grad_norm": 1.0385630130767822, "learning_rate": 3.0813901851892513e-05, "loss": 0.0182, "step": 64410 }, { "grad_norm": 0.3620709478855133, "learning_rate": 3.0798633956131316e-05, "loss": 0.0326, "step": 64420 }, { "grad_norm": 0.659765899181366, "learning_rate": 3.078336816020264e-05, "loss": 0.0204, "step": 64430 }, { "grad_norm": 0.48669126629829407, "learning_rate": 3.0768104465775935e-05, "loss": 0.0215, "step": 64440 }, { "grad_norm": 0.2043563276529312, "learning_rate": 3.0752842874520396e-05, "loss": 0.022, "step": 64450 }, { "grad_norm": 0.2842356860637665, "learning_rate": 3.073758338810503e-05, "loss": 0.0192, "step": 64460 }, { "grad_norm": 0.22232624888420105, "learning_rate": 3.072232600819859e-05, "loss": 0.0203, "step": 64470 }, { "grad_norm": 1.161749005317688, "learning_rate": 3.070707073646959e-05, "loss": 0.0217, "step": 64480 }, { "grad_norm": 0.35669463872909546, "learning_rate": 3.069181757458633e-05, "loss": 0.022, "step": 64490 }, { "grad_norm": 0.19348759949207306, "learning_rate": 3.067656652421687e-05, "loss": 0.0235, "step": 64500 }, { "grad_norm": 0.9814072251319885, "learning_rate": 3.0661317587029025e-05, "loss": 0.0221, "step": 64510 }, { "grad_norm": 0.33328089118003845, "learning_rate": 3.064607076469043e-05, "loss": 0.0239, "step": 64520 }, { "grad_norm": 0.3570963144302368, "learning_rate": 3.063082605886845e-05, "loss": 0.0193, "step": 64530 }, { "grad_norm": 0.2527351677417755, "learning_rate": 3.061558347123019e-05, "loss": 0.0314, "step": 64540 }, { "grad_norm": 0.1697062999010086, "learning_rate": 3.060034300344259e-05, "loss": 0.0187, "step": 64550 }, { "grad_norm": 0.3752523958683014, "learning_rate": 3.058510465717231e-05, "loss": 0.0245, "step": 64560 }, { "grad_norm": 0.24317048490047455, "learning_rate": 3.0569868434085804e-05, "loss": 0.0228, "step": 64570 }, { "grad_norm": 0.4308776557445526, "learning_rate": 3.055463433584924e-05, "loss": 0.02, "step": 64580 }, { "grad_norm": 0.2414538562297821, "learning_rate": 3.0539402364128656e-05, "loss": 0.0193, "step": 64590 }, { "grad_norm": 0.3068792521953583, "learning_rate": 3.052417252058977e-05, "loss": 0.0184, "step": 64600 }, { "grad_norm": 0.23464426398277283, "learning_rate": 3.0508944806898092e-05, "loss": 0.0252, "step": 64610 }, { "grad_norm": 0.5492590665817261, "learning_rate": 3.04937192247189e-05, "loss": 0.0269, "step": 64620 }, { "grad_norm": 0.19021858274936676, "learning_rate": 3.047849577571725e-05, "loss": 0.0179, "step": 64630 }, { "grad_norm": 0.29096898436546326, "learning_rate": 3.046327446155795e-05, "loss": 0.0182, "step": 64640 }, { "grad_norm": 0.29865723848342896, "learning_rate": 3.0448055283905587e-05, "loss": 0.0191, "step": 64650 }, { "grad_norm": 0.2822116017341614, "learning_rate": 3.0432838244424494e-05, "loss": 0.0198, "step": 64660 }, { "grad_norm": 0.22856676578521729, "learning_rate": 3.04176233447788e-05, "loss": 0.015, "step": 64670 }, { "grad_norm": 0.211885467171669, "learning_rate": 3.040241058663238e-05, "loss": 0.0162, "step": 64680 }, { "grad_norm": 0.2925024628639221, "learning_rate": 3.038719997164887e-05, "loss": 0.0219, "step": 64690 }, { "grad_norm": 0.3176572620868683, "learning_rate": 3.0371991501491683e-05, "loss": 0.0277, "step": 64700 }, { "grad_norm": 0.18398436903953552, "learning_rate": 3.0356785177823975e-05, "loss": 0.0228, "step": 64710 }, { "grad_norm": 0.3144860565662384, "learning_rate": 3.0341581002308726e-05, "loss": 0.0189, "step": 64720 }, { "grad_norm": 0.2394951432943344, "learning_rate": 3.0326378976608627e-05, "loss": 0.0177, "step": 64730 }, { "grad_norm": 0.41674596071243286, "learning_rate": 3.0311179102386146e-05, "loss": 0.0283, "step": 64740 }, { "grad_norm": 0.23310638964176178, "learning_rate": 3.0295981381303513e-05, "loss": 0.0148, "step": 64750 }, { "grad_norm": 0.25523731112480164, "learning_rate": 3.0280785815022727e-05, "loss": 0.0148, "step": 64760 }, { "grad_norm": 0.21776838600635529, "learning_rate": 3.0265592405205544e-05, "loss": 0.0159, "step": 64770 }, { "grad_norm": 0.38919341564178467, "learning_rate": 3.0250401153513517e-05, "loss": 0.0257, "step": 64780 }, { "grad_norm": 0.260515421628952, "learning_rate": 3.0235212061607936e-05, "loss": 0.0254, "step": 64790 }, { "grad_norm": 0.563777506351471, "learning_rate": 3.0220025131149842e-05, "loss": 0.0208, "step": 64800 }, { "grad_norm": 1.2333929538726807, "learning_rate": 3.020484036380007e-05, "loss": 0.0195, "step": 64810 }, { "grad_norm": 0.4731045365333557, "learning_rate": 3.0189657761219188e-05, "loss": 0.0236, "step": 64820 }, { "grad_norm": 0.2874304950237274, "learning_rate": 3.017447732506753e-05, "loss": 0.0174, "step": 64830 }, { "grad_norm": 0.21438509225845337, "learning_rate": 3.015929905700524e-05, "loss": 0.0156, "step": 64840 }, { "grad_norm": 0.21419183909893036, "learning_rate": 3.014412295869218e-05, "loss": 0.0163, "step": 64850 }, { "grad_norm": 0.40643465518951416, "learning_rate": 3.0128949031787977e-05, "loss": 0.017, "step": 64860 }, { "grad_norm": 0.29819369316101074, "learning_rate": 3.0113777277952022e-05, "loss": 0.028, "step": 64870 }, { "grad_norm": 0.2591439187526703, "learning_rate": 3.0098607698843496e-05, "loss": 0.0187, "step": 64880 }, { "grad_norm": 0.23870551586151123, "learning_rate": 3.0083440296121312e-05, "loss": 0.0187, "step": 64890 }, { "grad_norm": 1.0334450006484985, "learning_rate": 3.0068275071444142e-05, "loss": 0.0237, "step": 64900 }, { "grad_norm": 0.4634803533554077, "learning_rate": 3.0053112026470442e-05, "loss": 0.0285, "step": 64910 }, { "grad_norm": 0.21151742339134216, "learning_rate": 3.0037951162858434e-05, "loss": 0.0181, "step": 64920 }, { "grad_norm": 0.36919236183166504, "learning_rate": 3.002279248226607e-05, "loss": 0.0233, "step": 64930 }, { "grad_norm": 0.302689790725708, "learning_rate": 3.000763598635109e-05, "loss": 0.0192, "step": 64940 }, { "grad_norm": 0.3779006600379944, "learning_rate": 2.999248167677098e-05, "loss": 0.021, "step": 64950 }, { "grad_norm": 0.2787408232688904, "learning_rate": 2.9977329555182976e-05, "loss": 0.0184, "step": 64960 }, { "grad_norm": 0.3096378445625305, "learning_rate": 2.9962179623244124e-05, "loss": 0.0209, "step": 64970 }, { "grad_norm": 0.3378189206123352, "learning_rate": 2.9947031882611175e-05, "loss": 0.0205, "step": 64980 }, { "grad_norm": 0.17604954540729523, "learning_rate": 2.9931886334940685e-05, "loss": 0.0155, "step": 64990 }, { "grad_norm": 0.28788498044013977, "learning_rate": 2.9916742981888923e-05, "loss": 0.0182, "step": 65000 }, { "grad_norm": 0.29654496908187866, "learning_rate": 2.990160182511196e-05, "loss": 0.0256, "step": 65010 }, { "grad_norm": 0.3757404386997223, "learning_rate": 2.9886462866265574e-05, "loss": 0.0161, "step": 65020 }, { "grad_norm": 0.20418497920036316, "learning_rate": 2.9871326107005392e-05, "loss": 0.0205, "step": 65030 }, { "grad_norm": 0.4225316047668457, "learning_rate": 2.9856191548986735e-05, "loss": 0.0224, "step": 65040 }, { "grad_norm": 0.22974812984466553, "learning_rate": 2.9841059193864672e-05, "loss": 0.0204, "step": 65050 }, { "grad_norm": 0.2543290853500366, "learning_rate": 2.982592904329407e-05, "loss": 0.0245, "step": 65060 }, { "grad_norm": 0.26205334067344666, "learning_rate": 2.981080109892954e-05, "loss": 0.025, "step": 65070 }, { "grad_norm": 0.20032593607902527, "learning_rate": 2.9795675362425434e-05, "loss": 0.0177, "step": 65080 }, { "grad_norm": 0.20290382206439972, "learning_rate": 2.97805518354359e-05, "loss": 0.0175, "step": 65090 }, { "grad_norm": 0.32621458172798157, "learning_rate": 2.9765430519614825e-05, "loss": 0.0202, "step": 65100 }, { "grad_norm": 0.17279651761054993, "learning_rate": 2.975031141661584e-05, "loss": 0.0203, "step": 65110 }, { "grad_norm": 0.4086694121360779, "learning_rate": 2.9735194528092347e-05, "loss": 0.0238, "step": 65120 }, { "grad_norm": 0.27123862504959106, "learning_rate": 2.9720079855697525e-05, "loss": 0.0253, "step": 65130 }, { "grad_norm": 0.23798292875289917, "learning_rate": 2.970496740108427e-05, "loss": 0.0239, "step": 65140 }, { "grad_norm": 0.11616713553667068, "learning_rate": 2.968985716590527e-05, "loss": 0.0153, "step": 65150 }, { "grad_norm": 0.3471170663833618, "learning_rate": 2.9674749151812953e-05, "loss": 0.0145, "step": 65160 }, { "grad_norm": 0.2537281811237335, "learning_rate": 2.965964336045951e-05, "loss": 0.0179, "step": 65170 }, { "grad_norm": 0.8822228312492371, "learning_rate": 2.9644539793496894e-05, "loss": 0.0211, "step": 65180 }, { "grad_norm": 0.28671085834503174, "learning_rate": 2.9629438452576807e-05, "loss": 0.0171, "step": 65190 }, { "grad_norm": 0.283973753452301, "learning_rate": 2.9614339339350698e-05, "loss": 0.0181, "step": 65200 }, { "grad_norm": 0.2330150455236435, "learning_rate": 2.959924245546979e-05, "loss": 0.0163, "step": 65210 }, { "grad_norm": 0.3020176291465759, "learning_rate": 2.9584147802585038e-05, "loss": 0.0305, "step": 65220 }, { "grad_norm": 0.5319098830223083, "learning_rate": 2.956905538234721e-05, "loss": 0.0205, "step": 65230 }, { "grad_norm": 0.20542371273040771, "learning_rate": 2.9553965196406764e-05, "loss": 0.0217, "step": 65240 }, { "grad_norm": 0.34671053290367126, "learning_rate": 2.9538877246413943e-05, "loss": 0.0216, "step": 65250 }, { "grad_norm": 0.26570165157318115, "learning_rate": 2.952379153401874e-05, "loss": 0.0224, "step": 65260 }, { "grad_norm": 0.30268871784210205, "learning_rate": 2.9508708060870914e-05, "loss": 0.0219, "step": 65270 }, { "grad_norm": 0.2509499192237854, "learning_rate": 2.9493626828619948e-05, "loss": 0.0147, "step": 65280 }, { "grad_norm": 0.24522316455841064, "learning_rate": 2.9478547838915137e-05, "loss": 0.0192, "step": 65290 }, { "grad_norm": 0.15976554155349731, "learning_rate": 2.946347109340548e-05, "loss": 0.0207, "step": 65300 }, { "grad_norm": 0.2377464473247528, "learning_rate": 2.944839659373976e-05, "loss": 0.0219, "step": 65310 }, { "grad_norm": 0.39630958437919617, "learning_rate": 2.943332434156647e-05, "loss": 0.023, "step": 65320 }, { "grad_norm": 0.3310174345970154, "learning_rate": 2.9418254338533912e-05, "loss": 0.024, "step": 65330 }, { "grad_norm": 0.2496555596590042, "learning_rate": 2.940318658629011e-05, "loss": 0.023, "step": 65340 }, { "grad_norm": 0.2915968894958496, "learning_rate": 2.938812108648287e-05, "loss": 0.0163, "step": 65350 }, { "grad_norm": 0.25035855174064636, "learning_rate": 2.937305784075972e-05, "loss": 0.0204, "step": 65360 }, { "grad_norm": 0.42382436990737915, "learning_rate": 2.9357996850767937e-05, "loss": 0.0198, "step": 65370 }, { "grad_norm": 0.3417382538318634, "learning_rate": 2.9342938118154595e-05, "loss": 0.0168, "step": 65380 }, { "grad_norm": 0.2189183086156845, "learning_rate": 2.9327881644566485e-05, "loss": 0.0272, "step": 65390 }, { "grad_norm": 0.41424039006233215, "learning_rate": 2.9312827431650157e-05, "loss": 0.0212, "step": 65400 }, { "grad_norm": 0.17491641640663147, "learning_rate": 2.92977754810519e-05, "loss": 0.0186, "step": 65410 }, { "grad_norm": 0.23836182057857513, "learning_rate": 2.928272579441781e-05, "loss": 0.0153, "step": 65420 }, { "grad_norm": 0.3344769775867462, "learning_rate": 2.9267678373393676e-05, "loss": 0.0228, "step": 65430 }, { "grad_norm": 0.328485906124115, "learning_rate": 2.925263321962507e-05, "loss": 0.0232, "step": 65440 }, { "grad_norm": 0.31438398361206055, "learning_rate": 2.9237590334757292e-05, "loss": 0.0161, "step": 65450 }, { "grad_norm": 0.3559056222438812, "learning_rate": 2.922254972043543e-05, "loss": 0.0204, "step": 65460 }, { "grad_norm": 0.2874610424041748, "learning_rate": 2.9207511378304265e-05, "loss": 0.0203, "step": 65470 }, { "grad_norm": 0.23984813690185547, "learning_rate": 2.9192475310008428e-05, "loss": 0.0202, "step": 65480 }, { "grad_norm": 0.16770519316196442, "learning_rate": 2.917744151719219e-05, "loss": 0.0145, "step": 65490 }, { "grad_norm": 0.2529453933238983, "learning_rate": 2.916241000149964e-05, "loss": 0.0258, "step": 65500 }, { "grad_norm": 0.47670072317123413, "learning_rate": 2.9147380764574622e-05, "loss": 0.0164, "step": 65510 }, { "grad_norm": 0.23739507794380188, "learning_rate": 2.9132353808060674e-05, "loss": 0.0296, "step": 65520 }, { "grad_norm": 0.25682738423347473, "learning_rate": 2.9117329133601134e-05, "loss": 0.0297, "step": 65530 }, { "grad_norm": 0.20205554366111755, "learning_rate": 2.9102306742839093e-05, "loss": 0.0163, "step": 65540 }, { "grad_norm": 0.391862690448761, "learning_rate": 2.9087286637417375e-05, "loss": 0.0207, "step": 65550 }, { "grad_norm": 0.3016650676727295, "learning_rate": 2.9072268818978537e-05, "loss": 0.0255, "step": 65560 }, { "grad_norm": 0.6642484664916992, "learning_rate": 2.905725328916493e-05, "loss": 0.0241, "step": 65570 }, { "grad_norm": 0.17591358721256256, "learning_rate": 2.9042240049618608e-05, "loss": 0.0211, "step": 65580 }, { "grad_norm": 0.31647437810897827, "learning_rate": 2.9027229101981413e-05, "loss": 0.0191, "step": 65590 }, { "grad_norm": 0.24961094558238983, "learning_rate": 2.901222044789488e-05, "loss": 0.0169, "step": 65600 }, { "grad_norm": 0.3594634532928467, "learning_rate": 2.89972140890004e-05, "loss": 0.0182, "step": 65610 }, { "grad_norm": 0.3730088174343109, "learning_rate": 2.8982210026938988e-05, "loss": 0.0231, "step": 65620 }, { "grad_norm": 0.285197913646698, "learning_rate": 2.896720826335151e-05, "loss": 0.0317, "step": 65630 }, { "grad_norm": 0.18627992272377014, "learning_rate": 2.8952208799878493e-05, "loss": 0.0206, "step": 65640 }, { "grad_norm": 0.17663148045539856, "learning_rate": 2.89372116381603e-05, "loss": 0.0189, "step": 65650 }, { "grad_norm": 0.3102470934391022, "learning_rate": 2.8922216779836932e-05, "loss": 0.0145, "step": 65660 }, { "grad_norm": 0.3541792929172516, "learning_rate": 2.8907224226548285e-05, "loss": 0.0208, "step": 65670 }, { "grad_norm": 0.29946431517601013, "learning_rate": 2.8892233979933868e-05, "loss": 0.0181, "step": 65680 }, { "grad_norm": 0.2817249596118927, "learning_rate": 2.8877246041633026e-05, "loss": 0.0187, "step": 65690 }, { "grad_norm": 0.2953057885169983, "learning_rate": 2.886226041328478e-05, "loss": 0.0158, "step": 65700 }, { "grad_norm": 0.21931423246860504, "learning_rate": 2.8847277096527957e-05, "loss": 0.0181, "step": 65710 }, { "grad_norm": 0.28595229983329773, "learning_rate": 2.8832296093001105e-05, "loss": 0.0158, "step": 65720 }, { "grad_norm": 0.2708038091659546, "learning_rate": 2.8817317404342547e-05, "loss": 0.0226, "step": 65730 }, { "grad_norm": 0.26141947507858276, "learning_rate": 2.880234103219029e-05, "loss": 0.02, "step": 65740 }, { "grad_norm": 0.3321356773376465, "learning_rate": 2.878736697818215e-05, "loss": 0.0215, "step": 65750 }, { "grad_norm": 0.3274295926094055, "learning_rate": 2.877239524395568e-05, "loss": 0.0195, "step": 65760 }, { "grad_norm": 0.37060046195983887, "learning_rate": 2.8757425831148134e-05, "loss": 0.0206, "step": 65770 }, { "grad_norm": 0.3070657551288605, "learning_rate": 2.874245874139658e-05, "loss": 0.0206, "step": 65780 }, { "grad_norm": 0.17121084034442902, "learning_rate": 2.8727493976337737e-05, "loss": 0.0163, "step": 65790 }, { "grad_norm": 0.2876540422439575, "learning_rate": 2.8712531537608218e-05, "loss": 0.0131, "step": 65800 }, { "grad_norm": 0.29431626200675964, "learning_rate": 2.8697571426844217e-05, "loss": 0.0155, "step": 65810 }, { "grad_norm": 0.28165894746780396, "learning_rate": 2.86826136456818e-05, "loss": 0.0184, "step": 65820 }, { "grad_norm": 0.29424288868904114, "learning_rate": 2.8667658195756685e-05, "loss": 0.0172, "step": 65830 }, { "grad_norm": 0.392223060131073, "learning_rate": 2.8652705078704412e-05, "loss": 0.0259, "step": 65840 }, { "grad_norm": 0.2396213412284851, "learning_rate": 2.863775429616018e-05, "loss": 0.0151, "step": 65850 }, { "grad_norm": 0.2663852274417877, "learning_rate": 2.862280584975906e-05, "loss": 0.0182, "step": 65860 }, { "grad_norm": 0.37729308009147644, "learning_rate": 2.8607859741135735e-05, "loss": 0.0169, "step": 65870 }, { "grad_norm": 0.34278565645217896, "learning_rate": 2.8592915971924723e-05, "loss": 0.0262, "step": 65880 }, { "grad_norm": 0.3104858100414276, "learning_rate": 2.8577974543760212e-05, "loss": 0.017, "step": 65890 }, { "grad_norm": 0.16746315360069275, "learning_rate": 2.8563035458276222e-05, "loss": 0.0192, "step": 65900 }, { "grad_norm": 0.26428720355033875, "learning_rate": 2.8548098717106424e-05, "loss": 0.0282, "step": 65910 }, { "grad_norm": 0.2849508225917816, "learning_rate": 2.85331643218843e-05, "loss": 0.0149, "step": 65920 }, { "grad_norm": 0.5188335180282593, "learning_rate": 2.8518232274243046e-05, "loss": 0.0218, "step": 65930 }, { "grad_norm": 0.3922766149044037, "learning_rate": 2.8503302575815638e-05, "loss": 0.0259, "step": 65940 }, { "grad_norm": 0.28579697012901306, "learning_rate": 2.8488375228234716e-05, "loss": 0.0184, "step": 65950 }, { "grad_norm": 1.179726481437683, "learning_rate": 2.8473450233132738e-05, "loss": 0.0282, "step": 65960 }, { "grad_norm": 0.261425256729126, "learning_rate": 2.8458527592141898e-05, "loss": 0.0167, "step": 65970 }, { "grad_norm": 0.3956659138202667, "learning_rate": 2.8443607306894073e-05, "loss": 0.0305, "step": 65980 }, { "grad_norm": 0.3058694899082184, "learning_rate": 2.8428689379020946e-05, "loss": 0.0209, "step": 65990 }, { "grad_norm": 0.24645894765853882, "learning_rate": 2.841377381015391e-05, "loss": 0.0191, "step": 66000 }, { "grad_norm": 0.20921902358531952, "learning_rate": 2.8398860601924144e-05, "loss": 0.0211, "step": 66010 }, { "grad_norm": 0.41399046778678894, "learning_rate": 2.838394975596248e-05, "loss": 0.0167, "step": 66020 }, { "grad_norm": 0.2560094892978668, "learning_rate": 2.83690412738996e-05, "loss": 0.0234, "step": 66030 }, { "grad_norm": 0.3009384572505951, "learning_rate": 2.835413515736579e-05, "loss": 0.0214, "step": 66040 }, { "grad_norm": 0.64949631690979, "learning_rate": 2.8339231407991267e-05, "loss": 0.024, "step": 66050 }, { "grad_norm": 0.19475404918193817, "learning_rate": 2.832433002740581e-05, "loss": 0.0172, "step": 66060 }, { "grad_norm": 0.20614945888519287, "learning_rate": 2.8309431017239047e-05, "loss": 0.0162, "step": 66070 }, { "grad_norm": 0.2928430140018463, "learning_rate": 2.829453437912028e-05, "loss": 0.0229, "step": 66080 }, { "grad_norm": 0.3048059940338135, "learning_rate": 2.8279640114678617e-05, "loss": 0.0195, "step": 66090 }, { "grad_norm": 0.3091149926185608, "learning_rate": 2.8264748225542836e-05, "loss": 0.0202, "step": 66100 }, { "grad_norm": 0.28410348296165466, "learning_rate": 2.8249858713341516e-05, "loss": 0.0179, "step": 66110 }, { "grad_norm": 0.4346984326839447, "learning_rate": 2.823497157970294e-05, "loss": 0.0181, "step": 66120 }, { "grad_norm": 0.2710632085800171, "learning_rate": 2.8220086826255164e-05, "loss": 0.0163, "step": 66130 }, { "grad_norm": 0.3073345422744751, "learning_rate": 2.8205204454625934e-05, "loss": 0.0189, "step": 66140 }, { "grad_norm": 0.2699163556098938, "learning_rate": 2.8190324466442798e-05, "loss": 0.0196, "step": 66150 }, { "grad_norm": 0.22759655117988586, "learning_rate": 2.8175446863332965e-05, "loss": 0.0183, "step": 66160 }, { "grad_norm": 0.7955381870269775, "learning_rate": 2.816057164692345e-05, "loss": 0.0209, "step": 66170 }, { "grad_norm": 0.23064760863780975, "learning_rate": 2.8145698818840992e-05, "loss": 0.0191, "step": 66180 }, { "grad_norm": 0.2949633300304413, "learning_rate": 2.8130828380712065e-05, "loss": 0.0204, "step": 66190 }, { "grad_norm": 0.4738919734954834, "learning_rate": 2.811596033416285e-05, "loss": 0.0196, "step": 66200 }, { "grad_norm": 0.3395215570926666, "learning_rate": 2.8101094680819307e-05, "loss": 0.023, "step": 66210 }, { "grad_norm": 0.22291609644889832, "learning_rate": 2.808623142230715e-05, "loss": 0.022, "step": 66220 }, { "grad_norm": 0.371746689081192, "learning_rate": 2.8071370560251757e-05, "loss": 0.0213, "step": 66230 }, { "grad_norm": 0.3640212118625641, "learning_rate": 2.805651209627831e-05, "loss": 0.0183, "step": 66240 }, { "grad_norm": 0.2948135435581207, "learning_rate": 2.804165603201171e-05, "loss": 0.0176, "step": 66250 }, { "grad_norm": 0.21316660940647125, "learning_rate": 2.802680236907661e-05, "loss": 0.0181, "step": 66260 }, { "grad_norm": 0.2934064567089081, "learning_rate": 2.8011951109097344e-05, "loss": 0.0164, "step": 66270 }, { "grad_norm": 0.5783756375312805, "learning_rate": 2.7997102253698072e-05, "loss": 0.0228, "step": 66280 }, { "grad_norm": 0.2856064736843109, "learning_rate": 2.7982255804502588e-05, "loss": 0.0172, "step": 66290 }, { "grad_norm": 0.23902073502540588, "learning_rate": 2.7967411763134516e-05, "loss": 0.0185, "step": 66300 }, { "grad_norm": 0.37177494168281555, "learning_rate": 2.7952570131217166e-05, "loss": 0.0202, "step": 66310 }, { "grad_norm": 0.45064833760261536, "learning_rate": 2.7937730910373618e-05, "loss": 0.0224, "step": 66320 }, { "grad_norm": 0.2099834382534027, "learning_rate": 2.7922894102226628e-05, "loss": 0.0198, "step": 66330 }, { "grad_norm": 0.21526630222797394, "learning_rate": 2.790805970839876e-05, "loss": 0.0215, "step": 66340 }, { "grad_norm": 0.44404125213623047, "learning_rate": 2.789322773051225e-05, "loss": 0.0162, "step": 66350 }, { "grad_norm": 0.5401721000671387, "learning_rate": 2.7878398170189124e-05, "loss": 0.0237, "step": 66360 }, { "grad_norm": 0.39599063992500305, "learning_rate": 2.7863571029051112e-05, "loss": 0.0218, "step": 66370 }, { "grad_norm": 0.36978644132614136, "learning_rate": 2.784874630871971e-05, "loss": 0.0215, "step": 66380 }, { "grad_norm": 0.41758936643600464, "learning_rate": 2.7833924010816086e-05, "loss": 0.0135, "step": 66390 }, { "grad_norm": 0.21416489779949188, "learning_rate": 2.7819104136961226e-05, "loss": 0.0227, "step": 66400 }, { "grad_norm": 0.2027151882648468, "learning_rate": 2.780428668877577e-05, "loss": 0.0166, "step": 66410 }, { "grad_norm": 0.24870340526103973, "learning_rate": 2.7789471667880152e-05, "loss": 0.0208, "step": 66420 }, { "grad_norm": 0.2486693561077118, "learning_rate": 2.7774659075894523e-05, "loss": 0.0217, "step": 66430 }, { "grad_norm": 0.2832680344581604, "learning_rate": 2.7759848914438778e-05, "loss": 0.0266, "step": 66440 }, { "grad_norm": 0.24753399193286896, "learning_rate": 2.7745041185132496e-05, "loss": 0.018, "step": 66450 }, { "grad_norm": 0.5652157068252563, "learning_rate": 2.773023588959506e-05, "loss": 0.0201, "step": 66460 }, { "grad_norm": 0.2786242365837097, "learning_rate": 2.7715433029445555e-05, "loss": 0.0223, "step": 66470 }, { "grad_norm": 0.2873835265636444, "learning_rate": 2.7700632606302766e-05, "loss": 0.0193, "step": 66480 }, { "grad_norm": 0.3982633650302887, "learning_rate": 2.768583462178528e-05, "loss": 0.0194, "step": 66490 }, { "grad_norm": 0.1536494642496109, "learning_rate": 2.767103907751136e-05, "loss": 0.0143, "step": 66500 }, { "grad_norm": 0.26644080877304077, "learning_rate": 2.7656245975099065e-05, "loss": 0.0169, "step": 66510 }, { "grad_norm": 0.5415270328521729, "learning_rate": 2.7641455316166093e-05, "loss": 0.0213, "step": 66520 }, { "grad_norm": 0.22681409120559692, "learning_rate": 2.7626667102329973e-05, "loss": 0.0191, "step": 66530 }, { "grad_norm": 0.32532456517219543, "learning_rate": 2.761188133520788e-05, "loss": 0.0211, "step": 66540 }, { "grad_norm": 0.32457613945007324, "learning_rate": 2.7597098016416784e-05, "loss": 0.019, "step": 66550 }, { "grad_norm": 0.1616850048303604, "learning_rate": 2.7582317147573368e-05, "loss": 0.0136, "step": 66560 }, { "grad_norm": 0.2854107618331909, "learning_rate": 2.7567538730294064e-05, "loss": 0.0235, "step": 66570 }, { "grad_norm": 0.2744181156158447, "learning_rate": 2.7552762766194973e-05, "loss": 0.0188, "step": 66580 }, { "grad_norm": 0.34711670875549316, "learning_rate": 2.753798925689202e-05, "loss": 0.0309, "step": 66590 }, { "grad_norm": 0.3592779338359833, "learning_rate": 2.7523218204000774e-05, "loss": 0.0204, "step": 66600 }, { "grad_norm": 0.3913455903530121, "learning_rate": 2.7508449609136612e-05, "loss": 0.0267, "step": 66610 }, { "grad_norm": 0.2661187946796417, "learning_rate": 2.749368347391454e-05, "loss": 0.022, "step": 66620 }, { "grad_norm": 0.25073426961898804, "learning_rate": 2.7478919799949443e-05, "loss": 0.0262, "step": 66630 }, { "grad_norm": 0.25896602869033813, "learning_rate": 2.7464158588855803e-05, "loss": 0.0151, "step": 66640 }, { "grad_norm": 0.43614959716796875, "learning_rate": 2.7449399842247918e-05, "loss": 0.0271, "step": 66650 }, { "grad_norm": 0.37586531043052673, "learning_rate": 2.7434643561739736e-05, "loss": 0.019, "step": 66660 }, { "grad_norm": 0.18387891352176666, "learning_rate": 2.7419889748945016e-05, "loss": 0.0163, "step": 66670 }, { "grad_norm": 0.3239853084087372, "learning_rate": 2.7405138405477203e-05, "loss": 0.0191, "step": 66680 }, { "grad_norm": 0.3012247383594513, "learning_rate": 2.73903895329495e-05, "loss": 0.0205, "step": 66690 }, { "grad_norm": 0.24805378913879395, "learning_rate": 2.737564313297478e-05, "loss": 0.0169, "step": 66700 }, { "grad_norm": 0.24770113825798035, "learning_rate": 2.736089920716572e-05, "loss": 0.0164, "step": 66710 }, { "grad_norm": 0.33624106645584106, "learning_rate": 2.7346157757134706e-05, "loss": 0.0245, "step": 66720 }, { "grad_norm": 0.35602712631225586, "learning_rate": 2.733141878449379e-05, "loss": 0.0179, "step": 66730 }, { "grad_norm": 0.16417132318019867, "learning_rate": 2.7316682290854844e-05, "loss": 0.0192, "step": 66740 }, { "grad_norm": 0.5477456450462341, "learning_rate": 2.7301948277829414e-05, "loss": 0.0231, "step": 66750 }, { "grad_norm": 0.3773319125175476, "learning_rate": 2.7287216747028814e-05, "loss": 0.0215, "step": 66760 }, { "grad_norm": 0.24809560179710388, "learning_rate": 2.7272487700064025e-05, "loss": 0.0202, "step": 66770 }, { "grad_norm": 0.16484643518924713, "learning_rate": 2.725776113854583e-05, "loss": 0.0192, "step": 66780 }, { "grad_norm": 0.6503020524978638, "learning_rate": 2.7243037064084664e-05, "loss": 0.0219, "step": 66790 }, { "grad_norm": 0.44391778111457825, "learning_rate": 2.7228315478290766e-05, "loss": 0.0295, "step": 66800 }, { "grad_norm": 0.22286318242549896, "learning_rate": 2.721359638277402e-05, "loss": 0.0199, "step": 66810 }, { "grad_norm": 0.2723127007484436, "learning_rate": 2.7198879779144148e-05, "loss": 0.0156, "step": 66820 }, { "grad_norm": 0.15499188005924225, "learning_rate": 2.7184165669010485e-05, "loss": 0.0185, "step": 66830 }, { "grad_norm": 0.4447579085826874, "learning_rate": 2.716945405398218e-05, "loss": 0.0231, "step": 66840 }, { "grad_norm": 0.3193671405315399, "learning_rate": 2.715474493566803e-05, "loss": 0.0196, "step": 66850 }, { "grad_norm": 0.2869574725627899, "learning_rate": 2.7140038315676652e-05, "loss": 0.0164, "step": 66860 }, { "grad_norm": 0.28916558623313904, "learning_rate": 2.7125334195616275e-05, "loss": 0.0333, "step": 66870 }, { "grad_norm": 0.31412720680236816, "learning_rate": 2.7110632577094997e-05, "loss": 0.0177, "step": 66880 }, { "grad_norm": 0.2084241360425949, "learning_rate": 2.709593346172051e-05, "loss": 0.0186, "step": 66890 }, { "grad_norm": 0.26694029569625854, "learning_rate": 2.708123685110032e-05, "loss": 0.0192, "step": 66900 }, { "grad_norm": 0.3335808217525482, "learning_rate": 2.7066542746841588e-05, "loss": 0.0191, "step": 66910 }, { "grad_norm": 1.1712297201156616, "learning_rate": 2.7051851150551267e-05, "loss": 0.0182, "step": 66920 }, { "grad_norm": 0.18671253323554993, "learning_rate": 2.7037162063835996e-05, "loss": 0.0168, "step": 66930 }, { "grad_norm": 0.25419920682907104, "learning_rate": 2.702247548830218e-05, "loss": 0.0159, "step": 66940 }, { "grad_norm": 0.1903005689382553, "learning_rate": 2.7007791425555883e-05, "loss": 0.0175, "step": 66950 }, { "grad_norm": 0.24672415852546692, "learning_rate": 2.6993109877202945e-05, "loss": 0.0153, "step": 66960 }, { "grad_norm": 0.26314064860343933, "learning_rate": 2.6978430844848935e-05, "loss": 0.0216, "step": 66970 }, { "grad_norm": 0.34060904383659363, "learning_rate": 2.69637543300991e-05, "loss": 0.0156, "step": 66980 }, { "grad_norm": 0.24678167700767517, "learning_rate": 2.694908033455848e-05, "loss": 0.0198, "step": 66990 }, { "grad_norm": 0.23231087625026703, "learning_rate": 2.6934408859831733e-05, "loss": 0.0172, "step": 67000 }, { "grad_norm": 0.2218596488237381, "learning_rate": 2.69197399075234e-05, "loss": 0.0199, "step": 67010 }, { "grad_norm": 0.3332650661468506, "learning_rate": 2.6905073479237584e-05, "loss": 0.0181, "step": 67020 }, { "grad_norm": 0.2704487442970276, "learning_rate": 2.6890409576578235e-05, "loss": 0.0261, "step": 67030 }, { "grad_norm": 0.3181287944316864, "learning_rate": 2.687574820114892e-05, "loss": 0.0193, "step": 67040 }, { "grad_norm": 0.38019859790802, "learning_rate": 2.686108935455305e-05, "loss": 0.0284, "step": 67050 }, { "grad_norm": 0.2656024694442749, "learning_rate": 2.6846433038393603e-05, "loss": 0.0218, "step": 67060 }, { "grad_norm": 0.29898950457572937, "learning_rate": 2.6831779254273482e-05, "loss": 0.0179, "step": 67070 }, { "grad_norm": 0.31028392910957336, "learning_rate": 2.6817128003795122e-05, "loss": 0.0156, "step": 67080 }, { "grad_norm": 0.2003690004348755, "learning_rate": 2.680247928856081e-05, "loss": 0.0134, "step": 67090 }, { "grad_norm": 0.2961290180683136, "learning_rate": 2.6787833110172466e-05, "loss": 0.0154, "step": 67100 }, { "grad_norm": 0.2569975256919861, "learning_rate": 2.6773189470231807e-05, "loss": 0.0245, "step": 67110 }, { "grad_norm": 0.2872026264667511, "learning_rate": 2.675854837034022e-05, "loss": 0.0184, "step": 67120 }, { "grad_norm": 0.2095770239830017, "learning_rate": 2.6743909812098833e-05, "loss": 0.0207, "step": 67130 }, { "grad_norm": 0.35217946767807007, "learning_rate": 2.67292737971085e-05, "loss": 0.0272, "step": 67140 }, { "grad_norm": 0.2979171574115753, "learning_rate": 2.671464032696982e-05, "loss": 0.0189, "step": 67150 }, { "grad_norm": 0.3171044886112213, "learning_rate": 2.6700009403283055e-05, "loss": 0.0145, "step": 67160 }, { "grad_norm": 0.3035798966884613, "learning_rate": 2.6685381027648214e-05, "loss": 0.0176, "step": 67170 }, { "grad_norm": 0.37585335969924927, "learning_rate": 2.6670755201665086e-05, "loss": 0.0295, "step": 67180 }, { "grad_norm": 0.22893689572811127, "learning_rate": 2.6656131926933066e-05, "loss": 0.0193, "step": 67190 }, { "grad_norm": 0.2828795909881592, "learning_rate": 2.664151120505136e-05, "loss": 0.0157, "step": 67200 }, { "grad_norm": 0.33193641901016235, "learning_rate": 2.6626893037618873e-05, "loss": 0.0209, "step": 67210 }, { "grad_norm": 0.3711513578891754, "learning_rate": 2.6612277426234243e-05, "loss": 0.0202, "step": 67220 }, { "grad_norm": 0.3160250782966614, "learning_rate": 2.659766437249577e-05, "loss": 0.0293, "step": 67230 }, { "grad_norm": 0.20675790309906006, "learning_rate": 2.6583053878001557e-05, "loss": 0.0142, "step": 67240 }, { "grad_norm": 0.47498977184295654, "learning_rate": 2.6568445944349323e-05, "loss": 0.0252, "step": 67250 }, { "grad_norm": 0.24510474503040314, "learning_rate": 2.6553840573136657e-05, "loss": 0.0174, "step": 67260 }, { "grad_norm": 0.5391942858695984, "learning_rate": 2.6539237765960706e-05, "loss": 0.0209, "step": 67270 }, { "grad_norm": 0.2660903334617615, "learning_rate": 2.6524637524418465e-05, "loss": 0.021, "step": 67280 }, { "grad_norm": 0.15324299037456512, "learning_rate": 2.651003985010655e-05, "loss": 0.0189, "step": 67290 }, { "grad_norm": 0.2681022882461548, "learning_rate": 2.649544474462138e-05, "loss": 0.0194, "step": 67300 }, { "grad_norm": 0.32570239901542664, "learning_rate": 2.6480852209559015e-05, "loss": 0.0166, "step": 67310 }, { "grad_norm": 0.3485952913761139, "learning_rate": 2.6466262246515296e-05, "loss": 0.0186, "step": 67320 }, { "grad_norm": 0.22061686217784882, "learning_rate": 2.6451674857085746e-05, "loss": 0.0185, "step": 67330 }, { "grad_norm": 0.26116466522216797, "learning_rate": 2.6437090042865655e-05, "loss": 0.0196, "step": 67340 }, { "grad_norm": 0.243261456489563, "learning_rate": 2.6422507805449947e-05, "loss": 0.0167, "step": 67350 }, { "grad_norm": 0.1872369647026062, "learning_rate": 2.6407928146433357e-05, "loss": 0.0216, "step": 67360 }, { "grad_norm": 0.2739698886871338, "learning_rate": 2.639335106741026e-05, "loss": 0.0168, "step": 67370 }, { "grad_norm": 0.22833184897899628, "learning_rate": 2.6378776569974796e-05, "loss": 0.0219, "step": 67380 }, { "grad_norm": 0.3295900821685791, "learning_rate": 2.6364204655720814e-05, "loss": 0.0152, "step": 67390 }, { "grad_norm": 0.27772942185401917, "learning_rate": 2.6349635326241894e-05, "loss": 0.0181, "step": 67400 }, { "grad_norm": 0.207259863615036, "learning_rate": 2.633506858313127e-05, "loss": 0.0184, "step": 67410 }, { "grad_norm": 0.2261752337217331, "learning_rate": 2.6320504427981975e-05, "loss": 0.021, "step": 67420 }, { "grad_norm": 0.222348153591156, "learning_rate": 2.6305942862386734e-05, "loss": 0.0147, "step": 67430 }, { "grad_norm": 0.3114780783653259, "learning_rate": 2.629138388793793e-05, "loss": 0.0148, "step": 67440 }, { "grad_norm": 0.2752668559551239, "learning_rate": 2.6276827506227743e-05, "loss": 0.0223, "step": 67450 }, { "grad_norm": 0.24353580176830292, "learning_rate": 2.6262273718848037e-05, "loss": 0.017, "step": 67460 }, { "grad_norm": 0.320397287607193, "learning_rate": 2.6247722527390407e-05, "loss": 0.0275, "step": 67470 }, { "grad_norm": 0.23696665465831757, "learning_rate": 2.6233173933446104e-05, "loss": 0.0223, "step": 67480 }, { "grad_norm": 0.2095177322626114, "learning_rate": 2.6218627938606188e-05, "loss": 0.0252, "step": 67490 }, { "grad_norm": 0.38957467675209045, "learning_rate": 2.620408454446135e-05, "loss": 0.0194, "step": 67500 }, { "grad_norm": 0.23009037971496582, "learning_rate": 2.6189543752602046e-05, "loss": 0.0205, "step": 67510 }, { "grad_norm": 0.3566646873950958, "learning_rate": 2.6175005564618444e-05, "loss": 0.0169, "step": 67520 }, { "grad_norm": 0.3160209655761719, "learning_rate": 2.6160469982100428e-05, "loss": 0.0234, "step": 67530 }, { "grad_norm": 0.5837608575820923, "learning_rate": 2.6145937006637555e-05, "loss": 0.0236, "step": 67540 }, { "grad_norm": 0.34460005164146423, "learning_rate": 2.6131406639819166e-05, "loss": 0.0171, "step": 67550 }, { "grad_norm": 0.2081003487110138, "learning_rate": 2.611687888323424e-05, "loss": 0.0207, "step": 67560 }, { "grad_norm": 0.22874815762043, "learning_rate": 2.610235373847153e-05, "loss": 0.0222, "step": 67570 }, { "grad_norm": 0.35878321528434753, "learning_rate": 2.6087831207119494e-05, "loss": 0.0242, "step": 67580 }, { "grad_norm": 0.18009155988693237, "learning_rate": 2.60733112907663e-05, "loss": 0.0245, "step": 67590 }, { "grad_norm": 0.29914161562919617, "learning_rate": 2.6058793990999786e-05, "loss": 0.0286, "step": 67600 }, { "grad_norm": 0.2428458333015442, "learning_rate": 2.6044279309407593e-05, "loss": 0.0176, "step": 67610 }, { "grad_norm": 0.2689598500728607, "learning_rate": 2.6029767247576976e-05, "loss": 0.0228, "step": 67620 }, { "grad_norm": 0.39607617259025574, "learning_rate": 2.601525780709497e-05, "loss": 0.0222, "step": 67630 }, { "grad_norm": 0.4476298689842224, "learning_rate": 2.6000750989548318e-05, "loss": 0.0205, "step": 67640 }, { "grad_norm": 0.337566614151001, "learning_rate": 2.5986246796523462e-05, "loss": 0.0173, "step": 67650 }, { "grad_norm": 0.2688874304294586, "learning_rate": 2.5971745229606546e-05, "loss": 0.0133, "step": 67660 }, { "grad_norm": 0.38233131170272827, "learning_rate": 2.5957246290383442e-05, "loss": 0.028, "step": 67670 }, { "grad_norm": 0.2737202048301697, "learning_rate": 2.594274998043975e-05, "loss": 0.0183, "step": 67680 }, { "grad_norm": 0.2496260106563568, "learning_rate": 2.5928256301360733e-05, "loss": 0.0199, "step": 67690 }, { "grad_norm": 0.28156572580337524, "learning_rate": 2.591376525473142e-05, "loss": 0.0203, "step": 67700 }, { "grad_norm": 0.6025630831718445, "learning_rate": 2.5899276842136523e-05, "loss": 0.0216, "step": 67710 }, { "grad_norm": 0.43878626823425293, "learning_rate": 2.5884791065160495e-05, "loss": 0.026, "step": 67720 }, { "grad_norm": 0.33515995740890503, "learning_rate": 2.5870307925387436e-05, "loss": 0.0191, "step": 67730 }, { "grad_norm": 0.28831636905670166, "learning_rate": 2.585582742440125e-05, "loss": 0.0147, "step": 67740 }, { "grad_norm": 0.2617303133010864, "learning_rate": 2.5841349563785455e-05, "loss": 0.0225, "step": 67750 }, { "grad_norm": 0.2713344395160675, "learning_rate": 2.5826874345123355e-05, "loss": 0.0186, "step": 67760 }, { "grad_norm": 0.22171221673488617, "learning_rate": 2.581240176999792e-05, "loss": 0.0137, "step": 67770 }, { "grad_norm": 0.2690161466598511, "learning_rate": 2.579793183999189e-05, "loss": 0.0218, "step": 67780 }, { "grad_norm": 0.25374358892440796, "learning_rate": 2.5783464556687627e-05, "loss": 0.0197, "step": 67790 }, { "grad_norm": 0.18319526314735413, "learning_rate": 2.576899992166729e-05, "loss": 0.0186, "step": 67800 }, { "grad_norm": 0.2248666137456894, "learning_rate": 2.5754537936512668e-05, "loss": 0.0241, "step": 67810 }, { "grad_norm": 0.49587032198905945, "learning_rate": 2.574007860280535e-05, "loss": 0.0251, "step": 67820 }, { "grad_norm": 0.3290524184703827, "learning_rate": 2.5725621922126518e-05, "loss": 0.0195, "step": 67830 }, { "grad_norm": 0.40213221311569214, "learning_rate": 2.571116789605722e-05, "loss": 0.0175, "step": 67840 }, { "grad_norm": 0.40198519825935364, "learning_rate": 2.5696716526178054e-05, "loss": 0.0243, "step": 67850 }, { "grad_norm": 0.3018926680088043, "learning_rate": 2.568226781406946e-05, "loss": 0.0269, "step": 67860 }, { "grad_norm": 0.33093902468681335, "learning_rate": 2.5667821761311474e-05, "loss": 0.0206, "step": 67870 }, { "grad_norm": 0.23596301674842834, "learning_rate": 2.5653378369483917e-05, "loss": 0.019, "step": 67880 }, { "grad_norm": 0.537131667137146, "learning_rate": 2.5638937640166292e-05, "loss": 0.0173, "step": 67890 }, { "grad_norm": 0.2941918671131134, "learning_rate": 2.5624499574937842e-05, "loss": 0.0183, "step": 67900 }, { "grad_norm": 0.24161377549171448, "learning_rate": 2.5610064175377456e-05, "loss": 0.0253, "step": 67910 }, { "grad_norm": 0.22042585909366608, "learning_rate": 2.5595631443063782e-05, "loss": 0.0184, "step": 67920 }, { "grad_norm": 0.306058794260025, "learning_rate": 2.5581201379575177e-05, "loss": 0.0182, "step": 67930 }, { "grad_norm": 0.17308039963245392, "learning_rate": 2.5566773986489655e-05, "loss": 0.021, "step": 67940 }, { "grad_norm": 0.25601691007614136, "learning_rate": 2.5552349265385e-05, "loss": 0.0276, "step": 67950 }, { "grad_norm": 0.39383038878440857, "learning_rate": 2.5537927217838675e-05, "loss": 0.0222, "step": 67960 }, { "grad_norm": 0.4119652509689331, "learning_rate": 2.5523507845427864e-05, "loss": 0.0227, "step": 67970 }, { "grad_norm": 0.2874496579170227, "learning_rate": 2.5509091149729413e-05, "loss": 0.0164, "step": 67980 }, { "grad_norm": 0.2910393476486206, "learning_rate": 2.5494677132319962e-05, "loss": 0.0239, "step": 67990 }, { "grad_norm": 0.35272467136383057, "learning_rate": 2.548026579477575e-05, "loss": 0.0279, "step": 68000 }, { "grad_norm": 0.38723161816596985, "learning_rate": 2.5465857138672834e-05, "loss": 0.0199, "step": 68010 }, { "grad_norm": 0.2442411631345749, "learning_rate": 2.5451451165586848e-05, "loss": 0.024, "step": 68020 }, { "grad_norm": 0.20632179081439972, "learning_rate": 2.543704787709329e-05, "loss": 0.0159, "step": 68030 }, { "grad_norm": 1.5581650733947754, "learning_rate": 2.542264727476723e-05, "loss": 0.0167, "step": 68040 }, { "grad_norm": 0.2164406180381775, "learning_rate": 2.5408249360183535e-05, "loss": 0.0251, "step": 68050 }, { "grad_norm": 0.2295933961868286, "learning_rate": 2.5393854134916694e-05, "loss": 0.0153, "step": 68060 }, { "grad_norm": 0.1972021758556366, "learning_rate": 2.5379461600540993e-05, "loss": 0.0189, "step": 68070 }, { "grad_norm": 0.3683229088783264, "learning_rate": 2.5365071758630317e-05, "loss": 0.0177, "step": 68080 }, { "grad_norm": 0.454617440700531, "learning_rate": 2.535068461075839e-05, "loss": 0.0255, "step": 68090 }, { "grad_norm": 0.3484230041503906, "learning_rate": 2.5336300158498516e-05, "loss": 0.0238, "step": 68100 }, { "grad_norm": 0.4691184461116791, "learning_rate": 2.5321918403423793e-05, "loss": 0.0207, "step": 68110 }, { "grad_norm": 0.328931987285614, "learning_rate": 2.5307539347106957e-05, "loss": 0.0187, "step": 68120 }, { "grad_norm": 0.3253689706325531, "learning_rate": 2.529316299112049e-05, "loss": 0.0208, "step": 68130 }, { "grad_norm": 0.253264844417572, "learning_rate": 2.5278789337036567e-05, "loss": 0.016, "step": 68140 }, { "grad_norm": 0.34928348660469055, "learning_rate": 2.52644183864271e-05, "loss": 0.0206, "step": 68150 }, { "grad_norm": 0.21815235912799835, "learning_rate": 2.525005014086363e-05, "loss": 0.0184, "step": 68160 }, { "grad_norm": 0.5659304857254028, "learning_rate": 2.5235684601917465e-05, "loss": 0.0243, "step": 68170 }, { "grad_norm": 0.1661432683467865, "learning_rate": 2.5221321771159618e-05, "loss": 0.0164, "step": 68180 }, { "grad_norm": 0.2231474369764328, "learning_rate": 2.5206961650160743e-05, "loss": 0.0189, "step": 68190 }, { "grad_norm": 0.35699722170829773, "learning_rate": 2.519260424049129e-05, "loss": 0.0238, "step": 68200 }, { "grad_norm": 0.3858942687511444, "learning_rate": 2.5178249543721304e-05, "loss": 0.017, "step": 68210 }, { "grad_norm": 0.1927824318408966, "learning_rate": 2.5163897561420668e-05, "loss": 0.0206, "step": 68220 }, { "grad_norm": 0.244846910238266, "learning_rate": 2.514954829515883e-05, "loss": 0.0225, "step": 68230 }, { "grad_norm": 0.31092512607574463, "learning_rate": 2.513520174650505e-05, "loss": 0.0184, "step": 68240 }, { "grad_norm": 0.4066830277442932, "learning_rate": 2.5120857917028207e-05, "loss": 0.0176, "step": 68250 }, { "grad_norm": 0.2551647424697876, "learning_rate": 2.5106516808296953e-05, "loss": 0.0154, "step": 68260 }, { "grad_norm": 0.2301846295595169, "learning_rate": 2.5092178421879552e-05, "loss": 0.0193, "step": 68270 }, { "grad_norm": 0.2725205719470978, "learning_rate": 2.5077842759344116e-05, "loss": 0.03, "step": 68280 }, { "grad_norm": 0.17959845066070557, "learning_rate": 2.5063509822258304e-05, "loss": 0.0214, "step": 68290 }, { "grad_norm": 0.2085958570241928, "learning_rate": 2.5049179612189576e-05, "loss": 0.0197, "step": 68300 }, { "grad_norm": 0.5587625503540039, "learning_rate": 2.5034852130705043e-05, "loss": 0.0231, "step": 68310 }, { "grad_norm": 0.2993853986263275, "learning_rate": 2.5020527379371572e-05, "loss": 0.0186, "step": 68320 }, { "grad_norm": 0.22664058208465576, "learning_rate": 2.500620535975564e-05, "loss": 0.0191, "step": 68330 }, { "grad_norm": 0.23193731904029846, "learning_rate": 2.4991886073423515e-05, "loss": 0.0178, "step": 68340 }, { "grad_norm": 0.44622892141342163, "learning_rate": 2.4977569521941134e-05, "loss": 0.0188, "step": 68350 }, { "grad_norm": 0.22677607834339142, "learning_rate": 2.4963255706874143e-05, "loss": 0.0175, "step": 68360 }, { "grad_norm": 0.2313748002052307, "learning_rate": 2.4948944629787853e-05, "loss": 0.026, "step": 68370 }, { "grad_norm": 0.350570410490036, "learning_rate": 2.493463629224731e-05, "loss": 0.0219, "step": 68380 }, { "grad_norm": 0.322624146938324, "learning_rate": 2.492033069581728e-05, "loss": 0.0178, "step": 68390 }, { "grad_norm": 0.17941462993621826, "learning_rate": 2.4906027842062137e-05, "loss": 0.0183, "step": 68400 }, { "grad_norm": 0.2695440351963043, "learning_rate": 2.48917277325461e-05, "loss": 0.0265, "step": 68410 }, { "grad_norm": 0.22958296537399292, "learning_rate": 2.487743036883295e-05, "loss": 0.0191, "step": 68420 }, { "grad_norm": 0.4005647897720337, "learning_rate": 2.486313575248626e-05, "loss": 0.0226, "step": 68430 }, { "grad_norm": 0.3056768476963043, "learning_rate": 2.4848843885069235e-05, "loss": 0.0174, "step": 68440 }, { "grad_norm": 0.23569726943969727, "learning_rate": 2.483455476814484e-05, "loss": 0.0186, "step": 68450 }, { "grad_norm": 0.3411772549152374, "learning_rate": 2.4820268403275672e-05, "loss": 0.0197, "step": 68460 }, { "grad_norm": 0.39601659774780273, "learning_rate": 2.480598479202413e-05, "loss": 0.0241, "step": 68470 }, { "grad_norm": 0.3343936800956726, "learning_rate": 2.4791703935952193e-05, "loss": 0.0237, "step": 68480 }, { "grad_norm": 0.2985229194164276, "learning_rate": 2.477742583662163e-05, "loss": 0.0244, "step": 68490 }, { "grad_norm": 0.3703922927379608, "learning_rate": 2.476315049559384e-05, "loss": 0.021, "step": 68500 }, { "grad_norm": 0.3772840201854706, "learning_rate": 2.4748877914429987e-05, "loss": 0.0221, "step": 68510 }, { "grad_norm": 0.25637516379356384, "learning_rate": 2.4734608094690857e-05, "loss": 0.0207, "step": 68520 }, { "grad_norm": 0.35329827666282654, "learning_rate": 2.4720341037937006e-05, "loss": 0.0255, "step": 68530 }, { "grad_norm": 0.23943908512592316, "learning_rate": 2.4706076745728645e-05, "loss": 0.0273, "step": 68540 }, { "grad_norm": 0.46969959139823914, "learning_rate": 2.469181521962573e-05, "loss": 0.0264, "step": 68550 }, { "grad_norm": 0.15882593393325806, "learning_rate": 2.467755646118783e-05, "loss": 0.013, "step": 68560 }, { "grad_norm": 0.3409193456172943, "learning_rate": 2.46633004719743e-05, "loss": 0.0246, "step": 68570 }, { "grad_norm": 0.6552261114120483, "learning_rate": 2.4649047253544123e-05, "loss": 0.0171, "step": 68580 }, { "grad_norm": 0.2845366895198822, "learning_rate": 2.4634796807456016e-05, "loss": 0.0209, "step": 68590 }, { "grad_norm": 0.26646724343299866, "learning_rate": 2.46205491352684e-05, "loss": 0.0195, "step": 68600 }, { "grad_norm": 0.29691916704177856, "learning_rate": 2.4606304238539375e-05, "loss": 0.0203, "step": 68610 }, { "grad_norm": 0.26821884512901306, "learning_rate": 2.4592062118826753e-05, "loss": 0.0167, "step": 68620 }, { "grad_norm": 0.39063185453414917, "learning_rate": 2.4577822777688004e-05, "loss": 0.0225, "step": 68630 }, { "grad_norm": 0.3150888979434967, "learning_rate": 2.4563586216680347e-05, "loss": 0.0177, "step": 68640 }, { "grad_norm": 0.23646335303783417, "learning_rate": 2.4549352437360622e-05, "loss": 0.0208, "step": 68650 }, { "grad_norm": 0.28253111243247986, "learning_rate": 2.4535121441285493e-05, "loss": 0.0197, "step": 68660 }, { "grad_norm": 0.3080171048641205, "learning_rate": 2.4520893230011172e-05, "loss": 0.0226, "step": 68670 }, { "grad_norm": 0.1853613704442978, "learning_rate": 2.450666780509368e-05, "loss": 0.0244, "step": 68680 }, { "grad_norm": 0.29828202724456787, "learning_rate": 2.4492445168088656e-05, "loss": 0.0198, "step": 68690 }, { "grad_norm": 0.3097948431968689, "learning_rate": 2.447822532055149e-05, "loss": 0.0146, "step": 68700 }, { "grad_norm": 0.5789671540260315, "learning_rate": 2.4464008264037212e-05, "loss": 0.018, "step": 68710 }, { "grad_norm": 0.2744925618171692, "learning_rate": 2.4449794000100605e-05, "loss": 0.0204, "step": 68720 }, { "grad_norm": 0.32856321334838867, "learning_rate": 2.4435582530296108e-05, "loss": 0.0215, "step": 68730 }, { "grad_norm": 0.34143516421318054, "learning_rate": 2.4421373856177887e-05, "loss": 0.0197, "step": 68740 }, { "grad_norm": 0.28785568475723267, "learning_rate": 2.440716797929975e-05, "loss": 0.0217, "step": 68750 }, { "grad_norm": 0.2600513994693756, "learning_rate": 2.439296490121526e-05, "loss": 0.0161, "step": 68760 }, { "grad_norm": 0.260913223028183, "learning_rate": 2.4378764623477608e-05, "loss": 0.0156, "step": 68770 }, { "grad_norm": 0.38611724972724915, "learning_rate": 2.436456714763974e-05, "loss": 0.0171, "step": 68780 }, { "grad_norm": 0.2620750367641449, "learning_rate": 2.4350372475254262e-05, "loss": 0.0223, "step": 68790 }, { "grad_norm": 0.1861027181148529, "learning_rate": 2.433618060787351e-05, "loss": 0.0183, "step": 68800 }, { "grad_norm": 0.23309408128261566, "learning_rate": 2.432199154704944e-05, "loss": 0.0264, "step": 68810 }, { "grad_norm": 0.2087586671113968, "learning_rate": 2.430780529433377e-05, "loss": 0.0159, "step": 68820 }, { "grad_norm": 0.3091169595718384, "learning_rate": 2.42936218512779e-05, "loss": 0.028, "step": 68830 }, { "grad_norm": 0.2801132798194885, "learning_rate": 2.427944121943288e-05, "loss": 0.0199, "step": 68840 }, { "grad_norm": 0.1757907271385193, "learning_rate": 2.4265263400349497e-05, "loss": 0.0193, "step": 68850 }, { "grad_norm": 0.47609803080558777, "learning_rate": 2.4251088395578213e-05, "loss": 0.0206, "step": 68860 }, { "grad_norm": 0.4803657829761505, "learning_rate": 2.4236916206669214e-05, "loss": 0.0171, "step": 68870 }, { "grad_norm": 0.3266128599643707, "learning_rate": 2.4222746835172304e-05, "loss": 0.0185, "step": 68880 }, { "grad_norm": 0.32858455181121826, "learning_rate": 2.4208580282637065e-05, "loss": 0.02, "step": 68890 }, { "grad_norm": 0.3169168531894684, "learning_rate": 2.4194416550612687e-05, "loss": 0.0142, "step": 68900 }, { "grad_norm": 0.30029964447021484, "learning_rate": 2.418025564064812e-05, "loss": 0.0188, "step": 68910 }, { "grad_norm": 0.13877281546592712, "learning_rate": 2.4166097554291978e-05, "loss": 0.0257, "step": 68920 }, { "grad_norm": 0.990418016910553, "learning_rate": 2.4151942293092583e-05, "loss": 0.0218, "step": 68930 }, { "grad_norm": 0.24871014058589935, "learning_rate": 2.41377898585979e-05, "loss": 0.0191, "step": 68940 }, { "grad_norm": 0.47511976957321167, "learning_rate": 2.4123640252355652e-05, "loss": 0.0237, "step": 68950 }, { "grad_norm": 0.29726898670196533, "learning_rate": 2.410949347591318e-05, "loss": 0.0191, "step": 68960 }, { "grad_norm": 0.586586058139801, "learning_rate": 2.4095349530817578e-05, "loss": 0.0338, "step": 68970 }, { "grad_norm": 0.24220840632915497, "learning_rate": 2.4081208418615603e-05, "loss": 0.0225, "step": 68980 }, { "grad_norm": 0.31267762184143066, "learning_rate": 2.4067070140853725e-05, "loss": 0.0185, "step": 68990 }, { "grad_norm": 0.28269052505493164, "learning_rate": 2.405293469907805e-05, "loss": 0.0194, "step": 69000 }, { "grad_norm": 0.3065652847290039, "learning_rate": 2.4038802094834446e-05, "loss": 0.0217, "step": 69010 }, { "grad_norm": 0.268749475479126, "learning_rate": 2.4024672329668395e-05, "loss": 0.0185, "step": 69020 }, { "grad_norm": 0.27021488547325134, "learning_rate": 2.4010545405125125e-05, "loss": 0.0168, "step": 69030 }, { "grad_norm": 0.3011727035045624, "learning_rate": 2.3996421322749528e-05, "loss": 0.0188, "step": 69040 }, { "grad_norm": 0.3892870545387268, "learning_rate": 2.3982300084086224e-05, "loss": 0.0232, "step": 69050 }, { "grad_norm": 1.1486622095108032, "learning_rate": 2.396818169067944e-05, "loss": 0.0178, "step": 69060 }, { "grad_norm": 0.30990511178970337, "learning_rate": 2.3954066144073178e-05, "loss": 0.0214, "step": 69070 }, { "grad_norm": 0.4490416347980499, "learning_rate": 2.39399534458111e-05, "loss": 0.0196, "step": 69080 }, { "grad_norm": 0.43924105167388916, "learning_rate": 2.392584359743651e-05, "loss": 0.0169, "step": 69090 }, { "grad_norm": 0.25254425406455994, "learning_rate": 2.3911736600492463e-05, "loss": 0.0149, "step": 69100 }, { "grad_norm": 0.4577377736568451, "learning_rate": 2.3897632456521672e-05, "loss": 0.0178, "step": 69110 }, { "grad_norm": 0.23989146947860718, "learning_rate": 2.388353116706658e-05, "loss": 0.0223, "step": 69120 }, { "grad_norm": 0.2206391841173172, "learning_rate": 2.3869432733669232e-05, "loss": 0.0239, "step": 69130 }, { "grad_norm": 0.40079832077026367, "learning_rate": 2.385533715787145e-05, "loss": 0.0213, "step": 69140 }, { "grad_norm": 0.22563767433166504, "learning_rate": 2.3841244441214673e-05, "loss": 0.0178, "step": 69150 }, { "grad_norm": 0.3275146484375, "learning_rate": 2.3827154585240073e-05, "loss": 0.017, "step": 69160 }, { "grad_norm": 0.3282550871372223, "learning_rate": 2.3813067591488498e-05, "loss": 0.0232, "step": 69170 }, { "grad_norm": 0.26468971371650696, "learning_rate": 2.3798983461500494e-05, "loss": 0.0208, "step": 69180 }, { "grad_norm": 0.2772654592990875, "learning_rate": 2.3784902196816256e-05, "loss": 0.0177, "step": 69190 }, { "grad_norm": 0.23468533158302307, "learning_rate": 2.3770823798975716e-05, "loss": 0.0218, "step": 69200 }, { "grad_norm": 0.40850362181663513, "learning_rate": 2.3756748269518424e-05, "loss": 0.0173, "step": 69210 }, { "grad_norm": 0.2622707486152649, "learning_rate": 2.374267560998371e-05, "loss": 0.0193, "step": 69220 }, { "grad_norm": 0.3923466205596924, "learning_rate": 2.3728605821910477e-05, "loss": 0.0195, "step": 69230 }, { "grad_norm": 0.5440759062767029, "learning_rate": 2.3714538906837453e-05, "loss": 0.0289, "step": 69240 }, { "grad_norm": 0.34883877635002136, "learning_rate": 2.3700474866302913e-05, "loss": 0.0178, "step": 69250 }, { "grad_norm": 0.2054998129606247, "learning_rate": 2.3686413701844917e-05, "loss": 0.0181, "step": 69260 }, { "grad_norm": 0.25783881545066833, "learning_rate": 2.367235541500114e-05, "loss": 0.019, "step": 69270 }, { "grad_norm": 0.7663415670394897, "learning_rate": 2.3658300007308993e-05, "loss": 0.0166, "step": 69280 }, { "grad_norm": 0.20326459407806396, "learning_rate": 2.364424748030555e-05, "loss": 0.0117, "step": 69290 }, { "grad_norm": 0.23477981984615326, "learning_rate": 2.3630197835527595e-05, "loss": 0.0211, "step": 69300 }, { "grad_norm": 0.37673747539520264, "learning_rate": 2.3616151074511546e-05, "loss": 0.0274, "step": 69310 }, { "grad_norm": 0.6123353838920593, "learning_rate": 2.3602107198793545e-05, "loss": 0.0176, "step": 69320 }, { "grad_norm": 0.42369481921195984, "learning_rate": 2.3588066209909427e-05, "loss": 0.0221, "step": 69330 }, { "grad_norm": 0.5204058289527893, "learning_rate": 2.3574028109394662e-05, "loss": 0.0237, "step": 69340 }, { "grad_norm": 0.3379078805446625, "learning_rate": 2.3559992898784454e-05, "loss": 0.0252, "step": 69350 }, { "grad_norm": 0.1894274353981018, "learning_rate": 2.354596057961367e-05, "loss": 0.0143, "step": 69360 }, { "grad_norm": 0.2861477732658386, "learning_rate": 2.353193115341688e-05, "loss": 0.028, "step": 69370 }, { "grad_norm": 0.41648173332214355, "learning_rate": 2.3517904621728294e-05, "loss": 0.0231, "step": 69380 }, { "grad_norm": 0.23415830731391907, "learning_rate": 2.3503880986081855e-05, "loss": 0.0271, "step": 69390 }, { "grad_norm": 0.21648332476615906, "learning_rate": 2.3489860248011136e-05, "loss": 0.0176, "step": 69400 }, { "grad_norm": 0.3540962338447571, "learning_rate": 2.3475842409049464e-05, "loss": 0.0223, "step": 69410 }, { "grad_norm": 0.33062508702278137, "learning_rate": 2.346182747072975e-05, "loss": 0.021, "step": 69420 }, { "grad_norm": 0.2526485323905945, "learning_rate": 2.3447815434584718e-05, "loss": 0.0183, "step": 69430 }, { "grad_norm": 0.4238066077232361, "learning_rate": 2.3433806302146656e-05, "loss": 0.0192, "step": 69440 }, { "grad_norm": 0.4559951722621918, "learning_rate": 2.3419800074947612e-05, "loss": 0.0147, "step": 69450 }, { "grad_norm": 0.21772533655166626, "learning_rate": 2.3405796754519248e-05, "loss": 0.0172, "step": 69460 }, { "grad_norm": 0.3313491940498352, "learning_rate": 2.3391796342392992e-05, "loss": 0.0221, "step": 69470 }, { "grad_norm": 0.5276392102241516, "learning_rate": 2.337779884009984e-05, "loss": 0.0239, "step": 69480 }, { "grad_norm": 0.24889378249645233, "learning_rate": 2.3363804249170624e-05, "loss": 0.0212, "step": 69490 }, { "grad_norm": 0.38315171003341675, "learning_rate": 2.3349812571135703e-05, "loss": 0.0137, "step": 69500 }, { "grad_norm": 0.3146907687187195, "learning_rate": 2.333582380752523e-05, "loss": 0.022, "step": 69510 }, { "grad_norm": 0.27868664264678955, "learning_rate": 2.3321837959868963e-05, "loss": 0.0169, "step": 69520 }, { "grad_norm": 0.18422840535640717, "learning_rate": 2.3307855029696383e-05, "loss": 0.0152, "step": 69530 }, { "grad_norm": 0.32160505652427673, "learning_rate": 2.3293875018536648e-05, "loss": 0.0172, "step": 69540 }, { "grad_norm": 0.17854389548301697, "learning_rate": 2.327989792791861e-05, "loss": 0.0164, "step": 69550 }, { "grad_norm": 0.37804970145225525, "learning_rate": 2.326592375937074e-05, "loss": 0.0199, "step": 69560 }, { "grad_norm": 0.4007396101951599, "learning_rate": 2.3251952514421248e-05, "loss": 0.0159, "step": 69570 }, { "grad_norm": 0.3113359212875366, "learning_rate": 2.323798419459804e-05, "loss": 0.0187, "step": 69580 }, { "grad_norm": 0.2211443930864334, "learning_rate": 2.3224018801428625e-05, "loss": 0.0154, "step": 69590 }, { "grad_norm": 0.49055659770965576, "learning_rate": 2.3210056336440273e-05, "loss": 0.0204, "step": 69600 }, { "grad_norm": 0.17731407284736633, "learning_rate": 2.3196096801159844e-05, "loss": 0.0173, "step": 69610 }, { "grad_norm": 0.32900673151016235, "learning_rate": 2.3182140197114012e-05, "loss": 0.0177, "step": 69620 }, { "grad_norm": 0.3715018630027771, "learning_rate": 2.3168186525828985e-05, "loss": 0.0251, "step": 69630 }, { "grad_norm": 0.26348668336868286, "learning_rate": 2.3154235788830757e-05, "loss": 0.0183, "step": 69640 }, { "grad_norm": 0.6341849565505981, "learning_rate": 2.3140287987644927e-05, "loss": 0.0218, "step": 69650 }, { "grad_norm": 0.34937942028045654, "learning_rate": 2.3126343123796833e-05, "loss": 0.0197, "step": 69660 }, { "grad_norm": 0.319610595703125, "learning_rate": 2.3112401198811418e-05, "loss": 0.0256, "step": 69670 }, { "grad_norm": 0.36403658986091614, "learning_rate": 2.3098462214213418e-05, "loss": 0.0215, "step": 69680 }, { "grad_norm": 0.18413402140140533, "learning_rate": 2.308452617152712e-05, "loss": 0.0156, "step": 69690 }, { "grad_norm": 0.3312254548072815, "learning_rate": 2.3070593072276603e-05, "loss": 0.0229, "step": 69700 }, { "grad_norm": 0.1635832041501999, "learning_rate": 2.3056662917985518e-05, "loss": 0.017, "step": 69710 }, { "grad_norm": 0.15109901130199432, "learning_rate": 2.3042735710177283e-05, "loss": 0.0281, "step": 69720 }, { "grad_norm": 0.29819101095199585, "learning_rate": 2.302881145037492e-05, "loss": 0.0188, "step": 69730 }, { "grad_norm": 0.32963770627975464, "learning_rate": 2.3014890140101192e-05, "loss": 0.0208, "step": 69740 }, { "grad_norm": 0.715839684009552, "learning_rate": 2.30009717808785e-05, "loss": 0.0194, "step": 69750 }, { "grad_norm": 0.23747026920318604, "learning_rate": 2.2987056374228967e-05, "loss": 0.0167, "step": 69760 }, { "grad_norm": 0.46714144945144653, "learning_rate": 2.297314392167431e-05, "loss": 0.0179, "step": 69770 }, { "grad_norm": 0.4442415237426758, "learning_rate": 2.295923442473601e-05, "loss": 0.0194, "step": 69780 }, { "grad_norm": 0.2637367844581604, "learning_rate": 2.294532788493519e-05, "loss": 0.0173, "step": 69790 }, { "grad_norm": 0.22744999825954437, "learning_rate": 2.293142430379262e-05, "loss": 0.0162, "step": 69800 }, { "grad_norm": 0.17036394774913788, "learning_rate": 2.291752368282879e-05, "loss": 0.0187, "step": 69810 }, { "grad_norm": 0.30386215448379517, "learning_rate": 2.290362602356385e-05, "loss": 0.0159, "step": 69820 }, { "grad_norm": 0.26707005500793457, "learning_rate": 2.288973132751765e-05, "loss": 0.0153, "step": 69830 }, { "grad_norm": 0.26670265197753906, "learning_rate": 2.287583959620965e-05, "loss": 0.0234, "step": 69840 }, { "grad_norm": 0.3753073215484619, "learning_rate": 2.2861950831159074e-05, "loss": 0.0251, "step": 69850 }, { "grad_norm": 0.29622682929039, "learning_rate": 2.284806503388472e-05, "loss": 0.0189, "step": 69860 }, { "grad_norm": 0.2307002693414688, "learning_rate": 2.283418220590518e-05, "loss": 0.024, "step": 69870 }, { "grad_norm": 0.40877363085746765, "learning_rate": 2.282030234873862e-05, "loss": 0.02, "step": 69880 }, { "grad_norm": 0.21400192379951477, "learning_rate": 2.2806425463902948e-05, "loss": 0.0233, "step": 69890 }, { "grad_norm": 0.3483773171901703, "learning_rate": 2.279255155291568e-05, "loss": 0.0207, "step": 69900 }, { "grad_norm": 0.2561236023902893, "learning_rate": 2.277868061729409e-05, "loss": 0.0137, "step": 69910 }, { "grad_norm": 0.23095513880252838, "learning_rate": 2.276481265855504e-05, "loss": 0.0211, "step": 69920 }, { "grad_norm": 0.3367977738380432, "learning_rate": 2.2750947678215128e-05, "loss": 0.021, "step": 69930 }, { "grad_norm": 0.29554593563079834, "learning_rate": 2.2737085677790616e-05, "loss": 0.0183, "step": 69940 }, { "grad_norm": 0.24481606483459473, "learning_rate": 2.2723226658797443e-05, "loss": 0.0159, "step": 69950 }, { "grad_norm": 0.2677515149116516, "learning_rate": 2.270937062275117e-05, "loss": 0.0203, "step": 69960 }, { "grad_norm": 0.3060990273952484, "learning_rate": 2.2695517571167113e-05, "loss": 0.0207, "step": 69970 }, { "grad_norm": 0.2958669364452362, "learning_rate": 2.268166750556019e-05, "loss": 0.0155, "step": 69980 }, { "grad_norm": 0.2960006892681122, "learning_rate": 2.266782042744503e-05, "loss": 0.0132, "step": 69990 }, { "grad_norm": 0.22206246852874756, "learning_rate": 2.2653976338335935e-05, "loss": 0.0111, "step": 70000 }, { "grad_norm": 0.3564394414424896, "learning_rate": 2.2640135239746895e-05, "loss": 0.017, "step": 70010 }, { "grad_norm": 0.15893429517745972, "learning_rate": 2.2626297133191515e-05, "loss": 0.0158, "step": 70020 }, { "grad_norm": 0.24037741124629974, "learning_rate": 2.261246202018312e-05, "loss": 0.0159, "step": 70030 }, { "grad_norm": 0.1330816149711609, "learning_rate": 2.2598629902234718e-05, "loss": 0.023, "step": 70040 }, { "grad_norm": 0.26357948780059814, "learning_rate": 2.258480078085894e-05, "loss": 0.0203, "step": 70050 }, { "grad_norm": 0.31166478991508484, "learning_rate": 2.2570974657568126e-05, "loss": 0.0178, "step": 70060 }, { "grad_norm": 0.2669612765312195, "learning_rate": 2.2557151533874286e-05, "loss": 0.0182, "step": 70070 }, { "grad_norm": 0.2849476635456085, "learning_rate": 2.2543331411289116e-05, "loss": 0.0151, "step": 70080 }, { "grad_norm": 0.19273528456687927, "learning_rate": 2.252951429132392e-05, "loss": 0.0189, "step": 70090 }, { "grad_norm": 1.0883300304412842, "learning_rate": 2.2515700175489762e-05, "loss": 0.0229, "step": 70100 }, { "grad_norm": 0.5166059136390686, "learning_rate": 2.2501889065297288e-05, "loss": 0.0192, "step": 70110 }, { "grad_norm": 0.34975725412368774, "learning_rate": 2.2488080962256887e-05, "loss": 0.0218, "step": 70120 }, { "grad_norm": 0.2741130590438843, "learning_rate": 2.2474275867878585e-05, "loss": 0.0194, "step": 70130 }, { "grad_norm": 0.30660736560821533, "learning_rate": 2.246047378367211e-05, "loss": 0.0227, "step": 70140 }, { "grad_norm": 0.31214237213134766, "learning_rate": 2.2446674711146798e-05, "loss": 0.0195, "step": 70150 }, { "grad_norm": 0.48888760805130005, "learning_rate": 2.2432878651811734e-05, "loss": 0.0188, "step": 70160 }, { "grad_norm": 0.27501484751701355, "learning_rate": 2.2419085607175594e-05, "loss": 0.023, "step": 70170 }, { "grad_norm": 0.3389526307582855, "learning_rate": 2.240529557874678e-05, "loss": 0.0269, "step": 70180 }, { "grad_norm": 0.2873563766479492, "learning_rate": 2.2391508568033358e-05, "loss": 0.016, "step": 70190 }, { "grad_norm": 0.4299892485141754, "learning_rate": 2.237772457654307e-05, "loss": 0.0182, "step": 70200 }, { "grad_norm": 0.4089783728122711, "learning_rate": 2.236394360578327e-05, "loss": 0.0223, "step": 70210 }, { "grad_norm": 0.21557633578777313, "learning_rate": 2.2350165657261073e-05, "loss": 0.0178, "step": 70220 }, { "grad_norm": 0.1875888556241989, "learning_rate": 2.2336390732483164e-05, "loss": 0.0186, "step": 70230 }, { "grad_norm": 0.19723238050937653, "learning_rate": 2.2322618832955976e-05, "loss": 0.0221, "step": 70240 }, { "grad_norm": 0.26325538754463196, "learning_rate": 2.2308849960185578e-05, "loss": 0.0202, "step": 70250 }, { "grad_norm": 0.6908287405967712, "learning_rate": 2.229508411567774e-05, "loss": 0.0176, "step": 70260 }, { "grad_norm": 0.2775804102420807, "learning_rate": 2.2281321300937823e-05, "loss": 0.0176, "step": 70270 }, { "grad_norm": 0.299280047416687, "learning_rate": 2.2267561517470937e-05, "loss": 0.0133, "step": 70280 }, { "grad_norm": 0.8995985984802246, "learning_rate": 2.2253804766781845e-05, "loss": 0.0227, "step": 70290 }, { "grad_norm": 0.3111381530761719, "learning_rate": 2.2240051050374933e-05, "loss": 0.0156, "step": 70300 }, { "grad_norm": 0.29324331879615784, "learning_rate": 2.22263003697543e-05, "loss": 0.0185, "step": 70310 }, { "grad_norm": 0.3316953480243683, "learning_rate": 2.2212552726423702e-05, "loss": 0.0209, "step": 70320 }, { "grad_norm": 0.3193091154098511, "learning_rate": 2.2198808121886578e-05, "loss": 0.0175, "step": 70330 }, { "grad_norm": 0.22415579855442047, "learning_rate": 2.218506655764598e-05, "loss": 0.0192, "step": 70340 }, { "grad_norm": 0.2112867534160614, "learning_rate": 2.2171328035204704e-05, "loss": 0.0199, "step": 70350 }, { "grad_norm": 0.362601101398468, "learning_rate": 2.2157592556065133e-05, "loss": 0.0189, "step": 70360 }, { "grad_norm": 0.2420499175786972, "learning_rate": 2.2143860121729382e-05, "loss": 0.0159, "step": 70370 }, { "grad_norm": 0.17506764829158783, "learning_rate": 2.2130130733699206e-05, "loss": 0.0165, "step": 70380 }, { "grad_norm": 0.25139179825782776, "learning_rate": 2.2116404393476052e-05, "loss": 0.016, "step": 70390 }, { "grad_norm": 0.21747103333473206, "learning_rate": 2.2102681102560967e-05, "loss": 0.0155, "step": 70400 }, { "grad_norm": 0.22998766601085663, "learning_rate": 2.208896086245476e-05, "loss": 0.0152, "step": 70410 }, { "grad_norm": 0.17612870037555695, "learning_rate": 2.2075243674657808e-05, "loss": 0.0139, "step": 70420 }, { "grad_norm": 0.26648885011672974, "learning_rate": 2.206152954067024e-05, "loss": 0.0168, "step": 70430 }, { "grad_norm": 0.3237955868244171, "learning_rate": 2.2047818461991758e-05, "loss": 0.0228, "step": 70440 }, { "grad_norm": 0.6692147254943848, "learning_rate": 2.2034110440121864e-05, "loss": 0.0204, "step": 70450 }, { "grad_norm": 0.3336845934391022, "learning_rate": 2.2020405476559585e-05, "loss": 0.0221, "step": 70460 }, { "grad_norm": 0.42953506112098694, "learning_rate": 2.2006703572803722e-05, "loss": 0.023, "step": 70470 }, { "grad_norm": 0.24145689606666565, "learning_rate": 2.1993004730352646e-05, "loss": 0.0153, "step": 70480 }, { "grad_norm": 0.28348255157470703, "learning_rate": 2.1979308950704464e-05, "loss": 0.0193, "step": 70490 }, { "grad_norm": 0.26339012384414673, "learning_rate": 2.1965616235356923e-05, "loss": 0.021, "step": 70500 }, { "grad_norm": 0.24442481994628906, "learning_rate": 2.1951926585807465e-05, "loss": 0.0138, "step": 70510 }, { "grad_norm": 0.24400682747364044, "learning_rate": 2.193824000355313e-05, "loss": 0.0199, "step": 70520 }, { "grad_norm": 0.29244759678840637, "learning_rate": 2.192455649009067e-05, "loss": 0.0147, "step": 70530 }, { "grad_norm": 0.5326441526412964, "learning_rate": 2.1910876046916523e-05, "loss": 0.0229, "step": 70540 }, { "grad_norm": 0.4793338179588318, "learning_rate": 2.1897198675526725e-05, "loss": 0.0208, "step": 70550 }, { "grad_norm": 0.31491121649742126, "learning_rate": 2.1883524377417024e-05, "loss": 0.0196, "step": 70560 }, { "grad_norm": 0.3349579870700836, "learning_rate": 2.1869853154082825e-05, "loss": 0.0189, "step": 70570 }, { "grad_norm": 0.17340990900993347, "learning_rate": 2.1856185007019215e-05, "loss": 0.0183, "step": 70580 }, { "grad_norm": 0.47907888889312744, "learning_rate": 2.1842519937720874e-05, "loss": 0.0223, "step": 70590 }, { "grad_norm": 0.2577754259109497, "learning_rate": 2.182885794768224e-05, "loss": 0.018, "step": 70600 }, { "grad_norm": 0.2895599901676178, "learning_rate": 2.1815199038397326e-05, "loss": 0.0156, "step": 70610 }, { "grad_norm": 0.2525770366191864, "learning_rate": 2.1801543211359893e-05, "loss": 0.023, "step": 70620 }, { "grad_norm": 0.33160847425460815, "learning_rate": 2.178789046806326e-05, "loss": 0.017, "step": 70630 }, { "grad_norm": 0.4468347728252411, "learning_rate": 2.1774240810000547e-05, "loss": 0.0187, "step": 70640 }, { "grad_norm": 0.2949541509151459, "learning_rate": 2.17605942386644e-05, "loss": 0.0214, "step": 70650 }, { "grad_norm": 1.0020616054534912, "learning_rate": 2.174695075554723e-05, "loss": 0.0283, "step": 70660 }, { "grad_norm": 0.4325970411300659, "learning_rate": 2.173331036214103e-05, "loss": 0.033, "step": 70670 }, { "grad_norm": 0.24157193303108215, "learning_rate": 2.171967305993753e-05, "loss": 0.0236, "step": 70680 }, { "grad_norm": 0.20146740972995758, "learning_rate": 2.1706038850428024e-05, "loss": 0.0154, "step": 70690 }, { "grad_norm": 0.3569587469100952, "learning_rate": 2.169240773510361e-05, "loss": 0.0166, "step": 70700 }, { "grad_norm": 0.20562495291233063, "learning_rate": 2.1678779715454906e-05, "loss": 0.0144, "step": 70710 }, { "grad_norm": 0.2591053247451782, "learning_rate": 2.166515479297229e-05, "loss": 0.0167, "step": 70720 }, { "grad_norm": 0.19106651842594147, "learning_rate": 2.1651532969145732e-05, "loss": 0.0145, "step": 70730 }, { "grad_norm": 0.2926848828792572, "learning_rate": 2.1637914245464903e-05, "loss": 0.0286, "step": 70740 }, { "grad_norm": 0.5011599659919739, "learning_rate": 2.1624298623419133e-05, "loss": 0.0188, "step": 70750 }, { "grad_norm": 0.3590150475502014, "learning_rate": 2.1610686104497413e-05, "loss": 0.0201, "step": 70760 }, { "grad_norm": 0.7183102369308472, "learning_rate": 2.1597076690188366e-05, "loss": 0.0288, "step": 70770 }, { "grad_norm": 0.3405914306640625, "learning_rate": 2.1583470381980315e-05, "loss": 0.0208, "step": 70780 }, { "grad_norm": 0.18230797350406647, "learning_rate": 2.156986718136123e-05, "loss": 0.0283, "step": 70790 }, { "grad_norm": 0.39042359590530396, "learning_rate": 2.155626708981871e-05, "loss": 0.019, "step": 70800 }, { "grad_norm": 0.23595653474330902, "learning_rate": 2.1542670108840075e-05, "loss": 0.0157, "step": 70810 }, { "grad_norm": 0.3809085786342621, "learning_rate": 2.152907623991221e-05, "loss": 0.0196, "step": 70820 }, { "grad_norm": 0.39043042063713074, "learning_rate": 2.1515485484521802e-05, "loss": 0.0184, "step": 70830 }, { "grad_norm": 0.1847270429134369, "learning_rate": 2.150189784415506e-05, "loss": 0.0175, "step": 70840 }, { "grad_norm": 0.31037282943725586, "learning_rate": 2.148831332029794e-05, "loss": 0.0172, "step": 70850 }, { "grad_norm": 0.18041585385799408, "learning_rate": 2.147473191443599e-05, "loss": 0.0183, "step": 70860 }, { "grad_norm": 0.1962151974439621, "learning_rate": 2.1461153628054492e-05, "loss": 0.0205, "step": 70870 }, { "grad_norm": 0.513725757598877, "learning_rate": 2.1447578462638284e-05, "loss": 0.0232, "step": 70880 }, { "grad_norm": 0.23273305594921112, "learning_rate": 2.143400641967201e-05, "loss": 0.0196, "step": 70890 }, { "grad_norm": 1.1602708101272583, "learning_rate": 2.1420437500639823e-05, "loss": 0.0151, "step": 70900 }, { "grad_norm": 0.1978444755077362, "learning_rate": 2.1406871707025646e-05, "loss": 0.0219, "step": 70910 }, { "grad_norm": 0.24310359358787537, "learning_rate": 2.1393309040312963e-05, "loss": 0.0263, "step": 70920 }, { "grad_norm": 0.24791847169399261, "learning_rate": 2.137974950198502e-05, "loss": 0.0189, "step": 70930 }, { "grad_norm": 0.29900628328323364, "learning_rate": 2.1366193093524617e-05, "loss": 0.0199, "step": 70940 }, { "grad_norm": 0.4494020342826843, "learning_rate": 2.1352639816414288e-05, "loss": 0.0155, "step": 70950 }, { "grad_norm": 0.31887468695640564, "learning_rate": 2.1339089672136202e-05, "loss": 0.0217, "step": 70960 }, { "grad_norm": 0.2560453414916992, "learning_rate": 2.1325542662172193e-05, "loss": 0.017, "step": 70970 }, { "grad_norm": 0.23887400329113007, "learning_rate": 2.1311998788003712e-05, "loss": 0.0179, "step": 70980 }, { "grad_norm": 0.2719721794128418, "learning_rate": 2.1298458051111907e-05, "loss": 0.0166, "step": 70990 }, { "grad_norm": 0.2719492018222809, "learning_rate": 2.1284920452977602e-05, "loss": 0.0185, "step": 71000 }, { "grad_norm": 0.3003121614456177, "learning_rate": 2.12713859950812e-05, "loss": 0.0186, "step": 71010 }, { "grad_norm": 0.2725377380847931, "learning_rate": 2.1257854678902832e-05, "loss": 0.0156, "step": 71020 }, { "grad_norm": 0.2831055521965027, "learning_rate": 2.124432650592227e-05, "loss": 0.0192, "step": 71030 }, { "grad_norm": 0.29819318652153015, "learning_rate": 2.1230801477618937e-05, "loss": 0.0202, "step": 71040 }, { "grad_norm": 0.2816833257675171, "learning_rate": 2.121727959547189e-05, "loss": 0.0246, "step": 71050 }, { "grad_norm": 0.2824275493621826, "learning_rate": 2.1203760860959886e-05, "loss": 0.0178, "step": 71060 }, { "grad_norm": 0.2314707338809967, "learning_rate": 2.119024527556127e-05, "loss": 0.0206, "step": 71070 }, { "grad_norm": 0.2221374362707138, "learning_rate": 2.117673284075415e-05, "loss": 0.0162, "step": 71080 }, { "grad_norm": 0.19258148968219757, "learning_rate": 2.1163223558016165e-05, "loss": 0.0149, "step": 71090 }, { "grad_norm": 0.1507001370191574, "learning_rate": 2.1149717428824722e-05, "loss": 0.0184, "step": 71100 }, { "grad_norm": 0.343954473733902, "learning_rate": 2.1136214454656784e-05, "loss": 0.0221, "step": 71110 }, { "grad_norm": 0.2989552915096283, "learning_rate": 2.1122714636989056e-05, "loss": 0.0191, "step": 71120 }, { "grad_norm": 0.13516271114349365, "learning_rate": 2.1109217977297823e-05, "loss": 0.013, "step": 71130 }, { "grad_norm": 0.36804717779159546, "learning_rate": 2.1095724477059077e-05, "loss": 0.0191, "step": 71140 }, { "grad_norm": 0.3235495686531067, "learning_rate": 2.1082234137748434e-05, "loss": 0.0231, "step": 71150 }, { "grad_norm": 0.1975611299276352, "learning_rate": 2.106874696084122e-05, "loss": 0.0215, "step": 71160 }, { "grad_norm": 0.4165235757827759, "learning_rate": 2.1055262947812315e-05, "loss": 0.0175, "step": 71170 }, { "grad_norm": 0.33525630831718445, "learning_rate": 2.104178210013636e-05, "loss": 0.0231, "step": 71180 }, { "grad_norm": 0.3320627808570862, "learning_rate": 2.1028304419287554e-05, "loss": 0.0255, "step": 71190 }, { "grad_norm": 0.33400481939315796, "learning_rate": 2.101482990673982e-05, "loss": 0.018, "step": 71200 }, { "grad_norm": 0.32565319538116455, "learning_rate": 2.1001358563966707e-05, "loss": 0.023, "step": 71210 }, { "grad_norm": 0.172315776348114, "learning_rate": 2.098789039244144e-05, "loss": 0.0132, "step": 71220 }, { "grad_norm": 0.24297702312469482, "learning_rate": 2.097442539363685e-05, "loss": 0.0207, "step": 71230 }, { "grad_norm": 0.31573423743247986, "learning_rate": 2.0960963569025448e-05, "loss": 0.018, "step": 71240 }, { "grad_norm": 0.1478198915719986, "learning_rate": 2.0947504920079435e-05, "loss": 0.0128, "step": 71250 }, { "grad_norm": 0.32989954948425293, "learning_rate": 2.093404944827059e-05, "loss": 0.017, "step": 71260 }, { "grad_norm": 0.21249625086784363, "learning_rate": 2.0920597155070394e-05, "loss": 0.0207, "step": 71270 }, { "grad_norm": 0.2751559615135193, "learning_rate": 2.090714804194997e-05, "loss": 0.023, "step": 71280 }, { "grad_norm": 0.3751404285430908, "learning_rate": 2.0893702110380114e-05, "loss": 0.0127, "step": 71290 }, { "grad_norm": 0.3449455499649048, "learning_rate": 2.0880259361831216e-05, "loss": 0.0182, "step": 71300 }, { "grad_norm": 0.33182892203330994, "learning_rate": 2.0866819797773397e-05, "loss": 0.0184, "step": 71310 }, { "grad_norm": 0.42795488238334656, "learning_rate": 2.085338341967634e-05, "loss": 0.0174, "step": 71320 }, { "grad_norm": 0.1698152869939804, "learning_rate": 2.083995022900946e-05, "loss": 0.0141, "step": 71330 }, { "grad_norm": 0.17874252796173096, "learning_rate": 2.0826520227241768e-05, "loss": 0.0163, "step": 71340 }, { "grad_norm": 0.5278342366218567, "learning_rate": 2.081309341584199e-05, "loss": 0.0244, "step": 71350 }, { "grad_norm": 0.281633585691452, "learning_rate": 2.0799669796278413e-05, "loss": 0.0174, "step": 71360 }, { "grad_norm": 0.23603715002536774, "learning_rate": 2.078624937001906e-05, "loss": 0.0211, "step": 71370 }, { "grad_norm": 0.3564828038215637, "learning_rate": 2.0772832138531544e-05, "loss": 0.0349, "step": 71380 }, { "grad_norm": 0.2877536416053772, "learning_rate": 2.0759418103283158e-05, "loss": 0.023, "step": 71390 }, { "grad_norm": 0.19755014777183533, "learning_rate": 2.0746007265740846e-05, "loss": 0.02, "step": 71400 }, { "grad_norm": 0.2501530647277832, "learning_rate": 2.0732599627371215e-05, "loss": 0.016, "step": 71410 }, { "grad_norm": 0.16416186094284058, "learning_rate": 2.0719195189640468e-05, "loss": 0.0177, "step": 71420 }, { "grad_norm": 0.12716950476169586, "learning_rate": 2.0705793954014528e-05, "loss": 0.0139, "step": 71430 }, { "grad_norm": 0.32130417227745056, "learning_rate": 2.06923959219589e-05, "loss": 0.0173, "step": 71440 }, { "grad_norm": 0.6261999011039734, "learning_rate": 2.0679001094938793e-05, "loss": 0.0144, "step": 71450 }, { "grad_norm": 0.19474340975284576, "learning_rate": 2.066560947441904e-05, "loss": 0.0165, "step": 71460 }, { "grad_norm": 0.2609191834926605, "learning_rate": 2.0652221061864152e-05, "loss": 0.0154, "step": 71470 }, { "grad_norm": 0.4736449122428894, "learning_rate": 2.0638835858738226e-05, "loss": 0.0163, "step": 71480 }, { "grad_norm": 0.27065497636795044, "learning_rate": 2.0625453866505068e-05, "loss": 0.0181, "step": 71490 }, { "grad_norm": 0.20451007783412933, "learning_rate": 2.0612075086628123e-05, "loss": 0.0226, "step": 71500 }, { "grad_norm": 0.21459707617759705, "learning_rate": 2.059869952057045e-05, "loss": 0.0167, "step": 71510 }, { "grad_norm": 0.3595590889453888, "learning_rate": 2.0585327169794793e-05, "loss": 0.0324, "step": 71520 }, { "grad_norm": 0.26808127760887146, "learning_rate": 2.0571958035763534e-05, "loss": 0.0143, "step": 71530 }, { "grad_norm": 0.2258939892053604, "learning_rate": 2.0558592119938715e-05, "loss": 0.017, "step": 71540 }, { "grad_norm": 0.2001200020313263, "learning_rate": 2.054522942378198e-05, "loss": 0.0175, "step": 71550 }, { "grad_norm": 0.2649812400341034, "learning_rate": 2.053186994875469e-05, "loss": 0.0176, "step": 71560 }, { "grad_norm": 0.3459516763687134, "learning_rate": 2.0518513696317775e-05, "loss": 0.0199, "step": 71570 }, { "grad_norm": 0.27163949608802795, "learning_rate": 2.050516066793188e-05, "loss": 0.024, "step": 71580 }, { "grad_norm": 0.2509835660457611, "learning_rate": 2.0491810865057276e-05, "loss": 0.0183, "step": 71590 }, { "grad_norm": 0.4534763693809509, "learning_rate": 2.0478464289153882e-05, "loss": 0.0191, "step": 71600 }, { "grad_norm": 0.3999273180961609, "learning_rate": 2.0465120941681232e-05, "loss": 0.0235, "step": 71610 }, { "grad_norm": 0.2961980104446411, "learning_rate": 2.0451780824098566e-05, "loss": 0.0188, "step": 71620 }, { "grad_norm": 0.30641821026802063, "learning_rate": 2.043844393786471e-05, "loss": 0.0174, "step": 71630 }, { "grad_norm": 0.348889023065567, "learning_rate": 2.0425110284438193e-05, "loss": 0.0186, "step": 71640 }, { "grad_norm": 0.2998751699924469, "learning_rate": 2.041177986527712e-05, "loss": 0.0185, "step": 71650 }, { "grad_norm": 0.30509769916534424, "learning_rate": 2.0398452681839347e-05, "loss": 0.0146, "step": 71660 }, { "grad_norm": 0.2910352051258087, "learning_rate": 2.0385128735582272e-05, "loss": 0.0355, "step": 71670 }, { "grad_norm": 0.20855873823165894, "learning_rate": 2.0371808027963003e-05, "loss": 0.0187, "step": 71680 }, { "grad_norm": 0.8417187929153442, "learning_rate": 2.035849056043825e-05, "loss": 0.0211, "step": 71690 }, { "grad_norm": 0.40054839849472046, "learning_rate": 2.03451763344644e-05, "loss": 0.0197, "step": 71700 }, { "grad_norm": 0.22629281878471375, "learning_rate": 2.0331865351497482e-05, "loss": 0.0162, "step": 71710 }, { "grad_norm": 0.4259405732154846, "learning_rate": 2.0318557612993185e-05, "loss": 0.0187, "step": 71720 }, { "grad_norm": 0.284697949886322, "learning_rate": 2.0305253120406782e-05, "loss": 0.0182, "step": 71730 }, { "grad_norm": 0.25003698468208313, "learning_rate": 2.029195187519325e-05, "loss": 0.0184, "step": 71740 }, { "grad_norm": 0.41603952646255493, "learning_rate": 2.0278653878807217e-05, "loss": 0.0225, "step": 71750 }, { "grad_norm": 0.26770472526550293, "learning_rate": 2.0265359132702892e-05, "loss": 0.0271, "step": 71760 }, { "grad_norm": 0.24293074011802673, "learning_rate": 2.0252067638334182e-05, "loss": 0.0194, "step": 71770 }, { "grad_norm": 0.18898072838783264, "learning_rate": 2.023877939715464e-05, "loss": 0.0193, "step": 71780 }, { "grad_norm": 0.312640905380249, "learning_rate": 2.0225494410617448e-05, "loss": 0.0165, "step": 71790 }, { "grad_norm": 0.20552420616149902, "learning_rate": 2.02122126801754e-05, "loss": 0.0158, "step": 71800 }, { "grad_norm": 0.19070352613925934, "learning_rate": 2.0198934207281018e-05, "loss": 0.0195, "step": 71810 }, { "grad_norm": 0.17484456300735474, "learning_rate": 2.0185658993386357e-05, "loss": 0.0179, "step": 71820 }, { "grad_norm": 0.30199623107910156, "learning_rate": 2.0172387039943226e-05, "loss": 0.017, "step": 71830 }, { "grad_norm": 0.22126777470111847, "learning_rate": 2.0159118348402968e-05, "loss": 0.0256, "step": 71840 }, { "grad_norm": 0.16812840104103088, "learning_rate": 2.0145852920216697e-05, "loss": 0.0165, "step": 71850 }, { "grad_norm": 0.33602121472358704, "learning_rate": 2.0132590756835045e-05, "loss": 0.0187, "step": 71860 }, { "grad_norm": 0.2290026992559433, "learning_rate": 2.0119331859708386e-05, "loss": 0.0238, "step": 71870 }, { "grad_norm": 0.27577388286590576, "learning_rate": 2.0106076230286654e-05, "loss": 0.0196, "step": 71880 }, { "grad_norm": 1.0922491550445557, "learning_rate": 2.0092823870019488e-05, "loss": 0.0183, "step": 71890 }, { "grad_norm": 0.41005784273147583, "learning_rate": 2.0079574780356115e-05, "loss": 0.0166, "step": 71900 }, { "grad_norm": 0.2689926326274872, "learning_rate": 2.0066328962745486e-05, "loss": 0.0203, "step": 71910 }, { "grad_norm": 0.522101104259491, "learning_rate": 2.00530864186361e-05, "loss": 0.022, "step": 71920 }, { "grad_norm": 0.2023126482963562, "learning_rate": 2.003984714947618e-05, "loss": 0.0187, "step": 71930 }, { "grad_norm": 0.3024592995643616, "learning_rate": 2.0026611156713515e-05, "loss": 0.0202, "step": 71940 }, { "grad_norm": 0.3297637701034546, "learning_rate": 2.0013378441795584e-05, "loss": 0.014, "step": 71950 }, { "grad_norm": 0.19971974194049835, "learning_rate": 2.0000149006169506e-05, "loss": 0.0197, "step": 71960 }, { "grad_norm": 0.31815412640571594, "learning_rate": 1.9986922851282048e-05, "loss": 0.0214, "step": 71970 }, { "grad_norm": 0.33080071210861206, "learning_rate": 1.997369997857956e-05, "loss": 0.0193, "step": 71980 }, { "grad_norm": 0.1909387707710266, "learning_rate": 1.9960480389508103e-05, "loss": 0.0137, "step": 71990 }, { "grad_norm": 0.24053482711315155, "learning_rate": 1.9947264085513363e-05, "loss": 0.0178, "step": 72000 }, { "grad_norm": 0.31459274888038635, "learning_rate": 1.993405106804062e-05, "loss": 0.0163, "step": 72010 }, { "grad_norm": 0.41712862253189087, "learning_rate": 1.9920841338534867e-05, "loss": 0.0147, "step": 72020 }, { "grad_norm": 0.21633203327655792, "learning_rate": 1.9907634898440646e-05, "loss": 0.0152, "step": 72030 }, { "grad_norm": 0.27008479833602905, "learning_rate": 1.989443174920227e-05, "loss": 0.0136, "step": 72040 }, { "grad_norm": 0.241508349776268, "learning_rate": 1.988123189226355e-05, "loss": 0.027, "step": 72050 }, { "grad_norm": 0.25738397240638733, "learning_rate": 1.986803532906805e-05, "loss": 0.0245, "step": 72060 }, { "grad_norm": 0.28208521008491516, "learning_rate": 1.985484206105889e-05, "loss": 0.0166, "step": 72070 }, { "grad_norm": 0.2560287117958069, "learning_rate": 1.9841652089678887e-05, "loss": 0.0134, "step": 72080 }, { "grad_norm": 0.3537238538265228, "learning_rate": 1.9828465416370435e-05, "loss": 0.0252, "step": 72090 }, { "grad_norm": 0.3327716588973999, "learning_rate": 1.9815282042575685e-05, "loss": 0.0193, "step": 72100 }, { "grad_norm": 0.2127365916967392, "learning_rate": 1.9802101969736285e-05, "loss": 0.0179, "step": 72110 }, { "grad_norm": 0.30253419280052185, "learning_rate": 1.9788925199293633e-05, "loss": 0.0214, "step": 72120 }, { "grad_norm": 0.26643362641334534, "learning_rate": 1.977575173268868e-05, "loss": 0.0163, "step": 72130 }, { "grad_norm": 0.23057684302330017, "learning_rate": 1.9762581571362094e-05, "loss": 0.0178, "step": 72140 }, { "grad_norm": 0.6242797374725342, "learning_rate": 1.9749414716754112e-05, "loss": 0.0235, "step": 72150 }, { "grad_norm": 0.2956404983997345, "learning_rate": 1.973625117030465e-05, "loss": 0.0234, "step": 72160 }, { "grad_norm": 0.18466275930404663, "learning_rate": 1.972309093345326e-05, "loss": 0.0204, "step": 72170 }, { "grad_norm": 0.22692620754241943, "learning_rate": 1.970993400763915e-05, "loss": 0.022, "step": 72180 }, { "grad_norm": 0.21660465002059937, "learning_rate": 1.969678039430109e-05, "loss": 0.0247, "step": 72190 }, { "grad_norm": 0.3329309821128845, "learning_rate": 1.9683630094877564e-05, "loss": 0.0164, "step": 72200 }, { "grad_norm": 0.29400908946990967, "learning_rate": 1.9670483110806696e-05, "loss": 0.0177, "step": 72210 }, { "grad_norm": 0.18019497394561768, "learning_rate": 1.9657339443526172e-05, "loss": 0.0178, "step": 72220 }, { "grad_norm": 4.4068603515625, "learning_rate": 1.964419909447339e-05, "loss": 0.0173, "step": 72230 }, { "grad_norm": 0.23003414273262024, "learning_rate": 1.963106206508536e-05, "loss": 0.013, "step": 72240 }, { "grad_norm": 0.21896333992481232, "learning_rate": 1.9617928356798738e-05, "loss": 0.0199, "step": 72250 }, { "grad_norm": 0.27135220170021057, "learning_rate": 1.960479797104977e-05, "loss": 0.0182, "step": 72260 }, { "grad_norm": 0.2093077450990677, "learning_rate": 1.9591670909274418e-05, "loss": 0.0153, "step": 72270 }, { "grad_norm": 0.30661115050315857, "learning_rate": 1.9578547172908184e-05, "loss": 0.0202, "step": 72280 }, { "grad_norm": 0.2788950502872467, "learning_rate": 1.9565426763386325e-05, "loss": 0.0138, "step": 72290 }, { "grad_norm": 0.24594765901565552, "learning_rate": 1.9552309682143622e-05, "loss": 0.0193, "step": 72300 }, { "grad_norm": 0.47644761204719543, "learning_rate": 1.9539195930614574e-05, "loss": 0.0183, "step": 72310 }, { "grad_norm": 0.16926082968711853, "learning_rate": 1.952608551023324e-05, "loss": 0.0194, "step": 72320 }, { "grad_norm": 0.22884266078472137, "learning_rate": 1.9512978422433397e-05, "loss": 0.0191, "step": 72330 }, { "grad_norm": 0.44386282563209534, "learning_rate": 1.9499874668648376e-05, "loss": 0.0148, "step": 72340 }, { "grad_norm": 0.3186909854412079, "learning_rate": 1.9486774250311213e-05, "loss": 0.0167, "step": 72350 }, { "grad_norm": 0.2932024896144867, "learning_rate": 1.9473677168854537e-05, "loss": 0.0189, "step": 72360 }, { "grad_norm": 0.8558080792427063, "learning_rate": 1.946058342571065e-05, "loss": 0.0232, "step": 72370 }, { "grad_norm": 0.2740727663040161, "learning_rate": 1.9447493022311424e-05, "loss": 0.0139, "step": 72380 }, { "grad_norm": 0.3235585391521454, "learning_rate": 1.9434405960088442e-05, "loss": 0.0255, "step": 72390 }, { "grad_norm": 0.316819965839386, "learning_rate": 1.9421322240472857e-05, "loss": 0.0157, "step": 72400 }, { "grad_norm": 0.3566712439060211, "learning_rate": 1.9408241864895494e-05, "loss": 0.0203, "step": 72410 }, { "grad_norm": 0.2640587389469147, "learning_rate": 1.939516483478681e-05, "loss": 0.0161, "step": 72420 }, { "grad_norm": 0.3225348889827728, "learning_rate": 1.9382091151576902e-05, "loss": 0.0195, "step": 72430 }, { "grad_norm": 0.15015077590942383, "learning_rate": 1.9369020816695454e-05, "loss": 0.0207, "step": 72440 }, { "grad_norm": 0.18030372262001038, "learning_rate": 1.935595383157184e-05, "loss": 0.0148, "step": 72450 }, { "grad_norm": 0.5442133545875549, "learning_rate": 1.9342890197635057e-05, "loss": 0.0132, "step": 72460 }, { "grad_norm": 0.41706252098083496, "learning_rate": 1.9329829916313685e-05, "loss": 0.0164, "step": 72470 }, { "grad_norm": 0.37583500146865845, "learning_rate": 1.9316772989036013e-05, "loss": 0.0215, "step": 72480 }, { "grad_norm": 0.22477826476097107, "learning_rate": 1.930371941722991e-05, "loss": 0.0139, "step": 72490 }, { "grad_norm": 0.2058088481426239, "learning_rate": 1.929066920232292e-05, "loss": 0.0145, "step": 72500 }, { "grad_norm": 0.3549291789531708, "learning_rate": 1.9277622345742152e-05, "loss": 0.0172, "step": 72510 }, { "grad_norm": 0.2690136730670929, "learning_rate": 1.926457884891444e-05, "loss": 0.0187, "step": 72520 }, { "grad_norm": 0.22422024607658386, "learning_rate": 1.9251538713266147e-05, "loss": 0.0167, "step": 72530 }, { "grad_norm": 0.31975406408309937, "learning_rate": 1.9238501940223358e-05, "loss": 0.0205, "step": 72540 }, { "grad_norm": 0.30743542313575745, "learning_rate": 1.922546853121175e-05, "loss": 0.0215, "step": 72550 }, { "grad_norm": 0.2713295519351959, "learning_rate": 1.9212438487656647e-05, "loss": 0.0286, "step": 72560 }, { "grad_norm": 0.27695557475090027, "learning_rate": 1.9199411810982965e-05, "loss": 0.0156, "step": 72570 }, { "grad_norm": 0.45284873247146606, "learning_rate": 1.918638850261532e-05, "loss": 0.0135, "step": 72580 }, { "grad_norm": 0.3329150974750519, "learning_rate": 1.9173368563977885e-05, "loss": 0.0154, "step": 72590 }, { "grad_norm": 0.16649457812309265, "learning_rate": 1.9160351996494513e-05, "loss": 0.0152, "step": 72600 }, { "grad_norm": 0.5366971492767334, "learning_rate": 1.9147338801588677e-05, "loss": 0.0212, "step": 72610 }, { "grad_norm": 0.28952282667160034, "learning_rate": 1.913432898068351e-05, "loss": 0.0209, "step": 72620 }, { "grad_norm": 0.31760674715042114, "learning_rate": 1.912132253520169e-05, "loss": 0.0187, "step": 72630 }, { "grad_norm": 0.29260560870170593, "learning_rate": 1.910831946656564e-05, "loss": 0.0206, "step": 72640 }, { "grad_norm": 0.26189711689949036, "learning_rate": 1.909531977619731e-05, "loss": 0.0166, "step": 72650 }, { "grad_norm": 0.47597628831863403, "learning_rate": 1.908232346551834e-05, "loss": 0.0126, "step": 72660 }, { "grad_norm": 0.3543505072593689, "learning_rate": 1.9069330535949993e-05, "loss": 0.0186, "step": 72670 }, { "grad_norm": 0.42153123021125793, "learning_rate": 1.9056340988913167e-05, "loss": 0.0143, "step": 72680 }, { "grad_norm": 0.21289044618606567, "learning_rate": 1.9043354825828357e-05, "loss": 0.0175, "step": 72690 }, { "grad_norm": 0.26731663942337036, "learning_rate": 1.9030372048115724e-05, "loss": 0.0173, "step": 72700 }, { "grad_norm": 0.35997527837753296, "learning_rate": 1.9017392657195056e-05, "loss": 0.0243, "step": 72710 }, { "grad_norm": 0.2701779305934906, "learning_rate": 1.9004416654485724e-05, "loss": 0.0183, "step": 72720 }, { "grad_norm": 0.31174036860466003, "learning_rate": 1.8991444041406785e-05, "loss": 0.0223, "step": 72730 }, { "grad_norm": 0.10885217785835266, "learning_rate": 1.8978474819376912e-05, "loss": 0.0198, "step": 72740 }, { "grad_norm": 0.6667620539665222, "learning_rate": 1.8965508989814406e-05, "loss": 0.0197, "step": 72750 }, { "grad_norm": 0.27953803539276123, "learning_rate": 1.8952546554137158e-05, "loss": 0.0182, "step": 72760 }, { "grad_norm": 0.3678509593009949, "learning_rate": 1.8939587513762757e-05, "loss": 0.0157, "step": 72770 }, { "grad_norm": 0.24646368622779846, "learning_rate": 1.892663187010835e-05, "loss": 0.0197, "step": 72780 }, { "grad_norm": 0.2874421775341034, "learning_rate": 1.891367962459077e-05, "loss": 0.0231, "step": 72790 }, { "grad_norm": 0.7496454119682312, "learning_rate": 1.890073077862644e-05, "loss": 0.0142, "step": 72800 }, { "grad_norm": 0.2680983245372772, "learning_rate": 1.888778533363146e-05, "loss": 0.0145, "step": 72810 }, { "grad_norm": 0.14714589715003967, "learning_rate": 1.8874843291021478e-05, "loss": 0.0157, "step": 72820 }, { "grad_norm": 0.3796190917491913, "learning_rate": 1.8861904652211853e-05, "loss": 0.0171, "step": 72830 }, { "grad_norm": 0.3870601952075958, "learning_rate": 1.8848969418617507e-05, "loss": 0.0154, "step": 72840 }, { "grad_norm": 0.21899084746837616, "learning_rate": 1.8836037591653044e-05, "loss": 0.0148, "step": 72850 }, { "grad_norm": 0.29415085911750793, "learning_rate": 1.882310917273262e-05, "loss": 0.0152, "step": 72860 }, { "grad_norm": 0.21458356082439423, "learning_rate": 1.881018416327014e-05, "loss": 0.0201, "step": 72870 }, { "grad_norm": 0.3039790093898773, "learning_rate": 1.8797262564679014e-05, "loss": 0.0201, "step": 72880 }, { "grad_norm": 0.1666596382856369, "learning_rate": 1.8784344378372355e-05, "loss": 0.0138, "step": 72890 }, { "grad_norm": 0.2057957649230957, "learning_rate": 1.8771429605762843e-05, "loss": 0.0178, "step": 72900 }, { "grad_norm": 0.29275572299957275, "learning_rate": 1.8758518248262835e-05, "loss": 0.0157, "step": 72910 }, { "grad_norm": 0.32736194133758545, "learning_rate": 1.8745610307284305e-05, "loss": 0.0268, "step": 72920 }, { "grad_norm": 0.32168909907341003, "learning_rate": 1.8732705784238862e-05, "loss": 0.0181, "step": 72930 }, { "grad_norm": 0.2148871272802353, "learning_rate": 1.8719804680537682e-05, "loss": 0.0175, "step": 72940 }, { "grad_norm": 0.24217037856578827, "learning_rate": 1.870690699759164e-05, "loss": 0.0173, "step": 72950 }, { "grad_norm": 0.2956241965293884, "learning_rate": 1.8694012736811213e-05, "loss": 0.0212, "step": 72960 }, { "grad_norm": 0.30494630336761475, "learning_rate": 1.8681121899606463e-05, "loss": 0.0176, "step": 72970 }, { "grad_norm": 0.18122495710849762, "learning_rate": 1.8668234487387144e-05, "loss": 0.0175, "step": 72980 }, { "grad_norm": 0.1983032524585724, "learning_rate": 1.8655350501562592e-05, "loss": 0.013, "step": 72990 }, { "grad_norm": 0.244428813457489, "learning_rate": 1.8642469943541796e-05, "loss": 0.0262, "step": 73000 }, { "grad_norm": 0.18468090891838074, "learning_rate": 1.8629592814733328e-05, "loss": 0.0168, "step": 73010 }, { "grad_norm": 0.17566204071044922, "learning_rate": 1.8616719116545438e-05, "loss": 0.0191, "step": 73020 }, { "grad_norm": 0.27020806074142456, "learning_rate": 1.8603848850385952e-05, "loss": 0.0122, "step": 73030 }, { "grad_norm": 0.11288546025753021, "learning_rate": 1.859098201766236e-05, "loss": 0.0156, "step": 73040 }, { "grad_norm": 0.23325158655643463, "learning_rate": 1.8578118619781726e-05, "loss": 0.0196, "step": 73050 }, { "grad_norm": 0.19759495556354523, "learning_rate": 1.856525865815083e-05, "loss": 0.017, "step": 73060 }, { "grad_norm": 0.19254980981349945, "learning_rate": 1.8552402134175966e-05, "loss": 0.0132, "step": 73070 }, { "grad_norm": 0.2515554428100586, "learning_rate": 1.8539549049263143e-05, "loss": 0.0178, "step": 73080 }, { "grad_norm": 0.4276328980922699, "learning_rate": 1.8526699404817917e-05, "loss": 0.0194, "step": 73090 }, { "grad_norm": 0.3042193055152893, "learning_rate": 1.8513853202245546e-05, "loss": 0.0235, "step": 73100 }, { "grad_norm": 0.25495296716690063, "learning_rate": 1.850101044295081e-05, "loss": 0.0152, "step": 73110 }, { "grad_norm": 0.31319236755371094, "learning_rate": 1.848817112833825e-05, "loss": 0.0171, "step": 73120 }, { "grad_norm": 0.45683571696281433, "learning_rate": 1.8475335259811905e-05, "loss": 0.0194, "step": 73130 }, { "grad_norm": 0.41713443398475647, "learning_rate": 1.8462502838775508e-05, "loss": 0.0276, "step": 73140 }, { "grad_norm": 0.2778606712818146, "learning_rate": 1.8449673866632378e-05, "loss": 0.0229, "step": 73150 }, { "grad_norm": 0.1498597264289856, "learning_rate": 1.8436848344785473e-05, "loss": 0.0228, "step": 73160 }, { "grad_norm": 0.3066600263118744, "learning_rate": 1.8424026274637378e-05, "loss": 0.0142, "step": 73170 }, { "grad_norm": 0.2860051095485687, "learning_rate": 1.8411207657590312e-05, "loss": 0.0138, "step": 73180 }, { "grad_norm": 0.27425920963287354, "learning_rate": 1.839839249504607e-05, "loss": 0.0197, "step": 73190 }, { "grad_norm": 0.24336260557174683, "learning_rate": 1.8385580788406108e-05, "loss": 0.0158, "step": 73200 }, { "grad_norm": 0.24843405187129974, "learning_rate": 1.837277253907152e-05, "loss": 0.02, "step": 73210 }, { "grad_norm": 0.2958471179008484, "learning_rate": 1.8359967748442957e-05, "loss": 0.0165, "step": 73220 }, { "grad_norm": 0.3140551745891571, "learning_rate": 1.8347166417920774e-05, "loss": 0.0118, "step": 73230 }, { "grad_norm": 0.22479601204395294, "learning_rate": 1.833436854890484e-05, "loss": 0.0187, "step": 73240 }, { "grad_norm": 0.2643274664878845, "learning_rate": 1.8321574142794796e-05, "loss": 0.0145, "step": 73250 }, { "grad_norm": 0.14949437975883484, "learning_rate": 1.8308783200989748e-05, "loss": 0.0146, "step": 73260 }, { "grad_norm": 0.2211288958787918, "learning_rate": 1.829599572488855e-05, "loss": 0.0154, "step": 73270 }, { "grad_norm": 0.3550272583961487, "learning_rate": 1.828321171588957e-05, "loss": 0.0127, "step": 73280 }, { "grad_norm": 0.20336656272411346, "learning_rate": 1.827043117539089e-05, "loss": 0.0191, "step": 73290 }, { "grad_norm": 0.2959282696247101, "learning_rate": 1.8257654104790116e-05, "loss": 0.0195, "step": 73300 }, { "grad_norm": 0.2603301703929901, "learning_rate": 1.82448805054846e-05, "loss": 0.0182, "step": 73310 }, { "grad_norm": 0.19734987616539001, "learning_rate": 1.823211037887119e-05, "loss": 0.018, "step": 73320 }, { "grad_norm": 0.3400539755821228, "learning_rate": 1.821934372634645e-05, "loss": 0.0186, "step": 73330 }, { "grad_norm": 0.14440792798995972, "learning_rate": 1.8206580549306472e-05, "loss": 0.0129, "step": 73340 }, { "grad_norm": 0.3780641555786133, "learning_rate": 1.819382084914707e-05, "loss": 0.0173, "step": 73350 }, { "grad_norm": 0.3280933201313019, "learning_rate": 1.8181064627263577e-05, "loss": 0.0147, "step": 73360 }, { "grad_norm": 0.40071073174476624, "learning_rate": 1.8168311885051022e-05, "loss": 0.0231, "step": 73370 }, { "grad_norm": 0.2643623948097229, "learning_rate": 1.815556262390402e-05, "loss": 0.0153, "step": 73380 }, { "grad_norm": 0.27541226148605347, "learning_rate": 1.8142816845216826e-05, "loss": 0.0163, "step": 73390 }, { "grad_norm": 0.1581440418958664, "learning_rate": 1.8130074550383276e-05, "loss": 0.014, "step": 73400 }, { "grad_norm": 0.2774275243282318, "learning_rate": 1.8117335740796848e-05, "loss": 0.0179, "step": 73410 }, { "grad_norm": 0.33323749899864197, "learning_rate": 1.810460041785067e-05, "loss": 0.0208, "step": 73420 }, { "grad_norm": 0.27202391624450684, "learning_rate": 1.809186858293742e-05, "loss": 0.028, "step": 73430 }, { "grad_norm": 0.38744059205055237, "learning_rate": 1.8079140237449448e-05, "loss": 0.0211, "step": 73440 }, { "grad_norm": 0.46364670991897583, "learning_rate": 1.8066415382778707e-05, "loss": 0.0264, "step": 73450 }, { "grad_norm": 0.1894238293170929, "learning_rate": 1.8053694020316786e-05, "loss": 0.0137, "step": 73460 }, { "grad_norm": 0.27042055130004883, "learning_rate": 1.8040976151454837e-05, "loss": 0.0166, "step": 73470 }, { "grad_norm": 0.24827556312084198, "learning_rate": 1.8028261777583704e-05, "loss": 0.0235, "step": 73480 }, { "grad_norm": 0.2890218496322632, "learning_rate": 1.8015550900093758e-05, "loss": 0.0191, "step": 73490 }, { "grad_norm": 0.2681557238101959, "learning_rate": 1.800284352037512e-05, "loss": 0.017, "step": 73500 }, { "grad_norm": 0.21829932928085327, "learning_rate": 1.799013963981738e-05, "loss": 0.0157, "step": 73510 }, { "grad_norm": 0.3257463574409485, "learning_rate": 1.797743925980987e-05, "loss": 0.0218, "step": 73520 }, { "grad_norm": 0.22317901253700256, "learning_rate": 1.7964742381741435e-05, "loss": 0.0163, "step": 73530 }, { "grad_norm": 0.31996315717697144, "learning_rate": 1.7952049007000627e-05, "loss": 0.0182, "step": 73540 }, { "grad_norm": 0.3019838333129883, "learning_rate": 1.7939359136975535e-05, "loss": 0.0184, "step": 73550 }, { "grad_norm": 0.18476268649101257, "learning_rate": 1.7926672773053932e-05, "loss": 0.0153, "step": 73560 }, { "grad_norm": 0.3129042088985443, "learning_rate": 1.791398991662317e-05, "loss": 0.0246, "step": 73570 }, { "grad_norm": 0.17857897281646729, "learning_rate": 1.7901310569070244e-05, "loss": 0.0202, "step": 73580 }, { "grad_norm": 0.3122181296348572, "learning_rate": 1.7888634731781724e-05, "loss": 0.0257, "step": 73590 }, { "grad_norm": 0.23094697296619415, "learning_rate": 1.7875962406143844e-05, "loss": 0.0151, "step": 73600 }, { "grad_norm": 0.5782226920127869, "learning_rate": 1.7863293593542403e-05, "loss": 0.0321, "step": 73610 }, { "grad_norm": 0.3142014741897583, "learning_rate": 1.7850628295362853e-05, "loss": 0.016, "step": 73620 }, { "grad_norm": 0.27662041783332825, "learning_rate": 1.7837966512990263e-05, "loss": 0.0183, "step": 73630 }, { "grad_norm": 0.29064881801605225, "learning_rate": 1.782530824780931e-05, "loss": 0.0183, "step": 73640 }, { "grad_norm": 0.23741570115089417, "learning_rate": 1.7812653501204263e-05, "loss": 0.0217, "step": 73650 }, { "grad_norm": 0.2307901233434677, "learning_rate": 1.780000227455903e-05, "loss": 0.024, "step": 73660 }, { "grad_norm": 0.19319559633731842, "learning_rate": 1.778735456925716e-05, "loss": 0.0204, "step": 73670 }, { "grad_norm": 0.2443532645702362, "learning_rate": 1.7774710386681737e-05, "loss": 0.0133, "step": 73680 }, { "grad_norm": 0.45485028624534607, "learning_rate": 1.7762069728215545e-05, "loss": 0.0208, "step": 73690 }, { "grad_norm": 0.4577353596687317, "learning_rate": 1.7749432595240932e-05, "loss": 0.0172, "step": 73700 }, { "grad_norm": 0.2931225299835205, "learning_rate": 1.773679898913991e-05, "loss": 0.0224, "step": 73710 }, { "grad_norm": 0.24762003123760223, "learning_rate": 1.7724168911294015e-05, "loss": 0.0187, "step": 73720 }, { "grad_norm": 0.3241361379623413, "learning_rate": 1.7711542363084504e-05, "loss": 0.0156, "step": 73730 }, { "grad_norm": 0.22700472176074982, "learning_rate": 1.7698919345892158e-05, "loss": 0.014, "step": 73740 }, { "grad_norm": 0.3734742999076843, "learning_rate": 1.768629986109743e-05, "loss": 0.0139, "step": 73750 }, { "grad_norm": 0.5161715745925903, "learning_rate": 1.7673683910080357e-05, "loss": 0.0222, "step": 73760 }, { "grad_norm": 0.2040330171585083, "learning_rate": 1.7661071494220638e-05, "loss": 0.016, "step": 73770 }, { "grad_norm": 0.22447238862514496, "learning_rate": 1.7648462614897498e-05, "loss": 0.0163, "step": 73780 }, { "grad_norm": 0.24638159573078156, "learning_rate": 1.7635857273489863e-05, "loss": 0.0184, "step": 73790 }, { "grad_norm": 0.26623067259788513, "learning_rate": 1.7623255471376198e-05, "loss": 0.0173, "step": 73800 }, { "grad_norm": 0.20848043262958527, "learning_rate": 1.7610657209934646e-05, "loss": 0.0263, "step": 73810 }, { "grad_norm": 0.1993584781885147, "learning_rate": 1.7598062490542915e-05, "loss": 0.0166, "step": 73820 }, { "grad_norm": 1.0824154615402222, "learning_rate": 1.7585471314578377e-05, "loss": 0.0211, "step": 73830 }, { "grad_norm": 0.3433140814304352, "learning_rate": 1.7572883683417945e-05, "loss": 0.0185, "step": 73840 }, { "grad_norm": 0.3270759880542755, "learning_rate": 1.7560299598438212e-05, "loss": 0.0141, "step": 73850 }, { "grad_norm": 0.4510467052459717, "learning_rate": 1.7547719061015334e-05, "loss": 0.0183, "step": 73860 }, { "grad_norm": 0.28146758675575256, "learning_rate": 1.75351420725251e-05, "loss": 0.0168, "step": 73870 }, { "grad_norm": 0.7733357548713684, "learning_rate": 1.7522568634342923e-05, "loss": 0.018, "step": 73880 }, { "grad_norm": 0.3238314688205719, "learning_rate": 1.7509998747843825e-05, "loss": 0.0138, "step": 73890 }, { "grad_norm": 0.31314945220947266, "learning_rate": 1.7497432414402403e-05, "loss": 0.0179, "step": 73900 }, { "grad_norm": 0.5879201889038086, "learning_rate": 1.7484869635392903e-05, "loss": 0.0133, "step": 73910 }, { "grad_norm": 0.30693575739860535, "learning_rate": 1.747231041218919e-05, "loss": 0.0157, "step": 73920 }, { "grad_norm": 0.2570881247520447, "learning_rate": 1.7459754746164692e-05, "loss": 0.0142, "step": 73930 }, { "grad_norm": 0.20992445945739746, "learning_rate": 1.744720263869249e-05, "loss": 0.0232, "step": 73940 }, { "grad_norm": 0.5654670596122742, "learning_rate": 1.7434654091145263e-05, "loss": 0.02, "step": 73950 }, { "grad_norm": 0.31667348742485046, "learning_rate": 1.742210910489532e-05, "loss": 0.0174, "step": 73960 }, { "grad_norm": 0.15257538855075836, "learning_rate": 1.7409567681314532e-05, "loss": 0.0136, "step": 73970 }, { "grad_norm": 2.2086703777313232, "learning_rate": 1.7397029821774434e-05, "loss": 0.021, "step": 73980 }, { "grad_norm": 0.22868695855140686, "learning_rate": 1.7384495527646126e-05, "loss": 0.0153, "step": 73990 }, { "grad_norm": 0.21010565757751465, "learning_rate": 1.737196480030035e-05, "loss": 0.0176, "step": 74000 }, { "grad_norm": 0.40102317929267883, "learning_rate": 1.7359437641107447e-05, "loss": 0.0204, "step": 74010 }, { "grad_norm": 0.2806403636932373, "learning_rate": 1.7346914051437387e-05, "loss": 0.0199, "step": 74020 }, { "grad_norm": 0.32363420724868774, "learning_rate": 1.733439403265969e-05, "loss": 0.015, "step": 74030 }, { "grad_norm": 0.26073136925697327, "learning_rate": 1.732187758614357e-05, "loss": 0.0254, "step": 74040 }, { "grad_norm": 0.6206356883049011, "learning_rate": 1.730936471325777e-05, "loss": 0.0165, "step": 74050 }, { "grad_norm": 0.20276100933551788, "learning_rate": 1.7296855415370718e-05, "loss": 0.0207, "step": 74060 }, { "grad_norm": 0.38112929463386536, "learning_rate": 1.728434969385035e-05, "loss": 0.0151, "step": 74070 }, { "grad_norm": 0.1323668509721756, "learning_rate": 1.7271847550064346e-05, "loss": 0.0138, "step": 74080 }, { "grad_norm": 0.2678748369216919, "learning_rate": 1.725934898537987e-05, "loss": 0.0195, "step": 74090 }, { "grad_norm": 0.49298274517059326, "learning_rate": 1.7246854001163783e-05, "loss": 0.0174, "step": 74100 }, { "grad_norm": 0.30413925647735596, "learning_rate": 1.7234362598782476e-05, "loss": 0.016, "step": 74110 }, { "grad_norm": 0.30378320813179016, "learning_rate": 1.722187477960202e-05, "loss": 0.0215, "step": 74120 }, { "grad_norm": 0.1876029223203659, "learning_rate": 1.7209390544988057e-05, "loss": 0.0217, "step": 74130 }, { "grad_norm": 0.20012465119361877, "learning_rate": 1.7196909896305856e-05, "loss": 0.0257, "step": 74140 }, { "grad_norm": 0.25971919298171997, "learning_rate": 1.718443283492026e-05, "loss": 0.0193, "step": 74150 }, { "grad_norm": 0.3896145522594452, "learning_rate": 1.7171959362195746e-05, "loss": 0.0243, "step": 74160 }, { "grad_norm": 0.3545776903629303, "learning_rate": 1.715948947949642e-05, "loss": 0.0191, "step": 74170 }, { "grad_norm": 0.663962185382843, "learning_rate": 1.714702318818593e-05, "loss": 0.0211, "step": 74180 }, { "grad_norm": 0.29323986172676086, "learning_rate": 1.7134560489627592e-05, "loss": 0.0164, "step": 74190 }, { "grad_norm": 0.7218025326728821, "learning_rate": 1.712210138518431e-05, "loss": 0.0206, "step": 74200 }, { "grad_norm": 0.2759018838405609, "learning_rate": 1.71096458762186e-05, "loss": 0.0232, "step": 74210 }, { "grad_norm": 0.22188349068164825, "learning_rate": 1.7097193964092552e-05, "loss": 0.0146, "step": 74220 }, { "grad_norm": 0.33464205265045166, "learning_rate": 1.708474565016792e-05, "loss": 0.0172, "step": 74230 }, { "grad_norm": 0.26617231965065, "learning_rate": 1.7072300935806e-05, "loss": 0.0179, "step": 74240 }, { "grad_norm": 0.2711409330368042, "learning_rate": 1.705985982236776e-05, "loss": 0.0149, "step": 74250 }, { "grad_norm": 0.18895888328552246, "learning_rate": 1.7047422311213697e-05, "loss": 0.0164, "step": 74260 }, { "grad_norm": 0.21641133725643158, "learning_rate": 1.703498840370401e-05, "loss": 0.0195, "step": 74270 }, { "grad_norm": 0.35870689153671265, "learning_rate": 1.702255810119842e-05, "loss": 0.0162, "step": 74280 }, { "grad_norm": 0.2162015736103058, "learning_rate": 1.7010131405056305e-05, "loss": 0.0136, "step": 74290 }, { "grad_norm": 0.3394940197467804, "learning_rate": 1.6997708316636613e-05, "loss": 0.0167, "step": 74300 }, { "grad_norm": 0.2641557455062866, "learning_rate": 1.6985288837297932e-05, "loss": 0.0185, "step": 74310 }, { "grad_norm": 0.7076291441917419, "learning_rate": 1.6972872968398394e-05, "loss": 0.0168, "step": 74320 }, { "grad_norm": 0.2250082641839981, "learning_rate": 1.6960460711295845e-05, "loss": 0.0178, "step": 74330 }, { "grad_norm": 0.41829410195350647, "learning_rate": 1.6948052067347623e-05, "loss": 0.016, "step": 74340 }, { "grad_norm": 0.3907599449157715, "learning_rate": 1.693564703791075e-05, "loss": 0.0237, "step": 74350 }, { "grad_norm": 0.38459789752960205, "learning_rate": 1.6923245624341793e-05, "loss": 0.0162, "step": 74360 }, { "grad_norm": 0.23291391134262085, "learning_rate": 1.6910847827996962e-05, "loss": 0.0207, "step": 74370 }, { "grad_norm": 0.3232814371585846, "learning_rate": 1.689845365023206e-05, "loss": 0.0175, "step": 74380 }, { "grad_norm": 0.2759295403957367, "learning_rate": 1.6886063092402505e-05, "loss": 0.018, "step": 74390 }, { "grad_norm": 0.35852402448654175, "learning_rate": 1.6873676155863317e-05, "loss": 0.0216, "step": 74400 }, { "grad_norm": 0.32348185777664185, "learning_rate": 1.6861292841969085e-05, "loss": 0.0202, "step": 74410 }, { "grad_norm": 0.32085442543029785, "learning_rate": 1.684891315207406e-05, "loss": 0.0196, "step": 74420 }, { "grad_norm": 0.262622207403183, "learning_rate": 1.6836537087532034e-05, "loss": 0.015, "step": 74430 }, { "grad_norm": 0.17357724905014038, "learning_rate": 1.6824164649696465e-05, "loss": 0.0199, "step": 74440 }, { "grad_norm": 0.2546178996562958, "learning_rate": 1.681179583992034e-05, "loss": 0.014, "step": 74450 }, { "grad_norm": 0.2541526257991791, "learning_rate": 1.679943065955636e-05, "loss": 0.0166, "step": 74460 }, { "grad_norm": 0.35924842953681946, "learning_rate": 1.678706910995671e-05, "loss": 0.019, "step": 74470 }, { "grad_norm": 0.17313404381275177, "learning_rate": 1.6774711192473262e-05, "loss": 0.0203, "step": 74480 }, { "grad_norm": 0.2883133590221405, "learning_rate": 1.676235690845743e-05, "loss": 0.0158, "step": 74490 }, { "grad_norm": 0.43654823303222656, "learning_rate": 1.6750006259260285e-05, "loss": 0.0247, "step": 74500 }, { "grad_norm": 0.332170695066452, "learning_rate": 1.673765924623244e-05, "loss": 0.0198, "step": 74510 }, { "grad_norm": 0.3477267026901245, "learning_rate": 1.67253158707242e-05, "loss": 0.0165, "step": 74520 }, { "grad_norm": 0.27930131554603577, "learning_rate": 1.671297613408537e-05, "loss": 0.0138, "step": 74530 }, { "grad_norm": 0.18732354044914246, "learning_rate": 1.6700640037665444e-05, "loss": 0.0128, "step": 74540 }, { "grad_norm": 0.24673324823379517, "learning_rate": 1.668830758281344e-05, "loss": 0.0167, "step": 74550 }, { "grad_norm": 0.6554827690124512, "learning_rate": 1.6675978770878052e-05, "loss": 0.0188, "step": 74560 }, { "grad_norm": 0.21366815268993378, "learning_rate": 1.6663653603207506e-05, "loss": 0.0194, "step": 74570 }, { "grad_norm": 0.27742424607276917, "learning_rate": 1.6651332081149685e-05, "loss": 0.0167, "step": 74580 }, { "grad_norm": 0.7618899941444397, "learning_rate": 1.6639014206052046e-05, "loss": 0.021, "step": 74590 }, { "grad_norm": 0.1957099735736847, "learning_rate": 1.6626699979261657e-05, "loss": 0.0186, "step": 74600 }, { "grad_norm": 0.25194403529167175, "learning_rate": 1.6614389402125197e-05, "loss": 0.0153, "step": 74610 }, { "grad_norm": 0.3757357597351074, "learning_rate": 1.66020824759889e-05, "loss": 0.0176, "step": 74620 }, { "grad_norm": 0.2980576455593109, "learning_rate": 1.6589779202198675e-05, "loss": 0.0176, "step": 74630 }, { "grad_norm": 0.25792059302330017, "learning_rate": 1.6577479582099924e-05, "loss": 0.0206, "step": 74640 }, { "grad_norm": 0.4131096601486206, "learning_rate": 1.6565183617037788e-05, "loss": 0.0259, "step": 74650 }, { "grad_norm": 0.3862440884113312, "learning_rate": 1.6552891308356894e-05, "loss": 0.0239, "step": 74660 }, { "grad_norm": 0.7835953235626221, "learning_rate": 1.6540602657401526e-05, "loss": 0.0272, "step": 74670 }, { "grad_norm": 0.33093318343162537, "learning_rate": 1.6528317665515537e-05, "loss": 0.0284, "step": 74680 }, { "grad_norm": 0.17223039269447327, "learning_rate": 1.6516036334042416e-05, "loss": 0.0194, "step": 74690 }, { "grad_norm": 0.18159395456314087, "learning_rate": 1.650375866432519e-05, "loss": 0.0183, "step": 74700 }, { "grad_norm": 0.22891320288181305, "learning_rate": 1.649148465770659e-05, "loss": 0.0142, "step": 74710 }, { "grad_norm": 0.29868370294570923, "learning_rate": 1.647921431552883e-05, "loss": 0.0132, "step": 74720 }, { "grad_norm": 0.2987573742866516, "learning_rate": 1.646694763913381e-05, "loss": 0.0218, "step": 74730 }, { "grad_norm": 0.24513797461986542, "learning_rate": 1.6454684629862966e-05, "loss": 0.0147, "step": 74740 }, { "grad_norm": 0.3005158603191376, "learning_rate": 1.6442425289057388e-05, "loss": 0.0214, "step": 74750 }, { "grad_norm": 0.22305190563201904, "learning_rate": 1.6430169618057718e-05, "loss": 0.0148, "step": 74760 }, { "grad_norm": 0.2765381336212158, "learning_rate": 1.641791761820422e-05, "loss": 0.0169, "step": 74770 }, { "grad_norm": 0.1886412352323532, "learning_rate": 1.640566929083676e-05, "loss": 0.0172, "step": 74780 }, { "grad_norm": 0.20230072736740112, "learning_rate": 1.6393424637294823e-05, "loss": 0.0263, "step": 74790 }, { "grad_norm": 0.4958970546722412, "learning_rate": 1.638118365891742e-05, "loss": 0.0188, "step": 74800 }, { "grad_norm": 0.20448042452335358, "learning_rate": 1.636894635704323e-05, "loss": 0.0178, "step": 74810 }, { "grad_norm": 0.22384074330329895, "learning_rate": 1.6356712733010517e-05, "loss": 0.0156, "step": 74820 }, { "grad_norm": 0.21027806401252747, "learning_rate": 1.6344482788157096e-05, "loss": 0.019, "step": 74830 }, { "grad_norm": 0.31526243686676025, "learning_rate": 1.6332256523820444e-05, "loss": 0.0143, "step": 74840 }, { "grad_norm": 0.32587265968322754, "learning_rate": 1.6320033941337597e-05, "loss": 0.0178, "step": 74850 }, { "grad_norm": 0.2943936288356781, "learning_rate": 1.6307815042045216e-05, "loss": 0.0158, "step": 74860 }, { "grad_norm": 0.3294069170951843, "learning_rate": 1.6295599827279513e-05, "loss": 0.0182, "step": 74870 }, { "grad_norm": 0.3731003701686859, "learning_rate": 1.6283388298376356e-05, "loss": 0.0196, "step": 74880 }, { "grad_norm": 0.21997326612472534, "learning_rate": 1.6271180456671125e-05, "loss": 0.0136, "step": 74890 }, { "grad_norm": 0.15198589861392975, "learning_rate": 1.625897630349893e-05, "loss": 0.0175, "step": 74900 }, { "grad_norm": 0.3732962906360626, "learning_rate": 1.6246775840194335e-05, "loss": 0.0213, "step": 74910 }, { "grad_norm": 0.3048992455005646, "learning_rate": 1.6234579068091604e-05, "loss": 0.0176, "step": 74920 }, { "grad_norm": 0.4182574152946472, "learning_rate": 1.6222385988524524e-05, "loss": 0.0187, "step": 74930 }, { "grad_norm": 0.2805488109588623, "learning_rate": 1.6210196602826544e-05, "loss": 0.0209, "step": 74940 }, { "grad_norm": 0.20416037738323212, "learning_rate": 1.6198010912330646e-05, "loss": 0.0222, "step": 74950 }, { "grad_norm": 1.195189356803894, "learning_rate": 1.6185828918369456e-05, "loss": 0.0222, "step": 74960 }, { "grad_norm": 0.4324369430541992, "learning_rate": 1.617365062227517e-05, "loss": 0.0168, "step": 74970 }, { "grad_norm": 0.30752983689308167, "learning_rate": 1.616147602537962e-05, "loss": 0.017, "step": 74980 }, { "grad_norm": 0.19574929773807526, "learning_rate": 1.614930512901415e-05, "loss": 0.0173, "step": 74990 }, { "grad_norm": 0.4387262463569641, "learning_rate": 1.6137137934509804e-05, "loss": 0.0159, "step": 75000 }, { "grad_norm": 0.3845077157020569, "learning_rate": 1.6124974443197123e-05, "loss": 0.0172, "step": 75010 }, { "grad_norm": 0.13996782898902893, "learning_rate": 1.6112814656406306e-05, "loss": 0.016, "step": 75020 }, { "grad_norm": 0.11839041113853455, "learning_rate": 1.6100658575467125e-05, "loss": 0.013, "step": 75030 }, { "grad_norm": 0.2287592887878418, "learning_rate": 1.6088506201708974e-05, "loss": 0.0172, "step": 75040 }, { "grad_norm": 0.40383821725845337, "learning_rate": 1.6076357536460785e-05, "loss": 0.0174, "step": 75050 }, { "grad_norm": 0.25139060616493225, "learning_rate": 1.606421258105113e-05, "loss": 0.0152, "step": 75060 }, { "grad_norm": 0.31377360224723816, "learning_rate": 1.6052071336808183e-05, "loss": 0.0245, "step": 75070 }, { "grad_norm": 0.30852288007736206, "learning_rate": 1.6039933805059658e-05, "loss": 0.0223, "step": 75080 }, { "grad_norm": 0.2240176647901535, "learning_rate": 1.6027799987132915e-05, "loss": 0.0134, "step": 75090 }, { "grad_norm": 0.2882957458496094, "learning_rate": 1.6015669884354888e-05, "loss": 0.0225, "step": 75100 }, { "grad_norm": 0.3578874468803406, "learning_rate": 1.6003543498052125e-05, "loss": 0.0145, "step": 75110 }, { "grad_norm": 0.32388046383857727, "learning_rate": 1.5991420829550715e-05, "loss": 0.0205, "step": 75120 }, { "grad_norm": 0.3971042037010193, "learning_rate": 1.5979301880176405e-05, "loss": 0.0192, "step": 75130 }, { "grad_norm": 0.19839614629745483, "learning_rate": 1.5967186651254485e-05, "loss": 0.0158, "step": 75140 }, { "grad_norm": 0.1847434788942337, "learning_rate": 1.5955075144109853e-05, "loss": 0.0154, "step": 75150 }, { "grad_norm": 0.22736696898937225, "learning_rate": 1.594296736006703e-05, "loss": 0.022, "step": 75160 }, { "grad_norm": 0.36148717999458313, "learning_rate": 1.59308633004501e-05, "loss": 0.0249, "step": 75170 }, { "grad_norm": 0.21786998212337494, "learning_rate": 1.5918762966582724e-05, "loss": 0.0147, "step": 75180 }, { "grad_norm": 0.14610262215137482, "learning_rate": 1.5906666359788203e-05, "loss": 0.0146, "step": 75190 }, { "grad_norm": 0.21996968984603882, "learning_rate": 1.589457348138938e-05, "loss": 0.0145, "step": 75200 }, { "grad_norm": 0.4338802099227905, "learning_rate": 1.588248433270872e-05, "loss": 0.0225, "step": 75210 }, { "grad_norm": 0.3847919702529907, "learning_rate": 1.587039891506828e-05, "loss": 0.0162, "step": 75220 }, { "grad_norm": 0.3257681131362915, "learning_rate": 1.5858317229789715e-05, "loss": 0.0203, "step": 75230 }, { "grad_norm": 0.23320439457893372, "learning_rate": 1.584623927819423e-05, "loss": 0.0133, "step": 75240 }, { "grad_norm": 0.2898510694503784, "learning_rate": 1.583416506160269e-05, "loss": 0.0179, "step": 75250 }, { "grad_norm": 0.2230493426322937, "learning_rate": 1.5822094581335472e-05, "loss": 0.0192, "step": 75260 }, { "grad_norm": 0.202604740858078, "learning_rate": 1.5810027838712604e-05, "loss": 0.0201, "step": 75270 }, { "grad_norm": 0.622276782989502, "learning_rate": 1.5797964835053687e-05, "loss": 0.0136, "step": 75280 }, { "grad_norm": 0.3130583167076111, "learning_rate": 1.5785905571677933e-05, "loss": 0.0188, "step": 75290 }, { "grad_norm": 0.43801775574684143, "learning_rate": 1.5773850049904086e-05, "loss": 0.0182, "step": 75300 }, { "grad_norm": 0.24264433979988098, "learning_rate": 1.5761798271050543e-05, "loss": 0.0165, "step": 75310 }, { "grad_norm": 0.4583282172679901, "learning_rate": 1.5749750236435277e-05, "loss": 0.0221, "step": 75320 }, { "grad_norm": 0.29468104243278503, "learning_rate": 1.5737705947375815e-05, "loss": 0.0166, "step": 75330 }, { "grad_norm": 0.26856616139411926, "learning_rate": 1.5725665405189322e-05, "loss": 0.0225, "step": 75340 }, { "grad_norm": 0.24685654044151306, "learning_rate": 1.571362861119253e-05, "loss": 0.0181, "step": 75350 }, { "grad_norm": 0.2234760820865631, "learning_rate": 1.570159556670177e-05, "loss": 0.0162, "step": 75360 }, { "grad_norm": 0.15455564856529236, "learning_rate": 1.5689566273032946e-05, "loss": 0.013, "step": 75370 }, { "grad_norm": 0.3021993935108185, "learning_rate": 1.5677540731501583e-05, "loss": 0.0165, "step": 75380 }, { "grad_norm": 0.5108233690261841, "learning_rate": 1.5665518943422746e-05, "loss": 0.0215, "step": 75390 }, { "grad_norm": 0.27890563011169434, "learning_rate": 1.5653500910111134e-05, "loss": 0.0157, "step": 75400 }, { "grad_norm": 0.4747425615787506, "learning_rate": 1.5641486632881023e-05, "loss": 0.0153, "step": 75410 }, { "grad_norm": 0.3523936867713928, "learning_rate": 1.56294761130463e-05, "loss": 0.0162, "step": 75420 }, { "grad_norm": 0.21621276438236237, "learning_rate": 1.5617469351920373e-05, "loss": 0.013, "step": 75430 }, { "grad_norm": 0.24544553458690643, "learning_rate": 1.5605466350816315e-05, "loss": 0.0198, "step": 75440 }, { "grad_norm": 0.33931201696395874, "learning_rate": 1.559346711104674e-05, "loss": 0.0173, "step": 75450 }, { "grad_norm": 0.13855187594890594, "learning_rate": 1.5581471633923883e-05, "loss": 0.0191, "step": 75460 }, { "grad_norm": 0.3281133770942688, "learning_rate": 1.5569479920759505e-05, "loss": 0.0162, "step": 75470 }, { "grad_norm": 0.23995372653007507, "learning_rate": 1.5557491972865072e-05, "loss": 0.0157, "step": 75480 }, { "grad_norm": 0.22307230532169342, "learning_rate": 1.554550779155152e-05, "loss": 0.0127, "step": 75490 }, { "grad_norm": 0.18411697447299957, "learning_rate": 1.5533527378129453e-05, "loss": 0.0137, "step": 75500 }, { "grad_norm": 0.27177804708480835, "learning_rate": 1.552155073390899e-05, "loss": 0.0184, "step": 75510 }, { "grad_norm": 0.3052545189857483, "learning_rate": 1.5509577860199903e-05, "loss": 0.0213, "step": 75520 }, { "grad_norm": 0.21834200620651245, "learning_rate": 1.5497608758311525e-05, "loss": 0.0138, "step": 75530 }, { "grad_norm": 0.2436669021844864, "learning_rate": 1.5485643429552797e-05, "loss": 0.0181, "step": 75540 }, { "grad_norm": 0.20939640700817108, "learning_rate": 1.5473681875232197e-05, "loss": 0.0162, "step": 75550 }, { "grad_norm": 0.29491138458251953, "learning_rate": 1.5461724096657837e-05, "loss": 0.0127, "step": 75560 }, { "grad_norm": 0.36212238669395447, "learning_rate": 1.5449770095137416e-05, "loss": 0.0137, "step": 75570 }, { "grad_norm": 0.2183985710144043, "learning_rate": 1.5437819871978178e-05, "loss": 0.0175, "step": 75580 }, { "grad_norm": 0.46390268206596375, "learning_rate": 1.5425873428486992e-05, "loss": 0.0233, "step": 75590 }, { "grad_norm": 0.2468464970588684, "learning_rate": 1.5413930765970312e-05, "loss": 0.0162, "step": 75600 }, { "grad_norm": 0.23684656620025635, "learning_rate": 1.5401991885734175e-05, "loss": 0.0211, "step": 75610 }, { "grad_norm": 0.24304643273353577, "learning_rate": 1.5390056789084174e-05, "loss": 0.0188, "step": 75620 }, { "grad_norm": 0.32938480377197266, "learning_rate": 1.5378125477325545e-05, "loss": 0.0139, "step": 75630 }, { "grad_norm": 0.3564218282699585, "learning_rate": 1.5366197951763046e-05, "loss": 0.0156, "step": 75640 }, { "grad_norm": 0.39959365129470825, "learning_rate": 1.5354274213701075e-05, "loss": 0.021, "step": 75650 }, { "grad_norm": 0.7669853568077087, "learning_rate": 1.5342354264443566e-05, "loss": 0.0176, "step": 75660 }, { "grad_norm": 0.16297411918640137, "learning_rate": 1.533043810529411e-05, "loss": 0.0169, "step": 75670 }, { "grad_norm": 0.36582183837890625, "learning_rate": 1.5318525737555807e-05, "loss": 0.0211, "step": 75680 }, { "grad_norm": 0.5481477379798889, "learning_rate": 1.53066171625314e-05, "loss": 0.0207, "step": 75690 }, { "grad_norm": 0.2605963945388794, "learning_rate": 1.529471238152317e-05, "loss": 0.0223, "step": 75700 }, { "grad_norm": 0.20494714379310608, "learning_rate": 1.5282811395833023e-05, "loss": 0.0158, "step": 75710 }, { "grad_norm": 0.2918739318847656, "learning_rate": 1.5270914206762397e-05, "loss": 0.0147, "step": 75720 }, { "grad_norm": 0.9828976988792419, "learning_rate": 1.525902081561241e-05, "loss": 0.0187, "step": 75730 }, { "grad_norm": 0.2547916769981384, "learning_rate": 1.5247131223683664e-05, "loss": 0.0168, "step": 75740 }, { "grad_norm": 0.31045788526535034, "learning_rate": 1.5235245432276408e-05, "loss": 0.0154, "step": 75750 }, { "grad_norm": 0.36159899830818176, "learning_rate": 1.5223363442690424e-05, "loss": 0.0137, "step": 75760 }, { "grad_norm": 0.3283165693283081, "learning_rate": 1.5211485256225133e-05, "loss": 0.0194, "step": 75770 }, { "grad_norm": 0.3997088074684143, "learning_rate": 1.519961087417951e-05, "loss": 0.0143, "step": 75780 }, { "grad_norm": 0.14426067471504211, "learning_rate": 1.5187740297852137e-05, "loss": 0.0163, "step": 75790 }, { "grad_norm": 0.28059399127960205, "learning_rate": 1.517587352854113e-05, "loss": 0.0124, "step": 75800 }, { "grad_norm": 0.31189537048339844, "learning_rate": 1.516401056754423e-05, "loss": 0.0184, "step": 75810 }, { "grad_norm": 0.3370392620563507, "learning_rate": 1.5152151416158783e-05, "loss": 0.0211, "step": 75820 }, { "grad_norm": 0.3971266448497772, "learning_rate": 1.5140296075681642e-05, "loss": 0.0192, "step": 75830 }, { "grad_norm": 0.14611345529556274, "learning_rate": 1.5128444547409331e-05, "loss": 0.0154, "step": 75840 }, { "grad_norm": 0.37018364667892456, "learning_rate": 1.5116596832637864e-05, "loss": 0.0173, "step": 75850 }, { "grad_norm": 0.3302169144153595, "learning_rate": 1.5104752932662952e-05, "loss": 0.0144, "step": 75860 }, { "grad_norm": 0.35106319189071655, "learning_rate": 1.5092912848779778e-05, "loss": 0.0119, "step": 75870 }, { "grad_norm": 0.3139452040195465, "learning_rate": 1.5081076582283194e-05, "loss": 0.0152, "step": 75880 }, { "grad_norm": 0.304697722196579, "learning_rate": 1.5069244134467553e-05, "loss": 0.0164, "step": 75890 }, { "grad_norm": 0.35919588804244995, "learning_rate": 1.5057415506626876e-05, "loss": 0.0192, "step": 75900 }, { "grad_norm": 0.3073255717754364, "learning_rate": 1.5045590700054668e-05, "loss": 0.0209, "step": 75910 }, { "grad_norm": 0.14229796826839447, "learning_rate": 1.5033769716044144e-05, "loss": 0.0169, "step": 75920 }, { "grad_norm": 0.30012059211730957, "learning_rate": 1.5021952555887975e-05, "loss": 0.0157, "step": 75930 }, { "grad_norm": 0.391539067029953, "learning_rate": 1.50101392208785e-05, "loss": 0.0162, "step": 75940 }, { "grad_norm": 0.3325709104537964, "learning_rate": 1.4998329712307579e-05, "loss": 0.0158, "step": 75950 }, { "grad_norm": 0.319091260433197, "learning_rate": 1.4986524031466708e-05, "loss": 0.0139, "step": 75960 }, { "grad_norm": 0.2705199718475342, "learning_rate": 1.4974722179646916e-05, "loss": 0.0129, "step": 75970 }, { "grad_norm": 0.5060280561447144, "learning_rate": 1.4962924158138847e-05, "loss": 0.0202, "step": 75980 }, { "grad_norm": 0.28777432441711426, "learning_rate": 1.4951129968232714e-05, "loss": 0.0128, "step": 75990 }, { "grad_norm": 0.28886494040489197, "learning_rate": 1.4939339611218329e-05, "loss": 0.0184, "step": 76000 }, { "grad_norm": 0.27087512612342834, "learning_rate": 1.4927553088385032e-05, "loss": 0.0173, "step": 76010 }, { "grad_norm": 0.3819877803325653, "learning_rate": 1.4915770401021801e-05, "loss": 0.0177, "step": 76020 }, { "grad_norm": 0.2698383927345276, "learning_rate": 1.4903991550417185e-05, "loss": 0.0152, "step": 76030 }, { "grad_norm": 0.4124126434326172, "learning_rate": 1.4892216537859277e-05, "loss": 0.0168, "step": 76040 }, { "grad_norm": 0.3821916878223419, "learning_rate": 1.4880445364635775e-05, "loss": 0.0216, "step": 76050 }, { "grad_norm": 0.30716240406036377, "learning_rate": 1.4868678032033973e-05, "loss": 0.0192, "step": 76060 }, { "grad_norm": 0.2331835776567459, "learning_rate": 1.4856914541340739e-05, "loss": 0.0148, "step": 76070 }, { "grad_norm": 0.23976653814315796, "learning_rate": 1.4845154893842472e-05, "loss": 0.0219, "step": 76080 }, { "grad_norm": 1.2910815477371216, "learning_rate": 1.4833399090825234e-05, "loss": 0.0194, "step": 76090 }, { "grad_norm": 0.32738256454467773, "learning_rate": 1.4821647133574562e-05, "loss": 0.017, "step": 76100 }, { "grad_norm": 0.4475526213645935, "learning_rate": 1.4809899023375706e-05, "loss": 0.0166, "step": 76110 }, { "grad_norm": 0.37078914046287537, "learning_rate": 1.4798154761513366e-05, "loss": 0.0166, "step": 76120 }, { "grad_norm": 0.3397109806537628, "learning_rate": 1.4786414349271916e-05, "loss": 0.0179, "step": 76130 }, { "grad_norm": 0.2560087740421295, "learning_rate": 1.4774677787935226e-05, "loss": 0.0156, "step": 76140 }, { "grad_norm": 0.2665354609489441, "learning_rate": 1.4762945078786838e-05, "loss": 0.0183, "step": 76150 }, { "grad_norm": 0.2676752507686615, "learning_rate": 1.4751216223109777e-05, "loss": 0.0158, "step": 76160 }, { "grad_norm": 0.29883137345314026, "learning_rate": 1.4739491222186714e-05, "loss": 0.0151, "step": 76170 }, { "grad_norm": 0.18075008690357208, "learning_rate": 1.4727770077299879e-05, "loss": 0.0167, "step": 76180 }, { "grad_norm": 0.23022468388080597, "learning_rate": 1.4716052789731094e-05, "loss": 0.0123, "step": 76190 }, { "grad_norm": 0.1336791217327118, "learning_rate": 1.470433936076171e-05, "loss": 0.0129, "step": 76200 }, { "grad_norm": 0.3776613473892212, "learning_rate": 1.4692629791672719e-05, "loss": 0.016, "step": 76210 }, { "grad_norm": 0.17658545076847076, "learning_rate": 1.468092408374464e-05, "loss": 0.0131, "step": 76220 }, { "grad_norm": 0.333686888217926, "learning_rate": 1.4669222238257602e-05, "loss": 0.0182, "step": 76230 }, { "grad_norm": 0.22921380400657654, "learning_rate": 1.4657524256491301e-05, "loss": 0.015, "step": 76240 }, { "grad_norm": 0.3310670554637909, "learning_rate": 1.464583013972502e-05, "loss": 0.0176, "step": 76250 }, { "grad_norm": 0.35637056827545166, "learning_rate": 1.4634139889237591e-05, "loss": 0.0146, "step": 76260 }, { "grad_norm": 0.24187080562114716, "learning_rate": 1.4622453506307448e-05, "loss": 0.0128, "step": 76270 }, { "grad_norm": 0.3414997160434723, "learning_rate": 1.461077099221262e-05, "loss": 0.0158, "step": 76280 }, { "grad_norm": 0.24893313646316528, "learning_rate": 1.4599092348230652e-05, "loss": 0.0168, "step": 76290 }, { "grad_norm": 0.5121270418167114, "learning_rate": 1.4587417575638724e-05, "loss": 0.0134, "step": 76300 }, { "grad_norm": 0.4437333047389984, "learning_rate": 1.4575746675713563e-05, "loss": 0.0228, "step": 76310 }, { "grad_norm": 0.17163147032260895, "learning_rate": 1.4564079649731504e-05, "loss": 0.0189, "step": 76320 }, { "grad_norm": 0.5670202970504761, "learning_rate": 1.4552416498968402e-05, "loss": 0.0153, "step": 76330 }, { "grad_norm": 0.24656014144420624, "learning_rate": 1.4540757224699747e-05, "loss": 0.0124, "step": 76340 }, { "grad_norm": 0.3024652600288391, "learning_rate": 1.452910182820056e-05, "loss": 0.0139, "step": 76350 }, { "grad_norm": 0.3426825702190399, "learning_rate": 1.4517450310745468e-05, "loss": 0.0159, "step": 76360 }, { "grad_norm": 0.34072306752204895, "learning_rate": 1.450580267360866e-05, "loss": 0.0173, "step": 76370 }, { "grad_norm": 0.4862423539161682, "learning_rate": 1.4494158918063922e-05, "loss": 0.0139, "step": 76380 }, { "grad_norm": 0.3196312487125397, "learning_rate": 1.4482519045384568e-05, "loss": 0.0183, "step": 76390 }, { "grad_norm": 0.1909196525812149, "learning_rate": 1.4470883056843548e-05, "loss": 0.0191, "step": 76400 }, { "grad_norm": 0.2383088916540146, "learning_rate": 1.4459250953713322e-05, "loss": 0.0191, "step": 76410 }, { "grad_norm": 0.30394116044044495, "learning_rate": 1.444762273726598e-05, "loss": 0.0179, "step": 76420 }, { "grad_norm": 0.20805273950099945, "learning_rate": 1.443599840877316e-05, "loss": 0.0179, "step": 76430 }, { "grad_norm": 0.2735452950000763, "learning_rate": 1.4424377969506103e-05, "loss": 0.0181, "step": 76440 }, { "grad_norm": 0.16873352229595184, "learning_rate": 1.4412761420735565e-05, "loss": 0.0141, "step": 76450 }, { "grad_norm": 0.3461838662624359, "learning_rate": 1.4401148763731953e-05, "loss": 0.0168, "step": 76460 }, { "grad_norm": 0.22069427371025085, "learning_rate": 1.4389539999765173e-05, "loss": 0.0157, "step": 76470 }, { "grad_norm": 0.24804095923900604, "learning_rate": 1.4377935130104758e-05, "loss": 0.0174, "step": 76480 }, { "grad_norm": 0.2806553244590759, "learning_rate": 1.4366334156019801e-05, "loss": 0.0213, "step": 76490 }, { "grad_norm": 0.3426458239555359, "learning_rate": 1.435473707877898e-05, "loss": 0.015, "step": 76500 }, { "grad_norm": 0.2209652066230774, "learning_rate": 1.4343143899650508e-05, "loss": 0.0126, "step": 76510 }, { "grad_norm": 0.32503682374954224, "learning_rate": 1.4331554619902204e-05, "loss": 0.0177, "step": 76520 }, { "grad_norm": 0.36395466327667236, "learning_rate": 1.4319969240801484e-05, "loss": 0.0173, "step": 76530 }, { "grad_norm": 0.3870677649974823, "learning_rate": 1.430838776361526e-05, "loss": 0.0183, "step": 76540 }, { "grad_norm": 0.31225189566612244, "learning_rate": 1.4296810189610093e-05, "loss": 0.0167, "step": 76550 }, { "grad_norm": 0.44322624802589417, "learning_rate": 1.4285236520052087e-05, "loss": 0.0148, "step": 76560 }, { "grad_norm": 0.20743469893932343, "learning_rate": 1.4273666756206933e-05, "loss": 0.0157, "step": 76570 }, { "grad_norm": 0.26417306065559387, "learning_rate": 1.4262100899339848e-05, "loss": 0.0158, "step": 76580 }, { "grad_norm": 0.31201818585395813, "learning_rate": 1.4250538950715696e-05, "loss": 0.0203, "step": 76590 }, { "grad_norm": 0.19222065806388855, "learning_rate": 1.4238980911598842e-05, "loss": 0.0135, "step": 76600 }, { "grad_norm": 0.28968656063079834, "learning_rate": 1.4227426783253267e-05, "loss": 0.0219, "step": 76610 }, { "grad_norm": 0.4260767996311188, "learning_rate": 1.4215876566942521e-05, "loss": 0.0204, "step": 76620 }, { "grad_norm": 0.31404104828834534, "learning_rate": 1.4204330263929721e-05, "loss": 0.0123, "step": 76630 }, { "grad_norm": 0.19438014924526215, "learning_rate": 1.4192787875477537e-05, "loss": 0.0195, "step": 76640 }, { "grad_norm": 0.17853492498397827, "learning_rate": 1.4181249402848246e-05, "loss": 0.0137, "step": 76650 }, { "grad_norm": 0.32582226395606995, "learning_rate": 1.4169714847303656e-05, "loss": 0.0136, "step": 76660 }, { "grad_norm": 0.23923137784004211, "learning_rate": 1.4158184210105197e-05, "loss": 0.0252, "step": 76670 }, { "grad_norm": 0.22633519768714905, "learning_rate": 1.4146657492513787e-05, "loss": 0.0191, "step": 76680 }, { "grad_norm": 0.19450174272060394, "learning_rate": 1.4135134695790047e-05, "loss": 0.0159, "step": 76690 }, { "grad_norm": 0.446122407913208, "learning_rate": 1.4123615821194037e-05, "loss": 0.0182, "step": 76700 }, { "grad_norm": 0.10428318381309509, "learning_rate": 1.4112100869985473e-05, "loss": 0.0153, "step": 76710 }, { "grad_norm": 0.26366353034973145, "learning_rate": 1.4100589843423589e-05, "loss": 0.018, "step": 76720 }, { "grad_norm": 0.212649405002594, "learning_rate": 1.408908274276723e-05, "loss": 0.0194, "step": 76730 }, { "grad_norm": 0.30074772238731384, "learning_rate": 1.4077579569274786e-05, "loss": 0.0146, "step": 76740 }, { "grad_norm": 0.336117148399353, "learning_rate": 1.4066080324204245e-05, "loss": 0.0205, "step": 76750 }, { "grad_norm": 0.22520248591899872, "learning_rate": 1.405458500881312e-05, "loss": 0.0195, "step": 76760 }, { "grad_norm": 0.1810815930366516, "learning_rate": 1.4043093624358538e-05, "loss": 0.0154, "step": 76770 }, { "grad_norm": 0.3627561926841736, "learning_rate": 1.4031606172097189e-05, "loss": 0.0221, "step": 76780 }, { "grad_norm": 0.1823781579732895, "learning_rate": 1.40201226532853e-05, "loss": 0.0151, "step": 76790 }, { "grad_norm": 0.23944301903247833, "learning_rate": 1.4008643069178701e-05, "loss": 0.0143, "step": 76800 }, { "grad_norm": 0.23878277838230133, "learning_rate": 1.399716742103278e-05, "loss": 0.0124, "step": 76810 }, { "grad_norm": 0.276237428188324, "learning_rate": 1.398569571010252e-05, "loss": 0.0108, "step": 76820 }, { "grad_norm": 0.17766670882701874, "learning_rate": 1.3974227937642414e-05, "loss": 0.0162, "step": 76830 }, { "grad_norm": 0.24152088165283203, "learning_rate": 1.3962764104906595e-05, "loss": 0.0139, "step": 76840 }, { "grad_norm": 0.29572775959968567, "learning_rate": 1.3951304213148696e-05, "loss": 0.0199, "step": 76850 }, { "grad_norm": 0.22737494111061096, "learning_rate": 1.3939848263621979e-05, "loss": 0.0151, "step": 76860 }, { "grad_norm": 0.2126908004283905, "learning_rate": 1.3928396257579219e-05, "loss": 0.0178, "step": 76870 }, { "grad_norm": 0.2706156075000763, "learning_rate": 1.3916948196272833e-05, "loss": 0.0139, "step": 76880 }, { "grad_norm": 0.22221919894218445, "learning_rate": 1.3905504080954734e-05, "loss": 0.0229, "step": 76890 }, { "grad_norm": 0.2721996009349823, "learning_rate": 1.3894063912876454e-05, "loss": 0.015, "step": 76900 }, { "grad_norm": 0.5916497707366943, "learning_rate": 1.3882627693289046e-05, "loss": 0.0178, "step": 76910 }, { "grad_norm": 0.38005945086479187, "learning_rate": 1.3871195423443183e-05, "loss": 0.0166, "step": 76920 }, { "grad_norm": 0.43924447894096375, "learning_rate": 1.3859767104589044e-05, "loss": 0.0131, "step": 76930 }, { "grad_norm": 0.11794394254684448, "learning_rate": 1.3848342737976467e-05, "loss": 0.0119, "step": 76940 }, { "grad_norm": 0.3662508428096771, "learning_rate": 1.3836922324854757e-05, "loss": 0.0169, "step": 76950 }, { "grad_norm": 0.30192455649375916, "learning_rate": 1.382550586647287e-05, "loss": 0.0147, "step": 76960 }, { "grad_norm": 0.20663347840309143, "learning_rate": 1.3814093364079257e-05, "loss": 0.0161, "step": 76970 }, { "grad_norm": 0.37714841961860657, "learning_rate": 1.3802684818921991e-05, "loss": 0.0163, "step": 76980 }, { "grad_norm": 0.12749384343624115, "learning_rate": 1.3791280232248687e-05, "loss": 0.0184, "step": 76990 }, { "grad_norm": 0.2509436011314392, "learning_rate": 1.3779879605306555e-05, "loss": 0.0187, "step": 77000 }, { "grad_norm": 3.7571158409118652, "learning_rate": 1.376848293934232e-05, "loss": 0.0205, "step": 77010 }, { "grad_norm": 0.2793976366519928, "learning_rate": 1.3757090235602316e-05, "loss": 0.0163, "step": 77020 }, { "grad_norm": 0.24773858487606049, "learning_rate": 1.3745701495332447e-05, "loss": 0.0209, "step": 77030 }, { "grad_norm": 0.3038942217826843, "learning_rate": 1.3734316719778145e-05, "loss": 0.0181, "step": 77040 }, { "grad_norm": 0.3200295865535736, "learning_rate": 1.3722935910184454e-05, "loss": 0.0168, "step": 77050 }, { "grad_norm": 0.2916581928730011, "learning_rate": 1.3711559067795921e-05, "loss": 0.0168, "step": 77060 }, { "grad_norm": 0.29086020588874817, "learning_rate": 1.3700186193856756e-05, "loss": 0.015, "step": 77070 }, { "grad_norm": 0.18127627670764923, "learning_rate": 1.368881728961064e-05, "loss": 0.0127, "step": 77080 }, { "grad_norm": 0.2202625721693039, "learning_rate": 1.3677452356300885e-05, "loss": 0.0196, "step": 77090 }, { "grad_norm": 0.20243120193481445, "learning_rate": 1.3666091395170321e-05, "loss": 0.0168, "step": 77100 }, { "grad_norm": 0.42118996381759644, "learning_rate": 1.3654734407461384e-05, "loss": 0.0248, "step": 77110 }, { "grad_norm": 0.4259074926376343, "learning_rate": 1.3643381394416022e-05, "loss": 0.0204, "step": 77120 }, { "grad_norm": 0.22829373180866241, "learning_rate": 1.3632032357275838e-05, "loss": 0.0191, "step": 77130 }, { "grad_norm": 0.3396952152252197, "learning_rate": 1.3620687297281898e-05, "loss": 0.0173, "step": 77140 }, { "grad_norm": 0.29722321033477783, "learning_rate": 1.3609346215674917e-05, "loss": 0.0161, "step": 77150 }, { "grad_norm": 0.30729377269744873, "learning_rate": 1.3598009113695109e-05, "loss": 0.017, "step": 77160 }, { "grad_norm": 0.2850428819656372, "learning_rate": 1.358667599258231e-05, "loss": 0.0195, "step": 77170 }, { "grad_norm": 0.20832520723342896, "learning_rate": 1.357534685357586e-05, "loss": 0.0135, "step": 77180 }, { "grad_norm": 0.30274999141693115, "learning_rate": 1.3564021697914715e-05, "loss": 0.0158, "step": 77190 }, { "grad_norm": 0.21762384474277496, "learning_rate": 1.3552700526837381e-05, "loss": 0.0112, "step": 77200 }, { "grad_norm": 0.23845286667346954, "learning_rate": 1.3541383341581925e-05, "loss": 0.0149, "step": 77210 }, { "grad_norm": 0.5320225954055786, "learning_rate": 1.3530070143385965e-05, "loss": 0.0152, "step": 77220 }, { "grad_norm": 0.2572532892227173, "learning_rate": 1.3518760933486702e-05, "loss": 0.0165, "step": 77230 }, { "grad_norm": 0.311167448759079, "learning_rate": 1.3507455713120908e-05, "loss": 0.0204, "step": 77240 }, { "grad_norm": 0.23880864679813385, "learning_rate": 1.3496154483524876e-05, "loss": 0.0176, "step": 77250 }, { "grad_norm": 0.20605042576789856, "learning_rate": 1.3484857245934506e-05, "loss": 0.015, "step": 77260 }, { "grad_norm": 0.32350417971611023, "learning_rate": 1.3473564001585248e-05, "loss": 0.0142, "step": 77270 }, { "grad_norm": 0.13093094527721405, "learning_rate": 1.3462274751712129e-05, "loss": 0.016, "step": 77280 }, { "grad_norm": 0.25430551171302795, "learning_rate": 1.3450989497549698e-05, "loss": 0.0152, "step": 77290 }, { "grad_norm": 0.3010942041873932, "learning_rate": 1.3439708240332115e-05, "loss": 0.0182, "step": 77300 }, { "grad_norm": 0.2718355357646942, "learning_rate": 1.3428430981293043e-05, "loss": 0.0182, "step": 77310 }, { "grad_norm": 0.2454409897327423, "learning_rate": 1.3417157721665808e-05, "loss": 0.0188, "step": 77320 }, { "grad_norm": 0.4438251852989197, "learning_rate": 1.3405888462683186e-05, "loss": 0.0168, "step": 77330 }, { "grad_norm": 0.25583699345588684, "learning_rate": 1.3394623205577595e-05, "loss": 0.0164, "step": 77340 }, { "grad_norm": 0.310207337141037, "learning_rate": 1.3383361951580963e-05, "loss": 0.0141, "step": 77350 }, { "grad_norm": 0.15219582617282867, "learning_rate": 1.337210470192483e-05, "loss": 0.0161, "step": 77360 }, { "grad_norm": 0.7128049731254578, "learning_rate": 1.3360851457840245e-05, "loss": 0.017, "step": 77370 }, { "grad_norm": 0.23252253234386444, "learning_rate": 1.3349602220557856e-05, "loss": 0.0124, "step": 77380 }, { "grad_norm": 0.4338702857494354, "learning_rate": 1.3338356991307865e-05, "loss": 0.0179, "step": 77390 }, { "grad_norm": 0.3505968749523163, "learning_rate": 1.3327115771320048e-05, "loss": 0.0225, "step": 77400 }, { "grad_norm": 0.159199059009552, "learning_rate": 1.3315878561823697e-05, "loss": 0.0165, "step": 77410 }, { "grad_norm": 0.2232203185558319, "learning_rate": 1.330464536404773e-05, "loss": 0.0157, "step": 77420 }, { "grad_norm": 0.24799829721450806, "learning_rate": 1.3293416179220563e-05, "loss": 0.0178, "step": 77430 }, { "grad_norm": 0.3389975428581238, "learning_rate": 1.328219100857021e-05, "loss": 0.0125, "step": 77440 }, { "grad_norm": 0.2888442575931549, "learning_rate": 1.327096985332425e-05, "loss": 0.0178, "step": 77450 }, { "grad_norm": 0.21320393681526184, "learning_rate": 1.3259752714709816e-05, "loss": 0.0174, "step": 77460 }, { "grad_norm": 0.3737637996673584, "learning_rate": 1.3248539593953569e-05, "loss": 0.016, "step": 77470 }, { "grad_norm": 0.24499677121639252, "learning_rate": 1.323733049228178e-05, "loss": 0.0152, "step": 77480 }, { "grad_norm": 0.3487198054790497, "learning_rate": 1.3226125410920276e-05, "loss": 0.0189, "step": 77490 }, { "grad_norm": 0.3116990327835083, "learning_rate": 1.3214924351094388e-05, "loss": 0.0139, "step": 77500 }, { "grad_norm": 0.18774999678134918, "learning_rate": 1.3203727314029069e-05, "loss": 0.0164, "step": 77510 }, { "grad_norm": 0.2707309424877167, "learning_rate": 1.3192534300948811e-05, "loss": 0.0122, "step": 77520 }, { "grad_norm": 0.18799471855163574, "learning_rate": 1.3181345313077675e-05, "loss": 0.0137, "step": 77530 }, { "grad_norm": 0.24943490326404572, "learning_rate": 1.3170160351639243e-05, "loss": 0.0169, "step": 77540 }, { "grad_norm": 0.270783394575119, "learning_rate": 1.3158979417856721e-05, "loss": 0.0223, "step": 77550 }, { "grad_norm": 0.29369017481803894, "learning_rate": 1.31478025129528e-05, "loss": 0.0178, "step": 77560 }, { "grad_norm": 1.1680890321731567, "learning_rate": 1.3136629638149794e-05, "loss": 0.0211, "step": 77570 }, { "grad_norm": 0.2478427290916443, "learning_rate": 1.3125460794669547e-05, "loss": 0.0195, "step": 77580 }, { "grad_norm": 0.17881819605827332, "learning_rate": 1.3114295983733483e-05, "loss": 0.0173, "step": 77590 }, { "grad_norm": 0.27830833196640015, "learning_rate": 1.3103135206562534e-05, "loss": 0.0146, "step": 77600 }, { "grad_norm": 0.3372702896595001, "learning_rate": 1.3091978464377264e-05, "loss": 0.022, "step": 77610 }, { "grad_norm": 0.29725658893585205, "learning_rate": 1.3080825758397724e-05, "loss": 0.0151, "step": 77620 }, { "grad_norm": 0.5893287658691406, "learning_rate": 1.3069677089843568e-05, "loss": 0.0214, "step": 77630 }, { "grad_norm": 0.26730024814605713, "learning_rate": 1.305853245993401e-05, "loss": 0.0245, "step": 77640 }, { "grad_norm": 0.2668672800064087, "learning_rate": 1.3047391869887809e-05, "loss": 0.0134, "step": 77650 }, { "grad_norm": 0.25453615188598633, "learning_rate": 1.3036255320923263e-05, "loss": 0.017, "step": 77660 }, { "grad_norm": 0.2918712794780731, "learning_rate": 1.3025122814258273e-05, "loss": 0.0115, "step": 77670 }, { "grad_norm": 0.4984096586704254, "learning_rate": 1.3013994351110249e-05, "loss": 0.0219, "step": 77680 }, { "grad_norm": 0.13589756190776825, "learning_rate": 1.3002869932696194e-05, "loss": 0.0148, "step": 77690 }, { "grad_norm": 0.266146183013916, "learning_rate": 1.2991749560232664e-05, "loss": 0.0183, "step": 77700 }, { "grad_norm": 0.22260001301765442, "learning_rate": 1.2980633234935768e-05, "loss": 0.0191, "step": 77710 }, { "grad_norm": 0.3318093419075012, "learning_rate": 1.2969520958021153e-05, "loss": 0.0165, "step": 77720 }, { "grad_norm": 0.2571399509906769, "learning_rate": 1.295841273070405e-05, "loss": 0.0138, "step": 77730 }, { "grad_norm": 0.23401980102062225, "learning_rate": 1.2947308554199255e-05, "loss": 0.0145, "step": 77740 }, { "grad_norm": 0.431088387966156, "learning_rate": 1.2936208429721075e-05, "loss": 0.0175, "step": 77750 }, { "grad_norm": 0.2359999269247055, "learning_rate": 1.2925112358483416e-05, "loss": 0.0167, "step": 77760 }, { "grad_norm": 0.28570234775543213, "learning_rate": 1.2914020341699728e-05, "loss": 0.023, "step": 77770 }, { "grad_norm": 0.6448413133621216, "learning_rate": 1.2902932380583032e-05, "loss": 0.0189, "step": 77780 }, { "grad_norm": 0.21473003923892975, "learning_rate": 1.2891848476345863e-05, "loss": 0.015, "step": 77790 }, { "grad_norm": 0.1761571764945984, "learning_rate": 1.2880768630200368e-05, "loss": 0.0148, "step": 77800 }, { "grad_norm": 0.22116205096244812, "learning_rate": 1.286969284335819e-05, "loss": 0.0169, "step": 77810 }, { "grad_norm": 0.6418443918228149, "learning_rate": 1.2858621117030578e-05, "loss": 0.0209, "step": 77820 }, { "grad_norm": 0.26909583806991577, "learning_rate": 1.2847553452428324e-05, "loss": 0.012, "step": 77830 }, { "grad_norm": 0.5386825203895569, "learning_rate": 1.283648985076178e-05, "loss": 0.0135, "step": 77840 }, { "grad_norm": 0.27196288108825684, "learning_rate": 1.2825430313240817e-05, "loss": 0.0168, "step": 77850 }, { "grad_norm": 0.4977077543735504, "learning_rate": 1.281437484107491e-05, "loss": 0.0184, "step": 77860 }, { "grad_norm": 0.25028789043426514, "learning_rate": 1.2803323435473058e-05, "loss": 0.0228, "step": 77870 }, { "grad_norm": 0.21087031066417694, "learning_rate": 1.2792276097643841e-05, "loss": 0.0167, "step": 77880 }, { "grad_norm": 0.19656701385974884, "learning_rate": 1.2781232828795342e-05, "loss": 0.015, "step": 77890 }, { "grad_norm": 0.27879539132118225, "learning_rate": 1.2770193630135292e-05, "loss": 0.0242, "step": 77900 }, { "grad_norm": 0.2421031892299652, "learning_rate": 1.2759158502870877e-05, "loss": 0.0273, "step": 77910 }, { "grad_norm": 0.357342392206192, "learning_rate": 1.2748127448208913e-05, "loss": 0.0185, "step": 77920 }, { "grad_norm": 0.3573649525642395, "learning_rate": 1.2737100467355706e-05, "loss": 0.0195, "step": 77930 }, { "grad_norm": 0.2950400412082672, "learning_rate": 1.2726077561517169e-05, "loss": 0.0266, "step": 77940 }, { "grad_norm": 0.2447831630706787, "learning_rate": 1.2715058731898755e-05, "loss": 0.0138, "step": 77950 }, { "grad_norm": 0.36703363060951233, "learning_rate": 1.2704043979705471e-05, "loss": 0.0136, "step": 77960 }, { "grad_norm": 0.49194279313087463, "learning_rate": 1.269303330614185e-05, "loss": 0.0195, "step": 77970 }, { "grad_norm": 0.19577735662460327, "learning_rate": 1.2682026712412015e-05, "loss": 0.0171, "step": 77980 }, { "grad_norm": 0.27316027879714966, "learning_rate": 1.2671024199719644e-05, "loss": 0.0145, "step": 77990 }, { "grad_norm": 0.30687835812568665, "learning_rate": 1.2660025769267936e-05, "loss": 0.0174, "step": 78000 }, { "grad_norm": 0.27101942896842957, "learning_rate": 1.2649031422259667e-05, "loss": 0.0184, "step": 78010 }, { "grad_norm": 0.25158360600471497, "learning_rate": 1.2638041159897163e-05, "loss": 0.0145, "step": 78020 }, { "grad_norm": 0.2544253468513489, "learning_rate": 1.262705498338232e-05, "loss": 0.0163, "step": 78030 }, { "grad_norm": 0.32717031240463257, "learning_rate": 1.2616072893916542e-05, "loss": 0.026, "step": 78040 }, { "grad_norm": 0.3710085153579712, "learning_rate": 1.2605094892700836e-05, "loss": 0.0163, "step": 78050 }, { "grad_norm": 0.18560171127319336, "learning_rate": 1.259412098093572e-05, "loss": 0.0133, "step": 78060 }, { "grad_norm": 0.37427911162376404, "learning_rate": 1.258315115982131e-05, "loss": 0.0144, "step": 78070 }, { "grad_norm": 0.2941057085990906, "learning_rate": 1.2572185430557203e-05, "loss": 0.0136, "step": 78080 }, { "grad_norm": 0.31609174609184265, "learning_rate": 1.2561223794342659e-05, "loss": 0.015, "step": 78090 }, { "grad_norm": 0.18248510360717773, "learning_rate": 1.255026625237637e-05, "loss": 0.0151, "step": 78100 }, { "grad_norm": 0.20433467626571655, "learning_rate": 1.2539312805856678e-05, "loss": 0.0199, "step": 78110 }, { "grad_norm": 0.22002649307250977, "learning_rate": 1.25283634559814e-05, "loss": 0.0142, "step": 78120 }, { "grad_norm": 0.25479093194007874, "learning_rate": 1.2517418203947972e-05, "loss": 0.0158, "step": 78130 }, { "grad_norm": 0.33268383145332336, "learning_rate": 1.2506477050953308e-05, "loss": 0.014, "step": 78140 }, { "grad_norm": 0.35315263271331787, "learning_rate": 1.2495539998193972e-05, "loss": 0.0158, "step": 78150 }, { "grad_norm": 0.33143356442451477, "learning_rate": 1.2484607046865975e-05, "loss": 0.0212, "step": 78160 }, { "grad_norm": 0.20212121307849884, "learning_rate": 1.2473678198164968e-05, "loss": 0.016, "step": 78170 }, { "grad_norm": 0.3988669514656067, "learning_rate": 1.2462753453286075e-05, "loss": 0.0207, "step": 78180 }, { "grad_norm": 0.26587992906570435, "learning_rate": 1.2451832813424031e-05, "loss": 0.0287, "step": 78190 }, { "grad_norm": 0.27244481444358826, "learning_rate": 1.2440916279773095e-05, "loss": 0.0196, "step": 78200 }, { "grad_norm": 0.152750626206398, "learning_rate": 1.2430003853527101e-05, "loss": 0.0145, "step": 78210 }, { "grad_norm": 0.5253390669822693, "learning_rate": 1.2419095535879383e-05, "loss": 0.0212, "step": 78220 }, { "grad_norm": 0.6953704357147217, "learning_rate": 1.2408191328022878e-05, "loss": 0.0144, "step": 78230 }, { "grad_norm": 0.4685295522212982, "learning_rate": 1.2397291231150066e-05, "loss": 0.0199, "step": 78240 }, { "grad_norm": 0.46259233355522156, "learning_rate": 1.2386395246452937e-05, "loss": 0.0206, "step": 78250 }, { "grad_norm": 0.2899385988712311, "learning_rate": 1.2375503375123083e-05, "loss": 0.0172, "step": 78260 }, { "grad_norm": 0.23258237540721893, "learning_rate": 1.2364615618351583e-05, "loss": 0.0207, "step": 78270 }, { "grad_norm": 0.2681514620780945, "learning_rate": 1.2353731977329169e-05, "loss": 0.0158, "step": 78280 }, { "grad_norm": 0.2862253487110138, "learning_rate": 1.2342852453246007e-05, "loss": 0.0155, "step": 78290 }, { "grad_norm": 0.22693361341953278, "learning_rate": 1.2331977047291898e-05, "loss": 0.02, "step": 78300 }, { "grad_norm": 0.22671586275100708, "learning_rate": 1.2321105760656137e-05, "loss": 0.0152, "step": 78310 }, { "grad_norm": 0.2662108838558197, "learning_rate": 1.2310238594527613e-05, "loss": 0.0178, "step": 78320 }, { "grad_norm": 0.37257957458496094, "learning_rate": 1.2299375550094693e-05, "loss": 0.0156, "step": 78330 }, { "grad_norm": 0.23991549015045166, "learning_rate": 1.2288516628545421e-05, "loss": 0.0139, "step": 78340 }, { "grad_norm": 0.23114711046218872, "learning_rate": 1.2277661831067255e-05, "loss": 0.0185, "step": 78350 }, { "grad_norm": 0.30513864755630493, "learning_rate": 1.2266811158847286e-05, "loss": 0.0188, "step": 78360 }, { "grad_norm": 0.1830253005027771, "learning_rate": 1.2255964613072107e-05, "loss": 0.0138, "step": 78370 }, { "grad_norm": 0.18459483981132507, "learning_rate": 1.2245122194927905e-05, "loss": 0.0176, "step": 78380 }, { "grad_norm": 0.20630784332752228, "learning_rate": 1.2234283905600364e-05, "loss": 0.018, "step": 78390 }, { "grad_norm": 0.2781583368778229, "learning_rate": 1.222344974627475e-05, "loss": 0.0166, "step": 78400 }, { "grad_norm": 0.257133424282074, "learning_rate": 1.2212619718135875e-05, "loss": 0.015, "step": 78410 }, { "grad_norm": 0.17571984231472015, "learning_rate": 1.2201793822368118e-05, "loss": 0.0128, "step": 78420 }, { "grad_norm": 0.2091294378042221, "learning_rate": 1.2190972060155337e-05, "loss": 0.0118, "step": 78430 }, { "grad_norm": 0.5310021638870239, "learning_rate": 1.2180154432681013e-05, "loss": 0.0191, "step": 78440 }, { "grad_norm": 0.3725377023220062, "learning_rate": 1.2169340941128148e-05, "loss": 0.0151, "step": 78450 }, { "grad_norm": 0.5375611782073975, "learning_rate": 1.2158531586679267e-05, "loss": 0.0163, "step": 78460 }, { "grad_norm": 0.16261741518974304, "learning_rate": 1.2147726370516477e-05, "loss": 0.0134, "step": 78470 }, { "grad_norm": 0.45652836561203003, "learning_rate": 1.2136925293821422e-05, "loss": 0.0198, "step": 78480 }, { "grad_norm": 0.5256456136703491, "learning_rate": 1.2126128357775306e-05, "loss": 0.0164, "step": 78490 }, { "grad_norm": 0.27329912781715393, "learning_rate": 1.2115335563558838e-05, "loss": 0.0126, "step": 78500 }, { "grad_norm": 0.18946853280067444, "learning_rate": 1.2104546912352327e-05, "loss": 0.0137, "step": 78510 }, { "grad_norm": 0.30798500776290894, "learning_rate": 1.2093762405335557e-05, "loss": 0.0129, "step": 78520 }, { "grad_norm": 0.3422413468360901, "learning_rate": 1.2082982043687974e-05, "loss": 0.0127, "step": 78530 }, { "grad_norm": 0.2595866322517395, "learning_rate": 1.207220582858845e-05, "loss": 0.0152, "step": 78540 }, { "grad_norm": 0.18201692402362823, "learning_rate": 1.206143376121549e-05, "loss": 0.0122, "step": 78550 }, { "grad_norm": 0.21602904796600342, "learning_rate": 1.2050665842747078e-05, "loss": 0.0124, "step": 78560 }, { "grad_norm": 0.2398868054151535, "learning_rate": 1.2039902074360804e-05, "loss": 0.0127, "step": 78570 }, { "grad_norm": 0.19650951027870178, "learning_rate": 1.2029142457233755e-05, "loss": 0.0142, "step": 78580 }, { "grad_norm": 0.19949136674404144, "learning_rate": 1.2018386992542601e-05, "loss": 0.0128, "step": 78590 }, { "grad_norm": 0.16226765513420105, "learning_rate": 1.2007635681463542e-05, "loss": 0.014, "step": 78600 }, { "grad_norm": 0.14927053451538086, "learning_rate": 1.1996888525172334e-05, "loss": 0.0208, "step": 78610 }, { "grad_norm": 0.26371243596076965, "learning_rate": 1.198614552484425e-05, "loss": 0.017, "step": 78620 }, { "grad_norm": 0.29995718598365784, "learning_rate": 1.197540668165415e-05, "loss": 0.0177, "step": 78630 }, { "grad_norm": 0.3510265648365021, "learning_rate": 1.1964671996776395e-05, "loss": 0.0175, "step": 78640 }, { "grad_norm": 0.2436409890651703, "learning_rate": 1.1953941471384922e-05, "loss": 0.02, "step": 78650 }, { "grad_norm": 0.2282654494047165, "learning_rate": 1.1943215106653205e-05, "loss": 0.0187, "step": 78660 }, { "grad_norm": 0.1892627328634262, "learning_rate": 1.193249290375429e-05, "loss": 0.0172, "step": 78670 }, { "grad_norm": 0.34168118238449097, "learning_rate": 1.1921774863860696e-05, "loss": 0.0154, "step": 78680 }, { "grad_norm": 0.4539816975593567, "learning_rate": 1.1911060988144556e-05, "loss": 0.0168, "step": 78690 }, { "grad_norm": 0.2641466557979584, "learning_rate": 1.1900351277777533e-05, "loss": 0.0148, "step": 78700 }, { "grad_norm": 0.39502689242362976, "learning_rate": 1.1889645733930798e-05, "loss": 0.0211, "step": 78710 }, { "grad_norm": 0.21243757009506226, "learning_rate": 1.1878944357775107e-05, "loss": 0.0147, "step": 78720 }, { "grad_norm": 0.2833685278892517, "learning_rate": 1.1868247150480749e-05, "loss": 0.0173, "step": 78730 }, { "grad_norm": 0.26748165488243103, "learning_rate": 1.1857554113217567e-05, "loss": 0.0119, "step": 78740 }, { "grad_norm": 0.4048636853694916, "learning_rate": 1.1846865247154903e-05, "loss": 0.0182, "step": 78750 }, { "grad_norm": 0.13431519269943237, "learning_rate": 1.1836180553461711e-05, "loss": 0.015, "step": 78760 }, { "grad_norm": 0.3685758113861084, "learning_rate": 1.1825500033306419e-05, "loss": 0.0269, "step": 78770 }, { "grad_norm": 0.24493208527565002, "learning_rate": 1.1814823687857052e-05, "loss": 0.018, "step": 78780 }, { "grad_norm": 0.22376412153244019, "learning_rate": 1.1804151518281153e-05, "loss": 0.0172, "step": 78790 }, { "grad_norm": 0.3268091380596161, "learning_rate": 1.179348352574583e-05, "loss": 0.0287, "step": 78800 }, { "grad_norm": 0.10746534913778305, "learning_rate": 1.1782819711417697e-05, "loss": 0.0183, "step": 78810 }, { "grad_norm": 0.1808445006608963, "learning_rate": 1.1772160076462951e-05, "loss": 0.0135, "step": 78820 }, { "grad_norm": 0.2087317258119583, "learning_rate": 1.1761504622047293e-05, "loss": 0.0189, "step": 78830 }, { "grad_norm": 0.31282854080200195, "learning_rate": 1.1750853349335999e-05, "loss": 0.0142, "step": 78840 }, { "grad_norm": 0.26746392250061035, "learning_rate": 1.1740206259493874e-05, "loss": 0.0213, "step": 78850 }, { "grad_norm": 0.4401058554649353, "learning_rate": 1.1729563353685286e-05, "loss": 0.0151, "step": 78860 }, { "grad_norm": 0.2913140058517456, "learning_rate": 1.1718924633074091e-05, "loss": 0.0179, "step": 78870 }, { "grad_norm": 0.39637911319732666, "learning_rate": 1.1708290098823765e-05, "loss": 0.0173, "step": 78880 }, { "grad_norm": 0.23237842321395874, "learning_rate": 1.1697659752097245e-05, "loss": 0.013, "step": 78890 }, { "grad_norm": 0.36508113145828247, "learning_rate": 1.1687033594057068e-05, "loss": 0.0359, "step": 78900 }, { "grad_norm": 0.23754677176475525, "learning_rate": 1.1676411625865297e-05, "loss": 0.017, "step": 78910 }, { "grad_norm": 0.43731167912483215, "learning_rate": 1.1665793848683542e-05, "loss": 0.0174, "step": 78920 }, { "grad_norm": 0.5041829347610474, "learning_rate": 1.1655180263672927e-05, "loss": 0.0259, "step": 78930 }, { "grad_norm": 0.12900419533252716, "learning_rate": 1.1644570871994142e-05, "loss": 0.0118, "step": 78940 }, { "grad_norm": 0.15368197858333588, "learning_rate": 1.163396567480744e-05, "loss": 0.0158, "step": 78950 }, { "grad_norm": 0.3753712475299835, "learning_rate": 1.1623364673272552e-05, "loss": 0.0138, "step": 78960 }, { "grad_norm": 0.25530311465263367, "learning_rate": 1.1612767868548807e-05, "loss": 0.0166, "step": 78970 }, { "grad_norm": 0.31232279539108276, "learning_rate": 1.1602175261795057e-05, "loss": 0.0185, "step": 78980 }, { "grad_norm": 0.32641851902008057, "learning_rate": 1.1591586854169696e-05, "loss": 0.0133, "step": 78990 }, { "grad_norm": 0.24144446849822998, "learning_rate": 1.158100264683064e-05, "loss": 0.0176, "step": 79000 }, { "grad_norm": 0.3104321360588074, "learning_rate": 1.1570422640935386e-05, "loss": 0.0204, "step": 79010 }, { "grad_norm": 0.5088802576065063, "learning_rate": 1.155984683764092e-05, "loss": 0.0249, "step": 79020 }, { "grad_norm": 0.30036431550979614, "learning_rate": 1.1549275238103812e-05, "loss": 0.0127, "step": 79030 }, { "grad_norm": 0.25416064262390137, "learning_rate": 1.1538707843480152e-05, "loss": 0.0187, "step": 79040 }, { "grad_norm": 0.20144367218017578, "learning_rate": 1.1528144654925593e-05, "loss": 0.0197, "step": 79050 }, { "grad_norm": 0.1964641958475113, "learning_rate": 1.1517585673595277e-05, "loss": 0.015, "step": 79060 }, { "grad_norm": 0.3840346336364746, "learning_rate": 1.150703090064395e-05, "loss": 0.0186, "step": 79070 }, { "grad_norm": 0.2941552400588989, "learning_rate": 1.1496480337225835e-05, "loss": 0.0218, "step": 79080 }, { "grad_norm": 0.543059229850769, "learning_rate": 1.1485933984494751e-05, "loss": 0.0167, "step": 79090 }, { "grad_norm": 0.2287413328886032, "learning_rate": 1.1475391843604e-05, "loss": 0.0175, "step": 79100 }, { "grad_norm": 0.2484467774629593, "learning_rate": 1.1464853915706503e-05, "loss": 0.0231, "step": 79110 }, { "grad_norm": 0.35052672028541565, "learning_rate": 1.1454320201954626e-05, "loss": 0.0176, "step": 79120 }, { "grad_norm": 0.17642100155353546, "learning_rate": 1.1443790703500356e-05, "loss": 0.0133, "step": 79130 }, { "grad_norm": 0.18853893876075745, "learning_rate": 1.1433265421495154e-05, "loss": 0.0152, "step": 79140 }, { "grad_norm": 0.24388158321380615, "learning_rate": 1.1422744357090065e-05, "loss": 0.0252, "step": 79150 }, { "grad_norm": 0.31168168783187866, "learning_rate": 1.1412227511435647e-05, "loss": 0.0223, "step": 79160 }, { "grad_norm": 0.449044406414032, "learning_rate": 1.1401714885682025e-05, "loss": 0.0156, "step": 79170 }, { "grad_norm": 0.27726730704307556, "learning_rate": 1.1391206480978823e-05, "loss": 0.0157, "step": 79180 }, { "grad_norm": 0.18358537554740906, "learning_rate": 1.1380702298475227e-05, "loss": 0.0176, "step": 79190 }, { "grad_norm": 0.3902348577976227, "learning_rate": 1.1370202339319985e-05, "loss": 0.0157, "step": 79200 }, { "grad_norm": 0.37553009390830994, "learning_rate": 1.1359706604661319e-05, "loss": 0.0189, "step": 79210 }, { "grad_norm": 0.34183964133262634, "learning_rate": 1.1349215095647042e-05, "loss": 0.0156, "step": 79220 }, { "grad_norm": 0.17826393246650696, "learning_rate": 1.1338727813424493e-05, "loss": 0.016, "step": 79230 }, { "grad_norm": 0.2982666492462158, "learning_rate": 1.132824475914056e-05, "loss": 0.0143, "step": 79240 }, { "grad_norm": 0.35002586245536804, "learning_rate": 1.131776593394162e-05, "loss": 0.0139, "step": 79250 }, { "grad_norm": 0.33523494005203247, "learning_rate": 1.1307291338973657e-05, "loss": 0.0183, "step": 79260 }, { "grad_norm": 0.2815273702144623, "learning_rate": 1.1296820975382121e-05, "loss": 0.0127, "step": 79270 }, { "grad_norm": 0.2703414559364319, "learning_rate": 1.1286354844312074e-05, "loss": 0.0183, "step": 79280 }, { "grad_norm": 0.27172720432281494, "learning_rate": 1.1275892946908023e-05, "loss": 0.0203, "step": 79290 }, { "grad_norm": 0.2750532329082489, "learning_rate": 1.126543528431413e-05, "loss": 0.0222, "step": 79300 }, { "grad_norm": 0.31576550006866455, "learning_rate": 1.125498185767398e-05, "loss": 0.0145, "step": 79310 }, { "grad_norm": 0.5280277729034424, "learning_rate": 1.1244532668130781e-05, "loss": 0.0199, "step": 79320 }, { "grad_norm": 0.21203409135341644, "learning_rate": 1.1234087716827203e-05, "loss": 0.0224, "step": 79330 }, { "grad_norm": 0.9649298787117004, "learning_rate": 1.1223647004905524e-05, "loss": 0.0204, "step": 79340 }, { "grad_norm": 0.5033332705497742, "learning_rate": 1.1213210533507485e-05, "loss": 0.0201, "step": 79350 }, { "grad_norm": 0.5365344882011414, "learning_rate": 1.1202778303774447e-05, "loss": 0.0168, "step": 79360 }, { "grad_norm": 0.21535484492778778, "learning_rate": 1.1192350316847234e-05, "loss": 0.0171, "step": 79370 }, { "grad_norm": 0.6528973579406738, "learning_rate": 1.1181926573866258e-05, "loss": 0.0174, "step": 79380 }, { "grad_norm": 0.16077712178230286, "learning_rate": 1.1171507075971416e-05, "loss": 0.0207, "step": 79390 }, { "grad_norm": 0.20304375886917114, "learning_rate": 1.1161091824302177e-05, "loss": 0.0251, "step": 79400 }, { "grad_norm": 0.21942900121212006, "learning_rate": 1.115068081999755e-05, "loss": 0.0218, "step": 79410 }, { "grad_norm": 0.29388144612312317, "learning_rate": 1.1140274064196071e-05, "loss": 0.0203, "step": 79420 }, { "grad_norm": 0.3065429627895355, "learning_rate": 1.112987155803578e-05, "loss": 0.0171, "step": 79430 }, { "grad_norm": 0.3661942481994629, "learning_rate": 1.1119473302654298e-05, "loss": 0.0199, "step": 79440 }, { "grad_norm": 0.23412787914276123, "learning_rate": 1.1109079299188769e-05, "loss": 0.0159, "step": 79450 }, { "grad_norm": 0.23796531558036804, "learning_rate": 1.1098689548775847e-05, "loss": 0.0155, "step": 79460 }, { "grad_norm": 0.43125730752944946, "learning_rate": 1.1088304052551762e-05, "loss": 0.0185, "step": 79470 }, { "grad_norm": 0.2667502760887146, "learning_rate": 1.1077922811652215e-05, "loss": 0.0179, "step": 79480 }, { "grad_norm": 0.2643064558506012, "learning_rate": 1.1067545827212534e-05, "loss": 0.021, "step": 79490 }, { "grad_norm": 0.32513266801834106, "learning_rate": 1.1057173100367496e-05, "loss": 0.0153, "step": 79500 }, { "grad_norm": 0.201798215508461, "learning_rate": 1.1046804632251474e-05, "loss": 0.0127, "step": 79510 }, { "grad_norm": 0.16226555407047272, "learning_rate": 1.1036440423998318e-05, "loss": 0.0133, "step": 79520 }, { "grad_norm": 0.273367315530777, "learning_rate": 1.1026080476741469e-05, "loss": 0.024, "step": 79530 }, { "grad_norm": 0.17463752627372742, "learning_rate": 1.1015724791613836e-05, "loss": 0.0144, "step": 79540 }, { "grad_norm": 0.22235740721225739, "learning_rate": 1.1005373369747951e-05, "loss": 0.013, "step": 79550 }, { "grad_norm": 0.19777239859104156, "learning_rate": 1.09950262122758e-05, "loss": 0.0146, "step": 79560 }, { "grad_norm": 0.568867027759552, "learning_rate": 1.0984683320328948e-05, "loss": 0.0223, "step": 79570 }, { "grad_norm": 0.2727949619293213, "learning_rate": 1.0974344695038458e-05, "loss": 0.0149, "step": 79580 }, { "grad_norm": 0.386657178401947, "learning_rate": 1.0964010337534968e-05, "loss": 0.0167, "step": 79590 }, { "grad_norm": 0.23633308708667755, "learning_rate": 1.0953680248948611e-05, "loss": 0.0158, "step": 79600 }, { "grad_norm": 0.3049285411834717, "learning_rate": 1.0943354430409069e-05, "loss": 0.0129, "step": 79610 }, { "grad_norm": 0.24843940138816833, "learning_rate": 1.0933032883045574e-05, "loss": 0.0136, "step": 79620 }, { "grad_norm": 0.40247082710266113, "learning_rate": 1.0922715607986872e-05, "loss": 0.0186, "step": 79630 }, { "grad_norm": 0.19823133945465088, "learning_rate": 1.0912402606361234e-05, "loss": 0.0152, "step": 79640 }, { "grad_norm": 0.3957412838935852, "learning_rate": 1.090209387929647e-05, "loss": 0.015, "step": 79650 }, { "grad_norm": 0.25792834162712097, "learning_rate": 1.0891789427919957e-05, "loss": 0.0204, "step": 79660 }, { "grad_norm": 0.3647330701351166, "learning_rate": 1.088148925335854e-05, "loss": 0.0146, "step": 79670 }, { "grad_norm": 0.2804322838783264, "learning_rate": 1.087119335673864e-05, "loss": 0.0109, "step": 79680 }, { "grad_norm": 0.2561149597167969, "learning_rate": 1.0860901739186208e-05, "loss": 0.015, "step": 79690 }, { "grad_norm": 0.289247989654541, "learning_rate": 1.085061440182673e-05, "loss": 0.016, "step": 79700 }, { "grad_norm": 0.18644705414772034, "learning_rate": 1.0840331345785193e-05, "loss": 0.0127, "step": 79710 }, { "grad_norm": 0.5280910730361938, "learning_rate": 1.083005257218615e-05, "loss": 0.0163, "step": 79720 }, { "grad_norm": 0.3261829912662506, "learning_rate": 1.0819778082153642e-05, "loss": 0.0142, "step": 79730 }, { "grad_norm": 0.17594917118549347, "learning_rate": 1.0809507876811325e-05, "loss": 0.0171, "step": 79740 }, { "grad_norm": 0.1891293227672577, "learning_rate": 1.0799241957282292e-05, "loss": 0.015, "step": 79750 }, { "grad_norm": 0.38820943236351013, "learning_rate": 1.078898032468923e-05, "loss": 0.0187, "step": 79760 }, { "grad_norm": 0.24763821065425873, "learning_rate": 1.0778722980154315e-05, "loss": 0.0128, "step": 79770 }, { "grad_norm": 0.2079722285270691, "learning_rate": 1.0768469924799302e-05, "loss": 0.0125, "step": 79780 }, { "grad_norm": 0.26203784346580505, "learning_rate": 1.0758221159745419e-05, "loss": 0.0172, "step": 79790 }, { "grad_norm": 0.7272084355354309, "learning_rate": 1.074797668611347e-05, "loss": 0.0226, "step": 79800 }, { "grad_norm": 0.23492754995822906, "learning_rate": 1.0737736505023777e-05, "loss": 0.0161, "step": 79810 }, { "grad_norm": 0.16389359533786774, "learning_rate": 1.0727500617596203e-05, "loss": 0.017, "step": 79820 }, { "grad_norm": 0.23536506295204163, "learning_rate": 1.0717269024950104e-05, "loss": 0.0147, "step": 79830 }, { "grad_norm": 0.20311059057712555, "learning_rate": 1.070704172820442e-05, "loss": 0.0207, "step": 79840 }, { "grad_norm": 0.28610119223594666, "learning_rate": 1.0696818728477559e-05, "loss": 0.0169, "step": 79850 }, { "grad_norm": 0.32052817940711975, "learning_rate": 1.0686600026887511e-05, "loss": 0.0134, "step": 79860 }, { "grad_norm": 0.397326797246933, "learning_rate": 1.0676385624551777e-05, "loss": 0.0185, "step": 79870 }, { "grad_norm": 0.3590291440486908, "learning_rate": 1.0666175522587402e-05, "loss": 0.0184, "step": 79880 }, { "grad_norm": 0.2377038300037384, "learning_rate": 1.0655969722110925e-05, "loss": 0.0147, "step": 79890 }, { "grad_norm": 0.20366023480892181, "learning_rate": 1.0645768224238439e-05, "loss": 0.0168, "step": 79900 }, { "grad_norm": 0.4700666666030884, "learning_rate": 1.063557103008559e-05, "loss": 0.0193, "step": 79910 }, { "grad_norm": 0.3531879186630249, "learning_rate": 1.0625378140767494e-05, "loss": 0.0146, "step": 79920 }, { "grad_norm": 0.2287112921476364, "learning_rate": 1.0615189557398846e-05, "loss": 0.0212, "step": 79930 }, { "grad_norm": 0.5725235342979431, "learning_rate": 1.060500528109386e-05, "loss": 0.016, "step": 79940 }, { "grad_norm": 0.17052015662193298, "learning_rate": 1.0594825312966272e-05, "loss": 0.0166, "step": 79950 }, { "grad_norm": 0.3130640387535095, "learning_rate": 1.0584649654129336e-05, "loss": 0.019, "step": 79960 }, { "grad_norm": 0.31730589270591736, "learning_rate": 1.0574478305695863e-05, "loss": 0.0149, "step": 79970 }, { "grad_norm": 0.36363130807876587, "learning_rate": 1.0564311268778154e-05, "loss": 0.0162, "step": 79980 }, { "grad_norm": 0.37119027972221375, "learning_rate": 1.0554148544488073e-05, "loss": 0.0165, "step": 79990 }, { "grad_norm": 0.24061329662799835, "learning_rate": 1.0543990133937004e-05, "loss": 0.0209, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "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": 8, "trial_name": null, "trial_params": null }