{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 54979, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00018188762982229579, "grad_norm": 5.876733303070068, "learning_rate": 5.454545454545455e-08, "loss": 0.1151, "step": 10 }, { "epoch": 0.00036377525964459157, "grad_norm": 4.780970573425293, "learning_rate": 1.1515151515151516e-07, "loss": 0.1178, "step": 20 }, { "epoch": 0.0005456628894668873, "grad_norm": 3.683614730834961, "learning_rate": 1.7575757575757576e-07, "loss": 0.1159, "step": 30 }, { "epoch": 0.0007275505192891831, "grad_norm": 4.592695713043213, "learning_rate": 2.3636363636363637e-07, "loss": 0.1201, "step": 40 }, { "epoch": 0.000909438149111479, "grad_norm": 2.6691765785217285, "learning_rate": 2.96969696969697e-07, "loss": 0.1224, "step": 50 }, { "epoch": 0.0010913257789337747, "grad_norm": 2.3605756759643555, "learning_rate": 3.5757575757575764e-07, "loss": 0.1219, "step": 60 }, { "epoch": 0.0012732134087560705, "grad_norm": 3.01662540435791, "learning_rate": 4.181818181818182e-07, "loss": 0.1172, "step": 70 }, { "epoch": 0.0014551010385783663, "grad_norm": 2.2953855991363525, "learning_rate": 4.787878787878789e-07, "loss": 0.1179, "step": 80 }, { "epoch": 0.001636988668400662, "grad_norm": 2.7672362327575684, "learning_rate": 5.393939393939395e-07, "loss": 0.1154, "step": 90 }, { "epoch": 0.001818876298222958, "grad_norm": 2.4730336666107178, "learning_rate": 6.000000000000001e-07, "loss": 0.117, "step": 100 }, { "epoch": 0.0020007639280452535, "grad_norm": 3.292983293533325, "learning_rate": 6.606060606060606e-07, "loss": 0.1172, "step": 110 }, { "epoch": 0.0021826515578675493, "grad_norm": 3.204930305480957, "learning_rate": 7.212121212121213e-07, "loss": 0.1202, "step": 120 }, { "epoch": 0.002364539187689845, "grad_norm": 3.2186341285705566, "learning_rate": 7.818181818181819e-07, "loss": 0.1157, "step": 130 }, { "epoch": 0.002546426817512141, "grad_norm": 2.2171435356140137, "learning_rate": 8.424242424242425e-07, "loss": 0.1168, "step": 140 }, { "epoch": 0.0027283144473344368, "grad_norm": 3.4097282886505127, "learning_rate": 9.030303030303031e-07, "loss": 0.1193, "step": 150 }, { "epoch": 0.0029102020771567326, "grad_norm": 2.2514302730560303, "learning_rate": 9.636363636363636e-07, "loss": 0.1207, "step": 160 }, { "epoch": 0.0030920897069790284, "grad_norm": 2.446202039718628, "learning_rate": 1.0242424242424242e-06, "loss": 0.1174, "step": 170 }, { "epoch": 0.003273977336801324, "grad_norm": 5.393249034881592, "learning_rate": 1.084848484848485e-06, "loss": 0.1245, "step": 180 }, { "epoch": 0.00345586496662362, "grad_norm": 4.607318878173828, "learning_rate": 1.1454545454545457e-06, "loss": 0.1212, "step": 190 }, { "epoch": 0.003637752596445916, "grad_norm": 5.2917609214782715, "learning_rate": 1.206060606060606e-06, "loss": 0.1203, "step": 200 }, { "epoch": 0.0038196402262682116, "grad_norm": 3.2605690956115723, "learning_rate": 1.2666666666666669e-06, "loss": 0.1186, "step": 210 }, { "epoch": 0.004001527856090507, "grad_norm": 2.811206102371216, "learning_rate": 1.3272727272727273e-06, "loss": 0.1186, "step": 220 }, { "epoch": 0.004183415485912803, "grad_norm": 3.090862274169922, "learning_rate": 1.3878787878787881e-06, "loss": 0.1165, "step": 230 }, { "epoch": 0.004365303115735099, "grad_norm": 2.3865392208099365, "learning_rate": 1.4484848484848485e-06, "loss": 0.1214, "step": 240 }, { "epoch": 0.004547190745557395, "grad_norm": 2.5838823318481445, "learning_rate": 1.5090909090909091e-06, "loss": 0.1233, "step": 250 }, { "epoch": 0.00472907837537969, "grad_norm": 2.2738687992095947, "learning_rate": 1.56969696969697e-06, "loss": 0.1184, "step": 260 }, { "epoch": 0.0049109660052019865, "grad_norm": 4.266605854034424, "learning_rate": 1.6303030303030303e-06, "loss": 0.1224, "step": 270 }, { "epoch": 0.005092853635024282, "grad_norm": 4.353481292724609, "learning_rate": 1.6909090909090912e-06, "loss": 0.1213, "step": 280 }, { "epoch": 0.005274741264846578, "grad_norm": 5.161770343780518, "learning_rate": 1.7515151515151516e-06, "loss": 0.1246, "step": 290 }, { "epoch": 0.0054566288946688735, "grad_norm": 4.2574310302734375, "learning_rate": 1.8121212121212124e-06, "loss": 0.1204, "step": 300 }, { "epoch": 0.00563851652449117, "grad_norm": 3.01313853263855, "learning_rate": 1.872727272727273e-06, "loss": 0.1254, "step": 310 }, { "epoch": 0.005820404154313465, "grad_norm": 4.04472541809082, "learning_rate": 1.9333333333333336e-06, "loss": 0.1237, "step": 320 }, { "epoch": 0.0060022917841357605, "grad_norm": 3.0833067893981934, "learning_rate": 1.993939393939394e-06, "loss": 0.1281, "step": 330 }, { "epoch": 0.006184179413958057, "grad_norm": 4.189347743988037, "learning_rate": 2.054545454545455e-06, "loss": 0.1261, "step": 340 }, { "epoch": 0.006366067043780352, "grad_norm": 2.7407844066619873, "learning_rate": 2.1151515151515152e-06, "loss": 0.1228, "step": 350 }, { "epoch": 0.006547954673602648, "grad_norm": 4.66772985458374, "learning_rate": 2.175757575757576e-06, "loss": 0.1294, "step": 360 }, { "epoch": 0.006729842303424944, "grad_norm": 3.2758281230926514, "learning_rate": 2.2363636363636364e-06, "loss": 0.1249, "step": 370 }, { "epoch": 0.00691172993324724, "grad_norm": 4.250436782836914, "learning_rate": 2.2969696969696973e-06, "loss": 0.1271, "step": 380 }, { "epoch": 0.007093617563069535, "grad_norm": 5.138640403747559, "learning_rate": 2.3575757575757577e-06, "loss": 0.1282, "step": 390 }, { "epoch": 0.007275505192891832, "grad_norm": 3.753122568130493, "learning_rate": 2.4181818181818185e-06, "loss": 0.1322, "step": 400 }, { "epoch": 0.007457392822714127, "grad_norm": 3.447540521621704, "learning_rate": 2.478787878787879e-06, "loss": 0.1283, "step": 410 }, { "epoch": 0.007639280452536423, "grad_norm": 4.796869277954102, "learning_rate": 2.5393939393939397e-06, "loss": 0.1297, "step": 420 }, { "epoch": 0.007821168082358719, "grad_norm": 7.643021583557129, "learning_rate": 2.6e-06, "loss": 0.1287, "step": 430 }, { "epoch": 0.008003055712181014, "grad_norm": 3.2790892124176025, "learning_rate": 2.660606060606061e-06, "loss": 0.1336, "step": 440 }, { "epoch": 0.008184943342003311, "grad_norm": 5.4543843269348145, "learning_rate": 2.7212121212121213e-06, "loss": 0.1326, "step": 450 }, { "epoch": 0.008366830971825607, "grad_norm": 3.2349607944488525, "learning_rate": 2.781818181818182e-06, "loss": 0.1257, "step": 460 }, { "epoch": 0.008548718601647902, "grad_norm": 4.2942705154418945, "learning_rate": 2.8424242424242425e-06, "loss": 0.1288, "step": 470 }, { "epoch": 0.008730606231470197, "grad_norm": 3.6488354206085205, "learning_rate": 2.903030303030303e-06, "loss": 0.1337, "step": 480 }, { "epoch": 0.008912493861292493, "grad_norm": 5.0120015144348145, "learning_rate": 2.963636363636364e-06, "loss": 0.1374, "step": 490 }, { "epoch": 0.00909438149111479, "grad_norm": 3.2477922439575195, "learning_rate": 3.0242424242424246e-06, "loss": 0.1375, "step": 500 }, { "epoch": 0.009276269120937085, "grad_norm": 8.742966651916504, "learning_rate": 3.084848484848485e-06, "loss": 0.1403, "step": 510 }, { "epoch": 0.00945815675075938, "grad_norm": 3.294626474380493, "learning_rate": 3.145454545454546e-06, "loss": 0.1397, "step": 520 }, { "epoch": 0.009640044380581676, "grad_norm": 2.4122986793518066, "learning_rate": 3.2060606060606066e-06, "loss": 0.1379, "step": 530 }, { "epoch": 0.009821932010403973, "grad_norm": 2.8360445499420166, "learning_rate": 3.266666666666667e-06, "loss": 0.138, "step": 540 }, { "epoch": 0.010003819640226268, "grad_norm": 2.840850353240967, "learning_rate": 3.3272727272727274e-06, "loss": 0.1414, "step": 550 }, { "epoch": 0.010185707270048564, "grad_norm": 4.56731653213501, "learning_rate": 3.3878787878787882e-06, "loss": 0.1443, "step": 560 }, { "epoch": 0.01036759489987086, "grad_norm": 2.8126513957977295, "learning_rate": 3.4484848484848486e-06, "loss": 0.1449, "step": 570 }, { "epoch": 0.010549482529693156, "grad_norm": 2.8451006412506104, "learning_rate": 3.509090909090909e-06, "loss": 0.1429, "step": 580 }, { "epoch": 0.010731370159515452, "grad_norm": 2.4222652912139893, "learning_rate": 3.5696969696969703e-06, "loss": 0.1427, "step": 590 }, { "epoch": 0.010913257789337747, "grad_norm": 2.411315679550171, "learning_rate": 3.6303030303030307e-06, "loss": 0.1427, "step": 600 }, { "epoch": 0.011095145419160042, "grad_norm": 3.0026516914367676, "learning_rate": 3.690909090909091e-06, "loss": 0.1475, "step": 610 }, { "epoch": 0.01127703304898234, "grad_norm": 4.224020957946777, "learning_rate": 3.7515151515151515e-06, "loss": 0.149, "step": 620 }, { "epoch": 0.011458920678804635, "grad_norm": 2.231639862060547, "learning_rate": 3.8121212121212127e-06, "loss": 0.1524, "step": 630 }, { "epoch": 0.01164080830862693, "grad_norm": 3.71671199798584, "learning_rate": 3.872727272727273e-06, "loss": 0.152, "step": 640 }, { "epoch": 0.011822695938449226, "grad_norm": 3.882786989212036, "learning_rate": 3.9333333333333335e-06, "loss": 0.1538, "step": 650 }, { "epoch": 0.012004583568271521, "grad_norm": 3.012821674346924, "learning_rate": 3.993939393939394e-06, "loss": 0.1574, "step": 660 }, { "epoch": 0.012186471198093818, "grad_norm": 3.0974013805389404, "learning_rate": 4.054545454545455e-06, "loss": 0.1511, "step": 670 }, { "epoch": 0.012368358827916114, "grad_norm": 2.551675319671631, "learning_rate": 4.115151515151515e-06, "loss": 0.1534, "step": 680 }, { "epoch": 0.012550246457738409, "grad_norm": 3.405935049057007, "learning_rate": 4.175757575757576e-06, "loss": 0.153, "step": 690 }, { "epoch": 0.012732134087560704, "grad_norm": 2.814368963241577, "learning_rate": 4.236363636363637e-06, "loss": 0.1544, "step": 700 }, { "epoch": 0.012914021717383001, "grad_norm": 3.9631764888763428, "learning_rate": 4.296969696969698e-06, "loss": 0.1576, "step": 710 }, { "epoch": 0.013095909347205297, "grad_norm": 4.588418006896973, "learning_rate": 4.3575757575757576e-06, "loss": 0.1591, "step": 720 }, { "epoch": 0.013277796977027592, "grad_norm": 2.5307209491729736, "learning_rate": 4.418181818181818e-06, "loss": 0.1666, "step": 730 }, { "epoch": 0.013459684606849888, "grad_norm": 3.1999096870422363, "learning_rate": 4.478787878787879e-06, "loss": 0.1634, "step": 740 }, { "epoch": 0.013641572236672185, "grad_norm": 4.001588344573975, "learning_rate": 4.539393939393939e-06, "loss": 0.1651, "step": 750 }, { "epoch": 0.01382345986649448, "grad_norm": 2.62697696685791, "learning_rate": 4.600000000000001e-06, "loss": 0.1635, "step": 760 }, { "epoch": 0.014005347496316775, "grad_norm": 2.3145549297332764, "learning_rate": 4.660606060606061e-06, "loss": 0.1666, "step": 770 }, { "epoch": 0.01418723512613907, "grad_norm": 2.57094407081604, "learning_rate": 4.721212121212122e-06, "loss": 0.1655, "step": 780 }, { "epoch": 0.014369122755961366, "grad_norm": 2.779609441757202, "learning_rate": 4.7818181818181825e-06, "loss": 0.1653, "step": 790 }, { "epoch": 0.014551010385783663, "grad_norm": 3.416856288909912, "learning_rate": 4.842424242424243e-06, "loss": 0.1721, "step": 800 }, { "epoch": 0.014732898015605959, "grad_norm": 2.9525978565216064, "learning_rate": 4.903030303030303e-06, "loss": 0.1742, "step": 810 }, { "epoch": 0.014914785645428254, "grad_norm": 2.3522145748138428, "learning_rate": 4.963636363636364e-06, "loss": 0.1653, "step": 820 }, { "epoch": 0.01509667327525055, "grad_norm": 2.5971524715423584, "learning_rate": 5.024242424242425e-06, "loss": 0.1732, "step": 830 }, { "epoch": 0.015278560905072847, "grad_norm": 2.6136906147003174, "learning_rate": 5.084848484848486e-06, "loss": 0.1709, "step": 840 }, { "epoch": 0.015460448534895142, "grad_norm": 2.99776291847229, "learning_rate": 5.145454545454546e-06, "loss": 0.1768, "step": 850 }, { "epoch": 0.015642336164717437, "grad_norm": 3.0501508712768555, "learning_rate": 5.2060606060606065e-06, "loss": 0.1741, "step": 860 }, { "epoch": 0.015824223794539734, "grad_norm": 2.3400800228118896, "learning_rate": 5.2666666666666665e-06, "loss": 0.1711, "step": 870 }, { "epoch": 0.016006111424362028, "grad_norm": 2.6037988662719727, "learning_rate": 5.327272727272727e-06, "loss": 0.1756, "step": 880 }, { "epoch": 0.016187999054184325, "grad_norm": 3.502967596054077, "learning_rate": 5.387878787878789e-06, "loss": 0.1756, "step": 890 }, { "epoch": 0.016369886684006622, "grad_norm": 3.3316268920898438, "learning_rate": 5.448484848484848e-06, "loss": 0.1835, "step": 900 }, { "epoch": 0.016551774313828916, "grad_norm": 2.3901896476745605, "learning_rate": 5.50909090909091e-06, "loss": 0.1782, "step": 910 }, { "epoch": 0.016733661943651213, "grad_norm": 2.749833106994629, "learning_rate": 5.569696969696971e-06, "loss": 0.1776, "step": 920 }, { "epoch": 0.016915549573473507, "grad_norm": 2.7559974193573, "learning_rate": 5.630303030303031e-06, "loss": 0.181, "step": 930 }, { "epoch": 0.017097437203295804, "grad_norm": 2.5052711963653564, "learning_rate": 5.690909090909091e-06, "loss": 0.1862, "step": 940 }, { "epoch": 0.0172793248331181, "grad_norm": 2.573545217514038, "learning_rate": 5.751515151515152e-06, "loss": 0.1826, "step": 950 }, { "epoch": 0.017461212462940395, "grad_norm": 3.7804794311523438, "learning_rate": 5.812121212121212e-06, "loss": 0.1815, "step": 960 }, { "epoch": 0.01764310009276269, "grad_norm": 3.0479209423065186, "learning_rate": 5.872727272727273e-06, "loss": 0.1862, "step": 970 }, { "epoch": 0.017824987722584985, "grad_norm": 3.2225182056427, "learning_rate": 5.933333333333335e-06, "loss": 0.1884, "step": 980 }, { "epoch": 0.018006875352407282, "grad_norm": 3.020357131958008, "learning_rate": 5.993939393939394e-06, "loss": 0.1929, "step": 990 }, { "epoch": 0.01818876298222958, "grad_norm": 2.400305986404419, "learning_rate": 6.0545454545454555e-06, "loss": 0.1908, "step": 1000 }, { "epoch": 0.018370650612051873, "grad_norm": 2.480109214782715, "learning_rate": 6.115151515151516e-06, "loss": 0.1847, "step": 1010 }, { "epoch": 0.01855253824187417, "grad_norm": 3.367530107498169, "learning_rate": 6.175757575757576e-06, "loss": 0.1933, "step": 1020 }, { "epoch": 0.018734425871696467, "grad_norm": 2.765739917755127, "learning_rate": 6.236363636363637e-06, "loss": 0.1956, "step": 1030 }, { "epoch": 0.01891631350151876, "grad_norm": 2.7861576080322266, "learning_rate": 6.296969696969697e-06, "loss": 0.1943, "step": 1040 }, { "epoch": 0.019098201131341058, "grad_norm": 2.481764554977417, "learning_rate": 6.357575757575758e-06, "loss": 0.1927, "step": 1050 }, { "epoch": 0.019280088761163352, "grad_norm": 2.72752046585083, "learning_rate": 6.418181818181819e-06, "loss": 0.1954, "step": 1060 }, { "epoch": 0.01946197639098565, "grad_norm": 2.6103832721710205, "learning_rate": 6.478787878787879e-06, "loss": 0.2006, "step": 1070 }, { "epoch": 0.019643864020807946, "grad_norm": 2.674779176712036, "learning_rate": 6.5393939393939395e-06, "loss": 0.1979, "step": 1080 }, { "epoch": 0.01982575165063024, "grad_norm": 3.1836483478546143, "learning_rate": 6.600000000000001e-06, "loss": 0.1956, "step": 1090 }, { "epoch": 0.020007639280452537, "grad_norm": 3.2717857360839844, "learning_rate": 6.660606060606061e-06, "loss": 0.195, "step": 1100 }, { "epoch": 0.02018952691027483, "grad_norm": 2.6902787685394287, "learning_rate": 6.721212121212122e-06, "loss": 0.1984, "step": 1110 }, { "epoch": 0.020371414540097128, "grad_norm": 2.996378183364868, "learning_rate": 6.781818181818183e-06, "loss": 0.2037, "step": 1120 }, { "epoch": 0.020553302169919425, "grad_norm": 2.126051664352417, "learning_rate": 6.842424242424243e-06, "loss": 0.2036, "step": 1130 }, { "epoch": 0.02073518979974172, "grad_norm": 2.2756505012512207, "learning_rate": 6.903030303030304e-06, "loss": 0.2005, "step": 1140 }, { "epoch": 0.020917077429564016, "grad_norm": 2.734347105026245, "learning_rate": 6.963636363636364e-06, "loss": 0.1988, "step": 1150 }, { "epoch": 0.021098965059386313, "grad_norm": 2.327151298522949, "learning_rate": 7.024242424242424e-06, "loss": 0.2049, "step": 1160 }, { "epoch": 0.021280852689208606, "grad_norm": 2.478959083557129, "learning_rate": 7.084848484848485e-06, "loss": 0.2036, "step": 1170 }, { "epoch": 0.021462740319030903, "grad_norm": 3.202338695526123, "learning_rate": 7.145454545454547e-06, "loss": 0.2032, "step": 1180 }, { "epoch": 0.021644627948853197, "grad_norm": 2.571309804916382, "learning_rate": 7.206060606060606e-06, "loss": 0.2072, "step": 1190 }, { "epoch": 0.021826515578675494, "grad_norm": 2.5817041397094727, "learning_rate": 7.266666666666668e-06, "loss": 0.2065, "step": 1200 }, { "epoch": 0.02200840320849779, "grad_norm": 3.552062511444092, "learning_rate": 7.3272727272727285e-06, "loss": 0.2093, "step": 1210 }, { "epoch": 0.022190290838320085, "grad_norm": 3.1151270866394043, "learning_rate": 7.3878787878787885e-06, "loss": 0.2046, "step": 1220 }, { "epoch": 0.022372178468142382, "grad_norm": 3.065192699432373, "learning_rate": 7.448484848484849e-06, "loss": 0.2037, "step": 1230 }, { "epoch": 0.02255406609796468, "grad_norm": 2.38714337348938, "learning_rate": 7.509090909090909e-06, "loss": 0.2121, "step": 1240 }, { "epoch": 0.022735953727786973, "grad_norm": 2.8090269565582275, "learning_rate": 7.56969696969697e-06, "loss": 0.2112, "step": 1250 }, { "epoch": 0.02291784135760927, "grad_norm": 2.671116352081299, "learning_rate": 7.630303030303031e-06, "loss": 0.2081, "step": 1260 }, { "epoch": 0.023099728987431564, "grad_norm": 2.224801540374756, "learning_rate": 7.690909090909091e-06, "loss": 0.2145, "step": 1270 }, { "epoch": 0.02328161661725386, "grad_norm": 2.8272323608398438, "learning_rate": 7.751515151515153e-06, "loss": 0.2117, "step": 1280 }, { "epoch": 0.023463504247076158, "grad_norm": 2.3467135429382324, "learning_rate": 7.812121212121213e-06, "loss": 0.2162, "step": 1290 }, { "epoch": 0.02364539187689845, "grad_norm": 2.773315906524658, "learning_rate": 7.872727272727273e-06, "loss": 0.2158, "step": 1300 }, { "epoch": 0.02382727950672075, "grad_norm": 3.2185869216918945, "learning_rate": 7.933333333333334e-06, "loss": 0.2149, "step": 1310 }, { "epoch": 0.024009167136543042, "grad_norm": 4.196074962615967, "learning_rate": 7.993939393939396e-06, "loss": 0.2151, "step": 1320 }, { "epoch": 0.02419105476636534, "grad_norm": 2.5814170837402344, "learning_rate": 8.054545454545454e-06, "loss": 0.2071, "step": 1330 }, { "epoch": 0.024372942396187636, "grad_norm": 2.5502588748931885, "learning_rate": 8.115151515151516e-06, "loss": 0.2139, "step": 1340 }, { "epoch": 0.02455483002600993, "grad_norm": 2.3451342582702637, "learning_rate": 8.175757575757577e-06, "loss": 0.219, "step": 1350 }, { "epoch": 0.024736717655832227, "grad_norm": 2.5672507286071777, "learning_rate": 8.236363636363637e-06, "loss": 0.2159, "step": 1360 }, { "epoch": 0.024918605285654524, "grad_norm": 2.6987812519073486, "learning_rate": 8.296969696969697e-06, "loss": 0.2233, "step": 1370 }, { "epoch": 0.025100492915476818, "grad_norm": 2.6668167114257812, "learning_rate": 8.357575757575759e-06, "loss": 0.2198, "step": 1380 }, { "epoch": 0.025282380545299115, "grad_norm": 2.824317455291748, "learning_rate": 8.418181818181819e-06, "loss": 0.2293, "step": 1390 }, { "epoch": 0.02546426817512141, "grad_norm": 2.7933952808380127, "learning_rate": 8.478787878787879e-06, "loss": 0.2218, "step": 1400 }, { "epoch": 0.025646155804943706, "grad_norm": 2.5317513942718506, "learning_rate": 8.539393939393939e-06, "loss": 0.22, "step": 1410 }, { "epoch": 0.025828043434766003, "grad_norm": 2.351667642593384, "learning_rate": 8.6e-06, "loss": 0.2289, "step": 1420 }, { "epoch": 0.026009931064588297, "grad_norm": 3.0630593299865723, "learning_rate": 8.660606060606062e-06, "loss": 0.2241, "step": 1430 }, { "epoch": 0.026191818694410594, "grad_norm": 2.857048273086548, "learning_rate": 8.72121212121212e-06, "loss": 0.2188, "step": 1440 }, { "epoch": 0.026373706324232887, "grad_norm": 2.5042495727539062, "learning_rate": 8.781818181818182e-06, "loss": 0.2275, "step": 1450 }, { "epoch": 0.026555593954055184, "grad_norm": 2.7413737773895264, "learning_rate": 8.842424242424244e-06, "loss": 0.227, "step": 1460 }, { "epoch": 0.02673748158387748, "grad_norm": 2.6819159984588623, "learning_rate": 8.903030303030304e-06, "loss": 0.2353, "step": 1470 }, { "epoch": 0.026919369213699775, "grad_norm": 2.296743392944336, "learning_rate": 8.963636363636364e-06, "loss": 0.2201, "step": 1480 }, { "epoch": 0.027101256843522072, "grad_norm": 2.3405258655548096, "learning_rate": 9.024242424242426e-06, "loss": 0.2266, "step": 1490 }, { "epoch": 0.02728314447334437, "grad_norm": 2.599013090133667, "learning_rate": 9.084848484848486e-06, "loss": 0.2309, "step": 1500 }, { "epoch": 0.027465032103166663, "grad_norm": 3.263410806655884, "learning_rate": 9.145454545454546e-06, "loss": 0.2299, "step": 1510 }, { "epoch": 0.02764691973298896, "grad_norm": 2.7736713886260986, "learning_rate": 9.206060606060607e-06, "loss": 0.2304, "step": 1520 }, { "epoch": 0.027828807362811254, "grad_norm": 2.7861156463623047, "learning_rate": 9.266666666666667e-06, "loss": 0.231, "step": 1530 }, { "epoch": 0.02801069499263355, "grad_norm": 2.8155977725982666, "learning_rate": 9.327272727272729e-06, "loss": 0.2327, "step": 1540 }, { "epoch": 0.028192582622455848, "grad_norm": 2.1558234691619873, "learning_rate": 9.387878787878789e-06, "loss": 0.2232, "step": 1550 }, { "epoch": 0.02837447025227814, "grad_norm": 2.764052152633667, "learning_rate": 9.448484848484849e-06, "loss": 0.2372, "step": 1560 }, { "epoch": 0.02855635788210044, "grad_norm": 2.5895111560821533, "learning_rate": 9.50909090909091e-06, "loss": 0.2409, "step": 1570 }, { "epoch": 0.028738245511922732, "grad_norm": 2.3981637954711914, "learning_rate": 9.56969696969697e-06, "loss": 0.2311, "step": 1580 }, { "epoch": 0.02892013314174503, "grad_norm": 2.942044973373413, "learning_rate": 9.63030303030303e-06, "loss": 0.2291, "step": 1590 }, { "epoch": 0.029102020771567327, "grad_norm": 2.7201459407806396, "learning_rate": 9.690909090909092e-06, "loss": 0.2297, "step": 1600 }, { "epoch": 0.02928390840138962, "grad_norm": 3.1364593505859375, "learning_rate": 9.751515151515152e-06, "loss": 0.2379, "step": 1610 }, { "epoch": 0.029465796031211917, "grad_norm": 2.3275504112243652, "learning_rate": 9.812121212121212e-06, "loss": 0.2419, "step": 1620 }, { "epoch": 0.029647683661034215, "grad_norm": 2.7291622161865234, "learning_rate": 9.872727272727274e-06, "loss": 0.2444, "step": 1630 }, { "epoch": 0.029829571290856508, "grad_norm": 2.622133493423462, "learning_rate": 9.933333333333334e-06, "loss": 0.2378, "step": 1640 }, { "epoch": 0.030011458920678805, "grad_norm": 2.643294095993042, "learning_rate": 9.993939393939395e-06, "loss": 0.2449, "step": 1650 }, { "epoch": 0.0301933465505011, "grad_norm": 2.512709140777588, "learning_rate": 9.999999437803847e-06, "loss": 0.2407, "step": 1660 }, { "epoch": 0.030375234180323396, "grad_norm": 3.1786715984344482, "learning_rate": 9.99999749440994e-06, "loss": 0.2423, "step": 1670 }, { "epoch": 0.030557121810145693, "grad_norm": 2.9448277950286865, "learning_rate": 9.999994162878262e-06, "loss": 0.2392, "step": 1680 }, { "epoch": 0.030739009439967987, "grad_norm": 2.6393139362335205, "learning_rate": 9.999989443209965e-06, "loss": 0.243, "step": 1690 }, { "epoch": 0.030920897069790284, "grad_norm": 2.4845292568206787, "learning_rate": 9.999983335406693e-06, "loss": 0.244, "step": 1700 }, { "epoch": 0.03110278469961258, "grad_norm": 2.6240835189819336, "learning_rate": 9.999975839470558e-06, "loss": 0.2471, "step": 1710 }, { "epoch": 0.031284672329434875, "grad_norm": 2.3888332843780518, "learning_rate": 9.999966955404166e-06, "loss": 0.2419, "step": 1720 }, { "epoch": 0.03146655995925717, "grad_norm": 2.637495756149292, "learning_rate": 9.999956683210603e-06, "loss": 0.2499, "step": 1730 }, { "epoch": 0.03164844758907947, "grad_norm": 2.4855141639709473, "learning_rate": 9.999945022893427e-06, "loss": 0.2438, "step": 1740 }, { "epoch": 0.03183033521890176, "grad_norm": 2.3640244007110596, "learning_rate": 9.99993197445669e-06, "loss": 0.2472, "step": 1750 }, { "epoch": 0.032012222848724056, "grad_norm": 2.0164473056793213, "learning_rate": 9.999917537904916e-06, "loss": 0.2423, "step": 1760 }, { "epoch": 0.03219411047854635, "grad_norm": 2.6599628925323486, "learning_rate": 9.999901713243116e-06, "loss": 0.2457, "step": 1770 }, { "epoch": 0.03237599810836865, "grad_norm": 2.3170995712280273, "learning_rate": 9.999884500476786e-06, "loss": 0.2491, "step": 1780 }, { "epoch": 0.03255788573819095, "grad_norm": 2.725998640060425, "learning_rate": 9.999865899611894e-06, "loss": 0.2461, "step": 1790 }, { "epoch": 0.032739773368013245, "grad_norm": 2.2788147926330566, "learning_rate": 9.999845910654898e-06, "loss": 0.2502, "step": 1800 }, { "epoch": 0.032921660997835535, "grad_norm": 2.511791944503784, "learning_rate": 9.999824533612735e-06, "loss": 0.2471, "step": 1810 }, { "epoch": 0.03310354862765783, "grad_norm": 2.5130372047424316, "learning_rate": 9.999801768492822e-06, "loss": 0.2464, "step": 1820 }, { "epoch": 0.03328543625748013, "grad_norm": 2.455425500869751, "learning_rate": 9.999777615303062e-06, "loss": 0.2496, "step": 1830 }, { "epoch": 0.033467323887302426, "grad_norm": 2.541111469268799, "learning_rate": 9.999752074051832e-06, "loss": 0.2392, "step": 1840 }, { "epoch": 0.03364921151712472, "grad_norm": 2.7928929328918457, "learning_rate": 9.999725144748e-06, "loss": 0.2487, "step": 1850 }, { "epoch": 0.033831099146947013, "grad_norm": 2.2189643383026123, "learning_rate": 9.99969682740091e-06, "loss": 0.2544, "step": 1860 }, { "epoch": 0.03401298677676931, "grad_norm": 2.579400062561035, "learning_rate": 9.99966712202039e-06, "loss": 0.2515, "step": 1870 }, { "epoch": 0.03419487440659161, "grad_norm": 2.2661643028259277, "learning_rate": 9.999636028616749e-06, "loss": 0.2439, "step": 1880 }, { "epoch": 0.034376762036413905, "grad_norm": 2.5258700847625732, "learning_rate": 9.999603547200773e-06, "loss": 0.2472, "step": 1890 }, { "epoch": 0.0345586496662362, "grad_norm": 2.8430328369140625, "learning_rate": 9.99956967778374e-06, "loss": 0.2583, "step": 1900 }, { "epoch": 0.03474053729605849, "grad_norm": 2.662269353866577, "learning_rate": 9.999534420377401e-06, "loss": 0.2594, "step": 1910 }, { "epoch": 0.03492242492588079, "grad_norm": 2.963423013687134, "learning_rate": 9.999497774993994e-06, "loss": 0.2524, "step": 1920 }, { "epoch": 0.035104312555703086, "grad_norm": 2.175206184387207, "learning_rate": 9.999459741646234e-06, "loss": 0.2497, "step": 1930 }, { "epoch": 0.03528620018552538, "grad_norm": 2.3928956985473633, "learning_rate": 9.999420320347318e-06, "loss": 0.2508, "step": 1940 }, { "epoch": 0.03546808781534768, "grad_norm": 2.3240854740142822, "learning_rate": 9.99937951111093e-06, "loss": 0.2508, "step": 1950 }, { "epoch": 0.03564997544516997, "grad_norm": 2.5260486602783203, "learning_rate": 9.999337313951231e-06, "loss": 0.2503, "step": 1960 }, { "epoch": 0.03583186307499227, "grad_norm": 2.149907112121582, "learning_rate": 9.999293728882864e-06, "loss": 0.2464, "step": 1970 }, { "epoch": 0.036013750704814565, "grad_norm": 2.2896273136138916, "learning_rate": 9.999248755920958e-06, "loss": 0.2544, "step": 1980 }, { "epoch": 0.03619563833463686, "grad_norm": 2.3100383281707764, "learning_rate": 9.999202395081116e-06, "loss": 0.2592, "step": 1990 }, { "epoch": 0.03637752596445916, "grad_norm": 2.4681999683380127, "learning_rate": 9.999154646379427e-06, "loss": 0.246, "step": 2000 }, { "epoch": 0.036559413594281456, "grad_norm": 2.733447551727295, "learning_rate": 9.999105509832466e-06, "loss": 0.2562, "step": 2010 }, { "epoch": 0.036741301224103746, "grad_norm": 2.206465005874634, "learning_rate": 9.999054985457279e-06, "loss": 0.2489, "step": 2020 }, { "epoch": 0.036923188853926044, "grad_norm": 2.4155893325805664, "learning_rate": 9.999003073271405e-06, "loss": 0.2566, "step": 2030 }, { "epoch": 0.03710507648374834, "grad_norm": 2.471073865890503, "learning_rate": 9.998949773292855e-06, "loss": 0.2457, "step": 2040 }, { "epoch": 0.03728696411357064, "grad_norm": 2.521681785583496, "learning_rate": 9.998895085540129e-06, "loss": 0.2477, "step": 2050 }, { "epoch": 0.037468851743392935, "grad_norm": 2.284287929534912, "learning_rate": 9.998839010032204e-06, "loss": 0.2499, "step": 2060 }, { "epoch": 0.037650739373215225, "grad_norm": 2.6220502853393555, "learning_rate": 9.998781546788542e-06, "loss": 0.2583, "step": 2070 }, { "epoch": 0.03783262700303752, "grad_norm": 2.3022820949554443, "learning_rate": 9.998722695829082e-06, "loss": 0.2493, "step": 2080 }, { "epoch": 0.03801451463285982, "grad_norm": 2.1934311389923096, "learning_rate": 9.998662457174248e-06, "loss": 0.2481, "step": 2090 }, { "epoch": 0.038196402262682116, "grad_norm": 2.793304443359375, "learning_rate": 9.998600830844946e-06, "loss": 0.2531, "step": 2100 }, { "epoch": 0.038378289892504414, "grad_norm": 2.7151753902435303, "learning_rate": 9.998537816862564e-06, "loss": 0.2444, "step": 2110 }, { "epoch": 0.038560177522326704, "grad_norm": 2.384119987487793, "learning_rate": 9.998473415248967e-06, "loss": 0.2562, "step": 2120 }, { "epoch": 0.038742065152149, "grad_norm": 2.273545026779175, "learning_rate": 9.998407626026505e-06, "loss": 0.2597, "step": 2130 }, { "epoch": 0.0389239527819713, "grad_norm": 2.639721632003784, "learning_rate": 9.99834044921801e-06, "loss": 0.256, "step": 2140 }, { "epoch": 0.039105840411793595, "grad_norm": 2.518195390701294, "learning_rate": 9.998271884846795e-06, "loss": 0.2477, "step": 2150 }, { "epoch": 0.03928772804161589, "grad_norm": 3.183715343475342, "learning_rate": 9.998201932936653e-06, "loss": 0.2534, "step": 2160 }, { "epoch": 0.03946961567143818, "grad_norm": 2.6936371326446533, "learning_rate": 9.998130593511862e-06, "loss": 0.2531, "step": 2170 }, { "epoch": 0.03965150330126048, "grad_norm": 2.3837544918060303, "learning_rate": 9.998057866597176e-06, "loss": 0.2574, "step": 2180 }, { "epoch": 0.03983339093108278, "grad_norm": 3.021071434020996, "learning_rate": 9.997983752217838e-06, "loss": 0.2458, "step": 2190 }, { "epoch": 0.040015278560905074, "grad_norm": 2.625126600265503, "learning_rate": 9.997908250399561e-06, "loss": 0.2546, "step": 2200 }, { "epoch": 0.04019716619072737, "grad_norm": 2.2094852924346924, "learning_rate": 9.997831361168554e-06, "loss": 0.2447, "step": 2210 }, { "epoch": 0.04037905382054966, "grad_norm": 2.4440455436706543, "learning_rate": 9.9977530845515e-06, "loss": 0.2514, "step": 2220 }, { "epoch": 0.04056094145037196, "grad_norm": 2.135157823562622, "learning_rate": 9.99767342057556e-06, "loss": 0.248, "step": 2230 }, { "epoch": 0.040742829080194255, "grad_norm": 2.543926477432251, "learning_rate": 9.997592369268381e-06, "loss": 0.2526, "step": 2240 }, { "epoch": 0.04092471671001655, "grad_norm": 2.487163543701172, "learning_rate": 9.997509930658091e-06, "loss": 0.2547, "step": 2250 }, { "epoch": 0.04110660433983885, "grad_norm": 2.161539316177368, "learning_rate": 9.9974261047733e-06, "loss": 0.2626, "step": 2260 }, { "epoch": 0.04128849196966115, "grad_norm": 2.2146167755126953, "learning_rate": 9.997340891643097e-06, "loss": 0.2544, "step": 2270 }, { "epoch": 0.04147037959948344, "grad_norm": 2.9319753646850586, "learning_rate": 9.997254291297056e-06, "loss": 0.2606, "step": 2280 }, { "epoch": 0.041652267229305734, "grad_norm": 2.7116336822509766, "learning_rate": 9.997166303765226e-06, "loss": 0.2558, "step": 2290 }, { "epoch": 0.04183415485912803, "grad_norm": 2.5427331924438477, "learning_rate": 9.997076929078146e-06, "loss": 0.2576, "step": 2300 }, { "epoch": 0.04201604248895033, "grad_norm": 4.412798881530762, "learning_rate": 9.996986167266833e-06, "loss": 0.2565, "step": 2310 }, { "epoch": 0.042197930118772625, "grad_norm": 2.6133205890655518, "learning_rate": 9.99689401836278e-06, "loss": 0.2525, "step": 2320 }, { "epoch": 0.042379817748594915, "grad_norm": 2.3946893215179443, "learning_rate": 9.996800482397968e-06, "loss": 0.2494, "step": 2330 }, { "epoch": 0.04256170537841721, "grad_norm": 2.215367078781128, "learning_rate": 9.996705559404857e-06, "loss": 0.2532, "step": 2340 }, { "epoch": 0.04274359300823951, "grad_norm": 3.498199939727783, "learning_rate": 9.996609249416391e-06, "loss": 0.2469, "step": 2350 }, { "epoch": 0.04292548063806181, "grad_norm": 2.108877182006836, "learning_rate": 9.996511552465991e-06, "loss": 0.2521, "step": 2360 }, { "epoch": 0.043107368267884104, "grad_norm": 2.4160470962524414, "learning_rate": 9.99641246858756e-06, "loss": 0.254, "step": 2370 }, { "epoch": 0.043289255897706394, "grad_norm": 2.108677387237549, "learning_rate": 9.996311997815484e-06, "loss": 0.2575, "step": 2380 }, { "epoch": 0.04347114352752869, "grad_norm": 2.300039529800415, "learning_rate": 9.99621014018463e-06, "loss": 0.256, "step": 2390 }, { "epoch": 0.04365303115735099, "grad_norm": 2.3571081161499023, "learning_rate": 9.996106895730349e-06, "loss": 0.2626, "step": 2400 }, { "epoch": 0.043834918787173285, "grad_norm": 2.6703062057495117, "learning_rate": 9.996002264488466e-06, "loss": 0.2614, "step": 2410 }, { "epoch": 0.04401680641699558, "grad_norm": 2.220479965209961, "learning_rate": 9.995896246495294e-06, "loss": 0.2619, "step": 2420 }, { "epoch": 0.04419869404681787, "grad_norm": 2.495345115661621, "learning_rate": 9.995788841787625e-06, "loss": 0.2542, "step": 2430 }, { "epoch": 0.04438058167664017, "grad_norm": 2.1363518238067627, "learning_rate": 9.995680050402731e-06, "loss": 0.2608, "step": 2440 }, { "epoch": 0.04456246930646247, "grad_norm": 2.0159640312194824, "learning_rate": 9.995569872378368e-06, "loss": 0.2576, "step": 2450 }, { "epoch": 0.044744356936284764, "grad_norm": 2.6653597354888916, "learning_rate": 9.995458307752772e-06, "loss": 0.2535, "step": 2460 }, { "epoch": 0.04492624456610706, "grad_norm": 2.3818721771240234, "learning_rate": 9.995345356564655e-06, "loss": 0.2605, "step": 2470 }, { "epoch": 0.04510813219592936, "grad_norm": 2.637899875640869, "learning_rate": 9.99523101885322e-06, "loss": 0.2606, "step": 2480 }, { "epoch": 0.04529001982575165, "grad_norm": 2.55328369140625, "learning_rate": 9.995115294658147e-06, "loss": 0.26, "step": 2490 }, { "epoch": 0.045471907455573946, "grad_norm": 2.1691372394561768, "learning_rate": 9.994998184019591e-06, "loss": 0.2554, "step": 2500 }, { "epoch": 0.04565379508539624, "grad_norm": 2.2454211711883545, "learning_rate": 9.994879686978197e-06, "loss": 0.2467, "step": 2510 }, { "epoch": 0.04583568271521854, "grad_norm": 2.592961549758911, "learning_rate": 9.994759803575087e-06, "loss": 0.2572, "step": 2520 }, { "epoch": 0.04601757034504084, "grad_norm": 2.8489975929260254, "learning_rate": 9.994638533851864e-06, "loss": 0.2535, "step": 2530 }, { "epoch": 0.04619945797486313, "grad_norm": 2.2861077785491943, "learning_rate": 9.994515877850616e-06, "loss": 0.258, "step": 2540 }, { "epoch": 0.046381345604685424, "grad_norm": 2.378505229949951, "learning_rate": 9.994391835613905e-06, "loss": 0.2584, "step": 2550 }, { "epoch": 0.04656323323450772, "grad_norm": 2.0383517742156982, "learning_rate": 9.994266407184778e-06, "loss": 0.2541, "step": 2560 }, { "epoch": 0.04674512086433002, "grad_norm": 2.3268511295318604, "learning_rate": 9.994139592606765e-06, "loss": 0.2591, "step": 2570 }, { "epoch": 0.046927008494152316, "grad_norm": 2.3414666652679443, "learning_rate": 9.994011391923874e-06, "loss": 0.2597, "step": 2580 }, { "epoch": 0.047108896123974606, "grad_norm": 1.949152946472168, "learning_rate": 9.993881805180596e-06, "loss": 0.2616, "step": 2590 }, { "epoch": 0.0472907837537969, "grad_norm": 2.2557356357574463, "learning_rate": 9.993750832421904e-06, "loss": 0.2543, "step": 2600 }, { "epoch": 0.0474726713836192, "grad_norm": 2.161031484603882, "learning_rate": 9.993618473693245e-06, "loss": 0.2604, "step": 2610 }, { "epoch": 0.0476545590134415, "grad_norm": 2.305330991744995, "learning_rate": 9.993484729040553e-06, "loss": 0.2606, "step": 2620 }, { "epoch": 0.047836446643263794, "grad_norm": 2.2807087898254395, "learning_rate": 9.993349598510247e-06, "loss": 0.2643, "step": 2630 }, { "epoch": 0.048018334273086084, "grad_norm": 2.2481181621551514, "learning_rate": 9.993213082149216e-06, "loss": 0.2617, "step": 2640 }, { "epoch": 0.04820022190290838, "grad_norm": 2.30757474899292, "learning_rate": 9.993075180004841e-06, "loss": 0.2611, "step": 2650 }, { "epoch": 0.04838210953273068, "grad_norm": 2.3122236728668213, "learning_rate": 9.992935892124974e-06, "loss": 0.2641, "step": 2660 }, { "epoch": 0.048563997162552976, "grad_norm": 2.1791727542877197, "learning_rate": 9.992795218557958e-06, "loss": 0.2578, "step": 2670 }, { "epoch": 0.04874588479237527, "grad_norm": 2.26409912109375, "learning_rate": 9.992653159352608e-06, "loss": 0.2647, "step": 2680 }, { "epoch": 0.04892777242219756, "grad_norm": 2.238921642303467, "learning_rate": 9.992509714558223e-06, "loss": 0.2607, "step": 2690 }, { "epoch": 0.04910966005201986, "grad_norm": 2.29024600982666, "learning_rate": 9.992364884224584e-06, "loss": 0.2679, "step": 2700 }, { "epoch": 0.04929154768184216, "grad_norm": 2.243100166320801, "learning_rate": 9.992218668401954e-06, "loss": 0.2565, "step": 2710 }, { "epoch": 0.049473435311664454, "grad_norm": 2.0116207599639893, "learning_rate": 9.992071067141074e-06, "loss": 0.2594, "step": 2720 }, { "epoch": 0.04965532294148675, "grad_norm": 2.0418660640716553, "learning_rate": 9.991922080493167e-06, "loss": 0.2543, "step": 2730 }, { "epoch": 0.04983721057130905, "grad_norm": 2.3190555572509766, "learning_rate": 9.991771708509932e-06, "loss": 0.2521, "step": 2740 }, { "epoch": 0.05001909820113134, "grad_norm": 2.2598280906677246, "learning_rate": 9.991619951243561e-06, "loss": 0.2568, "step": 2750 }, { "epoch": 0.050200985830953636, "grad_norm": 2.283742904663086, "learning_rate": 9.991466808746714e-06, "loss": 0.2591, "step": 2760 }, { "epoch": 0.05038287346077593, "grad_norm": 2.1549720764160156, "learning_rate": 9.991312281072537e-06, "loss": 0.2567, "step": 2770 }, { "epoch": 0.05056476109059823, "grad_norm": 2.2427377700805664, "learning_rate": 9.99115636827466e-06, "loss": 0.261, "step": 2780 }, { "epoch": 0.05074664872042053, "grad_norm": 2.1626927852630615, "learning_rate": 9.990999070407186e-06, "loss": 0.2635, "step": 2790 }, { "epoch": 0.05092853635024282, "grad_norm": 2.3570544719696045, "learning_rate": 9.990840387524704e-06, "loss": 0.2631, "step": 2800 }, { "epoch": 0.051110423980065114, "grad_norm": 2.076906204223633, "learning_rate": 9.990680319682284e-06, "loss": 0.2623, "step": 2810 }, { "epoch": 0.05129231160988741, "grad_norm": 2.675663709640503, "learning_rate": 9.990518866935473e-06, "loss": 0.2598, "step": 2820 }, { "epoch": 0.05147419923970971, "grad_norm": 2.432823896408081, "learning_rate": 9.990356029340303e-06, "loss": 0.2636, "step": 2830 }, { "epoch": 0.051656086869532006, "grad_norm": 2.489528179168701, "learning_rate": 9.990191806953281e-06, "loss": 0.2621, "step": 2840 }, { "epoch": 0.051837974499354296, "grad_norm": 2.130462884902954, "learning_rate": 9.9900261998314e-06, "loss": 0.2568, "step": 2850 }, { "epoch": 0.05201986212917659, "grad_norm": 1.9805344343185425, "learning_rate": 9.989859208032134e-06, "loss": 0.2551, "step": 2860 }, { "epoch": 0.05220174975899889, "grad_norm": 2.8781046867370605, "learning_rate": 9.989690831613429e-06, "loss": 0.2676, "step": 2870 }, { "epoch": 0.05238363738882119, "grad_norm": 2.619541883468628, "learning_rate": 9.989521070633722e-06, "loss": 0.2653, "step": 2880 }, { "epoch": 0.052565525018643484, "grad_norm": 3.1417853832244873, "learning_rate": 9.989349925151925e-06, "loss": 0.2623, "step": 2890 }, { "epoch": 0.052747412648465775, "grad_norm": 2.4812614917755127, "learning_rate": 9.98917739522743e-06, "loss": 0.2704, "step": 2900 }, { "epoch": 0.05292930027828807, "grad_norm": 2.2866451740264893, "learning_rate": 9.989003480920112e-06, "loss": 0.2525, "step": 2910 }, { "epoch": 0.05311118790811037, "grad_norm": 2.2794110774993896, "learning_rate": 9.988828182290325e-06, "loss": 0.2501, "step": 2920 }, { "epoch": 0.053293075537932666, "grad_norm": 2.2236995697021484, "learning_rate": 9.988651499398905e-06, "loss": 0.2616, "step": 2930 }, { "epoch": 0.05347496316775496, "grad_norm": 2.155465841293335, "learning_rate": 9.988473432307165e-06, "loss": 0.2636, "step": 2940 }, { "epoch": 0.05365685079757726, "grad_norm": 2.208500623703003, "learning_rate": 9.9882939810769e-06, "loss": 0.2609, "step": 2950 }, { "epoch": 0.05383873842739955, "grad_norm": 2.0961525440216064, "learning_rate": 9.98811314577039e-06, "loss": 0.2587, "step": 2960 }, { "epoch": 0.05402062605722185, "grad_norm": 2.161388635635376, "learning_rate": 9.987930926450387e-06, "loss": 0.2622, "step": 2970 }, { "epoch": 0.054202513687044145, "grad_norm": 2.3338840007781982, "learning_rate": 9.98774732318013e-06, "loss": 0.2657, "step": 2980 }, { "epoch": 0.05438440131686644, "grad_norm": 2.2824997901916504, "learning_rate": 9.987562336023332e-06, "loss": 0.2579, "step": 2990 }, { "epoch": 0.05456628894668874, "grad_norm": 1.9325752258300781, "learning_rate": 9.987375965044193e-06, "loss": 0.2611, "step": 3000 }, { "epoch": 0.05474817657651103, "grad_norm": 2.2010979652404785, "learning_rate": 9.98718821030739e-06, "loss": 0.2536, "step": 3010 }, { "epoch": 0.054930064206333326, "grad_norm": 2.053678512573242, "learning_rate": 9.986999071878081e-06, "loss": 0.2576, "step": 3020 }, { "epoch": 0.05511195183615562, "grad_norm": 2.110184669494629, "learning_rate": 9.986808549821901e-06, "loss": 0.2631, "step": 3030 }, { "epoch": 0.05529383946597792, "grad_norm": 2.4985344409942627, "learning_rate": 9.98661664420497e-06, "loss": 0.2595, "step": 3040 }, { "epoch": 0.05547572709580022, "grad_norm": 2.296078681945801, "learning_rate": 9.986423355093885e-06, "loss": 0.2555, "step": 3050 }, { "epoch": 0.05565761472562251, "grad_norm": 2.44197940826416, "learning_rate": 9.986228682555723e-06, "loss": 0.2576, "step": 3060 }, { "epoch": 0.055839502355444805, "grad_norm": 2.5742697715759277, "learning_rate": 9.986032626658046e-06, "loss": 0.2613, "step": 3070 }, { "epoch": 0.0560213899852671, "grad_norm": 2.1652638912200928, "learning_rate": 9.985835187468889e-06, "loss": 0.2646, "step": 3080 }, { "epoch": 0.0562032776150894, "grad_norm": 2.0770223140716553, "learning_rate": 9.985636365056768e-06, "loss": 0.2757, "step": 3090 }, { "epoch": 0.056385165244911696, "grad_norm": 2.010807514190674, "learning_rate": 9.985436159490688e-06, "loss": 0.263, "step": 3100 }, { "epoch": 0.056567052874733986, "grad_norm": 2.348649740219116, "learning_rate": 9.98523457084012e-06, "loss": 0.2572, "step": 3110 }, { "epoch": 0.05674894050455628, "grad_norm": 2.420482873916626, "learning_rate": 9.985031599175027e-06, "loss": 0.2742, "step": 3120 }, { "epoch": 0.05693082813437858, "grad_norm": 2.2040185928344727, "learning_rate": 9.984827244565844e-06, "loss": 0.2689, "step": 3130 }, { "epoch": 0.05711271576420088, "grad_norm": 2.039229154586792, "learning_rate": 9.984621507083494e-06, "loss": 0.2619, "step": 3140 }, { "epoch": 0.057294603394023175, "grad_norm": 2.170764207839966, "learning_rate": 9.984414386799371e-06, "loss": 0.2599, "step": 3150 }, { "epoch": 0.057476491023845465, "grad_norm": 2.05330228805542, "learning_rate": 9.984205883785353e-06, "loss": 0.2676, "step": 3160 }, { "epoch": 0.05765837865366776, "grad_norm": 2.2553939819335938, "learning_rate": 9.983995998113799e-06, "loss": 0.2613, "step": 3170 }, { "epoch": 0.05784026628349006, "grad_norm": 2.566478967666626, "learning_rate": 9.983784729857545e-06, "loss": 0.2658, "step": 3180 }, { "epoch": 0.058022153913312356, "grad_norm": 2.56320858001709, "learning_rate": 9.983572079089911e-06, "loss": 0.2553, "step": 3190 }, { "epoch": 0.05820404154313465, "grad_norm": 2.784815549850464, "learning_rate": 9.983358045884693e-06, "loss": 0.2681, "step": 3200 }, { "epoch": 0.05838592917295695, "grad_norm": 2.690802574157715, "learning_rate": 9.983142630316167e-06, "loss": 0.2701, "step": 3210 }, { "epoch": 0.05856781680277924, "grad_norm": 2.3702690601348877, "learning_rate": 9.982925832459092e-06, "loss": 0.2643, "step": 3220 }, { "epoch": 0.05874970443260154, "grad_norm": 2.3712196350097656, "learning_rate": 9.9827076523887e-06, "loss": 0.2652, "step": 3230 }, { "epoch": 0.058931592062423835, "grad_norm": 2.2397680282592773, "learning_rate": 9.982488090180713e-06, "loss": 0.2561, "step": 3240 }, { "epoch": 0.05911347969224613, "grad_norm": 2.4473230838775635, "learning_rate": 9.982267145911323e-06, "loss": 0.2675, "step": 3250 }, { "epoch": 0.05929536732206843, "grad_norm": 1.9822324514389038, "learning_rate": 9.982044819657204e-06, "loss": 0.2667, "step": 3260 }, { "epoch": 0.05947725495189072, "grad_norm": 2.336728572845459, "learning_rate": 9.981821111495513e-06, "loss": 0.2552, "step": 3270 }, { "epoch": 0.059659142581713016, "grad_norm": 1.9782543182373047, "learning_rate": 9.981596021503885e-06, "loss": 0.2614, "step": 3280 }, { "epoch": 0.059841030211535313, "grad_norm": 2.45428204536438, "learning_rate": 9.981369549760433e-06, "loss": 0.2701, "step": 3290 }, { "epoch": 0.06002291784135761, "grad_norm": 2.5525901317596436, "learning_rate": 9.981141696343752e-06, "loss": 0.2661, "step": 3300 }, { "epoch": 0.06020480547117991, "grad_norm": 2.098222017288208, "learning_rate": 9.980912461332913e-06, "loss": 0.2604, "step": 3310 }, { "epoch": 0.0603866931010022, "grad_norm": 2.2277517318725586, "learning_rate": 9.980681844807469e-06, "loss": 0.2723, "step": 3320 }, { "epoch": 0.060568580730824495, "grad_norm": 2.202115535736084, "learning_rate": 9.980449846847453e-06, "loss": 0.2621, "step": 3330 }, { "epoch": 0.06075046836064679, "grad_norm": 2.2831459045410156, "learning_rate": 9.980216467533374e-06, "loss": 0.2727, "step": 3340 }, { "epoch": 0.06093235599046909, "grad_norm": 2.2909116744995117, "learning_rate": 9.979981706946226e-06, "loss": 0.27, "step": 3350 }, { "epoch": 0.061114243620291386, "grad_norm": 2.0527989864349365, "learning_rate": 9.979745565167477e-06, "loss": 0.2581, "step": 3360 }, { "epoch": 0.061296131250113677, "grad_norm": 2.2053298950195312, "learning_rate": 9.979508042279077e-06, "loss": 0.2688, "step": 3370 }, { "epoch": 0.061478018879935974, "grad_norm": 2.568100690841675, "learning_rate": 9.979269138363455e-06, "loss": 0.2626, "step": 3380 }, { "epoch": 0.06165990650975827, "grad_norm": 2.5659193992614746, "learning_rate": 9.979028853503517e-06, "loss": 0.2704, "step": 3390 }, { "epoch": 0.06184179413958057, "grad_norm": 2.1873576641082764, "learning_rate": 9.978787187782652e-06, "loss": 0.2611, "step": 3400 }, { "epoch": 0.062023681769402865, "grad_norm": 1.9845006465911865, "learning_rate": 9.978544141284726e-06, "loss": 0.2565, "step": 3410 }, { "epoch": 0.06220556939922516, "grad_norm": 2.306307077407837, "learning_rate": 9.978299714094087e-06, "loss": 0.2667, "step": 3420 }, { "epoch": 0.06238745702904745, "grad_norm": 2.2588934898376465, "learning_rate": 9.978053906295553e-06, "loss": 0.2594, "step": 3430 }, { "epoch": 0.06256934465886975, "grad_norm": 2.154123544692993, "learning_rate": 9.977806717974434e-06, "loss": 0.2634, "step": 3440 }, { "epoch": 0.06275123228869205, "grad_norm": 2.3066601753234863, "learning_rate": 9.977558149216511e-06, "loss": 0.2616, "step": 3450 }, { "epoch": 0.06293311991851434, "grad_norm": 2.328383684158325, "learning_rate": 9.977308200108046e-06, "loss": 0.2632, "step": 3460 }, { "epoch": 0.06311500754833664, "grad_norm": 2.2972147464752197, "learning_rate": 9.977056870735781e-06, "loss": 0.2611, "step": 3470 }, { "epoch": 0.06329689517815894, "grad_norm": 2.208423137664795, "learning_rate": 9.976804161186933e-06, "loss": 0.2625, "step": 3480 }, { "epoch": 0.06347878280798123, "grad_norm": 2.08139705657959, "learning_rate": 9.976550071549204e-06, "loss": 0.2613, "step": 3490 }, { "epoch": 0.06366067043780352, "grad_norm": 2.176004409790039, "learning_rate": 9.97629460191077e-06, "loss": 0.2678, "step": 3500 }, { "epoch": 0.06384255806762582, "grad_norm": 2.135936975479126, "learning_rate": 9.976037752360288e-06, "loss": 0.2647, "step": 3510 }, { "epoch": 0.06402444569744811, "grad_norm": 1.9473774433135986, "learning_rate": 9.975779522986896e-06, "loss": 0.268, "step": 3520 }, { "epoch": 0.06420633332727041, "grad_norm": 1.9846220016479492, "learning_rate": 9.975519913880208e-06, "loss": 0.263, "step": 3530 }, { "epoch": 0.0643882209570927, "grad_norm": 2.143720865249634, "learning_rate": 9.975258925130313e-06, "loss": 0.2622, "step": 3540 }, { "epoch": 0.064570108586915, "grad_norm": 2.051701545715332, "learning_rate": 9.97499655682779e-06, "loss": 0.2599, "step": 3550 }, { "epoch": 0.0647519962167373, "grad_norm": 2.2825229167938232, "learning_rate": 9.974732809063683e-06, "loss": 0.2652, "step": 3560 }, { "epoch": 0.0649338838465596, "grad_norm": 2.355632781982422, "learning_rate": 9.974467681929526e-06, "loss": 0.2629, "step": 3570 }, { "epoch": 0.0651157714763819, "grad_norm": 2.139585018157959, "learning_rate": 9.974201175517327e-06, "loss": 0.2701, "step": 3580 }, { "epoch": 0.06529765910620419, "grad_norm": 3.0646495819091797, "learning_rate": 9.97393328991957e-06, "loss": 0.2656, "step": 3590 }, { "epoch": 0.06547954673602649, "grad_norm": 2.4368131160736084, "learning_rate": 9.973664025229224e-06, "loss": 0.2671, "step": 3600 }, { "epoch": 0.06566143436584877, "grad_norm": 2.4334352016448975, "learning_rate": 9.973393381539732e-06, "loss": 0.265, "step": 3610 }, { "epoch": 0.06584332199567107, "grad_norm": 2.2475426197052, "learning_rate": 9.97312135894502e-06, "loss": 0.2686, "step": 3620 }, { "epoch": 0.06602520962549337, "grad_norm": 2.349309206008911, "learning_rate": 9.97284795753948e-06, "loss": 0.2696, "step": 3630 }, { "epoch": 0.06620709725531566, "grad_norm": 2.078340530395508, "learning_rate": 9.972573177418001e-06, "loss": 0.2614, "step": 3640 }, { "epoch": 0.06638898488513796, "grad_norm": 3.150269031524658, "learning_rate": 9.97229701867594e-06, "loss": 0.2694, "step": 3650 }, { "epoch": 0.06657087251496026, "grad_norm": 2.108267068862915, "learning_rate": 9.972019481409128e-06, "loss": 0.2656, "step": 3660 }, { "epoch": 0.06675276014478256, "grad_norm": 2.1291310787200928, "learning_rate": 9.971740565713884e-06, "loss": 0.2623, "step": 3670 }, { "epoch": 0.06693464777460485, "grad_norm": 2.588273763656616, "learning_rate": 9.971460271687004e-06, "loss": 0.2576, "step": 3680 }, { "epoch": 0.06711653540442715, "grad_norm": 2.0382542610168457, "learning_rate": 9.971178599425752e-06, "loss": 0.2628, "step": 3690 }, { "epoch": 0.06729842303424945, "grad_norm": 2.1772074699401855, "learning_rate": 9.970895549027887e-06, "loss": 0.2622, "step": 3700 }, { "epoch": 0.06748031066407173, "grad_norm": 2.692284107208252, "learning_rate": 9.970611120591631e-06, "loss": 0.2637, "step": 3710 }, { "epoch": 0.06766219829389403, "grad_norm": 2.493314027786255, "learning_rate": 9.970325314215693e-06, "loss": 0.2675, "step": 3720 }, { "epoch": 0.06784408592371632, "grad_norm": 2.23236346244812, "learning_rate": 9.970038129999257e-06, "loss": 0.2669, "step": 3730 }, { "epoch": 0.06802597355353862, "grad_norm": 2.0323848724365234, "learning_rate": 9.969749568041988e-06, "loss": 0.2671, "step": 3740 }, { "epoch": 0.06820786118336092, "grad_norm": 2.6642866134643555, "learning_rate": 9.969459628444025e-06, "loss": 0.2674, "step": 3750 }, { "epoch": 0.06838974881318322, "grad_norm": 2.2174201011657715, "learning_rate": 9.969168311305986e-06, "loss": 0.2724, "step": 3760 }, { "epoch": 0.06857163644300551, "grad_norm": 2.8481833934783936, "learning_rate": 9.96887561672897e-06, "loss": 0.2639, "step": 3770 }, { "epoch": 0.06875352407282781, "grad_norm": 2.0490846633911133, "learning_rate": 9.96858154481455e-06, "loss": 0.2618, "step": 3780 }, { "epoch": 0.0689354117026501, "grad_norm": 2.247692108154297, "learning_rate": 9.968286095664783e-06, "loss": 0.2718, "step": 3790 }, { "epoch": 0.0691172993324724, "grad_norm": 6.194633960723877, "learning_rate": 9.967989269382198e-06, "loss": 0.2636, "step": 3800 }, { "epoch": 0.0692991869622947, "grad_norm": 2.130823850631714, "learning_rate": 9.967691066069803e-06, "loss": 0.271, "step": 3810 }, { "epoch": 0.06948107459211698, "grad_norm": 2.2571115493774414, "learning_rate": 9.967391485831086e-06, "loss": 0.2664, "step": 3820 }, { "epoch": 0.06966296222193928, "grad_norm": 2.280886650085449, "learning_rate": 9.96709052877001e-06, "loss": 0.2631, "step": 3830 }, { "epoch": 0.06984484985176158, "grad_norm": 2.190044641494751, "learning_rate": 9.966788194991021e-06, "loss": 0.2704, "step": 3840 }, { "epoch": 0.07002673748158388, "grad_norm": 2.407681703567505, "learning_rate": 9.966484484599036e-06, "loss": 0.2665, "step": 3850 }, { "epoch": 0.07020862511140617, "grad_norm": 2.115851402282715, "learning_rate": 9.966179397699457e-06, "loss": 0.2677, "step": 3860 }, { "epoch": 0.07039051274122847, "grad_norm": 2.27629017829895, "learning_rate": 9.965872934398155e-06, "loss": 0.2617, "step": 3870 }, { "epoch": 0.07057240037105077, "grad_norm": 2.5646560192108154, "learning_rate": 9.965565094801485e-06, "loss": 0.2689, "step": 3880 }, { "epoch": 0.07075428800087306, "grad_norm": 2.4583852291107178, "learning_rate": 9.96525587901628e-06, "loss": 0.2665, "step": 3890 }, { "epoch": 0.07093617563069536, "grad_norm": 2.181593656539917, "learning_rate": 9.964945287149847e-06, "loss": 0.2698, "step": 3900 }, { "epoch": 0.07111806326051766, "grad_norm": 1.9303480386734009, "learning_rate": 9.964633319309972e-06, "loss": 0.2597, "step": 3910 }, { "epoch": 0.07129995089033994, "grad_norm": 2.1933610439300537, "learning_rate": 9.964319975604918e-06, "loss": 0.2584, "step": 3920 }, { "epoch": 0.07148183852016224, "grad_norm": 2.205679416656494, "learning_rate": 9.964005256143427e-06, "loss": 0.2636, "step": 3930 }, { "epoch": 0.07166372614998454, "grad_norm": 2.2825512886047363, "learning_rate": 9.96368916103472e-06, "loss": 0.2716, "step": 3940 }, { "epoch": 0.07184561377980683, "grad_norm": 2.116403579711914, "learning_rate": 9.963371690388489e-06, "loss": 0.2707, "step": 3950 }, { "epoch": 0.07202750140962913, "grad_norm": 2.308729410171509, "learning_rate": 9.963052844314908e-06, "loss": 0.2727, "step": 3960 }, { "epoch": 0.07220938903945143, "grad_norm": 2.1199777126312256, "learning_rate": 9.962732622924629e-06, "loss": 0.2629, "step": 3970 }, { "epoch": 0.07239127666927372, "grad_norm": 2.4700355529785156, "learning_rate": 9.96241102632878e-06, "loss": 0.2807, "step": 3980 }, { "epoch": 0.07257316429909602, "grad_norm": 2.043219804763794, "learning_rate": 9.962088054638964e-06, "loss": 0.2695, "step": 3990 }, { "epoch": 0.07275505192891832, "grad_norm": 1.9885350465774536, "learning_rate": 9.961763707967267e-06, "loss": 0.2706, "step": 4000 }, { "epoch": 0.07293693955874062, "grad_norm": 2.1748456954956055, "learning_rate": 9.961437986426244e-06, "loss": 0.265, "step": 4010 }, { "epoch": 0.07311882718856291, "grad_norm": 2.2957763671875, "learning_rate": 9.961110890128936e-06, "loss": 0.2698, "step": 4020 }, { "epoch": 0.0733007148183852, "grad_norm": 2.209665536880493, "learning_rate": 9.960782419188854e-06, "loss": 0.263, "step": 4030 }, { "epoch": 0.07348260244820749, "grad_norm": 2.1014902591705322, "learning_rate": 9.960452573719992e-06, "loss": 0.2585, "step": 4040 }, { "epoch": 0.07366449007802979, "grad_norm": 2.279738426208496, "learning_rate": 9.960121353836813e-06, "loss": 0.2585, "step": 4050 }, { "epoch": 0.07384637770785209, "grad_norm": 2.523153781890869, "learning_rate": 9.959788759654266e-06, "loss": 0.2721, "step": 4060 }, { "epoch": 0.07402826533767438, "grad_norm": 2.1873316764831543, "learning_rate": 9.95945479128777e-06, "loss": 0.2656, "step": 4070 }, { "epoch": 0.07421015296749668, "grad_norm": 2.491274118423462, "learning_rate": 9.959119448853224e-06, "loss": 0.2684, "step": 4080 }, { "epoch": 0.07439204059731898, "grad_norm": 2.141347646713257, "learning_rate": 9.958782732467006e-06, "loss": 0.2694, "step": 4090 }, { "epoch": 0.07457392822714128, "grad_norm": 2.464120864868164, "learning_rate": 9.958444642245966e-06, "loss": 0.2637, "step": 4100 }, { "epoch": 0.07475581585696357, "grad_norm": 2.268381118774414, "learning_rate": 9.958105178307431e-06, "loss": 0.2696, "step": 4110 }, { "epoch": 0.07493770348678587, "grad_norm": 3.6763522624969482, "learning_rate": 9.957764340769211e-06, "loss": 0.2636, "step": 4120 }, { "epoch": 0.07511959111660815, "grad_norm": 2.463346481323242, "learning_rate": 9.957422129749587e-06, "loss": 0.2717, "step": 4130 }, { "epoch": 0.07530147874643045, "grad_norm": 2.2569236755371094, "learning_rate": 9.957078545367316e-06, "loss": 0.2704, "step": 4140 }, { "epoch": 0.07548336637625275, "grad_norm": 2.19903564453125, "learning_rate": 9.956733587741637e-06, "loss": 0.2727, "step": 4150 }, { "epoch": 0.07566525400607504, "grad_norm": 2.2199020385742188, "learning_rate": 9.956387256992259e-06, "loss": 0.2674, "step": 4160 }, { "epoch": 0.07584714163589734, "grad_norm": 2.672415256500244, "learning_rate": 9.956039553239372e-06, "loss": 0.2681, "step": 4170 }, { "epoch": 0.07602902926571964, "grad_norm": 2.173099994659424, "learning_rate": 9.955690476603643e-06, "loss": 0.2684, "step": 4180 }, { "epoch": 0.07621091689554194, "grad_norm": 2.1958506107330322, "learning_rate": 9.955340027206213e-06, "loss": 0.2756, "step": 4190 }, { "epoch": 0.07639280452536423, "grad_norm": 2.3837409019470215, "learning_rate": 9.954988205168697e-06, "loss": 0.2684, "step": 4200 }, { "epoch": 0.07657469215518653, "grad_norm": 2.0851056575775146, "learning_rate": 9.954635010613194e-06, "loss": 0.2729, "step": 4210 }, { "epoch": 0.07675657978500883, "grad_norm": 2.69014573097229, "learning_rate": 9.95428044366227e-06, "loss": 0.2779, "step": 4220 }, { "epoch": 0.07693846741483111, "grad_norm": 1.9997787475585938, "learning_rate": 9.953924504438978e-06, "loss": 0.2653, "step": 4230 }, { "epoch": 0.07712035504465341, "grad_norm": 1.992789626121521, "learning_rate": 9.953567193066836e-06, "loss": 0.2718, "step": 4240 }, { "epoch": 0.0773022426744757, "grad_norm": 2.264871120452881, "learning_rate": 9.953208509669845e-06, "loss": 0.2588, "step": 4250 }, { "epoch": 0.077484130304298, "grad_norm": 1.931963562965393, "learning_rate": 9.952848454372482e-06, "loss": 0.2658, "step": 4260 }, { "epoch": 0.0776660179341203, "grad_norm": 2.115729331970215, "learning_rate": 9.952487027299695e-06, "loss": 0.272, "step": 4270 }, { "epoch": 0.0778479055639426, "grad_norm": 2.065002202987671, "learning_rate": 9.952124228576916e-06, "loss": 0.2655, "step": 4280 }, { "epoch": 0.0780297931937649, "grad_norm": 2.169032573699951, "learning_rate": 9.951760058330047e-06, "loss": 0.2771, "step": 4290 }, { "epoch": 0.07821168082358719, "grad_norm": 1.9951878786087036, "learning_rate": 9.951394516685467e-06, "loss": 0.2658, "step": 4300 }, { "epoch": 0.07839356845340949, "grad_norm": 2.3879683017730713, "learning_rate": 9.951027603770032e-06, "loss": 0.275, "step": 4310 }, { "epoch": 0.07857545608323178, "grad_norm": 1.9706722497940063, "learning_rate": 9.950659319711075e-06, "loss": 0.2714, "step": 4320 }, { "epoch": 0.07875734371305408, "grad_norm": 2.251642942428589, "learning_rate": 9.950289664636401e-06, "loss": 0.2589, "step": 4330 }, { "epoch": 0.07893923134287636, "grad_norm": 2.1722686290740967, "learning_rate": 9.949918638674296e-06, "loss": 0.2668, "step": 4340 }, { "epoch": 0.07912111897269866, "grad_norm": 2.0969953536987305, "learning_rate": 9.949546241953516e-06, "loss": 0.269, "step": 4350 }, { "epoch": 0.07930300660252096, "grad_norm": 1.9900615215301514, "learning_rate": 9.949172474603295e-06, "loss": 0.2622, "step": 4360 }, { "epoch": 0.07948489423234326, "grad_norm": 2.39467716217041, "learning_rate": 9.948797336753348e-06, "loss": 0.2654, "step": 4370 }, { "epoch": 0.07966678186216555, "grad_norm": 1.9490798711776733, "learning_rate": 9.948420828533855e-06, "loss": 0.2711, "step": 4380 }, { "epoch": 0.07984866949198785, "grad_norm": 2.0795390605926514, "learning_rate": 9.948042950075483e-06, "loss": 0.2639, "step": 4390 }, { "epoch": 0.08003055712181015, "grad_norm": 2.1249303817749023, "learning_rate": 9.947663701509363e-06, "loss": 0.2705, "step": 4400 }, { "epoch": 0.08021244475163244, "grad_norm": 1.9728280305862427, "learning_rate": 9.947283082967112e-06, "loss": 0.2645, "step": 4410 }, { "epoch": 0.08039433238145474, "grad_norm": 2.341526746749878, "learning_rate": 9.946901094580816e-06, "loss": 0.2744, "step": 4420 }, { "epoch": 0.08057622001127704, "grad_norm": 2.5215320587158203, "learning_rate": 9.946517736483039e-06, "loss": 0.2692, "step": 4430 }, { "epoch": 0.08075810764109932, "grad_norm": 2.1410155296325684, "learning_rate": 9.946133008806818e-06, "loss": 0.2694, "step": 4440 }, { "epoch": 0.08093999527092162, "grad_norm": 1.918274164199829, "learning_rate": 9.945746911685669e-06, "loss": 0.2657, "step": 4450 }, { "epoch": 0.08112188290074392, "grad_norm": 2.132948160171509, "learning_rate": 9.945359445253578e-06, "loss": 0.2667, "step": 4460 }, { "epoch": 0.08130377053056621, "grad_norm": 2.1490988731384277, "learning_rate": 9.944970609645013e-06, "loss": 0.2657, "step": 4470 }, { "epoch": 0.08148565816038851, "grad_norm": 2.2774269580841064, "learning_rate": 9.944580404994912e-06, "loss": 0.2684, "step": 4480 }, { "epoch": 0.08166754579021081, "grad_norm": 1.9507431983947754, "learning_rate": 9.944188831438689e-06, "loss": 0.2728, "step": 4490 }, { "epoch": 0.0818494334200331, "grad_norm": 2.1319475173950195, "learning_rate": 9.943795889112234e-06, "loss": 0.2728, "step": 4500 }, { "epoch": 0.0820313210498554, "grad_norm": 2.056828737258911, "learning_rate": 9.94340157815191e-06, "loss": 0.2765, "step": 4510 }, { "epoch": 0.0822132086796777, "grad_norm": 2.0936570167541504, "learning_rate": 9.943005898694559e-06, "loss": 0.2623, "step": 4520 }, { "epoch": 0.0823950963095, "grad_norm": 2.1678781509399414, "learning_rate": 9.942608850877493e-06, "loss": 0.2688, "step": 4530 }, { "epoch": 0.0825769839393223, "grad_norm": 2.123629331588745, "learning_rate": 9.942210434838506e-06, "loss": 0.2755, "step": 4540 }, { "epoch": 0.08275887156914458, "grad_norm": 2.478497266769409, "learning_rate": 9.941810650715854e-06, "loss": 0.2661, "step": 4550 }, { "epoch": 0.08294075919896687, "grad_norm": 2.288418769836426, "learning_rate": 9.941409498648287e-06, "loss": 0.265, "step": 4560 }, { "epoch": 0.08312264682878917, "grad_norm": 2.096135139465332, "learning_rate": 9.941006978775009e-06, "loss": 0.2682, "step": 4570 }, { "epoch": 0.08330453445861147, "grad_norm": 2.323246955871582, "learning_rate": 9.940603091235712e-06, "loss": 0.2734, "step": 4580 }, { "epoch": 0.08348642208843376, "grad_norm": 2.7104291915893555, "learning_rate": 9.940197836170558e-06, "loss": 0.2641, "step": 4590 }, { "epoch": 0.08366830971825606, "grad_norm": 2.184838056564331, "learning_rate": 9.939791213720188e-06, "loss": 0.2656, "step": 4600 }, { "epoch": 0.08385019734807836, "grad_norm": 2.1040074825286865, "learning_rate": 9.939383224025712e-06, "loss": 0.2654, "step": 4610 }, { "epoch": 0.08403208497790066, "grad_norm": 2.1614139080047607, "learning_rate": 9.938973867228714e-06, "loss": 0.2642, "step": 4620 }, { "epoch": 0.08421397260772295, "grad_norm": 1.9691828489303589, "learning_rate": 9.938563143471258e-06, "loss": 0.2686, "step": 4630 }, { "epoch": 0.08439586023754525, "grad_norm": 2.1312670707702637, "learning_rate": 9.938151052895878e-06, "loss": 0.269, "step": 4640 }, { "epoch": 0.08457774786736753, "grad_norm": 2.153400421142578, "learning_rate": 9.937737595645584e-06, "loss": 0.2701, "step": 4650 }, { "epoch": 0.08475963549718983, "grad_norm": 2.0017051696777344, "learning_rate": 9.937322771863861e-06, "loss": 0.2661, "step": 4660 }, { "epoch": 0.08494152312701213, "grad_norm": 1.9766541719436646, "learning_rate": 9.936906581694667e-06, "loss": 0.269, "step": 4670 }, { "epoch": 0.08512341075683443, "grad_norm": 2.0849990844726562, "learning_rate": 9.936489025282433e-06, "loss": 0.2736, "step": 4680 }, { "epoch": 0.08530529838665672, "grad_norm": 2.085813283920288, "learning_rate": 9.936070102772067e-06, "loss": 0.2798, "step": 4690 }, { "epoch": 0.08548718601647902, "grad_norm": 2.10787034034729, "learning_rate": 9.935649814308947e-06, "loss": 0.2611, "step": 4700 }, { "epoch": 0.08566907364630132, "grad_norm": 2.1545920372009277, "learning_rate": 9.935228160038933e-06, "loss": 0.271, "step": 4710 }, { "epoch": 0.08585096127612361, "grad_norm": 2.2246487140655518, "learning_rate": 9.934805140108347e-06, "loss": 0.271, "step": 4720 }, { "epoch": 0.08603284890594591, "grad_norm": 2.302215099334717, "learning_rate": 9.934380754663997e-06, "loss": 0.2671, "step": 4730 }, { "epoch": 0.08621473653576821, "grad_norm": 2.1134133338928223, "learning_rate": 9.933955003853157e-06, "loss": 0.2734, "step": 4740 }, { "epoch": 0.0863966241655905, "grad_norm": 2.2104434967041016, "learning_rate": 9.933527887823578e-06, "loss": 0.273, "step": 4750 }, { "epoch": 0.08657851179541279, "grad_norm": 2.363135814666748, "learning_rate": 9.933099406723483e-06, "loss": 0.2765, "step": 4760 }, { "epoch": 0.08676039942523509, "grad_norm": 1.8592411279678345, "learning_rate": 9.932669560701571e-06, "loss": 0.2678, "step": 4770 }, { "epoch": 0.08694228705505738, "grad_norm": 2.1407530307769775, "learning_rate": 9.932238349907014e-06, "loss": 0.2686, "step": 4780 }, { "epoch": 0.08712417468487968, "grad_norm": 2.040710926055908, "learning_rate": 9.931805774489457e-06, "loss": 0.267, "step": 4790 }, { "epoch": 0.08730606231470198, "grad_norm": 2.325152635574341, "learning_rate": 9.931371834599014e-06, "loss": 0.2703, "step": 4800 }, { "epoch": 0.08748794994452427, "grad_norm": 2.000645875930786, "learning_rate": 9.930936530386282e-06, "loss": 0.2733, "step": 4810 }, { "epoch": 0.08766983757434657, "grad_norm": 2.205035924911499, "learning_rate": 9.930499862002328e-06, "loss": 0.2643, "step": 4820 }, { "epoch": 0.08785172520416887, "grad_norm": 2.3274099826812744, "learning_rate": 9.930061829598687e-06, "loss": 0.27, "step": 4830 }, { "epoch": 0.08803361283399116, "grad_norm": 1.9341745376586914, "learning_rate": 9.929622433327372e-06, "loss": 0.2651, "step": 4840 }, { "epoch": 0.08821550046381346, "grad_norm": 2.3695449829101562, "learning_rate": 9.929181673340872e-06, "loss": 0.2749, "step": 4850 }, { "epoch": 0.08839738809363575, "grad_norm": 2.265350580215454, "learning_rate": 9.928739549792142e-06, "loss": 0.2714, "step": 4860 }, { "epoch": 0.08857927572345804, "grad_norm": 2.0112428665161133, "learning_rate": 9.928296062834619e-06, "loss": 0.268, "step": 4870 }, { "epoch": 0.08876116335328034, "grad_norm": 1.8931108713150024, "learning_rate": 9.927851212622202e-06, "loss": 0.2701, "step": 4880 }, { "epoch": 0.08894305098310264, "grad_norm": 1.9434469938278198, "learning_rate": 9.927404999309273e-06, "loss": 0.2641, "step": 4890 }, { "epoch": 0.08912493861292493, "grad_norm": 2.0935988426208496, "learning_rate": 9.926957423050684e-06, "loss": 0.2736, "step": 4900 }, { "epoch": 0.08930682624274723, "grad_norm": 2.2627434730529785, "learning_rate": 9.92650848400176e-06, "loss": 0.263, "step": 4910 }, { "epoch": 0.08948871387256953, "grad_norm": 2.0525052547454834, "learning_rate": 9.926058182318295e-06, "loss": 0.2624, "step": 4920 }, { "epoch": 0.08967060150239183, "grad_norm": 2.0932984352111816, "learning_rate": 9.925606518156564e-06, "loss": 0.2718, "step": 4930 }, { "epoch": 0.08985248913221412, "grad_norm": 2.228367328643799, "learning_rate": 9.925153491673306e-06, "loss": 0.2555, "step": 4940 }, { "epoch": 0.09003437676203642, "grad_norm": 2.117418050765991, "learning_rate": 9.92469910302574e-06, "loss": 0.2696, "step": 4950 }, { "epoch": 0.09021626439185872, "grad_norm": 2.4280197620391846, "learning_rate": 9.92424335237155e-06, "loss": 0.2732, "step": 4960 }, { "epoch": 0.090398152021681, "grad_norm": 1.9354093074798584, "learning_rate": 9.923786239868904e-06, "loss": 0.2782, "step": 4970 }, { "epoch": 0.0905800396515033, "grad_norm": 2.035464286804199, "learning_rate": 9.92332776567643e-06, "loss": 0.2743, "step": 4980 }, { "epoch": 0.0907619272813256, "grad_norm": 2.1137990951538086, "learning_rate": 9.922867929953237e-06, "loss": 0.2682, "step": 4990 }, { "epoch": 0.09094381491114789, "grad_norm": 2.312497138977051, "learning_rate": 9.922406732858903e-06, "loss": 0.2706, "step": 5000 }, { "epoch": 0.09112570254097019, "grad_norm": 2.166020154953003, "learning_rate": 9.921944174553481e-06, "loss": 0.2705, "step": 5010 }, { "epoch": 0.09130759017079249, "grad_norm": 1.987425446510315, "learning_rate": 9.921480255197495e-06, "loss": 0.2717, "step": 5020 }, { "epoch": 0.09148947780061478, "grad_norm": 2.1121134757995605, "learning_rate": 9.921014974951938e-06, "loss": 0.2692, "step": 5030 }, { "epoch": 0.09167136543043708, "grad_norm": 2.1555545330047607, "learning_rate": 9.92054833397828e-06, "loss": 0.2782, "step": 5040 }, { "epoch": 0.09185325306025938, "grad_norm": 2.023618698120117, "learning_rate": 9.920080332438464e-06, "loss": 0.2685, "step": 5050 }, { "epoch": 0.09203514069008167, "grad_norm": 2.0809028148651123, "learning_rate": 9.919610970494899e-06, "loss": 0.27, "step": 5060 }, { "epoch": 0.09221702831990396, "grad_norm": 1.9992759227752686, "learning_rate": 9.91914024831047e-06, "loss": 0.2619, "step": 5070 }, { "epoch": 0.09239891594972625, "grad_norm": 2.6855127811431885, "learning_rate": 9.918668166048537e-06, "loss": 0.2625, "step": 5080 }, { "epoch": 0.09258080357954855, "grad_norm": 2.1249120235443115, "learning_rate": 9.918194723872927e-06, "loss": 0.2655, "step": 5090 }, { "epoch": 0.09276269120937085, "grad_norm": 2.1699185371398926, "learning_rate": 9.917719921947941e-06, "loss": 0.2704, "step": 5100 }, { "epoch": 0.09294457883919315, "grad_norm": 1.8193974494934082, "learning_rate": 9.917243760438351e-06, "loss": 0.2608, "step": 5110 }, { "epoch": 0.09312646646901544, "grad_norm": 1.9998726844787598, "learning_rate": 9.916766239509404e-06, "loss": 0.2732, "step": 5120 }, { "epoch": 0.09330835409883774, "grad_norm": 2.080212354660034, "learning_rate": 9.916287359326816e-06, "loss": 0.2745, "step": 5130 }, { "epoch": 0.09349024172866004, "grad_norm": 2.1545395851135254, "learning_rate": 9.91580712005677e-06, "loss": 0.2712, "step": 5140 }, { "epoch": 0.09367212935848233, "grad_norm": 2.2914295196533203, "learning_rate": 9.915325521865931e-06, "loss": 0.2602, "step": 5150 }, { "epoch": 0.09385401698830463, "grad_norm": 2.3600196838378906, "learning_rate": 9.91484256492143e-06, "loss": 0.2652, "step": 5160 }, { "epoch": 0.09403590461812691, "grad_norm": 1.9958781003952026, "learning_rate": 9.914358249390867e-06, "loss": 0.2686, "step": 5170 }, { "epoch": 0.09421779224794921, "grad_norm": 2.5851707458496094, "learning_rate": 9.913872575442317e-06, "loss": 0.2662, "step": 5180 }, { "epoch": 0.09439967987777151, "grad_norm": 2.24420428276062, "learning_rate": 9.913385543244328e-06, "loss": 0.2732, "step": 5190 }, { "epoch": 0.0945815675075938, "grad_norm": 2.3371188640594482, "learning_rate": 9.912897152965915e-06, "loss": 0.2723, "step": 5200 }, { "epoch": 0.0947634551374161, "grad_norm": 2.3100273609161377, "learning_rate": 9.912407404776567e-06, "loss": 0.2663, "step": 5210 }, { "epoch": 0.0949453427672384, "grad_norm": 2.1801905632019043, "learning_rate": 9.911916298846242e-06, "loss": 0.2735, "step": 5220 }, { "epoch": 0.0951272303970607, "grad_norm": 1.9502242803573608, "learning_rate": 9.911423835345372e-06, "loss": 0.2675, "step": 5230 }, { "epoch": 0.095309118026883, "grad_norm": 2.2180044651031494, "learning_rate": 9.910930014444863e-06, "loss": 0.272, "step": 5240 }, { "epoch": 0.09549100565670529, "grad_norm": 2.2414679527282715, "learning_rate": 9.91043483631608e-06, "loss": 0.2665, "step": 5250 }, { "epoch": 0.09567289328652759, "grad_norm": 1.9704787731170654, "learning_rate": 9.909938301130872e-06, "loss": 0.2633, "step": 5260 }, { "epoch": 0.09585478091634989, "grad_norm": 2.2666220664978027, "learning_rate": 9.909440409061555e-06, "loss": 0.2697, "step": 5270 }, { "epoch": 0.09603666854617217, "grad_norm": 2.1401290893554688, "learning_rate": 9.908941160280909e-06, "loss": 0.2705, "step": 5280 }, { "epoch": 0.09621855617599447, "grad_norm": 2.129610061645508, "learning_rate": 9.908440554962198e-06, "loss": 0.2764, "step": 5290 }, { "epoch": 0.09640044380581676, "grad_norm": 2.0272693634033203, "learning_rate": 9.907938593279145e-06, "loss": 0.2678, "step": 5300 }, { "epoch": 0.09658233143563906, "grad_norm": 1.9134289026260376, "learning_rate": 9.907435275405948e-06, "loss": 0.2693, "step": 5310 }, { "epoch": 0.09676421906546136, "grad_norm": 1.9105654954910278, "learning_rate": 9.906930601517277e-06, "loss": 0.2691, "step": 5320 }, { "epoch": 0.09694610669528365, "grad_norm": 2.0640077590942383, "learning_rate": 9.906424571788271e-06, "loss": 0.2676, "step": 5330 }, { "epoch": 0.09712799432510595, "grad_norm": 2.021583318710327, "learning_rate": 9.90591718639454e-06, "loss": 0.2665, "step": 5340 }, { "epoch": 0.09730988195492825, "grad_norm": 2.141448736190796, "learning_rate": 9.905408445512163e-06, "loss": 0.2654, "step": 5350 }, { "epoch": 0.09749176958475055, "grad_norm": 2.3160240650177, "learning_rate": 9.904898349317694e-06, "loss": 0.2683, "step": 5360 }, { "epoch": 0.09767365721457284, "grad_norm": 1.9734972715377808, "learning_rate": 9.90438689798815e-06, "loss": 0.2668, "step": 5370 }, { "epoch": 0.09785554484439513, "grad_norm": 2.1574008464813232, "learning_rate": 9.903874091701024e-06, "loss": 0.2662, "step": 5380 }, { "epoch": 0.09803743247421742, "grad_norm": 2.0687415599823, "learning_rate": 9.90335993063428e-06, "loss": 0.2631, "step": 5390 }, { "epoch": 0.09821932010403972, "grad_norm": 2.3108668327331543, "learning_rate": 9.902844414966347e-06, "loss": 0.2749, "step": 5400 }, { "epoch": 0.09840120773386202, "grad_norm": 1.9238219261169434, "learning_rate": 9.902327544876124e-06, "loss": 0.2601, "step": 5410 }, { "epoch": 0.09858309536368431, "grad_norm": 1.923651099205017, "learning_rate": 9.90180932054299e-06, "loss": 0.2663, "step": 5420 }, { "epoch": 0.09876498299350661, "grad_norm": 2.4062352180480957, "learning_rate": 9.901289742146783e-06, "loss": 0.2724, "step": 5430 }, { "epoch": 0.09894687062332891, "grad_norm": 2.03061842918396, "learning_rate": 9.900768809867812e-06, "loss": 0.2698, "step": 5440 }, { "epoch": 0.0991287582531512, "grad_norm": 2.1078121662139893, "learning_rate": 9.900246523886862e-06, "loss": 0.2708, "step": 5450 }, { "epoch": 0.0993106458829735, "grad_norm": 1.8833906650543213, "learning_rate": 9.899722884385186e-06, "loss": 0.2641, "step": 5460 }, { "epoch": 0.0994925335127958, "grad_norm": 2.0742835998535156, "learning_rate": 9.899197891544502e-06, "loss": 0.2633, "step": 5470 }, { "epoch": 0.0996744211426181, "grad_norm": 1.9519597291946411, "learning_rate": 9.898671545546999e-06, "loss": 0.2721, "step": 5480 }, { "epoch": 0.09985630877244038, "grad_norm": 2.170928955078125, "learning_rate": 9.898143846575341e-06, "loss": 0.2746, "step": 5490 }, { "epoch": 0.10003819640226268, "grad_norm": 2.2291157245635986, "learning_rate": 9.897614794812656e-06, "loss": 0.273, "step": 5500 }, { "epoch": 0.10022008403208497, "grad_norm": 2.3735692501068115, "learning_rate": 9.897084390442544e-06, "loss": 0.2662, "step": 5510 }, { "epoch": 0.10040197166190727, "grad_norm": 2.2322700023651123, "learning_rate": 9.896552633649074e-06, "loss": 0.2764, "step": 5520 }, { "epoch": 0.10058385929172957, "grad_norm": 2.1907408237457275, "learning_rate": 9.896019524616783e-06, "loss": 0.2678, "step": 5530 }, { "epoch": 0.10076574692155187, "grad_norm": 2.1519410610198975, "learning_rate": 9.89548506353068e-06, "loss": 0.2666, "step": 5540 }, { "epoch": 0.10094763455137416, "grad_norm": 2.369323968887329, "learning_rate": 9.89494925057624e-06, "loss": 0.2712, "step": 5550 }, { "epoch": 0.10112952218119646, "grad_norm": 2.0390408039093018, "learning_rate": 9.894412085939408e-06, "loss": 0.2776, "step": 5560 }, { "epoch": 0.10131140981101876, "grad_norm": 1.8854340314865112, "learning_rate": 9.8938735698066e-06, "loss": 0.272, "step": 5570 }, { "epoch": 0.10149329744084105, "grad_norm": 2.137420892715454, "learning_rate": 9.8933337023647e-06, "loss": 0.2719, "step": 5580 }, { "epoch": 0.10167518507066334, "grad_norm": 2.2328431606292725, "learning_rate": 9.89279248380106e-06, "loss": 0.262, "step": 5590 }, { "epoch": 0.10185707270048563, "grad_norm": 2.3082571029663086, "learning_rate": 9.892249914303502e-06, "loss": 0.2671, "step": 5600 }, { "epoch": 0.10203896033030793, "grad_norm": 2.0838217735290527, "learning_rate": 9.891705994060317e-06, "loss": 0.274, "step": 5610 }, { "epoch": 0.10222084796013023, "grad_norm": 2.451049327850342, "learning_rate": 9.891160723260263e-06, "loss": 0.269, "step": 5620 }, { "epoch": 0.10240273558995253, "grad_norm": 1.7936525344848633, "learning_rate": 9.890614102092567e-06, "loss": 0.2748, "step": 5630 }, { "epoch": 0.10258462321977482, "grad_norm": 2.3081045150756836, "learning_rate": 9.890066130746929e-06, "loss": 0.2702, "step": 5640 }, { "epoch": 0.10276651084959712, "grad_norm": 2.374035120010376, "learning_rate": 9.889516809413512e-06, "loss": 0.2712, "step": 5650 }, { "epoch": 0.10294839847941942, "grad_norm": 2.1627538204193115, "learning_rate": 9.888966138282947e-06, "loss": 0.2657, "step": 5660 }, { "epoch": 0.10313028610924171, "grad_norm": 1.7945181131362915, "learning_rate": 9.88841411754634e-06, "loss": 0.2708, "step": 5670 }, { "epoch": 0.10331217373906401, "grad_norm": 2.0111374855041504, "learning_rate": 9.887860747395258e-06, "loss": 0.2744, "step": 5680 }, { "epoch": 0.10349406136888631, "grad_norm": 2.0123045444488525, "learning_rate": 9.887306028021743e-06, "loss": 0.2716, "step": 5690 }, { "epoch": 0.10367594899870859, "grad_norm": 1.8349015712738037, "learning_rate": 9.8867499596183e-06, "loss": 0.2702, "step": 5700 }, { "epoch": 0.10385783662853089, "grad_norm": 2.054389715194702, "learning_rate": 9.886192542377905e-06, "loss": 0.2666, "step": 5710 }, { "epoch": 0.10403972425835319, "grad_norm": 2.1223230361938477, "learning_rate": 9.885633776494e-06, "loss": 0.2722, "step": 5720 }, { "epoch": 0.10422161188817548, "grad_norm": 2.487995147705078, "learning_rate": 9.885073662160494e-06, "loss": 0.266, "step": 5730 }, { "epoch": 0.10440349951799778, "grad_norm": 2.1498637199401855, "learning_rate": 9.88451219957177e-06, "loss": 0.2771, "step": 5740 }, { "epoch": 0.10458538714782008, "grad_norm": 2.9072277545928955, "learning_rate": 9.883949388922673e-06, "loss": 0.2752, "step": 5750 }, { "epoch": 0.10476727477764237, "grad_norm": 1.7100896835327148, "learning_rate": 9.88338523040852e-06, "loss": 0.2589, "step": 5760 }, { "epoch": 0.10494916240746467, "grad_norm": 2.2005255222320557, "learning_rate": 9.882819724225091e-06, "loss": 0.2691, "step": 5770 }, { "epoch": 0.10513105003728697, "grad_norm": 2.1482181549072266, "learning_rate": 9.882252870568638e-06, "loss": 0.2702, "step": 5780 }, { "epoch": 0.10531293766710927, "grad_norm": 2.0267996788024902, "learning_rate": 9.881684669635876e-06, "loss": 0.272, "step": 5790 }, { "epoch": 0.10549482529693155, "grad_norm": 2.3969717025756836, "learning_rate": 9.881115121623992e-06, "loss": 0.2713, "step": 5800 }, { "epoch": 0.10567671292675385, "grad_norm": 2.0062918663024902, "learning_rate": 9.88054422673064e-06, "loss": 0.2657, "step": 5810 }, { "epoch": 0.10585860055657614, "grad_norm": 1.97930109500885, "learning_rate": 9.87997198515394e-06, "loss": 0.2777, "step": 5820 }, { "epoch": 0.10604048818639844, "grad_norm": 1.86002779006958, "learning_rate": 9.879398397092481e-06, "loss": 0.2751, "step": 5830 }, { "epoch": 0.10622237581622074, "grad_norm": 1.9831477403640747, "learning_rate": 9.878823462745314e-06, "loss": 0.266, "step": 5840 }, { "epoch": 0.10640426344604303, "grad_norm": 2.0226547718048096, "learning_rate": 9.878247182311965e-06, "loss": 0.2688, "step": 5850 }, { "epoch": 0.10658615107586533, "grad_norm": 2.120945930480957, "learning_rate": 9.87766955599242e-06, "loss": 0.2727, "step": 5860 }, { "epoch": 0.10676803870568763, "grad_norm": 1.7309359312057495, "learning_rate": 9.877090583987138e-06, "loss": 0.263, "step": 5870 }, { "epoch": 0.10694992633550993, "grad_norm": 1.6607890129089355, "learning_rate": 9.87651026649704e-06, "loss": 0.2649, "step": 5880 }, { "epoch": 0.10713181396533222, "grad_norm": 2.4255363941192627, "learning_rate": 9.875928603723517e-06, "loss": 0.2709, "step": 5890 }, { "epoch": 0.10731370159515452, "grad_norm": 2.0167882442474365, "learning_rate": 9.87534559586843e-06, "loss": 0.2739, "step": 5900 }, { "epoch": 0.1074955892249768, "grad_norm": 2.08882474899292, "learning_rate": 9.874761243134098e-06, "loss": 0.2695, "step": 5910 }, { "epoch": 0.1076774768547991, "grad_norm": 2.124006986618042, "learning_rate": 9.874175545723314e-06, "loss": 0.2771, "step": 5920 }, { "epoch": 0.1078593644846214, "grad_norm": 1.915802001953125, "learning_rate": 9.873588503839333e-06, "loss": 0.269, "step": 5930 }, { "epoch": 0.1080412521144437, "grad_norm": 1.8682143688201904, "learning_rate": 9.873000117685882e-06, "loss": 0.2616, "step": 5940 }, { "epoch": 0.10822313974426599, "grad_norm": 2.1641645431518555, "learning_rate": 9.872410387467149e-06, "loss": 0.2657, "step": 5950 }, { "epoch": 0.10840502737408829, "grad_norm": 2.021007776260376, "learning_rate": 9.871819313387792e-06, "loss": 0.2641, "step": 5960 }, { "epoch": 0.10858691500391059, "grad_norm": 2.1120288372039795, "learning_rate": 9.871226895652934e-06, "loss": 0.2722, "step": 5970 }, { "epoch": 0.10876880263373288, "grad_norm": 1.91517174243927, "learning_rate": 9.870633134468165e-06, "loss": 0.2638, "step": 5980 }, { "epoch": 0.10895069026355518, "grad_norm": 1.918696641921997, "learning_rate": 9.87003803003954e-06, "loss": 0.2696, "step": 5990 }, { "epoch": 0.10913257789337748, "grad_norm": 1.7498571872711182, "learning_rate": 9.86944158257358e-06, "loss": 0.2741, "step": 6000 }, { "epoch": 0.10931446552319976, "grad_norm": 2.5724616050720215, "learning_rate": 9.868843792277274e-06, "loss": 0.2732, "step": 6010 }, { "epoch": 0.10949635315302206, "grad_norm": 2.5691139698028564, "learning_rate": 9.868244659358077e-06, "loss": 0.2718, "step": 6020 }, { "epoch": 0.10967824078284436, "grad_norm": 2.2947797775268555, "learning_rate": 9.867644184023907e-06, "loss": 0.2732, "step": 6030 }, { "epoch": 0.10986012841266665, "grad_norm": 2.0955538749694824, "learning_rate": 9.86704236648315e-06, "loss": 0.2773, "step": 6040 }, { "epoch": 0.11004201604248895, "grad_norm": 2.054701089859009, "learning_rate": 9.86643920694466e-06, "loss": 0.2714, "step": 6050 }, { "epoch": 0.11022390367231125, "grad_norm": 1.9576607942581177, "learning_rate": 9.86583470561775e-06, "loss": 0.2662, "step": 6060 }, { "epoch": 0.11040579130213354, "grad_norm": 2.0634562969207764, "learning_rate": 9.865228862712206e-06, "loss": 0.2728, "step": 6070 }, { "epoch": 0.11058767893195584, "grad_norm": 2.1291534900665283, "learning_rate": 9.864621678438276e-06, "loss": 0.2753, "step": 6080 }, { "epoch": 0.11076956656177814, "grad_norm": 1.9722636938095093, "learning_rate": 9.864013153006676e-06, "loss": 0.2652, "step": 6090 }, { "epoch": 0.11095145419160043, "grad_norm": 1.8286100625991821, "learning_rate": 9.86340328662858e-06, "loss": 0.2713, "step": 6100 }, { "epoch": 0.11113334182142272, "grad_norm": 2.0306477546691895, "learning_rate": 9.862792079515636e-06, "loss": 0.269, "step": 6110 }, { "epoch": 0.11131522945124502, "grad_norm": 1.7443304061889648, "learning_rate": 9.862179531879954e-06, "loss": 0.2769, "step": 6120 }, { "epoch": 0.11149711708106731, "grad_norm": 2.2694969177246094, "learning_rate": 9.861565643934113e-06, "loss": 0.2665, "step": 6130 }, { "epoch": 0.11167900471088961, "grad_norm": 2.1398818492889404, "learning_rate": 9.860950415891145e-06, "loss": 0.2721, "step": 6140 }, { "epoch": 0.1118608923407119, "grad_norm": 1.8587054014205933, "learning_rate": 9.860333847964563e-06, "loss": 0.2563, "step": 6150 }, { "epoch": 0.1120427799705342, "grad_norm": 1.9754767417907715, "learning_rate": 9.859715940368334e-06, "loss": 0.2747, "step": 6160 }, { "epoch": 0.1122246676003565, "grad_norm": 1.7712266445159912, "learning_rate": 9.859096693316892e-06, "loss": 0.2694, "step": 6170 }, { "epoch": 0.1124065552301788, "grad_norm": 2.1411361694335938, "learning_rate": 9.85847610702514e-06, "loss": 0.2795, "step": 6180 }, { "epoch": 0.1125884428600011, "grad_norm": 2.127176284790039, "learning_rate": 9.857854181708442e-06, "loss": 0.2655, "step": 6190 }, { "epoch": 0.11277033048982339, "grad_norm": 1.899660587310791, "learning_rate": 9.857230917582628e-06, "loss": 0.2744, "step": 6200 }, { "epoch": 0.11295221811964569, "grad_norm": 1.8883416652679443, "learning_rate": 9.856606314863993e-06, "loss": 0.2735, "step": 6210 }, { "epoch": 0.11313410574946797, "grad_norm": 2.224287986755371, "learning_rate": 9.855980373769294e-06, "loss": 0.272, "step": 6220 }, { "epoch": 0.11331599337929027, "grad_norm": 2.0889804363250732, "learning_rate": 9.855353094515755e-06, "loss": 0.2783, "step": 6230 }, { "epoch": 0.11349788100911257, "grad_norm": 1.8856970071792603, "learning_rate": 9.854724477321063e-06, "loss": 0.2818, "step": 6240 }, { "epoch": 0.11367976863893486, "grad_norm": 2.367220401763916, "learning_rate": 9.854094522403372e-06, "loss": 0.2797, "step": 6250 }, { "epoch": 0.11386165626875716, "grad_norm": 2.131589412689209, "learning_rate": 9.853463229981295e-06, "loss": 0.2675, "step": 6260 }, { "epoch": 0.11404354389857946, "grad_norm": 1.9560551643371582, "learning_rate": 9.852830600273913e-06, "loss": 0.2597, "step": 6270 }, { "epoch": 0.11422543152840176, "grad_norm": 2.266486644744873, "learning_rate": 9.852196633500774e-06, "loss": 0.2753, "step": 6280 }, { "epoch": 0.11440731915822405, "grad_norm": 1.9914827346801758, "learning_rate": 9.851561329881883e-06, "loss": 0.2769, "step": 6290 }, { "epoch": 0.11458920678804635, "grad_norm": 1.9838916063308716, "learning_rate": 9.850924689637713e-06, "loss": 0.2698, "step": 6300 }, { "epoch": 0.11477109441786865, "grad_norm": 2.0777828693389893, "learning_rate": 9.850286712989202e-06, "loss": 0.2721, "step": 6310 }, { "epoch": 0.11495298204769093, "grad_norm": 1.7823212146759033, "learning_rate": 9.84964740015775e-06, "loss": 0.2662, "step": 6320 }, { "epoch": 0.11513486967751323, "grad_norm": 1.9486894607543945, "learning_rate": 9.849006751365216e-06, "loss": 0.275, "step": 6330 }, { "epoch": 0.11531675730733552, "grad_norm": 2.248166561126709, "learning_rate": 9.848364766833932e-06, "loss": 0.2788, "step": 6340 }, { "epoch": 0.11549864493715782, "grad_norm": 1.9253933429718018, "learning_rate": 9.847721446786686e-06, "loss": 0.2683, "step": 6350 }, { "epoch": 0.11568053256698012, "grad_norm": 2.005009889602661, "learning_rate": 9.847076791446736e-06, "loss": 0.2778, "step": 6360 }, { "epoch": 0.11586242019680242, "grad_norm": 2.063220977783203, "learning_rate": 9.846430801037796e-06, "loss": 0.2642, "step": 6370 }, { "epoch": 0.11604430782662471, "grad_norm": 2.323019504547119, "learning_rate": 9.84578347578405e-06, "loss": 0.2723, "step": 6380 }, { "epoch": 0.11622619545644701, "grad_norm": 1.843997597694397, "learning_rate": 9.845134815910141e-06, "loss": 0.2645, "step": 6390 }, { "epoch": 0.1164080830862693, "grad_norm": 2.035318374633789, "learning_rate": 9.844484821641175e-06, "loss": 0.2714, "step": 6400 }, { "epoch": 0.1165899707160916, "grad_norm": 2.1903116703033447, "learning_rate": 9.843833493202725e-06, "loss": 0.2729, "step": 6410 }, { "epoch": 0.1167718583459139, "grad_norm": 1.9219584465026855, "learning_rate": 9.843180830820824e-06, "loss": 0.2766, "step": 6420 }, { "epoch": 0.11695374597573618, "grad_norm": 2.1621315479278564, "learning_rate": 9.842526834721968e-06, "loss": 0.2726, "step": 6430 }, { "epoch": 0.11713563360555848, "grad_norm": 1.7539688348770142, "learning_rate": 9.841871505133116e-06, "loss": 0.2749, "step": 6440 }, { "epoch": 0.11731752123538078, "grad_norm": 1.9407087564468384, "learning_rate": 9.84121484228169e-06, "loss": 0.2667, "step": 6450 }, { "epoch": 0.11749940886520308, "grad_norm": 2.3421826362609863, "learning_rate": 9.840556846395575e-06, "loss": 0.2705, "step": 6460 }, { "epoch": 0.11768129649502537, "grad_norm": 1.6930785179138184, "learning_rate": 9.83989751770312e-06, "loss": 0.2695, "step": 6470 }, { "epoch": 0.11786318412484767, "grad_norm": 2.0107016563415527, "learning_rate": 9.839236856433132e-06, "loss": 0.2696, "step": 6480 }, { "epoch": 0.11804507175466997, "grad_norm": 2.1965296268463135, "learning_rate": 9.838574862814885e-06, "loss": 0.2737, "step": 6490 }, { "epoch": 0.11822695938449226, "grad_norm": 2.0577683448791504, "learning_rate": 9.837911537078114e-06, "loss": 0.27, "step": 6500 }, { "epoch": 0.11840884701431456, "grad_norm": 2.0963921546936035, "learning_rate": 9.837246879453015e-06, "loss": 0.2659, "step": 6510 }, { "epoch": 0.11859073464413686, "grad_norm": 1.895316243171692, "learning_rate": 9.836580890170248e-06, "loss": 0.2701, "step": 6520 }, { "epoch": 0.11877262227395914, "grad_norm": 2.6698243618011475, "learning_rate": 9.835913569460936e-06, "loss": 0.2752, "step": 6530 }, { "epoch": 0.11895450990378144, "grad_norm": 2.044926881790161, "learning_rate": 9.835244917556659e-06, "loss": 0.2709, "step": 6540 }, { "epoch": 0.11913639753360374, "grad_norm": 1.863968849182129, "learning_rate": 9.834574934689463e-06, "loss": 0.2603, "step": 6550 }, { "epoch": 0.11931828516342603, "grad_norm": 2.2572169303894043, "learning_rate": 9.833903621091858e-06, "loss": 0.2859, "step": 6560 }, { "epoch": 0.11950017279324833, "grad_norm": 2.0725252628326416, "learning_rate": 9.83323097699681e-06, "loss": 0.2715, "step": 6570 }, { "epoch": 0.11968206042307063, "grad_norm": 2.196953296661377, "learning_rate": 9.83255700263775e-06, "loss": 0.2731, "step": 6580 }, { "epoch": 0.11986394805289292, "grad_norm": 2.318014621734619, "learning_rate": 9.831881698248576e-06, "loss": 0.2766, "step": 6590 }, { "epoch": 0.12004583568271522, "grad_norm": 1.9593169689178467, "learning_rate": 9.831205064063633e-06, "loss": 0.2691, "step": 6600 }, { "epoch": 0.12022772331253752, "grad_norm": 1.8044135570526123, "learning_rate": 9.830527100317743e-06, "loss": 0.2676, "step": 6610 }, { "epoch": 0.12040961094235982, "grad_norm": 2.1179134845733643, "learning_rate": 9.829847807246182e-06, "loss": 0.2729, "step": 6620 }, { "epoch": 0.12059149857218211, "grad_norm": 2.0699362754821777, "learning_rate": 9.829167185084685e-06, "loss": 0.2719, "step": 6630 }, { "epoch": 0.1207733862020044, "grad_norm": 2.3758320808410645, "learning_rate": 9.828485234069457e-06, "loss": 0.2663, "step": 6640 }, { "epoch": 0.12095527383182669, "grad_norm": 1.9119694232940674, "learning_rate": 9.827801954437152e-06, "loss": 0.2709, "step": 6650 }, { "epoch": 0.12113716146164899, "grad_norm": 2.3850507736206055, "learning_rate": 9.8271173464249e-06, "loss": 0.2761, "step": 6660 }, { "epoch": 0.12131904909147129, "grad_norm": 2.1071863174438477, "learning_rate": 9.826431410270275e-06, "loss": 0.2715, "step": 6670 }, { "epoch": 0.12150093672129358, "grad_norm": 2.088996648788452, "learning_rate": 9.825744146211324e-06, "loss": 0.2764, "step": 6680 }, { "epoch": 0.12168282435111588, "grad_norm": 2.056175947189331, "learning_rate": 9.825055554486555e-06, "loss": 0.2705, "step": 6690 }, { "epoch": 0.12186471198093818, "grad_norm": 1.7988293170928955, "learning_rate": 9.824365635334927e-06, "loss": 0.2651, "step": 6700 }, { "epoch": 0.12204659961076048, "grad_norm": 2.968959331512451, "learning_rate": 9.823674388995872e-06, "loss": 0.274, "step": 6710 }, { "epoch": 0.12222848724058277, "grad_norm": 2.479417085647583, "learning_rate": 9.822981815709271e-06, "loss": 0.2745, "step": 6720 }, { "epoch": 0.12241037487040507, "grad_norm": 2.0367467403411865, "learning_rate": 9.822287915715475e-06, "loss": 0.2695, "step": 6730 }, { "epoch": 0.12259226250022735, "grad_norm": 1.9192004203796387, "learning_rate": 9.821592689255287e-06, "loss": 0.2706, "step": 6740 }, { "epoch": 0.12277415013004965, "grad_norm": 1.9723039865493774, "learning_rate": 9.82089613656998e-06, "loss": 0.2662, "step": 6750 }, { "epoch": 0.12295603775987195, "grad_norm": 2.0980241298675537, "learning_rate": 9.820198257901279e-06, "loss": 0.271, "step": 6760 }, { "epoch": 0.12313792538969424, "grad_norm": 2.195044755935669, "learning_rate": 9.819499053491371e-06, "loss": 0.2739, "step": 6770 }, { "epoch": 0.12331981301951654, "grad_norm": 2.096532106399536, "learning_rate": 9.818798523582906e-06, "loss": 0.2672, "step": 6780 }, { "epoch": 0.12350170064933884, "grad_norm": 1.7951505184173584, "learning_rate": 9.818096668418992e-06, "loss": 0.2704, "step": 6790 }, { "epoch": 0.12368358827916114, "grad_norm": 2.2349588871002197, "learning_rate": 9.817393488243199e-06, "loss": 0.2676, "step": 6800 }, { "epoch": 0.12386547590898343, "grad_norm": 2.0551187992095947, "learning_rate": 9.81668898329955e-06, "loss": 0.2791, "step": 6810 }, { "epoch": 0.12404736353880573, "grad_norm": 1.973987102508545, "learning_rate": 9.815983153832536e-06, "loss": 0.2752, "step": 6820 }, { "epoch": 0.12422925116862803, "grad_norm": 2.658951759338379, "learning_rate": 9.815276000087104e-06, "loss": 0.273, "step": 6830 }, { "epoch": 0.12441113879845032, "grad_norm": 2.2576608657836914, "learning_rate": 9.81456752230866e-06, "loss": 0.2774, "step": 6840 }, { "epoch": 0.12459302642827261, "grad_norm": 2.0820119380950928, "learning_rate": 9.813857720743069e-06, "loss": 0.2724, "step": 6850 }, { "epoch": 0.1247749140580949, "grad_norm": 1.7613835334777832, "learning_rate": 9.813146595636662e-06, "loss": 0.2692, "step": 6860 }, { "epoch": 0.1249568016879172, "grad_norm": 2.1212222576141357, "learning_rate": 9.81243414723622e-06, "loss": 0.2773, "step": 6870 }, { "epoch": 0.1251386893177395, "grad_norm": 2.0925939083099365, "learning_rate": 9.811720375788985e-06, "loss": 0.2737, "step": 6880 }, { "epoch": 0.1253205769475618, "grad_norm": 2.408587694168091, "learning_rate": 9.811005281542667e-06, "loss": 0.2698, "step": 6890 }, { "epoch": 0.1255024645773841, "grad_norm": 2.31778883934021, "learning_rate": 9.810288864745423e-06, "loss": 0.2737, "step": 6900 }, { "epoch": 0.1256843522072064, "grad_norm": 2.249347448348999, "learning_rate": 9.809571125645876e-06, "loss": 0.2675, "step": 6910 }, { "epoch": 0.1258662398370287, "grad_norm": 1.990417242050171, "learning_rate": 9.808852064493108e-06, "loss": 0.2709, "step": 6920 }, { "epoch": 0.12604812746685098, "grad_norm": 2.396634817123413, "learning_rate": 9.808131681536655e-06, "loss": 0.2767, "step": 6930 }, { "epoch": 0.12623001509667328, "grad_norm": 2.3744852542877197, "learning_rate": 9.80740997702652e-06, "loss": 0.2745, "step": 6940 }, { "epoch": 0.12641190272649558, "grad_norm": 1.8935449123382568, "learning_rate": 9.806686951213152e-06, "loss": 0.2693, "step": 6950 }, { "epoch": 0.12659379035631788, "grad_norm": 1.976562261581421, "learning_rate": 9.805962604347472e-06, "loss": 0.278, "step": 6960 }, { "epoch": 0.12677567798614017, "grad_norm": 2.412855625152588, "learning_rate": 9.80523693668085e-06, "loss": 0.2795, "step": 6970 }, { "epoch": 0.12695756561596247, "grad_norm": 1.7437015771865845, "learning_rate": 9.804509948465121e-06, "loss": 0.2701, "step": 6980 }, { "epoch": 0.12713945324578477, "grad_norm": 1.915553092956543, "learning_rate": 9.803781639952574e-06, "loss": 0.2718, "step": 6990 }, { "epoch": 0.12732134087560704, "grad_norm": 1.9514762163162231, "learning_rate": 9.803052011395956e-06, "loss": 0.2695, "step": 7000 }, { "epoch": 0.12750322850542933, "grad_norm": 1.8280830383300781, "learning_rate": 9.802321063048474e-06, "loss": 0.2698, "step": 7010 }, { "epoch": 0.12768511613525163, "grad_norm": 1.9753299951553345, "learning_rate": 9.801588795163794e-06, "loss": 0.2735, "step": 7020 }, { "epoch": 0.12786700376507393, "grad_norm": 2.0882484912872314, "learning_rate": 9.800855207996037e-06, "loss": 0.2703, "step": 7030 }, { "epoch": 0.12804889139489622, "grad_norm": 1.8476089239120483, "learning_rate": 9.800120301799781e-06, "loss": 0.2718, "step": 7040 }, { "epoch": 0.12823077902471852, "grad_norm": 2.0947279930114746, "learning_rate": 9.799384076830067e-06, "loss": 0.2736, "step": 7050 }, { "epoch": 0.12841266665454082, "grad_norm": 2.104279041290283, "learning_rate": 9.79864653334239e-06, "loss": 0.2756, "step": 7060 }, { "epoch": 0.12859455428436312, "grad_norm": 2.007946252822876, "learning_rate": 9.797907671592704e-06, "loss": 0.2693, "step": 7070 }, { "epoch": 0.1287764419141854, "grad_norm": 1.966328501701355, "learning_rate": 9.797167491837417e-06, "loss": 0.2703, "step": 7080 }, { "epoch": 0.1289583295440077, "grad_norm": 2.28542160987854, "learning_rate": 9.796425994333398e-06, "loss": 0.2787, "step": 7090 }, { "epoch": 0.12914021717383, "grad_norm": 1.8503503799438477, "learning_rate": 9.795683179337973e-06, "loss": 0.2697, "step": 7100 }, { "epoch": 0.1293221048036523, "grad_norm": 1.9124047756195068, "learning_rate": 9.794939047108925e-06, "loss": 0.2663, "step": 7110 }, { "epoch": 0.1295039924334746, "grad_norm": 2.0493764877319336, "learning_rate": 9.794193597904494e-06, "loss": 0.2727, "step": 7120 }, { "epoch": 0.1296858800632969, "grad_norm": 2.2535297870635986, "learning_rate": 9.793446831983375e-06, "loss": 0.2755, "step": 7130 }, { "epoch": 0.1298677676931192, "grad_norm": 1.9829463958740234, "learning_rate": 9.792698749604722e-06, "loss": 0.2665, "step": 7140 }, { "epoch": 0.1300496553229415, "grad_norm": 1.6937243938446045, "learning_rate": 9.791949351028145e-06, "loss": 0.265, "step": 7150 }, { "epoch": 0.1302315429527638, "grad_norm": 2.062657594680786, "learning_rate": 9.791198636513715e-06, "loss": 0.2773, "step": 7160 }, { "epoch": 0.1304134305825861, "grad_norm": 2.3050167560577393, "learning_rate": 9.79044660632195e-06, "loss": 0.2768, "step": 7170 }, { "epoch": 0.13059531821240838, "grad_norm": 2.13346266746521, "learning_rate": 9.789693260713835e-06, "loss": 0.2694, "step": 7180 }, { "epoch": 0.13077720584223068, "grad_norm": 2.1017186641693115, "learning_rate": 9.788938599950805e-06, "loss": 0.2656, "step": 7190 }, { "epoch": 0.13095909347205298, "grad_norm": 2.00433349609375, "learning_rate": 9.788182624294756e-06, "loss": 0.2685, "step": 7200 }, { "epoch": 0.13114098110187525, "grad_norm": 2.0764639377593994, "learning_rate": 9.787425334008034e-06, "loss": 0.2704, "step": 7210 }, { "epoch": 0.13132286873169755, "grad_norm": 1.9489102363586426, "learning_rate": 9.78666672935345e-06, "loss": 0.269, "step": 7220 }, { "epoch": 0.13150475636151984, "grad_norm": 2.024843454360962, "learning_rate": 9.785906810594261e-06, "loss": 0.2746, "step": 7230 }, { "epoch": 0.13168664399134214, "grad_norm": 1.8168283700942993, "learning_rate": 9.785145577994186e-06, "loss": 0.2811, "step": 7240 }, { "epoch": 0.13186853162116444, "grad_norm": 2.063729763031006, "learning_rate": 9.784383031817403e-06, "loss": 0.2703, "step": 7250 }, { "epoch": 0.13205041925098673, "grad_norm": 1.910118579864502, "learning_rate": 9.783619172328537e-06, "loss": 0.2753, "step": 7260 }, { "epoch": 0.13223230688080903, "grad_norm": 2.113741636276245, "learning_rate": 9.782853999792677e-06, "loss": 0.2661, "step": 7270 }, { "epoch": 0.13241419451063133, "grad_norm": 1.7757331132888794, "learning_rate": 9.782087514475363e-06, "loss": 0.2697, "step": 7280 }, { "epoch": 0.13259608214045362, "grad_norm": 2.0110747814178467, "learning_rate": 9.78131971664259e-06, "loss": 0.2793, "step": 7290 }, { "epoch": 0.13277796977027592, "grad_norm": 1.9604578018188477, "learning_rate": 9.780550606560818e-06, "loss": 0.2705, "step": 7300 }, { "epoch": 0.13295985740009822, "grad_norm": 2.31681227684021, "learning_rate": 9.779780184496945e-06, "loss": 0.2773, "step": 7310 }, { "epoch": 0.13314174502992052, "grad_norm": 2.2732443809509277, "learning_rate": 9.779008450718341e-06, "loss": 0.2648, "step": 7320 }, { "epoch": 0.1333236326597428, "grad_norm": 1.6980618238449097, "learning_rate": 9.77823540549282e-06, "loss": 0.2758, "step": 7330 }, { "epoch": 0.1335055202895651, "grad_norm": 2.1591973304748535, "learning_rate": 9.777461049088662e-06, "loss": 0.2693, "step": 7340 }, { "epoch": 0.1336874079193874, "grad_norm": 1.876729130744934, "learning_rate": 9.776685381774586e-06, "loss": 0.2725, "step": 7350 }, { "epoch": 0.1338692955492097, "grad_norm": 2.1398110389709473, "learning_rate": 9.775908403819784e-06, "loss": 0.2834, "step": 7360 }, { "epoch": 0.134051183179032, "grad_norm": 1.9038945436477661, "learning_rate": 9.775130115493889e-06, "loss": 0.2729, "step": 7370 }, { "epoch": 0.1342330708088543, "grad_norm": 2.269848585128784, "learning_rate": 9.774350517066998e-06, "loss": 0.2806, "step": 7380 }, { "epoch": 0.1344149584386766, "grad_norm": 2.206625461578369, "learning_rate": 9.773569608809655e-06, "loss": 0.2734, "step": 7390 }, { "epoch": 0.1345968460684989, "grad_norm": 1.9164247512817383, "learning_rate": 9.772787390992865e-06, "loss": 0.2821, "step": 7400 }, { "epoch": 0.1347787336983212, "grad_norm": 2.2497425079345703, "learning_rate": 9.772003863888082e-06, "loss": 0.2757, "step": 7410 }, { "epoch": 0.13496062132814346, "grad_norm": 1.9749068021774292, "learning_rate": 9.77121902776722e-06, "loss": 0.2705, "step": 7420 }, { "epoch": 0.13514250895796576, "grad_norm": 2.0749871730804443, "learning_rate": 9.77043288290264e-06, "loss": 0.273, "step": 7430 }, { "epoch": 0.13532439658778805, "grad_norm": 1.978787899017334, "learning_rate": 9.769645429567167e-06, "loss": 0.2662, "step": 7440 }, { "epoch": 0.13550628421761035, "grad_norm": 2.1563310623168945, "learning_rate": 9.768856668034072e-06, "loss": 0.2681, "step": 7450 }, { "epoch": 0.13568817184743265, "grad_norm": 1.8838967084884644, "learning_rate": 9.768066598577083e-06, "loss": 0.2746, "step": 7460 }, { "epoch": 0.13587005947725495, "grad_norm": 1.961010456085205, "learning_rate": 9.76727522147038e-06, "loss": 0.2667, "step": 7470 }, { "epoch": 0.13605194710707724, "grad_norm": 2.044339179992676, "learning_rate": 9.7664825369886e-06, "loss": 0.2726, "step": 7480 }, { "epoch": 0.13623383473689954, "grad_norm": 2.2387373447418213, "learning_rate": 9.765688545406831e-06, "loss": 0.2754, "step": 7490 }, { "epoch": 0.13641572236672184, "grad_norm": 2.238569974899292, "learning_rate": 9.764893247000616e-06, "loss": 0.2724, "step": 7500 }, { "epoch": 0.13659760999654413, "grad_norm": 1.9495487213134766, "learning_rate": 9.76409664204595e-06, "loss": 0.2807, "step": 7510 }, { "epoch": 0.13677949762636643, "grad_norm": 1.9459333419799805, "learning_rate": 9.763298730819286e-06, "loss": 0.276, "step": 7520 }, { "epoch": 0.13696138525618873, "grad_norm": 2.111726999282837, "learning_rate": 9.762499513597524e-06, "loss": 0.2744, "step": 7530 }, { "epoch": 0.13714327288601103, "grad_norm": 2.217513084411621, "learning_rate": 9.761698990658017e-06, "loss": 0.2695, "step": 7540 }, { "epoch": 0.13732516051583332, "grad_norm": 1.799977421760559, "learning_rate": 9.760897162278582e-06, "loss": 0.2688, "step": 7550 }, { "epoch": 0.13750704814565562, "grad_norm": 1.873760461807251, "learning_rate": 9.760094028737475e-06, "loss": 0.2712, "step": 7560 }, { "epoch": 0.13768893577547792, "grad_norm": 1.9231204986572266, "learning_rate": 9.759289590313414e-06, "loss": 0.271, "step": 7570 }, { "epoch": 0.1378708234053002, "grad_norm": 1.9535897970199585, "learning_rate": 9.758483847285565e-06, "loss": 0.2704, "step": 7580 }, { "epoch": 0.1380527110351225, "grad_norm": 2.100935697555542, "learning_rate": 9.75767679993355e-06, "loss": 0.2809, "step": 7590 }, { "epoch": 0.1382345986649448, "grad_norm": 1.9250428676605225, "learning_rate": 9.756868448537445e-06, "loss": 0.2747, "step": 7600 }, { "epoch": 0.1384164862947671, "grad_norm": 2.132201910018921, "learning_rate": 9.75605879337777e-06, "loss": 0.2739, "step": 7610 }, { "epoch": 0.1385983739245894, "grad_norm": 1.9315030574798584, "learning_rate": 9.755247834735509e-06, "loss": 0.2763, "step": 7620 }, { "epoch": 0.13878026155441167, "grad_norm": 2.006822347640991, "learning_rate": 9.75443557289209e-06, "loss": 0.2744, "step": 7630 }, { "epoch": 0.13896214918423397, "grad_norm": 1.9110196828842163, "learning_rate": 9.753622008129395e-06, "loss": 0.2696, "step": 7640 }, { "epoch": 0.13914403681405627, "grad_norm": 2.033297061920166, "learning_rate": 9.752807140729762e-06, "loss": 0.2763, "step": 7650 }, { "epoch": 0.13932592444387856, "grad_norm": 2.016798257827759, "learning_rate": 9.751990970975977e-06, "loss": 0.2777, "step": 7660 }, { "epoch": 0.13950781207370086, "grad_norm": 2.0587148666381836, "learning_rate": 9.751173499151278e-06, "loss": 0.2686, "step": 7670 }, { "epoch": 0.13968969970352316, "grad_norm": 1.989397406578064, "learning_rate": 9.750354725539359e-06, "loss": 0.2725, "step": 7680 }, { "epoch": 0.13987158733334545, "grad_norm": 2.485560178756714, "learning_rate": 9.749534650424359e-06, "loss": 0.2724, "step": 7690 }, { "epoch": 0.14005347496316775, "grad_norm": 2.244694232940674, "learning_rate": 9.748713274090876e-06, "loss": 0.2715, "step": 7700 }, { "epoch": 0.14023536259299005, "grad_norm": 1.853444218635559, "learning_rate": 9.747890596823953e-06, "loss": 0.2725, "step": 7710 }, { "epoch": 0.14041725022281235, "grad_norm": 2.1446988582611084, "learning_rate": 9.74706661890909e-06, "loss": 0.2849, "step": 7720 }, { "epoch": 0.14059913785263464, "grad_norm": 2.2862093448638916, "learning_rate": 9.746241340632233e-06, "loss": 0.2775, "step": 7730 }, { "epoch": 0.14078102548245694, "grad_norm": 1.7693822383880615, "learning_rate": 9.745414762279786e-06, "loss": 0.2785, "step": 7740 }, { "epoch": 0.14096291311227924, "grad_norm": 1.9367183446884155, "learning_rate": 9.744586884138598e-06, "loss": 0.2723, "step": 7750 }, { "epoch": 0.14114480074210153, "grad_norm": 2.223177671432495, "learning_rate": 9.743757706495972e-06, "loss": 0.275, "step": 7760 }, { "epoch": 0.14132668837192383, "grad_norm": 1.8188862800598145, "learning_rate": 9.74292722963966e-06, "loss": 0.2651, "step": 7770 }, { "epoch": 0.14150857600174613, "grad_norm": 2.0177745819091797, "learning_rate": 9.742095453857869e-06, "loss": 0.2749, "step": 7780 }, { "epoch": 0.14169046363156843, "grad_norm": 2.0186245441436768, "learning_rate": 9.741262379439253e-06, "loss": 0.2701, "step": 7790 }, { "epoch": 0.14187235126139072, "grad_norm": 1.9860202074050903, "learning_rate": 9.740428006672914e-06, "loss": 0.2766, "step": 7800 }, { "epoch": 0.14205423889121302, "grad_norm": 2.2777185440063477, "learning_rate": 9.739592335848415e-06, "loss": 0.2713, "step": 7810 }, { "epoch": 0.14223612652103532, "grad_norm": 1.9658492803573608, "learning_rate": 9.738755367255758e-06, "loss": 0.2701, "step": 7820 }, { "epoch": 0.1424180141508576, "grad_norm": 2.309454917907715, "learning_rate": 9.7379171011854e-06, "loss": 0.2747, "step": 7830 }, { "epoch": 0.14259990178067988, "grad_norm": 2.2083261013031006, "learning_rate": 9.737077537928252e-06, "loss": 0.2675, "step": 7840 }, { "epoch": 0.14278178941050218, "grad_norm": 2.0252718925476074, "learning_rate": 9.736236677775668e-06, "loss": 0.2737, "step": 7850 }, { "epoch": 0.14296367704032448, "grad_norm": 2.1313211917877197, "learning_rate": 9.735394521019456e-06, "loss": 0.2694, "step": 7860 }, { "epoch": 0.14314556467014677, "grad_norm": 2.1602728366851807, "learning_rate": 9.734551067951876e-06, "loss": 0.2745, "step": 7870 }, { "epoch": 0.14332745229996907, "grad_norm": 1.8900442123413086, "learning_rate": 9.733706318865634e-06, "loss": 0.2711, "step": 7880 }, { "epoch": 0.14350933992979137, "grad_norm": 2.1044933795928955, "learning_rate": 9.732860274053887e-06, "loss": 0.2729, "step": 7890 }, { "epoch": 0.14369122755961367, "grad_norm": 2.0659172534942627, "learning_rate": 9.732012933810242e-06, "loss": 0.2786, "step": 7900 }, { "epoch": 0.14387311518943596, "grad_norm": 1.8748278617858887, "learning_rate": 9.731164298428753e-06, "loss": 0.2647, "step": 7910 }, { "epoch": 0.14405500281925826, "grad_norm": 1.8489917516708374, "learning_rate": 9.73031436820393e-06, "loss": 0.2711, "step": 7920 }, { "epoch": 0.14423689044908056, "grad_norm": 1.9011132717132568, "learning_rate": 9.729463143430726e-06, "loss": 0.2704, "step": 7930 }, { "epoch": 0.14441877807890285, "grad_norm": 1.992617130279541, "learning_rate": 9.728610624404548e-06, "loss": 0.2761, "step": 7940 }, { "epoch": 0.14460066570872515, "grad_norm": 1.7454990148544312, "learning_rate": 9.727756811421246e-06, "loss": 0.2725, "step": 7950 }, { "epoch": 0.14478255333854745, "grad_norm": 2.0218071937561035, "learning_rate": 9.726901704777124e-06, "loss": 0.2686, "step": 7960 }, { "epoch": 0.14496444096836975, "grad_norm": 2.1577184200286865, "learning_rate": 9.726045304768935e-06, "loss": 0.2704, "step": 7970 }, { "epoch": 0.14514632859819204, "grad_norm": 2.257772207260132, "learning_rate": 9.725187611693878e-06, "loss": 0.2723, "step": 7980 }, { "epoch": 0.14532821622801434, "grad_norm": 2.177757978439331, "learning_rate": 9.724328625849605e-06, "loss": 0.2771, "step": 7990 }, { "epoch": 0.14551010385783664, "grad_norm": 2.112724781036377, "learning_rate": 9.723468347534208e-06, "loss": 0.2736, "step": 8000 }, { "epoch": 0.14569199148765893, "grad_norm": 1.9965767860412598, "learning_rate": 9.722606777046239e-06, "loss": 0.2708, "step": 8010 }, { "epoch": 0.14587387911748123, "grad_norm": 1.9999022483825684, "learning_rate": 9.721743914684691e-06, "loss": 0.2721, "step": 8020 }, { "epoch": 0.14605576674730353, "grad_norm": 2.458739995956421, "learning_rate": 9.720879760749007e-06, "loss": 0.2778, "step": 8030 }, { "epoch": 0.14623765437712583, "grad_norm": 1.7583467960357666, "learning_rate": 9.720014315539077e-06, "loss": 0.2727, "step": 8040 }, { "epoch": 0.1464195420069481, "grad_norm": 2.0864531993865967, "learning_rate": 9.719147579355244e-06, "loss": 0.2742, "step": 8050 }, { "epoch": 0.1466014296367704, "grad_norm": 2.3123183250427246, "learning_rate": 9.718279552498291e-06, "loss": 0.2709, "step": 8060 }, { "epoch": 0.1467833172665927, "grad_norm": 2.1994118690490723, "learning_rate": 9.717410235269458e-06, "loss": 0.274, "step": 8070 }, { "epoch": 0.14696520489641499, "grad_norm": 1.8450433015823364, "learning_rate": 9.716539627970424e-06, "loss": 0.2651, "step": 8080 }, { "epoch": 0.14714709252623728, "grad_norm": 1.8910679817199707, "learning_rate": 9.715667730903321e-06, "loss": 0.2776, "step": 8090 }, { "epoch": 0.14732898015605958, "grad_norm": 2.0053162574768066, "learning_rate": 9.714794544370732e-06, "loss": 0.2792, "step": 8100 }, { "epoch": 0.14751086778588188, "grad_norm": 2.0249104499816895, "learning_rate": 9.713920068675676e-06, "loss": 0.2697, "step": 8110 }, { "epoch": 0.14769275541570417, "grad_norm": 2.0345160961151123, "learning_rate": 9.713044304121632e-06, "loss": 0.2703, "step": 8120 }, { "epoch": 0.14787464304552647, "grad_norm": 1.8591153621673584, "learning_rate": 9.712167251012518e-06, "loss": 0.2749, "step": 8130 }, { "epoch": 0.14805653067534877, "grad_norm": 1.847113847732544, "learning_rate": 9.7112889096527e-06, "loss": 0.2737, "step": 8140 }, { "epoch": 0.14823841830517107, "grad_norm": 2.243380546569824, "learning_rate": 9.710409280346998e-06, "loss": 0.2746, "step": 8150 }, { "epoch": 0.14842030593499336, "grad_norm": 1.7040703296661377, "learning_rate": 9.709528363400669e-06, "loss": 0.2734, "step": 8160 }, { "epoch": 0.14860219356481566, "grad_norm": 1.9996528625488281, "learning_rate": 9.708646159119423e-06, "loss": 0.2844, "step": 8170 }, { "epoch": 0.14878408119463796, "grad_norm": 1.9428738355636597, "learning_rate": 9.707762667809419e-06, "loss": 0.2668, "step": 8180 }, { "epoch": 0.14896596882446025, "grad_norm": 1.9420526027679443, "learning_rate": 9.706877889777254e-06, "loss": 0.2806, "step": 8190 }, { "epoch": 0.14914785645428255, "grad_norm": 2.0141732692718506, "learning_rate": 9.70599182532998e-06, "loss": 0.273, "step": 8200 }, { "epoch": 0.14932974408410485, "grad_norm": 2.2449564933776855, "learning_rate": 9.70510447477509e-06, "loss": 0.2691, "step": 8210 }, { "epoch": 0.14951163171392715, "grad_norm": 2.162458896636963, "learning_rate": 9.704215838420528e-06, "loss": 0.2711, "step": 8220 }, { "epoch": 0.14969351934374944, "grad_norm": 1.9650862216949463, "learning_rate": 9.703325916574677e-06, "loss": 0.2706, "step": 8230 }, { "epoch": 0.14987540697357174, "grad_norm": 2.0945005416870117, "learning_rate": 9.702434709546376e-06, "loss": 0.2718, "step": 8240 }, { "epoch": 0.15005729460339404, "grad_norm": 2.030740261077881, "learning_rate": 9.701542217644902e-06, "loss": 0.2772, "step": 8250 }, { "epoch": 0.1502391822332163, "grad_norm": 2.142940044403076, "learning_rate": 9.700648441179982e-06, "loss": 0.2729, "step": 8260 }, { "epoch": 0.1504210698630386, "grad_norm": 2.0512921810150146, "learning_rate": 9.699753380461785e-06, "loss": 0.2681, "step": 8270 }, { "epoch": 0.1506029574928609, "grad_norm": 2.0757834911346436, "learning_rate": 9.69885703580093e-06, "loss": 0.2669, "step": 8280 }, { "epoch": 0.1507848451226832, "grad_norm": 1.7079254388809204, "learning_rate": 9.69795940750848e-06, "loss": 0.2769, "step": 8290 }, { "epoch": 0.1509667327525055, "grad_norm": 1.8910095691680908, "learning_rate": 9.69706049589594e-06, "loss": 0.2698, "step": 8300 }, { "epoch": 0.1511486203823278, "grad_norm": 1.9193789958953857, "learning_rate": 9.696160301275268e-06, "loss": 0.2691, "step": 8310 }, { "epoch": 0.1513305080121501, "grad_norm": 2.118091583251953, "learning_rate": 9.69525882395886e-06, "loss": 0.2706, "step": 8320 }, { "epoch": 0.15151239564197239, "grad_norm": 2.02308988571167, "learning_rate": 9.694356064259561e-06, "loss": 0.2705, "step": 8330 }, { "epoch": 0.15169428327179468, "grad_norm": 2.495108127593994, "learning_rate": 9.69345202249066e-06, "loss": 0.2645, "step": 8340 }, { "epoch": 0.15187617090161698, "grad_norm": 1.800036072731018, "learning_rate": 9.692546698965887e-06, "loss": 0.2746, "step": 8350 }, { "epoch": 0.15205805853143928, "grad_norm": 1.694621205329895, "learning_rate": 9.691640093999427e-06, "loss": 0.2678, "step": 8360 }, { "epoch": 0.15223994616126157, "grad_norm": 2.066009521484375, "learning_rate": 9.690732207905898e-06, "loss": 0.2787, "step": 8370 }, { "epoch": 0.15242183379108387, "grad_norm": 2.082476854324341, "learning_rate": 9.689823041000372e-06, "loss": 0.2804, "step": 8380 }, { "epoch": 0.15260372142090617, "grad_norm": 1.771616816520691, "learning_rate": 9.68891259359836e-06, "loss": 0.2707, "step": 8390 }, { "epoch": 0.15278560905072847, "grad_norm": 1.8816124200820923, "learning_rate": 9.688000866015816e-06, "loss": 0.2714, "step": 8400 }, { "epoch": 0.15296749668055076, "grad_norm": 2.0238237380981445, "learning_rate": 9.687087858569145e-06, "loss": 0.2768, "step": 8410 }, { "epoch": 0.15314938431037306, "grad_norm": 2.012746810913086, "learning_rate": 9.68617357157519e-06, "loss": 0.2723, "step": 8420 }, { "epoch": 0.15333127194019536, "grad_norm": 1.9547268152236938, "learning_rate": 9.68525800535124e-06, "loss": 0.2705, "step": 8430 }, { "epoch": 0.15351315957001765, "grad_norm": 1.9507250785827637, "learning_rate": 9.68434116021503e-06, "loss": 0.2736, "step": 8440 }, { "epoch": 0.15369504719983995, "grad_norm": 2.213735818862915, "learning_rate": 9.683423036484735e-06, "loss": 0.2747, "step": 8450 }, { "epoch": 0.15387693482966222, "grad_norm": 1.6955523490905762, "learning_rate": 9.68250363447898e-06, "loss": 0.2722, "step": 8460 }, { "epoch": 0.15405882245948452, "grad_norm": 1.7385085821151733, "learning_rate": 9.681582954516823e-06, "loss": 0.2669, "step": 8470 }, { "epoch": 0.15424071008930682, "grad_norm": 1.750905990600586, "learning_rate": 9.680660996917778e-06, "loss": 0.269, "step": 8480 }, { "epoch": 0.1544225977191291, "grad_norm": 2.132955551147461, "learning_rate": 9.679737762001793e-06, "loss": 0.275, "step": 8490 }, { "epoch": 0.1546044853489514, "grad_norm": 2.0018348693847656, "learning_rate": 9.678813250089261e-06, "loss": 0.2798, "step": 8500 }, { "epoch": 0.1547863729787737, "grad_norm": 1.7897703647613525, "learning_rate": 9.677887461501023e-06, "loss": 0.2745, "step": 8510 }, { "epoch": 0.154968260608596, "grad_norm": 1.8784393072128296, "learning_rate": 9.676960396558359e-06, "loss": 0.2804, "step": 8520 }, { "epoch": 0.1551501482384183, "grad_norm": 1.897304892539978, "learning_rate": 9.67603205558299e-06, "loss": 0.2652, "step": 8530 }, { "epoch": 0.1553320358682406, "grad_norm": 2.1000478267669678, "learning_rate": 9.675102438897086e-06, "loss": 0.2708, "step": 8540 }, { "epoch": 0.1555139234980629, "grad_norm": 1.973226547241211, "learning_rate": 9.674171546823253e-06, "loss": 0.2655, "step": 8550 }, { "epoch": 0.1556958111278852, "grad_norm": 2.093437671661377, "learning_rate": 9.673239379684546e-06, "loss": 0.2726, "step": 8560 }, { "epoch": 0.1558776987577075, "grad_norm": 2.179790735244751, "learning_rate": 9.672305937804456e-06, "loss": 0.2755, "step": 8570 }, { "epoch": 0.1560595863875298, "grad_norm": 1.819553017616272, "learning_rate": 9.671371221506925e-06, "loss": 0.2682, "step": 8580 }, { "epoch": 0.15624147401735208, "grad_norm": 1.7596627473831177, "learning_rate": 9.670435231116325e-06, "loss": 0.279, "step": 8590 }, { "epoch": 0.15642336164717438, "grad_norm": 1.949865698814392, "learning_rate": 9.669497966957483e-06, "loss": 0.2725, "step": 8600 }, { "epoch": 0.15660524927699668, "grad_norm": 2.012188196182251, "learning_rate": 9.668559429355657e-06, "loss": 0.2777, "step": 8610 }, { "epoch": 0.15678713690681897, "grad_norm": 1.8423269987106323, "learning_rate": 9.667619618636556e-06, "loss": 0.2729, "step": 8620 }, { "epoch": 0.15696902453664127, "grad_norm": 2.0667035579681396, "learning_rate": 9.666678535126327e-06, "loss": 0.2758, "step": 8630 }, { "epoch": 0.15715091216646357, "grad_norm": 2.082002878189087, "learning_rate": 9.665736179151558e-06, "loss": 0.2755, "step": 8640 }, { "epoch": 0.15733279979628587, "grad_norm": 1.9390238523483276, "learning_rate": 9.664792551039275e-06, "loss": 0.2709, "step": 8650 }, { "epoch": 0.15751468742610816, "grad_norm": 1.7984260320663452, "learning_rate": 9.663847651116955e-06, "loss": 0.2742, "step": 8660 }, { "epoch": 0.15769657505593043, "grad_norm": 1.8498462438583374, "learning_rate": 9.66290147971251e-06, "loss": 0.2742, "step": 8670 }, { "epoch": 0.15787846268575273, "grad_norm": 1.8986104726791382, "learning_rate": 9.661954037154293e-06, "loss": 0.2838, "step": 8680 }, { "epoch": 0.15806035031557503, "grad_norm": 2.097938060760498, "learning_rate": 9.661005323771098e-06, "loss": 0.2728, "step": 8690 }, { "epoch": 0.15824223794539732, "grad_norm": 1.9013657569885254, "learning_rate": 9.660055339892166e-06, "loss": 0.2724, "step": 8700 }, { "epoch": 0.15842412557521962, "grad_norm": 1.797692060470581, "learning_rate": 9.65910408584717e-06, "loss": 0.2756, "step": 8710 }, { "epoch": 0.15860601320504192, "grad_norm": 1.8362126350402832, "learning_rate": 9.65815156196623e-06, "loss": 0.2777, "step": 8720 }, { "epoch": 0.15878790083486422, "grad_norm": 1.9964643716812134, "learning_rate": 9.657197768579905e-06, "loss": 0.2745, "step": 8730 }, { "epoch": 0.1589697884646865, "grad_norm": 1.9255918264389038, "learning_rate": 9.656242706019193e-06, "loss": 0.2704, "step": 8740 }, { "epoch": 0.1591516760945088, "grad_norm": 2.402369737625122, "learning_rate": 9.655286374615534e-06, "loss": 0.2787, "step": 8750 }, { "epoch": 0.1593335637243311, "grad_norm": 2.0782716274261475, "learning_rate": 9.654328774700811e-06, "loss": 0.2702, "step": 8760 }, { "epoch": 0.1595154513541534, "grad_norm": 1.9463781118392944, "learning_rate": 9.653369906607341e-06, "loss": 0.28, "step": 8770 }, { "epoch": 0.1596973389839757, "grad_norm": 1.7673627138137817, "learning_rate": 9.652409770667886e-06, "loss": 0.2803, "step": 8780 }, { "epoch": 0.159879226613798, "grad_norm": 2.084846258163452, "learning_rate": 9.651448367215646e-06, "loss": 0.2781, "step": 8790 }, { "epoch": 0.1600611142436203, "grad_norm": 1.776212215423584, "learning_rate": 9.65048569658426e-06, "loss": 0.2702, "step": 8800 }, { "epoch": 0.1602430018734426, "grad_norm": 1.8076637983322144, "learning_rate": 9.649521759107812e-06, "loss": 0.271, "step": 8810 }, { "epoch": 0.1604248895032649, "grad_norm": 1.8371535539627075, "learning_rate": 9.648556555120817e-06, "loss": 0.2779, "step": 8820 }, { "epoch": 0.1606067771330872, "grad_norm": 2.280036687850952, "learning_rate": 9.647590084958238e-06, "loss": 0.2761, "step": 8830 }, { "epoch": 0.16078866476290948, "grad_norm": 1.7903146743774414, "learning_rate": 9.64662234895547e-06, "loss": 0.2678, "step": 8840 }, { "epoch": 0.16097055239273178, "grad_norm": 2.123106002807617, "learning_rate": 9.645653347448356e-06, "loss": 0.2729, "step": 8850 }, { "epoch": 0.16115244002255408, "grad_norm": 1.764780879020691, "learning_rate": 9.64468308077317e-06, "loss": 0.2689, "step": 8860 }, { "epoch": 0.16133432765237637, "grad_norm": 2.0765023231506348, "learning_rate": 9.643711549266625e-06, "loss": 0.2776, "step": 8870 }, { "epoch": 0.16151621528219864, "grad_norm": 2.270249605178833, "learning_rate": 9.642738753265882e-06, "loss": 0.2832, "step": 8880 }, { "epoch": 0.16169810291202094, "grad_norm": 1.918045997619629, "learning_rate": 9.641764693108532e-06, "loss": 0.2787, "step": 8890 }, { "epoch": 0.16187999054184324, "grad_norm": 2.1424458026885986, "learning_rate": 9.640789369132609e-06, "loss": 0.2756, "step": 8900 }, { "epoch": 0.16206187817166554, "grad_norm": 2.020528554916382, "learning_rate": 9.639812781676581e-06, "loss": 0.2746, "step": 8910 }, { "epoch": 0.16224376580148783, "grad_norm": 1.7173594236373901, "learning_rate": 9.638834931079362e-06, "loss": 0.2737, "step": 8920 }, { "epoch": 0.16242565343131013, "grad_norm": 1.8546167612075806, "learning_rate": 9.637855817680296e-06, "loss": 0.2666, "step": 8930 }, { "epoch": 0.16260754106113243, "grad_norm": 2.283088207244873, "learning_rate": 9.636875441819173e-06, "loss": 0.2683, "step": 8940 }, { "epoch": 0.16278942869095472, "grad_norm": 1.8376396894454956, "learning_rate": 9.635893803836213e-06, "loss": 0.2731, "step": 8950 }, { "epoch": 0.16297131632077702, "grad_norm": 2.0961368083953857, "learning_rate": 9.634910904072084e-06, "loss": 0.283, "step": 8960 }, { "epoch": 0.16315320395059932, "grad_norm": 2.1244053840637207, "learning_rate": 9.633926742867884e-06, "loss": 0.2721, "step": 8970 }, { "epoch": 0.16333509158042162, "grad_norm": 1.8164786100387573, "learning_rate": 9.632941320565146e-06, "loss": 0.276, "step": 8980 }, { "epoch": 0.1635169792102439, "grad_norm": 1.8274409770965576, "learning_rate": 9.631954637505854e-06, "loss": 0.2795, "step": 8990 }, { "epoch": 0.1636988668400662, "grad_norm": 2.123063087463379, "learning_rate": 9.630966694032415e-06, "loss": 0.2714, "step": 9000 }, { "epoch": 0.1638807544698885, "grad_norm": 2.144561529159546, "learning_rate": 9.629977490487684e-06, "loss": 0.2652, "step": 9010 }, { "epoch": 0.1640626420997108, "grad_norm": 2.094001054763794, "learning_rate": 9.628987027214945e-06, "loss": 0.2776, "step": 9020 }, { "epoch": 0.1642445297295331, "grad_norm": 1.9228192567825317, "learning_rate": 9.627995304557924e-06, "loss": 0.2711, "step": 9030 }, { "epoch": 0.1644264173593554, "grad_norm": 1.9204175472259521, "learning_rate": 9.627002322860787e-06, "loss": 0.2641, "step": 9040 }, { "epoch": 0.1646083049891777, "grad_norm": 2.061509609222412, "learning_rate": 9.62600808246813e-06, "loss": 0.2764, "step": 9050 }, { "epoch": 0.164790192619, "grad_norm": 1.911853551864624, "learning_rate": 9.625012583724988e-06, "loss": 0.2791, "step": 9060 }, { "epoch": 0.1649720802488223, "grad_norm": 2.1014468669891357, "learning_rate": 9.624015826976832e-06, "loss": 0.2795, "step": 9070 }, { "epoch": 0.1651539678786446, "grad_norm": 2.187317371368408, "learning_rate": 9.623017812569578e-06, "loss": 0.2765, "step": 9080 }, { "epoch": 0.16533585550846686, "grad_norm": 1.8406503200531006, "learning_rate": 9.622018540849566e-06, "loss": 0.2735, "step": 9090 }, { "epoch": 0.16551774313828915, "grad_norm": 1.990330457687378, "learning_rate": 9.621018012163577e-06, "loss": 0.2757, "step": 9100 }, { "epoch": 0.16569963076811145, "grad_norm": 1.8596837520599365, "learning_rate": 9.620016226858835e-06, "loss": 0.2809, "step": 9110 }, { "epoch": 0.16588151839793375, "grad_norm": 1.839220404624939, "learning_rate": 9.619013185282988e-06, "loss": 0.277, "step": 9120 }, { "epoch": 0.16606340602775604, "grad_norm": 1.9758014678955078, "learning_rate": 9.61800888778413e-06, "loss": 0.2708, "step": 9130 }, { "epoch": 0.16624529365757834, "grad_norm": 1.8037797212600708, "learning_rate": 9.617003334710784e-06, "loss": 0.2645, "step": 9140 }, { "epoch": 0.16642718128740064, "grad_norm": 1.982155442237854, "learning_rate": 9.615996526411915e-06, "loss": 0.2709, "step": 9150 }, { "epoch": 0.16660906891722294, "grad_norm": 1.8843717575073242, "learning_rate": 9.614988463236919e-06, "loss": 0.2726, "step": 9160 }, { "epoch": 0.16679095654704523, "grad_norm": 2.0037147998809814, "learning_rate": 9.61397914553563e-06, "loss": 0.2767, "step": 9170 }, { "epoch": 0.16697284417686753, "grad_norm": 1.8941482305526733, "learning_rate": 9.612968573658314e-06, "loss": 0.2722, "step": 9180 }, { "epoch": 0.16715473180668983, "grad_norm": 1.9671142101287842, "learning_rate": 9.611956747955675e-06, "loss": 0.277, "step": 9190 }, { "epoch": 0.16733661943651212, "grad_norm": 3.582712173461914, "learning_rate": 9.61094366877885e-06, "loss": 0.2683, "step": 9200 }, { "epoch": 0.16751850706633442, "grad_norm": 1.8300371170043945, "learning_rate": 9.609929336479419e-06, "loss": 0.2736, "step": 9210 }, { "epoch": 0.16770039469615672, "grad_norm": 1.9422295093536377, "learning_rate": 9.608913751409383e-06, "loss": 0.2748, "step": 9220 }, { "epoch": 0.16788228232597902, "grad_norm": 1.9039976596832275, "learning_rate": 9.607896913921192e-06, "loss": 0.2782, "step": 9230 }, { "epoch": 0.1680641699558013, "grad_norm": 1.7665455341339111, "learning_rate": 9.606878824367715e-06, "loss": 0.2705, "step": 9240 }, { "epoch": 0.1682460575856236, "grad_norm": 2.16220760345459, "learning_rate": 9.605859483102274e-06, "loss": 0.2685, "step": 9250 }, { "epoch": 0.1684279452154459, "grad_norm": 1.9492876529693604, "learning_rate": 9.60483889047861e-06, "loss": 0.2707, "step": 9260 }, { "epoch": 0.1686098328452682, "grad_norm": 1.7041864395141602, "learning_rate": 9.603817046850904e-06, "loss": 0.2703, "step": 9270 }, { "epoch": 0.1687917204750905, "grad_norm": 1.8844947814941406, "learning_rate": 9.602793952573774e-06, "loss": 0.2876, "step": 9280 }, { "epoch": 0.1689736081049128, "grad_norm": 2.0339581966400146, "learning_rate": 9.601769608002267e-06, "loss": 0.2686, "step": 9290 }, { "epoch": 0.16915549573473507, "grad_norm": 1.9342691898345947, "learning_rate": 9.600744013491868e-06, "loss": 0.2721, "step": 9300 }, { "epoch": 0.16933738336455736, "grad_norm": 1.76339590549469, "learning_rate": 9.599717169398492e-06, "loss": 0.2648, "step": 9310 }, { "epoch": 0.16951927099437966, "grad_norm": 1.9777650833129883, "learning_rate": 9.59868907607849e-06, "loss": 0.2769, "step": 9320 }, { "epoch": 0.16970115862420196, "grad_norm": 2.141381025314331, "learning_rate": 9.597659733888645e-06, "loss": 0.2777, "step": 9330 }, { "epoch": 0.16988304625402426, "grad_norm": 2.1173386573791504, "learning_rate": 9.596629143186178e-06, "loss": 0.2827, "step": 9340 }, { "epoch": 0.17006493388384655, "grad_norm": 1.8263686895370483, "learning_rate": 9.595597304328735e-06, "loss": 0.2753, "step": 9350 }, { "epoch": 0.17024682151366885, "grad_norm": 1.9561870098114014, "learning_rate": 9.594564217674402e-06, "loss": 0.2866, "step": 9360 }, { "epoch": 0.17042870914349115, "grad_norm": 1.9518007040023804, "learning_rate": 9.593529883581699e-06, "loss": 0.2653, "step": 9370 }, { "epoch": 0.17061059677331344, "grad_norm": 1.91704523563385, "learning_rate": 9.59249430240957e-06, "loss": 0.2772, "step": 9380 }, { "epoch": 0.17079248440313574, "grad_norm": 1.9121688604354858, "learning_rate": 9.5914574745174e-06, "loss": 0.2743, "step": 9390 }, { "epoch": 0.17097437203295804, "grad_norm": 2.026770830154419, "learning_rate": 9.590419400265005e-06, "loss": 0.2758, "step": 9400 }, { "epoch": 0.17115625966278034, "grad_norm": 1.9774518013000488, "learning_rate": 9.589380080012633e-06, "loss": 0.2698, "step": 9410 }, { "epoch": 0.17133814729260263, "grad_norm": 1.790171504020691, "learning_rate": 9.588339514120963e-06, "loss": 0.2729, "step": 9420 }, { "epoch": 0.17152003492242493, "grad_norm": 2.0977931022644043, "learning_rate": 9.587297702951109e-06, "loss": 0.2758, "step": 9430 }, { "epoch": 0.17170192255224723, "grad_norm": 1.783064603805542, "learning_rate": 9.586254646864612e-06, "loss": 0.2676, "step": 9440 }, { "epoch": 0.17188381018206952, "grad_norm": 1.755246877670288, "learning_rate": 9.585210346223453e-06, "loss": 0.2748, "step": 9450 }, { "epoch": 0.17206569781189182, "grad_norm": 1.7352960109710693, "learning_rate": 9.584164801390039e-06, "loss": 0.2676, "step": 9460 }, { "epoch": 0.17224758544171412, "grad_norm": 2.068819046020508, "learning_rate": 9.583118012727208e-06, "loss": 0.2725, "step": 9470 }, { "epoch": 0.17242947307153642, "grad_norm": 1.8458836078643799, "learning_rate": 9.582069980598235e-06, "loss": 0.2722, "step": 9480 }, { "epoch": 0.1726113607013587, "grad_norm": 1.9512110948562622, "learning_rate": 9.58102070536682e-06, "loss": 0.2769, "step": 9490 }, { "epoch": 0.172793248331181, "grad_norm": 2.253920555114746, "learning_rate": 9.579970187397103e-06, "loss": 0.2725, "step": 9500 }, { "epoch": 0.17297513596100328, "grad_norm": 2.0215065479278564, "learning_rate": 9.578918427053646e-06, "loss": 0.2734, "step": 9510 }, { "epoch": 0.17315702359082558, "grad_norm": 2.3101632595062256, "learning_rate": 9.577865424701448e-06, "loss": 0.2733, "step": 9520 }, { "epoch": 0.17333891122064787, "grad_norm": 2.165318489074707, "learning_rate": 9.576811180705937e-06, "loss": 0.2811, "step": 9530 }, { "epoch": 0.17352079885047017, "grad_norm": 1.810988426208496, "learning_rate": 9.575755695432972e-06, "loss": 0.2742, "step": 9540 }, { "epoch": 0.17370268648029247, "grad_norm": 1.6860508918762207, "learning_rate": 9.574698969248844e-06, "loss": 0.2711, "step": 9550 }, { "epoch": 0.17388457411011476, "grad_norm": 1.8202017545700073, "learning_rate": 9.57364100252027e-06, "loss": 0.2618, "step": 9560 }, { "epoch": 0.17406646173993706, "grad_norm": 2.096804618835449, "learning_rate": 9.572581795614407e-06, "loss": 0.2732, "step": 9570 }, { "epoch": 0.17424834936975936, "grad_norm": 1.9285768270492554, "learning_rate": 9.571521348898831e-06, "loss": 0.2773, "step": 9580 }, { "epoch": 0.17443023699958166, "grad_norm": 1.6708158254623413, "learning_rate": 9.570459662741558e-06, "loss": 0.273, "step": 9590 }, { "epoch": 0.17461212462940395, "grad_norm": 2.0844011306762695, "learning_rate": 9.569396737511026e-06, "loss": 0.2744, "step": 9600 }, { "epoch": 0.17479401225922625, "grad_norm": 2.1797995567321777, "learning_rate": 9.56833257357611e-06, "loss": 0.2649, "step": 9610 }, { "epoch": 0.17497589988904855, "grad_norm": 2.0049219131469727, "learning_rate": 9.56726717130611e-06, "loss": 0.267, "step": 9620 }, { "epoch": 0.17515778751887084, "grad_norm": 2.0283684730529785, "learning_rate": 9.566200531070759e-06, "loss": 0.2744, "step": 9630 }, { "epoch": 0.17533967514869314, "grad_norm": 1.9576160907745361, "learning_rate": 9.565132653240218e-06, "loss": 0.2694, "step": 9640 }, { "epoch": 0.17552156277851544, "grad_norm": 2.4907073974609375, "learning_rate": 9.564063538185076e-06, "loss": 0.2695, "step": 9650 }, { "epoch": 0.17570345040833774, "grad_norm": 1.9302901029586792, "learning_rate": 9.562993186276354e-06, "loss": 0.2767, "step": 9660 }, { "epoch": 0.17588533803816003, "grad_norm": 1.7408215999603271, "learning_rate": 9.5619215978855e-06, "loss": 0.2729, "step": 9670 }, { "epoch": 0.17606722566798233, "grad_norm": 2.374721050262451, "learning_rate": 9.560848773384394e-06, "loss": 0.2745, "step": 9680 }, { "epoch": 0.17624911329780463, "grad_norm": 1.7011867761611938, "learning_rate": 9.559774713145343e-06, "loss": 0.2739, "step": 9690 }, { "epoch": 0.17643100092762692, "grad_norm": 1.8902027606964111, "learning_rate": 9.558699417541084e-06, "loss": 0.2776, "step": 9700 }, { "epoch": 0.17661288855744922, "grad_norm": 2.086294174194336, "learning_rate": 9.557622886944779e-06, "loss": 0.2777, "step": 9710 }, { "epoch": 0.1767947761872715, "grad_norm": 1.6577656269073486, "learning_rate": 9.556545121730024e-06, "loss": 0.2746, "step": 9720 }, { "epoch": 0.1769766638170938, "grad_norm": 1.950271725654602, "learning_rate": 9.55546612227084e-06, "loss": 0.2801, "step": 9730 }, { "epoch": 0.17715855144691608, "grad_norm": 2.183258533477783, "learning_rate": 9.554385888941678e-06, "loss": 0.2755, "step": 9740 }, { "epoch": 0.17734043907673838, "grad_norm": 1.7564911842346191, "learning_rate": 9.553304422117414e-06, "loss": 0.269, "step": 9750 }, { "epoch": 0.17752232670656068, "grad_norm": 2.0111258029937744, "learning_rate": 9.552221722173356e-06, "loss": 0.2782, "step": 9760 }, { "epoch": 0.17770421433638298, "grad_norm": 2.1925253868103027, "learning_rate": 9.55113778948524e-06, "loss": 0.2698, "step": 9770 }, { "epoch": 0.17788610196620527, "grad_norm": 2.1718697547912598, "learning_rate": 9.550052624429227e-06, "loss": 0.2656, "step": 9780 }, { "epoch": 0.17806798959602757, "grad_norm": 1.722040057182312, "learning_rate": 9.548966227381903e-06, "loss": 0.2675, "step": 9790 }, { "epoch": 0.17824987722584987, "grad_norm": 1.697676181793213, "learning_rate": 9.547878598720292e-06, "loss": 0.2734, "step": 9800 }, { "epoch": 0.17843176485567216, "grad_norm": 1.8973422050476074, "learning_rate": 9.546789738821837e-06, "loss": 0.2767, "step": 9810 }, { "epoch": 0.17861365248549446, "grad_norm": 1.8632769584655762, "learning_rate": 9.545699648064408e-06, "loss": 0.2758, "step": 9820 }, { "epoch": 0.17879554011531676, "grad_norm": 1.771690011024475, "learning_rate": 9.544608326826303e-06, "loss": 0.2695, "step": 9830 }, { "epoch": 0.17897742774513906, "grad_norm": 2.0536015033721924, "learning_rate": 9.543515775486253e-06, "loss": 0.2729, "step": 9840 }, { "epoch": 0.17915931537496135, "grad_norm": 1.8740781545639038, "learning_rate": 9.54242199442341e-06, "loss": 0.2741, "step": 9850 }, { "epoch": 0.17934120300478365, "grad_norm": 1.9123297929763794, "learning_rate": 9.541326984017349e-06, "loss": 0.2785, "step": 9860 }, { "epoch": 0.17952309063460595, "grad_norm": 2.202456474304199, "learning_rate": 9.540230744648083e-06, "loss": 0.2707, "step": 9870 }, { "epoch": 0.17970497826442824, "grad_norm": 1.885122537612915, "learning_rate": 9.539133276696042e-06, "loss": 0.2673, "step": 9880 }, { "epoch": 0.17988686589425054, "grad_norm": 2.096921682357788, "learning_rate": 9.538034580542083e-06, "loss": 0.2811, "step": 9890 }, { "epoch": 0.18006875352407284, "grad_norm": 1.9052374362945557, "learning_rate": 9.536934656567497e-06, "loss": 0.2758, "step": 9900 }, { "epoch": 0.18025064115389514, "grad_norm": 2.2027785778045654, "learning_rate": 9.535833505153993e-06, "loss": 0.2685, "step": 9910 }, { "epoch": 0.18043252878371743, "grad_norm": 1.8513128757476807, "learning_rate": 9.534731126683708e-06, "loss": 0.2768, "step": 9920 }, { "epoch": 0.1806144164135397, "grad_norm": 1.9086543321609497, "learning_rate": 9.533627521539207e-06, "loss": 0.2744, "step": 9930 }, { "epoch": 0.180796304043362, "grad_norm": 1.849601149559021, "learning_rate": 9.53252269010348e-06, "loss": 0.2651, "step": 9940 }, { "epoch": 0.1809781916731843, "grad_norm": 2.3139688968658447, "learning_rate": 9.53141663275994e-06, "loss": 0.2737, "step": 9950 }, { "epoch": 0.1811600793030066, "grad_norm": 2.0035409927368164, "learning_rate": 9.530309349892425e-06, "loss": 0.2761, "step": 9960 }, { "epoch": 0.1813419669328289, "grad_norm": 2.0779783725738525, "learning_rate": 9.529200841885206e-06, "loss": 0.274, "step": 9970 }, { "epoch": 0.1815238545626512, "grad_norm": 1.9135199785232544, "learning_rate": 9.528091109122968e-06, "loss": 0.2761, "step": 9980 }, { "epoch": 0.18170574219247349, "grad_norm": 2.015873908996582, "learning_rate": 9.52698015199083e-06, "loss": 0.2766, "step": 9990 }, { "epoch": 0.18188762982229578, "grad_norm": 1.7870525121688843, "learning_rate": 9.525867970874336e-06, "loss": 0.2766, "step": 10000 }, { "epoch": 0.18206951745211808, "grad_norm": 1.9164520502090454, "learning_rate": 9.524754566159446e-06, "loss": 0.2757, "step": 10010 }, { "epoch": 0.18225140508194038, "grad_norm": 1.6098192930221558, "learning_rate": 9.523639938232549e-06, "loss": 0.2717, "step": 10020 }, { "epoch": 0.18243329271176267, "grad_norm": 1.832531213760376, "learning_rate": 9.522524087480465e-06, "loss": 0.2699, "step": 10030 }, { "epoch": 0.18261518034158497, "grad_norm": 2.1719374656677246, "learning_rate": 9.52140701429043e-06, "loss": 0.2774, "step": 10040 }, { "epoch": 0.18279706797140727, "grad_norm": 1.7979397773742676, "learning_rate": 9.520288719050104e-06, "loss": 0.2777, "step": 10050 }, { "epoch": 0.18297895560122956, "grad_norm": 1.7915912866592407, "learning_rate": 9.519169202147581e-06, "loss": 0.2679, "step": 10060 }, { "epoch": 0.18316084323105186, "grad_norm": 1.8024026155471802, "learning_rate": 9.518048463971368e-06, "loss": 0.2789, "step": 10070 }, { "epoch": 0.18334273086087416, "grad_norm": 2.041003704071045, "learning_rate": 9.516926504910399e-06, "loss": 0.274, "step": 10080 }, { "epoch": 0.18352461849069646, "grad_norm": 1.9346330165863037, "learning_rate": 9.515803325354035e-06, "loss": 0.27, "step": 10090 }, { "epoch": 0.18370650612051875, "grad_norm": 1.7015018463134766, "learning_rate": 9.514678925692058e-06, "loss": 0.2732, "step": 10100 }, { "epoch": 0.18388839375034105, "grad_norm": 1.8444437980651855, "learning_rate": 9.513553306314672e-06, "loss": 0.2737, "step": 10110 }, { "epoch": 0.18407028138016335, "grad_norm": 1.891257405281067, "learning_rate": 9.512426467612507e-06, "loss": 0.276, "step": 10120 }, { "epoch": 0.18425216900998564, "grad_norm": 1.739830493927002, "learning_rate": 9.511298409976617e-06, "loss": 0.2781, "step": 10130 }, { "epoch": 0.18443405663980791, "grad_norm": 1.949972152709961, "learning_rate": 9.510169133798474e-06, "loss": 0.2731, "step": 10140 }, { "epoch": 0.1846159442696302, "grad_norm": 1.943725347518921, "learning_rate": 9.509038639469979e-06, "loss": 0.2724, "step": 10150 }, { "epoch": 0.1847978318994525, "grad_norm": 1.8673515319824219, "learning_rate": 9.507906927383449e-06, "loss": 0.269, "step": 10160 }, { "epoch": 0.1849797195292748, "grad_norm": 2.1397430896759033, "learning_rate": 9.50677399793163e-06, "loss": 0.2795, "step": 10170 }, { "epoch": 0.1851616071590971, "grad_norm": 1.7960939407348633, "learning_rate": 9.505639851507686e-06, "loss": 0.2737, "step": 10180 }, { "epoch": 0.1853434947889194, "grad_norm": 2.8883700370788574, "learning_rate": 9.504504488505208e-06, "loss": 0.2713, "step": 10190 }, { "epoch": 0.1855253824187417, "grad_norm": 2.1294071674346924, "learning_rate": 9.503367909318203e-06, "loss": 0.2775, "step": 10200 }, { "epoch": 0.185707270048564, "grad_norm": 1.9537389278411865, "learning_rate": 9.502230114341105e-06, "loss": 0.2694, "step": 10210 }, { "epoch": 0.1858891576783863, "grad_norm": 1.8413572311401367, "learning_rate": 9.501091103968768e-06, "loss": 0.2694, "step": 10220 }, { "epoch": 0.1860710453082086, "grad_norm": 1.8749016523361206, "learning_rate": 9.499950878596468e-06, "loss": 0.2763, "step": 10230 }, { "epoch": 0.18625293293803089, "grad_norm": 1.9982633590698242, "learning_rate": 9.498809438619905e-06, "loss": 0.2666, "step": 10240 }, { "epoch": 0.18643482056785318, "grad_norm": 2.3183014392852783, "learning_rate": 9.497666784435194e-06, "loss": 0.2766, "step": 10250 }, { "epoch": 0.18661670819767548, "grad_norm": 1.9269524812698364, "learning_rate": 9.49652291643888e-06, "loss": 0.2792, "step": 10260 }, { "epoch": 0.18679859582749778, "grad_norm": 1.789507269859314, "learning_rate": 9.495377835027922e-06, "loss": 0.2738, "step": 10270 }, { "epoch": 0.18698048345732007, "grad_norm": 1.8226600885391235, "learning_rate": 9.494231540599703e-06, "loss": 0.271, "step": 10280 }, { "epoch": 0.18716237108714237, "grad_norm": 1.7554099559783936, "learning_rate": 9.493084033552026e-06, "loss": 0.2726, "step": 10290 }, { "epoch": 0.18734425871696467, "grad_norm": 1.9530220031738281, "learning_rate": 9.491935314283119e-06, "loss": 0.2775, "step": 10300 }, { "epoch": 0.18752614634678696, "grad_norm": 1.9381216764450073, "learning_rate": 9.490785383191625e-06, "loss": 0.2795, "step": 10310 }, { "epoch": 0.18770803397660926, "grad_norm": 1.9191685914993286, "learning_rate": 9.48963424067661e-06, "loss": 0.2737, "step": 10320 }, { "epoch": 0.18788992160643156, "grad_norm": 1.8311654329299927, "learning_rate": 9.488481887137562e-06, "loss": 0.273, "step": 10330 }, { "epoch": 0.18807180923625383, "grad_norm": 1.7584612369537354, "learning_rate": 9.487328322974388e-06, "loss": 0.2818, "step": 10340 }, { "epoch": 0.18825369686607613, "grad_norm": 1.8267720937728882, "learning_rate": 9.486173548587409e-06, "loss": 0.2739, "step": 10350 }, { "epoch": 0.18843558449589842, "grad_norm": 2.0363881587982178, "learning_rate": 9.48501756437738e-06, "loss": 0.2764, "step": 10360 }, { "epoch": 0.18861747212572072, "grad_norm": 1.9981052875518799, "learning_rate": 9.483860370745462e-06, "loss": 0.2726, "step": 10370 }, { "epoch": 0.18879935975554302, "grad_norm": 2.10805606842041, "learning_rate": 9.482701968093242e-06, "loss": 0.2753, "step": 10380 }, { "epoch": 0.18898124738536531, "grad_norm": 2.002072334289551, "learning_rate": 9.481542356822728e-06, "loss": 0.2768, "step": 10390 }, { "epoch": 0.1891631350151876, "grad_norm": 1.7902003526687622, "learning_rate": 9.480381537336344e-06, "loss": 0.2685, "step": 10400 }, { "epoch": 0.1893450226450099, "grad_norm": 1.9048466682434082, "learning_rate": 9.479219510036935e-06, "loss": 0.2679, "step": 10410 }, { "epoch": 0.1895269102748322, "grad_norm": 1.929429292678833, "learning_rate": 9.478056275327764e-06, "loss": 0.2796, "step": 10420 }, { "epoch": 0.1897087979046545, "grad_norm": 1.9428995847702026, "learning_rate": 9.476891833612514e-06, "loss": 0.266, "step": 10430 }, { "epoch": 0.1898906855344768, "grad_norm": 1.872316598892212, "learning_rate": 9.475726185295287e-06, "loss": 0.2734, "step": 10440 }, { "epoch": 0.1900725731642991, "grad_norm": 1.9746558666229248, "learning_rate": 9.474559330780603e-06, "loss": 0.275, "step": 10450 }, { "epoch": 0.1902544607941214, "grad_norm": 2.1079325675964355, "learning_rate": 9.4733912704734e-06, "loss": 0.2755, "step": 10460 }, { "epoch": 0.1904363484239437, "grad_norm": 1.9716049432754517, "learning_rate": 9.472222004779039e-06, "loss": 0.2711, "step": 10470 }, { "epoch": 0.190618236053766, "grad_norm": 1.9512261152267456, "learning_rate": 9.471051534103294e-06, "loss": 0.2743, "step": 10480 }, { "epoch": 0.19080012368358829, "grad_norm": 1.7707000970840454, "learning_rate": 9.469879858852355e-06, "loss": 0.2724, "step": 10490 }, { "epoch": 0.19098201131341058, "grad_norm": 2.1509060859680176, "learning_rate": 9.468706979432838e-06, "loss": 0.2722, "step": 10500 }, { "epoch": 0.19116389894323288, "grad_norm": 1.7995363473892212, "learning_rate": 9.467532896251772e-06, "loss": 0.267, "step": 10510 }, { "epoch": 0.19134578657305518, "grad_norm": 1.78023362159729, "learning_rate": 9.466357609716605e-06, "loss": 0.2642, "step": 10520 }, { "epoch": 0.19152767420287747, "grad_norm": 1.7914700508117676, "learning_rate": 9.465181120235202e-06, "loss": 0.265, "step": 10530 }, { "epoch": 0.19170956183269977, "grad_norm": 1.7304517030715942, "learning_rate": 9.464003428215844e-06, "loss": 0.2781, "step": 10540 }, { "epoch": 0.19189144946252204, "grad_norm": 1.8554359674453735, "learning_rate": 9.462824534067232e-06, "loss": 0.2588, "step": 10550 }, { "epoch": 0.19207333709234434, "grad_norm": 1.7533056735992432, "learning_rate": 9.461644438198483e-06, "loss": 0.2713, "step": 10560 }, { "epoch": 0.19225522472216663, "grad_norm": 1.8769350051879883, "learning_rate": 9.460463141019133e-06, "loss": 0.2712, "step": 10570 }, { "epoch": 0.19243711235198893, "grad_norm": 1.9449656009674072, "learning_rate": 9.459280642939132e-06, "loss": 0.2715, "step": 10580 }, { "epoch": 0.19261899998181123, "grad_norm": 1.9287248849868774, "learning_rate": 9.458096944368846e-06, "loss": 0.2764, "step": 10590 }, { "epoch": 0.19280088761163353, "grad_norm": 1.867408037185669, "learning_rate": 9.456912045719063e-06, "loss": 0.2738, "step": 10600 }, { "epoch": 0.19298277524145582, "grad_norm": 2.1422793865203857, "learning_rate": 9.45572594740098e-06, "loss": 0.2721, "step": 10610 }, { "epoch": 0.19316466287127812, "grad_norm": 1.9495258331298828, "learning_rate": 9.454538649826216e-06, "loss": 0.2749, "step": 10620 }, { "epoch": 0.19334655050110042, "grad_norm": 1.8994220495224, "learning_rate": 9.453350153406804e-06, "loss": 0.2818, "step": 10630 }, { "epoch": 0.19352843813092271, "grad_norm": 1.9024006128311157, "learning_rate": 9.452160458555194e-06, "loss": 0.2738, "step": 10640 }, { "epoch": 0.193710325760745, "grad_norm": 1.9405429363250732, "learning_rate": 9.450969565684252e-06, "loss": 0.2729, "step": 10650 }, { "epoch": 0.1938922133905673, "grad_norm": 1.7795639038085938, "learning_rate": 9.449777475207258e-06, "loss": 0.2682, "step": 10660 }, { "epoch": 0.1940741010203896, "grad_norm": 1.798181414604187, "learning_rate": 9.448584187537908e-06, "loss": 0.2664, "step": 10670 }, { "epoch": 0.1942559886502119, "grad_norm": 1.9236454963684082, "learning_rate": 9.447389703090317e-06, "loss": 0.2777, "step": 10680 }, { "epoch": 0.1944378762800342, "grad_norm": 1.8004556894302368, "learning_rate": 9.446194022279008e-06, "loss": 0.2777, "step": 10690 }, { "epoch": 0.1946197639098565, "grad_norm": 1.7399585247039795, "learning_rate": 9.444997145518927e-06, "loss": 0.2774, "step": 10700 }, { "epoch": 0.1948016515396788, "grad_norm": 2.332223653793335, "learning_rate": 9.443799073225432e-06, "loss": 0.2624, "step": 10710 }, { "epoch": 0.1949835391695011, "grad_norm": 1.9357693195343018, "learning_rate": 9.442599805814295e-06, "loss": 0.2701, "step": 10720 }, { "epoch": 0.1951654267993234, "grad_norm": 1.8994145393371582, "learning_rate": 9.441399343701702e-06, "loss": 0.2708, "step": 10730 }, { "epoch": 0.19534731442914569, "grad_norm": 1.8139945268630981, "learning_rate": 9.440197687304258e-06, "loss": 0.2787, "step": 10740 }, { "epoch": 0.19552920205896798, "grad_norm": 2.0681557655334473, "learning_rate": 9.438994837038976e-06, "loss": 0.2763, "step": 10750 }, { "epoch": 0.19571108968879025, "grad_norm": 1.9405529499053955, "learning_rate": 9.437790793323288e-06, "loss": 0.2776, "step": 10760 }, { "epoch": 0.19589297731861255, "grad_norm": 1.6982471942901611, "learning_rate": 9.436585556575039e-06, "loss": 0.2781, "step": 10770 }, { "epoch": 0.19607486494843485, "grad_norm": 1.968641996383667, "learning_rate": 9.435379127212489e-06, "loss": 0.2697, "step": 10780 }, { "epoch": 0.19625675257825714, "grad_norm": 1.8409310579299927, "learning_rate": 9.434171505654308e-06, "loss": 0.2758, "step": 10790 }, { "epoch": 0.19643864020807944, "grad_norm": 1.7277308702468872, "learning_rate": 9.432962692319585e-06, "loss": 0.2781, "step": 10800 }, { "epoch": 0.19662052783790174, "grad_norm": 2.2404773235321045, "learning_rate": 9.431752687627822e-06, "loss": 0.271, "step": 10810 }, { "epoch": 0.19680241546772403, "grad_norm": 1.7278773784637451, "learning_rate": 9.430541491998928e-06, "loss": 0.2731, "step": 10820 }, { "epoch": 0.19698430309754633, "grad_norm": 1.9006972312927246, "learning_rate": 9.42932910585323e-06, "loss": 0.2793, "step": 10830 }, { "epoch": 0.19716619072736863, "grad_norm": 1.6869897842407227, "learning_rate": 9.428115529611472e-06, "loss": 0.2733, "step": 10840 }, { "epoch": 0.19734807835719093, "grad_norm": 1.8430904150009155, "learning_rate": 9.426900763694805e-06, "loss": 0.2733, "step": 10850 }, { "epoch": 0.19752996598701322, "grad_norm": 1.8075217008590698, "learning_rate": 9.425684808524793e-06, "loss": 0.2837, "step": 10860 }, { "epoch": 0.19771185361683552, "grad_norm": 1.6880673170089722, "learning_rate": 9.424467664523418e-06, "loss": 0.2737, "step": 10870 }, { "epoch": 0.19789374124665782, "grad_norm": 2.1630122661590576, "learning_rate": 9.42324933211307e-06, "loss": 0.2756, "step": 10880 }, { "epoch": 0.19807562887648011, "grad_norm": 2.162285089492798, "learning_rate": 9.42202981171655e-06, "loss": 0.2778, "step": 10890 }, { "epoch": 0.1982575165063024, "grad_norm": 2.0496747493743896, "learning_rate": 9.420809103757078e-06, "loss": 0.2691, "step": 10900 }, { "epoch": 0.1984394041361247, "grad_norm": 1.7122371196746826, "learning_rate": 9.419587208658279e-06, "loss": 0.2682, "step": 10910 }, { "epoch": 0.198621291765947, "grad_norm": 2.315670967102051, "learning_rate": 9.418364126844195e-06, "loss": 0.2677, "step": 10920 }, { "epoch": 0.1988031793957693, "grad_norm": 2.0239105224609375, "learning_rate": 9.417139858739275e-06, "loss": 0.27, "step": 10930 }, { "epoch": 0.1989850670255916, "grad_norm": 1.8743823766708374, "learning_rate": 9.415914404768384e-06, "loss": 0.2731, "step": 10940 }, { "epoch": 0.1991669546554139, "grad_norm": 1.7944059371948242, "learning_rate": 9.414687765356798e-06, "loss": 0.2692, "step": 10950 }, { "epoch": 0.1993488422852362, "grad_norm": 1.9201269149780273, "learning_rate": 9.413459940930204e-06, "loss": 0.2743, "step": 10960 }, { "epoch": 0.19953072991505846, "grad_norm": 1.943484902381897, "learning_rate": 9.412230931914696e-06, "loss": 0.2777, "step": 10970 }, { "epoch": 0.19971261754488076, "grad_norm": 1.8951433897018433, "learning_rate": 9.411000738736787e-06, "loss": 0.2782, "step": 10980 }, { "epoch": 0.19989450517470306, "grad_norm": 2.0131232738494873, "learning_rate": 9.409769361823392e-06, "loss": 0.2744, "step": 10990 }, { "epoch": 0.20007639280452535, "grad_norm": 2.0924229621887207, "learning_rate": 9.408536801601845e-06, "loss": 0.2794, "step": 11000 }, { "epoch": 0.20025828043434765, "grad_norm": 2.090834617614746, "learning_rate": 9.407303058499885e-06, "loss": 0.2723, "step": 11010 }, { "epoch": 0.20044016806416995, "grad_norm": 1.7026249170303345, "learning_rate": 9.406068132945667e-06, "loss": 0.2742, "step": 11020 }, { "epoch": 0.20062205569399225, "grad_norm": 2.083750009536743, "learning_rate": 9.40483202536775e-06, "loss": 0.2742, "step": 11030 }, { "epoch": 0.20080394332381454, "grad_norm": 1.8714016675949097, "learning_rate": 9.403594736195108e-06, "loss": 0.273, "step": 11040 }, { "epoch": 0.20098583095363684, "grad_norm": 1.8905434608459473, "learning_rate": 9.40235626585712e-06, "loss": 0.273, "step": 11050 }, { "epoch": 0.20116771858345914, "grad_norm": 1.934120774269104, "learning_rate": 9.40111661478358e-06, "loss": 0.2726, "step": 11060 }, { "epoch": 0.20134960621328143, "grad_norm": 2.516869306564331, "learning_rate": 9.399875783404692e-06, "loss": 0.2734, "step": 11070 }, { "epoch": 0.20153149384310373, "grad_norm": 1.7838307619094849, "learning_rate": 9.398633772151064e-06, "loss": 0.2768, "step": 11080 }, { "epoch": 0.20171338147292603, "grad_norm": 1.8807979822158813, "learning_rate": 9.397390581453722e-06, "loss": 0.2814, "step": 11090 }, { "epoch": 0.20189526910274833, "grad_norm": 1.988893985748291, "learning_rate": 9.396146211744089e-06, "loss": 0.2668, "step": 11100 }, { "epoch": 0.20207715673257062, "grad_norm": 1.9086296558380127, "learning_rate": 9.39490066345401e-06, "loss": 0.2748, "step": 11110 }, { "epoch": 0.20225904436239292, "grad_norm": 1.9411386251449585, "learning_rate": 9.393653937015732e-06, "loss": 0.2691, "step": 11120 }, { "epoch": 0.20244093199221522, "grad_norm": 2.021340847015381, "learning_rate": 9.39240603286191e-06, "loss": 0.2681, "step": 11130 }, { "epoch": 0.20262281962203751, "grad_norm": 2.0120580196380615, "learning_rate": 9.391156951425613e-06, "loss": 0.2719, "step": 11140 }, { "epoch": 0.2028047072518598, "grad_norm": 1.6372696161270142, "learning_rate": 9.389906693140316e-06, "loss": 0.2767, "step": 11150 }, { "epoch": 0.2029865948816821, "grad_norm": 2.0706400871276855, "learning_rate": 9.388655258439899e-06, "loss": 0.2704, "step": 11160 }, { "epoch": 0.2031684825115044, "grad_norm": 1.9598315954208374, "learning_rate": 9.387402647758656e-06, "loss": 0.2787, "step": 11170 }, { "epoch": 0.20335037014132668, "grad_norm": 1.667887568473816, "learning_rate": 9.386148861531283e-06, "loss": 0.2684, "step": 11180 }, { "epoch": 0.20353225777114897, "grad_norm": 2.0744388103485107, "learning_rate": 9.384893900192892e-06, "loss": 0.2695, "step": 11190 }, { "epoch": 0.20371414540097127, "grad_norm": 1.8445208072662354, "learning_rate": 9.383637764178994e-06, "loss": 0.2773, "step": 11200 }, { "epoch": 0.20389603303079357, "grad_norm": 1.6974648237228394, "learning_rate": 9.382380453925513e-06, "loss": 0.2753, "step": 11210 }, { "epoch": 0.20407792066061586, "grad_norm": 1.840415120124817, "learning_rate": 9.381121969868779e-06, "loss": 0.2741, "step": 11220 }, { "epoch": 0.20425980829043816, "grad_norm": 2.148963451385498, "learning_rate": 9.37986231244553e-06, "loss": 0.2728, "step": 11230 }, { "epoch": 0.20444169592026046, "grad_norm": 1.8806569576263428, "learning_rate": 9.378601482092911e-06, "loss": 0.2776, "step": 11240 }, { "epoch": 0.20462358355008275, "grad_norm": 2.066775321960449, "learning_rate": 9.377339479248474e-06, "loss": 0.2781, "step": 11250 }, { "epoch": 0.20480547117990505, "grad_norm": 1.7773993015289307, "learning_rate": 9.376076304350177e-06, "loss": 0.2746, "step": 11260 }, { "epoch": 0.20498735880972735, "grad_norm": 1.9708168506622314, "learning_rate": 9.374811957836385e-06, "loss": 0.2722, "step": 11270 }, { "epoch": 0.20516924643954965, "grad_norm": 1.9675759077072144, "learning_rate": 9.373546440145873e-06, "loss": 0.2721, "step": 11280 }, { "epoch": 0.20535113406937194, "grad_norm": 1.8144842386245728, "learning_rate": 9.372279751717815e-06, "loss": 0.2655, "step": 11290 }, { "epoch": 0.20553302169919424, "grad_norm": 1.9125782251358032, "learning_rate": 9.371011892991799e-06, "loss": 0.2765, "step": 11300 }, { "epoch": 0.20571490932901654, "grad_norm": 1.8095166683197021, "learning_rate": 9.369742864407817e-06, "loss": 0.2659, "step": 11310 }, { "epoch": 0.20589679695883883, "grad_norm": 1.690028429031372, "learning_rate": 9.368472666406261e-06, "loss": 0.2726, "step": 11320 }, { "epoch": 0.20607868458866113, "grad_norm": 1.8992899656295776, "learning_rate": 9.36720129942794e-06, "loss": 0.2709, "step": 11330 }, { "epoch": 0.20626057221848343, "grad_norm": 1.7260117530822754, "learning_rate": 9.365928763914056e-06, "loss": 0.2727, "step": 11340 }, { "epoch": 0.20644245984830573, "grad_norm": 1.9133585691452026, "learning_rate": 9.364655060306227e-06, "loss": 0.2719, "step": 11350 }, { "epoch": 0.20662434747812802, "grad_norm": 1.8431439399719238, "learning_rate": 9.36338018904647e-06, "loss": 0.2808, "step": 11360 }, { "epoch": 0.20680623510795032, "grad_norm": 1.8168963193893433, "learning_rate": 9.362104150577214e-06, "loss": 0.2745, "step": 11370 }, { "epoch": 0.20698812273777262, "grad_norm": 1.8568031787872314, "learning_rate": 9.360826945341283e-06, "loss": 0.2751, "step": 11380 }, { "epoch": 0.2071700103675949, "grad_norm": 2.115534543991089, "learning_rate": 9.359548573781916e-06, "loss": 0.2691, "step": 11390 }, { "epoch": 0.20735189799741718, "grad_norm": 1.8806118965148926, "learning_rate": 9.358269036342748e-06, "loss": 0.2705, "step": 11400 }, { "epoch": 0.20753378562723948, "grad_norm": 1.835433006286621, "learning_rate": 9.356988333467826e-06, "loss": 0.273, "step": 11410 }, { "epoch": 0.20771567325706178, "grad_norm": 1.925215482711792, "learning_rate": 9.355706465601598e-06, "loss": 0.2763, "step": 11420 }, { "epoch": 0.20789756088688408, "grad_norm": 1.8704605102539062, "learning_rate": 9.354423433188913e-06, "loss": 0.2812, "step": 11430 }, { "epoch": 0.20807944851670637, "grad_norm": 2.2065653800964355, "learning_rate": 9.353139236675033e-06, "loss": 0.2748, "step": 11440 }, { "epoch": 0.20826133614652867, "grad_norm": 1.814016342163086, "learning_rate": 9.351853876505615e-06, "loss": 0.2725, "step": 11450 }, { "epoch": 0.20844322377635097, "grad_norm": 2.003926992416382, "learning_rate": 9.350567353126728e-06, "loss": 0.2689, "step": 11460 }, { "epoch": 0.20862511140617326, "grad_norm": 2.373053789138794, "learning_rate": 9.349279666984833e-06, "loss": 0.2834, "step": 11470 }, { "epoch": 0.20880699903599556, "grad_norm": 1.9096895456314087, "learning_rate": 9.347990818526808e-06, "loss": 0.2739, "step": 11480 }, { "epoch": 0.20898888666581786, "grad_norm": 1.9425276517868042, "learning_rate": 9.346700808199924e-06, "loss": 0.2714, "step": 11490 }, { "epoch": 0.20917077429564016, "grad_norm": 1.7807502746582031, "learning_rate": 9.345409636451863e-06, "loss": 0.2676, "step": 11500 }, { "epoch": 0.20935266192546245, "grad_norm": 1.8098808526992798, "learning_rate": 9.344117303730703e-06, "loss": 0.2689, "step": 11510 }, { "epoch": 0.20953454955528475, "grad_norm": 1.9336600303649902, "learning_rate": 9.342823810484932e-06, "loss": 0.2693, "step": 11520 }, { "epoch": 0.20971643718510705, "grad_norm": 1.8456264734268188, "learning_rate": 9.341529157163433e-06, "loss": 0.2784, "step": 11530 }, { "epoch": 0.20989832481492934, "grad_norm": 1.738369345664978, "learning_rate": 9.340233344215497e-06, "loss": 0.2752, "step": 11540 }, { "epoch": 0.21008021244475164, "grad_norm": 2.0807223320007324, "learning_rate": 9.338936372090816e-06, "loss": 0.2864, "step": 11550 }, { "epoch": 0.21026210007457394, "grad_norm": 2.1410999298095703, "learning_rate": 9.337638241239485e-06, "loss": 0.2794, "step": 11560 }, { "epoch": 0.21044398770439623, "grad_norm": 1.8419269323349, "learning_rate": 9.336338952112002e-06, "loss": 0.2745, "step": 11570 }, { "epoch": 0.21062587533421853, "grad_norm": 2.0261361598968506, "learning_rate": 9.33503850515926e-06, "loss": 0.27, "step": 11580 }, { "epoch": 0.21080776296404083, "grad_norm": 1.9840096235275269, "learning_rate": 9.333736900832565e-06, "loss": 0.2837, "step": 11590 }, { "epoch": 0.2109896505938631, "grad_norm": 1.736121416091919, "learning_rate": 9.332434139583617e-06, "loss": 0.2708, "step": 11600 }, { "epoch": 0.2111715382236854, "grad_norm": 1.9508411884307861, "learning_rate": 9.331130221864515e-06, "loss": 0.2753, "step": 11610 }, { "epoch": 0.2113534258535077, "grad_norm": 1.8387222290039062, "learning_rate": 9.32982514812777e-06, "loss": 0.2743, "step": 11620 }, { "epoch": 0.21153531348333, "grad_norm": 1.890883207321167, "learning_rate": 9.328518918826284e-06, "loss": 0.2721, "step": 11630 }, { "epoch": 0.2117172011131523, "grad_norm": 1.589691400527954, "learning_rate": 9.327211534413366e-06, "loss": 0.2693, "step": 11640 }, { "epoch": 0.21189908874297458, "grad_norm": 1.9391671419143677, "learning_rate": 9.325902995342721e-06, "loss": 0.2768, "step": 11650 }, { "epoch": 0.21208097637279688, "grad_norm": 1.959780216217041, "learning_rate": 9.32459330206846e-06, "loss": 0.2774, "step": 11660 }, { "epoch": 0.21226286400261918, "grad_norm": 1.6463429927825928, "learning_rate": 9.323282455045092e-06, "loss": 0.2794, "step": 11670 }, { "epoch": 0.21244475163244148, "grad_norm": 1.8171968460083008, "learning_rate": 9.321970454727525e-06, "loss": 0.2718, "step": 11680 }, { "epoch": 0.21262663926226377, "grad_norm": 1.796289324760437, "learning_rate": 9.320657301571067e-06, "loss": 0.2745, "step": 11690 }, { "epoch": 0.21280852689208607, "grad_norm": 1.8547602891921997, "learning_rate": 9.31934299603143e-06, "loss": 0.2752, "step": 11700 }, { "epoch": 0.21299041452190837, "grad_norm": 1.7526992559432983, "learning_rate": 9.318027538564725e-06, "loss": 0.2718, "step": 11710 }, { "epoch": 0.21317230215173066, "grad_norm": 1.7903738021850586, "learning_rate": 9.31671092962746e-06, "loss": 0.2775, "step": 11720 }, { "epoch": 0.21335418978155296, "grad_norm": 1.8952739238739014, "learning_rate": 9.315393169676542e-06, "loss": 0.2731, "step": 11730 }, { "epoch": 0.21353607741137526, "grad_norm": 1.927901029586792, "learning_rate": 9.31407425916928e-06, "loss": 0.2651, "step": 11740 }, { "epoch": 0.21371796504119756, "grad_norm": 1.8563554286956787, "learning_rate": 9.312754198563384e-06, "loss": 0.2611, "step": 11750 }, { "epoch": 0.21389985267101985, "grad_norm": 1.779923439025879, "learning_rate": 9.311432988316958e-06, "loss": 0.2652, "step": 11760 }, { "epoch": 0.21408174030084215, "grad_norm": 1.7967721223831177, "learning_rate": 9.31011062888851e-06, "loss": 0.2702, "step": 11770 }, { "epoch": 0.21426362793066445, "grad_norm": 1.665639042854309, "learning_rate": 9.308787120736944e-06, "loss": 0.2711, "step": 11780 }, { "epoch": 0.21444551556048674, "grad_norm": 2.052065134048462, "learning_rate": 9.307462464321562e-06, "loss": 0.2655, "step": 11790 }, { "epoch": 0.21462740319030904, "grad_norm": 1.7898550033569336, "learning_rate": 9.306136660102067e-06, "loss": 0.2797, "step": 11800 }, { "epoch": 0.2148092908201313, "grad_norm": 1.8339335918426514, "learning_rate": 9.304809708538557e-06, "loss": 0.2693, "step": 11810 }, { "epoch": 0.2149911784499536, "grad_norm": 1.7683700323104858, "learning_rate": 9.303481610091535e-06, "loss": 0.2743, "step": 11820 }, { "epoch": 0.2151730660797759, "grad_norm": 2.0376949310302734, "learning_rate": 9.30215236522189e-06, "loss": 0.2725, "step": 11830 }, { "epoch": 0.2153549537095982, "grad_norm": 2.0251524448394775, "learning_rate": 9.30082197439092e-06, "loss": 0.2732, "step": 11840 }, { "epoch": 0.2155368413394205, "grad_norm": 1.731840968132019, "learning_rate": 9.29949043806032e-06, "loss": 0.2722, "step": 11850 }, { "epoch": 0.2157187289692428, "grad_norm": 1.8186293840408325, "learning_rate": 9.298157756692173e-06, "loss": 0.265, "step": 11860 }, { "epoch": 0.2159006165990651, "grad_norm": 1.753422737121582, "learning_rate": 9.296823930748971e-06, "loss": 0.2705, "step": 11870 }, { "epoch": 0.2160825042288874, "grad_norm": 1.6979035139083862, "learning_rate": 9.295488960693593e-06, "loss": 0.2663, "step": 11880 }, { "epoch": 0.2162643918587097, "grad_norm": 2.1958999633789062, "learning_rate": 9.294152846989322e-06, "loss": 0.2732, "step": 11890 }, { "epoch": 0.21644627948853198, "grad_norm": 1.7942111492156982, "learning_rate": 9.292815590099837e-06, "loss": 0.2722, "step": 11900 }, { "epoch": 0.21662816711835428, "grad_norm": 2.0883402824401855, "learning_rate": 9.291477190489214e-06, "loss": 0.268, "step": 11910 }, { "epoch": 0.21681005474817658, "grad_norm": 1.6724151372909546, "learning_rate": 9.290137648621917e-06, "loss": 0.2701, "step": 11920 }, { "epoch": 0.21699194237799888, "grad_norm": 1.7269465923309326, "learning_rate": 9.288796964962821e-06, "loss": 0.2663, "step": 11930 }, { "epoch": 0.21717383000782117, "grad_norm": 1.7306690216064453, "learning_rate": 9.287455139977186e-06, "loss": 0.2669, "step": 11940 }, { "epoch": 0.21735571763764347, "grad_norm": 1.741295337677002, "learning_rate": 9.286112174130672e-06, "loss": 0.2703, "step": 11950 }, { "epoch": 0.21753760526746577, "grad_norm": 1.9263209104537964, "learning_rate": 9.284768067889335e-06, "loss": 0.2775, "step": 11960 }, { "epoch": 0.21771949289728806, "grad_norm": 2.0351462364196777, "learning_rate": 9.283422821719627e-06, "loss": 0.2742, "step": 11970 }, { "epoch": 0.21790138052711036, "grad_norm": 1.7681140899658203, "learning_rate": 9.282076436088396e-06, "loss": 0.275, "step": 11980 }, { "epoch": 0.21808326815693266, "grad_norm": 2.133812189102173, "learning_rate": 9.280728911462882e-06, "loss": 0.2703, "step": 11990 }, { "epoch": 0.21826515578675496, "grad_norm": 2.045640707015991, "learning_rate": 9.279380248310724e-06, "loss": 0.2733, "step": 12000 }, { "epoch": 0.21844704341657725, "grad_norm": 1.9449206590652466, "learning_rate": 9.278030447099955e-06, "loss": 0.2715, "step": 12010 }, { "epoch": 0.21862893104639952, "grad_norm": 2.149052858352661, "learning_rate": 9.276679508299001e-06, "loss": 0.2759, "step": 12020 }, { "epoch": 0.21881081867622182, "grad_norm": 1.857033371925354, "learning_rate": 9.275327432376689e-06, "loss": 0.2731, "step": 12030 }, { "epoch": 0.21899270630604412, "grad_norm": 1.6104307174682617, "learning_rate": 9.273974219802231e-06, "loss": 0.2704, "step": 12040 }, { "epoch": 0.2191745939358664, "grad_norm": 2.0256187915802, "learning_rate": 9.272619871045242e-06, "loss": 0.2727, "step": 12050 }, { "epoch": 0.2193564815656887, "grad_norm": 1.6821942329406738, "learning_rate": 9.271264386575727e-06, "loss": 0.2648, "step": 12060 }, { "epoch": 0.219538369195511, "grad_norm": 1.8060052394866943, "learning_rate": 9.269907766864086e-06, "loss": 0.2706, "step": 12070 }, { "epoch": 0.2197202568253333, "grad_norm": 1.6181248426437378, "learning_rate": 9.268550012381114e-06, "loss": 0.2764, "step": 12080 }, { "epoch": 0.2199021444551556, "grad_norm": 1.8796236515045166, "learning_rate": 9.267191123597994e-06, "loss": 0.2685, "step": 12090 }, { "epoch": 0.2200840320849779, "grad_norm": 2.0317063331604004, "learning_rate": 9.265831100986317e-06, "loss": 0.2808, "step": 12100 }, { "epoch": 0.2202659197148002, "grad_norm": 2.0092546939849854, "learning_rate": 9.26446994501805e-06, "loss": 0.2704, "step": 12110 }, { "epoch": 0.2204478073446225, "grad_norm": 1.967968463897705, "learning_rate": 9.263107656165564e-06, "loss": 0.268, "step": 12120 }, { "epoch": 0.2206296949744448, "grad_norm": 1.6516035795211792, "learning_rate": 9.26174423490162e-06, "loss": 0.2662, "step": 12130 }, { "epoch": 0.2208115826042671, "grad_norm": 1.9583278894424438, "learning_rate": 9.260379681699373e-06, "loss": 0.2768, "step": 12140 }, { "epoch": 0.22099347023408938, "grad_norm": 1.876579999923706, "learning_rate": 9.25901399703237e-06, "loss": 0.2739, "step": 12150 }, { "epoch": 0.22117535786391168, "grad_norm": 2.030467987060547, "learning_rate": 9.257647181374551e-06, "loss": 0.2775, "step": 12160 }, { "epoch": 0.22135724549373398, "grad_norm": 1.8784620761871338, "learning_rate": 9.256279235200247e-06, "loss": 0.273, "step": 12170 }, { "epoch": 0.22153913312355628, "grad_norm": 2.043222665786743, "learning_rate": 9.254910158984185e-06, "loss": 0.28, "step": 12180 }, { "epoch": 0.22172102075337857, "grad_norm": 1.7995871305465698, "learning_rate": 9.253539953201482e-06, "loss": 0.2756, "step": 12190 }, { "epoch": 0.22190290838320087, "grad_norm": 1.7770836353302002, "learning_rate": 9.252168618327645e-06, "loss": 0.2663, "step": 12200 }, { "epoch": 0.22208479601302317, "grad_norm": 2.3927841186523438, "learning_rate": 9.250796154838573e-06, "loss": 0.2768, "step": 12210 }, { "epoch": 0.22226668364284544, "grad_norm": 1.6489793062210083, "learning_rate": 9.249422563210564e-06, "loss": 0.2769, "step": 12220 }, { "epoch": 0.22244857127266773, "grad_norm": 1.9464770555496216, "learning_rate": 9.248047843920298e-06, "loss": 0.2726, "step": 12230 }, { "epoch": 0.22263045890249003, "grad_norm": 2.0069162845611572, "learning_rate": 9.246671997444851e-06, "loss": 0.278, "step": 12240 }, { "epoch": 0.22281234653231233, "grad_norm": 2.061629295349121, "learning_rate": 9.24529502426169e-06, "loss": 0.2719, "step": 12250 }, { "epoch": 0.22299423416213462, "grad_norm": 2.300499677658081, "learning_rate": 9.243916924848672e-06, "loss": 0.2706, "step": 12260 }, { "epoch": 0.22317612179195692, "grad_norm": 1.7445076704025269, "learning_rate": 9.242537699684046e-06, "loss": 0.2742, "step": 12270 }, { "epoch": 0.22335800942177922, "grad_norm": 1.9484809637069702, "learning_rate": 9.241157349246448e-06, "loss": 0.2771, "step": 12280 }, { "epoch": 0.22353989705160152, "grad_norm": 1.7764972448349, "learning_rate": 9.239775874014912e-06, "loss": 0.2704, "step": 12290 }, { "epoch": 0.2237217846814238, "grad_norm": 2.0144388675689697, "learning_rate": 9.238393274468852e-06, "loss": 0.2756, "step": 12300 }, { "epoch": 0.2239036723112461, "grad_norm": 1.7675139904022217, "learning_rate": 9.237009551088083e-06, "loss": 0.267, "step": 12310 }, { "epoch": 0.2240855599410684, "grad_norm": 1.7335245609283447, "learning_rate": 9.235624704352803e-06, "loss": 0.2734, "step": 12320 }, { "epoch": 0.2242674475708907, "grad_norm": 1.7823476791381836, "learning_rate": 9.234238734743603e-06, "loss": 0.28, "step": 12330 }, { "epoch": 0.224449335200713, "grad_norm": 1.7374162673950195, "learning_rate": 9.232851642741462e-06, "loss": 0.2774, "step": 12340 }, { "epoch": 0.2246312228305353, "grad_norm": 1.6057292222976685, "learning_rate": 9.231463428827746e-06, "loss": 0.2749, "step": 12350 }, { "epoch": 0.2248131104603576, "grad_norm": 1.9116148948669434, "learning_rate": 9.230074093484218e-06, "loss": 0.284, "step": 12360 }, { "epoch": 0.2249949980901799, "grad_norm": 1.950750470161438, "learning_rate": 9.228683637193024e-06, "loss": 0.2731, "step": 12370 }, { "epoch": 0.2251768857200022, "grad_norm": 1.8480029106140137, "learning_rate": 9.2272920604367e-06, "loss": 0.2811, "step": 12380 }, { "epoch": 0.2253587733498245, "grad_norm": 1.8499081134796143, "learning_rate": 9.22589936369817e-06, "loss": 0.2666, "step": 12390 }, { "epoch": 0.22554066097964678, "grad_norm": 1.8559907674789429, "learning_rate": 9.22450554746075e-06, "loss": 0.2769, "step": 12400 }, { "epoch": 0.22572254860946908, "grad_norm": 1.8302828073501587, "learning_rate": 9.223110612208142e-06, "loss": 0.2752, "step": 12410 }, { "epoch": 0.22590443623929138, "grad_norm": 2.1148841381073, "learning_rate": 9.221714558424438e-06, "loss": 0.2807, "step": 12420 }, { "epoch": 0.22608632386911365, "grad_norm": 1.77881920337677, "learning_rate": 9.220317386594113e-06, "loss": 0.2663, "step": 12430 }, { "epoch": 0.22626821149893594, "grad_norm": 1.636911153793335, "learning_rate": 9.21891909720204e-06, "loss": 0.2731, "step": 12440 }, { "epoch": 0.22645009912875824, "grad_norm": 1.7024197578430176, "learning_rate": 9.217519690733469e-06, "loss": 0.2717, "step": 12450 }, { "epoch": 0.22663198675858054, "grad_norm": 1.8405940532684326, "learning_rate": 9.216119167674045e-06, "loss": 0.2805, "step": 12460 }, { "epoch": 0.22681387438840284, "grad_norm": 1.7754671573638916, "learning_rate": 9.214717528509797e-06, "loss": 0.2736, "step": 12470 }, { "epoch": 0.22699576201822513, "grad_norm": 1.9494483470916748, "learning_rate": 9.213314773727142e-06, "loss": 0.276, "step": 12480 }, { "epoch": 0.22717764964804743, "grad_norm": 1.959725260734558, "learning_rate": 9.211910903812887e-06, "loss": 0.2702, "step": 12490 }, { "epoch": 0.22735953727786973, "grad_norm": 1.7560570240020752, "learning_rate": 9.210505919254222e-06, "loss": 0.2745, "step": 12500 }, { "epoch": 0.22754142490769202, "grad_norm": 1.9136221408843994, "learning_rate": 9.209099820538723e-06, "loss": 0.2781, "step": 12510 }, { "epoch": 0.22772331253751432, "grad_norm": 1.7814679145812988, "learning_rate": 9.20769260815436e-06, "loss": 0.2681, "step": 12520 }, { "epoch": 0.22790520016733662, "grad_norm": 1.69901442527771, "learning_rate": 9.20628428258948e-06, "loss": 0.277, "step": 12530 }, { "epoch": 0.22808708779715892, "grad_norm": 2.1332626342773438, "learning_rate": 9.204874844332823e-06, "loss": 0.2782, "step": 12540 }, { "epoch": 0.2282689754269812, "grad_norm": 1.7052695751190186, "learning_rate": 9.203464293873511e-06, "loss": 0.2787, "step": 12550 }, { "epoch": 0.2284508630568035, "grad_norm": 2.064547300338745, "learning_rate": 9.202052631701055e-06, "loss": 0.2776, "step": 12560 }, { "epoch": 0.2286327506866258, "grad_norm": 1.7723093032836914, "learning_rate": 9.200639858305353e-06, "loss": 0.27, "step": 12570 }, { "epoch": 0.2288146383164481, "grad_norm": 1.782307744026184, "learning_rate": 9.199225974176684e-06, "loss": 0.279, "step": 12580 }, { "epoch": 0.2289965259462704, "grad_norm": 1.8778737783432007, "learning_rate": 9.197810979805713e-06, "loss": 0.2817, "step": 12590 }, { "epoch": 0.2291784135760927, "grad_norm": 1.775260329246521, "learning_rate": 9.196394875683497e-06, "loss": 0.273, "step": 12600 }, { "epoch": 0.229360301205915, "grad_norm": 1.7114179134368896, "learning_rate": 9.194977662301468e-06, "loss": 0.2698, "step": 12610 }, { "epoch": 0.2295421888357373, "grad_norm": 1.8460768461227417, "learning_rate": 9.193559340151451e-06, "loss": 0.2767, "step": 12620 }, { "epoch": 0.2297240764655596, "grad_norm": 1.759171962738037, "learning_rate": 9.192139909725653e-06, "loss": 0.2767, "step": 12630 }, { "epoch": 0.22990596409538186, "grad_norm": 1.9579685926437378, "learning_rate": 9.190719371516667e-06, "loss": 0.2783, "step": 12640 }, { "epoch": 0.23008785172520416, "grad_norm": 1.8908634185791016, "learning_rate": 9.189297726017464e-06, "loss": 0.2764, "step": 12650 }, { "epoch": 0.23026973935502645, "grad_norm": 1.8195997476577759, "learning_rate": 9.18787497372141e-06, "loss": 0.2735, "step": 12660 }, { "epoch": 0.23045162698484875, "grad_norm": 1.6506339311599731, "learning_rate": 9.186451115122244e-06, "loss": 0.2694, "step": 12670 }, { "epoch": 0.23063351461467105, "grad_norm": 1.628507375717163, "learning_rate": 9.185026150714097e-06, "loss": 0.2667, "step": 12680 }, { "epoch": 0.23081540224449335, "grad_norm": 1.875654697418213, "learning_rate": 9.183600080991482e-06, "loss": 0.2653, "step": 12690 }, { "epoch": 0.23099728987431564, "grad_norm": 2.3001599311828613, "learning_rate": 9.182172906449295e-06, "loss": 0.2711, "step": 12700 }, { "epoch": 0.23117917750413794, "grad_norm": 2.235187292098999, "learning_rate": 9.180744627582811e-06, "loss": 0.2771, "step": 12710 }, { "epoch": 0.23136106513396024, "grad_norm": 2.078065872192383, "learning_rate": 9.179315244887697e-06, "loss": 0.2715, "step": 12720 }, { "epoch": 0.23154295276378253, "grad_norm": 1.8025132417678833, "learning_rate": 9.177884758859996e-06, "loss": 0.2701, "step": 12730 }, { "epoch": 0.23172484039360483, "grad_norm": 1.950846791267395, "learning_rate": 9.176453169996135e-06, "loss": 0.2724, "step": 12740 }, { "epoch": 0.23190672802342713, "grad_norm": 1.6701687574386597, "learning_rate": 9.175020478792928e-06, "loss": 0.2714, "step": 12750 }, { "epoch": 0.23208861565324942, "grad_norm": 1.624064564704895, "learning_rate": 9.173586685747567e-06, "loss": 0.2735, "step": 12760 }, { "epoch": 0.23227050328307172, "grad_norm": 1.8761203289031982, "learning_rate": 9.172151791357627e-06, "loss": 0.2748, "step": 12770 }, { "epoch": 0.23245239091289402, "grad_norm": 1.8243238925933838, "learning_rate": 9.170715796121068e-06, "loss": 0.2718, "step": 12780 }, { "epoch": 0.23263427854271632, "grad_norm": 1.9108163118362427, "learning_rate": 9.169278700536228e-06, "loss": 0.2695, "step": 12790 }, { "epoch": 0.2328161661725386, "grad_norm": 2.108973979949951, "learning_rate": 9.16784050510183e-06, "loss": 0.2674, "step": 12800 }, { "epoch": 0.2329980538023609, "grad_norm": 1.850645661354065, "learning_rate": 9.166401210316978e-06, "loss": 0.2745, "step": 12810 }, { "epoch": 0.2331799414321832, "grad_norm": 1.5718787908554077, "learning_rate": 9.164960816681158e-06, "loss": 0.2758, "step": 12820 }, { "epoch": 0.2333618290620055, "grad_norm": 1.7751846313476562, "learning_rate": 9.163519324694234e-06, "loss": 0.273, "step": 12830 }, { "epoch": 0.2335437166918278, "grad_norm": 1.9004830121994019, "learning_rate": 9.162076734856455e-06, "loss": 0.2725, "step": 12840 }, { "epoch": 0.23372560432165007, "grad_norm": 1.730285882949829, "learning_rate": 9.16063304766845e-06, "loss": 0.2765, "step": 12850 }, { "epoch": 0.23390749195147237, "grad_norm": 2.02489972114563, "learning_rate": 9.159188263631228e-06, "loss": 0.271, "step": 12860 }, { "epoch": 0.23408937958129467, "grad_norm": 1.9920841455459595, "learning_rate": 9.15774238324618e-06, "loss": 0.2752, "step": 12870 }, { "epoch": 0.23427126721111696, "grad_norm": 1.7362638711929321, "learning_rate": 9.156295407015075e-06, "loss": 0.2616, "step": 12880 }, { "epoch": 0.23445315484093926, "grad_norm": 2.038309335708618, "learning_rate": 9.154847335440066e-06, "loss": 0.2658, "step": 12890 }, { "epoch": 0.23463504247076156, "grad_norm": 1.7441821098327637, "learning_rate": 9.153398169023682e-06, "loss": 0.2709, "step": 12900 }, { "epoch": 0.23481693010058385, "grad_norm": 1.9977641105651855, "learning_rate": 9.151947908268837e-06, "loss": 0.2708, "step": 12910 }, { "epoch": 0.23499881773040615, "grad_norm": 1.9584201574325562, "learning_rate": 9.150496553678818e-06, "loss": 0.2751, "step": 12920 }, { "epoch": 0.23518070536022845, "grad_norm": 1.947704553604126, "learning_rate": 9.149044105757295e-06, "loss": 0.2741, "step": 12930 }, { "epoch": 0.23536259299005075, "grad_norm": 1.7571563720703125, "learning_rate": 9.147590565008323e-06, "loss": 0.272, "step": 12940 }, { "epoch": 0.23554448061987304, "grad_norm": 2.026949167251587, "learning_rate": 9.146135931936326e-06, "loss": 0.2777, "step": 12950 }, { "epoch": 0.23572636824969534, "grad_norm": 1.798315405845642, "learning_rate": 9.144680207046116e-06, "loss": 0.2754, "step": 12960 }, { "epoch": 0.23590825587951764, "grad_norm": 1.7100765705108643, "learning_rate": 9.143223390842874e-06, "loss": 0.2694, "step": 12970 }, { "epoch": 0.23609014350933993, "grad_norm": 1.784213662147522, "learning_rate": 9.141765483832172e-06, "loss": 0.2735, "step": 12980 }, { "epoch": 0.23627203113916223, "grad_norm": 1.9566267728805542, "learning_rate": 9.140306486519952e-06, "loss": 0.2763, "step": 12990 }, { "epoch": 0.23645391876898453, "grad_norm": 1.7198786735534668, "learning_rate": 9.138846399412537e-06, "loss": 0.28, "step": 13000 }, { "epoch": 0.23663580639880682, "grad_norm": 1.7697440385818481, "learning_rate": 9.137385223016627e-06, "loss": 0.2599, "step": 13010 }, { "epoch": 0.23681769402862912, "grad_norm": 1.921608805656433, "learning_rate": 9.135922957839299e-06, "loss": 0.2692, "step": 13020 }, { "epoch": 0.23699958165845142, "grad_norm": 1.9239137172698975, "learning_rate": 9.134459604388015e-06, "loss": 0.2717, "step": 13030 }, { "epoch": 0.23718146928827372, "grad_norm": 1.889373779296875, "learning_rate": 9.132995163170604e-06, "loss": 0.2866, "step": 13040 }, { "epoch": 0.237363356918096, "grad_norm": 1.9002997875213623, "learning_rate": 9.13152963469528e-06, "loss": 0.2684, "step": 13050 }, { "epoch": 0.23754524454791828, "grad_norm": 2.074984073638916, "learning_rate": 9.130063019470634e-06, "loss": 0.2705, "step": 13060 }, { "epoch": 0.23772713217774058, "grad_norm": 1.8789081573486328, "learning_rate": 9.12859531800563e-06, "loss": 0.2733, "step": 13070 }, { "epoch": 0.23790901980756288, "grad_norm": 1.9259274005889893, "learning_rate": 9.12712653080961e-06, "loss": 0.2719, "step": 13080 }, { "epoch": 0.23809090743738517, "grad_norm": 1.6735773086547852, "learning_rate": 9.125656658392299e-06, "loss": 0.2725, "step": 13090 }, { "epoch": 0.23827279506720747, "grad_norm": 2.0890579223632812, "learning_rate": 9.124185701263788e-06, "loss": 0.2693, "step": 13100 }, { "epoch": 0.23845468269702977, "grad_norm": 1.9287103414535522, "learning_rate": 9.122713659934554e-06, "loss": 0.2734, "step": 13110 }, { "epoch": 0.23863657032685207, "grad_norm": 1.7783693075180054, "learning_rate": 9.121240534915446e-06, "loss": 0.2676, "step": 13120 }, { "epoch": 0.23881845795667436, "grad_norm": 1.8101009130477905, "learning_rate": 9.119766326717688e-06, "loss": 0.2735, "step": 13130 }, { "epoch": 0.23900034558649666, "grad_norm": 2.033238410949707, "learning_rate": 9.11829103585288e-06, "loss": 0.2699, "step": 13140 }, { "epoch": 0.23918223321631896, "grad_norm": 1.8020979166030884, "learning_rate": 9.116814662833002e-06, "loss": 0.2757, "step": 13150 }, { "epoch": 0.23936412084614125, "grad_norm": 2.3389394283294678, "learning_rate": 9.115337208170405e-06, "loss": 0.2691, "step": 13160 }, { "epoch": 0.23954600847596355, "grad_norm": 1.9241260290145874, "learning_rate": 9.11385867237782e-06, "loss": 0.2689, "step": 13170 }, { "epoch": 0.23972789610578585, "grad_norm": 2.414499282836914, "learning_rate": 9.112379055968346e-06, "loss": 0.28, "step": 13180 }, { "epoch": 0.23990978373560815, "grad_norm": 1.9152687788009644, "learning_rate": 9.110898359455462e-06, "loss": 0.2757, "step": 13190 }, { "epoch": 0.24009167136543044, "grad_norm": 1.7827032804489136, "learning_rate": 9.109416583353024e-06, "loss": 0.273, "step": 13200 }, { "epoch": 0.24027355899525274, "grad_norm": 1.765580415725708, "learning_rate": 9.107933728175256e-06, "loss": 0.272, "step": 13210 }, { "epoch": 0.24045544662507504, "grad_norm": 1.8418359756469727, "learning_rate": 9.106449794436762e-06, "loss": 0.2707, "step": 13220 }, { "epoch": 0.24063733425489733, "grad_norm": 2.0075550079345703, "learning_rate": 9.104964782652517e-06, "loss": 0.2751, "step": 13230 }, { "epoch": 0.24081922188471963, "grad_norm": 1.8401769399642944, "learning_rate": 9.103478693337874e-06, "loss": 0.2686, "step": 13240 }, { "epoch": 0.24100110951454193, "grad_norm": 1.919350504875183, "learning_rate": 9.101991527008557e-06, "loss": 0.2725, "step": 13250 }, { "epoch": 0.24118299714436423, "grad_norm": 1.7573981285095215, "learning_rate": 9.100503284180662e-06, "loss": 0.266, "step": 13260 }, { "epoch": 0.2413648847741865, "grad_norm": 1.7389774322509766, "learning_rate": 9.099013965370663e-06, "loss": 0.2776, "step": 13270 }, { "epoch": 0.2415467724040088, "grad_norm": 1.829001784324646, "learning_rate": 9.097523571095404e-06, "loss": 0.282, "step": 13280 }, { "epoch": 0.2417286600338311, "grad_norm": 1.7949773073196411, "learning_rate": 9.096032101872104e-06, "loss": 0.2749, "step": 13290 }, { "epoch": 0.24191054766365339, "grad_norm": 1.920790195465088, "learning_rate": 9.094539558218355e-06, "loss": 0.2656, "step": 13300 }, { "epoch": 0.24209243529347568, "grad_norm": 1.7290561199188232, "learning_rate": 9.093045940652122e-06, "loss": 0.2672, "step": 13310 }, { "epoch": 0.24227432292329798, "grad_norm": 1.8126957416534424, "learning_rate": 9.091551249691739e-06, "loss": 0.2697, "step": 13320 }, { "epoch": 0.24245621055312028, "grad_norm": 1.9017740488052368, "learning_rate": 9.090055485855918e-06, "loss": 0.2706, "step": 13330 }, { "epoch": 0.24263809818294257, "grad_norm": 2.0063436031341553, "learning_rate": 9.088558649663738e-06, "loss": 0.2685, "step": 13340 }, { "epoch": 0.24281998581276487, "grad_norm": 2.0704269409179688, "learning_rate": 9.08706074163466e-06, "loss": 0.2725, "step": 13350 }, { "epoch": 0.24300187344258717, "grad_norm": 2.1321394443511963, "learning_rate": 9.085561762288504e-06, "loss": 0.2779, "step": 13360 }, { "epoch": 0.24318376107240947, "grad_norm": 2.100520610809326, "learning_rate": 9.08406171214547e-06, "loss": 0.2718, "step": 13370 }, { "epoch": 0.24336564870223176, "grad_norm": 2.080983877182007, "learning_rate": 9.082560591726123e-06, "loss": 0.2716, "step": 13380 }, { "epoch": 0.24354753633205406, "grad_norm": 1.639525294303894, "learning_rate": 9.081058401551412e-06, "loss": 0.2781, "step": 13390 }, { "epoch": 0.24372942396187636, "grad_norm": 2.024545669555664, "learning_rate": 9.079555142142642e-06, "loss": 0.2763, "step": 13400 }, { "epoch": 0.24391131159169865, "grad_norm": 1.7714779376983643, "learning_rate": 9.0780508140215e-06, "loss": 0.2679, "step": 13410 }, { "epoch": 0.24409319922152095, "grad_norm": 1.966191053390503, "learning_rate": 9.076545417710036e-06, "loss": 0.2771, "step": 13420 }, { "epoch": 0.24427508685134325, "grad_norm": 1.670017123222351, "learning_rate": 9.07503895373068e-06, "loss": 0.2786, "step": 13430 }, { "epoch": 0.24445697448116555, "grad_norm": 1.9615848064422607, "learning_rate": 9.073531422606223e-06, "loss": 0.2826, "step": 13440 }, { "epoch": 0.24463886211098784, "grad_norm": 1.6000488996505737, "learning_rate": 9.072022824859833e-06, "loss": 0.2685, "step": 13450 }, { "epoch": 0.24482074974081014, "grad_norm": 1.7565869092941284, "learning_rate": 9.070513161015045e-06, "loss": 0.2644, "step": 13460 }, { "epoch": 0.24500263737063244, "grad_norm": 2.2196924686431885, "learning_rate": 9.069002431595765e-06, "loss": 0.2732, "step": 13470 }, { "epoch": 0.2451845250004547, "grad_norm": 1.8010586500167847, "learning_rate": 9.067490637126265e-06, "loss": 0.2663, "step": 13480 }, { "epoch": 0.245366412630277, "grad_norm": 1.8808461427688599, "learning_rate": 9.065977778131198e-06, "loss": 0.2714, "step": 13490 }, { "epoch": 0.2455483002600993, "grad_norm": 1.7517722845077515, "learning_rate": 9.06446385513557e-06, "loss": 0.2734, "step": 13500 }, { "epoch": 0.2457301878899216, "grad_norm": 2.158367156982422, "learning_rate": 9.06294886866477e-06, "loss": 0.2851, "step": 13510 }, { "epoch": 0.2459120755197439, "grad_norm": 1.7290582656860352, "learning_rate": 9.061432819244548e-06, "loss": 0.2677, "step": 13520 }, { "epoch": 0.2460939631495662, "grad_norm": 1.8488789796829224, "learning_rate": 9.059915707401027e-06, "loss": 0.2679, "step": 13530 }, { "epoch": 0.2462758507793885, "grad_norm": 1.8831959962844849, "learning_rate": 9.058397533660697e-06, "loss": 0.2753, "step": 13540 }, { "epoch": 0.24645773840921079, "grad_norm": 2.01106595993042, "learning_rate": 9.056878298550418e-06, "loss": 0.2745, "step": 13550 }, { "epoch": 0.24663962603903308, "grad_norm": 1.9238888025283813, "learning_rate": 9.055358002597415e-06, "loss": 0.2823, "step": 13560 }, { "epoch": 0.24682151366885538, "grad_norm": 1.6950056552886963, "learning_rate": 9.053836646329286e-06, "loss": 0.2739, "step": 13570 }, { "epoch": 0.24700340129867768, "grad_norm": 1.9948365688323975, "learning_rate": 9.052314230273994e-06, "loss": 0.2778, "step": 13580 }, { "epoch": 0.24718528892849997, "grad_norm": 2.005326509475708, "learning_rate": 9.050790754959867e-06, "loss": 0.2783, "step": 13590 }, { "epoch": 0.24736717655832227, "grad_norm": 1.481015920639038, "learning_rate": 9.049266220915607e-06, "loss": 0.2801, "step": 13600 }, { "epoch": 0.24754906418814457, "grad_norm": 1.9426621198654175, "learning_rate": 9.047740628670279e-06, "loss": 0.272, "step": 13610 }, { "epoch": 0.24773095181796687, "grad_norm": 1.6672000885009766, "learning_rate": 9.046213978753317e-06, "loss": 0.278, "step": 13620 }, { "epoch": 0.24791283944778916, "grad_norm": 1.8051613569259644, "learning_rate": 9.04468627169452e-06, "loss": 0.2741, "step": 13630 }, { "epoch": 0.24809472707761146, "grad_norm": 1.6642589569091797, "learning_rate": 9.043157508024057e-06, "loss": 0.2803, "step": 13640 }, { "epoch": 0.24827661470743376, "grad_norm": 1.778029203414917, "learning_rate": 9.041627688272461e-06, "loss": 0.2654, "step": 13650 }, { "epoch": 0.24845850233725605, "grad_norm": 1.7516107559204102, "learning_rate": 9.040096812970633e-06, "loss": 0.2665, "step": 13660 }, { "epoch": 0.24864038996707835, "grad_norm": 1.739487648010254, "learning_rate": 9.038564882649839e-06, "loss": 0.2717, "step": 13670 }, { "epoch": 0.24882227759690065, "grad_norm": 1.777503490447998, "learning_rate": 9.037031897841714e-06, "loss": 0.2691, "step": 13680 }, { "epoch": 0.24900416522672292, "grad_norm": 1.8802329301834106, "learning_rate": 9.035497859078254e-06, "loss": 0.2781, "step": 13690 }, { "epoch": 0.24918605285654521, "grad_norm": 1.742026448249817, "learning_rate": 9.033962766891824e-06, "loss": 0.2759, "step": 13700 }, { "epoch": 0.2493679404863675, "grad_norm": 1.3851797580718994, "learning_rate": 9.032426621815158e-06, "loss": 0.2726, "step": 13710 }, { "epoch": 0.2495498281161898, "grad_norm": 1.7978825569152832, "learning_rate": 9.030889424381347e-06, "loss": 0.2796, "step": 13720 }, { "epoch": 0.2497317157460121, "grad_norm": 1.7305738925933838, "learning_rate": 9.029351175123853e-06, "loss": 0.2724, "step": 13730 }, { "epoch": 0.2499136033758344, "grad_norm": 1.6727195978164673, "learning_rate": 9.027811874576502e-06, "loss": 0.2691, "step": 13740 }, { "epoch": 0.2500954910056567, "grad_norm": 1.7010854482650757, "learning_rate": 9.026271523273486e-06, "loss": 0.2653, "step": 13750 }, { "epoch": 0.250277378635479, "grad_norm": 1.7834386825561523, "learning_rate": 9.024730121749359e-06, "loss": 0.2691, "step": 13760 }, { "epoch": 0.2504592662653013, "grad_norm": 1.9625029563903809, "learning_rate": 9.02318767053904e-06, "loss": 0.273, "step": 13770 }, { "epoch": 0.2506411538951236, "grad_norm": 1.755977988243103, "learning_rate": 9.021644170177815e-06, "loss": 0.2697, "step": 13780 }, { "epoch": 0.2508230415249459, "grad_norm": 1.8815405368804932, "learning_rate": 9.020099621201332e-06, "loss": 0.2819, "step": 13790 }, { "epoch": 0.2510049291547682, "grad_norm": 1.609426498413086, "learning_rate": 9.0185540241456e-06, "loss": 0.2662, "step": 13800 }, { "epoch": 0.2511868167845905, "grad_norm": 1.847530484199524, "learning_rate": 9.017007379547e-06, "loss": 0.2729, "step": 13810 }, { "epoch": 0.2513687044144128, "grad_norm": 1.5812369585037231, "learning_rate": 9.015459687942265e-06, "loss": 0.2727, "step": 13820 }, { "epoch": 0.2515505920442351, "grad_norm": 1.9705426692962646, "learning_rate": 9.013910949868503e-06, "loss": 0.2767, "step": 13830 }, { "epoch": 0.2517324796740574, "grad_norm": 1.8709354400634766, "learning_rate": 9.012361165863176e-06, "loss": 0.2688, "step": 13840 }, { "epoch": 0.25191436730387967, "grad_norm": 1.8682942390441895, "learning_rate": 9.010810336464117e-06, "loss": 0.2674, "step": 13850 }, { "epoch": 0.25209625493370197, "grad_norm": 1.7306082248687744, "learning_rate": 9.009258462209513e-06, "loss": 0.2699, "step": 13860 }, { "epoch": 0.25227814256352427, "grad_norm": 2.2202603816986084, "learning_rate": 9.00770554363792e-06, "loss": 0.2736, "step": 13870 }, { "epoch": 0.25246003019334656, "grad_norm": 1.7303143739700317, "learning_rate": 9.006151581288255e-06, "loss": 0.2688, "step": 13880 }, { "epoch": 0.25264191782316886, "grad_norm": 2.185300588607788, "learning_rate": 9.004596575699796e-06, "loss": 0.2764, "step": 13890 }, { "epoch": 0.25282380545299116, "grad_norm": 1.8854581117630005, "learning_rate": 9.003040527412186e-06, "loss": 0.2718, "step": 13900 }, { "epoch": 0.25300569308281345, "grad_norm": 1.6614826917648315, "learning_rate": 9.001483436965424e-06, "loss": 0.2603, "step": 13910 }, { "epoch": 0.25318758071263575, "grad_norm": 1.9011783599853516, "learning_rate": 8.999925304899879e-06, "loss": 0.2749, "step": 13920 }, { "epoch": 0.25336946834245805, "grad_norm": 1.7847542762756348, "learning_rate": 8.99836613175627e-06, "loss": 0.2728, "step": 13930 }, { "epoch": 0.25355135597228035, "grad_norm": 1.6698753833770752, "learning_rate": 8.996805918075691e-06, "loss": 0.2677, "step": 13940 }, { "epoch": 0.25373324360210264, "grad_norm": 1.9030448198318481, "learning_rate": 8.995244664399588e-06, "loss": 0.2774, "step": 13950 }, { "epoch": 0.25391513123192494, "grad_norm": 1.80971097946167, "learning_rate": 8.99368237126977e-06, "loss": 0.269, "step": 13960 }, { "epoch": 0.25409701886174724, "grad_norm": 1.8641363382339478, "learning_rate": 8.992119039228405e-06, "loss": 0.2691, "step": 13970 }, { "epoch": 0.25427890649156953, "grad_norm": 1.7382429838180542, "learning_rate": 8.990554668818026e-06, "loss": 0.2722, "step": 13980 }, { "epoch": 0.25446079412139183, "grad_norm": 1.9463986158370972, "learning_rate": 8.98898926058152e-06, "loss": 0.2769, "step": 13990 }, { "epoch": 0.2546426817512141, "grad_norm": 1.9630047082901, "learning_rate": 8.987422815062144e-06, "loss": 0.2712, "step": 14000 }, { "epoch": 0.25482456938103637, "grad_norm": 1.8461717367172241, "learning_rate": 8.985855332803504e-06, "loss": 0.2705, "step": 14010 }, { "epoch": 0.25500645701085867, "grad_norm": 1.7012630701065063, "learning_rate": 8.984286814349573e-06, "loss": 0.2729, "step": 14020 }, { "epoch": 0.25518834464068096, "grad_norm": 2.0259408950805664, "learning_rate": 8.982717260244679e-06, "loss": 0.2758, "step": 14030 }, { "epoch": 0.25537023227050326, "grad_norm": 1.7224067449569702, "learning_rate": 8.981146671033514e-06, "loss": 0.2705, "step": 14040 }, { "epoch": 0.25555211990032556, "grad_norm": 1.9237741231918335, "learning_rate": 8.979575047261126e-06, "loss": 0.2749, "step": 14050 }, { "epoch": 0.25573400753014786, "grad_norm": 1.5050947666168213, "learning_rate": 8.978002389472923e-06, "loss": 0.2697, "step": 14060 }, { "epoch": 0.25591589515997015, "grad_norm": 1.5392029285430908, "learning_rate": 8.976428698214674e-06, "loss": 0.2704, "step": 14070 }, { "epoch": 0.25609778278979245, "grad_norm": 2.1316094398498535, "learning_rate": 8.974853974032498e-06, "loss": 0.2794, "step": 14080 }, { "epoch": 0.25627967041961475, "grad_norm": 1.8270279169082642, "learning_rate": 8.973278217472884e-06, "loss": 0.267, "step": 14090 }, { "epoch": 0.25646155804943704, "grad_norm": 1.9356127977371216, "learning_rate": 8.971701429082675e-06, "loss": 0.2809, "step": 14100 }, { "epoch": 0.25664344567925934, "grad_norm": 1.7431124448776245, "learning_rate": 8.970123609409067e-06, "loss": 0.2687, "step": 14110 }, { "epoch": 0.25682533330908164, "grad_norm": 2.049654006958008, "learning_rate": 8.968544758999621e-06, "loss": 0.2754, "step": 14120 }, { "epoch": 0.25700722093890394, "grad_norm": 1.7952990531921387, "learning_rate": 8.966964878402254e-06, "loss": 0.2712, "step": 14130 }, { "epoch": 0.25718910856872623, "grad_norm": 1.7846381664276123, "learning_rate": 8.965383968165236e-06, "loss": 0.2621, "step": 14140 }, { "epoch": 0.25737099619854853, "grad_norm": 2.24041485786438, "learning_rate": 8.9638020288372e-06, "loss": 0.2743, "step": 14150 }, { "epoch": 0.2575528838283708, "grad_norm": 2.5044071674346924, "learning_rate": 8.962219060967133e-06, "loss": 0.2723, "step": 14160 }, { "epoch": 0.2577347714581931, "grad_norm": 1.719331979751587, "learning_rate": 8.960635065104381e-06, "loss": 0.2728, "step": 14170 }, { "epoch": 0.2579166590880154, "grad_norm": 2.0694642066955566, "learning_rate": 8.959050041798642e-06, "loss": 0.2736, "step": 14180 }, { "epoch": 0.2580985467178377, "grad_norm": 1.9994535446166992, "learning_rate": 8.957463991599978e-06, "loss": 0.2721, "step": 14190 }, { "epoch": 0.25828043434766, "grad_norm": 2.0530765056610107, "learning_rate": 8.955876915058801e-06, "loss": 0.2735, "step": 14200 }, { "epoch": 0.2584623219774823, "grad_norm": 1.8642618656158447, "learning_rate": 8.954288812725882e-06, "loss": 0.2727, "step": 14210 }, { "epoch": 0.2586442096073046, "grad_norm": 1.7390735149383545, "learning_rate": 8.952699685152348e-06, "loss": 0.2761, "step": 14220 }, { "epoch": 0.2588260972371269, "grad_norm": 1.8566875457763672, "learning_rate": 8.95110953288968e-06, "loss": 0.269, "step": 14230 }, { "epoch": 0.2590079848669492, "grad_norm": 1.9207974672317505, "learning_rate": 8.949518356489719e-06, "loss": 0.2652, "step": 14240 }, { "epoch": 0.2591898724967715, "grad_norm": 1.909471035003662, "learning_rate": 8.947926156504654e-06, "loss": 0.281, "step": 14250 }, { "epoch": 0.2593717601265938, "grad_norm": 1.9522607326507568, "learning_rate": 8.946332933487037e-06, "loss": 0.2769, "step": 14260 }, { "epoch": 0.2595536477564161, "grad_norm": 1.6393100023269653, "learning_rate": 8.944738687989772e-06, "loss": 0.274, "step": 14270 }, { "epoch": 0.2597355353862384, "grad_norm": 1.9315470457077026, "learning_rate": 8.943143420566116e-06, "loss": 0.273, "step": 14280 }, { "epoch": 0.2599174230160607, "grad_norm": 1.9166347980499268, "learning_rate": 8.94154713176968e-06, "loss": 0.2772, "step": 14290 }, { "epoch": 0.260099310645883, "grad_norm": 2.042541742324829, "learning_rate": 8.939949822154434e-06, "loss": 0.2753, "step": 14300 }, { "epoch": 0.2602811982757053, "grad_norm": 1.828870177268982, "learning_rate": 8.9383514922747e-06, "loss": 0.2659, "step": 14310 }, { "epoch": 0.2604630859055276, "grad_norm": 1.8214197158813477, "learning_rate": 8.936752142685156e-06, "loss": 0.2757, "step": 14320 }, { "epoch": 0.2606449735353499, "grad_norm": 1.808260202407837, "learning_rate": 8.935151773940826e-06, "loss": 0.2667, "step": 14330 }, { "epoch": 0.2608268611651722, "grad_norm": 1.7024041414260864, "learning_rate": 8.933550386597096e-06, "loss": 0.2738, "step": 14340 }, { "epoch": 0.26100874879499447, "grad_norm": 1.8048006296157837, "learning_rate": 8.931947981209705e-06, "loss": 0.2629, "step": 14350 }, { "epoch": 0.26119063642481677, "grad_norm": 1.7572041749954224, "learning_rate": 8.930344558334742e-06, "loss": 0.2648, "step": 14360 }, { "epoch": 0.26137252405463907, "grad_norm": 1.674434781074524, "learning_rate": 8.92874011852865e-06, "loss": 0.2698, "step": 14370 }, { "epoch": 0.26155441168446136, "grad_norm": 1.8418649435043335, "learning_rate": 8.927134662348224e-06, "loss": 0.2663, "step": 14380 }, { "epoch": 0.26173629931428366, "grad_norm": 1.6478322744369507, "learning_rate": 8.925528190350614e-06, "loss": 0.2688, "step": 14390 }, { "epoch": 0.26191818694410596, "grad_norm": 1.6001211404800415, "learning_rate": 8.923920703093322e-06, "loss": 0.2703, "step": 14400 }, { "epoch": 0.26210007457392825, "grad_norm": 1.9229553937911987, "learning_rate": 8.922312201134203e-06, "loss": 0.2754, "step": 14410 }, { "epoch": 0.2622819622037505, "grad_norm": 1.7133206129074097, "learning_rate": 8.92070268503146e-06, "loss": 0.2741, "step": 14420 }, { "epoch": 0.2624638498335728, "grad_norm": 1.9006825685501099, "learning_rate": 8.919092155343651e-06, "loss": 0.2656, "step": 14430 }, { "epoch": 0.2626457374633951, "grad_norm": 2.004538059234619, "learning_rate": 8.917480612629687e-06, "loss": 0.2641, "step": 14440 }, { "epoch": 0.2628276250932174, "grad_norm": 1.8331855535507202, "learning_rate": 8.91586805744883e-06, "loss": 0.2759, "step": 14450 }, { "epoch": 0.2630095127230397, "grad_norm": 1.705862283706665, "learning_rate": 8.914254490360693e-06, "loss": 0.2806, "step": 14460 }, { "epoch": 0.263191400352862, "grad_norm": 1.888244867324829, "learning_rate": 8.912639911925236e-06, "loss": 0.2753, "step": 14470 }, { "epoch": 0.2633732879826843, "grad_norm": 1.6042563915252686, "learning_rate": 8.911024322702777e-06, "loss": 0.2755, "step": 14480 }, { "epoch": 0.2635551756125066, "grad_norm": 1.7935091257095337, "learning_rate": 8.90940772325398e-06, "loss": 0.271, "step": 14490 }, { "epoch": 0.2637370632423289, "grad_norm": 2.147920608520508, "learning_rate": 8.907790114139862e-06, "loss": 0.2713, "step": 14500 }, { "epoch": 0.26391895087215117, "grad_norm": 1.6838860511779785, "learning_rate": 8.906171495921787e-06, "loss": 0.2666, "step": 14510 }, { "epoch": 0.26410083850197347, "grad_norm": 2.0382916927337646, "learning_rate": 8.904551869161473e-06, "loss": 0.2831, "step": 14520 }, { "epoch": 0.26428272613179576, "grad_norm": 2.114532232284546, "learning_rate": 8.90293123442099e-06, "loss": 0.2752, "step": 14530 }, { "epoch": 0.26446461376161806, "grad_norm": 1.9473316669464111, "learning_rate": 8.90130959226275e-06, "loss": 0.2662, "step": 14540 }, { "epoch": 0.26464650139144036, "grad_norm": 1.8372597694396973, "learning_rate": 8.899686943249521e-06, "loss": 0.2703, "step": 14550 }, { "epoch": 0.26482838902126266, "grad_norm": 1.7322118282318115, "learning_rate": 8.898063287944418e-06, "loss": 0.2725, "step": 14560 }, { "epoch": 0.26501027665108495, "grad_norm": 1.7019495964050293, "learning_rate": 8.896438626910905e-06, "loss": 0.2826, "step": 14570 }, { "epoch": 0.26519216428090725, "grad_norm": 1.7976843118667603, "learning_rate": 8.894812960712798e-06, "loss": 0.268, "step": 14580 }, { "epoch": 0.26537405191072955, "grad_norm": 1.7276408672332764, "learning_rate": 8.893186289914255e-06, "loss": 0.2742, "step": 14590 }, { "epoch": 0.26555593954055184, "grad_norm": 1.7939422130584717, "learning_rate": 8.891558615079791e-06, "loss": 0.2719, "step": 14600 }, { "epoch": 0.26573782717037414, "grad_norm": 1.7831395864486694, "learning_rate": 8.889929936774266e-06, "loss": 0.2762, "step": 14610 }, { "epoch": 0.26591971480019644, "grad_norm": 1.759138822555542, "learning_rate": 8.888300255562885e-06, "loss": 0.2732, "step": 14620 }, { "epoch": 0.26610160243001874, "grad_norm": 1.6650980710983276, "learning_rate": 8.886669572011208e-06, "loss": 0.2637, "step": 14630 }, { "epoch": 0.26628349005984103, "grad_norm": 1.883256196975708, "learning_rate": 8.885037886685135e-06, "loss": 0.2799, "step": 14640 }, { "epoch": 0.26646537768966333, "grad_norm": 1.8379348516464233, "learning_rate": 8.883405200150916e-06, "loss": 0.2697, "step": 14650 }, { "epoch": 0.2666472653194856, "grad_norm": 1.8757264614105225, "learning_rate": 8.881771512975153e-06, "loss": 0.2743, "step": 14660 }, { "epoch": 0.2668291529493079, "grad_norm": 1.8075147867202759, "learning_rate": 8.880136825724792e-06, "loss": 0.2636, "step": 14670 }, { "epoch": 0.2670110405791302, "grad_norm": 1.9301072359085083, "learning_rate": 8.878501138967126e-06, "loss": 0.2686, "step": 14680 }, { "epoch": 0.2671929282089525, "grad_norm": 1.9086004495620728, "learning_rate": 8.876864453269793e-06, "loss": 0.2743, "step": 14690 }, { "epoch": 0.2673748158387748, "grad_norm": 1.5953583717346191, "learning_rate": 8.875226769200783e-06, "loss": 0.2605, "step": 14700 }, { "epoch": 0.2675567034685971, "grad_norm": 1.7527942657470703, "learning_rate": 8.873588087328424e-06, "loss": 0.268, "step": 14710 }, { "epoch": 0.2677385910984194, "grad_norm": 1.92625892162323, "learning_rate": 8.871948408221402e-06, "loss": 0.2682, "step": 14720 }, { "epoch": 0.2679204787282417, "grad_norm": 1.9118071794509888, "learning_rate": 8.870307732448737e-06, "loss": 0.2678, "step": 14730 }, { "epoch": 0.268102366358064, "grad_norm": 1.7304034233093262, "learning_rate": 8.868666060579802e-06, "loss": 0.2678, "step": 14740 }, { "epoch": 0.2682842539878863, "grad_norm": 2.1667397022247314, "learning_rate": 8.867023393184315e-06, "loss": 0.2731, "step": 14750 }, { "epoch": 0.2684661416177086, "grad_norm": 1.6525498628616333, "learning_rate": 8.86537973083234e-06, "loss": 0.2749, "step": 14760 }, { "epoch": 0.2686480292475309, "grad_norm": 1.7403807640075684, "learning_rate": 8.863735074094282e-06, "loss": 0.2736, "step": 14770 }, { "epoch": 0.2688299168773532, "grad_norm": 1.9773097038269043, "learning_rate": 8.862089423540893e-06, "loss": 0.2681, "step": 14780 }, { "epoch": 0.2690118045071755, "grad_norm": 2.07108736038208, "learning_rate": 8.860442779743273e-06, "loss": 0.2721, "step": 14790 }, { "epoch": 0.2691936921369978, "grad_norm": 1.787192463874817, "learning_rate": 8.858795143272865e-06, "loss": 0.268, "step": 14800 }, { "epoch": 0.2693755797668201, "grad_norm": 1.9382777214050293, "learning_rate": 8.857146514701454e-06, "loss": 0.2755, "step": 14810 }, { "epoch": 0.2695574673966424, "grad_norm": 1.8266880512237549, "learning_rate": 8.855496894601173e-06, "loss": 0.276, "step": 14820 }, { "epoch": 0.2697393550264647, "grad_norm": 1.8361129760742188, "learning_rate": 8.853846283544497e-06, "loss": 0.2697, "step": 14830 }, { "epoch": 0.2699212426562869, "grad_norm": 1.8787060976028442, "learning_rate": 8.852194682104243e-06, "loss": 0.27, "step": 14840 }, { "epoch": 0.2701031302861092, "grad_norm": 1.548673152923584, "learning_rate": 8.850542090853576e-06, "loss": 0.2773, "step": 14850 }, { "epoch": 0.2702850179159315, "grad_norm": 2.086916208267212, "learning_rate": 8.848888510366002e-06, "loss": 0.2749, "step": 14860 }, { "epoch": 0.2704669055457538, "grad_norm": 1.7168408632278442, "learning_rate": 8.847233941215373e-06, "loss": 0.2692, "step": 14870 }, { "epoch": 0.2706487931755761, "grad_norm": 2.0008132457733154, "learning_rate": 8.84557838397588e-06, "loss": 0.2754, "step": 14880 }, { "epoch": 0.2708306808053984, "grad_norm": 1.6955533027648926, "learning_rate": 8.843921839222054e-06, "loss": 0.2698, "step": 14890 }, { "epoch": 0.2710125684352207, "grad_norm": 2.0896108150482178, "learning_rate": 8.842264307528781e-06, "loss": 0.272, "step": 14900 }, { "epoch": 0.271194456065043, "grad_norm": 1.750210165977478, "learning_rate": 8.84060578947128e-06, "loss": 0.2692, "step": 14910 }, { "epoch": 0.2713763436948653, "grad_norm": 1.8302279710769653, "learning_rate": 8.838946285625112e-06, "loss": 0.2698, "step": 14920 }, { "epoch": 0.2715582313246876, "grad_norm": 1.8876447677612305, "learning_rate": 8.837285796566184e-06, "loss": 0.2694, "step": 14930 }, { "epoch": 0.2717401189545099, "grad_norm": 1.6791057586669922, "learning_rate": 8.835624322870741e-06, "loss": 0.2695, "step": 14940 }, { "epoch": 0.2719220065843322, "grad_norm": 1.852291226387024, "learning_rate": 8.833961865115375e-06, "loss": 0.2761, "step": 14950 }, { "epoch": 0.2721038942141545, "grad_norm": 1.549614429473877, "learning_rate": 8.832298423877016e-06, "loss": 0.2725, "step": 14960 }, { "epoch": 0.2722857818439768, "grad_norm": 1.8144278526306152, "learning_rate": 8.830633999732933e-06, "loss": 0.2723, "step": 14970 }, { "epoch": 0.2724676694737991, "grad_norm": 1.760035514831543, "learning_rate": 8.828968593260743e-06, "loss": 0.2761, "step": 14980 }, { "epoch": 0.2726495571036214, "grad_norm": 1.8282285928726196, "learning_rate": 8.827302205038396e-06, "loss": 0.2656, "step": 14990 }, { "epoch": 0.2728314447334437, "grad_norm": 1.7346745729446411, "learning_rate": 8.825634835644188e-06, "loss": 0.2768, "step": 15000 }, { "epoch": 0.27301333236326597, "grad_norm": 1.837890863418579, "learning_rate": 8.823966485656755e-06, "loss": 0.2697, "step": 15010 }, { "epoch": 0.27319521999308827, "grad_norm": 1.7330036163330078, "learning_rate": 8.82229715565507e-06, "loss": 0.2729, "step": 15020 }, { "epoch": 0.27337710762291056, "grad_norm": 1.633461356163025, "learning_rate": 8.820626846218452e-06, "loss": 0.2707, "step": 15030 }, { "epoch": 0.27355899525273286, "grad_norm": 1.8497449159622192, "learning_rate": 8.81895555792655e-06, "loss": 0.269, "step": 15040 }, { "epoch": 0.27374088288255516, "grad_norm": 1.8120371103286743, "learning_rate": 8.817283291359367e-06, "loss": 0.2651, "step": 15050 }, { "epoch": 0.27392277051237746, "grad_norm": 1.792649507522583, "learning_rate": 8.815610047097232e-06, "loss": 0.2766, "step": 15060 }, { "epoch": 0.27410465814219975, "grad_norm": 1.6880427598953247, "learning_rate": 8.813935825720819e-06, "loss": 0.2732, "step": 15070 }, { "epoch": 0.27428654577202205, "grad_norm": 1.6092475652694702, "learning_rate": 8.812260627811142e-06, "loss": 0.2691, "step": 15080 }, { "epoch": 0.27446843340184435, "grad_norm": 1.9976555109024048, "learning_rate": 8.810584453949554e-06, "loss": 0.2696, "step": 15090 }, { "epoch": 0.27465032103166664, "grad_norm": 1.885633111000061, "learning_rate": 8.808907304717743e-06, "loss": 0.2663, "step": 15100 }, { "epoch": 0.27483220866148894, "grad_norm": 1.6500941514968872, "learning_rate": 8.807229180697739e-06, "loss": 0.2667, "step": 15110 }, { "epoch": 0.27501409629131124, "grad_norm": 1.9848288297653198, "learning_rate": 8.80555008247191e-06, "loss": 0.2728, "step": 15120 }, { "epoch": 0.27519598392113354, "grad_norm": 1.936705231666565, "learning_rate": 8.80387001062296e-06, "loss": 0.2763, "step": 15130 }, { "epoch": 0.27537787155095583, "grad_norm": 1.8035820722579956, "learning_rate": 8.802188965733933e-06, "loss": 0.2773, "step": 15140 }, { "epoch": 0.27555975918077813, "grad_norm": 1.8016527891159058, "learning_rate": 8.80050694838821e-06, "loss": 0.2743, "step": 15150 }, { "epoch": 0.2757416468106004, "grad_norm": 2.0319366455078125, "learning_rate": 8.798823959169509e-06, "loss": 0.2734, "step": 15160 }, { "epoch": 0.2759235344404227, "grad_norm": 1.5981788635253906, "learning_rate": 8.797139998661885e-06, "loss": 0.2638, "step": 15170 }, { "epoch": 0.276105422070245, "grad_norm": 1.5374460220336914, "learning_rate": 8.795455067449731e-06, "loss": 0.2715, "step": 15180 }, { "epoch": 0.2762873097000673, "grad_norm": 1.8707411289215088, "learning_rate": 8.793769166117776e-06, "loss": 0.2776, "step": 15190 }, { "epoch": 0.2764691973298896, "grad_norm": 2.092052698135376, "learning_rate": 8.792082295251088e-06, "loss": 0.2668, "step": 15200 }, { "epoch": 0.2766510849597119, "grad_norm": 1.7315542697906494, "learning_rate": 8.790394455435067e-06, "loss": 0.2681, "step": 15210 }, { "epoch": 0.2768329725895342, "grad_norm": 1.853711485862732, "learning_rate": 8.788705647255454e-06, "loss": 0.2706, "step": 15220 }, { "epoch": 0.2770148602193565, "grad_norm": 2.615706443786621, "learning_rate": 8.787015871298324e-06, "loss": 0.269, "step": 15230 }, { "epoch": 0.2771967478491788, "grad_norm": 1.677514910697937, "learning_rate": 8.785325128150087e-06, "loss": 0.2722, "step": 15240 }, { "epoch": 0.27737863547900105, "grad_norm": 1.887790322303772, "learning_rate": 8.78363341839749e-06, "loss": 0.2666, "step": 15250 }, { "epoch": 0.27756052310882334, "grad_norm": 1.652937889099121, "learning_rate": 8.781940742627613e-06, "loss": 0.2724, "step": 15260 }, { "epoch": 0.27774241073864564, "grad_norm": 1.7754359245300293, "learning_rate": 8.780247101427875e-06, "loss": 0.2788, "step": 15270 }, { "epoch": 0.27792429836846794, "grad_norm": 1.893696904182434, "learning_rate": 8.77855249538603e-06, "loss": 0.276, "step": 15280 }, { "epoch": 0.27810618599829023, "grad_norm": 1.7787847518920898, "learning_rate": 8.776856925090163e-06, "loss": 0.2672, "step": 15290 }, { "epoch": 0.27828807362811253, "grad_norm": 2.6819334030151367, "learning_rate": 8.775160391128693e-06, "loss": 0.2719, "step": 15300 }, { "epoch": 0.27846996125793483, "grad_norm": 2.215076208114624, "learning_rate": 8.77346289409038e-06, "loss": 0.2713, "step": 15310 }, { "epoch": 0.2786518488877571, "grad_norm": 2.1042890548706055, "learning_rate": 8.771764434564314e-06, "loss": 0.2687, "step": 15320 }, { "epoch": 0.2788337365175794, "grad_norm": 2.020458698272705, "learning_rate": 8.770065013139917e-06, "loss": 0.2642, "step": 15330 }, { "epoch": 0.2790156241474017, "grad_norm": 1.8389326333999634, "learning_rate": 8.76836463040695e-06, "loss": 0.2758, "step": 15340 }, { "epoch": 0.279197511777224, "grad_norm": 1.834970235824585, "learning_rate": 8.766663286955502e-06, "loss": 0.2738, "step": 15350 }, { "epoch": 0.2793793994070463, "grad_norm": 1.7737114429473877, "learning_rate": 8.764960983376001e-06, "loss": 0.2729, "step": 15360 }, { "epoch": 0.2795612870368686, "grad_norm": 1.867414951324463, "learning_rate": 8.763257720259202e-06, "loss": 0.2732, "step": 15370 }, { "epoch": 0.2797431746666909, "grad_norm": 1.7664620876312256, "learning_rate": 8.761553498196198e-06, "loss": 0.2693, "step": 15380 }, { "epoch": 0.2799250622965132, "grad_norm": 1.6396833658218384, "learning_rate": 8.759848317778415e-06, "loss": 0.2663, "step": 15390 }, { "epoch": 0.2801069499263355, "grad_norm": 1.918186902999878, "learning_rate": 8.758142179597607e-06, "loss": 0.2711, "step": 15400 }, { "epoch": 0.2802888375561578, "grad_norm": 1.597275733947754, "learning_rate": 8.756435084245862e-06, "loss": 0.2673, "step": 15410 }, { "epoch": 0.2804707251859801, "grad_norm": 1.6998951435089111, "learning_rate": 8.754727032315606e-06, "loss": 0.2693, "step": 15420 }, { "epoch": 0.2806526128158024, "grad_norm": 2.0231387615203857, "learning_rate": 8.753018024399587e-06, "loss": 0.2643, "step": 15430 }, { "epoch": 0.2808345004456247, "grad_norm": 1.5834314823150635, "learning_rate": 8.751308061090895e-06, "loss": 0.2722, "step": 15440 }, { "epoch": 0.281016388075447, "grad_norm": 1.5629477500915527, "learning_rate": 8.749597142982943e-06, "loss": 0.2693, "step": 15450 }, { "epoch": 0.2811982757052693, "grad_norm": 1.7561004161834717, "learning_rate": 8.74788527066948e-06, "loss": 0.2702, "step": 15460 }, { "epoch": 0.2813801633350916, "grad_norm": 2.082118272781372, "learning_rate": 8.746172444744586e-06, "loss": 0.2717, "step": 15470 }, { "epoch": 0.2815620509649139, "grad_norm": 1.806990146636963, "learning_rate": 8.744458665802668e-06, "loss": 0.2681, "step": 15480 }, { "epoch": 0.2817439385947362, "grad_norm": 1.8131160736083984, "learning_rate": 8.74274393443847e-06, "loss": 0.2696, "step": 15490 }, { "epoch": 0.2819258262245585, "grad_norm": 1.6759146451950073, "learning_rate": 8.741028251247058e-06, "loss": 0.2728, "step": 15500 }, { "epoch": 0.28210771385438077, "grad_norm": 1.8620915412902832, "learning_rate": 8.73931161682384e-06, "loss": 0.2625, "step": 15510 }, { "epoch": 0.28228960148420307, "grad_norm": 1.6715317964553833, "learning_rate": 8.737594031764546e-06, "loss": 0.275, "step": 15520 }, { "epoch": 0.28247148911402536, "grad_norm": 1.6930346488952637, "learning_rate": 8.735875496665237e-06, "loss": 0.2719, "step": 15530 }, { "epoch": 0.28265337674384766, "grad_norm": 1.8722572326660156, "learning_rate": 8.734156012122302e-06, "loss": 0.2738, "step": 15540 }, { "epoch": 0.28283526437366996, "grad_norm": 1.8941837549209595, "learning_rate": 8.732435578732465e-06, "loss": 0.272, "step": 15550 }, { "epoch": 0.28301715200349226, "grad_norm": 1.7664798498153687, "learning_rate": 8.730714197092774e-06, "loss": 0.2681, "step": 15560 }, { "epoch": 0.28319903963331455, "grad_norm": 1.8187108039855957, "learning_rate": 8.728991867800609e-06, "loss": 0.2765, "step": 15570 }, { "epoch": 0.28338092726313685, "grad_norm": 1.9802987575531006, "learning_rate": 8.727268591453675e-06, "loss": 0.2723, "step": 15580 }, { "epoch": 0.28356281489295915, "grad_norm": 1.7967326641082764, "learning_rate": 8.725544368650015e-06, "loss": 0.2722, "step": 15590 }, { "epoch": 0.28374470252278144, "grad_norm": 1.858877182006836, "learning_rate": 8.723819199987987e-06, "loss": 0.2652, "step": 15600 }, { "epoch": 0.28392659015260374, "grad_norm": 2.2118031978607178, "learning_rate": 8.722093086066288e-06, "loss": 0.2723, "step": 15610 }, { "epoch": 0.28410847778242604, "grad_norm": 1.651360273361206, "learning_rate": 8.72036602748394e-06, "loss": 0.2668, "step": 15620 }, { "epoch": 0.28429036541224834, "grad_norm": 1.791396141052246, "learning_rate": 8.718638024840292e-06, "loss": 0.2664, "step": 15630 }, { "epoch": 0.28447225304207063, "grad_norm": 1.7697302103042603, "learning_rate": 8.716909078735017e-06, "loss": 0.2719, "step": 15640 }, { "epoch": 0.28465414067189293, "grad_norm": 2.087526559829712, "learning_rate": 8.715179189768123e-06, "loss": 0.2714, "step": 15650 }, { "epoch": 0.2848360283017152, "grad_norm": 1.8158915042877197, "learning_rate": 8.713448358539943e-06, "loss": 0.2741, "step": 15660 }, { "epoch": 0.28501791593153747, "grad_norm": 1.8140453100204468, "learning_rate": 8.711716585651131e-06, "loss": 0.2735, "step": 15670 }, { "epoch": 0.28519980356135977, "grad_norm": 1.9313212633132935, "learning_rate": 8.709983871702675e-06, "loss": 0.2734, "step": 15680 }, { "epoch": 0.28538169119118206, "grad_norm": 1.9498099088668823, "learning_rate": 8.708250217295885e-06, "loss": 0.2653, "step": 15690 }, { "epoch": 0.28556357882100436, "grad_norm": 1.7619826793670654, "learning_rate": 8.7065156230324e-06, "loss": 0.2767, "step": 15700 }, { "epoch": 0.28574546645082666, "grad_norm": 1.8994309902191162, "learning_rate": 8.704780089514186e-06, "loss": 0.2783, "step": 15710 }, { "epoch": 0.28592735408064895, "grad_norm": 1.8281359672546387, "learning_rate": 8.703043617343533e-06, "loss": 0.2693, "step": 15720 }, { "epoch": 0.28610924171047125, "grad_norm": 2.0301361083984375, "learning_rate": 8.701306207123055e-06, "loss": 0.2693, "step": 15730 }, { "epoch": 0.28629112934029355, "grad_norm": 1.691141963005066, "learning_rate": 8.699567859455692e-06, "loss": 0.2675, "step": 15740 }, { "epoch": 0.28647301697011585, "grad_norm": 1.5921828746795654, "learning_rate": 8.697828574944716e-06, "loss": 0.2673, "step": 15750 }, { "epoch": 0.28665490459993814, "grad_norm": 1.8488514423370361, "learning_rate": 8.696088354193715e-06, "loss": 0.2622, "step": 15760 }, { "epoch": 0.28683679222976044, "grad_norm": 1.661576271057129, "learning_rate": 8.694347197806607e-06, "loss": 0.271, "step": 15770 }, { "epoch": 0.28701867985958274, "grad_norm": 1.7663037776947021, "learning_rate": 8.692605106387634e-06, "loss": 0.2605, "step": 15780 }, { "epoch": 0.28720056748940503, "grad_norm": 1.6850883960723877, "learning_rate": 8.690862080541362e-06, "loss": 0.2654, "step": 15790 }, { "epoch": 0.28738245511922733, "grad_norm": 1.8329260349273682, "learning_rate": 8.689118120872681e-06, "loss": 0.2758, "step": 15800 }, { "epoch": 0.28756434274904963, "grad_norm": 1.8841594457626343, "learning_rate": 8.687373227986807e-06, "loss": 0.279, "step": 15810 }, { "epoch": 0.2877462303788719, "grad_norm": 1.9412223100662231, "learning_rate": 8.685627402489276e-06, "loss": 0.2772, "step": 15820 }, { "epoch": 0.2879281180086942, "grad_norm": 1.7443537712097168, "learning_rate": 8.683880644985952e-06, "loss": 0.2778, "step": 15830 }, { "epoch": 0.2881100056385165, "grad_norm": 1.7777512073516846, "learning_rate": 8.682132956083019e-06, "loss": 0.2693, "step": 15840 }, { "epoch": 0.2882918932683388, "grad_norm": 1.9767982959747314, "learning_rate": 8.680384336386987e-06, "loss": 0.267, "step": 15850 }, { "epoch": 0.2884737808981611, "grad_norm": 1.816469669342041, "learning_rate": 8.678634786504684e-06, "loss": 0.2649, "step": 15860 }, { "epoch": 0.2886556685279834, "grad_norm": 1.9757472276687622, "learning_rate": 8.676884307043271e-06, "loss": 0.2694, "step": 15870 }, { "epoch": 0.2888375561578057, "grad_norm": 1.6784725189208984, "learning_rate": 8.675132898610218e-06, "loss": 0.271, "step": 15880 }, { "epoch": 0.289019443787628, "grad_norm": 1.8288112878799438, "learning_rate": 8.67338056181333e-06, "loss": 0.2666, "step": 15890 }, { "epoch": 0.2892013314174503, "grad_norm": 1.741296648979187, "learning_rate": 8.671627297260725e-06, "loss": 0.2663, "step": 15900 }, { "epoch": 0.2893832190472726, "grad_norm": 2.2620065212249756, "learning_rate": 8.669873105560848e-06, "loss": 0.2739, "step": 15910 }, { "epoch": 0.2895651066770949, "grad_norm": 1.7676191329956055, "learning_rate": 8.668117987322463e-06, "loss": 0.2735, "step": 15920 }, { "epoch": 0.2897469943069172, "grad_norm": 1.7007189989089966, "learning_rate": 8.66636194315466e-06, "loss": 0.2626, "step": 15930 }, { "epoch": 0.2899288819367395, "grad_norm": 1.763624668121338, "learning_rate": 8.664604973666845e-06, "loss": 0.2634, "step": 15940 }, { "epoch": 0.2901107695665618, "grad_norm": 1.6467622518539429, "learning_rate": 8.662847079468746e-06, "loss": 0.2757, "step": 15950 }, { "epoch": 0.2902926571963841, "grad_norm": 1.9086846113204956, "learning_rate": 8.661088261170415e-06, "loss": 0.27, "step": 15960 }, { "epoch": 0.2904745448262064, "grad_norm": 1.8461412191390991, "learning_rate": 8.659328519382225e-06, "loss": 0.2695, "step": 15970 }, { "epoch": 0.2906564324560287, "grad_norm": 1.880622148513794, "learning_rate": 8.657567854714861e-06, "loss": 0.2735, "step": 15980 }, { "epoch": 0.290838320085851, "grad_norm": 1.8540534973144531, "learning_rate": 8.655806267779342e-06, "loss": 0.2737, "step": 15990 }, { "epoch": 0.2910202077156733, "grad_norm": 1.7510467767715454, "learning_rate": 8.654043759186994e-06, "loss": 0.271, "step": 16000 }, { "epoch": 0.29120209534549557, "grad_norm": 2.0472488403320312, "learning_rate": 8.652280329549475e-06, "loss": 0.2764, "step": 16010 }, { "epoch": 0.29138398297531787, "grad_norm": 1.8119693994522095, "learning_rate": 8.650515979478748e-06, "loss": 0.2705, "step": 16020 }, { "epoch": 0.29156587060514016, "grad_norm": 1.5963205099105835, "learning_rate": 8.648750709587109e-06, "loss": 0.2644, "step": 16030 }, { "epoch": 0.29174775823496246, "grad_norm": 1.8231958150863647, "learning_rate": 8.646984520487167e-06, "loss": 0.2774, "step": 16040 }, { "epoch": 0.29192964586478476, "grad_norm": 1.8081388473510742, "learning_rate": 8.645217412791852e-06, "loss": 0.2693, "step": 16050 }, { "epoch": 0.29211153349460706, "grad_norm": 1.6767714023590088, "learning_rate": 8.643449387114408e-06, "loss": 0.2743, "step": 16060 }, { "epoch": 0.29229342112442935, "grad_norm": 1.7931911945343018, "learning_rate": 8.641680444068403e-06, "loss": 0.276, "step": 16070 }, { "epoch": 0.29247530875425165, "grad_norm": 1.7034413814544678, "learning_rate": 8.639910584267724e-06, "loss": 0.2639, "step": 16080 }, { "epoch": 0.2926571963840739, "grad_norm": 1.7084921598434448, "learning_rate": 8.63813980832657e-06, "loss": 0.2656, "step": 16090 }, { "epoch": 0.2928390840138962, "grad_norm": 1.9014605283737183, "learning_rate": 8.636368116859462e-06, "loss": 0.2751, "step": 16100 }, { "epoch": 0.2930209716437185, "grad_norm": 1.8590055704116821, "learning_rate": 8.634595510481241e-06, "loss": 0.2722, "step": 16110 }, { "epoch": 0.2932028592735408, "grad_norm": 1.7905917167663574, "learning_rate": 8.632821989807063e-06, "loss": 0.2693, "step": 16120 }, { "epoch": 0.2933847469033631, "grad_norm": 1.7895112037658691, "learning_rate": 8.631047555452397e-06, "loss": 0.2628, "step": 16130 }, { "epoch": 0.2935666345331854, "grad_norm": 1.641883134841919, "learning_rate": 8.629272208033037e-06, "loss": 0.2705, "step": 16140 }, { "epoch": 0.2937485221630077, "grad_norm": 1.6002143621444702, "learning_rate": 8.62749594816509e-06, "loss": 0.2708, "step": 16150 }, { "epoch": 0.29393040979282997, "grad_norm": 2.1332664489746094, "learning_rate": 8.625718776464976e-06, "loss": 0.268, "step": 16160 }, { "epoch": 0.29411229742265227, "grad_norm": 1.8993816375732422, "learning_rate": 8.623940693549439e-06, "loss": 0.2772, "step": 16170 }, { "epoch": 0.29429418505247457, "grad_norm": 1.9162184000015259, "learning_rate": 8.622161700035535e-06, "loss": 0.2726, "step": 16180 }, { "epoch": 0.29447607268229686, "grad_norm": 1.6448802947998047, "learning_rate": 8.620381796540633e-06, "loss": 0.2788, "step": 16190 }, { "epoch": 0.29465796031211916, "grad_norm": 1.8148847818374634, "learning_rate": 8.618600983682425e-06, "loss": 0.2759, "step": 16200 }, { "epoch": 0.29483984794194146, "grad_norm": 1.63584303855896, "learning_rate": 8.616819262078915e-06, "loss": 0.2661, "step": 16210 }, { "epoch": 0.29502173557176375, "grad_norm": 1.7894837856292725, "learning_rate": 8.61503663234842e-06, "loss": 0.2753, "step": 16220 }, { "epoch": 0.29520362320158605, "grad_norm": 1.8079034090042114, "learning_rate": 8.613253095109575e-06, "loss": 0.269, "step": 16230 }, { "epoch": 0.29538551083140835, "grad_norm": 1.783779501914978, "learning_rate": 8.611468650981327e-06, "loss": 0.268, "step": 16240 }, { "epoch": 0.29556739846123065, "grad_norm": 1.7352670431137085, "learning_rate": 8.609683300582945e-06, "loss": 0.2696, "step": 16250 }, { "epoch": 0.29574928609105294, "grad_norm": 1.745460033416748, "learning_rate": 8.607897044534004e-06, "loss": 0.278, "step": 16260 }, { "epoch": 0.29593117372087524, "grad_norm": 2.104820728302002, "learning_rate": 8.606109883454397e-06, "loss": 0.2752, "step": 16270 }, { "epoch": 0.29611306135069754, "grad_norm": 2.214723825454712, "learning_rate": 8.604321817964332e-06, "loss": 0.2637, "step": 16280 }, { "epoch": 0.29629494898051983, "grad_norm": 1.6713924407958984, "learning_rate": 8.602532848684328e-06, "loss": 0.2675, "step": 16290 }, { "epoch": 0.29647683661034213, "grad_norm": 2.3720040321350098, "learning_rate": 8.600742976235222e-06, "loss": 0.2705, "step": 16300 }, { "epoch": 0.29665872424016443, "grad_norm": 1.7470122575759888, "learning_rate": 8.598952201238158e-06, "loss": 0.2755, "step": 16310 }, { "epoch": 0.2968406118699867, "grad_norm": 1.9372084140777588, "learning_rate": 8.5971605243146e-06, "loss": 0.2754, "step": 16320 }, { "epoch": 0.297022499499809, "grad_norm": 1.8176859617233276, "learning_rate": 8.59536794608632e-06, "loss": 0.2681, "step": 16330 }, { "epoch": 0.2972043871296313, "grad_norm": 1.7375038862228394, "learning_rate": 8.593574467175407e-06, "loss": 0.2651, "step": 16340 }, { "epoch": 0.2973862747594536, "grad_norm": 2.16900897026062, "learning_rate": 8.59178008820426e-06, "loss": 0.2702, "step": 16350 }, { "epoch": 0.2975681623892759, "grad_norm": 1.5768966674804688, "learning_rate": 8.589984809795589e-06, "loss": 0.2697, "step": 16360 }, { "epoch": 0.2977500500190982, "grad_norm": 1.8247019052505493, "learning_rate": 8.588188632572419e-06, "loss": 0.2771, "step": 16370 }, { "epoch": 0.2979319376489205, "grad_norm": 1.714242696762085, "learning_rate": 8.586391557158086e-06, "loss": 0.2686, "step": 16380 }, { "epoch": 0.2981138252787428, "grad_norm": 2.091245174407959, "learning_rate": 8.584593584176237e-06, "loss": 0.2715, "step": 16390 }, { "epoch": 0.2982957129085651, "grad_norm": 1.478468656539917, "learning_rate": 8.58279471425083e-06, "loss": 0.2652, "step": 16400 }, { "epoch": 0.2984776005383874, "grad_norm": 1.6938995122909546, "learning_rate": 8.580994948006135e-06, "loss": 0.269, "step": 16410 }, { "epoch": 0.2986594881682097, "grad_norm": 1.7756412029266357, "learning_rate": 8.579194286066736e-06, "loss": 0.2671, "step": 16420 }, { "epoch": 0.298841375798032, "grad_norm": 1.8890210390090942, "learning_rate": 8.577392729057524e-06, "loss": 0.2696, "step": 16430 }, { "epoch": 0.2990232634278543, "grad_norm": 1.8928695917129517, "learning_rate": 8.5755902776037e-06, "loss": 0.2763, "step": 16440 }, { "epoch": 0.2992051510576766, "grad_norm": 1.531523585319519, "learning_rate": 8.573786932330779e-06, "loss": 0.2644, "step": 16450 }, { "epoch": 0.2993870386874989, "grad_norm": 1.5715446472167969, "learning_rate": 8.571982693864582e-06, "loss": 0.2685, "step": 16460 }, { "epoch": 0.2995689263173212, "grad_norm": 2.2516117095947266, "learning_rate": 8.570177562831247e-06, "loss": 0.2704, "step": 16470 }, { "epoch": 0.2997508139471435, "grad_norm": 1.9246903657913208, "learning_rate": 8.568371539857211e-06, "loss": 0.2699, "step": 16480 }, { "epoch": 0.2999327015769658, "grad_norm": 1.7765284776687622, "learning_rate": 8.566564625569233e-06, "loss": 0.2648, "step": 16490 }, { "epoch": 0.3001145892067881, "grad_norm": 1.6552984714508057, "learning_rate": 8.564756820594368e-06, "loss": 0.2715, "step": 16500 }, { "epoch": 0.3002964768366103, "grad_norm": 1.9084404706954956, "learning_rate": 8.562948125559992e-06, "loss": 0.2716, "step": 16510 }, { "epoch": 0.3004783644664326, "grad_norm": 1.7671000957489014, "learning_rate": 8.561138541093784e-06, "loss": 0.2719, "step": 16520 }, { "epoch": 0.3006602520962549, "grad_norm": 1.8123719692230225, "learning_rate": 8.55932806782373e-06, "loss": 0.2744, "step": 16530 }, { "epoch": 0.3008421397260772, "grad_norm": 1.872362732887268, "learning_rate": 8.55751670637813e-06, "loss": 0.2699, "step": 16540 }, { "epoch": 0.3010240273558995, "grad_norm": 1.7700779438018799, "learning_rate": 8.555704457385587e-06, "loss": 0.2689, "step": 16550 }, { "epoch": 0.3012059149857218, "grad_norm": 1.653733491897583, "learning_rate": 8.553891321475016e-06, "loss": 0.2681, "step": 16560 }, { "epoch": 0.3013878026155441, "grad_norm": 1.7761425971984863, "learning_rate": 8.552077299275637e-06, "loss": 0.2678, "step": 16570 }, { "epoch": 0.3015696902453664, "grad_norm": 1.9807618856430054, "learning_rate": 8.550262391416979e-06, "loss": 0.2711, "step": 16580 }, { "epoch": 0.3017515778751887, "grad_norm": 2.031745195388794, "learning_rate": 8.548446598528876e-06, "loss": 0.2804, "step": 16590 }, { "epoch": 0.301933465505011, "grad_norm": 1.5545111894607544, "learning_rate": 8.546629921241476e-06, "loss": 0.2664, "step": 16600 }, { "epoch": 0.3021153531348333, "grad_norm": 1.9626141786575317, "learning_rate": 8.544812360185222e-06, "loss": 0.2641, "step": 16610 }, { "epoch": 0.3022972407646556, "grad_norm": 1.7474576234817505, "learning_rate": 8.542993915990879e-06, "loss": 0.2668, "step": 16620 }, { "epoch": 0.3024791283944779, "grad_norm": 1.66206693649292, "learning_rate": 8.541174589289503e-06, "loss": 0.28, "step": 16630 }, { "epoch": 0.3026610160243002, "grad_norm": 1.9449723958969116, "learning_rate": 8.539354380712466e-06, "loss": 0.2718, "step": 16640 }, { "epoch": 0.3028429036541225, "grad_norm": 1.7889355421066284, "learning_rate": 8.537533290891444e-06, "loss": 0.2796, "step": 16650 }, { "epoch": 0.30302479128394477, "grad_norm": 1.9688464403152466, "learning_rate": 8.535711320458417e-06, "loss": 0.2745, "step": 16660 }, { "epoch": 0.30320667891376707, "grad_norm": 1.7477232217788696, "learning_rate": 8.533888470045674e-06, "loss": 0.2663, "step": 16670 }, { "epoch": 0.30338856654358937, "grad_norm": 1.9291878938674927, "learning_rate": 8.532064740285805e-06, "loss": 0.2759, "step": 16680 }, { "epoch": 0.30357045417341166, "grad_norm": 1.7114976644515991, "learning_rate": 8.53024013181171e-06, "loss": 0.271, "step": 16690 }, { "epoch": 0.30375234180323396, "grad_norm": 1.6495733261108398, "learning_rate": 8.528414645256587e-06, "loss": 0.268, "step": 16700 }, { "epoch": 0.30393422943305626, "grad_norm": 1.6727958917617798, "learning_rate": 8.526588281253947e-06, "loss": 0.2665, "step": 16710 }, { "epoch": 0.30411611706287855, "grad_norm": 1.6088627576828003, "learning_rate": 8.524761040437602e-06, "loss": 0.2665, "step": 16720 }, { "epoch": 0.30429800469270085, "grad_norm": 1.9844485521316528, "learning_rate": 8.522932923441665e-06, "loss": 0.2694, "step": 16730 }, { "epoch": 0.30447989232252315, "grad_norm": 1.8951237201690674, "learning_rate": 8.521103930900558e-06, "loss": 0.2699, "step": 16740 }, { "epoch": 0.30466177995234545, "grad_norm": 1.6571787595748901, "learning_rate": 8.519274063449002e-06, "loss": 0.2745, "step": 16750 }, { "epoch": 0.30484366758216774, "grad_norm": 1.8377057313919067, "learning_rate": 8.51744332172203e-06, "loss": 0.2656, "step": 16760 }, { "epoch": 0.30502555521199004, "grad_norm": 1.6869683265686035, "learning_rate": 8.515611706354968e-06, "loss": 0.2763, "step": 16770 }, { "epoch": 0.30520744284181234, "grad_norm": 1.8195377588272095, "learning_rate": 8.513779217983452e-06, "loss": 0.2754, "step": 16780 }, { "epoch": 0.30538933047163463, "grad_norm": 1.6624059677124023, "learning_rate": 8.511945857243418e-06, "loss": 0.2675, "step": 16790 }, { "epoch": 0.30557121810145693, "grad_norm": 1.7304959297180176, "learning_rate": 8.510111624771104e-06, "loss": 0.2758, "step": 16800 }, { "epoch": 0.30575310573127923, "grad_norm": 1.7821238040924072, "learning_rate": 8.508276521203058e-06, "loss": 0.2702, "step": 16810 }, { "epoch": 0.3059349933611015, "grad_norm": 1.8708544969558716, "learning_rate": 8.506440547176118e-06, "loss": 0.2714, "step": 16820 }, { "epoch": 0.3061168809909238, "grad_norm": 2.0409626960754395, "learning_rate": 8.504603703327436e-06, "loss": 0.2764, "step": 16830 }, { "epoch": 0.3062987686207461, "grad_norm": 1.8432759046554565, "learning_rate": 8.502765990294456e-06, "loss": 0.2701, "step": 16840 }, { "epoch": 0.3064806562505684, "grad_norm": 1.856852650642395, "learning_rate": 8.500927408714931e-06, "loss": 0.2746, "step": 16850 }, { "epoch": 0.3066625438803907, "grad_norm": 1.5718615055084229, "learning_rate": 8.499087959226912e-06, "loss": 0.265, "step": 16860 }, { "epoch": 0.306844431510213, "grad_norm": 1.7428364753723145, "learning_rate": 8.497247642468752e-06, "loss": 0.2688, "step": 16870 }, { "epoch": 0.3070263191400353, "grad_norm": 1.806641936302185, "learning_rate": 8.495406459079103e-06, "loss": 0.2748, "step": 16880 }, { "epoch": 0.3072082067698576, "grad_norm": 1.915995478630066, "learning_rate": 8.493564409696919e-06, "loss": 0.2659, "step": 16890 }, { "epoch": 0.3073900943996799, "grad_norm": 1.6505424976348877, "learning_rate": 8.49172149496146e-06, "loss": 0.2723, "step": 16900 }, { "epoch": 0.3075719820295022, "grad_norm": 1.6358137130737305, "learning_rate": 8.489877715512277e-06, "loss": 0.2678, "step": 16910 }, { "epoch": 0.30775386965932444, "grad_norm": 1.6353161334991455, "learning_rate": 8.488033071989223e-06, "loss": 0.275, "step": 16920 }, { "epoch": 0.30793575728914674, "grad_norm": 1.8037374019622803, "learning_rate": 8.48618756503246e-06, "loss": 0.2747, "step": 16930 }, { "epoch": 0.30811764491896904, "grad_norm": 1.7329192161560059, "learning_rate": 8.484341195282438e-06, "loss": 0.2639, "step": 16940 }, { "epoch": 0.30829953254879133, "grad_norm": 1.6963666677474976, "learning_rate": 8.482493963379914e-06, "loss": 0.2768, "step": 16950 }, { "epoch": 0.30848142017861363, "grad_norm": 2.0056605339050293, "learning_rate": 8.480645869965939e-06, "loss": 0.2715, "step": 16960 }, { "epoch": 0.3086633078084359, "grad_norm": 1.720537543296814, "learning_rate": 8.478796915681866e-06, "loss": 0.2665, "step": 16970 }, { "epoch": 0.3088451954382582, "grad_norm": 1.799531102180481, "learning_rate": 8.476947101169348e-06, "loss": 0.2723, "step": 16980 }, { "epoch": 0.3090270830680805, "grad_norm": 1.6935275793075562, "learning_rate": 8.475096427070333e-06, "loss": 0.2677, "step": 16990 }, { "epoch": 0.3092089706979028, "grad_norm": 2.0185720920562744, "learning_rate": 8.47324489402707e-06, "loss": 0.2698, "step": 17000 }, { "epoch": 0.3093908583277251, "grad_norm": 1.8859308958053589, "learning_rate": 8.471392502682104e-06, "loss": 0.2694, "step": 17010 }, { "epoch": 0.3095727459575474, "grad_norm": 1.8255337476730347, "learning_rate": 8.469539253678281e-06, "loss": 0.2661, "step": 17020 }, { "epoch": 0.3097546335873697, "grad_norm": 1.8749003410339355, "learning_rate": 8.46768514765874e-06, "loss": 0.2681, "step": 17030 }, { "epoch": 0.309936521217192, "grad_norm": 1.7183011770248413, "learning_rate": 8.465830185266924e-06, "loss": 0.2685, "step": 17040 }, { "epoch": 0.3101184088470143, "grad_norm": 1.9259191751480103, "learning_rate": 8.463974367146564e-06, "loss": 0.2745, "step": 17050 }, { "epoch": 0.3103002964768366, "grad_norm": 1.7431223392486572, "learning_rate": 8.462117693941698e-06, "loss": 0.2766, "step": 17060 }, { "epoch": 0.3104821841066589, "grad_norm": 1.6158863306045532, "learning_rate": 8.460260166296652e-06, "loss": 0.2597, "step": 17070 }, { "epoch": 0.3106640717364812, "grad_norm": 1.6540768146514893, "learning_rate": 8.458401784856055e-06, "loss": 0.2645, "step": 17080 }, { "epoch": 0.3108459593663035, "grad_norm": 1.8706053495407104, "learning_rate": 8.45654255026483e-06, "loss": 0.2677, "step": 17090 }, { "epoch": 0.3110278469961258, "grad_norm": 2.197874069213867, "learning_rate": 8.454682463168193e-06, "loss": 0.2727, "step": 17100 }, { "epoch": 0.3112097346259481, "grad_norm": 1.718917727470398, "learning_rate": 8.452821524211663e-06, "loss": 0.2722, "step": 17110 }, { "epoch": 0.3113916222557704, "grad_norm": 1.7869696617126465, "learning_rate": 8.450959734041046e-06, "loss": 0.2648, "step": 17120 }, { "epoch": 0.3115735098855927, "grad_norm": 1.8987525701522827, "learning_rate": 8.449097093302451e-06, "loss": 0.2739, "step": 17130 }, { "epoch": 0.311755397515415, "grad_norm": 1.6336997747421265, "learning_rate": 8.447233602642278e-06, "loss": 0.2717, "step": 17140 }, { "epoch": 0.3119372851452373, "grad_norm": 1.8327680826187134, "learning_rate": 8.445369262707222e-06, "loss": 0.2689, "step": 17150 }, { "epoch": 0.3121191727750596, "grad_norm": 1.8505804538726807, "learning_rate": 8.443504074144272e-06, "loss": 0.2753, "step": 17160 }, { "epoch": 0.31230106040488187, "grad_norm": 1.892967700958252, "learning_rate": 8.441638037600717e-06, "loss": 0.2662, "step": 17170 }, { "epoch": 0.31248294803470417, "grad_norm": 1.9072929620742798, "learning_rate": 8.439771153724134e-06, "loss": 0.275, "step": 17180 }, { "epoch": 0.31266483566452646, "grad_norm": 1.8511836528778076, "learning_rate": 8.4379034231624e-06, "loss": 0.2678, "step": 17190 }, { "epoch": 0.31284672329434876, "grad_norm": 2.01460337638855, "learning_rate": 8.436034846563676e-06, "loss": 0.2687, "step": 17200 }, { "epoch": 0.31302861092417106, "grad_norm": 1.7561367750167847, "learning_rate": 8.434165424576426e-06, "loss": 0.2665, "step": 17210 }, { "epoch": 0.31321049855399336, "grad_norm": 2.0245440006256104, "learning_rate": 8.432295157849404e-06, "loss": 0.272, "step": 17220 }, { "epoch": 0.31339238618381565, "grad_norm": 1.680908441543579, "learning_rate": 8.430424047031659e-06, "loss": 0.2665, "step": 17230 }, { "epoch": 0.31357427381363795, "grad_norm": 1.710769534111023, "learning_rate": 8.428552092772528e-06, "loss": 0.274, "step": 17240 }, { "epoch": 0.31375616144346025, "grad_norm": 1.889796257019043, "learning_rate": 8.426679295721648e-06, "loss": 0.2696, "step": 17250 }, { "epoch": 0.31393804907328254, "grad_norm": 1.724150538444519, "learning_rate": 8.424805656528942e-06, "loss": 0.2619, "step": 17260 }, { "epoch": 0.31411993670310484, "grad_norm": 1.750916600227356, "learning_rate": 8.422931175844626e-06, "loss": 0.2709, "step": 17270 }, { "epoch": 0.31430182433292714, "grad_norm": 1.9360909461975098, "learning_rate": 8.421055854319214e-06, "loss": 0.2665, "step": 17280 }, { "epoch": 0.31448371196274943, "grad_norm": 1.7757810354232788, "learning_rate": 8.419179692603504e-06, "loss": 0.2656, "step": 17290 }, { "epoch": 0.31466559959257173, "grad_norm": 1.7188013792037964, "learning_rate": 8.417302691348592e-06, "loss": 0.2695, "step": 17300 }, { "epoch": 0.31484748722239403, "grad_norm": 1.691807508468628, "learning_rate": 8.415424851205862e-06, "loss": 0.2694, "step": 17310 }, { "epoch": 0.3150293748522163, "grad_norm": 1.8605324029922485, "learning_rate": 8.413546172826986e-06, "loss": 0.2626, "step": 17320 }, { "epoch": 0.3152112624820386, "grad_norm": 1.5942559242248535, "learning_rate": 8.411666656863935e-06, "loss": 0.2643, "step": 17330 }, { "epoch": 0.31539315011186086, "grad_norm": 1.5705549716949463, "learning_rate": 8.409786303968965e-06, "loss": 0.2719, "step": 17340 }, { "epoch": 0.31557503774168316, "grad_norm": 2.1736111640930176, "learning_rate": 8.407905114794621e-06, "loss": 0.2691, "step": 17350 }, { "epoch": 0.31575692537150546, "grad_norm": 1.9853577613830566, "learning_rate": 8.406023089993744e-06, "loss": 0.2743, "step": 17360 }, { "epoch": 0.31593881300132776, "grad_norm": 1.6117346286773682, "learning_rate": 8.404140230219461e-06, "loss": 0.2687, "step": 17370 }, { "epoch": 0.31612070063115005, "grad_norm": 1.8996866941452026, "learning_rate": 8.402256536125188e-06, "loss": 0.2689, "step": 17380 }, { "epoch": 0.31630258826097235, "grad_norm": 1.8959630727767944, "learning_rate": 8.400372008364635e-06, "loss": 0.2633, "step": 17390 }, { "epoch": 0.31648447589079465, "grad_norm": 1.713854432106018, "learning_rate": 8.398486647591795e-06, "loss": 0.267, "step": 17400 }, { "epoch": 0.31666636352061694, "grad_norm": 1.573501467704773, "learning_rate": 8.396600454460957e-06, "loss": 0.2698, "step": 17410 }, { "epoch": 0.31684825115043924, "grad_norm": 1.89168381690979, "learning_rate": 8.394713429626695e-06, "loss": 0.2724, "step": 17420 }, { "epoch": 0.31703013878026154, "grad_norm": 1.7039189338684082, "learning_rate": 8.39282557374387e-06, "loss": 0.2731, "step": 17430 }, { "epoch": 0.31721202641008384, "grad_norm": 1.78823721408844, "learning_rate": 8.390936887467633e-06, "loss": 0.2725, "step": 17440 }, { "epoch": 0.31739391403990613, "grad_norm": 1.8221160173416138, "learning_rate": 8.389047371453427e-06, "loss": 0.2752, "step": 17450 }, { "epoch": 0.31757580166972843, "grad_norm": 1.623711109161377, "learning_rate": 8.387157026356976e-06, "loss": 0.2633, "step": 17460 }, { "epoch": 0.3177576892995507, "grad_norm": 1.9298077821731567, "learning_rate": 8.385265852834296e-06, "loss": 0.268, "step": 17470 }, { "epoch": 0.317939576929373, "grad_norm": 1.8997223377227783, "learning_rate": 8.383373851541691e-06, "loss": 0.2656, "step": 17480 }, { "epoch": 0.3181214645591953, "grad_norm": 1.6705416440963745, "learning_rate": 8.381481023135752e-06, "loss": 0.2776, "step": 17490 }, { "epoch": 0.3183033521890176, "grad_norm": 1.946698546409607, "learning_rate": 8.379587368273354e-06, "loss": 0.2724, "step": 17500 }, { "epoch": 0.3184852398188399, "grad_norm": 1.7014662027359009, "learning_rate": 8.37769288761166e-06, "loss": 0.2677, "step": 17510 }, { "epoch": 0.3186671274486622, "grad_norm": 1.8423265218734741, "learning_rate": 8.375797581808123e-06, "loss": 0.2712, "step": 17520 }, { "epoch": 0.3188490150784845, "grad_norm": 1.8786901235580444, "learning_rate": 8.373901451520478e-06, "loss": 0.2683, "step": 17530 }, { "epoch": 0.3190309027083068, "grad_norm": 1.560387372970581, "learning_rate": 8.372004497406749e-06, "loss": 0.2655, "step": 17540 }, { "epoch": 0.3192127903381291, "grad_norm": 1.705328345298767, "learning_rate": 8.370106720125244e-06, "loss": 0.2692, "step": 17550 }, { "epoch": 0.3193946779679514, "grad_norm": 2.012160301208496, "learning_rate": 8.36820812033456e-06, "loss": 0.2613, "step": 17560 }, { "epoch": 0.3195765655977737, "grad_norm": 1.7320992946624756, "learning_rate": 8.366308698693573e-06, "loss": 0.2712, "step": 17570 }, { "epoch": 0.319758453227596, "grad_norm": 1.534199833869934, "learning_rate": 8.364408455861451e-06, "loss": 0.2689, "step": 17580 }, { "epoch": 0.3199403408574183, "grad_norm": 1.828579306602478, "learning_rate": 8.362507392497643e-06, "loss": 0.2655, "step": 17590 }, { "epoch": 0.3201222284872406, "grad_norm": 2.2243027687072754, "learning_rate": 8.360605509261883e-06, "loss": 0.2649, "step": 17600 }, { "epoch": 0.3203041161170629, "grad_norm": 1.6375197172164917, "learning_rate": 8.358702806814192e-06, "loss": 0.2737, "step": 17610 }, { "epoch": 0.3204860037468852, "grad_norm": 1.743903398513794, "learning_rate": 8.356799285814875e-06, "loss": 0.2763, "step": 17620 }, { "epoch": 0.3206678913767075, "grad_norm": 1.760372519493103, "learning_rate": 8.354894946924514e-06, "loss": 0.2621, "step": 17630 }, { "epoch": 0.3208497790065298, "grad_norm": 1.7881368398666382, "learning_rate": 8.352989790803988e-06, "loss": 0.2649, "step": 17640 }, { "epoch": 0.3210316666363521, "grad_norm": 1.725089430809021, "learning_rate": 8.351083818114447e-06, "loss": 0.2734, "step": 17650 }, { "epoch": 0.3212135542661744, "grad_norm": 1.6019036769866943, "learning_rate": 8.349177029517333e-06, "loss": 0.268, "step": 17660 }, { "epoch": 0.32139544189599667, "grad_norm": 1.9626137018203735, "learning_rate": 8.347269425674362e-06, "loss": 0.2685, "step": 17670 }, { "epoch": 0.32157732952581897, "grad_norm": 1.6321464776992798, "learning_rate": 8.345361007247547e-06, "loss": 0.2673, "step": 17680 }, { "epoch": 0.32175921715564126, "grad_norm": 1.6685935258865356, "learning_rate": 8.343451774899169e-06, "loss": 0.2668, "step": 17690 }, { "epoch": 0.32194110478546356, "grad_norm": 1.8209408521652222, "learning_rate": 8.3415417292918e-06, "loss": 0.2649, "step": 17700 }, { "epoch": 0.32212299241528586, "grad_norm": 1.7171199321746826, "learning_rate": 8.339630871088293e-06, "loss": 0.2612, "step": 17710 }, { "epoch": 0.32230488004510816, "grad_norm": 2.063389301300049, "learning_rate": 8.337719200951779e-06, "loss": 0.2711, "step": 17720 }, { "epoch": 0.32248676767493045, "grad_norm": 1.626177191734314, "learning_rate": 8.335806719545677e-06, "loss": 0.2653, "step": 17730 }, { "epoch": 0.32266865530475275, "grad_norm": 1.8304898738861084, "learning_rate": 8.333893427533684e-06, "loss": 0.2723, "step": 17740 }, { "epoch": 0.32285054293457505, "grad_norm": 2.1485824584960938, "learning_rate": 8.331979325579775e-06, "loss": 0.2744, "step": 17750 }, { "epoch": 0.3230324305643973, "grad_norm": 1.7210336923599243, "learning_rate": 8.330064414348213e-06, "loss": 0.275, "step": 17760 }, { "epoch": 0.3232143181942196, "grad_norm": 1.93837308883667, "learning_rate": 8.328148694503539e-06, "loss": 0.273, "step": 17770 }, { "epoch": 0.3233962058240419, "grad_norm": 1.7995526790618896, "learning_rate": 8.326232166710574e-06, "loss": 0.268, "step": 17780 }, { "epoch": 0.3235780934538642, "grad_norm": 1.7017087936401367, "learning_rate": 8.324314831634418e-06, "loss": 0.2703, "step": 17790 }, { "epoch": 0.3237599810836865, "grad_norm": 1.7025460004806519, "learning_rate": 8.322396689940453e-06, "loss": 0.2645, "step": 17800 }, { "epoch": 0.3239418687135088, "grad_norm": 1.7240791320800781, "learning_rate": 8.32047774229434e-06, "loss": 0.2636, "step": 17810 }, { "epoch": 0.32412375634333107, "grad_norm": 1.8858330249786377, "learning_rate": 8.318557989362022e-06, "loss": 0.2637, "step": 17820 }, { "epoch": 0.32430564397315337, "grad_norm": 1.648013710975647, "learning_rate": 8.31663743180972e-06, "loss": 0.2654, "step": 17830 }, { "epoch": 0.32448753160297567, "grad_norm": 1.8663712739944458, "learning_rate": 8.314716070303933e-06, "loss": 0.2736, "step": 17840 }, { "epoch": 0.32466941923279796, "grad_norm": 1.7736581563949585, "learning_rate": 8.31279390551144e-06, "loss": 0.2712, "step": 17850 }, { "epoch": 0.32485130686262026, "grad_norm": 1.4665162563323975, "learning_rate": 8.3108709380993e-06, "loss": 0.2688, "step": 17860 }, { "epoch": 0.32503319449244256, "grad_norm": 1.6102604866027832, "learning_rate": 8.308947168734846e-06, "loss": 0.271, "step": 17870 }, { "epoch": 0.32521508212226485, "grad_norm": 1.81101393699646, "learning_rate": 8.307022598085699e-06, "loss": 0.2742, "step": 17880 }, { "epoch": 0.32539696975208715, "grad_norm": 1.9265875816345215, "learning_rate": 8.305097226819744e-06, "loss": 0.2774, "step": 17890 }, { "epoch": 0.32557885738190945, "grad_norm": 1.8511744737625122, "learning_rate": 8.30317105560516e-06, "loss": 0.2619, "step": 17900 }, { "epoch": 0.32576074501173174, "grad_norm": 1.7493246793746948, "learning_rate": 8.301244085110387e-06, "loss": 0.2677, "step": 17910 }, { "epoch": 0.32594263264155404, "grad_norm": 1.892838954925537, "learning_rate": 8.299316316004154e-06, "loss": 0.268, "step": 17920 }, { "epoch": 0.32612452027137634, "grad_norm": 1.7410292625427246, "learning_rate": 8.297387748955465e-06, "loss": 0.2679, "step": 17930 }, { "epoch": 0.32630640790119864, "grad_norm": 1.7477391958236694, "learning_rate": 8.2954583846336e-06, "loss": 0.2647, "step": 17940 }, { "epoch": 0.32648829553102093, "grad_norm": 1.8457207679748535, "learning_rate": 8.29352822370811e-06, "loss": 0.2667, "step": 17950 }, { "epoch": 0.32667018316084323, "grad_norm": 1.7903870344161987, "learning_rate": 8.291597266848833e-06, "loss": 0.2646, "step": 17960 }, { "epoch": 0.3268520707906655, "grad_norm": 1.6569454669952393, "learning_rate": 8.289665514725878e-06, "loss": 0.2612, "step": 17970 }, { "epoch": 0.3270339584204878, "grad_norm": 1.8847482204437256, "learning_rate": 8.287732968009628e-06, "loss": 0.2637, "step": 17980 }, { "epoch": 0.3272158460503101, "grad_norm": 1.4509360790252686, "learning_rate": 8.285799627370741e-06, "loss": 0.2679, "step": 17990 }, { "epoch": 0.3273977336801324, "grad_norm": 2.0952248573303223, "learning_rate": 8.283865493480158e-06, "loss": 0.2758, "step": 18000 }, { "epoch": 0.3275796213099547, "grad_norm": 1.6410502195358276, "learning_rate": 8.28193056700909e-06, "loss": 0.2734, "step": 18010 }, { "epoch": 0.327761508939777, "grad_norm": 1.9024614095687866, "learning_rate": 8.279994848629022e-06, "loss": 0.2741, "step": 18020 }, { "epoch": 0.3279433965695993, "grad_norm": 1.837601661682129, "learning_rate": 8.278058339011715e-06, "loss": 0.2707, "step": 18030 }, { "epoch": 0.3281252841994216, "grad_norm": 1.9234360456466675, "learning_rate": 8.276121038829206e-06, "loss": 0.2627, "step": 18040 }, { "epoch": 0.3283071718292439, "grad_norm": 1.7049081325531006, "learning_rate": 8.274182948753803e-06, "loss": 0.2683, "step": 18050 }, { "epoch": 0.3284890594590662, "grad_norm": 1.4943821430206299, "learning_rate": 8.272244069458095e-06, "loss": 0.2632, "step": 18060 }, { "epoch": 0.3286709470888885, "grad_norm": 1.9784094095230103, "learning_rate": 8.270304401614936e-06, "loss": 0.2647, "step": 18070 }, { "epoch": 0.3288528347187108, "grad_norm": 1.823173999786377, "learning_rate": 8.268363945897461e-06, "loss": 0.2723, "step": 18080 }, { "epoch": 0.3290347223485331, "grad_norm": 1.9103217124938965, "learning_rate": 8.266422702979075e-06, "loss": 0.27, "step": 18090 }, { "epoch": 0.3292166099783554, "grad_norm": 1.655578851699829, "learning_rate": 8.264480673533454e-06, "loss": 0.2642, "step": 18100 }, { "epoch": 0.3293984976081777, "grad_norm": 1.637680172920227, "learning_rate": 8.262537858234552e-06, "loss": 0.2695, "step": 18110 }, { "epoch": 0.329580385238, "grad_norm": 1.4565362930297852, "learning_rate": 8.260594257756592e-06, "loss": 0.27, "step": 18120 }, { "epoch": 0.3297622728678223, "grad_norm": 1.8409793376922607, "learning_rate": 8.25864987277407e-06, "loss": 0.2616, "step": 18130 }, { "epoch": 0.3299441604976446, "grad_norm": 1.5586967468261719, "learning_rate": 8.256704703961758e-06, "loss": 0.2651, "step": 18140 }, { "epoch": 0.3301260481274669, "grad_norm": 1.8315738439559937, "learning_rate": 8.254758751994693e-06, "loss": 0.2687, "step": 18150 }, { "epoch": 0.3303079357572892, "grad_norm": 1.7245064973831177, "learning_rate": 8.252812017548191e-06, "loss": 0.2663, "step": 18160 }, { "epoch": 0.33048982338711147, "grad_norm": 1.6562284231185913, "learning_rate": 8.250864501297838e-06, "loss": 0.2673, "step": 18170 }, { "epoch": 0.3306717110169337, "grad_norm": 2.025259256362915, "learning_rate": 8.248916203919484e-06, "loss": 0.2648, "step": 18180 }, { "epoch": 0.330853598646756, "grad_norm": 1.8241842985153198, "learning_rate": 8.246967126089257e-06, "loss": 0.2724, "step": 18190 }, { "epoch": 0.3310354862765783, "grad_norm": 1.846366047859192, "learning_rate": 8.245017268483559e-06, "loss": 0.2769, "step": 18200 }, { "epoch": 0.3312173739064006, "grad_norm": 1.6324641704559326, "learning_rate": 8.243066631779056e-06, "loss": 0.2657, "step": 18210 }, { "epoch": 0.3313992615362229, "grad_norm": 1.6214702129364014, "learning_rate": 8.241115216652684e-06, "loss": 0.258, "step": 18220 }, { "epoch": 0.3315811491660452, "grad_norm": 1.610123634338379, "learning_rate": 8.239163023781654e-06, "loss": 0.2607, "step": 18230 }, { "epoch": 0.3317630367958675, "grad_norm": 1.8203229904174805, "learning_rate": 8.237210053843444e-06, "loss": 0.2673, "step": 18240 }, { "epoch": 0.3319449244256898, "grad_norm": 1.7708474397659302, "learning_rate": 8.235256307515801e-06, "loss": 0.2698, "step": 18250 }, { "epoch": 0.3321268120555121, "grad_norm": 1.7932089567184448, "learning_rate": 8.233301785476746e-06, "loss": 0.2656, "step": 18260 }, { "epoch": 0.3323086996853344, "grad_norm": 1.9614633321762085, "learning_rate": 8.231346488404564e-06, "loss": 0.2681, "step": 18270 }, { "epoch": 0.3324905873151567, "grad_norm": 1.5292779207229614, "learning_rate": 8.22939041697781e-06, "loss": 0.2654, "step": 18280 }, { "epoch": 0.332672474944979, "grad_norm": 1.6557673215866089, "learning_rate": 8.22743357187531e-06, "loss": 0.2687, "step": 18290 }, { "epoch": 0.3328543625748013, "grad_norm": 1.693480372428894, "learning_rate": 8.225475953776156e-06, "loss": 0.2712, "step": 18300 }, { "epoch": 0.3330362502046236, "grad_norm": 1.7912392616271973, "learning_rate": 8.22351756335971e-06, "loss": 0.2634, "step": 18310 }, { "epoch": 0.33321813783444587, "grad_norm": 2.2468338012695312, "learning_rate": 8.221558401305602e-06, "loss": 0.2652, "step": 18320 }, { "epoch": 0.33340002546426817, "grad_norm": 1.6185134649276733, "learning_rate": 8.21959846829373e-06, "loss": 0.2695, "step": 18330 }, { "epoch": 0.33358191309409047, "grad_norm": 1.9489388465881348, "learning_rate": 8.217637765004254e-06, "loss": 0.2695, "step": 18340 }, { "epoch": 0.33376380072391276, "grad_norm": 1.8656140565872192, "learning_rate": 8.215676292117612e-06, "loss": 0.2719, "step": 18350 }, { "epoch": 0.33394568835373506, "grad_norm": 1.6311802864074707, "learning_rate": 8.213714050314496e-06, "loss": 0.2625, "step": 18360 }, { "epoch": 0.33412757598355736, "grad_norm": 1.8246026039123535, "learning_rate": 8.211751040275879e-06, "loss": 0.2696, "step": 18370 }, { "epoch": 0.33430946361337965, "grad_norm": 1.692513346672058, "learning_rate": 8.20978726268299e-06, "loss": 0.2753, "step": 18380 }, { "epoch": 0.33449135124320195, "grad_norm": 1.8024905920028687, "learning_rate": 8.207822718217328e-06, "loss": 0.2698, "step": 18390 }, { "epoch": 0.33467323887302425, "grad_norm": 1.8674801588058472, "learning_rate": 8.205857407560658e-06, "loss": 0.2757, "step": 18400 }, { "epoch": 0.33485512650284655, "grad_norm": 1.7579894065856934, "learning_rate": 8.203891331395012e-06, "loss": 0.2695, "step": 18410 }, { "epoch": 0.33503701413266884, "grad_norm": 1.7543578147888184, "learning_rate": 8.201924490402687e-06, "loss": 0.2746, "step": 18420 }, { "epoch": 0.33521890176249114, "grad_norm": 1.6305710077285767, "learning_rate": 8.199956885266242e-06, "loss": 0.2673, "step": 18430 }, { "epoch": 0.33540078939231344, "grad_norm": 1.9328386783599854, "learning_rate": 8.197988516668508e-06, "loss": 0.2645, "step": 18440 }, { "epoch": 0.33558267702213573, "grad_norm": 1.7740346193313599, "learning_rate": 8.196019385292573e-06, "loss": 0.2712, "step": 18450 }, { "epoch": 0.33576456465195803, "grad_norm": 1.8161479234695435, "learning_rate": 8.194049491821799e-06, "loss": 0.2746, "step": 18460 }, { "epoch": 0.33594645228178033, "grad_norm": 1.775051236152649, "learning_rate": 8.192078836939801e-06, "loss": 0.263, "step": 18470 }, { "epoch": 0.3361283399116026, "grad_norm": 1.972555160522461, "learning_rate": 8.19010742133047e-06, "loss": 0.2747, "step": 18480 }, { "epoch": 0.3363102275414249, "grad_norm": 1.5046793222427368, "learning_rate": 8.188135245677954e-06, "loss": 0.2629, "step": 18490 }, { "epoch": 0.3364921151712472, "grad_norm": 1.9561882019042969, "learning_rate": 8.186162310666666e-06, "loss": 0.2629, "step": 18500 }, { "epoch": 0.3366740028010695, "grad_norm": 1.8721929788589478, "learning_rate": 8.184188616981281e-06, "loss": 0.2738, "step": 18510 }, { "epoch": 0.3368558904308918, "grad_norm": 1.7283061742782593, "learning_rate": 8.182214165306741e-06, "loss": 0.2635, "step": 18520 }, { "epoch": 0.3370377780607141, "grad_norm": 1.8004591464996338, "learning_rate": 8.18023895632825e-06, "loss": 0.2693, "step": 18530 }, { "epoch": 0.3372196656905364, "grad_norm": 1.7982103824615479, "learning_rate": 8.178262990731272e-06, "loss": 0.2738, "step": 18540 }, { "epoch": 0.3374015533203587, "grad_norm": 1.7707053422927856, "learning_rate": 8.176286269201538e-06, "loss": 0.2658, "step": 18550 }, { "epoch": 0.337583440950181, "grad_norm": 1.9601140022277832, "learning_rate": 8.174308792425034e-06, "loss": 0.266, "step": 18560 }, { "epoch": 0.3377653285800033, "grad_norm": 2.0345799922943115, "learning_rate": 8.17233056108802e-06, "loss": 0.274, "step": 18570 }, { "epoch": 0.3379472162098256, "grad_norm": 1.599351406097412, "learning_rate": 8.170351575877006e-06, "loss": 0.2627, "step": 18580 }, { "epoch": 0.3381291038396479, "grad_norm": 1.983981966972351, "learning_rate": 8.168371837478768e-06, "loss": 0.2723, "step": 18590 }, { "epoch": 0.33831099146947013, "grad_norm": 1.7267260551452637, "learning_rate": 8.166391346580348e-06, "loss": 0.2626, "step": 18600 }, { "epoch": 0.33849287909929243, "grad_norm": 1.8020488023757935, "learning_rate": 8.16441010386904e-06, "loss": 0.271, "step": 18610 }, { "epoch": 0.33867476672911473, "grad_norm": 1.9002126455307007, "learning_rate": 8.162428110032409e-06, "loss": 0.2664, "step": 18620 }, { "epoch": 0.338856654358937, "grad_norm": 2.0641863346099854, "learning_rate": 8.160445365758271e-06, "loss": 0.2689, "step": 18630 }, { "epoch": 0.3390385419887593, "grad_norm": 1.7440769672393799, "learning_rate": 8.15846187173471e-06, "loss": 0.2711, "step": 18640 }, { "epoch": 0.3392204296185816, "grad_norm": 1.6792333126068115, "learning_rate": 8.156477628650065e-06, "loss": 0.2713, "step": 18650 }, { "epoch": 0.3394023172484039, "grad_norm": 1.6586326360702515, "learning_rate": 8.15449263719294e-06, "loss": 0.2671, "step": 18660 }, { "epoch": 0.3395842048782262, "grad_norm": 1.6031054258346558, "learning_rate": 8.152506898052191e-06, "loss": 0.2723, "step": 18670 }, { "epoch": 0.3397660925080485, "grad_norm": 1.8894691467285156, "learning_rate": 8.150520411916943e-06, "loss": 0.2701, "step": 18680 }, { "epoch": 0.3399479801378708, "grad_norm": 1.7643051147460938, "learning_rate": 8.148533179476575e-06, "loss": 0.2669, "step": 18690 }, { "epoch": 0.3401298677676931, "grad_norm": 1.7088966369628906, "learning_rate": 8.146545201420722e-06, "loss": 0.2692, "step": 18700 }, { "epoch": 0.3403117553975154, "grad_norm": 1.7181262969970703, "learning_rate": 8.144556478439285e-06, "loss": 0.2698, "step": 18710 }, { "epoch": 0.3404936430273377, "grad_norm": 1.6215527057647705, "learning_rate": 8.14256701122242e-06, "loss": 0.2639, "step": 18720 }, { "epoch": 0.34067553065716, "grad_norm": 1.8926949501037598, "learning_rate": 8.140576800460537e-06, "loss": 0.2699, "step": 18730 }, { "epoch": 0.3408574182869823, "grad_norm": 1.830185055732727, "learning_rate": 8.138585846844312e-06, "loss": 0.2674, "step": 18740 }, { "epoch": 0.3410393059168046, "grad_norm": 1.6932241916656494, "learning_rate": 8.136594151064673e-06, "loss": 0.268, "step": 18750 }, { "epoch": 0.3412211935466269, "grad_norm": 1.5851161479949951, "learning_rate": 8.13460171381281e-06, "loss": 0.2621, "step": 18760 }, { "epoch": 0.3414030811764492, "grad_norm": 3.850177526473999, "learning_rate": 8.132608535780163e-06, "loss": 0.257, "step": 18770 }, { "epoch": 0.3415849688062715, "grad_norm": 1.9509552717208862, "learning_rate": 8.130614617658439e-06, "loss": 0.2663, "step": 18780 }, { "epoch": 0.3417668564360938, "grad_norm": 1.8255066871643066, "learning_rate": 8.128619960139594e-06, "loss": 0.276, "step": 18790 }, { "epoch": 0.3419487440659161, "grad_norm": 1.6966034173965454, "learning_rate": 8.126624563915842e-06, "loss": 0.2664, "step": 18800 }, { "epoch": 0.3421306316957384, "grad_norm": 4.6753129959106445, "learning_rate": 8.124628429679658e-06, "loss": 0.2645, "step": 18810 }, { "epoch": 0.34231251932556067, "grad_norm": 1.809352159500122, "learning_rate": 8.122631558123769e-06, "loss": 0.263, "step": 18820 }, { "epoch": 0.34249440695538297, "grad_norm": 1.8184926509857178, "learning_rate": 8.120633949941155e-06, "loss": 0.2645, "step": 18830 }, { "epoch": 0.34267629458520527, "grad_norm": 1.852921962738037, "learning_rate": 8.118635605825062e-06, "loss": 0.2797, "step": 18840 }, { "epoch": 0.34285818221502756, "grad_norm": 1.7741526365280151, "learning_rate": 8.116636526468978e-06, "loss": 0.2677, "step": 18850 }, { "epoch": 0.34304006984484986, "grad_norm": 1.6634918451309204, "learning_rate": 8.114636712566655e-06, "loss": 0.2659, "step": 18860 }, { "epoch": 0.34322195747467216, "grad_norm": 1.5919853448867798, "learning_rate": 8.112636164812099e-06, "loss": 0.2683, "step": 18870 }, { "epoch": 0.34340384510449445, "grad_norm": 1.947690486907959, "learning_rate": 8.110634883899571e-06, "loss": 0.2668, "step": 18880 }, { "epoch": 0.34358573273431675, "grad_norm": 1.6984257698059082, "learning_rate": 8.10863287052358e-06, "loss": 0.2693, "step": 18890 }, { "epoch": 0.34376762036413905, "grad_norm": 1.8709865808486938, "learning_rate": 8.106630125378896e-06, "loss": 0.2706, "step": 18900 }, { "epoch": 0.34394950799396135, "grad_norm": 1.7090344429016113, "learning_rate": 8.104626649160545e-06, "loss": 0.2651, "step": 18910 }, { "epoch": 0.34413139562378364, "grad_norm": 1.826995611190796, "learning_rate": 8.102622442563796e-06, "loss": 0.2699, "step": 18920 }, { "epoch": 0.34431328325360594, "grad_norm": 1.650665283203125, "learning_rate": 8.100617506284183e-06, "loss": 0.2685, "step": 18930 }, { "epoch": 0.34449517088342824, "grad_norm": 1.7687606811523438, "learning_rate": 8.098611841017484e-06, "loss": 0.2691, "step": 18940 }, { "epoch": 0.34467705851325053, "grad_norm": 1.783839225769043, "learning_rate": 8.09660544745974e-06, "loss": 0.2616, "step": 18950 }, { "epoch": 0.34485894614307283, "grad_norm": 1.8444640636444092, "learning_rate": 8.094598326307235e-06, "loss": 0.2677, "step": 18960 }, { "epoch": 0.34504083377289513, "grad_norm": 1.5408271551132202, "learning_rate": 8.092590478256512e-06, "loss": 0.2757, "step": 18970 }, { "epoch": 0.3452227214027174, "grad_norm": 2.0140933990478516, "learning_rate": 8.090581904004363e-06, "loss": 0.2723, "step": 18980 }, { "epoch": 0.3454046090325397, "grad_norm": 1.7882782220840454, "learning_rate": 8.088572604247832e-06, "loss": 0.2621, "step": 18990 }, { "epoch": 0.345586496662362, "grad_norm": 1.9493714570999146, "learning_rate": 8.086562579684217e-06, "loss": 0.2642, "step": 19000 }, { "epoch": 0.34576838429218426, "grad_norm": 1.75930655002594, "learning_rate": 8.084551831011065e-06, "loss": 0.2672, "step": 19010 }, { "epoch": 0.34595027192200656, "grad_norm": 1.7578414678573608, "learning_rate": 8.082540358926175e-06, "loss": 0.2746, "step": 19020 }, { "epoch": 0.34613215955182886, "grad_norm": 1.7360261678695679, "learning_rate": 8.080528164127599e-06, "loss": 0.2622, "step": 19030 }, { "epoch": 0.34631404718165115, "grad_norm": 1.798301100730896, "learning_rate": 8.078515247313639e-06, "loss": 0.273, "step": 19040 }, { "epoch": 0.34649593481147345, "grad_norm": 1.9654287099838257, "learning_rate": 8.076501609182844e-06, "loss": 0.2684, "step": 19050 }, { "epoch": 0.34667782244129575, "grad_norm": 2.107332944869995, "learning_rate": 8.074487250434018e-06, "loss": 0.2756, "step": 19060 }, { "epoch": 0.34685971007111804, "grad_norm": 1.8855998516082764, "learning_rate": 8.072472171766214e-06, "loss": 0.2782, "step": 19070 }, { "epoch": 0.34704159770094034, "grad_norm": 1.7120180130004883, "learning_rate": 8.070456373878735e-06, "loss": 0.2627, "step": 19080 }, { "epoch": 0.34722348533076264, "grad_norm": 1.6591873168945312, "learning_rate": 8.068439857471128e-06, "loss": 0.2732, "step": 19090 }, { "epoch": 0.34740537296058493, "grad_norm": 1.7582883834838867, "learning_rate": 8.066422623243198e-06, "loss": 0.2649, "step": 19100 }, { "epoch": 0.34758726059040723, "grad_norm": 1.7825669050216675, "learning_rate": 8.064404671894997e-06, "loss": 0.2694, "step": 19110 }, { "epoch": 0.34776914822022953, "grad_norm": 1.7874218225479126, "learning_rate": 8.062386004126818e-06, "loss": 0.2686, "step": 19120 }, { "epoch": 0.3479510358500518, "grad_norm": 1.6099412441253662, "learning_rate": 8.060366620639215e-06, "loss": 0.2668, "step": 19130 }, { "epoch": 0.3481329234798741, "grad_norm": 1.9130234718322754, "learning_rate": 8.058346522132979e-06, "loss": 0.2678, "step": 19140 }, { "epoch": 0.3483148111096964, "grad_norm": 1.7845873832702637, "learning_rate": 8.056325709309159e-06, "loss": 0.2724, "step": 19150 }, { "epoch": 0.3484966987395187, "grad_norm": 1.7056001424789429, "learning_rate": 8.05430418286904e-06, "loss": 0.2629, "step": 19160 }, { "epoch": 0.348678586369341, "grad_norm": 1.6891430616378784, "learning_rate": 8.052281943514171e-06, "loss": 0.2603, "step": 19170 }, { "epoch": 0.3488604739991633, "grad_norm": 1.742738127708435, "learning_rate": 8.050258991946334e-06, "loss": 0.2693, "step": 19180 }, { "epoch": 0.3490423616289856, "grad_norm": 1.8213987350463867, "learning_rate": 8.048235328867561e-06, "loss": 0.2737, "step": 19190 }, { "epoch": 0.3492242492588079, "grad_norm": 1.7335463762283325, "learning_rate": 8.046210954980139e-06, "loss": 0.2746, "step": 19200 }, { "epoch": 0.3494061368886302, "grad_norm": 1.4243296384811401, "learning_rate": 8.044185870986589e-06, "loss": 0.2656, "step": 19210 }, { "epoch": 0.3495880245184525, "grad_norm": 1.7706300020217896, "learning_rate": 8.042160077589693e-06, "loss": 0.2641, "step": 19220 }, { "epoch": 0.3497699121482748, "grad_norm": 1.8934544324874878, "learning_rate": 8.040133575492466e-06, "loss": 0.2709, "step": 19230 }, { "epoch": 0.3499517997780971, "grad_norm": 1.6971802711486816, "learning_rate": 8.038106365398175e-06, "loss": 0.2659, "step": 19240 }, { "epoch": 0.3501336874079194, "grad_norm": 1.6817641258239746, "learning_rate": 8.036078448010334e-06, "loss": 0.2723, "step": 19250 }, { "epoch": 0.3503155750377417, "grad_norm": 1.7039215564727783, "learning_rate": 8.034049824032701e-06, "loss": 0.2664, "step": 19260 }, { "epoch": 0.350497462667564, "grad_norm": 1.6824240684509277, "learning_rate": 8.032020494169277e-06, "loss": 0.2699, "step": 19270 }, { "epoch": 0.3506793502973863, "grad_norm": 1.784736156463623, "learning_rate": 8.029990459124313e-06, "loss": 0.2655, "step": 19280 }, { "epoch": 0.3508612379272086, "grad_norm": 1.8442054986953735, "learning_rate": 8.027959719602296e-06, "loss": 0.2621, "step": 19290 }, { "epoch": 0.3510431255570309, "grad_norm": 1.4809743165969849, "learning_rate": 8.025928276307969e-06, "loss": 0.2618, "step": 19300 }, { "epoch": 0.3512250131868532, "grad_norm": 1.7457940578460693, "learning_rate": 8.023896129946306e-06, "loss": 0.2699, "step": 19310 }, { "epoch": 0.35140690081667547, "grad_norm": 1.6051439046859741, "learning_rate": 8.02186328122254e-06, "loss": 0.2577, "step": 19320 }, { "epoch": 0.35158878844649777, "grad_norm": 1.5249656438827515, "learning_rate": 8.019829730842136e-06, "loss": 0.2624, "step": 19330 }, { "epoch": 0.35177067607632007, "grad_norm": 1.6823769807815552, "learning_rate": 8.017795479510807e-06, "loss": 0.2646, "step": 19340 }, { "epoch": 0.35195256370614236, "grad_norm": 1.9106541872024536, "learning_rate": 8.01576052793451e-06, "loss": 0.2723, "step": 19350 }, { "epoch": 0.35213445133596466, "grad_norm": 2.0271952152252197, "learning_rate": 8.01372487681944e-06, "loss": 0.2636, "step": 19360 }, { "epoch": 0.35231633896578696, "grad_norm": 1.8481642007827759, "learning_rate": 8.011688526872044e-06, "loss": 0.2758, "step": 19370 }, { "epoch": 0.35249822659560925, "grad_norm": 1.8158880472183228, "learning_rate": 8.009651478799e-06, "loss": 0.2629, "step": 19380 }, { "epoch": 0.35268011422543155, "grad_norm": 1.6552129983901978, "learning_rate": 8.007613733307237e-06, "loss": 0.2582, "step": 19390 }, { "epoch": 0.35286200185525385, "grad_norm": 1.6896536350250244, "learning_rate": 8.005575291103926e-06, "loss": 0.266, "step": 19400 }, { "epoch": 0.35304388948507615, "grad_norm": 1.620726227760315, "learning_rate": 8.003536152896475e-06, "loss": 0.2645, "step": 19410 }, { "epoch": 0.35322577711489844, "grad_norm": 1.7454380989074707, "learning_rate": 8.001496319392531e-06, "loss": 0.2648, "step": 19420 }, { "epoch": 0.3534076647447207, "grad_norm": 1.7014884948730469, "learning_rate": 7.999455791299996e-06, "loss": 0.27, "step": 19430 }, { "epoch": 0.353589552374543, "grad_norm": 2.0165810585021973, "learning_rate": 7.997414569326996e-06, "loss": 0.2636, "step": 19440 }, { "epoch": 0.3537714400043653, "grad_norm": 1.8034156560897827, "learning_rate": 7.99537265418191e-06, "loss": 0.2669, "step": 19450 }, { "epoch": 0.3539533276341876, "grad_norm": 1.917822003364563, "learning_rate": 7.993330046573355e-06, "loss": 0.2655, "step": 19460 }, { "epoch": 0.3541352152640099, "grad_norm": 1.7873165607452393, "learning_rate": 7.991286747210181e-06, "loss": 0.2712, "step": 19470 }, { "epoch": 0.35431710289383217, "grad_norm": 1.7552499771118164, "learning_rate": 7.989242756801488e-06, "loss": 0.2659, "step": 19480 }, { "epoch": 0.35449899052365447, "grad_norm": 1.6211185455322266, "learning_rate": 7.98719807605661e-06, "loss": 0.2674, "step": 19490 }, { "epoch": 0.35468087815347676, "grad_norm": 1.7629934549331665, "learning_rate": 7.98515270568512e-06, "loss": 0.2552, "step": 19500 }, { "epoch": 0.35486276578329906, "grad_norm": 1.6470991373062134, "learning_rate": 7.983106646396837e-06, "loss": 0.2666, "step": 19510 }, { "epoch": 0.35504465341312136, "grad_norm": 1.9281684160232544, "learning_rate": 7.981059898901811e-06, "loss": 0.2677, "step": 19520 }, { "epoch": 0.35522654104294366, "grad_norm": 1.6442055702209473, "learning_rate": 7.979012463910335e-06, "loss": 0.2652, "step": 19530 }, { "epoch": 0.35540842867276595, "grad_norm": 1.6182316541671753, "learning_rate": 7.976964342132943e-06, "loss": 0.2684, "step": 19540 }, { "epoch": 0.35559031630258825, "grad_norm": 1.6929875612258911, "learning_rate": 7.9749155342804e-06, "loss": 0.2792, "step": 19550 }, { "epoch": 0.35577220393241055, "grad_norm": 1.5331625938415527, "learning_rate": 7.972866041063715e-06, "loss": 0.2664, "step": 19560 }, { "epoch": 0.35595409156223284, "grad_norm": 1.772447943687439, "learning_rate": 7.970815863194134e-06, "loss": 0.2686, "step": 19570 }, { "epoch": 0.35613597919205514, "grad_norm": 1.6652271747589111, "learning_rate": 7.968765001383137e-06, "loss": 0.2632, "step": 19580 }, { "epoch": 0.35631786682187744, "grad_norm": 1.6461851596832275, "learning_rate": 7.966713456342448e-06, "loss": 0.2663, "step": 19590 }, { "epoch": 0.35649975445169974, "grad_norm": 1.7712687253952026, "learning_rate": 7.964661228784021e-06, "loss": 0.2684, "step": 19600 }, { "epoch": 0.35668164208152203, "grad_norm": 1.553924798965454, "learning_rate": 7.962608319420052e-06, "loss": 0.2651, "step": 19610 }, { "epoch": 0.35686352971134433, "grad_norm": 1.7967501878738403, "learning_rate": 7.960554728962972e-06, "loss": 0.2713, "step": 19620 }, { "epoch": 0.3570454173411666, "grad_norm": 2.0417370796203613, "learning_rate": 7.958500458125445e-06, "loss": 0.2683, "step": 19630 }, { "epoch": 0.3572273049709889, "grad_norm": 1.9547628164291382, "learning_rate": 7.956445507620378e-06, "loss": 0.2648, "step": 19640 }, { "epoch": 0.3574091926008112, "grad_norm": 1.7738639116287231, "learning_rate": 7.954389878160908e-06, "loss": 0.2648, "step": 19650 }, { "epoch": 0.3575910802306335, "grad_norm": 1.7396105527877808, "learning_rate": 7.95233357046041e-06, "loss": 0.2774, "step": 19660 }, { "epoch": 0.3577729678604558, "grad_norm": 1.8559890985488892, "learning_rate": 7.950276585232495e-06, "loss": 0.2657, "step": 19670 }, { "epoch": 0.3579548554902781, "grad_norm": 1.8426591157913208, "learning_rate": 7.948218923191006e-06, "loss": 0.2629, "step": 19680 }, { "epoch": 0.3581367431201004, "grad_norm": 1.7033740282058716, "learning_rate": 7.946160585050025e-06, "loss": 0.2641, "step": 19690 }, { "epoch": 0.3583186307499227, "grad_norm": 1.9214856624603271, "learning_rate": 7.944101571523865e-06, "loss": 0.2668, "step": 19700 }, { "epoch": 0.358500518379745, "grad_norm": 1.6501766443252563, "learning_rate": 7.942041883327077e-06, "loss": 0.2733, "step": 19710 }, { "epoch": 0.3586824060095673, "grad_norm": 1.9803887605667114, "learning_rate": 7.939981521174444e-06, "loss": 0.2687, "step": 19720 }, { "epoch": 0.3588642936393896, "grad_norm": 1.8244247436523438, "learning_rate": 7.937920485780982e-06, "loss": 0.2638, "step": 19730 }, { "epoch": 0.3590461812692119, "grad_norm": 1.7289621829986572, "learning_rate": 7.935858777861943e-06, "loss": 0.2613, "step": 19740 }, { "epoch": 0.3592280688990342, "grad_norm": 1.6927841901779175, "learning_rate": 7.933796398132808e-06, "loss": 0.2672, "step": 19750 }, { "epoch": 0.3594099565288565, "grad_norm": 1.7630157470703125, "learning_rate": 7.931733347309298e-06, "loss": 0.2594, "step": 19760 }, { "epoch": 0.3595918441586788, "grad_norm": 2.0156173706054688, "learning_rate": 7.92966962610736e-06, "loss": 0.2728, "step": 19770 }, { "epoch": 0.3597737317885011, "grad_norm": 2.094928026199341, "learning_rate": 7.92760523524318e-06, "loss": 0.2691, "step": 19780 }, { "epoch": 0.3599556194183234, "grad_norm": 1.7985477447509766, "learning_rate": 7.92554017543317e-06, "loss": 0.2653, "step": 19790 }, { "epoch": 0.3601375070481457, "grad_norm": 1.8866866827011108, "learning_rate": 7.92347444739398e-06, "loss": 0.2721, "step": 19800 }, { "epoch": 0.360319394677968, "grad_norm": 1.7237496376037598, "learning_rate": 7.921408051842486e-06, "loss": 0.2738, "step": 19810 }, { "epoch": 0.36050128230779027, "grad_norm": 1.6516615152359009, "learning_rate": 7.919340989495803e-06, "loss": 0.2668, "step": 19820 }, { "epoch": 0.36068316993761257, "grad_norm": 1.5235028266906738, "learning_rate": 7.917273261071271e-06, "loss": 0.2707, "step": 19830 }, { "epoch": 0.36086505756743487, "grad_norm": 1.7319467067718506, "learning_rate": 7.915204867286465e-06, "loss": 0.2715, "step": 19840 }, { "epoch": 0.3610469451972571, "grad_norm": 1.7679342031478882, "learning_rate": 7.913135808859186e-06, "loss": 0.2716, "step": 19850 }, { "epoch": 0.3612288328270794, "grad_norm": 1.796040415763855, "learning_rate": 7.911066086507473e-06, "loss": 0.2688, "step": 19860 }, { "epoch": 0.3614107204569017, "grad_norm": 2.0058789253234863, "learning_rate": 7.908995700949588e-06, "loss": 0.2681, "step": 19870 }, { "epoch": 0.361592608086724, "grad_norm": 1.5828564167022705, "learning_rate": 7.906924652904027e-06, "loss": 0.2652, "step": 19880 }, { "epoch": 0.3617744957165463, "grad_norm": 1.9555833339691162, "learning_rate": 7.904852943089519e-06, "loss": 0.2617, "step": 19890 }, { "epoch": 0.3619563833463686, "grad_norm": 1.616082787513733, "learning_rate": 7.902780572225014e-06, "loss": 0.2692, "step": 19900 }, { "epoch": 0.3621382709761909, "grad_norm": 1.642417550086975, "learning_rate": 7.900707541029701e-06, "loss": 0.2592, "step": 19910 }, { "epoch": 0.3623201586060132, "grad_norm": 1.7206300497055054, "learning_rate": 7.898633850222991e-06, "loss": 0.2699, "step": 19920 }, { "epoch": 0.3625020462358355, "grad_norm": 1.4592714309692383, "learning_rate": 7.896559500524526e-06, "loss": 0.2677, "step": 19930 }, { "epoch": 0.3626839338656578, "grad_norm": 1.6835170984268188, "learning_rate": 7.894484492654181e-06, "loss": 0.2707, "step": 19940 }, { "epoch": 0.3628658214954801, "grad_norm": 1.9406410455703735, "learning_rate": 7.892408827332051e-06, "loss": 0.2723, "step": 19950 }, { "epoch": 0.3630477091253024, "grad_norm": 1.7036038637161255, "learning_rate": 7.890332505278467e-06, "loss": 0.2709, "step": 19960 }, { "epoch": 0.3632295967551247, "grad_norm": 1.8259098529815674, "learning_rate": 7.888255527213985e-06, "loss": 0.2642, "step": 19970 }, { "epoch": 0.36341148438494697, "grad_norm": 1.9415758848190308, "learning_rate": 7.886177893859385e-06, "loss": 0.2636, "step": 19980 }, { "epoch": 0.36359337201476927, "grad_norm": 1.788092851638794, "learning_rate": 7.884099605935681e-06, "loss": 0.2618, "step": 19990 }, { "epoch": 0.36377525964459156, "grad_norm": 2.2342987060546875, "learning_rate": 7.882020664164109e-06, "loss": 0.2692, "step": 20000 }, { "epoch": 0.36395714727441386, "grad_norm": 1.5898945331573486, "learning_rate": 7.879941069266136e-06, "loss": 0.2766, "step": 20010 }, { "epoch": 0.36413903490423616, "grad_norm": 1.9137252569198608, "learning_rate": 7.877860821963452e-06, "loss": 0.2611, "step": 20020 }, { "epoch": 0.36432092253405846, "grad_norm": 1.607248067855835, "learning_rate": 7.875779922977975e-06, "loss": 0.2644, "step": 20030 }, { "epoch": 0.36450281016388075, "grad_norm": 1.5571565628051758, "learning_rate": 7.873698373031848e-06, "loss": 0.2583, "step": 20040 }, { "epoch": 0.36468469779370305, "grad_norm": 1.8645843267440796, "learning_rate": 7.871616172847442e-06, "loss": 0.2655, "step": 20050 }, { "epoch": 0.36486658542352535, "grad_norm": 1.7471466064453125, "learning_rate": 7.869533323147355e-06, "loss": 0.2678, "step": 20060 }, { "epoch": 0.36504847305334764, "grad_norm": 1.7227317094802856, "learning_rate": 7.867449824654404e-06, "loss": 0.2719, "step": 20070 }, { "epoch": 0.36523036068316994, "grad_norm": 1.750684142112732, "learning_rate": 7.865365678091638e-06, "loss": 0.2669, "step": 20080 }, { "epoch": 0.36541224831299224, "grad_norm": 1.7662454843521118, "learning_rate": 7.863280884182327e-06, "loss": 0.2753, "step": 20090 }, { "epoch": 0.36559413594281454, "grad_norm": 1.9723824262619019, "learning_rate": 7.861195443649966e-06, "loss": 0.2697, "step": 20100 }, { "epoch": 0.36577602357263683, "grad_norm": 1.8316662311553955, "learning_rate": 7.859109357218277e-06, "loss": 0.2659, "step": 20110 }, { "epoch": 0.36595791120245913, "grad_norm": 1.8012858629226685, "learning_rate": 7.857022625611204e-06, "loss": 0.2657, "step": 20120 }, { "epoch": 0.3661397988322814, "grad_norm": 1.760227918624878, "learning_rate": 7.854935249552914e-06, "loss": 0.2683, "step": 20130 }, { "epoch": 0.3663216864621037, "grad_norm": 1.5332924127578735, "learning_rate": 7.852847229767798e-06, "loss": 0.2639, "step": 20140 }, { "epoch": 0.366503574091926, "grad_norm": 1.6832489967346191, "learning_rate": 7.850758566980473e-06, "loss": 0.267, "step": 20150 }, { "epoch": 0.3666854617217483, "grad_norm": 1.6019920110702515, "learning_rate": 7.848669261915779e-06, "loss": 0.2677, "step": 20160 }, { "epoch": 0.3668673493515706, "grad_norm": 1.8959670066833496, "learning_rate": 7.846579315298773e-06, "loss": 0.2662, "step": 20170 }, { "epoch": 0.3670492369813929, "grad_norm": 1.9301179647445679, "learning_rate": 7.844488727854742e-06, "loss": 0.262, "step": 20180 }, { "epoch": 0.3672311246112152, "grad_norm": 1.5959312915802002, "learning_rate": 7.842397500309192e-06, "loss": 0.2622, "step": 20190 }, { "epoch": 0.3674130122410375, "grad_norm": 1.9067891836166382, "learning_rate": 7.840305633387848e-06, "loss": 0.2615, "step": 20200 }, { "epoch": 0.3675948998708598, "grad_norm": 1.8157660961151123, "learning_rate": 7.838213127816666e-06, "loss": 0.2684, "step": 20210 }, { "epoch": 0.3677767875006821, "grad_norm": 1.8712553977966309, "learning_rate": 7.836119984321816e-06, "loss": 0.2678, "step": 20220 }, { "epoch": 0.3679586751305044, "grad_norm": 1.6528904438018799, "learning_rate": 7.83402620362969e-06, "loss": 0.2688, "step": 20230 }, { "epoch": 0.3681405627603267, "grad_norm": 1.7343589067459106, "learning_rate": 7.831931786466902e-06, "loss": 0.2699, "step": 20240 }, { "epoch": 0.368322450390149, "grad_norm": 1.6307051181793213, "learning_rate": 7.829836733560288e-06, "loss": 0.2703, "step": 20250 }, { "epoch": 0.3685043380199713, "grad_norm": 1.8685102462768555, "learning_rate": 7.827741045636907e-06, "loss": 0.2668, "step": 20260 }, { "epoch": 0.36868622564979353, "grad_norm": 1.5694949626922607, "learning_rate": 7.82564472342403e-06, "loss": 0.2701, "step": 20270 }, { "epoch": 0.36886811327961583, "grad_norm": 1.7147200107574463, "learning_rate": 7.823547767649156e-06, "loss": 0.2645, "step": 20280 }, { "epoch": 0.3690500009094381, "grad_norm": 1.6672979593276978, "learning_rate": 7.821450179040001e-06, "loss": 0.256, "step": 20290 }, { "epoch": 0.3692318885392604, "grad_norm": 1.5367040634155273, "learning_rate": 7.819351958324504e-06, "loss": 0.2639, "step": 20300 }, { "epoch": 0.3694137761690827, "grad_norm": 1.5847842693328857, "learning_rate": 7.817253106230817e-06, "loss": 0.2639, "step": 20310 }, { "epoch": 0.369595663798905, "grad_norm": 1.7817471027374268, "learning_rate": 7.815153623487314e-06, "loss": 0.261, "step": 20320 }, { "epoch": 0.3697775514287273, "grad_norm": 1.720489263534546, "learning_rate": 7.81305351082259e-06, "loss": 0.2713, "step": 20330 }, { "epoch": 0.3699594390585496, "grad_norm": 1.5868655443191528, "learning_rate": 7.810952768965457e-06, "loss": 0.2688, "step": 20340 }, { "epoch": 0.3701413266883719, "grad_norm": 1.6682578325271606, "learning_rate": 7.808851398644944e-06, "loss": 0.2605, "step": 20350 }, { "epoch": 0.3703232143181942, "grad_norm": 1.9641801118850708, "learning_rate": 7.8067494005903e-06, "loss": 0.2656, "step": 20360 }, { "epoch": 0.3705051019480165, "grad_norm": 1.885187029838562, "learning_rate": 7.80464677553099e-06, "loss": 0.2649, "step": 20370 }, { "epoch": 0.3706869895778388, "grad_norm": 1.4614664316177368, "learning_rate": 7.8025435241967e-06, "loss": 0.266, "step": 20380 }, { "epoch": 0.3708688772076611, "grad_norm": 1.6875547170639038, "learning_rate": 7.800439647317331e-06, "loss": 0.2614, "step": 20390 }, { "epoch": 0.3710507648374834, "grad_norm": 1.5589380264282227, "learning_rate": 7.798335145622999e-06, "loss": 0.2627, "step": 20400 }, { "epoch": 0.3712326524673057, "grad_norm": 1.8382956981658936, "learning_rate": 7.796230019844041e-06, "loss": 0.2614, "step": 20410 }, { "epoch": 0.371414540097128, "grad_norm": 2.1202392578125, "learning_rate": 7.794124270711006e-06, "loss": 0.2579, "step": 20420 }, { "epoch": 0.3715964277269503, "grad_norm": 1.5790187120437622, "learning_rate": 7.792017898954665e-06, "loss": 0.2642, "step": 20430 }, { "epoch": 0.3717783153567726, "grad_norm": 2.0159480571746826, "learning_rate": 7.789910905306002e-06, "loss": 0.2644, "step": 20440 }, { "epoch": 0.3719602029865949, "grad_norm": 1.8591891527175903, "learning_rate": 7.787803290496214e-06, "loss": 0.2671, "step": 20450 }, { "epoch": 0.3721420906164172, "grad_norm": 1.705283761024475, "learning_rate": 7.78569505525672e-06, "loss": 0.2666, "step": 20460 }, { "epoch": 0.3723239782462395, "grad_norm": 1.8430522680282593, "learning_rate": 7.783586200319144e-06, "loss": 0.2661, "step": 20470 }, { "epoch": 0.37250586587606177, "grad_norm": 1.67185640335083, "learning_rate": 7.78147672641534e-06, "loss": 0.2673, "step": 20480 }, { "epoch": 0.37268775350588407, "grad_norm": 2.007230043411255, "learning_rate": 7.779366634277364e-06, "loss": 0.261, "step": 20490 }, { "epoch": 0.37286964113570636, "grad_norm": 1.703642725944519, "learning_rate": 7.777255924637494e-06, "loss": 0.2652, "step": 20500 }, { "epoch": 0.37305152876552866, "grad_norm": 1.753648281097412, "learning_rate": 7.775144598228217e-06, "loss": 0.2685, "step": 20510 }, { "epoch": 0.37323341639535096, "grad_norm": 1.765081763267517, "learning_rate": 7.773032655782234e-06, "loss": 0.2685, "step": 20520 }, { "epoch": 0.37341530402517326, "grad_norm": 1.6673791408538818, "learning_rate": 7.770920098032468e-06, "loss": 0.2656, "step": 20530 }, { "epoch": 0.37359719165499555, "grad_norm": 1.406480312347412, "learning_rate": 7.768806925712046e-06, "loss": 0.2623, "step": 20540 }, { "epoch": 0.37377907928481785, "grad_norm": 1.7051317691802979, "learning_rate": 7.766693139554313e-06, "loss": 0.2652, "step": 20550 }, { "epoch": 0.37396096691464015, "grad_norm": 1.9095157384872437, "learning_rate": 7.764578740292827e-06, "loss": 0.2696, "step": 20560 }, { "epoch": 0.37414285454446244, "grad_norm": 1.646267056465149, "learning_rate": 7.762463728661356e-06, "loss": 0.266, "step": 20570 }, { "epoch": 0.37432474217428474, "grad_norm": 1.8604466915130615, "learning_rate": 7.760348105393883e-06, "loss": 0.2654, "step": 20580 }, { "epoch": 0.37450662980410704, "grad_norm": 1.8096489906311035, "learning_rate": 7.758231871224602e-06, "loss": 0.2666, "step": 20590 }, { "epoch": 0.37468851743392934, "grad_norm": 1.609092116355896, "learning_rate": 7.756115026887922e-06, "loss": 0.2719, "step": 20600 }, { "epoch": 0.37487040506375163, "grad_norm": 1.740100622177124, "learning_rate": 7.75399757311846e-06, "loss": 0.2646, "step": 20610 }, { "epoch": 0.37505229269357393, "grad_norm": 1.664339303970337, "learning_rate": 7.751879510651045e-06, "loss": 0.259, "step": 20620 }, { "epoch": 0.3752341803233962, "grad_norm": 1.8535163402557373, "learning_rate": 7.749760840220717e-06, "loss": 0.2579, "step": 20630 }, { "epoch": 0.3754160679532185, "grad_norm": 1.7382168769836426, "learning_rate": 7.747641562562733e-06, "loss": 0.259, "step": 20640 }, { "epoch": 0.3755979555830408, "grad_norm": 1.7861309051513672, "learning_rate": 7.745521678412551e-06, "loss": 0.2701, "step": 20650 }, { "epoch": 0.3757798432128631, "grad_norm": 1.5156443119049072, "learning_rate": 7.743401188505846e-06, "loss": 0.2621, "step": 20660 }, { "epoch": 0.3759617308426854, "grad_norm": 1.8175939321517944, "learning_rate": 7.741280093578501e-06, "loss": 0.2687, "step": 20670 }, { "epoch": 0.37614361847250766, "grad_norm": 1.7671836614608765, "learning_rate": 7.739158394366611e-06, "loss": 0.2716, "step": 20680 }, { "epoch": 0.37632550610232995, "grad_norm": 1.7070986032485962, "learning_rate": 7.737036091606477e-06, "loss": 0.2631, "step": 20690 }, { "epoch": 0.37650739373215225, "grad_norm": 1.570290207862854, "learning_rate": 7.734913186034612e-06, "loss": 0.2672, "step": 20700 }, { "epoch": 0.37668928136197455, "grad_norm": 1.742055058479309, "learning_rate": 7.73278967838774e-06, "loss": 0.2622, "step": 20710 }, { "epoch": 0.37687116899179685, "grad_norm": 2.1230714321136475, "learning_rate": 7.73066556940279e-06, "loss": 0.2622, "step": 20720 }, { "epoch": 0.37705305662161914, "grad_norm": 1.7661402225494385, "learning_rate": 7.7285408598169e-06, "loss": 0.2685, "step": 20730 }, { "epoch": 0.37723494425144144, "grad_norm": 1.579775094985962, "learning_rate": 7.72641555036742e-06, "loss": 0.2603, "step": 20740 }, { "epoch": 0.37741683188126374, "grad_norm": 1.7444466352462769, "learning_rate": 7.724289641791907e-06, "loss": 0.2639, "step": 20750 }, { "epoch": 0.37759871951108603, "grad_norm": 1.7103819847106934, "learning_rate": 7.722163134828122e-06, "loss": 0.2635, "step": 20760 }, { "epoch": 0.37778060714090833, "grad_norm": 1.819414734840393, "learning_rate": 7.720036030214039e-06, "loss": 0.2599, "step": 20770 }, { "epoch": 0.37796249477073063, "grad_norm": 1.5377068519592285, "learning_rate": 7.717908328687836e-06, "loss": 0.2658, "step": 20780 }, { "epoch": 0.3781443824005529, "grad_norm": 1.6223758459091187, "learning_rate": 7.7157800309879e-06, "loss": 0.2613, "step": 20790 }, { "epoch": 0.3783262700303752, "grad_norm": 1.5454790592193604, "learning_rate": 7.713651137852821e-06, "loss": 0.2667, "step": 20800 }, { "epoch": 0.3785081576601975, "grad_norm": 1.5975171327590942, "learning_rate": 7.711521650021405e-06, "loss": 0.2591, "step": 20810 }, { "epoch": 0.3786900452900198, "grad_norm": 1.4960099458694458, "learning_rate": 7.70939156823265e-06, "loss": 0.2656, "step": 20820 }, { "epoch": 0.3788719329198421, "grad_norm": 1.689795732498169, "learning_rate": 7.707260893225776e-06, "loss": 0.2609, "step": 20830 }, { "epoch": 0.3790538205496644, "grad_norm": 1.6433364152908325, "learning_rate": 7.705129625740194e-06, "loss": 0.2694, "step": 20840 }, { "epoch": 0.3792357081794867, "grad_norm": 1.6781716346740723, "learning_rate": 7.702997766515534e-06, "loss": 0.2636, "step": 20850 }, { "epoch": 0.379417595809309, "grad_norm": 1.9900877475738525, "learning_rate": 7.70086531629162e-06, "loss": 0.2625, "step": 20860 }, { "epoch": 0.3795994834391313, "grad_norm": 2.1116888523101807, "learning_rate": 7.69873227580849e-06, "loss": 0.2705, "step": 20870 }, { "epoch": 0.3797813710689536, "grad_norm": 1.5508040189743042, "learning_rate": 7.69659864580638e-06, "loss": 0.2639, "step": 20880 }, { "epoch": 0.3799632586987759, "grad_norm": 1.5948699712753296, "learning_rate": 7.694464427025737e-06, "loss": 0.261, "step": 20890 }, { "epoch": 0.3801451463285982, "grad_norm": 1.6207411289215088, "learning_rate": 7.692329620207206e-06, "loss": 0.2675, "step": 20900 }, { "epoch": 0.3803270339584205, "grad_norm": 1.5402151346206665, "learning_rate": 7.690194226091638e-06, "loss": 0.2653, "step": 20910 }, { "epoch": 0.3805089215882428, "grad_norm": 1.81956946849823, "learning_rate": 7.688058245420092e-06, "loss": 0.2672, "step": 20920 }, { "epoch": 0.3806908092180651, "grad_norm": 1.9283881187438965, "learning_rate": 7.685921678933825e-06, "loss": 0.2628, "step": 20930 }, { "epoch": 0.3808726968478874, "grad_norm": 1.8204073905944824, "learning_rate": 7.683784527374302e-06, "loss": 0.2643, "step": 20940 }, { "epoch": 0.3810545844777097, "grad_norm": 1.8113664388656616, "learning_rate": 7.681646791483187e-06, "loss": 0.2664, "step": 20950 }, { "epoch": 0.381236472107532, "grad_norm": 1.4469568729400635, "learning_rate": 7.679508472002346e-06, "loss": 0.2649, "step": 20960 }, { "epoch": 0.3814183597373543, "grad_norm": 1.6579619646072388, "learning_rate": 7.677369569673856e-06, "loss": 0.2649, "step": 20970 }, { "epoch": 0.38160024736717657, "grad_norm": 1.6956819295883179, "learning_rate": 7.675230085239986e-06, "loss": 0.2647, "step": 20980 }, { "epoch": 0.38178213499699887, "grad_norm": 1.6108160018920898, "learning_rate": 7.67309001944321e-06, "loss": 0.2659, "step": 20990 }, { "epoch": 0.38196402262682116, "grad_norm": 1.7158187627792358, "learning_rate": 7.670949373026207e-06, "loss": 0.265, "step": 21000 }, { "epoch": 0.38214591025664346, "grad_norm": 1.5772737264633179, "learning_rate": 7.668808146731858e-06, "loss": 0.2598, "step": 21010 }, { "epoch": 0.38232779788646576, "grad_norm": 1.8332700729370117, "learning_rate": 7.666666341303238e-06, "loss": 0.2685, "step": 21020 }, { "epoch": 0.38250968551628806, "grad_norm": 1.5997251272201538, "learning_rate": 7.664523957483632e-06, "loss": 0.2679, "step": 21030 }, { "epoch": 0.38269157314611035, "grad_norm": 1.949504017829895, "learning_rate": 7.662380996016519e-06, "loss": 0.2634, "step": 21040 }, { "epoch": 0.38287346077593265, "grad_norm": 1.4491320848464966, "learning_rate": 7.66023745764558e-06, "loss": 0.2581, "step": 21050 }, { "epoch": 0.38305534840575495, "grad_norm": 2.594693899154663, "learning_rate": 7.658093343114698e-06, "loss": 0.274, "step": 21060 }, { "epoch": 0.38323723603557724, "grad_norm": 1.6544212102890015, "learning_rate": 7.655948653167958e-06, "loss": 0.2715, "step": 21070 }, { "epoch": 0.38341912366539954, "grad_norm": 1.6347627639770508, "learning_rate": 7.653803388549638e-06, "loss": 0.2608, "step": 21080 }, { "epoch": 0.38360101129522184, "grad_norm": 1.6022043228149414, "learning_rate": 7.651657550004221e-06, "loss": 0.2638, "step": 21090 }, { "epoch": 0.3837828989250441, "grad_norm": 1.7370197772979736, "learning_rate": 7.649511138276387e-06, "loss": 0.2644, "step": 21100 }, { "epoch": 0.3839647865548664, "grad_norm": 1.909282922744751, "learning_rate": 7.647364154111015e-06, "loss": 0.2588, "step": 21110 }, { "epoch": 0.3841466741846887, "grad_norm": 1.7320462465286255, "learning_rate": 7.645216598253184e-06, "loss": 0.2618, "step": 21120 }, { "epoch": 0.38432856181451097, "grad_norm": 1.6371657848358154, "learning_rate": 7.643068471448168e-06, "loss": 0.2634, "step": 21130 }, { "epoch": 0.38451044944433327, "grad_norm": 1.6651579141616821, "learning_rate": 7.640919774441441e-06, "loss": 0.2544, "step": 21140 }, { "epoch": 0.38469233707415557, "grad_norm": 1.5182256698608398, "learning_rate": 7.638770507978679e-06, "loss": 0.2583, "step": 21150 }, { "epoch": 0.38487422470397786, "grad_norm": 1.8592920303344727, "learning_rate": 7.63662067280575e-06, "loss": 0.2679, "step": 21160 }, { "epoch": 0.38505611233380016, "grad_norm": 2.0336062908172607, "learning_rate": 7.634470269668721e-06, "loss": 0.2737, "step": 21170 }, { "epoch": 0.38523799996362246, "grad_norm": 2.10107684135437, "learning_rate": 7.632319299313859e-06, "loss": 0.2579, "step": 21180 }, { "epoch": 0.38541988759344475, "grad_norm": 1.5563379526138306, "learning_rate": 7.630167762487619e-06, "loss": 0.2581, "step": 21190 }, { "epoch": 0.38560177522326705, "grad_norm": 1.5381041765213013, "learning_rate": 7.628015659936666e-06, "loss": 0.2632, "step": 21200 }, { "epoch": 0.38578366285308935, "grad_norm": 1.633660078048706, "learning_rate": 7.6258629924078495e-06, "loss": 0.2595, "step": 21210 }, { "epoch": 0.38596555048291165, "grad_norm": 1.8849610090255737, "learning_rate": 7.623709760648223e-06, "loss": 0.2606, "step": 21220 }, { "epoch": 0.38614743811273394, "grad_norm": 1.9530928134918213, "learning_rate": 7.621555965405031e-06, "loss": 0.2672, "step": 21230 }, { "epoch": 0.38632932574255624, "grad_norm": 1.7186057567596436, "learning_rate": 7.6194016074257135e-06, "loss": 0.2682, "step": 21240 }, { "epoch": 0.38651121337237854, "grad_norm": 1.902069091796875, "learning_rate": 7.6172466874579096e-06, "loss": 0.2622, "step": 21250 }, { "epoch": 0.38669310100220083, "grad_norm": 1.805505633354187, "learning_rate": 7.61509120624945e-06, "loss": 0.2603, "step": 21260 }, { "epoch": 0.38687498863202313, "grad_norm": 1.6584793329238892, "learning_rate": 7.612935164548362e-06, "loss": 0.2641, "step": 21270 }, { "epoch": 0.38705687626184543, "grad_norm": 1.6730992794036865, "learning_rate": 7.610778563102867e-06, "loss": 0.2693, "step": 21280 }, { "epoch": 0.3872387638916677, "grad_norm": 1.8693163394927979, "learning_rate": 7.608621402661379e-06, "loss": 0.2682, "step": 21290 }, { "epoch": 0.38742065152149, "grad_norm": 1.567141056060791, "learning_rate": 7.60646368397251e-06, "loss": 0.2649, "step": 21300 }, { "epoch": 0.3876025391513123, "grad_norm": 1.6659691333770752, "learning_rate": 7.604305407785062e-06, "loss": 0.2659, "step": 21310 }, { "epoch": 0.3877844267811346, "grad_norm": 1.6289658546447754, "learning_rate": 7.602146574848032e-06, "loss": 0.2605, "step": 21320 }, { "epoch": 0.3879663144109569, "grad_norm": 1.8083112239837646, "learning_rate": 7.5999871859106075e-06, "loss": 0.2662, "step": 21330 }, { "epoch": 0.3881482020407792, "grad_norm": 1.719024419784546, "learning_rate": 7.597827241722175e-06, "loss": 0.2672, "step": 21340 }, { "epoch": 0.3883300896706015, "grad_norm": 1.947725772857666, "learning_rate": 7.595666743032306e-06, "loss": 0.2647, "step": 21350 }, { "epoch": 0.3885119773004238, "grad_norm": 1.649687647819519, "learning_rate": 7.593505690590771e-06, "loss": 0.2594, "step": 21360 }, { "epoch": 0.3886938649302461, "grad_norm": 1.9004639387130737, "learning_rate": 7.5913440851475305e-06, "loss": 0.2646, "step": 21370 }, { "epoch": 0.3888757525600684, "grad_norm": 1.59148108959198, "learning_rate": 7.589181927452734e-06, "loss": 0.2633, "step": 21380 }, { "epoch": 0.3890576401898907, "grad_norm": 1.8086068630218506, "learning_rate": 7.587019218256729e-06, "loss": 0.2553, "step": 21390 }, { "epoch": 0.389239527819713, "grad_norm": 1.8227794170379639, "learning_rate": 7.584855958310048e-06, "loss": 0.2709, "step": 21400 }, { "epoch": 0.3894214154495353, "grad_norm": 1.9718068838119507, "learning_rate": 7.582692148363416e-06, "loss": 0.2665, "step": 21410 }, { "epoch": 0.3896033030793576, "grad_norm": 1.6611214876174927, "learning_rate": 7.580527789167753e-06, "loss": 0.2643, "step": 21420 }, { "epoch": 0.3897851907091799, "grad_norm": 1.6950329542160034, "learning_rate": 7.578362881474164e-06, "loss": 0.2735, "step": 21430 }, { "epoch": 0.3899670783390022, "grad_norm": 1.7212209701538086, "learning_rate": 7.576197426033948e-06, "loss": 0.2649, "step": 21440 }, { "epoch": 0.3901489659688245, "grad_norm": 1.5802395343780518, "learning_rate": 7.574031423598595e-06, "loss": 0.2656, "step": 21450 }, { "epoch": 0.3903308535986468, "grad_norm": 1.6700574159622192, "learning_rate": 7.57186487491978e-06, "loss": 0.2666, "step": 21460 }, { "epoch": 0.3905127412284691, "grad_norm": 1.80098557472229, "learning_rate": 7.569697780749372e-06, "loss": 0.2681, "step": 21470 }, { "epoch": 0.39069462885829137, "grad_norm": 1.6656060218811035, "learning_rate": 7.567530141839427e-06, "loss": 0.2613, "step": 21480 }, { "epoch": 0.39087651648811367, "grad_norm": 1.5610135793685913, "learning_rate": 7.565361958942194e-06, "loss": 0.2654, "step": 21490 }, { "epoch": 0.39105840411793596, "grad_norm": 1.7718085050582886, "learning_rate": 7.563193232810102e-06, "loss": 0.2605, "step": 21500 }, { "epoch": 0.39124029174775826, "grad_norm": 1.647352695465088, "learning_rate": 7.5610239641957775e-06, "loss": 0.2634, "step": 21510 }, { "epoch": 0.3914221793775805, "grad_norm": 1.793739914894104, "learning_rate": 7.558854153852032e-06, "loss": 0.276, "step": 21520 }, { "epoch": 0.3916040670074028, "grad_norm": 1.8464970588684082, "learning_rate": 7.556683802531864e-06, "loss": 0.2634, "step": 21530 }, { "epoch": 0.3917859546372251, "grad_norm": 1.8392447233200073, "learning_rate": 7.554512910988463e-06, "loss": 0.2656, "step": 21540 }, { "epoch": 0.3919678422670474, "grad_norm": 1.816370964050293, "learning_rate": 7.552341479975199e-06, "loss": 0.2607, "step": 21550 }, { "epoch": 0.3921497298968697, "grad_norm": 1.538733959197998, "learning_rate": 7.550169510245638e-06, "loss": 0.2636, "step": 21560 }, { "epoch": 0.392331617526692, "grad_norm": 1.8174868822097778, "learning_rate": 7.547997002553526e-06, "loss": 0.2674, "step": 21570 }, { "epoch": 0.3925135051565143, "grad_norm": 1.9311929941177368, "learning_rate": 7.5458239576528e-06, "loss": 0.2662, "step": 21580 }, { "epoch": 0.3926953927863366, "grad_norm": 2.058086633682251, "learning_rate": 7.543650376297583e-06, "loss": 0.2625, "step": 21590 }, { "epoch": 0.3928772804161589, "grad_norm": 1.7169824838638306, "learning_rate": 7.541476259242181e-06, "loss": 0.2578, "step": 21600 }, { "epoch": 0.3930591680459812, "grad_norm": 1.8344030380249023, "learning_rate": 7.539301607241089e-06, "loss": 0.2623, "step": 21610 }, { "epoch": 0.3932410556758035, "grad_norm": 1.7528942823410034, "learning_rate": 7.537126421048985e-06, "loss": 0.2669, "step": 21620 }, { "epoch": 0.39342294330562577, "grad_norm": 1.9722665548324585, "learning_rate": 7.534950701420735e-06, "loss": 0.2649, "step": 21630 }, { "epoch": 0.39360483093544807, "grad_norm": 1.8792246580123901, "learning_rate": 7.532774449111392e-06, "loss": 0.2653, "step": 21640 }, { "epoch": 0.39378671856527037, "grad_norm": 1.5632494688034058, "learning_rate": 7.530597664876186e-06, "loss": 0.2618, "step": 21650 }, { "epoch": 0.39396860619509266, "grad_norm": 1.9905575513839722, "learning_rate": 7.528420349470541e-06, "loss": 0.2635, "step": 21660 }, { "epoch": 0.39415049382491496, "grad_norm": 1.8262239694595337, "learning_rate": 7.526242503650058e-06, "loss": 0.2502, "step": 21670 }, { "epoch": 0.39433238145473726, "grad_norm": 1.4856003522872925, "learning_rate": 7.524064128170524e-06, "loss": 0.2539, "step": 21680 }, { "epoch": 0.39451426908455955, "grad_norm": 1.6429799795150757, "learning_rate": 7.521885223787914e-06, "loss": 0.258, "step": 21690 }, { "epoch": 0.39469615671438185, "grad_norm": 1.549213171005249, "learning_rate": 7.519705791258381e-06, "loss": 0.2615, "step": 21700 }, { "epoch": 0.39487804434420415, "grad_norm": 1.811834692955017, "learning_rate": 7.5175258313382675e-06, "loss": 0.2597, "step": 21710 }, { "epoch": 0.39505993197402645, "grad_norm": 1.7459009885787964, "learning_rate": 7.51534534478409e-06, "loss": 0.2652, "step": 21720 }, { "epoch": 0.39524181960384874, "grad_norm": 1.7048277854919434, "learning_rate": 7.513164332352555e-06, "loss": 0.2628, "step": 21730 }, { "epoch": 0.39542370723367104, "grad_norm": 1.8092525005340576, "learning_rate": 7.510982794800551e-06, "loss": 0.2574, "step": 21740 }, { "epoch": 0.39560559486349334, "grad_norm": 1.9943004846572876, "learning_rate": 7.508800732885144e-06, "loss": 0.2613, "step": 21750 }, { "epoch": 0.39578748249331563, "grad_norm": 1.5456520318984985, "learning_rate": 7.506618147363589e-06, "loss": 0.268, "step": 21760 }, { "epoch": 0.39596937012313793, "grad_norm": 1.66286301612854, "learning_rate": 7.5044350389933155e-06, "loss": 0.2686, "step": 21770 }, { "epoch": 0.39615125775296023, "grad_norm": 1.7242809534072876, "learning_rate": 7.5022514085319385e-06, "loss": 0.2629, "step": 21780 }, { "epoch": 0.3963331453827825, "grad_norm": 1.730487585067749, "learning_rate": 7.500067256737256e-06, "loss": 0.2716, "step": 21790 }, { "epoch": 0.3965150330126048, "grad_norm": 1.973596453666687, "learning_rate": 7.497882584367239e-06, "loss": 0.2622, "step": 21800 }, { "epoch": 0.3966969206424271, "grad_norm": 1.9054383039474487, "learning_rate": 7.49569739218005e-06, "loss": 0.2605, "step": 21810 }, { "epoch": 0.3968788082722494, "grad_norm": 1.6594059467315674, "learning_rate": 7.493511680934024e-06, "loss": 0.2653, "step": 21820 }, { "epoch": 0.3970606959020717, "grad_norm": 1.8315117359161377, "learning_rate": 7.491325451387678e-06, "loss": 0.2672, "step": 21830 }, { "epoch": 0.397242583531894, "grad_norm": 1.7954097986221313, "learning_rate": 7.489138704299711e-06, "loss": 0.2616, "step": 21840 }, { "epoch": 0.3974244711617163, "grad_norm": 1.7754721641540527, "learning_rate": 7.486951440428998e-06, "loss": 0.2636, "step": 21850 }, { "epoch": 0.3976063587915386, "grad_norm": 2.151930332183838, "learning_rate": 7.484763660534599e-06, "loss": 0.2682, "step": 21860 }, { "epoch": 0.3977882464213609, "grad_norm": 1.808924674987793, "learning_rate": 7.482575365375746e-06, "loss": 0.2646, "step": 21870 }, { "epoch": 0.3979701340511832, "grad_norm": 1.6022930145263672, "learning_rate": 7.4803865557118524e-06, "loss": 0.2669, "step": 21880 }, { "epoch": 0.3981520216810055, "grad_norm": 1.6422566175460815, "learning_rate": 7.478197232302516e-06, "loss": 0.2606, "step": 21890 }, { "epoch": 0.3983339093108278, "grad_norm": 1.7295604944229126, "learning_rate": 7.476007395907504e-06, "loss": 0.2635, "step": 21900 }, { "epoch": 0.3985157969406501, "grad_norm": 1.711287021636963, "learning_rate": 7.4738170472867656e-06, "loss": 0.2643, "step": 21910 }, { "epoch": 0.3986976845704724, "grad_norm": 1.5771946907043457, "learning_rate": 7.471626187200428e-06, "loss": 0.2627, "step": 21920 }, { "epoch": 0.3988795722002947, "grad_norm": 1.749821662902832, "learning_rate": 7.4694348164087985e-06, "loss": 0.2625, "step": 21930 }, { "epoch": 0.3990614598301169, "grad_norm": 1.7415461540222168, "learning_rate": 7.467242935672355e-06, "loss": 0.2618, "step": 21940 }, { "epoch": 0.3992433474599392, "grad_norm": 1.5609982013702393, "learning_rate": 7.4650505457517565e-06, "loss": 0.2565, "step": 21950 }, { "epoch": 0.3994252350897615, "grad_norm": 1.702633261680603, "learning_rate": 7.462857647407841e-06, "loss": 0.2609, "step": 21960 }, { "epoch": 0.3996071227195838, "grad_norm": 1.543365478515625, "learning_rate": 7.460664241401617e-06, "loss": 0.2588, "step": 21970 }, { "epoch": 0.3997890103494061, "grad_norm": 1.570466160774231, "learning_rate": 7.458470328494273e-06, "loss": 0.2595, "step": 21980 }, { "epoch": 0.3999708979792284, "grad_norm": 1.8221886157989502, "learning_rate": 7.456275909447176e-06, "loss": 0.2619, "step": 21990 }, { "epoch": 0.4001527856090507, "grad_norm": 1.8005088567733765, "learning_rate": 7.454080985021861e-06, "loss": 0.26, "step": 22000 }, { "epoch": 0.400334673238873, "grad_norm": 1.6524486541748047, "learning_rate": 7.451885555980046e-06, "loss": 0.2722, "step": 22010 }, { "epoch": 0.4005165608686953, "grad_norm": 1.7118220329284668, "learning_rate": 7.44968962308362e-06, "loss": 0.2548, "step": 22020 }, { "epoch": 0.4006984484985176, "grad_norm": 1.8523120880126953, "learning_rate": 7.447493187094648e-06, "loss": 0.2699, "step": 22030 }, { "epoch": 0.4008803361283399, "grad_norm": 1.6891465187072754, "learning_rate": 7.44529624877537e-06, "loss": 0.256, "step": 22040 }, { "epoch": 0.4010622237581622, "grad_norm": 1.6645804643630981, "learning_rate": 7.443098808888197e-06, "loss": 0.2582, "step": 22050 }, { "epoch": 0.4012441113879845, "grad_norm": 1.7170202732086182, "learning_rate": 7.440900868195719e-06, "loss": 0.2554, "step": 22060 }, { "epoch": 0.4014259990178068, "grad_norm": 1.584946632385254, "learning_rate": 7.438702427460698e-06, "loss": 0.2529, "step": 22070 }, { "epoch": 0.4016078866476291, "grad_norm": 1.7264631986618042, "learning_rate": 7.436503487446068e-06, "loss": 0.2652, "step": 22080 }, { "epoch": 0.4017897742774514, "grad_norm": 1.6033161878585815, "learning_rate": 7.434304048914939e-06, "loss": 0.2603, "step": 22090 }, { "epoch": 0.4019716619072737, "grad_norm": 1.7151727676391602, "learning_rate": 7.432104112630591e-06, "loss": 0.2606, "step": 22100 }, { "epoch": 0.402153549537096, "grad_norm": 1.6288707256317139, "learning_rate": 7.429903679356477e-06, "loss": 0.2612, "step": 22110 }, { "epoch": 0.4023354371669183, "grad_norm": 1.7217378616333008, "learning_rate": 7.4277027498562255e-06, "loss": 0.2632, "step": 22120 }, { "epoch": 0.40251732479674057, "grad_norm": 1.5516083240509033, "learning_rate": 7.4255013248936355e-06, "loss": 0.2543, "step": 22130 }, { "epoch": 0.40269921242656287, "grad_norm": 1.78048574924469, "learning_rate": 7.4232994052326766e-06, "loss": 0.2593, "step": 22140 }, { "epoch": 0.40288110005638517, "grad_norm": 1.4978456497192383, "learning_rate": 7.421096991637493e-06, "loss": 0.2566, "step": 22150 }, { "epoch": 0.40306298768620746, "grad_norm": 1.654425024986267, "learning_rate": 7.418894084872397e-06, "loss": 0.2665, "step": 22160 }, { "epoch": 0.40324487531602976, "grad_norm": 1.9107543230056763, "learning_rate": 7.416690685701873e-06, "loss": 0.2622, "step": 22170 }, { "epoch": 0.40342676294585206, "grad_norm": 1.6833246946334839, "learning_rate": 7.414486794890577e-06, "loss": 0.2609, "step": 22180 }, { "epoch": 0.40360865057567435, "grad_norm": 1.735306739807129, "learning_rate": 7.4122824132033385e-06, "loss": 0.2645, "step": 22190 }, { "epoch": 0.40379053820549665, "grad_norm": 1.503095030784607, "learning_rate": 7.410077541405152e-06, "loss": 0.2654, "step": 22200 }, { "epoch": 0.40397242583531895, "grad_norm": 1.5784696340560913, "learning_rate": 7.407872180261183e-06, "loss": 0.2669, "step": 22210 }, { "epoch": 0.40415431346514125, "grad_norm": 1.856415033340454, "learning_rate": 7.4056663305367704e-06, "loss": 0.2726, "step": 22220 }, { "epoch": 0.40433620109496354, "grad_norm": 1.61407470703125, "learning_rate": 7.403459992997421e-06, "loss": 0.2617, "step": 22230 }, { "epoch": 0.40451808872478584, "grad_norm": 1.6668280363082886, "learning_rate": 7.401253168408807e-06, "loss": 0.2634, "step": 22240 }, { "epoch": 0.40469997635460814, "grad_norm": 1.817503571510315, "learning_rate": 7.399045857536779e-06, "loss": 0.2616, "step": 22250 }, { "epoch": 0.40488186398443043, "grad_norm": 1.6308964490890503, "learning_rate": 7.396838061147343e-06, "loss": 0.2559, "step": 22260 }, { "epoch": 0.40506375161425273, "grad_norm": 1.7172471284866333, "learning_rate": 7.394629780006687e-06, "loss": 0.2604, "step": 22270 }, { "epoch": 0.40524563924407503, "grad_norm": 1.8313478231430054, "learning_rate": 7.392421014881158e-06, "loss": 0.2631, "step": 22280 }, { "epoch": 0.4054275268738973, "grad_norm": 1.5881037712097168, "learning_rate": 7.390211766537275e-06, "loss": 0.2636, "step": 22290 }, { "epoch": 0.4056094145037196, "grad_norm": 1.564146876335144, "learning_rate": 7.388002035741722e-06, "loss": 0.2569, "step": 22300 }, { "epoch": 0.4057913021335419, "grad_norm": 1.5970245599746704, "learning_rate": 7.385791823261356e-06, "loss": 0.2665, "step": 22310 }, { "epoch": 0.4059731897633642, "grad_norm": 1.6829110383987427, "learning_rate": 7.383581129863195e-06, "loss": 0.2633, "step": 22320 }, { "epoch": 0.4061550773931865, "grad_norm": 1.8969204425811768, "learning_rate": 7.381369956314425e-06, "loss": 0.2621, "step": 22330 }, { "epoch": 0.4063369650230088, "grad_norm": 2.0180141925811768, "learning_rate": 7.379158303382401e-06, "loss": 0.2688, "step": 22340 }, { "epoch": 0.4065188526528311, "grad_norm": 1.5323492288589478, "learning_rate": 7.376946171834644e-06, "loss": 0.2615, "step": 22350 }, { "epoch": 0.40670074028265335, "grad_norm": 1.4518324136734009, "learning_rate": 7.37473356243884e-06, "loss": 0.2638, "step": 22360 }, { "epoch": 0.40688262791247565, "grad_norm": 1.604799509048462, "learning_rate": 7.37252047596284e-06, "loss": 0.2604, "step": 22370 }, { "epoch": 0.40706451554229794, "grad_norm": 1.6488336324691772, "learning_rate": 7.370306913174664e-06, "loss": 0.2657, "step": 22380 }, { "epoch": 0.40724640317212024, "grad_norm": 1.6509008407592773, "learning_rate": 7.36809287484249e-06, "loss": 0.2706, "step": 22390 }, { "epoch": 0.40742829080194254, "grad_norm": 1.6157000064849854, "learning_rate": 7.365878361734673e-06, "loss": 0.2653, "step": 22400 }, { "epoch": 0.40761017843176484, "grad_norm": 1.7007558345794678, "learning_rate": 7.363663374619721e-06, "loss": 0.2552, "step": 22410 }, { "epoch": 0.40779206606158713, "grad_norm": 1.8059935569763184, "learning_rate": 7.361447914266314e-06, "loss": 0.2617, "step": 22420 }, { "epoch": 0.40797395369140943, "grad_norm": 1.9607627391815186, "learning_rate": 7.359231981443292e-06, "loss": 0.2617, "step": 22430 }, { "epoch": 0.4081558413212317, "grad_norm": 1.7495924234390259, "learning_rate": 7.357015576919659e-06, "loss": 0.2653, "step": 22440 }, { "epoch": 0.408337728951054, "grad_norm": 1.7706396579742432, "learning_rate": 7.354798701464588e-06, "loss": 0.2548, "step": 22450 }, { "epoch": 0.4085196165808763, "grad_norm": 1.6404861211776733, "learning_rate": 7.352581355847409e-06, "loss": 0.263, "step": 22460 }, { "epoch": 0.4087015042106986, "grad_norm": 1.6823821067810059, "learning_rate": 7.350363540837618e-06, "loss": 0.2631, "step": 22470 }, { "epoch": 0.4088833918405209, "grad_norm": 1.531221866607666, "learning_rate": 7.348145257204874e-06, "loss": 0.2663, "step": 22480 }, { "epoch": 0.4090652794703432, "grad_norm": 1.5970299243927002, "learning_rate": 7.345926505718997e-06, "loss": 0.2603, "step": 22490 }, { "epoch": 0.4092471671001655, "grad_norm": 1.846611499786377, "learning_rate": 7.343707287149975e-06, "loss": 0.2515, "step": 22500 }, { "epoch": 0.4094290547299878, "grad_norm": 1.73561692237854, "learning_rate": 7.341487602267948e-06, "loss": 0.2633, "step": 22510 }, { "epoch": 0.4096109423598101, "grad_norm": 1.9377318620681763, "learning_rate": 7.339267451843228e-06, "loss": 0.2621, "step": 22520 }, { "epoch": 0.4097928299896324, "grad_norm": 1.6589305400848389, "learning_rate": 7.337046836646281e-06, "loss": 0.2605, "step": 22530 }, { "epoch": 0.4099747176194547, "grad_norm": 1.6283626556396484, "learning_rate": 7.33482575744774e-06, "loss": 0.2675, "step": 22540 }, { "epoch": 0.410156605249277, "grad_norm": 1.8639111518859863, "learning_rate": 7.332604215018396e-06, "loss": 0.2602, "step": 22550 }, { "epoch": 0.4103384928790993, "grad_norm": 1.8008445501327515, "learning_rate": 7.3303822101292e-06, "loss": 0.2676, "step": 22560 }, { "epoch": 0.4105203805089216, "grad_norm": 1.6590187549591064, "learning_rate": 7.328159743551265e-06, "loss": 0.2568, "step": 22570 }, { "epoch": 0.4107022681387439, "grad_norm": 1.785651445388794, "learning_rate": 7.3259368160558645e-06, "loss": 0.2602, "step": 22580 }, { "epoch": 0.4108841557685662, "grad_norm": 1.883351445198059, "learning_rate": 7.323713428414431e-06, "loss": 0.2667, "step": 22590 }, { "epoch": 0.4110660433983885, "grad_norm": 1.6446287631988525, "learning_rate": 7.321489581398557e-06, "loss": 0.2647, "step": 22600 }, { "epoch": 0.4112479310282108, "grad_norm": 1.7120455503463745, "learning_rate": 7.319265275779995e-06, "loss": 0.2618, "step": 22610 }, { "epoch": 0.4114298186580331, "grad_norm": 1.7611557245254517, "learning_rate": 7.317040512330656e-06, "loss": 0.2602, "step": 22620 }, { "epoch": 0.4116117062878554, "grad_norm": 1.6165378093719482, "learning_rate": 7.31481529182261e-06, "loss": 0.2608, "step": 22630 }, { "epoch": 0.41179359391767767, "grad_norm": 1.5006041526794434, "learning_rate": 7.312589615028086e-06, "loss": 0.2539, "step": 22640 }, { "epoch": 0.41197548154749997, "grad_norm": 1.54118812084198, "learning_rate": 7.3103634827194704e-06, "loss": 0.2618, "step": 22650 }, { "epoch": 0.41215736917732226, "grad_norm": 1.555252194404602, "learning_rate": 7.308136895669308e-06, "loss": 0.2498, "step": 22660 }, { "epoch": 0.41233925680714456, "grad_norm": 1.6434202194213867, "learning_rate": 7.3059098546503e-06, "loss": 0.2621, "step": 22670 }, { "epoch": 0.41252114443696686, "grad_norm": 1.6613385677337646, "learning_rate": 7.303682360435311e-06, "loss": 0.2585, "step": 22680 }, { "epoch": 0.41270303206678915, "grad_norm": 1.640413522720337, "learning_rate": 7.301454413797356e-06, "loss": 0.2629, "step": 22690 }, { "epoch": 0.41288491969661145, "grad_norm": 1.7014703750610352, "learning_rate": 7.299226015509612e-06, "loss": 0.2628, "step": 22700 }, { "epoch": 0.41306680732643375, "grad_norm": 1.6797107458114624, "learning_rate": 7.296997166345406e-06, "loss": 0.2592, "step": 22710 }, { "epoch": 0.41324869495625605, "grad_norm": 1.734316349029541, "learning_rate": 7.294767867078231e-06, "loss": 0.2587, "step": 22720 }, { "epoch": 0.41343058258607834, "grad_norm": 1.5630090236663818, "learning_rate": 7.292538118481728e-06, "loss": 0.2609, "step": 22730 }, { "epoch": 0.41361247021590064, "grad_norm": 1.7092863321304321, "learning_rate": 7.290307921329698e-06, "loss": 0.2668, "step": 22740 }, { "epoch": 0.41379435784572294, "grad_norm": 1.7730224132537842, "learning_rate": 7.288077276396097e-06, "loss": 0.2548, "step": 22750 }, { "epoch": 0.41397624547554523, "grad_norm": 1.6557178497314453, "learning_rate": 7.285846184455034e-06, "loss": 0.2643, "step": 22760 }, { "epoch": 0.4141581331053675, "grad_norm": 1.3327089548110962, "learning_rate": 7.283614646280779e-06, "loss": 0.2541, "step": 22770 }, { "epoch": 0.4143400207351898, "grad_norm": 1.8421602249145508, "learning_rate": 7.281382662647749e-06, "loss": 0.2663, "step": 22780 }, { "epoch": 0.41452190836501207, "grad_norm": 1.7708146572113037, "learning_rate": 7.2791502343305235e-06, "loss": 0.2618, "step": 22790 }, { "epoch": 0.41470379599483437, "grad_norm": 1.9023061990737915, "learning_rate": 7.276917362103829e-06, "loss": 0.2677, "step": 22800 }, { "epoch": 0.41488568362465666, "grad_norm": 1.718381404876709, "learning_rate": 7.274684046742551e-06, "loss": 0.2648, "step": 22810 }, { "epoch": 0.41506757125447896, "grad_norm": 1.7327731847763062, "learning_rate": 7.272450289021729e-06, "loss": 0.2646, "step": 22820 }, { "epoch": 0.41524945888430126, "grad_norm": 2.1224005222320557, "learning_rate": 7.270216089716552e-06, "loss": 0.263, "step": 22830 }, { "epoch": 0.41543134651412356, "grad_norm": 1.4343117475509644, "learning_rate": 7.267981449602365e-06, "loss": 0.2626, "step": 22840 }, { "epoch": 0.41561323414394585, "grad_norm": 1.7810622453689575, "learning_rate": 7.265746369454668e-06, "loss": 0.2699, "step": 22850 }, { "epoch": 0.41579512177376815, "grad_norm": 1.4534573554992676, "learning_rate": 7.263510850049108e-06, "loss": 0.2583, "step": 22860 }, { "epoch": 0.41597700940359045, "grad_norm": 1.7024257183074951, "learning_rate": 7.2612748921614905e-06, "loss": 0.2655, "step": 22870 }, { "epoch": 0.41615889703341274, "grad_norm": 1.6665180921554565, "learning_rate": 7.259038496567767e-06, "loss": 0.2585, "step": 22880 }, { "epoch": 0.41634078466323504, "grad_norm": 1.7265182733535767, "learning_rate": 7.256801664044047e-06, "loss": 0.2609, "step": 22890 }, { "epoch": 0.41652267229305734, "grad_norm": 1.8167197704315186, "learning_rate": 7.254564395366588e-06, "loss": 0.257, "step": 22900 }, { "epoch": 0.41670455992287964, "grad_norm": 1.7696117162704468, "learning_rate": 7.252326691311801e-06, "loss": 0.2577, "step": 22910 }, { "epoch": 0.41688644755270193, "grad_norm": 1.5539966821670532, "learning_rate": 7.250088552656245e-06, "loss": 0.2674, "step": 22920 }, { "epoch": 0.41706833518252423, "grad_norm": 1.7334880828857422, "learning_rate": 7.247849980176631e-06, "loss": 0.2595, "step": 22930 }, { "epoch": 0.4172502228123465, "grad_norm": 1.8514277935028076, "learning_rate": 7.245610974649825e-06, "loss": 0.2601, "step": 22940 }, { "epoch": 0.4174321104421688, "grad_norm": 1.5700068473815918, "learning_rate": 7.243371536852836e-06, "loss": 0.2592, "step": 22950 }, { "epoch": 0.4176139980719911, "grad_norm": 1.719199299812317, "learning_rate": 7.241131667562828e-06, "loss": 0.2672, "step": 22960 }, { "epoch": 0.4177958857018134, "grad_norm": 1.5360980033874512, "learning_rate": 7.238891367557114e-06, "loss": 0.2539, "step": 22970 }, { "epoch": 0.4179777733316357, "grad_norm": 1.8769913911819458, "learning_rate": 7.236650637613154e-06, "loss": 0.2512, "step": 22980 }, { "epoch": 0.418159660961458, "grad_norm": 1.5994954109191895, "learning_rate": 7.23440947850856e-06, "loss": 0.2591, "step": 22990 }, { "epoch": 0.4183415485912803, "grad_norm": 1.678484559059143, "learning_rate": 7.232167891021092e-06, "loss": 0.2551, "step": 23000 }, { "epoch": 0.4185234362211026, "grad_norm": 1.816329002380371, "learning_rate": 7.229925875928657e-06, "loss": 0.2636, "step": 23010 }, { "epoch": 0.4187053238509249, "grad_norm": 1.5608588457107544, "learning_rate": 7.227683434009314e-06, "loss": 0.2615, "step": 23020 }, { "epoch": 0.4188872114807472, "grad_norm": 1.9045665264129639, "learning_rate": 7.225440566041266e-06, "loss": 0.2685, "step": 23030 }, { "epoch": 0.4190690991105695, "grad_norm": 1.7593884468078613, "learning_rate": 7.2231972728028695e-06, "loss": 0.2606, "step": 23040 }, { "epoch": 0.4192509867403918, "grad_norm": 1.7941136360168457, "learning_rate": 7.220953555072621e-06, "loss": 0.2666, "step": 23050 }, { "epoch": 0.4194328743702141, "grad_norm": 1.5511022806167603, "learning_rate": 7.218709413629171e-06, "loss": 0.2605, "step": 23060 }, { "epoch": 0.4196147620000364, "grad_norm": 1.8234378099441528, "learning_rate": 7.216464849251313e-06, "loss": 0.2653, "step": 23070 }, { "epoch": 0.4197966496298587, "grad_norm": 1.66615891456604, "learning_rate": 7.214219862717988e-06, "loss": 0.2611, "step": 23080 }, { "epoch": 0.419978537259681, "grad_norm": 1.8985381126403809, "learning_rate": 7.211974454808286e-06, "loss": 0.2583, "step": 23090 }, { "epoch": 0.4201604248895033, "grad_norm": 1.4992234706878662, "learning_rate": 7.209728626301441e-06, "loss": 0.2555, "step": 23100 }, { "epoch": 0.4203423125193256, "grad_norm": 1.673504114151001, "learning_rate": 7.207482377976831e-06, "loss": 0.2631, "step": 23110 }, { "epoch": 0.4205242001491479, "grad_norm": 1.7547123432159424, "learning_rate": 7.205235710613986e-06, "loss": 0.2544, "step": 23120 }, { "epoch": 0.4207060877789702, "grad_norm": 1.6461541652679443, "learning_rate": 7.202988624992575e-06, "loss": 0.2587, "step": 23130 }, { "epoch": 0.42088797540879247, "grad_norm": 1.5552862882614136, "learning_rate": 7.2007411218924135e-06, "loss": 0.2532, "step": 23140 }, { "epoch": 0.42106986303861477, "grad_norm": 2.0108094215393066, "learning_rate": 7.198493202093467e-06, "loss": 0.2605, "step": 23150 }, { "epoch": 0.42125175066843706, "grad_norm": 1.4508616924285889, "learning_rate": 7.196244866375836e-06, "loss": 0.2668, "step": 23160 }, { "epoch": 0.42143363829825936, "grad_norm": 1.6513675451278687, "learning_rate": 7.1939961155197755e-06, "loss": 0.2628, "step": 23170 }, { "epoch": 0.42161552592808166, "grad_norm": 1.8484716415405273, "learning_rate": 7.191746950305677e-06, "loss": 0.2638, "step": 23180 }, { "epoch": 0.4217974135579039, "grad_norm": 1.756378173828125, "learning_rate": 7.189497371514082e-06, "loss": 0.2623, "step": 23190 }, { "epoch": 0.4219793011877262, "grad_norm": 1.8915191888809204, "learning_rate": 7.187247379925669e-06, "loss": 0.2599, "step": 23200 }, { "epoch": 0.4221611888175485, "grad_norm": 1.6380882263183594, "learning_rate": 7.184996976321266e-06, "loss": 0.2599, "step": 23210 }, { "epoch": 0.4223430764473708, "grad_norm": 1.7213866710662842, "learning_rate": 7.182746161481838e-06, "loss": 0.2617, "step": 23220 }, { "epoch": 0.4225249640771931, "grad_norm": 1.5589219331741333, "learning_rate": 7.1804949361884966e-06, "loss": 0.2614, "step": 23230 }, { "epoch": 0.4227068517070154, "grad_norm": 1.5882095098495483, "learning_rate": 7.178243301222498e-06, "loss": 0.2698, "step": 23240 }, { "epoch": 0.4228887393368377, "grad_norm": 1.521000623703003, "learning_rate": 7.175991257365236e-06, "loss": 0.2617, "step": 23250 }, { "epoch": 0.42307062696666, "grad_norm": 1.5384058952331543, "learning_rate": 7.173738805398245e-06, "loss": 0.255, "step": 23260 }, { "epoch": 0.4232525145964823, "grad_norm": 1.8060040473937988, "learning_rate": 7.171485946103207e-06, "loss": 0.2627, "step": 23270 }, { "epoch": 0.4234344022263046, "grad_norm": 1.7125217914581299, "learning_rate": 7.1692326802619406e-06, "loss": 0.2565, "step": 23280 }, { "epoch": 0.42361628985612687, "grad_norm": 1.6906009912490845, "learning_rate": 7.166979008656409e-06, "loss": 0.2667, "step": 23290 }, { "epoch": 0.42379817748594917, "grad_norm": 1.4905974864959717, "learning_rate": 7.164724932068713e-06, "loss": 0.2599, "step": 23300 }, { "epoch": 0.42398006511577147, "grad_norm": 1.8258556127548218, "learning_rate": 7.162470451281095e-06, "loss": 0.2594, "step": 23310 }, { "epoch": 0.42416195274559376, "grad_norm": 1.7071080207824707, "learning_rate": 7.16021556707594e-06, "loss": 0.2581, "step": 23320 }, { "epoch": 0.42434384037541606, "grad_norm": 1.63430655002594, "learning_rate": 7.157960280235768e-06, "loss": 0.258, "step": 23330 }, { "epoch": 0.42452572800523836, "grad_norm": 1.7545486688613892, "learning_rate": 7.155704591543245e-06, "loss": 0.2619, "step": 23340 }, { "epoch": 0.42470761563506065, "grad_norm": 2.2649600505828857, "learning_rate": 7.15344850178117e-06, "loss": 0.2673, "step": 23350 }, { "epoch": 0.42488950326488295, "grad_norm": 1.752982497215271, "learning_rate": 7.1511920117324855e-06, "loss": 0.267, "step": 23360 }, { "epoch": 0.42507139089470525, "grad_norm": 1.59830904006958, "learning_rate": 7.148935122180273e-06, "loss": 0.2591, "step": 23370 }, { "epoch": 0.42525327852452754, "grad_norm": 1.5176570415496826, "learning_rate": 7.146677833907749e-06, "loss": 0.2621, "step": 23380 }, { "epoch": 0.42543516615434984, "grad_norm": 1.8532017469406128, "learning_rate": 7.144420147698274e-06, "loss": 0.2582, "step": 23390 }, { "epoch": 0.42561705378417214, "grad_norm": 1.769713044166565, "learning_rate": 7.14216206433534e-06, "loss": 0.2559, "step": 23400 }, { "epoch": 0.42579894141399444, "grad_norm": 1.6185115575790405, "learning_rate": 7.139903584602581e-06, "loss": 0.2616, "step": 23410 }, { "epoch": 0.42598082904381673, "grad_norm": 1.7191033363342285, "learning_rate": 7.137644709283771e-06, "loss": 0.2602, "step": 23420 }, { "epoch": 0.42616271667363903, "grad_norm": 1.9944008588790894, "learning_rate": 7.135385439162813e-06, "loss": 0.2596, "step": 23430 }, { "epoch": 0.4263446043034613, "grad_norm": 1.8427743911743164, "learning_rate": 7.133125775023757e-06, "loss": 0.2552, "step": 23440 }, { "epoch": 0.4265264919332836, "grad_norm": 1.5880478620529175, "learning_rate": 7.13086571765078e-06, "loss": 0.2619, "step": 23450 }, { "epoch": 0.4267083795631059, "grad_norm": 1.6680949926376343, "learning_rate": 7.128605267828205e-06, "loss": 0.2634, "step": 23460 }, { "epoch": 0.4268902671929282, "grad_norm": 2.0171079635620117, "learning_rate": 7.1263444263404825e-06, "loss": 0.2607, "step": 23470 }, { "epoch": 0.4270721548227505, "grad_norm": 1.6441694498062134, "learning_rate": 7.1240831939722045e-06, "loss": 0.2605, "step": 23480 }, { "epoch": 0.4272540424525728, "grad_norm": 1.6284159421920776, "learning_rate": 7.1218215715080975e-06, "loss": 0.2578, "step": 23490 }, { "epoch": 0.4274359300823951, "grad_norm": 1.810858130455017, "learning_rate": 7.1195595597330205e-06, "loss": 0.258, "step": 23500 }, { "epoch": 0.4276178177122174, "grad_norm": 1.6118226051330566, "learning_rate": 7.117297159431973e-06, "loss": 0.2633, "step": 23510 }, { "epoch": 0.4277997053420397, "grad_norm": 1.8230342864990234, "learning_rate": 7.115034371390085e-06, "loss": 0.2549, "step": 23520 }, { "epoch": 0.427981592971862, "grad_norm": 1.9059803485870361, "learning_rate": 7.112771196392622e-06, "loss": 0.2648, "step": 23530 }, { "epoch": 0.4281634806016843, "grad_norm": 2.0286805629730225, "learning_rate": 7.110507635224986e-06, "loss": 0.2644, "step": 23540 }, { "epoch": 0.4283453682315066, "grad_norm": 1.6954351663589478, "learning_rate": 7.108243688672708e-06, "loss": 0.2576, "step": 23550 }, { "epoch": 0.4285272558613289, "grad_norm": 1.607082724571228, "learning_rate": 7.105979357521459e-06, "loss": 0.2485, "step": 23560 }, { "epoch": 0.4287091434911512, "grad_norm": 1.5333118438720703, "learning_rate": 7.103714642557038e-06, "loss": 0.2607, "step": 23570 }, { "epoch": 0.4288910311209735, "grad_norm": 1.7403535842895508, "learning_rate": 7.101449544565382e-06, "loss": 0.2619, "step": 23580 }, { "epoch": 0.4290729187507958, "grad_norm": 1.781952977180481, "learning_rate": 7.099184064332557e-06, "loss": 0.2588, "step": 23590 }, { "epoch": 0.4292548063806181, "grad_norm": 2.0243277549743652, "learning_rate": 7.096918202644763e-06, "loss": 0.2642, "step": 23600 }, { "epoch": 0.4294366940104403, "grad_norm": 2.0686635971069336, "learning_rate": 7.094651960288332e-06, "loss": 0.2612, "step": 23610 }, { "epoch": 0.4296185816402626, "grad_norm": 1.6793677806854248, "learning_rate": 7.09238533804973e-06, "loss": 0.2662, "step": 23620 }, { "epoch": 0.4298004692700849, "grad_norm": 1.5809094905853271, "learning_rate": 7.090118336715554e-06, "loss": 0.2585, "step": 23630 }, { "epoch": 0.4299823568999072, "grad_norm": 1.7940104007720947, "learning_rate": 7.087850957072531e-06, "loss": 0.2595, "step": 23640 }, { "epoch": 0.4301642445297295, "grad_norm": 1.8452444076538086, "learning_rate": 7.085583199907519e-06, "loss": 0.2654, "step": 23650 }, { "epoch": 0.4303461321595518, "grad_norm": 1.6052429676055908, "learning_rate": 7.083315066007511e-06, "loss": 0.2588, "step": 23660 }, { "epoch": 0.4305280197893741, "grad_norm": 1.5849964618682861, "learning_rate": 7.081046556159626e-06, "loss": 0.2552, "step": 23670 }, { "epoch": 0.4307099074191964, "grad_norm": 1.6717182397842407, "learning_rate": 7.0787776711511165e-06, "loss": 0.2587, "step": 23680 }, { "epoch": 0.4308917950490187, "grad_norm": 1.5357403755187988, "learning_rate": 7.076508411769365e-06, "loss": 0.2562, "step": 23690 }, { "epoch": 0.431073682678841, "grad_norm": 1.5411903858184814, "learning_rate": 7.074238778801882e-06, "loss": 0.2574, "step": 23700 }, { "epoch": 0.4312555703086633, "grad_norm": 1.7667765617370605, "learning_rate": 7.071968773036309e-06, "loss": 0.2548, "step": 23710 }, { "epoch": 0.4314374579384856, "grad_norm": 1.6345889568328857, "learning_rate": 7.069698395260416e-06, "loss": 0.2644, "step": 23720 }, { "epoch": 0.4316193455683079, "grad_norm": 1.6433453559875488, "learning_rate": 7.067427646262104e-06, "loss": 0.2648, "step": 23730 }, { "epoch": 0.4318012331981302, "grad_norm": 1.6951944828033447, "learning_rate": 7.065156526829402e-06, "loss": 0.2528, "step": 23740 }, { "epoch": 0.4319831208279525, "grad_norm": 1.6306946277618408, "learning_rate": 7.062885037750466e-06, "loss": 0.2582, "step": 23750 }, { "epoch": 0.4321650084577748, "grad_norm": 1.7072358131408691, "learning_rate": 7.0606131798135805e-06, "loss": 0.2598, "step": 23760 }, { "epoch": 0.4323468960875971, "grad_norm": 1.8231110572814941, "learning_rate": 7.058340953807159e-06, "loss": 0.263, "step": 23770 }, { "epoch": 0.4325287837174194, "grad_norm": 1.8130695819854736, "learning_rate": 7.056068360519747e-06, "loss": 0.257, "step": 23780 }, { "epoch": 0.43271067134724167, "grad_norm": 1.8182322978973389, "learning_rate": 7.053795400740009e-06, "loss": 0.2524, "step": 23790 }, { "epoch": 0.43289255897706397, "grad_norm": 1.8083937168121338, "learning_rate": 7.0515220752567415e-06, "loss": 0.2604, "step": 23800 }, { "epoch": 0.43307444660688627, "grad_norm": 1.837105393409729, "learning_rate": 7.049248384858868e-06, "loss": 0.2571, "step": 23810 }, { "epoch": 0.43325633423670856, "grad_norm": 1.7066290378570557, "learning_rate": 7.046974330335436e-06, "loss": 0.256, "step": 23820 }, { "epoch": 0.43343822186653086, "grad_norm": 1.9920514822006226, "learning_rate": 7.0446999124756244e-06, "loss": 0.2719, "step": 23830 }, { "epoch": 0.43362010949635316, "grad_norm": 1.9990757703781128, "learning_rate": 7.042425132068733e-06, "loss": 0.262, "step": 23840 }, { "epoch": 0.43380199712617545, "grad_norm": 1.8576642274856567, "learning_rate": 7.04014998990419e-06, "loss": 0.2603, "step": 23850 }, { "epoch": 0.43398388475599775, "grad_norm": 1.5895129442214966, "learning_rate": 7.037874486771548e-06, "loss": 0.2652, "step": 23860 }, { "epoch": 0.43416577238582005, "grad_norm": 1.7693026065826416, "learning_rate": 7.035598623460484e-06, "loss": 0.2585, "step": 23870 }, { "epoch": 0.43434766001564235, "grad_norm": 1.71629798412323, "learning_rate": 7.033322400760806e-06, "loss": 0.2617, "step": 23880 }, { "epoch": 0.43452954764546464, "grad_norm": 1.597542405128479, "learning_rate": 7.031045819462439e-06, "loss": 0.2679, "step": 23890 }, { "epoch": 0.43471143527528694, "grad_norm": 1.7809923887252808, "learning_rate": 7.028768880355435e-06, "loss": 0.2578, "step": 23900 }, { "epoch": 0.43489332290510924, "grad_norm": 1.634527325630188, "learning_rate": 7.0264915842299715e-06, "loss": 0.2566, "step": 23910 }, { "epoch": 0.43507521053493153, "grad_norm": 1.769689679145813, "learning_rate": 7.02421393187635e-06, "loss": 0.2526, "step": 23920 }, { "epoch": 0.43525709816475383, "grad_norm": 1.651004672050476, "learning_rate": 7.021935924084995e-06, "loss": 0.2575, "step": 23930 }, { "epoch": 0.4354389857945761, "grad_norm": 1.7002724409103394, "learning_rate": 7.019657561646453e-06, "loss": 0.2632, "step": 23940 }, { "epoch": 0.4356208734243984, "grad_norm": 1.5447951555252075, "learning_rate": 7.017378845351396e-06, "loss": 0.2552, "step": 23950 }, { "epoch": 0.4358027610542207, "grad_norm": 1.583295226097107, "learning_rate": 7.015099775990616e-06, "loss": 0.2592, "step": 23960 }, { "epoch": 0.435984648684043, "grad_norm": 1.7730849981307983, "learning_rate": 7.012820354355029e-06, "loss": 0.2594, "step": 23970 }, { "epoch": 0.4361665363138653, "grad_norm": 1.7654900550842285, "learning_rate": 7.010540581235676e-06, "loss": 0.2587, "step": 23980 }, { "epoch": 0.4363484239436876, "grad_norm": 1.6825916767120361, "learning_rate": 7.008260457423713e-06, "loss": 0.2623, "step": 23990 }, { "epoch": 0.4365303115735099, "grad_norm": 1.9724807739257812, "learning_rate": 7.005979983710427e-06, "loss": 0.2641, "step": 24000 }, { "epoch": 0.4367121992033322, "grad_norm": 1.5786068439483643, "learning_rate": 7.003699160887218e-06, "loss": 0.2626, "step": 24010 }, { "epoch": 0.4368940868331545, "grad_norm": 2.079246997833252, "learning_rate": 7.001417989745611e-06, "loss": 0.2586, "step": 24020 }, { "epoch": 0.43707597446297675, "grad_norm": 1.7108851671218872, "learning_rate": 6.999136471077252e-06, "loss": 0.261, "step": 24030 }, { "epoch": 0.43725786209279904, "grad_norm": 1.732879877090454, "learning_rate": 6.996854605673907e-06, "loss": 0.2581, "step": 24040 }, { "epoch": 0.43743974972262134, "grad_norm": 1.6867356300354004, "learning_rate": 6.994572394327462e-06, "loss": 0.2568, "step": 24050 }, { "epoch": 0.43762163735244364, "grad_norm": 1.6657767295837402, "learning_rate": 6.992289837829924e-06, "loss": 0.2557, "step": 24060 }, { "epoch": 0.43780352498226593, "grad_norm": 1.7175532579421997, "learning_rate": 6.990006936973419e-06, "loss": 0.2581, "step": 24070 }, { "epoch": 0.43798541261208823, "grad_norm": 1.9580870866775513, "learning_rate": 6.987723692550194e-06, "loss": 0.2576, "step": 24080 }, { "epoch": 0.43816730024191053, "grad_norm": 1.8978949785232544, "learning_rate": 6.98544010535261e-06, "loss": 0.2509, "step": 24090 }, { "epoch": 0.4383491878717328, "grad_norm": 1.573479175567627, "learning_rate": 6.983156176173156e-06, "loss": 0.2619, "step": 24100 }, { "epoch": 0.4385310755015551, "grad_norm": 1.4652496576309204, "learning_rate": 6.980871905804432e-06, "loss": 0.2548, "step": 24110 }, { "epoch": 0.4387129631313774, "grad_norm": 1.7417805194854736, "learning_rate": 6.978587295039157e-06, "loss": 0.2648, "step": 24120 }, { "epoch": 0.4388948507611997, "grad_norm": 1.7023935317993164, "learning_rate": 6.976302344670174e-06, "loss": 0.2545, "step": 24130 }, { "epoch": 0.439076738391022, "grad_norm": 1.7430983781814575, "learning_rate": 6.974017055490437e-06, "loss": 0.2627, "step": 24140 }, { "epoch": 0.4392586260208443, "grad_norm": 1.7450268268585205, "learning_rate": 6.971731428293022e-06, "loss": 0.2598, "step": 24150 }, { "epoch": 0.4394405136506666, "grad_norm": 1.7699657678604126, "learning_rate": 6.969445463871122e-06, "loss": 0.263, "step": 24160 }, { "epoch": 0.4396224012804889, "grad_norm": 1.5806900262832642, "learning_rate": 6.9671591630180445e-06, "loss": 0.2547, "step": 24170 }, { "epoch": 0.4398042889103112, "grad_norm": 1.7166205644607544, "learning_rate": 6.964872526527216e-06, "loss": 0.2692, "step": 24180 }, { "epoch": 0.4399861765401335, "grad_norm": 1.8208130598068237, "learning_rate": 6.962585555192174e-06, "loss": 0.2505, "step": 24190 }, { "epoch": 0.4401680641699558, "grad_norm": 1.5697458982467651, "learning_rate": 6.960298249806583e-06, "loss": 0.2531, "step": 24200 }, { "epoch": 0.4403499517997781, "grad_norm": 1.5280909538269043, "learning_rate": 6.958010611164216e-06, "loss": 0.2602, "step": 24210 }, { "epoch": 0.4405318394296004, "grad_norm": 1.6115089654922485, "learning_rate": 6.9557226400589596e-06, "loss": 0.2539, "step": 24220 }, { "epoch": 0.4407137270594227, "grad_norm": 1.6732498407363892, "learning_rate": 6.953434337284822e-06, "loss": 0.2599, "step": 24230 }, { "epoch": 0.440895614689245, "grad_norm": 1.4217745065689087, "learning_rate": 6.951145703635923e-06, "loss": 0.261, "step": 24240 }, { "epoch": 0.4410775023190673, "grad_norm": 1.5295265913009644, "learning_rate": 6.948856739906498e-06, "loss": 0.2583, "step": 24250 }, { "epoch": 0.4412593899488896, "grad_norm": 1.5609675645828247, "learning_rate": 6.946567446890895e-06, "loss": 0.2566, "step": 24260 }, { "epoch": 0.4414412775787119, "grad_norm": 1.6970956325531006, "learning_rate": 6.944277825383581e-06, "loss": 0.2589, "step": 24270 }, { "epoch": 0.4416231652085342, "grad_norm": 1.8068996667861938, "learning_rate": 6.9419878761791325e-06, "loss": 0.2551, "step": 24280 }, { "epoch": 0.44180505283835647, "grad_norm": 1.8861101865768433, "learning_rate": 6.9396976000722405e-06, "loss": 0.2605, "step": 24290 }, { "epoch": 0.44198694046817877, "grad_norm": 1.7718639373779297, "learning_rate": 6.93740699785771e-06, "loss": 0.26, "step": 24300 }, { "epoch": 0.44216882809800107, "grad_norm": 1.6604551076889038, "learning_rate": 6.935116070330461e-06, "loss": 0.2562, "step": 24310 }, { "epoch": 0.44235071572782336, "grad_norm": 1.7462385892868042, "learning_rate": 6.932824818285524e-06, "loss": 0.2627, "step": 24320 }, { "epoch": 0.44253260335764566, "grad_norm": 1.6673548221588135, "learning_rate": 6.930533242518043e-06, "loss": 0.2623, "step": 24330 }, { "epoch": 0.44271449098746796, "grad_norm": 1.5196261405944824, "learning_rate": 6.928241343823273e-06, "loss": 0.2559, "step": 24340 }, { "epoch": 0.44289637861729025, "grad_norm": 1.7657731771469116, "learning_rate": 6.925949122996584e-06, "loss": 0.2601, "step": 24350 }, { "epoch": 0.44307826624711255, "grad_norm": 1.4912601709365845, "learning_rate": 6.923656580833457e-06, "loss": 0.2535, "step": 24360 }, { "epoch": 0.44326015387693485, "grad_norm": 1.8413857221603394, "learning_rate": 6.92136371812948e-06, "loss": 0.2624, "step": 24370 }, { "epoch": 0.44344204150675715, "grad_norm": 1.5910555124282837, "learning_rate": 6.91907053568036e-06, "loss": 0.2528, "step": 24380 }, { "epoch": 0.44362392913657944, "grad_norm": 1.4936686754226685, "learning_rate": 6.916777034281906e-06, "loss": 0.2563, "step": 24390 }, { "epoch": 0.44380581676640174, "grad_norm": 1.729997992515564, "learning_rate": 6.9144832147300475e-06, "loss": 0.2538, "step": 24400 }, { "epoch": 0.44398770439622404, "grad_norm": 1.636481761932373, "learning_rate": 6.9121890778208145e-06, "loss": 0.2617, "step": 24410 }, { "epoch": 0.44416959202604633, "grad_norm": 1.8247597217559814, "learning_rate": 6.909894624350357e-06, "loss": 0.255, "step": 24420 }, { "epoch": 0.44435147965586863, "grad_norm": 1.487716794013977, "learning_rate": 6.907599855114926e-06, "loss": 0.2606, "step": 24430 }, { "epoch": 0.4445333672856909, "grad_norm": 1.5314017534255981, "learning_rate": 6.905304770910888e-06, "loss": 0.2529, "step": 24440 }, { "epoch": 0.44471525491551317, "grad_norm": 1.7001489400863647, "learning_rate": 6.903009372534717e-06, "loss": 0.2507, "step": 24450 }, { "epoch": 0.44489714254533547, "grad_norm": 1.527741551399231, "learning_rate": 6.900713660782992e-06, "loss": 0.2537, "step": 24460 }, { "epoch": 0.44507903017515776, "grad_norm": 1.632641315460205, "learning_rate": 6.89841763645241e-06, "loss": 0.258, "step": 24470 }, { "epoch": 0.44526091780498006, "grad_norm": 1.739601731300354, "learning_rate": 6.8961213003397675e-06, "loss": 0.2594, "step": 24480 }, { "epoch": 0.44544280543480236, "grad_norm": 1.643226981163025, "learning_rate": 6.893824653241974e-06, "loss": 0.2505, "step": 24490 }, { "epoch": 0.44562469306462466, "grad_norm": 1.7035881280899048, "learning_rate": 6.8915276959560454e-06, "loss": 0.2579, "step": 24500 }, { "epoch": 0.44580658069444695, "grad_norm": 1.5446864366531372, "learning_rate": 6.889230429279104e-06, "loss": 0.2561, "step": 24510 }, { "epoch": 0.44598846832426925, "grad_norm": 1.514877438545227, "learning_rate": 6.886932854008386e-06, "loss": 0.2533, "step": 24520 }, { "epoch": 0.44617035595409155, "grad_norm": 1.7692488431930542, "learning_rate": 6.884634970941224e-06, "loss": 0.2588, "step": 24530 }, { "epoch": 0.44635224358391384, "grad_norm": 1.7026398181915283, "learning_rate": 6.882336780875065e-06, "loss": 0.2527, "step": 24540 }, { "epoch": 0.44653413121373614, "grad_norm": 1.588701844215393, "learning_rate": 6.880038284607461e-06, "loss": 0.2595, "step": 24550 }, { "epoch": 0.44671601884355844, "grad_norm": 1.7857751846313477, "learning_rate": 6.87773948293607e-06, "loss": 0.2504, "step": 24560 }, { "epoch": 0.44689790647338073, "grad_norm": 1.6239644289016724, "learning_rate": 6.875440376658654e-06, "loss": 0.2655, "step": 24570 }, { "epoch": 0.44707979410320303, "grad_norm": 1.605722188949585, "learning_rate": 6.873140966573086e-06, "loss": 0.2605, "step": 24580 }, { "epoch": 0.44726168173302533, "grad_norm": 1.937291145324707, "learning_rate": 6.870841253477339e-06, "loss": 0.258, "step": 24590 }, { "epoch": 0.4474435693628476, "grad_norm": 2.219212055206299, "learning_rate": 6.8685412381694916e-06, "loss": 0.2572, "step": 24600 }, { "epoch": 0.4476254569926699, "grad_norm": 1.7578868865966797, "learning_rate": 6.866240921447731e-06, "loss": 0.2577, "step": 24610 }, { "epoch": 0.4478073446224922, "grad_norm": 1.8067165613174438, "learning_rate": 6.863940304110347e-06, "loss": 0.2674, "step": 24620 }, { "epoch": 0.4479892322523145, "grad_norm": 1.6823879480361938, "learning_rate": 6.861639386955732e-06, "loss": 0.2631, "step": 24630 }, { "epoch": 0.4481711198821368, "grad_norm": 1.4295493364334106, "learning_rate": 6.859338170782386e-06, "loss": 0.2614, "step": 24640 }, { "epoch": 0.4483530075119591, "grad_norm": 1.515064001083374, "learning_rate": 6.857036656388909e-06, "loss": 0.2541, "step": 24650 }, { "epoch": 0.4485348951417814, "grad_norm": 1.6178618669509888, "learning_rate": 6.854734844574007e-06, "loss": 0.2565, "step": 24660 }, { "epoch": 0.4487167827716037, "grad_norm": 1.5625019073486328, "learning_rate": 6.8524327361364875e-06, "loss": 0.2513, "step": 24670 }, { "epoch": 0.448898670401426, "grad_norm": 1.471646785736084, "learning_rate": 6.850130331875264e-06, "loss": 0.2599, "step": 24680 }, { "epoch": 0.4490805580312483, "grad_norm": 2.192232608795166, "learning_rate": 6.847827632589348e-06, "loss": 0.266, "step": 24690 }, { "epoch": 0.4492624456610706, "grad_norm": 1.6108287572860718, "learning_rate": 6.8455246390778585e-06, "loss": 0.264, "step": 24700 }, { "epoch": 0.4494443332908929, "grad_norm": 1.6938366889953613, "learning_rate": 6.843221352140012e-06, "loss": 0.26, "step": 24710 }, { "epoch": 0.4496262209207152, "grad_norm": 1.6689159870147705, "learning_rate": 6.840917772575132e-06, "loss": 0.2562, "step": 24720 }, { "epoch": 0.4498081085505375, "grad_norm": 1.8349446058273315, "learning_rate": 6.838613901182636e-06, "loss": 0.2543, "step": 24730 }, { "epoch": 0.4499899961803598, "grad_norm": 1.7295554876327515, "learning_rate": 6.836309738762049e-06, "loss": 0.261, "step": 24740 }, { "epoch": 0.4501718838101821, "grad_norm": 1.5465805530548096, "learning_rate": 6.834005286112997e-06, "loss": 0.2557, "step": 24750 }, { "epoch": 0.4503537714400044, "grad_norm": 1.952824592590332, "learning_rate": 6.831700544035203e-06, "loss": 0.2581, "step": 24760 }, { "epoch": 0.4505356590698267, "grad_norm": 1.5287601947784424, "learning_rate": 6.829395513328495e-06, "loss": 0.2509, "step": 24770 }, { "epoch": 0.450717546699649, "grad_norm": 1.9146968126296997, "learning_rate": 6.827090194792794e-06, "loss": 0.2585, "step": 24780 }, { "epoch": 0.45089943432947127, "grad_norm": 1.6278527975082397, "learning_rate": 6.824784589228131e-06, "loss": 0.254, "step": 24790 }, { "epoch": 0.45108132195929357, "grad_norm": 1.6493333578109741, "learning_rate": 6.822478697434626e-06, "loss": 0.26, "step": 24800 }, { "epoch": 0.45126320958911587, "grad_norm": 1.8265711069107056, "learning_rate": 6.820172520212504e-06, "loss": 0.2607, "step": 24810 }, { "epoch": 0.45144509721893816, "grad_norm": 1.7867045402526855, "learning_rate": 6.817866058362093e-06, "loss": 0.2607, "step": 24820 }, { "epoch": 0.45162698484876046, "grad_norm": 1.8799041509628296, "learning_rate": 6.815559312683809e-06, "loss": 0.2513, "step": 24830 }, { "epoch": 0.45180887247858276, "grad_norm": 1.6264467239379883, "learning_rate": 6.813252283978179e-06, "loss": 0.255, "step": 24840 }, { "epoch": 0.45199076010840505, "grad_norm": 1.6757241487503052, "learning_rate": 6.810944973045817e-06, "loss": 0.261, "step": 24850 }, { "epoch": 0.4521726477382273, "grad_norm": 1.4864046573638916, "learning_rate": 6.808637380687442e-06, "loss": 0.2591, "step": 24860 }, { "epoch": 0.4523545353680496, "grad_norm": 1.669597864151001, "learning_rate": 6.806329507703867e-06, "loss": 0.26, "step": 24870 }, { "epoch": 0.4525364229978719, "grad_norm": 1.7812600135803223, "learning_rate": 6.804021354896006e-06, "loss": 0.2558, "step": 24880 }, { "epoch": 0.4527183106276942, "grad_norm": 1.697568416595459, "learning_rate": 6.801712923064866e-06, "loss": 0.2505, "step": 24890 }, { "epoch": 0.4529001982575165, "grad_norm": 1.5243297815322876, "learning_rate": 6.799404213011555e-06, "loss": 0.2545, "step": 24900 }, { "epoch": 0.4530820858873388, "grad_norm": 1.5077449083328247, "learning_rate": 6.7970952255372735e-06, "loss": 0.2573, "step": 24910 }, { "epoch": 0.4532639735171611, "grad_norm": 1.623781442642212, "learning_rate": 6.794785961443322e-06, "loss": 0.2596, "step": 24920 }, { "epoch": 0.4534458611469834, "grad_norm": 1.5172909498214722, "learning_rate": 6.7924764215310936e-06, "loss": 0.2532, "step": 24930 }, { "epoch": 0.4536277487768057, "grad_norm": 1.6141313314437866, "learning_rate": 6.79016660660208e-06, "loss": 0.2615, "step": 24940 }, { "epoch": 0.45380963640662797, "grad_norm": 1.9648226499557495, "learning_rate": 6.787856517457865e-06, "loss": 0.2601, "step": 24950 }, { "epoch": 0.45399152403645027, "grad_norm": 1.9722754955291748, "learning_rate": 6.785546154900131e-06, "loss": 0.2511, "step": 24960 }, { "epoch": 0.45417341166627256, "grad_norm": 1.8580331802368164, "learning_rate": 6.783235519730655e-06, "loss": 0.2541, "step": 24970 }, { "epoch": 0.45435529929609486, "grad_norm": 1.69321608543396, "learning_rate": 6.780924612751305e-06, "loss": 0.2608, "step": 24980 }, { "epoch": 0.45453718692591716, "grad_norm": 1.6478092670440674, "learning_rate": 6.778613434764048e-06, "loss": 0.264, "step": 24990 }, { "epoch": 0.45471907455573946, "grad_norm": 1.8251203298568726, "learning_rate": 6.776301986570943e-06, "loss": 0.2522, "step": 25000 }, { "epoch": 0.45490096218556175, "grad_norm": 1.8467756509780884, "learning_rate": 6.773990268974138e-06, "loss": 0.2578, "step": 25010 }, { "epoch": 0.45508284981538405, "grad_norm": 1.876791000366211, "learning_rate": 6.771678282775886e-06, "loss": 0.2562, "step": 25020 }, { "epoch": 0.45526473744520635, "grad_norm": 1.6278146505355835, "learning_rate": 6.769366028778522e-06, "loss": 0.2577, "step": 25030 }, { "epoch": 0.45544662507502864, "grad_norm": 1.5702118873596191, "learning_rate": 6.76705350778448e-06, "loss": 0.248, "step": 25040 }, { "epoch": 0.45562851270485094, "grad_norm": 1.697059988975525, "learning_rate": 6.764740720596283e-06, "loss": 0.2582, "step": 25050 }, { "epoch": 0.45581040033467324, "grad_norm": 1.8452874422073364, "learning_rate": 6.762427668016551e-06, "loss": 0.2584, "step": 25060 }, { "epoch": 0.45599228796449554, "grad_norm": 1.6407160758972168, "learning_rate": 6.760114350847993e-06, "loss": 0.2614, "step": 25070 }, { "epoch": 0.45617417559431783, "grad_norm": 1.561628818511963, "learning_rate": 6.757800769893407e-06, "loss": 0.2558, "step": 25080 }, { "epoch": 0.45635606322414013, "grad_norm": 1.5526697635650635, "learning_rate": 6.75548692595569e-06, "loss": 0.2513, "step": 25090 }, { "epoch": 0.4565379508539624, "grad_norm": 1.6441733837127686, "learning_rate": 6.753172819837822e-06, "loss": 0.2545, "step": 25100 }, { "epoch": 0.4567198384837847, "grad_norm": 1.856651782989502, "learning_rate": 6.750858452342882e-06, "loss": 0.2619, "step": 25110 }, { "epoch": 0.456901726113607, "grad_norm": 1.6683262586593628, "learning_rate": 6.748543824274032e-06, "loss": 0.2514, "step": 25120 }, { "epoch": 0.4570836137434293, "grad_norm": 1.7004718780517578, "learning_rate": 6.746228936434531e-06, "loss": 0.2553, "step": 25130 }, { "epoch": 0.4572655013732516, "grad_norm": 1.5842589139938354, "learning_rate": 6.743913789627727e-06, "loss": 0.2584, "step": 25140 }, { "epoch": 0.4574473890030739, "grad_norm": 1.6368271112442017, "learning_rate": 6.7415983846570486e-06, "loss": 0.2555, "step": 25150 }, { "epoch": 0.4576292766328962, "grad_norm": 1.8812772035598755, "learning_rate": 6.73928272232603e-06, "loss": 0.2535, "step": 25160 }, { "epoch": 0.4578111642627185, "grad_norm": 1.6420806646347046, "learning_rate": 6.736966803438282e-06, "loss": 0.2538, "step": 25170 }, { "epoch": 0.4579930518925408, "grad_norm": 1.9163477420806885, "learning_rate": 6.734650628797509e-06, "loss": 0.2576, "step": 25180 }, { "epoch": 0.4581749395223631, "grad_norm": 1.7554240226745605, "learning_rate": 6.732334199207505e-06, "loss": 0.2611, "step": 25190 }, { "epoch": 0.4583568271521854, "grad_norm": 1.7175325155258179, "learning_rate": 6.730017515472151e-06, "loss": 0.2595, "step": 25200 }, { "epoch": 0.4585387147820077, "grad_norm": 1.7340543270111084, "learning_rate": 6.7277005783954155e-06, "loss": 0.2595, "step": 25210 }, { "epoch": 0.45872060241183, "grad_norm": 1.6893857717514038, "learning_rate": 6.725383388781357e-06, "loss": 0.2594, "step": 25220 }, { "epoch": 0.4589024900416523, "grad_norm": 1.70817232131958, "learning_rate": 6.72306594743412e-06, "loss": 0.2586, "step": 25230 }, { "epoch": 0.4590843776714746, "grad_norm": 1.6824100017547607, "learning_rate": 6.7207482551579396e-06, "loss": 0.2548, "step": 25240 }, { "epoch": 0.4592662653012969, "grad_norm": 1.7593399286270142, "learning_rate": 6.71843031275713e-06, "loss": 0.2593, "step": 25250 }, { "epoch": 0.4594481529311192, "grad_norm": 1.6708999872207642, "learning_rate": 6.716112121036103e-06, "loss": 0.2552, "step": 25260 }, { "epoch": 0.4596300405609415, "grad_norm": 1.7835047245025635, "learning_rate": 6.7137936807993475e-06, "loss": 0.2511, "step": 25270 }, { "epoch": 0.4598119281907637, "grad_norm": 1.8100974559783936, "learning_rate": 6.711474992851446e-06, "loss": 0.2512, "step": 25280 }, { "epoch": 0.459993815820586, "grad_norm": 1.5799822807312012, "learning_rate": 6.709156057997061e-06, "loss": 0.2551, "step": 25290 }, { "epoch": 0.4601757034504083, "grad_norm": 1.7194199562072754, "learning_rate": 6.706836877040943e-06, "loss": 0.2489, "step": 25300 }, { "epoch": 0.4603575910802306, "grad_norm": 1.5885584354400635, "learning_rate": 6.7045174507879316e-06, "loss": 0.251, "step": 25310 }, { "epoch": 0.4605394787100529, "grad_norm": 2.0879344940185547, "learning_rate": 6.702197780042945e-06, "loss": 0.26, "step": 25320 }, { "epoch": 0.4607213663398752, "grad_norm": 1.8664000034332275, "learning_rate": 6.69987786561099e-06, "loss": 0.2585, "step": 25330 }, { "epoch": 0.4609032539696975, "grad_norm": 1.8699272871017456, "learning_rate": 6.697557708297157e-06, "loss": 0.2524, "step": 25340 }, { "epoch": 0.4610851415995198, "grad_norm": 1.5203471183776855, "learning_rate": 6.695237308906623e-06, "loss": 0.2552, "step": 25350 }, { "epoch": 0.4612670292293421, "grad_norm": 1.667101502418518, "learning_rate": 6.692916668244644e-06, "loss": 0.2566, "step": 25360 }, { "epoch": 0.4614489168591644, "grad_norm": 1.663283348083496, "learning_rate": 6.690595787116565e-06, "loss": 0.2565, "step": 25370 }, { "epoch": 0.4616308044889867, "grad_norm": 1.5419948101043701, "learning_rate": 6.688274666327811e-06, "loss": 0.2598, "step": 25380 }, { "epoch": 0.461812692118809, "grad_norm": 1.6631609201431274, "learning_rate": 6.6859533066838925e-06, "loss": 0.2579, "step": 25390 }, { "epoch": 0.4619945797486313, "grad_norm": 1.4890174865722656, "learning_rate": 6.6836317089904e-06, "loss": 0.2518, "step": 25400 }, { "epoch": 0.4621764673784536, "grad_norm": 1.5177853107452393, "learning_rate": 6.681309874053011e-06, "loss": 0.2589, "step": 25410 }, { "epoch": 0.4623583550082759, "grad_norm": 1.7785732746124268, "learning_rate": 6.678987802677479e-06, "loss": 0.2636, "step": 25420 }, { "epoch": 0.4625402426380982, "grad_norm": 1.8615949153900146, "learning_rate": 6.676665495669646e-06, "loss": 0.264, "step": 25430 }, { "epoch": 0.4627221302679205, "grad_norm": 1.5739789009094238, "learning_rate": 6.674342953835433e-06, "loss": 0.2591, "step": 25440 }, { "epoch": 0.46290401789774277, "grad_norm": 1.799907922744751, "learning_rate": 6.672020177980843e-06, "loss": 0.2624, "step": 25450 }, { "epoch": 0.46308590552756507, "grad_norm": 1.6657679080963135, "learning_rate": 6.669697168911957e-06, "loss": 0.2592, "step": 25460 }, { "epoch": 0.46326779315738736, "grad_norm": 1.6266058683395386, "learning_rate": 6.667373927434941e-06, "loss": 0.2583, "step": 25470 }, { "epoch": 0.46344968078720966, "grad_norm": 1.6596373319625854, "learning_rate": 6.665050454356043e-06, "loss": 0.2602, "step": 25480 }, { "epoch": 0.46363156841703196, "grad_norm": 2.299347400665283, "learning_rate": 6.662726750481585e-06, "loss": 0.2568, "step": 25490 }, { "epoch": 0.46381345604685426, "grad_norm": 1.654019832611084, "learning_rate": 6.660402816617974e-06, "loss": 0.2535, "step": 25500 }, { "epoch": 0.46399534367667655, "grad_norm": 1.4890743494033813, "learning_rate": 6.658078653571696e-06, "loss": 0.2536, "step": 25510 }, { "epoch": 0.46417723130649885, "grad_norm": 1.9519829750061035, "learning_rate": 6.655754262149314e-06, "loss": 0.2502, "step": 25520 }, { "epoch": 0.46435911893632115, "grad_norm": 1.757631778717041, "learning_rate": 6.653429643157478e-06, "loss": 0.2602, "step": 25530 }, { "epoch": 0.46454100656614344, "grad_norm": 1.6403164863586426, "learning_rate": 6.651104797402903e-06, "loss": 0.2567, "step": 25540 }, { "epoch": 0.46472289419596574, "grad_norm": 1.9009634256362915, "learning_rate": 6.648779725692399e-06, "loss": 0.2539, "step": 25550 }, { "epoch": 0.46490478182578804, "grad_norm": 1.736086130142212, "learning_rate": 6.646454428832843e-06, "loss": 0.2523, "step": 25560 }, { "epoch": 0.46508666945561034, "grad_norm": 1.906761646270752, "learning_rate": 6.644128907631191e-06, "loss": 0.2696, "step": 25570 }, { "epoch": 0.46526855708543263, "grad_norm": 1.6342761516571045, "learning_rate": 6.641803162894485e-06, "loss": 0.2587, "step": 25580 }, { "epoch": 0.46545044471525493, "grad_norm": 1.522592544555664, "learning_rate": 6.6394771954298354e-06, "loss": 0.2489, "step": 25590 }, { "epoch": 0.4656323323450772, "grad_norm": 1.5317425727844238, "learning_rate": 6.637151006044435e-06, "loss": 0.2555, "step": 25600 }, { "epoch": 0.4658142199748995, "grad_norm": 1.6557449102401733, "learning_rate": 6.63482459554555e-06, "loss": 0.2561, "step": 25610 }, { "epoch": 0.4659961076047218, "grad_norm": 1.6254339218139648, "learning_rate": 6.632497964740527e-06, "loss": 0.259, "step": 25620 }, { "epoch": 0.4661779952345441, "grad_norm": 1.7135266065597534, "learning_rate": 6.63017111443679e-06, "loss": 0.2594, "step": 25630 }, { "epoch": 0.4663598828643664, "grad_norm": 2.026197910308838, "learning_rate": 6.627844045441832e-06, "loss": 0.2521, "step": 25640 }, { "epoch": 0.4665417704941887, "grad_norm": 1.6403559446334839, "learning_rate": 6.62551675856323e-06, "loss": 0.2552, "step": 25650 }, { "epoch": 0.466723658124011, "grad_norm": 1.7287286520004272, "learning_rate": 6.623189254608632e-06, "loss": 0.2558, "step": 25660 }, { "epoch": 0.4669055457538333, "grad_norm": 1.7451651096343994, "learning_rate": 6.620861534385761e-06, "loss": 0.2512, "step": 25670 }, { "epoch": 0.4670874333836556, "grad_norm": 1.6375327110290527, "learning_rate": 6.61853359870242e-06, "loss": 0.2563, "step": 25680 }, { "epoch": 0.4672693210134779, "grad_norm": 1.6252715587615967, "learning_rate": 6.61620544836648e-06, "loss": 0.2567, "step": 25690 }, { "epoch": 0.46745120864330014, "grad_norm": 1.8496596813201904, "learning_rate": 6.613877084185892e-06, "loss": 0.2612, "step": 25700 }, { "epoch": 0.46763309627312244, "grad_norm": 1.6058663129806519, "learning_rate": 6.611548506968679e-06, "loss": 0.2587, "step": 25710 }, { "epoch": 0.46781498390294474, "grad_norm": 1.5578477382659912, "learning_rate": 6.6092197175229366e-06, "loss": 0.2535, "step": 25720 }, { "epoch": 0.46799687153276703, "grad_norm": 1.8659363985061646, "learning_rate": 6.606890716656838e-06, "loss": 0.2562, "step": 25730 }, { "epoch": 0.46817875916258933, "grad_norm": 1.5342408418655396, "learning_rate": 6.604561505178624e-06, "loss": 0.2586, "step": 25740 }, { "epoch": 0.46836064679241163, "grad_norm": 1.5569432973861694, "learning_rate": 6.602232083896614e-06, "loss": 0.2533, "step": 25750 }, { "epoch": 0.4685425344222339, "grad_norm": 1.6291933059692383, "learning_rate": 6.599902453619196e-06, "loss": 0.2589, "step": 25760 }, { "epoch": 0.4687244220520562, "grad_norm": 1.2940446138381958, "learning_rate": 6.597572615154835e-06, "loss": 0.2446, "step": 25770 }, { "epoch": 0.4689063096818785, "grad_norm": 1.8390617370605469, "learning_rate": 6.595242569312064e-06, "loss": 0.2594, "step": 25780 }, { "epoch": 0.4690881973117008, "grad_norm": 1.6126302480697632, "learning_rate": 6.592912316899489e-06, "loss": 0.2545, "step": 25790 }, { "epoch": 0.4692700849415231, "grad_norm": 1.6349951028823853, "learning_rate": 6.590581858725792e-06, "loss": 0.254, "step": 25800 }, { "epoch": 0.4694519725713454, "grad_norm": 1.7571321725845337, "learning_rate": 6.588251195599717e-06, "loss": 0.2571, "step": 25810 }, { "epoch": 0.4696338602011677, "grad_norm": 1.7280138731002808, "learning_rate": 6.585920328330091e-06, "loss": 0.2593, "step": 25820 }, { "epoch": 0.46981574783099, "grad_norm": 1.6674140691757202, "learning_rate": 6.583589257725801e-06, "loss": 0.256, "step": 25830 }, { "epoch": 0.4699976354608123, "grad_norm": 1.9021083116531372, "learning_rate": 6.581257984595811e-06, "loss": 0.2499, "step": 25840 }, { "epoch": 0.4701795230906346, "grad_norm": 1.745432734489441, "learning_rate": 6.5789265097491525e-06, "loss": 0.2578, "step": 25850 }, { "epoch": 0.4703614107204569, "grad_norm": 1.6084610223770142, "learning_rate": 6.576594833994932e-06, "loss": 0.2495, "step": 25860 }, { "epoch": 0.4705432983502792, "grad_norm": 1.7408714294433594, "learning_rate": 6.574262958142316e-06, "loss": 0.2579, "step": 25870 }, { "epoch": 0.4707251859801015, "grad_norm": 1.8613688945770264, "learning_rate": 6.571930883000552e-06, "loss": 0.2625, "step": 25880 }, { "epoch": 0.4709070736099238, "grad_norm": 1.5254409313201904, "learning_rate": 6.5695986093789465e-06, "loss": 0.2595, "step": 25890 }, { "epoch": 0.4710889612397461, "grad_norm": 1.6205129623413086, "learning_rate": 6.567266138086883e-06, "loss": 0.2579, "step": 25900 }, { "epoch": 0.4712708488695684, "grad_norm": 1.9896262884140015, "learning_rate": 6.564933469933805e-06, "loss": 0.2577, "step": 25910 }, { "epoch": 0.4714527364993907, "grad_norm": 1.627240538597107, "learning_rate": 6.562600605729234e-06, "loss": 0.2503, "step": 25920 }, { "epoch": 0.471634624129213, "grad_norm": 1.5948129892349243, "learning_rate": 6.560267546282753e-06, "loss": 0.2585, "step": 25930 }, { "epoch": 0.4718165117590353, "grad_norm": 1.6386592388153076, "learning_rate": 6.557934292404013e-06, "loss": 0.2525, "step": 25940 }, { "epoch": 0.47199839938885757, "grad_norm": 1.714368224143982, "learning_rate": 6.555600844902736e-06, "loss": 0.2554, "step": 25950 }, { "epoch": 0.47218028701867987, "grad_norm": 1.4946590662002563, "learning_rate": 6.553267204588707e-06, "loss": 0.2569, "step": 25960 }, { "epoch": 0.47236217464850216, "grad_norm": 1.552980899810791, "learning_rate": 6.550933372271784e-06, "loss": 0.2523, "step": 25970 }, { "epoch": 0.47254406227832446, "grad_norm": 1.6275420188903809, "learning_rate": 6.548599348761884e-06, "loss": 0.2533, "step": 25980 }, { "epoch": 0.47272594990814676, "grad_norm": 1.522143840789795, "learning_rate": 6.546265134868994e-06, "loss": 0.259, "step": 25990 }, { "epoch": 0.47290783753796906, "grad_norm": 1.6570018529891968, "learning_rate": 6.54393073140317e-06, "loss": 0.2529, "step": 26000 }, { "epoch": 0.47308972516779135, "grad_norm": 1.5648494958877563, "learning_rate": 6.541596139174528e-06, "loss": 0.2541, "step": 26010 }, { "epoch": 0.47327161279761365, "grad_norm": 1.5716761350631714, "learning_rate": 6.539261358993254e-06, "loss": 0.2581, "step": 26020 }, { "epoch": 0.47345350042743595, "grad_norm": 1.5763497352600098, "learning_rate": 6.536926391669599e-06, "loss": 0.2594, "step": 26030 }, { "epoch": 0.47363538805725824, "grad_norm": 1.696830153465271, "learning_rate": 6.534591238013874e-06, "loss": 0.2566, "step": 26040 }, { "epoch": 0.47381727568708054, "grad_norm": 1.7752934694290161, "learning_rate": 6.532255898836461e-06, "loss": 0.2523, "step": 26050 }, { "epoch": 0.47399916331690284, "grad_norm": 1.7441697120666504, "learning_rate": 6.529920374947802e-06, "loss": 0.2514, "step": 26060 }, { "epoch": 0.47418105094672514, "grad_norm": 1.3989765644073486, "learning_rate": 6.5275846671584065e-06, "loss": 0.2561, "step": 26070 }, { "epoch": 0.47436293857654743, "grad_norm": 1.5847737789154053, "learning_rate": 6.5252487762788455e-06, "loss": 0.2526, "step": 26080 }, { "epoch": 0.47454482620636973, "grad_norm": 1.723466157913208, "learning_rate": 6.5229127031197525e-06, "loss": 0.2587, "step": 26090 }, { "epoch": 0.474726713836192, "grad_norm": 1.3558437824249268, "learning_rate": 6.520576448491829e-06, "loss": 0.2507, "step": 26100 }, { "epoch": 0.4749086014660143, "grad_norm": 1.9122035503387451, "learning_rate": 6.518240013205832e-06, "loss": 0.2619, "step": 26110 }, { "epoch": 0.47509048909583657, "grad_norm": 1.6036980152130127, "learning_rate": 6.515903398072589e-06, "loss": 0.2538, "step": 26120 }, { "epoch": 0.47527237672565886, "grad_norm": 1.5290218591690063, "learning_rate": 6.513566603902985e-06, "loss": 0.2564, "step": 26130 }, { "epoch": 0.47545426435548116, "grad_norm": 1.692419171333313, "learning_rate": 6.511229631507969e-06, "loss": 0.2585, "step": 26140 }, { "epoch": 0.47563615198530346, "grad_norm": 1.725437045097351, "learning_rate": 6.508892481698552e-06, "loss": 0.2517, "step": 26150 }, { "epoch": 0.47581803961512575, "grad_norm": 1.7519019842147827, "learning_rate": 6.506555155285806e-06, "loss": 0.252, "step": 26160 }, { "epoch": 0.47599992724494805, "grad_norm": 1.690828800201416, "learning_rate": 6.504217653080862e-06, "loss": 0.2556, "step": 26170 }, { "epoch": 0.47618181487477035, "grad_norm": 1.4973678588867188, "learning_rate": 6.501879975894917e-06, "loss": 0.2477, "step": 26180 }, { "epoch": 0.47636370250459265, "grad_norm": 1.9904648065567017, "learning_rate": 6.499542124539225e-06, "loss": 0.2573, "step": 26190 }, { "epoch": 0.47654559013441494, "grad_norm": 1.509968638420105, "learning_rate": 6.497204099825101e-06, "loss": 0.2534, "step": 26200 }, { "epoch": 0.47672747776423724, "grad_norm": 1.5223350524902344, "learning_rate": 6.49486590256392e-06, "loss": 0.2607, "step": 26210 }, { "epoch": 0.47690936539405954, "grad_norm": 1.5125833749771118, "learning_rate": 6.49252753356712e-06, "loss": 0.2581, "step": 26220 }, { "epoch": 0.47709125302388183, "grad_norm": 1.7082222700119019, "learning_rate": 6.490188993646193e-06, "loss": 0.2559, "step": 26230 }, { "epoch": 0.47727314065370413, "grad_norm": 1.9413002729415894, "learning_rate": 6.487850283612697e-06, "loss": 0.2592, "step": 26240 }, { "epoch": 0.47745502828352643, "grad_norm": 1.6592079401016235, "learning_rate": 6.485511404278242e-06, "loss": 0.2619, "step": 26250 }, { "epoch": 0.4776369159133487, "grad_norm": 1.5205318927764893, "learning_rate": 6.483172356454501e-06, "loss": 0.2482, "step": 26260 }, { "epoch": 0.477818803543171, "grad_norm": 1.7895303964614868, "learning_rate": 6.480833140953208e-06, "loss": 0.2542, "step": 26270 }, { "epoch": 0.4780006911729933, "grad_norm": 1.7535587549209595, "learning_rate": 6.478493758586147e-06, "loss": 0.2617, "step": 26280 }, { "epoch": 0.4781825788028156, "grad_norm": 1.697771430015564, "learning_rate": 6.476154210165168e-06, "loss": 0.2582, "step": 26290 }, { "epoch": 0.4783644664326379, "grad_norm": 1.5046156644821167, "learning_rate": 6.473814496502174e-06, "loss": 0.2465, "step": 26300 }, { "epoch": 0.4785463540624602, "grad_norm": 1.8414281606674194, "learning_rate": 6.471474618409126e-06, "loss": 0.2528, "step": 26310 }, { "epoch": 0.4787282416922825, "grad_norm": 1.855927586555481, "learning_rate": 6.4691345766980455e-06, "loss": 0.2477, "step": 26320 }, { "epoch": 0.4789101293221048, "grad_norm": 1.7250709533691406, "learning_rate": 6.466794372181004e-06, "loss": 0.2547, "step": 26330 }, { "epoch": 0.4790920169519271, "grad_norm": 1.5501047372817993, "learning_rate": 6.464454005670139e-06, "loss": 0.253, "step": 26340 }, { "epoch": 0.4792739045817494, "grad_norm": 1.5970101356506348, "learning_rate": 6.462113477977635e-06, "loss": 0.251, "step": 26350 }, { "epoch": 0.4794557922115717, "grad_norm": 1.7224488258361816, "learning_rate": 6.459772789915736e-06, "loss": 0.26, "step": 26360 }, { "epoch": 0.479637679841394, "grad_norm": 1.5426708459854126, "learning_rate": 6.457431942296744e-06, "loss": 0.2507, "step": 26370 }, { "epoch": 0.4798195674712163, "grad_norm": 1.7351865768432617, "learning_rate": 6.455090935933012e-06, "loss": 0.2531, "step": 26380 }, { "epoch": 0.4800014551010386, "grad_norm": 1.753657579421997, "learning_rate": 6.452749771636952e-06, "loss": 0.2562, "step": 26390 }, { "epoch": 0.4801833427308609, "grad_norm": 1.662170648574829, "learning_rate": 6.45040845022103e-06, "loss": 0.2501, "step": 26400 }, { "epoch": 0.4803652303606832, "grad_norm": 1.4300041198730469, "learning_rate": 6.448066972497762e-06, "loss": 0.2591, "step": 26410 }, { "epoch": 0.4805471179905055, "grad_norm": 1.4923090934753418, "learning_rate": 6.4457253392797255e-06, "loss": 0.2544, "step": 26420 }, { "epoch": 0.4807290056203278, "grad_norm": 1.6640247106552124, "learning_rate": 6.443383551379547e-06, "loss": 0.2528, "step": 26430 }, { "epoch": 0.4809108932501501, "grad_norm": 1.6461212635040283, "learning_rate": 6.441041609609909e-06, "loss": 0.2504, "step": 26440 }, { "epoch": 0.48109278087997237, "grad_norm": 1.7376182079315186, "learning_rate": 6.4386995147835466e-06, "loss": 0.2512, "step": 26450 }, { "epoch": 0.48127466850979467, "grad_norm": 1.4650858640670776, "learning_rate": 6.436357267713245e-06, "loss": 0.2491, "step": 26460 }, { "epoch": 0.48145655613961696, "grad_norm": 1.642574667930603, "learning_rate": 6.434014869211847e-06, "loss": 0.2597, "step": 26470 }, { "epoch": 0.48163844376943926, "grad_norm": 1.8151172399520874, "learning_rate": 6.431672320092246e-06, "loss": 0.2561, "step": 26480 }, { "epoch": 0.48182033139926156, "grad_norm": 1.6968491077423096, "learning_rate": 6.429329621167391e-06, "loss": 0.2508, "step": 26490 }, { "epoch": 0.48200221902908386, "grad_norm": 1.7312697172164917, "learning_rate": 6.4269867732502735e-06, "loss": 0.2555, "step": 26500 }, { "epoch": 0.48218410665890615, "grad_norm": 1.6552058458328247, "learning_rate": 6.4246437771539476e-06, "loss": 0.2609, "step": 26510 }, { "epoch": 0.48236599428872845, "grad_norm": 1.903763771057129, "learning_rate": 6.422300633691511e-06, "loss": 0.2593, "step": 26520 }, { "epoch": 0.4825478819185507, "grad_norm": 1.649354100227356, "learning_rate": 6.419957343676119e-06, "loss": 0.2482, "step": 26530 }, { "epoch": 0.482729769548373, "grad_norm": 1.7117300033569336, "learning_rate": 6.417613907920972e-06, "loss": 0.2532, "step": 26540 }, { "epoch": 0.4829116571781953, "grad_norm": 1.8833059072494507, "learning_rate": 6.415270327239326e-06, "loss": 0.26, "step": 26550 }, { "epoch": 0.4830935448080176, "grad_norm": 1.5429589748382568, "learning_rate": 6.41292660244448e-06, "loss": 0.2545, "step": 26560 }, { "epoch": 0.4832754324378399, "grad_norm": 1.6961249113082886, "learning_rate": 6.410582734349793e-06, "loss": 0.2529, "step": 26570 }, { "epoch": 0.4834573200676622, "grad_norm": 1.9985781908035278, "learning_rate": 6.408238723768665e-06, "loss": 0.252, "step": 26580 }, { "epoch": 0.4836392076974845, "grad_norm": 1.5959819555282593, "learning_rate": 6.40589457151455e-06, "loss": 0.2533, "step": 26590 }, { "epoch": 0.48382109532730677, "grad_norm": 1.730867862701416, "learning_rate": 6.403550278400951e-06, "loss": 0.2516, "step": 26600 }, { "epoch": 0.48400298295712907, "grad_norm": 1.5396891832351685, "learning_rate": 6.4012058452414185e-06, "loss": 0.2579, "step": 26610 }, { "epoch": 0.48418487058695137, "grad_norm": 1.5844835042953491, "learning_rate": 6.39886127284955e-06, "loss": 0.255, "step": 26620 }, { "epoch": 0.48436675821677366, "grad_norm": 1.6824878454208374, "learning_rate": 6.396516562038997e-06, "loss": 0.2588, "step": 26630 }, { "epoch": 0.48454864584659596, "grad_norm": 1.7164928913116455, "learning_rate": 6.3941717136234526e-06, "loss": 0.2523, "step": 26640 }, { "epoch": 0.48473053347641826, "grad_norm": 1.8169788122177124, "learning_rate": 6.391826728416661e-06, "loss": 0.2575, "step": 26650 }, { "epoch": 0.48491242110624055, "grad_norm": 1.9913438558578491, "learning_rate": 6.389481607232414e-06, "loss": 0.2581, "step": 26660 }, { "epoch": 0.48509430873606285, "grad_norm": 1.4283050298690796, "learning_rate": 6.38713635088455e-06, "loss": 0.2555, "step": 26670 }, { "epoch": 0.48527619636588515, "grad_norm": 1.600776195526123, "learning_rate": 6.3847909601869525e-06, "loss": 0.2546, "step": 26680 }, { "epoch": 0.48545808399570745, "grad_norm": 1.7205116748809814, "learning_rate": 6.382445435953556e-06, "loss": 0.2569, "step": 26690 }, { "epoch": 0.48563997162552974, "grad_norm": 1.750258207321167, "learning_rate": 6.380099778998337e-06, "loss": 0.2568, "step": 26700 }, { "epoch": 0.48582185925535204, "grad_norm": 1.6336965560913086, "learning_rate": 6.3777539901353215e-06, "loss": 0.2576, "step": 26710 }, { "epoch": 0.48600374688517434, "grad_norm": 1.6483572721481323, "learning_rate": 6.375408070178577e-06, "loss": 0.2586, "step": 26720 }, { "epoch": 0.48618563451499663, "grad_norm": 1.7365293502807617, "learning_rate": 6.3730620199422195e-06, "loss": 0.2501, "step": 26730 }, { "epoch": 0.48636752214481893, "grad_norm": 1.5670487880706787, "learning_rate": 6.370715840240412e-06, "loss": 0.2582, "step": 26740 }, { "epoch": 0.48654940977464123, "grad_norm": 1.5575616359710693, "learning_rate": 6.36836953188736e-06, "loss": 0.2501, "step": 26750 }, { "epoch": 0.4867312974044635, "grad_norm": 1.8665286302566528, "learning_rate": 6.36602309569731e-06, "loss": 0.2603, "step": 26760 }, { "epoch": 0.4869131850342858, "grad_norm": 1.6622121334075928, "learning_rate": 6.363676532484559e-06, "loss": 0.2578, "step": 26770 }, { "epoch": 0.4870950726641081, "grad_norm": 2.0481009483337402, "learning_rate": 6.361329843063448e-06, "loss": 0.2576, "step": 26780 }, { "epoch": 0.4872769602939304, "grad_norm": 1.8281711339950562, "learning_rate": 6.358983028248358e-06, "loss": 0.2542, "step": 26790 }, { "epoch": 0.4874588479237527, "grad_norm": 1.6463584899902344, "learning_rate": 6.356636088853712e-06, "loss": 0.2474, "step": 26800 }, { "epoch": 0.487640735553575, "grad_norm": 1.8716294765472412, "learning_rate": 6.354289025693982e-06, "loss": 0.2511, "step": 26810 }, { "epoch": 0.4878226231833973, "grad_norm": 1.3832165002822876, "learning_rate": 6.351941839583683e-06, "loss": 0.2475, "step": 26820 }, { "epoch": 0.4880045108132196, "grad_norm": 1.8266948461532593, "learning_rate": 6.3495945313373655e-06, "loss": 0.249, "step": 26830 }, { "epoch": 0.4881863984430419, "grad_norm": 1.5635591745376587, "learning_rate": 6.347247101769629e-06, "loss": 0.2493, "step": 26840 }, { "epoch": 0.4883682860728642, "grad_norm": 1.716919183731079, "learning_rate": 6.3448995516951114e-06, "loss": 0.2543, "step": 26850 }, { "epoch": 0.4885501737026865, "grad_norm": 1.588822364807129, "learning_rate": 6.342551881928496e-06, "loss": 0.2488, "step": 26860 }, { "epoch": 0.4887320613325088, "grad_norm": 1.5568591356277466, "learning_rate": 6.340204093284504e-06, "loss": 0.2496, "step": 26870 }, { "epoch": 0.4889139489623311, "grad_norm": 1.5851210355758667, "learning_rate": 6.337856186577901e-06, "loss": 0.256, "step": 26880 }, { "epoch": 0.4890958365921534, "grad_norm": 1.9514456987380981, "learning_rate": 6.335508162623492e-06, "loss": 0.257, "step": 26890 }, { "epoch": 0.4892777242219757, "grad_norm": 1.4798442125320435, "learning_rate": 6.333160022236118e-06, "loss": 0.2539, "step": 26900 }, { "epoch": 0.489459611851798, "grad_norm": 1.6771801710128784, "learning_rate": 6.330811766230671e-06, "loss": 0.2528, "step": 26910 }, { "epoch": 0.4896414994816203, "grad_norm": 1.6954429149627686, "learning_rate": 6.328463395422074e-06, "loss": 0.2503, "step": 26920 }, { "epoch": 0.4898233871114426, "grad_norm": 1.7941280603408813, "learning_rate": 6.326114910625295e-06, "loss": 0.2517, "step": 26930 }, { "epoch": 0.4900052747412649, "grad_norm": 1.6449493169784546, "learning_rate": 6.3237663126553385e-06, "loss": 0.2487, "step": 26940 }, { "epoch": 0.4901871623710871, "grad_norm": 1.8072386980056763, "learning_rate": 6.321417602327249e-06, "loss": 0.2606, "step": 26950 }, { "epoch": 0.4903690500009094, "grad_norm": 1.641465187072754, "learning_rate": 6.319068780456112e-06, "loss": 0.257, "step": 26960 }, { "epoch": 0.4905509376307317, "grad_norm": 1.5160104036331177, "learning_rate": 6.316719847857048e-06, "loss": 0.2571, "step": 26970 }, { "epoch": 0.490732825260554, "grad_norm": 1.8566166162490845, "learning_rate": 6.31437080534522e-06, "loss": 0.254, "step": 26980 }, { "epoch": 0.4909147128903763, "grad_norm": 1.8614758253097534, "learning_rate": 6.312021653735827e-06, "loss": 0.2516, "step": 26990 }, { "epoch": 0.4910966005201986, "grad_norm": 1.5409502983093262, "learning_rate": 6.309672393844104e-06, "loss": 0.2568, "step": 27000 }, { "epoch": 0.4912784881500209, "grad_norm": 1.5298179388046265, "learning_rate": 6.307323026485328e-06, "loss": 0.2631, "step": 27010 }, { "epoch": 0.4914603757798432, "grad_norm": 1.838226318359375, "learning_rate": 6.304973552474808e-06, "loss": 0.2611, "step": 27020 }, { "epoch": 0.4916422634096655, "grad_norm": 1.5697139501571655, "learning_rate": 6.302623972627897e-06, "loss": 0.2539, "step": 27030 }, { "epoch": 0.4918241510394878, "grad_norm": 1.712022304534912, "learning_rate": 6.3002742877599775e-06, "loss": 0.2581, "step": 27040 }, { "epoch": 0.4920060386693101, "grad_norm": 1.992040991783142, "learning_rate": 6.297924498686472e-06, "loss": 0.2499, "step": 27050 }, { "epoch": 0.4921879262991324, "grad_norm": 1.7539187669754028, "learning_rate": 6.295574606222838e-06, "loss": 0.2585, "step": 27060 }, { "epoch": 0.4923698139289547, "grad_norm": 1.866863489151001, "learning_rate": 6.293224611184573e-06, "loss": 0.2566, "step": 27070 }, { "epoch": 0.492551701558777, "grad_norm": 1.9154733419418335, "learning_rate": 6.290874514387203e-06, "loss": 0.2508, "step": 27080 }, { "epoch": 0.4927335891885993, "grad_norm": 1.7802720069885254, "learning_rate": 6.288524316646296e-06, "loss": 0.2501, "step": 27090 }, { "epoch": 0.49291547681842157, "grad_norm": 1.6577695608139038, "learning_rate": 6.28617401877745e-06, "loss": 0.2523, "step": 27100 }, { "epoch": 0.49309736444824387, "grad_norm": 1.7935168743133545, "learning_rate": 6.283823621596301e-06, "loss": 0.2472, "step": 27110 }, { "epoch": 0.49327925207806617, "grad_norm": 1.9505757093429565, "learning_rate": 6.281473125918518e-06, "loss": 0.2451, "step": 27120 }, { "epoch": 0.49346113970788846, "grad_norm": 1.5503737926483154, "learning_rate": 6.279122532559804e-06, "loss": 0.2533, "step": 27130 }, { "epoch": 0.49364302733771076, "grad_norm": 1.6249921321868896, "learning_rate": 6.276771842335897e-06, "loss": 0.2503, "step": 27140 }, { "epoch": 0.49382491496753306, "grad_norm": 1.5647251605987549, "learning_rate": 6.2744210560625675e-06, "loss": 0.2513, "step": 27150 }, { "epoch": 0.49400680259735535, "grad_norm": 1.7004214525222778, "learning_rate": 6.27207017455562e-06, "loss": 0.2522, "step": 27160 }, { "epoch": 0.49418869022717765, "grad_norm": 2.315734624862671, "learning_rate": 6.26971919863089e-06, "loss": 0.2518, "step": 27170 }, { "epoch": 0.49437057785699995, "grad_norm": 2.0240654945373535, "learning_rate": 6.267368129104252e-06, "loss": 0.2582, "step": 27180 }, { "epoch": 0.49455246548682225, "grad_norm": 1.6310930252075195, "learning_rate": 6.265016966791605e-06, "loss": 0.2492, "step": 27190 }, { "epoch": 0.49473435311664454, "grad_norm": 1.6598321199417114, "learning_rate": 6.262665712508884e-06, "loss": 0.2546, "step": 27200 }, { "epoch": 0.49491624074646684, "grad_norm": 1.4973392486572266, "learning_rate": 6.260314367072059e-06, "loss": 0.2535, "step": 27210 }, { "epoch": 0.49509812837628914, "grad_norm": 1.5635446310043335, "learning_rate": 6.2579629312971234e-06, "loss": 0.2481, "step": 27220 }, { "epoch": 0.49528001600611143, "grad_norm": 1.5276976823806763, "learning_rate": 6.255611406000112e-06, "loss": 0.2512, "step": 27230 }, { "epoch": 0.49546190363593373, "grad_norm": 1.785988211631775, "learning_rate": 6.253259791997082e-06, "loss": 0.2507, "step": 27240 }, { "epoch": 0.49564379126575603, "grad_norm": 1.7726504802703857, "learning_rate": 6.250908090104124e-06, "loss": 0.25, "step": 27250 }, { "epoch": 0.4958256788955783, "grad_norm": 1.668331265449524, "learning_rate": 6.248556301137365e-06, "loss": 0.2522, "step": 27260 }, { "epoch": 0.4960075665254006, "grad_norm": 1.6954641342163086, "learning_rate": 6.246204425912953e-06, "loss": 0.2552, "step": 27270 }, { "epoch": 0.4961894541552229, "grad_norm": 1.579798936843872, "learning_rate": 6.243852465247072e-06, "loss": 0.2566, "step": 27280 }, { "epoch": 0.4963713417850452, "grad_norm": 1.5814107656478882, "learning_rate": 6.241500419955934e-06, "loss": 0.2584, "step": 27290 }, { "epoch": 0.4965532294148675, "grad_norm": 1.7332746982574463, "learning_rate": 6.239148290855779e-06, "loss": 0.2528, "step": 27300 }, { "epoch": 0.4967351170446898, "grad_norm": 2.0022318363189697, "learning_rate": 6.236796078762879e-06, "loss": 0.2495, "step": 27310 }, { "epoch": 0.4969170046745121, "grad_norm": 1.745190978050232, "learning_rate": 6.234443784493531e-06, "loss": 0.2538, "step": 27320 }, { "epoch": 0.4970988923043344, "grad_norm": 1.6585222482681274, "learning_rate": 6.232091408864065e-06, "loss": 0.2583, "step": 27330 }, { "epoch": 0.4972807799341567, "grad_norm": 1.583263635635376, "learning_rate": 6.229738952690835e-06, "loss": 0.25, "step": 27340 }, { "epoch": 0.497462667563979, "grad_norm": 1.623914122581482, "learning_rate": 6.2273864167902256e-06, "loss": 0.262, "step": 27350 }, { "epoch": 0.4976445551938013, "grad_norm": 1.5719739198684692, "learning_rate": 6.225033801978647e-06, "loss": 0.2514, "step": 27360 }, { "epoch": 0.49782644282362354, "grad_norm": 1.6185269355773926, "learning_rate": 6.2226811090725405e-06, "loss": 0.2558, "step": 27370 }, { "epoch": 0.49800833045344584, "grad_norm": 1.6621936559677124, "learning_rate": 6.220328338888372e-06, "loss": 0.2551, "step": 27380 }, { "epoch": 0.49819021808326813, "grad_norm": 1.4732517004013062, "learning_rate": 6.217975492242632e-06, "loss": 0.248, "step": 27390 }, { "epoch": 0.49837210571309043, "grad_norm": 2.0537285804748535, "learning_rate": 6.21562256995184e-06, "loss": 0.254, "step": 27400 }, { "epoch": 0.4985539933429127, "grad_norm": 1.8530910015106201, "learning_rate": 6.213269572832546e-06, "loss": 0.259, "step": 27410 }, { "epoch": 0.498735880972735, "grad_norm": 1.7508877515792847, "learning_rate": 6.210916501701315e-06, "loss": 0.246, "step": 27420 }, { "epoch": 0.4989177686025573, "grad_norm": 1.6267504692077637, "learning_rate": 6.208563357374749e-06, "loss": 0.2439, "step": 27430 }, { "epoch": 0.4990996562323796, "grad_norm": 1.8121148347854614, "learning_rate": 6.2062101406694685e-06, "loss": 0.2482, "step": 27440 }, { "epoch": 0.4992815438622019, "grad_norm": 1.6939607858657837, "learning_rate": 6.203856852402121e-06, "loss": 0.2568, "step": 27450 }, { "epoch": 0.4994634314920242, "grad_norm": 1.671776294708252, "learning_rate": 6.201503493389382e-06, "loss": 0.2642, "step": 27460 }, { "epoch": 0.4996453191218465, "grad_norm": 1.4929838180541992, "learning_rate": 6.199150064447945e-06, "loss": 0.2566, "step": 27470 }, { "epoch": 0.4998272067516688, "grad_norm": 1.6434372663497925, "learning_rate": 6.196796566394534e-06, "loss": 0.2563, "step": 27480 }, { "epoch": 0.5000090943814911, "grad_norm": 1.6148639917373657, "learning_rate": 6.194443000045891e-06, "loss": 0.2515, "step": 27490 }, { "epoch": 0.5001909820113134, "grad_norm": 1.5178687572479248, "learning_rate": 6.19208936621879e-06, "loss": 0.2563, "step": 27500 }, { "epoch": 0.5003728696411357, "grad_norm": 1.4275343418121338, "learning_rate": 6.189735665730021e-06, "loss": 0.2502, "step": 27510 }, { "epoch": 0.500554757270958, "grad_norm": 1.5366615056991577, "learning_rate": 6.187381899396399e-06, "loss": 0.246, "step": 27520 }, { "epoch": 0.5007366449007803, "grad_norm": 1.4631476402282715, "learning_rate": 6.185028068034762e-06, "loss": 0.2536, "step": 27530 }, { "epoch": 0.5009185325306026, "grad_norm": 1.5529385805130005, "learning_rate": 6.1826741724619735e-06, "loss": 0.2518, "step": 27540 }, { "epoch": 0.5011004201604249, "grad_norm": 1.5317811965942383, "learning_rate": 6.180320213494916e-06, "loss": 0.2531, "step": 27550 }, { "epoch": 0.5012823077902472, "grad_norm": 1.6695886850357056, "learning_rate": 6.177966191950493e-06, "loss": 0.2508, "step": 27560 }, { "epoch": 0.5014641954200695, "grad_norm": 1.9448635578155518, "learning_rate": 6.175612108645634e-06, "loss": 0.248, "step": 27570 }, { "epoch": 0.5016460830498918, "grad_norm": 1.5857999324798584, "learning_rate": 6.1732579643972875e-06, "loss": 0.2523, "step": 27580 }, { "epoch": 0.5018279706797141, "grad_norm": 1.661697268486023, "learning_rate": 6.17090376002242e-06, "loss": 0.255, "step": 27590 }, { "epoch": 0.5020098583095364, "grad_norm": 1.6329433917999268, "learning_rate": 6.168549496338026e-06, "loss": 0.2521, "step": 27600 }, { "epoch": 0.5021917459393587, "grad_norm": 1.6956223249435425, "learning_rate": 6.1661951741611124e-06, "loss": 0.255, "step": 27610 }, { "epoch": 0.502373633569181, "grad_norm": 1.5853075981140137, "learning_rate": 6.163840794308715e-06, "loss": 0.2536, "step": 27620 }, { "epoch": 0.5025555211990033, "grad_norm": 1.9227761030197144, "learning_rate": 6.161486357597882e-06, "loss": 0.255, "step": 27630 }, { "epoch": 0.5027374088288256, "grad_norm": 1.8324581384658813, "learning_rate": 6.159131864845686e-06, "loss": 0.2558, "step": 27640 }, { "epoch": 0.5029192964586479, "grad_norm": 1.5495785474777222, "learning_rate": 6.156777316869218e-06, "loss": 0.247, "step": 27650 }, { "epoch": 0.5031011840884702, "grad_norm": 1.4700407981872559, "learning_rate": 6.154422714485586e-06, "loss": 0.2505, "step": 27660 }, { "epoch": 0.5032830717182925, "grad_norm": 1.7140799760818481, "learning_rate": 6.152068058511921e-06, "loss": 0.2494, "step": 27670 }, { "epoch": 0.5034649593481147, "grad_norm": 1.5895570516586304, "learning_rate": 6.149713349765369e-06, "loss": 0.2476, "step": 27680 }, { "epoch": 0.503646846977937, "grad_norm": 1.543782353401184, "learning_rate": 6.147358589063094e-06, "loss": 0.2496, "step": 27690 }, { "epoch": 0.5038287346077593, "grad_norm": 1.576270580291748, "learning_rate": 6.145003777222283e-06, "loss": 0.2517, "step": 27700 }, { "epoch": 0.5040106222375816, "grad_norm": 1.9441211223602295, "learning_rate": 6.142648915060132e-06, "loss": 0.2533, "step": 27710 }, { "epoch": 0.5041925098674039, "grad_norm": 1.7413835525512695, "learning_rate": 6.140294003393868e-06, "loss": 0.2571, "step": 27720 }, { "epoch": 0.5043743974972262, "grad_norm": 1.584546446800232, "learning_rate": 6.137939043040718e-06, "loss": 0.2489, "step": 27730 }, { "epoch": 0.5045562851270485, "grad_norm": 1.5731713771820068, "learning_rate": 6.13558403481794e-06, "loss": 0.2522, "step": 27740 }, { "epoch": 0.5047381727568708, "grad_norm": 1.8399882316589355, "learning_rate": 6.133228979542804e-06, "loss": 0.257, "step": 27750 }, { "epoch": 0.5049200603866931, "grad_norm": 1.656635046005249, "learning_rate": 6.1308738780325925e-06, "loss": 0.2553, "step": 27760 }, { "epoch": 0.5051019480165154, "grad_norm": 1.7557672262191772, "learning_rate": 6.12851873110461e-06, "loss": 0.2476, "step": 27770 }, { "epoch": 0.5052838356463377, "grad_norm": 1.6150859594345093, "learning_rate": 6.1261635395761746e-06, "loss": 0.2532, "step": 27780 }, { "epoch": 0.50546572327616, "grad_norm": 1.9174137115478516, "learning_rate": 6.123808304264616e-06, "loss": 0.2538, "step": 27790 }, { "epoch": 0.5056476109059823, "grad_norm": 1.7559070587158203, "learning_rate": 6.1214530259872854e-06, "loss": 0.2471, "step": 27800 }, { "epoch": 0.5058294985358046, "grad_norm": 1.6642966270446777, "learning_rate": 6.119097705561546e-06, "loss": 0.2513, "step": 27810 }, { "epoch": 0.5060113861656269, "grad_norm": 1.8800779581069946, "learning_rate": 6.116742343804774e-06, "loss": 0.244, "step": 27820 }, { "epoch": 0.5061932737954492, "grad_norm": 1.6106500625610352, "learning_rate": 6.114386941534361e-06, "loss": 0.2475, "step": 27830 }, { "epoch": 0.5063751614252715, "grad_norm": 1.4349795579910278, "learning_rate": 6.112031499567716e-06, "loss": 0.2563, "step": 27840 }, { "epoch": 0.5065570490550938, "grad_norm": 1.4453325271606445, "learning_rate": 6.109676018722258e-06, "loss": 0.251, "step": 27850 }, { "epoch": 0.5067389366849161, "grad_norm": 1.8032881021499634, "learning_rate": 6.107320499815419e-06, "loss": 0.2526, "step": 27860 }, { "epoch": 0.5069208243147384, "grad_norm": 1.512270450592041, "learning_rate": 6.1049649436646485e-06, "loss": 0.2501, "step": 27870 }, { "epoch": 0.5071027119445607, "grad_norm": 2.1545722484588623, "learning_rate": 6.1026093510874015e-06, "loss": 0.2547, "step": 27880 }, { "epoch": 0.507284599574383, "grad_norm": 1.974535346031189, "learning_rate": 6.100253722901156e-06, "loss": 0.2517, "step": 27890 }, { "epoch": 0.5074664872042053, "grad_norm": 1.5272411108016968, "learning_rate": 6.0978980599233925e-06, "loss": 0.252, "step": 27900 }, { "epoch": 0.5076483748340276, "grad_norm": 1.4532371759414673, "learning_rate": 6.095542362971608e-06, "loss": 0.2531, "step": 27910 }, { "epoch": 0.5078302624638499, "grad_norm": 1.4867517948150635, "learning_rate": 6.093186632863314e-06, "loss": 0.2548, "step": 27920 }, { "epoch": 0.5080121500936722, "grad_norm": 2.2128169536590576, "learning_rate": 6.090830870416028e-06, "loss": 0.2568, "step": 27930 }, { "epoch": 0.5081940377234945, "grad_norm": 1.7394620180130005, "learning_rate": 6.088475076447281e-06, "loss": 0.2483, "step": 27940 }, { "epoch": 0.5083759253533168, "grad_norm": 1.8515284061431885, "learning_rate": 6.0861192517746156e-06, "loss": 0.2533, "step": 27950 }, { "epoch": 0.5085578129831391, "grad_norm": 1.5345126390457153, "learning_rate": 6.083763397215585e-06, "loss": 0.253, "step": 27960 }, { "epoch": 0.5087397006129614, "grad_norm": 1.599887490272522, "learning_rate": 6.081407513587754e-06, "loss": 0.2508, "step": 27970 }, { "epoch": 0.5089215882427837, "grad_norm": 1.5257781744003296, "learning_rate": 6.079051601708693e-06, "loss": 0.2486, "step": 27980 }, { "epoch": 0.509103475872606, "grad_norm": 1.8190206289291382, "learning_rate": 6.076695662395986e-06, "loss": 0.2485, "step": 27990 }, { "epoch": 0.5092853635024281, "grad_norm": 1.591943621635437, "learning_rate": 6.074339696467226e-06, "loss": 0.2528, "step": 28000 }, { "epoch": 0.5094672511322504, "grad_norm": 1.643161416053772, "learning_rate": 6.071983704740014e-06, "loss": 0.2556, "step": 28010 }, { "epoch": 0.5096491387620727, "grad_norm": 1.8511171340942383, "learning_rate": 6.0696276880319636e-06, "loss": 0.2459, "step": 28020 }, { "epoch": 0.509831026391895, "grad_norm": 1.8285157680511475, "learning_rate": 6.0672716471606894e-06, "loss": 0.2542, "step": 28030 }, { "epoch": 0.5100129140217173, "grad_norm": 1.7867907285690308, "learning_rate": 6.0649155829438235e-06, "loss": 0.2464, "step": 28040 }, { "epoch": 0.5101948016515396, "grad_norm": 1.5665539503097534, "learning_rate": 6.062559496199001e-06, "loss": 0.2468, "step": 28050 }, { "epoch": 0.5103766892813619, "grad_norm": 1.538058876991272, "learning_rate": 6.060203387743862e-06, "loss": 0.2513, "step": 28060 }, { "epoch": 0.5105585769111842, "grad_norm": 1.796791434288025, "learning_rate": 6.057847258396065e-06, "loss": 0.2486, "step": 28070 }, { "epoch": 0.5107404645410065, "grad_norm": 1.8340249061584473, "learning_rate": 6.055491108973261e-06, "loss": 0.2499, "step": 28080 }, { "epoch": 0.5109223521708288, "grad_norm": 1.662392497062683, "learning_rate": 6.053134940293121e-06, "loss": 0.2509, "step": 28090 }, { "epoch": 0.5111042398006511, "grad_norm": 1.7514439821243286, "learning_rate": 6.050778753173314e-06, "loss": 0.2514, "step": 28100 }, { "epoch": 0.5112861274304734, "grad_norm": 1.614470362663269, "learning_rate": 6.048422548431519e-06, "loss": 0.2539, "step": 28110 }, { "epoch": 0.5114680150602957, "grad_norm": 1.8905019760131836, "learning_rate": 6.046066326885423e-06, "loss": 0.2542, "step": 28120 }, { "epoch": 0.511649902690118, "grad_norm": 1.5886870622634888, "learning_rate": 6.043710089352714e-06, "loss": 0.2486, "step": 28130 }, { "epoch": 0.5118317903199403, "grad_norm": 1.5729236602783203, "learning_rate": 6.041353836651088e-06, "loss": 0.2538, "step": 28140 }, { "epoch": 0.5120136779497626, "grad_norm": 1.651007056236267, "learning_rate": 6.0389975695982484e-06, "loss": 0.2512, "step": 28150 }, { "epoch": 0.5121955655795849, "grad_norm": 1.685370922088623, "learning_rate": 6.036641289011901e-06, "loss": 0.2471, "step": 28160 }, { "epoch": 0.5123774532094072, "grad_norm": 1.4535472393035889, "learning_rate": 6.034284995709756e-06, "loss": 0.248, "step": 28170 }, { "epoch": 0.5125593408392295, "grad_norm": 1.5880937576293945, "learning_rate": 6.031928690509528e-06, "loss": 0.2545, "step": 28180 }, { "epoch": 0.5127412284690518, "grad_norm": 1.4943636655807495, "learning_rate": 6.029572374228937e-06, "loss": 0.2449, "step": 28190 }, { "epoch": 0.5129231160988741, "grad_norm": 1.7186495065689087, "learning_rate": 6.027216047685709e-06, "loss": 0.251, "step": 28200 }, { "epoch": 0.5131050037286964, "grad_norm": 1.6143587827682495, "learning_rate": 6.0248597116975695e-06, "loss": 0.2507, "step": 28210 }, { "epoch": 0.5132868913585187, "grad_norm": 1.5314691066741943, "learning_rate": 6.022503367082247e-06, "loss": 0.2496, "step": 28220 }, { "epoch": 0.513468778988341, "grad_norm": 1.6237645149230957, "learning_rate": 6.020147014657476e-06, "loss": 0.2479, "step": 28230 }, { "epoch": 0.5136506666181633, "grad_norm": 1.778588056564331, "learning_rate": 6.0177906552409935e-06, "loss": 0.2556, "step": 28240 }, { "epoch": 0.5138325542479856, "grad_norm": 1.827941656112671, "learning_rate": 6.015434289650538e-06, "loss": 0.2494, "step": 28250 }, { "epoch": 0.5140144418778079, "grad_norm": 1.6456880569458008, "learning_rate": 6.013077918703848e-06, "loss": 0.2513, "step": 28260 }, { "epoch": 0.5141963295076302, "grad_norm": 1.6958799362182617, "learning_rate": 6.010721543218667e-06, "loss": 0.2557, "step": 28270 }, { "epoch": 0.5143782171374525, "grad_norm": 1.5531195402145386, "learning_rate": 6.008365164012738e-06, "loss": 0.2564, "step": 28280 }, { "epoch": 0.5145601047672748, "grad_norm": 1.6803406476974487, "learning_rate": 6.006008781903808e-06, "loss": 0.2473, "step": 28290 }, { "epoch": 0.5147419923970971, "grad_norm": 1.6246367692947388, "learning_rate": 6.003652397709622e-06, "loss": 0.2447, "step": 28300 }, { "epoch": 0.5149238800269194, "grad_norm": 1.7665361166000366, "learning_rate": 6.001296012247928e-06, "loss": 0.2492, "step": 28310 }, { "epoch": 0.5151057676567417, "grad_norm": 1.4759951829910278, "learning_rate": 5.998939626336473e-06, "loss": 0.2536, "step": 28320 }, { "epoch": 0.515287655286564, "grad_norm": 1.7069207429885864, "learning_rate": 5.996583240793004e-06, "loss": 0.2408, "step": 28330 }, { "epoch": 0.5154695429163862, "grad_norm": 1.480022668838501, "learning_rate": 5.994226856435269e-06, "loss": 0.2524, "step": 28340 }, { "epoch": 0.5156514305462085, "grad_norm": 1.7618958950042725, "learning_rate": 5.991870474081014e-06, "loss": 0.2468, "step": 28350 }, { "epoch": 0.5158333181760308, "grad_norm": 1.6927568912506104, "learning_rate": 5.989514094547988e-06, "loss": 0.2531, "step": 28360 }, { "epoch": 0.5160152058058531, "grad_norm": 1.722606897354126, "learning_rate": 5.987157718653933e-06, "loss": 0.2515, "step": 28370 }, { "epoch": 0.5161970934356754, "grad_norm": 1.723135232925415, "learning_rate": 5.984801347216595e-06, "loss": 0.2578, "step": 28380 }, { "epoch": 0.5163789810654977, "grad_norm": 1.5211652517318726, "learning_rate": 5.9824449810537156e-06, "loss": 0.2509, "step": 28390 }, { "epoch": 0.51656086869532, "grad_norm": 1.6230201721191406, "learning_rate": 5.980088620983038e-06, "loss": 0.25, "step": 28400 }, { "epoch": 0.5167427563251423, "grad_norm": 1.5877429246902466, "learning_rate": 5.9777322678222985e-06, "loss": 0.2487, "step": 28410 }, { "epoch": 0.5169246439549646, "grad_norm": 1.7243852615356445, "learning_rate": 5.975375922389234e-06, "loss": 0.2574, "step": 28420 }, { "epoch": 0.5171065315847869, "grad_norm": 1.7141191959381104, "learning_rate": 5.973019585501577e-06, "loss": 0.2585, "step": 28430 }, { "epoch": 0.5172884192146092, "grad_norm": 1.5358320474624634, "learning_rate": 5.97066325797706e-06, "loss": 0.2452, "step": 28440 }, { "epoch": 0.5174703068444315, "grad_norm": 1.678760051727295, "learning_rate": 5.968306940633408e-06, "loss": 0.2549, "step": 28450 }, { "epoch": 0.5176521944742538, "grad_norm": 1.5321565866470337, "learning_rate": 5.965950634288347e-06, "loss": 0.2491, "step": 28460 }, { "epoch": 0.5178340821040761, "grad_norm": 1.4173994064331055, "learning_rate": 5.963594339759598e-06, "loss": 0.2541, "step": 28470 }, { "epoch": 0.5180159697338984, "grad_norm": 1.6424896717071533, "learning_rate": 5.961238057864872e-06, "loss": 0.2544, "step": 28480 }, { "epoch": 0.5181978573637207, "grad_norm": 1.4406923055648804, "learning_rate": 5.958881789421886e-06, "loss": 0.2446, "step": 28490 }, { "epoch": 0.518379744993543, "grad_norm": 1.5553704500198364, "learning_rate": 5.956525535248341e-06, "loss": 0.2478, "step": 28500 }, { "epoch": 0.5185616326233653, "grad_norm": 1.730959177017212, "learning_rate": 5.954169296161944e-06, "loss": 0.2496, "step": 28510 }, { "epoch": 0.5187435202531876, "grad_norm": 1.5489145517349243, "learning_rate": 5.95181307298039e-06, "loss": 0.2514, "step": 28520 }, { "epoch": 0.5189254078830099, "grad_norm": 1.5761852264404297, "learning_rate": 5.9494568665213695e-06, "loss": 0.2516, "step": 28530 }, { "epoch": 0.5191072955128322, "grad_norm": 1.5120192766189575, "learning_rate": 5.947100677602568e-06, "loss": 0.2526, "step": 28540 }, { "epoch": 0.5192891831426545, "grad_norm": 1.4940146207809448, "learning_rate": 5.944744507041665e-06, "loss": 0.2562, "step": 28550 }, { "epoch": 0.5194710707724768, "grad_norm": 1.816493034362793, "learning_rate": 5.942388355656335e-06, "loss": 0.2579, "step": 28560 }, { "epoch": 0.5196529584022991, "grad_norm": 1.6247239112854004, "learning_rate": 5.9400322242642394e-06, "loss": 0.2506, "step": 28570 }, { "epoch": 0.5198348460321214, "grad_norm": 1.7014455795288086, "learning_rate": 5.937676113683042e-06, "loss": 0.2552, "step": 28580 }, { "epoch": 0.5200167336619437, "grad_norm": 1.5632561445236206, "learning_rate": 5.9353200247303935e-06, "loss": 0.2602, "step": 28590 }, { "epoch": 0.520198621291766, "grad_norm": 1.64927339553833, "learning_rate": 5.932963958223937e-06, "loss": 0.2537, "step": 28600 }, { "epoch": 0.5203805089215883, "grad_norm": 1.7362208366394043, "learning_rate": 5.930607914981312e-06, "loss": 0.2587, "step": 28610 }, { "epoch": 0.5205623965514106, "grad_norm": 1.5926878452301025, "learning_rate": 5.928251895820145e-06, "loss": 0.2487, "step": 28620 }, { "epoch": 0.5207442841812329, "grad_norm": 1.5896412134170532, "learning_rate": 5.925895901558057e-06, "loss": 0.2503, "step": 28630 }, { "epoch": 0.5209261718110552, "grad_norm": 2.0192463397979736, "learning_rate": 5.92353993301266e-06, "loss": 0.2567, "step": 28640 }, { "epoch": 0.5211080594408775, "grad_norm": 1.631351351737976, "learning_rate": 5.921183991001555e-06, "loss": 0.2539, "step": 28650 }, { "epoch": 0.5212899470706998, "grad_norm": 1.4202334880828857, "learning_rate": 5.91882807634234e-06, "loss": 0.2505, "step": 28660 }, { "epoch": 0.521471834700522, "grad_norm": 1.5418832302093506, "learning_rate": 5.916472189852593e-06, "loss": 0.2573, "step": 28670 }, { "epoch": 0.5216537223303443, "grad_norm": 1.8710976839065552, "learning_rate": 5.914116332349891e-06, "loss": 0.2442, "step": 28680 }, { "epoch": 0.5218356099601666, "grad_norm": 1.644284725189209, "learning_rate": 5.9117605046518e-06, "loss": 0.2482, "step": 28690 }, { "epoch": 0.5220174975899889, "grad_norm": 1.6543011665344238, "learning_rate": 5.909404707575872e-06, "loss": 0.2479, "step": 28700 }, { "epoch": 0.5221993852198112, "grad_norm": 1.5741186141967773, "learning_rate": 5.90704894193965e-06, "loss": 0.2511, "step": 28710 }, { "epoch": 0.5223812728496335, "grad_norm": 1.6234173774719238, "learning_rate": 5.9046932085606675e-06, "loss": 0.251, "step": 28720 }, { "epoch": 0.5225631604794558, "grad_norm": 1.6817890405654907, "learning_rate": 5.902337508256445e-06, "loss": 0.2484, "step": 28730 }, { "epoch": 0.5227450481092781, "grad_norm": 1.5553898811340332, "learning_rate": 5.8999818418444935e-06, "loss": 0.2578, "step": 28740 }, { "epoch": 0.5229269357391004, "grad_norm": 1.3907629251480103, "learning_rate": 5.897626210142307e-06, "loss": 0.2508, "step": 28750 }, { "epoch": 0.5231088233689227, "grad_norm": 1.686766266822815, "learning_rate": 5.895270613967375e-06, "loss": 0.2461, "step": 28760 }, { "epoch": 0.523290710998745, "grad_norm": 1.7154018878936768, "learning_rate": 5.892915054137169e-06, "loss": 0.2535, "step": 28770 }, { "epoch": 0.5234725986285673, "grad_norm": 1.7922029495239258, "learning_rate": 5.890559531469154e-06, "loss": 0.2489, "step": 28780 }, { "epoch": 0.5236544862583896, "grad_norm": 1.7294803857803345, "learning_rate": 5.888204046780772e-06, "loss": 0.2573, "step": 28790 }, { "epoch": 0.5238363738882119, "grad_norm": 2.0101089477539062, "learning_rate": 5.885848600889461e-06, "loss": 0.2518, "step": 28800 }, { "epoch": 0.5240182615180342, "grad_norm": 1.6905360221862793, "learning_rate": 5.883493194612643e-06, "loss": 0.2478, "step": 28810 }, { "epoch": 0.5242001491478565, "grad_norm": 1.7347999811172485, "learning_rate": 5.8811378287677235e-06, "loss": 0.247, "step": 28820 }, { "epoch": 0.5243820367776787, "grad_norm": 1.613142728805542, "learning_rate": 5.878782504172098e-06, "loss": 0.2464, "step": 28830 }, { "epoch": 0.524563924407501, "grad_norm": 1.6827285289764404, "learning_rate": 5.876427221643145e-06, "loss": 0.2469, "step": 28840 }, { "epoch": 0.5247458120373233, "grad_norm": 1.8094978332519531, "learning_rate": 5.874071981998229e-06, "loss": 0.2432, "step": 28850 }, { "epoch": 0.5249276996671456, "grad_norm": 1.820331335067749, "learning_rate": 5.871716786054701e-06, "loss": 0.2526, "step": 28860 }, { "epoch": 0.5251095872969679, "grad_norm": 1.7151222229003906, "learning_rate": 5.869361634629893e-06, "loss": 0.253, "step": 28870 }, { "epoch": 0.5252914749267902, "grad_norm": 1.6140753030776978, "learning_rate": 5.867006528541126e-06, "loss": 0.2496, "step": 28880 }, { "epoch": 0.5254733625566125, "grad_norm": 1.6155871152877808, "learning_rate": 5.864651468605704e-06, "loss": 0.2482, "step": 28890 }, { "epoch": 0.5256552501864348, "grad_norm": 1.4823789596557617, "learning_rate": 5.8622964556409135e-06, "loss": 0.2464, "step": 28900 }, { "epoch": 0.5258371378162571, "grad_norm": 1.4977233409881592, "learning_rate": 5.859941490464025e-06, "loss": 0.2545, "step": 28910 }, { "epoch": 0.5260190254460794, "grad_norm": 1.6373101472854614, "learning_rate": 5.857586573892294e-06, "loss": 0.2461, "step": 28920 }, { "epoch": 0.5262009130759017, "grad_norm": 1.6474307775497437, "learning_rate": 5.855231706742959e-06, "loss": 0.2485, "step": 28930 }, { "epoch": 0.526382800705724, "grad_norm": 1.625998854637146, "learning_rate": 5.852876889833238e-06, "loss": 0.2586, "step": 28940 }, { "epoch": 0.5265646883355463, "grad_norm": 1.780980110168457, "learning_rate": 5.850522123980336e-06, "loss": 0.248, "step": 28950 }, { "epoch": 0.5267465759653686, "grad_norm": 1.8471803665161133, "learning_rate": 5.848167410001439e-06, "loss": 0.2604, "step": 28960 }, { "epoch": 0.5269284635951909, "grad_norm": 1.7598106861114502, "learning_rate": 5.845812748713712e-06, "loss": 0.2438, "step": 28970 }, { "epoch": 0.5271103512250132, "grad_norm": 1.6446505784988403, "learning_rate": 5.843458140934306e-06, "loss": 0.2545, "step": 28980 }, { "epoch": 0.5272922388548354, "grad_norm": 1.8803670406341553, "learning_rate": 5.841103587480351e-06, "loss": 0.2521, "step": 28990 }, { "epoch": 0.5274741264846577, "grad_norm": 1.6409969329833984, "learning_rate": 5.838749089168959e-06, "loss": 0.2469, "step": 29000 }, { "epoch": 0.52765601411448, "grad_norm": 1.968094825744629, "learning_rate": 5.836394646817222e-06, "loss": 0.2511, "step": 29010 }, { "epoch": 0.5278379017443023, "grad_norm": 1.7107840776443481, "learning_rate": 5.834040261242212e-06, "loss": 0.2485, "step": 29020 }, { "epoch": 0.5280197893741246, "grad_norm": 1.712051272392273, "learning_rate": 5.831685933260985e-06, "loss": 0.2555, "step": 29030 }, { "epoch": 0.5282016770039469, "grad_norm": 1.4653308391571045, "learning_rate": 5.829331663690572e-06, "loss": 0.244, "step": 29040 }, { "epoch": 0.5283835646337692, "grad_norm": 1.5708281993865967, "learning_rate": 5.826977453347988e-06, "loss": 0.2485, "step": 29050 }, { "epoch": 0.5285654522635915, "grad_norm": 1.6463810205459595, "learning_rate": 5.824623303050225e-06, "loss": 0.2435, "step": 29060 }, { "epoch": 0.5287473398934138, "grad_norm": 1.5795314311981201, "learning_rate": 5.822269213614253e-06, "loss": 0.2485, "step": 29070 }, { "epoch": 0.5289292275232361, "grad_norm": 1.6581720113754272, "learning_rate": 5.819915185857026e-06, "loss": 0.2473, "step": 29080 }, { "epoch": 0.5291111151530584, "grad_norm": 1.9974408149719238, "learning_rate": 5.8175612205954704e-06, "loss": 0.2496, "step": 29090 }, { "epoch": 0.5292930027828807, "grad_norm": 1.6167153120040894, "learning_rate": 5.815207318646494e-06, "loss": 0.2414, "step": 29100 }, { "epoch": 0.529474890412703, "grad_norm": 1.597580075263977, "learning_rate": 5.812853480826983e-06, "loss": 0.2532, "step": 29110 }, { "epoch": 0.5296567780425253, "grad_norm": 1.8886264562606812, "learning_rate": 5.810499707953803e-06, "loss": 0.2475, "step": 29120 }, { "epoch": 0.5298386656723476, "grad_norm": 1.617472529411316, "learning_rate": 5.808146000843791e-06, "loss": 0.2525, "step": 29130 }, { "epoch": 0.5300205533021699, "grad_norm": 1.5576821565628052, "learning_rate": 5.805792360313765e-06, "loss": 0.2486, "step": 29140 }, { "epoch": 0.5302024409319922, "grad_norm": 1.8574141263961792, "learning_rate": 5.803438787180524e-06, "loss": 0.2515, "step": 29150 }, { "epoch": 0.5303843285618145, "grad_norm": 1.4668805599212646, "learning_rate": 5.801085282260835e-06, "loss": 0.248, "step": 29160 }, { "epoch": 0.5305662161916368, "grad_norm": 1.8084540367126465, "learning_rate": 5.798731846371448e-06, "loss": 0.2477, "step": 29170 }, { "epoch": 0.5307481038214591, "grad_norm": 1.7698315382003784, "learning_rate": 5.7963784803290875e-06, "loss": 0.2585, "step": 29180 }, { "epoch": 0.5309299914512814, "grad_norm": 1.8747092485427856, "learning_rate": 5.79402518495045e-06, "loss": 0.2433, "step": 29190 }, { "epoch": 0.5311118790811037, "grad_norm": 1.5883435010910034, "learning_rate": 5.791671961052215e-06, "loss": 0.2513, "step": 29200 }, { "epoch": 0.531293766710926, "grad_norm": 1.560390830039978, "learning_rate": 5.789318809451028e-06, "loss": 0.2547, "step": 29210 }, { "epoch": 0.5314756543407483, "grad_norm": 1.4489682912826538, "learning_rate": 5.786965730963515e-06, "loss": 0.2444, "step": 29220 }, { "epoch": 0.5316575419705706, "grad_norm": 1.8356773853302002, "learning_rate": 5.7846127264062805e-06, "loss": 0.253, "step": 29230 }, { "epoch": 0.5318394296003929, "grad_norm": 1.7019580602645874, "learning_rate": 5.782259796595893e-06, "loss": 0.2508, "step": 29240 }, { "epoch": 0.5320213172302152, "grad_norm": 1.642907738685608, "learning_rate": 5.779906942348905e-06, "loss": 0.244, "step": 29250 }, { "epoch": 0.5322032048600375, "grad_norm": 1.4067165851593018, "learning_rate": 5.777554164481835e-06, "loss": 0.2539, "step": 29260 }, { "epoch": 0.5323850924898598, "grad_norm": 1.6800825595855713, "learning_rate": 5.7752014638111815e-06, "loss": 0.2438, "step": 29270 }, { "epoch": 0.5325669801196821, "grad_norm": 1.3858803510665894, "learning_rate": 5.772848841153411e-06, "loss": 0.2466, "step": 29280 }, { "epoch": 0.5327488677495044, "grad_norm": 1.583370327949524, "learning_rate": 5.7704962973249655e-06, "loss": 0.2445, "step": 29290 }, { "epoch": 0.5329307553793267, "grad_norm": 1.5527572631835938, "learning_rate": 5.768143833142261e-06, "loss": 0.2481, "step": 29300 }, { "epoch": 0.533112643009149, "grad_norm": 1.5362396240234375, "learning_rate": 5.765791449421681e-06, "loss": 0.2565, "step": 29310 }, { "epoch": 0.5332945306389713, "grad_norm": 1.6855884790420532, "learning_rate": 5.763439146979587e-06, "loss": 0.2429, "step": 29320 }, { "epoch": 0.5334764182687936, "grad_norm": 1.621027946472168, "learning_rate": 5.7610869266323076e-06, "loss": 0.2491, "step": 29330 }, { "epoch": 0.5336583058986158, "grad_norm": 1.568507432937622, "learning_rate": 5.758734789196145e-06, "loss": 0.2541, "step": 29340 }, { "epoch": 0.5338401935284381, "grad_norm": 1.7042890787124634, "learning_rate": 5.7563827354873745e-06, "loss": 0.2504, "step": 29350 }, { "epoch": 0.5340220811582604, "grad_norm": 1.465028166770935, "learning_rate": 5.754030766322237e-06, "loss": 0.2502, "step": 29360 }, { "epoch": 0.5342039687880827, "grad_norm": 1.5872076749801636, "learning_rate": 5.75167888251695e-06, "loss": 0.252, "step": 29370 }, { "epoch": 0.534385856417905, "grad_norm": 1.7091124057769775, "learning_rate": 5.749327084887698e-06, "loss": 0.2505, "step": 29380 }, { "epoch": 0.5345677440477273, "grad_norm": 1.7392852306365967, "learning_rate": 5.746975374250635e-06, "loss": 0.2506, "step": 29390 }, { "epoch": 0.5347496316775496, "grad_norm": 1.4012095928192139, "learning_rate": 5.744623751421888e-06, "loss": 0.2524, "step": 29400 }, { "epoch": 0.5349315193073719, "grad_norm": 1.7729302644729614, "learning_rate": 5.742272217217551e-06, "loss": 0.247, "step": 29410 }, { "epoch": 0.5351134069371942, "grad_norm": 1.9010744094848633, "learning_rate": 5.739920772453686e-06, "loss": 0.2502, "step": 29420 }, { "epoch": 0.5352952945670165, "grad_norm": 1.7126411199569702, "learning_rate": 5.737569417946328e-06, "loss": 0.2492, "step": 29430 }, { "epoch": 0.5354771821968388, "grad_norm": 1.8829882144927979, "learning_rate": 5.735218154511479e-06, "loss": 0.2577, "step": 29440 }, { "epoch": 0.5356590698266611, "grad_norm": 1.6611924171447754, "learning_rate": 5.732866982965107e-06, "loss": 0.2438, "step": 29450 }, { "epoch": 0.5358409574564834, "grad_norm": 1.4220470190048218, "learning_rate": 5.7305159041231515e-06, "loss": 0.2489, "step": 29460 }, { "epoch": 0.5360228450863057, "grad_norm": 1.6747949123382568, "learning_rate": 5.728164918801517e-06, "loss": 0.2528, "step": 29470 }, { "epoch": 0.536204732716128, "grad_norm": 1.6412452459335327, "learning_rate": 5.725814027816078e-06, "loss": 0.2519, "step": 29480 }, { "epoch": 0.5363866203459503, "grad_norm": 1.5039174556732178, "learning_rate": 5.723463231982673e-06, "loss": 0.2429, "step": 29490 }, { "epoch": 0.5365685079757726, "grad_norm": 1.8403292894363403, "learning_rate": 5.7211125321171145e-06, "loss": 0.2561, "step": 29500 }, { "epoch": 0.5367503956055949, "grad_norm": 1.6720281839370728, "learning_rate": 5.718761929035172e-06, "loss": 0.2475, "step": 29510 }, { "epoch": 0.5369322832354172, "grad_norm": 1.5869802236557007, "learning_rate": 5.716411423552588e-06, "loss": 0.2535, "step": 29520 }, { "epoch": 0.5371141708652395, "grad_norm": 1.759123682975769, "learning_rate": 5.7140610164850705e-06, "loss": 0.2573, "step": 29530 }, { "epoch": 0.5372960584950618, "grad_norm": 1.7246074676513672, "learning_rate": 5.7117107086482896e-06, "loss": 0.2496, "step": 29540 }, { "epoch": 0.5374779461248841, "grad_norm": 1.6414995193481445, "learning_rate": 5.709360500857885e-06, "loss": 0.2452, "step": 29550 }, { "epoch": 0.5376598337547064, "grad_norm": 1.5498323440551758, "learning_rate": 5.707010393929462e-06, "loss": 0.248, "step": 29560 }, { "epoch": 0.5378417213845287, "grad_norm": 1.5664215087890625, "learning_rate": 5.7046603886785866e-06, "loss": 0.2464, "step": 29570 }, { "epoch": 0.538023609014351, "grad_norm": 1.6311269998550415, "learning_rate": 5.702310485920794e-06, "loss": 0.2478, "step": 29580 }, { "epoch": 0.5382054966441733, "grad_norm": 1.665647029876709, "learning_rate": 5.699960686471578e-06, "loss": 0.2559, "step": 29590 }, { "epoch": 0.5383873842739956, "grad_norm": 1.8010395765304565, "learning_rate": 5.697610991146406e-06, "loss": 0.2505, "step": 29600 }, { "epoch": 0.5385692719038179, "grad_norm": 1.821061372756958, "learning_rate": 5.6952614007607e-06, "loss": 0.2514, "step": 29610 }, { "epoch": 0.5387511595336402, "grad_norm": 1.7719630002975464, "learning_rate": 5.692911916129851e-06, "loss": 0.2495, "step": 29620 }, { "epoch": 0.5389330471634625, "grad_norm": 1.601934790611267, "learning_rate": 5.69056253806921e-06, "loss": 0.2494, "step": 29630 }, { "epoch": 0.5391149347932848, "grad_norm": 1.6494414806365967, "learning_rate": 5.688213267394094e-06, "loss": 0.2532, "step": 29640 }, { "epoch": 0.5392968224231071, "grad_norm": 1.8598002195358276, "learning_rate": 5.6858641049197815e-06, "loss": 0.2476, "step": 29650 }, { "epoch": 0.5394787100529294, "grad_norm": 1.6447867155075073, "learning_rate": 5.683515051461511e-06, "loss": 0.2554, "step": 29660 }, { "epoch": 0.5396605976827515, "grad_norm": 1.5638680458068848, "learning_rate": 5.681166107834489e-06, "loss": 0.245, "step": 29670 }, { "epoch": 0.5398424853125738, "grad_norm": 1.7238991260528564, "learning_rate": 5.678817274853877e-06, "loss": 0.2527, "step": 29680 }, { "epoch": 0.5400243729423961, "grad_norm": 1.6380161046981812, "learning_rate": 5.676468553334805e-06, "loss": 0.2485, "step": 29690 }, { "epoch": 0.5402062605722184, "grad_norm": 1.5627042055130005, "learning_rate": 5.674119944092357e-06, "loss": 0.2391, "step": 29700 }, { "epoch": 0.5403881482020407, "grad_norm": 1.6788673400878906, "learning_rate": 5.671771447941584e-06, "loss": 0.2475, "step": 29710 }, { "epoch": 0.540570035831863, "grad_norm": 1.8014169931411743, "learning_rate": 5.669423065697496e-06, "loss": 0.2532, "step": 29720 }, { "epoch": 0.5407519234616853, "grad_norm": 1.7036710977554321, "learning_rate": 5.66707479817506e-06, "loss": 0.2482, "step": 29730 }, { "epoch": 0.5409338110915076, "grad_norm": 1.7302627563476562, "learning_rate": 5.6647266461892115e-06, "loss": 0.2441, "step": 29740 }, { "epoch": 0.5411156987213299, "grad_norm": 1.7765907049179077, "learning_rate": 5.662378610554835e-06, "loss": 0.2542, "step": 29750 }, { "epoch": 0.5412975863511522, "grad_norm": 1.542041540145874, "learning_rate": 5.6600306920867844e-06, "loss": 0.24, "step": 29760 }, { "epoch": 0.5414794739809745, "grad_norm": 1.5126895904541016, "learning_rate": 5.657682891599865e-06, "loss": 0.2554, "step": 29770 }, { "epoch": 0.5416613616107968, "grad_norm": 1.4930263757705688, "learning_rate": 5.655335209908849e-06, "loss": 0.2487, "step": 29780 }, { "epoch": 0.5418432492406191, "grad_norm": 1.6440361738204956, "learning_rate": 5.652987647828461e-06, "loss": 0.2436, "step": 29790 }, { "epoch": 0.5420251368704414, "grad_norm": 1.7422982454299927, "learning_rate": 5.650640206173387e-06, "loss": 0.2456, "step": 29800 }, { "epoch": 0.5422070245002637, "grad_norm": 2.0083391666412354, "learning_rate": 5.648292885758268e-06, "loss": 0.254, "step": 29810 }, { "epoch": 0.542388912130086, "grad_norm": 1.5196517705917358, "learning_rate": 5.6459456873977115e-06, "loss": 0.2466, "step": 29820 }, { "epoch": 0.5425707997599083, "grad_norm": 1.6186455488204956, "learning_rate": 5.643598611906272e-06, "loss": 0.2472, "step": 29830 }, { "epoch": 0.5427526873897306, "grad_norm": 1.5902351140975952, "learning_rate": 5.641251660098468e-06, "loss": 0.2485, "step": 29840 }, { "epoch": 0.5429345750195529, "grad_norm": 1.6021696329116821, "learning_rate": 5.6389048327887716e-06, "loss": 0.2418, "step": 29850 }, { "epoch": 0.5431164626493752, "grad_norm": 1.539920687675476, "learning_rate": 5.636558130791613e-06, "loss": 0.2445, "step": 29860 }, { "epoch": 0.5432983502791975, "grad_norm": 1.7229737043380737, "learning_rate": 5.634211554921382e-06, "loss": 0.2521, "step": 29870 }, { "epoch": 0.5434802379090198, "grad_norm": 2.1440601348876953, "learning_rate": 5.631865105992418e-06, "loss": 0.2485, "step": 29880 }, { "epoch": 0.5436621255388421, "grad_norm": 1.7759722471237183, "learning_rate": 5.629518784819024e-06, "loss": 0.2455, "step": 29890 }, { "epoch": 0.5438440131686644, "grad_norm": 1.5447709560394287, "learning_rate": 5.627172592215451e-06, "loss": 0.2444, "step": 29900 }, { "epoch": 0.5440259007984867, "grad_norm": 1.6985243558883667, "learning_rate": 5.624826528995911e-06, "loss": 0.2527, "step": 29910 }, { "epoch": 0.544207788428309, "grad_norm": 1.4345487356185913, "learning_rate": 5.62248059597457e-06, "loss": 0.2468, "step": 29920 }, { "epoch": 0.5443896760581313, "grad_norm": 1.636495590209961, "learning_rate": 5.620134793965546e-06, "loss": 0.2447, "step": 29930 }, { "epoch": 0.5445715636879536, "grad_norm": 1.4587756395339966, "learning_rate": 5.617789123782916e-06, "loss": 0.2454, "step": 29940 }, { "epoch": 0.5447534513177759, "grad_norm": 1.7120718955993652, "learning_rate": 5.615443586240707e-06, "loss": 0.2488, "step": 29950 }, { "epoch": 0.5449353389475982, "grad_norm": 1.7309715747833252, "learning_rate": 5.613098182152904e-06, "loss": 0.248, "step": 29960 }, { "epoch": 0.5451172265774205, "grad_norm": 1.7647945880889893, "learning_rate": 5.610752912333441e-06, "loss": 0.2463, "step": 29970 }, { "epoch": 0.5452991142072428, "grad_norm": 1.7265647649765015, "learning_rate": 5.608407777596208e-06, "loss": 0.2483, "step": 29980 }, { "epoch": 0.545481001837065, "grad_norm": 1.5731103420257568, "learning_rate": 5.60606277875505e-06, "loss": 0.2426, "step": 29990 }, { "epoch": 0.5456628894668873, "grad_norm": 1.5492013692855835, "learning_rate": 5.6037179166237606e-06, "loss": 0.2384, "step": 30000 }, { "epoch": 0.5458447770967096, "grad_norm": 1.555044412612915, "learning_rate": 5.6013731920160915e-06, "loss": 0.247, "step": 30010 }, { "epoch": 0.5460266647265319, "grad_norm": 1.559145212173462, "learning_rate": 5.59902860574574e-06, "loss": 0.246, "step": 30020 }, { "epoch": 0.5462085523563542, "grad_norm": 1.711190104484558, "learning_rate": 5.596684158626361e-06, "loss": 0.2506, "step": 30030 }, { "epoch": 0.5463904399861765, "grad_norm": 1.7757201194763184, "learning_rate": 5.5943398514715566e-06, "loss": 0.2495, "step": 30040 }, { "epoch": 0.5465723276159988, "grad_norm": 1.6520837545394897, "learning_rate": 5.591995685094885e-06, "loss": 0.2485, "step": 30050 }, { "epoch": 0.5467542152458211, "grad_norm": 1.5731618404388428, "learning_rate": 5.589651660309853e-06, "loss": 0.2446, "step": 30060 }, { "epoch": 0.5469361028756434, "grad_norm": 1.51480233669281, "learning_rate": 5.587307777929918e-06, "loss": 0.2562, "step": 30070 }, { "epoch": 0.5471179905054657, "grad_norm": 1.927634596824646, "learning_rate": 5.584964038768486e-06, "loss": 0.255, "step": 30080 }, { "epoch": 0.547299878135288, "grad_norm": 1.4059242010116577, "learning_rate": 5.58262044363892e-06, "loss": 0.2486, "step": 30090 }, { "epoch": 0.5474817657651103, "grad_norm": 1.6470025777816772, "learning_rate": 5.580276993354525e-06, "loss": 0.2449, "step": 30100 }, { "epoch": 0.5476636533949326, "grad_norm": 1.6698163747787476, "learning_rate": 5.5779336887285625e-06, "loss": 0.2402, "step": 30110 }, { "epoch": 0.5478455410247549, "grad_norm": 1.7616002559661865, "learning_rate": 5.57559053057424e-06, "loss": 0.2466, "step": 30120 }, { "epoch": 0.5480274286545772, "grad_norm": 1.6377475261688232, "learning_rate": 5.573247519704712e-06, "loss": 0.2435, "step": 30130 }, { "epoch": 0.5482093162843995, "grad_norm": 1.7451447248458862, "learning_rate": 5.5709046569330875e-06, "loss": 0.2463, "step": 30140 }, { "epoch": 0.5483912039142218, "grad_norm": 1.6533030271530151, "learning_rate": 5.568561943072417e-06, "loss": 0.2514, "step": 30150 }, { "epoch": 0.5485730915440441, "grad_norm": 1.5492972135543823, "learning_rate": 5.566219378935709e-06, "loss": 0.247, "step": 30160 }, { "epoch": 0.5487549791738664, "grad_norm": 1.8269610404968262, "learning_rate": 5.56387696533591e-06, "loss": 0.2501, "step": 30170 }, { "epoch": 0.5489368668036887, "grad_norm": 1.7645307779312134, "learning_rate": 5.56153470308592e-06, "loss": 0.2465, "step": 30180 }, { "epoch": 0.549118754433511, "grad_norm": 1.7743544578552246, "learning_rate": 5.559192592998587e-06, "loss": 0.2492, "step": 30190 }, { "epoch": 0.5493006420633333, "grad_norm": 1.5146411657333374, "learning_rate": 5.556850635886701e-06, "loss": 0.2433, "step": 30200 }, { "epoch": 0.5494825296931556, "grad_norm": 1.798808217048645, "learning_rate": 5.554508832563006e-06, "loss": 0.2464, "step": 30210 }, { "epoch": 0.5496644173229779, "grad_norm": 1.716120719909668, "learning_rate": 5.5521671838401856e-06, "loss": 0.2541, "step": 30220 }, { "epoch": 0.5498463049528002, "grad_norm": 1.6600888967514038, "learning_rate": 5.549825690530876e-06, "loss": 0.2505, "step": 30230 }, { "epoch": 0.5500281925826225, "grad_norm": 1.6197490692138672, "learning_rate": 5.5474843534476545e-06, "loss": 0.2479, "step": 30240 }, { "epoch": 0.5502100802124448, "grad_norm": 1.519684910774231, "learning_rate": 5.545143173403046e-06, "loss": 0.2453, "step": 30250 }, { "epoch": 0.5503919678422671, "grad_norm": 1.649400234222412, "learning_rate": 5.542802151209523e-06, "loss": 0.2421, "step": 30260 }, { "epoch": 0.5505738554720894, "grad_norm": 1.7612255811691284, "learning_rate": 5.5404612876795e-06, "loss": 0.2487, "step": 30270 }, { "epoch": 0.5507557431019117, "grad_norm": 1.904625415802002, "learning_rate": 5.5381205836253375e-06, "loss": 0.2482, "step": 30280 }, { "epoch": 0.550937630731734, "grad_norm": 1.6640859842300415, "learning_rate": 5.535780039859341e-06, "loss": 0.2447, "step": 30290 }, { "epoch": 0.5511195183615563, "grad_norm": 1.5188733339309692, "learning_rate": 5.533439657193762e-06, "loss": 0.2454, "step": 30300 }, { "epoch": 0.5513014059913786, "grad_norm": 1.675742268562317, "learning_rate": 5.531099436440793e-06, "loss": 0.2432, "step": 30310 }, { "epoch": 0.5514832936212009, "grad_norm": 1.6823031902313232, "learning_rate": 5.52875937841257e-06, "loss": 0.2517, "step": 30320 }, { "epoch": 0.5516651812510232, "grad_norm": 1.655809760093689, "learning_rate": 5.526419483921177e-06, "loss": 0.2527, "step": 30330 }, { "epoch": 0.5518470688808454, "grad_norm": 1.6158369779586792, "learning_rate": 5.524079753778639e-06, "loss": 0.2406, "step": 30340 }, { "epoch": 0.5520289565106677, "grad_norm": 5.763541221618652, "learning_rate": 5.521740188796921e-06, "loss": 0.2508, "step": 30350 }, { "epoch": 0.55221084414049, "grad_norm": 1.7298872470855713, "learning_rate": 5.519400789787936e-06, "loss": 0.2432, "step": 30360 }, { "epoch": 0.5523927317703123, "grad_norm": 1.6519650220870972, "learning_rate": 5.517061557563532e-06, "loss": 0.2538, "step": 30370 }, { "epoch": 0.5525746194001346, "grad_norm": 1.758932113647461, "learning_rate": 5.514722492935508e-06, "loss": 0.2526, "step": 30380 }, { "epoch": 0.5527565070299569, "grad_norm": 1.6997936964035034, "learning_rate": 5.512383596715599e-06, "loss": 0.2459, "step": 30390 }, { "epoch": 0.5529383946597792, "grad_norm": 1.7433927059173584, "learning_rate": 5.510044869715483e-06, "loss": 0.2494, "step": 30400 }, { "epoch": 0.5531202822896015, "grad_norm": 2.035665273666382, "learning_rate": 5.507706312746779e-06, "loss": 0.2518, "step": 30410 }, { "epoch": 0.5533021699194238, "grad_norm": 1.7616043090820312, "learning_rate": 5.505367926621047e-06, "loss": 0.2518, "step": 30420 }, { "epoch": 0.5534840575492461, "grad_norm": 1.770274043083191, "learning_rate": 5.50302971214979e-06, "loss": 0.2436, "step": 30430 }, { "epoch": 0.5536659451790684, "grad_norm": 1.9598042964935303, "learning_rate": 5.500691670144448e-06, "loss": 0.2548, "step": 30440 }, { "epoch": 0.5538478328088907, "grad_norm": 1.435105562210083, "learning_rate": 5.498353801416402e-06, "loss": 0.2458, "step": 30450 }, { "epoch": 0.554029720438713, "grad_norm": 2.0694050788879395, "learning_rate": 5.496016106776975e-06, "loss": 0.2445, "step": 30460 }, { "epoch": 0.5542116080685353, "grad_norm": 1.606939673423767, "learning_rate": 5.493678587037425e-06, "loss": 0.2479, "step": 30470 }, { "epoch": 0.5543934956983576, "grad_norm": 1.5753785371780396, "learning_rate": 5.4913412430089574e-06, "loss": 0.2561, "step": 30480 }, { "epoch": 0.5545753833281799, "grad_norm": 1.5447063446044922, "learning_rate": 5.489004075502709e-06, "loss": 0.2438, "step": 30490 }, { "epoch": 0.5547572709580021, "grad_norm": 1.6582391262054443, "learning_rate": 5.486667085329755e-06, "loss": 0.2461, "step": 30500 }, { "epoch": 0.5549391585878244, "grad_norm": 1.8138723373413086, "learning_rate": 5.484330273301119e-06, "loss": 0.2492, "step": 30510 }, { "epoch": 0.5551210462176467, "grad_norm": 1.8980097770690918, "learning_rate": 5.481993640227748e-06, "loss": 0.2484, "step": 30520 }, { "epoch": 0.555302933847469, "grad_norm": 1.4801392555236816, "learning_rate": 5.47965718692054e-06, "loss": 0.2456, "step": 30530 }, { "epoch": 0.5554848214772913, "grad_norm": 1.716778039932251, "learning_rate": 5.477320914190322e-06, "loss": 0.2426, "step": 30540 }, { "epoch": 0.5556667091071136, "grad_norm": 1.8449753522872925, "learning_rate": 5.474984822847864e-06, "loss": 0.2524, "step": 30550 }, { "epoch": 0.5558485967369359, "grad_norm": 1.576230764389038, "learning_rate": 5.472648913703869e-06, "loss": 0.2436, "step": 30560 }, { "epoch": 0.5560304843667582, "grad_norm": 1.7377712726593018, "learning_rate": 5.4703131875689804e-06, "loss": 0.2458, "step": 30570 }, { "epoch": 0.5562123719965805, "grad_norm": 1.565962791442871, "learning_rate": 5.4679776452537746e-06, "loss": 0.2482, "step": 30580 }, { "epoch": 0.5563942596264028, "grad_norm": 1.4639439582824707, "learning_rate": 5.465642287568764e-06, "loss": 0.2461, "step": 30590 }, { "epoch": 0.5565761472562251, "grad_norm": 1.6844624280929565, "learning_rate": 5.463307115324402e-06, "loss": 0.2511, "step": 30600 }, { "epoch": 0.5567580348860474, "grad_norm": 2.023343086242676, "learning_rate": 5.460972129331075e-06, "loss": 0.2527, "step": 30610 }, { "epoch": 0.5569399225158697, "grad_norm": 1.6157524585723877, "learning_rate": 5.4586373303991e-06, "loss": 0.2401, "step": 30620 }, { "epoch": 0.557121810145692, "grad_norm": 1.4479868412017822, "learning_rate": 5.456302719338736e-06, "loss": 0.2424, "step": 30630 }, { "epoch": 0.5573036977755143, "grad_norm": 1.4573001861572266, "learning_rate": 5.453968296960174e-06, "loss": 0.2441, "step": 30640 }, { "epoch": 0.5574855854053365, "grad_norm": 1.7138762474060059, "learning_rate": 5.451634064073538e-06, "loss": 0.2422, "step": 30650 }, { "epoch": 0.5576674730351588, "grad_norm": 1.6859999895095825, "learning_rate": 5.449300021488888e-06, "loss": 0.247, "step": 30660 }, { "epoch": 0.5578493606649811, "grad_norm": 1.831881046295166, "learning_rate": 5.446966170016218e-06, "loss": 0.2509, "step": 30670 }, { "epoch": 0.5580312482948034, "grad_norm": 1.706407070159912, "learning_rate": 5.444632510465456e-06, "loss": 0.2472, "step": 30680 }, { "epoch": 0.5582131359246257, "grad_norm": 1.5402520895004272, "learning_rate": 5.442299043646462e-06, "loss": 0.2401, "step": 30690 }, { "epoch": 0.558395023554448, "grad_norm": 1.435646414756775, "learning_rate": 5.439965770369031e-06, "loss": 0.2478, "step": 30700 }, { "epoch": 0.5585769111842703, "grad_norm": 1.8345322608947754, "learning_rate": 5.437632691442889e-06, "loss": 0.2395, "step": 30710 }, { "epoch": 0.5587587988140926, "grad_norm": 1.492327094078064, "learning_rate": 5.435299807677693e-06, "loss": 0.243, "step": 30720 }, { "epoch": 0.5589406864439149, "grad_norm": 1.8519890308380127, "learning_rate": 5.432967119883038e-06, "loss": 0.2448, "step": 30730 }, { "epoch": 0.5591225740737372, "grad_norm": 1.8619674444198608, "learning_rate": 5.430634628868445e-06, "loss": 0.247, "step": 30740 }, { "epoch": 0.5593044617035595, "grad_norm": 1.592747688293457, "learning_rate": 5.428302335443372e-06, "loss": 0.2497, "step": 30750 }, { "epoch": 0.5594863493333818, "grad_norm": 1.5773985385894775, "learning_rate": 5.425970240417202e-06, "loss": 0.2417, "step": 30760 }, { "epoch": 0.5596682369632041, "grad_norm": 1.778259515762329, "learning_rate": 5.423638344599256e-06, "loss": 0.2454, "step": 30770 }, { "epoch": 0.5598501245930264, "grad_norm": 1.4957635402679443, "learning_rate": 5.4213066487987805e-06, "loss": 0.2501, "step": 30780 }, { "epoch": 0.5600320122228487, "grad_norm": 1.7415282726287842, "learning_rate": 5.418975153824957e-06, "loss": 0.2518, "step": 30790 }, { "epoch": 0.560213899852671, "grad_norm": 1.5715121030807495, "learning_rate": 5.416643860486892e-06, "loss": 0.2442, "step": 30800 }, { "epoch": 0.5603957874824933, "grad_norm": 1.4887337684631348, "learning_rate": 5.414312769593628e-06, "loss": 0.2424, "step": 30810 }, { "epoch": 0.5605776751123156, "grad_norm": 1.8489681482315063, "learning_rate": 5.411981881954133e-06, "loss": 0.2454, "step": 30820 }, { "epoch": 0.5607595627421379, "grad_norm": 1.6287978887557983, "learning_rate": 5.409651198377307e-06, "loss": 0.2508, "step": 30830 }, { "epoch": 0.5609414503719602, "grad_norm": 1.6293715238571167, "learning_rate": 5.407320719671976e-06, "loss": 0.2537, "step": 30840 }, { "epoch": 0.5611233380017825, "grad_norm": 1.5539792776107788, "learning_rate": 5.404990446646897e-06, "loss": 0.2383, "step": 30850 }, { "epoch": 0.5613052256316048, "grad_norm": 2.0170559883117676, "learning_rate": 5.402660380110757e-06, "loss": 0.2446, "step": 30860 }, { "epoch": 0.5614871132614271, "grad_norm": 2.1046934127807617, "learning_rate": 5.400330520872168e-06, "loss": 0.2575, "step": 30870 }, { "epoch": 0.5616690008912494, "grad_norm": 1.4719083309173584, "learning_rate": 5.398000869739674e-06, "loss": 0.2472, "step": 30880 }, { "epoch": 0.5618508885210717, "grad_norm": 1.7171549797058105, "learning_rate": 5.395671427521745e-06, "loss": 0.2486, "step": 30890 }, { "epoch": 0.562032776150894, "grad_norm": 1.7035393714904785, "learning_rate": 5.393342195026776e-06, "loss": 0.24, "step": 30900 }, { "epoch": 0.5622146637807163, "grad_norm": 1.6243705749511719, "learning_rate": 5.391013173063093e-06, "loss": 0.2463, "step": 30910 }, { "epoch": 0.5623965514105386, "grad_norm": 1.5602796077728271, "learning_rate": 5.3886843624389464e-06, "loss": 0.2441, "step": 30920 }, { "epoch": 0.5625784390403609, "grad_norm": 1.7890920639038086, "learning_rate": 5.386355763962515e-06, "loss": 0.2465, "step": 30930 }, { "epoch": 0.5627603266701832, "grad_norm": 1.5750818252563477, "learning_rate": 5.384027378441902e-06, "loss": 0.2445, "step": 30940 }, { "epoch": 0.5629422143000055, "grad_norm": 1.5181344747543335, "learning_rate": 5.381699206685139e-06, "loss": 0.243, "step": 30950 }, { "epoch": 0.5631241019298278, "grad_norm": 1.7626769542694092, "learning_rate": 5.379371249500181e-06, "loss": 0.2449, "step": 30960 }, { "epoch": 0.56330598955965, "grad_norm": 1.5077849626541138, "learning_rate": 5.3770435076949125e-06, "loss": 0.2372, "step": 30970 }, { "epoch": 0.5634878771894724, "grad_norm": 1.8718940019607544, "learning_rate": 5.374715982077139e-06, "loss": 0.2481, "step": 30980 }, { "epoch": 0.5636697648192946, "grad_norm": 1.5056146383285522, "learning_rate": 5.37238867345459e-06, "loss": 0.2483, "step": 30990 }, { "epoch": 0.563851652449117, "grad_norm": 1.8704304695129395, "learning_rate": 5.370061582634927e-06, "loss": 0.2473, "step": 31000 }, { "epoch": 0.5640335400789392, "grad_norm": 1.7574487924575806, "learning_rate": 5.3677347104257274e-06, "loss": 0.2472, "step": 31010 }, { "epoch": 0.5642154277087615, "grad_norm": 1.4230320453643799, "learning_rate": 5.365408057634499e-06, "loss": 0.2478, "step": 31020 }, { "epoch": 0.5643973153385838, "grad_norm": 1.8647326231002808, "learning_rate": 5.363081625068669e-06, "loss": 0.2467, "step": 31030 }, { "epoch": 0.5645792029684061, "grad_norm": 1.6629303693771362, "learning_rate": 5.360755413535591e-06, "loss": 0.2479, "step": 31040 }, { "epoch": 0.5647610905982284, "grad_norm": 1.7415740489959717, "learning_rate": 5.358429423842542e-06, "loss": 0.2509, "step": 31050 }, { "epoch": 0.5649429782280507, "grad_norm": 1.3605278730392456, "learning_rate": 5.356103656796719e-06, "loss": 0.2447, "step": 31060 }, { "epoch": 0.565124865857873, "grad_norm": 1.5983624458312988, "learning_rate": 5.353778113205245e-06, "loss": 0.2416, "step": 31070 }, { "epoch": 0.5653067534876953, "grad_norm": 1.6787457466125488, "learning_rate": 5.351452793875165e-06, "loss": 0.2532, "step": 31080 }, { "epoch": 0.5654886411175176, "grad_norm": 1.7930485010147095, "learning_rate": 5.349127699613444e-06, "loss": 0.2468, "step": 31090 }, { "epoch": 0.5656705287473399, "grad_norm": 1.5498206615447998, "learning_rate": 5.346802831226972e-06, "loss": 0.2492, "step": 31100 }, { "epoch": 0.5658524163771622, "grad_norm": 1.634857177734375, "learning_rate": 5.3444781895225564e-06, "loss": 0.2383, "step": 31110 }, { "epoch": 0.5660343040069845, "grad_norm": 1.525566816329956, "learning_rate": 5.34215377530693e-06, "loss": 0.2513, "step": 31120 }, { "epoch": 0.5662161916368068, "grad_norm": 2.0381224155426025, "learning_rate": 5.339829589386745e-06, "loss": 0.2486, "step": 31130 }, { "epoch": 0.5663980792666291, "grad_norm": 1.6233011484146118, "learning_rate": 5.337505632568574e-06, "loss": 0.2451, "step": 31140 }, { "epoch": 0.5665799668964514, "grad_norm": 1.8278348445892334, "learning_rate": 5.3351819056589096e-06, "loss": 0.2474, "step": 31150 }, { "epoch": 0.5667618545262737, "grad_norm": 1.5664994716644287, "learning_rate": 5.332858409464168e-06, "loss": 0.242, "step": 31160 }, { "epoch": 0.566943742156096, "grad_norm": 1.620962381362915, "learning_rate": 5.330535144790682e-06, "loss": 0.2463, "step": 31170 }, { "epoch": 0.5671256297859183, "grad_norm": 1.6366369724273682, "learning_rate": 5.328212112444705e-06, "loss": 0.2442, "step": 31180 }, { "epoch": 0.5673075174157406, "grad_norm": 1.4795818328857422, "learning_rate": 5.325889313232409e-06, "loss": 0.2442, "step": 31190 }, { "epoch": 0.5674894050455629, "grad_norm": 1.607767105102539, "learning_rate": 5.323566747959887e-06, "loss": 0.2516, "step": 31200 }, { "epoch": 0.5676712926753852, "grad_norm": 1.617714524269104, "learning_rate": 5.321244417433149e-06, "loss": 0.242, "step": 31210 }, { "epoch": 0.5678531803052075, "grad_norm": 1.7752474546432495, "learning_rate": 5.318922322458124e-06, "loss": 0.2492, "step": 31220 }, { "epoch": 0.5680350679350298, "grad_norm": 1.485776662826538, "learning_rate": 5.31660046384066e-06, "loss": 0.2433, "step": 31230 }, { "epoch": 0.5682169555648521, "grad_norm": 1.6531012058258057, "learning_rate": 5.314278842386521e-06, "loss": 0.2445, "step": 31240 }, { "epoch": 0.5683988431946744, "grad_norm": 1.5716959238052368, "learning_rate": 5.311957458901392e-06, "loss": 0.2431, "step": 31250 }, { "epoch": 0.5685807308244967, "grad_norm": 1.6719917058944702, "learning_rate": 5.3096363141908714e-06, "loss": 0.2474, "step": 31260 }, { "epoch": 0.568762618454319, "grad_norm": 1.5274966955184937, "learning_rate": 5.307315409060481e-06, "loss": 0.2502, "step": 31270 }, { "epoch": 0.5689445060841413, "grad_norm": 1.4356677532196045, "learning_rate": 5.304994744315649e-06, "loss": 0.2468, "step": 31280 }, { "epoch": 0.5691263937139636, "grad_norm": 1.650960087776184, "learning_rate": 5.302674320761732e-06, "loss": 0.2503, "step": 31290 }, { "epoch": 0.5693082813437859, "grad_norm": 1.641411542892456, "learning_rate": 5.300354139203996e-06, "loss": 0.2429, "step": 31300 }, { "epoch": 0.5694901689736082, "grad_norm": 1.6759140491485596, "learning_rate": 5.298034200447622e-06, "loss": 0.2449, "step": 31310 }, { "epoch": 0.5696720566034305, "grad_norm": 1.842369794845581, "learning_rate": 5.295714505297712e-06, "loss": 0.2445, "step": 31320 }, { "epoch": 0.5698539442332528, "grad_norm": 1.6736230850219727, "learning_rate": 5.293395054559278e-06, "loss": 0.246, "step": 31330 }, { "epoch": 0.5700358318630749, "grad_norm": 1.7178226709365845, "learning_rate": 5.291075849037254e-06, "loss": 0.2464, "step": 31340 }, { "epoch": 0.5702177194928972, "grad_norm": 1.5738674402236938, "learning_rate": 5.288756889536479e-06, "loss": 0.2443, "step": 31350 }, { "epoch": 0.5703996071227195, "grad_norm": 1.5932915210723877, "learning_rate": 5.286438176861719e-06, "loss": 0.2446, "step": 31360 }, { "epoch": 0.5705814947525418, "grad_norm": 1.6397435665130615, "learning_rate": 5.2841197118176426e-06, "loss": 0.2358, "step": 31370 }, { "epoch": 0.5707633823823641, "grad_norm": 1.7275387048721313, "learning_rate": 5.2818014952088385e-06, "loss": 0.2487, "step": 31380 }, { "epoch": 0.5709452700121864, "grad_norm": 1.6201766729354858, "learning_rate": 5.279483527839808e-06, "loss": 0.2484, "step": 31390 }, { "epoch": 0.5711271576420087, "grad_norm": 1.6508022546768188, "learning_rate": 5.277165810514968e-06, "loss": 0.2416, "step": 31400 }, { "epoch": 0.571309045271831, "grad_norm": 1.5905269384384155, "learning_rate": 5.274848344038646e-06, "loss": 0.2394, "step": 31410 }, { "epoch": 0.5714909329016533, "grad_norm": 2.018322706222534, "learning_rate": 5.272531129215079e-06, "loss": 0.2471, "step": 31420 }, { "epoch": 0.5716728205314756, "grad_norm": 1.5234766006469727, "learning_rate": 5.270214166848429e-06, "loss": 0.2399, "step": 31430 }, { "epoch": 0.5718547081612979, "grad_norm": 1.507820963859558, "learning_rate": 5.267897457742755e-06, "loss": 0.2445, "step": 31440 }, { "epoch": 0.5720365957911202, "grad_norm": 1.4596478939056396, "learning_rate": 5.265581002702041e-06, "loss": 0.2498, "step": 31450 }, { "epoch": 0.5722184834209425, "grad_norm": 1.8628661632537842, "learning_rate": 5.263264802530172e-06, "loss": 0.2441, "step": 31460 }, { "epoch": 0.5724003710507648, "grad_norm": 1.5080962181091309, "learning_rate": 5.260948858030954e-06, "loss": 0.2458, "step": 31470 }, { "epoch": 0.5725822586805871, "grad_norm": 1.4612164497375488, "learning_rate": 5.2586331700080966e-06, "loss": 0.2413, "step": 31480 }, { "epoch": 0.5727641463104094, "grad_norm": 1.5048826932907104, "learning_rate": 5.256317739265226e-06, "loss": 0.2482, "step": 31490 }, { "epoch": 0.5729460339402317, "grad_norm": 1.5051169395446777, "learning_rate": 5.254002566605878e-06, "loss": 0.2461, "step": 31500 }, { "epoch": 0.573127921570054, "grad_norm": 1.4559468030929565, "learning_rate": 5.251687652833494e-06, "loss": 0.2415, "step": 31510 }, { "epoch": 0.5733098091998763, "grad_norm": 1.5632097721099854, "learning_rate": 5.249372998751432e-06, "loss": 0.2434, "step": 31520 }, { "epoch": 0.5734916968296986, "grad_norm": 1.8371108770370483, "learning_rate": 5.247058605162955e-06, "loss": 0.2463, "step": 31530 }, { "epoch": 0.5736735844595209, "grad_norm": 2.2204020023345947, "learning_rate": 5.2447444728712415e-06, "loss": 0.2535, "step": 31540 }, { "epoch": 0.5738554720893432, "grad_norm": 1.6502611637115479, "learning_rate": 5.242430602679374e-06, "loss": 0.2453, "step": 31550 }, { "epoch": 0.5740373597191655, "grad_norm": 1.7881746292114258, "learning_rate": 5.240116995390343e-06, "loss": 0.2473, "step": 31560 }, { "epoch": 0.5742192473489878, "grad_norm": 1.7012512683868408, "learning_rate": 5.237803651807054e-06, "loss": 0.252, "step": 31570 }, { "epoch": 0.5744011349788101, "grad_norm": 1.7657350301742554, "learning_rate": 5.235490572732316e-06, "loss": 0.2519, "step": 31580 }, { "epoch": 0.5745830226086324, "grad_norm": 1.6935443878173828, "learning_rate": 5.2331777589688474e-06, "loss": 0.2471, "step": 31590 }, { "epoch": 0.5747649102384547, "grad_norm": 1.444085955619812, "learning_rate": 5.230865211319274e-06, "loss": 0.2487, "step": 31600 }, { "epoch": 0.574946797868277, "grad_norm": 1.833856225013733, "learning_rate": 5.22855293058613e-06, "loss": 0.2465, "step": 31610 }, { "epoch": 0.5751286854980993, "grad_norm": 1.5020118951797485, "learning_rate": 5.226240917571861e-06, "loss": 0.2465, "step": 31620 }, { "epoch": 0.5753105731279216, "grad_norm": 1.7880957126617432, "learning_rate": 5.2239291730788096e-06, "loss": 0.2492, "step": 31630 }, { "epoch": 0.5754924607577439, "grad_norm": 1.5525630712509155, "learning_rate": 5.221617697909236e-06, "loss": 0.2436, "step": 31640 }, { "epoch": 0.5756743483875661, "grad_norm": 1.5815961360931396, "learning_rate": 5.2193064928653e-06, "loss": 0.2448, "step": 31650 }, { "epoch": 0.5758562360173884, "grad_norm": 1.6669998168945312, "learning_rate": 5.21699555874907e-06, "loss": 0.2417, "step": 31660 }, { "epoch": 0.5760381236472107, "grad_norm": 1.8145921230316162, "learning_rate": 5.21468489636252e-06, "loss": 0.2522, "step": 31670 }, { "epoch": 0.576220011277033, "grad_norm": 1.5972278118133545, "learning_rate": 5.21237450650753e-06, "loss": 0.2428, "step": 31680 }, { "epoch": 0.5764018989068553, "grad_norm": 1.6267584562301636, "learning_rate": 5.210064389985887e-06, "loss": 0.2481, "step": 31690 }, { "epoch": 0.5765837865366776, "grad_norm": 1.7393306493759155, "learning_rate": 5.20775454759928e-06, "loss": 0.2429, "step": 31700 }, { "epoch": 0.5767656741664999, "grad_norm": 1.6464895009994507, "learning_rate": 5.205444980149303e-06, "loss": 0.2422, "step": 31710 }, { "epoch": 0.5769475617963222, "grad_norm": 1.486926555633545, "learning_rate": 5.2031356884374596e-06, "loss": 0.2387, "step": 31720 }, { "epoch": 0.5771294494261445, "grad_norm": 1.5699071884155273, "learning_rate": 5.200826673265152e-06, "loss": 0.2457, "step": 31730 }, { "epoch": 0.5773113370559668, "grad_norm": 1.8981149196624756, "learning_rate": 5.198517935433687e-06, "loss": 0.2514, "step": 31740 }, { "epoch": 0.5774932246857891, "grad_norm": 1.657317876815796, "learning_rate": 5.1962094757442795e-06, "loss": 0.2437, "step": 31750 }, { "epoch": 0.5776751123156114, "grad_norm": 1.4921995401382446, "learning_rate": 5.193901294998044e-06, "loss": 0.2407, "step": 31760 }, { "epoch": 0.5778569999454337, "grad_norm": 1.9248015880584717, "learning_rate": 5.1915933939959986e-06, "loss": 0.2449, "step": 31770 }, { "epoch": 0.578038887575256, "grad_norm": 1.5044682025909424, "learning_rate": 5.189285773539065e-06, "loss": 0.247, "step": 31780 }, { "epoch": 0.5782207752050783, "grad_norm": 1.6435177326202393, "learning_rate": 5.1869784344280675e-06, "loss": 0.2421, "step": 31790 }, { "epoch": 0.5784026628349006, "grad_norm": 1.807004451751709, "learning_rate": 5.184671377463732e-06, "loss": 0.2502, "step": 31800 }, { "epoch": 0.5785845504647229, "grad_norm": 1.815585970878601, "learning_rate": 5.1823646034466915e-06, "loss": 0.2377, "step": 31810 }, { "epoch": 0.5787664380945452, "grad_norm": 1.7584526538848877, "learning_rate": 5.180058113177471e-06, "loss": 0.2447, "step": 31820 }, { "epoch": 0.5789483257243675, "grad_norm": 1.8666753768920898, "learning_rate": 5.177751907456505e-06, "loss": 0.2524, "step": 31830 }, { "epoch": 0.5791302133541898, "grad_norm": 1.6685612201690674, "learning_rate": 5.175445987084128e-06, "loss": 0.2441, "step": 31840 }, { "epoch": 0.5793121009840121, "grad_norm": 1.8829401731491089, "learning_rate": 5.1731403528605706e-06, "loss": 0.2523, "step": 31850 }, { "epoch": 0.5794939886138344, "grad_norm": 1.5366345643997192, "learning_rate": 5.170835005585973e-06, "loss": 0.2418, "step": 31860 }, { "epoch": 0.5796758762436567, "grad_norm": 1.6691688299179077, "learning_rate": 5.168529946060365e-06, "loss": 0.2455, "step": 31870 }, { "epoch": 0.579857763873479, "grad_norm": 1.8357455730438232, "learning_rate": 5.166225175083687e-06, "loss": 0.2526, "step": 31880 }, { "epoch": 0.5800396515033013, "grad_norm": 1.805487036705017, "learning_rate": 5.163920693455772e-06, "loss": 0.2431, "step": 31890 }, { "epoch": 0.5802215391331236, "grad_norm": 2.0687975883483887, "learning_rate": 5.161616501976354e-06, "loss": 0.2472, "step": 31900 }, { "epoch": 0.5804034267629459, "grad_norm": 1.800674319267273, "learning_rate": 5.1593126014450686e-06, "loss": 0.2418, "step": 31910 }, { "epoch": 0.5805853143927682, "grad_norm": 1.519777774810791, "learning_rate": 5.15700899266145e-06, "loss": 0.2439, "step": 31920 }, { "epoch": 0.5807672020225905, "grad_norm": 1.4305551052093506, "learning_rate": 5.15470567642493e-06, "loss": 0.2404, "step": 31930 }, { "epoch": 0.5809490896524128, "grad_norm": 1.6010695695877075, "learning_rate": 5.152402653534838e-06, "loss": 0.2442, "step": 31940 }, { "epoch": 0.5811309772822351, "grad_norm": 1.6490657329559326, "learning_rate": 5.150099924790404e-06, "loss": 0.2481, "step": 31950 }, { "epoch": 0.5813128649120574, "grad_norm": 1.5831280946731567, "learning_rate": 5.147797490990753e-06, "loss": 0.2419, "step": 31960 }, { "epoch": 0.5814947525418797, "grad_norm": 1.5484228134155273, "learning_rate": 5.14549535293491e-06, "loss": 0.248, "step": 31970 }, { "epoch": 0.581676640171702, "grad_norm": 1.7961475849151611, "learning_rate": 5.143193511421798e-06, "loss": 0.2471, "step": 31980 }, { "epoch": 0.5818585278015243, "grad_norm": 1.5816153287887573, "learning_rate": 5.140891967250236e-06, "loss": 0.2417, "step": 31990 }, { "epoch": 0.5820404154313465, "grad_norm": 1.6106750965118408, "learning_rate": 5.138590721218937e-06, "loss": 0.2378, "step": 32000 }, { "epoch": 0.5822223030611688, "grad_norm": 1.7154803276062012, "learning_rate": 5.136289774126515e-06, "loss": 0.2496, "step": 32010 }, { "epoch": 0.5824041906909911, "grad_norm": 1.6120173931121826, "learning_rate": 5.133989126771477e-06, "loss": 0.2456, "step": 32020 }, { "epoch": 0.5825860783208134, "grad_norm": 1.525761365890503, "learning_rate": 5.131688779952231e-06, "loss": 0.2396, "step": 32030 }, { "epoch": 0.5827679659506357, "grad_norm": 1.5098758935928345, "learning_rate": 5.129388734467073e-06, "loss": 0.2473, "step": 32040 }, { "epoch": 0.582949853580458, "grad_norm": 1.5894217491149902, "learning_rate": 5.1270889911141995e-06, "loss": 0.2545, "step": 32050 }, { "epoch": 0.5831317412102803, "grad_norm": 1.5799627304077148, "learning_rate": 5.124789550691702e-06, "loss": 0.2367, "step": 32060 }, { "epoch": 0.5833136288401026, "grad_norm": 1.6591579914093018, "learning_rate": 5.122490413997566e-06, "loss": 0.2476, "step": 32070 }, { "epoch": 0.5834955164699249, "grad_norm": 1.96746027469635, "learning_rate": 5.1201915818296726e-06, "loss": 0.2379, "step": 32080 }, { "epoch": 0.5836774040997472, "grad_norm": 1.8087010383605957, "learning_rate": 5.117893054985793e-06, "loss": 0.2459, "step": 32090 }, { "epoch": 0.5838592917295695, "grad_norm": 1.7249137163162231, "learning_rate": 5.115594834263598e-06, "loss": 0.2413, "step": 32100 }, { "epoch": 0.5840411793593918, "grad_norm": 1.673807144165039, "learning_rate": 5.11329692046065e-06, "loss": 0.2484, "step": 32110 }, { "epoch": 0.5842230669892141, "grad_norm": 1.5768390893936157, "learning_rate": 5.110999314374403e-06, "loss": 0.2373, "step": 32120 }, { "epoch": 0.5844049546190364, "grad_norm": 1.7125407457351685, "learning_rate": 5.108702016802208e-06, "loss": 0.2418, "step": 32130 }, { "epoch": 0.5845868422488587, "grad_norm": 1.8713370561599731, "learning_rate": 5.106405028541306e-06, "loss": 0.2492, "step": 32140 }, { "epoch": 0.584768729878681, "grad_norm": 1.8414827585220337, "learning_rate": 5.104108350388832e-06, "loss": 0.2419, "step": 32150 }, { "epoch": 0.5849506175085033, "grad_norm": 1.6015037298202515, "learning_rate": 5.101811983141812e-06, "loss": 0.2467, "step": 32160 }, { "epoch": 0.5851325051383255, "grad_norm": 1.4999428987503052, "learning_rate": 5.099515927597165e-06, "loss": 0.2459, "step": 32170 }, { "epoch": 0.5853143927681478, "grad_norm": 1.646868109703064, "learning_rate": 5.0972201845517035e-06, "loss": 0.2487, "step": 32180 }, { "epoch": 0.5854962803979701, "grad_norm": 1.551987886428833, "learning_rate": 5.094924754802127e-06, "loss": 0.244, "step": 32190 }, { "epoch": 0.5856781680277924, "grad_norm": 1.5010571479797363, "learning_rate": 5.092629639145034e-06, "loss": 0.2488, "step": 32200 }, { "epoch": 0.5858600556576147, "grad_norm": 1.7333347797393799, "learning_rate": 5.090334838376904e-06, "loss": 0.2431, "step": 32210 }, { "epoch": 0.586041943287437, "grad_norm": 1.3925988674163818, "learning_rate": 5.088040353294115e-06, "loss": 0.2368, "step": 32220 }, { "epoch": 0.5862238309172593, "grad_norm": 1.8180700540542603, "learning_rate": 5.085746184692932e-06, "loss": 0.246, "step": 32230 }, { "epoch": 0.5864057185470816, "grad_norm": 1.6846833229064941, "learning_rate": 5.083452333369512e-06, "loss": 0.243, "step": 32240 }, { "epoch": 0.5865876061769039, "grad_norm": 1.4662691354751587, "learning_rate": 5.0811588001198986e-06, "loss": 0.2398, "step": 32250 }, { "epoch": 0.5867694938067262, "grad_norm": 1.7728859186172485, "learning_rate": 5.0788655857400304e-06, "loss": 0.2505, "step": 32260 }, { "epoch": 0.5869513814365485, "grad_norm": 2.0222933292388916, "learning_rate": 5.07657269102573e-06, "loss": 0.2407, "step": 32270 }, { "epoch": 0.5871332690663708, "grad_norm": 1.6106302738189697, "learning_rate": 5.074280116772715e-06, "loss": 0.2442, "step": 32280 }, { "epoch": 0.587315156696193, "grad_norm": 1.6832338571548462, "learning_rate": 5.071987863776584e-06, "loss": 0.245, "step": 32290 }, { "epoch": 0.5874970443260153, "grad_norm": 1.693297266960144, "learning_rate": 5.06969593283283e-06, "loss": 0.2439, "step": 32300 }, { "epoch": 0.5876789319558376, "grad_norm": 1.5660645961761475, "learning_rate": 5.0674043247368306e-06, "loss": 0.2507, "step": 32310 }, { "epoch": 0.5878608195856599, "grad_norm": 1.5096638202667236, "learning_rate": 5.065113040283857e-06, "loss": 0.2456, "step": 32320 }, { "epoch": 0.5880427072154822, "grad_norm": 1.754514455795288, "learning_rate": 5.0628220802690605e-06, "loss": 0.2377, "step": 32330 }, { "epoch": 0.5882245948453045, "grad_norm": 1.626570463180542, "learning_rate": 5.060531445487486e-06, "loss": 0.2505, "step": 32340 }, { "epoch": 0.5884064824751268, "grad_norm": 1.8577197790145874, "learning_rate": 5.058241136734062e-06, "loss": 0.2385, "step": 32350 }, { "epoch": 0.5885883701049491, "grad_norm": 1.508838176727295, "learning_rate": 5.055951154803606e-06, "loss": 0.2491, "step": 32360 }, { "epoch": 0.5887702577347714, "grad_norm": 1.6953033208847046, "learning_rate": 5.053661500490819e-06, "loss": 0.2479, "step": 32370 }, { "epoch": 0.5889521453645937, "grad_norm": 1.4166789054870605, "learning_rate": 5.051372174590294e-06, "loss": 0.2425, "step": 32380 }, { "epoch": 0.589134032994416, "grad_norm": 1.7386935949325562, "learning_rate": 5.049083177896503e-06, "loss": 0.2432, "step": 32390 }, { "epoch": 0.5893159206242383, "grad_norm": 1.655638575553894, "learning_rate": 5.046794511203809e-06, "loss": 0.2424, "step": 32400 }, { "epoch": 0.5894978082540606, "grad_norm": 1.7828532457351685, "learning_rate": 5.044506175306457e-06, "loss": 0.2427, "step": 32410 }, { "epoch": 0.5896796958838829, "grad_norm": 1.4402780532836914, "learning_rate": 5.042218170998582e-06, "loss": 0.2464, "step": 32420 }, { "epoch": 0.5898615835137052, "grad_norm": 1.6462291479110718, "learning_rate": 5.0399304990742e-06, "loss": 0.2351, "step": 32430 }, { "epoch": 0.5900434711435275, "grad_norm": 1.5812169313430786, "learning_rate": 5.037643160327208e-06, "loss": 0.2401, "step": 32440 }, { "epoch": 0.5902253587733498, "grad_norm": 1.6796984672546387, "learning_rate": 5.035356155551396e-06, "loss": 0.2438, "step": 32450 }, { "epoch": 0.5904072464031721, "grad_norm": 1.6128590106964111, "learning_rate": 5.033069485540434e-06, "loss": 0.2471, "step": 32460 }, { "epoch": 0.5905891340329944, "grad_norm": 1.6401640176773071, "learning_rate": 5.030783151087874e-06, "loss": 0.2444, "step": 32470 }, { "epoch": 0.5907710216628167, "grad_norm": 1.7412906885147095, "learning_rate": 5.028497152987156e-06, "loss": 0.2413, "step": 32480 }, { "epoch": 0.590952909292639, "grad_norm": 1.5573347806930542, "learning_rate": 5.026211492031596e-06, "loss": 0.2416, "step": 32490 }, { "epoch": 0.5911347969224613, "grad_norm": 1.5023516416549683, "learning_rate": 5.023926169014404e-06, "loss": 0.2463, "step": 32500 }, { "epoch": 0.5913166845522836, "grad_norm": 1.524918556213379, "learning_rate": 5.021641184728658e-06, "loss": 0.2446, "step": 32510 }, { "epoch": 0.5914985721821059, "grad_norm": 1.633177399635315, "learning_rate": 5.0193565399673326e-06, "loss": 0.2463, "step": 32520 }, { "epoch": 0.5916804598119282, "grad_norm": 1.5492827892303467, "learning_rate": 5.017072235523279e-06, "loss": 0.2378, "step": 32530 }, { "epoch": 0.5918623474417505, "grad_norm": 1.5966237783432007, "learning_rate": 5.0147882721892285e-06, "loss": 0.2395, "step": 32540 }, { "epoch": 0.5920442350715728, "grad_norm": 1.7611052989959717, "learning_rate": 5.012504650757794e-06, "loss": 0.2473, "step": 32550 }, { "epoch": 0.5922261227013951, "grad_norm": 1.980764627456665, "learning_rate": 5.010221372021476e-06, "loss": 0.2486, "step": 32560 }, { "epoch": 0.5924080103312174, "grad_norm": 2.2204697132110596, "learning_rate": 5.007938436772645e-06, "loss": 0.2489, "step": 32570 }, { "epoch": 0.5925898979610397, "grad_norm": 1.5080314874649048, "learning_rate": 5.005655845803563e-06, "loss": 0.2468, "step": 32580 }, { "epoch": 0.592771785590862, "grad_norm": 1.718171238899231, "learning_rate": 5.003373599906366e-06, "loss": 0.2419, "step": 32590 }, { "epoch": 0.5929536732206843, "grad_norm": 1.7438092231750488, "learning_rate": 5.001091699873074e-06, "loss": 0.2458, "step": 32600 }, { "epoch": 0.5931355608505066, "grad_norm": 1.8725066184997559, "learning_rate": 4.998810146495585e-06, "loss": 0.2455, "step": 32610 }, { "epoch": 0.5933174484803289, "grad_norm": 1.5113110542297363, "learning_rate": 4.996528940565677e-06, "loss": 0.2417, "step": 32620 }, { "epoch": 0.5934993361101512, "grad_norm": 1.795778751373291, "learning_rate": 4.994248082875008e-06, "loss": 0.2471, "step": 32630 }, { "epoch": 0.5936812237399735, "grad_norm": 1.6070858240127563, "learning_rate": 4.991967574215113e-06, "loss": 0.2491, "step": 32640 }, { "epoch": 0.5938631113697957, "grad_norm": 1.7281014919281006, "learning_rate": 4.98968741537741e-06, "loss": 0.2468, "step": 32650 }, { "epoch": 0.594044998999618, "grad_norm": 1.5942902565002441, "learning_rate": 4.987407607153191e-06, "loss": 0.2421, "step": 32660 }, { "epoch": 0.5942268866294403, "grad_norm": 1.7070560455322266, "learning_rate": 4.985128150333629e-06, "loss": 0.245, "step": 32670 }, { "epoch": 0.5944087742592626, "grad_norm": 1.6330397129058838, "learning_rate": 4.982849045709775e-06, "loss": 0.2453, "step": 32680 }, { "epoch": 0.5945906618890849, "grad_norm": 1.5252245664596558, "learning_rate": 4.980570294072558e-06, "loss": 0.241, "step": 32690 }, { "epoch": 0.5947725495189072, "grad_norm": 1.816597819328308, "learning_rate": 4.978291896212783e-06, "loss": 0.2445, "step": 32700 }, { "epoch": 0.5949544371487295, "grad_norm": 1.6120845079421997, "learning_rate": 4.976013852921131e-06, "loss": 0.2404, "step": 32710 }, { "epoch": 0.5951363247785518, "grad_norm": 1.585503101348877, "learning_rate": 4.973736164988166e-06, "loss": 0.2485, "step": 32720 }, { "epoch": 0.5953182124083741, "grad_norm": 1.6683989763259888, "learning_rate": 4.971458833204321e-06, "loss": 0.2382, "step": 32730 }, { "epoch": 0.5955001000381964, "grad_norm": 1.6237291097640991, "learning_rate": 4.969181858359913e-06, "loss": 0.2378, "step": 32740 }, { "epoch": 0.5956819876680187, "grad_norm": 1.6243606805801392, "learning_rate": 4.966905241245127e-06, "loss": 0.2435, "step": 32750 }, { "epoch": 0.595863875297841, "grad_norm": 1.4724311828613281, "learning_rate": 4.964628982650029e-06, "loss": 0.2365, "step": 32760 }, { "epoch": 0.5960457629276633, "grad_norm": 1.6828980445861816, "learning_rate": 4.9623530833645624e-06, "loss": 0.2407, "step": 32770 }, { "epoch": 0.5962276505574856, "grad_norm": 1.650145411491394, "learning_rate": 4.960077544178541e-06, "loss": 0.243, "step": 32780 }, { "epoch": 0.5964095381873079, "grad_norm": 1.8577244281768799, "learning_rate": 4.9578023658816565e-06, "loss": 0.2423, "step": 32790 }, { "epoch": 0.5965914258171302, "grad_norm": 1.563045859336853, "learning_rate": 4.955527549263471e-06, "loss": 0.2447, "step": 32800 }, { "epoch": 0.5967733134469525, "grad_norm": 1.4811038970947266, "learning_rate": 4.953253095113431e-06, "loss": 0.2377, "step": 32810 }, { "epoch": 0.5969552010767748, "grad_norm": 1.684252381324768, "learning_rate": 4.950979004220846e-06, "loss": 0.2404, "step": 32820 }, { "epoch": 0.5971370887065971, "grad_norm": 1.6488971710205078, "learning_rate": 4.948705277374908e-06, "loss": 0.2382, "step": 32830 }, { "epoch": 0.5973189763364194, "grad_norm": 1.7067914009094238, "learning_rate": 4.946431915364675e-06, "loss": 0.2384, "step": 32840 }, { "epoch": 0.5975008639662417, "grad_norm": 1.758481740951538, "learning_rate": 4.944158918979087e-06, "loss": 0.241, "step": 32850 }, { "epoch": 0.597682751596064, "grad_norm": 1.8750159740447998, "learning_rate": 4.941886289006949e-06, "loss": 0.2421, "step": 32860 }, { "epoch": 0.5978646392258863, "grad_norm": 1.8463245630264282, "learning_rate": 4.939614026236944e-06, "loss": 0.2412, "step": 32870 }, { "epoch": 0.5980465268557086, "grad_norm": 1.5703116655349731, "learning_rate": 4.937342131457625e-06, "loss": 0.242, "step": 32880 }, { "epoch": 0.5982284144855309, "grad_norm": 1.5049246549606323, "learning_rate": 4.9350706054574185e-06, "loss": 0.2471, "step": 32890 }, { "epoch": 0.5984103021153532, "grad_norm": 1.7129313945770264, "learning_rate": 4.932799449024624e-06, "loss": 0.2425, "step": 32900 }, { "epoch": 0.5985921897451755, "grad_norm": 1.816591739654541, "learning_rate": 4.930528662947409e-06, "loss": 0.244, "step": 32910 }, { "epoch": 0.5987740773749978, "grad_norm": 1.6089032888412476, "learning_rate": 4.928258248013818e-06, "loss": 0.24, "step": 32920 }, { "epoch": 0.5989559650048201, "grad_norm": 1.5756936073303223, "learning_rate": 4.925988205011759e-06, "loss": 0.2404, "step": 32930 }, { "epoch": 0.5991378526346424, "grad_norm": 1.7436996698379517, "learning_rate": 4.9237185347290214e-06, "loss": 0.236, "step": 32940 }, { "epoch": 0.5993197402644647, "grad_norm": 1.8471657037734985, "learning_rate": 4.921449237953257e-06, "loss": 0.2414, "step": 32950 }, { "epoch": 0.599501627894287, "grad_norm": 1.5572426319122314, "learning_rate": 4.919180315471989e-06, "loss": 0.2411, "step": 32960 }, { "epoch": 0.5996835155241093, "grad_norm": 1.5363867282867432, "learning_rate": 4.916911768072614e-06, "loss": 0.2415, "step": 32970 }, { "epoch": 0.5998654031539316, "grad_norm": 1.8263602256774902, "learning_rate": 4.914643596542395e-06, "loss": 0.2487, "step": 32980 }, { "epoch": 0.6000472907837539, "grad_norm": 1.8083959817886353, "learning_rate": 4.912375801668468e-06, "loss": 0.2426, "step": 32990 }, { "epoch": 0.6002291784135761, "grad_norm": 1.656080722808838, "learning_rate": 4.910108384237834e-06, "loss": 0.2484, "step": 33000 }, { "epoch": 0.6004110660433983, "grad_norm": 1.5476170778274536, "learning_rate": 4.907841345037367e-06, "loss": 0.2444, "step": 33010 }, { "epoch": 0.6005929536732206, "grad_norm": 1.5296739339828491, "learning_rate": 4.90557468485381e-06, "loss": 0.2407, "step": 33020 }, { "epoch": 0.6007748413030429, "grad_norm": 1.3939586877822876, "learning_rate": 4.90330840447377e-06, "loss": 0.2396, "step": 33030 }, { "epoch": 0.6009567289328652, "grad_norm": 1.6824944019317627, "learning_rate": 4.901042504683725e-06, "loss": 0.2458, "step": 33040 }, { "epoch": 0.6011386165626875, "grad_norm": 1.5207130908966064, "learning_rate": 4.898776986270021e-06, "loss": 0.2415, "step": 33050 }, { "epoch": 0.6013205041925098, "grad_norm": 1.514952540397644, "learning_rate": 4.896511850018873e-06, "loss": 0.2457, "step": 33060 }, { "epoch": 0.6015023918223321, "grad_norm": 1.5933853387832642, "learning_rate": 4.894247096716358e-06, "loss": 0.2335, "step": 33070 }, { "epoch": 0.6016842794521544, "grad_norm": 1.799322247505188, "learning_rate": 4.891982727148429e-06, "loss": 0.2424, "step": 33080 }, { "epoch": 0.6018661670819767, "grad_norm": 1.5660090446472168, "learning_rate": 4.889718742100896e-06, "loss": 0.2424, "step": 33090 }, { "epoch": 0.602048054711799, "grad_norm": 1.7699600458145142, "learning_rate": 4.8874551423594455e-06, "loss": 0.2453, "step": 33100 }, { "epoch": 0.6022299423416213, "grad_norm": 1.5634673833847046, "learning_rate": 4.8851919287096186e-06, "loss": 0.2371, "step": 33110 }, { "epoch": 0.6024118299714436, "grad_norm": 1.4367324113845825, "learning_rate": 4.882929101936835e-06, "loss": 0.2392, "step": 33120 }, { "epoch": 0.6025937176012659, "grad_norm": 1.5852688550949097, "learning_rate": 4.88066666282637e-06, "loss": 0.2446, "step": 33130 }, { "epoch": 0.6027756052310882, "grad_norm": 1.700419306755066, "learning_rate": 4.87840461216337e-06, "loss": 0.2484, "step": 33140 }, { "epoch": 0.6029574928609105, "grad_norm": 1.4809885025024414, "learning_rate": 4.876142950732845e-06, "loss": 0.2462, "step": 33150 }, { "epoch": 0.6031393804907328, "grad_norm": 1.9356333017349243, "learning_rate": 4.873881679319669e-06, "loss": 0.2457, "step": 33160 }, { "epoch": 0.6033212681205551, "grad_norm": 1.6121289730072021, "learning_rate": 4.871620798708583e-06, "loss": 0.239, "step": 33170 }, { "epoch": 0.6035031557503774, "grad_norm": 1.327017903327942, "learning_rate": 4.869360309684188e-06, "loss": 0.2407, "step": 33180 }, { "epoch": 0.6036850433801997, "grad_norm": 1.5155857801437378, "learning_rate": 4.867100213030957e-06, "loss": 0.2366, "step": 33190 }, { "epoch": 0.603866931010022, "grad_norm": 1.808019995689392, "learning_rate": 4.8648405095332164e-06, "loss": 0.2424, "step": 33200 }, { "epoch": 0.6040488186398443, "grad_norm": 1.7479857206344604, "learning_rate": 4.862581199975165e-06, "loss": 0.2463, "step": 33210 }, { "epoch": 0.6042307062696666, "grad_norm": 1.6887050867080688, "learning_rate": 4.860322285140862e-06, "loss": 0.2414, "step": 33220 }, { "epoch": 0.6044125938994889, "grad_norm": 1.7227073907852173, "learning_rate": 4.858063765814227e-06, "loss": 0.2478, "step": 33230 }, { "epoch": 0.6045944815293112, "grad_norm": 1.778282880783081, "learning_rate": 4.855805642779046e-06, "loss": 0.2517, "step": 33240 }, { "epoch": 0.6047763691591335, "grad_norm": 1.6957732439041138, "learning_rate": 4.853547916818964e-06, "loss": 0.2445, "step": 33250 }, { "epoch": 0.6049582567889558, "grad_norm": 1.529903531074524, "learning_rate": 4.851290588717492e-06, "loss": 0.2411, "step": 33260 }, { "epoch": 0.6051401444187781, "grad_norm": 1.5491935014724731, "learning_rate": 4.849033659258001e-06, "loss": 0.239, "step": 33270 }, { "epoch": 0.6053220320486004, "grad_norm": 1.63615083694458, "learning_rate": 4.846777129223723e-06, "loss": 0.2471, "step": 33280 }, { "epoch": 0.6055039196784227, "grad_norm": 1.678721308708191, "learning_rate": 4.844520999397753e-06, "loss": 0.2457, "step": 33290 }, { "epoch": 0.605685807308245, "grad_norm": 1.7792376279830933, "learning_rate": 4.842265270563045e-06, "loss": 0.2438, "step": 33300 }, { "epoch": 0.6058676949380672, "grad_norm": 1.4629743099212646, "learning_rate": 4.840009943502416e-06, "loss": 0.2379, "step": 33310 }, { "epoch": 0.6060495825678895, "grad_norm": 1.6599922180175781, "learning_rate": 4.837755018998542e-06, "loss": 0.2339, "step": 33320 }, { "epoch": 0.6062314701977118, "grad_norm": 1.4837884902954102, "learning_rate": 4.8355004978339625e-06, "loss": 0.2385, "step": 33330 }, { "epoch": 0.6064133578275341, "grad_norm": 1.6113208532333374, "learning_rate": 4.833246380791071e-06, "loss": 0.2471, "step": 33340 }, { "epoch": 0.6065952454573564, "grad_norm": 1.728805661201477, "learning_rate": 4.830992668652125e-06, "loss": 0.2414, "step": 33350 }, { "epoch": 0.6067771330871787, "grad_norm": 1.7450898885726929, "learning_rate": 4.828739362199242e-06, "loss": 0.237, "step": 33360 }, { "epoch": 0.606959020717001, "grad_norm": 1.3708575963974, "learning_rate": 4.8264864622143956e-06, "loss": 0.2381, "step": 33370 }, { "epoch": 0.6071409083468233, "grad_norm": 1.8511418104171753, "learning_rate": 4.8242339694794225e-06, "loss": 0.2407, "step": 33380 }, { "epoch": 0.6073227959766456, "grad_norm": 1.7683396339416504, "learning_rate": 4.821981884776016e-06, "loss": 0.2497, "step": 33390 }, { "epoch": 0.6075046836064679, "grad_norm": 1.6068861484527588, "learning_rate": 4.819730208885725e-06, "loss": 0.24, "step": 33400 }, { "epoch": 0.6076865712362902, "grad_norm": 1.7960857152938843, "learning_rate": 4.81747894258996e-06, "loss": 0.2352, "step": 33410 }, { "epoch": 0.6078684588661125, "grad_norm": 1.6196613311767578, "learning_rate": 4.81522808666999e-06, "loss": 0.2456, "step": 33420 }, { "epoch": 0.6080503464959348, "grad_norm": 1.4685659408569336, "learning_rate": 4.812977641906938e-06, "loss": 0.2372, "step": 33430 }, { "epoch": 0.6082322341257571, "grad_norm": 1.5599812269210815, "learning_rate": 4.810727609081787e-06, "loss": 0.2453, "step": 33440 }, { "epoch": 0.6084141217555794, "grad_norm": 1.8500854969024658, "learning_rate": 4.808477988975375e-06, "loss": 0.2359, "step": 33450 }, { "epoch": 0.6085960093854017, "grad_norm": 1.5858296155929565, "learning_rate": 4.806228782368402e-06, "loss": 0.234, "step": 33460 }, { "epoch": 0.608777897015224, "grad_norm": 2.029068946838379, "learning_rate": 4.803979990041417e-06, "loss": 0.2467, "step": 33470 }, { "epoch": 0.6089597846450463, "grad_norm": 1.4926445484161377, "learning_rate": 4.80173161277483e-06, "loss": 0.2423, "step": 33480 }, { "epoch": 0.6091416722748686, "grad_norm": 1.6191173791885376, "learning_rate": 4.799483651348908e-06, "loss": 0.2408, "step": 33490 }, { "epoch": 0.6093235599046909, "grad_norm": 1.6659523248672485, "learning_rate": 4.797236106543767e-06, "loss": 0.2404, "step": 33500 }, { "epoch": 0.6095054475345132, "grad_norm": 1.825563907623291, "learning_rate": 4.794988979139387e-06, "loss": 0.236, "step": 33510 }, { "epoch": 0.6096873351643355, "grad_norm": 1.4676662683486938, "learning_rate": 4.792742269915597e-06, "loss": 0.2435, "step": 33520 }, { "epoch": 0.6098692227941578, "grad_norm": 1.6901631355285645, "learning_rate": 4.790495979652083e-06, "loss": 0.2343, "step": 33530 }, { "epoch": 0.6100511104239801, "grad_norm": 1.5743681192398071, "learning_rate": 4.788250109128387e-06, "loss": 0.2382, "step": 33540 }, { "epoch": 0.6102329980538024, "grad_norm": 1.5836228132247925, "learning_rate": 4.786004659123903e-06, "loss": 0.2341, "step": 33550 }, { "epoch": 0.6104148856836247, "grad_norm": 1.704108476638794, "learning_rate": 4.783759630417879e-06, "loss": 0.2515, "step": 33560 }, { "epoch": 0.610596773313447, "grad_norm": 1.8372758626937866, "learning_rate": 4.781515023789419e-06, "loss": 0.247, "step": 33570 }, { "epoch": 0.6107786609432693, "grad_norm": 1.592516303062439, "learning_rate": 4.7792708400174805e-06, "loss": 0.2405, "step": 33580 }, { "epoch": 0.6109605485730916, "grad_norm": 1.580495834350586, "learning_rate": 4.77702707988087e-06, "loss": 0.2406, "step": 33590 }, { "epoch": 0.6111424362029139, "grad_norm": 1.6501171588897705, "learning_rate": 4.774783744158254e-06, "loss": 0.2455, "step": 33600 }, { "epoch": 0.6113243238327362, "grad_norm": 1.5636779069900513, "learning_rate": 4.772540833628143e-06, "loss": 0.247, "step": 33610 }, { "epoch": 0.6115062114625585, "grad_norm": 2.121617078781128, "learning_rate": 4.770298349068907e-06, "loss": 0.239, "step": 33620 }, { "epoch": 0.6116880990923808, "grad_norm": 1.4559310674667358, "learning_rate": 4.768056291258766e-06, "loss": 0.242, "step": 33630 }, { "epoch": 0.611869986722203, "grad_norm": 1.5545709133148193, "learning_rate": 4.765814660975794e-06, "loss": 0.2431, "step": 33640 }, { "epoch": 0.6120518743520253, "grad_norm": 1.6390504837036133, "learning_rate": 4.7635734589979095e-06, "loss": 0.2417, "step": 33650 }, { "epoch": 0.6122337619818476, "grad_norm": 1.9558972120285034, "learning_rate": 4.761332686102892e-06, "loss": 0.2456, "step": 33660 }, { "epoch": 0.6124156496116699, "grad_norm": 1.6971049308776855, "learning_rate": 4.759092343068365e-06, "loss": 0.2442, "step": 33670 }, { "epoch": 0.6125975372414922, "grad_norm": 1.8939284086227417, "learning_rate": 4.756852430671804e-06, "loss": 0.2358, "step": 33680 }, { "epoch": 0.6127794248713145, "grad_norm": 1.8695015907287598, "learning_rate": 4.754612949690539e-06, "loss": 0.2422, "step": 33690 }, { "epoch": 0.6129613125011368, "grad_norm": 1.7663874626159668, "learning_rate": 4.7523739009017456e-06, "loss": 0.2432, "step": 33700 }, { "epoch": 0.6131432001309591, "grad_norm": 1.4655174016952515, "learning_rate": 4.750135285082452e-06, "loss": 0.2408, "step": 33710 }, { "epoch": 0.6133250877607814, "grad_norm": 1.8976051807403564, "learning_rate": 4.747897103009535e-06, "loss": 0.2474, "step": 33720 }, { "epoch": 0.6135069753906037, "grad_norm": 1.8955023288726807, "learning_rate": 4.745659355459724e-06, "loss": 0.2407, "step": 33730 }, { "epoch": 0.613688863020426, "grad_norm": 1.7083512544631958, "learning_rate": 4.743422043209591e-06, "loss": 0.2381, "step": 33740 }, { "epoch": 0.6138707506502483, "grad_norm": 1.5324543714523315, "learning_rate": 4.741185167035563e-06, "loss": 0.2421, "step": 33750 }, { "epoch": 0.6140526382800706, "grad_norm": 1.986851692199707, "learning_rate": 4.738948727713915e-06, "loss": 0.24, "step": 33760 }, { "epoch": 0.6142345259098929, "grad_norm": 1.4030758142471313, "learning_rate": 4.736712726020764e-06, "loss": 0.2424, "step": 33770 }, { "epoch": 0.6144164135397152, "grad_norm": 1.8517976999282837, "learning_rate": 4.734477162732085e-06, "loss": 0.2399, "step": 33780 }, { "epoch": 0.6145983011695375, "grad_norm": 1.649462103843689, "learning_rate": 4.732242038623694e-06, "loss": 0.2462, "step": 33790 }, { "epoch": 0.6147801887993598, "grad_norm": 1.5561299324035645, "learning_rate": 4.730007354471257e-06, "loss": 0.2409, "step": 33800 }, { "epoch": 0.6149620764291821, "grad_norm": 1.7018394470214844, "learning_rate": 4.727773111050287e-06, "loss": 0.2456, "step": 33810 }, { "epoch": 0.6151439640590044, "grad_norm": 1.8638960123062134, "learning_rate": 4.725539309136142e-06, "loss": 0.2424, "step": 33820 }, { "epoch": 0.6153258516888267, "grad_norm": 1.6850372552871704, "learning_rate": 4.723305949504031e-06, "loss": 0.2417, "step": 33830 }, { "epoch": 0.6155077393186489, "grad_norm": 1.7201310396194458, "learning_rate": 4.721073032929006e-06, "loss": 0.2391, "step": 33840 }, { "epoch": 0.6156896269484712, "grad_norm": 1.4858553409576416, "learning_rate": 4.718840560185969e-06, "loss": 0.2397, "step": 33850 }, { "epoch": 0.6158715145782935, "grad_norm": 1.5495392084121704, "learning_rate": 4.716608532049661e-06, "loss": 0.2374, "step": 33860 }, { "epoch": 0.6160534022081158, "grad_norm": 1.3949028253555298, "learning_rate": 4.714376949294675e-06, "loss": 0.2453, "step": 33870 }, { "epoch": 0.6162352898379381, "grad_norm": 1.618809461593628, "learning_rate": 4.71214581269545e-06, "loss": 0.24, "step": 33880 }, { "epoch": 0.6164171774677604, "grad_norm": 1.4157332181930542, "learning_rate": 4.709915123026265e-06, "loss": 0.2432, "step": 33890 }, { "epoch": 0.6165990650975827, "grad_norm": 1.6074423789978027, "learning_rate": 4.707684881061245e-06, "loss": 0.2418, "step": 33900 }, { "epoch": 0.616780952727405, "grad_norm": 1.6395238637924194, "learning_rate": 4.705455087574367e-06, "loss": 0.2471, "step": 33910 }, { "epoch": 0.6169628403572273, "grad_norm": 1.3677653074264526, "learning_rate": 4.703225743339441e-06, "loss": 0.2357, "step": 33920 }, { "epoch": 0.6171447279870496, "grad_norm": 1.661681056022644, "learning_rate": 4.700996849130131e-06, "loss": 0.24, "step": 33930 }, { "epoch": 0.6173266156168719, "grad_norm": 1.9220619201660156, "learning_rate": 4.698768405719937e-06, "loss": 0.2409, "step": 33940 }, { "epoch": 0.6175085032466942, "grad_norm": 1.779252290725708, "learning_rate": 4.696540413882207e-06, "loss": 0.2346, "step": 33950 }, { "epoch": 0.6176903908765164, "grad_norm": 1.6462059020996094, "learning_rate": 4.694312874390133e-06, "loss": 0.239, "step": 33960 }, { "epoch": 0.6178722785063387, "grad_norm": 1.5252341032028198, "learning_rate": 4.6920857880167484e-06, "loss": 0.237, "step": 33970 }, { "epoch": 0.618054166136161, "grad_norm": 1.771949052810669, "learning_rate": 4.6898591555349275e-06, "loss": 0.2417, "step": 33980 }, { "epoch": 0.6182360537659833, "grad_norm": 1.8318979740142822, "learning_rate": 4.687632977717389e-06, "loss": 0.2359, "step": 33990 }, { "epoch": 0.6184179413958056, "grad_norm": 1.8545242547988892, "learning_rate": 4.6854072553366946e-06, "loss": 0.2449, "step": 34000 }, { "epoch": 0.6185998290256279, "grad_norm": 1.6919842958450317, "learning_rate": 4.683181989165249e-06, "loss": 0.2438, "step": 34010 }, { "epoch": 0.6187817166554502, "grad_norm": 1.6476613283157349, "learning_rate": 4.680957179975292e-06, "loss": 0.2338, "step": 34020 }, { "epoch": 0.6189636042852725, "grad_norm": 1.5216007232666016, "learning_rate": 4.678732828538914e-06, "loss": 0.2421, "step": 34030 }, { "epoch": 0.6191454919150948, "grad_norm": 1.6685469150543213, "learning_rate": 4.67650893562804e-06, "loss": 0.2289, "step": 34040 }, { "epoch": 0.6193273795449171, "grad_norm": 1.475750207901001, "learning_rate": 4.674285502014438e-06, "loss": 0.2445, "step": 34050 }, { "epoch": 0.6195092671747394, "grad_norm": 1.673102855682373, "learning_rate": 4.672062528469717e-06, "loss": 0.2392, "step": 34060 }, { "epoch": 0.6196911548045617, "grad_norm": 1.7885318994522095, "learning_rate": 4.669840015765327e-06, "loss": 0.2456, "step": 34070 }, { "epoch": 0.619873042434384, "grad_norm": 1.761447787284851, "learning_rate": 4.667617964672554e-06, "loss": 0.2409, "step": 34080 }, { "epoch": 0.6200549300642063, "grad_norm": 1.5922093391418457, "learning_rate": 4.665396375962529e-06, "loss": 0.2436, "step": 34090 }, { "epoch": 0.6202368176940286, "grad_norm": 1.6674751043319702, "learning_rate": 4.663175250406219e-06, "loss": 0.2398, "step": 34100 }, { "epoch": 0.6204187053238509, "grad_norm": 1.6290957927703857, "learning_rate": 4.660954588774431e-06, "loss": 0.2412, "step": 34110 }, { "epoch": 0.6206005929536732, "grad_norm": 1.4246951341629028, "learning_rate": 4.6587343918378146e-06, "loss": 0.236, "step": 34120 }, { "epoch": 0.6207824805834955, "grad_norm": 1.541675329208374, "learning_rate": 4.656514660366853e-06, "loss": 0.2359, "step": 34130 }, { "epoch": 0.6209643682133178, "grad_norm": 1.6145247220993042, "learning_rate": 4.65429539513187e-06, "loss": 0.2387, "step": 34140 }, { "epoch": 0.6211462558431401, "grad_norm": 1.5923584699630737, "learning_rate": 4.652076596903028e-06, "loss": 0.2441, "step": 34150 }, { "epoch": 0.6213281434729624, "grad_norm": 1.6687527894973755, "learning_rate": 4.649858266450327e-06, "loss": 0.2389, "step": 34160 }, { "epoch": 0.6215100311027847, "grad_norm": 1.6410337686538696, "learning_rate": 4.647640404543603e-06, "loss": 0.241, "step": 34170 }, { "epoch": 0.621691918732607, "grad_norm": 1.4801758527755737, "learning_rate": 4.645423011952531e-06, "loss": 0.2391, "step": 34180 }, { "epoch": 0.6218738063624293, "grad_norm": 1.7676002979278564, "learning_rate": 4.6432060894466255e-06, "loss": 0.243, "step": 34190 }, { "epoch": 0.6220556939922516, "grad_norm": 1.88124418258667, "learning_rate": 4.640989637795235e-06, "loss": 0.2423, "step": 34200 }, { "epoch": 0.6222375816220739, "grad_norm": 1.6932090520858765, "learning_rate": 4.638773657767543e-06, "loss": 0.2408, "step": 34210 }, { "epoch": 0.6224194692518962, "grad_norm": 1.7113335132598877, "learning_rate": 4.636558150132573e-06, "loss": 0.2461, "step": 34220 }, { "epoch": 0.6226013568817185, "grad_norm": 1.641739845275879, "learning_rate": 4.6343431156591835e-06, "loss": 0.2377, "step": 34230 }, { "epoch": 0.6227832445115408, "grad_norm": 1.5061615705490112, "learning_rate": 4.632128555116067e-06, "loss": 0.2397, "step": 34240 }, { "epoch": 0.6229651321413631, "grad_norm": 1.6858576536178589, "learning_rate": 4.629914469271752e-06, "loss": 0.236, "step": 34250 }, { "epoch": 0.6231470197711854, "grad_norm": 1.4242459535598755, "learning_rate": 4.627700858894604e-06, "loss": 0.2385, "step": 34260 }, { "epoch": 0.6233289074010077, "grad_norm": 1.6321117877960205, "learning_rate": 4.625487724752824e-06, "loss": 0.2391, "step": 34270 }, { "epoch": 0.62351079503083, "grad_norm": 2.072688341140747, "learning_rate": 4.623275067614443e-06, "loss": 0.2519, "step": 34280 }, { "epoch": 0.6236926826606523, "grad_norm": 1.6474965810775757, "learning_rate": 4.621062888247333e-06, "loss": 0.247, "step": 34290 }, { "epoch": 0.6238745702904746, "grad_norm": 1.3723793029785156, "learning_rate": 4.618851187419196e-06, "loss": 0.2423, "step": 34300 }, { "epoch": 0.6240564579202968, "grad_norm": 1.5008039474487305, "learning_rate": 4.616639965897566e-06, "loss": 0.232, "step": 34310 }, { "epoch": 0.6242383455501191, "grad_norm": 1.7176462411880493, "learning_rate": 4.614429224449818e-06, "loss": 0.238, "step": 34320 }, { "epoch": 0.6244202331799414, "grad_norm": 1.7501670122146606, "learning_rate": 4.6122189638431535e-06, "loss": 0.2478, "step": 34330 }, { "epoch": 0.6246021208097637, "grad_norm": 1.8477896451950073, "learning_rate": 4.610009184844608e-06, "loss": 0.2362, "step": 34340 }, { "epoch": 0.624784008439586, "grad_norm": 1.6035330295562744, "learning_rate": 4.607799888221055e-06, "loss": 0.2386, "step": 34350 }, { "epoch": 0.6249658960694083, "grad_norm": 1.5486146211624146, "learning_rate": 4.605591074739194e-06, "loss": 0.2338, "step": 34360 }, { "epoch": 0.6251477836992306, "grad_norm": 1.5520670413970947, "learning_rate": 4.603382745165562e-06, "loss": 0.2403, "step": 34370 }, { "epoch": 0.6253296713290529, "grad_norm": 1.716339349746704, "learning_rate": 4.601174900266524e-06, "loss": 0.2372, "step": 34380 }, { "epoch": 0.6255115589588752, "grad_norm": 1.9111133813858032, "learning_rate": 4.598967540808281e-06, "loss": 0.2484, "step": 34390 }, { "epoch": 0.6256934465886975, "grad_norm": 1.4965399503707886, "learning_rate": 4.596760667556861e-06, "loss": 0.2374, "step": 34400 }, { "epoch": 0.6258753342185198, "grad_norm": 1.4793572425842285, "learning_rate": 4.594554281278126e-06, "loss": 0.2434, "step": 34410 }, { "epoch": 0.6260572218483421, "grad_norm": 1.5718588829040527, "learning_rate": 4.59234838273777e-06, "loss": 0.238, "step": 34420 }, { "epoch": 0.6262391094781644, "grad_norm": 1.520565390586853, "learning_rate": 4.5901429727013136e-06, "loss": 0.2427, "step": 34430 }, { "epoch": 0.6264209971079867, "grad_norm": 1.6728506088256836, "learning_rate": 4.587938051934113e-06, "loss": 0.2408, "step": 34440 }, { "epoch": 0.626602884737809, "grad_norm": 1.9819400310516357, "learning_rate": 4.585733621201351e-06, "loss": 0.2373, "step": 34450 }, { "epoch": 0.6267847723676313, "grad_norm": 1.663589596748352, "learning_rate": 4.58352968126804e-06, "loss": 0.2494, "step": 34460 }, { "epoch": 0.6269666599974536, "grad_norm": 1.5852059125900269, "learning_rate": 4.581326232899028e-06, "loss": 0.2392, "step": 34470 }, { "epoch": 0.6271485476272759, "grad_norm": 1.645078182220459, "learning_rate": 4.579123276858985e-06, "loss": 0.2333, "step": 34480 }, { "epoch": 0.6273304352570982, "grad_norm": 1.684162974357605, "learning_rate": 4.576920813912411e-06, "loss": 0.2386, "step": 34490 }, { "epoch": 0.6275123228869205, "grad_norm": 1.5214858055114746, "learning_rate": 4.574718844823641e-06, "loss": 0.2412, "step": 34500 }, { "epoch": 0.6276942105167428, "grad_norm": 1.6977739334106445, "learning_rate": 4.572517370356833e-06, "loss": 0.2375, "step": 34510 }, { "epoch": 0.6278760981465651, "grad_norm": 1.6414052248001099, "learning_rate": 4.570316391275974e-06, "loss": 0.2348, "step": 34520 }, { "epoch": 0.6280579857763874, "grad_norm": 1.5384029150009155, "learning_rate": 4.568115908344881e-06, "loss": 0.2352, "step": 34530 }, { "epoch": 0.6282398734062097, "grad_norm": 1.7038182020187378, "learning_rate": 4.5659159223271975e-06, "loss": 0.2489, "step": 34540 }, { "epoch": 0.628421761036032, "grad_norm": 1.6013275384902954, "learning_rate": 4.563716433986396e-06, "loss": 0.2358, "step": 34550 }, { "epoch": 0.6286036486658543, "grad_norm": 1.7679864168167114, "learning_rate": 4.5615174440857736e-06, "loss": 0.2399, "step": 34560 }, { "epoch": 0.6287855362956766, "grad_norm": 1.598526120185852, "learning_rate": 4.559318953388456e-06, "loss": 0.2387, "step": 34570 }, { "epoch": 0.6289674239254989, "grad_norm": 1.8657137155532837, "learning_rate": 4.557120962657398e-06, "loss": 0.2406, "step": 34580 }, { "epoch": 0.6291493115553212, "grad_norm": 1.4935476779937744, "learning_rate": 4.554923472655376e-06, "loss": 0.2376, "step": 34590 }, { "epoch": 0.6293311991851435, "grad_norm": 1.4541130065917969, "learning_rate": 4.5527264841449955e-06, "loss": 0.2408, "step": 34600 }, { "epoch": 0.6295130868149658, "grad_norm": 1.7831206321716309, "learning_rate": 4.550529997888689e-06, "loss": 0.2313, "step": 34610 }, { "epoch": 0.6296949744447881, "grad_norm": 1.6108342409133911, "learning_rate": 4.548334014648713e-06, "loss": 0.2391, "step": 34620 }, { "epoch": 0.6298768620746104, "grad_norm": 1.557677149772644, "learning_rate": 4.546138535187148e-06, "loss": 0.2392, "step": 34630 }, { "epoch": 0.6300587497044327, "grad_norm": 1.670710563659668, "learning_rate": 4.543943560265904e-06, "loss": 0.2385, "step": 34640 }, { "epoch": 0.630240637334255, "grad_norm": 1.7195303440093994, "learning_rate": 4.541749090646712e-06, "loss": 0.2414, "step": 34650 }, { "epoch": 0.6304225249640772, "grad_norm": 1.515902042388916, "learning_rate": 4.53955512709113e-06, "loss": 0.2361, "step": 34660 }, { "epoch": 0.6306044125938995, "grad_norm": 1.6971460580825806, "learning_rate": 4.537361670360538e-06, "loss": 0.24, "step": 34670 }, { "epoch": 0.6307863002237217, "grad_norm": 1.6551028490066528, "learning_rate": 4.535168721216143e-06, "loss": 0.2415, "step": 34680 }, { "epoch": 0.630968187853544, "grad_norm": 1.6986185312271118, "learning_rate": 4.532976280418972e-06, "loss": 0.2389, "step": 34690 }, { "epoch": 0.6311500754833663, "grad_norm": 1.7337708473205566, "learning_rate": 4.530784348729881e-06, "loss": 0.2372, "step": 34700 }, { "epoch": 0.6313319631131886, "grad_norm": 1.4471923112869263, "learning_rate": 4.528592926909544e-06, "loss": 0.2375, "step": 34710 }, { "epoch": 0.6315138507430109, "grad_norm": 2.0391459465026855, "learning_rate": 4.526402015718462e-06, "loss": 0.239, "step": 34720 }, { "epoch": 0.6316957383728332, "grad_norm": 1.5685819387435913, "learning_rate": 4.524211615916955e-06, "loss": 0.2353, "step": 34730 }, { "epoch": 0.6318776260026555, "grad_norm": 1.8364397287368774, "learning_rate": 4.522021728265168e-06, "loss": 0.2395, "step": 34740 }, { "epoch": 0.6320595136324778, "grad_norm": 1.5756431818008423, "learning_rate": 4.51983235352307e-06, "loss": 0.2334, "step": 34750 }, { "epoch": 0.6322414012623001, "grad_norm": 1.6468877792358398, "learning_rate": 4.51764349245045e-06, "loss": 0.2481, "step": 34760 }, { "epoch": 0.6324232888921224, "grad_norm": 1.6170048713684082, "learning_rate": 4.515455145806916e-06, "loss": 0.2376, "step": 34770 }, { "epoch": 0.6326051765219447, "grad_norm": 1.5876015424728394, "learning_rate": 4.513267314351901e-06, "loss": 0.2384, "step": 34780 }, { "epoch": 0.632787064151767, "grad_norm": 1.5673484802246094, "learning_rate": 4.511079998844661e-06, "loss": 0.235, "step": 34790 }, { "epoch": 0.6329689517815893, "grad_norm": 1.8286389112472534, "learning_rate": 4.508893200044267e-06, "loss": 0.2407, "step": 34800 }, { "epoch": 0.6331508394114116, "grad_norm": 1.3747332096099854, "learning_rate": 4.506706918709615e-06, "loss": 0.2353, "step": 34810 }, { "epoch": 0.6333327270412339, "grad_norm": 1.684360384941101, "learning_rate": 4.50452115559942e-06, "loss": 0.2404, "step": 34820 }, { "epoch": 0.6335146146710562, "grad_norm": 1.5243250131607056, "learning_rate": 4.502335911472219e-06, "loss": 0.2337, "step": 34830 }, { "epoch": 0.6336965023008785, "grad_norm": 1.388210654258728, "learning_rate": 4.500151187086366e-06, "loss": 0.2369, "step": 34840 }, { "epoch": 0.6338783899307008, "grad_norm": 1.6233083009719849, "learning_rate": 4.497966983200036e-06, "loss": 0.2432, "step": 34850 }, { "epoch": 0.6340602775605231, "grad_norm": 1.4745090007781982, "learning_rate": 4.495783300571226e-06, "loss": 0.2315, "step": 34860 }, { "epoch": 0.6342421651903454, "grad_norm": 1.6023504734039307, "learning_rate": 4.493600139957746e-06, "loss": 0.2356, "step": 34870 }, { "epoch": 0.6344240528201677, "grad_norm": 1.7235206365585327, "learning_rate": 4.491417502117231e-06, "loss": 0.237, "step": 34880 }, { "epoch": 0.63460594044999, "grad_norm": 1.5142316818237305, "learning_rate": 4.489235387807132e-06, "loss": 0.2318, "step": 34890 }, { "epoch": 0.6347878280798123, "grad_norm": 1.7700202465057373, "learning_rate": 4.487053797784715e-06, "loss": 0.2372, "step": 34900 }, { "epoch": 0.6349697157096346, "grad_norm": 1.3787379264831543, "learning_rate": 4.484872732807071e-06, "loss": 0.2415, "step": 34910 }, { "epoch": 0.6351516033394569, "grad_norm": 1.5506428480148315, "learning_rate": 4.482692193631104e-06, "loss": 0.2407, "step": 34920 }, { "epoch": 0.6353334909692792, "grad_norm": 1.6538941860198975, "learning_rate": 4.480512181013535e-06, "loss": 0.2382, "step": 34930 }, { "epoch": 0.6355153785991015, "grad_norm": 1.4391694068908691, "learning_rate": 4.478332695710908e-06, "loss": 0.2462, "step": 34940 }, { "epoch": 0.6356972662289238, "grad_norm": 1.7074692249298096, "learning_rate": 4.476153738479575e-06, "loss": 0.2384, "step": 34950 }, { "epoch": 0.635879153858746, "grad_norm": 1.4871289730072021, "learning_rate": 4.473975310075713e-06, "loss": 0.2456, "step": 34960 }, { "epoch": 0.6360610414885683, "grad_norm": 1.4947360754013062, "learning_rate": 4.471797411255311e-06, "loss": 0.2335, "step": 34970 }, { "epoch": 0.6362429291183906, "grad_norm": 1.7491395473480225, "learning_rate": 4.469620042774176e-06, "loss": 0.2405, "step": 34980 }, { "epoch": 0.6364248167482129, "grad_norm": 1.8832077980041504, "learning_rate": 4.46744320538793e-06, "loss": 0.241, "step": 34990 }, { "epoch": 0.6366067043780352, "grad_norm": 1.5034409761428833, "learning_rate": 4.46526689985201e-06, "loss": 0.2429, "step": 35000 }, { "epoch": 0.6367885920078575, "grad_norm": 1.6697059869766235, "learning_rate": 4.46309112692167e-06, "loss": 0.2353, "step": 35010 }, { "epoch": 0.6369704796376798, "grad_norm": 1.7464020252227783, "learning_rate": 4.460915887351982e-06, "loss": 0.2389, "step": 35020 }, { "epoch": 0.6371523672675021, "grad_norm": 1.7947733402252197, "learning_rate": 4.458741181897823e-06, "loss": 0.2376, "step": 35030 }, { "epoch": 0.6373342548973244, "grad_norm": 1.6388744115829468, "learning_rate": 4.456567011313897e-06, "loss": 0.2413, "step": 35040 }, { "epoch": 0.6375161425271467, "grad_norm": 1.8273981809616089, "learning_rate": 4.4543933763547134e-06, "loss": 0.2404, "step": 35050 }, { "epoch": 0.637698030156969, "grad_norm": 1.7127268314361572, "learning_rate": 4.452220277774601e-06, "loss": 0.2422, "step": 35060 }, { "epoch": 0.6378799177867913, "grad_norm": 1.6029785871505737, "learning_rate": 4.450047716327698e-06, "loss": 0.2307, "step": 35070 }, { "epoch": 0.6380618054166136, "grad_norm": 1.5588276386260986, "learning_rate": 4.44787569276796e-06, "loss": 0.2344, "step": 35080 }, { "epoch": 0.6382436930464359, "grad_norm": 1.5859689712524414, "learning_rate": 4.445704207849154e-06, "loss": 0.2439, "step": 35090 }, { "epoch": 0.6384255806762582, "grad_norm": 1.8767293691635132, "learning_rate": 4.4435332623248605e-06, "loss": 0.2358, "step": 35100 }, { "epoch": 0.6386074683060805, "grad_norm": 2.0082614421844482, "learning_rate": 4.441362856948473e-06, "loss": 0.2417, "step": 35110 }, { "epoch": 0.6387893559359028, "grad_norm": 1.5053608417510986, "learning_rate": 4.439192992473197e-06, "loss": 0.2346, "step": 35120 }, { "epoch": 0.6389712435657251, "grad_norm": 1.835160732269287, "learning_rate": 4.437023669652051e-06, "loss": 0.239, "step": 35130 }, { "epoch": 0.6391531311955474, "grad_norm": 1.661138653755188, "learning_rate": 4.434854889237866e-06, "loss": 0.2311, "step": 35140 }, { "epoch": 0.6393350188253697, "grad_norm": 1.7556040287017822, "learning_rate": 4.432686651983282e-06, "loss": 0.2434, "step": 35150 }, { "epoch": 0.639516906455192, "grad_norm": 1.5505863428115845, "learning_rate": 4.430518958640755e-06, "loss": 0.2412, "step": 35160 }, { "epoch": 0.6396987940850143, "grad_norm": 1.4693282842636108, "learning_rate": 4.428351809962547e-06, "loss": 0.2349, "step": 35170 }, { "epoch": 0.6398806817148366, "grad_norm": 1.560219168663025, "learning_rate": 4.426185206700737e-06, "loss": 0.2365, "step": 35180 }, { "epoch": 0.6400625693446589, "grad_norm": 1.5313549041748047, "learning_rate": 4.4240191496072085e-06, "loss": 0.2393, "step": 35190 }, { "epoch": 0.6402444569744812, "grad_norm": 1.6754902601242065, "learning_rate": 4.421853639433659e-06, "loss": 0.2339, "step": 35200 }, { "epoch": 0.6404263446043035, "grad_norm": 1.809590458869934, "learning_rate": 4.419688676931597e-06, "loss": 0.2359, "step": 35210 }, { "epoch": 0.6406082322341258, "grad_norm": 1.6012390851974487, "learning_rate": 4.417524262852337e-06, "loss": 0.2394, "step": 35220 }, { "epoch": 0.6407901198639481, "grad_norm": 1.6614341735839844, "learning_rate": 4.415360397947009e-06, "loss": 0.2358, "step": 35230 }, { "epoch": 0.6409720074937704, "grad_norm": 1.8630151748657227, "learning_rate": 4.413197082966545e-06, "loss": 0.2395, "step": 35240 }, { "epoch": 0.6411538951235927, "grad_norm": 1.649672269821167, "learning_rate": 4.411034318661695e-06, "loss": 0.2395, "step": 35250 }, { "epoch": 0.641335782753415, "grad_norm": 1.7771202325820923, "learning_rate": 4.40887210578301e-06, "loss": 0.2271, "step": 35260 }, { "epoch": 0.6415176703832373, "grad_norm": 1.6678235530853271, "learning_rate": 4.406710445080854e-06, "loss": 0.2328, "step": 35270 }, { "epoch": 0.6416995580130596, "grad_norm": 1.5690922737121582, "learning_rate": 4.404549337305396e-06, "loss": 0.2351, "step": 35280 }, { "epoch": 0.6418814456428819, "grad_norm": 1.7024822235107422, "learning_rate": 4.4023887832066185e-06, "loss": 0.2416, "step": 35290 }, { "epoch": 0.6420633332727042, "grad_norm": 1.5230350494384766, "learning_rate": 4.400228783534307e-06, "loss": 0.2325, "step": 35300 }, { "epoch": 0.6422452209025264, "grad_norm": 1.8537375926971436, "learning_rate": 4.3980693390380565e-06, "loss": 0.24, "step": 35310 }, { "epoch": 0.6424271085323487, "grad_norm": 1.72139573097229, "learning_rate": 4.395910450467271e-06, "loss": 0.2409, "step": 35320 }, { "epoch": 0.642608996162171, "grad_norm": 1.5200196504592896, "learning_rate": 4.3937521185711555e-06, "loss": 0.242, "step": 35330 }, { "epoch": 0.6427908837919933, "grad_norm": 1.3149529695510864, "learning_rate": 4.391594344098729e-06, "loss": 0.2328, "step": 35340 }, { "epoch": 0.6429727714218156, "grad_norm": 1.732356071472168, "learning_rate": 4.389437127798813e-06, "loss": 0.2385, "step": 35350 }, { "epoch": 0.6431546590516379, "grad_norm": 1.7922861576080322, "learning_rate": 4.387280470420037e-06, "loss": 0.239, "step": 35360 }, { "epoch": 0.6433365466814602, "grad_norm": 1.7364299297332764, "learning_rate": 4.385124372710835e-06, "loss": 0.2355, "step": 35370 }, { "epoch": 0.6435184343112825, "grad_norm": 1.7171980142593384, "learning_rate": 4.38296883541945e-06, "loss": 0.2421, "step": 35380 }, { "epoch": 0.6437003219411048, "grad_norm": 2.0195508003234863, "learning_rate": 4.380813859293924e-06, "loss": 0.2366, "step": 35390 }, { "epoch": 0.6438822095709271, "grad_norm": 1.7796465158462524, "learning_rate": 4.378659445082112e-06, "loss": 0.2427, "step": 35400 }, { "epoch": 0.6440640972007494, "grad_norm": 1.6207078695297241, "learning_rate": 4.376505593531668e-06, "loss": 0.2355, "step": 35410 }, { "epoch": 0.6442459848305717, "grad_norm": 1.5916668176651, "learning_rate": 4.3743523053900545e-06, "loss": 0.2354, "step": 35420 }, { "epoch": 0.644427872460394, "grad_norm": 1.7301862239837646, "learning_rate": 4.372199581404538e-06, "loss": 0.2413, "step": 35430 }, { "epoch": 0.6446097600902163, "grad_norm": 1.803095817565918, "learning_rate": 4.370047422322186e-06, "loss": 0.2353, "step": 35440 }, { "epoch": 0.6447916477200386, "grad_norm": 1.467585563659668, "learning_rate": 4.367895828889874e-06, "loss": 0.2289, "step": 35450 }, { "epoch": 0.6449735353498609, "grad_norm": 1.920684814453125, "learning_rate": 4.365744801854278e-06, "loss": 0.2368, "step": 35460 }, { "epoch": 0.6451554229796832, "grad_norm": 1.7212369441986084, "learning_rate": 4.363594341961881e-06, "loss": 0.2396, "step": 35470 }, { "epoch": 0.6453373106095055, "grad_norm": 1.631714105606079, "learning_rate": 4.361444449958966e-06, "loss": 0.2362, "step": 35480 }, { "epoch": 0.6455191982393278, "grad_norm": 1.6330602169036865, "learning_rate": 4.35929512659162e-06, "loss": 0.2376, "step": 35490 }, { "epoch": 0.6457010858691501, "grad_norm": 1.7206764221191406, "learning_rate": 4.357146372605733e-06, "loss": 0.2384, "step": 35500 }, { "epoch": 0.6458829734989724, "grad_norm": 1.6833299398422241, "learning_rate": 4.354998188746995e-06, "loss": 0.2391, "step": 35510 }, { "epoch": 0.6460648611287946, "grad_norm": 1.6396733522415161, "learning_rate": 4.352850575760904e-06, "loss": 0.2372, "step": 35520 }, { "epoch": 0.6462467487586169, "grad_norm": 1.6733689308166504, "learning_rate": 4.350703534392755e-06, "loss": 0.2392, "step": 35530 }, { "epoch": 0.6464286363884392, "grad_norm": 1.844093680381775, "learning_rate": 4.3485570653876435e-06, "loss": 0.2504, "step": 35540 }, { "epoch": 0.6466105240182615, "grad_norm": 1.4026226997375488, "learning_rate": 4.346411169490471e-06, "loss": 0.2277, "step": 35550 }, { "epoch": 0.6467924116480838, "grad_norm": 1.6221991777420044, "learning_rate": 4.344265847445935e-06, "loss": 0.2353, "step": 35560 }, { "epoch": 0.6469742992779061, "grad_norm": 1.6393556594848633, "learning_rate": 4.34212109999854e-06, "loss": 0.2331, "step": 35570 }, { "epoch": 0.6471561869077284, "grad_norm": 1.5509982109069824, "learning_rate": 4.339976927892587e-06, "loss": 0.2359, "step": 35580 }, { "epoch": 0.6473380745375507, "grad_norm": 1.542065978050232, "learning_rate": 4.337833331872176e-06, "loss": 0.2394, "step": 35590 }, { "epoch": 0.647519962167373, "grad_norm": 1.509740948677063, "learning_rate": 4.335690312681211e-06, "loss": 0.2339, "step": 35600 }, { "epoch": 0.6477018497971953, "grad_norm": 1.8732882738113403, "learning_rate": 4.3335478710633934e-06, "loss": 0.2408, "step": 35610 }, { "epoch": 0.6478837374270175, "grad_norm": 1.5189504623413086, "learning_rate": 4.3314060077622255e-06, "loss": 0.2393, "step": 35620 }, { "epoch": 0.6480656250568398, "grad_norm": 1.6176422834396362, "learning_rate": 4.329264723521007e-06, "loss": 0.2356, "step": 35630 }, { "epoch": 0.6482475126866621, "grad_norm": 1.9019798040390015, "learning_rate": 4.327124019082838e-06, "loss": 0.2407, "step": 35640 }, { "epoch": 0.6484294003164844, "grad_norm": 1.6597651243209839, "learning_rate": 4.324983895190616e-06, "loss": 0.2375, "step": 35650 }, { "epoch": 0.6486112879463067, "grad_norm": 1.9416677951812744, "learning_rate": 4.32284435258704e-06, "loss": 0.2398, "step": 35660 }, { "epoch": 0.648793175576129, "grad_norm": 1.4401899576187134, "learning_rate": 4.320705392014605e-06, "loss": 0.2379, "step": 35670 }, { "epoch": 0.6489750632059513, "grad_norm": 1.863563060760498, "learning_rate": 4.318567014215604e-06, "loss": 0.2333, "step": 35680 }, { "epoch": 0.6491569508357736, "grad_norm": 1.4783889055252075, "learning_rate": 4.316429219932129e-06, "loss": 0.236, "step": 35690 }, { "epoch": 0.6493388384655959, "grad_norm": 1.8240243196487427, "learning_rate": 4.3142920099060665e-06, "loss": 0.2346, "step": 35700 }, { "epoch": 0.6495207260954182, "grad_norm": 1.58461332321167, "learning_rate": 4.312155384879105e-06, "loss": 0.235, "step": 35710 }, { "epoch": 0.6497026137252405, "grad_norm": 1.6267204284667969, "learning_rate": 4.310019345592728e-06, "loss": 0.2384, "step": 35720 }, { "epoch": 0.6498845013550628, "grad_norm": 1.576684594154358, "learning_rate": 4.307883892788212e-06, "loss": 0.2393, "step": 35730 }, { "epoch": 0.6500663889848851, "grad_norm": 1.4969137907028198, "learning_rate": 4.3057490272066364e-06, "loss": 0.2374, "step": 35740 }, { "epoch": 0.6502482766147074, "grad_norm": 1.5810168981552124, "learning_rate": 4.303614749588869e-06, "loss": 0.2363, "step": 35750 }, { "epoch": 0.6504301642445297, "grad_norm": 1.716822624206543, "learning_rate": 4.301481060675584e-06, "loss": 0.2411, "step": 35760 }, { "epoch": 0.650612051874352, "grad_norm": 1.5337903499603271, "learning_rate": 4.2993479612072405e-06, "loss": 0.2329, "step": 35770 }, { "epoch": 0.6507939395041743, "grad_norm": 1.6045070886611938, "learning_rate": 4.297215451924102e-06, "loss": 0.2286, "step": 35780 }, { "epoch": 0.6509758271339966, "grad_norm": 1.3870935440063477, "learning_rate": 4.295083533566221e-06, "loss": 0.2349, "step": 35790 }, { "epoch": 0.6511577147638189, "grad_norm": 1.548222541809082, "learning_rate": 4.292952206873444e-06, "loss": 0.2459, "step": 35800 }, { "epoch": 0.6513396023936412, "grad_norm": 1.8515039682388306, "learning_rate": 4.290821472585422e-06, "loss": 0.2376, "step": 35810 }, { "epoch": 0.6515214900234635, "grad_norm": 1.6403709650039673, "learning_rate": 4.288691331441589e-06, "loss": 0.2388, "step": 35820 }, { "epoch": 0.6517033776532858, "grad_norm": 1.6138057708740234, "learning_rate": 4.286561784181177e-06, "loss": 0.2419, "step": 35830 }, { "epoch": 0.6518852652831081, "grad_norm": 1.584830403327942, "learning_rate": 4.284432831543217e-06, "loss": 0.2345, "step": 35840 }, { "epoch": 0.6520671529129304, "grad_norm": 1.9941959381103516, "learning_rate": 4.282304474266525e-06, "loss": 0.2406, "step": 35850 }, { "epoch": 0.6522490405427527, "grad_norm": 1.617231011390686, "learning_rate": 4.2801767130897175e-06, "loss": 0.2394, "step": 35860 }, { "epoch": 0.652430928172575, "grad_norm": 1.5851857662200928, "learning_rate": 4.2780495487512015e-06, "loss": 0.2369, "step": 35870 }, { "epoch": 0.6526128158023973, "grad_norm": 2.2101097106933594, "learning_rate": 4.2759229819891755e-06, "loss": 0.237, "step": 35880 }, { "epoch": 0.6527947034322196, "grad_norm": 1.4305611848831177, "learning_rate": 4.2737970135416315e-06, "loss": 0.2355, "step": 35890 }, { "epoch": 0.6529765910620419, "grad_norm": 1.558733582496643, "learning_rate": 4.271671644146354e-06, "loss": 0.2475, "step": 35900 }, { "epoch": 0.6531584786918642, "grad_norm": 1.5892337560653687, "learning_rate": 4.269546874540923e-06, "loss": 0.2413, "step": 35910 }, { "epoch": 0.6533403663216865, "grad_norm": 1.5351895093917847, "learning_rate": 4.267422705462704e-06, "loss": 0.2416, "step": 35920 }, { "epoch": 0.6535222539515088, "grad_norm": 1.537174940109253, "learning_rate": 4.265299137648859e-06, "loss": 0.2364, "step": 35930 }, { "epoch": 0.653704141581331, "grad_norm": 1.6429611444473267, "learning_rate": 4.263176171836338e-06, "loss": 0.2312, "step": 35940 }, { "epoch": 0.6538860292111534, "grad_norm": 1.6118537187576294, "learning_rate": 4.261053808761883e-06, "loss": 0.2421, "step": 35950 }, { "epoch": 0.6540679168409756, "grad_norm": 1.6887657642364502, "learning_rate": 4.258932049162031e-06, "loss": 0.2365, "step": 35960 }, { "epoch": 0.654249804470798, "grad_norm": 1.3923894166946411, "learning_rate": 4.256810893773105e-06, "loss": 0.2405, "step": 35970 }, { "epoch": 0.6544316921006202, "grad_norm": 1.4629430770874023, "learning_rate": 4.254690343331218e-06, "loss": 0.2366, "step": 35980 }, { "epoch": 0.6546135797304425, "grad_norm": 1.5644056797027588, "learning_rate": 4.252570398572275e-06, "loss": 0.2378, "step": 35990 }, { "epoch": 0.6547954673602648, "grad_norm": 1.7078191041946411, "learning_rate": 4.2504510602319694e-06, "loss": 0.2361, "step": 36000 }, { "epoch": 0.6549773549900871, "grad_norm": 1.7219936847686768, "learning_rate": 4.2483323290457875e-06, "loss": 0.2492, "step": 36010 }, { "epoch": 0.6551592426199094, "grad_norm": 1.4759796857833862, "learning_rate": 4.246214205749002e-06, "loss": 0.2433, "step": 36020 }, { "epoch": 0.6553411302497317, "grad_norm": 1.7434996366500854, "learning_rate": 4.244096691076673e-06, "loss": 0.2371, "step": 36030 }, { "epoch": 0.655523017879554, "grad_norm": 1.4803670644760132, "learning_rate": 4.2419797857636545e-06, "loss": 0.2407, "step": 36040 }, { "epoch": 0.6557049055093763, "grad_norm": 1.481276035308838, "learning_rate": 4.239863490544581e-06, "loss": 0.2306, "step": 36050 }, { "epoch": 0.6558867931391986, "grad_norm": 1.5250709056854248, "learning_rate": 4.237747806153886e-06, "loss": 0.2342, "step": 36060 }, { "epoch": 0.6560686807690209, "grad_norm": 1.7232813835144043, "learning_rate": 4.235632733325782e-06, "loss": 0.2357, "step": 36070 }, { "epoch": 0.6562505683988432, "grad_norm": 1.7061187028884888, "learning_rate": 4.233518272794274e-06, "loss": 0.2386, "step": 36080 }, { "epoch": 0.6564324560286655, "grad_norm": 1.6222949028015137, "learning_rate": 4.231404425293151e-06, "loss": 0.2389, "step": 36090 }, { "epoch": 0.6566143436584878, "grad_norm": 1.8346209526062012, "learning_rate": 4.229291191555994e-06, "loss": 0.2391, "step": 36100 }, { "epoch": 0.6567962312883101, "grad_norm": 1.5065826177597046, "learning_rate": 4.227178572316167e-06, "loss": 0.2389, "step": 36110 }, { "epoch": 0.6569781189181324, "grad_norm": 1.27340829372406, "learning_rate": 4.225066568306823e-06, "loss": 0.23, "step": 36120 }, { "epoch": 0.6571600065479547, "grad_norm": 1.6099566221237183, "learning_rate": 4.222955180260899e-06, "loss": 0.2336, "step": 36130 }, { "epoch": 0.657341894177777, "grad_norm": 1.6589232683181763, "learning_rate": 4.2208444089111225e-06, "loss": 0.238, "step": 36140 }, { "epoch": 0.6575237818075993, "grad_norm": 1.4069366455078125, "learning_rate": 4.21873425499e-06, "loss": 0.2368, "step": 36150 }, { "epoch": 0.6577056694374216, "grad_norm": 1.7187268733978271, "learning_rate": 4.216624719229832e-06, "loss": 0.2296, "step": 36160 }, { "epoch": 0.6578875570672439, "grad_norm": 1.5550141334533691, "learning_rate": 4.2145158023627e-06, "loss": 0.2374, "step": 36170 }, { "epoch": 0.6580694446970662, "grad_norm": 1.5555415153503418, "learning_rate": 4.21240750512047e-06, "loss": 0.2335, "step": 36180 }, { "epoch": 0.6582513323268885, "grad_norm": 1.7765238285064697, "learning_rate": 4.210299828234795e-06, "loss": 0.2391, "step": 36190 }, { "epoch": 0.6584332199567108, "grad_norm": 1.5160225629806519, "learning_rate": 4.20819277243711e-06, "loss": 0.2343, "step": 36200 }, { "epoch": 0.6586151075865331, "grad_norm": 1.5511935949325562, "learning_rate": 4.206086338458638e-06, "loss": 0.2367, "step": 36210 }, { "epoch": 0.6587969952163554, "grad_norm": 1.8922092914581299, "learning_rate": 4.2039805270303866e-06, "loss": 0.2379, "step": 36220 }, { "epoch": 0.6589788828461777, "grad_norm": 1.6068605184555054, "learning_rate": 4.2018753388831425e-06, "loss": 0.2353, "step": 36230 }, { "epoch": 0.659160770476, "grad_norm": 1.6927746534347534, "learning_rate": 4.199770774747478e-06, "loss": 0.236, "step": 36240 }, { "epoch": 0.6593426581058223, "grad_norm": 1.6540027856826782, "learning_rate": 4.197666835353753e-06, "loss": 0.2336, "step": 36250 }, { "epoch": 0.6595245457356446, "grad_norm": 1.714683175086975, "learning_rate": 4.1955635214321065e-06, "loss": 0.237, "step": 36260 }, { "epoch": 0.6597064333654669, "grad_norm": 1.6757017374038696, "learning_rate": 4.193460833712461e-06, "loss": 0.2365, "step": 36270 }, { "epoch": 0.6598883209952892, "grad_norm": 1.6575994491577148, "learning_rate": 4.1913587729245205e-06, "loss": 0.2288, "step": 36280 }, { "epoch": 0.6600702086251115, "grad_norm": 1.646742343902588, "learning_rate": 4.189257339797774e-06, "loss": 0.2322, "step": 36290 }, { "epoch": 0.6602520962549338, "grad_norm": 1.5150055885314941, "learning_rate": 4.187156535061493e-06, "loss": 0.2367, "step": 36300 }, { "epoch": 0.660433983884756, "grad_norm": 1.8869123458862305, "learning_rate": 4.185056359444728e-06, "loss": 0.2397, "step": 36310 }, { "epoch": 0.6606158715145783, "grad_norm": 1.3744709491729736, "learning_rate": 4.182956813676315e-06, "loss": 0.2319, "step": 36320 }, { "epoch": 0.6607977591444006, "grad_norm": 1.5034881830215454, "learning_rate": 4.180857898484865e-06, "loss": 0.2378, "step": 36330 }, { "epoch": 0.6609796467742229, "grad_norm": 1.6327844858169556, "learning_rate": 4.178759614598776e-06, "loss": 0.2404, "step": 36340 }, { "epoch": 0.6611615344040451, "grad_norm": 1.694279670715332, "learning_rate": 4.176661962746227e-06, "loss": 0.2377, "step": 36350 }, { "epoch": 0.6613434220338674, "grad_norm": 1.4731638431549072, "learning_rate": 4.174564943655176e-06, "loss": 0.2352, "step": 36360 }, { "epoch": 0.6615253096636897, "grad_norm": 1.4561645984649658, "learning_rate": 4.172468558053359e-06, "loss": 0.2338, "step": 36370 }, { "epoch": 0.661707197293512, "grad_norm": 1.6582468748092651, "learning_rate": 4.170372806668294e-06, "loss": 0.2358, "step": 36380 }, { "epoch": 0.6618890849233343, "grad_norm": 1.476356029510498, "learning_rate": 4.168277690227284e-06, "loss": 0.2284, "step": 36390 }, { "epoch": 0.6620709725531566, "grad_norm": 1.6371877193450928, "learning_rate": 4.166183209457402e-06, "loss": 0.2405, "step": 36400 }, { "epoch": 0.6622528601829789, "grad_norm": 1.613621711730957, "learning_rate": 4.164089365085507e-06, "loss": 0.2394, "step": 36410 }, { "epoch": 0.6624347478128012, "grad_norm": 1.4393151998519897, "learning_rate": 4.161996157838237e-06, "loss": 0.238, "step": 36420 }, { "epoch": 0.6626166354426235, "grad_norm": 1.6156543493270874, "learning_rate": 4.159903588442007e-06, "loss": 0.2324, "step": 36430 }, { "epoch": 0.6627985230724458, "grad_norm": 1.7959401607513428, "learning_rate": 4.1578116576230085e-06, "loss": 0.2385, "step": 36440 }, { "epoch": 0.6629804107022681, "grad_norm": 1.3899948596954346, "learning_rate": 4.1557203661072185e-06, "loss": 0.2326, "step": 36450 }, { "epoch": 0.6631622983320904, "grad_norm": 1.5743433237075806, "learning_rate": 4.153629714620384e-06, "loss": 0.2367, "step": 36460 }, { "epoch": 0.6633441859619127, "grad_norm": 1.589131236076355, "learning_rate": 4.151539703888034e-06, "loss": 0.2357, "step": 36470 }, { "epoch": 0.663526073591735, "grad_norm": 1.4768692255020142, "learning_rate": 4.1494503346354774e-06, "loss": 0.2344, "step": 36480 }, { "epoch": 0.6637079612215573, "grad_norm": 1.385448694229126, "learning_rate": 4.147361607587793e-06, "loss": 0.2373, "step": 36490 }, { "epoch": 0.6638898488513796, "grad_norm": 1.6330894231796265, "learning_rate": 4.145273523469846e-06, "loss": 0.2327, "step": 36500 }, { "epoch": 0.6640717364812019, "grad_norm": 1.6317130327224731, "learning_rate": 4.143186083006272e-06, "loss": 0.2353, "step": 36510 }, { "epoch": 0.6642536241110242, "grad_norm": 1.6413300037384033, "learning_rate": 4.1410992869214865e-06, "loss": 0.235, "step": 36520 }, { "epoch": 0.6644355117408465, "grad_norm": 1.697766661643982, "learning_rate": 4.139013135939678e-06, "loss": 0.2382, "step": 36530 }, { "epoch": 0.6646173993706688, "grad_norm": 1.6241058111190796, "learning_rate": 4.136927630784812e-06, "loss": 0.2348, "step": 36540 }, { "epoch": 0.6647992870004911, "grad_norm": 1.3705792427062988, "learning_rate": 4.134842772180635e-06, "loss": 0.2304, "step": 36550 }, { "epoch": 0.6649811746303134, "grad_norm": 1.6359267234802246, "learning_rate": 4.132758560850663e-06, "loss": 0.2387, "step": 36560 }, { "epoch": 0.6651630622601357, "grad_norm": 1.5386621952056885, "learning_rate": 4.130674997518192e-06, "loss": 0.2324, "step": 36570 }, { "epoch": 0.665344949889958, "grad_norm": 1.6936745643615723, "learning_rate": 4.128592082906287e-06, "loss": 0.2398, "step": 36580 }, { "epoch": 0.6655268375197803, "grad_norm": 1.7429406642913818, "learning_rate": 4.126509817737792e-06, "loss": 0.2359, "step": 36590 }, { "epoch": 0.6657087251496026, "grad_norm": 1.4650548696517944, "learning_rate": 4.124428202735327e-06, "loss": 0.2323, "step": 36600 }, { "epoch": 0.6658906127794249, "grad_norm": 1.4957636594772339, "learning_rate": 4.122347238621286e-06, "loss": 0.2418, "step": 36610 }, { "epoch": 0.6660725004092471, "grad_norm": 1.7695200443267822, "learning_rate": 4.120266926117832e-06, "loss": 0.242, "step": 36620 }, { "epoch": 0.6662543880390694, "grad_norm": 1.748258352279663, "learning_rate": 4.118187265946906e-06, "loss": 0.2402, "step": 36630 }, { "epoch": 0.6664362756688917, "grad_norm": 1.6868304014205933, "learning_rate": 4.116108258830226e-06, "loss": 0.2321, "step": 36640 }, { "epoch": 0.666618163298714, "grad_norm": 1.5699268579483032, "learning_rate": 4.114029905489275e-06, "loss": 0.2322, "step": 36650 }, { "epoch": 0.6668000509285363, "grad_norm": 1.8072772026062012, "learning_rate": 4.111952206645316e-06, "loss": 0.2369, "step": 36660 }, { "epoch": 0.6669819385583586, "grad_norm": 1.7015644311904907, "learning_rate": 4.109875163019383e-06, "loss": 0.2309, "step": 36670 }, { "epoch": 0.6671638261881809, "grad_norm": 1.454583764076233, "learning_rate": 4.107798775332279e-06, "loss": 0.2295, "step": 36680 }, { "epoch": 0.6673457138180032, "grad_norm": 1.6292835474014282, "learning_rate": 4.105723044304585e-06, "loss": 0.2385, "step": 36690 }, { "epoch": 0.6675276014478255, "grad_norm": 1.6205466985702515, "learning_rate": 4.103647970656653e-06, "loss": 0.2323, "step": 36700 }, { "epoch": 0.6677094890776478, "grad_norm": 1.5392240285873413, "learning_rate": 4.101573555108604e-06, "loss": 0.2413, "step": 36710 }, { "epoch": 0.6678913767074701, "grad_norm": 1.6830856800079346, "learning_rate": 4.099499798380331e-06, "loss": 0.2368, "step": 36720 }, { "epoch": 0.6680732643372924, "grad_norm": 1.8239654302597046, "learning_rate": 4.097426701191501e-06, "loss": 0.2355, "step": 36730 }, { "epoch": 0.6682551519671147, "grad_norm": 1.7940762042999268, "learning_rate": 4.095354264261548e-06, "loss": 0.2358, "step": 36740 }, { "epoch": 0.668437039596937, "grad_norm": 1.4773577451705933, "learning_rate": 4.093282488309682e-06, "loss": 0.2347, "step": 36750 }, { "epoch": 0.6686189272267593, "grad_norm": 1.453567385673523, "learning_rate": 4.091211374054881e-06, "loss": 0.2344, "step": 36760 }, { "epoch": 0.6688008148565816, "grad_norm": 1.6955293416976929, "learning_rate": 4.089140922215893e-06, "loss": 0.2424, "step": 36770 }, { "epoch": 0.6689827024864039, "grad_norm": 2.25719952583313, "learning_rate": 4.0870711335112365e-06, "loss": 0.2425, "step": 36780 }, { "epoch": 0.6691645901162262, "grad_norm": 1.6270569562911987, "learning_rate": 4.085002008659198e-06, "loss": 0.2329, "step": 36790 }, { "epoch": 0.6693464777460485, "grad_norm": 1.7081178426742554, "learning_rate": 4.082933548377839e-06, "loss": 0.2293, "step": 36800 }, { "epoch": 0.6695283653758708, "grad_norm": 1.7463079690933228, "learning_rate": 4.080865753384985e-06, "loss": 0.237, "step": 36810 }, { "epoch": 0.6697102530056931, "grad_norm": 1.6229698657989502, "learning_rate": 4.078798624398233e-06, "loss": 0.2346, "step": 36820 }, { "epoch": 0.6698921406355154, "grad_norm": 1.6543521881103516, "learning_rate": 4.0767321621349474e-06, "loss": 0.2359, "step": 36830 }, { "epoch": 0.6700740282653377, "grad_norm": 1.412560224533081, "learning_rate": 4.0746663673122636e-06, "loss": 0.2288, "step": 36840 }, { "epoch": 0.67025591589516, "grad_norm": 1.60344398021698, "learning_rate": 4.072601240647084e-06, "loss": 0.2328, "step": 36850 }, { "epoch": 0.6704378035249823, "grad_norm": 1.5811963081359863, "learning_rate": 4.0705367828560785e-06, "loss": 0.2322, "step": 36860 }, { "epoch": 0.6706196911548046, "grad_norm": 1.6778242588043213, "learning_rate": 4.068472994655685e-06, "loss": 0.2293, "step": 36870 }, { "epoch": 0.6708015787846269, "grad_norm": 1.6231199502944946, "learning_rate": 4.066409876762108e-06, "loss": 0.2375, "step": 36880 }, { "epoch": 0.6709834664144492, "grad_norm": 1.4876571893692017, "learning_rate": 4.064347429891326e-06, "loss": 0.2319, "step": 36890 }, { "epoch": 0.6711653540442715, "grad_norm": 1.5922592878341675, "learning_rate": 4.062285654759077e-06, "loss": 0.2404, "step": 36900 }, { "epoch": 0.6713472416740938, "grad_norm": 1.814605951309204, "learning_rate": 4.060224552080866e-06, "loss": 0.2409, "step": 36910 }, { "epoch": 0.6715291293039161, "grad_norm": 1.5209298133850098, "learning_rate": 4.0581641225719705e-06, "loss": 0.2349, "step": 36920 }, { "epoch": 0.6717110169337384, "grad_norm": 1.7172188758850098, "learning_rate": 4.0561043669474275e-06, "loss": 0.2316, "step": 36930 }, { "epoch": 0.6718929045635607, "grad_norm": 1.73810613155365, "learning_rate": 4.054045285922046e-06, "loss": 0.2389, "step": 36940 }, { "epoch": 0.672074792193383, "grad_norm": 1.5693923234939575, "learning_rate": 4.051986880210399e-06, "loss": 0.2402, "step": 36950 }, { "epoch": 0.6722566798232052, "grad_norm": 1.8240548372268677, "learning_rate": 4.049929150526825e-06, "loss": 0.2302, "step": 36960 }, { "epoch": 0.6724385674530275, "grad_norm": 1.446181297302246, "learning_rate": 4.047872097585425e-06, "loss": 0.2335, "step": 36970 }, { "epoch": 0.6726204550828498, "grad_norm": 1.6454559564590454, "learning_rate": 4.04581572210007e-06, "loss": 0.231, "step": 36980 }, { "epoch": 0.6728023427126721, "grad_norm": 1.697892427444458, "learning_rate": 4.04376002478439e-06, "loss": 0.2365, "step": 36990 }, { "epoch": 0.6729842303424944, "grad_norm": 1.418923020362854, "learning_rate": 4.0417050063517875e-06, "loss": 0.2322, "step": 37000 }, { "epoch": 0.6731661179723167, "grad_norm": 1.6225695610046387, "learning_rate": 4.039650667515423e-06, "loss": 0.2409, "step": 37010 }, { "epoch": 0.673348005602139, "grad_norm": 1.5534909963607788, "learning_rate": 4.037597008988224e-06, "loss": 0.2356, "step": 37020 }, { "epoch": 0.6735298932319613, "grad_norm": 1.5790067911148071, "learning_rate": 4.03554403148288e-06, "loss": 0.2314, "step": 37030 }, { "epoch": 0.6737117808617836, "grad_norm": 1.5760855674743652, "learning_rate": 4.033491735711847e-06, "loss": 0.2323, "step": 37040 }, { "epoch": 0.6738936684916059, "grad_norm": 1.5739428997039795, "learning_rate": 4.031440122387342e-06, "loss": 0.2289, "step": 37050 }, { "epoch": 0.6740755561214282, "grad_norm": 1.399857997894287, "learning_rate": 4.0293891922213455e-06, "loss": 0.2344, "step": 37060 }, { "epoch": 0.6742574437512505, "grad_norm": 1.5287282466888428, "learning_rate": 4.0273389459256026e-06, "loss": 0.2371, "step": 37070 }, { "epoch": 0.6744393313810728, "grad_norm": 1.5963331460952759, "learning_rate": 4.025289384211617e-06, "loss": 0.2326, "step": 37080 }, { "epoch": 0.6746212190108951, "grad_norm": 1.6757148504257202, "learning_rate": 4.023240507790661e-06, "loss": 0.2282, "step": 37090 }, { "epoch": 0.6748031066407174, "grad_norm": 1.8936065435409546, "learning_rate": 4.021192317373764e-06, "loss": 0.2376, "step": 37100 }, { "epoch": 0.6749849942705397, "grad_norm": 1.5679763555526733, "learning_rate": 4.019144813671719e-06, "loss": 0.237, "step": 37110 }, { "epoch": 0.675166881900362, "grad_norm": 1.6264623403549194, "learning_rate": 4.017097997395083e-06, "loss": 0.2333, "step": 37120 }, { "epoch": 0.6753487695301843, "grad_norm": 1.6567474603652954, "learning_rate": 4.015051869254167e-06, "loss": 0.2311, "step": 37130 }, { "epoch": 0.6755306571600066, "grad_norm": 1.9365566968917847, "learning_rate": 4.013006429959055e-06, "loss": 0.2367, "step": 37140 }, { "epoch": 0.6757125447898289, "grad_norm": 1.5402438640594482, "learning_rate": 4.010961680219581e-06, "loss": 0.2291, "step": 37150 }, { "epoch": 0.6758944324196512, "grad_norm": 1.8027145862579346, "learning_rate": 4.008917620745345e-06, "loss": 0.2403, "step": 37160 }, { "epoch": 0.6760763200494735, "grad_norm": 1.5318578481674194, "learning_rate": 4.006874252245706e-06, "loss": 0.2293, "step": 37170 }, { "epoch": 0.6762582076792958, "grad_norm": 1.392268419265747, "learning_rate": 4.004831575429784e-06, "loss": 0.2347, "step": 37180 }, { "epoch": 0.676440095309118, "grad_norm": 1.8570098876953125, "learning_rate": 4.0027895910064595e-06, "loss": 0.2414, "step": 37190 }, { "epoch": 0.6766219829389403, "grad_norm": 1.6286484003067017, "learning_rate": 4.000748299684372e-06, "loss": 0.2345, "step": 37200 }, { "epoch": 0.6768038705687626, "grad_norm": 1.8365941047668457, "learning_rate": 3.998707702171917e-06, "loss": 0.2372, "step": 37210 }, { "epoch": 0.6769857581985849, "grad_norm": 1.6280345916748047, "learning_rate": 3.996667799177258e-06, "loss": 0.2316, "step": 37220 }, { "epoch": 0.6771676458284072, "grad_norm": 1.6245297193527222, "learning_rate": 3.994628591408307e-06, "loss": 0.2255, "step": 37230 }, { "epoch": 0.6773495334582295, "grad_norm": 1.76244056224823, "learning_rate": 3.992590079572744e-06, "loss": 0.24, "step": 37240 }, { "epoch": 0.6775314210880518, "grad_norm": 1.8369066715240479, "learning_rate": 3.990552264378001e-06, "loss": 0.2357, "step": 37250 }, { "epoch": 0.677713308717874, "grad_norm": 1.6122403144836426, "learning_rate": 3.988515146531271e-06, "loss": 0.2315, "step": 37260 }, { "epoch": 0.6778951963476963, "grad_norm": 1.419942021369934, "learning_rate": 3.986478726739504e-06, "loss": 0.2293, "step": 37270 }, { "epoch": 0.6780770839775186, "grad_norm": 1.6076029539108276, "learning_rate": 3.984443005709407e-06, "loss": 0.234, "step": 37280 }, { "epoch": 0.6782589716073409, "grad_norm": 1.5286948680877686, "learning_rate": 3.982407984147449e-06, "loss": 0.2347, "step": 37290 }, { "epoch": 0.6784408592371632, "grad_norm": 1.5347563028335571, "learning_rate": 3.980373662759851e-06, "loss": 0.2311, "step": 37300 }, { "epoch": 0.6786227468669855, "grad_norm": 1.5793254375457764, "learning_rate": 3.978340042252594e-06, "loss": 0.2296, "step": 37310 }, { "epoch": 0.6788046344968078, "grad_norm": 1.653449535369873, "learning_rate": 3.976307123331414e-06, "loss": 0.235, "step": 37320 }, { "epoch": 0.6789865221266301, "grad_norm": 1.7387100458145142, "learning_rate": 3.974274906701803e-06, "loss": 0.2374, "step": 37330 }, { "epoch": 0.6791684097564524, "grad_norm": 1.4326965808868408, "learning_rate": 3.972243393069013e-06, "loss": 0.23, "step": 37340 }, { "epoch": 0.6793502973862747, "grad_norm": 1.7696864604949951, "learning_rate": 3.970212583138049e-06, "loss": 0.2327, "step": 37350 }, { "epoch": 0.679532185016097, "grad_norm": 1.4643921852111816, "learning_rate": 3.968182477613672e-06, "loss": 0.2289, "step": 37360 }, { "epoch": 0.6797140726459193, "grad_norm": 1.5544707775115967, "learning_rate": 3.9661530772004e-06, "loss": 0.2308, "step": 37370 }, { "epoch": 0.6798959602757416, "grad_norm": 1.5018261671066284, "learning_rate": 3.964124382602503e-06, "loss": 0.23, "step": 37380 }, { "epoch": 0.6800778479055639, "grad_norm": 1.6293808221817017, "learning_rate": 3.96209639452401e-06, "loss": 0.2342, "step": 37390 }, { "epoch": 0.6802597355353862, "grad_norm": 1.7673211097717285, "learning_rate": 3.960069113668703e-06, "loss": 0.2324, "step": 37400 }, { "epoch": 0.6804416231652085, "grad_norm": 1.6128101348876953, "learning_rate": 3.95804254074012e-06, "loss": 0.2305, "step": 37410 }, { "epoch": 0.6806235107950308, "grad_norm": 1.6624560356140137, "learning_rate": 3.956016676441548e-06, "loss": 0.2394, "step": 37420 }, { "epoch": 0.6808053984248531, "grad_norm": 1.5195612907409668, "learning_rate": 3.9539915214760366e-06, "loss": 0.2324, "step": 37430 }, { "epoch": 0.6809872860546754, "grad_norm": 1.7004345655441284, "learning_rate": 3.951967076546383e-06, "loss": 0.2413, "step": 37440 }, { "epoch": 0.6811691736844977, "grad_norm": 1.7250142097473145, "learning_rate": 3.9499433423551406e-06, "loss": 0.238, "step": 37450 }, { "epoch": 0.68135106131432, "grad_norm": 1.7119930982589722, "learning_rate": 3.9479203196046134e-06, "loss": 0.2405, "step": 37460 }, { "epoch": 0.6815329489441423, "grad_norm": 1.5222562551498413, "learning_rate": 3.9458980089968605e-06, "loss": 0.2332, "step": 37470 }, { "epoch": 0.6817148365739646, "grad_norm": 1.6190292835235596, "learning_rate": 3.9438764112336936e-06, "loss": 0.2396, "step": 37480 }, { "epoch": 0.6818967242037869, "grad_norm": 1.6661221981048584, "learning_rate": 3.941855527016681e-06, "loss": 0.2277, "step": 37490 }, { "epoch": 0.6820786118336092, "grad_norm": 1.6784721612930298, "learning_rate": 3.939835357047137e-06, "loss": 0.2317, "step": 37500 }, { "epoch": 0.6822604994634315, "grad_norm": 1.856717824935913, "learning_rate": 3.937815902026129e-06, "loss": 0.2321, "step": 37510 }, { "epoch": 0.6824423870932538, "grad_norm": 1.6892873048782349, "learning_rate": 3.935797162654481e-06, "loss": 0.2299, "step": 37520 }, { "epoch": 0.6826242747230761, "grad_norm": 1.5455994606018066, "learning_rate": 3.933779139632763e-06, "loss": 0.2377, "step": 37530 }, { "epoch": 0.6828061623528984, "grad_norm": 1.4251238107681274, "learning_rate": 3.9317618336613e-06, "loss": 0.2295, "step": 37540 }, { "epoch": 0.6829880499827207, "grad_norm": 1.8959994316101074, "learning_rate": 3.929745245440168e-06, "loss": 0.2347, "step": 37550 }, { "epoch": 0.683169937612543, "grad_norm": 1.5539604425430298, "learning_rate": 3.92772937566919e-06, "loss": 0.2402, "step": 37560 }, { "epoch": 0.6833518252423653, "grad_norm": 1.7088817358016968, "learning_rate": 3.925714225047946e-06, "loss": 0.2395, "step": 37570 }, { "epoch": 0.6835337128721876, "grad_norm": 1.7770910263061523, "learning_rate": 3.923699794275761e-06, "loss": 0.2301, "step": 37580 }, { "epoch": 0.6837156005020099, "grad_norm": 1.7653735876083374, "learning_rate": 3.921686084051713e-06, "loss": 0.2299, "step": 37590 }, { "epoch": 0.6838974881318322, "grad_norm": 1.975982427597046, "learning_rate": 3.919673095074631e-06, "loss": 0.2397, "step": 37600 }, { "epoch": 0.6840793757616545, "grad_norm": 1.491562843322754, "learning_rate": 3.917660828043088e-06, "loss": 0.2326, "step": 37610 }, { "epoch": 0.6842612633914767, "grad_norm": 1.6037137508392334, "learning_rate": 3.915649283655413e-06, "loss": 0.2291, "step": 37620 }, { "epoch": 0.684443151021299, "grad_norm": 1.4939860105514526, "learning_rate": 3.9136384626096825e-06, "loss": 0.2353, "step": 37630 }, { "epoch": 0.6846250386511213, "grad_norm": 1.4401239156723022, "learning_rate": 3.9116283656037204e-06, "loss": 0.2362, "step": 37640 }, { "epoch": 0.6848069262809436, "grad_norm": 1.710833191871643, "learning_rate": 3.9096189933350985e-06, "loss": 0.232, "step": 37650 }, { "epoch": 0.6849888139107659, "grad_norm": 1.612045168876648, "learning_rate": 3.907610346501139e-06, "loss": 0.2341, "step": 37660 }, { "epoch": 0.6851707015405882, "grad_norm": 1.6596647500991821, "learning_rate": 3.905602425798912e-06, "loss": 0.2371, "step": 37670 }, { "epoch": 0.6853525891704105, "grad_norm": 1.4847307205200195, "learning_rate": 3.903595231925238e-06, "loss": 0.2314, "step": 37680 }, { "epoch": 0.6855344768002328, "grad_norm": 1.5489894151687622, "learning_rate": 3.90158876557668e-06, "loss": 0.2338, "step": 37690 }, { "epoch": 0.6857163644300551, "grad_norm": 1.4311063289642334, "learning_rate": 3.899583027449554e-06, "loss": 0.234, "step": 37700 }, { "epoch": 0.6858982520598774, "grad_norm": 1.7153013944625854, "learning_rate": 3.897578018239918e-06, "loss": 0.2289, "step": 37710 }, { "epoch": 0.6860801396896997, "grad_norm": 1.7383366823196411, "learning_rate": 3.895573738643579e-06, "loss": 0.2348, "step": 37720 }, { "epoch": 0.686262027319522, "grad_norm": 1.7097667455673218, "learning_rate": 3.893570189356095e-06, "loss": 0.2338, "step": 37730 }, { "epoch": 0.6864439149493443, "grad_norm": 1.7213491201400757, "learning_rate": 3.891567371072766e-06, "loss": 0.2336, "step": 37740 }, { "epoch": 0.6866258025791666, "grad_norm": 1.8945796489715576, "learning_rate": 3.889565284488638e-06, "loss": 0.2442, "step": 37750 }, { "epoch": 0.6868076902089889, "grad_norm": 1.6850038766860962, "learning_rate": 3.887563930298503e-06, "loss": 0.2341, "step": 37760 }, { "epoch": 0.6869895778388112, "grad_norm": 1.4625493288040161, "learning_rate": 3.885563309196905e-06, "loss": 0.2395, "step": 37770 }, { "epoch": 0.6871714654686335, "grad_norm": 1.5227839946746826, "learning_rate": 3.883563421878124e-06, "loss": 0.2331, "step": 37780 }, { "epoch": 0.6873533530984558, "grad_norm": 1.4926806688308716, "learning_rate": 3.881564269036193e-06, "loss": 0.2294, "step": 37790 }, { "epoch": 0.6875352407282781, "grad_norm": 1.4199090003967285, "learning_rate": 3.879565851364888e-06, "loss": 0.2384, "step": 37800 }, { "epoch": 0.6877171283581004, "grad_norm": 1.8124295473098755, "learning_rate": 3.877568169557728e-06, "loss": 0.2324, "step": 37810 }, { "epoch": 0.6878990159879227, "grad_norm": 1.5440409183502197, "learning_rate": 3.8755712243079765e-06, "loss": 0.2281, "step": 37820 }, { "epoch": 0.688080903617745, "grad_norm": 1.8262373208999634, "learning_rate": 3.873575016308645e-06, "loss": 0.2296, "step": 37830 }, { "epoch": 0.6882627912475673, "grad_norm": 1.7182132005691528, "learning_rate": 3.871579546252486e-06, "loss": 0.231, "step": 37840 }, { "epoch": 0.6884446788773896, "grad_norm": 1.4751484394073486, "learning_rate": 3.869584814831996e-06, "loss": 0.236, "step": 37850 }, { "epoch": 0.6886265665072119, "grad_norm": 1.5095540285110474, "learning_rate": 3.867590822739417e-06, "loss": 0.2284, "step": 37860 }, { "epoch": 0.6888084541370342, "grad_norm": 1.5121315717697144, "learning_rate": 3.865597570666732e-06, "loss": 0.2332, "step": 37870 }, { "epoch": 0.6889903417668565, "grad_norm": 1.7181339263916016, "learning_rate": 3.863605059305671e-06, "loss": 0.2367, "step": 37880 }, { "epoch": 0.6891722293966788, "grad_norm": 1.6357125043869019, "learning_rate": 3.861613289347702e-06, "loss": 0.2315, "step": 37890 }, { "epoch": 0.6893541170265011, "grad_norm": 1.7370637655258179, "learning_rate": 3.859622261484039e-06, "loss": 0.2292, "step": 37900 }, { "epoch": 0.6895360046563234, "grad_norm": 1.4477519989013672, "learning_rate": 3.857631976405639e-06, "loss": 0.2325, "step": 37910 }, { "epoch": 0.6897178922861457, "grad_norm": 1.724469780921936, "learning_rate": 3.855642434803198e-06, "loss": 0.235, "step": 37920 }, { "epoch": 0.689899779915968, "grad_norm": 1.744842529296875, "learning_rate": 3.853653637367156e-06, "loss": 0.2293, "step": 37930 }, { "epoch": 0.6900816675457903, "grad_norm": 1.5330880880355835, "learning_rate": 3.851665584787697e-06, "loss": 0.2297, "step": 37940 }, { "epoch": 0.6902635551756126, "grad_norm": 1.4893438816070557, "learning_rate": 3.849678277754742e-06, "loss": 0.2337, "step": 37950 }, { "epoch": 0.6904454428054349, "grad_norm": 1.544753909111023, "learning_rate": 3.847691716957957e-06, "loss": 0.2303, "step": 37960 }, { "epoch": 0.6906273304352571, "grad_norm": 1.641642689704895, "learning_rate": 3.845705903086742e-06, "loss": 0.2305, "step": 37970 }, { "epoch": 0.6908092180650794, "grad_norm": 1.5734165906906128, "learning_rate": 3.8437208368302515e-06, "loss": 0.2278, "step": 37980 }, { "epoch": 0.6909911056949017, "grad_norm": 1.542496681213379, "learning_rate": 3.841736518877368e-06, "loss": 0.2389, "step": 37990 }, { "epoch": 0.691172993324724, "grad_norm": 1.6024008989334106, "learning_rate": 3.839752949916719e-06, "loss": 0.2352, "step": 38000 }, { "epoch": 0.6913548809545463, "grad_norm": 1.6617779731750488, "learning_rate": 3.8377701306366695e-06, "loss": 0.2298, "step": 38010 }, { "epoch": 0.6915367685843685, "grad_norm": 1.6846224069595337, "learning_rate": 3.835788061725332e-06, "loss": 0.2316, "step": 38020 }, { "epoch": 0.6917186562141908, "grad_norm": 1.8125076293945312, "learning_rate": 3.8338067438705475e-06, "loss": 0.2389, "step": 38030 }, { "epoch": 0.6919005438440131, "grad_norm": 1.7671900987625122, "learning_rate": 3.8318261777599045e-06, "loss": 0.2389, "step": 38040 }, { "epoch": 0.6920824314738354, "grad_norm": 1.673972487449646, "learning_rate": 3.829846364080728e-06, "loss": 0.2311, "step": 38050 }, { "epoch": 0.6922643191036577, "grad_norm": 1.6495424509048462, "learning_rate": 3.827867303520083e-06, "loss": 0.236, "step": 38060 }, { "epoch": 0.69244620673348, "grad_norm": 1.5846751928329468, "learning_rate": 3.8258889967647695e-06, "loss": 0.2349, "step": 38070 }, { "epoch": 0.6926280943633023, "grad_norm": 1.518911361694336, "learning_rate": 3.82391144450133e-06, "loss": 0.2231, "step": 38080 }, { "epoch": 0.6928099819931246, "grad_norm": 1.697725772857666, "learning_rate": 3.8219346474160454e-06, "loss": 0.2356, "step": 38090 }, { "epoch": 0.6929918696229469, "grad_norm": 1.619160532951355, "learning_rate": 3.81995860619493e-06, "loss": 0.2393, "step": 38100 }, { "epoch": 0.6931737572527692, "grad_norm": 1.5078074932098389, "learning_rate": 3.81798332152374e-06, "loss": 0.2272, "step": 38110 }, { "epoch": 0.6933556448825915, "grad_norm": 1.6396231651306152, "learning_rate": 3.816008794087965e-06, "loss": 0.2326, "step": 38120 }, { "epoch": 0.6935375325124138, "grad_norm": 1.6575895547866821, "learning_rate": 3.8140350245728397e-06, "loss": 0.2413, "step": 38130 }, { "epoch": 0.6937194201422361, "grad_norm": 1.4341062307357788, "learning_rate": 3.812062013663327e-06, "loss": 0.2337, "step": 38140 }, { "epoch": 0.6939013077720584, "grad_norm": 1.5563515424728394, "learning_rate": 3.81008976204413e-06, "loss": 0.2288, "step": 38150 }, { "epoch": 0.6940831954018807, "grad_norm": 1.3189319372177124, "learning_rate": 3.80811827039969e-06, "loss": 0.2354, "step": 38160 }, { "epoch": 0.694265083031703, "grad_norm": 1.6120872497558594, "learning_rate": 3.80614753941418e-06, "loss": 0.2297, "step": 38170 }, { "epoch": 0.6944469706615253, "grad_norm": 1.8018306493759155, "learning_rate": 3.8041775697715155e-06, "loss": 0.2347, "step": 38180 }, { "epoch": 0.6946288582913476, "grad_norm": 1.5315794944763184, "learning_rate": 3.802208362155342e-06, "loss": 0.2296, "step": 38190 }, { "epoch": 0.6948107459211699, "grad_norm": 1.7121151685714722, "learning_rate": 3.8002399172490445e-06, "loss": 0.234, "step": 38200 }, { "epoch": 0.6949926335509922, "grad_norm": 1.478793740272522, "learning_rate": 3.798272235735738e-06, "loss": 0.2303, "step": 38210 }, { "epoch": 0.6951745211808145, "grad_norm": 1.649949312210083, "learning_rate": 3.7963053182982814e-06, "loss": 0.2336, "step": 38220 }, { "epoch": 0.6953564088106368, "grad_norm": 1.639299988746643, "learning_rate": 3.7943391656192598e-06, "loss": 0.2278, "step": 38230 }, { "epoch": 0.6955382964404591, "grad_norm": 1.837043046951294, "learning_rate": 3.792373778380998e-06, "loss": 0.2375, "step": 38240 }, { "epoch": 0.6957201840702814, "grad_norm": 1.5591641664505005, "learning_rate": 3.790409157265551e-06, "loss": 0.2285, "step": 38250 }, { "epoch": 0.6959020717001037, "grad_norm": 1.4238921403884888, "learning_rate": 3.788445302954712e-06, "loss": 0.2289, "step": 38260 }, { "epoch": 0.696083959329926, "grad_norm": 1.539294719696045, "learning_rate": 3.7864822161300076e-06, "loss": 0.2343, "step": 38270 }, { "epoch": 0.6962658469597482, "grad_norm": 1.8126156330108643, "learning_rate": 3.7845198974726956e-06, "loss": 0.2385, "step": 38280 }, { "epoch": 0.6964477345895705, "grad_norm": 1.4379500150680542, "learning_rate": 3.782558347663769e-06, "loss": 0.2291, "step": 38290 }, { "epoch": 0.6966296222193928, "grad_norm": 1.982388973236084, "learning_rate": 3.7805975673839533e-06, "loss": 0.2313, "step": 38300 }, { "epoch": 0.6968115098492151, "grad_norm": 1.4648336172103882, "learning_rate": 3.778637557313707e-06, "loss": 0.2246, "step": 38310 }, { "epoch": 0.6969933974790374, "grad_norm": 1.503151774406433, "learning_rate": 3.776678318133221e-06, "loss": 0.2318, "step": 38320 }, { "epoch": 0.6971752851088597, "grad_norm": 1.6394948959350586, "learning_rate": 3.7747198505224214e-06, "loss": 0.2306, "step": 38330 }, { "epoch": 0.697357172738682, "grad_norm": 1.4794031381607056, "learning_rate": 3.772762155160964e-06, "loss": 0.2347, "step": 38340 }, { "epoch": 0.6975390603685043, "grad_norm": 1.8471801280975342, "learning_rate": 3.770805232728235e-06, "loss": 0.2298, "step": 38350 }, { "epoch": 0.6977209479983266, "grad_norm": 1.371334433555603, "learning_rate": 3.7688490839033552e-06, "loss": 0.2332, "step": 38360 }, { "epoch": 0.6979028356281489, "grad_norm": 1.5441700220108032, "learning_rate": 3.7668937093651756e-06, "loss": 0.2293, "step": 38370 }, { "epoch": 0.6980847232579712, "grad_norm": 1.5151057243347168, "learning_rate": 3.7649391097922804e-06, "loss": 0.2323, "step": 38380 }, { "epoch": 0.6982666108877935, "grad_norm": 1.5354171991348267, "learning_rate": 3.7629852858629827e-06, "loss": 0.2351, "step": 38390 }, { "epoch": 0.6984484985176158, "grad_norm": 1.5565170049667358, "learning_rate": 3.7610322382553266e-06, "loss": 0.2255, "step": 38400 }, { "epoch": 0.6986303861474381, "grad_norm": 1.8153775930404663, "learning_rate": 3.759079967647087e-06, "loss": 0.2309, "step": 38410 }, { "epoch": 0.6988122737772604, "grad_norm": 1.4748708009719849, "learning_rate": 3.757128474715771e-06, "loss": 0.2315, "step": 38420 }, { "epoch": 0.6989941614070827, "grad_norm": 1.8670872449874878, "learning_rate": 3.755177760138613e-06, "loss": 0.235, "step": 38430 }, { "epoch": 0.699176049036905, "grad_norm": 1.724707007408142, "learning_rate": 3.75322782459258e-06, "loss": 0.2298, "step": 38440 }, { "epoch": 0.6993579366667273, "grad_norm": 1.6317079067230225, "learning_rate": 3.7512786687543655e-06, "loss": 0.2312, "step": 38450 }, { "epoch": 0.6995398242965496, "grad_norm": 1.6839632987976074, "learning_rate": 3.749330293300394e-06, "loss": 0.2345, "step": 38460 }, { "epoch": 0.6997217119263719, "grad_norm": 1.5725150108337402, "learning_rate": 3.747382698906821e-06, "loss": 0.2335, "step": 38470 }, { "epoch": 0.6999035995561942, "grad_norm": 1.7414135932922363, "learning_rate": 3.7454358862495284e-06, "loss": 0.2288, "step": 38480 }, { "epoch": 0.7000854871860165, "grad_norm": 1.5352978706359863, "learning_rate": 3.743489856004127e-06, "loss": 0.231, "step": 38490 }, { "epoch": 0.7002673748158388, "grad_norm": 1.6880759000778198, "learning_rate": 3.741544608845958e-06, "loss": 0.2347, "step": 38500 }, { "epoch": 0.7004492624456611, "grad_norm": 1.3766638040542603, "learning_rate": 3.739600145450088e-06, "loss": 0.2316, "step": 38510 }, { "epoch": 0.7006311500754834, "grad_norm": 1.7921630144119263, "learning_rate": 3.737656466491315e-06, "loss": 0.2308, "step": 38520 }, { "epoch": 0.7008130377053057, "grad_norm": 1.7640645503997803, "learning_rate": 3.735713572644163e-06, "loss": 0.239, "step": 38530 }, { "epoch": 0.700994925335128, "grad_norm": 1.392504334449768, "learning_rate": 3.733771464582882e-06, "loss": 0.2215, "step": 38540 }, { "epoch": 0.7011768129649503, "grad_norm": 1.7011175155639648, "learning_rate": 3.731830142981452e-06, "loss": 0.2351, "step": 38550 }, { "epoch": 0.7013587005947726, "grad_norm": 1.6048471927642822, "learning_rate": 3.729889608513576e-06, "loss": 0.2276, "step": 38560 }, { "epoch": 0.7015405882245949, "grad_norm": 1.5558810234069824, "learning_rate": 3.727949861852691e-06, "loss": 0.2298, "step": 38570 }, { "epoch": 0.7017224758544172, "grad_norm": 1.6301403045654297, "learning_rate": 3.726010903671955e-06, "loss": 0.2318, "step": 38580 }, { "epoch": 0.7019043634842395, "grad_norm": 1.6974676847457886, "learning_rate": 3.7240727346442508e-06, "loss": 0.2325, "step": 38590 }, { "epoch": 0.7020862511140618, "grad_norm": 1.539141058921814, "learning_rate": 3.7221353554421948e-06, "loss": 0.229, "step": 38600 }, { "epoch": 0.702268138743884, "grad_norm": 1.4468986988067627, "learning_rate": 3.72019876673812e-06, "loss": 0.23, "step": 38610 }, { "epoch": 0.7024500263737063, "grad_norm": 1.6177672147750854, "learning_rate": 3.7182629692040937e-06, "loss": 0.2342, "step": 38620 }, { "epoch": 0.7026319140035286, "grad_norm": 1.505551815032959, "learning_rate": 3.716327963511904e-06, "loss": 0.2297, "step": 38630 }, { "epoch": 0.7028138016333509, "grad_norm": 1.571563959121704, "learning_rate": 3.7143937503330647e-06, "loss": 0.2324, "step": 38640 }, { "epoch": 0.7029956892631732, "grad_norm": 1.7983750104904175, "learning_rate": 3.712460330338814e-06, "loss": 0.2371, "step": 38650 }, { "epoch": 0.7031775768929955, "grad_norm": 1.86733078956604, "learning_rate": 3.710527704200114e-06, "loss": 0.2314, "step": 38660 }, { "epoch": 0.7033594645228178, "grad_norm": 1.7437961101531982, "learning_rate": 3.708595872587657e-06, "loss": 0.2324, "step": 38670 }, { "epoch": 0.7035413521526401, "grad_norm": 1.5591180324554443, "learning_rate": 3.706664836171853e-06, "loss": 0.2196, "step": 38680 }, { "epoch": 0.7037232397824624, "grad_norm": 1.8341630697250366, "learning_rate": 3.70473459562284e-06, "loss": 0.2299, "step": 38690 }, { "epoch": 0.7039051274122847, "grad_norm": 1.739684820175171, "learning_rate": 3.702805151610476e-06, "loss": 0.2336, "step": 38700 }, { "epoch": 0.704087015042107, "grad_norm": 1.5761926174163818, "learning_rate": 3.700876504804346e-06, "loss": 0.2348, "step": 38710 }, { "epoch": 0.7042689026719293, "grad_norm": 2.0253257751464844, "learning_rate": 3.698948655873759e-06, "loss": 0.2327, "step": 38720 }, { "epoch": 0.7044507903017516, "grad_norm": 1.8174173831939697, "learning_rate": 3.697021605487744e-06, "loss": 0.2423, "step": 38730 }, { "epoch": 0.7046326779315739, "grad_norm": 1.4835727214813232, "learning_rate": 3.6950953543150538e-06, "loss": 0.2333, "step": 38740 }, { "epoch": 0.7048145655613962, "grad_norm": 1.4881632328033447, "learning_rate": 3.6931699030241653e-06, "loss": 0.2378, "step": 38750 }, { "epoch": 0.7049964531912185, "grad_norm": 1.7560468912124634, "learning_rate": 3.6912452522832747e-06, "loss": 0.2296, "step": 38760 }, { "epoch": 0.7051783408210408, "grad_norm": 1.948347806930542, "learning_rate": 3.6893214027603064e-06, "loss": 0.231, "step": 38770 }, { "epoch": 0.7053602284508631, "grad_norm": 1.4767361879348755, "learning_rate": 3.6873983551229e-06, "loss": 0.2378, "step": 38780 }, { "epoch": 0.7055421160806854, "grad_norm": 1.6362309455871582, "learning_rate": 3.6854761100384206e-06, "loss": 0.2368, "step": 38790 }, { "epoch": 0.7057240037105077, "grad_norm": 1.5816195011138916, "learning_rate": 3.683554668173952e-06, "loss": 0.2304, "step": 38800 }, { "epoch": 0.70590589134033, "grad_norm": 1.5121639966964722, "learning_rate": 3.681634030196305e-06, "loss": 0.2278, "step": 38810 }, { "epoch": 0.7060877789701523, "grad_norm": 1.697943925857544, "learning_rate": 3.6797141967720052e-06, "loss": 0.2293, "step": 38820 }, { "epoch": 0.7062696665999746, "grad_norm": 1.5413819551467896, "learning_rate": 3.6777951685673008e-06, "loss": 0.2344, "step": 38830 }, { "epoch": 0.7064515542297969, "grad_norm": 1.6251344680786133, "learning_rate": 3.6758769462481614e-06, "loss": 0.2353, "step": 38840 }, { "epoch": 0.7066334418596192, "grad_norm": 1.3809196949005127, "learning_rate": 3.673959530480277e-06, "loss": 0.2284, "step": 38850 }, { "epoch": 0.7068153294894414, "grad_norm": 1.3880391120910645, "learning_rate": 3.6720429219290553e-06, "loss": 0.2306, "step": 38860 }, { "epoch": 0.7069972171192637, "grad_norm": 1.3753941059112549, "learning_rate": 3.670127121259631e-06, "loss": 0.2262, "step": 38870 }, { "epoch": 0.707179104749086, "grad_norm": 2.5044376850128174, "learning_rate": 3.66821212913685e-06, "loss": 0.2285, "step": 38880 }, { "epoch": 0.7073609923789083, "grad_norm": 1.9220101833343506, "learning_rate": 3.666297946225281e-06, "loss": 0.2381, "step": 38890 }, { "epoch": 0.7075428800087306, "grad_norm": 1.494392991065979, "learning_rate": 3.6643845731892118e-06, "loss": 0.2233, "step": 38900 }, { "epoch": 0.7077247676385529, "grad_norm": 1.7012555599212646, "learning_rate": 3.662472010692648e-06, "loss": 0.2263, "step": 38910 }, { "epoch": 0.7079066552683752, "grad_norm": 1.5258604288101196, "learning_rate": 3.6605602593993185e-06, "loss": 0.2308, "step": 38920 }, { "epoch": 0.7080885428981974, "grad_norm": 1.5098202228546143, "learning_rate": 3.6586493199726647e-06, "loss": 0.2267, "step": 38930 }, { "epoch": 0.7082704305280197, "grad_norm": 1.4895315170288086, "learning_rate": 3.656739193075849e-06, "loss": 0.2277, "step": 38940 }, { "epoch": 0.708452318157842, "grad_norm": 1.5992547273635864, "learning_rate": 3.654829879371753e-06, "loss": 0.233, "step": 38950 }, { "epoch": 0.7086342057876643, "grad_norm": 1.7126734256744385, "learning_rate": 3.6529213795229712e-06, "loss": 0.2314, "step": 38960 }, { "epoch": 0.7088160934174866, "grad_norm": 1.6101570129394531, "learning_rate": 3.651013694191824e-06, "loss": 0.2307, "step": 38970 }, { "epoch": 0.7089979810473089, "grad_norm": 1.6353055238723755, "learning_rate": 3.6491068240403418e-06, "loss": 0.2316, "step": 38980 }, { "epoch": 0.7091798686771312, "grad_norm": 1.854366421699524, "learning_rate": 3.6472007697302743e-06, "loss": 0.2266, "step": 38990 }, { "epoch": 0.7093617563069535, "grad_norm": 1.5020897388458252, "learning_rate": 3.6452955319230877e-06, "loss": 0.2321, "step": 39000 }, { "epoch": 0.7095436439367758, "grad_norm": 1.402462363243103, "learning_rate": 3.643391111279968e-06, "loss": 0.2298, "step": 39010 }, { "epoch": 0.7097255315665981, "grad_norm": 1.4448200464248657, "learning_rate": 3.6414875084618153e-06, "loss": 0.2312, "step": 39020 }, { "epoch": 0.7099074191964204, "grad_norm": 1.6857494115829468, "learning_rate": 3.6395847241292436e-06, "loss": 0.2326, "step": 39030 }, { "epoch": 0.7100893068262427, "grad_norm": 1.3424025774002075, "learning_rate": 3.6376827589425857e-06, "loss": 0.2293, "step": 39040 }, { "epoch": 0.710271194456065, "grad_norm": 1.5388158559799194, "learning_rate": 3.6357816135618883e-06, "loss": 0.236, "step": 39050 }, { "epoch": 0.7104530820858873, "grad_norm": 1.438011884689331, "learning_rate": 3.6338812886469175e-06, "loss": 0.2287, "step": 39060 }, { "epoch": 0.7106349697157096, "grad_norm": 1.7384997606277466, "learning_rate": 3.6319817848571504e-06, "loss": 0.2356, "step": 39070 }, { "epoch": 0.7108168573455319, "grad_norm": 1.525881052017212, "learning_rate": 3.6300831028517815e-06, "loss": 0.2298, "step": 39080 }, { "epoch": 0.7109987449753542, "grad_norm": 1.494971752166748, "learning_rate": 3.628185243289718e-06, "loss": 0.233, "step": 39090 }, { "epoch": 0.7111806326051765, "grad_norm": 1.6810957193374634, "learning_rate": 3.6262882068295816e-06, "loss": 0.2272, "step": 39100 }, { "epoch": 0.7113625202349988, "grad_norm": 1.4338643550872803, "learning_rate": 3.6243919941297133e-06, "loss": 0.2251, "step": 39110 }, { "epoch": 0.7115444078648211, "grad_norm": 1.5483462810516357, "learning_rate": 3.6224966058481624e-06, "loss": 0.2286, "step": 39120 }, { "epoch": 0.7117262954946434, "grad_norm": 1.4004921913146973, "learning_rate": 3.6206020426426957e-06, "loss": 0.2318, "step": 39130 }, { "epoch": 0.7119081831244657, "grad_norm": 1.5195618867874146, "learning_rate": 3.6187083051707895e-06, "loss": 0.2287, "step": 39140 }, { "epoch": 0.712090070754288, "grad_norm": 1.580500602722168, "learning_rate": 3.61681539408964e-06, "loss": 0.2363, "step": 39150 }, { "epoch": 0.7122719583841103, "grad_norm": 1.4191662073135376, "learning_rate": 3.6149233100561498e-06, "loss": 0.2336, "step": 39160 }, { "epoch": 0.7124538460139326, "grad_norm": 1.6254645586013794, "learning_rate": 3.613032053726939e-06, "loss": 0.236, "step": 39170 }, { "epoch": 0.7126357336437549, "grad_norm": 1.7567315101623535, "learning_rate": 3.611141625758341e-06, "loss": 0.2347, "step": 39180 }, { "epoch": 0.7128176212735772, "grad_norm": 1.8064606189727783, "learning_rate": 3.6092520268063975e-06, "loss": 0.2273, "step": 39190 }, { "epoch": 0.7129995089033995, "grad_norm": 1.5686769485473633, "learning_rate": 3.6073632575268634e-06, "loss": 0.2348, "step": 39200 }, { "epoch": 0.7131813965332218, "grad_norm": 1.3668303489685059, "learning_rate": 3.6054753185752104e-06, "loss": 0.2327, "step": 39210 }, { "epoch": 0.7133632841630441, "grad_norm": 1.434196949005127, "learning_rate": 3.603588210606618e-06, "loss": 0.2311, "step": 39220 }, { "epoch": 0.7135451717928664, "grad_norm": 1.6092379093170166, "learning_rate": 3.6017019342759775e-06, "loss": 0.2286, "step": 39230 }, { "epoch": 0.7137270594226887, "grad_norm": 1.639337420463562, "learning_rate": 3.5998164902378906e-06, "loss": 0.2301, "step": 39240 }, { "epoch": 0.713908947052511, "grad_norm": 1.7780754566192627, "learning_rate": 3.597931879146673e-06, "loss": 0.2311, "step": 39250 }, { "epoch": 0.7140908346823333, "grad_norm": 1.6741893291473389, "learning_rate": 3.596048101656352e-06, "loss": 0.2308, "step": 39260 }, { "epoch": 0.7142727223121556, "grad_norm": 1.6602381467819214, "learning_rate": 3.5941651584206614e-06, "loss": 0.2254, "step": 39270 }, { "epoch": 0.7144546099419778, "grad_norm": 1.7589701414108276, "learning_rate": 3.5922830500930483e-06, "loss": 0.2283, "step": 39280 }, { "epoch": 0.7146364975718001, "grad_norm": 1.4406177997589111, "learning_rate": 3.5904017773266696e-06, "loss": 0.2275, "step": 39290 }, { "epoch": 0.7148183852016224, "grad_norm": 1.4541252851486206, "learning_rate": 3.5885213407743913e-06, "loss": 0.2278, "step": 39300 }, { "epoch": 0.7150002728314447, "grad_norm": 1.682936429977417, "learning_rate": 3.5866417410887917e-06, "loss": 0.2299, "step": 39310 }, { "epoch": 0.715182160461267, "grad_norm": 1.3597909212112427, "learning_rate": 3.584762978922156e-06, "loss": 0.2313, "step": 39320 }, { "epoch": 0.7153640480910893, "grad_norm": 1.6411367654800415, "learning_rate": 3.582885054926481e-06, "loss": 0.2281, "step": 39330 }, { "epoch": 0.7155459357209116, "grad_norm": 1.4917837381362915, "learning_rate": 3.5810079697534694e-06, "loss": 0.2295, "step": 39340 }, { "epoch": 0.7157278233507339, "grad_norm": 1.7936241626739502, "learning_rate": 3.5791317240545354e-06, "loss": 0.2277, "step": 39350 }, { "epoch": 0.7159097109805562, "grad_norm": 1.621938943862915, "learning_rate": 3.5772563184808028e-06, "loss": 0.2274, "step": 39360 }, { "epoch": 0.7160915986103785, "grad_norm": 1.6548649072647095, "learning_rate": 3.5753817536831e-06, "loss": 0.2289, "step": 39370 }, { "epoch": 0.7162734862402008, "grad_norm": 1.831284999847412, "learning_rate": 3.573508030311969e-06, "loss": 0.2397, "step": 39380 }, { "epoch": 0.7164553738700231, "grad_norm": 1.5375460386276245, "learning_rate": 3.5716351490176528e-06, "loss": 0.2361, "step": 39390 }, { "epoch": 0.7166372614998454, "grad_norm": 1.7184157371520996, "learning_rate": 3.5697631104501103e-06, "loss": 0.2277, "step": 39400 }, { "epoch": 0.7168191491296677, "grad_norm": 1.6375689506530762, "learning_rate": 3.5678919152589996e-06, "loss": 0.2244, "step": 39410 }, { "epoch": 0.71700103675949, "grad_norm": 1.5084221363067627, "learning_rate": 3.5660215640936947e-06, "loss": 0.2361, "step": 39420 }, { "epoch": 0.7171829243893123, "grad_norm": 1.481050968170166, "learning_rate": 3.564152057603269e-06, "loss": 0.2314, "step": 39430 }, { "epoch": 0.7173648120191346, "grad_norm": 1.4542697668075562, "learning_rate": 3.562283396436508e-06, "loss": 0.2317, "step": 39440 }, { "epoch": 0.7175466996489569, "grad_norm": 1.7269039154052734, "learning_rate": 3.5604155812418985e-06, "loss": 0.2289, "step": 39450 }, { "epoch": 0.7177285872787792, "grad_norm": 1.5469119548797607, "learning_rate": 3.558548612667642e-06, "loss": 0.2371, "step": 39460 }, { "epoch": 0.7179104749086015, "grad_norm": 1.6395949125289917, "learning_rate": 3.556682491361638e-06, "loss": 0.2316, "step": 39470 }, { "epoch": 0.7180923625384238, "grad_norm": 1.542067527770996, "learning_rate": 3.554817217971497e-06, "loss": 0.227, "step": 39480 }, { "epoch": 0.7182742501682461, "grad_norm": 1.808902621269226, "learning_rate": 3.552952793144533e-06, "loss": 0.2271, "step": 39490 }, { "epoch": 0.7184561377980684, "grad_norm": 1.6665210723876953, "learning_rate": 3.5510892175277633e-06, "loss": 0.2266, "step": 39500 }, { "epoch": 0.7186380254278907, "grad_norm": 1.465611457824707, "learning_rate": 3.5492264917679183e-06, "loss": 0.234, "step": 39510 }, { "epoch": 0.718819913057713, "grad_norm": 1.672702670097351, "learning_rate": 3.547364616511425e-06, "loss": 0.2297, "step": 39520 }, { "epoch": 0.7190018006875353, "grad_norm": 1.5185060501098633, "learning_rate": 3.5455035924044194e-06, "loss": 0.227, "step": 39530 }, { "epoch": 0.7191836883173576, "grad_norm": 1.3734583854675293, "learning_rate": 3.543643420092742e-06, "loss": 0.2281, "step": 39540 }, { "epoch": 0.7193655759471799, "grad_norm": 1.4581687450408936, "learning_rate": 3.5417841002219337e-06, "loss": 0.228, "step": 39550 }, { "epoch": 0.7195474635770022, "grad_norm": 1.5872080326080322, "learning_rate": 3.5399256334372485e-06, "loss": 0.2247, "step": 39560 }, { "epoch": 0.7197293512068245, "grad_norm": 1.6470266580581665, "learning_rate": 3.5380680203836353e-06, "loss": 0.2313, "step": 39570 }, { "epoch": 0.7199112388366468, "grad_norm": 1.7049634456634521, "learning_rate": 3.536211261705751e-06, "loss": 0.2334, "step": 39580 }, { "epoch": 0.7200931264664691, "grad_norm": 1.6086091995239258, "learning_rate": 3.534355358047954e-06, "loss": 0.2261, "step": 39590 }, { "epoch": 0.7202750140962914, "grad_norm": 1.6728771924972534, "learning_rate": 3.5325003100543093e-06, "loss": 0.2336, "step": 39600 }, { "epoch": 0.7204569017261137, "grad_norm": 1.4861699342727661, "learning_rate": 3.5306461183685812e-06, "loss": 0.2335, "step": 39610 }, { "epoch": 0.720638789355936, "grad_norm": 1.6889337301254272, "learning_rate": 3.5287927836342407e-06, "loss": 0.2333, "step": 39620 }, { "epoch": 0.7208206769857582, "grad_norm": 1.6567864418029785, "learning_rate": 3.5269403064944567e-06, "loss": 0.2301, "step": 39630 }, { "epoch": 0.7210025646155805, "grad_norm": 1.689330816268921, "learning_rate": 3.525088687592103e-06, "loss": 0.2268, "step": 39640 }, { "epoch": 0.7211844522454028, "grad_norm": 1.5431499481201172, "learning_rate": 3.5232379275697576e-06, "loss": 0.237, "step": 39650 }, { "epoch": 0.7213663398752251, "grad_norm": 1.705371618270874, "learning_rate": 3.5213880270696974e-06, "loss": 0.2377, "step": 39660 }, { "epoch": 0.7215482275050474, "grad_norm": 1.4513598680496216, "learning_rate": 3.5195389867339012e-06, "loss": 0.2286, "step": 39670 }, { "epoch": 0.7217301151348697, "grad_norm": 1.6676141023635864, "learning_rate": 3.5176908072040515e-06, "loss": 0.2276, "step": 39680 }, { "epoch": 0.7219120027646919, "grad_norm": 1.5901137590408325, "learning_rate": 3.5158434891215276e-06, "loss": 0.2331, "step": 39690 }, { "epoch": 0.7220938903945142, "grad_norm": 1.4521788358688354, "learning_rate": 3.513997033127416e-06, "loss": 0.2222, "step": 39700 }, { "epoch": 0.7222757780243365, "grad_norm": 1.629901647567749, "learning_rate": 3.512151439862501e-06, "loss": 0.2292, "step": 39710 }, { "epoch": 0.7224576656541588, "grad_norm": 1.6629183292388916, "learning_rate": 3.5103067099672673e-06, "loss": 0.2328, "step": 39720 }, { "epoch": 0.7226395532839811, "grad_norm": 1.7505450248718262, "learning_rate": 3.5084628440818978e-06, "loss": 0.2405, "step": 39730 }, { "epoch": 0.7228214409138034, "grad_norm": 1.7187684774398804, "learning_rate": 3.5066198428462793e-06, "loss": 0.2298, "step": 39740 }, { "epoch": 0.7230033285436257, "grad_norm": 1.5312607288360596, "learning_rate": 3.5047777068999976e-06, "loss": 0.2283, "step": 39750 }, { "epoch": 0.723185216173448, "grad_norm": 1.5967048406600952, "learning_rate": 3.502936436882338e-06, "loss": 0.2294, "step": 39760 }, { "epoch": 0.7233671038032703, "grad_norm": 1.3968302011489868, "learning_rate": 3.5010960334322847e-06, "loss": 0.2291, "step": 39770 }, { "epoch": 0.7235489914330926, "grad_norm": 1.5237973928451538, "learning_rate": 3.4992564971885205e-06, "loss": 0.2332, "step": 39780 }, { "epoch": 0.7237308790629149, "grad_norm": 1.773962378501892, "learning_rate": 3.497417828789428e-06, "loss": 0.2322, "step": 39790 }, { "epoch": 0.7239127666927372, "grad_norm": 1.495857834815979, "learning_rate": 3.495580028873091e-06, "loss": 0.2301, "step": 39800 }, { "epoch": 0.7240946543225595, "grad_norm": 1.7504196166992188, "learning_rate": 3.4937430980772883e-06, "loss": 0.2324, "step": 39810 }, { "epoch": 0.7242765419523818, "grad_norm": 1.5272889137268066, "learning_rate": 3.4919070370394992e-06, "loss": 0.229, "step": 39820 }, { "epoch": 0.7244584295822041, "grad_norm": 1.8057384490966797, "learning_rate": 3.4900718463968997e-06, "loss": 0.2262, "step": 39830 }, { "epoch": 0.7246403172120264, "grad_norm": 1.6802756786346436, "learning_rate": 3.4882375267863632e-06, "loss": 0.2339, "step": 39840 }, { "epoch": 0.7248222048418487, "grad_norm": 1.5091272592544556, "learning_rate": 3.4864040788444664e-06, "loss": 0.2301, "step": 39850 }, { "epoch": 0.725004092471671, "grad_norm": 1.4708820581436157, "learning_rate": 3.4845715032074758e-06, "loss": 0.2245, "step": 39860 }, { "epoch": 0.7251859801014933, "grad_norm": 1.7989814281463623, "learning_rate": 3.48273980051136e-06, "loss": 0.2244, "step": 39870 }, { "epoch": 0.7253678677313156, "grad_norm": 1.9554866552352905, "learning_rate": 3.4809089713917833e-06, "loss": 0.2293, "step": 39880 }, { "epoch": 0.7255497553611379, "grad_norm": 1.8624473810195923, "learning_rate": 3.4790790164841047e-06, "loss": 0.2273, "step": 39890 }, { "epoch": 0.7257316429909602, "grad_norm": 1.4795527458190918, "learning_rate": 3.4772499364233854e-06, "loss": 0.2258, "step": 39900 }, { "epoch": 0.7259135306207825, "grad_norm": 1.6969804763793945, "learning_rate": 3.4754217318443783e-06, "loss": 0.2307, "step": 39910 }, { "epoch": 0.7260954182506048, "grad_norm": 1.4941378831863403, "learning_rate": 3.4735944033815327e-06, "loss": 0.225, "step": 39920 }, { "epoch": 0.726277305880427, "grad_norm": 1.5829346179962158, "learning_rate": 3.471767951668995e-06, "loss": 0.2232, "step": 39930 }, { "epoch": 0.7264591935102493, "grad_norm": 1.7361671924591064, "learning_rate": 3.469942377340609e-06, "loss": 0.2296, "step": 39940 }, { "epoch": 0.7266410811400716, "grad_norm": 1.625732421875, "learning_rate": 3.4681176810299112e-06, "loss": 0.2314, "step": 39950 }, { "epoch": 0.7268229687698939, "grad_norm": 1.5632359981536865, "learning_rate": 3.4662938633701335e-06, "loss": 0.229, "step": 39960 }, { "epoch": 0.7270048563997162, "grad_norm": 1.695904016494751, "learning_rate": 3.4644709249942054e-06, "loss": 0.2261, "step": 39970 }, { "epoch": 0.7271867440295385, "grad_norm": 1.554611086845398, "learning_rate": 3.4626488665347502e-06, "loss": 0.2297, "step": 39980 }, { "epoch": 0.7273686316593608, "grad_norm": 1.9006733894348145, "learning_rate": 3.460827688624082e-06, "loss": 0.2305, "step": 39990 }, { "epoch": 0.7275505192891831, "grad_norm": 1.8506693840026855, "learning_rate": 3.459007391894217e-06, "loss": 0.2269, "step": 40000 }, { "epoch": 0.7277324069190054, "grad_norm": 1.7013295888900757, "learning_rate": 3.457187976976859e-06, "loss": 0.2258, "step": 40010 }, { "epoch": 0.7279142945488277, "grad_norm": 1.6095820665359497, "learning_rate": 3.455369444503408e-06, "loss": 0.2303, "step": 40020 }, { "epoch": 0.72809618217865, "grad_norm": 1.6976666450500488, "learning_rate": 3.453551795104958e-06, "loss": 0.2325, "step": 40030 }, { "epoch": 0.7282780698084723, "grad_norm": 1.576622724533081, "learning_rate": 3.451735029412295e-06, "loss": 0.2324, "step": 40040 }, { "epoch": 0.7284599574382946, "grad_norm": 1.712110161781311, "learning_rate": 3.4499191480559023e-06, "loss": 0.2293, "step": 40050 }, { "epoch": 0.7286418450681169, "grad_norm": 1.5115524530410767, "learning_rate": 3.448104151665952e-06, "loss": 0.2317, "step": 40060 }, { "epoch": 0.7288237326979392, "grad_norm": 1.4612706899642944, "learning_rate": 3.446290040872312e-06, "loss": 0.2278, "step": 40070 }, { "epoch": 0.7290056203277615, "grad_norm": 1.53774094581604, "learning_rate": 3.4444768163045405e-06, "loss": 0.2264, "step": 40080 }, { "epoch": 0.7291875079575838, "grad_norm": 1.5360826253890991, "learning_rate": 3.4426644785918878e-06, "loss": 0.2353, "step": 40090 }, { "epoch": 0.7293693955874061, "grad_norm": 1.574073076248169, "learning_rate": 3.440853028363301e-06, "loss": 0.234, "step": 40100 }, { "epoch": 0.7295512832172284, "grad_norm": 1.7739295959472656, "learning_rate": 3.4390424662474155e-06, "loss": 0.2296, "step": 40110 }, { "epoch": 0.7297331708470507, "grad_norm": 1.5634113550186157, "learning_rate": 3.4372327928725573e-06, "loss": 0.2238, "step": 40120 }, { "epoch": 0.729915058476873, "grad_norm": 1.4446649551391602, "learning_rate": 3.435424008866747e-06, "loss": 0.2257, "step": 40130 }, { "epoch": 0.7300969461066953, "grad_norm": 1.7083696126937866, "learning_rate": 3.4336161148576952e-06, "loss": 0.2315, "step": 40140 }, { "epoch": 0.7302788337365176, "grad_norm": 1.4161996841430664, "learning_rate": 3.4318091114728037e-06, "loss": 0.2269, "step": 40150 }, { "epoch": 0.7304607213663399, "grad_norm": 1.4343737363815308, "learning_rate": 3.4300029993391657e-06, "loss": 0.2266, "step": 40160 }, { "epoch": 0.7306426089961622, "grad_norm": 1.5869747400283813, "learning_rate": 3.428197779083565e-06, "loss": 0.2363, "step": 40170 }, { "epoch": 0.7308244966259845, "grad_norm": 1.499780535697937, "learning_rate": 3.4263934513324736e-06, "loss": 0.2328, "step": 40180 }, { "epoch": 0.7310063842558068, "grad_norm": 1.70485520362854, "learning_rate": 3.4245900167120582e-06, "loss": 0.2314, "step": 40190 }, { "epoch": 0.7311882718856291, "grad_norm": 1.72396981716156, "learning_rate": 3.4227874758481714e-06, "loss": 0.2307, "step": 40200 }, { "epoch": 0.7313701595154514, "grad_norm": 1.6176515817642212, "learning_rate": 3.420985829366359e-06, "loss": 0.2307, "step": 40210 }, { "epoch": 0.7315520471452737, "grad_norm": 1.6999177932739258, "learning_rate": 3.4191850778918535e-06, "loss": 0.2341, "step": 40220 }, { "epoch": 0.731733934775096, "grad_norm": 1.5535697937011719, "learning_rate": 3.417385222049576e-06, "loss": 0.233, "step": 40230 }, { "epoch": 0.7319158224049183, "grad_norm": 1.6614552736282349, "learning_rate": 3.415586262464143e-06, "loss": 0.2295, "step": 40240 }, { "epoch": 0.7320977100347406, "grad_norm": 1.8832679986953735, "learning_rate": 3.4137881997598545e-06, "loss": 0.2286, "step": 40250 }, { "epoch": 0.7322795976645629, "grad_norm": 1.499391794204712, "learning_rate": 3.4119910345607e-06, "loss": 0.2266, "step": 40260 }, { "epoch": 0.7324614852943852, "grad_norm": 1.3972996473312378, "learning_rate": 3.410194767490359e-06, "loss": 0.2257, "step": 40270 }, { "epoch": 0.7326433729242074, "grad_norm": 1.6804245710372925, "learning_rate": 3.4083993991721974e-06, "loss": 0.2286, "step": 40280 }, { "epoch": 0.7328252605540297, "grad_norm": 1.387999176979065, "learning_rate": 3.4066049302292694e-06, "loss": 0.2333, "step": 40290 }, { "epoch": 0.733007148183852, "grad_norm": 2.3093490600585938, "learning_rate": 3.40481136128432e-06, "loss": 0.2284, "step": 40300 }, { "epoch": 0.7331890358136743, "grad_norm": 1.7061376571655273, "learning_rate": 3.4030186929597784e-06, "loss": 0.2313, "step": 40310 }, { "epoch": 0.7333709234434966, "grad_norm": 1.5047361850738525, "learning_rate": 3.401226925877764e-06, "loss": 0.2207, "step": 40320 }, { "epoch": 0.7335528110733189, "grad_norm": 1.3827682733535767, "learning_rate": 3.3994360606600796e-06, "loss": 0.2285, "step": 40330 }, { "epoch": 0.7337346987031412, "grad_norm": 1.7182167768478394, "learning_rate": 3.39764609792822e-06, "loss": 0.229, "step": 40340 }, { "epoch": 0.7339165863329635, "grad_norm": 1.5398876667022705, "learning_rate": 3.395857038303364e-06, "loss": 0.2262, "step": 40350 }, { "epoch": 0.7340984739627858, "grad_norm": 1.5493121147155762, "learning_rate": 3.394068882406376e-06, "loss": 0.2269, "step": 40360 }, { "epoch": 0.7342803615926081, "grad_norm": 1.4677048921585083, "learning_rate": 3.39228163085781e-06, "loss": 0.2284, "step": 40370 }, { "epoch": 0.7344622492224304, "grad_norm": 1.411793828010559, "learning_rate": 3.3904952842779003e-06, "loss": 0.2269, "step": 40380 }, { "epoch": 0.7346441368522527, "grad_norm": 1.3949700593948364, "learning_rate": 3.3887098432865763e-06, "loss": 0.2236, "step": 40390 }, { "epoch": 0.734826024482075, "grad_norm": 1.688219666481018, "learning_rate": 3.3869253085034447e-06, "loss": 0.2252, "step": 40400 }, { "epoch": 0.7350079121118973, "grad_norm": 1.5014920234680176, "learning_rate": 3.385141680547801e-06, "loss": 0.2275, "step": 40410 }, { "epoch": 0.7351897997417196, "grad_norm": 1.7063504457473755, "learning_rate": 3.3833589600386275e-06, "loss": 0.23, "step": 40420 }, { "epoch": 0.7353716873715419, "grad_norm": 1.4360511302947998, "learning_rate": 3.3815771475945876e-06, "loss": 0.2263, "step": 40430 }, { "epoch": 0.7355535750013642, "grad_norm": 1.575986385345459, "learning_rate": 3.3797962438340346e-06, "loss": 0.2286, "step": 40440 }, { "epoch": 0.7357354626311865, "grad_norm": 1.7417843341827393, "learning_rate": 3.3780162493750023e-06, "loss": 0.232, "step": 40450 }, { "epoch": 0.7359173502610088, "grad_norm": 1.46807062625885, "learning_rate": 3.3762371648352114e-06, "loss": 0.2211, "step": 40460 }, { "epoch": 0.7360992378908311, "grad_norm": 1.7428525686264038, "learning_rate": 3.374458990832065e-06, "loss": 0.2352, "step": 40470 }, { "epoch": 0.7362811255206534, "grad_norm": 1.5470601320266724, "learning_rate": 3.3726817279826508e-06, "loss": 0.2257, "step": 40480 }, { "epoch": 0.7364630131504757, "grad_norm": 1.7576338052749634, "learning_rate": 3.3709053769037416e-06, "loss": 0.2298, "step": 40490 }, { "epoch": 0.736644900780298, "grad_norm": 1.9106241464614868, "learning_rate": 3.3691299382117926e-06, "loss": 0.231, "step": 40500 }, { "epoch": 0.7368267884101203, "grad_norm": 1.7451236248016357, "learning_rate": 3.3673554125229422e-06, "loss": 0.2342, "step": 40510 }, { "epoch": 0.7370086760399426, "grad_norm": 1.6848697662353516, "learning_rate": 3.3655818004530115e-06, "loss": 0.2278, "step": 40520 }, { "epoch": 0.7371905636697648, "grad_norm": 1.437981128692627, "learning_rate": 3.363809102617507e-06, "loss": 0.2353, "step": 40530 }, { "epoch": 0.7373724512995871, "grad_norm": 1.4890915155410767, "learning_rate": 3.3620373196316157e-06, "loss": 0.2304, "step": 40540 }, { "epoch": 0.7375543389294094, "grad_norm": 1.552687644958496, "learning_rate": 3.3602664521102093e-06, "loss": 0.2247, "step": 40550 }, { "epoch": 0.7377362265592317, "grad_norm": 1.5760536193847656, "learning_rate": 3.3584965006678373e-06, "loss": 0.2301, "step": 40560 }, { "epoch": 0.737918114189054, "grad_norm": 1.537656307220459, "learning_rate": 3.356727465918736e-06, "loss": 0.2271, "step": 40570 }, { "epoch": 0.7381000018188763, "grad_norm": 1.786529779434204, "learning_rate": 3.3549593484768194e-06, "loss": 0.2298, "step": 40580 }, { "epoch": 0.7382818894486985, "grad_norm": 1.9031106233596802, "learning_rate": 3.3531921489556897e-06, "loss": 0.2297, "step": 40590 }, { "epoch": 0.7384637770785208, "grad_norm": 1.9330304861068726, "learning_rate": 3.351425867968623e-06, "loss": 0.2363, "step": 40600 }, { "epoch": 0.7386456647083431, "grad_norm": 1.7010632753372192, "learning_rate": 3.349660506128582e-06, "loss": 0.2265, "step": 40610 }, { "epoch": 0.7388275523381654, "grad_norm": 1.6734060049057007, "learning_rate": 3.347896064048207e-06, "loss": 0.2261, "step": 40620 }, { "epoch": 0.7390094399679877, "grad_norm": 1.7370610237121582, "learning_rate": 3.3461325423398196e-06, "loss": 0.2259, "step": 40630 }, { "epoch": 0.73919132759781, "grad_norm": 1.7962417602539062, "learning_rate": 3.3443699416154264e-06, "loss": 0.2313, "step": 40640 }, { "epoch": 0.7393732152276323, "grad_norm": 1.4235504865646362, "learning_rate": 3.3426082624867086e-06, "loss": 0.2272, "step": 40650 }, { "epoch": 0.7395551028574546, "grad_norm": 1.3165321350097656, "learning_rate": 3.3408475055650304e-06, "loss": 0.2284, "step": 40660 }, { "epoch": 0.7397369904872769, "grad_norm": 1.6219496726989746, "learning_rate": 3.339087671461436e-06, "loss": 0.2292, "step": 40670 }, { "epoch": 0.7399188781170992, "grad_norm": 1.9441642761230469, "learning_rate": 3.337328760786646e-06, "loss": 0.2363, "step": 40680 }, { "epoch": 0.7401007657469215, "grad_norm": 1.5288000106811523, "learning_rate": 3.335570774151068e-06, "loss": 0.2221, "step": 40690 }, { "epoch": 0.7402826533767438, "grad_norm": 1.7903755903244019, "learning_rate": 3.333813712164781e-06, "loss": 0.2338, "step": 40700 }, { "epoch": 0.7404645410065661, "grad_norm": 1.69257652759552, "learning_rate": 3.3320575754375473e-06, "loss": 0.2266, "step": 40710 }, { "epoch": 0.7406464286363884, "grad_norm": 1.6193393468856812, "learning_rate": 3.330302364578807e-06, "loss": 0.233, "step": 40720 }, { "epoch": 0.7408283162662107, "grad_norm": 1.5101438760757446, "learning_rate": 3.328548080197679e-06, "loss": 0.2284, "step": 40730 }, { "epoch": 0.741010203896033, "grad_norm": 1.7546061277389526, "learning_rate": 3.326794722902962e-06, "loss": 0.2307, "step": 40740 }, { "epoch": 0.7411920915258553, "grad_norm": 1.7245675325393677, "learning_rate": 3.3250422933031314e-06, "loss": 0.2326, "step": 40750 }, { "epoch": 0.7413739791556776, "grad_norm": 1.5627977848052979, "learning_rate": 3.3232907920063397e-06, "loss": 0.2266, "step": 40760 }, { "epoch": 0.7415558667854999, "grad_norm": 1.4662491083145142, "learning_rate": 3.3215402196204184e-06, "loss": 0.2268, "step": 40770 }, { "epoch": 0.7417377544153222, "grad_norm": 1.611161231994629, "learning_rate": 3.319790576752878e-06, "loss": 0.2271, "step": 40780 }, { "epoch": 0.7419196420451445, "grad_norm": 1.4936320781707764, "learning_rate": 3.318041864010905e-06, "loss": 0.2311, "step": 40790 }, { "epoch": 0.7421015296749668, "grad_norm": 1.534979224205017, "learning_rate": 3.3162940820013644e-06, "loss": 0.2305, "step": 40800 }, { "epoch": 0.7422834173047891, "grad_norm": 1.649536371231079, "learning_rate": 3.3145472313307957e-06, "loss": 0.2296, "step": 40810 }, { "epoch": 0.7424653049346114, "grad_norm": 1.7187817096710205, "learning_rate": 3.312801312605417e-06, "loss": 0.2263, "step": 40820 }, { "epoch": 0.7426471925644337, "grad_norm": 1.643691062927246, "learning_rate": 3.3110563264311213e-06, "loss": 0.2291, "step": 40830 }, { "epoch": 0.742829080194256, "grad_norm": 1.6062812805175781, "learning_rate": 3.309312273413482e-06, "loss": 0.2296, "step": 40840 }, { "epoch": 0.7430109678240783, "grad_norm": 1.5221525430679321, "learning_rate": 3.307569154157744e-06, "loss": 0.2259, "step": 40850 }, { "epoch": 0.7431928554539006, "grad_norm": 1.721736192703247, "learning_rate": 3.3058269692688305e-06, "loss": 0.2233, "step": 40860 }, { "epoch": 0.7433747430837229, "grad_norm": 1.3326228857040405, "learning_rate": 3.304085719351341e-06, "loss": 0.2243, "step": 40870 }, { "epoch": 0.7435566307135452, "grad_norm": 1.8250069618225098, "learning_rate": 3.3023454050095473e-06, "loss": 0.2299, "step": 40880 }, { "epoch": 0.7437385183433675, "grad_norm": 1.678166151046753, "learning_rate": 3.3006060268474004e-06, "loss": 0.2252, "step": 40890 }, { "epoch": 0.7439204059731898, "grad_norm": 1.744805097579956, "learning_rate": 3.298867585468525e-06, "loss": 0.2296, "step": 40900 }, { "epoch": 0.744102293603012, "grad_norm": 1.8262431621551514, "learning_rate": 3.297130081476221e-06, "loss": 0.2243, "step": 40910 }, { "epoch": 0.7442841812328344, "grad_norm": 1.515555739402771, "learning_rate": 3.2953935154734594e-06, "loss": 0.2233, "step": 40920 }, { "epoch": 0.7444660688626566, "grad_norm": 1.606101155281067, "learning_rate": 3.2936578880628935e-06, "loss": 0.2281, "step": 40930 }, { "epoch": 0.744647956492479, "grad_norm": 1.5159915685653687, "learning_rate": 3.2919231998468425e-06, "loss": 0.2259, "step": 40940 }, { "epoch": 0.7448298441223012, "grad_norm": 1.5002261400222778, "learning_rate": 3.2901894514273055e-06, "loss": 0.2309, "step": 40950 }, { "epoch": 0.7450117317521235, "grad_norm": 1.8020626306533813, "learning_rate": 3.288456643405952e-06, "loss": 0.2257, "step": 40960 }, { "epoch": 0.7451936193819458, "grad_norm": 1.7683647871017456, "learning_rate": 3.2867247763841247e-06, "loss": 0.2236, "step": 40970 }, { "epoch": 0.7453755070117681, "grad_norm": 1.7142995595932007, "learning_rate": 3.2849938509628455e-06, "loss": 0.2279, "step": 40980 }, { "epoch": 0.7455573946415904, "grad_norm": 1.4909522533416748, "learning_rate": 3.283263867742804e-06, "loss": 0.2287, "step": 40990 }, { "epoch": 0.7457392822714127, "grad_norm": 1.6395071744918823, "learning_rate": 3.281534827324363e-06, "loss": 0.2183, "step": 41000 }, { "epoch": 0.745921169901235, "grad_norm": 1.4656262397766113, "learning_rate": 3.279806730307561e-06, "loss": 0.2284, "step": 41010 }, { "epoch": 0.7461030575310573, "grad_norm": 1.4541058540344238, "learning_rate": 3.278079577292105e-06, "loss": 0.2245, "step": 41020 }, { "epoch": 0.7462849451608796, "grad_norm": 1.6209406852722168, "learning_rate": 3.2763533688773808e-06, "loss": 0.2288, "step": 41030 }, { "epoch": 0.7464668327907019, "grad_norm": 1.4636616706848145, "learning_rate": 3.274628105662441e-06, "loss": 0.2264, "step": 41040 }, { "epoch": 0.7466487204205242, "grad_norm": 1.4808954000473022, "learning_rate": 3.2729037882460108e-06, "loss": 0.2297, "step": 41050 }, { "epoch": 0.7468306080503465, "grad_norm": 1.7775146961212158, "learning_rate": 3.2711804172264887e-06, "loss": 0.2241, "step": 41060 }, { "epoch": 0.7470124956801688, "grad_norm": 1.5317175388336182, "learning_rate": 3.2694579932019434e-06, "loss": 0.2252, "step": 41070 }, { "epoch": 0.7471943833099911, "grad_norm": 1.5781810283660889, "learning_rate": 3.267736516770116e-06, "loss": 0.2315, "step": 41080 }, { "epoch": 0.7473762709398134, "grad_norm": 1.448041558265686, "learning_rate": 3.2660159885284204e-06, "loss": 0.2262, "step": 41090 }, { "epoch": 0.7475581585696357, "grad_norm": 1.9008160829544067, "learning_rate": 3.264296409073937e-06, "loss": 0.2316, "step": 41100 }, { "epoch": 0.747740046199458, "grad_norm": 1.969994068145752, "learning_rate": 3.262577779003422e-06, "loss": 0.2301, "step": 41110 }, { "epoch": 0.7479219338292803, "grad_norm": 1.831913948059082, "learning_rate": 3.2608600989132966e-06, "loss": 0.2292, "step": 41120 }, { "epoch": 0.7481038214591026, "grad_norm": 1.7366083860397339, "learning_rate": 3.259143369399659e-06, "loss": 0.2286, "step": 41130 }, { "epoch": 0.7482857090889249, "grad_norm": 1.6025751829147339, "learning_rate": 3.2574275910582705e-06, "loss": 0.2272, "step": 41140 }, { "epoch": 0.7484675967187472, "grad_norm": 1.483383059501648, "learning_rate": 3.255712764484568e-06, "loss": 0.2251, "step": 41150 }, { "epoch": 0.7486494843485695, "grad_norm": 1.5069822072982788, "learning_rate": 3.253998890273653e-06, "loss": 0.225, "step": 41160 }, { "epoch": 0.7488313719783918, "grad_norm": 1.658043622970581, "learning_rate": 3.252285969020301e-06, "loss": 0.2242, "step": 41170 }, { "epoch": 0.7490132596082141, "grad_norm": 1.5943020582199097, "learning_rate": 3.2505740013189556e-06, "loss": 0.2238, "step": 41180 }, { "epoch": 0.7491951472380364, "grad_norm": 1.691799283027649, "learning_rate": 3.248862987763729e-06, "loss": 0.226, "step": 41190 }, { "epoch": 0.7493770348678587, "grad_norm": 1.705261468887329, "learning_rate": 3.2471529289484007e-06, "loss": 0.2262, "step": 41200 }, { "epoch": 0.749558922497681, "grad_norm": 1.558839201927185, "learning_rate": 3.2454438254664224e-06, "loss": 0.2276, "step": 41210 }, { "epoch": 0.7497408101275033, "grad_norm": 2.4446918964385986, "learning_rate": 3.2437356779109087e-06, "loss": 0.2235, "step": 41220 }, { "epoch": 0.7499226977573256, "grad_norm": 1.4466485977172852, "learning_rate": 3.24202848687465e-06, "loss": 0.2301, "step": 41230 }, { "epoch": 0.7501045853871479, "grad_norm": 1.6264324188232422, "learning_rate": 3.240322252950099e-06, "loss": 0.2298, "step": 41240 }, { "epoch": 0.7502864730169702, "grad_norm": 1.555150032043457, "learning_rate": 3.2386169767293783e-06, "loss": 0.224, "step": 41250 }, { "epoch": 0.7504683606467925, "grad_norm": 1.709165334701538, "learning_rate": 3.2369126588042775e-06, "loss": 0.229, "step": 41260 }, { "epoch": 0.7506502482766148, "grad_norm": 1.5412646532058716, "learning_rate": 3.235209299766253e-06, "loss": 0.2272, "step": 41270 }, { "epoch": 0.750832135906437, "grad_norm": 1.564255714416504, "learning_rate": 3.233506900206432e-06, "loss": 0.2284, "step": 41280 }, { "epoch": 0.7510140235362593, "grad_norm": 1.7024402618408203, "learning_rate": 3.231805460715605e-06, "loss": 0.2255, "step": 41290 }, { "epoch": 0.7511959111660816, "grad_norm": 1.682092308998108, "learning_rate": 3.2301049818842295e-06, "loss": 0.2248, "step": 41300 }, { "epoch": 0.7513777987959039, "grad_norm": 1.6049907207489014, "learning_rate": 3.2284054643024304e-06, "loss": 0.2252, "step": 41310 }, { "epoch": 0.7515596864257262, "grad_norm": 1.800464153289795, "learning_rate": 3.226706908560001e-06, "loss": 0.2273, "step": 41320 }, { "epoch": 0.7517415740555485, "grad_norm": 1.7059122323989868, "learning_rate": 3.225009315246399e-06, "loss": 0.2287, "step": 41330 }, { "epoch": 0.7519234616853708, "grad_norm": 1.4779480695724487, "learning_rate": 3.2233126849507458e-06, "loss": 0.235, "step": 41340 }, { "epoch": 0.7521053493151931, "grad_norm": 1.8433444499969482, "learning_rate": 3.221617018261833e-06, "loss": 0.2293, "step": 41350 }, { "epoch": 0.7522872369450153, "grad_norm": 1.9221056699752808, "learning_rate": 3.2199223157681153e-06, "loss": 0.2239, "step": 41360 }, { "epoch": 0.7524691245748376, "grad_norm": 1.6857057809829712, "learning_rate": 3.2182285780577127e-06, "loss": 0.2283, "step": 41370 }, { "epoch": 0.7526510122046599, "grad_norm": 1.3920960426330566, "learning_rate": 3.2165358057184116e-06, "loss": 0.2187, "step": 41380 }, { "epoch": 0.7528328998344822, "grad_norm": 1.6436946392059326, "learning_rate": 3.214843999337662e-06, "loss": 0.2257, "step": 41390 }, { "epoch": 0.7530147874643045, "grad_norm": 1.5145782232284546, "learning_rate": 3.21315315950258e-06, "loss": 0.2274, "step": 41400 }, { "epoch": 0.7531966750941268, "grad_norm": 1.7133100032806396, "learning_rate": 3.2114632867999452e-06, "loss": 0.2261, "step": 41410 }, { "epoch": 0.7533785627239491, "grad_norm": 1.4058102369308472, "learning_rate": 3.209774381816201e-06, "loss": 0.227, "step": 41420 }, { "epoch": 0.7535604503537714, "grad_norm": 1.5315316915512085, "learning_rate": 3.2080864451374583e-06, "loss": 0.2313, "step": 41430 }, { "epoch": 0.7537423379835937, "grad_norm": 1.5256694555282593, "learning_rate": 3.206399477349488e-06, "loss": 0.2261, "step": 41440 }, { "epoch": 0.753924225613416, "grad_norm": 1.728530764579773, "learning_rate": 3.2047134790377267e-06, "loss": 0.2308, "step": 41450 }, { "epoch": 0.7541061132432383, "grad_norm": 1.4194488525390625, "learning_rate": 3.2030284507872745e-06, "loss": 0.2261, "step": 41460 }, { "epoch": 0.7542880008730606, "grad_norm": 1.4520012140274048, "learning_rate": 3.2013443931828924e-06, "loss": 0.224, "step": 41470 }, { "epoch": 0.7544698885028829, "grad_norm": 1.44677734375, "learning_rate": 3.1996613068090114e-06, "loss": 0.2276, "step": 41480 }, { "epoch": 0.7546517761327052, "grad_norm": 1.5469682216644287, "learning_rate": 3.197979192249717e-06, "loss": 0.2264, "step": 41490 }, { "epoch": 0.7548336637625275, "grad_norm": 1.5177323818206787, "learning_rate": 3.1962980500887637e-06, "loss": 0.2263, "step": 41500 }, { "epoch": 0.7550155513923498, "grad_norm": 1.5385652780532837, "learning_rate": 3.1946178809095625e-06, "loss": 0.232, "step": 41510 }, { "epoch": 0.7551974390221721, "grad_norm": 1.705331802368164, "learning_rate": 3.1929386852951944e-06, "loss": 0.2311, "step": 41520 }, { "epoch": 0.7553793266519944, "grad_norm": 1.5453267097473145, "learning_rate": 3.191260463828396e-06, "loss": 0.2271, "step": 41530 }, { "epoch": 0.7555612142818167, "grad_norm": 1.5593972206115723, "learning_rate": 3.1895832170915697e-06, "loss": 0.2263, "step": 41540 }, { "epoch": 0.755743101911639, "grad_norm": 1.5889618396759033, "learning_rate": 3.1879069456667777e-06, "loss": 0.2275, "step": 41550 }, { "epoch": 0.7559249895414613, "grad_norm": 1.8623099327087402, "learning_rate": 3.1862316501357424e-06, "loss": 0.2271, "step": 41560 }, { "epoch": 0.7561068771712836, "grad_norm": 1.5764033794403076, "learning_rate": 3.1845573310798535e-06, "loss": 0.2281, "step": 41570 }, { "epoch": 0.7562887648011059, "grad_norm": 1.727687120437622, "learning_rate": 3.182883989080154e-06, "loss": 0.2352, "step": 41580 }, { "epoch": 0.7564706524309281, "grad_norm": 1.5287721157073975, "learning_rate": 3.1812116247173547e-06, "loss": 0.228, "step": 41590 }, { "epoch": 0.7566525400607504, "grad_norm": 1.6450011730194092, "learning_rate": 3.1795402385718212e-06, "loss": 0.2235, "step": 41600 }, { "epoch": 0.7568344276905727, "grad_norm": 1.5143601894378662, "learning_rate": 3.1778698312235824e-06, "loss": 0.2272, "step": 41610 }, { "epoch": 0.757016315320395, "grad_norm": 1.4348572492599487, "learning_rate": 3.176200403252329e-06, "loss": 0.2303, "step": 41620 }, { "epoch": 0.7571982029502173, "grad_norm": 1.5038989782333374, "learning_rate": 3.174531955237411e-06, "loss": 0.2276, "step": 41630 }, { "epoch": 0.7573800905800396, "grad_norm": 1.7140651941299438, "learning_rate": 3.1728644877578376e-06, "loss": 0.2364, "step": 41640 }, { "epoch": 0.7575619782098619, "grad_norm": 1.5472599267959595, "learning_rate": 3.1711980013922763e-06, "loss": 0.2289, "step": 41650 }, { "epoch": 0.7577438658396842, "grad_norm": 1.5463976860046387, "learning_rate": 3.169532496719056e-06, "loss": 0.2243, "step": 41660 }, { "epoch": 0.7579257534695065, "grad_norm": 1.779477834701538, "learning_rate": 3.1678679743161634e-06, "loss": 0.2288, "step": 41670 }, { "epoch": 0.7581076410993288, "grad_norm": 1.665790319442749, "learning_rate": 3.1662044347612485e-06, "loss": 0.2205, "step": 41680 }, { "epoch": 0.7582895287291511, "grad_norm": 1.7934366464614868, "learning_rate": 3.1645418786316164e-06, "loss": 0.2318, "step": 41690 }, { "epoch": 0.7584714163589734, "grad_norm": 1.8147112131118774, "learning_rate": 3.16288030650423e-06, "loss": 0.2275, "step": 41700 }, { "epoch": 0.7586533039887957, "grad_norm": 1.447569727897644, "learning_rate": 3.1612197189557124e-06, "loss": 0.2279, "step": 41710 }, { "epoch": 0.758835191618618, "grad_norm": 1.4215633869171143, "learning_rate": 3.1595601165623472e-06, "loss": 0.2188, "step": 41720 }, { "epoch": 0.7590170792484403, "grad_norm": 1.5483793020248413, "learning_rate": 3.1579014999000728e-06, "loss": 0.2244, "step": 41730 }, { "epoch": 0.7591989668782626, "grad_norm": 1.568713665008545, "learning_rate": 3.1562438695444865e-06, "loss": 0.223, "step": 41740 }, { "epoch": 0.7593808545080849, "grad_norm": 1.6501998901367188, "learning_rate": 3.1545872260708436e-06, "loss": 0.2162, "step": 41750 }, { "epoch": 0.7595627421379072, "grad_norm": 1.4736510515213013, "learning_rate": 3.1529315700540554e-06, "loss": 0.2233, "step": 41760 }, { "epoch": 0.7597446297677295, "grad_norm": 1.6534509658813477, "learning_rate": 3.151276902068695e-06, "loss": 0.2236, "step": 41770 }, { "epoch": 0.7599265173975518, "grad_norm": 1.5257247686386108, "learning_rate": 3.149623222688987e-06, "loss": 0.2256, "step": 41780 }, { "epoch": 0.7601084050273741, "grad_norm": 1.550372838973999, "learning_rate": 3.1479705324888172e-06, "loss": 0.2275, "step": 41790 }, { "epoch": 0.7602902926571964, "grad_norm": 1.7857885360717773, "learning_rate": 3.146318832041725e-06, "loss": 0.2228, "step": 41800 }, { "epoch": 0.7604721802870187, "grad_norm": 1.649653673171997, "learning_rate": 3.1446681219209065e-06, "loss": 0.2324, "step": 41810 }, { "epoch": 0.760654067916841, "grad_norm": 1.5172966718673706, "learning_rate": 3.1430184026992196e-06, "loss": 0.2242, "step": 41820 }, { "epoch": 0.7608359555466633, "grad_norm": 1.5811653137207031, "learning_rate": 3.1413696749491703e-06, "loss": 0.2313, "step": 41830 }, { "epoch": 0.7610178431764856, "grad_norm": 1.5342966318130493, "learning_rate": 3.1397219392429255e-06, "loss": 0.2238, "step": 41840 }, { "epoch": 0.7611997308063079, "grad_norm": 1.7292156219482422, "learning_rate": 3.138075196152307e-06, "loss": 0.2292, "step": 41850 }, { "epoch": 0.7613816184361302, "grad_norm": 1.8958659172058105, "learning_rate": 3.1364294462487897e-06, "loss": 0.2305, "step": 41860 }, { "epoch": 0.7615635060659525, "grad_norm": 1.5204015970230103, "learning_rate": 3.134784690103509e-06, "loss": 0.2274, "step": 41870 }, { "epoch": 0.7617453936957748, "grad_norm": 1.5755000114440918, "learning_rate": 3.1331409282872504e-06, "loss": 0.2205, "step": 41880 }, { "epoch": 0.7619272813255971, "grad_norm": 1.4617500305175781, "learning_rate": 3.1314981613704554e-06, "loss": 0.2251, "step": 41890 }, { "epoch": 0.7621091689554194, "grad_norm": 1.8036317825317383, "learning_rate": 3.1298563899232235e-06, "loss": 0.2258, "step": 41900 }, { "epoch": 0.7622910565852417, "grad_norm": 1.3277039527893066, "learning_rate": 3.1282156145153036e-06, "loss": 0.221, "step": 41910 }, { "epoch": 0.762472944215064, "grad_norm": 1.8417212963104248, "learning_rate": 3.1265758357161045e-06, "loss": 0.2268, "step": 41920 }, { "epoch": 0.7626548318448862, "grad_norm": 1.6140587329864502, "learning_rate": 3.1249370540946846e-06, "loss": 0.227, "step": 41930 }, { "epoch": 0.7628367194747085, "grad_norm": 1.5408518314361572, "learning_rate": 3.1232992702197584e-06, "loss": 0.2273, "step": 41940 }, { "epoch": 0.7630186071045308, "grad_norm": 1.4058839082717896, "learning_rate": 3.1216624846596934e-06, "loss": 0.2243, "step": 41950 }, { "epoch": 0.7632004947343531, "grad_norm": 1.3492591381072998, "learning_rate": 3.1200266979825095e-06, "loss": 0.2257, "step": 41960 }, { "epoch": 0.7633823823641754, "grad_norm": 1.3905291557312012, "learning_rate": 3.118391910755884e-06, "loss": 0.2239, "step": 41970 }, { "epoch": 0.7635642699939977, "grad_norm": 1.567042589187622, "learning_rate": 3.1167581235471432e-06, "loss": 0.2262, "step": 41980 }, { "epoch": 0.76374615762382, "grad_norm": 1.6218937635421753, "learning_rate": 3.1151253369232693e-06, "loss": 0.227, "step": 41990 }, { "epoch": 0.7639280452536423, "grad_norm": 1.657261610031128, "learning_rate": 3.113493551450894e-06, "loss": 0.2276, "step": 42000 }, { "epoch": 0.7641099328834646, "grad_norm": 1.3829878568649292, "learning_rate": 3.111862767696303e-06, "loss": 0.2239, "step": 42010 }, { "epoch": 0.7642918205132869, "grad_norm": 1.8736891746520996, "learning_rate": 3.1102329862254377e-06, "loss": 0.2319, "step": 42020 }, { "epoch": 0.7644737081431092, "grad_norm": 1.6273642778396606, "learning_rate": 3.108604207603887e-06, "loss": 0.2277, "step": 42030 }, { "epoch": 0.7646555957729315, "grad_norm": 1.6001158952713013, "learning_rate": 3.106976432396893e-06, "loss": 0.2308, "step": 42040 }, { "epoch": 0.7648374834027538, "grad_norm": 1.5642099380493164, "learning_rate": 3.1053496611693513e-06, "loss": 0.2302, "step": 42050 }, { "epoch": 0.7650193710325761, "grad_norm": 1.3734921216964722, "learning_rate": 3.1037238944858057e-06, "loss": 0.2229, "step": 42060 }, { "epoch": 0.7652012586623984, "grad_norm": 1.6404715776443481, "learning_rate": 3.102099132910457e-06, "loss": 0.2244, "step": 42070 }, { "epoch": 0.7653831462922207, "grad_norm": 1.753686785697937, "learning_rate": 3.1004753770071504e-06, "loss": 0.2234, "step": 42080 }, { "epoch": 0.765565033922043, "grad_norm": 1.445485234260559, "learning_rate": 3.098852627339388e-06, "loss": 0.2236, "step": 42090 }, { "epoch": 0.7657469215518653, "grad_norm": 1.5742558240890503, "learning_rate": 3.0972308844703184e-06, "loss": 0.2321, "step": 42100 }, { "epoch": 0.7659288091816876, "grad_norm": 1.3763035535812378, "learning_rate": 3.0956101489627444e-06, "loss": 0.2279, "step": 42110 }, { "epoch": 0.7661106968115099, "grad_norm": 1.4601948261260986, "learning_rate": 3.093990421379116e-06, "loss": 0.2271, "step": 42120 }, { "epoch": 0.7662925844413322, "grad_norm": 1.6341818571090698, "learning_rate": 3.092371702281536e-06, "loss": 0.2249, "step": 42130 }, { "epoch": 0.7664744720711545, "grad_norm": 1.3995336294174194, "learning_rate": 3.0907539922317555e-06, "loss": 0.2202, "step": 42140 }, { "epoch": 0.7666563597009768, "grad_norm": 1.6182501316070557, "learning_rate": 3.0891372917911733e-06, "loss": 0.2245, "step": 42150 }, { "epoch": 0.7668382473307991, "grad_norm": 1.4102394580841064, "learning_rate": 3.0875216015208444e-06, "loss": 0.2235, "step": 42160 }, { "epoch": 0.7670201349606214, "grad_norm": 1.582679033279419, "learning_rate": 3.085906921981467e-06, "loss": 0.2241, "step": 42170 }, { "epoch": 0.7672020225904437, "grad_norm": 1.6178699731826782, "learning_rate": 3.0842932537333923e-06, "loss": 0.2221, "step": 42180 }, { "epoch": 0.767383910220266, "grad_norm": 1.5782257318496704, "learning_rate": 3.0826805973366193e-06, "loss": 0.2256, "step": 42190 }, { "epoch": 0.7675657978500882, "grad_norm": 1.4761996269226074, "learning_rate": 3.0810689533507944e-06, "loss": 0.2287, "step": 42200 }, { "epoch": 0.7677476854799105, "grad_norm": 1.5119014978408813, "learning_rate": 3.079458322335214e-06, "loss": 0.2235, "step": 42210 }, { "epoch": 0.7679295731097328, "grad_norm": 1.5160459280014038, "learning_rate": 3.0778487048488236e-06, "loss": 0.2228, "step": 42220 }, { "epoch": 0.768111460739555, "grad_norm": 1.6392141580581665, "learning_rate": 3.0762401014502158e-06, "loss": 0.2242, "step": 42230 }, { "epoch": 0.7682933483693773, "grad_norm": 2.839362621307373, "learning_rate": 3.0746325126976324e-06, "loss": 0.2226, "step": 42240 }, { "epoch": 0.7684752359991996, "grad_norm": 1.6999958753585815, "learning_rate": 3.0730259391489616e-06, "loss": 0.2273, "step": 42250 }, { "epoch": 0.7686571236290219, "grad_norm": 1.7111214399337769, "learning_rate": 3.0714203813617395e-06, "loss": 0.2322, "step": 42260 }, { "epoch": 0.7688390112588442, "grad_norm": 1.6779775619506836, "learning_rate": 3.069815839893152e-06, "loss": 0.2279, "step": 42270 }, { "epoch": 0.7690208988886665, "grad_norm": 1.7394524812698364, "learning_rate": 3.0682123153000297e-06, "loss": 0.2238, "step": 42280 }, { "epoch": 0.7692027865184888, "grad_norm": 1.5193098783493042, "learning_rate": 3.066609808138852e-06, "loss": 0.2199, "step": 42290 }, { "epoch": 0.7693846741483111, "grad_norm": 1.6421772241592407, "learning_rate": 3.065008318965742e-06, "loss": 0.2314, "step": 42300 }, { "epoch": 0.7695665617781334, "grad_norm": 1.6395272016525269, "learning_rate": 3.063407848336475e-06, "loss": 0.231, "step": 42310 }, { "epoch": 0.7697484494079557, "grad_norm": 1.6416370868682861, "learning_rate": 3.061808396806468e-06, "loss": 0.234, "step": 42320 }, { "epoch": 0.769930337037778, "grad_norm": 1.7031155824661255, "learning_rate": 3.0602099649307872e-06, "loss": 0.228, "step": 42330 }, { "epoch": 0.7701122246676003, "grad_norm": 1.6196895837783813, "learning_rate": 3.058612553264143e-06, "loss": 0.2218, "step": 42340 }, { "epoch": 0.7702941122974226, "grad_norm": 1.5699883699417114, "learning_rate": 3.0570161623608925e-06, "loss": 0.2279, "step": 42350 }, { "epoch": 0.7704759999272449, "grad_norm": 1.5542292594909668, "learning_rate": 3.055420792775039e-06, "loss": 0.2326, "step": 42360 }, { "epoch": 0.7706578875570672, "grad_norm": 1.839769721031189, "learning_rate": 3.0538264450602317e-06, "loss": 0.226, "step": 42370 }, { "epoch": 0.7708397751868895, "grad_norm": 1.5752525329589844, "learning_rate": 3.0522331197697642e-06, "loss": 0.2246, "step": 42380 }, { "epoch": 0.7710216628167118, "grad_norm": 1.5795162916183472, "learning_rate": 3.050640817456575e-06, "loss": 0.2266, "step": 42390 }, { "epoch": 0.7712035504465341, "grad_norm": 1.6995458602905273, "learning_rate": 3.049049538673247e-06, "loss": 0.2249, "step": 42400 }, { "epoch": 0.7713854380763564, "grad_norm": 1.656883955001831, "learning_rate": 3.0474592839720116e-06, "loss": 0.2176, "step": 42410 }, { "epoch": 0.7715673257061787, "grad_norm": 1.4850218296051025, "learning_rate": 3.0458700539047426e-06, "loss": 0.2237, "step": 42420 }, { "epoch": 0.771749213336001, "grad_norm": 1.8708817958831787, "learning_rate": 3.0442818490229554e-06, "loss": 0.2329, "step": 42430 }, { "epoch": 0.7719311009658233, "grad_norm": 1.5584526062011719, "learning_rate": 3.042694669877814e-06, "loss": 0.2272, "step": 42440 }, { "epoch": 0.7721129885956456, "grad_norm": 1.5424078702926636, "learning_rate": 3.041108517020123e-06, "loss": 0.2219, "step": 42450 }, { "epoch": 0.7722948762254679, "grad_norm": 1.6326087713241577, "learning_rate": 3.0395233910003317e-06, "loss": 0.2364, "step": 42460 }, { "epoch": 0.7724767638552902, "grad_norm": 1.7039490938186646, "learning_rate": 3.037939292368538e-06, "loss": 0.2305, "step": 42470 }, { "epoch": 0.7726586514851125, "grad_norm": 1.5335170030593872, "learning_rate": 3.036356221674477e-06, "loss": 0.2249, "step": 42480 }, { "epoch": 0.7728405391149348, "grad_norm": 1.4866690635681152, "learning_rate": 3.0347741794675272e-06, "loss": 0.2224, "step": 42490 }, { "epoch": 0.7730224267447571, "grad_norm": 1.5464259386062622, "learning_rate": 3.0331931662967127e-06, "loss": 0.2237, "step": 42500 }, { "epoch": 0.7732043143745794, "grad_norm": 1.5354105234146118, "learning_rate": 3.0316131827107022e-06, "loss": 0.2257, "step": 42510 }, { "epoch": 0.7733862020044017, "grad_norm": 1.6487513780593872, "learning_rate": 3.0300342292578027e-06, "loss": 0.2197, "step": 42520 }, { "epoch": 0.773568089634224, "grad_norm": 1.3964065313339233, "learning_rate": 3.028456306485966e-06, "loss": 0.2261, "step": 42530 }, { "epoch": 0.7737499772640463, "grad_norm": 1.4715402126312256, "learning_rate": 3.0268794149427873e-06, "loss": 0.2272, "step": 42540 }, { "epoch": 0.7739318648938686, "grad_norm": 1.5327374935150146, "learning_rate": 3.0253035551754992e-06, "loss": 0.2275, "step": 42550 }, { "epoch": 0.7741137525236909, "grad_norm": 1.4172598123550415, "learning_rate": 3.0237287277309836e-06, "loss": 0.2218, "step": 42560 }, { "epoch": 0.7742956401535132, "grad_norm": 1.7768877744674683, "learning_rate": 3.0221549331557587e-06, "loss": 0.2194, "step": 42570 }, { "epoch": 0.7744775277833355, "grad_norm": 1.618207573890686, "learning_rate": 3.020582171995986e-06, "loss": 0.2308, "step": 42580 }, { "epoch": 0.7746594154131577, "grad_norm": 1.5510598421096802, "learning_rate": 3.0190104447974675e-06, "loss": 0.2282, "step": 42590 }, { "epoch": 0.77484130304298, "grad_norm": 1.5114662647247314, "learning_rate": 3.0174397521056465e-06, "loss": 0.2226, "step": 42600 }, { "epoch": 0.7750231906728023, "grad_norm": 2.0693676471710205, "learning_rate": 3.0158700944656093e-06, "loss": 0.2303, "step": 42610 }, { "epoch": 0.7752050783026246, "grad_norm": 1.635353684425354, "learning_rate": 3.014301472422081e-06, "loss": 0.2203, "step": 42620 }, { "epoch": 0.7753869659324469, "grad_norm": 1.4840179681777954, "learning_rate": 3.0127338865194276e-06, "loss": 0.2232, "step": 42630 }, { "epoch": 0.7755688535622692, "grad_norm": 1.699701189994812, "learning_rate": 3.011167337301656e-06, "loss": 0.2257, "step": 42640 }, { "epoch": 0.7757507411920915, "grad_norm": 1.6074998378753662, "learning_rate": 3.0096018253124117e-06, "loss": 0.2244, "step": 42650 }, { "epoch": 0.7759326288219138, "grad_norm": 1.597985029220581, "learning_rate": 3.0080373510949845e-06, "loss": 0.23, "step": 42660 }, { "epoch": 0.7761145164517361, "grad_norm": 1.452252745628357, "learning_rate": 3.0064739151922984e-06, "loss": 0.2269, "step": 42670 }, { "epoch": 0.7762964040815584, "grad_norm": 1.4596004486083984, "learning_rate": 3.004911518146922e-06, "loss": 0.2228, "step": 42680 }, { "epoch": 0.7764782917113807, "grad_norm": 1.5996555089950562, "learning_rate": 3.0033501605010585e-06, "loss": 0.2357, "step": 42690 }, { "epoch": 0.776660179341203, "grad_norm": 1.772981882095337, "learning_rate": 3.0017898427965553e-06, "loss": 0.2252, "step": 42700 }, { "epoch": 0.7768420669710253, "grad_norm": 1.8013094663619995, "learning_rate": 3.000230565574897e-06, "loss": 0.2287, "step": 42710 }, { "epoch": 0.7770239546008476, "grad_norm": 1.5652624368667603, "learning_rate": 2.9986723293772046e-06, "loss": 0.2214, "step": 42720 }, { "epoch": 0.7772058422306699, "grad_norm": 1.6401904821395874, "learning_rate": 2.997115134744243e-06, "loss": 0.2286, "step": 42730 }, { "epoch": 0.7773877298604922, "grad_norm": 1.4565964937210083, "learning_rate": 2.9955589822164112e-06, "loss": 0.224, "step": 42740 }, { "epoch": 0.7775696174903145, "grad_norm": 1.5585007667541504, "learning_rate": 2.994003872333747e-06, "loss": 0.2287, "step": 42750 }, { "epoch": 0.7777515051201368, "grad_norm": 1.7387639284133911, "learning_rate": 2.9924498056359295e-06, "loss": 0.2222, "step": 42760 }, { "epoch": 0.7779333927499591, "grad_norm": 1.6632596254348755, "learning_rate": 2.9908967826622726e-06, "loss": 0.226, "step": 42770 }, { "epoch": 0.7781152803797814, "grad_norm": 1.981288194656372, "learning_rate": 2.989344803951729e-06, "loss": 0.2255, "step": 42780 }, { "epoch": 0.7782971680096037, "grad_norm": 1.645169734954834, "learning_rate": 2.98779387004289e-06, "loss": 0.2258, "step": 42790 }, { "epoch": 0.778479055639426, "grad_norm": 1.8768880367279053, "learning_rate": 2.9862439814739815e-06, "loss": 0.2285, "step": 42800 }, { "epoch": 0.7786609432692483, "grad_norm": 1.6153594255447388, "learning_rate": 2.9846951387828705e-06, "loss": 0.2282, "step": 42810 }, { "epoch": 0.7788428308990706, "grad_norm": 1.7204914093017578, "learning_rate": 2.9831473425070588e-06, "loss": 0.2297, "step": 42820 }, { "epoch": 0.7790247185288929, "grad_norm": 1.3764595985412598, "learning_rate": 2.9816005931836845e-06, "loss": 0.2247, "step": 42830 }, { "epoch": 0.7792066061587152, "grad_norm": 1.552971363067627, "learning_rate": 2.9800548913495243e-06, "loss": 0.2195, "step": 42840 }, { "epoch": 0.7793884937885375, "grad_norm": 1.5393941402435303, "learning_rate": 2.978510237540988e-06, "loss": 0.2238, "step": 42850 }, { "epoch": 0.7795703814183598, "grad_norm": 1.5184656381607056, "learning_rate": 2.9769666322941272e-06, "loss": 0.2264, "step": 42860 }, { "epoch": 0.7797522690481821, "grad_norm": 1.5809845924377441, "learning_rate": 2.975424076144625e-06, "loss": 0.2222, "step": 42870 }, { "epoch": 0.7799341566780044, "grad_norm": 1.4929444789886475, "learning_rate": 2.9738825696278016e-06, "loss": 0.2264, "step": 42880 }, { "epoch": 0.7801160443078267, "grad_norm": 1.5562349557876587, "learning_rate": 2.9723421132786125e-06, "loss": 0.229, "step": 42890 }, { "epoch": 0.780297931937649, "grad_norm": 1.4021040201187134, "learning_rate": 2.9708027076316513e-06, "loss": 0.2215, "step": 42900 }, { "epoch": 0.7804798195674713, "grad_norm": 1.5819138288497925, "learning_rate": 2.969264353221144e-06, "loss": 0.2269, "step": 42910 }, { "epoch": 0.7806617071972936, "grad_norm": 1.5598760843276978, "learning_rate": 2.967727050580953e-06, "loss": 0.2222, "step": 42920 }, { "epoch": 0.7808435948271158, "grad_norm": 1.5301642417907715, "learning_rate": 2.9661908002445763e-06, "loss": 0.2224, "step": 42930 }, { "epoch": 0.7810254824569381, "grad_norm": 1.75581955909729, "learning_rate": 2.964655602745143e-06, "loss": 0.228, "step": 42940 }, { "epoch": 0.7812073700867604, "grad_norm": 1.5879426002502441, "learning_rate": 2.963121458615425e-06, "loss": 0.2226, "step": 42950 }, { "epoch": 0.7813892577165827, "grad_norm": 1.4016923904418945, "learning_rate": 2.96158836838782e-06, "loss": 0.2275, "step": 42960 }, { "epoch": 0.781571145346405, "grad_norm": 1.7532917261123657, "learning_rate": 2.9600563325943633e-06, "loss": 0.2222, "step": 42970 }, { "epoch": 0.7817530329762273, "grad_norm": 1.4906612634658813, "learning_rate": 2.958525351766727e-06, "loss": 0.2273, "step": 42980 }, { "epoch": 0.7819349206060496, "grad_norm": 1.5292712450027466, "learning_rate": 2.95699542643621e-06, "loss": 0.2217, "step": 42990 }, { "epoch": 0.7821168082358719, "grad_norm": 1.4805160760879517, "learning_rate": 2.9554665571337525e-06, "loss": 0.2169, "step": 43000 }, { "epoch": 0.7822986958656942, "grad_norm": 1.5987886190414429, "learning_rate": 2.953938744389926e-06, "loss": 0.2241, "step": 43010 }, { "epoch": 0.7824805834955165, "grad_norm": 1.4793434143066406, "learning_rate": 2.9524119887349334e-06, "loss": 0.226, "step": 43020 }, { "epoch": 0.7826624711253388, "grad_norm": 1.5781291723251343, "learning_rate": 2.9508862906986125e-06, "loss": 0.2245, "step": 43030 }, { "epoch": 0.782844358755161, "grad_norm": 1.4749236106872559, "learning_rate": 2.9493616508104315e-06, "loss": 0.2299, "step": 43040 }, { "epoch": 0.7830262463849833, "grad_norm": 1.4479331970214844, "learning_rate": 2.9478380695994934e-06, "loss": 0.2229, "step": 43050 }, { "epoch": 0.7832081340148056, "grad_norm": 1.5907857418060303, "learning_rate": 2.946315547594536e-06, "loss": 0.2246, "step": 43060 }, { "epoch": 0.7833900216446279, "grad_norm": 1.7501957416534424, "learning_rate": 2.9447940853239264e-06, "loss": 0.2262, "step": 43070 }, { "epoch": 0.7835719092744502, "grad_norm": 1.649501919746399, "learning_rate": 2.943273683315664e-06, "loss": 0.2222, "step": 43080 }, { "epoch": 0.7837537969042725, "grad_norm": 1.6185808181762695, "learning_rate": 2.9417543420973805e-06, "loss": 0.2235, "step": 43090 }, { "epoch": 0.7839356845340948, "grad_norm": 1.5840548276901245, "learning_rate": 2.940236062196341e-06, "loss": 0.2247, "step": 43100 }, { "epoch": 0.7841175721639171, "grad_norm": 1.4686129093170166, "learning_rate": 2.9387188441394408e-06, "loss": 0.217, "step": 43110 }, { "epoch": 0.7842994597937394, "grad_norm": 1.4070768356323242, "learning_rate": 2.9372026884532084e-06, "loss": 0.2248, "step": 43120 }, { "epoch": 0.7844813474235617, "grad_norm": 1.542201042175293, "learning_rate": 2.9356875956638008e-06, "loss": 0.2277, "step": 43130 }, { "epoch": 0.784663235053384, "grad_norm": 1.5359116792678833, "learning_rate": 2.9341735662970073e-06, "loss": 0.2255, "step": 43140 }, { "epoch": 0.7848451226832063, "grad_norm": 1.7011353969573975, "learning_rate": 2.93266060087825e-06, "loss": 0.2276, "step": 43150 }, { "epoch": 0.7850270103130286, "grad_norm": 1.74452543258667, "learning_rate": 2.93114869993258e-06, "loss": 0.2219, "step": 43160 }, { "epoch": 0.7852088979428509, "grad_norm": 1.6143320798873901, "learning_rate": 2.9296378639846797e-06, "loss": 0.2191, "step": 43170 }, { "epoch": 0.7853907855726732, "grad_norm": 1.5737248659133911, "learning_rate": 2.9281280935588606e-06, "loss": 0.2254, "step": 43180 }, { "epoch": 0.7855726732024955, "grad_norm": 1.4828940629959106, "learning_rate": 2.926619389179065e-06, "loss": 0.2242, "step": 43190 }, { "epoch": 0.7857545608323178, "grad_norm": 1.7650084495544434, "learning_rate": 2.9251117513688665e-06, "loss": 0.2216, "step": 43200 }, { "epoch": 0.7859364484621401, "grad_norm": 1.4697203636169434, "learning_rate": 2.9236051806514687e-06, "loss": 0.2266, "step": 43210 }, { "epoch": 0.7861183360919624, "grad_norm": 1.6550742387771606, "learning_rate": 2.9220996775497023e-06, "loss": 0.2291, "step": 43220 }, { "epoch": 0.7863002237217847, "grad_norm": 1.715627908706665, "learning_rate": 2.9205952425860277e-06, "loss": 0.2235, "step": 43230 }, { "epoch": 0.786482111351607, "grad_norm": 1.617796778678894, "learning_rate": 2.919091876282539e-06, "loss": 0.2219, "step": 43240 }, { "epoch": 0.7866639989814292, "grad_norm": 1.6077280044555664, "learning_rate": 2.9175895791609556e-06, "loss": 0.2306, "step": 43250 }, { "epoch": 0.7868458866112515, "grad_norm": 1.5231707096099854, "learning_rate": 2.9160883517426265e-06, "loss": 0.2277, "step": 43260 }, { "epoch": 0.7870277742410738, "grad_norm": 1.6581792831420898, "learning_rate": 2.9145881945485277e-06, "loss": 0.2246, "step": 43270 }, { "epoch": 0.7872096618708961, "grad_norm": 1.6049646139144897, "learning_rate": 2.9130891080992684e-06, "loss": 0.2243, "step": 43280 }, { "epoch": 0.7873915495007184, "grad_norm": 1.5201667547225952, "learning_rate": 2.911591092915082e-06, "loss": 0.2248, "step": 43290 }, { "epoch": 0.7875734371305407, "grad_norm": 1.485564112663269, "learning_rate": 2.910094149515833e-06, "loss": 0.2185, "step": 43300 }, { "epoch": 0.787755324760363, "grad_norm": 1.6930923461914062, "learning_rate": 2.908598278421012e-06, "loss": 0.2232, "step": 43310 }, { "epoch": 0.7879372123901853, "grad_norm": 1.729928970336914, "learning_rate": 2.9071034801497376e-06, "loss": 0.2212, "step": 43320 }, { "epoch": 0.7881191000200076, "grad_norm": 1.6475327014923096, "learning_rate": 2.9056097552207572e-06, "loss": 0.2262, "step": 43330 }, { "epoch": 0.7883009876498299, "grad_norm": 1.7008191347122192, "learning_rate": 2.9041171041524452e-06, "loss": 0.2261, "step": 43340 }, { "epoch": 0.7884828752796522, "grad_norm": 1.6102731227874756, "learning_rate": 2.902625527462802e-06, "loss": 0.2207, "step": 43350 }, { "epoch": 0.7886647629094745, "grad_norm": 1.49374258518219, "learning_rate": 2.9011350256694592e-06, "loss": 0.2304, "step": 43360 }, { "epoch": 0.7888466505392968, "grad_norm": 1.6847612857818604, "learning_rate": 2.89964559928967e-06, "loss": 0.2218, "step": 43370 }, { "epoch": 0.7890285381691191, "grad_norm": 1.315387487411499, "learning_rate": 2.8981572488403176e-06, "loss": 0.2186, "step": 43380 }, { "epoch": 0.7892104257989414, "grad_norm": 1.5319232940673828, "learning_rate": 2.8966699748379107e-06, "loss": 0.2153, "step": 43390 }, { "epoch": 0.7893923134287637, "grad_norm": 1.52704656124115, "learning_rate": 2.8951837777985862e-06, "loss": 0.23, "step": 43400 }, { "epoch": 0.789574201058586, "grad_norm": 2.0238494873046875, "learning_rate": 2.893698658238106e-06, "loss": 0.2259, "step": 43410 }, { "epoch": 0.7897560886884083, "grad_norm": 1.588160753250122, "learning_rate": 2.892214616671857e-06, "loss": 0.2198, "step": 43420 }, { "epoch": 0.7899379763182306, "grad_norm": 1.6149845123291016, "learning_rate": 2.8907316536148524e-06, "loss": 0.2191, "step": 43430 }, { "epoch": 0.7901198639480529, "grad_norm": 1.6071102619171143, "learning_rate": 2.8892497695817344e-06, "loss": 0.2205, "step": 43440 }, { "epoch": 0.7903017515778752, "grad_norm": 1.7389944791793823, "learning_rate": 2.887768965086766e-06, "loss": 0.2192, "step": 43450 }, { "epoch": 0.7904836392076975, "grad_norm": 1.5745720863342285, "learning_rate": 2.886289240643839e-06, "loss": 0.2299, "step": 43460 }, { "epoch": 0.7906655268375198, "grad_norm": 1.5986067056655884, "learning_rate": 2.8848105967664676e-06, "loss": 0.2251, "step": 43470 }, { "epoch": 0.7908474144673421, "grad_norm": 1.5638757944107056, "learning_rate": 2.883333033967792e-06, "loss": 0.2275, "step": 43480 }, { "epoch": 0.7910293020971644, "grad_norm": 1.752508282661438, "learning_rate": 2.881856552760581e-06, "loss": 0.2257, "step": 43490 }, { "epoch": 0.7912111897269867, "grad_norm": 1.473296880722046, "learning_rate": 2.8803811536572227e-06, "loss": 0.2231, "step": 43500 }, { "epoch": 0.791393077356809, "grad_norm": 1.44340980052948, "learning_rate": 2.8789068371697316e-06, "loss": 0.2185, "step": 43510 }, { "epoch": 0.7915749649866313, "grad_norm": 1.6513502597808838, "learning_rate": 2.877433603809746e-06, "loss": 0.23, "step": 43520 }, { "epoch": 0.7917568526164536, "grad_norm": 1.4450633525848389, "learning_rate": 2.8759614540885296e-06, "loss": 0.2229, "step": 43530 }, { "epoch": 0.7919387402462759, "grad_norm": 1.5262407064437866, "learning_rate": 2.87449038851697e-06, "loss": 0.2222, "step": 43540 }, { "epoch": 0.7921206278760982, "grad_norm": 1.5172697305679321, "learning_rate": 2.8730204076055755e-06, "loss": 0.222, "step": 43550 }, { "epoch": 0.7923025155059205, "grad_norm": 1.685287356376648, "learning_rate": 2.871551511864485e-06, "loss": 0.2298, "step": 43560 }, { "epoch": 0.7924844031357428, "grad_norm": 1.480605959892273, "learning_rate": 2.8700837018034534e-06, "loss": 0.2249, "step": 43570 }, { "epoch": 0.792666290765565, "grad_norm": 1.644003987312317, "learning_rate": 2.8686169779318608e-06, "loss": 0.2255, "step": 43580 }, { "epoch": 0.7928481783953873, "grad_norm": 1.541336178779602, "learning_rate": 2.8671513407587115e-06, "loss": 0.2269, "step": 43590 }, { "epoch": 0.7930300660252096, "grad_norm": 1.4250127077102661, "learning_rate": 2.8656867907926345e-06, "loss": 0.2252, "step": 43600 }, { "epoch": 0.7932119536550319, "grad_norm": 1.6509287357330322, "learning_rate": 2.8642233285418787e-06, "loss": 0.223, "step": 43610 }, { "epoch": 0.7933938412848542, "grad_norm": 1.402444839477539, "learning_rate": 2.862760954514314e-06, "loss": 0.2226, "step": 43620 }, { "epoch": 0.7935757289146765, "grad_norm": 1.5976272821426392, "learning_rate": 2.8612996692174367e-06, "loss": 0.2184, "step": 43630 }, { "epoch": 0.7937576165444988, "grad_norm": 1.7535666227340698, "learning_rate": 2.8598394731583635e-06, "loss": 0.2252, "step": 43640 }, { "epoch": 0.7939395041743211, "grad_norm": 1.715009093284607, "learning_rate": 2.858380366843834e-06, "loss": 0.2275, "step": 43650 }, { "epoch": 0.7941213918041434, "grad_norm": 1.4295657873153687, "learning_rate": 2.856922350780206e-06, "loss": 0.2219, "step": 43660 }, { "epoch": 0.7943032794339657, "grad_norm": 1.4344279766082764, "learning_rate": 2.8554654254734637e-06, "loss": 0.2198, "step": 43670 }, { "epoch": 0.794485167063788, "grad_norm": 1.666857123374939, "learning_rate": 2.854009591429209e-06, "loss": 0.2206, "step": 43680 }, { "epoch": 0.7946670546936103, "grad_norm": 1.5737017393112183, "learning_rate": 2.8525548491526684e-06, "loss": 0.2234, "step": 43690 }, { "epoch": 0.7948489423234326, "grad_norm": 1.5199137926101685, "learning_rate": 2.851101199148687e-06, "loss": 0.2222, "step": 43700 }, { "epoch": 0.7950308299532549, "grad_norm": 1.533660650253296, "learning_rate": 2.8496486419217327e-06, "loss": 0.2169, "step": 43710 }, { "epoch": 0.7952127175830772, "grad_norm": 1.6374728679656982, "learning_rate": 2.848197177975893e-06, "loss": 0.2213, "step": 43720 }, { "epoch": 0.7953946052128995, "grad_norm": 1.5039708614349365, "learning_rate": 2.846746807814873e-06, "loss": 0.2255, "step": 43730 }, { "epoch": 0.7955764928427218, "grad_norm": 1.6448384523391724, "learning_rate": 2.8452975319420062e-06, "loss": 0.2258, "step": 43740 }, { "epoch": 0.7957583804725441, "grad_norm": 1.7232133150100708, "learning_rate": 2.8438493508602393e-06, "loss": 0.2225, "step": 43750 }, { "epoch": 0.7959402681023664, "grad_norm": 1.5558747053146362, "learning_rate": 2.8424022650721415e-06, "loss": 0.2249, "step": 43760 }, { "epoch": 0.7961221557321887, "grad_norm": 1.5797094106674194, "learning_rate": 2.8409562750799026e-06, "loss": 0.2199, "step": 43770 }, { "epoch": 0.796304043362011, "grad_norm": 1.263487696647644, "learning_rate": 2.8395113813853286e-06, "loss": 0.2198, "step": 43780 }, { "epoch": 0.7964859309918333, "grad_norm": 1.6497843265533447, "learning_rate": 2.8380675844898507e-06, "loss": 0.2241, "step": 43790 }, { "epoch": 0.7966678186216556, "grad_norm": 1.6287074089050293, "learning_rate": 2.8366248848945144e-06, "loss": 0.2217, "step": 43800 }, { "epoch": 0.7968497062514779, "grad_norm": 1.4848452806472778, "learning_rate": 2.835183283099989e-06, "loss": 0.2195, "step": 43810 }, { "epoch": 0.7970315938813002, "grad_norm": 1.527714729309082, "learning_rate": 2.833742779606556e-06, "loss": 0.2213, "step": 43820 }, { "epoch": 0.7972134815111225, "grad_norm": 1.3014068603515625, "learning_rate": 2.8323033749141228e-06, "loss": 0.2237, "step": 43830 }, { "epoch": 0.7973953691409448, "grad_norm": 1.4647126197814941, "learning_rate": 2.8308650695222145e-06, "loss": 0.2234, "step": 43840 }, { "epoch": 0.7975772567707671, "grad_norm": 1.4026010036468506, "learning_rate": 2.82942786392997e-06, "loss": 0.2221, "step": 43850 }, { "epoch": 0.7977591444005894, "grad_norm": 1.5455846786499023, "learning_rate": 2.8279917586361516e-06, "loss": 0.2194, "step": 43860 }, { "epoch": 0.7979410320304116, "grad_norm": 1.5145128965377808, "learning_rate": 2.826556754139135e-06, "loss": 0.2252, "step": 43870 }, { "epoch": 0.7981229196602339, "grad_norm": 1.6196943521499634, "learning_rate": 2.8251228509369183e-06, "loss": 0.228, "step": 43880 }, { "epoch": 0.7983048072900562, "grad_norm": 1.6503673791885376, "learning_rate": 2.8236900495271158e-06, "loss": 0.2238, "step": 43890 }, { "epoch": 0.7984866949198784, "grad_norm": 1.5694082975387573, "learning_rate": 2.8222583504069582e-06, "loss": 0.224, "step": 43900 }, { "epoch": 0.7986685825497007, "grad_norm": 1.7611526250839233, "learning_rate": 2.8208277540732953e-06, "loss": 0.2237, "step": 43910 }, { "epoch": 0.798850470179523, "grad_norm": 1.7985469102859497, "learning_rate": 2.8193982610225946e-06, "loss": 0.2277, "step": 43920 }, { "epoch": 0.7990323578093453, "grad_norm": 1.546873688697815, "learning_rate": 2.8179698717509362e-06, "loss": 0.2211, "step": 43930 }, { "epoch": 0.7992142454391676, "grad_norm": 1.5399665832519531, "learning_rate": 2.816542586754025e-06, "loss": 0.2179, "step": 43940 }, { "epoch": 0.7993961330689899, "grad_norm": 1.8026998043060303, "learning_rate": 2.8151164065271763e-06, "loss": 0.2205, "step": 43950 }, { "epoch": 0.7995780206988122, "grad_norm": 1.5723596811294556, "learning_rate": 2.813691331565325e-06, "loss": 0.2184, "step": 43960 }, { "epoch": 0.7997599083286345, "grad_norm": 1.6394217014312744, "learning_rate": 2.8122673623630203e-06, "loss": 0.2329, "step": 43970 }, { "epoch": 0.7999417959584568, "grad_norm": 1.5772379636764526, "learning_rate": 2.810844499414428e-06, "loss": 0.2194, "step": 43980 }, { "epoch": 0.8001236835882791, "grad_norm": 1.767243504524231, "learning_rate": 2.8094227432133335e-06, "loss": 0.2234, "step": 43990 }, { "epoch": 0.8003055712181014, "grad_norm": 1.5920406579971313, "learning_rate": 2.808002094253134e-06, "loss": 0.2211, "step": 44000 }, { "epoch": 0.8004874588479237, "grad_norm": 1.7055844068527222, "learning_rate": 2.8065825530268442e-06, "loss": 0.2246, "step": 44010 }, { "epoch": 0.800669346477746, "grad_norm": 1.6793856620788574, "learning_rate": 2.805164120027093e-06, "loss": 0.2215, "step": 44020 }, { "epoch": 0.8008512341075683, "grad_norm": 1.4593816995620728, "learning_rate": 2.8037467957461284e-06, "loss": 0.2214, "step": 44030 }, { "epoch": 0.8010331217373906, "grad_norm": 1.5171387195587158, "learning_rate": 2.8023305806758093e-06, "loss": 0.2246, "step": 44040 }, { "epoch": 0.8012150093672129, "grad_norm": 1.5403754711151123, "learning_rate": 2.800915475307612e-06, "loss": 0.2273, "step": 44050 }, { "epoch": 0.8013968969970352, "grad_norm": 1.4078128337860107, "learning_rate": 2.799501480132626e-06, "loss": 0.2171, "step": 44060 }, { "epoch": 0.8015787846268575, "grad_norm": 1.5074537992477417, "learning_rate": 2.798088595641556e-06, "loss": 0.2188, "step": 44070 }, { "epoch": 0.8017606722566798, "grad_norm": 1.6876726150512695, "learning_rate": 2.7966768223247256e-06, "loss": 0.217, "step": 44080 }, { "epoch": 0.8019425598865021, "grad_norm": 1.7594119310379028, "learning_rate": 2.7952661606720654e-06, "loss": 0.2235, "step": 44090 }, { "epoch": 0.8021244475163244, "grad_norm": 1.7740134000778198, "learning_rate": 2.7938566111731245e-06, "loss": 0.2244, "step": 44100 }, { "epoch": 0.8023063351461467, "grad_norm": 1.456009030342102, "learning_rate": 2.792448174317068e-06, "loss": 0.2267, "step": 44110 }, { "epoch": 0.802488222775969, "grad_norm": 1.9524526596069336, "learning_rate": 2.79104085059267e-06, "loss": 0.2304, "step": 44120 }, { "epoch": 0.8026701104057913, "grad_norm": 1.6775113344192505, "learning_rate": 2.7896346404883202e-06, "loss": 0.2265, "step": 44130 }, { "epoch": 0.8028519980356136, "grad_norm": 1.5196272134780884, "learning_rate": 2.788229544492024e-06, "loss": 0.2223, "step": 44140 }, { "epoch": 0.8030338856654359, "grad_norm": 1.4368021488189697, "learning_rate": 2.7868255630913976e-06, "loss": 0.2222, "step": 44150 }, { "epoch": 0.8032157732952582, "grad_norm": 1.3528761863708496, "learning_rate": 2.785422696773671e-06, "loss": 0.2207, "step": 44160 }, { "epoch": 0.8033976609250805, "grad_norm": 1.4543849229812622, "learning_rate": 2.7840209460256874e-06, "loss": 0.2184, "step": 44170 }, { "epoch": 0.8035795485549028, "grad_norm": 1.4274356365203857, "learning_rate": 2.7826203113339024e-06, "loss": 0.2234, "step": 44180 }, { "epoch": 0.8037614361847251, "grad_norm": 1.473888635635376, "learning_rate": 2.7812207931843865e-06, "loss": 0.2235, "step": 44190 }, { "epoch": 0.8039433238145474, "grad_norm": 1.778382658958435, "learning_rate": 2.7798223920628197e-06, "loss": 0.2283, "step": 44200 }, { "epoch": 0.8041252114443697, "grad_norm": 1.958988904953003, "learning_rate": 2.7784251084544962e-06, "loss": 0.2294, "step": 44210 }, { "epoch": 0.804307099074192, "grad_norm": 1.5469648838043213, "learning_rate": 2.7770289428443208e-06, "loss": 0.2282, "step": 44220 }, { "epoch": 0.8044889867040143, "grad_norm": 1.6006443500518799, "learning_rate": 2.7756338957168126e-06, "loss": 0.2229, "step": 44230 }, { "epoch": 0.8046708743338365, "grad_norm": 1.5322225093841553, "learning_rate": 2.7742399675561003e-06, "loss": 0.2203, "step": 44240 }, { "epoch": 0.8048527619636588, "grad_norm": 1.8116118907928467, "learning_rate": 2.7728471588459273e-06, "loss": 0.2261, "step": 44250 }, { "epoch": 0.8050346495934811, "grad_norm": 1.750646710395813, "learning_rate": 2.771455470069644e-06, "loss": 0.2232, "step": 44260 }, { "epoch": 0.8052165372233034, "grad_norm": 1.453869342803955, "learning_rate": 2.7700649017102155e-06, "loss": 0.2247, "step": 44270 }, { "epoch": 0.8053984248531257, "grad_norm": 1.447540521621704, "learning_rate": 2.768675454250217e-06, "loss": 0.2264, "step": 44280 }, { "epoch": 0.805580312482948, "grad_norm": 1.6485536098480225, "learning_rate": 2.767287128171836e-06, "loss": 0.2252, "step": 44290 }, { "epoch": 0.8057622001127703, "grad_norm": 1.5859419107437134, "learning_rate": 2.7658999239568697e-06, "loss": 0.2264, "step": 44300 }, { "epoch": 0.8059440877425926, "grad_norm": 1.6988823413848877, "learning_rate": 2.7645138420867247e-06, "loss": 0.2241, "step": 44310 }, { "epoch": 0.8061259753724149, "grad_norm": 1.6945741176605225, "learning_rate": 2.7631288830424195e-06, "loss": 0.2216, "step": 44320 }, { "epoch": 0.8063078630022372, "grad_norm": 1.5651092529296875, "learning_rate": 2.7617450473045844e-06, "loss": 0.2277, "step": 44330 }, { "epoch": 0.8064897506320595, "grad_norm": 1.6481163501739502, "learning_rate": 2.760362335353457e-06, "loss": 0.2195, "step": 44340 }, { "epoch": 0.8066716382618818, "grad_norm": 1.8685877323150635, "learning_rate": 2.758980747668888e-06, "loss": 0.223, "step": 44350 }, { "epoch": 0.8068535258917041, "grad_norm": 1.3509308099746704, "learning_rate": 2.757600284730334e-06, "loss": 0.2208, "step": 44360 }, { "epoch": 0.8070354135215264, "grad_norm": 1.5949625968933105, "learning_rate": 2.7562209470168643e-06, "loss": 0.2243, "step": 44370 }, { "epoch": 0.8072173011513487, "grad_norm": 1.4665307998657227, "learning_rate": 2.754842735007156e-06, "loss": 0.231, "step": 44380 }, { "epoch": 0.807399188781171, "grad_norm": 1.6172082424163818, "learning_rate": 2.7534656491794996e-06, "loss": 0.223, "step": 44390 }, { "epoch": 0.8075810764109933, "grad_norm": 1.74868643283844, "learning_rate": 2.752089690011789e-06, "loss": 0.2277, "step": 44400 }, { "epoch": 0.8077629640408156, "grad_norm": 1.341217279434204, "learning_rate": 2.75071485798153e-06, "loss": 0.2228, "step": 44410 }, { "epoch": 0.8079448516706379, "grad_norm": 1.7043770551681519, "learning_rate": 2.7493411535658364e-06, "loss": 0.2273, "step": 44420 }, { "epoch": 0.8081267393004602, "grad_norm": 1.760082721710205, "learning_rate": 2.7479685772414336e-06, "loss": 0.2213, "step": 44430 }, { "epoch": 0.8083086269302825, "grad_norm": 1.4481010437011719, "learning_rate": 2.7465971294846517e-06, "loss": 0.2206, "step": 44440 }, { "epoch": 0.8084905145601048, "grad_norm": 1.650583267211914, "learning_rate": 2.7452268107714302e-06, "loss": 0.219, "step": 44450 }, { "epoch": 0.8086724021899271, "grad_norm": 1.4578981399536133, "learning_rate": 2.7438576215773173e-06, "loss": 0.2189, "step": 44460 }, { "epoch": 0.8088542898197494, "grad_norm": 1.6339765787124634, "learning_rate": 2.7424895623774684e-06, "loss": 0.2196, "step": 44470 }, { "epoch": 0.8090361774495717, "grad_norm": 1.4714179039001465, "learning_rate": 2.7411226336466497e-06, "loss": 0.22, "step": 44480 }, { "epoch": 0.809218065079394, "grad_norm": 1.6155695915222168, "learning_rate": 2.7397568358592313e-06, "loss": 0.2186, "step": 44490 }, { "epoch": 0.8093999527092163, "grad_norm": 1.7989859580993652, "learning_rate": 2.738392169489193e-06, "loss": 0.2176, "step": 44500 }, { "epoch": 0.8095818403390386, "grad_norm": 1.4286705255508423, "learning_rate": 2.7370286350101204e-06, "loss": 0.2209, "step": 44510 }, { "epoch": 0.8097637279688609, "grad_norm": 1.4793686866760254, "learning_rate": 2.7356662328952073e-06, "loss": 0.2229, "step": 44520 }, { "epoch": 0.8099456155986832, "grad_norm": 1.7915781736373901, "learning_rate": 2.734304963617256e-06, "loss": 0.2233, "step": 44530 }, { "epoch": 0.8101275032285055, "grad_norm": 1.632137656211853, "learning_rate": 2.7329448276486724e-06, "loss": 0.2225, "step": 44540 }, { "epoch": 0.8103093908583278, "grad_norm": 1.570643424987793, "learning_rate": 2.731585825461472e-06, "loss": 0.2226, "step": 44550 }, { "epoch": 0.8104912784881501, "grad_norm": 1.5774385929107666, "learning_rate": 2.730227957527275e-06, "loss": 0.2235, "step": 44560 }, { "epoch": 0.8106731661179724, "grad_norm": 1.5231455564498901, "learning_rate": 2.7288712243173075e-06, "loss": 0.225, "step": 44570 }, { "epoch": 0.8108550537477947, "grad_norm": 1.796142578125, "learning_rate": 2.727515626302406e-06, "loss": 0.2247, "step": 44580 }, { "epoch": 0.811036941377617, "grad_norm": 1.5439376831054688, "learning_rate": 2.726161163953007e-06, "loss": 0.2155, "step": 44590 }, { "epoch": 0.8112188290074392, "grad_norm": 1.544251799583435, "learning_rate": 2.7248078377391583e-06, "loss": 0.2191, "step": 44600 }, { "epoch": 0.8114007166372615, "grad_norm": 2.8411121368408203, "learning_rate": 2.7234556481305086e-06, "loss": 0.2232, "step": 44610 }, { "epoch": 0.8115826042670838, "grad_norm": 1.420611023902893, "learning_rate": 2.722104595596316e-06, "loss": 0.2212, "step": 44620 }, { "epoch": 0.8117644918969061, "grad_norm": 1.926038146018982, "learning_rate": 2.720754680605443e-06, "loss": 0.2216, "step": 44630 }, { "epoch": 0.8119463795267284, "grad_norm": 1.9240314960479736, "learning_rate": 2.719405903626356e-06, "loss": 0.2221, "step": 44640 }, { "epoch": 0.8121282671565507, "grad_norm": 1.5839893817901611, "learning_rate": 2.7180582651271265e-06, "loss": 0.2218, "step": 44650 }, { "epoch": 0.812310154786373, "grad_norm": 1.6030551195144653, "learning_rate": 2.7167117655754343e-06, "loss": 0.2213, "step": 44660 }, { "epoch": 0.8124920424161953, "grad_norm": 1.515621304512024, "learning_rate": 2.715366405438558e-06, "loss": 0.2163, "step": 44670 }, { "epoch": 0.8126739300460176, "grad_norm": 1.4662079811096191, "learning_rate": 2.7140221851833882e-06, "loss": 0.2221, "step": 44680 }, { "epoch": 0.8128558176758399, "grad_norm": 1.525153398513794, "learning_rate": 2.7126791052764133e-06, "loss": 0.2201, "step": 44690 }, { "epoch": 0.8130377053056622, "grad_norm": 1.791620135307312, "learning_rate": 2.7113371661837296e-06, "loss": 0.2217, "step": 44700 }, { "epoch": 0.8132195929354844, "grad_norm": 1.5764795541763306, "learning_rate": 2.7099963683710354e-06, "loss": 0.225, "step": 44710 }, { "epoch": 0.8134014805653067, "grad_norm": 1.5742896795272827, "learning_rate": 2.7086567123036343e-06, "loss": 0.2206, "step": 44720 }, { "epoch": 0.813583368195129, "grad_norm": 1.853286623954773, "learning_rate": 2.7073181984464347e-06, "loss": 0.2229, "step": 44730 }, { "epoch": 0.8137652558249513, "grad_norm": 1.7238266468048096, "learning_rate": 2.705980827263946e-06, "loss": 0.2182, "step": 44740 }, { "epoch": 0.8139471434547736, "grad_norm": 1.3976832628250122, "learning_rate": 2.7046445992202828e-06, "loss": 0.214, "step": 44750 }, { "epoch": 0.8141290310845959, "grad_norm": 1.6787030696868896, "learning_rate": 2.7033095147791623e-06, "loss": 0.2193, "step": 44760 }, { "epoch": 0.8143109187144182, "grad_norm": 1.5766364336013794, "learning_rate": 2.7019755744039045e-06, "loss": 0.2212, "step": 44770 }, { "epoch": 0.8144928063442405, "grad_norm": 1.4424915313720703, "learning_rate": 2.7006427785574345e-06, "loss": 0.2191, "step": 44780 }, { "epoch": 0.8146746939740628, "grad_norm": 1.723390817642212, "learning_rate": 2.6993111277022775e-06, "loss": 0.2211, "step": 44790 }, { "epoch": 0.8148565816038851, "grad_norm": 1.9696917533874512, "learning_rate": 2.6979806223005638e-06, "loss": 0.2234, "step": 44800 }, { "epoch": 0.8150384692337074, "grad_norm": 1.7177850008010864, "learning_rate": 2.6966512628140223e-06, "loss": 0.227, "step": 44810 }, { "epoch": 0.8152203568635297, "grad_norm": 1.732845664024353, "learning_rate": 2.6953230497039893e-06, "loss": 0.226, "step": 44820 }, { "epoch": 0.815402244493352, "grad_norm": 1.7561352252960205, "learning_rate": 2.6939959834314e-06, "loss": 0.2216, "step": 44830 }, { "epoch": 0.8155841321231743, "grad_norm": 1.436996579170227, "learning_rate": 2.6926700644567925e-06, "loss": 0.2166, "step": 44840 }, { "epoch": 0.8157660197529966, "grad_norm": 1.4266568422317505, "learning_rate": 2.691345293240306e-06, "loss": 0.2177, "step": 44850 }, { "epoch": 0.8159479073828189, "grad_norm": 1.7610318660736084, "learning_rate": 2.6900216702416828e-06, "loss": 0.2191, "step": 44860 }, { "epoch": 0.8161297950126412, "grad_norm": 1.5649834871292114, "learning_rate": 2.6886991959202652e-06, "loss": 0.2194, "step": 44870 }, { "epoch": 0.8163116826424635, "grad_norm": 1.940599799156189, "learning_rate": 2.6873778707349983e-06, "loss": 0.2306, "step": 44880 }, { "epoch": 0.8164935702722858, "grad_norm": 1.7465070486068726, "learning_rate": 2.686057695144428e-06, "loss": 0.2228, "step": 44890 }, { "epoch": 0.816675457902108, "grad_norm": 1.3582055568695068, "learning_rate": 2.6847386696066995e-06, "loss": 0.2184, "step": 44900 }, { "epoch": 0.8168573455319303, "grad_norm": 1.6862152814865112, "learning_rate": 2.6834207945795614e-06, "loss": 0.2223, "step": 44910 }, { "epoch": 0.8170392331617526, "grad_norm": 1.7375115156173706, "learning_rate": 2.6821040705203627e-06, "loss": 0.224, "step": 44920 }, { "epoch": 0.8172211207915749, "grad_norm": 1.514794111251831, "learning_rate": 2.680788497886051e-06, "loss": 0.2242, "step": 44930 }, { "epoch": 0.8174030084213972, "grad_norm": 1.703677773475647, "learning_rate": 2.6794740771331763e-06, "loss": 0.2203, "step": 44940 }, { "epoch": 0.8175848960512195, "grad_norm": 1.60550057888031, "learning_rate": 2.6781608087178885e-06, "loss": 0.2235, "step": 44950 }, { "epoch": 0.8177667836810418, "grad_norm": 1.825384259223938, "learning_rate": 2.676848693095937e-06, "loss": 0.2154, "step": 44960 }, { "epoch": 0.8179486713108641, "grad_norm": 1.4905916452407837, "learning_rate": 2.6755377307226694e-06, "loss": 0.226, "step": 44970 }, { "epoch": 0.8181305589406864, "grad_norm": 1.8254626989364624, "learning_rate": 2.6742279220530387e-06, "loss": 0.2305, "step": 44980 }, { "epoch": 0.8183124465705087, "grad_norm": 1.4971168041229248, "learning_rate": 2.6729192675415916e-06, "loss": 0.2209, "step": 44990 }, { "epoch": 0.818494334200331, "grad_norm": 1.7012851238250732, "learning_rate": 2.671611767642477e-06, "loss": 0.2197, "step": 45000 }, { "epoch": 0.8186762218301533, "grad_norm": 1.4867409467697144, "learning_rate": 2.670305422809442e-06, "loss": 0.2215, "step": 45010 }, { "epoch": 0.8188581094599756, "grad_norm": 1.419195532798767, "learning_rate": 2.6690002334958354e-06, "loss": 0.2215, "step": 45020 }, { "epoch": 0.8190399970897979, "grad_norm": 1.6450014114379883, "learning_rate": 2.6676962001546014e-06, "loss": 0.2205, "step": 45030 }, { "epoch": 0.8192218847196202, "grad_norm": 1.5068774223327637, "learning_rate": 2.666393323238286e-06, "loss": 0.2144, "step": 45040 }, { "epoch": 0.8194037723494425, "grad_norm": 1.626570701599121, "learning_rate": 2.6650916031990322e-06, "loss": 0.223, "step": 45050 }, { "epoch": 0.8195856599792648, "grad_norm": 1.3160313367843628, "learning_rate": 2.6637910404885804e-06, "loss": 0.2182, "step": 45060 }, { "epoch": 0.8197675476090871, "grad_norm": 1.4862666130065918, "learning_rate": 2.6624916355582732e-06, "loss": 0.2235, "step": 45070 }, { "epoch": 0.8199494352389094, "grad_norm": 1.459470272064209, "learning_rate": 2.6611933888590487e-06, "loss": 0.2219, "step": 45080 }, { "epoch": 0.8201313228687317, "grad_norm": 1.7187587022781372, "learning_rate": 2.6598963008414424e-06, "loss": 0.2167, "step": 45090 }, { "epoch": 0.820313210498554, "grad_norm": 1.5223971605300903, "learning_rate": 2.6586003719555896e-06, "loss": 0.2184, "step": 45100 }, { "epoch": 0.8204950981283763, "grad_norm": 1.8370494842529297, "learning_rate": 2.6573056026512215e-06, "loss": 0.2193, "step": 45110 }, { "epoch": 0.8206769857581986, "grad_norm": 1.6080551147460938, "learning_rate": 2.6560119933776707e-06, "loss": 0.2183, "step": 45120 }, { "epoch": 0.8208588733880209, "grad_norm": 1.7537709474563599, "learning_rate": 2.6547195445838614e-06, "loss": 0.2229, "step": 45130 }, { "epoch": 0.8210407610178432, "grad_norm": 1.600630521774292, "learning_rate": 2.6534282567183202e-06, "loss": 0.2189, "step": 45140 }, { "epoch": 0.8212226486476655, "grad_norm": 2.052199363708496, "learning_rate": 2.6521381302291678e-06, "loss": 0.2217, "step": 45150 }, { "epoch": 0.8214045362774878, "grad_norm": 1.7137939929962158, "learning_rate": 2.6508491655641214e-06, "loss": 0.2234, "step": 45160 }, { "epoch": 0.8215864239073101, "grad_norm": 1.7035177946090698, "learning_rate": 2.649561363170499e-06, "loss": 0.2186, "step": 45170 }, { "epoch": 0.8217683115371324, "grad_norm": 1.7127593755722046, "learning_rate": 2.6482747234952123e-06, "loss": 0.2205, "step": 45180 }, { "epoch": 0.8219501991669547, "grad_norm": 1.4728293418884277, "learning_rate": 2.646989246984768e-06, "loss": 0.2165, "step": 45190 }, { "epoch": 0.822132086796777, "grad_norm": 1.7955018281936646, "learning_rate": 2.645704934085272e-06, "loss": 0.224, "step": 45200 }, { "epoch": 0.8223139744265993, "grad_norm": 1.619748830795288, "learning_rate": 2.6444217852424253e-06, "loss": 0.2179, "step": 45210 }, { "epoch": 0.8224958620564216, "grad_norm": 1.5017597675323486, "learning_rate": 2.6431398009015265e-06, "loss": 0.2214, "step": 45220 }, { "epoch": 0.8226777496862439, "grad_norm": 1.7189174890518188, "learning_rate": 2.641858981507466e-06, "loss": 0.2202, "step": 45230 }, { "epoch": 0.8228596373160662, "grad_norm": 1.798473834991455, "learning_rate": 2.6405793275047353e-06, "loss": 0.2305, "step": 45240 }, { "epoch": 0.8230415249458884, "grad_norm": 1.6103712320327759, "learning_rate": 2.639300839337417e-06, "loss": 0.2216, "step": 45250 }, { "epoch": 0.8232234125757107, "grad_norm": 1.9024593830108643, "learning_rate": 2.6380235174491897e-06, "loss": 0.2211, "step": 45260 }, { "epoch": 0.823405300205533, "grad_norm": 1.517122507095337, "learning_rate": 2.636747362283331e-06, "loss": 0.2257, "step": 45270 }, { "epoch": 0.8235871878353553, "grad_norm": 1.509242057800293, "learning_rate": 2.63547237428271e-06, "loss": 0.2183, "step": 45280 }, { "epoch": 0.8237690754651776, "grad_norm": 1.6819661855697632, "learning_rate": 2.6341985538897906e-06, "loss": 0.2213, "step": 45290 }, { "epoch": 0.8239509630949999, "grad_norm": 1.5911716222763062, "learning_rate": 2.6329259015466345e-06, "loss": 0.2221, "step": 45300 }, { "epoch": 0.8241328507248222, "grad_norm": 1.5998164415359497, "learning_rate": 2.6316544176948934e-06, "loss": 0.2231, "step": 45310 }, { "epoch": 0.8243147383546445, "grad_norm": 1.6696820259094238, "learning_rate": 2.630384102775819e-06, "loss": 0.2192, "step": 45320 }, { "epoch": 0.8244966259844668, "grad_norm": 1.7189257144927979, "learning_rate": 2.6291149572302544e-06, "loss": 0.2183, "step": 45330 }, { "epoch": 0.8246785136142891, "grad_norm": 1.4429088830947876, "learning_rate": 2.6278469814986357e-06, "loss": 0.2219, "step": 45340 }, { "epoch": 0.8248604012441114, "grad_norm": 1.6223207712173462, "learning_rate": 2.6265801760209946e-06, "loss": 0.2253, "step": 45350 }, { "epoch": 0.8250422888739337, "grad_norm": 1.7426722049713135, "learning_rate": 2.625314541236957e-06, "loss": 0.2262, "step": 45360 }, { "epoch": 0.825224176503756, "grad_norm": 2.3376235961914062, "learning_rate": 2.6240500775857424e-06, "loss": 0.2175, "step": 45370 }, { "epoch": 0.8254060641335783, "grad_norm": 1.8039883375167847, "learning_rate": 2.6227867855061627e-06, "loss": 0.2188, "step": 45380 }, { "epoch": 0.8255879517634006, "grad_norm": 1.7391806840896606, "learning_rate": 2.6215246654366254e-06, "loss": 0.2264, "step": 45390 }, { "epoch": 0.8257698393932229, "grad_norm": 1.872894525527954, "learning_rate": 2.6202637178151275e-06, "loss": 0.2206, "step": 45400 }, { "epoch": 0.8259517270230452, "grad_norm": 1.3149220943450928, "learning_rate": 2.6190039430792635e-06, "loss": 0.2142, "step": 45410 }, { "epoch": 0.8261336146528675, "grad_norm": 1.7374378442764282, "learning_rate": 2.617745341666219e-06, "loss": 0.2201, "step": 45420 }, { "epoch": 0.8263155022826898, "grad_norm": 1.4413355588912964, "learning_rate": 2.6164879140127714e-06, "loss": 0.2238, "step": 45430 }, { "epoch": 0.8264973899125121, "grad_norm": 1.536802887916565, "learning_rate": 2.6152316605552924e-06, "loss": 0.2188, "step": 45440 }, { "epoch": 0.8266792775423344, "grad_norm": 1.6825134754180908, "learning_rate": 2.6139765817297434e-06, "loss": 0.2221, "step": 45450 }, { "epoch": 0.8268611651721567, "grad_norm": 1.910170316696167, "learning_rate": 2.6127226779716836e-06, "loss": 0.2183, "step": 45460 }, { "epoch": 0.827043052801979, "grad_norm": 1.4378530979156494, "learning_rate": 2.6114699497162587e-06, "loss": 0.2197, "step": 45470 }, { "epoch": 0.8272249404318013, "grad_norm": 1.6405071020126343, "learning_rate": 2.6102183973982086e-06, "loss": 0.2209, "step": 45480 }, { "epoch": 0.8274068280616236, "grad_norm": 1.4873217344284058, "learning_rate": 2.6089680214518674e-06, "loss": 0.2209, "step": 45490 }, { "epoch": 0.8275887156914459, "grad_norm": 1.550271987915039, "learning_rate": 2.6077188223111583e-06, "loss": 0.2195, "step": 45500 }, { "epoch": 0.8277706033212682, "grad_norm": 1.4853897094726562, "learning_rate": 2.6064708004095945e-06, "loss": 0.2214, "step": 45510 }, { "epoch": 0.8279524909510905, "grad_norm": 1.6911540031433105, "learning_rate": 2.6052239561802856e-06, "loss": 0.2202, "step": 45520 }, { "epoch": 0.8281343785809128, "grad_norm": 1.441166639328003, "learning_rate": 2.6039782900559285e-06, "loss": 0.2162, "step": 45530 }, { "epoch": 0.828316266210735, "grad_norm": 1.5016177892684937, "learning_rate": 2.602733802468813e-06, "loss": 0.2197, "step": 45540 }, { "epoch": 0.8284981538405572, "grad_norm": 1.4574086666107178, "learning_rate": 2.601490493850819e-06, "loss": 0.2252, "step": 45550 }, { "epoch": 0.8286800414703795, "grad_norm": 1.640177845954895, "learning_rate": 2.6002483646334165e-06, "loss": 0.2254, "step": 45560 }, { "epoch": 0.8288619291002018, "grad_norm": 1.56240713596344, "learning_rate": 2.59900741524767e-06, "loss": 0.2159, "step": 45570 }, { "epoch": 0.8290438167300241, "grad_norm": 1.5970180034637451, "learning_rate": 2.59776764612423e-06, "loss": 0.2227, "step": 45580 }, { "epoch": 0.8292257043598464, "grad_norm": 1.5315951108932495, "learning_rate": 2.5965290576933412e-06, "loss": 0.2101, "step": 45590 }, { "epoch": 0.8294075919896687, "grad_norm": 1.6785542964935303, "learning_rate": 2.5952916503848337e-06, "loss": 0.2225, "step": 45600 }, { "epoch": 0.829589479619491, "grad_norm": 1.5776458978652954, "learning_rate": 2.5940554246281333e-06, "loss": 0.2295, "step": 45610 }, { "epoch": 0.8297713672493133, "grad_norm": 1.7761632204055786, "learning_rate": 2.592820380852253e-06, "loss": 0.227, "step": 45620 }, { "epoch": 0.8299532548791356, "grad_norm": 1.6190102100372314, "learning_rate": 2.591586519485795e-06, "loss": 0.2154, "step": 45630 }, { "epoch": 0.8301351425089579, "grad_norm": 2.1331686973571777, "learning_rate": 2.5903538409569516e-06, "loss": 0.2205, "step": 45640 }, { "epoch": 0.8303170301387802, "grad_norm": 1.6402759552001953, "learning_rate": 2.5891223456935057e-06, "loss": 0.2201, "step": 45650 }, { "epoch": 0.8304989177686025, "grad_norm": 1.6089439392089844, "learning_rate": 2.5878920341228286e-06, "loss": 0.2199, "step": 45660 }, { "epoch": 0.8306808053984248, "grad_norm": 1.4964585304260254, "learning_rate": 2.586662906671882e-06, "loss": 0.2221, "step": 45670 }, { "epoch": 0.8308626930282471, "grad_norm": 1.6985679864883423, "learning_rate": 2.585434963767214e-06, "loss": 0.223, "step": 45680 }, { "epoch": 0.8310445806580694, "grad_norm": 1.607109546661377, "learning_rate": 2.584208205834964e-06, "loss": 0.2149, "step": 45690 }, { "epoch": 0.8312264682878917, "grad_norm": 1.425271987915039, "learning_rate": 2.5829826333008603e-06, "loss": 0.22, "step": 45700 }, { "epoch": 0.831408355917714, "grad_norm": 1.7867345809936523, "learning_rate": 2.581758246590218e-06, "loss": 0.2206, "step": 45710 }, { "epoch": 0.8315902435475363, "grad_norm": 1.5511635541915894, "learning_rate": 2.5805350461279435e-06, "loss": 0.2194, "step": 45720 }, { "epoch": 0.8317721311773586, "grad_norm": 1.8226492404937744, "learning_rate": 2.579313032338527e-06, "loss": 0.2241, "step": 45730 }, { "epoch": 0.8319540188071809, "grad_norm": 1.6600737571716309, "learning_rate": 2.5780922056460523e-06, "loss": 0.2201, "step": 45740 }, { "epoch": 0.8321359064370032, "grad_norm": 1.6084762811660767, "learning_rate": 2.5768725664741856e-06, "loss": 0.2177, "step": 45750 }, { "epoch": 0.8323177940668255, "grad_norm": 1.6620988845825195, "learning_rate": 2.5756541152461866e-06, "loss": 0.2222, "step": 45760 }, { "epoch": 0.8324996816966478, "grad_norm": 1.7406142950057983, "learning_rate": 2.5744368523848995e-06, "loss": 0.2231, "step": 45770 }, { "epoch": 0.8326815693264701, "grad_norm": 1.7591838836669922, "learning_rate": 2.5732207783127567e-06, "loss": 0.222, "step": 45780 }, { "epoch": 0.8328634569562924, "grad_norm": 1.445035457611084, "learning_rate": 2.572005893451778e-06, "loss": 0.221, "step": 45790 }, { "epoch": 0.8330453445861147, "grad_norm": 1.4194763898849487, "learning_rate": 2.570792198223569e-06, "loss": 0.2142, "step": 45800 }, { "epoch": 0.833227232215937, "grad_norm": 1.8514409065246582, "learning_rate": 2.569579693049327e-06, "loss": 0.2225, "step": 45810 }, { "epoch": 0.8334091198457593, "grad_norm": 1.4524284601211548, "learning_rate": 2.568368378349831e-06, "loss": 0.2248, "step": 45820 }, { "epoch": 0.8335910074755816, "grad_norm": 2.2880468368530273, "learning_rate": 2.56715825454545e-06, "loss": 0.2179, "step": 45830 }, { "epoch": 0.8337728951054039, "grad_norm": 1.762364149093628, "learning_rate": 2.5659493220561395e-06, "loss": 0.2258, "step": 45840 }, { "epoch": 0.8339547827352262, "grad_norm": 1.5219755172729492, "learning_rate": 2.564741581301438e-06, "loss": 0.2206, "step": 45850 }, { "epoch": 0.8341366703650485, "grad_norm": 1.5932985544204712, "learning_rate": 2.5635350327004777e-06, "loss": 0.2201, "step": 45860 }, { "epoch": 0.8343185579948708, "grad_norm": 1.666934847831726, "learning_rate": 2.562329676671971e-06, "loss": 0.2234, "step": 45870 }, { "epoch": 0.834500445624693, "grad_norm": 1.3618489503860474, "learning_rate": 2.5611255136342172e-06, "loss": 0.2192, "step": 45880 }, { "epoch": 0.8346823332545154, "grad_norm": 1.7931729555130005, "learning_rate": 2.5599225440051032e-06, "loss": 0.218, "step": 45890 }, { "epoch": 0.8348642208843376, "grad_norm": 1.7813940048217773, "learning_rate": 2.558720768202101e-06, "loss": 0.218, "step": 45900 }, { "epoch": 0.83504610851416, "grad_norm": 1.5229748487472534, "learning_rate": 2.55752018664227e-06, "loss": 0.2228, "step": 45910 }, { "epoch": 0.8352279961439822, "grad_norm": 1.6217128038406372, "learning_rate": 2.5563207997422517e-06, "loss": 0.2219, "step": 45920 }, { "epoch": 0.8354098837738045, "grad_norm": 1.4942569732666016, "learning_rate": 2.5551226079182763e-06, "loss": 0.2208, "step": 45930 }, { "epoch": 0.8355917714036268, "grad_norm": 1.4328118562698364, "learning_rate": 2.5539256115861566e-06, "loss": 0.2229, "step": 45940 }, { "epoch": 0.8357736590334491, "grad_norm": 1.6444305181503296, "learning_rate": 2.552729811161292e-06, "loss": 0.2244, "step": 45950 }, { "epoch": 0.8359555466632714, "grad_norm": 1.6650193929672241, "learning_rate": 2.551535207058668e-06, "loss": 0.2193, "step": 45960 }, { "epoch": 0.8361374342930937, "grad_norm": 1.6693511009216309, "learning_rate": 2.5503417996928527e-06, "loss": 0.22, "step": 45970 }, { "epoch": 0.836319321922916, "grad_norm": 1.5891889333724976, "learning_rate": 2.5491495894779994e-06, "loss": 0.2171, "step": 45980 }, { "epoch": 0.8365012095527383, "grad_norm": 1.4825189113616943, "learning_rate": 2.547958576827847e-06, "loss": 0.2171, "step": 45990 }, { "epoch": 0.8366830971825606, "grad_norm": 1.6337687969207764, "learning_rate": 2.546768762155718e-06, "loss": 0.2145, "step": 46000 }, { "epoch": 0.8368649848123829, "grad_norm": 1.6888765096664429, "learning_rate": 2.54558014587452e-06, "loss": 0.2213, "step": 46010 }, { "epoch": 0.8370468724422052, "grad_norm": 1.36631441116333, "learning_rate": 2.544392728396743e-06, "loss": 0.223, "step": 46020 }, { "epoch": 0.8372287600720275, "grad_norm": 1.7459758520126343, "learning_rate": 2.5432065101344616e-06, "loss": 0.2206, "step": 46030 }, { "epoch": 0.8374106477018498, "grad_norm": 2.088940382003784, "learning_rate": 2.542021491499337e-06, "loss": 0.2189, "step": 46040 }, { "epoch": 0.8375925353316721, "grad_norm": 1.4086999893188477, "learning_rate": 2.5408376729026086e-06, "loss": 0.2142, "step": 46050 }, { "epoch": 0.8377744229614944, "grad_norm": 1.614441990852356, "learning_rate": 2.539655054755105e-06, "loss": 0.2196, "step": 46060 }, { "epoch": 0.8379563105913167, "grad_norm": 1.4877558946609497, "learning_rate": 2.538473637467235e-06, "loss": 0.2205, "step": 46070 }, { "epoch": 0.838138198221139, "grad_norm": 1.458336353302002, "learning_rate": 2.5372934214489914e-06, "loss": 0.2169, "step": 46080 }, { "epoch": 0.8383200858509613, "grad_norm": 1.736435890197754, "learning_rate": 2.5361144071099496e-06, "loss": 0.2187, "step": 46090 }, { "epoch": 0.8385019734807836, "grad_norm": 1.6241075992584229, "learning_rate": 2.5349365948592675e-06, "loss": 0.23, "step": 46100 }, { "epoch": 0.8386838611106059, "grad_norm": 1.5988928079605103, "learning_rate": 2.533759985105689e-06, "loss": 0.2235, "step": 46110 }, { "epoch": 0.8388657487404282, "grad_norm": 1.5513267517089844, "learning_rate": 2.5325845782575377e-06, "loss": 0.2172, "step": 46120 }, { "epoch": 0.8390476363702505, "grad_norm": 1.8424766063690186, "learning_rate": 2.5314103747227204e-06, "loss": 0.2179, "step": 46130 }, { "epoch": 0.8392295240000728, "grad_norm": 1.5857994556427002, "learning_rate": 2.5302373749087255e-06, "loss": 0.2265, "step": 46140 }, { "epoch": 0.8394114116298951, "grad_norm": 1.5646429061889648, "learning_rate": 2.529065579222625e-06, "loss": 0.2139, "step": 46150 }, { "epoch": 0.8395932992597174, "grad_norm": 1.5306590795516968, "learning_rate": 2.5278949880710733e-06, "loss": 0.2224, "step": 46160 }, { "epoch": 0.8397751868895397, "grad_norm": 1.4946424961090088, "learning_rate": 2.526725601860306e-06, "loss": 0.2222, "step": 46170 }, { "epoch": 0.839957074519362, "grad_norm": 1.5833499431610107, "learning_rate": 2.5255574209961405e-06, "loss": 0.2197, "step": 46180 }, { "epoch": 0.8401389621491843, "grad_norm": 1.6160703897476196, "learning_rate": 2.5243904458839745e-06, "loss": 0.2243, "step": 46190 }, { "epoch": 0.8403208497790066, "grad_norm": 1.3714826107025146, "learning_rate": 2.523224676928792e-06, "loss": 0.22, "step": 46200 }, { "epoch": 0.8405027374088289, "grad_norm": 1.453411340713501, "learning_rate": 2.522060114535153e-06, "loss": 0.2171, "step": 46210 }, { "epoch": 0.8406846250386512, "grad_norm": 1.538541555404663, "learning_rate": 2.5208967591072014e-06, "loss": 0.2152, "step": 46220 }, { "epoch": 0.8408665126684735, "grad_norm": 1.637929081916809, "learning_rate": 2.5197346110486625e-06, "loss": 0.2194, "step": 46230 }, { "epoch": 0.8410484002982958, "grad_norm": 1.8073331117630005, "learning_rate": 2.51857367076284e-06, "loss": 0.2189, "step": 46240 }, { "epoch": 0.841230287928118, "grad_norm": 1.5738344192504883, "learning_rate": 2.517413938652623e-06, "loss": 0.2164, "step": 46250 }, { "epoch": 0.8414121755579403, "grad_norm": 1.748533844947815, "learning_rate": 2.516255415120478e-06, "loss": 0.2183, "step": 46260 }, { "epoch": 0.8415940631877626, "grad_norm": 1.534598469734192, "learning_rate": 2.5150981005684516e-06, "loss": 0.222, "step": 46270 }, { "epoch": 0.8417759508175849, "grad_norm": 1.7470864057540894, "learning_rate": 2.5139419953981726e-06, "loss": 0.2232, "step": 46280 }, { "epoch": 0.8419578384474072, "grad_norm": 1.7312983274459839, "learning_rate": 2.5127871000108493e-06, "loss": 0.2197, "step": 46290 }, { "epoch": 0.8421397260772295, "grad_norm": 1.6218007802963257, "learning_rate": 2.511633414807272e-06, "loss": 0.2228, "step": 46300 }, { "epoch": 0.8423216137070518, "grad_norm": 1.7451462745666504, "learning_rate": 2.510480940187808e-06, "loss": 0.2179, "step": 46310 }, { "epoch": 0.8425035013368741, "grad_norm": 1.4685285091400146, "learning_rate": 2.5093296765524073e-06, "loss": 0.2257, "step": 46320 }, { "epoch": 0.8426853889666964, "grad_norm": 1.2740005254745483, "learning_rate": 2.508179624300596e-06, "loss": 0.2176, "step": 46330 }, { "epoch": 0.8428672765965187, "grad_norm": 1.7241333723068237, "learning_rate": 2.5070307838314847e-06, "loss": 0.2206, "step": 46340 }, { "epoch": 0.843049164226341, "grad_norm": 1.5740649700164795, "learning_rate": 2.505883155543758e-06, "loss": 0.2208, "step": 46350 }, { "epoch": 0.8432310518561633, "grad_norm": 1.4280955791473389, "learning_rate": 2.504736739835685e-06, "loss": 0.2193, "step": 46360 }, { "epoch": 0.8434129394859856, "grad_norm": 1.6089574098587036, "learning_rate": 2.5035915371051123e-06, "loss": 0.2173, "step": 46370 }, { "epoch": 0.8435948271158078, "grad_norm": 1.649544358253479, "learning_rate": 2.5024475477494625e-06, "loss": 0.2185, "step": 46380 }, { "epoch": 0.8437767147456301, "grad_norm": 1.583290457725525, "learning_rate": 2.5013047721657402e-06, "loss": 0.2182, "step": 46390 }, { "epoch": 0.8439586023754524, "grad_norm": 1.5875389575958252, "learning_rate": 2.5001632107505287e-06, "loss": 0.2181, "step": 46400 }, { "epoch": 0.8441404900052747, "grad_norm": 1.6321688890457153, "learning_rate": 2.49902286389999e-06, "loss": 0.2204, "step": 46410 }, { "epoch": 0.844322377635097, "grad_norm": 1.4461219310760498, "learning_rate": 2.497883732009863e-06, "loss": 0.2198, "step": 46420 }, { "epoch": 0.8445042652649193, "grad_norm": 1.3250865936279297, "learning_rate": 2.4967458154754663e-06, "loss": 0.2188, "step": 46430 }, { "epoch": 0.8446861528947416, "grad_norm": 1.7512022256851196, "learning_rate": 2.4956091146916953e-06, "loss": 0.2198, "step": 46440 }, { "epoch": 0.8448680405245639, "grad_norm": 1.8719885349273682, "learning_rate": 2.4944736300530267e-06, "loss": 0.2243, "step": 46450 }, { "epoch": 0.8450499281543862, "grad_norm": 1.7234869003295898, "learning_rate": 2.4933393619535116e-06, "loss": 0.2151, "step": 46460 }, { "epoch": 0.8452318157842085, "grad_norm": 1.5953426361083984, "learning_rate": 2.492206310786781e-06, "loss": 0.2177, "step": 46470 }, { "epoch": 0.8454137034140308, "grad_norm": 1.49685537815094, "learning_rate": 2.491074476946042e-06, "loss": 0.2251, "step": 46480 }, { "epoch": 0.8455955910438531, "grad_norm": 1.6331065893173218, "learning_rate": 2.4899438608240804e-06, "loss": 0.2187, "step": 46490 }, { "epoch": 0.8457774786736754, "grad_norm": 1.5650495290756226, "learning_rate": 2.48881446281326e-06, "loss": 0.215, "step": 46500 }, { "epoch": 0.8459593663034977, "grad_norm": 1.7782893180847168, "learning_rate": 2.4876862833055204e-06, "loss": 0.2127, "step": 46510 }, { "epoch": 0.84614125393332, "grad_norm": 1.6833785772323608, "learning_rate": 2.4865593226923796e-06, "loss": 0.2156, "step": 46520 }, { "epoch": 0.8463231415631423, "grad_norm": 1.471179485321045, "learning_rate": 2.4854335813649307e-06, "loss": 0.2215, "step": 46530 }, { "epoch": 0.8465050291929646, "grad_norm": 1.6767085790634155, "learning_rate": 2.4843090597138464e-06, "loss": 0.2227, "step": 46540 }, { "epoch": 0.8466869168227869, "grad_norm": 1.4617966413497925, "learning_rate": 2.483185758129374e-06, "loss": 0.2206, "step": 46550 }, { "epoch": 0.8468688044526091, "grad_norm": 1.7027136087417603, "learning_rate": 2.4820636770013385e-06, "loss": 0.2152, "step": 46560 }, { "epoch": 0.8470506920824314, "grad_norm": 1.7018609046936035, "learning_rate": 2.480942816719141e-06, "loss": 0.2232, "step": 46570 }, { "epoch": 0.8472325797122537, "grad_norm": 1.788434386253357, "learning_rate": 2.4798231776717565e-06, "loss": 0.2086, "step": 46580 }, { "epoch": 0.847414467342076, "grad_norm": 1.5882635116577148, "learning_rate": 2.4787047602477415e-06, "loss": 0.2227, "step": 46590 }, { "epoch": 0.8475963549718983, "grad_norm": 1.6708132028579712, "learning_rate": 2.4775875648352253e-06, "loss": 0.2215, "step": 46600 }, { "epoch": 0.8477782426017206, "grad_norm": 1.475915551185608, "learning_rate": 2.476471591821913e-06, "loss": 0.2172, "step": 46610 }, { "epoch": 0.8479601302315429, "grad_norm": 1.6494638919830322, "learning_rate": 2.4753568415950848e-06, "loss": 0.2187, "step": 46620 }, { "epoch": 0.8481420178613652, "grad_norm": 1.6581273078918457, "learning_rate": 2.4742433145415994e-06, "loss": 0.2162, "step": 46630 }, { "epoch": 0.8483239054911875, "grad_norm": 1.4054946899414062, "learning_rate": 2.4731310110478875e-06, "loss": 0.217, "step": 46640 }, { "epoch": 0.8485057931210098, "grad_norm": 1.5314861536026, "learning_rate": 2.4720199314999584e-06, "loss": 0.2204, "step": 46650 }, { "epoch": 0.8486876807508321, "grad_norm": 1.5373157262802124, "learning_rate": 2.4709100762833953e-06, "loss": 0.2212, "step": 46660 }, { "epoch": 0.8488695683806544, "grad_norm": 1.3571586608886719, "learning_rate": 2.4698014457833557e-06, "loss": 0.2204, "step": 46670 }, { "epoch": 0.8490514560104767, "grad_norm": 1.684404969215393, "learning_rate": 2.4686940403845726e-06, "loss": 0.2209, "step": 46680 }, { "epoch": 0.849233343640299, "grad_norm": 1.4546865224838257, "learning_rate": 2.467587860471354e-06, "loss": 0.2169, "step": 46690 }, { "epoch": 0.8494152312701213, "grad_norm": 1.585585117340088, "learning_rate": 2.466482906427584e-06, "loss": 0.2141, "step": 46700 }, { "epoch": 0.8495971188999436, "grad_norm": 1.5344784259796143, "learning_rate": 2.4653791786367183e-06, "loss": 0.22, "step": 46710 }, { "epoch": 0.8497790065297659, "grad_norm": 1.5855134725570679, "learning_rate": 2.4642766774817905e-06, "loss": 0.2175, "step": 46720 }, { "epoch": 0.8499608941595882, "grad_norm": 1.6489709615707397, "learning_rate": 2.4631754033454037e-06, "loss": 0.2134, "step": 46730 }, { "epoch": 0.8501427817894105, "grad_norm": 1.841678261756897, "learning_rate": 2.462075356609742e-06, "loss": 0.2237, "step": 46740 }, { "epoch": 0.8503246694192328, "grad_norm": 1.5497024059295654, "learning_rate": 2.4609765376565574e-06, "loss": 0.2177, "step": 46750 }, { "epoch": 0.8505065570490551, "grad_norm": 1.474334478378296, "learning_rate": 2.459878946867179e-06, "loss": 0.2168, "step": 46760 }, { "epoch": 0.8506884446788774, "grad_norm": 1.528610110282898, "learning_rate": 2.458782584622508e-06, "loss": 0.2186, "step": 46770 }, { "epoch": 0.8508703323086997, "grad_norm": 1.5450189113616943, "learning_rate": 2.45768745130302e-06, "loss": 0.2253, "step": 46780 }, { "epoch": 0.851052219938522, "grad_norm": 1.4141016006469727, "learning_rate": 2.456593547288766e-06, "loss": 0.2132, "step": 46790 }, { "epoch": 0.8512341075683443, "grad_norm": 1.5778183937072754, "learning_rate": 2.4555008729593675e-06, "loss": 0.2233, "step": 46800 }, { "epoch": 0.8514159951981666, "grad_norm": 1.9362232685089111, "learning_rate": 2.45440942869402e-06, "loss": 0.219, "step": 46810 }, { "epoch": 0.8515978828279889, "grad_norm": 1.5379019975662231, "learning_rate": 2.4533192148714926e-06, "loss": 0.2154, "step": 46820 }, { "epoch": 0.8517797704578112, "grad_norm": 1.6528857946395874, "learning_rate": 2.452230231870127e-06, "loss": 0.2237, "step": 46830 }, { "epoch": 0.8519616580876335, "grad_norm": 1.770915150642395, "learning_rate": 2.4511424800678394e-06, "loss": 0.2161, "step": 46840 }, { "epoch": 0.8521435457174558, "grad_norm": 1.5583012104034424, "learning_rate": 2.4500559598421156e-06, "loss": 0.2178, "step": 46850 }, { "epoch": 0.8523254333472781, "grad_norm": 2.02091908454895, "learning_rate": 2.448970671570017e-06, "loss": 0.2227, "step": 46860 }, { "epoch": 0.8525073209771004, "grad_norm": 1.6759976148605347, "learning_rate": 2.4478866156281767e-06, "loss": 0.2205, "step": 46870 }, { "epoch": 0.8526892086069227, "grad_norm": 1.6698249578475952, "learning_rate": 2.4468037923927986e-06, "loss": 0.219, "step": 46880 }, { "epoch": 0.852871096236745, "grad_norm": 1.5587971210479736, "learning_rate": 2.445722202239659e-06, "loss": 0.2233, "step": 46890 }, { "epoch": 0.8530529838665672, "grad_norm": 1.5105293989181519, "learning_rate": 2.4446418455441097e-06, "loss": 0.2188, "step": 46900 }, { "epoch": 0.8532348714963895, "grad_norm": 1.8118854761123657, "learning_rate": 2.4435627226810707e-06, "loss": 0.2185, "step": 46910 }, { "epoch": 0.8534167591262118, "grad_norm": 1.5396713018417358, "learning_rate": 2.4424848340250346e-06, "loss": 0.2158, "step": 46920 }, { "epoch": 0.8535986467560341, "grad_norm": 1.4740480184555054, "learning_rate": 2.441408179950066e-06, "loss": 0.2197, "step": 46930 }, { "epoch": 0.8537805343858564, "grad_norm": 1.8889222145080566, "learning_rate": 2.440332760829802e-06, "loss": 0.2259, "step": 46940 }, { "epoch": 0.8539624220156787, "grad_norm": 1.492053747177124, "learning_rate": 2.4392585770374504e-06, "loss": 0.2216, "step": 46950 }, { "epoch": 0.854144309645501, "grad_norm": 1.5242769718170166, "learning_rate": 2.4381856289457895e-06, "loss": 0.2179, "step": 46960 }, { "epoch": 0.8543261972753233, "grad_norm": 1.6351228952407837, "learning_rate": 2.43711391692717e-06, "loss": 0.2209, "step": 46970 }, { "epoch": 0.8545080849051456, "grad_norm": 1.5266417264938354, "learning_rate": 2.4360434413535117e-06, "loss": 0.2174, "step": 46980 }, { "epoch": 0.8546899725349679, "grad_norm": 1.8363611698150635, "learning_rate": 2.4349742025963093e-06, "loss": 0.2178, "step": 46990 }, { "epoch": 0.8548718601647902, "grad_norm": 1.4672815799713135, "learning_rate": 2.4339062010266236e-06, "loss": 0.2149, "step": 47000 }, { "epoch": 0.8550537477946125, "grad_norm": 1.487020492553711, "learning_rate": 2.432839437015089e-06, "loss": 0.2168, "step": 47010 }, { "epoch": 0.8552356354244348, "grad_norm": 1.7355616092681885, "learning_rate": 2.4317739109319092e-06, "loss": 0.215, "step": 47020 }, { "epoch": 0.8554175230542571, "grad_norm": 1.582443118095398, "learning_rate": 2.4307096231468587e-06, "loss": 0.2154, "step": 47030 }, { "epoch": 0.8555994106840794, "grad_norm": 1.845543622970581, "learning_rate": 2.4296465740292834e-06, "loss": 0.22, "step": 47040 }, { "epoch": 0.8557812983139017, "grad_norm": 1.4626898765563965, "learning_rate": 2.4285847639480965e-06, "loss": 0.2142, "step": 47050 }, { "epoch": 0.855963185943724, "grad_norm": 1.9689143896102905, "learning_rate": 2.427524193271784e-06, "loss": 0.2177, "step": 47060 }, { "epoch": 0.8561450735735463, "grad_norm": 1.3770077228546143, "learning_rate": 2.4264648623684007e-06, "loss": 0.218, "step": 47070 }, { "epoch": 0.8563269612033686, "grad_norm": 1.8109238147735596, "learning_rate": 2.4254067716055695e-06, "loss": 0.2206, "step": 47080 }, { "epoch": 0.8565088488331909, "grad_norm": 1.6095725297927856, "learning_rate": 2.424349921350487e-06, "loss": 0.22, "step": 47090 }, { "epoch": 0.8566907364630132, "grad_norm": 1.7777554988861084, "learning_rate": 2.4232943119699158e-06, "loss": 0.2174, "step": 47100 }, { "epoch": 0.8568726240928355, "grad_norm": 1.6676456928253174, "learning_rate": 2.422239943830188e-06, "loss": 0.2219, "step": 47110 }, { "epoch": 0.8570545117226578, "grad_norm": 1.798563003540039, "learning_rate": 2.421186817297207e-06, "loss": 0.2159, "step": 47120 }, { "epoch": 0.8572363993524801, "grad_norm": 1.5027943849563599, "learning_rate": 2.420134932736444e-06, "loss": 0.2185, "step": 47130 }, { "epoch": 0.8574182869823024, "grad_norm": 1.7038503885269165, "learning_rate": 2.4190842905129393e-06, "loss": 0.2179, "step": 47140 }, { "epoch": 0.8576001746121247, "grad_norm": 1.4530085325241089, "learning_rate": 2.4180348909913027e-06, "loss": 0.219, "step": 47150 }, { "epoch": 0.857782062241947, "grad_norm": 1.6629583835601807, "learning_rate": 2.4169867345357113e-06, "loss": 0.2178, "step": 47160 }, { "epoch": 0.8579639498717693, "grad_norm": 1.587289571762085, "learning_rate": 2.415939821509913e-06, "loss": 0.2206, "step": 47170 }, { "epoch": 0.8581458375015916, "grad_norm": 1.4136275053024292, "learning_rate": 2.414894152277221e-06, "loss": 0.2176, "step": 47180 }, { "epoch": 0.8583277251314139, "grad_norm": 1.488362431526184, "learning_rate": 2.41384972720052e-06, "loss": 0.2261, "step": 47190 }, { "epoch": 0.8585096127612362, "grad_norm": 1.4429548978805542, "learning_rate": 2.412806546642262e-06, "loss": 0.2202, "step": 47200 }, { "epoch": 0.8586915003910583, "grad_norm": 1.7335416078567505, "learning_rate": 2.411764610964467e-06, "loss": 0.2193, "step": 47210 }, { "epoch": 0.8588733880208806, "grad_norm": 1.505759358406067, "learning_rate": 2.4107239205287214e-06, "loss": 0.2194, "step": 47220 }, { "epoch": 0.8590552756507029, "grad_norm": 1.5549830198287964, "learning_rate": 2.409684475696181e-06, "loss": 0.2184, "step": 47230 }, { "epoch": 0.8592371632805252, "grad_norm": 1.3998072147369385, "learning_rate": 2.4086462768275705e-06, "loss": 0.215, "step": 47240 }, { "epoch": 0.8594190509103475, "grad_norm": 1.5835505723953247, "learning_rate": 2.40760932428318e-06, "loss": 0.2158, "step": 47250 }, { "epoch": 0.8596009385401698, "grad_norm": 1.5188000202178955, "learning_rate": 2.4065736184228683e-06, "loss": 0.2146, "step": 47260 }, { "epoch": 0.8597828261699921, "grad_norm": 1.5598677396774292, "learning_rate": 2.4055391596060613e-06, "loss": 0.221, "step": 47270 }, { "epoch": 0.8599647137998144, "grad_norm": 1.7177484035491943, "learning_rate": 2.404505948191751e-06, "loss": 0.221, "step": 47280 }, { "epoch": 0.8601466014296367, "grad_norm": 1.5955262184143066, "learning_rate": 2.403473984538499e-06, "loss": 0.2264, "step": 47290 }, { "epoch": 0.860328489059459, "grad_norm": 1.4752774238586426, "learning_rate": 2.402443269004432e-06, "loss": 0.2157, "step": 47300 }, { "epoch": 0.8605103766892813, "grad_norm": 1.493731141090393, "learning_rate": 2.401413801947244e-06, "loss": 0.2176, "step": 47310 }, { "epoch": 0.8606922643191036, "grad_norm": 1.8592363595962524, "learning_rate": 2.4003855837241944e-06, "loss": 0.2171, "step": 47320 }, { "epoch": 0.8608741519489259, "grad_norm": 1.4667088985443115, "learning_rate": 2.3993586146921134e-06, "loss": 0.2135, "step": 47330 }, { "epoch": 0.8610560395787482, "grad_norm": 1.6936794519424438, "learning_rate": 2.398332895207392e-06, "loss": 0.212, "step": 47340 }, { "epoch": 0.8612379272085705, "grad_norm": 1.570433259010315, "learning_rate": 2.3973084256259928e-06, "loss": 0.2187, "step": 47350 }, { "epoch": 0.8614198148383928, "grad_norm": 1.8096070289611816, "learning_rate": 2.396285206303441e-06, "loss": 0.2202, "step": 47360 }, { "epoch": 0.8616017024682151, "grad_norm": 1.4470256567001343, "learning_rate": 2.3952632375948277e-06, "loss": 0.2204, "step": 47370 }, { "epoch": 0.8617835900980374, "grad_norm": 1.8379639387130737, "learning_rate": 2.3942425198548147e-06, "loss": 0.2182, "step": 47380 }, { "epoch": 0.8619654777278597, "grad_norm": 1.7470183372497559, "learning_rate": 2.3932230534376245e-06, "loss": 0.2218, "step": 47390 }, { "epoch": 0.862147365357682, "grad_norm": 1.524869441986084, "learning_rate": 2.3922048386970477e-06, "loss": 0.2137, "step": 47400 }, { "epoch": 0.8623292529875043, "grad_norm": 1.8196877241134644, "learning_rate": 2.3911878759864383e-06, "loss": 0.2197, "step": 47410 }, { "epoch": 0.8625111406173266, "grad_norm": 1.6333911418914795, "learning_rate": 2.390172165658721e-06, "loss": 0.2145, "step": 47420 }, { "epoch": 0.8626930282471489, "grad_norm": 1.4121137857437134, "learning_rate": 2.3891577080663797e-06, "loss": 0.2107, "step": 47430 }, { "epoch": 0.8628749158769712, "grad_norm": 1.488280177116394, "learning_rate": 2.3881445035614683e-06, "loss": 0.2194, "step": 47440 }, { "epoch": 0.8630568035067935, "grad_norm": 1.629680871963501, "learning_rate": 2.387132552495603e-06, "loss": 0.2243, "step": 47450 }, { "epoch": 0.8632386911366158, "grad_norm": 1.6300609111785889, "learning_rate": 2.3861218552199657e-06, "loss": 0.2236, "step": 47460 }, { "epoch": 0.8634205787664381, "grad_norm": 1.3787083625793457, "learning_rate": 2.3851124120853042e-06, "loss": 0.217, "step": 47470 }, { "epoch": 0.8636024663962604, "grad_norm": 1.6238939762115479, "learning_rate": 2.3841042234419284e-06, "loss": 0.2189, "step": 47480 }, { "epoch": 0.8637843540260827, "grad_norm": 1.6772152185440063, "learning_rate": 2.3830972896397165e-06, "loss": 0.22, "step": 47490 }, { "epoch": 0.863966241655905, "grad_norm": 1.5832314491271973, "learning_rate": 2.382091611028109e-06, "loss": 0.2162, "step": 47500 }, { "epoch": 0.8641481292857273, "grad_norm": 1.466549038887024, "learning_rate": 2.381087187956111e-06, "loss": 0.2167, "step": 47510 }, { "epoch": 0.8643300169155496, "grad_norm": 1.5919909477233887, "learning_rate": 2.380084020772292e-06, "loss": 0.2172, "step": 47520 }, { "epoch": 0.8645119045453719, "grad_norm": 1.370580792427063, "learning_rate": 2.379082109824786e-06, "loss": 0.2168, "step": 47530 }, { "epoch": 0.8646937921751942, "grad_norm": 1.8009626865386963, "learning_rate": 2.3780814554612915e-06, "loss": 0.2222, "step": 47540 }, { "epoch": 0.8648756798050165, "grad_norm": 2.908332109451294, "learning_rate": 2.3770820580290693e-06, "loss": 0.2209, "step": 47550 }, { "epoch": 0.8650575674348387, "grad_norm": 1.904571533203125, "learning_rate": 2.3760839178749452e-06, "loss": 0.2222, "step": 47560 }, { "epoch": 0.865239455064661, "grad_norm": 1.5211535692214966, "learning_rate": 2.375087035345308e-06, "loss": 0.2206, "step": 47570 }, { "epoch": 0.8654213426944833, "grad_norm": 1.6567026376724243, "learning_rate": 2.374091410786111e-06, "loss": 0.2152, "step": 47580 }, { "epoch": 0.8656032303243056, "grad_norm": 1.4654433727264404, "learning_rate": 2.3730970445428715e-06, "loss": 0.2173, "step": 47590 }, { "epoch": 0.8657851179541279, "grad_norm": 1.4674227237701416, "learning_rate": 2.372103936960667e-06, "loss": 0.2208, "step": 47600 }, { "epoch": 0.8659670055839502, "grad_norm": 1.6951502561569214, "learning_rate": 2.3711120883841416e-06, "loss": 0.2116, "step": 47610 }, { "epoch": 0.8661488932137725, "grad_norm": 1.5658916234970093, "learning_rate": 2.3701214991574996e-06, "loss": 0.2137, "step": 47620 }, { "epoch": 0.8663307808435948, "grad_norm": 1.4802206754684448, "learning_rate": 2.369132169624512e-06, "loss": 0.2219, "step": 47630 }, { "epoch": 0.8665126684734171, "grad_norm": 1.4666697978973389, "learning_rate": 2.368144100128509e-06, "loss": 0.2103, "step": 47640 }, { "epoch": 0.8666945561032394, "grad_norm": 1.5313905477523804, "learning_rate": 2.3671572910123855e-06, "loss": 0.2142, "step": 47650 }, { "epoch": 0.8668764437330617, "grad_norm": 1.5791854858398438, "learning_rate": 2.366171742618599e-06, "loss": 0.2177, "step": 47660 }, { "epoch": 0.867058331362884, "grad_norm": 1.7276084423065186, "learning_rate": 2.3651874552891666e-06, "loss": 0.2161, "step": 47670 }, { "epoch": 0.8672402189927063, "grad_norm": 1.625151515007019, "learning_rate": 2.3642044293656718e-06, "loss": 0.2209, "step": 47680 }, { "epoch": 0.8674221066225286, "grad_norm": 1.7878305912017822, "learning_rate": 2.36322266518926e-06, "loss": 0.2133, "step": 47690 }, { "epoch": 0.8676039942523509, "grad_norm": 1.6091423034667969, "learning_rate": 2.362242163100635e-06, "loss": 0.2168, "step": 47700 }, { "epoch": 0.8677858818821732, "grad_norm": 2.5581400394439697, "learning_rate": 2.3612629234400667e-06, "loss": 0.2214, "step": 47710 }, { "epoch": 0.8679677695119955, "grad_norm": 1.6099324226379395, "learning_rate": 2.3602849465473835e-06, "loss": 0.2152, "step": 47720 }, { "epoch": 0.8681496571418178, "grad_norm": 1.7335546016693115, "learning_rate": 2.3593082327619796e-06, "loss": 0.221, "step": 47730 }, { "epoch": 0.8683315447716401, "grad_norm": 1.4178905487060547, "learning_rate": 2.358332782422807e-06, "loss": 0.2223, "step": 47740 }, { "epoch": 0.8685134324014624, "grad_norm": 1.5112266540527344, "learning_rate": 2.3573585958683806e-06, "loss": 0.2118, "step": 47750 }, { "epoch": 0.8686953200312847, "grad_norm": 1.724450707435608, "learning_rate": 2.3563856734367772e-06, "loss": 0.2199, "step": 47760 }, { "epoch": 0.868877207661107, "grad_norm": 1.4649055004119873, "learning_rate": 2.355414015465635e-06, "loss": 0.2206, "step": 47770 }, { "epoch": 0.8690590952909293, "grad_norm": 1.5845561027526855, "learning_rate": 2.354443622292153e-06, "loss": 0.2178, "step": 47780 }, { "epoch": 0.8692409829207516, "grad_norm": 1.5292243957519531, "learning_rate": 2.353474494253091e-06, "loss": 0.2165, "step": 47790 }, { "epoch": 0.8694228705505739, "grad_norm": 1.5596344470977783, "learning_rate": 2.3525066316847693e-06, "loss": 0.2197, "step": 47800 }, { "epoch": 0.8696047581803962, "grad_norm": 1.6846836805343628, "learning_rate": 2.351540034923071e-06, "loss": 0.2155, "step": 47810 }, { "epoch": 0.8697866458102185, "grad_norm": 1.5527124404907227, "learning_rate": 2.350574704303438e-06, "loss": 0.2222, "step": 47820 }, { "epoch": 0.8699685334400408, "grad_norm": 1.5849783420562744, "learning_rate": 2.3496106401608736e-06, "loss": 0.2171, "step": 47830 }, { "epoch": 0.8701504210698631, "grad_norm": 1.6879748106002808, "learning_rate": 2.3486478428299415e-06, "loss": 0.2171, "step": 47840 }, { "epoch": 0.8703323086996854, "grad_norm": 1.66670823097229, "learning_rate": 2.347686312644766e-06, "loss": 0.2189, "step": 47850 }, { "epoch": 0.8705141963295077, "grad_norm": 1.651310920715332, "learning_rate": 2.3467260499390316e-06, "loss": 0.2155, "step": 47860 }, { "epoch": 0.87069608395933, "grad_norm": 1.945950984954834, "learning_rate": 2.345767055045981e-06, "loss": 0.2192, "step": 47870 }, { "epoch": 0.8708779715891523, "grad_norm": 1.4586187601089478, "learning_rate": 2.3448093282984207e-06, "loss": 0.2208, "step": 47880 }, { "epoch": 0.8710598592189746, "grad_norm": 1.4126148223876953, "learning_rate": 2.3438528700287145e-06, "loss": 0.2186, "step": 47890 }, { "epoch": 0.8712417468487968, "grad_norm": 1.7204301357269287, "learning_rate": 2.3428976805687865e-06, "loss": 0.2219, "step": 47900 }, { "epoch": 0.8714236344786191, "grad_norm": 1.4230365753173828, "learning_rate": 2.3419437602501194e-06, "loss": 0.2183, "step": 47910 }, { "epoch": 0.8716055221084414, "grad_norm": 1.6840287446975708, "learning_rate": 2.3409911094037585e-06, "loss": 0.2164, "step": 47920 }, { "epoch": 0.8717874097382637, "grad_norm": 1.712944507598877, "learning_rate": 2.340039728360306e-06, "loss": 0.2209, "step": 47930 }, { "epoch": 0.871969297368086, "grad_norm": 1.4489572048187256, "learning_rate": 2.3390896174499237e-06, "loss": 0.2152, "step": 47940 }, { "epoch": 0.8721511849979083, "grad_norm": 1.4974379539489746, "learning_rate": 2.3381407770023333e-06, "loss": 0.2149, "step": 47950 }, { "epoch": 0.8723330726277306, "grad_norm": 1.3884903192520142, "learning_rate": 2.337193207346815e-06, "loss": 0.2124, "step": 47960 }, { "epoch": 0.8725149602575529, "grad_norm": 1.7023890018463135, "learning_rate": 2.3362469088122077e-06, "loss": 0.2175, "step": 47970 }, { "epoch": 0.8726968478873752, "grad_norm": 1.5866307020187378, "learning_rate": 2.3353018817269125e-06, "loss": 0.2259, "step": 47980 }, { "epoch": 0.8728787355171975, "grad_norm": 1.617236852645874, "learning_rate": 2.334358126418884e-06, "loss": 0.214, "step": 47990 }, { "epoch": 0.8730606231470198, "grad_norm": 1.5860742330551147, "learning_rate": 2.3334156432156386e-06, "loss": 0.2194, "step": 48000 }, { "epoch": 0.8732425107768421, "grad_norm": 1.5623763799667358, "learning_rate": 2.332474432444251e-06, "loss": 0.2181, "step": 48010 }, { "epoch": 0.8734243984066644, "grad_norm": 1.5723611116409302, "learning_rate": 2.3315344944313532e-06, "loss": 0.2183, "step": 48020 }, { "epoch": 0.8736062860364867, "grad_norm": 1.4305757284164429, "learning_rate": 2.3305958295031373e-06, "loss": 0.2174, "step": 48030 }, { "epoch": 0.873788173666309, "grad_norm": 1.615243911743164, "learning_rate": 2.3296584379853524e-06, "loss": 0.2174, "step": 48040 }, { "epoch": 0.8739700612961312, "grad_norm": 1.598635196685791, "learning_rate": 2.328722320203305e-06, "loss": 0.2212, "step": 48050 }, { "epoch": 0.8741519489259535, "grad_norm": 1.5434132814407349, "learning_rate": 2.3277874764818612e-06, "loss": 0.2143, "step": 48060 }, { "epoch": 0.8743338365557758, "grad_norm": 1.6404497623443604, "learning_rate": 2.326853907145443e-06, "loss": 0.2166, "step": 48070 }, { "epoch": 0.8745157241855981, "grad_norm": 1.6273430585861206, "learning_rate": 2.3259216125180336e-06, "loss": 0.2163, "step": 48080 }, { "epoch": 0.8746976118154204, "grad_norm": 1.9097293615341187, "learning_rate": 2.3249905929231693e-06, "loss": 0.2172, "step": 48090 }, { "epoch": 0.8748794994452427, "grad_norm": 1.4671961069107056, "learning_rate": 2.3240608486839475e-06, "loss": 0.2152, "step": 48100 }, { "epoch": 0.875061387075065, "grad_norm": 2.0949974060058594, "learning_rate": 2.3231323801230205e-06, "loss": 0.2204, "step": 48110 }, { "epoch": 0.8752432747048873, "grad_norm": 1.4878032207489014, "learning_rate": 2.3222051875626004e-06, "loss": 0.2168, "step": 48120 }, { "epoch": 0.8754251623347096, "grad_norm": 1.6257480382919312, "learning_rate": 2.3212792713244547e-06, "loss": 0.2127, "step": 48130 }, { "epoch": 0.8756070499645319, "grad_norm": 1.7772722244262695, "learning_rate": 2.320354631729908e-06, "loss": 0.2139, "step": 48140 }, { "epoch": 0.8757889375943542, "grad_norm": 1.896026611328125, "learning_rate": 2.3194312690998427e-06, "loss": 0.2251, "step": 48150 }, { "epoch": 0.8759708252241765, "grad_norm": 1.8305730819702148, "learning_rate": 2.318509183754697e-06, "loss": 0.2192, "step": 48160 }, { "epoch": 0.8761527128539988, "grad_norm": 1.5334835052490234, "learning_rate": 2.3175883760144676e-06, "loss": 0.2137, "step": 48170 }, { "epoch": 0.8763346004838211, "grad_norm": 1.6222635507583618, "learning_rate": 2.316668846198706e-06, "loss": 0.2172, "step": 48180 }, { "epoch": 0.8765164881136434, "grad_norm": 1.6409903764724731, "learning_rate": 2.3157505946265203e-06, "loss": 0.2149, "step": 48190 }, { "epoch": 0.8766983757434657, "grad_norm": 1.5124157667160034, "learning_rate": 2.314833621616577e-06, "loss": 0.2169, "step": 48200 }, { "epoch": 0.876880263373288, "grad_norm": 1.5551868677139282, "learning_rate": 2.313917927487096e-06, "loss": 0.2159, "step": 48210 }, { "epoch": 0.8770621510031102, "grad_norm": 1.6489864587783813, "learning_rate": 2.3130035125558566e-06, "loss": 0.2122, "step": 48220 }, { "epoch": 0.8772440386329325, "grad_norm": 1.6299580335617065, "learning_rate": 2.312090377140191e-06, "loss": 0.2145, "step": 48230 }, { "epoch": 0.8774259262627548, "grad_norm": 1.5227888822555542, "learning_rate": 2.31117852155699e-06, "loss": 0.2156, "step": 48240 }, { "epoch": 0.8776078138925771, "grad_norm": 1.4735183715820312, "learning_rate": 2.3102679461226983e-06, "loss": 0.2145, "step": 48250 }, { "epoch": 0.8777897015223994, "grad_norm": 1.6101555824279785, "learning_rate": 2.3093586511533175e-06, "loss": 0.2201, "step": 48260 }, { "epoch": 0.8779715891522217, "grad_norm": 1.5715365409851074, "learning_rate": 2.3084506369644034e-06, "loss": 0.2175, "step": 48270 }, { "epoch": 0.878153476782044, "grad_norm": 1.3454794883728027, "learning_rate": 2.307543903871071e-06, "loss": 0.217, "step": 48280 }, { "epoch": 0.8783353644118663, "grad_norm": 1.5390565395355225, "learning_rate": 2.306638452187986e-06, "loss": 0.2209, "step": 48290 }, { "epoch": 0.8785172520416886, "grad_norm": 1.5743837356567383, "learning_rate": 2.3057342822293713e-06, "loss": 0.2117, "step": 48300 }, { "epoch": 0.8786991396715109, "grad_norm": 1.6095032691955566, "learning_rate": 2.3048313943090058e-06, "loss": 0.2142, "step": 48310 }, { "epoch": 0.8788810273013332, "grad_norm": 1.5865585803985596, "learning_rate": 2.3039297887402235e-06, "loss": 0.2176, "step": 48320 }, { "epoch": 0.8790629149311555, "grad_norm": 1.5612205266952515, "learning_rate": 2.303029465835912e-06, "loss": 0.2118, "step": 48330 }, { "epoch": 0.8792448025609778, "grad_norm": 1.6608513593673706, "learning_rate": 2.3021304259085154e-06, "loss": 0.2117, "step": 48340 }, { "epoch": 0.8794266901908001, "grad_norm": 1.2871698141098022, "learning_rate": 2.3012326692700306e-06, "loss": 0.219, "step": 48350 }, { "epoch": 0.8796085778206224, "grad_norm": 1.5790960788726807, "learning_rate": 2.30033619623201e-06, "loss": 0.2118, "step": 48360 }, { "epoch": 0.8797904654504447, "grad_norm": 1.595699429512024, "learning_rate": 2.299441007105563e-06, "loss": 0.2205, "step": 48370 }, { "epoch": 0.879972353080267, "grad_norm": 1.5646151304244995, "learning_rate": 2.2985471022013495e-06, "loss": 0.2204, "step": 48380 }, { "epoch": 0.8801542407100893, "grad_norm": 1.5497642755508423, "learning_rate": 2.297654481829586e-06, "loss": 0.2179, "step": 48390 }, { "epoch": 0.8803361283399116, "grad_norm": 1.8088572025299072, "learning_rate": 2.296763146300043e-06, "loss": 0.2151, "step": 48400 }, { "epoch": 0.8805180159697339, "grad_norm": 1.5708720684051514, "learning_rate": 2.295873095922043e-06, "loss": 0.2168, "step": 48410 }, { "epoch": 0.8806999035995562, "grad_norm": 1.650901436805725, "learning_rate": 2.2949843310044665e-06, "loss": 0.2165, "step": 48420 }, { "epoch": 0.8808817912293785, "grad_norm": 1.6483659744262695, "learning_rate": 2.2940968518557454e-06, "loss": 0.213, "step": 48430 }, { "epoch": 0.8810636788592008, "grad_norm": 1.4518420696258545, "learning_rate": 2.2932106587838644e-06, "loss": 0.2094, "step": 48440 }, { "epoch": 0.8812455664890231, "grad_norm": 1.777513861656189, "learning_rate": 2.292325752096365e-06, "loss": 0.2164, "step": 48450 }, { "epoch": 0.8814274541188454, "grad_norm": 1.5228859186172485, "learning_rate": 2.291442132100339e-06, "loss": 0.2133, "step": 48460 }, { "epoch": 0.8816093417486677, "grad_norm": 1.8106783628463745, "learning_rate": 2.2905597991024335e-06, "loss": 0.2206, "step": 48470 }, { "epoch": 0.88179122937849, "grad_norm": 1.5625507831573486, "learning_rate": 2.2896787534088494e-06, "loss": 0.2127, "step": 48480 }, { "epoch": 0.8819731170083123, "grad_norm": 1.6311596632003784, "learning_rate": 2.2887989953253397e-06, "loss": 0.2182, "step": 48490 }, { "epoch": 0.8821550046381346, "grad_norm": 1.5798946619033813, "learning_rate": 2.28792052515721e-06, "loss": 0.2167, "step": 48500 }, { "epoch": 0.8823368922679569, "grad_norm": 1.5797958374023438, "learning_rate": 2.287043343209321e-06, "loss": 0.2158, "step": 48510 }, { "epoch": 0.8825187798977792, "grad_norm": 1.4431475400924683, "learning_rate": 2.2861674497860853e-06, "loss": 0.2125, "step": 48520 }, { "epoch": 0.8827006675276015, "grad_norm": 1.5885894298553467, "learning_rate": 2.2852928451914677e-06, "loss": 0.2165, "step": 48530 }, { "epoch": 0.8828825551574238, "grad_norm": 1.5858372449874878, "learning_rate": 2.284419529728987e-06, "loss": 0.2103, "step": 48540 }, { "epoch": 0.883064442787246, "grad_norm": 1.6798572540283203, "learning_rate": 2.2835475037017133e-06, "loss": 0.2151, "step": 48550 }, { "epoch": 0.8832463304170683, "grad_norm": 1.8295904397964478, "learning_rate": 2.2826767674122694e-06, "loss": 0.219, "step": 48560 }, { "epoch": 0.8834282180468906, "grad_norm": 1.4538397789001465, "learning_rate": 2.2818073211628324e-06, "loss": 0.2157, "step": 48570 }, { "epoch": 0.8836101056767129, "grad_norm": 1.647520899772644, "learning_rate": 2.2809391652551294e-06, "loss": 0.2143, "step": 48580 }, { "epoch": 0.8837919933065352, "grad_norm": 1.73684823513031, "learning_rate": 2.2800722999904407e-06, "loss": 0.2223, "step": 48590 }, { "epoch": 0.8839738809363575, "grad_norm": 1.476023554801941, "learning_rate": 2.2792067256695985e-06, "loss": 0.2204, "step": 48600 }, { "epoch": 0.8841557685661798, "grad_norm": 1.5272600650787354, "learning_rate": 2.278342442592987e-06, "loss": 0.217, "step": 48610 }, { "epoch": 0.8843376561960021, "grad_norm": 1.6922507286071777, "learning_rate": 2.2774794510605424e-06, "loss": 0.2138, "step": 48620 }, { "epoch": 0.8845195438258244, "grad_norm": 1.6462255716323853, "learning_rate": 2.276617751371753e-06, "loss": 0.2136, "step": 48630 }, { "epoch": 0.8847014314556467, "grad_norm": 1.5136059522628784, "learning_rate": 2.2757573438256587e-06, "loss": 0.2273, "step": 48640 }, { "epoch": 0.884883319085469, "grad_norm": 1.5496056079864502, "learning_rate": 2.27489822872085e-06, "loss": 0.2132, "step": 48650 }, { "epoch": 0.8850652067152913, "grad_norm": 1.6040620803833008, "learning_rate": 2.2740404063554696e-06, "loss": 0.2139, "step": 48660 }, { "epoch": 0.8852470943451136, "grad_norm": 1.5355123281478882, "learning_rate": 2.273183877027212e-06, "loss": 0.2199, "step": 48670 }, { "epoch": 0.8854289819749359, "grad_norm": 1.5815424919128418, "learning_rate": 2.2723286410333234e-06, "loss": 0.2165, "step": 48680 }, { "epoch": 0.8856108696047582, "grad_norm": 1.50283944606781, "learning_rate": 2.2714746986705986e-06, "loss": 0.2174, "step": 48690 }, { "epoch": 0.8857927572345805, "grad_norm": 1.5871628522872925, "learning_rate": 2.270622050235386e-06, "loss": 0.2141, "step": 48700 }, { "epoch": 0.8859746448644028, "grad_norm": 1.5520185232162476, "learning_rate": 2.2697706960235842e-06, "loss": 0.2129, "step": 48710 }, { "epoch": 0.8861565324942251, "grad_norm": 1.5952019691467285, "learning_rate": 2.2689206363306434e-06, "loss": 0.2213, "step": 48720 }, { "epoch": 0.8863384201240474, "grad_norm": 1.9119555950164795, "learning_rate": 2.2680718714515627e-06, "loss": 0.2186, "step": 48730 }, { "epoch": 0.8865203077538697, "grad_norm": 1.5685157775878906, "learning_rate": 2.267224401680893e-06, "loss": 0.2186, "step": 48740 }, { "epoch": 0.886702195383692, "grad_norm": 1.4146095514297485, "learning_rate": 2.2663782273127355e-06, "loss": 0.2104, "step": 48750 }, { "epoch": 0.8868840830135143, "grad_norm": 1.5575730800628662, "learning_rate": 2.2655333486407427e-06, "loss": 0.2121, "step": 48760 }, { "epoch": 0.8870659706433366, "grad_norm": 1.7031009197235107, "learning_rate": 2.264689765958117e-06, "loss": 0.2128, "step": 48770 }, { "epoch": 0.8872478582731589, "grad_norm": 1.9377870559692383, "learning_rate": 2.263847479557609e-06, "loss": 0.2143, "step": 48780 }, { "epoch": 0.8874297459029812, "grad_norm": 1.6484438180923462, "learning_rate": 2.2630064897315242e-06, "loss": 0.2153, "step": 48790 }, { "epoch": 0.8876116335328035, "grad_norm": 1.4540479183197021, "learning_rate": 2.262166796771713e-06, "loss": 0.2185, "step": 48800 }, { "epoch": 0.8877935211626258, "grad_norm": 1.6640247106552124, "learning_rate": 2.2613284009695782e-06, "loss": 0.2239, "step": 48810 }, { "epoch": 0.8879754087924481, "grad_norm": 1.5256003141403198, "learning_rate": 2.260491302616073e-06, "loss": 0.2144, "step": 48820 }, { "epoch": 0.8881572964222704, "grad_norm": 1.5044360160827637, "learning_rate": 2.2596555020016995e-06, "loss": 0.2198, "step": 48830 }, { "epoch": 0.8883391840520927, "grad_norm": 1.3627371788024902, "learning_rate": 2.2588209994165093e-06, "loss": 0.2201, "step": 48840 }, { "epoch": 0.888521071681915, "grad_norm": 1.458184838294983, "learning_rate": 2.2579877951501032e-06, "loss": 0.2159, "step": 48850 }, { "epoch": 0.8887029593117373, "grad_norm": 1.5265560150146484, "learning_rate": 2.2571558894916316e-06, "loss": 0.2169, "step": 48860 }, { "epoch": 0.8888848469415596, "grad_norm": 1.4908156394958496, "learning_rate": 2.256325282729796e-06, "loss": 0.2194, "step": 48870 }, { "epoch": 0.8890667345713817, "grad_norm": 1.5431607961654663, "learning_rate": 2.2554959751528456e-06, "loss": 0.2156, "step": 48880 }, { "epoch": 0.889248622201204, "grad_norm": 1.7202205657958984, "learning_rate": 2.254667967048578e-06, "loss": 0.2144, "step": 48890 }, { "epoch": 0.8894305098310263, "grad_norm": 1.5927788019180298, "learning_rate": 2.25384125870434e-06, "loss": 0.2167, "step": 48900 }, { "epoch": 0.8896123974608486, "grad_norm": 1.442589521408081, "learning_rate": 2.25301585040703e-06, "loss": 0.2137, "step": 48910 }, { "epoch": 0.8897942850906709, "grad_norm": 1.5656592845916748, "learning_rate": 2.2521917424430924e-06, "loss": 0.2163, "step": 48920 }, { "epoch": 0.8899761727204932, "grad_norm": 1.7681816816329956, "learning_rate": 2.251368935098521e-06, "loss": 0.2101, "step": 48930 }, { "epoch": 0.8901580603503155, "grad_norm": 1.819075345993042, "learning_rate": 2.250547428658859e-06, "loss": 0.2181, "step": 48940 }, { "epoch": 0.8903399479801378, "grad_norm": 1.658298373222351, "learning_rate": 2.2497272234091965e-06, "loss": 0.2177, "step": 48950 }, { "epoch": 0.8905218356099601, "grad_norm": 1.5079752206802368, "learning_rate": 2.2489083196341744e-06, "loss": 0.2132, "step": 48960 }, { "epoch": 0.8907037232397824, "grad_norm": 1.6190603971481323, "learning_rate": 2.2480907176179806e-06, "loss": 0.213, "step": 48970 }, { "epoch": 0.8908856108696047, "grad_norm": 1.5183559656143188, "learning_rate": 2.2472744176443506e-06, "loss": 0.2123, "step": 48980 }, { "epoch": 0.891067498499427, "grad_norm": 1.514012098312378, "learning_rate": 2.2464594199965693e-06, "loss": 0.2135, "step": 48990 }, { "epoch": 0.8912493861292493, "grad_norm": 1.4223933219909668, "learning_rate": 2.2456457249574683e-06, "loss": 0.221, "step": 49000 }, { "epoch": 0.8914312737590716, "grad_norm": 1.4026131629943848, "learning_rate": 2.2448333328094296e-06, "loss": 0.2168, "step": 49010 }, { "epoch": 0.8916131613888939, "grad_norm": 1.4575672149658203, "learning_rate": 2.24402224383438e-06, "loss": 0.2161, "step": 49020 }, { "epoch": 0.8917950490187162, "grad_norm": 1.8208147287368774, "learning_rate": 2.2432124583137965e-06, "loss": 0.2214, "step": 49030 }, { "epoch": 0.8919769366485385, "grad_norm": 1.6939351558685303, "learning_rate": 2.242403976528702e-06, "loss": 0.2203, "step": 49040 }, { "epoch": 0.8921588242783608, "grad_norm": 1.835656762123108, "learning_rate": 2.2415967987596673e-06, "loss": 0.2129, "step": 49050 }, { "epoch": 0.8923407119081831, "grad_norm": 1.362054705619812, "learning_rate": 2.240790925286811e-06, "loss": 0.2156, "step": 49060 }, { "epoch": 0.8925225995380054, "grad_norm": 1.7379584312438965, "learning_rate": 2.2399863563898014e-06, "loss": 0.2148, "step": 49070 }, { "epoch": 0.8927044871678277, "grad_norm": 2.353651523590088, "learning_rate": 2.23918309234785e-06, "loss": 0.2152, "step": 49080 }, { "epoch": 0.89288637479765, "grad_norm": 1.5242611169815063, "learning_rate": 2.2383811334397178e-06, "loss": 0.2189, "step": 49090 }, { "epoch": 0.8930682624274723, "grad_norm": 1.609618902206421, "learning_rate": 2.2375804799437107e-06, "loss": 0.2124, "step": 49100 }, { "epoch": 0.8932501500572946, "grad_norm": 1.7054611444473267, "learning_rate": 2.236781132137685e-06, "loss": 0.2108, "step": 49110 }, { "epoch": 0.8934320376871169, "grad_norm": 1.6095832586288452, "learning_rate": 2.235983090299042e-06, "loss": 0.2164, "step": 49120 }, { "epoch": 0.8936139253169392, "grad_norm": 1.7654153108596802, "learning_rate": 2.2351863547047298e-06, "loss": 0.2163, "step": 49130 }, { "epoch": 0.8937958129467615, "grad_norm": 1.6826379299163818, "learning_rate": 2.2343909256312426e-06, "loss": 0.2179, "step": 49140 }, { "epoch": 0.8939777005765838, "grad_norm": 1.5546796321868896, "learning_rate": 2.2335968033546214e-06, "loss": 0.2127, "step": 49150 }, { "epoch": 0.8941595882064061, "grad_norm": 1.5013118982315063, "learning_rate": 2.232803988150456e-06, "loss": 0.2201, "step": 49160 }, { "epoch": 0.8943414758362284, "grad_norm": 1.4033942222595215, "learning_rate": 2.2320124802938793e-06, "loss": 0.2134, "step": 49170 }, { "epoch": 0.8945233634660507, "grad_norm": 1.9110299348831177, "learning_rate": 2.231222280059573e-06, "loss": 0.2168, "step": 49180 }, { "epoch": 0.894705251095873, "grad_norm": 1.4239728450775146, "learning_rate": 2.230433387721763e-06, "loss": 0.2053, "step": 49190 }, { "epoch": 0.8948871387256953, "grad_norm": 1.4018689393997192, "learning_rate": 2.229645803554222e-06, "loss": 0.2061, "step": 49200 }, { "epoch": 0.8950690263555175, "grad_norm": 1.7005221843719482, "learning_rate": 2.2288595278302704e-06, "loss": 0.2203, "step": 49210 }, { "epoch": 0.8952509139853398, "grad_norm": 1.638630986213684, "learning_rate": 2.228074560822772e-06, "loss": 0.2202, "step": 49220 }, { "epoch": 0.8954328016151621, "grad_norm": 1.3842589855194092, "learning_rate": 2.227290902804139e-06, "loss": 0.2178, "step": 49230 }, { "epoch": 0.8956146892449844, "grad_norm": 1.5513948202133179, "learning_rate": 2.2265085540463256e-06, "loss": 0.2122, "step": 49240 }, { "epoch": 0.8957965768748067, "grad_norm": 1.8409475088119507, "learning_rate": 2.2257275148208353e-06, "loss": 0.2156, "step": 49250 }, { "epoch": 0.895978464504629, "grad_norm": 1.4931317567825317, "learning_rate": 2.2249477853987154e-06, "loss": 0.2148, "step": 49260 }, { "epoch": 0.8961603521344513, "grad_norm": 1.798263669013977, "learning_rate": 2.2241693660505605e-06, "loss": 0.2183, "step": 49270 }, { "epoch": 0.8963422397642736, "grad_norm": 1.8095366954803467, "learning_rate": 2.223392257046507e-06, "loss": 0.22, "step": 49280 }, { "epoch": 0.8965241273940959, "grad_norm": 1.614747405052185, "learning_rate": 2.2226164586562395e-06, "loss": 0.2145, "step": 49290 }, { "epoch": 0.8967060150239182, "grad_norm": 1.6265164613723755, "learning_rate": 2.221841971148987e-06, "loss": 0.2189, "step": 49300 }, { "epoch": 0.8968879026537405, "grad_norm": 1.804789662361145, "learning_rate": 2.221068794793524e-06, "loss": 0.2185, "step": 49310 }, { "epoch": 0.8970697902835628, "grad_norm": 1.979533314704895, "learning_rate": 2.220296929858169e-06, "loss": 0.2213, "step": 49320 }, { "epoch": 0.8972516779133851, "grad_norm": 1.4259988069534302, "learning_rate": 2.2195263766107854e-06, "loss": 0.2193, "step": 49330 }, { "epoch": 0.8974335655432074, "grad_norm": 1.5908091068267822, "learning_rate": 2.218757135318782e-06, "loss": 0.2144, "step": 49340 }, { "epoch": 0.8976154531730297, "grad_norm": 1.604709506034851, "learning_rate": 2.217989206249112e-06, "loss": 0.2199, "step": 49350 }, { "epoch": 0.897797340802852, "grad_norm": 1.6444514989852905, "learning_rate": 2.2172225896682758e-06, "loss": 0.2148, "step": 49360 }, { "epoch": 0.8979792284326743, "grad_norm": 1.5288671255111694, "learning_rate": 2.216457285842313e-06, "loss": 0.212, "step": 49370 }, { "epoch": 0.8981611160624966, "grad_norm": 1.5035914182662964, "learning_rate": 2.2156932950368118e-06, "loss": 0.2148, "step": 49380 }, { "epoch": 0.8983430036923189, "grad_norm": 1.434545636177063, "learning_rate": 2.2149306175169037e-06, "loss": 0.2125, "step": 49390 }, { "epoch": 0.8985248913221412, "grad_norm": 1.7981197834014893, "learning_rate": 2.214169253547263e-06, "loss": 0.2148, "step": 49400 }, { "epoch": 0.8987067789519635, "grad_norm": 1.644718885421753, "learning_rate": 2.2134092033921105e-06, "loss": 0.2122, "step": 49410 }, { "epoch": 0.8988886665817858, "grad_norm": 1.4779000282287598, "learning_rate": 2.21265046731521e-06, "loss": 0.2211, "step": 49420 }, { "epoch": 0.8990705542116081, "grad_norm": 1.3953880071640015, "learning_rate": 2.2118930455798695e-06, "loss": 0.2132, "step": 49430 }, { "epoch": 0.8992524418414304, "grad_norm": 1.5219954252243042, "learning_rate": 2.211136938448939e-06, "loss": 0.2168, "step": 49440 }, { "epoch": 0.8994343294712527, "grad_norm": 1.4716719388961792, "learning_rate": 2.2103821461848144e-06, "loss": 0.2115, "step": 49450 }, { "epoch": 0.899616217101075, "grad_norm": 1.483673334121704, "learning_rate": 2.2096286690494363e-06, "loss": 0.2158, "step": 49460 }, { "epoch": 0.8997981047308973, "grad_norm": 1.4969524145126343, "learning_rate": 2.2088765073042857e-06, "loss": 0.2198, "step": 49470 }, { "epoch": 0.8999799923607196, "grad_norm": 1.7283666133880615, "learning_rate": 2.2081256612103897e-06, "loss": 0.2196, "step": 49480 }, { "epoch": 0.9001618799905419, "grad_norm": 1.7221510410308838, "learning_rate": 2.2073761310283177e-06, "loss": 0.2183, "step": 49490 }, { "epoch": 0.9003437676203642, "grad_norm": 1.8534198999404907, "learning_rate": 2.2066279170181825e-06, "loss": 0.2182, "step": 49500 }, { "epoch": 0.9005256552501865, "grad_norm": 1.750388264656067, "learning_rate": 2.2058810194396404e-06, "loss": 0.2115, "step": 49510 }, { "epoch": 0.9007075428800088, "grad_norm": 1.5831624269485474, "learning_rate": 2.205135438551891e-06, "loss": 0.2119, "step": 49520 }, { "epoch": 0.9008894305098311, "grad_norm": 1.7074732780456543, "learning_rate": 2.204391174613676e-06, "loss": 0.2238, "step": 49530 }, { "epoch": 0.9010713181396534, "grad_norm": 1.700405240058899, "learning_rate": 2.2036482278832814e-06, "loss": 0.2232, "step": 49540 }, { "epoch": 0.9012532057694757, "grad_norm": 1.6299225091934204, "learning_rate": 2.2029065986185358e-06, "loss": 0.2166, "step": 49550 }, { "epoch": 0.901435093399298, "grad_norm": 1.5532768964767456, "learning_rate": 2.202166287076809e-06, "loss": 0.2141, "step": 49560 }, { "epoch": 0.9016169810291202, "grad_norm": 1.5532926321029663, "learning_rate": 2.201427293515016e-06, "loss": 0.2128, "step": 49570 }, { "epoch": 0.9017988686589425, "grad_norm": 1.604063630104065, "learning_rate": 2.200689618189613e-06, "loss": 0.2166, "step": 49580 }, { "epoch": 0.9019807562887648, "grad_norm": 1.7669966220855713, "learning_rate": 2.1999532613565978e-06, "loss": 0.2148, "step": 49590 }, { "epoch": 0.9021626439185871, "grad_norm": 1.733925461769104, "learning_rate": 2.1992182232715124e-06, "loss": 0.2183, "step": 49600 }, { "epoch": 0.9023445315484094, "grad_norm": 1.5543144941329956, "learning_rate": 2.198484504189441e-06, "loss": 0.2159, "step": 49610 }, { "epoch": 0.9025264191782317, "grad_norm": 1.4838957786560059, "learning_rate": 2.1977521043650096e-06, "loss": 0.2213, "step": 49620 }, { "epoch": 0.902708306808054, "grad_norm": 1.9272964000701904, "learning_rate": 2.197021024052386e-06, "loss": 0.2127, "step": 49630 }, { "epoch": 0.9028901944378763, "grad_norm": 1.976576805114746, "learning_rate": 2.19629126350528e-06, "loss": 0.2156, "step": 49640 }, { "epoch": 0.9030720820676986, "grad_norm": 2.1740026473999023, "learning_rate": 2.195562822976944e-06, "loss": 0.2137, "step": 49650 }, { "epoch": 0.9032539696975209, "grad_norm": 1.5743670463562012, "learning_rate": 2.1948357027201722e-06, "loss": 0.2235, "step": 49660 }, { "epoch": 0.9034358573273432, "grad_norm": 1.604531168937683, "learning_rate": 2.1941099029873008e-06, "loss": 0.2196, "step": 49670 }, { "epoch": 0.9036177449571655, "grad_norm": 1.6129028797149658, "learning_rate": 2.193385424030207e-06, "loss": 0.2148, "step": 49680 }, { "epoch": 0.9037996325869878, "grad_norm": 1.6493451595306396, "learning_rate": 2.1926622661003106e-06, "loss": 0.2135, "step": 49690 }, { "epoch": 0.9039815202168101, "grad_norm": 1.5147918462753296, "learning_rate": 2.1919404294485723e-06, "loss": 0.2177, "step": 49700 }, { "epoch": 0.9041634078466324, "grad_norm": 1.4641519784927368, "learning_rate": 2.1912199143254943e-06, "loss": 0.211, "step": 49710 }, { "epoch": 0.9043452954764546, "grad_norm": 1.678545355796814, "learning_rate": 2.19050072098112e-06, "loss": 0.2157, "step": 49720 }, { "epoch": 0.9045271831062769, "grad_norm": 1.527331829071045, "learning_rate": 2.1897828496650353e-06, "loss": 0.2075, "step": 49730 }, { "epoch": 0.9047090707360992, "grad_norm": 1.6285661458969116, "learning_rate": 2.1890663006263657e-06, "loss": 0.2118, "step": 49740 }, { "epoch": 0.9048909583659215, "grad_norm": 1.500602126121521, "learning_rate": 2.188351074113779e-06, "loss": 0.2152, "step": 49750 }, { "epoch": 0.9050728459957438, "grad_norm": 1.562353491783142, "learning_rate": 2.1876371703754833e-06, "loss": 0.2135, "step": 49760 }, { "epoch": 0.9052547336255661, "grad_norm": 1.4914551973342896, "learning_rate": 2.1869245896592274e-06, "loss": 0.2236, "step": 49770 }, { "epoch": 0.9054366212553884, "grad_norm": 1.703267216682434, "learning_rate": 2.186213332212303e-06, "loss": 0.2221, "step": 49780 }, { "epoch": 0.9056185088852107, "grad_norm": 1.5058627128601074, "learning_rate": 2.1855033982815397e-06, "loss": 0.2115, "step": 49790 }, { "epoch": 0.905800396515033, "grad_norm": 1.6569452285766602, "learning_rate": 2.1847947881133093e-06, "loss": 0.2174, "step": 49800 }, { "epoch": 0.9059822841448553, "grad_norm": 1.5235955715179443, "learning_rate": 2.1840875019535256e-06, "loss": 0.2123, "step": 49810 }, { "epoch": 0.9061641717746776, "grad_norm": 1.3318418264389038, "learning_rate": 2.1833815400476393e-06, "loss": 0.2153, "step": 49820 }, { "epoch": 0.9063460594044999, "grad_norm": 1.482475996017456, "learning_rate": 2.1826769026406446e-06, "loss": 0.2143, "step": 49830 }, { "epoch": 0.9065279470343222, "grad_norm": 1.793595790863037, "learning_rate": 2.181973589977075e-06, "loss": 0.2128, "step": 49840 }, { "epoch": 0.9067098346641445, "grad_norm": 1.4660427570343018, "learning_rate": 2.1812716023010057e-06, "loss": 0.2183, "step": 49850 }, { "epoch": 0.9068917222939668, "grad_norm": 1.4744678735733032, "learning_rate": 2.180570939856049e-06, "loss": 0.2111, "step": 49860 }, { "epoch": 0.907073609923789, "grad_norm": 1.9802756309509277, "learning_rate": 2.1798716028853593e-06, "loss": 0.2167, "step": 49870 }, { "epoch": 0.9072554975536113, "grad_norm": 1.7732036113739014, "learning_rate": 2.179173591631631e-06, "loss": 0.2196, "step": 49880 }, { "epoch": 0.9074373851834336, "grad_norm": 1.723799228668213, "learning_rate": 2.178476906337098e-06, "loss": 0.218, "step": 49890 }, { "epoch": 0.9076192728132559, "grad_norm": 1.45048987865448, "learning_rate": 2.1777815472435353e-06, "loss": 0.2187, "step": 49900 }, { "epoch": 0.9078011604430782, "grad_norm": 1.4676223993301392, "learning_rate": 2.1770875145922557e-06, "loss": 0.2188, "step": 49910 }, { "epoch": 0.9079830480729005, "grad_norm": 1.4413344860076904, "learning_rate": 2.1763948086241128e-06, "loss": 0.2146, "step": 49920 }, { "epoch": 0.9081649357027228, "grad_norm": 1.7415993213653564, "learning_rate": 2.1757034295794995e-06, "loss": 0.2104, "step": 49930 }, { "epoch": 0.9083468233325451, "grad_norm": 1.691911220550537, "learning_rate": 2.1750133776983474e-06, "loss": 0.2212, "step": 49940 }, { "epoch": 0.9085287109623674, "grad_norm": 1.688471794128418, "learning_rate": 2.1743246532201304e-06, "loss": 0.2178, "step": 49950 }, { "epoch": 0.9087105985921897, "grad_norm": 1.4334405660629272, "learning_rate": 2.1736372563838583e-06, "loss": 0.2175, "step": 49960 }, { "epoch": 0.908892486222012, "grad_norm": 1.3679026365280151, "learning_rate": 2.1729511874280822e-06, "loss": 0.2157, "step": 49970 }, { "epoch": 0.9090743738518343, "grad_norm": 1.599624752998352, "learning_rate": 2.1722664465908914e-06, "loss": 0.2193, "step": 49980 }, { "epoch": 0.9092562614816566, "grad_norm": 1.6477738618850708, "learning_rate": 2.1715830341099146e-06, "loss": 0.2158, "step": 49990 }, { "epoch": 0.9094381491114789, "grad_norm": 1.4978662729263306, "learning_rate": 2.1709009502223204e-06, "loss": 0.2229, "step": 50000 }, { "epoch": 0.9096200367413012, "grad_norm": 1.4136052131652832, "learning_rate": 2.170220195164815e-06, "loss": 0.2118, "step": 50010 }, { "epoch": 0.9098019243711235, "grad_norm": 1.7001320123672485, "learning_rate": 2.1695407691736444e-06, "loss": 0.2156, "step": 50020 }, { "epoch": 0.9099838120009458, "grad_norm": 1.6481040716171265, "learning_rate": 2.168862672484592e-06, "loss": 0.2127, "step": 50030 }, { "epoch": 0.9101656996307681, "grad_norm": 1.5712233781814575, "learning_rate": 2.1681859053329816e-06, "loss": 0.2165, "step": 50040 }, { "epoch": 0.9103475872605904, "grad_norm": 1.571766972541809, "learning_rate": 2.1675104679536748e-06, "loss": 0.2124, "step": 50050 }, { "epoch": 0.9105294748904127, "grad_norm": 1.7027416229248047, "learning_rate": 2.1668363605810726e-06, "loss": 0.2167, "step": 50060 }, { "epoch": 0.910711362520235, "grad_norm": 1.6050999164581299, "learning_rate": 2.166163583449112e-06, "loss": 0.2147, "step": 50070 }, { "epoch": 0.9108932501500573, "grad_norm": 1.7384392023086548, "learning_rate": 2.1654921367912703e-06, "loss": 0.212, "step": 50080 }, { "epoch": 0.9110751377798796, "grad_norm": 1.5813875198364258, "learning_rate": 2.164822020840564e-06, "loss": 0.2181, "step": 50090 }, { "epoch": 0.9112570254097019, "grad_norm": 1.4210996627807617, "learning_rate": 2.1641532358295454e-06, "loss": 0.2125, "step": 50100 }, { "epoch": 0.9114389130395242, "grad_norm": 1.4332221746444702, "learning_rate": 2.163485781990307e-06, "loss": 0.2153, "step": 50110 }, { "epoch": 0.9116208006693465, "grad_norm": 1.6941217184066772, "learning_rate": 2.1628196595544776e-06, "loss": 0.2147, "step": 50120 }, { "epoch": 0.9118026882991688, "grad_norm": 1.8930811882019043, "learning_rate": 2.1621548687532245e-06, "loss": 0.2159, "step": 50130 }, { "epoch": 0.9119845759289911, "grad_norm": 1.9075202941894531, "learning_rate": 2.1614914098172543e-06, "loss": 0.2174, "step": 50140 }, { "epoch": 0.9121664635588134, "grad_norm": 1.560972809791565, "learning_rate": 2.1608292829768097e-06, "loss": 0.2235, "step": 50150 }, { "epoch": 0.9123483511886357, "grad_norm": 1.5418338775634766, "learning_rate": 2.1601684884616708e-06, "loss": 0.2163, "step": 50160 }, { "epoch": 0.912530238818458, "grad_norm": 1.820219874382019, "learning_rate": 2.1595090265011574e-06, "loss": 0.2152, "step": 50170 }, { "epoch": 0.9127121264482803, "grad_norm": 1.587208867073059, "learning_rate": 2.158850897324126e-06, "loss": 0.2198, "step": 50180 }, { "epoch": 0.9128940140781026, "grad_norm": 1.720674991607666, "learning_rate": 2.158194101158968e-06, "loss": 0.2203, "step": 50190 }, { "epoch": 0.9130759017079249, "grad_norm": 1.4921519756317139, "learning_rate": 2.157538638233617e-06, "loss": 0.2117, "step": 50200 }, { "epoch": 0.9132577893377472, "grad_norm": 2.011241912841797, "learning_rate": 2.1568845087755406e-06, "loss": 0.2163, "step": 50210 }, { "epoch": 0.9134396769675694, "grad_norm": 1.6936180591583252, "learning_rate": 2.1562317130117436e-06, "loss": 0.2175, "step": 50220 }, { "epoch": 0.9136215645973917, "grad_norm": 1.6963741779327393, "learning_rate": 2.1555802511687694e-06, "loss": 0.2201, "step": 50230 }, { "epoch": 0.913803452227214, "grad_norm": 1.713157296180725, "learning_rate": 2.1549301234726978e-06, "loss": 0.2201, "step": 50240 }, { "epoch": 0.9139853398570363, "grad_norm": 1.4089436531066895, "learning_rate": 2.154281330149145e-06, "loss": 0.2161, "step": 50250 }, { "epoch": 0.9141672274868586, "grad_norm": 1.4240070581436157, "learning_rate": 2.153633871423266e-06, "loss": 0.2093, "step": 50260 }, { "epoch": 0.9143491151166809, "grad_norm": 1.7982630729675293, "learning_rate": 2.1529877475197505e-06, "loss": 0.2224, "step": 50270 }, { "epoch": 0.9145310027465032, "grad_norm": 1.6001269817352295, "learning_rate": 2.1523429586628256e-06, "loss": 0.2158, "step": 50280 }, { "epoch": 0.9147128903763255, "grad_norm": 1.4012833833694458, "learning_rate": 2.151699505076256e-06, "loss": 0.2077, "step": 50290 }, { "epoch": 0.9148947780061478, "grad_norm": 1.573975682258606, "learning_rate": 2.151057386983343e-06, "loss": 0.2117, "step": 50300 }, { "epoch": 0.9150766656359701, "grad_norm": 1.42015540599823, "learning_rate": 2.1504166046069227e-06, "loss": 0.214, "step": 50310 }, { "epoch": 0.9152585532657924, "grad_norm": 1.2575422525405884, "learning_rate": 2.149777158169369e-06, "loss": 0.2131, "step": 50320 }, { "epoch": 0.9154404408956147, "grad_norm": 1.4981709718704224, "learning_rate": 2.149139047892592e-06, "loss": 0.2112, "step": 50330 }, { "epoch": 0.915622328525437, "grad_norm": 1.596022129058838, "learning_rate": 2.1485022739980385e-06, "loss": 0.2162, "step": 50340 }, { "epoch": 0.9158042161552593, "grad_norm": 1.395256757736206, "learning_rate": 2.147866836706691e-06, "loss": 0.2091, "step": 50350 }, { "epoch": 0.9159861037850816, "grad_norm": 1.4649969339370728, "learning_rate": 2.1472327362390684e-06, "loss": 0.2197, "step": 50360 }, { "epoch": 0.9161679914149039, "grad_norm": 1.6330574750900269, "learning_rate": 2.1465999728152254e-06, "loss": 0.2159, "step": 50370 }, { "epoch": 0.9163498790447262, "grad_norm": 1.4886186122894287, "learning_rate": 2.145968546654752e-06, "loss": 0.2127, "step": 50380 }, { "epoch": 0.9165317666745485, "grad_norm": 1.5352659225463867, "learning_rate": 2.1453384579767756e-06, "loss": 0.2163, "step": 50390 }, { "epoch": 0.9167136543043708, "grad_norm": 1.5954769849777222, "learning_rate": 2.14470970699996e-06, "loss": 0.2118, "step": 50400 }, { "epoch": 0.9168955419341931, "grad_norm": 1.6603959798812866, "learning_rate": 2.144082293942502e-06, "loss": 0.2198, "step": 50410 }, { "epoch": 0.9170774295640154, "grad_norm": 1.792300820350647, "learning_rate": 2.1434562190221355e-06, "loss": 0.2207, "step": 50420 }, { "epoch": 0.9172593171938377, "grad_norm": 1.5343786478042603, "learning_rate": 2.142831482456131e-06, "loss": 0.2128, "step": 50430 }, { "epoch": 0.91744120482366, "grad_norm": 1.775041937828064, "learning_rate": 2.1422080844612943e-06, "loss": 0.222, "step": 50440 }, { "epoch": 0.9176230924534823, "grad_norm": 1.5173028707504272, "learning_rate": 2.141586025253965e-06, "loss": 0.2142, "step": 50450 }, { "epoch": 0.9178049800833046, "grad_norm": 1.5705782175064087, "learning_rate": 2.1409653050500196e-06, "loss": 0.2184, "step": 50460 }, { "epoch": 0.9179868677131269, "grad_norm": 1.8471271991729736, "learning_rate": 2.14034592406487e-06, "loss": 0.2181, "step": 50470 }, { "epoch": 0.9181687553429492, "grad_norm": 1.6853325366973877, "learning_rate": 2.1397278825134614e-06, "loss": 0.2152, "step": 50480 }, { "epoch": 0.9183506429727715, "grad_norm": 1.731371283531189, "learning_rate": 2.139111180610278e-06, "loss": 0.2103, "step": 50490 }, { "epoch": 0.9185325306025938, "grad_norm": 1.5057346820831299, "learning_rate": 2.1384958185693343e-06, "loss": 0.2118, "step": 50500 }, { "epoch": 0.9187144182324161, "grad_norm": 1.715783715248108, "learning_rate": 2.137881796604184e-06, "loss": 0.2096, "step": 50510 }, { "epoch": 0.9188963058622384, "grad_norm": 1.4689819812774658, "learning_rate": 2.137269114927913e-06, "loss": 0.2144, "step": 50520 }, { "epoch": 0.9190781934920607, "grad_norm": 1.5270822048187256, "learning_rate": 2.1366577737531434e-06, "loss": 0.2083, "step": 50530 }, { "epoch": 0.919260081121883, "grad_norm": 1.469880223274231, "learning_rate": 2.1360477732920323e-06, "loss": 0.216, "step": 50540 }, { "epoch": 0.9194419687517053, "grad_norm": 1.5062416791915894, "learning_rate": 2.13543911375627e-06, "loss": 0.2097, "step": 50550 }, { "epoch": 0.9196238563815274, "grad_norm": 1.9055472612380981, "learning_rate": 2.134831795357083e-06, "loss": 0.2199, "step": 50560 }, { "epoch": 0.9198057440113497, "grad_norm": 1.734017014503479, "learning_rate": 2.1342258183052322e-06, "loss": 0.2198, "step": 50570 }, { "epoch": 0.919987631641172, "grad_norm": 1.5844762325286865, "learning_rate": 2.1336211828110114e-06, "loss": 0.2232, "step": 50580 }, { "epoch": 0.9201695192709943, "grad_norm": 1.8815439939498901, "learning_rate": 2.133017889084252e-06, "loss": 0.2131, "step": 50590 }, { "epoch": 0.9203514069008166, "grad_norm": 1.6211276054382324, "learning_rate": 2.1324159373343165e-06, "loss": 0.2181, "step": 50600 }, { "epoch": 0.9205332945306389, "grad_norm": 1.410794973373413, "learning_rate": 2.1318153277701027e-06, "loss": 0.2153, "step": 50610 }, { "epoch": 0.9207151821604612, "grad_norm": 1.7097445726394653, "learning_rate": 2.1312160606000435e-06, "loss": 0.211, "step": 50620 }, { "epoch": 0.9208970697902835, "grad_norm": 1.807395339012146, "learning_rate": 2.1306181360321056e-06, "loss": 0.2154, "step": 50630 }, { "epoch": 0.9210789574201058, "grad_norm": 1.598022699356079, "learning_rate": 2.13002155427379e-06, "loss": 0.2153, "step": 50640 }, { "epoch": 0.9212608450499281, "grad_norm": 1.971874475479126, "learning_rate": 2.12942631553213e-06, "loss": 0.2192, "step": 50650 }, { "epoch": 0.9214427326797504, "grad_norm": 1.6141102313995361, "learning_rate": 2.1288324200136945e-06, "loss": 0.2187, "step": 50660 }, { "epoch": 0.9216246203095727, "grad_norm": 1.7471818923950195, "learning_rate": 2.1282398679245857e-06, "loss": 0.2109, "step": 50670 }, { "epoch": 0.921806507939395, "grad_norm": 1.6613678932189941, "learning_rate": 2.1276486594704406e-06, "loss": 0.2122, "step": 50680 }, { "epoch": 0.9219883955692173, "grad_norm": 1.5311447381973267, "learning_rate": 2.127058794856429e-06, "loss": 0.2094, "step": 50690 }, { "epoch": 0.9221702831990396, "grad_norm": 1.5147730112075806, "learning_rate": 2.126470274287253e-06, "loss": 0.2157, "step": 50700 }, { "epoch": 0.9223521708288619, "grad_norm": 1.511496663093567, "learning_rate": 2.12588309796715e-06, "loss": 0.2096, "step": 50710 }, { "epoch": 0.9225340584586842, "grad_norm": 1.4777886867523193, "learning_rate": 2.1252972660998913e-06, "loss": 0.2173, "step": 50720 }, { "epoch": 0.9227159460885065, "grad_norm": 1.5351269245147705, "learning_rate": 2.1247127788887804e-06, "loss": 0.2138, "step": 50730 }, { "epoch": 0.9228978337183288, "grad_norm": 1.778604507446289, "learning_rate": 2.1241296365366547e-06, "loss": 0.2119, "step": 50740 }, { "epoch": 0.9230797213481511, "grad_norm": 1.457931637763977, "learning_rate": 2.1235478392458857e-06, "loss": 0.2188, "step": 50750 }, { "epoch": 0.9232616089779734, "grad_norm": 1.511454701423645, "learning_rate": 2.1229673872183763e-06, "loss": 0.2158, "step": 50760 }, { "epoch": 0.9234434966077957, "grad_norm": 1.7420461177825928, "learning_rate": 2.1223882806555625e-06, "loss": 0.213, "step": 50770 }, { "epoch": 0.923625384237618, "grad_norm": 1.3843843936920166, "learning_rate": 2.121810519758416e-06, "loss": 0.2097, "step": 50780 }, { "epoch": 0.9238072718674403, "grad_norm": 1.6379494667053223, "learning_rate": 2.1212341047274397e-06, "loss": 0.2169, "step": 50790 }, { "epoch": 0.9239891594972626, "grad_norm": 1.3711596727371216, "learning_rate": 2.1206590357626685e-06, "loss": 0.2146, "step": 50800 }, { "epoch": 0.9241710471270849, "grad_norm": 1.5877991914749146, "learning_rate": 2.120085313063672e-06, "loss": 0.211, "step": 50810 }, { "epoch": 0.9243529347569072, "grad_norm": 1.3405870199203491, "learning_rate": 2.1195129368295516e-06, "loss": 0.2129, "step": 50820 }, { "epoch": 0.9245348223867295, "grad_norm": 1.587302803993225, "learning_rate": 2.118941907258942e-06, "loss": 0.209, "step": 50830 }, { "epoch": 0.9247167100165518, "grad_norm": 1.5440970659255981, "learning_rate": 2.1183722245500106e-06, "loss": 0.2139, "step": 50840 }, { "epoch": 0.924898597646374, "grad_norm": 1.5478769540786743, "learning_rate": 2.117803888900457e-06, "loss": 0.2194, "step": 50850 }, { "epoch": 0.9250804852761964, "grad_norm": 1.4575668573379517, "learning_rate": 2.117236900507512e-06, "loss": 0.2158, "step": 50860 }, { "epoch": 0.9252623729060186, "grad_norm": 1.5289074182510376, "learning_rate": 2.1166712595679407e-06, "loss": 0.212, "step": 50870 }, { "epoch": 0.925444260535841, "grad_norm": 1.5669968128204346, "learning_rate": 2.1161069662780415e-06, "loss": 0.2126, "step": 50880 }, { "epoch": 0.9256261481656632, "grad_norm": 1.4298211336135864, "learning_rate": 2.115544020833643e-06, "loss": 0.2068, "step": 50890 }, { "epoch": 0.9258080357954855, "grad_norm": 1.8070645332336426, "learning_rate": 2.1149824234301055e-06, "loss": 0.2208, "step": 50900 }, { "epoch": 0.9259899234253078, "grad_norm": 1.7224944829940796, "learning_rate": 2.1144221742623246e-06, "loss": 0.2145, "step": 50910 }, { "epoch": 0.9261718110551301, "grad_norm": 1.820246696472168, "learning_rate": 2.1138632735247253e-06, "loss": 0.213, "step": 50920 }, { "epoch": 0.9263536986849524, "grad_norm": 1.6659518480300903, "learning_rate": 2.1133057214112655e-06, "loss": 0.2094, "step": 50930 }, { "epoch": 0.9265355863147747, "grad_norm": 1.855299472808838, "learning_rate": 2.1127495181154353e-06, "loss": 0.2128, "step": 50940 }, { "epoch": 0.926717473944597, "grad_norm": 1.6690791845321655, "learning_rate": 2.112194663830257e-06, "loss": 0.2085, "step": 50950 }, { "epoch": 0.9268993615744193, "grad_norm": 1.338478446006775, "learning_rate": 2.1116411587482836e-06, "loss": 0.212, "step": 50960 }, { "epoch": 0.9270812492042416, "grad_norm": 1.7496953010559082, "learning_rate": 2.1110890030616e-06, "loss": 0.2174, "step": 50970 }, { "epoch": 0.9272631368340639, "grad_norm": 1.7673001289367676, "learning_rate": 2.1105381969618246e-06, "loss": 0.2115, "step": 50980 }, { "epoch": 0.9274450244638862, "grad_norm": 1.7639652490615845, "learning_rate": 2.1099887406401053e-06, "loss": 0.2149, "step": 50990 }, { "epoch": 0.9276269120937085, "grad_norm": 1.4270669221878052, "learning_rate": 2.109440634287123e-06, "loss": 0.2154, "step": 51000 }, { "epoch": 0.9278087997235308, "grad_norm": 1.5078656673431396, "learning_rate": 2.1088938780930896e-06, "loss": 0.2134, "step": 51010 }, { "epoch": 0.9279906873533531, "grad_norm": 1.7076771259307861, "learning_rate": 2.1083484722477485e-06, "loss": 0.2121, "step": 51020 }, { "epoch": 0.9281725749831754, "grad_norm": 1.4566940069198608, "learning_rate": 2.107804416940374e-06, "loss": 0.2166, "step": 51030 }, { "epoch": 0.9283544626129977, "grad_norm": 1.6206134557724, "learning_rate": 2.1072617123597723e-06, "loss": 0.214, "step": 51040 }, { "epoch": 0.92853635024282, "grad_norm": 1.569254994392395, "learning_rate": 2.106720358694281e-06, "loss": 0.2165, "step": 51050 }, { "epoch": 0.9287182378726423, "grad_norm": 1.5295567512512207, "learning_rate": 2.106180356131769e-06, "loss": 0.2167, "step": 51060 }, { "epoch": 0.9289001255024646, "grad_norm": 1.455169916152954, "learning_rate": 2.105641704859634e-06, "loss": 0.2109, "step": 51070 }, { "epoch": 0.9290820131322869, "grad_norm": 1.341208815574646, "learning_rate": 2.1051044050648088e-06, "loss": 0.2119, "step": 51080 }, { "epoch": 0.9292639007621092, "grad_norm": 1.560876488685608, "learning_rate": 2.104568456933754e-06, "loss": 0.2156, "step": 51090 }, { "epoch": 0.9294457883919315, "grad_norm": 1.4910073280334473, "learning_rate": 2.1040338606524624e-06, "loss": 0.2157, "step": 51100 }, { "epoch": 0.9296276760217538, "grad_norm": 1.9762778282165527, "learning_rate": 2.103500616406457e-06, "loss": 0.215, "step": 51110 }, { "epoch": 0.9298095636515761, "grad_norm": 1.7921026945114136, "learning_rate": 2.1029687243807926e-06, "loss": 0.2127, "step": 51120 }, { "epoch": 0.9299914512813984, "grad_norm": 1.4947278499603271, "learning_rate": 2.1024381847600545e-06, "loss": 0.2161, "step": 51130 }, { "epoch": 0.9301733389112207, "grad_norm": 1.5096460580825806, "learning_rate": 2.101908997728358e-06, "loss": 0.2183, "step": 51140 }, { "epoch": 0.930355226541043, "grad_norm": 1.353597640991211, "learning_rate": 2.101381163469349e-06, "loss": 0.212, "step": 51150 }, { "epoch": 0.9305371141708653, "grad_norm": 1.5256788730621338, "learning_rate": 2.100854682166204e-06, "loss": 0.213, "step": 51160 }, { "epoch": 0.9307190018006876, "grad_norm": 1.5380598306655884, "learning_rate": 2.100329554001631e-06, "loss": 0.2113, "step": 51170 }, { "epoch": 0.9309008894305099, "grad_norm": 2.173736333847046, "learning_rate": 2.0998057791578676e-06, "loss": 0.2151, "step": 51180 }, { "epoch": 0.9310827770603322, "grad_norm": 1.7145205736160278, "learning_rate": 2.0992833578166813e-06, "loss": 0.2159, "step": 51190 }, { "epoch": 0.9312646646901545, "grad_norm": 1.564596176147461, "learning_rate": 2.098762290159371e-06, "loss": 0.2099, "step": 51200 }, { "epoch": 0.9314465523199768, "grad_norm": 1.4899507761001587, "learning_rate": 2.0982425763667644e-06, "loss": 0.2166, "step": 51210 }, { "epoch": 0.931628439949799, "grad_norm": 1.5956790447235107, "learning_rate": 2.0977242166192218e-06, "loss": 0.2193, "step": 51220 }, { "epoch": 0.9318103275796213, "grad_norm": 1.4445897340774536, "learning_rate": 2.0972072110966304e-06, "loss": 0.2129, "step": 51230 }, { "epoch": 0.9319922152094436, "grad_norm": 1.7969353199005127, "learning_rate": 2.0966915599784096e-06, "loss": 0.2151, "step": 51240 }, { "epoch": 0.9321741028392659, "grad_norm": 1.7442430257797241, "learning_rate": 2.0961772634435087e-06, "loss": 0.2111, "step": 51250 }, { "epoch": 0.9323559904690882, "grad_norm": 1.5903836488723755, "learning_rate": 2.095664321670405e-06, "loss": 0.2122, "step": 51260 }, { "epoch": 0.9325378780989105, "grad_norm": 1.491302728652954, "learning_rate": 2.095152734837109e-06, "loss": 0.2194, "step": 51270 }, { "epoch": 0.9327197657287328, "grad_norm": 2.2985031604766846, "learning_rate": 2.0946425031211574e-06, "loss": 0.2142, "step": 51280 }, { "epoch": 0.9329016533585551, "grad_norm": 1.6045117378234863, "learning_rate": 2.0941336266996194e-06, "loss": 0.2108, "step": 51290 }, { "epoch": 0.9330835409883774, "grad_norm": 1.8570703268051147, "learning_rate": 2.0936261057490925e-06, "loss": 0.2197, "step": 51300 }, { "epoch": 0.9332654286181997, "grad_norm": 1.6113274097442627, "learning_rate": 2.0931199404457035e-06, "loss": 0.2105, "step": 51310 }, { "epoch": 0.933447316248022, "grad_norm": 1.956642985343933, "learning_rate": 2.0926151309651095e-06, "loss": 0.211, "step": 51320 }, { "epoch": 0.9336292038778443, "grad_norm": 1.486788034439087, "learning_rate": 2.092111677482498e-06, "loss": 0.2184, "step": 51330 }, { "epoch": 0.9338110915076666, "grad_norm": 1.6131575107574463, "learning_rate": 2.0916095801725834e-06, "loss": 0.2131, "step": 51340 }, { "epoch": 0.9339929791374889, "grad_norm": 1.7520188093185425, "learning_rate": 2.091108839209611e-06, "loss": 0.2159, "step": 51350 }, { "epoch": 0.9341748667673112, "grad_norm": 1.5126012563705444, "learning_rate": 2.0906094547673565e-06, "loss": 0.2093, "step": 51360 }, { "epoch": 0.9343567543971335, "grad_norm": 1.4427599906921387, "learning_rate": 2.0901114270191224e-06, "loss": 0.2132, "step": 51370 }, { "epoch": 0.9345386420269558, "grad_norm": 1.636307716369629, "learning_rate": 2.0896147561377416e-06, "loss": 0.2134, "step": 51380 }, { "epoch": 0.934720529656778, "grad_norm": 1.5792710781097412, "learning_rate": 2.089119442295577e-06, "loss": 0.2116, "step": 51390 }, { "epoch": 0.9349024172866003, "grad_norm": 1.3742486238479614, "learning_rate": 2.0886254856645188e-06, "loss": 0.2145, "step": 51400 }, { "epoch": 0.9350843049164226, "grad_norm": 1.7296615839004517, "learning_rate": 2.0881328864159872e-06, "loss": 0.2095, "step": 51410 }, { "epoch": 0.9352661925462449, "grad_norm": 1.614981770515442, "learning_rate": 2.0876416447209313e-06, "loss": 0.2104, "step": 51420 }, { "epoch": 0.9354480801760672, "grad_norm": 1.5700891017913818, "learning_rate": 2.08715176074983e-06, "loss": 0.217, "step": 51430 }, { "epoch": 0.9356299678058895, "grad_norm": 1.7664406299591064, "learning_rate": 2.0866632346726883e-06, "loss": 0.21, "step": 51440 }, { "epoch": 0.9358118554357118, "grad_norm": 1.7216483354568481, "learning_rate": 2.0861760666590427e-06, "loss": 0.2138, "step": 51450 }, { "epoch": 0.9359937430655341, "grad_norm": 1.4654746055603027, "learning_rate": 2.0856902568779573e-06, "loss": 0.2147, "step": 51460 }, { "epoch": 0.9361756306953564, "grad_norm": 1.5052059888839722, "learning_rate": 2.0852058054980246e-06, "loss": 0.2147, "step": 51470 }, { "epoch": 0.9363575183251787, "grad_norm": 1.363581895828247, "learning_rate": 2.0847227126873667e-06, "loss": 0.2119, "step": 51480 }, { "epoch": 0.936539405955001, "grad_norm": 1.5730769634246826, "learning_rate": 2.0842409786136327e-06, "loss": 0.2141, "step": 51490 }, { "epoch": 0.9367212935848233, "grad_norm": 1.4292107820510864, "learning_rate": 2.0837606034440017e-06, "loss": 0.22, "step": 51500 }, { "epoch": 0.9369031812146456, "grad_norm": 1.4936333894729614, "learning_rate": 2.0832815873451793e-06, "loss": 0.2153, "step": 51510 }, { "epoch": 0.9370850688444679, "grad_norm": 1.5245020389556885, "learning_rate": 2.082803930483403e-06, "loss": 0.2159, "step": 51520 }, { "epoch": 0.9372669564742901, "grad_norm": 2.0585343837738037, "learning_rate": 2.0823276330244345e-06, "loss": 0.2162, "step": 51530 }, { "epoch": 0.9374488441041124, "grad_norm": 1.591489315032959, "learning_rate": 2.0818526951335656e-06, "loss": 0.2133, "step": 51540 }, { "epoch": 0.9376307317339347, "grad_norm": 1.4648133516311646, "learning_rate": 2.0813791169756165e-06, "loss": 0.2162, "step": 51550 }, { "epoch": 0.937812619363757, "grad_norm": 1.5270408391952515, "learning_rate": 2.0809068987149357e-06, "loss": 0.2101, "step": 51560 }, { "epoch": 0.9379945069935793, "grad_norm": 1.6623165607452393, "learning_rate": 2.0804360405153983e-06, "loss": 0.2145, "step": 51570 }, { "epoch": 0.9381763946234016, "grad_norm": 1.9897657632827759, "learning_rate": 2.07996654254041e-06, "loss": 0.2203, "step": 51580 }, { "epoch": 0.9383582822532239, "grad_norm": 1.554089069366455, "learning_rate": 2.079498404952901e-06, "loss": 0.2154, "step": 51590 }, { "epoch": 0.9385401698830462, "grad_norm": 1.8149625062942505, "learning_rate": 2.0790316279153326e-06, "loss": 0.2152, "step": 51600 }, { "epoch": 0.9387220575128685, "grad_norm": 1.5197043418884277, "learning_rate": 2.078566211589692e-06, "loss": 0.2181, "step": 51610 }, { "epoch": 0.9389039451426908, "grad_norm": 1.3645704984664917, "learning_rate": 2.078102156137495e-06, "loss": 0.2127, "step": 51620 }, { "epoch": 0.9390858327725131, "grad_norm": 1.9688973426818848, "learning_rate": 2.077639461719785e-06, "loss": 0.2211, "step": 51630 }, { "epoch": 0.9392677204023354, "grad_norm": 1.6355109214782715, "learning_rate": 2.077178128497133e-06, "loss": 0.2142, "step": 51640 }, { "epoch": 0.9394496080321577, "grad_norm": 1.405094861984253, "learning_rate": 2.0767181566296375e-06, "loss": 0.2167, "step": 51650 }, { "epoch": 0.93963149566198, "grad_norm": 1.9590537548065186, "learning_rate": 2.076259546276924e-06, "loss": 0.2148, "step": 51660 }, { "epoch": 0.9398133832918023, "grad_norm": 1.3730247020721436, "learning_rate": 2.0758022975981474e-06, "loss": 0.2129, "step": 51670 }, { "epoch": 0.9399952709216246, "grad_norm": 1.6191961765289307, "learning_rate": 2.0753464107519885e-06, "loss": 0.2145, "step": 51680 }, { "epoch": 0.9401771585514469, "grad_norm": 1.634648323059082, "learning_rate": 2.074891885896656e-06, "loss": 0.2195, "step": 51690 }, { "epoch": 0.9403590461812692, "grad_norm": 1.5185437202453613, "learning_rate": 2.074438723189884e-06, "loss": 0.216, "step": 51700 }, { "epoch": 0.9405409338110915, "grad_norm": 1.5957409143447876, "learning_rate": 2.0739869227889385e-06, "loss": 0.2145, "step": 51710 }, { "epoch": 0.9407228214409138, "grad_norm": 1.7346270084381104, "learning_rate": 2.0735364848506078e-06, "loss": 0.2148, "step": 51720 }, { "epoch": 0.9409047090707361, "grad_norm": 1.8671939373016357, "learning_rate": 2.0730874095312104e-06, "loss": 0.2127, "step": 51730 }, { "epoch": 0.9410865967005584, "grad_norm": 1.7579286098480225, "learning_rate": 2.072639696986591e-06, "loss": 0.2146, "step": 51740 }, { "epoch": 0.9412684843303807, "grad_norm": 1.5470640659332275, "learning_rate": 2.072193347372121e-06, "loss": 0.2152, "step": 51750 }, { "epoch": 0.941450371960203, "grad_norm": 1.5657879114151, "learning_rate": 2.0717483608426988e-06, "loss": 0.216, "step": 51760 }, { "epoch": 0.9416322595900253, "grad_norm": 1.6744608879089355, "learning_rate": 2.071304737552751e-06, "loss": 0.2173, "step": 51770 }, { "epoch": 0.9418141472198476, "grad_norm": 1.4457292556762695, "learning_rate": 2.07086247765623e-06, "loss": 0.2218, "step": 51780 }, { "epoch": 0.9419960348496699, "grad_norm": 1.5919594764709473, "learning_rate": 2.0704215813066153e-06, "loss": 0.2179, "step": 51790 }, { "epoch": 0.9421779224794922, "grad_norm": 1.6253888607025146, "learning_rate": 2.069982048656913e-06, "loss": 0.2141, "step": 51800 }, { "epoch": 0.9423598101093145, "grad_norm": 1.51896071434021, "learning_rate": 2.0695438798596566e-06, "loss": 0.2067, "step": 51810 }, { "epoch": 0.9425416977391368, "grad_norm": 1.5734539031982422, "learning_rate": 2.0691070750669048e-06, "loss": 0.2123, "step": 51820 }, { "epoch": 0.9427235853689591, "grad_norm": 1.569145917892456, "learning_rate": 2.0686716344302457e-06, "loss": 0.2121, "step": 51830 }, { "epoch": 0.9429054729987814, "grad_norm": 1.477557897567749, "learning_rate": 2.0682375581007914e-06, "loss": 0.2121, "step": 51840 }, { "epoch": 0.9430873606286037, "grad_norm": 1.4884390830993652, "learning_rate": 2.067804846229181e-06, "loss": 0.2079, "step": 51850 }, { "epoch": 0.943269248258426, "grad_norm": 1.6613463163375854, "learning_rate": 2.0673734989655812e-06, "loss": 0.212, "step": 51860 }, { "epoch": 0.9434511358882482, "grad_norm": 1.991605281829834, "learning_rate": 2.066943516459684e-06, "loss": 0.204, "step": 51870 }, { "epoch": 0.9436330235180705, "grad_norm": 1.8697813749313354, "learning_rate": 2.066514898860709e-06, "loss": 0.2207, "step": 51880 }, { "epoch": 0.9438149111478928, "grad_norm": 1.3647626638412476, "learning_rate": 2.0660876463174e-06, "loss": 0.2175, "step": 51890 }, { "epoch": 0.9439967987777151, "grad_norm": 1.9064949750900269, "learning_rate": 2.06566175897803e-06, "loss": 0.2157, "step": 51900 }, { "epoch": 0.9441786864075374, "grad_norm": 1.4686437845230103, "learning_rate": 2.0652372369903947e-06, "loss": 0.2185, "step": 51910 }, { "epoch": 0.9443605740373597, "grad_norm": 1.6960939168930054, "learning_rate": 2.06481408050182e-06, "loss": 0.217, "step": 51920 }, { "epoch": 0.944542461667182, "grad_norm": 1.559763789176941, "learning_rate": 2.064392289659155e-06, "loss": 0.2079, "step": 51930 }, { "epoch": 0.9447243492970043, "grad_norm": 1.5212302207946777, "learning_rate": 2.0639718646087758e-06, "loss": 0.2123, "step": 51940 }, { "epoch": 0.9449062369268266, "grad_norm": 1.5366591215133667, "learning_rate": 2.0635528054965843e-06, "loss": 0.2143, "step": 51950 }, { "epoch": 0.9450881245566489, "grad_norm": 1.6573872566223145, "learning_rate": 2.0631351124680082e-06, "loss": 0.2127, "step": 51960 }, { "epoch": 0.9452700121864712, "grad_norm": 1.4467015266418457, "learning_rate": 2.0627187856680024e-06, "loss": 0.2147, "step": 51970 }, { "epoch": 0.9454518998162935, "grad_norm": 1.289393663406372, "learning_rate": 2.0623038252410465e-06, "loss": 0.2132, "step": 51980 }, { "epoch": 0.9456337874461158, "grad_norm": 1.5093566179275513, "learning_rate": 2.0618902313311453e-06, "loss": 0.2125, "step": 51990 }, { "epoch": 0.9458156750759381, "grad_norm": 1.5324889421463013, "learning_rate": 2.0614780040818305e-06, "loss": 0.2089, "step": 52000 }, { "epoch": 0.9459975627057604, "grad_norm": 2.206815242767334, "learning_rate": 2.0610671436361603e-06, "loss": 0.2141, "step": 52010 }, { "epoch": 0.9461794503355827, "grad_norm": 1.5994696617126465, "learning_rate": 2.0606576501367165e-06, "loss": 0.2191, "step": 52020 }, { "epoch": 0.946361337965405, "grad_norm": 1.5028266906738281, "learning_rate": 2.0602495237256078e-06, "loss": 0.2127, "step": 52030 }, { "epoch": 0.9465432255952273, "grad_norm": 1.7158141136169434, "learning_rate": 2.0598427645444684e-06, "loss": 0.2103, "step": 52040 }, { "epoch": 0.9467251132250496, "grad_norm": 1.6664689779281616, "learning_rate": 2.0594373727344574e-06, "loss": 0.2197, "step": 52050 }, { "epoch": 0.9469070008548719, "grad_norm": 1.5897048711776733, "learning_rate": 2.05903334843626e-06, "loss": 0.2148, "step": 52060 }, { "epoch": 0.9470888884846942, "grad_norm": 1.5340644121170044, "learning_rate": 2.0586306917900875e-06, "loss": 0.2098, "step": 52070 }, { "epoch": 0.9472707761145165, "grad_norm": 1.4569259881973267, "learning_rate": 2.058229402935674e-06, "loss": 0.2072, "step": 52080 }, { "epoch": 0.9474526637443388, "grad_norm": 1.8624721765518188, "learning_rate": 2.0578294820122816e-06, "loss": 0.2113, "step": 52090 }, { "epoch": 0.9476345513741611, "grad_norm": 1.5389150381088257, "learning_rate": 2.0574309291586974e-06, "loss": 0.2171, "step": 52100 }, { "epoch": 0.9478164390039834, "grad_norm": 1.3157646656036377, "learning_rate": 2.057033744513231e-06, "loss": 0.2096, "step": 52110 }, { "epoch": 0.9479983266338057, "grad_norm": 1.6659090518951416, "learning_rate": 2.0566379282137215e-06, "loss": 0.2164, "step": 52120 }, { "epoch": 0.948180214263628, "grad_norm": 1.5797079801559448, "learning_rate": 2.05624348039753e-06, "loss": 0.2111, "step": 52130 }, { "epoch": 0.9483621018934503, "grad_norm": 1.6228766441345215, "learning_rate": 2.0558504012015435e-06, "loss": 0.2162, "step": 52140 }, { "epoch": 0.9485439895232726, "grad_norm": 1.6217526197433472, "learning_rate": 2.055458690762174e-06, "loss": 0.2088, "step": 52150 }, { "epoch": 0.9487258771530949, "grad_norm": 1.5737009048461914, "learning_rate": 2.0550683492153584e-06, "loss": 0.2186, "step": 52160 }, { "epoch": 0.9489077647829172, "grad_norm": 1.6241681575775146, "learning_rate": 2.0546793766965594e-06, "loss": 0.2128, "step": 52170 }, { "epoch": 0.9490896524127395, "grad_norm": 1.77498197555542, "learning_rate": 2.0542917733407636e-06, "loss": 0.2113, "step": 52180 }, { "epoch": 0.9492715400425618, "grad_norm": 1.4332072734832764, "learning_rate": 2.053905539282483e-06, "loss": 0.2136, "step": 52190 }, { "epoch": 0.949453427672384, "grad_norm": 1.4104589223861694, "learning_rate": 2.053520674655753e-06, "loss": 0.2112, "step": 52200 }, { "epoch": 0.9496353153022064, "grad_norm": 1.5245417356491089, "learning_rate": 2.053137179594137e-06, "loss": 0.2092, "step": 52210 }, { "epoch": 0.9498172029320286, "grad_norm": 1.5892328023910522, "learning_rate": 2.0527550542307198e-06, "loss": 0.2087, "step": 52220 }, { "epoch": 0.9499990905618508, "grad_norm": 1.673331618309021, "learning_rate": 2.0523742986981125e-06, "loss": 0.209, "step": 52230 }, { "epoch": 0.9501809781916731, "grad_norm": 1.5784764289855957, "learning_rate": 2.05199491312845e-06, "loss": 0.2164, "step": 52240 }, { "epoch": 0.9503628658214954, "grad_norm": 1.600792646408081, "learning_rate": 2.0516168976533925e-06, "loss": 0.2102, "step": 52250 }, { "epoch": 0.9505447534513177, "grad_norm": 1.5524204969406128, "learning_rate": 2.051240252404125e-06, "loss": 0.2178, "step": 52260 }, { "epoch": 0.95072664108114, "grad_norm": 2.0072855949401855, "learning_rate": 2.0508649775113564e-06, "loss": 0.2104, "step": 52270 }, { "epoch": 0.9509085287109623, "grad_norm": 1.9081599712371826, "learning_rate": 2.0504910731053195e-06, "loss": 0.2169, "step": 52280 }, { "epoch": 0.9510904163407846, "grad_norm": 1.6237502098083496, "learning_rate": 2.0501185393157714e-06, "loss": 0.2172, "step": 52290 }, { "epoch": 0.9512723039706069, "grad_norm": 1.6173839569091797, "learning_rate": 2.0497473762719958e-06, "loss": 0.2116, "step": 52300 }, { "epoch": 0.9514541916004292, "grad_norm": 1.4652540683746338, "learning_rate": 2.0493775841027987e-06, "loss": 0.2152, "step": 52310 }, { "epoch": 0.9516360792302515, "grad_norm": 1.596494197845459, "learning_rate": 2.0490091629365103e-06, "loss": 0.2157, "step": 52320 }, { "epoch": 0.9518179668600738, "grad_norm": 1.5986407995224, "learning_rate": 2.048642112900986e-06, "loss": 0.2159, "step": 52330 }, { "epoch": 0.9519998544898961, "grad_norm": 1.4605529308319092, "learning_rate": 2.0482764341236048e-06, "loss": 0.2123, "step": 52340 }, { "epoch": 0.9521817421197184, "grad_norm": 1.4638015031814575, "learning_rate": 2.0479121267312694e-06, "loss": 0.2116, "step": 52350 }, { "epoch": 0.9523636297495407, "grad_norm": 1.7762523889541626, "learning_rate": 2.0475491908504073e-06, "loss": 0.213, "step": 52360 }, { "epoch": 0.952545517379363, "grad_norm": 1.469164252281189, "learning_rate": 2.0471876266069705e-06, "loss": 0.2179, "step": 52370 }, { "epoch": 0.9527274050091853, "grad_norm": 1.5686920881271362, "learning_rate": 2.046827434126434e-06, "loss": 0.211, "step": 52380 }, { "epoch": 0.9529092926390076, "grad_norm": 1.6876786947250366, "learning_rate": 2.0464686135337965e-06, "loss": 0.2106, "step": 52390 }, { "epoch": 0.9530911802688299, "grad_norm": 1.6037756204605103, "learning_rate": 2.0461111649535817e-06, "loss": 0.2105, "step": 52400 }, { "epoch": 0.9532730678986522, "grad_norm": 1.4770326614379883, "learning_rate": 2.0457550885098364e-06, "loss": 0.2098, "step": 52410 }, { "epoch": 0.9534549555284745, "grad_norm": 1.5900073051452637, "learning_rate": 2.0454003843261323e-06, "loss": 0.2182, "step": 52420 }, { "epoch": 0.9536368431582968, "grad_norm": 1.7818630933761597, "learning_rate": 2.0450470525255627e-06, "loss": 0.215, "step": 52430 }, { "epoch": 0.9538187307881191, "grad_norm": 1.5357810258865356, "learning_rate": 2.044695093230747e-06, "loss": 0.2104, "step": 52440 }, { "epoch": 0.9540006184179414, "grad_norm": 1.5432544946670532, "learning_rate": 2.0443445065638266e-06, "loss": 0.2095, "step": 52450 }, { "epoch": 0.9541825060477637, "grad_norm": 1.5079010725021362, "learning_rate": 2.0439952926464684e-06, "loss": 0.2103, "step": 52460 }, { "epoch": 0.954364393677586, "grad_norm": 1.8452866077423096, "learning_rate": 2.04364745159986e-06, "loss": 0.2111, "step": 52470 }, { "epoch": 0.9545462813074083, "grad_norm": 1.5187337398529053, "learning_rate": 2.0433009835447153e-06, "loss": 0.2098, "step": 52480 }, { "epoch": 0.9547281689372306, "grad_norm": 1.7986438274383545, "learning_rate": 2.042955888601271e-06, "loss": 0.2173, "step": 52490 }, { "epoch": 0.9549100565670529, "grad_norm": 1.5371465682983398, "learning_rate": 2.042612166889286e-06, "loss": 0.2141, "step": 52500 }, { "epoch": 0.9550919441968752, "grad_norm": 1.3416426181793213, "learning_rate": 2.0422698185280445e-06, "loss": 0.2128, "step": 52510 }, { "epoch": 0.9552738318266975, "grad_norm": 1.678242802619934, "learning_rate": 2.0419288436363534e-06, "loss": 0.2139, "step": 52520 }, { "epoch": 0.9554557194565197, "grad_norm": 1.5509768724441528, "learning_rate": 2.041589242332542e-06, "loss": 0.2092, "step": 52530 }, { "epoch": 0.955637607086342, "grad_norm": 1.784528136253357, "learning_rate": 2.041251014734464e-06, "loss": 0.214, "step": 52540 }, { "epoch": 0.9558194947161643, "grad_norm": 1.4245223999023438, "learning_rate": 2.040914160959497e-06, "loss": 0.2182, "step": 52550 }, { "epoch": 0.9560013823459866, "grad_norm": 1.3570808172225952, "learning_rate": 2.0405786811245397e-06, "loss": 0.2147, "step": 52560 }, { "epoch": 0.9561832699758089, "grad_norm": 1.4353703260421753, "learning_rate": 2.0402445753460155e-06, "loss": 0.2118, "step": 52570 }, { "epoch": 0.9563651576056312, "grad_norm": 1.7563588619232178, "learning_rate": 2.039911843739871e-06, "loss": 0.2122, "step": 52580 }, { "epoch": 0.9565470452354535, "grad_norm": 1.6778290271759033, "learning_rate": 2.0395804864215756e-06, "loss": 0.2087, "step": 52590 }, { "epoch": 0.9567289328652758, "grad_norm": 1.6564286947250366, "learning_rate": 2.0392505035061215e-06, "loss": 0.2134, "step": 52600 }, { "epoch": 0.9569108204950981, "grad_norm": 1.4999020099639893, "learning_rate": 2.0389218951080246e-06, "loss": 0.2088, "step": 52610 }, { "epoch": 0.9570927081249204, "grad_norm": 1.6630356311798096, "learning_rate": 2.038594661341323e-06, "loss": 0.211, "step": 52620 }, { "epoch": 0.9572745957547427, "grad_norm": 1.420885443687439, "learning_rate": 2.038268802319578e-06, "loss": 0.2185, "step": 52630 }, { "epoch": 0.957456483384565, "grad_norm": 1.6790831089019775, "learning_rate": 2.037944318155875e-06, "loss": 0.2142, "step": 52640 }, { "epoch": 0.9576383710143873, "grad_norm": 1.5174791812896729, "learning_rate": 2.03762120896282e-06, "loss": 0.2126, "step": 52650 }, { "epoch": 0.9578202586442096, "grad_norm": 1.5314362049102783, "learning_rate": 2.037299474852544e-06, "loss": 0.2122, "step": 52660 }, { "epoch": 0.9580021462740319, "grad_norm": 1.8012624979019165, "learning_rate": 2.0369791159366993e-06, "loss": 0.2066, "step": 52670 }, { "epoch": 0.9581840339038542, "grad_norm": 1.6336514949798584, "learning_rate": 2.036660132326462e-06, "loss": 0.2159, "step": 52680 }, { "epoch": 0.9583659215336765, "grad_norm": 1.7178221940994263, "learning_rate": 2.0363425241325306e-06, "loss": 0.213, "step": 52690 }, { "epoch": 0.9585478091634988, "grad_norm": 1.6447243690490723, "learning_rate": 2.036026291465125e-06, "loss": 0.2134, "step": 52700 }, { "epoch": 0.9587296967933211, "grad_norm": 1.5625380277633667, "learning_rate": 2.03571143443399e-06, "loss": 0.211, "step": 52710 }, { "epoch": 0.9589115844231434, "grad_norm": 1.4429250955581665, "learning_rate": 2.035397953148392e-06, "loss": 0.2148, "step": 52720 }, { "epoch": 0.9590934720529657, "grad_norm": 1.6058486700057983, "learning_rate": 2.035085847717119e-06, "loss": 0.2117, "step": 52730 }, { "epoch": 0.959275359682788, "grad_norm": 1.7718349695205688, "learning_rate": 2.034775118248483e-06, "loss": 0.2164, "step": 52740 }, { "epoch": 0.9594572473126103, "grad_norm": 1.6358833312988281, "learning_rate": 2.034465764850318e-06, "loss": 0.2128, "step": 52750 }, { "epoch": 0.9596391349424326, "grad_norm": 1.5799076557159424, "learning_rate": 2.0341577876299797e-06, "loss": 0.2103, "step": 52760 }, { "epoch": 0.9598210225722549, "grad_norm": 1.671562671661377, "learning_rate": 2.033851186694348e-06, "loss": 0.2083, "step": 52770 }, { "epoch": 0.9600029102020772, "grad_norm": 1.5375436544418335, "learning_rate": 2.0335459621498225e-06, "loss": 0.2147, "step": 52780 }, { "epoch": 0.9601847978318995, "grad_norm": 1.6212810277938843, "learning_rate": 2.0332421141023283e-06, "loss": 0.21, "step": 52790 }, { "epoch": 0.9603666854617218, "grad_norm": 1.5517109632492065, "learning_rate": 2.03293964265731e-06, "loss": 0.2091, "step": 52800 }, { "epoch": 0.9605485730915441, "grad_norm": 1.3838728666305542, "learning_rate": 2.032638547919736e-06, "loss": 0.2076, "step": 52810 }, { "epoch": 0.9607304607213664, "grad_norm": 1.4372172355651855, "learning_rate": 2.032338829994097e-06, "loss": 0.2118, "step": 52820 }, { "epoch": 0.9609123483511887, "grad_norm": 1.8210010528564453, "learning_rate": 2.0320404889844053e-06, "loss": 0.2175, "step": 52830 }, { "epoch": 0.961094235981011, "grad_norm": 1.5177077054977417, "learning_rate": 2.031743524994195e-06, "loss": 0.21, "step": 52840 }, { "epoch": 0.9612761236108333, "grad_norm": 1.6663919687271118, "learning_rate": 2.031447938126524e-06, "loss": 0.2175, "step": 52850 }, { "epoch": 0.9614580112406556, "grad_norm": 1.6390336751937866, "learning_rate": 2.031153728483971e-06, "loss": 0.2133, "step": 52860 }, { "epoch": 0.9616398988704778, "grad_norm": 1.5988554954528809, "learning_rate": 2.0308608961686353e-06, "loss": 0.2146, "step": 52870 }, { "epoch": 0.9618217865003001, "grad_norm": 1.9096065759658813, "learning_rate": 2.030569441282142e-06, "loss": 0.2172, "step": 52880 }, { "epoch": 0.9620036741301224, "grad_norm": 1.8855921030044556, "learning_rate": 2.0302793639256347e-06, "loss": 0.2054, "step": 52890 }, { "epoch": 0.9621855617599447, "grad_norm": 1.4891284704208374, "learning_rate": 2.029990664199781e-06, "loss": 0.2126, "step": 52900 }, { "epoch": 0.962367449389767, "grad_norm": 1.591850757598877, "learning_rate": 2.0297033422047692e-06, "loss": 0.2131, "step": 52910 }, { "epoch": 0.9625493370195893, "grad_norm": 1.44329035282135, "learning_rate": 2.0294173980403105e-06, "loss": 0.2072, "step": 52920 }, { "epoch": 0.9627312246494116, "grad_norm": 1.8124598264694214, "learning_rate": 2.0291328318056376e-06, "loss": 0.2069, "step": 52930 }, { "epoch": 0.9629131122792339, "grad_norm": 1.4990973472595215, "learning_rate": 2.028849643599504e-06, "loss": 0.2132, "step": 52940 }, { "epoch": 0.9630949999090562, "grad_norm": 1.3908159732818604, "learning_rate": 2.0285678335201863e-06, "loss": 0.218, "step": 52950 }, { "epoch": 0.9632768875388785, "grad_norm": 1.6933542490005493, "learning_rate": 2.028287401665482e-06, "loss": 0.2135, "step": 52960 }, { "epoch": 0.9634587751687008, "grad_norm": 1.640644907951355, "learning_rate": 2.0280083481327115e-06, "loss": 0.2142, "step": 52970 }, { "epoch": 0.9636406627985231, "grad_norm": 1.376094937324524, "learning_rate": 2.027730673018715e-06, "loss": 0.2149, "step": 52980 }, { "epoch": 0.9638225504283454, "grad_norm": 1.5903375148773193, "learning_rate": 2.027454376419856e-06, "loss": 0.2119, "step": 52990 }, { "epoch": 0.9640044380581677, "grad_norm": 2.2288546562194824, "learning_rate": 2.027179458432019e-06, "loss": 0.2105, "step": 53000 }, { "epoch": 0.96418632568799, "grad_norm": 1.6484602689743042, "learning_rate": 2.0269059191506105e-06, "loss": 0.2141, "step": 53010 }, { "epoch": 0.9643682133178123, "grad_norm": 1.8340058326721191, "learning_rate": 2.026633758670557e-06, "loss": 0.2146, "step": 53020 }, { "epoch": 0.9645501009476346, "grad_norm": 1.5751073360443115, "learning_rate": 2.026362977086309e-06, "loss": 0.213, "step": 53030 }, { "epoch": 0.9647319885774569, "grad_norm": 1.4521722793579102, "learning_rate": 2.0260935744918358e-06, "loss": 0.2162, "step": 53040 }, { "epoch": 0.9649138762072792, "grad_norm": 1.401808738708496, "learning_rate": 2.02582555098063e-06, "loss": 0.2172, "step": 53050 }, { "epoch": 0.9650957638371014, "grad_norm": 1.4179785251617432, "learning_rate": 2.025558906645705e-06, "loss": 0.2174, "step": 53060 }, { "epoch": 0.9652776514669237, "grad_norm": 1.523332118988037, "learning_rate": 2.025293641579596e-06, "loss": 0.2134, "step": 53070 }, { "epoch": 0.965459539096746, "grad_norm": 1.5578234195709229, "learning_rate": 2.025029755874358e-06, "loss": 0.2163, "step": 53080 }, { "epoch": 0.9656414267265683, "grad_norm": 1.6146714687347412, "learning_rate": 2.02476724962157e-06, "loss": 0.2147, "step": 53090 }, { "epoch": 0.9658233143563906, "grad_norm": 1.617142677307129, "learning_rate": 2.0245061229123297e-06, "loss": 0.2079, "step": 53100 }, { "epoch": 0.9660052019862129, "grad_norm": 1.5444735288619995, "learning_rate": 2.0242463758372576e-06, "loss": 0.2233, "step": 53110 }, { "epoch": 0.9661870896160352, "grad_norm": 1.6221851110458374, "learning_rate": 2.0239880084864948e-06, "loss": 0.215, "step": 53120 }, { "epoch": 0.9663689772458575, "grad_norm": 1.7337911128997803, "learning_rate": 2.023731020949704e-06, "loss": 0.2181, "step": 53130 }, { "epoch": 0.9665508648756798, "grad_norm": 1.9777781963348389, "learning_rate": 2.023475413316068e-06, "loss": 0.2143, "step": 53140 }, { "epoch": 0.9667327525055021, "grad_norm": 1.6941969394683838, "learning_rate": 2.023221185674293e-06, "loss": 0.2134, "step": 53150 }, { "epoch": 0.9669146401353244, "grad_norm": 1.5509041547775269, "learning_rate": 2.022968338112603e-06, "loss": 0.2119, "step": 53160 }, { "epoch": 0.9670965277651467, "grad_norm": 1.7573155164718628, "learning_rate": 2.0227168707187457e-06, "loss": 0.2151, "step": 53170 }, { "epoch": 0.967278415394969, "grad_norm": 1.5867702960968018, "learning_rate": 2.0224667835799893e-06, "loss": 0.2118, "step": 53180 }, { "epoch": 0.9674603030247912, "grad_norm": 1.578321099281311, "learning_rate": 2.022218076783122e-06, "loss": 0.2153, "step": 53190 }, { "epoch": 0.9676421906546135, "grad_norm": 1.5501642227172852, "learning_rate": 2.0219707504144543e-06, "loss": 0.2158, "step": 53200 }, { "epoch": 0.9678240782844358, "grad_norm": 1.465165376663208, "learning_rate": 2.0217248045598175e-06, "loss": 0.2084, "step": 53210 }, { "epoch": 0.9680059659142581, "grad_norm": 1.7393033504486084, "learning_rate": 2.0214802393045617e-06, "loss": 0.2154, "step": 53220 }, { "epoch": 0.9681878535440804, "grad_norm": 1.6782667636871338, "learning_rate": 2.021237054733561e-06, "loss": 0.2175, "step": 53230 }, { "epoch": 0.9683697411739027, "grad_norm": 1.6932514905929565, "learning_rate": 2.0209952509312083e-06, "loss": 0.2074, "step": 53240 }, { "epoch": 0.968551628803725, "grad_norm": 1.6807295083999634, "learning_rate": 2.020754827981418e-06, "loss": 0.2107, "step": 53250 }, { "epoch": 0.9687335164335473, "grad_norm": 1.5327091217041016, "learning_rate": 2.0205157859676246e-06, "loss": 0.2101, "step": 53260 }, { "epoch": 0.9689154040633696, "grad_norm": 1.632326602935791, "learning_rate": 2.020278124972785e-06, "loss": 0.2096, "step": 53270 }, { "epoch": 0.9690972916931919, "grad_norm": 1.9346274137496948, "learning_rate": 2.0200418450793757e-06, "loss": 0.2169, "step": 53280 }, { "epoch": 0.9692791793230142, "grad_norm": 1.3549751043319702, "learning_rate": 2.0198069463693933e-06, "loss": 0.2121, "step": 53290 }, { "epoch": 0.9694610669528365, "grad_norm": 1.5334665775299072, "learning_rate": 2.019573428924356e-06, "loss": 0.2101, "step": 53300 }, { "epoch": 0.9696429545826588, "grad_norm": 1.3669434785842896, "learning_rate": 2.0193412928253025e-06, "loss": 0.2114, "step": 53310 }, { "epoch": 0.9698248422124811, "grad_norm": 1.5847642421722412, "learning_rate": 2.019110538152792e-06, "loss": 0.2151, "step": 53320 }, { "epoch": 0.9700067298423034, "grad_norm": 2.1516811847686768, "learning_rate": 2.018881164986905e-06, "loss": 0.2106, "step": 53330 }, { "epoch": 0.9701886174721257, "grad_norm": 1.4223577976226807, "learning_rate": 2.018653173407241e-06, "loss": 0.2058, "step": 53340 }, { "epoch": 0.970370505101948, "grad_norm": 1.6892526149749756, "learning_rate": 2.018426563492922e-06, "loss": 0.2146, "step": 53350 }, { "epoch": 0.9705523927317703, "grad_norm": 1.6498457193374634, "learning_rate": 2.018201335322588e-06, "loss": 0.2196, "step": 53360 }, { "epoch": 0.9707342803615926, "grad_norm": 1.5698055028915405, "learning_rate": 2.0179774889744026e-06, "loss": 0.2096, "step": 53370 }, { "epoch": 0.9709161679914149, "grad_norm": 1.8670353889465332, "learning_rate": 2.0177550245260476e-06, "loss": 0.2112, "step": 53380 }, { "epoch": 0.9710980556212372, "grad_norm": 1.6577906608581543, "learning_rate": 2.0175339420547254e-06, "loss": 0.2146, "step": 53390 }, { "epoch": 0.9712799432510595, "grad_norm": 1.484544038772583, "learning_rate": 2.0173142416371602e-06, "loss": 0.2108, "step": 53400 }, { "epoch": 0.9714618308808818, "grad_norm": 1.6384074687957764, "learning_rate": 2.017095923349595e-06, "loss": 0.2175, "step": 53410 }, { "epoch": 0.9716437185107041, "grad_norm": 1.5737968683242798, "learning_rate": 2.016878987267794e-06, "loss": 0.2103, "step": 53420 }, { "epoch": 0.9718256061405264, "grad_norm": 2.3047432899475098, "learning_rate": 2.0166634334670408e-06, "loss": 0.2159, "step": 53430 }, { "epoch": 0.9720074937703487, "grad_norm": 1.6005725860595703, "learning_rate": 2.0164492620221415e-06, "loss": 0.2153, "step": 53440 }, { "epoch": 0.972189381400171, "grad_norm": 1.45133376121521, "learning_rate": 2.01623647300742e-06, "loss": 0.2103, "step": 53450 }, { "epoch": 0.9723712690299933, "grad_norm": 1.7116947174072266, "learning_rate": 2.0160250664967216e-06, "loss": 0.2116, "step": 53460 }, { "epoch": 0.9725531566598156, "grad_norm": 1.6649397611618042, "learning_rate": 2.0158150425634124e-06, "loss": 0.2127, "step": 53470 }, { "epoch": 0.9727350442896379, "grad_norm": 1.610732078552246, "learning_rate": 2.015606401280377e-06, "loss": 0.2123, "step": 53480 }, { "epoch": 0.9729169319194602, "grad_norm": 1.6081775426864624, "learning_rate": 2.0153991427200215e-06, "loss": 0.2092, "step": 53490 }, { "epoch": 0.9730988195492825, "grad_norm": 1.805647611618042, "learning_rate": 2.015193266954272e-06, "loss": 0.2061, "step": 53500 }, { "epoch": 0.9732807071791048, "grad_norm": 1.3545130491256714, "learning_rate": 2.0149887740545747e-06, "loss": 0.2148, "step": 53510 }, { "epoch": 0.973462594808927, "grad_norm": 1.4244328737258911, "learning_rate": 2.014785664091895e-06, "loss": 0.2168, "step": 53520 }, { "epoch": 0.9736444824387493, "grad_norm": 1.7168967723846436, "learning_rate": 2.0145839371367194e-06, "loss": 0.2135, "step": 53530 }, { "epoch": 0.9738263700685716, "grad_norm": 1.6814241409301758, "learning_rate": 2.0143835932590543e-06, "loss": 0.2064, "step": 53540 }, { "epoch": 0.9740082576983939, "grad_norm": 1.4760215282440186, "learning_rate": 2.014184632528426e-06, "loss": 0.2131, "step": 53550 }, { "epoch": 0.9741901453282162, "grad_norm": 1.6346092224121094, "learning_rate": 2.0139870550138814e-06, "loss": 0.215, "step": 53560 }, { "epoch": 0.9743720329580385, "grad_norm": 1.530956506729126, "learning_rate": 2.0137908607839853e-06, "loss": 0.2154, "step": 53570 }, { "epoch": 0.9745539205878608, "grad_norm": 1.4516386985778809, "learning_rate": 2.0135960499068246e-06, "loss": 0.2082, "step": 53580 }, { "epoch": 0.9747358082176831, "grad_norm": 1.5337274074554443, "learning_rate": 2.0134026224500054e-06, "loss": 0.2062, "step": 53590 }, { "epoch": 0.9749176958475054, "grad_norm": 1.880128026008606, "learning_rate": 2.013210578480654e-06, "loss": 0.2112, "step": 53600 }, { "epoch": 0.9750995834773277, "grad_norm": 1.910699486732483, "learning_rate": 2.0130199180654164e-06, "loss": 0.2114, "step": 53610 }, { "epoch": 0.97528147110715, "grad_norm": 1.4044877290725708, "learning_rate": 2.0128306412704577e-06, "loss": 0.2077, "step": 53620 }, { "epoch": 0.9754633587369723, "grad_norm": 1.4494531154632568, "learning_rate": 2.012642748161464e-06, "loss": 0.2127, "step": 53630 }, { "epoch": 0.9756452463667946, "grad_norm": 1.6795005798339844, "learning_rate": 2.012456238803641e-06, "loss": 0.2157, "step": 53640 }, { "epoch": 0.9758271339966169, "grad_norm": 1.7789146900177002, "learning_rate": 2.012271113261713e-06, "loss": 0.2074, "step": 53650 }, { "epoch": 0.9760090216264392, "grad_norm": 1.5674182176589966, "learning_rate": 2.0120873715999257e-06, "loss": 0.2115, "step": 53660 }, { "epoch": 0.9761909092562615, "grad_norm": 1.5911133289337158, "learning_rate": 2.011905013882044e-06, "loss": 0.2131, "step": 53670 }, { "epoch": 0.9763727968860838, "grad_norm": 1.7950140237808228, "learning_rate": 2.011724040171351e-06, "loss": 0.2125, "step": 53680 }, { "epoch": 0.9765546845159061, "grad_norm": 1.3472710847854614, "learning_rate": 2.0115444505306524e-06, "loss": 0.2117, "step": 53690 }, { "epoch": 0.9767365721457284, "grad_norm": 1.4371317625045776, "learning_rate": 2.0113662450222716e-06, "loss": 0.2097, "step": 53700 }, { "epoch": 0.9769184597755507, "grad_norm": 1.6224696636199951, "learning_rate": 2.011189423708052e-06, "loss": 0.2112, "step": 53710 }, { "epoch": 0.977100347405373, "grad_norm": 1.7201018333435059, "learning_rate": 2.0110139866493562e-06, "loss": 0.2095, "step": 53720 }, { "epoch": 0.9772822350351953, "grad_norm": 1.7760016918182373, "learning_rate": 2.0108399339070675e-06, "loss": 0.2138, "step": 53730 }, { "epoch": 0.9774641226650176, "grad_norm": 1.6537659168243408, "learning_rate": 2.0106672655415884e-06, "loss": 0.2126, "step": 53740 }, { "epoch": 0.9776460102948399, "grad_norm": 1.5546841621398926, "learning_rate": 2.0104959816128404e-06, "loss": 0.2142, "step": 53750 }, { "epoch": 0.9778278979246622, "grad_norm": 1.5479320287704468, "learning_rate": 2.010326082180265e-06, "loss": 0.2118, "step": 53760 }, { "epoch": 0.9780097855544845, "grad_norm": 1.6534501314163208, "learning_rate": 2.0101575673028238e-06, "loss": 0.215, "step": 53770 }, { "epoch": 0.9781916731843068, "grad_norm": 1.438294529914856, "learning_rate": 2.0099904370389963e-06, "loss": 0.2079, "step": 53780 }, { "epoch": 0.9783735608141291, "grad_norm": 1.75930917263031, "learning_rate": 2.0098246914467833e-06, "loss": 0.2089, "step": 53790 }, { "epoch": 0.9785554484439514, "grad_norm": 1.548811912536621, "learning_rate": 2.0096603305837043e-06, "loss": 0.2171, "step": 53800 }, { "epoch": 0.9787373360737737, "grad_norm": 1.5662919282913208, "learning_rate": 2.009497354506797e-06, "loss": 0.2075, "step": 53810 }, { "epoch": 0.978919223703596, "grad_norm": 1.4805386066436768, "learning_rate": 2.009335763272621e-06, "loss": 0.213, "step": 53820 }, { "epoch": 0.9791011113334183, "grad_norm": 1.6475855112075806, "learning_rate": 2.009175556937254e-06, "loss": 0.2115, "step": 53830 }, { "epoch": 0.9792829989632406, "grad_norm": 1.4159873723983765, "learning_rate": 2.0090167355562927e-06, "loss": 0.2137, "step": 53840 }, { "epoch": 0.9794648865930629, "grad_norm": 1.596427321434021, "learning_rate": 2.008859299184854e-06, "loss": 0.21, "step": 53850 }, { "epoch": 0.9796467742228852, "grad_norm": 1.3872483968734741, "learning_rate": 2.0087032478775735e-06, "loss": 0.2131, "step": 53860 }, { "epoch": 0.9798286618527074, "grad_norm": 1.6004736423492432, "learning_rate": 2.008548581688606e-06, "loss": 0.2095, "step": 53870 }, { "epoch": 0.9800105494825297, "grad_norm": 1.6727675199508667, "learning_rate": 2.0083953006716264e-06, "loss": 0.2142, "step": 53880 }, { "epoch": 0.980192437112352, "grad_norm": 1.7615951299667358, "learning_rate": 2.0082434048798292e-06, "loss": 0.2172, "step": 53890 }, { "epoch": 0.9803743247421742, "grad_norm": 1.564741849899292, "learning_rate": 2.008092894365927e-06, "loss": 0.2246, "step": 53900 }, { "epoch": 0.9805562123719965, "grad_norm": 1.713159441947937, "learning_rate": 2.007943769182151e-06, "loss": 0.2129, "step": 53910 }, { "epoch": 0.9807381000018188, "grad_norm": 1.834788203239441, "learning_rate": 2.007796029380255e-06, "loss": 0.2149, "step": 53920 }, { "epoch": 0.9809199876316411, "grad_norm": 1.4996851682662964, "learning_rate": 2.007649675011508e-06, "loss": 0.2146, "step": 53930 }, { "epoch": 0.9811018752614634, "grad_norm": 1.7308238744735718, "learning_rate": 2.0075047061267015e-06, "loss": 0.2152, "step": 53940 }, { "epoch": 0.9812837628912857, "grad_norm": 1.6153265237808228, "learning_rate": 2.0073611227761433e-06, "loss": 0.2112, "step": 53950 }, { "epoch": 0.981465650521108, "grad_norm": 1.676879644393921, "learning_rate": 2.007218925009663e-06, "loss": 0.2108, "step": 53960 }, { "epoch": 0.9816475381509303, "grad_norm": 1.6669988632202148, "learning_rate": 2.007078112876607e-06, "loss": 0.2185, "step": 53970 }, { "epoch": 0.9818294257807526, "grad_norm": 1.958390712738037, "learning_rate": 2.006938686425843e-06, "loss": 0.2121, "step": 53980 }, { "epoch": 0.9820113134105749, "grad_norm": 1.604872703552246, "learning_rate": 2.006800645705757e-06, "loss": 0.2137, "step": 53990 }, { "epoch": 0.9821932010403972, "grad_norm": 1.59299635887146, "learning_rate": 2.0066639907642522e-06, "loss": 0.2098, "step": 54000 }, { "epoch": 0.9823750886702195, "grad_norm": 2.0484447479248047, "learning_rate": 2.006528721648755e-06, "loss": 0.2145, "step": 54010 }, { "epoch": 0.9825569763000418, "grad_norm": 1.512977957725525, "learning_rate": 2.0063948384062066e-06, "loss": 0.2126, "step": 54020 }, { "epoch": 0.9827388639298641, "grad_norm": 1.7267467975616455, "learning_rate": 2.006262341083069e-06, "loss": 0.211, "step": 54030 }, { "epoch": 0.9829207515596864, "grad_norm": 1.5911816358566284, "learning_rate": 2.006131229725325e-06, "loss": 0.2106, "step": 54040 }, { "epoch": 0.9831026391895087, "grad_norm": 1.791204810142517, "learning_rate": 2.0060015043784745e-06, "loss": 0.2112, "step": 54050 }, { "epoch": 0.983284526819331, "grad_norm": 1.4755839109420776, "learning_rate": 2.0058731650875354e-06, "loss": 0.2138, "step": 54060 }, { "epoch": 0.9834664144491533, "grad_norm": 1.656952142715454, "learning_rate": 2.0057462118970466e-06, "loss": 0.2114, "step": 54070 }, { "epoch": 0.9836483020789756, "grad_norm": 1.692993402481079, "learning_rate": 2.0056206448510647e-06, "loss": 0.2104, "step": 54080 }, { "epoch": 0.9838301897087979, "grad_norm": 1.4097809791564941, "learning_rate": 2.0054964639931674e-06, "loss": 0.2108, "step": 54090 }, { "epoch": 0.9840120773386202, "grad_norm": 1.4218165874481201, "learning_rate": 2.005373669366448e-06, "loss": 0.2071, "step": 54100 }, { "epoch": 0.9841939649684425, "grad_norm": 1.4621732234954834, "learning_rate": 2.0052522610135212e-06, "loss": 0.2111, "step": 54110 }, { "epoch": 0.9843758525982648, "grad_norm": 1.546377182006836, "learning_rate": 2.00513223897652e-06, "loss": 0.2063, "step": 54120 }, { "epoch": 0.9845577402280871, "grad_norm": 1.4669640064239502, "learning_rate": 2.0050136032970966e-06, "loss": 0.212, "step": 54130 }, { "epoch": 0.9847396278579094, "grad_norm": 1.5128979682922363, "learning_rate": 2.0048963540164208e-06, "loss": 0.2127, "step": 54140 }, { "epoch": 0.9849215154877317, "grad_norm": 1.5332446098327637, "learning_rate": 2.0047804911751824e-06, "loss": 0.2079, "step": 54150 }, { "epoch": 0.985103403117554, "grad_norm": 1.5511850118637085, "learning_rate": 2.00466601481359e-06, "loss": 0.2094, "step": 54160 }, { "epoch": 0.9852852907473763, "grad_norm": 1.5990674495697021, "learning_rate": 2.004552924971371e-06, "loss": 0.21, "step": 54170 }, { "epoch": 0.9854671783771985, "grad_norm": 1.6118334531784058, "learning_rate": 2.004441221687772e-06, "loss": 0.2081, "step": 54180 }, { "epoch": 0.9856490660070208, "grad_norm": 1.6771197319030762, "learning_rate": 2.0043309050015566e-06, "loss": 0.2124, "step": 54190 }, { "epoch": 0.9858309536368431, "grad_norm": 1.5905027389526367, "learning_rate": 2.0042219749510093e-06, "loss": 0.2092, "step": 54200 }, { "epoch": 0.9860128412666654, "grad_norm": 1.6574382781982422, "learning_rate": 2.0041144315739325e-06, "loss": 0.2075, "step": 54210 }, { "epoch": 0.9861947288964877, "grad_norm": 1.8627842664718628, "learning_rate": 2.0040082749076474e-06, "loss": 0.2094, "step": 54220 }, { "epoch": 0.98637661652631, "grad_norm": 1.3990075588226318, "learning_rate": 2.003903504988994e-06, "loss": 0.2044, "step": 54230 }, { "epoch": 0.9865585041561323, "grad_norm": 1.6408474445343018, "learning_rate": 2.0038001218543315e-06, "loss": 0.2116, "step": 54240 }, { "epoch": 0.9867403917859546, "grad_norm": 1.5397635698318481, "learning_rate": 2.003698125539537e-06, "loss": 0.2124, "step": 54250 }, { "epoch": 0.9869222794157769, "grad_norm": 1.6464295387268066, "learning_rate": 2.0035975160800066e-06, "loss": 0.2151, "step": 54260 }, { "epoch": 0.9871041670455992, "grad_norm": 1.553312063217163, "learning_rate": 2.0034982935106563e-06, "loss": 0.2144, "step": 54270 }, { "epoch": 0.9872860546754215, "grad_norm": 1.5275514125823975, "learning_rate": 2.0034004578659184e-06, "loss": 0.2064, "step": 54280 }, { "epoch": 0.9874679423052438, "grad_norm": 1.5056277513504028, "learning_rate": 2.0033040091797463e-06, "loss": 0.2052, "step": 54290 }, { "epoch": 0.9876498299350661, "grad_norm": 1.7777657508850098, "learning_rate": 2.0032089474856104e-06, "loss": 0.2134, "step": 54300 }, { "epoch": 0.9878317175648884, "grad_norm": 1.4062120914459229, "learning_rate": 2.0031152728165002e-06, "loss": 0.2066, "step": 54310 }, { "epoch": 0.9880136051947107, "grad_norm": 1.4958139657974243, "learning_rate": 2.0030229852049248e-06, "loss": 0.2092, "step": 54320 }, { "epoch": 0.988195492824533, "grad_norm": 1.5600354671478271, "learning_rate": 2.0029320846829113e-06, "loss": 0.2065, "step": 54330 }, { "epoch": 0.9883773804543553, "grad_norm": 1.8635714054107666, "learning_rate": 2.0028425712820044e-06, "loss": 0.2134, "step": 54340 }, { "epoch": 0.9885592680841776, "grad_norm": 1.2785776853561401, "learning_rate": 2.0027544450332693e-06, "loss": 0.2053, "step": 54350 }, { "epoch": 0.9887411557139999, "grad_norm": 1.9325377941131592, "learning_rate": 2.002667705967288e-06, "loss": 0.2097, "step": 54360 }, { "epoch": 0.9889230433438222, "grad_norm": 1.6926970481872559, "learning_rate": 2.002582354114163e-06, "loss": 0.2113, "step": 54370 }, { "epoch": 0.9891049309736445, "grad_norm": 1.519862174987793, "learning_rate": 2.002498389503513e-06, "loss": 0.2065, "step": 54380 }, { "epoch": 0.9892868186034668, "grad_norm": 1.584833025932312, "learning_rate": 2.002415812164478e-06, "loss": 0.208, "step": 54390 }, { "epoch": 0.9894687062332891, "grad_norm": 1.5123472213745117, "learning_rate": 2.002334622125714e-06, "loss": 0.2091, "step": 54400 }, { "epoch": 0.9896505938631114, "grad_norm": 1.4733960628509521, "learning_rate": 2.002254819415398e-06, "loss": 0.2105, "step": 54410 }, { "epoch": 0.9898324814929337, "grad_norm": 1.4161728620529175, "learning_rate": 2.002176404061223e-06, "loss": 0.2078, "step": 54420 }, { "epoch": 0.990014369122756, "grad_norm": 1.5188556909561157, "learning_rate": 2.0020993760904027e-06, "loss": 0.2061, "step": 54430 }, { "epoch": 0.9901962567525783, "grad_norm": 1.4689311981201172, "learning_rate": 2.0020237355296686e-06, "loss": 0.2059, "step": 54440 }, { "epoch": 0.9903781443824006, "grad_norm": 1.474869728088379, "learning_rate": 2.0019494824052697e-06, "loss": 0.2061, "step": 54450 }, { "epoch": 0.9905600320122229, "grad_norm": 1.6150691509246826, "learning_rate": 2.0018766167429755e-06, "loss": 0.2136, "step": 54460 }, { "epoch": 0.9907419196420452, "grad_norm": 1.4748198986053467, "learning_rate": 2.001805138568072e-06, "loss": 0.214, "step": 54470 }, { "epoch": 0.9909238072718675, "grad_norm": 1.6136471033096313, "learning_rate": 2.001735047905365e-06, "loss": 0.2151, "step": 54480 }, { "epoch": 0.9911056949016898, "grad_norm": 1.5139753818511963, "learning_rate": 2.0016663447791785e-06, "loss": 0.2086, "step": 54490 }, { "epoch": 0.9912875825315121, "grad_norm": 1.4934049844741821, "learning_rate": 2.001599029213355e-06, "loss": 0.2107, "step": 54500 }, { "epoch": 0.9914694701613344, "grad_norm": 1.5918073654174805, "learning_rate": 2.0015331012312547e-06, "loss": 0.21, "step": 54510 }, { "epoch": 0.9916513577911567, "grad_norm": 1.77572500705719, "learning_rate": 2.0014685608557575e-06, "loss": 0.2105, "step": 54520 }, { "epoch": 0.991833245420979, "grad_norm": 1.5469247102737427, "learning_rate": 2.00140540810926e-06, "loss": 0.2066, "step": 54530 }, { "epoch": 0.9920151330508012, "grad_norm": 1.675577163696289, "learning_rate": 2.0013436430136802e-06, "loss": 0.2082, "step": 54540 }, { "epoch": 0.9921970206806235, "grad_norm": 1.5973166227340698, "learning_rate": 2.0012832655904522e-06, "loss": 0.2187, "step": 54550 }, { "epoch": 0.9923789083104458, "grad_norm": 1.5149266719818115, "learning_rate": 2.001224275860528e-06, "loss": 0.2087, "step": 54560 }, { "epoch": 0.9925607959402681, "grad_norm": 1.5000344514846802, "learning_rate": 2.00116667384438e-06, "loss": 0.2174, "step": 54570 }, { "epoch": 0.9927426835700904, "grad_norm": 1.3353540897369385, "learning_rate": 2.001110459561998e-06, "loss": 0.2116, "step": 54580 }, { "epoch": 0.9929245711999127, "grad_norm": 1.567085862159729, "learning_rate": 2.00105563303289e-06, "loss": 0.2101, "step": 54590 }, { "epoch": 0.993106458829735, "grad_norm": 1.3988854885101318, "learning_rate": 2.001002194276083e-06, "loss": 0.2125, "step": 54600 }, { "epoch": 0.9932883464595573, "grad_norm": 1.5297785997390747, "learning_rate": 2.0009501433101217e-06, "loss": 0.2133, "step": 54610 }, { "epoch": 0.9934702340893796, "grad_norm": 1.4864294528961182, "learning_rate": 2.00089948015307e-06, "loss": 0.2122, "step": 54620 }, { "epoch": 0.9936521217192019, "grad_norm": 1.8264398574829102, "learning_rate": 2.0008502048225096e-06, "loss": 0.2139, "step": 54630 }, { "epoch": 0.9938340093490242, "grad_norm": 1.4639887809753418, "learning_rate": 2.0008023173355405e-06, "loss": 0.2073, "step": 54640 }, { "epoch": 0.9940158969788465, "grad_norm": 1.7296959161758423, "learning_rate": 2.0007558177087825e-06, "loss": 0.2145, "step": 54650 }, { "epoch": 0.9941977846086688, "grad_norm": 1.6712781190872192, "learning_rate": 2.000710705958371e-06, "loss": 0.2079, "step": 54660 }, { "epoch": 0.9943796722384911, "grad_norm": 1.486959457397461, "learning_rate": 2.0006669820999622e-06, "loss": 0.2104, "step": 54670 }, { "epoch": 0.9945615598683134, "grad_norm": 1.6992915868759155, "learning_rate": 2.00062464614873e-06, "loss": 0.2073, "step": 54680 }, { "epoch": 0.9947434474981357, "grad_norm": 1.558663249015808, "learning_rate": 2.000583698119365e-06, "loss": 0.2093, "step": 54690 }, { "epoch": 0.994925335127958, "grad_norm": 1.5898857116699219, "learning_rate": 2.0005441380260797e-06, "loss": 0.2055, "step": 54700 }, { "epoch": 0.9951072227577803, "grad_norm": 1.3374749422073364, "learning_rate": 2.0005059658826016e-06, "loss": 0.2115, "step": 54710 }, { "epoch": 0.9952891103876026, "grad_norm": 1.684146523475647, "learning_rate": 2.0004691817021777e-06, "loss": 0.2067, "step": 54720 }, { "epoch": 0.9954709980174248, "grad_norm": 1.4357985258102417, "learning_rate": 2.0004337854975734e-06, "loss": 0.2083, "step": 54730 }, { "epoch": 0.9956528856472471, "grad_norm": 1.57524836063385, "learning_rate": 2.0003997772810734e-06, "loss": 0.2148, "step": 54740 }, { "epoch": 0.9958347732770694, "grad_norm": 1.8838434219360352, "learning_rate": 2.000367157064478e-06, "loss": 0.2139, "step": 54750 }, { "epoch": 0.9960166609068917, "grad_norm": 1.4746474027633667, "learning_rate": 2.000335924859109e-06, "loss": 0.2098, "step": 54760 }, { "epoch": 0.996198548536714, "grad_norm": 1.4374456405639648, "learning_rate": 2.000306080675804e-06, "loss": 0.2092, "step": 54770 }, { "epoch": 0.9963804361665363, "grad_norm": 1.7636240720748901, "learning_rate": 2.000277624524922e-06, "loss": 0.2099, "step": 54780 }, { "epoch": 0.9965623237963586, "grad_norm": 1.5415031909942627, "learning_rate": 2.0002505564163357e-06, "loss": 0.208, "step": 54790 }, { "epoch": 0.9967442114261809, "grad_norm": 1.7566804885864258, "learning_rate": 2.0002248763594397e-06, "loss": 0.2082, "step": 54800 }, { "epoch": 0.9969260990560032, "grad_norm": 1.6372301578521729, "learning_rate": 2.000200584363146e-06, "loss": 0.2065, "step": 54810 }, { "epoch": 0.9971079866858255, "grad_norm": 1.6461862325668335, "learning_rate": 2.000177680435885e-06, "loss": 0.2123, "step": 54820 }, { "epoch": 0.9972898743156478, "grad_norm": 1.851456880569458, "learning_rate": 2.0001561645856044e-06, "loss": 0.2082, "step": 54830 }, { "epoch": 0.99747176194547, "grad_norm": 1.5484070777893066, "learning_rate": 2.000136036819772e-06, "loss": 0.2131, "step": 54840 }, { "epoch": 0.9976536495752923, "grad_norm": 1.575046181678772, "learning_rate": 2.0001172971453717e-06, "loss": 0.2112, "step": 54850 }, { "epoch": 0.9978355372051146, "grad_norm": 1.6345906257629395, "learning_rate": 2.000099945568908e-06, "loss": 0.2173, "step": 54860 }, { "epoch": 0.9980174248349369, "grad_norm": 1.6149917840957642, "learning_rate": 2.000083982096401e-06, "loss": 0.2127, "step": 54870 }, { "epoch": 0.9981993124647592, "grad_norm": 1.6498903036117554, "learning_rate": 2.0000694067333924e-06, "loss": 0.208, "step": 54880 }, { "epoch": 0.9983812000945815, "grad_norm": 1.5605581998825073, "learning_rate": 2.0000562194849385e-06, "loss": 0.2103, "step": 54890 }, { "epoch": 0.9985630877244038, "grad_norm": 1.7082101106643677, "learning_rate": 2.0000444203556176e-06, "loss": 0.2103, "step": 54900 }, { "epoch": 0.9987449753542261, "grad_norm": 1.4701950550079346, "learning_rate": 2.0000340093495226e-06, "loss": 0.2094, "step": 54910 }, { "epoch": 0.9989268629840484, "grad_norm": 1.9187380075454712, "learning_rate": 2.0000249864702677e-06, "loss": 0.2127, "step": 54920 }, { "epoch": 0.9991087506138707, "grad_norm": 1.586471676826477, "learning_rate": 2.0000173517209833e-06, "loss": 0.2074, "step": 54930 }, { "epoch": 0.999290638243693, "grad_norm": 1.5746173858642578, "learning_rate": 2.0000111051043196e-06, "loss": 0.2085, "step": 54940 }, { "epoch": 0.9994725258735153, "grad_norm": 1.856128454208374, "learning_rate": 2.000006246622444e-06, "loss": 0.2123, "step": 54950 }, { "epoch": 0.9996544135033376, "grad_norm": 1.5755693912506104, "learning_rate": 2.000002776277043e-06, "loss": 0.2121, "step": 54960 }, { "epoch": 0.9998363011331599, "grad_norm": 1.5777133703231812, "learning_rate": 2.0000006940693213e-06, "loss": 0.207, "step": 54970 }, { "epoch": 1.0, "step": 54979, "total_flos": 2980695134699520.0, "train_loss": 0.244969728950129, "train_runtime": 340379.8748, "train_samples_per_second": 20.675, "train_steps_per_second": 0.162 } ], "logging_steps": 10, "max_steps": 54979, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 50000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2980695134699520.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }