{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9551913142204285, "learning_rate": 9e-07, "loss": 1.3014, "step": 10 }, { "grad_norm": 0.6925719976425171, "learning_rate": 1.9e-06, "loss": 1.3222, "step": 20 }, { "grad_norm": 1.1202595233917236, "learning_rate": 2.9e-06, "loss": 1.2555, "step": 30 }, { "grad_norm": 0.9213563203811646, "learning_rate": 3.9e-06, "loss": 1.1668, "step": 40 }, { "grad_norm": 0.882110059261322, "learning_rate": 4.9000000000000005e-06, "loss": 1.1587, "step": 50 }, { "grad_norm": 0.7110579013824463, "learning_rate": 5.9e-06, "loss": 1.1063, "step": 60 }, { "grad_norm": 0.6221452951431274, "learning_rate": 6.900000000000001e-06, "loss": 1.0871, "step": 70 }, { "grad_norm": 0.6641867160797119, "learning_rate": 7.9e-06, "loss": 1.0572, "step": 80 }, { "grad_norm": 0.8032795786857605, "learning_rate": 8.9e-06, "loss": 1.0526, "step": 90 }, { "grad_norm": 0.7443200349807739, "learning_rate": 9.900000000000002e-06, "loss": 1.0451, "step": 100 }, { "grad_norm": 0.6724629402160645, "learning_rate": 1.09e-05, "loss": 1.0446, "step": 110 }, { "grad_norm": 0.7099994421005249, "learning_rate": 1.19e-05, "loss": 1.0357, "step": 120 }, { "grad_norm": 0.6829448342323303, "learning_rate": 1.29e-05, "loss": 1.025, "step": 130 }, { "grad_norm": 0.8098556995391846, "learning_rate": 1.3900000000000002e-05, "loss": 1.0025, "step": 140 }, { "grad_norm": 1.674937129020691, "learning_rate": 1.49e-05, "loss": 0.9478, "step": 150 }, { "grad_norm": 1.795446515083313, "learning_rate": 1.59e-05, "loss": 0.8653, "step": 160 }, { "grad_norm": 1.4311344623565674, "learning_rate": 1.69e-05, "loss": 0.7826, "step": 170 }, { "grad_norm": 2.002934694290161, "learning_rate": 1.79e-05, "loss": 0.7145, "step": 180 }, { "grad_norm": 1.7943265438079834, "learning_rate": 1.8900000000000002e-05, "loss": 0.6184, "step": 190 }, { "grad_norm": 1.7488774061203003, "learning_rate": 1.9900000000000003e-05, "loss": 0.5309, "step": 200 }, { "grad_norm": 1.9061145782470703, "learning_rate": 2.09e-05, "loss": 0.4664, "step": 210 }, { "grad_norm": 2.195199728012085, "learning_rate": 2.19e-05, "loss": 0.4032, "step": 220 }, { "grad_norm": 2.689815044403076, "learning_rate": 2.29e-05, "loss": 0.3339, "step": 230 }, { "grad_norm": 2.2053492069244385, "learning_rate": 2.39e-05, "loss": 0.2781, "step": 240 }, { "grad_norm": 2.095569610595703, "learning_rate": 2.4900000000000002e-05, "loss": 0.233, "step": 250 }, { "grad_norm": 2.3732612133026123, "learning_rate": 2.5900000000000003e-05, "loss": 0.1938, "step": 260 }, { "grad_norm": 2.2904515266418457, "learning_rate": 2.6900000000000003e-05, "loss": 0.1879, "step": 270 }, { "grad_norm": 1.7904020547866821, "learning_rate": 2.7900000000000004e-05, "loss": 0.1614, "step": 280 }, { "grad_norm": 2.3654394149780273, "learning_rate": 2.8899999999999998e-05, "loss": 0.1802, "step": 290 }, { "grad_norm": 2.3766984939575195, "learning_rate": 2.9900000000000002e-05, "loss": 0.1565, "step": 300 }, { "grad_norm": 2.1293435096740723, "learning_rate": 3.09e-05, "loss": 0.1624, "step": 310 }, { "grad_norm": 1.7178148031234741, "learning_rate": 3.19e-05, "loss": 0.1442, "step": 320 }, { "grad_norm": 2.282510280609131, "learning_rate": 3.29e-05, "loss": 0.1442, "step": 330 }, { "grad_norm": 1.9041420221328735, "learning_rate": 3.3900000000000004e-05, "loss": 0.1282, "step": 340 }, { "grad_norm": 1.7204537391662598, "learning_rate": 3.49e-05, "loss": 0.1145, "step": 350 }, { "grad_norm": 2.0920143127441406, "learning_rate": 3.59e-05, "loss": 0.123, "step": 360 }, { "grad_norm": 1.3838047981262207, "learning_rate": 3.69e-05, "loss": 0.1285, "step": 370 }, { "grad_norm": 1.8598517179489136, "learning_rate": 3.79e-05, "loss": 0.1198, "step": 380 }, { "grad_norm": 1.4385203123092651, "learning_rate": 3.8900000000000004e-05, "loss": 0.117, "step": 390 }, { "grad_norm": 2.0662553310394287, "learning_rate": 3.99e-05, "loss": 0.1036, "step": 400 }, { "grad_norm": 1.5192972421646118, "learning_rate": 4.09e-05, "loss": 0.1144, "step": 410 }, { "grad_norm": 2.0335099697113037, "learning_rate": 4.19e-05, "loss": 0.104, "step": 420 }, { "grad_norm": 1.6592355966567993, "learning_rate": 4.29e-05, "loss": 0.1107, "step": 430 }, { "grad_norm": 1.5217585563659668, "learning_rate": 4.39e-05, "loss": 0.1074, "step": 440 }, { "grad_norm": 1.1578655242919922, "learning_rate": 4.49e-05, "loss": 0.1072, "step": 450 }, { "grad_norm": 1.3794256448745728, "learning_rate": 4.5900000000000004e-05, "loss": 0.1064, "step": 460 }, { "grad_norm": 1.500340223312378, "learning_rate": 4.69e-05, "loss": 0.1057, "step": 470 }, { "grad_norm": 1.507816195487976, "learning_rate": 4.79e-05, "loss": 0.1062, "step": 480 }, { "grad_norm": 1.2124695777893066, "learning_rate": 4.89e-05, "loss": 0.0969, "step": 490 }, { "grad_norm": 1.200659155845642, "learning_rate": 4.99e-05, "loss": 0.0969, "step": 500 }, { "grad_norm": 1.790547490119934, "learning_rate": 5.0900000000000004e-05, "loss": 0.1125, "step": 510 }, { "grad_norm": 1.5206133127212524, "learning_rate": 5.19e-05, "loss": 0.1029, "step": 520 }, { "grad_norm": 1.3754093647003174, "learning_rate": 5.2900000000000005e-05, "loss": 0.1042, "step": 530 }, { "grad_norm": 1.0777167081832886, "learning_rate": 5.390000000000001e-05, "loss": 0.0867, "step": 540 }, { "grad_norm": 1.004372000694275, "learning_rate": 5.4900000000000006e-05, "loss": 0.091, "step": 550 }, { "grad_norm": 1.1158504486083984, "learning_rate": 5.590000000000001e-05, "loss": 0.0811, "step": 560 }, { "grad_norm": 1.4015352725982666, "learning_rate": 5.69e-05, "loss": 0.0882, "step": 570 }, { "grad_norm": 1.0852242708206177, "learning_rate": 5.79e-05, "loss": 0.0889, "step": 580 }, { "grad_norm": 1.306249737739563, "learning_rate": 5.89e-05, "loss": 0.0916, "step": 590 }, { "grad_norm": 1.0506494045257568, "learning_rate": 5.99e-05, "loss": 0.0921, "step": 600 }, { "grad_norm": 1.291216254234314, "learning_rate": 6.09e-05, "loss": 0.0986, "step": 610 }, { "grad_norm": 1.2934335470199585, "learning_rate": 6.19e-05, "loss": 0.0895, "step": 620 }, { "grad_norm": 0.8937121033668518, "learning_rate": 6.29e-05, "loss": 0.0836, "step": 630 }, { "grad_norm": 0.8294222950935364, "learning_rate": 6.390000000000001e-05, "loss": 0.0828, "step": 640 }, { "grad_norm": 1.2923587560653687, "learning_rate": 6.49e-05, "loss": 0.0813, "step": 650 }, { "grad_norm": 1.3499253988265991, "learning_rate": 6.59e-05, "loss": 0.0847, "step": 660 }, { "grad_norm": 1.268027901649475, "learning_rate": 6.690000000000001e-05, "loss": 0.099, "step": 670 }, { "grad_norm": 1.144601583480835, "learning_rate": 6.790000000000001e-05, "loss": 0.0833, "step": 680 }, { "grad_norm": 0.9558142423629761, "learning_rate": 6.89e-05, "loss": 0.0757, "step": 690 }, { "grad_norm": 1.0852667093276978, "learning_rate": 6.99e-05, "loss": 0.0798, "step": 700 }, { "grad_norm": 1.2283855676651, "learning_rate": 7.09e-05, "loss": 0.0931, "step": 710 }, { "grad_norm": 1.1257994174957275, "learning_rate": 7.19e-05, "loss": 0.0759, "step": 720 }, { "grad_norm": 1.0788460969924927, "learning_rate": 7.29e-05, "loss": 0.0796, "step": 730 }, { "grad_norm": 0.6946057081222534, "learning_rate": 7.390000000000001e-05, "loss": 0.0659, "step": 740 }, { "grad_norm": 0.9256339073181152, "learning_rate": 7.49e-05, "loss": 0.0765, "step": 750 }, { "grad_norm": 0.9185933470726013, "learning_rate": 7.59e-05, "loss": 0.0703, "step": 760 }, { "grad_norm": 0.6255671977996826, "learning_rate": 7.69e-05, "loss": 0.0765, "step": 770 }, { "grad_norm": 1.1388498544692993, "learning_rate": 7.790000000000001e-05, "loss": 0.0768, "step": 780 }, { "grad_norm": 0.8965200781822205, "learning_rate": 7.890000000000001e-05, "loss": 0.074, "step": 790 }, { "grad_norm": 0.9513829946517944, "learning_rate": 7.99e-05, "loss": 0.081, "step": 800 }, { "grad_norm": 0.7258803844451904, "learning_rate": 8.090000000000001e-05, "loss": 0.0823, "step": 810 }, { "grad_norm": 0.7350532412528992, "learning_rate": 8.19e-05, "loss": 0.073, "step": 820 }, { "grad_norm": 0.8834906220436096, "learning_rate": 8.29e-05, "loss": 0.0768, "step": 830 }, { "grad_norm": 0.5988731980323792, "learning_rate": 8.39e-05, "loss": 0.0692, "step": 840 }, { "grad_norm": 1.0358738899230957, "learning_rate": 8.49e-05, "loss": 0.0773, "step": 850 }, { "grad_norm": 0.8841147422790527, "learning_rate": 8.59e-05, "loss": 0.0664, "step": 860 }, { "grad_norm": 0.8775830864906311, "learning_rate": 8.69e-05, "loss": 0.0714, "step": 870 }, { "grad_norm": 1.0107133388519287, "learning_rate": 8.790000000000001e-05, "loss": 0.0697, "step": 880 }, { "grad_norm": 0.549202024936676, "learning_rate": 8.89e-05, "loss": 0.0676, "step": 890 }, { "grad_norm": 0.7344377636909485, "learning_rate": 8.99e-05, "loss": 0.0743, "step": 900 }, { "grad_norm": 0.7779461145401001, "learning_rate": 9.090000000000001e-05, "loss": 0.0596, "step": 910 }, { "grad_norm": 0.7221653461456299, "learning_rate": 9.190000000000001e-05, "loss": 0.0658, "step": 920 }, { "grad_norm": 0.6417852640151978, "learning_rate": 9.290000000000001e-05, "loss": 0.0584, "step": 930 }, { "grad_norm": 0.6676945686340332, "learning_rate": 9.39e-05, "loss": 0.0562, "step": 940 }, { "grad_norm": 0.7860474586486816, "learning_rate": 9.49e-05, "loss": 0.0691, "step": 950 }, { "grad_norm": 0.607173502445221, "learning_rate": 9.59e-05, "loss": 0.0567, "step": 960 }, { "grad_norm": 0.6575590372085571, "learning_rate": 9.69e-05, "loss": 0.0649, "step": 970 }, { "grad_norm": 0.7139838337898254, "learning_rate": 9.790000000000001e-05, "loss": 0.0691, "step": 980 }, { "grad_norm": 0.7679761648178101, "learning_rate": 9.89e-05, "loss": 0.0705, "step": 990 }, { "grad_norm": 0.8059155344963074, "learning_rate": 9.99e-05, "loss": 0.0701, "step": 1000 }, { "grad_norm": 0.7039210796356201, "learning_rate": 9.999994463727085e-05, "loss": 0.0682, "step": 1010 }, { "grad_norm": 0.6486578583717346, "learning_rate": 9.999975326009292e-05, "loss": 0.067, "step": 1020 }, { "grad_norm": 0.635718584060669, "learning_rate": 9.999942518549879e-05, "loss": 0.0695, "step": 1030 }, { "grad_norm": 0.6130461096763611, "learning_rate": 9.999896041438544e-05, "loss": 0.0651, "step": 1040 }, { "grad_norm": 0.5558497905731201, "learning_rate": 9.999835894802353e-05, "loss": 0.0585, "step": 1050 }, { "grad_norm": 0.7219929099082947, "learning_rate": 9.999762078805743e-05, "loss": 0.0719, "step": 1060 }, { "grad_norm": 0.6659247279167175, "learning_rate": 9.999674593650526e-05, "loss": 0.0613, "step": 1070 }, { "grad_norm": 0.7357816696166992, "learning_rate": 9.99957343957588e-05, "loss": 0.0624, "step": 1080 }, { "grad_norm": 0.8561655879020691, "learning_rate": 9.99945861685836e-05, "loss": 0.0515, "step": 1090 }, { "grad_norm": 0.44810089468955994, "learning_rate": 9.999330125811884e-05, "loss": 0.0581, "step": 1100 }, { "grad_norm": 0.5287982225418091, "learning_rate": 9.999187966787744e-05, "loss": 0.0618, "step": 1110 }, { "grad_norm": 0.6571439504623413, "learning_rate": 9.999032140174595e-05, "loss": 0.0615, "step": 1120 }, { "grad_norm": 0.6739640831947327, "learning_rate": 9.998862646398464e-05, "loss": 0.066, "step": 1130 }, { "grad_norm": 0.6945701837539673, "learning_rate": 9.998679485922739e-05, "loss": 0.0624, "step": 1140 }, { "grad_norm": 0.42567116022109985, "learning_rate": 9.998482659248174e-05, "loss": 0.0591, "step": 1150 }, { "grad_norm": 0.6355876922607422, "learning_rate": 9.998272166912883e-05, "loss": 0.0559, "step": 1160 }, { "grad_norm": 0.5608078837394714, "learning_rate": 9.998048009492347e-05, "loss": 0.0538, "step": 1170 }, { "grad_norm": 0.5491824746131897, "learning_rate": 9.997810187599403e-05, "loss": 0.0536, "step": 1180 }, { "grad_norm": 0.5426693558692932, "learning_rate": 9.997558701884249e-05, "loss": 0.0581, "step": 1190 }, { "grad_norm": 0.47622695565223694, "learning_rate": 9.997293553034433e-05, "loss": 0.0546, "step": 1200 }, { "grad_norm": 0.5745972990989685, "learning_rate": 9.997014741774866e-05, "loss": 0.0542, "step": 1210 }, { "grad_norm": 0.5194965600967407, "learning_rate": 9.996722268867803e-05, "loss": 0.0467, "step": 1220 }, { "grad_norm": 0.6288977861404419, "learning_rate": 9.996416135112858e-05, "loss": 0.0542, "step": 1230 }, { "grad_norm": 0.6587590575218201, "learning_rate": 9.996096341346988e-05, "loss": 0.0482, "step": 1240 }, { "grad_norm": 0.6076945662498474, "learning_rate": 9.995762888444495e-05, "loss": 0.0586, "step": 1250 }, { "grad_norm": 0.5410868525505066, "learning_rate": 9.995415777317027e-05, "loss": 0.0522, "step": 1260 }, { "grad_norm": 0.4868910312652588, "learning_rate": 9.995055008913574e-05, "loss": 0.0565, "step": 1270 }, { "grad_norm": 0.6964043378829956, "learning_rate": 9.994680584220463e-05, "loss": 0.0509, "step": 1280 }, { "grad_norm": 0.5489606857299805, "learning_rate": 9.994292504261355e-05, "loss": 0.0553, "step": 1290 }, { "grad_norm": 0.4580813944339752, "learning_rate": 9.993890770097247e-05, "loss": 0.0536, "step": 1300 }, { "grad_norm": 0.4927164614200592, "learning_rate": 9.993475382826467e-05, "loss": 0.0504, "step": 1310 }, { "grad_norm": 0.7281057834625244, "learning_rate": 9.993046343584664e-05, "loss": 0.0528, "step": 1320 }, { "grad_norm": 0.43749234080314636, "learning_rate": 9.992603653544816e-05, "loss": 0.046, "step": 1330 }, { "grad_norm": 0.6763569712638855, "learning_rate": 9.992147313917222e-05, "loss": 0.0473, "step": 1340 }, { "grad_norm": 0.5028315186500549, "learning_rate": 9.991677325949497e-05, "loss": 0.0422, "step": 1350 }, { "grad_norm": 0.5447657108306885, "learning_rate": 9.991193690926568e-05, "loss": 0.046, "step": 1360 }, { "grad_norm": 0.7757137417793274, "learning_rate": 9.990696410170678e-05, "loss": 0.0424, "step": 1370 }, { "grad_norm": 0.5636216402053833, "learning_rate": 9.990185485041371e-05, "loss": 0.0468, "step": 1380 }, { "grad_norm": 0.49852731823921204, "learning_rate": 9.989660916935498e-05, "loss": 0.0423, "step": 1390 }, { "grad_norm": 0.5682974457740784, "learning_rate": 9.989122707287208e-05, "loss": 0.0477, "step": 1400 }, { "grad_norm": 0.5859779715538025, "learning_rate": 9.988570857567945e-05, "loss": 0.0419, "step": 1410 }, { "grad_norm": 0.6125381588935852, "learning_rate": 9.988005369286446e-05, "loss": 0.0492, "step": 1420 }, { "grad_norm": 0.6116113066673279, "learning_rate": 9.987426243988734e-05, "loss": 0.0461, "step": 1430 }, { "grad_norm": 0.6193561553955078, "learning_rate": 9.986833483258114e-05, "loss": 0.0459, "step": 1440 }, { "grad_norm": 0.5412529110908508, "learning_rate": 9.986227088715173e-05, "loss": 0.0485, "step": 1450 }, { "grad_norm": 0.554725706577301, "learning_rate": 9.98560706201777e-05, "loss": 0.0456, "step": 1460 }, { "grad_norm": 0.5793951749801636, "learning_rate": 9.984973404861036e-05, "loss": 0.0487, "step": 1470 }, { "grad_norm": 0.5661277770996094, "learning_rate": 9.984326118977361e-05, "loss": 0.053, "step": 1480 }, { "grad_norm": 0.5117608904838562, "learning_rate": 9.983665206136406e-05, "loss": 0.0448, "step": 1490 }, { "grad_norm": 0.6818128824234009, "learning_rate": 9.982990668145075e-05, "loss": 0.0549, "step": 1500 }, { "grad_norm": 0.4928187429904938, "learning_rate": 9.982302506847534e-05, "loss": 0.0441, "step": 1510 }, { "grad_norm": 0.5571991801261902, "learning_rate": 9.981600724125189e-05, "loss": 0.0451, "step": 1520 }, { "grad_norm": 0.4613161087036133, "learning_rate": 9.980885321896685e-05, "loss": 0.046, "step": 1530 }, { "grad_norm": 0.6609203219413757, "learning_rate": 9.980156302117905e-05, "loss": 0.0499, "step": 1540 }, { "grad_norm": 0.40431761741638184, "learning_rate": 9.979413666781963e-05, "loss": 0.0498, "step": 1550 }, { "grad_norm": 0.4145183861255646, "learning_rate": 9.978657417919193e-05, "loss": 0.0456, "step": 1560 }, { "grad_norm": 0.4497072100639343, "learning_rate": 9.977887557597153e-05, "loss": 0.0446, "step": 1570 }, { "grad_norm": 0.6474009156227112, "learning_rate": 9.97710408792061e-05, "loss": 0.0455, "step": 1580 }, { "grad_norm": 0.5270069241523743, "learning_rate": 9.976307011031542e-05, "loss": 0.0455, "step": 1590 }, { "grad_norm": 0.5787636041641235, "learning_rate": 9.975496329109126e-05, "loss": 0.0459, "step": 1600 }, { "grad_norm": 0.4643445909023285, "learning_rate": 9.974672044369732e-05, "loss": 0.0395, "step": 1610 }, { "grad_norm": 0.4459812343120575, "learning_rate": 9.97383415906693e-05, "loss": 0.0394, "step": 1620 }, { "grad_norm": 0.6048677563667297, "learning_rate": 9.97298267549146e-05, "loss": 0.0451, "step": 1630 }, { "grad_norm": 0.5167340636253357, "learning_rate": 9.972117595971249e-05, "loss": 0.0446, "step": 1640 }, { "grad_norm": 0.5302894711494446, "learning_rate": 9.971238922871391e-05, "loss": 0.0446, "step": 1650 }, { "grad_norm": 0.36176666617393494, "learning_rate": 9.970346658594142e-05, "loss": 0.0336, "step": 1660 }, { "grad_norm": 0.6387920379638672, "learning_rate": 9.969440805578923e-05, "loss": 0.0454, "step": 1670 }, { "grad_norm": 0.4556961953639984, "learning_rate": 9.968521366302298e-05, "loss": 0.0482, "step": 1680 }, { "grad_norm": 0.5278778672218323, "learning_rate": 9.967588343277981e-05, "loss": 0.0476, "step": 1690 }, { "grad_norm": 0.6947939395904541, "learning_rate": 9.966641739056818e-05, "loss": 0.0408, "step": 1700 }, { "grad_norm": 0.4832363426685333, "learning_rate": 9.965681556226793e-05, "loss": 0.0421, "step": 1710 }, { "grad_norm": 0.5226179957389832, "learning_rate": 9.964707797413006e-05, "loss": 0.0502, "step": 1720 }, { "grad_norm": 0.5860890746116638, "learning_rate": 9.963720465277679e-05, "loss": 0.0447, "step": 1730 }, { "grad_norm": 0.3182539641857147, "learning_rate": 9.96271956252014e-05, "loss": 0.0379, "step": 1740 }, { "grad_norm": 0.5729960203170776, "learning_rate": 9.961705091876816e-05, "loss": 0.0421, "step": 1750 }, { "grad_norm": 0.3992767930030823, "learning_rate": 9.960677056121235e-05, "loss": 0.0432, "step": 1760 }, { "grad_norm": 0.5867078900337219, "learning_rate": 9.959635458064005e-05, "loss": 0.0427, "step": 1770 }, { "grad_norm": 0.42889097332954407, "learning_rate": 9.958580300552815e-05, "loss": 0.0425, "step": 1780 }, { "grad_norm": 0.418584942817688, "learning_rate": 9.957511586472426e-05, "loss": 0.0482, "step": 1790 }, { "grad_norm": 0.5036026239395142, "learning_rate": 9.956429318744662e-05, "loss": 0.0403, "step": 1800 }, { "grad_norm": 0.4868301451206207, "learning_rate": 9.955333500328404e-05, "loss": 0.0399, "step": 1810 }, { "grad_norm": 0.5466697216033936, "learning_rate": 9.95422413421957e-05, "loss": 0.0392, "step": 1820 }, { "grad_norm": 0.5338637232780457, "learning_rate": 9.953101223451133e-05, "loss": 0.0381, "step": 1830 }, { "grad_norm": 0.4507163465023041, "learning_rate": 9.951964771093085e-05, "loss": 0.0407, "step": 1840 }, { "grad_norm": 0.44600731134414673, "learning_rate": 9.950814780252442e-05, "loss": 0.0391, "step": 1850 }, { "grad_norm": 0.4139569103717804, "learning_rate": 9.949651254073236e-05, "loss": 0.0417, "step": 1860 }, { "grad_norm": 0.4858010411262512, "learning_rate": 9.948474195736504e-05, "loss": 0.0415, "step": 1870 }, { "grad_norm": 0.4814279079437256, "learning_rate": 9.947283608460277e-05, "loss": 0.0384, "step": 1880 }, { "grad_norm": 0.5872435569763184, "learning_rate": 9.946079495499577e-05, "loss": 0.0377, "step": 1890 }, { "grad_norm": 0.46607449650764465, "learning_rate": 9.944861860146401e-05, "loss": 0.0461, "step": 1900 }, { "grad_norm": 0.47768518328666687, "learning_rate": 9.943630705729719e-05, "loss": 0.0417, "step": 1910 }, { "grad_norm": 0.3898749351501465, "learning_rate": 9.942386035615459e-05, "loss": 0.04, "step": 1920 }, { "grad_norm": 0.5304390788078308, "learning_rate": 9.941127853206503e-05, "loss": 0.0409, "step": 1930 }, { "grad_norm": 0.46548062562942505, "learning_rate": 9.939856161942673e-05, "loss": 0.0405, "step": 1940 }, { "grad_norm": 0.4487619400024414, "learning_rate": 9.938570965300724e-05, "loss": 0.0416, "step": 1950 }, { "grad_norm": 0.3904606103897095, "learning_rate": 9.937272266794335e-05, "loss": 0.0342, "step": 1960 }, { "grad_norm": 0.4962557554244995, "learning_rate": 9.935960069974096e-05, "loss": 0.0407, "step": 1970 }, { "grad_norm": 0.5208577513694763, "learning_rate": 9.934634378427506e-05, "loss": 0.0451, "step": 1980 }, { "grad_norm": 0.4521556496620178, "learning_rate": 9.933295195778954e-05, "loss": 0.0387, "step": 1990 }, { "grad_norm": 0.482989639043808, "learning_rate": 9.931942525689715e-05, "loss": 0.0359, "step": 2000 }, { "grad_norm": 0.4423094689846039, "learning_rate": 9.930576371857936e-05, "loss": 0.0325, "step": 2010 }, { "grad_norm": 0.395851731300354, "learning_rate": 9.929196738018629e-05, "loss": 0.0362, "step": 2020 }, { "grad_norm": 0.4394199848175049, "learning_rate": 9.927803627943662e-05, "loss": 0.0397, "step": 2030 }, { "grad_norm": 0.4384741485118866, "learning_rate": 9.926397045441744e-05, "loss": 0.0421, "step": 2040 }, { "grad_norm": 0.38732296228408813, "learning_rate": 9.924976994358417e-05, "loss": 0.0358, "step": 2050 }, { "grad_norm": 0.3381761610507965, "learning_rate": 9.923543478576048e-05, "loss": 0.0336, "step": 2060 }, { "grad_norm": 0.3188372254371643, "learning_rate": 9.922096502013813e-05, "loss": 0.0343, "step": 2070 }, { "grad_norm": 0.2466750591993332, "learning_rate": 9.92063606862769e-05, "loss": 0.032, "step": 2080 }, { "grad_norm": 0.3608834445476532, "learning_rate": 9.919162182410453e-05, "loss": 0.031, "step": 2090 }, { "grad_norm": 0.37712398171424866, "learning_rate": 9.917674847391645e-05, "loss": 0.0322, "step": 2100 }, { "grad_norm": 0.4192401170730591, "learning_rate": 9.916174067637584e-05, "loss": 0.0321, "step": 2110 }, { "grad_norm": 0.47337812185287476, "learning_rate": 9.914659847251348e-05, "loss": 0.0304, "step": 2120 }, { "grad_norm": 0.4168669879436493, "learning_rate": 9.913132190372753e-05, "loss": 0.0349, "step": 2130 }, { "grad_norm": 0.5130103230476379, "learning_rate": 9.911591101178359e-05, "loss": 0.0369, "step": 2140 }, { "grad_norm": 0.5027634501457214, "learning_rate": 9.910036583881443e-05, "loss": 0.0426, "step": 2150 }, { "grad_norm": 0.4033798575401306, "learning_rate": 9.908468642731995e-05, "loss": 0.038, "step": 2160 }, { "grad_norm": 0.3905620276927948, "learning_rate": 9.906887282016707e-05, "loss": 0.0354, "step": 2170 }, { "grad_norm": 0.5748849511146545, "learning_rate": 9.90529250605896e-05, "loss": 0.0372, "step": 2180 }, { "grad_norm": 0.35952481627464294, "learning_rate": 9.903684319218809e-05, "loss": 0.0328, "step": 2190 }, { "grad_norm": 0.44712838530540466, "learning_rate": 9.902062725892976e-05, "loss": 0.0358, "step": 2200 }, { "grad_norm": 0.4204244911670685, "learning_rate": 9.900427730514834e-05, "loss": 0.0316, "step": 2210 }, { "grad_norm": 0.47271594405174255, "learning_rate": 9.8987793375544e-05, "loss": 0.0309, "step": 2220 }, { "grad_norm": 0.5245261788368225, "learning_rate": 9.897117551518318e-05, "loss": 0.0364, "step": 2230 }, { "grad_norm": 0.37689557671546936, "learning_rate": 9.895442376949844e-05, "loss": 0.0331, "step": 2240 }, { "grad_norm": 0.4505006968975067, "learning_rate": 9.893753818428845e-05, "loss": 0.0396, "step": 2250 }, { "grad_norm": 0.3736228346824646, "learning_rate": 9.892051880571773e-05, "loss": 0.0339, "step": 2260 }, { "grad_norm": 0.40342840552330017, "learning_rate": 9.890336568031663e-05, "loss": 0.0336, "step": 2270 }, { "grad_norm": 0.38836848735809326, "learning_rate": 9.888607885498113e-05, "loss": 0.0387, "step": 2280 }, { "grad_norm": 0.3188624083995819, "learning_rate": 9.886865837697275e-05, "loss": 0.0369, "step": 2290 }, { "grad_norm": 0.4409078359603882, "learning_rate": 9.88511042939184e-05, "loss": 0.0338, "step": 2300 }, { "grad_norm": 0.6200244426727295, "learning_rate": 9.883341665381028e-05, "loss": 0.0415, "step": 2310 }, { "grad_norm": 0.4343007802963257, "learning_rate": 9.881559550500575e-05, "loss": 0.035, "step": 2320 }, { "grad_norm": 0.3987315893173218, "learning_rate": 9.879764089622712e-05, "loss": 0.0467, "step": 2330 }, { "grad_norm": 0.402128666639328, "learning_rate": 9.87795528765616e-05, "loss": 0.0363, "step": 2340 }, { "grad_norm": 0.5304055213928223, "learning_rate": 9.876133149546118e-05, "loss": 0.0404, "step": 2350 }, { "grad_norm": 0.3402189314365387, "learning_rate": 9.874297680274238e-05, "loss": 0.0357, "step": 2360 }, { "grad_norm": 0.24317887425422668, "learning_rate": 9.872448884858624e-05, "loss": 0.0344, "step": 2370 }, { "grad_norm": 0.45474639534950256, "learning_rate": 9.870586768353815e-05, "loss": 0.032, "step": 2380 }, { "grad_norm": 0.4027414619922638, "learning_rate": 9.868711335850764e-05, "loss": 0.0359, "step": 2390 }, { "grad_norm": 0.3445315659046173, "learning_rate": 9.866822592476833e-05, "loss": 0.0376, "step": 2400 }, { "grad_norm": 0.3709997236728668, "learning_rate": 9.86492054339577e-05, "loss": 0.032, "step": 2410 }, { "grad_norm": 0.41493022441864014, "learning_rate": 9.863005193807711e-05, "loss": 0.0327, "step": 2420 }, { "grad_norm": 0.4496617913246155, "learning_rate": 9.861076548949143e-05, "loss": 0.0318, "step": 2430 }, { "grad_norm": 0.41117063164711, "learning_rate": 9.859134614092912e-05, "loss": 0.033, "step": 2440 }, { "grad_norm": 0.5059201121330261, "learning_rate": 9.857179394548191e-05, "loss": 0.0319, "step": 2450 }, { "grad_norm": 0.3489936888217926, "learning_rate": 9.855210895660477e-05, "loss": 0.0363, "step": 2460 }, { "grad_norm": 0.4177769124507904, "learning_rate": 9.853229122811568e-05, "loss": 0.0316, "step": 2470 }, { "grad_norm": 0.487947553396225, "learning_rate": 9.851234081419559e-05, "loss": 0.031, "step": 2480 }, { "grad_norm": 0.4785672426223755, "learning_rate": 9.849225776938814e-05, "loss": 0.0288, "step": 2490 }, { "grad_norm": 0.45622044801712036, "learning_rate": 9.847204214859964e-05, "loss": 0.0361, "step": 2500 }, { "grad_norm": 0.43176940083503723, "learning_rate": 9.845169400709879e-05, "loss": 0.0328, "step": 2510 }, { "grad_norm": 0.3938290476799011, "learning_rate": 9.843121340051664e-05, "loss": 0.0342, "step": 2520 }, { "grad_norm": 0.4284602999687195, "learning_rate": 9.841060038484641e-05, "loss": 0.0296, "step": 2530 }, { "grad_norm": 0.46028992533683777, "learning_rate": 9.838985501644328e-05, "loss": 0.0325, "step": 2540 }, { "grad_norm": 0.35803061723709106, "learning_rate": 9.83689773520243e-05, "loss": 0.0403, "step": 2550 }, { "grad_norm": 0.44375649094581604, "learning_rate": 9.834796744866819e-05, "loss": 0.0358, "step": 2560 }, { "grad_norm": 0.3268609941005707, "learning_rate": 9.832682536381525e-05, "loss": 0.0327, "step": 2570 }, { "grad_norm": 0.3023513853549957, "learning_rate": 9.830555115526711e-05, "loss": 0.0369, "step": 2580 }, { "grad_norm": 0.39501333236694336, "learning_rate": 9.828414488118667e-05, "loss": 0.0351, "step": 2590 }, { "grad_norm": 0.32449448108673096, "learning_rate": 9.826260660009785e-05, "loss": 0.0314, "step": 2600 }, { "grad_norm": 0.3975260853767395, "learning_rate": 9.824093637088547e-05, "loss": 0.03, "step": 2610 }, { "grad_norm": 0.3904464840888977, "learning_rate": 9.821913425279514e-05, "loss": 0.0335, "step": 2620 }, { "grad_norm": 0.429296612739563, "learning_rate": 9.8197200305433e-05, "loss": 0.0316, "step": 2630 }, { "grad_norm": 0.35084664821624756, "learning_rate": 9.817513458876564e-05, "loss": 0.0294, "step": 2640 }, { "grad_norm": 0.5197656750679016, "learning_rate": 9.815293716311987e-05, "loss": 0.0342, "step": 2650 }, { "grad_norm": 0.28495723009109497, "learning_rate": 9.813060808918262e-05, "loss": 0.0424, "step": 2660 }, { "grad_norm": 0.36250951886177063, "learning_rate": 9.810814742800069e-05, "loss": 0.0367, "step": 2670 }, { "grad_norm": 0.3054046928882599, "learning_rate": 9.808555524098074e-05, "loss": 0.0381, "step": 2680 }, { "grad_norm": 0.39500221610069275, "learning_rate": 9.806283158988887e-05, "loss": 0.0339, "step": 2690 }, { "grad_norm": 0.3321908116340637, "learning_rate": 9.803997653685072e-05, "loss": 0.0366, "step": 2700 }, { "grad_norm": 0.4531093239784241, "learning_rate": 9.801699014435112e-05, "loss": 0.0365, "step": 2710 }, { "grad_norm": 0.42526277899742126, "learning_rate": 9.799387247523398e-05, "loss": 0.0339, "step": 2720 }, { "grad_norm": 0.42415285110473633, "learning_rate": 9.797062359270215e-05, "loss": 0.0328, "step": 2730 }, { "grad_norm": 0.3714810907840729, "learning_rate": 9.794724356031715e-05, "loss": 0.0363, "step": 2740 }, { "grad_norm": 0.3706148564815521, "learning_rate": 9.792373244199913e-05, "loss": 0.0308, "step": 2750 }, { "grad_norm": 0.34478428959846497, "learning_rate": 9.790009030202658e-05, "loss": 0.0337, "step": 2760 }, { "grad_norm": 0.35362887382507324, "learning_rate": 9.78763172050362e-05, "loss": 0.031, "step": 2770 }, { "grad_norm": 0.42618703842163086, "learning_rate": 9.785241321602274e-05, "loss": 0.0303, "step": 2780 }, { "grad_norm": 0.28176233172416687, "learning_rate": 9.782837840033879e-05, "loss": 0.0304, "step": 2790 }, { "grad_norm": 0.4290007948875427, "learning_rate": 9.780421282369461e-05, "loss": 0.032, "step": 2800 }, { "grad_norm": 0.4026869237422943, "learning_rate": 9.777991655215797e-05, "loss": 0.0332, "step": 2810 }, { "grad_norm": 0.4786500334739685, "learning_rate": 9.775548965215394e-05, "loss": 0.0307, "step": 2820 }, { "grad_norm": 0.3217422068119049, "learning_rate": 9.773093219046474e-05, "loss": 0.031, "step": 2830 }, { "grad_norm": 0.32174065709114075, "learning_rate": 9.770624423422954e-05, "loss": 0.0327, "step": 2840 }, { "grad_norm": 0.40603211522102356, "learning_rate": 9.768142585094426e-05, "loss": 0.0324, "step": 2850 }, { "grad_norm": 0.3074100613594055, "learning_rate": 9.765647710846142e-05, "loss": 0.0319, "step": 2860 }, { "grad_norm": 0.33911457657814026, "learning_rate": 9.763139807498991e-05, "loss": 0.0294, "step": 2870 }, { "grad_norm": 0.3367382884025574, "learning_rate": 9.760618881909487e-05, "loss": 0.029, "step": 2880 }, { "grad_norm": 0.42009615898132324, "learning_rate": 9.758084940969744e-05, "loss": 0.03, "step": 2890 }, { "grad_norm": 0.2921687662601471, "learning_rate": 9.755537991607459e-05, "loss": 0.0323, "step": 2900 }, { "grad_norm": 0.38551318645477295, "learning_rate": 9.752978040785895e-05, "loss": 0.0404, "step": 2910 }, { "grad_norm": 0.32184872031211853, "learning_rate": 9.750405095503859e-05, "loss": 0.0325, "step": 2920 }, { "grad_norm": 0.43612349033355713, "learning_rate": 9.747819162795686e-05, "loss": 0.0322, "step": 2930 }, { "grad_norm": 0.2535778880119324, "learning_rate": 9.745220249731217e-05, "loss": 0.031, "step": 2940 }, { "grad_norm": 0.36052781343460083, "learning_rate": 9.742608363415781e-05, "loss": 0.0302, "step": 2950 }, { "grad_norm": 0.45721063017845154, "learning_rate": 9.739983510990176e-05, "loss": 0.0343, "step": 2960 }, { "grad_norm": 0.48942631483078003, "learning_rate": 9.737345699630647e-05, "loss": 0.0342, "step": 2970 }, { "grad_norm": 0.3808375298976898, "learning_rate": 9.734694936548869e-05, "loss": 0.0291, "step": 2980 }, { "grad_norm": 0.37008482217788696, "learning_rate": 9.732031228991932e-05, "loss": 0.0297, "step": 2990 }, { "grad_norm": 0.29760533571243286, "learning_rate": 9.729354584242302e-05, "loss": 0.0333, "step": 3000 }, { "grad_norm": 0.4348229169845581, "learning_rate": 9.726665009617832e-05, "loss": 0.0317, "step": 3010 }, { "grad_norm": 0.43272364139556885, "learning_rate": 9.723962512471714e-05, "loss": 0.0334, "step": 3020 }, { "grad_norm": 0.26072654128074646, "learning_rate": 9.72124710019247e-05, "loss": 0.0323, "step": 3030 }, { "grad_norm": 0.2522956132888794, "learning_rate": 9.718518780203934e-05, "loss": 0.0279, "step": 3040 }, { "grad_norm": 0.32629165053367615, "learning_rate": 9.715777559965228e-05, "loss": 0.0318, "step": 3050 }, { "grad_norm": 0.3961840569972992, "learning_rate": 9.713023446970746e-05, "loss": 0.0335, "step": 3060 }, { "grad_norm": 0.39364710450172424, "learning_rate": 9.710256448750126e-05, "loss": 0.0363, "step": 3070 }, { "grad_norm": 0.3732122778892517, "learning_rate": 9.707476572868235e-05, "loss": 0.0307, "step": 3080 }, { "grad_norm": 0.428852915763855, "learning_rate": 9.704683826925149e-05, "loss": 0.0313, "step": 3090 }, { "grad_norm": 0.3512025773525238, "learning_rate": 9.701878218556129e-05, "loss": 0.0316, "step": 3100 }, { "grad_norm": 0.28259706497192383, "learning_rate": 9.699059755431598e-05, "loss": 0.0323, "step": 3110 }, { "grad_norm": 0.3300177752971649, "learning_rate": 9.696228445257132e-05, "loss": 0.0295, "step": 3120 }, { "grad_norm": 0.3097679316997528, "learning_rate": 9.693384295773419e-05, "loss": 0.0325, "step": 3130 }, { "grad_norm": 0.2780998945236206, "learning_rate": 9.690527314756259e-05, "loss": 0.0285, "step": 3140 }, { "grad_norm": 0.29380565881729126, "learning_rate": 9.687657510016527e-05, "loss": 0.0292, "step": 3150 }, { "grad_norm": 0.4242912530899048, "learning_rate": 9.684774889400161e-05, "loss": 0.0274, "step": 3160 }, { "grad_norm": 0.40963923931121826, "learning_rate": 9.681879460788135e-05, "loss": 0.0331, "step": 3170 }, { "grad_norm": 0.4268638789653778, "learning_rate": 9.67897123209644e-05, "loss": 0.03, "step": 3180 }, { "grad_norm": 0.4310980439186096, "learning_rate": 9.676050211276062e-05, "loss": 0.0319, "step": 3190 }, { "grad_norm": 0.3833226263523102, "learning_rate": 9.673116406312962e-05, "loss": 0.0335, "step": 3200 }, { "grad_norm": 0.45712000131607056, "learning_rate": 9.67016982522805e-05, "loss": 0.0318, "step": 3210 }, { "grad_norm": 0.414510577917099, "learning_rate": 9.667210476077164e-05, "loss": 0.0314, "step": 3220 }, { "grad_norm": 0.31404972076416016, "learning_rate": 9.664238366951055e-05, "loss": 0.0302, "step": 3230 }, { "grad_norm": 0.3177141845226288, "learning_rate": 9.661253505975355e-05, "loss": 0.0273, "step": 3240 }, { "grad_norm": 0.46404072642326355, "learning_rate": 9.658255901310557e-05, "loss": 0.0258, "step": 3250 }, { "grad_norm": 0.32448244094848633, "learning_rate": 9.655245561152e-05, "loss": 0.0342, "step": 3260 }, { "grad_norm": 0.40415772795677185, "learning_rate": 9.65222249372984e-05, "loss": 0.0318, "step": 3270 }, { "grad_norm": 0.3887462317943573, "learning_rate": 9.649186707309026e-05, "loss": 0.0293, "step": 3280 }, { "grad_norm": 0.30810287594795227, "learning_rate": 9.646138210189283e-05, "loss": 0.03, "step": 3290 }, { "grad_norm": 0.46965813636779785, "learning_rate": 9.643077010705087e-05, "loss": 0.0272, "step": 3300 }, { "grad_norm": 0.3931674361228943, "learning_rate": 9.640003117225637e-05, "loss": 0.0269, "step": 3310 }, { "grad_norm": 0.3708631098270416, "learning_rate": 9.636916538154846e-05, "loss": 0.0271, "step": 3320 }, { "grad_norm": 0.3901376724243164, "learning_rate": 9.633817281931296e-05, "loss": 0.0287, "step": 3330 }, { "grad_norm": 0.34486785531044006, "learning_rate": 9.630705357028242e-05, "loss": 0.029, "step": 3340 }, { "grad_norm": 0.34120044112205505, "learning_rate": 9.627580771953563e-05, "loss": 0.0365, "step": 3350 }, { "grad_norm": 0.32826319336891174, "learning_rate": 9.624443535249759e-05, "loss": 0.0304, "step": 3360 }, { "grad_norm": 0.415689617395401, "learning_rate": 9.621293655493913e-05, "loss": 0.0382, "step": 3370 }, { "grad_norm": 0.33886831998825073, "learning_rate": 9.618131141297675e-05, "loss": 0.0343, "step": 3380 }, { "grad_norm": 0.3342231214046478, "learning_rate": 9.614956001307242e-05, "loss": 0.0342, "step": 3390 }, { "grad_norm": 0.3477620780467987, "learning_rate": 9.611768244203321e-05, "loss": 0.0392, "step": 3400 }, { "grad_norm": 0.28709501028060913, "learning_rate": 9.60856787870112e-05, "loss": 0.0318, "step": 3410 }, { "grad_norm": 0.34953030943870544, "learning_rate": 9.605354913550318e-05, "loss": 0.0325, "step": 3420 }, { "grad_norm": 0.3966991901397705, "learning_rate": 9.602129357535037e-05, "loss": 0.0277, "step": 3430 }, { "grad_norm": 0.3972017467021942, "learning_rate": 9.598891219473825e-05, "loss": 0.0313, "step": 3440 }, { "grad_norm": 0.29781433939933777, "learning_rate": 9.595640508219625e-05, "loss": 0.026, "step": 3450 }, { "grad_norm": 0.4594935476779938, "learning_rate": 9.592377232659761e-05, "loss": 0.0336, "step": 3460 }, { "grad_norm": 0.20357000827789307, "learning_rate": 9.589101401715904e-05, "loss": 0.0272, "step": 3470 }, { "grad_norm": 0.37121161818504333, "learning_rate": 9.585813024344045e-05, "loss": 0.0317, "step": 3480 }, { "grad_norm": 0.4270089566707611, "learning_rate": 9.58251210953449e-05, "loss": 0.0323, "step": 3490 }, { "grad_norm": 0.37827327847480774, "learning_rate": 9.579198666311809e-05, "loss": 0.0375, "step": 3500 }, { "grad_norm": 0.34917041659355164, "learning_rate": 9.575872703734832e-05, "loss": 0.0312, "step": 3510 }, { "grad_norm": 0.24639607965946198, "learning_rate": 9.572534230896611e-05, "loss": 0.028, "step": 3520 }, { "grad_norm": 0.2964099645614624, "learning_rate": 9.569183256924403e-05, "loss": 0.0269, "step": 3530 }, { "grad_norm": 0.35584557056427, "learning_rate": 9.565819790979646e-05, "loss": 0.0333, "step": 3540 }, { "grad_norm": 0.37006357312202454, "learning_rate": 9.562443842257925e-05, "loss": 0.0318, "step": 3550 }, { "grad_norm": 0.49795934557914734, "learning_rate": 9.559055419988956e-05, "loss": 0.0336, "step": 3560 }, { "grad_norm": 0.3791080117225647, "learning_rate": 9.555654533436557e-05, "loss": 0.0336, "step": 3570 }, { "grad_norm": 0.297141432762146, "learning_rate": 9.552241191898621e-05, "loss": 0.0318, "step": 3580 }, { "grad_norm": 0.3161041736602783, "learning_rate": 9.548815404707092e-05, "loss": 0.0285, "step": 3590 }, { "grad_norm": 0.3300725519657135, "learning_rate": 9.545377181227942e-05, "loss": 0.0308, "step": 3600 }, { "grad_norm": 0.3180306553840637, "learning_rate": 9.541926530861145e-05, "loss": 0.0292, "step": 3610 }, { "grad_norm": 0.32853659987449646, "learning_rate": 9.538463463040645e-05, "loss": 0.0306, "step": 3620 }, { "grad_norm": 0.3380567133426666, "learning_rate": 9.534987987234337e-05, "loss": 0.0304, "step": 3630 }, { "grad_norm": 0.34742414951324463, "learning_rate": 9.53150011294404e-05, "loss": 0.0346, "step": 3640 }, { "grad_norm": 0.4114944338798523, "learning_rate": 9.527999849705471e-05, "loss": 0.0341, "step": 3650 }, { "grad_norm": 0.3015265166759491, "learning_rate": 9.524487207088213e-05, "loss": 0.0278, "step": 3660 }, { "grad_norm": 0.35614529252052307, "learning_rate": 9.520962194695698e-05, "loss": 0.0302, "step": 3670 }, { "grad_norm": 0.3504146337509155, "learning_rate": 9.517424822165175e-05, "loss": 0.033, "step": 3680 }, { "grad_norm": 0.3307168185710907, "learning_rate": 9.513875099167685e-05, "loss": 0.0339, "step": 3690 }, { "grad_norm": 0.30637937784194946, "learning_rate": 9.510313035408035e-05, "loss": 0.0294, "step": 3700 }, { "grad_norm": 0.4314155578613281, "learning_rate": 9.506738640624775e-05, "loss": 0.034, "step": 3710 }, { "grad_norm": 0.27219170331954956, "learning_rate": 9.50315192459016e-05, "loss": 0.0326, "step": 3720 }, { "grad_norm": 0.3755035400390625, "learning_rate": 9.499552897110136e-05, "loss": 0.0291, "step": 3730 }, { "grad_norm": 0.34746628999710083, "learning_rate": 9.495941568024304e-05, "loss": 0.0325, "step": 3740 }, { "grad_norm": 0.38253656029701233, "learning_rate": 9.492317947205904e-05, "loss": 0.0304, "step": 3750 }, { "grad_norm": 0.2847307622432709, "learning_rate": 9.488682044561775e-05, "loss": 0.029, "step": 3760 }, { "grad_norm": 0.2772328555583954, "learning_rate": 9.485033870032335e-05, "loss": 0.0327, "step": 3770 }, { "grad_norm": 0.26286420226097107, "learning_rate": 9.481373433591556e-05, "loss": 0.0292, "step": 3780 }, { "grad_norm": 0.25149205327033997, "learning_rate": 9.47770074524693e-05, "loss": 0.026, "step": 3790 }, { "grad_norm": 0.26283958554267883, "learning_rate": 9.474015815039446e-05, "loss": 0.0297, "step": 3800 }, { "grad_norm": 0.265097051858902, "learning_rate": 9.470318653043565e-05, "loss": 0.025, "step": 3810 }, { "grad_norm": 0.2946009635925293, "learning_rate": 9.466609269367185e-05, "loss": 0.0301, "step": 3820 }, { "grad_norm": 0.34833046793937683, "learning_rate": 9.46288767415162e-05, "loss": 0.0295, "step": 3830 }, { "grad_norm": 0.3648533821105957, "learning_rate": 9.459153877571567e-05, "loss": 0.029, "step": 3840 }, { "grad_norm": 0.4258934259414673, "learning_rate": 9.455407889835087e-05, "loss": 0.0287, "step": 3850 }, { "grad_norm": 0.2937564253807068, "learning_rate": 9.451649721183564e-05, "loss": 0.0322, "step": 3860 }, { "grad_norm": 0.39758917689323425, "learning_rate": 9.447879381891692e-05, "loss": 0.0319, "step": 3870 }, { "grad_norm": 0.29667356610298157, "learning_rate": 9.444096882267428e-05, "loss": 0.0277, "step": 3880 }, { "grad_norm": 0.24447479844093323, "learning_rate": 9.440302232651988e-05, "loss": 0.0269, "step": 3890 }, { "grad_norm": 0.2567999064922333, "learning_rate": 9.436495443419795e-05, "loss": 0.0319, "step": 3900 }, { "grad_norm": 0.33196911215782166, "learning_rate": 9.432676524978466e-05, "loss": 0.0315, "step": 3910 }, { "grad_norm": 0.34205952286720276, "learning_rate": 9.42884548776878e-05, "loss": 0.0292, "step": 3920 }, { "grad_norm": 0.22506408393383026, "learning_rate": 9.425002342264646e-05, "loss": 0.031, "step": 3930 }, { "grad_norm": 0.3844817876815796, "learning_rate": 9.421147098973077e-05, "loss": 0.0308, "step": 3940 }, { "grad_norm": 0.279226690530777, "learning_rate": 9.41727976843416e-05, "loss": 0.0324, "step": 3950 }, { "grad_norm": 0.3045192062854767, "learning_rate": 9.413400361221029e-05, "loss": 0.0269, "step": 3960 }, { "grad_norm": 0.28793007135391235, "learning_rate": 9.409508887939835e-05, "loss": 0.0278, "step": 3970 }, { "grad_norm": 0.32803335785865784, "learning_rate": 9.40560535922972e-05, "loss": 0.0265, "step": 3980 }, { "grad_norm": 0.2827959656715393, "learning_rate": 9.40168978576278e-05, "loss": 0.0283, "step": 3990 }, { "grad_norm": 0.2905590236186981, "learning_rate": 9.397762178244043e-05, "loss": 0.0272, "step": 4000 }, { "grad_norm": 0.29874128103256226, "learning_rate": 9.393822547411439e-05, "loss": 0.0266, "step": 4010 }, { "grad_norm": 0.3884505033493042, "learning_rate": 9.389870904035769e-05, "loss": 0.0249, "step": 4020 }, { "grad_norm": 0.36586472392082214, "learning_rate": 9.385907258920672e-05, "loss": 0.0294, "step": 4030 }, { "grad_norm": 0.2418658584356308, "learning_rate": 9.381931622902607e-05, "loss": 0.0299, "step": 4040 }, { "grad_norm": 0.25352907180786133, "learning_rate": 9.377944006850807e-05, "loss": 0.026, "step": 4050 }, { "grad_norm": 0.38096845149993896, "learning_rate": 9.373944421667265e-05, "loss": 0.0335, "step": 4060 }, { "grad_norm": 0.24526287615299225, "learning_rate": 9.369932878286691e-05, "loss": 0.0282, "step": 4070 }, { "grad_norm": 0.3991052210330963, "learning_rate": 9.365909387676494e-05, "loss": 0.0263, "step": 4080 }, { "grad_norm": 0.41774439811706543, "learning_rate": 9.361873960836744e-05, "loss": 0.0276, "step": 4090 }, { "grad_norm": 0.4090806245803833, "learning_rate": 9.357826608800142e-05, "loss": 0.0312, "step": 4100 }, { "grad_norm": 0.31228333711624146, "learning_rate": 9.353767342631994e-05, "loss": 0.0295, "step": 4110 }, { "grad_norm": 0.3785773515701294, "learning_rate": 9.34969617343018e-05, "loss": 0.0264, "step": 4120 }, { "grad_norm": 0.35580289363861084, "learning_rate": 9.345613112325122e-05, "loss": 0.0269, "step": 4130 }, { "grad_norm": 0.3537365794181824, "learning_rate": 9.34151817047975e-05, "loss": 0.0286, "step": 4140 }, { "grad_norm": 0.33689165115356445, "learning_rate": 9.33741135908948e-05, "loss": 0.0278, "step": 4150 }, { "grad_norm": 0.3058110475540161, "learning_rate": 9.33329268938218e-05, "loss": 0.0243, "step": 4160 }, { "grad_norm": 0.41509923338890076, "learning_rate": 9.329162172618132e-05, "loss": 0.0297, "step": 4170 }, { "grad_norm": 0.2908613979816437, "learning_rate": 9.325019820090013e-05, "loss": 0.0303, "step": 4180 }, { "grad_norm": 0.2470216453075409, "learning_rate": 9.320865643122855e-05, "loss": 0.0308, "step": 4190 }, { "grad_norm": 0.39386704564094543, "learning_rate": 9.316699653074023e-05, "loss": 0.0257, "step": 4200 }, { "grad_norm": 0.39119914174079895, "learning_rate": 9.312521861333172e-05, "loss": 0.0252, "step": 4210 }, { "grad_norm": 0.3769645392894745, "learning_rate": 9.308332279322224e-05, "loss": 0.0286, "step": 4220 }, { "grad_norm": 0.3818323612213135, "learning_rate": 9.304130918495338e-05, "loss": 0.0256, "step": 4230 }, { "grad_norm": 0.3408069312572479, "learning_rate": 9.299917790338874e-05, "loss": 0.031, "step": 4240 }, { "grad_norm": 0.388755202293396, "learning_rate": 9.295692906371363e-05, "loss": 0.03, "step": 4250 }, { "grad_norm": 0.2997666895389557, "learning_rate": 9.291456278143476e-05, "loss": 0.0295, "step": 4260 }, { "grad_norm": 0.3474074900150299, "learning_rate": 9.287207917237994e-05, "loss": 0.0291, "step": 4270 }, { "grad_norm": 0.3130378723144531, "learning_rate": 9.282947835269773e-05, "loss": 0.0281, "step": 4280 }, { "grad_norm": 0.2597285509109497, "learning_rate": 9.278676043885715e-05, "loss": 0.0278, "step": 4290 }, { "grad_norm": 0.2979455292224884, "learning_rate": 9.274392554764733e-05, "loss": 0.0278, "step": 4300 }, { "grad_norm": 0.3137226700782776, "learning_rate": 9.270097379617723e-05, "loss": 0.024, "step": 4310 }, { "grad_norm": 0.2956026792526245, "learning_rate": 9.26579053018753e-05, "loss": 0.0276, "step": 4320 }, { "grad_norm": 0.33165881037712097, "learning_rate": 9.261472018248918e-05, "loss": 0.027, "step": 4330 }, { "grad_norm": 0.2637697160243988, "learning_rate": 9.25714185560853e-05, "loss": 0.027, "step": 4340 }, { "grad_norm": 0.32489845156669617, "learning_rate": 9.252800054104868e-05, "loss": 0.0254, "step": 4350 }, { "grad_norm": 0.3968600928783417, "learning_rate": 9.248446625608252e-05, "loss": 0.0247, "step": 4360 }, { "grad_norm": 0.32306909561157227, "learning_rate": 9.244081582020789e-05, "loss": 0.0277, "step": 4370 }, { "grad_norm": 0.3493836522102356, "learning_rate": 9.239704935276339e-05, "loss": 0.026, "step": 4380 }, { "grad_norm": 0.3256915211677551, "learning_rate": 9.235316697340489e-05, "loss": 0.0291, "step": 4390 }, { "grad_norm": 0.4577317237854004, "learning_rate": 9.230916880210512e-05, "loss": 0.0304, "step": 4400 }, { "grad_norm": 0.42831385135650635, "learning_rate": 9.226505495915342e-05, "loss": 0.0277, "step": 4410 }, { "grad_norm": 0.3322218060493469, "learning_rate": 9.222082556515536e-05, "loss": 0.0291, "step": 4420 }, { "grad_norm": 0.32897186279296875, "learning_rate": 9.217648074103242e-05, "loss": 0.028, "step": 4430 }, { "grad_norm": 0.3353818356990814, "learning_rate": 9.213202060802161e-05, "loss": 0.032, "step": 4440 }, { "grad_norm": 0.21227431297302246, "learning_rate": 9.208744528767528e-05, "loss": 0.023, "step": 4450 }, { "grad_norm": 0.32138609886169434, "learning_rate": 9.204275490186064e-05, "loss": 0.0267, "step": 4460 }, { "grad_norm": 0.30175480246543884, "learning_rate": 9.199794957275949e-05, "loss": 0.0241, "step": 4470 }, { "grad_norm": 0.2488914430141449, "learning_rate": 9.19530294228679e-05, "loss": 0.0231, "step": 4480 }, { "grad_norm": 0.2901773154735565, "learning_rate": 9.190799457499583e-05, "loss": 0.027, "step": 4490 }, { "grad_norm": 0.3169887363910675, "learning_rate": 9.186284515226686e-05, "loss": 0.0283, "step": 4500 }, { "grad_norm": 0.2995896339416504, "learning_rate": 9.181758127811777e-05, "loss": 0.0323, "step": 4510 }, { "grad_norm": 0.30276668071746826, "learning_rate": 9.177220307629825e-05, "loss": 0.0272, "step": 4520 }, { "grad_norm": 0.3941483497619629, "learning_rate": 9.172671067087059e-05, "loss": 0.0286, "step": 4530 }, { "grad_norm": 0.358067125082016, "learning_rate": 9.16811041862093e-05, "loss": 0.027, "step": 4540 }, { "grad_norm": 0.31476283073425293, "learning_rate": 9.163538374700076e-05, "loss": 0.0257, "step": 4550 }, { "grad_norm": 0.3502441644668579, "learning_rate": 9.158954947824287e-05, "loss": 0.0262, "step": 4560 }, { "grad_norm": 0.3656407594680786, "learning_rate": 9.154360150524482e-05, "loss": 0.0277, "step": 4570 }, { "grad_norm": 0.31957879662513733, "learning_rate": 9.14975399536266e-05, "loss": 0.0275, "step": 4580 }, { "grad_norm": 0.3076411485671997, "learning_rate": 9.14513649493187e-05, "loss": 0.027, "step": 4590 }, { "grad_norm": 0.42032983899116516, "learning_rate": 9.140507661856187e-05, "loss": 0.0266, "step": 4600 }, { "grad_norm": 0.3050939440727234, "learning_rate": 9.135867508790661e-05, "loss": 0.0285, "step": 4610 }, { "grad_norm": 0.2997128367424011, "learning_rate": 9.131216048421291e-05, "loss": 0.0323, "step": 4620 }, { "grad_norm": 0.3447135388851166, "learning_rate": 9.126553293464998e-05, "loss": 0.0288, "step": 4630 }, { "grad_norm": 0.344406396150589, "learning_rate": 9.121879256669572e-05, "loss": 0.0306, "step": 4640 }, { "grad_norm": 0.33332428336143494, "learning_rate": 9.117193950813652e-05, "loss": 0.0288, "step": 4650 }, { "grad_norm": 0.3839907646179199, "learning_rate": 9.112497388706685e-05, "loss": 0.0334, "step": 4660 }, { "grad_norm": 0.30797654390335083, "learning_rate": 9.10778958318889e-05, "loss": 0.0299, "step": 4670 }, { "grad_norm": 0.27200382947921753, "learning_rate": 9.103070547131232e-05, "loss": 0.0272, "step": 4680 }, { "grad_norm": 0.27666571736335754, "learning_rate": 9.098340293435375e-05, "loss": 0.0308, "step": 4690 }, { "grad_norm": 0.32322385907173157, "learning_rate": 9.093598835033649e-05, "loss": 0.0353, "step": 4700 }, { "grad_norm": 0.33282357454299927, "learning_rate": 9.088846184889021e-05, "loss": 0.0287, "step": 4710 }, { "grad_norm": 0.3156093657016754, "learning_rate": 9.084082355995057e-05, "loss": 0.0284, "step": 4720 }, { "grad_norm": 0.23986436426639557, "learning_rate": 9.079307361375882e-05, "loss": 0.0292, "step": 4730 }, { "grad_norm": 0.3906531035900116, "learning_rate": 9.074521214086149e-05, "loss": 0.0266, "step": 4740 }, { "grad_norm": 0.4088486433029175, "learning_rate": 9.069723927211001e-05, "loss": 0.0278, "step": 4750 }, { "grad_norm": 0.26060155034065247, "learning_rate": 9.064915513866037e-05, "loss": 0.0269, "step": 4760 }, { "grad_norm": 0.28228479623794556, "learning_rate": 9.060095987197279e-05, "loss": 0.0275, "step": 4770 }, { "grad_norm": 0.31312650442123413, "learning_rate": 9.055265360381126e-05, "loss": 0.0297, "step": 4780 }, { "grad_norm": 0.24215596914291382, "learning_rate": 9.050423646624326e-05, "loss": 0.0255, "step": 4790 }, { "grad_norm": 0.31612130999565125, "learning_rate": 9.045570859163943e-05, "loss": 0.0265, "step": 4800 }, { "grad_norm": 0.3255903422832489, "learning_rate": 9.04070701126731e-05, "loss": 0.0237, "step": 4810 }, { "grad_norm": 0.25042724609375, "learning_rate": 9.035832116232001e-05, "loss": 0.0226, "step": 4820 }, { "grad_norm": 0.4123761057853699, "learning_rate": 9.030946187385796e-05, "loss": 0.0262, "step": 4830 }, { "grad_norm": 0.2978915572166443, "learning_rate": 9.026049238086635e-05, "loss": 0.0341, "step": 4840 }, { "grad_norm": 0.26202860474586487, "learning_rate": 9.021141281722591e-05, "loss": 0.0258, "step": 4850 }, { "grad_norm": 0.29135826230049133, "learning_rate": 9.01622233171183e-05, "loss": 0.0259, "step": 4860 }, { "grad_norm": 0.3207217752933502, "learning_rate": 9.011292401502574e-05, "loss": 0.0258, "step": 4870 }, { "grad_norm": 0.2846567928791046, "learning_rate": 9.006351504573063e-05, "loss": 0.0318, "step": 4880 }, { "grad_norm": 0.29987066984176636, "learning_rate": 9.001399654431519e-05, "loss": 0.0274, "step": 4890 }, { "grad_norm": 0.43490856885910034, "learning_rate": 8.996436864616116e-05, "loss": 0.0271, "step": 4900 }, { "grad_norm": 0.2811910808086395, "learning_rate": 8.991463148694925e-05, "loss": 0.0255, "step": 4910 }, { "grad_norm": 0.3316333293914795, "learning_rate": 8.986478520265902e-05, "loss": 0.0268, "step": 4920 }, { "grad_norm": 0.2654367685317993, "learning_rate": 8.981482992956827e-05, "loss": 0.0272, "step": 4930 }, { "grad_norm": 0.3811144232749939, "learning_rate": 8.976476580425282e-05, "loss": 0.0256, "step": 4940 }, { "grad_norm": 0.2596016526222229, "learning_rate": 8.971459296358606e-05, "loss": 0.0254, "step": 4950 }, { "grad_norm": 0.31858256459236145, "learning_rate": 8.966431154473864e-05, "loss": 0.0238, "step": 4960 }, { "grad_norm": 0.30758175253868103, "learning_rate": 8.961392168517803e-05, "loss": 0.0318, "step": 4970 }, { "grad_norm": 0.2970453202724457, "learning_rate": 8.956342352266821e-05, "loss": 0.0309, "step": 4980 }, { "grad_norm": 0.29796522855758667, "learning_rate": 8.95128171952692e-05, "loss": 0.0243, "step": 4990 }, { "grad_norm": 0.3656720817089081, "learning_rate": 8.946210284133676e-05, "loss": 0.0257, "step": 5000 }, { "grad_norm": 0.4265839159488678, "learning_rate": 8.941128059952201e-05, "loss": 0.0285, "step": 5010 }, { "grad_norm": 0.3073383867740631, "learning_rate": 8.936035060877102e-05, "loss": 0.0268, "step": 5020 }, { "grad_norm": 0.24690712988376617, "learning_rate": 8.930931300832443e-05, "loss": 0.0244, "step": 5030 }, { "grad_norm": 0.2730584442615509, "learning_rate": 8.925816793771711e-05, "loss": 0.0254, "step": 5040 }, { "grad_norm": 0.33769068121910095, "learning_rate": 8.92069155367777e-05, "loss": 0.0267, "step": 5050 }, { "grad_norm": 0.3440764546394348, "learning_rate": 8.915555594562834e-05, "loss": 0.0267, "step": 5060 }, { "grad_norm": 0.2614094614982605, "learning_rate": 8.910408930468416e-05, "loss": 0.023, "step": 5070 }, { "grad_norm": 0.38997259736061096, "learning_rate": 8.905251575465303e-05, "loss": 0.0257, "step": 5080 }, { "grad_norm": 0.29437172412872314, "learning_rate": 8.900083543653502e-05, "loss": 0.0271, "step": 5090 }, { "grad_norm": 0.31064164638519287, "learning_rate": 8.894904849162218e-05, "loss": 0.0294, "step": 5100 }, { "grad_norm": 0.28452068567276, "learning_rate": 8.889715506149802e-05, "loss": 0.0252, "step": 5110 }, { "grad_norm": 0.33859190344810486, "learning_rate": 8.884515528803722e-05, "loss": 0.0329, "step": 5120 }, { "grad_norm": 0.3121359646320343, "learning_rate": 8.879304931340517e-05, "loss": 0.0268, "step": 5130 }, { "grad_norm": 0.3281187117099762, "learning_rate": 8.874083728005759e-05, "loss": 0.0304, "step": 5140 }, { "grad_norm": 0.3623501658439636, "learning_rate": 8.868851933074021e-05, "loss": 0.0273, "step": 5150 }, { "grad_norm": 0.29177701473236084, "learning_rate": 8.863609560848829e-05, "loss": 0.0283, "step": 5160 }, { "grad_norm": 0.3232070803642273, "learning_rate": 8.85835662566263e-05, "loss": 0.0239, "step": 5170 }, { "grad_norm": 0.4436071217060089, "learning_rate": 8.853093141876747e-05, "loss": 0.0266, "step": 5180 }, { "grad_norm": 0.2816045582294464, "learning_rate": 8.847819123881343e-05, "loss": 0.0278, "step": 5190 }, { "grad_norm": 0.31329259276390076, "learning_rate": 8.842534586095383e-05, "loss": 0.031, "step": 5200 }, { "grad_norm": 0.21701830625534058, "learning_rate": 8.837239542966593e-05, "loss": 0.0256, "step": 5210 }, { "grad_norm": 0.3068200945854187, "learning_rate": 8.831934008971417e-05, "loss": 0.0253, "step": 5220 }, { "grad_norm": 0.22425350546836853, "learning_rate": 8.826617998614982e-05, "loss": 0.0246, "step": 5230 }, { "grad_norm": 0.2649974822998047, "learning_rate": 8.821291526431056e-05, "loss": 0.0228, "step": 5240 }, { "grad_norm": 0.25373518466949463, "learning_rate": 8.815954606982015e-05, "loss": 0.0223, "step": 5250 }, { "grad_norm": 0.2015586644411087, "learning_rate": 8.810607254858789e-05, "loss": 0.0263, "step": 5260 }, { "grad_norm": 0.22784429788589478, "learning_rate": 8.805249484680838e-05, "loss": 0.0271, "step": 5270 }, { "grad_norm": 0.32793480157852173, "learning_rate": 8.799881311096096e-05, "loss": 0.0258, "step": 5280 }, { "grad_norm": 0.27490416169166565, "learning_rate": 8.794502748780949e-05, "loss": 0.0273, "step": 5290 }, { "grad_norm": 0.26882126927375793, "learning_rate": 8.78911381244018e-05, "loss": 0.0246, "step": 5300 }, { "grad_norm": 0.31763744354248047, "learning_rate": 8.783714516806933e-05, "loss": 0.0227, "step": 5310 }, { "grad_norm": 0.31146928668022156, "learning_rate": 8.77830487664268e-05, "loss": 0.0256, "step": 5320 }, { "grad_norm": 0.3122774660587311, "learning_rate": 8.772884906737167e-05, "loss": 0.0247, "step": 5330 }, { "grad_norm": 0.3552822768688202, "learning_rate": 8.767454621908387e-05, "loss": 0.0272, "step": 5340 }, { "grad_norm": 0.3332209587097168, "learning_rate": 8.76201403700253e-05, "loss": 0.0269, "step": 5350 }, { "grad_norm": 0.36409991979599, "learning_rate": 8.756563166893949e-05, "loss": 0.0285, "step": 5360 }, { "grad_norm": 0.41074737906455994, "learning_rate": 8.751102026485113e-05, "loss": 0.0319, "step": 5370 }, { "grad_norm": 0.29399991035461426, "learning_rate": 8.745630630706571e-05, "loss": 0.0285, "step": 5380 }, { "grad_norm": 0.23351778090000153, "learning_rate": 8.740148994516912e-05, "loss": 0.0246, "step": 5390 }, { "grad_norm": 0.274258553981781, "learning_rate": 8.73465713290272e-05, "loss": 0.0251, "step": 5400 }, { "grad_norm": 0.229307159781456, "learning_rate": 8.729155060878533e-05, "loss": 0.0255, "step": 5410 }, { "grad_norm": 0.3665757179260254, "learning_rate": 8.723642793486809e-05, "loss": 0.0238, "step": 5420 }, { "grad_norm": 0.3162052631378174, "learning_rate": 8.718120345797873e-05, "loss": 0.0238, "step": 5430 }, { "grad_norm": 0.31186363101005554, "learning_rate": 8.712587732909889e-05, "loss": 0.0236, "step": 5440 }, { "grad_norm": 0.4067315459251404, "learning_rate": 8.707044969948806e-05, "loss": 0.0282, "step": 5450 }, { "grad_norm": 0.23620837926864624, "learning_rate": 8.701492072068329e-05, "loss": 0.0239, "step": 5460 }, { "grad_norm": 0.3104446232318878, "learning_rate": 8.695929054449869e-05, "loss": 0.0252, "step": 5470 }, { "grad_norm": 0.32363754510879517, "learning_rate": 8.690355932302501e-05, "loss": 0.0235, "step": 5480 }, { "grad_norm": 0.33881083130836487, "learning_rate": 8.684772720862931e-05, "loss": 0.0244, "step": 5490 }, { "grad_norm": 0.27435627579689026, "learning_rate": 8.679179435395446e-05, "loss": 0.0259, "step": 5500 }, { "grad_norm": 0.18504434823989868, "learning_rate": 8.673576091191874e-05, "loss": 0.0235, "step": 5510 }, { "grad_norm": 0.23039261996746063, "learning_rate": 8.667962703571541e-05, "loss": 0.021, "step": 5520 }, { "grad_norm": 0.2918706238269806, "learning_rate": 8.662339287881238e-05, "loss": 0.0301, "step": 5530 }, { "grad_norm": 0.2959015369415283, "learning_rate": 8.656705859495169e-05, "loss": 0.024, "step": 5540 }, { "grad_norm": 0.23193950951099396, "learning_rate": 8.651062433814912e-05, "loss": 0.0238, "step": 5550 }, { "grad_norm": 0.28355249762535095, "learning_rate": 8.645409026269375e-05, "loss": 0.0242, "step": 5560 }, { "grad_norm": 0.30446746945381165, "learning_rate": 8.639745652314759e-05, "loss": 0.0236, "step": 5570 }, { "grad_norm": 0.29505982995033264, "learning_rate": 8.634072327434515e-05, "loss": 0.0248, "step": 5580 }, { "grad_norm": 0.2878054678440094, "learning_rate": 8.628389067139294e-05, "loss": 0.0235, "step": 5590 }, { "grad_norm": 0.3128759264945984, "learning_rate": 8.622695886966911e-05, "loss": 0.0276, "step": 5600 }, { "grad_norm": 0.3287160098552704, "learning_rate": 8.616992802482308e-05, "loss": 0.0262, "step": 5610 }, { "grad_norm": 0.2866574823856354, "learning_rate": 8.611279829277496e-05, "loss": 0.0252, "step": 5620 }, { "grad_norm": 0.2551356852054596, "learning_rate": 8.605556982971528e-05, "loss": 0.022, "step": 5630 }, { "grad_norm": 0.34121188521385193, "learning_rate": 8.599824279210447e-05, "loss": 0.0285, "step": 5640 }, { "grad_norm": 0.28182530403137207, "learning_rate": 8.594081733667243e-05, "loss": 0.0223, "step": 5650 }, { "grad_norm": 0.3255702555179596, "learning_rate": 8.58832936204182e-05, "loss": 0.0276, "step": 5660 }, { "grad_norm": 0.2512299120426178, "learning_rate": 8.582567180060942e-05, "loss": 0.0264, "step": 5670 }, { "grad_norm": 0.2800491750240326, "learning_rate": 8.576795203478194e-05, "loss": 0.0248, "step": 5680 }, { "grad_norm": 0.30090585350990295, "learning_rate": 8.571013448073939e-05, "loss": 0.0274, "step": 5690 }, { "grad_norm": 0.2180919349193573, "learning_rate": 8.565221929655275e-05, "loss": 0.0207, "step": 5700 }, { "grad_norm": 0.2863413393497467, "learning_rate": 8.559420664055992e-05, "loss": 0.0272, "step": 5710 }, { "grad_norm": 0.308532178401947, "learning_rate": 8.553609667136532e-05, "loss": 0.0302, "step": 5720 }, { "grad_norm": 0.2914881706237793, "learning_rate": 8.547788954783936e-05, "loss": 0.0262, "step": 5730 }, { "grad_norm": 0.25778794288635254, "learning_rate": 8.541958542911808e-05, "loss": 0.0235, "step": 5740 }, { "grad_norm": 0.247626394033432, "learning_rate": 8.536118447460275e-05, "loss": 0.0231, "step": 5750 }, { "grad_norm": 0.2599932551383972, "learning_rate": 8.530268684395932e-05, "loss": 0.0228, "step": 5760 }, { "grad_norm": 0.2675947844982147, "learning_rate": 8.524409269711807e-05, "loss": 0.025, "step": 5770 }, { "grad_norm": 0.3197520971298218, "learning_rate": 8.51854021942732e-05, "loss": 0.0261, "step": 5780 }, { "grad_norm": 0.23655539751052856, "learning_rate": 8.512661549588227e-05, "loss": 0.0239, "step": 5790 }, { "grad_norm": 0.265231192111969, "learning_rate": 8.506773276266588e-05, "loss": 0.0227, "step": 5800 }, { "grad_norm": 0.23522460460662842, "learning_rate": 8.500875415560721e-05, "loss": 0.0223, "step": 5810 }, { "grad_norm": 0.26978084444999695, "learning_rate": 8.494967983595144e-05, "loss": 0.0222, "step": 5820 }, { "grad_norm": 0.3708859980106354, "learning_rate": 8.489050996520558e-05, "loss": 0.0277, "step": 5830 }, { "grad_norm": 0.3048735558986664, "learning_rate": 8.483124470513775e-05, "loss": 0.0242, "step": 5840 }, { "grad_norm": 0.28182023763656616, "learning_rate": 8.477188421777692e-05, "loss": 0.0299, "step": 5850 }, { "grad_norm": 0.3493616580963135, "learning_rate": 8.47124286654124e-05, "loss": 0.0253, "step": 5860 }, { "grad_norm": 0.25177037715911865, "learning_rate": 8.465287821059341e-05, "loss": 0.0244, "step": 5870 }, { "grad_norm": 0.31861603260040283, "learning_rate": 8.45932330161286e-05, "loss": 0.0345, "step": 5880 }, { "grad_norm": 0.3486482501029968, "learning_rate": 8.453349324508567e-05, "loss": 0.023, "step": 5890 }, { "grad_norm": 0.4371389150619507, "learning_rate": 8.447365906079088e-05, "loss": 0.0287, "step": 5900 }, { "grad_norm": 0.46902981400489807, "learning_rate": 8.441373062682856e-05, "loss": 0.0258, "step": 5910 }, { "grad_norm": 0.31708288192749023, "learning_rate": 8.43537081070408e-05, "loss": 0.0271, "step": 5920 }, { "grad_norm": 0.33538079261779785, "learning_rate": 8.429359166552689e-05, "loss": 0.0238, "step": 5930 }, { "grad_norm": 0.2586621940135956, "learning_rate": 8.423338146664284e-05, "loss": 0.0248, "step": 5940 }, { "grad_norm": 0.28880825638771057, "learning_rate": 8.417307767500107e-05, "loss": 0.0228, "step": 5950 }, { "grad_norm": 0.3559423089027405, "learning_rate": 8.411268045546983e-05, "loss": 0.0226, "step": 5960 }, { "grad_norm": 0.28171002864837646, "learning_rate": 8.405218997317281e-05, "loss": 0.0216, "step": 5970 }, { "grad_norm": 0.3660081923007965, "learning_rate": 8.399160639348869e-05, "loss": 0.0239, "step": 5980 }, { "grad_norm": 0.39471328258514404, "learning_rate": 8.393092988205065e-05, "loss": 0.0222, "step": 5990 }, { "grad_norm": 0.22231453657150269, "learning_rate": 8.387016060474597e-05, "loss": 0.0228, "step": 6000 }, { "grad_norm": 0.3263501822948456, "learning_rate": 8.380929872771551e-05, "loss": 0.0209, "step": 6010 }, { "grad_norm": 0.2685607373714447, "learning_rate": 8.374834441735335e-05, "loss": 0.0232, "step": 6020 }, { "grad_norm": 0.3161412179470062, "learning_rate": 8.368729784030622e-05, "loss": 0.0213, "step": 6030 }, { "grad_norm": 0.3496015667915344, "learning_rate": 8.362615916347315e-05, "loss": 0.023, "step": 6040 }, { "grad_norm": 0.1971009075641632, "learning_rate": 8.356492855400493e-05, "loss": 0.0197, "step": 6050 }, { "grad_norm": 0.25681424140930176, "learning_rate": 8.350360617930371e-05, "loss": 0.0198, "step": 6060 }, { "grad_norm": 0.2757159173488617, "learning_rate": 8.344219220702255e-05, "loss": 0.0191, "step": 6070 }, { "grad_norm": 0.1987524777650833, "learning_rate": 8.338068680506485e-05, "loss": 0.0233, "step": 6080 }, { "grad_norm": 0.2464054971933365, "learning_rate": 8.33190901415841e-05, "loss": 0.0226, "step": 6090 }, { "grad_norm": 0.2086532711982727, "learning_rate": 8.325740238498317e-05, "loss": 0.0202, "step": 6100 }, { "grad_norm": 0.23794694244861603, "learning_rate": 8.319562370391406e-05, "loss": 0.0186, "step": 6110 }, { "grad_norm": 0.39384374022483826, "learning_rate": 8.31337542672773e-05, "loss": 0.022, "step": 6120 }, { "grad_norm": 0.25566935539245605, "learning_rate": 8.307179424422158e-05, "loss": 0.0227, "step": 6130 }, { "grad_norm": 0.2544749975204468, "learning_rate": 8.300974380414327e-05, "loss": 0.0185, "step": 6140 }, { "grad_norm": 0.30167677998542786, "learning_rate": 8.294760311668586e-05, "loss": 0.0194, "step": 6150 }, { "grad_norm": 0.2678512632846832, "learning_rate": 8.288537235173961e-05, "loss": 0.0235, "step": 6160 }, { "grad_norm": 0.3272489011287689, "learning_rate": 8.282305167944108e-05, "loss": 0.0237, "step": 6170 }, { "grad_norm": 0.2840566337108612, "learning_rate": 8.276064127017262e-05, "loss": 0.0227, "step": 6180 }, { "grad_norm": 0.2667955160140991, "learning_rate": 8.269814129456189e-05, "loss": 0.0226, "step": 6190 }, { "grad_norm": 0.28034839034080505, "learning_rate": 8.263555192348143e-05, "loss": 0.0241, "step": 6200 }, { "grad_norm": 0.37277883291244507, "learning_rate": 8.257287332804819e-05, "loss": 0.0249, "step": 6210 }, { "grad_norm": 0.2999477982521057, "learning_rate": 8.251010567962307e-05, "loss": 0.0236, "step": 6220 }, { "grad_norm": 0.20026935636997223, "learning_rate": 8.244724914981041e-05, "loss": 0.0219, "step": 6230 }, { "grad_norm": 0.23591923713684082, "learning_rate": 8.238430391045757e-05, "loss": 0.0196, "step": 6240 }, { "grad_norm": 0.2349013388156891, "learning_rate": 8.232127013365445e-05, "loss": 0.0235, "step": 6250 }, { "grad_norm": 0.27016481757164, "learning_rate": 8.225814799173295e-05, "loss": 0.0234, "step": 6260 }, { "grad_norm": 0.21662107110023499, "learning_rate": 8.219493765726663e-05, "loss": 0.0199, "step": 6270 }, { "grad_norm": 0.28961411118507385, "learning_rate": 8.21316393030701e-05, "loss": 0.0216, "step": 6280 }, { "grad_norm": 0.2765747010707855, "learning_rate": 8.206825310219865e-05, "loss": 0.0241, "step": 6290 }, { "grad_norm": 0.32264697551727295, "learning_rate": 8.200477922794776e-05, "loss": 0.0206, "step": 6300 }, { "grad_norm": 0.2277567982673645, "learning_rate": 8.194121785385256e-05, "loss": 0.0243, "step": 6310 }, { "grad_norm": 0.2591727674007416, "learning_rate": 8.187756915368741e-05, "loss": 0.0225, "step": 6320 }, { "grad_norm": 0.43888673186302185, "learning_rate": 8.181383330146544e-05, "loss": 0.0269, "step": 6330 }, { "grad_norm": 0.2814996540546417, "learning_rate": 8.175001047143804e-05, "loss": 0.0264, "step": 6340 }, { "grad_norm": 0.26310423016548157, "learning_rate": 8.168610083809438e-05, "loss": 0.023, "step": 6350 }, { "grad_norm": 0.22105643153190613, "learning_rate": 8.162210457616095e-05, "loss": 0.0225, "step": 6360 }, { "grad_norm": 0.2570667266845703, "learning_rate": 8.155802186060109e-05, "loss": 0.0259, "step": 6370 }, { "grad_norm": 0.19747498631477356, "learning_rate": 8.149385286661453e-05, "loss": 0.0233, "step": 6380 }, { "grad_norm": 0.2886122465133667, "learning_rate": 8.14295977696368e-05, "loss": 0.0237, "step": 6390 }, { "grad_norm": 0.21951311826705933, "learning_rate": 8.13652567453389e-05, "loss": 0.0214, "step": 6400 }, { "grad_norm": 0.25501635670661926, "learning_rate": 8.130082996962676e-05, "loss": 0.0235, "step": 6410 }, { "grad_norm": 0.3186947703361511, "learning_rate": 8.123631761864068e-05, "loss": 0.0227, "step": 6420 }, { "grad_norm": 0.33610808849334717, "learning_rate": 8.1171719868755e-05, "loss": 0.0207, "step": 6430 }, { "grad_norm": 0.19524377584457397, "learning_rate": 8.110703689657748e-05, "loss": 0.0226, "step": 6440 }, { "grad_norm": 0.3013477027416229, "learning_rate": 8.104226887894892e-05, "loss": 0.0205, "step": 6450 }, { "grad_norm": 0.22579389810562134, "learning_rate": 8.097741599294257e-05, "loss": 0.0201, "step": 6460 }, { "grad_norm": 0.253431499004364, "learning_rate": 8.091247841586378e-05, "loss": 0.0227, "step": 6470 }, { "grad_norm": 0.3011651635169983, "learning_rate": 8.084745632524939e-05, "loss": 0.0204, "step": 6480 }, { "grad_norm": 0.22242426872253418, "learning_rate": 8.07823498988673e-05, "loss": 0.0208, "step": 6490 }, { "grad_norm": 0.26831814646720886, "learning_rate": 8.071715931471602e-05, "loss": 0.0182, "step": 6500 }, { "grad_norm": 0.23434121906757355, "learning_rate": 8.06518847510241e-05, "loss": 0.0209, "step": 6510 }, { "grad_norm": 0.3345057964324951, "learning_rate": 8.058652638624971e-05, "loss": 0.0234, "step": 6520 }, { "grad_norm": 0.2448362410068512, "learning_rate": 8.052108439908013e-05, "loss": 0.0215, "step": 6530 }, { "grad_norm": 0.20532064139842987, "learning_rate": 8.045555896843125e-05, "loss": 0.022, "step": 6540 }, { "grad_norm": 0.22480297088623047, "learning_rate": 8.03899502734471e-05, "loss": 0.0207, "step": 6550 }, { "grad_norm": 0.33434534072875977, "learning_rate": 8.032425849349931e-05, "loss": 0.021, "step": 6560 }, { "grad_norm": 0.33099380135536194, "learning_rate": 8.025848380818674e-05, "loss": 0.0249, "step": 6570 }, { "grad_norm": 0.31918153166770935, "learning_rate": 8.019262639733487e-05, "loss": 0.0303, "step": 6580 }, { "grad_norm": 0.31886738538742065, "learning_rate": 8.012668644099531e-05, "loss": 0.0245, "step": 6590 }, { "grad_norm": 0.2409660816192627, "learning_rate": 8.006066411944542e-05, "loss": 0.0251, "step": 6600 }, { "grad_norm": 0.23869642615318298, "learning_rate": 7.999455961318769e-05, "loss": 0.0206, "step": 6610 }, { "grad_norm": 0.21888616681098938, "learning_rate": 7.992837310294932e-05, "loss": 0.0227, "step": 6620 }, { "grad_norm": 0.257647842168808, "learning_rate": 7.986210476968167e-05, "loss": 0.0242, "step": 6630 }, { "grad_norm": 0.3338513970375061, "learning_rate": 7.97957547945599e-05, "loss": 0.0223, "step": 6640 }, { "grad_norm": 0.26118963956832886, "learning_rate": 7.972932335898226e-05, "loss": 0.0201, "step": 6650 }, { "grad_norm": 0.29803118109703064, "learning_rate": 7.966281064456975e-05, "loss": 0.0218, "step": 6660 }, { "grad_norm": 0.23741863667964935, "learning_rate": 7.959621683316563e-05, "loss": 0.0239, "step": 6670 }, { "grad_norm": 0.2897915244102478, "learning_rate": 7.952954210683481e-05, "loss": 0.0237, "step": 6680 }, { "grad_norm": 0.23337016999721527, "learning_rate": 7.946278664786345e-05, "loss": 0.0205, "step": 6690 }, { "grad_norm": 0.36339133977890015, "learning_rate": 7.939595063875842e-05, "loss": 0.0203, "step": 6700 }, { "grad_norm": 0.22920475900173187, "learning_rate": 7.932903426224683e-05, "loss": 0.0212, "step": 6710 }, { "grad_norm": 0.30156245827674866, "learning_rate": 7.926203770127552e-05, "loss": 0.0214, "step": 6720 }, { "grad_norm": 0.4140009880065918, "learning_rate": 7.919496113901046e-05, "loss": 0.0223, "step": 6730 }, { "grad_norm": 0.2548609673976898, "learning_rate": 7.912780475883649e-05, "loss": 0.0219, "step": 6740 }, { "grad_norm": 0.18304933607578278, "learning_rate": 7.906056874435652e-05, "loss": 0.0193, "step": 6750 }, { "grad_norm": 0.2687060832977295, "learning_rate": 7.899325327939131e-05, "loss": 0.0182, "step": 6760 }, { "grad_norm": 0.21509476006031036, "learning_rate": 7.892585854797872e-05, "loss": 0.0171, "step": 6770 }, { "grad_norm": 0.34525665640830994, "learning_rate": 7.88583847343734e-05, "loss": 0.0197, "step": 6780 }, { "grad_norm": 0.2873477041721344, "learning_rate": 7.879083202304616e-05, "loss": 0.0211, "step": 6790 }, { "grad_norm": 0.3235916495323181, "learning_rate": 7.872320059868355e-05, "loss": 0.0256, "step": 6800 }, { "grad_norm": 0.32050007581710815, "learning_rate": 7.865549064618729e-05, "loss": 0.0199, "step": 6810 }, { "grad_norm": 0.23479898273944855, "learning_rate": 7.858770235067381e-05, "loss": 0.0202, "step": 6820 }, { "grad_norm": 0.24131444096565247, "learning_rate": 7.851983589747374e-05, "loss": 0.0184, "step": 6830 }, { "grad_norm": 0.21565231680870056, "learning_rate": 7.845189147213133e-05, "loss": 0.0198, "step": 6840 }, { "grad_norm": 0.263662725687027, "learning_rate": 7.838386926040407e-05, "loss": 0.0183, "step": 6850 }, { "grad_norm": 0.2966669797897339, "learning_rate": 7.83157694482621e-05, "loss": 0.0242, "step": 6860 }, { "grad_norm": 0.2648978531360626, "learning_rate": 7.824759222188768e-05, "loss": 0.0207, "step": 6870 }, { "grad_norm": 0.3747299611568451, "learning_rate": 7.817933776767478e-05, "loss": 0.0237, "step": 6880 }, { "grad_norm": 0.22540153563022614, "learning_rate": 7.811100627222842e-05, "loss": 0.0203, "step": 6890 }, { "grad_norm": 0.25776785612106323, "learning_rate": 7.804259792236435e-05, "loss": 0.0202, "step": 6900 }, { "grad_norm": 0.28414055705070496, "learning_rate": 7.797411290510835e-05, "loss": 0.0217, "step": 6910 }, { "grad_norm": 0.27722951769828796, "learning_rate": 7.790555140769586e-05, "loss": 0.0259, "step": 6920 }, { "grad_norm": 0.251611590385437, "learning_rate": 7.78369136175714e-05, "loss": 0.0183, "step": 6930 }, { "grad_norm": 0.26650911569595337, "learning_rate": 7.776819972238806e-05, "loss": 0.0183, "step": 6940 }, { "grad_norm": 0.24998006224632263, "learning_rate": 7.7699409910007e-05, "loss": 0.0214, "step": 6950 }, { "grad_norm": 0.2783333361148834, "learning_rate": 7.763054436849694e-05, "loss": 0.0209, "step": 6960 }, { "grad_norm": 0.2888818383216858, "learning_rate": 7.756160328613364e-05, "loss": 0.02, "step": 6970 }, { "grad_norm": 0.2426069974899292, "learning_rate": 7.749258685139942e-05, "loss": 0.0209, "step": 6980 }, { "grad_norm": 0.2351798713207245, "learning_rate": 7.742349525298253e-05, "loss": 0.0233, "step": 6990 }, { "grad_norm": 0.25346532464027405, "learning_rate": 7.735432867977679e-05, "loss": 0.0182, "step": 7000 }, { "grad_norm": 0.33036619424819946, "learning_rate": 7.728508732088096e-05, "loss": 0.0203, "step": 7010 }, { "grad_norm": 0.23617826402187347, "learning_rate": 7.721577136559825e-05, "loss": 0.0173, "step": 7020 }, { "grad_norm": 0.23685304820537567, "learning_rate": 7.714638100343588e-05, "loss": 0.0189, "step": 7030 }, { "grad_norm": 0.23842352628707886, "learning_rate": 7.707691642410444e-05, "loss": 0.0215, "step": 7040 }, { "grad_norm": 0.32530656456947327, "learning_rate": 7.70073778175174e-05, "loss": 0.0283, "step": 7050 }, { "grad_norm": 0.21224761009216309, "learning_rate": 7.69377653737907e-05, "loss": 0.021, "step": 7060 }, { "grad_norm": 0.3506028354167938, "learning_rate": 7.686807928324209e-05, "loss": 0.0226, "step": 7070 }, { "grad_norm": 0.2586401104927063, "learning_rate": 7.679831973639065e-05, "loss": 0.0195, "step": 7080 }, { "grad_norm": 0.20961783826351166, "learning_rate": 7.672848692395637e-05, "loss": 0.0186, "step": 7090 }, { "grad_norm": 0.2928829789161682, "learning_rate": 7.665858103685944e-05, "loss": 0.0194, "step": 7100 }, { "grad_norm": 0.25775185227394104, "learning_rate": 7.658860226621991e-05, "loss": 0.0206, "step": 7110 }, { "grad_norm": 0.2678923010826111, "learning_rate": 7.651855080335708e-05, "loss": 0.0218, "step": 7120 }, { "grad_norm": 0.29016000032424927, "learning_rate": 7.644842683978896e-05, "loss": 0.0194, "step": 7130 }, { "grad_norm": 0.21846239268779755, "learning_rate": 7.63782305672318e-05, "loss": 0.0219, "step": 7140 }, { "grad_norm": 0.2311229258775711, "learning_rate": 7.63079621775995e-05, "loss": 0.0236, "step": 7150 }, { "grad_norm": 0.31041696667671204, "learning_rate": 7.623762186300319e-05, "loss": 0.023, "step": 7160 }, { "grad_norm": 0.23054033517837524, "learning_rate": 7.616720981575057e-05, "loss": 0.0198, "step": 7170 }, { "grad_norm": 0.20487456023693085, "learning_rate": 7.609672622834552e-05, "loss": 0.0196, "step": 7180 }, { "grad_norm": 0.2788337469100952, "learning_rate": 7.602617129348747e-05, "loss": 0.0252, "step": 7190 }, { "grad_norm": 0.21366001665592194, "learning_rate": 7.595554520407088e-05, "loss": 0.0204, "step": 7200 }, { "grad_norm": 0.24048368632793427, "learning_rate": 7.588484815318484e-05, "loss": 0.0178, "step": 7210 }, { "grad_norm": 0.2383061647415161, "learning_rate": 7.581408033411234e-05, "loss": 0.0221, "step": 7220 }, { "grad_norm": 0.1990555375814438, "learning_rate": 7.574324194032995e-05, "loss": 0.0163, "step": 7230 }, { "grad_norm": 0.24994482100009918, "learning_rate": 7.567233316550705e-05, "loss": 0.0188, "step": 7240 }, { "grad_norm": 0.21251073479652405, "learning_rate": 7.560135420350562e-05, "loss": 0.0159, "step": 7250 }, { "grad_norm": 0.2570096552371979, "learning_rate": 7.553030524837935e-05, "loss": 0.0183, "step": 7260 }, { "grad_norm": 0.25550898909568787, "learning_rate": 7.545918649437341e-05, "loss": 0.0185, "step": 7270 }, { "grad_norm": 0.2440783530473709, "learning_rate": 7.538799813592377e-05, "loss": 0.0189, "step": 7280 }, { "grad_norm": 0.2735527753829956, "learning_rate": 7.531674036765662e-05, "loss": 0.019, "step": 7290 }, { "grad_norm": 0.2896132469177246, "learning_rate": 7.524541338438807e-05, "loss": 0.0192, "step": 7300 }, { "grad_norm": 0.23929251730442047, "learning_rate": 7.517401738112328e-05, "loss": 0.0239, "step": 7310 }, { "grad_norm": 0.3045233190059662, "learning_rate": 7.510255255305628e-05, "loss": 0.0206, "step": 7320 }, { "grad_norm": 0.2966775894165039, "learning_rate": 7.503101909556911e-05, "loss": 0.0237, "step": 7330 }, { "grad_norm": 0.3075738549232483, "learning_rate": 7.495941720423154e-05, "loss": 0.0205, "step": 7340 }, { "grad_norm": 0.2548248767852783, "learning_rate": 7.488774707480042e-05, "loss": 0.0198, "step": 7350 }, { "grad_norm": 0.2919084429740906, "learning_rate": 7.481600890321911e-05, "loss": 0.0188, "step": 7360 }, { "grad_norm": 0.20247814059257507, "learning_rate": 7.474420288561708e-05, "loss": 0.0178, "step": 7370 }, { "grad_norm": 0.23137032985687256, "learning_rate": 7.467232921830921e-05, "loss": 0.0187, "step": 7380 }, { "grad_norm": 0.24471186101436615, "learning_rate": 7.460038809779537e-05, "loss": 0.0243, "step": 7390 }, { "grad_norm": 0.24615512788295746, "learning_rate": 7.452837972075983e-05, "loss": 0.0187, "step": 7400 }, { "grad_norm": 0.22761079668998718, "learning_rate": 7.445630428407074e-05, "loss": 0.017, "step": 7410 }, { "grad_norm": 0.2261689305305481, "learning_rate": 7.43841619847796e-05, "loss": 0.0207, "step": 7420 }, { "grad_norm": 0.25447002053260803, "learning_rate": 7.431195302012072e-05, "loss": 0.0195, "step": 7430 }, { "grad_norm": 0.1927580088376999, "learning_rate": 7.423967758751061e-05, "loss": 0.0203, "step": 7440 }, { "grad_norm": 0.19569790363311768, "learning_rate": 7.416733588454758e-05, "loss": 0.017, "step": 7450 }, { "grad_norm": 0.24117732048034668, "learning_rate": 7.409492810901106e-05, "loss": 0.0197, "step": 7460 }, { "grad_norm": 0.2299519032239914, "learning_rate": 7.402245445886116e-05, "loss": 0.0194, "step": 7470 }, { "grad_norm": 0.29907312989234924, "learning_rate": 7.394991513223806e-05, "loss": 0.0218, "step": 7480 }, { "grad_norm": 0.2274298071861267, "learning_rate": 7.38773103274615e-05, "loss": 0.0212, "step": 7490 }, { "grad_norm": 0.3710629343986511, "learning_rate": 7.380464024303028e-05, "loss": 0.0195, "step": 7500 }, { "grad_norm": 0.3477567136287689, "learning_rate": 7.373190507762162e-05, "loss": 0.0162, "step": 7510 }, { "grad_norm": 0.23854491114616394, "learning_rate": 7.365910503009066e-05, "loss": 0.021, "step": 7520 }, { "grad_norm": 0.18822142481803894, "learning_rate": 7.358624029946996e-05, "loss": 0.0208, "step": 7530 }, { "grad_norm": 0.29330742359161377, "learning_rate": 7.351331108496893e-05, "loss": 0.0188, "step": 7540 }, { "grad_norm": 0.19375944137573242, "learning_rate": 7.344031758597325e-05, "loss": 0.0171, "step": 7550 }, { "grad_norm": 0.2597653269767761, "learning_rate": 7.336726000204435e-05, "loss": 0.0167, "step": 7560 }, { "grad_norm": 0.273719847202301, "learning_rate": 7.32941385329189e-05, "loss": 0.0209, "step": 7570 }, { "grad_norm": 0.2614143192768097, "learning_rate": 7.322095337850816e-05, "loss": 0.0198, "step": 7580 }, { "grad_norm": 0.27194517850875854, "learning_rate": 7.314770473889758e-05, "loss": 0.0191, "step": 7590 }, { "grad_norm": 0.24784107506275177, "learning_rate": 7.307439281434615e-05, "loss": 0.02, "step": 7600 }, { "grad_norm": 0.29697391390800476, "learning_rate": 7.300101780528585e-05, "loss": 0.0186, "step": 7610 }, { "grad_norm": 0.2820657789707184, "learning_rate": 7.292757991232117e-05, "loss": 0.0168, "step": 7620 }, { "grad_norm": 0.2498627007007599, "learning_rate": 7.285407933622848e-05, "loss": 0.0172, "step": 7630 }, { "grad_norm": 0.3507038652896881, "learning_rate": 7.278051627795557e-05, "loss": 0.0187, "step": 7640 }, { "grad_norm": 0.23955945670604706, "learning_rate": 7.270689093862105e-05, "loss": 0.0198, "step": 7650 }, { "grad_norm": 0.21441836655139923, "learning_rate": 7.263320351951374e-05, "loss": 0.0176, "step": 7660 }, { "grad_norm": 0.2514626085758209, "learning_rate": 7.255945422209227e-05, "loss": 0.0208, "step": 7670 }, { "grad_norm": 0.5019840002059937, "learning_rate": 7.248564324798437e-05, "loss": 0.018, "step": 7680 }, { "grad_norm": 0.23659074306488037, "learning_rate": 7.241177079898644e-05, "loss": 0.0207, "step": 7690 }, { "grad_norm": 0.20750008523464203, "learning_rate": 7.233783707706295e-05, "loss": 0.0196, "step": 7700 }, { "grad_norm": 0.26073890924453735, "learning_rate": 7.226384228434586e-05, "loss": 0.0203, "step": 7710 }, { "grad_norm": 0.28458890318870544, "learning_rate": 7.21897866231341e-05, "loss": 0.0182, "step": 7720 }, { "grad_norm": 0.24336105585098267, "learning_rate": 7.211567029589303e-05, "loss": 0.0196, "step": 7730 }, { "grad_norm": 0.1810448169708252, "learning_rate": 7.204149350525387e-05, "loss": 0.0173, "step": 7740 }, { "grad_norm": 0.24155084788799286, "learning_rate": 7.196725645401309e-05, "loss": 0.0165, "step": 7750 }, { "grad_norm": 0.2483537197113037, "learning_rate": 7.1892959345132e-05, "loss": 0.0171, "step": 7760 }, { "grad_norm": 0.24381069839000702, "learning_rate": 7.181860238173605e-05, "loss": 0.0152, "step": 7770 }, { "grad_norm": 0.2145576775074005, "learning_rate": 7.174418576711432e-05, "loss": 0.0192, "step": 7780 }, { "grad_norm": 0.2634003162384033, "learning_rate": 7.1669709704719e-05, "loss": 0.0173, "step": 7790 }, { "grad_norm": 0.22307248413562775, "learning_rate": 7.159517439816481e-05, "loss": 0.0177, "step": 7800 }, { "grad_norm": 0.19792577624320984, "learning_rate": 7.152058005122842e-05, "loss": 0.019, "step": 7810 }, { "grad_norm": 0.2338951677083969, "learning_rate": 7.144592686784793e-05, "loss": 0.0219, "step": 7820 }, { "grad_norm": 0.2148582637310028, "learning_rate": 7.137121505212229e-05, "loss": 0.0245, "step": 7830 }, { "grad_norm": 0.2876260578632355, "learning_rate": 7.129644480831077e-05, "loss": 0.0196, "step": 7840 }, { "grad_norm": 0.28695014119148254, "learning_rate": 7.122161634083234e-05, "loss": 0.0172, "step": 7850 }, { "grad_norm": 0.29757457971572876, "learning_rate": 7.114672985426516e-05, "loss": 0.0178, "step": 7860 }, { "grad_norm": 0.23931890726089478, "learning_rate": 7.107178555334606e-05, "loss": 0.0177, "step": 7870 }, { "grad_norm": 0.3041757047176361, "learning_rate": 7.099678364296989e-05, "loss": 0.0205, "step": 7880 }, { "grad_norm": 0.2287132292985916, "learning_rate": 7.0921724328189e-05, "loss": 0.0195, "step": 7890 }, { "grad_norm": 0.25023481249809265, "learning_rate": 7.084660781421268e-05, "loss": 0.0186, "step": 7900 }, { "grad_norm": 0.33957016468048096, "learning_rate": 7.077143430640662e-05, "loss": 0.0241, "step": 7910 }, { "grad_norm": 0.24979884922504425, "learning_rate": 7.069620401029232e-05, "loss": 0.0205, "step": 7920 }, { "grad_norm": 0.26833271980285645, "learning_rate": 7.062091713154655e-05, "loss": 0.0182, "step": 7930 }, { "grad_norm": 0.2358444333076477, "learning_rate": 7.054557387600075e-05, "loss": 0.0194, "step": 7940 }, { "grad_norm": 0.34401842951774597, "learning_rate": 7.04701744496405e-05, "loss": 0.0241, "step": 7950 }, { "grad_norm": 0.318040132522583, "learning_rate": 7.039471905860495e-05, "loss": 0.0254, "step": 7960 }, { "grad_norm": 0.35591861605644226, "learning_rate": 7.031920790918628e-05, "loss": 0.0205, "step": 7970 }, { "grad_norm": 0.3557029962539673, "learning_rate": 7.024364120782906e-05, "loss": 0.0245, "step": 7980 }, { "grad_norm": 0.20709972083568573, "learning_rate": 7.016801916112978e-05, "loss": 0.0199, "step": 7990 }, { "grad_norm": 0.2537086308002472, "learning_rate": 7.009234197583623e-05, "loss": 0.0188, "step": 8000 }, { "grad_norm": 0.32834625244140625, "learning_rate": 7.001660985884692e-05, "loss": 0.0226, "step": 8010 }, { "grad_norm": 0.22356341779232025, "learning_rate": 6.994082301721063e-05, "loss": 0.0199, "step": 8020 }, { "grad_norm": 0.26609379053115845, "learning_rate": 6.986498165812563e-05, "loss": 0.0206, "step": 8030 }, { "grad_norm": 0.2014419138431549, "learning_rate": 6.978908598893932e-05, "loss": 0.0174, "step": 8040 }, { "grad_norm": 0.24261440336704254, "learning_rate": 6.971313621714756e-05, "loss": 0.017, "step": 8050 }, { "grad_norm": 0.2565006613731384, "learning_rate": 6.96371325503941e-05, "loss": 0.0173, "step": 8060 }, { "grad_norm": 0.17474952340126038, "learning_rate": 6.956107519647014e-05, "loss": 0.0167, "step": 8070 }, { "grad_norm": 0.20978869497776031, "learning_rate": 6.94849643633135e-05, "loss": 0.0163, "step": 8080 }, { "grad_norm": 0.24848061800003052, "learning_rate": 6.940880025900834e-05, "loss": 0.0179, "step": 8090 }, { "grad_norm": 0.2776045799255371, "learning_rate": 6.933258309178438e-05, "loss": 0.0164, "step": 8100 }, { "grad_norm": 0.22770604491233826, "learning_rate": 6.925631307001646e-05, "loss": 0.0164, "step": 8110 }, { "grad_norm": 0.19094012677669525, "learning_rate": 6.91799904022239e-05, "loss": 0.0201, "step": 8120 }, { "grad_norm": 0.2235974669456482, "learning_rate": 6.910361529706997e-05, "loss": 0.0137, "step": 8130 }, { "grad_norm": 0.23095086216926575, "learning_rate": 6.902718796336131e-05, "loss": 0.0184, "step": 8140 }, { "grad_norm": 0.22299855947494507, "learning_rate": 6.895070861004729e-05, "loss": 0.0174, "step": 8150 }, { "grad_norm": 0.24088384211063385, "learning_rate": 6.887417744621956e-05, "loss": 0.0172, "step": 8160 }, { "grad_norm": 0.22246411442756653, "learning_rate": 6.87975946811114e-05, "loss": 0.0164, "step": 8170 }, { "grad_norm": 0.19309036433696747, "learning_rate": 6.872096052409718e-05, "loss": 0.0177, "step": 8180 }, { "grad_norm": 0.17469771206378937, "learning_rate": 6.864427518469174e-05, "loss": 0.0172, "step": 8190 }, { "grad_norm": 0.30553489923477173, "learning_rate": 6.856753887254986e-05, "loss": 0.0164, "step": 8200 }, { "grad_norm": 0.24126176536083221, "learning_rate": 6.849075179746572e-05, "loss": 0.0174, "step": 8210 }, { "grad_norm": 0.20547087490558624, "learning_rate": 6.841391416937221e-05, "loss": 0.0163, "step": 8220 }, { "grad_norm": 0.30071312189102173, "learning_rate": 6.833702619834053e-05, "loss": 0.0176, "step": 8230 }, { "grad_norm": 0.17966130375862122, "learning_rate": 6.82600880945794e-05, "loss": 0.0143, "step": 8240 }, { "grad_norm": 0.17524124681949615, "learning_rate": 6.818310006843468e-05, "loss": 0.017, "step": 8250 }, { "grad_norm": 0.26483047008514404, "learning_rate": 6.810606233038868e-05, "loss": 0.0163, "step": 8260 }, { "grad_norm": 0.30407842993736267, "learning_rate": 6.802897509105966e-05, "loss": 0.0225, "step": 8270 }, { "grad_norm": 0.2840201258659363, "learning_rate": 6.79518385612012e-05, "loss": 0.0179, "step": 8280 }, { "grad_norm": 0.2672426998615265, "learning_rate": 6.787465295170157e-05, "loss": 0.0203, "step": 8290 }, { "grad_norm": 0.2142908126115799, "learning_rate": 6.779741847358332e-05, "loss": 0.0168, "step": 8300 }, { "grad_norm": 0.3199232220649719, "learning_rate": 6.772013533800256e-05, "loss": 0.0208, "step": 8310 }, { "grad_norm": 0.2827606797218323, "learning_rate": 6.764280375624843e-05, "loss": 0.0175, "step": 8320 }, { "grad_norm": 0.23855434358119965, "learning_rate": 6.756542393974252e-05, "loss": 0.0159, "step": 8330 }, { "grad_norm": 0.20263145864009857, "learning_rate": 6.748799610003828e-05, "loss": 0.0149, "step": 8340 }, { "grad_norm": 0.16666564345359802, "learning_rate": 6.741052044882048e-05, "loss": 0.0156, "step": 8350 }, { "grad_norm": 0.2526615560054779, "learning_rate": 6.73329971979046e-05, "loss": 0.0162, "step": 8360 }, { "grad_norm": 0.2492632120847702, "learning_rate": 6.725542655923625e-05, "loss": 0.0148, "step": 8370 }, { "grad_norm": 0.17501485347747803, "learning_rate": 6.717780874489057e-05, "loss": 0.0166, "step": 8380 }, { "grad_norm": 0.1896304339170456, "learning_rate": 6.710014396707172e-05, "loss": 0.0168, "step": 8390 }, { "grad_norm": 0.23046039044857025, "learning_rate": 6.702243243811221e-05, "loss": 0.0144, "step": 8400 }, { "grad_norm": 0.25016841292381287, "learning_rate": 6.694467437047244e-05, "loss": 0.0149, "step": 8410 }, { "grad_norm": 0.1949721723794937, "learning_rate": 6.686686997673997e-05, "loss": 0.0152, "step": 8420 }, { "grad_norm": 0.24789372086524963, "learning_rate": 6.678901946962903e-05, "loss": 0.0162, "step": 8430 }, { "grad_norm": 0.24717895686626434, "learning_rate": 6.671112306197996e-05, "loss": 0.0146, "step": 8440 }, { "grad_norm": 0.19295766949653625, "learning_rate": 6.663318096675854e-05, "loss": 0.0149, "step": 8450 }, { "grad_norm": 0.23622924089431763, "learning_rate": 6.655519339705552e-05, "loss": 0.015, "step": 8460 }, { "grad_norm": 0.24216197431087494, "learning_rate": 6.647716056608588e-05, "loss": 0.0148, "step": 8470 }, { "grad_norm": 0.27196305990219116, "learning_rate": 6.639908268718843e-05, "loss": 0.0192, "step": 8480 }, { "grad_norm": 0.31252521276474, "learning_rate": 6.632095997382514e-05, "loss": 0.0172, "step": 8490 }, { "grad_norm": 0.2913345396518707, "learning_rate": 6.624279263958047e-05, "loss": 0.0152, "step": 8500 }, { "grad_norm": 0.2721444070339203, "learning_rate": 6.616458089816097e-05, "loss": 0.0157, "step": 8510 }, { "grad_norm": 0.2878122925758362, "learning_rate": 6.608632496339454e-05, "loss": 0.0163, "step": 8520 }, { "grad_norm": 0.23463095724582672, "learning_rate": 6.600802504922988e-05, "loss": 0.015, "step": 8530 }, { "grad_norm": 0.3177334666252136, "learning_rate": 6.592968136973604e-05, "loss": 0.0143, "step": 8540 }, { "grad_norm": 0.23913946747779846, "learning_rate": 6.585129413910159e-05, "loss": 0.0175, "step": 8550 }, { "grad_norm": 0.15877127647399902, "learning_rate": 6.577286357163424e-05, "loss": 0.014, "step": 8560 }, { "grad_norm": 0.2996457517147064, "learning_rate": 6.569438988176018e-05, "loss": 0.017, "step": 8570 }, { "grad_norm": 0.29986000061035156, "learning_rate": 6.561587328402347e-05, "loss": 0.0163, "step": 8580 }, { "grad_norm": 0.2384333312511444, "learning_rate": 6.553731399308549e-05, "loss": 0.0161, "step": 8590 }, { "grad_norm": 0.24613316357135773, "learning_rate": 6.545871222372436e-05, "loss": 0.0187, "step": 8600 }, { "grad_norm": 0.24091483652591705, "learning_rate": 6.538006819083426e-05, "loss": 0.0174, "step": 8610 }, { "grad_norm": 0.16167372465133667, "learning_rate": 6.530138210942505e-05, "loss": 0.0176, "step": 8620 }, { "grad_norm": 0.462427020072937, "learning_rate": 6.522265419462141e-05, "loss": 0.0162, "step": 8630 }, { "grad_norm": 0.19005243480205536, "learning_rate": 6.514388466166248e-05, "loss": 0.016, "step": 8640 }, { "grad_norm": 0.22093817591667175, "learning_rate": 6.506507372590119e-05, "loss": 0.0196, "step": 8650 }, { "grad_norm": 0.2605013847351074, "learning_rate": 6.498622160280355e-05, "loss": 0.0166, "step": 8660 }, { "grad_norm": 0.2998715043067932, "learning_rate": 6.490732850794832e-05, "loss": 0.0175, "step": 8670 }, { "grad_norm": 0.24319468438625336, "learning_rate": 6.482839465702616e-05, "loss": 0.0199, "step": 8680 }, { "grad_norm": 0.24138812720775604, "learning_rate": 6.474942026583923e-05, "loss": 0.0154, "step": 8690 }, { "grad_norm": 0.1867961436510086, "learning_rate": 6.467040555030052e-05, "loss": 0.0168, "step": 8700 }, { "grad_norm": 0.17850536108016968, "learning_rate": 6.459135072643321e-05, "loss": 0.0137, "step": 8710 }, { "grad_norm": 0.23338258266448975, "learning_rate": 6.451225601037019e-05, "loss": 0.0144, "step": 8720 }, { "grad_norm": 0.24381405115127563, "learning_rate": 6.443312161835338e-05, "loss": 0.0169, "step": 8730 }, { "grad_norm": 0.20882263779640198, "learning_rate": 6.43539477667332e-05, "loss": 0.0155, "step": 8740 }, { "grad_norm": 0.28676995635032654, "learning_rate": 6.427473467196793e-05, "loss": 0.0166, "step": 8750 }, { "grad_norm": 0.2546941339969635, "learning_rate": 6.419548255062315e-05, "loss": 0.0176, "step": 8760 }, { "grad_norm": 0.20132270455360413, "learning_rate": 6.411619161937112e-05, "loss": 0.015, "step": 8770 }, { "grad_norm": 0.3067421317100525, "learning_rate": 6.403686209499022e-05, "loss": 0.0172, "step": 8780 }, { "grad_norm": 0.26528966426849365, "learning_rate": 6.395749419436437e-05, "loss": 0.0191, "step": 8790 }, { "grad_norm": 0.2915380001068115, "learning_rate": 6.387808813448234e-05, "loss": 0.0171, "step": 8800 }, { "grad_norm": 0.22365249693393707, "learning_rate": 6.37986441324373e-05, "loss": 0.0176, "step": 8810 }, { "grad_norm": 0.29425835609436035, "learning_rate": 6.37191624054261e-05, "loss": 0.017, "step": 8820 }, { "grad_norm": 0.28182077407836914, "learning_rate": 6.363964317074872e-05, "loss": 0.0174, "step": 8830 }, { "grad_norm": 0.20635971426963806, "learning_rate": 6.356008664580776e-05, "loss": 0.0133, "step": 8840 }, { "grad_norm": 0.25797897577285767, "learning_rate": 6.348049304810771e-05, "loss": 0.0169, "step": 8850 }, { "grad_norm": 0.17963339388370514, "learning_rate": 6.340086259525442e-05, "loss": 0.0159, "step": 8860 }, { "grad_norm": 0.2213515192270279, "learning_rate": 6.332119550495448e-05, "loss": 0.0131, "step": 8870 }, { "grad_norm": 0.1862201690673828, "learning_rate": 6.324149199501473e-05, "loss": 0.0164, "step": 8880 }, { "grad_norm": 0.2558005452156067, "learning_rate": 6.316175228334146e-05, "loss": 0.0143, "step": 8890 }, { "grad_norm": 0.3173601031303406, "learning_rate": 6.308197658794003e-05, "loss": 0.0181, "step": 8900 }, { "grad_norm": 0.18817725777626038, "learning_rate": 6.300216512691417e-05, "loss": 0.0133, "step": 8910 }, { "grad_norm": 0.17183180153369904, "learning_rate": 6.292231811846532e-05, "loss": 0.014, "step": 8920 }, { "grad_norm": 0.22485992312431335, "learning_rate": 6.284243578089217e-05, "loss": 0.0124, "step": 8930 }, { "grad_norm": 0.23541121184825897, "learning_rate": 6.276251833258999e-05, "loss": 0.0144, "step": 8940 }, { "grad_norm": 0.17518948018550873, "learning_rate": 6.268256599205003e-05, "loss": 0.0156, "step": 8950 }, { "grad_norm": 0.31200090050697327, "learning_rate": 6.260257897785892e-05, "loss": 0.0166, "step": 8960 }, { "grad_norm": 0.2588314712047577, "learning_rate": 6.252255750869811e-05, "loss": 0.0156, "step": 8970 }, { "grad_norm": 0.1739206612110138, "learning_rate": 6.244250180334325e-05, "loss": 0.0169, "step": 8980 }, { "grad_norm": 0.28212660551071167, "learning_rate": 6.236241208066356e-05, "loss": 0.0158, "step": 8990 }, { "grad_norm": 0.23510552942752838, "learning_rate": 6.228228855962133e-05, "loss": 0.0193, "step": 9000 }, { "grad_norm": 0.16220645606517792, "learning_rate": 6.220213145927115e-05, "loss": 0.0154, "step": 9010 }, { "grad_norm": 0.1967046558856964, "learning_rate": 6.212194099875951e-05, "loss": 0.0144, "step": 9020 }, { "grad_norm": 0.21151095628738403, "learning_rate": 6.204171739732405e-05, "loss": 0.0131, "step": 9030 }, { "grad_norm": 0.20975472033023834, "learning_rate": 6.196146087429303e-05, "loss": 0.0143, "step": 9040 }, { "grad_norm": 0.18925878405570984, "learning_rate": 6.188117164908474e-05, "loss": 0.0146, "step": 9050 }, { "grad_norm": 0.2373703271150589, "learning_rate": 6.180084994120684e-05, "loss": 0.0162, "step": 9060 }, { "grad_norm": 0.24789804220199585, "learning_rate": 6.17204959702558e-05, "loss": 0.0155, "step": 9070 }, { "grad_norm": 0.24581287801265717, "learning_rate": 6.164010995591635e-05, "loss": 0.0166, "step": 9080 }, { "grad_norm": 0.1681068241596222, "learning_rate": 6.155969211796076e-05, "loss": 0.015, "step": 9090 }, { "grad_norm": 0.23939208686351776, "learning_rate": 6.147924267624829e-05, "loss": 0.0155, "step": 9100 }, { "grad_norm": 0.23828832805156708, "learning_rate": 6.13987618507247e-05, "loss": 0.017, "step": 9110 }, { "grad_norm": 0.26217731833457947, "learning_rate": 6.131824986142147e-05, "loss": 0.0156, "step": 9120 }, { "grad_norm": 0.16936837136745453, "learning_rate": 6.123770692845529e-05, "loss": 0.0115, "step": 9130 }, { "grad_norm": 0.2868308424949646, "learning_rate": 6.11571332720275e-05, "loss": 0.0138, "step": 9140 }, { "grad_norm": 0.26748451590538025, "learning_rate": 6.107652911242336e-05, "loss": 0.0183, "step": 9150 }, { "grad_norm": 0.22674813866615295, "learning_rate": 6.0995894670011586e-05, "loss": 0.014, "step": 9160 }, { "grad_norm": 0.21759876608848572, "learning_rate": 6.091523016524368e-05, "loss": 0.0152, "step": 9170 }, { "grad_norm": 0.25686028599739075, "learning_rate": 6.083453581865328e-05, "loss": 0.015, "step": 9180 }, { "grad_norm": 0.2304021567106247, "learning_rate": 6.075381185085568e-05, "loss": 0.0143, "step": 9190 }, { "grad_norm": 0.25461068749427795, "learning_rate": 6.067305848254709e-05, "loss": 0.0179, "step": 9200 }, { "grad_norm": 0.25499722361564636, "learning_rate": 6.059227593450418e-05, "loss": 0.0155, "step": 9210 }, { "grad_norm": 0.2619040906429291, "learning_rate": 6.051146442758333e-05, "loss": 0.0177, "step": 9220 }, { "grad_norm": 0.24324728548526764, "learning_rate": 6.043062418272012e-05, "loss": 0.0146, "step": 9230 }, { "grad_norm": 0.2236449271440506, "learning_rate": 6.0349755420928666e-05, "loss": 0.0162, "step": 9240 }, { "grad_norm": 0.21873562037944794, "learning_rate": 6.0268858363301105e-05, "loss": 0.0177, "step": 9250 }, { "grad_norm": 0.24383297562599182, "learning_rate": 6.018793323100689e-05, "loss": 0.0145, "step": 9260 }, { "grad_norm": 0.21697841584682465, "learning_rate": 6.0106980245292255e-05, "loss": 0.0145, "step": 9270 }, { "grad_norm": 0.1404578685760498, "learning_rate": 6.002599962747957e-05, "loss": 0.013, "step": 9280 }, { "grad_norm": 0.21760018169879913, "learning_rate": 5.994499159896673e-05, "loss": 0.0146, "step": 9290 }, { "grad_norm": 0.2171415239572525, "learning_rate": 5.9863956381226607e-05, "loss": 0.0167, "step": 9300 }, { "grad_norm": 0.3201409876346588, "learning_rate": 5.9782894195806394e-05, "loss": 0.0139, "step": 9310 }, { "grad_norm": 0.2838045656681061, "learning_rate": 5.9701805264327004e-05, "loss": 0.0152, "step": 9320 }, { "grad_norm": 0.25209781527519226, "learning_rate": 5.96206898084825e-05, "loss": 0.0154, "step": 9330 }, { "grad_norm": 0.24573639035224915, "learning_rate": 5.953954805003942e-05, "loss": 0.0149, "step": 9340 }, { "grad_norm": 0.20374000072479248, "learning_rate": 5.945838021083623e-05, "loss": 0.0124, "step": 9350 }, { "grad_norm": 0.2762211561203003, "learning_rate": 5.9377186512782714e-05, "loss": 0.017, "step": 9360 }, { "grad_norm": 0.1607363373041153, "learning_rate": 5.929596717785935e-05, "loss": 0.0174, "step": 9370 }, { "grad_norm": 0.1956639438867569, "learning_rate": 5.921472242811668e-05, "loss": 0.0179, "step": 9380 }, { "grad_norm": 0.3134216070175171, "learning_rate": 5.913345248567475e-05, "loss": 0.0165, "step": 9390 }, { "grad_norm": 0.3438674211502075, "learning_rate": 5.905215757272248e-05, "loss": 0.0152, "step": 9400 }, { "grad_norm": 0.23182900249958038, "learning_rate": 5.897083791151706e-05, "loss": 0.0149, "step": 9410 }, { "grad_norm": 0.2858206629753113, "learning_rate": 5.888949372438336e-05, "loss": 0.0188, "step": 9420 }, { "grad_norm": 0.28778859972953796, "learning_rate": 5.8808125233713255e-05, "loss": 0.0147, "step": 9430 }, { "grad_norm": 0.20862224698066711, "learning_rate": 5.872673266196509e-05, "loss": 0.0157, "step": 9440 }, { "grad_norm": 0.2550677955150604, "learning_rate": 5.864531623166305e-05, "loss": 0.0137, "step": 9450 }, { "grad_norm": 0.23168016970157623, "learning_rate": 5.856387616539656e-05, "loss": 0.016, "step": 9460 }, { "grad_norm": 0.19290634989738464, "learning_rate": 5.848241268581967e-05, "loss": 0.015, "step": 9470 }, { "grad_norm": 0.19897082448005676, "learning_rate": 5.840092601565037e-05, "loss": 0.0133, "step": 9480 }, { "grad_norm": 0.24782288074493408, "learning_rate": 5.8319416377670144e-05, "loss": 0.0145, "step": 9490 }, { "grad_norm": 0.2901589572429657, "learning_rate": 5.82378839947232e-05, "loss": 0.0172, "step": 9500 }, { "grad_norm": 0.16154535114765167, "learning_rate": 5.815632908971599e-05, "loss": 0.0163, "step": 9510 }, { "grad_norm": 0.24241513013839722, "learning_rate": 5.80747518856165e-05, "loss": 0.0172, "step": 9520 }, { "grad_norm": 0.2022661566734314, "learning_rate": 5.799315260545367e-05, "loss": 0.0136, "step": 9530 }, { "grad_norm": 0.19990874826908112, "learning_rate": 5.791153147231686e-05, "loss": 0.0165, "step": 9540 }, { "grad_norm": 0.28144073486328125, "learning_rate": 5.782988870935509e-05, "loss": 0.0154, "step": 9550 }, { "grad_norm": 0.23396556079387665, "learning_rate": 5.774822453977657e-05, "loss": 0.0159, "step": 9560 }, { "grad_norm": 0.3260382413864136, "learning_rate": 5.7666539186848036e-05, "loss": 0.0151, "step": 9570 }, { "grad_norm": 0.2949216365814209, "learning_rate": 5.758483287389411e-05, "loss": 0.0157, "step": 9580 }, { "grad_norm": 0.23506513237953186, "learning_rate": 5.7503105824296735e-05, "loss": 0.0131, "step": 9590 }, { "grad_norm": 0.24227792024612427, "learning_rate": 5.742135826149453e-05, "loss": 0.0158, "step": 9600 }, { "grad_norm": 0.3921871483325958, "learning_rate": 5.7339590408982223e-05, "loss": 0.0163, "step": 9610 }, { "grad_norm": 0.3276906907558441, "learning_rate": 5.725780249031e-05, "loss": 0.0163, "step": 9620 }, { "grad_norm": 0.2850809395313263, "learning_rate": 5.717599472908292e-05, "loss": 0.0184, "step": 9630 }, { "grad_norm": 0.1402866542339325, "learning_rate": 5.7094167348960237e-05, "loss": 0.0127, "step": 9640 }, { "grad_norm": 0.3249048590660095, "learning_rate": 5.7012320573654945e-05, "loss": 0.0169, "step": 9650 }, { "grad_norm": 0.20494817197322845, "learning_rate": 5.693045462693295e-05, "loss": 0.0143, "step": 9660 }, { "grad_norm": 0.2397540658712387, "learning_rate": 5.684856973261266e-05, "loss": 0.0132, "step": 9670 }, { "grad_norm": 0.24420522153377533, "learning_rate": 5.6766666114564215e-05, "loss": 0.0163, "step": 9680 }, { "grad_norm": 0.27738016843795776, "learning_rate": 5.668474399670899e-05, "loss": 0.0134, "step": 9690 }, { "grad_norm": 0.16654278337955475, "learning_rate": 5.660280360301896e-05, "loss": 0.0129, "step": 9700 }, { "grad_norm": 0.17653422057628632, "learning_rate": 5.652084515751599e-05, "loss": 0.0156, "step": 9710 }, { "grad_norm": 0.26727497577667236, "learning_rate": 5.643886888427137e-05, "loss": 0.0153, "step": 9720 }, { "grad_norm": 0.2528059184551239, "learning_rate": 5.6356875007405074e-05, "loss": 0.0147, "step": 9730 }, { "grad_norm": 0.19305428862571716, "learning_rate": 5.627486375108525e-05, "loss": 0.0133, "step": 9740 }, { "grad_norm": 0.20062606036663055, "learning_rate": 5.619283533952754e-05, "loss": 0.015, "step": 9750 }, { "grad_norm": 0.2507815361022949, "learning_rate": 5.6110789996994474e-05, "loss": 0.0136, "step": 9760 }, { "grad_norm": 0.14258165657520294, "learning_rate": 5.602872794779491e-05, "loss": 0.0122, "step": 9770 }, { "grad_norm": 0.25183331966400146, "learning_rate": 5.594664941628334e-05, "loss": 0.0146, "step": 9780 }, { "grad_norm": 0.20962822437286377, "learning_rate": 5.5864554626859324e-05, "loss": 0.0166, "step": 9790 }, { "grad_norm": 0.19947558641433716, "learning_rate": 5.578244380396691e-05, "loss": 0.0134, "step": 9800 }, { "grad_norm": 0.20691658556461334, "learning_rate": 5.570031717209394e-05, "loss": 0.0144, "step": 9810 }, { "grad_norm": 0.2477017194032669, "learning_rate": 5.561817495577147e-05, "loss": 0.012, "step": 9820 }, { "grad_norm": 0.20643216371536255, "learning_rate": 5.5536017379573215e-05, "loss": 0.0173, "step": 9830 }, { "grad_norm": 0.30732056498527527, "learning_rate": 5.545384466811483e-05, "loss": 0.0168, "step": 9840 }, { "grad_norm": 0.275785893201828, "learning_rate": 5.5371657046053384e-05, "loss": 0.0165, "step": 9850 }, { "grad_norm": 0.2695739269256592, "learning_rate": 5.528945473808669e-05, "loss": 0.0171, "step": 9860 }, { "grad_norm": 0.18648453056812286, "learning_rate": 5.520723796895272e-05, "loss": 0.0144, "step": 9870 }, { "grad_norm": 0.23861049115657806, "learning_rate": 5.512500696342897e-05, "loss": 0.0153, "step": 9880 }, { "grad_norm": 0.24855469167232513, "learning_rate": 5.504276194633188e-05, "loss": 0.0167, "step": 9890 }, { "grad_norm": 0.22213079035282135, "learning_rate": 5.49605031425162e-05, "loss": 0.0168, "step": 9900 }, { "grad_norm": 0.20375676453113556, "learning_rate": 5.487823077687434e-05, "loss": 0.016, "step": 9910 }, { "grad_norm": 0.2132408618927002, "learning_rate": 5.4795945074335806e-05, "loss": 0.0157, "step": 9920 }, { "grad_norm": 0.21304087340831757, "learning_rate": 5.471364625986657e-05, "loss": 0.0123, "step": 9930 }, { "grad_norm": 0.14666326344013214, "learning_rate": 5.463133455846845e-05, "loss": 0.0149, "step": 9940 }, { "grad_norm": 0.340154230594635, "learning_rate": 5.4549010195178505e-05, "loss": 0.0152, "step": 9950 }, { "grad_norm": 0.21569539606571198, "learning_rate": 5.446667339506838e-05, "loss": 0.0143, "step": 9960 }, { "grad_norm": 0.21765951812267303, "learning_rate": 5.4384324383243756e-05, "loss": 0.0133, "step": 9970 }, { "grad_norm": 0.22184298932552338, "learning_rate": 5.430196338484368e-05, "loss": 0.0132, "step": 9980 }, { "grad_norm": 0.19340775907039642, "learning_rate": 5.4219590625039975e-05, "loss": 0.0144, "step": 9990 }, { "grad_norm": 0.2250930666923523, "learning_rate": 5.413720632903664e-05, "loss": 0.0171, "step": 10000 }, { "grad_norm": 0.21738405525684357, "learning_rate": 5.405481072206917e-05, "loss": 0.0148, "step": 10010 }, { "grad_norm": 0.1984861195087433, "learning_rate": 5.397240402940402e-05, "loss": 0.0139, "step": 10020 }, { "grad_norm": 0.18186712265014648, "learning_rate": 5.388998647633794e-05, "loss": 0.0135, "step": 10030 }, { "grad_norm": 0.23618152737617493, "learning_rate": 5.380755828819737e-05, "loss": 0.0141, "step": 10040 }, { "grad_norm": 0.15989810228347778, "learning_rate": 5.3725119690337846e-05, "loss": 0.013, "step": 10050 }, { "grad_norm": 0.21025557816028595, "learning_rate": 5.3642670908143324e-05, "loss": 0.0144, "step": 10060 }, { "grad_norm": 0.18647345900535583, "learning_rate": 5.356021216702562e-05, "loss": 0.0155, "step": 10070 }, { "grad_norm": 0.20366987586021423, "learning_rate": 5.347774369242381e-05, "loss": 0.0138, "step": 10080 }, { "grad_norm": 0.16277140378952026, "learning_rate": 5.3395265709803545e-05, "loss": 0.0119, "step": 10090 }, { "grad_norm": 0.17942342162132263, "learning_rate": 5.331277844465647e-05, "loss": 0.0126, "step": 10100 }, { "grad_norm": 0.18167932331562042, "learning_rate": 5.323028212249963e-05, "loss": 0.0132, "step": 10110 }, { "grad_norm": 0.16819700598716736, "learning_rate": 5.314777696887481e-05, "loss": 0.0116, "step": 10120 }, { "grad_norm": 0.20540957152843475, "learning_rate": 5.306526320934796e-05, "loss": 0.0116, "step": 10130 }, { "grad_norm": 0.2136903703212738, "learning_rate": 5.298274106950854e-05, "loss": 0.0112, "step": 10140 }, { "grad_norm": 0.17087917029857635, "learning_rate": 5.290021077496893e-05, "loss": 0.0127, "step": 10150 }, { "grad_norm": 0.20682460069656372, "learning_rate": 5.2817672551363816e-05, "loss": 0.0127, "step": 10160 }, { "grad_norm": 0.20931626856327057, "learning_rate": 5.273512662434952e-05, "loss": 0.0138, "step": 10170 }, { "grad_norm": 0.19126757979393005, "learning_rate": 5.265257321960349e-05, "loss": 0.0126, "step": 10180 }, { "grad_norm": 0.2222648561000824, "learning_rate": 5.257001256282357e-05, "loss": 0.0143, "step": 10190 }, { "grad_norm": 0.16010713577270508, "learning_rate": 5.248744487972742e-05, "loss": 0.0126, "step": 10200 }, { "grad_norm": 0.21992014348506927, "learning_rate": 5.240487039605196e-05, "loss": 0.0148, "step": 10210 }, { "grad_norm": 0.25323227047920227, "learning_rate": 5.232228933755267e-05, "loss": 0.0127, "step": 10220 }, { "grad_norm": 0.2086285948753357, "learning_rate": 5.2239701930003006e-05, "loss": 0.0126, "step": 10230 }, { "grad_norm": 0.18220333755016327, "learning_rate": 5.215710839919379e-05, "loss": 0.0117, "step": 10240 }, { "grad_norm": 0.26334160566329956, "learning_rate": 5.207450897093257e-05, "loss": 0.0123, "step": 10250 }, { "grad_norm": 0.1777537763118744, "learning_rate": 5.1991903871043046e-05, "loss": 0.0134, "step": 10260 }, { "grad_norm": 0.16211360692977905, "learning_rate": 5.190929332536439e-05, "loss": 0.0132, "step": 10270 }, { "grad_norm": 0.3499620854854584, "learning_rate": 5.182667755975071e-05, "loss": 0.0185, "step": 10280 }, { "grad_norm": 0.2267434298992157, "learning_rate": 5.1744056800070315e-05, "loss": 0.0152, "step": 10290 }, { "grad_norm": 0.23871402442455292, "learning_rate": 5.166143127220524e-05, "loss": 0.0121, "step": 10300 }, { "grad_norm": 0.14621588587760925, "learning_rate": 5.1578801202050485e-05, "loss": 0.0129, "step": 10310 }, { "grad_norm": 0.18679270148277283, "learning_rate": 5.149616681551355e-05, "loss": 0.0163, "step": 10320 }, { "grad_norm": 0.14395032823085785, "learning_rate": 5.141352833851367e-05, "loss": 0.0104, "step": 10330 }, { "grad_norm": 0.18652021884918213, "learning_rate": 5.1330885996981285e-05, "loss": 0.0104, "step": 10340 }, { "grad_norm": 0.1289137601852417, "learning_rate": 5.124824001685741e-05, "loss": 0.0108, "step": 10350 }, { "grad_norm": 0.16982261836528778, "learning_rate": 5.116559062409298e-05, "loss": 0.01, "step": 10360 }, { "grad_norm": 0.24912673234939575, "learning_rate": 5.10829380446483e-05, "loss": 0.0114, "step": 10370 }, { "grad_norm": 0.2593719959259033, "learning_rate": 5.100028250449235e-05, "loss": 0.0118, "step": 10380 }, { "grad_norm": 0.20945888757705688, "learning_rate": 5.0917624229602234e-05, "loss": 0.0116, "step": 10390 }, { "grad_norm": 0.26518452167510986, "learning_rate": 5.0834963445962524e-05, "loss": 0.0137, "step": 10400 }, { "grad_norm": 0.21983595192432404, "learning_rate": 5.075230037956461e-05, "loss": 0.0121, "step": 10410 }, { "grad_norm": 0.21103666722774506, "learning_rate": 5.0669635256406213e-05, "loss": 0.0134, "step": 10420 }, { "grad_norm": 0.22496157884597778, "learning_rate": 5.058696830249058e-05, "loss": 0.0115, "step": 10430 }, { "grad_norm": 0.22121559083461761, "learning_rate": 5.050429974382602e-05, "loss": 0.0161, "step": 10440 }, { "grad_norm": 0.19921942055225372, "learning_rate": 5.042162980642523e-05, "loss": 0.0129, "step": 10450 }, { "grad_norm": 0.22929704189300537, "learning_rate": 5.033895871630462e-05, "loss": 0.0107, "step": 10460 }, { "grad_norm": 0.2015702873468399, "learning_rate": 5.025628669948386e-05, "loss": 0.0129, "step": 10470 }, { "grad_norm": 0.16210131347179413, "learning_rate": 5.017361398198502e-05, "loss": 0.0105, "step": 10480 }, { "grad_norm": 0.20006729662418365, "learning_rate": 5.009094078983221e-05, "loss": 0.0113, "step": 10490 }, { "grad_norm": 0.16510306298732758, "learning_rate": 5.000826734905073e-05, "loss": 0.0119, "step": 10500 }, { "grad_norm": 0.20159168541431427, "learning_rate": 4.9925593885666645e-05, "loss": 0.0135, "step": 10510 }, { "grad_norm": 0.2596547305583954, "learning_rate": 4.984292062570602e-05, "loss": 0.0135, "step": 10520 }, { "grad_norm": 0.19105841219425201, "learning_rate": 4.976024779519442e-05, "loss": 0.0142, "step": 10530 }, { "grad_norm": 0.2079208642244339, "learning_rate": 4.9677575620156194e-05, "loss": 0.0116, "step": 10540 }, { "grad_norm": 0.16703054308891296, "learning_rate": 4.959490432661391e-05, "loss": 0.0106, "step": 10550 }, { "grad_norm": 0.23712532222270966, "learning_rate": 4.9512234140587726e-05, "loss": 0.0124, "step": 10560 }, { "grad_norm": 0.1590341329574585, "learning_rate": 4.942956528809477e-05, "loss": 0.0114, "step": 10570 }, { "grad_norm": 0.23308712244033813, "learning_rate": 4.934689799514854e-05, "loss": 0.0106, "step": 10580 }, { "grad_norm": 0.2900393307209015, "learning_rate": 4.926423248775827e-05, "loss": 0.0125, "step": 10590 }, { "grad_norm": 0.15783965587615967, "learning_rate": 4.918156899192826e-05, "loss": 0.0115, "step": 10600 }, { "grad_norm": 0.15527956187725067, "learning_rate": 4.909890773365738e-05, "loss": 0.011, "step": 10610 }, { "grad_norm": 0.15634360909461975, "learning_rate": 4.9016248938938344e-05, "loss": 0.0115, "step": 10620 }, { "grad_norm": 0.18324986100196838, "learning_rate": 4.8933592833757156e-05, "loss": 0.0155, "step": 10630 }, { "grad_norm": 0.16489636898040771, "learning_rate": 4.8850939644092435e-05, "loss": 0.0112, "step": 10640 }, { "grad_norm": 0.27805811166763306, "learning_rate": 4.876828959591485e-05, "loss": 0.0116, "step": 10650 }, { "grad_norm": 0.16805268824100494, "learning_rate": 4.8685642915186474e-05, "loss": 0.0126, "step": 10660 }, { "grad_norm": 0.216538667678833, "learning_rate": 4.860299982786018e-05, "loss": 0.0111, "step": 10670 }, { "grad_norm": 0.27447572350502014, "learning_rate": 4.852036055987901e-05, "loss": 0.0143, "step": 10680 }, { "grad_norm": 0.222338005900383, "learning_rate": 4.843772533717558e-05, "loss": 0.0107, "step": 10690 }, { "grad_norm": 0.16961976885795593, "learning_rate": 4.835509438567142e-05, "loss": 0.0111, "step": 10700 }, { "grad_norm": 0.2749253809452057, "learning_rate": 4.827246793127639e-05, "loss": 0.0156, "step": 10710 }, { "grad_norm": 0.15790137648582458, "learning_rate": 4.818984619988807e-05, "loss": 0.0099, "step": 10720 }, { "grad_norm": 0.17354269325733185, "learning_rate": 4.810722941739115e-05, "loss": 0.0105, "step": 10730 }, { "grad_norm": 0.1941550225019455, "learning_rate": 4.8024617809656684e-05, "loss": 0.0127, "step": 10740 }, { "grad_norm": 0.3032527267932892, "learning_rate": 4.794201160254171e-05, "loss": 0.0103, "step": 10750 }, { "grad_norm": 0.25177672505378723, "learning_rate": 4.785941102188844e-05, "loss": 0.0137, "step": 10760 }, { "grad_norm": 0.20211432874202728, "learning_rate": 4.7776816293523686e-05, "loss": 0.0117, "step": 10770 }, { "grad_norm": 0.34308144450187683, "learning_rate": 4.769422764325832e-05, "loss": 0.0123, "step": 10780 }, { "grad_norm": 0.1899067461490631, "learning_rate": 4.76116452968865e-05, "loss": 0.0142, "step": 10790 }, { "grad_norm": 0.20432133972644806, "learning_rate": 4.752906948018525e-05, "loss": 0.0106, "step": 10800 }, { "grad_norm": 0.21609099209308624, "learning_rate": 4.7446500418913684e-05, "loss": 0.0099, "step": 10810 }, { "grad_norm": 0.46816977858543396, "learning_rate": 4.736393833881247e-05, "loss": 0.0122, "step": 10820 }, { "grad_norm": 0.2545136511325836, "learning_rate": 4.7281383465603194e-05, "loss": 0.0134, "step": 10830 }, { "grad_norm": 0.169282004237175, "learning_rate": 4.71988360249877e-05, "loss": 0.0131, "step": 10840 }, { "grad_norm": 0.20884770154953003, "learning_rate": 4.7116296242647554e-05, "loss": 0.0119, "step": 10850 }, { "grad_norm": 0.18723271787166595, "learning_rate": 4.703376434424336e-05, "loss": 0.0123, "step": 10860 }, { "grad_norm": 0.19101449847221375, "learning_rate": 4.695124055541421e-05, "loss": 0.0106, "step": 10870 }, { "grad_norm": 0.20921511948108673, "learning_rate": 4.6868725101776934e-05, "loss": 0.012, "step": 10880 }, { "grad_norm": 0.17719174921512604, "learning_rate": 4.678621820892567e-05, "loss": 0.0107, "step": 10890 }, { "grad_norm": 0.1617005318403244, "learning_rate": 4.670372010243111e-05, "loss": 0.0099, "step": 10900 }, { "grad_norm": 0.21239973604679108, "learning_rate": 4.662123100783992e-05, "loss": 0.0097, "step": 10910 }, { "grad_norm": 0.16977757215499878, "learning_rate": 4.653875115067415e-05, "loss": 0.0087, "step": 10920 }, { "grad_norm": 0.18096700310707092, "learning_rate": 4.6456280756430545e-05, "loss": 0.0128, "step": 10930 }, { "grad_norm": 0.30171436071395874, "learning_rate": 4.637382005058004e-05, "loss": 0.0111, "step": 10940 }, { "grad_norm": 0.22253403067588806, "learning_rate": 4.629136925856705e-05, "loss": 0.0119, "step": 10950 }, { "grad_norm": 0.24651950597763062, "learning_rate": 4.6208928605808895e-05, "loss": 0.0102, "step": 10960 }, { "grad_norm": 0.20224931836128235, "learning_rate": 4.612649831769519e-05, "loss": 0.0109, "step": 10970 }, { "grad_norm": 0.31111428141593933, "learning_rate": 4.604407861958715e-05, "loss": 0.0117, "step": 10980 }, { "grad_norm": 0.2261633425951004, "learning_rate": 4.5961669736817114e-05, "loss": 0.0123, "step": 10990 }, { "grad_norm": 0.23594127595424652, "learning_rate": 4.5879271894687814e-05, "loss": 0.0163, "step": 11000 }, { "grad_norm": 0.16583211719989777, "learning_rate": 4.5796885318471826e-05, "loss": 0.0115, "step": 11010 }, { "grad_norm": 0.2018720656633377, "learning_rate": 4.571451023341086e-05, "loss": 0.0118, "step": 11020 }, { "grad_norm": 0.18392331898212433, "learning_rate": 4.563214686471527e-05, "loss": 0.0124, "step": 11030 }, { "grad_norm": 0.12875346839427948, "learning_rate": 4.5549795437563365e-05, "loss": 0.0099, "step": 11040 }, { "grad_norm": 0.1749836951494217, "learning_rate": 4.546745617710081e-05, "loss": 0.0109, "step": 11050 }, { "grad_norm": 0.16593000292778015, "learning_rate": 4.5385129308440014e-05, "loss": 0.0095, "step": 11060 }, { "grad_norm": 0.1793351173400879, "learning_rate": 4.530281505665944e-05, "loss": 0.0134, "step": 11070 }, { "grad_norm": 0.15980024635791779, "learning_rate": 4.5220513646803134e-05, "loss": 0.0114, "step": 11080 }, { "grad_norm": 0.2374913990497589, "learning_rate": 4.513822530388003e-05, "loss": 0.012, "step": 11090 }, { "grad_norm": 0.20170755684375763, "learning_rate": 4.5055950252863296e-05, "loss": 0.0109, "step": 11100 }, { "grad_norm": 0.2541886270046234, "learning_rate": 4.4973688718689803e-05, "loss": 0.0147, "step": 11110 }, { "grad_norm": 0.21790188550949097, "learning_rate": 4.4891440926259406e-05, "loss": 0.0111, "step": 11120 }, { "grad_norm": 0.14384005963802338, "learning_rate": 4.480920710043443e-05, "loss": 0.0102, "step": 11130 }, { "grad_norm": 0.17508383095264435, "learning_rate": 4.4726987466039044e-05, "loss": 0.0141, "step": 11140 }, { "grad_norm": 0.19942712783813477, "learning_rate": 4.46447822478586e-05, "loss": 0.0174, "step": 11150 }, { "grad_norm": 0.23659269511699677, "learning_rate": 4.4562591670638974e-05, "loss": 0.0136, "step": 11160 }, { "grad_norm": 0.18859131634235382, "learning_rate": 4.4480415959086105e-05, "loss": 0.0125, "step": 11170 }, { "grad_norm": 0.22883790731430054, "learning_rate": 4.439825533786522e-05, "loss": 0.0108, "step": 11180 }, { "grad_norm": 0.17086724936962128, "learning_rate": 4.431611003160035e-05, "loss": 0.0146, "step": 11190 }, { "grad_norm": 0.19035977125167847, "learning_rate": 4.4233980264873636e-05, "loss": 0.0157, "step": 11200 }, { "grad_norm": 0.2053246647119522, "learning_rate": 4.4151866262224684e-05, "loss": 0.0169, "step": 11210 }, { "grad_norm": 0.25004956126213074, "learning_rate": 4.406976824815006e-05, "loss": 0.012, "step": 11220 }, { "grad_norm": 0.2684798836708069, "learning_rate": 4.3987686447102595e-05, "loss": 0.013, "step": 11230 }, { "grad_norm": 0.3282645046710968, "learning_rate": 4.3905621083490804e-05, "loss": 0.0102, "step": 11240 }, { "grad_norm": 0.19083893299102783, "learning_rate": 4.3823572381678286e-05, "loss": 0.0145, "step": 11250 }, { "grad_norm": 0.20477287471294403, "learning_rate": 4.374154056598301e-05, "loss": 0.0115, "step": 11260 }, { "grad_norm": 0.18212494254112244, "learning_rate": 4.3659525860676845e-05, "loss": 0.0109, "step": 11270 }, { "grad_norm": 0.1495860069990158, "learning_rate": 4.3577528489984854e-05, "loss": 0.0103, "step": 11280 }, { "grad_norm": 0.22396184504032135, "learning_rate": 4.349554867808476e-05, "loss": 0.0111, "step": 11290 }, { "grad_norm": 0.1798090785741806, "learning_rate": 4.34135866491062e-05, "loss": 0.0112, "step": 11300 }, { "grad_norm": 0.22386197745800018, "learning_rate": 4.333164262713022e-05, "loss": 0.0101, "step": 11310 }, { "grad_norm": 0.19709576666355133, "learning_rate": 4.324971683618868e-05, "loss": 0.0098, "step": 11320 }, { "grad_norm": 0.1794319897890091, "learning_rate": 4.316780950026354e-05, "loss": 0.0115, "step": 11330 }, { "grad_norm": 0.22672121226787567, "learning_rate": 4.308592084328637e-05, "loss": 0.012, "step": 11340 }, { "grad_norm": 0.15231290459632874, "learning_rate": 4.3004051089137576e-05, "loss": 0.0111, "step": 11350 }, { "grad_norm": 0.17297735810279846, "learning_rate": 4.292220046164597e-05, "loss": 0.0105, "step": 11360 }, { "grad_norm": 0.19104769825935364, "learning_rate": 4.2840369184588035e-05, "loss": 0.0133, "step": 11370 }, { "grad_norm": 0.1156485304236412, "learning_rate": 4.2758557481687345e-05, "loss": 0.0099, "step": 11380 }, { "grad_norm": 0.15822769701480865, "learning_rate": 4.267676557661403e-05, "loss": 0.0092, "step": 11390 }, { "grad_norm": 0.23529811203479767, "learning_rate": 4.2594993692983955e-05, "loss": 0.0098, "step": 11400 }, { "grad_norm": 0.17620892822742462, "learning_rate": 4.251324205435837e-05, "loss": 0.0105, "step": 11410 }, { "grad_norm": 0.15479299426078796, "learning_rate": 4.243151088424312e-05, "loss": 0.0107, "step": 11420 }, { "grad_norm": 0.21388503909111023, "learning_rate": 4.234980040608813e-05, "loss": 0.01, "step": 11430 }, { "grad_norm": 0.1620524525642395, "learning_rate": 4.22681108432867e-05, "loss": 0.0105, "step": 11440 }, { "grad_norm": 0.22485360503196716, "learning_rate": 4.2186442419174984e-05, "loss": 0.0118, "step": 11450 }, { "grad_norm": 0.2226731777191162, "learning_rate": 4.210479535703133e-05, "loss": 0.014, "step": 11460 }, { "grad_norm": 0.1750703901052475, "learning_rate": 4.202316988007567e-05, "loss": 0.0119, "step": 11470 }, { "grad_norm": 0.13731855154037476, "learning_rate": 4.194156621146901e-05, "loss": 0.0111, "step": 11480 }, { "grad_norm": 0.13277383148670197, "learning_rate": 4.1859984574312596e-05, "loss": 0.0095, "step": 11490 }, { "grad_norm": 0.1969424933195114, "learning_rate": 4.177842519164752e-05, "loss": 0.0123, "step": 11500 }, { "grad_norm": 0.18845291435718536, "learning_rate": 4.169688828645404e-05, "loss": 0.0092, "step": 11510 }, { "grad_norm": 0.2083635926246643, "learning_rate": 4.161537408165092e-05, "loss": 0.0109, "step": 11520 }, { "grad_norm": 0.20632536709308624, "learning_rate": 4.1533882800094924e-05, "loss": 0.0121, "step": 11530 }, { "grad_norm": 0.22038255631923676, "learning_rate": 4.145241466458005e-05, "loss": 0.0099, "step": 11540 }, { "grad_norm": 0.20307239890098572, "learning_rate": 4.13709698978371e-05, "loss": 0.0123, "step": 11550 }, { "grad_norm": 0.23687279224395752, "learning_rate": 4.1289548722532944e-05, "loss": 0.0116, "step": 11560 }, { "grad_norm": 0.11565456539392471, "learning_rate": 4.120815136126999e-05, "loss": 0.0098, "step": 11570 }, { "grad_norm": 0.2212340086698532, "learning_rate": 4.112677803658548e-05, "loss": 0.0094, "step": 11580 }, { "grad_norm": 0.22210803627967834, "learning_rate": 4.1045428970951e-05, "loss": 0.0115, "step": 11590 }, { "grad_norm": 0.17535655200481415, "learning_rate": 4.0964104386771785e-05, "loss": 0.0103, "step": 11600 }, { "grad_norm": 0.15506257116794586, "learning_rate": 4.0882804506386144e-05, "loss": 0.0095, "step": 11610 }, { "grad_norm": 0.1670108288526535, "learning_rate": 4.080152955206485e-05, "loss": 0.011, "step": 11620 }, { "grad_norm": 0.2137109786272049, "learning_rate": 4.0720279746010505e-05, "loss": 0.0094, "step": 11630 }, { "grad_norm": 0.17382527887821198, "learning_rate": 4.063905531035699e-05, "loss": 0.0133, "step": 11640 }, { "grad_norm": 0.10606616735458374, "learning_rate": 4.055785646716882e-05, "loss": 0.0106, "step": 11650 }, { "grad_norm": 0.1718030422925949, "learning_rate": 4.047668343844051e-05, "loss": 0.0105, "step": 11660 }, { "grad_norm": 0.14877130091190338, "learning_rate": 4.039553644609604e-05, "loss": 0.0115, "step": 11670 }, { "grad_norm": 0.15706849098205566, "learning_rate": 4.0314415711988176e-05, "loss": 0.0098, "step": 11680 }, { "grad_norm": 0.1750539392232895, "learning_rate": 4.023332145789792e-05, "loss": 0.0096, "step": 11690 }, { "grad_norm": 0.27187785506248474, "learning_rate": 4.015225390553385e-05, "loss": 0.0117, "step": 11700 }, { "grad_norm": 0.14765781164169312, "learning_rate": 4.007121327653158e-05, "loss": 0.0147, "step": 11710 }, { "grad_norm": 0.20647475123405457, "learning_rate": 3.9990199792453064e-05, "loss": 0.01, "step": 11720 }, { "grad_norm": 0.15453913807868958, "learning_rate": 3.9909213674786103e-05, "loss": 0.0103, "step": 11730 }, { "grad_norm": 0.14857546985149384, "learning_rate": 3.982825514494363e-05, "loss": 0.0094, "step": 11740 }, { "grad_norm": 0.1622646450996399, "learning_rate": 3.974732442426319e-05, "loss": 0.0092, "step": 11750 }, { "grad_norm": 0.2502105236053467, "learning_rate": 3.966642173400629e-05, "loss": 0.0107, "step": 11760 }, { "grad_norm": 0.19984343647956848, "learning_rate": 3.9585547295357764e-05, "loss": 0.0082, "step": 11770 }, { "grad_norm": 0.15005061030387878, "learning_rate": 3.950470132942526e-05, "loss": 0.0081, "step": 11780 }, { "grad_norm": 0.1921265721321106, "learning_rate": 3.942388405723856e-05, "loss": 0.0108, "step": 11790 }, { "grad_norm": 0.1351138800382614, "learning_rate": 3.9343095699749e-05, "loss": 0.0117, "step": 11800 }, { "grad_norm": 0.21033619344234467, "learning_rate": 3.9262336477828874e-05, "loss": 0.0099, "step": 11810 }, { "grad_norm": 0.242219939827919, "learning_rate": 3.9181606612270794e-05, "loss": 0.0111, "step": 11820 }, { "grad_norm": 0.1767548769712448, "learning_rate": 3.910090632378713e-05, "loss": 0.0104, "step": 11830 }, { "grad_norm": 0.21453575789928436, "learning_rate": 3.90202358330094e-05, "loss": 0.0099, "step": 11840 }, { "grad_norm": 0.11914826184511185, "learning_rate": 3.8939595360487656e-05, "loss": 0.0128, "step": 11850 }, { "grad_norm": 0.17922671139240265, "learning_rate": 3.885898512668984e-05, "loss": 0.012, "step": 11860 }, { "grad_norm": 0.19780080020427704, "learning_rate": 3.877840535200127e-05, "loss": 0.0123, "step": 11870 }, { "grad_norm": 0.20569980144500732, "learning_rate": 3.869785625672397e-05, "loss": 0.0094, "step": 11880 }, { "grad_norm": 0.2579275369644165, "learning_rate": 3.8617338061076094e-05, "loss": 0.0121, "step": 11890 }, { "grad_norm": 0.2740408182144165, "learning_rate": 3.853685098519132e-05, "loss": 0.0128, "step": 11900 }, { "grad_norm": 0.18375085294246674, "learning_rate": 3.845639524911823e-05, "loss": 0.0111, "step": 11910 }, { "grad_norm": 0.1636280119419098, "learning_rate": 3.837597107281974e-05, "loss": 0.0098, "step": 11920 }, { "grad_norm": 0.15587185323238373, "learning_rate": 3.829557867617247e-05, "loss": 0.0087, "step": 11930 }, { "grad_norm": 0.22542549669742584, "learning_rate": 3.821521827896618e-05, "loss": 0.0112, "step": 11940 }, { "grad_norm": 0.15630897879600525, "learning_rate": 3.81348901009031e-05, "loss": 0.011, "step": 11950 }, { "grad_norm": 0.17947183549404144, "learning_rate": 3.805459436159741e-05, "loss": 0.0115, "step": 11960 }, { "grad_norm": 0.16070300340652466, "learning_rate": 3.797433128057461e-05, "loss": 0.0125, "step": 11970 }, { "grad_norm": 0.2075878381729126, "learning_rate": 3.789410107727089e-05, "loss": 0.0095, "step": 11980 }, { "grad_norm": 0.16619177162647247, "learning_rate": 3.781390397103257e-05, "loss": 0.0086, "step": 11990 }, { "grad_norm": 0.18679627776145935, "learning_rate": 3.7733740181115455e-05, "loss": 0.0098, "step": 12000 }, { "grad_norm": 0.1669064611196518, "learning_rate": 3.7653609926684306e-05, "loss": 0.0108, "step": 12010 }, { "grad_norm": 0.3185606598854065, "learning_rate": 3.757351342681217e-05, "loss": 0.0132, "step": 12020 }, { "grad_norm": 0.19053444266319275, "learning_rate": 3.749345090047982e-05, "loss": 0.0109, "step": 12030 }, { "grad_norm": 0.1550602912902832, "learning_rate": 3.741342256657515e-05, "loss": 0.0087, "step": 12040 }, { "grad_norm": 0.18488168716430664, "learning_rate": 3.7333428643892567e-05, "loss": 0.0103, "step": 12050 }, { "grad_norm": 0.2508004605770111, "learning_rate": 3.725346935113239e-05, "loss": 0.0107, "step": 12060 }, { "grad_norm": 0.15030637383460999, "learning_rate": 3.717354490690029e-05, "loss": 0.0086, "step": 12070 }, { "grad_norm": 0.2779320478439331, "learning_rate": 3.709365552970664e-05, "loss": 0.0103, "step": 12080 }, { "grad_norm": 0.22192806005477905, "learning_rate": 3.7013801437965945e-05, "loss": 0.0092, "step": 12090 }, { "grad_norm": 0.22015003859996796, "learning_rate": 3.693398284999623e-05, "loss": 0.0108, "step": 12100 }, { "grad_norm": 0.14312106370925903, "learning_rate": 3.6854199984018484e-05, "loss": 0.0092, "step": 12110 }, { "grad_norm": 0.21019390225410461, "learning_rate": 3.677445305815601e-05, "loss": 0.0102, "step": 12120 }, { "grad_norm": 0.22302547097206116, "learning_rate": 3.669474229043387e-05, "loss": 0.0097, "step": 12130 }, { "grad_norm": 0.18242502212524414, "learning_rate": 3.6615067898778235e-05, "loss": 0.0089, "step": 12140 }, { "grad_norm": 0.15982046723365784, "learning_rate": 3.6535430101015866e-05, "loss": 0.0098, "step": 12150 }, { "grad_norm": 0.20597057044506073, "learning_rate": 3.645582911487345e-05, "loss": 0.0151, "step": 12160 }, { "grad_norm": 0.17153210937976837, "learning_rate": 3.637626515797706e-05, "loss": 0.0094, "step": 12170 }, { "grad_norm": 0.2367359697818756, "learning_rate": 3.629673844785152e-05, "loss": 0.0118, "step": 12180 }, { "grad_norm": 0.20402514934539795, "learning_rate": 3.621724920191979e-05, "loss": 0.011, "step": 12190 }, { "grad_norm": 0.1941729635000229, "learning_rate": 3.6137797637502444e-05, "loss": 0.008, "step": 12200 }, { "grad_norm": 0.29683181643486023, "learning_rate": 3.6058383971817035e-05, "loss": 0.0111, "step": 12210 }, { "grad_norm": 0.18733777105808258, "learning_rate": 3.59790084219775e-05, "loss": 0.01, "step": 12220 }, { "grad_norm": 0.20569513738155365, "learning_rate": 3.589967120499353e-05, "loss": 0.0114, "step": 12230 }, { "grad_norm": 0.1745534986257553, "learning_rate": 3.5820372537770075e-05, "loss": 0.0089, "step": 12240 }, { "grad_norm": 0.12937059998512268, "learning_rate": 3.5741112637106655e-05, "loss": 0.0109, "step": 12250 }, { "grad_norm": 0.22263334691524506, "learning_rate": 3.5661891719696804e-05, "loss": 0.0114, "step": 12260 }, { "grad_norm": 0.26705923676490784, "learning_rate": 3.5582710002127504e-05, "loss": 0.0108, "step": 12270 }, { "grad_norm": 0.29938268661499023, "learning_rate": 3.550356770087853e-05, "loss": 0.0113, "step": 12280 }, { "grad_norm": 0.15603797137737274, "learning_rate": 3.5424465032321914e-05, "loss": 0.009, "step": 12290 }, { "grad_norm": 0.23532390594482422, "learning_rate": 3.5345402212721335e-05, "loss": 0.0081, "step": 12300 }, { "grad_norm": 0.21208421885967255, "learning_rate": 3.526637945823152e-05, "loss": 0.0095, "step": 12310 }, { "grad_norm": 0.16370812058448792, "learning_rate": 3.518739698489767e-05, "loss": 0.0094, "step": 12320 }, { "grad_norm": 0.2524142265319824, "learning_rate": 3.510845500865485e-05, "loss": 0.0086, "step": 12330 }, { "grad_norm": 0.1159132644534111, "learning_rate": 3.502955374532739e-05, "loss": 0.0089, "step": 12340 }, { "grad_norm": 0.18857157230377197, "learning_rate": 3.495069341062836e-05, "loss": 0.0095, "step": 12350 }, { "grad_norm": 0.15494561195373535, "learning_rate": 3.4871874220158896e-05, "loss": 0.0082, "step": 12360 }, { "grad_norm": 0.17393533885478973, "learning_rate": 3.479309638940762e-05, "loss": 0.0077, "step": 12370 }, { "grad_norm": 0.27104252576828003, "learning_rate": 3.4714360133750146e-05, "loss": 0.0115, "step": 12380 }, { "grad_norm": 0.14085203409194946, "learning_rate": 3.463566566844839e-05, "loss": 0.0085, "step": 12390 }, { "grad_norm": 0.2343950718641281, "learning_rate": 3.4557013208650016e-05, "loss": 0.0105, "step": 12400 }, { "grad_norm": 0.23740044236183167, "learning_rate": 3.4478402969387857e-05, "loss": 0.0106, "step": 12410 }, { "grad_norm": 0.18042871356010437, "learning_rate": 3.4399835165579266e-05, "loss": 0.01, "step": 12420 }, { "grad_norm": 0.22215256094932556, "learning_rate": 3.4321310012025645e-05, "loss": 0.0104, "step": 12430 }, { "grad_norm": 0.18856659531593323, "learning_rate": 3.424282772341176e-05, "loss": 0.0118, "step": 12440 }, { "grad_norm": 0.18258292973041534, "learning_rate": 3.416438851430519e-05, "loss": 0.0089, "step": 12450 }, { "grad_norm": 0.23499912023544312, "learning_rate": 3.408599259915577e-05, "loss": 0.011, "step": 12460 }, { "grad_norm": 0.19045375287532806, "learning_rate": 3.400764019229487e-05, "loss": 0.0096, "step": 12470 }, { "grad_norm": 0.24957473576068878, "learning_rate": 3.3929331507935035e-05, "loss": 0.0107, "step": 12480 }, { "grad_norm": 0.14072176814079285, "learning_rate": 3.3851066760169196e-05, "loss": 0.0082, "step": 12490 }, { "grad_norm": 0.26666051149368286, "learning_rate": 3.377284616297021e-05, "loss": 0.009, "step": 12500 }, { "grad_norm": 0.21635226905345917, "learning_rate": 3.3694669930190166e-05, "loss": 0.01, "step": 12510 }, { "grad_norm": 0.19017444550991058, "learning_rate": 3.36165382755599e-05, "loss": 0.0099, "step": 12520 }, { "grad_norm": 0.1486937701702118, "learning_rate": 3.35384514126884e-05, "loss": 0.0096, "step": 12530 }, { "grad_norm": 0.19716839492321014, "learning_rate": 3.3460409555062154e-05, "loss": 0.0089, "step": 12540 }, { "grad_norm": 0.1379709094762802, "learning_rate": 3.3382412916044645e-05, "loss": 0.0094, "step": 12550 }, { "grad_norm": 0.12013270705938339, "learning_rate": 3.330446170887566e-05, "loss": 0.0081, "step": 12560 }, { "grad_norm": 0.183862566947937, "learning_rate": 3.3226556146670834e-05, "loss": 0.0112, "step": 12570 }, { "grad_norm": 0.19474278390407562, "learning_rate": 3.314869644242102e-05, "loss": 0.0091, "step": 12580 }, { "grad_norm": 0.18900476396083832, "learning_rate": 3.3070882808991674e-05, "loss": 0.0087, "step": 12590 }, { "grad_norm": 0.1760052591562271, "learning_rate": 3.2993115459122305e-05, "loss": 0.0089, "step": 12600 }, { "grad_norm": 0.16566202044487, "learning_rate": 3.2915394605425835e-05, "loss": 0.0085, "step": 12610 }, { "grad_norm": 0.1798708289861679, "learning_rate": 3.283772046038816e-05, "loss": 0.0075, "step": 12620 }, { "grad_norm": 0.18843935430049896, "learning_rate": 3.276009323636739e-05, "loss": 0.0101, "step": 12630 }, { "grad_norm": 0.14171066880226135, "learning_rate": 3.268251314559344e-05, "loss": 0.0075, "step": 12640 }, { "grad_norm": 0.17853738367557526, "learning_rate": 3.2604980400167254e-05, "loss": 0.0081, "step": 12650 }, { "grad_norm": 0.2422345131635666, "learning_rate": 3.252749521206042e-05, "loss": 0.0094, "step": 12660 }, { "grad_norm": 0.2428761124610901, "learning_rate": 3.2450057793114494e-05, "loss": 0.0109, "step": 12670 }, { "grad_norm": 0.1619350016117096, "learning_rate": 3.2372668355040435e-05, "loss": 0.0089, "step": 12680 }, { "grad_norm": 0.1933186799287796, "learning_rate": 3.2295327109418005e-05, "loss": 0.0127, "step": 12690 }, { "grad_norm": 0.28112977743148804, "learning_rate": 3.221803426769518e-05, "loss": 0.0137, "step": 12700 }, { "grad_norm": 0.1917015165090561, "learning_rate": 3.214079004118768e-05, "loss": 0.0092, "step": 12710 }, { "grad_norm": 0.23235733807086945, "learning_rate": 3.2063594641078234e-05, "loss": 0.0101, "step": 12720 }, { "grad_norm": 0.21012669801712036, "learning_rate": 3.198644827841616e-05, "loss": 0.0097, "step": 12730 }, { "grad_norm": 0.20292937755584717, "learning_rate": 3.1909351164116654e-05, "loss": 0.0096, "step": 12740 }, { "grad_norm": 0.20526261627674103, "learning_rate": 3.183230350896026e-05, "loss": 0.0098, "step": 12750 }, { "grad_norm": 0.1463351994752884, "learning_rate": 3.1755305523592337e-05, "loss": 0.0087, "step": 12760 }, { "grad_norm": 0.15577419102191925, "learning_rate": 3.167835741852245e-05, "loss": 0.0075, "step": 12770 }, { "grad_norm": 0.18636691570281982, "learning_rate": 3.160145940412378e-05, "loss": 0.0131, "step": 12780 }, { "grad_norm": 0.24122855067253113, "learning_rate": 3.1524611690632545e-05, "loss": 0.0115, "step": 12790 }, { "grad_norm": 0.21856479346752167, "learning_rate": 3.144781448814746e-05, "loss": 0.0114, "step": 12800 }, { "grad_norm": 0.19949841499328613, "learning_rate": 3.1371068006629145e-05, "loss": 0.0117, "step": 12810 }, { "grad_norm": 0.18744981288909912, "learning_rate": 3.129437245589956e-05, "loss": 0.0107, "step": 12820 }, { "grad_norm": 0.1762298196554184, "learning_rate": 3.121772804564143e-05, "loss": 0.0104, "step": 12830 }, { "grad_norm": 0.17405322194099426, "learning_rate": 3.11411349853976e-05, "loss": 0.0098, "step": 12840 }, { "grad_norm": 0.14769820868968964, "learning_rate": 3.10645934845706e-05, "loss": 0.0083, "step": 12850 }, { "grad_norm": 0.15435974299907684, "learning_rate": 3.098810375242196e-05, "loss": 0.0093, "step": 12860 }, { "grad_norm": 0.20624206960201263, "learning_rate": 3.0911665998071704e-05, "loss": 0.0082, "step": 12870 }, { "grad_norm": 0.17857100069522858, "learning_rate": 3.083528043049774e-05, "loss": 0.0089, "step": 12880 }, { "grad_norm": 0.3856310248374939, "learning_rate": 3.0758947258535255e-05, "loss": 0.0084, "step": 12890 }, { "grad_norm": 0.17701870203018188, "learning_rate": 3.068266669087625e-05, "loss": 0.0079, "step": 12900 }, { "grad_norm": 0.22868800163269043, "learning_rate": 3.060643893606887e-05, "loss": 0.0088, "step": 12910 }, { "grad_norm": 0.1557961255311966, "learning_rate": 3.053026420251693e-05, "loss": 0.0089, "step": 12920 }, { "grad_norm": 0.1718345582485199, "learning_rate": 3.0454142698479183e-05, "loss": 0.0085, "step": 12930 }, { "grad_norm": 0.17350423336029053, "learning_rate": 3.0378074632068954e-05, "loss": 0.0072, "step": 12940 }, { "grad_norm": 0.1360691636800766, "learning_rate": 3.0302060211253408e-05, "loss": 0.0076, "step": 12950 }, { "grad_norm": 0.2300570160150528, "learning_rate": 3.0226099643853073e-05, "loss": 0.0085, "step": 12960 }, { "grad_norm": 0.23320387303829193, "learning_rate": 3.0150193137541283e-05, "loss": 0.0087, "step": 12970 }, { "grad_norm": 0.15796206891536713, "learning_rate": 3.0074340899843467e-05, "loss": 0.0095, "step": 12980 }, { "grad_norm": 0.1602860689163208, "learning_rate": 2.999854313813677e-05, "loss": 0.0092, "step": 12990 }, { "grad_norm": 0.27343568205833435, "learning_rate": 2.9922800059649382e-05, "loss": 0.0096, "step": 13000 }, { "grad_norm": 0.1662762612104416, "learning_rate": 2.9847111871459976e-05, "loss": 0.0103, "step": 13010 }, { "grad_norm": 0.3069351017475128, "learning_rate": 2.977147878049721e-05, "loss": 0.0097, "step": 13020 }, { "grad_norm": 0.13450488448143005, "learning_rate": 2.9695900993539006e-05, "loss": 0.0088, "step": 13030 }, { "grad_norm": 0.1111426055431366, "learning_rate": 2.9620378717212183e-05, "loss": 0.0074, "step": 13040 }, { "grad_norm": 0.18716482818126678, "learning_rate": 2.9544912157991745e-05, "loss": 0.0081, "step": 13050 }, { "grad_norm": 0.18903988599777222, "learning_rate": 2.9469501522200405e-05, "loss": 0.0099, "step": 13060 }, { "grad_norm": 0.2508498430252075, "learning_rate": 2.9394147016007946e-05, "loss": 0.0073, "step": 13070 }, { "grad_norm": 0.18244940042495728, "learning_rate": 2.9318848845430702e-05, "loss": 0.0086, "step": 13080 }, { "grad_norm": 0.11170797795057297, "learning_rate": 2.9243607216331013e-05, "loss": 0.0092, "step": 13090 }, { "grad_norm": 0.1738201081752777, "learning_rate": 2.916842233441661e-05, "loss": 0.0088, "step": 13100 }, { "grad_norm": 0.13061141967773438, "learning_rate": 2.90932944052401e-05, "loss": 0.008, "step": 13110 }, { "grad_norm": 0.21040372550487518, "learning_rate": 2.9018223634198354e-05, "loss": 0.0097, "step": 13120 }, { "grad_norm": 0.21341542899608612, "learning_rate": 2.8943210226532025e-05, "loss": 0.0108, "step": 13130 }, { "grad_norm": 0.1655251532793045, "learning_rate": 2.8868254387324857e-05, "loss": 0.0078, "step": 13140 }, { "grad_norm": 0.19068247079849243, "learning_rate": 2.8793356321503306e-05, "loss": 0.0083, "step": 13150 }, { "grad_norm": 0.11730074882507324, "learning_rate": 2.87185162338358e-05, "loss": 0.0072, "step": 13160 }, { "grad_norm": 0.15405382215976715, "learning_rate": 2.8643734328932253e-05, "loss": 0.0078, "step": 13170 }, { "grad_norm": 0.15192165970802307, "learning_rate": 2.856901081124359e-05, "loss": 0.0084, "step": 13180 }, { "grad_norm": 0.1417967528104782, "learning_rate": 2.8494345885061002e-05, "loss": 0.0076, "step": 13190 }, { "grad_norm": 0.21760909259319305, "learning_rate": 2.8419739754515616e-05, "loss": 0.0102, "step": 13200 }, { "grad_norm": 0.20578059554100037, "learning_rate": 2.8345192623577666e-05, "loss": 0.009, "step": 13210 }, { "grad_norm": 0.1615130603313446, "learning_rate": 2.8270704696056193e-05, "loss": 0.0081, "step": 13220 }, { "grad_norm": 0.21402151882648468, "learning_rate": 2.8196276175598367e-05, "loss": 0.0098, "step": 13230 }, { "grad_norm": 0.14638644456863403, "learning_rate": 2.8121907265688884e-05, "loss": 0.0115, "step": 13240 }, { "grad_norm": 0.3924233317375183, "learning_rate": 2.804759816964957e-05, "loss": 0.0099, "step": 13250 }, { "grad_norm": 0.1542339324951172, "learning_rate": 2.797334909063857e-05, "loss": 0.009, "step": 13260 }, { "grad_norm": 0.23849767446517944, "learning_rate": 2.7899160231650056e-05, "loss": 0.0105, "step": 13270 }, { "grad_norm": 0.15615037083625793, "learning_rate": 2.7825031795513585e-05, "loss": 0.0086, "step": 13280 }, { "grad_norm": 0.16286875307559967, "learning_rate": 2.775096398489341e-05, "loss": 0.0122, "step": 13290 }, { "grad_norm": 0.1524781584739685, "learning_rate": 2.7676957002288163e-05, "loss": 0.0106, "step": 13300 }, { "grad_norm": 0.09859930723905563, "learning_rate": 2.760301105003003e-05, "loss": 0.0069, "step": 13310 }, { "grad_norm": 0.184215247631073, "learning_rate": 2.752912633028446e-05, "loss": 0.0086, "step": 13320 }, { "grad_norm": 0.213889017701149, "learning_rate": 2.7455303045049474e-05, "loss": 0.0097, "step": 13330 }, { "grad_norm": 0.14843671023845673, "learning_rate": 2.7381541396155098e-05, "loss": 0.008, "step": 13340 }, { "grad_norm": 0.18868941068649292, "learning_rate": 2.730784158526286e-05, "loss": 0.0072, "step": 13350 }, { "grad_norm": 0.1814301460981369, "learning_rate": 2.723420381386521e-05, "loss": 0.0089, "step": 13360 }, { "grad_norm": 0.18372131884098053, "learning_rate": 2.7160628283285018e-05, "loss": 0.0079, "step": 13370 }, { "grad_norm": 0.152345210313797, "learning_rate": 2.7087115194675007e-05, "loss": 0.0084, "step": 13380 }, { "grad_norm": 0.11616712808609009, "learning_rate": 2.701366474901712e-05, "loss": 0.0088, "step": 13390 }, { "grad_norm": 0.1265774965286255, "learning_rate": 2.6940277147122085e-05, "loss": 0.0063, "step": 13400 }, { "grad_norm": 0.23046642541885376, "learning_rate": 2.686695258962878e-05, "loss": 0.0072, "step": 13410 }, { "grad_norm": 0.17843882739543915, "learning_rate": 2.679369127700375e-05, "loss": 0.0093, "step": 13420 }, { "grad_norm": 0.12012670189142227, "learning_rate": 2.672049340954067e-05, "loss": 0.0087, "step": 13430 }, { "grad_norm": 0.14808817207813263, "learning_rate": 2.6647359187359676e-05, "loss": 0.0127, "step": 13440 }, { "grad_norm": 0.1583370417356491, "learning_rate": 2.6574288810406946e-05, "loss": 0.0083, "step": 13450 }, { "grad_norm": 0.1791803389787674, "learning_rate": 2.6501282478454083e-05, "loss": 0.0092, "step": 13460 }, { "grad_norm": 0.19236159324645996, "learning_rate": 2.6428340391097618e-05, "loss": 0.0108, "step": 13470 }, { "grad_norm": 0.10624723136425018, "learning_rate": 2.6355462747758485e-05, "loss": 0.0078, "step": 13480 }, { "grad_norm": 0.17199254035949707, "learning_rate": 2.6282649747681304e-05, "loss": 0.0118, "step": 13490 }, { "grad_norm": 0.16279897093772888, "learning_rate": 2.620990158993406e-05, "loss": 0.0111, "step": 13500 }, { "grad_norm": 0.1628580093383789, "learning_rate": 2.6137218473407477e-05, "loss": 0.0112, "step": 13510 }, { "grad_norm": 0.13751785457134247, "learning_rate": 2.606460059681436e-05, "loss": 0.0067, "step": 13520 }, { "grad_norm": 0.15811248123645782, "learning_rate": 2.599204815868928e-05, "loss": 0.0082, "step": 13530 }, { "grad_norm": 0.19726982712745667, "learning_rate": 2.5919561357387756e-05, "loss": 0.0127, "step": 13540 }, { "grad_norm": 0.4102162718772888, "learning_rate": 2.5847140391085972e-05, "loss": 0.0112, "step": 13550 }, { "grad_norm": 0.2823683023452759, "learning_rate": 2.5774785457780103e-05, "loss": 0.0107, "step": 13560 }, { "grad_norm": 0.1748407483100891, "learning_rate": 2.5702496755285753e-05, "loss": 0.0076, "step": 13570 }, { "grad_norm": 0.17432726919651031, "learning_rate": 2.5630274481237483e-05, "loss": 0.0133, "step": 13580 }, { "grad_norm": 0.13792356848716736, "learning_rate": 2.5558118833088197e-05, "loss": 0.0092, "step": 13590 }, { "grad_norm": 0.12939408421516418, "learning_rate": 2.548603000810872e-05, "loss": 0.0104, "step": 13600 }, { "grad_norm": 0.17408573627471924, "learning_rate": 2.5414008203387152e-05, "loss": 0.0088, "step": 13610 }, { "grad_norm": 0.11299771815538406, "learning_rate": 2.534205361582834e-05, "loss": 0.008, "step": 13620 }, { "grad_norm": 0.20833835005760193, "learning_rate": 2.527016644215338e-05, "loss": 0.009, "step": 13630 }, { "grad_norm": 0.11348573863506317, "learning_rate": 2.519834687889905e-05, "loss": 0.0091, "step": 13640 }, { "grad_norm": 0.15345452725887299, "learning_rate": 2.5126595122417295e-05, "loss": 0.0076, "step": 13650 }, { "grad_norm": 0.2562877833843231, "learning_rate": 2.5054911368874713e-05, "loss": 0.0099, "step": 13660 }, { "grad_norm": 0.16311323642730713, "learning_rate": 2.4983295814251916e-05, "loss": 0.0073, "step": 13670 }, { "grad_norm": 0.20037773251533508, "learning_rate": 2.4911748654343105e-05, "loss": 0.0104, "step": 13680 }, { "grad_norm": 0.23521381616592407, "learning_rate": 2.4840270084755463e-05, "loss": 0.0075, "step": 13690 }, { "grad_norm": 0.16355225443840027, "learning_rate": 2.4768860300908685e-05, "loss": 0.0079, "step": 13700 }, { "grad_norm": 0.1877547651529312, "learning_rate": 2.469751949803443e-05, "loss": 0.0095, "step": 13710 }, { "grad_norm": 0.1839171200990677, "learning_rate": 2.4626247871175666e-05, "loss": 0.0065, "step": 13720 }, { "grad_norm": 0.2924734950065613, "learning_rate": 2.4555045615186346e-05, "loss": 0.0077, "step": 13730 }, { "grad_norm": 0.1452537178993225, "learning_rate": 2.4483912924730677e-05, "loss": 0.0076, "step": 13740 }, { "grad_norm": 0.1971864402294159, "learning_rate": 2.4412849994282742e-05, "loss": 0.0103, "step": 13750 }, { "grad_norm": 0.22642765939235687, "learning_rate": 2.434185701812592e-05, "loss": 0.007, "step": 13760 }, { "grad_norm": 0.18533912301063538, "learning_rate": 2.4270934190352218e-05, "loss": 0.0071, "step": 13770 }, { "grad_norm": 0.18679581582546234, "learning_rate": 2.4200081704861998e-05, "loss": 0.0067, "step": 13780 }, { "grad_norm": 0.15628451108932495, "learning_rate": 2.412929975536321e-05, "loss": 0.0084, "step": 13790 }, { "grad_norm": 0.27595072984695435, "learning_rate": 2.4058588535371017e-05, "loss": 0.0094, "step": 13800 }, { "grad_norm": 0.21139909327030182, "learning_rate": 2.3987948238207243e-05, "loss": 0.0104, "step": 13810 }, { "grad_norm": 0.16834287345409393, "learning_rate": 2.3917379056999678e-05, "loss": 0.0072, "step": 13820 }, { "grad_norm": 0.17054033279418945, "learning_rate": 2.3846881184681824e-05, "loss": 0.0078, "step": 13830 }, { "grad_norm": 0.20645827054977417, "learning_rate": 2.377645481399214e-05, "loss": 0.0078, "step": 13840 }, { "grad_norm": 0.1429230123758316, "learning_rate": 2.3706100137473667e-05, "loss": 0.0075, "step": 13850 }, { "grad_norm": 0.16687805950641632, "learning_rate": 2.3635817347473394e-05, "loss": 0.0092, "step": 13860 }, { "grad_norm": 0.17313875257968903, "learning_rate": 2.3565606636141757e-05, "loss": 0.0073, "step": 13870 }, { "grad_norm": 0.10987868905067444, "learning_rate": 2.3495468195432203e-05, "loss": 0.0067, "step": 13880 }, { "grad_norm": 0.18193338811397552, "learning_rate": 2.3425402217100507e-05, "loss": 0.0081, "step": 13890 }, { "grad_norm": 0.29478874802589417, "learning_rate": 2.3355408892704424e-05, "loss": 0.0095, "step": 13900 }, { "grad_norm": 0.17304958403110504, "learning_rate": 2.3285488413603003e-05, "loss": 0.0068, "step": 13910 }, { "grad_norm": 0.15522640943527222, "learning_rate": 2.321564097095615e-05, "loss": 0.0073, "step": 13920 }, { "grad_norm": 0.13994340598583221, "learning_rate": 2.3145866755724142e-05, "loss": 0.0065, "step": 13930 }, { "grad_norm": 0.12542636692523956, "learning_rate": 2.307616595866699e-05, "loss": 0.0096, "step": 13940 }, { "grad_norm": 0.32555049657821655, "learning_rate": 2.3006538770344032e-05, "loss": 0.0085, "step": 13950 }, { "grad_norm": 0.11737692356109619, "learning_rate": 2.293698538111334e-05, "loss": 0.0079, "step": 13960 }, { "grad_norm": 0.20595699548721313, "learning_rate": 2.28675059811312e-05, "loss": 0.0071, "step": 13970 }, { "grad_norm": 0.11940669268369675, "learning_rate": 2.279810076035167e-05, "loss": 0.0075, "step": 13980 }, { "grad_norm": 0.16453303396701813, "learning_rate": 2.272876990852596e-05, "loss": 0.0068, "step": 13990 }, { "grad_norm": 0.12166702002286911, "learning_rate": 2.265951361520195e-05, "loss": 0.0063, "step": 14000 }, { "grad_norm": 0.20005439221858978, "learning_rate": 2.2590332069723748e-05, "loss": 0.0071, "step": 14010 }, { "grad_norm": 0.16831733286380768, "learning_rate": 2.2521225461231004e-05, "loss": 0.0081, "step": 14020 }, { "grad_norm": 0.10804498195648193, "learning_rate": 2.2452193978658597e-05, "loss": 0.0074, "step": 14030 }, { "grad_norm": 0.19373473525047302, "learning_rate": 2.238323781073594e-05, "loss": 0.0072, "step": 14040 }, { "grad_norm": 0.1723291575908661, "learning_rate": 2.2314357145986552e-05, "loss": 0.0061, "step": 14050 }, { "grad_norm": 0.1751866340637207, "learning_rate": 2.224555217272757e-05, "loss": 0.0064, "step": 14060 }, { "grad_norm": 0.19218294322490692, "learning_rate": 2.2176823079069127e-05, "loss": 0.0079, "step": 14070 }, { "grad_norm": 0.3986702561378479, "learning_rate": 2.210817005291398e-05, "loss": 0.0072, "step": 14080 }, { "grad_norm": 0.2290142923593521, "learning_rate": 2.203959328195686e-05, "loss": 0.007, "step": 14090 }, { "grad_norm": 0.26604822278022766, "learning_rate": 2.1971092953684026e-05, "loss": 0.0085, "step": 14100 }, { "grad_norm": 0.16322878003120422, "learning_rate": 2.1902669255372788e-05, "loss": 0.006, "step": 14110 }, { "grad_norm": 0.2603505849838257, "learning_rate": 2.1834322374090897e-05, "loss": 0.0078, "step": 14120 }, { "grad_norm": 0.18867948651313782, "learning_rate": 2.1766052496696153e-05, "loss": 0.0091, "step": 14130 }, { "grad_norm": 0.16376566886901855, "learning_rate": 2.169785980983577e-05, "loss": 0.0092, "step": 14140 }, { "grad_norm": 0.20239832997322083, "learning_rate": 2.162974449994593e-05, "loss": 0.0071, "step": 14150 }, { "grad_norm": 0.1524510234594345, "learning_rate": 2.1561706753251337e-05, "loss": 0.0073, "step": 14160 }, { "grad_norm": 0.1411779373884201, "learning_rate": 2.1493746755764544e-05, "loss": 0.008, "step": 14170 }, { "grad_norm": 0.1301325410604477, "learning_rate": 2.1425864693285635e-05, "loss": 0.0067, "step": 14180 }, { "grad_norm": 0.15414534509181976, "learning_rate": 2.1358060751401547e-05, "loss": 0.0068, "step": 14190 }, { "grad_norm": 0.20321588218212128, "learning_rate": 2.129033511548566e-05, "loss": 0.007, "step": 14200 }, { "grad_norm": 0.24943380057811737, "learning_rate": 2.1222687970697315e-05, "loss": 0.0074, "step": 14210 }, { "grad_norm": 0.21284577250480652, "learning_rate": 2.1155119501981173e-05, "loss": 0.0077, "step": 14220 }, { "grad_norm": 0.16672813892364502, "learning_rate": 2.1087629894066895e-05, "loss": 0.0085, "step": 14230 }, { "grad_norm": 0.18013784289360046, "learning_rate": 2.1020219331468473e-05, "loss": 0.0068, "step": 14240 }, { "grad_norm": 0.17006142437458038, "learning_rate": 2.095288799848379e-05, "loss": 0.0076, "step": 14250 }, { "grad_norm": 0.17850056290626526, "learning_rate": 2.088563607919417e-05, "loss": 0.009, "step": 14260 }, { "grad_norm": 0.1445813924074173, "learning_rate": 2.0818463757463786e-05, "loss": 0.0069, "step": 14270 }, { "grad_norm": 0.14390017092227936, "learning_rate": 2.0751371216939175e-05, "loss": 0.0074, "step": 14280 }, { "grad_norm": 0.10625940561294556, "learning_rate": 2.068435864104882e-05, "loss": 0.0056, "step": 14290 }, { "grad_norm": 0.09692031890153885, "learning_rate": 2.0617426213002506e-05, "loss": 0.007, "step": 14300 }, { "grad_norm": 0.16112850606441498, "learning_rate": 2.055057411579097e-05, "loss": 0.0058, "step": 14310 }, { "grad_norm": 0.15632018446922302, "learning_rate": 2.0483802532185286e-05, "loss": 0.0092, "step": 14320 }, { "grad_norm": 0.15759548544883728, "learning_rate": 2.041711164473638e-05, "loss": 0.0083, "step": 14330 }, { "grad_norm": 0.2144136130809784, "learning_rate": 2.0350501635774637e-05, "loss": 0.0088, "step": 14340 }, { "grad_norm": 0.13073790073394775, "learning_rate": 2.0283972687409247e-05, "loss": 0.0079, "step": 14350 }, { "grad_norm": 0.27075785398483276, "learning_rate": 2.021752498152784e-05, "loss": 0.0115, "step": 14360 }, { "grad_norm": 0.14081735908985138, "learning_rate": 2.015115869979589e-05, "loss": 0.0086, "step": 14370 }, { "grad_norm": 0.20877787470817566, "learning_rate": 2.0084874023656265e-05, "loss": 0.0084, "step": 14380 }, { "grad_norm": 0.1538696438074112, "learning_rate": 2.001867113432877e-05, "loss": 0.0081, "step": 14390 }, { "grad_norm": 0.1591489166021347, "learning_rate": 1.995255021280954e-05, "loss": 0.0074, "step": 14400 }, { "grad_norm": 0.19511725008487701, "learning_rate": 1.9886511439870688e-05, "loss": 0.008, "step": 14410 }, { "grad_norm": 0.13154065608978271, "learning_rate": 1.9820554996059675e-05, "loss": 0.0079, "step": 14420 }, { "grad_norm": 0.152265727519989, "learning_rate": 1.9754681061698893e-05, "loss": 0.0061, "step": 14430 }, { "grad_norm": 0.17777813971042633, "learning_rate": 1.9688889816885185e-05, "loss": 0.0068, "step": 14440 }, { "grad_norm": 0.21932631731033325, "learning_rate": 1.962318144148928e-05, "loss": 0.0097, "step": 14450 }, { "grad_norm": 0.12777146697044373, "learning_rate": 1.955755611515539e-05, "loss": 0.0068, "step": 14460 }, { "grad_norm": 0.1586160510778427, "learning_rate": 1.9492014017300642e-05, "loss": 0.0084, "step": 14470 }, { "grad_norm": 0.171687513589859, "learning_rate": 1.942655532711461e-05, "loss": 0.0103, "step": 14480 }, { "grad_norm": 0.20121882855892181, "learning_rate": 1.9361180223558882e-05, "loss": 0.0067, "step": 14490 }, { "grad_norm": 0.15869708359241486, "learning_rate": 1.929588888536647e-05, "loss": 0.005, "step": 14500 }, { "grad_norm": 0.17405566573143005, "learning_rate": 1.9230681491041425e-05, "loss": 0.0066, "step": 14510 }, { "grad_norm": 0.23950868844985962, "learning_rate": 1.9165558218858264e-05, "loss": 0.0062, "step": 14520 }, { "grad_norm": 0.14754101634025574, "learning_rate": 1.9100519246861505e-05, "loss": 0.007, "step": 14530 }, { "grad_norm": 0.1627737432718277, "learning_rate": 1.9035564752865248e-05, "loss": 0.0079, "step": 14540 }, { "grad_norm": 0.16454444825649261, "learning_rate": 1.897069491445258e-05, "loss": 0.0082, "step": 14550 }, { "grad_norm": 0.18162935972213745, "learning_rate": 1.890590990897515e-05, "loss": 0.0071, "step": 14560 }, { "grad_norm": 0.24299250543117523, "learning_rate": 1.884120991355272e-05, "loss": 0.0084, "step": 14570 }, { "grad_norm": 0.16916567087173462, "learning_rate": 1.8776595105072576e-05, "loss": 0.0089, "step": 14580 }, { "grad_norm": 0.2002318650484085, "learning_rate": 1.8712065660189166e-05, "loss": 0.0075, "step": 14590 }, { "grad_norm": 0.15641240775585175, "learning_rate": 1.8647621755323513e-05, "loss": 0.0061, "step": 14600 }, { "grad_norm": 0.1357610672712326, "learning_rate": 1.858326356666278e-05, "loss": 0.0079, "step": 14610 }, { "grad_norm": 0.1986888349056244, "learning_rate": 1.851899127015983e-05, "loss": 0.0068, "step": 14620 }, { "grad_norm": 0.15061034262180328, "learning_rate": 1.8454805041532626e-05, "loss": 0.01, "step": 14630 }, { "grad_norm": 0.17367121577262878, "learning_rate": 1.8390705056263906e-05, "loss": 0.0078, "step": 14640 }, { "grad_norm": 0.15947949886322021, "learning_rate": 1.832669148960057e-05, "loss": 0.0069, "step": 14650 }, { "grad_norm": 0.12651796638965607, "learning_rate": 1.8262764516553233e-05, "loss": 0.006, "step": 14660 }, { "grad_norm": 0.24611006677150726, "learning_rate": 1.8198924311895843e-05, "loss": 0.0056, "step": 14670 }, { "grad_norm": 0.16382458806037903, "learning_rate": 1.813517105016505e-05, "loss": 0.0073, "step": 14680 }, { "grad_norm": 0.1076081246137619, "learning_rate": 1.8071504905659888e-05, "loss": 0.006, "step": 14690 }, { "grad_norm": 0.10948217660188675, "learning_rate": 1.800792605244109e-05, "loss": 0.0068, "step": 14700 }, { "grad_norm": 0.09620071947574615, "learning_rate": 1.7944434664330844e-05, "loss": 0.0116, "step": 14710 }, { "grad_norm": 0.2767171263694763, "learning_rate": 1.7881030914912212e-05, "loss": 0.0069, "step": 14720 }, { "grad_norm": 0.1795174926519394, "learning_rate": 1.7817714977528577e-05, "loss": 0.007, "step": 14730 }, { "grad_norm": 0.28042730689048767, "learning_rate": 1.7754487025283332e-05, "loss": 0.0101, "step": 14740 }, { "grad_norm": 0.15795153379440308, "learning_rate": 1.7691347231039275e-05, "loss": 0.008, "step": 14750 }, { "grad_norm": 0.09051381796598434, "learning_rate": 1.7628295767418164e-05, "loss": 0.0086, "step": 14760 }, { "grad_norm": 0.275169312953949, "learning_rate": 1.7565332806800333e-05, "loss": 0.0096, "step": 14770 }, { "grad_norm": 0.20493553578853607, "learning_rate": 1.750245852132408e-05, "loss": 0.0076, "step": 14780 }, { "grad_norm": 0.1768176108598709, "learning_rate": 1.7439673082885323e-05, "loss": 0.0092, "step": 14790 }, { "grad_norm": 0.18474966287612915, "learning_rate": 1.7376976663137047e-05, "loss": 0.0056, "step": 14800 }, { "grad_norm": 0.21288339793682098, "learning_rate": 1.7314369433488853e-05, "loss": 0.0066, "step": 14810 }, { "grad_norm": 0.19863519072532654, "learning_rate": 1.7251851565106548e-05, "loss": 0.008, "step": 14820 }, { "grad_norm": 0.15599465370178223, "learning_rate": 1.7189423228911574e-05, "loss": 0.007, "step": 14830 }, { "grad_norm": 0.30430108308792114, "learning_rate": 1.7127084595580606e-05, "loss": 0.0073, "step": 14840 }, { "grad_norm": 0.14536628127098083, "learning_rate": 1.706483583554513e-05, "loss": 0.0062, "step": 14850 }, { "grad_norm": 0.1288166046142578, "learning_rate": 1.700267711899083e-05, "loss": 0.0062, "step": 14860 }, { "grad_norm": 0.2007361501455307, "learning_rate": 1.69406086158573e-05, "loss": 0.0061, "step": 14870 }, { "grad_norm": 0.1578177958726883, "learning_rate": 1.6878630495837455e-05, "loss": 0.006, "step": 14880 }, { "grad_norm": 0.09969499707221985, "learning_rate": 1.681674292837707e-05, "loss": 0.0078, "step": 14890 }, { "grad_norm": 0.10585860908031464, "learning_rate": 1.6754946082674444e-05, "loss": 0.006, "step": 14900 }, { "grad_norm": 0.2749101221561432, "learning_rate": 1.6693240127679748e-05, "loss": 0.0056, "step": 14910 }, { "grad_norm": 0.19092895090579987, "learning_rate": 1.663162523209475e-05, "loss": 0.0086, "step": 14920 }, { "grad_norm": 0.1306796371936798, "learning_rate": 1.6570101564372193e-05, "loss": 0.0072, "step": 14930 }, { "grad_norm": 0.20014528930187225, "learning_rate": 1.650866929271543e-05, "loss": 0.0071, "step": 14940 }, { "grad_norm": 0.1997898668050766, "learning_rate": 1.644732858507797e-05, "loss": 0.0072, "step": 14950 }, { "grad_norm": 0.16324928402900696, "learning_rate": 1.6386079609162943e-05, "loss": 0.0057, "step": 14960 }, { "grad_norm": 0.1201007068157196, "learning_rate": 1.6324922532422742e-05, "loss": 0.0049, "step": 14970 }, { "grad_norm": 0.12609346210956573, "learning_rate": 1.6263857522058434e-05, "loss": 0.0057, "step": 14980 }, { "grad_norm": 0.1812009960412979, "learning_rate": 1.6202884745019443e-05, "loss": 0.0077, "step": 14990 }, { "grad_norm": 0.17515036463737488, "learning_rate": 1.614200436800304e-05, "loss": 0.0066, "step": 15000 }, { "grad_norm": 0.28439024090766907, "learning_rate": 1.6081216557453814e-05, "loss": 0.0065, "step": 15010 }, { "grad_norm": 0.1405584067106247, "learning_rate": 1.6020521479563367e-05, "loss": 0.005, "step": 15020 }, { "grad_norm": 0.14033935964107513, "learning_rate": 1.5959919300269654e-05, "loss": 0.0071, "step": 15030 }, { "grad_norm": 0.15945085883140564, "learning_rate": 1.5899410185256764e-05, "loss": 0.007, "step": 15040 }, { "grad_norm": 0.1411488801240921, "learning_rate": 1.583899429995431e-05, "loss": 0.0066, "step": 15050 }, { "grad_norm": 0.13914041221141815, "learning_rate": 1.5778671809536993e-05, "loss": 0.0065, "step": 15060 }, { "grad_norm": 0.18932394683361053, "learning_rate": 1.5718442878924246e-05, "loss": 0.0065, "step": 15070 }, { "grad_norm": 0.15020237863063812, "learning_rate": 1.5658307672779593e-05, "loss": 0.0055, "step": 15080 }, { "grad_norm": 0.1578770875930786, "learning_rate": 1.5598266355510427e-05, "loss": 0.0076, "step": 15090 }, { "grad_norm": 0.14462171494960785, "learning_rate": 1.553831909126744e-05, "loss": 0.0061, "step": 15100 }, { "grad_norm": 0.1438542604446411, "learning_rate": 1.5478466043944135e-05, "loss": 0.008, "step": 15110 }, { "grad_norm": 0.13986705243587494, "learning_rate": 1.5418707377176468e-05, "loss": 0.0083, "step": 15120 }, { "grad_norm": 0.19286397099494934, "learning_rate": 1.535904325434233e-05, "loss": 0.007, "step": 15130 }, { "grad_norm": 0.11769445985555649, "learning_rate": 1.529947383856118e-05, "loss": 0.0055, "step": 15140 }, { "grad_norm": 0.10094045847654343, "learning_rate": 1.5239999292693524e-05, "loss": 0.006, "step": 15150 }, { "grad_norm": 0.14341425895690918, "learning_rate": 1.5180619779340505e-05, "loss": 0.0049, "step": 15160 }, { "grad_norm": 0.16170763969421387, "learning_rate": 1.5121335460843428e-05, "loss": 0.0061, "step": 15170 }, { "grad_norm": 0.14601518213748932, "learning_rate": 1.5062146499283347e-05, "loss": 0.0059, "step": 15180 }, { "grad_norm": 0.13407990336418152, "learning_rate": 1.5003053056480643e-05, "loss": 0.0074, "step": 15190 }, { "grad_norm": 0.18061061203479767, "learning_rate": 1.4944055293994551e-05, "loss": 0.0067, "step": 15200 }, { "grad_norm": 0.16719160974025726, "learning_rate": 1.4885153373122656e-05, "loss": 0.0059, "step": 15210 }, { "grad_norm": 0.15071117877960205, "learning_rate": 1.482634745490059e-05, "loss": 0.0059, "step": 15220 }, { "grad_norm": 0.135216623544693, "learning_rate": 1.4767637700101466e-05, "loss": 0.0059, "step": 15230 }, { "grad_norm": 0.135899618268013, "learning_rate": 1.4709024269235528e-05, "loss": 0.0066, "step": 15240 }, { "grad_norm": 0.17345286905765533, "learning_rate": 1.4650507322549684e-05, "loss": 0.0077, "step": 15250 }, { "grad_norm": 0.19219046831130981, "learning_rate": 1.4592087020026972e-05, "loss": 0.005, "step": 15260 }, { "grad_norm": 0.14325112104415894, "learning_rate": 1.4533763521386318e-05, "loss": 0.0066, "step": 15270 }, { "grad_norm": 0.16629421710968018, "learning_rate": 1.44755369860819e-05, "loss": 0.0051, "step": 15280 }, { "grad_norm": 0.1859646737575531, "learning_rate": 1.441740757330287e-05, "loss": 0.0081, "step": 15290 }, { "grad_norm": 0.15992504358291626, "learning_rate": 1.4359375441972844e-05, "loss": 0.0058, "step": 15300 }, { "grad_norm": 0.1612405925989151, "learning_rate": 1.4301440750749395e-05, "loss": 0.0063, "step": 15310 }, { "grad_norm": 0.14462994039058685, "learning_rate": 1.4243603658023808e-05, "loss": 0.0074, "step": 15320 }, { "grad_norm": 0.15201006829738617, "learning_rate": 1.4185864321920444e-05, "loss": 0.0077, "step": 15330 }, { "grad_norm": 0.14782527089118958, "learning_rate": 1.4128222900296485e-05, "loss": 0.007, "step": 15340 }, { "grad_norm": 0.1329716295003891, "learning_rate": 1.407067955074135e-05, "loss": 0.006, "step": 15350 }, { "grad_norm": 0.0698641762137413, "learning_rate": 1.4013234430576356e-05, "loss": 0.0059, "step": 15360 }, { "grad_norm": 0.14514224231243134, "learning_rate": 1.3955887696854286e-05, "loss": 0.0057, "step": 15370 }, { "grad_norm": 0.18291395902633667, "learning_rate": 1.38986395063589e-05, "loss": 0.0061, "step": 15380 }, { "grad_norm": 0.19162975251674652, "learning_rate": 1.3841490015604597e-05, "loss": 0.0091, "step": 15390 }, { "grad_norm": 0.15656042098999023, "learning_rate": 1.3784439380835879e-05, "loss": 0.0071, "step": 15400 }, { "grad_norm": 0.15763820707798004, "learning_rate": 1.3727487758026986e-05, "loss": 0.0056, "step": 15410 }, { "grad_norm": 0.16434554755687714, "learning_rate": 1.3670635302881525e-05, "loss": 0.0087, "step": 15420 }, { "grad_norm": 0.12442274391651154, "learning_rate": 1.3613882170831888e-05, "loss": 0.006, "step": 15430 }, { "grad_norm": 0.1326097995042801, "learning_rate": 1.355722851703901e-05, "loss": 0.0056, "step": 15440 }, { "grad_norm": 0.17795906960964203, "learning_rate": 1.3500674496391814e-05, "loss": 0.0076, "step": 15450 }, { "grad_norm": 0.15047936141490936, "learning_rate": 1.3444220263506795e-05, "loss": 0.0062, "step": 15460 }, { "grad_norm": 0.1731768399477005, "learning_rate": 1.3387865972727714e-05, "loss": 0.0063, "step": 15470 }, { "grad_norm": 0.19428685307502747, "learning_rate": 1.3331611778125036e-05, "loss": 0.0056, "step": 15480 }, { "grad_norm": 0.1581166833639145, "learning_rate": 1.3275457833495564e-05, "loss": 0.007, "step": 15490 }, { "grad_norm": 0.16964828968048096, "learning_rate": 1.3219404292362065e-05, "loss": 0.0061, "step": 15500 }, { "grad_norm": 0.1508956104516983, "learning_rate": 1.3163451307972751e-05, "loss": 0.0053, "step": 15510 }, { "grad_norm": 0.13400442898273468, "learning_rate": 1.3107599033300977e-05, "loss": 0.0059, "step": 15520 }, { "grad_norm": 0.16239358484745026, "learning_rate": 1.305184762104471e-05, "loss": 0.0092, "step": 15530 }, { "grad_norm": 0.2620585262775421, "learning_rate": 1.2996197223626178e-05, "loss": 0.0077, "step": 15540 }, { "grad_norm": 0.20591695606708527, "learning_rate": 1.2940647993191457e-05, "loss": 0.0064, "step": 15550 }, { "grad_norm": 0.11730386316776276, "learning_rate": 1.2885200081610005e-05, "loss": 0.0053, "step": 15560 }, { "grad_norm": 0.1091536357998848, "learning_rate": 1.2829853640474316e-05, "loss": 0.0064, "step": 15570 }, { "grad_norm": 0.15600907802581787, "learning_rate": 1.2774608821099438e-05, "loss": 0.0056, "step": 15580 }, { "grad_norm": 0.15474072098731995, "learning_rate": 1.2719465774522577e-05, "loss": 0.0064, "step": 15590 }, { "grad_norm": 0.23807787895202637, "learning_rate": 1.2664424651502755e-05, "loss": 0.0058, "step": 15600 }, { "grad_norm": 0.14216509461402893, "learning_rate": 1.260948560252026e-05, "loss": 0.0061, "step": 15610 }, { "grad_norm": 0.15795795619487762, "learning_rate": 1.2554648777776396e-05, "loss": 0.0054, "step": 15620 }, { "grad_norm": 0.0934879332780838, "learning_rate": 1.2499914327192919e-05, "loss": 0.0047, "step": 15630 }, { "grad_norm": 0.16751891374588013, "learning_rate": 1.2445282400411722e-05, "loss": 0.0054, "step": 15640 }, { "grad_norm": 0.16713222861289978, "learning_rate": 1.2390753146794437e-05, "loss": 0.0081, "step": 15650 }, { "grad_norm": 0.09406598657369614, "learning_rate": 1.2336326715421925e-05, "loss": 0.0072, "step": 15660 }, { "grad_norm": 0.12645189464092255, "learning_rate": 1.2282003255094005e-05, "loss": 0.0058, "step": 15670 }, { "grad_norm": 0.21070720255374908, "learning_rate": 1.2227782914328928e-05, "loss": 0.0064, "step": 15680 }, { "grad_norm": 0.11470688134431839, "learning_rate": 1.2173665841363018e-05, "loss": 0.0068, "step": 15690 }, { "grad_norm": 0.1499481499195099, "learning_rate": 1.211965218415032e-05, "loss": 0.0048, "step": 15700 }, { "grad_norm": 0.10851385444402695, "learning_rate": 1.2065742090362082e-05, "loss": 0.0061, "step": 15710 }, { "grad_norm": 0.07968513667583466, "learning_rate": 1.2011935707386457e-05, "loss": 0.0058, "step": 15720 }, { "grad_norm": 0.15797434747219086, "learning_rate": 1.1958233182328044e-05, "loss": 0.0048, "step": 15730 }, { "grad_norm": 0.16466160118579865, "learning_rate": 1.1904634662007474e-05, "loss": 0.0054, "step": 15740 }, { "grad_norm": 0.23402433097362518, "learning_rate": 1.1851140292961088e-05, "loss": 0.0048, "step": 15750 }, { "grad_norm": 0.13091973960399628, "learning_rate": 1.1797750221440424e-05, "loss": 0.0061, "step": 15760 }, { "grad_norm": 0.1462915539741516, "learning_rate": 1.1744464593411897e-05, "loss": 0.0059, "step": 15770 }, { "grad_norm": 0.1699388176202774, "learning_rate": 1.1691283554556399e-05, "loss": 0.0068, "step": 15780 }, { "grad_norm": 0.12896619737148285, "learning_rate": 1.1638207250268834e-05, "loss": 0.0065, "step": 15790 }, { "grad_norm": 0.13998913764953613, "learning_rate": 1.158523582565782e-05, "loss": 0.0065, "step": 15800 }, { "grad_norm": 0.13465294241905212, "learning_rate": 1.1532369425545192e-05, "loss": 0.0055, "step": 15810 }, { "grad_norm": 0.15440250933170319, "learning_rate": 1.1479608194465662e-05, "loss": 0.0069, "step": 15820 }, { "grad_norm": 0.12376557290554047, "learning_rate": 1.1426952276666442e-05, "loss": 0.0053, "step": 15830 }, { "grad_norm": 0.13334976136684418, "learning_rate": 1.1374401816106778e-05, "loss": 0.0054, "step": 15840 }, { "grad_norm": 0.11170188337564468, "learning_rate": 1.1321956956457646e-05, "loss": 0.0076, "step": 15850 }, { "grad_norm": 0.12342233210802078, "learning_rate": 1.1269617841101277e-05, "loss": 0.0052, "step": 15860 }, { "grad_norm": 0.12431229650974274, "learning_rate": 1.1217384613130804e-05, "loss": 0.0048, "step": 15870 }, { "grad_norm": 0.22500598430633545, "learning_rate": 1.11652574153499e-05, "loss": 0.0078, "step": 15880 }, { "grad_norm": 0.21914471685886383, "learning_rate": 1.1113236390272303e-05, "loss": 0.0057, "step": 15890 }, { "grad_norm": 0.1556463986635208, "learning_rate": 1.106132168012155e-05, "loss": 0.0074, "step": 15900 }, { "grad_norm": 0.11526963114738464, "learning_rate": 1.1009513426830448e-05, "loss": 0.0078, "step": 15910 }, { "grad_norm": 0.20678268373012543, "learning_rate": 1.0957811772040777e-05, "loss": 0.0062, "step": 15920 }, { "grad_norm": 0.112208291888237, "learning_rate": 1.0906216857102913e-05, "loss": 0.0097, "step": 15930 }, { "grad_norm": 0.19000187516212463, "learning_rate": 1.0854728823075355e-05, "loss": 0.0057, "step": 15940 }, { "grad_norm": 0.15828905999660492, "learning_rate": 1.0803347810724452e-05, "loss": 0.007, "step": 15950 }, { "grad_norm": 0.15289664268493652, "learning_rate": 1.0752073960523911e-05, "loss": 0.0076, "step": 15960 }, { "grad_norm": 0.1876230686903, "learning_rate": 1.070090741265447e-05, "loss": 0.0064, "step": 15970 }, { "grad_norm": 0.17958573997020721, "learning_rate": 1.0649848307003547e-05, "loss": 0.0065, "step": 15980 }, { "grad_norm": 0.1672486662864685, "learning_rate": 1.0598896783164757e-05, "loss": 0.0055, "step": 15990 }, { "grad_norm": 0.13452722132205963, "learning_rate": 1.0548052980437645e-05, "loss": 0.0054, "step": 16000 }, { "grad_norm": 0.15977191925048828, "learning_rate": 1.049731703782722e-05, "loss": 0.0054, "step": 16010 }, { "grad_norm": 0.16575075685977936, "learning_rate": 1.0446689094043587e-05, "loss": 0.0043, "step": 16020 }, { "grad_norm": 0.11192754656076431, "learning_rate": 1.039616928750165e-05, "loss": 0.0054, "step": 16030 }, { "grad_norm": 0.12973608076572418, "learning_rate": 1.0345757756320612e-05, "loss": 0.0058, "step": 16040 }, { "grad_norm": 0.11733056604862213, "learning_rate": 1.0295454638323666e-05, "loss": 0.006, "step": 16050 }, { "grad_norm": 0.12858331203460693, "learning_rate": 1.0245260071037632e-05, "loss": 0.0055, "step": 16060 }, { "grad_norm": 0.27599480748176575, "learning_rate": 1.0195174191692518e-05, "loss": 0.006, "step": 16070 }, { "grad_norm": 0.1324242204427719, "learning_rate": 1.014519713722124e-05, "loss": 0.0059, "step": 16080 }, { "grad_norm": 0.2230958193540573, "learning_rate": 1.0095329044259132e-05, "loss": 0.0058, "step": 16090 }, { "grad_norm": 0.171208918094635, "learning_rate": 1.004557004914365e-05, "loss": 0.0052, "step": 16100 }, { "grad_norm": 0.10891158878803253, "learning_rate": 9.995920287914007e-06, "loss": 0.0057, "step": 16110 }, { "grad_norm": 0.13817209005355835, "learning_rate": 9.946379896310737e-06, "loss": 0.0056, "step": 16120 }, { "grad_norm": 0.14293904602527618, "learning_rate": 9.896949009775396e-06, "loss": 0.005, "step": 16130 }, { "grad_norm": 0.22670146822929382, "learning_rate": 9.847627763450134e-06, "loss": 0.0063, "step": 16140 }, { "grad_norm": 0.1190618947148323, "learning_rate": 9.798416292177337e-06, "loss": 0.0065, "step": 16150 }, { "grad_norm": 0.08713887631893158, "learning_rate": 9.74931473049932e-06, "loss": 0.007, "step": 16160 }, { "grad_norm": 0.2843324840068817, "learning_rate": 9.700323212657847e-06, "loss": 0.0051, "step": 16170 }, { "grad_norm": 0.12646743655204773, "learning_rate": 9.65144187259388e-06, "loss": 0.0079, "step": 16180 }, { "grad_norm": 0.2146284431219101, "learning_rate": 9.602670843947132e-06, "loss": 0.0061, "step": 16190 }, { "grad_norm": 0.17945976555347443, "learning_rate": 9.554010260055713e-06, "loss": 0.007, "step": 16200 }, { "grad_norm": 0.1627890169620514, "learning_rate": 9.505460253955834e-06, "loss": 0.0055, "step": 16210 }, { "grad_norm": 0.22857825458049774, "learning_rate": 9.457020958381324e-06, "loss": 0.0063, "step": 16220 }, { "grad_norm": 0.22079415619373322, "learning_rate": 9.408692505763395e-06, "loss": 0.0065, "step": 16230 }, { "grad_norm": 0.13026170432567596, "learning_rate": 9.360475028230181e-06, "loss": 0.0061, "step": 16240 }, { "grad_norm": 0.1672443151473999, "learning_rate": 9.312368657606412e-06, "loss": 0.0055, "step": 16250 }, { "grad_norm": 0.11845371872186661, "learning_rate": 9.264373525413096e-06, "loss": 0.0069, "step": 16260 }, { "grad_norm": 0.13187696039676666, "learning_rate": 9.216489762867058e-06, "loss": 0.0052, "step": 16270 }, { "grad_norm": 0.1831694394350052, "learning_rate": 9.168717500880708e-06, "loss": 0.0052, "step": 16280 }, { "grad_norm": 0.10183826833963394, "learning_rate": 9.121056870061574e-06, "loss": 0.0066, "step": 16290 }, { "grad_norm": 0.14569179713726044, "learning_rate": 9.073508000711983e-06, "loss": 0.0049, "step": 16300 }, { "grad_norm": 0.18153166770935059, "learning_rate": 9.026071022828758e-06, "loss": 0.0059, "step": 16310 }, { "grad_norm": 0.12913790345191956, "learning_rate": 8.978746066102771e-06, "loss": 0.0051, "step": 16320 }, { "grad_norm": 0.18624036014080048, "learning_rate": 8.931533259918634e-06, "loss": 0.0069, "step": 16330 }, { "grad_norm": 0.1058429479598999, "learning_rate": 8.884432733354382e-06, "loss": 0.0048, "step": 16340 }, { "grad_norm": 0.10995589196681976, "learning_rate": 8.837444615181029e-06, "loss": 0.005, "step": 16350 }, { "grad_norm": 0.14928624033927917, "learning_rate": 8.790569033862323e-06, "loss": 0.0056, "step": 16360 }, { "grad_norm": 0.29094499349594116, "learning_rate": 8.7438061175543e-06, "loss": 0.005, "step": 16370 }, { "grad_norm": 0.12059822678565979, "learning_rate": 8.697155994104978e-06, "loss": 0.0075, "step": 16380 }, { "grad_norm": 0.17588664591312408, "learning_rate": 8.650618791054033e-06, "loss": 0.0066, "step": 16390 }, { "grad_norm": 0.1239367425441742, "learning_rate": 8.604194635632373e-06, "loss": 0.0061, "step": 16400 }, { "grad_norm": 0.14971938729286194, "learning_rate": 8.557883654761906e-06, "loss": 0.0064, "step": 16410 }, { "grad_norm": 0.10807310789823532, "learning_rate": 8.511685975055061e-06, "loss": 0.0045, "step": 16420 }, { "grad_norm": 0.10475943982601166, "learning_rate": 8.46560172281452e-06, "loss": 0.0048, "step": 16430 }, { "grad_norm": 0.11389699578285217, "learning_rate": 8.419631024032893e-06, "loss": 0.0053, "step": 16440 }, { "grad_norm": 0.11631321907043457, "learning_rate": 8.373774004392293e-06, "loss": 0.0047, "step": 16450 }, { "grad_norm": 0.1928280144929886, "learning_rate": 8.32803078926409e-06, "loss": 0.0046, "step": 16460 }, { "grad_norm": 0.20834466814994812, "learning_rate": 8.282401503708454e-06, "loss": 0.0055, "step": 16470 }, { "grad_norm": 0.10038771480321884, "learning_rate": 8.23688627247412e-06, "loss": 0.0046, "step": 16480 }, { "grad_norm": 0.1506493240594864, "learning_rate": 8.191485219998007e-06, "loss": 0.0057, "step": 16490 }, { "grad_norm": 0.15905316174030304, "learning_rate": 8.146198470404843e-06, "loss": 0.0056, "step": 16500 }, { "grad_norm": 0.20463865995407104, "learning_rate": 8.101026147506897e-06, "loss": 0.0051, "step": 16510 }, { "grad_norm": 0.17672953009605408, "learning_rate": 8.05596837480353e-06, "loss": 0.0045, "step": 16520 }, { "grad_norm": 0.0851769894361496, "learning_rate": 8.011025275480998e-06, "loss": 0.0058, "step": 16530 }, { "grad_norm": 0.12177392840385437, "learning_rate": 7.966196972412027e-06, "loss": 0.0052, "step": 16540 }, { "grad_norm": 0.1176517978310585, "learning_rate": 7.92148358815547e-06, "loss": 0.006, "step": 16550 }, { "grad_norm": 0.17061352729797363, "learning_rate": 7.87688524495604e-06, "loss": 0.0051, "step": 16560 }, { "grad_norm": 0.13741186261177063, "learning_rate": 7.83240206474386e-06, "loss": 0.0058, "step": 16570 }, { "grad_norm": 0.15751081705093384, "learning_rate": 7.788034169134272e-06, "loss": 0.0057, "step": 16580 }, { "grad_norm": 0.15973970293998718, "learning_rate": 7.743781679427414e-06, "loss": 0.0052, "step": 16590 }, { "grad_norm": 0.1201474517583847, "learning_rate": 7.699644716607895e-06, "loss": 0.005, "step": 16600 }, { "grad_norm": 0.15340934693813324, "learning_rate": 7.655623401344486e-06, "loss": 0.0053, "step": 16610 }, { "grad_norm": 0.10896404087543488, "learning_rate": 7.611717853989775e-06, "loss": 0.005, "step": 16620 }, { "grad_norm": 0.11655887961387634, "learning_rate": 7.567928194579854e-06, "loss": 0.005, "step": 16630 }, { "grad_norm": 0.14782212674617767, "learning_rate": 7.524254542833997e-06, "loss": 0.0042, "step": 16640 }, { "grad_norm": 0.07710693031549454, "learning_rate": 7.480697018154286e-06, "loss": 0.005, "step": 16650 }, { "grad_norm": 0.26042982935905457, "learning_rate": 7.437255739625332e-06, "loss": 0.0058, "step": 16660 }, { "grad_norm": 0.20510970056056976, "learning_rate": 7.393930826013923e-06, "loss": 0.0052, "step": 16670 }, { "grad_norm": 0.18918868899345398, "learning_rate": 7.350722395768722e-06, "loss": 0.0043, "step": 16680 }, { "grad_norm": 0.24705936014652252, "learning_rate": 7.307630567019963e-06, "loss": 0.0053, "step": 16690 }, { "grad_norm": 0.2932035028934479, "learning_rate": 7.264655457579e-06, "loss": 0.0074, "step": 16700 }, { "grad_norm": 0.10619460791349411, "learning_rate": 7.221797184938184e-06, "loss": 0.0043, "step": 16710 }, { "grad_norm": 0.13791345059871674, "learning_rate": 7.179055866270373e-06, "loss": 0.0075, "step": 16720 }, { "grad_norm": 0.12416204810142517, "learning_rate": 7.136431618428707e-06, "loss": 0.0041, "step": 16730 }, { "grad_norm": 0.08521026372909546, "learning_rate": 7.09392455794628e-06, "loss": 0.0046, "step": 16740 }, { "grad_norm": 0.2141842395067215, "learning_rate": 7.051534801035725e-06, "loss": 0.0047, "step": 16750 }, { "grad_norm": 0.16211766004562378, "learning_rate": 7.00926246358905e-06, "loss": 0.0049, "step": 16760 }, { "grad_norm": 0.10951259732246399, "learning_rate": 6.967107661177191e-06, "loss": 0.0065, "step": 16770 }, { "grad_norm": 0.10349685698747635, "learning_rate": 6.925070509049786e-06, "loss": 0.0048, "step": 16780 }, { "grad_norm": 0.24771879613399506, "learning_rate": 6.883151122134812e-06, "loss": 0.0055, "step": 16790 }, { "grad_norm": 0.0939963310956955, "learning_rate": 6.8413496150382394e-06, "loss": 0.0052, "step": 16800 }, { "grad_norm": 0.09511350840330124, "learning_rate": 6.7996661020438165e-06, "loss": 0.0073, "step": 16810 }, { "grad_norm": 0.24487431347370148, "learning_rate": 6.758100697112662e-06, "loss": 0.0053, "step": 16820 }, { "grad_norm": 0.13839127123355865, "learning_rate": 6.716653513883026e-06, "loss": 0.0048, "step": 16830 }, { "grad_norm": 0.19448305666446686, "learning_rate": 6.675324665669913e-06, "loss": 0.0045, "step": 16840 }, { "grad_norm": 0.15344883501529694, "learning_rate": 6.634114265464803e-06, "loss": 0.0052, "step": 16850 }, { "grad_norm": 0.11167632788419724, "learning_rate": 6.59302242593538e-06, "loss": 0.0044, "step": 16860 }, { "grad_norm": 0.08543732017278671, "learning_rate": 6.552049259425141e-06, "loss": 0.0048, "step": 16870 }, { "grad_norm": 0.24952682852745056, "learning_rate": 6.511194877953181e-06, "loss": 0.0058, "step": 16880 }, { "grad_norm": 0.2738325595855713, "learning_rate": 6.470459393213813e-06, "loss": 0.0061, "step": 16890 }, { "grad_norm": 0.18930433690547943, "learning_rate": 6.429842916576279e-06, "loss": 0.007, "step": 16900 }, { "grad_norm": 0.16963236033916473, "learning_rate": 6.389345559084503e-06, "loss": 0.0045, "step": 16910 }, { "grad_norm": 0.12538611888885498, "learning_rate": 6.348967431456682e-06, "loss": 0.0055, "step": 16920 }, { "grad_norm": 0.1983567774295807, "learning_rate": 6.30870864408511e-06, "loss": 0.005, "step": 16930 }, { "grad_norm": 0.26217392086982727, "learning_rate": 6.268569307035754e-06, "loss": 0.0051, "step": 16940 }, { "grad_norm": 0.07571061700582504, "learning_rate": 6.228549530048022e-06, "loss": 0.0041, "step": 16950 }, { "grad_norm": 0.09586335718631744, "learning_rate": 6.1886494225344814e-06, "loss": 0.0054, "step": 16960 }, { "grad_norm": 0.14684750139713287, "learning_rate": 6.148869093580479e-06, "loss": 0.0057, "step": 16970 }, { "grad_norm": 0.11763352900743484, "learning_rate": 6.109208651943921e-06, "loss": 0.0049, "step": 16980 }, { "grad_norm": 0.11983560770750046, "learning_rate": 6.069668206054946e-06, "loss": 0.0066, "step": 16990 }, { "grad_norm": 0.19931241869926453, "learning_rate": 6.0302478640156145e-06, "loss": 0.006, "step": 17000 }, { "grad_norm": 0.11649998277425766, "learning_rate": 5.990947733599644e-06, "loss": 0.0045, "step": 17010 }, { "grad_norm": 0.187097430229187, "learning_rate": 5.951767922252105e-06, "loss": 0.0054, "step": 17020 }, { "grad_norm": 0.1863829344511032, "learning_rate": 5.912708537089068e-06, "loss": 0.0051, "step": 17030 }, { "grad_norm": 0.12270959466695786, "learning_rate": 5.873769684897434e-06, "loss": 0.0053, "step": 17040 }, { "grad_norm": 0.09193172305822372, "learning_rate": 5.834951472134514e-06, "loss": 0.0052, "step": 17050 }, { "grad_norm": 0.08239881694316864, "learning_rate": 5.796254004927832e-06, "loss": 0.0046, "step": 17060 }, { "grad_norm": 0.22824274003505707, "learning_rate": 5.757677389074806e-06, "loss": 0.0062, "step": 17070 }, { "grad_norm": 0.1318599432706833, "learning_rate": 5.719221730042385e-06, "loss": 0.0065, "step": 17080 }, { "grad_norm": 0.10757116973400116, "learning_rate": 5.680887132966911e-06, "loss": 0.0037, "step": 17090 }, { "grad_norm": 0.12775897979736328, "learning_rate": 5.642673702653683e-06, "loss": 0.0058, "step": 17100 }, { "grad_norm": 0.2423734813928604, "learning_rate": 5.604581543576781e-06, "loss": 0.0061, "step": 17110 }, { "grad_norm": 0.1735844761133194, "learning_rate": 5.566610759878704e-06, "loss": 0.0053, "step": 17120 }, { "grad_norm": 0.2063962072134018, "learning_rate": 5.528761455370119e-06, "loss": 0.0061, "step": 17130 }, { "grad_norm": 0.13024362921714783, "learning_rate": 5.491033733529594e-06, "loss": 0.0072, "step": 17140 }, { "grad_norm": 0.12655507028102875, "learning_rate": 5.453427697503255e-06, "loss": 0.0044, "step": 17150 }, { "grad_norm": 0.16940344870090485, "learning_rate": 5.415943450104599e-06, "loss": 0.0055, "step": 17160 }, { "grad_norm": 0.11684836447238922, "learning_rate": 5.378581093814111e-06, "loss": 0.005, "step": 17170 }, { "grad_norm": 0.17797866463661194, "learning_rate": 5.3413407307790375e-06, "loss": 0.0047, "step": 17180 }, { "grad_norm": 0.20596137642860413, "learning_rate": 5.30422246281313e-06, "loss": 0.0042, "step": 17190 }, { "grad_norm": 0.14318975806236267, "learning_rate": 5.267226391396296e-06, "loss": 0.0046, "step": 17200 }, { "grad_norm": 0.11383084207773209, "learning_rate": 5.2303526176744e-06, "loss": 0.0064, "step": 17210 }, { "grad_norm": 0.10154753923416138, "learning_rate": 5.193601242458929e-06, "loss": 0.0043, "step": 17220 }, { "grad_norm": 0.08722593635320663, "learning_rate": 5.156972366226714e-06, "loss": 0.0047, "step": 17230 }, { "grad_norm": 0.11257325112819672, "learning_rate": 5.120466089119735e-06, "loss": 0.0048, "step": 17240 }, { "grad_norm": 0.10820606350898743, "learning_rate": 5.084082510944749e-06, "loss": 0.0076, "step": 17250 }, { "grad_norm": 0.23214158415794373, "learning_rate": 5.047821731173058e-06, "loss": 0.0057, "step": 17260 }, { "grad_norm": 0.12908749282360077, "learning_rate": 5.011683848940274e-06, "loss": 0.0057, "step": 17270 }, { "grad_norm": 0.22694729268550873, "learning_rate": 4.975668963045954e-06, "loss": 0.0057, "step": 17280 }, { "grad_norm": 0.14323537051677704, "learning_rate": 4.9397771719534525e-06, "loss": 0.0047, "step": 17290 }, { "grad_norm": 0.10725570470094681, "learning_rate": 4.904008573789548e-06, "loss": 0.0063, "step": 17300 }, { "grad_norm": 0.3616940975189209, "learning_rate": 4.8683632663442005e-06, "loss": 0.0063, "step": 17310 }, { "grad_norm": 0.1574936807155609, "learning_rate": 4.832841347070343e-06, "loss": 0.004, "step": 17320 }, { "grad_norm": 0.11101516336202621, "learning_rate": 4.797442913083539e-06, "loss": 0.0051, "step": 17330 }, { "grad_norm": 0.13963113725185394, "learning_rate": 4.7621680611617596e-06, "loss": 0.0052, "step": 17340 }, { "grad_norm": 0.1563306450843811, "learning_rate": 4.727016887745095e-06, "loss": 0.0055, "step": 17350 }, { "grad_norm": 0.1552247554063797, "learning_rate": 4.691989488935511e-06, "loss": 0.0059, "step": 17360 }, { "grad_norm": 0.12903225421905518, "learning_rate": 4.657085960496588e-06, "loss": 0.0042, "step": 17370 }, { "grad_norm": 0.08440220355987549, "learning_rate": 4.6223063978532265e-06, "loss": 0.0039, "step": 17380 }, { "grad_norm": 0.22230346500873566, "learning_rate": 4.587650896091439e-06, "loss": 0.0036, "step": 17390 }, { "grad_norm": 0.18031036853790283, "learning_rate": 4.553119549958035e-06, "loss": 0.0039, "step": 17400 }, { "grad_norm": 0.14127127826213837, "learning_rate": 4.518712453860385e-06, "loss": 0.0048, "step": 17410 }, { "grad_norm": 0.1608087122440338, "learning_rate": 4.484429701866205e-06, "loss": 0.0052, "step": 17420 }, { "grad_norm": 0.15428851544857025, "learning_rate": 4.4502713877031975e-06, "loss": 0.0065, "step": 17430 }, { "grad_norm": 0.09931740164756775, "learning_rate": 4.416237604758911e-06, "loss": 0.0034, "step": 17440 }, { "grad_norm": 0.14691193401813507, "learning_rate": 4.3823284460804025e-06, "loss": 0.0051, "step": 17450 }, { "grad_norm": 0.15339605510234833, "learning_rate": 4.348544004374011e-06, "loss": 0.0046, "step": 17460 }, { "grad_norm": 0.29708021879196167, "learning_rate": 4.314884372005123e-06, "loss": 0.0056, "step": 17470 }, { "grad_norm": 0.13721109926700592, "learning_rate": 4.281349640997867e-06, "loss": 0.0047, "step": 17480 }, { "grad_norm": 0.11771245300769806, "learning_rate": 4.247939903034942e-06, "loss": 0.0051, "step": 17490 }, { "grad_norm": 0.09580764919519424, "learning_rate": 4.214655249457284e-06, "loss": 0.0042, "step": 17500 }, { "grad_norm": 0.13237659633159637, "learning_rate": 4.181495771263855e-06, "loss": 0.0046, "step": 17510 }, { "grad_norm": 0.11580409109592438, "learning_rate": 4.148461559111427e-06, "loss": 0.0049, "step": 17520 }, { "grad_norm": 0.19971157610416412, "learning_rate": 4.115552703314252e-06, "loss": 0.005, "step": 17530 }, { "grad_norm": 0.13770025968551636, "learning_rate": 4.082769293843886e-06, "loss": 0.0062, "step": 17540 }, { "grad_norm": 0.15071751177310944, "learning_rate": 4.050111420328939e-06, "loss": 0.0053, "step": 17550 }, { "grad_norm": 0.15480288863182068, "learning_rate": 4.017579172054764e-06, "loss": 0.0051, "step": 17560 }, { "grad_norm": 0.08589605987071991, "learning_rate": 3.985172637963308e-06, "loss": 0.005, "step": 17570 }, { "grad_norm": 0.15905706584453583, "learning_rate": 3.952891906652784e-06, "loss": 0.0056, "step": 17580 }, { "grad_norm": 0.1610817313194275, "learning_rate": 3.920737066377478e-06, "loss": 0.004, "step": 17590 }, { "grad_norm": 0.08294785022735596, "learning_rate": 3.888708205047509e-06, "loss": 0.0034, "step": 17600 }, { "grad_norm": 0.20702502131462097, "learning_rate": 3.856805410228542e-06, "loss": 0.0045, "step": 17610 }, { "grad_norm": 0.1811138093471527, "learning_rate": 3.82502876914162e-06, "loss": 0.0045, "step": 17620 }, { "grad_norm": 0.2045951634645462, "learning_rate": 3.7933783686628586e-06, "loss": 0.0048, "step": 17630 }, { "grad_norm": 0.22259974479675293, "learning_rate": 3.7618542953232306e-06, "loss": 0.0048, "step": 17640 }, { "grad_norm": 0.06711810827255249, "learning_rate": 3.7304566353083658e-06, "loss": 0.0043, "step": 17650 }, { "grad_norm": 0.5942244529724121, "learning_rate": 3.6991854744582555e-06, "loss": 0.0047, "step": 17660 }, { "grad_norm": 0.12920062243938446, "learning_rate": 3.6680408982670777e-06, "loss": 0.0064, "step": 17670 }, { "grad_norm": 0.11750434339046478, "learning_rate": 3.637022991882899e-06, "loss": 0.004, "step": 17680 }, { "grad_norm": 0.09240257740020752, "learning_rate": 3.606131840107485e-06, "loss": 0.0041, "step": 17690 }, { "grad_norm": 0.0994911640882492, "learning_rate": 3.575367527396084e-06, "loss": 0.0067, "step": 17700 }, { "grad_norm": 0.09514772891998291, "learning_rate": 3.5447301378571386e-06, "loss": 0.0047, "step": 17710 }, { "grad_norm": 0.2185412049293518, "learning_rate": 3.514219755252113e-06, "loss": 0.0048, "step": 17720 }, { "grad_norm": 0.1583646535873413, "learning_rate": 3.4838364629952213e-06, "loss": 0.0054, "step": 17730 }, { "grad_norm": 0.14002589881420135, "learning_rate": 3.4535803441532123e-06, "loss": 0.0051, "step": 17740 }, { "grad_norm": 0.15753643214702606, "learning_rate": 3.4234514814451836e-06, "loss": 0.0048, "step": 17750 }, { "grad_norm": 0.20086023211479187, "learning_rate": 3.393449957242273e-06, "loss": 0.0051, "step": 17760 }, { "grad_norm": 0.12859241664409637, "learning_rate": 3.363575853567524e-06, "loss": 0.0055, "step": 17770 }, { "grad_norm": 0.06545715779066086, "learning_rate": 3.3338292520955826e-06, "loss": 0.0048, "step": 17780 }, { "grad_norm": 0.21723173558712006, "learning_rate": 3.304210234152516e-06, "loss": 0.0048, "step": 17790 }, { "grad_norm": 0.1383974701166153, "learning_rate": 3.2747188807155993e-06, "loss": 0.0058, "step": 17800 }, { "grad_norm": 0.14796775579452515, "learning_rate": 3.2453552724130643e-06, "loss": 0.0049, "step": 17810 }, { "grad_norm": 0.2962147891521454, "learning_rate": 3.216119489523889e-06, "loss": 0.005, "step": 17820 }, { "grad_norm": 0.14849650859832764, "learning_rate": 3.1870116119775917e-06, "loss": 0.0044, "step": 17830 }, { "grad_norm": 0.20346806943416595, "learning_rate": 3.158031719353999e-06, "loss": 0.0054, "step": 17840 }, { "grad_norm": 0.310509592294693, "learning_rate": 3.1291798908830273e-06, "loss": 0.0098, "step": 17850 }, { "grad_norm": 0.2674964368343353, "learning_rate": 3.1004562054444853e-06, "loss": 0.0049, "step": 17860 }, { "grad_norm": 0.19068869948387146, "learning_rate": 3.071860741567806e-06, "loss": 0.0042, "step": 17870 }, { "grad_norm": 0.17736060917377472, "learning_rate": 3.04339357743193e-06, "loss": 0.0057, "step": 17880 }, { "grad_norm": 0.08386991918087006, "learning_rate": 3.0150547908649628e-06, "loss": 0.0042, "step": 17890 }, { "grad_norm": 0.5458618402481079, "learning_rate": 2.9868444593440957e-06, "loss": 0.0054, "step": 17900 }, { "grad_norm": 0.2270624190568924, "learning_rate": 2.9587626599952846e-06, "loss": 0.0044, "step": 17910 }, { "grad_norm": 0.10960100591182709, "learning_rate": 2.930809469593082e-06, "loss": 0.0071, "step": 17920 }, { "grad_norm": 0.13843925297260284, "learning_rate": 2.9029849645604733e-06, "loss": 0.0047, "step": 17930 }, { "grad_norm": 0.2076551467180252, "learning_rate": 2.8752892209685632e-06, "loss": 0.0058, "step": 17940 }, { "grad_norm": 0.0740373283624649, "learning_rate": 2.847722314536483e-06, "loss": 0.0041, "step": 17950 }, { "grad_norm": 0.2878095507621765, "learning_rate": 2.820284320631078e-06, "loss": 0.0049, "step": 17960 }, { "grad_norm": 0.16466116905212402, "learning_rate": 2.792975314266788e-06, "loss": 0.0085, "step": 17970 }, { "grad_norm": 0.15636955201625824, "learning_rate": 2.7657953701054007e-06, "loss": 0.0054, "step": 17980 }, { "grad_norm": 0.09310764074325562, "learning_rate": 2.7387445624558306e-06, "loss": 0.0059, "step": 17990 }, { "grad_norm": 0.45640701055526733, "learning_rate": 2.7118229652739747e-06, "loss": 0.0058, "step": 18000 }, { "grad_norm": 0.23663705587387085, "learning_rate": 2.6850306521624236e-06, "loss": 0.0047, "step": 18010 }, { "grad_norm": 0.12767678499221802, "learning_rate": 2.6583676963703507e-06, "loss": 0.0062, "step": 18020 }, { "grad_norm": 0.10930272936820984, "learning_rate": 2.631834170793268e-06, "loss": 0.0065, "step": 18030 }, { "grad_norm": 0.10467953979969025, "learning_rate": 2.6054301479728036e-06, "loss": 0.0051, "step": 18040 }, { "grad_norm": 0.1390862911939621, "learning_rate": 2.579155700096575e-06, "loss": 0.0054, "step": 18050 }, { "grad_norm": 0.10999248921871185, "learning_rate": 2.5530108989978873e-06, "loss": 0.0063, "step": 18060 }, { "grad_norm": 0.09923200309276581, "learning_rate": 2.5269958161556416e-06, "loss": 0.0039, "step": 18070 }, { "grad_norm": 0.14087066054344177, "learning_rate": 2.5011105226940888e-06, "loss": 0.0047, "step": 18080 }, { "grad_norm": 0.15888482332229614, "learning_rate": 2.4753550893826248e-06, "loss": 0.0041, "step": 18090 }, { "grad_norm": 0.12528164684772491, "learning_rate": 2.4497295866356296e-06, "loss": 0.0045, "step": 18100 }, { "grad_norm": 0.07817265391349792, "learning_rate": 2.424234084512228e-06, "loss": 0.0046, "step": 18110 }, { "grad_norm": 0.0822952464222908, "learning_rate": 2.3988686527161687e-06, "loss": 0.0044, "step": 18120 }, { "grad_norm": 0.08261331915855408, "learning_rate": 2.373633360595573e-06, "loss": 0.0049, "step": 18130 }, { "grad_norm": 0.19211578369140625, "learning_rate": 2.3485282771427585e-06, "loss": 0.0051, "step": 18140 }, { "grad_norm": 0.07832220196723938, "learning_rate": 2.3235534709940665e-06, "loss": 0.0035, "step": 18150 }, { "grad_norm": 0.12068687379360199, "learning_rate": 2.2987090104296617e-06, "loss": 0.0054, "step": 18160 }, { "grad_norm": 0.10777637362480164, "learning_rate": 2.273994963373355e-06, "loss": 0.0043, "step": 18170 }, { "grad_norm": 0.22857287526130676, "learning_rate": 2.249411397392409e-06, "loss": 0.005, "step": 18180 }, { "grad_norm": 0.17759276926517487, "learning_rate": 2.2249583796973506e-06, "loss": 0.0047, "step": 18190 }, { "grad_norm": 0.09086990356445312, "learning_rate": 2.200635977141796e-06, "loss": 0.0046, "step": 18200 }, { "grad_norm": 0.11037593334913254, "learning_rate": 2.17644425622226e-06, "loss": 0.0041, "step": 18210 }, { "grad_norm": 0.15897154808044434, "learning_rate": 2.152383283077991e-06, "loss": 0.0051, "step": 18220 }, { "grad_norm": 0.2290656566619873, "learning_rate": 2.128453123490781e-06, "loss": 0.0061, "step": 18230 }, { "grad_norm": 0.2014833390712738, "learning_rate": 2.1046538428847462e-06, "loss": 0.0048, "step": 18240 }, { "grad_norm": 0.19485679268836975, "learning_rate": 2.0809855063262273e-06, "loss": 0.0048, "step": 18250 }, { "grad_norm": 0.05887121707201004, "learning_rate": 2.057448178523558e-06, "loss": 0.0039, "step": 18260 }, { "grad_norm": 0.23395031690597534, "learning_rate": 2.034041923826885e-06, "loss": 0.0041, "step": 18270 }, { "grad_norm": 0.06796801835298538, "learning_rate": 2.0107668062280204e-06, "loss": 0.0037, "step": 18280 }, { "grad_norm": 0.15631751716136932, "learning_rate": 1.9876228893602357e-06, "loss": 0.0054, "step": 18290 }, { "grad_norm": 0.14569973945617676, "learning_rate": 1.9646102364981266e-06, "loss": 0.0043, "step": 18300 }, { "grad_norm": 0.05591166764497757, "learning_rate": 1.9417289105574053e-06, "loss": 0.0044, "step": 18310 }, { "grad_norm": 0.2080976665019989, "learning_rate": 1.9189789740947427e-06, "loss": 0.0035, "step": 18320 }, { "grad_norm": 0.2441105842590332, "learning_rate": 1.896360489307597e-06, "loss": 0.0042, "step": 18330 }, { "grad_norm": 0.20226183533668518, "learning_rate": 1.8738735180340362e-06, "loss": 0.0048, "step": 18340 }, { "grad_norm": 0.10671108216047287, "learning_rate": 1.8515181217525824e-06, "loss": 0.0036, "step": 18350 }, { "grad_norm": 0.09736452251672745, "learning_rate": 1.8292943615820457e-06, "loss": 0.0046, "step": 18360 }, { "grad_norm": 0.3175564110279083, "learning_rate": 1.8072022982813296e-06, "loss": 0.0061, "step": 18370 }, { "grad_norm": 0.21738679707050323, "learning_rate": 1.7852419922492925e-06, "loss": 0.0044, "step": 18380 }, { "grad_norm": 0.2020753026008606, "learning_rate": 1.763413503524569e-06, "loss": 0.0043, "step": 18390 }, { "grad_norm": 0.515383780002594, "learning_rate": 1.7417168917854165e-06, "loss": 0.0045, "step": 18400 }, { "grad_norm": 0.2578742504119873, "learning_rate": 1.720152216349552e-06, "loss": 0.0067, "step": 18410 }, { "grad_norm": 0.07830578833818436, "learning_rate": 1.6987195361739595e-06, "loss": 0.0038, "step": 18420 }, { "grad_norm": 0.08631180971860886, "learning_rate": 1.6774189098547832e-06, "loss": 0.0047, "step": 18430 }, { "grad_norm": 0.25773003697395325, "learning_rate": 1.6562503956271069e-06, "loss": 0.0047, "step": 18440 }, { "grad_norm": 0.08599376678466797, "learning_rate": 1.6352140513648417e-06, "loss": 0.0042, "step": 18450 }, { "grad_norm": 0.29080089926719666, "learning_rate": 1.6143099345805712e-06, "loss": 0.0055, "step": 18460 }, { "grad_norm": 0.10286157578229904, "learning_rate": 1.5935381024253293e-06, "loss": 0.0038, "step": 18470 }, { "grad_norm": 0.1934097707271576, "learning_rate": 1.572898611688517e-06, "loss": 0.0044, "step": 18480 }, { "grad_norm": 0.11901362240314484, "learning_rate": 1.5523915187977133e-06, "loss": 0.0047, "step": 18490 }, { "grad_norm": 0.16781127452850342, "learning_rate": 1.532016879818532e-06, "loss": 0.0051, "step": 18500 }, { "grad_norm": 0.29636478424072266, "learning_rate": 1.51177475045447e-06, "loss": 0.0057, "step": 18510 }, { "grad_norm": 0.0915936604142189, "learning_rate": 1.4916651860467035e-06, "loss": 0.0046, "step": 18520 }, { "grad_norm": 0.4732084572315216, "learning_rate": 1.471688241574043e-06, "loss": 0.0061, "step": 18530 }, { "grad_norm": 0.18966174125671387, "learning_rate": 1.451843971652672e-06, "loss": 0.0068, "step": 18540 }, { "grad_norm": 0.10444244742393494, "learning_rate": 1.432132430536076e-06, "loss": 0.0048, "step": 18550 }, { "grad_norm": 0.08843202888965607, "learning_rate": 1.412553672114869e-06, "loss": 0.0068, "step": 18560 }, { "grad_norm": 0.21581582725048065, "learning_rate": 1.3931077499166056e-06, "loss": 0.0052, "step": 18570 }, { "grad_norm": 0.29907017946243286, "learning_rate": 1.3737947171057085e-06, "loss": 0.0042, "step": 18580 }, { "grad_norm": 0.1875777691602707, "learning_rate": 1.3546146264832582e-06, "loss": 0.0047, "step": 18590 }, { "grad_norm": 0.1782672107219696, "learning_rate": 1.3355675304869086e-06, "loss": 0.0051, "step": 18600 }, { "grad_norm": 0.19678841531276703, "learning_rate": 1.3166534811906827e-06, "loss": 0.0044, "step": 18610 }, { "grad_norm": 0.11261522769927979, "learning_rate": 1.2978725303048666e-06, "loss": 0.0043, "step": 18620 }, { "grad_norm": 0.14073891937732697, "learning_rate": 1.2792247291758762e-06, "loss": 0.0042, "step": 18630 }, { "grad_norm": 0.07456114143133163, "learning_rate": 1.2607101287860635e-06, "loss": 0.0036, "step": 18640 }, { "grad_norm": 0.0954209640622139, "learning_rate": 1.2423287797536654e-06, "loss": 0.004, "step": 18650 }, { "grad_norm": 0.17905673384666443, "learning_rate": 1.2240807323325776e-06, "loss": 0.0051, "step": 18660 }, { "grad_norm": 0.18040858209133148, "learning_rate": 1.205966036412254e-06, "loss": 0.0092, "step": 18670 }, { "grad_norm": 0.15471117198467255, "learning_rate": 1.1879847415175949e-06, "loss": 0.0047, "step": 18680 }, { "grad_norm": 0.10364064574241638, "learning_rate": 1.1701368968087712e-06, "loss": 0.0044, "step": 18690 }, { "grad_norm": 0.20020684599876404, "learning_rate": 1.1524225510811116e-06, "loss": 0.0059, "step": 18700 }, { "grad_norm": 0.13725629448890686, "learning_rate": 1.1348417527649535e-06, "loss": 0.0044, "step": 18710 }, { "grad_norm": 0.11646116524934769, "learning_rate": 1.1173945499255268e-06, "loss": 0.0044, "step": 18720 }, { "grad_norm": 0.12108121067285538, "learning_rate": 1.1000809902628307e-06, "loss": 0.0039, "step": 18730 }, { "grad_norm": 0.1474704146385193, "learning_rate": 1.082901121111468e-06, "loss": 0.0043, "step": 18740 }, { "grad_norm": 0.17346182465553284, "learning_rate": 1.0658549894405456e-06, "loss": 0.0041, "step": 18750 }, { "grad_norm": 0.14815430343151093, "learning_rate": 1.0489426418535342e-06, "loss": 0.0043, "step": 18760 }, { "grad_norm": 0.1399923712015152, "learning_rate": 1.0321641245881474e-06, "loss": 0.0045, "step": 18770 }, { "grad_norm": 0.3013243079185486, "learning_rate": 1.015519483516214e-06, "loss": 0.0044, "step": 18780 }, { "grad_norm": 0.4496837258338928, "learning_rate": 9.990087641435443e-07, "loss": 0.0055, "step": 18790 }, { "grad_norm": 0.21135134994983673, "learning_rate": 9.826320116098132e-07, "loss": 0.0055, "step": 18800 }, { "grad_norm": 0.10870014876127243, "learning_rate": 9.663892706884447e-07, "loss": 0.0045, "step": 18810 }, { "grad_norm": 0.06662442535161972, "learning_rate": 9.502805857864616e-07, "loss": 0.0048, "step": 18820 }, { "grad_norm": 0.20820352435112, "learning_rate": 9.34306000944396e-07, "loss": 0.0039, "step": 18830 }, { "grad_norm": 0.15246018767356873, "learning_rate": 9.184655598361624e-07, "loss": 0.0046, "step": 18840 }, { "grad_norm": 0.14060992002487183, "learning_rate": 9.027593057689076e-07, "loss": 0.0042, "step": 18850 }, { "grad_norm": 0.3176056742668152, "learning_rate": 8.871872816829441e-07, "loss": 0.0037, "step": 18860 }, { "grad_norm": 0.13642345368862152, "learning_rate": 8.717495301515777e-07, "loss": 0.0043, "step": 18870 }, { "grad_norm": 0.12462523579597473, "learning_rate": 8.564460933810415e-07, "loss": 0.0046, "step": 18880 }, { "grad_norm": 0.13640190660953522, "learning_rate": 8.412770132103453e-07, "loss": 0.004, "step": 18890 }, { "grad_norm": 0.1010136753320694, "learning_rate": 8.262423311111711e-07, "loss": 0.0044, "step": 18900 }, { "grad_norm": 0.21642367541790009, "learning_rate": 8.113420881877665e-07, "loss": 0.0041, "step": 18910 }, { "grad_norm": 0.20030346512794495, "learning_rate": 7.965763251768288e-07, "loss": 0.0052, "step": 18920 }, { "grad_norm": 0.07534200698137283, "learning_rate": 7.819450824473995e-07, "loss": 0.0046, "step": 18930 }, { "grad_norm": 0.07187651097774506, "learning_rate": 7.674484000007198e-07, "loss": 0.0039, "step": 18940 }, { "grad_norm": 0.09570746123790741, "learning_rate": 7.530863174701752e-07, "loss": 0.0041, "step": 18950 }, { "grad_norm": 0.14848357439041138, "learning_rate": 7.38858874121151e-07, "loss": 0.0046, "step": 18960 }, { "grad_norm": 0.10728520900011063, "learning_rate": 7.247661088509328e-07, "loss": 0.0041, "step": 18970 }, { "grad_norm": 0.07821942120790482, "learning_rate": 7.108080601886002e-07, "loss": 0.0037, "step": 18980 }, { "grad_norm": 0.10489551723003387, "learning_rate": 6.969847662949336e-07, "loss": 0.0041, "step": 18990 }, { "grad_norm": 0.09831903874874115, "learning_rate": 6.832962649622798e-07, "loss": 0.0035, "step": 19000 }, { "grad_norm": 0.23714427649974823, "learning_rate": 6.697425936144863e-07, "loss": 0.0045, "step": 19010 }, { "grad_norm": 0.09780099987983704, "learning_rate": 6.563237893067731e-07, "loss": 0.0046, "step": 19020 }, { "grad_norm": 0.287259578704834, "learning_rate": 6.430398887256328e-07, "loss": 0.0049, "step": 19030 }, { "grad_norm": 0.1303209811449051, "learning_rate": 6.298909281887478e-07, "loss": 0.0039, "step": 19040 }, { "grad_norm": 0.09827516227960587, "learning_rate": 6.168769436448673e-07, "loss": 0.0033, "step": 19050 }, { "grad_norm": 0.11640780419111252, "learning_rate": 6.03997970673742e-07, "loss": 0.0034, "step": 19060 }, { "grad_norm": 0.34408628940582275, "learning_rate": 5.912540444859782e-07, "loss": 0.004, "step": 19070 }, { "grad_norm": 0.1404014676809311, "learning_rate": 5.786451999229837e-07, "loss": 0.0044, "step": 19080 }, { "grad_norm": 0.12367918342351913, "learning_rate": 5.661714714568722e-07, "loss": 0.0037, "step": 19090 }, { "grad_norm": 0.15328967571258545, "learning_rate": 5.538328931903259e-07, "loss": 0.0043, "step": 19100 }, { "grad_norm": 0.07693419605493546, "learning_rate": 5.416294988565551e-07, "loss": 0.0037, "step": 19110 }, { "grad_norm": 0.08646579086780548, "learning_rate": 5.29561321819172e-07, "loss": 0.004, "step": 19120 }, { "grad_norm": 0.16110026836395264, "learning_rate": 5.176283950721061e-07, "loss": 0.0046, "step": 19130 }, { "grad_norm": 0.11052799224853516, "learning_rate": 5.058307512395332e-07, "loss": 0.0037, "step": 19140 }, { "grad_norm": 0.1054442748427391, "learning_rate": 4.941684225757526e-07, "loss": 0.0044, "step": 19150 }, { "grad_norm": 0.27489691972732544, "learning_rate": 4.826414409651314e-07, "loss": 0.0053, "step": 19160 }, { "grad_norm": 0.1843201220035553, "learning_rate": 4.712498379219943e-07, "loss": 0.0044, "step": 19170 }, { "grad_norm": 0.10936956852674484, "learning_rate": 4.599936445905506e-07, "loss": 0.0048, "step": 19180 }, { "grad_norm": 0.25090038776397705, "learning_rate": 4.4887289174480594e-07, "loss": 0.004, "step": 19190 }, { "grad_norm": 0.1647312343120575, "learning_rate": 4.378876097884621e-07, "loss": 0.0046, "step": 19200 }, { "grad_norm": 0.4449067711830139, "learning_rate": 4.2703782875487264e-07, "loss": 0.0036, "step": 19210 }, { "grad_norm": 0.06657539308071136, "learning_rate": 4.163235783069208e-07, "loss": 0.0038, "step": 19220 }, { "grad_norm": 0.14129798114299774, "learning_rate": 4.057448877369585e-07, "loss": 0.0051, "step": 19230 }, { "grad_norm": 0.16949862241744995, "learning_rate": 3.9530178596672295e-07, "loss": 0.0039, "step": 19240 }, { "grad_norm": 0.14871367812156677, "learning_rate": 3.849943015472479e-07, "loss": 0.0055, "step": 19250 }, { "grad_norm": 0.05025284364819527, "learning_rate": 3.748224626588137e-07, "loss": 0.0038, "step": 19260 }, { "grad_norm": 0.17078424990177155, "learning_rate": 3.647862971108307e-07, "loss": 0.0039, "step": 19270 }, { "grad_norm": 0.22514016926288605, "learning_rate": 3.5488583234179473e-07, "loss": 0.0051, "step": 19280 }, { "grad_norm": 0.15241055190563202, "learning_rate": 3.4512109541920413e-07, "loss": 0.0047, "step": 19290 }, { "grad_norm": 0.23233933746814728, "learning_rate": 3.354921130394706e-07, "loss": 0.0043, "step": 19300 }, { "grad_norm": 0.10467315465211868, "learning_rate": 3.259989115278639e-07, "loss": 0.0037, "step": 19310 }, { "grad_norm": 0.09275782853364944, "learning_rate": 3.1664151683843403e-07, "loss": 0.004, "step": 19320 }, { "grad_norm": 0.29165875911712646, "learning_rate": 3.074199545539447e-07, "loss": 0.0042, "step": 19330 }, { "grad_norm": 0.17750988900661469, "learning_rate": 2.983342498857955e-07, "loss": 0.0051, "step": 19340 }, { "grad_norm": 0.11714421957731247, "learning_rate": 2.893844276739499e-07, "loss": 0.005, "step": 19350 }, { "grad_norm": 0.0712306872010231, "learning_rate": 2.8057051238688514e-07, "loss": 0.0041, "step": 19360 }, { "grad_norm": 0.14554277062416077, "learning_rate": 2.71892528121509e-07, "loss": 0.0039, "step": 19370 }, { "grad_norm": 0.1637023240327835, "learning_rate": 2.633504986030988e-07, "loss": 0.0045, "step": 19380 }, { "grad_norm": 0.15247364342212677, "learning_rate": 2.549444471852347e-07, "loss": 0.0046, "step": 19390 }, { "grad_norm": 0.11917699128389359, "learning_rate": 2.4667439684974423e-07, "loss": 0.0044, "step": 19400 }, { "grad_norm": 0.20006094872951508, "learning_rate": 2.3854037020662467e-07, "loss": 0.005, "step": 19410 }, { "grad_norm": 0.1437601000070572, "learning_rate": 2.3054238949399288e-07, "loss": 0.0034, "step": 19420 }, { "grad_norm": 0.06230251491069794, "learning_rate": 2.2268047657802993e-07, "loss": 0.0043, "step": 19430 }, { "grad_norm": 0.14895881712436676, "learning_rate": 2.149546529529034e-07, "loss": 0.0049, "step": 19440 }, { "grad_norm": 0.13580268621444702, "learning_rate": 2.0736493974071736e-07, "loss": 0.0037, "step": 19450 }, { "grad_norm": 0.11051267385482788, "learning_rate": 1.9991135769145686e-07, "loss": 0.0035, "step": 19460 }, { "grad_norm": 0.07952316850423813, "learning_rate": 1.9259392718293245e-07, "loss": 0.0045, "step": 19470 }, { "grad_norm": 0.22933202981948853, "learning_rate": 1.8541266822072467e-07, "loss": 0.0057, "step": 19480 }, { "grad_norm": 0.14463379979133606, "learning_rate": 1.7836760043811184e-07, "loss": 0.0052, "step": 19490 }, { "grad_norm": 0.17319126427173615, "learning_rate": 1.7145874309604792e-07, "loss": 0.0036, "step": 19500 }, { "grad_norm": 0.0943792313337326, "learning_rate": 1.6468611508308474e-07, "loss": 0.0048, "step": 19510 }, { "grad_norm": 0.08833614736795425, "learning_rate": 1.5804973491532204e-07, "loss": 0.0045, "step": 19520 }, { "grad_norm": 0.08847440034151077, "learning_rate": 1.5154962073637424e-07, "loss": 0.0053, "step": 19530 }, { "grad_norm": 0.08224132657051086, "learning_rate": 1.4518579031730372e-07, "loss": 0.0037, "step": 19540 }, { "grad_norm": 0.159573495388031, "learning_rate": 1.389582610565876e-07, "loss": 0.0071, "step": 19550 }, { "grad_norm": 0.18006427586078644, "learning_rate": 1.3286704998003995e-07, "loss": 0.0058, "step": 19560 }, { "grad_norm": 0.15078650414943695, "learning_rate": 1.2691217374080632e-07, "loss": 0.005, "step": 19570 }, { "grad_norm": 0.26278990507125854, "learning_rate": 1.2109364861929705e-07, "loss": 0.0044, "step": 19580 }, { "grad_norm": 0.049890730530023575, "learning_rate": 1.1541149052312628e-07, "loss": 0.0039, "step": 19590 }, { "grad_norm": 0.1713542938232422, "learning_rate": 1.0986571498710074e-07, "loss": 0.0041, "step": 19600 }, { "grad_norm": 0.11811555176973343, "learning_rate": 1.0445633717316438e-07, "loss": 0.0032, "step": 19610 }, { "grad_norm": 0.11090978235006332, "learning_rate": 9.918337187034277e-08, "loss": 0.0035, "step": 19620 }, { "grad_norm": 0.10487181693315506, "learning_rate": 9.404683349472643e-08, "loss": 0.0046, "step": 19630 }, { "grad_norm": 0.08469382673501968, "learning_rate": 8.904673608940983e-08, "loss": 0.0035, "step": 19640 }, { "grad_norm": 0.08791341632604599, "learning_rate": 8.418309332447471e-08, "loss": 0.0034, "step": 19650 }, { "grad_norm": 0.11911951005458832, "learning_rate": 7.945591849692902e-08, "loss": 0.0038, "step": 19660 }, { "grad_norm": 0.10236451029777527, "learning_rate": 7.486522453069578e-08, "loss": 0.0046, "step": 19670 }, { "grad_norm": 0.35647299885749817, "learning_rate": 7.041102397655208e-08, "loss": 0.0041, "step": 19680 }, { "grad_norm": 0.1123572438955307, "learning_rate": 6.609332901210685e-08, "loss": 0.0033, "step": 19690 }, { "grad_norm": 0.20334173738956451, "learning_rate": 6.191215144178419e-08, "loss": 0.0073, "step": 19700 }, { "grad_norm": 0.18317662179470062, "learning_rate": 5.786750269675678e-08, "loss": 0.0037, "step": 19710 }, { "grad_norm": 0.1770678013563156, "learning_rate": 5.395939383494031e-08, "loss": 0.0043, "step": 19720 }, { "grad_norm": 0.05701444298028946, "learning_rate": 5.018783554095463e-08, "loss": 0.0043, "step": 19730 }, { "grad_norm": 0.13525618612766266, "learning_rate": 4.655283812610156e-08, "loss": 0.0049, "step": 19740 }, { "grad_norm": 0.08195001631975174, "learning_rate": 4.305441152831491e-08, "loss": 0.0041, "step": 19750 }, { "grad_norm": 0.10457485169172287, "learning_rate": 3.9692565312171584e-08, "loss": 0.0043, "step": 19760 }, { "grad_norm": 0.0753750130534172, "learning_rate": 3.6467308668824975e-08, "loss": 0.0038, "step": 19770 }, { "grad_norm": 0.14911320805549622, "learning_rate": 3.3378650416004964e-08, "loss": 0.0051, "step": 19780 }, { "grad_norm": 0.1687801480293274, "learning_rate": 3.042659899797906e-08, "loss": 0.0065, "step": 19790 }, { "grad_norm": 0.16779427230358124, "learning_rate": 2.76111624855524e-08, "loss": 0.004, "step": 19800 }, { "grad_norm": 0.2310183048248291, "learning_rate": 2.4932348576017784e-08, "loss": 0.0044, "step": 19810 }, { "grad_norm": 0.29713717103004456, "learning_rate": 2.239016459314458e-08, "loss": 0.0043, "step": 19820 }, { "grad_norm": 0.08210986852645874, "learning_rate": 1.9984617487173174e-08, "loss": 0.0058, "step": 19830 }, { "grad_norm": 0.1020374745130539, "learning_rate": 1.7715713834776105e-08, "loss": 0.0047, "step": 19840 }, { "grad_norm": 0.2653149366378784, "learning_rate": 1.5583459839046964e-08, "loss": 0.0066, "step": 19850 }, { "grad_norm": 0.13972827792167664, "learning_rate": 1.3587861329489304e-08, "loss": 0.0054, "step": 19860 }, { "grad_norm": 0.16129156947135925, "learning_rate": 1.1728923761994415e-08, "loss": 0.0038, "step": 19870 }, { "grad_norm": 0.0982513502240181, "learning_rate": 1.0006652218819135e-08, "loss": 0.0039, "step": 19880 }, { "grad_norm": 0.07048080861568451, "learning_rate": 8.421051408596947e-09, "loss": 0.0055, "step": 19890 }, { "grad_norm": 0.15688766539096832, "learning_rate": 6.972125666299123e-09, "loss": 0.0041, "step": 19900 }, { "grad_norm": 0.10301307588815689, "learning_rate": 5.659878953229169e-09, "loss": 0.0041, "step": 19910 }, { "grad_norm": 0.08453572541475296, "learning_rate": 4.48431485701728e-09, "loss": 0.0036, "step": 19920 }, { "grad_norm": 0.07852805405855179, "learning_rate": 3.4454365916203322e-09, "loss": 0.0042, "step": 19930 }, { "grad_norm": 0.17777632176876068, "learning_rate": 2.5432469972830332e-09, "loss": 0.004, "step": 19940 }, { "grad_norm": 0.14794519543647766, "learning_rate": 1.7777485405601203e-09, "loss": 0.0055, "step": 19950 }, { "grad_norm": 0.1273629516363144, "learning_rate": 1.1489433142941597e-09, "loss": 0.0044, "step": 19960 }, { "grad_norm": 0.22788840532302856, "learning_rate": 6.568330376210963e-10, "loss": 0.0044, "step": 19970 }, { "grad_norm": 0.11253508925437927, "learning_rate": 3.0141905594249787e-10, "loss": 0.0031, "step": 19980 }, { "grad_norm": 0.2629273533821106, "learning_rate": 8.270234094776008e-11, "loss": 0.0048, "step": 19990 }, { "grad_norm": 0.08234532177448273, "learning_rate": 6.834906085551041e-13, "loss": 0.0035, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }