{ "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": 0.15012216567993164, "learning_rate": 9e-07, "loss": 1.2086, "step": 10 }, { "grad_norm": 0.13072419166564941, "learning_rate": 1.9e-06, "loss": 1.2071, "step": 20 }, { "grad_norm": 0.12423919886350632, "learning_rate": 2.9e-06, "loss": 1.2035, "step": 30 }, { "grad_norm": 0.15493424236774445, "learning_rate": 3.9e-06, "loss": 1.1841, "step": 40 }, { "grad_norm": 0.2636421024799347, "learning_rate": 4.9000000000000005e-06, "loss": 1.1534, "step": 50 }, { "grad_norm": 0.1940675526857376, "learning_rate": 5.9e-06, "loss": 1.1253, "step": 60 }, { "grad_norm": 0.42185041308403015, "learning_rate": 6.900000000000001e-06, "loss": 1.11, "step": 70 }, { "grad_norm": 0.16094836592674255, "learning_rate": 7.9e-06, "loss": 1.1014, "step": 80 }, { "grad_norm": 0.21415859460830688, "learning_rate": 8.9e-06, "loss": 1.0986, "step": 90 }, { "grad_norm": 0.26082614064216614, "learning_rate": 9.900000000000002e-06, "loss": 1.0897, "step": 100 }, { "grad_norm": 0.24920392036437988, "learning_rate": 1.09e-05, "loss": 1.0845, "step": 110 }, { "grad_norm": 0.19736778736114502, "learning_rate": 1.19e-05, "loss": 1.0756, "step": 120 }, { "grad_norm": 0.17664603888988495, "learning_rate": 1.29e-05, "loss": 1.0589, "step": 130 }, { "grad_norm": 0.2617020905017853, "learning_rate": 1.3900000000000002e-05, "loss": 1.0399, "step": 140 }, { "grad_norm": 0.5237306952476501, "learning_rate": 1.49e-05, "loss": 1.0118, "step": 150 }, { "grad_norm": 0.29383063316345215, "learning_rate": 1.59e-05, "loss": 0.9761, "step": 160 }, { "grad_norm": 0.6056665778160095, "learning_rate": 1.69e-05, "loss": 0.9334, "step": 170 }, { "grad_norm": 0.527864396572113, "learning_rate": 1.79e-05, "loss": 0.8957, "step": 180 }, { "grad_norm": 0.7024576663970947, "learning_rate": 1.8900000000000002e-05, "loss": 0.8583, "step": 190 }, { "grad_norm": 0.6996781826019287, "learning_rate": 1.9900000000000003e-05, "loss": 0.8179, "step": 200 }, { "grad_norm": 0.8450202941894531, "learning_rate": 2.09e-05, "loss": 0.7829, "step": 210 }, { "grad_norm": 0.9583392143249512, "learning_rate": 2.19e-05, "loss": 0.745, "step": 220 }, { "grad_norm": 1.7142727375030518, "learning_rate": 2.29e-05, "loss": 0.7135, "step": 230 }, { "grad_norm": 0.8069450855255127, "learning_rate": 2.39e-05, "loss": 0.6912, "step": 240 }, { "grad_norm": 0.8630396723747253, "learning_rate": 2.4900000000000002e-05, "loss": 0.6688, "step": 250 }, { "grad_norm": 1.0455894470214844, "learning_rate": 2.5900000000000003e-05, "loss": 0.6463, "step": 260 }, { "grad_norm": 1.3055671453475952, "learning_rate": 2.6900000000000003e-05, "loss": 0.6241, "step": 270 }, { "grad_norm": 1.0451176166534424, "learning_rate": 2.7900000000000004e-05, "loss": 0.5995, "step": 280 }, { "grad_norm": 0.8477664589881897, "learning_rate": 2.8899999999999998e-05, "loss": 0.5817, "step": 290 }, { "grad_norm": 1.3464720249176025, "learning_rate": 2.9900000000000002e-05, "loss": 0.571, "step": 300 }, { "grad_norm": 0.90577632188797, "learning_rate": 3.09e-05, "loss": 0.5596, "step": 310 }, { "grad_norm": 0.9414107203483582, "learning_rate": 3.19e-05, "loss": 0.5423, "step": 320 }, { "grad_norm": 1.1524189710617065, "learning_rate": 3.29e-05, "loss": 0.5275, "step": 330 }, { "grad_norm": 0.7501125931739807, "learning_rate": 3.3900000000000004e-05, "loss": 0.5178, "step": 340 }, { "grad_norm": 1.073644995689392, "learning_rate": 3.49e-05, "loss": 0.4947, "step": 350 }, { "grad_norm": 1.12371027469635, "learning_rate": 3.59e-05, "loss": 0.4778, "step": 360 }, { "grad_norm": 1.2945274114608765, "learning_rate": 3.69e-05, "loss": 0.4703, "step": 370 }, { "grad_norm": 0.9674573540687561, "learning_rate": 3.79e-05, "loss": 0.4568, "step": 380 }, { "grad_norm": 1.2989366054534912, "learning_rate": 3.8900000000000004e-05, "loss": 0.4433, "step": 390 }, { "grad_norm": 1.2226121425628662, "learning_rate": 3.99e-05, "loss": 0.4272, "step": 400 }, { "grad_norm": 1.2388277053833008, "learning_rate": 4.09e-05, "loss": 0.4054, "step": 410 }, { "grad_norm": 1.370849609375, "learning_rate": 4.19e-05, "loss": 0.3921, "step": 420 }, { "grad_norm": 1.1839030981063843, "learning_rate": 4.29e-05, "loss": 0.3823, "step": 430 }, { "grad_norm": 0.9793280363082886, "learning_rate": 4.39e-05, "loss": 0.362, "step": 440 }, { "grad_norm": 1.08059823513031, "learning_rate": 4.49e-05, "loss": 0.3544, "step": 450 }, { "grad_norm": 1.2476121187210083, "learning_rate": 4.5900000000000004e-05, "loss": 0.3359, "step": 460 }, { "grad_norm": 1.5092262029647827, "learning_rate": 4.69e-05, "loss": 0.3235, "step": 470 }, { "grad_norm": 1.4717587232589722, "learning_rate": 4.79e-05, "loss": 0.31, "step": 480 }, { "grad_norm": 1.3592662811279297, "learning_rate": 4.89e-05, "loss": 0.2975, "step": 490 }, { "grad_norm": 1.1353168487548828, "learning_rate": 4.99e-05, "loss": 0.2877, "step": 500 }, { "grad_norm": 1.9561127424240112, "learning_rate": 5.0900000000000004e-05, "loss": 0.2692, "step": 510 }, { "grad_norm": 1.9443875551223755, "learning_rate": 5.19e-05, "loss": 0.253, "step": 520 }, { "grad_norm": 1.329230546951294, "learning_rate": 5.2900000000000005e-05, "loss": 0.2338, "step": 530 }, { "grad_norm": 1.4158693552017212, "learning_rate": 5.390000000000001e-05, "loss": 0.2161, "step": 540 }, { "grad_norm": 1.59760582447052, "learning_rate": 5.4900000000000006e-05, "loss": 0.2021, "step": 550 }, { "grad_norm": 1.1889493465423584, "learning_rate": 5.590000000000001e-05, "loss": 0.1897, "step": 560 }, { "grad_norm": 1.6009477376937866, "learning_rate": 5.69e-05, "loss": 0.1795, "step": 570 }, { "grad_norm": 1.6399871110916138, "learning_rate": 5.79e-05, "loss": 0.172, "step": 580 }, { "grad_norm": 1.272241473197937, "learning_rate": 5.89e-05, "loss": 0.1638, "step": 590 }, { "grad_norm": 1.4894850254058838, "learning_rate": 5.99e-05, "loss": 0.1546, "step": 600 }, { "grad_norm": 1.1664891242980957, "learning_rate": 6.09e-05, "loss": 0.1439, "step": 610 }, { "grad_norm": 1.2341861724853516, "learning_rate": 6.19e-05, "loss": 0.138, "step": 620 }, { "grad_norm": 1.2932838201522827, "learning_rate": 6.29e-05, "loss": 0.1372, "step": 630 }, { "grad_norm": 1.2948802709579468, "learning_rate": 6.390000000000001e-05, "loss": 0.1345, "step": 640 }, { "grad_norm": 1.1293632984161377, "learning_rate": 6.49e-05, "loss": 0.1281, "step": 650 }, { "grad_norm": 1.3699828386306763, "learning_rate": 6.59e-05, "loss": 0.1278, "step": 660 }, { "grad_norm": 1.3181347846984863, "learning_rate": 6.690000000000001e-05, "loss": 0.1265, "step": 670 }, { "grad_norm": 1.2064861059188843, "learning_rate": 6.790000000000001e-05, "loss": 0.1236, "step": 680 }, { "grad_norm": 1.164475440979004, "learning_rate": 6.89e-05, "loss": 0.1187, "step": 690 }, { "grad_norm": 1.2478479146957397, "learning_rate": 6.99e-05, "loss": 0.1205, "step": 700 }, { "grad_norm": 1.1712687015533447, "learning_rate": 7.09e-05, "loss": 0.1221, "step": 710 }, { "grad_norm": 1.3739711046218872, "learning_rate": 7.19e-05, "loss": 0.1204, "step": 720 }, { "grad_norm": 1.3635663986206055, "learning_rate": 7.29e-05, "loss": 0.1165, "step": 730 }, { "grad_norm": 1.1010574102401733, "learning_rate": 7.390000000000001e-05, "loss": 0.118, "step": 740 }, { "grad_norm": 1.2504950761795044, "learning_rate": 7.49e-05, "loss": 0.113, "step": 750 }, { "grad_norm": 1.144167184829712, "learning_rate": 7.59e-05, "loss": 0.111, "step": 760 }, { "grad_norm": 1.1996899843215942, "learning_rate": 7.69e-05, "loss": 0.1075, "step": 770 }, { "grad_norm": 1.0457687377929688, "learning_rate": 7.790000000000001e-05, "loss": 0.1131, "step": 780 }, { "grad_norm": 1.4280327558517456, "learning_rate": 7.890000000000001e-05, "loss": 0.1115, "step": 790 }, { "grad_norm": 1.16835618019104, "learning_rate": 7.99e-05, "loss": 0.1105, "step": 800 }, { "grad_norm": 1.403944492340088, "learning_rate": 8.090000000000001e-05, "loss": 0.1083, "step": 810 }, { "grad_norm": 1.2651947736740112, "learning_rate": 8.19e-05, "loss": 0.1099, "step": 820 }, { "grad_norm": 1.1956747770309448, "learning_rate": 8.29e-05, "loss": 0.1127, "step": 830 }, { "grad_norm": 1.141771912574768, "learning_rate": 8.39e-05, "loss": 0.1075, "step": 840 }, { "grad_norm": 1.2325164079666138, "learning_rate": 8.49e-05, "loss": 0.1075, "step": 850 }, { "grad_norm": 1.240704894065857, "learning_rate": 8.59e-05, "loss": 0.1038, "step": 860 }, { "grad_norm": 1.3105555772781372, "learning_rate": 8.69e-05, "loss": 0.1065, "step": 870 }, { "grad_norm": 1.3029900789260864, "learning_rate": 8.790000000000001e-05, "loss": 0.1032, "step": 880 }, { "grad_norm": 1.1861451864242554, "learning_rate": 8.89e-05, "loss": 0.1028, "step": 890 }, { "grad_norm": 1.3183419704437256, "learning_rate": 8.99e-05, "loss": 0.1032, "step": 900 }, { "grad_norm": 1.0315393209457397, "learning_rate": 9.090000000000001e-05, "loss": 0.0994, "step": 910 }, { "grad_norm": 1.1772143840789795, "learning_rate": 9.190000000000001e-05, "loss": 0.1014, "step": 920 }, { "grad_norm": 1.1044213771820068, "learning_rate": 9.290000000000001e-05, "loss": 0.0996, "step": 930 }, { "grad_norm": 1.0565844774246216, "learning_rate": 9.39e-05, "loss": 0.0997, "step": 940 }, { "grad_norm": 1.066573143005371, "learning_rate": 9.49e-05, "loss": 0.0996, "step": 950 }, { "grad_norm": 1.405110239982605, "learning_rate": 9.59e-05, "loss": 0.0963, "step": 960 }, { "grad_norm": 1.2910921573638916, "learning_rate": 9.69e-05, "loss": 0.0944, "step": 970 }, { "grad_norm": 1.1059085130691528, "learning_rate": 9.790000000000001e-05, "loss": 0.0967, "step": 980 }, { "grad_norm": 1.2163500785827637, "learning_rate": 9.89e-05, "loss": 0.0947, "step": 990 }, { "grad_norm": 0.8789966106414795, "learning_rate": 9.99e-05, "loss": 0.1027, "step": 1000 }, { "grad_norm": 1.1948117017745972, "learning_rate": 9.999994463727085e-05, "loss": 0.098, "step": 1010 }, { "grad_norm": 1.0235472917556763, "learning_rate": 9.999975326009292e-05, "loss": 0.0924, "step": 1020 }, { "grad_norm": 1.262692928314209, "learning_rate": 9.999942518549879e-05, "loss": 0.0937, "step": 1030 }, { "grad_norm": 1.0914654731750488, "learning_rate": 9.999896041438544e-05, "loss": 0.0937, "step": 1040 }, { "grad_norm": 0.9506348967552185, "learning_rate": 9.999835894802353e-05, "loss": 0.09, "step": 1050 }, { "grad_norm": 1.0387773513793945, "learning_rate": 9.999762078805743e-05, "loss": 0.0891, "step": 1060 }, { "grad_norm": 1.1099857091903687, "learning_rate": 9.999674593650526e-05, "loss": 0.0898, "step": 1070 }, { "grad_norm": 0.9533413648605347, "learning_rate": 9.99957343957588e-05, "loss": 0.09, "step": 1080 }, { "grad_norm": 0.8956478238105774, "learning_rate": 9.99945861685836e-05, "loss": 0.0866, "step": 1090 }, { "grad_norm": 0.8327865600585938, "learning_rate": 9.999330125811884e-05, "loss": 0.0911, "step": 1100 }, { "grad_norm": 1.0309516191482544, "learning_rate": 9.999187966787744e-05, "loss": 0.0888, "step": 1110 }, { "grad_norm": 1.0671669244766235, "learning_rate": 9.999032140174595e-05, "loss": 0.0853, "step": 1120 }, { "grad_norm": 1.0329535007476807, "learning_rate": 9.998862646398464e-05, "loss": 0.0847, "step": 1130 }, { "grad_norm": 1.174817442893982, "learning_rate": 9.998679485922739e-05, "loss": 0.0867, "step": 1140 }, { "grad_norm": 0.9075472950935364, "learning_rate": 9.998482659248174e-05, "loss": 0.0866, "step": 1150 }, { "grad_norm": 0.9608914852142334, "learning_rate": 9.998272166912883e-05, "loss": 0.0853, "step": 1160 }, { "grad_norm": 0.9264180064201355, "learning_rate": 9.998048009492347e-05, "loss": 0.0855, "step": 1170 }, { "grad_norm": 0.8515126705169678, "learning_rate": 9.997810187599403e-05, "loss": 0.0873, "step": 1180 }, { "grad_norm": 1.0085526704788208, "learning_rate": 9.997558701884249e-05, "loss": 0.084, "step": 1190 }, { "grad_norm": 0.8803794384002686, "learning_rate": 9.997293553034433e-05, "loss": 0.0857, "step": 1200 }, { "grad_norm": 0.8993883728981018, "learning_rate": 9.997014741774866e-05, "loss": 0.085, "step": 1210 }, { "grad_norm": 0.9325219988822937, "learning_rate": 9.996722268867803e-05, "loss": 0.08, "step": 1220 }, { "grad_norm": 0.7938389778137207, "learning_rate": 9.996416135112858e-05, "loss": 0.0804, "step": 1230 }, { "grad_norm": 1.0101356506347656, "learning_rate": 9.996096341346988e-05, "loss": 0.082, "step": 1240 }, { "grad_norm": 0.7713646292686462, "learning_rate": 9.995762888444495e-05, "loss": 0.0825, "step": 1250 }, { "grad_norm": 0.8901200890541077, "learning_rate": 9.995415777317027e-05, "loss": 0.0802, "step": 1260 }, { "grad_norm": 0.8909335732460022, "learning_rate": 9.995055008913574e-05, "loss": 0.078, "step": 1270 }, { "grad_norm": 0.8464662432670593, "learning_rate": 9.994680584220463e-05, "loss": 0.0768, "step": 1280 }, { "grad_norm": 0.8395467400550842, "learning_rate": 9.994292504261355e-05, "loss": 0.0803, "step": 1290 }, { "grad_norm": 0.9191763401031494, "learning_rate": 9.993890770097247e-05, "loss": 0.0778, "step": 1300 }, { "grad_norm": 0.9382724165916443, "learning_rate": 9.993475382826467e-05, "loss": 0.0778, "step": 1310 }, { "grad_norm": 0.8167632818222046, "learning_rate": 9.993046343584664e-05, "loss": 0.0768, "step": 1320 }, { "grad_norm": 0.964731752872467, "learning_rate": 9.992603653544816e-05, "loss": 0.0749, "step": 1330 }, { "grad_norm": 0.8039717674255371, "learning_rate": 9.992147313917222e-05, "loss": 0.0731, "step": 1340 }, { "grad_norm": 0.9752305746078491, "learning_rate": 9.991677325949497e-05, "loss": 0.0746, "step": 1350 }, { "grad_norm": 0.7187494039535522, "learning_rate": 9.991193690926568e-05, "loss": 0.0735, "step": 1360 }, { "grad_norm": 0.8025557398796082, "learning_rate": 9.990696410170678e-05, "loss": 0.0732, "step": 1370 }, { "grad_norm": 0.7585386633872986, "learning_rate": 9.990185485041371e-05, "loss": 0.0753, "step": 1380 }, { "grad_norm": 0.8446924090385437, "learning_rate": 9.989660916935498e-05, "loss": 0.0752, "step": 1390 }, { "grad_norm": 0.8321685791015625, "learning_rate": 9.989122707287208e-05, "loss": 0.0774, "step": 1400 }, { "grad_norm": 0.9697508215904236, "learning_rate": 9.988570857567945e-05, "loss": 0.0764, "step": 1410 }, { "grad_norm": 0.8045315742492676, "learning_rate": 9.988005369286446e-05, "loss": 0.0766, "step": 1420 }, { "grad_norm": 0.7006632089614868, "learning_rate": 9.987426243988734e-05, "loss": 0.0729, "step": 1430 }, { "grad_norm": 0.7440106272697449, "learning_rate": 9.986833483258114e-05, "loss": 0.0745, "step": 1440 }, { "grad_norm": 0.8354970812797546, "learning_rate": 9.986227088715173e-05, "loss": 0.0737, "step": 1450 }, { "grad_norm": 0.6917275190353394, "learning_rate": 9.98560706201777e-05, "loss": 0.0712, "step": 1460 }, { "grad_norm": 0.7299842834472656, "learning_rate": 9.984973404861036e-05, "loss": 0.0718, "step": 1470 }, { "grad_norm": 0.7468602657318115, "learning_rate": 9.984326118977361e-05, "loss": 0.0731, "step": 1480 }, { "grad_norm": 0.8825043439865112, "learning_rate": 9.983665206136406e-05, "loss": 0.0738, "step": 1490 }, { "grad_norm": 0.8674785494804382, "learning_rate": 9.982990668145075e-05, "loss": 0.0705, "step": 1500 }, { "grad_norm": 0.8119779229164124, "learning_rate": 9.982302506847534e-05, "loss": 0.0744, "step": 1510 }, { "grad_norm": 0.7920791506767273, "learning_rate": 9.981600724125189e-05, "loss": 0.0719, "step": 1520 }, { "grad_norm": 0.740151047706604, "learning_rate": 9.980885321896685e-05, "loss": 0.0735, "step": 1530 }, { "grad_norm": 0.7403775453567505, "learning_rate": 9.980156302117905e-05, "loss": 0.071, "step": 1540 }, { "grad_norm": 0.8703851699829102, "learning_rate": 9.979413666781963e-05, "loss": 0.0694, "step": 1550 }, { "grad_norm": 0.7278032302856445, "learning_rate": 9.978657417919193e-05, "loss": 0.0708, "step": 1560 }, { "grad_norm": 0.8014971017837524, "learning_rate": 9.977887557597153e-05, "loss": 0.0646, "step": 1570 }, { "grad_norm": 0.7528615593910217, "learning_rate": 9.97710408792061e-05, "loss": 0.0682, "step": 1580 }, { "grad_norm": 0.7048172354698181, "learning_rate": 9.976307011031542e-05, "loss": 0.0694, "step": 1590 }, { "grad_norm": 0.7487463355064392, "learning_rate": 9.975496329109126e-05, "loss": 0.07, "step": 1600 }, { "grad_norm": 0.8928484916687012, "learning_rate": 9.974672044369732e-05, "loss": 0.0736, "step": 1610 }, { "grad_norm": 0.8624300956726074, "learning_rate": 9.97383415906693e-05, "loss": 0.0692, "step": 1620 }, { "grad_norm": 0.7228650450706482, "learning_rate": 9.97298267549146e-05, "loss": 0.0719, "step": 1630 }, { "grad_norm": 0.9539430737495422, "learning_rate": 9.972117595971249e-05, "loss": 0.0692, "step": 1640 }, { "grad_norm": 0.6346696019172668, "learning_rate": 9.971238922871391e-05, "loss": 0.0688, "step": 1650 }, { "grad_norm": 0.7980417013168335, "learning_rate": 9.970346658594142e-05, "loss": 0.0747, "step": 1660 }, { "grad_norm": 0.8892179131507874, "learning_rate": 9.969440805578923e-05, "loss": 0.0692, "step": 1670 }, { "grad_norm": 0.8207274079322815, "learning_rate": 9.968521366302298e-05, "loss": 0.0677, "step": 1680 }, { "grad_norm": 0.8967740535736084, "learning_rate": 9.967588343277981e-05, "loss": 0.071, "step": 1690 }, { "grad_norm": 0.680429995059967, "learning_rate": 9.966641739056818e-05, "loss": 0.0671, "step": 1700 }, { "grad_norm": 0.5896378755569458, "learning_rate": 9.965681556226793e-05, "loss": 0.0695, "step": 1710 }, { "grad_norm": 0.6645721197128296, "learning_rate": 9.964707797413006e-05, "loss": 0.067, "step": 1720 }, { "grad_norm": 0.8198447227478027, "learning_rate": 9.963720465277679e-05, "loss": 0.067, "step": 1730 }, { "grad_norm": 0.7200887203216553, "learning_rate": 9.96271956252014e-05, "loss": 0.0658, "step": 1740 }, { "grad_norm": 0.7971823215484619, "learning_rate": 9.961705091876816e-05, "loss": 0.0649, "step": 1750 }, { "grad_norm": 0.7375441789627075, "learning_rate": 9.960677056121235e-05, "loss": 0.0657, "step": 1760 }, { "grad_norm": 0.7519879341125488, "learning_rate": 9.959635458064005e-05, "loss": 0.0646, "step": 1770 }, { "grad_norm": 0.7177464962005615, "learning_rate": 9.958580300552815e-05, "loss": 0.0659, "step": 1780 }, { "grad_norm": 0.8185701966285706, "learning_rate": 9.957511586472426e-05, "loss": 0.0645, "step": 1790 }, { "grad_norm": 0.7226177453994751, "learning_rate": 9.956429318744662e-05, "loss": 0.0664, "step": 1800 }, { "grad_norm": 0.6147686839103699, "learning_rate": 9.955333500328404e-05, "loss": 0.0663, "step": 1810 }, { "grad_norm": 0.8226826786994934, "learning_rate": 9.95422413421957e-05, "loss": 0.0654, "step": 1820 }, { "grad_norm": 0.6308465600013733, "learning_rate": 9.953101223451133e-05, "loss": 0.0671, "step": 1830 }, { "grad_norm": 0.6037870645523071, "learning_rate": 9.951964771093085e-05, "loss": 0.0652, "step": 1840 }, { "grad_norm": 0.679835319519043, "learning_rate": 9.950814780252442e-05, "loss": 0.067, "step": 1850 }, { "grad_norm": 0.6414503455162048, "learning_rate": 9.949651254073236e-05, "loss": 0.0625, "step": 1860 }, { "grad_norm": 0.6243733167648315, "learning_rate": 9.948474195736504e-05, "loss": 0.0631, "step": 1870 }, { "grad_norm": 0.7260075807571411, "learning_rate": 9.947283608460277e-05, "loss": 0.0616, "step": 1880 }, { "grad_norm": 0.6100785732269287, "learning_rate": 9.946079495499577e-05, "loss": 0.0657, "step": 1890 }, { "grad_norm": 0.6928042769432068, "learning_rate": 9.944861860146401e-05, "loss": 0.0619, "step": 1900 }, { "grad_norm": 0.6407603621482849, "learning_rate": 9.943630705729719e-05, "loss": 0.0636, "step": 1910 }, { "grad_norm": 0.6835041642189026, "learning_rate": 9.942386035615459e-05, "loss": 0.0659, "step": 1920 }, { "grad_norm": 0.674142599105835, "learning_rate": 9.941127853206503e-05, "loss": 0.0635, "step": 1930 }, { "grad_norm": 0.6852268576622009, "learning_rate": 9.939856161942673e-05, "loss": 0.0624, "step": 1940 }, { "grad_norm": 0.6523897647857666, "learning_rate": 9.938570965300724e-05, "loss": 0.0623, "step": 1950 }, { "grad_norm": 0.6527503132820129, "learning_rate": 9.937272266794335e-05, "loss": 0.0611, "step": 1960 }, { "grad_norm": 0.6659324169158936, "learning_rate": 9.935960069974096e-05, "loss": 0.0636, "step": 1970 }, { "grad_norm": 0.6519555449485779, "learning_rate": 9.934634378427506e-05, "loss": 0.0653, "step": 1980 }, { "grad_norm": 0.7410769462585449, "learning_rate": 9.933295195778954e-05, "loss": 0.0667, "step": 1990 }, { "grad_norm": 0.5415292382240295, "learning_rate": 9.931942525689715e-05, "loss": 0.0678, "step": 2000 }, { "grad_norm": 0.5994923710823059, "learning_rate": 9.930576371857936e-05, "loss": 0.0628, "step": 2010 }, { "grad_norm": 0.670676589012146, "learning_rate": 9.929196738018629e-05, "loss": 0.0624, "step": 2020 }, { "grad_norm": 0.5595282912254333, "learning_rate": 9.927803627943662e-05, "loss": 0.0614, "step": 2030 }, { "grad_norm": 0.5828508138656616, "learning_rate": 9.926397045441744e-05, "loss": 0.0604, "step": 2040 }, { "grad_norm": 0.5831713676452637, "learning_rate": 9.924976994358417e-05, "loss": 0.0605, "step": 2050 }, { "grad_norm": 0.6327888369560242, "learning_rate": 9.923543478576048e-05, "loss": 0.0592, "step": 2060 }, { "grad_norm": 0.6445356607437134, "learning_rate": 9.922096502013813e-05, "loss": 0.0605, "step": 2070 }, { "grad_norm": 0.5931155681610107, "learning_rate": 9.92063606862769e-05, "loss": 0.0608, "step": 2080 }, { "grad_norm": 0.601141095161438, "learning_rate": 9.919162182410453e-05, "loss": 0.059, "step": 2090 }, { "grad_norm": 0.559126615524292, "learning_rate": 9.917674847391645e-05, "loss": 0.0571, "step": 2100 }, { "grad_norm": 0.6964930295944214, "learning_rate": 9.916174067637584e-05, "loss": 0.0592, "step": 2110 }, { "grad_norm": 0.7474396824836731, "learning_rate": 9.914659847251348e-05, "loss": 0.0628, "step": 2120 }, { "grad_norm": 0.6328505277633667, "learning_rate": 9.913132190372753e-05, "loss": 0.0616, "step": 2130 }, { "grad_norm": 0.6215328574180603, "learning_rate": 9.911591101178359e-05, "loss": 0.0605, "step": 2140 }, { "grad_norm": 0.8151555061340332, "learning_rate": 9.910036583881443e-05, "loss": 0.0622, "step": 2150 }, { "grad_norm": 0.6360267400741577, "learning_rate": 9.908468642731995e-05, "loss": 0.0633, "step": 2160 }, { "grad_norm": 0.6325832605361938, "learning_rate": 9.906887282016707e-05, "loss": 0.0646, "step": 2170 }, { "grad_norm": 0.616167426109314, "learning_rate": 9.90529250605896e-05, "loss": 0.0586, "step": 2180 }, { "grad_norm": 0.5955585837364197, "learning_rate": 9.903684319218809e-05, "loss": 0.0586, "step": 2190 }, { "grad_norm": 0.712861180305481, "learning_rate": 9.902062725892976e-05, "loss": 0.0616, "step": 2200 }, { "grad_norm": 0.6084709167480469, "learning_rate": 9.900427730514834e-05, "loss": 0.0613, "step": 2210 }, { "grad_norm": 0.5973120331764221, "learning_rate": 9.8987793375544e-05, "loss": 0.0607, "step": 2220 }, { "grad_norm": 0.6146500706672668, "learning_rate": 9.897117551518318e-05, "loss": 0.0608, "step": 2230 }, { "grad_norm": 0.5947669148445129, "learning_rate": 9.895442376949844e-05, "loss": 0.0619, "step": 2240 }, { "grad_norm": 0.6229901909828186, "learning_rate": 9.893753818428845e-05, "loss": 0.0632, "step": 2250 }, { "grad_norm": 0.5743619203567505, "learning_rate": 9.892051880571773e-05, "loss": 0.062, "step": 2260 }, { "grad_norm": 0.7406059503555298, "learning_rate": 9.890336568031663e-05, "loss": 0.0597, "step": 2270 }, { "grad_norm": 0.6080294251441956, "learning_rate": 9.888607885498113e-05, "loss": 0.0575, "step": 2280 }, { "grad_norm": 0.5063527226448059, "learning_rate": 9.886865837697275e-05, "loss": 0.0575, "step": 2290 }, { "grad_norm": 0.5277106165885925, "learning_rate": 9.88511042939184e-05, "loss": 0.0588, "step": 2300 }, { "grad_norm": 0.6051730513572693, "learning_rate": 9.883341665381028e-05, "loss": 0.058, "step": 2310 }, { "grad_norm": 0.5652435421943665, "learning_rate": 9.881559550500575e-05, "loss": 0.0605, "step": 2320 }, { "grad_norm": 0.5267813801765442, "learning_rate": 9.879764089622712e-05, "loss": 0.0569, "step": 2330 }, { "grad_norm": 0.5755409002304077, "learning_rate": 9.87795528765616e-05, "loss": 0.0605, "step": 2340 }, { "grad_norm": 0.6392557621002197, "learning_rate": 9.876133149546118e-05, "loss": 0.0589, "step": 2350 }, { "grad_norm": 0.5705186724662781, "learning_rate": 9.874297680274238e-05, "loss": 0.0583, "step": 2360 }, { "grad_norm": 0.6585530042648315, "learning_rate": 9.872448884858624e-05, "loss": 0.0588, "step": 2370 }, { "grad_norm": 0.6803420782089233, "learning_rate": 9.870586768353815e-05, "loss": 0.0567, "step": 2380 }, { "grad_norm": 0.7465255856513977, "learning_rate": 9.868711335850764e-05, "loss": 0.0561, "step": 2390 }, { "grad_norm": 0.5848642587661743, "learning_rate": 9.866822592476833e-05, "loss": 0.0588, "step": 2400 }, { "grad_norm": 0.5381805300712585, "learning_rate": 9.86492054339577e-05, "loss": 0.0573, "step": 2410 }, { "grad_norm": 0.6431410908699036, "learning_rate": 9.863005193807711e-05, "loss": 0.0564, "step": 2420 }, { "grad_norm": 0.5730747580528259, "learning_rate": 9.861076548949143e-05, "loss": 0.0575, "step": 2430 }, { "grad_norm": 0.5729357600212097, "learning_rate": 9.859134614092912e-05, "loss": 0.0577, "step": 2440 }, { "grad_norm": 0.5879770517349243, "learning_rate": 9.857179394548191e-05, "loss": 0.0571, "step": 2450 }, { "grad_norm": 0.5518467426300049, "learning_rate": 9.855210895660477e-05, "loss": 0.0546, "step": 2460 }, { "grad_norm": 0.6395381689071655, "learning_rate": 9.853229122811568e-05, "loss": 0.0569, "step": 2470 }, { "grad_norm": 0.6454721093177795, "learning_rate": 9.851234081419559e-05, "loss": 0.0587, "step": 2480 }, { "grad_norm": 0.6770858764648438, "learning_rate": 9.849225776938814e-05, "loss": 0.0586, "step": 2490 }, { "grad_norm": 0.6783500909805298, "learning_rate": 9.847204214859964e-05, "loss": 0.056, "step": 2500 }, { "grad_norm": 0.7126483917236328, "learning_rate": 9.845169400709879e-05, "loss": 0.0565, "step": 2510 }, { "grad_norm": 0.5490720272064209, "learning_rate": 9.843121340051664e-05, "loss": 0.0574, "step": 2520 }, { "grad_norm": 0.6126254796981812, "learning_rate": 9.841060038484641e-05, "loss": 0.0569, "step": 2530 }, { "grad_norm": 0.549883246421814, "learning_rate": 9.838985501644328e-05, "loss": 0.0519, "step": 2540 }, { "grad_norm": 0.610352635383606, "learning_rate": 9.83689773520243e-05, "loss": 0.0564, "step": 2550 }, { "grad_norm": 0.588969349861145, "learning_rate": 9.834796744866819e-05, "loss": 0.0544, "step": 2560 }, { "grad_norm": 0.6367093920707703, "learning_rate": 9.832682536381525e-05, "loss": 0.056, "step": 2570 }, { "grad_norm": 0.696803092956543, "learning_rate": 9.830555115526711e-05, "loss": 0.0565, "step": 2580 }, { "grad_norm": 0.67190021276474, "learning_rate": 9.828414488118667e-05, "loss": 0.0576, "step": 2590 }, { "grad_norm": 0.5457078814506531, "learning_rate": 9.826260660009785e-05, "loss": 0.0555, "step": 2600 }, { "grad_norm": 0.5594436526298523, "learning_rate": 9.824093637088547e-05, "loss": 0.0528, "step": 2610 }, { "grad_norm": 0.5211487412452698, "learning_rate": 9.821913425279514e-05, "loss": 0.0538, "step": 2620 }, { "grad_norm": 0.5177614092826843, "learning_rate": 9.8197200305433e-05, "loss": 0.0545, "step": 2630 }, { "grad_norm": 0.5182483792304993, "learning_rate": 9.817513458876564e-05, "loss": 0.0556, "step": 2640 }, { "grad_norm": 0.6016985774040222, "learning_rate": 9.815293716311987e-05, "loss": 0.0527, "step": 2650 }, { "grad_norm": 0.5641167759895325, "learning_rate": 9.813060808918262e-05, "loss": 0.0561, "step": 2660 }, { "grad_norm": 0.6382514834403992, "learning_rate": 9.810814742800069e-05, "loss": 0.0544, "step": 2670 }, { "grad_norm": 0.7021539211273193, "learning_rate": 9.808555524098074e-05, "loss": 0.0559, "step": 2680 }, { "grad_norm": 0.5652006268501282, "learning_rate": 9.806283158988887e-05, "loss": 0.0544, "step": 2690 }, { "grad_norm": 0.5768893361091614, "learning_rate": 9.803997653685072e-05, "loss": 0.0516, "step": 2700 }, { "grad_norm": 0.6968298554420471, "learning_rate": 9.801699014435112e-05, "loss": 0.0534, "step": 2710 }, { "grad_norm": 0.618663489818573, "learning_rate": 9.799387247523398e-05, "loss": 0.0535, "step": 2720 }, { "grad_norm": 0.5248475670814514, "learning_rate": 9.797062359270215e-05, "loss": 0.0529, "step": 2730 }, { "grad_norm": 0.6241867542266846, "learning_rate": 9.794724356031715e-05, "loss": 0.0536, "step": 2740 }, { "grad_norm": 0.6985629200935364, "learning_rate": 9.792373244199913e-05, "loss": 0.0558, "step": 2750 }, { "grad_norm": 0.6230672001838684, "learning_rate": 9.790009030202658e-05, "loss": 0.0537, "step": 2760 }, { "grad_norm": 0.5703837871551514, "learning_rate": 9.78763172050362e-05, "loss": 0.0545, "step": 2770 }, { "grad_norm": 0.5615924596786499, "learning_rate": 9.785241321602274e-05, "loss": 0.0535, "step": 2780 }, { "grad_norm": 0.6150487065315247, "learning_rate": 9.782837840033879e-05, "loss": 0.0541, "step": 2790 }, { "grad_norm": 0.6442948579788208, "learning_rate": 9.780421282369461e-05, "loss": 0.0566, "step": 2800 }, { "grad_norm": 0.5317158102989197, "learning_rate": 9.777991655215797e-05, "loss": 0.0564, "step": 2810 }, { "grad_norm": 0.5907642245292664, "learning_rate": 9.775548965215394e-05, "loss": 0.0521, "step": 2820 }, { "grad_norm": 0.5747297406196594, "learning_rate": 9.773093219046474e-05, "loss": 0.0495, "step": 2830 }, { "grad_norm": 0.6061862111091614, "learning_rate": 9.770624423422954e-05, "loss": 0.0549, "step": 2840 }, { "grad_norm": 0.5587713122367859, "learning_rate": 9.768142585094426e-05, "loss": 0.056, "step": 2850 }, { "grad_norm": 0.646105170249939, "learning_rate": 9.765647710846142e-05, "loss": 0.0518, "step": 2860 }, { "grad_norm": 0.5185126066207886, "learning_rate": 9.763139807498991e-05, "loss": 0.0537, "step": 2870 }, { "grad_norm": 0.5357843637466431, "learning_rate": 9.760618881909487e-05, "loss": 0.0534, "step": 2880 }, { "grad_norm": 0.5286585688591003, "learning_rate": 9.758084940969744e-05, "loss": 0.0562, "step": 2890 }, { "grad_norm": 0.5728244781494141, "learning_rate": 9.755537991607459e-05, "loss": 0.0521, "step": 2900 }, { "grad_norm": 0.6184394955635071, "learning_rate": 9.752978040785895e-05, "loss": 0.0574, "step": 2910 }, { "grad_norm": 0.5012248754501343, "learning_rate": 9.750405095503859e-05, "loss": 0.0528, "step": 2920 }, { "grad_norm": 0.5458285212516785, "learning_rate": 9.747819162795686e-05, "loss": 0.0524, "step": 2930 }, { "grad_norm": 0.6669784188270569, "learning_rate": 9.745220249731217e-05, "loss": 0.0554, "step": 2940 }, { "grad_norm": 0.5164629817008972, "learning_rate": 9.742608363415781e-05, "loss": 0.0519, "step": 2950 }, { "grad_norm": 0.5527239441871643, "learning_rate": 9.739983510990176e-05, "loss": 0.0553, "step": 2960 }, { "grad_norm": 0.6122348308563232, "learning_rate": 9.737345699630647e-05, "loss": 0.0532, "step": 2970 }, { "grad_norm": 0.5157252550125122, "learning_rate": 9.734694936548869e-05, "loss": 0.0533, "step": 2980 }, { "grad_norm": 0.5645245313644409, "learning_rate": 9.732031228991932e-05, "loss": 0.0577, "step": 2990 }, { "grad_norm": 0.7678781747817993, "learning_rate": 9.729354584242302e-05, "loss": 0.0547, "step": 3000 }, { "grad_norm": 0.6804419755935669, "learning_rate": 9.726665009617832e-05, "loss": 0.0559, "step": 3010 }, { "grad_norm": 0.5836779475212097, "learning_rate": 9.723962512471714e-05, "loss": 0.0548, "step": 3020 }, { "grad_norm": 0.5465885996818542, "learning_rate": 9.72124710019247e-05, "loss": 0.0562, "step": 3030 }, { "grad_norm": 0.5152290463447571, "learning_rate": 9.718518780203934e-05, "loss": 0.0521, "step": 3040 }, { "grad_norm": 0.5115591883659363, "learning_rate": 9.715777559965228e-05, "loss": 0.0525, "step": 3050 }, { "grad_norm": 0.49400827288627625, "learning_rate": 9.713023446970746e-05, "loss": 0.0521, "step": 3060 }, { "grad_norm": 0.6357924342155457, "learning_rate": 9.710256448750126e-05, "loss": 0.05, "step": 3070 }, { "grad_norm": 0.6152380108833313, "learning_rate": 9.707476572868235e-05, "loss": 0.0542, "step": 3080 }, { "grad_norm": 0.6186644434928894, "learning_rate": 9.704683826925149e-05, "loss": 0.0519, "step": 3090 }, { "grad_norm": 0.5993762612342834, "learning_rate": 9.701878218556129e-05, "loss": 0.0501, "step": 3100 }, { "grad_norm": 0.578052818775177, "learning_rate": 9.699059755431598e-05, "loss": 0.0509, "step": 3110 }, { "grad_norm": 0.5247306227684021, "learning_rate": 9.696228445257132e-05, "loss": 0.051, "step": 3120 }, { "grad_norm": 0.6611071825027466, "learning_rate": 9.693384295773419e-05, "loss": 0.0524, "step": 3130 }, { "grad_norm": 0.5699040293693542, "learning_rate": 9.690527314756259e-05, "loss": 0.0526, "step": 3140 }, { "grad_norm": 0.5631635785102844, "learning_rate": 9.687657510016527e-05, "loss": 0.0507, "step": 3150 }, { "grad_norm": 0.4601803123950958, "learning_rate": 9.684774889400161e-05, "loss": 0.05, "step": 3160 }, { "grad_norm": 0.5025836825370789, "learning_rate": 9.681879460788135e-05, "loss": 0.049, "step": 3170 }, { "grad_norm": 0.5727059245109558, "learning_rate": 9.67897123209644e-05, "loss": 0.0526, "step": 3180 }, { "grad_norm": 0.649380087852478, "learning_rate": 9.676050211276062e-05, "loss": 0.0537, "step": 3190 }, { "grad_norm": 0.5575081706047058, "learning_rate": 9.673116406312962e-05, "loss": 0.0539, "step": 3200 }, { "grad_norm": 0.4866867959499359, "learning_rate": 9.67016982522805e-05, "loss": 0.0487, "step": 3210 }, { "grad_norm": 0.5518966913223267, "learning_rate": 9.667210476077164e-05, "loss": 0.0509, "step": 3220 }, { "grad_norm": 0.5296523571014404, "learning_rate": 9.664238366951055e-05, "loss": 0.0497, "step": 3230 }, { "grad_norm": 0.5967239141464233, "learning_rate": 9.661253505975355e-05, "loss": 0.0524, "step": 3240 }, { "grad_norm": 0.5917206406593323, "learning_rate": 9.658255901310557e-05, "loss": 0.0522, "step": 3250 }, { "grad_norm": 0.5756188035011292, "learning_rate": 9.655245561152e-05, "loss": 0.0504, "step": 3260 }, { "grad_norm": 0.5686755776405334, "learning_rate": 9.65222249372984e-05, "loss": 0.0536, "step": 3270 }, { "grad_norm": 0.6282801628112793, "learning_rate": 9.649186707309026e-05, "loss": 0.052, "step": 3280 }, { "grad_norm": 0.5592830777168274, "learning_rate": 9.646138210189283e-05, "loss": 0.0492, "step": 3290 }, { "grad_norm": 0.6329211592674255, "learning_rate": 9.643077010705087e-05, "loss": 0.0525, "step": 3300 }, { "grad_norm": 0.5022810697555542, "learning_rate": 9.640003117225637e-05, "loss": 0.0514, "step": 3310 }, { "grad_norm": 0.6291565895080566, "learning_rate": 9.636916538154846e-05, "loss": 0.0512, "step": 3320 }, { "grad_norm": 0.491485059261322, "learning_rate": 9.633817281931296e-05, "loss": 0.0524, "step": 3330 }, { "grad_norm": 0.5268076658248901, "learning_rate": 9.630705357028242e-05, "loss": 0.0517, "step": 3340 }, { "grad_norm": 0.5763431787490845, "learning_rate": 9.627580771953563e-05, "loss": 0.0496, "step": 3350 }, { "grad_norm": 0.5327632427215576, "learning_rate": 9.624443535249759e-05, "loss": 0.0538, "step": 3360 }, { "grad_norm": 0.5271499156951904, "learning_rate": 9.621293655493913e-05, "loss": 0.0493, "step": 3370 }, { "grad_norm": 0.5392346978187561, "learning_rate": 9.618131141297675e-05, "loss": 0.048, "step": 3380 }, { "grad_norm": 0.557559072971344, "learning_rate": 9.614956001307242e-05, "loss": 0.0495, "step": 3390 }, { "grad_norm": 0.514748215675354, "learning_rate": 9.611768244203321e-05, "loss": 0.051, "step": 3400 }, { "grad_norm": 0.6630014181137085, "learning_rate": 9.60856787870112e-05, "loss": 0.0477, "step": 3410 }, { "grad_norm": 0.48811548948287964, "learning_rate": 9.605354913550318e-05, "loss": 0.0524, "step": 3420 }, { "grad_norm": 0.47521334886550903, "learning_rate": 9.602129357535037e-05, "loss": 0.0495, "step": 3430 }, { "grad_norm": 0.46967071294784546, "learning_rate": 9.598891219473825e-05, "loss": 0.0519, "step": 3440 }, { "grad_norm": 0.5551022887229919, "learning_rate": 9.595640508219625e-05, "loss": 0.0468, "step": 3450 }, { "grad_norm": 0.5609787106513977, "learning_rate": 9.592377232659761e-05, "loss": 0.0506, "step": 3460 }, { "grad_norm": 0.4883131980895996, "learning_rate": 9.589101401715904e-05, "loss": 0.0498, "step": 3470 }, { "grad_norm": 0.5301612615585327, "learning_rate": 9.585813024344045e-05, "loss": 0.0509, "step": 3480 }, { "grad_norm": 0.5042786002159119, "learning_rate": 9.58251210953449e-05, "loss": 0.0485, "step": 3490 }, { "grad_norm": 0.5025367140769958, "learning_rate": 9.579198666311809e-05, "loss": 0.0498, "step": 3500 }, { "grad_norm": 0.5075255632400513, "learning_rate": 9.575872703734832e-05, "loss": 0.0487, "step": 3510 }, { "grad_norm": 0.43919530510902405, "learning_rate": 9.572534230896611e-05, "loss": 0.0465, "step": 3520 }, { "grad_norm": 0.5095721483230591, "learning_rate": 9.569183256924403e-05, "loss": 0.0487, "step": 3530 }, { "grad_norm": 0.512020468711853, "learning_rate": 9.565819790979646e-05, "loss": 0.0486, "step": 3540 }, { "grad_norm": 0.5008028149604797, "learning_rate": 9.562443842257925e-05, "loss": 0.0511, "step": 3550 }, { "grad_norm": 0.5364949107170105, "learning_rate": 9.559055419988956e-05, "loss": 0.0512, "step": 3560 }, { "grad_norm": 0.5370563864707947, "learning_rate": 9.555654533436557e-05, "loss": 0.0499, "step": 3570 }, { "grad_norm": 0.5545650124549866, "learning_rate": 9.552241191898621e-05, "loss": 0.0493, "step": 3580 }, { "grad_norm": 0.5982073545455933, "learning_rate": 9.548815404707092e-05, "loss": 0.0496, "step": 3590 }, { "grad_norm": 0.5401698350906372, "learning_rate": 9.545377181227942e-05, "loss": 0.0489, "step": 3600 }, { "grad_norm": 0.43389976024627686, "learning_rate": 9.541926530861145e-05, "loss": 0.0464, "step": 3610 }, { "grad_norm": 0.608935534954071, "learning_rate": 9.538463463040645e-05, "loss": 0.0467, "step": 3620 }, { "grad_norm": 0.4709791839122772, "learning_rate": 9.534987987234337e-05, "loss": 0.048, "step": 3630 }, { "grad_norm": 0.4395516812801361, "learning_rate": 9.53150011294404e-05, "loss": 0.0496, "step": 3640 }, { "grad_norm": 0.5766939520835876, "learning_rate": 9.527999849705471e-05, "loss": 0.0462, "step": 3650 }, { "grad_norm": 0.4720335602760315, "learning_rate": 9.524487207088213e-05, "loss": 0.0466, "step": 3660 }, { "grad_norm": 0.5506165027618408, "learning_rate": 9.520962194695698e-05, "loss": 0.0471, "step": 3670 }, { "grad_norm": 0.540799617767334, "learning_rate": 9.517424822165175e-05, "loss": 0.0471, "step": 3680 }, { "grad_norm": 0.49528366327285767, "learning_rate": 9.513875099167685e-05, "loss": 0.0483, "step": 3690 }, { "grad_norm": 0.5733321309089661, "learning_rate": 9.510313035408035e-05, "loss": 0.0471, "step": 3700 }, { "grad_norm": 0.5157331228256226, "learning_rate": 9.506738640624775e-05, "loss": 0.0474, "step": 3710 }, { "grad_norm": 0.5828023552894592, "learning_rate": 9.50315192459016e-05, "loss": 0.0485, "step": 3720 }, { "grad_norm": 0.5534822344779968, "learning_rate": 9.499552897110136e-05, "loss": 0.0497, "step": 3730 }, { "grad_norm": 0.4209916293621063, "learning_rate": 9.495941568024304e-05, "loss": 0.0491, "step": 3740 }, { "grad_norm": 0.4925754964351654, "learning_rate": 9.492317947205904e-05, "loss": 0.048, "step": 3750 }, { "grad_norm": 0.4909178912639618, "learning_rate": 9.488682044561775e-05, "loss": 0.0466, "step": 3760 }, { "grad_norm": 0.5199701189994812, "learning_rate": 9.485033870032335e-05, "loss": 0.0495, "step": 3770 }, { "grad_norm": 0.44645825028419495, "learning_rate": 9.481373433591556e-05, "loss": 0.0493, "step": 3780 }, { "grad_norm": 0.6020624041557312, "learning_rate": 9.47770074524693e-05, "loss": 0.0455, "step": 3790 }, { "grad_norm": 0.488620400428772, "learning_rate": 9.474015815039446e-05, "loss": 0.0491, "step": 3800 }, { "grad_norm": 0.5846852660179138, "learning_rate": 9.470318653043565e-05, "loss": 0.0488, "step": 3810 }, { "grad_norm": 0.518318235874176, "learning_rate": 9.466609269367185e-05, "loss": 0.0487, "step": 3820 }, { "grad_norm": 0.5965957641601562, "learning_rate": 9.46288767415162e-05, "loss": 0.0464, "step": 3830 }, { "grad_norm": 0.5559269189834595, "learning_rate": 9.459153877571567e-05, "loss": 0.0475, "step": 3840 }, { "grad_norm": 0.4260229766368866, "learning_rate": 9.455407889835087e-05, "loss": 0.0488, "step": 3850 }, { "grad_norm": 0.5213231444358826, "learning_rate": 9.451649721183564e-05, "loss": 0.0461, "step": 3860 }, { "grad_norm": 0.4626673460006714, "learning_rate": 9.447879381891692e-05, "loss": 0.0446, "step": 3870 }, { "grad_norm": 0.5997720956802368, "learning_rate": 9.444096882267428e-05, "loss": 0.0494, "step": 3880 }, { "grad_norm": 0.5883204936981201, "learning_rate": 9.440302232651988e-05, "loss": 0.0501, "step": 3890 }, { "grad_norm": 0.4910527765750885, "learning_rate": 9.436495443419795e-05, "loss": 0.0472, "step": 3900 }, { "grad_norm": 0.4996890127658844, "learning_rate": 9.432676524978466e-05, "loss": 0.0455, "step": 3910 }, { "grad_norm": 0.508808970451355, "learning_rate": 9.42884548776878e-05, "loss": 0.0448, "step": 3920 }, { "grad_norm": 0.48599308729171753, "learning_rate": 9.425002342264646e-05, "loss": 0.0444, "step": 3930 }, { "grad_norm": 0.48764660954475403, "learning_rate": 9.421147098973077e-05, "loss": 0.0457, "step": 3940 }, { "grad_norm": 0.5945606231689453, "learning_rate": 9.41727976843416e-05, "loss": 0.0501, "step": 3950 }, { "grad_norm": 0.42392799258232117, "learning_rate": 9.413400361221029e-05, "loss": 0.0466, "step": 3960 }, { "grad_norm": 0.5303824543952942, "learning_rate": 9.409508887939835e-05, "loss": 0.0497, "step": 3970 }, { "grad_norm": 0.5664520263671875, "learning_rate": 9.40560535922972e-05, "loss": 0.0488, "step": 3980 }, { "grad_norm": 0.49102678894996643, "learning_rate": 9.40168978576278e-05, "loss": 0.0461, "step": 3990 }, { "grad_norm": 0.45616212487220764, "learning_rate": 9.397762178244043e-05, "loss": 0.0454, "step": 4000 }, { "grad_norm": 0.3627954423427582, "learning_rate": 9.393822547411439e-05, "loss": 0.0441, "step": 4010 }, { "grad_norm": 0.526990532875061, "learning_rate": 9.389870904035769e-05, "loss": 0.0467, "step": 4020 }, { "grad_norm": 0.5587537288665771, "learning_rate": 9.385907258920672e-05, "loss": 0.0466, "step": 4030 }, { "grad_norm": 0.42573773860931396, "learning_rate": 9.381931622902607e-05, "loss": 0.0449, "step": 4040 }, { "grad_norm": 0.5555974841117859, "learning_rate": 9.377944006850807e-05, "loss": 0.0481, "step": 4050 }, { "grad_norm": 0.5161605477333069, "learning_rate": 9.373944421667265e-05, "loss": 0.0449, "step": 4060 }, { "grad_norm": 0.6617726683616638, "learning_rate": 9.369932878286691e-05, "loss": 0.0483, "step": 4070 }, { "grad_norm": 0.5100855231285095, "learning_rate": 9.365909387676494e-05, "loss": 0.0484, "step": 4080 }, { "grad_norm": 0.5112939476966858, "learning_rate": 9.361873960836744e-05, "loss": 0.046, "step": 4090 }, { "grad_norm": 0.4899284839630127, "learning_rate": 9.357826608800142e-05, "loss": 0.0444, "step": 4100 }, { "grad_norm": 0.5206869840621948, "learning_rate": 9.353767342631994e-05, "loss": 0.0447, "step": 4110 }, { "grad_norm": 0.5219537019729614, "learning_rate": 9.34969617343018e-05, "loss": 0.0448, "step": 4120 }, { "grad_norm": 0.5679187774658203, "learning_rate": 9.345613112325122e-05, "loss": 0.0449, "step": 4130 }, { "grad_norm": 0.4834611415863037, "learning_rate": 9.34151817047975e-05, "loss": 0.0424, "step": 4140 }, { "grad_norm": 0.504699170589447, "learning_rate": 9.33741135908948e-05, "loss": 0.0468, "step": 4150 }, { "grad_norm": 0.593715488910675, "learning_rate": 9.33329268938218e-05, "loss": 0.0445, "step": 4160 }, { "grad_norm": 0.5134977102279663, "learning_rate": 9.329162172618132e-05, "loss": 0.0454, "step": 4170 }, { "grad_norm": 0.5556760430335999, "learning_rate": 9.325019820090013e-05, "loss": 0.0444, "step": 4180 }, { "grad_norm": 0.4709920883178711, "learning_rate": 9.320865643122855e-05, "loss": 0.0439, "step": 4190 }, { "grad_norm": 0.5165664553642273, "learning_rate": 9.316699653074023e-05, "loss": 0.0445, "step": 4200 }, { "grad_norm": 0.3771870732307434, "learning_rate": 9.312521861333172e-05, "loss": 0.045, "step": 4210 }, { "grad_norm": 0.5048888325691223, "learning_rate": 9.308332279322224e-05, "loss": 0.0434, "step": 4220 }, { "grad_norm": 0.47928208112716675, "learning_rate": 9.304130918495338e-05, "loss": 0.0453, "step": 4230 }, { "grad_norm": 0.43420812487602234, "learning_rate": 9.299917790338874e-05, "loss": 0.0425, "step": 4240 }, { "grad_norm": 0.5437456369400024, "learning_rate": 9.295692906371363e-05, "loss": 0.0449, "step": 4250 }, { "grad_norm": 0.5281121134757996, "learning_rate": 9.291456278143476e-05, "loss": 0.0457, "step": 4260 }, { "grad_norm": 0.4478260278701782, "learning_rate": 9.287207917237994e-05, "loss": 0.0436, "step": 4270 }, { "grad_norm": 0.44330844283103943, "learning_rate": 9.282947835269773e-05, "loss": 0.0456, "step": 4280 }, { "grad_norm": 0.436909943819046, "learning_rate": 9.278676043885715e-05, "loss": 0.0462, "step": 4290 }, { "grad_norm": 0.4699770212173462, "learning_rate": 9.274392554764733e-05, "loss": 0.0467, "step": 4300 }, { "grad_norm": 0.48411276936531067, "learning_rate": 9.270097379617723e-05, "loss": 0.0462, "step": 4310 }, { "grad_norm": 0.476259708404541, "learning_rate": 9.26579053018753e-05, "loss": 0.0449, "step": 4320 }, { "grad_norm": 0.3954092860221863, "learning_rate": 9.261472018248918e-05, "loss": 0.0452, "step": 4330 }, { "grad_norm": 0.46465450525283813, "learning_rate": 9.25714185560853e-05, "loss": 0.0431, "step": 4340 }, { "grad_norm": 0.5479564070701599, "learning_rate": 9.252800054104868e-05, "loss": 0.0462, "step": 4350 }, { "grad_norm": 0.6036524772644043, "learning_rate": 9.248446625608252e-05, "loss": 0.0414, "step": 4360 }, { "grad_norm": 0.49141258001327515, "learning_rate": 9.244081582020789e-05, "loss": 0.0455, "step": 4370 }, { "grad_norm": 0.4593587815761566, "learning_rate": 9.239704935276339e-05, "loss": 0.0434, "step": 4380 }, { "grad_norm": 0.5062708258628845, "learning_rate": 9.235316697340489e-05, "loss": 0.0467, "step": 4390 }, { "grad_norm": 0.43314996361732483, "learning_rate": 9.230916880210512e-05, "loss": 0.0459, "step": 4400 }, { "grad_norm": 0.5277203321456909, "learning_rate": 9.226505495915342e-05, "loss": 0.0464, "step": 4410 }, { "grad_norm": 0.4898311197757721, "learning_rate": 9.222082556515536e-05, "loss": 0.0453, "step": 4420 }, { "grad_norm": 0.4869357943534851, "learning_rate": 9.217648074103242e-05, "loss": 0.0436, "step": 4430 }, { "grad_norm": 0.42760169506073, "learning_rate": 9.213202060802161e-05, "loss": 0.0413, "step": 4440 }, { "grad_norm": 0.5749702453613281, "learning_rate": 9.208744528767528e-05, "loss": 0.0426, "step": 4450 }, { "grad_norm": 0.484538733959198, "learning_rate": 9.204275490186064e-05, "loss": 0.0461, "step": 4460 }, { "grad_norm": 0.46094122529029846, "learning_rate": 9.199794957275949e-05, "loss": 0.0446, "step": 4470 }, { "grad_norm": 0.4943442642688751, "learning_rate": 9.19530294228679e-05, "loss": 0.0432, "step": 4480 }, { "grad_norm": 0.5320891737937927, "learning_rate": 9.190799457499583e-05, "loss": 0.0431, "step": 4490 }, { "grad_norm": 0.5658668875694275, "learning_rate": 9.186284515226686e-05, "loss": 0.0418, "step": 4500 }, { "grad_norm": 0.4938783645629883, "learning_rate": 9.181758127811777e-05, "loss": 0.0429, "step": 4510 }, { "grad_norm": 0.5220164060592651, "learning_rate": 9.177220307629825e-05, "loss": 0.0443, "step": 4520 }, { "grad_norm": 0.5482075810432434, "learning_rate": 9.172671067087059e-05, "loss": 0.043, "step": 4530 }, { "grad_norm": 0.3709627389907837, "learning_rate": 9.16811041862093e-05, "loss": 0.0441, "step": 4540 }, { "grad_norm": 0.5296226739883423, "learning_rate": 9.163538374700076e-05, "loss": 0.0441, "step": 4550 }, { "grad_norm": 0.4832737147808075, "learning_rate": 9.158954947824287e-05, "loss": 0.0438, "step": 4560 }, { "grad_norm": 0.5564871430397034, "learning_rate": 9.154360150524482e-05, "loss": 0.0442, "step": 4570 }, { "grad_norm": 0.3699457347393036, "learning_rate": 9.14975399536266e-05, "loss": 0.044, "step": 4580 }, { "grad_norm": 0.45649245381355286, "learning_rate": 9.14513649493187e-05, "loss": 0.0441, "step": 4590 }, { "grad_norm": 0.44787317514419556, "learning_rate": 9.140507661856187e-05, "loss": 0.0451, "step": 4600 }, { "grad_norm": 0.48270872235298157, "learning_rate": 9.135867508790661e-05, "loss": 0.0435, "step": 4610 }, { "grad_norm": 0.5807814002037048, "learning_rate": 9.131216048421291e-05, "loss": 0.0418, "step": 4620 }, { "grad_norm": 0.4520087242126465, "learning_rate": 9.126553293464998e-05, "loss": 0.044, "step": 4630 }, { "grad_norm": 0.44217219948768616, "learning_rate": 9.121879256669572e-05, "loss": 0.0444, "step": 4640 }, { "grad_norm": 0.4815622866153717, "learning_rate": 9.117193950813652e-05, "loss": 0.043, "step": 4650 }, { "grad_norm": 0.49311602115631104, "learning_rate": 9.112497388706685e-05, "loss": 0.0413, "step": 4660 }, { "grad_norm": 0.5161907076835632, "learning_rate": 9.10778958318889e-05, "loss": 0.0403, "step": 4670 }, { "grad_norm": 0.4166286289691925, "learning_rate": 9.103070547131232e-05, "loss": 0.0464, "step": 4680 }, { "grad_norm": 0.455807626247406, "learning_rate": 9.098340293435375e-05, "loss": 0.0437, "step": 4690 }, { "grad_norm": 0.575040876865387, "learning_rate": 9.093598835033649e-05, "loss": 0.0438, "step": 4700 }, { "grad_norm": 0.48000290989875793, "learning_rate": 9.088846184889021e-05, "loss": 0.0444, "step": 4710 }, { "grad_norm": 0.4801058769226074, "learning_rate": 9.084082355995057e-05, "loss": 0.0428, "step": 4720 }, { "grad_norm": 0.5418894290924072, "learning_rate": 9.079307361375882e-05, "loss": 0.0433, "step": 4730 }, { "grad_norm": 0.5534800291061401, "learning_rate": 9.074521214086149e-05, "loss": 0.0425, "step": 4740 }, { "grad_norm": 0.3854370415210724, "learning_rate": 9.069723927211001e-05, "loss": 0.044, "step": 4750 }, { "grad_norm": 0.5107142329216003, "learning_rate": 9.064915513866037e-05, "loss": 0.0443, "step": 4760 }, { "grad_norm": 0.48281845450401306, "learning_rate": 9.060095987197279e-05, "loss": 0.0448, "step": 4770 }, { "grad_norm": 0.524905800819397, "learning_rate": 9.055265360381126e-05, "loss": 0.0451, "step": 4780 }, { "grad_norm": 0.42521294951438904, "learning_rate": 9.050423646624326e-05, "loss": 0.0441, "step": 4790 }, { "grad_norm": 0.42200207710266113, "learning_rate": 9.045570859163943e-05, "loss": 0.0432, "step": 4800 }, { "grad_norm": 0.4576437473297119, "learning_rate": 9.04070701126731e-05, "loss": 0.0417, "step": 4810 }, { "grad_norm": 0.3940509557723999, "learning_rate": 9.035832116232001e-05, "loss": 0.0454, "step": 4820 }, { "grad_norm": 0.4551846385002136, "learning_rate": 9.030946187385796e-05, "loss": 0.0423, "step": 4830 }, { "grad_norm": 0.5149896740913391, "learning_rate": 9.026049238086635e-05, "loss": 0.0413, "step": 4840 }, { "grad_norm": 0.4057488739490509, "learning_rate": 9.021141281722591e-05, "loss": 0.0439, "step": 4850 }, { "grad_norm": 0.49988746643066406, "learning_rate": 9.01622233171183e-05, "loss": 0.0442, "step": 4860 }, { "grad_norm": 0.4672064781188965, "learning_rate": 9.011292401502574e-05, "loss": 0.0405, "step": 4870 }, { "grad_norm": 0.42719805240631104, "learning_rate": 9.006351504573063e-05, "loss": 0.0429, "step": 4880 }, { "grad_norm": 0.5068572163581848, "learning_rate": 9.001399654431519e-05, "loss": 0.0443, "step": 4890 }, { "grad_norm": 0.4347115159034729, "learning_rate": 8.996436864616116e-05, "loss": 0.0432, "step": 4900 }, { "grad_norm": 0.5151947736740112, "learning_rate": 8.991463148694925e-05, "loss": 0.0426, "step": 4910 }, { "grad_norm": 0.4219285547733307, "learning_rate": 8.986478520265902e-05, "loss": 0.0457, "step": 4920 }, { "grad_norm": 0.46755605936050415, "learning_rate": 8.981482992956827e-05, "loss": 0.0406, "step": 4930 }, { "grad_norm": 0.5359017848968506, "learning_rate": 8.976476580425282e-05, "loss": 0.0432, "step": 4940 }, { "grad_norm": 0.43240243196487427, "learning_rate": 8.971459296358606e-05, "loss": 0.043, "step": 4950 }, { "grad_norm": 0.42066359519958496, "learning_rate": 8.966431154473864e-05, "loss": 0.0406, "step": 4960 }, { "grad_norm": 0.44336849451065063, "learning_rate": 8.961392168517803e-05, "loss": 0.041, "step": 4970 }, { "grad_norm": 0.4278302490711212, "learning_rate": 8.956342352266821e-05, "loss": 0.0419, "step": 4980 }, { "grad_norm": 0.4722197949886322, "learning_rate": 8.95128171952692e-05, "loss": 0.0447, "step": 4990 }, { "grad_norm": 0.5200199484825134, "learning_rate": 8.946210284133676e-05, "loss": 0.0405, "step": 5000 }, { "grad_norm": 0.39662498235702515, "learning_rate": 8.941128059952201e-05, "loss": 0.04, "step": 5010 }, { "grad_norm": 0.4512505531311035, "learning_rate": 8.936035060877102e-05, "loss": 0.0385, "step": 5020 }, { "grad_norm": 0.43291807174682617, "learning_rate": 8.930931300832443e-05, "loss": 0.04, "step": 5030 }, { "grad_norm": 0.620459794998169, "learning_rate": 8.925816793771711e-05, "loss": 0.042, "step": 5040 }, { "grad_norm": 0.44059115648269653, "learning_rate": 8.92069155367777e-05, "loss": 0.0415, "step": 5050 }, { "grad_norm": 0.43802881240844727, "learning_rate": 8.915555594562834e-05, "loss": 0.0445, "step": 5060 }, { "grad_norm": 0.5628476142883301, "learning_rate": 8.910408930468416e-05, "loss": 0.0401, "step": 5070 }, { "grad_norm": 0.4604913294315338, "learning_rate": 8.905251575465303e-05, "loss": 0.0406, "step": 5080 }, { "grad_norm": 0.4277662932872772, "learning_rate": 8.900083543653502e-05, "loss": 0.041, "step": 5090 }, { "grad_norm": 0.4302856922149658, "learning_rate": 8.894904849162218e-05, "loss": 0.0427, "step": 5100 }, { "grad_norm": 0.4249647855758667, "learning_rate": 8.889715506149802e-05, "loss": 0.0398, "step": 5110 }, { "grad_norm": 0.4415731728076935, "learning_rate": 8.884515528803722e-05, "loss": 0.0413, "step": 5120 }, { "grad_norm": 0.4602738618850708, "learning_rate": 8.879304931340517e-05, "loss": 0.0398, "step": 5130 }, { "grad_norm": 0.4362502992153168, "learning_rate": 8.874083728005759e-05, "loss": 0.0422, "step": 5140 }, { "grad_norm": 0.442611426115036, "learning_rate": 8.868851933074021e-05, "loss": 0.042, "step": 5150 }, { "grad_norm": 0.5155687928199768, "learning_rate": 8.863609560848829e-05, "loss": 0.0405, "step": 5160 }, { "grad_norm": 0.5138674378395081, "learning_rate": 8.85835662566263e-05, "loss": 0.0389, "step": 5170 }, { "grad_norm": 0.4577963948249817, "learning_rate": 8.853093141876747e-05, "loss": 0.0404, "step": 5180 }, { "grad_norm": 0.4621758759021759, "learning_rate": 8.847819123881343e-05, "loss": 0.0386, "step": 5190 }, { "grad_norm": 0.43907004594802856, "learning_rate": 8.842534586095383e-05, "loss": 0.0379, "step": 5200 }, { "grad_norm": 0.46490663290023804, "learning_rate": 8.837239542966593e-05, "loss": 0.0398, "step": 5210 }, { "grad_norm": 0.48866644501686096, "learning_rate": 8.831934008971417e-05, "loss": 0.04, "step": 5220 }, { "grad_norm": 0.46526607871055603, "learning_rate": 8.826617998614982e-05, "loss": 0.0436, "step": 5230 }, { "grad_norm": 0.4442521035671234, "learning_rate": 8.821291526431056e-05, "loss": 0.0406, "step": 5240 }, { "grad_norm": 0.5185173153877258, "learning_rate": 8.815954606982015e-05, "loss": 0.0413, "step": 5250 }, { "grad_norm": 0.49182066321372986, "learning_rate": 8.810607254858789e-05, "loss": 0.0423, "step": 5260 }, { "grad_norm": 0.4638088345527649, "learning_rate": 8.805249484680838e-05, "loss": 0.038, "step": 5270 }, { "grad_norm": 0.46812355518341064, "learning_rate": 8.799881311096096e-05, "loss": 0.0385, "step": 5280 }, { "grad_norm": 0.47700005769729614, "learning_rate": 8.794502748780949e-05, "loss": 0.0396, "step": 5290 }, { "grad_norm": 0.4278887212276459, "learning_rate": 8.78911381244018e-05, "loss": 0.0407, "step": 5300 }, { "grad_norm": 0.4923304319381714, "learning_rate": 8.783714516806933e-05, "loss": 0.0417, "step": 5310 }, { "grad_norm": 0.470355749130249, "learning_rate": 8.77830487664268e-05, "loss": 0.0383, "step": 5320 }, { "grad_norm": 0.37395450472831726, "learning_rate": 8.772884906737167e-05, "loss": 0.0404, "step": 5330 }, { "grad_norm": 0.4390724003314972, "learning_rate": 8.767454621908387e-05, "loss": 0.0408, "step": 5340 }, { "grad_norm": 0.4128201901912689, "learning_rate": 8.76201403700253e-05, "loss": 0.0377, "step": 5350 }, { "grad_norm": 0.44246816635131836, "learning_rate": 8.756563166893949e-05, "loss": 0.041, "step": 5360 }, { "grad_norm": 0.533183217048645, "learning_rate": 8.751102026485113e-05, "loss": 0.0417, "step": 5370 }, { "grad_norm": 0.470833420753479, "learning_rate": 8.745630630706571e-05, "loss": 0.0426, "step": 5380 }, { "grad_norm": 0.4990265369415283, "learning_rate": 8.740148994516912e-05, "loss": 0.04, "step": 5390 }, { "grad_norm": 0.4298129677772522, "learning_rate": 8.73465713290272e-05, "loss": 0.0383, "step": 5400 }, { "grad_norm": 0.5823467969894409, "learning_rate": 8.729155060878533e-05, "loss": 0.0398, "step": 5410 }, { "grad_norm": 0.445436954498291, "learning_rate": 8.723642793486809e-05, "loss": 0.039, "step": 5420 }, { "grad_norm": 0.3757471740245819, "learning_rate": 8.718120345797873e-05, "loss": 0.0397, "step": 5430 }, { "grad_norm": 0.4835352301597595, "learning_rate": 8.712587732909889e-05, "loss": 0.0401, "step": 5440 }, { "grad_norm": 0.4055708050727844, "learning_rate": 8.707044969948806e-05, "loss": 0.0378, "step": 5450 }, { "grad_norm": 0.4023078382015228, "learning_rate": 8.701492072068329e-05, "loss": 0.0389, "step": 5460 }, { "grad_norm": 0.48503735661506653, "learning_rate": 8.695929054449869e-05, "loss": 0.0401, "step": 5470 }, { "grad_norm": 0.5127048492431641, "learning_rate": 8.690355932302501e-05, "loss": 0.0398, "step": 5480 }, { "grad_norm": 0.466054230928421, "learning_rate": 8.684772720862931e-05, "loss": 0.0422, "step": 5490 }, { "grad_norm": 0.3630341589450836, "learning_rate": 8.679179435395446e-05, "loss": 0.0422, "step": 5500 }, { "grad_norm": 0.4958881139755249, "learning_rate": 8.673576091191874e-05, "loss": 0.0389, "step": 5510 }, { "grad_norm": 0.47542330622673035, "learning_rate": 8.667962703571541e-05, "loss": 0.0381, "step": 5520 }, { "grad_norm": 0.48895490169525146, "learning_rate": 8.662339287881238e-05, "loss": 0.0395, "step": 5530 }, { "grad_norm": 0.436758816242218, "learning_rate": 8.656705859495169e-05, "loss": 0.0401, "step": 5540 }, { "grad_norm": 0.46079763770103455, "learning_rate": 8.651062433814912e-05, "loss": 0.0398, "step": 5550 }, { "grad_norm": 0.4107966721057892, "learning_rate": 8.645409026269375e-05, "loss": 0.0402, "step": 5560 }, { "grad_norm": 0.41389384865760803, "learning_rate": 8.639745652314759e-05, "loss": 0.0411, "step": 5570 }, { "grad_norm": 0.451793372631073, "learning_rate": 8.634072327434515e-05, "loss": 0.0398, "step": 5580 }, { "grad_norm": 0.4580284059047699, "learning_rate": 8.628389067139294e-05, "loss": 0.0407, "step": 5590 }, { "grad_norm": 0.4872185289859772, "learning_rate": 8.622695886966911e-05, "loss": 0.0411, "step": 5600 }, { "grad_norm": 0.3741706311702728, "learning_rate": 8.616992802482308e-05, "loss": 0.0421, "step": 5610 }, { "grad_norm": 0.4313085377216339, "learning_rate": 8.611279829277496e-05, "loss": 0.0391, "step": 5620 }, { "grad_norm": 0.42592012882232666, "learning_rate": 8.605556982971528e-05, "loss": 0.0384, "step": 5630 }, { "grad_norm": 0.4015299081802368, "learning_rate": 8.599824279210447e-05, "loss": 0.0379, "step": 5640 }, { "grad_norm": 0.39964044094085693, "learning_rate": 8.594081733667243e-05, "loss": 0.0381, "step": 5650 }, { "grad_norm": 0.43630242347717285, "learning_rate": 8.58832936204182e-05, "loss": 0.0401, "step": 5660 }, { "grad_norm": 0.4856785535812378, "learning_rate": 8.582567180060942e-05, "loss": 0.0399, "step": 5670 }, { "grad_norm": 0.3906029164791107, "learning_rate": 8.576795203478194e-05, "loss": 0.041, "step": 5680 }, { "grad_norm": 0.3879716098308563, "learning_rate": 8.571013448073939e-05, "loss": 0.0383, "step": 5690 }, { "grad_norm": 0.36499521136283875, "learning_rate": 8.565221929655275e-05, "loss": 0.037, "step": 5700 }, { "grad_norm": 0.3831605911254883, "learning_rate": 8.559420664055992e-05, "loss": 0.0347, "step": 5710 }, { "grad_norm": 0.5628010630607605, "learning_rate": 8.553609667136532e-05, "loss": 0.0382, "step": 5720 }, { "grad_norm": 0.4884410500526428, "learning_rate": 8.547788954783936e-05, "loss": 0.038, "step": 5730 }, { "grad_norm": 0.420221209526062, "learning_rate": 8.541958542911808e-05, "loss": 0.0385, "step": 5740 }, { "grad_norm": 0.3182365894317627, "learning_rate": 8.536118447460275e-05, "loss": 0.0391, "step": 5750 }, { "grad_norm": 0.39842677116394043, "learning_rate": 8.530268684395932e-05, "loss": 0.0385, "step": 5760 }, { "grad_norm": 0.5056148767471313, "learning_rate": 8.524409269711807e-05, "loss": 0.0385, "step": 5770 }, { "grad_norm": 0.5514624118804932, "learning_rate": 8.51854021942732e-05, "loss": 0.0407, "step": 5780 }, { "grad_norm": 0.4214351177215576, "learning_rate": 8.512661549588227e-05, "loss": 0.0399, "step": 5790 }, { "grad_norm": 0.41320428252220154, "learning_rate": 8.506773276266588e-05, "loss": 0.0395, "step": 5800 }, { "grad_norm": 0.34123003482818604, "learning_rate": 8.500875415560721e-05, "loss": 0.038, "step": 5810 }, { "grad_norm": 0.42489731311798096, "learning_rate": 8.494967983595144e-05, "loss": 0.0389, "step": 5820 }, { "grad_norm": 0.37367987632751465, "learning_rate": 8.489050996520558e-05, "loss": 0.0371, "step": 5830 }, { "grad_norm": 0.3735558092594147, "learning_rate": 8.483124470513775e-05, "loss": 0.0376, "step": 5840 }, { "grad_norm": 0.4131052792072296, "learning_rate": 8.477188421777692e-05, "loss": 0.0376, "step": 5850 }, { "grad_norm": 0.41150277853012085, "learning_rate": 8.47124286654124e-05, "loss": 0.0387, "step": 5860 }, { "grad_norm": 0.4659402668476105, "learning_rate": 8.465287821059341e-05, "loss": 0.0387, "step": 5870 }, { "grad_norm": 0.3859495222568512, "learning_rate": 8.45932330161286e-05, "loss": 0.0385, "step": 5880 }, { "grad_norm": 0.4939848780632019, "learning_rate": 8.453349324508567e-05, "loss": 0.0403, "step": 5890 }, { "grad_norm": 0.424917995929718, "learning_rate": 8.447365906079088e-05, "loss": 0.0379, "step": 5900 }, { "grad_norm": 0.3580872714519501, "learning_rate": 8.441373062682856e-05, "loss": 0.0368, "step": 5910 }, { "grad_norm": 0.44607481360435486, "learning_rate": 8.43537081070408e-05, "loss": 0.0389, "step": 5920 }, { "grad_norm": 0.4138437807559967, "learning_rate": 8.429359166552689e-05, "loss": 0.037, "step": 5930 }, { "grad_norm": 0.40142327547073364, "learning_rate": 8.423338146664284e-05, "loss": 0.0373, "step": 5940 }, { "grad_norm": 0.4808826148509979, "learning_rate": 8.417307767500107e-05, "loss": 0.0379, "step": 5950 }, { "grad_norm": 0.5105547904968262, "learning_rate": 8.411268045546983e-05, "loss": 0.04, "step": 5960 }, { "grad_norm": 0.34908849000930786, "learning_rate": 8.405218997317281e-05, "loss": 0.0381, "step": 5970 }, { "grad_norm": 0.4811083674430847, "learning_rate": 8.399160639348869e-05, "loss": 0.0369, "step": 5980 }, { "grad_norm": 0.43989846110343933, "learning_rate": 8.393092988205065e-05, "loss": 0.0381, "step": 5990 }, { "grad_norm": 0.43242347240448, "learning_rate": 8.387016060474597e-05, "loss": 0.0374, "step": 6000 }, { "grad_norm": 0.48375701904296875, "learning_rate": 8.380929872771551e-05, "loss": 0.0363, "step": 6010 }, { "grad_norm": 0.431797593832016, "learning_rate": 8.374834441735335e-05, "loss": 0.038, "step": 6020 }, { "grad_norm": 0.34289318323135376, "learning_rate": 8.368729784030622e-05, "loss": 0.0359, "step": 6030 }, { "grad_norm": 0.44675248861312866, "learning_rate": 8.362615916347315e-05, "loss": 0.0353, "step": 6040 }, { "grad_norm": 0.4955903887748718, "learning_rate": 8.356492855400493e-05, "loss": 0.0372, "step": 6050 }, { "grad_norm": 0.403791606426239, "learning_rate": 8.350360617930371e-05, "loss": 0.0365, "step": 6060 }, { "grad_norm": 0.31619811058044434, "learning_rate": 8.344219220702255e-05, "loss": 0.036, "step": 6070 }, { "grad_norm": 0.5609193444252014, "learning_rate": 8.338068680506485e-05, "loss": 0.0395, "step": 6080 }, { "grad_norm": 0.4349352717399597, "learning_rate": 8.33190901415841e-05, "loss": 0.0388, "step": 6090 }, { "grad_norm": 0.3991135060787201, "learning_rate": 8.325740238498317e-05, "loss": 0.0364, "step": 6100 }, { "grad_norm": 0.41887056827545166, "learning_rate": 8.319562370391406e-05, "loss": 0.0386, "step": 6110 }, { "grad_norm": 0.4059099555015564, "learning_rate": 8.31337542672773e-05, "loss": 0.0369, "step": 6120 }, { "grad_norm": 0.4385959804058075, "learning_rate": 8.307179424422158e-05, "loss": 0.0364, "step": 6130 }, { "grad_norm": 0.38501816987991333, "learning_rate": 8.300974380414327e-05, "loss": 0.0359, "step": 6140 }, { "grad_norm": 0.427293986082077, "learning_rate": 8.294760311668586e-05, "loss": 0.0362, "step": 6150 }, { "grad_norm": 0.3641018867492676, "learning_rate": 8.288537235173961e-05, "loss": 0.0347, "step": 6160 }, { "grad_norm": 0.3719545602798462, "learning_rate": 8.282305167944108e-05, "loss": 0.0341, "step": 6170 }, { "grad_norm": 0.5745158195495605, "learning_rate": 8.276064127017262e-05, "loss": 0.0376, "step": 6180 }, { "grad_norm": 0.38939550518989563, "learning_rate": 8.269814129456189e-05, "loss": 0.0353, "step": 6190 }, { "grad_norm": 0.3649023771286011, "learning_rate": 8.263555192348143e-05, "loss": 0.0375, "step": 6200 }, { "grad_norm": 0.4111265540122986, "learning_rate": 8.257287332804819e-05, "loss": 0.0382, "step": 6210 }, { "grad_norm": 0.42668649554252625, "learning_rate": 8.251010567962307e-05, "loss": 0.0396, "step": 6220 }, { "grad_norm": 0.4298035502433777, "learning_rate": 8.244724914981041e-05, "loss": 0.0387, "step": 6230 }, { "grad_norm": 0.3768647015094757, "learning_rate": 8.238430391045757e-05, "loss": 0.0382, "step": 6240 }, { "grad_norm": 0.38798126578330994, "learning_rate": 8.232127013365445e-05, "loss": 0.0384, "step": 6250 }, { "grad_norm": 0.38849231600761414, "learning_rate": 8.225814799173295e-05, "loss": 0.0378, "step": 6260 }, { "grad_norm": 0.38603171706199646, "learning_rate": 8.219493765726663e-05, "loss": 0.0376, "step": 6270 }, { "grad_norm": 0.3577180504798889, "learning_rate": 8.21316393030701e-05, "loss": 0.0368, "step": 6280 }, { "grad_norm": 0.50895094871521, "learning_rate": 8.206825310219865e-05, "loss": 0.0375, "step": 6290 }, { "grad_norm": 0.4111712872982025, "learning_rate": 8.200477922794776e-05, "loss": 0.037, "step": 6300 }, { "grad_norm": 0.41497135162353516, "learning_rate": 8.194121785385256e-05, "loss": 0.0365, "step": 6310 }, { "grad_norm": 0.42979851365089417, "learning_rate": 8.187756915368741e-05, "loss": 0.036, "step": 6320 }, { "grad_norm": 0.359604150056839, "learning_rate": 8.181383330146544e-05, "loss": 0.0348, "step": 6330 }, { "grad_norm": 0.4255097508430481, "learning_rate": 8.175001047143804e-05, "loss": 0.036, "step": 6340 }, { "grad_norm": 0.3181275725364685, "learning_rate": 8.168610083809438e-05, "loss": 0.0353, "step": 6350 }, { "grad_norm": 0.34049099683761597, "learning_rate": 8.162210457616095e-05, "loss": 0.0366, "step": 6360 }, { "grad_norm": 0.45616477727890015, "learning_rate": 8.155802186060109e-05, "loss": 0.0371, "step": 6370 }, { "grad_norm": 0.4055083990097046, "learning_rate": 8.149385286661453e-05, "loss": 0.037, "step": 6380 }, { "grad_norm": 0.3632270097732544, "learning_rate": 8.14295977696368e-05, "loss": 0.0365, "step": 6390 }, { "grad_norm": 0.4278525412082672, "learning_rate": 8.13652567453389e-05, "loss": 0.0384, "step": 6400 }, { "grad_norm": 0.4336703419685364, "learning_rate": 8.130082996962676e-05, "loss": 0.0359, "step": 6410 }, { "grad_norm": 0.4452275037765503, "learning_rate": 8.123631761864068e-05, "loss": 0.0363, "step": 6420 }, { "grad_norm": 0.3490341603755951, "learning_rate": 8.1171719868755e-05, "loss": 0.037, "step": 6430 }, { "grad_norm": 0.4245625436306, "learning_rate": 8.110703689657748e-05, "loss": 0.0358, "step": 6440 }, { "grad_norm": 0.3515666425228119, "learning_rate": 8.104226887894892e-05, "loss": 0.0374, "step": 6450 }, { "grad_norm": 0.45157885551452637, "learning_rate": 8.097741599294257e-05, "loss": 0.0364, "step": 6460 }, { "grad_norm": 0.36290258169174194, "learning_rate": 8.091247841586378e-05, "loss": 0.036, "step": 6470 }, { "grad_norm": 0.431040495634079, "learning_rate": 8.084745632524939e-05, "loss": 0.0363, "step": 6480 }, { "grad_norm": 0.41045570373535156, "learning_rate": 8.07823498988673e-05, "loss": 0.0362, "step": 6490 }, { "grad_norm": 0.46383172273635864, "learning_rate": 8.071715931471602e-05, "loss": 0.0379, "step": 6500 }, { "grad_norm": 0.38879528641700745, "learning_rate": 8.06518847510241e-05, "loss": 0.0367, "step": 6510 }, { "grad_norm": 0.36130836606025696, "learning_rate": 8.058652638624971e-05, "loss": 0.0368, "step": 6520 }, { "grad_norm": 0.49286437034606934, "learning_rate": 8.052108439908013e-05, "loss": 0.0362, "step": 6530 }, { "grad_norm": 0.3841950595378876, "learning_rate": 8.045555896843125e-05, "loss": 0.0336, "step": 6540 }, { "grad_norm": 0.40306082367897034, "learning_rate": 8.03899502734471e-05, "loss": 0.0353, "step": 6550 }, { "grad_norm": 0.38004013895988464, "learning_rate": 8.032425849349931e-05, "loss": 0.0353, "step": 6560 }, { "grad_norm": 0.365662544965744, "learning_rate": 8.025848380818674e-05, "loss": 0.0373, "step": 6570 }, { "grad_norm": 0.4262871742248535, "learning_rate": 8.019262639733487e-05, "loss": 0.0355, "step": 6580 }, { "grad_norm": 0.4187777042388916, "learning_rate": 8.012668644099531e-05, "loss": 0.0372, "step": 6590 }, { "grad_norm": 0.3868800699710846, "learning_rate": 8.006066411944542e-05, "loss": 0.0379, "step": 6600 }, { "grad_norm": 0.45861124992370605, "learning_rate": 7.999455961318769e-05, "loss": 0.0342, "step": 6610 }, { "grad_norm": 0.4273201525211334, "learning_rate": 7.992837310294932e-05, "loss": 0.0354, "step": 6620 }, { "grad_norm": 0.47681984305381775, "learning_rate": 7.986210476968167e-05, "loss": 0.0374, "step": 6630 }, { "grad_norm": 0.41647207736968994, "learning_rate": 7.97957547945599e-05, "loss": 0.0364, "step": 6640 }, { "grad_norm": 0.49211931228637695, "learning_rate": 7.972932335898226e-05, "loss": 0.039, "step": 6650 }, { "grad_norm": 0.3217320740222931, "learning_rate": 7.966281064456975e-05, "loss": 0.035, "step": 6660 }, { "grad_norm": 0.4061133861541748, "learning_rate": 7.959621683316563e-05, "loss": 0.0358, "step": 6670 }, { "grad_norm": 0.399582177400589, "learning_rate": 7.952954210683481e-05, "loss": 0.0357, "step": 6680 }, { "grad_norm": 0.3294697403907776, "learning_rate": 7.946278664786345e-05, "loss": 0.0375, "step": 6690 }, { "grad_norm": 0.350180983543396, "learning_rate": 7.939595063875842e-05, "loss": 0.0372, "step": 6700 }, { "grad_norm": 0.37327510118484497, "learning_rate": 7.932903426224683e-05, "loss": 0.0344, "step": 6710 }, { "grad_norm": 0.373664915561676, "learning_rate": 7.926203770127552e-05, "loss": 0.0346, "step": 6720 }, { "grad_norm": 0.37155434489250183, "learning_rate": 7.919496113901046e-05, "loss": 0.0335, "step": 6730 }, { "grad_norm": 0.40722668170928955, "learning_rate": 7.912780475883649e-05, "loss": 0.0338, "step": 6740 }, { "grad_norm": 0.3765784502029419, "learning_rate": 7.906056874435652e-05, "loss": 0.0347, "step": 6750 }, { "grad_norm": 0.35448965430259705, "learning_rate": 7.899325327939131e-05, "loss": 0.0361, "step": 6760 }, { "grad_norm": 0.39404189586639404, "learning_rate": 7.892585854797872e-05, "loss": 0.0378, "step": 6770 }, { "grad_norm": 0.38724619150161743, "learning_rate": 7.88583847343734e-05, "loss": 0.0336, "step": 6780 }, { "grad_norm": 0.3638536334037781, "learning_rate": 7.879083202304616e-05, "loss": 0.033, "step": 6790 }, { "grad_norm": 0.4080106019973755, "learning_rate": 7.872320059868355e-05, "loss": 0.0359, "step": 6800 }, { "grad_norm": 0.37898164987564087, "learning_rate": 7.865549064618729e-05, "loss": 0.0344, "step": 6810 }, { "grad_norm": 0.3711118698120117, "learning_rate": 7.858770235067381e-05, "loss": 0.0351, "step": 6820 }, { "grad_norm": 0.405647337436676, "learning_rate": 7.851983589747374e-05, "loss": 0.0352, "step": 6830 }, { "grad_norm": 0.4660595953464508, "learning_rate": 7.845189147213133e-05, "loss": 0.0344, "step": 6840 }, { "grad_norm": 0.40214425325393677, "learning_rate": 7.838386926040407e-05, "loss": 0.0328, "step": 6850 }, { "grad_norm": 0.46578049659729004, "learning_rate": 7.83157694482621e-05, "loss": 0.0355, "step": 6860 }, { "grad_norm": 0.37721648812294006, "learning_rate": 7.824759222188768e-05, "loss": 0.0361, "step": 6870 }, { "grad_norm": 0.4541361331939697, "learning_rate": 7.817933776767478e-05, "loss": 0.0363, "step": 6880 }, { "grad_norm": 0.37245070934295654, "learning_rate": 7.811100627222842e-05, "loss": 0.0354, "step": 6890 }, { "grad_norm": 0.3632197082042694, "learning_rate": 7.804259792236435e-05, "loss": 0.0334, "step": 6900 }, { "grad_norm": 0.4251933991909027, "learning_rate": 7.797411290510835e-05, "loss": 0.0351, "step": 6910 }, { "grad_norm": 0.35473453998565674, "learning_rate": 7.790555140769586e-05, "loss": 0.0352, "step": 6920 }, { "grad_norm": 0.40998026728630066, "learning_rate": 7.78369136175714e-05, "loss": 0.0372, "step": 6930 }, { "grad_norm": 0.35263049602508545, "learning_rate": 7.776819972238806e-05, "loss": 0.0354, "step": 6940 }, { "grad_norm": 0.4343003034591675, "learning_rate": 7.7699409910007e-05, "loss": 0.036, "step": 6950 }, { "grad_norm": 0.4120432138442993, "learning_rate": 7.763054436849694e-05, "loss": 0.0347, "step": 6960 }, { "grad_norm": 0.37748095393180847, "learning_rate": 7.756160328613364e-05, "loss": 0.0341, "step": 6970 }, { "grad_norm": 0.48622989654541016, "learning_rate": 7.749258685139942e-05, "loss": 0.0378, "step": 6980 }, { "grad_norm": 0.4021105468273163, "learning_rate": 7.742349525298253e-05, "loss": 0.0343, "step": 6990 }, { "grad_norm": 0.3868004381656647, "learning_rate": 7.735432867977679e-05, "loss": 0.0362, "step": 7000 }, { "grad_norm": 0.3710803687572479, "learning_rate": 7.728508732088096e-05, "loss": 0.0352, "step": 7010 }, { "grad_norm": 0.411406546831131, "learning_rate": 7.721577136559825e-05, "loss": 0.0357, "step": 7020 }, { "grad_norm": 0.4609767198562622, "learning_rate": 7.714638100343588e-05, "loss": 0.0374, "step": 7030 }, { "grad_norm": 0.38829782605171204, "learning_rate": 7.707691642410444e-05, "loss": 0.0362, "step": 7040 }, { "grad_norm": 0.419757217168808, "learning_rate": 7.70073778175174e-05, "loss": 0.0334, "step": 7050 }, { "grad_norm": 0.41437044739723206, "learning_rate": 7.69377653737907e-05, "loss": 0.0325, "step": 7060 }, { "grad_norm": 0.4515761733055115, "learning_rate": 7.686807928324209e-05, "loss": 0.0371, "step": 7070 }, { "grad_norm": 0.4891188442707062, "learning_rate": 7.679831973639065e-05, "loss": 0.0367, "step": 7080 }, { "grad_norm": 0.29925650358200073, "learning_rate": 7.672848692395637e-05, "loss": 0.0342, "step": 7090 }, { "grad_norm": 0.32905447483062744, "learning_rate": 7.665858103685944e-05, "loss": 0.0344, "step": 7100 }, { "grad_norm": 0.3577474355697632, "learning_rate": 7.658860226621991e-05, "loss": 0.0318, "step": 7110 }, { "grad_norm": 0.3971805274486542, "learning_rate": 7.651855080335708e-05, "loss": 0.0331, "step": 7120 }, { "grad_norm": 0.37195658683776855, "learning_rate": 7.644842683978896e-05, "loss": 0.0339, "step": 7130 }, { "grad_norm": 0.40617552399635315, "learning_rate": 7.63782305672318e-05, "loss": 0.0322, "step": 7140 }, { "grad_norm": 0.4543386399745941, "learning_rate": 7.63079621775995e-05, "loss": 0.0348, "step": 7150 }, { "grad_norm": 0.5070335268974304, "learning_rate": 7.623762186300319e-05, "loss": 0.0349, "step": 7160 }, { "grad_norm": 0.4141789674758911, "learning_rate": 7.616720981575057e-05, "loss": 0.0357, "step": 7170 }, { "grad_norm": 0.3738696575164795, "learning_rate": 7.609672622834552e-05, "loss": 0.0355, "step": 7180 }, { "grad_norm": 0.3759477734565735, "learning_rate": 7.602617129348747e-05, "loss": 0.0351, "step": 7190 }, { "grad_norm": 0.3712978661060333, "learning_rate": 7.595554520407088e-05, "loss": 0.0344, "step": 7200 }, { "grad_norm": 0.37921613454818726, "learning_rate": 7.588484815318484e-05, "loss": 0.0345, "step": 7210 }, { "grad_norm": 0.47082775831222534, "learning_rate": 7.581408033411234e-05, "loss": 0.0378, "step": 7220 }, { "grad_norm": 0.4012286365032196, "learning_rate": 7.574324194032995e-05, "loss": 0.0342, "step": 7230 }, { "grad_norm": 0.36344462633132935, "learning_rate": 7.567233316550705e-05, "loss": 0.0351, "step": 7240 }, { "grad_norm": 0.35988062620162964, "learning_rate": 7.560135420350562e-05, "loss": 0.0326, "step": 7250 }, { "grad_norm": 0.4449901580810547, "learning_rate": 7.553030524837935e-05, "loss": 0.0346, "step": 7260 }, { "grad_norm": 0.4368852376937866, "learning_rate": 7.545918649437341e-05, "loss": 0.036, "step": 7270 }, { "grad_norm": 0.4328041076660156, "learning_rate": 7.538799813592377e-05, "loss": 0.0349, "step": 7280 }, { "grad_norm": 0.33543661236763, "learning_rate": 7.531674036765662e-05, "loss": 0.0344, "step": 7290 }, { "grad_norm": 0.3932455778121948, "learning_rate": 7.524541338438807e-05, "loss": 0.0327, "step": 7300 }, { "grad_norm": 0.339933305978775, "learning_rate": 7.517401738112328e-05, "loss": 0.0341, "step": 7310 }, { "grad_norm": 0.36566051840782166, "learning_rate": 7.510255255305628e-05, "loss": 0.033, "step": 7320 }, { "grad_norm": 0.5061442255973816, "learning_rate": 7.503101909556911e-05, "loss": 0.0339, "step": 7330 }, { "grad_norm": 0.4760403335094452, "learning_rate": 7.495941720423154e-05, "loss": 0.0346, "step": 7340 }, { "grad_norm": 0.4327556788921356, "learning_rate": 7.488774707480042e-05, "loss": 0.0335, "step": 7350 }, { "grad_norm": 0.3216523230075836, "learning_rate": 7.481600890321911e-05, "loss": 0.0355, "step": 7360 }, { "grad_norm": 0.36823445558547974, "learning_rate": 7.474420288561708e-05, "loss": 0.0337, "step": 7370 }, { "grad_norm": 0.4510650336742401, "learning_rate": 7.467232921830921e-05, "loss": 0.0345, "step": 7380 }, { "grad_norm": 0.3118990361690521, "learning_rate": 7.460038809779537e-05, "loss": 0.0337, "step": 7390 }, { "grad_norm": 0.3244684040546417, "learning_rate": 7.452837972075983e-05, "loss": 0.0341, "step": 7400 }, { "grad_norm": 0.3592819571495056, "learning_rate": 7.445630428407074e-05, "loss": 0.0338, "step": 7410 }, { "grad_norm": 0.3477317988872528, "learning_rate": 7.43841619847796e-05, "loss": 0.0323, "step": 7420 }, { "grad_norm": 0.40442895889282227, "learning_rate": 7.431195302012072e-05, "loss": 0.0336, "step": 7430 }, { "grad_norm": 0.3610159456729889, "learning_rate": 7.423967758751061e-05, "loss": 0.0325, "step": 7440 }, { "grad_norm": 0.34850987792015076, "learning_rate": 7.416733588454758e-05, "loss": 0.0354, "step": 7450 }, { "grad_norm": 0.41649460792541504, "learning_rate": 7.409492810901106e-05, "loss": 0.0327, "step": 7460 }, { "grad_norm": 0.366507351398468, "learning_rate": 7.402245445886116e-05, "loss": 0.0349, "step": 7470 }, { "grad_norm": 0.3944462537765503, "learning_rate": 7.394991513223806e-05, "loss": 0.0349, "step": 7480 }, { "grad_norm": 0.29914093017578125, "learning_rate": 7.38773103274615e-05, "loss": 0.0332, "step": 7490 }, { "grad_norm": 0.3260498642921448, "learning_rate": 7.380464024303028e-05, "loss": 0.031, "step": 7500 }, { "grad_norm": 0.37803712487220764, "learning_rate": 7.373190507762162e-05, "loss": 0.0335, "step": 7510 }, { "grad_norm": 0.3602581024169922, "learning_rate": 7.365910503009066e-05, "loss": 0.0336, "step": 7520 }, { "grad_norm": 0.3542674481868744, "learning_rate": 7.358624029946996e-05, "loss": 0.0342, "step": 7530 }, { "grad_norm": 0.4673260748386383, "learning_rate": 7.351331108496893e-05, "loss": 0.0346, "step": 7540 }, { "grad_norm": 0.36193886399269104, "learning_rate": 7.344031758597325e-05, "loss": 0.0318, "step": 7550 }, { "grad_norm": 0.3531333804130554, "learning_rate": 7.336726000204435e-05, "loss": 0.0326, "step": 7560 }, { "grad_norm": 0.365803986787796, "learning_rate": 7.32941385329189e-05, "loss": 0.0332, "step": 7570 }, { "grad_norm": 0.3608737587928772, "learning_rate": 7.322095337850816e-05, "loss": 0.0332, "step": 7580 }, { "grad_norm": 0.40053385496139526, "learning_rate": 7.314770473889758e-05, "loss": 0.0339, "step": 7590 }, { "grad_norm": 0.35238954424858093, "learning_rate": 7.307439281434615e-05, "loss": 0.0332, "step": 7600 }, { "grad_norm": 0.3912058174610138, "learning_rate": 7.300101780528585e-05, "loss": 0.0317, "step": 7610 }, { "grad_norm": 0.3912275731563568, "learning_rate": 7.292757991232117e-05, "loss": 0.0309, "step": 7620 }, { "grad_norm": 0.4632697105407715, "learning_rate": 7.285407933622848e-05, "loss": 0.0311, "step": 7630 }, { "grad_norm": 0.3496100604534149, "learning_rate": 7.278051627795557e-05, "loss": 0.0321, "step": 7640 }, { "grad_norm": 0.3689875304698944, "learning_rate": 7.270689093862105e-05, "loss": 0.0319, "step": 7650 }, { "grad_norm": 0.38931483030319214, "learning_rate": 7.263320351951374e-05, "loss": 0.0316, "step": 7660 }, { "grad_norm": 0.3756372928619385, "learning_rate": 7.255945422209227e-05, "loss": 0.0322, "step": 7670 }, { "grad_norm": 0.32288503646850586, "learning_rate": 7.248564324798437e-05, "loss": 0.0334, "step": 7680 }, { "grad_norm": 0.36744463443756104, "learning_rate": 7.241177079898644e-05, "loss": 0.0335, "step": 7690 }, { "grad_norm": 0.309794545173645, "learning_rate": 7.233783707706295e-05, "loss": 0.0304, "step": 7700 }, { "grad_norm": 0.3395470082759857, "learning_rate": 7.226384228434586e-05, "loss": 0.0307, "step": 7710 }, { "grad_norm": 0.365632027387619, "learning_rate": 7.21897866231341e-05, "loss": 0.0298, "step": 7720 }, { "grad_norm": 0.3479767441749573, "learning_rate": 7.211567029589303e-05, "loss": 0.0325, "step": 7730 }, { "grad_norm": 0.33217042684555054, "learning_rate": 7.204149350525387e-05, "loss": 0.034, "step": 7740 }, { "grad_norm": 0.3533134162425995, "learning_rate": 7.196725645401309e-05, "loss": 0.0316, "step": 7750 }, { "grad_norm": 0.470449835062027, "learning_rate": 7.1892959345132e-05, "loss": 0.0341, "step": 7760 }, { "grad_norm": 0.45280715823173523, "learning_rate": 7.181860238173605e-05, "loss": 0.0341, "step": 7770 }, { "grad_norm": 0.41947415471076965, "learning_rate": 7.174418576711432e-05, "loss": 0.0351, "step": 7780 }, { "grad_norm": 0.44200241565704346, "learning_rate": 7.1669709704719e-05, "loss": 0.0316, "step": 7790 }, { "grad_norm": 0.39203381538391113, "learning_rate": 7.159517439816481e-05, "loss": 0.0347, "step": 7800 }, { "grad_norm": 0.3449641466140747, "learning_rate": 7.152058005122842e-05, "loss": 0.0314, "step": 7810 }, { "grad_norm": 0.35387441515922546, "learning_rate": 7.144592686784793e-05, "loss": 0.0337, "step": 7820 }, { "grad_norm": 0.3497450351715088, "learning_rate": 7.137121505212229e-05, "loss": 0.0332, "step": 7830 }, { "grad_norm": 0.4479694664478302, "learning_rate": 7.129644480831077e-05, "loss": 0.0333, "step": 7840 }, { "grad_norm": 0.4464643597602844, "learning_rate": 7.122161634083234e-05, "loss": 0.0333, "step": 7850 }, { "grad_norm": 0.28151991963386536, "learning_rate": 7.114672985426516e-05, "loss": 0.031, "step": 7860 }, { "grad_norm": 0.40099674463272095, "learning_rate": 7.107178555334606e-05, "loss": 0.0312, "step": 7870 }, { "grad_norm": 0.3103621304035187, "learning_rate": 7.099678364296989e-05, "loss": 0.0307, "step": 7880 }, { "grad_norm": 0.3409649133682251, "learning_rate": 7.0921724328189e-05, "loss": 0.0321, "step": 7890 }, { "grad_norm": 0.3594605624675751, "learning_rate": 7.084660781421268e-05, "loss": 0.0298, "step": 7900 }, { "grad_norm": 0.33482658863067627, "learning_rate": 7.077143430640662e-05, "loss": 0.0324, "step": 7910 }, { "grad_norm": 0.3721316158771515, "learning_rate": 7.069620401029232e-05, "loss": 0.0313, "step": 7920 }, { "grad_norm": 0.3950805366039276, "learning_rate": 7.062091713154655e-05, "loss": 0.0306, "step": 7930 }, { "grad_norm": 0.33270686864852905, "learning_rate": 7.054557387600075e-05, "loss": 0.0312, "step": 7940 }, { "grad_norm": 0.4002479612827301, "learning_rate": 7.04701744496405e-05, "loss": 0.0336, "step": 7950 }, { "grad_norm": 0.36822566390037537, "learning_rate": 7.039471905860495e-05, "loss": 0.0329, "step": 7960 }, { "grad_norm": 0.36939260363578796, "learning_rate": 7.031920790918628e-05, "loss": 0.0337, "step": 7970 }, { "grad_norm": 0.3197311460971832, "learning_rate": 7.024364120782906e-05, "loss": 0.0327, "step": 7980 }, { "grad_norm": 0.37314367294311523, "learning_rate": 7.016801916112978e-05, "loss": 0.0332, "step": 7990 }, { "grad_norm": 0.36284559965133667, "learning_rate": 7.009234197583623e-05, "loss": 0.0328, "step": 8000 }, { "grad_norm": 0.34743204712867737, "learning_rate": 7.001660985884692e-05, "loss": 0.0343, "step": 8010 }, { "grad_norm": 0.3687737286090851, "learning_rate": 6.994082301721063e-05, "loss": 0.0311, "step": 8020 }, { "grad_norm": 0.3445630967617035, "learning_rate": 6.986498165812563e-05, "loss": 0.0332, "step": 8030 }, { "grad_norm": 0.3472203314304352, "learning_rate": 6.978908598893932e-05, "loss": 0.0326, "step": 8040 }, { "grad_norm": 0.31326931715011597, "learning_rate": 6.971313621714756e-05, "loss": 0.0312, "step": 8050 }, { "grad_norm": 0.3364187479019165, "learning_rate": 6.96371325503941e-05, "loss": 0.0325, "step": 8060 }, { "grad_norm": 0.41103580594062805, "learning_rate": 6.956107519647014e-05, "loss": 0.0308, "step": 8070 }, { "grad_norm": 0.31624966859817505, "learning_rate": 6.94849643633135e-05, "loss": 0.033, "step": 8080 }, { "grad_norm": 0.38413509726524353, "learning_rate": 6.940880025900834e-05, "loss": 0.0322, "step": 8090 }, { "grad_norm": 0.36665037274360657, "learning_rate": 6.933258309178438e-05, "loss": 0.0317, "step": 8100 }, { "grad_norm": 0.38233819603919983, "learning_rate": 6.925631307001646e-05, "loss": 0.0344, "step": 8110 }, { "grad_norm": 0.33661946654319763, "learning_rate": 6.91799904022239e-05, "loss": 0.0318, "step": 8120 }, { "grad_norm": 0.2750566303730011, "learning_rate": 6.910361529706997e-05, "loss": 0.0329, "step": 8130 }, { "grad_norm": 0.3372498154640198, "learning_rate": 6.902718796336131e-05, "loss": 0.0326, "step": 8140 }, { "grad_norm": 0.3901897072792053, "learning_rate": 6.895070861004729e-05, "loss": 0.0328, "step": 8150 }, { "grad_norm": 0.38243699073791504, "learning_rate": 6.887417744621956e-05, "loss": 0.0321, "step": 8160 }, { "grad_norm": 0.33180657029151917, "learning_rate": 6.87975946811114e-05, "loss": 0.0319, "step": 8170 }, { "grad_norm": 0.46084290742874146, "learning_rate": 6.872096052409718e-05, "loss": 0.032, "step": 8180 }, { "grad_norm": 0.3516644239425659, "learning_rate": 6.864427518469174e-05, "loss": 0.0327, "step": 8190 }, { "grad_norm": 0.2888367772102356, "learning_rate": 6.856753887254986e-05, "loss": 0.0336, "step": 8200 }, { "grad_norm": 0.33609098196029663, "learning_rate": 6.849075179746572e-05, "loss": 0.0302, "step": 8210 }, { "grad_norm": 0.30836594104766846, "learning_rate": 6.841391416937221e-05, "loss": 0.0305, "step": 8220 }, { "grad_norm": 0.25423118472099304, "learning_rate": 6.833702619834053e-05, "loss": 0.0308, "step": 8230 }, { "grad_norm": 0.33365368843078613, "learning_rate": 6.82600880945794e-05, "loss": 0.0307, "step": 8240 }, { "grad_norm": 0.3425358235836029, "learning_rate": 6.818310006843468e-05, "loss": 0.0311, "step": 8250 }, { "grad_norm": 0.3232588768005371, "learning_rate": 6.810606233038868e-05, "loss": 0.0296, "step": 8260 }, { "grad_norm": 0.3485594689846039, "learning_rate": 6.802897509105966e-05, "loss": 0.0295, "step": 8270 }, { "grad_norm": 0.35349369049072266, "learning_rate": 6.79518385612012e-05, "loss": 0.0292, "step": 8280 }, { "grad_norm": 0.31503310799598694, "learning_rate": 6.787465295170157e-05, "loss": 0.0304, "step": 8290 }, { "grad_norm": 0.3930913209915161, "learning_rate": 6.779741847358332e-05, "loss": 0.0299, "step": 8300 }, { "grad_norm": 0.43458643555641174, "learning_rate": 6.772013533800256e-05, "loss": 0.033, "step": 8310 }, { "grad_norm": 0.37892040610313416, "learning_rate": 6.764280375624843e-05, "loss": 0.0305, "step": 8320 }, { "grad_norm": 0.3377176821231842, "learning_rate": 6.756542393974252e-05, "loss": 0.0318, "step": 8330 }, { "grad_norm": 0.3355208933353424, "learning_rate": 6.748799610003828e-05, "loss": 0.0321, "step": 8340 }, { "grad_norm": 0.33350318670272827, "learning_rate": 6.741052044882048e-05, "loss": 0.0294, "step": 8350 }, { "grad_norm": 0.31123825907707214, "learning_rate": 6.73329971979046e-05, "loss": 0.0318, "step": 8360 }, { "grad_norm": 0.3478899300098419, "learning_rate": 6.725542655923625e-05, "loss": 0.0328, "step": 8370 }, { "grad_norm": 0.33028343319892883, "learning_rate": 6.717780874489057e-05, "loss": 0.0319, "step": 8380 }, { "grad_norm": 0.37385088205337524, "learning_rate": 6.710014396707172e-05, "loss": 0.0323, "step": 8390 }, { "grad_norm": 0.36853834986686707, "learning_rate": 6.702243243811221e-05, "loss": 0.0299, "step": 8400 }, { "grad_norm": 0.3731279671192169, "learning_rate": 6.694467437047244e-05, "loss": 0.0319, "step": 8410 }, { "grad_norm": 0.33576029539108276, "learning_rate": 6.686686997673997e-05, "loss": 0.0304, "step": 8420 }, { "grad_norm": 0.3119388222694397, "learning_rate": 6.678901946962903e-05, "loss": 0.0301, "step": 8430 }, { "grad_norm": 0.31896984577178955, "learning_rate": 6.671112306197996e-05, "loss": 0.0301, "step": 8440 }, { "grad_norm": 0.3219243884086609, "learning_rate": 6.663318096675854e-05, "loss": 0.0299, "step": 8450 }, { "grad_norm": 0.3318690061569214, "learning_rate": 6.655519339705552e-05, "loss": 0.0306, "step": 8460 }, { "grad_norm": 0.4354896545410156, "learning_rate": 6.647716056608588e-05, "loss": 0.032, "step": 8470 }, { "grad_norm": 0.394074946641922, "learning_rate": 6.639908268718843e-05, "loss": 0.0328, "step": 8480 }, { "grad_norm": 0.2799919545650482, "learning_rate": 6.632095997382514e-05, "loss": 0.0314, "step": 8490 }, { "grad_norm": 0.3137683868408203, "learning_rate": 6.624279263958047e-05, "loss": 0.0299, "step": 8500 }, { "grad_norm": 0.2940962314605713, "learning_rate": 6.616458089816097e-05, "loss": 0.031, "step": 8510 }, { "grad_norm": 0.31405109167099, "learning_rate": 6.608632496339454e-05, "loss": 0.0302, "step": 8520 }, { "grad_norm": 0.3612689673900604, "learning_rate": 6.600802504922988e-05, "loss": 0.0291, "step": 8530 }, { "grad_norm": 0.3423871397972107, "learning_rate": 6.592968136973604e-05, "loss": 0.0312, "step": 8540 }, { "grad_norm": 0.3506467938423157, "learning_rate": 6.585129413910159e-05, "loss": 0.0317, "step": 8550 }, { "grad_norm": 0.34871047735214233, "learning_rate": 6.577286357163424e-05, "loss": 0.0299, "step": 8560 }, { "grad_norm": 0.3316114544868469, "learning_rate": 6.569438988176018e-05, "loss": 0.031, "step": 8570 }, { "grad_norm": 0.31105852127075195, "learning_rate": 6.561587328402347e-05, "loss": 0.0316, "step": 8580 }, { "grad_norm": 0.3708553910255432, "learning_rate": 6.553731399308549e-05, "loss": 0.0347, "step": 8590 }, { "grad_norm": 0.2936054766178131, "learning_rate": 6.545871222372436e-05, "loss": 0.0303, "step": 8600 }, { "grad_norm": 0.3593730926513672, "learning_rate": 6.538006819083426e-05, "loss": 0.0308, "step": 8610 }, { "grad_norm": 0.3295312523841858, "learning_rate": 6.530138210942505e-05, "loss": 0.0304, "step": 8620 }, { "grad_norm": 0.372282475233078, "learning_rate": 6.522265419462141e-05, "loss": 0.0301, "step": 8630 }, { "grad_norm": 0.38826271891593933, "learning_rate": 6.514388466166248e-05, "loss": 0.0333, "step": 8640 }, { "grad_norm": 0.38165098428726196, "learning_rate": 6.506507372590119e-05, "loss": 0.033, "step": 8650 }, { "grad_norm": 0.3342225253582001, "learning_rate": 6.498622160280355e-05, "loss": 0.0313, "step": 8660 }, { "grad_norm": 0.2735998034477234, "learning_rate": 6.490732850794832e-05, "loss": 0.0302, "step": 8670 }, { "grad_norm": 0.3015640676021576, "learning_rate": 6.482839465702616e-05, "loss": 0.0288, "step": 8680 }, { "grad_norm": 0.3979356586933136, "learning_rate": 6.474942026583923e-05, "loss": 0.0327, "step": 8690 }, { "grad_norm": 0.33981460332870483, "learning_rate": 6.467040555030052e-05, "loss": 0.0306, "step": 8700 }, { "grad_norm": 0.3290560245513916, "learning_rate": 6.459135072643321e-05, "loss": 0.0305, "step": 8710 }, { "grad_norm": 0.29458317160606384, "learning_rate": 6.451225601037019e-05, "loss": 0.0304, "step": 8720 }, { "grad_norm": 0.3136504590511322, "learning_rate": 6.443312161835338e-05, "loss": 0.0311, "step": 8730 }, { "grad_norm": 0.3427092134952545, "learning_rate": 6.43539477667332e-05, "loss": 0.0308, "step": 8740 }, { "grad_norm": 0.3662634789943695, "learning_rate": 6.427473467196793e-05, "loss": 0.0304, "step": 8750 }, { "grad_norm": 0.3307614326477051, "learning_rate": 6.419548255062315e-05, "loss": 0.0289, "step": 8760 }, { "grad_norm": 0.33490097522735596, "learning_rate": 6.411619161937112e-05, "loss": 0.0307, "step": 8770 }, { "grad_norm": 0.3141922056674957, "learning_rate": 6.403686209499022e-05, "loss": 0.0294, "step": 8780 }, { "grad_norm": 0.2485896199941635, "learning_rate": 6.395749419436437e-05, "loss": 0.0281, "step": 8790 }, { "grad_norm": 0.34049034118652344, "learning_rate": 6.387808813448234e-05, "loss": 0.0273, "step": 8800 }, { "grad_norm": 0.3435012698173523, "learning_rate": 6.37986441324373e-05, "loss": 0.0317, "step": 8810 }, { "grad_norm": 0.33055153489112854, "learning_rate": 6.37191624054261e-05, "loss": 0.0284, "step": 8820 }, { "grad_norm": 0.2825639247894287, "learning_rate": 6.363964317074872e-05, "loss": 0.0296, "step": 8830 }, { "grad_norm": 0.3263731300830841, "learning_rate": 6.356008664580776e-05, "loss": 0.0307, "step": 8840 }, { "grad_norm": 0.37316974997520447, "learning_rate": 6.348049304810771e-05, "loss": 0.03, "step": 8850 }, { "grad_norm": 0.3351162075996399, "learning_rate": 6.340086259525442e-05, "loss": 0.0316, "step": 8860 }, { "grad_norm": 0.2722276449203491, "learning_rate": 6.332119550495448e-05, "loss": 0.0285, "step": 8870 }, { "grad_norm": 0.3693591058254242, "learning_rate": 6.324149199501473e-05, "loss": 0.0311, "step": 8880 }, { "grad_norm": 0.34115368127822876, "learning_rate": 6.316175228334146e-05, "loss": 0.0302, "step": 8890 }, { "grad_norm": 0.3268924057483673, "learning_rate": 6.308197658794003e-05, "loss": 0.0293, "step": 8900 }, { "grad_norm": 0.29654431343078613, "learning_rate": 6.300216512691417e-05, "loss": 0.0288, "step": 8910 }, { "grad_norm": 0.3543151617050171, "learning_rate": 6.292231811846532e-05, "loss": 0.03, "step": 8920 }, { "grad_norm": 0.3144560754299164, "learning_rate": 6.284243578089217e-05, "loss": 0.0281, "step": 8930 }, { "grad_norm": 0.29155558347702026, "learning_rate": 6.276251833258999e-05, "loss": 0.0282, "step": 8940 }, { "grad_norm": 0.30916574597358704, "learning_rate": 6.268256599205003e-05, "loss": 0.0278, "step": 8950 }, { "grad_norm": 0.33211639523506165, "learning_rate": 6.260257897785892e-05, "loss": 0.0282, "step": 8960 }, { "grad_norm": 0.36915016174316406, "learning_rate": 6.252255750869811e-05, "loss": 0.0287, "step": 8970 }, { "grad_norm": 0.2836068272590637, "learning_rate": 6.244250180334325e-05, "loss": 0.0292, "step": 8980 }, { "grad_norm": 0.292288213968277, "learning_rate": 6.236241208066356e-05, "loss": 0.0276, "step": 8990 }, { "grad_norm": 0.32288089394569397, "learning_rate": 6.228228855962133e-05, "loss": 0.031, "step": 9000 }, { "grad_norm": 0.254332035779953, "learning_rate": 6.220213145927115e-05, "loss": 0.0286, "step": 9010 }, { "grad_norm": 0.2761654853820801, "learning_rate": 6.212194099875951e-05, "loss": 0.0285, "step": 9020 }, { "grad_norm": 0.3462757468223572, "learning_rate": 6.204171739732405e-05, "loss": 0.0289, "step": 9030 }, { "grad_norm": 0.32191091775894165, "learning_rate": 6.196146087429303e-05, "loss": 0.0268, "step": 9040 }, { "grad_norm": 0.3476082682609558, "learning_rate": 6.188117164908474e-05, "loss": 0.0295, "step": 9050 }, { "grad_norm": 0.314022958278656, "learning_rate": 6.180084994120684e-05, "loss": 0.0305, "step": 9060 }, { "grad_norm": 0.308940052986145, "learning_rate": 6.17204959702558e-05, "loss": 0.0303, "step": 9070 }, { "grad_norm": 0.30428552627563477, "learning_rate": 6.164010995591635e-05, "loss": 0.0298, "step": 9080 }, { "grad_norm": 0.33882665634155273, "learning_rate": 6.155969211796076e-05, "loss": 0.029, "step": 9090 }, { "grad_norm": 0.3185388445854187, "learning_rate": 6.147924267624829e-05, "loss": 0.0329, "step": 9100 }, { "grad_norm": 0.3034641146659851, "learning_rate": 6.13987618507247e-05, "loss": 0.0292, "step": 9110 }, { "grad_norm": 0.3364293575286865, "learning_rate": 6.131824986142147e-05, "loss": 0.0291, "step": 9120 }, { "grad_norm": 0.2786848545074463, "learning_rate": 6.123770692845529e-05, "loss": 0.0292, "step": 9130 }, { "grad_norm": 0.3995400071144104, "learning_rate": 6.11571332720275e-05, "loss": 0.029, "step": 9140 }, { "grad_norm": 0.3171330988407135, "learning_rate": 6.107652911242336e-05, "loss": 0.0293, "step": 9150 }, { "grad_norm": 0.2955140471458435, "learning_rate": 6.0995894670011586e-05, "loss": 0.0299, "step": 9160 }, { "grad_norm": 0.30255284905433655, "learning_rate": 6.091523016524368e-05, "loss": 0.0278, "step": 9170 }, { "grad_norm": 0.3838665783405304, "learning_rate": 6.083453581865328e-05, "loss": 0.0285, "step": 9180 }, { "grad_norm": 0.36146125197410583, "learning_rate": 6.075381185085568e-05, "loss": 0.0285, "step": 9190 }, { "grad_norm": 0.2929210662841797, "learning_rate": 6.067305848254709e-05, "loss": 0.0282, "step": 9200 }, { "grad_norm": 0.3012180030345917, "learning_rate": 6.059227593450418e-05, "loss": 0.0286, "step": 9210 }, { "grad_norm": 0.3448551595211029, "learning_rate": 6.051146442758333e-05, "loss": 0.0298, "step": 9220 }, { "grad_norm": 0.36912456154823303, "learning_rate": 6.043062418272012e-05, "loss": 0.0297, "step": 9230 }, { "grad_norm": 0.3273569643497467, "learning_rate": 6.0349755420928666e-05, "loss": 0.0284, "step": 9240 }, { "grad_norm": 0.31341856718063354, "learning_rate": 6.0268858363301105e-05, "loss": 0.0282, "step": 9250 }, { "grad_norm": 0.27713340520858765, "learning_rate": 6.018793323100689e-05, "loss": 0.0307, "step": 9260 }, { "grad_norm": 0.3550170361995697, "learning_rate": 6.0106980245292255e-05, "loss": 0.03, "step": 9270 }, { "grad_norm": 0.38089385628700256, "learning_rate": 6.002599962747957e-05, "loss": 0.0283, "step": 9280 }, { "grad_norm": 0.33127492666244507, "learning_rate": 5.994499159896673e-05, "loss": 0.0288, "step": 9290 }, { "grad_norm": 0.2906492352485657, "learning_rate": 5.9863956381226607e-05, "loss": 0.0297, "step": 9300 }, { "grad_norm": 0.3067975640296936, "learning_rate": 5.9782894195806394e-05, "loss": 0.0284, "step": 9310 }, { "grad_norm": 0.3185523450374603, "learning_rate": 5.9701805264327004e-05, "loss": 0.0292, "step": 9320 }, { "grad_norm": 0.3392626643180847, "learning_rate": 5.96206898084825e-05, "loss": 0.0305, "step": 9330 }, { "grad_norm": 0.3309928774833679, "learning_rate": 5.953954805003942e-05, "loss": 0.029, "step": 9340 }, { "grad_norm": 0.3263966143131256, "learning_rate": 5.945838021083623e-05, "loss": 0.0289, "step": 9350 }, { "grad_norm": 0.2927856147289276, "learning_rate": 5.9377186512782714e-05, "loss": 0.0293, "step": 9360 }, { "grad_norm": 0.282087504863739, "learning_rate": 5.929596717785935e-05, "loss": 0.0293, "step": 9370 }, { "grad_norm": 0.32162657380104065, "learning_rate": 5.921472242811668e-05, "loss": 0.0311, "step": 9380 }, { "grad_norm": 0.3156202733516693, "learning_rate": 5.913345248567475e-05, "loss": 0.03, "step": 9390 }, { "grad_norm": 0.3575294017791748, "learning_rate": 5.905215757272248e-05, "loss": 0.0306, "step": 9400 }, { "grad_norm": 0.38162529468536377, "learning_rate": 5.897083791151706e-05, "loss": 0.0301, "step": 9410 }, { "grad_norm": 0.4070315659046173, "learning_rate": 5.888949372438336e-05, "loss": 0.0286, "step": 9420 }, { "grad_norm": 0.27342599630355835, "learning_rate": 5.8808125233713255e-05, "loss": 0.0303, "step": 9430 }, { "grad_norm": 0.3448978662490845, "learning_rate": 5.872673266196509e-05, "loss": 0.0304, "step": 9440 }, { "grad_norm": 0.3333384692668915, "learning_rate": 5.864531623166305e-05, "loss": 0.0287, "step": 9450 }, { "grad_norm": 0.28400591015815735, "learning_rate": 5.856387616539656e-05, "loss": 0.0291, "step": 9460 }, { "grad_norm": 0.26151156425476074, "learning_rate": 5.848241268581967e-05, "loss": 0.0297, "step": 9470 }, { "grad_norm": 0.31600454449653625, "learning_rate": 5.840092601565037e-05, "loss": 0.0281, "step": 9480 }, { "grad_norm": 0.3280591070652008, "learning_rate": 5.8319416377670144e-05, "loss": 0.0295, "step": 9490 }, { "grad_norm": 0.3647940158843994, "learning_rate": 5.82378839947232e-05, "loss": 0.0289, "step": 9500 }, { "grad_norm": 0.3250693380832672, "learning_rate": 5.815632908971599e-05, "loss": 0.0271, "step": 9510 }, { "grad_norm": 0.3595479130744934, "learning_rate": 5.80747518856165e-05, "loss": 0.0307, "step": 9520 }, { "grad_norm": 0.3656829297542572, "learning_rate": 5.799315260545367e-05, "loss": 0.0277, "step": 9530 }, { "grad_norm": 0.3162411153316498, "learning_rate": 5.791153147231686e-05, "loss": 0.0297, "step": 9540 }, { "grad_norm": 0.33892256021499634, "learning_rate": 5.782988870935509e-05, "loss": 0.028, "step": 9550 }, { "grad_norm": 0.3145725727081299, "learning_rate": 5.774822453977657e-05, "loss": 0.0269, "step": 9560 }, { "grad_norm": 0.348692387342453, "learning_rate": 5.7666539186848036e-05, "loss": 0.0275, "step": 9570 }, { "grad_norm": 0.29440540075302124, "learning_rate": 5.758483287389411e-05, "loss": 0.0286, "step": 9580 }, { "grad_norm": 0.2720487415790558, "learning_rate": 5.7503105824296735e-05, "loss": 0.0279, "step": 9590 }, { "grad_norm": 0.33431124687194824, "learning_rate": 5.742135826149453e-05, "loss": 0.0275, "step": 9600 }, { "grad_norm": 0.29012277722358704, "learning_rate": 5.7339590408982223e-05, "loss": 0.0271, "step": 9610 }, { "grad_norm": 0.278516560792923, "learning_rate": 5.725780249031e-05, "loss": 0.0291, "step": 9620 }, { "grad_norm": 0.3895028531551361, "learning_rate": 5.717599472908292e-05, "loss": 0.0277, "step": 9630 }, { "grad_norm": 0.27406060695648193, "learning_rate": 5.7094167348960237e-05, "loss": 0.0279, "step": 9640 }, { "grad_norm": 0.3252066969871521, "learning_rate": 5.7012320573654945e-05, "loss": 0.0286, "step": 9650 }, { "grad_norm": 0.3357921242713928, "learning_rate": 5.693045462693295e-05, "loss": 0.0275, "step": 9660 }, { "grad_norm": 0.3069935739040375, "learning_rate": 5.684856973261266e-05, "loss": 0.0288, "step": 9670 }, { "grad_norm": 0.40459972620010376, "learning_rate": 5.6766666114564215e-05, "loss": 0.0271, "step": 9680 }, { "grad_norm": 0.32910487055778503, "learning_rate": 5.668474399670899e-05, "loss": 0.0299, "step": 9690 }, { "grad_norm": 0.31085216999053955, "learning_rate": 5.660280360301896e-05, "loss": 0.0264, "step": 9700 }, { "grad_norm": 0.25815442204475403, "learning_rate": 5.652084515751599e-05, "loss": 0.0278, "step": 9710 }, { "grad_norm": 0.23869329690933228, "learning_rate": 5.643886888427137e-05, "loss": 0.0276, "step": 9720 }, { "grad_norm": 0.28807342052459717, "learning_rate": 5.6356875007405074e-05, "loss": 0.0272, "step": 9730 }, { "grad_norm": 0.3306489586830139, "learning_rate": 5.627486375108525e-05, "loss": 0.0274, "step": 9740 }, { "grad_norm": 0.36949414014816284, "learning_rate": 5.619283533952754e-05, "loss": 0.0276, "step": 9750 }, { "grad_norm": 0.3529314696788788, "learning_rate": 5.6110789996994474e-05, "loss": 0.027, "step": 9760 }, { "grad_norm": 0.3026208281517029, "learning_rate": 5.602872794779491e-05, "loss": 0.027, "step": 9770 }, { "grad_norm": 0.3088509738445282, "learning_rate": 5.594664941628334e-05, "loss": 0.0266, "step": 9780 }, { "grad_norm": 0.28330138325691223, "learning_rate": 5.5864554626859324e-05, "loss": 0.0277, "step": 9790 }, { "grad_norm": 0.3052361011505127, "learning_rate": 5.578244380396691e-05, "loss": 0.0286, "step": 9800 }, { "grad_norm": 0.31326109170913696, "learning_rate": 5.570031717209394e-05, "loss": 0.0282, "step": 9810 }, { "grad_norm": 0.36193203926086426, "learning_rate": 5.561817495577147e-05, "loss": 0.0299, "step": 9820 }, { "grad_norm": 0.3179263174533844, "learning_rate": 5.5536017379573215e-05, "loss": 0.0295, "step": 9830 }, { "grad_norm": 0.3034178614616394, "learning_rate": 5.545384466811483e-05, "loss": 0.0295, "step": 9840 }, { "grad_norm": 0.32758182287216187, "learning_rate": 5.5371657046053384e-05, "loss": 0.0287, "step": 9850 }, { "grad_norm": 0.3008367419242859, "learning_rate": 5.528945473808669e-05, "loss": 0.0285, "step": 9860 }, { "grad_norm": 0.24443472921848297, "learning_rate": 5.520723796895272e-05, "loss": 0.0278, "step": 9870 }, { "grad_norm": 0.3161213994026184, "learning_rate": 5.512500696342897e-05, "loss": 0.0258, "step": 9880 }, { "grad_norm": 0.3906218409538269, "learning_rate": 5.504276194633188e-05, "loss": 0.0275, "step": 9890 }, { "grad_norm": 0.2700369656085968, "learning_rate": 5.49605031425162e-05, "loss": 0.0282, "step": 9900 }, { "grad_norm": 0.23842576146125793, "learning_rate": 5.487823077687434e-05, "loss": 0.0282, "step": 9910 }, { "grad_norm": 0.2773781716823578, "learning_rate": 5.4795945074335806e-05, "loss": 0.0278, "step": 9920 }, { "grad_norm": 0.2632705867290497, "learning_rate": 5.471364625986657e-05, "loss": 0.0282, "step": 9930 }, { "grad_norm": 0.2873260974884033, "learning_rate": 5.463133455846845e-05, "loss": 0.0277, "step": 9940 }, { "grad_norm": 0.3415659964084625, "learning_rate": 5.4549010195178505e-05, "loss": 0.0281, "step": 9950 }, { "grad_norm": 0.28778907656669617, "learning_rate": 5.446667339506838e-05, "loss": 0.0282, "step": 9960 }, { "grad_norm": 0.31027838587760925, "learning_rate": 5.4384324383243756e-05, "loss": 0.0285, "step": 9970 }, { "grad_norm": 0.2910730540752411, "learning_rate": 5.430196338484368e-05, "loss": 0.0278, "step": 9980 }, { "grad_norm": 0.3089366555213928, "learning_rate": 5.4219590625039975e-05, "loss": 0.0254, "step": 9990 }, { "grad_norm": 0.3694930970668793, "learning_rate": 5.413720632903664e-05, "loss": 0.0292, "step": 10000 }, { "grad_norm": 0.29372188448905945, "learning_rate": 5.405481072206917e-05, "loss": 0.0295, "step": 10010 }, { "grad_norm": 0.26854264736175537, "learning_rate": 5.397240402940402e-05, "loss": 0.0297, "step": 10020 }, { "grad_norm": 0.21098779141902924, "learning_rate": 5.388998647633794e-05, "loss": 0.026, "step": 10030 }, { "grad_norm": 0.3128412067890167, "learning_rate": 5.380755828819737e-05, "loss": 0.0286, "step": 10040 }, { "grad_norm": 0.2690325975418091, "learning_rate": 5.3725119690337846e-05, "loss": 0.0269, "step": 10050 }, { "grad_norm": 0.313642680644989, "learning_rate": 5.3642670908143324e-05, "loss": 0.0258, "step": 10060 }, { "grad_norm": 0.30513402819633484, "learning_rate": 5.356021216702562e-05, "loss": 0.0277, "step": 10070 }, { "grad_norm": 0.3040320575237274, "learning_rate": 5.347774369242381e-05, "loss": 0.0296, "step": 10080 }, { "grad_norm": 0.3620952367782593, "learning_rate": 5.3395265709803545e-05, "loss": 0.0281, "step": 10090 }, { "grad_norm": 0.32502326369285583, "learning_rate": 5.331277844465647e-05, "loss": 0.0281, "step": 10100 }, { "grad_norm": 0.25353220105171204, "learning_rate": 5.323028212249963e-05, "loss": 0.0283, "step": 10110 }, { "grad_norm": 0.3063226342201233, "learning_rate": 5.314777696887481e-05, "loss": 0.0267, "step": 10120 }, { "grad_norm": 0.29310786724090576, "learning_rate": 5.306526320934796e-05, "loss": 0.0274, "step": 10130 }, { "grad_norm": 0.2712216377258301, "learning_rate": 5.298274106950854e-05, "loss": 0.0279, "step": 10140 }, { "grad_norm": 0.3502790033817291, "learning_rate": 5.290021077496893e-05, "loss": 0.0263, "step": 10150 }, { "grad_norm": 0.3084924817085266, "learning_rate": 5.2817672551363816e-05, "loss": 0.027, "step": 10160 }, { "grad_norm": 0.25670862197875977, "learning_rate": 5.273512662434952e-05, "loss": 0.0268, "step": 10170 }, { "grad_norm": 0.28704017400741577, "learning_rate": 5.265257321960349e-05, "loss": 0.0286, "step": 10180 }, { "grad_norm": 0.2519485652446747, "learning_rate": 5.257001256282357e-05, "loss": 0.0272, "step": 10190 }, { "grad_norm": 0.2839933931827545, "learning_rate": 5.248744487972742e-05, "loss": 0.0258, "step": 10200 }, { "grad_norm": 0.24591894447803497, "learning_rate": 5.240487039605196e-05, "loss": 0.0269, "step": 10210 }, { "grad_norm": 0.320033997297287, "learning_rate": 5.232228933755267e-05, "loss": 0.027, "step": 10220 }, { "grad_norm": 0.31525081396102905, "learning_rate": 5.2239701930003006e-05, "loss": 0.0268, "step": 10230 }, { "grad_norm": 0.30413728952407837, "learning_rate": 5.215710839919379e-05, "loss": 0.027, "step": 10240 }, { "grad_norm": 0.285496324300766, "learning_rate": 5.207450897093257e-05, "loss": 0.0266, "step": 10250 }, { "grad_norm": 0.24853327870368958, "learning_rate": 5.1991903871043046e-05, "loss": 0.0269, "step": 10260 }, { "grad_norm": 0.2545941472053528, "learning_rate": 5.190929332536439e-05, "loss": 0.0259, "step": 10270 }, { "grad_norm": 0.2264733612537384, "learning_rate": 5.182667755975071e-05, "loss": 0.0272, "step": 10280 }, { "grad_norm": 0.3716336786746979, "learning_rate": 5.1744056800070315e-05, "loss": 0.0262, "step": 10290 }, { "grad_norm": 0.2962269186973572, "learning_rate": 5.166143127220524e-05, "loss": 0.0283, "step": 10300 }, { "grad_norm": 0.26524618268013, "learning_rate": 5.1578801202050485e-05, "loss": 0.0263, "step": 10310 }, { "grad_norm": 0.26325875520706177, "learning_rate": 5.149616681551355e-05, "loss": 0.0268, "step": 10320 }, { "grad_norm": 0.31458163261413574, "learning_rate": 5.141352833851367e-05, "loss": 0.0264, "step": 10330 }, { "grad_norm": 0.2847176194190979, "learning_rate": 5.1330885996981285e-05, "loss": 0.0268, "step": 10340 }, { "grad_norm": 0.3532206118106842, "learning_rate": 5.124824001685741e-05, "loss": 0.029, "step": 10350 }, { "grad_norm": 0.29204443097114563, "learning_rate": 5.116559062409298e-05, "loss": 0.0272, "step": 10360 }, { "grad_norm": 0.25431933999061584, "learning_rate": 5.10829380446483e-05, "loss": 0.0264, "step": 10370 }, { "grad_norm": 0.3363592326641083, "learning_rate": 5.100028250449235e-05, "loss": 0.0266, "step": 10380 }, { "grad_norm": 0.27360281348228455, "learning_rate": 5.0917624229602234e-05, "loss": 0.0259, "step": 10390 }, { "grad_norm": 0.2973903715610504, "learning_rate": 5.0834963445962524e-05, "loss": 0.0261, "step": 10400 }, { "grad_norm": 0.3245631158351898, "learning_rate": 5.075230037956461e-05, "loss": 0.0263, "step": 10410 }, { "grad_norm": 0.29883265495300293, "learning_rate": 5.0669635256406213e-05, "loss": 0.0256, "step": 10420 }, { "grad_norm": 0.3094143867492676, "learning_rate": 5.058696830249058e-05, "loss": 0.0271, "step": 10430 }, { "grad_norm": 0.2976891100406647, "learning_rate": 5.050429974382602e-05, "loss": 0.0266, "step": 10440 }, { "grad_norm": 0.2858240604400635, "learning_rate": 5.042162980642523e-05, "loss": 0.0282, "step": 10450 }, { "grad_norm": 0.26014000177383423, "learning_rate": 5.033895871630462e-05, "loss": 0.0258, "step": 10460 }, { "grad_norm": 0.2775442898273468, "learning_rate": 5.025628669948386e-05, "loss": 0.0258, "step": 10470 }, { "grad_norm": 0.2859240174293518, "learning_rate": 5.017361398198502e-05, "loss": 0.0264, "step": 10480 }, { "grad_norm": 0.28274884819984436, "learning_rate": 5.009094078983221e-05, "loss": 0.0273, "step": 10490 }, { "grad_norm": 0.2919665575027466, "learning_rate": 5.000826734905073e-05, "loss": 0.0254, "step": 10500 }, { "grad_norm": 0.2655998468399048, "learning_rate": 4.9925593885666645e-05, "loss": 0.0261, "step": 10510 }, { "grad_norm": 0.30897948145866394, "learning_rate": 4.984292062570602e-05, "loss": 0.0284, "step": 10520 }, { "grad_norm": 0.2610779106616974, "learning_rate": 4.976024779519442e-05, "loss": 0.0255, "step": 10530 }, { "grad_norm": 0.29545003175735474, "learning_rate": 4.9677575620156194e-05, "loss": 0.0264, "step": 10540 }, { "grad_norm": 0.2925003170967102, "learning_rate": 4.959490432661391e-05, "loss": 0.0272, "step": 10550 }, { "grad_norm": 0.2587912976741791, "learning_rate": 4.9512234140587726e-05, "loss": 0.0252, "step": 10560 }, { "grad_norm": 0.24825049936771393, "learning_rate": 4.942956528809477e-05, "loss": 0.0285, "step": 10570 }, { "grad_norm": 0.26931285858154297, "learning_rate": 4.934689799514854e-05, "loss": 0.0266, "step": 10580 }, { "grad_norm": 0.3153868615627289, "learning_rate": 4.926423248775827e-05, "loss": 0.0263, "step": 10590 }, { "grad_norm": 0.3746855854988098, "learning_rate": 4.918156899192826e-05, "loss": 0.0283, "step": 10600 }, { "grad_norm": 0.2746714651584625, "learning_rate": 4.909890773365738e-05, "loss": 0.0298, "step": 10610 }, { "grad_norm": 0.21346977353096008, "learning_rate": 4.9016248938938344e-05, "loss": 0.0267, "step": 10620 }, { "grad_norm": 0.2732856273651123, "learning_rate": 4.8933592833757156e-05, "loss": 0.0269, "step": 10630 }, { "grad_norm": 0.2723126709461212, "learning_rate": 4.8850939644092435e-05, "loss": 0.0277, "step": 10640 }, { "grad_norm": 0.2625075876712799, "learning_rate": 4.876828959591485e-05, "loss": 0.0263, "step": 10650 }, { "grad_norm": 0.35670316219329834, "learning_rate": 4.8685642915186474e-05, "loss": 0.0275, "step": 10660 }, { "grad_norm": 0.3121415376663208, "learning_rate": 4.860299982786018e-05, "loss": 0.0261, "step": 10670 }, { "grad_norm": 0.22707031667232513, "learning_rate": 4.852036055987901e-05, "loss": 0.0264, "step": 10680 }, { "grad_norm": 0.2499329298734665, "learning_rate": 4.843772533717558e-05, "loss": 0.0251, "step": 10690 }, { "grad_norm": 0.2842705547809601, "learning_rate": 4.835509438567142e-05, "loss": 0.025, "step": 10700 }, { "grad_norm": 0.26373982429504395, "learning_rate": 4.827246793127639e-05, "loss": 0.0256, "step": 10710 }, { "grad_norm": 0.23659338057041168, "learning_rate": 4.818984619988807e-05, "loss": 0.0255, "step": 10720 }, { "grad_norm": 0.2719283998012543, "learning_rate": 4.810722941739115e-05, "loss": 0.0261, "step": 10730 }, { "grad_norm": 0.27317193150520325, "learning_rate": 4.8024617809656684e-05, "loss": 0.0244, "step": 10740 }, { "grad_norm": 0.24555061757564545, "learning_rate": 4.794201160254171e-05, "loss": 0.0256, "step": 10750 }, { "grad_norm": 0.30347776412963867, "learning_rate": 4.785941102188844e-05, "loss": 0.0275, "step": 10760 }, { "grad_norm": 0.2649945318698883, "learning_rate": 4.7776816293523686e-05, "loss": 0.0271, "step": 10770 }, { "grad_norm": 0.2999151349067688, "learning_rate": 4.769422764325832e-05, "loss": 0.0261, "step": 10780 }, { "grad_norm": 0.2858549952507019, "learning_rate": 4.76116452968865e-05, "loss": 0.0264, "step": 10790 }, { "grad_norm": 0.24151720106601715, "learning_rate": 4.752906948018525e-05, "loss": 0.0268, "step": 10800 }, { "grad_norm": 0.3064921796321869, "learning_rate": 4.7446500418913684e-05, "loss": 0.027, "step": 10810 }, { "grad_norm": 0.36086323857307434, "learning_rate": 4.736393833881247e-05, "loss": 0.0286, "step": 10820 }, { "grad_norm": 0.2800183892250061, "learning_rate": 4.7281383465603194e-05, "loss": 0.0264, "step": 10830 }, { "grad_norm": 0.2452600747346878, "learning_rate": 4.71988360249877e-05, "loss": 0.0271, "step": 10840 }, { "grad_norm": 0.31657466292381287, "learning_rate": 4.7116296242647554e-05, "loss": 0.0269, "step": 10850 }, { "grad_norm": 0.2230883687734604, "learning_rate": 4.703376434424336e-05, "loss": 0.0256, "step": 10860 }, { "grad_norm": 0.25808900594711304, "learning_rate": 4.695124055541421e-05, "loss": 0.0247, "step": 10870 }, { "grad_norm": 0.24242731928825378, "learning_rate": 4.6868725101776934e-05, "loss": 0.0257, "step": 10880 }, { "grad_norm": 0.23183222115039825, "learning_rate": 4.678621820892567e-05, "loss": 0.027, "step": 10890 }, { "grad_norm": 0.2436673939228058, "learning_rate": 4.670372010243111e-05, "loss": 0.0266, "step": 10900 }, { "grad_norm": 0.28864219784736633, "learning_rate": 4.662123100783992e-05, "loss": 0.0275, "step": 10910 }, { "grad_norm": 0.2710740268230438, "learning_rate": 4.653875115067415e-05, "loss": 0.0254, "step": 10920 }, { "grad_norm": 0.2811516225337982, "learning_rate": 4.6456280756430545e-05, "loss": 0.0272, "step": 10930 }, { "grad_norm": 0.4286440312862396, "learning_rate": 4.637382005058004e-05, "loss": 0.0271, "step": 10940 }, { "grad_norm": 0.3003618121147156, "learning_rate": 4.629136925856705e-05, "loss": 0.0283, "step": 10950 }, { "grad_norm": 0.2558567225933075, "learning_rate": 4.6208928605808895e-05, "loss": 0.0263, "step": 10960 }, { "grad_norm": 0.29975736141204834, "learning_rate": 4.612649831769519e-05, "loss": 0.0268, "step": 10970 }, { "grad_norm": 0.3471018075942993, "learning_rate": 4.604407861958715e-05, "loss": 0.0282, "step": 10980 }, { "grad_norm": 0.2733312249183655, "learning_rate": 4.5961669736817114e-05, "loss": 0.0259, "step": 10990 }, { "grad_norm": 0.263622909784317, "learning_rate": 4.5879271894687814e-05, "loss": 0.0268, "step": 11000 }, { "grad_norm": 0.2831194996833801, "learning_rate": 4.5796885318471826e-05, "loss": 0.024, "step": 11010 }, { "grad_norm": 0.27422842383384705, "learning_rate": 4.571451023341086e-05, "loss": 0.0263, "step": 11020 }, { "grad_norm": 0.24844121932983398, "learning_rate": 4.563214686471527e-05, "loss": 0.0274, "step": 11030 }, { "grad_norm": 0.2774341106414795, "learning_rate": 4.5549795437563365e-05, "loss": 0.0284, "step": 11040 }, { "grad_norm": 0.2616455554962158, "learning_rate": 4.546745617710081e-05, "loss": 0.0252, "step": 11050 }, { "grad_norm": 0.28466013073921204, "learning_rate": 4.5385129308440014e-05, "loss": 0.028, "step": 11060 }, { "grad_norm": 0.33024710416793823, "learning_rate": 4.530281505665944e-05, "loss": 0.0274, "step": 11070 }, { "grad_norm": 0.2456883192062378, "learning_rate": 4.5220513646803134e-05, "loss": 0.0261, "step": 11080 }, { "grad_norm": 0.21814896166324615, "learning_rate": 4.513822530388003e-05, "loss": 0.0253, "step": 11090 }, { "grad_norm": 0.2041274756193161, "learning_rate": 4.5055950252863296e-05, "loss": 0.024, "step": 11100 }, { "grad_norm": 0.2642790675163269, "learning_rate": 4.4973688718689803e-05, "loss": 0.025, "step": 11110 }, { "grad_norm": 0.2435254454612732, "learning_rate": 4.4891440926259406e-05, "loss": 0.0246, "step": 11120 }, { "grad_norm": 0.26154738664627075, "learning_rate": 4.480920710043443e-05, "loss": 0.0245, "step": 11130 }, { "grad_norm": 0.24319909512996674, "learning_rate": 4.4726987466039044e-05, "loss": 0.0266, "step": 11140 }, { "grad_norm": 0.2646913528442383, "learning_rate": 4.46447822478586e-05, "loss": 0.0248, "step": 11150 }, { "grad_norm": 0.23865778744220734, "learning_rate": 4.4562591670638974e-05, "loss": 0.0246, "step": 11160 }, { "grad_norm": 0.2762450575828552, "learning_rate": 4.4480415959086105e-05, "loss": 0.0261, "step": 11170 }, { "grad_norm": 0.2475234866142273, "learning_rate": 4.439825533786522e-05, "loss": 0.0274, "step": 11180 }, { "grad_norm": 0.27529993653297424, "learning_rate": 4.431611003160035e-05, "loss": 0.0241, "step": 11190 }, { "grad_norm": 0.24605320394039154, "learning_rate": 4.4233980264873636e-05, "loss": 0.0243, "step": 11200 }, { "grad_norm": 0.2894233167171478, "learning_rate": 4.4151866262224684e-05, "loss": 0.025, "step": 11210 }, { "grad_norm": 0.2499406486749649, "learning_rate": 4.406976824815006e-05, "loss": 0.0253, "step": 11220 }, { "grad_norm": 0.20934346318244934, "learning_rate": 4.3987686447102595e-05, "loss": 0.0271, "step": 11230 }, { "grad_norm": 0.22174270451068878, "learning_rate": 4.3905621083490804e-05, "loss": 0.0249, "step": 11240 }, { "grad_norm": 0.24905836582183838, "learning_rate": 4.3823572381678286e-05, "loss": 0.0234, "step": 11250 }, { "grad_norm": 0.2458183616399765, "learning_rate": 4.374154056598301e-05, "loss": 0.0266, "step": 11260 }, { "grad_norm": 0.2258196324110031, "learning_rate": 4.3659525860676845e-05, "loss": 0.0245, "step": 11270 }, { "grad_norm": 0.24206264317035675, "learning_rate": 4.3577528489984854e-05, "loss": 0.026, "step": 11280 }, { "grad_norm": 0.3505096137523651, "learning_rate": 4.349554867808476e-05, "loss": 0.0254, "step": 11290 }, { "grad_norm": 0.2943549156188965, "learning_rate": 4.34135866491062e-05, "loss": 0.0251, "step": 11300 }, { "grad_norm": 0.22776412963867188, "learning_rate": 4.333164262713022e-05, "loss": 0.0239, "step": 11310 }, { "grad_norm": 0.3522234857082367, "learning_rate": 4.324971683618868e-05, "loss": 0.0261, "step": 11320 }, { "grad_norm": 0.25832250714302063, "learning_rate": 4.316780950026354e-05, "loss": 0.0265, "step": 11330 }, { "grad_norm": 0.23458081483840942, "learning_rate": 4.308592084328637e-05, "loss": 0.0248, "step": 11340 }, { "grad_norm": 0.3063584566116333, "learning_rate": 4.3004051089137576e-05, "loss": 0.0241, "step": 11350 }, { "grad_norm": 0.2466963529586792, "learning_rate": 4.292220046164597e-05, "loss": 0.024, "step": 11360 }, { "grad_norm": 0.23143546283245087, "learning_rate": 4.2840369184588035e-05, "loss": 0.0269, "step": 11370 }, { "grad_norm": 0.2598622143268585, "learning_rate": 4.2758557481687345e-05, "loss": 0.0258, "step": 11380 }, { "grad_norm": 0.25042787194252014, "learning_rate": 4.267676557661403e-05, "loss": 0.0253, "step": 11390 }, { "grad_norm": 0.27981042861938477, "learning_rate": 4.2594993692983955e-05, "loss": 0.0258, "step": 11400 }, { "grad_norm": 0.23158715665340424, "learning_rate": 4.251324205435837e-05, "loss": 0.0258, "step": 11410 }, { "grad_norm": 0.20966002345085144, "learning_rate": 4.243151088424312e-05, "loss": 0.025, "step": 11420 }, { "grad_norm": 0.22721266746520996, "learning_rate": 4.234980040608813e-05, "loss": 0.024, "step": 11430 }, { "grad_norm": 0.2267540693283081, "learning_rate": 4.22681108432867e-05, "loss": 0.0241, "step": 11440 }, { "grad_norm": 0.26713839173316956, "learning_rate": 4.2186442419174984e-05, "loss": 0.024, "step": 11450 }, { "grad_norm": 0.23806089162826538, "learning_rate": 4.210479535703133e-05, "loss": 0.0244, "step": 11460 }, { "grad_norm": 0.24438728392124176, "learning_rate": 4.202316988007567e-05, "loss": 0.0237, "step": 11470 }, { "grad_norm": 0.28668519854545593, "learning_rate": 4.194156621146901e-05, "loss": 0.0252, "step": 11480 }, { "grad_norm": 0.2527896761894226, "learning_rate": 4.1859984574312596e-05, "loss": 0.0239, "step": 11490 }, { "grad_norm": 0.25434234738349915, "learning_rate": 4.177842519164752e-05, "loss": 0.0259, "step": 11500 }, { "grad_norm": 0.24488703906536102, "learning_rate": 4.169688828645404e-05, "loss": 0.0254, "step": 11510 }, { "grad_norm": 0.2654699385166168, "learning_rate": 4.161537408165092e-05, "loss": 0.0239, "step": 11520 }, { "grad_norm": 0.21970440447330475, "learning_rate": 4.1533882800094924e-05, "loss": 0.0235, "step": 11530 }, { "grad_norm": 0.2611706852912903, "learning_rate": 4.145241466458005e-05, "loss": 0.0234, "step": 11540 }, { "grad_norm": 0.2876072824001312, "learning_rate": 4.13709698978371e-05, "loss": 0.024, "step": 11550 }, { "grad_norm": 0.27616211771965027, "learning_rate": 4.1289548722532944e-05, "loss": 0.0244, "step": 11560 }, { "grad_norm": 0.2684192657470703, "learning_rate": 4.120815136126999e-05, "loss": 0.0231, "step": 11570 }, { "grad_norm": 0.26480668783187866, "learning_rate": 4.112677803658548e-05, "loss": 0.0239, "step": 11580 }, { "grad_norm": 0.20038700103759766, "learning_rate": 4.1045428970951e-05, "loss": 0.0239, "step": 11590 }, { "grad_norm": 0.2235405594110489, "learning_rate": 4.0964104386771785e-05, "loss": 0.0254, "step": 11600 }, { "grad_norm": 0.22211749851703644, "learning_rate": 4.0882804506386144e-05, "loss": 0.0246, "step": 11610 }, { "grad_norm": 0.254679411649704, "learning_rate": 4.080152955206485e-05, "loss": 0.0279, "step": 11620 }, { "grad_norm": 0.2376973181962967, "learning_rate": 4.0720279746010505e-05, "loss": 0.0225, "step": 11630 }, { "grad_norm": 0.2576953172683716, "learning_rate": 4.063905531035699e-05, "loss": 0.0252, "step": 11640 }, { "grad_norm": 0.2636263072490692, "learning_rate": 4.055785646716882e-05, "loss": 0.0238, "step": 11650 }, { "grad_norm": 0.1774149388074875, "learning_rate": 4.047668343844051e-05, "loss": 0.0237, "step": 11660 }, { "grad_norm": 0.25247856974601746, "learning_rate": 4.039553644609604e-05, "loss": 0.0253, "step": 11670 }, { "grad_norm": 0.28483960032463074, "learning_rate": 4.0314415711988176e-05, "loss": 0.0236, "step": 11680 }, { "grad_norm": 0.21961231529712677, "learning_rate": 4.023332145789792e-05, "loss": 0.0263, "step": 11690 }, { "grad_norm": 0.2707476317882538, "learning_rate": 4.015225390553385e-05, "loss": 0.0246, "step": 11700 }, { "grad_norm": 0.23409664630889893, "learning_rate": 4.007121327653158e-05, "loss": 0.0247, "step": 11710 }, { "grad_norm": 0.2347470372915268, "learning_rate": 3.9990199792453064e-05, "loss": 0.0243, "step": 11720 }, { "grad_norm": 0.2636200785636902, "learning_rate": 3.9909213674786103e-05, "loss": 0.0256, "step": 11730 }, { "grad_norm": 0.28403228521347046, "learning_rate": 3.982825514494363e-05, "loss": 0.0243, "step": 11740 }, { "grad_norm": 0.23869837820529938, "learning_rate": 3.974732442426319e-05, "loss": 0.0249, "step": 11750 }, { "grad_norm": 0.2306026667356491, "learning_rate": 3.966642173400629e-05, "loss": 0.0236, "step": 11760 }, { "grad_norm": 0.2752315402030945, "learning_rate": 3.9585547295357764e-05, "loss": 0.0236, "step": 11770 }, { "grad_norm": 0.27126824855804443, "learning_rate": 3.950470132942526e-05, "loss": 0.024, "step": 11780 }, { "grad_norm": 0.2925502061843872, "learning_rate": 3.942388405723856e-05, "loss": 0.0247, "step": 11790 }, { "grad_norm": 0.24411644041538239, "learning_rate": 3.9343095699749e-05, "loss": 0.0253, "step": 11800 }, { "grad_norm": 0.25158149003982544, "learning_rate": 3.9262336477828874e-05, "loss": 0.0243, "step": 11810 }, { "grad_norm": 0.20255249738693237, "learning_rate": 3.9181606612270794e-05, "loss": 0.0238, "step": 11820 }, { "grad_norm": 0.2747310996055603, "learning_rate": 3.910090632378713e-05, "loss": 0.0237, "step": 11830 }, { "grad_norm": 0.2253623902797699, "learning_rate": 3.90202358330094e-05, "loss": 0.0244, "step": 11840 }, { "grad_norm": 0.23513925075531006, "learning_rate": 3.8939595360487656e-05, "loss": 0.0252, "step": 11850 }, { "grad_norm": 0.23476095497608185, "learning_rate": 3.885898512668984e-05, "loss": 0.0238, "step": 11860 }, { "grad_norm": 0.21912851929664612, "learning_rate": 3.877840535200127e-05, "loss": 0.0237, "step": 11870 }, { "grad_norm": 0.2150072306394577, "learning_rate": 3.869785625672397e-05, "loss": 0.0257, "step": 11880 }, { "grad_norm": 0.2747061550617218, "learning_rate": 3.8617338061076094e-05, "loss": 0.0253, "step": 11890 }, { "grad_norm": 0.19871577620506287, "learning_rate": 3.853685098519132e-05, "loss": 0.0246, "step": 11900 }, { "grad_norm": 0.25453925132751465, "learning_rate": 3.845639524911823e-05, "loss": 0.0252, "step": 11910 }, { "grad_norm": 0.27074214816093445, "learning_rate": 3.837597107281974e-05, "loss": 0.0235, "step": 11920 }, { "grad_norm": 0.24596640467643738, "learning_rate": 3.829557867617247e-05, "loss": 0.0235, "step": 11930 }, { "grad_norm": 0.2652173936367035, "learning_rate": 3.821521827896618e-05, "loss": 0.0243, "step": 11940 }, { "grad_norm": 0.23743368685245514, "learning_rate": 3.81348901009031e-05, "loss": 0.0241, "step": 11950 }, { "grad_norm": 0.19287081062793732, "learning_rate": 3.805459436159741e-05, "loss": 0.0234, "step": 11960 }, { "grad_norm": 0.23280374705791473, "learning_rate": 3.797433128057461e-05, "loss": 0.0235, "step": 11970 }, { "grad_norm": 0.30371859669685364, "learning_rate": 3.789410107727089e-05, "loss": 0.0255, "step": 11980 }, { "grad_norm": 0.2843475937843323, "learning_rate": 3.781390397103257e-05, "loss": 0.0246, "step": 11990 }, { "grad_norm": 0.2325042337179184, "learning_rate": 3.7733740181115455e-05, "loss": 0.0237, "step": 12000 }, { "grad_norm": 0.2393411546945572, "learning_rate": 3.7653609926684306e-05, "loss": 0.023, "step": 12010 }, { "grad_norm": 0.242447629570961, "learning_rate": 3.757351342681217e-05, "loss": 0.0228, "step": 12020 }, { "grad_norm": 0.21674779057502747, "learning_rate": 3.749345090047982e-05, "loss": 0.024, "step": 12030 }, { "grad_norm": 0.22891637682914734, "learning_rate": 3.741342256657515e-05, "loss": 0.0231, "step": 12040 }, { "grad_norm": 0.19386912882328033, "learning_rate": 3.7333428643892567e-05, "loss": 0.0238, "step": 12050 }, { "grad_norm": 0.2304595708847046, "learning_rate": 3.725346935113239e-05, "loss": 0.0241, "step": 12060 }, { "grad_norm": 0.24560977518558502, "learning_rate": 3.717354490690029e-05, "loss": 0.0247, "step": 12070 }, { "grad_norm": 0.25254616141319275, "learning_rate": 3.709365552970664e-05, "loss": 0.0247, "step": 12080 }, { "grad_norm": 0.2197575569152832, "learning_rate": 3.7013801437965945e-05, "loss": 0.0224, "step": 12090 }, { "grad_norm": 0.2088940441608429, "learning_rate": 3.693398284999623e-05, "loss": 0.0231, "step": 12100 }, { "grad_norm": 0.2185819000005722, "learning_rate": 3.6854199984018484e-05, "loss": 0.0239, "step": 12110 }, { "grad_norm": 0.24083296954631805, "learning_rate": 3.677445305815601e-05, "loss": 0.024, "step": 12120 }, { "grad_norm": 0.2469249814748764, "learning_rate": 3.669474229043387e-05, "loss": 0.0243, "step": 12130 }, { "grad_norm": 0.2418401688337326, "learning_rate": 3.6615067898778235e-05, "loss": 0.0258, "step": 12140 }, { "grad_norm": 0.2880762815475464, "learning_rate": 3.6535430101015866e-05, "loss": 0.0249, "step": 12150 }, { "grad_norm": 0.2143857479095459, "learning_rate": 3.645582911487345e-05, "loss": 0.0244, "step": 12160 }, { "grad_norm": 0.24245190620422363, "learning_rate": 3.637626515797706e-05, "loss": 0.0234, "step": 12170 }, { "grad_norm": 0.21837551891803741, "learning_rate": 3.629673844785152e-05, "loss": 0.0237, "step": 12180 }, { "grad_norm": 0.2658310830593109, "learning_rate": 3.621724920191979e-05, "loss": 0.0241, "step": 12190 }, { "grad_norm": 0.24084746837615967, "learning_rate": 3.6137797637502444e-05, "loss": 0.0264, "step": 12200 }, { "grad_norm": 0.18568439781665802, "learning_rate": 3.6058383971817035e-05, "loss": 0.0236, "step": 12210 }, { "grad_norm": 0.2138763815164566, "learning_rate": 3.59790084219775e-05, "loss": 0.0231, "step": 12220 }, { "grad_norm": 0.19918783009052277, "learning_rate": 3.589967120499353e-05, "loss": 0.0215, "step": 12230 }, { "grad_norm": 0.2026434987783432, "learning_rate": 3.5820372537770075e-05, "loss": 0.0239, "step": 12240 }, { "grad_norm": 0.21180933713912964, "learning_rate": 3.5741112637106655e-05, "loss": 0.0234, "step": 12250 }, { "grad_norm": 0.261705219745636, "learning_rate": 3.5661891719696804e-05, "loss": 0.0251, "step": 12260 }, { "grad_norm": 0.24347835779190063, "learning_rate": 3.5582710002127504e-05, "loss": 0.0234, "step": 12270 }, { "grad_norm": 0.2533838152885437, "learning_rate": 3.550356770087853e-05, "loss": 0.0236, "step": 12280 }, { "grad_norm": 0.24861827492713928, "learning_rate": 3.5424465032321914e-05, "loss": 0.0234, "step": 12290 }, { "grad_norm": 0.24010534584522247, "learning_rate": 3.5345402212721335e-05, "loss": 0.0244, "step": 12300 }, { "grad_norm": 0.21945035457611084, "learning_rate": 3.526637945823152e-05, "loss": 0.0239, "step": 12310 }, { "grad_norm": 0.225992813706398, "learning_rate": 3.518739698489767e-05, "loss": 0.0237, "step": 12320 }, { "grad_norm": 0.20420624315738678, "learning_rate": 3.510845500865485e-05, "loss": 0.0231, "step": 12330 }, { "grad_norm": 0.22775393724441528, "learning_rate": 3.502955374532739e-05, "loss": 0.0239, "step": 12340 }, { "grad_norm": 0.20464055240154266, "learning_rate": 3.495069341062836e-05, "loss": 0.0219, "step": 12350 }, { "grad_norm": 0.1965724676847458, "learning_rate": 3.4871874220158896e-05, "loss": 0.0251, "step": 12360 }, { "grad_norm": 0.20349079370498657, "learning_rate": 3.479309638940762e-05, "loss": 0.0224, "step": 12370 }, { "grad_norm": 0.27577853202819824, "learning_rate": 3.4714360133750146e-05, "loss": 0.0227, "step": 12380 }, { "grad_norm": 0.1947135478258133, "learning_rate": 3.463566566844839e-05, "loss": 0.0239, "step": 12390 }, { "grad_norm": 0.21639291942119598, "learning_rate": 3.4557013208650016e-05, "loss": 0.0242, "step": 12400 }, { "grad_norm": 0.24349351227283478, "learning_rate": 3.4478402969387857e-05, "loss": 0.024, "step": 12410 }, { "grad_norm": 0.1771192103624344, "learning_rate": 3.4399835165579266e-05, "loss": 0.0224, "step": 12420 }, { "grad_norm": 0.21613696217536926, "learning_rate": 3.4321310012025645e-05, "loss": 0.0227, "step": 12430 }, { "grad_norm": 0.1914314478635788, "learning_rate": 3.424282772341176e-05, "loss": 0.0239, "step": 12440 }, { "grad_norm": 0.26054608821868896, "learning_rate": 3.416438851430519e-05, "loss": 0.0238, "step": 12450 }, { "grad_norm": 0.26302576065063477, "learning_rate": 3.408599259915577e-05, "loss": 0.0248, "step": 12460 }, { "grad_norm": 0.2208419293165207, "learning_rate": 3.400764019229487e-05, "loss": 0.0241, "step": 12470 }, { "grad_norm": 0.22556054592132568, "learning_rate": 3.3929331507935035e-05, "loss": 0.0239, "step": 12480 }, { "grad_norm": 0.2576669454574585, "learning_rate": 3.3851066760169196e-05, "loss": 0.0237, "step": 12490 }, { "grad_norm": 0.2228955775499344, "learning_rate": 3.377284616297021e-05, "loss": 0.0226, "step": 12500 }, { "grad_norm": 0.21516799926757812, "learning_rate": 3.3694669930190166e-05, "loss": 0.0225, "step": 12510 }, { "grad_norm": 0.25476589798927307, "learning_rate": 3.36165382755599e-05, "loss": 0.0243, "step": 12520 }, { "grad_norm": 0.24828772246837616, "learning_rate": 3.35384514126884e-05, "loss": 0.0229, "step": 12530 }, { "grad_norm": 0.26145341992378235, "learning_rate": 3.3460409555062154e-05, "loss": 0.0241, "step": 12540 }, { "grad_norm": 0.2099943310022354, "learning_rate": 3.3382412916044645e-05, "loss": 0.0241, "step": 12550 }, { "grad_norm": 0.2080882340669632, "learning_rate": 3.330446170887566e-05, "loss": 0.0228, "step": 12560 }, { "grad_norm": 0.24270547926425934, "learning_rate": 3.3226556146670834e-05, "loss": 0.0228, "step": 12570 }, { "grad_norm": 0.2450834959745407, "learning_rate": 3.314869644242102e-05, "loss": 0.0225, "step": 12580 }, { "grad_norm": 0.20512452721595764, "learning_rate": 3.3070882808991674e-05, "loss": 0.0224, "step": 12590 }, { "grad_norm": 0.24536307156085968, "learning_rate": 3.2993115459122305e-05, "loss": 0.0223, "step": 12600 }, { "grad_norm": 0.18863624334335327, "learning_rate": 3.2915394605425835e-05, "loss": 0.0226, "step": 12610 }, { "grad_norm": 0.26306861639022827, "learning_rate": 3.283772046038816e-05, "loss": 0.0249, "step": 12620 }, { "grad_norm": 0.20362767577171326, "learning_rate": 3.276009323636739e-05, "loss": 0.0229, "step": 12630 }, { "grad_norm": 0.21615828573703766, "learning_rate": 3.268251314559344e-05, "loss": 0.0225, "step": 12640 }, { "grad_norm": 0.23042799532413483, "learning_rate": 3.2604980400167254e-05, "loss": 0.0233, "step": 12650 }, { "grad_norm": 0.24369099736213684, "learning_rate": 3.252749521206042e-05, "loss": 0.0245, "step": 12660 }, { "grad_norm": 0.19415032863616943, "learning_rate": 3.2450057793114494e-05, "loss": 0.0228, "step": 12670 }, { "grad_norm": 0.2636665999889374, "learning_rate": 3.2372668355040435e-05, "loss": 0.0228, "step": 12680 }, { "grad_norm": 0.20225577056407928, "learning_rate": 3.2295327109418005e-05, "loss": 0.0225, "step": 12690 }, { "grad_norm": 0.236003577709198, "learning_rate": 3.221803426769518e-05, "loss": 0.0258, "step": 12700 }, { "grad_norm": 0.1943429857492447, "learning_rate": 3.214079004118768e-05, "loss": 0.0229, "step": 12710 }, { "grad_norm": 0.2631860077381134, "learning_rate": 3.2063594641078234e-05, "loss": 0.0242, "step": 12720 }, { "grad_norm": 0.19716256856918335, "learning_rate": 3.198644827841616e-05, "loss": 0.023, "step": 12730 }, { "grad_norm": 0.20239505171775818, "learning_rate": 3.1909351164116654e-05, "loss": 0.0229, "step": 12740 }, { "grad_norm": 0.17767761647701263, "learning_rate": 3.183230350896026e-05, "loss": 0.0235, "step": 12750 }, { "grad_norm": 0.24143946170806885, "learning_rate": 3.1755305523592337e-05, "loss": 0.022, "step": 12760 }, { "grad_norm": 0.17240728437900543, "learning_rate": 3.167835741852245e-05, "loss": 0.0226, "step": 12770 }, { "grad_norm": 0.2482723593711853, "learning_rate": 3.160145940412378e-05, "loss": 0.0224, "step": 12780 }, { "grad_norm": 0.22401189804077148, "learning_rate": 3.1524611690632545e-05, "loss": 0.0228, "step": 12790 }, { "grad_norm": 0.22106897830963135, "learning_rate": 3.144781448814746e-05, "loss": 0.0244, "step": 12800 }, { "grad_norm": 0.21688224375247955, "learning_rate": 3.1371068006629145e-05, "loss": 0.0242, "step": 12810 }, { "grad_norm": 0.28668397665023804, "learning_rate": 3.129437245589956e-05, "loss": 0.0235, "step": 12820 }, { "grad_norm": 0.20468005537986755, "learning_rate": 3.121772804564143e-05, "loss": 0.0236, "step": 12830 }, { "grad_norm": 0.16789987683296204, "learning_rate": 3.11411349853976e-05, "loss": 0.0216, "step": 12840 }, { "grad_norm": 0.1526648849248886, "learning_rate": 3.10645934845706e-05, "loss": 0.0229, "step": 12850 }, { "grad_norm": 0.21197731792926788, "learning_rate": 3.098810375242196e-05, "loss": 0.0219, "step": 12860 }, { "grad_norm": 0.21264412999153137, "learning_rate": 3.0911665998071704e-05, "loss": 0.0227, "step": 12870 }, { "grad_norm": 0.210529163479805, "learning_rate": 3.083528043049774e-05, "loss": 0.0216, "step": 12880 }, { "grad_norm": 0.24606485664844513, "learning_rate": 3.0758947258535255e-05, "loss": 0.0236, "step": 12890 }, { "grad_norm": 0.17687533795833588, "learning_rate": 3.068266669087625e-05, "loss": 0.0232, "step": 12900 }, { "grad_norm": 0.2415573000907898, "learning_rate": 3.060643893606887e-05, "loss": 0.0257, "step": 12910 }, { "grad_norm": 0.18574582040309906, "learning_rate": 3.053026420251693e-05, "loss": 0.022, "step": 12920 }, { "grad_norm": 0.21233494579792023, "learning_rate": 3.0454142698479183e-05, "loss": 0.0229, "step": 12930 }, { "grad_norm": 0.1969909965991974, "learning_rate": 3.0378074632068954e-05, "loss": 0.0228, "step": 12940 }, { "grad_norm": 0.16348156332969666, "learning_rate": 3.0302060211253408e-05, "loss": 0.0214, "step": 12950 }, { "grad_norm": 0.23678553104400635, "learning_rate": 3.0226099643853073e-05, "loss": 0.0223, "step": 12960 }, { "grad_norm": 0.1820821464061737, "learning_rate": 3.0150193137541283e-05, "loss": 0.0228, "step": 12970 }, { "grad_norm": 0.2132912129163742, "learning_rate": 3.0074340899843467e-05, "loss": 0.0211, "step": 12980 }, { "grad_norm": 0.1702083945274353, "learning_rate": 2.999854313813677e-05, "loss": 0.0231, "step": 12990 }, { "grad_norm": 0.20193415880203247, "learning_rate": 2.9922800059649382e-05, "loss": 0.0228, "step": 13000 }, { "grad_norm": 0.2184297889471054, "learning_rate": 2.9847111871459976e-05, "loss": 0.0216, "step": 13010 }, { "grad_norm": 0.20431411266326904, "learning_rate": 2.977147878049721e-05, "loss": 0.0206, "step": 13020 }, { "grad_norm": 0.20736168324947357, "learning_rate": 2.9695900993539006e-05, "loss": 0.0217, "step": 13030 }, { "grad_norm": 0.1635676473379135, "learning_rate": 2.9620378717212183e-05, "loss": 0.0211, "step": 13040 }, { "grad_norm": 0.1925518661737442, "learning_rate": 2.9544912157991745e-05, "loss": 0.0218, "step": 13050 }, { "grad_norm": 0.21357300877571106, "learning_rate": 2.9469501522200405e-05, "loss": 0.0229, "step": 13060 }, { "grad_norm": 0.22109025716781616, "learning_rate": 2.9394147016007946e-05, "loss": 0.0222, "step": 13070 }, { "grad_norm": 0.17300154268741608, "learning_rate": 2.9318848845430702e-05, "loss": 0.0224, "step": 13080 }, { "grad_norm": 0.255881130695343, "learning_rate": 2.9243607216331013e-05, "loss": 0.0219, "step": 13090 }, { "grad_norm": 0.2567136585712433, "learning_rate": 2.916842233441661e-05, "loss": 0.0235, "step": 13100 }, { "grad_norm": 0.23944227397441864, "learning_rate": 2.90932944052401e-05, "loss": 0.0237, "step": 13110 }, { "grad_norm": 0.2081209123134613, "learning_rate": 2.9018223634198354e-05, "loss": 0.022, "step": 13120 }, { "grad_norm": 0.19258525967597961, "learning_rate": 2.8943210226532025e-05, "loss": 0.0235, "step": 13130 }, { "grad_norm": 0.18348468840122223, "learning_rate": 2.8868254387324857e-05, "loss": 0.0213, "step": 13140 }, { "grad_norm": 0.21855105459690094, "learning_rate": 2.8793356321503306e-05, "loss": 0.0229, "step": 13150 }, { "grad_norm": 0.2553718090057373, "learning_rate": 2.87185162338358e-05, "loss": 0.0227, "step": 13160 }, { "grad_norm": 0.20410236716270447, "learning_rate": 2.8643734328932253e-05, "loss": 0.0248, "step": 13170 }, { "grad_norm": 0.19844947755336761, "learning_rate": 2.856901081124359e-05, "loss": 0.0226, "step": 13180 }, { "grad_norm": 0.21894066035747528, "learning_rate": 2.8494345885061002e-05, "loss": 0.0214, "step": 13190 }, { "grad_norm": 0.28154054284095764, "learning_rate": 2.8419739754515616e-05, "loss": 0.0212, "step": 13200 }, { "grad_norm": 0.23150375485420227, "learning_rate": 2.8345192623577666e-05, "loss": 0.0222, "step": 13210 }, { "grad_norm": 0.2084595412015915, "learning_rate": 2.8270704696056193e-05, "loss": 0.0212, "step": 13220 }, { "grad_norm": 0.1955718845129013, "learning_rate": 2.8196276175598367e-05, "loss": 0.0213, "step": 13230 }, { "grad_norm": 0.1906796097755432, "learning_rate": 2.8121907265688884e-05, "loss": 0.0222, "step": 13240 }, { "grad_norm": 0.1775696724653244, "learning_rate": 2.804759816964957e-05, "loss": 0.0216, "step": 13250 }, { "grad_norm": 0.1779080033302307, "learning_rate": 2.797334909063857e-05, "loss": 0.0229, "step": 13260 }, { "grad_norm": 0.19127808511257172, "learning_rate": 2.7899160231650056e-05, "loss": 0.0224, "step": 13270 }, { "grad_norm": 0.24541473388671875, "learning_rate": 2.7825031795513585e-05, "loss": 0.0219, "step": 13280 }, { "grad_norm": 0.1974967122077942, "learning_rate": 2.775096398489341e-05, "loss": 0.0207, "step": 13290 }, { "grad_norm": 0.23240505158901215, "learning_rate": 2.7676957002288163e-05, "loss": 0.0225, "step": 13300 }, { "grad_norm": 0.20563489198684692, "learning_rate": 2.760301105003003e-05, "loss": 0.0227, "step": 13310 }, { "grad_norm": 0.23215380311012268, "learning_rate": 2.752912633028446e-05, "loss": 0.0227, "step": 13320 }, { "grad_norm": 0.21937181055545807, "learning_rate": 2.7455303045049474e-05, "loss": 0.0244, "step": 13330 }, { "grad_norm": 0.16980494558811188, "learning_rate": 2.7381541396155098e-05, "loss": 0.0227, "step": 13340 }, { "grad_norm": 0.20515039563179016, "learning_rate": 2.730784158526286e-05, "loss": 0.0218, "step": 13350 }, { "grad_norm": 0.17633619904518127, "learning_rate": 2.723420381386521e-05, "loss": 0.0221, "step": 13360 }, { "grad_norm": 0.2556562125682831, "learning_rate": 2.7160628283285018e-05, "loss": 0.0217, "step": 13370 }, { "grad_norm": 0.2931564450263977, "learning_rate": 2.7087115194675007e-05, "loss": 0.0239, "step": 13380 }, { "grad_norm": 0.21385416388511658, "learning_rate": 2.701366474901712e-05, "loss": 0.0218, "step": 13390 }, { "grad_norm": 0.30802226066589355, "learning_rate": 2.6940277147122085e-05, "loss": 0.0221, "step": 13400 }, { "grad_norm": 0.2483556717634201, "learning_rate": 2.686695258962878e-05, "loss": 0.0229, "step": 13410 }, { "grad_norm": 0.21626685559749603, "learning_rate": 2.679369127700375e-05, "loss": 0.022, "step": 13420 }, { "grad_norm": 0.17082610726356506, "learning_rate": 2.672049340954067e-05, "loss": 0.023, "step": 13430 }, { "grad_norm": 0.22597908973693848, "learning_rate": 2.6647359187359676e-05, "loss": 0.0213, "step": 13440 }, { "grad_norm": 0.17067371308803558, "learning_rate": 2.6574288810406946e-05, "loss": 0.0217, "step": 13450 }, { "grad_norm": 0.18510067462921143, "learning_rate": 2.6501282478454083e-05, "loss": 0.0215, "step": 13460 }, { "grad_norm": 0.20813614130020142, "learning_rate": 2.6428340391097618e-05, "loss": 0.0229, "step": 13470 }, { "grad_norm": 0.2504822909832001, "learning_rate": 2.6355462747758485e-05, "loss": 0.023, "step": 13480 }, { "grad_norm": 0.18876208364963531, "learning_rate": 2.6282649747681304e-05, "loss": 0.0217, "step": 13490 }, { "grad_norm": 0.189689502120018, "learning_rate": 2.620990158993406e-05, "loss": 0.0231, "step": 13500 }, { "grad_norm": 0.19888155162334442, "learning_rate": 2.6137218473407477e-05, "loss": 0.022, "step": 13510 }, { "grad_norm": 0.17806392908096313, "learning_rate": 2.606460059681436e-05, "loss": 0.022, "step": 13520 }, { "grad_norm": 0.18764039874076843, "learning_rate": 2.599204815868928e-05, "loss": 0.0232, "step": 13530 }, { "grad_norm": 0.18877901136875153, "learning_rate": 2.5919561357387756e-05, "loss": 0.0221, "step": 13540 }, { "grad_norm": 0.18175508081912994, "learning_rate": 2.5847140391085972e-05, "loss": 0.021, "step": 13550 }, { "grad_norm": 0.19493460655212402, "learning_rate": 2.5774785457780103e-05, "loss": 0.0217, "step": 13560 }, { "grad_norm": 0.18832387030124664, "learning_rate": 2.5702496755285753e-05, "loss": 0.0209, "step": 13570 }, { "grad_norm": 0.16636501252651215, "learning_rate": 2.5630274481237483e-05, "loss": 0.0213, "step": 13580 }, { "grad_norm": 0.17416414618492126, "learning_rate": 2.5558118833088197e-05, "loss": 0.0225, "step": 13590 }, { "grad_norm": 0.1850271224975586, "learning_rate": 2.548603000810872e-05, "loss": 0.0214, "step": 13600 }, { "grad_norm": 0.2147357016801834, "learning_rate": 2.5414008203387152e-05, "loss": 0.0215, "step": 13610 }, { "grad_norm": 0.20776668190956116, "learning_rate": 2.534205361582834e-05, "loss": 0.0232, "step": 13620 }, { "grad_norm": 0.18746121227741241, "learning_rate": 2.527016644215338e-05, "loss": 0.0213, "step": 13630 }, { "grad_norm": 0.22120405733585358, "learning_rate": 2.519834687889905e-05, "loss": 0.0214, "step": 13640 }, { "grad_norm": 0.19154590368270874, "learning_rate": 2.5126595122417295e-05, "loss": 0.0226, "step": 13650 }, { "grad_norm": 0.1750413030385971, "learning_rate": 2.5054911368874713e-05, "loss": 0.0217, "step": 13660 }, { "grad_norm": 0.18256334960460663, "learning_rate": 2.4983295814251916e-05, "loss": 0.0222, "step": 13670 }, { "grad_norm": 0.19815151393413544, "learning_rate": 2.4911748654343105e-05, "loss": 0.0204, "step": 13680 }, { "grad_norm": 0.18328066170215607, "learning_rate": 2.4840270084755463e-05, "loss": 0.023, "step": 13690 }, { "grad_norm": 0.164686918258667, "learning_rate": 2.4768860300908685e-05, "loss": 0.0211, "step": 13700 }, { "grad_norm": 0.17100219428539276, "learning_rate": 2.469751949803443e-05, "loss": 0.022, "step": 13710 }, { "grad_norm": 0.1595728099346161, "learning_rate": 2.4626247871175666e-05, "loss": 0.021, "step": 13720 }, { "grad_norm": 0.1632814109325409, "learning_rate": 2.4555045615186346e-05, "loss": 0.0233, "step": 13730 }, { "grad_norm": 0.22475282847881317, "learning_rate": 2.4483912924730677e-05, "loss": 0.0219, "step": 13740 }, { "grad_norm": 0.19110484421253204, "learning_rate": 2.4412849994282742e-05, "loss": 0.0209, "step": 13750 }, { "grad_norm": 0.22388555109500885, "learning_rate": 2.434185701812592e-05, "loss": 0.0223, "step": 13760 }, { "grad_norm": 0.18791374564170837, "learning_rate": 2.4270934190352218e-05, "loss": 0.0215, "step": 13770 }, { "grad_norm": 0.22562669217586517, "learning_rate": 2.4200081704861998e-05, "loss": 0.0227, "step": 13780 }, { "grad_norm": 0.2013666033744812, "learning_rate": 2.412929975536321e-05, "loss": 0.0216, "step": 13790 }, { "grad_norm": 0.21182924509048462, "learning_rate": 2.4058588535371017e-05, "loss": 0.0231, "step": 13800 }, { "grad_norm": 0.1796230673789978, "learning_rate": 2.3987948238207243e-05, "loss": 0.0203, "step": 13810 }, { "grad_norm": 0.17952768504619598, "learning_rate": 2.3917379056999678e-05, "loss": 0.0215, "step": 13820 }, { "grad_norm": 0.17965689301490784, "learning_rate": 2.3846881184681824e-05, "loss": 0.022, "step": 13830 }, { "grad_norm": 0.19630467891693115, "learning_rate": 2.377645481399214e-05, "loss": 0.0216, "step": 13840 }, { "grad_norm": 0.20185938477516174, "learning_rate": 2.3706100137473667e-05, "loss": 0.0236, "step": 13850 }, { "grad_norm": 0.225504070520401, "learning_rate": 2.3635817347473394e-05, "loss": 0.0205, "step": 13860 }, { "grad_norm": 0.26334401965141296, "learning_rate": 2.3565606636141757e-05, "loss": 0.0214, "step": 13870 }, { "grad_norm": 0.20825116336345673, "learning_rate": 2.3495468195432203e-05, "loss": 0.0218, "step": 13880 }, { "grad_norm": 0.26213446259498596, "learning_rate": 2.3425402217100507e-05, "loss": 0.0218, "step": 13890 }, { "grad_norm": 0.22175198793411255, "learning_rate": 2.3355408892704424e-05, "loss": 0.0235, "step": 13900 }, { "grad_norm": 0.22209376096725464, "learning_rate": 2.3285488413603003e-05, "loss": 0.0245, "step": 13910 }, { "grad_norm": 0.19226107001304626, "learning_rate": 2.321564097095615e-05, "loss": 0.0224, "step": 13920 }, { "grad_norm": 0.19645434617996216, "learning_rate": 2.3145866755724142e-05, "loss": 0.0216, "step": 13930 }, { "grad_norm": 0.22097283601760864, "learning_rate": 2.307616595866699e-05, "loss": 0.0222, "step": 13940 }, { "grad_norm": 0.16634298861026764, "learning_rate": 2.3006538770344032e-05, "loss": 0.0204, "step": 13950 }, { "grad_norm": 0.17928001284599304, "learning_rate": 2.293698538111334e-05, "loss": 0.021, "step": 13960 }, { "grad_norm": 0.20549143850803375, "learning_rate": 2.28675059811312e-05, "loss": 0.0224, "step": 13970 }, { "grad_norm": 0.19577045738697052, "learning_rate": 2.279810076035167e-05, "loss": 0.0212, "step": 13980 }, { "grad_norm": 0.20418968796730042, "learning_rate": 2.272876990852596e-05, "loss": 0.022, "step": 13990 }, { "grad_norm": 0.15012609958648682, "learning_rate": 2.265951361520195e-05, "loss": 0.0207, "step": 14000 }, { "grad_norm": 0.21215562522411346, "learning_rate": 2.2590332069723748e-05, "loss": 0.0224, "step": 14010 }, { "grad_norm": 0.1909102201461792, "learning_rate": 2.2521225461231004e-05, "loss": 0.0213, "step": 14020 }, { "grad_norm": 0.1943381279706955, "learning_rate": 2.2452193978658597e-05, "loss": 0.0221, "step": 14030 }, { "grad_norm": 0.20277070999145508, "learning_rate": 2.238323781073594e-05, "loss": 0.0219, "step": 14040 }, { "grad_norm": 0.14334891736507416, "learning_rate": 2.2314357145986552e-05, "loss": 0.0209, "step": 14050 }, { "grad_norm": 0.19120126962661743, "learning_rate": 2.224555217272757e-05, "loss": 0.0225, "step": 14060 }, { "grad_norm": 0.24032415449619293, "learning_rate": 2.2176823079069127e-05, "loss": 0.0221, "step": 14070 }, { "grad_norm": 0.23620891571044922, "learning_rate": 2.210817005291398e-05, "loss": 0.0226, "step": 14080 }, { "grad_norm": 0.1875101774930954, "learning_rate": 2.203959328195686e-05, "loss": 0.0225, "step": 14090 }, { "grad_norm": 0.20050492882728577, "learning_rate": 2.1971092953684026e-05, "loss": 0.0228, "step": 14100 }, { "grad_norm": 0.1531722992658615, "learning_rate": 2.1902669255372788e-05, "loss": 0.0223, "step": 14110 }, { "grad_norm": 0.18600891530513763, "learning_rate": 2.1834322374090897e-05, "loss": 0.0201, "step": 14120 }, { "grad_norm": 0.21166935563087463, "learning_rate": 2.1766052496696153e-05, "loss": 0.0208, "step": 14130 }, { "grad_norm": 0.1817096620798111, "learning_rate": 2.169785980983577e-05, "loss": 0.0214, "step": 14140 }, { "grad_norm": 0.19963404536247253, "learning_rate": 2.162974449994593e-05, "loss": 0.0211, "step": 14150 }, { "grad_norm": 0.1629282534122467, "learning_rate": 2.1561706753251337e-05, "loss": 0.021, "step": 14160 }, { "grad_norm": 0.177133709192276, "learning_rate": 2.1493746755764544e-05, "loss": 0.0217, "step": 14170 }, { "grad_norm": 0.18018849194049835, "learning_rate": 2.1425864693285635e-05, "loss": 0.0202, "step": 14180 }, { "grad_norm": 0.21749944984912872, "learning_rate": 2.1358060751401547e-05, "loss": 0.0208, "step": 14190 }, { "grad_norm": 0.2162315994501114, "learning_rate": 2.129033511548566e-05, "loss": 0.021, "step": 14200 }, { "grad_norm": 0.18218906223773956, "learning_rate": 2.1222687970697315e-05, "loss": 0.0224, "step": 14210 }, { "grad_norm": 0.14657311141490936, "learning_rate": 2.1155119501981173e-05, "loss": 0.0219, "step": 14220 }, { "grad_norm": 0.24183398485183716, "learning_rate": 2.1087629894066895e-05, "loss": 0.02, "step": 14230 }, { "grad_norm": 0.18855106830596924, "learning_rate": 2.1020219331468473e-05, "loss": 0.021, "step": 14240 }, { "grad_norm": 0.17855308949947357, "learning_rate": 2.095288799848379e-05, "loss": 0.0217, "step": 14250 }, { "grad_norm": 0.2082131803035736, "learning_rate": 2.088563607919417e-05, "loss": 0.0213, "step": 14260 }, { "grad_norm": 0.34249991178512573, "learning_rate": 2.0818463757463786e-05, "loss": 0.0211, "step": 14270 }, { "grad_norm": 0.24091827869415283, "learning_rate": 2.0751371216939175e-05, "loss": 0.0212, "step": 14280 }, { "grad_norm": 0.22116713225841522, "learning_rate": 2.068435864104882e-05, "loss": 0.0215, "step": 14290 }, { "grad_norm": 0.236334890127182, "learning_rate": 2.0617426213002506e-05, "loss": 0.0221, "step": 14300 }, { "grad_norm": 0.3374742269515991, "learning_rate": 2.055057411579097e-05, "loss": 0.0222, "step": 14310 }, { "grad_norm": 0.2693263590335846, "learning_rate": 2.0483802532185286e-05, "loss": 0.0224, "step": 14320 }, { "grad_norm": 0.2818041443824768, "learning_rate": 2.041711164473638e-05, "loss": 0.0208, "step": 14330 }, { "grad_norm": 0.3213637173175812, "learning_rate": 2.0350501635774637e-05, "loss": 0.0219, "step": 14340 }, { "grad_norm": 0.23030894994735718, "learning_rate": 2.0283972687409247e-05, "loss": 0.0213, "step": 14350 }, { "grad_norm": 0.19465239346027374, "learning_rate": 2.021752498152784e-05, "loss": 0.0216, "step": 14360 }, { "grad_norm": 0.15640775859355927, "learning_rate": 2.015115869979589e-05, "loss": 0.0207, "step": 14370 }, { "grad_norm": 0.20456399023532867, "learning_rate": 2.0084874023656265e-05, "loss": 0.0195, "step": 14380 }, { "grad_norm": 0.19709578156471252, "learning_rate": 2.001867113432877e-05, "loss": 0.021, "step": 14390 }, { "grad_norm": 0.20980525016784668, "learning_rate": 1.995255021280954e-05, "loss": 0.0202, "step": 14400 }, { "grad_norm": 0.19574403762817383, "learning_rate": 1.9886511439870688e-05, "loss": 0.0212, "step": 14410 }, { "grad_norm": 0.16698411107063293, "learning_rate": 1.9820554996059675e-05, "loss": 0.0202, "step": 14420 }, { "grad_norm": 0.20160529017448425, "learning_rate": 1.9754681061698893e-05, "loss": 0.0202, "step": 14430 }, { "grad_norm": 0.19934529066085815, "learning_rate": 1.9688889816885185e-05, "loss": 0.0202, "step": 14440 }, { "grad_norm": 0.197042778134346, "learning_rate": 1.962318144148928e-05, "loss": 0.0207, "step": 14450 }, { "grad_norm": 0.16163069009780884, "learning_rate": 1.955755611515539e-05, "loss": 0.0199, "step": 14460 }, { "grad_norm": 0.17621831595897675, "learning_rate": 1.9492014017300642e-05, "loss": 0.0198, "step": 14470 }, { "grad_norm": 0.2289516031742096, "learning_rate": 1.942655532711461e-05, "loss": 0.0213, "step": 14480 }, { "grad_norm": 0.1652049422264099, "learning_rate": 1.9361180223558882e-05, "loss": 0.0198, "step": 14490 }, { "grad_norm": 0.17474450170993805, "learning_rate": 1.929588888536647e-05, "loss": 0.0203, "step": 14500 }, { "grad_norm": 0.17470572888851166, "learning_rate": 1.9230681491041425e-05, "loss": 0.0195, "step": 14510 }, { "grad_norm": 0.18699367344379425, "learning_rate": 1.9165558218858264e-05, "loss": 0.0211, "step": 14520 }, { "grad_norm": 0.1729167252779007, "learning_rate": 1.9100519246861505e-05, "loss": 0.0213, "step": 14530 }, { "grad_norm": 0.26194968819618225, "learning_rate": 1.9035564752865248e-05, "loss": 0.0204, "step": 14540 }, { "grad_norm": 0.24537846446037292, "learning_rate": 1.897069491445258e-05, "loss": 0.0218, "step": 14550 }, { "grad_norm": 0.2189158946275711, "learning_rate": 1.890590990897515e-05, "loss": 0.0221, "step": 14560 }, { "grad_norm": 0.2909720838069916, "learning_rate": 1.884120991355272e-05, "loss": 0.0191, "step": 14570 }, { "grad_norm": 0.21551169455051422, "learning_rate": 1.8776595105072576e-05, "loss": 0.0219, "step": 14580 }, { "grad_norm": 0.21516075730323792, "learning_rate": 1.8712065660189166e-05, "loss": 0.0217, "step": 14590 }, { "grad_norm": 0.24602806568145752, "learning_rate": 1.8647621755323513e-05, "loss": 0.0207, "step": 14600 }, { "grad_norm": 0.25051191449165344, "learning_rate": 1.858326356666278e-05, "loss": 0.0206, "step": 14610 }, { "grad_norm": 0.1863940954208374, "learning_rate": 1.851899127015983e-05, "loss": 0.0212, "step": 14620 }, { "grad_norm": 0.20068129897117615, "learning_rate": 1.8454805041532626e-05, "loss": 0.0209, "step": 14630 }, { "grad_norm": 0.2235233634710312, "learning_rate": 1.8390705056263906e-05, "loss": 0.0213, "step": 14640 }, { "grad_norm": 0.21838922798633575, "learning_rate": 1.832669148960057e-05, "loss": 0.0216, "step": 14650 }, { "grad_norm": 0.197084441781044, "learning_rate": 1.8262764516553233e-05, "loss": 0.0198, "step": 14660 }, { "grad_norm": 0.19519749283790588, "learning_rate": 1.8198924311895843e-05, "loss": 0.0214, "step": 14670 }, { "grad_norm": 0.19629782438278198, "learning_rate": 1.813517105016505e-05, "loss": 0.02, "step": 14680 }, { "grad_norm": 0.18188932538032532, "learning_rate": 1.8071504905659888e-05, "loss": 0.0203, "step": 14690 }, { "grad_norm": 0.2267570048570633, "learning_rate": 1.800792605244109e-05, "loss": 0.0224, "step": 14700 }, { "grad_norm": 0.20180481672286987, "learning_rate": 1.7944434664330844e-05, "loss": 0.0212, "step": 14710 }, { "grad_norm": 0.22436325252056122, "learning_rate": 1.7881030914912212e-05, "loss": 0.0196, "step": 14720 }, { "grad_norm": 0.19949530065059662, "learning_rate": 1.7817714977528577e-05, "loss": 0.0216, "step": 14730 }, { "grad_norm": 0.2209596335887909, "learning_rate": 1.7754487025283332e-05, "loss": 0.0216, "step": 14740 }, { "grad_norm": 0.20057496428489685, "learning_rate": 1.7691347231039275e-05, "loss": 0.0192, "step": 14750 }, { "grad_norm": 0.2236972153186798, "learning_rate": 1.7628295767418164e-05, "loss": 0.0207, "step": 14760 }, { "grad_norm": 0.19766856729984283, "learning_rate": 1.7565332806800333e-05, "loss": 0.0226, "step": 14770 }, { "grad_norm": 0.22181640565395355, "learning_rate": 1.750245852132408e-05, "loss": 0.0189, "step": 14780 }, { "grad_norm": 0.21095508337020874, "learning_rate": 1.7439673082885323e-05, "loss": 0.02, "step": 14790 }, { "grad_norm": 0.22101515531539917, "learning_rate": 1.7376976663137047e-05, "loss": 0.022, "step": 14800 }, { "grad_norm": 0.22741739451885223, "learning_rate": 1.7314369433488853e-05, "loss": 0.0195, "step": 14810 }, { "grad_norm": 0.2214757651090622, "learning_rate": 1.7251851565106548e-05, "loss": 0.02, "step": 14820 }, { "grad_norm": 0.1894490122795105, "learning_rate": 1.7189423228911574e-05, "loss": 0.0202, "step": 14830 }, { "grad_norm": 0.2075107842683792, "learning_rate": 1.7127084595580606e-05, "loss": 0.0208, "step": 14840 }, { "grad_norm": 0.20117327570915222, "learning_rate": 1.706483583554513e-05, "loss": 0.0195, "step": 14850 }, { "grad_norm": 0.19263648986816406, "learning_rate": 1.700267711899083e-05, "loss": 0.0217, "step": 14860 }, { "grad_norm": 0.16756251454353333, "learning_rate": 1.69406086158573e-05, "loss": 0.0207, "step": 14870 }, { "grad_norm": 0.18142400681972504, "learning_rate": 1.6878630495837455e-05, "loss": 0.0216, "step": 14880 }, { "grad_norm": 0.2275918424129486, "learning_rate": 1.681674292837707e-05, "loss": 0.0221, "step": 14890 }, { "grad_norm": 0.23513822257518768, "learning_rate": 1.6754946082674444e-05, "loss": 0.0218, "step": 14900 }, { "grad_norm": 0.21210579574108124, "learning_rate": 1.6693240127679748e-05, "loss": 0.0212, "step": 14910 }, { "grad_norm": 0.19317179918289185, "learning_rate": 1.663162523209475e-05, "loss": 0.0195, "step": 14920 }, { "grad_norm": 0.1689218282699585, "learning_rate": 1.6570101564372193e-05, "loss": 0.0193, "step": 14930 }, { "grad_norm": 0.1754661500453949, "learning_rate": 1.650866929271543e-05, "loss": 0.0193, "step": 14940 }, { "grad_norm": 0.20064830780029297, "learning_rate": 1.644732858507797e-05, "loss": 0.0202, "step": 14950 }, { "grad_norm": 0.1914495825767517, "learning_rate": 1.6386079609162943e-05, "loss": 0.0214, "step": 14960 }, { "grad_norm": 0.15126840770244598, "learning_rate": 1.6324922532422742e-05, "loss": 0.0199, "step": 14970 }, { "grad_norm": 0.15962336957454681, "learning_rate": 1.6263857522058434e-05, "loss": 0.021, "step": 14980 }, { "grad_norm": 0.17981533706188202, "learning_rate": 1.6202884745019443e-05, "loss": 0.0203, "step": 14990 }, { "grad_norm": 0.2141706496477127, "learning_rate": 1.614200436800304e-05, "loss": 0.0201, "step": 15000 }, { "grad_norm": 0.16968941688537598, "learning_rate": 1.6081216557453814e-05, "loss": 0.0203, "step": 15010 }, { "grad_norm": 0.20406045019626617, "learning_rate": 1.6020521479563367e-05, "loss": 0.0203, "step": 15020 }, { "grad_norm": 0.1813531219959259, "learning_rate": 1.5959919300269654e-05, "loss": 0.02, "step": 15030 }, { "grad_norm": 0.1753879338502884, "learning_rate": 1.5899410185256764e-05, "loss": 0.0203, "step": 15040 }, { "grad_norm": 0.20152154564857483, "learning_rate": 1.583899429995431e-05, "loss": 0.0199, "step": 15050 }, { "grad_norm": 0.22421516478061676, "learning_rate": 1.5778671809536993e-05, "loss": 0.0213, "step": 15060 }, { "grad_norm": 0.1662084013223648, "learning_rate": 1.5718442878924246e-05, "loss": 0.0214, "step": 15070 }, { "grad_norm": 0.14828598499298096, "learning_rate": 1.5658307672779593e-05, "loss": 0.0196, "step": 15080 }, { "grad_norm": 0.1767648309469223, "learning_rate": 1.5598266355510427e-05, "loss": 0.0207, "step": 15090 }, { "grad_norm": 0.1629311591386795, "learning_rate": 1.553831909126744e-05, "loss": 0.0203, "step": 15100 }, { "grad_norm": 0.1344102919101715, "learning_rate": 1.5478466043944135e-05, "loss": 0.0209, "step": 15110 }, { "grad_norm": 0.1583583503961563, "learning_rate": 1.5418707377176468e-05, "loss": 0.0202, "step": 15120 }, { "grad_norm": 0.1958785206079483, "learning_rate": 1.535904325434233e-05, "loss": 0.0223, "step": 15130 }, { "grad_norm": 0.16113868355751038, "learning_rate": 1.529947383856118e-05, "loss": 0.0195, "step": 15140 }, { "grad_norm": 0.18747664988040924, "learning_rate": 1.5239999292693524e-05, "loss": 0.0209, "step": 15150 }, { "grad_norm": 0.1634334921836853, "learning_rate": 1.5180619779340505e-05, "loss": 0.0211, "step": 15160 }, { "grad_norm": 0.20533563196659088, "learning_rate": 1.5121335460843428e-05, "loss": 0.0202, "step": 15170 }, { "grad_norm": 0.1839279979467392, "learning_rate": 1.5062146499283347e-05, "loss": 0.0196, "step": 15180 }, { "grad_norm": 0.21900957822799683, "learning_rate": 1.5003053056480643e-05, "loss": 0.0222, "step": 15190 }, { "grad_norm": 0.18252944946289062, "learning_rate": 1.4944055293994551e-05, "loss": 0.0203, "step": 15200 }, { "grad_norm": 0.19765298068523407, "learning_rate": 1.4885153373122656e-05, "loss": 0.0204, "step": 15210 }, { "grad_norm": 0.2241915762424469, "learning_rate": 1.482634745490059e-05, "loss": 0.0202, "step": 15220 }, { "grad_norm": 0.1998424082994461, "learning_rate": 1.4767637700101466e-05, "loss": 0.0208, "step": 15230 }, { "grad_norm": 0.22726739943027496, "learning_rate": 1.4709024269235528e-05, "loss": 0.021, "step": 15240 }, { "grad_norm": 0.21139967441558838, "learning_rate": 1.4650507322549684e-05, "loss": 0.0197, "step": 15250 }, { "grad_norm": 0.21264822781085968, "learning_rate": 1.4592087020026972e-05, "loss": 0.0216, "step": 15260 }, { "grad_norm": 0.19440479576587677, "learning_rate": 1.4533763521386318e-05, "loss": 0.0197, "step": 15270 }, { "grad_norm": 0.161437526345253, "learning_rate": 1.44755369860819e-05, "loss": 0.0186, "step": 15280 }, { "grad_norm": 0.1659521609544754, "learning_rate": 1.441740757330287e-05, "loss": 0.0201, "step": 15290 }, { "grad_norm": 0.19472941756248474, "learning_rate": 1.4359375441972844e-05, "loss": 0.0204, "step": 15300 }, { "grad_norm": 0.16127830743789673, "learning_rate": 1.4301440750749395e-05, "loss": 0.0214, "step": 15310 }, { "grad_norm": 0.13027265667915344, "learning_rate": 1.4243603658023808e-05, "loss": 0.0208, "step": 15320 }, { "grad_norm": 0.1744663566350937, "learning_rate": 1.4185864321920444e-05, "loss": 0.0202, "step": 15330 }, { "grad_norm": 0.18297117948532104, "learning_rate": 1.4128222900296485e-05, "loss": 0.0209, "step": 15340 }, { "grad_norm": 0.17192034423351288, "learning_rate": 1.407067955074135e-05, "loss": 0.0199, "step": 15350 }, { "grad_norm": 0.1604640632867813, "learning_rate": 1.4013234430576356e-05, "loss": 0.0192, "step": 15360 }, { "grad_norm": 0.16699129343032837, "learning_rate": 1.3955887696854286e-05, "loss": 0.0205, "step": 15370 }, { "grad_norm": 0.1830444186925888, "learning_rate": 1.38986395063589e-05, "loss": 0.0196, "step": 15380 }, { "grad_norm": 0.12765322625637054, "learning_rate": 1.3841490015604597e-05, "loss": 0.0202, "step": 15390 }, { "grad_norm": 0.15817393362522125, "learning_rate": 1.3784439380835879e-05, "loss": 0.0191, "step": 15400 }, { "grad_norm": 0.1403321623802185, "learning_rate": 1.3727487758026986e-05, "loss": 0.0188, "step": 15410 }, { "grad_norm": 0.15586677193641663, "learning_rate": 1.3670635302881525e-05, "loss": 0.0202, "step": 15420 }, { "grad_norm": 0.12670113146305084, "learning_rate": 1.3613882170831888e-05, "loss": 0.0186, "step": 15430 }, { "grad_norm": 0.18510116636753082, "learning_rate": 1.355722851703901e-05, "loss": 0.021, "step": 15440 }, { "grad_norm": 0.2101944386959076, "learning_rate": 1.3500674496391814e-05, "loss": 0.0195, "step": 15450 }, { "grad_norm": 0.18954844772815704, "learning_rate": 1.3444220263506795e-05, "loss": 0.0224, "step": 15460 }, { "grad_norm": 0.20376065373420715, "learning_rate": 1.3387865972727714e-05, "loss": 0.0198, "step": 15470 }, { "grad_norm": 0.14803212881088257, "learning_rate": 1.3331611778125036e-05, "loss": 0.0185, "step": 15480 }, { "grad_norm": 0.15743452310562134, "learning_rate": 1.3275457833495564e-05, "loss": 0.0202, "step": 15490 }, { "grad_norm": 0.18332582712173462, "learning_rate": 1.3219404292362065e-05, "loss": 0.0181, "step": 15500 }, { "grad_norm": 0.1497395783662796, "learning_rate": 1.3163451307972751e-05, "loss": 0.0192, "step": 15510 }, { "grad_norm": 0.15261942148208618, "learning_rate": 1.3107599033300977e-05, "loss": 0.0207, "step": 15520 }, { "grad_norm": 0.15652009844779968, "learning_rate": 1.305184762104471e-05, "loss": 0.0191, "step": 15530 }, { "grad_norm": 0.1713467240333557, "learning_rate": 1.2996197223626178e-05, "loss": 0.0191, "step": 15540 }, { "grad_norm": 0.20641596615314484, "learning_rate": 1.2940647993191457e-05, "loss": 0.0197, "step": 15550 }, { "grad_norm": 0.18791630864143372, "learning_rate": 1.2885200081610005e-05, "loss": 0.0204, "step": 15560 }, { "grad_norm": 0.16861027479171753, "learning_rate": 1.2829853640474316e-05, "loss": 0.0203, "step": 15570 }, { "grad_norm": 0.13847310841083527, "learning_rate": 1.2774608821099438e-05, "loss": 0.0192, "step": 15580 }, { "grad_norm": 0.16946300864219666, "learning_rate": 1.2719465774522577e-05, "loss": 0.0208, "step": 15590 }, { "grad_norm": 0.18134766817092896, "learning_rate": 1.2664424651502755e-05, "loss": 0.0214, "step": 15600 }, { "grad_norm": 0.18080337345600128, "learning_rate": 1.260948560252026e-05, "loss": 0.0203, "step": 15610 }, { "grad_norm": 0.16824504733085632, "learning_rate": 1.2554648777776396e-05, "loss": 0.0197, "step": 15620 }, { "grad_norm": 0.14809824526309967, "learning_rate": 1.2499914327192919e-05, "loss": 0.0197, "step": 15630 }, { "grad_norm": 0.1447165459394455, "learning_rate": 1.2445282400411722e-05, "loss": 0.0207, "step": 15640 }, { "grad_norm": 0.14481312036514282, "learning_rate": 1.2390753146794437e-05, "loss": 0.0197, "step": 15650 }, { "grad_norm": 0.16595686972141266, "learning_rate": 1.2336326715421925e-05, "loss": 0.0207, "step": 15660 }, { "grad_norm": 0.1779286116361618, "learning_rate": 1.2282003255094005e-05, "loss": 0.0196, "step": 15670 }, { "grad_norm": 0.1857452094554901, "learning_rate": 1.2227782914328928e-05, "loss": 0.0212, "step": 15680 }, { "grad_norm": 0.20521990954875946, "learning_rate": 1.2173665841363018e-05, "loss": 0.022, "step": 15690 }, { "grad_norm": 0.20392106473445892, "learning_rate": 1.211965218415032e-05, "loss": 0.0209, "step": 15700 }, { "grad_norm": 0.18570278584957123, "learning_rate": 1.2065742090362082e-05, "loss": 0.0193, "step": 15710 }, { "grad_norm": 0.17304109036922455, "learning_rate": 1.2011935707386457e-05, "loss": 0.0195, "step": 15720 }, { "grad_norm": 0.17562566697597504, "learning_rate": 1.1958233182328044e-05, "loss": 0.02, "step": 15730 }, { "grad_norm": 0.18849711120128632, "learning_rate": 1.1904634662007474e-05, "loss": 0.0194, "step": 15740 }, { "grad_norm": 0.18120568990707397, "learning_rate": 1.1851140292961088e-05, "loss": 0.0214, "step": 15750 }, { "grad_norm": 0.17639996111392975, "learning_rate": 1.1797750221440424e-05, "loss": 0.0211, "step": 15760 }, { "grad_norm": 0.15852057933807373, "learning_rate": 1.1744464593411897e-05, "loss": 0.019, "step": 15770 }, { "grad_norm": 0.1936522126197815, "learning_rate": 1.1691283554556399e-05, "loss": 0.0198, "step": 15780 }, { "grad_norm": 0.1340925544500351, "learning_rate": 1.1638207250268834e-05, "loss": 0.0212, "step": 15790 }, { "grad_norm": 0.20613309741020203, "learning_rate": 1.158523582565782e-05, "loss": 0.0204, "step": 15800 }, { "grad_norm": 0.16609272360801697, "learning_rate": 1.1532369425545192e-05, "loss": 0.0213, "step": 15810 }, { "grad_norm": 0.15243923664093018, "learning_rate": 1.1479608194465662e-05, "loss": 0.0199, "step": 15820 }, { "grad_norm": 0.16645246744155884, "learning_rate": 1.1426952276666442e-05, "loss": 0.0194, "step": 15830 }, { "grad_norm": 0.15648725628852844, "learning_rate": 1.1374401816106778e-05, "loss": 0.019, "step": 15840 }, { "grad_norm": 0.19573137164115906, "learning_rate": 1.1321956956457646e-05, "loss": 0.0209, "step": 15850 }, { "grad_norm": 0.1505841463804245, "learning_rate": 1.1269617841101277e-05, "loss": 0.0197, "step": 15860 }, { "grad_norm": 0.1494966298341751, "learning_rate": 1.1217384613130804e-05, "loss": 0.0191, "step": 15870 }, { "grad_norm": 0.13882553577423096, "learning_rate": 1.11652574153499e-05, "loss": 0.0191, "step": 15880 }, { "grad_norm": 0.13584761321544647, "learning_rate": 1.1113236390272303e-05, "loss": 0.0196, "step": 15890 }, { "grad_norm": 0.17278319597244263, "learning_rate": 1.106132168012155e-05, "loss": 0.0197, "step": 15900 }, { "grad_norm": 0.16721129417419434, "learning_rate": 1.1009513426830448e-05, "loss": 0.019, "step": 15910 }, { "grad_norm": 0.1583346724510193, "learning_rate": 1.0957811772040777e-05, "loss": 0.021, "step": 15920 }, { "grad_norm": 0.14587350189685822, "learning_rate": 1.0906216857102913e-05, "loss": 0.0211, "step": 15930 }, { "grad_norm": 0.22549010813236237, "learning_rate": 1.0854728823075355e-05, "loss": 0.0216, "step": 15940 }, { "grad_norm": 0.19456681609153748, "learning_rate": 1.0803347810724452e-05, "loss": 0.0202, "step": 15950 }, { "grad_norm": 0.15969687700271606, "learning_rate": 1.0752073960523911e-05, "loss": 0.0199, "step": 15960 }, { "grad_norm": 0.14000806212425232, "learning_rate": 1.070090741265447e-05, "loss": 0.0205, "step": 15970 }, { "grad_norm": 0.16552968323230743, "learning_rate": 1.0649848307003547e-05, "loss": 0.0207, "step": 15980 }, { "grad_norm": 0.16761423647403717, "learning_rate": 1.0598896783164757e-05, "loss": 0.0187, "step": 15990 }, { "grad_norm": 0.18307462334632874, "learning_rate": 1.0548052980437645e-05, "loss": 0.0206, "step": 16000 }, { "grad_norm": 0.15799720585346222, "learning_rate": 1.049731703782722e-05, "loss": 0.0199, "step": 16010 }, { "grad_norm": 0.13839060068130493, "learning_rate": 1.0446689094043587e-05, "loss": 0.0198, "step": 16020 }, { "grad_norm": 0.15229645371437073, "learning_rate": 1.039616928750165e-05, "loss": 0.0199, "step": 16030 }, { "grad_norm": 0.1501142531633377, "learning_rate": 1.0345757756320612e-05, "loss": 0.0205, "step": 16040 }, { "grad_norm": 0.1507476270198822, "learning_rate": 1.0295454638323666e-05, "loss": 0.0195, "step": 16050 }, { "grad_norm": 0.13357940316200256, "learning_rate": 1.0245260071037632e-05, "loss": 0.0188, "step": 16060 }, { "grad_norm": 0.15263362228870392, "learning_rate": 1.0195174191692518e-05, "loss": 0.0179, "step": 16070 }, { "grad_norm": 0.17054200172424316, "learning_rate": 1.014519713722124e-05, "loss": 0.0203, "step": 16080 }, { "grad_norm": 0.16000963747501373, "learning_rate": 1.0095329044259132e-05, "loss": 0.0191, "step": 16090 }, { "grad_norm": 0.17435100674629211, "learning_rate": 1.004557004914365e-05, "loss": 0.0203, "step": 16100 }, { "grad_norm": 0.19281384348869324, "learning_rate": 9.995920287914007e-06, "loss": 0.0197, "step": 16110 }, { "grad_norm": 0.16330058872699738, "learning_rate": 9.946379896310737e-06, "loss": 0.0187, "step": 16120 }, { "grad_norm": 0.1337546855211258, "learning_rate": 9.896949009775396e-06, "loss": 0.0188, "step": 16130 }, { "grad_norm": 0.15508702397346497, "learning_rate": 9.847627763450134e-06, "loss": 0.0196, "step": 16140 }, { "grad_norm": 0.1487896889448166, "learning_rate": 9.798416292177337e-06, "loss": 0.019, "step": 16150 }, { "grad_norm": 0.18217651546001434, "learning_rate": 9.74931473049932e-06, "loss": 0.0184, "step": 16160 }, { "grad_norm": 0.14415104687213898, "learning_rate": 9.700323212657847e-06, "loss": 0.0193, "step": 16170 }, { "grad_norm": 0.15772981941699982, "learning_rate": 9.65144187259388e-06, "loss": 0.0185, "step": 16180 }, { "grad_norm": 0.14109084010124207, "learning_rate": 9.602670843947132e-06, "loss": 0.0185, "step": 16190 }, { "grad_norm": 0.1745998114347458, "learning_rate": 9.554010260055713e-06, "loss": 0.0189, "step": 16200 }, { "grad_norm": 0.17752870917320251, "learning_rate": 9.505460253955834e-06, "loss": 0.0201, "step": 16210 }, { "grad_norm": 0.1673152595758438, "learning_rate": 9.457020958381324e-06, "loss": 0.019, "step": 16220 }, { "grad_norm": 0.1732870489358902, "learning_rate": 9.408692505763395e-06, "loss": 0.0171, "step": 16230 }, { "grad_norm": 0.11677542328834534, "learning_rate": 9.360475028230181e-06, "loss": 0.0183, "step": 16240 }, { "grad_norm": 0.13998177647590637, "learning_rate": 9.312368657606412e-06, "loss": 0.0202, "step": 16250 }, { "grad_norm": 0.14570988714694977, "learning_rate": 9.264373525413096e-06, "loss": 0.0196, "step": 16260 }, { "grad_norm": 0.15782517194747925, "learning_rate": 9.216489762867058e-06, "loss": 0.0194, "step": 16270 }, { "grad_norm": 0.14912141859531403, "learning_rate": 9.168717500880708e-06, "loss": 0.0181, "step": 16280 }, { "grad_norm": 0.14191263914108276, "learning_rate": 9.121056870061574e-06, "loss": 0.0197, "step": 16290 }, { "grad_norm": 0.1731659471988678, "learning_rate": 9.073508000711983e-06, "loss": 0.0201, "step": 16300 }, { "grad_norm": 0.15012440085411072, "learning_rate": 9.026071022828758e-06, "loss": 0.0194, "step": 16310 }, { "grad_norm": 0.21935097873210907, "learning_rate": 8.978746066102771e-06, "loss": 0.0207, "step": 16320 }, { "grad_norm": 0.25112712383270264, "learning_rate": 8.931533259918634e-06, "loss": 0.0214, "step": 16330 }, { "grad_norm": 0.17706634104251862, "learning_rate": 8.884432733354382e-06, "loss": 0.0188, "step": 16340 }, { "grad_norm": 0.2115912139415741, "learning_rate": 8.837444615181029e-06, "loss": 0.0194, "step": 16350 }, { "grad_norm": 0.17479270696640015, "learning_rate": 8.790569033862323e-06, "loss": 0.0192, "step": 16360 }, { "grad_norm": 0.17822231352329254, "learning_rate": 8.7438061175543e-06, "loss": 0.0198, "step": 16370 }, { "grad_norm": 0.1446305811405182, "learning_rate": 8.697155994104978e-06, "loss": 0.019, "step": 16380 }, { "grad_norm": 0.1319979578256607, "learning_rate": 8.650618791054033e-06, "loss": 0.0181, "step": 16390 }, { "grad_norm": 0.14336855709552765, "learning_rate": 8.604194635632373e-06, "loss": 0.0184, "step": 16400 }, { "grad_norm": 0.1425846666097641, "learning_rate": 8.557883654761906e-06, "loss": 0.0193, "step": 16410 }, { "grad_norm": 0.13647519052028656, "learning_rate": 8.511685975055061e-06, "loss": 0.0187, "step": 16420 }, { "grad_norm": 0.14065441489219666, "learning_rate": 8.46560172281452e-06, "loss": 0.0207, "step": 16430 }, { "grad_norm": 0.1366286426782608, "learning_rate": 8.419631024032893e-06, "loss": 0.0196, "step": 16440 }, { "grad_norm": 0.13725334405899048, "learning_rate": 8.373774004392293e-06, "loss": 0.0211, "step": 16450 }, { "grad_norm": 0.12092507630586624, "learning_rate": 8.32803078926409e-06, "loss": 0.0187, "step": 16460 }, { "grad_norm": 0.18251851201057434, "learning_rate": 8.282401503708454e-06, "loss": 0.0189, "step": 16470 }, { "grad_norm": 0.13532008230686188, "learning_rate": 8.23688627247412e-06, "loss": 0.0198, "step": 16480 }, { "grad_norm": 0.13276885449886322, "learning_rate": 8.191485219998007e-06, "loss": 0.0185, "step": 16490 }, { "grad_norm": 0.130792036652565, "learning_rate": 8.146198470404843e-06, "loss": 0.0208, "step": 16500 }, { "grad_norm": 0.12545868754386902, "learning_rate": 8.101026147506897e-06, "loss": 0.0189, "step": 16510 }, { "grad_norm": 0.11476390063762665, "learning_rate": 8.05596837480353e-06, "loss": 0.0198, "step": 16520 }, { "grad_norm": 0.11856121569871902, "learning_rate": 8.011025275480998e-06, "loss": 0.0194, "step": 16530 }, { "grad_norm": 0.1298731416463852, "learning_rate": 7.966196972412027e-06, "loss": 0.0196, "step": 16540 }, { "grad_norm": 0.11277741193771362, "learning_rate": 7.92148358815547e-06, "loss": 0.0182, "step": 16550 }, { "grad_norm": 0.1217537671327591, "learning_rate": 7.87688524495604e-06, "loss": 0.0192, "step": 16560 }, { "grad_norm": 0.1233842745423317, "learning_rate": 7.83240206474386e-06, "loss": 0.0197, "step": 16570 }, { "grad_norm": 0.11316624283790588, "learning_rate": 7.788034169134272e-06, "loss": 0.0194, "step": 16580 }, { "grad_norm": 0.12220564484596252, "learning_rate": 7.743781679427414e-06, "loss": 0.0195, "step": 16590 }, { "grad_norm": 0.1390571892261505, "learning_rate": 7.699644716607895e-06, "loss": 0.0194, "step": 16600 }, { "grad_norm": 0.1419750154018402, "learning_rate": 7.655623401344486e-06, "loss": 0.0182, "step": 16610 }, { "grad_norm": 0.14665065705776215, "learning_rate": 7.611717853989775e-06, "loss": 0.0189, "step": 16620 }, { "grad_norm": 0.12608684599399567, "learning_rate": 7.567928194579854e-06, "loss": 0.0187, "step": 16630 }, { "grad_norm": 0.18072271347045898, "learning_rate": 7.524254542833997e-06, "loss": 0.0205, "step": 16640 }, { "grad_norm": 0.12376997619867325, "learning_rate": 7.480697018154286e-06, "loss": 0.0186, "step": 16650 }, { "grad_norm": 0.1204204112291336, "learning_rate": 7.437255739625332e-06, "loss": 0.0194, "step": 16660 }, { "grad_norm": 0.1356189250946045, "learning_rate": 7.393930826013923e-06, "loss": 0.0185, "step": 16670 }, { "grad_norm": 0.1394031047821045, "learning_rate": 7.350722395768722e-06, "loss": 0.0191, "step": 16680 }, { "grad_norm": 0.14268043637275696, "learning_rate": 7.307630567019963e-06, "loss": 0.0195, "step": 16690 }, { "grad_norm": 0.13907508552074432, "learning_rate": 7.264655457579e-06, "loss": 0.0183, "step": 16700 }, { "grad_norm": 0.1018611416220665, "learning_rate": 7.221797184938184e-06, "loss": 0.0195, "step": 16710 }, { "grad_norm": 0.1230030432343483, "learning_rate": 7.179055866270373e-06, "loss": 0.019, "step": 16720 }, { "grad_norm": 0.12636303901672363, "learning_rate": 7.136431618428707e-06, "loss": 0.0193, "step": 16730 }, { "grad_norm": 0.10615220665931702, "learning_rate": 7.09392455794628e-06, "loss": 0.0196, "step": 16740 }, { "grad_norm": 0.10752319544553757, "learning_rate": 7.051534801035725e-06, "loss": 0.0195, "step": 16750 }, { "grad_norm": 0.11410488933324814, "learning_rate": 7.00926246358905e-06, "loss": 0.0181, "step": 16760 }, { "grad_norm": 0.1200530081987381, "learning_rate": 6.967107661177191e-06, "loss": 0.0182, "step": 16770 }, { "grad_norm": 0.12442640960216522, "learning_rate": 6.925070509049786e-06, "loss": 0.0186, "step": 16780 }, { "grad_norm": 0.11781477183103561, "learning_rate": 6.883151122134812e-06, "loss": 0.0206, "step": 16790 }, { "grad_norm": 0.10902592539787292, "learning_rate": 6.8413496150382394e-06, "loss": 0.0195, "step": 16800 }, { "grad_norm": 0.11057227849960327, "learning_rate": 6.7996661020438165e-06, "loss": 0.0188, "step": 16810 }, { "grad_norm": 0.11770366877317429, "learning_rate": 6.758100697112662e-06, "loss": 0.0185, "step": 16820 }, { "grad_norm": 0.09819748252630234, "learning_rate": 6.716653513883026e-06, "loss": 0.0192, "step": 16830 }, { "grad_norm": 0.10984419286251068, "learning_rate": 6.675324665669913e-06, "loss": 0.0177, "step": 16840 }, { "grad_norm": 0.09591743350028992, "learning_rate": 6.634114265464803e-06, "loss": 0.0179, "step": 16850 }, { "grad_norm": 0.12017423659563065, "learning_rate": 6.59302242593538e-06, "loss": 0.02, "step": 16860 }, { "grad_norm": 0.10011997818946838, "learning_rate": 6.552049259425141e-06, "loss": 0.019, "step": 16870 }, { "grad_norm": 0.10804496705532074, "learning_rate": 6.511194877953181e-06, "loss": 0.0194, "step": 16880 }, { "grad_norm": 0.12753018736839294, "learning_rate": 6.470459393213813e-06, "loss": 0.0191, "step": 16890 }, { "grad_norm": 0.11749093234539032, "learning_rate": 6.429842916576279e-06, "loss": 0.0186, "step": 16900 }, { "grad_norm": 0.10887564718723297, "learning_rate": 6.389345559084503e-06, "loss": 0.0195, "step": 16910 }, { "grad_norm": 0.11989623308181763, "learning_rate": 6.348967431456682e-06, "loss": 0.0187, "step": 16920 }, { "grad_norm": 0.08975622802972794, "learning_rate": 6.30870864408511e-06, "loss": 0.0181, "step": 16930 }, { "grad_norm": 0.10270331799983978, "learning_rate": 6.268569307035754e-06, "loss": 0.0198, "step": 16940 }, { "grad_norm": 0.11460033804178238, "learning_rate": 6.228549530048022e-06, "loss": 0.0191, "step": 16950 }, { "grad_norm": 0.11567949503660202, "learning_rate": 6.1886494225344814e-06, "loss": 0.0189, "step": 16960 }, { "grad_norm": 0.10597874224185944, "learning_rate": 6.148869093580479e-06, "loss": 0.0182, "step": 16970 }, { "grad_norm": 0.16211888194084167, "learning_rate": 6.109208651943921e-06, "loss": 0.0208, "step": 16980 }, { "grad_norm": 0.1231410875916481, "learning_rate": 6.069668206054946e-06, "loss": 0.0206, "step": 16990 }, { "grad_norm": 0.1196521669626236, "learning_rate": 6.0302478640156145e-06, "loss": 0.0186, "step": 17000 }, { "grad_norm": 0.10815589874982834, "learning_rate": 5.990947733599644e-06, "loss": 0.0184, "step": 17010 }, { "grad_norm": 0.14697270095348358, "learning_rate": 5.951767922252105e-06, "loss": 0.0199, "step": 17020 }, { "grad_norm": 0.124472476541996, "learning_rate": 5.912708537089068e-06, "loss": 0.0202, "step": 17030 }, { "grad_norm": 0.11188208311796188, "learning_rate": 5.873769684897434e-06, "loss": 0.0184, "step": 17040 }, { "grad_norm": 0.12566852569580078, "learning_rate": 5.834951472134514e-06, "loss": 0.0212, "step": 17050 }, { "grad_norm": 0.09102421998977661, "learning_rate": 5.796254004927832e-06, "loss": 0.0186, "step": 17060 }, { "grad_norm": 0.10012388229370117, "learning_rate": 5.757677389074806e-06, "loss": 0.0183, "step": 17070 }, { "grad_norm": 0.09278150647878647, "learning_rate": 5.719221730042385e-06, "loss": 0.0199, "step": 17080 }, { "grad_norm": 0.10310041904449463, "learning_rate": 5.680887132966911e-06, "loss": 0.0201, "step": 17090 }, { "grad_norm": 0.0953638032078743, "learning_rate": 5.642673702653683e-06, "loss": 0.0186, "step": 17100 }, { "grad_norm": 0.11209260672330856, "learning_rate": 5.604581543576781e-06, "loss": 0.0204, "step": 17110 }, { "grad_norm": 0.1093745231628418, "learning_rate": 5.566610759878704e-06, "loss": 0.0186, "step": 17120 }, { "grad_norm": 0.107696034014225, "learning_rate": 5.528761455370119e-06, "loss": 0.019, "step": 17130 }, { "grad_norm": 0.09608832746744156, "learning_rate": 5.491033733529594e-06, "loss": 0.0182, "step": 17140 }, { "grad_norm": 0.10013222694396973, "learning_rate": 5.453427697503255e-06, "loss": 0.0186, "step": 17150 }, { "grad_norm": 0.09796315431594849, "learning_rate": 5.415943450104599e-06, "loss": 0.0191, "step": 17160 }, { "grad_norm": 0.11197696626186371, "learning_rate": 5.378581093814111e-06, "loss": 0.0181, "step": 17170 }, { "grad_norm": 0.10383418202400208, "learning_rate": 5.3413407307790375e-06, "loss": 0.0182, "step": 17180 }, { "grad_norm": 0.10765958577394485, "learning_rate": 5.30422246281313e-06, "loss": 0.0191, "step": 17190 }, { "grad_norm": 0.10390859097242355, "learning_rate": 5.267226391396296e-06, "loss": 0.0185, "step": 17200 }, { "grad_norm": 0.11334461718797684, "learning_rate": 5.2303526176744e-06, "loss": 0.0193, "step": 17210 }, { "grad_norm": 0.11591918021440506, "learning_rate": 5.193601242458929e-06, "loss": 0.0203, "step": 17220 }, { "grad_norm": 0.13865213096141815, "learning_rate": 5.156972366226714e-06, "loss": 0.0186, "step": 17230 }, { "grad_norm": 0.11398731917142868, "learning_rate": 5.120466089119735e-06, "loss": 0.0195, "step": 17240 }, { "grad_norm": 0.10961835086345673, "learning_rate": 5.084082510944749e-06, "loss": 0.0201, "step": 17250 }, { "grad_norm": 0.1330420970916748, "learning_rate": 5.047821731173058e-06, "loss": 0.0194, "step": 17260 }, { "grad_norm": 0.11129545420408249, "learning_rate": 5.011683848940274e-06, "loss": 0.0199, "step": 17270 }, { "grad_norm": 0.11073155701160431, "learning_rate": 4.975668963045954e-06, "loss": 0.0183, "step": 17280 }, { "grad_norm": 0.11386819928884506, "learning_rate": 4.9397771719534525e-06, "loss": 0.0213, "step": 17290 }, { "grad_norm": 0.09753624349832535, "learning_rate": 4.904008573789548e-06, "loss": 0.0201, "step": 17300 }, { "grad_norm": 0.10075170546770096, "learning_rate": 4.8683632663442005e-06, "loss": 0.0194, "step": 17310 }, { "grad_norm": 0.09593509882688522, "learning_rate": 4.832841347070343e-06, "loss": 0.0191, "step": 17320 }, { "grad_norm": 0.09085100144147873, "learning_rate": 4.797442913083539e-06, "loss": 0.0179, "step": 17330 }, { "grad_norm": 0.09883932769298553, "learning_rate": 4.7621680611617596e-06, "loss": 0.0188, "step": 17340 }, { "grad_norm": 0.09658411145210266, "learning_rate": 4.727016887745095e-06, "loss": 0.0195, "step": 17350 }, { "grad_norm": 0.12021001428365707, "learning_rate": 4.691989488935511e-06, "loss": 0.0191, "step": 17360 }, { "grad_norm": 0.10778813809156418, "learning_rate": 4.657085960496588e-06, "loss": 0.0189, "step": 17370 }, { "grad_norm": 0.09863677620887756, "learning_rate": 4.6223063978532265e-06, "loss": 0.019, "step": 17380 }, { "grad_norm": 0.12138848006725311, "learning_rate": 4.587650896091439e-06, "loss": 0.0194, "step": 17390 }, { "grad_norm": 0.0982840433716774, "learning_rate": 4.553119549958035e-06, "loss": 0.0187, "step": 17400 }, { "grad_norm": 0.09998755902051926, "learning_rate": 4.518712453860385e-06, "loss": 0.019, "step": 17410 }, { "grad_norm": 0.09160250425338745, "learning_rate": 4.484429701866205e-06, "loss": 0.0175, "step": 17420 }, { "grad_norm": 0.12484665215015411, "learning_rate": 4.4502713877031975e-06, "loss": 0.0195, "step": 17430 }, { "grad_norm": 0.08996372669935226, "learning_rate": 4.416237604758911e-06, "loss": 0.0192, "step": 17440 }, { "grad_norm": 0.10497836768627167, "learning_rate": 4.3823284460804025e-06, "loss": 0.0178, "step": 17450 }, { "grad_norm": 0.09456887096166611, "learning_rate": 4.348544004374011e-06, "loss": 0.0188, "step": 17460 }, { "grad_norm": 0.0914742723107338, "learning_rate": 4.314884372005123e-06, "loss": 0.0193, "step": 17470 }, { "grad_norm": 0.09114857763051987, "learning_rate": 4.281349640997867e-06, "loss": 0.0195, "step": 17480 }, { "grad_norm": 0.09412012249231339, "learning_rate": 4.247939903034942e-06, "loss": 0.0189, "step": 17490 }, { "grad_norm": 0.07358711212873459, "learning_rate": 4.214655249457284e-06, "loss": 0.0192, "step": 17500 }, { "grad_norm": 0.09024911373853683, "learning_rate": 4.181495771263855e-06, "loss": 0.0185, "step": 17510 }, { "grad_norm": 0.10063929110765457, "learning_rate": 4.148461559111427e-06, "loss": 0.0177, "step": 17520 }, { "grad_norm": 0.10230059176683426, "learning_rate": 4.115552703314252e-06, "loss": 0.0179, "step": 17530 }, { "grad_norm": 0.10042743384838104, "learning_rate": 4.082769293843886e-06, "loss": 0.0184, "step": 17540 }, { "grad_norm": 0.08458146452903748, "learning_rate": 4.050111420328939e-06, "loss": 0.0185, "step": 17550 }, { "grad_norm": 0.12238188087940216, "learning_rate": 4.017579172054764e-06, "loss": 0.0212, "step": 17560 }, { "grad_norm": 0.099589504301548, "learning_rate": 3.985172637963308e-06, "loss": 0.0192, "step": 17570 }, { "grad_norm": 0.09939547628164291, "learning_rate": 3.952891906652784e-06, "loss": 0.0192, "step": 17580 }, { "grad_norm": 0.10164329409599304, "learning_rate": 3.920737066377478e-06, "loss": 0.0198, "step": 17590 }, { "grad_norm": 0.10007759183645248, "learning_rate": 3.888708205047509e-06, "loss": 0.0187, "step": 17600 }, { "grad_norm": 0.10535383224487305, "learning_rate": 3.856805410228542e-06, "loss": 0.0197, "step": 17610 }, { "grad_norm": 0.10492052882909775, "learning_rate": 3.82502876914162e-06, "loss": 0.0203, "step": 17620 }, { "grad_norm": 0.0826592817902565, "learning_rate": 3.7933783686628586e-06, "loss": 0.0171, "step": 17630 }, { "grad_norm": 0.08985013514757156, "learning_rate": 3.7618542953232306e-06, "loss": 0.0193, "step": 17640 }, { "grad_norm": 0.08294995129108429, "learning_rate": 3.7304566353083658e-06, "loss": 0.0185, "step": 17650 }, { "grad_norm": 0.08955778181552887, "learning_rate": 3.6991854744582555e-06, "loss": 0.0189, "step": 17660 }, { "grad_norm": 0.09362080693244934, "learning_rate": 3.6680408982670777e-06, "loss": 0.0196, "step": 17670 }, { "grad_norm": 0.07552852481603622, "learning_rate": 3.637022991882899e-06, "loss": 0.0189, "step": 17680 }, { "grad_norm": 0.09246214479207993, "learning_rate": 3.606131840107485e-06, "loss": 0.0186, "step": 17690 }, { "grad_norm": 0.08261837810277939, "learning_rate": 3.575367527396084e-06, "loss": 0.0192, "step": 17700 }, { "grad_norm": 0.09520620852708817, "learning_rate": 3.5447301378571386e-06, "loss": 0.0186, "step": 17710 }, { "grad_norm": 0.11964000016450882, "learning_rate": 3.514219755252113e-06, "loss": 0.0187, "step": 17720 }, { "grad_norm": 0.09163133054971695, "learning_rate": 3.4838364629952213e-06, "loss": 0.0182, "step": 17730 }, { "grad_norm": 0.10012180358171463, "learning_rate": 3.4535803441532123e-06, "loss": 0.0189, "step": 17740 }, { "grad_norm": 0.09245579689741135, "learning_rate": 3.4234514814451836e-06, "loss": 0.0183, "step": 17750 }, { "grad_norm": 0.08221224695444107, "learning_rate": 3.393449957242273e-06, "loss": 0.0178, "step": 17760 }, { "grad_norm": 0.0949811190366745, "learning_rate": 3.363575853567524e-06, "loss": 0.0183, "step": 17770 }, { "grad_norm": 0.10598493367433548, "learning_rate": 3.3338292520955826e-06, "loss": 0.0196, "step": 17780 }, { "grad_norm": 0.09300761669874191, "learning_rate": 3.304210234152516e-06, "loss": 0.0185, "step": 17790 }, { "grad_norm": 0.0809948742389679, "learning_rate": 3.2747188807155993e-06, "loss": 0.0195, "step": 17800 }, { "grad_norm": 0.07418692857027054, "learning_rate": 3.2453552724130643e-06, "loss": 0.0186, "step": 17810 }, { "grad_norm": 0.08977346867322922, "learning_rate": 3.216119489523889e-06, "loss": 0.0185, "step": 17820 }, { "grad_norm": 0.08624174445867538, "learning_rate": 3.1870116119775917e-06, "loss": 0.0188, "step": 17830 }, { "grad_norm": 0.08672924339771271, "learning_rate": 3.158031719353999e-06, "loss": 0.0178, "step": 17840 }, { "grad_norm": 0.08659668266773224, "learning_rate": 3.1291798908830273e-06, "loss": 0.0192, "step": 17850 }, { "grad_norm": 0.08481490612030029, "learning_rate": 3.1004562054444853e-06, "loss": 0.0183, "step": 17860 }, { "grad_norm": 0.13561566174030304, "learning_rate": 3.071860741567806e-06, "loss": 0.0185, "step": 17870 }, { "grad_norm": 0.08047522604465485, "learning_rate": 3.04339357743193e-06, "loss": 0.0183, "step": 17880 }, { "grad_norm": 0.10032394528388977, "learning_rate": 3.0150547908649628e-06, "loss": 0.0181, "step": 17890 }, { "grad_norm": 0.07632776349782944, "learning_rate": 2.9868444593440957e-06, "loss": 0.0177, "step": 17900 }, { "grad_norm": 0.09150762110948563, "learning_rate": 2.9587626599952846e-06, "loss": 0.0183, "step": 17910 }, { "grad_norm": 0.07407736033201218, "learning_rate": 2.930809469593082e-06, "loss": 0.0185, "step": 17920 }, { "grad_norm": 0.09304295480251312, "learning_rate": 2.9029849645604733e-06, "loss": 0.0188, "step": 17930 }, { "grad_norm": 0.1199328601360321, "learning_rate": 2.8752892209685632e-06, "loss": 0.0187, "step": 17940 }, { "grad_norm": 0.0806715339422226, "learning_rate": 2.847722314536483e-06, "loss": 0.0188, "step": 17950 }, { "grad_norm": 0.0846756100654602, "learning_rate": 2.820284320631078e-06, "loss": 0.019, "step": 17960 }, { "grad_norm": 0.08223669230937958, "learning_rate": 2.792975314266788e-06, "loss": 0.0185, "step": 17970 }, { "grad_norm": 0.08667353540658951, "learning_rate": 2.7657953701054007e-06, "loss": 0.0177, "step": 17980 }, { "grad_norm": 0.09585791081190109, "learning_rate": 2.7387445624558306e-06, "loss": 0.0185, "step": 17990 }, { "grad_norm": 0.10648784041404724, "learning_rate": 2.7118229652739747e-06, "loss": 0.0197, "step": 18000 }, { "grad_norm": 0.07420063018798828, "learning_rate": 2.6850306521624236e-06, "loss": 0.0193, "step": 18010 }, { "grad_norm": 0.0903848335146904, "learning_rate": 2.6583676963703507e-06, "loss": 0.0183, "step": 18020 }, { "grad_norm": 0.08648047596216202, "learning_rate": 2.631834170793268e-06, "loss": 0.0188, "step": 18030 }, { "grad_norm": 0.070640429854393, "learning_rate": 2.6054301479728036e-06, "loss": 0.0176, "step": 18040 }, { "grad_norm": 0.08904149383306503, "learning_rate": 2.579155700096575e-06, "loss": 0.0181, "step": 18050 }, { "grad_norm": 0.08976573497056961, "learning_rate": 2.5530108989978873e-06, "loss": 0.0184, "step": 18060 }, { "grad_norm": 0.07876861840486526, "learning_rate": 2.5269958161556416e-06, "loss": 0.0195, "step": 18070 }, { "grad_norm": 0.08020199835300446, "learning_rate": 2.5011105226940888e-06, "loss": 0.0193, "step": 18080 }, { "grad_norm": 0.0750158354640007, "learning_rate": 2.4753550893826248e-06, "loss": 0.0178, "step": 18090 }, { "grad_norm": 0.0816555917263031, "learning_rate": 2.4497295866356296e-06, "loss": 0.0187, "step": 18100 }, { "grad_norm": 0.0850660651922226, "learning_rate": 2.424234084512228e-06, "loss": 0.0189, "step": 18110 }, { "grad_norm": 0.07758436352014542, "learning_rate": 2.3988686527161687e-06, "loss": 0.0191, "step": 18120 }, { "grad_norm": 0.08557108044624329, "learning_rate": 2.373633360595573e-06, "loss": 0.0188, "step": 18130 }, { "grad_norm": 0.08034805208444595, "learning_rate": 2.3485282771427585e-06, "loss": 0.0178, "step": 18140 }, { "grad_norm": 0.09413416683673859, "learning_rate": 2.3235534709940665e-06, "loss": 0.0197, "step": 18150 }, { "grad_norm": 0.08051977306604385, "learning_rate": 2.2987090104296617e-06, "loss": 0.0188, "step": 18160 }, { "grad_norm": 0.08600842207670212, "learning_rate": 2.273994963373355e-06, "loss": 0.0194, "step": 18170 }, { "grad_norm": 0.12128717452287674, "learning_rate": 2.249411397392409e-06, "loss": 0.0182, "step": 18180 }, { "grad_norm": 0.07293324917554855, "learning_rate": 2.2249583796973506e-06, "loss": 0.0179, "step": 18190 }, { "grad_norm": 0.07597894966602325, "learning_rate": 2.200635977141796e-06, "loss": 0.0201, "step": 18200 }, { "grad_norm": 0.07461533695459366, "learning_rate": 2.17644425622226e-06, "loss": 0.0173, "step": 18210 }, { "grad_norm": 0.0839361622929573, "learning_rate": 2.152383283077991e-06, "loss": 0.0178, "step": 18220 }, { "grad_norm": 0.08921470493078232, "learning_rate": 2.128453123490781e-06, "loss": 0.019, "step": 18230 }, { "grad_norm": 0.08448687195777893, "learning_rate": 2.1046538428847462e-06, "loss": 0.018, "step": 18240 }, { "grad_norm": 0.06634507328271866, "learning_rate": 2.0809855063262273e-06, "loss": 0.0191, "step": 18250 }, { "grad_norm": 0.08953968435525894, "learning_rate": 2.057448178523558e-06, "loss": 0.0192, "step": 18260 }, { "grad_norm": 0.08316390961408615, "learning_rate": 2.034041923826885e-06, "loss": 0.0199, "step": 18270 }, { "grad_norm": 0.07167964428663254, "learning_rate": 2.0107668062280204e-06, "loss": 0.0206, "step": 18280 }, { "grad_norm": 0.09577219188213348, "learning_rate": 1.9876228893602357e-06, "loss": 0.0193, "step": 18290 }, { "grad_norm": 0.09482000768184662, "learning_rate": 1.9646102364981266e-06, "loss": 0.0184, "step": 18300 }, { "grad_norm": 0.06764829903841019, "learning_rate": 1.9417289105574053e-06, "loss": 0.0193, "step": 18310 }, { "grad_norm": 0.08158379793167114, "learning_rate": 1.9189789740947427e-06, "loss": 0.0187, "step": 18320 }, { "grad_norm": 0.08493087440729141, "learning_rate": 1.896360489307597e-06, "loss": 0.0177, "step": 18330 }, { "grad_norm": 0.12112103402614594, "learning_rate": 1.8738735180340362e-06, "loss": 0.0189, "step": 18340 }, { "grad_norm": 0.07676908373832703, "learning_rate": 1.8515181217525824e-06, "loss": 0.0189, "step": 18350 }, { "grad_norm": 0.0844210535287857, "learning_rate": 1.8292943615820457e-06, "loss": 0.0189, "step": 18360 }, { "grad_norm": 0.09637466818094254, "learning_rate": 1.8072022982813296e-06, "loss": 0.0179, "step": 18370 }, { "grad_norm": 0.07009586691856384, "learning_rate": 1.7852419922492925e-06, "loss": 0.019, "step": 18380 }, { "grad_norm": 0.07139284163713455, "learning_rate": 1.763413503524569e-06, "loss": 0.0179, "step": 18390 }, { "grad_norm": 0.05992857739329338, "learning_rate": 1.7417168917854165e-06, "loss": 0.0195, "step": 18400 }, { "grad_norm": 0.0720355212688446, "learning_rate": 1.720152216349552e-06, "loss": 0.0194, "step": 18410 }, { "grad_norm": 0.11513974517583847, "learning_rate": 1.6987195361739595e-06, "loss": 0.0192, "step": 18420 }, { "grad_norm": 0.0796152800321579, "learning_rate": 1.6774189098547832e-06, "loss": 0.0177, "step": 18430 }, { "grad_norm": 0.056552786380052567, "learning_rate": 1.6562503956271069e-06, "loss": 0.018, "step": 18440 }, { "grad_norm": 0.07640979439020157, "learning_rate": 1.6352140513648417e-06, "loss": 0.0192, "step": 18450 }, { "grad_norm": 0.0677565410733223, "learning_rate": 1.6143099345805712e-06, "loss": 0.0177, "step": 18460 }, { "grad_norm": 0.07272911816835403, "learning_rate": 1.5935381024253293e-06, "loss": 0.0157, "step": 18470 }, { "grad_norm": 0.09101029485464096, "learning_rate": 1.572898611688517e-06, "loss": 0.0183, "step": 18480 }, { "grad_norm": 0.07309176027774811, "learning_rate": 1.5523915187977133e-06, "loss": 0.0176, "step": 18490 }, { "grad_norm": 0.07091354578733444, "learning_rate": 1.532016879818532e-06, "loss": 0.018, "step": 18500 }, { "grad_norm": 0.060966987162828445, "learning_rate": 1.51177475045447e-06, "loss": 0.0181, "step": 18510 }, { "grad_norm": 0.08735939115285873, "learning_rate": 1.4916651860467035e-06, "loss": 0.0196, "step": 18520 }, { "grad_norm": 0.06948551535606384, "learning_rate": 1.471688241574043e-06, "loss": 0.0188, "step": 18530 }, { "grad_norm": 0.1262359470129013, "learning_rate": 1.451843971652672e-06, "loss": 0.018, "step": 18540 }, { "grad_norm": 0.08656743168830872, "learning_rate": 1.432132430536076e-06, "loss": 0.0192, "step": 18550 }, { "grad_norm": 0.07310928404331207, "learning_rate": 1.412553672114869e-06, "loss": 0.0193, "step": 18560 }, { "grad_norm": 0.10886044800281525, "learning_rate": 1.3931077499166056e-06, "loss": 0.0176, "step": 18570 }, { "grad_norm": 0.07889262586832047, "learning_rate": 1.3737947171057085e-06, "loss": 0.0181, "step": 18580 }, { "grad_norm": 0.07615923881530762, "learning_rate": 1.3546146264832582e-06, "loss": 0.0184, "step": 18590 }, { "grad_norm": 0.10115508735179901, "learning_rate": 1.3355675304869086e-06, "loss": 0.0192, "step": 18600 }, { "grad_norm": 0.07058317214250565, "learning_rate": 1.3166534811906827e-06, "loss": 0.0187, "step": 18610 }, { "grad_norm": 0.06487547606229782, "learning_rate": 1.2978725303048666e-06, "loss": 0.0177, "step": 18620 }, { "grad_norm": 0.09874937683343887, "learning_rate": 1.2792247291758762e-06, "loss": 0.0182, "step": 18630 }, { "grad_norm": 0.07286360114812851, "learning_rate": 1.2607101287860635e-06, "loss": 0.019, "step": 18640 }, { "grad_norm": 0.0769769474864006, "learning_rate": 1.2423287797536654e-06, "loss": 0.0194, "step": 18650 }, { "grad_norm": 0.0660451129078865, "learning_rate": 1.2240807323325776e-06, "loss": 0.0174, "step": 18660 }, { "grad_norm": 0.07427334040403366, "learning_rate": 1.205966036412254e-06, "loss": 0.0191, "step": 18670 }, { "grad_norm": 0.08705832064151764, "learning_rate": 1.1879847415175949e-06, "loss": 0.0192, "step": 18680 }, { "grad_norm": 0.05896363407373428, "learning_rate": 1.1701368968087712e-06, "loss": 0.0182, "step": 18690 }, { "grad_norm": 0.07255297154188156, "learning_rate": 1.1524225510811116e-06, "loss": 0.0176, "step": 18700 }, { "grad_norm": 0.06544996798038483, "learning_rate": 1.1348417527649535e-06, "loss": 0.0175, "step": 18710 }, { "grad_norm": 0.07191231101751328, "learning_rate": 1.1173945499255268e-06, "loss": 0.017, "step": 18720 }, { "grad_norm": 0.0616837739944458, "learning_rate": 1.1000809902628307e-06, "loss": 0.0199, "step": 18730 }, { "grad_norm": 0.06197264418005943, "learning_rate": 1.082901121111468e-06, "loss": 0.0178, "step": 18740 }, { "grad_norm": 0.07012811303138733, "learning_rate": 1.0658549894405456e-06, "loss": 0.0177, "step": 18750 }, { "grad_norm": 0.06340237706899643, "learning_rate": 1.0489426418535342e-06, "loss": 0.0176, "step": 18760 }, { "grad_norm": 0.07743152976036072, "learning_rate": 1.0321641245881474e-06, "loss": 0.0187, "step": 18770 }, { "grad_norm": 0.10269147902727127, "learning_rate": 1.015519483516214e-06, "loss": 0.0197, "step": 18780 }, { "grad_norm": 0.06675872951745987, "learning_rate": 9.990087641435443e-07, "loss": 0.0174, "step": 18790 }, { "grad_norm": 0.06241551786661148, "learning_rate": 9.826320116098132e-07, "loss": 0.0178, "step": 18800 }, { "grad_norm": 0.07533831149339676, "learning_rate": 9.663892706884447e-07, "loss": 0.0184, "step": 18810 }, { "grad_norm": 0.07732561975717545, "learning_rate": 9.502805857864616e-07, "loss": 0.0177, "step": 18820 }, { "grad_norm": 0.06552597880363464, "learning_rate": 9.34306000944396e-07, "loss": 0.0181, "step": 18830 }, { "grad_norm": 0.06827986985445023, "learning_rate": 9.184655598361624e-07, "loss": 0.0161, "step": 18840 }, { "grad_norm": 0.06298546493053436, "learning_rate": 9.027593057689076e-07, "loss": 0.0169, "step": 18850 }, { "grad_norm": 0.07931306958198547, "learning_rate": 8.871872816829441e-07, "loss": 0.019, "step": 18860 }, { "grad_norm": 0.07324095815420151, "learning_rate": 8.717495301515777e-07, "loss": 0.0177, "step": 18870 }, { "grad_norm": 0.06831870973110199, "learning_rate": 8.564460933810415e-07, "loss": 0.0194, "step": 18880 }, { "grad_norm": 0.11744377017021179, "learning_rate": 8.412770132103453e-07, "loss": 0.0176, "step": 18890 }, { "grad_norm": 0.05584131181240082, "learning_rate": 8.262423311111711e-07, "loss": 0.0177, "step": 18900 }, { "grad_norm": 0.05580657720565796, "learning_rate": 8.113420881877665e-07, "loss": 0.0184, "step": 18910 }, { "grad_norm": 0.07608523219823837, "learning_rate": 7.965763251768288e-07, "loss": 0.0201, "step": 18920 }, { "grad_norm": 0.06218058988451958, "learning_rate": 7.819450824473995e-07, "loss": 0.0189, "step": 18930 }, { "grad_norm": 0.06790728121995926, "learning_rate": 7.674484000007198e-07, "loss": 0.0183, "step": 18940 }, { "grad_norm": 0.060115665197372437, "learning_rate": 7.530863174701752e-07, "loss": 0.0185, "step": 18950 }, { "grad_norm": 0.07820990681648254, "learning_rate": 7.38858874121151e-07, "loss": 0.0185, "step": 18960 }, { "grad_norm": 0.050593093037605286, "learning_rate": 7.247661088509328e-07, "loss": 0.0177, "step": 18970 }, { "grad_norm": 0.06430694460868835, "learning_rate": 7.108080601886002e-07, "loss": 0.0174, "step": 18980 }, { "grad_norm": 0.056897200644016266, "learning_rate": 6.969847662949336e-07, "loss": 0.018, "step": 18990 }, { "grad_norm": 0.07234598696231842, "learning_rate": 6.832962649622798e-07, "loss": 0.0196, "step": 19000 }, { "grad_norm": 0.06761138886213303, "learning_rate": 6.697425936144863e-07, "loss": 0.0187, "step": 19010 }, { "grad_norm": 0.07058118283748627, "learning_rate": 6.563237893067731e-07, "loss": 0.0185, "step": 19020 }, { "grad_norm": 0.07036527246236801, "learning_rate": 6.430398887256328e-07, "loss": 0.02, "step": 19030 }, { "grad_norm": 0.05924783647060394, "learning_rate": 6.298909281887478e-07, "loss": 0.0201, "step": 19040 }, { "grad_norm": 0.06524398922920227, "learning_rate": 6.168769436448673e-07, "loss": 0.0183, "step": 19050 }, { "grad_norm": 0.05983167141675949, "learning_rate": 6.03997970673742e-07, "loss": 0.0184, "step": 19060 }, { "grad_norm": 0.0779258981347084, "learning_rate": 5.912540444859782e-07, "loss": 0.0188, "step": 19070 }, { "grad_norm": 0.05197395384311676, "learning_rate": 5.786451999229837e-07, "loss": 0.0177, "step": 19080 }, { "grad_norm": 0.060394302010536194, "learning_rate": 5.661714714568722e-07, "loss": 0.018, "step": 19090 }, { "grad_norm": 0.05897321552038193, "learning_rate": 5.538328931903259e-07, "loss": 0.0178, "step": 19100 }, { "grad_norm": 0.07204906642436981, "learning_rate": 5.416294988565551e-07, "loss": 0.0174, "step": 19110 }, { "grad_norm": 0.07259289175271988, "learning_rate": 5.29561321819172e-07, "loss": 0.0168, "step": 19120 }, { "grad_norm": 0.07323791831731796, "learning_rate": 5.176283950721061e-07, "loss": 0.0197, "step": 19130 }, { "grad_norm": 0.053083620965480804, "learning_rate": 5.058307512395332e-07, "loss": 0.0183, "step": 19140 }, { "grad_norm": 0.07854636013507843, "learning_rate": 4.941684225757526e-07, "loss": 0.0188, "step": 19150 }, { "grad_norm": 0.061381272971630096, "learning_rate": 4.826414409651314e-07, "loss": 0.018, "step": 19160 }, { "grad_norm": 0.057782500982284546, "learning_rate": 4.712498379219943e-07, "loss": 0.0178, "step": 19170 }, { "grad_norm": 0.05670284852385521, "learning_rate": 4.599936445905506e-07, "loss": 0.0174, "step": 19180 }, { "grad_norm": 0.12929992377758026, "learning_rate": 4.4887289174480594e-07, "loss": 0.0189, "step": 19190 }, { "grad_norm": 0.05720289424061775, "learning_rate": 4.378876097884621e-07, "loss": 0.0184, "step": 19200 }, { "grad_norm": 0.06051964312791824, "learning_rate": 4.2703782875487264e-07, "loss": 0.0176, "step": 19210 }, { "grad_norm": 0.0493638701736927, "learning_rate": 4.163235783069208e-07, "loss": 0.0177, "step": 19220 }, { "grad_norm": 0.0652337372303009, "learning_rate": 4.057448877369585e-07, "loss": 0.0198, "step": 19230 }, { "grad_norm": 0.05942024663090706, "learning_rate": 3.9530178596672295e-07, "loss": 0.0177, "step": 19240 }, { "grad_norm": 0.05567534267902374, "learning_rate": 3.849943015472479e-07, "loss": 0.0199, "step": 19250 }, { "grad_norm": 0.05920228362083435, "learning_rate": 3.748224626588137e-07, "loss": 0.0192, "step": 19260 }, { "grad_norm": 0.06272899359464645, "learning_rate": 3.647862971108307e-07, "loss": 0.0188, "step": 19270 }, { "grad_norm": 0.07572190463542938, "learning_rate": 3.5488583234179473e-07, "loss": 0.0187, "step": 19280 }, { "grad_norm": 0.08448875695466995, "learning_rate": 3.4512109541920413e-07, "loss": 0.019, "step": 19290 }, { "grad_norm": 0.052559271454811096, "learning_rate": 3.354921130394706e-07, "loss": 0.0178, "step": 19300 }, { "grad_norm": 0.0641639456152916, "learning_rate": 3.259989115278639e-07, "loss": 0.0203, "step": 19310 }, { "grad_norm": 0.06436393409967422, "learning_rate": 3.1664151683843403e-07, "loss": 0.0182, "step": 19320 }, { "grad_norm": 0.04778827354311943, "learning_rate": 3.074199545539447e-07, "loss": 0.0173, "step": 19330 }, { "grad_norm": 0.05337690934538841, "learning_rate": 2.983342498857955e-07, "loss": 0.017, "step": 19340 }, { "grad_norm": 0.0658155083656311, "learning_rate": 2.893844276739499e-07, "loss": 0.0189, "step": 19350 }, { "grad_norm": 0.06418932974338531, "learning_rate": 2.8057051238688514e-07, "loss": 0.0179, "step": 19360 }, { "grad_norm": 0.05661708861589432, "learning_rate": 2.71892528121509e-07, "loss": 0.018, "step": 19370 }, { "grad_norm": 0.08219263702630997, "learning_rate": 2.633504986030988e-07, "loss": 0.018, "step": 19380 }, { "grad_norm": 0.06125827506184578, "learning_rate": 2.549444471852347e-07, "loss": 0.0195, "step": 19390 }, { "grad_norm": 0.0712885931134224, "learning_rate": 2.4667439684974423e-07, "loss": 0.018, "step": 19400 }, { "grad_norm": 0.0522608645260334, "learning_rate": 2.3854037020662467e-07, "loss": 0.0177, "step": 19410 }, { "grad_norm": 0.05308903008699417, "learning_rate": 2.3054238949399288e-07, "loss": 0.0198, "step": 19420 }, { "grad_norm": 0.06117355450987816, "learning_rate": 2.2268047657802993e-07, "loss": 0.0184, "step": 19430 }, { "grad_norm": 0.07344716787338257, "learning_rate": 2.149546529529034e-07, "loss": 0.0171, "step": 19440 }, { "grad_norm": 0.08279003947973251, "learning_rate": 2.0736493974071736e-07, "loss": 0.019, "step": 19450 }, { "grad_norm": 0.06080082431435585, "learning_rate": 1.9991135769145686e-07, "loss": 0.0185, "step": 19460 }, { "grad_norm": 0.04789438843727112, "learning_rate": 1.9259392718293245e-07, "loss": 0.017, "step": 19470 }, { "grad_norm": 0.0527624674141407, "learning_rate": 1.8541266822072467e-07, "loss": 0.0172, "step": 19480 }, { "grad_norm": 0.07215801626443863, "learning_rate": 1.7836760043811184e-07, "loss": 0.0191, "step": 19490 }, { "grad_norm": 0.0700642392039299, "learning_rate": 1.7145874309604792e-07, "loss": 0.0196, "step": 19500 }, { "grad_norm": 0.04787475988268852, "learning_rate": 1.6468611508308474e-07, "loss": 0.0195, "step": 19510 }, { "grad_norm": 0.061453379690647125, "learning_rate": 1.5804973491532204e-07, "loss": 0.0185, "step": 19520 }, { "grad_norm": 0.08077406883239746, "learning_rate": 1.5154962073637424e-07, "loss": 0.018, "step": 19530 }, { "grad_norm": 0.05786279961466789, "learning_rate": 1.4518579031730372e-07, "loss": 0.0189, "step": 19540 }, { "grad_norm": 0.05931727960705757, "learning_rate": 1.389582610565876e-07, "loss": 0.0193, "step": 19550 }, { "grad_norm": 0.06307802349328995, "learning_rate": 1.3286704998003995e-07, "loss": 0.0194, "step": 19560 }, { "grad_norm": 0.08200068771839142, "learning_rate": 1.2691217374080632e-07, "loss": 0.0192, "step": 19570 }, { "grad_norm": 0.05883583426475525, "learning_rate": 1.2109364861929705e-07, "loss": 0.0186, "step": 19580 }, { "grad_norm": 0.05385186895728111, "learning_rate": 1.1541149052312628e-07, "loss": 0.0178, "step": 19590 }, { "grad_norm": 0.06371750682592392, "learning_rate": 1.0986571498710074e-07, "loss": 0.0185, "step": 19600 }, { "grad_norm": 0.05983789637684822, "learning_rate": 1.0445633717316438e-07, "loss": 0.017, "step": 19610 }, { "grad_norm": 0.054410241544246674, "learning_rate": 9.918337187034277e-08, "loss": 0.0179, "step": 19620 }, { "grad_norm": 0.046379316598176956, "learning_rate": 9.404683349472643e-08, "loss": 0.0173, "step": 19630 }, { "grad_norm": 0.05279647558927536, "learning_rate": 8.904673608940983e-08, "loss": 0.0189, "step": 19640 }, { "grad_norm": 0.05891500413417816, "learning_rate": 8.418309332447471e-08, "loss": 0.0191, "step": 19650 }, { "grad_norm": 0.05261486396193504, "learning_rate": 7.945591849692902e-08, "loss": 0.0172, "step": 19660 }, { "grad_norm": 0.08951441943645477, "learning_rate": 7.486522453069578e-08, "loss": 0.0181, "step": 19670 }, { "grad_norm": 0.05398193374276161, "learning_rate": 7.041102397655208e-08, "loss": 0.0183, "step": 19680 }, { "grad_norm": 0.08699457347393036, "learning_rate": 6.609332901210685e-08, "loss": 0.0191, "step": 19690 }, { "grad_norm": 0.10914719849824905, "learning_rate": 6.191215144178419e-08, "loss": 0.0179, "step": 19700 }, { "grad_norm": 0.049424994736909866, "learning_rate": 5.786750269675678e-08, "loss": 0.0188, "step": 19710 }, { "grad_norm": 0.06691493839025497, "learning_rate": 5.395939383494031e-08, "loss": 0.02, "step": 19720 }, { "grad_norm": 0.062497660517692566, "learning_rate": 5.018783554095463e-08, "loss": 0.018, "step": 19730 }, { "grad_norm": 0.095695860683918, "learning_rate": 4.655283812610156e-08, "loss": 0.0187, "step": 19740 }, { "grad_norm": 0.0481402762234211, "learning_rate": 4.305441152831491e-08, "loss": 0.0198, "step": 19750 }, { "grad_norm": 0.06425575166940689, "learning_rate": 3.9692565312171584e-08, "loss": 0.0171, "step": 19760 }, { "grad_norm": 0.049779508262872696, "learning_rate": 3.6467308668824975e-08, "loss": 0.0183, "step": 19770 }, { "grad_norm": 0.05047006532549858, "learning_rate": 3.3378650416004964e-08, "loss": 0.0171, "step": 19780 }, { "grad_norm": 0.047848593443632126, "learning_rate": 3.042659899797906e-08, "loss": 0.0178, "step": 19790 }, { "grad_norm": 0.04910007864236832, "learning_rate": 2.76111624855524e-08, "loss": 0.0194, "step": 19800 }, { "grad_norm": 0.04693181812763214, "learning_rate": 2.4932348576017784e-08, "loss": 0.0169, "step": 19810 }, { "grad_norm": 0.05605169013142586, "learning_rate": 2.239016459314458e-08, "loss": 0.019, "step": 19820 }, { "grad_norm": 0.092982716858387, "learning_rate": 1.9984617487173174e-08, "loss": 0.0192, "step": 19830 }, { "grad_norm": 0.06333128362894058, "learning_rate": 1.7715713834776105e-08, "loss": 0.0185, "step": 19840 }, { "grad_norm": 0.05863652378320694, "learning_rate": 1.5583459839046964e-08, "loss": 0.0194, "step": 19850 }, { "grad_norm": 0.05939078703522682, "learning_rate": 1.3587861329489304e-08, "loss": 0.0185, "step": 19860 }, { "grad_norm": 0.05109507590532303, "learning_rate": 1.1728923761994415e-08, "loss": 0.0184, "step": 19870 }, { "grad_norm": 0.06612105667591095, "learning_rate": 1.0006652218819135e-08, "loss": 0.019, "step": 19880 }, { "grad_norm": 0.07438040524721146, "learning_rate": 8.421051408596947e-09, "loss": 0.0194, "step": 19890 }, { "grad_norm": 0.061179425567388535, "learning_rate": 6.972125666299123e-09, "loss": 0.0197, "step": 19900 }, { "grad_norm": 0.06516063213348389, "learning_rate": 5.659878953229169e-09, "loss": 0.0195, "step": 19910 }, { "grad_norm": 0.054037563502788544, "learning_rate": 4.48431485701728e-09, "loss": 0.0193, "step": 19920 }, { "grad_norm": 0.055296141654253006, "learning_rate": 3.4454365916203322e-09, "loss": 0.0179, "step": 19930 }, { "grad_norm": 0.04656418785452843, "learning_rate": 2.5432469972830332e-09, "loss": 0.0187, "step": 19940 }, { "grad_norm": 0.05092895030975342, "learning_rate": 1.7777485405601203e-09, "loss": 0.0173, "step": 19950 }, { "grad_norm": 0.05153869092464447, "learning_rate": 1.1489433142941597e-09, "loss": 0.0173, "step": 19960 }, { "grad_norm": 0.06036381423473358, "learning_rate": 6.568330376210963e-10, "loss": 0.0196, "step": 19970 }, { "grad_norm": 0.047707609832286835, "learning_rate": 3.0141905594249787e-10, "loss": 0.0184, "step": 19980 }, { "grad_norm": 0.05357581004500389, "learning_rate": 8.270234094776008e-11, "loss": 0.0178, "step": 19990 }, { "grad_norm": 0.03908345848321915, "learning_rate": 6.834906085551041e-13, "loss": 0.0188, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 4000, "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": 256, "trial_name": null, "trial_params": null }