{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 90705, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 3.9727091789245605, "learning_rate": 1.9841269841269841e-07, "loss": 1.2449, "step": 10 }, { "grad_norm": 4.266497611999512, "learning_rate": 4.1887125220458554e-07, "loss": 1.2445, "step": 20 }, { "grad_norm": 3.6945888996124268, "learning_rate": 6.393298059964726e-07, "loss": 1.2359, "step": 30 }, { "grad_norm": 3.3150298595428467, "learning_rate": 8.597883597883597e-07, "loss": 1.2168, "step": 40 }, { "grad_norm": 3.1260156631469727, "learning_rate": 1.0802469135802469e-06, "loss": 1.2199, "step": 50 }, { "grad_norm": 2.0918667316436768, "learning_rate": 1.300705467372134e-06, "loss": 1.1848, "step": 60 }, { "grad_norm": 1.5751596689224243, "learning_rate": 1.5211640211640211e-06, "loss": 1.1305, "step": 70 }, { "grad_norm": 1.1673942804336548, "learning_rate": 1.7416225749559084e-06, "loss": 1.0973, "step": 80 }, { "grad_norm": 0.7059447169303894, "learning_rate": 1.9620811287477954e-06, "loss": 1.0496, "step": 90 }, { "grad_norm": 0.5040328502655029, "learning_rate": 2.1825396825396824e-06, "loss": 1.0441, "step": 100 }, { "grad_norm": 0.4679824411869049, "learning_rate": 2.40299823633157e-06, "loss": 1.0418, "step": 110 }, { "grad_norm": 0.4639356732368469, "learning_rate": 2.623456790123457e-06, "loss": 1.0312, "step": 120 }, { "grad_norm": 0.39873185753822327, "learning_rate": 2.843915343915344e-06, "loss": 1.0398, "step": 130 }, { "grad_norm": 0.4860520362854004, "learning_rate": 3.064373897707231e-06, "loss": 1.0363, "step": 140 }, { "grad_norm": 0.38556888699531555, "learning_rate": 3.284832451499118e-06, "loss": 1.0352, "step": 150 }, { "grad_norm": 0.42066851258277893, "learning_rate": 3.5052910052910054e-06, "loss": 1.032, "step": 160 }, { "grad_norm": 0.5018773674964905, "learning_rate": 3.7257495590828924e-06, "loss": 1.018, "step": 170 }, { "grad_norm": 0.28947851061820984, "learning_rate": 3.946208112874779e-06, "loss": 1.0234, "step": 180 }, { "grad_norm": 0.3536842465400696, "learning_rate": 4.166666666666667e-06, "loss": 1.0225, "step": 190 }, { "grad_norm": 0.2965538203716278, "learning_rate": 4.3871252204585535e-06, "loss": 1.024, "step": 200 }, { "grad_norm": 0.4923820495605469, "learning_rate": 4.607583774250441e-06, "loss": 1.0289, "step": 210 }, { "grad_norm": 0.365470290184021, "learning_rate": 4.828042328042328e-06, "loss": 1.0336, "step": 220 }, { "grad_norm": 0.39292263984680176, "learning_rate": 5.048500881834215e-06, "loss": 1.015, "step": 230 }, { "grad_norm": 0.4249347746372223, "learning_rate": 5.268959435626102e-06, "loss": 1.0213, "step": 240 }, { "grad_norm": 0.3931747078895569, "learning_rate": 5.489417989417989e-06, "loss": 1.0242, "step": 250 }, { "grad_norm": 0.32198604941368103, "learning_rate": 5.7098765432098764e-06, "loss": 1.0197, "step": 260 }, { "grad_norm": 0.4586363732814789, "learning_rate": 5.930335097001764e-06, "loss": 1.0174, "step": 270 }, { "grad_norm": 0.38224002718925476, "learning_rate": 6.1507936507936505e-06, "loss": 1.0215, "step": 280 }, { "grad_norm": 0.5327089428901672, "learning_rate": 6.371252204585539e-06, "loss": 1.007, "step": 290 }, { "grad_norm": 0.4740910232067108, "learning_rate": 6.591710758377425e-06, "loss": 1.015, "step": 300 }, { "grad_norm": 0.641400933265686, "learning_rate": 6.812169312169313e-06, "loss": 0.9861, "step": 310 }, { "grad_norm": 1.1272286176681519, "learning_rate": 7.0326278659612e-06, "loss": 0.9471, "step": 320 }, { "grad_norm": 1.6132850646972656, "learning_rate": 7.253086419753087e-06, "loss": 0.8686, "step": 330 }, { "grad_norm": 1.6038436889648438, "learning_rate": 7.473544973544974e-06, "loss": 0.7928, "step": 340 }, { "grad_norm": 1.6522324085235596, "learning_rate": 7.694003527336861e-06, "loss": 0.7508, "step": 350 }, { "grad_norm": 1.944132685661316, "learning_rate": 7.91446208112875e-06, "loss": 0.6986, "step": 360 }, { "grad_norm": 2.1045079231262207, "learning_rate": 8.134920634920636e-06, "loss": 0.6576, "step": 370 }, { "grad_norm": 2.4142909049987793, "learning_rate": 8.355379188712522e-06, "loss": 0.6143, "step": 380 }, { "grad_norm": 2.552396774291992, "learning_rate": 8.575837742504409e-06, "loss": 0.5536, "step": 390 }, { "grad_norm": 2.4402475357055664, "learning_rate": 8.796296296296297e-06, "loss": 0.5001, "step": 400 }, { "grad_norm": 2.7372348308563232, "learning_rate": 9.016754850088184e-06, "loss": 0.451, "step": 410 }, { "grad_norm": 2.2060861587524414, "learning_rate": 9.23721340388007e-06, "loss": 0.3952, "step": 420 }, { "grad_norm": 2.6041111946105957, "learning_rate": 9.457671957671959e-06, "loss": 0.3596, "step": 430 }, { "grad_norm": 1.960633635520935, "learning_rate": 9.678130511463845e-06, "loss": 0.3229, "step": 440 }, { "grad_norm": 2.9828975200653076, "learning_rate": 9.898589065255732e-06, "loss": 0.2854, "step": 450 }, { "grad_norm": 2.2687416076660156, "learning_rate": 1.011904761904762e-05, "loss": 0.2631, "step": 460 }, { "grad_norm": 2.5895698070526123, "learning_rate": 1.0339506172839507e-05, "loss": 0.2501, "step": 470 }, { "grad_norm": 2.266566753387451, "learning_rate": 1.0559964726631393e-05, "loss": 0.2278, "step": 480 }, { "grad_norm": 1.858057975769043, "learning_rate": 1.0780423280423282e-05, "loss": 0.2138, "step": 490 }, { "grad_norm": 1.7918211221694946, "learning_rate": 1.1000881834215168e-05, "loss": 0.1991, "step": 500 }, { "grad_norm": 1.7447489500045776, "learning_rate": 1.1221340388007055e-05, "loss": 0.1919, "step": 510 }, { "grad_norm": 2.0552680492401123, "learning_rate": 1.1441798941798943e-05, "loss": 0.1766, "step": 520 }, { "grad_norm": 1.9516521692276, "learning_rate": 1.166225749559083e-05, "loss": 0.1831, "step": 530 }, { "grad_norm": 1.8374879360198975, "learning_rate": 1.1882716049382716e-05, "loss": 0.1528, "step": 540 }, { "grad_norm": 1.7183077335357666, "learning_rate": 1.2103174603174603e-05, "loss": 0.1598, "step": 550 }, { "grad_norm": 1.8435531854629517, "learning_rate": 1.2323633156966491e-05, "loss": 0.1456, "step": 560 }, { "grad_norm": 2.405951738357544, "learning_rate": 1.254409171075838e-05, "loss": 0.1381, "step": 570 }, { "grad_norm": 1.5730798244476318, "learning_rate": 1.2764550264550265e-05, "loss": 0.1436, "step": 580 }, { "grad_norm": 2.3104043006896973, "learning_rate": 1.2985008818342155e-05, "loss": 0.1487, "step": 590 }, { "grad_norm": 2.621455192565918, "learning_rate": 1.320546737213404e-05, "loss": 0.139, "step": 600 }, { "grad_norm": 2.1750593185424805, "learning_rate": 1.3425925925925928e-05, "loss": 0.1535, "step": 610 }, { "grad_norm": 1.6293586492538452, "learning_rate": 1.3646384479717814e-05, "loss": 0.126, "step": 620 }, { "grad_norm": 2.298038959503174, "learning_rate": 1.3866843033509703e-05, "loss": 0.1127, "step": 630 }, { "grad_norm": 1.73967444896698, "learning_rate": 1.4087301587301587e-05, "loss": 0.1113, "step": 640 }, { "grad_norm": 2.3229684829711914, "learning_rate": 1.4307760141093476e-05, "loss": 0.1125, "step": 650 }, { "grad_norm": 1.6949951648712158, "learning_rate": 1.4528218694885362e-05, "loss": 0.1081, "step": 660 }, { "grad_norm": 2.1708574295043945, "learning_rate": 1.474867724867725e-05, "loss": 0.1296, "step": 670 }, { "grad_norm": 2.8138720989227295, "learning_rate": 1.4969135802469136e-05, "loss": 0.1121, "step": 680 }, { "grad_norm": 2.0427732467651367, "learning_rate": 1.5189594356261026e-05, "loss": 0.0985, "step": 690 }, { "grad_norm": 1.7366427183151245, "learning_rate": 1.5410052910052912e-05, "loss": 0.105, "step": 700 }, { "grad_norm": 1.9069232940673828, "learning_rate": 1.56305114638448e-05, "loss": 0.1147, "step": 710 }, { "grad_norm": 1.7431962490081787, "learning_rate": 1.5850970017636685e-05, "loss": 0.0987, "step": 720 }, { "grad_norm": 2.1397457122802734, "learning_rate": 1.6071428571428572e-05, "loss": 0.0988, "step": 730 }, { "grad_norm": 1.791387915611267, "learning_rate": 1.629188712522046e-05, "loss": 0.0814, "step": 740 }, { "grad_norm": 1.335062861442566, "learning_rate": 1.651234567901235e-05, "loss": 0.0944, "step": 750 }, { "grad_norm": 2.1201820373535156, "learning_rate": 1.6732804232804232e-05, "loss": 0.0839, "step": 760 }, { "grad_norm": 1.3060195446014404, "learning_rate": 1.6953262786596122e-05, "loss": 0.0901, "step": 770 }, { "grad_norm": 1.2846031188964844, "learning_rate": 1.717372134038801e-05, "loss": 0.0874, "step": 780 }, { "grad_norm": 1.8046228885650635, "learning_rate": 1.7394179894179895e-05, "loss": 0.0838, "step": 790 }, { "grad_norm": 1.6489754915237427, "learning_rate": 1.761463844797178e-05, "loss": 0.0873, "step": 800 }, { "grad_norm": 1.0265909433364868, "learning_rate": 1.783509700176367e-05, "loss": 0.0885, "step": 810 }, { "grad_norm": 1.3547688722610474, "learning_rate": 1.8055555555555555e-05, "loss": 0.0777, "step": 820 }, { "grad_norm": 1.3912655115127563, "learning_rate": 1.8276014109347445e-05, "loss": 0.0835, "step": 830 }, { "grad_norm": 1.6364963054656982, "learning_rate": 1.849647266313933e-05, "loss": 0.079, "step": 840 }, { "grad_norm": 1.4650019407272339, "learning_rate": 1.8716931216931218e-05, "loss": 0.0884, "step": 850 }, { "grad_norm": 1.5996640920639038, "learning_rate": 1.8937389770723105e-05, "loss": 0.0744, "step": 860 }, { "grad_norm": 1.6634843349456787, "learning_rate": 1.9157848324514994e-05, "loss": 0.0807, "step": 870 }, { "grad_norm": 1.8248449563980103, "learning_rate": 1.9378306878306878e-05, "loss": 0.0696, "step": 880 }, { "grad_norm": 1.3528789281845093, "learning_rate": 1.9598765432098768e-05, "loss": 0.0736, "step": 890 }, { "grad_norm": 1.6499396562576294, "learning_rate": 1.9819223985890654e-05, "loss": 0.0823, "step": 900 }, { "grad_norm": 1.7581647634506226, "learning_rate": 2.003968253968254e-05, "loss": 0.0715, "step": 910 }, { "grad_norm": 1.8056023120880127, "learning_rate": 2.0260141093474427e-05, "loss": 0.0808, "step": 920 }, { "grad_norm": 1.8105623722076416, "learning_rate": 2.0480599647266317e-05, "loss": 0.0697, "step": 930 }, { "grad_norm": 1.3996994495391846, "learning_rate": 2.07010582010582e-05, "loss": 0.0673, "step": 940 }, { "grad_norm": 1.6322094202041626, "learning_rate": 2.092151675485009e-05, "loss": 0.0805, "step": 950 }, { "grad_norm": 1.0672597885131836, "learning_rate": 2.1141975308641977e-05, "loss": 0.0638, "step": 960 }, { "grad_norm": 2.172330141067505, "learning_rate": 2.1362433862433864e-05, "loss": 0.0635, "step": 970 }, { "grad_norm": 1.4284693002700806, "learning_rate": 2.158289241622575e-05, "loss": 0.0695, "step": 980 }, { "grad_norm": 1.7590469121932983, "learning_rate": 2.1803350970017637e-05, "loss": 0.0719, "step": 990 }, { "grad_norm": 1.0337010622024536, "learning_rate": 2.2023809523809524e-05, "loss": 0.0659, "step": 1000 }, { "grad_norm": 1.187981367111206, "learning_rate": 2.2244268077601414e-05, "loss": 0.0643, "step": 1010 }, { "grad_norm": 1.4629205465316772, "learning_rate": 2.24647266313933e-05, "loss": 0.0626, "step": 1020 }, { "grad_norm": 1.3576452732086182, "learning_rate": 2.2685185185185187e-05, "loss": 0.069, "step": 1030 }, { "grad_norm": 1.5236841440200806, "learning_rate": 2.2905643738977073e-05, "loss": 0.0657, "step": 1040 }, { "grad_norm": 1.1200459003448486, "learning_rate": 2.312610229276896e-05, "loss": 0.0609, "step": 1050 }, { "grad_norm": 1.7263450622558594, "learning_rate": 2.3346560846560847e-05, "loss": 0.0733, "step": 1060 }, { "grad_norm": 1.4978276491165161, "learning_rate": 2.3567019400352737e-05, "loss": 0.0663, "step": 1070 }, { "grad_norm": 1.6703027486801147, "learning_rate": 2.378747795414462e-05, "loss": 0.0695, "step": 1080 }, { "grad_norm": 1.7204174995422363, "learning_rate": 2.400793650793651e-05, "loss": 0.0534, "step": 1090 }, { "grad_norm": 1.1813308000564575, "learning_rate": 2.4228395061728396e-05, "loss": 0.0595, "step": 1100 }, { "grad_norm": 0.7972854375839233, "learning_rate": 2.4448853615520283e-05, "loss": 0.0673, "step": 1110 }, { "grad_norm": 1.6913692951202393, "learning_rate": 2.466931216931217e-05, "loss": 0.062, "step": 1120 }, { "grad_norm": 1.4429078102111816, "learning_rate": 2.488977072310406e-05, "loss": 0.0705, "step": 1130 }, { "grad_norm": 1.6779141426086426, "learning_rate": 2.5110229276895946e-05, "loss": 0.0637, "step": 1140 }, { "grad_norm": 0.9283114075660706, "learning_rate": 2.5330687830687833e-05, "loss": 0.0591, "step": 1150 }, { "grad_norm": 1.3795076608657837, "learning_rate": 2.555114638447972e-05, "loss": 0.0625, "step": 1160 }, { "grad_norm": 1.6687039136886597, "learning_rate": 2.5771604938271603e-05, "loss": 0.0639, "step": 1170 }, { "grad_norm": 0.9644541144371033, "learning_rate": 2.5992063492063496e-05, "loss": 0.0637, "step": 1180 }, { "grad_norm": 1.5841455459594727, "learning_rate": 2.6212522045855383e-05, "loss": 0.0611, "step": 1190 }, { "grad_norm": 0.8746373653411865, "learning_rate": 2.6432980599647266e-05, "loss": 0.0638, "step": 1200 }, { "grad_norm": 1.3063255548477173, "learning_rate": 2.6653439153439152e-05, "loss": 0.0623, "step": 1210 }, { "grad_norm": 1.0743155479431152, "learning_rate": 2.6873897707231042e-05, "loss": 0.0525, "step": 1220 }, { "grad_norm": 1.418043851852417, "learning_rate": 2.709435626102293e-05, "loss": 0.063, "step": 1230 }, { "grad_norm": 1.3658545017242432, "learning_rate": 2.7314814814814816e-05, "loss": 0.0564, "step": 1240 }, { "grad_norm": 0.8988314867019653, "learning_rate": 2.7535273368606702e-05, "loss": 0.06, "step": 1250 }, { "grad_norm": 1.39566969871521, "learning_rate": 2.7755731922398592e-05, "loss": 0.0537, "step": 1260 }, { "grad_norm": 1.4576269388198853, "learning_rate": 2.797619047619048e-05, "loss": 0.0653, "step": 1270 }, { "grad_norm": 1.6392078399658203, "learning_rate": 2.8196649029982365e-05, "loss": 0.062, "step": 1280 }, { "grad_norm": 1.279990315437317, "learning_rate": 2.841710758377425e-05, "loss": 0.0659, "step": 1290 }, { "grad_norm": 1.5828242301940918, "learning_rate": 2.8637566137566142e-05, "loss": 0.0605, "step": 1300 }, { "grad_norm": 0.9892917275428772, "learning_rate": 2.8858024691358025e-05, "loss": 0.0581, "step": 1310 }, { "grad_norm": 1.1595244407653809, "learning_rate": 2.9078483245149912e-05, "loss": 0.0617, "step": 1320 }, { "grad_norm": 0.9096443057060242, "learning_rate": 2.92989417989418e-05, "loss": 0.0504, "step": 1330 }, { "grad_norm": 0.7367215752601624, "learning_rate": 2.951940035273369e-05, "loss": 0.0557, "step": 1340 }, { "grad_norm": 0.8230398893356323, "learning_rate": 2.9739858906525575e-05, "loss": 0.0508, "step": 1350 }, { "grad_norm": 1.3531360626220703, "learning_rate": 2.996031746031746e-05, "loss": 0.0532, "step": 1360 }, { "grad_norm": 1.1673325300216675, "learning_rate": 3.0180776014109348e-05, "loss": 0.0447, "step": 1370 }, { "grad_norm": 0.9227582812309265, "learning_rate": 3.0401234567901238e-05, "loss": 0.0495, "step": 1380 }, { "grad_norm": 1.4739028215408325, "learning_rate": 3.062169312169312e-05, "loss": 0.0505, "step": 1390 }, { "grad_norm": 1.1722733974456787, "learning_rate": 3.084215167548501e-05, "loss": 0.0611, "step": 1400 }, { "grad_norm": 1.165026068687439, "learning_rate": 3.1062610229276895e-05, "loss": 0.0466, "step": 1410 }, { "grad_norm": 1.074580430984497, "learning_rate": 3.1283068783068784e-05, "loss": 0.0605, "step": 1420 }, { "grad_norm": 0.920509934425354, "learning_rate": 3.1503527336860674e-05, "loss": 0.0543, "step": 1430 }, { "grad_norm": 0.9097546935081482, "learning_rate": 3.172398589065256e-05, "loss": 0.0536, "step": 1440 }, { "grad_norm": 1.31371009349823, "learning_rate": 3.194444444444444e-05, "loss": 0.0568, "step": 1450 }, { "grad_norm": 0.7599324584007263, "learning_rate": 3.216490299823634e-05, "loss": 0.055, "step": 1460 }, { "grad_norm": 1.4091113805770874, "learning_rate": 3.238536155202822e-05, "loss": 0.0531, "step": 1470 }, { "grad_norm": 0.9330068826675415, "learning_rate": 3.2605820105820104e-05, "loss": 0.0567, "step": 1480 }, { "grad_norm": 0.7672646641731262, "learning_rate": 3.2826278659611994e-05, "loss": 0.0513, "step": 1490 }, { "grad_norm": 1.315729022026062, "learning_rate": 3.3046737213403884e-05, "loss": 0.0578, "step": 1500 }, { "grad_norm": 1.0171236991882324, "learning_rate": 3.326719576719577e-05, "loss": 0.0552, "step": 1510 }, { "grad_norm": 1.0306004285812378, "learning_rate": 3.348765432098766e-05, "loss": 0.0508, "step": 1520 }, { "grad_norm": 0.8343230485916138, "learning_rate": 3.370811287477954e-05, "loss": 0.0555, "step": 1530 }, { "grad_norm": 1.6432008743286133, "learning_rate": 3.392857142857143e-05, "loss": 0.0604, "step": 1540 }, { "grad_norm": 0.8494104146957397, "learning_rate": 3.414902998236332e-05, "loss": 0.0524, "step": 1550 }, { "grad_norm": 0.8459282517433167, "learning_rate": 3.4369488536155204e-05, "loss": 0.0505, "step": 1560 }, { "grad_norm": 1.357969880104065, "learning_rate": 3.458994708994709e-05, "loss": 0.0495, "step": 1570 }, { "grad_norm": 1.0906956195831299, "learning_rate": 3.4810405643738984e-05, "loss": 0.0492, "step": 1580 }, { "grad_norm": 0.8925685286521912, "learning_rate": 3.503086419753087e-05, "loss": 0.05, "step": 1590 }, { "grad_norm": 1.3753013610839844, "learning_rate": 3.525132275132275e-05, "loss": 0.0492, "step": 1600 }, { "grad_norm": 1.1666048765182495, "learning_rate": 3.547178130511464e-05, "loss": 0.0489, "step": 1610 }, { "grad_norm": 0.9918048977851868, "learning_rate": 3.569223985890653e-05, "loss": 0.0426, "step": 1620 }, { "grad_norm": 0.5872971415519714, "learning_rate": 3.591269841269841e-05, "loss": 0.0475, "step": 1630 }, { "grad_norm": 0.8629538416862488, "learning_rate": 3.61331569664903e-05, "loss": 0.0499, "step": 1640 }, { "grad_norm": 0.8983898162841797, "learning_rate": 3.6353615520282186e-05, "loss": 0.0527, "step": 1650 }, { "grad_norm": 0.7907662987709045, "learning_rate": 3.6574074074074076e-05, "loss": 0.0372, "step": 1660 }, { "grad_norm": 0.8957839012145996, "learning_rate": 3.6794532627865966e-05, "loss": 0.0485, "step": 1670 }, { "grad_norm": 0.6047958731651306, "learning_rate": 3.701499118165785e-05, "loss": 0.0531, "step": 1680 }, { "grad_norm": 0.8826355934143066, "learning_rate": 3.723544973544973e-05, "loss": 0.0534, "step": 1690 }, { "grad_norm": 1.4728583097457886, "learning_rate": 3.745590828924162e-05, "loss": 0.0455, "step": 1700 }, { "grad_norm": 0.924312174320221, "learning_rate": 3.767636684303351e-05, "loss": 0.044, "step": 1710 }, { "grad_norm": 0.8607197403907776, "learning_rate": 3.7896825396825396e-05, "loss": 0.051, "step": 1720 }, { "grad_norm": 0.8497306108474731, "learning_rate": 3.8117283950617286e-05, "loss": 0.0519, "step": 1730 }, { "grad_norm": 0.6050167679786682, "learning_rate": 3.8337742504409176e-05, "loss": 0.0479, "step": 1740 }, { "grad_norm": 0.5061044096946716, "learning_rate": 3.855820105820106e-05, "loss": 0.0459, "step": 1750 }, { "grad_norm": 0.7951064705848694, "learning_rate": 3.877865961199294e-05, "loss": 0.0515, "step": 1760 }, { "grad_norm": 0.8302925825119019, "learning_rate": 3.899911816578484e-05, "loss": 0.0467, "step": 1770 }, { "grad_norm": 0.44537433981895447, "learning_rate": 3.921957671957672e-05, "loss": 0.044, "step": 1780 }, { "grad_norm": 0.9236002564430237, "learning_rate": 3.9440035273368606e-05, "loss": 0.0505, "step": 1790 }, { "grad_norm": 0.9867004156112671, "learning_rate": 3.9660493827160496e-05, "loss": 0.0493, "step": 1800 }, { "grad_norm": 0.9377702474594116, "learning_rate": 3.9880952380952386e-05, "loss": 0.0484, "step": 1810 }, { "grad_norm": 0.7115578055381775, "learning_rate": 4.010141093474427e-05, "loss": 0.042, "step": 1820 }, { "grad_norm": 0.5101749300956726, "learning_rate": 4.032186948853616e-05, "loss": 0.0427, "step": 1830 }, { "grad_norm": 0.8565178513526917, "learning_rate": 4.054232804232804e-05, "loss": 0.0415, "step": 1840 }, { "grad_norm": 1.031072735786438, "learning_rate": 4.076278659611993e-05, "loss": 0.0441, "step": 1850 }, { "grad_norm": 0.938869833946228, "learning_rate": 4.098324514991182e-05, "loss": 0.0494, "step": 1860 }, { "grad_norm": 0.6841536164283752, "learning_rate": 4.1203703703703705e-05, "loss": 0.0506, "step": 1870 }, { "grad_norm": 0.7050836682319641, "learning_rate": 4.142416225749559e-05, "loss": 0.0412, "step": 1880 }, { "grad_norm": 0.8972555994987488, "learning_rate": 4.1644620811287485e-05, "loss": 0.0405, "step": 1890 }, { "grad_norm": 0.7760329842567444, "learning_rate": 4.186507936507937e-05, "loss": 0.0583, "step": 1900 }, { "grad_norm": 0.7573826313018799, "learning_rate": 4.208553791887125e-05, "loss": 0.046, "step": 1910 }, { "grad_norm": 0.49431484937667847, "learning_rate": 4.230599647266314e-05, "loss": 0.0417, "step": 1920 }, { "grad_norm": 1.104238748550415, "learning_rate": 4.252645502645503e-05, "loss": 0.0454, "step": 1930 }, { "grad_norm": 0.7912407517433167, "learning_rate": 4.2746913580246915e-05, "loss": 0.0417, "step": 1940 }, { "grad_norm": 0.6641307473182678, "learning_rate": 4.2967372134038805e-05, "loss": 0.0441, "step": 1950 }, { "grad_norm": 0.5976917147636414, "learning_rate": 4.318783068783069e-05, "loss": 0.0378, "step": 1960 }, { "grad_norm": 0.691142201423645, "learning_rate": 4.340828924162258e-05, "loss": 0.0444, "step": 1970 }, { "grad_norm": 1.313568115234375, "learning_rate": 4.362874779541447e-05, "loss": 0.0549, "step": 1980 }, { "grad_norm": 0.7250801920890808, "learning_rate": 4.384920634920635e-05, "loss": 0.0444, "step": 1990 }, { "grad_norm": 0.669163703918457, "learning_rate": 4.4069664902998234e-05, "loss": 0.0402, "step": 2000 }, { "grad_norm": 0.8633808493614197, "learning_rate": 4.429012345679013e-05, "loss": 0.0405, "step": 2010 }, { "grad_norm": 0.9168856739997864, "learning_rate": 4.4510582010582014e-05, "loss": 0.0446, "step": 2020 }, { "grad_norm": 0.6814913153648376, "learning_rate": 4.47310405643739e-05, "loss": 0.0478, "step": 2030 }, { "grad_norm": 0.5873094201087952, "learning_rate": 4.495149911816579e-05, "loss": 0.0497, "step": 2040 }, { "grad_norm": 0.8535837531089783, "learning_rate": 4.517195767195768e-05, "loss": 0.0412, "step": 2050 }, { "grad_norm": 0.557858943939209, "learning_rate": 4.539241622574956e-05, "loss": 0.0382, "step": 2060 }, { "grad_norm": 0.8151389956474304, "learning_rate": 4.561287477954145e-05, "loss": 0.0416, "step": 2070 }, { "grad_norm": 0.5355256795883179, "learning_rate": 4.5833333333333334e-05, "loss": 0.037, "step": 2080 }, { "grad_norm": 0.437349408864975, "learning_rate": 4.6053791887125224e-05, "loss": 0.0346, "step": 2090 }, { "grad_norm": 0.706866443157196, "learning_rate": 4.6274250440917114e-05, "loss": 0.0346, "step": 2100 }, { "grad_norm": 0.561486542224884, "learning_rate": 4.6494708994709e-05, "loss": 0.0368, "step": 2110 }, { "grad_norm": 0.629368245601654, "learning_rate": 4.671516754850088e-05, "loss": 0.0336, "step": 2120 }, { "grad_norm": 0.5728518962860107, "learning_rate": 4.693562610229277e-05, "loss": 0.0391, "step": 2130 }, { "grad_norm": 1.0224108695983887, "learning_rate": 4.715608465608466e-05, "loss": 0.0437, "step": 2140 }, { "grad_norm": 0.6586576104164124, "learning_rate": 4.7376543209876543e-05, "loss": 0.0512, "step": 2150 }, { "grad_norm": 0.6802510023117065, "learning_rate": 4.7597001763668433e-05, "loss": 0.0385, "step": 2160 }, { "grad_norm": 0.5301350951194763, "learning_rate": 4.781746031746032e-05, "loss": 0.0346, "step": 2170 }, { "grad_norm": 1.1564081907272339, "learning_rate": 4.8037918871252207e-05, "loss": 0.0465, "step": 2180 }, { "grad_norm": 0.736979067325592, "learning_rate": 4.8258377425044097e-05, "loss": 0.0378, "step": 2190 }, { "grad_norm": 0.8070315718650818, "learning_rate": 4.847883597883598e-05, "loss": 0.0392, "step": 2200 }, { "grad_norm": 0.7366067171096802, "learning_rate": 4.869929453262787e-05, "loss": 0.0369, "step": 2210 }, { "grad_norm": 0.6569851040840149, "learning_rate": 4.891975308641975e-05, "loss": 0.0365, "step": 2220 }, { "grad_norm": 0.9452429413795471, "learning_rate": 4.914021164021164e-05, "loss": 0.0379, "step": 2230 }, { "grad_norm": 0.5302839279174805, "learning_rate": 4.9360670194003526e-05, "loss": 0.033, "step": 2240 }, { "grad_norm": 0.657779335975647, "learning_rate": 4.9581128747795416e-05, "loss": 0.0414, "step": 2250 }, { "grad_norm": 0.7192308902740479, "learning_rate": 4.9801587301587306e-05, "loss": 0.0381, "step": 2260 }, { "grad_norm": 0.6906507015228271, "learning_rate": 5.0022045855379196e-05, "loss": 0.0354, "step": 2270 }, { "grad_norm": 1.0288872718811035, "learning_rate": 5.024250440917108e-05, "loss": 0.0444, "step": 2280 }, { "grad_norm": 0.7283273935317993, "learning_rate": 5.046296296296297e-05, "loss": 0.0417, "step": 2290 }, { "grad_norm": 0.6848441958427429, "learning_rate": 5.0683421516754846e-05, "loss": 0.0374, "step": 2300 }, { "grad_norm": 0.5878885388374329, "learning_rate": 5.0903880070546736e-05, "loss": 0.0421, "step": 2310 }, { "grad_norm": 0.8764672875404358, "learning_rate": 5.112433862433863e-05, "loss": 0.0344, "step": 2320 }, { "grad_norm": 0.39648154377937317, "learning_rate": 5.134479717813051e-05, "loss": 0.0485, "step": 2330 }, { "grad_norm": 0.5573259592056274, "learning_rate": 5.15652557319224e-05, "loss": 0.0368, "step": 2340 }, { "grad_norm": 0.9108426570892334, "learning_rate": 5.1785714285714296e-05, "loss": 0.0449, "step": 2350 }, { "grad_norm": 0.5590903162956238, "learning_rate": 5.200617283950617e-05, "loss": 0.0372, "step": 2360 }, { "grad_norm": 0.5880994200706482, "learning_rate": 5.222663139329806e-05, "loss": 0.0381, "step": 2370 }, { "grad_norm": 0.7257521152496338, "learning_rate": 5.2447089947089945e-05, "loss": 0.0333, "step": 2380 }, { "grad_norm": 0.9818815588951111, "learning_rate": 5.2667548500881835e-05, "loss": 0.0391, "step": 2390 }, { "grad_norm": 0.6356369853019714, "learning_rate": 5.2888007054673725e-05, "loss": 0.0407, "step": 2400 }, { "grad_norm": 0.5782774090766907, "learning_rate": 5.310846560846561e-05, "loss": 0.0322, "step": 2410 }, { "grad_norm": 0.6343235373497009, "learning_rate": 5.33289241622575e-05, "loss": 0.0365, "step": 2420 }, { "grad_norm": 0.5656561255455017, "learning_rate": 5.354938271604939e-05, "loss": 0.0377, "step": 2430 }, { "grad_norm": 0.6332433223724365, "learning_rate": 5.376984126984127e-05, "loss": 0.0374, "step": 2440 }, { "grad_norm": 0.5313704013824463, "learning_rate": 5.399029982363316e-05, "loss": 0.0329, "step": 2450 }, { "grad_norm": 0.6799907684326172, "learning_rate": 5.4210758377425045e-05, "loss": 0.0394, "step": 2460 }, { "grad_norm": 0.7366439700126648, "learning_rate": 5.4431216931216935e-05, "loss": 0.0395, "step": 2470 }, { "grad_norm": 0.7594524621963501, "learning_rate": 5.4651675485008825e-05, "loss": 0.0377, "step": 2480 }, { "grad_norm": 0.5178806781768799, "learning_rate": 5.48721340388007e-05, "loss": 0.0301, "step": 2490 }, { "grad_norm": 0.7261418700218201, "learning_rate": 5.50925925925926e-05, "loss": 0.0374, "step": 2500 }, { "grad_norm": 0.6458377838134766, "learning_rate": 5.531305114638449e-05, "loss": 0.0367, "step": 2510 }, { "grad_norm": 0.5536720156669617, "learning_rate": 5.5533509700176364e-05, "loss": 0.0327, "step": 2520 }, { "grad_norm": 0.4521956741809845, "learning_rate": 5.575396825396826e-05, "loss": 0.0322, "step": 2530 }, { "grad_norm": 0.5007415413856506, "learning_rate": 5.597442680776014e-05, "loss": 0.0373, "step": 2540 }, { "grad_norm": 0.7257701754570007, "learning_rate": 5.619488536155203e-05, "loss": 0.0424, "step": 2550 }, { "grad_norm": 0.6484182476997375, "learning_rate": 5.6415343915343924e-05, "loss": 0.0388, "step": 2560 }, { "grad_norm": 0.5204418301582336, "learning_rate": 5.66358024691358e-05, "loss": 0.0383, "step": 2570 }, { "grad_norm": 0.5221874713897705, "learning_rate": 5.685626102292769e-05, "loss": 0.0366, "step": 2580 }, { "grad_norm": 0.5247634649276733, "learning_rate": 5.707671957671958e-05, "loss": 0.0368, "step": 2590 }, { "grad_norm": 0.36856523156166077, "learning_rate": 5.7297178130511464e-05, "loss": 0.0317, "step": 2600 }, { "grad_norm": 0.5370011925697327, "learning_rate": 5.7517636684303354e-05, "loss": 0.0341, "step": 2610 }, { "grad_norm": 0.7906946539878845, "learning_rate": 5.773809523809524e-05, "loss": 0.0349, "step": 2620 }, { "grad_norm": 0.258973628282547, "learning_rate": 5.795855379188713e-05, "loss": 0.0328, "step": 2630 }, { "grad_norm": 0.6492506265640259, "learning_rate": 5.817901234567902e-05, "loss": 0.0317, "step": 2640 }, { "grad_norm": 0.5406093001365662, "learning_rate": 5.83994708994709e-05, "loss": 0.0371, "step": 2650 }, { "grad_norm": 0.4607742726802826, "learning_rate": 5.861992945326279e-05, "loss": 0.0442, "step": 2660 }, { "grad_norm": 0.4770677983760834, "learning_rate": 5.884038800705468e-05, "loss": 0.0352, "step": 2670 }, { "grad_norm": 0.6915843486785889, "learning_rate": 5.9060846560846564e-05, "loss": 0.041, "step": 2680 }, { "grad_norm": 0.4500851333141327, "learning_rate": 5.9281305114638454e-05, "loss": 0.0357, "step": 2690 }, { "grad_norm": 0.590701162815094, "learning_rate": 5.950176366843033e-05, "loss": 0.0372, "step": 2700 }, { "grad_norm": 0.5730337500572205, "learning_rate": 5.972222222222223e-05, "loss": 0.0401, "step": 2710 }, { "grad_norm": 0.3938755691051483, "learning_rate": 5.994268077601412e-05, "loss": 0.0365, "step": 2720 }, { "grad_norm": 0.43746277689933777, "learning_rate": 6.016313932980599e-05, "loss": 0.0356, "step": 2730 }, { "grad_norm": 0.3833053410053253, "learning_rate": 6.038359788359789e-05, "loss": 0.0319, "step": 2740 }, { "grad_norm": 0.4860956072807312, "learning_rate": 6.060405643738978e-05, "loss": 0.0315, "step": 2750 }, { "grad_norm": 0.4610036611557007, "learning_rate": 6.0824514991181656e-05, "loss": 0.0309, "step": 2760 }, { "grad_norm": 0.6027385592460632, "learning_rate": 6.104497354497355e-05, "loss": 0.0325, "step": 2770 }, { "grad_norm": 0.3683805465698242, "learning_rate": 6.126543209876542e-05, "loss": 0.0329, "step": 2780 }, { "grad_norm": 0.6089707016944885, "learning_rate": 6.148589065255732e-05, "loss": 0.031, "step": 2790 }, { "grad_norm": 0.4812466502189636, "learning_rate": 6.170634920634922e-05, "loss": 0.0332, "step": 2800 }, { "grad_norm": 0.4886496365070343, "learning_rate": 6.192680776014109e-05, "loss": 0.0368, "step": 2810 }, { "grad_norm": 0.45839378237724304, "learning_rate": 6.214726631393298e-05, "loss": 0.0291, "step": 2820 }, { "grad_norm": 0.6510221362113953, "learning_rate": 6.236772486772488e-05, "loss": 0.0292, "step": 2830 }, { "grad_norm": 0.4064684510231018, "learning_rate": 6.258818342151675e-05, "loss": 0.0336, "step": 2840 }, { "grad_norm": 0.6008754372596741, "learning_rate": 6.280864197530865e-05, "loss": 0.035, "step": 2850 }, { "grad_norm": 0.4285314381122589, "learning_rate": 6.302910052910054e-05, "loss": 0.0357, "step": 2860 }, { "grad_norm": 0.49698328971862793, "learning_rate": 6.324955908289241e-05, "loss": 0.0328, "step": 2870 }, { "grad_norm": 0.42896974086761475, "learning_rate": 6.347001763668431e-05, "loss": 0.0317, "step": 2880 }, { "grad_norm": 0.5287501811981201, "learning_rate": 6.369047619047619e-05, "loss": 0.03, "step": 2890 }, { "grad_norm": 0.676975429058075, "learning_rate": 6.391093474426808e-05, "loss": 0.0315, "step": 2900 }, { "grad_norm": 0.45902353525161743, "learning_rate": 6.413139329805997e-05, "loss": 0.0427, "step": 2910 }, { "grad_norm": 0.4295649230480194, "learning_rate": 6.435185185185186e-05, "loss": 0.0328, "step": 2920 }, { "grad_norm": 0.33384954929351807, "learning_rate": 6.457231040564374e-05, "loss": 0.032, "step": 2930 }, { "grad_norm": 0.34692829847335815, "learning_rate": 6.479276895943564e-05, "loss": 0.0325, "step": 2940 }, { "grad_norm": 0.5482107400894165, "learning_rate": 6.501322751322752e-05, "loss": 0.0312, "step": 2950 }, { "grad_norm": 0.4473950266838074, "learning_rate": 6.52336860670194e-05, "loss": 0.0375, "step": 2960 }, { "grad_norm": 0.3560939431190491, "learning_rate": 6.545414462081129e-05, "loss": 0.0326, "step": 2970 }, { "grad_norm": 0.4232942461967468, "learning_rate": 6.567460317460318e-05, "loss": 0.0293, "step": 2980 }, { "grad_norm": 0.39909809827804565, "learning_rate": 6.589506172839507e-05, "loss": 0.0302, "step": 2990 }, { "grad_norm": 0.5054033398628235, "learning_rate": 6.611552028218695e-05, "loss": 0.0386, "step": 3000 }, { "grad_norm": 0.5319614410400391, "learning_rate": 6.633597883597885e-05, "loss": 0.0328, "step": 3010 }, { "grad_norm": 0.518134593963623, "learning_rate": 6.655643738977073e-05, "loss": 0.0366, "step": 3020 }, { "grad_norm": 0.4065493047237396, "learning_rate": 6.677689594356261e-05, "loss": 0.0276, "step": 3030 }, { "grad_norm": 0.41736024618148804, "learning_rate": 6.699735449735451e-05, "loss": 0.028, "step": 3040 }, { "grad_norm": 0.4613221287727356, "learning_rate": 6.721781305114638e-05, "loss": 0.0341, "step": 3050 }, { "grad_norm": 0.2946951687335968, "learning_rate": 6.743827160493827e-05, "loss": 0.0324, "step": 3060 }, { "grad_norm": 0.5000953674316406, "learning_rate": 6.765873015873017e-05, "loss": 0.0359, "step": 3070 }, { "grad_norm": 0.4207751154899597, "learning_rate": 6.787918871252204e-05, "loss": 0.0308, "step": 3080 }, { "grad_norm": 0.30412307381629944, "learning_rate": 6.809964726631394e-05, "loss": 0.034, "step": 3090 }, { "grad_norm": 0.35921937227249146, "learning_rate": 6.832010582010583e-05, "loss": 0.0288, "step": 3100 }, { "grad_norm": 0.4605396091938019, "learning_rate": 6.85405643738977e-05, "loss": 0.0355, "step": 3110 }, { "grad_norm": 0.47208788990974426, "learning_rate": 6.87610229276896e-05, "loss": 0.0295, "step": 3120 }, { "grad_norm": 0.4570513069629669, "learning_rate": 6.898148148148148e-05, "loss": 0.0335, "step": 3130 }, { "grad_norm": 0.537216067314148, "learning_rate": 6.920194003527337e-05, "loss": 0.0353, "step": 3140 }, { "grad_norm": 0.3355444371700287, "learning_rate": 6.942239858906526e-05, "loss": 0.0362, "step": 3150 }, { "grad_norm": 0.5311845541000366, "learning_rate": 6.964285714285715e-05, "loss": 0.0293, "step": 3160 }, { "grad_norm": 0.5636377930641174, "learning_rate": 6.986331569664903e-05, "loss": 0.0328, "step": 3170 }, { "grad_norm": 0.3425399661064148, "learning_rate": 7.008377425044093e-05, "loss": 0.0286, "step": 3180 }, { "grad_norm": 0.43383997678756714, "learning_rate": 7.030423280423281e-05, "loss": 0.0305, "step": 3190 }, { "grad_norm": 0.4824570119380951, "learning_rate": 7.05246913580247e-05, "loss": 0.0345, "step": 3200 }, { "grad_norm": 0.4983125925064087, "learning_rate": 7.074514991181658e-05, "loss": 0.0322, "step": 3210 }, { "grad_norm": 0.36462217569351196, "learning_rate": 7.096560846560847e-05, "loss": 0.0269, "step": 3220 }, { "grad_norm": 0.5150817036628723, "learning_rate": 7.118606701940036e-05, "loss": 0.0284, "step": 3230 }, { "grad_norm": 0.2510686218738556, "learning_rate": 7.140652557319224e-05, "loss": 0.0267, "step": 3240 }, { "grad_norm": 0.41770869493484497, "learning_rate": 7.162698412698414e-05, "loss": 0.027, "step": 3250 }, { "grad_norm": 0.5390652418136597, "learning_rate": 7.184744268077602e-05, "loss": 0.0373, "step": 3260 }, { "grad_norm": 0.48923686146736145, "learning_rate": 7.20679012345679e-05, "loss": 0.0321, "step": 3270 }, { "grad_norm": 0.43046075105667114, "learning_rate": 7.22883597883598e-05, "loss": 0.0293, "step": 3280 }, { "grad_norm": 0.2752998471260071, "learning_rate": 7.250881834215167e-05, "loss": 0.0293, "step": 3290 }, { "grad_norm": 0.5566617250442505, "learning_rate": 7.272927689594357e-05, "loss": 0.0293, "step": 3300 }, { "grad_norm": 0.698322057723999, "learning_rate": 7.294973544973546e-05, "loss": 0.0304, "step": 3310 }, { "grad_norm": 0.46611639857292175, "learning_rate": 7.317019400352733e-05, "loss": 0.03, "step": 3320 }, { "grad_norm": 0.45372897386550903, "learning_rate": 7.339065255731923e-05, "loss": 0.0322, "step": 3330 }, { "grad_norm": 0.30704405903816223, "learning_rate": 7.361111111111111e-05, "loss": 0.0303, "step": 3340 }, { "grad_norm": 0.43227705359458923, "learning_rate": 7.3831569664903e-05, "loss": 0.0339, "step": 3350 }, { "grad_norm": 0.2767896354198456, "learning_rate": 7.405202821869489e-05, "loss": 0.0286, "step": 3360 }, { "grad_norm": 0.5246477127075195, "learning_rate": 7.427248677248678e-05, "loss": 0.0436, "step": 3370 }, { "grad_norm": 0.3208570182323456, "learning_rate": 7.449294532627866e-05, "loss": 0.033, "step": 3380 }, { "grad_norm": 0.3258393108844757, "learning_rate": 7.471340388007056e-05, "loss": 0.032, "step": 3390 }, { "grad_norm": 0.6356700658798218, "learning_rate": 7.493386243386244e-05, "loss": 0.0378, "step": 3400 }, { "grad_norm": 0.37633436918258667, "learning_rate": 7.515432098765432e-05, "loss": 0.0321, "step": 3410 }, { "grad_norm": 0.41480451822280884, "learning_rate": 7.537477954144622e-05, "loss": 0.0303, "step": 3420 }, { "grad_norm": 0.35956037044525146, "learning_rate": 7.55952380952381e-05, "loss": 0.0293, "step": 3430 }, { "grad_norm": 0.472808301448822, "learning_rate": 7.581569664902999e-05, "loss": 0.0322, "step": 3440 }, { "grad_norm": 0.34884876012802124, "learning_rate": 7.603615520282187e-05, "loss": 0.0262, "step": 3450 }, { "grad_norm": 0.4175480902194977, "learning_rate": 7.625661375661377e-05, "loss": 0.0323, "step": 3460 }, { "grad_norm": 0.3296893835067749, "learning_rate": 7.647707231040565e-05, "loss": 0.0261, "step": 3470 }, { "grad_norm": 0.4557775557041168, "learning_rate": 7.669753086419753e-05, "loss": 0.0311, "step": 3480 }, { "grad_norm": 0.3504185378551483, "learning_rate": 7.691798941798943e-05, "loss": 0.0301, "step": 3490 }, { "grad_norm": 0.4998333752155304, "learning_rate": 7.713844797178131e-05, "loss": 0.0377, "step": 3500 }, { "grad_norm": 0.4116503596305847, "learning_rate": 7.73589065255732e-05, "loss": 0.0317, "step": 3510 }, { "grad_norm": 0.4546218812465668, "learning_rate": 7.757936507936508e-05, "loss": 0.0303, "step": 3520 }, { "grad_norm": 0.37392115592956543, "learning_rate": 7.779982363315696e-05, "loss": 0.0294, "step": 3530 }, { "grad_norm": 0.3766471743583679, "learning_rate": 7.802028218694886e-05, "loss": 0.0351, "step": 3540 }, { "grad_norm": 0.5632163286209106, "learning_rate": 7.824074074074074e-05, "loss": 0.0299, "step": 3550 }, { "grad_norm": 0.3817369043827057, "learning_rate": 7.846119929453262e-05, "loss": 0.029, "step": 3560 }, { "grad_norm": 0.5337211489677429, "learning_rate": 7.868165784832452e-05, "loss": 0.034, "step": 3570 }, { "grad_norm": 0.42646437883377075, "learning_rate": 7.89021164021164e-05, "loss": 0.0318, "step": 3580 }, { "grad_norm": 0.45616504549980164, "learning_rate": 7.912257495590829e-05, "loss": 0.0288, "step": 3590 }, { "grad_norm": 0.3136594593524933, "learning_rate": 7.934303350970018e-05, "loss": 0.0326, "step": 3600 }, { "grad_norm": 0.3336106538772583, "learning_rate": 7.956349206349207e-05, "loss": 0.0258, "step": 3610 }, { "grad_norm": 0.328382670879364, "learning_rate": 7.978395061728395e-05, "loss": 0.0274, "step": 3620 }, { "grad_norm": 0.4112264811992645, "learning_rate": 8.000440917107585e-05, "loss": 0.0312, "step": 3630 }, { "grad_norm": 0.3936302661895752, "learning_rate": 8.022486772486773e-05, "loss": 0.033, "step": 3640 }, { "grad_norm": 0.4836055040359497, "learning_rate": 8.044532627865961e-05, "loss": 0.0339, "step": 3650 }, { "grad_norm": 0.23360410332679749, "learning_rate": 8.066578483245151e-05, "loss": 0.0271, "step": 3660 }, { "grad_norm": 0.36382749676704407, "learning_rate": 8.08862433862434e-05, "loss": 0.0297, "step": 3670 }, { "grad_norm": 0.41395890712738037, "learning_rate": 8.110670194003528e-05, "loss": 0.0312, "step": 3680 }, { "grad_norm": 0.22548474371433258, "learning_rate": 8.132716049382716e-05, "loss": 0.0349, "step": 3690 }, { "grad_norm": 0.3954477906227112, "learning_rate": 8.154761904761904e-05, "loss": 0.0294, "step": 3700 }, { "grad_norm": 0.30830076336860657, "learning_rate": 8.176807760141094e-05, "loss": 0.0323, "step": 3710 }, { "grad_norm": 0.23829622566699982, "learning_rate": 8.198853615520282e-05, "loss": 0.031, "step": 3720 }, { "grad_norm": 0.2872510552406311, "learning_rate": 8.220899470899471e-05, "loss": 0.0265, "step": 3730 }, { "grad_norm": 0.3808799684047699, "learning_rate": 8.24294532627866e-05, "loss": 0.031, "step": 3740 }, { "grad_norm": 0.43583858013153076, "learning_rate": 8.264991181657849e-05, "loss": 0.0319, "step": 3750 }, { "grad_norm": 0.20851323008537292, "learning_rate": 8.287037037037037e-05, "loss": 0.0282, "step": 3760 }, { "grad_norm": 0.32150620222091675, "learning_rate": 8.309082892416225e-05, "loss": 0.0332, "step": 3770 }, { "grad_norm": 0.33523237705230713, "learning_rate": 8.331128747795415e-05, "loss": 0.0286, "step": 3780 }, { "grad_norm": 0.27417242527008057, "learning_rate": 8.353174603174603e-05, "loss": 0.0251, "step": 3790 }, { "grad_norm": 0.2775508463382721, "learning_rate": 8.375220458553792e-05, "loss": 0.0312, "step": 3800 }, { "grad_norm": 0.3230460584163666, "learning_rate": 8.397266313932981e-05, "loss": 0.0245, "step": 3810 }, { "grad_norm": 0.21209140121936798, "learning_rate": 8.41931216931217e-05, "loss": 0.0291, "step": 3820 }, { "grad_norm": 0.29508453607559204, "learning_rate": 8.441358024691358e-05, "loss": 0.0304, "step": 3830 }, { "grad_norm": 0.34769919514656067, "learning_rate": 8.463403880070548e-05, "loss": 0.0268, "step": 3840 }, { "grad_norm": 0.3206215798854828, "learning_rate": 8.485449735449736e-05, "loss": 0.0282, "step": 3850 }, { "grad_norm": 0.35007739067077637, "learning_rate": 8.507495590828924e-05, "loss": 0.0319, "step": 3860 }, { "grad_norm": 0.43878090381622314, "learning_rate": 8.529541446208114e-05, "loss": 0.0283, "step": 3870 }, { "grad_norm": 0.33012616634368896, "learning_rate": 8.551587301587301e-05, "loss": 0.0309, "step": 3880 }, { "grad_norm": 0.4760889410972595, "learning_rate": 8.57363315696649e-05, "loss": 0.0285, "step": 3890 }, { "grad_norm": 0.5450194478034973, "learning_rate": 8.59567901234568e-05, "loss": 0.0319, "step": 3900 }, { "grad_norm": 0.37741073966026306, "learning_rate": 8.617724867724867e-05, "loss": 0.0283, "step": 3910 }, { "grad_norm": 0.40245258808135986, "learning_rate": 8.639770723104057e-05, "loss": 0.0306, "step": 3920 }, { "grad_norm": 0.3583366870880127, "learning_rate": 8.661816578483245e-05, "loss": 0.029, "step": 3930 }, { "grad_norm": 0.3080029785633087, "learning_rate": 8.683862433862434e-05, "loss": 0.0325, "step": 3940 }, { "grad_norm": 0.386202335357666, "learning_rate": 8.705908289241623e-05, "loss": 0.0335, "step": 3950 }, { "grad_norm": 0.4623749554157257, "learning_rate": 8.727954144620812e-05, "loss": 0.0279, "step": 3960 }, { "grad_norm": 0.43474286794662476, "learning_rate": 8.75e-05, "loss": 0.0327, "step": 3970 }, { "grad_norm": 0.3571276366710663, "learning_rate": 8.77204585537919e-05, "loss": 0.0268, "step": 3980 }, { "grad_norm": 0.37350979447364807, "learning_rate": 8.794091710758378e-05, "loss": 0.0279, "step": 3990 }, { "grad_norm": 0.3493541479110718, "learning_rate": 8.816137566137566e-05, "loss": 0.0294, "step": 4000 }, { "grad_norm": 0.36407196521759033, "learning_rate": 8.838183421516756e-05, "loss": 0.0297, "step": 4010 }, { "grad_norm": 0.291076123714447, "learning_rate": 8.860229276895944e-05, "loss": 0.0267, "step": 4020 }, { "grad_norm": 0.3058566153049469, "learning_rate": 8.882275132275133e-05, "loss": 0.0307, "step": 4030 }, { "grad_norm": 0.28948384523391724, "learning_rate": 8.904320987654321e-05, "loss": 0.032, "step": 4040 }, { "grad_norm": 0.3258579969406128, "learning_rate": 8.92636684303351e-05, "loss": 0.0267, "step": 4050 }, { "grad_norm": 0.31115788221359253, "learning_rate": 8.948412698412699e-05, "loss": 0.0308, "step": 4060 }, { "grad_norm": 0.2519753575325012, "learning_rate": 8.970458553791887e-05, "loss": 0.0315, "step": 4070 }, { "grad_norm": 0.2787935435771942, "learning_rate": 8.992504409171077e-05, "loss": 0.0277, "step": 4080 }, { "grad_norm": 0.46763861179351807, "learning_rate": 9.014550264550265e-05, "loss": 0.028, "step": 4090 }, { "grad_norm": 0.2785889208316803, "learning_rate": 9.036596119929454e-05, "loss": 0.0318, "step": 4100 }, { "grad_norm": 0.3804284334182739, "learning_rate": 9.058641975308643e-05, "loss": 0.0301, "step": 4110 }, { "grad_norm": 0.34066805243492126, "learning_rate": 9.08068783068783e-05, "loss": 0.0281, "step": 4120 }, { "grad_norm": 0.2528081238269806, "learning_rate": 9.10273368606702e-05, "loss": 0.0282, "step": 4130 }, { "grad_norm": 0.2751939296722412, "learning_rate": 9.12477954144621e-05, "loss": 0.0335, "step": 4140 }, { "grad_norm": 0.3156932592391968, "learning_rate": 9.146825396825396e-05, "loss": 0.0292, "step": 4150 }, { "grad_norm": 0.3844470977783203, "learning_rate": 9.168871252204586e-05, "loss": 0.0267, "step": 4160 }, { "grad_norm": 0.5035740733146667, "learning_rate": 9.190917107583776e-05, "loss": 0.0347, "step": 4170 }, { "grad_norm": 0.33674898743629456, "learning_rate": 9.212962962962963e-05, "loss": 0.0344, "step": 4180 }, { "grad_norm": 0.22722302377223969, "learning_rate": 9.235008818342152e-05, "loss": 0.0312, "step": 4190 }, { "grad_norm": 0.3420249819755554, "learning_rate": 9.257054673721341e-05, "loss": 0.0261, "step": 4200 }, { "grad_norm": 0.2761017978191376, "learning_rate": 9.279100529100529e-05, "loss": 0.0243, "step": 4210 }, { "grad_norm": 0.3194929361343384, "learning_rate": 9.301146384479719e-05, "loss": 0.0262, "step": 4220 }, { "grad_norm": 0.2260126918554306, "learning_rate": 9.323192239858907e-05, "loss": 0.025, "step": 4230 }, { "grad_norm": 0.31774210929870605, "learning_rate": 9.345238095238095e-05, "loss": 0.0319, "step": 4240 }, { "grad_norm": 0.3132788836956024, "learning_rate": 9.367283950617285e-05, "loss": 0.0322, "step": 4250 }, { "grad_norm": 0.3816041052341461, "learning_rate": 9.389329805996473e-05, "loss": 0.0335, "step": 4260 }, { "grad_norm": 0.23684382438659668, "learning_rate": 9.411375661375662e-05, "loss": 0.0301, "step": 4270 }, { "grad_norm": 0.27095678448677063, "learning_rate": 9.43342151675485e-05, "loss": 0.0303, "step": 4280 }, { "grad_norm": 0.30275025963783264, "learning_rate": 9.45546737213404e-05, "loss": 0.0262, "step": 4290 }, { "grad_norm": 0.3106994032859802, "learning_rate": 9.477513227513228e-05, "loss": 0.0273, "step": 4300 }, { "grad_norm": 0.37933823466300964, "learning_rate": 9.499559082892416e-05, "loss": 0.0257, "step": 4310 }, { "grad_norm": 0.34993776679039, "learning_rate": 9.521604938271606e-05, "loss": 0.0293, "step": 4320 }, { "grad_norm": 0.2474062442779541, "learning_rate": 9.543650793650794e-05, "loss": 0.0267, "step": 4330 }, { "grad_norm": 0.388630211353302, "learning_rate": 9.565696649029983e-05, "loss": 0.0306, "step": 4340 }, { "grad_norm": 0.34848752617836, "learning_rate": 9.587742504409172e-05, "loss": 0.0282, "step": 4350 }, { "grad_norm": 0.3217245042324066, "learning_rate": 9.60978835978836e-05, "loss": 0.0264, "step": 4360 }, { "grad_norm": 0.2863377034664154, "learning_rate": 9.631834215167549e-05, "loss": 0.023, "step": 4370 }, { "grad_norm": 0.21386080980300903, "learning_rate": 9.653880070546739e-05, "loss": 0.0261, "step": 4380 }, { "grad_norm": 0.3120568096637726, "learning_rate": 9.675925925925926e-05, "loss": 0.03, "step": 4390 }, { "grad_norm": 0.39826852083206177, "learning_rate": 9.697971781305115e-05, "loss": 0.0343, "step": 4400 }, { "grad_norm": 0.28577497601509094, "learning_rate": 9.720017636684304e-05, "loss": 0.0284, "step": 4410 }, { "grad_norm": 0.2567664086818695, "learning_rate": 9.742063492063492e-05, "loss": 0.0302, "step": 4420 }, { "grad_norm": 0.2691606283187866, "learning_rate": 9.764109347442682e-05, "loss": 0.0288, "step": 4430 }, { "grad_norm": 0.21750569343566895, "learning_rate": 9.78615520282187e-05, "loss": 0.0345, "step": 4440 }, { "grad_norm": 0.2635342478752136, "learning_rate": 9.808201058201058e-05, "loss": 0.0229, "step": 4450 }, { "grad_norm": 0.3777622580528259, "learning_rate": 9.830246913580248e-05, "loss": 0.028, "step": 4460 }, { "grad_norm": 0.3233034610748291, "learning_rate": 9.852292768959436e-05, "loss": 0.0272, "step": 4470 }, { "grad_norm": 0.18069671094417572, "learning_rate": 9.874338624338625e-05, "loss": 0.0263, "step": 4480 }, { "grad_norm": 0.41449233889579773, "learning_rate": 9.896384479717814e-05, "loss": 0.0287, "step": 4490 }, { "grad_norm": 0.39910176396369934, "learning_rate": 9.918430335097003e-05, "loss": 0.0291, "step": 4500 }, { "grad_norm": 0.37015393376350403, "learning_rate": 9.940476190476191e-05, "loss": 0.0354, "step": 4510 }, { "grad_norm": 0.2943476736545563, "learning_rate": 9.962522045855379e-05, "loss": 0.0282, "step": 4520 }, { "grad_norm": 0.23606519401073456, "learning_rate": 9.984567901234569e-05, "loss": 0.0282, "step": 4530 }, { "grad_norm": 0.2693200707435608, "learning_rate": 9.999999970092498e-05, "loss": 0.0243, "step": 4540 }, { "grad_norm": 0.30157095193862915, "learning_rate": 9.999999438403562e-05, "loss": 0.0263, "step": 4550 }, { "grad_norm": 0.45228511095046997, "learning_rate": 9.999998242103527e-05, "loss": 0.0258, "step": 4560 }, { "grad_norm": 0.3370663821697235, "learning_rate": 9.999996381192551e-05, "loss": 0.0288, "step": 4570 }, { "grad_norm": 0.3403978645801544, "learning_rate": 9.999993855670882e-05, "loss": 0.0254, "step": 4580 }, { "grad_norm": 0.2614413797855377, "learning_rate": 9.999990665538856e-05, "loss": 0.0251, "step": 4590 }, { "grad_norm": 0.2987765967845917, "learning_rate": 9.999986810796895e-05, "loss": 0.0264, "step": 4600 }, { "grad_norm": 0.27884024381637573, "learning_rate": 9.999982291445513e-05, "loss": 0.0266, "step": 4610 }, { "grad_norm": 0.19106809794902802, "learning_rate": 9.999977107485313e-05, "loss": 0.031, "step": 4620 }, { "grad_norm": 0.4677678048610687, "learning_rate": 9.999971258916978e-05, "loss": 0.0313, "step": 4630 }, { "grad_norm": 0.2854192554950714, "learning_rate": 9.999964745741291e-05, "loss": 0.0295, "step": 4640 }, { "grad_norm": 0.3738241493701935, "learning_rate": 9.999957567959113e-05, "loss": 0.0297, "step": 4650 }, { "grad_norm": 0.27395254373550415, "learning_rate": 9.999949725571402e-05, "loss": 0.0392, "step": 4660 }, { "grad_norm": 0.36524292826652527, "learning_rate": 9.9999412185792e-05, "loss": 0.0297, "step": 4670 }, { "grad_norm": 0.3077434003353119, "learning_rate": 9.999932046983635e-05, "loss": 0.023, "step": 4680 }, { "grad_norm": 0.2249855250120163, "learning_rate": 9.99992221078593e-05, "loss": 0.0322, "step": 4690 }, { "grad_norm": 0.29686838388442993, "learning_rate": 9.999911709987391e-05, "loss": 0.0318, "step": 4700 }, { "grad_norm": 0.28234079480171204, "learning_rate": 9.999900544589412e-05, "loss": 0.0272, "step": 4710 }, { "grad_norm": 0.26331669092178345, "learning_rate": 9.999888714593477e-05, "loss": 0.0313, "step": 4720 }, { "grad_norm": 0.3140276074409485, "learning_rate": 9.999876220001162e-05, "loss": 0.0269, "step": 4730 }, { "grad_norm": 0.38838621973991394, "learning_rate": 9.999863060814125e-05, "loss": 0.0272, "step": 4740 }, { "grad_norm": 0.26013314723968506, "learning_rate": 9.999849237034115e-05, "loss": 0.0261, "step": 4750 }, { "grad_norm": 0.24803310632705688, "learning_rate": 9.999834748662971e-05, "loss": 0.0248, "step": 4760 }, { "grad_norm": 0.17012913525104523, "learning_rate": 9.999819595702619e-05, "loss": 0.0247, "step": 4770 }, { "grad_norm": 0.37854474782943726, "learning_rate": 9.99980377815507e-05, "loss": 0.0269, "step": 4780 }, { "grad_norm": 0.2810017466545105, "learning_rate": 9.999787296022432e-05, "loss": 0.0266, "step": 4790 }, { "grad_norm": 0.23475205898284912, "learning_rate": 9.99977014930689e-05, "loss": 0.0303, "step": 4800 }, { "grad_norm": 0.27067941427230835, "learning_rate": 9.999752338010726e-05, "loss": 0.0257, "step": 4810 }, { "grad_norm": 0.248803049325943, "learning_rate": 9.999733862136308e-05, "loss": 0.0256, "step": 4820 }, { "grad_norm": 0.39793941378593445, "learning_rate": 9.99971472168609e-05, "loss": 0.0334, "step": 4830 }, { "grad_norm": 0.30359119176864624, "learning_rate": 9.999694916662619e-05, "loss": 0.035, "step": 4840 }, { "grad_norm": 0.3373635709285736, "learning_rate": 9.999674447068525e-05, "loss": 0.0295, "step": 4850 }, { "grad_norm": 0.270368367433548, "learning_rate": 9.999653312906532e-05, "loss": 0.0271, "step": 4860 }, { "grad_norm": 0.3919520974159241, "learning_rate": 9.999631514179444e-05, "loss": 0.0242, "step": 4870 }, { "grad_norm": 0.29250219464302063, "learning_rate": 9.999609050890163e-05, "loss": 0.0308, "step": 4880 }, { "grad_norm": 0.25735899806022644, "learning_rate": 9.999585923041672e-05, "loss": 0.0287, "step": 4890 }, { "grad_norm": 0.17462670803070068, "learning_rate": 9.999562130637048e-05, "loss": 0.0263, "step": 4900 }, { "grad_norm": 0.2160707712173462, "learning_rate": 9.999537673679452e-05, "loss": 0.0238, "step": 4910 }, { "grad_norm": 0.2890188694000244, "learning_rate": 9.999512552172135e-05, "loss": 0.0236, "step": 4920 }, { "grad_norm": 0.2721152901649475, "learning_rate": 9.999486766118436e-05, "loss": 0.0271, "step": 4930 }, { "grad_norm": 0.30923956632614136, "learning_rate": 9.999460315521782e-05, "loss": 0.026, "step": 4940 }, { "grad_norm": 0.3883291184902191, "learning_rate": 9.999433200385691e-05, "loss": 0.0298, "step": 4950 }, { "grad_norm": 0.27644261717796326, "learning_rate": 9.999405420713764e-05, "loss": 0.0235, "step": 4960 }, { "grad_norm": 0.22725538909435272, "learning_rate": 9.999376976509696e-05, "loss": 0.0312, "step": 4970 }, { "grad_norm": 0.32390880584716797, "learning_rate": 9.999347867777269e-05, "loss": 0.0307, "step": 4980 }, { "grad_norm": 0.20231372117996216, "learning_rate": 9.999318094520349e-05, "loss": 0.0272, "step": 4990 }, { "grad_norm": 0.36317533254623413, "learning_rate": 9.999287656742895e-05, "loss": 0.0292, "step": 5000 }, { "grad_norm": 0.40086036920547485, "learning_rate": 9.999256554448952e-05, "loss": 0.0302, "step": 5010 }, { "grad_norm": 0.33906540274620056, "learning_rate": 9.999224787642657e-05, "loss": 0.029, "step": 5020 }, { "grad_norm": 0.35007861256599426, "learning_rate": 9.999192356328229e-05, "loss": 0.0283, "step": 5030 }, { "grad_norm": 0.1871664673089981, "learning_rate": 9.99915926050998e-05, "loss": 0.0273, "step": 5040 }, { "grad_norm": 0.25610899925231934, "learning_rate": 9.99912550019231e-05, "loss": 0.0287, "step": 5050 }, { "grad_norm": 0.3225916624069214, "learning_rate": 9.999091075379706e-05, "loss": 0.0268, "step": 5060 }, { "grad_norm": 0.29616814851760864, "learning_rate": 9.999055986076744e-05, "loss": 0.0218, "step": 5070 }, { "grad_norm": 0.2695976495742798, "learning_rate": 9.999020232288087e-05, "loss": 0.0277, "step": 5080 }, { "grad_norm": 0.30051907896995544, "learning_rate": 9.99898381401849e-05, "loss": 0.0258, "step": 5090 }, { "grad_norm": 0.2591812014579773, "learning_rate": 9.998946731272791e-05, "loss": 0.0297, "step": 5100 }, { "grad_norm": 0.18123438954353333, "learning_rate": 9.998908984055921e-05, "loss": 0.0258, "step": 5110 }, { "grad_norm": 0.3043799102306366, "learning_rate": 9.998870572372897e-05, "loss": 0.0345, "step": 5120 }, { "grad_norm": 0.24570991098880768, "learning_rate": 9.998831496228824e-05, "loss": 0.0247, "step": 5130 }, { "grad_norm": 0.35273098945617676, "learning_rate": 9.998791755628898e-05, "loss": 0.0273, "step": 5140 }, { "grad_norm": 0.30366250872612, "learning_rate": 9.998751350578399e-05, "loss": 0.0262, "step": 5150 }, { "grad_norm": 0.33098113536834717, "learning_rate": 9.998710281082698e-05, "loss": 0.0267, "step": 5160 }, { "grad_norm": 0.18093867599964142, "learning_rate": 9.998668547147257e-05, "loss": 0.0221, "step": 5170 }, { "grad_norm": 0.29693710803985596, "learning_rate": 9.998626148777619e-05, "loss": 0.0258, "step": 5180 }, { "grad_norm": 0.4045669436454773, "learning_rate": 9.998583085979423e-05, "loss": 0.0279, "step": 5190 }, { "grad_norm": 0.30083319544792175, "learning_rate": 9.998539358758392e-05, "loss": 0.0324, "step": 5200 }, { "grad_norm": 0.28027957677841187, "learning_rate": 9.998494967120338e-05, "loss": 0.0293, "step": 5210 }, { "grad_norm": 0.30479350686073303, "learning_rate": 9.998449911071162e-05, "loss": 0.0285, "step": 5220 }, { "grad_norm": 0.2328883409500122, "learning_rate": 9.998404190616853e-05, "loss": 0.0259, "step": 5230 }, { "grad_norm": 0.22633248567581177, "learning_rate": 9.998357805763488e-05, "loss": 0.024, "step": 5240 }, { "grad_norm": 0.18916067481040955, "learning_rate": 9.998310756517231e-05, "loss": 0.0205, "step": 5250 }, { "grad_norm": 0.3042478859424591, "learning_rate": 9.99826304288434e-05, "loss": 0.029, "step": 5260 }, { "grad_norm": 0.15191003680229187, "learning_rate": 9.998214664871155e-05, "loss": 0.0277, "step": 5270 }, { "grad_norm": 0.27522599697113037, "learning_rate": 9.998165622484103e-05, "loss": 0.0266, "step": 5280 }, { "grad_norm": 0.19681106507778168, "learning_rate": 9.998115915729709e-05, "loss": 0.0259, "step": 5290 }, { "grad_norm": 0.19915805757045746, "learning_rate": 9.998065544614577e-05, "loss": 0.025, "step": 5300 }, { "grad_norm": 0.18867741525173187, "learning_rate": 9.998014509145402e-05, "loss": 0.0258, "step": 5310 }, { "grad_norm": 0.263865202665329, "learning_rate": 9.997962809328969e-05, "loss": 0.028, "step": 5320 }, { "grad_norm": 0.19892366230487823, "learning_rate": 9.997910445172148e-05, "loss": 0.0247, "step": 5330 }, { "grad_norm": 0.28076764941215515, "learning_rate": 9.997857416681901e-05, "loss": 0.0288, "step": 5340 }, { "grad_norm": 0.25362300872802734, "learning_rate": 9.997803723865279e-05, "loss": 0.0249, "step": 5350 }, { "grad_norm": 0.21423280239105225, "learning_rate": 9.997749366729415e-05, "loss": 0.0263, "step": 5360 }, { "grad_norm": 0.23283863067626953, "learning_rate": 9.997694345281535e-05, "loss": 0.028, "step": 5370 }, { "grad_norm": 0.15979328751564026, "learning_rate": 9.997638659528953e-05, "loss": 0.0268, "step": 5380 }, { "grad_norm": 0.23935280740261078, "learning_rate": 9.997582309479073e-05, "loss": 0.0248, "step": 5390 }, { "grad_norm": 0.2979908287525177, "learning_rate": 9.997525295139383e-05, "loss": 0.0261, "step": 5400 }, { "grad_norm": 0.42018577456474304, "learning_rate": 9.997467616517462e-05, "loss": 0.0307, "step": 5410 }, { "grad_norm": 0.33350685238838196, "learning_rate": 9.997409273620975e-05, "loss": 0.0227, "step": 5420 }, { "grad_norm": 0.2492145448923111, "learning_rate": 9.997350266457679e-05, "loss": 0.0257, "step": 5430 }, { "grad_norm": 0.21456393599510193, "learning_rate": 9.997290595035418e-05, "loss": 0.0233, "step": 5440 }, { "grad_norm": 0.3385896384716034, "learning_rate": 9.997230259362123e-05, "loss": 0.0329, "step": 5450 }, { "grad_norm": 0.1732434779405594, "learning_rate": 9.997169259445814e-05, "loss": 0.0231, "step": 5460 }, { "grad_norm": 0.21403250098228455, "learning_rate": 9.997107595294597e-05, "loss": 0.0253, "step": 5470 }, { "grad_norm": 0.15211467444896698, "learning_rate": 9.997045266916671e-05, "loss": 0.0241, "step": 5480 }, { "grad_norm": 0.14283227920532227, "learning_rate": 9.996982274320321e-05, "loss": 0.0208, "step": 5490 }, { "grad_norm": 0.2873881757259369, "learning_rate": 9.996918617513918e-05, "loss": 0.0216, "step": 5500 }, { "grad_norm": 0.2697436213493347, "learning_rate": 9.996854296505927e-05, "loss": 0.0268, "step": 5510 }, { "grad_norm": 0.1592351347208023, "learning_rate": 9.996789311304894e-05, "loss": 0.0193, "step": 5520 }, { "grad_norm": 0.2895340919494629, "learning_rate": 9.996723661919459e-05, "loss": 0.0322, "step": 5530 }, { "grad_norm": 0.2646966874599457, "learning_rate": 9.996657348358347e-05, "loss": 0.0241, "step": 5540 }, { "grad_norm": 0.2528928220272064, "learning_rate": 9.996590370630374e-05, "loss": 0.0203, "step": 5550 }, { "grad_norm": 0.33302372694015503, "learning_rate": 9.996522728744442e-05, "loss": 0.0246, "step": 5560 }, { "grad_norm": 0.33563536405563354, "learning_rate": 9.996454422709542e-05, "loss": 0.023, "step": 5570 }, { "grad_norm": 0.24636615812778473, "learning_rate": 9.996385452534753e-05, "loss": 0.0242, "step": 5580 }, { "grad_norm": 0.2749592661857605, "learning_rate": 9.996315818229244e-05, "loss": 0.0238, "step": 5590 }, { "grad_norm": 0.27457544207572937, "learning_rate": 9.99624551980227e-05, "loss": 0.0262, "step": 5600 }, { "grad_norm": 0.44096410274505615, "learning_rate": 9.996174557263176e-05, "loss": 0.0295, "step": 5610 }, { "grad_norm": 0.24932125210762024, "learning_rate": 9.996102930621394e-05, "loss": 0.0276, "step": 5620 }, { "grad_norm": 0.2755689024925232, "learning_rate": 9.996030639886444e-05, "loss": 0.0261, "step": 5630 }, { "grad_norm": 0.2611924409866333, "learning_rate": 9.995957685067936e-05, "loss": 0.0306, "step": 5640 }, { "grad_norm": 0.24760571122169495, "learning_rate": 9.995884066175566e-05, "loss": 0.0198, "step": 5650 }, { "grad_norm": 0.17219536006450653, "learning_rate": 9.995809783219121e-05, "loss": 0.0283, "step": 5660 }, { "grad_norm": 0.15443222224712372, "learning_rate": 9.995734836208476e-05, "loss": 0.0252, "step": 5670 }, { "grad_norm": 0.26081785559654236, "learning_rate": 9.99565922515359e-05, "loss": 0.0245, "step": 5680 }, { "grad_norm": 0.3308599293231964, "learning_rate": 9.995582950064517e-05, "loss": 0.0272, "step": 5690 }, { "grad_norm": 0.35326841473579407, "learning_rate": 9.995506010951392e-05, "loss": 0.0214, "step": 5700 }, { "grad_norm": 0.22491246461868286, "learning_rate": 9.995428407824445e-05, "loss": 0.0256, "step": 5710 }, { "grad_norm": 0.248326376080513, "learning_rate": 9.995350140693989e-05, "loss": 0.0268, "step": 5720 }, { "grad_norm": 0.27141961455345154, "learning_rate": 9.995271209570427e-05, "loss": 0.0305, "step": 5730 }, { "grad_norm": 0.309098482131958, "learning_rate": 9.995191614464255e-05, "loss": 0.0241, "step": 5740 }, { "grad_norm": 0.18473894894123077, "learning_rate": 9.995111355386047e-05, "loss": 0.0226, "step": 5750 }, { "grad_norm": 0.30636826157569885, "learning_rate": 9.995030432346475e-05, "loss": 0.0245, "step": 5760 }, { "grad_norm": 0.2096962332725525, "learning_rate": 9.994948845356294e-05, "loss": 0.0241, "step": 5770 }, { "grad_norm": 0.17710116505622864, "learning_rate": 9.99486659442635e-05, "loss": 0.0269, "step": 5780 }, { "grad_norm": 0.24542704224586487, "learning_rate": 9.994783679567576e-05, "loss": 0.0284, "step": 5790 }, { "grad_norm": 0.31881365180015564, "learning_rate": 9.994700100790993e-05, "loss": 0.0249, "step": 5800 }, { "grad_norm": 0.1968272626399994, "learning_rate": 9.994615858107707e-05, "loss": 0.0235, "step": 5810 }, { "grad_norm": 0.31535688042640686, "learning_rate": 9.99453095152892e-05, "loss": 0.0214, "step": 5820 }, { "grad_norm": 0.2599996328353882, "learning_rate": 9.994445381065918e-05, "loss": 0.0293, "step": 5830 }, { "grad_norm": 0.2572386562824249, "learning_rate": 9.994359146730073e-05, "loss": 0.025, "step": 5840 }, { "grad_norm": 0.2859001159667969, "learning_rate": 9.99427224853285e-05, "loss": 0.0314, "step": 5850 }, { "grad_norm": 0.1686020940542221, "learning_rate": 9.994184686485794e-05, "loss": 0.0256, "step": 5860 }, { "grad_norm": 0.15978603065013885, "learning_rate": 9.994096460600553e-05, "loss": 0.0223, "step": 5870 }, { "grad_norm": 0.23577526211738586, "learning_rate": 9.994007570888845e-05, "loss": 0.033, "step": 5880 }, { "grad_norm": 0.17041218280792236, "learning_rate": 9.993918017362491e-05, "loss": 0.0221, "step": 5890 }, { "grad_norm": 0.23909159004688263, "learning_rate": 9.993827800033395e-05, "loss": 0.0274, "step": 5900 }, { "grad_norm": 0.22166059911251068, "learning_rate": 9.993736918913546e-05, "loss": 0.0262, "step": 5910 }, { "grad_norm": 0.2614532709121704, "learning_rate": 9.993645374015024e-05, "loss": 0.028, "step": 5920 }, { "grad_norm": 0.2850078344345093, "learning_rate": 9.99355316535e-05, "loss": 0.0218, "step": 5930 }, { "grad_norm": 0.176982119679451, "learning_rate": 9.993460292930729e-05, "loss": 0.0237, "step": 5940 }, { "grad_norm": 0.25900301337242126, "learning_rate": 9.993366756769556e-05, "loss": 0.0235, "step": 5950 }, { "grad_norm": 0.2082744985818863, "learning_rate": 9.993272556878914e-05, "loss": 0.024, "step": 5960 }, { "grad_norm": 0.14169947803020477, "learning_rate": 9.993177693271324e-05, "loss": 0.0208, "step": 5970 }, { "grad_norm": 0.2387237399816513, "learning_rate": 9.993082165959395e-05, "loss": 0.0249, "step": 5980 }, { "grad_norm": 0.28521203994750977, "learning_rate": 9.992985974955827e-05, "loss": 0.0223, "step": 5990 }, { "grad_norm": 0.16798384487628937, "learning_rate": 9.992889120273406e-05, "loss": 0.0242, "step": 6000 }, { "grad_norm": 0.32389378547668457, "learning_rate": 9.992791601925002e-05, "loss": 0.0227, "step": 6010 }, { "grad_norm": 0.28837573528289795, "learning_rate": 9.99269341992358e-05, "loss": 0.0243, "step": 6020 }, { "grad_norm": 0.20296016335487366, "learning_rate": 9.992594574282191e-05, "loss": 0.0184, "step": 6030 }, { "grad_norm": 0.17123167216777802, "learning_rate": 9.992495065013973e-05, "loss": 0.0255, "step": 6040 }, { "grad_norm": 0.18195882439613342, "learning_rate": 9.992394892132154e-05, "loss": 0.0224, "step": 6050 }, { "grad_norm": 0.23930130898952484, "learning_rate": 9.992294055650046e-05, "loss": 0.0221, "step": 6060 }, { "grad_norm": 0.27495548129081726, "learning_rate": 9.992192555581058e-05, "loss": 0.0257, "step": 6070 }, { "grad_norm": 0.20902997255325317, "learning_rate": 9.992090391938677e-05, "loss": 0.028, "step": 6080 }, { "grad_norm": 0.2500196695327759, "learning_rate": 9.991987564736484e-05, "loss": 0.0254, "step": 6090 }, { "grad_norm": 0.3068868815898895, "learning_rate": 9.99188407398815e-05, "loss": 0.0201, "step": 6100 }, { "grad_norm": 0.1993526816368103, "learning_rate": 9.991779919707425e-05, "loss": 0.0207, "step": 6110 }, { "grad_norm": 0.2679353952407837, "learning_rate": 9.99167510190816e-05, "loss": 0.0244, "step": 6120 }, { "grad_norm": 0.17651771008968353, "learning_rate": 9.991569620604283e-05, "loss": 0.0244, "step": 6130 }, { "grad_norm": 0.14901167154312134, "learning_rate": 9.991463475809817e-05, "loss": 0.0197, "step": 6140 }, { "grad_norm": 0.24463613331317902, "learning_rate": 9.991356667538869e-05, "loss": 0.0209, "step": 6150 }, { "grad_norm": 0.21300193667411804, "learning_rate": 9.991249195805639e-05, "loss": 0.0249, "step": 6160 }, { "grad_norm": 0.12934879958629608, "learning_rate": 9.99114106062441e-05, "loss": 0.024, "step": 6170 }, { "grad_norm": 0.1906438022851944, "learning_rate": 9.991032262009558e-05, "loss": 0.0233, "step": 6180 }, { "grad_norm": 0.21864426136016846, "learning_rate": 9.990922799975542e-05, "loss": 0.0263, "step": 6190 }, { "grad_norm": 0.21605359017848969, "learning_rate": 9.990812674536913e-05, "loss": 0.0229, "step": 6200 }, { "grad_norm": 0.2519116997718811, "learning_rate": 9.990701885708311e-05, "loss": 0.0251, "step": 6210 }, { "grad_norm": 0.19187623262405396, "learning_rate": 9.990590433504459e-05, "loss": 0.0235, "step": 6220 }, { "grad_norm": 0.23193465173244476, "learning_rate": 9.990478317940175e-05, "loss": 0.0247, "step": 6230 }, { "grad_norm": 0.19324029982089996, "learning_rate": 9.990365539030359e-05, "loss": 0.0234, "step": 6240 }, { "grad_norm": 0.2840346395969391, "learning_rate": 9.990252096790002e-05, "loss": 0.0253, "step": 6250 }, { "grad_norm": 0.22846174240112305, "learning_rate": 9.990137991234184e-05, "loss": 0.0249, "step": 6260 }, { "grad_norm": 0.18812412023544312, "learning_rate": 9.990023222378071e-05, "loss": 0.023, "step": 6270 }, { "grad_norm": 0.2007352113723755, "learning_rate": 9.989907790236921e-05, "loss": 0.0219, "step": 6280 }, { "grad_norm": 0.23102955520153046, "learning_rate": 9.989791694826074e-05, "loss": 0.0254, "step": 6290 }, { "grad_norm": 0.3354489505290985, "learning_rate": 9.989674936160964e-05, "loss": 0.022, "step": 6300 }, { "grad_norm": 0.21505333483219147, "learning_rate": 9.989557514257109e-05, "loss": 0.0221, "step": 6310 }, { "grad_norm": 0.21754711866378784, "learning_rate": 9.989439429130119e-05, "loss": 0.0221, "step": 6320 }, { "grad_norm": 0.1794452667236328, "learning_rate": 9.989320680795691e-05, "loss": 0.0216, "step": 6330 }, { "grad_norm": 0.2474789023399353, "learning_rate": 9.989201269269604e-05, "loss": 0.0231, "step": 6340 }, { "grad_norm": 0.12894026935100555, "learning_rate": 9.989081194567737e-05, "loss": 0.0182, "step": 6350 }, { "grad_norm": 0.20950670540332794, "learning_rate": 9.988960456706045e-05, "loss": 0.0231, "step": 6360 }, { "grad_norm": 0.1646740883588791, "learning_rate": 9.98883905570058e-05, "loss": 0.0252, "step": 6370 }, { "grad_norm": 0.3448674976825714, "learning_rate": 9.988716991567479e-05, "loss": 0.0217, "step": 6380 }, { "grad_norm": 0.28670087456703186, "learning_rate": 9.988594264322966e-05, "loss": 0.0242, "step": 6390 }, { "grad_norm": 0.2414359152317047, "learning_rate": 9.988470873983355e-05, "loss": 0.0247, "step": 6400 }, { "grad_norm": 0.2534656524658203, "learning_rate": 9.988346820565047e-05, "loss": 0.0232, "step": 6410 }, { "grad_norm": 0.2128477245569229, "learning_rate": 9.98822210408453e-05, "loss": 0.0242, "step": 6420 }, { "grad_norm": 0.153532937169075, "learning_rate": 9.988096724558382e-05, "loss": 0.0181, "step": 6430 }, { "grad_norm": 0.2845677435398102, "learning_rate": 9.987970682003272e-05, "loss": 0.0241, "step": 6440 }, { "grad_norm": 0.20618702471256256, "learning_rate": 9.987843976435948e-05, "loss": 0.0193, "step": 6450 }, { "grad_norm": 0.2725130021572113, "learning_rate": 9.987716607873259e-05, "loss": 0.0213, "step": 6460 }, { "grad_norm": 0.14160636067390442, "learning_rate": 9.98758857633213e-05, "loss": 0.0237, "step": 6470 }, { "grad_norm": 0.1773141473531723, "learning_rate": 9.98745988182958e-05, "loss": 0.0193, "step": 6480 }, { "grad_norm": 0.2235368937253952, "learning_rate": 9.987330524382716e-05, "loss": 0.0218, "step": 6490 }, { "grad_norm": 0.24228239059448242, "learning_rate": 9.987200504008733e-05, "loss": 0.0216, "step": 6500 }, { "grad_norm": 0.2852589786052704, "learning_rate": 9.987069820724914e-05, "loss": 0.0235, "step": 6510 }, { "grad_norm": 0.24002377688884735, "learning_rate": 9.986938474548629e-05, "loss": 0.0246, "step": 6520 }, { "grad_norm": 0.19459016621112823, "learning_rate": 9.986806465497336e-05, "loss": 0.0242, "step": 6530 }, { "grad_norm": 0.25737935304641724, "learning_rate": 9.986673793588581e-05, "loss": 0.0219, "step": 6540 }, { "grad_norm": 0.26740702986717224, "learning_rate": 9.986540458840003e-05, "loss": 0.0208, "step": 6550 }, { "grad_norm": 0.2798624634742737, "learning_rate": 9.986406461269323e-05, "loss": 0.0224, "step": 6560 }, { "grad_norm": 0.21678535640239716, "learning_rate": 9.986271800894351e-05, "loss": 0.0277, "step": 6570 }, { "grad_norm": 0.25450870394706726, "learning_rate": 9.986136477732987e-05, "loss": 0.0311, "step": 6580 }, { "grad_norm": 0.18609493970870972, "learning_rate": 9.986000491803219e-05, "loss": 0.0231, "step": 6590 }, { "grad_norm": 0.2132185399532318, "learning_rate": 9.985863843123123e-05, "loss": 0.0212, "step": 6600 }, { "grad_norm": 0.28147464990615845, "learning_rate": 9.985726531710862e-05, "loss": 0.0237, "step": 6610 }, { "grad_norm": 0.21340112388134003, "learning_rate": 9.985588557584689e-05, "loss": 0.0232, "step": 6620 }, { "grad_norm": 0.19912089407444, "learning_rate": 9.98544992076294e-05, "loss": 0.0261, "step": 6630 }, { "grad_norm": 0.21415619552135468, "learning_rate": 9.985310621264048e-05, "loss": 0.0225, "step": 6640 }, { "grad_norm": 0.2620002329349518, "learning_rate": 9.985170659106526e-05, "loss": 0.0194, "step": 6650 }, { "grad_norm": 0.25145694613456726, "learning_rate": 9.985030034308977e-05, "loss": 0.0223, "step": 6660 }, { "grad_norm": 0.19182083010673523, "learning_rate": 9.984888746890098e-05, "loss": 0.0207, "step": 6670 }, { "grad_norm": 0.2287394106388092, "learning_rate": 9.984746796868664e-05, "loss": 0.0249, "step": 6680 }, { "grad_norm": 0.18933095037937164, "learning_rate": 9.984604184263546e-05, "loss": 0.0227, "step": 6690 }, { "grad_norm": 0.17460067570209503, "learning_rate": 9.984460909093698e-05, "loss": 0.0202, "step": 6700 }, { "grad_norm": 0.17353948950767517, "learning_rate": 9.984316971378168e-05, "loss": 0.0214, "step": 6710 }, { "grad_norm": 0.24662521481513977, "learning_rate": 9.984172371136087e-05, "loss": 0.0213, "step": 6720 }, { "grad_norm": 0.24231332540512085, "learning_rate": 9.984027108386675e-05, "loss": 0.0231, "step": 6730 }, { "grad_norm": 0.25907954573631287, "learning_rate": 9.98388118314924e-05, "loss": 0.0241, "step": 6740 }, { "grad_norm": 0.19229936599731445, "learning_rate": 9.983734595443181e-05, "loss": 0.0228, "step": 6750 }, { "grad_norm": 0.2698969542980194, "learning_rate": 9.983587345287982e-05, "loss": 0.0221, "step": 6760 }, { "grad_norm": 0.1994880884885788, "learning_rate": 9.983439432703213e-05, "loss": 0.0189, "step": 6770 }, { "grad_norm": 0.23246780037879944, "learning_rate": 9.983290857708538e-05, "loss": 0.0199, "step": 6780 }, { "grad_norm": 0.30211251974105835, "learning_rate": 9.983141620323706e-05, "loss": 0.0223, "step": 6790 }, { "grad_norm": 0.36340847611427307, "learning_rate": 9.982991720568553e-05, "loss": 0.0303, "step": 6800 }, { "grad_norm": 0.34181177616119385, "learning_rate": 9.982841158463003e-05, "loss": 0.0212, "step": 6810 }, { "grad_norm": 0.23758503794670105, "learning_rate": 9.98268993402707e-05, "loss": 0.0209, "step": 6820 }, { "grad_norm": 0.17120790481567383, "learning_rate": 9.982538047280856e-05, "loss": 0.0181, "step": 6830 }, { "grad_norm": 0.1968587189912796, "learning_rate": 9.98238549824455e-05, "loss": 0.0203, "step": 6840 }, { "grad_norm": 0.17743870615959167, "learning_rate": 9.982232286938427e-05, "loss": 0.02, "step": 6850 }, { "grad_norm": 0.23600877821445465, "learning_rate": 9.982078413382855e-05, "loss": 0.0253, "step": 6860 }, { "grad_norm": 0.23463770747184753, "learning_rate": 9.981923877598285e-05, "loss": 0.0255, "step": 6870 }, { "grad_norm": 0.21961627900600433, "learning_rate": 9.981768679605259e-05, "loss": 0.025, "step": 6880 }, { "grad_norm": 0.15284568071365356, "learning_rate": 9.981612819424407e-05, "loss": 0.0237, "step": 6890 }, { "grad_norm": 0.17351043224334717, "learning_rate": 9.981456297076445e-05, "loss": 0.0223, "step": 6900 }, { "grad_norm": 0.19893111288547516, "learning_rate": 9.981299112582181e-05, "loss": 0.026, "step": 6910 }, { "grad_norm": 0.26945334672927856, "learning_rate": 9.981141265962504e-05, "loss": 0.0289, "step": 6920 }, { "grad_norm": 0.2214251011610031, "learning_rate": 9.980982757238397e-05, "loss": 0.0195, "step": 6930 }, { "grad_norm": 0.14704231917858124, "learning_rate": 9.980823586430932e-05, "loss": 0.0202, "step": 6940 }, { "grad_norm": 0.14472949504852295, "learning_rate": 9.980663753561264e-05, "loss": 0.0274, "step": 6950 }, { "grad_norm": 0.2915170192718506, "learning_rate": 9.980503258650638e-05, "loss": 0.0231, "step": 6960 }, { "grad_norm": 0.22163347899913788, "learning_rate": 9.980342101720388e-05, "loss": 0.022, "step": 6970 }, { "grad_norm": 0.18467183411121368, "learning_rate": 9.980180282791935e-05, "loss": 0.0198, "step": 6980 }, { "grad_norm": 0.251004159450531, "learning_rate": 9.980017801886789e-05, "loss": 0.0228, "step": 6990 }, { "grad_norm": 0.27328211069107056, "learning_rate": 9.979854659026549e-05, "loss": 0.0217, "step": 7000 }, { "grad_norm": 0.16636747121810913, "learning_rate": 9.979690854232895e-05, "loss": 0.0229, "step": 7010 }, { "grad_norm": 0.17554926872253418, "learning_rate": 9.979526387527604e-05, "loss": 0.0174, "step": 7020 }, { "grad_norm": 0.24472680687904358, "learning_rate": 9.979361258932539e-05, "loss": 0.0216, "step": 7030 }, { "grad_norm": 0.29241833090782166, "learning_rate": 9.979195468469647e-05, "loss": 0.0246, "step": 7040 }, { "grad_norm": 0.1951007843017578, "learning_rate": 9.979029016160963e-05, "loss": 0.0285, "step": 7050 }, { "grad_norm": 0.14513200521469116, "learning_rate": 9.978861902028616e-05, "loss": 0.02, "step": 7060 }, { "grad_norm": 0.18114402890205383, "learning_rate": 9.978694126094818e-05, "loss": 0.0243, "step": 7070 }, { "grad_norm": 0.22578930854797363, "learning_rate": 9.97852568838187e-05, "loss": 0.0175, "step": 7080 }, { "grad_norm": 0.17170864343643188, "learning_rate": 9.978356588912162e-05, "loss": 0.0206, "step": 7090 }, { "grad_norm": 0.18740487098693848, "learning_rate": 9.978186827708168e-05, "loss": 0.0248, "step": 7100 }, { "grad_norm": 0.1978156715631485, "learning_rate": 9.978016404792455e-05, "loss": 0.0211, "step": 7110 }, { "grad_norm": 0.25046488642692566, "learning_rate": 9.977845320187678e-05, "loss": 0.0239, "step": 7120 }, { "grad_norm": 0.3040243685245514, "learning_rate": 9.977673573916573e-05, "loss": 0.0227, "step": 7130 }, { "grad_norm": 0.1926129162311554, "learning_rate": 9.977501166001975e-05, "loss": 0.0231, "step": 7140 }, { "grad_norm": 0.23601660132408142, "learning_rate": 9.977328096466798e-05, "loss": 0.0227, "step": 7150 }, { "grad_norm": 0.10720379650592804, "learning_rate": 9.977154365334045e-05, "loss": 0.0248, "step": 7160 }, { "grad_norm": 0.27171581983566284, "learning_rate": 9.97697997262681e-05, "loss": 0.0215, "step": 7170 }, { "grad_norm": 0.17842461168766022, "learning_rate": 9.976804918368276e-05, "loss": 0.0212, "step": 7180 }, { "grad_norm": 0.23035359382629395, "learning_rate": 9.976629202581708e-05, "loss": 0.0268, "step": 7190 }, { "grad_norm": 0.20792731642723083, "learning_rate": 9.976452825290465e-05, "loss": 0.0194, "step": 7200 }, { "grad_norm": 0.17355109751224518, "learning_rate": 9.97627578651799e-05, "loss": 0.0182, "step": 7210 }, { "grad_norm": 0.2362479418516159, "learning_rate": 9.976098086287818e-05, "loss": 0.0246, "step": 7220 }, { "grad_norm": 0.21698296070098877, "learning_rate": 9.975919724623565e-05, "loss": 0.0196, "step": 7230 }, { "grad_norm": 0.24291442334651947, "learning_rate": 9.975740701548941e-05, "loss": 0.0201, "step": 7240 }, { "grad_norm": 0.3601558208465576, "learning_rate": 9.975561017087744e-05, "loss": 0.021, "step": 7250 }, { "grad_norm": 0.2726258337497711, "learning_rate": 9.975380671263858e-05, "loss": 0.0241, "step": 7260 }, { "grad_norm": 0.26007431745529175, "learning_rate": 9.975199664101252e-05, "loss": 0.0221, "step": 7270 }, { "grad_norm": 0.27130964398384094, "learning_rate": 9.975017995623988e-05, "loss": 0.0217, "step": 7280 }, { "grad_norm": 0.22829356789588928, "learning_rate": 9.974835665856213e-05, "loss": 0.022, "step": 7290 }, { "grad_norm": 0.14233966171741486, "learning_rate": 9.974652674822165e-05, "loss": 0.0241, "step": 7300 }, { "grad_norm": 0.14446929097175598, "learning_rate": 9.974469022546164e-05, "loss": 0.0257, "step": 7310 }, { "grad_norm": 0.13078276813030243, "learning_rate": 9.974284709052623e-05, "loss": 0.0202, "step": 7320 }, { "grad_norm": 0.14037875831127167, "learning_rate": 9.974099734366042e-05, "loss": 0.0205, "step": 7330 }, { "grad_norm": 0.19826462864875793, "learning_rate": 9.973914098511008e-05, "loss": 0.0189, "step": 7340 }, { "grad_norm": 0.17061476409435272, "learning_rate": 9.973727801512195e-05, "loss": 0.0185, "step": 7350 }, { "grad_norm": 0.29741838574409485, "learning_rate": 9.973540843394367e-05, "loss": 0.0222, "step": 7360 }, { "grad_norm": 0.1874169111251831, "learning_rate": 9.973353224182376e-05, "loss": 0.0215, "step": 7370 }, { "grad_norm": 0.11372797191143036, "learning_rate": 9.973164943901157e-05, "loss": 0.0194, "step": 7380 }, { "grad_norm": 0.23038190603256226, "learning_rate": 9.972976002575742e-05, "loss": 0.0243, "step": 7390 }, { "grad_norm": 0.22128954529762268, "learning_rate": 9.97278640023124e-05, "loss": 0.0218, "step": 7400 }, { "grad_norm": 0.2161611020565033, "learning_rate": 9.97259613689286e-05, "loss": 0.0177, "step": 7410 }, { "grad_norm": 0.12873412668704987, "learning_rate": 9.972405212585884e-05, "loss": 0.0235, "step": 7420 }, { "grad_norm": 0.2082854062318802, "learning_rate": 9.972213627335696e-05, "loss": 0.0195, "step": 7430 }, { "grad_norm": 0.1938987821340561, "learning_rate": 9.97202138116776e-05, "loss": 0.0227, "step": 7440 }, { "grad_norm": 0.1488087922334671, "learning_rate": 9.971828474107632e-05, "loss": 0.0184, "step": 7450 }, { "grad_norm": 0.21949107944965363, "learning_rate": 9.97163490618095e-05, "loss": 0.0226, "step": 7460 }, { "grad_norm": 0.16912822425365448, "learning_rate": 9.971440677413445e-05, "loss": 0.021, "step": 7470 }, { "grad_norm": 0.14798003435134888, "learning_rate": 9.971245787830935e-05, "loss": 0.0216, "step": 7480 }, { "grad_norm": 0.1331186592578888, "learning_rate": 9.971050237459325e-05, "loss": 0.0246, "step": 7490 }, { "grad_norm": 0.1720438003540039, "learning_rate": 9.970854026324608e-05, "loss": 0.021, "step": 7500 }, { "grad_norm": 0.18169455230236053, "learning_rate": 9.970657154452862e-05, "loss": 0.018, "step": 7510 }, { "grad_norm": 0.22353291511535645, "learning_rate": 9.97045962187026e-05, "loss": 0.0225, "step": 7520 }, { "grad_norm": 0.2049826830625534, "learning_rate": 9.970261428603059e-05, "loss": 0.025, "step": 7530 }, { "grad_norm": 0.21547292172908783, "learning_rate": 9.970062574677597e-05, "loss": 0.0185, "step": 7540 }, { "grad_norm": 0.15293604135513306, "learning_rate": 9.969863060120311e-05, "loss": 0.02, "step": 7550 }, { "grad_norm": 0.200698584318161, "learning_rate": 9.969662884957722e-05, "loss": 0.0185, "step": 7560 }, { "grad_norm": 0.2548874616622925, "learning_rate": 9.969462049216433e-05, "loss": 0.0233, "step": 7570 }, { "grad_norm": 0.29575079679489136, "learning_rate": 9.969260552923143e-05, "loss": 0.026, "step": 7580 }, { "grad_norm": 0.18493056297302246, "learning_rate": 9.969058396104636e-05, "loss": 0.0249, "step": 7590 }, { "grad_norm": 0.17440152168273926, "learning_rate": 9.96885557878778e-05, "loss": 0.0226, "step": 7600 }, { "grad_norm": 0.1505824327468872, "learning_rate": 9.968652100999535e-05, "loss": 0.0169, "step": 7610 }, { "grad_norm": 0.23863095045089722, "learning_rate": 9.968447962766952e-05, "loss": 0.0193, "step": 7620 }, { "grad_norm": 0.2016505002975464, "learning_rate": 9.968243164117158e-05, "loss": 0.02, "step": 7630 }, { "grad_norm": 0.18297936022281647, "learning_rate": 9.96803770507738e-05, "loss": 0.022, "step": 7640 }, { "grad_norm": 0.15989503264427185, "learning_rate": 9.967831585674929e-05, "loss": 0.0216, "step": 7650 }, { "grad_norm": 0.1680835485458374, "learning_rate": 9.967624805937199e-05, "loss": 0.0209, "step": 7660 }, { "grad_norm": 0.24122051894664764, "learning_rate": 9.967417365891678e-05, "loss": 0.0226, "step": 7670 }, { "grad_norm": 0.18453988432884216, "learning_rate": 9.96720926556594e-05, "loss": 0.0186, "step": 7680 }, { "grad_norm": 0.3329004645347595, "learning_rate": 9.967000504987645e-05, "loss": 0.022, "step": 7690 }, { "grad_norm": 0.1932893544435501, "learning_rate": 9.966791084184541e-05, "loss": 0.0209, "step": 7700 }, { "grad_norm": 0.22027307748794556, "learning_rate": 9.966581003184468e-05, "loss": 0.0239, "step": 7710 }, { "grad_norm": 0.1454644650220871, "learning_rate": 9.966370262015348e-05, "loss": 0.0185, "step": 7720 }, { "grad_norm": 0.139683336019516, "learning_rate": 9.966158860705191e-05, "loss": 0.0218, "step": 7730 }, { "grad_norm": 0.18768306076526642, "learning_rate": 9.965946799282103e-05, "loss": 0.0181, "step": 7740 }, { "grad_norm": 0.14192132651805878, "learning_rate": 9.965734077774265e-05, "loss": 0.02, "step": 7750 }, { "grad_norm": 0.15519434213638306, "learning_rate": 9.965520696209955e-05, "loss": 0.0165, "step": 7760 }, { "grad_norm": 0.16246941685676575, "learning_rate": 9.965306654617539e-05, "loss": 0.0155, "step": 7770 }, { "grad_norm": 0.2502850294113159, "learning_rate": 9.965091953025464e-05, "loss": 0.0222, "step": 7780 }, { "grad_norm": 0.19194726645946503, "learning_rate": 9.964876591462272e-05, "loss": 0.0206, "step": 7790 }, { "grad_norm": 0.2571253180503845, "learning_rate": 9.964660569956586e-05, "loss": 0.0205, "step": 7800 }, { "grad_norm": 0.26261821389198303, "learning_rate": 9.964443888537123e-05, "loss": 0.0221, "step": 7810 }, { "grad_norm": 0.21205304563045502, "learning_rate": 9.964226547232683e-05, "loss": 0.0235, "step": 7820 }, { "grad_norm": 0.11720824241638184, "learning_rate": 9.964008546072155e-05, "loss": 0.0277, "step": 7830 }, { "grad_norm": 0.25857993960380554, "learning_rate": 9.963789885084516e-05, "loss": 0.0222, "step": 7840 }, { "grad_norm": 0.22598819434642792, "learning_rate": 9.963570564298834e-05, "loss": 0.0211, "step": 7850 }, { "grad_norm": 0.2127516269683838, "learning_rate": 9.963350583744257e-05, "loss": 0.0177, "step": 7860 }, { "grad_norm": 0.18768300116062164, "learning_rate": 9.963129943450032e-05, "loss": 0.0183, "step": 7870 }, { "grad_norm": 0.15506187081336975, "learning_rate": 9.962908643445479e-05, "loss": 0.0206, "step": 7880 }, { "grad_norm": 0.18503610789775848, "learning_rate": 9.962686683760019e-05, "loss": 0.0208, "step": 7890 }, { "grad_norm": 0.19954077899456024, "learning_rate": 9.962464064423153e-05, "loss": 0.0187, "step": 7900 }, { "grad_norm": 0.17540717124938965, "learning_rate": 9.962240785464475e-05, "loss": 0.0219, "step": 7910 }, { "grad_norm": 0.16141821444034576, "learning_rate": 9.96201684691366e-05, "loss": 0.0187, "step": 7920 }, { "grad_norm": 0.26675087213516235, "learning_rate": 9.961792248800477e-05, "loss": 0.0212, "step": 7930 }, { "grad_norm": 0.21705879271030426, "learning_rate": 9.96156699115478e-05, "loss": 0.0217, "step": 7940 }, { "grad_norm": 0.2245873510837555, "learning_rate": 9.961341074006508e-05, "loss": 0.0261, "step": 7950 }, { "grad_norm": 0.19738559424877167, "learning_rate": 9.961114497385693e-05, "loss": 0.028, "step": 7960 }, { "grad_norm": 0.18209177255630493, "learning_rate": 9.960887261322452e-05, "loss": 0.0232, "step": 7970 }, { "grad_norm": 0.2320372313261032, "learning_rate": 9.96065936584699e-05, "loss": 0.0303, "step": 7980 }, { "grad_norm": 0.12798309326171875, "learning_rate": 9.960430810989596e-05, "loss": 0.0205, "step": 7990 }, { "grad_norm": 0.21158529818058014, "learning_rate": 9.960201596780654e-05, "loss": 0.0199, "step": 8000 }, { "grad_norm": 0.21195393800735474, "learning_rate": 9.959971723250632e-05, "loss": 0.0209, "step": 8010 }, { "grad_norm": 0.2851486802101135, "learning_rate": 9.95974119043008e-05, "loss": 0.0221, "step": 8020 }, { "grad_norm": 0.1093270480632782, "learning_rate": 9.959509998349648e-05, "loss": 0.023, "step": 8030 }, { "grad_norm": 0.23897668719291687, "learning_rate": 9.959278147040062e-05, "loss": 0.0191, "step": 8040 }, { "grad_norm": 0.16815946996212006, "learning_rate": 9.959045636532139e-05, "loss": 0.0179, "step": 8050 }, { "grad_norm": 0.22999155521392822, "learning_rate": 9.958812466856788e-05, "loss": 0.0223, "step": 8060 }, { "grad_norm": 0.15029039978981018, "learning_rate": 9.958578638045003e-05, "loss": 0.0189, "step": 8070 }, { "grad_norm": 0.11871293932199478, "learning_rate": 9.958344150127862e-05, "loss": 0.0189, "step": 8080 }, { "grad_norm": 0.1265449970960617, "learning_rate": 9.958109003136536e-05, "loss": 0.0183, "step": 8090 }, { "grad_norm": 0.14702185988426208, "learning_rate": 9.95787319710228e-05, "loss": 0.0152, "step": 8100 }, { "grad_norm": 0.1596256047487259, "learning_rate": 9.957636732056438e-05, "loss": 0.0178, "step": 8110 }, { "grad_norm": 0.12828028202056885, "learning_rate": 9.957399608030444e-05, "loss": 0.0186, "step": 8120 }, { "grad_norm": 0.17785131931304932, "learning_rate": 9.957161825055814e-05, "loss": 0.0161, "step": 8130 }, { "grad_norm": 0.18243032693862915, "learning_rate": 9.956923383164154e-05, "loss": 0.0215, "step": 8140 }, { "grad_norm": 0.16348546743392944, "learning_rate": 9.956684282387161e-05, "loss": 0.0214, "step": 8150 }, { "grad_norm": 0.12205395102500916, "learning_rate": 9.956444522756614e-05, "loss": 0.0219, "step": 8160 }, { "grad_norm": 0.21506065130233765, "learning_rate": 9.956204104304386e-05, "loss": 0.0206, "step": 8170 }, { "grad_norm": 0.1664665937423706, "learning_rate": 9.955963027062432e-05, "loss": 0.0231, "step": 8180 }, { "grad_norm": 0.187059685587883, "learning_rate": 9.955721291062794e-05, "loss": 0.0212, "step": 8190 }, { "grad_norm": 0.15794233977794647, "learning_rate": 9.955478896337609e-05, "loss": 0.0194, "step": 8200 }, { "grad_norm": 0.21558675169944763, "learning_rate": 9.955235842919093e-05, "loss": 0.0197, "step": 8210 }, { "grad_norm": 0.1778411716222763, "learning_rate": 9.954992130839553e-05, "loss": 0.0218, "step": 8220 }, { "grad_norm": 0.1597873419523239, "learning_rate": 9.954747760131387e-05, "loss": 0.0212, "step": 8230 }, { "grad_norm": 0.19327488541603088, "learning_rate": 9.954502730827074e-05, "loss": 0.0184, "step": 8240 }, { "grad_norm": 0.2400939017534256, "learning_rate": 9.954257042959187e-05, "loss": 0.0192, "step": 8250 }, { "grad_norm": 0.21796920895576477, "learning_rate": 9.954010696560381e-05, "loss": 0.0181, "step": 8260 }, { "grad_norm": 0.2225862294435501, "learning_rate": 9.953763691663401e-05, "loss": 0.0268, "step": 8270 }, { "grad_norm": 0.1551787108182907, "learning_rate": 9.95351602830108e-05, "loss": 0.0198, "step": 8280 }, { "grad_norm": 0.1892603188753128, "learning_rate": 9.953267706506338e-05, "loss": 0.0201, "step": 8290 }, { "grad_norm": 0.2202761173248291, "learning_rate": 9.953018726312182e-05, "loss": 0.0164, "step": 8300 }, { "grad_norm": 0.24105797708034515, "learning_rate": 9.952769087751707e-05, "loss": 0.0254, "step": 8310 }, { "grad_norm": 0.19372202455997467, "learning_rate": 9.952518790858098e-05, "loss": 0.0218, "step": 8320 }, { "grad_norm": 0.2201070487499237, "learning_rate": 9.952267835664622e-05, "loss": 0.0221, "step": 8330 }, { "grad_norm": 0.20117786526679993, "learning_rate": 9.952016222204638e-05, "loss": 0.0184, "step": 8340 }, { "grad_norm": 0.16619335114955902, "learning_rate": 9.95176395051159e-05, "loss": 0.0213, "step": 8350 }, { "grad_norm": 0.1788148581981659, "learning_rate": 9.95151102061901e-05, "loss": 0.0192, "step": 8360 }, { "grad_norm": 0.18428801000118256, "learning_rate": 9.95125743256052e-05, "loss": 0.0156, "step": 8370 }, { "grad_norm": 0.17269007861614227, "learning_rate": 9.951003186369827e-05, "loss": 0.021, "step": 8380 }, { "grad_norm": 0.16873174905776978, "learning_rate": 9.950748282080726e-05, "loss": 0.0194, "step": 8390 }, { "grad_norm": 0.09199657291173935, "learning_rate": 9.950492719727098e-05, "loss": 0.0171, "step": 8400 }, { "grad_norm": 0.22855350375175476, "learning_rate": 9.950236499342913e-05, "loss": 0.0198, "step": 8410 }, { "grad_norm": 0.21165381371974945, "learning_rate": 9.94997962096223e-05, "loss": 0.0184, "step": 8420 }, { "grad_norm": 0.1732073873281479, "learning_rate": 9.949722084619193e-05, "loss": 0.0181, "step": 8430 }, { "grad_norm": 0.21739359200000763, "learning_rate": 9.949463890348034e-05, "loss": 0.0182, "step": 8440 }, { "grad_norm": 0.21892696619033813, "learning_rate": 9.949205038183075e-05, "loss": 0.0197, "step": 8450 }, { "grad_norm": 0.15750735998153687, "learning_rate": 9.948945528158719e-05, "loss": 0.0146, "step": 8460 }, { "grad_norm": 0.20527909696102142, "learning_rate": 9.948685360309465e-05, "loss": 0.0187, "step": 8470 }, { "grad_norm": 0.29787254333496094, "learning_rate": 9.948424534669893e-05, "loss": 0.0248, "step": 8480 }, { "grad_norm": 0.2961162030696869, "learning_rate": 9.948163051274671e-05, "loss": 0.0242, "step": 8490 }, { "grad_norm": 0.1848241537809372, "learning_rate": 9.94790091015856e-05, "loss": 0.0182, "step": 8500 }, { "grad_norm": 0.15311864018440247, "learning_rate": 9.9476381113564e-05, "loss": 0.0206, "step": 8510 }, { "grad_norm": 0.19476012885570526, "learning_rate": 9.947374654903125e-05, "loss": 0.0187, "step": 8520 }, { "grad_norm": 0.223210409283638, "learning_rate": 9.947110540833755e-05, "loss": 0.0152, "step": 8530 }, { "grad_norm": 0.16255830228328705, "learning_rate": 9.946845769183395e-05, "loss": 0.0167, "step": 8540 }, { "grad_norm": 0.16612568497657776, "learning_rate": 9.946580339987241e-05, "loss": 0.0184, "step": 8550 }, { "grad_norm": 0.20754683017730713, "learning_rate": 9.946314253280571e-05, "loss": 0.0157, "step": 8560 }, { "grad_norm": 0.19858548045158386, "learning_rate": 9.946047509098757e-05, "loss": 0.0214, "step": 8570 }, { "grad_norm": 0.1304282695055008, "learning_rate": 9.945780107477254e-05, "loss": 0.0179, "step": 8580 }, { "grad_norm": 0.1460692286491394, "learning_rate": 9.945512048451606e-05, "loss": 0.019, "step": 8590 }, { "grad_norm": 0.15658390522003174, "learning_rate": 9.945243332057443e-05, "loss": 0.0223, "step": 8600 }, { "grad_norm": 0.2061644047498703, "learning_rate": 9.944973958330485e-05, "loss": 0.0204, "step": 8610 }, { "grad_norm": 0.15186980366706848, "learning_rate": 9.944703927306538e-05, "loss": 0.0202, "step": 8620 }, { "grad_norm": 0.19370487332344055, "learning_rate": 9.944433239021494e-05, "loss": 0.0177, "step": 8630 }, { "grad_norm": 0.15346036851406097, "learning_rate": 9.944161893511332e-05, "loss": 0.0169, "step": 8640 }, { "grad_norm": 0.15172408521175385, "learning_rate": 9.943889890812123e-05, "loss": 0.0217, "step": 8650 }, { "grad_norm": 0.15423600375652313, "learning_rate": 9.943617230960018e-05, "loss": 0.0186, "step": 8660 }, { "grad_norm": 0.2509293258190155, "learning_rate": 9.943343913991266e-05, "loss": 0.0217, "step": 8670 }, { "grad_norm": 0.1836799532175064, "learning_rate": 9.943069939942191e-05, "loss": 0.0238, "step": 8680 }, { "grad_norm": 0.24654896557331085, "learning_rate": 9.942795308849212e-05, "loss": 0.0182, "step": 8690 }, { "grad_norm": 0.15819969773292542, "learning_rate": 9.942520020748834e-05, "loss": 0.0203, "step": 8700 }, { "grad_norm": 0.18347391486167908, "learning_rate": 9.942244075677649e-05, "loss": 0.0157, "step": 8710 }, { "grad_norm": 0.2135680764913559, "learning_rate": 9.941967473672338e-05, "loss": 0.0197, "step": 8720 }, { "grad_norm": 0.12352743744850159, "learning_rate": 9.941690214769663e-05, "loss": 0.0207, "step": 8730 }, { "grad_norm": 0.17744281888008118, "learning_rate": 9.941412299006482e-05, "loss": 0.0186, "step": 8740 }, { "grad_norm": 0.19614547491073608, "learning_rate": 9.941133726419734e-05, "loss": 0.0219, "step": 8750 }, { "grad_norm": 0.23817111551761627, "learning_rate": 9.940854497046447e-05, "loss": 0.022, "step": 8760 }, { "grad_norm": 0.20493383705615997, "learning_rate": 9.940574610923741e-05, "loss": 0.0177, "step": 8770 }, { "grad_norm": 0.24739564955234528, "learning_rate": 9.940294068088813e-05, "loss": 0.023, "step": 8780 }, { "grad_norm": 0.1620938628911972, "learning_rate": 9.940012868578958e-05, "loss": 0.0226, "step": 8790 }, { "grad_norm": 0.15466728806495667, "learning_rate": 9.939731012431552e-05, "loss": 0.02, "step": 8800 }, { "grad_norm": 0.17312395572662354, "learning_rate": 9.939448499684061e-05, "loss": 0.0225, "step": 8810 }, { "grad_norm": 0.17915277183055878, "learning_rate": 9.939165330374037e-05, "loss": 0.0228, "step": 8820 }, { "grad_norm": 0.20053714513778687, "learning_rate": 9.938881504539115e-05, "loss": 0.0206, "step": 8830 }, { "grad_norm": 0.2756570279598236, "learning_rate": 9.938597022217029e-05, "loss": 0.0161, "step": 8840 }, { "grad_norm": 0.19484713673591614, "learning_rate": 9.938311883445587e-05, "loss": 0.0208, "step": 8850 }, { "grad_norm": 0.13732397556304932, "learning_rate": 9.938026088262696e-05, "loss": 0.0195, "step": 8860 }, { "grad_norm": 0.22546632587909698, "learning_rate": 9.937739636706338e-05, "loss": 0.0161, "step": 8870 }, { "grad_norm": 0.15628300607204437, "learning_rate": 9.937452528814594e-05, "loss": 0.019, "step": 8880 }, { "grad_norm": 0.23857538402080536, "learning_rate": 9.937164764625625e-05, "loss": 0.0198, "step": 8890 }, { "grad_norm": 0.21537122130393982, "learning_rate": 9.936876344177681e-05, "loss": 0.0231, "step": 8900 }, { "grad_norm": 0.13858000934123993, "learning_rate": 9.936587267509101e-05, "loss": 0.0163, "step": 8910 }, { "grad_norm": 0.16507023572921753, "learning_rate": 9.936297534658308e-05, "loss": 0.016, "step": 8920 }, { "grad_norm": 0.12139920890331268, "learning_rate": 9.936007145663814e-05, "loss": 0.0161, "step": 8930 }, { "grad_norm": 0.19780175387859344, "learning_rate": 9.935716100564219e-05, "loss": 0.0193, "step": 8940 }, { "grad_norm": 0.1993655562400818, "learning_rate": 9.93542439939821e-05, "loss": 0.0203, "step": 8950 }, { "grad_norm": 0.24585393071174622, "learning_rate": 9.935132042204558e-05, "loss": 0.0246, "step": 8960 }, { "grad_norm": 0.21267767250537872, "learning_rate": 9.934839029022127e-05, "loss": 0.0159, "step": 8970 }, { "grad_norm": 0.22372104227542877, "learning_rate": 9.934545359889864e-05, "loss": 0.0199, "step": 8980 }, { "grad_norm": 0.15055544674396515, "learning_rate": 9.934251034846802e-05, "loss": 0.0202, "step": 8990 }, { "grad_norm": 0.1695021390914917, "learning_rate": 9.933956053932065e-05, "loss": 0.0241, "step": 9000 }, { "grad_norm": 0.13975760340690613, "learning_rate": 9.933660417184863e-05, "loss": 0.0175, "step": 9010 }, { "grad_norm": 0.1713019758462906, "learning_rate": 9.933364124644493e-05, "loss": 0.0166, "step": 9020 }, { "grad_norm": 0.14152520895004272, "learning_rate": 9.933067176350337e-05, "loss": 0.0167, "step": 9030 }, { "grad_norm": 0.1480080634355545, "learning_rate": 9.932769572341867e-05, "loss": 0.016, "step": 9040 }, { "grad_norm": 0.12113139033317566, "learning_rate": 9.932471312658643e-05, "loss": 0.0177, "step": 9050 }, { "grad_norm": 0.1355171948671341, "learning_rate": 9.932172397340304e-05, "loss": 0.0182, "step": 9060 }, { "grad_norm": 0.20701663196086884, "learning_rate": 9.931872826426591e-05, "loss": 0.0188, "step": 9070 }, { "grad_norm": 0.1570008397102356, "learning_rate": 9.93157259995732e-05, "loss": 0.0192, "step": 9080 }, { "grad_norm": 0.15087154507637024, "learning_rate": 9.931271717972396e-05, "loss": 0.0179, "step": 9090 }, { "grad_norm": 0.15742790699005127, "learning_rate": 9.930970180511814e-05, "loss": 0.0195, "step": 9100 }, { "grad_norm": 0.13652536273002625, "learning_rate": 9.930667987615657e-05, "loss": 0.0193, "step": 9110 }, { "grad_norm": 0.22846300899982452, "learning_rate": 9.930365139324091e-05, "loss": 0.0193, "step": 9120 }, { "grad_norm": 0.1606762558221817, "learning_rate": 9.93006163567737e-05, "loss": 0.0142, "step": 9130 }, { "grad_norm": 0.16974948346614838, "learning_rate": 9.92975747671584e-05, "loss": 0.0185, "step": 9140 }, { "grad_norm": 0.19969569146633148, "learning_rate": 9.929452662479929e-05, "loss": 0.0215, "step": 9150 }, { "grad_norm": 0.19285519421100616, "learning_rate": 9.929147193010153e-05, "loss": 0.019, "step": 9160 }, { "grad_norm": 0.2214736044406891, "learning_rate": 9.928841068347116e-05, "loss": 0.024, "step": 9170 }, { "grad_norm": 0.21521727740764618, "learning_rate": 9.928534288531508e-05, "loss": 0.02, "step": 9180 }, { "grad_norm": 0.10966306179761887, "learning_rate": 9.928226853604108e-05, "loss": 0.017, "step": 9190 }, { "grad_norm": 0.1865493655204773, "learning_rate": 9.927918763605781e-05, "loss": 0.0191, "step": 9200 }, { "grad_norm": 0.16314129531383514, "learning_rate": 9.927610018577477e-05, "loss": 0.0173, "step": 9210 }, { "grad_norm": 0.11949130147695541, "learning_rate": 9.927300618560239e-05, "loss": 0.0215, "step": 9220 }, { "grad_norm": 0.20855805277824402, "learning_rate": 9.926990563595189e-05, "loss": 0.0203, "step": 9230 }, { "grad_norm": 0.14536745846271515, "learning_rate": 9.926679853723544e-05, "loss": 0.0186, "step": 9240 }, { "grad_norm": 0.16685067117214203, "learning_rate": 9.9263684889866e-05, "loss": 0.0214, "step": 9250 }, { "grad_norm": 0.21656987071037292, "learning_rate": 9.926056469425748e-05, "loss": 0.0187, "step": 9260 }, { "grad_norm": 0.2148478925228119, "learning_rate": 9.92574379508246e-05, "loss": 0.0201, "step": 9270 }, { "grad_norm": 0.1811847984790802, "learning_rate": 9.9254304659983e-05, "loss": 0.0172, "step": 9280 }, { "grad_norm": 0.12272396683692932, "learning_rate": 9.925116482214914e-05, "loss": 0.0175, "step": 9290 }, { "grad_norm": 0.22089874744415283, "learning_rate": 9.924801843774037e-05, "loss": 0.0195, "step": 9300 }, { "grad_norm": 0.2850658893585205, "learning_rate": 9.924486550717492e-05, "loss": 0.0199, "step": 9310 }, { "grad_norm": 0.2600962221622467, "learning_rate": 9.92417060308719e-05, "loss": 0.0201, "step": 9320 }, { "grad_norm": 0.2550548315048218, "learning_rate": 9.923854000925125e-05, "loss": 0.023, "step": 9330 }, { "grad_norm": 0.19607794284820557, "learning_rate": 9.923536744273385e-05, "loss": 0.0207, "step": 9340 }, { "grad_norm": 0.21093952655792236, "learning_rate": 9.923218833174134e-05, "loss": 0.0174, "step": 9350 }, { "grad_norm": 0.1900164633989334, "learning_rate": 9.922900267669636e-05, "loss": 0.0243, "step": 9360 }, { "grad_norm": 0.21176418662071228, "learning_rate": 9.922581047802231e-05, "loss": 0.0241, "step": 9370 }, { "grad_norm": 0.1534925252199173, "learning_rate": 9.92226117361435e-05, "loss": 0.022, "step": 9380 }, { "grad_norm": 0.1965819001197815, "learning_rate": 9.921940645148515e-05, "loss": 0.0198, "step": 9390 }, { "grad_norm": 0.23976635932922363, "learning_rate": 9.92161946244733e-05, "loss": 0.0178, "step": 9400 }, { "grad_norm": 0.1918146014213562, "learning_rate": 9.921297625553485e-05, "loss": 0.0185, "step": 9410 }, { "grad_norm": 0.1362549066543579, "learning_rate": 9.920975134509762e-05, "loss": 0.0198, "step": 9420 }, { "grad_norm": 0.1880139857530594, "learning_rate": 9.920651989359027e-05, "loss": 0.0177, "step": 9430 }, { "grad_norm": 0.20801420509815216, "learning_rate": 9.92032819014423e-05, "loss": 0.0216, "step": 9440 }, { "grad_norm": 0.14536702632904053, "learning_rate": 9.920003736908416e-05, "loss": 0.0169, "step": 9450 }, { "grad_norm": 0.18448825180530548, "learning_rate": 9.919678629694709e-05, "loss": 0.0219, "step": 9460 }, { "grad_norm": 0.21322469413280487, "learning_rate": 9.919352868546323e-05, "loss": 0.0197, "step": 9470 }, { "grad_norm": 0.15648020803928375, "learning_rate": 9.91902645350656e-05, "loss": 0.0199, "step": 9480 }, { "grad_norm": 0.20756733417510986, "learning_rate": 9.918699384618805e-05, "loss": 0.0224, "step": 9490 }, { "grad_norm": 0.18985487520694733, "learning_rate": 9.918371661926538e-05, "loss": 0.0205, "step": 9500 }, { "grad_norm": 0.14659175276756287, "learning_rate": 9.918043285473317e-05, "loss": 0.0186, "step": 9510 }, { "grad_norm": 0.12663695216178894, "learning_rate": 9.917714255302791e-05, "loss": 0.0218, "step": 9520 }, { "grad_norm": 0.1463843584060669, "learning_rate": 9.917384571458696e-05, "loss": 0.0195, "step": 9530 }, { "grad_norm": 0.15251773595809937, "learning_rate": 9.917054233984854e-05, "loss": 0.0238, "step": 9540 }, { "grad_norm": 0.15187685191631317, "learning_rate": 9.916723242925173e-05, "loss": 0.021, "step": 9550 }, { "grad_norm": 0.20710310339927673, "learning_rate": 9.916391598323651e-05, "loss": 0.0216, "step": 9560 }, { "grad_norm": 0.20567847788333893, "learning_rate": 9.91605930022437e-05, "loss": 0.0211, "step": 9570 }, { "grad_norm": 0.23919610679149628, "learning_rate": 9.9157263486715e-05, "loss": 0.0225, "step": 9580 }, { "grad_norm": 0.07820995151996613, "learning_rate": 9.915392743709298e-05, "loss": 0.0155, "step": 9590 }, { "grad_norm": 0.20738615095615387, "learning_rate": 9.915058485382108e-05, "loss": 0.0187, "step": 9600 }, { "grad_norm": 0.19671180844306946, "learning_rate": 9.914723573734356e-05, "loss": 0.0229, "step": 9610 }, { "grad_norm": 0.2233743816614151, "learning_rate": 9.914388008810566e-05, "loss": 0.0223, "step": 9620 }, { "grad_norm": 0.1674029380083084, "learning_rate": 9.914051790655337e-05, "loss": 0.024, "step": 9630 }, { "grad_norm": 0.2644875943660736, "learning_rate": 9.913714919313363e-05, "loss": 0.0177, "step": 9640 }, { "grad_norm": 0.2776275873184204, "learning_rate": 9.91337739482942e-05, "loss": 0.0224, "step": 9650 }, { "grad_norm": 0.2514394223690033, "learning_rate": 9.913039217248372e-05, "loss": 0.0204, "step": 9660 }, { "grad_norm": 0.21443839371204376, "learning_rate": 9.912700386615172e-05, "loss": 0.0192, "step": 9670 }, { "grad_norm": 0.16790707409381866, "learning_rate": 9.912360902974855e-05, "loss": 0.0195, "step": 9680 }, { "grad_norm": 0.11504894495010376, "learning_rate": 9.91202076637255e-05, "loss": 0.0158, "step": 9690 }, { "grad_norm": 0.11161217838525772, "learning_rate": 9.911679976853467e-05, "loss": 0.0177, "step": 9700 }, { "grad_norm": 0.10286598652601242, "learning_rate": 9.911338534462905e-05, "loss": 0.02, "step": 9710 }, { "grad_norm": 0.21529532968997955, "learning_rate": 9.910996439246247e-05, "loss": 0.0215, "step": 9720 }, { "grad_norm": 0.14428575336933136, "learning_rate": 9.910653691248968e-05, "loss": 0.0202, "step": 9730 }, { "grad_norm": 0.15255965292453766, "learning_rate": 9.910310290516627e-05, "loss": 0.0175, "step": 9740 }, { "grad_norm": 0.1696741282939911, "learning_rate": 9.909966237094865e-05, "loss": 0.0183, "step": 9750 }, { "grad_norm": 0.15457502007484436, "learning_rate": 9.90962153102942e-05, "loss": 0.017, "step": 9760 }, { "grad_norm": 0.19617430865764618, "learning_rate": 9.909276172366107e-05, "loss": 0.0188, "step": 9770 }, { "grad_norm": 0.21092666685581207, "learning_rate": 9.908930161150834e-05, "loss": 0.0174, "step": 9780 }, { "grad_norm": 0.13520777225494385, "learning_rate": 9.908583497429593e-05, "loss": 0.0188, "step": 9790 }, { "grad_norm": 0.17040935158729553, "learning_rate": 9.908236181248463e-05, "loss": 0.0148, "step": 9800 }, { "grad_norm": 0.23609909415245056, "learning_rate": 9.907888212653611e-05, "loss": 0.0198, "step": 9810 }, { "grad_norm": 0.14853458106517792, "learning_rate": 9.907539591691291e-05, "loss": 0.0154, "step": 9820 }, { "grad_norm": 0.16988438367843628, "learning_rate": 9.907190318407838e-05, "loss": 0.0223, "step": 9830 }, { "grad_norm": 0.13445447385311127, "learning_rate": 9.90684039284968e-05, "loss": 0.019, "step": 9840 }, { "grad_norm": 0.12470310181379318, "learning_rate": 9.906489815063333e-05, "loss": 0.0205, "step": 9850 }, { "grad_norm": 0.14663207530975342, "learning_rate": 9.906138585095395e-05, "loss": 0.0184, "step": 9860 }, { "grad_norm": 0.16024228930473328, "learning_rate": 9.905786702992549e-05, "loss": 0.0153, "step": 9870 }, { "grad_norm": 0.14968286454677582, "learning_rate": 9.905434168801572e-05, "loss": 0.0174, "step": 9880 }, { "grad_norm": 0.1957681030035019, "learning_rate": 9.905080982569321e-05, "loss": 0.0225, "step": 9890 }, { "grad_norm": 0.19849705696105957, "learning_rate": 9.904727144342744e-05, "loss": 0.0214, "step": 9900 }, { "grad_norm": 0.16518628597259521, "learning_rate": 9.904372654168874e-05, "loss": 0.0188, "step": 9910 }, { "grad_norm": 0.14035657048225403, "learning_rate": 9.904017512094831e-05, "loss": 0.0223, "step": 9920 }, { "grad_norm": 0.16782493889331818, "learning_rate": 9.903661718167818e-05, "loss": 0.0146, "step": 9930 }, { "grad_norm": 0.1809711456298828, "learning_rate": 9.903305272435133e-05, "loss": 0.0223, "step": 9940 }, { "grad_norm": 0.18775130808353424, "learning_rate": 9.902948174944152e-05, "loss": 0.022, "step": 9950 }, { "grad_norm": 0.153536856174469, "learning_rate": 9.90259042574234e-05, "loss": 0.0189, "step": 9960 }, { "grad_norm": 0.1507744938135147, "learning_rate": 9.902232024877254e-05, "loss": 0.0189, "step": 9970 }, { "grad_norm": 0.21685929596424103, "learning_rate": 9.90187297239653e-05, "loss": 0.0192, "step": 9980 }, { "grad_norm": 0.14479804039001465, "learning_rate": 9.901513268347895e-05, "loss": 0.0187, "step": 9990 }, { "grad_norm": 0.14513489603996277, "learning_rate": 9.901152912779164e-05, "loss": 0.0224, "step": 10000 }, { "grad_norm": 0.13540318608283997, "learning_rate": 9.900791905738233e-05, "loss": 0.0184, "step": 10010 }, { "grad_norm": 0.16828399896621704, "learning_rate": 9.900430247273089e-05, "loss": 0.0185, "step": 10020 }, { "grad_norm": 0.18047425150871277, "learning_rate": 9.900067937431805e-05, "loss": 0.0202, "step": 10030 }, { "grad_norm": 0.16305848956108093, "learning_rate": 9.899704976262539e-05, "loss": 0.0204, "step": 10040 }, { "grad_norm": 0.15307728946208954, "learning_rate": 9.899341363813537e-05, "loss": 0.0146, "step": 10050 }, { "grad_norm": 0.17332494258880615, "learning_rate": 9.898977100133134e-05, "loss": 0.0221, "step": 10060 }, { "grad_norm": 0.20953896641731262, "learning_rate": 9.898612185269744e-05, "loss": 0.0199, "step": 10070 }, { "grad_norm": 0.13448789715766907, "learning_rate": 9.898246619271874e-05, "loss": 0.0173, "step": 10080 }, { "grad_norm": 0.14839880168437958, "learning_rate": 9.897880402188118e-05, "loss": 0.0161, "step": 10090 }, { "grad_norm": 0.20986586809158325, "learning_rate": 9.897513534067151e-05, "loss": 0.0171, "step": 10100 }, { "grad_norm": 0.18478646874427795, "learning_rate": 9.897146014957742e-05, "loss": 0.0182, "step": 10110 }, { "grad_norm": 0.18059396743774414, "learning_rate": 9.896777844908738e-05, "loss": 0.0163, "step": 10120 }, { "grad_norm": 0.17330215871334076, "learning_rate": 9.896409023969078e-05, "loss": 0.016, "step": 10130 }, { "grad_norm": 0.12410493940114975, "learning_rate": 9.89603955218779e-05, "loss": 0.0151, "step": 10140 }, { "grad_norm": 0.17392072081565857, "learning_rate": 9.895669429613982e-05, "loss": 0.0168, "step": 10150 }, { "grad_norm": 0.20250344276428223, "learning_rate": 9.895298656296852e-05, "loss": 0.0171, "step": 10160 }, { "grad_norm": 0.14450973272323608, "learning_rate": 9.894927232285684e-05, "loss": 0.0187, "step": 10170 }, { "grad_norm": 0.15960846841335297, "learning_rate": 9.894555157629849e-05, "loss": 0.0166, "step": 10180 }, { "grad_norm": 0.18447965383529663, "learning_rate": 9.894182432378802e-05, "loss": 0.0167, "step": 10190 }, { "grad_norm": 0.1813008338212967, "learning_rate": 9.893809056582091e-05, "loss": 0.0187, "step": 10200 }, { "grad_norm": 0.1755169928073883, "learning_rate": 9.893435030289342e-05, "loss": 0.0205, "step": 10210 }, { "grad_norm": 0.17264272272586823, "learning_rate": 9.893060353550272e-05, "loss": 0.0158, "step": 10220 }, { "grad_norm": 0.21062198281288147, "learning_rate": 9.892685026414683e-05, "loss": 0.0186, "step": 10230 }, { "grad_norm": 0.13298042118549347, "learning_rate": 9.892309048932468e-05, "loss": 0.0161, "step": 10240 }, { "grad_norm": 0.190851628780365, "learning_rate": 9.8919324211536e-05, "loss": 0.0157, "step": 10250 }, { "grad_norm": 0.1704181581735611, "learning_rate": 9.891555143128141e-05, "loss": 0.0216, "step": 10260 }, { "grad_norm": 0.16961990296840668, "learning_rate": 9.89117721490624e-05, "loss": 0.0185, "step": 10270 }, { "grad_norm": 0.19153565168380737, "learning_rate": 9.890798636538132e-05, "loss": 0.025, "step": 10280 }, { "grad_norm": 0.13150626420974731, "learning_rate": 9.890419408074141e-05, "loss": 0.0198, "step": 10290 }, { "grad_norm": 0.13737238943576813, "learning_rate": 9.890039529564671e-05, "loss": 0.0169, "step": 10300 }, { "grad_norm": 0.18708333373069763, "learning_rate": 9.889659001060218e-05, "loss": 0.0167, "step": 10310 }, { "grad_norm": 0.14009033143520355, "learning_rate": 9.889277822611363e-05, "loss": 0.0195, "step": 10320 }, { "grad_norm": 0.16128556430339813, "learning_rate": 9.888895994268774e-05, "loss": 0.0151, "step": 10330 }, { "grad_norm": 0.16388118267059326, "learning_rate": 9.888513516083202e-05, "loss": 0.0159, "step": 10340 }, { "grad_norm": 0.14417748153209686, "learning_rate": 9.888130388105487e-05, "loss": 0.018, "step": 10350 }, { "grad_norm": 0.11653763800859451, "learning_rate": 9.887746610386558e-05, "loss": 0.0146, "step": 10360 }, { "grad_norm": 0.2308042049407959, "learning_rate": 9.887362182977425e-05, "loss": 0.0217, "step": 10370 }, { "grad_norm": 0.15395678579807281, "learning_rate": 9.886977105929189e-05, "loss": 0.0192, "step": 10380 }, { "grad_norm": 0.2132197767496109, "learning_rate": 9.886591379293034e-05, "loss": 0.016, "step": 10390 }, { "grad_norm": 0.14144928753376007, "learning_rate": 9.886205003120233e-05, "loss": 0.0179, "step": 10400 }, { "grad_norm": 0.1171068474650383, "learning_rate": 9.88581797746214e-05, "loss": 0.0182, "step": 10410 }, { "grad_norm": 0.1101209968328476, "learning_rate": 9.885430302370205e-05, "loss": 0.0186, "step": 10420 }, { "grad_norm": 0.18149322271347046, "learning_rate": 9.885041977895952e-05, "loss": 0.0165, "step": 10430 }, { "grad_norm": 0.17347191274166107, "learning_rate": 9.884653004091004e-05, "loss": 0.0187, "step": 10440 }, { "grad_norm": 0.23863796889781952, "learning_rate": 9.884263381007064e-05, "loss": 0.0185, "step": 10450 }, { "grad_norm": 0.157740980386734, "learning_rate": 9.883873108695918e-05, "loss": 0.0169, "step": 10460 }, { "grad_norm": 0.23168693482875824, "learning_rate": 9.883482187209442e-05, "loss": 0.0206, "step": 10470 }, { "grad_norm": 0.10868564248085022, "learning_rate": 9.883090616599601e-05, "loss": 0.0173, "step": 10480 }, { "grad_norm": 0.20547407865524292, "learning_rate": 9.882698396918442e-05, "loss": 0.0184, "step": 10490 }, { "grad_norm": 0.258035808801651, "learning_rate": 9.882305528218101e-05, "loss": 0.0181, "step": 10500 }, { "grad_norm": 0.15859529376029968, "learning_rate": 9.881912010550795e-05, "loss": 0.021, "step": 10510 }, { "grad_norm": 0.13599100708961487, "learning_rate": 9.881517843968835e-05, "loss": 0.0215, "step": 10520 }, { "grad_norm": 0.20496207475662231, "learning_rate": 9.881123028524615e-05, "loss": 0.0201, "step": 10530 }, { "grad_norm": 0.17919081449508667, "learning_rate": 9.880727564270613e-05, "loss": 0.0187, "step": 10540 }, { "grad_norm": 0.15748535096645355, "learning_rate": 9.880331451259394e-05, "loss": 0.0161, "step": 10550 }, { "grad_norm": 0.23824861645698547, "learning_rate": 9.879934689543612e-05, "loss": 0.0169, "step": 10560 }, { "grad_norm": 0.18897147476673126, "learning_rate": 9.879537279176006e-05, "loss": 0.0157, "step": 10570 }, { "grad_norm": 0.17256636917591095, "learning_rate": 9.879139220209397e-05, "loss": 0.0152, "step": 10580 }, { "grad_norm": 0.191074401140213, "learning_rate": 9.878740512696701e-05, "loss": 0.0185, "step": 10590 }, { "grad_norm": 0.1784287691116333, "learning_rate": 9.878341156690911e-05, "loss": 0.0165, "step": 10600 }, { "grad_norm": 0.1286739856004715, "learning_rate": 9.877941152245114e-05, "loss": 0.0184, "step": 10610 }, { "grad_norm": 0.16123248636722565, "learning_rate": 9.877540499412475e-05, "loss": 0.0165, "step": 10620 }, { "grad_norm": 0.12359566986560822, "learning_rate": 9.877139198246253e-05, "loss": 0.0177, "step": 10630 }, { "grad_norm": 0.15857356786727905, "learning_rate": 9.876737248799788e-05, "loss": 0.016, "step": 10640 }, { "grad_norm": 0.19610624015331268, "learning_rate": 9.87633465112651e-05, "loss": 0.0152, "step": 10650 }, { "grad_norm": 0.09657291322946548, "learning_rate": 9.875931405279932e-05, "loss": 0.0174, "step": 10660 }, { "grad_norm": 0.10266852378845215, "learning_rate": 9.875527511313655e-05, "loss": 0.0149, "step": 10670 }, { "grad_norm": 0.0903034657239914, "learning_rate": 9.875122969281364e-05, "loss": 0.0204, "step": 10680 }, { "grad_norm": 0.11197059601545334, "learning_rate": 9.874717779236833e-05, "loss": 0.014, "step": 10690 }, { "grad_norm": 0.13699525594711304, "learning_rate": 9.87431194123392e-05, "loss": 0.016, "step": 10700 }, { "grad_norm": 0.26714232563972473, "learning_rate": 9.87390545532657e-05, "loss": 0.0232, "step": 10710 }, { "grad_norm": 0.1263362318277359, "learning_rate": 9.873498321568815e-05, "loss": 0.0156, "step": 10720 }, { "grad_norm": 0.115691177546978, "learning_rate": 9.873090540014772e-05, "loss": 0.0147, "step": 10730 }, { "grad_norm": 0.13678234815597534, "learning_rate": 9.872682110718641e-05, "loss": 0.0175, "step": 10740 }, { "grad_norm": 0.17310023307800293, "learning_rate": 9.872273033734715e-05, "loss": 0.0175, "step": 10750 }, { "grad_norm": 0.13149891793727875, "learning_rate": 9.871863309117368e-05, "loss": 0.017, "step": 10760 }, { "grad_norm": 0.1981220245361328, "learning_rate": 9.871452936921063e-05, "loss": 0.0176, "step": 10770 }, { "grad_norm": 0.17580285668373108, "learning_rate": 9.871041917200347e-05, "loss": 0.0166, "step": 10780 }, { "grad_norm": 0.13197563588619232, "learning_rate": 9.870630250009851e-05, "loss": 0.0183, "step": 10790 }, { "grad_norm": 0.18152372539043427, "learning_rate": 9.870217935404299e-05, "loss": 0.0166, "step": 10800 }, { "grad_norm": 0.16534262895584106, "learning_rate": 9.869804973438493e-05, "loss": 0.0151, "step": 10810 }, { "grad_norm": 0.17121589183807373, "learning_rate": 9.869391364167327e-05, "loss": 0.0194, "step": 10820 }, { "grad_norm": 0.11225254088640213, "learning_rate": 9.868977107645779e-05, "loss": 0.0179, "step": 10830 }, { "grad_norm": 0.13587583601474762, "learning_rate": 9.868562203928912e-05, "loss": 0.0163, "step": 10840 }, { "grad_norm": 0.12130776047706604, "learning_rate": 9.868146653071875e-05, "loss": 0.0155, "step": 10850 }, { "grad_norm": 0.2346871793270111, "learning_rate": 9.867730455129906e-05, "loss": 0.0149, "step": 10860 }, { "grad_norm": 0.13557744026184082, "learning_rate": 9.867313610158326e-05, "loss": 0.0166, "step": 10870 }, { "grad_norm": 0.12069445103406906, "learning_rate": 9.866896118212543e-05, "loss": 0.0148, "step": 10880 }, { "grad_norm": 0.16601921617984772, "learning_rate": 9.866477979348051e-05, "loss": 0.0149, "step": 10890 }, { "grad_norm": 0.14868593215942383, "learning_rate": 9.86605919362043e-05, "loss": 0.0186, "step": 10900 }, { "grad_norm": 0.139649897813797, "learning_rate": 9.865639761085346e-05, "loss": 0.0181, "step": 10910 }, { "grad_norm": 0.14254550635814667, "learning_rate": 9.86521968179855e-05, "loss": 0.0141, "step": 10920 }, { "grad_norm": 0.20976530015468597, "learning_rate": 9.864798955815881e-05, "loss": 0.02, "step": 10930 }, { "grad_norm": 0.20362670719623566, "learning_rate": 9.864377583193263e-05, "loss": 0.0181, "step": 10940 }, { "grad_norm": 0.24052469432353973, "learning_rate": 9.863955563986705e-05, "loss": 0.0186, "step": 10950 }, { "grad_norm": 0.10306339710950851, "learning_rate": 9.863532898252302e-05, "loss": 0.0179, "step": 10960 }, { "grad_norm": 0.11891356110572815, "learning_rate": 9.863109586046237e-05, "loss": 0.0175, "step": 10970 }, { "grad_norm": 0.17944957315921783, "learning_rate": 9.86268562742478e-05, "loss": 0.0184, "step": 10980 }, { "grad_norm": 0.1719767153263092, "learning_rate": 9.862261022444279e-05, "loss": 0.0204, "step": 10990 }, { "grad_norm": 0.15667186677455902, "learning_rate": 9.861835771161177e-05, "loss": 0.0236, "step": 11000 }, { "grad_norm": 0.19629740715026855, "learning_rate": 9.861409873631998e-05, "loss": 0.0154, "step": 11010 }, { "grad_norm": 0.1897650510072708, "learning_rate": 9.860983329913354e-05, "loss": 0.0161, "step": 11020 }, { "grad_norm": 0.16961941123008728, "learning_rate": 9.860556140061943e-05, "loss": 0.0217, "step": 11030 }, { "grad_norm": 0.1309482902288437, "learning_rate": 9.860128304134546e-05, "loss": 0.0139, "step": 11040 }, { "grad_norm": 0.15743054449558258, "learning_rate": 9.859699822188035e-05, "loss": 0.0165, "step": 11050 }, { "grad_norm": 0.12438806146383286, "learning_rate": 9.85927069427936e-05, "loss": 0.0132, "step": 11060 }, { "grad_norm": 0.09827075898647308, "learning_rate": 9.858840920465566e-05, "loss": 0.0184, "step": 11070 }, { "grad_norm": 0.15732558071613312, "learning_rate": 9.858410500803778e-05, "loss": 0.0168, "step": 11080 }, { "grad_norm": 0.10824697464704514, "learning_rate": 9.857979435351208e-05, "loss": 0.018, "step": 11090 }, { "grad_norm": 0.16149915754795074, "learning_rate": 9.857547724165152e-05, "loss": 0.018, "step": 11100 }, { "grad_norm": 0.12490206211805344, "learning_rate": 9.857115367303e-05, "loss": 0.0224, "step": 11110 }, { "grad_norm": 0.15117000043392181, "learning_rate": 9.856682364822217e-05, "loss": 0.0178, "step": 11120 }, { "grad_norm": 0.2202039659023285, "learning_rate": 9.856248716780358e-05, "loss": 0.0157, "step": 11130 }, { "grad_norm": 0.17330864071846008, "learning_rate": 9.855814423235069e-05, "loss": 0.0202, "step": 11140 }, { "grad_norm": 0.14217106997966766, "learning_rate": 9.855379484244074e-05, "loss": 0.0227, "step": 11150 }, { "grad_norm": 0.12819333374500275, "learning_rate": 9.854943899865185e-05, "loss": 0.0221, "step": 11160 }, { "grad_norm": 0.20195798575878143, "learning_rate": 9.854507670156306e-05, "loss": 0.0188, "step": 11170 }, { "grad_norm": 0.19126029312610626, "learning_rate": 9.854070795175415e-05, "loss": 0.0166, "step": 11180 }, { "grad_norm": 0.2013355791568756, "learning_rate": 9.853633274980587e-05, "loss": 0.0246, "step": 11190 }, { "grad_norm": 0.20427575707435608, "learning_rate": 9.853195109629976e-05, "loss": 0.022, "step": 11200 }, { "grad_norm": 0.17429088056087494, "learning_rate": 9.852756299181826e-05, "loss": 0.0218, "step": 11210 }, { "grad_norm": 0.1336418241262436, "learning_rate": 9.852316843694463e-05, "loss": 0.0178, "step": 11220 }, { "grad_norm": 0.1436985731124878, "learning_rate": 9.8518767432263e-05, "loss": 0.0193, "step": 11230 }, { "grad_norm": 0.09055299311876297, "learning_rate": 9.851435997835836e-05, "loss": 0.0143, "step": 11240 }, { "grad_norm": 0.14784426987171173, "learning_rate": 9.850994607581658e-05, "loss": 0.0205, "step": 11250 }, { "grad_norm": 0.2094230204820633, "learning_rate": 9.850552572522436e-05, "loss": 0.0168, "step": 11260 }, { "grad_norm": 0.151676207780838, "learning_rate": 9.850109892716923e-05, "loss": 0.016, "step": 11270 }, { "grad_norm": 0.1625036597251892, "learning_rate": 9.849666568223965e-05, "loss": 0.0159, "step": 11280 }, { "grad_norm": 0.21141672134399414, "learning_rate": 9.84922259910249e-05, "loss": 0.0189, "step": 11290 }, { "grad_norm": 0.2355823516845703, "learning_rate": 9.848777985411508e-05, "loss": 0.019, "step": 11300 }, { "grad_norm": 0.154935821890831, "learning_rate": 9.848332727210118e-05, "loss": 0.0162, "step": 11310 }, { "grad_norm": 0.14841005206108093, "learning_rate": 9.84788682455751e-05, "loss": 0.0157, "step": 11320 }, { "grad_norm": 0.17447148263454437, "learning_rate": 9.84744027751295e-05, "loss": 0.0187, "step": 11330 }, { "grad_norm": 0.15201225876808167, "learning_rate": 9.846993086135792e-05, "loss": 0.0206, "step": 11340 }, { "grad_norm": 0.11921828240156174, "learning_rate": 9.846545250485482e-05, "loss": 0.0167, "step": 11350 }, { "grad_norm": 0.11958276480436325, "learning_rate": 9.846096770621547e-05, "loss": 0.0144, "step": 11360 }, { "grad_norm": 0.15566162765026093, "learning_rate": 9.845647646603597e-05, "loss": 0.0196, "step": 11370 }, { "grad_norm": 0.16593754291534424, "learning_rate": 9.845197878491335e-05, "loss": 0.0143, "step": 11380 }, { "grad_norm": 0.10852315276861191, "learning_rate": 9.844747466344539e-05, "loss": 0.0141, "step": 11390 }, { "grad_norm": 0.19197338819503784, "learning_rate": 9.844296410223083e-05, "loss": 0.0131, "step": 11400 }, { "grad_norm": 0.2131531685590744, "learning_rate": 9.843844710186925e-05, "loss": 0.0167, "step": 11410 }, { "grad_norm": 0.17680968344211578, "learning_rate": 9.8433923662961e-05, "loss": 0.0183, "step": 11420 }, { "grad_norm": 0.12735892832279205, "learning_rate": 9.842939378610738e-05, "loss": 0.0155, "step": 11430 }, { "grad_norm": 0.19633276760578156, "learning_rate": 9.842485747191048e-05, "loss": 0.0163, "step": 11440 }, { "grad_norm": 0.20909219980239868, "learning_rate": 9.842031472097333e-05, "loss": 0.0174, "step": 11450 }, { "grad_norm": 0.2054215967655182, "learning_rate": 9.841576553389971e-05, "loss": 0.0229, "step": 11460 }, { "grad_norm": 0.17817141115665436, "learning_rate": 9.841120991129435e-05, "loss": 0.0188, "step": 11470 }, { "grad_norm": 0.14537853002548218, "learning_rate": 9.840664785376276e-05, "loss": 0.0185, "step": 11480 }, { "grad_norm": 0.18088160455226898, "learning_rate": 9.840207936191137e-05, "loss": 0.0156, "step": 11490 }, { "grad_norm": 0.13529551029205322, "learning_rate": 9.839750443634742e-05, "loss": 0.0178, "step": 11500 }, { "grad_norm": 0.17790286242961884, "learning_rate": 9.8392923077679e-05, "loss": 0.0169, "step": 11510 }, { "grad_norm": 0.12445485591888428, "learning_rate": 9.83883352865151e-05, "loss": 0.0174, "step": 11520 }, { "grad_norm": 0.13131456077098846, "learning_rate": 9.838374106346553e-05, "loss": 0.0189, "step": 11530 }, { "grad_norm": 0.22245138883590698, "learning_rate": 9.837914040914095e-05, "loss": 0.0143, "step": 11540 }, { "grad_norm": 0.20563754439353943, "learning_rate": 9.837453332415292e-05, "loss": 0.0178, "step": 11550 }, { "grad_norm": 0.21233361959457397, "learning_rate": 9.836991980911382e-05, "loss": 0.0178, "step": 11560 }, { "grad_norm": 0.2030424326658249, "learning_rate": 9.836529986463688e-05, "loss": 0.0176, "step": 11570 }, { "grad_norm": 0.21054987609386444, "learning_rate": 9.836067349133616e-05, "loss": 0.0154, "step": 11580 }, { "grad_norm": 0.2096216380596161, "learning_rate": 9.835604068982666e-05, "loss": 0.0171, "step": 11590 }, { "grad_norm": 0.18359149992465973, "learning_rate": 9.835140146072414e-05, "loss": 0.0154, "step": 11600 }, { "grad_norm": 0.15023301541805267, "learning_rate": 9.83467558046453e-05, "loss": 0.0153, "step": 11610 }, { "grad_norm": 0.19365930557250977, "learning_rate": 9.834210372220762e-05, "loss": 0.0176, "step": 11620 }, { "grad_norm": 0.2173481434583664, "learning_rate": 9.833744521402948e-05, "loss": 0.0143, "step": 11630 }, { "grad_norm": 0.17353621125221252, "learning_rate": 9.833278028073009e-05, "loss": 0.0205, "step": 11640 }, { "grad_norm": 0.19923406839370728, "learning_rate": 9.832810892292953e-05, "loss": 0.0148, "step": 11650 }, { "grad_norm": 0.21655096113681793, "learning_rate": 9.83234311412487e-05, "loss": 0.0152, "step": 11660 }, { "grad_norm": 0.14929360151290894, "learning_rate": 9.831874693630945e-05, "loss": 0.0175, "step": 11670 }, { "grad_norm": 0.1342879831790924, "learning_rate": 9.831405630873434e-05, "loss": 0.014, "step": 11680 }, { "grad_norm": 0.09389326721429825, "learning_rate": 9.83093592591469e-05, "loss": 0.0163, "step": 11690 }, { "grad_norm": 0.10263054072856903, "learning_rate": 9.830465578817145e-05, "loss": 0.0161, "step": 11700 }, { "grad_norm": 0.109204962849617, "learning_rate": 9.82999458964332e-05, "loss": 0.0205, "step": 11710 }, { "grad_norm": 0.18020181357860565, "learning_rate": 9.829522958455819e-05, "loss": 0.0134, "step": 11720 }, { "grad_norm": 0.11472059041261673, "learning_rate": 9.829050685317333e-05, "loss": 0.0154, "step": 11730 }, { "grad_norm": 0.18864989280700684, "learning_rate": 9.828577770290639e-05, "loss": 0.0154, "step": 11740 }, { "grad_norm": 0.18584659695625305, "learning_rate": 9.828104213438596e-05, "loss": 0.0166, "step": 11750 }, { "grad_norm": 0.17490431666374207, "learning_rate": 9.827630014824151e-05, "loss": 0.0182, "step": 11760 }, { "grad_norm": 0.14184430241584778, "learning_rate": 9.827155174510334e-05, "loss": 0.013, "step": 11770 }, { "grad_norm": 0.10925900191068649, "learning_rate": 9.826679692560263e-05, "loss": 0.013, "step": 11780 }, { "grad_norm": 0.2027689665555954, "learning_rate": 9.826203569037141e-05, "loss": 0.0172, "step": 11790 }, { "grad_norm": 0.14337995648384094, "learning_rate": 9.825726804004253e-05, "loss": 0.0161, "step": 11800 }, { "grad_norm": 0.1600741446018219, "learning_rate": 9.825249397524976e-05, "loss": 0.0142, "step": 11810 }, { "grad_norm": 0.10313127934932709, "learning_rate": 9.824771349662763e-05, "loss": 0.0165, "step": 11820 }, { "grad_norm": 0.16814599931240082, "learning_rate": 9.82429266048116e-05, "loss": 0.0168, "step": 11830 }, { "grad_norm": 0.13768674433231354, "learning_rate": 9.823813330043795e-05, "loss": 0.0158, "step": 11840 }, { "grad_norm": 0.13844852149486542, "learning_rate": 9.823333358414382e-05, "loss": 0.0186, "step": 11850 }, { "grad_norm": 0.17914243042469025, "learning_rate": 9.822852745656719e-05, "loss": 0.0205, "step": 11860 }, { "grad_norm": 0.17964361608028412, "learning_rate": 9.82237149183469e-05, "loss": 0.0212, "step": 11870 }, { "grad_norm": 0.1907847374677658, "learning_rate": 9.821889597012264e-05, "loss": 0.015, "step": 11880 }, { "grad_norm": 0.20532183349132538, "learning_rate": 9.821407061253498e-05, "loss": 0.0181, "step": 11890 }, { "grad_norm": 0.20001351833343506, "learning_rate": 9.820923884622528e-05, "loss": 0.019, "step": 11900 }, { "grad_norm": 0.14930211007595062, "learning_rate": 9.820440067183582e-05, "loss": 0.0184, "step": 11910 }, { "grad_norm": 0.1055416688323021, "learning_rate": 9.81995560900097e-05, "loss": 0.014, "step": 11920 }, { "grad_norm": 0.20055480301380157, "learning_rate": 9.819470510139086e-05, "loss": 0.0142, "step": 11930 }, { "grad_norm": 0.17686061561107635, "learning_rate": 9.818984770662411e-05, "loss": 0.0181, "step": 11940 }, { "grad_norm": 0.1838676631450653, "learning_rate": 9.81849839063551e-05, "loss": 0.0156, "step": 11950 }, { "grad_norm": 0.12055360525846481, "learning_rate": 9.818011370123034e-05, "loss": 0.0168, "step": 11960 }, { "grad_norm": 0.17560319602489471, "learning_rate": 9.81752370918972e-05, "loss": 0.0191, "step": 11970 }, { "grad_norm": 0.2679031491279602, "learning_rate": 9.817035407900387e-05, "loss": 0.0164, "step": 11980 }, { "grad_norm": 0.16278637945652008, "learning_rate": 9.816546466319943e-05, "loss": 0.015, "step": 11990 }, { "grad_norm": 0.13758839666843414, "learning_rate": 9.816056884513377e-05, "loss": 0.0174, "step": 12000 }, { "grad_norm": 0.14542782306671143, "learning_rate": 9.815566662545766e-05, "loss": 0.0174, "step": 12010 }, { "grad_norm": 0.13682781159877777, "learning_rate": 9.815075800482271e-05, "loss": 0.0199, "step": 12020 }, { "grad_norm": 0.18759523332118988, "learning_rate": 9.814584298388143e-05, "loss": 0.0159, "step": 12030 }, { "grad_norm": 0.11767801642417908, "learning_rate": 9.814092156328709e-05, "loss": 0.0132, "step": 12040 }, { "grad_norm": 0.16466158628463745, "learning_rate": 9.813599374369384e-05, "loss": 0.0182, "step": 12050 }, { "grad_norm": 0.17070740461349487, "learning_rate": 9.813105952575675e-05, "loss": 0.0191, "step": 12060 }, { "grad_norm": 0.22355064749717712, "learning_rate": 9.812611891013165e-05, "loss": 0.015, "step": 12070 }, { "grad_norm": 0.1512124389410019, "learning_rate": 9.812117189747528e-05, "loss": 0.0151, "step": 12080 }, { "grad_norm": 0.11573900282382965, "learning_rate": 9.811621848844517e-05, "loss": 0.012, "step": 12090 }, { "grad_norm": 0.1452537328004837, "learning_rate": 9.811125868369978e-05, "loss": 0.0168, "step": 12100 }, { "grad_norm": 0.1508302241563797, "learning_rate": 9.810629248389836e-05, "loss": 0.0138, "step": 12110 }, { "grad_norm": 0.1895166039466858, "learning_rate": 9.810131988970103e-05, "loss": 0.0161, "step": 12120 }, { "grad_norm": 0.1581396758556366, "learning_rate": 9.809634090176876e-05, "loss": 0.0136, "step": 12130 }, { "grad_norm": 0.24489708244800568, "learning_rate": 9.809135552076338e-05, "loss": 0.0165, "step": 12140 }, { "grad_norm": 0.1369009017944336, "learning_rate": 9.808636374734752e-05, "loss": 0.0163, "step": 12150 }, { "grad_norm": 0.1245192140340805, "learning_rate": 9.808136558218475e-05, "loss": 0.0227, "step": 12160 }, { "grad_norm": 0.11369010806083679, "learning_rate": 9.807636102593938e-05, "loss": 0.0149, "step": 12170 }, { "grad_norm": 0.1343180239200592, "learning_rate": 9.807135007927667e-05, "loss": 0.0182, "step": 12180 }, { "grad_norm": 0.10898847132921219, "learning_rate": 9.806633274286269e-05, "loss": 0.0182, "step": 12190 }, { "grad_norm": 0.11302303522825241, "learning_rate": 9.806130901736432e-05, "loss": 0.0138, "step": 12200 }, { "grad_norm": 0.20815467834472656, "learning_rate": 9.805627890344935e-05, "loss": 0.0178, "step": 12210 }, { "grad_norm": 0.13435164093971252, "learning_rate": 9.805124240178637e-05, "loss": 0.0152, "step": 12220 }, { "grad_norm": 0.1700487732887268, "learning_rate": 9.80461995130449e-05, "loss": 0.0187, "step": 12230 }, { "grad_norm": 0.15386641025543213, "learning_rate": 9.804115023789517e-05, "loss": 0.0135, "step": 12240 }, { "grad_norm": 0.22499416768550873, "learning_rate": 9.803609457700842e-05, "loss": 0.0156, "step": 12250 }, { "grad_norm": 0.20969776809215546, "learning_rate": 9.80310325310566e-05, "loss": 0.0168, "step": 12260 }, { "grad_norm": 0.20889787375926971, "learning_rate": 9.802596410071259e-05, "loss": 0.0158, "step": 12270 }, { "grad_norm": 0.13788944482803345, "learning_rate": 9.802088928665012e-05, "loss": 0.0166, "step": 12280 }, { "grad_norm": 0.1720416396856308, "learning_rate": 9.80158080895437e-05, "loss": 0.0151, "step": 12290 }, { "grad_norm": 0.09556452929973602, "learning_rate": 9.801072051006878e-05, "loss": 0.0127, "step": 12300 }, { "grad_norm": 0.13022537529468536, "learning_rate": 9.800562654890157e-05, "loss": 0.0149, "step": 12310 }, { "grad_norm": 0.14356820285320282, "learning_rate": 9.80005262067192e-05, "loss": 0.0146, "step": 12320 }, { "grad_norm": 0.16995291411876678, "learning_rate": 9.799541948419963e-05, "loss": 0.0148, "step": 12330 }, { "grad_norm": 0.21986576914787292, "learning_rate": 9.799030638202162e-05, "loss": 0.0145, "step": 12340 }, { "grad_norm": 0.19433064758777618, "learning_rate": 9.798518690086483e-05, "loss": 0.0172, "step": 12350 }, { "grad_norm": 0.21013621985912323, "learning_rate": 9.798006104140975e-05, "loss": 0.0154, "step": 12360 }, { "grad_norm": 0.16144007444381714, "learning_rate": 9.797492880433774e-05, "loss": 0.0127, "step": 12370 }, { "grad_norm": 0.11613766849040985, "learning_rate": 9.796979019033097e-05, "loss": 0.0211, "step": 12380 }, { "grad_norm": 0.15912482142448425, "learning_rate": 9.796464520007249e-05, "loss": 0.0148, "step": 12390 }, { "grad_norm": 0.16171365976333618, "learning_rate": 9.795949383424616e-05, "loss": 0.0189, "step": 12400 }, { "grad_norm": 0.19967658817768097, "learning_rate": 9.795433609353673e-05, "loss": 0.0161, "step": 12410 }, { "grad_norm": 0.11777558922767639, "learning_rate": 9.794917197862976e-05, "loss": 0.0156, "step": 12420 }, { "grad_norm": 0.12051840871572495, "learning_rate": 9.794400149021172e-05, "loss": 0.0178, "step": 12430 }, { "grad_norm": 0.17323453724384308, "learning_rate": 9.793882462896983e-05, "loss": 0.0159, "step": 12440 }, { "grad_norm": 0.11789634078741074, "learning_rate": 9.793364139559224e-05, "loss": 0.0138, "step": 12450 }, { "grad_norm": 0.15239350497722626, "learning_rate": 9.792845179076791e-05, "loss": 0.0162, "step": 12460 }, { "grad_norm": 0.11618350446224213, "learning_rate": 9.792325581518664e-05, "loss": 0.0152, "step": 12470 }, { "grad_norm": 0.18432050943374634, "learning_rate": 9.79180534695391e-05, "loss": 0.0186, "step": 12480 }, { "grad_norm": 0.10140036791563034, "learning_rate": 9.791284475451683e-05, "loss": 0.0163, "step": 12490 }, { "grad_norm": 0.20130154490470886, "learning_rate": 9.790762967081212e-05, "loss": 0.0156, "step": 12500 }, { "grad_norm": 0.22380827367305756, "learning_rate": 9.790240821911824e-05, "loss": 0.0149, "step": 12510 }, { "grad_norm": 0.21064086258411407, "learning_rate": 9.789718040012919e-05, "loss": 0.0156, "step": 12520 }, { "grad_norm": 0.13735580444335938, "learning_rate": 9.789194621453988e-05, "loss": 0.0155, "step": 12530 }, { "grad_norm": 0.12468552589416504, "learning_rate": 9.788670566304603e-05, "loss": 0.015, "step": 12540 }, { "grad_norm": 0.09322162717580795, "learning_rate": 9.788145874634426e-05, "loss": 0.0156, "step": 12550 }, { "grad_norm": 0.16368430852890015, "learning_rate": 9.787620546513196e-05, "loss": 0.0129, "step": 12560 }, { "grad_norm": 0.16245050728321075, "learning_rate": 9.787094582010745e-05, "loss": 0.0184, "step": 12570 }, { "grad_norm": 0.1573290079832077, "learning_rate": 9.786567981196983e-05, "loss": 0.016, "step": 12580 }, { "grad_norm": 0.1837443858385086, "learning_rate": 9.786040744141909e-05, "loss": 0.0152, "step": 12590 }, { "grad_norm": 0.13828647136688232, "learning_rate": 9.785512870915601e-05, "loss": 0.017, "step": 12600 }, { "grad_norm": 0.12844276428222656, "learning_rate": 9.784984361588227e-05, "loss": 0.014, "step": 12610 }, { "grad_norm": 0.16216817498207092, "learning_rate": 9.78445521623004e-05, "loss": 0.015, "step": 12620 }, { "grad_norm": 0.127259761095047, "learning_rate": 9.783925434911372e-05, "loss": 0.0171, "step": 12630 }, { "grad_norm": 0.12006533145904541, "learning_rate": 9.783395017702643e-05, "loss": 0.0129, "step": 12640 }, { "grad_norm": 0.19317756593227386, "learning_rate": 9.782863964674357e-05, "loss": 0.015, "step": 12650 }, { "grad_norm": 0.13987818360328674, "learning_rate": 9.782332275897105e-05, "loss": 0.0159, "step": 12660 }, { "grad_norm": 0.2151240110397339, "learning_rate": 9.781799951441559e-05, "loss": 0.0186, "step": 12670 }, { "grad_norm": 0.12777483463287354, "learning_rate": 9.781266991378476e-05, "loss": 0.0125, "step": 12680 }, { "grad_norm": 0.10695236176252365, "learning_rate": 9.780733395778698e-05, "loss": 0.0176, "step": 12690 }, { "grad_norm": 0.1803872436285019, "learning_rate": 9.780199164713154e-05, "loss": 0.019, "step": 12700 }, { "grad_norm": 0.12743505835533142, "learning_rate": 9.779664298252853e-05, "loss": 0.0124, "step": 12710 }, { "grad_norm": 0.17272458970546722, "learning_rate": 9.779128796468893e-05, "loss": 0.0172, "step": 12720 }, { "grad_norm": 0.18939939141273499, "learning_rate": 9.77859265943245e-05, "loss": 0.014, "step": 12730 }, { "grad_norm": 0.1557309627532959, "learning_rate": 9.778055887214793e-05, "loss": 0.0157, "step": 12740 }, { "grad_norm": 0.11105551570653915, "learning_rate": 9.777518479887269e-05, "loss": 0.013, "step": 12750 }, { "grad_norm": 0.1684393733739853, "learning_rate": 9.77698043752131e-05, "loss": 0.0142, "step": 12760 }, { "grad_norm": 0.19197268784046173, "learning_rate": 9.776441760188437e-05, "loss": 0.0127, "step": 12770 }, { "grad_norm": 0.21469469368457794, "learning_rate": 9.775902447960249e-05, "loss": 0.0207, "step": 12780 }, { "grad_norm": 0.26988089084625244, "learning_rate": 9.775362500908435e-05, "loss": 0.0204, "step": 12790 }, { "grad_norm": 0.23172752559185028, "learning_rate": 9.774821919104764e-05, "loss": 0.0184, "step": 12800 }, { "grad_norm": 0.2061227560043335, "learning_rate": 9.774280702621093e-05, "loss": 0.0211, "step": 12810 }, { "grad_norm": 0.16675373911857605, "learning_rate": 9.773738851529362e-05, "loss": 0.0168, "step": 12820 }, { "grad_norm": 0.12979140877723694, "learning_rate": 9.773196365901591e-05, "loss": 0.017, "step": 12830 }, { "grad_norm": 0.11690196394920349, "learning_rate": 9.772653245809894e-05, "loss": 0.0168, "step": 12840 }, { "grad_norm": 0.16695788502693176, "learning_rate": 9.77210949132646e-05, "loss": 0.0125, "step": 12850 }, { "grad_norm": 0.11750418692827225, "learning_rate": 9.771565102523568e-05, "loss": 0.0146, "step": 12860 }, { "grad_norm": 0.21630162000656128, "learning_rate": 9.771020079473579e-05, "loss": 0.016, "step": 12870 }, { "grad_norm": 0.1705772876739502, "learning_rate": 9.770474422248938e-05, "loss": 0.0153, "step": 12880 }, { "grad_norm": 0.1624675989151001, "learning_rate": 9.769928130922174e-05, "loss": 0.0151, "step": 12890 }, { "grad_norm": 0.13740946352481842, "learning_rate": 9.769381205565905e-05, "loss": 0.0154, "step": 12900 }, { "grad_norm": 0.09476683288812637, "learning_rate": 9.768833646252824e-05, "loss": 0.0144, "step": 12910 }, { "grad_norm": 0.1346244066953659, "learning_rate": 9.768285453055719e-05, "loss": 0.0176, "step": 12920 }, { "grad_norm": 0.11476311087608337, "learning_rate": 9.767736626047455e-05, "loss": 0.0141, "step": 12930 }, { "grad_norm": 0.11084911972284317, "learning_rate": 9.767187165300983e-05, "loss": 0.0109, "step": 12940 }, { "grad_norm": 0.18415765464305878, "learning_rate": 9.766637070889339e-05, "loss": 0.0146, "step": 12950 }, { "grad_norm": 0.11389992386102676, "learning_rate": 9.766086342885642e-05, "loss": 0.016, "step": 12960 }, { "grad_norm": 0.15102580189704895, "learning_rate": 9.765534981363097e-05, "loss": 0.015, "step": 12970 }, { "grad_norm": 0.16426347196102142, "learning_rate": 9.764982986394992e-05, "loss": 0.0167, "step": 12980 }, { "grad_norm": 0.14726175367832184, "learning_rate": 9.764430358054699e-05, "loss": 0.0161, "step": 12990 }, { "grad_norm": 0.18921302258968353, "learning_rate": 9.763877096415675e-05, "loss": 0.0156, "step": 13000 }, { "grad_norm": 0.16387130320072174, "learning_rate": 9.76332320155146e-05, "loss": 0.0177, "step": 13010 }, { "grad_norm": 0.21363678574562073, "learning_rate": 9.76276867353568e-05, "loss": 0.0163, "step": 13020 }, { "grad_norm": 0.13296011090278625, "learning_rate": 9.762213512442045e-05, "loss": 0.0179, "step": 13030 }, { "grad_norm": 0.1502024382352829, "learning_rate": 9.761657718344344e-05, "loss": 0.0144, "step": 13040 }, { "grad_norm": 0.0900251716375351, "learning_rate": 9.76110129131646e-05, "loss": 0.0145, "step": 13050 }, { "grad_norm": 0.12078005075454712, "learning_rate": 9.76054423143235e-05, "loss": 0.0171, "step": 13060 }, { "grad_norm": 0.12148915231227875, "learning_rate": 9.759986538766062e-05, "loss": 0.0095, "step": 13070 }, { "grad_norm": 0.08177720755338669, "learning_rate": 9.759428213391725e-05, "loss": 0.0123, "step": 13080 }, { "grad_norm": 0.17816261947155, "learning_rate": 9.758869255383554e-05, "loss": 0.0154, "step": 13090 }, { "grad_norm": 0.14671485126018524, "learning_rate": 9.758309664815845e-05, "loss": 0.0142, "step": 13100 }, { "grad_norm": 0.10288151353597641, "learning_rate": 9.757749441762981e-05, "loss": 0.0149, "step": 13110 }, { "grad_norm": 0.1899467259645462, "learning_rate": 9.757188586299428e-05, "loss": 0.0173, "step": 13120 }, { "grad_norm": 0.1341048926115036, "learning_rate": 9.756627098499737e-05, "loss": 0.018, "step": 13130 }, { "grad_norm": 0.1547253280878067, "learning_rate": 9.756064978438542e-05, "loss": 0.0155, "step": 13140 }, { "grad_norm": 0.14451906085014343, "learning_rate": 9.75550222619056e-05, "loss": 0.0141, "step": 13150 }, { "grad_norm": 0.11399590969085693, "learning_rate": 9.754938841830594e-05, "loss": 0.0116, "step": 13160 }, { "grad_norm": 0.1404435634613037, "learning_rate": 9.75437482543353e-05, "loss": 0.0168, "step": 13170 }, { "grad_norm": 0.15882669389247894, "learning_rate": 9.75381017707434e-05, "loss": 0.0141, "step": 13180 }, { "grad_norm": 0.131956547498703, "learning_rate": 9.753244896828075e-05, "loss": 0.0118, "step": 13190 }, { "grad_norm": 0.1656184047460556, "learning_rate": 9.752678984769879e-05, "loss": 0.0205, "step": 13200 }, { "grad_norm": 0.14150027930736542, "learning_rate": 9.752112440974968e-05, "loss": 0.0136, "step": 13210 }, { "grad_norm": 0.1521233469247818, "learning_rate": 9.75154526551865e-05, "loss": 0.0162, "step": 13220 }, { "grad_norm": 0.09635952115058899, "learning_rate": 9.750977458476318e-05, "loss": 0.0187, "step": 13230 }, { "grad_norm": 0.10984025150537491, "learning_rate": 9.750409019923443e-05, "loss": 0.018, "step": 13240 }, { "grad_norm": 0.09988366067409515, "learning_rate": 9.749839949935585e-05, "loss": 0.018, "step": 13250 }, { "grad_norm": 0.19472336769104004, "learning_rate": 9.749270248588384e-05, "loss": 0.0202, "step": 13260 }, { "grad_norm": 0.21544045209884644, "learning_rate": 9.748699915957571e-05, "loss": 0.0214, "step": 13270 }, { "grad_norm": 0.22026348114013672, "learning_rate": 9.748128952118949e-05, "loss": 0.0155, "step": 13280 }, { "grad_norm": 0.14066345989704132, "learning_rate": 9.747557357148417e-05, "loss": 0.0207, "step": 13290 }, { "grad_norm": 0.1922696828842163, "learning_rate": 9.74698513112195e-05, "loss": 0.0204, "step": 13300 }, { "grad_norm": 0.10688565671443939, "learning_rate": 9.74641227411561e-05, "loss": 0.0127, "step": 13310 }, { "grad_norm": 0.17034465074539185, "learning_rate": 9.745838786205544e-05, "loss": 0.0139, "step": 13320 }, { "grad_norm": 0.10689014941453934, "learning_rate": 9.745264667467979e-05, "loss": 0.0171, "step": 13330 }, { "grad_norm": 0.15569919347763062, "learning_rate": 9.744689917979231e-05, "loss": 0.0163, "step": 13340 }, { "grad_norm": 0.1331900805234909, "learning_rate": 9.744114537815693e-05, "loss": 0.0121, "step": 13350 }, { "grad_norm": 0.10856851935386658, "learning_rate": 9.743538527053849e-05, "loss": 0.0157, "step": 13360 }, { "grad_norm": 0.09278349578380585, "learning_rate": 9.742961885770264e-05, "loss": 0.0135, "step": 13370 }, { "grad_norm": 0.06966540217399597, "learning_rate": 9.742384614041582e-05, "loss": 0.015, "step": 13380 }, { "grad_norm": 0.10370887070894241, "learning_rate": 9.741806711944541e-05, "loss": 0.0176, "step": 13390 }, { "grad_norm": 0.11139868944883347, "learning_rate": 9.741228179555952e-05, "loss": 0.0119, "step": 13400 }, { "grad_norm": 0.10917849093675613, "learning_rate": 9.74064901695272e-05, "loss": 0.0134, "step": 13410 }, { "grad_norm": 0.21166740357875824, "learning_rate": 9.740069224211825e-05, "loss": 0.0138, "step": 13420 }, { "grad_norm": 0.15755893290042877, "learning_rate": 9.739488801410334e-05, "loss": 0.0158, "step": 13430 }, { "grad_norm": 0.19719769060611725, "learning_rate": 9.738907748625401e-05, "loss": 0.0195, "step": 13440 }, { "grad_norm": 0.1761169284582138, "learning_rate": 9.738326065934258e-05, "loss": 0.0182, "step": 13450 }, { "grad_norm": 0.1351499706506729, "learning_rate": 9.737743753414224e-05, "loss": 0.0159, "step": 13460 }, { "grad_norm": 0.08183567970991135, "learning_rate": 9.737160811142703e-05, "loss": 0.0138, "step": 13470 }, { "grad_norm": 0.15815380215644836, "learning_rate": 9.736577239197179e-05, "loss": 0.0125, "step": 13480 }, { "grad_norm": 0.18230362236499786, "learning_rate": 9.735993037655225e-05, "loss": 0.023, "step": 13490 }, { "grad_norm": 0.14431911706924438, "learning_rate": 9.735408206594489e-05, "loss": 0.0176, "step": 13500 }, { "grad_norm": 0.14422321319580078, "learning_rate": 9.734822746092713e-05, "loss": 0.0196, "step": 13510 }, { "grad_norm": 0.1539468616247177, "learning_rate": 9.734236656227716e-05, "loss": 0.012, "step": 13520 }, { "grad_norm": 0.13923034071922302, "learning_rate": 9.7336499370774e-05, "loss": 0.0139, "step": 13530 }, { "grad_norm": 0.12513555586338043, "learning_rate": 9.733062588719757e-05, "loss": 0.0188, "step": 13540 }, { "grad_norm": 0.17566625773906708, "learning_rate": 9.732474611232857e-05, "loss": 0.0163, "step": 13550 }, { "grad_norm": 0.13752558827400208, "learning_rate": 9.731886004694854e-05, "loss": 0.0186, "step": 13560 }, { "grad_norm": 0.20318828523159027, "learning_rate": 9.731296769183989e-05, "loss": 0.0127, "step": 13570 }, { "grad_norm": 0.17126062512397766, "learning_rate": 9.730706904778582e-05, "loss": 0.0183, "step": 13580 }, { "grad_norm": 0.1327134072780609, "learning_rate": 9.730116411557043e-05, "loss": 0.0164, "step": 13590 }, { "grad_norm": 0.12935008108615875, "learning_rate": 9.729525289597857e-05, "loss": 0.0158, "step": 13600 }, { "grad_norm": 0.11287324130535126, "learning_rate": 9.728933538979601e-05, "loss": 0.0164, "step": 13610 }, { "grad_norm": 0.11404147744178772, "learning_rate": 9.72834115978093e-05, "loss": 0.0155, "step": 13620 }, { "grad_norm": 0.13763442635536194, "learning_rate": 9.727748152080584e-05, "loss": 0.0177, "step": 13630 }, { "grad_norm": 0.16292661428451538, "learning_rate": 9.727154515957388e-05, "loss": 0.0189, "step": 13640 }, { "grad_norm": 0.13327611982822418, "learning_rate": 9.726560251490248e-05, "loss": 0.0133, "step": 13650 }, { "grad_norm": 0.15021583437919617, "learning_rate": 9.725965358758158e-05, "loss": 0.0147, "step": 13660 }, { "grad_norm": 0.10089121758937836, "learning_rate": 9.72536983784019e-05, "loss": 0.0124, "step": 13670 }, { "grad_norm": 0.14542436599731445, "learning_rate": 9.724773688815501e-05, "loss": 0.0181, "step": 13680 }, { "grad_norm": 0.20943543314933777, "learning_rate": 9.724176911763336e-05, "loss": 0.0132, "step": 13690 }, { "grad_norm": 0.14412625133991241, "learning_rate": 9.723579506763017e-05, "loss": 0.0156, "step": 13700 }, { "grad_norm": 0.12391990423202515, "learning_rate": 9.722981473893952e-05, "loss": 0.0166, "step": 13710 }, { "grad_norm": 0.2131601721048355, "learning_rate": 9.722382813235635e-05, "loss": 0.0126, "step": 13720 }, { "grad_norm": 0.14163383841514587, "learning_rate": 9.721783524867643e-05, "loss": 0.0156, "step": 13730 }, { "grad_norm": 0.146232470870018, "learning_rate": 9.721183608869628e-05, "loss": 0.014, "step": 13740 }, { "grad_norm": 0.18839721381664276, "learning_rate": 9.72058306532134e-05, "loss": 0.0164, "step": 13750 }, { "grad_norm": 0.1798640936613083, "learning_rate": 9.719981894302598e-05, "loss": 0.0135, "step": 13760 }, { "grad_norm": 0.19354890286922455, "learning_rate": 9.719380095893316e-05, "loss": 0.0126, "step": 13770 }, { "grad_norm": 0.1959683895111084, "learning_rate": 9.718777670173483e-05, "loss": 0.0137, "step": 13780 }, { "grad_norm": 0.1736048460006714, "learning_rate": 9.718174617223177e-05, "loss": 0.0142, "step": 13790 }, { "grad_norm": 0.15461434423923492, "learning_rate": 9.717570937122556e-05, "loss": 0.0178, "step": 13800 }, { "grad_norm": 0.201029971241951, "learning_rate": 9.716966629951863e-05, "loss": 0.0194, "step": 13810 }, { "grad_norm": 0.14460057020187378, "learning_rate": 9.716361695791423e-05, "loss": 0.0158, "step": 13820 }, { "grad_norm": 0.19367650151252747, "learning_rate": 9.715756134721646e-05, "loss": 0.0181, "step": 13830 }, { "grad_norm": 0.11343488842248917, "learning_rate": 9.715149946823026e-05, "loss": 0.0143, "step": 13840 }, { "grad_norm": 0.15990495681762695, "learning_rate": 9.714543132176135e-05, "loss": 0.0134, "step": 13850 }, { "grad_norm": 0.10973229259252548, "learning_rate": 9.713935690861636e-05, "loss": 0.0184, "step": 13860 }, { "grad_norm": 0.11373625695705414, "learning_rate": 9.713327622960267e-05, "loss": 0.012, "step": 13870 }, { "grad_norm": 0.12623679637908936, "learning_rate": 9.71271892855286e-05, "loss": 0.0137, "step": 13880 }, { "grad_norm": 0.1460621953010559, "learning_rate": 9.71210960772032e-05, "loss": 0.015, "step": 13890 }, { "grad_norm": 0.17315031588077545, "learning_rate": 9.711499660543637e-05, "loss": 0.0158, "step": 13900 }, { "grad_norm": 0.17821790277957916, "learning_rate": 9.710889087103892e-05, "loss": 0.0135, "step": 13910 }, { "grad_norm": 0.23294785618782043, "learning_rate": 9.71027788748224e-05, "loss": 0.0187, "step": 13920 }, { "grad_norm": 0.14622975885868073, "learning_rate": 9.709666061759926e-05, "loss": 0.0141, "step": 13930 }, { "grad_norm": 0.1188267320394516, "learning_rate": 9.709053610018273e-05, "loss": 0.0143, "step": 13940 }, { "grad_norm": 0.17928551137447357, "learning_rate": 9.708440532338688e-05, "loss": 0.0181, "step": 13950 }, { "grad_norm": 0.1873060017824173, "learning_rate": 9.707826828802666e-05, "loss": 0.015, "step": 13960 }, { "grad_norm": 0.1343936026096344, "learning_rate": 9.707212499491781e-05, "loss": 0.0136, "step": 13970 }, { "grad_norm": 0.10445930808782578, "learning_rate": 9.70659754448769e-05, "loss": 0.0142, "step": 13980 }, { "grad_norm": 0.14674480259418488, "learning_rate": 9.705981963872135e-05, "loss": 0.0157, "step": 13990 }, { "grad_norm": 0.1460658758878708, "learning_rate": 9.705365757726939e-05, "loss": 0.012, "step": 14000 }, { "grad_norm": 0.10259534418582916, "learning_rate": 9.704748926134011e-05, "loss": 0.0128, "step": 14010 }, { "grad_norm": 0.15460582077503204, "learning_rate": 9.704131469175342e-05, "loss": 0.0146, "step": 14020 }, { "grad_norm": 0.09289385378360748, "learning_rate": 9.703513386933003e-05, "loss": 0.0136, "step": 14030 }, { "grad_norm": 0.11010687053203583, "learning_rate": 9.702894679489154e-05, "loss": 0.0134, "step": 14040 }, { "grad_norm": 0.09635987877845764, "learning_rate": 9.702275346926033e-05, "loss": 0.012, "step": 14050 }, { "grad_norm": 0.12191697955131531, "learning_rate": 9.701655389325965e-05, "loss": 0.0128, "step": 14060 }, { "grad_norm": 0.13215938210487366, "learning_rate": 9.701034806771353e-05, "loss": 0.0135, "step": 14070 }, { "grad_norm": 0.200346440076828, "learning_rate": 9.700413599344691e-05, "loss": 0.0124, "step": 14080 }, { "grad_norm": 0.15999913215637207, "learning_rate": 9.699791767128547e-05, "loss": 0.0142, "step": 14090 }, { "grad_norm": 0.1330602914094925, "learning_rate": 9.699169310205576e-05, "loss": 0.0143, "step": 14100 }, { "grad_norm": 0.11621222645044327, "learning_rate": 9.69854622865852e-05, "loss": 0.0112, "step": 14110 }, { "grad_norm": 0.11369504779577255, "learning_rate": 9.6979225225702e-05, "loss": 0.0128, "step": 14120 }, { "grad_norm": 0.16595244407653809, "learning_rate": 9.697298192023516e-05, "loss": 0.0187, "step": 14130 }, { "grad_norm": 0.13889439404010773, "learning_rate": 9.69667323710146e-05, "loss": 0.0182, "step": 14140 }, { "grad_norm": 0.10890266299247742, "learning_rate": 9.6960476578871e-05, "loss": 0.0129, "step": 14150 }, { "grad_norm": 0.1502906084060669, "learning_rate": 9.695421454463592e-05, "loss": 0.014, "step": 14160 }, { "grad_norm": 0.10809926688671112, "learning_rate": 9.694794626914169e-05, "loss": 0.0161, "step": 14170 }, { "grad_norm": 0.15883736312389374, "learning_rate": 9.694167175322152e-05, "loss": 0.0144, "step": 14180 }, { "grad_norm": 0.08665461093187332, "learning_rate": 9.693539099770944e-05, "loss": 0.0125, "step": 14190 }, { "grad_norm": 0.15500475466251373, "learning_rate": 9.692910400344028e-05, "loss": 0.0135, "step": 14200 }, { "grad_norm": 0.13446924090385437, "learning_rate": 9.692281077124973e-05, "loss": 0.0131, "step": 14210 }, { "grad_norm": 0.11280497163534164, "learning_rate": 9.691651130197432e-05, "loss": 0.0161, "step": 14220 }, { "grad_norm": 0.17027392983436584, "learning_rate": 9.691020559645137e-05, "loss": 0.0173, "step": 14230 }, { "grad_norm": 0.17788052558898926, "learning_rate": 9.690389365551904e-05, "loss": 0.0149, "step": 14240 }, { "grad_norm": 0.16478104889392853, "learning_rate": 9.689757548001635e-05, "loss": 0.0183, "step": 14250 }, { "grad_norm": 0.12123090028762817, "learning_rate": 9.689125107078312e-05, "loss": 0.0173, "step": 14260 }, { "grad_norm": 0.13306500017642975, "learning_rate": 9.688492042866e-05, "loss": 0.0125, "step": 14270 }, { "grad_norm": 0.16796524822711945, "learning_rate": 9.687858355448847e-05, "loss": 0.013, "step": 14280 }, { "grad_norm": 0.14305895566940308, "learning_rate": 9.687224044911085e-05, "loss": 0.0106, "step": 14290 }, { "grad_norm": 0.11985379457473755, "learning_rate": 9.686589111337027e-05, "loss": 0.0143, "step": 14300 }, { "grad_norm": 0.1396239548921585, "learning_rate": 9.68595355481107e-05, "loss": 0.0135, "step": 14310 }, { "grad_norm": 0.13495469093322754, "learning_rate": 9.685317375417696e-05, "loss": 0.0195, "step": 14320 }, { "grad_norm": 0.10488463938236237, "learning_rate": 9.684680573241463e-05, "loss": 0.0184, "step": 14330 }, { "grad_norm": 0.1406366527080536, "learning_rate": 9.684043148367019e-05, "loss": 0.0152, "step": 14340 }, { "grad_norm": 0.10077331960201263, "learning_rate": 9.68340510087909e-05, "loss": 0.0161, "step": 14350 }, { "grad_norm": 0.17212145030498505, "learning_rate": 9.682766430862491e-05, "loss": 0.0168, "step": 14360 }, { "grad_norm": 0.19685308635234833, "learning_rate": 9.68212713840211e-05, "loss": 0.0165, "step": 14370 }, { "grad_norm": 0.14790447056293488, "learning_rate": 9.681487223582928e-05, "loss": 0.0166, "step": 14380 }, { "grad_norm": 0.11069867014884949, "learning_rate": 9.68084668649e-05, "loss": 0.0146, "step": 14390 }, { "grad_norm": 0.13845574855804443, "learning_rate": 9.68020552720847e-05, "loss": 0.0123, "step": 14400 }, { "grad_norm": 0.1429387331008911, "learning_rate": 9.679563745823561e-05, "loss": 0.0181, "step": 14410 }, { "grad_norm": 0.1578810214996338, "learning_rate": 9.678921342420581e-05, "loss": 0.0149, "step": 14420 }, { "grad_norm": 0.18037806451320648, "learning_rate": 9.678278317084919e-05, "loss": 0.0125, "step": 14430 }, { "grad_norm": 0.1695297509431839, "learning_rate": 9.677634669902047e-05, "loss": 0.0127, "step": 14440 }, { "grad_norm": 0.14874596893787384, "learning_rate": 9.676990400957521e-05, "loss": 0.0207, "step": 14450 }, { "grad_norm": 0.12264546751976013, "learning_rate": 9.676345510336979e-05, "loss": 0.0139, "step": 14460 }, { "grad_norm": 0.13327835500240326, "learning_rate": 9.675699998126142e-05, "loss": 0.0181, "step": 14470 }, { "grad_norm": 0.16415950655937195, "learning_rate": 9.67505386441081e-05, "loss": 0.0178, "step": 14480 }, { "grad_norm": 0.113854318857193, "learning_rate": 9.67440710927687e-05, "loss": 0.0149, "step": 14490 }, { "grad_norm": 0.14208346605300903, "learning_rate": 9.67375973281029e-05, "loss": 0.0114, "step": 14500 }, { "grad_norm": 0.13906176388263702, "learning_rate": 9.673111735097122e-05, "loss": 0.0151, "step": 14510 }, { "grad_norm": 0.1479986011981964, "learning_rate": 9.672463116223498e-05, "loss": 0.013, "step": 14520 }, { "grad_norm": 0.20048309862613678, "learning_rate": 9.671813876275632e-05, "loss": 0.0147, "step": 14530 }, { "grad_norm": 0.11143423616886139, "learning_rate": 9.671164015339827e-05, "loss": 0.0153, "step": 14540 }, { "grad_norm": 0.1622883826494217, "learning_rate": 9.670513533502462e-05, "loss": 0.0162, "step": 14550 }, { "grad_norm": 0.13439787924289703, "learning_rate": 9.669862430849999e-05, "loss": 0.0134, "step": 14560 }, { "grad_norm": 0.20635974407196045, "learning_rate": 9.669210707468983e-05, "loss": 0.0153, "step": 14570 }, { "grad_norm": 0.1056271567940712, "learning_rate": 9.668558363446048e-05, "loss": 0.0154, "step": 14580 }, { "grad_norm": 0.11143117398023605, "learning_rate": 9.667905398867899e-05, "loss": 0.0122, "step": 14590 }, { "grad_norm": 0.1370008885860443, "learning_rate": 9.667251813821334e-05, "loss": 0.0156, "step": 14600 }, { "grad_norm": 0.2195812314748764, "learning_rate": 9.666597608393226e-05, "loss": 0.0173, "step": 14610 }, { "grad_norm": 0.14707452058792114, "learning_rate": 9.665942782670533e-05, "loss": 0.0131, "step": 14620 }, { "grad_norm": 0.15059907734394073, "learning_rate": 9.665287336740301e-05, "loss": 0.0182, "step": 14630 }, { "grad_norm": 0.14788644015789032, "learning_rate": 9.664631270689648e-05, "loss": 0.0155, "step": 14640 }, { "grad_norm": 0.1798136830329895, "learning_rate": 9.663974584605781e-05, "loss": 0.0137, "step": 14650 }, { "grad_norm": 0.13354991376399994, "learning_rate": 9.66331727857599e-05, "loss": 0.0142, "step": 14660 }, { "grad_norm": 0.11695249378681183, "learning_rate": 9.662659352687644e-05, "loss": 0.0123, "step": 14670 }, { "grad_norm": 0.13146379590034485, "learning_rate": 9.662000807028196e-05, "loss": 0.0129, "step": 14680 }, { "grad_norm": 0.09417645633220673, "learning_rate": 9.661341641685182e-05, "loss": 0.0121, "step": 14690 }, { "grad_norm": 0.12875592708587646, "learning_rate": 9.660681856746219e-05, "loss": 0.0144, "step": 14700 }, { "grad_norm": 0.13285766541957855, "learning_rate": 9.660021452299007e-05, "loss": 0.0239, "step": 14710 }, { "grad_norm": 0.10902765393257141, "learning_rate": 9.65936042843133e-05, "loss": 0.0148, "step": 14720 }, { "grad_norm": 0.10423196107149124, "learning_rate": 9.65869878523105e-05, "loss": 0.014, "step": 14730 }, { "grad_norm": 0.13590362668037415, "learning_rate": 9.658036522786118e-05, "loss": 0.0188, "step": 14740 }, { "grad_norm": 0.10690294951200485, "learning_rate": 9.657373641184559e-05, "loss": 0.0176, "step": 14750 }, { "grad_norm": 0.16264870762825012, "learning_rate": 9.656710140514489e-05, "loss": 0.0152, "step": 14760 }, { "grad_norm": 0.1490030437707901, "learning_rate": 9.6560460208641e-05, "loss": 0.0185, "step": 14770 }, { "grad_norm": 0.18110597133636475, "learning_rate": 9.655381282321666e-05, "loss": 0.0152, "step": 14780 }, { "grad_norm": 0.11715500056743622, "learning_rate": 9.654715924975548e-05, "loss": 0.0176, "step": 14790 }, { "grad_norm": 0.10428190976381302, "learning_rate": 9.654049948914188e-05, "loss": 0.0101, "step": 14800 }, { "grad_norm": 0.11112337559461594, "learning_rate": 9.653383354226107e-05, "loss": 0.0133, "step": 14810 }, { "grad_norm": 0.12856687605381012, "learning_rate": 9.65271614099991e-05, "loss": 0.0128, "step": 14820 }, { "grad_norm": 0.11232297867536545, "learning_rate": 9.652048309324286e-05, "loss": 0.0136, "step": 14830 }, { "grad_norm": 0.13562750816345215, "learning_rate": 9.651379859288003e-05, "loss": 0.0135, "step": 14840 }, { "grad_norm": 0.09483135491609573, "learning_rate": 9.650710790979914e-05, "loss": 0.0138, "step": 14850 }, { "grad_norm": 0.12360306829214096, "learning_rate": 9.650041104488953e-05, "loss": 0.0139, "step": 14860 }, { "grad_norm": 0.13850560784339905, "learning_rate": 9.649370799904138e-05, "loss": 0.0143, "step": 14870 }, { "grad_norm": 0.1682085543870926, "learning_rate": 9.648699877314562e-05, "loss": 0.0146, "step": 14880 }, { "grad_norm": 0.13662128150463104, "learning_rate": 9.648028336809411e-05, "loss": 0.0146, "step": 14890 }, { "grad_norm": 0.1497308611869812, "learning_rate": 9.647356178477946e-05, "loss": 0.0139, "step": 14900 }, { "grad_norm": 0.11456011235713959, "learning_rate": 9.64668340240951e-05, "loss": 0.0124, "step": 14910 }, { "grad_norm": 0.18163849413394928, "learning_rate": 9.646010008693532e-05, "loss": 0.0204, "step": 14920 }, { "grad_norm": 0.08249898999929428, "learning_rate": 9.64533599741952e-05, "loss": 0.0134, "step": 14930 }, { "grad_norm": 0.13181157410144806, "learning_rate": 9.644661368677066e-05, "loss": 0.0159, "step": 14940 }, { "grad_norm": 0.10477741062641144, "learning_rate": 9.643986122555843e-05, "loss": 0.0136, "step": 14950 }, { "grad_norm": 0.10190045833587646, "learning_rate": 9.643310259145603e-05, "loss": 0.0111, "step": 14960 }, { "grad_norm": 0.1354116052389145, "learning_rate": 9.64263377853619e-05, "loss": 0.0128, "step": 14970 }, { "grad_norm": 0.08748377114534378, "learning_rate": 9.641956680817517e-05, "loss": 0.0121, "step": 14980 }, { "grad_norm": 0.11254963278770447, "learning_rate": 9.641278966079587e-05, "loss": 0.0173, "step": 14990 }, { "grad_norm": 0.16904132068157196, "learning_rate": 9.640600634412486e-05, "loss": 0.0131, "step": 15000 }, { "grad_norm": 0.14433659613132477, "learning_rate": 9.639921685906377e-05, "loss": 0.0133, "step": 15010 }, { "grad_norm": 0.14092221856117249, "learning_rate": 9.639242120651507e-05, "loss": 0.0111, "step": 15020 }, { "grad_norm": 0.17175011336803436, "learning_rate": 9.638561938738206e-05, "loss": 0.0165, "step": 15030 }, { "grad_norm": 0.13550424575805664, "learning_rate": 9.637881140256886e-05, "loss": 0.0156, "step": 15040 }, { "grad_norm": 0.0883827954530716, "learning_rate": 9.63719972529804e-05, "loss": 0.0139, "step": 15050 }, { "grad_norm": 0.1045226976275444, "learning_rate": 9.636517693952243e-05, "loss": 0.012, "step": 15060 }, { "grad_norm": 0.1712837964296341, "learning_rate": 9.635835046310151e-05, "loss": 0.0157, "step": 15070 }, { "grad_norm": 0.12362801283597946, "learning_rate": 9.635151782462506e-05, "loss": 0.0103, "step": 15080 }, { "grad_norm": 0.11777137219905853, "learning_rate": 9.634467902500125e-05, "loss": 0.0131, "step": 15090 }, { "grad_norm": 0.11442451924085617, "learning_rate": 9.633783406513913e-05, "loss": 0.013, "step": 15100 }, { "grad_norm": 0.15757381916046143, "learning_rate": 9.633098294594856e-05, "loss": 0.0165, "step": 15110 }, { "grad_norm": 0.13073886930942535, "learning_rate": 9.632412566834018e-05, "loss": 0.0165, "step": 15120 }, { "grad_norm": 0.14481447637081146, "learning_rate": 9.631726223322548e-05, "loss": 0.0146, "step": 15130 }, { "grad_norm": 0.12705174088478088, "learning_rate": 9.631039264151678e-05, "loss": 0.0122, "step": 15140 }, { "grad_norm": 0.06798305362462997, "learning_rate": 9.63035168941272e-05, "loss": 0.0132, "step": 15150 }, { "grad_norm": 0.12751474976539612, "learning_rate": 9.629663499197065e-05, "loss": 0.0106, "step": 15160 }, { "grad_norm": 0.17450088262557983, "learning_rate": 9.628974693596193e-05, "loss": 0.0129, "step": 15170 }, { "grad_norm": 0.15278978645801544, "learning_rate": 9.628285272701659e-05, "loss": 0.0122, "step": 15180 }, { "grad_norm": 0.1337270587682724, "learning_rate": 9.627595236605102e-05, "loss": 0.0147, "step": 15190 }, { "grad_norm": 0.10605212301015854, "learning_rate": 9.626904585398245e-05, "loss": 0.0153, "step": 15200 }, { "grad_norm": 0.09813129901885986, "learning_rate": 9.626213319172888e-05, "loss": 0.0113, "step": 15210 }, { "grad_norm": 0.12800166010856628, "learning_rate": 9.625521438020919e-05, "loss": 0.0132, "step": 15220 }, { "grad_norm": 0.08782783150672913, "learning_rate": 9.624828942034303e-05, "loss": 0.0122, "step": 15230 }, { "grad_norm": 0.11251337826251984, "learning_rate": 9.624135831305088e-05, "loss": 0.0139, "step": 15240 }, { "grad_norm": 0.10510606318712234, "learning_rate": 9.623442105925403e-05, "loss": 0.0178, "step": 15250 }, { "grad_norm": 0.12176679074764252, "learning_rate": 9.622747765987461e-05, "loss": 0.0118, "step": 15260 }, { "grad_norm": 0.10419471561908722, "learning_rate": 9.622052811583555e-05, "loss": 0.0141, "step": 15270 }, { "grad_norm": 0.13334517180919647, "learning_rate": 9.62135724280606e-05, "loss": 0.0122, "step": 15280 }, { "grad_norm": 0.08366252481937408, "learning_rate": 9.62066105974743e-05, "loss": 0.014, "step": 15290 }, { "grad_norm": 0.08555340021848679, "learning_rate": 9.619964262500206e-05, "loss": 0.0132, "step": 15300 }, { "grad_norm": 0.12852202355861664, "learning_rate": 9.619266851157007e-05, "loss": 0.0147, "step": 15310 }, { "grad_norm": 0.15045779943466187, "learning_rate": 9.618568825810536e-05, "loss": 0.0183, "step": 15320 }, { "grad_norm": 0.1421641856431961, "learning_rate": 9.617870186553574e-05, "loss": 0.0157, "step": 15330 }, { "grad_norm": 0.12059955298900604, "learning_rate": 9.617170933478985e-05, "loss": 0.0148, "step": 15340 }, { "grad_norm": 0.09279285371303558, "learning_rate": 9.616471066679718e-05, "loss": 0.0134, "step": 15350 }, { "grad_norm": 0.17029953002929688, "learning_rate": 9.6157705862488e-05, "loss": 0.0124, "step": 15360 }, { "grad_norm": 0.16595128178596497, "learning_rate": 9.615069492279339e-05, "loss": 0.0123, "step": 15370 }, { "grad_norm": 0.12323315441608429, "learning_rate": 9.614367784864526e-05, "loss": 0.0127, "step": 15380 }, { "grad_norm": 0.11835898458957672, "learning_rate": 9.613665464097636e-05, "loss": 0.0177, "step": 15390 }, { "grad_norm": 0.10898202657699585, "learning_rate": 9.612962530072021e-05, "loss": 0.0142, "step": 15400 }, { "grad_norm": 0.11487638205289841, "learning_rate": 9.612258982881117e-05, "loss": 0.012, "step": 15410 }, { "grad_norm": 0.10510455816984177, "learning_rate": 9.611554822618442e-05, "loss": 0.0163, "step": 15420 }, { "grad_norm": 0.12491632252931595, "learning_rate": 9.610850049377592e-05, "loss": 0.012, "step": 15430 }, { "grad_norm": 0.09094955772161484, "learning_rate": 9.610144663252249e-05, "loss": 0.0129, "step": 15440 }, { "grad_norm": 0.12469124048948288, "learning_rate": 9.609438664336174e-05, "loss": 0.0145, "step": 15450 }, { "grad_norm": 0.11672188341617584, "learning_rate": 9.608732052723213e-05, "loss": 0.0141, "step": 15460 }, { "grad_norm": 0.10936018079519272, "learning_rate": 9.608024828507286e-05, "loss": 0.0137, "step": 15470 }, { "grad_norm": 0.14606337249279022, "learning_rate": 9.607316991782398e-05, "loss": 0.0112, "step": 15480 }, { "grad_norm": 0.12878194451332092, "learning_rate": 9.60660854264264e-05, "loss": 0.0162, "step": 15490 }, { "grad_norm": 0.14385823905467987, "learning_rate": 9.60589948118218e-05, "loss": 0.0119, "step": 15500 }, { "grad_norm": 0.13449744880199432, "learning_rate": 9.605189807495267e-05, "loss": 0.0126, "step": 15510 }, { "grad_norm": 0.13939356803894043, "learning_rate": 9.604479521676232e-05, "loss": 0.0138, "step": 15520 }, { "grad_norm": 0.16228748857975006, "learning_rate": 9.603768623819489e-05, "loss": 0.0207, "step": 15530 }, { "grad_norm": 0.16309474408626556, "learning_rate": 9.603057114019532e-05, "loss": 0.0161, "step": 15540 }, { "grad_norm": 0.14933042228221893, "learning_rate": 9.602344992370936e-05, "loss": 0.0134, "step": 15550 }, { "grad_norm": 0.15449801087379456, "learning_rate": 9.601632258968356e-05, "loss": 0.0124, "step": 15560 }, { "grad_norm": 0.18439215421676636, "learning_rate": 9.600918913906534e-05, "loss": 0.0111, "step": 15570 }, { "grad_norm": 0.10016704350709915, "learning_rate": 9.600204957280286e-05, "loss": 0.0124, "step": 15580 }, { "grad_norm": 0.16417846083641052, "learning_rate": 9.599490389184515e-05, "loss": 0.0133, "step": 15590 }, { "grad_norm": 0.13780531287193298, "learning_rate": 9.598775209714203e-05, "loss": 0.0114, "step": 15600 }, { "grad_norm": 0.13189375400543213, "learning_rate": 9.598059418964411e-05, "loss": 0.0123, "step": 15610 }, { "grad_norm": 0.142081618309021, "learning_rate": 9.597343017030285e-05, "loss": 0.0122, "step": 15620 }, { "grad_norm": 0.1047079861164093, "learning_rate": 9.596626004007051e-05, "loss": 0.009, "step": 15630 }, { "grad_norm": 0.12482710927724838, "learning_rate": 9.595908379990015e-05, "loss": 0.0117, "step": 15640 }, { "grad_norm": 0.1073862835764885, "learning_rate": 9.595190145074566e-05, "loss": 0.0172, "step": 15650 }, { "grad_norm": 0.11540693789720535, "learning_rate": 9.594471299356174e-05, "loss": 0.0151, "step": 15660 }, { "grad_norm": 0.17347964644432068, "learning_rate": 9.593751842930388e-05, "loss": 0.0131, "step": 15670 }, { "grad_norm": 0.24672451615333557, "learning_rate": 9.593031775892841e-05, "loss": 0.0138, "step": 15680 }, { "grad_norm": 0.16282741725444794, "learning_rate": 9.592311098339245e-05, "loss": 0.0159, "step": 15690 }, { "grad_norm": 0.15860438346862793, "learning_rate": 9.591589810365395e-05, "loss": 0.0124, "step": 15700 }, { "grad_norm": 0.11431993544101715, "learning_rate": 9.590867912067165e-05, "loss": 0.0127, "step": 15710 }, { "grad_norm": 0.13955052196979523, "learning_rate": 9.590145403540513e-05, "loss": 0.0142, "step": 15720 }, { "grad_norm": 0.07865919172763824, "learning_rate": 9.589422284881475e-05, "loss": 0.0116, "step": 15730 }, { "grad_norm": 0.148652583360672, "learning_rate": 9.588698556186169e-05, "loss": 0.0135, "step": 15740 }, { "grad_norm": 0.22653265297412872, "learning_rate": 9.587974217550798e-05, "loss": 0.0129, "step": 15750 }, { "grad_norm": 0.17078354954719543, "learning_rate": 9.587249269071637e-05, "loss": 0.0126, "step": 15760 }, { "grad_norm": 0.2207668125629425, "learning_rate": 9.586523710845053e-05, "loss": 0.016, "step": 15770 }, { "grad_norm": 0.09262394160032272, "learning_rate": 9.585797542967487e-05, "loss": 0.014, "step": 15780 }, { "grad_norm": 0.13590382039546967, "learning_rate": 9.585070765535462e-05, "loss": 0.0137, "step": 15790 }, { "grad_norm": 0.1427859514951706, "learning_rate": 9.584343378645586e-05, "loss": 0.0146, "step": 15800 }, { "grad_norm": 0.15742269158363342, "learning_rate": 9.583615382394541e-05, "loss": 0.012, "step": 15810 }, { "grad_norm": 0.0990947037935257, "learning_rate": 9.582886776879095e-05, "loss": 0.0157, "step": 15820 }, { "grad_norm": 0.17884385585784912, "learning_rate": 9.582157562196097e-05, "loss": 0.0137, "step": 15830 }, { "grad_norm": 0.1628747433423996, "learning_rate": 9.581427738442474e-05, "loss": 0.0123, "step": 15840 }, { "grad_norm": 0.07966665178537369, "learning_rate": 9.580697305715239e-05, "loss": 0.0152, "step": 15850 }, { "grad_norm": 0.08510421961545944, "learning_rate": 9.57996626411148e-05, "loss": 0.0104, "step": 15860 }, { "grad_norm": 0.1536339819431305, "learning_rate": 9.579234613728369e-05, "loss": 0.0107, "step": 15870 }, { "grad_norm": 0.09245456010103226, "learning_rate": 9.57850235466316e-05, "loss": 0.017, "step": 15880 }, { "grad_norm": 0.19959698617458344, "learning_rate": 9.577769487013185e-05, "loss": 0.0125, "step": 15890 }, { "grad_norm": 0.1801130175590515, "learning_rate": 9.57703601087586e-05, "loss": 0.0138, "step": 15900 }, { "grad_norm": 0.0754949152469635, "learning_rate": 9.576301926348676e-05, "loss": 0.0156, "step": 15910 }, { "grad_norm": 0.09187955409288406, "learning_rate": 9.575567233529216e-05, "loss": 0.0116, "step": 15920 }, { "grad_norm": 0.17072375118732452, "learning_rate": 9.574831932515131e-05, "loss": 0.0199, "step": 15930 }, { "grad_norm": 0.12595033645629883, "learning_rate": 9.574096023404163e-05, "loss": 0.0124, "step": 15940 }, { "grad_norm": 0.1050480306148529, "learning_rate": 9.573359506294128e-05, "loss": 0.0126, "step": 15950 }, { "grad_norm": 0.18872305750846863, "learning_rate": 9.572622381282927e-05, "loss": 0.0157, "step": 15960 }, { "grad_norm": 0.2603624761104584, "learning_rate": 9.57188464846854e-05, "loss": 0.0154, "step": 15970 }, { "grad_norm": 0.21667389571666718, "learning_rate": 9.571146307949027e-05, "loss": 0.0162, "step": 15980 }, { "grad_norm": 0.1268446147441864, "learning_rate": 9.570407359822531e-05, "loss": 0.0113, "step": 15990 }, { "grad_norm": 0.13667769730091095, "learning_rate": 9.569667804187274e-05, "loss": 0.0126, "step": 16000 }, { "grad_norm": 0.16168120503425598, "learning_rate": 9.56892764114156e-05, "loss": 0.0114, "step": 16010 }, { "grad_norm": 0.15963198244571686, "learning_rate": 9.568186870783773e-05, "loss": 0.0104, "step": 16020 }, { "grad_norm": 0.16294582188129425, "learning_rate": 9.56744549321238e-05, "loss": 0.011, "step": 16030 }, { "grad_norm": 0.08771342784166336, "learning_rate": 9.56670350852592e-05, "loss": 0.0136, "step": 16040 }, { "grad_norm": 0.11540767550468445, "learning_rate": 9.565960916823026e-05, "loss": 0.013, "step": 16050 }, { "grad_norm": 0.1994771510362625, "learning_rate": 9.565217718202403e-05, "loss": 0.0114, "step": 16060 }, { "grad_norm": 0.13675111532211304, "learning_rate": 9.564473912762837e-05, "loss": 0.0135, "step": 16070 }, { "grad_norm": 0.15759330987930298, "learning_rate": 9.563729500603198e-05, "loss": 0.0145, "step": 16080 }, { "grad_norm": 0.12147770822048187, "learning_rate": 9.562984481822433e-05, "loss": 0.014, "step": 16090 }, { "grad_norm": 0.0913865864276886, "learning_rate": 9.562238856519574e-05, "loss": 0.0122, "step": 16100 }, { "grad_norm": 0.12191737443208694, "learning_rate": 9.561492624793731e-05, "loss": 0.0139, "step": 16110 }, { "grad_norm": 0.16803483664989471, "learning_rate": 9.560745786744092e-05, "loss": 0.0135, "step": 16120 }, { "grad_norm": 0.1588539034128189, "learning_rate": 9.559998342469932e-05, "loss": 0.0111, "step": 16130 }, { "grad_norm": 0.16791574656963348, "learning_rate": 9.5592502920706e-05, "loss": 0.0138, "step": 16140 }, { "grad_norm": 0.11383505910634995, "learning_rate": 9.55850163564553e-05, "loss": 0.0122, "step": 16150 }, { "grad_norm": 0.14169113337993622, "learning_rate": 9.557752373294236e-05, "loss": 0.0159, "step": 16160 }, { "grad_norm": 0.13500066101551056, "learning_rate": 9.557002505116308e-05, "loss": 0.011, "step": 16170 }, { "grad_norm": 0.1171693503856659, "learning_rate": 9.556252031211424e-05, "loss": 0.0157, "step": 16180 }, { "grad_norm": 0.11469975858926773, "learning_rate": 9.555500951679338e-05, "loss": 0.0112, "step": 16190 }, { "grad_norm": 0.1651957929134369, "learning_rate": 9.554749266619884e-05, "loss": 0.0144, "step": 16200 }, { "grad_norm": 0.0808137059211731, "learning_rate": 9.553996976132977e-05, "loss": 0.0096, "step": 16210 }, { "grad_norm": 0.10749141126871109, "learning_rate": 9.553244080318615e-05, "loss": 0.0121, "step": 16220 }, { "grad_norm": 0.13922055065631866, "learning_rate": 9.552490579276874e-05, "loss": 0.0137, "step": 16230 }, { "grad_norm": 0.1252659261226654, "learning_rate": 9.55173647310791e-05, "loss": 0.0124, "step": 16240 }, { "grad_norm": 0.1697847694158554, "learning_rate": 9.550981761911964e-05, "loss": 0.0136, "step": 16250 }, { "grad_norm": 0.14188437163829803, "learning_rate": 9.550226445789349e-05, "loss": 0.0148, "step": 16260 }, { "grad_norm": 0.13563939929008484, "learning_rate": 9.549470524840466e-05, "loss": 0.012, "step": 16270 }, { "grad_norm": 0.12919646501541138, "learning_rate": 9.548713999165793e-05, "loss": 0.0123, "step": 16280 }, { "grad_norm": 0.13462862372398376, "learning_rate": 9.547956868865889e-05, "loss": 0.0162, "step": 16290 }, { "grad_norm": 0.15109197795391083, "learning_rate": 9.547199134041394e-05, "loss": 0.0125, "step": 16300 }, { "grad_norm": 0.10736038535833359, "learning_rate": 9.546440794793028e-05, "loss": 0.0163, "step": 16310 }, { "grad_norm": 0.11981872469186783, "learning_rate": 9.545681851221592e-05, "loss": 0.012, "step": 16320 }, { "grad_norm": 0.14343544840812683, "learning_rate": 9.544922303427963e-05, "loss": 0.0152, "step": 16330 }, { "grad_norm": 0.1257682740688324, "learning_rate": 9.544162151513105e-05, "loss": 0.0122, "step": 16340 }, { "grad_norm": 0.12665215134620667, "learning_rate": 9.543401395578059e-05, "loss": 0.0127, "step": 16350 }, { "grad_norm": 0.12074924260377884, "learning_rate": 9.542640035723944e-05, "loss": 0.0145, "step": 16360 }, { "grad_norm": 0.15803857147693634, "learning_rate": 9.541878072051965e-05, "loss": 0.0141, "step": 16370 }, { "grad_norm": 0.14714090526103973, "learning_rate": 9.5411155046634e-05, "loss": 0.0158, "step": 16380 }, { "grad_norm": 0.12435600161552429, "learning_rate": 9.540352333659614e-05, "loss": 0.014, "step": 16390 }, { "grad_norm": 0.12224281579256058, "learning_rate": 9.539588559142047e-05, "loss": 0.0122, "step": 16400 }, { "grad_norm": 0.13325545191764832, "learning_rate": 9.538824181212226e-05, "loss": 0.0159, "step": 16410 }, { "grad_norm": 0.13903139531612396, "learning_rate": 9.538059199971749e-05, "loss": 0.011, "step": 16420 }, { "grad_norm": 0.12535320222377777, "learning_rate": 9.537293615522302e-05, "loss": 0.0107, "step": 16430 }, { "grad_norm": 0.1241285502910614, "learning_rate": 9.536527427965647e-05, "loss": 0.0121, "step": 16440 }, { "grad_norm": 0.10245367884635925, "learning_rate": 9.535760637403628e-05, "loss": 0.0106, "step": 16450 }, { "grad_norm": 0.18222539126873016, "learning_rate": 9.534993243938169e-05, "loss": 0.015, "step": 16460 }, { "grad_norm": 0.10448634624481201, "learning_rate": 9.534225247671271e-05, "loss": 0.0132, "step": 16470 }, { "grad_norm": 0.07525420188903809, "learning_rate": 9.533456648705021e-05, "loss": 0.0128, "step": 16480 }, { "grad_norm": 0.11818958818912506, "learning_rate": 9.532687447141581e-05, "loss": 0.0146, "step": 16490 }, { "grad_norm": 0.12562446296215057, "learning_rate": 9.531917643083195e-05, "loss": 0.0134, "step": 16500 }, { "grad_norm": 0.104550801217556, "learning_rate": 9.531147236632188e-05, "loss": 0.0107, "step": 16510 }, { "grad_norm": 0.07235922664403915, "learning_rate": 9.530376227890964e-05, "loss": 0.0117, "step": 16520 }, { "grad_norm": 0.09574998915195465, "learning_rate": 9.529604616962006e-05, "loss": 0.0103, "step": 16530 }, { "grad_norm": 0.12383521348237991, "learning_rate": 9.528832403947879e-05, "loss": 0.0122, "step": 16540 }, { "grad_norm": 0.09187500178813934, "learning_rate": 9.528059588951228e-05, "loss": 0.016, "step": 16550 }, { "grad_norm": 0.13238021731376648, "learning_rate": 9.527286172074777e-05, "loss": 0.0107, "step": 16560 }, { "grad_norm": 0.11579092592000961, "learning_rate": 9.526512153421329e-05, "loss": 0.0131, "step": 16570 }, { "grad_norm": 0.1641036570072174, "learning_rate": 9.52573753309377e-05, "loss": 0.0145, "step": 16580 }, { "grad_norm": 0.1257411241531372, "learning_rate": 9.524962311195065e-05, "loss": 0.0121, "step": 16590 }, { "grad_norm": 0.1630779206752777, "learning_rate": 9.524186487828255e-05, "loss": 0.0129, "step": 16600 }, { "grad_norm": 0.1577000916004181, "learning_rate": 9.523410063096466e-05, "loss": 0.0149, "step": 16610 }, { "grad_norm": 0.21313786506652832, "learning_rate": 9.522633037102902e-05, "loss": 0.0146, "step": 16620 }, { "grad_norm": 0.14986389875411987, "learning_rate": 9.521855409950849e-05, "loss": 0.0122, "step": 16630 }, { "grad_norm": 0.1282113492488861, "learning_rate": 9.521077181743667e-05, "loss": 0.0119, "step": 16640 }, { "grad_norm": 0.12915416061878204, "learning_rate": 9.520298352584802e-05, "loss": 0.0133, "step": 16650 }, { "grad_norm": 0.12271340936422348, "learning_rate": 9.519518922577779e-05, "loss": 0.0127, "step": 16660 }, { "grad_norm": 0.16792629659175873, "learning_rate": 9.518738891826197e-05, "loss": 0.013, "step": 16670 }, { "grad_norm": 0.08030863106250763, "learning_rate": 9.517958260433746e-05, "loss": 0.0113, "step": 16680 }, { "grad_norm": 0.1038566306233406, "learning_rate": 9.517177028504185e-05, "loss": 0.0132, "step": 16690 }, { "grad_norm": 0.17155350744724274, "learning_rate": 9.516395196141357e-05, "loss": 0.0167, "step": 16700 }, { "grad_norm": 0.11438057571649551, "learning_rate": 9.515612763449186e-05, "loss": 0.0099, "step": 16710 }, { "grad_norm": 0.13415327668190002, "learning_rate": 9.514829730531676e-05, "loss": 0.015, "step": 16720 }, { "grad_norm": 0.11882433295249939, "learning_rate": 9.514046097492907e-05, "loss": 0.0119, "step": 16730 }, { "grad_norm": 0.09993068873882294, "learning_rate": 9.513261864437041e-05, "loss": 0.0097, "step": 16740 }, { "grad_norm": 0.1105896532535553, "learning_rate": 9.512477031468325e-05, "loss": 0.0104, "step": 16750 }, { "grad_norm": 0.14159776270389557, "learning_rate": 9.511691598691075e-05, "loss": 0.0117, "step": 16760 }, { "grad_norm": 0.0942469909787178, "learning_rate": 9.510905566209693e-05, "loss": 0.0153, "step": 16770 }, { "grad_norm": 0.1312398612499237, "learning_rate": 9.510118934128665e-05, "loss": 0.0129, "step": 16780 }, { "grad_norm": 0.15783093869686127, "learning_rate": 9.509331702552547e-05, "loss": 0.0128, "step": 16790 }, { "grad_norm": 0.13466836512088776, "learning_rate": 9.50854387158598e-05, "loss": 0.0134, "step": 16800 }, { "grad_norm": 0.14783242344856262, "learning_rate": 9.507755441333685e-05, "loss": 0.0162, "step": 16810 }, { "grad_norm": 0.11685100197792053, "learning_rate": 9.506966411900466e-05, "loss": 0.0156, "step": 16820 }, { "grad_norm": 0.1000606119632721, "learning_rate": 9.506176783391196e-05, "loss": 0.0116, "step": 16830 }, { "grad_norm": 0.1373935341835022, "learning_rate": 9.505386555910838e-05, "loss": 0.0133, "step": 16840 }, { "grad_norm": 0.10688044130802155, "learning_rate": 9.50459572956443e-05, "loss": 0.0119, "step": 16850 }, { "grad_norm": 0.1790466606616974, "learning_rate": 9.50380430445709e-05, "loss": 0.0122, "step": 16860 }, { "grad_norm": 0.13318131864070892, "learning_rate": 9.503012280694017e-05, "loss": 0.0097, "step": 16870 }, { "grad_norm": 0.11918117851018906, "learning_rate": 9.502219658380487e-05, "loss": 0.0155, "step": 16880 }, { "grad_norm": 0.11057846248149872, "learning_rate": 9.501426437621859e-05, "loss": 0.0174, "step": 16890 }, { "grad_norm": 0.1177368089556694, "learning_rate": 9.500632618523569e-05, "loss": 0.0194, "step": 16900 }, { "grad_norm": 0.17023521661758423, "learning_rate": 9.499838201191131e-05, "loss": 0.0121, "step": 16910 }, { "grad_norm": 0.12731079757213593, "learning_rate": 9.499043185730145e-05, "loss": 0.0129, "step": 16920 }, { "grad_norm": 0.10501370579004288, "learning_rate": 9.498247572246281e-05, "loss": 0.0113, "step": 16930 }, { "grad_norm": 0.14817732572555542, "learning_rate": 9.4974513608453e-05, "loss": 0.0132, "step": 16940 }, { "grad_norm": 0.10805071145296097, "learning_rate": 9.496654551633031e-05, "loss": 0.0108, "step": 16950 }, { "grad_norm": 0.10888286679983139, "learning_rate": 9.49585714471539e-05, "loss": 0.0133, "step": 16960 }, { "grad_norm": 0.11668006330728531, "learning_rate": 9.495059140198368e-05, "loss": 0.0152, "step": 16970 }, { "grad_norm": 0.10836157947778702, "learning_rate": 9.49426053818804e-05, "loss": 0.0111, "step": 16980 }, { "grad_norm": 0.08586586266756058, "learning_rate": 9.493461338790558e-05, "loss": 0.014, "step": 16990 }, { "grad_norm": 0.1377040445804596, "learning_rate": 9.492661542112151e-05, "loss": 0.0173, "step": 17000 }, { "grad_norm": 0.16052591800689697, "learning_rate": 9.491861148259133e-05, "loss": 0.0111, "step": 17010 }, { "grad_norm": 0.15086904168128967, "learning_rate": 9.49106015733789e-05, "loss": 0.0125, "step": 17020 }, { "grad_norm": 0.07667435705661774, "learning_rate": 9.490258569454897e-05, "loss": 0.0116, "step": 17030 }, { "grad_norm": 0.13686802983283997, "learning_rate": 9.489456384716697e-05, "loss": 0.0134, "step": 17040 }, { "grad_norm": 0.09414160251617432, "learning_rate": 9.488653603229922e-05, "loss": 0.012, "step": 17050 }, { "grad_norm": 0.16831353306770325, "learning_rate": 9.487850225101277e-05, "loss": 0.0102, "step": 17060 }, { "grad_norm": 0.1823829710483551, "learning_rate": 9.487046250437553e-05, "loss": 0.0156, "step": 17070 }, { "grad_norm": 0.10759501904249191, "learning_rate": 9.486241679345611e-05, "loss": 0.016, "step": 17080 }, { "grad_norm": 0.1414782702922821, "learning_rate": 9.4854365119324e-05, "loss": 0.0103, "step": 17090 }, { "grad_norm": 0.06701111793518066, "learning_rate": 9.484630748304943e-05, "loss": 0.0151, "step": 17100 }, { "grad_norm": 0.18183445930480957, "learning_rate": 9.483824388570344e-05, "loss": 0.0156, "step": 17110 }, { "grad_norm": 0.13705667853355408, "learning_rate": 9.483017432835786e-05, "loss": 0.013, "step": 17120 }, { "grad_norm": 0.1829671412706375, "learning_rate": 9.482209881208533e-05, "loss": 0.0183, "step": 17130 }, { "grad_norm": 0.06298325210809708, "learning_rate": 9.481401733795925e-05, "loss": 0.0166, "step": 17140 }, { "grad_norm": 0.12562766671180725, "learning_rate": 9.480592990705381e-05, "loss": 0.0113, "step": 17150 }, { "grad_norm": 0.16408772766590118, "learning_rate": 9.479783652044405e-05, "loss": 0.0121, "step": 17160 }, { "grad_norm": 0.17132721841335297, "learning_rate": 9.478973717920574e-05, "loss": 0.0111, "step": 17170 }, { "grad_norm": 0.1852424144744873, "learning_rate": 9.478163188441546e-05, "loss": 0.0143, "step": 17180 }, { "grad_norm": 0.1725136935710907, "learning_rate": 9.477352063715059e-05, "loss": 0.0115, "step": 17190 }, { "grad_norm": 0.15680602192878723, "learning_rate": 9.47654034384893e-05, "loss": 0.0123, "step": 17200 }, { "grad_norm": 0.07009869068861008, "learning_rate": 9.475728028951054e-05, "loss": 0.0139, "step": 17210 }, { "grad_norm": 0.09612014144659042, "learning_rate": 9.474915119129403e-05, "loss": 0.0117, "step": 17220 }, { "grad_norm": 0.1059911847114563, "learning_rate": 9.474101614492037e-05, "loss": 0.0118, "step": 17230 }, { "grad_norm": 0.06305976212024689, "learning_rate": 9.473287515147084e-05, "loss": 0.0122, "step": 17240 }, { "grad_norm": 0.07999637722969055, "learning_rate": 9.472472821202755e-05, "loss": 0.0148, "step": 17250 }, { "grad_norm": 0.11956793814897537, "learning_rate": 9.471657532767346e-05, "loss": 0.0131, "step": 17260 }, { "grad_norm": 0.1087665855884552, "learning_rate": 9.470841649949222e-05, "loss": 0.0137, "step": 17270 }, { "grad_norm": 0.1588239073753357, "learning_rate": 9.470025172856835e-05, "loss": 0.0139, "step": 17280 }, { "grad_norm": 0.14486300945281982, "learning_rate": 9.469208101598713e-05, "loss": 0.0124, "step": 17290 }, { "grad_norm": 0.12286315113306046, "learning_rate": 9.468390436283461e-05, "loss": 0.0148, "step": 17300 }, { "grad_norm": 0.14170007407665253, "learning_rate": 9.467572177019766e-05, "loss": 0.0148, "step": 17310 }, { "grad_norm": 0.10124240815639496, "learning_rate": 9.466753323916392e-05, "loss": 0.0101, "step": 17320 }, { "grad_norm": 0.0873115137219429, "learning_rate": 9.465933877082184e-05, "loss": 0.0131, "step": 17330 }, { "grad_norm": 0.13942520320415497, "learning_rate": 9.465113836626064e-05, "loss": 0.0156, "step": 17340 }, { "grad_norm": 0.12580521404743195, "learning_rate": 9.464293202657033e-05, "loss": 0.0143, "step": 17350 }, { "grad_norm": 0.13485631346702576, "learning_rate": 9.463471975284173e-05, "loss": 0.0141, "step": 17360 }, { "grad_norm": 0.12004990875720978, "learning_rate": 9.462650154616644e-05, "loss": 0.0113, "step": 17370 }, { "grad_norm": 0.12073970586061478, "learning_rate": 9.461827740763681e-05, "loss": 0.0106, "step": 17380 }, { "grad_norm": 0.13552191853523254, "learning_rate": 9.461004733834604e-05, "loss": 0.0115, "step": 17390 }, { "grad_norm": 0.11078450828790665, "learning_rate": 9.460181133938809e-05, "loss": 0.0121, "step": 17400 }, { "grad_norm": 0.14854267239570618, "learning_rate": 9.459356941185768e-05, "loss": 0.0137, "step": 17410 }, { "grad_norm": 0.12860356271266937, "learning_rate": 9.458532155685036e-05, "loss": 0.0139, "step": 17420 }, { "grad_norm": 0.16743378341197968, "learning_rate": 9.457706777546246e-05, "loss": 0.0156, "step": 17430 }, { "grad_norm": 0.16530515253543854, "learning_rate": 9.456880806879109e-05, "loss": 0.0107, "step": 17440 }, { "grad_norm": 0.13771027326583862, "learning_rate": 9.456054243793413e-05, "loss": 0.0121, "step": 17450 }, { "grad_norm": 0.12945955991744995, "learning_rate": 9.45522708839903e-05, "loss": 0.0128, "step": 17460 }, { "grad_norm": 0.1164821907877922, "learning_rate": 9.454399340805904e-05, "loss": 0.0122, "step": 17470 }, { "grad_norm": 0.08481424301862717, "learning_rate": 9.453571001124064e-05, "loss": 0.0134, "step": 17480 }, { "grad_norm": 0.14258743822574615, "learning_rate": 9.452742069463612e-05, "loss": 0.0118, "step": 17490 }, { "grad_norm": 0.10890686511993408, "learning_rate": 9.451912545934733e-05, "loss": 0.0089, "step": 17500 }, { "grad_norm": 0.07910569757223129, "learning_rate": 9.451082430647689e-05, "loss": 0.0095, "step": 17510 }, { "grad_norm": 0.10479875653982162, "learning_rate": 9.450251723712822e-05, "loss": 0.0111, "step": 17520 }, { "grad_norm": 0.125475212931633, "learning_rate": 9.449420425240548e-05, "loss": 0.0116, "step": 17530 }, { "grad_norm": 0.13323816657066345, "learning_rate": 9.448588535341368e-05, "loss": 0.0145, "step": 17540 }, { "grad_norm": 0.15647567808628082, "learning_rate": 9.447756054125858e-05, "loss": 0.0161, "step": 17550 }, { "grad_norm": 0.15321116149425507, "learning_rate": 9.446922981704673e-05, "loss": 0.0123, "step": 17560 }, { "grad_norm": 0.06534083932638168, "learning_rate": 9.446089318188547e-05, "loss": 0.0108, "step": 17570 }, { "grad_norm": 0.15628865361213684, "learning_rate": 9.445255063688292e-05, "loss": 0.0155, "step": 17580 }, { "grad_norm": 0.1347440630197525, "learning_rate": 9.444420218314798e-05, "loss": 0.0104, "step": 17590 }, { "grad_norm": 0.06651027500629425, "learning_rate": 9.443584782179038e-05, "loss": 0.0095, "step": 17600 }, { "grad_norm": 0.12447704374790192, "learning_rate": 9.442748755392057e-05, "loss": 0.0109, "step": 17610 }, { "grad_norm": 0.12854903936386108, "learning_rate": 9.441912138064981e-05, "loss": 0.0111, "step": 17620 }, { "grad_norm": 0.12866131961345673, "learning_rate": 9.441074930309017e-05, "loss": 0.0127, "step": 17630 }, { "grad_norm": 0.11953849345445633, "learning_rate": 9.44023713223545e-05, "loss": 0.0141, "step": 17640 }, { "grad_norm": 0.1249605342745781, "learning_rate": 9.439398743955638e-05, "loss": 0.011, "step": 17650 }, { "grad_norm": 0.1299656480550766, "learning_rate": 9.438559765581023e-05, "loss": 0.0129, "step": 17660 }, { "grad_norm": 0.10081227868795395, "learning_rate": 9.437720197223125e-05, "loss": 0.0183, "step": 17670 }, { "grad_norm": 0.1287946254014969, "learning_rate": 9.43688003899354e-05, "loss": 0.012, "step": 17680 }, { "grad_norm": 0.1309032142162323, "learning_rate": 9.436039291003946e-05, "loss": 0.0113, "step": 17690 }, { "grad_norm": 0.11856762319803238, "learning_rate": 9.435197953366095e-05, "loss": 0.0194, "step": 17700 }, { "grad_norm": 0.13376016914844513, "learning_rate": 9.434356026191819e-05, "loss": 0.0123, "step": 17710 }, { "grad_norm": 0.10630069673061371, "learning_rate": 9.433513509593029e-05, "loss": 0.0122, "step": 17720 }, { "grad_norm": 0.1789953112602234, "learning_rate": 9.432670403681716e-05, "loss": 0.0152, "step": 17730 }, { "grad_norm": 0.14446449279785156, "learning_rate": 9.431826708569946e-05, "loss": 0.0137, "step": 17740 }, { "grad_norm": 0.13635095953941345, "learning_rate": 9.430982424369865e-05, "loss": 0.0104, "step": 17750 }, { "grad_norm": 0.13597936928272247, "learning_rate": 9.430137551193698e-05, "loss": 0.0125, "step": 17760 }, { "grad_norm": 0.10039380192756653, "learning_rate": 9.429292089153748e-05, "loss": 0.0147, "step": 17770 }, { "grad_norm": 0.13022322952747345, "learning_rate": 9.428446038362392e-05, "loss": 0.0138, "step": 17780 }, { "grad_norm": 0.1360592395067215, "learning_rate": 9.427599398932092e-05, "loss": 0.0122, "step": 17790 }, { "grad_norm": 0.13928471505641937, "learning_rate": 9.426752170975386e-05, "loss": 0.0172, "step": 17800 }, { "grad_norm": 0.102018803358078, "learning_rate": 9.425904354604887e-05, "loss": 0.0134, "step": 17810 }, { "grad_norm": 0.1365673989057541, "learning_rate": 9.42505594993329e-05, "loss": 0.011, "step": 17820 }, { "grad_norm": 0.09961608052253723, "learning_rate": 9.424206957073366e-05, "loss": 0.0123, "step": 17830 }, { "grad_norm": 0.10134779661893845, "learning_rate": 9.423357376137967e-05, "loss": 0.015, "step": 17840 }, { "grad_norm": 0.21148373186588287, "learning_rate": 9.422507207240019e-05, "loss": 0.0213, "step": 17850 }, { "grad_norm": 0.11573764681816101, "learning_rate": 9.421656450492529e-05, "loss": 0.0129, "step": 17860 }, { "grad_norm": 0.1553853303194046, "learning_rate": 9.420805106008584e-05, "loss": 0.0118, "step": 17870 }, { "grad_norm": 0.16432109475135803, "learning_rate": 9.419953173901341e-05, "loss": 0.0139, "step": 17880 }, { "grad_norm": 0.08350963890552521, "learning_rate": 9.419100654284045e-05, "loss": 0.0116, "step": 17890 }, { "grad_norm": 0.10198260843753815, "learning_rate": 9.418247547270014e-05, "loss": 0.0142, "step": 17900 }, { "grad_norm": 0.1319095641374588, "learning_rate": 9.417393852972645e-05, "loss": 0.0112, "step": 17910 }, { "grad_norm": 0.07761556655168533, "learning_rate": 9.416539571505413e-05, "loss": 0.0162, "step": 17920 }, { "grad_norm": 0.1300189346075058, "learning_rate": 9.41568470298187e-05, "loss": 0.0142, "step": 17930 }, { "grad_norm": 0.1337466984987259, "learning_rate": 9.414829247515649e-05, "loss": 0.0147, "step": 17940 }, { "grad_norm": 0.09571731090545654, "learning_rate": 9.413973205220457e-05, "loss": 0.0103, "step": 17950 }, { "grad_norm": 0.15340900421142578, "learning_rate": 9.41311657621008e-05, "loss": 0.0137, "step": 17960 }, { "grad_norm": 0.2094734013080597, "learning_rate": 9.412259360598386e-05, "loss": 0.0188, "step": 17970 }, { "grad_norm": 0.14677974581718445, "learning_rate": 9.411401558499318e-05, "loss": 0.0121, "step": 17980 }, { "grad_norm": 0.09144726395606995, "learning_rate": 9.410543170026895e-05, "loss": 0.0161, "step": 17990 }, { "grad_norm": 0.12170079350471497, "learning_rate": 9.409684195295217e-05, "loss": 0.0141, "step": 18000 }, { "grad_norm": 0.14675703644752502, "learning_rate": 9.40882463441846e-05, "loss": 0.0132, "step": 18010 }, { "grad_norm": 0.12936079502105713, "learning_rate": 9.40796448751088e-05, "loss": 0.0172, "step": 18020 }, { "grad_norm": 0.11963652074337006, "learning_rate": 9.407103754686808e-05, "loss": 0.0158, "step": 18030 }, { "grad_norm": 0.08720950037240982, "learning_rate": 9.406242436060657e-05, "loss": 0.0139, "step": 18040 }, { "grad_norm": 0.13274365663528442, "learning_rate": 9.405380531746913e-05, "loss": 0.0159, "step": 18050 }, { "grad_norm": 0.1250864714384079, "learning_rate": 9.404518041860143e-05, "loss": 0.0148, "step": 18060 }, { "grad_norm": 0.0906008780002594, "learning_rate": 9.403654966514992e-05, "loss": 0.0115, "step": 18070 }, { "grad_norm": 0.1399320513010025, "learning_rate": 9.402791305826181e-05, "loss": 0.0123, "step": 18080 }, { "grad_norm": 0.11261443793773651, "learning_rate": 9.401927059908509e-05, "loss": 0.011, "step": 18090 }, { "grad_norm": 0.12063785642385483, "learning_rate": 9.401062228876856e-05, "loss": 0.0128, "step": 18100 }, { "grad_norm": 0.11437803506851196, "learning_rate": 9.400196812846174e-05, "loss": 0.0128, "step": 18110 }, { "grad_norm": 0.08361956477165222, "learning_rate": 9.399330811931498e-05, "loss": 0.0136, "step": 18120 }, { "grad_norm": 0.07465573400259018, "learning_rate": 9.398464226247938e-05, "loss": 0.0096, "step": 18130 }, { "grad_norm": 0.09686412662267685, "learning_rate": 9.397597055910685e-05, "loss": 0.0154, "step": 18140 }, { "grad_norm": 0.11119115352630615, "learning_rate": 9.396729301035001e-05, "loss": 0.0095, "step": 18150 }, { "grad_norm": 0.12715302407741547, "learning_rate": 9.395860961736234e-05, "loss": 0.0133, "step": 18160 }, { "grad_norm": 0.07523179054260254, "learning_rate": 9.394992038129802e-05, "loss": 0.0103, "step": 18170 }, { "grad_norm": 0.10688462108373642, "learning_rate": 9.394122530331206e-05, "loss": 0.0151, "step": 18180 }, { "grad_norm": 0.11355981975793839, "learning_rate": 9.393252438456024e-05, "loss": 0.0127, "step": 18190 }, { "grad_norm": 0.09153039753437042, "learning_rate": 9.392381762619909e-05, "loss": 0.0096, "step": 18200 }, { "grad_norm": 0.14339634776115417, "learning_rate": 9.391510502938594e-05, "loss": 0.0084, "step": 18210 }, { "grad_norm": 0.1357395052909851, "learning_rate": 9.390638659527888e-05, "loss": 0.0121, "step": 18220 }, { "grad_norm": 0.10951324552297592, "learning_rate": 9.389766232503679e-05, "loss": 0.0132, "step": 18230 }, { "grad_norm": 0.1070990338921547, "learning_rate": 9.388893221981932e-05, "loss": 0.0121, "step": 18240 }, { "grad_norm": 0.10128344595432281, "learning_rate": 9.388019628078689e-05, "loss": 0.0134, "step": 18250 }, { "grad_norm": 0.09857170283794403, "learning_rate": 9.387145450910068e-05, "loss": 0.0112, "step": 18260 }, { "grad_norm": 0.11914873123168945, "learning_rate": 9.386270690592272e-05, "loss": 0.0127, "step": 18270 }, { "grad_norm": 0.07601271569728851, "learning_rate": 9.385395347241571e-05, "loss": 0.0094, "step": 18280 }, { "grad_norm": 0.17697326838970184, "learning_rate": 9.38451942097432e-05, "loss": 0.0151, "step": 18290 }, { "grad_norm": 0.1892663538455963, "learning_rate": 9.383642911906949e-05, "loss": 0.0148, "step": 18300 }, { "grad_norm": 0.08445852994918823, "learning_rate": 9.382765820155966e-05, "loss": 0.0114, "step": 18310 }, { "grad_norm": 0.10364193469285965, "learning_rate": 9.381888145837954e-05, "loss": 0.0091, "step": 18320 }, { "grad_norm": 0.1501377671957016, "learning_rate": 9.381009889069576e-05, "loss": 0.0138, "step": 18330 }, { "grad_norm": 0.16116918623447418, "learning_rate": 9.380131049967575e-05, "loss": 0.0121, "step": 18340 }, { "grad_norm": 0.13822045922279358, "learning_rate": 9.379251628648764e-05, "loss": 0.0096, "step": 18350 }, { "grad_norm": 0.17319147288799286, "learning_rate": 9.37837162523004e-05, "loss": 0.0168, "step": 18360 }, { "grad_norm": 0.15856240689754486, "learning_rate": 9.377491039828374e-05, "loss": 0.0166, "step": 18370 }, { "grad_norm": 0.17579546570777893, "learning_rate": 9.376609872560816e-05, "loss": 0.0123, "step": 18380 }, { "grad_norm": 0.09568292647600174, "learning_rate": 9.375728123544492e-05, "loss": 0.0129, "step": 18390 }, { "grad_norm": 0.09573199599981308, "learning_rate": 9.374845792896605e-05, "loss": 0.0147, "step": 18400 }, { "grad_norm": 0.09775874018669128, "learning_rate": 9.37396288073444e-05, "loss": 0.0151, "step": 18410 }, { "grad_norm": 0.12910211086273193, "learning_rate": 9.373079387175353e-05, "loss": 0.0124, "step": 18420 }, { "grad_norm": 0.09617377817630768, "learning_rate": 9.372195312336781e-05, "loss": 0.0133, "step": 18430 }, { "grad_norm": 0.16188037395477295, "learning_rate": 9.371310656336236e-05, "loss": 0.0122, "step": 18440 }, { "grad_norm": 0.11792804300785065, "learning_rate": 9.370425419291309e-05, "loss": 0.0126, "step": 18450 }, { "grad_norm": 0.14940567314624786, "learning_rate": 9.369539601319669e-05, "loss": 0.0156, "step": 18460 }, { "grad_norm": 0.14439795911312103, "learning_rate": 9.368653202539059e-05, "loss": 0.0102, "step": 18470 }, { "grad_norm": 0.09173562377691269, "learning_rate": 9.367766223067303e-05, "loss": 0.0101, "step": 18480 }, { "grad_norm": 0.1077805832028389, "learning_rate": 9.366878663022297e-05, "loss": 0.0138, "step": 18490 }, { "grad_norm": 0.11713068187236786, "learning_rate": 9.365990522522021e-05, "loss": 0.0137, "step": 18500 }, { "grad_norm": 0.11394604295492172, "learning_rate": 9.365101801684527e-05, "loss": 0.0164, "step": 18510 }, { "grad_norm": 0.13679073750972748, "learning_rate": 9.364212500627945e-05, "loss": 0.0124, "step": 18520 }, { "grad_norm": 0.10125166922807693, "learning_rate": 9.363322619470484e-05, "loss": 0.0151, "step": 18530 }, { "grad_norm": 0.1270345002412796, "learning_rate": 9.362432158330432e-05, "loss": 0.0184, "step": 18540 }, { "grad_norm": 0.10831897705793381, "learning_rate": 9.361541117326144e-05, "loss": 0.0115, "step": 18550 }, { "grad_norm": 0.11369405686855316, "learning_rate": 9.360649496576066e-05, "loss": 0.0154, "step": 18560 }, { "grad_norm": 0.10094425827264786, "learning_rate": 9.35975729619871e-05, "loss": 0.0136, "step": 18570 }, { "grad_norm": 0.09003140032291412, "learning_rate": 9.358864516312671e-05, "loss": 0.0103, "step": 18580 }, { "grad_norm": 0.2199879288673401, "learning_rate": 9.357971157036618e-05, "loss": 0.0125, "step": 18590 }, { "grad_norm": 0.1057523787021637, "learning_rate": 9.3570772184893e-05, "loss": 0.0124, "step": 18600 }, { "grad_norm": 0.09955661743879318, "learning_rate": 9.35618270078954e-05, "loss": 0.0156, "step": 18610 }, { "grad_norm": 0.10845330357551575, "learning_rate": 9.35528760405624e-05, "loss": 0.0104, "step": 18620 }, { "grad_norm": 0.15951205790042877, "learning_rate": 9.354391928408379e-05, "loss": 0.0124, "step": 18630 }, { "grad_norm": 0.12411057204008102, "learning_rate": 9.35349567396501e-05, "loss": 0.0093, "step": 18640 }, { "grad_norm": 0.09079435467720032, "learning_rate": 9.352598840845266e-05, "loss": 0.0117, "step": 18650 }, { "grad_norm": 0.11308930069208145, "learning_rate": 9.351701429168358e-05, "loss": 0.0121, "step": 18660 }, { "grad_norm": 0.08803748339414597, "learning_rate": 9.350803439053571e-05, "loss": 0.0199, "step": 18670 }, { "grad_norm": 0.091986745595932, "learning_rate": 9.349904870620265e-05, "loss": 0.0134, "step": 18680 }, { "grad_norm": 0.09487676620483398, "learning_rate": 9.349005723987883e-05, "loss": 0.0144, "step": 18690 }, { "grad_norm": 0.0995766818523407, "learning_rate": 9.34810599927594e-05, "loss": 0.0148, "step": 18700 }, { "grad_norm": 0.10361775010824203, "learning_rate": 9.347205696604031e-05, "loss": 0.0135, "step": 18710 }, { "grad_norm": 0.14565448462963104, "learning_rate": 9.346304816091826e-05, "loss": 0.0123, "step": 18720 }, { "grad_norm": 0.09941337257623672, "learning_rate": 9.34540335785907e-05, "loss": 0.0113, "step": 18730 }, { "grad_norm": 0.07942698895931244, "learning_rate": 9.344501322025589e-05, "loss": 0.0129, "step": 18740 }, { "grad_norm": 0.07837896794080734, "learning_rate": 9.343598708711281e-05, "loss": 0.0092, "step": 18750 }, { "grad_norm": 0.07124246656894684, "learning_rate": 9.342695518036128e-05, "loss": 0.0117, "step": 18760 }, { "grad_norm": 0.14290401339530945, "learning_rate": 9.34179175012018e-05, "loss": 0.0121, "step": 18770 }, { "grad_norm": 0.1935611069202423, "learning_rate": 9.340887405083568e-05, "loss": 0.0135, "step": 18780 }, { "grad_norm": 1.0653661489486694, "learning_rate": 9.339982483046502e-05, "loss": 0.0164, "step": 18790 }, { "grad_norm": 0.1998583823442459, "learning_rate": 9.339076984129264e-05, "loss": 0.0308, "step": 18800 }, { "grad_norm": 0.15936166048049927, "learning_rate": 9.338170908452217e-05, "loss": 0.0206, "step": 18810 }, { "grad_norm": 0.11521827429533005, "learning_rate": 9.337264256135798e-05, "loss": 0.017, "step": 18820 }, { "grad_norm": 0.16054867208003998, "learning_rate": 9.336357027300519e-05, "loss": 0.0139, "step": 18830 }, { "grad_norm": 0.10225139558315277, "learning_rate": 9.335449222066973e-05, "loss": 0.0113, "step": 18840 }, { "grad_norm": 0.11230377852916718, "learning_rate": 9.334540840555828e-05, "loss": 0.0132, "step": 18850 }, { "grad_norm": 0.15036237239837646, "learning_rate": 9.333631882887826e-05, "loss": 0.0107, "step": 18860 }, { "grad_norm": 0.12456382066011429, "learning_rate": 9.33272234918379e-05, "loss": 0.0114, "step": 18870 }, { "grad_norm": 0.08922632038593292, "learning_rate": 9.331812239564615e-05, "loss": 0.0127, "step": 18880 }, { "grad_norm": 0.11001817882061005, "learning_rate": 9.330901554151278e-05, "loss": 0.0096, "step": 18890 }, { "grad_norm": 0.11649661511182785, "learning_rate": 9.329990293064825e-05, "loss": 0.0137, "step": 18900 }, { "grad_norm": 0.2060580551624298, "learning_rate": 9.329078456426388e-05, "loss": 0.0124, "step": 18910 }, { "grad_norm": 0.17463190853595734, "learning_rate": 9.328166044357165e-05, "loss": 0.0158, "step": 18920 }, { "grad_norm": 0.12299712002277374, "learning_rate": 9.32725305697844e-05, "loss": 0.012, "step": 18930 }, { "grad_norm": 0.06393668055534363, "learning_rate": 9.326339494411566e-05, "loss": 0.0123, "step": 18940 }, { "grad_norm": 0.07199973613023758, "learning_rate": 9.32542535677798e-05, "loss": 0.01, "step": 18950 }, { "grad_norm": 0.09362590312957764, "learning_rate": 9.324510644199186e-05, "loss": 0.0141, "step": 18960 }, { "grad_norm": 0.1029452532529831, "learning_rate": 9.323595356796775e-05, "loss": 0.0112, "step": 18970 }, { "grad_norm": 0.055100876837968826, "learning_rate": 9.322679494692403e-05, "loss": 0.0113, "step": 18980 }, { "grad_norm": 0.08388252556324005, "learning_rate": 9.321763058007816e-05, "loss": 0.0082, "step": 18990 }, { "grad_norm": 0.12510955333709717, "learning_rate": 9.320846046864822e-05, "loss": 0.0106, "step": 19000 }, { "grad_norm": 0.13326601684093475, "learning_rate": 9.319928461385314e-05, "loss": 0.0119, "step": 19010 }, { "grad_norm": 0.13003067672252655, "learning_rate": 9.319010301691262e-05, "loss": 0.0142, "step": 19020 }, { "grad_norm": 0.11528167128562927, "learning_rate": 9.318091567904708e-05, "loss": 0.0157, "step": 19030 }, { "grad_norm": 0.11891065537929535, "learning_rate": 9.317172260147772e-05, "loss": 0.012, "step": 19040 }, { "grad_norm": 0.06550620496273041, "learning_rate": 9.316252378542652e-05, "loss": 0.0108, "step": 19050 }, { "grad_norm": 0.11575612425804138, "learning_rate": 9.315331923211619e-05, "loss": 0.0102, "step": 19060 }, { "grad_norm": 0.1407509297132492, "learning_rate": 9.314410894277021e-05, "loss": 0.0134, "step": 19070 }, { "grad_norm": 0.09109805524349213, "learning_rate": 9.313489291861286e-05, "loss": 0.0115, "step": 19080 }, { "grad_norm": 0.13921038806438446, "learning_rate": 9.312567116086914e-05, "loss": 0.0115, "step": 19090 }, { "grad_norm": 0.1014690026640892, "learning_rate": 9.311644367076482e-05, "loss": 0.0122, "step": 19100 }, { "grad_norm": 0.08732319623231888, "learning_rate": 9.310721044952645e-05, "loss": 0.012, "step": 19110 }, { "grad_norm": 0.1052546575665474, "learning_rate": 9.309797149838133e-05, "loss": 0.0102, "step": 19120 }, { "grad_norm": 0.08180686086416245, "learning_rate": 9.308872681855752e-05, "loss": 0.0133, "step": 19130 }, { "grad_norm": 0.13135790824890137, "learning_rate": 9.307947641128384e-05, "loss": 0.0128, "step": 19140 }, { "grad_norm": 0.16308791935443878, "learning_rate": 9.307022027778989e-05, "loss": 0.014, "step": 19150 }, { "grad_norm": 0.18486811220645905, "learning_rate": 9.306095841930598e-05, "loss": 0.0147, "step": 19160 }, { "grad_norm": 0.08936220407485962, "learning_rate": 9.305169083706326e-05, "loss": 0.0138, "step": 19170 }, { "grad_norm": 0.17666545510292053, "learning_rate": 9.304241753229356e-05, "loss": 0.0119, "step": 19180 }, { "grad_norm": 0.1878988891839981, "learning_rate": 9.303313850622955e-05, "loss": 0.0124, "step": 19190 }, { "grad_norm": 0.14667300879955292, "learning_rate": 9.302385376010457e-05, "loss": 0.0131, "step": 19200 }, { "grad_norm": 0.15862174332141876, "learning_rate": 9.30145632951528e-05, "loss": 0.0103, "step": 19210 }, { "grad_norm": 0.16124840080738068, "learning_rate": 9.300526711260917e-05, "loss": 0.0119, "step": 19220 }, { "grad_norm": 0.08629567921161652, "learning_rate": 9.29959652137093e-05, "loss": 0.0146, "step": 19230 }, { "grad_norm": 0.07619472593069077, "learning_rate": 9.298665759968965e-05, "loss": 0.0117, "step": 19240 }, { "grad_norm": 0.16241998970508575, "learning_rate": 9.297734427178739e-05, "loss": 0.0138, "step": 19250 }, { "grad_norm": 0.14068040251731873, "learning_rate": 9.296802523124049e-05, "loss": 0.0123, "step": 19260 }, { "grad_norm": 0.16501498222351074, "learning_rate": 9.295870047928764e-05, "loss": 0.0127, "step": 19270 }, { "grad_norm": 0.12209724634885788, "learning_rate": 9.294937001716832e-05, "loss": 0.0145, "step": 19280 }, { "grad_norm": 0.10324767976999283, "learning_rate": 9.294003384612275e-05, "loss": 0.0131, "step": 19290 }, { "grad_norm": 0.1032118946313858, "learning_rate": 9.293069196739191e-05, "loss": 0.0128, "step": 19300 }, { "grad_norm": 0.12758700549602509, "learning_rate": 9.292134438221755e-05, "loss": 0.0108, "step": 19310 }, { "grad_norm": 0.11538774520158768, "learning_rate": 9.291199109184219e-05, "loss": 0.0102, "step": 19320 }, { "grad_norm": 0.07805410027503967, "learning_rate": 9.290263209750905e-05, "loss": 0.0124, "step": 19330 }, { "grad_norm": 0.09848297387361526, "learning_rate": 9.289326740046216e-05, "loss": 0.0132, "step": 19340 }, { "grad_norm": 0.10501472651958466, "learning_rate": 9.288389700194632e-05, "loss": 0.0131, "step": 19350 }, { "grad_norm": 0.09409178048372269, "learning_rate": 9.287452090320705e-05, "loss": 0.0108, "step": 19360 }, { "grad_norm": 0.09932349622249603, "learning_rate": 9.286513910549063e-05, "loss": 0.0112, "step": 19370 }, { "grad_norm": 0.06155438721179962, "learning_rate": 9.285575161004415e-05, "loss": 0.0113, "step": 19380 }, { "grad_norm": 0.09682013094425201, "learning_rate": 9.284635841811535e-05, "loss": 0.0139, "step": 19390 }, { "grad_norm": 0.09642504155635834, "learning_rate": 9.283695953095287e-05, "loss": 0.0101, "step": 19400 }, { "grad_norm": 0.12595263123512268, "learning_rate": 9.282755494980595e-05, "loss": 0.0134, "step": 19410 }, { "grad_norm": 0.11846555024385452, "learning_rate": 9.281814467592473e-05, "loss": 0.0109, "step": 19420 }, { "grad_norm": 0.11164312064647675, "learning_rate": 9.280872871056003e-05, "loss": 0.011, "step": 19430 }, { "grad_norm": 0.09623057395219803, "learning_rate": 9.279930705496345e-05, "loss": 0.0121, "step": 19440 }, { "grad_norm": 0.1210617646574974, "learning_rate": 9.278987971038731e-05, "loss": 0.0102, "step": 19450 }, { "grad_norm": 0.0530785396695137, "learning_rate": 9.278044667808471e-05, "loss": 0.0095, "step": 19460 }, { "grad_norm": 0.0806111991405487, "learning_rate": 9.277100795930956e-05, "loss": 0.0134, "step": 19470 }, { "grad_norm": 0.09271027892827988, "learning_rate": 9.276156355531643e-05, "loss": 0.0159, "step": 19480 }, { "grad_norm": 0.10190948098897934, "learning_rate": 9.27521134673607e-05, "loss": 0.016, "step": 19490 }, { "grad_norm": 0.11085374653339386, "learning_rate": 9.274265769669852e-05, "loss": 0.0107, "step": 19500 }, { "grad_norm": 0.1165449321269989, "learning_rate": 9.273319624458674e-05, "loss": 0.0129, "step": 19510 }, { "grad_norm": 0.11958472430706024, "learning_rate": 9.272372911228301e-05, "loss": 0.0095, "step": 19520 }, { "grad_norm": 0.1343296766281128, "learning_rate": 9.271425630104572e-05, "loss": 0.0145, "step": 19530 }, { "grad_norm": 0.11442488431930542, "learning_rate": 9.270477781213402e-05, "loss": 0.0105, "step": 19540 }, { "grad_norm": 0.14975585043430328, "learning_rate": 9.269529364680782e-05, "loss": 0.0161, "step": 19550 }, { "grad_norm": 0.13132824003696442, "learning_rate": 9.268580380632777e-05, "loss": 0.0147, "step": 19560 }, { "grad_norm": 0.12311417609453201, "learning_rate": 9.267630829195528e-05, "loss": 0.0145, "step": 19570 }, { "grad_norm": 0.11950835585594177, "learning_rate": 9.266680710495252e-05, "loss": 0.0118, "step": 19580 }, { "grad_norm": 0.09869784116744995, "learning_rate": 9.265730024658239e-05, "loss": 0.0103, "step": 19590 }, { "grad_norm": 0.0952729806303978, "learning_rate": 9.264778771810858e-05, "loss": 0.0105, "step": 19600 }, { "grad_norm": 0.12430835515260696, "learning_rate": 9.263826952079552e-05, "loss": 0.0091, "step": 19610 }, { "grad_norm": 0.11421866714954376, "learning_rate": 9.262874565590838e-05, "loss": 0.0121, "step": 19620 }, { "grad_norm": 0.10340642929077148, "learning_rate": 9.26192161247131e-05, "loss": 0.0098, "step": 19630 }, { "grad_norm": 0.12892182171344757, "learning_rate": 9.260968092847634e-05, "loss": 0.0104, "step": 19640 }, { "grad_norm": 0.05902743339538574, "learning_rate": 9.26001400684656e-05, "loss": 0.0098, "step": 19650 }, { "grad_norm": 0.10814337432384491, "learning_rate": 9.259059354594901e-05, "loss": 0.0141, "step": 19660 }, { "grad_norm": 0.20138108730316162, "learning_rate": 9.258104136219556e-05, "loss": 0.0149, "step": 19670 }, { "grad_norm": 0.14777441322803497, "learning_rate": 9.257148351847491e-05, "loss": 0.0103, "step": 19680 }, { "grad_norm": 0.13565103709697723, "learning_rate": 9.256192001605753e-05, "loss": 0.0105, "step": 19690 }, { "grad_norm": 0.12502361834049225, "learning_rate": 9.255235085621464e-05, "loss": 0.0194, "step": 19700 }, { "grad_norm": 0.11379539966583252, "learning_rate": 9.254277604021815e-05, "loss": 0.011, "step": 19710 }, { "grad_norm": 0.1187247708439827, "learning_rate": 9.25331955693408e-05, "loss": 0.0132, "step": 19720 }, { "grad_norm": 0.07226172834634781, "learning_rate": 9.252360944485603e-05, "loss": 0.0131, "step": 19730 }, { "grad_norm": 0.1412927210330963, "learning_rate": 9.251401766803807e-05, "loss": 0.0134, "step": 19740 }, { "grad_norm": 0.14134253561496735, "learning_rate": 9.250442024016186e-05, "loss": 0.0135, "step": 19750 }, { "grad_norm": 0.0720527172088623, "learning_rate": 9.249481716250312e-05, "loss": 0.0116, "step": 19760 }, { "grad_norm": 0.09439201653003693, "learning_rate": 9.248520843633832e-05, "loss": 0.0112, "step": 19770 }, { "grad_norm": 0.1002228707075119, "learning_rate": 9.247559406294465e-05, "loss": 0.0138, "step": 19780 }, { "grad_norm": 0.11687085032463074, "learning_rate": 9.24659740436001e-05, "loss": 0.0158, "step": 19790 }, { "grad_norm": 0.12804436683654785, "learning_rate": 9.245634837958337e-05, "loss": 0.0103, "step": 19800 }, { "grad_norm": 0.10665424913167953, "learning_rate": 9.244671707217391e-05, "loss": 0.0116, "step": 19810 }, { "grad_norm": 0.13359318673610687, "learning_rate": 9.243708012265197e-05, "loss": 0.0111, "step": 19820 }, { "grad_norm": 0.08487138897180557, "learning_rate": 9.24274375322985e-05, "loss": 0.0136, "step": 19830 }, { "grad_norm": 0.11740714311599731, "learning_rate": 9.241778930239522e-05, "loss": 0.0176, "step": 19840 }, { "grad_norm": 0.18885816633701324, "learning_rate": 9.240813543422457e-05, "loss": 0.0154, "step": 19850 }, { "grad_norm": 0.16206206381320953, "learning_rate": 9.23984759290698e-05, "loss": 0.0135, "step": 19860 }, { "grad_norm": 0.20700497925281525, "learning_rate": 9.238881078821483e-05, "loss": 0.0124, "step": 19870 }, { "grad_norm": 0.10391691327095032, "learning_rate": 9.237914001294441e-05, "loss": 0.0107, "step": 19880 }, { "grad_norm": 0.11267366260290146, "learning_rate": 9.236946360454398e-05, "loss": 0.0123, "step": 19890 }, { "grad_norm": 0.11368998140096664, "learning_rate": 9.235978156429976e-05, "loss": 0.0121, "step": 19900 }, { "grad_norm": 0.10192184150218964, "learning_rate": 9.235009389349871e-05, "loss": 0.0095, "step": 19910 }, { "grad_norm": 0.1347690373659134, "learning_rate": 9.234040059342852e-05, "loss": 0.0105, "step": 19920 }, { "grad_norm": 0.06493555009365082, "learning_rate": 9.233070166537767e-05, "loss": 0.0089, "step": 19930 }, { "grad_norm": 0.11350279301404953, "learning_rate": 9.232099711063533e-05, "loss": 0.0094, "step": 19940 }, { "grad_norm": 0.11022733896970749, "learning_rate": 9.23112869304915e-05, "loss": 0.0117, "step": 19950 }, { "grad_norm": 0.10586649924516678, "learning_rate": 9.230157112623684e-05, "loss": 0.0105, "step": 19960 }, { "grad_norm": 0.12422637641429901, "learning_rate": 9.229184969916278e-05, "loss": 0.0124, "step": 19970 }, { "grad_norm": 0.09758557379245758, "learning_rate": 9.228212265056158e-05, "loss": 0.0099, "step": 19980 }, { "grad_norm": 0.14796660840511322, "learning_rate": 9.227238998172611e-05, "loss": 0.0124, "step": 19990 }, { "grad_norm": 0.1382729411125183, "learning_rate": 9.22626516939501e-05, "loss": 0.0097, "step": 20000 }, { "grad_norm": 0.09884510189294815, "learning_rate": 9.225290778852798e-05, "loss": 0.0112, "step": 20010 }, { "grad_norm": 0.0762394592165947, "learning_rate": 9.224315826675493e-05, "loss": 0.01, "step": 20020 }, { "grad_norm": 0.06306883692741394, "learning_rate": 9.223340312992686e-05, "loss": 0.0096, "step": 20030 }, { "grad_norm": 0.10509440302848816, "learning_rate": 9.222364237934046e-05, "loss": 0.0119, "step": 20040 }, { "grad_norm": 0.14806804060935974, "learning_rate": 9.221387601629316e-05, "loss": 0.0103, "step": 20050 }, { "grad_norm": 0.11242355406284332, "learning_rate": 9.220410404208311e-05, "loss": 0.0139, "step": 20060 }, { "grad_norm": 0.12264809757471085, "learning_rate": 9.219432645800923e-05, "loss": 0.0111, "step": 20070 }, { "grad_norm": 0.09110995382070541, "learning_rate": 9.218454326537118e-05, "loss": 0.0106, "step": 20080 }, { "grad_norm": 0.1657199263572693, "learning_rate": 9.217475446546936e-05, "loss": 0.013, "step": 20090 }, { "grad_norm": 0.12626568973064423, "learning_rate": 9.216496005960492e-05, "loss": 0.0101, "step": 20100 }, { "grad_norm": 0.13210059702396393, "learning_rate": 9.215516004907976e-05, "loss": 0.0115, "step": 20110 }, { "grad_norm": 0.0524178110063076, "learning_rate": 9.214535443519649e-05, "loss": 0.0094, "step": 20120 }, { "grad_norm": 0.14180269837379456, "learning_rate": 9.213554321925854e-05, "loss": 0.0083, "step": 20130 }, { "grad_norm": 0.15954242646694183, "learning_rate": 9.212572640257003e-05, "loss": 0.011, "step": 20140 }, { "grad_norm": 0.10361927002668381, "learning_rate": 9.211590398643579e-05, "loss": 0.0142, "step": 20150 }, { "grad_norm": 0.13641977310180664, "learning_rate": 9.210607597216149e-05, "loss": 0.0152, "step": 20160 }, { "grad_norm": 0.11063767969608307, "learning_rate": 9.209624236105345e-05, "loss": 0.0109, "step": 20170 }, { "grad_norm": 0.13560616970062256, "learning_rate": 9.208640315441881e-05, "loss": 0.0164, "step": 20180 }, { "grad_norm": 0.12187725305557251, "learning_rate": 9.207655835356541e-05, "loss": 0.0111, "step": 20190 }, { "grad_norm": 0.07640862464904785, "learning_rate": 9.206670795980182e-05, "loss": 0.0158, "step": 20200 }, { "grad_norm": 0.13282889127731323, "learning_rate": 9.205685197443741e-05, "loss": 0.012, "step": 20210 }, { "grad_norm": 0.1566106528043747, "learning_rate": 9.204699039878222e-05, "loss": 0.0126, "step": 20220 }, { "grad_norm": 0.18180985748767853, "learning_rate": 9.203712323414712e-05, "loss": 0.0104, "step": 20230 }, { "grad_norm": 0.11752326041460037, "learning_rate": 9.202725048184365e-05, "loss": 0.012, "step": 20240 }, { "grad_norm": 0.10441605001688004, "learning_rate": 9.20173721431841e-05, "loss": 0.0078, "step": 20250 }, { "grad_norm": 0.09413743764162064, "learning_rate": 9.200748821948155e-05, "loss": 0.011, "step": 20260 }, { "grad_norm": 0.11337558180093765, "learning_rate": 9.199759871204978e-05, "loss": 0.0115, "step": 20270 }, { "grad_norm": 0.09642261266708374, "learning_rate": 9.198770362220332e-05, "loss": 0.0116, "step": 20280 }, { "grad_norm": 0.07472699135541916, "learning_rate": 9.197780295125749e-05, "loss": 0.0089, "step": 20290 }, { "grad_norm": 0.08487249165773392, "learning_rate": 9.196789670052824e-05, "loss": 0.0123, "step": 20300 }, { "grad_norm": 0.09297779202461243, "learning_rate": 9.195798487133238e-05, "loss": 0.0104, "step": 20310 }, { "grad_norm": 0.12910683453083038, "learning_rate": 9.194806746498741e-05, "loss": 0.0101, "step": 20320 }, { "grad_norm": 0.10694025456905365, "learning_rate": 9.193814448281155e-05, "loss": 0.0104, "step": 20330 }, { "grad_norm": 0.1789100021123886, "learning_rate": 9.192821592612381e-05, "loss": 0.0204, "step": 20340 }, { "grad_norm": 0.14869044721126556, "learning_rate": 9.191828179624389e-05, "loss": 0.011, "step": 20350 }, { "grad_norm": 0.10555194318294525, "learning_rate": 9.190834209449229e-05, "loss": 0.0128, "step": 20360 }, { "grad_norm": 0.09882283210754395, "learning_rate": 9.189839682219019e-05, "loss": 0.0097, "step": 20370 }, { "grad_norm": 0.0993318259716034, "learning_rate": 9.188844598065953e-05, "loss": 0.0092, "step": 20380 }, { "grad_norm": 0.07357437163591385, "learning_rate": 9.187848957122303e-05, "loss": 0.0125, "step": 20390 }, { "grad_norm": 0.13355495035648346, "learning_rate": 9.18685275952041e-05, "loss": 0.0093, "step": 20400 }, { "grad_norm": 0.0979224368929863, "learning_rate": 9.18585600539269e-05, "loss": 0.0123, "step": 20410 }, { "grad_norm": 0.10763214528560638, "learning_rate": 9.184858694871636e-05, "loss": 0.0091, "step": 20420 }, { "grad_norm": 0.07357458770275116, "learning_rate": 9.183860828089811e-05, "loss": 0.011, "step": 20430 }, { "grad_norm": 0.1308152675628662, "learning_rate": 9.182862405179855e-05, "loss": 0.013, "step": 20440 }, { "grad_norm": 0.1747599095106125, "learning_rate": 9.181863426274479e-05, "loss": 0.0099, "step": 20450 }, { "grad_norm": 0.07062019407749176, "learning_rate": 9.180863891506471e-05, "loss": 0.0114, "step": 20460 }, { "grad_norm": 0.12001199275255203, "learning_rate": 9.179863801008692e-05, "loss": 0.0122, "step": 20470 }, { "grad_norm": 0.14059437811374664, "learning_rate": 9.178863154914073e-05, "loss": 0.0104, "step": 20480 }, { "grad_norm": 0.11939772963523865, "learning_rate": 9.177861953355624e-05, "loss": 0.0101, "step": 20490 }, { "grad_norm": 0.09582175314426422, "learning_rate": 9.176860196466428e-05, "loss": 0.0103, "step": 20500 }, { "grad_norm": 0.0802951231598854, "learning_rate": 9.175857884379639e-05, "loss": 0.0138, "step": 20510 }, { "grad_norm": 0.15403540432453156, "learning_rate": 9.174855017228489e-05, "loss": 0.0103, "step": 20520 }, { "grad_norm": 0.09810323268175125, "learning_rate": 9.173851595146278e-05, "loss": 0.0103, "step": 20530 }, { "grad_norm": 0.0973416119813919, "learning_rate": 9.172847618266386e-05, "loss": 0.0139, "step": 20540 }, { "grad_norm": 0.08770640194416046, "learning_rate": 9.171843086722263e-05, "loss": 0.0112, "step": 20550 }, { "grad_norm": 0.10356995463371277, "learning_rate": 9.170838000647432e-05, "loss": 0.0129, "step": 20560 }, { "grad_norm": 0.10630559176206589, "learning_rate": 9.169832360175493e-05, "loss": 0.0101, "step": 20570 }, { "grad_norm": 0.11867409944534302, "learning_rate": 9.168826165440118e-05, "loss": 0.0139, "step": 20580 }, { "grad_norm": 0.1312924176454544, "learning_rate": 9.167819416575052e-05, "loss": 0.0109, "step": 20590 }, { "grad_norm": 0.12310940772294998, "learning_rate": 9.166812113714115e-05, "loss": 0.0156, "step": 20600 }, { "grad_norm": 0.11944089829921722, "learning_rate": 9.1658042569912e-05, "loss": 0.0134, "step": 20610 }, { "grad_norm": 0.09317050874233246, "learning_rate": 9.16479584654027e-05, "loss": 0.01, "step": 20620 }, { "grad_norm": 0.07600709795951843, "learning_rate": 9.16378688249537e-05, "loss": 0.0088, "step": 20630 }, { "grad_norm": 0.09259427338838577, "learning_rate": 9.162777364990614e-05, "loss": 0.0095, "step": 20640 }, { "grad_norm": 0.14417342841625214, "learning_rate": 9.161767294160186e-05, "loss": 0.0113, "step": 20650 }, { "grad_norm": 0.11829544603824615, "learning_rate": 9.160756670138346e-05, "loss": 0.0155, "step": 20660 }, { "grad_norm": 0.10580919682979584, "learning_rate": 9.159745493059433e-05, "loss": 0.0114, "step": 20670 }, { "grad_norm": 0.16195964813232422, "learning_rate": 9.158733763057851e-05, "loss": 0.0102, "step": 20680 }, { "grad_norm": 0.11233118176460266, "learning_rate": 9.157721480268084e-05, "loss": 0.0112, "step": 20690 }, { "grad_norm": 0.09982585906982422, "learning_rate": 9.156708644824685e-05, "loss": 0.0117, "step": 20700 }, { "grad_norm": 0.09963732957839966, "learning_rate": 9.155695256862285e-05, "loss": 0.0092, "step": 20710 }, { "grad_norm": 0.09373459964990616, "learning_rate": 9.154681316515582e-05, "loss": 0.0111, "step": 20720 }, { "grad_norm": 0.11228100210428238, "learning_rate": 9.153666823919354e-05, "loss": 0.0113, "step": 20730 }, { "grad_norm": 0.07940027117729187, "learning_rate": 9.152651779208448e-05, "loss": 0.0094, "step": 20740 }, { "grad_norm": 0.10071222484111786, "learning_rate": 9.151636182517786e-05, "loss": 0.0085, "step": 20750 }, { "grad_norm": 0.14590097963809967, "learning_rate": 9.150620033982367e-05, "loss": 0.0103, "step": 20760 }, { "grad_norm": 0.0865062028169632, "learning_rate": 9.149603333737256e-05, "loss": 0.0094, "step": 20770 }, { "grad_norm": 0.10751762986183167, "learning_rate": 9.148586081917593e-05, "loss": 0.0142, "step": 20780 }, { "grad_norm": 0.11425992846488953, "learning_rate": 9.1475682786586e-05, "loss": 0.0125, "step": 20790 }, { "grad_norm": 0.13743534684181213, "learning_rate": 9.14654992409556e-05, "loss": 0.0133, "step": 20800 }, { "grad_norm": 0.12147554010152817, "learning_rate": 9.145531018363838e-05, "loss": 0.0115, "step": 20810 }, { "grad_norm": 0.07265900075435638, "learning_rate": 9.144511561598867e-05, "loss": 0.0117, "step": 20820 }, { "grad_norm": 0.11115970462560654, "learning_rate": 9.143491553936157e-05, "loss": 0.0105, "step": 20830 }, { "grad_norm": 0.13386961817741394, "learning_rate": 9.142470995511291e-05, "loss": 0.0111, "step": 20840 }, { "grad_norm": 0.08921946585178375, "learning_rate": 9.141449886459921e-05, "loss": 0.0102, "step": 20850 }, { "grad_norm": 0.09781576693058014, "learning_rate": 9.140428226917777e-05, "loss": 0.0078, "step": 20860 }, { "grad_norm": 0.08435523509979248, "learning_rate": 9.13940601702066e-05, "loss": 0.009, "step": 20870 }, { "grad_norm": 0.0817621499300003, "learning_rate": 9.138383256904443e-05, "loss": 0.0084, "step": 20880 }, { "grad_norm": 0.11432819813489914, "learning_rate": 9.137359946705075e-05, "loss": 0.01, "step": 20890 }, { "grad_norm": 0.09424123913049698, "learning_rate": 9.136336086558576e-05, "loss": 0.0118, "step": 20900 }, { "grad_norm": 0.11722097545862198, "learning_rate": 9.13531167660104e-05, "loss": 0.0141, "step": 20910 }, { "grad_norm": 0.21093763411045074, "learning_rate": 9.134286716968634e-05, "loss": 0.0149, "step": 20920 }, { "grad_norm": 0.09773897379636765, "learning_rate": 9.133261207797599e-05, "loss": 0.0131, "step": 20930 }, { "grad_norm": 0.07772034406661987, "learning_rate": 9.132235149224246e-05, "loss": 0.0124, "step": 20940 }, { "grad_norm": 0.08316221088171005, "learning_rate": 9.131208541384963e-05, "loss": 0.0114, "step": 20950 }, { "grad_norm": 0.08865959942340851, "learning_rate": 9.130181384416205e-05, "loss": 0.0118, "step": 20960 }, { "grad_norm": 0.07977885007858276, "learning_rate": 9.12915367845451e-05, "loss": 0.0124, "step": 20970 }, { "grad_norm": 0.12625621259212494, "learning_rate": 9.128125423636477e-05, "loss": 0.0109, "step": 20980 }, { "grad_norm": 0.09251454472541809, "learning_rate": 9.12709662009879e-05, "loss": 0.0145, "step": 20990 }, { "grad_norm": 0.11403211951255798, "learning_rate": 9.126067267978196e-05, "loss": 0.0164, "step": 21000 }, { "grad_norm": 0.13026775419712067, "learning_rate": 9.125037367411518e-05, "loss": 0.0107, "step": 21010 }, { "grad_norm": 0.10881204903125763, "learning_rate": 9.124006918535654e-05, "loss": 0.0096, "step": 21020 }, { "grad_norm": 0.10105312615633011, "learning_rate": 9.122975921487576e-05, "loss": 0.0116, "step": 21030 }, { "grad_norm": 0.10334097594022751, "learning_rate": 9.121944376404321e-05, "loss": 0.0089, "step": 21040 }, { "grad_norm": 0.12557066977024078, "learning_rate": 9.120912283423009e-05, "loss": 0.0123, "step": 21050 }, { "grad_norm": 0.11633086949586868, "learning_rate": 9.119879642680827e-05, "loss": 0.0135, "step": 21060 }, { "grad_norm": 0.11624511331319809, "learning_rate": 9.118846454315033e-05, "loss": 0.0094, "step": 21070 }, { "grad_norm": 0.13664278388023376, "learning_rate": 9.117812718462966e-05, "loss": 0.0152, "step": 21080 }, { "grad_norm": 0.08363683521747589, "learning_rate": 9.116778435262028e-05, "loss": 0.0102, "step": 21090 }, { "grad_norm": 0.1327618509531021, "learning_rate": 9.1157436048497e-05, "loss": 0.0124, "step": 21100 }, { "grad_norm": 0.1385578215122223, "learning_rate": 9.114708227363535e-05, "loss": 0.0111, "step": 21110 }, { "grad_norm": 0.10130726546049118, "learning_rate": 9.113672302941155e-05, "loss": 0.0118, "step": 21120 }, { "grad_norm": 0.06760429590940475, "learning_rate": 9.11263583172026e-05, "loss": 0.011, "step": 21130 }, { "grad_norm": 0.14909814298152924, "learning_rate": 9.111598813838618e-05, "loss": 0.0118, "step": 21140 }, { "grad_norm": 0.10134337097406387, "learning_rate": 9.110561249434073e-05, "loss": 0.0112, "step": 21150 }, { "grad_norm": 0.12219025939702988, "learning_rate": 9.10952313864454e-05, "loss": 0.0092, "step": 21160 }, { "grad_norm": 0.1008937656879425, "learning_rate": 9.108484481608006e-05, "loss": 0.0118, "step": 21170 }, { "grad_norm": 0.08895538747310638, "learning_rate": 9.107445278462535e-05, "loss": 0.0099, "step": 21180 }, { "grad_norm": 0.07827528566122055, "learning_rate": 9.106405529346255e-05, "loss": 0.0087, "step": 21190 }, { "grad_norm": 0.12161416560411453, "learning_rate": 9.105365234397379e-05, "loss": 0.0114, "step": 21200 }, { "grad_norm": 0.10803432017564774, "learning_rate": 9.104324393754177e-05, "loss": 0.0096, "step": 21210 }, { "grad_norm": 0.09776046872138977, "learning_rate": 9.103283007555006e-05, "loss": 0.0106, "step": 21220 }, { "grad_norm": 0.10965567827224731, "learning_rate": 9.102241075938286e-05, "loss": 0.0129, "step": 21230 }, { "grad_norm": 0.09289386868476868, "learning_rate": 9.101198599042516e-05, "loss": 0.0117, "step": 21240 }, { "grad_norm": 0.07044804096221924, "learning_rate": 9.10015557700626e-05, "loss": 0.0112, "step": 21250 }, { "grad_norm": 0.13602864742279053, "learning_rate": 9.099112009968162e-05, "loss": 0.0118, "step": 21260 }, { "grad_norm": 0.13829582929611206, "learning_rate": 9.098067898066935e-05, "loss": 0.0145, "step": 21270 }, { "grad_norm": 0.1213698461651802, "learning_rate": 9.097023241441365e-05, "loss": 0.0131, "step": 21280 }, { "grad_norm": 0.12360496073961258, "learning_rate": 9.095978040230309e-05, "loss": 0.0099, "step": 21290 }, { "grad_norm": 0.13624690473079681, "learning_rate": 9.094932294572696e-05, "loss": 0.0139, "step": 21300 }, { "grad_norm": 0.11658009886741638, "learning_rate": 9.093886004607532e-05, "loss": 0.0082, "step": 21310 }, { "grad_norm": 0.17908507585525513, "learning_rate": 9.092839170473891e-05, "loss": 0.0091, "step": 21320 }, { "grad_norm": 0.11550583690404892, "learning_rate": 9.091791792310921e-05, "loss": 0.011, "step": 21330 }, { "grad_norm": 0.1428518444299698, "learning_rate": 9.09074387025784e-05, "loss": 0.0113, "step": 21340 }, { "grad_norm": 0.07171012461185455, "learning_rate": 9.089695404453942e-05, "loss": 0.0102, "step": 21350 }, { "grad_norm": 0.11274971812963486, "learning_rate": 9.088646395038592e-05, "loss": 0.0164, "step": 21360 }, { "grad_norm": 0.07281015068292618, "learning_rate": 9.087596842151224e-05, "loss": 0.0126, "step": 21370 }, { "grad_norm": 0.10224150866270065, "learning_rate": 9.086546745931348e-05, "loss": 0.0117, "step": 21380 }, { "grad_norm": 0.12634338438510895, "learning_rate": 9.085496106518546e-05, "loss": 0.009, "step": 21390 }, { "grad_norm": 0.09960873425006866, "learning_rate": 9.08444492405247e-05, "loss": 0.0126, "step": 21400 }, { "grad_norm": 0.1541280299425125, "learning_rate": 9.083393198672849e-05, "loss": 0.0114, "step": 21410 }, { "grad_norm": 0.1494412124156952, "learning_rate": 9.082340930519478e-05, "loss": 0.0129, "step": 21420 }, { "grad_norm": 0.10890863090753555, "learning_rate": 9.081288119732225e-05, "loss": 0.0119, "step": 21430 }, { "grad_norm": 0.15097567439079285, "learning_rate": 9.080234766451034e-05, "loss": 0.0123, "step": 21440 }, { "grad_norm": 0.14680518209934235, "learning_rate": 9.07918087081592e-05, "loss": 0.0132, "step": 21450 }, { "grad_norm": 0.1321668028831482, "learning_rate": 9.078126432966967e-05, "loss": 0.0107, "step": 21460 }, { "grad_norm": 0.11833278834819794, "learning_rate": 9.077071453044334e-05, "loss": 0.0137, "step": 21470 }, { "grad_norm": 0.09900262951850891, "learning_rate": 9.076015931188253e-05, "loss": 0.0098, "step": 21480 }, { "grad_norm": 0.07802830636501312, "learning_rate": 9.074959867539025e-05, "loss": 0.0101, "step": 21490 }, { "grad_norm": 0.07674061506986618, "learning_rate": 9.073903262237023e-05, "loss": 0.0116, "step": 21500 }, { "grad_norm": 0.07820919901132584, "learning_rate": 9.072846115422695e-05, "loss": 0.0113, "step": 21510 }, { "grad_norm": 0.08969347178936005, "learning_rate": 9.071788427236561e-05, "loss": 0.0134, "step": 21520 }, { "grad_norm": 0.11498096585273743, "learning_rate": 9.070730197819206e-05, "loss": 0.0099, "step": 21530 }, { "grad_norm": 0.06407320499420166, "learning_rate": 9.069671427311299e-05, "loss": 0.0113, "step": 21540 }, { "grad_norm": 0.06978126615285873, "learning_rate": 9.068612115853567e-05, "loss": 0.0078, "step": 21550 }, { "grad_norm": 0.09507925063371658, "learning_rate": 9.067552263586823e-05, "loss": 0.0084, "step": 21560 }, { "grad_norm": 0.05507725849747658, "learning_rate": 9.066491870651939e-05, "loss": 0.0098, "step": 21570 }, { "grad_norm": 0.07067324221134186, "learning_rate": 9.065430937189868e-05, "loss": 0.0118, "step": 21580 }, { "grad_norm": 0.15163448452949524, "learning_rate": 9.064369463341632e-05, "loss": 0.0144, "step": 21590 }, { "grad_norm": 0.17338384687900543, "learning_rate": 9.063307449248322e-05, "loss": 0.0112, "step": 21600 }, { "grad_norm": 0.11972885578870773, "learning_rate": 9.062244895051106e-05, "loss": 0.0132, "step": 21610 }, { "grad_norm": 0.14022143185138702, "learning_rate": 9.061181800891221e-05, "loss": 0.0129, "step": 21620 }, { "grad_norm": 0.13932812213897705, "learning_rate": 9.060118166909972e-05, "loss": 0.0121, "step": 21630 }, { "grad_norm": 0.12472618371248245, "learning_rate": 9.059053993248744e-05, "loss": 0.0107, "step": 21640 }, { "grad_norm": 0.05947092920541763, "learning_rate": 9.057989280048986e-05, "loss": 0.0102, "step": 21650 }, { "grad_norm": 0.11508332937955856, "learning_rate": 9.056924027452228e-05, "loss": 0.0093, "step": 21660 }, { "grad_norm": 0.12143776565790176, "learning_rate": 9.055858235600056e-05, "loss": 0.0115, "step": 21670 }, { "grad_norm": 0.19565078616142273, "learning_rate": 9.054791904634146e-05, "loss": 0.0111, "step": 21680 }, { "grad_norm": 0.09494640678167343, "learning_rate": 9.053725034696234e-05, "loss": 0.0127, "step": 21690 }, { "grad_norm": 0.12730297446250916, "learning_rate": 9.05265762592813e-05, "loss": 0.012, "step": 21700 }, { "grad_norm": 0.14349564909934998, "learning_rate": 9.051589678471717e-05, "loss": 0.0107, "step": 21710 }, { "grad_norm": 0.08069951087236404, "learning_rate": 9.05052119246895e-05, "loss": 0.0126, "step": 21720 }, { "grad_norm": 0.1013464406132698, "learning_rate": 9.049452168061852e-05, "loss": 0.012, "step": 21730 }, { "grad_norm": 0.07604783773422241, "learning_rate": 9.048382605392522e-05, "loss": 0.0095, "step": 21740 }, { "grad_norm": 0.07511118054389954, "learning_rate": 9.04731250460313e-05, "loss": 0.0107, "step": 21750 }, { "grad_norm": 0.17211170494556427, "learning_rate": 9.046241865835914e-05, "loss": 0.0251, "step": 21760 }, { "grad_norm": 0.1572858989238739, "learning_rate": 9.045170689233187e-05, "loss": 0.0165, "step": 21770 }, { "grad_norm": 0.1557115614414215, "learning_rate": 9.04409897493733e-05, "loss": 0.0139, "step": 21780 }, { "grad_norm": 0.09365618228912354, "learning_rate": 9.043026723090799e-05, "loss": 0.0145, "step": 21790 }, { "grad_norm": 0.10410155355930328, "learning_rate": 9.04195393383612e-05, "loss": 0.0124, "step": 21800 }, { "grad_norm": 0.1855163723230362, "learning_rate": 9.040880607315892e-05, "loss": 0.0151, "step": 21810 }, { "grad_norm": 0.12620146572589874, "learning_rate": 9.039806743672783e-05, "loss": 0.0102, "step": 21820 }, { "grad_norm": 0.12174353748559952, "learning_rate": 9.038732343049532e-05, "loss": 0.0121, "step": 21830 }, { "grad_norm": 0.1300855576992035, "learning_rate": 9.037657405588952e-05, "loss": 0.0107, "step": 21840 }, { "grad_norm": 0.1479620486497879, "learning_rate": 9.036581931433927e-05, "loss": 0.0134, "step": 21850 }, { "grad_norm": 0.13257049024105072, "learning_rate": 9.035505920727409e-05, "loss": 0.01, "step": 21860 }, { "grad_norm": 0.14193138480186462, "learning_rate": 9.034429373612425e-05, "loss": 0.0119, "step": 21870 }, { "grad_norm": 0.09820772707462311, "learning_rate": 9.033352290232073e-05, "loss": 0.0101, "step": 21880 }, { "grad_norm": 0.15104760229587555, "learning_rate": 9.03227467072952e-05, "loss": 0.0104, "step": 21890 }, { "grad_norm": 0.08531121164560318, "learning_rate": 9.031196515248006e-05, "loss": 0.0087, "step": 21900 }, { "grad_norm": 0.07359869033098221, "learning_rate": 9.030117823930843e-05, "loss": 0.0121, "step": 21910 }, { "grad_norm": 0.17590533196926117, "learning_rate": 9.029038596921411e-05, "loss": 0.0087, "step": 21920 }, { "grad_norm": 0.1151110827922821, "learning_rate": 9.027958834363165e-05, "loss": 0.0152, "step": 21930 }, { "grad_norm": 0.13461089134216309, "learning_rate": 9.026878536399627e-05, "loss": 0.0161, "step": 21940 }, { "grad_norm": 0.08288256078958511, "learning_rate": 9.025797703174396e-05, "loss": 0.012, "step": 21950 }, { "grad_norm": 0.10386107116937637, "learning_rate": 9.024716334831136e-05, "loss": 0.01, "step": 21960 }, { "grad_norm": 0.09088345617055893, "learning_rate": 9.023634431513587e-05, "loss": 0.0081, "step": 21970 }, { "grad_norm": 0.1230614185333252, "learning_rate": 9.022551993365556e-05, "loss": 0.0182, "step": 21980 }, { "grad_norm": 0.14010651409626007, "learning_rate": 9.021469020530925e-05, "loss": 0.011, "step": 21990 }, { "grad_norm": 0.10082114487886429, "learning_rate": 9.020385513153644e-05, "loss": 0.0103, "step": 22000 }, { "grad_norm": 0.110671266913414, "learning_rate": 9.019301471377735e-05, "loss": 0.0125, "step": 22010 }, { "grad_norm": 0.09175477176904678, "learning_rate": 9.018216895347291e-05, "loss": 0.0112, "step": 22020 }, { "grad_norm": 0.10288412123918533, "learning_rate": 9.017131785206478e-05, "loss": 0.0113, "step": 22030 }, { "grad_norm": 0.09406669437885284, "learning_rate": 9.016046141099531e-05, "loss": 0.0094, "step": 22040 }, { "grad_norm": 0.08220430463552475, "learning_rate": 9.014959963170755e-05, "loss": 0.0117, "step": 22050 }, { "grad_norm": 0.09907730668783188, "learning_rate": 9.013873251564526e-05, "loss": 0.0086, "step": 22060 }, { "grad_norm": 0.07756613194942474, "learning_rate": 9.012786006425296e-05, "loss": 0.0084, "step": 22070 }, { "grad_norm": 0.0999765396118164, "learning_rate": 9.011698227897581e-05, "loss": 0.011, "step": 22080 }, { "grad_norm": 0.16106478869915009, "learning_rate": 9.010609916125971e-05, "loss": 0.0113, "step": 22090 }, { "grad_norm": 0.13476914167404175, "learning_rate": 9.009521071255129e-05, "loss": 0.0114, "step": 22100 }, { "grad_norm": 0.12337588518857956, "learning_rate": 9.008431693429786e-05, "loss": 0.0138, "step": 22110 }, { "grad_norm": 0.10827170312404633, "learning_rate": 9.007341782794743e-05, "loss": 0.0112, "step": 22120 }, { "grad_norm": 0.10053785145282745, "learning_rate": 9.006251339494874e-05, "loss": 0.0107, "step": 22130 }, { "grad_norm": 0.08972079306840897, "learning_rate": 9.005160363675123e-05, "loss": 0.0098, "step": 22140 }, { "grad_norm": 0.0804406926035881, "learning_rate": 9.004068855480505e-05, "loss": 0.0111, "step": 22150 }, { "grad_norm": 0.09791327267885208, "learning_rate": 9.002976815056108e-05, "loss": 0.0107, "step": 22160 }, { "grad_norm": 0.0704628974199295, "learning_rate": 9.001884242547084e-05, "loss": 0.0085, "step": 22170 }, { "grad_norm": 0.10238973796367645, "learning_rate": 9.000791138098665e-05, "loss": 0.0103, "step": 22180 }, { "grad_norm": 0.09145395457744598, "learning_rate": 8.999697501856146e-05, "loss": 0.0111, "step": 22190 }, { "grad_norm": 0.06359481811523438, "learning_rate": 8.998603333964897e-05, "loss": 0.0116, "step": 22200 }, { "grad_norm": 0.1114407554268837, "learning_rate": 8.997508634570354e-05, "loss": 0.0088, "step": 22210 }, { "grad_norm": 0.10529591143131256, "learning_rate": 8.996413403818032e-05, "loss": 0.0125, "step": 22220 }, { "grad_norm": 0.18952426314353943, "learning_rate": 8.995317641853507e-05, "loss": 0.0112, "step": 22230 }, { "grad_norm": 0.07702072709798813, "learning_rate": 8.994221348822432e-05, "loss": 0.0118, "step": 22240 }, { "grad_norm": 0.10981113463640213, "learning_rate": 8.993124524870529e-05, "loss": 0.0117, "step": 22250 }, { "grad_norm": 0.07089941948652267, "learning_rate": 8.992027170143589e-05, "loss": 0.0129, "step": 22260 }, { "grad_norm": 0.09396149963140488, "learning_rate": 8.990929284787477e-05, "loss": 0.01, "step": 22270 }, { "grad_norm": 0.10423041135072708, "learning_rate": 8.989830868948126e-05, "loss": 0.0104, "step": 22280 }, { "grad_norm": 0.08885090798139572, "learning_rate": 8.988731922771538e-05, "loss": 0.0092, "step": 22290 }, { "grad_norm": 0.12385033816099167, "learning_rate": 8.987632446403789e-05, "loss": 0.0085, "step": 22300 }, { "grad_norm": 0.10928826034069061, "learning_rate": 8.986532439991022e-05, "loss": 0.0086, "step": 22310 }, { "grad_norm": 0.10967499017715454, "learning_rate": 8.985431903679454e-05, "loss": 0.0131, "step": 22320 }, { "grad_norm": 0.12346624583005905, "learning_rate": 8.984330837615371e-05, "loss": 0.0131, "step": 22330 }, { "grad_norm": 0.13317568600177765, "learning_rate": 8.983229241945129e-05, "loss": 0.0112, "step": 22340 }, { "grad_norm": 0.09886227548122406, "learning_rate": 8.982127116815152e-05, "loss": 0.0114, "step": 22350 }, { "grad_norm": 0.12358283251523972, "learning_rate": 8.98102446237194e-05, "loss": 0.0098, "step": 22360 }, { "grad_norm": 0.08788347989320755, "learning_rate": 8.97992127876206e-05, "loss": 0.0115, "step": 22370 }, { "grad_norm": 0.07595673948526382, "learning_rate": 8.97881756613215e-05, "loss": 0.0119, "step": 22380 }, { "grad_norm": 0.12659220397472382, "learning_rate": 8.977713324628913e-05, "loss": 0.0102, "step": 22390 }, { "grad_norm": 0.1090291365981102, "learning_rate": 8.976608554399135e-05, "loss": 0.0126, "step": 22400 }, { "grad_norm": 0.10794559866189957, "learning_rate": 8.975503255589659e-05, "loss": 0.0147, "step": 22410 }, { "grad_norm": 0.0859600082039833, "learning_rate": 8.974397428347406e-05, "loss": 0.0106, "step": 22420 }, { "grad_norm": 0.08143297582864761, "learning_rate": 8.973291072819363e-05, "loss": 0.0112, "step": 22430 }, { "grad_norm": 0.06105221435427666, "learning_rate": 8.972184189152593e-05, "loss": 0.0103, "step": 22440 }, { "grad_norm": 0.15926991403102875, "learning_rate": 8.971076777494221e-05, "loss": 0.0166, "step": 22450 }, { "grad_norm": 0.12411271780729294, "learning_rate": 8.96996883799145e-05, "loss": 0.0141, "step": 22460 }, { "grad_norm": 0.10376838594675064, "learning_rate": 8.968860370791549e-05, "loss": 0.0093, "step": 22470 }, { "grad_norm": 0.05445469170808792, "learning_rate": 8.967751376041857e-05, "loss": 0.01, "step": 22480 }, { "grad_norm": 0.1741127371788025, "learning_rate": 8.966641853889784e-05, "loss": 0.0128, "step": 22490 }, { "grad_norm": 0.115618035197258, "learning_rate": 8.965531804482811e-05, "loss": 0.0105, "step": 22500 }, { "grad_norm": 0.10263980180025101, "learning_rate": 8.964421227968489e-05, "loss": 0.0101, "step": 22510 }, { "grad_norm": 0.1495552510023117, "learning_rate": 8.963310124494437e-05, "loss": 0.0112, "step": 22520 }, { "grad_norm": 0.1238865777850151, "learning_rate": 8.962198494208345e-05, "loss": 0.0132, "step": 22530 }, { "grad_norm": 0.08893744647502899, "learning_rate": 8.961086337257973e-05, "loss": 0.0091, "step": 22540 }, { "grad_norm": 0.10554929077625275, "learning_rate": 8.959973653791155e-05, "loss": 0.0092, "step": 22550 }, { "grad_norm": 0.0725860446691513, "learning_rate": 8.958860443955788e-05, "loss": 0.0088, "step": 22560 }, { "grad_norm": 0.09280454367399216, "learning_rate": 8.957746707899843e-05, "loss": 0.0121, "step": 22570 }, { "grad_norm": 0.08789325505495071, "learning_rate": 8.95663244577136e-05, "loss": 0.0086, "step": 22580 }, { "grad_norm": 0.09038699418306351, "learning_rate": 8.955517657718451e-05, "loss": 0.0089, "step": 22590 }, { "grad_norm": 0.06707315146923065, "learning_rate": 8.954402343889293e-05, "loss": 0.0092, "step": 22600 }, { "grad_norm": 0.10234690457582474, "learning_rate": 8.95328650443214e-05, "loss": 0.0106, "step": 22610 }, { "grad_norm": 0.08972364664077759, "learning_rate": 8.952170139495306e-05, "loss": 0.0108, "step": 22620 }, { "grad_norm": 0.10576923191547394, "learning_rate": 8.951053249227188e-05, "loss": 0.0079, "step": 22630 }, { "grad_norm": 0.15152809023857117, "learning_rate": 8.94993583377624e-05, "loss": 0.011, "step": 22640 }, { "grad_norm": 0.1236908957362175, "learning_rate": 8.948817893290993e-05, "loss": 0.0109, "step": 22650 }, { "grad_norm": 0.09727641195058823, "learning_rate": 8.947699427920047e-05, "loss": 0.0106, "step": 22660 }, { "grad_norm": 0.12777815759181976, "learning_rate": 8.946580437812071e-05, "loss": 0.0098, "step": 22670 }, { "grad_norm": 0.07846240699291229, "learning_rate": 8.945460923115802e-05, "loss": 0.0115, "step": 22680 }, { "grad_norm": 0.08302772790193558, "learning_rate": 8.944340883980052e-05, "loss": 0.0124, "step": 22690 }, { "grad_norm": 0.1351221203804016, "learning_rate": 8.943220320553693e-05, "loss": 0.0148, "step": 22700 }, { "grad_norm": 0.152975395321846, "learning_rate": 8.942099232985679e-05, "loss": 0.0114, "step": 22710 }, { "grad_norm": 0.1420615166425705, "learning_rate": 8.940977621425023e-05, "loss": 0.0115, "step": 22720 }, { "grad_norm": 0.093377023935318, "learning_rate": 8.939855486020814e-05, "loss": 0.0138, "step": 22730 }, { "grad_norm": 0.07609062641859055, "learning_rate": 8.93873282692221e-05, "loss": 0.0122, "step": 22740 }, { "grad_norm": 0.10598601400852203, "learning_rate": 8.937609644278435e-05, "loss": 0.0119, "step": 22750 }, { "grad_norm": 0.15294526517391205, "learning_rate": 8.936485938238786e-05, "loss": 0.013, "step": 22760 }, { "grad_norm": 0.09868106245994568, "learning_rate": 8.93536170895263e-05, "loss": 0.0107, "step": 22770 }, { "grad_norm": 0.12846045196056366, "learning_rate": 8.934236956569398e-05, "loss": 0.0117, "step": 22780 }, { "grad_norm": 0.07149338722229004, "learning_rate": 8.933111681238599e-05, "loss": 0.0138, "step": 22790 }, { "grad_norm": 0.10248308628797531, "learning_rate": 8.931985883109804e-05, "loss": 0.0122, "step": 22800 }, { "grad_norm": 0.1320371925830841, "learning_rate": 8.930859562332659e-05, "loss": 0.0121, "step": 22810 }, { "grad_norm": 0.10274186730384827, "learning_rate": 8.929732719056876e-05, "loss": 0.0119, "step": 22820 }, { "grad_norm": 0.11194203048944473, "learning_rate": 8.928605353432236e-05, "loss": 0.0095, "step": 22830 }, { "grad_norm": 0.09722136706113815, "learning_rate": 8.927477465608594e-05, "loss": 0.0113, "step": 22840 }, { "grad_norm": 0.10100774466991425, "learning_rate": 8.926349055735868e-05, "loss": 0.0084, "step": 22850 }, { "grad_norm": 0.11428359895944595, "learning_rate": 8.925220123964052e-05, "loss": 0.012, "step": 22860 }, { "grad_norm": 0.09243544191122055, "learning_rate": 8.924090670443204e-05, "loss": 0.0093, "step": 22870 }, { "grad_norm": 0.1797240972518921, "learning_rate": 8.922960695323455e-05, "loss": 0.0112, "step": 22880 }, { "grad_norm": 0.07638410478830338, "learning_rate": 8.921830198755002e-05, "loss": 0.013, "step": 22890 }, { "grad_norm": 0.09502945840358734, "learning_rate": 8.920699180888114e-05, "loss": 0.0119, "step": 22900 }, { "grad_norm": 0.11613523215055466, "learning_rate": 8.91956764187313e-05, "loss": 0.0126, "step": 22910 }, { "grad_norm": 0.1743987500667572, "learning_rate": 8.918435581860454e-05, "loss": 0.0124, "step": 22920 }, { "grad_norm": 0.17144343256950378, "learning_rate": 8.917303001000564e-05, "loss": 0.0111, "step": 22930 }, { "grad_norm": 0.11436669528484344, "learning_rate": 8.916169899444003e-05, "loss": 0.0088, "step": 22940 }, { "grad_norm": 0.081133633852005, "learning_rate": 8.915036277341387e-05, "loss": 0.0111, "step": 22950 }, { "grad_norm": 0.10948224365711212, "learning_rate": 8.913902134843401e-05, "loss": 0.0103, "step": 22960 }, { "grad_norm": 0.08277591317892075, "learning_rate": 8.912767472100795e-05, "loss": 0.0068, "step": 22970 }, { "grad_norm": 0.1522381752729416, "learning_rate": 8.911632289264392e-05, "loss": 0.0115, "step": 22980 }, { "grad_norm": 0.09619861096143723, "learning_rate": 8.910496586485082e-05, "loss": 0.0092, "step": 22990 }, { "grad_norm": 0.14540301263332367, "learning_rate": 8.909360363913828e-05, "loss": 0.0105, "step": 23000 }, { "grad_norm": 0.13895590603351593, "learning_rate": 8.908223621701657e-05, "loss": 0.013, "step": 23010 }, { "grad_norm": 0.2056083083152771, "learning_rate": 8.907086359999666e-05, "loss": 0.0124, "step": 23020 }, { "grad_norm": 0.09710328280925751, "learning_rate": 8.905948578959027e-05, "loss": 0.0115, "step": 23030 }, { "grad_norm": 0.09693629294633865, "learning_rate": 8.904810278730972e-05, "loss": 0.0119, "step": 23040 }, { "grad_norm": 0.1663317084312439, "learning_rate": 8.903671459466807e-05, "loss": 0.0171, "step": 23050 }, { "grad_norm": 0.11433617770671844, "learning_rate": 8.902532121317908e-05, "loss": 0.0096, "step": 23060 }, { "grad_norm": 0.11635725945234299, "learning_rate": 8.901392264435718e-05, "loss": 0.0126, "step": 23070 }, { "grad_norm": 0.12371137738227844, "learning_rate": 8.900251888971749e-05, "loss": 0.013, "step": 23080 }, { "grad_norm": 0.09102948009967804, "learning_rate": 8.899110995077582e-05, "loss": 0.0114, "step": 23090 }, { "grad_norm": 0.0842125415802002, "learning_rate": 8.897969582904868e-05, "loss": 0.0139, "step": 23100 }, { "grad_norm": 0.07174074649810791, "learning_rate": 8.896827652605327e-05, "loss": 0.0108, "step": 23110 }, { "grad_norm": 0.0976305902004242, "learning_rate": 8.895685204330743e-05, "loss": 0.0131, "step": 23120 }, { "grad_norm": 0.16564711928367615, "learning_rate": 8.894542238232975e-05, "loss": 0.0105, "step": 23130 }, { "grad_norm": 0.10879655182361603, "learning_rate": 8.89339875446395e-05, "loss": 0.0121, "step": 23140 }, { "grad_norm": 0.12416006624698639, "learning_rate": 8.892254753175661e-05, "loss": 0.0145, "step": 23150 }, { "grad_norm": 0.1371523141860962, "learning_rate": 8.89111023452017e-05, "loss": 0.0142, "step": 23160 }, { "grad_norm": 0.09437865018844604, "learning_rate": 8.889965198649613e-05, "loss": 0.0098, "step": 23170 }, { "grad_norm": 0.06037476658821106, "learning_rate": 8.888819645716187e-05, "loss": 0.0124, "step": 23180 }, { "grad_norm": 0.10191698372364044, "learning_rate": 8.887673575872162e-05, "loss": 0.0087, "step": 23190 }, { "grad_norm": 0.11917097866535187, "learning_rate": 8.886526989269876e-05, "loss": 0.0137, "step": 23200 }, { "grad_norm": 0.10806215554475784, "learning_rate": 8.885379886061738e-05, "loss": 0.0149, "step": 23210 }, { "grad_norm": 0.11225096136331558, "learning_rate": 8.884232266400223e-05, "loss": 0.0106, "step": 23220 }, { "grad_norm": 0.14985723793506622, "learning_rate": 8.883084130437873e-05, "loss": 0.0181, "step": 23230 }, { "grad_norm": 0.16914305090904236, "learning_rate": 8.881935478327304e-05, "loss": 0.012, "step": 23240 }, { "grad_norm": 0.07884226739406586, "learning_rate": 8.880786310221193e-05, "loss": 0.0141, "step": 23250 }, { "grad_norm": 0.1307639628648758, "learning_rate": 8.879636626272294e-05, "loss": 0.0157, "step": 23260 }, { "grad_norm": 0.09666730463504791, "learning_rate": 8.878486426633424e-05, "loss": 0.0101, "step": 23270 }, { "grad_norm": 0.07681035995483398, "learning_rate": 8.877335711457469e-05, "loss": 0.0118, "step": 23280 }, { "grad_norm": 0.10233893990516663, "learning_rate": 8.876184480897386e-05, "loss": 0.0139, "step": 23290 }, { "grad_norm": 0.11018527299165726, "learning_rate": 8.875032735106199e-05, "loss": 0.0108, "step": 23300 }, { "grad_norm": 0.07923837006092072, "learning_rate": 8.873880474237002e-05, "loss": 0.0098, "step": 23310 }, { "grad_norm": 0.14094042778015137, "learning_rate": 8.872727698442953e-05, "loss": 0.0105, "step": 23320 }, { "grad_norm": 0.07354374974966049, "learning_rate": 8.871574407877283e-05, "loss": 0.0118, "step": 23330 }, { "grad_norm": 0.13545644283294678, "learning_rate": 8.870420602693292e-05, "loss": 0.0102, "step": 23340 }, { "grad_norm": 0.12367410212755203, "learning_rate": 8.869266283044344e-05, "loss": 0.0117, "step": 23350 }, { "grad_norm": 0.10015957802534103, "learning_rate": 8.868111449083874e-05, "loss": 0.0075, "step": 23360 }, { "grad_norm": 0.11862911283969879, "learning_rate": 8.866956100965384e-05, "loss": 0.0106, "step": 23370 }, { "grad_norm": 0.09573939442634583, "learning_rate": 8.865800238842449e-05, "loss": 0.0104, "step": 23380 }, { "grad_norm": 0.10088092088699341, "learning_rate": 8.864643862868705e-05, "loss": 0.011, "step": 23390 }, { "grad_norm": 0.09613319486379623, "learning_rate": 8.863486973197864e-05, "loss": 0.0116, "step": 23400 }, { "grad_norm": 0.07186497747898102, "learning_rate": 8.862329569983697e-05, "loss": 0.0095, "step": 23410 }, { "grad_norm": 0.12298198789358139, "learning_rate": 8.861171653380054e-05, "loss": 0.0108, "step": 23420 }, { "grad_norm": 0.22984595596790314, "learning_rate": 8.860013223540844e-05, "loss": 0.0107, "step": 23430 }, { "grad_norm": 0.11776621639728546, "learning_rate": 8.85885428062005e-05, "loss": 0.0105, "step": 23440 }, { "grad_norm": 0.09379152208566666, "learning_rate": 8.85769482477172e-05, "loss": 0.0128, "step": 23450 }, { "grad_norm": 0.11851824074983597, "learning_rate": 8.856534856149974e-05, "loss": 0.0147, "step": 23460 }, { "grad_norm": 0.08274345099925995, "learning_rate": 8.855374374908996e-05, "loss": 0.0108, "step": 23470 }, { "grad_norm": 0.12926875054836273, "learning_rate": 8.854213381203038e-05, "loss": 0.0129, "step": 23480 }, { "grad_norm": 0.09968124330043793, "learning_rate": 8.853051875186425e-05, "loss": 0.0124, "step": 23490 }, { "grad_norm": 0.12401869148015976, "learning_rate": 8.851889857013545e-05, "loss": 0.0128, "step": 23500 }, { "grad_norm": 0.08840759843587875, "learning_rate": 8.850727326838857e-05, "loss": 0.0094, "step": 23510 }, { "grad_norm": 0.14376798272132874, "learning_rate": 8.849564284816887e-05, "loss": 0.0121, "step": 23520 }, { "grad_norm": 0.12947382032871246, "learning_rate": 8.848400731102229e-05, "loss": 0.0105, "step": 23530 }, { "grad_norm": 0.15246392786502838, "learning_rate": 8.847236665849545e-05, "loss": 0.0126, "step": 23540 }, { "grad_norm": 0.062317654490470886, "learning_rate": 8.846072089213564e-05, "loss": 0.0079, "step": 23550 }, { "grad_norm": 0.05955887213349342, "learning_rate": 8.844907001349087e-05, "loss": 0.0099, "step": 23560 }, { "grad_norm": 0.07608837634325027, "learning_rate": 8.843741402410978e-05, "loss": 0.01, "step": 23570 }, { "grad_norm": 0.08407198637723923, "learning_rate": 8.84257529255417e-05, "loss": 0.0114, "step": 23580 }, { "grad_norm": 0.12135849893093109, "learning_rate": 8.841408671933669e-05, "loss": 0.0117, "step": 23590 }, { "grad_norm": 0.10731419175863266, "learning_rate": 8.84024154070454e-05, "loss": 0.0118, "step": 23600 }, { "grad_norm": 0.1209663450717926, "learning_rate": 8.839073899021924e-05, "loss": 0.0122, "step": 23610 }, { "grad_norm": 0.11071094870567322, "learning_rate": 8.837905747041024e-05, "loss": 0.0121, "step": 23620 }, { "grad_norm": 0.07962099462747574, "learning_rate": 8.836737084917117e-05, "loss": 0.009, "step": 23630 }, { "grad_norm": 0.12302429974079132, "learning_rate": 8.83556791280554e-05, "loss": 0.0126, "step": 23640 }, { "grad_norm": 0.11995186656713486, "learning_rate": 8.834398230861705e-05, "loss": 0.011, "step": 23650 }, { "grad_norm": 0.0863175168633461, "learning_rate": 8.833228039241088e-05, "loss": 0.0115, "step": 23660 }, { "grad_norm": 0.12000519037246704, "learning_rate": 8.832057338099234e-05, "loss": 0.0143, "step": 23670 }, { "grad_norm": 0.1237211525440216, "learning_rate": 8.830886127591751e-05, "loss": 0.0112, "step": 23680 }, { "grad_norm": 0.12489096075296402, "learning_rate": 8.829714407874325e-05, "loss": 0.0122, "step": 23690 }, { "grad_norm": 0.11568303406238556, "learning_rate": 8.8285421791027e-05, "loss": 0.0121, "step": 23700 }, { "grad_norm": 0.13666532933712006, "learning_rate": 8.827369441432693e-05, "loss": 0.0095, "step": 23710 }, { "grad_norm": 0.11518479883670807, "learning_rate": 8.826196195020184e-05, "loss": 0.0122, "step": 23720 }, { "grad_norm": 0.07991962879896164, "learning_rate": 8.825022440021128e-05, "loss": 0.016, "step": 23730 }, { "grad_norm": 0.12223455309867859, "learning_rate": 8.82384817659154e-05, "loss": 0.0107, "step": 23740 }, { "grad_norm": 0.08893639594316483, "learning_rate": 8.822673404887507e-05, "loss": 0.0131, "step": 23750 }, { "grad_norm": 0.12703180313110352, "learning_rate": 8.82149812506518e-05, "loss": 0.0121, "step": 23760 }, { "grad_norm": 0.15197204053401947, "learning_rate": 8.820322337280784e-05, "loss": 0.0115, "step": 23770 }, { "grad_norm": 0.1340334564447403, "learning_rate": 8.819146041690603e-05, "loss": 0.0113, "step": 23780 }, { "grad_norm": 0.10530664771795273, "learning_rate": 8.817969238450997e-05, "loss": 0.012, "step": 23790 }, { "grad_norm": 0.04850873351097107, "learning_rate": 8.816791927718385e-05, "loss": 0.0104, "step": 23800 }, { "grad_norm": 0.1209673061966896, "learning_rate": 8.815614109649259e-05, "loss": 0.0118, "step": 23810 }, { "grad_norm": 0.08306638151407242, "learning_rate": 8.814435784400179e-05, "loss": 0.0105, "step": 23820 }, { "grad_norm": 0.1293523758649826, "learning_rate": 8.81325695212777e-05, "loss": 0.0122, "step": 23830 }, { "grad_norm": 0.1246030330657959, "learning_rate": 8.812077612988725e-05, "loss": 0.0106, "step": 23840 }, { "grad_norm": 0.1159704327583313, "learning_rate": 8.810897767139804e-05, "loss": 0.017, "step": 23850 }, { "grad_norm": 0.13439856469631195, "learning_rate": 8.809717414737834e-05, "loss": 0.0135, "step": 23860 }, { "grad_norm": 0.11574216187000275, "learning_rate": 8.808536555939713e-05, "loss": 0.0123, "step": 23870 }, { "grad_norm": 0.1523267775774002, "learning_rate": 8.807355190902398e-05, "loss": 0.0158, "step": 23880 }, { "grad_norm": 0.08053167909383774, "learning_rate": 8.806173319782926e-05, "loss": 0.0121, "step": 23890 }, { "grad_norm": 0.08163592964410782, "learning_rate": 8.804990942738387e-05, "loss": 0.0082, "step": 23900 }, { "grad_norm": 0.1091628223657608, "learning_rate": 8.803808059925948e-05, "loss": 0.0126, "step": 23910 }, { "grad_norm": 0.0936041921377182, "learning_rate": 8.802624671502843e-05, "loss": 0.013, "step": 23920 }, { "grad_norm": 0.0946982130408287, "learning_rate": 8.801440777626366e-05, "loss": 0.011, "step": 23930 }, { "grad_norm": 0.0780741423368454, "learning_rate": 8.800256378453886e-05, "loss": 0.0096, "step": 23940 }, { "grad_norm": 0.09547632932662964, "learning_rate": 8.799071474142834e-05, "loss": 0.0116, "step": 23950 }, { "grad_norm": 0.0899035707116127, "learning_rate": 8.797886064850714e-05, "loss": 0.0101, "step": 23960 }, { "grad_norm": 0.10507351905107498, "learning_rate": 8.796700150735088e-05, "loss": 0.0075, "step": 23970 }, { "grad_norm": 0.07775106281042099, "learning_rate": 8.795513731953593e-05, "loss": 0.0096, "step": 23980 }, { "grad_norm": 0.05671700835227966, "learning_rate": 8.794326808663931e-05, "loss": 0.0128, "step": 23990 }, { "grad_norm": 0.12822620570659637, "learning_rate": 8.793139381023869e-05, "loss": 0.0127, "step": 24000 }, { "grad_norm": 0.1066797524690628, "learning_rate": 8.791951449191242e-05, "loss": 0.0129, "step": 24010 }, { "grad_norm": 0.12741078436374664, "learning_rate": 8.790763013323954e-05, "loss": 0.0121, "step": 24020 }, { "grad_norm": 0.11296366900205612, "learning_rate": 8.789574073579976e-05, "loss": 0.0107, "step": 24030 }, { "grad_norm": 0.06544226408004761, "learning_rate": 8.788384630117344e-05, "loss": 0.0111, "step": 24040 }, { "grad_norm": 0.13794340193271637, "learning_rate": 8.787194683094157e-05, "loss": 0.0102, "step": 24050 }, { "grad_norm": 0.10451137274503708, "learning_rate": 8.786004232668591e-05, "loss": 0.0113, "step": 24060 }, { "grad_norm": 0.0967903733253479, "learning_rate": 8.78481327899888e-05, "loss": 0.0104, "step": 24070 }, { "grad_norm": 0.056636519730091095, "learning_rate": 8.78362182224333e-05, "loss": 0.0102, "step": 24080 }, { "grad_norm": 0.5837519764900208, "learning_rate": 8.78242986256031e-05, "loss": 0.0169, "step": 24090 }, { "grad_norm": 0.25368615984916687, "learning_rate": 8.78123740010826e-05, "loss": 0.0177, "step": 24100 }, { "grad_norm": 0.24732884764671326, "learning_rate": 8.780044435045685e-05, "loss": 0.0145, "step": 24110 }, { "grad_norm": 0.0959518700838089, "learning_rate": 8.778850967531157e-05, "loss": 0.0139, "step": 24120 }, { "grad_norm": 0.10572148859500885, "learning_rate": 8.77765699772331e-05, "loss": 0.0109, "step": 24130 }, { "grad_norm": 0.13172991573810577, "learning_rate": 8.776462525780854e-05, "loss": 0.0098, "step": 24140 }, { "grad_norm": 0.1196398064494133, "learning_rate": 8.775267551862558e-05, "loss": 0.0108, "step": 24150 }, { "grad_norm": 0.11646309494972229, "learning_rate": 8.774072076127262e-05, "loss": 0.011, "step": 24160 }, { "grad_norm": 0.09965363144874573, "learning_rate": 8.772876098733872e-05, "loss": 0.0099, "step": 24170 }, { "grad_norm": 0.09386014938354492, "learning_rate": 8.771679619841358e-05, "loss": 0.0101, "step": 24180 }, { "grad_norm": 0.16596029698848724, "learning_rate": 8.77048263960876e-05, "loss": 0.0133, "step": 24190 }, { "grad_norm": 0.10121237486600876, "learning_rate": 8.769285158195185e-05, "loss": 0.0097, "step": 24200 }, { "grad_norm": 0.12330924719572067, "learning_rate": 8.768087175759798e-05, "loss": 0.0101, "step": 24210 }, { "grad_norm": 0.09743556380271912, "learning_rate": 8.766888692461845e-05, "loss": 0.0113, "step": 24220 }, { "grad_norm": 0.1337355375289917, "learning_rate": 8.76568970846063e-05, "loss": 0.0103, "step": 24230 }, { "grad_norm": 0.17990122735500336, "learning_rate": 8.76449022391552e-05, "loss": 0.0118, "step": 24240 }, { "grad_norm": 0.08727477490901947, "learning_rate": 8.763290238985959e-05, "loss": 0.0099, "step": 24250 }, { "grad_norm": 0.08549439907073975, "learning_rate": 8.762089753831447e-05, "loss": 0.0123, "step": 24260 }, { "grad_norm": 0.09699919819831848, "learning_rate": 8.760888768611556e-05, "loss": 0.0091, "step": 24270 }, { "grad_norm": 0.10837988555431366, "learning_rate": 8.759687283485927e-05, "loss": 0.0095, "step": 24280 }, { "grad_norm": 0.05708635225892067, "learning_rate": 8.758485298614262e-05, "loss": 0.0081, "step": 24290 }, { "grad_norm": 0.10883548855781555, "learning_rate": 8.75728281415633e-05, "loss": 0.0103, "step": 24300 }, { "grad_norm": 0.1329871565103531, "learning_rate": 8.75607983027197e-05, "loss": 0.0095, "step": 24310 }, { "grad_norm": 0.09953723847866058, "learning_rate": 8.754876347121084e-05, "loss": 0.0136, "step": 24320 }, { "grad_norm": 0.11779941618442535, "learning_rate": 8.753672364863643e-05, "loss": 0.0095, "step": 24330 }, { "grad_norm": 0.1169932410120964, "learning_rate": 8.752467883659681e-05, "loss": 0.0098, "step": 24340 }, { "grad_norm": 0.09299564361572266, "learning_rate": 8.751262903669301e-05, "loss": 0.0117, "step": 24350 }, { "grad_norm": 0.1430898904800415, "learning_rate": 8.750057425052674e-05, "loss": 0.0113, "step": 24360 }, { "grad_norm": 0.15431582927703857, "learning_rate": 8.748851447970032e-05, "loss": 0.0087, "step": 24370 }, { "grad_norm": 0.1049719750881195, "learning_rate": 8.747644972581678e-05, "loss": 0.0112, "step": 24380 }, { "grad_norm": 0.08457124978303909, "learning_rate": 8.746437999047978e-05, "loss": 0.014, "step": 24390 }, { "grad_norm": 0.12174201756715775, "learning_rate": 8.745230527529367e-05, "loss": 0.0102, "step": 24400 }, { "grad_norm": 0.06886475533246994, "learning_rate": 8.744022558186344e-05, "loss": 0.0108, "step": 24410 }, { "grad_norm": 0.14387284219264984, "learning_rate": 8.742814091179474e-05, "loss": 0.0149, "step": 24420 }, { "grad_norm": 0.08064499497413635, "learning_rate": 8.741605126669391e-05, "loss": 0.0094, "step": 24430 }, { "grad_norm": 0.12565936148166656, "learning_rate": 8.740395664816791e-05, "loss": 0.0098, "step": 24440 }, { "grad_norm": 0.09809320420026779, "learning_rate": 8.739185705782443e-05, "loss": 0.013, "step": 24450 }, { "grad_norm": 0.12409012019634247, "learning_rate": 8.737975249727173e-05, "loss": 0.0094, "step": 24460 }, { "grad_norm": 0.13215366005897522, "learning_rate": 8.736764296811878e-05, "loss": 0.0143, "step": 24470 }, { "grad_norm": 0.06843545287847519, "learning_rate": 8.735552847197521e-05, "loss": 0.0099, "step": 24480 }, { "grad_norm": 0.07617081701755524, "learning_rate": 8.734340901045133e-05, "loss": 0.0091, "step": 24490 }, { "grad_norm": 0.09136761724948883, "learning_rate": 8.733128458515805e-05, "loss": 0.0087, "step": 24500 }, { "grad_norm": 0.08689837157726288, "learning_rate": 8.731915519770699e-05, "loss": 0.0089, "step": 24510 }, { "grad_norm": 0.09124656021595001, "learning_rate": 8.730702084971044e-05, "loss": 0.0115, "step": 24520 }, { "grad_norm": 0.17168238759040833, "learning_rate": 8.729488154278127e-05, "loss": 0.0135, "step": 24530 }, { "grad_norm": 0.13806352019309998, "learning_rate": 8.728273727853312e-05, "loss": 0.012, "step": 24540 }, { "grad_norm": 0.09293775260448456, "learning_rate": 8.72705880585802e-05, "loss": 0.0104, "step": 24550 }, { "grad_norm": 0.10847349464893341, "learning_rate": 8.72584338845374e-05, "loss": 0.0086, "step": 24560 }, { "grad_norm": 0.06442643702030182, "learning_rate": 8.724627475802032e-05, "loss": 0.0076, "step": 24570 }, { "grad_norm": 0.0683404728770256, "learning_rate": 8.723411068064515e-05, "loss": 0.0082, "step": 24580 }, { "grad_norm": 0.08922379463911057, "learning_rate": 8.722194165402878e-05, "loss": 0.0143, "step": 24590 }, { "grad_norm": 0.11558442562818527, "learning_rate": 8.720976767978874e-05, "loss": 0.0134, "step": 24600 }, { "grad_norm": 0.10853038728237152, "learning_rate": 8.719758875954322e-05, "loss": 0.0118, "step": 24610 }, { "grad_norm": 0.162513867020607, "learning_rate": 8.718540489491106e-05, "loss": 0.0113, "step": 24620 }, { "grad_norm": 0.09276753664016724, "learning_rate": 8.717321608751179e-05, "loss": 0.0106, "step": 24630 }, { "grad_norm": 0.11648862063884735, "learning_rate": 8.716102233896554e-05, "loss": 0.0088, "step": 24640 }, { "grad_norm": 0.1422768235206604, "learning_rate": 8.714882365089316e-05, "loss": 0.0103, "step": 24650 }, { "grad_norm": 0.09156211465597153, "learning_rate": 8.713662002491612e-05, "loss": 0.0092, "step": 24660 }, { "grad_norm": 0.09065581858158112, "learning_rate": 8.712441146265655e-05, "loss": 0.0108, "step": 24670 }, { "grad_norm": 0.13076044619083405, "learning_rate": 8.711219796573724e-05, "loss": 0.011, "step": 24680 }, { "grad_norm": 0.08531752973794937, "learning_rate": 8.709997953578164e-05, "loss": 0.0125, "step": 24690 }, { "grad_norm": 0.0893181562423706, "learning_rate": 8.708775617441383e-05, "loss": 0.0098, "step": 24700 }, { "grad_norm": 0.09232218563556671, "learning_rate": 8.707552788325859e-05, "loss": 0.0079, "step": 24710 }, { "grad_norm": 0.07353078573942184, "learning_rate": 8.706329466394132e-05, "loss": 0.0105, "step": 24720 }, { "grad_norm": 0.09146689623594284, "learning_rate": 8.70510565180881e-05, "loss": 0.0102, "step": 24730 }, { "grad_norm": 0.1085626482963562, "learning_rate": 8.703881344732563e-05, "loss": 0.0092, "step": 24740 }, { "grad_norm": 0.06946602463722229, "learning_rate": 8.702656545328132e-05, "loss": 0.0102, "step": 24750 }, { "grad_norm": 0.06085527688264847, "learning_rate": 8.701431253758317e-05, "loss": 0.0067, "step": 24760 }, { "grad_norm": 0.07852979004383087, "learning_rate": 8.700205470185985e-05, "loss": 0.0089, "step": 24770 }, { "grad_norm": 0.11647950857877731, "learning_rate": 8.698979194774075e-05, "loss": 0.0102, "step": 24780 }, { "grad_norm": 0.11358276009559631, "learning_rate": 8.697752427685583e-05, "loss": 0.0109, "step": 24790 }, { "grad_norm": 0.14420297741889954, "learning_rate": 8.696525169083574e-05, "loss": 0.0106, "step": 24800 }, { "grad_norm": 0.09046170860528946, "learning_rate": 8.695297419131178e-05, "loss": 0.0084, "step": 24810 }, { "grad_norm": 0.08284541964530945, "learning_rate": 8.694069177991592e-05, "loss": 0.0067, "step": 24820 }, { "grad_norm": 0.06834322959184647, "learning_rate": 8.692840445828074e-05, "loss": 0.0102, "step": 24830 }, { "grad_norm": 0.07266576588153839, "learning_rate": 8.69161122280395e-05, "loss": 0.0106, "step": 24840 }, { "grad_norm": 0.09450621157884598, "learning_rate": 8.690381509082612e-05, "loss": 0.0128, "step": 24850 }, { "grad_norm": 0.16413205862045288, "learning_rate": 8.689151304827519e-05, "loss": 0.013, "step": 24860 }, { "grad_norm": 0.10838603228330612, "learning_rate": 8.687920610202188e-05, "loss": 0.0064, "step": 24870 }, { "grad_norm": 0.08405206352472305, "learning_rate": 8.686689425370204e-05, "loss": 0.0098, "step": 24880 }, { "grad_norm": 0.09323127567768097, "learning_rate": 8.685457750495227e-05, "loss": 0.0101, "step": 24890 }, { "grad_norm": 0.09796541184186935, "learning_rate": 8.684225585740966e-05, "loss": 0.0089, "step": 24900 }, { "grad_norm": 0.12461675703525543, "learning_rate": 8.682992931271208e-05, "loss": 0.0091, "step": 24910 }, { "grad_norm": 0.07493144273757935, "learning_rate": 8.681759787249799e-05, "loss": 0.008, "step": 24920 }, { "grad_norm": 0.11143644154071808, "learning_rate": 8.680526153840648e-05, "loss": 0.0096, "step": 24930 }, { "grad_norm": 0.115418441593647, "learning_rate": 8.679292031207739e-05, "loss": 0.0133, "step": 24940 }, { "grad_norm": 0.09890905022621155, "learning_rate": 8.678057419515108e-05, "loss": 0.0107, "step": 24950 }, { "grad_norm": 0.09695038199424744, "learning_rate": 8.676822318926865e-05, "loss": 0.0078, "step": 24960 }, { "grad_norm": 0.07741279155015945, "learning_rate": 8.675586729607183e-05, "loss": 0.0173, "step": 24970 }, { "grad_norm": 0.09648280590772629, "learning_rate": 8.674350651720297e-05, "loss": 0.0098, "step": 24980 }, { "grad_norm": 0.080413818359375, "learning_rate": 8.673114085430511e-05, "loss": 0.0098, "step": 24990 }, { "grad_norm": 0.11454790830612183, "learning_rate": 8.671877030902191e-05, "loss": 0.0082, "step": 25000 }, { "grad_norm": 0.13375581800937653, "learning_rate": 8.670639488299771e-05, "loss": 0.0137, "step": 25010 }, { "grad_norm": 0.0860946774482727, "learning_rate": 8.669401457787747e-05, "loss": 0.0108, "step": 25020 }, { "grad_norm": 0.06750927120447159, "learning_rate": 8.668162939530679e-05, "loss": 0.0083, "step": 25030 }, { "grad_norm": 0.08212205767631531, "learning_rate": 8.666923933693197e-05, "loss": 0.0083, "step": 25040 }, { "grad_norm": 0.08601599186658859, "learning_rate": 8.66568444043999e-05, "loss": 0.0101, "step": 25050 }, { "grad_norm": 0.10748908668756485, "learning_rate": 8.664444459935814e-05, "loss": 0.0119, "step": 25060 }, { "grad_norm": 0.062315575778484344, "learning_rate": 8.66320399234549e-05, "loss": 0.0093, "step": 25070 }, { "grad_norm": 0.07116454839706421, "learning_rate": 8.661963037833907e-05, "loss": 0.0088, "step": 25080 }, { "grad_norm": 0.08061981201171875, "learning_rate": 8.66072159656601e-05, "loss": 0.0104, "step": 25090 }, { "grad_norm": 0.1201389729976654, "learning_rate": 8.65947966870682e-05, "loss": 0.0097, "step": 25100 }, { "grad_norm": 0.10008510947227478, "learning_rate": 8.658237254421411e-05, "loss": 0.0135, "step": 25110 }, { "grad_norm": 0.1017477959394455, "learning_rate": 8.656994353874932e-05, "loss": 0.0123, "step": 25120 }, { "grad_norm": 0.14361146092414856, "learning_rate": 8.655750967232591e-05, "loss": 0.0089, "step": 25130 }, { "grad_norm": 0.09088841825723648, "learning_rate": 8.654507094659661e-05, "loss": 0.0104, "step": 25140 }, { "grad_norm": 0.16565260291099548, "learning_rate": 8.65326273632148e-05, "loss": 0.0138, "step": 25150 }, { "grad_norm": 0.09017443656921387, "learning_rate": 8.652017892383451e-05, "loss": 0.0111, "step": 25160 }, { "grad_norm": 0.10036864131689072, "learning_rate": 8.650772563011043e-05, "loss": 0.0114, "step": 25170 }, { "grad_norm": 0.07557643204927444, "learning_rate": 8.649526748369785e-05, "loss": 0.0094, "step": 25180 }, { "grad_norm": 0.09109235554933548, "learning_rate": 8.648280448625276e-05, "loss": 0.0085, "step": 25190 }, { "grad_norm": 0.0916498452425003, "learning_rate": 8.647033663943177e-05, "loss": 0.0118, "step": 25200 }, { "grad_norm": 0.09836559742689133, "learning_rate": 8.645786394489214e-05, "loss": 0.0066, "step": 25210 }, { "grad_norm": 0.07543464750051498, "learning_rate": 8.644538640429175e-05, "loss": 0.0096, "step": 25220 }, { "grad_norm": 0.07502483576536179, "learning_rate": 8.643290401928912e-05, "loss": 0.0096, "step": 25230 }, { "grad_norm": 0.07927656173706055, "learning_rate": 8.642041679154349e-05, "loss": 0.0074, "step": 25240 }, { "grad_norm": 0.06741873174905777, "learning_rate": 8.640792472271465e-05, "loss": 0.013, "step": 25250 }, { "grad_norm": 0.09801926463842392, "learning_rate": 8.63954278144631e-05, "loss": 0.0098, "step": 25260 }, { "grad_norm": 0.1116248220205307, "learning_rate": 8.638292606844993e-05, "loss": 0.0138, "step": 25270 }, { "grad_norm": 0.10642655193805695, "learning_rate": 8.637041948633693e-05, "loss": 0.0108, "step": 25280 }, { "grad_norm": 0.10473860800266266, "learning_rate": 8.635790806978648e-05, "loss": 0.0091, "step": 25290 }, { "grad_norm": 0.09529545158147812, "learning_rate": 8.634539182046164e-05, "loss": 0.0087, "step": 25300 }, { "grad_norm": 0.10923240333795547, "learning_rate": 8.633287074002609e-05, "loss": 0.01, "step": 25310 }, { "grad_norm": 0.09828155487775803, "learning_rate": 8.632034483014415e-05, "loss": 0.011, "step": 25320 }, { "grad_norm": 0.08437125384807587, "learning_rate": 8.63078140924808e-05, "loss": 0.0104, "step": 25330 }, { "grad_norm": 0.18529540300369263, "learning_rate": 8.629527852870169e-05, "loss": 0.0107, "step": 25340 }, { "grad_norm": 0.15361006557941437, "learning_rate": 8.6282738140473e-05, "loss": 0.0111, "step": 25350 }, { "grad_norm": 0.13401687145233154, "learning_rate": 8.62701929294617e-05, "loss": 0.0088, "step": 25360 }, { "grad_norm": 0.0860699787735939, "learning_rate": 8.625764289733529e-05, "loss": 0.0145, "step": 25370 }, { "grad_norm": 0.1107417568564415, "learning_rate": 8.624508804576195e-05, "loss": 0.0133, "step": 25380 }, { "grad_norm": 0.07889178395271301, "learning_rate": 8.623252837641049e-05, "loss": 0.0099, "step": 25390 }, { "grad_norm": 0.10313977301120758, "learning_rate": 8.62199638909504e-05, "loss": 0.008, "step": 25400 }, { "grad_norm": 0.14636069536209106, "learning_rate": 8.620739459105176e-05, "loss": 0.0096, "step": 25410 }, { "grad_norm": 0.12221911549568176, "learning_rate": 8.61948204783853e-05, "loss": 0.0115, "step": 25420 }, { "grad_norm": 0.1792653203010559, "learning_rate": 8.618224155462242e-05, "loss": 0.0125, "step": 25430 }, { "grad_norm": 0.112305648624897, "learning_rate": 8.616965782143512e-05, "loss": 0.0149, "step": 25440 }, { "grad_norm": 0.12047109752893448, "learning_rate": 8.615706928049608e-05, "loss": 0.0121, "step": 25450 }, { "grad_norm": 0.17464911937713623, "learning_rate": 8.614447593347859e-05, "loss": 0.014, "step": 25460 }, { "grad_norm": 0.08481896668672562, "learning_rate": 8.613187778205656e-05, "loss": 0.0092, "step": 25470 }, { "grad_norm": 0.0684802234172821, "learning_rate": 8.611927482790458e-05, "loss": 0.0091, "step": 25480 }, { "grad_norm": 0.07370562106370926, "learning_rate": 8.61066670726979e-05, "loss": 0.0087, "step": 25490 }, { "grad_norm": 0.09286471456289291, "learning_rate": 8.60940545181123e-05, "loss": 0.0122, "step": 25500 }, { "grad_norm": 0.13465437293052673, "learning_rate": 8.608143716582432e-05, "loss": 0.0145, "step": 25510 }, { "grad_norm": 0.10008290410041809, "learning_rate": 8.60688150175111e-05, "loss": 0.0116, "step": 25520 }, { "grad_norm": 0.1615535467863083, "learning_rate": 8.605618807485033e-05, "loss": 0.0137, "step": 25530 }, { "grad_norm": 0.10197732597589493, "learning_rate": 8.60435563395205e-05, "loss": 0.011, "step": 25540 }, { "grad_norm": 0.1316554993391037, "learning_rate": 8.60309198132006e-05, "loss": 0.0098, "step": 25550 }, { "grad_norm": 0.15295995771884918, "learning_rate": 8.60182784975703e-05, "loss": 0.0137, "step": 25560 }, { "grad_norm": 0.10631439834833145, "learning_rate": 8.600563239430993e-05, "loss": 0.012, "step": 25570 }, { "grad_norm": 0.10083634406328201, "learning_rate": 8.599298150510045e-05, "loss": 0.0103, "step": 25580 }, { "grad_norm": 0.09943493455648422, "learning_rate": 8.598032583162341e-05, "loss": 0.0112, "step": 25590 }, { "grad_norm": 0.11130569130182266, "learning_rate": 8.596766537556104e-05, "loss": 0.0108, "step": 25600 }, { "grad_norm": 0.1090841293334961, "learning_rate": 8.595500013859622e-05, "loss": 0.0093, "step": 25610 }, { "grad_norm": 0.14861838519573212, "learning_rate": 8.594233012241242e-05, "loss": 0.0095, "step": 25620 }, { "grad_norm": 0.1183653399348259, "learning_rate": 8.592965532869377e-05, "loss": 0.0127, "step": 25630 }, { "grad_norm": 0.09721916913986206, "learning_rate": 8.591697575912505e-05, "loss": 0.0084, "step": 25640 }, { "grad_norm": 0.10794182121753693, "learning_rate": 8.590429141539162e-05, "loss": 0.0095, "step": 25650 }, { "grad_norm": 0.07065355777740479, "learning_rate": 8.589160229917955e-05, "loss": 0.0089, "step": 25660 }, { "grad_norm": 0.07061689347028732, "learning_rate": 8.58789084121755e-05, "loss": 0.0092, "step": 25670 }, { "grad_norm": 0.11978023499250412, "learning_rate": 8.586620975606673e-05, "loss": 0.0099, "step": 25680 }, { "grad_norm": 0.08802060782909393, "learning_rate": 8.585350633254121e-05, "loss": 0.0104, "step": 25690 }, { "grad_norm": 0.11614319682121277, "learning_rate": 8.584079814328751e-05, "loss": 0.0123, "step": 25700 }, { "grad_norm": 0.09954864531755447, "learning_rate": 8.582808518999481e-05, "loss": 0.0098, "step": 25710 }, { "grad_norm": 0.0987982228398323, "learning_rate": 8.581536747435295e-05, "loss": 0.0108, "step": 25720 }, { "grad_norm": 0.13625265657901764, "learning_rate": 8.580264499805242e-05, "loss": 0.012, "step": 25730 }, { "grad_norm": 0.11908283084630966, "learning_rate": 8.578991776278428e-05, "loss": 0.0105, "step": 25740 }, { "grad_norm": 0.12033718079328537, "learning_rate": 8.577718577024031e-05, "loss": 0.0072, "step": 25750 }, { "grad_norm": 0.05955667793750763, "learning_rate": 8.576444902211283e-05, "loss": 0.0109, "step": 25760 }, { "grad_norm": 0.1040055975317955, "learning_rate": 8.575170752009487e-05, "loss": 0.0083, "step": 25770 }, { "grad_norm": 0.099680595099926, "learning_rate": 8.573896126588003e-05, "loss": 0.0125, "step": 25780 }, { "grad_norm": 0.0885453075170517, "learning_rate": 8.572621026116259e-05, "loss": 0.009, "step": 25790 }, { "grad_norm": 0.12128982692956924, "learning_rate": 8.571345450763744e-05, "loss": 0.01, "step": 25800 }, { "grad_norm": 0.14099213480949402, "learning_rate": 8.57006940070001e-05, "loss": 0.0082, "step": 25810 }, { "grad_norm": 0.09606894850730896, "learning_rate": 8.568792876094674e-05, "loss": 0.0082, "step": 25820 }, { "grad_norm": 0.12853844463825226, "learning_rate": 8.567515877117411e-05, "loss": 0.01, "step": 25830 }, { "grad_norm": 0.1064995601773262, "learning_rate": 8.566238403937965e-05, "loss": 0.0114, "step": 25840 }, { "grad_norm": 0.08979878574609756, "learning_rate": 8.564960456726141e-05, "loss": 0.0128, "step": 25850 }, { "grad_norm": 0.09960247576236725, "learning_rate": 8.563682035651806e-05, "loss": 0.0132, "step": 25860 }, { "grad_norm": 0.11689089983701706, "learning_rate": 8.562403140884889e-05, "loss": 0.0098, "step": 25870 }, { "grad_norm": 0.10289923846721649, "learning_rate": 8.561123772595385e-05, "loss": 0.0097, "step": 25880 }, { "grad_norm": 0.11811418831348419, "learning_rate": 8.559843930953352e-05, "loss": 0.0115, "step": 25890 }, { "grad_norm": 0.10668386518955231, "learning_rate": 8.558563616128907e-05, "loss": 0.0114, "step": 25900 }, { "grad_norm": 0.11034148931503296, "learning_rate": 8.557282828292232e-05, "loss": 0.0119, "step": 25910 }, { "grad_norm": 0.1002427339553833, "learning_rate": 8.556001567613575e-05, "loss": 0.0095, "step": 25920 }, { "grad_norm": 0.07700560241937637, "learning_rate": 8.554719834263241e-05, "loss": 0.0079, "step": 25930 }, { "grad_norm": 0.11170227825641632, "learning_rate": 8.553437628411605e-05, "loss": 0.0088, "step": 25940 }, { "grad_norm": 0.11555604636669159, "learning_rate": 8.552154950229094e-05, "loss": 0.0101, "step": 25950 }, { "grad_norm": 0.09398743510246277, "learning_rate": 8.550871799886212e-05, "loss": 0.0083, "step": 25960 }, { "grad_norm": 0.1594805270433426, "learning_rate": 8.549588177553511e-05, "loss": 0.013, "step": 25970 }, { "grad_norm": 0.14963370561599731, "learning_rate": 8.548304083401619e-05, "loss": 0.0087, "step": 25980 }, { "grad_norm": 0.06497850269079208, "learning_rate": 8.547019517601217e-05, "loss": 0.014, "step": 25990 }, { "grad_norm": 0.1198951005935669, "learning_rate": 8.545734480323053e-05, "loss": 0.0116, "step": 26000 }, { "grad_norm": 0.07686161249876022, "learning_rate": 8.544448971737939e-05, "loss": 0.0092, "step": 26010 }, { "grad_norm": 0.10544871538877487, "learning_rate": 8.543162992016746e-05, "loss": 0.0109, "step": 26020 }, { "grad_norm": 0.09862758964300156, "learning_rate": 8.54187654133041e-05, "loss": 0.0116, "step": 26030 }, { "grad_norm": 0.10101021081209183, "learning_rate": 8.540589619849927e-05, "loss": 0.0105, "step": 26040 }, { "grad_norm": 0.09090051800012589, "learning_rate": 8.539302227746359e-05, "loss": 0.0084, "step": 26050 }, { "grad_norm": 0.10179600119590759, "learning_rate": 8.538014365190829e-05, "loss": 0.0123, "step": 26060 }, { "grad_norm": 0.16423721611499786, "learning_rate": 8.536726032354523e-05, "loss": 0.0071, "step": 26070 }, { "grad_norm": 0.106484554708004, "learning_rate": 8.53543722940869e-05, "loss": 0.012, "step": 26080 }, { "grad_norm": 0.11128359287977219, "learning_rate": 8.534147956524637e-05, "loss": 0.0095, "step": 26090 }, { "grad_norm": 0.12909486889839172, "learning_rate": 8.53285821387374e-05, "loss": 0.0133, "step": 26100 }, { "grad_norm": 0.1372722089290619, "learning_rate": 8.531568001627432e-05, "loss": 0.0112, "step": 26110 }, { "grad_norm": 0.11839879304170609, "learning_rate": 8.530277319957215e-05, "loss": 0.01, "step": 26120 }, { "grad_norm": 0.12195441126823425, "learning_rate": 8.528986169034647e-05, "loss": 0.0088, "step": 26130 }, { "grad_norm": 0.09567153453826904, "learning_rate": 8.527694549031348e-05, "loss": 0.0078, "step": 26140 }, { "grad_norm": 0.1029144749045372, "learning_rate": 8.526402460119008e-05, "loss": 0.0132, "step": 26150 }, { "grad_norm": 0.08951260149478912, "learning_rate": 8.52510990246937e-05, "loss": 0.0107, "step": 26160 }, { "grad_norm": 0.07198506593704224, "learning_rate": 8.523816876254246e-05, "loss": 0.0098, "step": 26170 }, { "grad_norm": 0.09164810925722122, "learning_rate": 8.522523381645506e-05, "loss": 0.0122, "step": 26180 }, { "grad_norm": 0.06950429826974869, "learning_rate": 8.521229418815089e-05, "loss": 0.0065, "step": 26190 }, { "grad_norm": 0.11575779318809509, "learning_rate": 8.519934987934987e-05, "loss": 0.0127, "step": 26200 }, { "grad_norm": 0.09342413395643234, "learning_rate": 8.518640089177259e-05, "loss": 0.0111, "step": 26210 }, { "grad_norm": 0.13599710166454315, "learning_rate": 8.517344722714027e-05, "loss": 0.0112, "step": 26220 }, { "grad_norm": 0.07276458293199539, "learning_rate": 8.516048888717473e-05, "loss": 0.0117, "step": 26230 }, { "grad_norm": 0.07725398987531662, "learning_rate": 8.514752587359843e-05, "loss": 0.0083, "step": 26240 }, { "grad_norm": 0.09910321980714798, "learning_rate": 8.513455818813444e-05, "loss": 0.0095, "step": 26250 }, { "grad_norm": 0.10428117960691452, "learning_rate": 8.512158583250645e-05, "loss": 0.0094, "step": 26260 }, { "grad_norm": 0.06631171703338623, "learning_rate": 8.510860880843879e-05, "loss": 0.0103, "step": 26270 }, { "grad_norm": 0.10099279135465622, "learning_rate": 8.509562711765637e-05, "loss": 0.0078, "step": 26280 }, { "grad_norm": 0.1253102719783783, "learning_rate": 8.508264076188476e-05, "loss": 0.0105, "step": 26290 }, { "grad_norm": 0.09456022828817368, "learning_rate": 8.506964974285015e-05, "loss": 0.0087, "step": 26300 }, { "grad_norm": 0.06711684167385101, "learning_rate": 8.505665406227929e-05, "loss": 0.0096, "step": 26310 }, { "grad_norm": 0.10382568836212158, "learning_rate": 8.504365372189964e-05, "loss": 0.0073, "step": 26320 }, { "grad_norm": 0.05835402011871338, "learning_rate": 8.503064872343922e-05, "loss": 0.0113, "step": 26330 }, { "grad_norm": 0.06084451451897621, "learning_rate": 8.50176390686267e-05, "loss": 0.0086, "step": 26340 }, { "grad_norm": 0.12459378689527512, "learning_rate": 8.500462475919131e-05, "loss": 0.0109, "step": 26350 }, { "grad_norm": 0.127391517162323, "learning_rate": 8.499160579686298e-05, "loss": 0.0085, "step": 26360 }, { "grad_norm": 0.11495426297187805, "learning_rate": 8.497858218337217e-05, "loss": 0.0108, "step": 26370 }, { "grad_norm": 0.09365793317556381, "learning_rate": 8.496555392045009e-05, "loss": 0.013, "step": 26380 }, { "grad_norm": 0.09970036894083023, "learning_rate": 8.495252100982843e-05, "loss": 0.0071, "step": 26390 }, { "grad_norm": 0.06877957284450531, "learning_rate": 8.493948345323954e-05, "loss": 0.009, "step": 26400 }, { "grad_norm": 0.1125492975115776, "learning_rate": 8.492644125241643e-05, "loss": 0.011, "step": 26410 }, { "grad_norm": 0.09856465458869934, "learning_rate": 8.49133944090927e-05, "loss": 0.009, "step": 26420 }, { "grad_norm": 0.07410816848278046, "learning_rate": 8.490034292500255e-05, "loss": 0.0101, "step": 26430 }, { "grad_norm": 0.10733115673065186, "learning_rate": 8.488728680188083e-05, "loss": 0.0101, "step": 26440 }, { "grad_norm": 0.08112209290266037, "learning_rate": 8.487422604146297e-05, "loss": 0.0121, "step": 26450 }, { "grad_norm": 0.10590997338294983, "learning_rate": 8.486116064548505e-05, "loss": 0.0106, "step": 26460 }, { "grad_norm": 0.06159749627113342, "learning_rate": 8.484809061568372e-05, "loss": 0.0083, "step": 26470 }, { "grad_norm": 0.1384294629096985, "learning_rate": 8.483501595379635e-05, "loss": 0.0084, "step": 26480 }, { "grad_norm": 0.10660099238157272, "learning_rate": 8.482193666156078e-05, "loss": 0.0081, "step": 26490 }, { "grad_norm": 0.05185071751475334, "learning_rate": 8.480885274071557e-05, "loss": 0.0111, "step": 26500 }, { "grad_norm": 0.13569901883602142, "learning_rate": 8.479576419299986e-05, "loss": 0.011, "step": 26510 }, { "grad_norm": 0.12927618622779846, "learning_rate": 8.47826710201534e-05, "loss": 0.014, "step": 26520 }, { "grad_norm": 0.10309478640556335, "learning_rate": 8.476957322391659e-05, "loss": 0.0073, "step": 26530 }, { "grad_norm": 0.06939897686243057, "learning_rate": 8.475647080603039e-05, "loss": 0.0119, "step": 26540 }, { "grad_norm": 0.14068524539470673, "learning_rate": 8.474336376823641e-05, "loss": 0.0114, "step": 26550 }, { "grad_norm": 0.06721869856119156, "learning_rate": 8.473025211227688e-05, "loss": 0.0086, "step": 26560 }, { "grad_norm": 0.09374596923589706, "learning_rate": 8.471713583989461e-05, "loss": 0.0109, "step": 26570 }, { "grad_norm": 0.08475144952535629, "learning_rate": 8.470401495283306e-05, "loss": 0.0136, "step": 26580 }, { "grad_norm": 0.06687549501657486, "learning_rate": 8.469088945283627e-05, "loss": 0.009, "step": 26590 }, { "grad_norm": 0.0830700621008873, "learning_rate": 8.467775934164894e-05, "loss": 0.0133, "step": 26600 }, { "grad_norm": 0.14068423211574554, "learning_rate": 8.466462462101634e-05, "loss": 0.0108, "step": 26610 }, { "grad_norm": 0.0805605947971344, "learning_rate": 8.465148529268436e-05, "loss": 0.0084, "step": 26620 }, { "grad_norm": 0.1033690795302391, "learning_rate": 8.463834135839951e-05, "loss": 0.011, "step": 26630 }, { "grad_norm": 0.15032246708869934, "learning_rate": 8.462519281990892e-05, "loss": 0.0153, "step": 26640 }, { "grad_norm": 0.13206735253334045, "learning_rate": 8.461203967896032e-05, "loss": 0.0141, "step": 26650 }, { "grad_norm": 0.10716190189123154, "learning_rate": 8.459888193730203e-05, "loss": 0.0113, "step": 26660 }, { "grad_norm": 0.15499010682106018, "learning_rate": 8.458571959668304e-05, "loss": 0.0101, "step": 26670 }, { "grad_norm": 0.09221966564655304, "learning_rate": 8.457255265885292e-05, "loss": 0.0104, "step": 26680 }, { "grad_norm": 0.1035752072930336, "learning_rate": 8.455938112556182e-05, "loss": 0.0085, "step": 26690 }, { "grad_norm": 0.08754794299602509, "learning_rate": 8.454620499856053e-05, "loss": 0.0097, "step": 26700 }, { "grad_norm": 0.13806100189685822, "learning_rate": 8.453302427960048e-05, "loss": 0.0128, "step": 26710 }, { "grad_norm": 0.09161384403705597, "learning_rate": 8.451983897043367e-05, "loss": 0.0121, "step": 26720 }, { "grad_norm": 0.11721217632293701, "learning_rate": 8.450664907281272e-05, "loss": 0.0087, "step": 26730 }, { "grad_norm": 0.049038439989089966, "learning_rate": 8.449345458849084e-05, "loss": 0.0076, "step": 26740 }, { "grad_norm": 0.10252448171377182, "learning_rate": 8.44802555192219e-05, "loss": 0.0103, "step": 26750 }, { "grad_norm": 0.09891343861818314, "learning_rate": 8.446705186676033e-05, "loss": 0.0099, "step": 26760 }, { "grad_norm": 0.07702626287937164, "learning_rate": 8.44538436328612e-05, "loss": 0.0097, "step": 26770 }, { "grad_norm": 0.13015630841255188, "learning_rate": 8.444063081928016e-05, "loss": 0.0091, "step": 26780 }, { "grad_norm": 0.09921552985906601, "learning_rate": 8.442741342777351e-05, "loss": 0.0099, "step": 26790 }, { "grad_norm": 0.13285158574581146, "learning_rate": 8.441419146009815e-05, "loss": 0.0122, "step": 26800 }, { "grad_norm": 0.07752494513988495, "learning_rate": 8.44009649180115e-05, "loss": 0.0106, "step": 26810 }, { "grad_norm": 0.1382560282945633, "learning_rate": 8.438773380327174e-05, "loss": 0.0097, "step": 26820 }, { "grad_norm": 0.09410783648490906, "learning_rate": 8.437449811763755e-05, "loss": 0.009, "step": 26830 }, { "grad_norm": 0.06797752529382706, "learning_rate": 8.436125786286825e-05, "loss": 0.0137, "step": 26840 }, { "grad_norm": 0.10093577206134796, "learning_rate": 8.434801304072375e-05, "loss": 0.0115, "step": 26850 }, { "grad_norm": 0.1088542491197586, "learning_rate": 8.43347636529646e-05, "loss": 0.013, "step": 26860 }, { "grad_norm": 0.10293717682361603, "learning_rate": 8.432150970135193e-05, "loss": 0.01, "step": 26870 }, { "grad_norm": 0.12450189143419266, "learning_rate": 8.430825118764748e-05, "loss": 0.0136, "step": 26880 }, { "grad_norm": 0.07586189359426498, "learning_rate": 8.42949881136136e-05, "loss": 0.0083, "step": 26890 }, { "grad_norm": 0.08106004446744919, "learning_rate": 8.428172048101328e-05, "loss": 0.0116, "step": 26900 }, { "grad_norm": 0.1283414363861084, "learning_rate": 8.426844829161005e-05, "loss": 0.0106, "step": 26910 }, { "grad_norm": 0.06214028224349022, "learning_rate": 8.425517154716807e-05, "loss": 0.0087, "step": 26920 }, { "grad_norm": 0.08030548691749573, "learning_rate": 8.424189024945215e-05, "loss": 0.0073, "step": 26930 }, { "grad_norm": 0.057929281145334244, "learning_rate": 8.422860440022764e-05, "loss": 0.0146, "step": 26940 }, { "grad_norm": 0.09860936552286148, "learning_rate": 8.421531400126053e-05, "loss": 0.0112, "step": 26950 }, { "grad_norm": 0.13214455544948578, "learning_rate": 8.420201905431741e-05, "loss": 0.0115, "step": 26960 }, { "grad_norm": 0.08595852553844452, "learning_rate": 8.418871956116551e-05, "loss": 0.0094, "step": 26970 }, { "grad_norm": 0.058906588703393936, "learning_rate": 8.417541552357258e-05, "loss": 0.0084, "step": 26980 }, { "grad_norm": 0.08279543370008469, "learning_rate": 8.416210694330705e-05, "loss": 0.0113, "step": 26990 }, { "grad_norm": 0.09681646525859833, "learning_rate": 8.41487938221379e-05, "loss": 0.0087, "step": 27000 }, { "grad_norm": 0.07692262530326843, "learning_rate": 8.413547616183477e-05, "loss": 0.0122, "step": 27010 }, { "grad_norm": 0.0693737044930458, "learning_rate": 8.412215396416785e-05, "loss": 0.0081, "step": 27020 }, { "grad_norm": 0.08969562500715256, "learning_rate": 8.410882723090796e-05, "loss": 0.0083, "step": 27030 }, { "grad_norm": 0.09312170743942261, "learning_rate": 8.409549596382653e-05, "loss": 0.0083, "step": 27040 }, { "grad_norm": 0.1502491682767868, "learning_rate": 8.408216016469559e-05, "loss": 0.01, "step": 27050 }, { "grad_norm": 0.07238147407770157, "learning_rate": 8.406881983528775e-05, "loss": 0.0083, "step": 27060 }, { "grad_norm": 0.09174090623855591, "learning_rate": 8.405547497737623e-05, "loss": 0.01, "step": 27070 }, { "grad_norm": 0.06596962362527847, "learning_rate": 8.404212559273485e-05, "loss": 0.0112, "step": 27080 }, { "grad_norm": 0.08680492639541626, "learning_rate": 8.402877168313807e-05, "loss": 0.013, "step": 27090 }, { "grad_norm": 0.05213061720132828, "learning_rate": 8.401541325036091e-05, "loss": 0.0119, "step": 27100 }, { "grad_norm": 0.06920523196458817, "learning_rate": 8.4002050296179e-05, "loss": 0.0119, "step": 27110 }, { "grad_norm": 0.18542134761810303, "learning_rate": 8.398868282236858e-05, "loss": 0.0088, "step": 27120 }, { "grad_norm": 0.09448912739753723, "learning_rate": 8.397531083070646e-05, "loss": 0.0113, "step": 27130 }, { "grad_norm": 0.1220979243516922, "learning_rate": 8.39619343229701e-05, "loss": 0.0097, "step": 27140 }, { "grad_norm": 0.09076374769210815, "learning_rate": 8.394855330093754e-05, "loss": 0.0096, "step": 27150 }, { "grad_norm": 0.1139400377869606, "learning_rate": 8.393516776638739e-05, "loss": 0.012, "step": 27160 }, { "grad_norm": 0.12659864127635956, "learning_rate": 8.39217777210989e-05, "loss": 0.0089, "step": 27170 }, { "grad_norm": 0.11044338345527649, "learning_rate": 8.390838316685191e-05, "loss": 0.0158, "step": 27180 }, { "grad_norm": 0.07721253484487534, "learning_rate": 8.389498410542685e-05, "loss": 0.012, "step": 27190 }, { "grad_norm": 0.13218428194522858, "learning_rate": 8.388158053860476e-05, "loss": 0.011, "step": 27200 }, { "grad_norm": 0.09271993488073349, "learning_rate": 8.386817246816723e-05, "loss": 0.008, "step": 27210 }, { "grad_norm": 0.07504317164421082, "learning_rate": 8.385475989589655e-05, "loss": 0.0088, "step": 27220 }, { "grad_norm": 0.0716228038072586, "learning_rate": 8.384134282357552e-05, "loss": 0.008, "step": 27230 }, { "grad_norm": 0.1184549480676651, "learning_rate": 8.382792125298757e-05, "loss": 0.0086, "step": 27240 }, { "grad_norm": 0.10205277800559998, "learning_rate": 8.381449518591672e-05, "loss": 0.0101, "step": 27250 }, { "grad_norm": 0.07523203641176224, "learning_rate": 8.380106462414759e-05, "loss": 0.0113, "step": 27260 }, { "grad_norm": 0.09526703506708145, "learning_rate": 8.378762956946544e-05, "loss": 0.009, "step": 27270 }, { "grad_norm": 0.08401166647672653, "learning_rate": 8.377419002365603e-05, "loss": 0.01, "step": 27280 }, { "grad_norm": 0.08970766514539719, "learning_rate": 8.37607459885058e-05, "loss": 0.0108, "step": 27290 }, { "grad_norm": 0.12609387934207916, "learning_rate": 8.374729746580177e-05, "loss": 0.0117, "step": 27300 }, { "grad_norm": 0.08944875746965408, "learning_rate": 8.373384445733153e-05, "loss": 0.0079, "step": 27310 }, { "grad_norm": 0.06324076652526855, "learning_rate": 8.372038696488332e-05, "loss": 0.0091, "step": 27320 }, { "grad_norm": 0.07255896180868149, "learning_rate": 8.370692499024587e-05, "loss": 0.0094, "step": 27330 }, { "grad_norm": 0.09114528447389603, "learning_rate": 8.369345853520865e-05, "loss": 0.0094, "step": 27340 }, { "grad_norm": 0.05586252361536026, "learning_rate": 8.367998760156162e-05, "loss": 0.0117, "step": 27350 }, { "grad_norm": 0.12425155937671661, "learning_rate": 8.366651219109533e-05, "loss": 0.0126, "step": 27360 }, { "grad_norm": 0.12560489773750305, "learning_rate": 8.365303230560105e-05, "loss": 0.012, "step": 27370 }, { "grad_norm": 0.08940625190734863, "learning_rate": 8.363954794687048e-05, "loss": 0.0095, "step": 27380 }, { "grad_norm": 0.11699927598237991, "learning_rate": 8.362605911669602e-05, "loss": 0.0109, "step": 27390 }, { "grad_norm": 0.10709183663129807, "learning_rate": 8.361256581687062e-05, "loss": 0.0123, "step": 27400 }, { "grad_norm": 0.10570713877677917, "learning_rate": 8.359906804918786e-05, "loss": 0.0099, "step": 27410 }, { "grad_norm": 0.08134374022483826, "learning_rate": 8.35855658154419e-05, "loss": 0.0103, "step": 27420 }, { "grad_norm": 0.06365364044904709, "learning_rate": 8.357205911742744e-05, "loss": 0.0074, "step": 27430 }, { "grad_norm": 0.06377578526735306, "learning_rate": 8.355854795693987e-05, "loss": 0.0074, "step": 27440 }, { "grad_norm": 0.07629897445440292, "learning_rate": 8.35450323357751e-05, "loss": 0.0115, "step": 27450 }, { "grad_norm": 0.07180720567703247, "learning_rate": 8.353151225572968e-05, "loss": 0.016, "step": 27460 }, { "grad_norm": 0.09265264123678207, "learning_rate": 8.351798771860071e-05, "loss": 0.0098, "step": 27470 }, { "grad_norm": 0.11987277865409851, "learning_rate": 8.350445872618588e-05, "loss": 0.0085, "step": 27480 }, { "grad_norm": 0.11137008666992188, "learning_rate": 8.349092528028353e-05, "loss": 0.0132, "step": 27490 }, { "grad_norm": 0.10151451081037521, "learning_rate": 8.347738738269255e-05, "loss": 0.0119, "step": 27500 }, { "grad_norm": 0.14547103643417358, "learning_rate": 8.34638450352124e-05, "loss": 0.0097, "step": 27510 }, { "grad_norm": 0.10777736455202103, "learning_rate": 8.345029823964321e-05, "loss": 0.0115, "step": 27520 }, { "grad_norm": 0.11561178416013718, "learning_rate": 8.343674699778561e-05, "loss": 0.0066, "step": 27530 }, { "grad_norm": 0.08751921355724335, "learning_rate": 8.342319131144087e-05, "loss": 0.0088, "step": 27540 }, { "grad_norm": 0.10213208943605423, "learning_rate": 8.340963118241084e-05, "loss": 0.0083, "step": 27550 }, { "grad_norm": 0.08949514478445053, "learning_rate": 8.339606661249795e-05, "loss": 0.0094, "step": 27560 }, { "grad_norm": 0.1476498395204544, "learning_rate": 8.338249760350527e-05, "loss": 0.0118, "step": 27570 }, { "grad_norm": 0.12795059382915497, "learning_rate": 8.33689241572364e-05, "loss": 0.0084, "step": 27580 }, { "grad_norm": 0.09442561864852905, "learning_rate": 8.335534627549555e-05, "loss": 0.0074, "step": 27590 }, { "grad_norm": 0.11129459738731384, "learning_rate": 8.334176396008752e-05, "loss": 0.0094, "step": 27600 }, { "grad_norm": 0.10740846395492554, "learning_rate": 8.332817721281774e-05, "loss": 0.0109, "step": 27610 }, { "grad_norm": 0.06626289337873459, "learning_rate": 8.331458603549213e-05, "loss": 0.0086, "step": 27620 }, { "grad_norm": 0.10421054065227509, "learning_rate": 8.330099042991728e-05, "loss": 0.0104, "step": 27630 }, { "grad_norm": 0.1160256639122963, "learning_rate": 8.328739039790039e-05, "loss": 0.0098, "step": 27640 }, { "grad_norm": 0.0943794995546341, "learning_rate": 8.327378594124918e-05, "loss": 0.0137, "step": 27650 }, { "grad_norm": 0.09593942761421204, "learning_rate": 8.326017706177194e-05, "loss": 0.0092, "step": 27660 }, { "grad_norm": 0.11788050085306168, "learning_rate": 8.324656376127765e-05, "loss": 0.0111, "step": 27670 }, { "grad_norm": 0.10523426532745361, "learning_rate": 8.323294604157581e-05, "loss": 0.011, "step": 27680 }, { "grad_norm": 0.1223926991224289, "learning_rate": 8.321932390447651e-05, "loss": 0.0081, "step": 27690 }, { "grad_norm": 0.13781778514385223, "learning_rate": 8.320569735179042e-05, "loss": 0.0106, "step": 27700 }, { "grad_norm": 0.08396849036216736, "learning_rate": 8.319206638532883e-05, "loss": 0.0095, "step": 27710 }, { "grad_norm": 0.07727079093456268, "learning_rate": 8.317843100690361e-05, "loss": 0.0101, "step": 27720 }, { "grad_norm": 0.09194964915513992, "learning_rate": 8.316479121832719e-05, "loss": 0.0139, "step": 27730 }, { "grad_norm": 0.11277569085359573, "learning_rate": 8.31511470214126e-05, "loss": 0.0117, "step": 27740 }, { "grad_norm": 0.1758480817079544, "learning_rate": 8.313749841797345e-05, "loss": 0.0117, "step": 27750 }, { "grad_norm": 0.08451154828071594, "learning_rate": 8.312384540982396e-05, "loss": 0.0105, "step": 27760 }, { "grad_norm": 0.08705407381057739, "learning_rate": 8.31101879987789e-05, "loss": 0.0107, "step": 27770 }, { "grad_norm": 0.10326755046844482, "learning_rate": 8.309652618665365e-05, "loss": 0.0106, "step": 27780 }, { "grad_norm": 0.07390110939741135, "learning_rate": 8.308285997526418e-05, "loss": 0.0095, "step": 27790 }, { "grad_norm": 0.06694996356964111, "learning_rate": 8.306918936642703e-05, "loss": 0.0081, "step": 27800 }, { "grad_norm": 0.0819869264960289, "learning_rate": 8.305551436195931e-05, "loss": 0.0138, "step": 27810 }, { "grad_norm": 0.13785745203495026, "learning_rate": 8.304183496367877e-05, "loss": 0.0106, "step": 27820 }, { "grad_norm": 0.1030069887638092, "learning_rate": 8.302815117340366e-05, "loss": 0.0094, "step": 27830 }, { "grad_norm": 0.08661693334579468, "learning_rate": 8.301446299295288e-05, "loss": 0.0106, "step": 27840 }, { "grad_norm": 0.06827589124441147, "learning_rate": 8.300077042414592e-05, "loss": 0.0108, "step": 27850 }, { "grad_norm": 0.09771367907524109, "learning_rate": 8.298707346880276e-05, "loss": 0.0079, "step": 27860 }, { "grad_norm": 0.10083090513944626, "learning_rate": 8.297337212874413e-05, "loss": 0.0098, "step": 27870 }, { "grad_norm": 0.09681977331638336, "learning_rate": 8.295966640579114e-05, "loss": 0.0076, "step": 27880 }, { "grad_norm": 0.11403065919876099, "learning_rate": 8.294595630176563e-05, "loss": 0.0108, "step": 27890 }, { "grad_norm": 0.08633747696876526, "learning_rate": 8.293224181849e-05, "loss": 0.0088, "step": 27900 }, { "grad_norm": 0.12028142809867859, "learning_rate": 8.291852295778718e-05, "loss": 0.0116, "step": 27910 }, { "grad_norm": 0.08298647403717041, "learning_rate": 8.29047997214807e-05, "loss": 0.0102, "step": 27920 }, { "grad_norm": 0.10998567938804626, "learning_rate": 8.289107211139473e-05, "loss": 0.0126, "step": 27930 }, { "grad_norm": 0.06879378110170364, "learning_rate": 8.287734012935393e-05, "loss": 0.0106, "step": 27940 }, { "grad_norm": 0.08409539610147476, "learning_rate": 8.28636037771836e-05, "loss": 0.0097, "step": 27950 }, { "grad_norm": 0.1189541220664978, "learning_rate": 8.284986305670963e-05, "loss": 0.0086, "step": 27960 }, { "grad_norm": 0.0943315327167511, "learning_rate": 8.283611796975844e-05, "loss": 0.0091, "step": 27970 }, { "grad_norm": 0.09698384255170822, "learning_rate": 8.282236851815706e-05, "loss": 0.0092, "step": 27980 }, { "grad_norm": 0.0886671170592308, "learning_rate": 8.280861470373309e-05, "loss": 0.0082, "step": 27990 }, { "grad_norm": 0.0767369270324707, "learning_rate": 8.279485652831474e-05, "loss": 0.0085, "step": 28000 }, { "grad_norm": 0.14092057943344116, "learning_rate": 8.278109399373077e-05, "loss": 0.0071, "step": 28010 }, { "grad_norm": 0.1769407093524933, "learning_rate": 8.276732710181052e-05, "loss": 0.0148, "step": 28020 }, { "grad_norm": 0.08320345729589462, "learning_rate": 8.275355585438391e-05, "loss": 0.0118, "step": 28030 }, { "grad_norm": 0.05734670162200928, "learning_rate": 8.273978025328147e-05, "loss": 0.0111, "step": 28040 }, { "grad_norm": 0.064540795981884, "learning_rate": 8.272600030033428e-05, "loss": 0.0077, "step": 28050 }, { "grad_norm": 0.09825851768255234, "learning_rate": 8.271221599737398e-05, "loss": 0.0094, "step": 28060 }, { "grad_norm": 0.0849364623427391, "learning_rate": 8.269842734623281e-05, "loss": 0.0068, "step": 28070 }, { "grad_norm": 0.07237266004085541, "learning_rate": 8.26846343487436e-05, "loss": 0.0079, "step": 28080 }, { "grad_norm": 0.09855203330516815, "learning_rate": 8.267083700673976e-05, "loss": 0.0105, "step": 28090 }, { "grad_norm": 0.10422968119382858, "learning_rate": 8.265703532205525e-05, "loss": 0.0115, "step": 28100 }, { "grad_norm": 0.06826901435852051, "learning_rate": 8.264322929652461e-05, "loss": 0.0086, "step": 28110 }, { "grad_norm": 0.0871778354048729, "learning_rate": 8.262941893198297e-05, "loss": 0.01, "step": 28120 }, { "grad_norm": 0.06170464679598808, "learning_rate": 8.261560423026606e-05, "loss": 0.008, "step": 28130 }, { "grad_norm": 0.08655782788991928, "learning_rate": 8.260178519321014e-05, "loss": 0.0063, "step": 28140 }, { "grad_norm": 0.07284072786569595, "learning_rate": 8.258796182265207e-05, "loss": 0.0083, "step": 28150 }, { "grad_norm": 0.06582773476839066, "learning_rate": 8.257413412042928e-05, "loss": 0.0075, "step": 28160 }, { "grad_norm": 0.07772252708673477, "learning_rate": 8.256030208837979e-05, "loss": 0.0089, "step": 28170 }, { "grad_norm": 0.08266618847846985, "learning_rate": 8.254646572834217e-05, "loss": 0.0086, "step": 28180 }, { "grad_norm": 0.06441047042608261, "learning_rate": 8.253262504215558e-05, "loss": 0.0068, "step": 28190 }, { "grad_norm": 0.07170157134532928, "learning_rate": 8.251878003165979e-05, "loss": 0.0116, "step": 28200 }, { "grad_norm": 0.09358406066894531, "learning_rate": 8.250493069869505e-05, "loss": 0.0087, "step": 28210 }, { "grad_norm": 0.09887120872735977, "learning_rate": 8.249107704510228e-05, "loss": 0.0101, "step": 28220 }, { "grad_norm": 0.11657541990280151, "learning_rate": 8.247721907272295e-05, "loss": 0.012, "step": 28230 }, { "grad_norm": 0.08437523990869522, "learning_rate": 8.246335678339908e-05, "loss": 0.0089, "step": 28240 }, { "grad_norm": 0.10987073928117752, "learning_rate": 8.244949017897328e-05, "loss": 0.0112, "step": 28250 }, { "grad_norm": 0.14442628622055054, "learning_rate": 8.24356192612887e-05, "loss": 0.0099, "step": 28260 }, { "grad_norm": 0.08570786565542221, "learning_rate": 8.242174403218913e-05, "loss": 0.0098, "step": 28270 }, { "grad_norm": 0.09176193922758102, "learning_rate": 8.24078644935189e-05, "loss": 0.0084, "step": 28280 }, { "grad_norm": 0.10676486790180206, "learning_rate": 8.239398064712285e-05, "loss": 0.0075, "step": 28290 }, { "grad_norm": 0.0875450074672699, "learning_rate": 8.238009249484653e-05, "loss": 0.008, "step": 28300 }, { "grad_norm": 0.13245755434036255, "learning_rate": 8.236620003853595e-05, "loss": 0.0082, "step": 28310 }, { "grad_norm": 0.13670024275779724, "learning_rate": 8.235230328003771e-05, "loss": 0.0121, "step": 28320 }, { "grad_norm": 0.11288584023714066, "learning_rate": 8.233840222119903e-05, "loss": 0.0144, "step": 28330 }, { "grad_norm": 0.09973260760307312, "learning_rate": 8.232449686386766e-05, "loss": 0.0096, "step": 28340 }, { "grad_norm": 0.07865501940250397, "learning_rate": 8.23105872098919e-05, "loss": 0.0111, "step": 28350 }, { "grad_norm": 0.11345742642879486, "learning_rate": 8.22966732611207e-05, "loss": 0.009, "step": 28360 }, { "grad_norm": 0.09748728573322296, "learning_rate": 8.228275501940348e-05, "loss": 0.0099, "step": 28370 }, { "grad_norm": 0.08556656539440155, "learning_rate": 8.226883248659034e-05, "loss": 0.0093, "step": 28380 }, { "grad_norm": 0.1641949862241745, "learning_rate": 8.225490566453187e-05, "loss": 0.0089, "step": 28390 }, { "grad_norm": 0.07153428345918655, "learning_rate": 8.224097455507923e-05, "loss": 0.011, "step": 28400 }, { "grad_norm": 0.1025247871875763, "learning_rate": 8.22270391600842e-05, "loss": 0.0114, "step": 28410 }, { "grad_norm": 0.12150559574365616, "learning_rate": 8.221309948139912e-05, "loss": 0.0103, "step": 28420 }, { "grad_norm": 0.09431496262550354, "learning_rate": 8.219915552087687e-05, "loss": 0.0101, "step": 28430 }, { "grad_norm": 0.09213066101074219, "learning_rate": 8.218520728037089e-05, "loss": 0.0117, "step": 28440 }, { "grad_norm": 0.10590429604053497, "learning_rate": 8.217125476173524e-05, "loss": 0.0081, "step": 28450 }, { "grad_norm": 0.11228516697883606, "learning_rate": 8.215729796682449e-05, "loss": 0.012, "step": 28460 }, { "grad_norm": 0.0969759002327919, "learning_rate": 8.214333689749385e-05, "loss": 0.0094, "step": 28470 }, { "grad_norm": 0.1360156238079071, "learning_rate": 8.2129371555599e-05, "loss": 0.0109, "step": 28480 }, { "grad_norm": 0.0998532623052597, "learning_rate": 8.21154019429963e-05, "loss": 0.0124, "step": 28490 }, { "grad_norm": 0.14611215889453888, "learning_rate": 8.21014280615426e-05, "loss": 0.0099, "step": 28500 }, { "grad_norm": 0.0942913219332695, "learning_rate": 8.208744991309533e-05, "loss": 0.01, "step": 28510 }, { "grad_norm": 0.05266847833991051, "learning_rate": 8.20734674995125e-05, "loss": 0.0096, "step": 28520 }, { "grad_norm": 0.07142814248800278, "learning_rate": 8.205948082265271e-05, "loss": 0.0089, "step": 28530 }, { "grad_norm": 0.1136288195848465, "learning_rate": 8.204548988437507e-05, "loss": 0.0111, "step": 28540 }, { "grad_norm": 0.13756516575813293, "learning_rate": 8.203149468653929e-05, "loss": 0.0099, "step": 28550 }, { "grad_norm": 0.07409700751304626, "learning_rate": 8.201749523100566e-05, "loss": 0.0072, "step": 28560 }, { "grad_norm": 0.06343889981508255, "learning_rate": 8.2003491519635e-05, "loss": 0.0075, "step": 28570 }, { "grad_norm": 0.08918444812297821, "learning_rate": 8.198948355428872e-05, "loss": 0.01, "step": 28580 }, { "grad_norm": 0.08008188754320145, "learning_rate": 8.197547133682879e-05, "loss": 0.0089, "step": 28590 }, { "grad_norm": 0.12996964156627655, "learning_rate": 8.196145486911776e-05, "loss": 0.0115, "step": 28600 }, { "grad_norm": 0.06666003912687302, "learning_rate": 8.194743415301872e-05, "loss": 0.0095, "step": 28610 }, { "grad_norm": 0.11358999460935593, "learning_rate": 8.193340919039534e-05, "loss": 0.016, "step": 28620 }, { "grad_norm": 0.07177507877349854, "learning_rate": 8.191937998311182e-05, "loss": 0.0106, "step": 28630 }, { "grad_norm": 0.08478746563196182, "learning_rate": 8.190534653303299e-05, "loss": 0.0094, "step": 28640 }, { "grad_norm": 0.12178526818752289, "learning_rate": 8.189130884202419e-05, "loss": 0.0086, "step": 28650 }, { "grad_norm": 0.1116456538438797, "learning_rate": 8.187726691195134e-05, "loss": 0.007, "step": 28660 }, { "grad_norm": 0.1056450828909874, "learning_rate": 8.186322074468093e-05, "loss": 0.0115, "step": 28670 }, { "grad_norm": 0.05912921577692032, "learning_rate": 8.184917034208002e-05, "loss": 0.0102, "step": 28680 }, { "grad_norm": 0.08448104560375214, "learning_rate": 8.183511570601621e-05, "loss": 0.009, "step": 28690 }, { "grad_norm": 0.08245061337947845, "learning_rate": 8.182105683835766e-05, "loss": 0.0073, "step": 28700 }, { "grad_norm": 0.10068901628255844, "learning_rate": 8.180699374097312e-05, "loss": 0.0089, "step": 28710 }, { "grad_norm": 0.06066100671887398, "learning_rate": 8.179292641573188e-05, "loss": 0.0087, "step": 28720 }, { "grad_norm": 0.1038837879896164, "learning_rate": 8.177885486450383e-05, "loss": 0.0104, "step": 28730 }, { "grad_norm": 0.08766169846057892, "learning_rate": 8.176477908915934e-05, "loss": 0.0076, "step": 28740 }, { "grad_norm": 0.07542601972818375, "learning_rate": 8.175069909156945e-05, "loss": 0.0088, "step": 28750 }, { "grad_norm": 0.09545218199491501, "learning_rate": 8.173661487360566e-05, "loss": 0.0106, "step": 28760 }, { "grad_norm": 0.08609121292829514, "learning_rate": 8.17225264371401e-05, "loss": 0.0104, "step": 28770 }, { "grad_norm": 0.09577690064907074, "learning_rate": 8.170843378404542e-05, "loss": 0.0124, "step": 28780 }, { "grad_norm": 0.08254366368055344, "learning_rate": 8.169433691619487e-05, "loss": 0.0087, "step": 28790 }, { "grad_norm": 0.07581166923046112, "learning_rate": 8.168023583546222e-05, "loss": 0.009, "step": 28800 }, { "grad_norm": 0.10773879289627075, "learning_rate": 8.16661305437218e-05, "loss": 0.0083, "step": 28810 }, { "grad_norm": 0.07426997274160385, "learning_rate": 8.165202104284857e-05, "loss": 0.008, "step": 28820 }, { "grad_norm": 0.07976783812046051, "learning_rate": 8.163790733471796e-05, "loss": 0.0065, "step": 28830 }, { "grad_norm": 0.07899907231330872, "learning_rate": 8.162378942120599e-05, "loss": 0.0092, "step": 28840 }, { "grad_norm": 0.09535816311836243, "learning_rate": 8.160966730418924e-05, "loss": 0.0102, "step": 28850 }, { "grad_norm": 0.07650669664144516, "learning_rate": 8.159554098554489e-05, "loss": 0.0089, "step": 28860 }, { "grad_norm": 0.07047534734010696, "learning_rate": 8.158141046715062e-05, "loss": 0.0123, "step": 28870 }, { "grad_norm": 0.08060191571712494, "learning_rate": 8.156727575088468e-05, "loss": 0.0072, "step": 28880 }, { "grad_norm": 0.09831029921770096, "learning_rate": 8.15531368386259e-05, "loss": 0.0083, "step": 28890 }, { "grad_norm": 0.12276532500982285, "learning_rate": 8.153899373225365e-05, "loss": 0.0085, "step": 28900 }, { "grad_norm": 0.09840243309736252, "learning_rate": 8.152484643364788e-05, "loss": 0.0102, "step": 28910 }, { "grad_norm": 0.09304498881101608, "learning_rate": 8.151069494468905e-05, "loss": 0.0103, "step": 28920 }, { "grad_norm": 0.09000647813081741, "learning_rate": 8.149653926725823e-05, "loss": 0.0113, "step": 28930 }, { "grad_norm": 0.09049932658672333, "learning_rate": 8.148237940323702e-05, "loss": 0.0061, "step": 28940 }, { "grad_norm": 0.09687227010726929, "learning_rate": 8.14682153545076e-05, "loss": 0.0074, "step": 28950 }, { "grad_norm": 0.08411905169487, "learning_rate": 8.145404712295263e-05, "loss": 0.0076, "step": 28960 }, { "grad_norm": 0.07750435173511505, "learning_rate": 8.143987471045544e-05, "loss": 0.0096, "step": 28970 }, { "grad_norm": 0.10441742837429047, "learning_rate": 8.142569811889983e-05, "loss": 0.0093, "step": 28980 }, { "grad_norm": 0.1149798035621643, "learning_rate": 8.14115173501702e-05, "loss": 0.0141, "step": 28990 }, { "grad_norm": 0.08715546131134033, "learning_rate": 8.13973324061515e-05, "loss": 0.0081, "step": 29000 }, { "grad_norm": 0.08154135942459106, "learning_rate": 8.138314328872918e-05, "loss": 0.0106, "step": 29010 }, { "grad_norm": 0.06837329268455505, "learning_rate": 8.136894999978934e-05, "loss": 0.0091, "step": 29020 }, { "grad_norm": 0.13693088293075562, "learning_rate": 8.135475254121854e-05, "loss": 0.0064, "step": 29030 }, { "grad_norm": 0.11280364543199539, "learning_rate": 8.134055091490395e-05, "loss": 0.0111, "step": 29040 }, { "grad_norm": 0.14211423695087433, "learning_rate": 8.132634512273332e-05, "loss": 0.0075, "step": 29050 }, { "grad_norm": 0.13465003669261932, "learning_rate": 8.131213516659486e-05, "loss": 0.0104, "step": 29060 }, { "grad_norm": 0.09360141307115555, "learning_rate": 8.129792104837742e-05, "loss": 0.0076, "step": 29070 }, { "grad_norm": 0.09287966787815094, "learning_rate": 8.128370276997037e-05, "loss": 0.008, "step": 29080 }, { "grad_norm": 0.10948826372623444, "learning_rate": 8.126948033326364e-05, "loss": 0.0114, "step": 29090 }, { "grad_norm": 0.13845062255859375, "learning_rate": 8.125525374014769e-05, "loss": 0.0096, "step": 29100 }, { "grad_norm": 0.1690751612186432, "learning_rate": 8.124102299251356e-05, "loss": 0.0139, "step": 29110 }, { "grad_norm": 0.1257592737674713, "learning_rate": 8.122678809225283e-05, "loss": 0.0084, "step": 29120 }, { "grad_norm": 0.09674861282110214, "learning_rate": 8.121254904125766e-05, "loss": 0.0074, "step": 29130 }, { "grad_norm": 0.08060771226882935, "learning_rate": 8.11983058414207e-05, "loss": 0.009, "step": 29140 }, { "grad_norm": 0.07649558782577515, "learning_rate": 8.11840584946352e-05, "loss": 0.007, "step": 29150 }, { "grad_norm": 0.06948535144329071, "learning_rate": 8.116980700279495e-05, "loss": 0.0095, "step": 29160 }, { "grad_norm": 0.10386242717504501, "learning_rate": 8.11555513677943e-05, "loss": 0.0092, "step": 29170 }, { "grad_norm": 0.07920247316360474, "learning_rate": 8.114129159152812e-05, "loss": 0.01, "step": 29180 }, { "grad_norm": 0.06787914782762527, "learning_rate": 8.112702767589188e-05, "loss": 0.0065, "step": 29190 }, { "grad_norm": 0.0873769074678421, "learning_rate": 8.111275962278155e-05, "loss": 0.0116, "step": 29200 }, { "grad_norm": 0.09125498682260513, "learning_rate": 8.109848743409367e-05, "loss": 0.0088, "step": 29210 }, { "grad_norm": 0.11742549389600754, "learning_rate": 8.108421111172535e-05, "loss": 0.0126, "step": 29220 }, { "grad_norm": 0.14291730523109436, "learning_rate": 8.10699306575742e-05, "loss": 0.0102, "step": 29230 }, { "grad_norm": 0.08528514206409454, "learning_rate": 8.105564607353846e-05, "loss": 0.009, "step": 29240 }, { "grad_norm": 0.1004851683974266, "learning_rate": 8.104135736151681e-05, "loss": 0.0068, "step": 29250 }, { "grad_norm": 0.14388976991176605, "learning_rate": 8.102706452340858e-05, "loss": 0.011, "step": 29260 }, { "grad_norm": 0.11684127897024155, "learning_rate": 8.101276756111359e-05, "loss": 0.0112, "step": 29270 }, { "grad_norm": 0.12799212336540222, "learning_rate": 8.099846647653222e-05, "loss": 0.0077, "step": 29280 }, { "grad_norm": 0.10810190439224243, "learning_rate": 8.098416127156542e-05, "loss": 0.01, "step": 29290 }, { "grad_norm": 0.10323517769575119, "learning_rate": 8.096985194811464e-05, "loss": 0.0073, "step": 29300 }, { "grad_norm": 0.08163919299840927, "learning_rate": 8.095553850808193e-05, "loss": 0.0086, "step": 29310 }, { "grad_norm": 0.12185795605182648, "learning_rate": 8.094122095336987e-05, "loss": 0.01, "step": 29320 }, { "grad_norm": 0.11943917721509933, "learning_rate": 8.092689928588155e-05, "loss": 0.0122, "step": 29330 }, { "grad_norm": 0.08936722576618195, "learning_rate": 8.091257350752069e-05, "loss": 0.008, "step": 29340 }, { "grad_norm": 0.12004336714744568, "learning_rate": 8.089824362019145e-05, "loss": 0.0114, "step": 29350 }, { "grad_norm": 0.10904107987880707, "learning_rate": 8.088390962579861e-05, "loss": 0.0077, "step": 29360 }, { "grad_norm": 0.07865209877490997, "learning_rate": 8.08695715262475e-05, "loss": 0.0066, "step": 29370 }, { "grad_norm": 0.09464263170957565, "learning_rate": 8.085522932344394e-05, "loss": 0.0103, "step": 29380 }, { "grad_norm": 0.0733603984117508, "learning_rate": 8.084088301929434e-05, "loss": 0.0104, "step": 29390 }, { "grad_norm": 0.11425170302391052, "learning_rate": 8.082653261570565e-05, "loss": 0.0111, "step": 29400 }, { "grad_norm": 0.10635003447532654, "learning_rate": 8.081217811458535e-05, "loss": 0.0085, "step": 29410 }, { "grad_norm": 0.1286846399307251, "learning_rate": 8.079781951784147e-05, "loss": 0.0086, "step": 29420 }, { "grad_norm": 0.11330016702413559, "learning_rate": 8.078345682738258e-05, "loss": 0.0098, "step": 29430 }, { "grad_norm": 0.12335922569036484, "learning_rate": 8.076909004511782e-05, "loss": 0.0084, "step": 29440 }, { "grad_norm": 0.11196984350681305, "learning_rate": 8.075471917295684e-05, "loss": 0.0098, "step": 29450 }, { "grad_norm": 0.07324041426181793, "learning_rate": 8.074034421280986e-05, "loss": 0.0113, "step": 29460 }, { "grad_norm": 0.05036013945937157, "learning_rate": 8.072596516658764e-05, "loss": 0.0074, "step": 29470 }, { "grad_norm": 0.08245670795440674, "learning_rate": 8.071158203620144e-05, "loss": 0.0091, "step": 29480 }, { "grad_norm": 0.09228549152612686, "learning_rate": 8.069719482356312e-05, "loss": 0.0098, "step": 29490 }, { "grad_norm": 0.07656966149806976, "learning_rate": 8.068280353058506e-05, "loss": 0.0064, "step": 29500 }, { "grad_norm": 0.09156869351863861, "learning_rate": 8.06684081591802e-05, "loss": 0.0117, "step": 29510 }, { "grad_norm": 0.11333348602056503, "learning_rate": 8.065400871126197e-05, "loss": 0.0097, "step": 29520 }, { "grad_norm": 0.09093157947063446, "learning_rate": 8.063960518874439e-05, "loss": 0.0095, "step": 29530 }, { "grad_norm": 0.08313403278589249, "learning_rate": 8.062519759354201e-05, "loss": 0.0074, "step": 29540 }, { "grad_norm": 0.06961674988269806, "learning_rate": 8.061078592756995e-05, "loss": 0.0075, "step": 29550 }, { "grad_norm": 0.09244231134653091, "learning_rate": 8.059637019274378e-05, "loss": 0.0107, "step": 29560 }, { "grad_norm": 0.08455915749073029, "learning_rate": 8.058195039097971e-05, "loss": 0.0084, "step": 29570 }, { "grad_norm": 0.09190814197063446, "learning_rate": 8.056752652419446e-05, "loss": 0.0107, "step": 29580 }, { "grad_norm": 0.11635535210371017, "learning_rate": 8.055309859430526e-05, "loss": 0.0108, "step": 29590 }, { "grad_norm": 0.09647969901561737, "learning_rate": 8.053866660322991e-05, "loss": 0.0118, "step": 29600 }, { "grad_norm": 0.14912031590938568, "learning_rate": 8.052423055288676e-05, "loss": 0.0101, "step": 29610 }, { "grad_norm": 0.0721297413110733, "learning_rate": 8.050979044519466e-05, "loss": 0.0084, "step": 29620 }, { "grad_norm": 0.09254085272550583, "learning_rate": 8.049534628207302e-05, "loss": 0.009, "step": 29630 }, { "grad_norm": 0.10279138386249542, "learning_rate": 8.04808980654418e-05, "loss": 0.009, "step": 29640 }, { "grad_norm": 0.07519859820604324, "learning_rate": 8.04664457972215e-05, "loss": 0.0103, "step": 29650 }, { "grad_norm": 0.11790375411510468, "learning_rate": 8.045198947933312e-05, "loss": 0.0101, "step": 29660 }, { "grad_norm": 0.15393409132957458, "learning_rate": 8.043752911369824e-05, "loss": 0.0109, "step": 29670 }, { "grad_norm": 0.12823620438575745, "learning_rate": 8.042306470223897e-05, "loss": 0.0107, "step": 29680 }, { "grad_norm": 0.13504669070243835, "learning_rate": 8.040859624687795e-05, "loss": 0.0088, "step": 29690 }, { "grad_norm": 0.09701218456029892, "learning_rate": 8.039412374953837e-05, "loss": 0.0076, "step": 29700 }, { "grad_norm": 0.12337461113929749, "learning_rate": 8.037964721214392e-05, "loss": 0.0104, "step": 29710 }, { "grad_norm": 0.07216765731573105, "learning_rate": 8.036516663661886e-05, "loss": 0.0092, "step": 29720 }, { "grad_norm": 0.10120765119791031, "learning_rate": 8.0350682024888e-05, "loss": 0.0096, "step": 29730 }, { "grad_norm": 0.1605822741985321, "learning_rate": 8.033619337887666e-05, "loss": 0.0097, "step": 29740 }, { "grad_norm": 0.12001733481884003, "learning_rate": 8.032170070051068e-05, "loss": 0.0086, "step": 29750 }, { "grad_norm": 0.14134392142295837, "learning_rate": 8.030720399171648e-05, "loss": 0.0141, "step": 29760 }, { "grad_norm": 0.15607358515262604, "learning_rate": 8.0292703254421e-05, "loss": 0.0085, "step": 29770 }, { "grad_norm": 0.14476172626018524, "learning_rate": 8.02781984905517e-05, "loss": 0.0102, "step": 29780 }, { "grad_norm": 0.11840096861124039, "learning_rate": 8.02636897020366e-05, "loss": 0.0113, "step": 29790 }, { "grad_norm": 0.12528708577156067, "learning_rate": 8.02491768908042e-05, "loss": 0.0114, "step": 29800 }, { "grad_norm": 0.0880979374051094, "learning_rate": 8.023466005878364e-05, "loss": 0.0101, "step": 29810 }, { "grad_norm": 0.0965082049369812, "learning_rate": 8.022013920790447e-05, "loss": 0.0126, "step": 29820 }, { "grad_norm": 0.08531758934259415, "learning_rate": 8.020561434009686e-05, "loss": 0.0082, "step": 29830 }, { "grad_norm": 0.08556083589792252, "learning_rate": 8.019108545729149e-05, "loss": 0.0114, "step": 29840 }, { "grad_norm": 0.09588164836168289, "learning_rate": 8.017655256141957e-05, "loss": 0.0094, "step": 29850 }, { "grad_norm": 0.15608641505241394, "learning_rate": 8.016201565441283e-05, "loss": 0.0123, "step": 29860 }, { "grad_norm": 0.06319836527109146, "learning_rate": 8.014747473820356e-05, "loss": 0.0124, "step": 29870 }, { "grad_norm": 0.05900290235877037, "learning_rate": 8.01329298147246e-05, "loss": 0.0071, "step": 29880 }, { "grad_norm": 0.11029605567455292, "learning_rate": 8.011838088590923e-05, "loss": 0.0078, "step": 29890 }, { "grad_norm": 0.11684218049049377, "learning_rate": 8.010382795369138e-05, "loss": 0.0117, "step": 29900 }, { "grad_norm": 0.1169179230928421, "learning_rate": 8.008927102000544e-05, "loss": 0.009, "step": 29910 }, { "grad_norm": 0.1484064906835556, "learning_rate": 8.007471008678635e-05, "loss": 0.0076, "step": 29920 }, { "grad_norm": 0.10862529277801514, "learning_rate": 8.006014515596956e-05, "loss": 0.0097, "step": 29930 }, { "grad_norm": 0.1207486018538475, "learning_rate": 8.004557622949112e-05, "loss": 0.0088, "step": 29940 }, { "grad_norm": 0.1540881097316742, "learning_rate": 8.003100330928753e-05, "loss": 0.0103, "step": 29950 }, { "grad_norm": 0.11485716700553894, "learning_rate": 8.001642639729586e-05, "loss": 0.0113, "step": 29960 }, { "grad_norm": 0.09725905954837799, "learning_rate": 8.000184549545373e-05, "loss": 0.0095, "step": 29970 }, { "grad_norm": 0.12122983485460281, "learning_rate": 7.998726060569922e-05, "loss": 0.0078, "step": 29980 }, { "grad_norm": 0.07613930851221085, "learning_rate": 7.997267172997103e-05, "loss": 0.0088, "step": 29990 }, { "grad_norm": 0.09773487597703934, "learning_rate": 7.995807887020832e-05, "loss": 0.0098, "step": 30000 }, { "grad_norm": 0.11440173536539078, "learning_rate": 7.99434820283508e-05, "loss": 0.013, "step": 30010 }, { "grad_norm": 0.13243043422698975, "learning_rate": 7.992888120633874e-05, "loss": 0.0106, "step": 30020 }, { "grad_norm": 0.09256302565336227, "learning_rate": 7.991427640611292e-05, "loss": 0.0105, "step": 30030 }, { "grad_norm": 0.1167430728673935, "learning_rate": 7.98996676296146e-05, "loss": 0.0107, "step": 30040 }, { "grad_norm": 0.07259654253721237, "learning_rate": 7.988505487878566e-05, "loss": 0.0077, "step": 30050 }, { "grad_norm": 0.09575348347425461, "learning_rate": 7.987043815556842e-05, "loss": 0.0101, "step": 30060 }, { "grad_norm": 0.08635278791189194, "learning_rate": 7.985581746190578e-05, "loss": 0.0092, "step": 30070 }, { "grad_norm": 0.05695224553346634, "learning_rate": 7.984119279974116e-05, "loss": 0.007, "step": 30080 }, { "grad_norm": 0.11133143305778503, "learning_rate": 7.98265641710185e-05, "loss": 0.0109, "step": 30090 }, { "grad_norm": 0.10861673951148987, "learning_rate": 7.981193157768229e-05, "loss": 0.0096, "step": 30100 }, { "grad_norm": 0.08403109014034271, "learning_rate": 7.979729502167748e-05, "loss": 0.0108, "step": 30110 }, { "grad_norm": 0.10442891716957092, "learning_rate": 7.978265450494964e-05, "loss": 0.009, "step": 30120 }, { "grad_norm": 0.11428078263998032, "learning_rate": 7.97680100294448e-05, "loss": 0.013, "step": 30130 }, { "grad_norm": 0.11233368515968323, "learning_rate": 7.975336159710953e-05, "loss": 0.0104, "step": 30140 }, { "grad_norm": 0.11589328944683075, "learning_rate": 7.973870920989094e-05, "loss": 0.0132, "step": 30150 }, { "grad_norm": 0.14362147450447083, "learning_rate": 7.972405286973668e-05, "loss": 0.012, "step": 30160 }, { "grad_norm": 0.10799778252840042, "learning_rate": 7.970939257859487e-05, "loss": 0.0131, "step": 30170 }, { "grad_norm": 0.08752886205911636, "learning_rate": 7.969472833841421e-05, "loss": 0.0125, "step": 30180 }, { "grad_norm": 0.08197042346000671, "learning_rate": 7.968006015114387e-05, "loss": 0.0113, "step": 30190 }, { "grad_norm": 0.07543439418077469, "learning_rate": 7.966538801873362e-05, "loss": 0.0094, "step": 30200 }, { "grad_norm": 0.05909155681729317, "learning_rate": 7.965071194313369e-05, "loss": 0.0095, "step": 30210 }, { "grad_norm": 0.08447954058647156, "learning_rate": 7.963603192629487e-05, "loss": 0.0111, "step": 30220 }, { "grad_norm": 0.1153082475066185, "learning_rate": 7.962134797016844e-05, "loss": 0.01, "step": 30230 }, { "grad_norm": 0.07841707766056061, "learning_rate": 7.960666007670625e-05, "loss": 0.0094, "step": 30240 }, { "grad_norm": 0.06790165603160858, "learning_rate": 7.959196824786064e-05, "loss": 0.0064, "step": 30250 }, { "grad_norm": 0.11758274585008621, "learning_rate": 7.957727248558447e-05, "loss": 0.0098, "step": 30260 }, { "grad_norm": 0.1299402117729187, "learning_rate": 7.956257279183112e-05, "loss": 0.0094, "step": 30270 }, { "grad_norm": 0.12329623103141785, "learning_rate": 7.954786916855454e-05, "loss": 0.0108, "step": 30280 }, { "grad_norm": 0.08497153222560883, "learning_rate": 7.953316161770916e-05, "loss": 0.0085, "step": 30290 }, { "grad_norm": 0.08393549919128418, "learning_rate": 7.951845014124994e-05, "loss": 0.0083, "step": 30300 }, { "grad_norm": 0.11041302233934402, "learning_rate": 7.950373474113234e-05, "loss": 0.0092, "step": 30310 }, { "grad_norm": 0.09011387825012207, "learning_rate": 7.948901541931239e-05, "loss": 0.009, "step": 30320 }, { "grad_norm": 0.1114831417798996, "learning_rate": 7.94742921777466e-05, "loss": 0.0126, "step": 30330 }, { "grad_norm": 0.13965964317321777, "learning_rate": 7.945956501839202e-05, "loss": 0.0087, "step": 30340 }, { "grad_norm": 0.05836295336484909, "learning_rate": 7.944483394320623e-05, "loss": 0.0081, "step": 30350 }, { "grad_norm": 0.12724266946315765, "learning_rate": 7.94300989541473e-05, "loss": 0.0108, "step": 30360 }, { "grad_norm": 0.1115584522485733, "learning_rate": 7.941536005317384e-05, "loss": 0.0107, "step": 30370 }, { "grad_norm": 0.10466811805963516, "learning_rate": 7.940061724224499e-05, "loss": 0.014, "step": 30380 }, { "grad_norm": 0.07647629827260971, "learning_rate": 7.938587052332038e-05, "loss": 0.0107, "step": 30390 }, { "grad_norm": 0.08723464608192444, "learning_rate": 7.937111989836018e-05, "loss": 0.0108, "step": 30400 }, { "grad_norm": 0.1263715624809265, "learning_rate": 7.935636536932509e-05, "loss": 0.0088, "step": 30410 }, { "grad_norm": 0.08860370516777039, "learning_rate": 7.934160693817631e-05, "loss": 0.0111, "step": 30420 }, { "grad_norm": 0.061787959188222885, "learning_rate": 7.932684460687556e-05, "loss": 0.0084, "step": 30430 }, { "grad_norm": 0.08034326136112213, "learning_rate": 7.93120783773851e-05, "loss": 0.0101, "step": 30440 }, { "grad_norm": 0.11525937914848328, "learning_rate": 7.929730825166764e-05, "loss": 0.0094, "step": 30450 }, { "grad_norm": 0.08992043137550354, "learning_rate": 7.92825342316865e-05, "loss": 0.0121, "step": 30460 }, { "grad_norm": 0.10495658963918686, "learning_rate": 7.926775631940549e-05, "loss": 0.01, "step": 30470 }, { "grad_norm": 0.10160989314317703, "learning_rate": 7.925297451678886e-05, "loss": 0.0092, "step": 30480 }, { "grad_norm": 0.13904719054698944, "learning_rate": 7.923818882580151e-05, "loss": 0.0102, "step": 30490 }, { "grad_norm": 0.11445001512765884, "learning_rate": 7.922339924840874e-05, "loss": 0.0082, "step": 30500 }, { "grad_norm": 0.06671202182769775, "learning_rate": 7.920860578657645e-05, "loss": 0.0088, "step": 30510 }, { "grad_norm": 0.08742320537567139, "learning_rate": 7.919380844227097e-05, "loss": 0.0087, "step": 30520 }, { "grad_norm": 0.07712867856025696, "learning_rate": 7.917900721745925e-05, "loss": 0.0136, "step": 30530 }, { "grad_norm": 0.09279125183820724, "learning_rate": 7.916420211410868e-05, "loss": 0.0096, "step": 30540 }, { "grad_norm": 0.1284012496471405, "learning_rate": 7.914939313418719e-05, "loss": 0.0101, "step": 30550 }, { "grad_norm": 0.11745645105838776, "learning_rate": 7.91345802796632e-05, "loss": 0.0097, "step": 30560 }, { "grad_norm": 0.08272481709718704, "learning_rate": 7.91197635525057e-05, "loss": 0.0114, "step": 30570 }, { "grad_norm": 0.08174553513526917, "learning_rate": 7.910494295468416e-05, "loss": 0.0098, "step": 30580 }, { "grad_norm": 0.104304239153862, "learning_rate": 7.909011848816854e-05, "loss": 0.0124, "step": 30590 }, { "grad_norm": 0.10693308711051941, "learning_rate": 7.907529015492936e-05, "loss": 0.0118, "step": 30600 }, { "grad_norm": 0.07038572430610657, "learning_rate": 7.906045795693762e-05, "loss": 0.0111, "step": 30610 }, { "grad_norm": 0.08000587671995163, "learning_rate": 7.90456218961649e-05, "loss": 0.0072, "step": 30620 }, { "grad_norm": 0.06026219576597214, "learning_rate": 7.903078197458316e-05, "loss": 0.0057, "step": 30630 }, { "grad_norm": 0.10585861653089523, "learning_rate": 7.901593819416502e-05, "loss": 0.0066, "step": 30640 }, { "grad_norm": 0.09584634751081467, "learning_rate": 7.900109055688352e-05, "loss": 0.0075, "step": 30650 }, { "grad_norm": 0.06840531527996063, "learning_rate": 7.898623906471226e-05, "loss": 0.0117, "step": 30660 }, { "grad_norm": 0.08825749158859253, "learning_rate": 7.89713837196253e-05, "loss": 0.0085, "step": 30670 }, { "grad_norm": 0.10010726749897003, "learning_rate": 7.895652452359728e-05, "loss": 0.0092, "step": 30680 }, { "grad_norm": 0.0978522077202797, "learning_rate": 7.894166147860331e-05, "loss": 0.0102, "step": 30690 }, { "grad_norm": 0.09795647114515305, "learning_rate": 7.892679458661902e-05, "loss": 0.0088, "step": 30700 }, { "grad_norm": 0.04510179162025452, "learning_rate": 7.891192384962053e-05, "loss": 0.0103, "step": 30710 }, { "grad_norm": 0.08419062942266464, "learning_rate": 7.889704926958452e-05, "loss": 0.0087, "step": 30720 }, { "grad_norm": 0.08249789476394653, "learning_rate": 7.888217084848814e-05, "loss": 0.0128, "step": 30730 }, { "grad_norm": 0.08407502621412277, "learning_rate": 7.886728858830905e-05, "loss": 0.0086, "step": 30740 }, { "grad_norm": 0.07706359773874283, "learning_rate": 7.885240249102546e-05, "loss": 0.0093, "step": 30750 }, { "grad_norm": 0.07383841276168823, "learning_rate": 7.883751255861604e-05, "loss": 0.0075, "step": 30760 }, { "grad_norm": 0.10283675044775009, "learning_rate": 7.882261879306e-05, "loss": 0.0137, "step": 30770 }, { "grad_norm": 0.11932647973299026, "learning_rate": 7.880772119633706e-05, "loss": 0.0088, "step": 30780 }, { "grad_norm": 0.1211123839020729, "learning_rate": 7.879281977042744e-05, "loss": 0.0127, "step": 30790 }, { "grad_norm": 0.11842736601829529, "learning_rate": 7.877791451731186e-05, "loss": 0.0078, "step": 30800 }, { "grad_norm": 0.07595274597406387, "learning_rate": 7.876300543897157e-05, "loss": 0.0062, "step": 30810 }, { "grad_norm": 0.07477406412363052, "learning_rate": 7.874809253738831e-05, "loss": 0.0072, "step": 30820 }, { "grad_norm": 0.08312606066465378, "learning_rate": 7.873317581454435e-05, "loss": 0.0115, "step": 30830 }, { "grad_norm": 0.106267049908638, "learning_rate": 7.871825527242245e-05, "loss": 0.0118, "step": 30840 }, { "grad_norm": 0.0732346624135971, "learning_rate": 7.870333091300585e-05, "loss": 0.0101, "step": 30850 }, { "grad_norm": 0.0981028825044632, "learning_rate": 7.868840273827837e-05, "loss": 0.0079, "step": 30860 }, { "grad_norm": 0.1362134963274002, "learning_rate": 7.86734707502243e-05, "loss": 0.014, "step": 30870 }, { "grad_norm": 0.12401212006807327, "learning_rate": 7.86585349508284e-05, "loss": 0.0103, "step": 30880 }, { "grad_norm": 0.049873072654008865, "learning_rate": 7.8643595342076e-05, "loss": 0.0079, "step": 30890 }, { "grad_norm": 0.14423328638076782, "learning_rate": 7.862865192595287e-05, "loss": 0.0088, "step": 30900 }, { "grad_norm": 0.09891124069690704, "learning_rate": 7.861370470444535e-05, "loss": 0.0102, "step": 30910 }, { "grad_norm": 0.06991274654865265, "learning_rate": 7.859875367954025e-05, "loss": 0.0079, "step": 30920 }, { "grad_norm": 0.0845687985420227, "learning_rate": 7.85837988532249e-05, "loss": 0.0088, "step": 30930 }, { "grad_norm": 0.08652437478303909, "learning_rate": 7.856884022748713e-05, "loss": 0.0072, "step": 30940 }, { "grad_norm": 0.06743136048316956, "learning_rate": 7.855387780431527e-05, "loss": 0.0125, "step": 30950 }, { "grad_norm": 0.057939741760492325, "learning_rate": 7.853891158569815e-05, "loss": 0.0089, "step": 30960 }, { "grad_norm": 0.07639895379543304, "learning_rate": 7.852394157362513e-05, "loss": 0.0105, "step": 30970 }, { "grad_norm": 0.08238600939512253, "learning_rate": 7.850896777008603e-05, "loss": 0.0119, "step": 30980 }, { "grad_norm": 0.09161147475242615, "learning_rate": 7.849399017707125e-05, "loss": 0.0121, "step": 30990 }, { "grad_norm": 0.14456604421138763, "learning_rate": 7.847900879657158e-05, "loss": 0.0121, "step": 31000 }, { "grad_norm": 0.09271831065416336, "learning_rate": 7.846402363057844e-05, "loss": 0.0089, "step": 31010 }, { "grad_norm": 0.07989078015089035, "learning_rate": 7.844903468108364e-05, "loss": 0.0103, "step": 31020 }, { "grad_norm": 0.1576087474822998, "learning_rate": 7.843404195007959e-05, "loss": 0.0104, "step": 31030 }, { "grad_norm": 0.12426378577947617, "learning_rate": 7.841904543955911e-05, "loss": 0.0113, "step": 31040 }, { "grad_norm": 0.10132981091737747, "learning_rate": 7.840404515151563e-05, "loss": 0.008, "step": 31050 }, { "grad_norm": 0.131244495511055, "learning_rate": 7.838904108794297e-05, "loss": 0.0092, "step": 31060 }, { "grad_norm": 0.09755739569664001, "learning_rate": 7.837403325083554e-05, "loss": 0.01, "step": 31070 }, { "grad_norm": 0.10101649910211563, "learning_rate": 7.835902164218818e-05, "loss": 0.0089, "step": 31080 }, { "grad_norm": 0.1043759360909462, "learning_rate": 7.834400626399629e-05, "loss": 0.0106, "step": 31090 }, { "grad_norm": 0.06388406455516815, "learning_rate": 7.832898711825573e-05, "loss": 0.0083, "step": 31100 }, { "grad_norm": 0.12133220583200455, "learning_rate": 7.831396420696291e-05, "loss": 0.0138, "step": 31110 }, { "grad_norm": 0.10196302086114883, "learning_rate": 7.829893753211467e-05, "loss": 0.0079, "step": 31120 }, { "grad_norm": 0.1094527468085289, "learning_rate": 7.828390709570842e-05, "loss": 0.0113, "step": 31130 }, { "grad_norm": 0.10709397494792938, "learning_rate": 7.826887289974203e-05, "loss": 0.01, "step": 31140 }, { "grad_norm": 0.10267851501703262, "learning_rate": 7.825383494621385e-05, "loss": 0.012, "step": 31150 }, { "grad_norm": 0.08131921291351318, "learning_rate": 7.823879323712282e-05, "loss": 0.0092, "step": 31160 }, { "grad_norm": 0.06062131002545357, "learning_rate": 7.822374777446827e-05, "loss": 0.0075, "step": 31170 }, { "grad_norm": 0.07089122384786606, "learning_rate": 7.820869856025009e-05, "loss": 0.0076, "step": 31180 }, { "grad_norm": 0.0884275734424591, "learning_rate": 7.819364559646865e-05, "loss": 0.0088, "step": 31190 }, { "grad_norm": 0.13411520421504974, "learning_rate": 7.817858888512481e-05, "loss": 0.0085, "step": 31200 }, { "grad_norm": 0.11503110826015472, "learning_rate": 7.816352842822e-05, "loss": 0.0077, "step": 31210 }, { "grad_norm": 0.13316798210144043, "learning_rate": 7.814846422775602e-05, "loss": 0.0114, "step": 31220 }, { "grad_norm": 0.07818348705768585, "learning_rate": 7.813339628573527e-05, "loss": 0.0069, "step": 31230 }, { "grad_norm": 0.07332823425531387, "learning_rate": 7.811832460416062e-05, "loss": 0.0103, "step": 31240 }, { "grad_norm": 0.10433588176965714, "learning_rate": 7.81032491850354e-05, "loss": 0.0089, "step": 31250 }, { "grad_norm": 0.09320426732301712, "learning_rate": 7.808817003036353e-05, "loss": 0.0067, "step": 31260 }, { "grad_norm": 0.092918261885643, "learning_rate": 7.807308714214929e-05, "loss": 0.0085, "step": 31270 }, { "grad_norm": 0.08978880941867828, "learning_rate": 7.805800052239757e-05, "loss": 0.0083, "step": 31280 }, { "grad_norm": 0.1254810243844986, "learning_rate": 7.804291017311374e-05, "loss": 0.0094, "step": 31290 }, { "grad_norm": 0.0939515084028244, "learning_rate": 7.802781609630359e-05, "loss": 0.0079, "step": 31300 }, { "grad_norm": 0.08094002306461334, "learning_rate": 7.801271829397348e-05, "loss": 0.0067, "step": 31310 }, { "grad_norm": 0.12303639948368073, "learning_rate": 7.799761676813027e-05, "loss": 0.0117, "step": 31320 }, { "grad_norm": 0.11400330066680908, "learning_rate": 7.798251152078127e-05, "loss": 0.0086, "step": 31330 }, { "grad_norm": 0.08187156915664673, "learning_rate": 7.796740255393429e-05, "loss": 0.008, "step": 31340 }, { "grad_norm": 0.0618634931743145, "learning_rate": 7.795228986959766e-05, "loss": 0.006, "step": 31350 }, { "grad_norm": 0.14339083433151245, "learning_rate": 7.793717346978019e-05, "loss": 0.0096, "step": 31360 }, { "grad_norm": 0.1280158907175064, "learning_rate": 7.792205335649119e-05, "loss": 0.0085, "step": 31370 }, { "grad_norm": 0.09528309106826782, "learning_rate": 7.790692953174043e-05, "loss": 0.007, "step": 31380 }, { "grad_norm": 0.1139478087425232, "learning_rate": 7.789180199753826e-05, "loss": 0.0077, "step": 31390 }, { "grad_norm": 0.09247075766324997, "learning_rate": 7.787667075589542e-05, "loss": 0.0087, "step": 31400 }, { "grad_norm": 0.06895725429058075, "learning_rate": 7.78615358088232e-05, "loss": 0.0134, "step": 31410 }, { "grad_norm": 0.08998006582260132, "learning_rate": 7.784639715833339e-05, "loss": 0.008, "step": 31420 }, { "grad_norm": 0.033888235688209534, "learning_rate": 7.783125480643823e-05, "loss": 0.008, "step": 31430 }, { "grad_norm": 0.04850073903799057, "learning_rate": 7.781610875515046e-05, "loss": 0.0081, "step": 31440 }, { "grad_norm": 0.042684491723775864, "learning_rate": 7.780095900648336e-05, "loss": 0.0097, "step": 31450 }, { "grad_norm": 0.06538329273462296, "learning_rate": 7.778580556245065e-05, "loss": 0.0074, "step": 31460 }, { "grad_norm": 0.06351519376039505, "learning_rate": 7.777064842506658e-05, "loss": 0.0108, "step": 31470 }, { "grad_norm": 0.0865776464343071, "learning_rate": 7.775548759634587e-05, "loss": 0.0068, "step": 31480 }, { "grad_norm": 0.12053528428077698, "learning_rate": 7.774032307830369e-05, "loss": 0.0107, "step": 31490 }, { "grad_norm": 0.10874726623296738, "learning_rate": 7.772515487295577e-05, "loss": 0.0096, "step": 31500 }, { "grad_norm": 0.11479905247688293, "learning_rate": 7.770998298231833e-05, "loss": 0.0097, "step": 31510 }, { "grad_norm": 0.10958413034677505, "learning_rate": 7.7694807408408e-05, "loss": 0.0079, "step": 31520 }, { "grad_norm": 0.10802620649337769, "learning_rate": 7.767962815324197e-05, "loss": 0.0103, "step": 31530 }, { "grad_norm": 0.1151379644870758, "learning_rate": 7.766444521883792e-05, "loss": 0.0079, "step": 31540 }, { "grad_norm": 0.07824462652206421, "learning_rate": 7.764925860721399e-05, "loss": 0.0093, "step": 31550 }, { "grad_norm": 0.07428024709224701, "learning_rate": 7.763406832038878e-05, "loss": 0.0107, "step": 31560 }, { "grad_norm": 0.1137048602104187, "learning_rate": 7.761887436038149e-05, "loss": 0.0123, "step": 31570 }, { "grad_norm": 0.08232323825359344, "learning_rate": 7.760367672921167e-05, "loss": 0.0085, "step": 31580 }, { "grad_norm": 0.06725896149873734, "learning_rate": 7.758847542889946e-05, "loss": 0.0075, "step": 31590 }, { "grad_norm": 0.0906829684972763, "learning_rate": 7.757327046146544e-05, "loss": 0.0152, "step": 31600 }, { "grad_norm": 0.15179084241390228, "learning_rate": 7.755806182893067e-05, "loss": 0.011, "step": 31610 }, { "grad_norm": 0.0702844113111496, "learning_rate": 7.754284953331674e-05, "loss": 0.0074, "step": 31620 }, { "grad_norm": 0.08879297971725464, "learning_rate": 7.75276335766457e-05, "loss": 0.0079, "step": 31630 }, { "grad_norm": 0.12859106063842773, "learning_rate": 7.751241396094007e-05, "loss": 0.0116, "step": 31640 }, { "grad_norm": 0.10136640816926956, "learning_rate": 7.74971906882229e-05, "loss": 0.0087, "step": 31650 }, { "grad_norm": 0.08024616539478302, "learning_rate": 7.748196376051769e-05, "loss": 0.0115, "step": 31660 }, { "grad_norm": 0.10032397508621216, "learning_rate": 7.746673317984844e-05, "loss": 0.01, "step": 31670 }, { "grad_norm": 0.10464852303266525, "learning_rate": 7.745149894823963e-05, "loss": 0.0067, "step": 31680 }, { "grad_norm": 0.08961798995733261, "learning_rate": 7.743626106771621e-05, "loss": 0.0112, "step": 31690 }, { "grad_norm": 0.09970398992300034, "learning_rate": 7.742101954030369e-05, "loss": 0.0091, "step": 31700 }, { "grad_norm": 0.08967697620391846, "learning_rate": 7.740577436802793e-05, "loss": 0.0114, "step": 31710 }, { "grad_norm": 0.11519458144903183, "learning_rate": 7.739052555291541e-05, "loss": 0.011, "step": 31720 }, { "grad_norm": 0.08310215920209885, "learning_rate": 7.737527309699304e-05, "loss": 0.009, "step": 31730 }, { "grad_norm": 0.09564045071601868, "learning_rate": 7.736001700228816e-05, "loss": 0.0093, "step": 31740 }, { "grad_norm": 0.061292897909879684, "learning_rate": 7.734475727082868e-05, "loss": 0.0091, "step": 31750 }, { "grad_norm": 0.07210659980773926, "learning_rate": 7.732949390464296e-05, "loss": 0.0109, "step": 31760 }, { "grad_norm": 0.0651240199804306, "learning_rate": 7.731422690575982e-05, "loss": 0.0098, "step": 31770 }, { "grad_norm": 0.0746467337012291, "learning_rate": 7.72989562762086e-05, "loss": 0.0118, "step": 31780 }, { "grad_norm": 0.08546622842550278, "learning_rate": 7.72836820180191e-05, "loss": 0.0082, "step": 31790 }, { "grad_norm": 0.13091251254081726, "learning_rate": 7.72684041332216e-05, "loss": 0.0103, "step": 31800 }, { "grad_norm": 0.05618453025817871, "learning_rate": 7.725312262384691e-05, "loss": 0.0072, "step": 31810 }, { "grad_norm": 0.10700783133506775, "learning_rate": 7.723783749192624e-05, "loss": 0.0108, "step": 31820 }, { "grad_norm": 0.11517587304115295, "learning_rate": 7.722254873949133e-05, "loss": 0.009, "step": 31830 }, { "grad_norm": 0.08954855054616928, "learning_rate": 7.72072563685744e-05, "loss": 0.009, "step": 31840 }, { "grad_norm": 0.09240727871656418, "learning_rate": 7.719196038120815e-05, "loss": 0.0095, "step": 31850 }, { "grad_norm": 0.09488608688116074, "learning_rate": 7.717666077942576e-05, "loss": 0.009, "step": 31860 }, { "grad_norm": 0.06617506593465805, "learning_rate": 7.716135756526088e-05, "loss": 0.0076, "step": 31870 }, { "grad_norm": 0.14973150193691254, "learning_rate": 7.714605074074766e-05, "loss": 0.0082, "step": 31880 }, { "grad_norm": 0.09169896692037582, "learning_rate": 7.71307403079207e-05, "loss": 0.0081, "step": 31890 }, { "grad_norm": 0.10587135702371597, "learning_rate": 7.711542626881508e-05, "loss": 0.0115, "step": 31900 }, { "grad_norm": 0.10550928860902786, "learning_rate": 7.710010862546641e-05, "loss": 0.0072, "step": 31910 }, { "grad_norm": 0.09428669512271881, "learning_rate": 7.708478737991077e-05, "loss": 0.0078, "step": 31920 }, { "grad_norm": 0.0622686967253685, "learning_rate": 7.706946253418462e-05, "loss": 0.0075, "step": 31930 }, { "grad_norm": 0.07771027088165283, "learning_rate": 7.705413409032504e-05, "loss": 0.0118, "step": 31940 }, { "grad_norm": 0.06516958028078079, "learning_rate": 7.703880205036948e-05, "loss": 0.0075, "step": 31950 }, { "grad_norm": 0.12775850296020508, "learning_rate": 7.702346641635593e-05, "loss": 0.0142, "step": 31960 }, { "grad_norm": 0.11791341006755829, "learning_rate": 7.700812719032283e-05, "loss": 0.0068, "step": 31970 }, { "grad_norm": 0.07802922278642654, "learning_rate": 7.699278437430909e-05, "loss": 0.011, "step": 31980 }, { "grad_norm": 0.06521362066268921, "learning_rate": 7.697743797035413e-05, "loss": 0.0076, "step": 31990 }, { "grad_norm": 0.10117983818054199, "learning_rate": 7.696208798049783e-05, "loss": 0.0107, "step": 32000 }, { "grad_norm": 0.09094954282045364, "learning_rate": 7.694673440678052e-05, "loss": 0.0073, "step": 32010 }, { "grad_norm": 0.08278235048055649, "learning_rate": 7.693137725124309e-05, "loss": 0.0092, "step": 32020 }, { "grad_norm": 0.10010629892349243, "learning_rate": 7.691601651592678e-05, "loss": 0.0109, "step": 32030 }, { "grad_norm": 0.06987957656383514, "learning_rate": 7.690065220287339e-05, "loss": 0.008, "step": 32040 }, { "grad_norm": 0.06714632362127304, "learning_rate": 7.68852843141252e-05, "loss": 0.0071, "step": 32050 }, { "grad_norm": 0.10972794145345688, "learning_rate": 7.686991285172494e-05, "loss": 0.0102, "step": 32060 }, { "grad_norm": 0.10789260268211365, "learning_rate": 7.68545378177158e-05, "loss": 0.0074, "step": 32070 }, { "grad_norm": 0.056521132588386536, "learning_rate": 7.683915921414149e-05, "loss": 0.0115, "step": 32080 }, { "grad_norm": 0.1105593591928482, "learning_rate": 7.682377704304614e-05, "loss": 0.0094, "step": 32090 }, { "grad_norm": 0.08280995488166809, "learning_rate": 7.680839130647442e-05, "loss": 0.0083, "step": 32100 }, { "grad_norm": 0.0910603478550911, "learning_rate": 7.679300200647139e-05, "loss": 0.009, "step": 32110 }, { "grad_norm": 0.12845319509506226, "learning_rate": 7.677760914508266e-05, "loss": 0.0108, "step": 32120 }, { "grad_norm": 0.06070461869239807, "learning_rate": 7.67622127243543e-05, "loss": 0.0069, "step": 32130 }, { "grad_norm": 0.10788394510746002, "learning_rate": 7.674681274633277e-05, "loss": 0.0085, "step": 32140 }, { "grad_norm": 0.07590651512145996, "learning_rate": 7.673140921306515e-05, "loss": 0.008, "step": 32150 }, { "grad_norm": 0.10666953027248383, "learning_rate": 7.671600212659885e-05, "loss": 0.0123, "step": 32160 }, { "grad_norm": 0.08281240612268448, "learning_rate": 7.670059148898185e-05, "loss": 0.0137, "step": 32170 }, { "grad_norm": 0.08348134905099869, "learning_rate": 7.668517730226257e-05, "loss": 0.0105, "step": 32180 }, { "grad_norm": 0.08889737725257874, "learning_rate": 7.666975956848987e-05, "loss": 0.0071, "step": 32190 }, { "grad_norm": 0.0667438879609108, "learning_rate": 7.665433828971312e-05, "loss": 0.0063, "step": 32200 }, { "grad_norm": 0.10295794159173965, "learning_rate": 7.663891346798217e-05, "loss": 0.0079, "step": 32210 }, { "grad_norm": 0.09012711048126221, "learning_rate": 7.66234851053473e-05, "loss": 0.0087, "step": 32220 }, { "grad_norm": 0.14047551155090332, "learning_rate": 7.660805320385928e-05, "loss": 0.0085, "step": 32230 }, { "grad_norm": 0.08641752600669861, "learning_rate": 7.659261776556939e-05, "loss": 0.009, "step": 32240 }, { "grad_norm": 0.0997653678059578, "learning_rate": 7.657717879252928e-05, "loss": 0.0106, "step": 32250 }, { "grad_norm": 0.09428686648607254, "learning_rate": 7.656173628679118e-05, "loss": 0.0076, "step": 32260 }, { "grad_norm": 0.08691470324993134, "learning_rate": 7.654629025040774e-05, "loss": 0.0084, "step": 32270 }, { "grad_norm": 0.10315432399511337, "learning_rate": 7.653084068543206e-05, "loss": 0.0093, "step": 32280 }, { "grad_norm": 0.0797269269824028, "learning_rate": 7.651538759391776e-05, "loss": 0.0102, "step": 32290 }, { "grad_norm": 0.07248938828706741, "learning_rate": 7.649993097791887e-05, "loss": 0.0076, "step": 32300 }, { "grad_norm": 0.06679555773735046, "learning_rate": 7.648447083948992e-05, "loss": 0.01, "step": 32310 }, { "grad_norm": 0.10776881128549576, "learning_rate": 7.646900718068594e-05, "loss": 0.0082, "step": 32320 }, { "grad_norm": 0.06671225279569626, "learning_rate": 7.645354000356235e-05, "loss": 0.0109, "step": 32330 }, { "grad_norm": 0.08972541987895966, "learning_rate": 7.64380693101751e-05, "loss": 0.0098, "step": 32340 }, { "grad_norm": 0.07420498877763748, "learning_rate": 7.642259510258061e-05, "loss": 0.0082, "step": 32350 }, { "grad_norm": 0.05759917572140694, "learning_rate": 7.640711738283571e-05, "loss": 0.0105, "step": 32360 }, { "grad_norm": 0.11226546764373779, "learning_rate": 7.639163615299776e-05, "loss": 0.0124, "step": 32370 }, { "grad_norm": 0.11894556879997253, "learning_rate": 7.637615141512454e-05, "loss": 0.009, "step": 32380 }, { "grad_norm": 0.09252656996250153, "learning_rate": 7.636066317127433e-05, "loss": 0.0116, "step": 32390 }, { "grad_norm": 0.09615025669336319, "learning_rate": 7.634517142350589e-05, "loss": 0.0127, "step": 32400 }, { "grad_norm": 0.09615232050418854, "learning_rate": 7.632967617387834e-05, "loss": 0.0085, "step": 32410 }, { "grad_norm": 0.06367949396371841, "learning_rate": 7.63141774244514e-05, "loss": 0.0086, "step": 32420 }, { "grad_norm": 0.05319445952773094, "learning_rate": 7.62986751772852e-05, "loss": 0.009, "step": 32430 }, { "grad_norm": 0.15612557530403137, "learning_rate": 7.628316943444032e-05, "loss": 0.0099, "step": 32440 }, { "grad_norm": 0.10429955273866653, "learning_rate": 7.62676601979778e-05, "loss": 0.0078, "step": 32450 }, { "grad_norm": 0.10914868861436844, "learning_rate": 7.625214746995919e-05, "loss": 0.0085, "step": 32460 }, { "grad_norm": 0.11301961541175842, "learning_rate": 7.623663125244648e-05, "loss": 0.0088, "step": 32470 }, { "grad_norm": 0.07902410626411438, "learning_rate": 7.622111154750208e-05, "loss": 0.0053, "step": 32480 }, { "grad_norm": 0.1247534453868866, "learning_rate": 7.620558835718896e-05, "loss": 0.011, "step": 32490 }, { "grad_norm": 0.08731172233819962, "learning_rate": 7.619006168357046e-05, "loss": 0.0099, "step": 32500 }, { "grad_norm": 0.087250255048275, "learning_rate": 7.617453152871043e-05, "loss": 0.0112, "step": 32510 }, { "grad_norm": 0.08694043755531311, "learning_rate": 7.615899789467318e-05, "loss": 0.0095, "step": 32520 }, { "grad_norm": 0.08442787081003189, "learning_rate": 7.614346078352346e-05, "loss": 0.0105, "step": 32530 }, { "grad_norm": 0.0969095528125763, "learning_rate": 7.612792019732651e-05, "loss": 0.0112, "step": 32540 }, { "grad_norm": 0.07184348255395889, "learning_rate": 7.6112376138148e-05, "loss": 0.0081, "step": 32550 }, { "grad_norm": 0.08143015950918198, "learning_rate": 7.609682860805409e-05, "loss": 0.0093, "step": 32560 }, { "grad_norm": 0.10636290907859802, "learning_rate": 7.608127760911142e-05, "loss": 0.0111, "step": 32570 }, { "grad_norm": 0.15291999280452728, "learning_rate": 7.606572314338702e-05, "loss": 0.0148, "step": 32580 }, { "grad_norm": 0.09973360598087311, "learning_rate": 7.605016521294847e-05, "loss": 0.0101, "step": 32590 }, { "grad_norm": 0.1261577606201172, "learning_rate": 7.603460381986372e-05, "loss": 0.0126, "step": 32600 }, { "grad_norm": 0.070195771753788, "learning_rate": 7.601903896620126e-05, "loss": 0.0078, "step": 32610 }, { "grad_norm": 0.12804612517356873, "learning_rate": 7.600347065403e-05, "loss": 0.0099, "step": 32620 }, { "grad_norm": 0.10904739797115326, "learning_rate": 7.598789888541928e-05, "loss": 0.0095, "step": 32630 }, { "grad_norm": 0.15115654468536377, "learning_rate": 7.597232366243896e-05, "loss": 0.0101, "step": 32640 }, { "grad_norm": 0.11216413974761963, "learning_rate": 7.595674498715935e-05, "loss": 0.0122, "step": 32650 }, { "grad_norm": 0.10061585903167725, "learning_rate": 7.594116286165118e-05, "loss": 0.0082, "step": 32660 }, { "grad_norm": 0.13598735630512238, "learning_rate": 7.592557728798566e-05, "loss": 0.0107, "step": 32670 }, { "grad_norm": 0.1009397953748703, "learning_rate": 7.590998826823448e-05, "loss": 0.0087, "step": 32680 }, { "grad_norm": 0.064400315284729, "learning_rate": 7.589439580446975e-05, "loss": 0.007, "step": 32690 }, { "grad_norm": 0.10354627668857574, "learning_rate": 7.587879989876405e-05, "loss": 0.0082, "step": 32700 }, { "grad_norm": 0.08428608626127243, "learning_rate": 7.586320055319043e-05, "loss": 0.0105, "step": 32710 }, { "grad_norm": 0.13385802507400513, "learning_rate": 7.58475977698224e-05, "loss": 0.0094, "step": 32720 }, { "grad_norm": 0.08575288951396942, "learning_rate": 7.583199155073391e-05, "loss": 0.0058, "step": 32730 }, { "grad_norm": 0.06899704784154892, "learning_rate": 7.581638189799937e-05, "loss": 0.0092, "step": 32740 }, { "grad_norm": 0.10265452414751053, "learning_rate": 7.580076881369363e-05, "loss": 0.0082, "step": 32750 }, { "grad_norm": 0.08469172567129135, "learning_rate": 7.578515229989207e-05, "loss": 0.0081, "step": 32760 }, { "grad_norm": 0.11055074632167816, "learning_rate": 7.576953235867041e-05, "loss": 0.0112, "step": 32770 }, { "grad_norm": 0.09910941123962402, "learning_rate": 7.575390899210494e-05, "loss": 0.01, "step": 32780 }, { "grad_norm": 0.08629605174064636, "learning_rate": 7.573828220227233e-05, "loss": 0.0064, "step": 32790 }, { "grad_norm": 0.055529750883579254, "learning_rate": 7.57226519912497e-05, "loss": 0.009, "step": 32800 }, { "grad_norm": 0.06051813066005707, "learning_rate": 7.570701836111471e-05, "loss": 0.0095, "step": 32810 }, { "grad_norm": 0.1058911681175232, "learning_rate": 7.569138131394538e-05, "loss": 0.0072, "step": 32820 }, { "grad_norm": 0.08268065750598907, "learning_rate": 7.567574085182021e-05, "loss": 0.008, "step": 32830 }, { "grad_norm": 0.1171107292175293, "learning_rate": 7.56600969768182e-05, "loss": 0.0106, "step": 32840 }, { "grad_norm": 0.08653327822685242, "learning_rate": 7.564444969101875e-05, "loss": 0.0137, "step": 32850 }, { "grad_norm": 0.10133235156536102, "learning_rate": 7.562879899650172e-05, "loss": 0.008, "step": 32860 }, { "grad_norm": 0.0938936322927475, "learning_rate": 7.561314489534748e-05, "loss": 0.0084, "step": 32870 }, { "grad_norm": 0.0661231204867363, "learning_rate": 7.559748738963675e-05, "loss": 0.0086, "step": 32880 }, { "grad_norm": 0.07081689685583115, "learning_rate": 7.55818264814508e-05, "loss": 0.0082, "step": 32890 }, { "grad_norm": 0.04920732229948044, "learning_rate": 7.55661621728713e-05, "loss": 0.0109, "step": 32900 }, { "grad_norm": 0.0779019445180893, "learning_rate": 7.555049446598039e-05, "loss": 0.0082, "step": 32910 }, { "grad_norm": 0.10376658290624619, "learning_rate": 7.553482336286066e-05, "loss": 0.0071, "step": 32920 }, { "grad_norm": 0.06243118643760681, "learning_rate": 7.551914886559512e-05, "loss": 0.0072, "step": 32930 }, { "grad_norm": 0.0822802186012268, "learning_rate": 7.550347097626729e-05, "loss": 0.0091, "step": 32940 }, { "grad_norm": 0.09842346608638763, "learning_rate": 7.548778969696111e-05, "loss": 0.0107, "step": 32950 }, { "grad_norm": 0.07124418765306473, "learning_rate": 7.547210502976094e-05, "loss": 0.0104, "step": 32960 }, { "grad_norm": 0.10133186727762222, "learning_rate": 7.545641697675166e-05, "loss": 0.0104, "step": 32970 }, { "grad_norm": 0.09596298635005951, "learning_rate": 7.544072554001854e-05, "loss": 0.0116, "step": 32980 }, { "grad_norm": 0.08658809214830399, "learning_rate": 7.542503072164733e-05, "loss": 0.0063, "step": 32990 }, { "grad_norm": 0.06921810656785965, "learning_rate": 7.54093325237242e-05, "loss": 0.0067, "step": 33000 }, { "grad_norm": 0.04785818234086037, "learning_rate": 7.53936309483358e-05, "loss": 0.0082, "step": 33010 }, { "grad_norm": 0.07413991540670395, "learning_rate": 7.537792599756923e-05, "loss": 0.0078, "step": 33020 }, { "grad_norm": 0.06760519742965698, "learning_rate": 7.536221767351204e-05, "loss": 0.0082, "step": 33030 }, { "grad_norm": 0.07362443953752518, "learning_rate": 7.534650597825217e-05, "loss": 0.0068, "step": 33040 }, { "grad_norm": 0.09971409291028976, "learning_rate": 7.533079091387808e-05, "loss": 0.0086, "step": 33050 }, { "grad_norm": 0.1152743548154831, "learning_rate": 7.531507248247867e-05, "loss": 0.0086, "step": 33060 }, { "grad_norm": 0.09328044950962067, "learning_rate": 7.529935068614322e-05, "loss": 0.0079, "step": 33070 }, { "grad_norm": 0.10695822536945343, "learning_rate": 7.528362552696156e-05, "loss": 0.0068, "step": 33080 }, { "grad_norm": 0.08360933512449265, "learning_rate": 7.526789700702389e-05, "loss": 0.0067, "step": 33090 }, { "grad_norm": 0.06228380650281906, "learning_rate": 7.525216512842084e-05, "loss": 0.0084, "step": 33100 }, { "grad_norm": 0.07122138887643814, "learning_rate": 7.52364298932436e-05, "loss": 0.01, "step": 33110 }, { "grad_norm": 0.06351695954799652, "learning_rate": 7.522069130358369e-05, "loss": 0.006, "step": 33120 }, { "grad_norm": 0.08725892752408981, "learning_rate": 7.520494936153313e-05, "loss": 0.0064, "step": 33130 }, { "grad_norm": 0.07890800386667252, "learning_rate": 7.518920406918438e-05, "loss": 0.009, "step": 33140 }, { "grad_norm": 0.0809209793806076, "learning_rate": 7.517345542863031e-05, "loss": 0.0124, "step": 33150 }, { "grad_norm": 0.0753326490521431, "learning_rate": 7.515770344196429e-05, "loss": 0.0072, "step": 33160 }, { "grad_norm": 0.06279833614826202, "learning_rate": 7.51419481112801e-05, "loss": 0.0091, "step": 33170 }, { "grad_norm": 0.05581223592162132, "learning_rate": 7.5126189438672e-05, "loss": 0.009, "step": 33180 }, { "grad_norm": 0.12905308604240417, "learning_rate": 7.511042742623461e-05, "loss": 0.0119, "step": 33190 }, { "grad_norm": 0.0904090479016304, "learning_rate": 7.509466207606312e-05, "loss": 0.0106, "step": 33200 }, { "grad_norm": 0.12955406308174133, "learning_rate": 7.507889339025304e-05, "loss": 0.0106, "step": 33210 }, { "grad_norm": 0.09812350571155548, "learning_rate": 7.506312137090041e-05, "loss": 0.0093, "step": 33220 }, { "grad_norm": 0.0778229832649231, "learning_rate": 7.504734602010168e-05, "loss": 0.0086, "step": 33230 }, { "grad_norm": 0.07375641912221909, "learning_rate": 7.503156733995374e-05, "loss": 0.0085, "step": 33240 }, { "grad_norm": 0.09496892243623734, "learning_rate": 7.501578533255392e-05, "loss": 0.0085, "step": 33250 }, { "grad_norm": 0.09643231332302094, "learning_rate": 7.500000000000001e-05, "loss": 0.0084, "step": 33260 }, { "grad_norm": 0.10777045041322708, "learning_rate": 7.498421134439022e-05, "loss": 0.0116, "step": 33270 }, { "grad_norm": 0.09930646419525146, "learning_rate": 7.496841936782323e-05, "loss": 0.0098, "step": 33280 }, { "grad_norm": 0.05665484070777893, "learning_rate": 7.495262407239813e-05, "loss": 0.0067, "step": 33290 }, { "grad_norm": 0.10148321837186813, "learning_rate": 7.493682546021447e-05, "loss": 0.0101, "step": 33300 }, { "grad_norm": 0.07249783724546432, "learning_rate": 7.492102353337224e-05, "loss": 0.0079, "step": 33310 }, { "grad_norm": 0.0646917000412941, "learning_rate": 7.490521829397186e-05, "loss": 0.0097, "step": 33320 }, { "grad_norm": 0.13167692720890045, "learning_rate": 7.488940974411423e-05, "loss": 0.0119, "step": 33330 }, { "grad_norm": 0.11656375974416733, "learning_rate": 7.487359788590062e-05, "loss": 0.0097, "step": 33340 }, { "grad_norm": 0.08895544707775116, "learning_rate": 7.485778272143277e-05, "loss": 0.008, "step": 33350 }, { "grad_norm": 0.0729711651802063, "learning_rate": 7.48419642528129e-05, "loss": 0.0096, "step": 33360 }, { "grad_norm": 0.11391673982143402, "learning_rate": 7.482614248214363e-05, "loss": 0.0188, "step": 33370 }, { "grad_norm": 0.046564310789108276, "learning_rate": 7.4810317411528e-05, "loss": 0.0104, "step": 33380 }, { "grad_norm": 0.11294296383857727, "learning_rate": 7.479448904306954e-05, "loss": 0.0099, "step": 33390 }, { "grad_norm": 0.09515587240457535, "learning_rate": 7.477865737887218e-05, "loss": 0.0102, "step": 33400 }, { "grad_norm": 0.09599169343709946, "learning_rate": 7.476282242104031e-05, "loss": 0.0079, "step": 33410 }, { "grad_norm": 0.0802430808544159, "learning_rate": 7.474698417167873e-05, "loss": 0.0096, "step": 33420 }, { "grad_norm": 0.08980825543403625, "learning_rate": 7.473114263289272e-05, "loss": 0.0107, "step": 33430 }, { "grad_norm": 0.09550303965806961, "learning_rate": 7.471529780678797e-05, "loss": 0.0083, "step": 33440 }, { "grad_norm": 0.04190600663423538, "learning_rate": 7.469944969547058e-05, "loss": 0.006, "step": 33450 }, { "grad_norm": 0.11124730110168457, "learning_rate": 7.468359830104714e-05, "loss": 0.0086, "step": 33460 }, { "grad_norm": 0.0912514179944992, "learning_rate": 7.466774362562465e-05, "loss": 0.0074, "step": 33470 }, { "grad_norm": 0.09451259672641754, "learning_rate": 7.465188567131056e-05, "loss": 0.0084, "step": 33480 }, { "grad_norm": 0.08291326463222504, "learning_rate": 7.46360244402127e-05, "loss": 0.0092, "step": 33490 }, { "grad_norm": 0.08736609667539597, "learning_rate": 7.462015993443943e-05, "loss": 0.0119, "step": 33500 }, { "grad_norm": 0.0849887952208519, "learning_rate": 7.460429215609949e-05, "loss": 0.0095, "step": 33510 }, { "grad_norm": 0.13950391113758087, "learning_rate": 7.458842110730204e-05, "loss": 0.0074, "step": 33520 }, { "grad_norm": 0.11324569582939148, "learning_rate": 7.45725467901567e-05, "loss": 0.0103, "step": 33530 }, { "grad_norm": 0.0910949632525444, "learning_rate": 7.455666920677353e-05, "loss": 0.0096, "step": 33540 }, { "grad_norm": 0.060086917132139206, "learning_rate": 7.4540788359263e-05, "loss": 0.0104, "step": 33550 }, { "grad_norm": 0.0639014020562172, "learning_rate": 7.452490424973604e-05, "loss": 0.0099, "step": 33560 }, { "grad_norm": 0.09644422680139542, "learning_rate": 7.450901688030399e-05, "loss": 0.0078, "step": 33570 }, { "grad_norm": 0.06662028282880783, "learning_rate": 7.449312625307865e-05, "loss": 0.011, "step": 33580 }, { "grad_norm": 0.10902249068021774, "learning_rate": 7.447723237017223e-05, "loss": 0.0097, "step": 33590 }, { "grad_norm": 0.10157003253698349, "learning_rate": 7.446133523369737e-05, "loss": 0.0111, "step": 33600 }, { "grad_norm": 0.12320840358734131, "learning_rate": 7.444543484576717e-05, "loss": 0.0092, "step": 33610 }, { "grad_norm": 0.08077140152454376, "learning_rate": 7.442953120849513e-05, "loss": 0.0125, "step": 33620 }, { "grad_norm": 0.05953584238886833, "learning_rate": 7.44136243239952e-05, "loss": 0.0092, "step": 33630 }, { "grad_norm": 0.07813020050525665, "learning_rate": 7.439771419438179e-05, "loss": 0.0112, "step": 33640 }, { "grad_norm": 0.10817597806453705, "learning_rate": 7.438180082176966e-05, "loss": 0.0116, "step": 33650 }, { "grad_norm": 0.1105927899479866, "learning_rate": 7.43658842082741e-05, "loss": 0.0089, "step": 33660 }, { "grad_norm": 0.07868307083845139, "learning_rate": 7.434996435601073e-05, "loss": 0.009, "step": 33670 }, { "grad_norm": 0.07724133133888245, "learning_rate": 7.433404126709567e-05, "loss": 0.0075, "step": 33680 }, { "grad_norm": 0.07306515425443649, "learning_rate": 7.431811494364549e-05, "loss": 0.0093, "step": 33690 }, { "grad_norm": 0.06760270148515701, "learning_rate": 7.43021853877771e-05, "loss": 0.0086, "step": 33700 }, { "grad_norm": 0.08952978253364563, "learning_rate": 7.428625260160792e-05, "loss": 0.0076, "step": 33710 }, { "grad_norm": 0.07709483802318573, "learning_rate": 7.427031658725578e-05, "loss": 0.0081, "step": 33720 }, { "grad_norm": 0.08452904224395752, "learning_rate": 7.42543773468389e-05, "loss": 0.0071, "step": 33730 }, { "grad_norm": 0.081514872610569, "learning_rate": 7.423843488247598e-05, "loss": 0.0113, "step": 33740 }, { "grad_norm": 0.11512254178524017, "learning_rate": 7.422248919628614e-05, "loss": 0.007, "step": 33750 }, { "grad_norm": 0.09950600564479828, "learning_rate": 7.42065402903889e-05, "loss": 0.0093, "step": 33760 }, { "grad_norm": 0.10626505315303802, "learning_rate": 7.419058816690421e-05, "loss": 0.0089, "step": 33770 }, { "grad_norm": 0.07349022477865219, "learning_rate": 7.417463282795249e-05, "loss": 0.0089, "step": 33780 }, { "grad_norm": 0.08558528125286102, "learning_rate": 7.415867427565452e-05, "loss": 0.0071, "step": 33790 }, { "grad_norm": 0.09533628821372986, "learning_rate": 7.414271251213161e-05, "loss": 0.0144, "step": 33800 }, { "grad_norm": 0.08683335781097412, "learning_rate": 7.412674753950538e-05, "loss": 0.0069, "step": 33810 }, { "grad_norm": 0.06858877092599869, "learning_rate": 7.411077935989795e-05, "loss": 0.0091, "step": 33820 }, { "grad_norm": 0.09564042836427689, "learning_rate": 7.409480797543184e-05, "loss": 0.0071, "step": 33830 }, { "grad_norm": 0.08753766119480133, "learning_rate": 7.407883338823001e-05, "loss": 0.0086, "step": 33840 }, { "grad_norm": 0.0675082728266716, "learning_rate": 7.406285560041583e-05, "loss": 0.0089, "step": 33850 }, { "grad_norm": 0.07014720886945724, "learning_rate": 7.404687461411311e-05, "loss": 0.0076, "step": 33860 }, { "grad_norm": 0.04747544974088669, "learning_rate": 7.403089043144606e-05, "loss": 0.0073, "step": 33870 }, { "grad_norm": 0.09545435756444931, "learning_rate": 7.401490305453935e-05, "loss": 0.0078, "step": 33880 }, { "grad_norm": 0.11682278662919998, "learning_rate": 7.399891248551807e-05, "loss": 0.0079, "step": 33890 }, { "grad_norm": 0.0777849480509758, "learning_rate": 7.398291872650768e-05, "loss": 0.0098, "step": 33900 }, { "grad_norm": 0.06890033185482025, "learning_rate": 7.396692177963415e-05, "loss": 0.008, "step": 33910 }, { "grad_norm": 0.0587676465511322, "learning_rate": 7.395092164702382e-05, "loss": 0.0077, "step": 33920 }, { "grad_norm": 0.05527477711439133, "learning_rate": 7.393491833080346e-05, "loss": 0.0059, "step": 33930 }, { "grad_norm": 0.08314820379018784, "learning_rate": 7.391891183310024e-05, "loss": 0.0064, "step": 33940 }, { "grad_norm": 0.0649200975894928, "learning_rate": 7.390290215604183e-05, "loss": 0.0112, "step": 33950 }, { "grad_norm": 0.08432219922542572, "learning_rate": 7.388688930175625e-05, "loss": 0.0075, "step": 33960 }, { "grad_norm": 0.10906276106834412, "learning_rate": 7.387087327237193e-05, "loss": 0.0138, "step": 33970 }, { "grad_norm": 0.10456465184688568, "learning_rate": 7.385485407001778e-05, "loss": 0.0093, "step": 33980 }, { "grad_norm": 0.09371547400951385, "learning_rate": 7.383883169682315e-05, "loss": 0.0071, "step": 33990 }, { "grad_norm": 0.11655107140541077, "learning_rate": 7.382280615491772e-05, "loss": 0.0128, "step": 34000 }, { "grad_norm": 0.07721009105443954, "learning_rate": 7.380677744643165e-05, "loss": 0.0083, "step": 34010 }, { "grad_norm": 0.07493811845779419, "learning_rate": 7.379074557349552e-05, "loss": 0.0112, "step": 34020 }, { "grad_norm": 0.07245147973299026, "learning_rate": 7.377471053824031e-05, "loss": 0.0112, "step": 34030 }, { "grad_norm": 0.09843674302101135, "learning_rate": 7.375867234279744e-05, "loss": 0.0084, "step": 34040 }, { "grad_norm": 0.08828306198120117, "learning_rate": 7.374263098929876e-05, "loss": 0.0074, "step": 34050 }, { "grad_norm": 0.08859068155288696, "learning_rate": 7.37265864798765e-05, "loss": 0.009, "step": 34060 }, { "grad_norm": 0.12974466383457184, "learning_rate": 7.371053881666334e-05, "loss": 0.0111, "step": 34070 }, { "grad_norm": 0.09164510667324066, "learning_rate": 7.369448800179237e-05, "loss": 0.0076, "step": 34080 }, { "grad_norm": 0.08727714419364929, "learning_rate": 7.367843403739707e-05, "loss": 0.0081, "step": 34090 }, { "grad_norm": 0.036506734788417816, "learning_rate": 7.366237692561144e-05, "loss": 0.0069, "step": 34100 }, { "grad_norm": 0.09030873328447342, "learning_rate": 7.364631666856978e-05, "loss": 0.0106, "step": 34110 }, { "grad_norm": 0.04086552560329437, "learning_rate": 7.363025326840684e-05, "loss": 0.0111, "step": 34120 }, { "grad_norm": 0.10912550985813141, "learning_rate": 7.361418672725783e-05, "loss": 0.0066, "step": 34130 }, { "grad_norm": 0.05416533350944519, "learning_rate": 7.359811704725837e-05, "loss": 0.0075, "step": 34140 }, { "grad_norm": 0.05834665894508362, "learning_rate": 7.358204423054443e-05, "loss": 0.0107, "step": 34150 }, { "grad_norm": 0.06761221587657928, "learning_rate": 7.356596827925248e-05, "loss": 0.0082, "step": 34160 }, { "grad_norm": 0.09476009011268616, "learning_rate": 7.354988919551934e-05, "loss": 0.0097, "step": 34170 }, { "grad_norm": 0.07958562672138214, "learning_rate": 7.353380698148232e-05, "loss": 0.0083, "step": 34180 }, { "grad_norm": 0.060808468610048294, "learning_rate": 7.351772163927908e-05, "loss": 0.0085, "step": 34190 }, { "grad_norm": 0.10385594516992569, "learning_rate": 7.35016331710477e-05, "loss": 0.0083, "step": 34200 }, { "grad_norm": 0.06957827508449554, "learning_rate": 7.348554157892672e-05, "loss": 0.0112, "step": 34210 }, { "grad_norm": 0.1070139929652214, "learning_rate": 7.346944686505507e-05, "loss": 0.01, "step": 34220 }, { "grad_norm": 0.08337713778018951, "learning_rate": 7.34533490315721e-05, "loss": 0.0064, "step": 34230 }, { "grad_norm": 0.07640913873910904, "learning_rate": 7.343724808061754e-05, "loss": 0.0086, "step": 34240 }, { "grad_norm": 0.10936295241117477, "learning_rate": 7.34211440143316e-05, "loss": 0.0088, "step": 34250 }, { "grad_norm": 0.05702364817261696, "learning_rate": 7.340503683485486e-05, "loss": 0.0081, "step": 34260 }, { "grad_norm": 0.07766411453485489, "learning_rate": 7.338892654432829e-05, "loss": 0.0113, "step": 34270 }, { "grad_norm": 0.09525519609451294, "learning_rate": 7.337281314489335e-05, "loss": 0.0108, "step": 34280 }, { "grad_norm": 0.09131424129009247, "learning_rate": 7.335669663869186e-05, "loss": 0.0093, "step": 34290 }, { "grad_norm": 0.07746623456478119, "learning_rate": 7.334057702786605e-05, "loss": 0.0068, "step": 34300 }, { "grad_norm": 0.08690491318702698, "learning_rate": 7.332445431455857e-05, "loss": 0.0105, "step": 34310 }, { "grad_norm": 0.06238456815481186, "learning_rate": 7.330832850091249e-05, "loss": 0.0086, "step": 34320 }, { "grad_norm": 0.09741490334272385, "learning_rate": 7.32921995890713e-05, "loss": 0.0083, "step": 34330 }, { "grad_norm": 0.097975954413414, "learning_rate": 7.327606758117889e-05, "loss": 0.0078, "step": 34340 }, { "grad_norm": 0.09109662473201752, "learning_rate": 7.325993247937956e-05, "loss": 0.0106, "step": 34350 }, { "grad_norm": 0.10874511301517487, "learning_rate": 7.324379428581802e-05, "loss": 0.0101, "step": 34360 }, { "grad_norm": 0.07265249639749527, "learning_rate": 7.322765300263938e-05, "loss": 0.0074, "step": 34370 }, { "grad_norm": 0.09098045527935028, "learning_rate": 7.321150863198922e-05, "loss": 0.0101, "step": 34380 }, { "grad_norm": 0.09944642335176468, "learning_rate": 7.319536117601342e-05, "loss": 0.0071, "step": 34390 }, { "grad_norm": 0.10402769595384598, "learning_rate": 7.317921063685839e-05, "loss": 0.01, "step": 34400 }, { "grad_norm": 0.12287586182355881, "learning_rate": 7.31630570166709e-05, "loss": 0.0084, "step": 34410 }, { "grad_norm": 0.07409270852804184, "learning_rate": 7.314690031759808e-05, "loss": 0.008, "step": 34420 }, { "grad_norm": 0.11314738541841507, "learning_rate": 7.313074054178755e-05, "loss": 0.0127, "step": 34430 }, { "grad_norm": 0.1402904987335205, "learning_rate": 7.311457769138729e-05, "loss": 0.0099, "step": 34440 }, { "grad_norm": 0.12011305242776871, "learning_rate": 7.309841176854569e-05, "loss": 0.0087, "step": 34450 }, { "grad_norm": 0.11267036199569702, "learning_rate": 7.308224277541159e-05, "loss": 0.0096, "step": 34460 }, { "grad_norm": 0.10652711242437363, "learning_rate": 7.30660707141342e-05, "loss": 0.0081, "step": 34470 }, { "grad_norm": 0.11821628361940384, "learning_rate": 7.304989558686313e-05, "loss": 0.0082, "step": 34480 }, { "grad_norm": 0.07625678181648254, "learning_rate": 7.303371739574844e-05, "loss": 0.0098, "step": 34490 }, { "grad_norm": 0.07981479167938232, "learning_rate": 7.301753614294052e-05, "loss": 0.0081, "step": 34500 }, { "grad_norm": 0.12748456001281738, "learning_rate": 7.30013518305903e-05, "loss": 0.0085, "step": 34510 }, { "grad_norm": 0.09661752730607986, "learning_rate": 7.298516446084896e-05, "loss": 0.0083, "step": 34520 }, { "grad_norm": 0.07507671415805817, "learning_rate": 7.29689740358682e-05, "loss": 0.0068, "step": 34530 }, { "grad_norm": 0.08053531497716904, "learning_rate": 7.295278055780007e-05, "loss": 0.0066, "step": 34540 }, { "grad_norm": 0.10063629597425461, "learning_rate": 7.293658402879705e-05, "loss": 0.0078, "step": 34550 }, { "grad_norm": 0.07866835594177246, "learning_rate": 7.292038445101204e-05, "loss": 0.0086, "step": 34560 }, { "grad_norm": 0.07033589482307434, "learning_rate": 7.290418182659828e-05, "loss": 0.007, "step": 34570 }, { "grad_norm": 0.054009612649679184, "learning_rate": 7.28879761577095e-05, "loss": 0.0071, "step": 34580 }, { "grad_norm": 0.09156437963247299, "learning_rate": 7.287176744649977e-05, "loss": 0.01, "step": 34590 }, { "grad_norm": 0.0528741292655468, "learning_rate": 7.28555556951236e-05, "loss": 0.0093, "step": 34600 }, { "grad_norm": 0.06756453961133957, "learning_rate": 7.283934090573588e-05, "loss": 0.0076, "step": 34610 }, { "grad_norm": 0.07797611504793167, "learning_rate": 7.282312308049193e-05, "loss": 0.0114, "step": 34620 }, { "grad_norm": 0.04569454863667488, "learning_rate": 7.280690222154745e-05, "loss": 0.0095, "step": 34630 }, { "grad_norm": 0.06031522899866104, "learning_rate": 7.279067833105855e-05, "loss": 0.0077, "step": 34640 }, { "grad_norm": 0.09100725501775742, "learning_rate": 7.277445141118175e-05, "loss": 0.0089, "step": 34650 }, { "grad_norm": 0.10315293073654175, "learning_rate": 7.275822146407398e-05, "loss": 0.0099, "step": 34660 }, { "grad_norm": 0.06827384978532791, "learning_rate": 7.274198849189256e-05, "loss": 0.0061, "step": 34670 }, { "grad_norm": 0.09387808293104172, "learning_rate": 7.272575249679519e-05, "loss": 0.0071, "step": 34680 }, { "grad_norm": 0.09030361473560333, "learning_rate": 7.270951348093999e-05, "loss": 0.0082, "step": 34690 }, { "grad_norm": 0.07704869657754898, "learning_rate": 7.269327144648555e-05, "loss": 0.007, "step": 34700 }, { "grad_norm": 0.08707718551158905, "learning_rate": 7.267702639559072e-05, "loss": 0.0062, "step": 34710 }, { "grad_norm": 0.0764506608247757, "learning_rate": 7.266077833041486e-05, "loss": 0.0085, "step": 34720 }, { "grad_norm": 0.07693370431661606, "learning_rate": 7.264452725311772e-05, "loss": 0.0095, "step": 34730 }, { "grad_norm": 0.07965018600225449, "learning_rate": 7.262827316585938e-05, "loss": 0.0089, "step": 34740 }, { "grad_norm": 0.06334257870912552, "learning_rate": 7.261201607080043e-05, "loss": 0.009, "step": 34750 }, { "grad_norm": 0.09157079458236694, "learning_rate": 7.259575597010175e-05, "loss": 0.0088, "step": 34760 }, { "grad_norm": 0.09193441271781921, "learning_rate": 7.257949286592469e-05, "loss": 0.0067, "step": 34770 }, { "grad_norm": 0.09606565535068512, "learning_rate": 7.256322676043099e-05, "loss": 0.0105, "step": 34780 }, { "grad_norm": 0.09967239201068878, "learning_rate": 7.254695765578274e-05, "loss": 0.01, "step": 34790 }, { "grad_norm": 0.1403583586215973, "learning_rate": 7.253068555414249e-05, "loss": 0.0154, "step": 34800 }, { "grad_norm": 0.10447537153959274, "learning_rate": 7.251441045767317e-05, "loss": 0.0066, "step": 34810 }, { "grad_norm": 0.1027037501335144, "learning_rate": 7.249813236853811e-05, "loss": 0.0097, "step": 34820 }, { "grad_norm": 0.07387927174568176, "learning_rate": 7.2481851288901e-05, "loss": 0.0065, "step": 34830 }, { "grad_norm": 0.09871829301118851, "learning_rate": 7.246556722092597e-05, "loss": 0.0096, "step": 34840 }, { "grad_norm": 0.06548144668340683, "learning_rate": 7.244928016677754e-05, "loss": 0.0092, "step": 34850 }, { "grad_norm": 0.08829405158758163, "learning_rate": 7.243299012862061e-05, "loss": 0.009, "step": 34860 }, { "grad_norm": 0.05509807914495468, "learning_rate": 7.241669710862051e-05, "loss": 0.0061, "step": 34870 }, { "grad_norm": 0.06266231089830399, "learning_rate": 7.240040110894293e-05, "loss": 0.008, "step": 34880 }, { "grad_norm": 0.05111311376094818, "learning_rate": 7.238410213175397e-05, "loss": 0.0069, "step": 34890 }, { "grad_norm": 0.08463827520608902, "learning_rate": 7.236780017922013e-05, "loss": 0.0071, "step": 34900 }, { "grad_norm": 0.13874612748622894, "learning_rate": 7.235149525350829e-05, "loss": 0.0088, "step": 34910 }, { "grad_norm": 0.16499999165534973, "learning_rate": 7.233518735678576e-05, "loss": 0.0109, "step": 34920 }, { "grad_norm": 0.12766298651695251, "learning_rate": 7.23188764912202e-05, "loss": 0.0105, "step": 34930 }, { "grad_norm": 0.10461036115884781, "learning_rate": 7.23025626589797e-05, "loss": 0.0083, "step": 34940 }, { "grad_norm": 0.07142335176467896, "learning_rate": 7.228624586223273e-05, "loss": 0.007, "step": 34950 }, { "grad_norm": 0.12977531552314758, "learning_rate": 7.226992610314816e-05, "loss": 0.0078, "step": 34960 }, { "grad_norm": 0.06261751800775528, "learning_rate": 7.225360338389523e-05, "loss": 0.0081, "step": 34970 }, { "grad_norm": 0.08371587097644806, "learning_rate": 7.22372777066436e-05, "loss": 0.0104, "step": 34980 }, { "grad_norm": 0.06568609923124313, "learning_rate": 7.222094907356333e-05, "loss": 0.0058, "step": 34990 }, { "grad_norm": 0.06435814499855042, "learning_rate": 7.220461748682486e-05, "loss": 0.0084, "step": 35000 }, { "grad_norm": 0.07885490357875824, "learning_rate": 7.218828294859898e-05, "loss": 0.0053, "step": 35010 }, { "grad_norm": 0.07744737714529037, "learning_rate": 7.217194546105695e-05, "loss": 0.0082, "step": 35020 }, { "grad_norm": 0.06941525638103485, "learning_rate": 7.215560502637038e-05, "loss": 0.0093, "step": 35030 }, { "grad_norm": 0.09205649048089981, "learning_rate": 7.213926164671129e-05, "loss": 0.0093, "step": 35040 }, { "grad_norm": 0.1391592174768448, "learning_rate": 7.212291532425203e-05, "loss": 0.0119, "step": 35050 }, { "grad_norm": 0.11763410270214081, "learning_rate": 7.210656606116546e-05, "loss": 0.0106, "step": 35060 }, { "grad_norm": 0.06810466945171356, "learning_rate": 7.20902138596247e-05, "loss": 0.0105, "step": 35070 }, { "grad_norm": 0.10295013338327408, "learning_rate": 7.207385872180335e-05, "loss": 0.0095, "step": 35080 }, { "grad_norm": 0.10172868520021439, "learning_rate": 7.205750064987536e-05, "loss": 0.0094, "step": 35090 }, { "grad_norm": 0.10910629481077194, "learning_rate": 7.20411396460151e-05, "loss": 0.0093, "step": 35100 }, { "grad_norm": 0.055453088134527206, "learning_rate": 7.20247757123973e-05, "loss": 0.0085, "step": 35110 }, { "grad_norm": 0.07602134346961975, "learning_rate": 7.200840885119709e-05, "loss": 0.0095, "step": 35120 }, { "grad_norm": 0.09080955386161804, "learning_rate": 7.199203906458997e-05, "loss": 0.0103, "step": 35130 }, { "grad_norm": 0.09569746255874634, "learning_rate": 7.197566635475188e-05, "loss": 0.0102, "step": 35140 }, { "grad_norm": 0.0768779069185257, "learning_rate": 7.195929072385911e-05, "loss": 0.0113, "step": 35150 }, { "grad_norm": 0.08383350819349289, "learning_rate": 7.194291217408832e-05, "loss": 0.0082, "step": 35160 }, { "grad_norm": 0.10508305579423904, "learning_rate": 7.192653070761662e-05, "loss": 0.0117, "step": 35170 }, { "grad_norm": 0.12952473759651184, "learning_rate": 7.191014632662145e-05, "loss": 0.0072, "step": 35180 }, { "grad_norm": 0.13895805180072784, "learning_rate": 7.189375903328067e-05, "loss": 0.0095, "step": 35190 }, { "grad_norm": 0.11330769211053848, "learning_rate": 7.187736882977251e-05, "loss": 0.0105, "step": 35200 }, { "grad_norm": 0.07136110216379166, "learning_rate": 7.186097571827558e-05, "loss": 0.008, "step": 35210 }, { "grad_norm": 0.08132078498601913, "learning_rate": 7.184457970096892e-05, "loss": 0.0086, "step": 35220 }, { "grad_norm": 0.08813179284334183, "learning_rate": 7.18281807800319e-05, "loss": 0.0066, "step": 35230 }, { "grad_norm": 0.11599569022655487, "learning_rate": 7.181177895764429e-05, "loss": 0.0068, "step": 35240 }, { "grad_norm": 0.12601138651371002, "learning_rate": 7.179537423598629e-05, "loss": 0.0129, "step": 35250 }, { "grad_norm": 0.0828147679567337, "learning_rate": 7.177896661723843e-05, "loss": 0.011, "step": 35260 }, { "grad_norm": 0.12230803072452545, "learning_rate": 7.176255610358165e-05, "loss": 0.0099, "step": 35270 }, { "grad_norm": 0.10617672652006149, "learning_rate": 7.174614269719728e-05, "loss": 0.0099, "step": 35280 }, { "grad_norm": 0.1235966831445694, "learning_rate": 7.172972640026702e-05, "loss": 0.0098, "step": 35290 }, { "grad_norm": 0.10098989307880402, "learning_rate": 7.171330721497298e-05, "loss": 0.0103, "step": 35300 }, { "grad_norm": 0.1248214840888977, "learning_rate": 7.16968851434976e-05, "loss": 0.0116, "step": 35310 }, { "grad_norm": 0.13953977823257446, "learning_rate": 7.168046018802375e-05, "loss": 0.0093, "step": 35320 }, { "grad_norm": 0.07533708959817886, "learning_rate": 7.16640323507347e-05, "loss": 0.0081, "step": 35330 }, { "grad_norm": 0.1472221463918686, "learning_rate": 7.164760163381403e-05, "loss": 0.0086, "step": 35340 }, { "grad_norm": 0.12881186604499817, "learning_rate": 7.163116803944578e-05, "loss": 0.0081, "step": 35350 }, { "grad_norm": 0.0754442811012268, "learning_rate": 7.161473156981434e-05, "loss": 0.0082, "step": 35360 }, { "grad_norm": 0.073973648250103, "learning_rate": 7.159829222710447e-05, "loss": 0.0136, "step": 35370 }, { "grad_norm": 0.08231991529464722, "learning_rate": 7.158185001350132e-05, "loss": 0.0112, "step": 35380 }, { "grad_norm": 0.09614882618188858, "learning_rate": 7.156540493119044e-05, "loss": 0.0087, "step": 35390 }, { "grad_norm": 0.10404762625694275, "learning_rate": 7.154895698235774e-05, "loss": 0.007, "step": 35400 }, { "grad_norm": 0.060414887964725494, "learning_rate": 7.153250616918952e-05, "loss": 0.0068, "step": 35410 }, { "grad_norm": 0.04798613488674164, "learning_rate": 7.151605249387246e-05, "loss": 0.0065, "step": 35420 }, { "grad_norm": 0.07149998843669891, "learning_rate": 7.149959595859359e-05, "loss": 0.0072, "step": 35430 }, { "grad_norm": 0.13062770664691925, "learning_rate": 7.14831365655404e-05, "loss": 0.0093, "step": 35440 }, { "grad_norm": 0.0644158124923706, "learning_rate": 7.146667431690068e-05, "loss": 0.0093, "step": 35450 }, { "grad_norm": 0.06686301529407501, "learning_rate": 7.145020921486264e-05, "loss": 0.0066, "step": 35460 }, { "grad_norm": 0.03871317580342293, "learning_rate": 7.143374126161484e-05, "loss": 0.0079, "step": 35470 }, { "grad_norm": 0.10335280746221542, "learning_rate": 7.141727045934627e-05, "loss": 0.0069, "step": 35480 }, { "grad_norm": 0.06933179497718811, "learning_rate": 7.140079681024622e-05, "loss": 0.0079, "step": 35490 }, { "grad_norm": 0.0649435892701149, "learning_rate": 7.138432031650443e-05, "loss": 0.0084, "step": 35500 }, { "grad_norm": 0.09196845442056656, "learning_rate": 7.136784098031102e-05, "loss": 0.0082, "step": 35510 }, { "grad_norm": 0.08843008428812027, "learning_rate": 7.135135880385642e-05, "loss": 0.0059, "step": 35520 }, { "grad_norm": 0.1572529673576355, "learning_rate": 7.133487378933148e-05, "loss": 0.007, "step": 35530 }, { "grad_norm": 0.10437791794538498, "learning_rate": 7.131838593892743e-05, "loss": 0.0073, "step": 35540 }, { "grad_norm": 0.09671199321746826, "learning_rate": 7.13018952548359e-05, "loss": 0.0079, "step": 35550 }, { "grad_norm": 0.07993657141923904, "learning_rate": 7.128540173924882e-05, "loss": 0.0074, "step": 35560 }, { "grad_norm": 0.11351191252470016, "learning_rate": 7.126890539435859e-05, "loss": 0.0082, "step": 35570 }, { "grad_norm": 0.12737658619880676, "learning_rate": 7.125240622235792e-05, "loss": 0.0093, "step": 35580 }, { "grad_norm": 0.08444200456142426, "learning_rate": 7.12359042254399e-05, "loss": 0.0066, "step": 35590 }, { "grad_norm": 0.07754205167293549, "learning_rate": 7.121939940579804e-05, "loss": 0.0091, "step": 35600 }, { "grad_norm": 0.0820685476064682, "learning_rate": 7.120289176562619e-05, "loss": 0.0105, "step": 35610 }, { "grad_norm": 0.08148876577615738, "learning_rate": 7.118638130711858e-05, "loss": 0.0086, "step": 35620 }, { "grad_norm": 0.03618447110056877, "learning_rate": 7.116986803246982e-05, "loss": 0.0082, "step": 35630 }, { "grad_norm": 0.07607679069042206, "learning_rate": 7.11533519438749e-05, "loss": 0.0109, "step": 35640 }, { "grad_norm": 0.08485531061887741, "learning_rate": 7.113683304352913e-05, "loss": 0.0082, "step": 35650 }, { "grad_norm": 0.11151372641324997, "learning_rate": 7.112031133362831e-05, "loss": 0.0069, "step": 35660 }, { "grad_norm": 0.1297292858362198, "learning_rate": 7.110378681636847e-05, "loss": 0.0099, "step": 35670 }, { "grad_norm": 0.09879175573587418, "learning_rate": 7.108725949394614e-05, "loss": 0.0077, "step": 35680 }, { "grad_norm": 0.07994987070560455, "learning_rate": 7.107072936855817e-05, "loss": 0.0077, "step": 35690 }, { "grad_norm": 0.0684487372636795, "learning_rate": 7.105419644240173e-05, "loss": 0.0099, "step": 35700 }, { "grad_norm": 0.09027010947465897, "learning_rate": 7.103766071767446e-05, "loss": 0.0078, "step": 35710 }, { "grad_norm": 0.08150777220726013, "learning_rate": 7.102112219657431e-05, "loss": 0.008, "step": 35720 }, { "grad_norm": 0.08791422843933105, "learning_rate": 7.100458088129962e-05, "loss": 0.0114, "step": 35730 }, { "grad_norm": 0.07832130044698715, "learning_rate": 7.098803677404909e-05, "loss": 0.0073, "step": 35740 }, { "grad_norm": 0.08881255984306335, "learning_rate": 7.097148987702181e-05, "loss": 0.007, "step": 35750 }, { "grad_norm": 0.08866631984710693, "learning_rate": 7.095494019241722e-05, "loss": 0.0091, "step": 35760 }, { "grad_norm": 0.10706900805234909, "learning_rate": 7.093838772243516e-05, "loss": 0.0089, "step": 35770 }, { "grad_norm": 0.06379691511392593, "learning_rate": 7.09218324692758e-05, "loss": 0.0069, "step": 35780 }, { "grad_norm": 0.09321291744709015, "learning_rate": 7.09052744351397e-05, "loss": 0.0072, "step": 35790 }, { "grad_norm": 0.07731746882200241, "learning_rate": 7.088871362222781e-05, "loss": 0.0083, "step": 35800 }, { "grad_norm": 0.09996787458658218, "learning_rate": 7.087215003274143e-05, "loss": 0.0083, "step": 35810 }, { "grad_norm": 0.06991636008024216, "learning_rate": 7.085558366888223e-05, "loss": 0.0072, "step": 35820 }, { "grad_norm": 0.05716767907142639, "learning_rate": 7.083901453285221e-05, "loss": 0.0119, "step": 35830 }, { "grad_norm": 0.13140816986560822, "learning_rate": 7.082244262685382e-05, "loss": 0.0082, "step": 35840 }, { "grad_norm": 0.07000504434108734, "learning_rate": 7.080586795308984e-05, "loss": 0.0085, "step": 35850 }, { "grad_norm": 0.13254591822624207, "learning_rate": 7.078929051376336e-05, "loss": 0.0078, "step": 35860 }, { "grad_norm": 0.09048014134168625, "learning_rate": 7.077271031107793e-05, "loss": 0.0092, "step": 35870 }, { "grad_norm": 0.0817372277379036, "learning_rate": 7.075612734723742e-05, "loss": 0.0077, "step": 35880 }, { "grad_norm": 0.07133448868989944, "learning_rate": 7.073954162444609e-05, "loss": 0.0066, "step": 35890 }, { "grad_norm": 0.06304362416267395, "learning_rate": 7.072295314490852e-05, "loss": 0.0078, "step": 35900 }, { "grad_norm": 0.09751981496810913, "learning_rate": 7.070636191082971e-05, "loss": 0.0071, "step": 35910 }, { "grad_norm": 0.09211141616106033, "learning_rate": 7.0689767924415e-05, "loss": 0.0077, "step": 35920 }, { "grad_norm": 0.12100861221551895, "learning_rate": 7.067317118787012e-05, "loss": 0.0104, "step": 35930 }, { "grad_norm": 0.07895390689373016, "learning_rate": 7.06565717034011e-05, "loss": 0.0099, "step": 35940 }, { "grad_norm": 0.0685167983174324, "learning_rate": 7.063996947321438e-05, "loss": 0.0069, "step": 35950 }, { "grad_norm": 0.07044051587581635, "learning_rate": 7.062336449951684e-05, "loss": 0.0063, "step": 35960 }, { "grad_norm": 0.08895476907491684, "learning_rate": 7.060675678451556e-05, "loss": 0.0065, "step": 35970 }, { "grad_norm": 0.0731094479560852, "learning_rate": 7.059014633041813e-05, "loss": 0.0092, "step": 35980 }, { "grad_norm": 0.06221918761730194, "learning_rate": 7.057353313943244e-05, "loss": 0.0072, "step": 35990 }, { "grad_norm": 0.09161228686571121, "learning_rate": 7.055691721376674e-05, "loss": 0.0103, "step": 36000 }, { "grad_norm": 0.09650254994630814, "learning_rate": 7.054029855562965e-05, "loss": 0.0082, "step": 36010 }, { "grad_norm": 0.11616703122854233, "learning_rate": 7.052367716723018e-05, "loss": 0.0081, "step": 36020 }, { "grad_norm": 0.05362202599644661, "learning_rate": 7.050705305077767e-05, "loss": 0.0085, "step": 36030 }, { "grad_norm": 0.09417697042226791, "learning_rate": 7.049042620848185e-05, "loss": 0.0087, "step": 36040 }, { "grad_norm": 0.08113718777894974, "learning_rate": 7.047379664255278e-05, "loss": 0.0085, "step": 36050 }, { "grad_norm": 0.11330759525299072, "learning_rate": 7.045716435520088e-05, "loss": 0.0092, "step": 36060 }, { "grad_norm": 0.11407840996980667, "learning_rate": 7.044052934863701e-05, "loss": 0.0093, "step": 36070 }, { "grad_norm": 0.09100725501775742, "learning_rate": 7.042389162507226e-05, "loss": 0.0089, "step": 36080 }, { "grad_norm": 0.08965455740690231, "learning_rate": 7.040725118671823e-05, "loss": 0.0119, "step": 36090 }, { "grad_norm": 0.06894996017217636, "learning_rate": 7.039060803578673e-05, "loss": 0.0079, "step": 36100 }, { "grad_norm": 0.0967545285820961, "learning_rate": 7.037396217449005e-05, "loss": 0.0077, "step": 36110 }, { "grad_norm": 0.10136600583791733, "learning_rate": 7.035731360504078e-05, "loss": 0.0073, "step": 36120 }, { "grad_norm": 0.044328249990940094, "learning_rate": 7.03406623296519e-05, "loss": 0.0086, "step": 36130 }, { "grad_norm": 0.09063518792390823, "learning_rate": 7.032400835053672e-05, "loss": 0.0074, "step": 36140 }, { "grad_norm": 0.06786669790744781, "learning_rate": 7.030735166990894e-05, "loss": 0.0086, "step": 36150 }, { "grad_norm": 0.11748550832271576, "learning_rate": 7.029069228998258e-05, "loss": 0.0075, "step": 36160 }, { "grad_norm": 0.10802765190601349, "learning_rate": 7.027403021297204e-05, "loss": 0.0085, "step": 36170 }, { "grad_norm": 0.046769920736551285, "learning_rate": 7.025736544109212e-05, "loss": 0.0115, "step": 36180 }, { "grad_norm": 0.09191359579563141, "learning_rate": 7.024069797655789e-05, "loss": 0.0082, "step": 36190 }, { "grad_norm": 0.06730924546718597, "learning_rate": 7.022402782158486e-05, "loss": 0.0061, "step": 36200 }, { "grad_norm": 0.09373071044683456, "learning_rate": 7.020735497838885e-05, "loss": 0.0079, "step": 36210 }, { "grad_norm": 0.07442416995763779, "learning_rate": 7.019067944918606e-05, "loss": 0.0074, "step": 36220 }, { "grad_norm": 0.1297074258327484, "learning_rate": 7.017400123619303e-05, "loss": 0.0076, "step": 36230 }, { "grad_norm": 0.07240854203701019, "learning_rate": 7.015732034162666e-05, "loss": 0.0092, "step": 36240 }, { "grad_norm": 0.07388704270124435, "learning_rate": 7.014063676770424e-05, "loss": 0.0061, "step": 36250 }, { "grad_norm": 0.09984459728002548, "learning_rate": 7.012395051664336e-05, "loss": 0.0101, "step": 36260 }, { "grad_norm": 0.10956718772649765, "learning_rate": 7.0107261590662e-05, "loss": 0.0093, "step": 36270 }, { "grad_norm": 0.14034539461135864, "learning_rate": 7.009056999197848e-05, "loss": 0.0131, "step": 36280 }, { "grad_norm": 0.1317913830280304, "learning_rate": 7.007387572281152e-05, "loss": 0.0077, "step": 36290 }, { "grad_norm": 0.1095149964094162, "learning_rate": 7.005717878538012e-05, "loss": 0.0073, "step": 36300 }, { "grad_norm": 0.06374325603246689, "learning_rate": 7.00404791819037e-05, "loss": 0.0078, "step": 36310 }, { "grad_norm": 0.07864585518836975, "learning_rate": 7.002377691460199e-05, "loss": 0.0077, "step": 36320 }, { "grad_norm": 0.12031165510416031, "learning_rate": 7.000707198569511e-05, "loss": 0.012, "step": 36330 }, { "grad_norm": 0.1109401285648346, "learning_rate": 6.999036439740351e-05, "loss": 0.0092, "step": 36340 }, { "grad_norm": 0.10232444107532501, "learning_rate": 6.997365415194797e-05, "loss": 0.0108, "step": 36350 }, { "grad_norm": 0.11270467936992645, "learning_rate": 6.995694125154972e-05, "loss": 0.0082, "step": 36360 }, { "grad_norm": 0.09514545649290085, "learning_rate": 6.994022569843021e-05, "loss": 0.0096, "step": 36370 }, { "grad_norm": 0.07061028480529785, "learning_rate": 6.992350749481135e-05, "loss": 0.0078, "step": 36380 }, { "grad_norm": 0.0956130176782608, "learning_rate": 6.990678664291535e-05, "loss": 0.009, "step": 36390 }, { "grad_norm": 0.076134592294693, "learning_rate": 6.989006314496477e-05, "loss": 0.0068, "step": 36400 }, { "grad_norm": 0.07944320142269135, "learning_rate": 6.987333700318255e-05, "loss": 0.0081, "step": 36410 }, { "grad_norm": 0.05996865779161453, "learning_rate": 6.985660821979196e-05, "loss": 0.0072, "step": 36420 }, { "grad_norm": 0.0691591203212738, "learning_rate": 6.983987679701662e-05, "loss": 0.0122, "step": 36430 }, { "grad_norm": 0.09562978893518448, "learning_rate": 6.982314273708052e-05, "loss": 0.0087, "step": 36440 }, { "grad_norm": 0.07955796271562576, "learning_rate": 6.9806406042208e-05, "loss": 0.0091, "step": 36450 }, { "grad_norm": 0.0853070616722107, "learning_rate": 6.97896667146237e-05, "loss": 0.0092, "step": 36460 }, { "grad_norm": 0.06742710620164871, "learning_rate": 6.977292475655269e-05, "loss": 0.0084, "step": 36470 }, { "grad_norm": 0.0954577624797821, "learning_rate": 6.975618017022032e-05, "loss": 0.0086, "step": 36480 }, { "grad_norm": 0.10805349051952362, "learning_rate": 6.973943295785234e-05, "loss": 0.009, "step": 36490 }, { "grad_norm": 0.05627075955271721, "learning_rate": 6.972268312167481e-05, "loss": 0.0083, "step": 36500 }, { "grad_norm": 0.11030501127243042, "learning_rate": 6.970593066391415e-05, "loss": 0.0107, "step": 36510 }, { "grad_norm": 0.10087023675441742, "learning_rate": 6.968917558679717e-05, "loss": 0.0078, "step": 36520 }, { "grad_norm": 0.102816142141819, "learning_rate": 6.967241789255095e-05, "loss": 0.009, "step": 36530 }, { "grad_norm": 0.07883570343255997, "learning_rate": 6.965565758340298e-05, "loss": 0.0096, "step": 36540 }, { "grad_norm": 0.08873573690652847, "learning_rate": 6.963889466158107e-05, "loss": 0.0104, "step": 36550 }, { "grad_norm": 0.08044390380382538, "learning_rate": 6.962212912931341e-05, "loss": 0.0082, "step": 36560 }, { "grad_norm": 0.11981818825006485, "learning_rate": 6.960536098882848e-05, "loss": 0.0094, "step": 36570 }, { "grad_norm": 0.060121215879917145, "learning_rate": 6.958859024235514e-05, "loss": 0.008, "step": 36580 }, { "grad_norm": 0.1094861850142479, "learning_rate": 6.957181689212263e-05, "loss": 0.0105, "step": 36590 }, { "grad_norm": 0.07358111441135406, "learning_rate": 6.955504094036047e-05, "loss": 0.0086, "step": 36600 }, { "grad_norm": 0.09681835025548935, "learning_rate": 6.953826238929856e-05, "loss": 0.008, "step": 36610 }, { "grad_norm": 0.10322663187980652, "learning_rate": 6.952148124116714e-05, "loss": 0.0109, "step": 36620 }, { "grad_norm": 0.07848669588565826, "learning_rate": 6.950469749819682e-05, "loss": 0.0068, "step": 36630 }, { "grad_norm": 0.08048442006111145, "learning_rate": 6.948791116261851e-05, "loss": 0.0095, "step": 36640 }, { "grad_norm": 0.0687023475766182, "learning_rate": 6.94711222366635e-05, "loss": 0.0118, "step": 36650 }, { "grad_norm": 0.10357369482517242, "learning_rate": 6.945433072256339e-05, "loss": 0.0106, "step": 36660 }, { "grad_norm": 0.06078996881842613, "learning_rate": 6.94375366225502e-05, "loss": 0.0081, "step": 36670 }, { "grad_norm": 0.07146536558866501, "learning_rate": 6.942073993885615e-05, "loss": 0.0095, "step": 36680 }, { "grad_norm": 0.07705572992563248, "learning_rate": 6.940394067371397e-05, "loss": 0.0067, "step": 36690 }, { "grad_norm": 0.12465834617614746, "learning_rate": 6.938713882935664e-05, "loss": 0.0084, "step": 36700 }, { "grad_norm": 0.09491200000047684, "learning_rate": 6.937033440801748e-05, "loss": 0.0098, "step": 36710 }, { "grad_norm": 0.05547349154949188, "learning_rate": 6.935352741193018e-05, "loss": 0.0059, "step": 36720 }, { "grad_norm": 0.09725972265005112, "learning_rate": 6.933671784332875e-05, "loss": 0.0095, "step": 36730 }, { "grad_norm": 0.07552940398454666, "learning_rate": 6.93199057044476e-05, "loss": 0.0073, "step": 36740 }, { "grad_norm": 0.08890487253665924, "learning_rate": 6.930309099752139e-05, "loss": 0.0127, "step": 36750 }, { "grad_norm": 0.1058254987001419, "learning_rate": 6.928627372478518e-05, "loss": 0.0088, "step": 36760 }, { "grad_norm": 0.08649569004774094, "learning_rate": 6.926945388847436e-05, "loss": 0.0069, "step": 36770 }, { "grad_norm": 0.0907905101776123, "learning_rate": 6.925263149082467e-05, "loss": 0.0105, "step": 36780 }, { "grad_norm": 0.05659808591008186, "learning_rate": 6.923580653407218e-05, "loss": 0.0085, "step": 36790 }, { "grad_norm": 0.09856461733579636, "learning_rate": 6.921897902045328e-05, "loss": 0.0087, "step": 36800 }, { "grad_norm": 0.10693054646253586, "learning_rate": 6.920214895220474e-05, "loss": 0.013, "step": 36810 }, { "grad_norm": 0.049727849662303925, "learning_rate": 6.918531633156367e-05, "loss": 0.0096, "step": 36820 }, { "grad_norm": 0.1547115594148636, "learning_rate": 6.916848116076745e-05, "loss": 0.0066, "step": 36830 }, { "grad_norm": 0.11982704699039459, "learning_rate": 6.915164344205389e-05, "loss": 0.0085, "step": 36840 }, { "grad_norm": 0.08923845738172531, "learning_rate": 6.913480317766106e-05, "loss": 0.0131, "step": 36850 }, { "grad_norm": 0.11319869011640549, "learning_rate": 6.911796036982745e-05, "loss": 0.0091, "step": 36860 }, { "grad_norm": 0.13471822440624237, "learning_rate": 6.91011150207918e-05, "loss": 0.0073, "step": 36870 }, { "grad_norm": 0.0865391418337822, "learning_rate": 6.908426713279328e-05, "loss": 0.0068, "step": 36880 }, { "grad_norm": 0.1352127194404602, "learning_rate": 6.906741670807131e-05, "loss": 0.0087, "step": 36890 }, { "grad_norm": 0.12036032974720001, "learning_rate": 6.905056374886569e-05, "loss": 0.0076, "step": 36900 }, { "grad_norm": 0.0777793601155281, "learning_rate": 6.903370825741657e-05, "loss": 0.0077, "step": 36910 }, { "grad_norm": 0.10584138333797455, "learning_rate": 6.90168502359644e-05, "loss": 0.0096, "step": 36920 }, { "grad_norm": 0.09470030665397644, "learning_rate": 6.899998968675e-05, "loss": 0.0094, "step": 36930 }, { "grad_norm": 0.043130990117788315, "learning_rate": 6.898312661201452e-05, "loss": 0.0067, "step": 36940 }, { "grad_norm": 0.0666317567229271, "learning_rate": 6.896626101399941e-05, "loss": 0.0079, "step": 36950 }, { "grad_norm": 0.1133568286895752, "learning_rate": 6.89493928949465e-05, "loss": 0.0076, "step": 36960 }, { "grad_norm": 0.0821882113814354, "learning_rate": 6.893252225709794e-05, "loss": 0.0084, "step": 36970 }, { "grad_norm": 0.09662975370883942, "learning_rate": 6.891564910269621e-05, "loss": 0.0076, "step": 36980 }, { "grad_norm": 0.05955641716718674, "learning_rate": 6.889877343398412e-05, "loss": 0.0084, "step": 36990 }, { "grad_norm": 0.10174603015184402, "learning_rate": 6.888189525320484e-05, "loss": 0.0076, "step": 37000 }, { "grad_norm": 0.0945497453212738, "learning_rate": 6.886501456260184e-05, "loss": 0.0097, "step": 37010 }, { "grad_norm": 0.07826759666204453, "learning_rate": 6.884813136441894e-05, "loss": 0.0104, "step": 37020 }, { "grad_norm": 0.06753826141357422, "learning_rate": 6.88312456609003e-05, "loss": 0.0103, "step": 37030 }, { "grad_norm": 0.06891679018735886, "learning_rate": 6.881435745429039e-05, "loss": 0.0085, "step": 37040 }, { "grad_norm": 0.08738237619400024, "learning_rate": 6.879746674683405e-05, "loss": 0.0067, "step": 37050 }, { "grad_norm": 0.08420731872320175, "learning_rate": 6.878057354077641e-05, "loss": 0.0081, "step": 37060 }, { "grad_norm": 0.10697801411151886, "learning_rate": 6.876367783836297e-05, "loss": 0.0095, "step": 37070 }, { "grad_norm": 0.08498971164226532, "learning_rate": 6.874677964183955e-05, "loss": 0.0109, "step": 37080 }, { "grad_norm": 0.05315164104104042, "learning_rate": 6.872987895345227e-05, "loss": 0.0079, "step": 37090 }, { "grad_norm": 0.09175831824541092, "learning_rate": 6.871297577544761e-05, "loss": 0.0086, "step": 37100 }, { "grad_norm": 0.07117197662591934, "learning_rate": 6.86960701100724e-05, "loss": 0.0085, "step": 37110 }, { "grad_norm": 0.05011255294084549, "learning_rate": 6.867916195957377e-05, "loss": 0.007, "step": 37120 }, { "grad_norm": 0.05163322016596794, "learning_rate": 6.866225132619918e-05, "loss": 0.0086, "step": 37130 }, { "grad_norm": 0.1093757301568985, "learning_rate": 6.864533821219644e-05, "loss": 0.0123, "step": 37140 }, { "grad_norm": 0.08670884370803833, "learning_rate": 6.862842261981366e-05, "loss": 0.0072, "step": 37150 }, { "grad_norm": 0.06011553108692169, "learning_rate": 6.861150455129934e-05, "loss": 0.0052, "step": 37160 }, { "grad_norm": 0.06013115867972374, "learning_rate": 6.859458400890222e-05, "loss": 0.0089, "step": 37170 }, { "grad_norm": 0.06258372962474823, "learning_rate": 6.857766099487144e-05, "loss": 0.0084, "step": 37180 }, { "grad_norm": 0.07834348827600479, "learning_rate": 6.856073551145645e-05, "loss": 0.0091, "step": 37190 }, { "grad_norm": 0.07240147143602371, "learning_rate": 6.854380756090702e-05, "loss": 0.0082, "step": 37200 }, { "grad_norm": 0.06108317896723747, "learning_rate": 6.852687714547321e-05, "loss": 0.0104, "step": 37210 }, { "grad_norm": 0.06725631654262543, "learning_rate": 6.850994426740551e-05, "loss": 0.0071, "step": 37220 }, { "grad_norm": 0.11191081255674362, "learning_rate": 6.849300892895464e-05, "loss": 0.0077, "step": 37230 }, { "grad_norm": 0.04889184981584549, "learning_rate": 6.847607113237171e-05, "loss": 0.0063, "step": 37240 }, { "grad_norm": 0.06303760409355164, "learning_rate": 6.845913087990811e-05, "loss": 0.0089, "step": 37250 }, { "grad_norm": 0.08774460107088089, "learning_rate": 6.844218817381557e-05, "loss": 0.0076, "step": 37260 }, { "grad_norm": 0.04414036497473717, "learning_rate": 6.842524301634618e-05, "loss": 0.0057, "step": 37270 }, { "grad_norm": 0.05943996459245682, "learning_rate": 6.840829540975228e-05, "loss": 0.0092, "step": 37280 }, { "grad_norm": 0.05031190067529678, "learning_rate": 6.839134535628664e-05, "loss": 0.0061, "step": 37290 }, { "grad_norm": 0.03878283128142357, "learning_rate": 6.837439285820225e-05, "loss": 0.0067, "step": 37300 }, { "grad_norm": 0.07679332047700882, "learning_rate": 6.835743791775251e-05, "loss": 0.0094, "step": 37310 }, { "grad_norm": 0.07587996870279312, "learning_rate": 6.834048053719109e-05, "loss": 0.006, "step": 37320 }, { "grad_norm": 0.06681464612483978, "learning_rate": 6.832352071877202e-05, "loss": 0.0053, "step": 37330 }, { "grad_norm": 0.09079189598560333, "learning_rate": 6.83065584647496e-05, "loss": 0.0058, "step": 37340 }, { "grad_norm": 0.12221890687942505, "learning_rate": 6.828959377737854e-05, "loss": 0.0099, "step": 37350 }, { "grad_norm": 0.06721244752407074, "learning_rate": 6.827262665891378e-05, "loss": 0.0078, "step": 37360 }, { "grad_norm": 0.12097384035587311, "learning_rate": 6.825565711161067e-05, "loss": 0.0073, "step": 37370 }, { "grad_norm": 0.12070044130086899, "learning_rate": 6.823868513772481e-05, "loss": 0.0107, "step": 37380 }, { "grad_norm": 0.08443859219551086, "learning_rate": 6.822171073951215e-05, "loss": 0.0061, "step": 37390 }, { "grad_norm": 0.09473170340061188, "learning_rate": 6.820473391922896e-05, "loss": 0.0087, "step": 37400 }, { "grad_norm": 0.06867605447769165, "learning_rate": 6.81877546791319e-05, "loss": 0.0064, "step": 37410 }, { "grad_norm": 0.12412045896053314, "learning_rate": 6.817077302147781e-05, "loss": 0.008, "step": 37420 }, { "grad_norm": 0.10358088463544846, "learning_rate": 6.815378894852395e-05, "loss": 0.0075, "step": 37430 }, { "grad_norm": 0.09996317327022552, "learning_rate": 6.813680246252792e-05, "loss": 0.0078, "step": 37440 }, { "grad_norm": 0.11844933778047562, "learning_rate": 6.811981356574754e-05, "loss": 0.008, "step": 37450 }, { "grad_norm": 0.0948706865310669, "learning_rate": 6.810282226044107e-05, "loss": 0.0061, "step": 37460 }, { "grad_norm": 0.09053956717252731, "learning_rate": 6.8085828548867e-05, "loss": 0.0104, "step": 37470 }, { "grad_norm": 0.07256996631622314, "learning_rate": 6.806883243328418e-05, "loss": 0.0074, "step": 37480 }, { "grad_norm": 0.06584764271974564, "learning_rate": 6.805183391595178e-05, "loss": 0.0089, "step": 37490 }, { "grad_norm": 0.06274767965078354, "learning_rate": 6.803483299912926e-05, "loss": 0.0078, "step": 37500 }, { "grad_norm": 0.09953071922063828, "learning_rate": 6.801782968507642e-05, "loss": 0.0071, "step": 37510 }, { "grad_norm": 0.07514460384845734, "learning_rate": 6.800082397605342e-05, "loss": 0.0065, "step": 37520 }, { "grad_norm": 0.11533594131469727, "learning_rate": 6.798381587432065e-05, "loss": 0.0076, "step": 37530 }, { "grad_norm": 0.1096930131316185, "learning_rate": 6.796680538213888e-05, "loss": 0.0085, "step": 37540 }, { "grad_norm": 0.07854968309402466, "learning_rate": 6.794979250176919e-05, "loss": 0.009, "step": 37550 }, { "grad_norm": 0.1158631443977356, "learning_rate": 6.793277723547295e-05, "loss": 0.0067, "step": 37560 }, { "grad_norm": 0.09352250397205353, "learning_rate": 6.791575958551191e-05, "loss": 0.0106, "step": 37570 }, { "grad_norm": 0.06014394387602806, "learning_rate": 6.789873955414804e-05, "loss": 0.0061, "step": 37580 }, { "grad_norm": 0.12030401080846786, "learning_rate": 6.788171714364371e-05, "loss": 0.009, "step": 37590 }, { "grad_norm": 0.08097858726978302, "learning_rate": 6.786469235626161e-05, "loss": 0.0091, "step": 37600 }, { "grad_norm": 0.0786476582288742, "learning_rate": 6.784766519426464e-05, "loss": 0.0092, "step": 37610 }, { "grad_norm": 0.05021728575229645, "learning_rate": 6.783063565991612e-05, "loss": 0.0074, "step": 37620 }, { "grad_norm": 0.0709703117609024, "learning_rate": 6.781360375547968e-05, "loss": 0.0064, "step": 37630 }, { "grad_norm": 0.05852317065000534, "learning_rate": 6.779656948321921e-05, "loss": 0.0093, "step": 37640 }, { "grad_norm": 0.06845971196889877, "learning_rate": 6.777953284539894e-05, "loss": 0.0086, "step": 37650 }, { "grad_norm": 0.07559092342853546, "learning_rate": 6.776249384428344e-05, "loss": 0.0083, "step": 37660 }, { "grad_norm": 0.12193787097930908, "learning_rate": 6.774545248213756e-05, "loss": 0.0083, "step": 37670 }, { "grad_norm": 0.0956074595451355, "learning_rate": 6.772840876122649e-05, "loss": 0.0087, "step": 37680 }, { "grad_norm": 0.056520041078329086, "learning_rate": 6.771136268381568e-05, "loss": 0.0064, "step": 37690 }, { "grad_norm": 0.10694726556539536, "learning_rate": 6.769431425217096e-05, "loss": 0.0115, "step": 37700 }, { "grad_norm": 0.0886588990688324, "learning_rate": 6.767726346855848e-05, "loss": 0.0081, "step": 37710 }, { "grad_norm": 0.06299006938934326, "learning_rate": 6.76602103352446e-05, "loss": 0.008, "step": 37720 }, { "grad_norm": 0.07586823403835297, "learning_rate": 6.764315485449609e-05, "loss": 0.0078, "step": 37730 }, { "grad_norm": 0.056104354560375214, "learning_rate": 6.762609702858003e-05, "loss": 0.0077, "step": 37740 }, { "grad_norm": 0.08994238823652267, "learning_rate": 6.760903685976376e-05, "loss": 0.0062, "step": 37750 }, { "grad_norm": 0.06240953877568245, "learning_rate": 6.759197435031493e-05, "loss": 0.009, "step": 37760 }, { "grad_norm": 0.06669165939092636, "learning_rate": 6.757490950250156e-05, "loss": 0.0101, "step": 37770 }, { "grad_norm": 0.06556902080774307, "learning_rate": 6.755784231859197e-05, "loss": 0.0065, "step": 37780 }, { "grad_norm": 0.11247236281633377, "learning_rate": 6.754077280085473e-05, "loss": 0.0102, "step": 37790 }, { "grad_norm": 0.08086694777011871, "learning_rate": 6.752370095155873e-05, "loss": 0.0064, "step": 37800 }, { "grad_norm": 0.062436699867248535, "learning_rate": 6.750662677297327e-05, "loss": 0.0059, "step": 37810 }, { "grad_norm": 0.0727086290717125, "learning_rate": 6.748955026736785e-05, "loss": 0.0079, "step": 37820 }, { "grad_norm": 0.07265880703926086, "learning_rate": 6.747247143701232e-05, "loss": 0.0082, "step": 37830 }, { "grad_norm": 0.09955130517482758, "learning_rate": 6.745539028417683e-05, "loss": 0.0104, "step": 37840 }, { "grad_norm": 0.06553969532251358, "learning_rate": 6.743830681113185e-05, "loss": 0.0071, "step": 37850 }, { "grad_norm": 0.06928043067455292, "learning_rate": 6.742122102014818e-05, "loss": 0.008, "step": 37860 }, { "grad_norm": 0.07778820395469666, "learning_rate": 6.740413291349685e-05, "loss": 0.0078, "step": 37870 }, { "grad_norm": 0.048899419605731964, "learning_rate": 6.73870424934493e-05, "loss": 0.005, "step": 37880 }, { "grad_norm": 0.056174252182245255, "learning_rate": 6.736994976227718e-05, "loss": 0.0076, "step": 37890 }, { "grad_norm": 0.09404437988996506, "learning_rate": 6.735285472225254e-05, "loss": 0.0094, "step": 37900 }, { "grad_norm": 0.111930251121521, "learning_rate": 6.733575737564766e-05, "loss": 0.0062, "step": 37910 }, { "grad_norm": 0.085015669465065, "learning_rate": 6.731865772473516e-05, "loss": 0.0093, "step": 37920 }, { "grad_norm": 0.06770452857017517, "learning_rate": 6.730155577178798e-05, "loss": 0.0101, "step": 37930 }, { "grad_norm": 0.07384662330150604, "learning_rate": 6.728445151907933e-05, "loss": 0.0092, "step": 37940 }, { "grad_norm": 0.08234405517578125, "learning_rate": 6.726734496888276e-05, "loss": 0.0062, "step": 37950 }, { "grad_norm": 0.0683758407831192, "learning_rate": 6.72502361234721e-05, "loss": 0.0112, "step": 37960 }, { "grad_norm": 0.0544305182993412, "learning_rate": 6.723312498512151e-05, "loss": 0.0084, "step": 37970 }, { "grad_norm": 0.12225553393363953, "learning_rate": 6.721601155610541e-05, "loss": 0.0099, "step": 37980 }, { "grad_norm": 0.11448413878679276, "learning_rate": 6.719889583869861e-05, "loss": 0.0083, "step": 37990 }, { "grad_norm": 0.08596745878458023, "learning_rate": 6.718177783517612e-05, "loss": 0.0068, "step": 38000 }, { "grad_norm": 0.08619800955057144, "learning_rate": 6.716465754781334e-05, "loss": 0.0074, "step": 38010 }, { "grad_norm": 0.062385909259319305, "learning_rate": 6.714753497888591e-05, "loss": 0.0066, "step": 38020 }, { "grad_norm": 0.05890148878097534, "learning_rate": 6.713041013066978e-05, "loss": 0.0061, "step": 38030 }, { "grad_norm": 0.11691924184560776, "learning_rate": 6.711328300544129e-05, "loss": 0.0078, "step": 38040 }, { "grad_norm": 0.0754266232252121, "learning_rate": 6.709615360547695e-05, "loss": 0.0082, "step": 38050 }, { "grad_norm": 0.10126645863056183, "learning_rate": 6.707902193305367e-05, "loss": 0.01, "step": 38060 }, { "grad_norm": 0.11266569793224335, "learning_rate": 6.706188799044863e-05, "loss": 0.0094, "step": 38070 }, { "grad_norm": 0.08318397402763367, "learning_rate": 6.70447517799393e-05, "loss": 0.012, "step": 38080 }, { "grad_norm": 0.08380749076604843, "learning_rate": 6.702761330380347e-05, "loss": 0.0091, "step": 38090 }, { "grad_norm": 0.06319395452737808, "learning_rate": 6.701047256431924e-05, "loss": 0.0086, "step": 38100 }, { "grad_norm": 0.11257658153772354, "learning_rate": 6.699332956376496e-05, "loss": 0.0132, "step": 38110 }, { "grad_norm": 0.09245412051677704, "learning_rate": 6.697618430441936e-05, "loss": 0.0109, "step": 38120 }, { "grad_norm": 0.05271298810839653, "learning_rate": 6.69590367885614e-05, "loss": 0.0099, "step": 38130 }, { "grad_norm": 0.09097958356142044, "learning_rate": 6.694188701847033e-05, "loss": 0.0075, "step": 38140 }, { "grad_norm": 0.07103300839662552, "learning_rate": 6.692473499642582e-05, "loss": 0.009, "step": 38150 }, { "grad_norm": 0.06683411449193954, "learning_rate": 6.690758072470769e-05, "loss": 0.008, "step": 38160 }, { "grad_norm": 0.13225233554840088, "learning_rate": 6.689042420559614e-05, "loss": 0.0075, "step": 38170 }, { "grad_norm": 0.14169491827487946, "learning_rate": 6.687326544137166e-05, "loss": 0.0103, "step": 38180 }, { "grad_norm": 0.08069960027933121, "learning_rate": 6.685610443431502e-05, "loss": 0.01, "step": 38190 }, { "grad_norm": 0.08524820953607559, "learning_rate": 6.683894118670733e-05, "loss": 0.0106, "step": 38200 }, { "grad_norm": 0.07583096623420715, "learning_rate": 6.682177570082992e-05, "loss": 0.0146, "step": 38210 }, { "grad_norm": 0.12624873220920563, "learning_rate": 6.68046079789645e-05, "loss": 0.0096, "step": 38220 }, { "grad_norm": 0.11590844392776489, "learning_rate": 6.678743802339304e-05, "loss": 0.0108, "step": 38230 }, { "grad_norm": 0.11722099035978317, "learning_rate": 6.677026583639779e-05, "loss": 0.0086, "step": 38240 }, { "grad_norm": 0.07416753470897675, "learning_rate": 6.675309142026131e-05, "loss": 0.0088, "step": 38250 }, { "grad_norm": 0.07575259357690811, "learning_rate": 6.67359147772665e-05, "loss": 0.0089, "step": 38260 }, { "grad_norm": 0.07452613115310669, "learning_rate": 6.671873590969648e-05, "loss": 0.0081, "step": 38270 }, { "grad_norm": 0.09853768348693848, "learning_rate": 6.670155481983473e-05, "loss": 0.0126, "step": 38280 }, { "grad_norm": 0.12036710232496262, "learning_rate": 6.668437150996498e-05, "loss": 0.01, "step": 38290 }, { "grad_norm": 0.08048511296510696, "learning_rate": 6.666718598237127e-05, "loss": 0.0088, "step": 38300 }, { "grad_norm": 0.120839923620224, "learning_rate": 6.664999823933797e-05, "loss": 0.0105, "step": 38310 }, { "grad_norm": 0.08830559253692627, "learning_rate": 6.663280828314969e-05, "loss": 0.0056, "step": 38320 }, { "grad_norm": 0.10719076544046402, "learning_rate": 6.661561611609134e-05, "loss": 0.0068, "step": 38330 }, { "grad_norm": 0.05465337261557579, "learning_rate": 6.659842174044819e-05, "loss": 0.007, "step": 38340 }, { "grad_norm": 0.09209486097097397, "learning_rate": 6.658122515850572e-05, "loss": 0.01, "step": 38350 }, { "grad_norm": 0.05368725210428238, "learning_rate": 6.656402637254974e-05, "loss": 0.0075, "step": 38360 }, { "grad_norm": 0.09109025448560715, "learning_rate": 6.654682538486635e-05, "loss": 0.0113, "step": 38370 }, { "grad_norm": 0.06937707960605621, "learning_rate": 6.652962219774196e-05, "loss": 0.0094, "step": 38380 }, { "grad_norm": 0.0654439851641655, "learning_rate": 6.651241681346322e-05, "loss": 0.0105, "step": 38390 }, { "grad_norm": 0.04654552415013313, "learning_rate": 6.649520923431718e-05, "loss": 0.0086, "step": 38400 }, { "grad_norm": 0.0867776870727539, "learning_rate": 6.647799946259103e-05, "loss": 0.0064, "step": 38410 }, { "grad_norm": 0.08292561769485474, "learning_rate": 6.646078750057237e-05, "loss": 0.0077, "step": 38420 }, { "grad_norm": 0.058433517813682556, "learning_rate": 6.644357335054905e-05, "loss": 0.0094, "step": 38430 }, { "grad_norm": 0.07587524503469467, "learning_rate": 6.64263570148092e-05, "loss": 0.0083, "step": 38440 }, { "grad_norm": 0.11545754969120026, "learning_rate": 6.640913849564129e-05, "loss": 0.0089, "step": 38450 }, { "grad_norm": 0.12786200642585754, "learning_rate": 6.6391917795334e-05, "loss": 0.0074, "step": 38460 }, { "grad_norm": 0.06050293147563934, "learning_rate": 6.637469491617636e-05, "loss": 0.0067, "step": 38470 }, { "grad_norm": 0.09473413974046707, "learning_rate": 6.635746986045767e-05, "loss": 0.0101, "step": 38480 }, { "grad_norm": 0.14913690090179443, "learning_rate": 6.634024263046755e-05, "loss": 0.0086, "step": 38490 }, { "grad_norm": 0.1035320982336998, "learning_rate": 6.632301322849586e-05, "loss": 0.0145, "step": 38500 }, { "grad_norm": 0.10235659033060074, "learning_rate": 6.630578165683275e-05, "loss": 0.008, "step": 38510 }, { "grad_norm": 0.0966181755065918, "learning_rate": 6.628854791776872e-05, "loss": 0.0104, "step": 38520 }, { "grad_norm": 0.08498836308717728, "learning_rate": 6.62713120135945e-05, "loss": 0.0065, "step": 38530 }, { "grad_norm": 0.0875912755727768, "learning_rate": 6.625407394660111e-05, "loss": 0.0065, "step": 38540 }, { "grad_norm": 0.07256269454956055, "learning_rate": 6.623683371907988e-05, "loss": 0.0095, "step": 38550 }, { "grad_norm": 0.11812624335289001, "learning_rate": 6.621959133332243e-05, "loss": 0.0115, "step": 38560 }, { "grad_norm": 0.10177316516637802, "learning_rate": 6.620234679162066e-05, "loss": 0.0075, "step": 38570 }, { "grad_norm": 0.06052793562412262, "learning_rate": 6.618510009626674e-05, "loss": 0.0072, "step": 38580 }, { "grad_norm": 0.06285776197910309, "learning_rate": 6.616785124955314e-05, "loss": 0.0106, "step": 38590 }, { "grad_norm": 0.09938086569309235, "learning_rate": 6.615060025377263e-05, "loss": 0.008, "step": 38600 }, { "grad_norm": 0.1157355085015297, "learning_rate": 6.613334711121824e-05, "loss": 0.0097, "step": 38610 }, { "grad_norm": 0.08024630695581436, "learning_rate": 6.611609182418326e-05, "loss": 0.007, "step": 38620 }, { "grad_norm": 0.1034490317106247, "learning_rate": 6.609883439496136e-05, "loss": 0.011, "step": 38630 }, { "grad_norm": 0.08287197351455688, "learning_rate": 6.608157482584642e-05, "loss": 0.0108, "step": 38640 }, { "grad_norm": 0.10224293917417526, "learning_rate": 6.606431311913262e-05, "loss": 0.0073, "step": 38650 }, { "grad_norm": 0.07438401132822037, "learning_rate": 6.60470492771144e-05, "loss": 0.0067, "step": 38660 }, { "grad_norm": 0.09718012064695358, "learning_rate": 6.602978330208654e-05, "loss": 0.0084, "step": 38670 }, { "grad_norm": 0.08772219717502594, "learning_rate": 6.601251519634405e-05, "loss": 0.0074, "step": 38680 }, { "grad_norm": 0.08796469122171402, "learning_rate": 6.599524496218226e-05, "loss": 0.0071, "step": 38690 }, { "grad_norm": 0.06481295824050903, "learning_rate": 6.597797260189675e-05, "loss": 0.0072, "step": 38700 }, { "grad_norm": 0.10740674287080765, "learning_rate": 6.596069811778342e-05, "loss": 0.0078, "step": 38710 }, { "grad_norm": 0.09241297841072083, "learning_rate": 6.594342151213844e-05, "loss": 0.0062, "step": 38720 }, { "grad_norm": 0.08262161165475845, "learning_rate": 6.592614278725822e-05, "loss": 0.0096, "step": 38730 }, { "grad_norm": 0.08509314805269241, "learning_rate": 6.590886194543951e-05, "loss": 0.0132, "step": 38740 }, { "grad_norm": 0.09338142722845078, "learning_rate": 6.589157898897932e-05, "loss": 0.0085, "step": 38750 }, { "grad_norm": 0.06717520207166672, "learning_rate": 6.587429392017493e-05, "loss": 0.0086, "step": 38760 }, { "grad_norm": 0.11932142078876495, "learning_rate": 6.585700674132392e-05, "loss": 0.0098, "step": 38770 }, { "grad_norm": 0.11585943400859833, "learning_rate": 6.583971745472413e-05, "loss": 0.0097, "step": 38780 }, { "grad_norm": 0.10750219970941544, "learning_rate": 6.58224260626737e-05, "loss": 0.0112, "step": 38790 }, { "grad_norm": 0.094743512570858, "learning_rate": 6.580513256747105e-05, "loss": 0.0077, "step": 38800 }, { "grad_norm": 0.09980218857526779, "learning_rate": 6.578783697141484e-05, "loss": 0.0118, "step": 38810 }, { "grad_norm": 0.05641765147447586, "learning_rate": 6.577053927680405e-05, "loss": 0.0083, "step": 38820 }, { "grad_norm": 0.08261173963546753, "learning_rate": 6.575323948593796e-05, "loss": 0.0065, "step": 38830 }, { "grad_norm": 0.11845619231462479, "learning_rate": 6.573593760111604e-05, "loss": 0.0094, "step": 38840 }, { "grad_norm": 0.07325800508260727, "learning_rate": 6.571863362463815e-05, "loss": 0.0073, "step": 38850 }, { "grad_norm": 0.0824207067489624, "learning_rate": 6.570132755880434e-05, "loss": 0.0073, "step": 38860 }, { "grad_norm": 0.07137097418308258, "learning_rate": 6.568401940591498e-05, "loss": 0.0115, "step": 38870 }, { "grad_norm": 0.06405159085988998, "learning_rate": 6.56667091682707e-05, "loss": 0.0088, "step": 38880 }, { "grad_norm": 0.09166977554559708, "learning_rate": 6.564939684817244e-05, "loss": 0.0062, "step": 38890 }, { "grad_norm": 0.0643981397151947, "learning_rate": 6.563208244792136e-05, "loss": 0.0061, "step": 38900 }, { "grad_norm": 0.09763398766517639, "learning_rate": 6.561476596981896e-05, "loss": 0.008, "step": 38910 }, { "grad_norm": 0.07583257555961609, "learning_rate": 6.559744741616696e-05, "loss": 0.0077, "step": 38920 }, { "grad_norm": 0.09385839104652405, "learning_rate": 6.558012678926739e-05, "loss": 0.0067, "step": 38930 }, { "grad_norm": 0.13577789068222046, "learning_rate": 6.556280409142256e-05, "loss": 0.0073, "step": 38940 }, { "grad_norm": 0.12632326781749725, "learning_rate": 6.554547932493502e-05, "loss": 0.0068, "step": 38950 }, { "grad_norm": 0.0962429791688919, "learning_rate": 6.552815249210762e-05, "loss": 0.0097, "step": 38960 }, { "grad_norm": 0.059107523411512375, "learning_rate": 6.551082359524349e-05, "loss": 0.0069, "step": 38970 }, { "grad_norm": 0.09747492522001266, "learning_rate": 6.549349263664603e-05, "loss": 0.0083, "step": 38980 }, { "grad_norm": 0.07712098211050034, "learning_rate": 6.54761596186189e-05, "loss": 0.0071, "step": 38990 }, { "grad_norm": 0.1013716608285904, "learning_rate": 6.545882454346604e-05, "loss": 0.0074, "step": 39000 }, { "grad_norm": 0.12747429311275482, "learning_rate": 6.544148741349167e-05, "loss": 0.0135, "step": 39010 }, { "grad_norm": 0.11907556653022766, "learning_rate": 6.54241482310003e-05, "loss": 0.0103, "step": 39020 }, { "grad_norm": 0.07296089828014374, "learning_rate": 6.540680699829666e-05, "loss": 0.0067, "step": 39030 }, { "grad_norm": 0.08428299427032471, "learning_rate": 6.538946371768581e-05, "loss": 0.0089, "step": 39040 }, { "grad_norm": 0.12061871588230133, "learning_rate": 6.537211839147304e-05, "loss": 0.0104, "step": 39050 }, { "grad_norm": 0.10805148631334305, "learning_rate": 6.535477102196395e-05, "loss": 0.0076, "step": 39060 }, { "grad_norm": 0.09075508266687393, "learning_rate": 6.533742161146437e-05, "loss": 0.0065, "step": 39070 }, { "grad_norm": 0.07124684751033783, "learning_rate": 6.532007016228044e-05, "loss": 0.005, "step": 39080 }, { "grad_norm": 0.07447582483291626, "learning_rate": 6.530271667671853e-05, "loss": 0.0079, "step": 39090 }, { "grad_norm": 0.09295971691608429, "learning_rate": 6.528536115708534e-05, "loss": 0.0091, "step": 39100 }, { "grad_norm": 0.07026001811027527, "learning_rate": 6.526800360568777e-05, "loss": 0.0106, "step": 39110 }, { "grad_norm": 0.06284094601869583, "learning_rate": 6.525064402483304e-05, "loss": 0.0071, "step": 39120 }, { "grad_norm": 0.06530087441205978, "learning_rate": 6.523328241682863e-05, "loss": 0.0075, "step": 39130 }, { "grad_norm": 0.10596907883882523, "learning_rate": 6.521591878398226e-05, "loss": 0.0117, "step": 39140 }, { "grad_norm": 0.11388742923736572, "learning_rate": 6.519855312860197e-05, "loss": 0.007, "step": 39150 }, { "grad_norm": 0.1047310084104538, "learning_rate": 6.518118545299606e-05, "loss": 0.0084, "step": 39160 }, { "grad_norm": 0.058642297983169556, "learning_rate": 6.516381575947302e-05, "loss": 0.0074, "step": 39170 }, { "grad_norm": 0.06770835816860199, "learning_rate": 6.514644405034168e-05, "loss": 0.0057, "step": 39180 }, { "grad_norm": 0.141516774892807, "learning_rate": 6.512907032791118e-05, "loss": 0.0077, "step": 39190 }, { "grad_norm": 0.11283325403928757, "learning_rate": 6.511169459449082e-05, "loss": 0.0094, "step": 39200 }, { "grad_norm": 0.09188342839479446, "learning_rate": 6.509431685239026e-05, "loss": 0.0071, "step": 39210 }, { "grad_norm": 0.07570643723011017, "learning_rate": 6.507693710391935e-05, "loss": 0.0085, "step": 39220 }, { "grad_norm": 0.07038857042789459, "learning_rate": 6.505955535138828e-05, "loss": 0.0067, "step": 39230 }, { "grad_norm": 0.06667089462280273, "learning_rate": 6.504217159710746e-05, "loss": 0.0072, "step": 39240 }, { "grad_norm": 0.06712468713521957, "learning_rate": 6.502478584338756e-05, "loss": 0.0095, "step": 39250 }, { "grad_norm": 0.09474257379770279, "learning_rate": 6.500739809253955e-05, "loss": 0.0072, "step": 39260 }, { "grad_norm": 0.0924653634428978, "learning_rate": 6.499000834687465e-05, "loss": 0.0068, "step": 39270 }, { "grad_norm": 0.09458119422197342, "learning_rate": 6.497261660870434e-05, "loss": 0.0064, "step": 39280 }, { "grad_norm": 0.081211157143116, "learning_rate": 6.495522288034035e-05, "loss": 0.0098, "step": 39290 }, { "grad_norm": 0.06550891697406769, "learning_rate": 6.493782716409474e-05, "loss": 0.0063, "step": 39300 }, { "grad_norm": 0.062220677733421326, "learning_rate": 6.492042946227974e-05, "loss": 0.0092, "step": 39310 }, { "grad_norm": 0.06904816627502441, "learning_rate": 6.490302977720792e-05, "loss": 0.0113, "step": 39320 }, { "grad_norm": 0.06835070252418518, "learning_rate": 6.488562811119206e-05, "loss": 0.0069, "step": 39330 }, { "grad_norm": 0.08002263307571411, "learning_rate": 6.486822446654526e-05, "loss": 0.0126, "step": 39340 }, { "grad_norm": 0.07446366548538208, "learning_rate": 6.485081884558084e-05, "loss": 0.0072, "step": 39350 }, { "grad_norm": 0.09194230288267136, "learning_rate": 6.483341125061238e-05, "loss": 0.0065, "step": 39360 }, { "grad_norm": 0.07453200966119766, "learning_rate": 6.481600168395373e-05, "loss": 0.0055, "step": 39370 }, { "grad_norm": 0.05269410088658333, "learning_rate": 6.479859014791904e-05, "loss": 0.008, "step": 39380 }, { "grad_norm": 0.05069803446531296, "learning_rate": 6.478117664482269e-05, "loss": 0.0061, "step": 39390 }, { "grad_norm": 0.05843590572476387, "learning_rate": 6.476376117697929e-05, "loss": 0.0062, "step": 39400 }, { "grad_norm": 0.10355322808027267, "learning_rate": 6.474634374670376e-05, "loss": 0.0066, "step": 39410 }, { "grad_norm": 0.07138184458017349, "learning_rate": 6.472892435631126e-05, "loss": 0.0057, "step": 39420 }, { "grad_norm": 0.04973659664392471, "learning_rate": 6.47115030081172e-05, "loss": 0.0073, "step": 39430 }, { "grad_norm": 0.06615117192268372, "learning_rate": 6.46940797044373e-05, "loss": 0.0079, "step": 39440 }, { "grad_norm": 0.09506656974554062, "learning_rate": 6.467665444758749e-05, "loss": 0.0079, "step": 39450 }, { "grad_norm": 0.07418422400951385, "learning_rate": 6.465922723988397e-05, "loss": 0.0068, "step": 39460 }, { "grad_norm": 0.0885431095957756, "learning_rate": 6.46417980836432e-05, "loss": 0.0073, "step": 39470 }, { "grad_norm": 0.06464112550020218, "learning_rate": 6.462436698118189e-05, "loss": 0.008, "step": 39480 }, { "grad_norm": 0.13807697594165802, "learning_rate": 6.460693393481707e-05, "loss": 0.0093, "step": 39490 }, { "grad_norm": 0.10508783161640167, "learning_rate": 6.458949894686592e-05, "loss": 0.0088, "step": 39500 }, { "grad_norm": 0.0697803869843483, "learning_rate": 6.457206201964597e-05, "loss": 0.0098, "step": 39510 }, { "grad_norm": 0.05346653237938881, "learning_rate": 6.455462315547497e-05, "loss": 0.0119, "step": 39520 }, { "grad_norm": 0.0925048366189003, "learning_rate": 6.453718235667094e-05, "loss": 0.0075, "step": 39530 }, { "grad_norm": 0.1107984185218811, "learning_rate": 6.451973962555214e-05, "loss": 0.0096, "step": 39540 }, { "grad_norm": 0.0676523745059967, "learning_rate": 6.450229496443708e-05, "loss": 0.0071, "step": 39550 }, { "grad_norm": 0.07344026118516922, "learning_rate": 6.448484837564458e-05, "loss": 0.0068, "step": 39560 }, { "grad_norm": 0.10694350302219391, "learning_rate": 6.446739986149367e-05, "loss": 0.0112, "step": 39570 }, { "grad_norm": 0.09022445231676102, "learning_rate": 6.444994942430363e-05, "loss": 0.0077, "step": 39580 }, { "grad_norm": 0.06944653391838074, "learning_rate": 6.4432497066394e-05, "loss": 0.0094, "step": 39590 }, { "grad_norm": 0.08280327171087265, "learning_rate": 6.441504279008463e-05, "loss": 0.0078, "step": 39600 }, { "grad_norm": 0.11659647524356842, "learning_rate": 6.439758659769555e-05, "loss": 0.0094, "step": 39610 }, { "grad_norm": 0.07196104526519775, "learning_rate": 6.438012849154706e-05, "loss": 0.0081, "step": 39620 }, { "grad_norm": 0.05764380469918251, "learning_rate": 6.436266847395977e-05, "loss": 0.0079, "step": 39630 }, { "grad_norm": 0.08318324387073517, "learning_rate": 6.43452065472545e-05, "loss": 0.0072, "step": 39640 }, { "grad_norm": 0.07624868303537369, "learning_rate": 6.43277427137523e-05, "loss": 0.0083, "step": 39650 }, { "grad_norm": 0.08740174770355225, "learning_rate": 6.431027697577452e-05, "loss": 0.0095, "step": 39660 }, { "grad_norm": 0.09187567234039307, "learning_rate": 6.429280933564275e-05, "loss": 0.0076, "step": 39670 }, { "grad_norm": 0.07798799872398376, "learning_rate": 6.427533979567881e-05, "loss": 0.0076, "step": 39680 }, { "grad_norm": 0.06222902983427048, "learning_rate": 6.42578683582048e-05, "loss": 0.0071, "step": 39690 }, { "grad_norm": 0.052967701107263565, "learning_rate": 6.424039502554306e-05, "loss": 0.0066, "step": 39700 }, { "grad_norm": 0.08980324119329453, "learning_rate": 6.422291980001622e-05, "loss": 0.009, "step": 39710 }, { "grad_norm": 0.06843192130327225, "learning_rate": 6.420544268394706e-05, "loss": 0.0083, "step": 39720 }, { "grad_norm": 0.05380164086818695, "learning_rate": 6.418796367965871e-05, "loss": 0.0094, "step": 39730 }, { "grad_norm": 0.07367569953203201, "learning_rate": 6.417048278947454e-05, "loss": 0.0077, "step": 39740 }, { "grad_norm": 0.0394727885723114, "learning_rate": 6.41530000157181e-05, "loss": 0.0074, "step": 39750 }, { "grad_norm": 0.1030198261141777, "learning_rate": 6.41355153607133e-05, "loss": 0.0073, "step": 39760 }, { "grad_norm": 0.09026467800140381, "learning_rate": 6.411802882678418e-05, "loss": 0.007, "step": 39770 }, { "grad_norm": 0.09710322320461273, "learning_rate": 6.410054041625513e-05, "loss": 0.0075, "step": 39780 }, { "grad_norm": 0.09165684133768082, "learning_rate": 6.408305013145075e-05, "loss": 0.009, "step": 39790 }, { "grad_norm": 0.06988545507192612, "learning_rate": 6.406555797469584e-05, "loss": 0.0058, "step": 39800 }, { "grad_norm": 0.12089096754789352, "learning_rate": 6.404806394831555e-05, "loss": 0.0121, "step": 39810 }, { "grad_norm": 0.06934870034456253, "learning_rate": 6.40305680546352e-05, "loss": 0.0076, "step": 39820 }, { "grad_norm": 0.06501897424459457, "learning_rate": 6.401307029598038e-05, "loss": 0.0078, "step": 39830 }, { "grad_norm": 0.06456924229860306, "learning_rate": 6.399557067467695e-05, "loss": 0.01, "step": 39840 }, { "grad_norm": 0.13911615312099457, "learning_rate": 6.397806919305098e-05, "loss": 0.0111, "step": 39850 }, { "grad_norm": 0.11651995778083801, "learning_rate": 6.396056585342881e-05, "loss": 0.0112, "step": 39860 }, { "grad_norm": 0.07682178169488907, "learning_rate": 6.394306065813705e-05, "loss": 0.0081, "step": 39870 }, { "grad_norm": 0.0907202735543251, "learning_rate": 6.392555360950246e-05, "loss": 0.0064, "step": 39880 }, { "grad_norm": 0.09986193478107452, "learning_rate": 6.390804470985219e-05, "loss": 0.0077, "step": 39890 }, { "grad_norm": 0.0783512070775032, "learning_rate": 6.389053396151354e-05, "loss": 0.0093, "step": 39900 }, { "grad_norm": 0.0742025151848793, "learning_rate": 6.387302136681407e-05, "loss": 0.0106, "step": 39910 }, { "grad_norm": 0.08888280391693115, "learning_rate": 6.385550692808159e-05, "loss": 0.0109, "step": 39920 }, { "grad_norm": 0.06071489304304123, "learning_rate": 6.383799064764416e-05, "loss": 0.0064, "step": 39930 }, { "grad_norm": 0.13186310231685638, "learning_rate": 6.382047252783009e-05, "loss": 0.0072, "step": 39940 }, { "grad_norm": 0.05728396028280258, "learning_rate": 6.380295257096792e-05, "loss": 0.0082, "step": 39950 }, { "grad_norm": 0.10934489220380783, "learning_rate": 6.378543077938646e-05, "loss": 0.0072, "step": 39960 }, { "grad_norm": 0.06620492786169052, "learning_rate": 6.376790715541473e-05, "loss": 0.0056, "step": 39970 }, { "grad_norm": 0.1111692264676094, "learning_rate": 6.3750381701382e-05, "loss": 0.0094, "step": 39980 }, { "grad_norm": 0.10147394239902496, "learning_rate": 6.37328544196178e-05, "loss": 0.009, "step": 39990 }, { "grad_norm": 0.1064109057188034, "learning_rate": 6.37153253124519e-05, "loss": 0.0086, "step": 40000 }, { "grad_norm": 0.09403026103973389, "learning_rate": 6.369779438221433e-05, "loss": 0.0059, "step": 40010 }, { "grad_norm": 0.0674983561038971, "learning_rate": 6.36802616312353e-05, "loss": 0.0072, "step": 40020 }, { "grad_norm": 0.08621898293495178, "learning_rate": 6.366272706184531e-05, "loss": 0.0082, "step": 40030 }, { "grad_norm": 0.0923430472612381, "learning_rate": 6.364519067637513e-05, "loss": 0.0075, "step": 40040 }, { "grad_norm": 0.0675194039940834, "learning_rate": 6.362765247715568e-05, "loss": 0.0084, "step": 40050 }, { "grad_norm": 0.08380679041147232, "learning_rate": 6.361011246651823e-05, "loss": 0.0071, "step": 40060 }, { "grad_norm": 0.06877037137746811, "learning_rate": 6.35925706467942e-05, "loss": 0.0073, "step": 40070 }, { "grad_norm": 0.05254466086626053, "learning_rate": 6.357502702031531e-05, "loss": 0.0086, "step": 40080 }, { "grad_norm": 0.055889762938022614, "learning_rate": 6.355748158941349e-05, "loss": 0.0068, "step": 40090 }, { "grad_norm": 0.09967955946922302, "learning_rate": 6.353993435642092e-05, "loss": 0.0064, "step": 40100 }, { "grad_norm": 0.06777455657720566, "learning_rate": 6.352238532366999e-05, "loss": 0.0056, "step": 40110 }, { "grad_norm": 0.09709642082452774, "learning_rate": 6.350483449349341e-05, "loss": 0.0076, "step": 40120 }, { "grad_norm": 0.05299577862024307, "learning_rate": 6.348728186822404e-05, "loss": 0.0057, "step": 40130 }, { "grad_norm": 0.07010839879512787, "learning_rate": 6.346972745019502e-05, "loss": 0.0085, "step": 40140 }, { "grad_norm": 0.10830499976873398, "learning_rate": 6.345217124173973e-05, "loss": 0.0067, "step": 40150 }, { "grad_norm": 0.06824865937232971, "learning_rate": 6.343461324519176e-05, "loss": 0.0093, "step": 40160 }, { "grad_norm": 0.08982426673173904, "learning_rate": 6.3417053462885e-05, "loss": 0.0091, "step": 40170 }, { "grad_norm": 0.04612194746732712, "learning_rate": 6.339949189715347e-05, "loss": 0.0064, "step": 40180 }, { "grad_norm": 0.07849040627479553, "learning_rate": 6.338192855033156e-05, "loss": 0.0065, "step": 40190 }, { "grad_norm": 0.1202956885099411, "learning_rate": 6.336436342475379e-05, "loss": 0.0099, "step": 40200 }, { "grad_norm": 0.09320255368947983, "learning_rate": 6.334679652275497e-05, "loss": 0.0087, "step": 40210 }, { "grad_norm": 0.11417946964502335, "learning_rate": 6.332922784667012e-05, "loss": 0.0085, "step": 40220 }, { "grad_norm": 0.07424434274435043, "learning_rate": 6.331165739883451e-05, "loss": 0.0069, "step": 40230 }, { "grad_norm": 0.07827644050121307, "learning_rate": 6.329408518158366e-05, "loss": 0.008, "step": 40240 }, { "grad_norm": 0.07232929766178131, "learning_rate": 6.327651119725329e-05, "loss": 0.009, "step": 40250 }, { "grad_norm": 0.06547466665506363, "learning_rate": 6.325893544817939e-05, "loss": 0.0081, "step": 40260 }, { "grad_norm": 0.09425748884677887, "learning_rate": 6.324135793669816e-05, "loss": 0.007, "step": 40270 }, { "grad_norm": 0.0921710729598999, "learning_rate": 6.322377866514603e-05, "loss": 0.0071, "step": 40280 }, { "grad_norm": 0.07233410328626633, "learning_rate": 6.320619763585968e-05, "loss": 0.0053, "step": 40290 }, { "grad_norm": 0.0713125616312027, "learning_rate": 6.318861485117603e-05, "loss": 0.0088, "step": 40300 }, { "grad_norm": 0.08011849224567413, "learning_rate": 6.317103031343223e-05, "loss": 0.0092, "step": 40310 }, { "grad_norm": 0.08991288393735886, "learning_rate": 6.315344402496564e-05, "loss": 0.008, "step": 40320 }, { "grad_norm": 0.07348714768886566, "learning_rate": 6.313585598811387e-05, "loss": 0.0071, "step": 40330 }, { "grad_norm": 0.08431585133075714, "learning_rate": 6.311826620521476e-05, "loss": 0.0114, "step": 40340 }, { "grad_norm": 0.08819885551929474, "learning_rate": 6.31006746786064e-05, "loss": 0.0089, "step": 40350 }, { "grad_norm": 0.12837953865528107, "learning_rate": 6.308308141062708e-05, "loss": 0.0081, "step": 40360 }, { "grad_norm": 0.06490875780582428, "learning_rate": 6.306548640361532e-05, "loss": 0.0083, "step": 40370 }, { "grad_norm": 0.10242890566587448, "learning_rate": 6.304788965990993e-05, "loss": 0.0071, "step": 40380 }, { "grad_norm": 0.06830273568630219, "learning_rate": 6.303029118184988e-05, "loss": 0.0079, "step": 40390 }, { "grad_norm": 0.07722336053848267, "learning_rate": 6.30126909717744e-05, "loss": 0.0065, "step": 40400 }, { "grad_norm": 0.06334064900875092, "learning_rate": 6.299508903202295e-05, "loss": 0.0088, "step": 40410 }, { "grad_norm": 0.05134838446974754, "learning_rate": 6.297748536493525e-05, "loss": 0.0094, "step": 40420 }, { "grad_norm": 0.06804074347019196, "learning_rate": 6.295987997285118e-05, "loss": 0.0093, "step": 40430 }, { "grad_norm": 0.07573133707046509, "learning_rate": 6.29422728581109e-05, "loss": 0.0093, "step": 40440 }, { "grad_norm": 0.09261585026979446, "learning_rate": 6.292466402305477e-05, "loss": 0.0073, "step": 40450 }, { "grad_norm": 0.07492463290691376, "learning_rate": 6.290705347002343e-05, "loss": 0.0082, "step": 40460 }, { "grad_norm": 0.07731151580810547, "learning_rate": 6.288944120135768e-05, "loss": 0.0066, "step": 40470 }, { "grad_norm": 0.06437815725803375, "learning_rate": 6.287182721939861e-05, "loss": 0.0075, "step": 40480 }, { "grad_norm": 0.04061596840620041, "learning_rate": 6.285421152648749e-05, "loss": 0.008, "step": 40490 }, { "grad_norm": 0.07280567288398743, "learning_rate": 6.283659412496588e-05, "loss": 0.0065, "step": 40500 }, { "grad_norm": 0.06034303829073906, "learning_rate": 6.281897501717546e-05, "loss": 0.0078, "step": 40510 }, { "grad_norm": 0.07600675523281097, "learning_rate": 6.280135420545823e-05, "loss": 0.0083, "step": 40520 }, { "grad_norm": 0.10474274307489395, "learning_rate": 6.27837316921564e-05, "loss": 0.0082, "step": 40530 }, { "grad_norm": 0.0918198898434639, "learning_rate": 6.276610747961238e-05, "loss": 0.0071, "step": 40540 }, { "grad_norm": 0.03757094964385033, "learning_rate": 6.274848157016881e-05, "loss": 0.0066, "step": 40550 }, { "grad_norm": 0.08708794414997101, "learning_rate": 6.27308539661686e-05, "loss": 0.0076, "step": 40560 }, { "grad_norm": 0.066868357360363, "learning_rate": 6.271322466995482e-05, "loss": 0.0064, "step": 40570 }, { "grad_norm": 0.08601290732622147, "learning_rate": 6.269559368387082e-05, "loss": 0.0094, "step": 40580 }, { "grad_norm": 0.06848090142011642, "learning_rate": 6.26779610102601e-05, "loss": 0.0093, "step": 40590 }, { "grad_norm": 0.05105799436569214, "learning_rate": 6.266032665146648e-05, "loss": 0.0079, "step": 40600 }, { "grad_norm": 0.09836756438016891, "learning_rate": 6.264269060983397e-05, "loss": 0.0064, "step": 40610 }, { "grad_norm": 0.08434228599071503, "learning_rate": 6.262505288770675e-05, "loss": 0.0099, "step": 40620 }, { "grad_norm": 0.0958184078335762, "learning_rate": 6.260741348742928e-05, "loss": 0.0067, "step": 40630 }, { "grad_norm": 0.08386557549238205, "learning_rate": 6.258977241134625e-05, "loss": 0.005, "step": 40640 }, { "grad_norm": 0.05233658105134964, "learning_rate": 6.257212966180253e-05, "loss": 0.0078, "step": 40650 }, { "grad_norm": 0.09119444340467453, "learning_rate": 6.255448524114322e-05, "loss": 0.0066, "step": 40660 }, { "grad_norm": 0.07342341542243958, "learning_rate": 6.253683915171369e-05, "loss": 0.0074, "step": 40670 }, { "grad_norm": 0.07332558929920197, "learning_rate": 6.251919139585946e-05, "loss": 0.0079, "step": 40680 }, { "grad_norm": 0.06719912588596344, "learning_rate": 6.250154197592637e-05, "loss": 0.0075, "step": 40690 }, { "grad_norm": 0.09123475104570389, "learning_rate": 6.248389089426035e-05, "loss": 0.007, "step": 40700 }, { "grad_norm": 0.08933620899915695, "learning_rate": 6.246623815320765e-05, "loss": 0.0068, "step": 40710 }, { "grad_norm": 0.07551207393407822, "learning_rate": 6.244858375511475e-05, "loss": 0.0105, "step": 40720 }, { "grad_norm": 0.0905386358499527, "learning_rate": 6.243092770232824e-05, "loss": 0.0082, "step": 40730 }, { "grad_norm": 0.0738099217414856, "learning_rate": 6.241326999719505e-05, "loss": 0.0077, "step": 40740 }, { "grad_norm": 0.1229795590043068, "learning_rate": 6.239561064206228e-05, "loss": 0.0114, "step": 40750 }, { "grad_norm": 0.10982271283864975, "learning_rate": 6.237794963927722e-05, "loss": 0.0114, "step": 40760 }, { "grad_norm": 0.06220202520489693, "learning_rate": 6.236028699118744e-05, "loss": 0.0044, "step": 40770 }, { "grad_norm": 0.10823339223861694, "learning_rate": 6.234262270014069e-05, "loss": 0.0088, "step": 40780 }, { "grad_norm": 0.10153109580278397, "learning_rate": 6.232495676848495e-05, "loss": 0.0119, "step": 40790 }, { "grad_norm": 0.12622161209583282, "learning_rate": 6.230728919856841e-05, "loss": 0.0074, "step": 40800 }, { "grad_norm": 0.06880446523427963, "learning_rate": 6.228961999273948e-05, "loss": 0.0085, "step": 40810 }, { "grad_norm": 0.12497532367706299, "learning_rate": 6.227194915334678e-05, "loss": 0.0061, "step": 40820 }, { "grad_norm": 0.10155455023050308, "learning_rate": 6.225427668273921e-05, "loss": 0.007, "step": 40830 }, { "grad_norm": 0.11738941818475723, "learning_rate": 6.223660258326577e-05, "loss": 0.0072, "step": 40840 }, { "grad_norm": 0.11940356343984604, "learning_rate": 6.221892685727578e-05, "loss": 0.007, "step": 40850 }, { "grad_norm": 0.10811838507652283, "learning_rate": 6.220124950711871e-05, "loss": 0.0072, "step": 40860 }, { "grad_norm": 0.10773178935050964, "learning_rate": 6.218357053514429e-05, "loss": 0.0093, "step": 40870 }, { "grad_norm": 0.07092864066362381, "learning_rate": 6.216588994370244e-05, "loss": 0.0066, "step": 40880 }, { "grad_norm": 0.0900164395570755, "learning_rate": 6.214820773514333e-05, "loss": 0.0081, "step": 40890 }, { "grad_norm": 0.09190156310796738, "learning_rate": 6.213052391181727e-05, "loss": 0.0075, "step": 40900 }, { "grad_norm": 0.08969218283891678, "learning_rate": 6.21128384760749e-05, "loss": 0.0081, "step": 40910 }, { "grad_norm": 0.05380887910723686, "learning_rate": 6.209515143026693e-05, "loss": 0.007, "step": 40920 }, { "grad_norm": 0.08097700029611588, "learning_rate": 6.207746277674441e-05, "loss": 0.0096, "step": 40930 }, { "grad_norm": 0.0842425674200058, "learning_rate": 6.205977251785857e-05, "loss": 0.0067, "step": 40940 }, { "grad_norm": 0.07191229611635208, "learning_rate": 6.204208065596079e-05, "loss": 0.0108, "step": 40950 }, { "grad_norm": 0.08561170101165771, "learning_rate": 6.202438719340275e-05, "loss": 0.0097, "step": 40960 }, { "grad_norm": 0.08192303776741028, "learning_rate": 6.200669213253628e-05, "loss": 0.0073, "step": 40970 }, { "grad_norm": 0.10718468576669693, "learning_rate": 6.198899547571346e-05, "loss": 0.008, "step": 40980 }, { "grad_norm": 0.10084526985883713, "learning_rate": 6.197129722528656e-05, "loss": 0.0076, "step": 40990 }, { "grad_norm": 0.07979357242584229, "learning_rate": 6.19535973836081e-05, "loss": 0.0059, "step": 41000 }, { "grad_norm": 0.10057612508535385, "learning_rate": 6.193589595303074e-05, "loss": 0.0068, "step": 41010 }, { "grad_norm": 0.11003877222537994, "learning_rate": 6.191819293590744e-05, "loss": 0.0065, "step": 41020 }, { "grad_norm": 0.12600477039813995, "learning_rate": 6.190048833459128e-05, "loss": 0.0069, "step": 41030 }, { "grad_norm": 0.09780317544937134, "learning_rate": 6.188278215143561e-05, "loss": 0.0062, "step": 41040 }, { "grad_norm": 0.09399113059043884, "learning_rate": 6.186507438879398e-05, "loss": 0.0083, "step": 41050 }, { "grad_norm": 0.09542534500360489, "learning_rate": 6.184736504902015e-05, "loss": 0.0081, "step": 41060 }, { "grad_norm": 0.05561593919992447, "learning_rate": 6.182965413446809e-05, "loss": 0.0085, "step": 41070 }, { "grad_norm": 0.1083793193101883, "learning_rate": 6.181194164749195e-05, "loss": 0.0094, "step": 41080 }, { "grad_norm": 0.08964836597442627, "learning_rate": 6.179422759044613e-05, "loss": 0.007, "step": 41090 }, { "grad_norm": 0.07064773887395859, "learning_rate": 6.177651196568523e-05, "loss": 0.0066, "step": 41100 }, { "grad_norm": 0.14032121002674103, "learning_rate": 6.175879477556403e-05, "loss": 0.0095, "step": 41110 }, { "grad_norm": 0.08203326165676117, "learning_rate": 6.174107602243755e-05, "loss": 0.0084, "step": 41120 }, { "grad_norm": 0.08568380028009415, "learning_rate": 6.172335570866102e-05, "loss": 0.0081, "step": 41130 }, { "grad_norm": 0.08717844635248184, "learning_rate": 6.170563383658984e-05, "loss": 0.0074, "step": 41140 }, { "grad_norm": 0.08382012695074081, "learning_rate": 6.168791040857963e-05, "loss": 0.009, "step": 41150 }, { "grad_norm": 0.09281866997480392, "learning_rate": 6.167018542698628e-05, "loss": 0.0078, "step": 41160 }, { "grad_norm": 0.12409144639968872, "learning_rate": 6.165245889416578e-05, "loss": 0.0068, "step": 41170 }, { "grad_norm": 0.05346833914518356, "learning_rate": 6.16347308124744e-05, "loss": 0.0078, "step": 41180 }, { "grad_norm": 0.07431412488222122, "learning_rate": 6.16170011842686e-05, "loss": 0.0073, "step": 41190 }, { "grad_norm": 0.08781518042087555, "learning_rate": 6.159927001190505e-05, "loss": 0.0086, "step": 41200 }, { "grad_norm": 0.11480323225259781, "learning_rate": 6.15815372977406e-05, "loss": 0.0098, "step": 41210 }, { "grad_norm": 0.06752396374940872, "learning_rate": 6.156380304413232e-05, "loss": 0.0081, "step": 41220 }, { "grad_norm": 0.03606092184782028, "learning_rate": 6.154606725343752e-05, "loss": 0.0059, "step": 41230 }, { "grad_norm": 0.058556314557790756, "learning_rate": 6.152832992801365e-05, "loss": 0.006, "step": 41240 }, { "grad_norm": 0.05341140180826187, "learning_rate": 6.151059107021839e-05, "loss": 0.008, "step": 41250 }, { "grad_norm": 0.07722097635269165, "learning_rate": 6.149285068240965e-05, "loss": 0.0087, "step": 41260 }, { "grad_norm": 0.05795702338218689, "learning_rate": 6.14751087669455e-05, "loss": 0.0057, "step": 41270 }, { "grad_norm": 0.08118646591901779, "learning_rate": 6.145736532618425e-05, "loss": 0.0084, "step": 41280 }, { "grad_norm": 0.09644917398691177, "learning_rate": 6.14396203624844e-05, "loss": 0.0056, "step": 41290 }, { "grad_norm": 0.08438591659069061, "learning_rate": 6.142187387820464e-05, "loss": 0.0079, "step": 41300 }, { "grad_norm": 0.09518436342477798, "learning_rate": 6.140412587570387e-05, "loss": 0.0086, "step": 41310 }, { "grad_norm": 0.06878586113452911, "learning_rate": 6.138637635734123e-05, "loss": 0.0081, "step": 41320 }, { "grad_norm": 0.07197091728448868, "learning_rate": 6.136862532547597e-05, "loss": 0.0092, "step": 41330 }, { "grad_norm": 0.06031981483101845, "learning_rate": 6.135087278246762e-05, "loss": 0.005, "step": 41340 }, { "grad_norm": 0.0841837078332901, "learning_rate": 6.133311873067592e-05, "loss": 0.0084, "step": 41350 }, { "grad_norm": 0.06713452190160751, "learning_rate": 6.131536317246074e-05, "loss": 0.0065, "step": 41360 }, { "grad_norm": 0.07797381281852722, "learning_rate": 6.129760611018219e-05, "loss": 0.0094, "step": 41370 }, { "grad_norm": 0.10660041868686676, "learning_rate": 6.127984754620059e-05, "loss": 0.0093, "step": 41380 }, { "grad_norm": 0.07892429828643799, "learning_rate": 6.126208748287644e-05, "loss": 0.0098, "step": 41390 }, { "grad_norm": 0.051455289125442505, "learning_rate": 6.124432592257045e-05, "loss": 0.0063, "step": 41400 }, { "grad_norm": 0.07452793419361115, "learning_rate": 6.122656286764354e-05, "loss": 0.0091, "step": 41410 }, { "grad_norm": 0.10212372243404388, "learning_rate": 6.12087983204568e-05, "loss": 0.0079, "step": 41420 }, { "grad_norm": 0.08206582814455032, "learning_rate": 6.119103228337155e-05, "loss": 0.0109, "step": 41430 }, { "grad_norm": 0.08114787191152573, "learning_rate": 6.117326475874925e-05, "loss": 0.0084, "step": 41440 }, { "grad_norm": 0.0668124407529831, "learning_rate": 6.115549574895164e-05, "loss": 0.0088, "step": 41450 }, { "grad_norm": 0.10620235651731491, "learning_rate": 6.113772525634061e-05, "loss": 0.0064, "step": 41460 }, { "grad_norm": 0.1318274438381195, "learning_rate": 6.111995328327825e-05, "loss": 0.0076, "step": 41470 }, { "grad_norm": 0.07385097444057465, "learning_rate": 6.110217983212684e-05, "loss": 0.0058, "step": 41480 }, { "grad_norm": 0.05439101159572601, "learning_rate": 6.108440490524887e-05, "loss": 0.0086, "step": 41490 }, { "grad_norm": 0.07873532921075821, "learning_rate": 6.106662850500705e-05, "loss": 0.0065, "step": 41500 }, { "grad_norm": 0.07183508574962616, "learning_rate": 6.104885063376422e-05, "loss": 0.0048, "step": 41510 }, { "grad_norm": 0.07198204100131989, "learning_rate": 6.1031071293883454e-05, "loss": 0.0056, "step": 41520 }, { "grad_norm": 0.08487090468406677, "learning_rate": 6.101329048772805e-05, "loss": 0.0097, "step": 41530 }, { "grad_norm": 0.0823783203959465, "learning_rate": 6.099550821766148e-05, "loss": 0.0073, "step": 41540 }, { "grad_norm": 0.09715534746646881, "learning_rate": 6.097772448604736e-05, "loss": 0.0083, "step": 41550 }, { "grad_norm": 0.06846606731414795, "learning_rate": 6.095993929524957e-05, "loss": 0.0072, "step": 41560 }, { "grad_norm": 0.08812138438224792, "learning_rate": 6.0942152647632165e-05, "loss": 0.0069, "step": 41570 }, { "grad_norm": 0.07780857384204865, "learning_rate": 6.0924364545559377e-05, "loss": 0.0059, "step": 41580 }, { "grad_norm": 0.0901888906955719, "learning_rate": 6.0906574991395625e-05, "loss": 0.0056, "step": 41590 }, { "grad_norm": 0.05141626298427582, "learning_rate": 6.088878398750556e-05, "loss": 0.008, "step": 41600 }, { "grad_norm": 0.061260152608156204, "learning_rate": 6.087099153625399e-05, "loss": 0.0073, "step": 41610 }, { "grad_norm": 0.09358756989240646, "learning_rate": 6.0853197640005946e-05, "loss": 0.0107, "step": 41620 }, { "grad_norm": 0.05122482031583786, "learning_rate": 6.0835402301126585e-05, "loss": 0.0057, "step": 41630 }, { "grad_norm": 0.06992552429437637, "learning_rate": 6.081760552198135e-05, "loss": 0.0083, "step": 41640 }, { "grad_norm": 0.05943657085299492, "learning_rate": 6.079980730493582e-05, "loss": 0.0085, "step": 41650 }, { "grad_norm": 0.07352367043495178, "learning_rate": 6.0782007652355754e-05, "loss": 0.0087, "step": 41660 }, { "grad_norm": 0.04971829429268837, "learning_rate": 6.076420656660714e-05, "loss": 0.0062, "step": 41670 }, { "grad_norm": 0.04065997153520584, "learning_rate": 6.0746404050056126e-05, "loss": 0.0059, "step": 41680 }, { "grad_norm": 0.08489320427179337, "learning_rate": 6.072860010506908e-05, "loss": 0.0115, "step": 41690 }, { "grad_norm": 0.06705520302057266, "learning_rate": 6.071079473401252e-05, "loss": 0.0053, "step": 41700 }, { "grad_norm": 0.06686214357614517, "learning_rate": 6.0692987939253186e-05, "loss": 0.0073, "step": 41710 }, { "grad_norm": 0.13373760879039764, "learning_rate": 6.0675179723158005e-05, "loss": 0.0083, "step": 41720 }, { "grad_norm": 0.15203188359737396, "learning_rate": 6.065737008809409e-05, "loss": 0.0097, "step": 41730 }, { "grad_norm": 0.11881786584854126, "learning_rate": 6.0639559036428694e-05, "loss": 0.0076, "step": 41740 }, { "grad_norm": 0.11210290342569351, "learning_rate": 6.062174657052935e-05, "loss": 0.0069, "step": 41750 }, { "grad_norm": 0.08898791670799255, "learning_rate": 6.060393269276372e-05, "loss": 0.0057, "step": 41760 }, { "grad_norm": 0.10272186994552612, "learning_rate": 6.058611740549964e-05, "loss": 0.0076, "step": 41770 }, { "grad_norm": 0.08426837623119354, "learning_rate": 6.056830071110519e-05, "loss": 0.0087, "step": 41780 }, { "grad_norm": 0.04252663627266884, "learning_rate": 6.0550482611948575e-05, "loss": 0.0067, "step": 41790 }, { "grad_norm": 0.09138588607311249, "learning_rate": 6.0532663110398245e-05, "loss": 0.0093, "step": 41800 }, { "grad_norm": 0.104544498026371, "learning_rate": 6.05148422088228e-05, "loss": 0.0091, "step": 41810 }, { "grad_norm": 0.07473570108413696, "learning_rate": 6.049701990959101e-05, "loss": 0.0077, "step": 41820 }, { "grad_norm": 0.0574052520096302, "learning_rate": 6.047919621507189e-05, "loss": 0.0091, "step": 41830 }, { "grad_norm": 0.061968736350536346, "learning_rate": 6.0461371127634595e-05, "loss": 0.0076, "step": 41840 }, { "grad_norm": 0.08835636079311371, "learning_rate": 6.044354464964845e-05, "loss": 0.007, "step": 41850 }, { "grad_norm": 0.09153417497873306, "learning_rate": 6.042571678348302e-05, "loss": 0.0076, "step": 41860 }, { "grad_norm": 0.08793171495199203, "learning_rate": 6.0407887531508024e-05, "loss": 0.0115, "step": 41870 }, { "grad_norm": 0.07957662642002106, "learning_rate": 6.0390056896093335e-05, "loss": 0.0069, "step": 41880 }, { "grad_norm": 0.08281532675027847, "learning_rate": 6.0372224879609086e-05, "loss": 0.0078, "step": 41890 }, { "grad_norm": 0.06638901680707932, "learning_rate": 6.035439148442551e-05, "loss": 0.0097, "step": 41900 }, { "grad_norm": 0.0706615075469017, "learning_rate": 6.033655671291307e-05, "loss": 0.0062, "step": 41910 }, { "grad_norm": 0.05792837217450142, "learning_rate": 6.031872056744243e-05, "loss": 0.0091, "step": 41920 }, { "grad_norm": 0.07089108228683472, "learning_rate": 6.0300883050384383e-05, "loss": 0.0094, "step": 41930 }, { "grad_norm": 0.07674936205148697, "learning_rate": 6.0283044164109936e-05, "loss": 0.0069, "step": 41940 }, { "grad_norm": 0.10409490764141083, "learning_rate": 6.0265203910990286e-05, "loss": 0.0085, "step": 41950 }, { "grad_norm": 0.09649091958999634, "learning_rate": 6.024736229339678e-05, "loss": 0.0076, "step": 41960 }, { "grad_norm": 0.08757033199071884, "learning_rate": 6.022951931370098e-05, "loss": 0.0072, "step": 41970 }, { "grad_norm": 0.09848757088184357, "learning_rate": 6.021167497427462e-05, "loss": 0.0097, "step": 41980 }, { "grad_norm": 0.09130680561065674, "learning_rate": 6.019382927748959e-05, "loss": 0.007, "step": 41990 }, { "grad_norm": 0.06836891174316406, "learning_rate": 6.017598222571799e-05, "loss": 0.0071, "step": 42000 }, { "grad_norm": 0.14062954485416412, "learning_rate": 6.01581338213321e-05, "loss": 0.0093, "step": 42010 }, { "grad_norm": 0.13874205946922302, "learning_rate": 6.0140284066704346e-05, "loss": 0.0094, "step": 42020 }, { "grad_norm": 0.06210605427622795, "learning_rate": 6.0122432964207385e-05, "loss": 0.0091, "step": 42030 }, { "grad_norm": 0.07623313367366791, "learning_rate": 6.010458051621399e-05, "loss": 0.0089, "step": 42040 }, { "grad_norm": 0.06701653450727463, "learning_rate": 6.008672672509719e-05, "loss": 0.0074, "step": 42050 }, { "grad_norm": 0.043332021683454514, "learning_rate": 6.0068871593230134e-05, "loss": 0.008, "step": 42060 }, { "grad_norm": 0.11675823479890823, "learning_rate": 6.0051015122986154e-05, "loss": 0.011, "step": 42070 }, { "grad_norm": 0.05891294777393341, "learning_rate": 6.003315731673878e-05, "loss": 0.0072, "step": 42080 }, { "grad_norm": 0.12938247621059418, "learning_rate": 6.001529817686172e-05, "loss": 0.0088, "step": 42090 }, { "grad_norm": 0.05249287188053131, "learning_rate": 5.9997437705728835e-05, "loss": 0.006, "step": 42100 }, { "grad_norm": 0.057633403688669205, "learning_rate": 5.997957590571419e-05, "loss": 0.0068, "step": 42110 }, { "grad_norm": 0.08139112591743469, "learning_rate": 5.996171277919202e-05, "loss": 0.0082, "step": 42120 }, { "grad_norm": 0.057471275329589844, "learning_rate": 5.99438483285367e-05, "loss": 0.0053, "step": 42130 }, { "grad_norm": 0.08266280591487885, "learning_rate": 5.9925982556122876e-05, "loss": 0.0075, "step": 42140 }, { "grad_norm": 0.08432383835315704, "learning_rate": 5.990811546432522e-05, "loss": 0.0094, "step": 42150 }, { "grad_norm": 0.07217147201299667, "learning_rate": 5.989024705551875e-05, "loss": 0.0065, "step": 42160 }, { "grad_norm": 0.06188526749610901, "learning_rate": 5.9872377332078525e-05, "loss": 0.0078, "step": 42170 }, { "grad_norm": 0.07689252495765686, "learning_rate": 5.985450629637984e-05, "loss": 0.0065, "step": 42180 }, { "grad_norm": 0.12097175419330597, "learning_rate": 5.983663395079816e-05, "loss": 0.0065, "step": 42190 }, { "grad_norm": 0.09234974533319473, "learning_rate": 5.98187602977091e-05, "loss": 0.008, "step": 42200 }, { "grad_norm": 0.06896070390939713, "learning_rate": 5.9800885339488485e-05, "loss": 0.0067, "step": 42210 }, { "grad_norm": 0.05912899225950241, "learning_rate": 5.978300907851229e-05, "loss": 0.0056, "step": 42220 }, { "grad_norm": 0.0752287283539772, "learning_rate": 5.976513151715666e-05, "loss": 0.0048, "step": 42230 }, { "grad_norm": 0.061542633920907974, "learning_rate": 5.9747252657797925e-05, "loss": 0.0094, "step": 42240 }, { "grad_norm": 0.1226009950041771, "learning_rate": 5.972937250281259e-05, "loss": 0.0094, "step": 42250 }, { "grad_norm": 0.0854891762137413, "learning_rate": 5.9711491054577295e-05, "loss": 0.0069, "step": 42260 }, { "grad_norm": 0.08926590532064438, "learning_rate": 5.969360831546892e-05, "loss": 0.0059, "step": 42270 }, { "grad_norm": 0.06884216517210007, "learning_rate": 5.967572428786447e-05, "loss": 0.0079, "step": 42280 }, { "grad_norm": 0.08646894991397858, "learning_rate": 5.9657838974141124e-05, "loss": 0.0081, "step": 42290 }, { "grad_norm": 0.10457495599985123, "learning_rate": 5.9639952376676236e-05, "loss": 0.0077, "step": 42300 }, { "grad_norm": 0.0529123917222023, "learning_rate": 5.9622064497847316e-05, "loss": 0.0072, "step": 42310 }, { "grad_norm": 0.07770746201276779, "learning_rate": 5.96041753400321e-05, "loss": 0.0062, "step": 42320 }, { "grad_norm": 0.07101620733737946, "learning_rate": 5.958628490560841e-05, "loss": 0.0075, "step": 42330 }, { "grad_norm": 0.0949445366859436, "learning_rate": 5.956839319695433e-05, "loss": 0.0068, "step": 42340 }, { "grad_norm": 0.09134974330663681, "learning_rate": 5.955050021644803e-05, "loss": 0.006, "step": 42350 }, { "grad_norm": 0.11317643523216248, "learning_rate": 5.9532605966467913e-05, "loss": 0.009, "step": 42360 }, { "grad_norm": 0.13561087846755981, "learning_rate": 5.951471044939247e-05, "loss": 0.0064, "step": 42370 }, { "grad_norm": 0.08567222207784653, "learning_rate": 5.949681366760048e-05, "loss": 0.0095, "step": 42380 }, { "grad_norm": 0.05301831662654877, "learning_rate": 5.9478915623470796e-05, "loss": 0.0077, "step": 42390 }, { "grad_norm": 0.054736871272325516, "learning_rate": 5.9461016319382456e-05, "loss": 0.0056, "step": 42400 }, { "grad_norm": 0.03214066103100777, "learning_rate": 5.944311575771467e-05, "loss": 0.0071, "step": 42410 }, { "grad_norm": 0.08935530483722687, "learning_rate": 5.9425213940846844e-05, "loss": 0.0081, "step": 42420 }, { "grad_norm": 0.09596665948629379, "learning_rate": 5.9407310871158516e-05, "loss": 0.0059, "step": 42430 }, { "grad_norm": 0.08683944493532181, "learning_rate": 5.9389406551029404e-05, "loss": 0.0085, "step": 42440 }, { "grad_norm": 0.06696340441703796, "learning_rate": 5.937150098283938e-05, "loss": 0.0071, "step": 42450 }, { "grad_norm": 0.1216723620891571, "learning_rate": 5.935359416896851e-05, "loss": 0.0096, "step": 42460 }, { "grad_norm": 0.05949382483959198, "learning_rate": 5.933568611179701e-05, "loss": 0.0069, "step": 42470 }, { "grad_norm": 0.0780639797449112, "learning_rate": 5.9317776813705214e-05, "loss": 0.0094, "step": 42480 }, { "grad_norm": 0.07347501814365387, "learning_rate": 5.9299866277073725e-05, "loss": 0.0084, "step": 42490 }, { "grad_norm": 0.07546699047088623, "learning_rate": 5.9281954504283234e-05, "loss": 0.0061, "step": 42500 }, { "grad_norm": 0.09888351708650589, "learning_rate": 5.926404149771459e-05, "loss": 0.0075, "step": 42510 }, { "grad_norm": 0.055201541632413864, "learning_rate": 5.924612725974885e-05, "loss": 0.0056, "step": 42520 }, { "grad_norm": 0.09381360560655594, "learning_rate": 5.9228211792767205e-05, "loss": 0.0078, "step": 42530 }, { "grad_norm": 0.05895679444074631, "learning_rate": 5.921029509915104e-05, "loss": 0.0054, "step": 42540 }, { "grad_norm": 0.07456561177968979, "learning_rate": 5.919237718128185e-05, "loss": 0.0056, "step": 42550 }, { "grad_norm": 0.081369549036026, "learning_rate": 5.9174458041541334e-05, "loss": 0.0084, "step": 42560 }, { "grad_norm": 0.07190432399511337, "learning_rate": 5.9156537682311375e-05, "loss": 0.0069, "step": 42570 }, { "grad_norm": 0.06023877486586571, "learning_rate": 5.913861610597395e-05, "loss": 0.0066, "step": 42580 }, { "grad_norm": 0.049697600305080414, "learning_rate": 5.9120693314911256e-05, "loss": 0.0074, "step": 42590 }, { "grad_norm": 0.05934084579348564, "learning_rate": 5.910276931150561e-05, "loss": 0.0064, "step": 42600 }, { "grad_norm": 0.059603020548820496, "learning_rate": 5.9084844098139535e-05, "loss": 0.0091, "step": 42610 }, { "grad_norm": 0.08092786371707916, "learning_rate": 5.9066917677195674e-05, "loss": 0.0084, "step": 42620 }, { "grad_norm": 0.05472004413604736, "learning_rate": 5.9048990051056865e-05, "loss": 0.0061, "step": 42630 }, { "grad_norm": 0.06522544473409653, "learning_rate": 5.903106122210606e-05, "loss": 0.0083, "step": 42640 }, { "grad_norm": 0.07881634682416916, "learning_rate": 5.9013131192726413e-05, "loss": 0.0091, "step": 42650 }, { "grad_norm": 0.10510013997554779, "learning_rate": 5.899519996530124e-05, "loss": 0.0068, "step": 42660 }, { "grad_norm": 0.08524639904499054, "learning_rate": 5.897726754221396e-05, "loss": 0.0078, "step": 42670 }, { "grad_norm": 0.08070056885480881, "learning_rate": 5.8959333925848225e-05, "loss": 0.0074, "step": 42680 }, { "grad_norm": 0.10112465173006058, "learning_rate": 5.894139911858781e-05, "loss": 0.0086, "step": 42690 }, { "grad_norm": 0.15502259135246277, "learning_rate": 5.892346312281664e-05, "loss": 0.0103, "step": 42700 }, { "grad_norm": 0.09911628067493439, "learning_rate": 5.8905525940918806e-05, "loss": 0.0069, "step": 42710 }, { "grad_norm": 0.09914828836917877, "learning_rate": 5.8887587575278546e-05, "loss": 0.0101, "step": 42720 }, { "grad_norm": 0.069398894906044, "learning_rate": 5.8869648028280297e-05, "loss": 0.0058, "step": 42730 }, { "grad_norm": 0.0588625892996788, "learning_rate": 5.885170730230861e-05, "loss": 0.0067, "step": 42740 }, { "grad_norm": 0.12645864486694336, "learning_rate": 5.88337653997482e-05, "loss": 0.0076, "step": 42750 }, { "grad_norm": 0.09223634749650955, "learning_rate": 5.881582232298395e-05, "loss": 0.0064, "step": 42760 }, { "grad_norm": 0.09204288572072983, "learning_rate": 5.8797878074400916e-05, "loss": 0.0069, "step": 42770 }, { "grad_norm": 0.09113964438438416, "learning_rate": 5.877993265638424e-05, "loss": 0.0059, "step": 42780 }, { "grad_norm": 0.09155179560184479, "learning_rate": 5.87619860713193e-05, "loss": 0.0089, "step": 42790 }, { "grad_norm": 0.06420192867517471, "learning_rate": 5.87440383215916e-05, "loss": 0.0072, "step": 42800 }, { "grad_norm": 0.06713675707578659, "learning_rate": 5.872608940958677e-05, "loss": 0.0058, "step": 42810 }, { "grad_norm": 0.07673566788434982, "learning_rate": 5.870813933769064e-05, "loss": 0.0093, "step": 42820 }, { "grad_norm": 0.06794346868991852, "learning_rate": 5.869018810828917e-05, "loss": 0.0072, "step": 42830 }, { "grad_norm": 0.07227909564971924, "learning_rate": 5.8672235723768464e-05, "loss": 0.0075, "step": 42840 }, { "grad_norm": 0.06290467083454132, "learning_rate": 5.865428218651481e-05, "loss": 0.0064, "step": 42850 }, { "grad_norm": 0.05903439223766327, "learning_rate": 5.863632749891463e-05, "loss": 0.0083, "step": 42860 }, { "grad_norm": 0.0814918577671051, "learning_rate": 5.86183716633545e-05, "loss": 0.0067, "step": 42870 }, { "grad_norm": 0.0982724130153656, "learning_rate": 5.860041468222115e-05, "loss": 0.0079, "step": 42880 }, { "grad_norm": 0.09564139693975449, "learning_rate": 5.8582456557901446e-05, "loss": 0.0063, "step": 42890 }, { "grad_norm": 0.06327648460865021, "learning_rate": 5.856449729278244e-05, "loss": 0.0069, "step": 42900 }, { "grad_norm": 0.09591991454362869, "learning_rate": 5.854653688925131e-05, "loss": 0.0056, "step": 42910 }, { "grad_norm": 0.07841532677412033, "learning_rate": 5.852857534969542e-05, "loss": 0.0088, "step": 42920 }, { "grad_norm": 0.0919947400689125, "learning_rate": 5.851061267650221e-05, "loss": 0.0066, "step": 42930 }, { "grad_norm": 0.07332465052604675, "learning_rate": 5.849264887205934e-05, "loss": 0.012, "step": 42940 }, { "grad_norm": 0.06039239838719368, "learning_rate": 5.847468393875461e-05, "loss": 0.0057, "step": 42950 }, { "grad_norm": 0.0684371292591095, "learning_rate": 5.845671787897596e-05, "loss": 0.0067, "step": 42960 }, { "grad_norm": 0.08476302027702332, "learning_rate": 5.8438750695111446e-05, "loss": 0.0079, "step": 42970 }, { "grad_norm": 0.08335918188095093, "learning_rate": 5.842078238954935e-05, "loss": 0.006, "step": 42980 }, { "grad_norm": 0.09605667740106583, "learning_rate": 5.8402812964678047e-05, "loss": 0.0059, "step": 42990 }, { "grad_norm": 0.06080355495214462, "learning_rate": 5.8384842422886036e-05, "loss": 0.0063, "step": 43000 }, { "grad_norm": 0.09038732945919037, "learning_rate": 5.836687076656204e-05, "loss": 0.0082, "step": 43010 }, { "grad_norm": 0.0765174850821495, "learning_rate": 5.8348897998094895e-05, "loss": 0.0084, "step": 43020 }, { "grad_norm": 0.08133544027805328, "learning_rate": 5.833092411987357e-05, "loss": 0.0075, "step": 43030 }, { "grad_norm": 0.07033193111419678, "learning_rate": 5.831294913428719e-05, "loss": 0.0104, "step": 43040 }, { "grad_norm": 0.045616790652275085, "learning_rate": 5.829497304372502e-05, "loss": 0.0062, "step": 43050 }, { "grad_norm": 0.08916758745908737, "learning_rate": 5.827699585057651e-05, "loss": 0.0077, "step": 43060 }, { "grad_norm": 0.08247289806604385, "learning_rate": 5.825901755723121e-05, "loss": 0.0078, "step": 43070 }, { "grad_norm": 0.054228972643613815, "learning_rate": 5.824103816607882e-05, "loss": 0.006, "step": 43080 }, { "grad_norm": 0.07236592471599579, "learning_rate": 5.822305767950924e-05, "loss": 0.0077, "step": 43090 }, { "grad_norm": 0.06961770355701447, "learning_rate": 5.820507609991246e-05, "loss": 0.0058, "step": 43100 }, { "grad_norm": 0.0801418125629425, "learning_rate": 5.818709342967861e-05, "loss": 0.0071, "step": 43110 }, { "grad_norm": 0.08986333757638931, "learning_rate": 5.8169109671198e-05, "loss": 0.0087, "step": 43120 }, { "grad_norm": 0.04526587203145027, "learning_rate": 5.815112482686109e-05, "loss": 0.009, "step": 43130 }, { "grad_norm": 0.06586103141307831, "learning_rate": 5.813313889905846e-05, "loss": 0.0055, "step": 43140 }, { "grad_norm": 0.07273788750171661, "learning_rate": 5.811515189018081e-05, "loss": 0.0077, "step": 43150 }, { "grad_norm": 0.06665375828742981, "learning_rate": 5.809716380261905e-05, "loss": 0.0071, "step": 43160 }, { "grad_norm": 0.06055194512009621, "learning_rate": 5.807917463876417e-05, "loss": 0.0099, "step": 43170 }, { "grad_norm": 0.05862284079194069, "learning_rate": 5.8061184401007354e-05, "loss": 0.0056, "step": 43180 }, { "grad_norm": 0.07747390121221542, "learning_rate": 5.804319309173987e-05, "loss": 0.0056, "step": 43190 }, { "grad_norm": 0.06730396300554276, "learning_rate": 5.802520071335319e-05, "loss": 0.0069, "step": 43200 }, { "grad_norm": 0.0463830940425396, "learning_rate": 5.800720726823892e-05, "loss": 0.0058, "step": 43210 }, { "grad_norm": 0.12302117049694061, "learning_rate": 5.7989212758788746e-05, "loss": 0.0097, "step": 43220 }, { "grad_norm": 0.10685357451438904, "learning_rate": 5.797121718739456e-05, "loss": 0.0088, "step": 43230 }, { "grad_norm": 0.07918243110179901, "learning_rate": 5.795322055644836e-05, "loss": 0.0057, "step": 43240 }, { "grad_norm": 0.0871962383389473, "learning_rate": 5.793522286834232e-05, "loss": 0.008, "step": 43250 }, { "grad_norm": 0.09530888497829437, "learning_rate": 5.7917224125468714e-05, "loss": 0.007, "step": 43260 }, { "grad_norm": 0.054652806371450424, "learning_rate": 5.789922433021999e-05, "loss": 0.0066, "step": 43270 }, { "grad_norm": 0.09622877091169357, "learning_rate": 5.788122348498871e-05, "loss": 0.0084, "step": 43280 }, { "grad_norm": 0.10026924312114716, "learning_rate": 5.786322159216761e-05, "loss": 0.007, "step": 43290 }, { "grad_norm": 0.09877081960439682, "learning_rate": 5.7845218654149494e-05, "loss": 0.0073, "step": 43300 }, { "grad_norm": 0.06120322272181511, "learning_rate": 5.782721467332739e-05, "loss": 0.0065, "step": 43310 }, { "grad_norm": 0.07388298958539963, "learning_rate": 5.7809209652094433e-05, "loss": 0.0058, "step": 43320 }, { "grad_norm": 0.08105607330799103, "learning_rate": 5.779120359284387e-05, "loss": 0.0063, "step": 43330 }, { "grad_norm": 0.06470894813537598, "learning_rate": 5.7773196497969105e-05, "loss": 0.0073, "step": 43340 }, { "grad_norm": 0.08171232789754868, "learning_rate": 5.775518836986369e-05, "loss": 0.0076, "step": 43350 }, { "grad_norm": 0.10652138292789459, "learning_rate": 5.7737179210921296e-05, "loss": 0.0066, "step": 43360 }, { "grad_norm": 0.07341723144054413, "learning_rate": 5.771916902353577e-05, "loss": 0.0067, "step": 43370 }, { "grad_norm": 0.042394641786813736, "learning_rate": 5.7701157810101035e-05, "loss": 0.0054, "step": 43380 }, { "grad_norm": 0.06394042074680328, "learning_rate": 5.768314557301119e-05, "loss": 0.0087, "step": 43390 }, { "grad_norm": 0.10420826822519302, "learning_rate": 5.766513231466049e-05, "loss": 0.0088, "step": 43400 }, { "grad_norm": 0.07775327563285828, "learning_rate": 5.764711803744323e-05, "loss": 0.0058, "step": 43410 }, { "grad_norm": 0.07640038430690765, "learning_rate": 5.762910274375397e-05, "loss": 0.0067, "step": 43420 }, { "grad_norm": 0.06640774756669998, "learning_rate": 5.7611086435987336e-05, "loss": 0.0089, "step": 43430 }, { "grad_norm": 0.06040383130311966, "learning_rate": 5.759306911653807e-05, "loss": 0.0097, "step": 43440 }, { "grad_norm": 0.12160859256982803, "learning_rate": 5.7575050787801086e-05, "loss": 0.0075, "step": 43450 }, { "grad_norm": 0.14518818259239197, "learning_rate": 5.7557031452171425e-05, "loss": 0.0079, "step": 43460 }, { "grad_norm": 0.07202260941267014, "learning_rate": 5.753901111204425e-05, "loss": 0.0062, "step": 43470 }, { "grad_norm": 0.06380145251750946, "learning_rate": 5.752098976981488e-05, "loss": 0.007, "step": 43480 }, { "grad_norm": 0.07042112201452255, "learning_rate": 5.750296742787873e-05, "loss": 0.01, "step": 43490 }, { "grad_norm": 0.0940849632024765, "learning_rate": 5.7484944088631376e-05, "loss": 0.0073, "step": 43500 }, { "grad_norm": 0.09288869798183441, "learning_rate": 5.746691975446855e-05, "loss": 0.007, "step": 43510 }, { "grad_norm": 0.08716415613889694, "learning_rate": 5.7448894427786036e-05, "loss": 0.0095, "step": 43520 }, { "grad_norm": 0.07519975304603577, "learning_rate": 5.743086811097984e-05, "loss": 0.0062, "step": 43530 }, { "grad_norm": 0.08172541856765747, "learning_rate": 5.741284080644606e-05, "loss": 0.0068, "step": 43540 }, { "grad_norm": 0.07034952938556671, "learning_rate": 5.7394812516580886e-05, "loss": 0.009, "step": 43550 }, { "grad_norm": 0.10685840994119644, "learning_rate": 5.737678324378072e-05, "loss": 0.0092, "step": 43560 }, { "grad_norm": 0.09277951717376709, "learning_rate": 5.7358752990442035e-05, "loss": 0.0051, "step": 43570 }, { "grad_norm": 0.06342945247888565, "learning_rate": 5.734072175896146e-05, "loss": 0.0055, "step": 43580 }, { "grad_norm": 0.05704028531908989, "learning_rate": 5.7322689551735756e-05, "loss": 0.0062, "step": 43590 }, { "grad_norm": 0.05479081720113754, "learning_rate": 5.730465637116176e-05, "loss": 0.0076, "step": 43600 }, { "grad_norm": 0.06140510365366936, "learning_rate": 5.728662221963653e-05, "loss": 0.0081, "step": 43610 }, { "grad_norm": 0.07815416902303696, "learning_rate": 5.7268587099557204e-05, "loss": 0.007, "step": 43620 }, { "grad_norm": 0.04763088375329971, "learning_rate": 5.725055101332102e-05, "loss": 0.0064, "step": 43630 }, { "grad_norm": 0.07200782001018524, "learning_rate": 5.7232513963325394e-05, "loss": 0.0075, "step": 43640 }, { "grad_norm": 0.14621977508068085, "learning_rate": 5.721447595196785e-05, "loss": 0.0077, "step": 43650 }, { "grad_norm": 0.07184924185276031, "learning_rate": 5.719643698164604e-05, "loss": 0.0068, "step": 43660 }, { "grad_norm": 0.08337552100419998, "learning_rate": 5.717839705475774e-05, "loss": 0.0058, "step": 43670 }, { "grad_norm": 0.0987178161740303, "learning_rate": 5.716035617370086e-05, "loss": 0.0089, "step": 43680 }, { "grad_norm": 0.07281497120857239, "learning_rate": 5.714231434087343e-05, "loss": 0.0083, "step": 43690 }, { "grad_norm": 0.0887969583272934, "learning_rate": 5.712427155867364e-05, "loss": 0.0085, "step": 43700 }, { "grad_norm": 0.0488411970436573, "learning_rate": 5.710622782949972e-05, "loss": 0.0077, "step": 43710 }, { "grad_norm": 0.0654890239238739, "learning_rate": 5.708818315575012e-05, "loss": 0.0071, "step": 43720 }, { "grad_norm": 0.07044011354446411, "learning_rate": 5.7070137539823397e-05, "loss": 0.0094, "step": 43730 }, { "grad_norm": 0.0583866685628891, "learning_rate": 5.705209098411818e-05, "loss": 0.008, "step": 43740 }, { "grad_norm": 0.11008156836032867, "learning_rate": 5.703404349103326e-05, "loss": 0.0071, "step": 43750 }, { "grad_norm": 0.10886643081903458, "learning_rate": 5.7015995062967566e-05, "loss": 0.0075, "step": 43760 }, { "grad_norm": 0.07359769195318222, "learning_rate": 5.699794570232012e-05, "loss": 0.0076, "step": 43770 }, { "grad_norm": 0.1195511445403099, "learning_rate": 5.69798954114901e-05, "loss": 0.008, "step": 43780 }, { "grad_norm": 0.06369908899068832, "learning_rate": 5.6961844192876776e-05, "loss": 0.0056, "step": 43790 }, { "grad_norm": 0.09878435730934143, "learning_rate": 5.6943792048879574e-05, "loss": 0.0068, "step": 43800 }, { "grad_norm": 0.0412568636238575, "learning_rate": 5.6925738981898014e-05, "loss": 0.0062, "step": 43810 }, { "grad_norm": 0.08828923106193542, "learning_rate": 5.690768499433173e-05, "loss": 0.0092, "step": 43820 }, { "grad_norm": 0.09206803143024445, "learning_rate": 5.688963008858054e-05, "loss": 0.0056, "step": 43830 }, { "grad_norm": 0.06349874287843704, "learning_rate": 5.687157426704432e-05, "loss": 0.0068, "step": 43840 }, { "grad_norm": 0.07326716929674149, "learning_rate": 5.685351753212309e-05, "loss": 0.0088, "step": 43850 }, { "grad_norm": 0.08277294039726257, "learning_rate": 5.683545988621698e-05, "loss": 0.0101, "step": 43860 }, { "grad_norm": 0.08420851081609726, "learning_rate": 5.681740133172627e-05, "loss": 0.009, "step": 43870 }, { "grad_norm": 0.06562846899032593, "learning_rate": 5.679934187105135e-05, "loss": 0.0065, "step": 43880 }, { "grad_norm": 0.05404626950621605, "learning_rate": 5.67812815065927e-05, "loss": 0.0077, "step": 43890 }, { "grad_norm": 0.058166395872831345, "learning_rate": 5.6763220240750955e-05, "loss": 0.0069, "step": 43900 }, { "grad_norm": 0.08066453784704208, "learning_rate": 5.674515807592687e-05, "loss": 0.0054, "step": 43910 }, { "grad_norm": 0.09096395969390869, "learning_rate": 5.672709501452132e-05, "loss": 0.0072, "step": 43920 }, { "grad_norm": 0.07335983961820602, "learning_rate": 5.6709031058935224e-05, "loss": 0.0083, "step": 43930 }, { "grad_norm": 0.06345116347074509, "learning_rate": 5.669096621156974e-05, "loss": 0.0074, "step": 43940 }, { "grad_norm": 0.08164843171834946, "learning_rate": 5.667290047482609e-05, "loss": 0.0083, "step": 43950 }, { "grad_norm": 0.09232509881258011, "learning_rate": 5.6654833851105585e-05, "loss": 0.0062, "step": 43960 }, { "grad_norm": 0.06272619217634201, "learning_rate": 5.663676634280969e-05, "loss": 0.0073, "step": 43970 }, { "grad_norm": 0.08639220893383026, "learning_rate": 5.661869795233997e-05, "loss": 0.0071, "step": 43980 }, { "grad_norm": 0.08234505355358124, "learning_rate": 5.660062868209815e-05, "loss": 0.0092, "step": 43990 }, { "grad_norm": 0.11751789599657059, "learning_rate": 5.6582558534486e-05, "loss": 0.0063, "step": 44000 }, { "grad_norm": 0.08646419644355774, "learning_rate": 5.656448751190544e-05, "loss": 0.0049, "step": 44010 }, { "grad_norm": 0.09333472698926926, "learning_rate": 5.6546415616758556e-05, "loss": 0.0096, "step": 44020 }, { "grad_norm": 0.10310626775026321, "learning_rate": 5.652834285144748e-05, "loss": 0.008, "step": 44030 }, { "grad_norm": 0.06916351616382599, "learning_rate": 5.651026921837447e-05, "loss": 0.0081, "step": 44040 }, { "grad_norm": 0.09180818498134613, "learning_rate": 5.649219471994191e-05, "loss": 0.0083, "step": 44050 }, { "grad_norm": 0.08583766222000122, "learning_rate": 5.647411935855235e-05, "loss": 0.0067, "step": 44060 }, { "grad_norm": 0.10489469021558762, "learning_rate": 5.645604313660836e-05, "loss": 0.008, "step": 44070 }, { "grad_norm": 0.07417765259742737, "learning_rate": 5.643796605651268e-05, "loss": 0.007, "step": 44080 }, { "grad_norm": 0.11182787269353867, "learning_rate": 5.641988812066818e-05, "loss": 0.0074, "step": 44090 }, { "grad_norm": 0.07986921072006226, "learning_rate": 5.6401809331477796e-05, "loss": 0.0071, "step": 44100 }, { "grad_norm": 0.07069890201091766, "learning_rate": 5.638372969134461e-05, "loss": 0.0053, "step": 44110 }, { "grad_norm": 0.06265095621347427, "learning_rate": 5.63656492026718e-05, "loss": 0.0048, "step": 44120 }, { "grad_norm": 0.1287991851568222, "learning_rate": 5.634756786786267e-05, "loss": 0.0073, "step": 44130 }, { "grad_norm": 0.11866484582424164, "learning_rate": 5.6329485689320646e-05, "loss": 0.0087, "step": 44140 }, { "grad_norm": 0.05505963787436485, "learning_rate": 5.6311402669449234e-05, "loss": 0.0075, "step": 44150 }, { "grad_norm": 0.07607721537351608, "learning_rate": 5.629331881065208e-05, "loss": 0.007, "step": 44160 }, { "grad_norm": 0.09753919392824173, "learning_rate": 5.6275234115332907e-05, "loss": 0.0066, "step": 44170 }, { "grad_norm": 0.04909829795360565, "learning_rate": 5.62571485858956e-05, "loss": 0.0075, "step": 44180 }, { "grad_norm": 0.08309068530797958, "learning_rate": 5.6239062224744124e-05, "loss": 0.0084, "step": 44190 }, { "grad_norm": 0.05066640302538872, "learning_rate": 5.622097503428254e-05, "loss": 0.0049, "step": 44200 }, { "grad_norm": 0.08416607975959778, "learning_rate": 5.620288701691506e-05, "loss": 0.0064, "step": 44210 }, { "grad_norm": 0.0954299196600914, "learning_rate": 5.6184798175045985e-05, "loss": 0.0083, "step": 44220 }, { "grad_norm": 0.09826643764972687, "learning_rate": 5.616670851107969e-05, "loss": 0.0053, "step": 44230 }, { "grad_norm": 0.07301640510559082, "learning_rate": 5.614861802742073e-05, "loss": 0.0064, "step": 44240 }, { "grad_norm": 0.08466791361570358, "learning_rate": 5.6130526726473743e-05, "loss": 0.0084, "step": 44250 }, { "grad_norm": 0.057349082082509995, "learning_rate": 5.6112434610643414e-05, "loss": 0.0087, "step": 44260 }, { "grad_norm": 0.06273962557315826, "learning_rate": 5.609434168233463e-05, "loss": 0.0076, "step": 44270 }, { "grad_norm": 0.08876007050275803, "learning_rate": 5.607624794395232e-05, "loss": 0.0095, "step": 44280 }, { "grad_norm": 0.04642558842897415, "learning_rate": 5.605815339790155e-05, "loss": 0.0069, "step": 44290 }, { "grad_norm": 0.08369201421737671, "learning_rate": 5.6040058046587494e-05, "loss": 0.0073, "step": 44300 }, { "grad_norm": 0.06699608266353607, "learning_rate": 5.6021961892415434e-05, "loss": 0.01, "step": 44310 }, { "grad_norm": 0.06983835995197296, "learning_rate": 5.600386493779073e-05, "loss": 0.0073, "step": 44320 }, { "grad_norm": 0.09906043857336044, "learning_rate": 5.5985767185118886e-05, "loss": 0.0076, "step": 44330 }, { "grad_norm": 0.056340355426073074, "learning_rate": 5.5967668636805484e-05, "loss": 0.0079, "step": 44340 }, { "grad_norm": 0.058759402483701706, "learning_rate": 5.594956929525623e-05, "loss": 0.0063, "step": 44350 }, { "grad_norm": 0.10714942961931229, "learning_rate": 5.593146916287695e-05, "loss": 0.0087, "step": 44360 }, { "grad_norm": 0.08484596014022827, "learning_rate": 5.5913368242073503e-05, "loss": 0.006, "step": 44370 }, { "grad_norm": 0.04218662902712822, "learning_rate": 5.589526653525196e-05, "loss": 0.0071, "step": 44380 }, { "grad_norm": 0.04299992695450783, "learning_rate": 5.5877164044818386e-05, "loss": 0.0071, "step": 44390 }, { "grad_norm": 0.06859147548675537, "learning_rate": 5.585906077317905e-05, "loss": 0.0081, "step": 44400 }, { "grad_norm": 0.07817967981100082, "learning_rate": 5.584095672274027e-05, "loss": 0.0072, "step": 44410 }, { "grad_norm": 0.04708757624030113, "learning_rate": 5.5822851895908446e-05, "loss": 0.0109, "step": 44420 }, { "grad_norm": 0.08556722104549408, "learning_rate": 5.5804746295090135e-05, "loss": 0.0084, "step": 44430 }, { "grad_norm": 0.04249926283955574, "learning_rate": 5.5786639922691996e-05, "loss": 0.0066, "step": 44440 }, { "grad_norm": 0.08026977628469467, "learning_rate": 5.576853278112072e-05, "loss": 0.0059, "step": 44450 }, { "grad_norm": 0.13581345975399017, "learning_rate": 5.575042487278319e-05, "loss": 0.0081, "step": 44460 }, { "grad_norm": 0.09814068675041199, "learning_rate": 5.5732316200086345e-05, "loss": 0.0086, "step": 44470 }, { "grad_norm": 0.07825446128845215, "learning_rate": 5.5714206765437205e-05, "loss": 0.0059, "step": 44480 }, { "grad_norm": 0.0827651396393776, "learning_rate": 5.569609657124293e-05, "loss": 0.0082, "step": 44490 }, { "grad_norm": 0.0665382519364357, "learning_rate": 5.567798561991077e-05, "loss": 0.0071, "step": 44500 }, { "grad_norm": 0.06012656167149544, "learning_rate": 5.565987391384807e-05, "loss": 0.0052, "step": 44510 }, { "grad_norm": 0.0921282097697258, "learning_rate": 5.5641761455462294e-05, "loss": 0.006, "step": 44520 }, { "grad_norm": 0.0835333988070488, "learning_rate": 5.562364824716095e-05, "loss": 0.0088, "step": 44530 }, { "grad_norm": 0.07157327234745026, "learning_rate": 5.5605534291351735e-05, "loss": 0.0071, "step": 44540 }, { "grad_norm": 0.06736703217029572, "learning_rate": 5.558741959044238e-05, "loss": 0.0081, "step": 44550 }, { "grad_norm": 0.08904670178890228, "learning_rate": 5.5569304146840704e-05, "loss": 0.0084, "step": 44560 }, { "grad_norm": 0.05402239039540291, "learning_rate": 5.555118796295467e-05, "loss": 0.0047, "step": 44570 }, { "grad_norm": 0.07156667113304138, "learning_rate": 5.5533071041192363e-05, "loss": 0.007, "step": 44580 }, { "grad_norm": 0.05897459015250206, "learning_rate": 5.5514953383961855e-05, "loss": 0.0139, "step": 44590 }, { "grad_norm": 0.06397632509469986, "learning_rate": 5.549683499367143e-05, "loss": 0.005, "step": 44600 }, { "grad_norm": 0.05402546375989914, "learning_rate": 5.54787158727294e-05, "loss": 0.0067, "step": 44610 }, { "grad_norm": 0.0623779259622097, "learning_rate": 5.5460596023544234e-05, "loss": 0.0086, "step": 44620 }, { "grad_norm": 0.04373196139931679, "learning_rate": 5.544247544852443e-05, "loss": 0.0051, "step": 44630 }, { "grad_norm": 0.08280989527702332, "learning_rate": 5.542435415007862e-05, "loss": 0.0089, "step": 44640 }, { "grad_norm": 0.12160422652959824, "learning_rate": 5.540623213061553e-05, "loss": 0.0102, "step": 44650 }, { "grad_norm": 0.08365751057863235, "learning_rate": 5.5388109392544e-05, "loss": 0.0072, "step": 44660 }, { "grad_norm": 0.0804172232747078, "learning_rate": 5.5369985938272916e-05, "loss": 0.0079, "step": 44670 }, { "grad_norm": 0.07729905098676682, "learning_rate": 5.53518617702113e-05, "loss": 0.0076, "step": 44680 }, { "grad_norm": 0.11327142268419266, "learning_rate": 5.533373689076826e-05, "loss": 0.0074, "step": 44690 }, { "grad_norm": 0.0958595722913742, "learning_rate": 5.531561130235299e-05, "loss": 0.0111, "step": 44700 }, { "grad_norm": 0.05859139934182167, "learning_rate": 5.529748500737478e-05, "loss": 0.0061, "step": 44710 }, { "grad_norm": 0.0722694844007492, "learning_rate": 5.527935800824304e-05, "loss": 0.0058, "step": 44720 }, { "grad_norm": 0.09421131014823914, "learning_rate": 5.526123030736722e-05, "loss": 0.0067, "step": 44730 }, { "grad_norm": 0.11405275762081146, "learning_rate": 5.524310190715693e-05, "loss": 0.0093, "step": 44740 }, { "grad_norm": 0.05439465865492821, "learning_rate": 5.522497281002178e-05, "loss": 0.0081, "step": 44750 }, { "grad_norm": 0.08710590749979019, "learning_rate": 5.520684301837159e-05, "loss": 0.0065, "step": 44760 }, { "grad_norm": 0.12539248168468475, "learning_rate": 5.51887125346162e-05, "loss": 0.0068, "step": 44770 }, { "grad_norm": 0.04220633953809738, "learning_rate": 5.517058136116553e-05, "loss": 0.0099, "step": 44780 }, { "grad_norm": 0.06439028680324554, "learning_rate": 5.5152449500429625e-05, "loss": 0.0076, "step": 44790 }, { "grad_norm": 0.07618479430675507, "learning_rate": 5.5134316954818624e-05, "loss": 0.0079, "step": 44800 }, { "grad_norm": 0.09060002118349075, "learning_rate": 5.511618372674274e-05, "loss": 0.0078, "step": 44810 }, { "grad_norm": 0.054964881390333176, "learning_rate": 5.5098049818612275e-05, "loss": 0.0126, "step": 44820 }, { "grad_norm": 0.08535881340503693, "learning_rate": 5.507991523283764e-05, "loss": 0.0107, "step": 44830 }, { "grad_norm": 0.0766395702958107, "learning_rate": 5.506177997182931e-05, "loss": 0.0068, "step": 44840 }, { "grad_norm": 0.07975893467664719, "learning_rate": 5.50436440379979e-05, "loss": 0.0073, "step": 44850 }, { "grad_norm": 0.06867361813783646, "learning_rate": 5.502550743375402e-05, "loss": 0.0068, "step": 44860 }, { "grad_norm": 0.07419664412736893, "learning_rate": 5.5007370161508464e-05, "loss": 0.01, "step": 44870 }, { "grad_norm": 0.06107497215270996, "learning_rate": 5.4989232223672094e-05, "loss": 0.0071, "step": 44880 }, { "grad_norm": 0.05459928512573242, "learning_rate": 5.4971093622655824e-05, "loss": 0.0099, "step": 44890 }, { "grad_norm": 0.07943646609783173, "learning_rate": 5.495295436087067e-05, "loss": 0.0079, "step": 44900 }, { "grad_norm": 0.0449024997651577, "learning_rate": 5.4934814440727756e-05, "loss": 0.0092, "step": 44910 }, { "grad_norm": 0.055077649652957916, "learning_rate": 5.491667386463828e-05, "loss": 0.0093, "step": 44920 }, { "grad_norm": 0.058858927339315414, "learning_rate": 5.489853263501353e-05, "loss": 0.0058, "step": 44930 }, { "grad_norm": 0.04954277351498604, "learning_rate": 5.488039075426485e-05, "loss": 0.0067, "step": 44940 }, { "grad_norm": 0.0622282400727272, "learning_rate": 5.4862248224803746e-05, "loss": 0.0062, "step": 44950 }, { "grad_norm": 0.03871217370033264, "learning_rate": 5.484410504904175e-05, "loss": 0.0059, "step": 44960 }, { "grad_norm": 0.07634010910987854, "learning_rate": 5.4825961229390454e-05, "loss": 0.0041, "step": 44970 }, { "grad_norm": 0.10668034106492996, "learning_rate": 5.480781676826162e-05, "loss": 0.0049, "step": 44980 }, { "grad_norm": 0.09192630648612976, "learning_rate": 5.478967166806706e-05, "loss": 0.0075, "step": 44990 }, { "grad_norm": 0.07453491538763046, "learning_rate": 5.4771525931218614e-05, "loss": 0.0065, "step": 45000 }, { "grad_norm": 0.07302483916282654, "learning_rate": 5.4753379560128284e-05, "loss": 0.0056, "step": 45010 }, { "grad_norm": 0.07665961235761642, "learning_rate": 5.4735232557208114e-05, "loss": 0.0077, "step": 45020 }, { "grad_norm": 0.08014876395463943, "learning_rate": 5.471708492487024e-05, "loss": 0.0094, "step": 45030 }, { "grad_norm": 0.12767058610916138, "learning_rate": 5.469893666552692e-05, "loss": 0.0081, "step": 45040 }, { "grad_norm": 0.073882095515728, "learning_rate": 5.4680787781590414e-05, "loss": 0.0052, "step": 45050 }, { "grad_norm": 0.05582574009895325, "learning_rate": 5.466263827547316e-05, "loss": 0.0051, "step": 45060 }, { "grad_norm": 0.09921327233314514, "learning_rate": 5.464448814958759e-05, "loss": 0.0087, "step": 45070 }, { "grad_norm": 0.12063699960708618, "learning_rate": 5.4626337406346286e-05, "loss": 0.0058, "step": 45080 }, { "grad_norm": 0.0713353231549263, "learning_rate": 5.460818604816186e-05, "loss": 0.0106, "step": 45090 }, { "grad_norm": 0.06832390278577805, "learning_rate": 5.459003407744705e-05, "loss": 0.0073, "step": 45100 }, { "grad_norm": 0.08535032719373703, "learning_rate": 5.457188149661465e-05, "loss": 0.0062, "step": 45110 }, { "grad_norm": 0.048872459679841995, "learning_rate": 5.455372830807755e-05, "loss": 0.0061, "step": 45120 }, { "grad_norm": 0.04969144985079765, "learning_rate": 5.45355745142487e-05, "loss": 0.0081, "step": 45130 }, { "grad_norm": 0.0687323808670044, "learning_rate": 5.451742011754115e-05, "loss": 0.0082, "step": 45140 }, { "grad_norm": 0.08407674729824066, "learning_rate": 5.4499265120368036e-05, "loss": 0.0076, "step": 45150 }, { "grad_norm": 0.049767106771469116, "learning_rate": 5.4481109525142525e-05, "loss": 0.0083, "step": 45160 }, { "grad_norm": 0.08880241215229034, "learning_rate": 5.4462953334277936e-05, "loss": 0.0073, "step": 45170 }, { "grad_norm": 0.07758323848247528, "learning_rate": 5.4444796550187615e-05, "loss": 0.0068, "step": 45180 }, { "grad_norm": 0.059157613664865494, "learning_rate": 5.4426639175284996e-05, "loss": 0.0066, "step": 45190 }, { "grad_norm": 0.06789611279964447, "learning_rate": 5.440848121198361e-05, "loss": 0.0057, "step": 45200 }, { "grad_norm": 0.0938388779759407, "learning_rate": 5.439032266269704e-05, "loss": 0.0126, "step": 45210 }, { "grad_norm": 0.09078191220760345, "learning_rate": 5.437216352983898e-05, "loss": 0.0066, "step": 45220 }, { "grad_norm": 0.08939811587333679, "learning_rate": 5.435400381582316e-05, "loss": 0.0067, "step": 45230 }, { "grad_norm": 0.0857631042599678, "learning_rate": 5.433584352306343e-05, "loss": 0.009, "step": 45240 }, { "grad_norm": 0.07633711397647858, "learning_rate": 5.431768265397369e-05, "loss": 0.0089, "step": 45250 }, { "grad_norm": 0.06087524816393852, "learning_rate": 5.429952121096793e-05, "loss": 0.0065, "step": 45260 }, { "grad_norm": 0.09974120557308197, "learning_rate": 5.428135919646019e-05, "loss": 0.0072, "step": 45270 }, { "grad_norm": 0.07971850037574768, "learning_rate": 5.426319661286462e-05, "loss": 0.0072, "step": 45280 }, { "grad_norm": 0.08882735669612885, "learning_rate": 5.4245033462595426e-05, "loss": 0.0089, "step": 45290 }, { "grad_norm": 0.08849997073411942, "learning_rate": 5.422686974806691e-05, "loss": 0.009, "step": 45300 }, { "grad_norm": 0.08750153332948685, "learning_rate": 5.420870547169341e-05, "loss": 0.0081, "step": 45310 }, { "grad_norm": 0.0634915754199028, "learning_rate": 5.419054063588936e-05, "loss": 0.0074, "step": 45320 }, { "grad_norm": 0.06265734136104584, "learning_rate": 5.4172375243069306e-05, "loss": 0.0122, "step": 45330 }, { "grad_norm": 0.049216967076063156, "learning_rate": 5.415420929564779e-05, "loss": 0.006, "step": 45340 }, { "grad_norm": 0.07449370622634888, "learning_rate": 5.413604279603951e-05, "loss": 0.0057, "step": 45350 }, { "grad_norm": 0.07246586680412292, "learning_rate": 5.411787574665916e-05, "loss": 0.008, "step": 45360 }, { "grad_norm": 0.11843593418598175, "learning_rate": 5.409970814992158e-05, "loss": 0.0073, "step": 45370 }, { "grad_norm": 0.08683083206415176, "learning_rate": 5.4081540008241614e-05, "loss": 0.0086, "step": 45380 }, { "grad_norm": 0.07332879304885864, "learning_rate": 5.406337132403424e-05, "loss": 0.007, "step": 45390 }, { "grad_norm": 0.07945165783166885, "learning_rate": 5.404520209971447e-05, "loss": 0.0079, "step": 45400 }, { "grad_norm": 0.13395369052886963, "learning_rate": 5.402703233769739e-05, "loss": 0.011, "step": 45410 }, { "grad_norm": 0.09089073538780212, "learning_rate": 5.4008862040398166e-05, "loss": 0.0085, "step": 45420 }, { "grad_norm": 0.07328876107931137, "learning_rate": 5.3990691210232046e-05, "loss": 0.0079, "step": 45430 }, { "grad_norm": 0.07989934086799622, "learning_rate": 5.3972519849614335e-05, "loss": 0.0075, "step": 45440 }, { "grad_norm": 0.07092548161745071, "learning_rate": 5.395434796096042e-05, "loss": 0.0075, "step": 45450 }, { "grad_norm": 0.07042798399925232, "learning_rate": 5.393617554668571e-05, "loss": 0.0057, "step": 45460 }, { "grad_norm": 0.062427818775177, "learning_rate": 5.3918002609205754e-05, "loss": 0.0079, "step": 45470 }, { "grad_norm": 0.047296129167079926, "learning_rate": 5.3899829150936156e-05, "loss": 0.0081, "step": 45480 }, { "grad_norm": 0.04410512000322342, "learning_rate": 5.388165517429252e-05, "loss": 0.0063, "step": 45490 }, { "grad_norm": 0.1077074408531189, "learning_rate": 5.3863480681690614e-05, "loss": 0.0088, "step": 45500 }, { "grad_norm": 0.10116493701934814, "learning_rate": 5.384530567554624e-05, "loss": 0.0086, "step": 45510 }, { "grad_norm": 0.1216929480433464, "learning_rate": 5.382713015827523e-05, "loss": 0.0066, "step": 45520 }, { "grad_norm": 0.06994609534740448, "learning_rate": 5.380895413229352e-05, "loss": 0.0087, "step": 45530 }, { "grad_norm": 0.07504040002822876, "learning_rate": 5.379077760001712e-05, "loss": 0.0076, "step": 45540 }, { "grad_norm": 0.07522204518318176, "learning_rate": 5.377260056386209e-05, "loss": 0.0083, "step": 45550 }, { "grad_norm": 0.088833287358284, "learning_rate": 5.375442302624457e-05, "loss": 0.0054, "step": 45560 }, { "grad_norm": 0.08827938139438629, "learning_rate": 5.373624498958073e-05, "loss": 0.0071, "step": 45570 }, { "grad_norm": 0.08198053389787674, "learning_rate": 5.371806645628687e-05, "loss": 0.0079, "step": 45580 }, { "grad_norm": 0.06254351884126663, "learning_rate": 5.369988742877931e-05, "loss": 0.0053, "step": 45590 }, { "grad_norm": 0.10107381641864777, "learning_rate": 5.368170790947443e-05, "loss": 0.0065, "step": 45600 }, { "grad_norm": 0.09212932735681534, "learning_rate": 5.3663527900788713e-05, "loss": 0.0076, "step": 45610 }, { "grad_norm": 0.05995596572756767, "learning_rate": 5.364534740513869e-05, "loss": 0.0076, "step": 45620 }, { "grad_norm": 0.06154199317097664, "learning_rate": 5.362716642494093e-05, "loss": 0.0066, "step": 45630 }, { "grad_norm": 0.06533452868461609, "learning_rate": 5.360898496261211e-05, "loss": 0.0039, "step": 45640 }, { "grad_norm": 0.043316688388586044, "learning_rate": 5.359080302056894e-05, "loss": 0.0038, "step": 45650 }, { "grad_norm": 0.06534332782030106, "learning_rate": 5.35726206012282e-05, "loss": 0.0054, "step": 45660 }, { "grad_norm": 0.10881175100803375, "learning_rate": 5.3554437707006764e-05, "loss": 0.007, "step": 45670 }, { "grad_norm": 0.13021445274353027, "learning_rate": 5.35362543403215e-05, "loss": 0.0061, "step": 45680 }, { "grad_norm": 0.13208727538585663, "learning_rate": 5.3518070503589415e-05, "loss": 0.0067, "step": 45690 }, { "grad_norm": 0.04773696884512901, "learning_rate": 5.349988619922756e-05, "loss": 0.0062, "step": 45700 }, { "grad_norm": 0.06112554669380188, "learning_rate": 5.3481701429652986e-05, "loss": 0.0066, "step": 45710 }, { "grad_norm": 0.04846980795264244, "learning_rate": 5.346351619728287e-05, "loss": 0.0058, "step": 45720 }, { "grad_norm": 0.07899535447359085, "learning_rate": 5.3445330504534453e-05, "loss": 0.007, "step": 45730 }, { "grad_norm": 0.10554506629705429, "learning_rate": 5.3427144353825e-05, "loss": 0.0082, "step": 45740 }, { "grad_norm": 0.08212502300739288, "learning_rate": 5.3408957747571864e-05, "loss": 0.008, "step": 45750 }, { "grad_norm": 0.07218405604362488, "learning_rate": 5.339077068819245e-05, "loss": 0.0073, "step": 45760 }, { "grad_norm": 0.05166017264127731, "learning_rate": 5.3372583178104206e-05, "loss": 0.0069, "step": 45770 }, { "grad_norm": 0.07508896291255951, "learning_rate": 5.335439521972468e-05, "loss": 0.0066, "step": 45780 }, { "grad_norm": 0.06714817881584167, "learning_rate": 5.333620681547143e-05, "loss": 0.0072, "step": 45790 }, { "grad_norm": 0.05795477330684662, "learning_rate": 5.3318017967762125e-05, "loss": 0.0076, "step": 45800 }, { "grad_norm": 0.062057893723249435, "learning_rate": 5.3299828679014466e-05, "loss": 0.0077, "step": 45810 }, { "grad_norm": 0.0689549446105957, "learning_rate": 5.32816389516462e-05, "loss": 0.0064, "step": 45820 }, { "grad_norm": 0.0899970754981041, "learning_rate": 5.3263448788075145e-05, "loss": 0.0066, "step": 45830 }, { "grad_norm": 0.06496608257293701, "learning_rate": 5.3245258190719173e-05, "loss": 0.0099, "step": 45840 }, { "grad_norm": 0.08469516038894653, "learning_rate": 5.322706716199624e-05, "loss": 0.0075, "step": 45850 }, { "grad_norm": 0.06920498609542847, "learning_rate": 5.320887570432434e-05, "loss": 0.0104, "step": 45860 }, { "grad_norm": 0.05356007069349289, "learning_rate": 5.3190683820121476e-05, "loss": 0.0096, "step": 45870 }, { "grad_norm": 0.08703769743442535, "learning_rate": 5.317249151180581e-05, "loss": 0.0061, "step": 45880 }, { "grad_norm": 0.0998724177479744, "learning_rate": 5.3154298781795494e-05, "loss": 0.0058, "step": 45890 }, { "grad_norm": 0.04662032425403595, "learning_rate": 5.3136105632508716e-05, "loss": 0.006, "step": 45900 }, { "grad_norm": 0.052724193781614304, "learning_rate": 5.311791206636376e-05, "loss": 0.0058, "step": 45910 }, { "grad_norm": 0.055797990411520004, "learning_rate": 5.309971808577899e-05, "loss": 0.0073, "step": 45920 }, { "grad_norm": 0.07074816524982452, "learning_rate": 5.308152369317275e-05, "loss": 0.0052, "step": 45930 }, { "grad_norm": 0.0525236539542675, "learning_rate": 5.3063328890963505e-05, "loss": 0.0067, "step": 45940 }, { "grad_norm": 0.1407896727323532, "learning_rate": 5.3045133681569725e-05, "loss": 0.0069, "step": 45950 }, { "grad_norm": 0.07238501310348511, "learning_rate": 5.302693806740997e-05, "loss": 0.0078, "step": 45960 }, { "grad_norm": 0.05688853561878204, "learning_rate": 5.300874205090286e-05, "loss": 0.0082, "step": 45970 }, { "grad_norm": 0.04786833003163338, "learning_rate": 5.2990545634467004e-05, "loss": 0.007, "step": 45980 }, { "grad_norm": 0.0532427541911602, "learning_rate": 5.2972348820521154e-05, "loss": 0.007, "step": 45990 }, { "grad_norm": 0.07630398124456406, "learning_rate": 5.295415161148406e-05, "loss": 0.0057, "step": 46000 }, { "grad_norm": 0.051101140677928925, "learning_rate": 5.293595400977451e-05, "loss": 0.0068, "step": 46010 }, { "grad_norm": 0.09637555480003357, "learning_rate": 5.291775601781141e-05, "loss": 0.0069, "step": 46020 }, { "grad_norm": 0.043244294822216034, "learning_rate": 5.2899557638013676e-05, "loss": 0.0079, "step": 46030 }, { "grad_norm": 0.06160846725106239, "learning_rate": 5.2881358872800235e-05, "loss": 0.0054, "step": 46040 }, { "grad_norm": 0.08185255527496338, "learning_rate": 5.286315972459014e-05, "loss": 0.0074, "step": 46050 }, { "grad_norm": 0.06526622921228409, "learning_rate": 5.2844960195802464e-05, "loss": 0.008, "step": 46060 }, { "grad_norm": 0.0778493583202362, "learning_rate": 5.2826760288856305e-05, "loss": 0.0076, "step": 46070 }, { "grad_norm": 0.09746330976486206, "learning_rate": 5.280856000617087e-05, "loss": 0.0084, "step": 46080 }, { "grad_norm": 0.06166834384202957, "learning_rate": 5.279035935016534e-05, "loss": 0.0079, "step": 46090 }, { "grad_norm": 0.0732831284403801, "learning_rate": 5.277215832325901e-05, "loss": 0.0084, "step": 46100 }, { "grad_norm": 0.06325434148311615, "learning_rate": 5.275395692787123e-05, "loss": 0.0075, "step": 46110 }, { "grad_norm": 0.09442085027694702, "learning_rate": 5.273575516642131e-05, "loss": 0.0065, "step": 46120 }, { "grad_norm": 0.059984780848026276, "learning_rate": 5.271755304132871e-05, "loss": 0.0059, "step": 46130 }, { "grad_norm": 0.06268497556447983, "learning_rate": 5.2699350555012886e-05, "loss": 0.0069, "step": 46140 }, { "grad_norm": 0.09144871681928635, "learning_rate": 5.2681147709893354e-05, "loss": 0.0066, "step": 46150 }, { "grad_norm": 0.12983213365077972, "learning_rate": 5.2662944508389663e-05, "loss": 0.01, "step": 46160 }, { "grad_norm": 0.06357091665267944, "learning_rate": 5.2644740952921455e-05, "loss": 0.0098, "step": 46170 }, { "grad_norm": 0.0808517336845398, "learning_rate": 5.262653704590836e-05, "loss": 0.0053, "step": 46180 }, { "grad_norm": 0.12502433359622955, "learning_rate": 5.260833278977011e-05, "loss": 0.0061, "step": 46190 }, { "grad_norm": 0.0893416777253151, "learning_rate": 5.25901281869264e-05, "loss": 0.0097, "step": 46200 }, { "grad_norm": 0.08990603685379028, "learning_rate": 5.257192323979708e-05, "loss": 0.009, "step": 46210 }, { "grad_norm": 0.05455229803919792, "learning_rate": 5.2553717950801996e-05, "loss": 0.006, "step": 46220 }, { "grad_norm": 0.0754256471991539, "learning_rate": 5.253551232236099e-05, "loss": 0.0089, "step": 46230 }, { "grad_norm": 0.06807009875774384, "learning_rate": 5.251730635689403e-05, "loss": 0.0066, "step": 46240 }, { "grad_norm": 0.07772421836853027, "learning_rate": 5.249910005682107e-05, "loss": 0.0085, "step": 46250 }, { "grad_norm": 0.09422530233860016, "learning_rate": 5.248089342456216e-05, "loss": 0.005, "step": 46260 }, { "grad_norm": 0.08579711616039276, "learning_rate": 5.246268646253735e-05, "loss": 0.0062, "step": 46270 }, { "grad_norm": 0.0684296190738678, "learning_rate": 5.2444479173166736e-05, "loss": 0.0053, "step": 46280 }, { "grad_norm": 0.053761228919029236, "learning_rate": 5.24262715588705e-05, "loss": 0.0096, "step": 46290 }, { "grad_norm": 0.06729363650083542, "learning_rate": 5.240806362206884e-05, "loss": 0.0103, "step": 46300 }, { "grad_norm": 0.07120797783136368, "learning_rate": 5.238985536518195e-05, "loss": 0.0084, "step": 46310 }, { "grad_norm": 0.07547333091497421, "learning_rate": 5.237164679063016e-05, "loss": 0.0078, "step": 46320 }, { "grad_norm": 0.11186636239290237, "learning_rate": 5.235343790083378e-05, "loss": 0.0066, "step": 46330 }, { "grad_norm": 0.07382407784461975, "learning_rate": 5.233522869821318e-05, "loss": 0.0064, "step": 46340 }, { "grad_norm": 0.11074943840503693, "learning_rate": 5.231701918518875e-05, "loss": 0.0083, "step": 46350 }, { "grad_norm": 0.050464071333408356, "learning_rate": 5.2298809364180954e-05, "loss": 0.0065, "step": 46360 }, { "grad_norm": 0.07339891791343689, "learning_rate": 5.2280599237610285e-05, "loss": 0.0069, "step": 46370 }, { "grad_norm": 0.06591318547725677, "learning_rate": 5.226238880789728e-05, "loss": 0.007, "step": 46380 }, { "grad_norm": 0.06473727524280548, "learning_rate": 5.224417807746247e-05, "loss": 0.0077, "step": 46390 }, { "grad_norm": 0.07430321723222733, "learning_rate": 5.22259670487265e-05, "loss": 0.0048, "step": 46400 }, { "grad_norm": 0.09732505679130554, "learning_rate": 5.220775572411003e-05, "loss": 0.0057, "step": 46410 }, { "grad_norm": 0.04243217781186104, "learning_rate": 5.218954410603372e-05, "loss": 0.0056, "step": 46420 }, { "grad_norm": 0.06471579521894455, "learning_rate": 5.217133219691832e-05, "loss": 0.0058, "step": 46430 }, { "grad_norm": 0.06406912207603455, "learning_rate": 5.215311999918461e-05, "loss": 0.0084, "step": 46440 }, { "grad_norm": 0.05139555409550667, "learning_rate": 5.213490751525336e-05, "loss": 0.0056, "step": 46450 }, { "grad_norm": 0.07701889425516129, "learning_rate": 5.211669474754544e-05, "loss": 0.0068, "step": 46460 }, { "grad_norm": 0.08024778962135315, "learning_rate": 5.20984816984817e-05, "loss": 0.0057, "step": 46470 }, { "grad_norm": 0.06383541226387024, "learning_rate": 5.2080268370483106e-05, "loss": 0.0086, "step": 46480 }, { "grad_norm": 0.046704329550266266, "learning_rate": 5.206205476597059e-05, "loss": 0.0072, "step": 46490 }, { "grad_norm": 0.06776026636362076, "learning_rate": 5.204384088736513e-05, "loss": 0.0056, "step": 46500 }, { "grad_norm": 0.06618288904428482, "learning_rate": 5.202562673708778e-05, "loss": 0.0062, "step": 46510 }, { "grad_norm": 0.0945400521159172, "learning_rate": 5.200741231755961e-05, "loss": 0.0085, "step": 46520 }, { "grad_norm": 0.08350029587745667, "learning_rate": 5.1989197631201694e-05, "loss": 0.0071, "step": 46530 }, { "grad_norm": 0.048279114067554474, "learning_rate": 5.19709826804352e-05, "loss": 0.006, "step": 46540 }, { "grad_norm": 0.0904061421751976, "learning_rate": 5.1952767467681264e-05, "loss": 0.0077, "step": 46550 }, { "grad_norm": 0.08308377861976624, "learning_rate": 5.1934551995361124e-05, "loss": 0.0084, "step": 46560 }, { "grad_norm": 0.06687168031930923, "learning_rate": 5.1916336265896004e-05, "loss": 0.0067, "step": 46570 }, { "grad_norm": 0.06651753187179565, "learning_rate": 5.189812028170717e-05, "loss": 0.0073, "step": 46580 }, { "grad_norm": 0.06332309544086456, "learning_rate": 5.187990404521596e-05, "loss": 0.007, "step": 46590 }, { "grad_norm": 0.09681890904903412, "learning_rate": 5.186168755884372e-05, "loss": 0.0072, "step": 46600 }, { "grad_norm": 0.05648145079612732, "learning_rate": 5.184347082501179e-05, "loss": 0.009, "step": 46610 }, { "grad_norm": 0.07855778932571411, "learning_rate": 5.182525384614159e-05, "loss": 0.0059, "step": 46620 }, { "grad_norm": 0.03631838783621788, "learning_rate": 5.180703662465459e-05, "loss": 0.0063, "step": 46630 }, { "grad_norm": 0.06327839195728302, "learning_rate": 5.1788819162972236e-05, "loss": 0.0076, "step": 46640 }, { "grad_norm": 0.10629343241453171, "learning_rate": 5.177060146351603e-05, "loss": 0.0078, "step": 46650 }, { "grad_norm": 0.05413186922669411, "learning_rate": 5.175238352870754e-05, "loss": 0.0054, "step": 46660 }, { "grad_norm": 0.0833636149764061, "learning_rate": 5.173416536096829e-05, "loss": 0.0094, "step": 46670 }, { "grad_norm": 0.07486066222190857, "learning_rate": 5.171594696271992e-05, "loss": 0.0065, "step": 46680 }, { "grad_norm": 0.07257881760597229, "learning_rate": 5.1697728336384046e-05, "loss": 0.006, "step": 46690 }, { "grad_norm": 0.0715080127120018, "learning_rate": 5.167950948438233e-05, "loss": 0.0055, "step": 46700 }, { "grad_norm": 0.07058237493038177, "learning_rate": 5.1661290409136455e-05, "loss": 0.0073, "step": 46710 }, { "grad_norm": 0.08208327740430832, "learning_rate": 5.164307111306813e-05, "loss": 0.006, "step": 46720 }, { "grad_norm": 0.08158934116363525, "learning_rate": 5.162485159859913e-05, "loss": 0.0076, "step": 46730 }, { "grad_norm": 0.07017636299133301, "learning_rate": 5.160663186815125e-05, "loss": 0.0096, "step": 46740 }, { "grad_norm": 0.13881614804267883, "learning_rate": 5.158841192414625e-05, "loss": 0.0092, "step": 46750 }, { "grad_norm": 0.056882429867982864, "learning_rate": 5.1570191769005985e-05, "loss": 0.0056, "step": 46760 }, { "grad_norm": 0.09540653973817825, "learning_rate": 5.155197140515232e-05, "loss": 0.0079, "step": 46770 }, { "grad_norm": 0.09066738933324814, "learning_rate": 5.1533750835007154e-05, "loss": 0.0066, "step": 46780 }, { "grad_norm": 0.04008954018354416, "learning_rate": 5.1515530060992403e-05, "loss": 0.0052, "step": 46790 }, { "grad_norm": 0.04714065417647362, "learning_rate": 5.1497309085529996e-05, "loss": 0.0078, "step": 46800 }, { "grad_norm": 0.05967254191637039, "learning_rate": 5.1479087911041924e-05, "loss": 0.0085, "step": 46810 }, { "grad_norm": 0.0686473697423935, "learning_rate": 5.146086653995019e-05, "loss": 0.0055, "step": 46820 }, { "grad_norm": 0.07994519174098969, "learning_rate": 5.14426449746768e-05, "loss": 0.0114, "step": 46830 }, { "grad_norm": 0.08155888319015503, "learning_rate": 5.142442321764381e-05, "loss": 0.0067, "step": 46840 }, { "grad_norm": 0.13033483922481537, "learning_rate": 5.1406201271273325e-05, "loss": 0.0107, "step": 46850 }, { "grad_norm": 0.140716090798378, "learning_rate": 5.138797913798742e-05, "loss": 0.0073, "step": 46860 }, { "grad_norm": 0.03977000340819359, "learning_rate": 5.136975682020822e-05, "loss": 0.0076, "step": 46870 }, { "grad_norm": 0.04880116507411003, "learning_rate": 5.135153432035787e-05, "loss": 0.0081, "step": 46880 }, { "grad_norm": 0.056952208280563354, "learning_rate": 5.1333311640858564e-05, "loss": 0.0044, "step": 46890 }, { "grad_norm": 0.07086973637342453, "learning_rate": 5.131508878413252e-05, "loss": 0.0095, "step": 46900 }, { "grad_norm": 0.06824041903018951, "learning_rate": 5.12968657526019e-05, "loss": 0.0064, "step": 46910 }, { "grad_norm": 0.05822199210524559, "learning_rate": 5.127864254868899e-05, "loss": 0.0089, "step": 46920 }, { "grad_norm": 0.08994782716035843, "learning_rate": 5.1260419174816074e-05, "loss": 0.0078, "step": 46930 }, { "grad_norm": 0.0530967079102993, "learning_rate": 5.12421956334054e-05, "loss": 0.0112, "step": 46940 }, { "grad_norm": 0.06580893695354462, "learning_rate": 5.1223971926879325e-05, "loss": 0.0091, "step": 46950 }, { "grad_norm": 0.061484139412641525, "learning_rate": 5.120574805766018e-05, "loss": 0.0068, "step": 46960 }, { "grad_norm": 0.05725453048944473, "learning_rate": 5.1187524028170286e-05, "loss": 0.0059, "step": 46970 }, { "grad_norm": 0.07976574450731277, "learning_rate": 5.116929984083204e-05, "loss": 0.0052, "step": 46980 }, { "grad_norm": 0.06478923559188843, "learning_rate": 5.115107549806786e-05, "loss": 0.007, "step": 46990 }, { "grad_norm": 0.09229247272014618, "learning_rate": 5.113285100230014e-05, "loss": 0.008, "step": 47000 }, { "grad_norm": 0.04040129482746124, "learning_rate": 5.111462635595136e-05, "loss": 0.0058, "step": 47010 }, { "grad_norm": 0.06233051046729088, "learning_rate": 5.109640156144392e-05, "loss": 0.0091, "step": 47020 }, { "grad_norm": 0.08585204184055328, "learning_rate": 5.1078176621200346e-05, "loss": 0.008, "step": 47030 }, { "grad_norm": 0.09149965643882751, "learning_rate": 5.105995153764313e-05, "loss": 0.0086, "step": 47040 }, { "grad_norm": 0.08395494520664215, "learning_rate": 5.104172631319477e-05, "loss": 0.0087, "step": 47050 }, { "grad_norm": 0.08059188723564148, "learning_rate": 5.1023500950277844e-05, "loss": 0.0062, "step": 47060 }, { "grad_norm": 0.06436216086149216, "learning_rate": 5.1005275451314885e-05, "loss": 0.0087, "step": 47070 }, { "grad_norm": 0.0809556320309639, "learning_rate": 5.098704981872845e-05, "loss": 0.0105, "step": 47080 }, { "grad_norm": 0.08292308449745178, "learning_rate": 5.096882405494116e-05, "loss": 0.0111, "step": 47090 }, { "grad_norm": 0.05576145648956299, "learning_rate": 5.0950598162375593e-05, "loss": 0.0063, "step": 47100 }, { "grad_norm": 0.09429191797971725, "learning_rate": 5.093237214345441e-05, "loss": 0.0085, "step": 47110 }, { "grad_norm": 0.07221267372369766, "learning_rate": 5.091414600060025e-05, "loss": 0.0066, "step": 47120 }, { "grad_norm": 0.09672082960605621, "learning_rate": 5.089591973623573e-05, "loss": 0.0088, "step": 47130 }, { "grad_norm": 0.1110045313835144, "learning_rate": 5.087769335278356e-05, "loss": 0.0077, "step": 47140 }, { "grad_norm": 0.06544502079486847, "learning_rate": 5.085946685266646e-05, "loss": 0.0056, "step": 47150 }, { "grad_norm": 0.11739269644021988, "learning_rate": 5.0841240238307075e-05, "loss": 0.0079, "step": 47160 }, { "grad_norm": 0.06879210472106934, "learning_rate": 5.082301351212817e-05, "loss": 0.0055, "step": 47170 }, { "grad_norm": 0.09328487515449524, "learning_rate": 5.080478667655246e-05, "loss": 0.0062, "step": 47180 }, { "grad_norm": 0.10128811001777649, "learning_rate": 5.0786559734002716e-05, "loss": 0.0094, "step": 47190 }, { "grad_norm": 0.07732772827148438, "learning_rate": 5.076833268690169e-05, "loss": 0.0063, "step": 47200 }, { "grad_norm": 0.05390651524066925, "learning_rate": 5.075010553767217e-05, "loss": 0.0074, "step": 47210 }, { "grad_norm": 0.09569913148880005, "learning_rate": 5.073187828873693e-05, "loss": 0.0057, "step": 47220 }, { "grad_norm": 0.08457035571336746, "learning_rate": 5.071365094251882e-05, "loss": 0.0088, "step": 47230 }, { "grad_norm": 0.05404318496584892, "learning_rate": 5.0695423501440596e-05, "loss": 0.0059, "step": 47240 }, { "grad_norm": 0.08101148158311844, "learning_rate": 5.067719596792515e-05, "loss": 0.0084, "step": 47250 }, { "grad_norm": 0.10626942664384842, "learning_rate": 5.0658968344395305e-05, "loss": 0.007, "step": 47260 }, { "grad_norm": 0.047681063413619995, "learning_rate": 5.064074063327391e-05, "loss": 0.0055, "step": 47270 }, { "grad_norm": 0.04374793916940689, "learning_rate": 5.0622512836983834e-05, "loss": 0.0069, "step": 47280 }, { "grad_norm": 0.06582118570804596, "learning_rate": 5.060428495794797e-05, "loss": 0.0069, "step": 47290 }, { "grad_norm": 0.08406510949134827, "learning_rate": 5.0586056998589184e-05, "loss": 0.0082, "step": 47300 }, { "grad_norm": 0.05564846470952034, "learning_rate": 5.05678289613304e-05, "loss": 0.008, "step": 47310 }, { "grad_norm": 0.09198559075593948, "learning_rate": 5.0549600848594525e-05, "loss": 0.0108, "step": 47320 }, { "grad_norm": 0.094743512570858, "learning_rate": 5.0531372662804466e-05, "loss": 0.0072, "step": 47330 }, { "grad_norm": 0.0648484155535698, "learning_rate": 5.051314440638318e-05, "loss": 0.0078, "step": 47340 }, { "grad_norm": 0.05696447938680649, "learning_rate": 5.0494916081753574e-05, "loss": 0.0056, "step": 47350 }, { "grad_norm": 0.04812847077846527, "learning_rate": 5.047668769133863e-05, "loss": 0.0065, "step": 47360 }, { "grad_norm": 0.07198251038789749, "learning_rate": 5.045845923756129e-05, "loss": 0.0051, "step": 47370 }, { "grad_norm": 0.08430884778499603, "learning_rate": 5.044023072284453e-05, "loss": 0.0067, "step": 47380 }, { "grad_norm": 0.08047688007354736, "learning_rate": 5.042200214961131e-05, "loss": 0.0059, "step": 47390 }, { "grad_norm": 0.06910017132759094, "learning_rate": 5.040377352028463e-05, "loss": 0.0074, "step": 47400 }, { "grad_norm": 0.04910062253475189, "learning_rate": 5.0385544837287455e-05, "loss": 0.0058, "step": 47410 }, { "grad_norm": 0.08326208591461182, "learning_rate": 5.036731610304283e-05, "loss": 0.0084, "step": 47420 }, { "grad_norm": 0.09281016886234283, "learning_rate": 5.034908731997369e-05, "loss": 0.01, "step": 47430 }, { "grad_norm": 0.0699615478515625, "learning_rate": 5.03308584905031e-05, "loss": 0.0043, "step": 47440 }, { "grad_norm": 0.03438702970743179, "learning_rate": 5.0312629617054065e-05, "loss": 0.0065, "step": 47450 }, { "grad_norm": 0.10290224850177765, "learning_rate": 5.029440070204958e-05, "loss": 0.0087, "step": 47460 }, { "grad_norm": 0.09767445176839828, "learning_rate": 5.027617174791272e-05, "loss": 0.0076, "step": 47470 }, { "grad_norm": 0.10912852734327316, "learning_rate": 5.025794275706649e-05, "loss": 0.0076, "step": 47480 }, { "grad_norm": 0.07690226286649704, "learning_rate": 5.0239713731933924e-05, "loss": 0.0092, "step": 47490 }, { "grad_norm": 0.05475664511322975, "learning_rate": 5.022148467493808e-05, "loss": 0.0073, "step": 47500 }, { "grad_norm": 0.07096634805202484, "learning_rate": 5.0203255588501996e-05, "loss": 0.0098, "step": 47510 }, { "grad_norm": 0.05598174408078194, "learning_rate": 5.0185026475048724e-05, "loss": 0.0096, "step": 47520 }, { "grad_norm": 0.06435032188892365, "learning_rate": 5.016679733700133e-05, "loss": 0.0047, "step": 47530 }, { "grad_norm": 0.060181379318237305, "learning_rate": 5.014856817678284e-05, "loss": 0.0088, "step": 47540 }, { "grad_norm": 0.088088758289814, "learning_rate": 5.013033899681634e-05, "loss": 0.0082, "step": 47550 }, { "grad_norm": 0.07209912687540054, "learning_rate": 5.011210979952491e-05, "loss": 0.0063, "step": 47560 }, { "grad_norm": 0.07986940443515778, "learning_rate": 5.0093880587331565e-05, "loss": 0.0082, "step": 47570 }, { "grad_norm": 0.06514979153871536, "learning_rate": 5.007565136265944e-05, "loss": 0.0048, "step": 47580 }, { "grad_norm": 0.10741971433162689, "learning_rate": 5.005742212793154e-05, "loss": 0.0062, "step": 47590 }, { "grad_norm": 0.047756828367710114, "learning_rate": 5.003919288557096e-05, "loss": 0.007, "step": 47600 }, { "grad_norm": 0.06501179188489914, "learning_rate": 5.002096363800078e-05, "loss": 0.0079, "step": 47610 }, { "grad_norm": 0.05760550498962402, "learning_rate": 5.000273438764408e-05, "loss": 0.0074, "step": 47620 }, { "grad_norm": 0.062169820070266724, "learning_rate": 4.9984505136923895e-05, "loss": 0.0072, "step": 47630 }, { "grad_norm": 0.08490251004695892, "learning_rate": 4.996627588826335e-05, "loss": 0.0064, "step": 47640 }, { "grad_norm": 0.09446816891431808, "learning_rate": 4.9948046644085475e-05, "loss": 0.0089, "step": 47650 }, { "grad_norm": 0.09320757538080215, "learning_rate": 4.9929817406813354e-05, "loss": 0.0069, "step": 47660 }, { "grad_norm": 0.09176398068666458, "learning_rate": 4.991158817887005e-05, "loss": 0.0068, "step": 47670 }, { "grad_norm": 0.058269940316677094, "learning_rate": 4.989335896267865e-05, "loss": 0.0073, "step": 47680 }, { "grad_norm": 0.08215751498937607, "learning_rate": 4.987512976066222e-05, "loss": 0.0082, "step": 47690 }, { "grad_norm": 0.11050788313150406, "learning_rate": 4.9856900575243826e-05, "loss": 0.0098, "step": 47700 }, { "grad_norm": 0.046629901975393295, "learning_rate": 4.983867140884652e-05, "loss": 0.008, "step": 47710 }, { "grad_norm": 0.09439145028591156, "learning_rate": 4.9820442263893375e-05, "loss": 0.0068, "step": 47720 }, { "grad_norm": 0.06353418529033661, "learning_rate": 4.9802213142807456e-05, "loss": 0.0057, "step": 47730 }, { "grad_norm": 0.04935462400317192, "learning_rate": 4.978398404801178e-05, "loss": 0.0074, "step": 47740 }, { "grad_norm": 0.060318730771541595, "learning_rate": 4.9765754981929434e-05, "loss": 0.0055, "step": 47750 }, { "grad_norm": 0.06493780761957169, "learning_rate": 4.974752594698347e-05, "loss": 0.0063, "step": 47760 }, { "grad_norm": 0.06930487602949142, "learning_rate": 4.972929694559692e-05, "loss": 0.0065, "step": 47770 }, { "grad_norm": 0.07301840931177139, "learning_rate": 4.971106798019283e-05, "loss": 0.0068, "step": 47780 }, { "grad_norm": 0.08826743066310883, "learning_rate": 4.9692839053194224e-05, "loss": 0.0078, "step": 47790 }, { "grad_norm": 0.07718604803085327, "learning_rate": 4.967461016702414e-05, "loss": 0.0068, "step": 47800 }, { "grad_norm": 0.05543869733810425, "learning_rate": 4.9656381324105606e-05, "loss": 0.0067, "step": 47810 }, { "grad_norm": 0.07221885770559311, "learning_rate": 4.96381525268616e-05, "loss": 0.0066, "step": 47820 }, { "grad_norm": 0.07851896435022354, "learning_rate": 4.9619923777715205e-05, "loss": 0.0051, "step": 47830 }, { "grad_norm": 0.07415106892585754, "learning_rate": 4.960169507908938e-05, "loss": 0.0059, "step": 47840 }, { "grad_norm": 0.11416212469339371, "learning_rate": 4.958346643340713e-05, "loss": 0.0061, "step": 47850 }, { "grad_norm": 0.09502055495977402, "learning_rate": 4.956523784309147e-05, "loss": 0.0088, "step": 47860 }, { "grad_norm": 0.07035728543996811, "learning_rate": 4.954700931056536e-05, "loss": 0.007, "step": 47870 }, { "grad_norm": 0.11441236734390259, "learning_rate": 4.9528780838251775e-05, "loss": 0.0063, "step": 47880 }, { "grad_norm": 0.07492035627365112, "learning_rate": 4.951055242857369e-05, "loss": 0.0074, "step": 47890 }, { "grad_norm": 0.10732366889715195, "learning_rate": 4.949232408395407e-05, "loss": 0.0082, "step": 47900 }, { "grad_norm": 0.08185718208551407, "learning_rate": 4.947409580681589e-05, "loss": 0.009, "step": 47910 }, { "grad_norm": 0.08195389807224274, "learning_rate": 4.9455867599582055e-05, "loss": 0.0084, "step": 47920 }, { "grad_norm": 0.11290782690048218, "learning_rate": 4.9437639464675515e-05, "loss": 0.0134, "step": 47930 }, { "grad_norm": 0.07307815551757812, "learning_rate": 4.941941140451921e-05, "loss": 0.0066, "step": 47940 }, { "grad_norm": 0.059862807393074036, "learning_rate": 4.940118342153603e-05, "loss": 0.007, "step": 47950 }, { "grad_norm": 0.11351371556520462, "learning_rate": 4.9382955518148866e-05, "loss": 0.0072, "step": 47960 }, { "grad_norm": 0.0831717997789383, "learning_rate": 4.936472769678065e-05, "loss": 0.0081, "step": 47970 }, { "grad_norm": 0.0626196414232254, "learning_rate": 4.934649995985425e-05, "loss": 0.0055, "step": 47980 }, { "grad_norm": 0.048291128128767014, "learning_rate": 4.932827230979254e-05, "loss": 0.0055, "step": 47990 }, { "grad_norm": 0.08438920229673386, "learning_rate": 4.931004474901837e-05, "loss": 0.0069, "step": 48000 }, { "grad_norm": 0.10751702636480331, "learning_rate": 4.9291817279954595e-05, "loss": 0.0089, "step": 48010 }, { "grad_norm": 0.09116177260875702, "learning_rate": 4.927358990502405e-05, "loss": 0.0081, "step": 48020 }, { "grad_norm": 0.07772884517908096, "learning_rate": 4.925536262664956e-05, "loss": 0.007, "step": 48030 }, { "grad_norm": 0.0854448452591896, "learning_rate": 4.9237135447253914e-05, "loss": 0.0059, "step": 48040 }, { "grad_norm": 0.07259881496429443, "learning_rate": 4.9218908369259954e-05, "loss": 0.0069, "step": 48050 }, { "grad_norm": 0.05660081282258034, "learning_rate": 4.9200681395090434e-05, "loss": 0.0057, "step": 48060 }, { "grad_norm": 0.09695465117692947, "learning_rate": 4.9182454527168125e-05, "loss": 0.0068, "step": 48070 }, { "grad_norm": 0.03952109441161156, "learning_rate": 4.91642277679158e-05, "loss": 0.0053, "step": 48080 }, { "grad_norm": 0.08546946197748184, "learning_rate": 4.914600111975618e-05, "loss": 0.008, "step": 48090 }, { "grad_norm": 0.08077775686979294, "learning_rate": 4.912777458511199e-05, "loss": 0.0061, "step": 48100 }, { "grad_norm": 0.10642832517623901, "learning_rate": 4.910954816640595e-05, "loss": 0.0072, "step": 48110 }, { "grad_norm": 0.11705507338047028, "learning_rate": 4.909132186606076e-05, "loss": 0.0082, "step": 48120 }, { "grad_norm": 0.08575913310050964, "learning_rate": 4.907309568649912e-05, "loss": 0.0079, "step": 48130 }, { "grad_norm": 0.06015022099018097, "learning_rate": 4.9054869630143655e-05, "loss": 0.0064, "step": 48140 }, { "grad_norm": 0.06781186163425446, "learning_rate": 4.9036643699417034e-05, "loss": 0.0066, "step": 48150 }, { "grad_norm": 0.07460106909275055, "learning_rate": 4.90184178967419e-05, "loss": 0.0066, "step": 48160 }, { "grad_norm": 0.08967330306768417, "learning_rate": 4.900019222454083e-05, "loss": 0.0066, "step": 48170 }, { "grad_norm": 0.06075970083475113, "learning_rate": 4.8981966685236454e-05, "loss": 0.0066, "step": 48180 }, { "grad_norm": 0.0871950164437294, "learning_rate": 4.8963741281251326e-05, "loss": 0.0069, "step": 48190 }, { "grad_norm": 0.0780804306268692, "learning_rate": 4.894551601500804e-05, "loss": 0.0087, "step": 48200 }, { "grad_norm": 0.06523773819208145, "learning_rate": 4.892729088892912e-05, "loss": 0.0052, "step": 48210 }, { "grad_norm": 0.048262130469083786, "learning_rate": 4.8909065905437095e-05, "loss": 0.0107, "step": 48220 }, { "grad_norm": 0.08289351314306259, "learning_rate": 4.889084106695446e-05, "loss": 0.005, "step": 48230 }, { "grad_norm": 0.05783778429031372, "learning_rate": 4.8872616375903716e-05, "loss": 0.0069, "step": 48240 }, { "grad_norm": 0.06284831464290619, "learning_rate": 4.885439183470732e-05, "loss": 0.0055, "step": 48250 }, { "grad_norm": 0.08269534260034561, "learning_rate": 4.8836167445787716e-05, "loss": 0.0084, "step": 48260 }, { "grad_norm": 0.06161041930317879, "learning_rate": 4.881794321156734e-05, "loss": 0.0069, "step": 48270 }, { "grad_norm": 0.022421477362513542, "learning_rate": 4.87997191344686e-05, "loss": 0.0057, "step": 48280 }, { "grad_norm": 0.12054417282342911, "learning_rate": 4.878149521691386e-05, "loss": 0.0057, "step": 48290 }, { "grad_norm": 0.07514173537492752, "learning_rate": 4.8763271461325526e-05, "loss": 0.0061, "step": 48300 }, { "grad_norm": 0.07877231389284134, "learning_rate": 4.8745047870125895e-05, "loss": 0.0105, "step": 48310 }, { "grad_norm": 0.05832017585635185, "learning_rate": 4.872682444573731e-05, "loss": 0.0063, "step": 48320 }, { "grad_norm": 0.10981418192386627, "learning_rate": 4.8708601190582074e-05, "loss": 0.0089, "step": 48330 }, { "grad_norm": 0.1160883828997612, "learning_rate": 4.869037810708242e-05, "loss": 0.0076, "step": 48340 }, { "grad_norm": 0.12298493087291718, "learning_rate": 4.867215519766066e-05, "loss": 0.0069, "step": 48350 }, { "grad_norm": 0.09563469141721725, "learning_rate": 4.8653932464739e-05, "loss": 0.0079, "step": 48360 }, { "grad_norm": 0.09245789051055908, "learning_rate": 4.863570991073964e-05, "loss": 0.0045, "step": 48370 }, { "grad_norm": 0.09333565831184387, "learning_rate": 4.861748753808478e-05, "loss": 0.0058, "step": 48380 }, { "grad_norm": 0.08964531868696213, "learning_rate": 4.8599265349196554e-05, "loss": 0.0105, "step": 48390 }, { "grad_norm": 0.049682069569826126, "learning_rate": 4.858104334649711e-05, "loss": 0.0069, "step": 48400 }, { "grad_norm": 0.07920219749212265, "learning_rate": 4.856282153240854e-05, "loss": 0.007, "step": 48410 }, { "grad_norm": 0.08468768745660782, "learning_rate": 4.854459990935296e-05, "loss": 0.0071, "step": 48420 }, { "grad_norm": 0.07589757442474365, "learning_rate": 4.8526378479752424e-05, "loss": 0.0066, "step": 48430 }, { "grad_norm": 0.04152835160493851, "learning_rate": 4.850815724602894e-05, "loss": 0.0081, "step": 48440 }, { "grad_norm": 0.05839909240603447, "learning_rate": 4.8489936210604534e-05, "loss": 0.0064, "step": 48450 }, { "grad_norm": 0.06788447499275208, "learning_rate": 4.847171537590118e-05, "loss": 0.0056, "step": 48460 }, { "grad_norm": 0.04455859214067459, "learning_rate": 4.845349474434083e-05, "loss": 0.0062, "step": 48470 }, { "grad_norm": 0.049590058624744415, "learning_rate": 4.843527431834541e-05, "loss": 0.006, "step": 48480 }, { "grad_norm": 0.0534646101295948, "learning_rate": 4.841705410033683e-05, "loss": 0.0062, "step": 48490 }, { "grad_norm": 0.06701887398958206, "learning_rate": 4.8398834092736956e-05, "loss": 0.0069, "step": 48500 }, { "grad_norm": 0.04341571033000946, "learning_rate": 4.8380614297967645e-05, "loss": 0.0065, "step": 48510 }, { "grad_norm": 0.05627654120326042, "learning_rate": 4.8362394718450696e-05, "loss": 0.0058, "step": 48520 }, { "grad_norm": 0.06161586940288544, "learning_rate": 4.834417535660789e-05, "loss": 0.0058, "step": 48530 }, { "grad_norm": 0.05082882568240166, "learning_rate": 4.8325956214861004e-05, "loss": 0.0078, "step": 48540 }, { "grad_norm": 0.06158502772450447, "learning_rate": 4.8307737295631764e-05, "loss": 0.0063, "step": 48550 }, { "grad_norm": 0.07998410612344742, "learning_rate": 4.828951860134184e-05, "loss": 0.0068, "step": 48560 }, { "grad_norm": 0.07478500157594681, "learning_rate": 4.827130013441295e-05, "loss": 0.0074, "step": 48570 }, { "grad_norm": 0.07941513508558273, "learning_rate": 4.82530818972667e-05, "loss": 0.0083, "step": 48580 }, { "grad_norm": 0.06897471100091934, "learning_rate": 4.8234863892324714e-05, "loss": 0.0097, "step": 48590 }, { "grad_norm": 0.046738576143980026, "learning_rate": 4.821664612200857e-05, "loss": 0.0082, "step": 48600 }, { "grad_norm": 0.08075769245624542, "learning_rate": 4.8198428588739804e-05, "loss": 0.0114, "step": 48610 }, { "grad_norm": 0.06892480701208115, "learning_rate": 4.818021129493993e-05, "loss": 0.0065, "step": 48620 }, { "grad_norm": 0.0597827285528183, "learning_rate": 4.816199424303043e-05, "loss": 0.0075, "step": 48630 }, { "grad_norm": 0.0670408308506012, "learning_rate": 4.814377743543277e-05, "loss": 0.0057, "step": 48640 }, { "grad_norm": 0.08306091278791428, "learning_rate": 4.812556087456838e-05, "loss": 0.0059, "step": 48650 }, { "grad_norm": 0.08033991605043411, "learning_rate": 4.810734456285861e-05, "loss": 0.0078, "step": 48660 }, { "grad_norm": 0.08614088594913483, "learning_rate": 4.808912850272483e-05, "loss": 0.008, "step": 48670 }, { "grad_norm": 0.09255029261112213, "learning_rate": 4.807091269658838e-05, "loss": 0.0087, "step": 48680 }, { "grad_norm": 0.09235899895429611, "learning_rate": 4.805269714687051e-05, "loss": 0.0081, "step": 48690 }, { "grad_norm": 0.09283168613910675, "learning_rate": 4.8034481855992486e-05, "loss": 0.0103, "step": 48700 }, { "grad_norm": 0.06443515419960022, "learning_rate": 4.801626682637552e-05, "loss": 0.008, "step": 48710 }, { "grad_norm": 0.08483109623193741, "learning_rate": 4.799805206044081e-05, "loss": 0.0072, "step": 48720 }, { "grad_norm": 0.07731694728136063, "learning_rate": 4.79798375606095e-05, "loss": 0.0079, "step": 48730 }, { "grad_norm": 0.07277095317840576, "learning_rate": 4.796162332930269e-05, "loss": 0.0051, "step": 48740 }, { "grad_norm": 0.05383147671818733, "learning_rate": 4.7943409368941453e-05, "loss": 0.005, "step": 48750 }, { "grad_norm": 0.11771665513515472, "learning_rate": 4.792519568194686e-05, "loss": 0.0067, "step": 48760 }, { "grad_norm": 0.07706035673618317, "learning_rate": 4.790698227073988e-05, "loss": 0.0048, "step": 48770 }, { "grad_norm": 0.11320635676383972, "learning_rate": 4.7888769137741486e-05, "loss": 0.0084, "step": 48780 }, { "grad_norm": 0.06440705060958862, "learning_rate": 4.7870556285372625e-05, "loss": 0.0069, "step": 48790 }, { "grad_norm": 0.08550761640071869, "learning_rate": 4.785234371605418e-05, "loss": 0.0097, "step": 48800 }, { "grad_norm": 0.08970063924789429, "learning_rate": 4.783413143220701e-05, "loss": 0.0113, "step": 48810 }, { "grad_norm": 0.07243616878986359, "learning_rate": 4.781591943625194e-05, "loss": 0.0069, "step": 48820 }, { "grad_norm": 0.07493747025728226, "learning_rate": 4.7797707730609736e-05, "loss": 0.0055, "step": 48830 }, { "grad_norm": 0.09406819939613342, "learning_rate": 4.777949631770114e-05, "loss": 0.0066, "step": 48840 }, { "grad_norm": 0.05456986650824547, "learning_rate": 4.776128519994686e-05, "loss": 0.0071, "step": 48850 }, { "grad_norm": 0.06209106743335724, "learning_rate": 4.774307437976753e-05, "loss": 0.0078, "step": 48860 }, { "grad_norm": 0.12343081086874008, "learning_rate": 4.772486385958382e-05, "loss": 0.009, "step": 48870 }, { "grad_norm": 0.09366708993911743, "learning_rate": 4.7706653641816285e-05, "loss": 0.0097, "step": 48880 }, { "grad_norm": 0.10997837781906128, "learning_rate": 4.768844372888547e-05, "loss": 0.0112, "step": 48890 }, { "grad_norm": 0.05315606668591499, "learning_rate": 4.7670234123211895e-05, "loss": 0.0098, "step": 48900 }, { "grad_norm": 0.08913916349411011, "learning_rate": 4.7652024827216004e-05, "loss": 0.0052, "step": 48910 }, { "grad_norm": 0.0610865093767643, "learning_rate": 4.7633815843318205e-05, "loss": 0.0056, "step": 48920 }, { "grad_norm": 0.04515889659523964, "learning_rate": 4.761560717393889e-05, "loss": 0.0045, "step": 48930 }, { "grad_norm": 0.04718434065580368, "learning_rate": 4.7597398821498415e-05, "loss": 0.0059, "step": 48940 }, { "grad_norm": 0.0817815288901329, "learning_rate": 4.757919078841706e-05, "loss": 0.0082, "step": 48950 }, { "grad_norm": 0.0627439022064209, "learning_rate": 4.7560983077115066e-05, "loss": 0.0061, "step": 48960 }, { "grad_norm": 0.03865377977490425, "learning_rate": 4.7542775690012656e-05, "loss": 0.0041, "step": 48970 }, { "grad_norm": 0.0512348972260952, "learning_rate": 4.7524568629530006e-05, "loss": 0.0072, "step": 48980 }, { "grad_norm": 0.057274241000413895, "learning_rate": 4.7506361898087204e-05, "loss": 0.0077, "step": 48990 }, { "grad_norm": 0.06019372120499611, "learning_rate": 4.7488155498104346e-05, "loss": 0.0075, "step": 49000 }, { "grad_norm": 0.090818852186203, "learning_rate": 4.7469949432001486e-05, "loss": 0.0082, "step": 49010 }, { "grad_norm": 0.053072210401296616, "learning_rate": 4.74517437021986e-05, "loss": 0.0076, "step": 49020 }, { "grad_norm": 0.04301295801997185, "learning_rate": 4.743353831111565e-05, "loss": 0.0077, "step": 49030 }, { "grad_norm": 0.05422268435359001, "learning_rate": 4.741533326117251e-05, "loss": 0.0056, "step": 49040 }, { "grad_norm": 0.0745190903544426, "learning_rate": 4.739712855478905e-05, "loss": 0.0072, "step": 49050 }, { "grad_norm": 0.05884234979748726, "learning_rate": 4.7378924194385076e-05, "loss": 0.005, "step": 49060 }, { "grad_norm": 0.08940406888723373, "learning_rate": 4.736072018238037e-05, "loss": 0.0051, "step": 49070 }, { "grad_norm": 0.09473156183958054, "learning_rate": 4.7342516521194605e-05, "loss": 0.0056, "step": 49080 }, { "grad_norm": 0.0603315569460392, "learning_rate": 4.7324313213247516e-05, "loss": 0.007, "step": 49090 }, { "grad_norm": 0.08453834056854248, "learning_rate": 4.730611026095868e-05, "loss": 0.0061, "step": 49100 }, { "grad_norm": 0.05069399252533913, "learning_rate": 4.7287907666747694e-05, "loss": 0.0049, "step": 49110 }, { "grad_norm": 0.0761268213391304, "learning_rate": 4.726970543303409e-05, "loss": 0.0083, "step": 49120 }, { "grad_norm": 0.046744346618652344, "learning_rate": 4.725150356223733e-05, "loss": 0.0053, "step": 49130 }, { "grad_norm": 0.06601674854755402, "learning_rate": 4.7233302056776855e-05, "loss": 0.0072, "step": 49140 }, { "grad_norm": 0.05933254957199097, "learning_rate": 4.721510091907205e-05, "loss": 0.0068, "step": 49150 }, { "grad_norm": 0.09278364479541779, "learning_rate": 4.719690015154227e-05, "loss": 0.0071, "step": 49160 }, { "grad_norm": 0.06360932439565659, "learning_rate": 4.7178699756606786e-05, "loss": 0.0065, "step": 49170 }, { "grad_norm": 0.05562107264995575, "learning_rate": 4.716049973668483e-05, "loss": 0.0051, "step": 49180 }, { "grad_norm": 0.07026876509189606, "learning_rate": 4.71423000941956e-05, "loss": 0.0061, "step": 49190 }, { "grad_norm": 0.07668758183717728, "learning_rate": 4.712410083155823e-05, "loss": 0.0071, "step": 49200 }, { "grad_norm": 0.07070165872573853, "learning_rate": 4.710590195119179e-05, "loss": 0.0058, "step": 49210 }, { "grad_norm": 0.03620533272624016, "learning_rate": 4.708770345551534e-05, "loss": 0.0063, "step": 49220 }, { "grad_norm": 0.0487855039536953, "learning_rate": 4.706950534694785e-05, "loss": 0.0045, "step": 49230 }, { "grad_norm": 0.06501299142837524, "learning_rate": 4.705130762790826e-05, "loss": 0.009, "step": 49240 }, { "grad_norm": 0.04982997849583626, "learning_rate": 4.703311030081546e-05, "loss": 0.0069, "step": 49250 }, { "grad_norm": 0.08640258759260178, "learning_rate": 4.701491336808826e-05, "loss": 0.0058, "step": 49260 }, { "grad_norm": 0.09737274050712585, "learning_rate": 4.699671683214545e-05, "loss": 0.006, "step": 49270 }, { "grad_norm": 0.051981545984745026, "learning_rate": 4.697852069540576e-05, "loss": 0.0059, "step": 49280 }, { "grad_norm": 0.04331053048372269, "learning_rate": 4.6960324960287836e-05, "loss": 0.005, "step": 49290 }, { "grad_norm": 0.055271536111831665, "learning_rate": 4.6942129629210305e-05, "loss": 0.0062, "step": 49300 }, { "grad_norm": 0.055946070700883865, "learning_rate": 4.692393470459175e-05, "loss": 0.0082, "step": 49310 }, { "grad_norm": 0.04787050932645798, "learning_rate": 4.6905740188850675e-05, "loss": 0.0066, "step": 49320 }, { "grad_norm": 0.07705751061439514, "learning_rate": 4.688754608440552e-05, "loss": 0.0057, "step": 49330 }, { "grad_norm": 0.0741015374660492, "learning_rate": 4.686935239367471e-05, "loss": 0.0048, "step": 49340 }, { "grad_norm": 0.05839516222476959, "learning_rate": 4.685115911907656e-05, "loss": 0.0064, "step": 49350 }, { "grad_norm": 0.06475833058357239, "learning_rate": 4.683296626302938e-05, "loss": 0.0033, "step": 49360 }, { "grad_norm": 0.0630643218755722, "learning_rate": 4.681477382795142e-05, "loss": 0.0065, "step": 49370 }, { "grad_norm": 0.07250706106424332, "learning_rate": 4.6796581816260796e-05, "loss": 0.0064, "step": 49380 }, { "grad_norm": 0.05249764025211334, "learning_rate": 4.677839023037572e-05, "loss": 0.0063, "step": 49390 }, { "grad_norm": 0.12230601161718369, "learning_rate": 4.67601990727142e-05, "loss": 0.0062, "step": 49400 }, { "grad_norm": 0.13072003424167633, "learning_rate": 4.674200834569426e-05, "loss": 0.0099, "step": 49410 }, { "grad_norm": 0.12487006932497025, "learning_rate": 4.672381805173386e-05, "loss": 0.0074, "step": 49420 }, { "grad_norm": 0.08157870918512344, "learning_rate": 4.670562819325088e-05, "loss": 0.0084, "step": 49430 }, { "grad_norm": 0.05498315021395683, "learning_rate": 4.668743877266316e-05, "loss": 0.0042, "step": 49440 }, { "grad_norm": 0.04901730269193649, "learning_rate": 4.666924979238848e-05, "loss": 0.0048, "step": 49450 }, { "grad_norm": 0.10382439196109772, "learning_rate": 4.665106125484456e-05, "loss": 0.0077, "step": 49460 }, { "grad_norm": 0.07762549817562103, "learning_rate": 4.663287316244907e-05, "loss": 0.0059, "step": 49470 }, { "grad_norm": 0.062216222286224365, "learning_rate": 4.66146855176196e-05, "loss": 0.0081, "step": 49480 }, { "grad_norm": 0.06291145831346512, "learning_rate": 4.6596498322773703e-05, "loss": 0.0062, "step": 49490 }, { "grad_norm": 0.036275334656238556, "learning_rate": 4.657831158032886e-05, "loss": 0.0071, "step": 49500 }, { "grad_norm": 0.06222278252243996, "learning_rate": 4.6560125292702484e-05, "loss": 0.0047, "step": 49510 }, { "grad_norm": 0.04300931468605995, "learning_rate": 4.654193946231192e-05, "loss": 0.0062, "step": 49520 }, { "grad_norm": 0.06963717192411423, "learning_rate": 4.652375409157451e-05, "loss": 0.0066, "step": 49530 }, { "grad_norm": 0.0784623846411705, "learning_rate": 4.6505569182907474e-05, "loss": 0.0098, "step": 49540 }, { "grad_norm": 0.07380230724811554, "learning_rate": 4.6487384738728e-05, "loss": 0.006, "step": 49550 }, { "grad_norm": 0.05045042186975479, "learning_rate": 4.646920076145319e-05, "loss": 0.0073, "step": 49560 }, { "grad_norm": 0.04290878400206566, "learning_rate": 4.645101725350011e-05, "loss": 0.0075, "step": 49570 }, { "grad_norm": 0.04156486690044403, "learning_rate": 4.643283421728574e-05, "loss": 0.0071, "step": 49580 }, { "grad_norm": 0.08946424722671509, "learning_rate": 4.6414651655227026e-05, "loss": 0.007, "step": 49590 }, { "grad_norm": 0.0654483512043953, "learning_rate": 4.639646956974081e-05, "loss": 0.0088, "step": 49600 }, { "grad_norm": 0.07596833258867264, "learning_rate": 4.637828796324393e-05, "loss": 0.0058, "step": 49610 }, { "grad_norm": 0.04746253043413162, "learning_rate": 4.6360106838153094e-05, "loss": 0.0083, "step": 49620 }, { "grad_norm": 0.07436587661504745, "learning_rate": 4.6341926196885e-05, "loss": 0.0085, "step": 49630 }, { "grad_norm": 0.07412426918745041, "learning_rate": 4.6323746041856266e-05, "loss": 0.0077, "step": 49640 }, { "grad_norm": 0.10433415323495865, "learning_rate": 4.63055663754834e-05, "loss": 0.007, "step": 49650 }, { "grad_norm": 0.07639652490615845, "learning_rate": 4.628738720018292e-05, "loss": 0.0079, "step": 49660 }, { "grad_norm": 0.0942913144826889, "learning_rate": 4.6269208518371224e-05, "loss": 0.0051, "step": 49670 }, { "grad_norm": 0.07960514724254608, "learning_rate": 4.625103033246467e-05, "loss": 0.0064, "step": 49680 }, { "grad_norm": 0.06327646970748901, "learning_rate": 4.6232852644879554e-05, "loss": 0.0062, "step": 49690 }, { "grad_norm": 0.05219157785177231, "learning_rate": 4.621467545803208e-05, "loss": 0.0048, "step": 49700 }, { "grad_norm": 0.06692366302013397, "learning_rate": 4.619649877433839e-05, "loss": 0.0114, "step": 49710 }, { "grad_norm": 0.07705585658550262, "learning_rate": 4.6178322596214604e-05, "loss": 0.0075, "step": 49720 }, { "grad_norm": 0.04310782253742218, "learning_rate": 4.616014692607671e-05, "loss": 0.005, "step": 49730 }, { "grad_norm": 0.07332867383956909, "learning_rate": 4.614197176634067e-05, "loss": 0.0082, "step": 49740 }, { "grad_norm": 0.057251691818237305, "learning_rate": 4.612379711942235e-05, "loss": 0.0054, "step": 49750 }, { "grad_norm": 0.07793586701154709, "learning_rate": 4.610562298773759e-05, "loss": 0.0086, "step": 49760 }, { "grad_norm": 0.06738781929016113, "learning_rate": 4.608744937370214e-05, "loss": 0.0079, "step": 49770 }, { "grad_norm": 0.06740141659975052, "learning_rate": 4.606927627973165e-05, "loss": 0.0053, "step": 49780 }, { "grad_norm": 0.11007394641637802, "learning_rate": 4.605110370824174e-05, "loss": 0.0059, "step": 49790 }, { "grad_norm": 0.07906889915466309, "learning_rate": 4.6032931661647965e-05, "loss": 0.0055, "step": 49800 }, { "grad_norm": 0.07666973769664764, "learning_rate": 4.601476014236576e-05, "loss": 0.0089, "step": 49810 }, { "grad_norm": 0.10181837528944016, "learning_rate": 4.599658915281054e-05, "loss": 0.0064, "step": 49820 }, { "grad_norm": 0.07043027132749557, "learning_rate": 4.5978418695397654e-05, "loss": 0.0068, "step": 49830 }, { "grad_norm": 0.07151878625154495, "learning_rate": 4.5960248772542334e-05, "loss": 0.0077, "step": 49840 }, { "grad_norm": 0.08329413086175919, "learning_rate": 4.5942079386659784e-05, "loss": 0.0069, "step": 49850 }, { "grad_norm": 0.07613720744848251, "learning_rate": 4.592391054016511e-05, "loss": 0.0065, "step": 49860 }, { "grad_norm": 0.09253179281949997, "learning_rate": 4.5905742235473355e-05, "loss": 0.0072, "step": 49870 }, { "grad_norm": 0.1072717159986496, "learning_rate": 4.588757447499948e-05, "loss": 0.0102, "step": 49880 }, { "grad_norm": 0.07620637863874435, "learning_rate": 4.586940726115843e-05, "loss": 0.0057, "step": 49890 }, { "grad_norm": 0.04914412647485733, "learning_rate": 4.585124059636494e-05, "loss": 0.0039, "step": 49900 }, { "grad_norm": 0.06417404115200043, "learning_rate": 4.583307448303387e-05, "loss": 0.0089, "step": 49910 }, { "grad_norm": 0.09262726455926895, "learning_rate": 4.581490892357984e-05, "loss": 0.0081, "step": 49920 }, { "grad_norm": 0.09677628427743912, "learning_rate": 4.579674392041747e-05, "loss": 0.0072, "step": 49930 }, { "grad_norm": 0.07406016439199448, "learning_rate": 4.5778579475961296e-05, "loss": 0.0076, "step": 49940 }, { "grad_norm": 0.11873366683721542, "learning_rate": 4.5760415592625766e-05, "loss": 0.0052, "step": 49950 }, { "grad_norm": 0.12049860507249832, "learning_rate": 4.574225227282527e-05, "loss": 0.0071, "step": 49960 }, { "grad_norm": 0.07244856655597687, "learning_rate": 4.572408951897411e-05, "loss": 0.0074, "step": 49970 }, { "grad_norm": 0.06703764200210571, "learning_rate": 4.5705927333486534e-05, "loss": 0.0054, "step": 49980 }, { "grad_norm": 0.053220540285110474, "learning_rate": 4.5687765718776707e-05, "loss": 0.0099, "step": 49990 }, { "grad_norm": 0.08407139778137207, "learning_rate": 4.566960467725868e-05, "loss": 0.0085, "step": 50000 }, { "grad_norm": 0.050910867750644684, "learning_rate": 4.5651444211346486e-05, "loss": 0.0074, "step": 50010 }, { "grad_norm": 0.06943152844905853, "learning_rate": 4.563328432345405e-05, "loss": 0.0044, "step": 50020 }, { "grad_norm": 0.051859110593795776, "learning_rate": 4.561512501599521e-05, "loss": 0.0067, "step": 50030 }, { "grad_norm": 0.06575202196836472, "learning_rate": 4.5596966291383756e-05, "loss": 0.0058, "step": 50040 }, { "grad_norm": 0.05809391289949417, "learning_rate": 4.5578808152033374e-05, "loss": 0.0058, "step": 50050 }, { "grad_norm": 0.0583714060485363, "learning_rate": 4.5560650600357704e-05, "loss": 0.0077, "step": 50060 }, { "grad_norm": 0.07549944519996643, "learning_rate": 4.554249363877029e-05, "loss": 0.007, "step": 50070 }, { "grad_norm": 0.07750450074672699, "learning_rate": 4.552433726968457e-05, "loss": 0.0093, "step": 50080 }, { "grad_norm": 0.08156979829072952, "learning_rate": 4.550618149551394e-05, "loss": 0.0066, "step": 50090 }, { "grad_norm": 0.057851642370224, "learning_rate": 4.548802631867171e-05, "loss": 0.0067, "step": 50100 }, { "grad_norm": 0.08828996866941452, "learning_rate": 4.5469871741571115e-05, "loss": 0.0082, "step": 50110 }, { "grad_norm": 0.10654224455356598, "learning_rate": 4.545171776662527e-05, "loss": 0.0069, "step": 50120 }, { "grad_norm": 0.09817486256361008, "learning_rate": 4.5433564396247284e-05, "loss": 0.005, "step": 50130 }, { "grad_norm": 0.07339508831501007, "learning_rate": 4.541541163285012e-05, "loss": 0.0076, "step": 50140 }, { "grad_norm": 0.07411355525255203, "learning_rate": 4.5397259478846686e-05, "loss": 0.0098, "step": 50150 }, { "grad_norm": 0.059245891869068146, "learning_rate": 4.537910793664982e-05, "loss": 0.0091, "step": 50160 }, { "grad_norm": 0.04426252096891403, "learning_rate": 4.5360957008672246e-05, "loss": 0.0062, "step": 50170 }, { "grad_norm": 0.0593612976372242, "learning_rate": 4.534280669732663e-05, "loss": 0.0064, "step": 50180 }, { "grad_norm": 0.05961135774850845, "learning_rate": 4.532465700502555e-05, "loss": 0.0044, "step": 50190 }, { "grad_norm": 0.08288717269897461, "learning_rate": 4.530650793418151e-05, "loss": 0.0098, "step": 50200 }, { "grad_norm": 0.0854291319847107, "learning_rate": 4.528835948720695e-05, "loss": 0.0091, "step": 50210 }, { "grad_norm": 0.07802649587392807, "learning_rate": 4.527021166651415e-05, "loss": 0.0069, "step": 50220 }, { "grad_norm": 0.05367555096745491, "learning_rate": 4.52520644745154e-05, "loss": 0.0062, "step": 50230 }, { "grad_norm": 0.059711355715990067, "learning_rate": 4.5233917913622856e-05, "loss": 0.0059, "step": 50240 }, { "grad_norm": 0.06504691392183304, "learning_rate": 4.521577198624858e-05, "loss": 0.0058, "step": 50250 }, { "grad_norm": 0.06913445144891739, "learning_rate": 4.519762669480459e-05, "loss": 0.0084, "step": 50260 }, { "grad_norm": 0.07958691567182541, "learning_rate": 4.517948204170278e-05, "loss": 0.0084, "step": 50270 }, { "grad_norm": 0.04575883597135544, "learning_rate": 4.5161338029355006e-05, "loss": 0.0048, "step": 50280 }, { "grad_norm": 0.06569086760282516, "learning_rate": 4.5143194660173e-05, "loss": 0.0104, "step": 50290 }, { "grad_norm": 0.03742564469575882, "learning_rate": 4.51250519365684e-05, "loss": 0.0036, "step": 50300 }, { "grad_norm": 0.05928478389978409, "learning_rate": 4.51069098609528e-05, "loss": 0.0067, "step": 50310 }, { "grad_norm": 0.06867394596338272, "learning_rate": 4.508876843573768e-05, "loss": 0.007, "step": 50320 }, { "grad_norm": 0.05034889653325081, "learning_rate": 4.507062766333443e-05, "loss": 0.0073, "step": 50330 }, { "grad_norm": 0.07595613598823547, "learning_rate": 4.5052487546154356e-05, "loss": 0.0078, "step": 50340 }, { "grad_norm": 0.10660766065120697, "learning_rate": 4.50343480866087e-05, "loss": 0.0047, "step": 50350 }, { "grad_norm": 0.09935717284679413, "learning_rate": 4.50162092871086e-05, "loss": 0.0097, "step": 50360 }, { "grad_norm": 0.0713280662894249, "learning_rate": 4.49980711500651e-05, "loss": 0.0075, "step": 50370 }, { "grad_norm": 0.06190372630953789, "learning_rate": 4.497993367788917e-05, "loss": 0.0068, "step": 50380 }, { "grad_norm": 0.12392719835042953, "learning_rate": 4.496179687299167e-05, "loss": 0.0098, "step": 50390 }, { "grad_norm": 0.07953919470310211, "learning_rate": 4.494366073778338e-05, "loss": 0.0056, "step": 50400 }, { "grad_norm": 0.07625029981136322, "learning_rate": 4.492552527467502e-05, "loss": 0.0083, "step": 50410 }, { "grad_norm": 0.07373959571123123, "learning_rate": 4.490739048607715e-05, "loss": 0.0068, "step": 50420 }, { "grad_norm": 0.08153747767210007, "learning_rate": 4.488925637440035e-05, "loss": 0.0065, "step": 50430 }, { "grad_norm": 0.08716010302305222, "learning_rate": 4.4871122942055006e-05, "loss": 0.009, "step": 50440 }, { "grad_norm": 0.06048639863729477, "learning_rate": 4.485299019145146e-05, "loss": 0.0052, "step": 50450 }, { "grad_norm": 0.06867397576570511, "learning_rate": 4.483485812499997e-05, "loss": 0.0068, "step": 50460 }, { "grad_norm": 0.07243765145540237, "learning_rate": 4.481672674511067e-05, "loss": 0.0065, "step": 50470 }, { "grad_norm": 0.05465313419699669, "learning_rate": 4.479859605419364e-05, "loss": 0.0058, "step": 50480 }, { "grad_norm": 0.09606928378343582, "learning_rate": 4.4780466054658836e-05, "loss": 0.008, "step": 50490 }, { "grad_norm": 0.05994314327836037, "learning_rate": 4.4762336748916156e-05, "loss": 0.0054, "step": 50500 }, { "grad_norm": 0.06334836781024933, "learning_rate": 4.47442081393754e-05, "loss": 0.0066, "step": 50510 }, { "grad_norm": 0.08927721530199051, "learning_rate": 4.472608022844622e-05, "loss": 0.0058, "step": 50520 }, { "grad_norm": 0.08548203855752945, "learning_rate": 4.470795301853825e-05, "loss": 0.0075, "step": 50530 }, { "grad_norm": 0.06869442760944366, "learning_rate": 4.4689826512061007e-05, "loss": 0.0077, "step": 50540 }, { "grad_norm": 0.04900399595499039, "learning_rate": 4.4671700711423875e-05, "loss": 0.0038, "step": 50550 }, { "grad_norm": 0.0848984494805336, "learning_rate": 4.4653575619036195e-05, "loss": 0.0066, "step": 50560 }, { "grad_norm": 0.04425948113203049, "learning_rate": 4.463545123730718e-05, "loss": 0.0057, "step": 50570 }, { "grad_norm": 0.057809360325336456, "learning_rate": 4.461732756864599e-05, "loss": 0.0078, "step": 50580 }, { "grad_norm": 0.0600271075963974, "learning_rate": 4.459920461546166e-05, "loss": 0.0051, "step": 50590 }, { "grad_norm": 0.07115443050861359, "learning_rate": 4.45810823801631e-05, "loss": 0.0046, "step": 50600 }, { "grad_norm": 0.03897529095411301, "learning_rate": 4.456296086515919e-05, "loss": 0.0053, "step": 50610 }, { "grad_norm": 0.07040911167860031, "learning_rate": 4.454484007285868e-05, "loss": 0.0064, "step": 50620 }, { "grad_norm": 0.04753012955188751, "learning_rate": 4.452672000567022e-05, "loss": 0.006, "step": 50630 }, { "grad_norm": 0.09334179013967514, "learning_rate": 4.450860066600235e-05, "loss": 0.0061, "step": 50640 }, { "grad_norm": 0.05720347911119461, "learning_rate": 4.449048205626357e-05, "loss": 0.0065, "step": 50650 }, { "grad_norm": 0.08195862919092178, "learning_rate": 4.4472364178862214e-05, "loss": 0.0084, "step": 50660 }, { "grad_norm": 0.06659863889217377, "learning_rate": 4.445424703620657e-05, "loss": 0.006, "step": 50670 }, { "grad_norm": 0.08590637147426605, "learning_rate": 4.443613063070482e-05, "loss": 0.0058, "step": 50680 }, { "grad_norm": 0.09153661131858826, "learning_rate": 4.441801496476501e-05, "loss": 0.0078, "step": 50690 }, { "grad_norm": 0.048925478011369705, "learning_rate": 4.439990004079512e-05, "loss": 0.0075, "step": 50700 }, { "grad_norm": 0.04889398068189621, "learning_rate": 4.438178586120303e-05, "loss": 0.0064, "step": 50710 }, { "grad_norm": 0.041189685463905334, "learning_rate": 4.436367242839652e-05, "loss": 0.0059, "step": 50720 }, { "grad_norm": 0.051753707230091095, "learning_rate": 4.434555974478329e-05, "loss": 0.0053, "step": 50730 }, { "grad_norm": 0.07186621427536011, "learning_rate": 4.432744781277088e-05, "loss": 0.0073, "step": 50740 }, { "grad_norm": 0.06276149302721024, "learning_rate": 4.430933663476678e-05, "loss": 0.0077, "step": 50750 }, { "grad_norm": 0.07195217907428741, "learning_rate": 4.429122621317839e-05, "loss": 0.006, "step": 50760 }, { "grad_norm": 0.04384496435523033, "learning_rate": 4.427311655041295e-05, "loss": 0.0049, "step": 50770 }, { "grad_norm": 0.0601639449596405, "learning_rate": 4.425500764887767e-05, "loss": 0.0068, "step": 50780 }, { "grad_norm": 0.07198213040828705, "learning_rate": 4.423689951097959e-05, "loss": 0.005, "step": 50790 }, { "grad_norm": 0.05457199737429619, "learning_rate": 4.4218792139125726e-05, "loss": 0.0065, "step": 50800 }, { "grad_norm": 0.07088581472635269, "learning_rate": 4.420068553572293e-05, "loss": 0.0076, "step": 50810 }, { "grad_norm": 0.06960242241621017, "learning_rate": 4.4182579703177976e-05, "loss": 0.0082, "step": 50820 }, { "grad_norm": 0.046134352684020996, "learning_rate": 4.416447464389752e-05, "loss": 0.0078, "step": 50830 }, { "grad_norm": 0.038461025804281235, "learning_rate": 4.414637036028816e-05, "loss": 0.0085, "step": 50840 }, { "grad_norm": 0.06892014294862747, "learning_rate": 4.412826685475631e-05, "loss": 0.0089, "step": 50850 }, { "grad_norm": 0.05148256570100784, "learning_rate": 4.411016412970835e-05, "loss": 0.0047, "step": 50860 }, { "grad_norm": 0.05934172123670578, "learning_rate": 4.4092062187550544e-05, "loss": 0.0053, "step": 50870 }, { "grad_norm": 0.049611710011959076, "learning_rate": 4.407396103068903e-05, "loss": 0.0055, "step": 50880 }, { "grad_norm": 0.06106935814023018, "learning_rate": 4.405586066152988e-05, "loss": 0.0103, "step": 50890 }, { "grad_norm": 0.051142677664756775, "learning_rate": 4.4037761082479e-05, "loss": 0.0069, "step": 50900 }, { "grad_norm": 0.04644802585244179, "learning_rate": 4.4019662295942244e-05, "loss": 0.0052, "step": 50910 }, { "grad_norm": 0.08864807337522507, "learning_rate": 4.4001564304325335e-05, "loss": 0.0048, "step": 50920 }, { "grad_norm": 0.07973041385412216, "learning_rate": 4.398346711003392e-05, "loss": 0.0103, "step": 50930 }, { "grad_norm": 0.07428412139415741, "learning_rate": 4.396537071547347e-05, "loss": 0.005, "step": 50940 }, { "grad_norm": 0.07049733400344849, "learning_rate": 4.394727512304947e-05, "loss": 0.0062, "step": 50950 }, { "grad_norm": 0.08613643050193787, "learning_rate": 4.392918033516717e-05, "loss": 0.0091, "step": 50960 }, { "grad_norm": 0.08741357177495956, "learning_rate": 4.391108635423179e-05, "loss": 0.0072, "step": 50970 }, { "grad_norm": 0.07245013117790222, "learning_rate": 4.389299318264843e-05, "loss": 0.0062, "step": 50980 }, { "grad_norm": 0.07033879309892654, "learning_rate": 4.387490082282205e-05, "loss": 0.0087, "step": 50990 }, { "grad_norm": 0.0522434376180172, "learning_rate": 4.3856809277157544e-05, "loss": 0.0086, "step": 51000 }, { "grad_norm": 0.04056696221232414, "learning_rate": 4.383871854805967e-05, "loss": 0.004, "step": 51010 }, { "grad_norm": 0.0803174376487732, "learning_rate": 4.38206286379331e-05, "loss": 0.0072, "step": 51020 }, { "grad_norm": 0.08459322899580002, "learning_rate": 4.38025395491824e-05, "loss": 0.0077, "step": 51030 }, { "grad_norm": 0.0566687136888504, "learning_rate": 4.3784451284211985e-05, "loss": 0.0059, "step": 51040 }, { "grad_norm": 0.05291832238435745, "learning_rate": 4.376636384542619e-05, "loss": 0.0095, "step": 51050 }, { "grad_norm": 0.0545189194381237, "learning_rate": 4.3748277235229255e-05, "loss": 0.0067, "step": 51060 }, { "grad_norm": 0.07042457163333893, "learning_rate": 4.373019145602528e-05, "loss": 0.0062, "step": 51070 }, { "grad_norm": 0.06830843538045883, "learning_rate": 4.371210651021826e-05, "loss": 0.0066, "step": 51080 }, { "grad_norm": 0.0834026113152504, "learning_rate": 4.369402240021209e-05, "loss": 0.009, "step": 51090 }, { "grad_norm": 0.058622367680072784, "learning_rate": 4.367593912841056e-05, "loss": 0.0048, "step": 51100 }, { "grad_norm": 0.08014710992574692, "learning_rate": 4.365785669721735e-05, "loss": 0.0077, "step": 51110 }, { "grad_norm": 0.06330198049545288, "learning_rate": 4.363977510903598e-05, "loss": 0.0073, "step": 51120 }, { "grad_norm": 0.1253284364938736, "learning_rate": 4.3621694366269935e-05, "loss": 0.009, "step": 51130 }, { "grad_norm": 0.06748294830322266, "learning_rate": 4.360361447132252e-05, "loss": 0.0053, "step": 51140 }, { "grad_norm": 0.07342597842216492, "learning_rate": 4.358553542659698e-05, "loss": 0.0051, "step": 51150 }, { "grad_norm": 0.05926293134689331, "learning_rate": 4.356745723449638e-05, "loss": 0.0061, "step": 51160 }, { "grad_norm": 0.08986330032348633, "learning_rate": 4.354937989742375e-05, "loss": 0.0098, "step": 51170 }, { "grad_norm": 0.06888590008020401, "learning_rate": 4.353130341778197e-05, "loss": 0.0069, "step": 51180 }, { "grad_norm": 0.03438662737607956, "learning_rate": 4.351322779797379e-05, "loss": 0.006, "step": 51190 }, { "grad_norm": 0.06982314586639404, "learning_rate": 4.3495153040401874e-05, "loss": 0.0067, "step": 51200 }, { "grad_norm": 0.07396002113819122, "learning_rate": 4.3477079147468755e-05, "loss": 0.0051, "step": 51210 }, { "grad_norm": 0.06961985677480698, "learning_rate": 4.345900612157684e-05, "loss": 0.0069, "step": 51220 }, { "grad_norm": 0.07388361543416977, "learning_rate": 4.344093396512845e-05, "loss": 0.006, "step": 51230 }, { "grad_norm": 0.0964105874300003, "learning_rate": 4.342286268052578e-05, "loss": 0.006, "step": 51240 }, { "grad_norm": 0.06053665280342102, "learning_rate": 4.3404792270170916e-05, "loss": 0.0062, "step": 51250 }, { "grad_norm": 0.08712360262870789, "learning_rate": 4.338672273646579e-05, "loss": 0.008, "step": 51260 }, { "grad_norm": 0.041003331542015076, "learning_rate": 4.336865408181227e-05, "loss": 0.0059, "step": 51270 }, { "grad_norm": 0.08486644178628922, "learning_rate": 4.335058630861206e-05, "loss": 0.005, "step": 51280 }, { "grad_norm": 0.05330023914575577, "learning_rate": 4.3332519419266785e-05, "loss": 0.0053, "step": 51290 }, { "grad_norm": 0.06978372484445572, "learning_rate": 4.331445341617792e-05, "loss": 0.0053, "step": 51300 }, { "grad_norm": 0.04747111722826958, "learning_rate": 4.3296388301746836e-05, "loss": 0.0052, "step": 51310 }, { "grad_norm": 0.08174312114715576, "learning_rate": 4.3278324078374814e-05, "loss": 0.005, "step": 51320 }, { "grad_norm": 0.06662020087242126, "learning_rate": 4.326026074846299e-05, "loss": 0.0054, "step": 51330 }, { "grad_norm": 0.054624564945697784, "learning_rate": 4.324219831441236e-05, "loss": 0.007, "step": 51340 }, { "grad_norm": 0.044200073927640915, "learning_rate": 4.322413677862384e-05, "loss": 0.0052, "step": 51350 }, { "grad_norm": 0.10254237055778503, "learning_rate": 4.3206076143498204e-05, "loss": 0.0095, "step": 51360 }, { "grad_norm": 0.034658487886190414, "learning_rate": 4.31880164114361e-05, "loss": 0.006, "step": 51370 }, { "grad_norm": 0.09162634611129761, "learning_rate": 4.316995758483807e-05, "loss": 0.0078, "step": 51380 }, { "grad_norm": 0.08716365694999695, "learning_rate": 4.315189966610455e-05, "loss": 0.0079, "step": 51390 }, { "grad_norm": 0.08648983389139175, "learning_rate": 4.313384265763584e-05, "loss": 0.0062, "step": 51400 }, { "grad_norm": 0.05151040479540825, "learning_rate": 4.311578656183211e-05, "loss": 0.007, "step": 51410 }, { "grad_norm": 0.05754830315709114, "learning_rate": 4.309773138109341e-05, "loss": 0.0076, "step": 51420 }, { "grad_norm": 0.05718066915869713, "learning_rate": 4.307967711781968e-05, "loss": 0.006, "step": 51430 }, { "grad_norm": 0.054503876715898514, "learning_rate": 4.306162377441073e-05, "loss": 0.0064, "step": 51440 }, { "grad_norm": 0.06894563138484955, "learning_rate": 4.304357135326627e-05, "loss": 0.0055, "step": 51450 }, { "grad_norm": 0.05727982893586159, "learning_rate": 4.3025519856785815e-05, "loss": 0.0044, "step": 51460 }, { "grad_norm": 0.09186641871929169, "learning_rate": 4.300746928736888e-05, "loss": 0.0072, "step": 51470 }, { "grad_norm": 0.09247996658086777, "learning_rate": 4.298941964741475e-05, "loss": 0.0064, "step": 51480 }, { "grad_norm": 0.10081897675991058, "learning_rate": 4.297137093932263e-05, "loss": 0.0066, "step": 51490 }, { "grad_norm": 0.06554768979549408, "learning_rate": 4.29533231654916e-05, "loss": 0.005, "step": 51500 }, { "grad_norm": 0.08008963614702225, "learning_rate": 4.293527632832059e-05, "loss": 0.0058, "step": 51510 }, { "grad_norm": 0.0735098123550415, "learning_rate": 4.2917230430208446e-05, "loss": 0.0074, "step": 51520 }, { "grad_norm": 0.04603971913456917, "learning_rate": 4.2899185473553844e-05, "loss": 0.0059, "step": 51530 }, { "grad_norm": 0.08503637462854385, "learning_rate": 4.2881141460755393e-05, "loss": 0.0073, "step": 51540 }, { "grad_norm": 0.061168331652879715, "learning_rate": 4.286309839421154e-05, "loss": 0.0061, "step": 51550 }, { "grad_norm": 0.04524838924407959, "learning_rate": 4.2845056276320586e-05, "loss": 0.0052, "step": 51560 }, { "grad_norm": 0.09360959380865097, "learning_rate": 4.2827015109480746e-05, "loss": 0.0081, "step": 51570 }, { "grad_norm": 0.06971558183431625, "learning_rate": 4.280897489609009e-05, "loss": 0.0117, "step": 51580 }, { "grad_norm": 0.06299920380115509, "learning_rate": 4.279093563854656e-05, "loss": 0.0068, "step": 51590 }, { "grad_norm": 0.0910012423992157, "learning_rate": 4.2772897339247975e-05, "loss": 0.0068, "step": 51600 }, { "grad_norm": 0.058699559420347214, "learning_rate": 4.275486000059202e-05, "loss": 0.006, "step": 51610 }, { "grad_norm": 0.04814250394701958, "learning_rate": 4.273682362497627e-05, "loss": 0.0048, "step": 51620 }, { "grad_norm": 0.03434707596898079, "learning_rate": 4.2718788214798166e-05, "loss": 0.0052, "step": 51630 }, { "grad_norm": 0.06096399948000908, "learning_rate": 4.2700753772455004e-05, "loss": 0.0067, "step": 51640 }, { "grad_norm": 0.06530001014471054, "learning_rate": 4.2682720300343956e-05, "loss": 0.0062, "step": 51650 }, { "grad_norm": 0.04995157569646835, "learning_rate": 4.266468780086209e-05, "loss": 0.005, "step": 51660 }, { "grad_norm": 0.05802062526345253, "learning_rate": 4.2646656276406304e-05, "loss": 0.0069, "step": 51670 }, { "grad_norm": 0.07963918894529343, "learning_rate": 4.262862572937339e-05, "loss": 0.0068, "step": 51680 }, { "grad_norm": 0.06506510823965073, "learning_rate": 4.261059616216003e-05, "loss": 0.0053, "step": 51690 }, { "grad_norm": 0.06284596771001816, "learning_rate": 4.259256757716275e-05, "loss": 0.0055, "step": 51700 }, { "grad_norm": 0.08350216597318649, "learning_rate": 4.257453997677793e-05, "loss": 0.0053, "step": 51710 }, { "grad_norm": 0.05475998669862747, "learning_rate": 4.255651336340187e-05, "loss": 0.007, "step": 51720 }, { "grad_norm": 0.05188184231519699, "learning_rate": 4.253848773943067e-05, "loss": 0.0046, "step": 51730 }, { "grad_norm": 0.08657897263765335, "learning_rate": 4.2520463107260364e-05, "loss": 0.0073, "step": 51740 }, { "grad_norm": 0.047539789229631424, "learning_rate": 4.25024394692868e-05, "loss": 0.0052, "step": 51750 }, { "grad_norm": 0.08358705788850784, "learning_rate": 4.248441682790576e-05, "loss": 0.0062, "step": 51760 }, { "grad_norm": 0.048611804842948914, "learning_rate": 4.246639518551283e-05, "loss": 0.0061, "step": 51770 }, { "grad_norm": 0.046830859035253525, "learning_rate": 4.244837454450348e-05, "loss": 0.0075, "step": 51780 }, { "grad_norm": 0.05281677842140198, "learning_rate": 4.243035490727307e-05, "loss": 0.0063, "step": 51790 }, { "grad_norm": 0.0590788871049881, "learning_rate": 4.2412336276216806e-05, "loss": 0.0076, "step": 51800 }, { "grad_norm": 0.05854613333940506, "learning_rate": 4.2394318653729753e-05, "loss": 0.005, "step": 51810 }, { "grad_norm": 0.06066964939236641, "learning_rate": 4.237630204220687e-05, "loss": 0.007, "step": 51820 }, { "grad_norm": 0.04158991202712059, "learning_rate": 4.235828644404294e-05, "loss": 0.0041, "step": 51830 }, { "grad_norm": 0.0893908441066742, "learning_rate": 4.2340271861632664e-05, "loss": 0.0072, "step": 51840 }, { "grad_norm": 0.062293801456689835, "learning_rate": 4.232225829737058e-05, "loss": 0.0076, "step": 51850 }, { "grad_norm": 0.06222836673259735, "learning_rate": 4.230424575365107e-05, "loss": 0.0058, "step": 51860 }, { "grad_norm": 0.05647571384906769, "learning_rate": 4.228623423286843e-05, "loss": 0.0084, "step": 51870 }, { "grad_norm": 0.03769377991557121, "learning_rate": 4.2268223737416766e-05, "loss": 0.0045, "step": 51880 }, { "grad_norm": 0.06888506561517715, "learning_rate": 4.225021426969008e-05, "loss": 0.0058, "step": 51890 }, { "grad_norm": 0.06565240770578384, "learning_rate": 4.2232205832082225e-05, "loss": 0.0054, "step": 51900 }, { "grad_norm": 0.05283620208501816, "learning_rate": 4.221419842698694e-05, "loss": 0.0083, "step": 51910 }, { "grad_norm": 0.04620078206062317, "learning_rate": 4.219619205679779e-05, "loss": 0.0054, "step": 51920 }, { "grad_norm": 0.053647562861442566, "learning_rate": 4.2178186723908255e-05, "loss": 0.0062, "step": 51930 }, { "grad_norm": 0.05838284268975258, "learning_rate": 4.216018243071161e-05, "loss": 0.0066, "step": 51940 }, { "grad_norm": 0.08247865736484528, "learning_rate": 4.214217917960104e-05, "loss": 0.0098, "step": 51950 }, { "grad_norm": 0.051902804523706436, "learning_rate": 4.212417697296957e-05, "loss": 0.0091, "step": 51960 }, { "grad_norm": 0.07478515058755875, "learning_rate": 4.2106175813210114e-05, "loss": 0.0089, "step": 51970 }, { "grad_norm": 0.032618217170238495, "learning_rate": 4.208817570271537e-05, "loss": 0.0075, "step": 51980 }, { "grad_norm": 0.05481573939323425, "learning_rate": 4.2070176643878034e-05, "loss": 0.0052, "step": 51990 }, { "grad_norm": 0.07000754028558731, "learning_rate": 4.205217863909052e-05, "loss": 0.0053, "step": 52000 }, { "grad_norm": 0.08331555873155594, "learning_rate": 4.20341816907452e-05, "loss": 0.0058, "step": 52010 }, { "grad_norm": 0.0688098818063736, "learning_rate": 4.201618580123425e-05, "loss": 0.0086, "step": 52020 }, { "grad_norm": 0.07645206898450851, "learning_rate": 4.1998190972949714e-05, "loss": 0.0081, "step": 52030 }, { "grad_norm": 0.05065500736236572, "learning_rate": 4.1980197208283525e-05, "loss": 0.0049, "step": 52040 }, { "grad_norm": 0.06039836257696152, "learning_rate": 4.1962204509627436e-05, "loss": 0.0076, "step": 52050 }, { "grad_norm": 0.08530935645103455, "learning_rate": 4.194421287937309e-05, "loss": 0.005, "step": 52060 }, { "grad_norm": 0.06908897310495377, "learning_rate": 4.192622231991199e-05, "loss": 0.0085, "step": 52070 }, { "grad_norm": 0.10548146069049835, "learning_rate": 4.190823283363545e-05, "loss": 0.0058, "step": 52080 }, { "grad_norm": 0.05472683161497116, "learning_rate": 4.1890244422934685e-05, "loss": 0.0055, "step": 52090 }, { "grad_norm": 0.08267782628536224, "learning_rate": 4.187225709020077e-05, "loss": 0.0051, "step": 52100 }, { "grad_norm": 0.06717188656330109, "learning_rate": 4.185427083782459e-05, "loss": 0.0071, "step": 52110 }, { "grad_norm": 0.12494102120399475, "learning_rate": 4.183628566819694e-05, "loss": 0.0084, "step": 52120 }, { "grad_norm": 0.08281964063644409, "learning_rate": 4.181830158370843e-05, "loss": 0.0062, "step": 52130 }, { "grad_norm": 0.0524328351020813, "learning_rate": 4.180031858674957e-05, "loss": 0.0054, "step": 52140 }, { "grad_norm": 0.06304338574409485, "learning_rate": 4.1782336679710696e-05, "loss": 0.0063, "step": 52150 }, { "grad_norm": 0.08221440762281418, "learning_rate": 4.1764355864981995e-05, "loss": 0.0059, "step": 52160 }, { "grad_norm": 0.03998551890254021, "learning_rate": 4.174637614495351e-05, "loss": 0.0044, "step": 52170 }, { "grad_norm": 0.0665125697851181, "learning_rate": 4.1728397522015174e-05, "loss": 0.0057, "step": 52180 }, { "grad_norm": 0.0435970276594162, "learning_rate": 4.1710419998556704e-05, "loss": 0.0054, "step": 52190 }, { "grad_norm": 0.06328108161687851, "learning_rate": 4.169244357696773e-05, "loss": 0.0058, "step": 52200 }, { "grad_norm": 0.09260622411966324, "learning_rate": 4.167446825963773e-05, "loss": 0.0079, "step": 52210 }, { "grad_norm": 0.08351971209049225, "learning_rate": 4.165649404895602e-05, "loss": 0.0064, "step": 52220 }, { "grad_norm": 0.06641225516796112, "learning_rate": 4.163852094731177e-05, "loss": 0.0061, "step": 52230 }, { "grad_norm": 0.06933318078517914, "learning_rate": 4.1620548957094e-05, "loss": 0.0074, "step": 52240 }, { "grad_norm": 0.08448295295238495, "learning_rate": 4.160257808069159e-05, "loss": 0.0055, "step": 52250 }, { "grad_norm": 0.07052769511938095, "learning_rate": 4.158460832049327e-05, "loss": 0.0057, "step": 52260 }, { "grad_norm": 0.08466403931379318, "learning_rate": 4.15666396788876e-05, "loss": 0.0063, "step": 52270 }, { "grad_norm": 0.05547564849257469, "learning_rate": 4.1548672158263045e-05, "loss": 0.0049, "step": 52280 }, { "grad_norm": 0.0735708624124527, "learning_rate": 4.1530705761007894e-05, "loss": 0.0042, "step": 52290 }, { "grad_norm": 0.07983797043561935, "learning_rate": 4.1512740489510236e-05, "loss": 0.0081, "step": 52300 }, { "grad_norm": 0.05151091143488884, "learning_rate": 4.1494776346158095e-05, "loss": 0.0068, "step": 52310 }, { "grad_norm": 0.04226725921034813, "learning_rate": 4.147681333333929e-05, "loss": 0.0053, "step": 52320 }, { "grad_norm": 0.062216419726610184, "learning_rate": 4.14588514534415e-05, "loss": 0.0078, "step": 52330 }, { "grad_norm": 0.041267745196819305, "learning_rate": 4.1440890708852266e-05, "loss": 0.005, "step": 52340 }, { "grad_norm": 0.06350260972976685, "learning_rate": 4.142293110195895e-05, "loss": 0.0064, "step": 52350 }, { "grad_norm": 0.05202043056488037, "learning_rate": 4.140497263514881e-05, "loss": 0.0052, "step": 52360 }, { "grad_norm": 0.05497937276959419, "learning_rate": 4.138701531080894e-05, "loss": 0.0047, "step": 52370 }, { "grad_norm": 0.054156120866537094, "learning_rate": 4.1369059131326215e-05, "loss": 0.0069, "step": 52380 }, { "grad_norm": 0.055366579443216324, "learning_rate": 4.135110409908745e-05, "loss": 0.0059, "step": 52390 }, { "grad_norm": 0.08014438301324844, "learning_rate": 4.133315021647927e-05, "loss": 0.0078, "step": 52400 }, { "grad_norm": 0.04236774891614914, "learning_rate": 4.1315197485888116e-05, "loss": 0.0054, "step": 52410 }, { "grad_norm": 0.03982416167855263, "learning_rate": 4.129724590970032e-05, "loss": 0.0054, "step": 52420 }, { "grad_norm": 0.07905229926109314, "learning_rate": 4.127929549030205e-05, "loss": 0.0051, "step": 52430 }, { "grad_norm": 0.04737206920981407, "learning_rate": 4.126134623007931e-05, "loss": 0.0072, "step": 52440 }, { "grad_norm": 0.05995599925518036, "learning_rate": 4.124339813141798e-05, "loss": 0.0062, "step": 52450 }, { "grad_norm": 0.047782979905605316, "learning_rate": 4.122545119670372e-05, "loss": 0.0083, "step": 52460 }, { "grad_norm": 0.049618177115917206, "learning_rate": 4.120750542832211e-05, "loss": 0.0061, "step": 52470 }, { "grad_norm": 0.07722964137792587, "learning_rate": 4.118956082865852e-05, "loss": 0.0074, "step": 52480 }, { "grad_norm": 0.05618089810013771, "learning_rate": 4.1171617400098205e-05, "loss": 0.0049, "step": 52490 }, { "grad_norm": 0.0720100998878479, "learning_rate": 4.1153675145026213e-05, "loss": 0.0065, "step": 52500 }, { "grad_norm": 0.038702819496393204, "learning_rate": 4.113573406582751e-05, "loss": 0.013, "step": 52510 }, { "grad_norm": 0.10475294291973114, "learning_rate": 4.111779416488683e-05, "loss": 0.0085, "step": 52520 }, { "grad_norm": 0.06148015707731247, "learning_rate": 4.1099855444588817e-05, "loss": 0.0059, "step": 52530 }, { "grad_norm": 0.07451361417770386, "learning_rate": 4.108191790731792e-05, "loss": 0.0084, "step": 52540 }, { "grad_norm": 0.06968697905540466, "learning_rate": 4.106398155545841e-05, "loss": 0.0078, "step": 52550 }, { "grad_norm": 0.07255423814058304, "learning_rate": 4.1046046391394435e-05, "loss": 0.0067, "step": 52560 }, { "grad_norm": 0.06750009208917618, "learning_rate": 4.102811241750999e-05, "loss": 0.006, "step": 52570 }, { "grad_norm": 0.06101636216044426, "learning_rate": 4.101017963618888e-05, "loss": 0.006, "step": 52580 }, { "grad_norm": 0.06515524536371231, "learning_rate": 4.0992248049814816e-05, "loss": 0.0043, "step": 52590 }, { "grad_norm": 0.058813877403736115, "learning_rate": 4.0974317660771254e-05, "loss": 0.0044, "step": 52600 }, { "grad_norm": 0.04410206526517868, "learning_rate": 4.0956388471441564e-05, "loss": 0.011, "step": 52610 }, { "grad_norm": 0.06461222469806671, "learning_rate": 4.093846048420894e-05, "loss": 0.0079, "step": 52620 }, { "grad_norm": 0.05325799062848091, "learning_rate": 4.092053370145639e-05, "loss": 0.0066, "step": 52630 }, { "grad_norm": 0.04279647022485733, "learning_rate": 4.0902608125566785e-05, "loss": 0.0059, "step": 52640 }, { "grad_norm": 0.0640413835644722, "learning_rate": 4.088468375892283e-05, "loss": 0.0078, "step": 52650 }, { "grad_norm": 0.04951292648911476, "learning_rate": 4.08667606039071e-05, "loss": 0.006, "step": 52660 }, { "grad_norm": 0.08051583915948868, "learning_rate": 4.084883866290197e-05, "loss": 0.0126, "step": 52670 }, { "grad_norm": 0.07643447816371918, "learning_rate": 4.083091793828964e-05, "loss": 0.0072, "step": 52680 }, { "grad_norm": 0.06507064402103424, "learning_rate": 4.08129984324522e-05, "loss": 0.0077, "step": 52690 }, { "grad_norm": 0.04835865646600723, "learning_rate": 4.079508014777154e-05, "loss": 0.0065, "step": 52700 }, { "grad_norm": 0.09937907010316849, "learning_rate": 4.077716308662939e-05, "loss": 0.0053, "step": 52710 }, { "grad_norm": 0.09459301084280014, "learning_rate": 4.075924725140733e-05, "loss": 0.0052, "step": 52720 }, { "grad_norm": 0.04295165464282036, "learning_rate": 4.0741332644486787e-05, "loss": 0.0074, "step": 52730 }, { "grad_norm": 0.06865917891263962, "learning_rate": 4.0723419268248994e-05, "loss": 0.0069, "step": 52740 }, { "grad_norm": 0.05132397636771202, "learning_rate": 4.070550712507505e-05, "loss": 0.0054, "step": 52750 }, { "grad_norm": 0.06438814848661423, "learning_rate": 4.0687596217345885e-05, "loss": 0.0086, "step": 52760 }, { "grad_norm": 0.07597538828849792, "learning_rate": 4.0669686547442235e-05, "loss": 0.0079, "step": 52770 }, { "grad_norm": 0.07513218373060226, "learning_rate": 4.065177811774469e-05, "loss": 0.0106, "step": 52780 }, { "grad_norm": 0.05932406336069107, "learning_rate": 4.0633870930633685e-05, "loss": 0.006, "step": 52790 }, { "grad_norm": 0.07504177838563919, "learning_rate": 4.0615964988489496e-05, "loss": 0.0111, "step": 52800 }, { "grad_norm": 0.05415453389286995, "learning_rate": 4.0598060293692226e-05, "loss": 0.005, "step": 52810 }, { "grad_norm": 0.05217229574918747, "learning_rate": 4.058015684862178e-05, "loss": 0.0055, "step": 52820 }, { "grad_norm": 0.09354820847511292, "learning_rate": 4.0562254655657945e-05, "loss": 0.0057, "step": 52830 }, { "grad_norm": 0.057274386286735535, "learning_rate": 4.054435371718032e-05, "loss": 0.0054, "step": 52840 }, { "grad_norm": 0.05732050910592079, "learning_rate": 4.0526454035568316e-05, "loss": 0.0055, "step": 52850 }, { "grad_norm": 0.08089985698461533, "learning_rate": 4.050855561320122e-05, "loss": 0.0077, "step": 52860 }, { "grad_norm": 0.08807229995727539, "learning_rate": 4.049065845245811e-05, "loss": 0.0088, "step": 52870 }, { "grad_norm": 0.05857241526246071, "learning_rate": 4.0472762555717946e-05, "loss": 0.008, "step": 52880 }, { "grad_norm": 0.06784036755561829, "learning_rate": 4.045486792535949e-05, "loss": 0.0052, "step": 52890 }, { "grad_norm": 0.06801487505435944, "learning_rate": 4.04369745637613e-05, "loss": 0.0053, "step": 52900 }, { "grad_norm": 0.06241336092352867, "learning_rate": 4.041908247330183e-05, "loss": 0.0069, "step": 52910 }, { "grad_norm": 0.06446722894906998, "learning_rate": 4.0401191656359337e-05, "loss": 0.0041, "step": 52920 }, { "grad_norm": 0.07319605350494385, "learning_rate": 4.0383302115311896e-05, "loss": 0.006, "step": 52930 }, { "grad_norm": 0.07782988250255585, "learning_rate": 4.0365413852537415e-05, "loss": 0.0092, "step": 52940 }, { "grad_norm": 0.2064768522977829, "learning_rate": 4.034752687041366e-05, "loss": 0.0091, "step": 52950 }, { "grad_norm": 0.0938154086470604, "learning_rate": 4.032964117131822e-05, "loss": 0.0071, "step": 52960 }, { "grad_norm": 0.07328855246305466, "learning_rate": 4.031175675762849e-05, "loss": 0.0091, "step": 52970 }, { "grad_norm": 0.05556865409016609, "learning_rate": 4.02938736317217e-05, "loss": 0.0066, "step": 52980 }, { "grad_norm": 0.08170878142118454, "learning_rate": 4.027599179597491e-05, "loss": 0.0071, "step": 52990 }, { "grad_norm": 0.0677240788936615, "learning_rate": 4.0258111252765037e-05, "loss": 0.0092, "step": 53000 }, { "grad_norm": 0.12003376334905624, "learning_rate": 4.0240232004468785e-05, "loss": 0.0096, "step": 53010 }, { "grad_norm": 0.034911494702100754, "learning_rate": 4.022235405346269e-05, "loss": 0.0086, "step": 53020 }, { "grad_norm": 0.0581689327955246, "learning_rate": 4.020447740212316e-05, "loss": 0.0072, "step": 53030 }, { "grad_norm": 0.05684121325612068, "learning_rate": 4.018660205282638e-05, "loss": 0.0055, "step": 53040 }, { "grad_norm": 0.07808619737625122, "learning_rate": 4.016872800794839e-05, "loss": 0.0109, "step": 53050 }, { "grad_norm": 0.04626360535621643, "learning_rate": 4.0150855269865055e-05, "loss": 0.0068, "step": 53060 }, { "grad_norm": 0.0713275894522667, "learning_rate": 4.013298384095204e-05, "loss": 0.004, "step": 53070 }, { "grad_norm": 0.04332338646054268, "learning_rate": 4.011511372358486e-05, "loss": 0.0073, "step": 53080 }, { "grad_norm": 0.07210011780261993, "learning_rate": 4.0097244920138857e-05, "loss": 0.0054, "step": 53090 }, { "grad_norm": 0.07814577966928482, "learning_rate": 4.007937743298919e-05, "loss": 0.0075, "step": 53100 }, { "grad_norm": 0.05102339759469032, "learning_rate": 4.006151126451086e-05, "loss": 0.0065, "step": 53110 }, { "grad_norm": 0.0654999390244484, "learning_rate": 4.0043646417078654e-05, "loss": 0.0065, "step": 53120 }, { "grad_norm": 0.08537451922893524, "learning_rate": 4.002578289306721e-05, "loss": 0.008, "step": 53130 }, { "grad_norm": 0.09144893288612366, "learning_rate": 4.000792069485102e-05, "loss": 0.0086, "step": 53140 }, { "grad_norm": 0.07313575595617294, "learning_rate": 3.999005982480432e-05, "loss": 0.0117, "step": 53150 }, { "grad_norm": 0.04042283073067665, "learning_rate": 3.9972200285301246e-05, "loss": 0.0057, "step": 53160 }, { "grad_norm": 0.04444727301597595, "learning_rate": 3.9954342078715706e-05, "loss": 0.0083, "step": 53170 }, { "grad_norm": 0.057290032505989075, "learning_rate": 3.993648520742148e-05, "loss": 0.0046, "step": 53180 }, { "grad_norm": 0.05895084887742996, "learning_rate": 3.9918629673792145e-05, "loss": 0.0097, "step": 53190 }, { "grad_norm": 0.08408107608556747, "learning_rate": 3.990077548020107e-05, "loss": 0.0076, "step": 53200 }, { "grad_norm": 0.08259552717208862, "learning_rate": 3.988292262902149e-05, "loss": 0.0073, "step": 53210 }, { "grad_norm": 0.08483507484197617, "learning_rate": 3.986507112262646e-05, "loss": 0.0061, "step": 53220 }, { "grad_norm": 0.044652074575424194, "learning_rate": 3.984722096338881e-05, "loss": 0.0041, "step": 53230 }, { "grad_norm": 0.07018347829580307, "learning_rate": 3.9829372153681225e-05, "loss": 0.0082, "step": 53240 }, { "grad_norm": 0.05207684636116028, "learning_rate": 3.981152469587625e-05, "loss": 0.0059, "step": 53250 }, { "grad_norm": 0.046097658574581146, "learning_rate": 3.979367859234617e-05, "loss": 0.0097, "step": 53260 }, { "grad_norm": 0.08571844547986984, "learning_rate": 3.977583384546314e-05, "loss": 0.0087, "step": 53270 }, { "grad_norm": 0.07710438221693039, "learning_rate": 3.975799045759914e-05, "loss": 0.0063, "step": 53280 }, { "grad_norm": 0.07786072790622711, "learning_rate": 3.974014843112593e-05, "loss": 0.0073, "step": 53290 }, { "grad_norm": 0.02953372895717621, "learning_rate": 3.972230776841512e-05, "loss": 0.0061, "step": 53300 }, { "grad_norm": 0.04787693917751312, "learning_rate": 3.9704468471838116e-05, "loss": 0.0071, "step": 53310 }, { "grad_norm": 0.047296371310949326, "learning_rate": 3.9686630543766186e-05, "loss": 0.0061, "step": 53320 }, { "grad_norm": 0.05569107085466385, "learning_rate": 3.966879398657038e-05, "loss": 0.0066, "step": 53330 }, { "grad_norm": 0.04106690362095833, "learning_rate": 3.965095880262156e-05, "loss": 0.0044, "step": 53340 }, { "grad_norm": 0.07313629984855652, "learning_rate": 3.963312499429042e-05, "loss": 0.0075, "step": 53350 }, { "grad_norm": 0.04916708543896675, "learning_rate": 3.9615292563947494e-05, "loss": 0.0061, "step": 53360 }, { "grad_norm": 0.08197824656963348, "learning_rate": 3.9597461513963073e-05, "loss": 0.0063, "step": 53370 }, { "grad_norm": 0.06471794843673706, "learning_rate": 3.9579631846707315e-05, "loss": 0.0065, "step": 53380 }, { "grad_norm": 0.08101881295442581, "learning_rate": 3.956180356455017e-05, "loss": 0.0056, "step": 53390 }, { "grad_norm": 0.08175875246524811, "learning_rate": 3.954397666986144e-05, "loss": 0.0068, "step": 53400 }, { "grad_norm": 0.061105601489543915, "learning_rate": 3.9526151165010706e-05, "loss": 0.006, "step": 53410 }, { "grad_norm": 0.05495353043079376, "learning_rate": 3.950832705236735e-05, "loss": 0.009, "step": 53420 }, { "grad_norm": 0.06425543874502182, "learning_rate": 3.9490504334300616e-05, "loss": 0.0061, "step": 53430 }, { "grad_norm": 0.05940813571214676, "learning_rate": 3.947268301317955e-05, "loss": 0.0062, "step": 53440 }, { "grad_norm": 0.07560314983129501, "learning_rate": 3.945486309137296e-05, "loss": 0.0096, "step": 53450 }, { "grad_norm": 0.11307425051927567, "learning_rate": 3.943704457124954e-05, "loss": 0.0101, "step": 53460 }, { "grad_norm": 0.06319893151521683, "learning_rate": 3.9419227455177756e-05, "loss": 0.0053, "step": 53470 }, { "grad_norm": 0.048064567148685455, "learning_rate": 3.940141174552592e-05, "loss": 0.0068, "step": 53480 }, { "grad_norm": 0.059029512107372284, "learning_rate": 3.938359744466213e-05, "loss": 0.007, "step": 53490 }, { "grad_norm": 0.050941165536642075, "learning_rate": 3.9365784554954274e-05, "loss": 0.0053, "step": 53500 }, { "grad_norm": 0.06156853213906288, "learning_rate": 3.9347973078770116e-05, "loss": 0.0075, "step": 53510 }, { "grad_norm": 0.06765297055244446, "learning_rate": 3.9330163018477174e-05, "loss": 0.0055, "step": 53520 }, { "grad_norm": 0.0627594068646431, "learning_rate": 3.9312354376442815e-05, "loss": 0.0089, "step": 53530 }, { "grad_norm": 0.053720634430646896, "learning_rate": 3.929454715503419e-05, "loss": 0.0061, "step": 53540 }, { "grad_norm": 0.057356443256139755, "learning_rate": 3.927674135661829e-05, "loss": 0.007, "step": 53550 }, { "grad_norm": 0.05784618854522705, "learning_rate": 3.925893698356188e-05, "loss": 0.0044, "step": 53560 }, { "grad_norm": 0.06045255437493324, "learning_rate": 3.924113403823158e-05, "loss": 0.0139, "step": 53570 }, { "grad_norm": 0.05277158319950104, "learning_rate": 3.922333252299381e-05, "loss": 0.0062, "step": 53580 }, { "grad_norm": 0.07135389745235443, "learning_rate": 3.920553244021474e-05, "loss": 0.0064, "step": 53590 }, { "grad_norm": 0.08686123788356781, "learning_rate": 3.918773379226043e-05, "loss": 0.0074, "step": 53600 }, { "grad_norm": 0.09501869231462479, "learning_rate": 3.9169936581496695e-05, "loss": 0.0054, "step": 53610 }, { "grad_norm": 0.06198809668421745, "learning_rate": 3.9152140810289204e-05, "loss": 0.0061, "step": 53620 }, { "grad_norm": 0.08636999130249023, "learning_rate": 3.9134346481003415e-05, "loss": 0.0072, "step": 53630 }, { "grad_norm": 0.05557987838983536, "learning_rate": 3.9116553596004564e-05, "loss": 0.0059, "step": 53640 }, { "grad_norm": 0.07889041304588318, "learning_rate": 3.909876215765773e-05, "loss": 0.0078, "step": 53650 }, { "grad_norm": 0.07501038163900375, "learning_rate": 3.9080972168327806e-05, "loss": 0.0055, "step": 53660 }, { "grad_norm": 0.05744731053709984, "learning_rate": 3.906318363037945e-05, "loss": 0.0045, "step": 53670 }, { "grad_norm": 0.0636778473854065, "learning_rate": 3.904539654617717e-05, "loss": 0.0061, "step": 53680 }, { "grad_norm": 0.09096202999353409, "learning_rate": 3.9027610918085255e-05, "loss": 0.0054, "step": 53690 }, { "grad_norm": 0.0982607826590538, "learning_rate": 3.900982674846783e-05, "loss": 0.0071, "step": 53700 }, { "grad_norm": 0.04162370413541794, "learning_rate": 3.89920440396888e-05, "loss": 0.0074, "step": 53710 }, { "grad_norm": 0.05389947444200516, "learning_rate": 3.8974262794111874e-05, "loss": 0.0063, "step": 53720 }, { "grad_norm": 0.06563758850097656, "learning_rate": 3.895648301410058e-05, "loss": 0.0044, "step": 53730 }, { "grad_norm": 0.04971809312701225, "learning_rate": 3.893870470201825e-05, "loss": 0.0067, "step": 53740 }, { "grad_norm": 0.07182811945676804, "learning_rate": 3.8920927860227995e-05, "loss": 0.0038, "step": 53750 }, { "grad_norm": 0.09739582985639572, "learning_rate": 3.890315249109277e-05, "loss": 0.0074, "step": 53760 }, { "grad_norm": 0.04961298406124115, "learning_rate": 3.888537859697532e-05, "loss": 0.0065, "step": 53770 }, { "grad_norm": 0.060935840010643005, "learning_rate": 3.886760618023819e-05, "loss": 0.0048, "step": 53780 }, { "grad_norm": 0.04064391180872917, "learning_rate": 3.8849835243243737e-05, "loss": 0.0039, "step": 53790 }, { "grad_norm": 0.056495726108551025, "learning_rate": 3.883206578835409e-05, "loss": 0.0073, "step": 53800 }, { "grad_norm": 0.06321373581886292, "learning_rate": 3.881429781793122e-05, "loss": 0.0084, "step": 53810 }, { "grad_norm": 0.045385539531707764, "learning_rate": 3.8796531334336886e-05, "loss": 0.0047, "step": 53820 }, { "grad_norm": 0.04348039627075195, "learning_rate": 3.877876633993263e-05, "loss": 0.0065, "step": 53830 }, { "grad_norm": 0.06423735618591309, "learning_rate": 3.876100283707984e-05, "loss": 0.0055, "step": 53840 }, { "grad_norm": 0.14161266386508942, "learning_rate": 3.874324082813968e-05, "loss": 0.0053, "step": 53850 }, { "grad_norm": 0.038913238793611526, "learning_rate": 3.8725480315473104e-05, "loss": 0.0076, "step": 53860 }, { "grad_norm": 0.07923915237188339, "learning_rate": 3.870772130144088e-05, "loss": 0.0055, "step": 53870 }, { "grad_norm": 0.069142647087574, "learning_rate": 3.8689963788403586e-05, "loss": 0.0052, "step": 53880 }, { "grad_norm": 0.06240665167570114, "learning_rate": 3.867220777872157e-05, "loss": 0.0062, "step": 53890 }, { "grad_norm": 0.07095392793416977, "learning_rate": 3.865445327475501e-05, "loss": 0.0054, "step": 53900 }, { "grad_norm": 0.07954873144626617, "learning_rate": 3.863670027886387e-05, "loss": 0.0055, "step": 53910 }, { "grad_norm": 0.05535302311182022, "learning_rate": 3.861894879340794e-05, "loss": 0.0069, "step": 53920 }, { "grad_norm": 0.06294272094964981, "learning_rate": 3.860119882074678e-05, "loss": 0.0047, "step": 53930 }, { "grad_norm": 0.06867550313472748, "learning_rate": 3.858345036323972e-05, "loss": 0.0045, "step": 53940 }, { "grad_norm": 0.06277239322662354, "learning_rate": 3.8565703423245976e-05, "loss": 0.0061, "step": 53950 }, { "grad_norm": 0.083697110414505, "learning_rate": 3.854795800312449e-05, "loss": 0.0087, "step": 53960 }, { "grad_norm": 0.07066282629966736, "learning_rate": 3.853021410523401e-05, "loss": 0.0058, "step": 53970 }, { "grad_norm": 0.05739464610815048, "learning_rate": 3.851247173193309e-05, "loss": 0.0087, "step": 53980 }, { "grad_norm": 0.09126699715852737, "learning_rate": 3.8494730885580114e-05, "loss": 0.0055, "step": 53990 }, { "grad_norm": 0.05294397473335266, "learning_rate": 3.8476991568533223e-05, "loss": 0.0079, "step": 54000 }, { "grad_norm": 0.06332172453403473, "learning_rate": 3.845925378315037e-05, "loss": 0.0054, "step": 54010 }, { "grad_norm": 0.05269749462604523, "learning_rate": 3.8441517531789295e-05, "loss": 0.0061, "step": 54020 }, { "grad_norm": 0.04715985804796219, "learning_rate": 3.842378281680754e-05, "loss": 0.0061, "step": 54030 }, { "grad_norm": 0.08019746094942093, "learning_rate": 3.8406049640562444e-05, "loss": 0.0075, "step": 54040 }, { "grad_norm": 0.07125520706176758, "learning_rate": 3.838831800541115e-05, "loss": 0.0092, "step": 54050 }, { "grad_norm": 0.07323004305362701, "learning_rate": 3.8370587913710553e-05, "loss": 0.0081, "step": 54060 }, { "grad_norm": 0.05491552874445915, "learning_rate": 3.8352859367817414e-05, "loss": 0.0064, "step": 54070 }, { "grad_norm": 0.05657734349370003, "learning_rate": 3.8335132370088237e-05, "loss": 0.0081, "step": 54080 }, { "grad_norm": 0.08721249550580978, "learning_rate": 3.831740692287934e-05, "loss": 0.0064, "step": 54090 }, { "grad_norm": 0.052417851984500885, "learning_rate": 3.829968302854683e-05, "loss": 0.0068, "step": 54100 }, { "grad_norm": 0.056363970041275024, "learning_rate": 3.8281960689446595e-05, "loss": 0.0067, "step": 54110 }, { "grad_norm": 0.07238122820854187, "learning_rate": 3.826423990793433e-05, "loss": 0.0058, "step": 54120 }, { "grad_norm": 0.07854525744915009, "learning_rate": 3.824652068636551e-05, "loss": 0.0076, "step": 54130 }, { "grad_norm": 0.09063974767923355, "learning_rate": 3.822880302709544e-05, "loss": 0.0092, "step": 54140 }, { "grad_norm": 0.05021487921476364, "learning_rate": 3.8211086932479204e-05, "loss": 0.0076, "step": 54150 }, { "grad_norm": 0.04379791021347046, "learning_rate": 3.8193372404871614e-05, "loss": 0.0088, "step": 54160 }, { "grad_norm": 0.06531158089637756, "learning_rate": 3.8175659446627357e-05, "loss": 0.0052, "step": 54170 }, { "grad_norm": 0.052071355283260345, "learning_rate": 3.815794806010088e-05, "loss": 0.0088, "step": 54180 }, { "grad_norm": 0.028015170246362686, "learning_rate": 3.81402382476464e-05, "loss": 0.0055, "step": 54190 }, { "grad_norm": 0.03927842527627945, "learning_rate": 3.812253001161796e-05, "loss": 0.0063, "step": 54200 }, { "grad_norm": 0.051587410271167755, "learning_rate": 3.810482335436937e-05, "loss": 0.005, "step": 54210 }, { "grad_norm": 0.06393599510192871, "learning_rate": 3.8087118278254246e-05, "loss": 0.0059, "step": 54220 }, { "grad_norm": 0.07034339755773544, "learning_rate": 3.8069414785626e-05, "loss": 0.0071, "step": 54230 }, { "grad_norm": 0.037205882370471954, "learning_rate": 3.8051712878837794e-05, "loss": 0.0049, "step": 54240 }, { "grad_norm": 0.07660278677940369, "learning_rate": 3.803401256024261e-05, "loss": 0.0091, "step": 54250 }, { "grad_norm": 0.039955984801054, "learning_rate": 3.8016313832193234e-05, "loss": 0.0054, "step": 54260 }, { "grad_norm": 0.04344431310892105, "learning_rate": 3.799861669704219e-05, "loss": 0.006, "step": 54270 }, { "grad_norm": 0.05453392490744591, "learning_rate": 3.7980921157141835e-05, "loss": 0.0064, "step": 54280 }, { "grad_norm": 0.06409188359975815, "learning_rate": 3.796322721484431e-05, "loss": 0.0059, "step": 54290 }, { "grad_norm": 0.0505584217607975, "learning_rate": 3.794553487250152e-05, "loss": 0.0058, "step": 54300 }, { "grad_norm": 0.07700835913419724, "learning_rate": 3.7927844132465185e-05, "loss": 0.0058, "step": 54310 }, { "grad_norm": 0.040324773639440536, "learning_rate": 3.791015499708678e-05, "loss": 0.0042, "step": 54320 }, { "grad_norm": 0.06788656860589981, "learning_rate": 3.7892467468717594e-05, "loss": 0.0047, "step": 54330 }, { "grad_norm": 0.059447724372148514, "learning_rate": 3.787478154970869e-05, "loss": 0.0076, "step": 54340 }, { "grad_norm": 0.07609475404024124, "learning_rate": 3.785709724241091e-05, "loss": 0.0088, "step": 54350 }, { "grad_norm": 0.08437756448984146, "learning_rate": 3.783941454917491e-05, "loss": 0.0084, "step": 54360 }, { "grad_norm": 0.0975368395447731, "learning_rate": 3.7821733472351116e-05, "loss": 0.0071, "step": 54370 }, { "grad_norm": 0.09432704746723175, "learning_rate": 3.7804054014289724e-05, "loss": 0.0056, "step": 54380 }, { "grad_norm": 0.07959302514791489, "learning_rate": 3.778637617734072e-05, "loss": 0.0072, "step": 54390 }, { "grad_norm": 0.07151325792074203, "learning_rate": 3.77686999638539e-05, "loss": 0.0056, "step": 54400 }, { "grad_norm": 0.10048749297857285, "learning_rate": 3.775102537617882e-05, "loss": 0.0073, "step": 54410 }, { "grad_norm": 0.08260341733694077, "learning_rate": 3.773335241666481e-05, "loss": 0.0076, "step": 54420 }, { "grad_norm": 0.09780918806791306, "learning_rate": 3.7715681087661e-05, "loss": 0.0057, "step": 54430 }, { "grad_norm": 0.06261282414197922, "learning_rate": 3.769801139151633e-05, "loss": 0.0066, "step": 54440 }, { "grad_norm": 0.10400092601776123, "learning_rate": 3.7680343330579494e-05, "loss": 0.0051, "step": 54450 }, { "grad_norm": 0.05866154655814171, "learning_rate": 3.766267690719893e-05, "loss": 0.0051, "step": 54460 }, { "grad_norm": 0.053816549479961395, "learning_rate": 3.764501212372294e-05, "loss": 0.008, "step": 54470 }, { "grad_norm": 0.08959925919771194, "learning_rate": 3.762734898249955e-05, "loss": 0.0054, "step": 54480 }, { "grad_norm": 0.09057619422674179, "learning_rate": 3.760968748587658e-05, "loss": 0.0078, "step": 54490 }, { "grad_norm": 0.061311449855566025, "learning_rate": 3.759202763620163e-05, "loss": 0.0077, "step": 54500 }, { "grad_norm": 0.06538785248994827, "learning_rate": 3.7574369435822106e-05, "loss": 0.0072, "step": 54510 }, { "grad_norm": 0.0837896391749382, "learning_rate": 3.755671288708516e-05, "loss": 0.005, "step": 54520 }, { "grad_norm": 0.0491933599114418, "learning_rate": 3.753905799233777e-05, "loss": 0.006, "step": 54530 }, { "grad_norm": 0.048448268324136734, "learning_rate": 3.7521404753926625e-05, "loss": 0.0044, "step": 54540 }, { "grad_norm": 0.05267883464694023, "learning_rate": 3.7503753174198244e-05, "loss": 0.0063, "step": 54550 }, { "grad_norm": 0.05708488076925278, "learning_rate": 3.748610325549893e-05, "loss": 0.0075, "step": 54560 }, { "grad_norm": 0.05127522721886635, "learning_rate": 3.746845500017473e-05, "loss": 0.0063, "step": 54570 }, { "grad_norm": 0.043411415070295334, "learning_rate": 3.745080841057149e-05, "loss": 0.0058, "step": 54580 }, { "grad_norm": 0.040174949914216995, "learning_rate": 3.743316348903485e-05, "loss": 0.0074, "step": 54590 }, { "grad_norm": 0.08243823051452637, "learning_rate": 3.7415520237910205e-05, "loss": 0.0057, "step": 54600 }, { "grad_norm": 0.05973808839917183, "learning_rate": 3.739787865954274e-05, "loss": 0.0057, "step": 54610 }, { "grad_norm": 0.08172667026519775, "learning_rate": 3.738023875627741e-05, "loss": 0.0066, "step": 54620 }, { "grad_norm": 0.05704101547598839, "learning_rate": 3.7362600530458944e-05, "loss": 0.0055, "step": 54630 }, { "grad_norm": 0.06724218279123306, "learning_rate": 3.734496398443186e-05, "loss": 0.007, "step": 54640 }, { "grad_norm": 0.06278739869594574, "learning_rate": 3.732732912054042e-05, "loss": 0.0059, "step": 54650 }, { "grad_norm": 0.07184560596942902, "learning_rate": 3.730969594112874e-05, "loss": 0.0074, "step": 54660 }, { "grad_norm": 0.0837545171380043, "learning_rate": 3.729206444854065e-05, "loss": 0.0046, "step": 54670 }, { "grad_norm": 0.05272138863801956, "learning_rate": 3.727443464511974e-05, "loss": 0.0075, "step": 54680 }, { "grad_norm": 0.03980136290192604, "learning_rate": 3.725680653320941e-05, "loss": 0.0048, "step": 54690 }, { "grad_norm": 0.05352245643734932, "learning_rate": 3.723918011515285e-05, "loss": 0.0064, "step": 54700 }, { "grad_norm": 0.058937303721904755, "learning_rate": 3.722155539329298e-05, "loss": 0.0046, "step": 54710 }, { "grad_norm": 0.0587463341653347, "learning_rate": 3.7203932369972525e-05, "loss": 0.0065, "step": 54720 }, { "grad_norm": 0.06358733773231506, "learning_rate": 3.718631104753397e-05, "loss": 0.0056, "step": 54730 }, { "grad_norm": 0.05528755486011505, "learning_rate": 3.716869142831959e-05, "loss": 0.0049, "step": 54740 }, { "grad_norm": 0.07001674920320511, "learning_rate": 3.715107351467144e-05, "loss": 0.0058, "step": 54750 }, { "grad_norm": 0.07618959993124008, "learning_rate": 3.7133457308931295e-05, "loss": 0.0043, "step": 54760 }, { "grad_norm": 0.059612467885017395, "learning_rate": 3.711584281344077e-05, "loss": 0.0046, "step": 54770 }, { "grad_norm": 0.044831544160842896, "learning_rate": 3.709823003054122e-05, "loss": 0.0037, "step": 54780 }, { "grad_norm": 0.04760809615254402, "learning_rate": 3.708061896257375e-05, "loss": 0.0048, "step": 54790 }, { "grad_norm": 0.05978546291589737, "learning_rate": 3.7063009611879286e-05, "loss": 0.0073, "step": 54800 }, { "grad_norm": 0.08528533577919006, "learning_rate": 3.7045401980798497e-05, "loss": 0.0062, "step": 54810 }, { "grad_norm": 0.05983975529670715, "learning_rate": 3.7027796071671836e-05, "loss": 0.0053, "step": 54820 }, { "grad_norm": 0.041549742221832275, "learning_rate": 3.701019188683952e-05, "loss": 0.005, "step": 54830 }, { "grad_norm": 0.07442861050367355, "learning_rate": 3.699258942864151e-05, "loss": 0.0056, "step": 54840 }, { "grad_norm": 0.05975331366062164, "learning_rate": 3.697498869941759e-05, "loss": 0.007, "step": 54850 }, { "grad_norm": 0.07186248153448105, "learning_rate": 3.695738970150728e-05, "loss": 0.0085, "step": 54860 }, { "grad_norm": 0.05978985130786896, "learning_rate": 3.693979243724988e-05, "loss": 0.0068, "step": 54870 }, { "grad_norm": 0.044809550046920776, "learning_rate": 3.6922196908984455e-05, "loss": 0.0067, "step": 54880 }, { "grad_norm": 0.07230690866708755, "learning_rate": 3.690460311904986e-05, "loss": 0.0094, "step": 54890 }, { "grad_norm": 0.056721244007349014, "learning_rate": 3.688701106978467e-05, "loss": 0.006, "step": 54900 }, { "grad_norm": 0.08902405202388763, "learning_rate": 3.686942076352728e-05, "loss": 0.007, "step": 54910 }, { "grad_norm": 0.038653090596199036, "learning_rate": 3.685183220261583e-05, "loss": 0.0079, "step": 54920 }, { "grad_norm": 0.04209073260426521, "learning_rate": 3.6834245389388226e-05, "loss": 0.0073, "step": 54930 }, { "grad_norm": 0.034607551991939545, "learning_rate": 3.681666032618215e-05, "loss": 0.0056, "step": 54940 }, { "grad_norm": 0.05244215950369835, "learning_rate": 3.679907701533503e-05, "loss": 0.0049, "step": 54950 }, { "grad_norm": 0.066725954413414, "learning_rate": 3.678149545918411e-05, "loss": 0.0057, "step": 54960 }, { "grad_norm": 0.06654828041791916, "learning_rate": 3.6763915660066365e-05, "loss": 0.0076, "step": 54970 }, { "grad_norm": 0.03740251064300537, "learning_rate": 3.6746337620318526e-05, "loss": 0.0047, "step": 54980 }, { "grad_norm": 0.036226507276296616, "learning_rate": 3.672876134227711e-05, "loss": 0.005, "step": 54990 }, { "grad_norm": 0.058136217296123505, "learning_rate": 3.67111868282784e-05, "loss": 0.0055, "step": 55000 }, { "grad_norm": 0.056914545595645905, "learning_rate": 3.669361408065843e-05, "loss": 0.0069, "step": 55010 }, { "grad_norm": 0.03532738983631134, "learning_rate": 3.6676043101753e-05, "loss": 0.0048, "step": 55020 }, { "grad_norm": 0.0840948075056076, "learning_rate": 3.6658473893897716e-05, "loss": 0.0079, "step": 55030 }, { "grad_norm": 0.05910210311412811, "learning_rate": 3.664090645942789e-05, "loss": 0.0083, "step": 55040 }, { "grad_norm": 0.06712934374809265, "learning_rate": 3.662334080067865e-05, "loss": 0.0061, "step": 55050 }, { "grad_norm": 0.06262889504432678, "learning_rate": 3.6605776919984834e-05, "loss": 0.0058, "step": 55060 }, { "grad_norm": 0.07017488032579422, "learning_rate": 3.658821481968109e-05, "loss": 0.0056, "step": 55070 }, { "grad_norm": 0.048145879060029984, "learning_rate": 3.657065450210181e-05, "loss": 0.007, "step": 55080 }, { "grad_norm": 0.06894958019256592, "learning_rate": 3.655309596958113e-05, "loss": 0.0063, "step": 55090 }, { "grad_norm": 0.054378800094127655, "learning_rate": 3.653553922445299e-05, "loss": 0.0053, "step": 55100 }, { "grad_norm": 0.05961569398641586, "learning_rate": 3.651798426905107e-05, "loss": 0.0055, "step": 55110 }, { "grad_norm": 0.08512715995311737, "learning_rate": 3.650043110570881e-05, "loss": 0.0097, "step": 55120 }, { "grad_norm": 0.12313748896121979, "learning_rate": 3.648287973675942e-05, "loss": 0.0078, "step": 55130 }, { "grad_norm": 0.08938107639551163, "learning_rate": 3.6465330164535886e-05, "loss": 0.0077, "step": 55140 }, { "grad_norm": 0.04570973291993141, "learning_rate": 3.6447782391370887e-05, "loss": 0.0041, "step": 55150 }, { "grad_norm": 0.09126157313585281, "learning_rate": 3.6430236419596955e-05, "loss": 0.0059, "step": 55160 }, { "grad_norm": 0.05832364410161972, "learning_rate": 3.64126922515463e-05, "loss": 0.0053, "step": 55170 }, { "grad_norm": 0.04925181344151497, "learning_rate": 3.639514988955098e-05, "loss": 0.0061, "step": 55180 }, { "grad_norm": 0.06598000973463058, "learning_rate": 3.637760933594274e-05, "loss": 0.0068, "step": 55190 }, { "grad_norm": 0.1053946241736412, "learning_rate": 3.6360070593053115e-05, "loss": 0.007, "step": 55200 }, { "grad_norm": 0.09262193739414215, "learning_rate": 3.6342533663213384e-05, "loss": 0.0059, "step": 55210 }, { "grad_norm": 0.09749659150838852, "learning_rate": 3.6324998548754615e-05, "loss": 0.0067, "step": 55220 }, { "grad_norm": 0.06865457445383072, "learning_rate": 3.630746525200759e-05, "loss": 0.0074, "step": 55230 }, { "grad_norm": 0.09911804646253586, "learning_rate": 3.6289933775302873e-05, "loss": 0.0077, "step": 55240 }, { "grad_norm": 0.05489727109670639, "learning_rate": 3.6272404120970806e-05, "loss": 0.0059, "step": 55250 }, { "grad_norm": 0.06808803230524063, "learning_rate": 3.6254876291341467e-05, "loss": 0.0056, "step": 55260 }, { "grad_norm": 0.046957992017269135, "learning_rate": 3.62373502887447e-05, "loss": 0.0069, "step": 55270 }, { "grad_norm": 0.06679075211286545, "learning_rate": 3.621982611551007e-05, "loss": 0.0086, "step": 55280 }, { "grad_norm": 0.07306843250989914, "learning_rate": 3.620230377396697e-05, "loss": 0.0056, "step": 55290 }, { "grad_norm": 0.06737945228815079, "learning_rate": 3.6184783266444486e-05, "loss": 0.006, "step": 55300 }, { "grad_norm": 0.07729583978652954, "learning_rate": 3.6167264595271475e-05, "loss": 0.0056, "step": 55310 }, { "grad_norm": 0.05436200648546219, "learning_rate": 3.6149747762776566e-05, "loss": 0.0079, "step": 55320 }, { "grad_norm": 0.08169415593147278, "learning_rate": 3.6132232771288144e-05, "loss": 0.0054, "step": 55330 }, { "grad_norm": 0.0976620763540268, "learning_rate": 3.611471962313433e-05, "loss": 0.0059, "step": 55340 }, { "grad_norm": 0.08941112458705902, "learning_rate": 3.6097208320643034e-05, "loss": 0.0054, "step": 55350 }, { "grad_norm": 0.08962540328502655, "learning_rate": 3.607969886614186e-05, "loss": 0.007, "step": 55360 }, { "grad_norm": 0.05950105935335159, "learning_rate": 3.606219126195823e-05, "loss": 0.0094, "step": 55370 }, { "grad_norm": 0.062483515590429306, "learning_rate": 3.60446855104193e-05, "loss": 0.0056, "step": 55380 }, { "grad_norm": 0.09304115921258926, "learning_rate": 3.602718161385193e-05, "loss": 0.0058, "step": 55390 }, { "grad_norm": 0.06949125975370407, "learning_rate": 3.600967957458283e-05, "loss": 0.0046, "step": 55400 }, { "grad_norm": 0.09642451256513596, "learning_rate": 3.599217939493839e-05, "loss": 0.0087, "step": 55410 }, { "grad_norm": 0.057160839438438416, "learning_rate": 3.597468107724476e-05, "loss": 0.0097, "step": 55420 }, { "grad_norm": 0.09547062963247299, "learning_rate": 3.5957184623827865e-05, "loss": 0.0088, "step": 55430 }, { "grad_norm": 0.06709519773721695, "learning_rate": 3.593969003701339e-05, "loss": 0.0084, "step": 55440 }, { "grad_norm": 0.04854750633239746, "learning_rate": 3.592219731912673e-05, "loss": 0.0052, "step": 55450 }, { "grad_norm": 0.06006234511733055, "learning_rate": 3.590470647249306e-05, "loss": 0.0071, "step": 55460 }, { "grad_norm": 0.039962828159332275, "learning_rate": 3.58872174994373e-05, "loss": 0.0058, "step": 55470 }, { "grad_norm": 0.0790807381272316, "learning_rate": 3.586973040228414e-05, "loss": 0.0064, "step": 55480 }, { "grad_norm": 0.059869930148124695, "learning_rate": 3.5852245183358e-05, "loss": 0.0062, "step": 55490 }, { "grad_norm": 0.05503839626908302, "learning_rate": 3.5834761844983044e-05, "loss": 0.0071, "step": 55500 }, { "grad_norm": 0.09746197611093521, "learning_rate": 3.58172803894832e-05, "loss": 0.0058, "step": 55510 }, { "grad_norm": 0.06514082849025726, "learning_rate": 3.5799800819182154e-05, "loss": 0.0084, "step": 55520 }, { "grad_norm": 0.049752023071050644, "learning_rate": 3.578232313640331e-05, "loss": 0.0049, "step": 55530 }, { "grad_norm": 0.06874454766511917, "learning_rate": 3.5764847343469836e-05, "loss": 0.0069, "step": 55540 }, { "grad_norm": 0.07823340594768524, "learning_rate": 3.574737344270468e-05, "loss": 0.011, "step": 55550 }, { "grad_norm": 0.06154141575098038, "learning_rate": 3.57299014364305e-05, "loss": 0.0048, "step": 55560 }, { "grad_norm": 0.042567115277051926, "learning_rate": 3.571243132696972e-05, "loss": 0.0082, "step": 55570 }, { "grad_norm": 0.07388624548912048, "learning_rate": 3.569496311664449e-05, "loss": 0.0065, "step": 55580 }, { "grad_norm": 0.05831398442387581, "learning_rate": 3.567749680777673e-05, "loss": 0.0097, "step": 55590 }, { "grad_norm": 0.07628022879362106, "learning_rate": 3.5660032402688125e-05, "loss": 0.0057, "step": 55600 }, { "grad_norm": 0.09414385259151459, "learning_rate": 3.564256990370004e-05, "loss": 0.0045, "step": 55610 }, { "grad_norm": 0.07224497199058533, "learning_rate": 3.5625109313133644e-05, "loss": 0.0073, "step": 55620 }, { "grad_norm": 0.07252707332372665, "learning_rate": 3.560765063330985e-05, "loss": 0.007, "step": 55630 }, { "grad_norm": 0.06570247560739517, "learning_rate": 3.559019386654931e-05, "loss": 0.0079, "step": 55640 }, { "grad_norm": 0.062808096408844, "learning_rate": 3.55727390151724e-05, "loss": 0.0065, "step": 55650 }, { "grad_norm": 0.04112448915839195, "learning_rate": 3.555528608149927e-05, "loss": 0.0066, "step": 55660 }, { "grad_norm": 0.06485474854707718, "learning_rate": 3.553783506784979e-05, "loss": 0.0038, "step": 55670 }, { "grad_norm": 0.047710273414850235, "learning_rate": 3.552038597654359e-05, "loss": 0.0056, "step": 55680 }, { "grad_norm": 0.08484107255935669, "learning_rate": 3.5502938809900045e-05, "loss": 0.0056, "step": 55690 }, { "grad_norm": 0.03634876012802124, "learning_rate": 3.548549357023828e-05, "loss": 0.0046, "step": 55700 }, { "grad_norm": 0.045609746128320694, "learning_rate": 3.546805025987715e-05, "loss": 0.0047, "step": 55710 }, { "grad_norm": 0.07602091133594513, "learning_rate": 3.545060888113525e-05, "loss": 0.0058, "step": 55720 }, { "grad_norm": 0.05559998005628586, "learning_rate": 3.543316943633094e-05, "loss": 0.0068, "step": 55730 }, { "grad_norm": 0.06320271641016006, "learning_rate": 3.541573192778231e-05, "loss": 0.0052, "step": 55740 }, { "grad_norm": 0.05968358367681503, "learning_rate": 3.539829635780717e-05, "loss": 0.0076, "step": 55750 }, { "grad_norm": 0.07123459875583649, "learning_rate": 3.538086272872313e-05, "loss": 0.0092, "step": 55760 }, { "grad_norm": 0.049145761877298355, "learning_rate": 3.536343104284746e-05, "loss": 0.0066, "step": 55770 }, { "grad_norm": 0.08815789222717285, "learning_rate": 3.5346001302497264e-05, "loss": 0.009, "step": 55780 }, { "grad_norm": 0.0593557134270668, "learning_rate": 3.5328573509989334e-05, "loss": 0.0078, "step": 55790 }, { "grad_norm": 0.04356683790683746, "learning_rate": 3.531114766764019e-05, "loss": 0.0079, "step": 55800 }, { "grad_norm": 0.05995059385895729, "learning_rate": 3.529372377776613e-05, "loss": 0.0064, "step": 55810 }, { "grad_norm": 0.06108870357275009, "learning_rate": 3.527630184268318e-05, "loss": 0.0095, "step": 55820 }, { "grad_norm": 0.05610988661646843, "learning_rate": 3.5258881864707086e-05, "loss": 0.0074, "step": 55830 }, { "grad_norm": 0.06893851608037949, "learning_rate": 3.5241463846153344e-05, "loss": 0.0087, "step": 55840 }, { "grad_norm": 0.06596194952726364, "learning_rate": 3.522404778933723e-05, "loss": 0.0048, "step": 55850 }, { "grad_norm": 0.07300224900245667, "learning_rate": 3.52066336965737e-05, "loss": 0.0058, "step": 55860 }, { "grad_norm": 0.06095558777451515, "learning_rate": 3.51892215701775e-05, "loss": 0.0059, "step": 55870 }, { "grad_norm": 0.11027462035417557, "learning_rate": 3.517181141246305e-05, "loss": 0.0073, "step": 55880 }, { "grad_norm": 0.058002445846796036, "learning_rate": 3.515440322574457e-05, "loss": 0.0077, "step": 55890 }, { "grad_norm": 0.07131741940975189, "learning_rate": 3.513699701233599e-05, "loss": 0.0056, "step": 55900 }, { "grad_norm": 0.06845853477716446, "learning_rate": 3.511959277455097e-05, "loss": 0.0056, "step": 55910 }, { "grad_norm": 0.06717099994421005, "learning_rate": 3.510219051470296e-05, "loss": 0.0086, "step": 55920 }, { "grad_norm": 0.06634271889925003, "learning_rate": 3.5084790235105066e-05, "loss": 0.0056, "step": 55930 }, { "grad_norm": 0.06225180625915527, "learning_rate": 3.50673919380702e-05, "loss": 0.0048, "step": 55940 }, { "grad_norm": 0.058442071080207825, "learning_rate": 3.504999562591096e-05, "loss": 0.0074, "step": 55950 }, { "grad_norm": 0.058961790055036545, "learning_rate": 3.5032601300939724e-05, "loss": 0.0043, "step": 55960 }, { "grad_norm": 0.07474518567323685, "learning_rate": 3.501520896546857e-05, "loss": 0.0073, "step": 55970 }, { "grad_norm": 0.06907336413860321, "learning_rate": 3.499781862180933e-05, "loss": 0.0106, "step": 55980 }, { "grad_norm": 0.06468256562948227, "learning_rate": 3.4980430272273554e-05, "loss": 0.0077, "step": 55990 }, { "grad_norm": 0.08576156944036484, "learning_rate": 3.4963043919172556e-05, "loss": 0.0069, "step": 56000 }, { "grad_norm": 0.06643717736005783, "learning_rate": 3.4945659564817386e-05, "loss": 0.0074, "step": 56010 }, { "grad_norm": 0.08073663711547852, "learning_rate": 3.4928277211518777e-05, "loss": 0.0068, "step": 56020 }, { "grad_norm": 0.033101875334978104, "learning_rate": 3.4910896861587244e-05, "loss": 0.0057, "step": 56030 }, { "grad_norm": 0.06126493588089943, "learning_rate": 3.489351851733303e-05, "loss": 0.0085, "step": 56040 }, { "grad_norm": 0.04748626425862312, "learning_rate": 3.48761421810661e-05, "loss": 0.0056, "step": 56050 }, { "grad_norm": 0.05934471637010574, "learning_rate": 3.4858767855096134e-05, "loss": 0.0048, "step": 56060 }, { "grad_norm": 0.05606921762228012, "learning_rate": 3.484139554173259e-05, "loss": 0.0069, "step": 56070 }, { "grad_norm": 0.07453880459070206, "learning_rate": 3.4824025243284626e-05, "loss": 0.0084, "step": 56080 }, { "grad_norm": 0.05264711007475853, "learning_rate": 3.480665696206116e-05, "loss": 0.0049, "step": 56090 }, { "grad_norm": 0.04918116331100464, "learning_rate": 3.4789290700370793e-05, "loss": 0.0081, "step": 56100 }, { "grad_norm": 0.07744982093572617, "learning_rate": 3.4771926460521896e-05, "loss": 0.0099, "step": 56110 }, { "grad_norm": 0.09579673409461975, "learning_rate": 3.4754564244822584e-05, "loss": 0.006, "step": 56120 }, { "grad_norm": 0.10273150354623795, "learning_rate": 3.473720405558065e-05, "loss": 0.0056, "step": 56130 }, { "grad_norm": 0.0434364378452301, "learning_rate": 3.4719845895103655e-05, "loss": 0.0052, "step": 56140 }, { "grad_norm": 0.052897859364748, "learning_rate": 3.470248976569891e-05, "loss": 0.0085, "step": 56150 }, { "grad_norm": 0.057626497000455856, "learning_rate": 3.46851356696734e-05, "loss": 0.0055, "step": 56160 }, { "grad_norm": 0.03783398121595383, "learning_rate": 3.466778360933389e-05, "loss": 0.0068, "step": 56170 }, { "grad_norm": 0.08450217545032501, "learning_rate": 3.4650433586986866e-05, "loss": 0.0103, "step": 56180 }, { "grad_norm": 0.07271207869052887, "learning_rate": 3.46330856049385e-05, "loss": 0.006, "step": 56190 }, { "grad_norm": 0.044636957347393036, "learning_rate": 3.461573966549473e-05, "loss": 0.0047, "step": 56200 }, { "grad_norm": 0.040657322853803635, "learning_rate": 3.459839577096122e-05, "loss": 0.0093, "step": 56210 }, { "grad_norm": 0.0567576102912426, "learning_rate": 3.458105392364338e-05, "loss": 0.0089, "step": 56220 }, { "grad_norm": 0.05872470512986183, "learning_rate": 3.456371412584632e-05, "loss": 0.0055, "step": 56230 }, { "grad_norm": 0.06173586845397949, "learning_rate": 3.454637637987487e-05, "loss": 0.0084, "step": 56240 }, { "grad_norm": 0.05276792496442795, "learning_rate": 3.4529040688033605e-05, "loss": 0.008, "step": 56250 }, { "grad_norm": 0.04117046669125557, "learning_rate": 3.451170705262684e-05, "loss": 0.0072, "step": 56260 }, { "grad_norm": 0.09281236678361893, "learning_rate": 3.449437547595857e-05, "loss": 0.0087, "step": 56270 }, { "grad_norm": 0.057971738278865814, "learning_rate": 3.4477045960332574e-05, "loss": 0.0066, "step": 56280 }, { "grad_norm": 0.09603414684534073, "learning_rate": 3.445971850805231e-05, "loss": 0.007, "step": 56290 }, { "grad_norm": 0.04189387336373329, "learning_rate": 3.4442393121421e-05, "loss": 0.0076, "step": 56300 }, { "grad_norm": 0.05981616675853729, "learning_rate": 3.4425069802741586e-05, "loss": 0.0051, "step": 56310 }, { "grad_norm": 0.04628530144691467, "learning_rate": 3.4407748554316674e-05, "loss": 0.0064, "step": 56320 }, { "grad_norm": 0.06351269781589508, "learning_rate": 3.4390429378448684e-05, "loss": 0.0063, "step": 56330 }, { "grad_norm": 0.035105928778648376, "learning_rate": 3.437311227743971e-05, "loss": 0.0059, "step": 56340 }, { "grad_norm": 0.0561131127178669, "learning_rate": 3.4355797253591574e-05, "loss": 0.0075, "step": 56350 }, { "grad_norm": 0.0513286218047142, "learning_rate": 3.433848430920581e-05, "loss": 0.0063, "step": 56360 }, { "grad_norm": 0.0739155039191246, "learning_rate": 3.4321173446583734e-05, "loss": 0.0065, "step": 56370 }, { "grad_norm": 0.11824086308479309, "learning_rate": 3.430386466802632e-05, "loss": 0.0058, "step": 56380 }, { "grad_norm": 0.08616815507411957, "learning_rate": 3.42865579758343e-05, "loss": 0.0049, "step": 56390 }, { "grad_norm": 0.03341479226946831, "learning_rate": 3.4269253372308106e-05, "loss": 0.0043, "step": 56400 }, { "grad_norm": 0.08456042408943176, "learning_rate": 3.425195085974791e-05, "loss": 0.0047, "step": 56410 }, { "grad_norm": 0.07142073661088943, "learning_rate": 3.42346504404536e-05, "loss": 0.0051, "step": 56420 }, { "grad_norm": 0.08587011694908142, "learning_rate": 3.42173521167248e-05, "loss": 0.0055, "step": 56430 }, { "grad_norm": 0.04917142167687416, "learning_rate": 3.42000558908608e-05, "loss": 0.0047, "step": 56440 }, { "grad_norm": 0.08905211091041565, "learning_rate": 3.4182761765160695e-05, "loss": 0.0049, "step": 56450 }, { "grad_norm": 0.06861906498670578, "learning_rate": 3.416546974192325e-05, "loss": 0.0064, "step": 56460 }, { "grad_norm": 0.0512971393764019, "learning_rate": 3.414817982344696e-05, "loss": 0.0058, "step": 56470 }, { "grad_norm": 0.0521206259727478, "learning_rate": 3.4130892012030036e-05, "loss": 0.0064, "step": 56480 }, { "grad_norm": 0.07022189348936081, "learning_rate": 3.41136063099704e-05, "loss": 0.0088, "step": 56490 }, { "grad_norm": 0.058866605162620544, "learning_rate": 3.409632271956573e-05, "loss": 0.0076, "step": 56500 }, { "grad_norm": 0.07958804070949554, "learning_rate": 3.4079041243113366e-05, "loss": 0.0059, "step": 56510 }, { "grad_norm": 0.07564366608858109, "learning_rate": 3.406176188291044e-05, "loss": 0.0111, "step": 56520 }, { "grad_norm": 0.0819779708981514, "learning_rate": 3.404448464125375e-05, "loss": 0.0078, "step": 56530 }, { "grad_norm": 0.0373232364654541, "learning_rate": 3.402720952043981e-05, "loss": 0.0053, "step": 56540 }, { "grad_norm": 0.05435990169644356, "learning_rate": 3.4009936522764875e-05, "loss": 0.0049, "step": 56550 }, { "grad_norm": 0.06468697637319565, "learning_rate": 3.399266565052493e-05, "loss": 0.0065, "step": 56560 }, { "grad_norm": 0.04604044556617737, "learning_rate": 3.397539690601563e-05, "loss": 0.0075, "step": 56570 }, { "grad_norm": 0.049846772104501724, "learning_rate": 3.3958130291532374e-05, "loss": 0.0048, "step": 56580 }, { "grad_norm": 0.06137274578213692, "learning_rate": 3.39408658093703e-05, "loss": 0.0061, "step": 56590 }, { "grad_norm": 0.055898915976285934, "learning_rate": 3.392360346182424e-05, "loss": 0.0051, "step": 56600 }, { "grad_norm": 0.06672581285238266, "learning_rate": 3.390634325118874e-05, "loss": 0.0084, "step": 56610 }, { "grad_norm": 0.039171554148197174, "learning_rate": 3.388908517975805e-05, "loss": 0.0072, "step": 56620 }, { "grad_norm": 0.0848672091960907, "learning_rate": 3.387182924982617e-05, "loss": 0.0061, "step": 56630 }, { "grad_norm": 0.0449136346578598, "learning_rate": 3.38545754636868e-05, "loss": 0.0054, "step": 56640 }, { "grad_norm": 0.04745182767510414, "learning_rate": 3.3837323823633335e-05, "loss": 0.0061, "step": 56650 }, { "grad_norm": 0.07061821222305298, "learning_rate": 3.3820074331958894e-05, "loss": 0.0058, "step": 56660 }, { "grad_norm": 0.0476793497800827, "learning_rate": 3.3802826990956346e-05, "loss": 0.0087, "step": 56670 }, { "grad_norm": 0.07789476215839386, "learning_rate": 3.378558180291823e-05, "loss": 0.0052, "step": 56680 }, { "grad_norm": 0.04515375941991806, "learning_rate": 3.3768338770136834e-05, "loss": 0.0066, "step": 56690 }, { "grad_norm": 0.06221180036664009, "learning_rate": 3.375109789490411e-05, "loss": 0.005, "step": 56700 }, { "grad_norm": 0.08384466171264648, "learning_rate": 3.3733859179511774e-05, "loss": 0.0068, "step": 56710 }, { "grad_norm": 0.09963040798902512, "learning_rate": 3.371662262625123e-05, "loss": 0.0066, "step": 56720 }, { "grad_norm": 0.06788252294063568, "learning_rate": 3.3699388237413576e-05, "loss": 0.006, "step": 56730 }, { "grad_norm": 0.04671267792582512, "learning_rate": 3.368215601528969e-05, "loss": 0.004, "step": 56740 }, { "grad_norm": 0.04486500844359398, "learning_rate": 3.3664925962170094e-05, "loss": 0.0062, "step": 56750 }, { "grad_norm": 0.03168084844946861, "learning_rate": 3.364769808034505e-05, "loss": 0.007, "step": 56760 }, { "grad_norm": 0.03455647453665733, "learning_rate": 3.363047237210451e-05, "loss": 0.0057, "step": 56770 }, { "grad_norm": 0.057588476687669754, "learning_rate": 3.361324883973818e-05, "loss": 0.0085, "step": 56780 }, { "grad_norm": 0.042054157704114914, "learning_rate": 3.359602748553543e-05, "loss": 0.0062, "step": 56790 }, { "grad_norm": 0.05667293071746826, "learning_rate": 3.357880831178537e-05, "loss": 0.0051, "step": 56800 }, { "grad_norm": 0.07951052486896515, "learning_rate": 3.3561591320776794e-05, "loss": 0.0065, "step": 56810 }, { "grad_norm": 0.044687096029520035, "learning_rate": 3.354437651479825e-05, "loss": 0.0057, "step": 56820 }, { "grad_norm": 0.046976104378700256, "learning_rate": 3.352716389613796e-05, "loss": 0.0054, "step": 56830 }, { "grad_norm": 0.04700803384184837, "learning_rate": 3.350995346708385e-05, "loss": 0.0052, "step": 56840 }, { "grad_norm": 0.07340484112501144, "learning_rate": 3.349274522992359e-05, "loss": 0.0056, "step": 56850 }, { "grad_norm": 0.039011456072330475, "learning_rate": 3.3475539186944523e-05, "loss": 0.006, "step": 56860 }, { "grad_norm": 0.07590200752019882, "learning_rate": 3.345833534043372e-05, "loss": 0.0043, "step": 56870 }, { "grad_norm": 0.04118480533361435, "learning_rate": 3.3441133692677934e-05, "loss": 0.0065, "step": 56880 }, { "grad_norm": 0.07246040552854538, "learning_rate": 3.342393424596368e-05, "loss": 0.0053, "step": 56890 }, { "grad_norm": 0.05302250012755394, "learning_rate": 3.340673700257713e-05, "loss": 0.0052, "step": 56900 }, { "grad_norm": 0.06427241116762161, "learning_rate": 3.3389541964804186e-05, "loss": 0.0063, "step": 56910 }, { "grad_norm": 0.04031845182180405, "learning_rate": 3.3372349134930444e-05, "loss": 0.0041, "step": 56920 }, { "grad_norm": 0.0452989861369133, "learning_rate": 3.335515851524121e-05, "loss": 0.007, "step": 56930 }, { "grad_norm": 0.051019567996263504, "learning_rate": 3.333797010802151e-05, "loss": 0.0055, "step": 56940 }, { "grad_norm": 0.07263842970132828, "learning_rate": 3.332078391555608e-05, "loss": 0.0064, "step": 56950 }, { "grad_norm": 0.06333918124437332, "learning_rate": 3.330359994012929e-05, "loss": 0.0058, "step": 56960 }, { "grad_norm": 0.07042280584573746, "learning_rate": 3.3286418184025324e-05, "loss": 0.0073, "step": 56970 }, { "grad_norm": 0.07247230410575867, "learning_rate": 3.326923864952801e-05, "loss": 0.0045, "step": 56980 }, { "grad_norm": 0.11894761770963669, "learning_rate": 3.3252061338920894e-05, "loss": 0.0076, "step": 56990 }, { "grad_norm": 0.058491598814725876, "learning_rate": 3.323488625448722e-05, "loss": 0.0084, "step": 57000 }, { "grad_norm": 0.08268789201974869, "learning_rate": 3.321771339850992e-05, "loss": 0.0072, "step": 57010 }, { "grad_norm": 0.051218096166849136, "learning_rate": 3.320054277327166e-05, "loss": 0.0059, "step": 57020 }, { "grad_norm": 0.04638213664293289, "learning_rate": 3.318337438105479e-05, "loss": 0.0063, "step": 57030 }, { "grad_norm": 0.037989210337400436, "learning_rate": 3.3166208224141394e-05, "loss": 0.0045, "step": 57040 }, { "grad_norm": 0.06104429438710213, "learning_rate": 3.3149044304813227e-05, "loss": 0.0051, "step": 57050 }, { "grad_norm": 0.04534027725458145, "learning_rate": 3.313188262535175e-05, "loss": 0.0059, "step": 57060 }, { "grad_norm": 0.06154631823301315, "learning_rate": 3.311472318803813e-05, "loss": 0.0058, "step": 57070 }, { "grad_norm": 0.062374047935009, "learning_rate": 3.3097565995153244e-05, "loss": 0.0068, "step": 57080 }, { "grad_norm": 0.0638267919421196, "learning_rate": 3.3080411048977653e-05, "loss": 0.0065, "step": 57090 }, { "grad_norm": 0.05420402064919472, "learning_rate": 3.306325835179164e-05, "loss": 0.004, "step": 57100 }, { "grad_norm": 0.04731832444667816, "learning_rate": 3.304610790587518e-05, "loss": 0.0069, "step": 57110 }, { "grad_norm": 0.05710204690694809, "learning_rate": 3.302895971350795e-05, "loss": 0.006, "step": 57120 }, { "grad_norm": 0.0669003278017044, "learning_rate": 3.3011813776969326e-05, "loss": 0.0088, "step": 57130 }, { "grad_norm": 0.05706474557518959, "learning_rate": 3.2994670098538374e-05, "loss": 0.0059, "step": 57140 }, { "grad_norm": 0.05973603203892708, "learning_rate": 3.297752868049388e-05, "loss": 0.0105, "step": 57150 }, { "grad_norm": 0.04959335923194885, "learning_rate": 3.296038952511433e-05, "loss": 0.0076, "step": 57160 }, { "grad_norm": 0.04213808849453926, "learning_rate": 3.2943252634677866e-05, "loss": 0.0041, "step": 57170 }, { "grad_norm": 0.03713872656226158, "learning_rate": 3.292611801146238e-05, "loss": 0.0071, "step": 57180 }, { "grad_norm": 0.05904383212327957, "learning_rate": 3.290898565774546e-05, "loss": 0.0071, "step": 57190 }, { "grad_norm": 0.046164728701114655, "learning_rate": 3.289185557580435e-05, "loss": 0.0086, "step": 57200 }, { "grad_norm": 0.05536703020334244, "learning_rate": 3.287472776791606e-05, "loss": 0.0044, "step": 57210 }, { "grad_norm": 0.04981476441025734, "learning_rate": 3.2857602236357206e-05, "loss": 0.0056, "step": 57220 }, { "grad_norm": 0.07584971189498901, "learning_rate": 3.284047898340418e-05, "loss": 0.0062, "step": 57230 }, { "grad_norm": 0.06251054257154465, "learning_rate": 3.282335801133303e-05, "loss": 0.0065, "step": 57240 }, { "grad_norm": 0.057031042873859406, "learning_rate": 3.280623932241951e-05, "loss": 0.0069, "step": 57250 }, { "grad_norm": 0.07025604695081711, "learning_rate": 3.27891229189391e-05, "loss": 0.0065, "step": 57260 }, { "grad_norm": 0.0394587367773056, "learning_rate": 3.2772008803166946e-05, "loss": 0.0041, "step": 57270 }, { "grad_norm": 0.040343113243579865, "learning_rate": 3.275489697737788e-05, "loss": 0.0065, "step": 57280 }, { "grad_norm": 0.03819020092487335, "learning_rate": 3.273778744384644e-05, "loss": 0.0046, "step": 57290 }, { "grad_norm": 0.0818929523229599, "learning_rate": 3.272068020484689e-05, "loss": 0.006, "step": 57300 }, { "grad_norm": 0.053542591631412506, "learning_rate": 3.270357526265313e-05, "loss": 0.0058, "step": 57310 }, { "grad_norm": 0.07315284013748169, "learning_rate": 3.26864726195388e-05, "loss": 0.0053, "step": 57320 }, { "grad_norm": 0.042411718517541885, "learning_rate": 3.266937227777721e-05, "loss": 0.0068, "step": 57330 }, { "grad_norm": 0.08974552154541016, "learning_rate": 3.2652274239641404e-05, "loss": 0.0102, "step": 57340 }, { "grad_norm": 0.03742947056889534, "learning_rate": 3.2635178507404084e-05, "loss": 0.005, "step": 57350 }, { "grad_norm": 0.04316761717200279, "learning_rate": 3.2618085083337645e-05, "loss": 0.0042, "step": 57360 }, { "grad_norm": 0.05810673534870148, "learning_rate": 3.260099396971417e-05, "loss": 0.0068, "step": 57370 }, { "grad_norm": 0.05240597948431969, "learning_rate": 3.2583905168805486e-05, "loss": 0.0054, "step": 57380 }, { "grad_norm": 0.0541398823261261, "learning_rate": 3.256681868288303e-05, "loss": 0.0055, "step": 57390 }, { "grad_norm": 0.04690299928188324, "learning_rate": 3.2549734514217986e-05, "loss": 0.0057, "step": 57400 }, { "grad_norm": 0.054113246500492096, "learning_rate": 3.253265266508124e-05, "loss": 0.0057, "step": 57410 }, { "grad_norm": 0.08670862764120102, "learning_rate": 3.2515573137743336e-05, "loss": 0.0066, "step": 57420 }, { "grad_norm": 0.06415161490440369, "learning_rate": 3.2498495934474544e-05, "loss": 0.0055, "step": 57430 }, { "grad_norm": 0.08606522530317307, "learning_rate": 3.248142105754477e-05, "loss": 0.0064, "step": 57440 }, { "grad_norm": 0.061018310487270355, "learning_rate": 3.246434850922367e-05, "loss": 0.0064, "step": 57450 }, { "grad_norm": 0.047121062874794006, "learning_rate": 3.244727829178056e-05, "loss": 0.0058, "step": 57460 }, { "grad_norm": 0.054099854081869125, "learning_rate": 3.243021040748444e-05, "loss": 0.0067, "step": 57470 }, { "grad_norm": 0.06685948371887207, "learning_rate": 3.2413144858604004e-05, "loss": 0.0053, "step": 57480 }, { "grad_norm": 0.1562899351119995, "learning_rate": 3.239608164740767e-05, "loss": 0.0074, "step": 57490 }, { "grad_norm": 0.0762251615524292, "learning_rate": 3.2379020776163514e-05, "loss": 0.0094, "step": 57500 }, { "grad_norm": 0.04706162214279175, "learning_rate": 3.2361962247139286e-05, "loss": 0.0098, "step": 57510 }, { "grad_norm": 0.06888092309236526, "learning_rate": 3.2344906062602476e-05, "loss": 0.0066, "step": 57520 }, { "grad_norm": 0.03989217430353165, "learning_rate": 3.232785222482019e-05, "loss": 0.005, "step": 57530 }, { "grad_norm": 0.05622145161032677, "learning_rate": 3.231080073605929e-05, "loss": 0.0069, "step": 57540 }, { "grad_norm": 0.07880410552024841, "learning_rate": 3.2293751598586276e-05, "loss": 0.0061, "step": 57550 }, { "grad_norm": 0.07644088566303253, "learning_rate": 3.227670481466738e-05, "loss": 0.007, "step": 57560 }, { "grad_norm": 0.07642596960067749, "learning_rate": 3.22596603865685e-05, "loss": 0.0058, "step": 57570 }, { "grad_norm": 0.059678517282009125, "learning_rate": 3.2242618316555204e-05, "loss": 0.0043, "step": 57580 }, { "grad_norm": 0.05047741159796715, "learning_rate": 3.2225578606892756e-05, "loss": 0.0062, "step": 57590 }, { "grad_norm": 0.08448737859725952, "learning_rate": 3.220854125984615e-05, "loss": 0.0056, "step": 57600 }, { "grad_norm": 0.05131746456027031, "learning_rate": 3.219150627767997e-05, "loss": 0.008, "step": 57610 }, { "grad_norm": 0.07404276728630066, "learning_rate": 3.217447366265858e-05, "loss": 0.0059, "step": 57620 }, { "grad_norm": 0.053051695227622986, "learning_rate": 3.2157443417046e-05, "loss": 0.0079, "step": 57630 }, { "grad_norm": 0.06517040729522705, "learning_rate": 3.214041554310592e-05, "loss": 0.0073, "step": 57640 }, { "grad_norm": 0.06718453764915466, "learning_rate": 3.212339004310172e-05, "loss": 0.0063, "step": 57650 }, { "grad_norm": 0.06295911222696304, "learning_rate": 3.210636691929646e-05, "loss": 0.0043, "step": 57660 }, { "grad_norm": 0.05551194027066231, "learning_rate": 3.20893461739529e-05, "loss": 0.0055, "step": 57670 }, { "grad_norm": 0.10440965741872787, "learning_rate": 3.2072327809333484e-05, "loss": 0.0076, "step": 57680 }, { "grad_norm": 0.08317975699901581, "learning_rate": 3.2055311827700316e-05, "loss": 0.0079, "step": 57690 }, { "grad_norm": 0.051497336477041245, "learning_rate": 3.2038298231315186e-05, "loss": 0.0061, "step": 57700 }, { "grad_norm": 0.08115284889936447, "learning_rate": 3.202128702243962e-05, "loss": 0.0091, "step": 57710 }, { "grad_norm": 0.06113869324326515, "learning_rate": 3.200427820333476e-05, "loss": 0.0043, "step": 57720 }, { "grad_norm": 0.04921473562717438, "learning_rate": 3.1987271776261466e-05, "loss": 0.0068, "step": 57730 }, { "grad_norm": 0.052132368087768555, "learning_rate": 3.197026774348025e-05, "loss": 0.0118, "step": 57740 }, { "grad_norm": 0.03623104840517044, "learning_rate": 3.195326610725135e-05, "loss": 0.0057, "step": 57750 }, { "grad_norm": 0.08383197337388992, "learning_rate": 3.193626686983465e-05, "loss": 0.0097, "step": 57760 }, { "grad_norm": 0.027542881667613983, "learning_rate": 3.191927003348971e-05, "loss": 0.0039, "step": 57770 }, { "grad_norm": 0.04757262393832207, "learning_rate": 3.190227560047582e-05, "loss": 0.005, "step": 57780 }, { "grad_norm": 0.04732019454240799, "learning_rate": 3.188528357305192e-05, "loss": 0.0052, "step": 57790 }, { "grad_norm": 0.0702483281493187, "learning_rate": 3.186829395347659e-05, "loss": 0.0087, "step": 57800 }, { "grad_norm": 0.05601602420210838, "learning_rate": 3.185130674400816e-05, "loss": 0.0067, "step": 57810 }, { "grad_norm": 0.05273683741688728, "learning_rate": 3.18343219469046e-05, "loss": 0.0084, "step": 57820 }, { "grad_norm": 0.054061997681856155, "learning_rate": 3.181733956442357e-05, "loss": 0.0053, "step": 57830 }, { "grad_norm": 0.04472365230321884, "learning_rate": 3.180035959882238e-05, "loss": 0.0058, "step": 57840 }, { "grad_norm": 0.05292631685733795, "learning_rate": 3.178338205235807e-05, "loss": 0.0055, "step": 57850 }, { "grad_norm": 0.09820600599050522, "learning_rate": 3.176640692728733e-05, "loss": 0.0083, "step": 57860 }, { "grad_norm": 0.084278404712677, "learning_rate": 3.174943422586655e-05, "loss": 0.005, "step": 57870 }, { "grad_norm": 0.049664322286844254, "learning_rate": 3.1732463950351756e-05, "loss": 0.0049, "step": 57880 }, { "grad_norm": 0.03804457560181618, "learning_rate": 3.1715496102998665e-05, "loss": 0.0071, "step": 57890 }, { "grad_norm": 0.051846694201231, "learning_rate": 3.1698530686062715e-05, "loss": 0.0067, "step": 57900 }, { "grad_norm": 0.05914030596613884, "learning_rate": 3.168156770179896e-05, "loss": 0.0067, "step": 57910 }, { "grad_norm": 0.06984177976846695, "learning_rate": 3.166460715246215e-05, "loss": 0.0084, "step": 57920 }, { "grad_norm": 0.062248893082141876, "learning_rate": 3.1647649040306746e-05, "loss": 0.0068, "step": 57930 }, { "grad_norm": 0.04636772349476814, "learning_rate": 3.163069336758685e-05, "loss": 0.0057, "step": 57940 }, { "grad_norm": 0.08088839054107666, "learning_rate": 3.1613740136556254e-05, "loss": 0.0049, "step": 57950 }, { "grad_norm": 0.05500953271985054, "learning_rate": 3.15967893494684e-05, "loss": 0.006, "step": 57960 }, { "grad_norm": 0.05364494025707245, "learning_rate": 3.157984100857643e-05, "loss": 0.0066, "step": 57970 }, { "grad_norm": 0.08011262118816376, "learning_rate": 3.156289511613318e-05, "loss": 0.0061, "step": 57980 }, { "grad_norm": 0.0894855409860611, "learning_rate": 3.15459516743911e-05, "loss": 0.0065, "step": 57990 }, { "grad_norm": 0.055141352117061615, "learning_rate": 3.1529010685602355e-05, "loss": 0.0048, "step": 58000 }, { "grad_norm": 0.05228957161307335, "learning_rate": 3.15120721520188e-05, "loss": 0.0075, "step": 58010 }, { "grad_norm": 0.04934690520167351, "learning_rate": 3.149513607589194e-05, "loss": 0.0047, "step": 58020 }, { "grad_norm": 0.06037350371479988, "learning_rate": 3.147820245947294e-05, "loss": 0.0063, "step": 58030 }, { "grad_norm": 0.06711124628782272, "learning_rate": 3.146127130501268e-05, "loss": 0.0052, "step": 58040 }, { "grad_norm": 0.04077303037047386, "learning_rate": 3.1444342614761654e-05, "loss": 0.0057, "step": 58050 }, { "grad_norm": 0.05202368274331093, "learning_rate": 3.142741639097009e-05, "loss": 0.0057, "step": 58060 }, { "grad_norm": 0.08124766498804092, "learning_rate": 3.1410492635887825e-05, "loss": 0.0067, "step": 58070 }, { "grad_norm": 0.044913288205862045, "learning_rate": 3.139357135176444e-05, "loss": 0.006, "step": 58080 }, { "grad_norm": 0.0446372926235199, "learning_rate": 3.137665254084914e-05, "loss": 0.0062, "step": 58090 }, { "grad_norm": 0.05192272737622261, "learning_rate": 3.13597362053908e-05, "loss": 0.0078, "step": 58100 }, { "grad_norm": 0.06850770860910416, "learning_rate": 3.1342822347637976e-05, "loss": 0.0074, "step": 58110 }, { "grad_norm": 0.07147887349128723, "learning_rate": 3.132591096983892e-05, "loss": 0.0061, "step": 58120 }, { "grad_norm": 0.04267694428563118, "learning_rate": 3.13090020742415e-05, "loss": 0.0069, "step": 58130 }, { "grad_norm": 0.06312411278486252, "learning_rate": 3.1292095663093276e-05, "loss": 0.0059, "step": 58140 }, { "grad_norm": 0.03442303463816643, "learning_rate": 3.1275191738641526e-05, "loss": 0.0042, "step": 58150 }, { "grad_norm": 0.11848074942827225, "learning_rate": 3.125829030313314e-05, "loss": 0.0104, "step": 58160 }, { "grad_norm": 0.0774240717291832, "learning_rate": 3.124139135881471e-05, "loss": 0.0048, "step": 58170 }, { "grad_norm": 0.04623868688941002, "learning_rate": 3.122449490793244e-05, "loss": 0.0045, "step": 58180 }, { "grad_norm": 0.06393708288669586, "learning_rate": 3.1207600952732274e-05, "loss": 0.006, "step": 58190 }, { "grad_norm": 0.05257521569728851, "learning_rate": 3.11907094954598e-05, "loss": 0.0051, "step": 58200 }, { "grad_norm": 0.0456417016685009, "learning_rate": 3.117382053836024e-05, "loss": 0.0055, "step": 58210 }, { "grad_norm": 0.06147370859980583, "learning_rate": 3.115693408367853e-05, "loss": 0.0063, "step": 58220 }, { "grad_norm": 0.07772059738636017, "learning_rate": 3.114005013365925e-05, "loss": 0.0072, "step": 58230 }, { "grad_norm": 0.08553088456392288, "learning_rate": 3.112316869054666e-05, "loss": 0.0064, "step": 58240 }, { "grad_norm": 0.07122699916362762, "learning_rate": 3.110628975658469e-05, "loss": 0.0065, "step": 58250 }, { "grad_norm": 0.07042883336544037, "learning_rate": 3.1089413334016895e-05, "loss": 0.0055, "step": 58260 }, { "grad_norm": 0.057677336037158966, "learning_rate": 3.107253942508655e-05, "loss": 0.0074, "step": 58270 }, { "grad_norm": 0.06270434707403183, "learning_rate": 3.105566803203656e-05, "loss": 0.0065, "step": 58280 }, { "grad_norm": 0.045889053493738174, "learning_rate": 3.103879915710951e-05, "loss": 0.0082, "step": 58290 }, { "grad_norm": 0.07728685438632965, "learning_rate": 3.102193280254767e-05, "loss": 0.0052, "step": 58300 }, { "grad_norm": 0.06662626564502716, "learning_rate": 3.100506897059292e-05, "loss": 0.0078, "step": 58310 }, { "grad_norm": 0.06283221393823624, "learning_rate": 3.098820766348686e-05, "loss": 0.0068, "step": 58320 }, { "grad_norm": 0.05340716987848282, "learning_rate": 3.0971348883470716e-05, "loss": 0.0057, "step": 58330 }, { "grad_norm": 0.08022817224264145, "learning_rate": 3.0954492632785424e-05, "loss": 0.0076, "step": 58340 }, { "grad_norm": 0.08456141501665115, "learning_rate": 3.093763891367153e-05, "loss": 0.007, "step": 58350 }, { "grad_norm": 0.04977916553616524, "learning_rate": 3.092078772836927e-05, "loss": 0.0067, "step": 58360 }, { "grad_norm": 0.07791499048471451, "learning_rate": 3.0903939079118536e-05, "loss": 0.0065, "step": 58370 }, { "grad_norm": 0.07477842271327972, "learning_rate": 3.0887092968158905e-05, "loss": 0.0053, "step": 58380 }, { "grad_norm": 0.051917824894189835, "learning_rate": 3.08702493977296e-05, "loss": 0.0042, "step": 58390 }, { "grad_norm": 0.04642859473824501, "learning_rate": 3.0853408370069495e-05, "loss": 0.0042, "step": 58400 }, { "grad_norm": 0.042850248515605927, "learning_rate": 3.083656988741713e-05, "loss": 0.0053, "step": 58410 }, { "grad_norm": 0.0687752291560173, "learning_rate": 3.0819733952010744e-05, "loss": 0.0091, "step": 58420 }, { "grad_norm": 0.05578305199742317, "learning_rate": 3.0802900566088174e-05, "loss": 0.0063, "step": 58430 }, { "grad_norm": 0.06773930788040161, "learning_rate": 3.078606973188695e-05, "loss": 0.0054, "step": 58440 }, { "grad_norm": 0.04134071245789528, "learning_rate": 3.076924145164429e-05, "loss": 0.004, "step": 58450 }, { "grad_norm": 0.05064934492111206, "learning_rate": 3.075241572759704e-05, "loss": 0.0074, "step": 58460 }, { "grad_norm": 0.05089692771434784, "learning_rate": 3.073559256198171e-05, "loss": 0.0052, "step": 58470 }, { "grad_norm": 0.05830927938222885, "learning_rate": 3.071877195703446e-05, "loss": 0.0069, "step": 58480 }, { "grad_norm": 0.0642385259270668, "learning_rate": 3.070195391499114e-05, "loss": 0.005, "step": 58490 }, { "grad_norm": 0.0461038276553154, "learning_rate": 3.068513843808725e-05, "loss": 0.0045, "step": 58500 }, { "grad_norm": 0.07106690853834152, "learning_rate": 3.0668325528557904e-05, "loss": 0.0092, "step": 58510 }, { "grad_norm": 0.050115760415792465, "learning_rate": 3.065151518863793e-05, "loss": 0.005, "step": 58520 }, { "grad_norm": 0.07071726769208908, "learning_rate": 3.0634707420561805e-05, "loss": 0.005, "step": 58530 }, { "grad_norm": 0.08187642693519592, "learning_rate": 3.0617902226563643e-05, "loss": 0.0093, "step": 58540 }, { "grad_norm": 0.04291773587465286, "learning_rate": 3.0601099608877245e-05, "loss": 0.0046, "step": 58550 }, { "grad_norm": 0.09288325905799866, "learning_rate": 3.058429956973604e-05, "loss": 0.0083, "step": 58560 }, { "grad_norm": 0.0512162521481514, "learning_rate": 3.0567502111373124e-05, "loss": 0.0047, "step": 58570 }, { "grad_norm": 0.09033168107271194, "learning_rate": 3.0550707236021256e-05, "loss": 0.0084, "step": 58580 }, { "grad_norm": 0.03686519339680672, "learning_rate": 3.0533914945912843e-05, "loss": 0.0052, "step": 58590 }, { "grad_norm": 0.09942790865898132, "learning_rate": 3.0517125243279976e-05, "loss": 0.0076, "step": 58600 }, { "grad_norm": 0.13189652562141418, "learning_rate": 3.050033813035436e-05, "loss": 0.0066, "step": 58610 }, { "grad_norm": 0.05882579833269119, "learning_rate": 3.0483553609367376e-05, "loss": 0.0037, "step": 58620 }, { "grad_norm": 0.09736449271440506, "learning_rate": 3.0466771682550067e-05, "loss": 0.0076, "step": 58630 }, { "grad_norm": 0.0822896957397461, "learning_rate": 3.044999235213312e-05, "loss": 0.0065, "step": 58640 }, { "grad_norm": 0.0744331106543541, "learning_rate": 3.043321562034688e-05, "loss": 0.0059, "step": 58650 }, { "grad_norm": 0.043427929282188416, "learning_rate": 3.0416441489421337e-05, "loss": 0.0048, "step": 58660 }, { "grad_norm": 0.06089520454406738, "learning_rate": 3.039966996158617e-05, "loss": 0.0078, "step": 58670 }, { "grad_norm": 0.04473692551255226, "learning_rate": 3.0382901039070678e-05, "loss": 0.0072, "step": 58680 }, { "grad_norm": 0.047004301100969315, "learning_rate": 3.036613472410382e-05, "loss": 0.0083, "step": 58690 }, { "grad_norm": 0.07601402699947357, "learning_rate": 3.034937101891422e-05, "loss": 0.0072, "step": 58700 }, { "grad_norm": 0.06486090272665024, "learning_rate": 3.033260992573013e-05, "loss": 0.0061, "step": 58710 }, { "grad_norm": 0.03941848501563072, "learning_rate": 3.0315851446779496e-05, "loss": 0.0044, "step": 58720 }, { "grad_norm": 0.054786354303359985, "learning_rate": 3.0299095584289872e-05, "loss": 0.0049, "step": 58730 }, { "grad_norm": 0.08157308399677277, "learning_rate": 3.028234234048848e-05, "loss": 0.0054, "step": 58740 }, { "grad_norm": 0.07057943195104599, "learning_rate": 3.0265591717602226e-05, "loss": 0.0051, "step": 58750 }, { "grad_norm": 0.04133124276995659, "learning_rate": 3.0248843717857623e-05, "loss": 0.0046, "step": 58760 }, { "grad_norm": 0.030627215281128883, "learning_rate": 3.0232098343480865e-05, "loss": 0.0046, "step": 58770 }, { "grad_norm": 0.07642646133899689, "learning_rate": 3.021535559669776e-05, "loss": 0.008, "step": 58780 }, { "grad_norm": 0.05338303744792938, "learning_rate": 3.0198615479733807e-05, "loss": 0.0047, "step": 58790 }, { "grad_norm": 0.07888446003198624, "learning_rate": 3.0181877994814135e-05, "loss": 0.0055, "step": 58800 }, { "grad_norm": 0.05515763536095619, "learning_rate": 3.0165143144163528e-05, "loss": 0.004, "step": 58810 }, { "grad_norm": 0.04488963261246681, "learning_rate": 3.0148410930006443e-05, "loss": 0.007, "step": 58820 }, { "grad_norm": 0.057641398161649704, "learning_rate": 3.013168135456693e-05, "loss": 0.0064, "step": 58830 }, { "grad_norm": 0.05126282945275307, "learning_rate": 3.011495442006873e-05, "loss": 0.0064, "step": 58840 }, { "grad_norm": 0.04367637634277344, "learning_rate": 3.009823012873524e-05, "loss": 0.0056, "step": 58850 }, { "grad_norm": 0.1060466319322586, "learning_rate": 3.0081508482789483e-05, "loss": 0.011, "step": 58860 }, { "grad_norm": 0.08152269572019577, "learning_rate": 3.0064789484454125e-05, "loss": 0.0051, "step": 58870 }, { "grad_norm": 0.04699864983558655, "learning_rate": 3.0048073135951506e-05, "loss": 0.0031, "step": 58880 }, { "grad_norm": 0.07256992906332016, "learning_rate": 3.003135943950358e-05, "loss": 0.0054, "step": 58890 }, { "grad_norm": 0.04864875227212906, "learning_rate": 3.0014648397332002e-05, "loss": 0.0044, "step": 58900 }, { "grad_norm": 0.05052800104022026, "learning_rate": 2.9997940011658032e-05, "loss": 0.0061, "step": 58910 }, { "grad_norm": 0.08554944396018982, "learning_rate": 2.9981234284702563e-05, "loss": 0.0105, "step": 58920 }, { "grad_norm": 0.07247111201286316, "learning_rate": 2.9964531218686176e-05, "loss": 0.0051, "step": 58930 }, { "grad_norm": 0.06051166355609894, "learning_rate": 2.9947830815829087e-05, "loss": 0.0079, "step": 58940 }, { "grad_norm": 0.07121331244707108, "learning_rate": 2.9931133078351126e-05, "loss": 0.0062, "step": 58950 }, { "grad_norm": 0.04069345444440842, "learning_rate": 2.99144380084718e-05, "loss": 0.0056, "step": 58960 }, { "grad_norm": 0.049108460545539856, "learning_rate": 2.9897745608410278e-05, "loss": 0.0069, "step": 58970 }, { "grad_norm": 0.06336810439825058, "learning_rate": 2.9881055880385335e-05, "loss": 0.0068, "step": 58980 }, { "grad_norm": 0.054894331842660904, "learning_rate": 2.986436882661542e-05, "loss": 0.0076, "step": 58990 }, { "grad_norm": 0.04974106326699257, "learning_rate": 2.984768444931858e-05, "loss": 0.0055, "step": 59000 }, { "grad_norm": 0.047506190836429596, "learning_rate": 2.9831002750712574e-05, "loss": 0.0041, "step": 59010 }, { "grad_norm": 0.06110323220491409, "learning_rate": 2.981432373301476e-05, "loss": 0.0072, "step": 59020 }, { "grad_norm": 0.05182589590549469, "learning_rate": 2.979764739844214e-05, "loss": 0.008, "step": 59030 }, { "grad_norm": 0.08075997233390808, "learning_rate": 2.9780973749211373e-05, "loss": 0.0057, "step": 59040 }, { "grad_norm": 0.04952871426939964, "learning_rate": 2.9764302787538767e-05, "loss": 0.0057, "step": 59050 }, { "grad_norm": 0.03436402603983879, "learning_rate": 2.9747634515640254e-05, "loss": 0.0051, "step": 59060 }, { "grad_norm": 0.06866759061813354, "learning_rate": 2.9730968935731428e-05, "loss": 0.0061, "step": 59070 }, { "grad_norm": 0.0511699914932251, "learning_rate": 2.971430605002753e-05, "loss": 0.0058, "step": 59080 }, { "grad_norm": 0.04429391399025917, "learning_rate": 2.9697645860743383e-05, "loss": 0.0043, "step": 59090 }, { "grad_norm": 0.06473986804485321, "learning_rate": 2.9680988370093533e-05, "loss": 0.0053, "step": 59100 }, { "grad_norm": 0.08032837510108948, "learning_rate": 2.9664333580292105e-05, "loss": 0.0055, "step": 59110 }, { "grad_norm": 0.07522331178188324, "learning_rate": 2.9647681493552914e-05, "loss": 0.005, "step": 59120 }, { "grad_norm": 0.06702088564634323, "learning_rate": 2.9631032112089397e-05, "loss": 0.0063, "step": 59130 }, { "grad_norm": 0.05420155078172684, "learning_rate": 2.96143854381146e-05, "loss": 0.0049, "step": 59140 }, { "grad_norm": 0.042179543524980545, "learning_rate": 2.9597741473841257e-05, "loss": 0.0089, "step": 59150 }, { "grad_norm": 0.047934696078300476, "learning_rate": 2.9581100221481718e-05, "loss": 0.0073, "step": 59160 }, { "grad_norm": 0.04493454471230507, "learning_rate": 2.9564461683247964e-05, "loss": 0.0079, "step": 59170 }, { "grad_norm": 0.037359848618507385, "learning_rate": 2.9547825861351626e-05, "loss": 0.0031, "step": 59180 }, { "grad_norm": 0.06794185191392899, "learning_rate": 2.9531192758003988e-05, "loss": 0.0069, "step": 59190 }, { "grad_norm": 0.03274911269545555, "learning_rate": 2.9514562375415956e-05, "loss": 0.0047, "step": 59200 }, { "grad_norm": 0.0644359439611435, "learning_rate": 2.949793471579809e-05, "loss": 0.0052, "step": 59210 }, { "grad_norm": 0.05561237037181854, "learning_rate": 2.9481309781360545e-05, "loss": 0.0053, "step": 59220 }, { "grad_norm": 0.07565868645906448, "learning_rate": 2.9464687574313167e-05, "loss": 0.005, "step": 59230 }, { "grad_norm": 0.05033515393733978, "learning_rate": 2.9448068096865417e-05, "loss": 0.0066, "step": 59240 }, { "grad_norm": 0.031245291233062744, "learning_rate": 2.943145135122638e-05, "loss": 0.0054, "step": 59250 }, { "grad_norm": 0.07357128709554672, "learning_rate": 2.9414837339604785e-05, "loss": 0.0106, "step": 59260 }, { "grad_norm": 0.03623727709054947, "learning_rate": 2.9398226064209028e-05, "loss": 0.0063, "step": 59270 }, { "grad_norm": 0.06386580318212509, "learning_rate": 2.9381617527247106e-05, "loss": 0.0072, "step": 59280 }, { "grad_norm": 0.05570901185274124, "learning_rate": 2.9365011730926673e-05, "loss": 0.0035, "step": 59290 }, { "grad_norm": 0.08534705638885498, "learning_rate": 2.934840867745498e-05, "loss": 0.0064, "step": 59300 }, { "grad_norm": 0.08638383448123932, "learning_rate": 2.9331808369038965e-05, "loss": 0.0051, "step": 59310 }, { "grad_norm": 0.061995431780815125, "learning_rate": 2.9315210807885172e-05, "loss": 0.0055, "step": 59320 }, { "grad_norm": 0.04819771647453308, "learning_rate": 2.929861599619978e-05, "loss": 0.0046, "step": 59330 }, { "grad_norm": 0.042263638228178024, "learning_rate": 2.9282023936188628e-05, "loss": 0.0039, "step": 59340 }, { "grad_norm": 0.07057434320449829, "learning_rate": 2.9265434630057153e-05, "loss": 0.005, "step": 59350 }, { "grad_norm": 0.08253936469554901, "learning_rate": 2.924884808001045e-05, "loss": 0.0067, "step": 59360 }, { "grad_norm": 0.059069886803627014, "learning_rate": 2.923226428825323e-05, "loss": 0.0075, "step": 59370 }, { "grad_norm": 0.042268551886081696, "learning_rate": 2.9215683256989866e-05, "loss": 0.0039, "step": 59380 }, { "grad_norm": 0.09539886564016342, "learning_rate": 2.9199104988424326e-05, "loss": 0.0084, "step": 59390 }, { "grad_norm": 0.060251880437135696, "learning_rate": 2.9182529484760245e-05, "loss": 0.0065, "step": 59400 }, { "grad_norm": 0.06520213186740875, "learning_rate": 2.916595674820084e-05, "loss": 0.0076, "step": 59410 }, { "grad_norm": 0.08489805459976196, "learning_rate": 2.914938678094905e-05, "loss": 0.0052, "step": 59420 }, { "grad_norm": 0.06950151175260544, "learning_rate": 2.9132819585207367e-05, "loss": 0.0041, "step": 59430 }, { "grad_norm": 0.057731326669454575, "learning_rate": 2.9116255163177958e-05, "loss": 0.0056, "step": 59440 }, { "grad_norm": 0.05496474355459213, "learning_rate": 2.9099693517062565e-05, "loss": 0.0052, "step": 59450 }, { "grad_norm": 0.05007191747426987, "learning_rate": 2.9083134649062627e-05, "loss": 0.0053, "step": 59460 }, { "grad_norm": 0.05518543720245361, "learning_rate": 2.906657856137917e-05, "loss": 0.0046, "step": 59470 }, { "grad_norm": 0.08900569379329681, "learning_rate": 2.9050025256212853e-05, "loss": 0.005, "step": 59480 }, { "grad_norm": 0.06534279137849808, "learning_rate": 2.9033474735764048e-05, "loss": 0.0058, "step": 59490 }, { "grad_norm": 0.08571375161409378, "learning_rate": 2.901692700223262e-05, "loss": 0.0061, "step": 59500 }, { "grad_norm": 0.05536825954914093, "learning_rate": 2.9000382057818142e-05, "loss": 0.0062, "step": 59510 }, { "grad_norm": 0.05126935988664627, "learning_rate": 2.8983839904719813e-05, "loss": 0.0056, "step": 59520 }, { "grad_norm": 0.05013841763138771, "learning_rate": 2.8967300545136456e-05, "loss": 0.0053, "step": 59530 }, { "grad_norm": 0.07145841419696808, "learning_rate": 2.8950763981266517e-05, "loss": 0.0054, "step": 59540 }, { "grad_norm": 0.07858828455209732, "learning_rate": 2.893423021530807e-05, "loss": 0.0063, "step": 59550 }, { "grad_norm": 0.11250687390565872, "learning_rate": 2.891769924945883e-05, "loss": 0.0082, "step": 59560 }, { "grad_norm": 0.1008998230099678, "learning_rate": 2.890117108591611e-05, "loss": 0.0068, "step": 59570 }, { "grad_norm": 0.0451677069067955, "learning_rate": 2.888464572687689e-05, "loss": 0.0045, "step": 59580 }, { "grad_norm": 0.05384256690740585, "learning_rate": 2.8868123174537754e-05, "loss": 0.0077, "step": 59590 }, { "grad_norm": 0.04732833057641983, "learning_rate": 2.885160343109491e-05, "loss": 0.0065, "step": 59600 }, { "grad_norm": 0.04619608446955681, "learning_rate": 2.8835086498744217e-05, "loss": 0.0134, "step": 59610 }, { "grad_norm": 0.05573653057217598, "learning_rate": 2.8818572379681115e-05, "loss": 0.0042, "step": 59620 }, { "grad_norm": 0.055143702775239944, "learning_rate": 2.8802061076100682e-05, "loss": 0.0053, "step": 59630 }, { "grad_norm": 0.07648712396621704, "learning_rate": 2.8785552590197694e-05, "loss": 0.0068, "step": 59640 }, { "grad_norm": 0.05842749774456024, "learning_rate": 2.8769046924166465e-05, "loss": 0.006, "step": 59650 }, { "grad_norm": 0.022936247289180756, "learning_rate": 2.875254408020099e-05, "loss": 0.005, "step": 59660 }, { "grad_norm": 0.05509881302714348, "learning_rate": 2.8736044060494816e-05, "loss": 0.0054, "step": 59670 }, { "grad_norm": 0.06369739770889282, "learning_rate": 2.8719546867241193e-05, "loss": 0.0074, "step": 59680 }, { "grad_norm": 0.06654095649719238, "learning_rate": 2.8703052502632954e-05, "loss": 0.0074, "step": 59690 }, { "grad_norm": 0.06713452935218811, "learning_rate": 2.8686560968862565e-05, "loss": 0.0068, "step": 59700 }, { "grad_norm": 0.0640784204006195, "learning_rate": 2.867007226812215e-05, "loss": 0.0073, "step": 59710 }, { "grad_norm": 0.05701136961579323, "learning_rate": 2.8653586402603394e-05, "loss": 0.0067, "step": 59720 }, { "grad_norm": 0.07173151522874832, "learning_rate": 2.8637103374497632e-05, "loss": 0.0077, "step": 59730 }, { "grad_norm": 0.05851895362138748, "learning_rate": 2.8620623185995833e-05, "loss": 0.005, "step": 59740 }, { "grad_norm": 0.08333992213010788, "learning_rate": 2.8604145839288575e-05, "loss": 0.0083, "step": 59750 }, { "grad_norm": 0.07220294326543808, "learning_rate": 2.858767133656608e-05, "loss": 0.0046, "step": 59760 }, { "grad_norm": 0.08462963998317719, "learning_rate": 2.857119968001816e-05, "loss": 0.0072, "step": 59770 }, { "grad_norm": 0.07981514185667038, "learning_rate": 2.8554730871834273e-05, "loss": 0.008, "step": 59780 }, { "grad_norm": 0.06207243725657463, "learning_rate": 2.8538264914203482e-05, "loss": 0.0059, "step": 59790 }, { "grad_norm": 0.03685876727104187, "learning_rate": 2.8521801809314487e-05, "loss": 0.0044, "step": 59800 }, { "grad_norm": 0.06480024755001068, "learning_rate": 2.8505341559355596e-05, "loss": 0.0054, "step": 59810 }, { "grad_norm": 0.06243707612156868, "learning_rate": 2.8488884166514745e-05, "loss": 0.0052, "step": 59820 }, { "grad_norm": 0.06143984571099281, "learning_rate": 2.8472429632979503e-05, "loss": 0.0083, "step": 59830 }, { "grad_norm": 0.025678008794784546, "learning_rate": 2.845597796093701e-05, "loss": 0.0067, "step": 59840 }, { "grad_norm": 0.05397840216755867, "learning_rate": 2.8439529152574053e-05, "loss": 0.0077, "step": 59850 }, { "grad_norm": 0.053277723491191864, "learning_rate": 2.8423083210077082e-05, "loss": 0.008, "step": 59860 }, { "grad_norm": 0.057214174419641495, "learning_rate": 2.8406640135632112e-05, "loss": 0.005, "step": 59870 }, { "grad_norm": 0.04329302906990051, "learning_rate": 2.839019993142482e-05, "loss": 0.0072, "step": 59880 }, { "grad_norm": 0.08812329173088074, "learning_rate": 2.8373762599640426e-05, "loss": 0.0068, "step": 59890 }, { "grad_norm": 0.0670248344540596, "learning_rate": 2.835732814246384e-05, "loss": 0.0086, "step": 59900 }, { "grad_norm": 0.06515990197658539, "learning_rate": 2.8340896562079565e-05, "loss": 0.0052, "step": 59910 }, { "grad_norm": 0.03475719317793846, "learning_rate": 2.832446786067173e-05, "loss": 0.0035, "step": 59920 }, { "grad_norm": 0.06631974875926971, "learning_rate": 2.8308042040424067e-05, "loss": 0.0071, "step": 59930 }, { "grad_norm": 0.064963698387146, "learning_rate": 2.829161910351993e-05, "loss": 0.0047, "step": 59940 }, { "grad_norm": 0.047271665185689926, "learning_rate": 2.82751990521423e-05, "loss": 0.0047, "step": 59950 }, { "grad_norm": 0.07301522046327591, "learning_rate": 2.8258781888473762e-05, "loss": 0.0057, "step": 59960 }, { "grad_norm": 0.058446019887924194, "learning_rate": 2.824236761469653e-05, "loss": 0.0048, "step": 59970 }, { "grad_norm": 0.052307311445474625, "learning_rate": 2.8225956232992425e-05, "loss": 0.0042, "step": 59980 }, { "grad_norm": 0.055689387023448944, "learning_rate": 2.820954774554287e-05, "loss": 0.005, "step": 59990 }, { "grad_norm": 0.05502066761255264, "learning_rate": 2.8193142154528928e-05, "loss": 0.0044, "step": 60000 }, { "grad_norm": 0.06644283980131149, "learning_rate": 2.8176739462131273e-05, "loss": 0.0056, "step": 60010 }, { "grad_norm": 0.0495360791683197, "learning_rate": 2.816033967053018e-05, "loss": 0.0075, "step": 60020 }, { "grad_norm": 0.05236392468214035, "learning_rate": 2.8143942781905548e-05, "loss": 0.0068, "step": 60030 }, { "grad_norm": 0.03861304372549057, "learning_rate": 2.8127548798436888e-05, "loss": 0.0054, "step": 60040 }, { "grad_norm": 0.07251908630132675, "learning_rate": 2.811115772230334e-05, "loss": 0.0073, "step": 60050 }, { "grad_norm": 0.05702132731676102, "learning_rate": 2.8094769555683608e-05, "loss": 0.0039, "step": 60060 }, { "grad_norm": 0.07367632538080215, "learning_rate": 2.8078384300756068e-05, "loss": 0.0088, "step": 60070 }, { "grad_norm": 0.056473951786756516, "learning_rate": 2.806200195969865e-05, "loss": 0.0038, "step": 60080 }, { "grad_norm": 0.06943564116954803, "learning_rate": 2.8045622534688987e-05, "loss": 0.0062, "step": 60090 }, { "grad_norm": 0.04878902807831764, "learning_rate": 2.8029246027904255e-05, "loss": 0.0054, "step": 60100 }, { "grad_norm": 0.0531153678894043, "learning_rate": 2.801287244152123e-05, "loss": 0.0056, "step": 60110 }, { "grad_norm": 0.03681117668747902, "learning_rate": 2.799650177771634e-05, "loss": 0.006, "step": 60120 }, { "grad_norm": 0.039919402450323105, "learning_rate": 2.7980134038665607e-05, "loss": 0.0065, "step": 60130 }, { "grad_norm": 0.07966423034667969, "learning_rate": 2.7963769226544667e-05, "loss": 0.0114, "step": 60140 }, { "grad_norm": 0.06191862002015114, "learning_rate": 2.7947407343528776e-05, "loss": 0.0061, "step": 60150 }, { "grad_norm": 0.08643033355474472, "learning_rate": 2.793104839179278e-05, "loss": 0.0059, "step": 60160 }, { "grad_norm": 0.04339706525206566, "learning_rate": 2.7914692373511152e-05, "loss": 0.0044, "step": 60170 }, { "grad_norm": 0.06222870945930481, "learning_rate": 2.789833929085798e-05, "loss": 0.0092, "step": 60180 }, { "grad_norm": 0.06787431985139847, "learning_rate": 2.7881989146006937e-05, "loss": 0.0121, "step": 60190 }, { "grad_norm": 0.0737740695476532, "learning_rate": 2.7865641941131337e-05, "loss": 0.0068, "step": 60200 }, { "grad_norm": 0.052482616156339645, "learning_rate": 2.784929767840409e-05, "loss": 0.006, "step": 60210 }, { "grad_norm": 0.177388995885849, "learning_rate": 2.7832956359997665e-05, "loss": 0.0062, "step": 60220 }, { "grad_norm": 0.045159805566072464, "learning_rate": 2.781661798808424e-05, "loss": 0.0037, "step": 60230 }, { "grad_norm": 0.07947120815515518, "learning_rate": 2.780028256483552e-05, "loss": 0.0066, "step": 60240 }, { "grad_norm": 0.08881215006113052, "learning_rate": 2.7783950092422872e-05, "loss": 0.0084, "step": 60250 }, { "grad_norm": 0.0675395280122757, "learning_rate": 2.7767620573017216e-05, "loss": 0.0066, "step": 60260 }, { "grad_norm": 0.06631403416395187, "learning_rate": 2.7751294008789143e-05, "loss": 0.0055, "step": 60270 }, { "grad_norm": 0.07893184572458267, "learning_rate": 2.7734970401908778e-05, "loss": 0.0068, "step": 60280 }, { "grad_norm": 0.0537802092730999, "learning_rate": 2.77186497545459e-05, "loss": 0.0046, "step": 60290 }, { "grad_norm": 0.07472416758537292, "learning_rate": 2.770233206886988e-05, "loss": 0.0077, "step": 60300 }, { "grad_norm": 0.052298568189144135, "learning_rate": 2.768601734704973e-05, "loss": 0.0059, "step": 60310 }, { "grad_norm": 0.056512244045734406, "learning_rate": 2.7669705591254046e-05, "loss": 0.004, "step": 60320 }, { "grad_norm": 0.026640234515070915, "learning_rate": 2.7653396803650984e-05, "loss": 0.007, "step": 60330 }, { "grad_norm": 0.046761032193899155, "learning_rate": 2.7637090986408352e-05, "loss": 0.0052, "step": 60340 }, { "grad_norm": 0.04777713492512703, "learning_rate": 2.762078814169357e-05, "loss": 0.0064, "step": 60350 }, { "grad_norm": 0.053730469197034836, "learning_rate": 2.7604488271673645e-05, "loss": 0.0067, "step": 60360 }, { "grad_norm": 0.04356369748711586, "learning_rate": 2.758819137851519e-05, "loss": 0.0033, "step": 60370 }, { "grad_norm": 0.020841628313064575, "learning_rate": 2.7571897464384422e-05, "loss": 0.0042, "step": 60380 }, { "grad_norm": 0.06793289631605148, "learning_rate": 2.755560653144717e-05, "loss": 0.0092, "step": 60390 }, { "grad_norm": 0.08877081423997879, "learning_rate": 2.7539318581868854e-05, "loss": 0.0049, "step": 60400 }, { "grad_norm": 0.06573320180177689, "learning_rate": 2.7523033617814513e-05, "loss": 0.0055, "step": 60410 }, { "grad_norm": 0.07448014616966248, "learning_rate": 2.750675164144878e-05, "loss": 0.0082, "step": 60420 }, { "grad_norm": 0.034339118748903275, "learning_rate": 2.7490472654935908e-05, "loss": 0.0056, "step": 60430 }, { "grad_norm": 0.07061786949634552, "learning_rate": 2.7474196660439688e-05, "loss": 0.005, "step": 60440 }, { "grad_norm": 0.06389445066452026, "learning_rate": 2.745792366012358e-05, "loss": 0.0039, "step": 60450 }, { "grad_norm": 0.0974789708852768, "learning_rate": 2.7441653656150657e-05, "loss": 0.0081, "step": 60460 }, { "grad_norm": 0.047074876725673676, "learning_rate": 2.7425386650683543e-05, "loss": 0.0082, "step": 60470 }, { "grad_norm": 0.07838036119937897, "learning_rate": 2.740912264588451e-05, "loss": 0.006, "step": 60480 }, { "grad_norm": 0.0866861566901207, "learning_rate": 2.7392861643915358e-05, "loss": 0.0075, "step": 60490 }, { "grad_norm": 0.038326140493154526, "learning_rate": 2.737660364693757e-05, "loss": 0.0067, "step": 60500 }, { "grad_norm": 0.08217483013868332, "learning_rate": 2.736034865711218e-05, "loss": 0.0052, "step": 60510 }, { "grad_norm": 0.06264480203390121, "learning_rate": 2.734409667659982e-05, "loss": 0.0053, "step": 60520 }, { "grad_norm": 0.06063520163297653, "learning_rate": 2.7327847707560805e-05, "loss": 0.0074, "step": 60530 }, { "grad_norm": 0.06250149756669998, "learning_rate": 2.7311601752154924e-05, "loss": 0.0078, "step": 60540 }, { "grad_norm": 0.06633003801107407, "learning_rate": 2.729535881254165e-05, "loss": 0.0056, "step": 60550 }, { "grad_norm": 0.045082103461027145, "learning_rate": 2.7279118890880018e-05, "loss": 0.005, "step": 60560 }, { "grad_norm": 0.0803227424621582, "learning_rate": 2.7262881989328683e-05, "loss": 0.0055, "step": 60570 }, { "grad_norm": 0.0406312420964241, "learning_rate": 2.7246648110045893e-05, "loss": 0.0045, "step": 60580 }, { "grad_norm": 0.08019378036260605, "learning_rate": 2.7230417255189484e-05, "loss": 0.008, "step": 60590 }, { "grad_norm": 0.06527983397245407, "learning_rate": 2.7214189426916902e-05, "loss": 0.0063, "step": 60600 }, { "grad_norm": 0.041619084775447845, "learning_rate": 2.7197964627385186e-05, "loss": 0.0043, "step": 60610 }, { "grad_norm": 0.06504760682582855, "learning_rate": 2.718174285875097e-05, "loss": 0.0064, "step": 60620 }, { "grad_norm": 0.08106529712677002, "learning_rate": 2.716552412317048e-05, "loss": 0.0057, "step": 60630 }, { "grad_norm": 0.08275970071554184, "learning_rate": 2.7149308422799568e-05, "loss": 0.0054, "step": 60640 }, { "grad_norm": 0.04020128771662712, "learning_rate": 2.713309575979367e-05, "loss": 0.0069, "step": 60650 }, { "grad_norm": 0.06778594106435776, "learning_rate": 2.7116886136307767e-05, "loss": 0.011, "step": 60660 }, { "grad_norm": 0.07392517477273941, "learning_rate": 2.7100679554496477e-05, "loss": 0.0066, "step": 60670 }, { "grad_norm": 0.046699102967977524, "learning_rate": 2.7084476016514058e-05, "loss": 0.0072, "step": 60680 }, { "grad_norm": 0.07202314585447311, "learning_rate": 2.706827552451431e-05, "loss": 0.0081, "step": 60690 }, { "grad_norm": 0.03634055331349373, "learning_rate": 2.705207808065064e-05, "loss": 0.0048, "step": 60700 }, { "grad_norm": 0.03658583387732506, "learning_rate": 2.703588368707602e-05, "loss": 0.0056, "step": 60710 }, { "grad_norm": 0.07435853779315948, "learning_rate": 2.7019692345943058e-05, "loss": 0.007, "step": 60720 }, { "grad_norm": 0.04427274316549301, "learning_rate": 2.7003504059403956e-05, "loss": 0.0039, "step": 60730 }, { "grad_norm": 0.07028200477361679, "learning_rate": 2.6987318829610463e-05, "loss": 0.0052, "step": 60740 }, { "grad_norm": 0.036376625299453735, "learning_rate": 2.6971136658714025e-05, "loss": 0.0039, "step": 60750 }, { "grad_norm": 0.07632006704807281, "learning_rate": 2.6954957548865543e-05, "loss": 0.0062, "step": 60760 }, { "grad_norm": 0.06332021951675415, "learning_rate": 2.6938781502215614e-05, "loss": 0.005, "step": 60770 }, { "grad_norm": 0.038659531623125076, "learning_rate": 2.6922608520914387e-05, "loss": 0.0063, "step": 60780 }, { "grad_norm": 0.07241734117269516, "learning_rate": 2.690643860711161e-05, "loss": 0.0074, "step": 60790 }, { "grad_norm": 0.029189232736825943, "learning_rate": 2.6890271762956626e-05, "loss": 0.0108, "step": 60800 }, { "grad_norm": 0.05269453302025795, "learning_rate": 2.6874107990598364e-05, "loss": 0.0066, "step": 60810 }, { "grad_norm": 0.05584034323692322, "learning_rate": 2.6857947292185355e-05, "loss": 0.0085, "step": 60820 }, { "grad_norm": 0.0457429476082325, "learning_rate": 2.6841789669865707e-05, "loss": 0.007, "step": 60830 }, { "grad_norm": 0.10955051332712173, "learning_rate": 2.6825635125787136e-05, "loss": 0.007, "step": 60840 }, { "grad_norm": 0.036631952971220016, "learning_rate": 2.6809483662096935e-05, "loss": 0.0045, "step": 60850 }, { "grad_norm": 0.05939513072371483, "learning_rate": 2.6793335280941996e-05, "loss": 0.006, "step": 60860 }, { "grad_norm": 0.07318940758705139, "learning_rate": 2.6777189984468808e-05, "loss": 0.0051, "step": 60870 }, { "grad_norm": 0.0638289526104927, "learning_rate": 2.676104777482342e-05, "loss": 0.0039, "step": 60880 }, { "grad_norm": 0.044555921107530594, "learning_rate": 2.6744908654151467e-05, "loss": 0.0047, "step": 60890 }, { "grad_norm": 0.07940872013568878, "learning_rate": 2.6728772624598257e-05, "loss": 0.0071, "step": 60900 }, { "grad_norm": 0.07779117673635483, "learning_rate": 2.6712639688308594e-05, "loss": 0.006, "step": 60910 }, { "grad_norm": 0.04574961960315704, "learning_rate": 2.669650984742693e-05, "loss": 0.0054, "step": 60920 }, { "grad_norm": 0.04185720905661583, "learning_rate": 2.668038310409724e-05, "loss": 0.0042, "step": 60930 }, { "grad_norm": 0.06525000184774399, "learning_rate": 2.6664259460463135e-05, "loss": 0.0057, "step": 60940 }, { "grad_norm": 0.05266006663441658, "learning_rate": 2.6648138918667825e-05, "loss": 0.0061, "step": 60950 }, { "grad_norm": 0.06155917048454285, "learning_rate": 2.6632021480854074e-05, "loss": 0.0059, "step": 60960 }, { "grad_norm": 0.07814351469278336, "learning_rate": 2.6615907149164253e-05, "loss": 0.0053, "step": 60970 }, { "grad_norm": 0.053204819560050964, "learning_rate": 2.659979592574032e-05, "loss": 0.0036, "step": 60980 }, { "grad_norm": 0.05421912670135498, "learning_rate": 2.6583687812723805e-05, "loss": 0.0058, "step": 60990 }, { "grad_norm": 0.05006387457251549, "learning_rate": 2.656758281225584e-05, "loss": 0.0036, "step": 61000 }, { "grad_norm": 0.021345416083931923, "learning_rate": 2.6551480926477133e-05, "loss": 0.0036, "step": 61010 }, { "grad_norm": 0.03829103335738182, "learning_rate": 2.6535382157527993e-05, "loss": 0.0045, "step": 61020 }, { "grad_norm": 0.07089244574308395, "learning_rate": 2.6519286507548292e-05, "loss": 0.0087, "step": 61030 }, { "grad_norm": 0.0573926717042923, "learning_rate": 2.650319397867751e-05, "loss": 0.0062, "step": 61040 }, { "grad_norm": 0.044009558856487274, "learning_rate": 2.6487104573054693e-05, "loss": 0.0042, "step": 61050 }, { "grad_norm": 0.05552152171730995, "learning_rate": 2.6471018292818484e-05, "loss": 0.0053, "step": 61060 }, { "grad_norm": 0.047331273555755615, "learning_rate": 2.6454935140107112e-05, "loss": 0.006, "step": 61070 }, { "grad_norm": 0.09590822458267212, "learning_rate": 2.643885511705838e-05, "loss": 0.0084, "step": 61080 }, { "grad_norm": 0.05881207808852196, "learning_rate": 2.6422778225809704e-05, "loss": 0.0043, "step": 61090 }, { "grad_norm": 0.0602593719959259, "learning_rate": 2.6406704468498016e-05, "loss": 0.0061, "step": 61100 }, { "grad_norm": 0.07439202815294266, "learning_rate": 2.63906338472599e-05, "loss": 0.0067, "step": 61110 }, { "grad_norm": 0.07181286811828613, "learning_rate": 2.637456636423148e-05, "loss": 0.0053, "step": 61120 }, { "grad_norm": 0.05770034343004227, "learning_rate": 2.6358502021548515e-05, "loss": 0.0066, "step": 61130 }, { "grad_norm": 0.08811555802822113, "learning_rate": 2.6342440821346314e-05, "loss": 0.0068, "step": 61140 }, { "grad_norm": 0.042820461094379425, "learning_rate": 2.632638276575974e-05, "loss": 0.0028, "step": 61150 }, { "grad_norm": 0.03835967928171158, "learning_rate": 2.631032785692327e-05, "loss": 0.0047, "step": 61160 }, { "grad_norm": 0.06266910582780838, "learning_rate": 2.6294276096970972e-05, "loss": 0.0049, "step": 61170 }, { "grad_norm": 0.05202789604663849, "learning_rate": 2.6278227488036466e-05, "loss": 0.0066, "step": 61180 }, { "grad_norm": 0.045291148126125336, "learning_rate": 2.626218203225299e-05, "loss": 0.0042, "step": 61190 }, { "grad_norm": 0.07138784229755402, "learning_rate": 2.6246139731753318e-05, "loss": 0.0074, "step": 61200 }, { "grad_norm": 0.06919398158788681, "learning_rate": 2.6230100588669848e-05, "loss": 0.0072, "step": 61210 }, { "grad_norm": 0.06542912870645523, "learning_rate": 2.6214064605134525e-05, "loss": 0.0044, "step": 61220 }, { "grad_norm": 0.0527866892516613, "learning_rate": 2.61980317832789e-05, "loss": 0.0057, "step": 61230 }, { "grad_norm": 0.07867668569087982, "learning_rate": 2.618200212523408e-05, "loss": 0.0052, "step": 61240 }, { "grad_norm": 0.058345064520835876, "learning_rate": 2.616597563313079e-05, "loss": 0.0058, "step": 61250 }, { "grad_norm": 0.08296900242567062, "learning_rate": 2.6149952309099247e-05, "loss": 0.0047, "step": 61260 }, { "grad_norm": 0.03198401629924774, "learning_rate": 2.6133932155269357e-05, "loss": 0.0044, "step": 61270 }, { "grad_norm": 0.049025796353816986, "learning_rate": 2.611791517377055e-05, "loss": 0.0057, "step": 61280 }, { "grad_norm": 0.05130220949649811, "learning_rate": 2.6101901366731833e-05, "loss": 0.0041, "step": 61290 }, { "grad_norm": 0.06862293183803558, "learning_rate": 2.6085890736281814e-05, "loss": 0.0044, "step": 61300 }, { "grad_norm": 0.05143120512366295, "learning_rate": 2.6069883284548622e-05, "loss": 0.0051, "step": 61310 }, { "grad_norm": 0.08416828513145447, "learning_rate": 2.6053879013660033e-05, "loss": 0.0058, "step": 61320 }, { "grad_norm": 0.04463537037372589, "learning_rate": 2.6037877925743352e-05, "loss": 0.0072, "step": 61330 }, { "grad_norm": 0.03902745246887207, "learning_rate": 2.6021880022925476e-05, "loss": 0.0053, "step": 61340 }, { "grad_norm": 0.06547941267490387, "learning_rate": 2.600588530733291e-05, "loss": 0.0067, "step": 61350 }, { "grad_norm": 0.08562013506889343, "learning_rate": 2.5989893781091725e-05, "loss": 0.0064, "step": 61360 }, { "grad_norm": 0.04405587166547775, "learning_rate": 2.5973905446327483e-05, "loss": 0.0061, "step": 61370 }, { "grad_norm": 0.060678984969854355, "learning_rate": 2.595792030516544e-05, "loss": 0.0044, "step": 61380 }, { "grad_norm": 0.0549674928188324, "learning_rate": 2.5941938359730356e-05, "loss": 0.0046, "step": 61390 }, { "grad_norm": 0.06485576927661896, "learning_rate": 2.5925959612146588e-05, "loss": 0.0053, "step": 61400 }, { "grad_norm": 0.05459176003932953, "learning_rate": 2.5909984064538073e-05, "loss": 0.0068, "step": 61410 }, { "grad_norm": 0.06449038535356522, "learning_rate": 2.5894011719028315e-05, "loss": 0.0064, "step": 61420 }, { "grad_norm": 0.07129473239183426, "learning_rate": 2.587804257774039e-05, "loss": 0.0048, "step": 61430 }, { "grad_norm": 0.07963170856237411, "learning_rate": 2.586207664279696e-05, "loss": 0.0048, "step": 61440 }, { "grad_norm": 0.04190078005194664, "learning_rate": 2.584611391632025e-05, "loss": 0.0069, "step": 61450 }, { "grad_norm": 0.037276092916727066, "learning_rate": 2.583015440043206e-05, "loss": 0.0082, "step": 61460 }, { "grad_norm": 0.06230070814490318, "learning_rate": 2.5814198097253784e-05, "loss": 0.0044, "step": 61470 }, { "grad_norm": 0.032477591186761856, "learning_rate": 2.579824500890633e-05, "loss": 0.0065, "step": 61480 }, { "grad_norm": 0.07367005199193954, "learning_rate": 2.5782295137510214e-05, "loss": 0.0072, "step": 61490 }, { "grad_norm": 0.04512561112642288, "learning_rate": 2.576634848518557e-05, "loss": 0.0046, "step": 61500 }, { "grad_norm": 0.03736181929707527, "learning_rate": 2.5750405054052053e-05, "loss": 0.0051, "step": 61510 }, { "grad_norm": 0.04862625524401665, "learning_rate": 2.57344648462289e-05, "loss": 0.0062, "step": 61520 }, { "grad_norm": 0.06855522096157074, "learning_rate": 2.5718527863834894e-05, "loss": 0.0048, "step": 61530 }, { "grad_norm": 0.051773373037576675, "learning_rate": 2.570259410898844e-05, "loss": 0.0053, "step": 61540 }, { "grad_norm": 0.06292670965194702, "learning_rate": 2.5686663583807475e-05, "loss": 0.0062, "step": 61550 }, { "grad_norm": 0.05315954610705376, "learning_rate": 2.56707362904095e-05, "loss": 0.0056, "step": 61560 }, { "grad_norm": 0.06515851616859436, "learning_rate": 2.5654812230911674e-05, "loss": 0.0058, "step": 61570 }, { "grad_norm": 0.03582388907670975, "learning_rate": 2.5638891407430594e-05, "loss": 0.0045, "step": 61580 }, { "grad_norm": 0.029497921466827393, "learning_rate": 2.5622973822082512e-05, "loss": 0.0037, "step": 61590 }, { "grad_norm": 0.061585746705532074, "learning_rate": 2.560705947698323e-05, "loss": 0.0061, "step": 61600 }, { "grad_norm": 0.04749953746795654, "learning_rate": 2.559114837424812e-05, "loss": 0.0045, "step": 61610 }, { "grad_norm": 0.04174920916557312, "learning_rate": 2.5575240515992117e-05, "loss": 0.0059, "step": 61620 }, { "grad_norm": 0.06138996034860611, "learning_rate": 2.555933590432973e-05, "loss": 0.0064, "step": 61630 }, { "grad_norm": 0.04900798574090004, "learning_rate": 2.554343454137504e-05, "loss": 0.0039, "step": 61640 }, { "grad_norm": 0.07047425955533981, "learning_rate": 2.5527536429241682e-05, "loss": 0.0066, "step": 61650 }, { "grad_norm": 0.08701980859041214, "learning_rate": 2.5511641570042887e-05, "loss": 0.0052, "step": 61660 }, { "grad_norm": 0.06631315499544144, "learning_rate": 2.5495749965891414e-05, "loss": 0.0051, "step": 61670 }, { "grad_norm": 0.08119373768568039, "learning_rate": 2.5479861618899625e-05, "loss": 0.0045, "step": 61680 }, { "grad_norm": 0.04986213892698288, "learning_rate": 2.5463976531179447e-05, "loss": 0.0039, "step": 61690 }, { "grad_norm": 0.05017503723502159, "learning_rate": 2.5448094704842317e-05, "loss": 0.0058, "step": 61700 }, { "grad_norm": 0.07067850977182388, "learning_rate": 2.5432216141999297e-05, "loss": 0.0047, "step": 61710 }, { "grad_norm": 0.043165724724531174, "learning_rate": 2.5416340844761034e-05, "loss": 0.0056, "step": 61720 }, { "grad_norm": 0.036153312772512436, "learning_rate": 2.540046881523769e-05, "loss": 0.0043, "step": 61730 }, { "grad_norm": 0.0495373010635376, "learning_rate": 2.538460005553902e-05, "loss": 0.0044, "step": 61740 }, { "grad_norm": 0.06132906302809715, "learning_rate": 2.536873456777431e-05, "loss": 0.0062, "step": 61750 }, { "grad_norm": 0.04361267015337944, "learning_rate": 2.535287235405245e-05, "loss": 0.0046, "step": 61760 }, { "grad_norm": 0.050626058131456375, "learning_rate": 2.533701341648188e-05, "loss": 0.0051, "step": 61770 }, { "grad_norm": 0.05558054521679878, "learning_rate": 2.532115775717058e-05, "loss": 0.0087, "step": 61780 }, { "grad_norm": 0.07146046310663223, "learning_rate": 2.5305305378226195e-05, "loss": 0.0092, "step": 61790 }, { "grad_norm": 0.057514164596796036, "learning_rate": 2.528945628175578e-05, "loss": 0.0085, "step": 61800 }, { "grad_norm": 0.05706026032567024, "learning_rate": 2.5273610469866064e-05, "loss": 0.0034, "step": 61810 }, { "grad_norm": 0.057824790477752686, "learning_rate": 2.5257767944663314e-05, "loss": 0.0059, "step": 61820 }, { "grad_norm": 0.061551667749881744, "learning_rate": 2.524192870825333e-05, "loss": 0.0051, "step": 61830 }, { "grad_norm": 0.03779697045683861, "learning_rate": 2.5226092762741528e-05, "loss": 0.004, "step": 61840 }, { "grad_norm": 0.07117164880037308, "learning_rate": 2.521026011023283e-05, "loss": 0.0054, "step": 61850 }, { "grad_norm": 0.08533305674791336, "learning_rate": 2.5194430752831766e-05, "loss": 0.0051, "step": 61860 }, { "grad_norm": 0.04345238208770752, "learning_rate": 2.5178604692642406e-05, "loss": 0.0089, "step": 61870 }, { "grad_norm": 0.06141040846705437, "learning_rate": 2.5162781931768386e-05, "loss": 0.0054, "step": 61880 }, { "grad_norm": 0.08510149270296097, "learning_rate": 2.5146962472312898e-05, "loss": 0.0066, "step": 61890 }, { "grad_norm": 0.09472587704658508, "learning_rate": 2.5131146316378695e-05, "loss": 0.0073, "step": 61900 }, { "grad_norm": 0.09399184584617615, "learning_rate": 2.5115333466068126e-05, "loss": 0.0076, "step": 61910 }, { "grad_norm": 0.07038792967796326, "learning_rate": 2.5099523923483033e-05, "loss": 0.0073, "step": 61920 }, { "grad_norm": 0.05166786164045334, "learning_rate": 2.5083717690724846e-05, "loss": 0.0034, "step": 61930 }, { "grad_norm": 0.044976502656936646, "learning_rate": 2.5067914769894608e-05, "loss": 0.0048, "step": 61940 }, { "grad_norm": 0.06336253136396408, "learning_rate": 2.5052115163092858e-05, "loss": 0.0057, "step": 61950 }, { "grad_norm": 0.05581371858716011, "learning_rate": 2.503631887241974e-05, "loss": 0.005, "step": 61960 }, { "grad_norm": 0.04251878708600998, "learning_rate": 2.502052589997489e-05, "loss": 0.0058, "step": 61970 }, { "grad_norm": 0.04983269050717354, "learning_rate": 2.5004736247857564e-05, "loss": 0.0063, "step": 61980 }, { "grad_norm": 0.04920193925499916, "learning_rate": 2.498894991816655e-05, "loss": 0.0062, "step": 61990 }, { "grad_norm": 0.04674026370048523, "learning_rate": 2.4973166913000212e-05, "loss": 0.0052, "step": 62000 }, { "grad_norm": 0.05738428235054016, "learning_rate": 2.4957387234456464e-05, "loss": 0.0056, "step": 62010 }, { "grad_norm": 0.045797888189554214, "learning_rate": 2.4941610884632773e-05, "loss": 0.0051, "step": 62020 }, { "grad_norm": 0.052536945790052414, "learning_rate": 2.4925837865626163e-05, "loss": 0.0061, "step": 62030 }, { "grad_norm": 0.04593424126505852, "learning_rate": 2.4910068179533224e-05, "loss": 0.0057, "step": 62040 }, { "grad_norm": 0.0611237958073616, "learning_rate": 2.4894301828450094e-05, "loss": 0.0046, "step": 62050 }, { "grad_norm": 0.08869931101799011, "learning_rate": 2.4878538814472475e-05, "loss": 0.0085, "step": 62060 }, { "grad_norm": 0.04352150484919548, "learning_rate": 2.4862779139695643e-05, "loss": 0.0042, "step": 62070 }, { "grad_norm": 0.06363070011138916, "learning_rate": 2.4847022806214343e-05, "loss": 0.0076, "step": 62080 }, { "grad_norm": 0.039726097136735916, "learning_rate": 2.4831269816123003e-05, "loss": 0.007, "step": 62090 }, { "grad_norm": 0.040569454431533813, "learning_rate": 2.4815520171515533e-05, "loss": 0.0045, "step": 62100 }, { "grad_norm": 0.06149234622716904, "learning_rate": 2.4799773874485406e-05, "loss": 0.0093, "step": 62110 }, { "grad_norm": 0.06613161414861679, "learning_rate": 2.4784030927125656e-05, "loss": 0.0068, "step": 62120 }, { "grad_norm": 0.03507125377655029, "learning_rate": 2.4768291331528885e-05, "loss": 0.0038, "step": 62130 }, { "grad_norm": 0.058497581630945206, "learning_rate": 2.4752555089787204e-05, "loss": 0.0037, "step": 62140 }, { "grad_norm": 0.04999630153179169, "learning_rate": 2.4736822203992327e-05, "loss": 0.01, "step": 62150 }, { "grad_norm": 0.06161521002650261, "learning_rate": 2.4721092676235487e-05, "loss": 0.0057, "step": 62160 }, { "grad_norm": 0.04328903928399086, "learning_rate": 2.4705366508607512e-05, "loss": 0.0068, "step": 62170 }, { "grad_norm": 0.049277666956186295, "learning_rate": 2.4689643703198783e-05, "loss": 0.0048, "step": 62180 }, { "grad_norm": 0.08708149194717407, "learning_rate": 2.4673924262099153e-05, "loss": 0.0055, "step": 62190 }, { "grad_norm": 0.036873385310173035, "learning_rate": 2.4658208187398114e-05, "loss": 0.0077, "step": 62200 }, { "grad_norm": 0.05130646377801895, "learning_rate": 2.4642495481184675e-05, "loss": 0.0066, "step": 62210 }, { "grad_norm": 0.07789957523345947, "learning_rate": 2.462678614554741e-05, "loss": 0.0054, "step": 62220 }, { "grad_norm": 0.07366449385881424, "learning_rate": 2.4611080182574436e-05, "loss": 0.0088, "step": 62230 }, { "grad_norm": 0.06019037589430809, "learning_rate": 2.459537759435343e-05, "loss": 0.0064, "step": 62240 }, { "grad_norm": 0.07279374450445175, "learning_rate": 2.457967838297161e-05, "loss": 0.0057, "step": 62250 }, { "grad_norm": 0.07769717276096344, "learning_rate": 2.4563982550515747e-05, "loss": 0.0075, "step": 62260 }, { "grad_norm": 0.05977536365389824, "learning_rate": 2.4548290099072168e-05, "loss": 0.0048, "step": 62270 }, { "grad_norm": 0.030352208763360977, "learning_rate": 2.453260103072676e-05, "loss": 0.0043, "step": 62280 }, { "grad_norm": 0.05385905131697655, "learning_rate": 2.451691534756495e-05, "loss": 0.0066, "step": 62290 }, { "grad_norm": 0.06103238835930824, "learning_rate": 2.4501233051671692e-05, "loss": 0.0048, "step": 62300 }, { "grad_norm": 0.04655753821134567, "learning_rate": 2.44855541451315e-05, "loss": 0.0065, "step": 62310 }, { "grad_norm": 0.039193470031023026, "learning_rate": 2.446987863002849e-05, "loss": 0.0049, "step": 62320 }, { "grad_norm": 0.03558797389268875, "learning_rate": 2.4454206508446275e-05, "loss": 0.0066, "step": 62330 }, { "grad_norm": 0.0458650179207325, "learning_rate": 2.443853778246804e-05, "loss": 0.0051, "step": 62340 }, { "grad_norm": 0.06466018408536911, "learning_rate": 2.442287245417647e-05, "loss": 0.0077, "step": 62350 }, { "grad_norm": 0.06569761037826538, "learning_rate": 2.4407210525653858e-05, "loss": 0.0044, "step": 62360 }, { "grad_norm": 0.04680816829204559, "learning_rate": 2.439155199898202e-05, "loss": 0.006, "step": 62370 }, { "grad_norm": 0.06965028494596481, "learning_rate": 2.43758968762423e-05, "loss": 0.0099, "step": 62380 }, { "grad_norm": 0.04258418455719948, "learning_rate": 2.4360245159515678e-05, "loss": 0.0058, "step": 62390 }, { "grad_norm": 0.06427954882383347, "learning_rate": 2.4344596850882546e-05, "loss": 0.0057, "step": 62400 }, { "grad_norm": 0.07271815836429596, "learning_rate": 2.432895195242294e-05, "loss": 0.0055, "step": 62410 }, { "grad_norm": 0.03247437626123428, "learning_rate": 2.4313310466216415e-05, "loss": 0.0042, "step": 62420 }, { "grad_norm": 0.10676665604114532, "learning_rate": 2.4297672394342063e-05, "loss": 0.0091, "step": 62430 }, { "grad_norm": 0.042279064655303955, "learning_rate": 2.4282037738878543e-05, "loss": 0.0037, "step": 62440 }, { "grad_norm": 0.06198855862021446, "learning_rate": 2.426640650190405e-05, "loss": 0.0056, "step": 62450 }, { "grad_norm": 0.028441209346055984, "learning_rate": 2.4250778685496305e-05, "loss": 0.0047, "step": 62460 }, { "grad_norm": 0.05016225576400757, "learning_rate": 2.423515429173261e-05, "loss": 0.0041, "step": 62470 }, { "grad_norm": 0.06236933544278145, "learning_rate": 2.4219533322689786e-05, "loss": 0.0048, "step": 62480 }, { "grad_norm": 0.05718262493610382, "learning_rate": 2.4203915780444213e-05, "loss": 0.0033, "step": 62490 }, { "grad_norm": 0.04624754562973976, "learning_rate": 2.4188301667071805e-05, "loss": 0.0085, "step": 62500 }, { "grad_norm": 0.05772100389003754, "learning_rate": 2.4172690984648038e-05, "loss": 0.005, "step": 62510 }, { "grad_norm": 0.03522234410047531, "learning_rate": 2.4157083735247898e-05, "loss": 0.0044, "step": 62520 }, { "grad_norm": 0.0618368461728096, "learning_rate": 2.4141479920945918e-05, "loss": 0.0066, "step": 62530 }, { "grad_norm": 0.06590364128351212, "learning_rate": 2.4125879543816238e-05, "loss": 0.005, "step": 62540 }, { "grad_norm": 0.08726231008768082, "learning_rate": 2.411028260593248e-05, "loss": 0.0059, "step": 62550 }, { "grad_norm": 0.08017653971910477, "learning_rate": 2.409468910936783e-05, "loss": 0.0048, "step": 62560 }, { "grad_norm": 0.07036904990673065, "learning_rate": 2.407909905619498e-05, "loss": 0.0073, "step": 62570 }, { "grad_norm": 0.0581653006374836, "learning_rate": 2.406351244848622e-05, "loss": 0.0046, "step": 62580 }, { "grad_norm": 0.04113338887691498, "learning_rate": 2.404792928831335e-05, "loss": 0.0048, "step": 62590 }, { "grad_norm": 0.05721550062298775, "learning_rate": 2.40323495777477e-05, "loss": 0.0041, "step": 62600 }, { "grad_norm": 0.07709189504384995, "learning_rate": 2.401677331886021e-05, "loss": 0.0073, "step": 62610 }, { "grad_norm": 0.04175807163119316, "learning_rate": 2.4001200513721268e-05, "loss": 0.0048, "step": 62620 }, { "grad_norm": 0.05680704489350319, "learning_rate": 2.3985631164400856e-05, "loss": 0.006, "step": 62630 }, { "grad_norm": 0.06758512556552887, "learning_rate": 2.397006527296849e-05, "loss": 0.0053, "step": 62640 }, { "grad_norm": 0.05948789417743683, "learning_rate": 2.395450284149322e-05, "loss": 0.01, "step": 62650 }, { "grad_norm": 0.037141188979148865, "learning_rate": 2.3938943872043645e-05, "loss": 0.0051, "step": 62660 }, { "grad_norm": 0.028676392510533333, "learning_rate": 2.3923388366687887e-05, "loss": 0.0062, "step": 62670 }, { "grad_norm": 0.05766542628407478, "learning_rate": 2.390783632749363e-05, "loss": 0.0073, "step": 62680 }, { "grad_norm": 0.06691606342792511, "learning_rate": 2.3892287756528086e-05, "loss": 0.005, "step": 62690 }, { "grad_norm": 0.054184556007385254, "learning_rate": 2.3876742655857997e-05, "loss": 0.0043, "step": 62700 }, { "grad_norm": 0.046968474984169006, "learning_rate": 2.3861201027549656e-05, "loss": 0.0048, "step": 62710 }, { "grad_norm": 0.04336582496762276, "learning_rate": 2.3845662873668895e-05, "loss": 0.0054, "step": 62720 }, { "grad_norm": 0.05494697019457817, "learning_rate": 2.3830128196281092e-05, "loss": 0.007, "step": 62730 }, { "grad_norm": 0.07946690917015076, "learning_rate": 2.3814596997451118e-05, "loss": 0.0077, "step": 62740 }, { "grad_norm": 0.04942215234041214, "learning_rate": 2.3799069279243414e-05, "loss": 0.0044, "step": 62750 }, { "grad_norm": 0.028234373778104782, "learning_rate": 2.3783545043722e-05, "loss": 0.0041, "step": 62760 }, { "grad_norm": 0.03795314207673073, "learning_rate": 2.3768024292950362e-05, "loss": 0.0095, "step": 62770 }, { "grad_norm": 0.03427683934569359, "learning_rate": 2.375250702899159e-05, "loss": 0.0044, "step": 62780 }, { "grad_norm": 0.07812812924385071, "learning_rate": 2.373699325390822e-05, "loss": 0.0064, "step": 62790 }, { "grad_norm": 0.04361698403954506, "learning_rate": 2.3721482969762398e-05, "loss": 0.0051, "step": 62800 }, { "grad_norm": 0.0481078177690506, "learning_rate": 2.370597617861579e-05, "loss": 0.0058, "step": 62810 }, { "grad_norm": 0.07503014802932739, "learning_rate": 2.3690472882529597e-05, "loss": 0.01, "step": 62820 }, { "grad_norm": 0.03803412988781929, "learning_rate": 2.3674973083564546e-05, "loss": 0.0036, "step": 62830 }, { "grad_norm": 0.03212109953165054, "learning_rate": 2.3659476783780905e-05, "loss": 0.0059, "step": 62840 }, { "grad_norm": 0.052163027226924896, "learning_rate": 2.364398398523848e-05, "loss": 0.0057, "step": 62850 }, { "grad_norm": 0.04077053442597389, "learning_rate": 2.3628494689996606e-05, "loss": 0.0079, "step": 62860 }, { "grad_norm": 0.060306526720523834, "learning_rate": 2.361300890011415e-05, "loss": 0.0052, "step": 62870 }, { "grad_norm": 0.058840807527303696, "learning_rate": 2.3597526617649534e-05, "loss": 0.0049, "step": 62880 }, { "grad_norm": 0.07909586280584335, "learning_rate": 2.3582047844660676e-05, "loss": 0.0066, "step": 62890 }, { "grad_norm": 0.07801289856433868, "learning_rate": 2.3566572583205067e-05, "loss": 0.0049, "step": 62900 }, { "grad_norm": 0.05762400105595589, "learning_rate": 2.35511008353397e-05, "loss": 0.004, "step": 62910 }, { "grad_norm": 0.037142619490623474, "learning_rate": 2.3535632603121126e-05, "loss": 0.0046, "step": 62920 }, { "grad_norm": 0.06274470686912537, "learning_rate": 2.352016788860541e-05, "loss": 0.0055, "step": 62930 }, { "grad_norm": 0.08021005988121033, "learning_rate": 2.3504706693848155e-05, "loss": 0.0064, "step": 62940 }, { "grad_norm": 0.05745331570506096, "learning_rate": 2.3489249020904525e-05, "loss": 0.0053, "step": 62950 }, { "grad_norm": 0.04294300824403763, "learning_rate": 2.347379487182914e-05, "loss": 0.008, "step": 62960 }, { "grad_norm": 0.049624960869550705, "learning_rate": 2.3458344248676208e-05, "loss": 0.0054, "step": 62970 }, { "grad_norm": 0.04921801760792732, "learning_rate": 2.3442897153499494e-05, "loss": 0.0049, "step": 62980 }, { "grad_norm": 0.04296448454260826, "learning_rate": 2.342745358835224e-05, "loss": 0.0053, "step": 62990 }, { "grad_norm": 0.02735563926398754, "learning_rate": 2.341201355528726e-05, "loss": 0.0048, "step": 63000 }, { "grad_norm": 0.040861427783966064, "learning_rate": 2.3396577056356845e-05, "loss": 0.0051, "step": 63010 }, { "grad_norm": 0.05153246968984604, "learning_rate": 2.338114409361286e-05, "loss": 0.0054, "step": 63020 }, { "grad_norm": 0.05021194368600845, "learning_rate": 2.336571466910669e-05, "loss": 0.0045, "step": 63030 }, { "grad_norm": 0.08317241072654724, "learning_rate": 2.3350288784889257e-05, "loss": 0.007, "step": 63040 }, { "grad_norm": 0.04847182705998421, "learning_rate": 2.3334866443011e-05, "loss": 0.0067, "step": 63050 }, { "grad_norm": 0.08760488033294678, "learning_rate": 2.3319447645521886e-05, "loss": 0.0047, "step": 63060 }, { "grad_norm": 0.03586263582110405, "learning_rate": 2.3304032394471416e-05, "loss": 0.0041, "step": 63070 }, { "grad_norm": 0.05594753101468086, "learning_rate": 2.3288620691908625e-05, "loss": 0.0052, "step": 63080 }, { "grad_norm": 0.035868749022483826, "learning_rate": 2.3273212539882065e-05, "loss": 0.0069, "step": 63090 }, { "grad_norm": 0.04135463759303093, "learning_rate": 2.325780794043983e-05, "loss": 0.0065, "step": 63100 }, { "grad_norm": 0.07379740476608276, "learning_rate": 2.3242406895629548e-05, "loss": 0.0045, "step": 63110 }, { "grad_norm": 0.029207875952124596, "learning_rate": 2.3227009407498302e-05, "loss": 0.0055, "step": 63120 }, { "grad_norm": 0.07102864235639572, "learning_rate": 2.3211615478092823e-05, "loss": 0.005, "step": 63130 }, { "grad_norm": 0.05647137761116028, "learning_rate": 2.319622510945928e-05, "loss": 0.006, "step": 63140 }, { "grad_norm": 0.08409806340932846, "learning_rate": 2.31808383036434e-05, "loss": 0.0071, "step": 63150 }, { "grad_norm": 0.04131869599223137, "learning_rate": 2.3165455062690434e-05, "loss": 0.0048, "step": 63160 }, { "grad_norm": 0.040867097675800323, "learning_rate": 2.3150075388645166e-05, "loss": 0.0061, "step": 63170 }, { "grad_norm": 0.04611004516482353, "learning_rate": 2.313469928355187e-05, "loss": 0.0047, "step": 63180 }, { "grad_norm": 0.09768707305192947, "learning_rate": 2.3119326749454384e-05, "loss": 0.0055, "step": 63190 }, { "grad_norm": 0.048540305346250534, "learning_rate": 2.310395778839604e-05, "loss": 0.0034, "step": 63200 }, { "grad_norm": 0.03539038822054863, "learning_rate": 2.3088592402419762e-05, "loss": 0.0053, "step": 63210 }, { "grad_norm": 0.04855000972747803, "learning_rate": 2.307323059356794e-05, "loss": 0.0069, "step": 63220 }, { "grad_norm": 0.0565280057489872, "learning_rate": 2.3057872363882473e-05, "loss": 0.0061, "step": 63230 }, { "grad_norm": 0.050604891031980515, "learning_rate": 2.3042517715404815e-05, "loss": 0.0064, "step": 63240 }, { "grad_norm": 0.03746378421783447, "learning_rate": 2.3027166650175957e-05, "loss": 0.0049, "step": 63250 }, { "grad_norm": 0.04464656487107277, "learning_rate": 2.3011819170236388e-05, "loss": 0.0065, "step": 63260 }, { "grad_norm": 0.04659015312790871, "learning_rate": 2.299647527762613e-05, "loss": 0.0048, "step": 63270 }, { "grad_norm": 0.05468259006738663, "learning_rate": 2.2981134974384726e-05, "loss": 0.0052, "step": 63280 }, { "grad_norm": 0.06532806158065796, "learning_rate": 2.2965798262551246e-05, "loss": 0.0069, "step": 63290 }, { "grad_norm": 0.05639590322971344, "learning_rate": 2.295046514416428e-05, "loss": 0.0065, "step": 63300 }, { "grad_norm": 0.04042036458849907, "learning_rate": 2.293513562126194e-05, "loss": 0.0058, "step": 63310 }, { "grad_norm": 0.06415963917970657, "learning_rate": 2.291980969588186e-05, "loss": 0.0045, "step": 63320 }, { "grad_norm": 0.04561664164066315, "learning_rate": 2.290448737006122e-05, "loss": 0.0042, "step": 63330 }, { "grad_norm": 0.06092609837651253, "learning_rate": 2.2889168645836646e-05, "loss": 0.0057, "step": 63340 }, { "grad_norm": 0.05901948735117912, "learning_rate": 2.2873853525244354e-05, "loss": 0.0047, "step": 63350 }, { "grad_norm": 0.053812939673662186, "learning_rate": 2.285854201032009e-05, "loss": 0.0051, "step": 63360 }, { "grad_norm": 0.05727718397974968, "learning_rate": 2.2843234103099083e-05, "loss": 0.0049, "step": 63370 }, { "grad_norm": 0.0596567839384079, "learning_rate": 2.282792980561611e-05, "loss": 0.0053, "step": 63380 }, { "grad_norm": 0.04031321033835411, "learning_rate": 2.2812629119905417e-05, "loss": 0.0049, "step": 63390 }, { "grad_norm": 0.07357780635356903, "learning_rate": 2.2797332048000812e-05, "loss": 0.0061, "step": 63400 }, { "grad_norm": 0.04880078136920929, "learning_rate": 2.278203859193564e-05, "loss": 0.0064, "step": 63410 }, { "grad_norm": 0.05445953831076622, "learning_rate": 2.27667487537427e-05, "loss": 0.0053, "step": 63420 }, { "grad_norm": 0.060257863253355026, "learning_rate": 2.2751462535454422e-05, "loss": 0.0052, "step": 63430 }, { "grad_norm": 0.047650255262851715, "learning_rate": 2.2736179939102613e-05, "loss": 0.0048, "step": 63440 }, { "grad_norm": 0.04580908268690109, "learning_rate": 2.272090096671871e-05, "loss": 0.0063, "step": 63450 }, { "grad_norm": 0.042479246854782104, "learning_rate": 2.2705625620333608e-05, "loss": 0.0046, "step": 63460 }, { "grad_norm": 0.07922504842281342, "learning_rate": 2.2690353901977752e-05, "loss": 0.0069, "step": 63470 }, { "grad_norm": 0.0823858305811882, "learning_rate": 2.267508581368109e-05, "loss": 0.0085, "step": 63480 }, { "grad_norm": 0.056128326803445816, "learning_rate": 2.2659821357473088e-05, "loss": 0.0057, "step": 63490 }, { "grad_norm": 0.057784225791692734, "learning_rate": 2.2644560535382737e-05, "loss": 0.0074, "step": 63500 }, { "grad_norm": 0.0417727530002594, "learning_rate": 2.2629303349438535e-05, "loss": 0.0041, "step": 63510 }, { "grad_norm": 0.04947176203131676, "learning_rate": 2.2614049801668497e-05, "loss": 0.0064, "step": 63520 }, { "grad_norm": 0.05908907577395439, "learning_rate": 2.259879989410017e-05, "loss": 0.0069, "step": 63530 }, { "grad_norm": 0.06363611668348312, "learning_rate": 2.2583553628760597e-05, "loss": 0.0054, "step": 63540 }, { "grad_norm": 0.03251321613788605, "learning_rate": 2.2568311007676367e-05, "loss": 0.0065, "step": 63550 }, { "grad_norm": 0.03199226036667824, "learning_rate": 2.2553072032873528e-05, "loss": 0.0066, "step": 63560 }, { "grad_norm": 0.04701106995344162, "learning_rate": 2.2537836706377673e-05, "loss": 0.0098, "step": 63570 }, { "grad_norm": 0.05868149921298027, "learning_rate": 2.252260503021396e-05, "loss": 0.0065, "step": 63580 }, { "grad_norm": 0.056709159165620804, "learning_rate": 2.2507377006407e-05, "loss": 0.0056, "step": 63590 }, { "grad_norm": 0.0335078090429306, "learning_rate": 2.249215263698094e-05, "loss": 0.0041, "step": 63600 }, { "grad_norm": 0.07807176560163498, "learning_rate": 2.2476931923959415e-05, "loss": 0.0049, "step": 63610 }, { "grad_norm": 0.049365367740392685, "learning_rate": 2.246171486936561e-05, "loss": 0.0064, "step": 63620 }, { "grad_norm": 0.08433502167463303, "learning_rate": 2.244650147522221e-05, "loss": 0.0084, "step": 63630 }, { "grad_norm": 0.07229898124933243, "learning_rate": 2.243129174355139e-05, "loss": 0.0051, "step": 63640 }, { "grad_norm": 0.05373765155673027, "learning_rate": 2.2416085676374927e-05, "loss": 0.0097, "step": 63650 }, { "grad_norm": 0.05458454415202141, "learning_rate": 2.2400883275713986e-05, "loss": 0.0042, "step": 63660 }, { "grad_norm": 0.051844265311956406, "learning_rate": 2.2385684543589314e-05, "loss": 0.0056, "step": 63670 }, { "grad_norm": 0.05782318487763405, "learning_rate": 2.2370489482021173e-05, "loss": 0.0076, "step": 63680 }, { "grad_norm": 0.07633274793624878, "learning_rate": 2.2355298093029314e-05, "loss": 0.0059, "step": 63690 }, { "grad_norm": 0.06853263825178146, "learning_rate": 2.234011037863302e-05, "loss": 0.0095, "step": 63700 }, { "grad_norm": 0.039097048342227936, "learning_rate": 2.232492634085107e-05, "loss": 0.0055, "step": 63710 }, { "grad_norm": 0.07336744666099548, "learning_rate": 2.2309745981701764e-05, "loss": 0.0076, "step": 63720 }, { "grad_norm": 0.07867378741502762, "learning_rate": 2.22945693032029e-05, "loss": 0.0047, "step": 63730 }, { "grad_norm": 0.05127411335706711, "learning_rate": 2.227939630737181e-05, "loss": 0.0072, "step": 63740 }, { "grad_norm": 0.04509323462843895, "learning_rate": 2.2264226996225313e-05, "loss": 0.0066, "step": 63750 }, { "grad_norm": 0.06211860477924347, "learning_rate": 2.224906137177975e-05, "loss": 0.007, "step": 63760 }, { "grad_norm": 0.05012645199894905, "learning_rate": 2.2233899436050987e-05, "loss": 0.0072, "step": 63770 }, { "grad_norm": 0.03491188585758209, "learning_rate": 2.221874119105435e-05, "loss": 0.0076, "step": 63780 }, { "grad_norm": 0.03479337692260742, "learning_rate": 2.2203586638804697e-05, "loss": 0.0064, "step": 63790 }, { "grad_norm": 0.07048704475164413, "learning_rate": 2.2188435781316453e-05, "loss": 0.0058, "step": 63800 }, { "grad_norm": 0.046108145266771317, "learning_rate": 2.2173288620603484e-05, "loss": 0.0058, "step": 63810 }, { "grad_norm": 0.060635440051555634, "learning_rate": 2.2158145158679195e-05, "loss": 0.006, "step": 63820 }, { "grad_norm": 0.07628335803747177, "learning_rate": 2.2143005397556454e-05, "loss": 0.004, "step": 63830 }, { "grad_norm": 0.04310767725110054, "learning_rate": 2.2127869339247697e-05, "loss": 0.0034, "step": 63840 }, { "grad_norm": 0.05472598224878311, "learning_rate": 2.2112736985764836e-05, "loss": 0.005, "step": 63850 }, { "grad_norm": 0.030281774699687958, "learning_rate": 2.2097608339119302e-05, "loss": 0.0044, "step": 63860 }, { "grad_norm": 0.05574169382452965, "learning_rate": 2.208248340132203e-05, "loss": 0.007, "step": 63870 }, { "grad_norm": 0.09195204824209213, "learning_rate": 2.2067362174383454e-05, "loss": 0.0056, "step": 63880 }, { "grad_norm": 0.05194179713726044, "learning_rate": 2.2052244660313522e-05, "loss": 0.0091, "step": 63890 }, { "grad_norm": 0.07147505134344101, "learning_rate": 2.2037130861121692e-05, "loss": 0.0083, "step": 63900 }, { "grad_norm": 0.04665535315871239, "learning_rate": 2.202202077881692e-05, "loss": 0.0051, "step": 63910 }, { "grad_norm": 0.030158357694745064, "learning_rate": 2.2006914415407674e-05, "loss": 0.005, "step": 63920 }, { "grad_norm": 0.050769295543432236, "learning_rate": 2.1991811772901927e-05, "loss": 0.0056, "step": 63930 }, { "grad_norm": 0.04633264243602753, "learning_rate": 2.1976712853307156e-05, "loss": 0.0082, "step": 63940 }, { "grad_norm": 0.05440134555101395, "learning_rate": 2.196161765863034e-05, "loss": 0.006, "step": 63950 }, { "grad_norm": 0.04435209929943085, "learning_rate": 2.1946526190877965e-05, "loss": 0.0057, "step": 63960 }, { "grad_norm": 0.04766702279448509, "learning_rate": 2.193143845205603e-05, "loss": 0.0077, "step": 63970 }, { "grad_norm": 0.053215473890304565, "learning_rate": 2.1916354444170023e-05, "loss": 0.0043, "step": 63980 }, { "grad_norm": 0.06387105584144592, "learning_rate": 2.1901274169224967e-05, "loss": 0.0042, "step": 63990 }, { "grad_norm": 0.07915130257606506, "learning_rate": 2.1886197629225326e-05, "loss": 0.0051, "step": 64000 }, { "grad_norm": 0.0288010872900486, "learning_rate": 2.187112482617511e-05, "loss": 0.0054, "step": 64010 }, { "grad_norm": 0.07733071595430374, "learning_rate": 2.1856055762077866e-05, "loss": 0.0059, "step": 64020 }, { "grad_norm": 0.03076087310910225, "learning_rate": 2.1840990438936586e-05, "loss": 0.004, "step": 64030 }, { "grad_norm": 0.024196075275540352, "learning_rate": 2.1825928858753807e-05, "loss": 0.0094, "step": 64040 }, { "grad_norm": 0.06667477637529373, "learning_rate": 2.1810871023531516e-05, "loss": 0.0077, "step": 64050 }, { "grad_norm": 0.0501326285302639, "learning_rate": 2.1795816935271245e-05, "loss": 0.0067, "step": 64060 }, { "grad_norm": 0.048434849828481674, "learning_rate": 2.1780766595974022e-05, "loss": 0.0072, "step": 64070 }, { "grad_norm": 0.03511356934905052, "learning_rate": 2.176572000764037e-05, "loss": 0.0051, "step": 64080 }, { "grad_norm": 0.040983039885759354, "learning_rate": 2.175067717227032e-05, "loss": 0.0059, "step": 64090 }, { "grad_norm": 0.07157811522483826, "learning_rate": 2.1735638091863386e-05, "loss": 0.0061, "step": 64100 }, { "grad_norm": 0.06681091338396072, "learning_rate": 2.172060276841861e-05, "loss": 0.0055, "step": 64110 }, { "grad_norm": 0.04808767884969711, "learning_rate": 2.170557120393451e-05, "loss": 0.0042, "step": 64120 }, { "grad_norm": 0.08886067569255829, "learning_rate": 2.1690543400409126e-05, "loss": 0.0069, "step": 64130 }, { "grad_norm": 0.033144764602184296, "learning_rate": 2.167551935983998e-05, "loss": 0.0037, "step": 64140 }, { "grad_norm": 0.05302438139915466, "learning_rate": 2.1660499084224122e-05, "loss": 0.0037, "step": 64150 }, { "grad_norm": 0.04915217310190201, "learning_rate": 2.1645482575558022e-05, "loss": 0.0071, "step": 64160 }, { "grad_norm": 0.04104466736316681, "learning_rate": 2.1630469835837768e-05, "loss": 0.0057, "step": 64170 }, { "grad_norm": 0.06695683300495148, "learning_rate": 2.1615460867058867e-05, "loss": 0.005, "step": 64180 }, { "grad_norm": 0.0353938490152359, "learning_rate": 2.1600455671216342e-05, "loss": 0.0063, "step": 64190 }, { "grad_norm": 0.08922175318002701, "learning_rate": 2.1585454250304737e-05, "loss": 0.0058, "step": 64200 }, { "grad_norm": 0.0470687672495842, "learning_rate": 2.157045660631804e-05, "loss": 0.0036, "step": 64210 }, { "grad_norm": 0.04471772164106369, "learning_rate": 2.1555462741249787e-05, "loss": 0.0095, "step": 64220 }, { "grad_norm": 0.05611778050661087, "learning_rate": 2.1540472657092997e-05, "loss": 0.0057, "step": 64230 }, { "grad_norm": 0.06546790152788162, "learning_rate": 2.152548635584017e-05, "loss": 0.0049, "step": 64240 }, { "grad_norm": 0.059169892221689224, "learning_rate": 2.1510503839483348e-05, "loss": 0.0072, "step": 64250 }, { "grad_norm": 0.0385211780667305, "learning_rate": 2.1495525110014048e-05, "loss": 0.0056, "step": 64260 }, { "grad_norm": 0.02753029204905033, "learning_rate": 2.1480550169423234e-05, "loss": 0.0043, "step": 64270 }, { "grad_norm": 0.04363272339105606, "learning_rate": 2.1465579019701433e-05, "loss": 0.0057, "step": 64280 }, { "grad_norm": 0.042169734835624695, "learning_rate": 2.145061166283863e-05, "loss": 0.0065, "step": 64290 }, { "grad_norm": 0.06717311590909958, "learning_rate": 2.143564810082434e-05, "loss": 0.0043, "step": 64300 }, { "grad_norm": 0.04892963171005249, "learning_rate": 2.1420688335647536e-05, "loss": 0.0045, "step": 64310 }, { "grad_norm": 0.029828380793333054, "learning_rate": 2.1405732369296717e-05, "loss": 0.0077, "step": 64320 }, { "grad_norm": 0.04441197216510773, "learning_rate": 2.139078020375985e-05, "loss": 0.0047, "step": 64330 }, { "grad_norm": 0.08758074790239334, "learning_rate": 2.1375831841024425e-05, "loss": 0.0064, "step": 64340 }, { "grad_norm": 0.07263020426034927, "learning_rate": 2.1360887283077396e-05, "loss": 0.0055, "step": 64350 }, { "grad_norm": 0.04225365072488785, "learning_rate": 2.1345946531905236e-05, "loss": 0.004, "step": 64360 }, { "grad_norm": 0.06528284400701523, "learning_rate": 2.1331009589493922e-05, "loss": 0.0042, "step": 64370 }, { "grad_norm": 0.0688774362206459, "learning_rate": 2.1316076457828872e-05, "loss": 0.0055, "step": 64380 }, { "grad_norm": 0.03255524858832359, "learning_rate": 2.1301147138895016e-05, "loss": 0.0053, "step": 64390 }, { "grad_norm": 0.02856575883924961, "learning_rate": 2.1286221634676844e-05, "loss": 0.006, "step": 64400 }, { "grad_norm": 0.13237544894218445, "learning_rate": 2.1271299947158267e-05, "loss": 0.0058, "step": 64410 }, { "grad_norm": 0.027046967297792435, "learning_rate": 2.1256382078322724e-05, "loss": 0.0043, "step": 64420 }, { "grad_norm": 0.07042522728443146, "learning_rate": 2.1241468030153095e-05, "loss": 0.0067, "step": 64430 }, { "grad_norm": 0.06135185435414314, "learning_rate": 2.1226557804631803e-05, "loss": 0.0046, "step": 64440 }, { "grad_norm": 0.050831686705350876, "learning_rate": 2.1211651403740756e-05, "loss": 0.0055, "step": 64450 }, { "grad_norm": 0.10926344245672226, "learning_rate": 2.1196748829461328e-05, "loss": 0.0067, "step": 64460 }, { "grad_norm": 0.03669844567775726, "learning_rate": 2.1181850083774453e-05, "loss": 0.0065, "step": 64470 }, { "grad_norm": 0.05244641378521919, "learning_rate": 2.116695516866045e-05, "loss": 0.0045, "step": 64480 }, { "grad_norm": 0.040500327944755554, "learning_rate": 2.115206408609921e-05, "loss": 0.0042, "step": 64490 }, { "grad_norm": 0.07680094987154007, "learning_rate": 2.1137176838070077e-05, "loss": 0.0062, "step": 64500 }, { "grad_norm": 0.059603817760944366, "learning_rate": 2.1122293426551908e-05, "loss": 0.0058, "step": 64510 }, { "grad_norm": 0.04355907812714577, "learning_rate": 2.1107413853523035e-05, "loss": 0.0059, "step": 64520 }, { "grad_norm": 0.07510403543710709, "learning_rate": 2.109253812096128e-05, "loss": 0.0054, "step": 64530 }, { "grad_norm": 0.04881104826927185, "learning_rate": 2.1077666230843966e-05, "loss": 0.0047, "step": 64540 }, { "grad_norm": 0.06134257838129997, "learning_rate": 2.1062798185147896e-05, "loss": 0.0057, "step": 64550 }, { "grad_norm": 0.0644170418381691, "learning_rate": 2.1047933985849368e-05, "loss": 0.0042, "step": 64560 }, { "grad_norm": 0.05292896553874016, "learning_rate": 2.103307363492415e-05, "loss": 0.0063, "step": 64570 }, { "grad_norm": 0.05036536231637001, "learning_rate": 2.1018217134347524e-05, "loss": 0.0051, "step": 64580 }, { "grad_norm": 0.07496623694896698, "learning_rate": 2.100336448609427e-05, "loss": 0.0049, "step": 64590 }, { "grad_norm": 0.04104249179363251, "learning_rate": 2.0988515692138595e-05, "loss": 0.0072, "step": 64600 }, { "grad_norm": 0.08431030064821243, "learning_rate": 2.0973670754454228e-05, "loss": 0.0075, "step": 64610 }, { "grad_norm": 0.04281897470355034, "learning_rate": 2.0958829675014436e-05, "loss": 0.005, "step": 64620 }, { "grad_norm": 0.03992373123764992, "learning_rate": 2.0943992455791905e-05, "loss": 0.0069, "step": 64630 }, { "grad_norm": 0.028137264773249626, "learning_rate": 2.0929159098758862e-05, "loss": 0.0046, "step": 64640 }, { "grad_norm": 0.05833766236901283, "learning_rate": 2.0914329605886933e-05, "loss": 0.005, "step": 64650 }, { "grad_norm": 0.08089729398488998, "learning_rate": 2.0899503979147316e-05, "loss": 0.0064, "step": 64660 }, { "grad_norm": 0.09597761929035187, "learning_rate": 2.088468222051067e-05, "loss": 0.0083, "step": 64670 }, { "grad_norm": 0.06741323322057724, "learning_rate": 2.086986433194712e-05, "loss": 0.0053, "step": 64680 }, { "grad_norm": 0.04819035157561302, "learning_rate": 2.0855050315426334e-05, "loss": 0.0066, "step": 64690 }, { "grad_norm": 0.05421127751469612, "learning_rate": 2.084024017291738e-05, "loss": 0.0047, "step": 64700 }, { "grad_norm": 0.04381706938147545, "learning_rate": 2.0825433906388868e-05, "loss": 0.0035, "step": 64710 }, { "grad_norm": 0.04831201210618019, "learning_rate": 2.081063151780889e-05, "loss": 0.0059, "step": 64720 }, { "grad_norm": 0.041758663952350616, "learning_rate": 2.0795833009145003e-05, "loss": 0.0042, "step": 64730 }, { "grad_norm": 0.05724142864346504, "learning_rate": 2.078103838236426e-05, "loss": 0.0047, "step": 64740 }, { "grad_norm": 0.047791849821805954, "learning_rate": 2.0766247639433195e-05, "loss": 0.0052, "step": 64750 }, { "grad_norm": 0.05477696657180786, "learning_rate": 2.0751460782317834e-05, "loss": 0.0074, "step": 64760 }, { "grad_norm": 0.026717809960246086, "learning_rate": 2.0736677812983664e-05, "loss": 0.0058, "step": 64770 }, { "grad_norm": 0.030726201832294464, "learning_rate": 2.0721898733395685e-05, "loss": 0.0043, "step": 64780 }, { "grad_norm": 0.0664835274219513, "learning_rate": 2.070712354551836e-05, "loss": 0.0066, "step": 64790 }, { "grad_norm": 0.04897609353065491, "learning_rate": 2.069235225131564e-05, "loss": 0.005, "step": 64800 }, { "grad_norm": 0.03908294811844826, "learning_rate": 2.0677584852750975e-05, "loss": 0.0062, "step": 64810 }, { "grad_norm": 0.04687197506427765, "learning_rate": 2.0662821351787242e-05, "loss": 0.0045, "step": 64820 }, { "grad_norm": 0.07815449684858322, "learning_rate": 2.064806175038685e-05, "loss": 0.0108, "step": 64830 }, { "grad_norm": 0.06800070405006409, "learning_rate": 2.0633306050511696e-05, "loss": 0.0091, "step": 64840 }, { "grad_norm": 0.06012950837612152, "learning_rate": 2.0618554254123146e-05, "loss": 0.0072, "step": 64850 }, { "grad_norm": 0.06052400544285774, "learning_rate": 2.0603806363182036e-05, "loss": 0.0066, "step": 64860 }, { "grad_norm": 0.05036541074514389, "learning_rate": 2.0589062379648673e-05, "loss": 0.006, "step": 64870 }, { "grad_norm": 0.06917441636323929, "learning_rate": 2.0574322305482863e-05, "loss": 0.0056, "step": 64880 }, { "grad_norm": 0.05073929578065872, "learning_rate": 2.0559586142643893e-05, "loss": 0.0059, "step": 64890 }, { "grad_norm": 0.04585650563240051, "learning_rate": 2.054485389309053e-05, "loss": 0.0043, "step": 64900 }, { "grad_norm": 0.04698082432150841, "learning_rate": 2.0530125558781022e-05, "loss": 0.0041, "step": 64910 }, { "grad_norm": 0.03414018824696541, "learning_rate": 2.0515401141673084e-05, "loss": 0.0053, "step": 64920 }, { "grad_norm": 0.022027358412742615, "learning_rate": 2.050068064372392e-05, "loss": 0.0054, "step": 64930 }, { "grad_norm": 0.053984321653842926, "learning_rate": 2.048596406689021e-05, "loss": 0.0061, "step": 64940 }, { "grad_norm": 0.023706527426838875, "learning_rate": 2.0471251413128124e-05, "loss": 0.0055, "step": 64950 }, { "grad_norm": 0.0562567412853241, "learning_rate": 2.045654268439329e-05, "loss": 0.0047, "step": 64960 }, { "grad_norm": 0.06775089353322983, "learning_rate": 2.0441837882640846e-05, "loss": 0.0048, "step": 64970 }, { "grad_norm": 0.05396319180727005, "learning_rate": 2.0427137009825336e-05, "loss": 0.0054, "step": 64980 }, { "grad_norm": 0.08995328098535538, "learning_rate": 2.0412440067900886e-05, "loss": 0.009, "step": 64990 }, { "grad_norm": 0.06275437027215958, "learning_rate": 2.0397747058821022e-05, "loss": 0.0055, "step": 65000 }, { "grad_norm": 0.06199713796377182, "learning_rate": 2.038305798453878e-05, "loss": 0.0064, "step": 65010 }, { "grad_norm": 0.044270820915699005, "learning_rate": 2.0368372847006662e-05, "loss": 0.0045, "step": 65020 }, { "grad_norm": 0.04005991667509079, "learning_rate": 2.0353691648176665e-05, "loss": 0.007, "step": 65030 }, { "grad_norm": 0.03142046928405762, "learning_rate": 2.033901439000021e-05, "loss": 0.0056, "step": 65040 }, { "grad_norm": 0.029490845277905464, "learning_rate": 2.032434107442824e-05, "loss": 0.0069, "step": 65050 }, { "grad_norm": 0.05935296788811684, "learning_rate": 2.0309671703411186e-05, "loss": 0.0054, "step": 65060 }, { "grad_norm": 0.04722663387656212, "learning_rate": 2.029500627889893e-05, "loss": 0.0056, "step": 65070 }, { "grad_norm": 0.049208153039216995, "learning_rate": 2.0280344802840846e-05, "loss": 0.0064, "step": 65080 }, { "grad_norm": 0.03927651047706604, "learning_rate": 2.026568727718573e-05, "loss": 0.0055, "step": 65090 }, { "grad_norm": 0.040073104202747345, "learning_rate": 2.025103370388191e-05, "loss": 0.0063, "step": 65100 }, { "grad_norm": 0.058212101459503174, "learning_rate": 2.023638408487718e-05, "loss": 0.007, "step": 65110 }, { "grad_norm": 0.05714912712574005, "learning_rate": 2.0221738422118795e-05, "loss": 0.0067, "step": 65120 }, { "grad_norm": 0.046520378440618515, "learning_rate": 2.0207096717553493e-05, "loss": 0.0081, "step": 65130 }, { "grad_norm": 0.045920487493276596, "learning_rate": 2.0192458973127475e-05, "loss": 0.007, "step": 65140 }, { "grad_norm": 0.04820643365383148, "learning_rate": 2.017782519078643e-05, "loss": 0.0066, "step": 65150 }, { "grad_norm": 0.051409050822257996, "learning_rate": 2.016319537247551e-05, "loss": 0.0054, "step": 65160 }, { "grad_norm": 0.0805073082447052, "learning_rate": 2.0148569520139342e-05, "loss": 0.0059, "step": 65170 }, { "grad_norm": 0.06668145954608917, "learning_rate": 2.0133947635722034e-05, "loss": 0.003, "step": 65180 }, { "grad_norm": 0.06484125554561615, "learning_rate": 2.0119329721167163e-05, "loss": 0.006, "step": 65190 }, { "grad_norm": 0.05052676424384117, "learning_rate": 2.010471577841773e-05, "loss": 0.0045, "step": 65200 }, { "grad_norm": 0.03334025666117668, "learning_rate": 2.0090105809416315e-05, "loss": 0.0082, "step": 65210 }, { "grad_norm": 0.05985872447490692, "learning_rate": 2.0075499816104876e-05, "loss": 0.0064, "step": 65220 }, { "grad_norm": 0.07132764160633087, "learning_rate": 2.0060897800424876e-05, "loss": 0.0049, "step": 65230 }, { "grad_norm": 0.0624660924077034, "learning_rate": 2.004629976431727e-05, "loss": 0.0054, "step": 65240 }, { "grad_norm": 0.052251316606998444, "learning_rate": 2.0031705709722425e-05, "loss": 0.0066, "step": 65250 }, { "grad_norm": 0.052066341042518616, "learning_rate": 2.0017115638580236e-05, "loss": 0.0048, "step": 65260 }, { "grad_norm": 0.04164468124508858, "learning_rate": 2.0002529552830046e-05, "loss": 0.003, "step": 65270 }, { "grad_norm": 0.04486474767327309, "learning_rate": 1.998794745441065e-05, "loss": 0.0079, "step": 65280 }, { "grad_norm": 0.05443452298641205, "learning_rate": 1.9973369345260383e-05, "loss": 0.0043, "step": 65290 }, { "grad_norm": 0.05540773645043373, "learning_rate": 1.9958795227316957e-05, "loss": 0.0079, "step": 65300 }, { "grad_norm": 0.04544485732913017, "learning_rate": 1.9944225102517607e-05, "loss": 0.0038, "step": 65310 }, { "grad_norm": 0.031886812299489975, "learning_rate": 1.992965897279902e-05, "loss": 0.0048, "step": 65320 }, { "grad_norm": 0.0328783243894577, "learning_rate": 1.9915096840097375e-05, "loss": 0.0048, "step": 65330 }, { "grad_norm": 0.060570523142814636, "learning_rate": 1.9900538706348294e-05, "loss": 0.0046, "step": 65340 }, { "grad_norm": 0.04890760779380798, "learning_rate": 1.988598457348687e-05, "loss": 0.0056, "step": 65350 }, { "grad_norm": 0.06626826524734497, "learning_rate": 1.9871434443447688e-05, "loss": 0.0066, "step": 65360 }, { "grad_norm": 0.065765880048275, "learning_rate": 1.985688831816477e-05, "loss": 0.0043, "step": 65370 }, { "grad_norm": 0.03844212740659714, "learning_rate": 1.9842346199571622e-05, "loss": 0.0049, "step": 65380 }, { "grad_norm": 0.041703514754772186, "learning_rate": 1.9827808089601214e-05, "loss": 0.0053, "step": 65390 }, { "grad_norm": 0.074385866522789, "learning_rate": 1.9813273990185993e-05, "loss": 0.0069, "step": 65400 }, { "grad_norm": 0.03563438355922699, "learning_rate": 1.9798743903257864e-05, "loss": 0.0071, "step": 65410 }, { "grad_norm": 0.08557343482971191, "learning_rate": 1.9784217830748186e-05, "loss": 0.0055, "step": 65420 }, { "grad_norm": 0.03946897014975548, "learning_rate": 1.9769695774587782e-05, "loss": 0.0052, "step": 65430 }, { "grad_norm": 0.0534655787050724, "learning_rate": 1.975517773670699e-05, "loss": 0.0043, "step": 65440 }, { "grad_norm": 0.07304932177066803, "learning_rate": 1.9740663719035574e-05, "loss": 0.0058, "step": 65450 }, { "grad_norm": 0.07714059948921204, "learning_rate": 1.9726153723502776e-05, "loss": 0.0066, "step": 65460 }, { "grad_norm": 0.06389681249856949, "learning_rate": 1.9711647752037265e-05, "loss": 0.0063, "step": 65470 }, { "grad_norm": 0.039799198508262634, "learning_rate": 1.969714580656723e-05, "loss": 0.0043, "step": 65480 }, { "grad_norm": 0.04374902695417404, "learning_rate": 1.9682647889020295e-05, "loss": 0.0073, "step": 65490 }, { "grad_norm": 0.07789991796016693, "learning_rate": 1.966815400132354e-05, "loss": 0.0048, "step": 65500 }, { "grad_norm": 0.06114752218127251, "learning_rate": 1.9653664145403583e-05, "loss": 0.0068, "step": 65510 }, { "grad_norm": 0.05707956850528717, "learning_rate": 1.963917832318639e-05, "loss": 0.006, "step": 65520 }, { "grad_norm": 0.062210261821746826, "learning_rate": 1.9624696536597464e-05, "loss": 0.0082, "step": 65530 }, { "grad_norm": 0.05353374406695366, "learning_rate": 1.961021878756176e-05, "loss": 0.0056, "step": 65540 }, { "grad_norm": 0.05777953937649727, "learning_rate": 1.9595745078003697e-05, "loss": 0.0079, "step": 65550 }, { "grad_norm": 0.05699213966727257, "learning_rate": 1.958127540984715e-05, "loss": 0.0053, "step": 65560 }, { "grad_norm": 0.08130497485399246, "learning_rate": 1.9566809785015456e-05, "loss": 0.0079, "step": 65570 }, { "grad_norm": 0.048830438405275345, "learning_rate": 1.9552348205431413e-05, "loss": 0.0062, "step": 65580 }, { "grad_norm": 0.05768207088112831, "learning_rate": 1.95378906730173e-05, "loss": 0.0046, "step": 65590 }, { "grad_norm": 0.03401189669966698, "learning_rate": 1.952343718969484e-05, "loss": 0.0034, "step": 65600 }, { "grad_norm": 0.06608910113573074, "learning_rate": 1.9508987757385216e-05, "loss": 0.006, "step": 65610 }, { "grad_norm": 0.057970866560935974, "learning_rate": 1.949454237800909e-05, "loss": 0.0045, "step": 65620 }, { "grad_norm": 0.08476462960243225, "learning_rate": 1.9480101053486576e-05, "loss": 0.0044, "step": 65630 }, { "grad_norm": 0.041443560272455215, "learning_rate": 1.9465663785737226e-05, "loss": 0.0043, "step": 65640 }, { "grad_norm": 0.04898235574364662, "learning_rate": 1.9451230576680068e-05, "loss": 0.0044, "step": 65650 }, { "grad_norm": 0.05693214759230614, "learning_rate": 1.9436801428233637e-05, "loss": 0.0037, "step": 65660 }, { "grad_norm": 0.05095553398132324, "learning_rate": 1.9422376342315863e-05, "loss": 0.0069, "step": 65670 }, { "grad_norm": 0.04238869994878769, "learning_rate": 1.940795532084418e-05, "loss": 0.0061, "step": 65680 }, { "grad_norm": 0.09574975818395615, "learning_rate": 1.9393538365735424e-05, "loss": 0.0057, "step": 65690 }, { "grad_norm": 0.031607627868652344, "learning_rate": 1.9379125478905958e-05, "loss": 0.0076, "step": 65700 }, { "grad_norm": 0.0257858969271183, "learning_rate": 1.936471666227156e-05, "loss": 0.0027, "step": 65710 }, { "grad_norm": 0.06907524168491364, "learning_rate": 1.9350311917747476e-05, "loss": 0.0065, "step": 65720 }, { "grad_norm": 0.05159110203385353, "learning_rate": 1.9335911247248466e-05, "loss": 0.0055, "step": 65730 }, { "grad_norm": 0.05554944649338722, "learning_rate": 1.932151465268865e-05, "loss": 0.0064, "step": 65740 }, { "grad_norm": 0.06711466610431671, "learning_rate": 1.9307122135981663e-05, "loss": 0.0046, "step": 65750 }, { "grad_norm": 0.08586609363555908, "learning_rate": 1.92927336990406e-05, "loss": 0.0055, "step": 65760 }, { "grad_norm": 0.05497932434082031, "learning_rate": 1.9278349343778e-05, "loss": 0.0071, "step": 65770 }, { "grad_norm": 0.07577919960021973, "learning_rate": 1.926396907210587e-05, "loss": 0.0043, "step": 65780 }, { "grad_norm": 0.045239418745040894, "learning_rate": 1.9249592885935657e-05, "loss": 0.0068, "step": 65790 }, { "grad_norm": 0.04452167823910713, "learning_rate": 1.9235220787178282e-05, "loss": 0.0056, "step": 65800 }, { "grad_norm": 0.05067760497331619, "learning_rate": 1.922085277774412e-05, "loss": 0.0058, "step": 65810 }, { "grad_norm": 0.025595111772418022, "learning_rate": 1.9206488859542987e-05, "loss": 0.007, "step": 65820 }, { "grad_norm": 0.047335024923086166, "learning_rate": 1.919212903448418e-05, "loss": 0.0046, "step": 65830 }, { "grad_norm": 0.04865461587905884, "learning_rate": 1.9177773304476436e-05, "loss": 0.0033, "step": 65840 }, { "grad_norm": 0.06787668913602829, "learning_rate": 1.916342167142796e-05, "loss": 0.0055, "step": 65850 }, { "grad_norm": 0.06106413155794144, "learning_rate": 1.914907413724638e-05, "loss": 0.0062, "step": 65860 }, { "grad_norm": 0.05136648565530777, "learning_rate": 1.9134730703838792e-05, "loss": 0.0041, "step": 65870 }, { "grad_norm": 0.03675634786486626, "learning_rate": 1.9120391373111802e-05, "loss": 0.0043, "step": 65880 }, { "grad_norm": 0.03722289577126503, "learning_rate": 1.91060561469714e-05, "loss": 0.004, "step": 65890 }, { "grad_norm": 0.08752502501010895, "learning_rate": 1.909172502732308e-05, "loss": 0.0065, "step": 65900 }, { "grad_norm": 0.04481630399823189, "learning_rate": 1.9077398016071728e-05, "loss": 0.0044, "step": 65910 }, { "grad_norm": 0.023629093542695045, "learning_rate": 1.9063075115121737e-05, "loss": 0.0079, "step": 65920 }, { "grad_norm": 0.028834450989961624, "learning_rate": 1.9048756326376942e-05, "loss": 0.0053, "step": 65930 }, { "grad_norm": 0.033018458634614944, "learning_rate": 1.9034441651740635e-05, "loss": 0.0063, "step": 65940 }, { "grad_norm": 0.025449827313423157, "learning_rate": 1.9020131093115544e-05, "loss": 0.0032, "step": 65950 }, { "grad_norm": 0.05887044593691826, "learning_rate": 1.9005824652403864e-05, "loss": 0.004, "step": 65960 }, { "grad_norm": 0.03504398092627525, "learning_rate": 1.8991522331507245e-05, "loss": 0.0074, "step": 65970 }, { "grad_norm": 0.04534995183348656, "learning_rate": 1.897722413232677e-05, "loss": 0.005, "step": 65980 }, { "grad_norm": 0.03921996057033539, "learning_rate": 1.8962930056763e-05, "loss": 0.0049, "step": 65990 }, { "grad_norm": 0.04776621609926224, "learning_rate": 1.8948640106715932e-05, "loss": 0.0052, "step": 66000 }, { "grad_norm": 0.07488488405942917, "learning_rate": 1.893435428408504e-05, "loss": 0.0053, "step": 66010 }, { "grad_norm": 0.038565777242183685, "learning_rate": 1.8920072590769162e-05, "loss": 0.0048, "step": 66020 }, { "grad_norm": 0.04658007621765137, "learning_rate": 1.8905795028666713e-05, "loss": 0.0048, "step": 66030 }, { "grad_norm": 0.06259360909461975, "learning_rate": 1.8891521599675482e-05, "loss": 0.0038, "step": 66040 }, { "grad_norm": 0.07053906470537186, "learning_rate": 1.887725230569273e-05, "loss": 0.0053, "step": 66050 }, { "grad_norm": 0.06702244281768799, "learning_rate": 1.8862987148615153e-05, "loss": 0.005, "step": 66060 }, { "grad_norm": 0.03970443829894066, "learning_rate": 1.884872613033893e-05, "loss": 0.0031, "step": 66070 }, { "grad_norm": 0.03384030610322952, "learning_rate": 1.8834469252759628e-05, "loss": 0.0052, "step": 66080 }, { "grad_norm": 0.037195414304733276, "learning_rate": 1.882021651777231e-05, "loss": 0.0042, "step": 66090 }, { "grad_norm": 0.039370764046907425, "learning_rate": 1.8805967927271517e-05, "loss": 0.0056, "step": 66100 }, { "grad_norm": 0.03668012470006943, "learning_rate": 1.879172348315118e-05, "loss": 0.0036, "step": 66110 }, { "grad_norm": 0.0509507954120636, "learning_rate": 1.877748318730472e-05, "loss": 0.0053, "step": 66120 }, { "grad_norm": 0.032409101724624634, "learning_rate": 1.876324704162496e-05, "loss": 0.0033, "step": 66130 }, { "grad_norm": 0.051028311252593994, "learning_rate": 1.874901504800421e-05, "loss": 0.0039, "step": 66140 }, { "grad_norm": 0.051477283239364624, "learning_rate": 1.8734787208334225e-05, "loss": 0.0053, "step": 66150 }, { "grad_norm": 0.08065918833017349, "learning_rate": 1.8720563524506196e-05, "loss": 0.0073, "step": 66160 }, { "grad_norm": 0.06150718033313751, "learning_rate": 1.8706343998410767e-05, "loss": 0.0059, "step": 66170 }, { "grad_norm": 0.033660974353551865, "learning_rate": 1.8692128631938038e-05, "loss": 0.0045, "step": 66180 }, { "grad_norm": 0.028766633942723274, "learning_rate": 1.867791742697754e-05, "loss": 0.0063, "step": 66190 }, { "grad_norm": 0.0873110219836235, "learning_rate": 1.866371038541826e-05, "loss": 0.0049, "step": 66200 }, { "grad_norm": 0.05788842588663101, "learning_rate": 1.864950750914863e-05, "loss": 0.0052, "step": 66210 }, { "grad_norm": 0.05890841409564018, "learning_rate": 1.8635308800056523e-05, "loss": 0.0052, "step": 66220 }, { "grad_norm": 0.035173702985048294, "learning_rate": 1.862111426002928e-05, "loss": 0.0035, "step": 66230 }, { "grad_norm": 0.05684050917625427, "learning_rate": 1.860692389095363e-05, "loss": 0.0049, "step": 66240 }, { "grad_norm": 0.05487486720085144, "learning_rate": 1.8592737694715834e-05, "loss": 0.0045, "step": 66250 }, { "grad_norm": 0.024004491046071053, "learning_rate": 1.857855567320153e-05, "loss": 0.0069, "step": 66260 }, { "grad_norm": 0.048742085695266724, "learning_rate": 1.8564377828295838e-05, "loss": 0.004, "step": 66270 }, { "grad_norm": 0.07509780675172806, "learning_rate": 1.8550204161883312e-05, "loss": 0.004, "step": 66280 }, { "grad_norm": 0.049569614231586456, "learning_rate": 1.8536034675847917e-05, "loss": 0.0046, "step": 66290 }, { "grad_norm": 0.038739681243896484, "learning_rate": 1.8521869372073115e-05, "loss": 0.0068, "step": 66300 }, { "grad_norm": 0.0488414466381073, "learning_rate": 1.8507708252441786e-05, "loss": 0.0049, "step": 66310 }, { "grad_norm": 0.03204362466931343, "learning_rate": 1.849355131883624e-05, "loss": 0.0033, "step": 66320 }, { "grad_norm": 0.052626144140958786, "learning_rate": 1.8479398573138297e-05, "loss": 0.0045, "step": 66330 }, { "grad_norm": 0.058672986924648285, "learning_rate": 1.8465250017229125e-05, "loss": 0.0067, "step": 66340 }, { "grad_norm": 0.06363749504089355, "learning_rate": 1.8451105652989402e-05, "loss": 0.0053, "step": 66350 }, { "grad_norm": 0.07433822751045227, "learning_rate": 1.843696548229922e-05, "loss": 0.0046, "step": 66360 }, { "grad_norm": 0.04843563959002495, "learning_rate": 1.8422829507038125e-05, "loss": 0.0047, "step": 66370 }, { "grad_norm": 0.05979331582784653, "learning_rate": 1.8408697729085107e-05, "loss": 0.0053, "step": 66380 }, { "grad_norm": 0.048393554985523224, "learning_rate": 1.839457015031859e-05, "loss": 0.0048, "step": 66390 }, { "grad_norm": 0.04331037029623985, "learning_rate": 1.8380446772616444e-05, "loss": 0.005, "step": 66400 }, { "grad_norm": 0.05262232571840286, "learning_rate": 1.8366327597855974e-05, "loss": 0.0097, "step": 66410 }, { "grad_norm": 0.055290475487709045, "learning_rate": 1.8352212627913946e-05, "loss": 0.0045, "step": 66420 }, { "grad_norm": 0.04779213294386864, "learning_rate": 1.8338101864666536e-05, "loss": 0.0061, "step": 66430 }, { "grad_norm": 0.019090311601758003, "learning_rate": 1.832399530998939e-05, "loss": 0.0044, "step": 66440 }, { "grad_norm": 0.04602600634098053, "learning_rate": 1.8309892965757595e-05, "loss": 0.0053, "step": 66450 }, { "grad_norm": 0.04523500055074692, "learning_rate": 1.8295794833845643e-05, "loss": 0.0057, "step": 66460 }, { "grad_norm": 0.056708429008722305, "learning_rate": 1.828170091612747e-05, "loss": 0.0045, "step": 66470 }, { "grad_norm": 0.07578369230031967, "learning_rate": 1.8267611214476523e-05, "loss": 0.0084, "step": 66480 }, { "grad_norm": 0.0659320056438446, "learning_rate": 1.8253525730765602e-05, "loss": 0.0064, "step": 66490 }, { "grad_norm": 0.0652320459485054, "learning_rate": 1.8239444466867017e-05, "loss": 0.0053, "step": 66500 }, { "grad_norm": 0.03920958936214447, "learning_rate": 1.8225367424652433e-05, "loss": 0.0062, "step": 66510 }, { "grad_norm": 0.05766172707080841, "learning_rate": 1.821129460599303e-05, "loss": 0.0058, "step": 66520 }, { "grad_norm": 0.04111403226852417, "learning_rate": 1.8197226012759383e-05, "loss": 0.0049, "step": 66530 }, { "grad_norm": 0.025272348895668983, "learning_rate": 1.8183161646821517e-05, "loss": 0.0065, "step": 66540 }, { "grad_norm": 0.026808002963662148, "learning_rate": 1.8169101510048947e-05, "loss": 0.0028, "step": 66550 }, { "grad_norm": 0.056620676070451736, "learning_rate": 1.8155045604310528e-05, "loss": 0.0051, "step": 66560 }, { "grad_norm": 0.04392124339938164, "learning_rate": 1.8140993931474616e-05, "loss": 0.0056, "step": 66570 }, { "grad_norm": 0.049581244587898254, "learning_rate": 1.812694649340899e-05, "loss": 0.0082, "step": 66580 }, { "grad_norm": 0.048917412757873535, "learning_rate": 1.8112903291980872e-05, "loss": 0.0055, "step": 66590 }, { "grad_norm": 0.07443082332611084, "learning_rate": 1.809886432905691e-05, "loss": 0.0044, "step": 66600 }, { "grad_norm": 0.0470687635242939, "learning_rate": 1.8084829606503196e-05, "loss": 0.0052, "step": 66610 }, { "grad_norm": 0.044812966138124466, "learning_rate": 1.807079912618526e-05, "loss": 0.0061, "step": 66620 }, { "grad_norm": 0.04350171610713005, "learning_rate": 1.805677288996806e-05, "loss": 0.0053, "step": 66630 }, { "grad_norm": 0.04202380031347275, "learning_rate": 1.8042750899716e-05, "loss": 0.0049, "step": 66640 }, { "grad_norm": 0.06744678318500519, "learning_rate": 1.802873315729291e-05, "loss": 0.0066, "step": 66650 }, { "grad_norm": 0.03938886895775795, "learning_rate": 1.8014719664562065e-05, "loss": 0.0048, "step": 66660 }, { "grad_norm": 0.0382169634103775, "learning_rate": 1.8000710423386176e-05, "loss": 0.0055, "step": 66670 }, { "grad_norm": 0.05114872753620148, "learning_rate": 1.798670543562736e-05, "loss": 0.0042, "step": 66680 }, { "grad_norm": 0.04431029036641121, "learning_rate": 1.7972704703147185e-05, "loss": 0.0068, "step": 66690 }, { "grad_norm": 0.04631525278091431, "learning_rate": 1.79587082278067e-05, "loss": 0.0082, "step": 66700 }, { "grad_norm": 0.04887477681040764, "learning_rate": 1.7944716011466317e-05, "loss": 0.0058, "step": 66710 }, { "grad_norm": 0.03670163080096245, "learning_rate": 1.7930728055985945e-05, "loss": 0.0059, "step": 66720 }, { "grad_norm": 0.05863518267869949, "learning_rate": 1.7916744363224853e-05, "loss": 0.004, "step": 66730 }, { "grad_norm": 0.05518059432506561, "learning_rate": 1.79027649350418e-05, "loss": 0.0049, "step": 66740 }, { "grad_norm": 0.05197841301560402, "learning_rate": 1.7888789773294972e-05, "loss": 0.0053, "step": 66750 }, { "grad_norm": 0.04759608581662178, "learning_rate": 1.7874818879841948e-05, "loss": 0.0043, "step": 66760 }, { "grad_norm": 0.039315562695264816, "learning_rate": 1.7860852256539834e-05, "loss": 0.0054, "step": 66770 }, { "grad_norm": 0.05314944311976433, "learning_rate": 1.784688990524505e-05, "loss": 0.0033, "step": 66780 }, { "grad_norm": 0.032402485609054565, "learning_rate": 1.7832931827813514e-05, "loss": 0.0036, "step": 66790 }, { "grad_norm": 0.04964769259095192, "learning_rate": 1.7818978026100575e-05, "loss": 0.0044, "step": 66800 }, { "grad_norm": 0.07529346644878387, "learning_rate": 1.780502850196099e-05, "loss": 0.0078, "step": 66810 }, { "grad_norm": 0.02731354720890522, "learning_rate": 1.7791083257248975e-05, "loss": 0.004, "step": 66820 }, { "grad_norm": 0.07066137343645096, "learning_rate": 1.7777142293818156e-05, "loss": 0.0045, "step": 66830 }, { "grad_norm": 0.04196614772081375, "learning_rate": 1.7763205613521593e-05, "loss": 0.0062, "step": 66840 }, { "grad_norm": 0.046983979642391205, "learning_rate": 1.7749273218211793e-05, "loss": 0.0049, "step": 66850 }, { "grad_norm": 0.05254117399454117, "learning_rate": 1.773534510974067e-05, "loss": 0.0063, "step": 66860 }, { "grad_norm": 0.07285624742507935, "learning_rate": 1.7721421289959585e-05, "loss": 0.0089, "step": 66870 }, { "grad_norm": 0.0671527311205864, "learning_rate": 1.7707501760719313e-05, "loss": 0.0072, "step": 66880 }, { "grad_norm": 0.04890971630811691, "learning_rate": 1.7693586523870103e-05, "loss": 0.0051, "step": 66890 }, { "grad_norm": 0.03839029371738434, "learning_rate": 1.7679675581261555e-05, "loss": 0.0061, "step": 66900 }, { "grad_norm": 0.03330603614449501, "learning_rate": 1.7665768934742743e-05, "loss": 0.0071, "step": 66910 }, { "grad_norm": 0.03842267766594887, "learning_rate": 1.7651866586162202e-05, "loss": 0.005, "step": 66920 }, { "grad_norm": 0.06132609397172928, "learning_rate": 1.7637968537367855e-05, "loss": 0.0063, "step": 66930 }, { "grad_norm": 0.06272751092910767, "learning_rate": 1.7624074790207067e-05, "loss": 0.0048, "step": 66940 }, { "grad_norm": 0.07194368541240692, "learning_rate": 1.76101853465266e-05, "loss": 0.0061, "step": 66950 }, { "grad_norm": 0.06019449234008789, "learning_rate": 1.759630020817269e-05, "loss": 0.0066, "step": 66960 }, { "grad_norm": 0.07438939809799194, "learning_rate": 1.7582419376990965e-05, "loss": 0.0078, "step": 66970 }, { "grad_norm": 0.04021076858043671, "learning_rate": 1.7568542854826516e-05, "loss": 0.0056, "step": 66980 }, { "grad_norm": 0.0546422004699707, "learning_rate": 1.7554670643523823e-05, "loss": 0.0048, "step": 66990 }, { "grad_norm": 0.062944695353508, "learning_rate": 1.7540802744926827e-05, "loss": 0.0048, "step": 67000 }, { "grad_norm": 0.0751379132270813, "learning_rate": 1.7526939160878865e-05, "loss": 0.0067, "step": 67010 }, { "grad_norm": 0.06599177420139313, "learning_rate": 1.7513079893222727e-05, "loss": 0.0059, "step": 67020 }, { "grad_norm": 0.061775900423526764, "learning_rate": 1.749922494380062e-05, "loss": 0.0043, "step": 67030 }, { "grad_norm": 0.056616563349962234, "learning_rate": 1.748537431445416e-05, "loss": 0.0064, "step": 67040 }, { "grad_norm": 0.08062038570642471, "learning_rate": 1.7471528007024434e-05, "loss": 0.0053, "step": 67050 }, { "grad_norm": 0.07771238684654236, "learning_rate": 1.745768602335187e-05, "loss": 0.0067, "step": 67060 }, { "grad_norm": 0.04037763178348541, "learning_rate": 1.7443848365276428e-05, "loss": 0.0046, "step": 67070 }, { "grad_norm": 0.03660900890827179, "learning_rate": 1.7430015034637415e-05, "loss": 0.0052, "step": 67080 }, { "grad_norm": 0.058127567172050476, "learning_rate": 1.7416186033273595e-05, "loss": 0.0041, "step": 67090 }, { "grad_norm": 0.03799685835838318, "learning_rate": 1.7402361363023167e-05, "loss": 0.0074, "step": 67100 }, { "grad_norm": 0.05583292990922928, "learning_rate": 1.7388541025723703e-05, "loss": 0.0048, "step": 67110 }, { "grad_norm": 0.07745757699012756, "learning_rate": 1.7374725023212247e-05, "loss": 0.0059, "step": 67120 }, { "grad_norm": 0.04532614350318909, "learning_rate": 1.7360913357325237e-05, "loss": 0.0044, "step": 67130 }, { "grad_norm": 0.053270403295755386, "learning_rate": 1.734710602989859e-05, "loss": 0.0031, "step": 67140 }, { "grad_norm": 0.05830032005906105, "learning_rate": 1.7333303042767578e-05, "loss": 0.0048, "step": 67150 }, { "grad_norm": 0.060111649334430695, "learning_rate": 1.7319504397766957e-05, "loss": 0.0066, "step": 67160 }, { "grad_norm": 0.05880456045269966, "learning_rate": 1.7305710096730827e-05, "loss": 0.0046, "step": 67170 }, { "grad_norm": 0.04452425613999367, "learning_rate": 1.729192014149278e-05, "loss": 0.0044, "step": 67180 }, { "grad_norm": 0.06567259132862091, "learning_rate": 1.7278134533885808e-05, "loss": 0.0036, "step": 67190 }, { "grad_norm": 0.04122673720121384, "learning_rate": 1.7264353275742325e-05, "loss": 0.0066, "step": 67200 }, { "grad_norm": 0.028611790388822556, "learning_rate": 1.725057636889416e-05, "loss": 0.0032, "step": 67210 }, { "grad_norm": 0.05012064054608345, "learning_rate": 1.723680381517258e-05, "loss": 0.0056, "step": 67220 }, { "grad_norm": 0.035191215574741364, "learning_rate": 1.7223035616408258e-05, "loss": 0.0062, "step": 67230 }, { "grad_norm": 0.036321695894002914, "learning_rate": 1.720927177443129e-05, "loss": 0.0072, "step": 67240 }, { "grad_norm": 0.04880531504750252, "learning_rate": 1.7195512291071208e-05, "loss": 0.0076, "step": 67250 }, { "grad_norm": 0.06184229627251625, "learning_rate": 1.718175716815695e-05, "loss": 0.008, "step": 67260 }, { "grad_norm": 0.0652971863746643, "learning_rate": 1.716800640751688e-05, "loss": 0.0036, "step": 67270 }, { "grad_norm": 0.04003286361694336, "learning_rate": 1.7154260010978746e-05, "loss": 0.0077, "step": 67280 }, { "grad_norm": 0.07037733495235443, "learning_rate": 1.7140517980369797e-05, "loss": 0.0057, "step": 67290 }, { "grad_norm": 0.10923387110233307, "learning_rate": 1.712678031751663e-05, "loss": 0.007, "step": 67300 }, { "grad_norm": 0.05688732862472534, "learning_rate": 1.71130470242453e-05, "loss": 0.0037, "step": 67310 }, { "grad_norm": 0.04538830369710922, "learning_rate": 1.709931810238127e-05, "loss": 0.0069, "step": 67320 }, { "grad_norm": 0.03889656439423561, "learning_rate": 1.708559355374939e-05, "loss": 0.0046, "step": 67330 }, { "grad_norm": 0.0553462915122509, "learning_rate": 1.7071873380173975e-05, "loss": 0.0053, "step": 67340 }, { "grad_norm": 0.06719104945659637, "learning_rate": 1.705815758347874e-05, "loss": 0.004, "step": 67350 }, { "grad_norm": 0.05040353909134865, "learning_rate": 1.7044446165486804e-05, "loss": 0.0057, "step": 67360 }, { "grad_norm": 0.04859483614563942, "learning_rate": 1.7030739128020773e-05, "loss": 0.005, "step": 67370 }, { "grad_norm": 0.03080279752612114, "learning_rate": 1.701703647290256e-05, "loss": 0.0046, "step": 67380 }, { "grad_norm": 0.0812281146645546, "learning_rate": 1.7003338201953574e-05, "loss": 0.0101, "step": 67390 }, { "grad_norm": 0.04826796054840088, "learning_rate": 1.6989644316994618e-05, "loss": 0.0062, "step": 67400 }, { "grad_norm": 0.021111762151122093, "learning_rate": 1.6975954819845913e-05, "loss": 0.0055, "step": 67410 }, { "grad_norm": 0.07347052544355392, "learning_rate": 1.69622697123271e-05, "loss": 0.0047, "step": 67420 }, { "grad_norm": 0.06796500086784363, "learning_rate": 1.694858899625723e-05, "loss": 0.0052, "step": 67430 }, { "grad_norm": 0.043784402310848236, "learning_rate": 1.6934912673454777e-05, "loss": 0.0055, "step": 67440 }, { "grad_norm": 0.027994586154818535, "learning_rate": 1.692124074573763e-05, "loss": 0.004, "step": 67450 }, { "grad_norm": 0.05204547196626663, "learning_rate": 1.690757321492309e-05, "loss": 0.006, "step": 67460 }, { "grad_norm": 0.04929715022444725, "learning_rate": 1.689391008282788e-05, "loss": 0.0036, "step": 67470 }, { "grad_norm": 0.03768029808998108, "learning_rate": 1.6880251351268122e-05, "loss": 0.0065, "step": 67480 }, { "grad_norm": 0.038442812860012054, "learning_rate": 1.6866597022059398e-05, "loss": 0.007, "step": 67490 }, { "grad_norm": 0.049640048295259476, "learning_rate": 1.685294709701663e-05, "loss": 0.004, "step": 67500 }, { "grad_norm": 0.0541226826608181, "learning_rate": 1.683930157795419e-05, "loss": 0.0055, "step": 67510 }, { "grad_norm": 0.04164014011621475, "learning_rate": 1.6825660466685922e-05, "loss": 0.0061, "step": 67520 }, { "grad_norm": 0.08026048541069031, "learning_rate": 1.6812023765024997e-05, "loss": 0.0042, "step": 67530 }, { "grad_norm": 0.05477013811469078, "learning_rate": 1.6798391474784065e-05, "loss": 0.0045, "step": 67540 }, { "grad_norm": 0.052065685391426086, "learning_rate": 1.6784763597775123e-05, "loss": 0.0065, "step": 67550 }, { "grad_norm": 0.05742282420396805, "learning_rate": 1.6771140135809636e-05, "loss": 0.0063, "step": 67560 }, { "grad_norm": 0.06020806357264519, "learning_rate": 1.6757521090698463e-05, "loss": 0.0069, "step": 67570 }, { "grad_norm": 0.05828091502189636, "learning_rate": 1.6743906464251867e-05, "loss": 0.0051, "step": 67580 }, { "grad_norm": 0.049086686223745346, "learning_rate": 1.673029625827958e-05, "loss": 0.0062, "step": 67590 }, { "grad_norm": 0.05565018579363823, "learning_rate": 1.671669047459065e-05, "loss": 0.0042, "step": 67600 }, { "grad_norm": 0.05715664103627205, "learning_rate": 1.670308911499361e-05, "loss": 0.0051, "step": 67610 }, { "grad_norm": 0.04733588546514511, "learning_rate": 1.668949218129638e-05, "loss": 0.0045, "step": 67620 }, { "grad_norm": 0.04080100730061531, "learning_rate": 1.667589967530629e-05, "loss": 0.0039, "step": 67630 }, { "grad_norm": 0.05045752599835396, "learning_rate": 1.6662311598830093e-05, "loss": 0.0059, "step": 67640 }, { "grad_norm": 0.05562487989664078, "learning_rate": 1.6648727953673947e-05, "loss": 0.0061, "step": 67650 }, { "grad_norm": 0.05161752551794052, "learning_rate": 1.6635148741643415e-05, "loss": 0.0068, "step": 67660 }, { "grad_norm": 0.05928441509604454, "learning_rate": 1.662157396454348e-05, "loss": 0.0058, "step": 67670 }, { "grad_norm": 0.05515994876623154, "learning_rate": 1.660800362417853e-05, "loss": 0.0049, "step": 67680 }, { "grad_norm": 0.04450356960296631, "learning_rate": 1.6594437722352367e-05, "loss": 0.0046, "step": 67690 }, { "grad_norm": 0.03476749360561371, "learning_rate": 1.65808762608682e-05, "loss": 0.0046, "step": 67700 }, { "grad_norm": 0.06602528691291809, "learning_rate": 1.656731924152866e-05, "loss": 0.0073, "step": 67710 }, { "grad_norm": 0.04301108419895172, "learning_rate": 1.6553766666135757e-05, "loss": 0.0039, "step": 67720 }, { "grad_norm": 0.03989121690392494, "learning_rate": 1.6540218536490913e-05, "loss": 0.0073, "step": 67730 }, { "grad_norm": 0.05087096616625786, "learning_rate": 1.6526674854395023e-05, "loss": 0.0061, "step": 67740 }, { "grad_norm": 0.05648822337388992, "learning_rate": 1.6513135621648317e-05, "loss": 0.0048, "step": 67750 }, { "grad_norm": 0.042424675077199936, "learning_rate": 1.6499600840050482e-05, "loss": 0.0052, "step": 67760 }, { "grad_norm": 0.0435546338558197, "learning_rate": 1.6486070511400552e-05, "loss": 0.0081, "step": 67770 }, { "grad_norm": 0.05087330564856529, "learning_rate": 1.647254463749704e-05, "loss": 0.0052, "step": 67780 }, { "grad_norm": 0.06407801061868668, "learning_rate": 1.6459023220137814e-05, "loss": 0.0068, "step": 67790 }, { "grad_norm": 0.05981019511818886, "learning_rate": 1.644550626112017e-05, "loss": 0.0066, "step": 67800 }, { "grad_norm": 0.08027210831642151, "learning_rate": 1.6431993762240862e-05, "loss": 0.0049, "step": 67810 }, { "grad_norm": 0.04132973775267601, "learning_rate": 1.641848572529594e-05, "loss": 0.0044, "step": 67820 }, { "grad_norm": 0.027477964758872986, "learning_rate": 1.640498215208095e-05, "loss": 0.0035, "step": 67830 }, { "grad_norm": 0.06347782164812088, "learning_rate": 1.639148304439081e-05, "loss": 0.005, "step": 67840 }, { "grad_norm": 0.0356961227953434, "learning_rate": 1.6377988404019857e-05, "loss": 0.0043, "step": 67850 }, { "grad_norm": 0.05224838852882385, "learning_rate": 1.636449823276183e-05, "loss": 0.006, "step": 67860 }, { "grad_norm": 0.055355995893478394, "learning_rate": 1.635101253240988e-05, "loss": 0.0048, "step": 67870 }, { "grad_norm": 0.05092830955982208, "learning_rate": 1.633753130475652e-05, "loss": 0.0101, "step": 67880 }, { "grad_norm": 0.03733619302511215, "learning_rate": 1.6324054551593743e-05, "loss": 0.0055, "step": 67890 }, { "grad_norm": 0.04439973086118698, "learning_rate": 1.63105822747129e-05, "loss": 0.0059, "step": 67900 }, { "grad_norm": 0.06125544756650925, "learning_rate": 1.6297114475904752e-05, "loss": 0.0049, "step": 67910 }, { "grad_norm": 0.048041921108961105, "learning_rate": 1.6283651156959468e-05, "loss": 0.0054, "step": 67920 }, { "grad_norm": 0.05882396176457405, "learning_rate": 1.6270192319666643e-05, "loss": 0.0059, "step": 67930 }, { "grad_norm": 0.036345578730106354, "learning_rate": 1.625673796581522e-05, "loss": 0.0053, "step": 67940 }, { "grad_norm": 0.04186469316482544, "learning_rate": 1.6243288097193586e-05, "loss": 0.0069, "step": 67950 }, { "grad_norm": 0.04234343022108078, "learning_rate": 1.6229842715589554e-05, "loss": 0.0048, "step": 67960 }, { "grad_norm": 0.04139634221792221, "learning_rate": 1.62164018227903e-05, "loss": 0.0035, "step": 67970 }, { "grad_norm": 0.043829623609781265, "learning_rate": 1.6202965420582438e-05, "loss": 0.0059, "step": 67980 }, { "grad_norm": 0.03784389793872833, "learning_rate": 1.618953351075193e-05, "loss": 0.0064, "step": 67990 }, { "grad_norm": 0.026179468259215355, "learning_rate": 1.6176106095084187e-05, "loss": 0.0048, "step": 68000 }, { "grad_norm": 0.061493489891290665, "learning_rate": 1.6162683175364013e-05, "loss": 0.0051, "step": 68010 }, { "grad_norm": 0.07169006764888763, "learning_rate": 1.6149264753375616e-05, "loss": 0.0051, "step": 68020 }, { "grad_norm": 0.04127248749136925, "learning_rate": 1.6135850830902603e-05, "loss": 0.0034, "step": 68030 }, { "grad_norm": 0.03873637318611145, "learning_rate": 1.612244140972798e-05, "loss": 0.0051, "step": 68040 }, { "grad_norm": 0.07511721551418304, "learning_rate": 1.6109036491634155e-05, "loss": 0.0065, "step": 68050 }, { "grad_norm": 0.08137097954750061, "learning_rate": 1.6095636078402942e-05, "loss": 0.0064, "step": 68060 }, { "grad_norm": 0.10957256704568863, "learning_rate": 1.6082240171815554e-05, "loss": 0.0058, "step": 68070 }, { "grad_norm": 0.05500634387135506, "learning_rate": 1.6068848773652605e-05, "loss": 0.0058, "step": 68080 }, { "grad_norm": 0.06646101921796799, "learning_rate": 1.605546188569413e-05, "loss": 0.0083, "step": 68090 }, { "grad_norm": 0.06608156859874725, "learning_rate": 1.6042079509719484e-05, "loss": 0.0056, "step": 68100 }, { "grad_norm": 0.03938865661621094, "learning_rate": 1.6028701647507544e-05, "loss": 0.0065, "step": 68110 }, { "grad_norm": 0.10115008056163788, "learning_rate": 1.6015328300836503e-05, "loss": 0.0048, "step": 68120 }, { "grad_norm": 0.07825599610805511, "learning_rate": 1.6001959471483968e-05, "loss": 0.0046, "step": 68130 }, { "grad_norm": 0.0373404286801815, "learning_rate": 1.5988595161226986e-05, "loss": 0.0036, "step": 68140 }, { "grad_norm": 0.038156721740961075, "learning_rate": 1.597523537184193e-05, "loss": 0.0043, "step": 68150 }, { "grad_norm": 0.06696882098913193, "learning_rate": 1.596188010510463e-05, "loss": 0.0056, "step": 68160 }, { "grad_norm": 0.040918201208114624, "learning_rate": 1.5948529362790284e-05, "loss": 0.0091, "step": 68170 }, { "grad_norm": 0.03865089267492294, "learning_rate": 1.5935183146673526e-05, "loss": 0.0036, "step": 68180 }, { "grad_norm": 0.05096305161714554, "learning_rate": 1.5921841458528373e-05, "loss": 0.004, "step": 68190 }, { "grad_norm": 0.06725315749645233, "learning_rate": 1.59085043001282e-05, "loss": 0.0062, "step": 68200 }, { "grad_norm": 0.047154683619737625, "learning_rate": 1.589517167324583e-05, "loss": 0.006, "step": 68210 }, { "grad_norm": 0.036743663251399994, "learning_rate": 1.5881843579653454e-05, "loss": 0.0036, "step": 68220 }, { "grad_norm": 0.024742230772972107, "learning_rate": 1.5868520021122685e-05, "loss": 0.0057, "step": 68230 }, { "grad_norm": 0.029744522646069527, "learning_rate": 1.585520099942452e-05, "loss": 0.0044, "step": 68240 }, { "grad_norm": 0.05693335086107254, "learning_rate": 1.584188651632934e-05, "loss": 0.0044, "step": 68250 }, { "grad_norm": 0.075560063123703, "learning_rate": 1.5828576573606945e-05, "loss": 0.0054, "step": 68260 }, { "grad_norm": 0.06318759173154831, "learning_rate": 1.5815271173026525e-05, "loss": 0.0052, "step": 68270 }, { "grad_norm": 0.09167025983333588, "learning_rate": 1.580197031635666e-05, "loss": 0.0091, "step": 68280 }, { "grad_norm": 0.05682803690433502, "learning_rate": 1.5788674005365323e-05, "loss": 0.0053, "step": 68290 }, { "grad_norm": 0.08890408277511597, "learning_rate": 1.5775382241819896e-05, "loss": 0.0046, "step": 68300 }, { "grad_norm": 0.03678121417760849, "learning_rate": 1.5762095027487174e-05, "loss": 0.0053, "step": 68310 }, { "grad_norm": 0.05386961251497269, "learning_rate": 1.5748812364133257e-05, "loss": 0.0046, "step": 68320 }, { "grad_norm": 0.05849495530128479, "learning_rate": 1.573553425352377e-05, "loss": 0.0045, "step": 68330 }, { "grad_norm": 0.06930658221244812, "learning_rate": 1.5722260697423646e-05, "loss": 0.0054, "step": 68340 }, { "grad_norm": 0.03086763806641102, "learning_rate": 1.570899169759724e-05, "loss": 0.0053, "step": 68350 }, { "grad_norm": 0.06823495030403137, "learning_rate": 1.5695727255808307e-05, "loss": 0.0041, "step": 68360 }, { "grad_norm": 0.06512080132961273, "learning_rate": 1.568246737381996e-05, "loss": 0.005, "step": 68370 }, { "grad_norm": 0.03893674537539482, "learning_rate": 1.5669212053394743e-05, "loss": 0.0046, "step": 68380 }, { "grad_norm": 0.030587738379836082, "learning_rate": 1.5655961296294592e-05, "loss": 0.0057, "step": 68390 }, { "grad_norm": 0.030360147356987, "learning_rate": 1.5642715104280797e-05, "loss": 0.008, "step": 68400 }, { "grad_norm": 0.07229603826999664, "learning_rate": 1.5629473479114127e-05, "loss": 0.0067, "step": 68410 }, { "grad_norm": 0.055848512798547745, "learning_rate": 1.5616236422554647e-05, "loss": 0.0041, "step": 68420 }, { "grad_norm": 0.049037687480449677, "learning_rate": 1.560300393636186e-05, "loss": 0.0049, "step": 68430 }, { "grad_norm": 0.051102880388498306, "learning_rate": 1.5589776022294665e-05, "loss": 0.0034, "step": 68440 }, { "grad_norm": 0.04238881543278694, "learning_rate": 1.5576552682111344e-05, "loss": 0.0043, "step": 68450 }, { "grad_norm": 0.04088970646262169, "learning_rate": 1.5563333917569573e-05, "loss": 0.007, "step": 68460 }, { "grad_norm": 0.04843713715672493, "learning_rate": 1.555011973042642e-05, "loss": 0.0046, "step": 68470 }, { "grad_norm": 0.04746459797024727, "learning_rate": 1.5536910122438346e-05, "loss": 0.0043, "step": 68480 }, { "grad_norm": 0.043322931975126266, "learning_rate": 1.55237050953612e-05, "loss": 0.0048, "step": 68490 }, { "grad_norm": 0.06655670702457428, "learning_rate": 1.5510504650950225e-05, "loss": 0.0041, "step": 68500 }, { "grad_norm": 0.031497593969106674, "learning_rate": 1.5497308790960047e-05, "loss": 0.0049, "step": 68510 }, { "grad_norm": 0.03633252903819084, "learning_rate": 1.5484117517144697e-05, "loss": 0.0063, "step": 68520 }, { "grad_norm": 0.04389514401555061, "learning_rate": 1.5470930831257602e-05, "loss": 0.005, "step": 68530 }, { "grad_norm": 0.04144848883152008, "learning_rate": 1.545774873505153e-05, "loss": 0.0068, "step": 68540 }, { "grad_norm": 0.045956335961818695, "learning_rate": 1.5444571230278682e-05, "loss": 0.0049, "step": 68550 }, { "grad_norm": 0.054849911481142044, "learning_rate": 1.5431398318690672e-05, "loss": 0.0066, "step": 68560 }, { "grad_norm": 0.06906777620315552, "learning_rate": 1.5418230002038452e-05, "loss": 0.0055, "step": 68570 }, { "grad_norm": 0.05883195996284485, "learning_rate": 1.54050662820724e-05, "loss": 0.0046, "step": 68580 }, { "grad_norm": 0.07077249884605408, "learning_rate": 1.539190716054224e-05, "loss": 0.0072, "step": 68590 }, { "grad_norm": 0.06355012208223343, "learning_rate": 1.5378752639197125e-05, "loss": 0.0049, "step": 68600 }, { "grad_norm": 0.045821454375982285, "learning_rate": 1.536560271978559e-05, "loss": 0.0065, "step": 68610 }, { "grad_norm": 0.05032436177134514, "learning_rate": 1.535245740405552e-05, "loss": 0.0091, "step": 68620 }, { "grad_norm": 0.05471422150731087, "learning_rate": 1.533931669375428e-05, "loss": 0.0059, "step": 68630 }, { "grad_norm": 0.06509029120206833, "learning_rate": 1.5326180590628513e-05, "loss": 0.0065, "step": 68640 }, { "grad_norm": 0.08477886021137238, "learning_rate": 1.531304909642432e-05, "loss": 0.0075, "step": 68650 }, { "grad_norm": 0.06911418586969376, "learning_rate": 1.5299922212887158e-05, "loss": 0.0062, "step": 68660 }, { "grad_norm": 0.0676775649189949, "learning_rate": 1.528679994176188e-05, "loss": 0.0079, "step": 68670 }, { "grad_norm": 0.044449593871831894, "learning_rate": 1.5273682284792745e-05, "loss": 0.0058, "step": 68680 }, { "grad_norm": 0.0596536360681057, "learning_rate": 1.526056924372336e-05, "loss": 0.0046, "step": 68690 }, { "grad_norm": 0.05039506033062935, "learning_rate": 1.5247460820296755e-05, "loss": 0.01, "step": 68700 }, { "grad_norm": 0.029610885307192802, "learning_rate": 1.523435701625533e-05, "loss": 0.005, "step": 68710 }, { "grad_norm": 0.05700180307030678, "learning_rate": 1.5221257833340863e-05, "loss": 0.0053, "step": 68720 }, { "grad_norm": 0.04426533728837967, "learning_rate": 1.5208163273294535e-05, "loss": 0.004, "step": 68730 }, { "grad_norm": 0.05073579028248787, "learning_rate": 1.5195073337856896e-05, "loss": 0.0061, "step": 68740 }, { "grad_norm": 0.037436388432979584, "learning_rate": 1.5181988028767912e-05, "loss": 0.0038, "step": 68750 }, { "grad_norm": 0.06346946209669113, "learning_rate": 1.5168907347766875e-05, "loss": 0.0086, "step": 68760 }, { "grad_norm": 0.08478993922472, "learning_rate": 1.5155831296592504e-05, "loss": 0.0063, "step": 68770 }, { "grad_norm": 0.021032745018601418, "learning_rate": 1.5142759876982926e-05, "loss": 0.0043, "step": 68780 }, { "grad_norm": 0.06448666006326675, "learning_rate": 1.5129693090675612e-05, "loss": 0.0055, "step": 68790 }, { "grad_norm": 0.04574082791805267, "learning_rate": 1.5116630939407434e-05, "loss": 0.0045, "step": 68800 }, { "grad_norm": 0.09795603156089783, "learning_rate": 1.5103573424914618e-05, "loss": 0.0069, "step": 68810 }, { "grad_norm": 0.05473718047142029, "learning_rate": 1.5090520548932807e-05, "loss": 0.0066, "step": 68820 }, { "grad_norm": 0.06407332420349121, "learning_rate": 1.5077472313197033e-05, "loss": 0.0036, "step": 68830 }, { "grad_norm": 0.045551422983407974, "learning_rate": 1.5064428719441665e-05, "loss": 0.0041, "step": 68840 }, { "grad_norm": 0.03977459669113159, "learning_rate": 1.5051389769400542e-05, "loss": 0.0038, "step": 68850 }, { "grad_norm": 0.026629768311977386, "learning_rate": 1.503835546480678e-05, "loss": 0.0044, "step": 68860 }, { "grad_norm": 0.054060861468315125, "learning_rate": 1.5025325807392949e-05, "loss": 0.0042, "step": 68870 }, { "grad_norm": 0.06304468959569931, "learning_rate": 1.5012300798890971e-05, "loss": 0.0073, "step": 68880 }, { "grad_norm": 0.04549575597047806, "learning_rate": 1.4999280441032165e-05, "loss": 0.0051, "step": 68890 }, { "grad_norm": 0.05655055120587349, "learning_rate": 1.4986264735547229e-05, "loss": 0.0043, "step": 68900 }, { "grad_norm": 0.037268124520778656, "learning_rate": 1.497325368416625e-05, "loss": 0.0042, "step": 68910 }, { "grad_norm": 0.061693549156188965, "learning_rate": 1.496024728861864e-05, "loss": 0.0045, "step": 68920 }, { "grad_norm": 0.0391877144575119, "learning_rate": 1.4947245550633288e-05, "loss": 0.0038, "step": 68930 }, { "grad_norm": 0.06362495571374893, "learning_rate": 1.4934248471938401e-05, "loss": 0.0054, "step": 68940 }, { "grad_norm": 0.05289130285382271, "learning_rate": 1.4921256054261568e-05, "loss": 0.0054, "step": 68950 }, { "grad_norm": 0.033441510051488876, "learning_rate": 1.490826829932978e-05, "loss": 0.0044, "step": 68960 }, { "grad_norm": 0.02726912312209606, "learning_rate": 1.4895285208869414e-05, "loss": 0.005, "step": 68970 }, { "grad_norm": 0.057852331548929214, "learning_rate": 1.488230678460617e-05, "loss": 0.0053, "step": 68980 }, { "grad_norm": 0.04390237480401993, "learning_rate": 1.4869333028265187e-05, "loss": 0.0073, "step": 68990 }, { "grad_norm": 0.08995168656110764, "learning_rate": 1.4856363941570978e-05, "loss": 0.0063, "step": 69000 }, { "grad_norm": 0.04646981880068779, "learning_rate": 1.484339952624742e-05, "loss": 0.0051, "step": 69010 }, { "grad_norm": 0.059034597128629684, "learning_rate": 1.4830439784017785e-05, "loss": 0.0041, "step": 69020 }, { "grad_norm": 0.0494370236992836, "learning_rate": 1.4817484716604679e-05, "loss": 0.0047, "step": 69030 }, { "grad_norm": 0.026505738496780396, "learning_rate": 1.480453432573013e-05, "loss": 0.0036, "step": 69040 }, { "grad_norm": 0.07597063481807709, "learning_rate": 1.4791588613115537e-05, "loss": 0.006, "step": 69050 }, { "grad_norm": 0.06296717375516891, "learning_rate": 1.4778647580481675e-05, "loss": 0.0067, "step": 69060 }, { "grad_norm": 0.06265432387590408, "learning_rate": 1.476571122954869e-05, "loss": 0.0097, "step": 69070 }, { "grad_norm": 0.054443359375, "learning_rate": 1.4752779562036111e-05, "loss": 0.0042, "step": 69080 }, { "grad_norm": 0.05683179572224617, "learning_rate": 1.4739852579662855e-05, "loss": 0.0072, "step": 69090 }, { "grad_norm": 0.07459606975317001, "learning_rate": 1.4726930284147189e-05, "loss": 0.0072, "step": 69100 }, { "grad_norm": 0.06607648730278015, "learning_rate": 1.4714012677206785e-05, "loss": 0.0047, "step": 69110 }, { "grad_norm": 0.09153879433870316, "learning_rate": 1.4701099760558674e-05, "loss": 0.0082, "step": 69120 }, { "grad_norm": 0.04253171384334564, "learning_rate": 1.4688191535919287e-05, "loss": 0.0043, "step": 69130 }, { "grad_norm": 0.04164274036884308, "learning_rate": 1.4675288005004373e-05, "loss": 0.0052, "step": 69140 }, { "grad_norm": 0.0423600859940052, "learning_rate": 1.4662389169529133e-05, "loss": 0.0058, "step": 69150 }, { "grad_norm": 0.03641931340098381, "learning_rate": 1.4649495031208105e-05, "loss": 0.0048, "step": 69160 }, { "grad_norm": 0.057549525052309036, "learning_rate": 1.4636605591755203e-05, "loss": 0.0069, "step": 69170 }, { "grad_norm": 0.07395163178443909, "learning_rate": 1.4623720852883732e-05, "loss": 0.0089, "step": 69180 }, { "grad_norm": 0.040483541786670685, "learning_rate": 1.4610840816306331e-05, "loss": 0.0042, "step": 69190 }, { "grad_norm": 0.0613168366253376, "learning_rate": 1.459796548373506e-05, "loss": 0.0085, "step": 69200 }, { "grad_norm": 0.05362778902053833, "learning_rate": 1.4585094856881337e-05, "loss": 0.0056, "step": 69210 }, { "grad_norm": 0.05743512511253357, "learning_rate": 1.457222893745594e-05, "loss": 0.0043, "step": 69220 }, { "grad_norm": 0.062272924929857254, "learning_rate": 1.4559367727169082e-05, "loss": 0.0048, "step": 69230 }, { "grad_norm": 0.04761142283678055, "learning_rate": 1.4546511227730253e-05, "loss": 0.0063, "step": 69240 }, { "grad_norm": 0.03629567101597786, "learning_rate": 1.4533659440848385e-05, "loss": 0.0032, "step": 69250 }, { "grad_norm": 0.03679240494966507, "learning_rate": 1.4520812368231773e-05, "loss": 0.0059, "step": 69260 }, { "grad_norm": 0.04988124594092369, "learning_rate": 1.4507970011588068e-05, "loss": 0.0044, "step": 69270 }, { "grad_norm": 0.06859276443719864, "learning_rate": 1.4495132372624315e-05, "loss": 0.0066, "step": 69280 }, { "grad_norm": 0.041571926325559616, "learning_rate": 1.4482299453046911e-05, "loss": 0.0067, "step": 69290 }, { "grad_norm": 0.047110751271247864, "learning_rate": 1.4469471254561646e-05, "loss": 0.0048, "step": 69300 }, { "grad_norm": 0.0598016157746315, "learning_rate": 1.4456647778873666e-05, "loss": 0.0048, "step": 69310 }, { "grad_norm": 0.061711691319942474, "learning_rate": 1.44438290276875e-05, "loss": 0.0045, "step": 69320 }, { "grad_norm": 0.04381881281733513, "learning_rate": 1.4431015002707043e-05, "loss": 0.0041, "step": 69330 }, { "grad_norm": 0.04876139760017395, "learning_rate": 1.4418205705635568e-05, "loss": 0.0057, "step": 69340 }, { "grad_norm": 0.07422176748514175, "learning_rate": 1.4405401138175723e-05, "loss": 0.0058, "step": 69350 }, { "grad_norm": 0.07018113136291504, "learning_rate": 1.4392601302029485e-05, "loss": 0.0047, "step": 69360 }, { "grad_norm": 0.029957471415400505, "learning_rate": 1.4379806198898266e-05, "loss": 0.0047, "step": 69370 }, { "grad_norm": 0.050689712166786194, "learning_rate": 1.4367015830482822e-05, "loss": 0.0048, "step": 69380 }, { "grad_norm": 0.037981271743774414, "learning_rate": 1.4354230198483265e-05, "loss": 0.0048, "step": 69390 }, { "grad_norm": 0.04957934468984604, "learning_rate": 1.4341449304599115e-05, "loss": 0.0047, "step": 69400 }, { "grad_norm": 0.04127772897481918, "learning_rate": 1.4328673150529198e-05, "loss": 0.0036, "step": 69410 }, { "grad_norm": 0.050837788730859756, "learning_rate": 1.4315901737971766e-05, "loss": 0.0048, "step": 69420 }, { "grad_norm": 0.04257180541753769, "learning_rate": 1.4303135068624429e-05, "loss": 0.0048, "step": 69430 }, { "grad_norm": 0.044984593987464905, "learning_rate": 1.4290373144184138e-05, "loss": 0.0047, "step": 69440 }, { "grad_norm": 0.03999800607562065, "learning_rate": 1.4277615966347285e-05, "loss": 0.0045, "step": 69450 }, { "grad_norm": 0.048238735646009445, "learning_rate": 1.4264863536809541e-05, "loss": 0.0053, "step": 69460 }, { "grad_norm": 0.05417922884225845, "learning_rate": 1.4252115857266001e-05, "loss": 0.0081, "step": 69470 }, { "grad_norm": 0.057893577963113785, "learning_rate": 1.4239372929411116e-05, "loss": 0.0045, "step": 69480 }, { "grad_norm": 0.062330227345228195, "learning_rate": 1.42266347549387e-05, "loss": 0.006, "step": 69490 }, { "grad_norm": 0.10363699495792389, "learning_rate": 1.421390133554194e-05, "loss": 0.0072, "step": 69500 }, { "grad_norm": 0.06521676480770111, "learning_rate": 1.4201172672913393e-05, "loss": 0.0062, "step": 69510 }, { "grad_norm": 0.042346853762865067, "learning_rate": 1.4188448768744983e-05, "loss": 0.0085, "step": 69520 }, { "grad_norm": 0.04079548642039299, "learning_rate": 1.4175729624727996e-05, "loss": 0.0046, "step": 69530 }, { "grad_norm": 0.038909394294023514, "learning_rate": 1.4163015242553091e-05, "loss": 0.0051, "step": 69540 }, { "grad_norm": 0.03294113650918007, "learning_rate": 1.4150305623910292e-05, "loss": 0.0064, "step": 69550 }, { "grad_norm": 0.018848787993192673, "learning_rate": 1.413760077048899e-05, "loss": 0.0038, "step": 69560 }, { "grad_norm": 0.06882932782173157, "learning_rate": 1.4124900683977954e-05, "loss": 0.0057, "step": 69570 }, { "grad_norm": 0.04192287474870682, "learning_rate": 1.4112205366065284e-05, "loss": 0.006, "step": 69580 }, { "grad_norm": 0.03336401656270027, "learning_rate": 1.4099514818438465e-05, "loss": 0.0043, "step": 69590 }, { "grad_norm": 0.03805458918213844, "learning_rate": 1.4086829042784388e-05, "loss": 0.0047, "step": 69600 }, { "grad_norm": 0.03810331970453262, "learning_rate": 1.4074148040789254e-05, "loss": 0.0062, "step": 69610 }, { "grad_norm": 0.024654781445860863, "learning_rate": 1.4061471814138672e-05, "loss": 0.0057, "step": 69620 }, { "grad_norm": 0.044685546308755875, "learning_rate": 1.4048800364517556e-05, "loss": 0.0056, "step": 69630 }, { "grad_norm": 0.03803635016083717, "learning_rate": 1.4036133693610244e-05, "loss": 0.0052, "step": 69640 }, { "grad_norm": 0.050122883170843124, "learning_rate": 1.4023471803100419e-05, "loss": 0.0056, "step": 69650 }, { "grad_norm": 0.062437281012535095, "learning_rate": 1.4010814694671103e-05, "loss": 0.0053, "step": 69660 }, { "grad_norm": 0.046061400324106216, "learning_rate": 1.399816237000477e-05, "loss": 0.0054, "step": 69670 }, { "grad_norm": 0.025047417730093002, "learning_rate": 1.398551483078313e-05, "loss": 0.0037, "step": 69680 }, { "grad_norm": 0.03529828414320946, "learning_rate": 1.3972872078687355e-05, "loss": 0.0079, "step": 69690 }, { "grad_norm": 0.07235989719629288, "learning_rate": 1.3960234115397936e-05, "loss": 0.0046, "step": 69700 }, { "grad_norm": 0.07672321796417236, "learning_rate": 1.3947600942594747e-05, "loss": 0.0073, "step": 69710 }, { "grad_norm": 0.026925165206193924, "learning_rate": 1.393497256195701e-05, "loss": 0.0066, "step": 69720 }, { "grad_norm": 0.06700332462787628, "learning_rate": 1.3922348975163325e-05, "loss": 0.005, "step": 69730 }, { "grad_norm": 0.04364544153213501, "learning_rate": 1.390973018389164e-05, "loss": 0.0039, "step": 69740 }, { "grad_norm": 0.05064399540424347, "learning_rate": 1.3897116189819282e-05, "loss": 0.004, "step": 69750 }, { "grad_norm": 0.06913779675960541, "learning_rate": 1.388450699462292e-05, "loss": 0.0061, "step": 69760 }, { "grad_norm": 0.039196498692035675, "learning_rate": 1.3871902599978604e-05, "loss": 0.0059, "step": 69770 }, { "grad_norm": 0.0456833653151989, "learning_rate": 1.3859303007561742e-05, "loss": 0.0054, "step": 69780 }, { "grad_norm": 0.021633384749293327, "learning_rate": 1.3846708219047105e-05, "loss": 0.0052, "step": 69790 }, { "grad_norm": 0.042459942400455475, "learning_rate": 1.3834118236108795e-05, "loss": 0.0047, "step": 69800 }, { "grad_norm": 0.03353472799062729, "learning_rate": 1.38215330604203e-05, "loss": 0.0034, "step": 69810 }, { "grad_norm": 0.04324200749397278, "learning_rate": 1.3808952693654503e-05, "loss": 0.0071, "step": 69820 }, { "grad_norm": 0.054902467876672745, "learning_rate": 1.3796377137483596e-05, "loss": 0.0037, "step": 69830 }, { "grad_norm": 0.05438758805394173, "learning_rate": 1.3783806393579168e-05, "loss": 0.0063, "step": 69840 }, { "grad_norm": 0.050971660763025284, "learning_rate": 1.3771240463612117e-05, "loss": 0.0046, "step": 69850 }, { "grad_norm": 0.04459832236170769, "learning_rate": 1.3758679349252752e-05, "loss": 0.0057, "step": 69860 }, { "grad_norm": 0.08350330591201782, "learning_rate": 1.3746123052170723e-05, "loss": 0.0076, "step": 69870 }, { "grad_norm": 0.04782082885503769, "learning_rate": 1.3733571574035025e-05, "loss": 0.0046, "step": 69880 }, { "grad_norm": 0.04127809405326843, "learning_rate": 1.3721024916514074e-05, "loss": 0.0051, "step": 69890 }, { "grad_norm": 0.053062692284584045, "learning_rate": 1.3708483081275559e-05, "loss": 0.0049, "step": 69900 }, { "grad_norm": 0.06184406951069832, "learning_rate": 1.3695946069986581e-05, "loss": 0.0052, "step": 69910 }, { "grad_norm": 0.04460417479276657, "learning_rate": 1.3683413884313584e-05, "loss": 0.0048, "step": 69920 }, { "grad_norm": 0.04260364919900894, "learning_rate": 1.3670886525922377e-05, "loss": 0.0079, "step": 69930 }, { "grad_norm": 0.045571912080049515, "learning_rate": 1.3658363996478124e-05, "loss": 0.0068, "step": 69940 }, { "grad_norm": 0.042757418006658554, "learning_rate": 1.3645846297645365e-05, "loss": 0.0042, "step": 69950 }, { "grad_norm": 0.03353376314043999, "learning_rate": 1.3633333431087936e-05, "loss": 0.0039, "step": 69960 }, { "grad_norm": 0.05461700260639191, "learning_rate": 1.3620825398469117e-05, "loss": 0.0045, "step": 69970 }, { "grad_norm": 0.05379272997379303, "learning_rate": 1.3608322201451484e-05, "loss": 0.0058, "step": 69980 }, { "grad_norm": 0.039112403988838196, "learning_rate": 1.3595823841697003e-05, "loss": 0.0051, "step": 69990 }, { "grad_norm": 0.06458666920661926, "learning_rate": 1.3583330320866988e-05, "loss": 0.0067, "step": 70000 }, { "grad_norm": 0.09304322302341461, "learning_rate": 1.3570841640622078e-05, "loss": 0.0056, "step": 70010 }, { "grad_norm": 0.04438828304409981, "learning_rate": 1.3558357802622313e-05, "loss": 0.0053, "step": 70020 }, { "grad_norm": 0.07475867122411728, "learning_rate": 1.3545878808527056e-05, "loss": 0.0061, "step": 70030 }, { "grad_norm": 0.0692673847079277, "learning_rate": 1.3533404659995075e-05, "loss": 0.0049, "step": 70040 }, { "grad_norm": 0.06737036257982254, "learning_rate": 1.3520935358684445e-05, "loss": 0.008, "step": 70050 }, { "grad_norm": 0.05779752507805824, "learning_rate": 1.350847090625263e-05, "loss": 0.0049, "step": 70060 }, { "grad_norm": 0.03170088678598404, "learning_rate": 1.3496011304356405e-05, "loss": 0.005, "step": 70070 }, { "grad_norm": 0.0328601710498333, "learning_rate": 1.3483556554651933e-05, "loss": 0.0042, "step": 70080 }, { "grad_norm": 0.032735224813222885, "learning_rate": 1.3471106658794735e-05, "loss": 0.0068, "step": 70090 }, { "grad_norm": 0.0686759278178215, "learning_rate": 1.3458661618439684e-05, "loss": 0.0059, "step": 70100 }, { "grad_norm": 0.0477946400642395, "learning_rate": 1.3446221435240991e-05, "loss": 0.0037, "step": 70110 }, { "grad_norm": 0.04479878768324852, "learning_rate": 1.3433786110852242e-05, "loss": 0.0049, "step": 70120 }, { "grad_norm": 0.0349620059132576, "learning_rate": 1.3421355646926365e-05, "loss": 0.0057, "step": 70130 }, { "grad_norm": 0.03788212686777115, "learning_rate": 1.3408930045115641e-05, "loss": 0.0045, "step": 70140 }, { "grad_norm": 0.051324520260095596, "learning_rate": 1.3396509307071714e-05, "loss": 0.0063, "step": 70150 }, { "grad_norm": 0.038895148783922195, "learning_rate": 1.3384093434445577e-05, "loss": 0.0044, "step": 70160 }, { "grad_norm": 0.03708236292004585, "learning_rate": 1.3371682428887583e-05, "loss": 0.0037, "step": 70170 }, { "grad_norm": 0.030307991430163383, "learning_rate": 1.3359276292047396e-05, "loss": 0.0049, "step": 70180 }, { "grad_norm": 0.041084591299295425, "learning_rate": 1.3346875025574107e-05, "loss": 0.0048, "step": 70190 }, { "grad_norm": 0.07959337532520294, "learning_rate": 1.3334478631116105e-05, "loss": 0.0075, "step": 70200 }, { "grad_norm": 0.0453668050467968, "learning_rate": 1.332208711032114e-05, "loss": 0.0043, "step": 70210 }, { "grad_norm": 0.041294265538454056, "learning_rate": 1.3309700464836349e-05, "loss": 0.0046, "step": 70220 }, { "grad_norm": 0.051133833825588226, "learning_rate": 1.3297318696308159e-05, "loss": 0.0049, "step": 70230 }, { "grad_norm": 0.03084128350019455, "learning_rate": 1.3284941806382389e-05, "loss": 0.0055, "step": 70240 }, { "grad_norm": 0.029172947630286217, "learning_rate": 1.3272569796704209e-05, "loss": 0.0034, "step": 70250 }, { "grad_norm": 0.06619424372911453, "learning_rate": 1.326020266891812e-05, "loss": 0.0048, "step": 70260 }, { "grad_norm": 0.03342046961188316, "learning_rate": 1.3247840424668035e-05, "loss": 0.0039, "step": 70270 }, { "grad_norm": 0.08596254140138626, "learning_rate": 1.3235483065597115e-05, "loss": 0.0045, "step": 70280 }, { "grad_norm": 0.064583919942379, "learning_rate": 1.3223130593347954e-05, "loss": 0.0049, "step": 70290 }, { "grad_norm": 0.05450188368558884, "learning_rate": 1.3210783009562466e-05, "loss": 0.0041, "step": 70300 }, { "grad_norm": 0.056120406836271286, "learning_rate": 1.3198440315881927e-05, "loss": 0.007, "step": 70310 }, { "grad_norm": 0.06401824951171875, "learning_rate": 1.318610251394694e-05, "loss": 0.0065, "step": 70320 }, { "grad_norm": 0.043883904814720154, "learning_rate": 1.3173769605397484e-05, "loss": 0.0037, "step": 70330 }, { "grad_norm": 0.03817017748951912, "learning_rate": 1.316144159187288e-05, "loss": 0.0037, "step": 70340 }, { "grad_norm": 0.025924600660800934, "learning_rate": 1.3149118475011784e-05, "loss": 0.0059, "step": 70350 }, { "grad_norm": 0.04416351020336151, "learning_rate": 1.313680025645222e-05, "loss": 0.0042, "step": 70360 }, { "grad_norm": 0.0405660942196846, "learning_rate": 1.3124486937831553e-05, "loss": 0.0039, "step": 70370 }, { "grad_norm": 0.0527043417096138, "learning_rate": 1.3112178520786494e-05, "loss": 0.0042, "step": 70380 }, { "grad_norm": 0.08165755122900009, "learning_rate": 1.3099875006953122e-05, "loss": 0.0061, "step": 70390 }, { "grad_norm": 0.032913632690906525, "learning_rate": 1.30875763979668e-05, "loss": 0.0045, "step": 70400 }, { "grad_norm": 0.06328361481428146, "learning_rate": 1.3075282695462331e-05, "loss": 0.0064, "step": 70410 }, { "grad_norm": 0.06269118934869766, "learning_rate": 1.3062993901073811e-05, "loss": 0.0068, "step": 70420 }, { "grad_norm": 0.02706998772919178, "learning_rate": 1.3050710016434692e-05, "loss": 0.0045, "step": 70430 }, { "grad_norm": 0.055998384952545166, "learning_rate": 1.3038431043177796e-05, "loss": 0.0044, "step": 70440 }, { "grad_norm": 0.06565053761005402, "learning_rate": 1.3026156982935228e-05, "loss": 0.0046, "step": 70450 }, { "grad_norm": 0.06288137286901474, "learning_rate": 1.3013887837338506e-05, "loss": 0.006, "step": 70460 }, { "grad_norm": 0.05721662566065788, "learning_rate": 1.3001623608018477e-05, "loss": 0.0082, "step": 70470 }, { "grad_norm": 0.052525509148836136, "learning_rate": 1.2989364296605312e-05, "loss": 0.0041, "step": 70480 }, { "grad_norm": 0.07201110571622849, "learning_rate": 1.2977109904728586e-05, "loss": 0.0046, "step": 70490 }, { "grad_norm": 0.04361940920352936, "learning_rate": 1.296486043401714e-05, "loss": 0.0031, "step": 70500 }, { "grad_norm": 0.035579632967710495, "learning_rate": 1.2952615886099213e-05, "loss": 0.0044, "step": 70510 }, { "grad_norm": 0.026720011606812477, "learning_rate": 1.2940376262602377e-05, "loss": 0.0047, "step": 70520 }, { "grad_norm": 0.041393086314201355, "learning_rate": 1.2928141565153557e-05, "loss": 0.0039, "step": 70530 }, { "grad_norm": 0.034128233790397644, "learning_rate": 1.2915911795379009e-05, "loss": 0.0045, "step": 70540 }, { "grad_norm": 0.03364385664463043, "learning_rate": 1.2903686954904348e-05, "loss": 0.0071, "step": 70550 }, { "grad_norm": 0.05671615153551102, "learning_rate": 1.2891467045354521e-05, "loss": 0.0062, "step": 70560 }, { "grad_norm": 0.060884054750204086, "learning_rate": 1.2879252068353825e-05, "loss": 0.0052, "step": 70570 }, { "grad_norm": 0.045232344418764114, "learning_rate": 1.2867042025525911e-05, "loss": 0.0053, "step": 70580 }, { "grad_norm": 0.04473768547177315, "learning_rate": 1.2854836918493757e-05, "loss": 0.0053, "step": 70590 }, { "grad_norm": 0.03170591592788696, "learning_rate": 1.2842636748879694e-05, "loss": 0.0063, "step": 70600 }, { "grad_norm": 0.060898665338754654, "learning_rate": 1.2830441518305409e-05, "loss": 0.0047, "step": 70610 }, { "grad_norm": 0.04070783406496048, "learning_rate": 1.2818251228391897e-05, "loss": 0.0042, "step": 70620 }, { "grad_norm": 0.03717246651649475, "learning_rate": 1.280606588075951e-05, "loss": 0.0046, "step": 70630 }, { "grad_norm": 0.03466878458857536, "learning_rate": 1.2793885477027989e-05, "loss": 0.0059, "step": 70640 }, { "grad_norm": 0.0404946468770504, "learning_rate": 1.2781710018816362e-05, "loss": 0.0065, "step": 70650 }, { "grad_norm": 0.04311511293053627, "learning_rate": 1.2769539507743028e-05, "loss": 0.0059, "step": 70660 }, { "grad_norm": 0.07141322642564774, "learning_rate": 1.2757373945425699e-05, "loss": 0.0067, "step": 70670 }, { "grad_norm": 0.03213423117995262, "learning_rate": 1.2745213333481459e-05, "loss": 0.0037, "step": 70680 }, { "grad_norm": 0.04966013506054878, "learning_rate": 1.2733057673526721e-05, "loss": 0.0047, "step": 70690 }, { "grad_norm": 0.042619913816452026, "learning_rate": 1.2720906967177226e-05, "loss": 0.0051, "step": 70700 }, { "grad_norm": 0.038930926471948624, "learning_rate": 1.2708761216048126e-05, "loss": 0.0053, "step": 70710 }, { "grad_norm": 0.03579794242978096, "learning_rate": 1.2696620421753813e-05, "loss": 0.0061, "step": 70720 }, { "grad_norm": 0.10517767816781998, "learning_rate": 1.268448458590808e-05, "loss": 0.0068, "step": 70730 }, { "grad_norm": 0.02056042104959488, "learning_rate": 1.2672353710124058e-05, "loss": 0.0047, "step": 70740 }, { "grad_norm": 0.06602365523576736, "learning_rate": 1.2660227796014196e-05, "loss": 0.0039, "step": 70750 }, { "grad_norm": 0.04265971481800079, "learning_rate": 1.264810684519031e-05, "loss": 0.0073, "step": 70760 }, { "grad_norm": 0.060128554701805115, "learning_rate": 1.2635990859263552e-05, "loss": 0.004, "step": 70770 }, { "grad_norm": 0.024320336058735847, "learning_rate": 1.2623879839844365e-05, "loss": 0.0053, "step": 70780 }, { "grad_norm": 0.052729532122612, "learning_rate": 1.261177378854262e-05, "loss": 0.0056, "step": 70790 }, { "grad_norm": 0.04585950821638107, "learning_rate": 1.2599672706967453e-05, "loss": 0.0062, "step": 70800 }, { "grad_norm": 0.0678902268409729, "learning_rate": 1.258757659672738e-05, "loss": 0.0074, "step": 70810 }, { "grad_norm": 0.057004086673259735, "learning_rate": 1.2575485459430241e-05, "loss": 0.0045, "step": 70820 }, { "grad_norm": 0.04407302662730217, "learning_rate": 1.2563399296683225e-05, "loss": 0.006, "step": 70830 }, { "grad_norm": 0.09801755845546722, "learning_rate": 1.255131811009283e-05, "loss": 0.0057, "step": 70840 }, { "grad_norm": 0.047409478574991226, "learning_rate": 1.2539241901264908e-05, "loss": 0.0054, "step": 70850 }, { "grad_norm": 0.03682522848248482, "learning_rate": 1.2527170671804689e-05, "loss": 0.0046, "step": 70860 }, { "grad_norm": 0.02662607468664646, "learning_rate": 1.2515104423316693e-05, "loss": 0.0046, "step": 70870 }, { "grad_norm": 0.024154704064130783, "learning_rate": 1.2503043157404808e-05, "loss": 0.0062, "step": 70880 }, { "grad_norm": 0.03510661423206329, "learning_rate": 1.2490986875672212e-05, "loss": 0.0034, "step": 70890 }, { "grad_norm": 0.04932800680398941, "learning_rate": 1.2478935579721468e-05, "loss": 0.0063, "step": 70900 }, { "grad_norm": 0.034875091165304184, "learning_rate": 1.2466889271154463e-05, "loss": 0.0051, "step": 70910 }, { "grad_norm": 0.0684254989027977, "learning_rate": 1.2454847951572402e-05, "loss": 0.0046, "step": 70920 }, { "grad_norm": 0.02222476899623871, "learning_rate": 1.244281162257589e-05, "loss": 0.0053, "step": 70930 }, { "grad_norm": 0.034924883395433426, "learning_rate": 1.2430780285764781e-05, "loss": 0.0061, "step": 70940 }, { "grad_norm": 0.03239697217941284, "learning_rate": 1.2418753942738321e-05, "loss": 0.0036, "step": 70950 }, { "grad_norm": 0.04270988330245018, "learning_rate": 1.2406732595095073e-05, "loss": 0.0053, "step": 70960 }, { "grad_norm": 0.04791078343987465, "learning_rate": 1.2394716244432947e-05, "loss": 0.0058, "step": 70970 }, { "grad_norm": 0.0367693230509758, "learning_rate": 1.2382704892349185e-05, "loss": 0.005, "step": 70980 }, { "grad_norm": 0.035710789263248444, "learning_rate": 1.2370698540440368e-05, "loss": 0.0043, "step": 70990 }, { "grad_norm": 0.0378575399518013, "learning_rate": 1.2358697190302371e-05, "loss": 0.0054, "step": 71000 }, { "grad_norm": 0.0638367086648941, "learning_rate": 1.2346700843530485e-05, "loss": 0.0037, "step": 71010 }, { "grad_norm": 0.09013234078884125, "learning_rate": 1.2334709501719272e-05, "loss": 0.0059, "step": 71020 }, { "grad_norm": 0.028720568865537643, "learning_rate": 1.2322723166462646e-05, "loss": 0.0059, "step": 71030 }, { "grad_norm": 0.03924498334527016, "learning_rate": 1.2310741839353885e-05, "loss": 0.0089, "step": 71040 }, { "grad_norm": 0.06897449493408203, "learning_rate": 1.2298765521985528e-05, "loss": 0.0051, "step": 71050 }, { "grad_norm": 0.05221148207783699, "learning_rate": 1.2286794215949516e-05, "loss": 0.0042, "step": 71060 }, { "grad_norm": 0.04200970008969307, "learning_rate": 1.227482792283709e-05, "loss": 0.0056, "step": 71070 }, { "grad_norm": 0.07353803515434265, "learning_rate": 1.2262866644238863e-05, "loss": 0.0052, "step": 71080 }, { "grad_norm": 0.057804711163043976, "learning_rate": 1.2250910381744756e-05, "loss": 0.0051, "step": 71090 }, { "grad_norm": 0.06903175264596939, "learning_rate": 1.223895913694399e-05, "loss": 0.0051, "step": 71100 }, { "grad_norm": 0.04881742596626282, "learning_rate": 1.2227012911425162e-05, "loss": 0.0059, "step": 71110 }, { "grad_norm": 0.050632428377866745, "learning_rate": 1.2215071706776204e-05, "loss": 0.0055, "step": 71120 }, { "grad_norm": 0.04985510930418968, "learning_rate": 1.2203135524584358e-05, "loss": 0.0052, "step": 71130 }, { "grad_norm": 0.043959878385066986, "learning_rate": 1.219120436643621e-05, "loss": 0.0036, "step": 71140 }, { "grad_norm": 0.030854444950819016, "learning_rate": 1.2179278233917674e-05, "loss": 0.0053, "step": 71150 }, { "grad_norm": 0.038929786533117294, "learning_rate": 1.2167357128614005e-05, "loss": 0.004, "step": 71160 }, { "grad_norm": 0.03095882758498192, "learning_rate": 1.2155441052109773e-05, "loss": 0.0054, "step": 71170 }, { "grad_norm": 0.04182612523436546, "learning_rate": 1.2143530005988896e-05, "loss": 0.0039, "step": 71180 }, { "grad_norm": 0.0674801915884018, "learning_rate": 1.2131623991834617e-05, "loss": 0.0069, "step": 71190 }, { "grad_norm": 0.0363929383456707, "learning_rate": 1.2119723011229505e-05, "loss": 0.0034, "step": 71200 }, { "grad_norm": 0.04623514413833618, "learning_rate": 1.210782706575549e-05, "loss": 0.004, "step": 71210 }, { "grad_norm": 0.05554686859250069, "learning_rate": 1.2095936156993754e-05, "loss": 0.0078, "step": 71220 }, { "grad_norm": 0.0448189377784729, "learning_rate": 1.2084050286524912e-05, "loss": 0.0049, "step": 71230 }, { "grad_norm": 0.03320841118693352, "learning_rate": 1.207216945592884e-05, "loss": 0.0044, "step": 71240 }, { "grad_norm": 0.06532932072877884, "learning_rate": 1.2060293666784766e-05, "loss": 0.0042, "step": 71250 }, { "grad_norm": 0.04119464010000229, "learning_rate": 1.2048422920671271e-05, "loss": 0.0048, "step": 71260 }, { "grad_norm": 0.02722538262605667, "learning_rate": 1.20365572191662e-05, "loss": 0.006, "step": 71270 }, { "grad_norm": 0.03125382214784622, "learning_rate": 1.2024696563846788e-05, "loss": 0.0037, "step": 71280 }, { "grad_norm": 0.05512522906064987, "learning_rate": 1.2012840956289583e-05, "loss": 0.0045, "step": 71290 }, { "grad_norm": 0.03900831192731857, "learning_rate": 1.2000990398070438e-05, "loss": 0.0063, "step": 71300 }, { "grad_norm": 0.060353923588991165, "learning_rate": 1.1989144890764603e-05, "loss": 0.0046, "step": 71310 }, { "grad_norm": 0.08014962077140808, "learning_rate": 1.1977304435946562e-05, "loss": 0.0057, "step": 71320 }, { "grad_norm": 0.06798439472913742, "learning_rate": 1.1965469035190197e-05, "loss": 0.0062, "step": 71330 }, { "grad_norm": 0.034419696778059006, "learning_rate": 1.1953638690068686e-05, "loss": 0.0038, "step": 71340 }, { "grad_norm": 0.046093229204416275, "learning_rate": 1.194181340215455e-05, "loss": 0.0057, "step": 71350 }, { "grad_norm": 0.04034063592553139, "learning_rate": 1.1929993173019637e-05, "loss": 0.0047, "step": 71360 }, { "grad_norm": 0.06701186299324036, "learning_rate": 1.191817800423511e-05, "loss": 0.0062, "step": 71370 }, { "grad_norm": 0.04514788091182709, "learning_rate": 1.1906367897371473e-05, "loss": 0.0054, "step": 71380 }, { "grad_norm": 0.03238672390580177, "learning_rate": 1.1894562853998547e-05, "loss": 0.0043, "step": 71390 }, { "grad_norm": 0.02860374003648758, "learning_rate": 1.1882762875685494e-05, "loss": 0.004, "step": 71400 }, { "grad_norm": 0.031678784638643265, "learning_rate": 1.1870967964000784e-05, "loss": 0.0084, "step": 71410 }, { "grad_norm": 0.05008308216929436, "learning_rate": 1.1859178120512227e-05, "loss": 0.0034, "step": 71420 }, { "grad_norm": 0.03562179580330849, "learning_rate": 1.1847393346786966e-05, "loss": 0.0071, "step": 71430 }, { "grad_norm": 0.05802258476614952, "learning_rate": 1.1835613644391424e-05, "loss": 0.0031, "step": 71440 }, { "grad_norm": 0.05588585138320923, "learning_rate": 1.1823839014891424e-05, "loss": 0.0044, "step": 71450 }, { "grad_norm": 0.053183943033218384, "learning_rate": 1.181206945985206e-05, "loss": 0.005, "step": 71460 }, { "grad_norm": 0.039195865392684937, "learning_rate": 1.1800304980837768e-05, "loss": 0.0046, "step": 71470 }, { "grad_norm": 0.03639660403132439, "learning_rate": 1.1788545579412324e-05, "loss": 0.0063, "step": 71480 }, { "grad_norm": 0.03778846934437752, "learning_rate": 1.1776791257138792e-05, "loss": 0.0045, "step": 71490 }, { "grad_norm": 0.04128585383296013, "learning_rate": 1.1765042015579585e-05, "loss": 0.0044, "step": 71500 }, { "grad_norm": 0.04921842738986015, "learning_rate": 1.1753297856296447e-05, "loss": 0.0053, "step": 71510 }, { "grad_norm": 0.03665896877646446, "learning_rate": 1.174155878085042e-05, "loss": 0.0047, "step": 71520 }, { "grad_norm": 0.03334325924515724, "learning_rate": 1.172982479080193e-05, "loss": 0.0042, "step": 71530 }, { "grad_norm": 0.04811786115169525, "learning_rate": 1.1718095887710644e-05, "loss": 0.0055, "step": 71540 }, { "grad_norm": 0.04433295130729675, "learning_rate": 1.1706372073135613e-05, "loss": 0.0038, "step": 71550 }, { "grad_norm": 0.0433746837079525, "learning_rate": 1.1694653348635181e-05, "loss": 0.0045, "step": 71560 }, { "grad_norm": 0.0619228221476078, "learning_rate": 1.1682939715767043e-05, "loss": 0.0041, "step": 71570 }, { "grad_norm": 0.03520987555384636, "learning_rate": 1.1671231176088188e-05, "loss": 0.0038, "step": 71580 }, { "grad_norm": 0.03430072218179703, "learning_rate": 1.1659527731154945e-05, "loss": 0.0057, "step": 71590 }, { "grad_norm": 0.054822973906993866, "learning_rate": 1.1647829382522962e-05, "loss": 0.0057, "step": 71600 }, { "grad_norm": 0.05062299221754074, "learning_rate": 1.1636136131747215e-05, "loss": 0.0059, "step": 71610 }, { "grad_norm": 0.03951700031757355, "learning_rate": 1.1624447980381986e-05, "loss": 0.0049, "step": 71620 }, { "grad_norm": 0.03447399660944939, "learning_rate": 1.1612764929980902e-05, "loss": 0.0054, "step": 71630 }, { "grad_norm": 0.04714520275592804, "learning_rate": 1.1601086982096892e-05, "loss": 0.005, "step": 71640 }, { "grad_norm": 0.06576358526945114, "learning_rate": 1.1589414138282235e-05, "loss": 0.0042, "step": 71650 }, { "grad_norm": 0.0469796247780323, "learning_rate": 1.157774640008848e-05, "loss": 0.0042, "step": 71660 }, { "grad_norm": 0.02552618645131588, "learning_rate": 1.1566083769066527e-05, "loss": 0.005, "step": 71670 }, { "grad_norm": 0.029530778527259827, "learning_rate": 1.1554426246766625e-05, "loss": 0.0046, "step": 71680 }, { "grad_norm": 0.06259217858314514, "learning_rate": 1.1542773834738313e-05, "loss": 0.0073, "step": 71690 }, { "grad_norm": 0.05524260550737381, "learning_rate": 1.1531126534530468e-05, "loss": 0.0045, "step": 71700 }, { "grad_norm": 0.07506271451711655, "learning_rate": 1.151948434769124e-05, "loss": 0.0049, "step": 71710 }, { "grad_norm": 0.04913688823580742, "learning_rate": 1.1507847275768146e-05, "loss": 0.0046, "step": 71720 }, { "grad_norm": 0.07808740437030792, "learning_rate": 1.149621532030803e-05, "loss": 0.007, "step": 71730 }, { "grad_norm": 0.039726871997117996, "learning_rate": 1.1484588482857e-05, "loss": 0.0059, "step": 71740 }, { "grad_norm": 0.03244185447692871, "learning_rate": 1.147296676496058e-05, "loss": 0.0034, "step": 71750 }, { "grad_norm": 0.03947385028004646, "learning_rate": 1.146135016816351e-05, "loss": 0.0045, "step": 71760 }, { "grad_norm": 0.03683972358703613, "learning_rate": 1.1449738694009903e-05, "loss": 0.0046, "step": 71770 }, { "grad_norm": 0.04876541346311569, "learning_rate": 1.1438132344043185e-05, "loss": 0.0047, "step": 71780 }, { "grad_norm": 0.023176634684205055, "learning_rate": 1.1426531119806095e-05, "loss": 0.0047, "step": 71790 }, { "grad_norm": 0.05196303129196167, "learning_rate": 1.1414935022840695e-05, "loss": 0.0065, "step": 71800 }, { "grad_norm": 0.06183145195245743, "learning_rate": 1.1403344054688386e-05, "loss": 0.0038, "step": 71810 }, { "grad_norm": 0.09456321597099304, "learning_rate": 1.1391758216889808e-05, "loss": 0.0071, "step": 71820 }, { "grad_norm": 0.07768724113702774, "learning_rate": 1.1380177510985035e-05, "loss": 0.0063, "step": 71830 }, { "grad_norm": 0.03755032643675804, "learning_rate": 1.1368601938513374e-05, "loss": 0.0039, "step": 71840 }, { "grad_norm": 0.02996799908578396, "learning_rate": 1.1357031501013483e-05, "loss": 0.0047, "step": 71850 }, { "grad_norm": 0.06956250965595245, "learning_rate": 1.1345466200023324e-05, "loss": 0.0048, "step": 71860 }, { "grad_norm": 0.043634142726659775, "learning_rate": 1.1333906037080205e-05, "loss": 0.0037, "step": 71870 }, { "grad_norm": 0.046764083206653595, "learning_rate": 1.132235101372069e-05, "loss": 0.0071, "step": 71880 }, { "grad_norm": 0.05481364578008652, "learning_rate": 1.1310801131480707e-05, "loss": 0.0065, "step": 71890 }, { "grad_norm": 0.019448788836598396, "learning_rate": 1.1299256391895519e-05, "loss": 0.0053, "step": 71900 }, { "grad_norm": 0.037584371864795685, "learning_rate": 1.1287716796499664e-05, "loss": 0.0047, "step": 71910 }, { "grad_norm": 0.04547618329524994, "learning_rate": 1.1276182346827025e-05, "loss": 0.0045, "step": 71920 }, { "grad_norm": 0.036856699734926224, "learning_rate": 1.1264653044410756e-05, "loss": 0.0062, "step": 71930 }, { "grad_norm": 0.08990896493196487, "learning_rate": 1.1253128890783377e-05, "loss": 0.0057, "step": 71940 }, { "grad_norm": 0.0339709036052227, "learning_rate": 1.1241609887476696e-05, "loss": 0.0087, "step": 71950 }, { "grad_norm": 0.0490742102265358, "learning_rate": 1.1230096036021836e-05, "loss": 0.0059, "step": 71960 }, { "grad_norm": 0.04202624410390854, "learning_rate": 1.1218587337949287e-05, "loss": 0.0053, "step": 71970 }, { "grad_norm": 0.03926268219947815, "learning_rate": 1.120708379478877e-05, "loss": 0.0037, "step": 71980 }, { "grad_norm": 0.02724621072411537, "learning_rate": 1.119558540806937e-05, "loss": 0.0084, "step": 71990 }, { "grad_norm": 0.056941188871860504, "learning_rate": 1.1184092179319477e-05, "loss": 0.0061, "step": 72000 }, { "grad_norm": 0.04732701927423477, "learning_rate": 1.1172604110066808e-05, "loss": 0.0044, "step": 72010 }, { "grad_norm": 0.06586398184299469, "learning_rate": 1.1161121201838376e-05, "loss": 0.0035, "step": 72020 }, { "grad_norm": 0.04135485365986824, "learning_rate": 1.1149643456160524e-05, "loss": 0.0054, "step": 72030 }, { "grad_norm": 0.04176510497927666, "learning_rate": 1.1138170874558863e-05, "loss": 0.0044, "step": 72040 }, { "grad_norm": 0.06026120111346245, "learning_rate": 1.1126703458558396e-05, "loss": 0.0057, "step": 72050 }, { "grad_norm": 0.032831426709890366, "learning_rate": 1.1115241209683386e-05, "loss": 0.0046, "step": 72060 }, { "grad_norm": 0.033640820533037186, "learning_rate": 1.1103784129457412e-05, "loss": 0.0052, "step": 72070 }, { "grad_norm": 0.05603833124041557, "learning_rate": 1.1092332219403395e-05, "loss": 0.0042, "step": 72080 }, { "grad_norm": 0.03439028561115265, "learning_rate": 1.1080885481043524e-05, "loss": 0.0054, "step": 72090 }, { "grad_norm": 0.0657341480255127, "learning_rate": 1.106944391589933e-05, "loss": 0.0042, "step": 72100 }, { "grad_norm": 0.037437066435813904, "learning_rate": 1.1058007525491648e-05, "loss": 0.0063, "step": 72110 }, { "grad_norm": 0.029778165742754936, "learning_rate": 1.1046576311340646e-05, "loss": 0.0034, "step": 72120 }, { "grad_norm": 0.04110686108469963, "learning_rate": 1.1035150274965794e-05, "loss": 0.0065, "step": 72130 }, { "grad_norm": 0.0660443902015686, "learning_rate": 1.1023729417885836e-05, "loss": 0.0037, "step": 72140 }, { "grad_norm": 0.03994560241699219, "learning_rate": 1.1012313741618869e-05, "loss": 0.0067, "step": 72150 }, { "grad_norm": 0.029576655477285385, "learning_rate": 1.1000903247682298e-05, "loss": 0.0041, "step": 72160 }, { "grad_norm": 0.03521827980875969, "learning_rate": 1.0989497937592818e-05, "loss": 0.006, "step": 72170 }, { "grad_norm": 0.024053847417235374, "learning_rate": 1.0978097812866461e-05, "loss": 0.0037, "step": 72180 }, { "grad_norm": 0.021587520837783813, "learning_rate": 1.0966702875018553e-05, "loss": 0.006, "step": 72190 }, { "grad_norm": 0.06108308956027031, "learning_rate": 1.0955313125563733e-05, "loss": 0.0045, "step": 72200 }, { "grad_norm": 0.044588204473257065, "learning_rate": 1.0943928566015954e-05, "loss": 0.0042, "step": 72210 }, { "grad_norm": 0.04381110891699791, "learning_rate": 1.0932549197888475e-05, "loss": 0.0063, "step": 72220 }, { "grad_norm": 0.025038423016667366, "learning_rate": 1.0921175022693869e-05, "loss": 0.0048, "step": 72230 }, { "grad_norm": 0.03742818906903267, "learning_rate": 1.0909806041944014e-05, "loss": 0.0039, "step": 72240 }, { "grad_norm": 0.02798573486506939, "learning_rate": 1.0898442257150121e-05, "loss": 0.0061, "step": 72250 }, { "grad_norm": 0.04178954288363457, "learning_rate": 1.088708366982264e-05, "loss": 0.0041, "step": 72260 }, { "grad_norm": 0.04020051285624504, "learning_rate": 1.0875730281471424e-05, "loss": 0.0049, "step": 72270 }, { "grad_norm": 0.03188132122159004, "learning_rate": 1.0864382093605573e-05, "loss": 0.0037, "step": 72280 }, { "grad_norm": 0.02221558429300785, "learning_rate": 1.0853039107733525e-05, "loss": 0.0054, "step": 72290 }, { "grad_norm": 0.043785467743873596, "learning_rate": 1.0841701325363018e-05, "loss": 0.0046, "step": 72300 }, { "grad_norm": 0.03951245918869972, "learning_rate": 1.0830368748001073e-05, "loss": 0.0055, "step": 72310 }, { "grad_norm": 0.05003019794821739, "learning_rate": 1.0819041377154054e-05, "loss": 0.0043, "step": 72320 }, { "grad_norm": 0.04904118552803993, "learning_rate": 1.0807719214327622e-05, "loss": 0.0041, "step": 72330 }, { "grad_norm": 0.04615493491292, "learning_rate": 1.0796402261026727e-05, "loss": 0.0068, "step": 72340 }, { "grad_norm": 0.06747802346944809, "learning_rate": 1.0785090518755691e-05, "loss": 0.0048, "step": 72350 }, { "grad_norm": 0.06538506597280502, "learning_rate": 1.0773783989018043e-05, "loss": 0.0066, "step": 72360 }, { "grad_norm": 0.03557389974594116, "learning_rate": 1.0762482673316698e-05, "loss": 0.0046, "step": 72370 }, { "grad_norm": 0.024889998137950897, "learning_rate": 1.0751186573153849e-05, "loss": 0.0053, "step": 72380 }, { "grad_norm": 0.05632805824279785, "learning_rate": 1.0739895690030993e-05, "loss": 0.0047, "step": 72390 }, { "grad_norm": 0.04195193573832512, "learning_rate": 1.0728610025448949e-05, "loss": 0.0056, "step": 72400 }, { "grad_norm": 0.0769243910908699, "learning_rate": 1.0717329580907825e-05, "loss": 0.0066, "step": 72410 }, { "grad_norm": 0.07154975086450577, "learning_rate": 1.0706054357907048e-05, "loss": 0.0039, "step": 72420 }, { "grad_norm": 0.05458126589655876, "learning_rate": 1.0694784357945337e-05, "loss": 0.0042, "step": 72430 }, { "grad_norm": 0.06750836223363876, "learning_rate": 1.0683519582520735e-05, "loss": 0.0052, "step": 72440 }, { "grad_norm": 0.09029118716716766, "learning_rate": 1.0672260033130582e-05, "loss": 0.0053, "step": 72450 }, { "grad_norm": 0.08927179127931595, "learning_rate": 1.0661005711271515e-05, "loss": 0.0064, "step": 72460 }, { "grad_norm": 0.04086821526288986, "learning_rate": 1.0649756618439504e-05, "loss": 0.0055, "step": 72470 }, { "grad_norm": 0.05310561880469322, "learning_rate": 1.0638512756129754e-05, "loss": 0.0075, "step": 72480 }, { "grad_norm": 0.042637668550014496, "learning_rate": 1.0627274125836878e-05, "loss": 0.0048, "step": 72490 }, { "grad_norm": 0.06474874913692474, "learning_rate": 1.0616040729054716e-05, "loss": 0.0056, "step": 72500 }, { "grad_norm": 0.036541420966386795, "learning_rate": 1.0604812567276433e-05, "loss": 0.0039, "step": 72510 }, { "grad_norm": 0.027175555005669594, "learning_rate": 1.0593589641994527e-05, "loss": 0.005, "step": 72520 }, { "grad_norm": 0.0719136968255043, "learning_rate": 1.0582371954700743e-05, "loss": 0.0066, "step": 72530 }, { "grad_norm": 0.049710072576999664, "learning_rate": 1.0571159506886168e-05, "loss": 0.0071, "step": 72540 }, { "grad_norm": 0.0875149667263031, "learning_rate": 1.0559952300041199e-05, "loss": 0.0074, "step": 72550 }, { "grad_norm": 0.0727081224322319, "learning_rate": 1.0548750335655494e-05, "loss": 0.0053, "step": 72560 }, { "grad_norm": 0.05563460290431976, "learning_rate": 1.0537553615218094e-05, "loss": 0.0075, "step": 72570 }, { "grad_norm": 0.03447651490569115, "learning_rate": 1.0526362140217244e-05, "loss": 0.0061, "step": 72580 }, { "grad_norm": 0.038978081196546555, "learning_rate": 1.0515175912140562e-05, "loss": 0.0037, "step": 72590 }, { "grad_norm": 0.04481533169746399, "learning_rate": 1.0503994932474936e-05, "loss": 0.006, "step": 72600 }, { "grad_norm": 0.044789642095565796, "learning_rate": 1.0492819202706578e-05, "loss": 0.0058, "step": 72610 }, { "grad_norm": 0.04087573289871216, "learning_rate": 1.0481648724320986e-05, "loss": 0.0048, "step": 72620 }, { "grad_norm": 0.029614746570587158, "learning_rate": 1.0470483498802964e-05, "loss": 0.0038, "step": 72630 }, { "grad_norm": 0.040846820920705795, "learning_rate": 1.0459323527636616e-05, "loss": 0.0038, "step": 72640 }, { "grad_norm": 0.05030803382396698, "learning_rate": 1.0448168812305359e-05, "loss": 0.0048, "step": 72650 }, { "grad_norm": 0.04317418485879898, "learning_rate": 1.0437019354291893e-05, "loss": 0.007, "step": 72660 }, { "grad_norm": 0.049164459109306335, "learning_rate": 1.042587515507824e-05, "loss": 0.0049, "step": 72670 }, { "grad_norm": 0.07486271858215332, "learning_rate": 1.0414736216145699e-05, "loss": 0.0042, "step": 72680 }, { "grad_norm": 0.023565175011754036, "learning_rate": 1.0403602538974905e-05, "loss": 0.0066, "step": 72690 }, { "grad_norm": 0.04453406110405922, "learning_rate": 1.0392474125045742e-05, "loss": 0.0045, "step": 72700 }, { "grad_norm": 0.0584585927426815, "learning_rate": 1.0381350975837422e-05, "loss": 0.0035, "step": 72710 }, { "grad_norm": 0.03764143958687782, "learning_rate": 1.037023309282849e-05, "loss": 0.004, "step": 72720 }, { "grad_norm": 0.0445064976811409, "learning_rate": 1.0359120477496742e-05, "loss": 0.0042, "step": 72730 }, { "grad_norm": 0.04533439129590988, "learning_rate": 1.0348013131319306e-05, "loss": 0.0078, "step": 72740 }, { "grad_norm": 0.05616028234362602, "learning_rate": 1.0336911055772569e-05, "loss": 0.005, "step": 72750 }, { "grad_norm": 0.045583613216876984, "learning_rate": 1.0325814252332255e-05, "loss": 0.0065, "step": 72760 }, { "grad_norm": 0.08885642886161804, "learning_rate": 1.031472272247338e-05, "loss": 0.0052, "step": 72770 }, { "grad_norm": 0.05786168575286865, "learning_rate": 1.0303636467670242e-05, "loss": 0.0083, "step": 72780 }, { "grad_norm": 0.05442601069808006, "learning_rate": 1.0292555489396483e-05, "loss": 0.0068, "step": 72790 }, { "grad_norm": 0.06868429481983185, "learning_rate": 1.0281479789124981e-05, "loss": 0.0067, "step": 72800 }, { "grad_norm": 0.0770077109336853, "learning_rate": 1.0270409368327954e-05, "loss": 0.005, "step": 72810 }, { "grad_norm": 0.04915914312005043, "learning_rate": 1.0259344228476903e-05, "loss": 0.0043, "step": 72820 }, { "grad_norm": 0.05167567357420921, "learning_rate": 1.0248284371042633e-05, "loss": 0.0059, "step": 72830 }, { "grad_norm": 0.03656020388007164, "learning_rate": 1.0237229797495245e-05, "loss": 0.0049, "step": 72840 }, { "grad_norm": 0.03577810525894165, "learning_rate": 1.022618050930415e-05, "loss": 0.0035, "step": 72850 }, { "grad_norm": 0.04034369811415672, "learning_rate": 1.0215136507938006e-05, "loss": 0.0047, "step": 72860 }, { "grad_norm": 0.03611002862453461, "learning_rate": 1.0204097794864848e-05, "loss": 0.0059, "step": 72870 }, { "grad_norm": 0.04259568080306053, "learning_rate": 1.0193064371551946e-05, "loss": 0.0057, "step": 72880 }, { "grad_norm": 0.06387913227081299, "learning_rate": 1.0182036239465893e-05, "loss": 0.0044, "step": 72890 }, { "grad_norm": 0.04474957659840584, "learning_rate": 1.0171013400072587e-05, "loss": 0.0042, "step": 72900 }, { "grad_norm": 0.042925141751766205, "learning_rate": 1.0159995854837179e-05, "loss": 0.0089, "step": 72910 }, { "grad_norm": 0.07120762020349503, "learning_rate": 1.014898360522416e-05, "loss": 0.0053, "step": 72920 }, { "grad_norm": 0.0609232522547245, "learning_rate": 1.0137976652697295e-05, "loss": 0.0046, "step": 72930 }, { "grad_norm": 0.03912857919931412, "learning_rate": 1.0126974998719669e-05, "loss": 0.0037, "step": 72940 }, { "grad_norm": 0.04096972197294235, "learning_rate": 1.0115978644753637e-05, "loss": 0.0067, "step": 72950 }, { "grad_norm": 0.06338727474212646, "learning_rate": 1.0104987592260873e-05, "loss": 0.0041, "step": 72960 }, { "grad_norm": 0.056041568517684937, "learning_rate": 1.0094001842702311e-05, "loss": 0.0079, "step": 72970 }, { "grad_norm": 0.03984595462679863, "learning_rate": 1.0083021397538206e-05, "loss": 0.0072, "step": 72980 }, { "grad_norm": 0.059880685061216354, "learning_rate": 1.0072046258228108e-05, "loss": 0.0049, "step": 72990 }, { "grad_norm": 0.0632898136973381, "learning_rate": 1.0061076426230842e-05, "loss": 0.0049, "step": 73000 }, { "grad_norm": 0.031350914388895035, "learning_rate": 1.0050111903004584e-05, "loss": 0.0048, "step": 73010 }, { "grad_norm": 0.03425813093781471, "learning_rate": 1.0039152690006721e-05, "loss": 0.0037, "step": 73020 }, { "grad_norm": 0.04016946628689766, "learning_rate": 1.0028198788693988e-05, "loss": 0.0037, "step": 73030 }, { "grad_norm": 0.0369226299226284, "learning_rate": 1.001725020052241e-05, "loss": 0.0044, "step": 73040 }, { "grad_norm": 0.04229864105582237, "learning_rate": 1.000630692694729e-05, "loss": 0.0067, "step": 73050 }, { "grad_norm": 0.03788512572646141, "learning_rate": 9.995368969423235e-06, "loss": 0.0059, "step": 73060 }, { "grad_norm": 0.04772481322288513, "learning_rate": 9.984436329404156e-06, "loss": 0.0043, "step": 73070 }, { "grad_norm": 0.04220070689916611, "learning_rate": 9.973509008343202e-06, "loss": 0.0042, "step": 73080 }, { "grad_norm": 0.048117708414793015, "learning_rate": 9.962587007692904e-06, "loss": 0.0066, "step": 73090 }, { "grad_norm": 0.07284665107727051, "learning_rate": 9.951670328905016e-06, "loss": 0.0055, "step": 73100 }, { "grad_norm": 0.060986608266830444, "learning_rate": 9.940758973430614e-06, "loss": 0.0036, "step": 73110 }, { "grad_norm": 0.018514906987547874, "learning_rate": 9.929852942720069e-06, "loss": 0.0043, "step": 73120 }, { "grad_norm": 0.020387187600135803, "learning_rate": 9.918952238223007e-06, "loss": 0.0048, "step": 73130 }, { "grad_norm": 0.05801015347242355, "learning_rate": 9.908056861388393e-06, "loss": 0.0044, "step": 73140 }, { "grad_norm": 0.05714850500226021, "learning_rate": 9.89716681366445e-06, "loss": 0.0063, "step": 73150 }, { "grad_norm": 0.07163115590810776, "learning_rate": 9.886282096498733e-06, "loss": 0.005, "step": 73160 }, { "grad_norm": 0.03451218456029892, "learning_rate": 9.875402711338062e-06, "loss": 0.0069, "step": 73170 }, { "grad_norm": 0.028523076325654984, "learning_rate": 9.864528659628524e-06, "loss": 0.0049, "step": 73180 }, { "grad_norm": 0.023861432448029518, "learning_rate": 9.853659942815536e-06, "loss": 0.0051, "step": 73190 }, { "grad_norm": 0.04430123046040535, "learning_rate": 9.842796562343787e-06, "loss": 0.0038, "step": 73200 }, { "grad_norm": 0.03547649458050728, "learning_rate": 9.831938519657268e-06, "loss": 0.0039, "step": 73210 }, { "grad_norm": 0.02894207462668419, "learning_rate": 9.821085816199254e-06, "loss": 0.004, "step": 73220 }, { "grad_norm": 0.030368220061063766, "learning_rate": 9.810238453412307e-06, "loss": 0.0051, "step": 73230 }, { "grad_norm": 0.051633041352033615, "learning_rate": 9.799396432738284e-06, "loss": 0.0079, "step": 73240 }, { "grad_norm": 0.027855614200234413, "learning_rate": 9.788559755618337e-06, "loss": 0.0072, "step": 73250 }, { "grad_norm": 0.06333045661449432, "learning_rate": 9.777728423492888e-06, "loss": 0.0049, "step": 73260 }, { "grad_norm": 0.05831648036837578, "learning_rate": 9.766902437801673e-06, "loss": 0.0088, "step": 73270 }, { "grad_norm": 0.07614382356405258, "learning_rate": 9.756081799983702e-06, "loss": 0.0063, "step": 73280 }, { "grad_norm": 0.04079156368970871, "learning_rate": 9.745266511477291e-06, "loss": 0.0041, "step": 73290 }, { "grad_norm": 0.060754869133234024, "learning_rate": 9.734456573719997e-06, "loss": 0.0059, "step": 73300 }, { "grad_norm": 0.07024139165878296, "learning_rate": 9.723651988148735e-06, "loss": 0.0061, "step": 73310 }, { "grad_norm": 0.0395672582089901, "learning_rate": 9.712852756199663e-06, "loss": 0.0054, "step": 73320 }, { "grad_norm": 0.030712511390447617, "learning_rate": 9.702058879308245e-06, "loss": 0.0044, "step": 73330 }, { "grad_norm": 0.03599005565047264, "learning_rate": 9.691270358909233e-06, "loss": 0.0054, "step": 73340 }, { "grad_norm": 0.031379442662000656, "learning_rate": 9.680487196436633e-06, "loss": 0.0066, "step": 73350 }, { "grad_norm": 0.023376546800136566, "learning_rate": 9.669709393323789e-06, "loss": 0.0044, "step": 73360 }, { "grad_norm": 0.02127612940967083, "learning_rate": 9.658936951003306e-06, "loss": 0.0037, "step": 73370 }, { "grad_norm": 0.06306906044483185, "learning_rate": 9.648169870907064e-06, "loss": 0.0045, "step": 73380 }, { "grad_norm": 0.029326384887099266, "learning_rate": 9.6374081544663e-06, "loss": 0.0042, "step": 73390 }, { "grad_norm": 0.045427143573760986, "learning_rate": 9.62665180311143e-06, "loss": 0.0059, "step": 73400 }, { "grad_norm": 0.03157667815685272, "learning_rate": 9.615900818272234e-06, "loss": 0.0076, "step": 73410 }, { "grad_norm": 0.04327220469713211, "learning_rate": 9.605155201377753e-06, "loss": 0.007, "step": 73420 }, { "grad_norm": 0.04948648437857628, "learning_rate": 9.594414953856324e-06, "loss": 0.0033, "step": 73430 }, { "grad_norm": 0.02816995047032833, "learning_rate": 9.58368007713556e-06, "loss": 0.0036, "step": 73440 }, { "grad_norm": 0.028081171214580536, "learning_rate": 9.57295057264237e-06, "loss": 0.0048, "step": 73450 }, { "grad_norm": 0.05676239728927612, "learning_rate": 9.562226441802935e-06, "loss": 0.0055, "step": 73460 }, { "grad_norm": 0.04507880285382271, "learning_rate": 9.551507686042737e-06, "loss": 0.005, "step": 73470 }, { "grad_norm": 0.02298741601407528, "learning_rate": 9.540794306786532e-06, "loss": 0.0039, "step": 73480 }, { "grad_norm": 0.03455613926053047, "learning_rate": 9.530086305458374e-06, "loss": 0.0035, "step": 73490 }, { "grad_norm": 0.06767751276493073, "learning_rate": 9.519383683481593e-06, "loss": 0.0082, "step": 73500 }, { "grad_norm": 0.03699648752808571, "learning_rate": 9.508686442278808e-06, "loss": 0.0048, "step": 73510 }, { "grad_norm": 0.02859484776854515, "learning_rate": 9.497994583271897e-06, "loss": 0.0041, "step": 73520 }, { "grad_norm": 0.0649053081870079, "learning_rate": 9.48730810788207e-06, "loss": 0.0056, "step": 73530 }, { "grad_norm": 0.05715521425008774, "learning_rate": 9.476627017529793e-06, "loss": 0.0076, "step": 73540 }, { "grad_norm": 0.04976151883602142, "learning_rate": 9.46595131363482e-06, "loss": 0.0043, "step": 73550 }, { "grad_norm": 0.031628988683223724, "learning_rate": 9.4552809976162e-06, "loss": 0.006, "step": 73560 }, { "grad_norm": 0.05285529047250748, "learning_rate": 9.444616070892226e-06, "loss": 0.0053, "step": 73570 }, { "grad_norm": 0.04464636743068695, "learning_rate": 9.433956534880523e-06, "loss": 0.005, "step": 73580 }, { "grad_norm": 0.038420721888542175, "learning_rate": 9.42330239099798e-06, "loss": 0.0037, "step": 73590 }, { "grad_norm": 0.04788149520754814, "learning_rate": 9.412653640660756e-06, "loss": 0.0041, "step": 73600 }, { "grad_norm": 0.08259319514036179, "learning_rate": 9.40201028528434e-06, "loss": 0.0052, "step": 73610 }, { "grad_norm": 0.05356624722480774, "learning_rate": 9.391372326283437e-06, "loss": 0.0112, "step": 73620 }, { "grad_norm": 0.04531686007976532, "learning_rate": 9.380739765072082e-06, "loss": 0.0035, "step": 73630 }, { "grad_norm": 0.046644434332847595, "learning_rate": 9.370112603063574e-06, "loss": 0.0041, "step": 73640 }, { "grad_norm": 0.0547071173787117, "learning_rate": 9.359490841670504e-06, "loss": 0.0044, "step": 73650 }, { "grad_norm": 0.03175542131066322, "learning_rate": 9.348874482304743e-06, "loss": 0.0051, "step": 73660 }, { "grad_norm": 0.029287949204444885, "learning_rate": 9.338263526377433e-06, "loss": 0.006, "step": 73670 }, { "grad_norm": 0.05116914212703705, "learning_rate": 9.327657975299009e-06, "loss": 0.0036, "step": 73680 }, { "grad_norm": 0.07145094871520996, "learning_rate": 9.317057830479186e-06, "loss": 0.003, "step": 73690 }, { "grad_norm": 0.06871771067380905, "learning_rate": 9.306463093326956e-06, "loss": 0.0054, "step": 73700 }, { "grad_norm": 0.036745306104421616, "learning_rate": 9.295873765250601e-06, "loss": 0.0075, "step": 73710 }, { "grad_norm": 0.053730156272649765, "learning_rate": 9.285289847657669e-06, "loss": 0.0041, "step": 73720 }, { "grad_norm": 0.08418073505163193, "learning_rate": 9.274711341955022e-06, "loss": 0.0063, "step": 73730 }, { "grad_norm": 0.05447673797607422, "learning_rate": 9.264138249548743e-06, "loss": 0.0058, "step": 73740 }, { "grad_norm": 0.08525437861680984, "learning_rate": 9.253570571844238e-06, "loss": 0.0043, "step": 73750 }, { "grad_norm": 0.04159612953662872, "learning_rate": 9.24300831024621e-06, "loss": 0.0044, "step": 73760 }, { "grad_norm": 0.02577117644250393, "learning_rate": 9.232451466158604e-06, "loss": 0.0049, "step": 73770 }, { "grad_norm": 0.08246755599975586, "learning_rate": 9.22190004098467e-06, "loss": 0.0061, "step": 73780 }, { "grad_norm": 0.03352740779519081, "learning_rate": 9.211354036126906e-06, "loss": 0.0038, "step": 73790 }, { "grad_norm": 0.05447700247168541, "learning_rate": 9.200813452987122e-06, "loss": 0.0074, "step": 73800 }, { "grad_norm": 0.046707212924957275, "learning_rate": 9.190278292966398e-06, "loss": 0.0055, "step": 73810 }, { "grad_norm": 0.027615834027528763, "learning_rate": 9.179748557465074e-06, "loss": 0.0034, "step": 73820 }, { "grad_norm": 0.07574109733104706, "learning_rate": 9.169224247882824e-06, "loss": 0.0046, "step": 73830 }, { "grad_norm": 0.03873208165168762, "learning_rate": 9.15870536561853e-06, "loss": 0.0054, "step": 73840 }, { "grad_norm": 0.04048497974872589, "learning_rate": 9.148191912070392e-06, "loss": 0.0048, "step": 73850 }, { "grad_norm": 0.03833555430173874, "learning_rate": 9.137683888635884e-06, "loss": 0.0067, "step": 73860 }, { "grad_norm": 0.0783536359667778, "learning_rate": 9.127181296711761e-06, "loss": 0.006, "step": 73870 }, { "grad_norm": 0.02893603965640068, "learning_rate": 9.116684137694043e-06, "loss": 0.0051, "step": 73880 }, { "grad_norm": 0.05142270401120186, "learning_rate": 9.106192412978055e-06, "loss": 0.0057, "step": 73890 }, { "grad_norm": 0.05191854014992714, "learning_rate": 9.095706123958337e-06, "loss": 0.0045, "step": 73900 }, { "grad_norm": 0.03321772441267967, "learning_rate": 9.085225272028786e-06, "loss": 0.0056, "step": 73910 }, { "grad_norm": 0.04211946204304695, "learning_rate": 9.074749858582537e-06, "loss": 0.0048, "step": 73920 }, { "grad_norm": 0.04730159044265747, "learning_rate": 9.064279885012e-06, "loss": 0.0043, "step": 73930 }, { "grad_norm": 0.08301757276058197, "learning_rate": 9.053815352708877e-06, "loss": 0.0048, "step": 73940 }, { "grad_norm": 0.04825715720653534, "learning_rate": 9.043356263064117e-06, "loss": 0.0056, "step": 73950 }, { "grad_norm": 0.06866651028394699, "learning_rate": 9.03290261746798e-06, "loss": 0.0042, "step": 73960 }, { "grad_norm": 0.029391204938292503, "learning_rate": 9.022454417309966e-06, "loss": 0.0045, "step": 73970 }, { "grad_norm": 0.03187971189618111, "learning_rate": 9.012011663978908e-06, "loss": 0.0036, "step": 73980 }, { "grad_norm": 0.05352276191115379, "learning_rate": 9.001574358862875e-06, "loss": 0.0043, "step": 73990 }, { "grad_norm": 0.037119705229997635, "learning_rate": 8.991142503349188e-06, "loss": 0.0078, "step": 74000 }, { "grad_norm": 0.0710432231426239, "learning_rate": 8.980716098824493e-06, "loss": 0.0049, "step": 74010 }, { "grad_norm": 0.05535764992237091, "learning_rate": 8.970295146674689e-06, "loss": 0.005, "step": 74020 }, { "grad_norm": 0.03718024119734764, "learning_rate": 8.959879648284951e-06, "loss": 0.0036, "step": 74030 }, { "grad_norm": 0.0757509246468544, "learning_rate": 8.949469605039717e-06, "loss": 0.0046, "step": 74040 }, { "grad_norm": 0.028759798035025597, "learning_rate": 8.93906501832275e-06, "loss": 0.0047, "step": 74050 }, { "grad_norm": 0.037390463054180145, "learning_rate": 8.928665889517019e-06, "loss": 0.0062, "step": 74060 }, { "grad_norm": 0.073991559445858, "learning_rate": 8.918272220004808e-06, "loss": 0.0075, "step": 74070 }, { "grad_norm": 0.03521152213215828, "learning_rate": 8.907884011167661e-06, "loss": 0.0094, "step": 74080 }, { "grad_norm": 0.04089372977614403, "learning_rate": 8.897501264386415e-06, "loss": 0.0044, "step": 74090 }, { "grad_norm": 0.056517209857702255, "learning_rate": 8.887123981041157e-06, "loss": 0.0042, "step": 74100 }, { "grad_norm": 0.02967160753905773, "learning_rate": 8.87675216251127e-06, "loss": 0.0078, "step": 74110 }, { "grad_norm": 0.09974581003189087, "learning_rate": 8.866385810175377e-06, "loss": 0.0055, "step": 74120 }, { "grad_norm": 0.05375918000936508, "learning_rate": 8.856024925411415e-06, "loss": 0.0074, "step": 74130 }, { "grad_norm": 0.027545025572180748, "learning_rate": 8.845669509596577e-06, "loss": 0.0044, "step": 74140 }, { "grad_norm": 0.025613047182559967, "learning_rate": 8.83531956410732e-06, "loss": 0.0063, "step": 74150 }, { "grad_norm": 0.01808200217783451, "learning_rate": 8.824975090319399e-06, "loss": 0.0034, "step": 74160 }, { "grad_norm": 0.048718854784965515, "learning_rate": 8.814636089607797e-06, "loss": 0.0041, "step": 74170 }, { "grad_norm": 0.03924618661403656, "learning_rate": 8.80430256334681e-06, "loss": 0.0052, "step": 74180 }, { "grad_norm": 0.022773712873458862, "learning_rate": 8.793974512909986e-06, "loss": 0.0059, "step": 74190 }, { "grad_norm": 0.04378452152013779, "learning_rate": 8.783651939670174e-06, "loss": 0.0043, "step": 74200 }, { "grad_norm": 0.031109442934393883, "learning_rate": 8.77333484499947e-06, "loss": 0.0036, "step": 74210 }, { "grad_norm": 0.03966517001390457, "learning_rate": 8.763023230269219e-06, "loss": 0.0049, "step": 74220 }, { "grad_norm": 0.07054948061704636, "learning_rate": 8.752717096850088e-06, "loss": 0.0061, "step": 74230 }, { "grad_norm": 0.04608270153403282, "learning_rate": 8.742416446111979e-06, "loss": 0.0046, "step": 74240 }, { "grad_norm": 0.03666411712765694, "learning_rate": 8.732121279424082e-06, "loss": 0.0061, "step": 74250 }, { "grad_norm": 0.041202105581760406, "learning_rate": 8.721831598154855e-06, "loss": 0.006, "step": 74260 }, { "grad_norm": 0.03496989980340004, "learning_rate": 8.711547403672026e-06, "loss": 0.0048, "step": 74270 }, { "grad_norm": 0.04510011896491051, "learning_rate": 8.701268697342591e-06, "loss": 0.0036, "step": 74280 }, { "grad_norm": 0.05631083995103836, "learning_rate": 8.690995480532816e-06, "loss": 0.0047, "step": 74290 }, { "grad_norm": 0.03876495361328125, "learning_rate": 8.680727754608243e-06, "loss": 0.0037, "step": 74300 }, { "grad_norm": 0.0589628666639328, "learning_rate": 8.670465520933684e-06, "loss": 0.0061, "step": 74310 }, { "grad_norm": 0.0416066013276577, "learning_rate": 8.660208780873214e-06, "loss": 0.0056, "step": 74320 }, { "grad_norm": 0.05993422493338585, "learning_rate": 8.649957535790194e-06, "loss": 0.0043, "step": 74330 }, { "grad_norm": 0.01878458820283413, "learning_rate": 8.639711787047205e-06, "loss": 0.0055, "step": 74340 }, { "grad_norm": 0.02639780379831791, "learning_rate": 8.629471536006178e-06, "loss": 0.0059, "step": 74350 }, { "grad_norm": 0.03691307082772255, "learning_rate": 8.61923678402825e-06, "loss": 0.0048, "step": 74360 }, { "grad_norm": 0.04556005448102951, "learning_rate": 8.609007532473851e-06, "loss": 0.0041, "step": 74370 }, { "grad_norm": 0.03509214147925377, "learning_rate": 8.598783782702685e-06, "loss": 0.0087, "step": 74380 }, { "grad_norm": 0.0319373793900013, "learning_rate": 8.588565536073695e-06, "loss": 0.0053, "step": 74390 }, { "grad_norm": 0.029834046959877014, "learning_rate": 8.578352793945126e-06, "loss": 0.0048, "step": 74400 }, { "grad_norm": 0.03444921225309372, "learning_rate": 8.568145557674478e-06, "loss": 0.0066, "step": 74410 }, { "grad_norm": 0.019323894754052162, "learning_rate": 8.557943828618503e-06, "loss": 0.0034, "step": 74420 }, { "grad_norm": 0.07223069667816162, "learning_rate": 8.547747608133277e-06, "loss": 0.0046, "step": 74430 }, { "grad_norm": 0.04724092036485672, "learning_rate": 8.537556897574067e-06, "loss": 0.0054, "step": 74440 }, { "grad_norm": 0.054101720452308655, "learning_rate": 8.52737169829546e-06, "loss": 0.0044, "step": 74450 }, { "grad_norm": 0.09209024906158447, "learning_rate": 8.517192011651292e-06, "loss": 0.0055, "step": 74460 }, { "grad_norm": 0.02766970917582512, "learning_rate": 8.507017838994674e-06, "loss": 0.0044, "step": 74470 }, { "grad_norm": 0.045343831181526184, "learning_rate": 8.49684918167798e-06, "loss": 0.0058, "step": 74480 }, { "grad_norm": 0.058892183005809784, "learning_rate": 8.486686041052844e-06, "loss": 0.0035, "step": 74490 }, { "grad_norm": 0.06572242081165314, "learning_rate": 8.476528418470176e-06, "loss": 0.0083, "step": 74500 }, { "grad_norm": 0.04915647581219673, "learning_rate": 8.46637631528015e-06, "loss": 0.0043, "step": 74510 }, { "grad_norm": 0.03097693622112274, "learning_rate": 8.456229732832211e-06, "loss": 0.0065, "step": 74520 }, { "grad_norm": 0.05552227050065994, "learning_rate": 8.446088672475062e-06, "loss": 0.0049, "step": 74530 }, { "grad_norm": 0.038864050060510635, "learning_rate": 8.435953135556673e-06, "loss": 0.0069, "step": 74540 }, { "grad_norm": 0.03542092815041542, "learning_rate": 8.425823123424297e-06, "loss": 0.0048, "step": 74550 }, { "grad_norm": 0.05739814415574074, "learning_rate": 8.415698637424401e-06, "loss": 0.006, "step": 74560 }, { "grad_norm": 0.053277455270290375, "learning_rate": 8.405579678902786e-06, "loss": 0.0058, "step": 74570 }, { "grad_norm": 0.06786993891000748, "learning_rate": 8.395466249204486e-06, "loss": 0.0069, "step": 74580 }, { "grad_norm": 0.025337625294923782, "learning_rate": 8.385358349673794e-06, "loss": 0.005, "step": 74590 }, { "grad_norm": 0.06970897316932678, "learning_rate": 8.375255981654283e-06, "loss": 0.0066, "step": 74600 }, { "grad_norm": 0.033638909459114075, "learning_rate": 8.36515914648876e-06, "loss": 0.0051, "step": 74610 }, { "grad_norm": 0.03505147621035576, "learning_rate": 8.355067845519337e-06, "loss": 0.0043, "step": 74620 }, { "grad_norm": 0.054093461483716965, "learning_rate": 8.344982080087366e-06, "loss": 0.0055, "step": 74630 }, { "grad_norm": 0.021456018090248108, "learning_rate": 8.33490185153346e-06, "loss": 0.0043, "step": 74640 }, { "grad_norm": 0.045804496854543686, "learning_rate": 8.32482716119754e-06, "loss": 0.0038, "step": 74650 }, { "grad_norm": 0.05335022881627083, "learning_rate": 8.314758010418721e-06, "loss": 0.0067, "step": 74660 }, { "grad_norm": 0.04409153759479523, "learning_rate": 8.304694400535428e-06, "loss": 0.0033, "step": 74670 }, { "grad_norm": 0.059875521808862686, "learning_rate": 8.294636332885348e-06, "loss": 0.0055, "step": 74680 }, { "grad_norm": 0.0496944785118103, "learning_rate": 8.284583808805402e-06, "loss": 0.0037, "step": 74690 }, { "grad_norm": 0.03075329400599003, "learning_rate": 8.274536829631813e-06, "loss": 0.0055, "step": 74700 }, { "grad_norm": 0.06484965980052948, "learning_rate": 8.264495396700039e-06, "loss": 0.0052, "step": 74710 }, { "grad_norm": 0.028590867295861244, "learning_rate": 8.254459511344808e-06, "loss": 0.0051, "step": 74720 }, { "grad_norm": 0.04300783574581146, "learning_rate": 8.24442917490012e-06, "loss": 0.0045, "step": 74730 }, { "grad_norm": 0.04472048580646515, "learning_rate": 8.234404388699225e-06, "loss": 0.0045, "step": 74740 }, { "grad_norm": 0.024451257660984993, "learning_rate": 8.224385154074637e-06, "loss": 0.0043, "step": 74750 }, { "grad_norm": 0.04600077122449875, "learning_rate": 8.21437147235814e-06, "loss": 0.0053, "step": 74760 }, { "grad_norm": 0.03360579162836075, "learning_rate": 8.204363344880789e-06, "loss": 0.0041, "step": 74770 }, { "grad_norm": 0.028289509937167168, "learning_rate": 8.194360772972853e-06, "loss": 0.0058, "step": 74780 }, { "grad_norm": 0.04255935177206993, "learning_rate": 8.184363757963904e-06, "loss": 0.0054, "step": 74790 }, { "grad_norm": 0.04173467680811882, "learning_rate": 8.174372301182792e-06, "loss": 0.0054, "step": 74800 }, { "grad_norm": 0.04882439598441124, "learning_rate": 8.164386403957586e-06, "loss": 0.005, "step": 74810 }, { "grad_norm": 0.05955023318529129, "learning_rate": 8.15440606761565e-06, "loss": 0.0057, "step": 74820 }, { "grad_norm": 0.03227177634835243, "learning_rate": 8.144431293483568e-06, "loss": 0.005, "step": 74830 }, { "grad_norm": 0.02920331433415413, "learning_rate": 8.134462082887223e-06, "loss": 0.0071, "step": 74840 }, { "grad_norm": 0.04246053844690323, "learning_rate": 8.12449843715174e-06, "loss": 0.0035, "step": 74850 }, { "grad_norm": 0.02253386378288269, "learning_rate": 8.114540357601502e-06, "loss": 0.0037, "step": 74860 }, { "grad_norm": 0.03781207650899887, "learning_rate": 8.10458784556019e-06, "loss": 0.005, "step": 74870 }, { "grad_norm": 0.06621157377958298, "learning_rate": 8.094640902350687e-06, "loss": 0.0065, "step": 74880 }, { "grad_norm": 0.07993698120117188, "learning_rate": 8.084699529295164e-06, "loss": 0.0082, "step": 74890 }, { "grad_norm": 0.03468841314315796, "learning_rate": 8.074763727715063e-06, "loss": 0.0045, "step": 74900 }, { "grad_norm": 0.07449110597372055, "learning_rate": 8.064833498931057e-06, "loss": 0.0056, "step": 74910 }, { "grad_norm": 0.0551803894340992, "learning_rate": 8.054908844263109e-06, "loss": 0.0049, "step": 74920 }, { "grad_norm": 0.04431247338652611, "learning_rate": 8.04498976503043e-06, "loss": 0.0052, "step": 74930 }, { "grad_norm": 0.043800000101327896, "learning_rate": 8.03507626255145e-06, "loss": 0.0041, "step": 74940 }, { "grad_norm": 0.03442343324422836, "learning_rate": 8.025168338143935e-06, "loss": 0.004, "step": 74950 }, { "grad_norm": 0.030982578173279762, "learning_rate": 8.015265993124853e-06, "loss": 0.0034, "step": 74960 }, { "grad_norm": 0.03239533677697182, "learning_rate": 8.00536922881045e-06, "loss": 0.0047, "step": 74970 }, { "grad_norm": 0.04582226276397705, "learning_rate": 7.99547804651623e-06, "loss": 0.0052, "step": 74980 }, { "grad_norm": 0.040086157619953156, "learning_rate": 7.985592447556934e-06, "loss": 0.0061, "step": 74990 }, { "grad_norm": 0.048221755772829056, "learning_rate": 7.975712433246585e-06, "loss": 0.0054, "step": 75000 }, { "grad_norm": 0.06940966844558716, "learning_rate": 7.965838004898446e-06, "loss": 0.0104, "step": 75010 }, { "grad_norm": 0.0590156652033329, "learning_rate": 7.955969163825078e-06, "loss": 0.0049, "step": 75020 }, { "grad_norm": 0.04613563045859337, "learning_rate": 7.94610591133826e-06, "loss": 0.0042, "step": 75030 }, { "grad_norm": 0.05748813599348068, "learning_rate": 7.936248248749023e-06, "loss": 0.008, "step": 75040 }, { "grad_norm": 0.054213136434555054, "learning_rate": 7.926396177367672e-06, "loss": 0.0063, "step": 75050 }, { "grad_norm": 0.058205511420965195, "learning_rate": 7.916549698503772e-06, "loss": 0.0051, "step": 75060 }, { "grad_norm": 0.04688413813710213, "learning_rate": 7.90670881346614e-06, "loss": 0.0068, "step": 75070 }, { "grad_norm": 0.05015378072857857, "learning_rate": 7.896873523562847e-06, "loss": 0.0047, "step": 75080 }, { "grad_norm": 0.03316226974129677, "learning_rate": 7.887043830101221e-06, "loss": 0.0062, "step": 75090 }, { "grad_norm": 0.027271084487438202, "learning_rate": 7.877219734387842e-06, "loss": 0.0029, "step": 75100 }, { "grad_norm": 0.03635568171739578, "learning_rate": 7.867401237728561e-06, "loss": 0.0043, "step": 75110 }, { "grad_norm": 0.039240289479494095, "learning_rate": 7.857588341428467e-06, "loss": 0.0061, "step": 75120 }, { "grad_norm": 0.048392653465270996, "learning_rate": 7.847781046791919e-06, "loss": 0.0052, "step": 75130 }, { "grad_norm": 0.03972431272268295, "learning_rate": 7.837979355122515e-06, "loss": 0.0064, "step": 75140 }, { "grad_norm": 0.039737120270729065, "learning_rate": 7.828183267723138e-06, "loss": 0.0042, "step": 75150 }, { "grad_norm": 0.03680223971605301, "learning_rate": 7.818392785895862e-06, "loss": 0.0043, "step": 75160 }, { "grad_norm": 0.07477748394012451, "learning_rate": 7.8086079109421e-06, "loss": 0.0084, "step": 75170 }, { "grad_norm": 0.02572595700621605, "learning_rate": 7.798828644162475e-06, "loss": 0.0053, "step": 75180 }, { "grad_norm": 0.03920239955186844, "learning_rate": 7.789054986856852e-06, "loss": 0.0052, "step": 75190 }, { "grad_norm": 0.03667908534407616, "learning_rate": 7.779286940324399e-06, "loss": 0.0092, "step": 75200 }, { "grad_norm": 0.03666095808148384, "learning_rate": 7.769524505863468e-06, "loss": 0.0044, "step": 75210 }, { "grad_norm": 0.07357925176620483, "learning_rate": 7.759767684771719e-06, "loss": 0.0066, "step": 75220 }, { "grad_norm": 0.08198200911283493, "learning_rate": 7.750016478346039e-06, "loss": 0.0047, "step": 75230 }, { "grad_norm": 0.05708606168627739, "learning_rate": 7.740270887882606e-06, "loss": 0.0043, "step": 75240 }, { "grad_norm": 0.04552428796887398, "learning_rate": 7.730530914676826e-06, "loss": 0.0059, "step": 75250 }, { "grad_norm": 0.05861145630478859, "learning_rate": 7.72079656002333e-06, "loss": 0.0054, "step": 75260 }, { "grad_norm": 0.05683796852827072, "learning_rate": 7.711067825216044e-06, "loss": 0.0039, "step": 75270 }, { "grad_norm": 0.06389142572879791, "learning_rate": 7.701344711548142e-06, "loss": 0.0056, "step": 75280 }, { "grad_norm": 0.055534422397613525, "learning_rate": 7.691627220312026e-06, "loss": 0.005, "step": 75290 }, { "grad_norm": 0.06258269399404526, "learning_rate": 7.681915352799375e-06, "loss": 0.004, "step": 75300 }, { "grad_norm": 0.02584609016776085, "learning_rate": 7.672209110301115e-06, "loss": 0.0061, "step": 75310 }, { "grad_norm": 0.04682119935750961, "learning_rate": 7.662508494107417e-06, "loss": 0.0047, "step": 75320 }, { "grad_norm": 0.04149412363767624, "learning_rate": 7.652813505507706e-06, "loss": 0.0047, "step": 75330 }, { "grad_norm": 0.020470594987273216, "learning_rate": 7.64312414579067e-06, "loss": 0.0041, "step": 75340 }, { "grad_norm": 0.04256720468401909, "learning_rate": 7.633440416244231e-06, "loss": 0.0049, "step": 75350 }, { "grad_norm": 0.02398316189646721, "learning_rate": 7.623762318155581e-06, "loss": 0.0041, "step": 75360 }, { "grad_norm": 0.028215257450938225, "learning_rate": 7.6140898528111625e-06, "loss": 0.0032, "step": 75370 }, { "grad_norm": 0.04772451892495155, "learning_rate": 7.6044230214966205e-06, "loss": 0.0072, "step": 75380 }, { "grad_norm": 0.0552842952311039, "learning_rate": 7.594761825496932e-06, "loss": 0.0037, "step": 75390 }, { "grad_norm": 0.08506376296281815, "learning_rate": 7.5851062660962755e-06, "loss": 0.0057, "step": 75400 }, { "grad_norm": 0.05131809413433075, "learning_rate": 7.575456344578085e-06, "loss": 0.0062, "step": 75410 }, { "grad_norm": 0.0664634183049202, "learning_rate": 7.56581206222507e-06, "loss": 0.0059, "step": 75420 }, { "grad_norm": 0.06319248676300049, "learning_rate": 7.556173420319129e-06, "loss": 0.0051, "step": 75430 }, { "grad_norm": 0.06407992541790009, "learning_rate": 7.546540420141479e-06, "loss": 0.004, "step": 75440 }, { "grad_norm": 0.06686537712812424, "learning_rate": 7.5369130629725556e-06, "loss": 0.0051, "step": 75450 }, { "grad_norm": 0.0359320230782032, "learning_rate": 7.527291350092036e-06, "loss": 0.0047, "step": 75460 }, { "grad_norm": 0.0737074464559555, "learning_rate": 7.517675282778897e-06, "loss": 0.007, "step": 75470 }, { "grad_norm": 0.05186541751027107, "learning_rate": 7.508064862311292e-06, "loss": 0.0073, "step": 75480 }, { "grad_norm": 0.028806092217564583, "learning_rate": 7.498460089966669e-06, "loss": 0.0059, "step": 75490 }, { "grad_norm": 0.03820313513278961, "learning_rate": 7.488860967021716e-06, "loss": 0.0044, "step": 75500 }, { "grad_norm": 0.02477964013814926, "learning_rate": 7.479267494752373e-06, "loss": 0.0053, "step": 75510 }, { "grad_norm": 0.05360395833849907, "learning_rate": 7.469679674433822e-06, "loss": 0.0043, "step": 75520 }, { "grad_norm": 0.08551445603370667, "learning_rate": 7.4600975073405025e-06, "loss": 0.0058, "step": 75530 }, { "grad_norm": 0.06188831850886345, "learning_rate": 7.450520994746091e-06, "loss": 0.0062, "step": 75540 }, { "grad_norm": 0.04111146926879883, "learning_rate": 7.4409501379235194e-06, "loss": 0.0058, "step": 75550 }, { "grad_norm": 0.042853713035583496, "learning_rate": 7.4313849381449755e-06, "loss": 0.0037, "step": 75560 }, { "grad_norm": 0.0781339630484581, "learning_rate": 7.421825396681881e-06, "loss": 0.0051, "step": 75570 }, { "grad_norm": 0.029026556760072708, "learning_rate": 7.412271514804914e-06, "loss": 0.0054, "step": 75580 }, { "grad_norm": 0.021592725068330765, "learning_rate": 7.402723293784003e-06, "loss": 0.0038, "step": 75590 }, { "grad_norm": 0.04657074064016342, "learning_rate": 7.393180734888305e-06, "loss": 0.0044, "step": 75600 }, { "grad_norm": 0.06242665648460388, "learning_rate": 7.383643839386234e-06, "loss": 0.0057, "step": 75610 }, { "grad_norm": 0.029661759734153748, "learning_rate": 7.374112608545475e-06, "loss": 0.0045, "step": 75620 }, { "grad_norm": 0.04546399787068367, "learning_rate": 7.364587043632931e-06, "loss": 0.0047, "step": 75630 }, { "grad_norm": 0.0480523407459259, "learning_rate": 7.355067145914779e-06, "loss": 0.0044, "step": 75640 }, { "grad_norm": 0.03800565376877785, "learning_rate": 7.345552916656395e-06, "loss": 0.007, "step": 75650 }, { "grad_norm": 0.05228203535079956, "learning_rate": 7.336044357122446e-06, "loss": 0.005, "step": 75660 }, { "grad_norm": 0.046632543206214905, "learning_rate": 7.3265414685768304e-06, "loss": 0.0047, "step": 75670 }, { "grad_norm": 0.05696051940321922, "learning_rate": 7.317044252282679e-06, "loss": 0.005, "step": 75680 }, { "grad_norm": 0.02749800682067871, "learning_rate": 7.307552709502424e-06, "loss": 0.0041, "step": 75690 }, { "grad_norm": 0.019318057224154472, "learning_rate": 7.298066841497664e-06, "loss": 0.0071, "step": 75700 }, { "grad_norm": 0.03724436089396477, "learning_rate": 7.288586649529294e-06, "loss": 0.0033, "step": 75710 }, { "grad_norm": 0.05696148797869682, "learning_rate": 7.279112134857441e-06, "loss": 0.0047, "step": 75720 }, { "grad_norm": 0.07251472771167755, "learning_rate": 7.26964329874148e-06, "loss": 0.0057, "step": 75730 }, { "grad_norm": 0.033719323575496674, "learning_rate": 7.2601801424400266e-06, "loss": 0.0055, "step": 75740 }, { "grad_norm": 0.031947165727615356, "learning_rate": 7.250722667210952e-06, "loss": 0.0051, "step": 75750 }, { "grad_norm": 0.07008755952119827, "learning_rate": 7.241270874311357e-06, "loss": 0.0068, "step": 75760 }, { "grad_norm": 0.03174985945224762, "learning_rate": 7.231824764997597e-06, "loss": 0.0029, "step": 75770 }, { "grad_norm": 0.03619011119008064, "learning_rate": 7.222384340525279e-06, "loss": 0.0043, "step": 75780 }, { "grad_norm": 0.05215685814619064, "learning_rate": 7.212949602149233e-06, "loss": 0.0051, "step": 75790 }, { "grad_norm": 0.060697924345731735, "learning_rate": 7.20352055112356e-06, "loss": 0.0051, "step": 75800 }, { "grad_norm": 0.04019270837306976, "learning_rate": 7.194097188701571e-06, "loss": 0.0036, "step": 75810 }, { "grad_norm": 0.02169851027429104, "learning_rate": 7.18467951613585e-06, "loss": 0.0042, "step": 75820 }, { "grad_norm": 0.022531798109412193, "learning_rate": 7.175267534678204e-06, "loss": 0.0052, "step": 75830 }, { "grad_norm": 0.0642370730638504, "learning_rate": 7.165861245579719e-06, "loss": 0.009, "step": 75840 }, { "grad_norm": 0.04913169518113136, "learning_rate": 7.156460650090685e-06, "loss": 0.0056, "step": 75850 }, { "grad_norm": 0.038957323879003525, "learning_rate": 7.147065749460663e-06, "loss": 0.0056, "step": 75860 }, { "grad_norm": 0.018924769014120102, "learning_rate": 7.137676544938426e-06, "loss": 0.0059, "step": 75870 }, { "grad_norm": 0.07548098266124725, "learning_rate": 7.1282930377720104e-06, "loss": 0.0035, "step": 75880 }, { "grad_norm": 0.10938473790884018, "learning_rate": 7.1189152292087e-06, "loss": 0.0051, "step": 75890 }, { "grad_norm": 0.03226826712489128, "learning_rate": 7.109543120494999e-06, "loss": 0.0045, "step": 75900 }, { "grad_norm": 0.03909013420343399, "learning_rate": 7.100176712876711e-06, "loss": 0.0067, "step": 75910 }, { "grad_norm": 0.02314167097210884, "learning_rate": 7.090816007598794e-06, "loss": 0.0047, "step": 75920 }, { "grad_norm": 0.044973865151405334, "learning_rate": 7.08146100590551e-06, "loss": 0.0054, "step": 75930 }, { "grad_norm": 0.05114179849624634, "learning_rate": 7.072111709040346e-06, "loss": 0.0065, "step": 75940 }, { "grad_norm": 0.029272574931383133, "learning_rate": 7.062768118246038e-06, "loss": 0.0049, "step": 75950 }, { "grad_norm": 0.028111616149544716, "learning_rate": 7.053430234764547e-06, "loss": 0.006, "step": 75960 }, { "grad_norm": 0.037441425025463104, "learning_rate": 7.0440980598371064e-06, "loss": 0.004, "step": 75970 }, { "grad_norm": 0.020188389346003532, "learning_rate": 7.034771594704126e-06, "loss": 0.0034, "step": 75980 }, { "grad_norm": 0.028369879350066185, "learning_rate": 7.025450840605341e-06, "loss": 0.0039, "step": 75990 }, { "grad_norm": 0.027810223400592804, "learning_rate": 7.016135798779677e-06, "loss": 0.0049, "step": 76000 }, { "grad_norm": 0.022124920040369034, "learning_rate": 7.0068264704653e-06, "loss": 0.0024, "step": 76010 }, { "grad_norm": 0.047623198479413986, "learning_rate": 6.997522856899652e-06, "loss": 0.0054, "step": 76020 }, { "grad_norm": 0.023188890889286995, "learning_rate": 6.988224959319356e-06, "loss": 0.0033, "step": 76030 }, { "grad_norm": 0.06198997050523758, "learning_rate": 6.978932778960329e-06, "loss": 0.0063, "step": 76040 }, { "grad_norm": 0.04875580593943596, "learning_rate": 6.9696463170576875e-06, "loss": 0.0048, "step": 76050 }, { "grad_norm": 0.027772266417741776, "learning_rate": 6.96036557484584e-06, "loss": 0.0037, "step": 76060 }, { "grad_norm": 0.040946099907159805, "learning_rate": 6.951090553558403e-06, "loss": 0.0052, "step": 76070 }, { "grad_norm": 0.0480913370847702, "learning_rate": 6.941821254428204e-06, "loss": 0.0044, "step": 76080 }, { "grad_norm": 0.06700048595666885, "learning_rate": 6.9325576786873566e-06, "loss": 0.0054, "step": 76090 }, { "grad_norm": 0.04094499349594116, "learning_rate": 6.9232998275671925e-06, "loss": 0.0035, "step": 76100 }, { "grad_norm": 0.03192877396941185, "learning_rate": 6.9140477022982876e-06, "loss": 0.0054, "step": 76110 }, { "grad_norm": 0.06267412006855011, "learning_rate": 6.904801304110453e-06, "loss": 0.0054, "step": 76120 }, { "grad_norm": 0.05334009602665901, "learning_rate": 6.895560634232745e-06, "loss": 0.0049, "step": 76130 }, { "grad_norm": 0.024626271799206734, "learning_rate": 6.886325693893447e-06, "loss": 0.0053, "step": 76140 }, { "grad_norm": 0.032392676919698715, "learning_rate": 6.877096484320095e-06, "loss": 0.0071, "step": 76150 }, { "grad_norm": 0.030695674940943718, "learning_rate": 6.867873006739456e-06, "loss": 0.0053, "step": 76160 }, { "grad_norm": 0.07084089517593384, "learning_rate": 6.858655262377528e-06, "loss": 0.0042, "step": 76170 }, { "grad_norm": 0.0379735603928566, "learning_rate": 6.849443252459564e-06, "loss": 0.0048, "step": 76180 }, { "grad_norm": 0.022663798183202744, "learning_rate": 6.840236978210046e-06, "loss": 0.0042, "step": 76190 }, { "grad_norm": 0.036500900983810425, "learning_rate": 6.831036440852667e-06, "loss": 0.0048, "step": 76200 }, { "grad_norm": 0.08443273603916168, "learning_rate": 6.82184164161041e-06, "loss": 0.0039, "step": 76210 }, { "grad_norm": 0.03755684942007065, "learning_rate": 6.812652581705464e-06, "loss": 0.006, "step": 76220 }, { "grad_norm": 0.03613072633743286, "learning_rate": 6.803469262359252e-06, "loss": 0.0041, "step": 76230 }, { "grad_norm": 0.02987617813050747, "learning_rate": 6.794291684792459e-06, "loss": 0.0045, "step": 76240 }, { "grad_norm": 0.06245339661836624, "learning_rate": 6.785119850224964e-06, "loss": 0.0073, "step": 76250 }, { "grad_norm": 0.04331975430250168, "learning_rate": 6.775953759875919e-06, "loss": 0.0046, "step": 76260 }, { "grad_norm": 0.027294985949993134, "learning_rate": 6.76679341496369e-06, "loss": 0.0029, "step": 76270 }, { "grad_norm": 0.043540842831134796, "learning_rate": 6.7576388167059135e-06, "loss": 0.0049, "step": 76280 }, { "grad_norm": 0.07835447043180466, "learning_rate": 6.748489966319438e-06, "loss": 0.0069, "step": 76290 }, { "grad_norm": 0.060136254876852036, "learning_rate": 6.73934686502033e-06, "loss": 0.0049, "step": 76300 }, { "grad_norm": 0.0798502117395401, "learning_rate": 6.730209514023916e-06, "loss": 0.0045, "step": 76310 }, { "grad_norm": 0.03624771162867546, "learning_rate": 6.72107791454476e-06, "loss": 0.0082, "step": 76320 }, { "grad_norm": 0.05695093423128128, "learning_rate": 6.711952067796645e-06, "loss": 0.0036, "step": 76330 }, { "grad_norm": 0.05004732310771942, "learning_rate": 6.702831974992613e-06, "loss": 0.0057, "step": 76340 }, { "grad_norm": 0.04234212636947632, "learning_rate": 6.693717637344915e-06, "loss": 0.0065, "step": 76350 }, { "grad_norm": 0.04003765806555748, "learning_rate": 6.684609056065056e-06, "loss": 0.004, "step": 76360 }, { "grad_norm": 0.0467393733561039, "learning_rate": 6.6755062323637655e-06, "loss": 0.0052, "step": 76370 }, { "grad_norm": 0.041869860142469406, "learning_rate": 6.66640916745101e-06, "loss": 0.0065, "step": 76380 }, { "grad_norm": 0.03665275499224663, "learning_rate": 6.657317862535994e-06, "loss": 0.0039, "step": 76390 }, { "grad_norm": 0.0313953198492527, "learning_rate": 6.648232318827158e-06, "loss": 0.0056, "step": 76400 }, { "grad_norm": 0.03947559744119644, "learning_rate": 6.639152537532173e-06, "loss": 0.0063, "step": 76410 }, { "grad_norm": 0.05297950282692909, "learning_rate": 6.630078519857924e-06, "loss": 0.0059, "step": 76420 }, { "grad_norm": 0.06285605579614639, "learning_rate": 6.621010267010574e-06, "loss": 0.0036, "step": 76430 }, { "grad_norm": 0.026499375700950623, "learning_rate": 6.611947780195488e-06, "loss": 0.0045, "step": 76440 }, { "grad_norm": 0.08140718191862106, "learning_rate": 6.602891060617267e-06, "loss": 0.0062, "step": 76450 }, { "grad_norm": 0.0346200093626976, "learning_rate": 6.593840109479765e-06, "loss": 0.0037, "step": 76460 }, { "grad_norm": 0.04397008568048477, "learning_rate": 6.584794927986038e-06, "loss": 0.0029, "step": 76470 }, { "grad_norm": 0.027545679360628128, "learning_rate": 6.575755517338395e-06, "loss": 0.0046, "step": 76480 }, { "grad_norm": 0.023528926074504852, "learning_rate": 6.566721878738375e-06, "loss": 0.0033, "step": 76490 }, { "grad_norm": 0.056600891053676605, "learning_rate": 6.557694013386745e-06, "loss": 0.0049, "step": 76500 }, { "grad_norm": 0.023359348997473717, "learning_rate": 6.548671922483535e-06, "loss": 0.0041, "step": 76510 }, { "grad_norm": 0.04417835921049118, "learning_rate": 6.539655607227952e-06, "loss": 0.0062, "step": 76520 }, { "grad_norm": 0.042016901075839996, "learning_rate": 6.5306450688184695e-06, "loss": 0.0061, "step": 76530 }, { "grad_norm": 0.03055964969098568, "learning_rate": 6.521640308452798e-06, "loss": 0.0046, "step": 76540 }, { "grad_norm": 0.059102874249219894, "learning_rate": 6.512641327327868e-06, "loss": 0.0052, "step": 76550 }, { "grad_norm": 0.04118349403142929, "learning_rate": 6.503648126639839e-06, "loss": 0.0032, "step": 76560 }, { "grad_norm": 0.024575289338827133, "learning_rate": 6.49466070758411e-06, "loss": 0.0041, "step": 76570 }, { "grad_norm": 0.03922621160745621, "learning_rate": 6.485679071355316e-06, "loss": 0.0053, "step": 76580 }, { "grad_norm": 0.03403659537434578, "learning_rate": 6.476703219147301e-06, "loss": 0.0071, "step": 76590 }, { "grad_norm": 0.10549323260784149, "learning_rate": 6.467733152153171e-06, "loss": 0.0085, "step": 76600 }, { "grad_norm": 0.03240520879626274, "learning_rate": 6.458768871565235e-06, "loss": 0.0031, "step": 76610 }, { "grad_norm": 0.07545827329158783, "learning_rate": 6.4498103785750556e-06, "loss": 0.0061, "step": 76620 }, { "grad_norm": 0.02765350043773651, "learning_rate": 6.440857674373419e-06, "loss": 0.0031, "step": 76630 }, { "grad_norm": 0.06296524405479431, "learning_rate": 6.431910760150317e-06, "loss": 0.0086, "step": 76640 }, { "grad_norm": 0.0633467510342598, "learning_rate": 6.422969637095e-06, "loss": 0.0042, "step": 76650 }, { "grad_norm": 0.04565569385886192, "learning_rate": 6.414034306395955e-06, "loss": 0.0045, "step": 76660 }, { "grad_norm": 0.04517500102519989, "learning_rate": 6.405104769240888e-06, "loss": 0.006, "step": 76670 }, { "grad_norm": 0.05856623873114586, "learning_rate": 6.396181026816728e-06, "loss": 0.0054, "step": 76680 }, { "grad_norm": 0.027812477201223373, "learning_rate": 6.38726308030963e-06, "loss": 0.0045, "step": 76690 }, { "grad_norm": 0.03385591134428978, "learning_rate": 6.378350930904986e-06, "loss": 0.0053, "step": 76700 }, { "grad_norm": 0.031010756269097328, "learning_rate": 6.369444579787431e-06, "loss": 0.0036, "step": 76710 }, { "grad_norm": 0.0357915461063385, "learning_rate": 6.360544028140803e-06, "loss": 0.0039, "step": 76720 }, { "grad_norm": 0.03782500699162483, "learning_rate": 6.351649277148214e-06, "loss": 0.0039, "step": 76730 }, { "grad_norm": 0.03452138230204582, "learning_rate": 6.342760327991937e-06, "loss": 0.0055, "step": 76740 }, { "grad_norm": 0.05017942562699318, "learning_rate": 6.333877181853526e-06, "loss": 0.0034, "step": 76750 }, { "grad_norm": 0.027236448600888252, "learning_rate": 6.324999839913753e-06, "loss": 0.003, "step": 76760 }, { "grad_norm": 0.060576893389225006, "learning_rate": 6.316128303352608e-06, "loss": 0.0059, "step": 76770 }, { "grad_norm": 0.04669351503252983, "learning_rate": 6.307262573349321e-06, "loss": 0.0068, "step": 76780 }, { "grad_norm": 0.03714882582426071, "learning_rate": 6.298402651082336e-06, "loss": 0.0044, "step": 76790 }, { "grad_norm": 0.03434183821082115, "learning_rate": 6.289548537729345e-06, "loss": 0.004, "step": 76800 }, { "grad_norm": 0.030263707041740417, "learning_rate": 6.280700234467246e-06, "loss": 0.0044, "step": 76810 }, { "grad_norm": 0.09337683022022247, "learning_rate": 6.271857742472181e-06, "loss": 0.0037, "step": 76820 }, { "grad_norm": 0.04033391922712326, "learning_rate": 6.2630210629195106e-06, "loss": 0.0046, "step": 76830 }, { "grad_norm": 0.05936197564005852, "learning_rate": 6.254190196983845e-06, "loss": 0.0054, "step": 76840 }, { "grad_norm": 0.030500128865242004, "learning_rate": 6.245365145838971e-06, "loss": 0.0044, "step": 76850 }, { "grad_norm": 0.07431758940219879, "learning_rate": 6.236545910657948e-06, "loss": 0.0067, "step": 76860 }, { "grad_norm": 0.024935532361268997, "learning_rate": 6.227732492613036e-06, "loss": 0.0048, "step": 76870 }, { "grad_norm": 0.033285729587078094, "learning_rate": 6.21892489287576e-06, "loss": 0.0043, "step": 76880 }, { "grad_norm": 0.06044740974903107, "learning_rate": 6.210123112616845e-06, "loss": 0.0086, "step": 76890 }, { "grad_norm": 0.06277193129062653, "learning_rate": 6.20132715300622e-06, "loss": 0.0062, "step": 76900 }, { "grad_norm": 0.06689194589853287, "learning_rate": 6.192537015213079e-06, "loss": 0.0058, "step": 76910 }, { "grad_norm": 0.09280726313591003, "learning_rate": 6.18375270040582e-06, "loss": 0.0049, "step": 76920 }, { "grad_norm": 0.053123630583286285, "learning_rate": 6.174974209752071e-06, "loss": 0.0046, "step": 76930 }, { "grad_norm": 0.07998450100421906, "learning_rate": 6.166201544418693e-06, "loss": 0.0041, "step": 76940 }, { "grad_norm": 0.05555927753448486, "learning_rate": 6.157434705571785e-06, "loss": 0.0053, "step": 76950 }, { "grad_norm": 0.06580611318349838, "learning_rate": 6.148673694376633e-06, "loss": 0.0057, "step": 76960 }, { "grad_norm": 0.0686301738023758, "learning_rate": 6.139918511997772e-06, "loss": 0.0062, "step": 76970 }, { "grad_norm": 0.034096457064151764, "learning_rate": 6.131169159598965e-06, "loss": 0.005, "step": 76980 }, { "grad_norm": 0.02702881209552288, "learning_rate": 6.122425638343199e-06, "loss": 0.0045, "step": 76990 }, { "grad_norm": 0.03271442651748657, "learning_rate": 6.113687949392677e-06, "loss": 0.0044, "step": 77000 }, { "grad_norm": 0.04179936274886131, "learning_rate": 6.1049560939088425e-06, "loss": 0.0048, "step": 77010 }, { "grad_norm": 0.10194947570562363, "learning_rate": 6.096230073052328e-06, "loss": 0.006, "step": 77020 }, { "grad_norm": 0.07940157502889633, "learning_rate": 6.087509887983039e-06, "loss": 0.0053, "step": 77030 }, { "grad_norm": 0.04772323742508888, "learning_rate": 6.078795539860077e-06, "loss": 0.0045, "step": 77040 }, { "grad_norm": 0.04213402420282364, "learning_rate": 6.07008702984177e-06, "loss": 0.0061, "step": 77050 }, { "grad_norm": 0.033269479870796204, "learning_rate": 6.0613843590856865e-06, "loss": 0.0041, "step": 77060 }, { "grad_norm": 0.05863236263394356, "learning_rate": 6.05268752874858e-06, "loss": 0.0049, "step": 77070 }, { "grad_norm": 0.029210137203335762, "learning_rate": 6.0439965399864625e-06, "loss": 0.0046, "step": 77080 }, { "grad_norm": 0.0476003922522068, "learning_rate": 6.035311393954552e-06, "loss": 0.0043, "step": 77090 }, { "grad_norm": 0.05645238235592842, "learning_rate": 6.026632091807321e-06, "loss": 0.0054, "step": 77100 }, { "grad_norm": 0.04738551005721092, "learning_rate": 6.017958634698434e-06, "loss": 0.0062, "step": 77110 }, { "grad_norm": 0.08269540220499039, "learning_rate": 6.0092910237807776e-06, "loss": 0.0057, "step": 77120 }, { "grad_norm": 0.05575118213891983, "learning_rate": 6.000629260206464e-06, "loss": 0.0048, "step": 77130 }, { "grad_norm": 0.03328818455338478, "learning_rate": 5.99197334512685e-06, "loss": 0.006, "step": 77140 }, { "grad_norm": 0.054480113089084625, "learning_rate": 5.983323279692488e-06, "loss": 0.0035, "step": 77150 }, { "grad_norm": 0.05660250410437584, "learning_rate": 5.974679065053168e-06, "loss": 0.0053, "step": 77160 }, { "grad_norm": 0.03720390051603317, "learning_rate": 5.9660407023579e-06, "loss": 0.007, "step": 77170 }, { "grad_norm": 0.029139021411538124, "learning_rate": 5.95740819275491e-06, "loss": 0.0064, "step": 77180 }, { "grad_norm": 0.055984798818826675, "learning_rate": 5.948781537391657e-06, "loss": 0.0038, "step": 77190 }, { "grad_norm": 0.043470315635204315, "learning_rate": 5.940160737414813e-06, "loss": 0.0061, "step": 77200 }, { "grad_norm": 0.0774092897772789, "learning_rate": 5.931545793970272e-06, "loss": 0.0053, "step": 77210 }, { "grad_norm": 0.09806311875581741, "learning_rate": 5.9229367082031476e-06, "loss": 0.0082, "step": 77220 }, { "grad_norm": 0.03990694507956505, "learning_rate": 5.9143334812578e-06, "loss": 0.0039, "step": 77230 }, { "grad_norm": 0.04986606538295746, "learning_rate": 5.905736114277749e-06, "loss": 0.0047, "step": 77240 }, { "grad_norm": 0.01856190338730812, "learning_rate": 5.897144608405808e-06, "loss": 0.0051, "step": 77250 }, { "grad_norm": 0.017562678083777428, "learning_rate": 5.888558964783974e-06, "loss": 0.0025, "step": 77260 }, { "grad_norm": 0.07070863991975784, "learning_rate": 5.8799791845534635e-06, "loss": 0.0053, "step": 77270 }, { "grad_norm": 0.07344222068786621, "learning_rate": 5.871405268854735e-06, "loss": 0.0044, "step": 77280 }, { "grad_norm": 0.03476812317967415, "learning_rate": 5.862837218827433e-06, "loss": 0.0033, "step": 77290 }, { "grad_norm": 0.056106872856616974, "learning_rate": 5.854275035610446e-06, "loss": 0.0071, "step": 77300 }, { "grad_norm": 0.05179232358932495, "learning_rate": 5.845718720341875e-06, "loss": 0.0044, "step": 77310 }, { "grad_norm": 0.03139353543519974, "learning_rate": 5.837168274159061e-06, "loss": 0.0039, "step": 77320 }, { "grad_norm": 0.04725346341729164, "learning_rate": 5.8286236981985485e-06, "loss": 0.0054, "step": 77330 }, { "grad_norm": 0.055853284895420074, "learning_rate": 5.820084993596086e-06, "loss": 0.005, "step": 77340 }, { "grad_norm": 0.036103010177612305, "learning_rate": 5.811552161486661e-06, "loss": 0.0049, "step": 77350 }, { "grad_norm": 0.044242750853300095, "learning_rate": 5.803025203004475e-06, "loss": 0.0054, "step": 77360 }, { "grad_norm": 0.052948080003261566, "learning_rate": 5.794504119282956e-06, "loss": 0.0031, "step": 77370 }, { "grad_norm": 0.058676984161138535, "learning_rate": 5.7859889114547486e-06, "loss": 0.0052, "step": 77380 }, { "grad_norm": 0.030183158814907074, "learning_rate": 5.7774795806517065e-06, "loss": 0.0036, "step": 77390 }, { "grad_norm": 0.04010331258177757, "learning_rate": 5.76897612800491e-06, "loss": 0.0071, "step": 77400 }, { "grad_norm": 0.04660144820809364, "learning_rate": 5.760478554644655e-06, "loss": 0.0037, "step": 77410 }, { "grad_norm": 0.044320832937955856, "learning_rate": 5.751986861700465e-06, "loss": 0.0029, "step": 77420 }, { "grad_norm": 0.05193449929356575, "learning_rate": 5.74350105030107e-06, "loss": 0.0046, "step": 77430 }, { "grad_norm": 0.08573053032159805, "learning_rate": 5.735021121574424e-06, "loss": 0.0039, "step": 77440 }, { "grad_norm": 0.04406895488500595, "learning_rate": 5.726547076647709e-06, "loss": 0.0046, "step": 77450 }, { "grad_norm": 0.07749998569488525, "learning_rate": 5.7180789166472845e-06, "loss": 0.0071, "step": 77460 }, { "grad_norm": 0.0672372430562973, "learning_rate": 5.709616642698784e-06, "loss": 0.0063, "step": 77470 }, { "grad_norm": 0.025454705581068993, "learning_rate": 5.701160255927024e-06, "loss": 0.0046, "step": 77480 }, { "grad_norm": 0.07034289836883545, "learning_rate": 5.692709757456044e-06, "loss": 0.0077, "step": 77490 }, { "grad_norm": 0.05858471617102623, "learning_rate": 5.684265148409113e-06, "loss": 0.0046, "step": 77500 }, { "grad_norm": 0.07922431826591492, "learning_rate": 5.675826429908693e-06, "loss": 0.0068, "step": 77510 }, { "grad_norm": 0.03613445535302162, "learning_rate": 5.667393603076482e-06, "loss": 0.0046, "step": 77520 }, { "grad_norm": 0.059681039303541183, "learning_rate": 5.658966669033394e-06, "loss": 0.005, "step": 77530 }, { "grad_norm": 0.021854180842638016, "learning_rate": 5.650545628899545e-06, "loss": 0.0026, "step": 77540 }, { "grad_norm": 0.04186895489692688, "learning_rate": 5.6421304837943054e-06, "loss": 0.0032, "step": 77550 }, { "grad_norm": 0.03619854897260666, "learning_rate": 5.633721234836203e-06, "loss": 0.0039, "step": 77560 }, { "grad_norm": 0.053694888949394226, "learning_rate": 5.625317883143033e-06, "loss": 0.0041, "step": 77570 }, { "grad_norm": 0.03132740780711174, "learning_rate": 5.616920429831779e-06, "loss": 0.0044, "step": 77580 }, { "grad_norm": 0.04922398179769516, "learning_rate": 5.608528876018659e-06, "loss": 0.0038, "step": 77590 }, { "grad_norm": 0.019231600686907768, "learning_rate": 5.600143222819082e-06, "loss": 0.0049, "step": 77600 }, { "grad_norm": 0.05452171713113785, "learning_rate": 5.591763471347705e-06, "loss": 0.0053, "step": 77610 }, { "grad_norm": 0.04908487945795059, "learning_rate": 5.583389622718372e-06, "loss": 0.0045, "step": 77620 }, { "grad_norm": 0.026803428307175636, "learning_rate": 5.575021678044157e-06, "loss": 0.0074, "step": 77630 }, { "grad_norm": 0.02365250326693058, "learning_rate": 5.566659638437343e-06, "loss": 0.0041, "step": 77640 }, { "grad_norm": 0.032731421291828156, "learning_rate": 5.558303505009433e-06, "loss": 0.0064, "step": 77650 }, { "grad_norm": 0.03371419385075569, "learning_rate": 5.549953278871145e-06, "loss": 0.0074, "step": 77660 }, { "grad_norm": 0.01872422732412815, "learning_rate": 5.541608961132422e-06, "loss": 0.0039, "step": 77670 }, { "grad_norm": 0.05092547461390495, "learning_rate": 5.5332705529023806e-06, "loss": 0.0055, "step": 77680 }, { "grad_norm": 0.034678470343351364, "learning_rate": 5.524938055289386e-06, "loss": 0.0046, "step": 77690 }, { "grad_norm": 0.030852336436510086, "learning_rate": 5.516611469401034e-06, "loss": 0.0088, "step": 77700 }, { "grad_norm": 0.06650891155004501, "learning_rate": 5.508290796344101e-06, "loss": 0.0036, "step": 77710 }, { "grad_norm": 0.020132241770625114, "learning_rate": 5.499976037224603e-06, "loss": 0.0038, "step": 77720 }, { "grad_norm": 0.07059924304485321, "learning_rate": 5.491667193147726e-06, "loss": 0.0097, "step": 77730 }, { "grad_norm": 0.05921740457415581, "learning_rate": 5.483364265217921e-06, "loss": 0.0036, "step": 77740 }, { "grad_norm": 0.04243317246437073, "learning_rate": 5.475067254538829e-06, "loss": 0.0049, "step": 77750 }, { "grad_norm": 0.030951092019677162, "learning_rate": 5.4667761622132884e-06, "loss": 0.0063, "step": 77760 }, { "grad_norm": 0.05676986649632454, "learning_rate": 5.458490989343417e-06, "loss": 0.0047, "step": 77770 }, { "grad_norm": 0.045945215970277786, "learning_rate": 5.450211737030453e-06, "loss": 0.0054, "step": 77780 }, { "grad_norm": 0.03712261840701103, "learning_rate": 5.441938406374908e-06, "loss": 0.0062, "step": 77790 }, { "grad_norm": 0.03527211770415306, "learning_rate": 5.433670998476498e-06, "loss": 0.0033, "step": 77800 }, { "grad_norm": 0.04956187680363655, "learning_rate": 5.425409514434138e-06, "loss": 0.0044, "step": 77810 }, { "grad_norm": 0.032543737441301346, "learning_rate": 5.417153955345961e-06, "loss": 0.0054, "step": 77820 }, { "grad_norm": 0.07853583246469498, "learning_rate": 5.408904322309321e-06, "loss": 0.0039, "step": 77830 }, { "grad_norm": 0.04161381721496582, "learning_rate": 5.400660616420772e-06, "loss": 0.0038, "step": 77840 }, { "grad_norm": 0.04363355040550232, "learning_rate": 5.39242283877609e-06, "loss": 0.005, "step": 77850 }, { "grad_norm": 0.02939414046704769, "learning_rate": 5.3841909904702614e-06, "loss": 0.0054, "step": 77860 }, { "grad_norm": 0.036492086946964264, "learning_rate": 5.375965072597472e-06, "loss": 0.006, "step": 77870 }, { "grad_norm": 0.028625786304473877, "learning_rate": 5.367745086251147e-06, "loss": 0.0048, "step": 77880 }, { "grad_norm": 0.03255963325500488, "learning_rate": 5.359531032523885e-06, "loss": 0.0046, "step": 77890 }, { "grad_norm": 0.029240304604172707, "learning_rate": 5.351322912507523e-06, "loss": 0.0039, "step": 77900 }, { "grad_norm": 0.021024532616138458, "learning_rate": 5.343120727293094e-06, "loss": 0.0035, "step": 77910 }, { "grad_norm": 0.07678341865539551, "learning_rate": 5.334924477970876e-06, "loss": 0.0051, "step": 77920 }, { "grad_norm": 0.03276742622256279, "learning_rate": 5.326734165630321e-06, "loss": 0.0028, "step": 77930 }, { "grad_norm": 0.046725861728191376, "learning_rate": 5.318549791360095e-06, "loss": 0.0035, "step": 77940 }, { "grad_norm": 0.03636466711759567, "learning_rate": 5.310371356248089e-06, "loss": 0.0039, "step": 77950 }, { "grad_norm": 0.03664112836122513, "learning_rate": 5.302198861381397e-06, "loss": 0.0034, "step": 77960 }, { "grad_norm": 0.03671417012810707, "learning_rate": 5.294032307846331e-06, "loss": 0.0046, "step": 77970 }, { "grad_norm": 0.046201616525650024, "learning_rate": 5.285871696728389e-06, "loss": 0.0092, "step": 77980 }, { "grad_norm": 0.062112871557474136, "learning_rate": 5.277717029112334e-06, "loss": 0.0043, "step": 77990 }, { "grad_norm": 0.07617311924695969, "learning_rate": 5.269568306082068e-06, "loss": 0.0059, "step": 78000 }, { "grad_norm": 0.053292952477931976, "learning_rate": 5.2614255287207535e-06, "loss": 0.0044, "step": 78010 }, { "grad_norm": 0.0519927442073822, "learning_rate": 5.2532886981107475e-06, "loss": 0.0033, "step": 78020 }, { "grad_norm": 0.051634352654218674, "learning_rate": 5.245157815333607e-06, "loss": 0.0056, "step": 78030 }, { "grad_norm": 0.07710522413253784, "learning_rate": 5.237032881470111e-06, "loss": 0.0044, "step": 78040 }, { "grad_norm": 0.05155012756586075, "learning_rate": 5.2289138976002575e-06, "loss": 0.0048, "step": 78050 }, { "grad_norm": 0.03600061312317848, "learning_rate": 5.220800864803205e-06, "loss": 0.0049, "step": 78060 }, { "grad_norm": 0.03967643156647682, "learning_rate": 5.212693784157391e-06, "loss": 0.0042, "step": 78070 }, { "grad_norm": 0.03446853905916214, "learning_rate": 5.204592656740409e-06, "loss": 0.0039, "step": 78080 }, { "grad_norm": 0.03411877527832985, "learning_rate": 5.196497483629087e-06, "loss": 0.0044, "step": 78090 }, { "grad_norm": 0.04424538463354111, "learning_rate": 5.188408265899463e-06, "loss": 0.005, "step": 78100 }, { "grad_norm": 0.05277180299162865, "learning_rate": 5.180325004626752e-06, "loss": 0.0034, "step": 78110 }, { "grad_norm": 0.025326620787382126, "learning_rate": 5.172247700885402e-06, "loss": 0.004, "step": 78120 }, { "grad_norm": 0.027682781219482422, "learning_rate": 5.164176355749067e-06, "loss": 0.0044, "step": 78130 }, { "grad_norm": 0.05265122279524803, "learning_rate": 5.156110970290623e-06, "loss": 0.0052, "step": 78140 }, { "grad_norm": 0.1125103011727333, "learning_rate": 5.148051545582138e-06, "loss": 0.003, "step": 78150 }, { "grad_norm": 0.028302378952503204, "learning_rate": 5.139998082694875e-06, "loss": 0.0041, "step": 78160 }, { "grad_norm": 0.04572167247533798, "learning_rate": 5.131950582699324e-06, "loss": 0.0043, "step": 78170 }, { "grad_norm": 0.03670794144272804, "learning_rate": 5.123909046665176e-06, "loss": 0.0034, "step": 78180 }, { "grad_norm": 0.04139126464724541, "learning_rate": 5.115873475661331e-06, "loss": 0.0037, "step": 78190 }, { "grad_norm": 0.028902631253004074, "learning_rate": 5.107843870755891e-06, "loss": 0.0034, "step": 78200 }, { "grad_norm": 0.05716776102781296, "learning_rate": 5.099820233016178e-06, "loss": 0.0045, "step": 78210 }, { "grad_norm": 0.05554571747779846, "learning_rate": 5.091802563508702e-06, "loss": 0.0048, "step": 78220 }, { "grad_norm": 0.058518167585134506, "learning_rate": 5.083790863299198e-06, "loss": 0.0048, "step": 78230 }, { "grad_norm": 0.04395858570933342, "learning_rate": 5.075785133452593e-06, "loss": 0.0048, "step": 78240 }, { "grad_norm": 0.036118511110544205, "learning_rate": 5.067785375033029e-06, "loss": 0.0034, "step": 78250 }, { "grad_norm": 0.04634775221347809, "learning_rate": 5.0597915891038514e-06, "loss": 0.0046, "step": 78260 }, { "grad_norm": 0.01870211958885193, "learning_rate": 5.051803776727626e-06, "loss": 0.0036, "step": 78270 }, { "grad_norm": 0.015699302777647972, "learning_rate": 5.0438219389660754e-06, "loss": 0.0066, "step": 78280 }, { "grad_norm": 0.033051397651433945, "learning_rate": 5.035846076880191e-06, "loss": 0.0063, "step": 78290 }, { "grad_norm": 0.03875492885708809, "learning_rate": 5.02787619153014e-06, "loss": 0.0043, "step": 78300 }, { "grad_norm": 0.0545818954706192, "learning_rate": 5.019912283975292e-06, "loss": 0.0047, "step": 78310 }, { "grad_norm": 0.0806768462061882, "learning_rate": 5.01195435527424e-06, "loss": 0.004, "step": 78320 }, { "grad_norm": 0.03520932421088219, "learning_rate": 5.004002406484748e-06, "loss": 0.0049, "step": 78330 }, { "grad_norm": 0.036521729081869125, "learning_rate": 4.9960564386638156e-06, "loss": 0.0046, "step": 78340 }, { "grad_norm": 0.030630964785814285, "learning_rate": 4.988116452867631e-06, "loss": 0.0049, "step": 78350 }, { "grad_norm": 0.046664848923683167, "learning_rate": 4.980182450151616e-06, "loss": 0.0062, "step": 78360 }, { "grad_norm": 0.0316145122051239, "learning_rate": 4.972254431570373e-06, "loss": 0.0041, "step": 78370 }, { "grad_norm": 0.039593107998371124, "learning_rate": 4.96433239817769e-06, "loss": 0.0046, "step": 78380 }, { "grad_norm": 0.018045438453555107, "learning_rate": 4.9564163510266e-06, "loss": 0.0041, "step": 78390 }, { "grad_norm": 0.041997913271188736, "learning_rate": 4.948506291169314e-06, "loss": 0.0045, "step": 78400 }, { "grad_norm": 0.04351992532610893, "learning_rate": 4.940602219657253e-06, "loss": 0.0071, "step": 78410 }, { "grad_norm": 0.04930026829242706, "learning_rate": 4.932704137541055e-06, "loss": 0.0045, "step": 78420 }, { "grad_norm": 0.026008613407611847, "learning_rate": 4.924812045870536e-06, "loss": 0.0052, "step": 78430 }, { "grad_norm": 0.018490992486476898, "learning_rate": 4.916925945694745e-06, "loss": 0.0064, "step": 78440 }, { "grad_norm": 0.04164555296301842, "learning_rate": 4.90904583806191e-06, "loss": 0.0044, "step": 78450 }, { "grad_norm": 0.056280992925167084, "learning_rate": 4.901171724019477e-06, "loss": 0.0048, "step": 78460 }, { "grad_norm": 0.049799174070358276, "learning_rate": 4.8933036046140855e-06, "loss": 0.0037, "step": 78470 }, { "grad_norm": 0.03865492343902588, "learning_rate": 4.885441480891589e-06, "loss": 0.0047, "step": 78480 }, { "grad_norm": 0.050663433969020844, "learning_rate": 4.877585353897051e-06, "loss": 0.0032, "step": 78490 }, { "grad_norm": 0.02871277555823326, "learning_rate": 4.869735224674687e-06, "loss": 0.0059, "step": 78500 }, { "grad_norm": 0.041244685649871826, "learning_rate": 4.8618910942679894e-06, "loss": 0.0061, "step": 78510 }, { "grad_norm": 0.04565625265240669, "learning_rate": 4.854052963719607e-06, "loss": 0.0047, "step": 78520 }, { "grad_norm": 0.07787472754716873, "learning_rate": 4.846220834071402e-06, "loss": 0.0053, "step": 78530 }, { "grad_norm": 0.030867667868733406, "learning_rate": 4.838394706364446e-06, "loss": 0.007, "step": 78540 }, { "grad_norm": 0.026064811274409294, "learning_rate": 4.8305745816389855e-06, "loss": 0.0046, "step": 78550 }, { "grad_norm": 0.04224718362092972, "learning_rate": 4.822760460934506e-06, "loss": 0.0036, "step": 78560 }, { "grad_norm": 0.03230265527963638, "learning_rate": 4.814952345289669e-06, "loss": 0.0048, "step": 78570 }, { "grad_norm": 0.02655969187617302, "learning_rate": 4.807150235742336e-06, "loss": 0.0033, "step": 78580 }, { "grad_norm": 0.031072841957211494, "learning_rate": 4.79935413332962e-06, "loss": 0.005, "step": 78590 }, { "grad_norm": 0.05384165793657303, "learning_rate": 4.791564039087759e-06, "loss": 0.0036, "step": 78600 }, { "grad_norm": 0.04366093501448631, "learning_rate": 4.783779954052242e-06, "loss": 0.0106, "step": 78610 }, { "grad_norm": 0.020918654277920723, "learning_rate": 4.7760018792577464e-06, "loss": 0.0069, "step": 78620 }, { "grad_norm": 0.02499304711818695, "learning_rate": 4.768229815738151e-06, "loss": 0.0049, "step": 78630 }, { "grad_norm": 0.05831701681017876, "learning_rate": 4.76046376452654e-06, "loss": 0.007, "step": 78640 }, { "grad_norm": 0.034956056624650955, "learning_rate": 4.752703726655189e-06, "loss": 0.0055, "step": 78650 }, { "grad_norm": 0.019993310794234276, "learning_rate": 4.744949703155583e-06, "loss": 0.004, "step": 78660 }, { "grad_norm": 0.030898552387952805, "learning_rate": 4.737201695058402e-06, "loss": 0.0061, "step": 78670 }, { "grad_norm": 0.036016978323459625, "learning_rate": 4.729459703393524e-06, "loss": 0.0052, "step": 78680 }, { "grad_norm": 0.05364121496677399, "learning_rate": 4.721723729190042e-06, "loss": 0.0047, "step": 78690 }, { "grad_norm": 0.021172555163502693, "learning_rate": 4.713993773476244e-06, "loss": 0.0054, "step": 78700 }, { "grad_norm": 0.017980096861720085, "learning_rate": 4.706269837279592e-06, "loss": 0.0072, "step": 78710 }, { "grad_norm": 0.06841576844453812, "learning_rate": 4.698551921626781e-06, "loss": 0.0063, "step": 78720 }, { "grad_norm": 0.052944157272577286, "learning_rate": 4.6908400275436805e-06, "loss": 0.0069, "step": 78730 }, { "grad_norm": 0.039042916148900986, "learning_rate": 4.683134156055397e-06, "loss": 0.0066, "step": 78740 }, { "grad_norm": 0.030366593971848488, "learning_rate": 4.6754343081861995e-06, "loss": 0.0045, "step": 78750 }, { "grad_norm": 0.0263153538107872, "learning_rate": 4.667740484959577e-06, "loss": 0.005, "step": 78760 }, { "grad_norm": 0.06774482876062393, "learning_rate": 4.660052687398192e-06, "loss": 0.0058, "step": 78770 }, { "grad_norm": 0.09021540731191635, "learning_rate": 4.652370916523935e-06, "loss": 0.0045, "step": 78780 }, { "grad_norm": 0.07959743589162827, "learning_rate": 4.644695173357882e-06, "loss": 0.0048, "step": 78790 }, { "grad_norm": 0.017431939020752907, "learning_rate": 4.637025458920302e-06, "loss": 0.0046, "step": 78800 }, { "grad_norm": 0.06658395379781723, "learning_rate": 4.6293617742307e-06, "loss": 0.0064, "step": 78810 }, { "grad_norm": 0.057872604578733444, "learning_rate": 4.6217041203077184e-06, "loss": 0.0058, "step": 78820 }, { "grad_norm": 0.056231338530778885, "learning_rate": 4.614052498169241e-06, "loss": 0.0056, "step": 78830 }, { "grad_norm": 0.042593762278556824, "learning_rate": 4.606406908832333e-06, "loss": 0.0052, "step": 78840 }, { "grad_norm": 0.047211624681949615, "learning_rate": 4.598767353313277e-06, "loss": 0.0045, "step": 78850 }, { "grad_norm": 0.023977164179086685, "learning_rate": 4.591133832627525e-06, "loss": 0.0037, "step": 78860 }, { "grad_norm": 0.041615184396505356, "learning_rate": 4.583506347789751e-06, "loss": 0.0059, "step": 78870 }, { "grad_norm": 0.036775000393390656, "learning_rate": 4.575884899813815e-06, "loss": 0.0035, "step": 78880 }, { "grad_norm": 0.051958758383989334, "learning_rate": 4.568269489712773e-06, "loss": 0.0052, "step": 78890 }, { "grad_norm": 0.047770895063877106, "learning_rate": 4.560660118498889e-06, "loss": 0.0064, "step": 78900 }, { "grad_norm": 0.05535177141427994, "learning_rate": 4.553056787183607e-06, "loss": 0.0062, "step": 78910 }, { "grad_norm": 0.026907365769147873, "learning_rate": 4.545459496777605e-06, "loss": 0.0038, "step": 78920 }, { "grad_norm": 0.03757676109671593, "learning_rate": 4.537868248290694e-06, "loss": 0.0039, "step": 78930 }, { "grad_norm": 0.041611701250076294, "learning_rate": 4.530283042731947e-06, "loss": 0.0041, "step": 78940 }, { "grad_norm": 0.03478601947426796, "learning_rate": 4.522703881109586e-06, "loss": 0.0046, "step": 78950 }, { "grad_norm": 0.044989921152591705, "learning_rate": 4.515130764431069e-06, "loss": 0.0037, "step": 78960 }, { "grad_norm": 0.042583633214235306, "learning_rate": 4.507563693703032e-06, "loss": 0.0045, "step": 78970 }, { "grad_norm": 0.023142134770751, "learning_rate": 4.500002669931291e-06, "loss": 0.0036, "step": 78980 }, { "grad_norm": 0.02673116698861122, "learning_rate": 4.4924476941208885e-06, "loss": 0.0036, "step": 78990 }, { "grad_norm": 0.02658468671143055, "learning_rate": 4.484898767276036e-06, "loss": 0.004, "step": 79000 }, { "grad_norm": 0.10736721754074097, "learning_rate": 4.477355890400164e-06, "loss": 0.0061, "step": 79010 }, { "grad_norm": 0.05638361722230911, "learning_rate": 4.469819064495884e-06, "loss": 0.0049, "step": 79020 }, { "grad_norm": 0.02931906469166279, "learning_rate": 4.462288290565014e-06, "loss": 0.0039, "step": 79030 }, { "grad_norm": 0.02911829575896263, "learning_rate": 4.4547635696085495e-06, "loss": 0.0044, "step": 79040 }, { "grad_norm": 0.03672976419329643, "learning_rate": 4.447244902626702e-06, "loss": 0.0064, "step": 79050 }, { "grad_norm": 0.03233041241765022, "learning_rate": 4.43973229061887e-06, "loss": 0.0052, "step": 79060 }, { "grad_norm": 0.02573045715689659, "learning_rate": 4.43222573458364e-06, "loss": 0.0075, "step": 79070 }, { "grad_norm": 0.06735765188932419, "learning_rate": 4.424725235518806e-06, "loss": 0.0053, "step": 79080 }, { "grad_norm": 0.03924836963415146, "learning_rate": 4.417230794421362e-06, "loss": 0.0032, "step": 79090 }, { "grad_norm": 0.03168421983718872, "learning_rate": 4.409742412287455e-06, "loss": 0.0048, "step": 79100 }, { "grad_norm": 0.022300129756331444, "learning_rate": 4.402260090112481e-06, "loss": 0.0065, "step": 79110 }, { "grad_norm": 0.021449124440550804, "learning_rate": 4.3947838288910025e-06, "loss": 0.0043, "step": 79120 }, { "grad_norm": 0.04906446859240532, "learning_rate": 4.3873136296167794e-06, "loss": 0.0042, "step": 79130 }, { "grad_norm": 0.04349445551633835, "learning_rate": 4.379849493282778e-06, "loss": 0.0089, "step": 79140 }, { "grad_norm": 0.01943458616733551, "learning_rate": 4.372391420881133e-06, "loss": 0.0054, "step": 79150 }, { "grad_norm": 0.029863696545362473, "learning_rate": 4.36493941340319e-06, "loss": 0.0026, "step": 79160 }, { "grad_norm": 0.03878837451338768, "learning_rate": 4.35749347183948e-06, "loss": 0.0047, "step": 79170 }, { "grad_norm": 0.06260888278484344, "learning_rate": 4.350053597179754e-06, "loss": 0.0034, "step": 79180 }, { "grad_norm": 0.04093259572982788, "learning_rate": 4.3426197904129325e-06, "loss": 0.005, "step": 79190 }, { "grad_norm": 0.03889314830303192, "learning_rate": 4.3351920525271204e-06, "loss": 0.0065, "step": 79200 }, { "grad_norm": 0.061890773475170135, "learning_rate": 4.327770384509638e-06, "loss": 0.0052, "step": 79210 }, { "grad_norm": 0.044927362352609634, "learning_rate": 4.320354787346992e-06, "loss": 0.0052, "step": 79220 }, { "grad_norm": 0.032034311443567276, "learning_rate": 4.31294526202487e-06, "loss": 0.0049, "step": 79230 }, { "grad_norm": 0.05133732780814171, "learning_rate": 4.305541809528174e-06, "loss": 0.0073, "step": 79240 }, { "grad_norm": 0.04087623208761215, "learning_rate": 4.298144430840978e-06, "loss": 0.0032, "step": 79250 }, { "grad_norm": 0.027480311691761017, "learning_rate": 4.290753126946567e-06, "loss": 0.0057, "step": 79260 }, { "grad_norm": 0.027368921786546707, "learning_rate": 4.283367898827406e-06, "loss": 0.0042, "step": 79270 }, { "grad_norm": 0.0729445144534111, "learning_rate": 4.275988747465154e-06, "loss": 0.0046, "step": 79280 }, { "grad_norm": 0.06007622554898262, "learning_rate": 4.268615673840665e-06, "loss": 0.0059, "step": 79290 }, { "grad_norm": 0.03774746134877205, "learning_rate": 4.261248678933988e-06, "loss": 0.0057, "step": 79300 }, { "grad_norm": 0.024452485144138336, "learning_rate": 4.253887763724362e-06, "loss": 0.0055, "step": 79310 }, { "grad_norm": 0.05963724106550217, "learning_rate": 4.2465329291902e-06, "loss": 0.0049, "step": 79320 }, { "grad_norm": 0.05606890842318535, "learning_rate": 4.239184176309141e-06, "loss": 0.0049, "step": 79330 }, { "grad_norm": 0.03531128913164139, "learning_rate": 4.231841506057993e-06, "loss": 0.004, "step": 79340 }, { "grad_norm": 0.024526435881853104, "learning_rate": 4.224504919412758e-06, "loss": 0.0036, "step": 79350 }, { "grad_norm": 0.03769420087337494, "learning_rate": 4.21717441734864e-06, "loss": 0.0068, "step": 79360 }, { "grad_norm": 0.0676422044634819, "learning_rate": 4.20985000084001e-06, "loss": 0.0063, "step": 79370 }, { "grad_norm": 0.04566363990306854, "learning_rate": 4.202531670860455e-06, "loss": 0.009, "step": 79380 }, { "grad_norm": 0.032911546528339386, "learning_rate": 4.195219428382735e-06, "loss": 0.0066, "step": 79390 }, { "grad_norm": 0.042822111397981644, "learning_rate": 4.187913274378824e-06, "loss": 0.0049, "step": 79400 }, { "grad_norm": 0.03542497754096985, "learning_rate": 4.180613209819872e-06, "loss": 0.0049, "step": 79410 }, { "grad_norm": 0.03775244578719139, "learning_rate": 4.17331923567621e-06, "loss": 0.0052, "step": 79420 }, { "grad_norm": 0.019621269777417183, "learning_rate": 4.166031352917371e-06, "loss": 0.0041, "step": 79430 }, { "grad_norm": 0.026025034487247467, "learning_rate": 4.158749562512077e-06, "loss": 0.0041, "step": 79440 }, { "grad_norm": 0.07561201602220535, "learning_rate": 4.151473865428241e-06, "loss": 0.0051, "step": 79450 }, { "grad_norm": 0.03529367595911026, "learning_rate": 4.144204262632967e-06, "loss": 0.0056, "step": 79460 }, { "grad_norm": 0.026784786954522133, "learning_rate": 4.136940755092544e-06, "loss": 0.0053, "step": 79470 }, { "grad_norm": 0.09852053970098495, "learning_rate": 4.129683343772456e-06, "loss": 0.0055, "step": 79480 }, { "grad_norm": 0.040388744324445724, "learning_rate": 4.122432029637374e-06, "loss": 0.0045, "step": 79490 }, { "grad_norm": 0.040356870740652084, "learning_rate": 4.115186813651156e-06, "loss": 0.0031, "step": 79500 }, { "grad_norm": 0.03307368978857994, "learning_rate": 4.107947696776854e-06, "loss": 0.0049, "step": 79510 }, { "grad_norm": 0.034798700362443924, "learning_rate": 4.1007146799767086e-06, "loss": 0.005, "step": 79520 }, { "grad_norm": 0.057006143033504486, "learning_rate": 4.093487764212161e-06, "loss": 0.0054, "step": 79530 }, { "grad_norm": 0.024134546518325806, "learning_rate": 4.08626695044379e-06, "loss": 0.0054, "step": 79540 }, { "grad_norm": 0.031148651614785194, "learning_rate": 4.0790522396314424e-06, "loss": 0.0042, "step": 79550 }, { "grad_norm": 0.02390776015818119, "learning_rate": 4.071843632734101e-06, "loss": 0.0048, "step": 79560 }, { "grad_norm": 0.04094658046960831, "learning_rate": 4.064641130709945e-06, "loss": 0.006, "step": 79570 }, { "grad_norm": 0.028769446536898613, "learning_rate": 4.057444734516358e-06, "loss": 0.0066, "step": 79580 }, { "grad_norm": 0.04536861181259155, "learning_rate": 4.050254445109891e-06, "loss": 0.0045, "step": 79590 }, { "grad_norm": 0.02489742450416088, "learning_rate": 4.043070263446291e-06, "loss": 0.0065, "step": 79600 }, { "grad_norm": 0.050165798515081406, "learning_rate": 4.0358921904805e-06, "loss": 0.0046, "step": 79610 }, { "grad_norm": 0.03554569184780121, "learning_rate": 4.028720227166638e-06, "loss": 0.0035, "step": 79620 }, { "grad_norm": 0.03681721165776253, "learning_rate": 4.021554374458036e-06, "loss": 0.0063, "step": 79630 }, { "grad_norm": 0.045328471809625626, "learning_rate": 4.014394633307178e-06, "loss": 0.011, "step": 79640 }, { "grad_norm": 0.028801215812563896, "learning_rate": 4.007241004665762e-06, "loss": 0.0043, "step": 79650 }, { "grad_norm": 0.06494829803705215, "learning_rate": 4.0000934894846575e-06, "loss": 0.0047, "step": 79660 }, { "grad_norm": 0.049515802413225174, "learning_rate": 3.992952088713931e-06, "loss": 0.0037, "step": 79670 }, { "grad_norm": 0.03169627487659454, "learning_rate": 3.985816803302833e-06, "loss": 0.0045, "step": 79680 }, { "grad_norm": 0.03739927336573601, "learning_rate": 3.978687634199801e-06, "loss": 0.0062, "step": 79690 }, { "grad_norm": 0.03235626220703125, "learning_rate": 3.971564582352466e-06, "loss": 0.0051, "step": 79700 }, { "grad_norm": 0.06279649585485458, "learning_rate": 3.964447648707631e-06, "loss": 0.0048, "step": 79710 }, { "grad_norm": 0.04691339656710625, "learning_rate": 3.957336834211301e-06, "loss": 0.0035, "step": 79720 }, { "grad_norm": 0.028268801048398018, "learning_rate": 3.950232139808657e-06, "loss": 0.0041, "step": 79730 }, { "grad_norm": 0.03642566502094269, "learning_rate": 3.9431335664440845e-06, "loss": 0.0043, "step": 79740 }, { "grad_norm": 0.06645052134990692, "learning_rate": 3.936041115061123e-06, "loss": 0.0075, "step": 79750 }, { "grad_norm": 0.02754032425582409, "learning_rate": 3.9289547866025234e-06, "loss": 0.0043, "step": 79760 }, { "grad_norm": 0.06240328773856163, "learning_rate": 3.921874582010204e-06, "loss": 0.005, "step": 79770 }, { "grad_norm": 0.029932202771306038, "learning_rate": 3.914800502225308e-06, "loss": 0.0043, "step": 79780 }, { "grad_norm": 0.03739093244075775, "learning_rate": 3.907732548188137e-06, "loss": 0.006, "step": 79790 }, { "grad_norm": 0.026914680376648903, "learning_rate": 3.900670720838151e-06, "loss": 0.0042, "step": 79800 }, { "grad_norm": 0.04150897264480591, "learning_rate": 3.893615021114044e-06, "loss": 0.0054, "step": 79810 }, { "grad_norm": 0.04714987054467201, "learning_rate": 3.8865654499536715e-06, "loss": 0.0034, "step": 79820 }, { "grad_norm": 0.04911576583981514, "learning_rate": 3.879522008294073e-06, "loss": 0.0037, "step": 79830 }, { "grad_norm": 0.03121614083647728, "learning_rate": 3.8724846970714824e-06, "loss": 0.0034, "step": 79840 }, { "grad_norm": 0.023580016568303108, "learning_rate": 3.865453517221323e-06, "loss": 0.0029, "step": 79850 }, { "grad_norm": 0.06158105283975601, "learning_rate": 3.858428469678188e-06, "loss": 0.0063, "step": 79860 }, { "grad_norm": 0.03461684659123421, "learning_rate": 3.8514095553758625e-06, "loss": 0.0026, "step": 79870 }, { "grad_norm": 0.023646749556064606, "learning_rate": 3.844396775247311e-06, "loss": 0.0043, "step": 79880 }, { "grad_norm": 0.04972982034087181, "learning_rate": 3.8373901302246936e-06, "loss": 0.0031, "step": 79890 }, { "grad_norm": 0.023720933124423027, "learning_rate": 3.83038962123935e-06, "loss": 0.0046, "step": 79900 }, { "grad_norm": 0.0400337316095829, "learning_rate": 3.8233952492218e-06, "loss": 0.005, "step": 79910 }, { "grad_norm": 0.04808149114251137, "learning_rate": 3.8164070151017584e-06, "loss": 0.0043, "step": 79920 }, { "grad_norm": 0.031102126464247704, "learning_rate": 3.8094249198081077e-06, "loss": 0.0056, "step": 79930 }, { "grad_norm": 0.061754729598760605, "learning_rate": 3.8024489642689287e-06, "loss": 0.0043, "step": 79940 }, { "grad_norm": 0.05487218499183655, "learning_rate": 3.7954791494114793e-06, "loss": 0.0055, "step": 79950 }, { "grad_norm": 0.02181941084563732, "learning_rate": 3.7885154761622077e-06, "loss": 0.0052, "step": 79960 }, { "grad_norm": 0.046842463314533234, "learning_rate": 3.7815579454467342e-06, "loss": 0.0066, "step": 79970 }, { "grad_norm": 0.03301732987165451, "learning_rate": 3.7746065581898693e-06, "loss": 0.0035, "step": 79980 }, { "grad_norm": 0.08667956292629242, "learning_rate": 3.7676613153156014e-06, "loss": 0.0086, "step": 79990 }, { "grad_norm": 0.04685444384813309, "learning_rate": 3.7607222177471257e-06, "loss": 0.0058, "step": 80000 }, { "grad_norm": 0.028416389599442482, "learning_rate": 3.7537892664067988e-06, "loss": 0.0043, "step": 80010 }, { "grad_norm": 0.046636201441287994, "learning_rate": 3.7468624622161553e-06, "loss": 0.0056, "step": 80020 }, { "grad_norm": 0.022297251969575882, "learning_rate": 3.7399418060959144e-06, "loss": 0.004, "step": 80030 }, { "grad_norm": 0.038566213101148605, "learning_rate": 3.733027298966002e-06, "loss": 0.0044, "step": 80040 }, { "grad_norm": 0.0463092066347599, "learning_rate": 3.726118941745499e-06, "loss": 0.0062, "step": 80050 }, { "grad_norm": 0.09851107746362686, "learning_rate": 3.7192167353526873e-06, "loss": 0.006, "step": 80060 }, { "grad_norm": 0.05951541289687157, "learning_rate": 3.712320680705017e-06, "loss": 0.005, "step": 80070 }, { "grad_norm": 0.024944622069597244, "learning_rate": 3.7054307787191324e-06, "loss": 0.0049, "step": 80080 }, { "grad_norm": 0.047860804945230484, "learning_rate": 3.698547030310845e-06, "loss": 0.006, "step": 80090 }, { "grad_norm": 0.02299089916050434, "learning_rate": 3.6916694363951733e-06, "loss": 0.0045, "step": 80100 }, { "grad_norm": 0.029491858556866646, "learning_rate": 3.6847979978862913e-06, "loss": 0.0055, "step": 80110 }, { "grad_norm": 0.055182430893182755, "learning_rate": 3.6779327156975686e-06, "loss": 0.0046, "step": 80120 }, { "grad_norm": 0.030412087216973305, "learning_rate": 3.6710735907415583e-06, "loss": 0.0068, "step": 80130 }, { "grad_norm": 0.025867067277431488, "learning_rate": 3.664220623929976e-06, "loss": 0.0031, "step": 80140 }, { "grad_norm": 0.04805777966976166, "learning_rate": 3.6573738161737483e-06, "loss": 0.0079, "step": 80150 }, { "grad_norm": 0.030678916722536087, "learning_rate": 3.650533168382969e-06, "loss": 0.0102, "step": 80160 }, { "grad_norm": 0.048630308359861374, "learning_rate": 3.6436986814669004e-06, "loss": 0.0048, "step": 80170 }, { "grad_norm": 0.04834649711847305, "learning_rate": 3.6368703563340155e-06, "loss": 0.0055, "step": 80180 }, { "grad_norm": 0.04217454418540001, "learning_rate": 3.6300481938919327e-06, "loss": 0.0055, "step": 80190 }, { "grad_norm": 0.054022978991270065, "learning_rate": 3.6232321950474714e-06, "loss": 0.0071, "step": 80200 }, { "grad_norm": 0.0769679918885231, "learning_rate": 3.6164223607066284e-06, "loss": 0.0056, "step": 80210 }, { "grad_norm": 0.02639409340918064, "learning_rate": 3.6096186917745967e-06, "loss": 0.0035, "step": 80220 }, { "grad_norm": 0.02494625560939312, "learning_rate": 3.602821189155736e-06, "loss": 0.0025, "step": 80230 }, { "grad_norm": 0.013726224191486835, "learning_rate": 3.596029853753563e-06, "loss": 0.0045, "step": 80240 }, { "grad_norm": 0.024603594094514847, "learning_rate": 3.5892446864708105e-06, "loss": 0.0044, "step": 80250 }, { "grad_norm": 0.0466141439974308, "learning_rate": 3.58246568820938e-06, "loss": 0.0045, "step": 80260 }, { "grad_norm": 0.03058919683098793, "learning_rate": 3.575692859870344e-06, "loss": 0.0047, "step": 80270 }, { "grad_norm": 0.051142618060112, "learning_rate": 3.5689262023539727e-06, "loss": 0.0049, "step": 80280 }, { "grad_norm": 0.03470684960484505, "learning_rate": 3.5621657165596958e-06, "loss": 0.0057, "step": 80290 }, { "grad_norm": 0.05331423133611679, "learning_rate": 3.555411403386133e-06, "loss": 0.0053, "step": 80300 }, { "grad_norm": 0.030687682330608368, "learning_rate": 3.5486632637310892e-06, "loss": 0.0033, "step": 80310 }, { "grad_norm": 0.052352648228406906, "learning_rate": 3.5419212984915352e-06, "loss": 0.0078, "step": 80320 }, { "grad_norm": 0.07280631363391876, "learning_rate": 3.5351855085636322e-06, "loss": 0.0044, "step": 80330 }, { "grad_norm": 0.04880250245332718, "learning_rate": 3.5284558948427136e-06, "loss": 0.0068, "step": 80340 }, { "grad_norm": 0.022112984210252762, "learning_rate": 3.5217324582233036e-06, "loss": 0.0027, "step": 80350 }, { "grad_norm": 0.03615075722336769, "learning_rate": 3.5150151995990755e-06, "loss": 0.0028, "step": 80360 }, { "grad_norm": 0.0731503963470459, "learning_rate": 3.508304119862921e-06, "loss": 0.0066, "step": 80370 }, { "grad_norm": 0.03031747229397297, "learning_rate": 3.5015992199068926e-06, "loss": 0.0057, "step": 80380 }, { "grad_norm": 0.047789692878723145, "learning_rate": 3.4949005006222112e-06, "loss": 0.0038, "step": 80390 }, { "grad_norm": 0.04045327007770538, "learning_rate": 3.488207962899298e-06, "loss": 0.0039, "step": 80400 }, { "grad_norm": 0.037697963416576385, "learning_rate": 3.481521607627719e-06, "loss": 0.003, "step": 80410 }, { "grad_norm": 0.030845006927847862, "learning_rate": 3.474841435696258e-06, "loss": 0.0061, "step": 80420 }, { "grad_norm": 0.026436686515808105, "learning_rate": 3.4681674479928428e-06, "loss": 0.0042, "step": 80430 }, { "grad_norm": 0.019271817058324814, "learning_rate": 3.461499645404609e-06, "loss": 0.0044, "step": 80440 }, { "grad_norm": 0.04153026267886162, "learning_rate": 3.4548380288178584e-06, "loss": 0.0034, "step": 80450 }, { "grad_norm": 0.019805504009127617, "learning_rate": 3.448182599118055e-06, "loss": 0.0035, "step": 80460 }, { "grad_norm": 0.026896696537733078, "learning_rate": 3.441533357189858e-06, "loss": 0.0073, "step": 80470 }, { "grad_norm": 0.054159946739673615, "learning_rate": 3.434890303917104e-06, "loss": 0.0055, "step": 80480 }, { "grad_norm": 0.0858197808265686, "learning_rate": 3.428253440182794e-06, "loss": 0.0074, "step": 80490 }, { "grad_norm": 0.032825667411088943, "learning_rate": 3.4216227668691204e-06, "loss": 0.005, "step": 80500 }, { "grad_norm": 0.01979079842567444, "learning_rate": 3.4149982848574456e-06, "loss": 0.0052, "step": 80510 }, { "grad_norm": 0.03862792253494263, "learning_rate": 3.40837999502831e-06, "loss": 0.0039, "step": 80520 }, { "grad_norm": 0.05775384232401848, "learning_rate": 3.4017678982614376e-06, "loss": 0.0036, "step": 80530 }, { "grad_norm": 0.05421155318617821, "learning_rate": 3.3951619954357138e-06, "loss": 0.004, "step": 80540 }, { "grad_norm": 0.023509124293923378, "learning_rate": 3.38856228742922e-06, "loss": 0.004, "step": 80550 }, { "grad_norm": 0.05392851680517197, "learning_rate": 3.381968775119193e-06, "loss": 0.0038, "step": 80560 }, { "grad_norm": 0.021883267909288406, "learning_rate": 3.375381459382071e-06, "loss": 0.0059, "step": 80570 }, { "grad_norm": 0.04584988206624985, "learning_rate": 3.368800341093431e-06, "loss": 0.0063, "step": 80580 }, { "grad_norm": 0.06068442016839981, "learning_rate": 3.3622254211280735e-06, "loss": 0.0054, "step": 80590 }, { "grad_norm": 0.05062718689441681, "learning_rate": 3.355656700359949e-06, "loss": 0.0075, "step": 80600 }, { "grad_norm": 0.04195979982614517, "learning_rate": 3.349094179662171e-06, "loss": 0.0089, "step": 80610 }, { "grad_norm": 0.03166806325316429, "learning_rate": 3.342537859907069e-06, "loss": 0.0043, "step": 80620 }, { "grad_norm": 0.032905276864767075, "learning_rate": 3.335987741966101e-06, "loss": 0.0031, "step": 80630 }, { "grad_norm": 0.05874001607298851, "learning_rate": 3.329443826709927e-06, "loss": 0.0036, "step": 80640 }, { "grad_norm": 0.03626449778676033, "learning_rate": 3.3229061150083897e-06, "loss": 0.0044, "step": 80650 }, { "grad_norm": 0.04879063740372658, "learning_rate": 3.316374607730477e-06, "loss": 0.0045, "step": 80660 }, { "grad_norm": 0.02959793247282505, "learning_rate": 3.3098493057444e-06, "loss": 0.0053, "step": 80670 }, { "grad_norm": 0.043143291026353836, "learning_rate": 3.3033302099174934e-06, "loss": 0.0043, "step": 80680 }, { "grad_norm": 0.021732652559876442, "learning_rate": 3.2968173211162968e-06, "loss": 0.0045, "step": 80690 }, { "grad_norm": 0.03137722611427307, "learning_rate": 3.2903106402065233e-06, "loss": 0.0058, "step": 80700 }, { "grad_norm": 0.044522494077682495, "learning_rate": 3.2838101680530474e-06, "loss": 0.0074, "step": 80710 }, { "grad_norm": 0.05875556543469429, "learning_rate": 3.2773159055199287e-06, "loss": 0.0051, "step": 80720 }, { "grad_norm": 0.052582066506147385, "learning_rate": 3.2708278534703983e-06, "loss": 0.0058, "step": 80730 }, { "grad_norm": 0.06144608184695244, "learning_rate": 3.2643460127668666e-06, "loss": 0.0056, "step": 80740 }, { "grad_norm": 0.031279709190130234, "learning_rate": 3.257870384270911e-06, "loss": 0.0036, "step": 80750 }, { "grad_norm": 0.07399897277355194, "learning_rate": 3.251400968843288e-06, "loss": 0.0047, "step": 80760 }, { "grad_norm": 0.07421277463436127, "learning_rate": 3.244937767343931e-06, "loss": 0.0063, "step": 80770 }, { "grad_norm": 0.03427628055214882, "learning_rate": 3.238480780631942e-06, "loss": 0.003, "step": 80780 }, { "grad_norm": 0.044413045048713684, "learning_rate": 3.2320300095655843e-06, "loss": 0.0043, "step": 80790 }, { "grad_norm": 0.08192680776119232, "learning_rate": 3.2255854550023225e-06, "loss": 0.0062, "step": 80800 }, { "grad_norm": 0.0234648659825325, "learning_rate": 3.219147117798771e-06, "loss": 0.0034, "step": 80810 }, { "grad_norm": 0.044630274176597595, "learning_rate": 3.21271499881074e-06, "loss": 0.0062, "step": 80820 }, { "grad_norm": 0.057266853749752045, "learning_rate": 3.2062890988932002e-06, "loss": 0.0045, "step": 80830 }, { "grad_norm": 0.019996214658021927, "learning_rate": 3.199869418900292e-06, "loss": 0.0054, "step": 80840 }, { "grad_norm": 0.036072321236133575, "learning_rate": 3.1934559596853263e-06, "loss": 0.0065, "step": 80850 }, { "grad_norm": 0.06666350364685059, "learning_rate": 3.1870487221008105e-06, "loss": 0.0041, "step": 80860 }, { "grad_norm": 0.037823636084795, "learning_rate": 3.1806477069983966e-06, "loss": 0.0051, "step": 80870 }, { "grad_norm": 0.03821565583348274, "learning_rate": 3.1742529152289147e-06, "loss": 0.0062, "step": 80880 }, { "grad_norm": 0.04558565840125084, "learning_rate": 3.167864347642402e-06, "loss": 0.004, "step": 80890 }, { "grad_norm": 0.021330509334802628, "learning_rate": 3.1614820050880232e-06, "loss": 0.0028, "step": 80900 }, { "grad_norm": 0.02902086451649666, "learning_rate": 3.1551058884141282e-06, "loss": 0.0035, "step": 80910 }, { "grad_norm": 0.028726067394018173, "learning_rate": 3.148735998468255e-06, "loss": 0.0054, "step": 80920 }, { "grad_norm": 0.038677897304296494, "learning_rate": 3.142372336097099e-06, "loss": 0.005, "step": 80930 }, { "grad_norm": 0.0700678676366806, "learning_rate": 3.136014902146539e-06, "loss": 0.0052, "step": 80940 }, { "grad_norm": 0.02651931345462799, "learning_rate": 3.1296636974616154e-06, "loss": 0.0048, "step": 80950 }, { "grad_norm": 0.057229600846767426, "learning_rate": 3.123318722886531e-06, "loss": 0.0036, "step": 80960 }, { "grad_norm": 0.0441194586455822, "learning_rate": 3.1169799792646947e-06, "loss": 0.0069, "step": 80970 }, { "grad_norm": 0.026291754096746445, "learning_rate": 3.1106474674386545e-06, "loss": 0.0041, "step": 80980 }, { "grad_norm": 0.03458801284432411, "learning_rate": 3.1043211882501477e-06, "loss": 0.0041, "step": 80990 }, { "grad_norm": 0.03424926847219467, "learning_rate": 3.0980011425400855e-06, "loss": 0.0034, "step": 81000 }, { "grad_norm": 0.029730241745710373, "learning_rate": 3.091687331148518e-06, "loss": 0.0041, "step": 81010 }, { "grad_norm": 0.06740561127662659, "learning_rate": 3.0853797549147136e-06, "loss": 0.0055, "step": 81020 }, { "grad_norm": 0.027354244142770767, "learning_rate": 3.0790784146770666e-06, "loss": 0.0045, "step": 81030 }, { "grad_norm": 0.026061877608299255, "learning_rate": 3.072783311273192e-06, "loss": 0.0041, "step": 81040 }, { "grad_norm": 0.03165636211633682, "learning_rate": 3.066494445539847e-06, "loss": 0.0041, "step": 81050 }, { "grad_norm": 0.05383659154176712, "learning_rate": 3.0602118183129357e-06, "loss": 0.0036, "step": 81060 }, { "grad_norm": 0.06645873188972473, "learning_rate": 3.053935430427579e-06, "loss": 0.0074, "step": 81070 }, { "grad_norm": 0.03570808097720146, "learning_rate": 3.0476652827180486e-06, "loss": 0.004, "step": 81080 }, { "grad_norm": 0.020398318767547607, "learning_rate": 3.0414013760177772e-06, "loss": 0.0032, "step": 81090 }, { "grad_norm": 0.06454195082187653, "learning_rate": 3.0351437111593826e-06, "loss": 0.0043, "step": 81100 }, { "grad_norm": 0.04483721777796745, "learning_rate": 3.02889228897465e-06, "loss": 0.0038, "step": 81110 }, { "grad_norm": 0.078842893242836, "learning_rate": 3.022647110294524e-06, "loss": 0.0058, "step": 81120 }, { "grad_norm": 0.035235341638326645, "learning_rate": 3.016408175949137e-06, "loss": 0.0047, "step": 81130 }, { "grad_norm": 0.06925246119499207, "learning_rate": 3.01017548676778e-06, "loss": 0.0045, "step": 81140 }, { "grad_norm": 0.01968534290790558, "learning_rate": 3.003949043578913e-06, "loss": 0.004, "step": 81150 }, { "grad_norm": 0.06276430934667587, "learning_rate": 2.9977288472101737e-06, "loss": 0.006, "step": 81160 }, { "grad_norm": 0.0355696901679039, "learning_rate": 2.9915148984883625e-06, "loss": 0.004, "step": 81170 }, { "grad_norm": 0.01895274966955185, "learning_rate": 2.98530719823944e-06, "loss": 0.0029, "step": 81180 }, { "grad_norm": 0.0336642861366272, "learning_rate": 2.979105747288563e-06, "loss": 0.005, "step": 81190 }, { "grad_norm": 0.040337320417165756, "learning_rate": 2.9729105464600336e-06, "loss": 0.006, "step": 81200 }, { "grad_norm": 0.05569040775299072, "learning_rate": 2.9667215965773375e-06, "loss": 0.0044, "step": 81210 }, { "grad_norm": 0.07183659821748734, "learning_rate": 2.9605388984631277e-06, "loss": 0.0078, "step": 81220 }, { "grad_norm": 0.05208978056907654, "learning_rate": 2.9543624529392077e-06, "loss": 0.0045, "step": 81230 }, { "grad_norm": 0.03168375417590141, "learning_rate": 2.9481922608265765e-06, "loss": 0.0071, "step": 81240 }, { "grad_norm": 0.04510045424103737, "learning_rate": 2.9420283229453726e-06, "loss": 0.0034, "step": 81250 }, { "grad_norm": 0.033499572426080704, "learning_rate": 2.935870640114946e-06, "loss": 0.0049, "step": 81260 }, { "grad_norm": 0.025728754699230194, "learning_rate": 2.9297192131537807e-06, "loss": 0.0048, "step": 81270 }, { "grad_norm": 0.02124030701816082, "learning_rate": 2.923574042879529e-06, "loss": 0.0039, "step": 81280 }, { "grad_norm": 0.026371533051133156, "learning_rate": 2.9174351301090254e-06, "loss": 0.0036, "step": 81290 }, { "grad_norm": 0.037196021527051926, "learning_rate": 2.911302475658273e-06, "loss": 0.0039, "step": 81300 }, { "grad_norm": 0.03638237714767456, "learning_rate": 2.9051760803424256e-06, "loss": 0.0066, "step": 81310 }, { "grad_norm": 0.04639982804656029, "learning_rate": 2.8990559449758313e-06, "loss": 0.0048, "step": 81320 }, { "grad_norm": 0.022206388413906097, "learning_rate": 2.8929420703719845e-06, "loss": 0.0033, "step": 81330 }, { "grad_norm": 0.03941357135772705, "learning_rate": 2.886834457343557e-06, "loss": 0.0042, "step": 81340 }, { "grad_norm": 0.03084251843392849, "learning_rate": 2.880733106702388e-06, "loss": 0.0056, "step": 81350 }, { "grad_norm": 0.07318037003278732, "learning_rate": 2.874638019259479e-06, "loss": 0.0055, "step": 81360 }, { "grad_norm": 0.03179068863391876, "learning_rate": 2.8685491958250046e-06, "loss": 0.0046, "step": 81370 }, { "grad_norm": 0.026758462190628052, "learning_rate": 2.8624666372083054e-06, "loss": 0.0043, "step": 81380 }, { "grad_norm": 0.04450499266386032, "learning_rate": 2.8563903442178963e-06, "loss": 0.0038, "step": 81390 }, { "grad_norm": 0.04294769465923309, "learning_rate": 2.8503203176614255e-06, "loss": 0.005, "step": 81400 }, { "grad_norm": 0.03053668513894081, "learning_rate": 2.8442565583457646e-06, "loss": 0.0055, "step": 81410 }, { "grad_norm": 0.019266264513134956, "learning_rate": 2.8381990670769075e-06, "loss": 0.0045, "step": 81420 }, { "grad_norm": 0.018802857026457787, "learning_rate": 2.8321478446600325e-06, "loss": 0.008, "step": 81430 }, { "grad_norm": 0.05110630393028259, "learning_rate": 2.8261028918994845e-06, "loss": 0.0047, "step": 81440 }, { "grad_norm": 0.0715148076415062, "learning_rate": 2.8200642095987663e-06, "loss": 0.0057, "step": 81450 }, { "grad_norm": 0.01884750835597515, "learning_rate": 2.8140317985605567e-06, "loss": 0.0055, "step": 81460 }, { "grad_norm": 0.0752706304192543, "learning_rate": 2.808005659586693e-06, "loss": 0.0052, "step": 81470 }, { "grad_norm": 0.06766694039106369, "learning_rate": 2.801985793478179e-06, "loss": 0.0044, "step": 81480 }, { "grad_norm": 0.048084210604429245, "learning_rate": 2.795972201035207e-06, "loss": 0.0035, "step": 81490 }, { "grad_norm": 0.05705493688583374, "learning_rate": 2.789964883057106e-06, "loss": 0.0045, "step": 81500 }, { "grad_norm": 0.028090346604585648, "learning_rate": 2.783963840342374e-06, "loss": 0.006, "step": 81510 }, { "grad_norm": 0.04677457734942436, "learning_rate": 2.7779690736886965e-06, "loss": 0.0084, "step": 81520 }, { "grad_norm": 0.0337982214987278, "learning_rate": 2.7719805838929026e-06, "loss": 0.0049, "step": 81530 }, { "grad_norm": 0.06355522572994232, "learning_rate": 2.7659983717509997e-06, "loss": 0.0062, "step": 81540 }, { "grad_norm": 0.06518282741308212, "learning_rate": 2.7600224380581574e-06, "loss": 0.0055, "step": 81550 }, { "grad_norm": 0.03158562630414963, "learning_rate": 2.7540527836087126e-06, "loss": 0.0042, "step": 81560 }, { "grad_norm": 0.025975419208407402, "learning_rate": 2.748089409196153e-06, "loss": 0.0036, "step": 81570 }, { "grad_norm": 0.019450725987553596, "learning_rate": 2.7421323156131608e-06, "loss": 0.0047, "step": 81580 }, { "grad_norm": 0.031358249485492706, "learning_rate": 2.7361815036515527e-06, "loss": 0.0027, "step": 81590 }, { "grad_norm": 0.03296709805727005, "learning_rate": 2.730236974102335e-06, "loss": 0.0048, "step": 81600 }, { "grad_norm": 0.04069359973073006, "learning_rate": 2.724298727755653e-06, "loss": 0.0043, "step": 81610 }, { "grad_norm": 0.053792502731084824, "learning_rate": 2.7183667654008373e-06, "loss": 0.0042, "step": 81620 }, { "grad_norm": 0.06394972652196884, "learning_rate": 2.7124410878263896e-06, "loss": 0.0053, "step": 81630 }, { "grad_norm": 0.0318615660071373, "learning_rate": 2.706521695819947e-06, "loss": 0.0056, "step": 81640 }, { "grad_norm": 0.026783965528011322, "learning_rate": 2.700608590168341e-06, "loss": 0.004, "step": 81650 }, { "grad_norm": 0.03780268877744675, "learning_rate": 2.6947017716575597e-06, "loss": 0.0067, "step": 81660 }, { "grad_norm": 0.031300339847803116, "learning_rate": 2.68880124107273e-06, "loss": 0.0046, "step": 81670 }, { "grad_norm": 0.11060918122529984, "learning_rate": 2.682906999198176e-06, "loss": 0.0048, "step": 81680 }, { "grad_norm": 0.05694485455751419, "learning_rate": 2.677019046817375e-06, "loss": 0.0041, "step": 81690 }, { "grad_norm": 0.06401043385267258, "learning_rate": 2.6711373847129517e-06, "loss": 0.0045, "step": 81700 }, { "grad_norm": 0.05619070678949356, "learning_rate": 2.665262013666742e-06, "loss": 0.0039, "step": 81710 }, { "grad_norm": 0.044815462082624435, "learning_rate": 2.6593929344596813e-06, "loss": 0.0045, "step": 81720 }, { "grad_norm": 0.017296764999628067, "learning_rate": 2.653530147871913e-06, "loss": 0.0045, "step": 81730 }, { "grad_norm": 0.04523545131087303, "learning_rate": 2.6476736546827365e-06, "loss": 0.0041, "step": 81740 }, { "grad_norm": 0.04888172820210457, "learning_rate": 2.6418234556706e-06, "loss": 0.0083, "step": 81750 }, { "grad_norm": 0.03130670264363289, "learning_rate": 2.6359795516131323e-06, "loss": 0.0048, "step": 81760 }, { "grad_norm": 0.06370729953050613, "learning_rate": 2.6301419432871177e-06, "loss": 0.005, "step": 81770 }, { "grad_norm": 0.030515113845467567, "learning_rate": 2.6243106314685017e-06, "loss": 0.0057, "step": 81780 }, { "grad_norm": 0.052531369030475616, "learning_rate": 2.618485616932398e-06, "loss": 0.0049, "step": 81790 }, { "grad_norm": 0.020106805488467216, "learning_rate": 2.6126669004530756e-06, "loss": 0.0035, "step": 81800 }, { "grad_norm": 0.02745947241783142, "learning_rate": 2.606854482803972e-06, "loss": 0.0055, "step": 81810 }, { "grad_norm": 0.050215307623147964, "learning_rate": 2.601048364757702e-06, "loss": 0.0048, "step": 81820 }, { "grad_norm": 0.07787709683179855, "learning_rate": 2.5952485470860045e-06, "loss": 0.0034, "step": 81830 }, { "grad_norm": 0.04231589660048485, "learning_rate": 2.5894550305598118e-06, "loss": 0.0045, "step": 81840 }, { "grad_norm": 0.04598706215620041, "learning_rate": 2.5836678159492145e-06, "loss": 0.0045, "step": 81850 }, { "grad_norm": 0.06361506879329681, "learning_rate": 2.577886904023463e-06, "loss": 0.0058, "step": 81860 }, { "grad_norm": 0.031189484521746635, "learning_rate": 2.572112295550977e-06, "loss": 0.0048, "step": 81870 }, { "grad_norm": 0.06651746481657028, "learning_rate": 2.5663439912993083e-06, "loss": 0.0056, "step": 81880 }, { "grad_norm": 0.059765011072158813, "learning_rate": 2.5605819920352103e-06, "loss": 0.008, "step": 81890 }, { "grad_norm": 0.048873171210289, "learning_rate": 2.5548262985245764e-06, "loss": 0.0044, "step": 81900 }, { "grad_norm": 0.037625595927238464, "learning_rate": 2.5490769115324666e-06, "loss": 0.004, "step": 81910 }, { "grad_norm": 0.04894713684916496, "learning_rate": 2.543333831823097e-06, "loss": 0.006, "step": 81920 }, { "grad_norm": 0.07627056539058685, "learning_rate": 2.5375970601598576e-06, "loss": 0.0067, "step": 81930 }, { "grad_norm": 0.029969269409775734, "learning_rate": 2.531866597305288e-06, "loss": 0.0046, "step": 81940 }, { "grad_norm": 0.04154716432094574, "learning_rate": 2.5261424440211013e-06, "loss": 0.0045, "step": 81950 }, { "grad_norm": 0.03620174154639244, "learning_rate": 2.5204246010681555e-06, "loss": 0.0045, "step": 81960 }, { "grad_norm": 0.04822814092040062, "learning_rate": 2.514713069206487e-06, "loss": 0.0039, "step": 81970 }, { "grad_norm": 0.059378061443567276, "learning_rate": 2.509007849195277e-06, "loss": 0.0051, "step": 81980 }, { "grad_norm": 0.020348545163869858, "learning_rate": 2.5033089417928923e-06, "loss": 0.0056, "step": 81990 }, { "grad_norm": 0.015269431285560131, "learning_rate": 2.4976163477568203e-06, "loss": 0.0046, "step": 82000 }, { "grad_norm": 0.018037989735603333, "learning_rate": 2.491930067843751e-06, "loss": 0.0046, "step": 82010 }, { "grad_norm": 0.025395262986421585, "learning_rate": 2.4862501028095063e-06, "loss": 0.0059, "step": 82020 }, { "grad_norm": 0.0315755270421505, "learning_rate": 2.480576453409095e-06, "loss": 0.0044, "step": 82030 }, { "grad_norm": 0.03002466820180416, "learning_rate": 2.4749091203966624e-06, "loss": 0.0046, "step": 82040 }, { "grad_norm": 0.024013984948396683, "learning_rate": 2.4692481045255234e-06, "loss": 0.0056, "step": 82050 }, { "grad_norm": 0.03293535113334656, "learning_rate": 2.463593406548148e-06, "loss": 0.0037, "step": 82060 }, { "grad_norm": 0.023286471143364906, "learning_rate": 2.4579450272161685e-06, "loss": 0.0067, "step": 82070 }, { "grad_norm": 0.015670204535126686, "learning_rate": 2.4523029672803955e-06, "loss": 0.0043, "step": 82080 }, { "grad_norm": 0.08765196800231934, "learning_rate": 2.4466672274907785e-06, "loss": 0.0091, "step": 82090 }, { "grad_norm": 0.04593198001384735, "learning_rate": 2.44103780859643e-06, "loss": 0.0037, "step": 82100 }, { "grad_norm": 0.02667536400258541, "learning_rate": 2.435414711345624e-06, "loss": 0.0031, "step": 82110 }, { "grad_norm": 0.01828683353960514, "learning_rate": 2.4297979364857947e-06, "loss": 0.0049, "step": 82120 }, { "grad_norm": 0.02422233298420906, "learning_rate": 2.424187484763535e-06, "loss": 0.0039, "step": 82130 }, { "grad_norm": 0.050180479884147644, "learning_rate": 2.418583356924603e-06, "loss": 0.0042, "step": 82140 }, { "grad_norm": 0.061344560235738754, "learning_rate": 2.4129855537139145e-06, "loss": 0.0029, "step": 82150 }, { "grad_norm": 0.05019068345427513, "learning_rate": 2.4073940758755354e-06, "loss": 0.0048, "step": 82160 }, { "grad_norm": 0.027966173365712166, "learning_rate": 2.4018089241526987e-06, "loss": 0.005, "step": 82170 }, { "grad_norm": 0.021065913140773773, "learning_rate": 2.3962300992877994e-06, "loss": 0.0041, "step": 82180 }, { "grad_norm": 0.014870807528495789, "learning_rate": 2.3906576020223837e-06, "loss": 0.0052, "step": 82190 }, { "grad_norm": 0.026860542595386505, "learning_rate": 2.3850914330971585e-06, "loss": 0.0043, "step": 82200 }, { "grad_norm": 0.054136648774147034, "learning_rate": 2.3795315932520046e-06, "loss": 0.0102, "step": 82210 }, { "grad_norm": 0.0310309249907732, "learning_rate": 2.373978083225925e-06, "loss": 0.0035, "step": 82220 }, { "grad_norm": 0.06143074482679367, "learning_rate": 2.3684309037571183e-06, "loss": 0.0063, "step": 82230 }, { "grad_norm": 0.06041119992733002, "learning_rate": 2.3628900555829335e-06, "loss": 0.0069, "step": 82240 }, { "grad_norm": 0.059575825929641724, "learning_rate": 2.357355539439865e-06, "loss": 0.0049, "step": 82250 }, { "grad_norm": 0.06377801299095154, "learning_rate": 2.3518273560635794e-06, "loss": 0.0036, "step": 82260 }, { "grad_norm": 0.04935471713542938, "learning_rate": 2.346305506188884e-06, "loss": 0.0049, "step": 82270 }, { "grad_norm": 0.06472312659025192, "learning_rate": 2.3407899905497633e-06, "loss": 0.0055, "step": 82280 }, { "grad_norm": 0.04538726061582565, "learning_rate": 2.3352808098793422e-06, "loss": 0.0057, "step": 82290 }, { "grad_norm": 0.04050499200820923, "learning_rate": 2.329777964909929e-06, "loss": 0.0031, "step": 82300 }, { "grad_norm": 0.03393828496336937, "learning_rate": 2.324281456372973e-06, "loss": 0.0035, "step": 82310 }, { "grad_norm": 0.0363842211663723, "learning_rate": 2.3187912849990668e-06, "loss": 0.0057, "step": 82320 }, { "grad_norm": 0.02235463261604309, "learning_rate": 2.3133074515179877e-06, "loss": 0.0041, "step": 82330 }, { "grad_norm": 0.04340537264943123, "learning_rate": 2.3078299566586524e-06, "loss": 0.0049, "step": 82340 }, { "grad_norm": 0.03392195329070091, "learning_rate": 2.3023588011491516e-06, "loss": 0.0036, "step": 82350 }, { "grad_norm": 0.07477684319019318, "learning_rate": 2.2968939857167136e-06, "loss": 0.01, "step": 82360 }, { "grad_norm": 0.05763104185461998, "learning_rate": 2.2914355110877416e-06, "loss": 0.0058, "step": 82370 }, { "grad_norm": 0.0225739236921072, "learning_rate": 2.285983377987788e-06, "loss": 0.005, "step": 82380 }, { "grad_norm": 0.0505005419254303, "learning_rate": 2.280537587141551e-06, "loss": 0.0039, "step": 82390 }, { "grad_norm": 0.04768940061330795, "learning_rate": 2.2750981392729133e-06, "loss": 0.0047, "step": 82400 }, { "grad_norm": 0.030054301023483276, "learning_rate": 2.2696650351048897e-06, "loss": 0.0042, "step": 82410 }, { "grad_norm": 0.02417384460568428, "learning_rate": 2.264238275359659e-06, "loss": 0.0045, "step": 82420 }, { "grad_norm": 0.03872252255678177, "learning_rate": 2.2588178607585666e-06, "loss": 0.0049, "step": 82430 }, { "grad_norm": 0.05293301120400429, "learning_rate": 2.2534037920220907e-06, "loss": 0.0046, "step": 82440 }, { "grad_norm": 0.04450776427984238, "learning_rate": 2.2479960698698956e-06, "loss": 0.0054, "step": 82450 }, { "grad_norm": 0.05565742403268814, "learning_rate": 2.242594695020783e-06, "loss": 0.0037, "step": 82460 }, { "grad_norm": 0.06470488756895065, "learning_rate": 2.2371996681927186e-06, "loss": 0.0041, "step": 82470 }, { "grad_norm": 0.07451796531677246, "learning_rate": 2.2318109901028283e-06, "loss": 0.0043, "step": 82480 }, { "grad_norm": 0.04745158180594444, "learning_rate": 2.226428661467367e-06, "loss": 0.0044, "step": 82490 }, { "grad_norm": 0.06269429624080658, "learning_rate": 2.2210526830017785e-06, "loss": 0.0048, "step": 82500 }, { "grad_norm": 0.04320323094725609, "learning_rate": 2.2156830554206474e-06, "loss": 0.0053, "step": 82510 }, { "grad_norm": 0.04955970495939255, "learning_rate": 2.210319779437708e-06, "loss": 0.0057, "step": 82520 }, { "grad_norm": 0.03611617162823677, "learning_rate": 2.204962855765885e-06, "loss": 0.0055, "step": 82530 }, { "grad_norm": 0.04858999326825142, "learning_rate": 2.1996122851172087e-06, "loss": 0.0047, "step": 82540 }, { "grad_norm": 0.01968955621123314, "learning_rate": 2.1942680682028924e-06, "loss": 0.0032, "step": 82550 }, { "grad_norm": 0.05378133803606033, "learning_rate": 2.1889302057333083e-06, "loss": 0.0074, "step": 82560 }, { "grad_norm": 0.030077306553721428, "learning_rate": 2.1835986984179768e-06, "loss": 0.0055, "step": 82570 }, { "grad_norm": 0.027714943513274193, "learning_rate": 2.17827354696557e-06, "loss": 0.0044, "step": 82580 }, { "grad_norm": 0.023278934881091118, "learning_rate": 2.172954752083917e-06, "loss": 0.0064, "step": 82590 }, { "grad_norm": 0.068580761551857, "learning_rate": 2.167642314480006e-06, "loss": 0.0046, "step": 82600 }, { "grad_norm": 0.07601236552000046, "learning_rate": 2.16233623485998e-06, "loss": 0.0046, "step": 82610 }, { "grad_norm": 0.02039729803800583, "learning_rate": 2.1570365139291338e-06, "loss": 0.0047, "step": 82620 }, { "grad_norm": 0.028266919776797295, "learning_rate": 2.151743152391916e-06, "loss": 0.0041, "step": 82630 }, { "grad_norm": 0.016836995258927345, "learning_rate": 2.146456150951942e-06, "loss": 0.0066, "step": 82640 }, { "grad_norm": 0.04537055268883705, "learning_rate": 2.1411755103119604e-06, "loss": 0.0055, "step": 82650 }, { "grad_norm": 0.03328142687678337, "learning_rate": 2.135901231173881e-06, "loss": 0.006, "step": 82660 }, { "grad_norm": 0.020433271303772926, "learning_rate": 2.1306333142387836e-06, "loss": 0.0047, "step": 82670 }, { "grad_norm": 0.03879780322313309, "learning_rate": 2.1253717602068955e-06, "loss": 0.0032, "step": 82680 }, { "grad_norm": 0.031072726473212242, "learning_rate": 2.1201165697775916e-06, "loss": 0.0039, "step": 82690 }, { "grad_norm": 0.025787489488720894, "learning_rate": 2.1148677436493904e-06, "loss": 0.007, "step": 82700 }, { "grad_norm": 0.034825488924980164, "learning_rate": 2.109625282519989e-06, "loss": 0.0038, "step": 82710 }, { "grad_norm": 0.08130300790071487, "learning_rate": 2.1043891870862254e-06, "loss": 0.0047, "step": 82720 }, { "grad_norm": 0.02986251190304756, "learning_rate": 2.099159458044092e-06, "loss": 0.006, "step": 82730 }, { "grad_norm": 0.0215001218020916, "learning_rate": 2.0939360960887276e-06, "loss": 0.0045, "step": 82740 }, { "grad_norm": 0.03401508554816246, "learning_rate": 2.0887191019144603e-06, "loss": 0.003, "step": 82750 }, { "grad_norm": 0.020462138578295708, "learning_rate": 2.0835084762147185e-06, "loss": 0.0061, "step": 82760 }, { "grad_norm": 0.07936698943376541, "learning_rate": 2.0783042196821146e-06, "loss": 0.0071, "step": 82770 }, { "grad_norm": 0.049617454409599304, "learning_rate": 2.073106333008418e-06, "loss": 0.006, "step": 82780 }, { "grad_norm": 0.029658034443855286, "learning_rate": 2.0679148168845363e-06, "loss": 0.0043, "step": 82790 }, { "grad_norm": 0.035064447671175, "learning_rate": 2.0627296720005406e-06, "loss": 0.0087, "step": 82800 }, { "grad_norm": 0.05672342702746391, "learning_rate": 2.0575508990456516e-06, "loss": 0.0046, "step": 82810 }, { "grad_norm": 0.04923676326870918, "learning_rate": 2.0523784987082404e-06, "loss": 0.0045, "step": 82820 }, { "grad_norm": 0.02378048561513424, "learning_rate": 2.0472124716758413e-06, "loss": 0.0041, "step": 82830 }, { "grad_norm": 0.023086659610271454, "learning_rate": 2.0420528186351316e-06, "loss": 0.002, "step": 82840 }, { "grad_norm": 0.028193416073918343, "learning_rate": 2.036899540271936e-06, "loss": 0.0046, "step": 82850 }, { "grad_norm": 0.03719387948513031, "learning_rate": 2.0317526372712557e-06, "loss": 0.0045, "step": 82860 }, { "grad_norm": 0.016626494005322456, "learning_rate": 2.0266121103172108e-06, "loss": 0.004, "step": 82870 }, { "grad_norm": 0.03917234018445015, "learning_rate": 2.0214779600931043e-06, "loss": 0.0066, "step": 82880 }, { "grad_norm": 0.04381706938147545, "learning_rate": 2.016350187281363e-06, "loss": 0.0044, "step": 82890 }, { "grad_norm": 0.022454068064689636, "learning_rate": 2.011228792563602e-06, "loss": 0.0043, "step": 82900 }, { "grad_norm": 0.05126727744936943, "learning_rate": 2.0061137766205606e-06, "loss": 0.0038, "step": 82910 }, { "grad_norm": 0.04608731344342232, "learning_rate": 2.0010051401321337e-06, "loss": 0.0031, "step": 82920 }, { "grad_norm": 0.03792480751872063, "learning_rate": 1.9959028837773787e-06, "loss": 0.008, "step": 82930 }, { "grad_norm": 0.041169799864292145, "learning_rate": 1.9908070082344965e-06, "loss": 0.0038, "step": 82940 }, { "grad_norm": 0.07332029938697815, "learning_rate": 1.9857175141808404e-06, "loss": 0.0052, "step": 82950 }, { "grad_norm": 0.017251618206501007, "learning_rate": 1.980634402292919e-06, "loss": 0.0043, "step": 82960 }, { "grad_norm": 0.026956165209412575, "learning_rate": 1.975557673246392e-06, "loss": 0.0058, "step": 82970 }, { "grad_norm": 0.060105014592409134, "learning_rate": 1.970487327716064e-06, "loss": 0.0058, "step": 82980 }, { "grad_norm": 0.04992195591330528, "learning_rate": 1.965423366375907e-06, "loss": 0.0037, "step": 82990 }, { "grad_norm": 0.030540717765688896, "learning_rate": 1.960365789899027e-06, "loss": 0.0038, "step": 83000 }, { "grad_norm": 0.03209869563579559, "learning_rate": 1.955314598957686e-06, "loss": 0.0035, "step": 83010 }, { "grad_norm": 0.052679140120744705, "learning_rate": 1.9502697942233085e-06, "loss": 0.007, "step": 83020 }, { "grad_norm": 0.028856974095106125, "learning_rate": 1.9452313763664577e-06, "loss": 0.006, "step": 83030 }, { "grad_norm": 0.037377484142780304, "learning_rate": 1.940199346056837e-06, "loss": 0.003, "step": 83040 }, { "grad_norm": 0.040519919246435165, "learning_rate": 1.9351737039633387e-06, "loss": 0.0043, "step": 83050 }, { "grad_norm": 0.02897861786186695, "learning_rate": 1.930154450753968e-06, "loss": 0.0049, "step": 83060 }, { "grad_norm": 0.022231660783290863, "learning_rate": 1.9251415870958966e-06, "loss": 0.0026, "step": 83070 }, { "grad_norm": 0.055708907544612885, "learning_rate": 1.920135113655458e-06, "loss": 0.0047, "step": 83080 }, { "grad_norm": 0.03430239483714104, "learning_rate": 1.915135031098103e-06, "loss": 0.004, "step": 83090 }, { "grad_norm": 0.02143060602247715, "learning_rate": 1.9101413400884673e-06, "loss": 0.0054, "step": 83100 }, { "grad_norm": 0.08059486001729965, "learning_rate": 1.9051540412903136e-06, "loss": 0.0052, "step": 83110 }, { "grad_norm": 0.029467785730957985, "learning_rate": 1.9001731353665786e-06, "loss": 0.0047, "step": 83120 }, { "grad_norm": 0.013373183086514473, "learning_rate": 1.8951986229793329e-06, "loss": 0.0034, "step": 83130 }, { "grad_norm": 0.05552797392010689, "learning_rate": 1.890230504789786e-06, "loss": 0.005, "step": 83140 }, { "grad_norm": 0.03310956060886383, "learning_rate": 1.885268781458327e-06, "loss": 0.005, "step": 83150 }, { "grad_norm": 0.02323441579937935, "learning_rate": 1.8803134536444667e-06, "loss": 0.0043, "step": 83160 }, { "grad_norm": 0.02289758436381817, "learning_rate": 1.8753645220068895e-06, "loss": 0.0041, "step": 83170 }, { "grad_norm": 0.05058930069208145, "learning_rate": 1.8704219872034134e-06, "loss": 0.0055, "step": 83180 }, { "grad_norm": 0.028299525380134583, "learning_rate": 1.8654858498910077e-06, "loss": 0.0038, "step": 83190 }, { "grad_norm": 0.030502697452902794, "learning_rate": 1.8605561107257974e-06, "loss": 0.0044, "step": 83200 }, { "grad_norm": 0.04179242625832558, "learning_rate": 1.8556327703630584e-06, "loss": 0.0047, "step": 83210 }, { "grad_norm": 0.030848246067762375, "learning_rate": 1.8507158294572113e-06, "loss": 0.0029, "step": 83220 }, { "grad_norm": 0.04287030175328255, "learning_rate": 1.8458052886618227e-06, "loss": 0.0036, "step": 83230 }, { "grad_norm": 0.05818634852766991, "learning_rate": 1.840901148629609e-06, "loss": 0.0067, "step": 83240 }, { "grad_norm": 0.019582068547606468, "learning_rate": 1.83600341001246e-06, "loss": 0.006, "step": 83250 }, { "grad_norm": 0.021327929571270943, "learning_rate": 1.8311120734613606e-06, "loss": 0.0037, "step": 83260 }, { "grad_norm": 0.022552113980054855, "learning_rate": 1.8262271396265018e-06, "loss": 0.0049, "step": 83270 }, { "grad_norm": 0.0694199949502945, "learning_rate": 1.8213486091571973e-06, "loss": 0.0051, "step": 83280 }, { "grad_norm": 0.0535440556704998, "learning_rate": 1.816476482701912e-06, "loss": 0.0041, "step": 83290 }, { "grad_norm": 0.06953739374876022, "learning_rate": 1.8116107609082612e-06, "loss": 0.004, "step": 83300 }, { "grad_norm": 0.01738974079489708, "learning_rate": 1.8067514444229994e-06, "loss": 0.0069, "step": 83310 }, { "grad_norm": 0.015893300995230675, "learning_rate": 1.8018985338920436e-06, "loss": 0.0033, "step": 83320 }, { "grad_norm": 0.05747238174080849, "learning_rate": 1.7970520299604387e-06, "loss": 0.0057, "step": 83330 }, { "grad_norm": 0.05760171264410019, "learning_rate": 1.7922119332724197e-06, "loss": 0.0039, "step": 83340 }, { "grad_norm": 0.02548847906291485, "learning_rate": 1.7873782444713326e-06, "loss": 0.0035, "step": 83350 }, { "grad_norm": 0.05347827821969986, "learning_rate": 1.7825509641996752e-06, "loss": 0.004, "step": 83360 }, { "grad_norm": 0.023734979331493378, "learning_rate": 1.777730093099106e-06, "loss": 0.0052, "step": 83370 }, { "grad_norm": 0.02950754202902317, "learning_rate": 1.7729156318104179e-06, "loss": 0.0042, "step": 83380 }, { "grad_norm": 0.02686496451497078, "learning_rate": 1.7681075809735714e-06, "loss": 0.0037, "step": 83390 }, { "grad_norm": 0.08417069911956787, "learning_rate": 1.7633059412276553e-06, "loss": 0.0067, "step": 83400 }, { "grad_norm": 0.023212919011712074, "learning_rate": 1.7585107132109203e-06, "loss": 0.0059, "step": 83410 }, { "grad_norm": 0.04326402768492699, "learning_rate": 1.753721897560756e-06, "loss": 0.0031, "step": 83420 }, { "grad_norm": 0.03727416321635246, "learning_rate": 1.7489394949136983e-06, "loss": 0.0042, "step": 83430 }, { "grad_norm": 0.017257576808333397, "learning_rate": 1.744163505905444e-06, "loss": 0.0043, "step": 83440 }, { "grad_norm": 0.020992565900087357, "learning_rate": 1.7393939311708185e-06, "loss": 0.004, "step": 83450 }, { "grad_norm": 0.05706974118947983, "learning_rate": 1.7346307713438147e-06, "loss": 0.0056, "step": 83460 }, { "grad_norm": 0.025508860126137733, "learning_rate": 1.7298740270575543e-06, "loss": 0.0064, "step": 83470 }, { "grad_norm": 0.028774317353963852, "learning_rate": 1.7251236989443143e-06, "loss": 0.0032, "step": 83480 }, { "grad_norm": 0.04161326214671135, "learning_rate": 1.7203797876355177e-06, "loss": 0.0027, "step": 83490 }, { "grad_norm": 0.031786367297172546, "learning_rate": 1.715642293761749e-06, "loss": 0.0057, "step": 83500 }, { "grad_norm": 0.03649488463997841, "learning_rate": 1.7109112179527098e-06, "loss": 0.004, "step": 83510 }, { "grad_norm": 0.061796531081199646, "learning_rate": 1.7061865608372807e-06, "loss": 0.0051, "step": 83520 }, { "grad_norm": 0.0228867270052433, "learning_rate": 1.7014683230434593e-06, "loss": 0.0069, "step": 83530 }, { "grad_norm": 0.04359079524874687, "learning_rate": 1.6967565051984101e-06, "loss": 0.0038, "step": 83540 }, { "grad_norm": 0.03013456240296364, "learning_rate": 1.6920511079284384e-06, "loss": 0.0071, "step": 83550 }, { "grad_norm": 0.06592834740877151, "learning_rate": 1.6873521318589935e-06, "loss": 0.0056, "step": 83560 }, { "grad_norm": 0.05172634869813919, "learning_rate": 1.682659577614687e-06, "loss": 0.0041, "step": 83570 }, { "grad_norm": 0.027888676151633263, "learning_rate": 1.6779734458192475e-06, "loss": 0.0049, "step": 83580 }, { "grad_norm": 0.07510567456483841, "learning_rate": 1.6732937370955715e-06, "loss": 0.0044, "step": 83590 }, { "grad_norm": 0.04071696102619171, "learning_rate": 1.6686204520656945e-06, "loss": 0.0039, "step": 83600 }, { "grad_norm": 0.022154314443469048, "learning_rate": 1.6639535913508031e-06, "loss": 0.004, "step": 83610 }, { "grad_norm": 0.0607805959880352, "learning_rate": 1.6592931555712288e-06, "loss": 0.0042, "step": 83620 }, { "grad_norm": 0.028608761727809906, "learning_rate": 1.6546391453464426e-06, "loss": 0.003, "step": 83630 }, { "grad_norm": 0.06626938283443451, "learning_rate": 1.6499915612950723e-06, "loss": 0.0054, "step": 83640 }, { "grad_norm": 0.05446016043424606, "learning_rate": 1.6453504040348787e-06, "loss": 0.0034, "step": 83650 }, { "grad_norm": 0.02983025647699833, "learning_rate": 1.6407156741827746e-06, "loss": 0.0077, "step": 83660 }, { "grad_norm": 0.04040644317865372, "learning_rate": 1.636087372354822e-06, "loss": 0.0067, "step": 83670 }, { "grad_norm": 0.03808322921395302, "learning_rate": 1.6314654991662292e-06, "loss": 0.0036, "step": 83680 }, { "grad_norm": 0.03116566315293312, "learning_rate": 1.6268500552313326e-06, "loss": 0.0061, "step": 83690 }, { "grad_norm": 0.04381275177001953, "learning_rate": 1.6222410411636358e-06, "loss": 0.0047, "step": 83700 }, { "grad_norm": 0.027163183316588402, "learning_rate": 1.617638457575782e-06, "loss": 0.0057, "step": 83710 }, { "grad_norm": 0.025798019021749496, "learning_rate": 1.6130423050795484e-06, "loss": 0.003, "step": 83720 }, { "grad_norm": 0.03183206170797348, "learning_rate": 1.6084525842858856e-06, "loss": 0.0044, "step": 83730 }, { "grad_norm": 0.03105108253657818, "learning_rate": 1.603869295804844e-06, "loss": 0.0039, "step": 83740 }, { "grad_norm": 0.05034538730978966, "learning_rate": 1.5992924402456588e-06, "loss": 0.0067, "step": 83750 }, { "grad_norm": 0.022328106686472893, "learning_rate": 1.5947220182166877e-06, "loss": 0.0048, "step": 83760 }, { "grad_norm": 0.020207734778523445, "learning_rate": 1.5901580303254504e-06, "loss": 0.0049, "step": 83770 }, { "grad_norm": 0.024466298520565033, "learning_rate": 1.5856004771785948e-06, "loss": 0.0042, "step": 83780 }, { "grad_norm": 0.051330309361219406, "learning_rate": 1.581049359381931e-06, "loss": 0.0053, "step": 83790 }, { "grad_norm": 0.07336290180683136, "learning_rate": 1.5765046775403968e-06, "loss": 0.0061, "step": 83800 }, { "grad_norm": 0.0434468649327755, "learning_rate": 1.5719664322580763e-06, "loss": 0.0043, "step": 83810 }, { "grad_norm": 0.01816047541797161, "learning_rate": 1.5674346241382142e-06, "loss": 0.004, "step": 83820 }, { "grad_norm": 0.01741781085729599, "learning_rate": 1.562909253783179e-06, "loss": 0.0036, "step": 83830 }, { "grad_norm": 0.050848137587308884, "learning_rate": 1.5583903217945005e-06, "loss": 0.0029, "step": 83840 }, { "grad_norm": 0.0705425888299942, "learning_rate": 1.553877828772843e-06, "loss": 0.0055, "step": 83850 }, { "grad_norm": 0.05071704462170601, "learning_rate": 1.5493717753180149e-06, "loss": 0.0046, "step": 83860 }, { "grad_norm": 0.06970267742872238, "learning_rate": 1.5448721620289764e-06, "loss": 0.0055, "step": 83870 }, { "grad_norm": 0.03671732172369957, "learning_rate": 1.5403789895038156e-06, "loss": 0.004, "step": 83880 }, { "grad_norm": 0.05322708189487457, "learning_rate": 1.5358922583397829e-06, "loss": 0.0048, "step": 83890 }, { "grad_norm": 0.04800549894571304, "learning_rate": 1.531411969133273e-06, "loss": 0.006, "step": 83900 }, { "grad_norm": 0.0412217453122139, "learning_rate": 1.5269381224797986e-06, "loss": 0.0044, "step": 83910 }, { "grad_norm": 0.03295020014047623, "learning_rate": 1.5224707189740395e-06, "loss": 0.0046, "step": 83920 }, { "grad_norm": 0.06056177616119385, "learning_rate": 1.5180097592098097e-06, "loss": 0.0056, "step": 83930 }, { "grad_norm": 0.025878051295876503, "learning_rate": 1.5135552437800792e-06, "loss": 0.0074, "step": 83940 }, { "grad_norm": 0.03562050685286522, "learning_rate": 1.5091071732769524e-06, "loss": 0.0043, "step": 83950 }, { "grad_norm": 0.06460302323102951, "learning_rate": 1.5046655482916673e-06, "loss": 0.0051, "step": 83960 }, { "grad_norm": 0.043831001967191696, "learning_rate": 1.5002303694146237e-06, "loss": 0.0028, "step": 83970 }, { "grad_norm": 0.023910755291581154, "learning_rate": 1.4958016372353446e-06, "loss": 0.0036, "step": 83980 }, { "grad_norm": 0.02352670393884182, "learning_rate": 1.49137935234252e-06, "loss": 0.0043, "step": 83990 }, { "grad_norm": 0.03423989191651344, "learning_rate": 1.4869635153239581e-06, "loss": 0.0031, "step": 84000 }, { "grad_norm": 0.020715558901429176, "learning_rate": 1.4825541267666275e-06, "loss": 0.006, "step": 84010 }, { "grad_norm": 0.0214772317558527, "learning_rate": 1.4781511872566378e-06, "loss": 0.0049, "step": 84020 }, { "grad_norm": 0.025074170902371407, "learning_rate": 1.4737546973792315e-06, "loss": 0.0048, "step": 84030 }, { "grad_norm": 0.02264384552836418, "learning_rate": 1.4693646577188024e-06, "loss": 0.005, "step": 84040 }, { "grad_norm": 0.03584546595811844, "learning_rate": 1.4649810688588784e-06, "loss": 0.0032, "step": 84050 }, { "grad_norm": 0.057944051921367645, "learning_rate": 1.4606039313821485e-06, "loss": 0.0041, "step": 84060 }, { "grad_norm": 0.016965866088867188, "learning_rate": 1.4562332458704254e-06, "loss": 0.0039, "step": 84070 }, { "grad_norm": 0.04228769615292549, "learning_rate": 1.4518690129046609e-06, "loss": 0.0049, "step": 84080 }, { "grad_norm": 0.01929073967039585, "learning_rate": 1.4475112330649688e-06, "loss": 0.0025, "step": 84090 }, { "grad_norm": 0.0268859826028347, "learning_rate": 1.4431599069305968e-06, "loss": 0.0079, "step": 84100 }, { "grad_norm": 0.01836470514535904, "learning_rate": 1.4388150350799268e-06, "loss": 0.0055, "step": 84110 }, { "grad_norm": 0.05622231960296631, "learning_rate": 1.4344766180904967e-06, "loss": 0.0058, "step": 84120 }, { "grad_norm": 0.021035948768258095, "learning_rate": 1.430144656538973e-06, "loss": 0.0074, "step": 84130 }, { "grad_norm": 0.02550818957388401, "learning_rate": 1.425819151001162e-06, "loss": 0.0102, "step": 84140 }, { "grad_norm": 0.04463467746973038, "learning_rate": 1.4215001020520258e-06, "loss": 0.0039, "step": 84150 }, { "grad_norm": 0.036153361201286316, "learning_rate": 1.4171875102656719e-06, "loss": 0.0057, "step": 84160 }, { "grad_norm": 0.02058819867670536, "learning_rate": 1.412881376215336e-06, "loss": 0.0033, "step": 84170 }, { "grad_norm": 0.059149354696273804, "learning_rate": 1.4085817004733881e-06, "loss": 0.006, "step": 84180 }, { "grad_norm": 0.03769516944885254, "learning_rate": 1.4042884836113546e-06, "loss": 0.0035, "step": 84190 }, { "grad_norm": 0.029066601768136024, "learning_rate": 1.400001726199901e-06, "loss": 0.0037, "step": 84200 }, { "grad_norm": 0.04891059547662735, "learning_rate": 1.3957214288088382e-06, "loss": 0.0054, "step": 84210 }, { "grad_norm": 0.032636500895023346, "learning_rate": 1.3914475920071057e-06, "loss": 0.0072, "step": 84220 }, { "grad_norm": 0.03831084817647934, "learning_rate": 1.3871802163627989e-06, "loss": 0.0044, "step": 84230 }, { "grad_norm": 0.02204516902565956, "learning_rate": 1.3829193024431364e-06, "loss": 0.0037, "step": 84240 }, { "grad_norm": 0.05142366886138916, "learning_rate": 1.3786648508144984e-06, "loss": 0.0044, "step": 84250 }, { "grad_norm": 0.023198997601866722, "learning_rate": 1.3744168620423882e-06, "loss": 0.0047, "step": 84260 }, { "grad_norm": 0.021658679470419884, "learning_rate": 1.3701753366914649e-06, "loss": 0.0047, "step": 84270 }, { "grad_norm": 0.05891082063317299, "learning_rate": 1.3659402753255169e-06, "loss": 0.0048, "step": 84280 }, { "grad_norm": 0.03266018256545067, "learning_rate": 1.3617116785074825e-06, "loss": 0.0037, "step": 84290 }, { "grad_norm": 0.02353162132203579, "learning_rate": 1.357489546799423e-06, "loss": 0.0056, "step": 84300 }, { "grad_norm": 0.027875207364559174, "learning_rate": 1.3532738807625733e-06, "loss": 0.0044, "step": 84310 }, { "grad_norm": 0.024512270465493202, "learning_rate": 1.3490646809572738e-06, "loss": 0.0052, "step": 84320 }, { "grad_norm": 0.0427505262196064, "learning_rate": 1.3448619479430324e-06, "loss": 0.0057, "step": 84330 }, { "grad_norm": 0.027349194511771202, "learning_rate": 1.3406656822784802e-06, "loss": 0.0047, "step": 84340 }, { "grad_norm": 0.014648516662418842, "learning_rate": 1.3364758845213875e-06, "loss": 0.0043, "step": 84350 }, { "grad_norm": 0.040373917669057846, "learning_rate": 1.3322925552286812e-06, "loss": 0.0044, "step": 84360 }, { "grad_norm": 0.031381987035274506, "learning_rate": 1.3281156949564112e-06, "loss": 0.0047, "step": 84370 }, { "grad_norm": 0.028678536415100098, "learning_rate": 1.3239453042597826e-06, "loss": 0.0039, "step": 84380 }, { "grad_norm": 0.06210649386048317, "learning_rate": 1.3197813836931361e-06, "loss": 0.0057, "step": 84390 }, { "grad_norm": 0.030803726986050606, "learning_rate": 1.3156239338099397e-06, "loss": 0.0049, "step": 84400 }, { "grad_norm": 0.019549550488591194, "learning_rate": 1.3114729551628123e-06, "loss": 0.0042, "step": 84410 }, { "grad_norm": 0.049551840871572495, "learning_rate": 1.3073284483035131e-06, "loss": 0.0048, "step": 84420 }, { "grad_norm": 0.043178360909223557, "learning_rate": 1.3031904137829398e-06, "loss": 0.0043, "step": 84430 }, { "grad_norm": 0.028236154466867447, "learning_rate": 1.2990588521511249e-06, "loss": 0.0058, "step": 84440 }, { "grad_norm": 0.03563448786735535, "learning_rate": 1.2949337639572568e-06, "loss": 0.0044, "step": 84450 }, { "grad_norm": 0.03203743323683739, "learning_rate": 1.2908151497496356e-06, "loss": 0.0033, "step": 84460 }, { "grad_norm": 0.019889995455741882, "learning_rate": 1.2867030100757294e-06, "loss": 0.0043, "step": 84470 }, { "grad_norm": 0.06651626527309418, "learning_rate": 1.2825973454821282e-06, "loss": 0.0041, "step": 84480 }, { "grad_norm": 0.05299830064177513, "learning_rate": 1.2784981565145626e-06, "loss": 0.0038, "step": 84490 }, { "grad_norm": 0.04686484485864639, "learning_rate": 1.2744054437179131e-06, "loss": 0.0052, "step": 84500 }, { "grad_norm": 0.03250629082322121, "learning_rate": 1.2703192076361891e-06, "loss": 0.0061, "step": 84510 }, { "grad_norm": 0.037332430481910706, "learning_rate": 1.266239448812534e-06, "loss": 0.0065, "step": 84520 }, { "grad_norm": 0.029907003045082092, "learning_rate": 1.2621661677892528e-06, "loss": 0.0049, "step": 84530 }, { "grad_norm": 0.029535407200455666, "learning_rate": 1.258099365107762e-06, "loss": 0.0051, "step": 84540 }, { "grad_norm": 0.07099306583404541, "learning_rate": 1.254039041308641e-06, "loss": 0.0046, "step": 84550 }, { "grad_norm": 0.052670229226350784, "learning_rate": 1.2499851969316024e-06, "loss": 0.0041, "step": 84560 }, { "grad_norm": 0.031884852796792984, "learning_rate": 1.2459378325154702e-06, "loss": 0.0045, "step": 84570 }, { "grad_norm": 0.04396549239754677, "learning_rate": 1.2418969485982423e-06, "loss": 0.0045, "step": 84580 }, { "grad_norm": 0.05206827446818352, "learning_rate": 1.2378625457170445e-06, "loss": 0.0043, "step": 84590 }, { "grad_norm": 0.06078862398862839, "learning_rate": 1.2338346244081256e-06, "loss": 0.0046, "step": 84600 }, { "grad_norm": 0.030285673215985298, "learning_rate": 1.2298131852069072e-06, "loss": 0.0045, "step": 84610 }, { "grad_norm": 0.04761093854904175, "learning_rate": 1.2257982286479174e-06, "loss": 0.0059, "step": 84620 }, { "grad_norm": 0.05202842131257057, "learning_rate": 1.2217897552648239e-06, "loss": 0.0052, "step": 84630 }, { "grad_norm": 0.037098173052072525, "learning_rate": 1.2177877655904556e-06, "loss": 0.0044, "step": 84640 }, { "grad_norm": 0.020927103236317635, "learning_rate": 1.2137922601567543e-06, "loss": 0.0063, "step": 84650 }, { "grad_norm": 0.05478740483522415, "learning_rate": 1.2098032394948223e-06, "loss": 0.0047, "step": 84660 }, { "grad_norm": 0.03019506111741066, "learning_rate": 1.2058207041348856e-06, "loss": 0.0032, "step": 84670 }, { "grad_norm": 0.07055933028459549, "learning_rate": 1.20184465460631e-06, "loss": 0.0044, "step": 84680 }, { "grad_norm": 0.02249056100845337, "learning_rate": 1.1978750914376003e-06, "loss": 0.0049, "step": 84690 }, { "grad_norm": 0.034276265650987625, "learning_rate": 1.193912015156401e-06, "loss": 0.0064, "step": 84700 }, { "grad_norm": 0.05003572255373001, "learning_rate": 1.1899554262894908e-06, "loss": 0.0034, "step": 84710 }, { "grad_norm": 0.04360263794660568, "learning_rate": 1.186005325362799e-06, "loss": 0.0042, "step": 84720 }, { "grad_norm": 0.030183471739292145, "learning_rate": 1.1820617129013666e-06, "loss": 0.0059, "step": 84730 }, { "grad_norm": 0.06811028718948364, "learning_rate": 1.1781245894293857e-06, "loss": 0.0088, "step": 84740 }, { "grad_norm": 0.08664794266223907, "learning_rate": 1.1741939554702097e-06, "loss": 0.0068, "step": 84750 }, { "grad_norm": 0.024038510397076607, "learning_rate": 1.1702698115462872e-06, "loss": 0.0043, "step": 84760 }, { "grad_norm": 0.0503433458507061, "learning_rate": 1.1663521581792348e-06, "loss": 0.0048, "step": 84770 }, { "grad_norm": 0.037508200854063034, "learning_rate": 1.1624409958897908e-06, "loss": 0.0042, "step": 84780 }, { "grad_norm": 0.0300627201795578, "learning_rate": 1.1585363251978342e-06, "loss": 0.0031, "step": 84790 }, { "grad_norm": 0.03399338573217392, "learning_rate": 1.1546381466223833e-06, "loss": 0.0044, "step": 84800 }, { "grad_norm": 0.07516330480575562, "learning_rate": 1.1507464606815954e-06, "loss": 0.0062, "step": 84810 }, { "grad_norm": 0.0624440498650074, "learning_rate": 1.1468612678927625e-06, "loss": 0.0042, "step": 84820 }, { "grad_norm": 0.04062352329492569, "learning_rate": 1.142982568772305e-06, "loss": 0.006, "step": 84830 }, { "grad_norm": 0.033400218933820724, "learning_rate": 1.1391103638357992e-06, "loss": 0.0052, "step": 84840 }, { "grad_norm": 0.05759777873754501, "learning_rate": 1.1352446535979443e-06, "loss": 0.0068, "step": 84850 }, { "grad_norm": 0.02371343970298767, "learning_rate": 1.1313854385725741e-06, "loss": 0.0054, "step": 84860 }, { "grad_norm": 0.0472407341003418, "learning_rate": 1.1275327192726726e-06, "loss": 0.0065, "step": 84870 }, { "grad_norm": 0.020820895209908485, "learning_rate": 1.123686496210341e-06, "loss": 0.0056, "step": 84880 }, { "grad_norm": 0.042409755289554596, "learning_rate": 1.1198467698968374e-06, "loss": 0.0042, "step": 84890 }, { "grad_norm": 0.03137718886137009, "learning_rate": 1.116013540842542e-06, "loss": 0.0069, "step": 84900 }, { "grad_norm": 0.017310474067926407, "learning_rate": 1.1121868095569754e-06, "loss": 0.0034, "step": 84910 }, { "grad_norm": 0.031002243980765343, "learning_rate": 1.1083665765487972e-06, "loss": 0.0042, "step": 84920 }, { "grad_norm": 0.06766834855079651, "learning_rate": 1.104552842325801e-06, "loss": 0.0059, "step": 84930 }, { "grad_norm": 0.027872135862708092, "learning_rate": 1.1007456073949262e-06, "loss": 0.0062, "step": 84940 }, { "grad_norm": 0.04157332330942154, "learning_rate": 1.0969448722622234e-06, "loss": 0.0034, "step": 84950 }, { "grad_norm": 0.0311770960688591, "learning_rate": 1.0931506374328992e-06, "loss": 0.0066, "step": 84960 }, { "grad_norm": 0.04748283699154854, "learning_rate": 1.0893629034112895e-06, "loss": 0.0055, "step": 84970 }, { "grad_norm": 0.060664813965559006, "learning_rate": 1.0855816707008747e-06, "loss": 0.0045, "step": 84980 }, { "grad_norm": 0.06922247260808945, "learning_rate": 1.0818069398042697e-06, "loss": 0.0065, "step": 84990 }, { "grad_norm": 0.044726815074682236, "learning_rate": 1.0780387112232116e-06, "loss": 0.0039, "step": 85000 }, { "grad_norm": 0.017876839265227318, "learning_rate": 1.0742769854585778e-06, "loss": 0.0026, "step": 85010 }, { "grad_norm": 0.047101736068725586, "learning_rate": 1.0705217630103959e-06, "loss": 0.004, "step": 85020 }, { "grad_norm": 0.05892931669950485, "learning_rate": 1.0667730443778112e-06, "loss": 0.007, "step": 85030 }, { "grad_norm": 0.03040623851120472, "learning_rate": 1.063030830059114e-06, "loss": 0.0049, "step": 85040 }, { "grad_norm": 0.016214460134506226, "learning_rate": 1.0592951205517287e-06, "loss": 0.0055, "step": 85050 }, { "grad_norm": 0.045900847762823105, "learning_rate": 1.055565916352208e-06, "loss": 0.0051, "step": 85060 }, { "grad_norm": 0.06799096614122391, "learning_rate": 1.0518432179562553e-06, "loss": 0.0059, "step": 85070 }, { "grad_norm": 0.02935830131173134, "learning_rate": 1.0481270258586973e-06, "loss": 0.0052, "step": 85080 }, { "grad_norm": 0.044458698481321335, "learning_rate": 1.044417340553494e-06, "loss": 0.0073, "step": 85090 }, { "grad_norm": 0.023868586868047714, "learning_rate": 1.0407141625337512e-06, "loss": 0.004, "step": 85100 }, { "grad_norm": 0.019587373360991478, "learning_rate": 1.037017492291703e-06, "loss": 0.0058, "step": 85110 }, { "grad_norm": 0.027554361149668694, "learning_rate": 1.0333273303187118e-06, "loss": 0.0055, "step": 85120 }, { "grad_norm": 0.04561428725719452, "learning_rate": 1.0296436771052853e-06, "loss": 0.0055, "step": 85130 }, { "grad_norm": 0.031292472034692764, "learning_rate": 1.0259665331410706e-06, "loss": 0.0035, "step": 85140 }, { "grad_norm": 0.038108084350824356, "learning_rate": 1.0222958989148323e-06, "loss": 0.004, "step": 85150 }, { "grad_norm": 0.03475523367524147, "learning_rate": 1.0186317749144857e-06, "loss": 0.0055, "step": 85160 }, { "grad_norm": 0.021657776087522507, "learning_rate": 1.014974161627069e-06, "loss": 0.0048, "step": 85170 }, { "grad_norm": 0.04801961034536362, "learning_rate": 1.0113230595387657e-06, "loss": 0.0043, "step": 85180 }, { "grad_norm": 0.040807876735925674, "learning_rate": 1.0076784691348818e-06, "loss": 0.0045, "step": 85190 }, { "grad_norm": 0.022236166521906853, "learning_rate": 1.0040403908998686e-06, "loss": 0.005, "step": 85200 }, { "grad_norm": 0.05328996852040291, "learning_rate": 1.0004088253173116e-06, "loss": 0.0042, "step": 85210 }, { "grad_norm": 0.06187993288040161, "learning_rate": 9.967837728699193e-07, "loss": 0.0053, "step": 85220 }, { "grad_norm": 0.029368488118052483, "learning_rate": 9.931652340395448e-07, "loss": 0.0033, "step": 85230 }, { "grad_norm": 0.01562986709177494, "learning_rate": 9.895532093071757e-07, "loss": 0.0046, "step": 85240 }, { "grad_norm": 0.05410890281200409, "learning_rate": 9.859476991529226e-07, "loss": 0.0057, "step": 85250 }, { "grad_norm": 0.05805293470621109, "learning_rate": 9.823487040560464e-07, "loss": 0.006, "step": 85260 }, { "grad_norm": 0.01786484196782112, "learning_rate": 9.787562244949255e-07, "loss": 0.0058, "step": 85270 }, { "grad_norm": 0.059214673936367035, "learning_rate": 9.751702609470835e-07, "loss": 0.0049, "step": 85280 }, { "grad_norm": 0.06130485609173775, "learning_rate": 9.715908138891783e-07, "loss": 0.0032, "step": 85290 }, { "grad_norm": 0.04676489904522896, "learning_rate": 9.680178837969956e-07, "loss": 0.0053, "step": 85300 }, { "grad_norm": 0.041589368134737015, "learning_rate": 9.64451471145461e-07, "loss": 0.0072, "step": 85310 }, { "grad_norm": 0.030925296247005463, "learning_rate": 9.60891576408618e-07, "loss": 0.0032, "step": 85320 }, { "grad_norm": 0.027358558028936386, "learning_rate": 9.57338200059671e-07, "loss": 0.0051, "step": 85330 }, { "grad_norm": 0.035922821611166, "learning_rate": 9.537913425709311e-07, "loss": 0.0041, "step": 85340 }, { "grad_norm": 0.03175503388047218, "learning_rate": 9.502510044138602e-07, "loss": 0.0057, "step": 85350 }, { "grad_norm": 0.016166478395462036, "learning_rate": 9.467171860590429e-07, "loss": 0.0052, "step": 85360 }, { "grad_norm": 0.045774366706609726, "learning_rate": 9.431898879762092e-07, "loss": 0.0055, "step": 85370 }, { "grad_norm": 0.02004963718354702, "learning_rate": 9.39669110634217e-07, "loss": 0.0051, "step": 85380 }, { "grad_norm": 0.035994913429021835, "learning_rate": 9.361548545010479e-07, "loss": 0.0058, "step": 85390 }, { "grad_norm": 0.05847076326608658, "learning_rate": 9.32647120043828e-07, "loss": 0.0048, "step": 85400 }, { "grad_norm": 0.031701650470495224, "learning_rate": 9.291459077288067e-07, "loss": 0.0057, "step": 85410 }, { "grad_norm": 0.053478579968214035, "learning_rate": 9.256512180213839e-07, "loss": 0.0042, "step": 85420 }, { "grad_norm": 0.051211509853601456, "learning_rate": 9.221630513860824e-07, "loss": 0.0046, "step": 85430 }, { "grad_norm": 0.024439703673124313, "learning_rate": 9.186814082865424e-07, "loss": 0.004, "step": 85440 }, { "grad_norm": 0.08873621374368668, "learning_rate": 9.152062891855662e-07, "loss": 0.0057, "step": 85450 }, { "grad_norm": 0.051291488111019135, "learning_rate": 9.117376945450728e-07, "loss": 0.0041, "step": 85460 }, { "grad_norm": 0.042030125856399536, "learning_rate": 9.082756248261048e-07, "loss": 0.004, "step": 85470 }, { "grad_norm": 0.0486554317176342, "learning_rate": 9.048200804888607e-07, "loss": 0.0049, "step": 85480 }, { "grad_norm": 0.020330771803855896, "learning_rate": 9.013710619926507e-07, "loss": 0.0058, "step": 85490 }, { "grad_norm": 0.029893822968006134, "learning_rate": 8.979285697959305e-07, "loss": 0.0046, "step": 85500 }, { "grad_norm": 0.03035529889166355, "learning_rate": 8.94492604356284e-07, "loss": 0.0028, "step": 85510 }, { "grad_norm": 0.031191624701023102, "learning_rate": 8.910631661304236e-07, "loss": 0.0039, "step": 85520 }, { "grad_norm": 0.02032921090722084, "learning_rate": 8.876402555742014e-07, "loss": 0.0028, "step": 85530 }, { "grad_norm": 0.06385728716850281, "learning_rate": 8.842238731425978e-07, "loss": 0.0057, "step": 85540 }, { "grad_norm": 0.03771044686436653, "learning_rate": 8.808140192897274e-07, "loss": 0.0061, "step": 85550 }, { "grad_norm": 0.0719406008720398, "learning_rate": 8.774106944688277e-07, "loss": 0.0057, "step": 85560 }, { "grad_norm": 0.020552871748805046, "learning_rate": 8.740138991322811e-07, "loss": 0.0071, "step": 85570 }, { "grad_norm": 0.06838308274745941, "learning_rate": 8.706236337316043e-07, "loss": 0.0045, "step": 85580 }, { "grad_norm": 0.0396365150809288, "learning_rate": 8.672398987174368e-07, "loss": 0.0028, "step": 85590 }, { "grad_norm": 0.014718500897288322, "learning_rate": 8.638626945395411e-07, "loss": 0.0046, "step": 85600 }, { "grad_norm": 0.05325590446591377, "learning_rate": 8.604920216468304e-07, "loss": 0.005, "step": 85610 }, { "grad_norm": 0.036138180643320084, "learning_rate": 8.571278804873462e-07, "loss": 0.0042, "step": 85620 }, { "grad_norm": 0.02897372469305992, "learning_rate": 8.537702715082474e-07, "loss": 0.0055, "step": 85630 }, { "grad_norm": 0.057734716683626175, "learning_rate": 8.504191951558382e-07, "loss": 0.005, "step": 85640 }, { "grad_norm": 0.04120597988367081, "learning_rate": 8.470746518755624e-07, "loss": 0.0034, "step": 85650 }, { "grad_norm": 0.028122926130890846, "learning_rate": 8.437366421119752e-07, "loss": 0.0041, "step": 85660 }, { "grad_norm": 0.03436555340886116, "learning_rate": 8.404051663087664e-07, "loss": 0.0049, "step": 85670 }, { "grad_norm": 0.03188968077301979, "learning_rate": 8.370802249087706e-07, "loss": 0.0058, "step": 85680 }, { "grad_norm": 0.044002071022987366, "learning_rate": 8.337618183539453e-07, "loss": 0.0065, "step": 85690 }, { "grad_norm": 0.07670425623655319, "learning_rate": 8.304499470853821e-07, "loss": 0.0057, "step": 85700 }, { "grad_norm": 0.02101736143231392, "learning_rate": 8.271446115433013e-07, "loss": 0.0034, "step": 85710 }, { "grad_norm": 0.02717933990061283, "learning_rate": 8.238458121670511e-07, "loss": 0.005, "step": 85720 }, { "grad_norm": 0.06788770854473114, "learning_rate": 8.2055354939512e-07, "loss": 0.0038, "step": 85730 }, { "grad_norm": 0.02011154778301716, "learning_rate": 8.172678236651243e-07, "loss": 0.0041, "step": 85740 }, { "grad_norm": 0.054494887590408325, "learning_rate": 8.139886354138038e-07, "loss": 0.0042, "step": 85750 }, { "grad_norm": 0.04356217011809349, "learning_rate": 8.107159850770485e-07, "loss": 0.0048, "step": 85760 }, { "grad_norm": 0.04280930384993553, "learning_rate": 8.074498730898494e-07, "loss": 0.0033, "step": 85770 }, { "grad_norm": 0.028010457754135132, "learning_rate": 8.041902998863482e-07, "loss": 0.0063, "step": 85780 }, { "grad_norm": 0.025732828304171562, "learning_rate": 8.00937265899826e-07, "loss": 0.0053, "step": 85790 }, { "grad_norm": 0.052282657474279404, "learning_rate": 7.976907715626758e-07, "loss": 0.005, "step": 85800 }, { "grad_norm": 0.01947810873389244, "learning_rate": 7.944508173064357e-07, "loss": 0.0049, "step": 85810 }, { "grad_norm": 0.03515009209513664, "learning_rate": 7.912174035617559e-07, "loss": 0.0032, "step": 85820 }, { "grad_norm": 0.0539536289870739, "learning_rate": 7.879905307584368e-07, "loss": 0.0102, "step": 85830 }, { "grad_norm": 0.026193246245384216, "learning_rate": 7.84770199325402e-07, "loss": 0.0037, "step": 85840 }, { "grad_norm": 0.0649101510643959, "learning_rate": 7.81556409690698e-07, "loss": 0.0047, "step": 85850 }, { "grad_norm": 0.029328124597668648, "learning_rate": 7.783491622815164e-07, "loss": 0.0032, "step": 85860 }, { "grad_norm": 0.06164994090795517, "learning_rate": 7.751484575241719e-07, "loss": 0.0042, "step": 85870 }, { "grad_norm": 0.04486147686839104, "learning_rate": 7.719542958441018e-07, "loss": 0.0035, "step": 85880 }, { "grad_norm": 0.02741813287138939, "learning_rate": 7.687666776658886e-07, "loss": 0.0041, "step": 85890 }, { "grad_norm": 0.059076860547065735, "learning_rate": 7.655856034132325e-07, "loss": 0.0043, "step": 85900 }, { "grad_norm": 0.053878191858530045, "learning_rate": 7.624110735089784e-07, "loss": 0.0031, "step": 85910 }, { "grad_norm": 0.04633161798119545, "learning_rate": 7.592430883750778e-07, "loss": 0.0056, "step": 85920 }, { "grad_norm": 0.028596555814146996, "learning_rate": 7.560816484326383e-07, "loss": 0.008, "step": 85930 }, { "grad_norm": 0.021837443113327026, "learning_rate": 7.529267541018792e-07, "loss": 0.0036, "step": 85940 }, { "grad_norm": 0.04352568835020065, "learning_rate": 7.49778405802154e-07, "loss": 0.0049, "step": 85950 }, { "grad_norm": 0.02514874003827572, "learning_rate": 7.466366039519556e-07, "loss": 0.0034, "step": 85960 }, { "grad_norm": 0.04676235094666481, "learning_rate": 7.435013489689002e-07, "loss": 0.0046, "step": 85970 }, { "grad_norm": 0.05120566859841347, "learning_rate": 7.403726412697265e-07, "loss": 0.0049, "step": 85980 }, { "grad_norm": 0.03610750287771225, "learning_rate": 7.372504812703073e-07, "loss": 0.0064, "step": 85990 }, { "grad_norm": 0.022917235270142555, "learning_rate": 7.341348693856553e-07, "loss": 0.0039, "step": 86000 }, { "grad_norm": 0.05281160771846771, "learning_rate": 7.310258060298891e-07, "loss": 0.0032, "step": 86010 }, { "grad_norm": 0.09649229794740677, "learning_rate": 7.279232916162948e-07, "loss": 0.0062, "step": 86020 }, { "grad_norm": 0.028868425637483597, "learning_rate": 7.248273265572536e-07, "loss": 0.0047, "step": 86030 }, { "grad_norm": 0.024648936465382576, "learning_rate": 7.217379112642863e-07, "loss": 0.0064, "step": 86040 }, { "grad_norm": 0.052989210933446884, "learning_rate": 7.186550461480479e-07, "loss": 0.0036, "step": 86050 }, { "grad_norm": 0.04725837707519531, "learning_rate": 7.155787316183216e-07, "loss": 0.0046, "step": 86060 }, { "grad_norm": 0.020874032750725746, "learning_rate": 7.125089680840136e-07, "loss": 0.0067, "step": 86070 }, { "grad_norm": 0.019701287150382996, "learning_rate": 7.094457559531642e-07, "loss": 0.0053, "step": 86080 }, { "grad_norm": 0.019537437707185745, "learning_rate": 7.063890956329478e-07, "loss": 0.0044, "step": 86090 }, { "grad_norm": 0.048355016857385635, "learning_rate": 7.033389875296559e-07, "loss": 0.0051, "step": 86100 }, { "grad_norm": 0.02866346389055252, "learning_rate": 7.002954320487198e-07, "loss": 0.0046, "step": 86110 }, { "grad_norm": 0.033358294516801834, "learning_rate": 6.972584295946882e-07, "loss": 0.0054, "step": 86120 }, { "grad_norm": 0.02902344986796379, "learning_rate": 6.942279805712603e-07, "loss": 0.0053, "step": 86130 }, { "grad_norm": 0.05047035589814186, "learning_rate": 6.912040853812418e-07, "loss": 0.0066, "step": 86140 }, { "grad_norm": 0.07035980373620987, "learning_rate": 6.881867444265777e-07, "loss": 0.0041, "step": 86150 }, { "grad_norm": 0.04842603951692581, "learning_rate": 6.851759581083306e-07, "loss": 0.0047, "step": 86160 }, { "grad_norm": 0.019430220127105713, "learning_rate": 6.82171726826708e-07, "loss": 0.0071, "step": 86170 }, { "grad_norm": 0.030998459085822105, "learning_rate": 6.791740509810463e-07, "loss": 0.0039, "step": 86180 }, { "grad_norm": 0.05374986305832863, "learning_rate": 6.761829309697931e-07, "loss": 0.0046, "step": 86190 }, { "grad_norm": 0.061376601457595825, "learning_rate": 6.731983671905417e-07, "loss": 0.0058, "step": 86200 }, { "grad_norm": 0.024419136345386505, "learning_rate": 6.702203600400025e-07, "loss": 0.0037, "step": 86210 }, { "grad_norm": 0.01938471570611, "learning_rate": 6.672489099140144e-07, "loss": 0.0026, "step": 86220 }, { "grad_norm": 0.01563810370862484, "learning_rate": 6.64284017207556e-07, "loss": 0.0052, "step": 86230 }, { "grad_norm": 0.026399344205856323, "learning_rate": 6.613256823147284e-07, "loss": 0.0054, "step": 86240 }, { "grad_norm": 0.04054974764585495, "learning_rate": 6.583739056287619e-07, "loss": 0.0038, "step": 86250 }, { "grad_norm": 0.03567458316683769, "learning_rate": 6.554286875420035e-07, "loss": 0.0054, "step": 86260 }, { "grad_norm": 0.0425528958439827, "learning_rate": 6.524900284459456e-07, "loss": 0.0049, "step": 86270 }, { "grad_norm": 0.023969070985913277, "learning_rate": 6.495579287311982e-07, "loss": 0.0047, "step": 86280 }, { "grad_norm": 0.028951134532690048, "learning_rate": 6.466323887875048e-07, "loss": 0.0047, "step": 86290 }, { "grad_norm": 0.06974896043539047, "learning_rate": 6.437134090037323e-07, "loss": 0.0057, "step": 86300 }, { "grad_norm": 0.016532307490706444, "learning_rate": 6.408009897678813e-07, "loss": 0.0056, "step": 86310 }, { "grad_norm": 0.060314908623695374, "learning_rate": 6.378951314670755e-07, "loss": 0.006, "step": 86320 }, { "grad_norm": 0.037388406693935394, "learning_rate": 6.349958344875617e-07, "loss": 0.0038, "step": 86330 }, { "grad_norm": 0.02885439805686474, "learning_rate": 6.32103099214737e-07, "loss": 0.0084, "step": 86340 }, { "grad_norm": 0.016330135986208916, "learning_rate": 6.292169260330938e-07, "loss": 0.0036, "step": 86350 }, { "grad_norm": 0.06201426312327385, "learning_rate": 6.263373153262753e-07, "loss": 0.0067, "step": 86360 }, { "grad_norm": 0.043731689453125, "learning_rate": 6.234642674770475e-07, "loss": 0.0033, "step": 86370 }, { "grad_norm": 0.06089938431978226, "learning_rate": 6.205977828672993e-07, "loss": 0.0046, "step": 86380 }, { "grad_norm": 0.06754724681377411, "learning_rate": 6.177378618780538e-07, "loss": 0.0066, "step": 86390 }, { "grad_norm": 0.03423088788986206, "learning_rate": 6.148845048894513e-07, "loss": 0.006, "step": 86400 }, { "grad_norm": 0.039261214435100555, "learning_rate": 6.120377122807719e-07, "loss": 0.0027, "step": 86410 }, { "grad_norm": 0.07355909049510956, "learning_rate": 6.091974844304238e-07, "loss": 0.006, "step": 86420 }, { "grad_norm": 0.04739426448941231, "learning_rate": 6.063638217159218e-07, "loss": 0.0072, "step": 86430 }, { "grad_norm": 0.036514762789011, "learning_rate": 6.035367245139311e-07, "loss": 0.0051, "step": 86440 }, { "grad_norm": 0.05225841701030731, "learning_rate": 6.007161932002348e-07, "loss": 0.0045, "step": 86450 }, { "grad_norm": 0.021310100331902504, "learning_rate": 5.979022281497437e-07, "loss": 0.0076, "step": 86460 }, { "grad_norm": 0.04065753519535065, "learning_rate": 5.950948297364978e-07, "loss": 0.0072, "step": 86470 }, { "grad_norm": 0.046092476695775986, "learning_rate": 5.922939983336651e-07, "loss": 0.0044, "step": 86480 }, { "grad_norm": 0.043341297656297684, "learning_rate": 5.894997343135311e-07, "loss": 0.0041, "step": 86490 }, { "grad_norm": 0.022730126976966858, "learning_rate": 5.867120380475211e-07, "loss": 0.0045, "step": 86500 }, { "grad_norm": 0.023470953106880188, "learning_rate": 5.839309099061774e-07, "loss": 0.0041, "step": 86510 }, { "grad_norm": 0.045928649604320526, "learning_rate": 5.811563502591766e-07, "loss": 0.0043, "step": 86520 }, { "grad_norm": 0.057955577969551086, "learning_rate": 5.783883594753236e-07, "loss": 0.0049, "step": 86530 }, { "grad_norm": 0.03486400842666626, "learning_rate": 5.75626937922541e-07, "loss": 0.0074, "step": 86540 }, { "grad_norm": 0.038689177483320236, "learning_rate": 5.728720859678794e-07, "loss": 0.0055, "step": 86550 }, { "grad_norm": 0.030048461630940437, "learning_rate": 5.701238039775292e-07, "loss": 0.0075, "step": 86560 }, { "grad_norm": 0.07206372916698456, "learning_rate": 5.673820923167928e-07, "loss": 0.0048, "step": 86570 }, { "grad_norm": 0.04447779059410095, "learning_rate": 5.646469513501063e-07, "loss": 0.0035, "step": 86580 }, { "grad_norm": 0.02444351278245449, "learning_rate": 5.619183814410289e-07, "loss": 0.0036, "step": 86590 }, { "grad_norm": 0.06556013971567154, "learning_rate": 5.591963829522429e-07, "loss": 0.0063, "step": 86600 }, { "grad_norm": 0.046575501561164856, "learning_rate": 5.564809562455753e-07, "loss": 0.0044, "step": 86610 }, { "grad_norm": 0.02964073047041893, "learning_rate": 5.537721016819597e-07, "loss": 0.0031, "step": 86620 }, { "grad_norm": 0.017778733745217323, "learning_rate": 5.510698196214636e-07, "loss": 0.0058, "step": 86630 }, { "grad_norm": 0.03471004590392113, "learning_rate": 5.483741104232776e-07, "loss": 0.0053, "step": 86640 }, { "grad_norm": 0.06350671499967575, "learning_rate": 5.456849744457259e-07, "loss": 0.0069, "step": 86650 }, { "grad_norm": 0.04307149723172188, "learning_rate": 5.430024120462506e-07, "loss": 0.0039, "step": 86660 }, { "grad_norm": 0.04789598658680916, "learning_rate": 5.403264235814276e-07, "loss": 0.0047, "step": 86670 }, { "grad_norm": 0.04961100593209267, "learning_rate": 5.376570094069444e-07, "loss": 0.0058, "step": 86680 }, { "grad_norm": 0.05813879147171974, "learning_rate": 5.34994169877645e-07, "loss": 0.0059, "step": 86690 }, { "grad_norm": 0.03143748641014099, "learning_rate": 5.323379053474686e-07, "loss": 0.0054, "step": 86700 }, { "grad_norm": 0.04097145423293114, "learning_rate": 5.296882161694938e-07, "loss": 0.0038, "step": 86710 }, { "grad_norm": 0.017028525471687317, "learning_rate": 5.270451026959167e-07, "loss": 0.0047, "step": 86720 }, { "grad_norm": 0.07395593822002411, "learning_rate": 5.244085652780728e-07, "loss": 0.0075, "step": 86730 }, { "grad_norm": 0.03361857682466507, "learning_rate": 5.217786042664152e-07, "loss": 0.0049, "step": 86740 }, { "grad_norm": 0.057096488773822784, "learning_rate": 5.191552200105254e-07, "loss": 0.0039, "step": 86750 }, { "grad_norm": 0.05764615163207054, "learning_rate": 5.165384128591078e-07, "loss": 0.0068, "step": 86760 }, { "grad_norm": 0.012747948057949543, "learning_rate": 5.139281831599951e-07, "loss": 0.0034, "step": 86770 }, { "grad_norm": 0.0361027754843235, "learning_rate": 5.113245312601433e-07, "loss": 0.0052, "step": 86780 }, { "grad_norm": 0.012214954942464828, "learning_rate": 5.087274575056367e-07, "loss": 0.0039, "step": 86790 }, { "grad_norm": 0.025987010449171066, "learning_rate": 5.061369622416879e-07, "loss": 0.0033, "step": 86800 }, { "grad_norm": 0.053825635462999344, "learning_rate": 5.035530458126215e-07, "loss": 0.0049, "step": 86810 }, { "grad_norm": 0.020077351480722427, "learning_rate": 5.009757085619016e-07, "loss": 0.0049, "step": 86820 }, { "grad_norm": 0.03649259731173515, "learning_rate": 4.98404950832121e-07, "loss": 0.0051, "step": 86830 }, { "grad_norm": 0.04669400304555893, "learning_rate": 4.958407729649838e-07, "loss": 0.0037, "step": 86840 }, { "grad_norm": 0.029293200001120567, "learning_rate": 4.93283175301329e-07, "loss": 0.0037, "step": 86850 }, { "grad_norm": 0.053153492510318756, "learning_rate": 4.907321581811176e-07, "loss": 0.0047, "step": 86860 }, { "grad_norm": 0.017795436084270477, "learning_rate": 4.881877219434283e-07, "loss": 0.004, "step": 86870 }, { "grad_norm": 0.025786306709051132, "learning_rate": 4.856498669264853e-07, "loss": 0.0042, "step": 86880 }, { "grad_norm": 0.05603308603167534, "learning_rate": 4.831185934676241e-07, "loss": 0.0036, "step": 86890 }, { "grad_norm": 0.012407447211444378, "learning_rate": 4.805939019032979e-07, "loss": 0.0037, "step": 86900 }, { "grad_norm": 0.027208013460040092, "learning_rate": 4.780757925691048e-07, "loss": 0.0041, "step": 86910 }, { "grad_norm": 0.032055649906396866, "learning_rate": 4.75564265799755e-07, "loss": 0.0036, "step": 86920 }, { "grad_norm": 0.019325409084558487, "learning_rate": 4.730593219290813e-07, "loss": 0.0035, "step": 86930 }, { "grad_norm": 0.03381694108247757, "learning_rate": 4.705609612900508e-07, "loss": 0.0048, "step": 86940 }, { "grad_norm": 0.023464620113372803, "learning_rate": 4.6806918421475333e-07, "loss": 0.0046, "step": 86950 }, { "grad_norm": 0.012520911172032356, "learning_rate": 4.655839910343962e-07, "loss": 0.0055, "step": 86960 }, { "grad_norm": 0.01828000135719776, "learning_rate": 4.631053820793152e-07, "loss": 0.003, "step": 86970 }, { "grad_norm": 0.03829207643866539, "learning_rate": 4.606333576789801e-07, "loss": 0.0061, "step": 86980 }, { "grad_norm": 0.04633694514632225, "learning_rate": 4.5816791816197244e-07, "loss": 0.0056, "step": 86990 }, { "grad_norm": 0.026880092918872833, "learning_rate": 4.55709063856008e-07, "loss": 0.0036, "step": 87000 }, { "grad_norm": 0.03147450089454651, "learning_rate": 4.5325679508791963e-07, "loss": 0.006, "step": 87010 }, { "grad_norm": 0.0298606026917696, "learning_rate": 4.5081111218367444e-07, "loss": 0.0039, "step": 87020 }, { "grad_norm": 0.04795082286000252, "learning_rate": 4.483720154683513e-07, "loss": 0.003, "step": 87030 }, { "grad_norm": 0.04989201948046684, "learning_rate": 4.459395052661575e-07, "loss": 0.0055, "step": 87040 }, { "grad_norm": 0.060760244727134705, "learning_rate": 4.4351358190043435e-07, "loss": 0.0039, "step": 87050 }, { "grad_norm": 0.034355998039245605, "learning_rate": 4.4109424569364066e-07, "loss": 0.0046, "step": 87060 }, { "grad_norm": 0.08030426502227783, "learning_rate": 4.3868149696735804e-07, "loss": 0.0056, "step": 87070 }, { "grad_norm": 0.04377884045243263, "learning_rate": 4.3627533604229665e-07, "loss": 0.0038, "step": 87080 }, { "grad_norm": 0.06748659163713455, "learning_rate": 4.338757632382839e-07, "loss": 0.0055, "step": 87090 }, { "grad_norm": 0.04154304042458534, "learning_rate": 4.314827788742815e-07, "loss": 0.0066, "step": 87100 }, { "grad_norm": 0.0368623286485672, "learning_rate": 4.290963832683681e-07, "loss": 0.0052, "step": 87110 }, { "grad_norm": 0.018260881304740906, "learning_rate": 4.267165767377512e-07, "loss": 0.0073, "step": 87120 }, { "grad_norm": 0.039620138704776764, "learning_rate": 4.243433595987556e-07, "loss": 0.0036, "step": 87130 }, { "grad_norm": 0.034972183406353, "learning_rate": 4.2197673216683995e-07, "loss": 0.0057, "step": 87140 }, { "grad_norm": 0.07178875058889389, "learning_rate": 4.1961669475657493e-07, "loss": 0.0045, "step": 87150 }, { "grad_norm": 0.026101481169462204, "learning_rate": 4.1726324768166516e-07, "loss": 0.0045, "step": 87160 }, { "grad_norm": 0.05162988230586052, "learning_rate": 4.1491639125493807e-07, "loss": 0.0073, "step": 87170 }, { "grad_norm": 0.021927360445261, "learning_rate": 4.1257612578834424e-07, "loss": 0.0035, "step": 87180 }, { "grad_norm": 0.04001893475651741, "learning_rate": 4.102424515929515e-07, "loss": 0.0047, "step": 87190 }, { "grad_norm": 0.02384888380765915, "learning_rate": 4.079153689789561e-07, "loss": 0.0073, "step": 87200 }, { "grad_norm": 0.04692738130688667, "learning_rate": 4.055948782556884e-07, "loss": 0.008, "step": 87210 }, { "grad_norm": 0.03313271701335907, "learning_rate": 4.0328097973158506e-07, "loss": 0.0038, "step": 87220 }, { "grad_norm": 0.01933399774134159, "learning_rate": 4.009736737142167e-07, "loss": 0.0048, "step": 87230 }, { "grad_norm": 0.019231418147683144, "learning_rate": 3.9867296051028235e-07, "loss": 0.0048, "step": 87240 }, { "grad_norm": 0.01700221747159958, "learning_rate": 3.963788404255875e-07, "loss": 0.0053, "step": 87250 }, { "grad_norm": 0.02292218804359436, "learning_rate": 3.9409131376507145e-07, "loss": 0.0046, "step": 87260 }, { "grad_norm": 0.04043884202837944, "learning_rate": 3.9181038083280217e-07, "loss": 0.006, "step": 87270 }, { "grad_norm": 0.03869851678609848, "learning_rate": 3.8953604193197046e-07, "loss": 0.0038, "step": 87280 }, { "grad_norm": 0.04106512665748596, "learning_rate": 3.8726829736488445e-07, "loss": 0.0056, "step": 87290 }, { "grad_norm": 0.04332079365849495, "learning_rate": 3.850071474329753e-07, "loss": 0.0047, "step": 87300 }, { "grad_norm": 0.022061293944716454, "learning_rate": 3.8275259243679693e-07, "loss": 0.0047, "step": 87310 }, { "grad_norm": 0.027580298483371735, "learning_rate": 3.80504632676032e-07, "loss": 0.005, "step": 87320 }, { "grad_norm": 0.023056184872984886, "learning_rate": 3.782632684494858e-07, "loss": 0.0042, "step": 87330 }, { "grad_norm": 0.04900350421667099, "learning_rate": 3.760285000550867e-07, "loss": 0.0038, "step": 87340 }, { "grad_norm": 0.03604743629693985, "learning_rate": 3.738003277898805e-07, "loss": 0.0032, "step": 87350 }, { "grad_norm": 0.03143826872110367, "learning_rate": 3.715787519500469e-07, "loss": 0.0051, "step": 87360 }, { "grad_norm": 0.03273957222700119, "learning_rate": 3.693637728308774e-07, "loss": 0.006, "step": 87370 }, { "grad_norm": 0.06780905276536942, "learning_rate": 3.671553907267977e-07, "loss": 0.0064, "step": 87380 }, { "grad_norm": 0.08284365385770798, "learning_rate": 3.649536059313452e-07, "loss": 0.0051, "step": 87390 }, { "grad_norm": 0.04200746491551399, "learning_rate": 3.627584187371913e-07, "loss": 0.0043, "step": 87400 }, { "grad_norm": 0.04280472546815872, "learning_rate": 3.605698294361193e-07, "loss": 0.0038, "step": 87410 }, { "grad_norm": 0.04573797434568405, "learning_rate": 3.583878383190464e-07, "loss": 0.0047, "step": 87420 }, { "grad_norm": 0.021841490641236305, "learning_rate": 3.5621244567600187e-07, "loss": 0.0042, "step": 87430 }, { "grad_norm": 0.015283158048987389, "learning_rate": 3.5404365179614873e-07, "loss": 0.0053, "step": 87440 }, { "grad_norm": 0.05802382901310921, "learning_rate": 3.5188145696776753e-07, "loss": 0.0036, "step": 87450 }, { "grad_norm": 0.03754276782274246, "learning_rate": 3.4972586147826727e-07, "loss": 0.0051, "step": 87460 }, { "grad_norm": 0.02775260992348194, "learning_rate": 3.4757686561416316e-07, "loss": 0.0042, "step": 87470 }, { "grad_norm": 0.021267535164952278, "learning_rate": 3.4543446966111006e-07, "loss": 0.0028, "step": 87480 }, { "grad_norm": 0.09003613144159317, "learning_rate": 3.432986739038746e-07, "loss": 0.0048, "step": 87490 }, { "grad_norm": 0.04570307955145836, "learning_rate": 3.4116947862636307e-07, "loss": 0.0055, "step": 87500 }, { "grad_norm": 0.05971677601337433, "learning_rate": 3.390468841115879e-07, "loss": 0.0059, "step": 87510 }, { "grad_norm": 0.03578362613916397, "learning_rate": 3.369308906416846e-07, "loss": 0.005, "step": 87520 }, { "grad_norm": 0.04295724630355835, "learning_rate": 3.34821498497917e-07, "loss": 0.005, "step": 87530 }, { "grad_norm": 0.028471050783991814, "learning_rate": 3.327187079606775e-07, "loss": 0.0036, "step": 87540 }, { "grad_norm": 0.029719101265072823, "learning_rate": 3.3062251930946477e-07, "loss": 0.0049, "step": 87550 }, { "grad_norm": 0.020068179816007614, "learning_rate": 3.2853293282291145e-07, "loss": 0.005, "step": 87560 }, { "grad_norm": 0.029762843623757362, "learning_rate": 3.264499487787731e-07, "loss": 0.003, "step": 87570 }, { "grad_norm": 0.03401516377925873, "learning_rate": 3.243735674539228e-07, "loss": 0.0042, "step": 87580 }, { "grad_norm": 0.1069394126534462, "learning_rate": 3.2230378912435635e-07, "loss": 0.0066, "step": 87590 }, { "grad_norm": 0.026009704917669296, "learning_rate": 3.202406140651981e-07, "loss": 0.0081, "step": 87600 }, { "grad_norm": 0.02167973294854164, "learning_rate": 3.181840425506788e-07, "loss": 0.0036, "step": 87610 }, { "grad_norm": 0.02677765302360058, "learning_rate": 3.1613407485417965e-07, "loss": 0.0077, "step": 87620 }, { "grad_norm": 0.04008402302861214, "learning_rate": 3.1409071124817167e-07, "loss": 0.0046, "step": 87630 }, { "grad_norm": 0.05285220965743065, "learning_rate": 3.120539520042654e-07, "loss": 0.0036, "step": 87640 }, { "grad_norm": 0.043220825493335724, "learning_rate": 3.1002379739319966e-07, "loss": 0.0036, "step": 87650 }, { "grad_norm": 0.03880148008465767, "learning_rate": 3.080002476848254e-07, "loss": 0.0046, "step": 87660 }, { "grad_norm": 0.036047689616680145, "learning_rate": 3.0598330314811076e-07, "loss": 0.0064, "step": 87670 }, { "grad_norm": 0.009212750010192394, "learning_rate": 3.039729640511579e-07, "loss": 0.0055, "step": 87680 }, { "grad_norm": 0.02123129367828369, "learning_rate": 3.01969230661181e-07, "loss": 0.0075, "step": 87690 }, { "grad_norm": 0.03356665000319481, "learning_rate": 2.9997210324452794e-07, "loss": 0.0045, "step": 87700 }, { "grad_norm": 0.0388617217540741, "learning_rate": 2.979815820666532e-07, "loss": 0.0043, "step": 87710 }, { "grad_norm": 0.03283189982175827, "learning_rate": 2.95997667392145e-07, "loss": 0.004, "step": 87720 }, { "grad_norm": 0.053004804998636246, "learning_rate": 2.9402035948470906e-07, "loss": 0.0041, "step": 87730 }, { "grad_norm": 0.03751280903816223, "learning_rate": 2.920496586071797e-07, "loss": 0.0077, "step": 87740 }, { "grad_norm": 0.03219742700457573, "learning_rate": 2.900855650214973e-07, "loss": 0.0053, "step": 87750 }, { "grad_norm": 0.04451189190149307, "learning_rate": 2.881280789887419e-07, "loss": 0.0048, "step": 87760 }, { "grad_norm": 0.03700892627239227, "learning_rate": 2.8617720076909993e-07, "loss": 0.0046, "step": 87770 }, { "grad_norm": 0.03781032934784889, "learning_rate": 2.8423293062188607e-07, "loss": 0.0078, "step": 87780 }, { "grad_norm": 0.03461810201406479, "learning_rate": 2.822952688055436e-07, "loss": 0.0043, "step": 87790 }, { "grad_norm": 0.018440691754221916, "learning_rate": 2.803642155776276e-07, "loss": 0.0056, "step": 87800 }, { "grad_norm": 0.054139960557222366, "learning_rate": 2.7843977119482167e-07, "loss": 0.0075, "step": 87810 }, { "grad_norm": 0.027272269129753113, "learning_rate": 2.765219359129212e-07, "loss": 0.0038, "step": 87820 }, { "grad_norm": 0.06057518348097801, "learning_rate": 2.7461070998685e-07, "loss": 0.0073, "step": 87830 }, { "grad_norm": 0.03383812680840492, "learning_rate": 2.7270609367065493e-07, "loss": 0.0025, "step": 87840 }, { "grad_norm": 0.023376721888780594, "learning_rate": 2.708080872175056e-07, "loss": 0.0056, "step": 87850 }, { "grad_norm": 0.03738837689161301, "learning_rate": 2.689166908796781e-07, "loss": 0.006, "step": 87860 }, { "grad_norm": 0.04466749727725983, "learning_rate": 2.670319049085879e-07, "loss": 0.0052, "step": 87870 }, { "grad_norm": 0.030340442433953285, "learning_rate": 2.651537295547679e-07, "loss": 0.0031, "step": 87880 }, { "grad_norm": 0.024409931153059006, "learning_rate": 2.63282165067863e-07, "loss": 0.0053, "step": 87890 }, { "grad_norm": 0.057081859558820724, "learning_rate": 2.614172116966518e-07, "loss": 0.004, "step": 87900 }, { "grad_norm": 0.03191825747489929, "learning_rate": 2.59558869689025e-07, "loss": 0.005, "step": 87910 }, { "grad_norm": 0.026729246601462364, "learning_rate": 2.577071392919961e-07, "loss": 0.0053, "step": 87920 }, { "grad_norm": 0.0431627482175827, "learning_rate": 2.5586202075170705e-07, "loss": 0.0051, "step": 87930 }, { "grad_norm": 0.022892436012625694, "learning_rate": 2.540235143134062e-07, "loss": 0.0036, "step": 87940 }, { "grad_norm": 0.03687024489045143, "learning_rate": 2.521916202214758e-07, "loss": 0.0075, "step": 87950 }, { "grad_norm": 0.0401715449988842, "learning_rate": 2.5036633871942107e-07, "loss": 0.0041, "step": 87960 }, { "grad_norm": 0.02956019714474678, "learning_rate": 2.4854767004985345e-07, "loss": 0.0043, "step": 87970 }, { "grad_norm": 0.030756250023841858, "learning_rate": 2.467356144545241e-07, "loss": 0.0059, "step": 87980 }, { "grad_norm": 0.04977327212691307, "learning_rate": 2.449301721742847e-07, "loss": 0.0047, "step": 87990 }, { "grad_norm": 0.02235277183353901, "learning_rate": 2.431313434491267e-07, "loss": 0.0033, "step": 88000 }, { "grad_norm": 0.06165095791220665, "learning_rate": 2.4133912851814766e-07, "loss": 0.0038, "step": 88010 }, { "grad_norm": 0.01189684309065342, "learning_rate": 2.395535276195793e-07, "loss": 0.003, "step": 88020 }, { "grad_norm": 0.06607947498559952, "learning_rate": 2.3777454099077057e-07, "loss": 0.0046, "step": 88030 }, { "grad_norm": 0.06111801788210869, "learning_rate": 2.3600216886817682e-07, "loss": 0.0036, "step": 88040 }, { "grad_norm": 0.02130376547574997, "learning_rate": 2.3423641148739294e-07, "loss": 0.0063, "step": 88050 }, { "grad_norm": 0.03594416007399559, "learning_rate": 2.3247726908313117e-07, "loss": 0.0032, "step": 88060 }, { "grad_norm": 0.04965613782405853, "learning_rate": 2.3072474188921e-07, "loss": 0.0069, "step": 88070 }, { "grad_norm": 0.031039828434586525, "learning_rate": 2.2897883013858755e-07, "loss": 0.0041, "step": 88080 }, { "grad_norm": 0.05478687211871147, "learning_rate": 2.272395340633282e-07, "loss": 0.004, "step": 88090 }, { "grad_norm": 0.04262098670005798, "learning_rate": 2.2550685389463033e-07, "loss": 0.0066, "step": 88100 }, { "grad_norm": 0.024471240118145943, "learning_rate": 2.2378078986280416e-07, "loss": 0.0048, "step": 88110 }, { "grad_norm": 0.0288477074354887, "learning_rate": 2.220613421972717e-07, "loss": 0.0058, "step": 88120 }, { "grad_norm": 0.025383638218045235, "learning_rate": 2.2034851112660015e-07, "loss": 0.0046, "step": 88130 }, { "grad_norm": 0.02230171672999859, "learning_rate": 2.186422968784574e-07, "loss": 0.0036, "step": 88140 }, { "grad_norm": 0.024600505828857422, "learning_rate": 2.169426996796342e-07, "loss": 0.0069, "step": 88150 }, { "grad_norm": 0.02233063615858555, "learning_rate": 2.152497197560499e-07, "loss": 0.0034, "step": 88160 }, { "grad_norm": 0.05246276408433914, "learning_rate": 2.1356335733273004e-07, "loss": 0.0078, "step": 88170 }, { "grad_norm": 0.039027463644742966, "learning_rate": 2.1188361263383972e-07, "loss": 0.0041, "step": 88180 }, { "grad_norm": 0.032601095736026764, "learning_rate": 2.102104858826559e-07, "loss": 0.0035, "step": 88190 }, { "grad_norm": 0.047036249190568924, "learning_rate": 2.0854397730156182e-07, "loss": 0.0084, "step": 88200 }, { "grad_norm": 0.04283345863223076, "learning_rate": 2.0688408711208584e-07, "loss": 0.0059, "step": 88210 }, { "grad_norm": 0.02936520241200924, "learning_rate": 2.05230815534857e-07, "loss": 0.0069, "step": 88220 }, { "grad_norm": 0.0490303672850132, "learning_rate": 2.0358416278963287e-07, "loss": 0.0039, "step": 88230 }, { "grad_norm": 0.040233511477708817, "learning_rate": 2.0194412909528836e-07, "loss": 0.0062, "step": 88240 }, { "grad_norm": 0.04094894602894783, "learning_rate": 2.0031071466983243e-07, "loss": 0.0048, "step": 88250 }, { "grad_norm": 0.02265976369380951, "learning_rate": 1.986839197303636e-07, "loss": 0.0043, "step": 88260 }, { "grad_norm": 0.06521878391504288, "learning_rate": 1.970637444931367e-07, "loss": 0.0046, "step": 88270 }, { "grad_norm": 0.025592736899852753, "learning_rate": 1.954501891734961e-07, "loss": 0.0041, "step": 88280 }, { "grad_norm": 0.07998423278331757, "learning_rate": 1.9384325398592585e-07, "loss": 0.0071, "step": 88290 }, { "grad_norm": 0.016997499391436577, "learning_rate": 1.922429391440217e-07, "loss": 0.0058, "step": 88300 }, { "grad_norm": 0.02521650120615959, "learning_rate": 1.906492448605024e-07, "loss": 0.0066, "step": 88310 }, { "grad_norm": 0.043050821870565414, "learning_rate": 1.890621713471985e-07, "loss": 0.0061, "step": 88320 }, { "grad_norm": 0.06023360788822174, "learning_rate": 1.8748171881508013e-07, "loss": 0.0052, "step": 88330 }, { "grad_norm": 0.03989120572805405, "learning_rate": 1.8590788747421816e-07, "loss": 0.0044, "step": 88340 }, { "grad_norm": 0.02679947018623352, "learning_rate": 1.843406775338008e-07, "loss": 0.0032, "step": 88350 }, { "grad_norm": 0.023149890825152397, "learning_rate": 1.8278008920215583e-07, "loss": 0.0045, "step": 88360 }, { "grad_norm": 0.019078711047768593, "learning_rate": 1.8122612268672289e-07, "loss": 0.004, "step": 88370 }, { "grad_norm": 0.03015836887061596, "learning_rate": 1.7967877819404788e-07, "loss": 0.0028, "step": 88380 }, { "grad_norm": 0.03414860740303993, "learning_rate": 1.7813805592981625e-07, "loss": 0.005, "step": 88390 }, { "grad_norm": 0.028985144570469856, "learning_rate": 1.7660395609881419e-07, "loss": 0.0036, "step": 88400 }, { "grad_norm": 0.04845672845840454, "learning_rate": 1.7507647890497303e-07, "loss": 0.0037, "step": 88410 }, { "grad_norm": 0.024705920368433, "learning_rate": 1.735556245513137e-07, "loss": 0.0076, "step": 88420 }, { "grad_norm": 0.03541578724980354, "learning_rate": 1.720413932399967e-07, "loss": 0.0061, "step": 88430 }, { "grad_norm": 0.02322196401655674, "learning_rate": 1.7053378517230545e-07, "loss": 0.0049, "step": 88440 }, { "grad_norm": 0.03696156665682793, "learning_rate": 1.690328005486186e-07, "loss": 0.0031, "step": 88450 }, { "grad_norm": 0.03366697579622269, "learning_rate": 1.6753843956845983e-07, "loss": 0.0069, "step": 88460 }, { "grad_norm": 0.031876176595687866, "learning_rate": 1.6605070243046472e-07, "loss": 0.0059, "step": 88470 }, { "grad_norm": 0.06260684877634048, "learning_rate": 1.6456958933238064e-07, "loss": 0.0079, "step": 88480 }, { "grad_norm": 0.0646451935172081, "learning_rate": 1.63095100471089e-07, "loss": 0.0065, "step": 88490 }, { "grad_norm": 0.07203849405050278, "learning_rate": 1.616272360425719e-07, "loss": 0.0051, "step": 88500 }, { "grad_norm": 0.0307699516415596, "learning_rate": 1.601659962419455e-07, "loss": 0.0046, "step": 88510 }, { "grad_norm": 0.03688235208392143, "learning_rate": 1.5871138126344888e-07, "loss": 0.0062, "step": 88520 }, { "grad_norm": 0.017966104671359062, "learning_rate": 1.5726339130041623e-07, "loss": 0.0068, "step": 88530 }, { "grad_norm": 0.03945039212703705, "learning_rate": 1.5582202654532695e-07, "loss": 0.0052, "step": 88540 }, { "grad_norm": 0.06481501460075378, "learning_rate": 1.5438728718977778e-07, "loss": 0.0063, "step": 88550 }, { "grad_norm": 0.02734001725912094, "learning_rate": 1.529591734244662e-07, "loss": 0.006, "step": 88560 }, { "grad_norm": 0.036224763840436935, "learning_rate": 1.5153768543922365e-07, "loss": 0.0044, "step": 88570 }, { "grad_norm": 0.0431830957531929, "learning_rate": 1.5012282342299898e-07, "loss": 0.0053, "step": 88580 }, { "grad_norm": 0.021441057324409485, "learning_rate": 1.4871458756385292e-07, "loss": 0.0046, "step": 88590 }, { "grad_norm": 0.05277891829609871, "learning_rate": 1.4731297804898015e-07, "loss": 0.004, "step": 88600 }, { "grad_norm": 0.03827638551592827, "learning_rate": 1.459179950646816e-07, "loss": 0.0032, "step": 88610 }, { "grad_norm": 0.07080472260713577, "learning_rate": 1.4452963879638127e-07, "loss": 0.0045, "step": 88620 }, { "grad_norm": 0.05577760562300682, "learning_rate": 1.4314790942862588e-07, "loss": 0.0047, "step": 88630 }, { "grad_norm": 0.03849051892757416, "learning_rate": 1.4177280714506858e-07, "loss": 0.005, "step": 88640 }, { "grad_norm": 0.0598660446703434, "learning_rate": 1.4040433212850202e-07, "loss": 0.0042, "step": 88650 }, { "grad_norm": 0.02883506566286087, "learning_rate": 1.3904248456082514e-07, "loss": 0.0059, "step": 88660 }, { "grad_norm": 0.03300761058926582, "learning_rate": 1.376872646230487e-07, "loss": 0.0042, "step": 88670 }, { "grad_norm": 0.041085198521614075, "learning_rate": 1.3633867249532307e-07, "loss": 0.0043, "step": 88680 }, { "grad_norm": 0.029564402997493744, "learning_rate": 1.3499670835689927e-07, "loss": 0.0037, "step": 88690 }, { "grad_norm": 0.0347583182156086, "learning_rate": 1.3366137238615128e-07, "loss": 0.0051, "step": 88700 }, { "grad_norm": 0.036434754729270935, "learning_rate": 1.3233266476058714e-07, "loss": 0.0065, "step": 88710 }, { "grad_norm": 0.04972991347312927, "learning_rate": 1.3101058565681e-07, "loss": 0.004, "step": 88720 }, { "grad_norm": 0.023258639499545097, "learning_rate": 1.296951352505571e-07, "loss": 0.0046, "step": 88730 }, { "grad_norm": 0.04126861318945885, "learning_rate": 1.2838631371668296e-07, "loss": 0.0034, "step": 88740 }, { "grad_norm": 0.020543543621897697, "learning_rate": 1.270841212291596e-07, "loss": 0.0044, "step": 88750 }, { "grad_norm": 0.02119075506925583, "learning_rate": 1.2578855796107068e-07, "loss": 0.0033, "step": 88760 }, { "grad_norm": 0.03588896617293358, "learning_rate": 1.2449962408463412e-07, "loss": 0.0036, "step": 88770 }, { "grad_norm": 0.06258660554885864, "learning_rate": 1.232173197711739e-07, "loss": 0.0051, "step": 88780 }, { "grad_norm": 0.02164917252957821, "learning_rate": 1.219416451911315e-07, "loss": 0.0049, "step": 88790 }, { "grad_norm": 0.020704003050923347, "learning_rate": 1.2067260051408236e-07, "loss": 0.0051, "step": 88800 }, { "grad_norm": 0.060986362397670746, "learning_rate": 1.1941018590870268e-07, "loss": 0.0049, "step": 88810 }, { "grad_norm": 0.055283933877944946, "learning_rate": 1.1815440154280267e-07, "loss": 0.0038, "step": 88820 }, { "grad_norm": 0.014412830583751202, "learning_rate": 1.1690524758329879e-07, "loss": 0.0065, "step": 88830 }, { "grad_norm": 0.051990360021591187, "learning_rate": 1.1566272419623048e-07, "loss": 0.0052, "step": 88840 }, { "grad_norm": 0.03373177349567413, "learning_rate": 1.1442683154676558e-07, "loss": 0.0046, "step": 88850 }, { "grad_norm": 0.024608558043837547, "learning_rate": 1.1319756979916719e-07, "loss": 0.0045, "step": 88860 }, { "grad_norm": 0.03832833841443062, "learning_rate": 1.1197493911684343e-07, "loss": 0.0047, "step": 88870 }, { "grad_norm": 0.015555170364677906, "learning_rate": 1.1075893966230877e-07, "loss": 0.0045, "step": 88880 }, { "grad_norm": 0.050700873136520386, "learning_rate": 1.0954957159718393e-07, "loss": 0.0051, "step": 88890 }, { "grad_norm": 0.020021405071020126, "learning_rate": 1.0834683508223476e-07, "loss": 0.0055, "step": 88900 }, { "grad_norm": 0.05371219664812088, "learning_rate": 1.071507302773278e-07, "loss": 0.0047, "step": 88910 }, { "grad_norm": 0.014311007224023342, "learning_rate": 1.0596125734144701e-07, "loss": 0.0038, "step": 88920 }, { "grad_norm": 0.021467581391334534, "learning_rate": 1.0477841643271036e-07, "loss": 0.0049, "step": 88930 }, { "grad_norm": 0.03970040753483772, "learning_rate": 1.0360220770833096e-07, "loss": 0.0052, "step": 88940 }, { "grad_norm": 0.015489908866584301, "learning_rate": 1.0243263132466152e-07, "loss": 0.0031, "step": 88950 }, { "grad_norm": 0.051185913383960724, "learning_rate": 1.0126968743716103e-07, "loss": 0.005, "step": 88960 }, { "grad_norm": 0.03811538591980934, "learning_rate": 1.0011337620041695e-07, "loss": 0.0026, "step": 88970 }, { "grad_norm": 0.029385700821876526, "learning_rate": 9.896369776811743e-08, "loss": 0.0041, "step": 88980 }, { "grad_norm": 0.05002405494451523, "learning_rate": 9.782065229309023e-08, "loss": 0.0069, "step": 88990 }, { "grad_norm": 0.03500908240675926, "learning_rate": 9.668423992726383e-08, "loss": 0.0064, "step": 89000 }, { "grad_norm": 0.03941868990659714, "learning_rate": 9.555446082170072e-08, "loss": 0.0044, "step": 89010 }, { "grad_norm": 0.0202497448772192, "learning_rate": 9.443131512656967e-08, "loss": 0.0038, "step": 89020 }, { "grad_norm": 0.020079685375094414, "learning_rate": 9.331480299116235e-08, "loss": 0.0047, "step": 89030 }, { "grad_norm": 0.026071326807141304, "learning_rate": 9.22049245638823e-08, "loss": 0.0033, "step": 89040 }, { "grad_norm": 0.047112468630075455, "learning_rate": 9.110167999226705e-08, "loss": 0.0077, "step": 89050 }, { "grad_norm": 0.028888393193483353, "learning_rate": 9.000506942295484e-08, "loss": 0.0054, "step": 89060 }, { "grad_norm": 0.02677677385509014, "learning_rate": 8.891509300171242e-08, "loss": 0.0035, "step": 89070 }, { "grad_norm": 0.06353937834501266, "learning_rate": 8.783175087342388e-08, "loss": 0.0062, "step": 89080 }, { "grad_norm": 0.03486209362745285, "learning_rate": 8.675504318208516e-08, "loss": 0.0045, "step": 89090 }, { "grad_norm": 0.09556280076503754, "learning_rate": 8.568497007082066e-08, "loss": 0.0052, "step": 89100 }, { "grad_norm": 0.058539148420095444, "learning_rate": 8.462153168186104e-08, "loss": 0.0041, "step": 89110 }, { "grad_norm": 0.015824129804968834, "learning_rate": 8.356472815655992e-08, "loss": 0.0047, "step": 89120 }, { "grad_norm": 0.03647729754447937, "learning_rate": 8.251455963539378e-08, "loss": 0.0056, "step": 89130 }, { "grad_norm": 0.037271372973918915, "learning_rate": 8.147102625795656e-08, "loss": 0.009, "step": 89140 }, { "grad_norm": 0.03928940370678902, "learning_rate": 8.043412816294838e-08, "loss": 0.0055, "step": 89150 }, { "grad_norm": 0.04870634526014328, "learning_rate": 7.940386548820344e-08, "loss": 0.0044, "step": 89160 }, { "grad_norm": 0.03032740391790867, "learning_rate": 7.838023837066222e-08, "loss": 0.0041, "step": 89170 }, { "grad_norm": 0.017717931419610977, "learning_rate": 7.736324694639363e-08, "loss": 0.0039, "step": 89180 }, { "grad_norm": 0.038031384348869324, "learning_rate": 7.635289135056733e-08, "loss": 0.0033, "step": 89190 }, { "grad_norm": 0.032648030668497086, "learning_rate": 7.534917171749257e-08, "loss": 0.0037, "step": 89200 }, { "grad_norm": 0.025838209316134453, "learning_rate": 7.435208818058481e-08, "loss": 0.0062, "step": 89210 }, { "grad_norm": 0.021259643137454987, "learning_rate": 7.33616408723714e-08, "loss": 0.003, "step": 89220 }, { "grad_norm": 0.013028550893068314, "learning_rate": 7.237782992450814e-08, "loss": 0.0038, "step": 89230 }, { "grad_norm": 0.025707582011818886, "learning_rate": 7.14006554677682e-08, "loss": 0.0041, "step": 89240 }, { "grad_norm": 0.01905757561326027, "learning_rate": 7.04301176320421e-08, "loss": 0.0052, "step": 89250 }, { "grad_norm": 0.04449828714132309, "learning_rate": 6.946621654632668e-08, "loss": 0.0051, "step": 89260 }, { "grad_norm": 0.01910753734409809, "learning_rate": 6.850895233875277e-08, "loss": 0.0027, "step": 89270 }, { "grad_norm": 0.03386765718460083, "learning_rate": 6.755832513656302e-08, "loss": 0.0037, "step": 89280 }, { "grad_norm": 0.03759947419166565, "learning_rate": 6.661433506611192e-08, "loss": 0.0051, "step": 89290 }, { "grad_norm": 0.022931676357984543, "learning_rate": 6.567698225287688e-08, "loss": 0.0035, "step": 89300 }, { "grad_norm": 0.04280846565961838, "learning_rate": 6.474626682145823e-08, "loss": 0.0039, "step": 89310 }, { "grad_norm": 0.011665872298181057, "learning_rate": 6.382218889556257e-08, "loss": 0.0046, "step": 89320 }, { "grad_norm": 0.06015216186642647, "learning_rate": 6.290474859803053e-08, "loss": 0.005, "step": 89330 }, { "grad_norm": 0.049314383417367935, "learning_rate": 6.199394605079788e-08, "loss": 0.004, "step": 89340 }, { "grad_norm": 0.0740196704864502, "learning_rate": 6.108978137493449e-08, "loss": 0.0099, "step": 89350 }, { "grad_norm": 0.06824371218681335, "learning_rate": 6.01922546906275e-08, "loss": 0.005, "step": 89360 }, { "grad_norm": 0.037277914583683014, "learning_rate": 5.930136611717596e-08, "loss": 0.0079, "step": 89370 }, { "grad_norm": 0.04428383708000183, "learning_rate": 5.84171157730018e-08, "loss": 0.0043, "step": 89380 }, { "grad_norm": 0.04547812044620514, "learning_rate": 5.753950377563322e-08, "loss": 0.0043, "step": 89390 }, { "grad_norm": 0.024254214018583298, "learning_rate": 5.666853024173246e-08, "loss": 0.0058, "step": 89400 }, { "grad_norm": 0.03420224040746689, "learning_rate": 5.580419528706804e-08, "loss": 0.0031, "step": 89410 }, { "grad_norm": 0.046446822583675385, "learning_rate": 5.494649902653137e-08, "loss": 0.0048, "step": 89420 }, { "grad_norm": 0.07403919845819473, "learning_rate": 5.409544157413127e-08, "loss": 0.0073, "step": 89430 }, { "grad_norm": 0.03226625919342041, "learning_rate": 5.3251023042982795e-08, "loss": 0.0071, "step": 89440 }, { "grad_norm": 0.035832975059747696, "learning_rate": 5.2413243545335054e-08, "loss": 0.0055, "step": 89450 }, { "grad_norm": 0.056840088218450546, "learning_rate": 5.158210319254897e-08, "loss": 0.0036, "step": 89460 }, { "grad_norm": 0.042525775730609894, "learning_rate": 5.075760209509728e-08, "loss": 0.0042, "step": 89470 }, { "grad_norm": 0.06044713035225868, "learning_rate": 4.993974036257565e-08, "loss": 0.0056, "step": 89480 }, { "grad_norm": 0.049474358558654785, "learning_rate": 4.9128518103697116e-08, "loss": 0.0038, "step": 89490 }, { "grad_norm": 0.05493398755788803, "learning_rate": 4.832393542628655e-08, "loss": 0.0044, "step": 89500 }, { "grad_norm": 0.0255409087985754, "learning_rate": 4.7525992437302824e-08, "loss": 0.0039, "step": 89510 }, { "grad_norm": 0.03225776553153992, "learning_rate": 4.67346892428e-08, "loss": 0.0033, "step": 89520 }, { "grad_norm": 0.12093020975589752, "learning_rate": 4.5950025947960605e-08, "loss": 0.0057, "step": 89530 }, { "grad_norm": 0.035454295575618744, "learning_rate": 4.517200265708454e-08, "loss": 0.003, "step": 89540 }, { "grad_norm": 0.03251883015036583, "learning_rate": 4.440061947359464e-08, "loss": 0.0058, "step": 89550 }, { "grad_norm": 0.04635651037096977, "learning_rate": 4.3635876500014436e-08, "loss": 0.0031, "step": 89560 }, { "grad_norm": 0.045625414699316025, "learning_rate": 4.287777383800151e-08, "loss": 0.0042, "step": 89570 }, { "grad_norm": 0.06132134422659874, "learning_rate": 4.212631158832525e-08, "loss": 0.0038, "step": 89580 }, { "grad_norm": 0.023356836289167404, "learning_rate": 4.138148985087242e-08, "loss": 0.0032, "step": 89590 }, { "grad_norm": 0.016875257715582848, "learning_rate": 4.064330872464162e-08, "loss": 0.0054, "step": 89600 }, { "grad_norm": 0.032485298812389374, "learning_rate": 3.991176830775434e-08, "loss": 0.0054, "step": 89610 }, { "grad_norm": 0.06460948288440704, "learning_rate": 3.9186868697455025e-08, "loss": 0.0057, "step": 89620 }, { "grad_norm": 0.054429683834314346, "learning_rate": 3.8468609990088834e-08, "loss": 0.0042, "step": 89630 }, { "grad_norm": 0.016712790355086327, "learning_rate": 3.775699228113494e-08, "loss": 0.0046, "step": 89640 }, { "grad_norm": 0.08838094770908356, "learning_rate": 3.7052015665184345e-08, "loss": 0.0055, "step": 89650 }, { "grad_norm": 0.04083576053380966, "learning_rate": 3.635368023593988e-08, "loss": 0.0042, "step": 89660 }, { "grad_norm": 0.043382760137319565, "learning_rate": 3.5661986086227286e-08, "loss": 0.0057, "step": 89670 }, { "grad_norm": 0.0156793761998415, "learning_rate": 3.497693330798968e-08, "loss": 0.0034, "step": 89680 }, { "grad_norm": 0.04754823073744774, "learning_rate": 3.429852199228201e-08, "loss": 0.0057, "step": 89690 }, { "grad_norm": 0.028747644275426865, "learning_rate": 3.3626752229287684e-08, "loss": 0.0052, "step": 89700 }, { "grad_norm": 0.051978591829538345, "learning_rate": 3.2961624108290846e-08, "loss": 0.0043, "step": 89710 }, { "grad_norm": 0.0227502454072237, "learning_rate": 3.23031377177041e-08, "loss": 0.0045, "step": 89720 }, { "grad_norm": 0.019127167761325836, "learning_rate": 3.165129314505744e-08, "loss": 0.0035, "step": 89730 }, { "grad_norm": 0.031387798488140106, "learning_rate": 3.100609047699821e-08, "loss": 0.0041, "step": 89740 }, { "grad_norm": 0.04377195984125137, "learning_rate": 3.036752979928559e-08, "loss": 0.0045, "step": 89750 }, { "grad_norm": 0.024558430537581444, "learning_rate": 2.973561119679058e-08, "loss": 0.0041, "step": 89760 }, { "grad_norm": 0.025791721418499947, "learning_rate": 2.911033475352376e-08, "loss": 0.0064, "step": 89770 }, { "grad_norm": 0.028442079201340675, "learning_rate": 2.849170055258532e-08, "loss": 0.0042, "step": 89780 }, { "grad_norm": 0.06426016986370087, "learning_rate": 2.7879708676215034e-08, "loss": 0.0066, "step": 89790 }, { "grad_norm": 0.08143831789493561, "learning_rate": 2.7274359205753387e-08, "loss": 0.0058, "step": 89800 }, { "grad_norm": 0.031230349093675613, "learning_rate": 2.6675652221669345e-08, "loss": 0.004, "step": 89810 }, { "grad_norm": 0.08012601733207703, "learning_rate": 2.6083587803543696e-08, "loss": 0.0067, "step": 89820 }, { "grad_norm": 0.05587950348854065, "learning_rate": 2.549816603006905e-08, "loss": 0.0042, "step": 89830 }, { "grad_norm": 0.0191962830722332, "learning_rate": 2.4919386979072033e-08, "loss": 0.0089, "step": 89840 }, { "grad_norm": 0.03317103162407875, "learning_rate": 2.434725072747446e-08, "loss": 0.005, "step": 89850 }, { "grad_norm": 0.035309262573719025, "learning_rate": 2.3781757351332147e-08, "loss": 0.0046, "step": 89860 }, { "grad_norm": 0.045077111572027206, "learning_rate": 2.32229069258072e-08, "loss": 0.0049, "step": 89870 }, { "grad_norm": 0.04436549171805382, "learning_rate": 2.2670699525190188e-08, "loss": 0.0061, "step": 89880 }, { "grad_norm": 0.011819809675216675, "learning_rate": 2.2125135222877958e-08, "loss": 0.0041, "step": 89890 }, { "grad_norm": 0.022159889340400696, "learning_rate": 2.1586214091384728e-08, "loss": 0.0048, "step": 89900 }, { "grad_norm": 0.02917082980275154, "learning_rate": 2.1053936202347635e-08, "loss": 0.0038, "step": 89910 }, { "grad_norm": 0.04674477130174637, "learning_rate": 2.0528301626521194e-08, "loss": 0.0038, "step": 89920 }, { "grad_norm": 0.01749066822230816, "learning_rate": 2.0009310433771743e-08, "loss": 0.0036, "step": 89930 }, { "grad_norm": 0.019605306908488274, "learning_rate": 1.9496962693088538e-08, "loss": 0.0036, "step": 89940 }, { "grad_norm": 0.04134095460176468, "learning_rate": 1.8991258472561558e-08, "loss": 0.0045, "step": 89950 }, { "grad_norm": 0.02316887676715851, "learning_rate": 1.8492197839425907e-08, "loss": 0.0035, "step": 89960 }, { "grad_norm": 0.04573064669966698, "learning_rate": 1.799978086000631e-08, "loss": 0.0075, "step": 89970 }, { "grad_norm": 0.04844974726438522, "learning_rate": 1.7514007599761517e-08, "loss": 0.0037, "step": 89980 }, { "grad_norm": 0.07764352858066559, "learning_rate": 1.7034878123262098e-08, "loss": 0.0042, "step": 89990 }, { "grad_norm": 0.03796860948204994, "learning_rate": 1.656239249419045e-08, "loss": 0.0073, "step": 90000 }, { "grad_norm": 0.049469757825136185, "learning_rate": 1.609655077535188e-08, "loss": 0.0053, "step": 90010 }, { "grad_norm": 0.03517837077379227, "learning_rate": 1.563735302867464e-08, "loss": 0.0059, "step": 90020 }, { "grad_norm": 0.07905103266239166, "learning_rate": 1.518479931518213e-08, "loss": 0.008, "step": 90030 }, { "grad_norm": 0.0562443844974041, "learning_rate": 1.473888969504289e-08, "loss": 0.0043, "step": 90040 }, { "grad_norm": 0.0611247681081295, "learning_rate": 1.4299624227520625e-08, "loss": 0.0045, "step": 90050 }, { "grad_norm": 0.037716783583164215, "learning_rate": 1.3867002971001963e-08, "loss": 0.0047, "step": 90060 }, { "grad_norm": 0.04921155422925949, "learning_rate": 1.3441025982996458e-08, "loss": 0.0052, "step": 90070 }, { "grad_norm": 0.04916059225797653, "learning_rate": 1.3021693320125484e-08, "loss": 0.0085, "step": 90080 }, { "grad_norm": 0.03612435609102249, "learning_rate": 1.2609005038127786e-08, "loss": 0.0054, "step": 90090 }, { "grad_norm": 0.04946867749094963, "learning_rate": 1.2202961191853935e-08, "loss": 0.004, "step": 90100 }, { "grad_norm": 0.02931441180408001, "learning_rate": 1.1803561835282972e-08, "loss": 0.0044, "step": 90110 }, { "grad_norm": 0.04017852246761322, "learning_rate": 1.1410807021500214e-08, "loss": 0.0055, "step": 90120 }, { "grad_norm": 0.06645800918340683, "learning_rate": 1.1024696802708345e-08, "loss": 0.0056, "step": 90130 }, { "grad_norm": 0.026177236810326576, "learning_rate": 1.0645231230238528e-08, "loss": 0.0065, "step": 90140 }, { "grad_norm": 0.06903083622455597, "learning_rate": 1.0272410354522643e-08, "loss": 0.0048, "step": 90150 }, { "grad_norm": 0.03293542563915253, "learning_rate": 9.906234225121046e-09, "loss": 0.0044, "step": 90160 }, { "grad_norm": 0.07550399005413055, "learning_rate": 9.546702890705917e-09, "loss": 0.0039, "step": 90170 }, { "grad_norm": 0.019022781401872635, "learning_rate": 9.193816399066802e-09, "loss": 0.0049, "step": 90180 }, { "grad_norm": 0.02250160463154316, "learning_rate": 8.847574797105074e-09, "loss": 0.0056, "step": 90190 }, { "grad_norm": 0.033133115619421005, "learning_rate": 8.507978130856132e-09, "loss": 0.0055, "step": 90200 }, { "grad_norm": 0.03652287274599075, "learning_rate": 8.175026445444989e-09, "loss": 0.0048, "step": 90210 }, { "grad_norm": 0.025036009028553963, "learning_rate": 7.848719785136237e-09, "loss": 0.0048, "step": 90220 }, { "grad_norm": 0.03109295852482319, "learning_rate": 7.529058193306293e-09, "loss": 0.0045, "step": 90230 }, { "grad_norm": 0.08443628251552582, "learning_rate": 7.216041712443389e-09, "loss": 0.0064, "step": 90240 }, { "grad_norm": 0.023881906643509865, "learning_rate": 6.909670384147582e-09, "loss": 0.0061, "step": 90250 }, { "grad_norm": 0.06114954128861427, "learning_rate": 6.609944249147404e-09, "loss": 0.0035, "step": 90260 }, { "grad_norm": 0.05152490362524986, "learning_rate": 6.31686334728876e-09, "loss": 0.0072, "step": 90270 }, { "grad_norm": 0.025229455903172493, "learning_rate": 6.0304277175182725e-09, "loss": 0.0032, "step": 90280 }, { "grad_norm": 0.017831869423389435, "learning_rate": 5.750637397916592e-09, "loss": 0.0043, "step": 90290 }, { "grad_norm": 0.03081241063773632, "learning_rate": 5.4774924256706386e-09, "loss": 0.0045, "step": 90300 }, { "grad_norm": 0.022270694375038147, "learning_rate": 5.210992837090256e-09, "loss": 0.0043, "step": 90310 }, { "grad_norm": 0.04215458407998085, "learning_rate": 4.951138667597111e-09, "loss": 0.0041, "step": 90320 }, { "grad_norm": 0.021381590515375137, "learning_rate": 4.69792995173024e-09, "loss": 0.0031, "step": 90330 }, { "grad_norm": 0.041996438056230545, "learning_rate": 4.4513667231516065e-09, "loss": 0.0042, "step": 90340 }, { "grad_norm": 0.03198778256773949, "learning_rate": 4.211449014629442e-09, "loss": 0.0045, "step": 90350 }, { "grad_norm": 0.033288806676864624, "learning_rate": 3.978176858054905e-09, "loss": 0.0034, "step": 90360 }, { "grad_norm": 0.026343831792473793, "learning_rate": 3.7515502844420736e-09, "loss": 0.0046, "step": 90370 }, { "grad_norm": 0.04221360385417938, "learning_rate": 3.531569323905748e-09, "loss": 0.0045, "step": 90380 }, { "grad_norm": 0.041550152003765106, "learning_rate": 3.3182340056947535e-09, "loss": 0.0054, "step": 90390 }, { "grad_norm": 0.07771416008472443, "learning_rate": 3.1115443581530845e-09, "loss": 0.0061, "step": 90400 }, { "grad_norm": 0.06284703314304352, "learning_rate": 2.9115004087698626e-09, "loss": 0.0072, "step": 90410 }, { "grad_norm": 0.06363146007061005, "learning_rate": 2.7181021841238276e-09, "loss": 0.0036, "step": 90420 }, { "grad_norm": 0.05404174327850342, "learning_rate": 2.531349709927744e-09, "loss": 0.0072, "step": 90430 }, { "grad_norm": 0.02176554501056671, "learning_rate": 2.351243011006199e-09, "loss": 0.0039, "step": 90440 }, { "grad_norm": 0.05290614441037178, "learning_rate": 2.1777821112900498e-09, "loss": 0.0036, "step": 90450 }, { "grad_norm": 0.02525869756937027, "learning_rate": 2.010967033849731e-09, "loss": 0.0034, "step": 90460 }, { "grad_norm": 0.034487035125494, "learning_rate": 1.8507978008508454e-09, "loss": 0.0067, "step": 90470 }, { "grad_norm": 0.02608615905046463, "learning_rate": 1.697274433581919e-09, "loss": 0.0041, "step": 90480 }, { "grad_norm": 0.061240240931510925, "learning_rate": 1.5503969524488515e-09, "loss": 0.0067, "step": 90490 }, { "grad_norm": 0.038672760128974915, "learning_rate": 1.4101653769860169e-09, "loss": 0.0051, "step": 90500 }, { "grad_norm": 0.030051106587052345, "learning_rate": 1.276579725817406e-09, "loss": 0.0053, "step": 90510 }, { "grad_norm": 0.01668558456003666, "learning_rate": 1.1496400167121391e-09, "loss": 0.005, "step": 90520 }, { "grad_norm": 0.01699800416827202, "learning_rate": 1.0293462665400544e-09, "loss": 0.0056, "step": 90530 }, { "grad_norm": 0.06869451701641083, "learning_rate": 9.156984912883637e-10, "loss": 0.007, "step": 90540 }, { "grad_norm": 0.04333395138382912, "learning_rate": 8.086967060616513e-10, "loss": 0.004, "step": 90550 }, { "grad_norm": 0.017635513097047806, "learning_rate": 7.083409250874251e-10, "loss": 0.0041, "step": 90560 }, { "grad_norm": 0.050212327390909195, "learning_rate": 6.146311617050148e-10, "loss": 0.0056, "step": 90570 }, { "grad_norm": 0.052339114248752594, "learning_rate": 5.27567428371123e-10, "loss": 0.0049, "step": 90580 }, { "grad_norm": 0.043125513941049576, "learning_rate": 4.471497366542732e-10, "loss": 0.008, "step": 90590 }, { "grad_norm": 0.02282026968896389, "learning_rate": 3.7337809724036223e-10, "loss": 0.0049, "step": 90600 }, { "grad_norm": 0.04917481169104576, "learning_rate": 3.062525199493127e-10, "loss": 0.0049, "step": 90610 }, { "grad_norm": 0.035581864416599274, "learning_rate": 2.457730136906644e-10, "loss": 0.0079, "step": 90620 }, { "grad_norm": 0.024479491636157036, "learning_rate": 1.9193958650798316e-10, "loss": 0.0068, "step": 90630 }, { "grad_norm": 0.019637800753116608, "learning_rate": 1.4475224555665635e-10, "loss": 0.0028, "step": 90640 }, { "grad_norm": 0.0233025923371315, "learning_rate": 1.0421099710944404e-10, "loss": 0.003, "step": 90650 }, { "grad_norm": 0.016905788332223892, "learning_rate": 7.031584655647905e-11, "loss": 0.0046, "step": 90660 }, { "grad_norm": 0.07892818748950958, "learning_rate": 4.306679839971573e-11, "loss": 0.0055, "step": 90670 }, { "grad_norm": 0.05648059770464897, "learning_rate": 2.2463856264032247e-11, "loss": 0.0038, "step": 90680 }, { "grad_norm": 0.026061004027724266, "learning_rate": 8.507022886128369e-12, "loss": 0.0073, "step": 90690 }, { "grad_norm": 0.029781728982925415, "learning_rate": 1.1963001200765434e-12, "loss": 0.0045, "step": 90700 } ], "logging_steps": 10, "max_steps": 90705, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 9070, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }