{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0034846999891105, "eval_steps": 500, "global_step": 9200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0021779374931939453, "grad_norm": 8.702981948852539, "learning_rate": 1.956521739130435e-06, "loss": 0.8381, "step": 10 }, { "epoch": 0.004355874986387891, "grad_norm": 2.2589502334594727, "learning_rate": 4.130434782608695e-06, "loss": 0.5872, "step": 20 }, { "epoch": 0.006533812479581836, "grad_norm": 2.637632131576538, "learning_rate": 6.304347826086957e-06, "loss": 0.3281, "step": 30 }, { "epoch": 0.008711749972775781, "grad_norm": 1.8374388217926025, "learning_rate": 8.478260869565217e-06, "loss": 0.2372, "step": 40 }, { "epoch": 0.010889687465969727, "grad_norm": 1.901687741279602, "learning_rate": 1.0652173913043479e-05, "loss": 0.1911, "step": 50 }, { "epoch": 0.013067624959163672, "grad_norm": 1.3057066202163696, "learning_rate": 1.2826086956521741e-05, "loss": 0.1595, "step": 60 }, { "epoch": 0.015245562452357617, "grad_norm": 1.224269986152649, "learning_rate": 1.5e-05, "loss": 0.1345, "step": 70 }, { "epoch": 0.017423499945551563, "grad_norm": 0.7691900730133057, "learning_rate": 1.7173913043478263e-05, "loss": 0.1211, "step": 80 }, { "epoch": 0.019601437438745508, "grad_norm": 0.4880311191082001, "learning_rate": 1.9347826086956523e-05, "loss": 0.1057, "step": 90 }, { "epoch": 0.021779374931939453, "grad_norm": 0.7204518914222717, "learning_rate": 2.1521739130434784e-05, "loss": 0.0957, "step": 100 }, { "epoch": 0.0239573124251334, "grad_norm": 0.7122465968132019, "learning_rate": 2.3695652173913045e-05, "loss": 0.0912, "step": 110 }, { "epoch": 0.026135249918327344, "grad_norm": 0.5246570706367493, "learning_rate": 2.5869565217391305e-05, "loss": 0.0851, "step": 120 }, { "epoch": 0.02831318741152129, "grad_norm": 0.3882908821105957, "learning_rate": 2.8043478260869566e-05, "loss": 0.0777, "step": 130 }, { "epoch": 0.030491124904715235, "grad_norm": 0.26690909266471863, "learning_rate": 3.0217391304347827e-05, "loss": 0.0727, "step": 140 }, { "epoch": 0.03266906239790918, "grad_norm": 0.48248469829559326, "learning_rate": 3.239130434782609e-05, "loss": 0.0669, "step": 150 }, { "epoch": 0.034846999891103125, "grad_norm": 0.5324282050132751, "learning_rate": 3.456521739130435e-05, "loss": 0.0631, "step": 160 }, { "epoch": 0.03702493738429707, "grad_norm": 0.5717258453369141, "learning_rate": 3.673913043478261e-05, "loss": 0.062, "step": 170 }, { "epoch": 0.039202874877491016, "grad_norm": 0.4997138977050781, "learning_rate": 3.8913043478260866e-05, "loss": 0.0585, "step": 180 }, { "epoch": 0.04138081237068496, "grad_norm": 0.8486296534538269, "learning_rate": 4.1086956521739134e-05, "loss": 0.0501, "step": 190 }, { "epoch": 0.04355874986387891, "grad_norm": 0.5152239203453064, "learning_rate": 4.3260869565217394e-05, "loss": 0.0491, "step": 200 }, { "epoch": 0.04573668735707285, "grad_norm": 0.4378140866756439, "learning_rate": 4.5434782608695655e-05, "loss": 0.0484, "step": 210 }, { "epoch": 0.0479146248502668, "grad_norm": 0.392478883266449, "learning_rate": 4.7608695652173916e-05, "loss": 0.0465, "step": 220 }, { "epoch": 0.05009256234346074, "grad_norm": 0.4341851770877838, "learning_rate": 4.9782608695652176e-05, "loss": 0.043, "step": 230 }, { "epoch": 0.05227049983665469, "grad_norm": 0.4033254086971283, "learning_rate": 5.195652173913044e-05, "loss": 0.0411, "step": 240 }, { "epoch": 0.05444843732984863, "grad_norm": 0.470051109790802, "learning_rate": 5.41304347826087e-05, "loss": 0.0412, "step": 250 }, { "epoch": 0.05662637482304258, "grad_norm": 0.48915666341781616, "learning_rate": 5.630434782608696e-05, "loss": 0.0374, "step": 260 }, { "epoch": 0.05880431231623652, "grad_norm": 0.5334729552268982, "learning_rate": 5.847826086956521e-05, "loss": 0.0396, "step": 270 }, { "epoch": 0.06098224980943047, "grad_norm": 0.3755224943161011, "learning_rate": 6.0652173913043487e-05, "loss": 0.034, "step": 280 }, { "epoch": 0.06316018730262442, "grad_norm": 0.5257115960121155, "learning_rate": 6.282608695652175e-05, "loss": 0.0366, "step": 290 }, { "epoch": 0.06533812479581835, "grad_norm": 0.6483582854270935, "learning_rate": 6.500000000000001e-05, "loss": 0.0337, "step": 300 }, { "epoch": 0.0675160622890123, "grad_norm": 0.6292915344238281, "learning_rate": 6.717391304347827e-05, "loss": 0.0331, "step": 310 }, { "epoch": 0.06969399978220625, "grad_norm": 0.6189426779747009, "learning_rate": 6.934782608695653e-05, "loss": 0.0339, "step": 320 }, { "epoch": 0.0718719372754002, "grad_norm": 0.5627177357673645, "learning_rate": 7.152173913043479e-05, "loss": 0.0317, "step": 330 }, { "epoch": 0.07404987476859413, "grad_norm": 0.6323244571685791, "learning_rate": 7.369565217391304e-05, "loss": 0.0306, "step": 340 }, { "epoch": 0.07622781226178808, "grad_norm": 0.39575356245040894, "learning_rate": 7.58695652173913e-05, "loss": 0.0296, "step": 350 }, { "epoch": 0.07840574975498203, "grad_norm": 0.4582425653934479, "learning_rate": 7.804347826086957e-05, "loss": 0.0313, "step": 360 }, { "epoch": 0.08058368724817598, "grad_norm": 0.472897469997406, "learning_rate": 8.021739130434783e-05, "loss": 0.0321, "step": 370 }, { "epoch": 0.08276162474136992, "grad_norm": 0.4483104944229126, "learning_rate": 8.23913043478261e-05, "loss": 0.0298, "step": 380 }, { "epoch": 0.08493956223456386, "grad_norm": 0.5729120373725891, "learning_rate": 8.456521739130435e-05, "loss": 0.0305, "step": 390 }, { "epoch": 0.08711749972775781, "grad_norm": 0.4654761254787445, "learning_rate": 8.673913043478261e-05, "loss": 0.0285, "step": 400 }, { "epoch": 0.08929543722095176, "grad_norm": 0.394075483083725, "learning_rate": 8.891304347826088e-05, "loss": 0.0264, "step": 410 }, { "epoch": 0.0914733747141457, "grad_norm": 0.6974279880523682, "learning_rate": 9.108695652173914e-05, "loss": 0.0272, "step": 420 }, { "epoch": 0.09365131220733965, "grad_norm": 0.43244045972824097, "learning_rate": 9.32608695652174e-05, "loss": 0.0275, "step": 430 }, { "epoch": 0.0958292497005336, "grad_norm": 0.35818201303482056, "learning_rate": 9.543478260869566e-05, "loss": 0.0272, "step": 440 }, { "epoch": 0.09800718719372754, "grad_norm": 0.38727718591690063, "learning_rate": 9.760869565217392e-05, "loss": 0.0265, "step": 450 }, { "epoch": 0.10018512468692148, "grad_norm": 0.4803049862384796, "learning_rate": 9.978260869565218e-05, "loss": 0.0248, "step": 460 }, { "epoch": 0.10236306218011543, "grad_norm": 0.6433846354484558, "learning_rate": 9.999973836157333e-05, "loss": 0.0251, "step": 470 }, { "epoch": 0.10454099967330938, "grad_norm": 0.5021262168884277, "learning_rate": 9.999883393595947e-05, "loss": 0.0263, "step": 480 }, { "epoch": 0.10671893716650332, "grad_norm": 0.34864214062690735, "learning_rate": 9.999728350473721e-05, "loss": 0.0247, "step": 490 }, { "epoch": 0.10889687465969726, "grad_norm": 0.2652354836463928, "learning_rate": 9.99950870879387e-05, "loss": 0.0248, "step": 500 }, { "epoch": 0.11107481215289121, "grad_norm": 0.47971311211586, "learning_rate": 9.99922447139426e-05, "loss": 0.0248, "step": 510 }, { "epoch": 0.11325274964608516, "grad_norm": 0.3711201846599579, "learning_rate": 9.998875641947354e-05, "loss": 0.0224, "step": 520 }, { "epoch": 0.1154306871392791, "grad_norm": 0.31161391735076904, "learning_rate": 9.998462224960175e-05, "loss": 0.0218, "step": 530 }, { "epoch": 0.11760862463247304, "grad_norm": 0.3941214084625244, "learning_rate": 9.997984225774238e-05, "loss": 0.0245, "step": 540 }, { "epoch": 0.11978656212566699, "grad_norm": 0.2373579889535904, "learning_rate": 9.99744165056549e-05, "loss": 0.0221, "step": 550 }, { "epoch": 0.12196449961886094, "grad_norm": 0.22790683805942535, "learning_rate": 9.99683450634423e-05, "loss": 0.02, "step": 560 }, { "epoch": 0.12414243711205489, "grad_norm": 0.28962817788124084, "learning_rate": 9.996162800955011e-05, "loss": 0.0223, "step": 570 }, { "epoch": 0.12632037460524884, "grad_norm": 0.23208728432655334, "learning_rate": 9.995426543076545e-05, "loss": 0.0205, "step": 580 }, { "epoch": 0.12849831209844279, "grad_norm": 0.3005686402320862, "learning_rate": 9.994625742221586e-05, "loss": 0.0223, "step": 590 }, { "epoch": 0.1306762495916367, "grad_norm": 0.33873653411865234, "learning_rate": 9.993760408736814e-05, "loss": 0.0214, "step": 600 }, { "epoch": 0.13285418708483065, "grad_norm": 0.3097042441368103, "learning_rate": 9.992830553802696e-05, "loss": 0.0196, "step": 610 }, { "epoch": 0.1350321245780246, "grad_norm": 0.2924380302429199, "learning_rate": 9.991836189433342e-05, "loss": 0.0208, "step": 620 }, { "epoch": 0.13721006207121855, "grad_norm": 0.4635681211948395, "learning_rate": 9.990777328476348e-05, "loss": 0.0199, "step": 630 }, { "epoch": 0.1393879995644125, "grad_norm": 0.43763524293899536, "learning_rate": 9.98965398461264e-05, "loss": 0.0211, "step": 640 }, { "epoch": 0.14156593705760645, "grad_norm": 0.43816155195236206, "learning_rate": 9.988466172356282e-05, "loss": 0.0208, "step": 650 }, { "epoch": 0.1437438745508004, "grad_norm": 0.4274309277534485, "learning_rate": 9.9872139070543e-05, "loss": 0.02, "step": 660 }, { "epoch": 0.14592181204399435, "grad_norm": 0.3821732699871063, "learning_rate": 9.985897204886481e-05, "loss": 0.024, "step": 670 }, { "epoch": 0.14809974953718827, "grad_norm": 0.3163106441497803, "learning_rate": 9.984516082865159e-05, "loss": 0.0204, "step": 680 }, { "epoch": 0.15027768703038222, "grad_norm": 0.362935870885849, "learning_rate": 9.983070558835002e-05, "loss": 0.0191, "step": 690 }, { "epoch": 0.15245562452357617, "grad_norm": 0.5344079732894897, "learning_rate": 9.981560651472781e-05, "loss": 0.0188, "step": 700 }, { "epoch": 0.15463356201677012, "grad_norm": 0.5458508133888245, "learning_rate": 9.97998638028712e-05, "loss": 0.0206, "step": 710 }, { "epoch": 0.15681149950996406, "grad_norm": 0.4091506004333496, "learning_rate": 9.978347765618257e-05, "loss": 0.0192, "step": 720 }, { "epoch": 0.158989437003158, "grad_norm": 0.2918073534965515, "learning_rate": 9.976644828637767e-05, "loss": 0.0206, "step": 730 }, { "epoch": 0.16116737449635196, "grad_norm": 0.28235873579978943, "learning_rate": 9.974877591348304e-05, "loss": 0.0183, "step": 740 }, { "epoch": 0.1633453119895459, "grad_norm": 0.2747727334499359, "learning_rate": 9.973046076583301e-05, "loss": 0.0196, "step": 750 }, { "epoch": 0.16552324948273983, "grad_norm": 0.36752331256866455, "learning_rate": 9.97115030800669e-05, "loss": 0.0179, "step": 760 }, { "epoch": 0.16770118697593378, "grad_norm": 0.44407200813293457, "learning_rate": 9.969190310112579e-05, "loss": 0.0212, "step": 770 }, { "epoch": 0.16987912446912773, "grad_norm": 0.2927279770374298, "learning_rate": 9.967166108224957e-05, "loss": 0.0203, "step": 780 }, { "epoch": 0.17205706196232168, "grad_norm": 0.4532650411128998, "learning_rate": 9.965077728497348e-05, "loss": 0.0189, "step": 790 }, { "epoch": 0.17423499945551563, "grad_norm": 0.3554689884185791, "learning_rate": 9.96292519791248e-05, "loss": 0.0182, "step": 800 }, { "epoch": 0.17641293694870958, "grad_norm": 0.30251172184944153, "learning_rate": 9.96070854428194e-05, "loss": 0.0162, "step": 810 }, { "epoch": 0.17859087444190352, "grad_norm": 0.2571502923965454, "learning_rate": 9.958427796245808e-05, "loss": 0.0171, "step": 820 }, { "epoch": 0.18076881193509747, "grad_norm": 0.30262139439582825, "learning_rate": 9.956082983272293e-05, "loss": 0.0167, "step": 830 }, { "epoch": 0.1829467494282914, "grad_norm": 0.38527029752731323, "learning_rate": 9.953674135657345e-05, "loss": 0.0184, "step": 840 }, { "epoch": 0.18512468692148534, "grad_norm": 0.31793850660324097, "learning_rate": 9.951201284524275e-05, "loss": 0.0205, "step": 850 }, { "epoch": 0.1873026244146793, "grad_norm": 0.24737121164798737, "learning_rate": 9.94866446182334e-05, "loss": 0.0181, "step": 860 }, { "epoch": 0.18948056190787324, "grad_norm": 0.2765544652938843, "learning_rate": 9.94606370033134e-05, "loss": 0.0171, "step": 870 }, { "epoch": 0.1916584994010672, "grad_norm": 0.2321496307849884, "learning_rate": 9.943399033651189e-05, "loss": 0.0172, "step": 880 }, { "epoch": 0.19383643689426114, "grad_norm": 0.3293866515159607, "learning_rate": 9.94067049621148e-05, "loss": 0.0181, "step": 890 }, { "epoch": 0.1960143743874551, "grad_norm": 0.4550676941871643, "learning_rate": 9.937878123266044e-05, "loss": 0.0188, "step": 900 }, { "epoch": 0.19819231188064904, "grad_norm": 0.4054636061191559, "learning_rate": 9.9350219508935e-05, "loss": 0.0171, "step": 910 }, { "epoch": 0.20037024937384296, "grad_norm": 0.4267730116844177, "learning_rate": 9.93210201599677e-05, "loss": 0.0168, "step": 920 }, { "epoch": 0.2025481868670369, "grad_norm": 0.2319348007440567, "learning_rate": 9.929118356302621e-05, "loss": 0.0184, "step": 930 }, { "epoch": 0.20472612436023085, "grad_norm": 0.34213143587112427, "learning_rate": 9.926071010361173e-05, "loss": 0.0168, "step": 940 }, { "epoch": 0.2069040618534248, "grad_norm": 0.299695760011673, "learning_rate": 9.922960017545395e-05, "loss": 0.0179, "step": 950 }, { "epoch": 0.20908199934661875, "grad_norm": 0.17575781047344208, "learning_rate": 9.919785418050598e-05, "loss": 0.0167, "step": 960 }, { "epoch": 0.2112599368398127, "grad_norm": 0.25471341609954834, "learning_rate": 9.916547252893923e-05, "loss": 0.0167, "step": 970 }, { "epoch": 0.21343787433300665, "grad_norm": 0.2856571674346924, "learning_rate": 9.9132455639138e-05, "loss": 0.0169, "step": 980 }, { "epoch": 0.2156158118262006, "grad_norm": 0.25572776794433594, "learning_rate": 9.90988039376942e-05, "loss": 0.0176, "step": 990 }, { "epoch": 0.21779374931939452, "grad_norm": 0.24768424034118652, "learning_rate": 9.906451785940167e-05, "loss": 0.0162, "step": 1000 }, { "epoch": 0.21997168681258847, "grad_norm": 0.33404502272605896, "learning_rate": 9.902959784725077e-05, "loss": 0.0168, "step": 1010 }, { "epoch": 0.22214962430578242, "grad_norm": 0.27414974570274353, "learning_rate": 9.899404435242246e-05, "loss": 0.0164, "step": 1020 }, { "epoch": 0.22432756179897637, "grad_norm": 0.2570934295654297, "learning_rate": 9.895785783428262e-05, "loss": 0.0168, "step": 1030 }, { "epoch": 0.22650549929217031, "grad_norm": 0.1893799602985382, "learning_rate": 9.8921038760376e-05, "loss": 0.0162, "step": 1040 }, { "epoch": 0.22868343678536426, "grad_norm": 0.24636855721473694, "learning_rate": 9.888358760642029e-05, "loss": 0.0165, "step": 1050 }, { "epoch": 0.2308613742785582, "grad_norm": 0.29858240485191345, "learning_rate": 9.884550485629987e-05, "loss": 0.0149, "step": 1060 }, { "epoch": 0.23303931177175216, "grad_norm": 0.2941649854183197, "learning_rate": 9.88067910020596e-05, "loss": 0.0162, "step": 1070 }, { "epoch": 0.23521724926494608, "grad_norm": 0.2786182761192322, "learning_rate": 9.876744654389854e-05, "loss": 0.0155, "step": 1080 }, { "epoch": 0.23739518675814003, "grad_norm": 0.3152257204055786, "learning_rate": 9.872747199016328e-05, "loss": 0.0145, "step": 1090 }, { "epoch": 0.23957312425133398, "grad_norm": 0.30433782935142517, "learning_rate": 9.868686785734165e-05, "loss": 0.0174, "step": 1100 }, { "epoch": 0.24175106174452793, "grad_norm": 0.3568665087223053, "learning_rate": 9.86456346700558e-05, "loss": 0.0178, "step": 1110 }, { "epoch": 0.24392899923772188, "grad_norm": 0.33190423250198364, "learning_rate": 9.860377296105556e-05, "loss": 0.0153, "step": 1120 }, { "epoch": 0.24610693673091583, "grad_norm": 0.25492173433303833, "learning_rate": 9.856128327121155e-05, "loss": 0.0181, "step": 1130 }, { "epoch": 0.24828487422410978, "grad_norm": 0.2693391442298889, "learning_rate": 9.85181661495081e-05, "loss": 0.0163, "step": 1140 }, { "epoch": 0.2504628117173037, "grad_norm": 0.24566078186035156, "learning_rate": 9.847442215303626e-05, "loss": 0.0169, "step": 1150 }, { "epoch": 0.2526407492104977, "grad_norm": 0.18821237981319427, "learning_rate": 9.843005184698655e-05, "loss": 0.0178, "step": 1160 }, { "epoch": 0.2548186867036916, "grad_norm": 0.1899777352809906, "learning_rate": 9.838505580464168e-05, "loss": 0.0165, "step": 1170 }, { "epoch": 0.25699662419688557, "grad_norm": 0.23964695632457733, "learning_rate": 9.833943460736912e-05, "loss": 0.0186, "step": 1180 }, { "epoch": 0.2591745616900795, "grad_norm": 0.2561779022216797, "learning_rate": 9.829318884461359e-05, "loss": 0.0172, "step": 1190 }, { "epoch": 0.2613524991832734, "grad_norm": 0.27848705649375916, "learning_rate": 9.824631911388948e-05, "loss": 0.0173, "step": 1200 }, { "epoch": 0.26353043667646736, "grad_norm": 0.23972542583942413, "learning_rate": 9.819882602077309e-05, "loss": 0.0189, "step": 1210 }, { "epoch": 0.2657083741696613, "grad_norm": 0.24309852719306946, "learning_rate": 9.815071017889482e-05, "loss": 0.0179, "step": 1220 }, { "epoch": 0.26788631166285526, "grad_norm": 0.2942643165588379, "learning_rate": 9.810197220993123e-05, "loss": 0.0167, "step": 1230 }, { "epoch": 0.2700642491560492, "grad_norm": 0.14913658797740936, "learning_rate": 9.805261274359705e-05, "loss": 0.0159, "step": 1240 }, { "epoch": 0.27224218664924316, "grad_norm": 0.43585604429244995, "learning_rate": 9.800263241763698e-05, "loss": 0.0179, "step": 1250 }, { "epoch": 0.2744201241424371, "grad_norm": 0.16332468390464783, "learning_rate": 9.795203187781751e-05, "loss": 0.017, "step": 1260 }, { "epoch": 0.27659806163563105, "grad_norm": 0.25408506393432617, "learning_rate": 9.790081177791852e-05, "loss": 0.0152, "step": 1270 }, { "epoch": 0.278775999128825, "grad_norm": 0.303528368473053, "learning_rate": 9.784897277972491e-05, "loss": 0.0168, "step": 1280 }, { "epoch": 0.28095393662201895, "grad_norm": 0.2518719732761383, "learning_rate": 9.779651555301794e-05, "loss": 0.0138, "step": 1290 }, { "epoch": 0.2831318741152129, "grad_norm": 0.3640092611312866, "learning_rate": 9.77434407755667e-05, "loss": 0.0145, "step": 1300 }, { "epoch": 0.28530981160840685, "grad_norm": 0.2869015038013458, "learning_rate": 9.768974913311922e-05, "loss": 0.0155, "step": 1310 }, { "epoch": 0.2874877491016008, "grad_norm": 0.27514615654945374, "learning_rate": 9.763544131939374e-05, "loss": 0.0155, "step": 1320 }, { "epoch": 0.28966568659479475, "grad_norm": 0.25864166021347046, "learning_rate": 9.758051803606971e-05, "loss": 0.0153, "step": 1330 }, { "epoch": 0.2918436240879887, "grad_norm": 0.25849685072898865, "learning_rate": 9.75249799927786e-05, "loss": 0.0148, "step": 1340 }, { "epoch": 0.29402156158118264, "grad_norm": 0.19816963374614716, "learning_rate": 9.746882790709491e-05, "loss": 0.0149, "step": 1350 }, { "epoch": 0.29619949907437654, "grad_norm": 0.14698155224323273, "learning_rate": 9.741206250452683e-05, "loss": 0.013, "step": 1360 }, { "epoch": 0.2983774365675705, "grad_norm": 0.24931158125400543, "learning_rate": 9.735468451850681e-05, "loss": 0.0149, "step": 1370 }, { "epoch": 0.30055537406076444, "grad_norm": 0.2916138768196106, "learning_rate": 9.729669469038216e-05, "loss": 0.0127, "step": 1380 }, { "epoch": 0.3027333115539584, "grad_norm": 0.22157330811023712, "learning_rate": 9.723809376940544e-05, "loss": 0.0146, "step": 1390 }, { "epoch": 0.30491124904715233, "grad_norm": 0.2587496042251587, "learning_rate": 9.717888251272477e-05, "loss": 0.0136, "step": 1400 }, { "epoch": 0.3070891865403463, "grad_norm": 0.25592541694641113, "learning_rate": 9.71190616853741e-05, "loss": 0.0153, "step": 1410 }, { "epoch": 0.30926712403354023, "grad_norm": 0.23283854126930237, "learning_rate": 9.705863206026321e-05, "loss": 0.0129, "step": 1420 }, { "epoch": 0.3114450615267342, "grad_norm": 0.28844064474105835, "learning_rate": 9.699759441816787e-05, "loss": 0.0161, "step": 1430 }, { "epoch": 0.31362299901992813, "grad_norm": 0.1569090634584427, "learning_rate": 9.693594954771965e-05, "loss": 0.0151, "step": 1440 }, { "epoch": 0.3158009365131221, "grad_norm": 0.3367060720920563, "learning_rate": 9.687369824539577e-05, "loss": 0.014, "step": 1450 }, { "epoch": 0.317978874006316, "grad_norm": 0.26260900497436523, "learning_rate": 9.68108413155088e-05, "loss": 0.0147, "step": 1460 }, { "epoch": 0.32015681149951, "grad_norm": 0.15449059009552002, "learning_rate": 9.674737957019624e-05, "loss": 0.0167, "step": 1470 }, { "epoch": 0.3223347489927039, "grad_norm": 0.26870304346084595, "learning_rate": 9.66833138294101e-05, "loss": 0.0145, "step": 1480 }, { "epoch": 0.32451268648589787, "grad_norm": 0.38192567229270935, "learning_rate": 9.661864492090625e-05, "loss": 0.0136, "step": 1490 }, { "epoch": 0.3266906239790918, "grad_norm": 0.410439133644104, "learning_rate": 9.655337368023371e-05, "loss": 0.0135, "step": 1500 }, { "epoch": 0.32886856147228577, "grad_norm": 0.4695938229560852, "learning_rate": 9.64875009507239e-05, "loss": 0.0142, "step": 1510 }, { "epoch": 0.33104649896547966, "grad_norm": 0.24386045336723328, "learning_rate": 9.642102758347973e-05, "loss": 0.0142, "step": 1520 }, { "epoch": 0.3332244364586736, "grad_norm": 0.19882968068122864, "learning_rate": 9.63539544373646e-05, "loss": 0.0136, "step": 1530 }, { "epoch": 0.33540237395186756, "grad_norm": 0.254899263381958, "learning_rate": 9.628628237899126e-05, "loss": 0.0146, "step": 1540 }, { "epoch": 0.3375803114450615, "grad_norm": 0.17491990327835083, "learning_rate": 9.621801228271073e-05, "loss": 0.0152, "step": 1550 }, { "epoch": 0.33975824893825546, "grad_norm": 0.3666568994522095, "learning_rate": 9.614914503060083e-05, "loss": 0.0142, "step": 1560 }, { "epoch": 0.3419361864314494, "grad_norm": 0.2654305398464203, "learning_rate": 9.607968151245498e-05, "loss": 0.0155, "step": 1570 }, { "epoch": 0.34411412392464336, "grad_norm": 0.18583115935325623, "learning_rate": 9.600962262577053e-05, "loss": 0.0144, "step": 1580 }, { "epoch": 0.3462920614178373, "grad_norm": 0.20424164831638336, "learning_rate": 9.593896927573728e-05, "loss": 0.0141, "step": 1590 }, { "epoch": 0.34846999891103125, "grad_norm": 0.20820754766464233, "learning_rate": 9.586772237522573e-05, "loss": 0.0135, "step": 1600 }, { "epoch": 0.3506479364042252, "grad_norm": 0.22122822701931, "learning_rate": 9.579588284477526e-05, "loss": 0.0128, "step": 1610 }, { "epoch": 0.35282587389741915, "grad_norm": 0.23795032501220703, "learning_rate": 9.572345161258235e-05, "loss": 0.0123, "step": 1620 }, { "epoch": 0.3550038113906131, "grad_norm": 0.3005674481391907, "learning_rate": 9.565042961448844e-05, "loss": 0.0131, "step": 1630 }, { "epoch": 0.35718174888380705, "grad_norm": 0.24608589708805084, "learning_rate": 9.557681779396797e-05, "loss": 0.0133, "step": 1640 }, { "epoch": 0.359359686377001, "grad_norm": 0.21148374676704407, "learning_rate": 9.550261710211608e-05, "loss": 0.015, "step": 1650 }, { "epoch": 0.36153762387019495, "grad_norm": 0.3020222783088684, "learning_rate": 9.542782849763637e-05, "loss": 0.0134, "step": 1660 }, { "epoch": 0.3637155613633889, "grad_norm": 0.24635276198387146, "learning_rate": 9.535245294682857e-05, "loss": 0.0137, "step": 1670 }, { "epoch": 0.3658934988565828, "grad_norm": 0.28255170583724976, "learning_rate": 9.527649142357596e-05, "loss": 0.0132, "step": 1680 }, { "epoch": 0.36807143634977674, "grad_norm": 0.1997251659631729, "learning_rate": 9.519994490933279e-05, "loss": 0.0138, "step": 1690 }, { "epoch": 0.3702493738429707, "grad_norm": 0.2821415960788727, "learning_rate": 9.51228143931117e-05, "loss": 0.0137, "step": 1700 }, { "epoch": 0.37242731133616463, "grad_norm": 0.23824472725391388, "learning_rate": 9.504510087147088e-05, "loss": 0.0121, "step": 1710 }, { "epoch": 0.3746052488293586, "grad_norm": 0.3249528408050537, "learning_rate": 9.496680534850113e-05, "loss": 0.0157, "step": 1720 }, { "epoch": 0.37678318632255253, "grad_norm": 0.3454100489616394, "learning_rate": 9.488792883581299e-05, "loss": 0.0157, "step": 1730 }, { "epoch": 0.3789611238157465, "grad_norm": 0.3697403073310852, "learning_rate": 9.480847235252361e-05, "loss": 0.014, "step": 1740 }, { "epoch": 0.38113906130894043, "grad_norm": 0.2749633193016052, "learning_rate": 9.472843692524363e-05, "loss": 0.0147, "step": 1750 }, { "epoch": 0.3833169988021344, "grad_norm": 0.2249903827905655, "learning_rate": 9.464782358806383e-05, "loss": 0.0137, "step": 1760 }, { "epoch": 0.3854949362953283, "grad_norm": 0.21319381892681122, "learning_rate": 9.45666333825419e-05, "loss": 0.0147, "step": 1770 }, { "epoch": 0.3876728737885223, "grad_norm": 0.23701409995555878, "learning_rate": 9.448486735768884e-05, "loss": 0.0134, "step": 1780 }, { "epoch": 0.3898508112817162, "grad_norm": 0.1817920058965683, "learning_rate": 9.440252656995551e-05, "loss": 0.0155, "step": 1790 }, { "epoch": 0.3920287487749102, "grad_norm": 0.1825178563594818, "learning_rate": 9.431961208321892e-05, "loss": 0.0143, "step": 1800 }, { "epoch": 0.3942066862681041, "grad_norm": 0.3201183080673218, "learning_rate": 9.423612496876855e-05, "loss": 0.0126, "step": 1810 }, { "epoch": 0.39638462376129807, "grad_norm": 0.3739440143108368, "learning_rate": 9.415206630529241e-05, "loss": 0.013, "step": 1820 }, { "epoch": 0.398562561254492, "grad_norm": 0.35186412930488586, "learning_rate": 9.406743717886321e-05, "loss": 0.0148, "step": 1830 }, { "epoch": 0.4007404987476859, "grad_norm": 0.34517958760261536, "learning_rate": 9.398223868292424e-05, "loss": 0.0122, "step": 1840 }, { "epoch": 0.40291843624087986, "grad_norm": 0.24918566644191742, "learning_rate": 9.389647191827533e-05, "loss": 0.0138, "step": 1850 }, { "epoch": 0.4050963737340738, "grad_norm": 0.24154038727283478, "learning_rate": 9.38101379930585e-05, "loss": 0.0131, "step": 1860 }, { "epoch": 0.40727431122726776, "grad_norm": 0.19564920663833618, "learning_rate": 9.372323802274379e-05, "loss": 0.0129, "step": 1870 }, { "epoch": 0.4094522487204617, "grad_norm": 0.1982380896806717, "learning_rate": 9.363577313011473e-05, "loss": 0.0131, "step": 1880 }, { "epoch": 0.41163018621365566, "grad_norm": 0.2124156355857849, "learning_rate": 9.354774444525391e-05, "loss": 0.0118, "step": 1890 }, { "epoch": 0.4138081237068496, "grad_norm": 0.2862381041049957, "learning_rate": 9.345915310552835e-05, "loss": 0.0131, "step": 1900 }, { "epoch": 0.41598606120004356, "grad_norm": 0.27992814779281616, "learning_rate": 9.337000025557476e-05, "loss": 0.0124, "step": 1910 }, { "epoch": 0.4181639986932375, "grad_norm": 0.2013150304555893, "learning_rate": 9.328028704728486e-05, "loss": 0.013, "step": 1920 }, { "epoch": 0.42034193618643145, "grad_norm": 0.2200571894645691, "learning_rate": 9.319001463979036e-05, "loss": 0.0122, "step": 1930 }, { "epoch": 0.4225198736796254, "grad_norm": 0.16203193366527557, "learning_rate": 9.309918419944812e-05, "loss": 0.0125, "step": 1940 }, { "epoch": 0.42469781117281935, "grad_norm": 0.16349908709526062, "learning_rate": 9.300779689982498e-05, "loss": 0.0116, "step": 1950 }, { "epoch": 0.4268757486660133, "grad_norm": 0.21969835460186005, "learning_rate": 9.291585392168262e-05, "loss": 0.015, "step": 1960 }, { "epoch": 0.42905368615920725, "grad_norm": 0.2813135087490082, "learning_rate": 9.282335645296236e-05, "loss": 0.0129, "step": 1970 }, { "epoch": 0.4312316236524012, "grad_norm": 0.20814667642116547, "learning_rate": 9.273030568876972e-05, "loss": 0.0127, "step": 1980 }, { "epoch": 0.43340956114559515, "grad_norm": 0.186491921544075, "learning_rate": 9.263670283135908e-05, "loss": 0.0157, "step": 1990 }, { "epoch": 0.43558749863878904, "grad_norm": 0.1562117636203766, "learning_rate": 9.254254909011804e-05, "loss": 0.0136, "step": 2000 }, { "epoch": 0.437765436131983, "grad_norm": 0.22019344568252563, "learning_rate": 9.244784568155186e-05, "loss": 0.0132, "step": 2010 }, { "epoch": 0.43994337362517694, "grad_norm": 0.11756141483783722, "learning_rate": 9.235259382926775e-05, "loss": 0.0117, "step": 2020 }, { "epoch": 0.4421213111183709, "grad_norm": 0.24443650245666504, "learning_rate": 9.225679476395904e-05, "loss": 0.0121, "step": 2030 }, { "epoch": 0.44429924861156483, "grad_norm": 0.15402700006961823, "learning_rate": 9.216044972338924e-05, "loss": 0.0116, "step": 2040 }, { "epoch": 0.4464771861047588, "grad_norm": 0.2157009243965149, "learning_rate": 9.206355995237614e-05, "loss": 0.0123, "step": 2050 }, { "epoch": 0.44865512359795273, "grad_norm": 0.2419036626815796, "learning_rate": 9.196612670277561e-05, "loss": 0.0128, "step": 2060 }, { "epoch": 0.4508330610911467, "grad_norm": 0.2521984577178955, "learning_rate": 9.186815123346555e-05, "loss": 0.0129, "step": 2070 }, { "epoch": 0.45301099858434063, "grad_norm": 0.23165738582611084, "learning_rate": 9.176963481032951e-05, "loss": 0.0129, "step": 2080 }, { "epoch": 0.4551889360775346, "grad_norm": 0.22679926455020905, "learning_rate": 9.167057870624045e-05, "loss": 0.0117, "step": 2090 }, { "epoch": 0.4573668735707285, "grad_norm": 0.17972847819328308, "learning_rate": 9.157098420104416e-05, "loss": 0.0113, "step": 2100 }, { "epoch": 0.4595448110639225, "grad_norm": 0.18063072860240936, "learning_rate": 9.147085258154284e-05, "loss": 0.0115, "step": 2110 }, { "epoch": 0.4617227485571164, "grad_norm": 0.20134253799915314, "learning_rate": 9.137018514147842e-05, "loss": 0.0126, "step": 2120 }, { "epoch": 0.4639006860503104, "grad_norm": 0.2901334762573242, "learning_rate": 9.126898318151585e-05, "loss": 0.012, "step": 2130 }, { "epoch": 0.4660786235435043, "grad_norm": 0.21262285113334656, "learning_rate": 9.116724800922629e-05, "loss": 0.0113, "step": 2140 }, { "epoch": 0.46825656103669827, "grad_norm": 0.17349544167518616, "learning_rate": 9.106498093907024e-05, "loss": 0.0121, "step": 2150 }, { "epoch": 0.47043449852989216, "grad_norm": 0.18809765577316284, "learning_rate": 9.096218329238053e-05, "loss": 0.0126, "step": 2160 }, { "epoch": 0.4726124360230861, "grad_norm": 0.2807271182537079, "learning_rate": 9.085885639734527e-05, "loss": 0.0125, "step": 2170 }, { "epoch": 0.47479037351628006, "grad_norm": 0.3028620779514313, "learning_rate": 9.075500158899067e-05, "loss": 0.0117, "step": 2180 }, { "epoch": 0.476968311009474, "grad_norm": 0.2349446713924408, "learning_rate": 9.065062020916377e-05, "loss": 0.013, "step": 2190 }, { "epoch": 0.47914624850266796, "grad_norm": 0.2204253226518631, "learning_rate": 9.054571360651517e-05, "loss": 0.0114, "step": 2200 }, { "epoch": 0.4813241859958619, "grad_norm": 0.19165095686912537, "learning_rate": 9.044028313648157e-05, "loss": 0.0126, "step": 2210 }, { "epoch": 0.48350212348905586, "grad_norm": 0.15540415048599243, "learning_rate": 9.033433016126822e-05, "loss": 0.0122, "step": 2220 }, { "epoch": 0.4856800609822498, "grad_norm": 0.15101611614227295, "learning_rate": 9.022785604983139e-05, "loss": 0.0114, "step": 2230 }, { "epoch": 0.48785799847544375, "grad_norm": 0.20519086718559265, "learning_rate": 9.01208621778606e-05, "loss": 0.0118, "step": 2240 }, { "epoch": 0.4900359359686377, "grad_norm": 0.18452398478984833, "learning_rate": 9.001334992776094e-05, "loss": 0.0119, "step": 2250 }, { "epoch": 0.49221387346183165, "grad_norm": 0.2507144808769226, "learning_rate": 8.990532068863513e-05, "loss": 0.0101, "step": 2260 }, { "epoch": 0.4943918109550256, "grad_norm": 0.29505401849746704, "learning_rate": 8.979677585626559e-05, "loss": 0.0111, "step": 2270 }, { "epoch": 0.49656974844821955, "grad_norm": 0.3609241545200348, "learning_rate": 8.968771683309645e-05, "loss": 0.0114, "step": 2280 }, { "epoch": 0.4987476859414135, "grad_norm": 0.3754465579986572, "learning_rate": 8.95781450282154e-05, "loss": 0.0117, "step": 2290 }, { "epoch": 0.5009256234346074, "grad_norm": 0.25874945521354675, "learning_rate": 8.946806185733543e-05, "loss": 0.0124, "step": 2300 }, { "epoch": 0.5031035609278014, "grad_norm": 0.2572658360004425, "learning_rate": 8.935746874277667e-05, "loss": 0.012, "step": 2310 }, { "epoch": 0.5052814984209953, "grad_norm": 0.23539291322231293, "learning_rate": 8.924636711344784e-05, "loss": 0.0116, "step": 2320 }, { "epoch": 0.5074594359141893, "grad_norm": 0.2633945047855377, "learning_rate": 8.913475840482797e-05, "loss": 0.012, "step": 2330 }, { "epoch": 0.5096373734073832, "grad_norm": 0.25647762417793274, "learning_rate": 8.902264405894771e-05, "loss": 0.0122, "step": 2340 }, { "epoch": 0.5118153109005772, "grad_norm": 0.19141608476638794, "learning_rate": 8.891002552437076e-05, "loss": 0.013, "step": 2350 }, { "epoch": 0.5139932483937711, "grad_norm": 0.2109457105398178, "learning_rate": 8.879690425617517e-05, "loss": 0.013, "step": 2360 }, { "epoch": 0.5161711858869651, "grad_norm": 0.20676490664482117, "learning_rate": 8.868328171593448e-05, "loss": 0.0123, "step": 2370 }, { "epoch": 0.518349123380159, "grad_norm": 0.257219135761261, "learning_rate": 8.85691593716989e-05, "loss": 0.0131, "step": 2380 }, { "epoch": 0.520527060873353, "grad_norm": 0.2181825041770935, "learning_rate": 8.845453869797631e-05, "loss": 0.0129, "step": 2390 }, { "epoch": 0.5227049983665468, "grad_norm": 0.20766574144363403, "learning_rate": 8.833942117571321e-05, "loss": 0.0111, "step": 2400 }, { "epoch": 0.5248829358597408, "grad_norm": 0.162784144282341, "learning_rate": 8.822380829227559e-05, "loss": 0.0122, "step": 2410 }, { "epoch": 0.5270608733529347, "grad_norm": 0.24111758172512054, "learning_rate": 8.810770154142969e-05, "loss": 0.0127, "step": 2420 }, { "epoch": 0.5292388108461287, "grad_norm": 0.29185184836387634, "learning_rate": 8.799110242332276e-05, "loss": 0.0126, "step": 2430 }, { "epoch": 0.5314167483393226, "grad_norm": 0.29338541626930237, "learning_rate": 8.787401244446361e-05, "loss": 0.0119, "step": 2440 }, { "epoch": 0.5335946858325166, "grad_norm": 0.2108069658279419, "learning_rate": 8.775643311770318e-05, "loss": 0.0115, "step": 2450 }, { "epoch": 0.5357726233257105, "grad_norm": 0.3178095519542694, "learning_rate": 8.7638365962215e-05, "loss": 0.0125, "step": 2460 }, { "epoch": 0.5379505608189045, "grad_norm": 0.15945272147655487, "learning_rate": 8.751981250347552e-05, "loss": 0.0128, "step": 2470 }, { "epoch": 0.5401284983120984, "grad_norm": 0.18829770386219025, "learning_rate": 8.740077427324446e-05, "loss": 0.0136, "step": 2480 }, { "epoch": 0.5423064358052924, "grad_norm": 0.20409831404685974, "learning_rate": 8.728125280954498e-05, "loss": 0.0113, "step": 2490 }, { "epoch": 0.5444843732984863, "grad_norm": 0.2616059184074402, "learning_rate": 8.716124965664379e-05, "loss": 0.0119, "step": 2500 }, { "epoch": 0.5466623107916803, "grad_norm": 0.1775239109992981, "learning_rate": 8.704076636503128e-05, "loss": 0.0128, "step": 2510 }, { "epoch": 0.5488402482848742, "grad_norm": 0.22858548164367676, "learning_rate": 8.691980449140135e-05, "loss": 0.0124, "step": 2520 }, { "epoch": 0.5510181857780682, "grad_norm": 0.29392021894454956, "learning_rate": 8.679836559863148e-05, "loss": 0.0115, "step": 2530 }, { "epoch": 0.5531961232712621, "grad_norm": 0.23020786046981812, "learning_rate": 8.667645125576235e-05, "loss": 0.011, "step": 2540 }, { "epoch": 0.5553740607644561, "grad_norm": 0.12346918135881424, "learning_rate": 8.655406303797767e-05, "loss": 0.0108, "step": 2550 }, { "epoch": 0.55755199825765, "grad_norm": 0.12898415327072144, "learning_rate": 8.643120252658381e-05, "loss": 0.0119, "step": 2560 }, { "epoch": 0.559729935750844, "grad_norm": 0.17225253582000732, "learning_rate": 8.630787130898943e-05, "loss": 0.0121, "step": 2570 }, { "epoch": 0.5619078732440379, "grad_norm": 0.1442943513393402, "learning_rate": 8.618407097868482e-05, "loss": 0.0119, "step": 2580 }, { "epoch": 0.5640858107372319, "grad_norm": 0.2083996832370758, "learning_rate": 8.605980313522142e-05, "loss": 0.0092, "step": 2590 }, { "epoch": 0.5662637482304258, "grad_norm": 0.2155020833015442, "learning_rate": 8.59350693841912e-05, "loss": 0.0109, "step": 2600 }, { "epoch": 0.5684416857236197, "grad_norm": 0.18167933821678162, "learning_rate": 8.580987133720576e-05, "loss": 0.0112, "step": 2610 }, { "epoch": 0.5706196232168137, "grad_norm": 0.18125946819782257, "learning_rate": 8.568421061187567e-05, "loss": 0.0097, "step": 2620 }, { "epoch": 0.5727975607100076, "grad_norm": 0.1901247203350067, "learning_rate": 8.55580888317894e-05, "loss": 0.0113, "step": 2630 }, { "epoch": 0.5749754982032016, "grad_norm": 0.1511031538248062, "learning_rate": 8.543150762649257e-05, "loss": 0.0102, "step": 2640 }, { "epoch": 0.5771534356963955, "grad_norm": 0.17375598847866058, "learning_rate": 8.530446863146664e-05, "loss": 0.0106, "step": 2650 }, { "epoch": 0.5793313731895895, "grad_norm": 0.27421876788139343, "learning_rate": 8.517697348810798e-05, "loss": 0.01, "step": 2660 }, { "epoch": 0.5815093106827834, "grad_norm": 0.213004931807518, "learning_rate": 8.504902384370657e-05, "loss": 0.0107, "step": 2670 }, { "epoch": 0.5836872481759774, "grad_norm": 0.19220004975795746, "learning_rate": 8.492062135142469e-05, "loss": 0.011, "step": 2680 }, { "epoch": 0.5858651856691713, "grad_norm": 0.21476730704307556, "learning_rate": 8.479176767027566e-05, "loss": 0.0104, "step": 2690 }, { "epoch": 0.5880431231623653, "grad_norm": 0.24100437760353088, "learning_rate": 8.466246446510231e-05, "loss": 0.0111, "step": 2700 }, { "epoch": 0.5902210606555592, "grad_norm": 0.1968834400177002, "learning_rate": 8.45327134065555e-05, "loss": 0.011, "step": 2710 }, { "epoch": 0.5923989981487531, "grad_norm": 0.26684340834617615, "learning_rate": 8.44025161710726e-05, "loss": 0.0102, "step": 2720 }, { "epoch": 0.594576935641947, "grad_norm": 0.14660967886447906, "learning_rate": 8.42718744408557e-05, "loss": 0.0092, "step": 2730 }, { "epoch": 0.596754873135141, "grad_norm": 0.1794154793024063, "learning_rate": 8.414078990384995e-05, "loss": 0.0105, "step": 2740 }, { "epoch": 0.5989328106283349, "grad_norm": 0.13432596623897552, "learning_rate": 8.400926425372182e-05, "loss": 0.0114, "step": 2750 }, { "epoch": 0.6011107481215289, "grad_norm": 0.13757142424583435, "learning_rate": 8.387729918983706e-05, "loss": 0.0099, "step": 2760 }, { "epoch": 0.6032886856147228, "grad_norm": 0.3109431564807892, "learning_rate": 8.374489641723886e-05, "loss": 0.0109, "step": 2770 }, { "epoch": 0.6054666231079168, "grad_norm": 0.28936275839805603, "learning_rate": 8.36120576466258e-05, "loss": 0.0113, "step": 2780 }, { "epoch": 0.6076445606011107, "grad_norm": 0.24570229649543762, "learning_rate": 8.347878459432973e-05, "loss": 0.0135, "step": 2790 }, { "epoch": 0.6098224980943047, "grad_norm": 0.15275812149047852, "learning_rate": 8.334507898229357e-05, "loss": 0.0128, "step": 2800 }, { "epoch": 0.6120004355874986, "grad_norm": 0.20751699805259705, "learning_rate": 8.321094253804915e-05, "loss": 0.0117, "step": 2810 }, { "epoch": 0.6141783730806926, "grad_norm": 0.24923484027385712, "learning_rate": 8.307637699469477e-05, "loss": 0.0105, "step": 2820 }, { "epoch": 0.6163563105738865, "grad_norm": 0.1519884169101715, "learning_rate": 8.29413840908729e-05, "loss": 0.0103, "step": 2830 }, { "epoch": 0.6185342480670805, "grad_norm": 0.1287517100572586, "learning_rate": 8.280596557074767e-05, "loss": 0.0112, "step": 2840 }, { "epoch": 0.6207121855602744, "grad_norm": 0.20145294070243835, "learning_rate": 8.267012318398234e-05, "loss": 0.0106, "step": 2850 }, { "epoch": 0.6228901230534684, "grad_norm": 0.18479537963867188, "learning_rate": 8.253385868571672e-05, "loss": 0.01, "step": 2860 }, { "epoch": 0.6250680605466623, "grad_norm": 0.15948829054832458, "learning_rate": 8.23971738365445e-05, "loss": 0.0123, "step": 2870 }, { "epoch": 0.6272459980398563, "grad_norm": 0.1715373694896698, "learning_rate": 8.226007040249044e-05, "loss": 0.0105, "step": 2880 }, { "epoch": 0.6294239355330502, "grad_norm": 0.17077109217643738, "learning_rate": 8.212255015498758e-05, "loss": 0.0113, "step": 2890 }, { "epoch": 0.6316018730262442, "grad_norm": 0.18706509470939636, "learning_rate": 8.198461487085438e-05, "loss": 0.0104, "step": 2900 }, { "epoch": 0.6337798105194381, "grad_norm": 0.30177706480026245, "learning_rate": 8.184626633227178e-05, "loss": 0.0125, "step": 2910 }, { "epoch": 0.635957748012632, "grad_norm": 0.3015088140964508, "learning_rate": 8.170750632676006e-05, "loss": 0.0112, "step": 2920 }, { "epoch": 0.638135685505826, "grad_norm": 0.18643289804458618, "learning_rate": 8.156833664715589e-05, "loss": 0.012, "step": 2930 }, { "epoch": 0.64031362299902, "grad_norm": 0.18556185066699982, "learning_rate": 8.142875909158906e-05, "loss": 0.0115, "step": 2940 }, { "epoch": 0.6424915604922139, "grad_norm": 0.15762163698673248, "learning_rate": 8.128877546345933e-05, "loss": 0.0097, "step": 2950 }, { "epoch": 0.6446694979854078, "grad_norm": 0.23215970396995544, "learning_rate": 8.114838757141304e-05, "loss": 0.0124, "step": 2960 }, { "epoch": 0.6468474354786018, "grad_norm": 0.2265307605266571, "learning_rate": 8.100759722931981e-05, "loss": 0.0125, "step": 2970 }, { "epoch": 0.6490253729717957, "grad_norm": 0.2771134376525879, "learning_rate": 8.086640625624905e-05, "loss": 0.0118, "step": 2980 }, { "epoch": 0.6512033104649897, "grad_norm": 0.1871901899576187, "learning_rate": 8.072481647644656e-05, "loss": 0.0105, "step": 2990 }, { "epoch": 0.6533812479581836, "grad_norm": 0.2593614459037781, "learning_rate": 8.05828297193108e-05, "loss": 0.0105, "step": 3000 }, { "epoch": 0.6555591854513776, "grad_norm": 0.22635385394096375, "learning_rate": 8.044044781936942e-05, "loss": 0.0105, "step": 3010 }, { "epoch": 0.6577371229445715, "grad_norm": 0.23502841591835022, "learning_rate": 8.029767261625538e-05, "loss": 0.0112, "step": 3020 }, { "epoch": 0.6599150604377655, "grad_norm": 0.14734604954719543, "learning_rate": 8.015450595468341e-05, "loss": 0.0101, "step": 3030 }, { "epoch": 0.6620929979309593, "grad_norm": 0.17127743363380432, "learning_rate": 8.001094968442592e-05, "loss": 0.0125, "step": 3040 }, { "epoch": 0.6642709354241533, "grad_norm": 0.22311829030513763, "learning_rate": 7.986700566028933e-05, "loss": 0.0108, "step": 3050 }, { "epoch": 0.6664488729173472, "grad_norm": 0.1292509287595749, "learning_rate": 7.972267574208991e-05, "loss": 0.0119, "step": 3060 }, { "epoch": 0.6686268104105412, "grad_norm": 0.20118913054466248, "learning_rate": 7.957796179462989e-05, "loss": 0.0114, "step": 3070 }, { "epoch": 0.6708047479037351, "grad_norm": 0.20826192200183868, "learning_rate": 7.943286568767331e-05, "loss": 0.0093, "step": 3080 }, { "epoch": 0.6729826853969291, "grad_norm": 0.18899671733379364, "learning_rate": 7.92873892959219e-05, "loss": 0.0097, "step": 3090 }, { "epoch": 0.675160622890123, "grad_norm": 0.23138172924518585, "learning_rate": 7.914153449899076e-05, "loss": 0.0106, "step": 3100 }, { "epoch": 0.677338560383317, "grad_norm": 0.28532886505126953, "learning_rate": 7.899530318138417e-05, "loss": 0.011, "step": 3110 }, { "epoch": 0.6795164978765109, "grad_norm": 0.23720648884773254, "learning_rate": 7.884869723247124e-05, "loss": 0.0102, "step": 3120 }, { "epoch": 0.6816944353697049, "grad_norm": 0.22022375464439392, "learning_rate": 7.870171854646144e-05, "loss": 0.0092, "step": 3130 }, { "epoch": 0.6838723728628988, "grad_norm": 0.16302189230918884, "learning_rate": 7.855436902238017e-05, "loss": 0.0094, "step": 3140 }, { "epoch": 0.6860503103560928, "grad_norm": 0.2520020306110382, "learning_rate": 7.840665056404421e-05, "loss": 0.0114, "step": 3150 }, { "epoch": 0.6882282478492867, "grad_norm": 0.21204477548599243, "learning_rate": 7.825856508003713e-05, "loss": 0.0099, "step": 3160 }, { "epoch": 0.6904061853424807, "grad_norm": 0.23702597618103027, "learning_rate": 7.811011448368459e-05, "loss": 0.0103, "step": 3170 }, { "epoch": 0.6925841228356746, "grad_norm": 0.24510057270526886, "learning_rate": 7.796130069302969e-05, "loss": 0.0115, "step": 3180 }, { "epoch": 0.6947620603288686, "grad_norm": 0.27067533135414124, "learning_rate": 7.781212563080816e-05, "loss": 0.0101, "step": 3190 }, { "epoch": 0.6969399978220625, "grad_norm": 0.3111419677734375, "learning_rate": 7.766259122442344e-05, "loss": 0.0123, "step": 3200 }, { "epoch": 0.6991179353152565, "grad_norm": 0.24340444803237915, "learning_rate": 7.751269940592194e-05, "loss": 0.0118, "step": 3210 }, { "epoch": 0.7012958728084504, "grad_norm": 0.29663288593292236, "learning_rate": 7.73624521119679e-05, "loss": 0.0106, "step": 3220 }, { "epoch": 0.7034738103016444, "grad_norm": 0.2705160081386566, "learning_rate": 7.721185128381849e-05, "loss": 0.0105, "step": 3230 }, { "epoch": 0.7056517477948383, "grad_norm": 0.26855409145355225, "learning_rate": 7.706089886729869e-05, "loss": 0.01, "step": 3240 }, { "epoch": 0.7078296852880323, "grad_norm": 0.31059056520462036, "learning_rate": 7.690959681277614e-05, "loss": 0.0123, "step": 3250 }, { "epoch": 0.7100076227812262, "grad_norm": 0.16126227378845215, "learning_rate": 7.675794707513595e-05, "loss": 0.0122, "step": 3260 }, { "epoch": 0.7121855602744201, "grad_norm": 0.3061024248600006, "learning_rate": 7.660595161375541e-05, "loss": 0.0102, "step": 3270 }, { "epoch": 0.7143634977676141, "grad_norm": 0.2468884289264679, "learning_rate": 7.645361239247876e-05, "loss": 0.0116, "step": 3280 }, { "epoch": 0.716541435260808, "grad_norm": 0.2789005935192108, "learning_rate": 7.630093137959171e-05, "loss": 0.0107, "step": 3290 }, { "epoch": 0.718719372754002, "grad_norm": 0.23722700774669647, "learning_rate": 7.614791054779609e-05, "loss": 0.0109, "step": 3300 }, { "epoch": 0.7208973102471959, "grad_norm": 0.22962240874767303, "learning_rate": 7.599455187418432e-05, "loss": 0.0114, "step": 3310 }, { "epoch": 0.7230752477403899, "grad_norm": 0.20258665084838867, "learning_rate": 7.584085734021384e-05, "loss": 0.0119, "step": 3320 }, { "epoch": 0.7252531852335838, "grad_norm": 0.1530407965183258, "learning_rate": 7.568682893168164e-05, "loss": 0.0112, "step": 3330 }, { "epoch": 0.7274311227267778, "grad_norm": 0.13580188155174255, "learning_rate": 7.553246863869841e-05, "loss": 0.01, "step": 3340 }, { "epoch": 0.7296090602199717, "grad_norm": 0.18131376802921295, "learning_rate": 7.537777845566296e-05, "loss": 0.0098, "step": 3350 }, { "epoch": 0.7317869977131656, "grad_norm": 0.11071394383907318, "learning_rate": 7.522276038123643e-05, "loss": 0.0109, "step": 3360 }, { "epoch": 0.7339649352063595, "grad_norm": 0.15888631343841553, "learning_rate": 7.506741641831645e-05, "loss": 0.0092, "step": 3370 }, { "epoch": 0.7361428726995535, "grad_norm": 0.1531762182712555, "learning_rate": 7.491174857401122e-05, "loss": 0.0119, "step": 3380 }, { "epoch": 0.7383208101927474, "grad_norm": 0.23579220473766327, "learning_rate": 7.475575885961368e-05, "loss": 0.0113, "step": 3390 }, { "epoch": 0.7404987476859414, "grad_norm": 0.20043380558490753, "learning_rate": 7.459944929057543e-05, "loss": 0.0106, "step": 3400 }, { "epoch": 0.7426766851791353, "grad_norm": 0.1395815908908844, "learning_rate": 7.444282188648075e-05, "loss": 0.0112, "step": 3410 }, { "epoch": 0.7448546226723293, "grad_norm": 0.1925949603319168, "learning_rate": 7.428587867102045e-05, "loss": 0.0099, "step": 3420 }, { "epoch": 0.7470325601655232, "grad_norm": 0.24926616251468658, "learning_rate": 7.412862167196577e-05, "loss": 0.0102, "step": 3430 }, { "epoch": 0.7492104976587172, "grad_norm": 0.20951345562934875, "learning_rate": 7.397105292114217e-05, "loss": 0.0104, "step": 3440 }, { "epoch": 0.7513884351519111, "grad_norm": 0.18474841117858887, "learning_rate": 7.381317445440305e-05, "loss": 0.0099, "step": 3450 }, { "epoch": 0.7535663726451051, "grad_norm": 0.24569986760616302, "learning_rate": 7.365498831160349e-05, "loss": 0.011, "step": 3460 }, { "epoch": 0.755744310138299, "grad_norm": 0.21855206787586212, "learning_rate": 7.349649653657386e-05, "loss": 0.0099, "step": 3470 }, { "epoch": 0.757922247631493, "grad_norm": 0.22094163298606873, "learning_rate": 7.333770117709344e-05, "loss": 0.0098, "step": 3480 }, { "epoch": 0.7601001851246869, "grad_norm": 0.24347667396068573, "learning_rate": 7.317860428486391e-05, "loss": 0.0096, "step": 3490 }, { "epoch": 0.7622781226178809, "grad_norm": 0.1839141547679901, "learning_rate": 7.301920791548294e-05, "loss": 0.0089, "step": 3500 }, { "epoch": 0.7644560601110748, "grad_norm": 0.23142606019973755, "learning_rate": 7.28595141284175e-05, "loss": 0.0105, "step": 3510 }, { "epoch": 0.7666339976042688, "grad_norm": 0.219207763671875, "learning_rate": 7.269952498697734e-05, "loss": 0.0097, "step": 3520 }, { "epoch": 0.7688119350974627, "grad_norm": 0.1916622817516327, "learning_rate": 7.253924255828836e-05, "loss": 0.0088, "step": 3530 }, { "epoch": 0.7709898725906567, "grad_norm": 0.19172070920467377, "learning_rate": 7.237866891326579e-05, "loss": 0.0096, "step": 3540 }, { "epoch": 0.7731678100838506, "grad_norm": 0.1597731113433838, "learning_rate": 7.221780612658752e-05, "loss": 0.0097, "step": 3550 }, { "epoch": 0.7753457475770446, "grad_norm": 0.22632285952568054, "learning_rate": 7.205665627666726e-05, "loss": 0.0097, "step": 3560 }, { "epoch": 0.7775236850702385, "grad_norm": 0.2983749806880951, "learning_rate": 7.189522144562773e-05, "loss": 0.0102, "step": 3570 }, { "epoch": 0.7797016225634325, "grad_norm": 0.1546776443719864, "learning_rate": 7.173350371927369e-05, "loss": 0.0095, "step": 3580 }, { "epoch": 0.7818795600566264, "grad_norm": 0.26190662384033203, "learning_rate": 7.157150518706503e-05, "loss": 0.0108, "step": 3590 }, { "epoch": 0.7840574975498203, "grad_norm": 0.24352820217609406, "learning_rate": 7.140922794208976e-05, "loss": 0.0102, "step": 3600 }, { "epoch": 0.7862354350430143, "grad_norm": 0.22204279899597168, "learning_rate": 7.124667408103698e-05, "loss": 0.0102, "step": 3610 }, { "epoch": 0.7884133725362082, "grad_norm": 0.21124747395515442, "learning_rate": 7.108384570416984e-05, "loss": 0.0124, "step": 3620 }, { "epoch": 0.7905913100294022, "grad_norm": 0.2160303145647049, "learning_rate": 7.092074491529823e-05, "loss": 0.0092, "step": 3630 }, { "epoch": 0.7927692475225961, "grad_norm": 0.2168467938899994, "learning_rate": 7.075737382175182e-05, "loss": 0.0111, "step": 3640 }, { "epoch": 0.7949471850157901, "grad_norm": 0.16817857325077057, "learning_rate": 7.059373453435269e-05, "loss": 0.0101, "step": 3650 }, { "epoch": 0.797125122508984, "grad_norm": 0.17954842746257782, "learning_rate": 7.042982916738807e-05, "loss": 0.0106, "step": 3660 }, { "epoch": 0.799303060002178, "grad_norm": 0.15504342317581177, "learning_rate": 7.026565983858311e-05, "loss": 0.0088, "step": 3670 }, { "epoch": 0.8014809974953718, "grad_norm": 0.18250958621501923, "learning_rate": 7.010122866907341e-05, "loss": 0.0115, "step": 3680 }, { "epoch": 0.8036589349885658, "grad_norm": 0.16664089262485504, "learning_rate": 6.993653778337763e-05, "loss": 0.0106, "step": 3690 }, { "epoch": 0.8058368724817597, "grad_norm": 0.181007981300354, "learning_rate": 6.977158930937019e-05, "loss": 0.01, "step": 3700 }, { "epoch": 0.8080148099749537, "grad_norm": 0.21096596121788025, "learning_rate": 6.960638537825352e-05, "loss": 0.0097, "step": 3710 }, { "epoch": 0.8101927474681476, "grad_norm": 0.12322518974542618, "learning_rate": 6.944092812453072e-05, "loss": 0.0097, "step": 3720 }, { "epoch": 0.8123706849613416, "grad_norm": 0.2257552593946457, "learning_rate": 6.927521968597795e-05, "loss": 0.01, "step": 3730 }, { "epoch": 0.8145486224545355, "grad_norm": 0.2484917938709259, "learning_rate": 6.910926220361675e-05, "loss": 0.0088, "step": 3740 }, { "epoch": 0.8167265599477295, "grad_norm": 0.23486147820949554, "learning_rate": 6.894305782168638e-05, "loss": 0.0093, "step": 3750 }, { "epoch": 0.8189044974409234, "grad_norm": 0.1993435025215149, "learning_rate": 6.877660868761623e-05, "loss": 0.0104, "step": 3760 }, { "epoch": 0.8210824349341174, "grad_norm": 0.13605287671089172, "learning_rate": 6.86099169519979e-05, "loss": 0.0099, "step": 3770 }, { "epoch": 0.8232603724273113, "grad_norm": 0.23245613276958466, "learning_rate": 6.844298476855754e-05, "loss": 0.0114, "step": 3780 }, { "epoch": 0.8254383099205053, "grad_norm": 0.2080039381980896, "learning_rate": 6.827581429412801e-05, "loss": 0.0097, "step": 3790 }, { "epoch": 0.8276162474136992, "grad_norm": 0.1880807876586914, "learning_rate": 6.810840768862096e-05, "loss": 0.0099, "step": 3800 }, { "epoch": 0.8297941849068932, "grad_norm": 0.22683408856391907, "learning_rate": 6.794076711499893e-05, "loss": 0.0091, "step": 3810 }, { "epoch": 0.8319721224000871, "grad_norm": 0.13539546728134155, "learning_rate": 6.777289473924746e-05, "loss": 0.0094, "step": 3820 }, { "epoch": 0.8341500598932811, "grad_norm": 0.22389768064022064, "learning_rate": 6.760479273034705e-05, "loss": 0.0108, "step": 3830 }, { "epoch": 0.836327997386475, "grad_norm": 0.2073911428451538, "learning_rate": 6.743646326024517e-05, "loss": 0.0091, "step": 3840 }, { "epoch": 0.838505934879669, "grad_norm": 0.1815386712551117, "learning_rate": 6.72679085038282e-05, "loss": 0.0096, "step": 3850 }, { "epoch": 0.8406838723728629, "grad_norm": 0.2157953828573227, "learning_rate": 6.709913063889326e-05, "loss": 0.0094, "step": 3860 }, { "epoch": 0.8428618098660569, "grad_norm": 0.22398988902568817, "learning_rate": 6.693013184612015e-05, "loss": 0.0094, "step": 3870 }, { "epoch": 0.8450397473592508, "grad_norm": 0.1814090758562088, "learning_rate": 6.676091430904317e-05, "loss": 0.0097, "step": 3880 }, { "epoch": 0.8472176848524448, "grad_norm": 0.2258855104446411, "learning_rate": 6.659148021402286e-05, "loss": 0.0111, "step": 3890 }, { "epoch": 0.8493956223456387, "grad_norm": 0.17605619132518768, "learning_rate": 6.642183175021779e-05, "loss": 0.009, "step": 3900 }, { "epoch": 0.8515735598388326, "grad_norm": 0.1938215047121048, "learning_rate": 6.625197110955625e-05, "loss": 0.0089, "step": 3910 }, { "epoch": 0.8537514973320266, "grad_norm": 0.1933971643447876, "learning_rate": 6.608190048670797e-05, "loss": 0.0097, "step": 3920 }, { "epoch": 0.8559294348252205, "grad_norm": 0.17037704586982727, "learning_rate": 6.591162207905571e-05, "loss": 0.0096, "step": 3930 }, { "epoch": 0.8581073723184145, "grad_norm": 0.19444690644741058, "learning_rate": 6.574113808666689e-05, "loss": 0.0098, "step": 3940 }, { "epoch": 0.8602853098116084, "grad_norm": 0.19196923077106476, "learning_rate": 6.55704507122652e-05, "loss": 0.0108, "step": 3950 }, { "epoch": 0.8624632473048024, "grad_norm": 0.09434881061315536, "learning_rate": 6.53995621612021e-05, "loss": 0.0091, "step": 3960 }, { "epoch": 0.8646411847979963, "grad_norm": 0.1585153490304947, "learning_rate": 6.522847464142833e-05, "loss": 0.0108, "step": 3970 }, { "epoch": 0.8668191222911903, "grad_norm": 0.20595288276672363, "learning_rate": 6.505719036346539e-05, "loss": 0.0102, "step": 3980 }, { "epoch": 0.8689970597843842, "grad_norm": 0.2814238667488098, "learning_rate": 6.488571154037695e-05, "loss": 0.0117, "step": 3990 }, { "epoch": 0.8711749972775781, "grad_norm": 0.18257386982440948, "learning_rate": 6.471404038774033e-05, "loss": 0.0088, "step": 4000 }, { "epoch": 0.873352934770772, "grad_norm": 0.16891984641551971, "learning_rate": 6.454217912361781e-05, "loss": 0.0083, "step": 4010 }, { "epoch": 0.875530872263966, "grad_norm": 0.15388254821300507, "learning_rate": 6.437012996852799e-05, "loss": 0.0097, "step": 4020 }, { "epoch": 0.8777088097571599, "grad_norm": 0.21748176217079163, "learning_rate": 6.419789514541705e-05, "loss": 0.0097, "step": 4030 }, { "epoch": 0.8798867472503539, "grad_norm": 0.19647681713104248, "learning_rate": 6.402547687963018e-05, "loss": 0.0094, "step": 4040 }, { "epoch": 0.8820646847435478, "grad_norm": 0.1627378761768341, "learning_rate": 6.385287739888263e-05, "loss": 0.0098, "step": 4050 }, { "epoch": 0.8842426222367418, "grad_norm": 0.1662755012512207, "learning_rate": 6.368009893323109e-05, "loss": 0.0092, "step": 4060 }, { "epoch": 0.8864205597299357, "grad_norm": 0.21434028446674347, "learning_rate": 6.350714371504476e-05, "loss": 0.0093, "step": 4070 }, { "epoch": 0.8885984972231297, "grad_norm": 0.20301590859889984, "learning_rate": 6.333401397897656e-05, "loss": 0.0098, "step": 4080 }, { "epoch": 0.8907764347163236, "grad_norm": 0.161471426486969, "learning_rate": 6.31607119619343e-05, "loss": 0.0092, "step": 4090 }, { "epoch": 0.8929543722095176, "grad_norm": 0.32318904995918274, "learning_rate": 6.298723990305166e-05, "loss": 0.0091, "step": 4100 }, { "epoch": 0.8951323097027115, "grad_norm": 0.20659998059272766, "learning_rate": 6.281360004365938e-05, "loss": 0.0093, "step": 4110 }, { "epoch": 0.8973102471959055, "grad_norm": 0.1987421065568924, "learning_rate": 6.263979462725627e-05, "loss": 0.0095, "step": 4120 }, { "epoch": 0.8994881846890994, "grad_norm": 0.16830500960350037, "learning_rate": 6.246582589948015e-05, "loss": 0.009, "step": 4130 }, { "epoch": 0.9016661221822934, "grad_norm": 0.16572082042694092, "learning_rate": 6.229169610807891e-05, "loss": 0.0086, "step": 4140 }, { "epoch": 0.9038440596754873, "grad_norm": 0.2040412425994873, "learning_rate": 6.21174075028815e-05, "loss": 0.0092, "step": 4150 }, { "epoch": 0.9060219971686813, "grad_norm": 0.25126901268959045, "learning_rate": 6.194296233576873e-05, "loss": 0.0091, "step": 4160 }, { "epoch": 0.9081999346618752, "grad_norm": 0.14921554923057556, "learning_rate": 6.176836286064428e-05, "loss": 0.0088, "step": 4170 }, { "epoch": 0.9103778721550692, "grad_norm": 0.25366389751434326, "learning_rate": 6.159361133340559e-05, "loss": 0.0099, "step": 4180 }, { "epoch": 0.9125558096482631, "grad_norm": 0.12061012536287308, "learning_rate": 6.141871001191463e-05, "loss": 0.0081, "step": 4190 }, { "epoch": 0.914733747141457, "grad_norm": 0.16168254613876343, "learning_rate": 6.124366115596881e-05, "loss": 0.0088, "step": 4200 }, { "epoch": 0.916911684634651, "grad_norm": 0.15729384124279022, "learning_rate": 6.106846702727172e-05, "loss": 0.0092, "step": 4210 }, { "epoch": 0.919089622127845, "grad_norm": 0.32289597392082214, "learning_rate": 6.0893129889403955e-05, "loss": 0.0087, "step": 4220 }, { "epoch": 0.9212675596210389, "grad_norm": 0.20793640613555908, "learning_rate": 6.0717652007793826e-05, "loss": 0.0105, "step": 4230 }, { "epoch": 0.9234454971142328, "grad_norm": 0.212624192237854, "learning_rate": 6.0542035649688114e-05, "loss": 0.0088, "step": 4240 }, { "epoch": 0.9256234346074268, "grad_norm": 0.22177855670452118, "learning_rate": 6.03662830841228e-05, "loss": 0.0102, "step": 4250 }, { "epoch": 0.9278013721006207, "grad_norm": 0.2538255453109741, "learning_rate": 6.019039658189369e-05, "loss": 0.0084, "step": 4260 }, { "epoch": 0.9299793095938147, "grad_norm": 0.20407214760780334, "learning_rate": 6.0014378415527116e-05, "loss": 0.009, "step": 4270 }, { "epoch": 0.9321572470870086, "grad_norm": 0.1658613681793213, "learning_rate": 5.9838230859250586e-05, "loss": 0.0085, "step": 4280 }, { "epoch": 0.9343351845802026, "grad_norm": 0.17914806306362152, "learning_rate": 5.966195618896333e-05, "loss": 0.0105, "step": 4290 }, { "epoch": 0.9365131220733965, "grad_norm": 0.15912717580795288, "learning_rate": 5.948555668220702e-05, "loss": 0.0085, "step": 4300 }, { "epoch": 0.9386910595665905, "grad_norm": 0.21457307040691376, "learning_rate": 5.9309034618136195e-05, "loss": 0.0089, "step": 4310 }, { "epoch": 0.9408689970597843, "grad_norm": 0.2808639407157898, "learning_rate": 5.913239227748892e-05, "loss": 0.0102, "step": 4320 }, { "epoch": 0.9430469345529783, "grad_norm": 0.2178078591823578, "learning_rate": 5.895563194255729e-05, "loss": 0.0093, "step": 4330 }, { "epoch": 0.9452248720461722, "grad_norm": 0.14186891913414001, "learning_rate": 5.877875589715789e-05, "loss": 0.0093, "step": 4340 }, { "epoch": 0.9474028095393662, "grad_norm": 0.1486235111951828, "learning_rate": 5.860176642660237e-05, "loss": 0.0077, "step": 4350 }, { "epoch": 0.9495807470325601, "grad_norm": 0.30821213126182556, "learning_rate": 5.842466581766789e-05, "loss": 0.0099, "step": 4360 }, { "epoch": 0.9517586845257541, "grad_norm": 0.20526987314224243, "learning_rate": 5.824745635856753e-05, "loss": 0.0086, "step": 4370 }, { "epoch": 0.953936622018948, "grad_norm": 0.13682414591312408, "learning_rate": 5.80701403389208e-05, "loss": 0.0087, "step": 4380 }, { "epoch": 0.956114559512142, "grad_norm": 0.27883806824684143, "learning_rate": 5.789272004972396e-05, "loss": 0.0106, "step": 4390 }, { "epoch": 0.9582924970053359, "grad_norm": 0.12711714208126068, "learning_rate": 5.7715197783320516e-05, "loss": 0.0089, "step": 4400 }, { "epoch": 0.9604704344985299, "grad_norm": 0.11983494460582733, "learning_rate": 5.753757583337158e-05, "loss": 0.0088, "step": 4410 }, { "epoch": 0.9626483719917238, "grad_norm": 0.13640055060386658, "learning_rate": 5.735985649482616e-05, "loss": 0.0097, "step": 4420 }, { "epoch": 0.9648263094849178, "grad_norm": 0.2632695436477661, "learning_rate": 5.718204206389162e-05, "loss": 0.0092, "step": 4430 }, { "epoch": 0.9670042469781117, "grad_norm": 0.17099197208881378, "learning_rate": 5.70041348380039e-05, "loss": 0.0091, "step": 4440 }, { "epoch": 0.9691821844713057, "grad_norm": 0.239213764667511, "learning_rate": 5.682613711579794e-05, "loss": 0.0102, "step": 4450 }, { "epoch": 0.9713601219644996, "grad_norm": 0.16065600514411926, "learning_rate": 5.664805119707791e-05, "loss": 0.0093, "step": 4460 }, { "epoch": 0.9735380594576936, "grad_norm": 0.1607472151517868, "learning_rate": 5.646987938278753e-05, "loss": 0.0103, "step": 4470 }, { "epoch": 0.9757159969508875, "grad_norm": 0.13043415546417236, "learning_rate": 5.629162397498029e-05, "loss": 0.0097, "step": 4480 }, { "epoch": 0.9778939344440815, "grad_norm": 0.18604792654514313, "learning_rate": 5.6113287276789775e-05, "loss": 0.0091, "step": 4490 }, { "epoch": 0.9800718719372754, "grad_norm": 0.1953439563512802, "learning_rate": 5.593487159239986e-05, "loss": 0.0102, "step": 4500 }, { "epoch": 0.9822498094304694, "grad_norm": 0.1606867015361786, "learning_rate": 5.575637922701499e-05, "loss": 0.0097, "step": 4510 }, { "epoch": 0.9844277469236633, "grad_norm": 0.2450137883424759, "learning_rate": 5.5577812486830284e-05, "loss": 0.0091, "step": 4520 }, { "epoch": 0.9866056844168573, "grad_norm": 0.18296198546886444, "learning_rate": 5.539917367900188e-05, "loss": 0.008, "step": 4530 }, { "epoch": 0.9887836219100512, "grad_norm": 0.1558394730091095, "learning_rate": 5.522046511161706e-05, "loss": 0.009, "step": 4540 }, { "epoch": 0.9909615594032452, "grad_norm": 0.1733635663986206, "learning_rate": 5.504168909366438e-05, "loss": 0.009, "step": 4550 }, { "epoch": 0.9931394968964391, "grad_norm": 0.1972990334033966, "learning_rate": 5.486284793500393e-05, "loss": 0.009, "step": 4560 }, { "epoch": 0.995317434389633, "grad_norm": 0.13507431745529175, "learning_rate": 5.468394394633742e-05, "loss": 0.008, "step": 4570 }, { "epoch": 0.997495371882827, "grad_norm": 0.15703797340393066, "learning_rate": 5.4504979439178374e-05, "loss": 0.009, "step": 4580 }, { "epoch": 0.999673309376021, "grad_norm": 0.18476291000843048, "learning_rate": 5.43259567258222e-05, "loss": 0.0082, "step": 4590 }, { "epoch": 1.0017423499945552, "grad_norm": 0.16002224385738373, "learning_rate": 5.41468781193164e-05, "loss": 0.0091, "step": 4600 }, { "epoch": 1.003920287487749, "grad_norm": 0.18123571574687958, "learning_rate": 5.3967745933430604e-05, "loss": 0.0091, "step": 4610 }, { "epoch": 1.0060982249809431, "grad_norm": 0.2654893696308136, "learning_rate": 5.378856248262673e-05, "loss": 0.0078, "step": 4620 }, { "epoch": 1.008276162474137, "grad_norm": 0.17253746092319489, "learning_rate": 5.360933008202905e-05, "loss": 0.0101, "step": 4630 }, { "epoch": 1.010454099967331, "grad_norm": 0.19651098549365997, "learning_rate": 5.343005104739429e-05, "loss": 0.0103, "step": 4640 }, { "epoch": 1.0126320374605249, "grad_norm": 0.1622696965932846, "learning_rate": 5.325072769508172e-05, "loss": 0.0085, "step": 4650 }, { "epoch": 1.0148099749537187, "grad_norm": 0.0935782715678215, "learning_rate": 5.307136234202318e-05, "loss": 0.0081, "step": 4660 }, { "epoch": 1.0169879124469128, "grad_norm": 0.10772078484296799, "learning_rate": 5.2891957305693205e-05, "loss": 0.0102, "step": 4670 }, { "epoch": 1.0191658499401066, "grad_norm": 0.2894693613052368, "learning_rate": 5.271251490407905e-05, "loss": 0.0088, "step": 4680 }, { "epoch": 1.0213437874333007, "grad_norm": 0.16669273376464844, "learning_rate": 5.2533037455650736e-05, "loss": 0.0088, "step": 4690 }, { "epoch": 1.0235217249264945, "grad_norm": 0.25250449776649475, "learning_rate": 5.235352727933113e-05, "loss": 0.0091, "step": 4700 }, { "epoch": 1.0256996624196886, "grad_norm": 0.2245578169822693, "learning_rate": 5.2173986694465895e-05, "loss": 0.0082, "step": 4710 }, { "epoch": 1.0278775999128824, "grad_norm": 0.170179545879364, "learning_rate": 5.199441802079364e-05, "loss": 0.0079, "step": 4720 }, { "epoch": 1.0300555374060765, "grad_norm": 0.11900405585765839, "learning_rate": 5.18148235784159e-05, "loss": 0.0092, "step": 4730 }, { "epoch": 1.0322334748992703, "grad_norm": 0.16673199832439423, "learning_rate": 5.163520568776712e-05, "loss": 0.0085, "step": 4740 }, { "epoch": 1.0344114123924644, "grad_norm": 0.18400241434574127, "learning_rate": 5.14555666695847e-05, "loss": 0.008, "step": 4750 }, { "epoch": 1.0365893498856582, "grad_norm": 0.206729456782341, "learning_rate": 5.127590884487906e-05, "loss": 0.0091, "step": 4760 }, { "epoch": 1.0387672873788523, "grad_norm": 0.19932995736598969, "learning_rate": 5.109623453490356e-05, "loss": 0.0084, "step": 4770 }, { "epoch": 1.040945224872046, "grad_norm": 0.19285811483860016, "learning_rate": 5.091654606112461e-05, "loss": 0.0079, "step": 4780 }, { "epoch": 1.0431231623652402, "grad_norm": 0.2269679605960846, "learning_rate": 5.0736845745191555e-05, "loss": 0.0086, "step": 4790 }, { "epoch": 1.045301099858434, "grad_norm": 0.19612117111682892, "learning_rate": 5.055713590890679e-05, "loss": 0.0089, "step": 4800 }, { "epoch": 1.047479037351628, "grad_norm": 0.2492666393518448, "learning_rate": 5.0377418874195714e-05, "loss": 0.0077, "step": 4810 }, { "epoch": 1.049656974844822, "grad_norm": 0.15281422436237335, "learning_rate": 5.019769696307674e-05, "loss": 0.0081, "step": 4820 }, { "epoch": 1.051834912338016, "grad_norm": 0.16830147802829742, "learning_rate": 5.001797249763124e-05, "loss": 0.0078, "step": 4830 }, { "epoch": 1.0540128498312098, "grad_norm": 0.1108197569847107, "learning_rate": 4.983824779997363e-05, "loss": 0.0077, "step": 4840 }, { "epoch": 1.0561907873244039, "grad_norm": 0.1993359923362732, "learning_rate": 4.965852519222134e-05, "loss": 0.0077, "step": 4850 }, { "epoch": 1.0583687248175977, "grad_norm": 0.166915163397789, "learning_rate": 4.9478806996464766e-05, "loss": 0.0076, "step": 4860 }, { "epoch": 1.0605466623107918, "grad_norm": 0.18394790589809418, "learning_rate": 4.9299095534737304e-05, "loss": 0.0089, "step": 4870 }, { "epoch": 1.0627245998039856, "grad_norm": 0.2215089648962021, "learning_rate": 4.911939312898535e-05, "loss": 0.0086, "step": 4880 }, { "epoch": 1.0649025372971797, "grad_norm": 0.2000982016324997, "learning_rate": 4.893970210103829e-05, "loss": 0.0088, "step": 4890 }, { "epoch": 1.0670804747903735, "grad_norm": 0.16642798483371735, "learning_rate": 4.87600247725785e-05, "loss": 0.0084, "step": 4900 }, { "epoch": 1.0692584122835675, "grad_norm": 0.13275150954723358, "learning_rate": 4.858036346511138e-05, "loss": 0.0071, "step": 4910 }, { "epoch": 1.0714363497767614, "grad_norm": 0.16688916087150574, "learning_rate": 4.840072049993527e-05, "loss": 0.0079, "step": 4920 }, { "epoch": 1.0736142872699554, "grad_norm": 0.10775057226419449, "learning_rate": 4.8221098198111614e-05, "loss": 0.009, "step": 4930 }, { "epoch": 1.0757922247631493, "grad_norm": 0.10888474434614182, "learning_rate": 4.804149888043477e-05, "loss": 0.0087, "step": 4940 }, { "epoch": 1.0779701622563433, "grad_norm": 0.2054320126771927, "learning_rate": 4.786192486740223e-05, "loss": 0.0088, "step": 4950 }, { "epoch": 1.0801480997495372, "grad_norm": 0.2288559526205063, "learning_rate": 4.7682378479184464e-05, "loss": 0.0088, "step": 4960 }, { "epoch": 1.0823260372427312, "grad_norm": 0.14645184576511383, "learning_rate": 4.750286203559505e-05, "loss": 0.0089, "step": 4970 }, { "epoch": 1.084503974735925, "grad_norm": 0.18956804275512695, "learning_rate": 4.732337785606069e-05, "loss": 0.0078, "step": 4980 }, { "epoch": 1.086681912229119, "grad_norm": 0.14879284799098969, "learning_rate": 4.714392825959117e-05, "loss": 0.0074, "step": 4990 }, { "epoch": 1.088859849722313, "grad_norm": 0.2719704806804657, "learning_rate": 4.696451556474951e-05, "loss": 0.0084, "step": 5000 }, { "epoch": 1.0910377872155068, "grad_norm": 0.2144235074520111, "learning_rate": 4.6785142089621875e-05, "loss": 0.007, "step": 5010 }, { "epoch": 1.0932157247087009, "grad_norm": 0.24391882121562958, "learning_rate": 4.660581015178777e-05, "loss": 0.0081, "step": 5020 }, { "epoch": 1.0953936622018947, "grad_norm": 0.14946295320987701, "learning_rate": 4.642652206828994e-05, "loss": 0.0074, "step": 5030 }, { "epoch": 1.0975715996950888, "grad_norm": 0.191061332821846, "learning_rate": 4.624728015560461e-05, "loss": 0.0092, "step": 5040 }, { "epoch": 1.0997495371882826, "grad_norm": 0.17625565826892853, "learning_rate": 4.606808672961137e-05, "loss": 0.0097, "step": 5050 }, { "epoch": 1.1019274746814767, "grad_norm": 0.18322786688804626, "learning_rate": 4.5888944105563386e-05, "loss": 0.0078, "step": 5060 }, { "epoch": 1.1041054121746705, "grad_norm": 0.21730530261993408, "learning_rate": 4.570985459805742e-05, "loss": 0.0074, "step": 5070 }, { "epoch": 1.1062833496678646, "grad_norm": 0.10806269198656082, "learning_rate": 4.5530820521004e-05, "loss": 0.0069, "step": 5080 }, { "epoch": 1.1084612871610584, "grad_norm": 0.22006118297576904, "learning_rate": 4.535184418759738e-05, "loss": 0.0093, "step": 5090 }, { "epoch": 1.1106392246542525, "grad_norm": 0.10638593882322311, "learning_rate": 4.517292791028585e-05, "loss": 0.007, "step": 5100 }, { "epoch": 1.1128171621474463, "grad_norm": 0.21001309156417847, "learning_rate": 4.499407400074163e-05, "loss": 0.0079, "step": 5110 }, { "epoch": 1.1149950996406404, "grad_norm": 0.1483108252286911, "learning_rate": 4.481528476983122e-05, "loss": 0.0085, "step": 5120 }, { "epoch": 1.1171730371338342, "grad_norm": 0.11377395689487457, "learning_rate": 4.463656252758542e-05, "loss": 0.008, "step": 5130 }, { "epoch": 1.1193509746270283, "grad_norm": 0.19740240275859833, "learning_rate": 4.4457909583169483e-05, "loss": 0.0079, "step": 5140 }, { "epoch": 1.121528912120222, "grad_norm": 0.15365475416183472, "learning_rate": 4.4279328244853343e-05, "loss": 0.0078, "step": 5150 }, { "epoch": 1.1237068496134162, "grad_norm": 0.15118981897830963, "learning_rate": 4.410082081998169e-05, "loss": 0.0069, "step": 5160 }, { "epoch": 1.12588478710661, "grad_norm": 0.12912306189537048, "learning_rate": 4.39223896149443e-05, "loss": 0.0075, "step": 5170 }, { "epoch": 1.128062724599804, "grad_norm": 0.20978759229183197, "learning_rate": 4.3744036935146104e-05, "loss": 0.008, "step": 5180 }, { "epoch": 1.130240662092998, "grad_norm": 0.2227969616651535, "learning_rate": 4.3565765084977496e-05, "loss": 0.007, "step": 5190 }, { "epoch": 1.132418599586192, "grad_norm": 0.15055429935455322, "learning_rate": 4.3387576367784456e-05, "loss": 0.0085, "step": 5200 }, { "epoch": 1.1345965370793858, "grad_norm": 0.16581185162067413, "learning_rate": 4.320947308583893e-05, "loss": 0.0076, "step": 5210 }, { "epoch": 1.1367744745725799, "grad_norm": 0.15262490510940552, "learning_rate": 4.303145754030893e-05, "loss": 0.0073, "step": 5220 }, { "epoch": 1.1389524120657737, "grad_norm": 0.21703152358531952, "learning_rate": 4.285353203122893e-05, "loss": 0.0076, "step": 5230 }, { "epoch": 1.1411303495589677, "grad_norm": 0.10583783686161041, "learning_rate": 4.267569885747008e-05, "loss": 0.007, "step": 5240 }, { "epoch": 1.1433082870521616, "grad_norm": 0.17903271317481995, "learning_rate": 4.2497960316710514e-05, "loss": 0.0087, "step": 5250 }, { "epoch": 1.1454862245453556, "grad_norm": 0.10162434726953506, "learning_rate": 4.232031870540565e-05, "loss": 0.0074, "step": 5260 }, { "epoch": 1.1476641620385495, "grad_norm": 0.158236563205719, "learning_rate": 4.214277631875857e-05, "loss": 0.0085, "step": 5270 }, { "epoch": 1.1498420995317433, "grad_norm": 0.21938243508338928, "learning_rate": 4.196533545069036e-05, "loss": 0.0071, "step": 5280 }, { "epoch": 1.1520200370249374, "grad_norm": 0.1561320573091507, "learning_rate": 4.1787998393810345e-05, "loss": 0.0077, "step": 5290 }, { "epoch": 1.1541979745181314, "grad_norm": 0.17417271435260773, "learning_rate": 4.161076743938668e-05, "loss": 0.0072, "step": 5300 }, { "epoch": 1.1563759120113253, "grad_norm": 0.1613398641347885, "learning_rate": 4.1433644877316546e-05, "loss": 0.0073, "step": 5310 }, { "epoch": 1.1585538495045191, "grad_norm": 0.12310460209846497, "learning_rate": 4.125663299609671e-05, "loss": 0.0086, "step": 5320 }, { "epoch": 1.1607317869977132, "grad_norm": 0.16984106600284576, "learning_rate": 4.107973408279382e-05, "loss": 0.0072, "step": 5330 }, { "epoch": 1.162909724490907, "grad_norm": 0.1483551561832428, "learning_rate": 4.090295042301504e-05, "loss": 0.0077, "step": 5340 }, { "epoch": 1.165087661984101, "grad_norm": 0.18259865045547485, "learning_rate": 4.07262843008783e-05, "loss": 0.0078, "step": 5350 }, { "epoch": 1.167265599477295, "grad_norm": 0.2085978239774704, "learning_rate": 4.0549737998983e-05, "loss": 0.0079, "step": 5360 }, { "epoch": 1.169443536970489, "grad_norm": 0.15245895087718964, "learning_rate": 4.037331379838031e-05, "loss": 0.0084, "step": 5370 }, { "epoch": 1.1716214744636828, "grad_norm": 0.19410376250743866, "learning_rate": 4.0197013978543896e-05, "loss": 0.0081, "step": 5380 }, { "epoch": 1.1737994119568769, "grad_norm": 0.2189418375492096, "learning_rate": 4.0020840817340286e-05, "loss": 0.0064, "step": 5390 }, { "epoch": 1.1759773494500707, "grad_norm": 0.19394533336162567, "learning_rate": 3.984479659099962e-05, "loss": 0.0077, "step": 5400 }, { "epoch": 1.1781552869432648, "grad_norm": 0.15754857659339905, "learning_rate": 3.966888357408605e-05, "loss": 0.0075, "step": 5410 }, { "epoch": 1.1803332244364586, "grad_norm": 0.2520867884159088, "learning_rate": 3.949310403946849e-05, "loss": 0.0069, "step": 5420 }, { "epoch": 1.1825111619296527, "grad_norm": 0.13636243343353271, "learning_rate": 3.931746025829126e-05, "loss": 0.0074, "step": 5430 }, { "epoch": 1.1846890994228465, "grad_norm": 0.10027901828289032, "learning_rate": 3.9141954499944575e-05, "loss": 0.0071, "step": 5440 }, { "epoch": 1.1868670369160406, "grad_norm": 0.1434406191110611, "learning_rate": 3.896658903203547e-05, "loss": 0.0075, "step": 5450 }, { "epoch": 1.1890449744092344, "grad_norm": 0.18343040347099304, "learning_rate": 3.8791366120358235e-05, "loss": 0.0078, "step": 5460 }, { "epoch": 1.1912229119024285, "grad_norm": 0.19430644810199738, "learning_rate": 3.86162880288654e-05, "loss": 0.0072, "step": 5470 }, { "epoch": 1.1934008493956223, "grad_norm": 0.13038694858551025, "learning_rate": 3.8441357019638266e-05, "loss": 0.0071, "step": 5480 }, { "epoch": 1.1955787868888164, "grad_norm": 0.18733017146587372, "learning_rate": 3.826657535285781e-05, "loss": 0.0082, "step": 5490 }, { "epoch": 1.1977567243820102, "grad_norm": 0.20870471000671387, "learning_rate": 3.8091945286775435e-05, "loss": 0.0081, "step": 5500 }, { "epoch": 1.1999346618752043, "grad_norm": 0.18687936663627625, "learning_rate": 3.7917469077683824e-05, "loss": 0.0072, "step": 5510 }, { "epoch": 1.202112599368398, "grad_norm": 0.14487966895103455, "learning_rate": 3.77431489798877e-05, "loss": 0.0065, "step": 5520 }, { "epoch": 1.2042905368615922, "grad_norm": 0.12217055261135101, "learning_rate": 3.756898724567485e-05, "loss": 0.0077, "step": 5530 }, { "epoch": 1.206468474354786, "grad_norm": 0.12225670367479324, "learning_rate": 3.739498612528686e-05, "loss": 0.0065, "step": 5540 }, { "epoch": 1.20864641184798, "grad_norm": 0.15228663384914398, "learning_rate": 3.722114786689018e-05, "loss": 0.0075, "step": 5550 }, { "epoch": 1.2108243493411739, "grad_norm": 0.1891603022813797, "learning_rate": 3.704747471654696e-05, "loss": 0.0073, "step": 5560 }, { "epoch": 1.213002286834368, "grad_norm": 0.17855939269065857, "learning_rate": 3.687396891818615e-05, "loss": 0.0078, "step": 5570 }, { "epoch": 1.2151802243275618, "grad_norm": 0.14619100093841553, "learning_rate": 3.670063271357438e-05, "loss": 0.0059, "step": 5580 }, { "epoch": 1.2173581618207558, "grad_norm": 0.16030114889144897, "learning_rate": 3.65274683422871e-05, "loss": 0.0075, "step": 5590 }, { "epoch": 1.2195360993139497, "grad_norm": 0.1406458169221878, "learning_rate": 3.63544780416796e-05, "loss": 0.0078, "step": 5600 }, { "epoch": 1.2217140368071435, "grad_norm": 0.1633547991514206, "learning_rate": 3.618166404685805e-05, "loss": 0.0071, "step": 5610 }, { "epoch": 1.2238919743003376, "grad_norm": 0.16294194757938385, "learning_rate": 3.600902859065076e-05, "loss": 0.0083, "step": 5620 }, { "epoch": 1.2260699117935316, "grad_norm": 0.3183712959289551, "learning_rate": 3.583657390357919e-05, "loss": 0.0067, "step": 5630 }, { "epoch": 1.2282478492867255, "grad_norm": 0.16475562751293182, "learning_rate": 3.5664302213829214e-05, "loss": 0.0063, "step": 5640 }, { "epoch": 1.2304257867799193, "grad_norm": 0.14080336689949036, "learning_rate": 3.549221574722228e-05, "loss": 0.0063, "step": 5650 }, { "epoch": 1.2326037242731134, "grad_norm": 0.14986054599285126, "learning_rate": 3.532031672718672e-05, "loss": 0.0064, "step": 5660 }, { "epoch": 1.2347816617663072, "grad_norm": 0.22976188361644745, "learning_rate": 3.5148607374728924e-05, "loss": 0.007, "step": 5670 }, { "epoch": 1.2369595992595013, "grad_norm": 0.15098370611667633, "learning_rate": 3.4977089908404744e-05, "loss": 0.0075, "step": 5680 }, { "epoch": 1.2391375367526951, "grad_norm": 0.07116223126649857, "learning_rate": 3.480576654429075e-05, "loss": 0.0074, "step": 5690 }, { "epoch": 1.2413154742458892, "grad_norm": 0.2963467538356781, "learning_rate": 3.463463949595567e-05, "loss": 0.0075, "step": 5700 }, { "epoch": 1.243493411739083, "grad_norm": 0.2506103515625, "learning_rate": 3.44637109744317e-05, "loss": 0.0077, "step": 5710 }, { "epoch": 1.245671349232277, "grad_norm": 0.19080466032028198, "learning_rate": 3.4292983188186034e-05, "loss": 0.0065, "step": 5720 }, { "epoch": 1.247849286725471, "grad_norm": 0.1616036295890808, "learning_rate": 3.412245834309224e-05, "loss": 0.0074, "step": 5730 }, { "epoch": 1.250027224218665, "grad_norm": 0.2033299058675766, "learning_rate": 3.395213864240185e-05, "loss": 0.0072, "step": 5740 }, { "epoch": 1.2522051617118588, "grad_norm": 0.24676287174224854, "learning_rate": 3.378202628671584e-05, "loss": 0.0071, "step": 5750 }, { "epoch": 1.2543830992050529, "grad_norm": 0.16114264726638794, "learning_rate": 3.3612123473956146e-05, "loss": 0.0075, "step": 5760 }, { "epoch": 1.2565610366982467, "grad_norm": 0.18414832651615143, "learning_rate": 3.344243239933739e-05, "loss": 0.0091, "step": 5770 }, { "epoch": 1.2587389741914408, "grad_norm": 0.2218603491783142, "learning_rate": 3.327295525533839e-05, "loss": 0.0075, "step": 5780 }, { "epoch": 1.2609169116846346, "grad_norm": 0.1624598205089569, "learning_rate": 3.310369423167396e-05, "loss": 0.0081, "step": 5790 }, { "epoch": 1.2630948491778287, "grad_norm": 0.1396883726119995, "learning_rate": 3.293465151526649e-05, "loss": 0.0074, "step": 5800 }, { "epoch": 1.2652727866710225, "grad_norm": 0.12150052189826965, "learning_rate": 3.2765829290217784e-05, "loss": 0.0087, "step": 5810 }, { "epoch": 1.2674507241642166, "grad_norm": 0.14573368430137634, "learning_rate": 3.2597229737780774e-05, "loss": 0.008, "step": 5820 }, { "epoch": 1.2696286616574104, "grad_norm": 0.1662302017211914, "learning_rate": 3.242885503633144e-05, "loss": 0.0069, "step": 5830 }, { "epoch": 1.2718065991506045, "grad_norm": 0.1439054310321808, "learning_rate": 3.226070736134049e-05, "loss": 0.0068, "step": 5840 }, { "epoch": 1.2739845366437983, "grad_norm": 0.1672573834657669, "learning_rate": 3.209278888534545e-05, "loss": 0.0075, "step": 5850 }, { "epoch": 1.2761624741369921, "grad_norm": 0.13138443231582642, "learning_rate": 3.192510177792245e-05, "loss": 0.007, "step": 5860 }, { "epoch": 1.2783404116301862, "grad_norm": 0.1558171808719635, "learning_rate": 3.175764820565826e-05, "loss": 0.0071, "step": 5870 }, { "epoch": 1.2805183491233803, "grad_norm": 0.16756413877010345, "learning_rate": 3.1590430332122246e-05, "loss": 0.0069, "step": 5880 }, { "epoch": 1.282696286616574, "grad_norm": 0.14233723282814026, "learning_rate": 3.14234503178385e-05, "loss": 0.0063, "step": 5890 }, { "epoch": 1.284874224109768, "grad_norm": 0.13999633491039276, "learning_rate": 3.125671032025787e-05, "loss": 0.0071, "step": 5900 }, { "epoch": 1.287052161602962, "grad_norm": 0.1227533295750618, "learning_rate": 3.109021249373003e-05, "loss": 0.0074, "step": 5910 }, { "epoch": 1.289230099096156, "grad_norm": 0.20161081850528717, "learning_rate": 3.092395898947579e-05, "loss": 0.0079, "step": 5920 }, { "epoch": 1.2914080365893499, "grad_norm": 0.11331768333911896, "learning_rate": 3.0757951955559123e-05, "loss": 0.0066, "step": 5930 }, { "epoch": 1.2935859740825437, "grad_norm": 0.1756715625524521, "learning_rate": 3.059219353685957e-05, "loss": 0.0083, "step": 5940 }, { "epoch": 1.2957639115757378, "grad_norm": 0.19707661867141724, "learning_rate": 3.0426685875044446e-05, "loss": 0.0067, "step": 5950 }, { "epoch": 1.2979418490689318, "grad_norm": 0.14970429241657257, "learning_rate": 3.0261431108541173e-05, "loss": 0.0075, "step": 5960 }, { "epoch": 1.3001197865621257, "grad_norm": 0.15001006424427032, "learning_rate": 3.0096431372509647e-05, "loss": 0.0093, "step": 5970 }, { "epoch": 1.3022977240553195, "grad_norm": 0.14086420834064484, "learning_rate": 2.9931688798814716e-05, "loss": 0.0079, "step": 5980 }, { "epoch": 1.3044756615485136, "grad_norm": 0.14562717080116272, "learning_rate": 2.9767205515998518e-05, "loss": 0.0071, "step": 5990 }, { "epoch": 1.3066535990417076, "grad_norm": 0.12243077903985977, "learning_rate": 2.9602983649253113e-05, "loss": 0.0074, "step": 6000 }, { "epoch": 1.3088315365349015, "grad_norm": 0.17876362800598145, "learning_rate": 2.9439025320392888e-05, "loss": 0.0083, "step": 6010 }, { "epoch": 1.3110094740280953, "grad_norm": 0.17103134095668793, "learning_rate": 2.927533264782728e-05, "loss": 0.0067, "step": 6020 }, { "epoch": 1.3131874115212894, "grad_norm": 0.16949225962162018, "learning_rate": 2.9111907746533275e-05, "loss": 0.0071, "step": 6030 }, { "epoch": 1.3153653490144832, "grad_norm": 0.14292246103286743, "learning_rate": 2.894875272802817e-05, "loss": 0.0065, "step": 6040 }, { "epoch": 1.3175432865076773, "grad_norm": 0.13834917545318604, "learning_rate": 2.878586970034232e-05, "loss": 0.0067, "step": 6050 }, { "epoch": 1.319721224000871, "grad_norm": 0.09846922755241394, "learning_rate": 2.8623260767991737e-05, "loss": 0.0066, "step": 6060 }, { "epoch": 1.3218991614940652, "grad_norm": 0.0991157814860344, "learning_rate": 2.8460928031951083e-05, "loss": 0.0065, "step": 6070 }, { "epoch": 1.324077098987259, "grad_norm": 0.10126027464866638, "learning_rate": 2.829887358962643e-05, "loss": 0.0069, "step": 6080 }, { "epoch": 1.326255036480453, "grad_norm": 0.14820335805416107, "learning_rate": 2.8137099534828176e-05, "loss": 0.0064, "step": 6090 }, { "epoch": 1.328432973973647, "grad_norm": 0.14515072107315063, "learning_rate": 2.797560795774401e-05, "loss": 0.0074, "step": 6100 }, { "epoch": 1.330610911466841, "grad_norm": 0.12842248380184174, "learning_rate": 2.7814400944911873e-05, "loss": 0.0067, "step": 6110 }, { "epoch": 1.3327888489600348, "grad_norm": 0.12815093994140625, "learning_rate": 2.7653480579193036e-05, "loss": 0.0067, "step": 6120 }, { "epoch": 1.3349667864532289, "grad_norm": 0.22363591194152832, "learning_rate": 2.7492848939745162e-05, "loss": 0.0061, "step": 6130 }, { "epoch": 1.3371447239464227, "grad_norm": 0.2403387576341629, "learning_rate": 2.7332508101995447e-05, "loss": 0.0071, "step": 6140 }, { "epoch": 1.3393226614396168, "grad_norm": 0.17800240218639374, "learning_rate": 2.717246013761382e-05, "loss": 0.0068, "step": 6150 }, { "epoch": 1.3415005989328106, "grad_norm": 0.18616099655628204, "learning_rate": 2.7012707114486148e-05, "loss": 0.0068, "step": 6160 }, { "epoch": 1.3436785364260047, "grad_norm": 0.1361059695482254, "learning_rate": 2.685325109668756e-05, "loss": 0.0068, "step": 6170 }, { "epoch": 1.3458564739191985, "grad_norm": 0.237016499042511, "learning_rate": 2.669409414445574e-05, "loss": 0.0068, "step": 6180 }, { "epoch": 1.3480344114123923, "grad_norm": 0.1593666970729828, "learning_rate": 2.65352383141643e-05, "loss": 0.0058, "step": 6190 }, { "epoch": 1.3502123489055864, "grad_norm": 0.12591980397701263, "learning_rate": 2.6376685658296273e-05, "loss": 0.0066, "step": 6200 }, { "epoch": 1.3523902863987805, "grad_norm": 0.201295867562294, "learning_rate": 2.621843822541753e-05, "loss": 0.0057, "step": 6210 }, { "epoch": 1.3545682238919743, "grad_norm": 0.20959442853927612, "learning_rate": 2.606049806015033e-05, "loss": 0.0066, "step": 6220 }, { "epoch": 1.3567461613851681, "grad_norm": 0.12013950198888779, "learning_rate": 2.590286720314692e-05, "loss": 0.0058, "step": 6230 }, { "epoch": 1.3589240988783622, "grad_norm": 0.10039467364549637, "learning_rate": 2.5745547691063144e-05, "loss": 0.0066, "step": 6240 }, { "epoch": 1.3611020363715562, "grad_norm": 0.09179899841547012, "learning_rate": 2.5588541556532152e-05, "loss": 0.0071, "step": 6250 }, { "epoch": 1.36327997386475, "grad_norm": 0.19506333768367767, "learning_rate": 2.5431850828138125e-05, "loss": 0.0069, "step": 6260 }, { "epoch": 1.365457911357944, "grad_norm": 0.14047041535377502, "learning_rate": 2.527547753039007e-05, "loss": 0.0061, "step": 6270 }, { "epoch": 1.367635848851138, "grad_norm": 0.17929664254188538, "learning_rate": 2.511942368369566e-05, "loss": 0.0063, "step": 6280 }, { "epoch": 1.369813786344332, "grad_norm": 0.1176358312368393, "learning_rate": 2.4963691304335067e-05, "loss": 0.0068, "step": 6290 }, { "epoch": 1.3719917238375259, "grad_norm": 0.2112642526626587, "learning_rate": 2.48082824044351e-05, "loss": 0.0063, "step": 6300 }, { "epoch": 1.3741696613307197, "grad_norm": 0.1680610477924347, "learning_rate": 2.4653198991942938e-05, "loss": 0.0055, "step": 6310 }, { "epoch": 1.3763475988239138, "grad_norm": 0.1417052298784256, "learning_rate": 2.4498443070600464e-05, "loss": 0.0066, "step": 6320 }, { "epoch": 1.3785255363171078, "grad_norm": 0.15466494858264923, "learning_rate": 2.43440166399181e-05, "loss": 0.0061, "step": 6330 }, { "epoch": 1.3807034738103017, "grad_norm": 0.11344225704669952, "learning_rate": 2.4189921695149238e-05, "loss": 0.0062, "step": 6340 }, { "epoch": 1.3828814113034955, "grad_norm": 0.13439317047595978, "learning_rate": 2.4036160227264282e-05, "loss": 0.007, "step": 6350 }, { "epoch": 1.3850593487966896, "grad_norm": 0.20675961673259735, "learning_rate": 2.388273422292492e-05, "loss": 0.0072, "step": 6360 }, { "epoch": 1.3872372862898834, "grad_norm": 0.14077794551849365, "learning_rate": 2.3729645664458638e-05, "loss": 0.0068, "step": 6370 }, { "epoch": 1.3894152237830775, "grad_norm": 0.1917889565229416, "learning_rate": 2.3576896529832844e-05, "loss": 0.0078, "step": 6380 }, { "epoch": 1.3915931612762713, "grad_norm": 0.15326690673828125, "learning_rate": 2.342448879262959e-05, "loss": 0.0068, "step": 6390 }, { "epoch": 1.3937710987694654, "grad_norm": 0.16057105362415314, "learning_rate": 2.3272424422019783e-05, "loss": 0.0068, "step": 6400 }, { "epoch": 1.3959490362626592, "grad_norm": 0.12924149632453918, "learning_rate": 2.3120705382738045e-05, "loss": 0.0061, "step": 6410 }, { "epoch": 1.3981269737558533, "grad_norm": 0.15427018702030182, "learning_rate": 2.2969333635057024e-05, "loss": 0.0065, "step": 6420 }, { "epoch": 1.400304911249047, "grad_norm": 0.14138518273830414, "learning_rate": 2.2818311134762364e-05, "loss": 0.0063, "step": 6430 }, { "epoch": 1.4024828487422412, "grad_norm": 0.18323691189289093, "learning_rate": 2.2667639833127156e-05, "loss": 0.0068, "step": 6440 }, { "epoch": 1.404660786235435, "grad_norm": 0.12226750701665878, "learning_rate": 2.2517321676886945e-05, "loss": 0.0067, "step": 6450 }, { "epoch": 1.406838723728629, "grad_norm": 0.1077219545841217, "learning_rate": 2.236735860821445e-05, "loss": 0.0055, "step": 6460 }, { "epoch": 1.409016661221823, "grad_norm": 0.17939791083335876, "learning_rate": 2.2217752564694543e-05, "loss": 0.0067, "step": 6470 }, { "epoch": 1.411194598715017, "grad_norm": 0.23838293552398682, "learning_rate": 2.2068505479299157e-05, "loss": 0.0061, "step": 6480 }, { "epoch": 1.4133725362082108, "grad_norm": 0.08796317130327225, "learning_rate": 2.191961928036233e-05, "loss": 0.0053, "step": 6490 }, { "epoch": 1.4155504737014049, "grad_norm": 0.245268315076828, "learning_rate": 2.177109589155537e-05, "loss": 0.0069, "step": 6500 }, { "epoch": 1.4177284111945987, "grad_norm": 0.15249638259410858, "learning_rate": 2.1622937231861822e-05, "loss": 0.0066, "step": 6510 }, { "epoch": 1.4199063486877925, "grad_norm": 0.11680476367473602, "learning_rate": 2.1475145215552843e-05, "loss": 0.0062, "step": 6520 }, { "epoch": 1.4220842861809866, "grad_norm": 0.14956925809383392, "learning_rate": 2.1327721752162404e-05, "loss": 0.007, "step": 6530 }, { "epoch": 1.4242622236741806, "grad_norm": 0.1575952172279358, "learning_rate": 2.1180668746462602e-05, "loss": 0.0056, "step": 6540 }, { "epoch": 1.4264401611673745, "grad_norm": 0.1346137970685959, "learning_rate": 2.103398809843908e-05, "loss": 0.0066, "step": 6550 }, { "epoch": 1.4286180986605683, "grad_norm": 0.13053518533706665, "learning_rate": 2.0887681703266453e-05, "loss": 0.0063, "step": 6560 }, { "epoch": 1.4307960361537624, "grad_norm": 0.12916165590286255, "learning_rate": 2.074175145128385e-05, "loss": 0.0056, "step": 6570 }, { "epoch": 1.4329739736469564, "grad_norm": 0.10324876755475998, "learning_rate": 2.059619922797047e-05, "loss": 0.0061, "step": 6580 }, { "epoch": 1.4351519111401503, "grad_norm": 0.11840296536684036, "learning_rate": 2.0451026913921217e-05, "loss": 0.0056, "step": 6590 }, { "epoch": 1.4373298486333441, "grad_norm": 0.1157921776175499, "learning_rate": 2.030623638482242e-05, "loss": 0.0058, "step": 6600 }, { "epoch": 1.4395077861265382, "grad_norm": 0.13884562253952026, "learning_rate": 2.016182951142757e-05, "loss": 0.0054, "step": 6610 }, { "epoch": 1.4416857236197322, "grad_norm": 0.1162991151213646, "learning_rate": 2.0017808159533203e-05, "loss": 0.0058, "step": 6620 }, { "epoch": 1.443863661112926, "grad_norm": 0.17985719442367554, "learning_rate": 1.9874174189954724e-05, "loss": 0.0061, "step": 6630 }, { "epoch": 1.44604159860612, "grad_norm": 0.1988246589899063, "learning_rate": 1.9730929458502418e-05, "loss": 0.0057, "step": 6640 }, { "epoch": 1.448219536099314, "grad_norm": 0.12605798244476318, "learning_rate": 1.958807581595744e-05, "loss": 0.0062, "step": 6650 }, { "epoch": 1.4503974735925078, "grad_norm": 0.16082799434661865, "learning_rate": 1.9445615108047923e-05, "loss": 0.0066, "step": 6660 }, { "epoch": 1.4525754110857019, "grad_norm": 0.1606266051530838, "learning_rate": 1.930354917542511e-05, "loss": 0.0064, "step": 6670 }, { "epoch": 1.4547533485788957, "grad_norm": 0.2217327058315277, "learning_rate": 1.9161879853639607e-05, "loss": 0.0069, "step": 6680 }, { "epoch": 1.4569312860720898, "grad_norm": 0.15943476557731628, "learning_rate": 1.9020608973117627e-05, "loss": 0.0058, "step": 6690 }, { "epoch": 1.4591092235652836, "grad_norm": 0.11603707075119019, "learning_rate": 1.887973835913735e-05, "loss": 0.0057, "step": 6700 }, { "epoch": 1.4612871610584777, "grad_norm": 0.14399226009845734, "learning_rate": 1.873926983180539e-05, "loss": 0.0059, "step": 6710 }, { "epoch": 1.4634650985516715, "grad_norm": 0.17077799141407013, "learning_rate": 1.8599205206033138e-05, "loss": 0.007, "step": 6720 }, { "epoch": 1.4656430360448656, "grad_norm": 0.14094921946525574, "learning_rate": 1.8459546291513545e-05, "loss": 0.0056, "step": 6730 }, { "epoch": 1.4678209735380594, "grad_norm": 0.12991905212402344, "learning_rate": 1.8320294892697478e-05, "loss": 0.0061, "step": 6740 }, { "epoch": 1.4699989110312535, "grad_norm": 0.1356325000524521, "learning_rate": 1.8181452808770637e-05, "loss": 0.0062, "step": 6750 }, { "epoch": 1.4721768485244473, "grad_norm": 0.1286916732788086, "learning_rate": 1.80430218336301e-05, "loss": 0.0055, "step": 6760 }, { "epoch": 1.4743547860176414, "grad_norm": 0.13185356557369232, "learning_rate": 1.7905003755861377e-05, "loss": 0.0062, "step": 6770 }, { "epoch": 1.4765327235108352, "grad_norm": 0.14180611073970795, "learning_rate": 1.776740035871502e-05, "loss": 0.0068, "step": 6780 }, { "epoch": 1.4787106610040293, "grad_norm": 0.10832282155752182, "learning_rate": 1.7630213420083884e-05, "loss": 0.0062, "step": 6790 }, { "epoch": 1.480888598497223, "grad_norm": 0.12900234758853912, "learning_rate": 1.749344471247993e-05, "loss": 0.0067, "step": 6800 }, { "epoch": 1.4830665359904172, "grad_norm": 0.12022525072097778, "learning_rate": 1.735709600301137e-05, "loss": 0.0055, "step": 6810 }, { "epoch": 1.485244473483611, "grad_norm": 0.218561053276062, "learning_rate": 1.722116905335997e-05, "loss": 0.0056, "step": 6820 }, { "epoch": 1.487422410976805, "grad_norm": 0.15504102408885956, "learning_rate": 1.7085665619758055e-05, "loss": 0.0063, "step": 6830 }, { "epoch": 1.489600348469999, "grad_norm": 0.14231227338314056, "learning_rate": 1.6950587452966093e-05, "loss": 0.0064, "step": 6840 }, { "epoch": 1.4917782859631927, "grad_norm": 0.14866498112678528, "learning_rate": 1.6815936298249785e-05, "loss": 0.0064, "step": 6850 }, { "epoch": 1.4939562234563868, "grad_norm": 0.15301434695720673, "learning_rate": 1.668171389535778e-05, "loss": 0.0059, "step": 6860 }, { "epoch": 1.4961341609495808, "grad_norm": 0.13548678159713745, "learning_rate": 1.6547921978498988e-05, "loss": 0.007, "step": 6870 }, { "epoch": 1.4983120984427747, "grad_norm": 0.19671176373958588, "learning_rate": 1.64145622763203e-05, "loss": 0.0066, "step": 6880 }, { "epoch": 1.5004900359359685, "grad_norm": 0.15893584489822388, "learning_rate": 1.6281636511884218e-05, "loss": 0.0054, "step": 6890 }, { "epoch": 1.5026679734291626, "grad_norm": 0.17707988619804382, "learning_rate": 1.6149146402646588e-05, "loss": 0.0053, "step": 6900 }, { "epoch": 1.5048459109223566, "grad_norm": 0.14699822664260864, "learning_rate": 1.601709366043439e-05, "loss": 0.0056, "step": 6910 }, { "epoch": 1.5070238484155505, "grad_norm": 0.1524478793144226, "learning_rate": 1.5885479991423667e-05, "loss": 0.0067, "step": 6920 }, { "epoch": 1.5092017859087443, "grad_norm": 0.183628648519516, "learning_rate": 1.5754307096117445e-05, "loss": 0.0053, "step": 6930 }, { "epoch": 1.5113797234019384, "grad_norm": 0.14608708024024963, "learning_rate": 1.5623576669323743e-05, "loss": 0.0058, "step": 6940 }, { "epoch": 1.5135576608951324, "grad_norm": 0.12225823849439621, "learning_rate": 1.5493290400133782e-05, "loss": 0.0051, "step": 6950 }, { "epoch": 1.5157355983883263, "grad_norm": 0.13646119832992554, "learning_rate": 1.536344997189999e-05, "loss": 0.0061, "step": 6960 }, { "epoch": 1.5179135358815201, "grad_norm": 0.13831979036331177, "learning_rate": 1.5234057062214402e-05, "loss": 0.0046, "step": 6970 }, { "epoch": 1.5200914733747142, "grad_norm": 0.11302287131547928, "learning_rate": 1.5105113342886912e-05, "loss": 0.0052, "step": 6980 }, { "epoch": 1.5222694108679082, "grad_norm": 0.12660376727581024, "learning_rate": 1.4976620479923687e-05, "loss": 0.0058, "step": 6990 }, { "epoch": 1.524447348361102, "grad_norm": 0.12537828087806702, "learning_rate": 1.4848580133505651e-05, "loss": 0.0048, "step": 7000 }, { "epoch": 1.526625285854296, "grad_norm": 0.11998041719198227, "learning_rate": 1.4720993957967022e-05, "loss": 0.0061, "step": 7010 }, { "epoch": 1.52880322334749, "grad_norm": 0.24933016300201416, "learning_rate": 1.4593863601773944e-05, "loss": 0.005, "step": 7020 }, { "epoch": 1.530981160840684, "grad_norm": 0.1901293247938156, "learning_rate": 1.4467190707503187e-05, "loss": 0.0052, "step": 7030 }, { "epoch": 1.5331590983338779, "grad_norm": 0.15064118802547455, "learning_rate": 1.4340976911820925e-05, "loss": 0.0041, "step": 7040 }, { "epoch": 1.5353370358270717, "grad_norm": 0.12917497754096985, "learning_rate": 1.4215223845461578e-05, "loss": 0.005, "step": 7050 }, { "epoch": 1.5375149733202655, "grad_norm": 0.08987925946712494, "learning_rate": 1.4089933133206762e-05, "loss": 0.0054, "step": 7060 }, { "epoch": 1.5396929108134596, "grad_norm": 0.10574358701705933, "learning_rate": 1.3965106393864297e-05, "loss": 0.0053, "step": 7070 }, { "epoch": 1.5418708483066537, "grad_norm": 0.11784743517637253, "learning_rate": 1.3840745240247261e-05, "loss": 0.0048, "step": 7080 }, { "epoch": 1.5440487857998475, "grad_norm": 0.10242784768342972, "learning_rate": 1.3716851279153176e-05, "loss": 0.0055, "step": 7090 }, { "epoch": 1.5462267232930413, "grad_norm": 0.11379238218069077, "learning_rate": 1.359342611134326e-05, "loss": 0.0059, "step": 7100 }, { "epoch": 1.5484046607862354, "grad_norm": 0.21122296154499054, "learning_rate": 1.347047133152171e-05, "loss": 0.0064, "step": 7110 }, { "epoch": 1.5505825982794295, "grad_norm": 0.09893481433391571, "learning_rate": 1.3347988528315119e-05, "loss": 0.0059, "step": 7120 }, { "epoch": 1.5527605357726233, "grad_norm": 0.15016481280326843, "learning_rate": 1.3225979284251954e-05, "loss": 0.0059, "step": 7130 }, { "epoch": 1.5549384732658171, "grad_norm": 0.1129075437784195, "learning_rate": 1.3104445175742097e-05, "loss": 0.0054, "step": 7140 }, { "epoch": 1.5571164107590112, "grad_norm": 0.11966872215270996, "learning_rate": 1.2983387773056482e-05, "loss": 0.005, "step": 7150 }, { "epoch": 1.5592943482522053, "grad_norm": 0.12486154586076736, "learning_rate": 1.2862808640306823e-05, "loss": 0.0064, "step": 7160 }, { "epoch": 1.561472285745399, "grad_norm": 0.16171671450138092, "learning_rate": 1.2742709335425329e-05, "loss": 0.0053, "step": 7170 }, { "epoch": 1.563650223238593, "grad_norm": 0.15467683970928192, "learning_rate": 1.2623091410144739e-05, "loss": 0.0048, "step": 7180 }, { "epoch": 1.565828160731787, "grad_norm": 0.17628200352191925, "learning_rate": 1.2503956409978047e-05, "loss": 0.0048, "step": 7190 }, { "epoch": 1.568006098224981, "grad_norm": 0.11883492022752762, "learning_rate": 1.2385305874198776e-05, "loss": 0.0054, "step": 7200 }, { "epoch": 1.5701840357181749, "grad_norm": 0.10365116596221924, "learning_rate": 1.2267141335820847e-05, "loss": 0.0045, "step": 7210 }, { "epoch": 1.5723619732113687, "grad_norm": 0.20156355202198029, "learning_rate": 1.2149464321579024e-05, "loss": 0.0056, "step": 7220 }, { "epoch": 1.5745399107045628, "grad_norm": 0.17337171733379364, "learning_rate": 1.2032276351908923e-05, "loss": 0.0053, "step": 7230 }, { "epoch": 1.5767178481977568, "grad_norm": 0.14272476732730865, "learning_rate": 1.1915578940927608e-05, "loss": 0.0048, "step": 7240 }, { "epoch": 1.5788957856909507, "grad_norm": 0.12799839675426483, "learning_rate": 1.1799373596413893e-05, "loss": 0.0058, "step": 7250 }, { "epoch": 1.5810737231841445, "grad_norm": 0.14727646112442017, "learning_rate": 1.1683661819788827e-05, "loss": 0.0052, "step": 7260 }, { "epoch": 1.5832516606773386, "grad_norm": 0.12912233173847198, "learning_rate": 1.1568445106096471e-05, "loss": 0.0055, "step": 7270 }, { "epoch": 1.5854295981705326, "grad_norm": 0.165940061211586, "learning_rate": 1.1453724943984346e-05, "loss": 0.0064, "step": 7280 }, { "epoch": 1.5876075356637265, "grad_norm": 0.2137843519449234, "learning_rate": 1.1339502815684438e-05, "loss": 0.0047, "step": 7290 }, { "epoch": 1.5897854731569203, "grad_norm": 0.14913132786750793, "learning_rate": 1.122578019699383e-05, "loss": 0.0053, "step": 7300 }, { "epoch": 1.5919634106501144, "grad_norm": 0.163711816072464, "learning_rate": 1.111255855725582e-05, "loss": 0.0052, "step": 7310 }, { "epoch": 1.5941413481433084, "grad_norm": 0.17178109288215637, "learning_rate": 1.099983935934077e-05, "loss": 0.0055, "step": 7320 }, { "epoch": 1.5963192856365023, "grad_norm": 0.18735261261463165, "learning_rate": 1.088762405962735e-05, "loss": 0.0052, "step": 7330 }, { "epoch": 1.598497223129696, "grad_norm": 0.13000036776065826, "learning_rate": 1.0775914107983615e-05, "loss": 0.005, "step": 7340 }, { "epoch": 1.6006751606228902, "grad_norm": 0.14790494740009308, "learning_rate": 1.0664710947748347e-05, "loss": 0.0054, "step": 7350 }, { "epoch": 1.6028530981160842, "grad_norm": 0.13647720217704773, "learning_rate": 1.0554016015712347e-05, "loss": 0.0048, "step": 7360 }, { "epoch": 1.605031035609278, "grad_norm": 0.08765874058008194, "learning_rate": 1.0443830742099914e-05, "loss": 0.0053, "step": 7370 }, { "epoch": 1.607208973102472, "grad_norm": 0.12826956808567047, "learning_rate": 1.0334156550550345e-05, "loss": 0.0055, "step": 7380 }, { "epoch": 1.6093869105956657, "grad_norm": 0.17395393550395966, "learning_rate": 1.0224994858099524e-05, "loss": 0.0053, "step": 7390 }, { "epoch": 1.6115648480888598, "grad_norm": 0.155074805021286, "learning_rate": 1.0116347075161698e-05, "loss": 0.0061, "step": 7400 }, { "epoch": 1.6137427855820539, "grad_norm": 0.11691328138113022, "learning_rate": 1.0008214605511101e-05, "loss": 0.0046, "step": 7410 }, { "epoch": 1.6159207230752477, "grad_norm": 0.12568417191505432, "learning_rate": 9.900598846263981e-06, "loss": 0.0056, "step": 7420 }, { "epoch": 1.6180986605684415, "grad_norm": 0.11361382901668549, "learning_rate": 9.793501187860432e-06, "loss": 0.0047, "step": 7430 }, { "epoch": 1.6202765980616356, "grad_norm": 0.09501596540212631, "learning_rate": 9.68692301404649e-06, "loss": 0.0056, "step": 7440 }, { "epoch": 1.6224545355548297, "grad_norm": 0.11473933607339859, "learning_rate": 9.580865701856222e-06, "loss": 0.0053, "step": 7450 }, { "epoch": 1.6246324730480235, "grad_norm": 0.16465674340724945, "learning_rate": 9.475330621593969e-06, "loss": 0.0053, "step": 7460 }, { "epoch": 1.6268104105412173, "grad_norm": 0.0834634080529213, "learning_rate": 9.370319136816591e-06, "loss": 0.0062, "step": 7470 }, { "epoch": 1.6289883480344114, "grad_norm": 0.10178032517433167, "learning_rate": 9.265832604315905e-06, "loss": 0.0048, "step": 7480 }, { "epoch": 1.6311662855276055, "grad_norm": 0.1961919069290161, "learning_rate": 9.161872374101105e-06, "loss": 0.0052, "step": 7490 }, { "epoch": 1.6333442230207993, "grad_norm": 0.16450706124305725, "learning_rate": 9.058439789381357e-06, "loss": 0.0049, "step": 7500 }, { "epoch": 1.6355221605139931, "grad_norm": 0.19026803970336914, "learning_rate": 8.955536186548425e-06, "loss": 0.0061, "step": 7510 }, { "epoch": 1.6377000980071872, "grad_norm": 0.08654351532459259, "learning_rate": 8.853162895159418e-06, "loss": 0.0044, "step": 7520 }, { "epoch": 1.6398780355003812, "grad_norm": 0.11829744279384613, "learning_rate": 8.751321237919547e-06, "loss": 0.0057, "step": 7530 }, { "epoch": 1.642055972993575, "grad_norm": 0.167857363820076, "learning_rate": 8.65001253066518e-06, "loss": 0.0066, "step": 7540 }, { "epoch": 1.644233910486769, "grad_norm": 0.09933420270681381, "learning_rate": 8.549238082346695e-06, "loss": 0.0062, "step": 7550 }, { "epoch": 1.646411847979963, "grad_norm": 0.12215885519981384, "learning_rate": 8.44899919501162e-06, "loss": 0.0057, "step": 7560 }, { "epoch": 1.648589785473157, "grad_norm": 0.18838058412075043, "learning_rate": 8.349297163787839e-06, "loss": 0.0048, "step": 7570 }, { "epoch": 1.6507677229663509, "grad_norm": 0.1070280522108078, "learning_rate": 8.250133276866818e-06, "loss": 0.0048, "step": 7580 }, { "epoch": 1.6529456604595447, "grad_norm": 0.14078475534915924, "learning_rate": 8.151508815486974e-06, "loss": 0.0052, "step": 7590 }, { "epoch": 1.6551235979527388, "grad_norm": 0.2071860283613205, "learning_rate": 8.05342505391714e-06, "loss": 0.0058, "step": 7600 }, { "epoch": 1.6573015354459328, "grad_norm": 0.2031438946723938, "learning_rate": 7.955883259440072e-06, "loss": 0.0045, "step": 7610 }, { "epoch": 1.6594794729391267, "grad_norm": 0.17647232115268707, "learning_rate": 7.858884692336055e-06, "loss": 0.0052, "step": 7620 }, { "epoch": 1.6616574104323205, "grad_norm": 0.11669361591339111, "learning_rate": 7.762430605866734e-06, "loss": 0.0049, "step": 7630 }, { "epoch": 1.6638353479255146, "grad_norm": 0.11189775913953781, "learning_rate": 7.66652224625874e-06, "loss": 0.0059, "step": 7640 }, { "epoch": 1.6660132854187086, "grad_norm": 0.16998837888240814, "learning_rate": 7.571160852687786e-06, "loss": 0.0068, "step": 7650 }, { "epoch": 1.6681912229119025, "grad_norm": 0.15501129627227783, "learning_rate": 7.476347657262456e-06, "loss": 0.0044, "step": 7660 }, { "epoch": 1.6703691604050963, "grad_norm": 0.11749287694692612, "learning_rate": 7.382083885008484e-06, "loss": 0.005, "step": 7670 }, { "epoch": 1.6725470978982904, "grad_norm": 0.09791196137666702, "learning_rate": 7.288370753852758e-06, "loss": 0.0054, "step": 7680 }, { "epoch": 1.6747250353914844, "grad_norm": 0.10001195967197418, "learning_rate": 7.195209474607678e-06, "loss": 0.0054, "step": 7690 }, { "epoch": 1.6769029728846783, "grad_norm": 0.1074577197432518, "learning_rate": 7.1026012509555265e-06, "loss": 0.0061, "step": 7700 }, { "epoch": 1.679080910377872, "grad_norm": 0.15040123462677002, "learning_rate": 7.0105472794328e-06, "loss": 0.0053, "step": 7710 }, { "epoch": 1.681258847871066, "grad_norm": 0.16351404786109924, "learning_rate": 6.919048749414919e-06, "loss": 0.0052, "step": 7720 }, { "epoch": 1.68343678536426, "grad_norm": 0.11897435784339905, "learning_rate": 6.828106843100679e-06, "loss": 0.0048, "step": 7730 }, { "epoch": 1.685614722857454, "grad_norm": 0.20059573650360107, "learning_rate": 6.737722735497165e-06, "loss": 0.0055, "step": 7740 }, { "epoch": 1.687792660350648, "grad_norm": 0.4155087471008301, "learning_rate": 6.6478975944043675e-06, "loss": 0.006, "step": 7750 }, { "epoch": 1.6899705978438417, "grad_norm": 0.1162080243229866, "learning_rate": 6.558632580400298e-06, "loss": 0.0054, "step": 7760 }, { "epoch": 1.6921485353370358, "grad_norm": 0.14566195011138916, "learning_rate": 6.4699288468258126e-06, "loss": 0.0051, "step": 7770 }, { "epoch": 1.6943264728302299, "grad_norm": 0.1610737144947052, "learning_rate": 6.38178753976984e-06, "loss": 0.0063, "step": 7780 }, { "epoch": 1.6965044103234237, "grad_norm": 0.21307019889354706, "learning_rate": 6.2942097980545176e-06, "loss": 0.0061, "step": 7790 }, { "epoch": 1.6986823478166175, "grad_norm": 0.10614914447069168, "learning_rate": 6.207196753220473e-06, "loss": 0.0043, "step": 7800 }, { "epoch": 1.7008602853098116, "grad_norm": 0.09863957017660141, "learning_rate": 6.120749529512226e-06, "loss": 0.0044, "step": 7810 }, { "epoch": 1.7030382228030057, "grad_norm": 0.12667401134967804, "learning_rate": 6.034869243863656e-06, "loss": 0.0047, "step": 7820 }, { "epoch": 1.7052161602961995, "grad_norm": 0.13384513556957245, "learning_rate": 5.94955700588356e-06, "loss": 0.0048, "step": 7830 }, { "epoch": 1.7073940977893933, "grad_norm": 0.14361801743507385, "learning_rate": 5.864813917841333e-06, "loss": 0.0047, "step": 7840 }, { "epoch": 1.7095720352825874, "grad_norm": 0.15779605507850647, "learning_rate": 5.780641074652704e-06, "loss": 0.0046, "step": 7850 }, { "epoch": 1.7117499727757814, "grad_norm": 0.1296186000108719, "learning_rate": 5.697039563865625e-06, "loss": 0.0054, "step": 7860 }, { "epoch": 1.7139279102689753, "grad_norm": 0.0921684131026268, "learning_rate": 5.614010465646169e-06, "loss": 0.0055, "step": 7870 }, { "epoch": 1.7161058477621691, "grad_norm": 0.1396203637123108, "learning_rate": 5.531554852764626e-06, "loss": 0.0051, "step": 7880 }, { "epoch": 1.7182837852553632, "grad_norm": 0.09174776077270508, "learning_rate": 5.449673790581611e-06, "loss": 0.0056, "step": 7890 }, { "epoch": 1.7204617227485572, "grad_norm": 0.1453590840101242, "learning_rate": 5.368368337034296e-06, "loss": 0.0061, "step": 7900 }, { "epoch": 1.722639660241751, "grad_norm": 0.08000554889440536, "learning_rate": 5.287639542622763e-06, "loss": 0.0047, "step": 7910 }, { "epoch": 1.724817597734945, "grad_norm": 0.10845358669757843, "learning_rate": 5.20748845039642e-06, "loss": 0.0055, "step": 7920 }, { "epoch": 1.726995535228139, "grad_norm": 0.10388531535863876, "learning_rate": 5.127916095940516e-06, "loss": 0.0054, "step": 7930 }, { "epoch": 1.729173472721333, "grad_norm": 0.14150455594062805, "learning_rate": 5.048923507362768e-06, "loss": 0.0047, "step": 7940 }, { "epoch": 1.7313514102145269, "grad_norm": 0.08367276191711426, "learning_rate": 4.970511705280096e-06, "loss": 0.0051, "step": 7950 }, { "epoch": 1.7335293477077207, "grad_norm": 0.071738101541996, "learning_rate": 4.892681702805396e-06, "loss": 0.0047, "step": 7960 }, { "epoch": 1.7357072852009148, "grad_norm": 0.09582991153001785, "learning_rate": 4.815434505534494e-06, "loss": 0.0048, "step": 7970 }, { "epoch": 1.7378852226941088, "grad_norm": 0.11175650358200073, "learning_rate": 4.738771111533091e-06, "loss": 0.0047, "step": 7980 }, { "epoch": 1.7400631601873027, "grad_norm": 0.14590439200401306, "learning_rate": 4.662692511323968e-06, "loss": 0.0048, "step": 7990 }, { "epoch": 1.7422410976804965, "grad_norm": 0.1824604868888855, "learning_rate": 4.587199687874083e-06, "loss": 0.0055, "step": 8000 }, { "epoch": 1.7444190351736906, "grad_norm": 0.09297237545251846, "learning_rate": 4.512293616581936e-06, "loss": 0.0043, "step": 8010 }, { "epoch": 1.7465969726668844, "grad_norm": 0.1086159348487854, "learning_rate": 4.437975265264932e-06, "loss": 0.0045, "step": 8020 }, { "epoch": 1.7487749101600785, "grad_norm": 0.17781151831150055, "learning_rate": 4.36424559414691e-06, "loss": 0.0055, "step": 8030 }, { "epoch": 1.7509528476532723, "grad_norm": 0.14021795988082886, "learning_rate": 4.291105555845698e-06, "loss": 0.0052, "step": 8040 }, { "epoch": 1.7531307851464661, "grad_norm": 0.1870332658290863, "learning_rate": 4.218556095360843e-06, "loss": 0.0053, "step": 8050 }, { "epoch": 1.7553087226396602, "grad_norm": 0.17223937809467316, "learning_rate": 4.146598150061371e-06, "loss": 0.0047, "step": 8060 }, { "epoch": 1.7574866601328543, "grad_norm": 0.1115616112947464, "learning_rate": 4.075232649673677e-06, "loss": 0.0051, "step": 8070 }, { "epoch": 1.759664597626048, "grad_norm": 0.15488360822200775, "learning_rate": 4.004460516269554e-06, "loss": 0.0048, "step": 8080 }, { "epoch": 1.761842535119242, "grad_norm": 0.102665014564991, "learning_rate": 3.934282664254208e-06, "loss": 0.0049, "step": 8090 }, { "epoch": 1.764020472612436, "grad_norm": 0.14327962696552277, "learning_rate": 3.864700000354526e-06, "loss": 0.0059, "step": 8100 }, { "epoch": 1.76619841010563, "grad_norm": 0.1204613596200943, "learning_rate": 3.7957134236072722e-06, "loss": 0.005, "step": 8110 }, { "epoch": 1.768376347598824, "grad_norm": 0.08825606107711792, "learning_rate": 3.7273238253475785e-06, "loss": 0.0047, "step": 8120 }, { "epoch": 1.7705542850920177, "grad_norm": 0.10067594796419144, "learning_rate": 3.659532089197304e-06, "loss": 0.0052, "step": 8130 }, { "epoch": 1.7727322225852118, "grad_norm": 0.10493259876966476, "learning_rate": 3.592339091053709e-06, "loss": 0.0052, "step": 8140 }, { "epoch": 1.7749101600784059, "grad_norm": 0.11088038980960846, "learning_rate": 3.5257456990781424e-06, "loss": 0.0068, "step": 8150 }, { "epoch": 1.7770880975715997, "grad_norm": 0.09026432037353516, "learning_rate": 3.459752773684727e-06, "loss": 0.0046, "step": 8160 }, { "epoch": 1.7792660350647935, "grad_norm": 0.10548806190490723, "learning_rate": 3.3943611675293807e-06, "loss": 0.0052, "step": 8170 }, { "epoch": 1.7814439725579876, "grad_norm": 0.12848129868507385, "learning_rate": 3.329571725498665e-06, "loss": 0.0044, "step": 8180 }, { "epoch": 1.7836219100511816, "grad_norm": 0.0845961943268776, "learning_rate": 3.2653852846989852e-06, "loss": 0.0043, "step": 8190 }, { "epoch": 1.7857998475443755, "grad_norm": 0.11423788219690323, "learning_rate": 3.2018026744456754e-06, "loss": 0.0054, "step": 8200 }, { "epoch": 1.7879777850375693, "grad_norm": 0.6692479252815247, "learning_rate": 3.13882471625237e-06, "loss": 0.0046, "step": 8210 }, { "epoch": 1.7901557225307634, "grad_norm": 0.11326146125793457, "learning_rate": 3.0764522238203143e-06, "loss": 0.0048, "step": 8220 }, { "epoch": 1.7923336600239574, "grad_norm": 0.126267209649086, "learning_rate": 3.0146860030279058e-06, "loss": 0.0044, "step": 8230 }, { "epoch": 1.7945115975171513, "grad_norm": 0.13366961479187012, "learning_rate": 2.953526851920263e-06, "loss": 0.0049, "step": 8240 }, { "epoch": 1.7966895350103451, "grad_norm": 0.1864164024591446, "learning_rate": 2.892975560698896e-06, "loss": 0.0045, "step": 8250 }, { "epoch": 1.7988674725035392, "grad_norm": 0.14489804208278656, "learning_rate": 2.8330329117115304e-06, "loss": 0.0041, "step": 8260 }, { "epoch": 1.8010454099967332, "grad_norm": 0.17568811774253845, "learning_rate": 2.7736996794419767e-06, "loss": 0.0044, "step": 8270 }, { "epoch": 1.803223347489927, "grad_norm": 0.1365700215101242, "learning_rate": 2.7149766305001268e-06, "loss": 0.0044, "step": 8280 }, { "epoch": 1.805401284983121, "grad_norm": 0.13300161063671112, "learning_rate": 2.6568645236120627e-06, "loss": 0.0053, "step": 8290 }, { "epoch": 1.807579222476315, "grad_norm": 0.09710974991321564, "learning_rate": 2.599364109610225e-06, "loss": 0.0053, "step": 8300 }, { "epoch": 1.809757159969509, "grad_norm": 0.09724448621273041, "learning_rate": 2.542476131423743e-06, "loss": 0.0046, "step": 8310 }, { "epoch": 1.8119350974627029, "grad_norm": 0.13140472769737244, "learning_rate": 2.4862013240688152e-06, "loss": 0.0054, "step": 8320 }, { "epoch": 1.8141130349558967, "grad_norm": 0.13002344965934753, "learning_rate": 2.4305404146392207e-06, "loss": 0.0047, "step": 8330 }, { "epoch": 1.8162909724490905, "grad_norm": 0.13739478588104248, "learning_rate": 2.3754941222969283e-06, "loss": 0.0051, "step": 8340 }, { "epoch": 1.8184689099422846, "grad_norm": 0.10706328600645065, "learning_rate": 2.321063158262793e-06, "loss": 0.0057, "step": 8350 }, { "epoch": 1.8206468474354787, "grad_norm": 0.11440842598676682, "learning_rate": 2.267248225807378e-06, "loss": 0.0045, "step": 8360 }, { "epoch": 1.8228247849286725, "grad_norm": 0.087703175842762, "learning_rate": 2.2140500202418644e-06, "loss": 0.0044, "step": 8370 }, { "epoch": 1.8250027224218663, "grad_norm": 0.15237921476364136, "learning_rate": 2.161469228909063e-06, "loss": 0.0052, "step": 8380 }, { "epoch": 1.8271806599150604, "grad_norm": 0.06693877279758453, "learning_rate": 2.109506531174538e-06, "loss": 0.0045, "step": 8390 }, { "epoch": 1.8293585974082545, "grad_norm": 0.08744368702173233, "learning_rate": 2.0581625984178343e-06, "loss": 0.0049, "step": 8400 }, { "epoch": 1.8315365349014483, "grad_norm": 0.12937608361244202, "learning_rate": 2.0074380940237893e-06, "loss": 0.005, "step": 8410 }, { "epoch": 1.8337144723946421, "grad_norm": 0.10375534743070602, "learning_rate": 1.957333673373979e-06, "loss": 0.0054, "step": 8420 }, { "epoch": 1.8358924098878362, "grad_norm": 0.11480173468589783, "learning_rate": 1.907849983838217e-06, "loss": 0.0041, "step": 8430 }, { "epoch": 1.8380703473810303, "grad_norm": 0.10196216404438019, "learning_rate": 1.8589876647662597e-06, "loss": 0.005, "step": 8440 }, { "epoch": 1.840248284874224, "grad_norm": 0.1354270875453949, "learning_rate": 1.810747347479458e-06, "loss": 0.0049, "step": 8450 }, { "epoch": 1.842426222367418, "grad_norm": 0.11422459781169891, "learning_rate": 1.7631296552626685e-06, "loss": 0.0043, "step": 8460 }, { "epoch": 1.844604159860612, "grad_norm": 0.19297891855239868, "learning_rate": 1.7161352033561505e-06, "loss": 0.0042, "step": 8470 }, { "epoch": 1.846782097353806, "grad_norm": 0.08900764584541321, "learning_rate": 1.6697645989476651e-06, "loss": 0.0045, "step": 8480 }, { "epoch": 1.848960034847, "grad_norm": 0.11755415052175522, "learning_rate": 1.624018441164593e-06, "loss": 0.0036, "step": 8490 }, { "epoch": 1.8511379723401937, "grad_norm": 0.11477325111627579, "learning_rate": 1.5788973210661973e-06, "loss": 0.005, "step": 8500 }, { "epoch": 1.8533159098333878, "grad_norm": 0.08399289101362228, "learning_rate": 1.534401821636028e-06, "loss": 0.0049, "step": 8510 }, { "epoch": 1.8554938473265818, "grad_norm": 0.06838883459568024, "learning_rate": 1.4905325177743123e-06, "loss": 0.0047, "step": 8520 }, { "epoch": 1.8576717848197757, "grad_norm": 0.12109494209289551, "learning_rate": 1.4472899762906167e-06, "loss": 0.0053, "step": 8530 }, { "epoch": 1.8598497223129695, "grad_norm": 0.13868533074855804, "learning_rate": 1.4046747558964403e-06, "loss": 0.0044, "step": 8540 }, { "epoch": 1.8620276598061636, "grad_norm": 0.11102143675088882, "learning_rate": 1.362687407198071e-06, "loss": 0.0039, "step": 8550 }, { "epoch": 1.8642055972993576, "grad_norm": 0.12908202409744263, "learning_rate": 1.321328472689387e-06, "loss": 0.0046, "step": 8560 }, { "epoch": 1.8663835347925515, "grad_norm": 0.08044327050447464, "learning_rate": 1.280598486744955e-06, "loss": 0.0045, "step": 8570 }, { "epoch": 1.8685614722857453, "grad_norm": 0.13346053659915924, "learning_rate": 1.2404979756130142e-06, "loss": 0.0048, "step": 8580 }, { "epoch": 1.8707394097789394, "grad_norm": 0.07873251289129257, "learning_rate": 1.2010274574087544e-06, "loss": 0.0045, "step": 8590 }, { "epoch": 1.8729173472721334, "grad_norm": 0.1406606286764145, "learning_rate": 1.1621874421075984e-06, "loss": 0.0046, "step": 8600 }, { "epoch": 1.8750952847653273, "grad_norm": 0.12321413308382034, "learning_rate": 1.1239784315385859e-06, "loss": 0.005, "step": 8610 }, { "epoch": 1.8772732222585211, "grad_norm": 0.1397770792245865, "learning_rate": 1.0864009193779613e-06, "loss": 0.0043, "step": 8620 }, { "epoch": 1.8794511597517152, "grad_norm": 0.09104544669389725, "learning_rate": 1.0494553911426841e-06, "loss": 0.0047, "step": 8630 }, { "epoch": 1.8816290972449092, "grad_norm": 0.10594195872545242, "learning_rate": 1.0131423241842907e-06, "loss": 0.0039, "step": 8640 }, { "epoch": 1.883807034738103, "grad_norm": 0.07154612988233566, "learning_rate": 9.774621876825983e-07, "loss": 0.0044, "step": 8650 }, { "epoch": 1.885984972231297, "grad_norm": 0.10314908623695374, "learning_rate": 9.424154426397269e-07, "loss": 0.0048, "step": 8660 }, { "epoch": 1.8881629097244907, "grad_norm": 0.062456533312797546, "learning_rate": 9.080025418741255e-07, "loss": 0.0043, "step": 8670 }, { "epoch": 1.8903408472176848, "grad_norm": 0.11628817766904831, "learning_rate": 8.742239300146838e-07, "loss": 0.0041, "step": 8680 }, { "epoch": 1.8925187847108789, "grad_norm": 0.0872725322842598, "learning_rate": 8.410800434950517e-07, "loss": 0.005, "step": 8690 }, { "epoch": 1.8946967222040727, "grad_norm": 0.09369292855262756, "learning_rate": 8.085713105479343e-07, "loss": 0.0061, "step": 8700 }, { "epoch": 1.8968746596972665, "grad_norm": 0.12869521975517273, "learning_rate": 7.766981511996119e-07, "loss": 0.0044, "step": 8710 }, { "epoch": 1.8990525971904606, "grad_norm": 0.12514403462409973, "learning_rate": 7.454609772644783e-07, "loss": 0.004, "step": 8720 }, { "epoch": 1.9012305346836547, "grad_norm": 0.07966753095388412, "learning_rate": 7.148601923397391e-07, "loss": 0.0045, "step": 8730 }, { "epoch": 1.9034084721768485, "grad_norm": 0.07699485123157501, "learning_rate": 6.848961918001829e-07, "loss": 0.0046, "step": 8740 }, { "epoch": 1.9055864096700423, "grad_norm": 0.09614022076129913, "learning_rate": 6.555693627930958e-07, "loss": 0.0047, "step": 8750 }, { "epoch": 1.9077643471632364, "grad_norm": 0.07762955874204636, "learning_rate": 6.268800842332278e-07, "loss": 0.0044, "step": 8760 }, { "epoch": 1.9099422846564305, "grad_norm": 0.13291290402412415, "learning_rate": 5.988287267979231e-07, "loss": 0.0058, "step": 8770 }, { "epoch": 1.9121202221496243, "grad_norm": 0.1213541105389595, "learning_rate": 5.714156529223191e-07, "loss": 0.005, "step": 8780 }, { "epoch": 1.9142981596428181, "grad_norm": 0.20027388632297516, "learning_rate": 5.446412167946613e-07, "loss": 0.0052, "step": 8790 }, { "epoch": 1.9164760971360122, "grad_norm": 0.11805108934640884, "learning_rate": 5.185057643517455e-07, "loss": 0.0049, "step": 8800 }, { "epoch": 1.9186540346292063, "grad_norm": 0.13726966083049774, "learning_rate": 4.930096332744105e-07, "loss": 0.0037, "step": 8810 }, { "epoch": 1.9208319721224, "grad_norm": 0.09815724194049835, "learning_rate": 4.6815315298322506e-07, "loss": 0.0057, "step": 8820 }, { "epoch": 1.923009909615594, "grad_norm": 0.07099764794111252, "learning_rate": 4.439366446341797e-07, "loss": 0.0048, "step": 8830 }, { "epoch": 1.925187847108788, "grad_norm": 0.1317773163318634, "learning_rate": 4.2036042111458506e-07, "loss": 0.0042, "step": 8840 }, { "epoch": 1.927365784601982, "grad_norm": 0.10230360925197601, "learning_rate": 3.974247870389969e-07, "loss": 0.0049, "step": 8850 }, { "epoch": 1.9295437220951759, "grad_norm": 0.11070272326469421, "learning_rate": 3.7513003874528606e-07, "loss": 0.0039, "step": 8860 }, { "epoch": 1.9317216595883697, "grad_norm": 0.13296164572238922, "learning_rate": 3.5347646429082505e-07, "loss": 0.004, "step": 8870 }, { "epoch": 1.9338995970815638, "grad_norm": 0.09902767091989517, "learning_rate": 3.3246434344874067e-07, "loss": 0.0041, "step": 8880 }, { "epoch": 1.9360775345747578, "grad_norm": 0.15179429948329926, "learning_rate": 3.1209394770433384e-07, "loss": 0.004, "step": 8890 }, { "epoch": 1.9382554720679517, "grad_norm": 0.13276688754558563, "learning_rate": 2.923655402515324e-07, "loss": 0.0043, "step": 8900 }, { "epoch": 1.9404334095611455, "grad_norm": 0.13763545453548431, "learning_rate": 2.7327937598952134e-07, "loss": 0.0051, "step": 8910 }, { "epoch": 1.9426113470543396, "grad_norm": 0.06456010043621063, "learning_rate": 2.5483570151942913e-07, "loss": 0.0041, "step": 8920 }, { "epoch": 1.9447892845475336, "grad_norm": 0.08242253214120865, "learning_rate": 2.370347551411689e-07, "loss": 0.0044, "step": 8930 }, { "epoch": 1.9469672220407275, "grad_norm": 0.13994471728801727, "learning_rate": 2.1987676685031878e-07, "loss": 0.0046, "step": 8940 }, { "epoch": 1.9491451595339213, "grad_norm": 0.159123495221138, "learning_rate": 2.0336195833517424e-07, "loss": 0.0044, "step": 8950 }, { "epoch": 1.9513230970271154, "grad_norm": 0.08478546142578125, "learning_rate": 1.8749054297391155e-07, "loss": 0.0045, "step": 8960 }, { "epoch": 1.9535010345203094, "grad_norm": 0.11139588057994843, "learning_rate": 1.7226272583176218e-07, "loss": 0.0044, "step": 8970 }, { "epoch": 1.9556789720135033, "grad_norm": 0.20084549486637115, "learning_rate": 1.5767870365842597e-07, "loss": 0.0041, "step": 8980 }, { "epoch": 1.957856909506697, "grad_norm": 0.08986944705247879, "learning_rate": 1.4373866488549548e-07, "loss": 0.0047, "step": 8990 }, { "epoch": 1.960034846999891, "grad_norm": 0.13636291027069092, "learning_rate": 1.3044278962404676e-07, "loss": 0.004, "step": 9000 }, { "epoch": 1.962212784493085, "grad_norm": 0.0729387030005455, "learning_rate": 1.1779124966227462e-07, "loss": 0.0047, "step": 9010 }, { "epoch": 1.964390721986279, "grad_norm": 0.09830494225025177, "learning_rate": 1.0578420846331649e-07, "loss": 0.0044, "step": 9020 }, { "epoch": 1.966568659479473, "grad_norm": 0.089515820145607, "learning_rate": 9.442182116309872e-08, "loss": 0.0044, "step": 9030 }, { "epoch": 1.9687465969726667, "grad_norm": 0.10633354634046555, "learning_rate": 8.37042345683714e-08, "loss": 0.0048, "step": 9040 }, { "epoch": 1.9709245344658608, "grad_norm": 0.11465106159448624, "learning_rate": 7.363158715478213e-08, "loss": 0.0042, "step": 9050 }, { "epoch": 1.9731024719590549, "grad_norm": 0.11644294857978821, "learning_rate": 6.420400906509417e-08, "loss": 0.0042, "step": 9060 }, { "epoch": 1.9752804094522487, "grad_norm": 0.1346789002418518, "learning_rate": 5.542162210752655e-08, "loss": 0.0046, "step": 9070 }, { "epoch": 1.9774583469454425, "grad_norm": 0.08295319229364395, "learning_rate": 4.7284539754144374e-08, "loss": 0.0037, "step": 9080 }, { "epoch": 1.9796362844386366, "grad_norm": 0.10331818461418152, "learning_rate": 3.9792867139415434e-08, "loss": 0.0057, "step": 9090 }, { "epoch": 1.9818142219318307, "grad_norm": 0.12241711467504501, "learning_rate": 3.2946701058850226e-08, "loss": 0.0039, "step": 9100 }, { "epoch": 1.9839921594250245, "grad_norm": 0.07530458271503448, "learning_rate": 2.674612996774184e-08, "loss": 0.0047, "step": 9110 }, { "epoch": 1.9861700969182183, "grad_norm": 0.10693301260471344, "learning_rate": 2.119123398003353e-08, "loss": 0.0046, "step": 9120 }, { "epoch": 1.9883480344114124, "grad_norm": 0.0734107717871666, "learning_rate": 1.6282084867280668e-08, "loss": 0.0042, "step": 9130 }, { "epoch": 1.9905259719046065, "grad_norm": 0.09543687850236893, "learning_rate": 1.2018746057718134e-08, "loss": 0.005, "step": 9140 }, { "epoch": 1.9927039093978003, "grad_norm": 0.0988156795501709, "learning_rate": 8.401272635449875e-09, "loss": 0.0049, "step": 9150 }, { "epoch": 1.9948818468909941, "grad_norm": 0.11033082753419876, "learning_rate": 5.429711339721699e-09, "loss": 0.0043, "step": 9160 }, { "epoch": 1.9970597843841882, "grad_norm": 0.09224304556846619, "learning_rate": 3.104100564338408e-09, "loss": 0.005, "step": 9170 }, { "epoch": 1.9992377218773822, "grad_norm": 0.12420514971017838, "learning_rate": 1.4244703571475449e-09, "loss": 0.0043, "step": 9180 }, { "epoch": 2.001306762495916, "grad_norm": 0.13912928104400635, "learning_rate": 3.908424196730209e-10, "loss": 0.0045, "step": 9190 }, { "epoch": 2.0034846999891105, "grad_norm": 0.15209396183490753, "learning_rate": 3.2301068153550717e-12, "loss": 0.0049, "step": 9200 }, { "epoch": 2.0034846999891105, "step": 9200, "total_flos": 0.0, "train_loss": 0.013528645742360665, "train_runtime": 20376.5501, "train_samples_per_second": 115.584, "train_steps_per_second": 0.451 } ], "logging_steps": 10, "max_steps": 9200, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 4600, "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": 64, "trial_name": null, "trial_params": null }