{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 4.0, "eval_steps": 500, "global_step": 4376, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.009140767824497258, "grad_norm": 20.285207748413086, "learning_rate": 4.10958904109589e-06, "loss": 1.4218, "step": 10 }, { "epoch": 0.018281535648994516, "grad_norm": 3.7156178951263428, "learning_rate": 8.675799086757991e-06, "loss": 0.6623, "step": 20 }, { "epoch": 0.027422303473491772, "grad_norm": 3.670661687850952, "learning_rate": 1.3242009132420092e-05, "loss": 0.3818, "step": 30 }, { "epoch": 0.03656307129798903, "grad_norm": 1.6582924127578735, "learning_rate": 1.780821917808219e-05, "loss": 0.2857, "step": 40 }, { "epoch": 0.04570383912248629, "grad_norm": 3.376695394515991, "learning_rate": 2.237442922374429e-05, "loss": 0.2503, "step": 50 }, { "epoch": 0.054844606946983544, "grad_norm": 2.420069456100464, "learning_rate": 2.6940639269406392e-05, "loss": 0.211, "step": 60 }, { "epoch": 0.06398537477148081, "grad_norm": 2.5535237789154053, "learning_rate": 3.1506849315068496e-05, "loss": 0.1969, "step": 70 }, { "epoch": 0.07312614259597806, "grad_norm": 1.6121954917907715, "learning_rate": 3.60730593607306e-05, "loss": 0.1838, "step": 80 }, { "epoch": 0.08226691042047532, "grad_norm": 2.0982680320739746, "learning_rate": 4.063926940639269e-05, "loss": 0.1935, "step": 90 }, { "epoch": 0.09140767824497258, "grad_norm": 2.1739630699157715, "learning_rate": 4.520547945205479e-05, "loss": 0.1721, "step": 100 }, { "epoch": 0.10054844606946983, "grad_norm": 2.0316410064697266, "learning_rate": 4.977168949771689e-05, "loss": 0.1454, "step": 110 }, { "epoch": 0.10968921389396709, "grad_norm": 2.113408327102661, "learning_rate": 5.4337899543379e-05, "loss": 0.1445, "step": 120 }, { "epoch": 0.11882998171846434, "grad_norm": 1.6204296350479126, "learning_rate": 5.89041095890411e-05, "loss": 0.1178, "step": 130 }, { "epoch": 0.12797074954296161, "grad_norm": 1.2460007667541504, "learning_rate": 6.34703196347032e-05, "loss": 0.1113, "step": 140 }, { "epoch": 0.13711151736745886, "grad_norm": 2.4157071113586426, "learning_rate": 6.803652968036531e-05, "loss": 0.1038, "step": 150 }, { "epoch": 0.14625228519195613, "grad_norm": 1.6945436000823975, "learning_rate": 7.26027397260274e-05, "loss": 0.1148, "step": 160 }, { "epoch": 0.15539305301645337, "grad_norm": 1.4777846336364746, "learning_rate": 7.716894977168951e-05, "loss": 0.1147, "step": 170 }, { "epoch": 0.16453382084095064, "grad_norm": 1.7918751239776611, "learning_rate": 8.17351598173516e-05, "loss": 0.1149, "step": 180 }, { "epoch": 0.1736745886654479, "grad_norm": 1.584354043006897, "learning_rate": 8.630136986301371e-05, "loss": 0.113, "step": 190 }, { "epoch": 0.18281535648994515, "grad_norm": 1.4646424055099487, "learning_rate": 9.08675799086758e-05, "loss": 0.0993, "step": 200 }, { "epoch": 0.19195612431444242, "grad_norm": 1.1241822242736816, "learning_rate": 9.543378995433791e-05, "loss": 0.1105, "step": 210 }, { "epoch": 0.20109689213893966, "grad_norm": 1.0951234102249146, "learning_rate": 0.0001, "loss": 0.0891, "step": 220 }, { "epoch": 0.21023765996343693, "grad_norm": 0.955017626285553, "learning_rate": 9.999857216626715e-05, "loss": 0.0948, "step": 230 }, { "epoch": 0.21937842778793418, "grad_norm": 1.671467900276184, "learning_rate": 9.999428874661695e-05, "loss": 0.089, "step": 240 }, { "epoch": 0.22851919561243145, "grad_norm": 1.0147258043289185, "learning_rate": 9.998714998568987e-05, "loss": 0.0923, "step": 250 }, { "epoch": 0.2376599634369287, "grad_norm": 1.6253561973571777, "learning_rate": 9.997715629120446e-05, "loss": 0.081, "step": 260 }, { "epoch": 0.24680073126142596, "grad_norm": 1.460111141204834, "learning_rate": 9.996430823393403e-05, "loss": 0.086, "step": 270 }, { "epoch": 0.25594149908592323, "grad_norm": 1.5847985744476318, "learning_rate": 9.994860654767421e-05, "loss": 0.0882, "step": 280 }, { "epoch": 0.26508226691042047, "grad_norm": 0.8328115344047546, "learning_rate": 9.99300521292009e-05, "loss": 0.0884, "step": 290 }, { "epoch": 0.2742230347349177, "grad_norm": 0.7161694765090942, "learning_rate": 9.990864603821906e-05, "loss": 0.0874, "step": 300 }, { "epoch": 0.283363802559415, "grad_norm": 1.3507859706878662, "learning_rate": 9.988438949730223e-05, "loss": 0.0902, "step": 310 }, { "epoch": 0.29250457038391225, "grad_norm": 1.1614631414413452, "learning_rate": 9.985728389182275e-05, "loss": 0.0773, "step": 320 }, { "epoch": 0.3016453382084095, "grad_norm": 1.271041989326477, "learning_rate": 9.982733076987248e-05, "loss": 0.0974, "step": 330 }, { "epoch": 0.31078610603290674, "grad_norm": 1.0431146621704102, "learning_rate": 9.979453184217456e-05, "loss": 0.0872, "step": 340 }, { "epoch": 0.31992687385740404, "grad_norm": 2.105883836746216, "learning_rate": 9.975888898198562e-05, "loss": 0.0936, "step": 350 }, { "epoch": 0.3290676416819013, "grad_norm": 1.3111586570739746, "learning_rate": 9.972040422498877e-05, "loss": 0.102, "step": 360 }, { "epoch": 0.3382084095063985, "grad_norm": 1.21346914768219, "learning_rate": 9.967907976917738e-05, "loss": 0.0754, "step": 370 }, { "epoch": 0.3473491773308958, "grad_norm": 0.7348060607910156, "learning_rate": 9.963491797472953e-05, "loss": 0.0703, "step": 380 }, { "epoch": 0.35648994515539306, "grad_norm": 0.7553760409355164, "learning_rate": 9.958792136387322e-05, "loss": 0.083, "step": 390 }, { "epoch": 0.3656307129798903, "grad_norm": 1.1796748638153076, "learning_rate": 9.95380926207423e-05, "loss": 0.0642, "step": 400 }, { "epoch": 0.37477148080438755, "grad_norm": 0.9751859903335571, "learning_rate": 9.948543459122317e-05, "loss": 0.0735, "step": 410 }, { "epoch": 0.38391224862888484, "grad_norm": 1.2366355657577515, "learning_rate": 9.942995028279227e-05, "loss": 0.0748, "step": 420 }, { "epoch": 0.3930530164533821, "grad_norm": 1.4610956907272339, "learning_rate": 9.93716428643443e-05, "loss": 0.0973, "step": 430 }, { "epoch": 0.40219378427787933, "grad_norm": 1.440199375152588, "learning_rate": 9.931051566601119e-05, "loss": 0.0765, "step": 440 }, { "epoch": 0.4113345521023766, "grad_norm": 1.6553508043289185, "learning_rate": 9.924657217897201e-05, "loss": 0.0922, "step": 450 }, { "epoch": 0.42047531992687387, "grad_norm": 1.534443736076355, "learning_rate": 9.917981605525345e-05, "loss": 0.1115, "step": 460 }, { "epoch": 0.4296160877513711, "grad_norm": 1.0473862886428833, "learning_rate": 9.911025110752132e-05, "loss": 0.0704, "step": 470 }, { "epoch": 0.43875685557586835, "grad_norm": 1.1698426008224487, "learning_rate": 9.90378813088628e-05, "loss": 0.0788, "step": 480 }, { "epoch": 0.44789762340036565, "grad_norm": 1.1977952718734741, "learning_rate": 9.896271079255943e-05, "loss": 0.0967, "step": 490 }, { "epoch": 0.4570383912248629, "grad_norm": 1.3484407663345337, "learning_rate": 9.888474385185122e-05, "loss": 0.0762, "step": 500 }, { "epoch": 0.46617915904936014, "grad_norm": 0.7931845784187317, "learning_rate": 9.880398493969128e-05, "loss": 0.0629, "step": 510 }, { "epoch": 0.4753199268738574, "grad_norm": 0.7555088400840759, "learning_rate": 9.872043866849156e-05, "loss": 0.072, "step": 520 }, { "epoch": 0.4844606946983547, "grad_norm": 0.9508174061775208, "learning_rate": 9.863410980985943e-05, "loss": 0.0688, "step": 530 }, { "epoch": 0.4936014625228519, "grad_norm": 0.9366075992584229, "learning_rate": 9.854500329432514e-05, "loss": 0.0635, "step": 540 }, { "epoch": 0.5027422303473492, "grad_norm": 0.7485318779945374, "learning_rate": 9.845312421106026e-05, "loss": 0.0612, "step": 550 }, { "epoch": 0.5118829981718465, "grad_norm": 1.1440995931625366, "learning_rate": 9.835847780758695e-05, "loss": 0.0657, "step": 560 }, { "epoch": 0.5210237659963437, "grad_norm": 0.6450702548027039, "learning_rate": 9.826106948947831e-05, "loss": 0.0756, "step": 570 }, { "epoch": 0.5301645338208409, "grad_norm": 0.9916141033172607, "learning_rate": 9.816090482004966e-05, "loss": 0.0689, "step": 580 }, { "epoch": 0.5393053016453382, "grad_norm": 0.9260968565940857, "learning_rate": 9.805798952004074e-05, "loss": 0.0741, "step": 590 }, { "epoch": 0.5484460694698354, "grad_norm": 1.099038004875183, "learning_rate": 9.795232946728902e-05, "loss": 0.0664, "step": 600 }, { "epoch": 0.5575868372943327, "grad_norm": 0.8096833825111389, "learning_rate": 9.784393069639402e-05, "loss": 0.0726, "step": 610 }, { "epoch": 0.56672760511883, "grad_norm": 1.4272693395614624, "learning_rate": 9.77327993983726e-05, "loss": 0.0733, "step": 620 }, { "epoch": 0.5758683729433273, "grad_norm": 1.050479769706726, "learning_rate": 9.76189419203054e-05, "loss": 0.067, "step": 630 }, { "epoch": 0.5850091407678245, "grad_norm": 0.8118894100189209, "learning_rate": 9.750236476497435e-05, "loss": 0.0709, "step": 640 }, { "epoch": 0.5941499085923218, "grad_norm": 0.5546473264694214, "learning_rate": 9.738307459049121e-05, "loss": 0.0536, "step": 650 }, { "epoch": 0.603290676416819, "grad_norm": 0.9993944764137268, "learning_rate": 9.726107820991741e-05, "loss": 0.0812, "step": 660 }, { "epoch": 0.6124314442413162, "grad_norm": 1.0287402868270874, "learning_rate": 9.713638259087486e-05, "loss": 0.0695, "step": 670 }, { "epoch": 0.6215722120658135, "grad_norm": 0.9676907062530518, "learning_rate": 9.7008994855148e-05, "loss": 0.0649, "step": 680 }, { "epoch": 0.6307129798903108, "grad_norm": 0.886712908744812, "learning_rate": 9.687892227827707e-05, "loss": 0.0786, "step": 690 }, { "epoch": 0.6398537477148081, "grad_norm": 0.5730246305465698, "learning_rate": 9.674617228914259e-05, "loss": 0.058, "step": 700 }, { "epoch": 0.6489945155393053, "grad_norm": 1.1880249977111816, "learning_rate": 9.661075246954106e-05, "loss": 0.0701, "step": 710 }, { "epoch": 0.6581352833638026, "grad_norm": 0.8326805233955383, "learning_rate": 9.647267055375195e-05, "loss": 0.0683, "step": 720 }, { "epoch": 0.6672760511882998, "grad_norm": 0.7834450006484985, "learning_rate": 9.633193442809596e-05, "loss": 0.0625, "step": 730 }, { "epoch": 0.676416819012797, "grad_norm": 0.8641069531440735, "learning_rate": 9.618855213048457e-05, "loss": 0.0915, "step": 740 }, { "epoch": 0.6855575868372943, "grad_norm": 1.1982171535491943, "learning_rate": 9.604253184996103e-05, "loss": 0.0744, "step": 750 }, { "epoch": 0.6946983546617916, "grad_norm": 0.9463611841201782, "learning_rate": 9.589388192623266e-05, "loss": 0.0578, "step": 760 }, { "epoch": 0.7038391224862889, "grad_norm": 0.7067205905914307, "learning_rate": 9.574261084919445e-05, "loss": 0.0774, "step": 770 }, { "epoch": 0.7129798903107861, "grad_norm": 0.9819389581680298, "learning_rate": 9.558872725844426e-05, "loss": 0.0636, "step": 780 }, { "epoch": 0.7221206581352834, "grad_norm": 0.7785099148750305, "learning_rate": 9.543223994278938e-05, "loss": 0.0649, "step": 790 }, { "epoch": 0.7312614259597806, "grad_norm": 0.6871359944343567, "learning_rate": 9.527315783974454e-05, "loss": 0.0711, "step": 800 }, { "epoch": 0.7404021937842779, "grad_norm": 0.9093775153160095, "learning_rate": 9.511149003502145e-05, "loss": 0.0866, "step": 810 }, { "epoch": 0.7495429616087751, "grad_norm": 0.8019959926605225, "learning_rate": 9.494724576200989e-05, "loss": 0.0688, "step": 820 }, { "epoch": 0.7586837294332724, "grad_norm": 0.6367955803871155, "learning_rate": 9.478043440125043e-05, "loss": 0.0499, "step": 830 }, { "epoch": 0.7678244972577697, "grad_norm": 1.0420273542404175, "learning_rate": 9.461106547989858e-05, "loss": 0.0696, "step": 840 }, { "epoch": 0.7769652650822669, "grad_norm": 1.0322197675704956, "learning_rate": 9.44391486711807e-05, "loss": 0.078, "step": 850 }, { "epoch": 0.7861060329067642, "grad_norm": 0.6654338836669922, "learning_rate": 9.426469379384153e-05, "loss": 0.0608, "step": 860 }, { "epoch": 0.7952468007312614, "grad_norm": 0.6521769762039185, "learning_rate": 9.408771081158346e-05, "loss": 0.0542, "step": 870 }, { "epoch": 0.8043875685557587, "grad_norm": 0.588214099407196, "learning_rate": 9.390820983249733e-05, "loss": 0.0558, "step": 880 }, { "epoch": 0.8135283363802559, "grad_norm": 1.0544887781143188, "learning_rate": 9.372620110848528e-05, "loss": 0.0558, "step": 890 }, { "epoch": 0.8226691042047533, "grad_norm": 0.4711853265762329, "learning_rate": 9.354169503467515e-05, "loss": 0.0617, "step": 900 }, { "epoch": 0.8318098720292505, "grad_norm": 0.8516566157341003, "learning_rate": 9.335470214882678e-05, "loss": 0.0577, "step": 910 }, { "epoch": 0.8409506398537477, "grad_norm": 0.9411584734916687, "learning_rate": 9.316523313073018e-05, "loss": 0.0578, "step": 920 }, { "epoch": 0.850091407678245, "grad_norm": 0.6325905919075012, "learning_rate": 9.297329880159556e-05, "loss": 0.0549, "step": 930 }, { "epoch": 0.8592321755027422, "grad_norm": 0.9557574391365051, "learning_rate": 9.277891012343529e-05, "loss": 0.0542, "step": 940 }, { "epoch": 0.8683729433272395, "grad_norm": 0.9808653593063354, "learning_rate": 9.258207819843787e-05, "loss": 0.0665, "step": 950 }, { "epoch": 0.8775137111517367, "grad_norm": 0.924566388130188, "learning_rate": 9.23828142683338e-05, "loss": 0.0786, "step": 960 }, { "epoch": 0.886654478976234, "grad_norm": 1.108108401298523, "learning_rate": 9.21811297137535e-05, "loss": 0.0766, "step": 970 }, { "epoch": 0.8957952468007313, "grad_norm": 1.2968391180038452, "learning_rate": 9.197703605357739e-05, "loss": 0.0605, "step": 980 }, { "epoch": 0.9049360146252285, "grad_norm": 0.9167346954345703, "learning_rate": 9.177054494427799e-05, "loss": 0.0504, "step": 990 }, { "epoch": 0.9140767824497258, "grad_norm": 0.8108971118927002, "learning_rate": 9.156166817925415e-05, "loss": 0.0526, "step": 1000 }, { "epoch": 0.923217550274223, "grad_norm": 0.8036976456642151, "learning_rate": 9.135041768815751e-05, "loss": 0.0469, "step": 1010 }, { "epoch": 0.9323583180987203, "grad_norm": 0.9675208926200867, "learning_rate": 9.113680553621116e-05, "loss": 0.054, "step": 1020 }, { "epoch": 0.9414990859232175, "grad_norm": 0.7497916221618652, "learning_rate": 9.092084392352056e-05, "loss": 0.0649, "step": 1030 }, { "epoch": 0.9506398537477148, "grad_norm": 0.5699302554130554, "learning_rate": 9.070254518437673e-05, "loss": 0.0505, "step": 1040 }, { "epoch": 0.9597806215722121, "grad_norm": 0.539658784866333, "learning_rate": 9.048192178655181e-05, "loss": 0.0435, "step": 1050 }, { "epoch": 0.9689213893967094, "grad_norm": 0.6254842877388, "learning_rate": 9.025898633058699e-05, "loss": 0.0541, "step": 1060 }, { "epoch": 0.9780621572212066, "grad_norm": 0.6805806756019592, "learning_rate": 9.003375154907284e-05, "loss": 0.0714, "step": 1070 }, { "epoch": 0.9872029250457038, "grad_norm": 0.6484934687614441, "learning_rate": 8.980623030592212e-05, "loss": 0.065, "step": 1080 }, { "epoch": 0.9963436928702011, "grad_norm": 0.5721938014030457, "learning_rate": 8.957643559563507e-05, "loss": 0.0735, "step": 1090 }, { "epoch": 1.0054844606946984, "grad_norm": 1.0545032024383545, "learning_rate": 8.934438054255723e-05, "loss": 0.0585, "step": 1100 }, { "epoch": 1.0146252285191957, "grad_norm": 0.5121464729309082, "learning_rate": 8.91100784001299e-05, "loss": 0.0545, "step": 1110 }, { "epoch": 1.023765996343693, "grad_norm": 0.890046238899231, "learning_rate": 8.887354255013323e-05, "loss": 0.063, "step": 1120 }, { "epoch": 1.0329067641681902, "grad_norm": 0.4641508162021637, "learning_rate": 8.863478650192181e-05, "loss": 0.0622, "step": 1130 }, { "epoch": 1.0420475319926874, "grad_norm": 0.7307084202766418, "learning_rate": 8.839382389165323e-05, "loss": 0.0586, "step": 1140 }, { "epoch": 1.0511882998171846, "grad_norm": 0.7155199646949768, "learning_rate": 8.815066848150922e-05, "loss": 0.052, "step": 1150 }, { "epoch": 1.0603290676416819, "grad_norm": 0.675208568572998, "learning_rate": 8.790533415890967e-05, "loss": 0.057, "step": 1160 }, { "epoch": 1.0694698354661791, "grad_norm": 0.7103818655014038, "learning_rate": 8.765783493571942e-05, "loss": 0.0517, "step": 1170 }, { "epoch": 1.0786106032906764, "grad_norm": 0.8737835884094238, "learning_rate": 8.740818494744809e-05, "loss": 0.0376, "step": 1180 }, { "epoch": 1.0877513711151736, "grad_norm": 1.203967809677124, "learning_rate": 8.715639845244263e-05, "loss": 0.0538, "step": 1190 }, { "epoch": 1.0968921389396709, "grad_norm": 0.6871981620788574, "learning_rate": 8.690248983107313e-05, "loss": 0.0586, "step": 1200 }, { "epoch": 1.106032906764168, "grad_norm": 1.3813148736953735, "learning_rate": 8.664647358491134e-05, "loss": 0.0486, "step": 1210 }, { "epoch": 1.1151736745886653, "grad_norm": 0.9246250987052917, "learning_rate": 8.638836433590257e-05, "loss": 0.0463, "step": 1220 }, { "epoch": 1.1243144424131628, "grad_norm": 1.0106674432754517, "learning_rate": 8.61281768255305e-05, "loss": 0.0466, "step": 1230 }, { "epoch": 1.13345521023766, "grad_norm": 0.7251119017601013, "learning_rate": 8.586592591397533e-05, "loss": 0.0575, "step": 1240 }, { "epoch": 1.1425959780621573, "grad_norm": 0.8430503606796265, "learning_rate": 8.560162657926494e-05, "loss": 0.0427, "step": 1250 }, { "epoch": 1.1517367458866545, "grad_norm": 0.739719808101654, "learning_rate": 8.533529391641959e-05, "loss": 0.0425, "step": 1260 }, { "epoch": 1.1608775137111518, "grad_norm": 0.8659994006156921, "learning_rate": 8.506694313658969e-05, "loss": 0.048, "step": 1270 }, { "epoch": 1.170018281535649, "grad_norm": 0.54444819688797, "learning_rate": 8.479658956618704e-05, "loss": 0.0572, "step": 1280 }, { "epoch": 1.1791590493601463, "grad_norm": 1.301878809928894, "learning_rate": 8.452424864600956e-05, "loss": 0.0611, "step": 1290 }, { "epoch": 1.1882998171846435, "grad_norm": 0.42106685042381287, "learning_rate": 8.424993593035936e-05, "loss": 0.0446, "step": 1300 }, { "epoch": 1.1974405850091407, "grad_norm": 0.6645190715789795, "learning_rate": 8.397366708615439e-05, "loss": 0.0618, "step": 1310 }, { "epoch": 1.206581352833638, "grad_norm": 0.6044383645057678, "learning_rate": 8.369545789203363e-05, "loss": 0.0408, "step": 1320 }, { "epoch": 1.2157221206581352, "grad_norm": 0.5589725375175476, "learning_rate": 8.341532423745599e-05, "loss": 0.059, "step": 1330 }, { "epoch": 1.2248628884826325, "grad_norm": 0.9428887963294983, "learning_rate": 8.313328212179274e-05, "loss": 0.0573, "step": 1340 }, { "epoch": 1.2340036563071297, "grad_norm": 0.6069498062133789, "learning_rate": 8.284934765341374e-05, "loss": 0.0539, "step": 1350 }, { "epoch": 1.2431444241316272, "grad_norm": 0.8920032978057861, "learning_rate": 8.256353704876748e-05, "loss": 0.0699, "step": 1360 }, { "epoch": 1.2522851919561244, "grad_norm": 1.1463648080825806, "learning_rate": 8.227586663145485e-05, "loss": 0.0442, "step": 1370 }, { "epoch": 1.2614259597806217, "grad_norm": 0.5741071701049805, "learning_rate": 8.198635283129688e-05, "loss": 0.053, "step": 1380 }, { "epoch": 1.270566727605119, "grad_norm": 0.6575504541397095, "learning_rate": 8.169501218339638e-05, "loss": 0.05, "step": 1390 }, { "epoch": 1.2797074954296161, "grad_norm": 0.9514917135238647, "learning_rate": 8.140186132719352e-05, "loss": 0.0578, "step": 1400 }, { "epoch": 1.2888482632541134, "grad_norm": 0.4858059585094452, "learning_rate": 8.110691700551558e-05, "loss": 0.0478, "step": 1410 }, { "epoch": 1.2979890310786106, "grad_norm": 0.6736142635345459, "learning_rate": 8.081019606362063e-05, "loss": 0.051, "step": 1420 }, { "epoch": 1.3071297989031079, "grad_norm": 0.5692024827003479, "learning_rate": 8.051171544823543e-05, "loss": 0.0457, "step": 1430 }, { "epoch": 1.3162705667276051, "grad_norm": 0.7224279642105103, "learning_rate": 8.02114922065877e-05, "loss": 0.0537, "step": 1440 }, { "epoch": 1.3254113345521024, "grad_norm": 0.7441707849502563, "learning_rate": 7.990954348543227e-05, "loss": 0.0441, "step": 1450 }, { "epoch": 1.3345521023765996, "grad_norm": 0.8292146325111389, "learning_rate": 7.960588653007191e-05, "loss": 0.0463, "step": 1460 }, { "epoch": 1.3436928702010968, "grad_norm": 0.5339452028274536, "learning_rate": 7.930053868337242e-05, "loss": 0.0536, "step": 1470 }, { "epoch": 1.352833638025594, "grad_norm": 0.47000357508659363, "learning_rate": 7.899351738477201e-05, "loss": 0.0465, "step": 1480 }, { "epoch": 1.3619744058500913, "grad_norm": 0.4280283451080322, "learning_rate": 7.868484016928537e-05, "loss": 0.0457, "step": 1490 }, { "epoch": 1.3711151736745886, "grad_norm": 1.0179851055145264, "learning_rate": 7.837452466650213e-05, "loss": 0.0474, "step": 1500 }, { "epoch": 1.3802559414990858, "grad_norm": 0.6735445261001587, "learning_rate": 7.806258859957998e-05, "loss": 0.055, "step": 1510 }, { "epoch": 1.389396709323583, "grad_norm": 1.0807559490203857, "learning_rate": 7.774904978423247e-05, "loss": 0.0522, "step": 1520 }, { "epoch": 1.3985374771480805, "grad_norm": 0.5602183938026428, "learning_rate": 7.743392612771152e-05, "loss": 0.0386, "step": 1530 }, { "epoch": 1.4076782449725778, "grad_norm": 0.887178361415863, "learning_rate": 7.711723562778456e-05, "loss": 0.0476, "step": 1540 }, { "epoch": 1.416819012797075, "grad_norm": 1.0711039304733276, "learning_rate": 7.679899637170677e-05, "loss": 0.0444, "step": 1550 }, { "epoch": 1.4259597806215722, "grad_norm": 1.0682564973831177, "learning_rate": 7.647922653518796e-05, "loss": 0.0522, "step": 1560 }, { "epoch": 1.4351005484460695, "grad_norm": 0.5387830138206482, "learning_rate": 7.615794438135444e-05, "loss": 0.0555, "step": 1570 }, { "epoch": 1.4442413162705667, "grad_norm": 0.7840970754623413, "learning_rate": 7.583516825970616e-05, "loss": 0.0483, "step": 1580 }, { "epoch": 1.453382084095064, "grad_norm": 0.9905489683151245, "learning_rate": 7.551091660506847e-05, "loss": 0.0471, "step": 1590 }, { "epoch": 1.4625228519195612, "grad_norm": 0.5009207129478455, "learning_rate": 7.518520793653939e-05, "loss": 0.0463, "step": 1600 }, { "epoch": 1.4716636197440585, "grad_norm": 0.5666080713272095, "learning_rate": 7.485806085643188e-05, "loss": 0.0438, "step": 1610 }, { "epoch": 1.4808043875685557, "grad_norm": 0.8383708596229553, "learning_rate": 7.452949404921142e-05, "loss": 0.0504, "step": 1620 }, { "epoch": 1.489945155393053, "grad_norm": 0.6216220855712891, "learning_rate": 7.419952628042881e-05, "loss": 0.0464, "step": 1630 }, { "epoch": 1.4990859232175504, "grad_norm": 0.5239822864532471, "learning_rate": 7.386817639564851e-05, "loss": 0.0431, "step": 1640 }, { "epoch": 1.5082266910420477, "grad_norm": 1.0513837337493896, "learning_rate": 7.353546331937224e-05, "loss": 0.0571, "step": 1650 }, { "epoch": 1.517367458866545, "grad_norm": 0.7348514199256897, "learning_rate": 7.320140605395815e-05, "loss": 0.0526, "step": 1660 }, { "epoch": 1.5265082266910421, "grad_norm": 0.7277615070343018, "learning_rate": 7.286602367853552e-05, "loss": 0.0379, "step": 1670 }, { "epoch": 1.5356489945155394, "grad_norm": 0.719916045665741, "learning_rate": 7.252933534791511e-05, "loss": 0.0411, "step": 1680 }, { "epoch": 1.5447897623400366, "grad_norm": 0.8909785747528076, "learning_rate": 7.219136029149515e-05, "loss": 0.0526, "step": 1690 }, { "epoch": 1.5539305301645339, "grad_norm": 0.9339496493339539, "learning_rate": 7.185211781216314e-05, "loss": 0.0699, "step": 1700 }, { "epoch": 1.563071297989031, "grad_norm": 0.7735287547111511, "learning_rate": 7.151162728519324e-05, "loss": 0.0571, "step": 1710 }, { "epoch": 1.5722120658135283, "grad_norm": 0.8301142454147339, "learning_rate": 7.11699081571399e-05, "loss": 0.0484, "step": 1720 }, { "epoch": 1.5813528336380256, "grad_norm": 0.7117460370063782, "learning_rate": 7.082697994472705e-05, "loss": 0.0505, "step": 1730 }, { "epoch": 1.5904936014625228, "grad_norm": 0.39248979091644287, "learning_rate": 7.048286223373342e-05, "loss": 0.0329, "step": 1740 }, { "epoch": 1.59963436928702, "grad_norm": 0.6386615037918091, "learning_rate": 7.013757467787412e-05, "loss": 0.0382, "step": 1750 }, { "epoch": 1.6087751371115173, "grad_norm": 0.8316029906272888, "learning_rate": 6.979113699767789e-05, "loss": 0.0443, "step": 1760 }, { "epoch": 1.6179159049360146, "grad_norm": 0.6195833086967468, "learning_rate": 6.9443568979361e-05, "loss": 0.0468, "step": 1770 }, { "epoch": 1.6270566727605118, "grad_norm": 0.5122663974761963, "learning_rate": 6.909489047369706e-05, "loss": 0.0486, "step": 1780 }, { "epoch": 1.636197440585009, "grad_norm": 0.7890117168426514, "learning_rate": 6.87451213948834e-05, "loss": 0.0368, "step": 1790 }, { "epoch": 1.6453382084095063, "grad_norm": 1.125509262084961, "learning_rate": 6.839428171940357e-05, "loss": 0.0485, "step": 1800 }, { "epoch": 1.6544789762340035, "grad_norm": 0.6382961869239807, "learning_rate": 6.804239148488651e-05, "loss": 0.0551, "step": 1810 }, { "epoch": 1.6636197440585008, "grad_norm": 0.5489874482154846, "learning_rate": 6.768947078896212e-05, "loss": 0.0543, "step": 1820 }, { "epoch": 1.672760511882998, "grad_norm": 0.5988185405731201, "learning_rate": 6.733553978811338e-05, "loss": 0.0408, "step": 1830 }, { "epoch": 1.6819012797074955, "grad_norm": 0.5746557116508484, "learning_rate": 6.698061869652518e-05, "loss": 0.052, "step": 1840 }, { "epoch": 1.6910420475319927, "grad_norm": 0.4615539014339447, "learning_rate": 6.662472778492979e-05, "loss": 0.0356, "step": 1850 }, { "epoch": 1.70018281535649, "grad_norm": 0.5085066556930542, "learning_rate": 6.626788737944916e-05, "loss": 0.0477, "step": 1860 }, { "epoch": 1.7093235831809872, "grad_norm": 0.5613322257995605, "learning_rate": 6.591011786043404e-05, "loss": 0.0394, "step": 1870 }, { "epoch": 1.7184643510054844, "grad_norm": 0.37369537353515625, "learning_rate": 6.55514396612999e-05, "loss": 0.0475, "step": 1880 }, { "epoch": 1.7276051188299817, "grad_norm": 0.6342534422874451, "learning_rate": 6.519187326736007e-05, "loss": 0.0451, "step": 1890 }, { "epoch": 1.736745886654479, "grad_norm": 0.7583989500999451, "learning_rate": 6.48314392146556e-05, "loss": 0.0468, "step": 1900 }, { "epoch": 1.7458866544789764, "grad_norm": 0.6060802340507507, "learning_rate": 6.447015808878243e-05, "loss": 0.0581, "step": 1910 }, { "epoch": 1.7550274223034736, "grad_norm": 0.5234028100967407, "learning_rate": 6.41080505237157e-05, "loss": 0.0409, "step": 1920 }, { "epoch": 1.7641681901279709, "grad_norm": 0.7409939169883728, "learning_rate": 6.374513720063126e-05, "loss": 0.0512, "step": 1930 }, { "epoch": 1.7733089579524681, "grad_norm": 0.5225390195846558, "learning_rate": 6.338143884672451e-05, "loss": 0.0518, "step": 1940 }, { "epoch": 1.7824497257769654, "grad_norm": 0.45462241768836975, "learning_rate": 6.301697623402658e-05, "loss": 0.0432, "step": 1950 }, { "epoch": 1.7915904936014626, "grad_norm": 0.3406662344932556, "learning_rate": 6.265177017821799e-05, "loss": 0.0507, "step": 1960 }, { "epoch": 1.8007312614259599, "grad_norm": 0.8838239908218384, "learning_rate": 6.228584153743975e-05, "loss": 0.0474, "step": 1970 }, { "epoch": 1.809872029250457, "grad_norm": 0.6049211025238037, "learning_rate": 6.19192112111022e-05, "loss": 0.0435, "step": 1980 }, { "epoch": 1.8190127970749543, "grad_norm": 1.0864945650100708, "learning_rate": 6.155190013869119e-05, "loss": 0.0492, "step": 1990 }, { "epoch": 1.8281535648994516, "grad_norm": 0.734801173210144, "learning_rate": 6.118392929857231e-05, "loss": 0.0477, "step": 2000 }, { "epoch": 1.8372943327239488, "grad_norm": 0.9690449237823486, "learning_rate": 6.08153197067927e-05, "loss": 0.0489, "step": 2010 }, { "epoch": 1.846435100548446, "grad_norm": 0.6723517179489136, "learning_rate": 6.044609241588073e-05, "loss": 0.0383, "step": 2020 }, { "epoch": 1.8555758683729433, "grad_norm": 1.0849159955978394, "learning_rate": 6.0076268513643644e-05, "loss": 0.0458, "step": 2030 }, { "epoch": 1.8647166361974405, "grad_norm": 0.6273264288902283, "learning_rate": 5.9705869121963165e-05, "loss": 0.0374, "step": 2040 }, { "epoch": 1.8738574040219378, "grad_norm": 0.9236069321632385, "learning_rate": 5.933491539558913e-05, "loss": 0.0464, "step": 2050 }, { "epoch": 1.882998171846435, "grad_norm": 0.5545753240585327, "learning_rate": 5.8963428520931275e-05, "loss": 0.045, "step": 2060 }, { "epoch": 1.8921389396709323, "grad_norm": 0.43973255157470703, "learning_rate": 5.8591429714849263e-05, "loss": 0.0474, "step": 2070 }, { "epoch": 1.9012797074954295, "grad_norm": 0.4769677519798279, "learning_rate": 5.821894022344083e-05, "loss": 0.044, "step": 2080 }, { "epoch": 1.9104204753199268, "grad_norm": 0.4825699031352997, "learning_rate": 5.7845981320828436e-05, "loss": 0.0379, "step": 2090 }, { "epoch": 1.919561243144424, "grad_norm": 0.4520387649536133, "learning_rate": 5.747257430794415e-05, "loss": 0.0511, "step": 2100 }, { "epoch": 1.9287020109689212, "grad_norm": 1.0386220216751099, "learning_rate": 5.709874051131312e-05, "loss": 0.0451, "step": 2110 }, { "epoch": 1.9378427787934185, "grad_norm": 0.5633272528648376, "learning_rate": 5.672450128183561e-05, "loss": 0.047, "step": 2120 }, { "epoch": 1.946983546617916, "grad_norm": 0.5927779674530029, "learning_rate": 5.63498779935674e-05, "loss": 0.0408, "step": 2130 }, { "epoch": 1.9561243144424132, "grad_norm": 0.4667109549045563, "learning_rate": 5.597489204249925e-05, "loss": 0.041, "step": 2140 }, { "epoch": 1.9652650822669104, "grad_norm": 0.7210867404937744, "learning_rate": 5.559956484533477e-05, "loss": 0.0407, "step": 2150 }, { "epoch": 1.9744058500914077, "grad_norm": 0.6026551723480225, "learning_rate": 5.522391783826726e-05, "loss": 0.0452, "step": 2160 }, { "epoch": 1.983546617915905, "grad_norm": 0.559966504573822, "learning_rate": 5.484797247575547e-05, "loss": 0.0394, "step": 2170 }, { "epoch": 1.9926873857404022, "grad_norm": 0.6461185812950134, "learning_rate": 5.4471750229298215e-05, "loss": 0.0366, "step": 2180 }, { "epoch": 2.0018281535648996, "grad_norm": 0.3840448558330536, "learning_rate": 5.409527258620806e-05, "loss": 0.0326, "step": 2190 }, { "epoch": 2.010968921389397, "grad_norm": 0.9953444600105286, "learning_rate": 5.3718561048384166e-05, "loss": 0.0353, "step": 2200 }, { "epoch": 2.020109689213894, "grad_norm": 0.6786929368972778, "learning_rate": 5.334163713108417e-05, "loss": 0.046, "step": 2210 }, { "epoch": 2.0292504570383914, "grad_norm": 0.4201532304286957, "learning_rate": 5.29645223616954e-05, "loss": 0.035, "step": 2220 }, { "epoch": 2.0383912248628886, "grad_norm": 0.6718032956123352, "learning_rate": 5.258723827850547e-05, "loss": 0.0365, "step": 2230 }, { "epoch": 2.047531992687386, "grad_norm": 0.6162596344947815, "learning_rate": 5.2209806429471955e-05, "loss": 0.0455, "step": 2240 }, { "epoch": 2.056672760511883, "grad_norm": 0.6207318305969238, "learning_rate": 5.183224837099192e-05, "loss": 0.035, "step": 2250 }, { "epoch": 2.0658135283363803, "grad_norm": 0.6675350666046143, "learning_rate": 5.145458566667065e-05, "loss": 0.0339, "step": 2260 }, { "epoch": 2.0749542961608776, "grad_norm": 0.6197575330734253, "learning_rate": 5.107683988609009e-05, "loss": 0.0401, "step": 2270 }, { "epoch": 2.084095063985375, "grad_norm": 0.7717021107673645, "learning_rate": 5.069903260357696e-05, "loss": 0.0454, "step": 2280 }, { "epoch": 2.093235831809872, "grad_norm": 0.44513338804244995, "learning_rate": 5.0321185396970564e-05, "loss": 0.0445, "step": 2290 }, { "epoch": 2.1023765996343693, "grad_norm": 0.47282537817955017, "learning_rate": 4.994331984639039e-05, "loss": 0.044, "step": 2300 }, { "epoch": 2.1115173674588665, "grad_norm": 0.506184995174408, "learning_rate": 4.956545753300363e-05, "loss": 0.0391, "step": 2310 }, { "epoch": 2.1206581352833638, "grad_norm": 0.5166071653366089, "learning_rate": 4.9187620037792575e-05, "loss": 0.0402, "step": 2320 }, { "epoch": 2.129798903107861, "grad_norm": 0.6371522545814514, "learning_rate": 4.880982894032208e-05, "loss": 0.0385, "step": 2330 }, { "epoch": 2.1389396709323583, "grad_norm": 0.6055366396903992, "learning_rate": 4.8432105817507066e-05, "loss": 0.0336, "step": 2340 }, { "epoch": 2.1480804387568555, "grad_norm": 0.28509601950645447, "learning_rate": 4.8054472242380174e-05, "loss": 0.0326, "step": 2350 }, { "epoch": 2.1572212065813527, "grad_norm": 0.7012027502059937, "learning_rate": 4.767694978285972e-05, "loss": 0.0439, "step": 2360 }, { "epoch": 2.16636197440585, "grad_norm": 0.3986015021800995, "learning_rate": 4.729956000051779e-05, "loss": 0.0343, "step": 2370 }, { "epoch": 2.1755027422303472, "grad_norm": 0.5725045800209045, "learning_rate": 4.6922324449348826e-05, "loss": 0.0352, "step": 2380 }, { "epoch": 2.1846435100548445, "grad_norm": 0.8768548965454102, "learning_rate": 4.6545264674538683e-05, "loss": 0.0367, "step": 2390 }, { "epoch": 2.1937842778793417, "grad_norm": 0.5288705229759216, "learning_rate": 4.616840221123397e-05, "loss": 0.0358, "step": 2400 }, { "epoch": 2.202925045703839, "grad_norm": 0.546050488948822, "learning_rate": 4.579175858331219e-05, "loss": 0.0464, "step": 2410 }, { "epoch": 2.212065813528336, "grad_norm": 0.5290732383728027, "learning_rate": 4.5415355302152443e-05, "loss": 0.0359, "step": 2420 }, { "epoch": 2.2212065813528334, "grad_norm": 0.5098981261253357, "learning_rate": 4.5039213865406816e-05, "loss": 0.0377, "step": 2430 }, { "epoch": 2.2303473491773307, "grad_norm": 0.7096160650253296, "learning_rate": 4.466335575577254e-05, "loss": 0.0496, "step": 2440 }, { "epoch": 2.2394881170018284, "grad_norm": 0.4775889813899994, "learning_rate": 4.428780243976517e-05, "loss": 0.0392, "step": 2450 }, { "epoch": 2.2486288848263256, "grad_norm": 0.5815747380256653, "learning_rate": 4.3912575366492415e-05, "loss": 0.0401, "step": 2460 }, { "epoch": 2.257769652650823, "grad_norm": 0.5321184396743774, "learning_rate": 4.353769596642917e-05, "loss": 0.0486, "step": 2470 }, { "epoch": 2.26691042047532, "grad_norm": 0.6674132943153381, "learning_rate": 4.316318565019356e-05, "loss": 0.0373, "step": 2480 }, { "epoch": 2.2760511882998173, "grad_norm": 0.47233355045318604, "learning_rate": 4.278906580732411e-05, "loss": 0.0304, "step": 2490 }, { "epoch": 2.2851919561243146, "grad_norm": 0.5845870971679688, "learning_rate": 4.241535780505809e-05, "loss": 0.0395, "step": 2500 }, { "epoch": 2.294332723948812, "grad_norm": 0.6557559967041016, "learning_rate": 4.204208298711116e-05, "loss": 0.0256, "step": 2510 }, { "epoch": 2.303473491773309, "grad_norm": 0.7204177975654602, "learning_rate": 4.166926267245842e-05, "loss": 0.0396, "step": 2520 }, { "epoch": 2.3126142595978063, "grad_norm": 0.759373128414154, "learning_rate": 4.1296918154116706e-05, "loss": 0.0299, "step": 2530 }, { "epoch": 2.3217550274223036, "grad_norm": 0.5927398204803467, "learning_rate": 4.092507069792855e-05, "loss": 0.031, "step": 2540 }, { "epoch": 2.330895795246801, "grad_norm": 0.7468788623809814, "learning_rate": 4.055374154134764e-05, "loss": 0.0336, "step": 2550 }, { "epoch": 2.340036563071298, "grad_norm": 0.6309873461723328, "learning_rate": 4.0182951892225784e-05, "loss": 0.0298, "step": 2560 }, { "epoch": 2.3491773308957953, "grad_norm": 0.594954252243042, "learning_rate": 3.9812722927601724e-05, "loss": 0.037, "step": 2570 }, { "epoch": 2.3583180987202925, "grad_norm": 0.4929609000682831, "learning_rate": 3.944307579249168e-05, "loss": 0.0335, "step": 2580 }, { "epoch": 2.3674588665447898, "grad_norm": 0.5985615849494934, "learning_rate": 3.907403159868158e-05, "loss": 0.0276, "step": 2590 }, { "epoch": 2.376599634369287, "grad_norm": 0.762140154838562, "learning_rate": 3.870561142352137e-05, "loss": 0.0379, "step": 2600 }, { "epoch": 2.3857404021937842, "grad_norm": 0.7068864107131958, "learning_rate": 3.833783630872122e-05, "loss": 0.0256, "step": 2610 }, { "epoch": 2.3948811700182815, "grad_norm": 0.9355035424232483, "learning_rate": 3.797072725914974e-05, "loss": 0.0289, "step": 2620 }, { "epoch": 2.4040219378427787, "grad_norm": 0.5562652945518494, "learning_rate": 3.7604305241634284e-05, "loss": 0.0442, "step": 2630 }, { "epoch": 2.413162705667276, "grad_norm": 0.9532639384269714, "learning_rate": 3.723859118376357e-05, "loss": 0.0435, "step": 2640 }, { "epoch": 2.422303473491773, "grad_norm": 0.443515807390213, "learning_rate": 3.6873605972692303e-05, "loss": 0.029, "step": 2650 }, { "epoch": 2.4314442413162705, "grad_norm": 0.7470449209213257, "learning_rate": 3.650937045394835e-05, "loss": 0.0325, "step": 2660 }, { "epoch": 2.4405850091407677, "grad_norm": 0.9110393524169922, "learning_rate": 3.614590543024214e-05, "loss": 0.0282, "step": 2670 }, { "epoch": 2.449725776965265, "grad_norm": 0.4086456000804901, "learning_rate": 3.578323166027853e-05, "loss": 0.0418, "step": 2680 }, { "epoch": 2.458866544789762, "grad_norm": 0.38568276166915894, "learning_rate": 3.54213698575712e-05, "loss": 0.0406, "step": 2690 }, { "epoch": 2.4680073126142594, "grad_norm": 0.5677224397659302, "learning_rate": 3.506034068925973e-05, "loss": 0.0479, "step": 2700 }, { "epoch": 2.4771480804387567, "grad_norm": 0.7311156392097473, "learning_rate": 3.470016477492911e-05, "loss": 0.0278, "step": 2710 }, { "epoch": 2.4862888482632544, "grad_norm": 0.8331183791160583, "learning_rate": 3.434086268543213e-05, "loss": 0.0342, "step": 2720 }, { "epoch": 2.495429616087751, "grad_norm": 0.5231893062591553, "learning_rate": 3.398245494171457e-05, "loss": 0.0313, "step": 2730 }, { "epoch": 2.504570383912249, "grad_norm": 0.40512651205062866, "learning_rate": 3.362496201364308e-05, "loss": 0.027, "step": 2740 }, { "epoch": 2.5137111517367456, "grad_norm": 0.7198314070701599, "learning_rate": 3.326840431883613e-05, "loss": 0.0386, "step": 2750 }, { "epoch": 2.5228519195612433, "grad_norm": 0.5212320685386658, "learning_rate": 3.291280222149788e-05, "loss": 0.0404, "step": 2760 }, { "epoch": 2.53199268738574, "grad_norm": 0.3318735659122467, "learning_rate": 3.255817603125518e-05, "loss": 0.0358, "step": 2770 }, { "epoch": 2.541133455210238, "grad_norm": 0.4588581323623657, "learning_rate": 3.220454600199748e-05, "loss": 0.0318, "step": 2780 }, { "epoch": 2.550274223034735, "grad_norm": 0.47819092869758606, "learning_rate": 3.185193233072016e-05, "loss": 0.0342, "step": 2790 }, { "epoch": 2.5594149908592323, "grad_norm": 0.5531491041183472, "learning_rate": 3.150035515637102e-05, "loss": 0.0335, "step": 2800 }, { "epoch": 2.5685557586837295, "grad_norm": 0.6975427269935608, "learning_rate": 3.114983455870002e-05, "loss": 0.036, "step": 2810 }, { "epoch": 2.577696526508227, "grad_norm": 0.6512779593467712, "learning_rate": 3.080039055711248e-05, "loss": 0.0322, "step": 2820 }, { "epoch": 2.586837294332724, "grad_norm": 0.4503866136074066, "learning_rate": 3.045204310952576e-05, "loss": 0.0342, "step": 2830 }, { "epoch": 2.5959780621572213, "grad_norm": 0.4288831353187561, "learning_rate": 3.0104812111229307e-05, "loss": 0.0332, "step": 2840 }, { "epoch": 2.6051188299817185, "grad_norm": 0.5384790301322937, "learning_rate": 2.9758717393748392e-05, "loss": 0.0343, "step": 2850 }, { "epoch": 2.6142595978062158, "grad_norm": 0.44517871737480164, "learning_rate": 2.941377872371155e-05, "loss": 0.044, "step": 2860 }, { "epoch": 2.623400365630713, "grad_norm": 0.5115548372268677, "learning_rate": 2.9070015801721518e-05, "loss": 0.0343, "step": 2870 }, { "epoch": 2.6325411334552102, "grad_norm": 0.610212504863739, "learning_rate": 2.8727448261230128e-05, "loss": 0.033, "step": 2880 }, { "epoch": 2.6416819012797075, "grad_norm": 0.6658434867858887, "learning_rate": 2.8386095667417008e-05, "loss": 0.0368, "step": 2890 }, { "epoch": 2.6508226691042047, "grad_norm": 0.6432914733886719, "learning_rate": 2.8045977516072075e-05, "loss": 0.0322, "step": 2900 }, { "epoch": 2.659963436928702, "grad_norm": 0.5268338322639465, "learning_rate": 2.77071132324821e-05, "loss": 0.027, "step": 2910 }, { "epoch": 2.669104204753199, "grad_norm": 0.6469492316246033, "learning_rate": 2.7369522170321317e-05, "loss": 0.0292, "step": 2920 }, { "epoch": 2.6782449725776964, "grad_norm": 0.6078864932060242, "learning_rate": 2.7033223610545967e-05, "loss": 0.0245, "step": 2930 }, { "epoch": 2.6873857404021937, "grad_norm": 0.6451284289360046, "learning_rate": 2.6698236760293138e-05, "loss": 0.0309, "step": 2940 }, { "epoch": 2.696526508226691, "grad_norm": 0.6047863364219666, "learning_rate": 2.6364580751783872e-05, "loss": 0.0321, "step": 2950 }, { "epoch": 2.705667276051188, "grad_norm": 0.53343665599823, "learning_rate": 2.6032274641230325e-05, "loss": 0.0302, "step": 2960 }, { "epoch": 2.7148080438756854, "grad_norm": 0.9398259520530701, "learning_rate": 2.570133740774743e-05, "loss": 0.0387, "step": 2970 }, { "epoch": 2.7239488117001827, "grad_norm": 0.7435660362243652, "learning_rate": 2.537178795226905e-05, "loss": 0.0326, "step": 2980 }, { "epoch": 2.7330895795246803, "grad_norm": 0.8232837915420532, "learning_rate": 2.5043645096468333e-05, "loss": 0.0345, "step": 2990 }, { "epoch": 2.742230347349177, "grad_norm": 0.664114236831665, "learning_rate": 2.4716927581682832e-05, "loss": 0.0343, "step": 3000 }, { "epoch": 2.751371115173675, "grad_norm": 0.4456695318222046, "learning_rate": 2.4391654067844083e-05, "loss": 0.0262, "step": 3010 }, { "epoch": 2.7605118829981716, "grad_norm": 0.8034543991088867, "learning_rate": 2.4067843132411922e-05, "loss": 0.0269, "step": 3020 }, { "epoch": 2.7696526508226693, "grad_norm": 0.5905967950820923, "learning_rate": 2.374551326931338e-05, "loss": 0.029, "step": 3030 }, { "epoch": 2.778793418647166, "grad_norm": 0.5978665351867676, "learning_rate": 2.3424682887886573e-05, "loss": 0.0386, "step": 3040 }, { "epoch": 2.787934186471664, "grad_norm": 0.38325613737106323, "learning_rate": 2.3105370311829124e-05, "loss": 0.0327, "step": 3050 }, { "epoch": 2.797074954296161, "grad_norm": 0.8944759964942932, "learning_rate": 2.2787593778151722e-05, "loss": 0.0313, "step": 3060 }, { "epoch": 2.8062157221206583, "grad_norm": 0.47093465924263, "learning_rate": 2.247137143613653e-05, "loss": 0.0299, "step": 3070 }, { "epoch": 2.8153564899451555, "grad_norm": 0.5948256254196167, "learning_rate": 2.2156721346300657e-05, "loss": 0.0293, "step": 3080 }, { "epoch": 2.8244972577696528, "grad_norm": 0.48618465662002563, "learning_rate": 2.1843661479364587e-05, "loss": 0.0289, "step": 3090 }, { "epoch": 2.83363802559415, "grad_norm": 0.5810682773590088, "learning_rate": 2.1532209715225833e-05, "loss": 0.0264, "step": 3100 }, { "epoch": 2.8427787934186473, "grad_norm": 0.5437697172164917, "learning_rate": 2.1222383841937826e-05, "loss": 0.0325, "step": 3110 }, { "epoch": 2.8519195612431445, "grad_norm": 0.655825138092041, "learning_rate": 2.0914201554693884e-05, "loss": 0.0266, "step": 3120 }, { "epoch": 2.8610603290676417, "grad_norm": 0.4643450379371643, "learning_rate": 2.0607680454816612e-05, "loss": 0.0218, "step": 3130 }, { "epoch": 2.870201096892139, "grad_norm": 0.4340280592441559, "learning_rate": 2.0302838048752688e-05, "loss": 0.0275, "step": 3140 }, { "epoch": 2.8793418647166362, "grad_norm": 0.4632952809333801, "learning_rate": 1.9999691747072925e-05, "loss": 0.0372, "step": 3150 }, { "epoch": 2.8884826325411335, "grad_norm": 0.2842860519886017, "learning_rate": 1.9698258863477925e-05, "loss": 0.0284, "step": 3160 }, { "epoch": 2.8976234003656307, "grad_norm": 0.4197808802127838, "learning_rate": 1.9398556613809294e-05, "loss": 0.0449, "step": 3170 }, { "epoch": 2.906764168190128, "grad_norm": 0.5161760449409485, "learning_rate": 1.9100602115066295e-05, "loss": 0.0294, "step": 3180 }, { "epoch": 2.915904936014625, "grad_norm": 0.5271576642990112, "learning_rate": 1.880441238442827e-05, "loss": 0.0243, "step": 3190 }, { "epoch": 2.9250457038391224, "grad_norm": 0.5610905289649963, "learning_rate": 1.8510004338282816e-05, "loss": 0.0233, "step": 3200 }, { "epoch": 2.9341864716636197, "grad_norm": 0.4908517897129059, "learning_rate": 1.8217394791259483e-05, "loss": 0.0221, "step": 3210 }, { "epoch": 2.943327239488117, "grad_norm": 0.3808526396751404, "learning_rate": 1.792660045526954e-05, "loss": 0.0286, "step": 3220 }, { "epoch": 2.952468007312614, "grad_norm": 0.4248492419719696, "learning_rate": 1.763763793855149e-05, "loss": 0.0275, "step": 3230 }, { "epoch": 2.9616087751371114, "grad_norm": 0.3431572914123535, "learning_rate": 1.7350523744722493e-05, "loss": 0.0363, "step": 3240 }, { "epoch": 2.9707495429616086, "grad_norm": 0.5953290462493896, "learning_rate": 1.7065274271835756e-05, "loss": 0.033, "step": 3250 }, { "epoch": 2.979890310786106, "grad_norm": 0.852656364440918, "learning_rate": 1.6781905811444122e-05, "loss": 0.0343, "step": 3260 }, { "epoch": 2.989031078610603, "grad_norm": 0.7636460065841675, "learning_rate": 1.6500434547669425e-05, "loss": 0.0353, "step": 3270 }, { "epoch": 2.998171846435101, "grad_norm": 0.38765236735343933, "learning_rate": 1.6220876556278273e-05, "loss": 0.0236, "step": 3280 }, { "epoch": 3.0073126142595976, "grad_norm": 0.18574023246765137, "learning_rate": 1.5943247803763887e-05, "loss": 0.0256, "step": 3290 }, { "epoch": 3.016453382084095, "grad_norm": 0.42090004682540894, "learning_rate": 1.5667564146434172e-05, "loss": 0.0222, "step": 3300 }, { "epoch": 3.025594149908592, "grad_norm": 0.3421524167060852, "learning_rate": 1.539384132950618e-05, "loss": 0.0179, "step": 3310 }, { "epoch": 3.03473491773309, "grad_norm": 0.558494508266449, "learning_rate": 1.5122094986206758e-05, "loss": 0.0247, "step": 3320 }, { "epoch": 3.043875685557587, "grad_norm": 0.4569318890571594, "learning_rate": 1.485234063687973e-05, "loss": 0.0217, "step": 3330 }, { "epoch": 3.0530164533820843, "grad_norm": 1.014539122581482, "learning_rate": 1.4584593688099473e-05, "loss": 0.0247, "step": 3340 }, { "epoch": 3.0621572212065815, "grad_norm": 0.4421018362045288, "learning_rate": 1.4318869431791e-05, "loss": 0.0189, "step": 3350 }, { "epoch": 3.0712979890310788, "grad_norm": 0.45253804326057434, "learning_rate": 1.40551830443566e-05, "loss": 0.0276, "step": 3360 }, { "epoch": 3.080438756855576, "grad_norm": 0.3116052448749542, "learning_rate": 1.3793549585809023e-05, "loss": 0.0257, "step": 3370 }, { "epoch": 3.0895795246800732, "grad_norm": 0.5444056391716003, "learning_rate": 1.3533983998911353e-05, "loss": 0.0251, "step": 3380 }, { "epoch": 3.0987202925045705, "grad_norm": 0.2725338041782379, "learning_rate": 1.3276501108323664e-05, "loss": 0.026, "step": 3390 }, { "epoch": 3.1078610603290677, "grad_norm": 0.4318808317184448, "learning_rate": 1.3021115619756214e-05, "loss": 0.0226, "step": 3400 }, { "epoch": 3.117001828153565, "grad_norm": 0.5103743076324463, "learning_rate": 1.2767842119129597e-05, "loss": 0.0246, "step": 3410 }, { "epoch": 3.126142595978062, "grad_norm": 0.7406987547874451, "learning_rate": 1.2516695071741762e-05, "loss": 0.0216, "step": 3420 }, { "epoch": 3.1352833638025595, "grad_norm": 0.6604713797569275, "learning_rate": 1.2267688821441742e-05, "loss": 0.0345, "step": 3430 }, { "epoch": 3.1444241316270567, "grad_norm": 0.42933446168899536, "learning_rate": 1.2020837589810474e-05, "loss": 0.0226, "step": 3440 }, { "epoch": 3.153564899451554, "grad_norm": 0.4417478144168854, "learning_rate": 1.177615547534861e-05, "loss": 0.0216, "step": 3450 }, { "epoch": 3.162705667276051, "grad_norm": 0.41561007499694824, "learning_rate": 1.153365645267121e-05, "loss": 0.0314, "step": 3460 }, { "epoch": 3.1718464351005484, "grad_norm": 0.49010348320007324, "learning_rate": 1.1293354371709647e-05, "loss": 0.027, "step": 3470 }, { "epoch": 3.1809872029250457, "grad_norm": 0.6788333654403687, "learning_rate": 1.1055262956920647e-05, "loss": 0.0188, "step": 3480 }, { "epoch": 3.190127970749543, "grad_norm": 0.6015557646751404, "learning_rate": 1.081939580650233e-05, "loss": 0.0343, "step": 3490 }, { "epoch": 3.19926873857404, "grad_norm": 0.4312863051891327, "learning_rate": 1.0585766391617641e-05, "loss": 0.0199, "step": 3500 }, { "epoch": 3.2084095063985374, "grad_norm": 0.6486406326293945, "learning_rate": 1.0354388055624987e-05, "loss": 0.0218, "step": 3510 }, { "epoch": 3.2175502742230346, "grad_norm": 0.23588594794273376, "learning_rate": 1.0125274013316078e-05, "loss": 0.02, "step": 3520 }, { "epoch": 3.226691042047532, "grad_norm": 0.44049379229545593, "learning_rate": 9.898437350161238e-06, "loss": 0.018, "step": 3530 }, { "epoch": 3.235831809872029, "grad_norm": 0.3360377848148346, "learning_rate": 9.67389102156207e-06, "loss": 0.0195, "step": 3540 }, { "epoch": 3.2449725776965264, "grad_norm": 0.5655617117881775, "learning_rate": 9.451647852111472e-06, "loss": 0.0222, "step": 3550 }, { "epoch": 3.2541133455210236, "grad_norm": 0.599947452545166, "learning_rate": 9.231720534861187e-06, "loss": 0.0262, "step": 3560 }, { "epoch": 3.263254113345521, "grad_norm": 0.4483935236930847, "learning_rate": 9.014121630596944e-06, "loss": 0.0286, "step": 3570 }, { "epoch": 3.272394881170018, "grad_norm": 0.5928664803504944, "learning_rate": 8.798863567120952e-06, "loss": 0.0222, "step": 3580 }, { "epoch": 3.2815356489945158, "grad_norm": 0.42674168944358826, "learning_rate": 8.585958638542191e-06, "loss": 0.0173, "step": 3590 }, { "epoch": 3.2906764168190126, "grad_norm": 0.4651152491569519, "learning_rate": 8.375419004574215e-06, "loss": 0.0357, "step": 3600 }, { "epoch": 3.2998171846435103, "grad_norm": 0.6086673736572266, "learning_rate": 8.16725668984068e-06, "loss": 0.017, "step": 3610 }, { "epoch": 3.3089579524680075, "grad_norm": 0.2967171370983124, "learning_rate": 7.961483583188573e-06, "loss": 0.0315, "step": 3620 }, { "epoch": 3.3180987202925047, "grad_norm": 0.3449037969112396, "learning_rate": 7.758111437009235e-06, "loss": 0.0257, "step": 3630 }, { "epoch": 3.327239488117002, "grad_norm": 0.7032669186592102, "learning_rate": 7.557151866567086e-06, "loss": 0.0257, "step": 3640 }, { "epoch": 3.3363802559414992, "grad_norm": 0.5395845174789429, "learning_rate": 7.358616349336261e-06, "loss": 0.0273, "step": 3650 }, { "epoch": 3.3455210237659965, "grad_norm": 0.46958377957344055, "learning_rate": 7.162516224345101e-06, "loss": 0.0178, "step": 3660 }, { "epoch": 3.3546617915904937, "grad_norm": 0.4303223788738251, "learning_rate": 6.968862691528566e-06, "loss": 0.0202, "step": 3670 }, { "epoch": 3.363802559414991, "grad_norm": 0.6079144477844238, "learning_rate": 6.777666811088518e-06, "loss": 0.0197, "step": 3680 }, { "epoch": 3.372943327239488, "grad_norm": 0.38781973719596863, "learning_rate": 6.588939502862041e-06, "loss": 0.029, "step": 3690 }, { "epoch": 3.3820840950639854, "grad_norm": 0.3500356078147888, "learning_rate": 6.402691545697842e-06, "loss": 0.0239, "step": 3700 }, { "epoch": 3.3912248628884827, "grad_norm": 0.637479841709137, "learning_rate": 6.218933576840547e-06, "loss": 0.0208, "step": 3710 }, { "epoch": 3.40036563071298, "grad_norm": 0.414711594581604, "learning_rate": 6.037676091323208e-06, "loss": 0.0259, "step": 3720 }, { "epoch": 3.409506398537477, "grad_norm": 0.7444067001342773, "learning_rate": 5.858929441367939e-06, "loss": 0.033, "step": 3730 }, { "epoch": 3.4186471663619744, "grad_norm": 0.35689577460289, "learning_rate": 5.682703835794578e-06, "loss": 0.0263, "step": 3740 }, { "epoch": 3.4277879341864717, "grad_norm": 0.39409419894218445, "learning_rate": 5.5090093394376875e-06, "loss": 0.0278, "step": 3750 }, { "epoch": 3.436928702010969, "grad_norm": 0.4039927124977112, "learning_rate": 5.337855872571734e-06, "loss": 0.0208, "step": 3760 }, { "epoch": 3.446069469835466, "grad_norm": 0.6069719791412354, "learning_rate": 5.1692532103444494e-06, "loss": 0.0296, "step": 3770 }, { "epoch": 3.4552102376599634, "grad_norm": 0.4256989657878876, "learning_rate": 5.003210982218559e-06, "loss": 0.0191, "step": 3780 }, { "epoch": 3.4643510054844606, "grad_norm": 0.36178430914878845, "learning_rate": 4.839738671421862e-06, "loss": 0.0235, "step": 3790 }, { "epoch": 3.473491773308958, "grad_norm": 0.29829517006874084, "learning_rate": 4.678845614405541e-06, "loss": 0.0173, "step": 3800 }, { "epoch": 3.482632541133455, "grad_norm": 0.5337382555007935, "learning_rate": 4.5205410003109515e-06, "loss": 0.0265, "step": 3810 }, { "epoch": 3.4917733089579523, "grad_norm": 0.49504226446151733, "learning_rate": 4.364833870444834e-06, "loss": 0.0252, "step": 3820 }, { "epoch": 3.5009140767824496, "grad_norm": 0.6263189911842346, "learning_rate": 4.211733117762884e-06, "loss": 0.0357, "step": 3830 }, { "epoch": 3.510054844606947, "grad_norm": 0.5205623507499695, "learning_rate": 4.061247486361852e-06, "loss": 0.0246, "step": 3840 }, { "epoch": 3.519195612431444, "grad_norm": 0.3644954264163971, "learning_rate": 3.913385570980194e-06, "loss": 0.0198, "step": 3850 }, { "epoch": 3.5283363802559418, "grad_norm": 0.5801895260810852, "learning_rate": 3.7681558165071186e-06, "loss": 0.0179, "step": 3860 }, { "epoch": 3.5374771480804386, "grad_norm": 0.44283440709114075, "learning_rate": 3.625566517500334e-06, "loss": 0.0205, "step": 3870 }, { "epoch": 3.5466179159049362, "grad_norm": 0.6655595898628235, "learning_rate": 3.4856258177122735e-06, "loss": 0.018, "step": 3880 }, { "epoch": 3.555758683729433, "grad_norm": 0.5158295631408691, "learning_rate": 3.348341709625008e-06, "loss": 0.0195, "step": 3890 }, { "epoch": 3.5648994515539307, "grad_norm": 0.4960062503814697, "learning_rate": 3.2137220339937658e-06, "loss": 0.0211, "step": 3900 }, { "epoch": 3.5740402193784275, "grad_norm": 0.6395726799964905, "learning_rate": 3.081774479399102e-06, "loss": 0.0203, "step": 3910 }, { "epoch": 3.583180987202925, "grad_norm": 0.430395245552063, "learning_rate": 2.9525065818077956e-06, "loss": 0.0219, "step": 3920 }, { "epoch": 3.5923217550274225, "grad_norm": 0.5035272836685181, "learning_rate": 2.8259257241424253e-06, "loss": 0.0182, "step": 3930 }, { "epoch": 3.6014625228519197, "grad_norm": 0.5497689843177795, "learning_rate": 2.7020391358597464e-06, "loss": 0.029, "step": 3940 }, { "epoch": 3.610603290676417, "grad_norm": 0.5781487226486206, "learning_rate": 2.5808538925377433e-06, "loss": 0.0187, "step": 3950 }, { "epoch": 3.619744058500914, "grad_norm": 0.26463502645492554, "learning_rate": 2.462376915471548e-06, "loss": 0.0238, "step": 3960 }, { "epoch": 3.6288848263254114, "grad_norm": 0.5389289855957031, "learning_rate": 2.3466149712781384e-06, "loss": 0.0196, "step": 3970 }, { "epoch": 3.6380255941499087, "grad_norm": 0.22814074158668518, "learning_rate": 2.2335746715098747e-06, "loss": 0.0134, "step": 3980 }, { "epoch": 3.647166361974406, "grad_norm": 0.21715475618839264, "learning_rate": 2.1232624722768822e-06, "loss": 0.0172, "step": 3990 }, { "epoch": 3.656307129798903, "grad_norm": 0.8281272053718567, "learning_rate": 2.0156846738783264e-06, "loss": 0.0263, "step": 4000 }, { "epoch": 3.6654478976234004, "grad_norm": 0.6629803776741028, "learning_rate": 1.910847420442591e-06, "loss": 0.0244, "step": 4010 }, { "epoch": 3.6745886654478976, "grad_norm": 0.5557454228401184, "learning_rate": 1.8087566995763472e-06, "loss": 0.0252, "step": 4020 }, { "epoch": 3.683729433272395, "grad_norm": 0.5342206358909607, "learning_rate": 1.7094183420225952e-06, "loss": 0.0183, "step": 4030 }, { "epoch": 3.692870201096892, "grad_norm": 0.6376899480819702, "learning_rate": 1.6128380213276618e-06, "loss": 0.0267, "step": 4040 }, { "epoch": 3.7020109689213894, "grad_norm": 0.5518518686294556, "learning_rate": 1.5190212535171334e-06, "loss": 0.0187, "step": 4050 }, { "epoch": 3.7111517367458866, "grad_norm": 0.580907940864563, "learning_rate": 1.4279733967808352e-06, "loss": 0.0287, "step": 4060 }, { "epoch": 3.720292504570384, "grad_norm": 0.3451583683490753, "learning_rate": 1.3396996511668203e-06, "loss": 0.0227, "step": 4070 }, { "epoch": 3.729433272394881, "grad_norm": 0.5293395519256592, "learning_rate": 1.2542050582843578e-06, "loss": 0.0329, "step": 4080 }, { "epoch": 3.7385740402193783, "grad_norm": 0.4432827830314636, "learning_rate": 1.1714945010159905e-06, "loss": 0.0275, "step": 4090 }, { "epoch": 3.7477148080438756, "grad_norm": 0.4323786199092865, "learning_rate": 1.0915727032386747e-06, "loss": 0.022, "step": 4100 }, { "epoch": 3.756855575868373, "grad_norm": 0.19359391927719116, "learning_rate": 1.0144442295539625e-06, "loss": 0.0196, "step": 4110 }, { "epoch": 3.76599634369287, "grad_norm": 0.49890777468681335, "learning_rate": 9.401134850273053e-07, "loss": 0.0205, "step": 4120 }, { "epoch": 3.7751371115173673, "grad_norm": 0.5493078827857971, "learning_rate": 8.685847149364878e-07, "loss": 0.02, "step": 4130 }, { "epoch": 3.7842778793418645, "grad_norm": 0.5801013708114624, "learning_rate": 7.99862004529145e-07, "loss": 0.0278, "step": 4140 }, { "epoch": 3.7934186471663622, "grad_norm": 0.4189988374710083, "learning_rate": 7.339492787894364e-07, "loss": 0.0199, "step": 4150 }, { "epoch": 3.802559414990859, "grad_norm": 0.5373661518096924, "learning_rate": 6.708503022138879e-07, "loss": 0.0214, "step": 4160 }, { "epoch": 3.8117001828153567, "grad_norm": 0.5810131430625916, "learning_rate": 6.105686785964015e-07, "loss": 0.0243, "step": 4170 }, { "epoch": 3.8208409506398535, "grad_norm": 0.6303432583808899, "learning_rate": 5.531078508223986e-07, "loss": 0.0209, "step": 4180 }, { "epoch": 3.829981718464351, "grad_norm": 0.44799119234085083, "learning_rate": 4.984711006722042e-07, "loss": 0.0228, "step": 4190 }, { "epoch": 3.839122486288848, "grad_norm": 0.3489096760749817, "learning_rate": 4.4666154863361453e-07, "loss": 0.0131, "step": 4200 }, { "epoch": 3.8482632541133457, "grad_norm": 0.368489146232605, "learning_rate": 3.9768215372367745e-07, "loss": 0.0232, "step": 4210 }, { "epoch": 3.857404021937843, "grad_norm": 0.34034988284111023, "learning_rate": 3.5153571331968394e-07, "loss": 0.0188, "step": 4220 }, { "epoch": 3.86654478976234, "grad_norm": 0.4425741732120514, "learning_rate": 3.082248629994011e-07, "loss": 0.0145, "step": 4230 }, { "epoch": 3.8756855575868374, "grad_norm": 0.5476856827735901, "learning_rate": 2.677520763905539e-07, "loss": 0.0227, "step": 4240 }, { "epoch": 3.8848263254113347, "grad_norm": 0.49523550271987915, "learning_rate": 2.301196650295434e-07, "loss": 0.0377, "step": 4250 }, { "epoch": 3.893967093235832, "grad_norm": 0.4109029471874237, "learning_rate": 1.95329778229425e-07, "loss": 0.0268, "step": 4260 }, { "epoch": 3.903107861060329, "grad_norm": 0.48295220732688904, "learning_rate": 1.6338440295715073e-07, "loss": 0.0255, "step": 4270 }, { "epoch": 3.9122486288848264, "grad_norm": 0.3963299095630646, "learning_rate": 1.3428536372009914e-07, "loss": 0.0163, "step": 4280 }, { "epoch": 3.9213893967093236, "grad_norm": 0.6109392642974854, "learning_rate": 1.0803432246186963e-07, "loss": 0.0303, "step": 4290 }, { "epoch": 3.930530164533821, "grad_norm": 0.5676685571670532, "learning_rate": 8.46327784673473e-08, "loss": 0.025, "step": 4300 }, { "epoch": 3.939670932358318, "grad_norm": 0.4061093032360077, "learning_rate": 6.408206827708818e-08, "loss": 0.0219, "step": 4310 }, { "epoch": 3.9488117001828154, "grad_norm": 0.4048132598400116, "learning_rate": 4.6383365610980224e-08, "loss": 0.0274, "step": 4320 }, { "epoch": 3.9579524680073126, "grad_norm": 0.9811397194862366, "learning_rate": 3.153768130121359e-08, "loss": 0.0187, "step": 4330 }, { "epoch": 3.96709323583181, "grad_norm": 0.4111926257610321, "learning_rate": 1.954586323454355e-08, "loss": 0.025, "step": 4340 }, { "epoch": 3.976234003656307, "grad_norm": 0.37575072050094604, "learning_rate": 1.0408596303862483e-08, "loss": 0.0206, "step": 4350 }, { "epoch": 3.9853747714808043, "grad_norm": 0.4568402171134949, "learning_rate": 4.126402369086791e-09, "loss": 0.0202, "step": 4360 }, { "epoch": 3.9945155393053016, "grad_norm": 0.6871432065963745, "learning_rate": 6.996402273529157e-10, "loss": 0.0248, "step": 4370 }, { "epoch": 4.0, "step": 4376, "total_flos": 0.0, "train_loss": 0.05302349528062823, "train_runtime": 2534.7991, "train_samples_per_second": 34.507, "train_steps_per_second": 1.726 } ], "logging_steps": 10, "max_steps": 4376, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 2500, "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": 20, "trial_name": null, "trial_params": null }