{ "best_metric": null, "best_model_checkpoint": null, "epoch": 10.0, "eval_steps": 500, "global_step": 4570, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02188183807439825, "grad_norm": 7.835834503173828, "learning_rate": 4.3668122270742355e-06, "loss": 1.1502, "step": 10 }, { "epoch": 0.0437636761487965, "grad_norm": 5.089853286743164, "learning_rate": 8.733624454148471e-06, "loss": 1.0181, "step": 20 }, { "epoch": 0.06564551422319474, "grad_norm": 2.486280918121338, "learning_rate": 1.3100436681222708e-05, "loss": 0.6402, "step": 30 }, { "epoch": 0.087527352297593, "grad_norm": 2.2392938137054443, "learning_rate": 1.7467248908296942e-05, "loss": 0.4384, "step": 40 }, { "epoch": 0.10940919037199125, "grad_norm": 2.0446348190307617, "learning_rate": 2.183406113537118e-05, "loss": 0.3294, "step": 50 }, { "epoch": 0.13129102844638948, "grad_norm": 1.6032756567001343, "learning_rate": 2.6200873362445416e-05, "loss": 0.2657, "step": 60 }, { "epoch": 0.15317286652078774, "grad_norm": 1.2855844497680664, "learning_rate": 3.056768558951965e-05, "loss": 0.2188, "step": 70 }, { "epoch": 0.175054704595186, "grad_norm": 2.869873046875, "learning_rate": 3.4934497816593884e-05, "loss": 0.1833, "step": 80 }, { "epoch": 0.19693654266958424, "grad_norm": 1.311296820640564, "learning_rate": 3.930131004366812e-05, "loss": 0.1776, "step": 90 }, { "epoch": 0.2188183807439825, "grad_norm": 0.9431630373001099, "learning_rate": 4.366812227074236e-05, "loss": 0.1691, "step": 100 }, { "epoch": 0.24070021881838075, "grad_norm": 0.9962027668952942, "learning_rate": 4.8034934497816594e-05, "loss": 0.1551, "step": 110 }, { "epoch": 0.26258205689277897, "grad_norm": 0.8225952982902527, "learning_rate": 5.240174672489083e-05, "loss": 0.1378, "step": 120 }, { "epoch": 0.2844638949671772, "grad_norm": 0.662451982498169, "learning_rate": 5.6768558951965065e-05, "loss": 0.1357, "step": 130 }, { "epoch": 0.3063457330415755, "grad_norm": 0.7660549283027649, "learning_rate": 6.11353711790393e-05, "loss": 0.1335, "step": 140 }, { "epoch": 0.3282275711159737, "grad_norm": 0.6745500564575195, "learning_rate": 6.550218340611354e-05, "loss": 0.1284, "step": 150 }, { "epoch": 0.350109409190372, "grad_norm": 0.7069956064224243, "learning_rate": 6.986899563318777e-05, "loss": 0.1116, "step": 160 }, { "epoch": 0.37199124726477023, "grad_norm": 1.5347901582717896, "learning_rate": 7.423580786026201e-05, "loss": 0.1157, "step": 170 }, { "epoch": 0.3938730853391685, "grad_norm": 0.6622304320335388, "learning_rate": 7.860262008733625e-05, "loss": 0.1137, "step": 180 }, { "epoch": 0.41575492341356673, "grad_norm": 0.8169113993644714, "learning_rate": 8.296943231441049e-05, "loss": 0.1099, "step": 190 }, { "epoch": 0.437636761487965, "grad_norm": 1.2949796915054321, "learning_rate": 8.733624454148472e-05, "loss": 0.1047, "step": 200 }, { "epoch": 0.45951859956236324, "grad_norm": 1.0924999713897705, "learning_rate": 9.170305676855896e-05, "loss": 0.1077, "step": 210 }, { "epoch": 0.4814004376367615, "grad_norm": 1.032711386680603, "learning_rate": 9.606986899563319e-05, "loss": 0.093, "step": 220 }, { "epoch": 0.5032822757111597, "grad_norm": 0.8792279958724976, "learning_rate": 9.999998690636703e-05, "loss": 0.0948, "step": 230 }, { "epoch": 0.5251641137855579, "grad_norm": 0.7732746601104736, "learning_rate": 9.99984156787068e-05, "loss": 0.0929, "step": 240 }, { "epoch": 0.5470459518599562, "grad_norm": 1.3531904220581055, "learning_rate": 9.999422581874277e-05, "loss": 0.1034, "step": 250 }, { "epoch": 0.5689277899343544, "grad_norm": 42.917991638183594, "learning_rate": 9.998741754591594e-05, "loss": 0.0949, "step": 260 }, { "epoch": 0.5908096280087527, "grad_norm": 13.660039901733398, "learning_rate": 9.997799121680488e-05, "loss": 0.1149, "step": 270 }, { "epoch": 0.612691466083151, "grad_norm": 5.179759979248047, "learning_rate": 9.996594732510703e-05, "loss": 0.1119, "step": 280 }, { "epoch": 0.6345733041575492, "grad_norm": 7.586740970611572, "learning_rate": 9.995128650161283e-05, "loss": 0.0992, "step": 290 }, { "epoch": 0.6564551422319475, "grad_norm": 3.8440942764282227, "learning_rate": 9.993400951417277e-05, "loss": 0.1064, "step": 300 }, { "epoch": 0.6783369803063457, "grad_norm": 2.224613666534424, "learning_rate": 9.991411726765704e-05, "loss": 0.1033, "step": 310 }, { "epoch": 0.700218818380744, "grad_norm": 63.815773010253906, "learning_rate": 9.989161080390825e-05, "loss": 0.1072, "step": 320 }, { "epoch": 0.7221006564551422, "grad_norm": 52.77833938598633, "learning_rate": 9.986649130168682e-05, "loss": 0.0984, "step": 330 }, { "epoch": 0.7439824945295405, "grad_norm": 15.396129608154297, "learning_rate": 9.983876007660922e-05, "loss": 0.0979, "step": 340 }, { "epoch": 0.7658643326039387, "grad_norm": 1.6834031343460083, "learning_rate": 9.980841858107912e-05, "loss": 0.0956, "step": 350 }, { "epoch": 0.787746170678337, "grad_norm": 2.0496537685394287, "learning_rate": 9.977546840421128e-05, "loss": 0.0932, "step": 360 }, { "epoch": 0.8096280087527352, "grad_norm": 1.0929217338562012, "learning_rate": 9.973991127174833e-05, "loss": 0.0864, "step": 370 }, { "epoch": 0.8315098468271335, "grad_norm": 7.530645847320557, "learning_rate": 9.970174904597038e-05, "loss": 0.095, "step": 380 }, { "epoch": 0.8533916849015317, "grad_norm": 1.3216899633407593, "learning_rate": 9.966098372559754e-05, "loss": 0.1115, "step": 390 }, { "epoch": 0.87527352297593, "grad_norm": 1.4777535200119019, "learning_rate": 9.961761744568512e-05, "loss": 0.1035, "step": 400 }, { "epoch": 0.8971553610503282, "grad_norm": 1.4249495267868042, "learning_rate": 9.957165247751192e-05, "loss": 0.0945, "step": 410 }, { "epoch": 0.9190371991247265, "grad_norm": 5.402189254760742, "learning_rate": 9.952309122846128e-05, "loss": 0.0972, "step": 420 }, { "epoch": 0.9409190371991247, "grad_norm": 2.5458638668060303, "learning_rate": 9.947193624189485e-05, "loss": 0.0933, "step": 430 }, { "epoch": 0.962800875273523, "grad_norm": 1.0579487085342407, "learning_rate": 9.941819019701954e-05, "loss": 0.1069, "step": 440 }, { "epoch": 0.9846827133479212, "grad_norm": 2.8812167644500732, "learning_rate": 9.936185590874716e-05, "loss": 0.0873, "step": 450 }, { "epoch": 1.0065645514223194, "grad_norm": 1.7286851406097412, "learning_rate": 9.93029363275469e-05, "loss": 0.098, "step": 460 }, { "epoch": 1.0284463894967177, "grad_norm": 2.46840763092041, "learning_rate": 9.924143453929095e-05, "loss": 0.097, "step": 470 }, { "epoch": 1.0503282275711159, "grad_norm": 3.4682812690734863, "learning_rate": 9.917735376509274e-05, "loss": 0.0899, "step": 480 }, { "epoch": 1.0722100656455142, "grad_norm": 1.5925973653793335, "learning_rate": 9.911069736113832e-05, "loss": 0.0864, "step": 490 }, { "epoch": 1.0940919037199124, "grad_norm": 1.5992704629898071, "learning_rate": 9.904146881851059e-05, "loss": 0.0819, "step": 500 }, { "epoch": 1.1159737417943107, "grad_norm": 1.0037175416946411, "learning_rate": 9.896967176300638e-05, "loss": 0.0833, "step": 510 }, { "epoch": 1.1378555798687089, "grad_norm": 1.681389570236206, "learning_rate": 9.88953099549466e-05, "loss": 0.0934, "step": 520 }, { "epoch": 1.1597374179431073, "grad_norm": 1.4893583059310913, "learning_rate": 9.881838728897936e-05, "loss": 0.093, "step": 530 }, { "epoch": 1.1816192560175054, "grad_norm": 1.2456260919570923, "learning_rate": 9.873890779387584e-05, "loss": 0.1026, "step": 540 }, { "epoch": 1.2035010940919038, "grad_norm": 0.8705942630767822, "learning_rate": 9.865687563231943e-05, "loss": 0.092, "step": 550 }, { "epoch": 1.225382932166302, "grad_norm": 2.0043041706085205, "learning_rate": 9.857229510068761e-05, "loss": 0.0949, "step": 560 }, { "epoch": 1.2472647702407003, "grad_norm": 1.574006199836731, "learning_rate": 9.848517062882703e-05, "loss": 0.0843, "step": 570 }, { "epoch": 1.2691466083150984, "grad_norm": 0.98636794090271, "learning_rate": 9.839550677982135e-05, "loss": 0.0875, "step": 580 }, { "epoch": 1.2910284463894968, "grad_norm": 0.5821505784988403, "learning_rate": 9.830330824975245e-05, "loss": 0.0923, "step": 590 }, { "epoch": 1.312910284463895, "grad_norm": 0.5067301392555237, "learning_rate": 9.820857986745431e-05, "loss": 0.0812, "step": 600 }, { "epoch": 1.3347921225382933, "grad_norm": 0.5871260762214661, "learning_rate": 9.811132659426016e-05, "loss": 0.0804, "step": 610 }, { "epoch": 1.3566739606126914, "grad_norm": 0.6894322633743286, "learning_rate": 9.801155352374268e-05, "loss": 0.0909, "step": 620 }, { "epoch": 1.3785557986870898, "grad_norm": 0.7414800524711609, "learning_rate": 9.790926588144714e-05, "loss": 0.0901, "step": 630 }, { "epoch": 1.400437636761488, "grad_norm": 0.633738100528717, "learning_rate": 9.780446902461777e-05, "loss": 0.092, "step": 640 }, { "epoch": 1.4223194748358863, "grad_norm": 0.9708178639411926, "learning_rate": 9.769716844191722e-05, "loss": 0.0905, "step": 650 }, { "epoch": 1.4442013129102844, "grad_norm": 0.6483299732208252, "learning_rate": 9.758736975313895e-05, "loss": 0.0835, "step": 660 }, { "epoch": 1.4660831509846828, "grad_norm": 0.6239913702011108, "learning_rate": 9.747507870891311e-05, "loss": 0.084, "step": 670 }, { "epoch": 1.487964989059081, "grad_norm": 0.5222530364990234, "learning_rate": 9.736030119040508e-05, "loss": 0.0883, "step": 680 }, { "epoch": 1.509846827133479, "grad_norm": 0.7400548458099365, "learning_rate": 9.724304320900779e-05, "loss": 0.0894, "step": 690 }, { "epoch": 1.5317286652078774, "grad_norm": 0.5438825488090515, "learning_rate": 9.712331090602654e-05, "loss": 0.0848, "step": 700 }, { "epoch": 1.5536105032822758, "grad_norm": 0.351309210062027, "learning_rate": 9.700111055235754e-05, "loss": 0.0796, "step": 710 }, { "epoch": 1.575492341356674, "grad_norm": 0.5478343963623047, "learning_rate": 9.687644854815952e-05, "loss": 0.0812, "step": 720 }, { "epoch": 1.597374179431072, "grad_norm": 0.5228798985481262, "learning_rate": 9.674933142251837e-05, "loss": 0.0802, "step": 730 }, { "epoch": 1.6192560175054704, "grad_norm": 0.7279865145683289, "learning_rate": 9.661976583310524e-05, "loss": 0.0834, "step": 740 }, { "epoch": 1.6411378555798688, "grad_norm": 0.653016209602356, "learning_rate": 9.648775856582795e-05, "loss": 0.0818, "step": 750 }, { "epoch": 1.663019693654267, "grad_norm": 0.5135919451713562, "learning_rate": 9.635331653447545e-05, "loss": 0.075, "step": 760 }, { "epoch": 1.684901531728665, "grad_norm": 0.6700458526611328, "learning_rate": 9.621644678035577e-05, "loss": 0.0752, "step": 770 }, { "epoch": 1.7067833698030634, "grad_norm": 0.7173268795013428, "learning_rate": 9.607715647192726e-05, "loss": 0.0829, "step": 780 }, { "epoch": 1.7286652078774618, "grad_norm": 0.5226753354072571, "learning_rate": 9.593545290442311e-05, "loss": 0.0796, "step": 790 }, { "epoch": 1.75054704595186, "grad_norm": 0.6492887139320374, "learning_rate": 9.579134349946926e-05, "loss": 0.0759, "step": 800 }, { "epoch": 1.772428884026258, "grad_norm": 0.6919318437576294, "learning_rate": 9.564483580469571e-05, "loss": 0.0815, "step": 810 }, { "epoch": 1.7943107221006565, "grad_norm": 0.8271254301071167, "learning_rate": 9.549593749334128e-05, "loss": 0.0796, "step": 820 }, { "epoch": 1.8161925601750548, "grad_norm": 0.4964686930179596, "learning_rate": 9.534465636385162e-05, "loss": 0.087, "step": 830 }, { "epoch": 1.838074398249453, "grad_norm": 0.5696905255317688, "learning_rate": 9.519100033947083e-05, "loss": 0.0822, "step": 840 }, { "epoch": 1.859956236323851, "grad_norm": 0.849829375743866, "learning_rate": 9.503497746782653e-05, "loss": 0.0727, "step": 850 }, { "epoch": 1.8818380743982495, "grad_norm": 0.5354284644126892, "learning_rate": 9.487659592050824e-05, "loss": 0.0902, "step": 860 }, { "epoch": 1.9037199124726478, "grad_norm": 0.38756123185157776, "learning_rate": 9.471586399263956e-05, "loss": 0.0775, "step": 870 }, { "epoch": 1.925601750547046, "grad_norm": 0.7026869058609009, "learning_rate": 9.45527901024436e-05, "loss": 0.0788, "step": 880 }, { "epoch": 1.947483588621444, "grad_norm": 0.5695210695266724, "learning_rate": 9.43873827908021e-05, "loss": 0.0757, "step": 890 }, { "epoch": 1.9693654266958425, "grad_norm": 0.40557512640953064, "learning_rate": 9.421965072080817e-05, "loss": 0.0722, "step": 900 }, { "epoch": 1.9912472647702408, "grad_norm": 0.42085573077201843, "learning_rate": 9.404960267731251e-05, "loss": 0.0689, "step": 910 }, { "epoch": 2.0131291028446388, "grad_norm": 0.4493652880191803, "learning_rate": 9.387724756646332e-05, "loss": 0.0774, "step": 920 }, { "epoch": 2.035010940919037, "grad_norm": 0.49804505705833435, "learning_rate": 9.370259441523987e-05, "loss": 0.0771, "step": 930 }, { "epoch": 2.0568927789934355, "grad_norm": 0.6483808755874634, "learning_rate": 9.352565237097965e-05, "loss": 0.078, "step": 940 }, { "epoch": 2.078774617067834, "grad_norm": 0.5233718156814575, "learning_rate": 9.334643070089937e-05, "loss": 0.0709, "step": 950 }, { "epoch": 2.1006564551422318, "grad_norm": 0.6516150832176208, "learning_rate": 9.316493879160957e-05, "loss": 0.079, "step": 960 }, { "epoch": 2.12253829321663, "grad_norm": 0.692732036113739, "learning_rate": 9.298118614862297e-05, "loss": 0.0775, "step": 970 }, { "epoch": 2.1444201312910285, "grad_norm": 0.5700138807296753, "learning_rate": 9.279518239585667e-05, "loss": 0.0782, "step": 980 }, { "epoch": 2.166301969365427, "grad_norm": 0.3617083728313446, "learning_rate": 9.260693727512803e-05, "loss": 0.0714, "step": 990 }, { "epoch": 2.1881838074398248, "grad_norm": 0.4216122031211853, "learning_rate": 9.241646064564457e-05, "loss": 0.0672, "step": 1000 }, { "epoch": 2.210065645514223, "grad_norm": 0.38087522983551025, "learning_rate": 9.222376248348747e-05, "loss": 0.08, "step": 1010 }, { "epoch": 2.2319474835886215, "grad_norm": 0.8693896532058716, "learning_rate": 9.20288528810892e-05, "loss": 0.0778, "step": 1020 }, { "epoch": 2.25382932166302, "grad_norm": 0.6329444050788879, "learning_rate": 9.183174204670483e-05, "loss": 0.0722, "step": 1030 }, { "epoch": 2.2757111597374178, "grad_norm": 0.9332349896430969, "learning_rate": 9.163244030387743e-05, "loss": 0.0705, "step": 1040 }, { "epoch": 2.297592997811816, "grad_norm": 0.8268052935600281, "learning_rate": 9.143095809089741e-05, "loss": 0.0719, "step": 1050 }, { "epoch": 2.3194748358862145, "grad_norm": 1.8830317258834839, "learning_rate": 9.122730596025579e-05, "loss": 0.0733, "step": 1060 }, { "epoch": 2.341356673960613, "grad_norm": 0.4748838245868683, "learning_rate": 9.102149457809146e-05, "loss": 0.0692, "step": 1070 }, { "epoch": 2.363238512035011, "grad_norm": 0.6225442290306091, "learning_rate": 9.081353472363265e-05, "loss": 0.0724, "step": 1080 }, { "epoch": 2.385120350109409, "grad_norm": 0.3627307415008545, "learning_rate": 9.060343728863239e-05, "loss": 0.064, "step": 1090 }, { "epoch": 2.4070021881838075, "grad_norm": 0.5193670392036438, "learning_rate": 9.039121327679791e-05, "loss": 0.079, "step": 1100 }, { "epoch": 2.428884026258206, "grad_norm": 0.8696744441986084, "learning_rate": 9.017687380321449e-05, "loss": 0.0715, "step": 1110 }, { "epoch": 2.450765864332604, "grad_norm": 0.6547001004219055, "learning_rate": 8.996043009376319e-05, "loss": 0.0691, "step": 1120 }, { "epoch": 2.472647702407002, "grad_norm": 0.5175451040267944, "learning_rate": 8.974189348453306e-05, "loss": 0.0749, "step": 1130 }, { "epoch": 2.4945295404814005, "grad_norm": 0.5802105665206909, "learning_rate": 8.952127542122722e-05, "loss": 0.0717, "step": 1140 }, { "epoch": 2.516411378555799, "grad_norm": 0.7577764987945557, "learning_rate": 8.929858745856354e-05, "loss": 0.0816, "step": 1150 }, { "epoch": 2.538293216630197, "grad_norm": 0.5405493378639221, "learning_rate": 8.907384125966945e-05, "loss": 0.0738, "step": 1160 }, { "epoch": 2.560175054704595, "grad_norm": 0.3944590985774994, "learning_rate": 8.884704859547108e-05, "loss": 0.071, "step": 1170 }, { "epoch": 2.5820568927789935, "grad_norm": 0.5056865215301514, "learning_rate": 8.861822134407672e-05, "loss": 0.0692, "step": 1180 }, { "epoch": 2.6039387308533914, "grad_norm": 0.4500432312488556, "learning_rate": 8.838737149015477e-05, "loss": 0.0723, "step": 1190 }, { "epoch": 2.62582056892779, "grad_norm": 0.36721524596214294, "learning_rate": 8.815451112430604e-05, "loss": 0.0686, "step": 1200 }, { "epoch": 2.647702407002188, "grad_norm": 0.7576287984848022, "learning_rate": 8.79196524424305e-05, "loss": 0.0769, "step": 1210 }, { "epoch": 2.6695842450765865, "grad_norm": 0.6370827555656433, "learning_rate": 8.768280774508853e-05, "loss": 0.0815, "step": 1220 }, { "epoch": 2.691466083150985, "grad_norm": 0.3080037236213684, "learning_rate": 8.74439894368567e-05, "loss": 0.0607, "step": 1230 }, { "epoch": 2.713347921225383, "grad_norm": 0.46508321166038513, "learning_rate": 8.720321002567807e-05, "loss": 0.0696, "step": 1240 }, { "epoch": 2.735229759299781, "grad_norm": 0.6237578988075256, "learning_rate": 8.696048212220711e-05, "loss": 0.0716, "step": 1250 }, { "epoch": 2.7571115973741795, "grad_norm": 0.4095085561275482, "learning_rate": 8.671581843914923e-05, "loss": 0.0635, "step": 1260 }, { "epoch": 2.7789934354485775, "grad_norm": 0.6405256986618042, "learning_rate": 8.646923179059494e-05, "loss": 0.0648, "step": 1270 }, { "epoch": 2.800875273522976, "grad_norm": 0.48811593651771545, "learning_rate": 8.622073509134874e-05, "loss": 0.0645, "step": 1280 }, { "epoch": 2.822757111597374, "grad_norm": 0.5660180449485779, "learning_rate": 8.597034135625268e-05, "loss": 0.0697, "step": 1290 }, { "epoch": 2.8446389496717726, "grad_norm": 0.5524961352348328, "learning_rate": 8.571806369950481e-05, "loss": 0.0721, "step": 1300 }, { "epoch": 2.866520787746171, "grad_norm": 0.38668200373649597, "learning_rate": 8.54639153339722e-05, "loss": 0.0666, "step": 1310 }, { "epoch": 2.888402625820569, "grad_norm": 0.5835091471672058, "learning_rate": 8.520790957049904e-05, "loss": 0.0613, "step": 1320 }, { "epoch": 2.910284463894967, "grad_norm": 0.5212987661361694, "learning_rate": 8.49500598172094e-05, "loss": 0.0732, "step": 1330 }, { "epoch": 2.9321663019693656, "grad_norm": 0.28042611479759216, "learning_rate": 8.46903795788051e-05, "loss": 0.0582, "step": 1340 }, { "epoch": 2.9540481400437635, "grad_norm": 0.5811691880226135, "learning_rate": 8.442888245585828e-05, "loss": 0.0683, "step": 1350 }, { "epoch": 2.975929978118162, "grad_norm": 0.4917229115962982, "learning_rate": 8.41655821440992e-05, "loss": 0.0669, "step": 1360 }, { "epoch": 2.99781181619256, "grad_norm": 0.5715236663818359, "learning_rate": 8.390049243369886e-05, "loss": 0.064, "step": 1370 }, { "epoch": 3.0196936542669586, "grad_norm": 0.7416530251502991, "learning_rate": 8.363362720854677e-05, "loss": 0.0701, "step": 1380 }, { "epoch": 3.0415754923413565, "grad_norm": 0.6221567988395691, "learning_rate": 8.33650004455238e-05, "loss": 0.0707, "step": 1390 }, { "epoch": 3.063457330415755, "grad_norm": 0.5047647953033447, "learning_rate": 8.309462621377013e-05, "loss": 0.0787, "step": 1400 }, { "epoch": 3.085339168490153, "grad_norm": 0.4984498918056488, "learning_rate": 8.28225186739484e-05, "loss": 0.0629, "step": 1410 }, { "epoch": 3.1072210065645516, "grad_norm": 0.37575212121009827, "learning_rate": 8.254869207750206e-05, "loss": 0.0638, "step": 1420 }, { "epoch": 3.1291028446389495, "grad_norm": 0.5403359532356262, "learning_rate": 8.227316076590898e-05, "loss": 0.0645, "step": 1430 }, { "epoch": 3.150984682713348, "grad_norm": 0.4255600869655609, "learning_rate": 8.19959391699303e-05, "loss": 0.0615, "step": 1440 }, { "epoch": 3.1728665207877462, "grad_norm": 0.41742101311683655, "learning_rate": 8.171704180885457e-05, "loss": 0.0614, "step": 1450 }, { "epoch": 3.1947483588621446, "grad_norm": 0.36597940325737, "learning_rate": 8.143648328973746e-05, "loss": 0.0662, "step": 1460 }, { "epoch": 3.2166301969365425, "grad_norm": 0.34130623936653137, "learning_rate": 8.115427830663658e-05, "loss": 0.0705, "step": 1470 }, { "epoch": 3.238512035010941, "grad_norm": 0.45353439450263977, "learning_rate": 8.087044163984197e-05, "loss": 0.062, "step": 1480 }, { "epoch": 3.2603938730853392, "grad_norm": 0.525982677936554, "learning_rate": 8.058498815510196e-05, "loss": 0.0691, "step": 1490 }, { "epoch": 3.2822757111597376, "grad_norm": 0.520045280456543, "learning_rate": 8.029793280284459e-05, "loss": 0.0612, "step": 1500 }, { "epoch": 3.3041575492341355, "grad_norm": 0.5012158155441284, "learning_rate": 8.000929061739464e-05, "loss": 0.067, "step": 1510 }, { "epoch": 3.326039387308534, "grad_norm": 0.36634936928749084, "learning_rate": 7.971907671618612e-05, "loss": 0.0672, "step": 1520 }, { "epoch": 3.3479212253829322, "grad_norm": 0.43757519125938416, "learning_rate": 7.942730629897063e-05, "loss": 0.067, "step": 1530 }, { "epoch": 3.3698030634573306, "grad_norm": 0.44617760181427, "learning_rate": 7.913399464702113e-05, "loss": 0.0623, "step": 1540 }, { "epoch": 3.3916849015317285, "grad_norm": 0.6319367289543152, "learning_rate": 7.883915712233177e-05, "loss": 0.063, "step": 1550 }, { "epoch": 3.413566739606127, "grad_norm": 0.3538963496685028, "learning_rate": 7.854280916681314e-05, "loss": 0.066, "step": 1560 }, { "epoch": 3.4354485776805253, "grad_norm": 0.4608595371246338, "learning_rate": 7.824496630148364e-05, "loss": 0.0621, "step": 1570 }, { "epoch": 3.4573304157549236, "grad_norm": 0.5707418322563171, "learning_rate": 7.794564412565655e-05, "loss": 0.0696, "step": 1580 }, { "epoch": 3.4792122538293215, "grad_norm": 0.37221798300743103, "learning_rate": 7.764485831612299e-05, "loss": 0.066, "step": 1590 }, { "epoch": 3.50109409190372, "grad_norm": 0.6436975598335266, "learning_rate": 7.734262462633085e-05, "loss": 0.0597, "step": 1600 }, { "epoch": 3.5229759299781183, "grad_norm": 0.7743731141090393, "learning_rate": 7.703895888555978e-05, "loss": 0.0696, "step": 1610 }, { "epoch": 3.544857768052516, "grad_norm": 0.561302900314331, "learning_rate": 7.673387699809211e-05, "loss": 0.0613, "step": 1620 }, { "epoch": 3.5667396061269145, "grad_norm": 0.42251160740852356, "learning_rate": 7.642739494237986e-05, "loss": 0.0642, "step": 1630 }, { "epoch": 3.588621444201313, "grad_norm": 0.2583286166191101, "learning_rate": 7.611952877020787e-05, "loss": 0.0648, "step": 1640 }, { "epoch": 3.6105032822757113, "grad_norm": 0.6360892057418823, "learning_rate": 7.581029460585313e-05, "loss": 0.0621, "step": 1650 }, { "epoch": 3.6323851203501096, "grad_norm": 0.4245406687259674, "learning_rate": 7.54997086452403e-05, "loss": 0.0618, "step": 1660 }, { "epoch": 3.6542669584245075, "grad_norm": 0.3884272277355194, "learning_rate": 7.518778715509341e-05, "loss": 0.0651, "step": 1670 }, { "epoch": 3.676148796498906, "grad_norm": 0.4515193700790405, "learning_rate": 7.487454647208387e-05, "loss": 0.0614, "step": 1680 }, { "epoch": 3.6980306345733043, "grad_norm": 0.32720717787742615, "learning_rate": 7.456000300197498e-05, "loss": 0.0632, "step": 1690 }, { "epoch": 3.719912472647702, "grad_norm": 0.3803008198738098, "learning_rate": 7.424417321876258e-05, "loss": 0.0688, "step": 1700 }, { "epoch": 3.7417943107221006, "grad_norm": 0.4703556299209595, "learning_rate": 7.392707366381226e-05, "loss": 0.0565, "step": 1710 }, { "epoch": 3.763676148796499, "grad_norm": 0.3876951336860657, "learning_rate": 7.360872094499303e-05, "loss": 0.0629, "step": 1720 }, { "epoch": 3.7855579868708973, "grad_norm": 0.6968600749969482, "learning_rate": 7.328913173580745e-05, "loss": 0.0654, "step": 1730 }, { "epoch": 3.8074398249452956, "grad_norm": 0.3983824551105499, "learning_rate": 7.29683227745185e-05, "loss": 0.0665, "step": 1740 }, { "epoch": 3.8293216630196936, "grad_norm": 0.6152142286300659, "learning_rate": 7.264631086327273e-05, "loss": 0.0665, "step": 1750 }, { "epoch": 3.851203501094092, "grad_norm": 0.45532792806625366, "learning_rate": 7.232311286722038e-05, "loss": 0.0644, "step": 1760 }, { "epoch": 3.8730853391684903, "grad_norm": 0.5433072447776794, "learning_rate": 7.199874571363212e-05, "loss": 0.0669, "step": 1770 }, { "epoch": 3.894967177242888, "grad_norm": 0.393799364566803, "learning_rate": 7.167322639101234e-05, "loss": 0.0609, "step": 1780 }, { "epoch": 3.9168490153172866, "grad_norm": 0.37970104813575745, "learning_rate": 7.134657194820957e-05, "loss": 0.0613, "step": 1790 }, { "epoch": 3.938730853391685, "grad_norm": 0.4533461928367615, "learning_rate": 7.101879949352343e-05, "loss": 0.0661, "step": 1800 }, { "epoch": 3.9606126914660833, "grad_norm": 0.5303936004638672, "learning_rate": 7.068992619380859e-05, "loss": 0.0604, "step": 1810 }, { "epoch": 3.9824945295404817, "grad_norm": 0.43903806805610657, "learning_rate": 7.035996927357583e-05, "loss": 0.0575, "step": 1820 }, { "epoch": 4.00437636761488, "grad_norm": 0.4638700485229492, "learning_rate": 7.00289460140897e-05, "loss": 0.0675, "step": 1830 }, { "epoch": 4.0262582056892775, "grad_norm": 0.6604804396629333, "learning_rate": 6.969687375246355e-05, "loss": 0.066, "step": 1840 }, { "epoch": 4.048140043763676, "grad_norm": 0.39075350761413574, "learning_rate": 6.936376988075155e-05, "loss": 0.0642, "step": 1850 }, { "epoch": 4.070021881838074, "grad_norm": 0.5355797410011292, "learning_rate": 6.902965184503763e-05, "loss": 0.0618, "step": 1860 }, { "epoch": 4.091903719912473, "grad_norm": 0.70420241355896, "learning_rate": 6.869453714452194e-05, "loss": 0.0613, "step": 1870 }, { "epoch": 4.113785557986871, "grad_norm": 0.5154333114624023, "learning_rate": 6.835844333060423e-05, "loss": 0.058, "step": 1880 }, { "epoch": 4.135667396061269, "grad_norm": 0.5481790900230408, "learning_rate": 6.802138800596462e-05, "loss": 0.0602, "step": 1890 }, { "epoch": 4.157549234135668, "grad_norm": 1.0096722841262817, "learning_rate": 6.768338882364168e-05, "loss": 0.0693, "step": 1900 }, { "epoch": 4.179431072210066, "grad_norm": 0.4854465126991272, "learning_rate": 6.734446348610792e-05, "loss": 0.0612, "step": 1910 }, { "epoch": 4.2013129102844635, "grad_norm": 0.33401617407798767, "learning_rate": 6.700462974434251e-05, "loss": 0.0612, "step": 1920 }, { "epoch": 4.223194748358862, "grad_norm": 0.44405895471572876, "learning_rate": 6.666390539690179e-05, "loss": 0.0644, "step": 1930 }, { "epoch": 4.24507658643326, "grad_norm": 0.43215858936309814, "learning_rate": 6.632230828898677e-05, "loss": 0.0592, "step": 1940 }, { "epoch": 4.266958424507659, "grad_norm": 0.3804273307323456, "learning_rate": 6.597985631150884e-05, "loss": 0.0565, "step": 1950 }, { "epoch": 4.288840262582057, "grad_norm": 0.43090665340423584, "learning_rate": 6.56365674001525e-05, "loss": 0.0579, "step": 1960 }, { "epoch": 4.310722100656455, "grad_norm": 0.5440065264701843, "learning_rate": 6.529245953443615e-05, "loss": 0.0575, "step": 1970 }, { "epoch": 4.332603938730854, "grad_norm": 0.36599451303482056, "learning_rate": 6.49475507367703e-05, "loss": 0.0598, "step": 1980 }, { "epoch": 4.354485776805252, "grad_norm": 0.33587977290153503, "learning_rate": 6.460185907151371e-05, "loss": 0.0583, "step": 1990 }, { "epoch": 4.3763676148796495, "grad_norm": 0.3317202627658844, "learning_rate": 6.425540264402735e-05, "loss": 0.0601, "step": 2000 }, { "epoch": 4.398249452954048, "grad_norm": 0.36118966341018677, "learning_rate": 6.390819959972603e-05, "loss": 0.0531, "step": 2010 }, { "epoch": 4.420131291028446, "grad_norm": 0.3242454528808594, "learning_rate": 6.356026812312809e-05, "loss": 0.0585, "step": 2020 }, { "epoch": 4.442013129102845, "grad_norm": 0.4215313494205475, "learning_rate": 6.321162643690304e-05, "loss": 0.0637, "step": 2030 }, { "epoch": 4.463894967177243, "grad_norm": 0.34491801261901855, "learning_rate": 6.286229280091711e-05, "loss": 0.0642, "step": 2040 }, { "epoch": 4.485776805251641, "grad_norm": 0.5286457538604736, "learning_rate": 6.251228551127691e-05, "loss": 0.0654, "step": 2050 }, { "epoch": 4.50765864332604, "grad_norm": 0.357379674911499, "learning_rate": 6.216162289937119e-05, "loss": 0.0529, "step": 2060 }, { "epoch": 4.529540481400438, "grad_norm": 0.4211725890636444, "learning_rate": 6.181032333091078e-05, "loss": 0.0574, "step": 2070 }, { "epoch": 4.5514223194748356, "grad_norm": 0.4051927328109741, "learning_rate": 6.145840520496666e-05, "loss": 0.0555, "step": 2080 }, { "epoch": 4.573304157549234, "grad_norm": 0.4655281603336334, "learning_rate": 6.11058869530063e-05, "loss": 0.061, "step": 2090 }, { "epoch": 4.595185995623632, "grad_norm": 0.3950660526752472, "learning_rate": 6.0752787037928336e-05, "loss": 0.061, "step": 2100 }, { "epoch": 4.61706783369803, "grad_norm": 0.5274171233177185, "learning_rate": 6.039912395309568e-05, "loss": 0.0605, "step": 2110 }, { "epoch": 4.638949671772429, "grad_norm": 0.3764810860157013, "learning_rate": 6.004491622136681e-05, "loss": 0.0555, "step": 2120 }, { "epoch": 4.660831509846827, "grad_norm": 0.3669235110282898, "learning_rate": 5.969018239412573e-05, "loss": 0.0657, "step": 2130 }, { "epoch": 4.682713347921226, "grad_norm": 0.4793848395347595, "learning_rate": 5.933494105031032e-05, "loss": 0.0625, "step": 2140 }, { "epoch": 4.704595185995624, "grad_norm": 0.7370563745498657, "learning_rate": 5.897921079543931e-05, "loss": 0.0553, "step": 2150 }, { "epoch": 4.726477024070022, "grad_norm": 0.32018744945526123, "learning_rate": 5.862301026063777e-05, "loss": 0.055, "step": 2160 }, { "epoch": 4.74835886214442, "grad_norm": 0.39296776056289673, "learning_rate": 5.826635810166137e-05, "loss": 0.052, "step": 2170 }, { "epoch": 4.770240700218818, "grad_norm": 0.4568535089492798, "learning_rate": 5.7909272997919305e-05, "loss": 0.056, "step": 2180 }, { "epoch": 4.792122538293217, "grad_norm": 0.39834004640579224, "learning_rate": 5.7551773651495953e-05, "loss": 0.0586, "step": 2190 }, { "epoch": 4.814004376367615, "grad_norm": 0.47946035861968994, "learning_rate": 5.719387878617131e-05, "loss": 0.0609, "step": 2200 }, { "epoch": 4.835886214442013, "grad_norm": 0.3279586434364319, "learning_rate": 5.683560714644044e-05, "loss": 0.059, "step": 2210 }, { "epoch": 4.857768052516412, "grad_norm": 0.48775798082351685, "learning_rate": 5.647697749653172e-05, "loss": 0.0565, "step": 2220 }, { "epoch": 4.87964989059081, "grad_norm": 0.42671310901641846, "learning_rate": 5.6118008619424036e-05, "loss": 0.0605, "step": 2230 }, { "epoch": 4.901531728665208, "grad_norm": 0.3488156497478485, "learning_rate": 5.5758719315863076e-05, "loss": 0.0548, "step": 2240 }, { "epoch": 4.923413566739606, "grad_norm": 0.4897012710571289, "learning_rate": 5.539912840337661e-05, "loss": 0.0587, "step": 2250 }, { "epoch": 4.945295404814004, "grad_norm": 0.40971580147743225, "learning_rate": 5.5039254715289e-05, "loss": 0.0608, "step": 2260 }, { "epoch": 4.967177242888402, "grad_norm": 0.5318540334701538, "learning_rate": 5.46791170997348e-05, "loss": 0.0497, "step": 2270 }, { "epoch": 4.989059080962801, "grad_norm": 0.42819133400917053, "learning_rate": 5.4318734418671525e-05, "loss": 0.0636, "step": 2280 }, { "epoch": 5.010940919037199, "grad_norm": 0.4624245762825012, "learning_rate": 5.395812554689188e-05, "loss": 0.0613, "step": 2290 }, { "epoch": 5.032822757111598, "grad_norm": 0.4868052005767822, "learning_rate": 5.359730937103512e-05, "loss": 0.0523, "step": 2300 }, { "epoch": 5.054704595185996, "grad_norm": 0.6195753216743469, "learning_rate": 5.323630478859789e-05, "loss": 0.0604, "step": 2310 }, { "epoch": 5.076586433260394, "grad_norm": 0.38792890310287476, "learning_rate": 5.2875130706944574e-05, "loss": 0.0539, "step": 2320 }, { "epoch": 5.098468271334792, "grad_norm": 0.5414655208587646, "learning_rate": 5.251380604231687e-05, "loss": 0.0602, "step": 2330 }, { "epoch": 5.12035010940919, "grad_norm": 0.5284174680709839, "learning_rate": 5.215234971884321e-05, "loss": 0.0569, "step": 2340 }, { "epoch": 5.142231947483588, "grad_norm": 0.4137500822544098, "learning_rate": 5.1790780667547564e-05, "loss": 0.0595, "step": 2350 }, { "epoch": 5.164113785557987, "grad_norm": 0.456150084733963, "learning_rate": 5.1429117825357894e-05, "loss": 0.0652, "step": 2360 }, { "epoch": 5.185995623632385, "grad_norm": 0.45625388622283936, "learning_rate": 5.1067380134114484e-05, "loss": 0.0638, "step": 2370 }, { "epoch": 5.207877461706784, "grad_norm": 0.4778393507003784, "learning_rate": 5.0705586539577687e-05, "loss": 0.0562, "step": 2380 }, { "epoch": 5.229759299781182, "grad_norm": 0.45723408460617065, "learning_rate": 5.034375599043582e-05, "loss": 0.0578, "step": 2390 }, { "epoch": 5.25164113785558, "grad_norm": 0.4665261507034302, "learning_rate": 4.998190743731265e-05, "loss": 0.0483, "step": 2400 }, { "epoch": 5.273522975929978, "grad_norm": 0.37840256094932556, "learning_rate": 4.96200598317749e-05, "loss": 0.06, "step": 2410 }, { "epoch": 5.295404814004376, "grad_norm": 0.4239036738872528, "learning_rate": 4.925823212533963e-05, "loss": 0.0575, "step": 2420 }, { "epoch": 5.317286652078774, "grad_norm": 0.45726969838142395, "learning_rate": 4.8896443268481715e-05, "loss": 0.0523, "step": 2430 }, { "epoch": 5.339168490153173, "grad_norm": 0.3009950816631317, "learning_rate": 4.853471220964137e-05, "loss": 0.0573, "step": 2440 }, { "epoch": 5.361050328227571, "grad_norm": 0.5033520460128784, "learning_rate": 4.8173057894231585e-05, "loss": 0.0576, "step": 2450 }, { "epoch": 5.38293216630197, "grad_norm": 0.6416246294975281, "learning_rate": 4.7811499263646024e-05, "loss": 0.0566, "step": 2460 }, { "epoch": 5.404814004376368, "grad_norm": 0.45454248785972595, "learning_rate": 4.7450055254266875e-05, "loss": 0.0603, "step": 2470 }, { "epoch": 5.426695842450766, "grad_norm": 0.4150170087814331, "learning_rate": 4.708874479647319e-05, "loss": 0.061, "step": 2480 }, { "epoch": 5.4485776805251644, "grad_norm": 0.7388213276863098, "learning_rate": 4.6727586813649324e-05, "loss": 0.0559, "step": 2490 }, { "epoch": 5.470459518599562, "grad_norm": 0.42447715997695923, "learning_rate": 4.6366600221193815e-05, "loss": 0.0523, "step": 2500 }, { "epoch": 5.49234135667396, "grad_norm": 0.35511597990989685, "learning_rate": 4.60058039255288e-05, "loss": 0.0494, "step": 2510 }, { "epoch": 5.514223194748359, "grad_norm": 0.47702649235725403, "learning_rate": 4.5645216823109727e-05, "loss": 0.0621, "step": 2520 }, { "epoch": 5.536105032822757, "grad_norm": 0.5418275594711304, "learning_rate": 4.5284857799435726e-05, "loss": 0.0569, "step": 2530 }, { "epoch": 5.557986870897155, "grad_norm": 0.2867294251918793, "learning_rate": 4.4924745728060444e-05, "loss": 0.0575, "step": 2540 }, { "epoch": 5.579868708971554, "grad_norm": 0.3692294657230377, "learning_rate": 4.456489946960353e-05, "loss": 0.0621, "step": 2550 }, { "epoch": 5.601750547045952, "grad_norm": 0.5166872143745422, "learning_rate": 4.4205337870762944e-05, "loss": 0.0514, "step": 2560 }, { "epoch": 5.6236323851203505, "grad_norm": 0.25781774520874023, "learning_rate": 4.384607976332773e-05, "loss": 0.0526, "step": 2570 }, { "epoch": 5.645514223194748, "grad_norm": 0.458324670791626, "learning_rate": 4.348714396319182e-05, "loss": 0.0618, "step": 2580 }, { "epoch": 5.667396061269146, "grad_norm": 0.5077418088912964, "learning_rate": 4.3128549269368525e-05, "loss": 0.057, "step": 2590 }, { "epoch": 5.689277899343545, "grad_norm": 0.3836349844932556, "learning_rate": 4.27703144630059e-05, "loss": 0.0548, "step": 2600 }, { "epoch": 5.711159737417943, "grad_norm": 0.29659900069236755, "learning_rate": 4.2412458306403224e-05, "loss": 0.0535, "step": 2610 }, { "epoch": 5.733041575492342, "grad_norm": 0.41197437047958374, "learning_rate": 4.2054999542028174e-05, "loss": 0.0553, "step": 2620 }, { "epoch": 5.75492341356674, "grad_norm": 0.49191319942474365, "learning_rate": 4.169795689153535e-05, "loss": 0.0565, "step": 2630 }, { "epoch": 5.776805251641138, "grad_norm": 0.5994260907173157, "learning_rate": 4.134134905478566e-05, "loss": 0.0608, "step": 2640 }, { "epoch": 5.7986870897155365, "grad_norm": 0.5599576234817505, "learning_rate": 4.0985194708866904e-05, "loss": 0.0552, "step": 2650 }, { "epoch": 5.820568927789934, "grad_norm": 0.48255977034568787, "learning_rate": 4.06295125071157e-05, "loss": 0.0587, "step": 2660 }, { "epoch": 5.842450765864332, "grad_norm": 0.5754146575927734, "learning_rate": 4.027432107814032e-05, "loss": 0.0604, "step": 2670 }, { "epoch": 5.864332603938731, "grad_norm": 0.6725056767463684, "learning_rate": 3.991963902484527e-05, "loss": 0.0594, "step": 2680 }, { "epoch": 5.886214442013129, "grad_norm": 0.5410199761390686, "learning_rate": 3.956548492345681e-05, "loss": 0.0606, "step": 2690 }, { "epoch": 5.908096280087527, "grad_norm": 0.2633671164512634, "learning_rate": 3.921187732255006e-05, "loss": 0.053, "step": 2700 }, { "epoch": 5.929978118161926, "grad_norm": 0.3890124261379242, "learning_rate": 3.8858834742077624e-05, "loss": 0.056, "step": 2710 }, { "epoch": 5.951859956236324, "grad_norm": 0.4105083644390106, "learning_rate": 3.8506375672399466e-05, "loss": 0.0513, "step": 2720 }, { "epoch": 5.9737417943107225, "grad_norm": 0.4586251378059387, "learning_rate": 3.815451857331465e-05, "loss": 0.0557, "step": 2730 }, { "epoch": 5.99562363238512, "grad_norm": 0.23562565445899963, "learning_rate": 3.780328187309443e-05, "loss": 0.0536, "step": 2740 }, { "epoch": 6.017505470459518, "grad_norm": 0.4445761740207672, "learning_rate": 3.7452683967517057e-05, "loss": 0.0475, "step": 2750 }, { "epoch": 6.039387308533917, "grad_norm": 0.7048770785331726, "learning_rate": 3.710274321890443e-05, "loss": 0.0506, "step": 2760 }, { "epoch": 6.061269146608315, "grad_norm": 0.41764745116233826, "learning_rate": 3.6753477955160244e-05, "loss": 0.0533, "step": 2770 }, { "epoch": 6.083150984682713, "grad_norm": 0.30127957463264465, "learning_rate": 3.640490646881018e-05, "loss": 0.0571, "step": 2780 }, { "epoch": 6.105032822757112, "grad_norm": 0.3346746563911438, "learning_rate": 3.605704701604381e-05, "loss": 0.051, "step": 2790 }, { "epoch": 6.12691466083151, "grad_norm": 0.5310165882110596, "learning_rate": 3.5709917815758385e-05, "loss": 0.0546, "step": 2800 }, { "epoch": 6.1487964989059085, "grad_norm": 0.3638152778148651, "learning_rate": 3.536353704860479e-05, "loss": 0.049, "step": 2810 }, { "epoch": 6.170678336980306, "grad_norm": 0.37336069345474243, "learning_rate": 3.5017922856035155e-05, "loss": 0.0582, "step": 2820 }, { "epoch": 6.192560175054704, "grad_norm": 0.5262767672538757, "learning_rate": 3.467309333935283e-05, "loss": 0.0517, "step": 2830 }, { "epoch": 6.214442013129103, "grad_norm": 0.28704574704170227, "learning_rate": 3.4329066558764355e-05, "loss": 0.0458, "step": 2840 }, { "epoch": 6.236323851203501, "grad_norm": 0.4320286810398102, "learning_rate": 3.398586053243341e-05, "loss": 0.0489, "step": 2850 }, { "epoch": 6.258205689277899, "grad_norm": 0.43195730447769165, "learning_rate": 3.364349323553738e-05, "loss": 0.054, "step": 2860 }, { "epoch": 6.280087527352298, "grad_norm": 0.3144959509372711, "learning_rate": 3.330198259932567e-05, "loss": 0.0537, "step": 2870 }, { "epoch": 6.301969365426696, "grad_norm": 0.4727848172187805, "learning_rate": 3.2961346510180716e-05, "loss": 0.0599, "step": 2880 }, { "epoch": 6.3238512035010945, "grad_norm": 0.2918757498264313, "learning_rate": 3.26216028086812e-05, "loss": 0.057, "step": 2890 }, { "epoch": 6.3457330415754925, "grad_norm": 0.8075538873672485, "learning_rate": 3.2282769288667525e-05, "loss": 0.0583, "step": 2900 }, { "epoch": 6.36761487964989, "grad_norm": 0.25191056728363037, "learning_rate": 3.194486369631007e-05, "loss": 0.0558, "step": 2910 }, { "epoch": 6.389496717724289, "grad_norm": 0.3775458335876465, "learning_rate": 3.1607903729179576e-05, "loss": 0.0508, "step": 2920 }, { "epoch": 6.411378555798687, "grad_norm": 0.4212424159049988, "learning_rate": 3.127190703532036e-05, "loss": 0.0581, "step": 2930 }, { "epoch": 6.433260393873085, "grad_norm": 0.3469920754432678, "learning_rate": 3.0936891212325966e-05, "loss": 0.0537, "step": 2940 }, { "epoch": 6.455142231947484, "grad_norm": 0.3886861801147461, "learning_rate": 3.0602873806417484e-05, "loss": 0.0601, "step": 2950 }, { "epoch": 6.477024070021882, "grad_norm": 0.4389936923980713, "learning_rate": 3.0269872311524626e-05, "loss": 0.0554, "step": 2960 }, { "epoch": 6.49890590809628, "grad_norm": 0.4016413390636444, "learning_rate": 2.9937904168369446e-05, "loss": 0.0505, "step": 2970 }, { "epoch": 6.5207877461706785, "grad_norm": 0.4214434325695038, "learning_rate": 2.9606986763552935e-05, "loss": 0.0484, "step": 2980 }, { "epoch": 6.542669584245076, "grad_norm": 0.432650625705719, "learning_rate": 2.927713742864442e-05, "loss": 0.0544, "step": 2990 }, { "epoch": 6.564551422319475, "grad_norm": 0.22795255482196808, "learning_rate": 2.8948373439273734e-05, "loss": 0.0496, "step": 3000 }, { "epoch": 6.586433260393873, "grad_norm": 0.3206609785556793, "learning_rate": 2.862071201422659e-05, "loss": 0.0478, "step": 3010 }, { "epoch": 6.608315098468271, "grad_norm": 0.4075492024421692, "learning_rate": 2.8294170314542567e-05, "loss": 0.0528, "step": 3020 }, { "epoch": 6.63019693654267, "grad_norm": 0.2718966007232666, "learning_rate": 2.796876544261645e-05, "loss": 0.0492, "step": 3030 }, { "epoch": 6.652078774617068, "grad_norm": 0.33011844754219055, "learning_rate": 2.7644514441302466e-05, "loss": 0.0595, "step": 3040 }, { "epoch": 6.673960612691467, "grad_norm": 0.3843638002872467, "learning_rate": 2.732143429302165e-05, "loss": 0.0598, "step": 3050 }, { "epoch": 6.6958424507658645, "grad_norm": 0.2999420166015625, "learning_rate": 2.699954191887244e-05, "loss": 0.0517, "step": 3060 }, { "epoch": 6.717724288840262, "grad_norm": 0.4345828890800476, "learning_rate": 2.6678854177744416e-05, "loss": 0.0623, "step": 3070 }, { "epoch": 6.739606126914661, "grad_norm": 0.4258360266685486, "learning_rate": 2.6359387865435346e-05, "loss": 0.0461, "step": 3080 }, { "epoch": 6.761487964989059, "grad_norm": 0.44895169138908386, "learning_rate": 2.6041159713771522e-05, "loss": 0.0558, "step": 3090 }, { "epoch": 6.783369803063457, "grad_norm": 0.47770532965660095, "learning_rate": 2.5724186389731364e-05, "loss": 0.0556, "step": 3100 }, { "epoch": 6.805251641137856, "grad_norm": 0.32761141657829285, "learning_rate": 2.5408484494572704e-05, "loss": 0.0593, "step": 3110 }, { "epoch": 6.827133479212254, "grad_norm": 0.47967493534088135, "learning_rate": 2.509407056296303e-05, "loss": 0.0563, "step": 3120 }, { "epoch": 6.849015317286652, "grad_norm": 0.3491036295890808, "learning_rate": 2.4780961062113684e-05, "loss": 0.0527, "step": 3130 }, { "epoch": 6.8708971553610505, "grad_norm": 0.32572147250175476, "learning_rate": 2.4469172390917417e-05, "loss": 0.0497, "step": 3140 }, { "epoch": 6.892778993435448, "grad_norm": 0.42841556668281555, "learning_rate": 2.4158720879089337e-05, "loss": 0.0487, "step": 3150 }, { "epoch": 6.914660831509847, "grad_norm": 0.3758542835712433, "learning_rate": 2.384962278631182e-05, "loss": 0.0582, "step": 3160 }, { "epoch": 6.936542669584245, "grad_norm": 0.32348281145095825, "learning_rate": 2.354189430138285e-05, "loss": 0.0558, "step": 3170 }, { "epoch": 6.958424507658643, "grad_norm": 0.3324337601661682, "learning_rate": 2.323555154136815e-05, "loss": 0.0486, "step": 3180 }, { "epoch": 6.980306345733042, "grad_norm": 0.3877069652080536, "learning_rate": 2.293061055075707e-05, "loss": 0.0534, "step": 3190 }, { "epoch": 7.00218818380744, "grad_norm": 0.3602635860443115, "learning_rate": 2.2627087300622208e-05, "loss": 0.0528, "step": 3200 }, { "epoch": 7.024070021881838, "grad_norm": 0.4806849956512451, "learning_rate": 2.2324997687783102e-05, "loss": 0.0529, "step": 3210 }, { "epoch": 7.0459518599562365, "grad_norm": 0.4764333963394165, "learning_rate": 2.20243575339734e-05, "loss": 0.0532, "step": 3220 }, { "epoch": 7.067833698030634, "grad_norm": 0.4383291006088257, "learning_rate": 2.172518258501241e-05, "loss": 0.0482, "step": 3230 }, { "epoch": 7.089715536105033, "grad_norm": 0.4792105257511139, "learning_rate": 2.1427488509980402e-05, "loss": 0.0508, "step": 3240 }, { "epoch": 7.111597374179431, "grad_norm": 0.4630696475505829, "learning_rate": 2.1131290900397793e-05, "loss": 0.0534, "step": 3250 }, { "epoch": 7.133479212253829, "grad_norm": 0.5735715627670288, "learning_rate": 2.0836605269408722e-05, "loss": 0.0517, "step": 3260 }, { "epoch": 7.155361050328228, "grad_norm": 0.3468039929866791, "learning_rate": 2.05434470509685e-05, "loss": 0.0485, "step": 3270 }, { "epoch": 7.177242888402626, "grad_norm": 0.27370214462280273, "learning_rate": 2.025183159903526e-05, "loss": 0.0564, "step": 3280 }, { "epoch": 7.199124726477024, "grad_norm": 0.43867847323417664, "learning_rate": 1.9961774186765796e-05, "loss": 0.0454, "step": 3290 }, { "epoch": 7.2210065645514225, "grad_norm": 0.36127227544784546, "learning_rate": 1.9673290005715683e-05, "loss": 0.0539, "step": 3300 }, { "epoch": 7.2428884026258205, "grad_norm": 0.3754895329475403, "learning_rate": 1.93863941650436e-05, "loss": 0.0466, "step": 3310 }, { "epoch": 7.264770240700219, "grad_norm": 0.3637754023075104, "learning_rate": 1.910110169071995e-05, "loss": 0.0494, "step": 3320 }, { "epoch": 7.286652078774617, "grad_norm": 0.5620509386062622, "learning_rate": 1.8817427524739988e-05, "loss": 0.046, "step": 3330 }, { "epoch": 7.308533916849015, "grad_norm": 0.5723752379417419, "learning_rate": 1.8535386524341225e-05, "loss": 0.0489, "step": 3340 }, { "epoch": 7.330415754923414, "grad_norm": 0.5323576927185059, "learning_rate": 1.8254993461225172e-05, "loss": 0.0522, "step": 3350 }, { "epoch": 7.352297592997812, "grad_norm": 0.40650686621665955, "learning_rate": 1.797626302078383e-05, "loss": 0.0489, "step": 3360 }, { "epoch": 7.37417943107221, "grad_norm": 0.4388726055622101, "learning_rate": 1.7699209801330468e-05, "loss": 0.0486, "step": 3370 }, { "epoch": 7.3960612691466086, "grad_norm": 0.292429119348526, "learning_rate": 1.7423848313335084e-05, "loss": 0.0464, "step": 3380 }, { "epoch": 7.4179431072210065, "grad_norm": 0.37241464853286743, "learning_rate": 1.715019297866441e-05, "loss": 0.0509, "step": 3390 }, { "epoch": 7.439824945295404, "grad_norm": 0.30430710315704346, "learning_rate": 1.6878258129826575e-05, "loss": 0.0501, "step": 3400 }, { "epoch": 7.461706783369803, "grad_norm": 0.3546319305896759, "learning_rate": 1.660805800922049e-05, "loss": 0.0563, "step": 3410 }, { "epoch": 7.483588621444201, "grad_norm": 0.3407646417617798, "learning_rate": 1.6339606768389792e-05, "loss": 0.048, "step": 3420 }, { "epoch": 7.5054704595186, "grad_norm": 0.31406334042549133, "learning_rate": 1.6072918467281872e-05, "loss": 0.0481, "step": 3430 }, { "epoch": 7.527352297592998, "grad_norm": 0.4938783645629883, "learning_rate": 1.5808007073511317e-05, "loss": 0.0477, "step": 3440 }, { "epoch": 7.549234135667396, "grad_norm": 0.3935304880142212, "learning_rate": 1.55448864616284e-05, "loss": 0.0478, "step": 3450 }, { "epoch": 7.571115973741795, "grad_norm": 0.5982438921928406, "learning_rate": 1.528357041239248e-05, "loss": 0.0552, "step": 3460 }, { "epoch": 7.5929978118161925, "grad_norm": 0.5672005414962769, "learning_rate": 1.502407261205019e-05, "loss": 0.0483, "step": 3470 }, { "epoch": 7.614879649890591, "grad_norm": 0.539139986038208, "learning_rate": 1.4766406651618626e-05, "loss": 0.0503, "step": 3480 }, { "epoch": 7.636761487964989, "grad_norm": 0.3828607201576233, "learning_rate": 1.4510586026173557e-05, "loss": 0.0547, "step": 3490 }, { "epoch": 7.658643326039387, "grad_norm": 0.40856969356536865, "learning_rate": 1.4256624134142604e-05, "loss": 0.046, "step": 3500 }, { "epoch": 7.680525164113786, "grad_norm": 0.8226433396339417, "learning_rate": 1.4004534276603538e-05, "loss": 0.0531, "step": 3510 }, { "epoch": 7.702407002188184, "grad_norm": 0.393865704536438, "learning_rate": 1.3754329656587556e-05, "loss": 0.0487, "step": 3520 }, { "epoch": 7.724288840262582, "grad_norm": 0.2557559907436371, "learning_rate": 1.3506023378387944e-05, "loss": 0.0459, "step": 3530 }, { "epoch": 7.746170678336981, "grad_norm": 0.3025050461292267, "learning_rate": 1.3259628446873601e-05, "loss": 0.0516, "step": 3540 }, { "epoch": 7.7680525164113785, "grad_norm": 0.46884235739707947, "learning_rate": 1.3015157766807939e-05, "loss": 0.0541, "step": 3550 }, { "epoch": 7.789934354485776, "grad_norm": 0.4090341627597809, "learning_rate": 1.2772624142173123e-05, "loss": 0.0547, "step": 3560 }, { "epoch": 7.811816192560175, "grad_norm": 0.39957642555236816, "learning_rate": 1.2532040275499301e-05, "loss": 0.0463, "step": 3570 }, { "epoch": 7.833698030634573, "grad_norm": 0.4493909776210785, "learning_rate": 1.2293418767199449e-05, "loss": 0.0522, "step": 3580 }, { "epoch": 7.855579868708972, "grad_norm": 0.40386393666267395, "learning_rate": 1.2056772114909365e-05, "loss": 0.0483, "step": 3590 }, { "epoch": 7.87746170678337, "grad_norm": 0.4808233380317688, "learning_rate": 1.182211271283315e-05, "loss": 0.0485, "step": 3600 }, { "epoch": 7.899343544857768, "grad_norm": 0.35203149914741516, "learning_rate": 1.1589452851094062e-05, "loss": 0.0489, "step": 3610 }, { "epoch": 7.921225382932167, "grad_norm": 0.37248197197914124, "learning_rate": 1.135880471509076e-05, "loss": 0.055, "step": 3620 }, { "epoch": 7.9431072210065645, "grad_norm": 0.36168205738067627, "learning_rate": 1.1130180384859256e-05, "loss": 0.0445, "step": 3630 }, { "epoch": 7.964989059080962, "grad_norm": 0.6294784545898438, "learning_rate": 1.0903591834440097e-05, "loss": 0.0494, "step": 3640 }, { "epoch": 7.986870897155361, "grad_norm": 0.46873342990875244, "learning_rate": 1.0679050931251238e-05, "loss": 0.0508, "step": 3650 }, { "epoch": 8.00875273522976, "grad_norm": 0.4308367967605591, "learning_rate": 1.0456569435466617e-05, "loss": 0.0524, "step": 3660 }, { "epoch": 8.030634573304157, "grad_norm": 0.45153898000717163, "learning_rate": 1.0236158999400053e-05, "loss": 0.0475, "step": 3670 }, { "epoch": 8.052516411378555, "grad_norm": 0.6072208881378174, "learning_rate": 1.0017831166895104e-05, "loss": 0.0478, "step": 3680 }, { "epoch": 8.074398249452955, "grad_norm": 0.34941720962524414, "learning_rate": 9.801597372720395e-06, "loss": 0.0466, "step": 3690 }, { "epoch": 8.096280087527353, "grad_norm": 0.32349899411201477, "learning_rate": 9.587468941970751e-06, "loss": 0.0434, "step": 3700 }, { "epoch": 8.11816192560175, "grad_norm": 0.4987752437591553, "learning_rate": 9.375457089474043e-06, "loss": 0.0457, "step": 3710 }, { "epoch": 8.140043763676148, "grad_norm": 0.3384380638599396, "learning_rate": 9.165572919203847e-06, "loss": 0.0539, "step": 3720 }, { "epoch": 8.161925601750546, "grad_norm": 0.4982927739620209, "learning_rate": 8.957827423697823e-06, "loss": 0.0482, "step": 3730 }, { "epoch": 8.183807439824946, "grad_norm": 0.26903849840164185, "learning_rate": 8.75223148348207e-06, "loss": 0.0529, "step": 3740 }, { "epoch": 8.205689277899344, "grad_norm": 0.49081680178642273, "learning_rate": 8.548795866501164e-06, "loss": 0.0444, "step": 3750 }, { "epoch": 8.227571115973742, "grad_norm": 0.37901756167411804, "learning_rate": 8.347531227554323e-06, "loss": 0.0457, "step": 3760 }, { "epoch": 8.24945295404814, "grad_norm": 0.40646421909332275, "learning_rate": 8.148448107737227e-06, "loss": 0.0447, "step": 3770 }, { "epoch": 8.271334792122538, "grad_norm": 0.6366869211196899, "learning_rate": 7.951556933890047e-06, "loss": 0.0555, "step": 3780 }, { "epoch": 8.293216630196937, "grad_norm": 0.508129358291626, "learning_rate": 7.756868018051322e-06, "loss": 0.0461, "step": 3790 }, { "epoch": 8.315098468271335, "grad_norm": 0.317501425743103, "learning_rate": 7.56439155691781e-06, "loss": 0.0456, "step": 3800 }, { "epoch": 8.336980306345733, "grad_norm": 0.34288403391838074, "learning_rate": 7.374137631310512e-06, "loss": 0.0498, "step": 3810 }, { "epoch": 8.358862144420131, "grad_norm": 0.44967618584632874, "learning_rate": 7.186116205646687e-06, "loss": 0.0475, "step": 3820 }, { "epoch": 8.38074398249453, "grad_norm": 0.6020100712776184, "learning_rate": 7.000337127417938e-06, "loss": 0.0489, "step": 3830 }, { "epoch": 8.402625820568927, "grad_norm": 0.37260666489601135, "learning_rate": 6.816810126674494e-06, "loss": 0.0483, "step": 3840 }, { "epoch": 8.424507658643327, "grad_norm": 0.4538766145706177, "learning_rate": 6.635544815515576e-06, "loss": 0.0448, "step": 3850 }, { "epoch": 8.446389496717725, "grad_norm": 0.4207398593425751, "learning_rate": 6.456550687586016e-06, "loss": 0.0484, "step": 3860 }, { "epoch": 8.468271334792123, "grad_norm": 0.3660431504249573, "learning_rate": 6.2798371175789405e-06, "loss": 0.0557, "step": 3870 }, { "epoch": 8.49015317286652, "grad_norm": 0.3074619472026825, "learning_rate": 6.105413360744883e-06, "loss": 0.0506, "step": 3880 }, { "epoch": 8.512035010940918, "grad_norm": 0.41881874203681946, "learning_rate": 5.933288552407018e-06, "loss": 0.0468, "step": 3890 }, { "epoch": 8.533916849015318, "grad_norm": 0.6764019131660461, "learning_rate": 5.763471707482638e-06, "loss": 0.0459, "step": 3900 }, { "epoch": 8.555798687089716, "grad_norm": 0.5430788993835449, "learning_rate": 5.5959717200110896e-06, "loss": 0.047, "step": 3910 }, { "epoch": 8.577680525164114, "grad_norm": 0.256409227848053, "learning_rate": 5.430797362687906e-06, "loss": 0.0458, "step": 3920 }, { "epoch": 8.599562363238512, "grad_norm": 0.3550792932510376, "learning_rate": 5.267957286405368e-06, "loss": 0.0461, "step": 3930 }, { "epoch": 8.62144420131291, "grad_norm": 0.27031660079956055, "learning_rate": 5.107460019799387e-06, "loss": 0.044, "step": 3940 }, { "epoch": 8.643326039387308, "grad_norm": 0.5709757208824158, "learning_rate": 4.949313968802871e-06, "loss": 0.0519, "step": 3950 }, { "epoch": 8.665207877461707, "grad_norm": 0.42753544449806213, "learning_rate": 4.793527416205429e-06, "loss": 0.0422, "step": 3960 }, { "epoch": 8.687089715536105, "grad_norm": 0.6058006286621094, "learning_rate": 4.640108521219561e-06, "loss": 0.0518, "step": 3970 }, { "epoch": 8.708971553610503, "grad_norm": 0.3088013827800751, "learning_rate": 4.489065319053376e-06, "loss": 0.0475, "step": 3980 }, { "epoch": 8.730853391684901, "grad_norm": 0.6196874976158142, "learning_rate": 4.340405720489721e-06, "loss": 0.0455, "step": 3990 }, { "epoch": 8.752735229759299, "grad_norm": 0.3027147054672241, "learning_rate": 4.194137511471824e-06, "loss": 0.0433, "step": 4000 }, { "epoch": 8.774617067833699, "grad_norm": 0.34994643926620483, "learning_rate": 4.0502683526955745e-06, "loss": 0.0438, "step": 4010 }, { "epoch": 8.796498905908097, "grad_norm": 0.3832307457923889, "learning_rate": 3.908805779208269e-06, "loss": 0.0415, "step": 4020 }, { "epoch": 8.818380743982495, "grad_norm": 0.2572057545185089, "learning_rate": 3.7697572000139626e-06, "loss": 0.0491, "step": 4030 }, { "epoch": 8.840262582056893, "grad_norm": 0.5285372138023376, "learning_rate": 3.633129897685439e-06, "loss": 0.0498, "step": 4040 }, { "epoch": 8.86214442013129, "grad_norm": 0.44389864802360535, "learning_rate": 3.4989310279827938e-06, "loss": 0.0495, "step": 4050 }, { "epoch": 8.88402625820569, "grad_norm": 0.3135242760181427, "learning_rate": 3.3671676194786504e-06, "loss": 0.0454, "step": 4060 }, { "epoch": 8.905908096280088, "grad_norm": 0.29390549659729004, "learning_rate": 3.2378465731900064e-06, "loss": 0.0438, "step": 4070 }, { "epoch": 8.927789934354486, "grad_norm": 0.4052577018737793, "learning_rate": 3.1109746622168922e-06, "loss": 0.0454, "step": 4080 }, { "epoch": 8.949671772428884, "grad_norm": 0.7273414134979248, "learning_rate": 2.9865585313875565e-06, "loss": 0.0463, "step": 4090 }, { "epoch": 8.971553610503282, "grad_norm": 0.3757275640964508, "learning_rate": 2.8646046969104335e-06, "loss": 0.0457, "step": 4100 }, { "epoch": 8.99343544857768, "grad_norm": 0.4105775058269501, "learning_rate": 2.7451195460329315e-06, "loss": 0.0497, "step": 4110 }, { "epoch": 9.01531728665208, "grad_norm": 0.3794170916080475, "learning_rate": 2.6281093367068743e-06, "loss": 0.0423, "step": 4120 }, { "epoch": 9.037199124726477, "grad_norm": 0.28719809651374817, "learning_rate": 2.513580197260712e-06, "loss": 0.0441, "step": 4130 }, { "epoch": 9.059080962800875, "grad_norm": 0.38277560472488403, "learning_rate": 2.401538126078606e-06, "loss": 0.0423, "step": 4140 }, { "epoch": 9.080962800875273, "grad_norm": 0.43671509623527527, "learning_rate": 2.2919889912862313e-06, "loss": 0.0436, "step": 4150 }, { "epoch": 9.102844638949671, "grad_norm": 0.3907598853111267, "learning_rate": 2.1849385304434644e-06, "loss": 0.0485, "step": 4160 }, { "epoch": 9.12472647702407, "grad_norm": 0.38813626766204834, "learning_rate": 2.080392350243837e-06, "loss": 0.0443, "step": 4170 }, { "epoch": 9.146608315098469, "grad_norm": 0.4304605722427368, "learning_rate": 1.978355926220965e-06, "loss": 0.0478, "step": 4180 }, { "epoch": 9.168490153172867, "grad_norm": 0.3653641939163208, "learning_rate": 1.8788346024617044e-06, "loss": 0.0481, "step": 4190 }, { "epoch": 9.190371991247265, "grad_norm": 0.49059316515922546, "learning_rate": 1.7818335913262708e-06, "loss": 0.0482, "step": 4200 }, { "epoch": 9.212253829321662, "grad_norm": 0.6487129330635071, "learning_rate": 1.6873579731752798e-06, "loss": 0.0439, "step": 4210 }, { "epoch": 9.23413566739606, "grad_norm": 0.3429771959781647, "learning_rate": 1.5954126961036208e-06, "loss": 0.0452, "step": 4220 }, { "epoch": 9.25601750547046, "grad_norm": 0.23548723757266998, "learning_rate": 1.5060025756813423e-06, "loss": 0.0419, "step": 4230 }, { "epoch": 9.277899343544858, "grad_norm": 0.3810988962650299, "learning_rate": 1.4191322947014197e-06, "loss": 0.0467, "step": 4240 }, { "epoch": 9.299781181619256, "grad_norm": 0.4621908664703369, "learning_rate": 1.3348064029344908e-06, "loss": 0.0521, "step": 4250 }, { "epoch": 9.321663019693654, "grad_norm": 0.6988676190376282, "learning_rate": 1.2530293168905916e-06, "loss": 0.0502, "step": 4260 }, { "epoch": 9.343544857768052, "grad_norm": 0.2640683352947235, "learning_rate": 1.1738053195878174e-06, "loss": 0.0494, "step": 4270 }, { "epoch": 9.365426695842451, "grad_norm": 0.3481920063495636, "learning_rate": 1.097138560328015e-06, "loss": 0.0477, "step": 4280 }, { "epoch": 9.38730853391685, "grad_norm": 0.42929619550704956, "learning_rate": 1.023033054479472e-06, "loss": 0.0487, "step": 4290 }, { "epoch": 9.409190371991247, "grad_norm": 0.5334005951881409, "learning_rate": 9.51492683266586e-07, "loss": 0.051, "step": 4300 }, { "epoch": 9.431072210065645, "grad_norm": 0.3841243088245392, "learning_rate": 8.825211935666433e-07, "loss": 0.0478, "step": 4310 }, { "epoch": 9.452954048140043, "grad_norm": 0.34734562039375305, "learning_rate": 8.161221977135092e-07, "loss": 0.0438, "step": 4320 }, { "epoch": 9.474835886214443, "grad_norm": 0.6244210004806519, "learning_rate": 7.522991733084905e-07, "loss": 0.0473, "step": 4330 }, { "epoch": 9.49671772428884, "grad_norm": 0.4562375247478485, "learning_rate": 6.910554630381815e-07, "loss": 0.0486, "step": 4340 }, { "epoch": 9.518599562363239, "grad_norm": 0.8920162916183472, "learning_rate": 6.323942744993761e-07, "loss": 0.0479, "step": 4350 }, { "epoch": 9.540481400437637, "grad_norm": 0.46890518069267273, "learning_rate": 5.763186800310849e-07, "loss": 0.0473, "step": 4360 }, { "epoch": 9.562363238512035, "grad_norm": 0.37694522738456726, "learning_rate": 5.228316165536429e-07, "loss": 0.0464, "step": 4370 }, { "epoch": 9.584245076586432, "grad_norm": 0.42408043146133423, "learning_rate": 4.7193588541484813e-07, "loss": 0.0494, "step": 4380 }, { "epoch": 9.606126914660832, "grad_norm": 0.3436243534088135, "learning_rate": 4.2363415224329075e-07, "loss": 0.0446, "step": 4390 }, { "epoch": 9.62800875273523, "grad_norm": 0.32050296664237976, "learning_rate": 3.779289468086922e-07, "loss": 0.0506, "step": 4400 }, { "epoch": 9.649890590809628, "grad_norm": 0.5982491970062256, "learning_rate": 3.3482266288946686e-07, "loss": 0.0458, "step": 4410 }, { "epoch": 9.671772428884026, "grad_norm": 0.5433162450790405, "learning_rate": 2.9431755814729457e-07, "loss": 0.0465, "step": 4420 }, { "epoch": 9.693654266958424, "grad_norm": 0.4410666823387146, "learning_rate": 2.5641575400890403e-07, "loss": 0.0452, "step": 4430 }, { "epoch": 9.715536105032823, "grad_norm": 0.2922728955745697, "learning_rate": 2.2111923555498405e-07, "loss": 0.0464, "step": 4440 }, { "epoch": 9.737417943107221, "grad_norm": 0.4933525323867798, "learning_rate": 1.8842985141617197e-07, "loss": 0.048, "step": 4450 }, { "epoch": 9.75929978118162, "grad_norm": 0.7095863223075867, "learning_rate": 1.5834931367625928e-07, "loss": 0.0462, "step": 4460 }, { "epoch": 9.781181619256017, "grad_norm": 0.5306093096733093, "learning_rate": 1.3087919778252966e-07, "loss": 0.0496, "step": 4470 }, { "epoch": 9.803063457330415, "grad_norm": 0.31279975175857544, "learning_rate": 1.060209424632308e-07, "loss": 0.0431, "step": 4480 }, { "epoch": 9.824945295404813, "grad_norm": 0.5052768588066101, "learning_rate": 8.377584965221785e-08, "loss": 0.0493, "step": 4490 }, { "epoch": 9.846827133479213, "grad_norm": 0.4014386534690857, "learning_rate": 6.414508442078026e-08, "loss": 0.0478, "step": 4500 }, { "epoch": 9.86870897155361, "grad_norm": 0.2997543215751648, "learning_rate": 4.7129674916618346e-08, "loss": 0.0451, "step": 4510 }, { "epoch": 9.890590809628009, "grad_norm": 0.5213238000869751, "learning_rate": 3.273051230999191e-08, "loss": 0.0436, "step": 4520 }, { "epoch": 9.912472647702407, "grad_norm": 0.5938975811004639, "learning_rate": 2.0948350747046487e-08, "loss": 0.0417, "step": 4530 }, { "epoch": 9.934354485776804, "grad_norm": 0.26146259903907776, "learning_rate": 1.1783807310300488e-08, "loss": 0.0456, "step": 4540 }, { "epoch": 9.956236323851204, "grad_norm": 0.238664910197258, "learning_rate": 5.237361986365486e-09, "loss": 0.0456, "step": 4550 }, { "epoch": 9.978118161925602, "grad_norm": 0.3414470851421356, "learning_rate": 1.30935764076634e-09, "loss": 0.0471, "step": 4560 }, { "epoch": 10.0, "grad_norm": 0.492483526468277, "learning_rate": 0.0, "loss": 0.0444, "step": 4570 }, { "epoch": 10.0, "step": 4570, "total_flos": 4.682127244666608e+17, "train_loss": 0.07182003624199255, "train_runtime": 4395.4484, "train_samples_per_second": 50.928, "train_steps_per_second": 1.04 } ], "logging_steps": 10, "max_steps": 4570, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4.682127244666608e+17, "train_batch_size": 49, "trial_name": null, "trial_params": null }