{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2093, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0023894862604540022, "grad_norm": 1.6310280561447144, "learning_rate": 6.349206349206349e-06, "loss": 2.949186325073242, "step": 5 }, { "epoch": 0.0047789725209080045, "grad_norm": 1.1911828517913818, "learning_rate": 1.4285714285714285e-05, "loss": 2.8994724273681642, "step": 10 }, { "epoch": 0.007168458781362007, "grad_norm": 0.4962977468967438, "learning_rate": 2.2222222222222223e-05, "loss": 2.7720775604248047, "step": 15 }, { "epoch": 0.009557945041816009, "grad_norm": 0.3306114077568054, "learning_rate": 3.0158730158730158e-05, "loss": 2.6373291015625, "step": 20 }, { "epoch": 0.011947431302270013, "grad_norm": 0.3313261568546295, "learning_rate": 3.809523809523809e-05, "loss": 2.508778381347656, "step": 25 }, { "epoch": 0.014336917562724014, "grad_norm": 0.2656371295452118, "learning_rate": 4.603174603174603e-05, "loss": 2.427614212036133, "step": 30 }, { "epoch": 0.016726403823178016, "grad_norm": 0.2168935239315033, "learning_rate": 5.396825396825397e-05, "loss": 2.317409896850586, "step": 35 }, { "epoch": 0.019115890083632018, "grad_norm": 0.19373714923858643, "learning_rate": 6.19047619047619e-05, "loss": 2.2381568908691407, "step": 40 }, { "epoch": 0.021505376344086023, "grad_norm": 0.18095631897449493, "learning_rate": 6.984126984126984e-05, "loss": 2.1587665557861326, "step": 45 }, { "epoch": 0.023894862604540025, "grad_norm": 0.22285227477550507, "learning_rate": 7.777777777777778e-05, "loss": 2.0901668548583983, "step": 50 }, { "epoch": 0.026284348864994027, "grad_norm": 0.18036071956157684, "learning_rate": 8.571428571428571e-05, "loss": 2.0653894424438475, "step": 55 }, { "epoch": 0.02867383512544803, "grad_norm": 0.17877984046936035, "learning_rate": 9.365079365079366e-05, "loss": 2.0302642822265624, "step": 60 }, { "epoch": 0.03106332138590203, "grad_norm": 0.20096184313297272, "learning_rate": 9.99999401247153e-05, "loss": 1.925634765625, "step": 65 }, { "epoch": 0.03345280764635603, "grad_norm": 0.18011413514614105, "learning_rate": 9.999784450480743e-05, "loss": 1.9200904846191407, "step": 70 }, { "epoch": 0.035842293906810034, "grad_norm": 0.17637914419174194, "learning_rate": 9.999275526406429e-05, "loss": 1.8704423904418945, "step": 75 }, { "epoch": 0.038231780167264036, "grad_norm": 0.22258557379245758, "learning_rate": 9.99846727072041e-05, "loss": 1.8649562835693358, "step": 80 }, { "epoch": 0.04062126642771804, "grad_norm": 0.25038355588912964, "learning_rate": 9.997359731816998e-05, "loss": 1.815376091003418, "step": 85 }, { "epoch": 0.043010752688172046, "grad_norm": 0.20416149497032166, "learning_rate": 9.995952976010081e-05, "loss": 1.7895587921142577, "step": 90 }, { "epoch": 0.04540023894862605, "grad_norm": 0.18203961849212646, "learning_rate": 9.994247087529158e-05, "loss": 1.7791255950927733, "step": 95 }, { "epoch": 0.04778972520908005, "grad_norm": 0.23035629093647003, "learning_rate": 9.9922421685143e-05, "loss": 1.784566879272461, "step": 100 }, { "epoch": 0.05017921146953405, "grad_norm": 0.20310822129249573, "learning_rate": 9.989938339010031e-05, "loss": 1.7500259399414062, "step": 105 }, { "epoch": 0.052568697729988054, "grad_norm": 0.21497943997383118, "learning_rate": 9.987335736958134e-05, "loss": 1.7058338165283202, "step": 110 }, { "epoch": 0.054958183990442055, "grad_norm": 0.19419395923614502, "learning_rate": 9.984434518189404e-05, "loss": 1.7215375900268555, "step": 115 }, { "epoch": 0.05734767025089606, "grad_norm": 0.2528029978275299, "learning_rate": 9.981234856414307e-05, "loss": 1.7244646072387695, "step": 120 }, { "epoch": 0.05973715651135006, "grad_norm": 0.23098044097423553, "learning_rate": 9.977736943212586e-05, "loss": 1.6946876525878907, "step": 125 }, { "epoch": 0.06212664277180406, "grad_norm": 0.24221175909042358, "learning_rate": 9.973940988021786e-05, "loss": 1.6946578979492188, "step": 130 }, { "epoch": 0.06451612903225806, "grad_norm": 0.26225730776786804, "learning_rate": 9.969847218124717e-05, "loss": 1.6926054000854491, "step": 135 }, { "epoch": 0.06690561529271206, "grad_norm": 0.2752564251422882, "learning_rate": 9.965455878635842e-05, "loss": 1.6920223236083984, "step": 140 }, { "epoch": 0.06929510155316607, "grad_norm": 0.23380827903747559, "learning_rate": 9.960767232486605e-05, "loss": 1.6452816009521485, "step": 145 }, { "epoch": 0.07168458781362007, "grad_norm": 0.23962828516960144, "learning_rate": 9.955781560409684e-05, "loss": 1.6491233825683593, "step": 150 }, { "epoch": 0.07407407407407407, "grad_norm": 0.21389207243919373, "learning_rate": 9.950499160922183e-05, "loss": 1.6587303161621094, "step": 155 }, { "epoch": 0.07646356033452807, "grad_norm": 0.23845191299915314, "learning_rate": 9.944920350307762e-05, "loss": 1.652520751953125, "step": 160 }, { "epoch": 0.07885304659498207, "grad_norm": 0.23529696464538574, "learning_rate": 9.939045462597694e-05, "loss": 1.6252538681030273, "step": 165 }, { "epoch": 0.08124253285543608, "grad_norm": 0.2364918291568756, "learning_rate": 9.932874849550868e-05, "loss": 1.620565414428711, "step": 170 }, { "epoch": 0.08363201911589008, "grad_norm": 0.24278566241264343, "learning_rate": 9.926408880632727e-05, "loss": 1.643486785888672, "step": 175 }, { "epoch": 0.08602150537634409, "grad_norm": 0.31502866744995117, "learning_rate": 9.919647942993148e-05, "loss": 1.5687498092651366, "step": 180 }, { "epoch": 0.0884109916367981, "grad_norm": 0.2938617467880249, "learning_rate": 9.912592441443259e-05, "loss": 1.597340965270996, "step": 185 }, { "epoch": 0.0908004778972521, "grad_norm": 0.2505309581756592, "learning_rate": 9.905242798431196e-05, "loss": 1.6328857421875, "step": 190 }, { "epoch": 0.0931899641577061, "grad_norm": 0.21380369365215302, "learning_rate": 9.897599454016823e-05, "loss": 1.6188299179077148, "step": 195 }, { "epoch": 0.0955794504181601, "grad_norm": 0.23083806037902832, "learning_rate": 9.88966286584537e-05, "loss": 1.570663833618164, "step": 200 }, { "epoch": 0.0979689366786141, "grad_norm": 0.25395503640174866, "learning_rate": 9.881433509120037e-05, "loss": 1.632042694091797, "step": 205 }, { "epoch": 0.1003584229390681, "grad_norm": 0.21581228077411652, "learning_rate": 9.87291187657354e-05, "loss": 1.630429458618164, "step": 210 }, { "epoch": 0.1027479091995221, "grad_norm": 0.23620952665805817, "learning_rate": 9.864098478438611e-05, "loss": 1.5788474082946777, "step": 215 }, { "epoch": 0.10513739545997611, "grad_norm": 0.25899890065193176, "learning_rate": 9.854993842417444e-05, "loss": 1.5768406867980957, "step": 220 }, { "epoch": 0.10752688172043011, "grad_norm": 0.2653222382068634, "learning_rate": 9.845598513650103e-05, "loss": 1.5978818893432618, "step": 225 }, { "epoch": 0.10991636798088411, "grad_norm": 0.26710769534111023, "learning_rate": 9.835913054681877e-05, "loss": 1.5802912712097168, "step": 230 }, { "epoch": 0.11230585424133811, "grad_norm": 0.27862873673439026, "learning_rate": 9.8259380454296e-05, "loss": 1.5534930229187012, "step": 235 }, { "epoch": 0.11469534050179211, "grad_norm": 0.2402317225933075, "learning_rate": 9.815674083146936e-05, "loss": 1.5752450942993164, "step": 240 }, { "epoch": 0.11708482676224612, "grad_norm": 0.26304373145103455, "learning_rate": 9.8051217823886e-05, "loss": 1.5972543716430665, "step": 245 }, { "epoch": 0.11947431302270012, "grad_norm": 0.24205845594406128, "learning_rate": 9.794281774973578e-05, "loss": 1.5491343498229981, "step": 250 }, { "epoch": 0.12186379928315412, "grad_norm": 0.2603611946105957, "learning_rate": 9.783154709947294e-05, "loss": 1.5475645065307617, "step": 255 }, { "epoch": 0.12425328554360812, "grad_norm": 0.27649691700935364, "learning_rate": 9.771741253542741e-05, "loss": 1.5449405670166017, "step": 260 }, { "epoch": 0.12664277180406214, "grad_norm": 0.24233537912368774, "learning_rate": 9.760042089140598e-05, "loss": 1.56983642578125, "step": 265 }, { "epoch": 0.12903225806451613, "grad_norm": 0.2796640694141388, "learning_rate": 9.74805791722831e-05, "loss": 1.513641929626465, "step": 270 }, { "epoch": 0.13142174432497014, "grad_norm": 0.2519722878932953, "learning_rate": 9.735789455358145e-05, "loss": 1.606661033630371, "step": 275 }, { "epoch": 0.13381123058542413, "grad_norm": 0.3342560827732086, "learning_rate": 9.72323743810423e-05, "loss": 1.5392388343811034, "step": 280 }, { "epoch": 0.13620071684587814, "grad_norm": 0.27568939328193665, "learning_rate": 9.710402617018575e-05, "loss": 1.5748723983764648, "step": 285 }, { "epoch": 0.13859020310633213, "grad_norm": 0.2670389711856842, "learning_rate": 9.697285760586063e-05, "loss": 1.5220499992370606, "step": 290 }, { "epoch": 0.14097968936678615, "grad_norm": 0.22553808987140656, "learning_rate": 9.683887654178445e-05, "loss": 1.538209819793701, "step": 295 }, { "epoch": 0.14336917562724014, "grad_norm": 0.2573106586933136, "learning_rate": 9.670209100007318e-05, "loss": 1.5534125328063966, "step": 300 }, { "epoch": 0.14575866188769415, "grad_norm": 0.24078048765659332, "learning_rate": 9.656250917076081e-05, "loss": 1.560771369934082, "step": 305 }, { "epoch": 0.14814814814814814, "grad_norm": 0.2561890184879303, "learning_rate": 9.642013941130909e-05, "loss": 1.5603245735168456, "step": 310 }, { "epoch": 0.15053763440860216, "grad_norm": 0.22784937918186188, "learning_rate": 9.627499024610707e-05, "loss": 1.503566074371338, "step": 315 }, { "epoch": 0.15292712066905614, "grad_norm": 0.31219249963760376, "learning_rate": 9.612707036596071e-05, "loss": 1.529867172241211, "step": 320 }, { "epoch": 0.15531660692951016, "grad_norm": 0.23493337631225586, "learning_rate": 9.597638862757255e-05, "loss": 1.5141928672790528, "step": 325 }, { "epoch": 0.15770609318996415, "grad_norm": 0.24456432461738586, "learning_rate": 9.582295405301131e-05, "loss": 1.5292648315429687, "step": 330 }, { "epoch": 0.16009557945041816, "grad_norm": 0.2798076868057251, "learning_rate": 9.566677582917185e-05, "loss": 1.5751214981079102, "step": 335 }, { "epoch": 0.16248506571087215, "grad_norm": 0.23931868374347687, "learning_rate": 9.550786330722502e-05, "loss": 1.5360325813293456, "step": 340 }, { "epoch": 0.16487455197132617, "grad_norm": 0.27479445934295654, "learning_rate": 9.534622600205769e-05, "loss": 1.5060487747192384, "step": 345 }, { "epoch": 0.16726403823178015, "grad_norm": 0.24661928415298462, "learning_rate": 9.51818735917032e-05, "loss": 1.518015480041504, "step": 350 }, { "epoch": 0.16965352449223417, "grad_norm": 0.22562278807163239, "learning_rate": 9.501481591676177e-05, "loss": 1.5276756286621094, "step": 355 }, { "epoch": 0.17204301075268819, "grad_norm": 0.26462316513061523, "learning_rate": 9.484506297981137e-05, "loss": 1.5334238052368163, "step": 360 }, { "epoch": 0.17443249701314217, "grad_norm": 0.22379732131958008, "learning_rate": 9.467262494480869e-05, "loss": 1.5265101432800292, "step": 365 }, { "epoch": 0.1768219832735962, "grad_norm": 0.24132543802261353, "learning_rate": 9.449751213648075e-05, "loss": 1.536444854736328, "step": 370 }, { "epoch": 0.17921146953405018, "grad_norm": 0.24024462699890137, "learning_rate": 9.431973503970655e-05, "loss": 1.5176925659179688, "step": 375 }, { "epoch": 0.1816009557945042, "grad_norm": 0.26123809814453125, "learning_rate": 9.413930429888945e-05, "loss": 1.4924360275268556, "step": 380 }, { "epoch": 0.18399044205495818, "grad_norm": 0.2500710189342499, "learning_rate": 9.39562307173196e-05, "loss": 1.505857276916504, "step": 385 }, { "epoch": 0.1863799283154122, "grad_norm": 0.24786685407161713, "learning_rate": 9.377052525652739e-05, "loss": 1.5032339096069336, "step": 390 }, { "epoch": 0.18876941457586618, "grad_norm": 0.2467229813337326, "learning_rate": 9.358219903562685e-05, "loss": 1.496539306640625, "step": 395 }, { "epoch": 0.1911589008363202, "grad_norm": 0.23587468266487122, "learning_rate": 9.339126333065007e-05, "loss": 1.4925189018249512, "step": 400 }, { "epoch": 0.1935483870967742, "grad_norm": 0.2529931664466858, "learning_rate": 9.3197729573872e-05, "loss": 1.4608583450317383, "step": 405 }, { "epoch": 0.1959378733572282, "grad_norm": 0.21555373072624207, "learning_rate": 9.300160935312591e-05, "loss": 1.5208477973937988, "step": 410 }, { "epoch": 0.1983273596176822, "grad_norm": 0.21678751707077026, "learning_rate": 9.280291441110963e-05, "loss": 1.521120834350586, "step": 415 }, { "epoch": 0.2007168458781362, "grad_norm": 0.24857617914676666, "learning_rate": 9.260165664468237e-05, "loss": 1.4850285530090332, "step": 420 }, { "epoch": 0.2031063321385902, "grad_norm": 0.2462393194437027, "learning_rate": 9.239784810415249e-05, "loss": 1.4648225784301758, "step": 425 }, { "epoch": 0.2054958183990442, "grad_norm": 0.24047738313674927, "learning_rate": 9.219150099255591e-05, "loss": 1.47837495803833, "step": 430 }, { "epoch": 0.2078853046594982, "grad_norm": 0.2716653347015381, "learning_rate": 9.198262766492554e-05, "loss": 1.437627124786377, "step": 435 }, { "epoch": 0.21027479091995221, "grad_norm": 0.24831922352313995, "learning_rate": 9.17712406275514e-05, "loss": 1.4409489631652832, "step": 440 }, { "epoch": 0.2126642771804062, "grad_norm": 0.25995934009552, "learning_rate": 9.155735253723193e-05, "loss": 1.5035175323486327, "step": 445 }, { "epoch": 0.21505376344086022, "grad_norm": 0.22890552878379822, "learning_rate": 9.13409762005161e-05, "loss": 1.4517516136169433, "step": 450 }, { "epoch": 0.2174432497013142, "grad_norm": 0.3182690739631653, "learning_rate": 9.112212457293659e-05, "loss": 1.513246726989746, "step": 455 }, { "epoch": 0.21983273596176822, "grad_norm": 0.2540818154811859, "learning_rate": 9.090081075823416e-05, "loss": 1.4569884300231934, "step": 460 }, { "epoch": 0.2222222222222222, "grad_norm": 0.2668378949165344, "learning_rate": 9.067704800757301e-05, "loss": 1.4742938995361328, "step": 465 }, { "epoch": 0.22461170848267623, "grad_norm": 0.25422269105911255, "learning_rate": 9.045084971874738e-05, "loss": 1.5090333938598632, "step": 470 }, { "epoch": 0.2270011947431302, "grad_norm": 0.23948684334754944, "learning_rate": 9.02222294353793e-05, "loss": 1.4315285682678223, "step": 475 }, { "epoch": 0.22939068100358423, "grad_norm": 0.2671644985675812, "learning_rate": 8.99912008461078e-05, "loss": 1.5001513481140136, "step": 480 }, { "epoch": 0.23178016726403824, "grad_norm": 0.21840688586235046, "learning_rate": 8.975777778376915e-05, "loss": 1.4276542663574219, "step": 485 }, { "epoch": 0.23416965352449223, "grad_norm": 0.29224705696105957, "learning_rate": 8.952197422456875e-05, "loss": 1.4305816650390626, "step": 490 }, { "epoch": 0.23655913978494625, "grad_norm": 0.2511056363582611, "learning_rate": 8.928380428724419e-05, "loss": 1.4829061508178711, "step": 495 }, { "epoch": 0.23894862604540024, "grad_norm": 0.2419624626636505, "learning_rate": 8.904328223221998e-05, "loss": 1.4800556182861329, "step": 500 }, { "epoch": 0.24133811230585425, "grad_norm": 0.20764684677124023, "learning_rate": 8.880042246075365e-05, "loss": 1.498704719543457, "step": 505 }, { "epoch": 0.24372759856630824, "grad_norm": 0.25269007682800293, "learning_rate": 8.855523951407353e-05, "loss": 1.465312385559082, "step": 510 }, { "epoch": 0.24611708482676226, "grad_norm": 0.22692237794399261, "learning_rate": 8.830774807250803e-05, "loss": 1.4672849655151368, "step": 515 }, { "epoch": 0.24850657108721624, "grad_norm": 0.26884037256240845, "learning_rate": 8.805796295460669e-05, "loss": 1.4294979095458984, "step": 520 }, { "epoch": 0.25089605734767023, "grad_norm": 0.28854894638061523, "learning_rate": 8.780589911625293e-05, "loss": 1.4606175422668457, "step": 525 }, { "epoch": 0.2532855436081243, "grad_norm": 0.2832304537296295, "learning_rate": 8.755157164976855e-05, "loss": 1.4829923629760742, "step": 530 }, { "epoch": 0.25567502986857826, "grad_norm": 0.26198551058769226, "learning_rate": 8.729499578301005e-05, "loss": 1.4265382766723633, "step": 535 }, { "epoch": 0.25806451612903225, "grad_norm": 0.2522987425327301, "learning_rate": 8.703618687845696e-05, "loss": 1.446976375579834, "step": 540 }, { "epoch": 0.26045400238948624, "grad_norm": 0.26261770725250244, "learning_rate": 8.677516043229186e-05, "loss": 1.442029857635498, "step": 545 }, { "epoch": 0.2628434886499403, "grad_norm": 0.25842398405075073, "learning_rate": 8.65119320734727e-05, "loss": 1.4650904655456543, "step": 550 }, { "epoch": 0.26523297491039427, "grad_norm": 0.259551078081131, "learning_rate": 8.624651756279687e-05, "loss": 1.4772340774536132, "step": 555 }, { "epoch": 0.26762246117084826, "grad_norm": 0.2567656636238098, "learning_rate": 8.59789327919577e-05, "loss": 1.4593741416931152, "step": 560 }, { "epoch": 0.27001194743130225, "grad_norm": 0.25208425521850586, "learning_rate": 8.570919378259274e-05, "loss": 1.465700340270996, "step": 565 }, { "epoch": 0.2724014336917563, "grad_norm": 0.2418772131204605, "learning_rate": 8.543731668532463e-05, "loss": 1.4598386764526368, "step": 570 }, { "epoch": 0.2747909199522103, "grad_norm": 0.26329582929611206, "learning_rate": 8.5163317778794e-05, "loss": 1.4464704513549804, "step": 575 }, { "epoch": 0.27718040621266427, "grad_norm": 0.25123128294944763, "learning_rate": 8.488721346868482e-05, "loss": 1.4359591484069825, "step": 580 }, { "epoch": 0.27956989247311825, "grad_norm": 0.22458429634571075, "learning_rate": 8.460902028674204e-05, "loss": 1.4677635192871095, "step": 585 }, { "epoch": 0.2819593787335723, "grad_norm": 0.23584137856960297, "learning_rate": 8.432875488978189e-05, "loss": 1.4590428352355957, "step": 590 }, { "epoch": 0.2843488649940263, "grad_norm": 0.2357102632522583, "learning_rate": 8.404643405869441e-05, "loss": 1.4029648780822754, "step": 595 }, { "epoch": 0.2867383512544803, "grad_norm": 0.28613921999931335, "learning_rate": 8.376207469743877e-05, "loss": 1.4268519401550293, "step": 600 }, { "epoch": 0.2891278375149343, "grad_norm": 0.296101450920105, "learning_rate": 8.347569383203113e-05, "loss": 1.4516454696655274, "step": 605 }, { "epoch": 0.2915173237753883, "grad_norm": 0.2902820110321045, "learning_rate": 8.318730860952522e-05, "loss": 1.4078670501708985, "step": 610 }, { "epoch": 0.2939068100358423, "grad_norm": 0.2812166213989258, "learning_rate": 8.289693629698564e-05, "loss": 1.446097183227539, "step": 615 }, { "epoch": 0.2962962962962963, "grad_norm": 0.28198304772377014, "learning_rate": 8.260459428045397e-05, "loss": 1.4084587097167969, "step": 620 }, { "epoch": 0.2986857825567503, "grad_norm": 0.28590261936187744, "learning_rate": 8.231030006390786e-05, "loss": 1.438043212890625, "step": 625 }, { "epoch": 0.3010752688172043, "grad_norm": 0.2735881209373474, "learning_rate": 8.20140712682129e-05, "loss": 1.4516472816467285, "step": 630 }, { "epoch": 0.3034647550776583, "grad_norm": 0.23204857110977173, "learning_rate": 8.171592563006763e-05, "loss": 1.4332650184631348, "step": 635 }, { "epoch": 0.3058542413381123, "grad_norm": 0.2775753140449524, "learning_rate": 8.141588100094149e-05, "loss": 1.4567228317260743, "step": 640 }, { "epoch": 0.30824372759856633, "grad_norm": 0.2524227499961853, "learning_rate": 8.111395534600603e-05, "loss": 1.3886870384216308, "step": 645 }, { "epoch": 0.3106332138590203, "grad_norm": 0.2950308322906494, "learning_rate": 8.081016674305923e-05, "loss": 1.4061997413635254, "step": 650 }, { "epoch": 0.3130227001194743, "grad_norm": 0.22507043182849884, "learning_rate": 8.050453338144302e-05, "loss": 1.4376940727233887, "step": 655 }, { "epoch": 0.3154121863799283, "grad_norm": 0.24150978028774261, "learning_rate": 8.019707356095432e-05, "loss": 1.4593083381652832, "step": 660 }, { "epoch": 0.31780167264038234, "grad_norm": 0.24970196187496185, "learning_rate": 7.98878056907492e-05, "loss": 1.3895917892456056, "step": 665 }, { "epoch": 0.3201911589008363, "grad_norm": 0.22895695269107819, "learning_rate": 7.95767482882408e-05, "loss": 1.379880142211914, "step": 670 }, { "epoch": 0.3225806451612903, "grad_norm": 0.24954067170619965, "learning_rate": 7.92639199779904e-05, "loss": 1.4173902511596679, "step": 675 }, { "epoch": 0.3249701314217443, "grad_norm": 0.24520763754844666, "learning_rate": 7.894933949059245e-05, "loss": 1.4556282043457032, "step": 680 }, { "epoch": 0.32735961768219835, "grad_norm": 0.25326064229011536, "learning_rate": 7.863302566155295e-05, "loss": 1.4287520408630372, "step": 685 }, { "epoch": 0.32974910394265233, "grad_norm": 0.2682090997695923, "learning_rate": 7.831499743016176e-05, "loss": 1.4502517700195312, "step": 690 }, { "epoch": 0.3321385902031063, "grad_norm": 0.27123990654945374, "learning_rate": 7.799527383835858e-05, "loss": 1.400387668609619, "step": 695 }, { "epoch": 0.3345280764635603, "grad_norm": 0.3303516209125519, "learning_rate": 7.767387402959281e-05, "loss": 1.4342352867126464, "step": 700 }, { "epoch": 0.33691756272401435, "grad_norm": 0.27083998918533325, "learning_rate": 7.735081724767732e-05, "loss": 1.4210332870483398, "step": 705 }, { "epoch": 0.33930704898446834, "grad_norm": 0.3033396303653717, "learning_rate": 7.702612283563624e-05, "loss": 1.4115476608276367, "step": 710 }, { "epoch": 0.34169653524492233, "grad_norm": 0.2991386950016022, "learning_rate": 7.669981023454682e-05, "loss": 1.438405990600586, "step": 715 }, { "epoch": 0.34408602150537637, "grad_norm": 0.2793933153152466, "learning_rate": 7.637189898237536e-05, "loss": 1.4340227127075196, "step": 720 }, { "epoch": 0.34647550776583036, "grad_norm": 0.2670401930809021, "learning_rate": 7.604240871280743e-05, "loss": 1.4393259048461915, "step": 725 }, { "epoch": 0.34886499402628435, "grad_norm": 0.28978416323661804, "learning_rate": 7.571135915407219e-05, "loss": 1.4036520004272461, "step": 730 }, { "epoch": 0.35125448028673834, "grad_norm": 0.23741275072097778, "learning_rate": 7.537877012776132e-05, "loss": 1.360525417327881, "step": 735 }, { "epoch": 0.3536439665471924, "grad_norm": 0.29458534717559814, "learning_rate": 7.504466154764205e-05, "loss": 1.390071487426758, "step": 740 }, { "epoch": 0.35603345280764637, "grad_norm": 0.220574751496315, "learning_rate": 7.470905341846492e-05, "loss": 1.3987879753112793, "step": 745 }, { "epoch": 0.35842293906810035, "grad_norm": 0.23055969178676605, "learning_rate": 7.437196583476596e-05, "loss": 1.3894487380981446, "step": 750 }, { "epoch": 0.36081242532855434, "grad_norm": 0.2939671277999878, "learning_rate": 7.403341897966356e-05, "loss": 1.4039226531982423, "step": 755 }, { "epoch": 0.3632019115890084, "grad_norm": 0.24370169639587402, "learning_rate": 7.369343312364993e-05, "loss": 1.412961769104004, "step": 760 }, { "epoch": 0.3655913978494624, "grad_norm": 0.26102086901664734, "learning_rate": 7.335202862337754e-05, "loss": 1.402806854248047, "step": 765 }, { "epoch": 0.36798088410991636, "grad_norm": 0.4320904314517975, "learning_rate": 7.300922592044006e-05, "loss": 1.4030139923095704, "step": 770 }, { "epoch": 0.37037037037037035, "grad_norm": 0.21553704142570496, "learning_rate": 7.266504554014867e-05, "loss": 1.3706230163574218, "step": 775 }, { "epoch": 0.3727598566308244, "grad_norm": 0.23417530953884125, "learning_rate": 7.231950809030287e-05, "loss": 1.3858160018920898, "step": 780 }, { "epoch": 0.3751493428912784, "grad_norm": 0.2949879467487335, "learning_rate": 7.197263425995682e-05, "loss": 1.4113401412963866, "step": 785 }, { "epoch": 0.37753882915173237, "grad_norm": 0.23589320480823517, "learning_rate": 7.162444481818034e-05, "loss": 1.4031970977783204, "step": 790 }, { "epoch": 0.37992831541218636, "grad_norm": 0.29415807127952576, "learning_rate": 7.127496061281551e-05, "loss": 1.3802370071411132, "step": 795 }, { "epoch": 0.3823178016726404, "grad_norm": 0.24823541939258575, "learning_rate": 7.092420256922845e-05, "loss": 1.4349376678466796, "step": 800 }, { "epoch": 0.3847072879330944, "grad_norm": 0.2479880154132843, "learning_rate": 7.057219168905624e-05, "loss": 1.391560173034668, "step": 805 }, { "epoch": 0.3870967741935484, "grad_norm": 0.27021872997283936, "learning_rate": 7.021894904894963e-05, "loss": 1.4005847930908204, "step": 810 }, { "epoch": 0.38948626045400236, "grad_norm": 0.24428631365299225, "learning_rate": 6.986449579931092e-05, "loss": 1.4160762786865235, "step": 815 }, { "epoch": 0.3918757467144564, "grad_norm": 0.2466736137866974, "learning_rate": 6.950885316302773e-05, "loss": 1.3857328414916992, "step": 820 }, { "epoch": 0.3942652329749104, "grad_norm": 0.22851239144802094, "learning_rate": 6.915204243420215e-05, "loss": 1.3662797927856445, "step": 825 }, { "epoch": 0.3966547192353644, "grad_norm": 0.27387139201164246, "learning_rate": 6.879408497687582e-05, "loss": 1.3770858764648437, "step": 830 }, { "epoch": 0.39904420549581837, "grad_norm": 0.24753716588020325, "learning_rate": 6.843500222375074e-05, "loss": 1.3732675552368163, "step": 835 }, { "epoch": 0.4014336917562724, "grad_norm": 0.313991516828537, "learning_rate": 6.807481567490604e-05, "loss": 1.3654424667358398, "step": 840 }, { "epoch": 0.4038231780167264, "grad_norm": 0.2923700511455536, "learning_rate": 6.771354689651053e-05, "loss": 1.3375659942626954, "step": 845 }, { "epoch": 0.4062126642771804, "grad_norm": 0.2623784840106964, "learning_rate": 6.73512175195316e-05, "loss": 1.4283991813659669, "step": 850 }, { "epoch": 0.40860215053763443, "grad_norm": 0.25396597385406494, "learning_rate": 6.698784923843992e-05, "loss": 1.4071483612060547, "step": 855 }, { "epoch": 0.4109916367980884, "grad_norm": 0.2575069069862366, "learning_rate": 6.662346380991058e-05, "loss": 1.410196590423584, "step": 860 }, { "epoch": 0.4133811230585424, "grad_norm": 0.26217254996299744, "learning_rate": 6.625808305152033e-05, "loss": 1.376619529724121, "step": 865 }, { "epoch": 0.4157706093189964, "grad_norm": 0.31529921293258667, "learning_rate": 6.589172884044135e-05, "loss": 1.4235527038574218, "step": 870 }, { "epoch": 0.41816009557945044, "grad_norm": 0.26250818371772766, "learning_rate": 6.552442311213122e-05, "loss": 1.3581882476806642, "step": 875 }, { "epoch": 0.42054958183990443, "grad_norm": 0.2841402292251587, "learning_rate": 6.515618785901967e-05, "loss": 1.3833487510681153, "step": 880 }, { "epoch": 0.4229390681003584, "grad_norm": 0.30380716919898987, "learning_rate": 6.478704512919173e-05, "loss": 1.4346580505371094, "step": 885 }, { "epoch": 0.4253285543608124, "grad_norm": 0.2277163565158844, "learning_rate": 6.441701702506754e-05, "loss": 1.367258644104004, "step": 890 }, { "epoch": 0.42771804062126645, "grad_norm": 0.28610748052597046, "learning_rate": 6.40461257020791e-05, "loss": 1.3959890365600587, "step": 895 }, { "epoch": 0.43010752688172044, "grad_norm": 0.32751911878585815, "learning_rate": 6.36743933673436e-05, "loss": 1.3987584114074707, "step": 900 }, { "epoch": 0.4324970131421744, "grad_norm": 0.24882109463214874, "learning_rate": 6.330184227833376e-05, "loss": 1.3455117225646973, "step": 905 }, { "epoch": 0.4348864994026284, "grad_norm": 0.2577092945575714, "learning_rate": 6.29284947415453e-05, "loss": 1.3549080848693849, "step": 910 }, { "epoch": 0.43727598566308246, "grad_norm": 0.26310569047927856, "learning_rate": 6.255437311116119e-05, "loss": 1.4157590866088867, "step": 915 }, { "epoch": 0.43966547192353644, "grad_norm": 0.30136528611183167, "learning_rate": 6.217949978771328e-05, "loss": 1.411697769165039, "step": 920 }, { "epoch": 0.44205495818399043, "grad_norm": 0.2392907589673996, "learning_rate": 6.1803897216741e-05, "loss": 1.3838436126708984, "step": 925 }, { "epoch": 0.4444444444444444, "grad_norm": 0.24898621439933777, "learning_rate": 6.142758788744754e-05, "loss": 1.3894274711608887, "step": 930 }, { "epoch": 0.44683393070489846, "grad_norm": 0.24590671062469482, "learning_rate": 6.105059433135317e-05, "loss": 1.3372700691223145, "step": 935 }, { "epoch": 0.44922341696535245, "grad_norm": 0.2970089614391327, "learning_rate": 6.06729391209463e-05, "loss": 1.394719409942627, "step": 940 }, { "epoch": 0.45161290322580644, "grad_norm": 0.2693709433078766, "learning_rate": 6.029464486833186e-05, "loss": 1.393956756591797, "step": 945 }, { "epoch": 0.4540023894862604, "grad_norm": 0.2572726607322693, "learning_rate": 5.991573422387742e-05, "loss": 1.3696969985961913, "step": 950 }, { "epoch": 0.45639187574671447, "grad_norm": 0.24961479008197784, "learning_rate": 5.953622987485703e-05, "loss": 1.4150132179260253, "step": 955 }, { "epoch": 0.45878136200716846, "grad_norm": 0.2659885585308075, "learning_rate": 5.9156154544092815e-05, "loss": 1.365439224243164, "step": 960 }, { "epoch": 0.46117084826762245, "grad_norm": 0.3113667070865631, "learning_rate": 5.877553098859439e-05, "loss": 1.359626865386963, "step": 965 }, { "epoch": 0.4635603345280765, "grad_norm": 0.29236137866973877, "learning_rate": 5.8394381998196344e-05, "loss": 1.4090446472167968, "step": 970 }, { "epoch": 0.4659498207885305, "grad_norm": 0.2842124402523041, "learning_rate": 5.801273039419368e-05, "loss": 1.3704862594604492, "step": 975 }, { "epoch": 0.46833930704898447, "grad_norm": 0.30265432596206665, "learning_rate": 5.763059902797542e-05, "loss": 1.3788330078125, "step": 980 }, { "epoch": 0.47072879330943845, "grad_norm": 0.24866238236427307, "learning_rate": 5.724801077965629e-05, "loss": 1.375232982635498, "step": 985 }, { "epoch": 0.4731182795698925, "grad_norm": 0.2759285271167755, "learning_rate": 5.686498855670687e-05, "loss": 1.3548823356628419, "step": 990 }, { "epoch": 0.4755077658303465, "grad_norm": 0.2486870288848877, "learning_rate": 5.6481555292581946e-05, "loss": 1.3632122039794923, "step": 995 }, { "epoch": 0.4778972520908005, "grad_norm": 0.26392608880996704, "learning_rate": 5.609773394534741e-05, "loss": 1.350174331665039, "step": 1000 }, { "epoch": 0.48028673835125446, "grad_norm": 0.2669070363044739, "learning_rate": 5.571354749630564e-05, "loss": 1.370138168334961, "step": 1005 }, { "epoch": 0.4826762246117085, "grad_norm": 0.2636848986148834, "learning_rate": 5.532901894861946e-05, "loss": 1.3845187187194825, "step": 1010 }, { "epoch": 0.4850657108721625, "grad_norm": 0.30728474259376526, "learning_rate": 5.494417132593488e-05, "loss": 1.3009090423583984, "step": 1015 }, { "epoch": 0.4874551971326165, "grad_norm": 0.26170051097869873, "learning_rate": 5.455902767100246e-05, "loss": 1.3479591369628907, "step": 1020 }, { "epoch": 0.48984468339307047, "grad_norm": 0.2369968444108963, "learning_rate": 5.4173611044297766e-05, "loss": 1.3598762512207032, "step": 1025 }, { "epoch": 0.4922341696535245, "grad_norm": 0.23840807378292084, "learning_rate": 5.378794452264053e-05, "loss": 1.3730680465698242, "step": 1030 }, { "epoch": 0.4946236559139785, "grad_norm": 0.2579615116119385, "learning_rate": 5.3402051197812884e-05, "loss": 1.3374858856201173, "step": 1035 }, { "epoch": 0.4970131421744325, "grad_norm": 0.29306015372276306, "learning_rate": 5.30159541751769e-05, "loss": 1.3605210304260253, "step": 1040 }, { "epoch": 0.4994026284348865, "grad_norm": 0.3821243941783905, "learning_rate": 5.2629676572290956e-05, "loss": 1.2784107208251954, "step": 1045 }, { "epoch": 0.5017921146953405, "grad_norm": 0.2783209979534149, "learning_rate": 5.2243241517525754e-05, "loss": 1.3636322975158692, "step": 1050 }, { "epoch": 0.5041816009557945, "grad_norm": 0.30755525827407837, "learning_rate": 5.185667214867937e-05, "loss": 1.3342845916748047, "step": 1055 }, { "epoch": 0.5065710872162486, "grad_norm": 0.24514539539813995, "learning_rate": 5.146999161159195e-05, "loss": 1.3436594009399414, "step": 1060 }, { "epoch": 0.5089605734767025, "grad_norm": 0.24278080463409424, "learning_rate": 5.108322305875988e-05, "loss": 1.407780075073242, "step": 1065 }, { "epoch": 0.5113500597371565, "grad_norm": 0.24821552634239197, "learning_rate": 5.0696389647949415e-05, "loss": 1.399796485900879, "step": 1070 }, { "epoch": 0.5137395459976105, "grad_norm": 0.2761519253253937, "learning_rate": 5.030951454081023e-05, "loss": 1.3828035354614259, "step": 1075 }, { "epoch": 0.5161290322580645, "grad_norm": 0.2488783895969391, "learning_rate": 4.992262090148856e-05, "loss": 1.3616230964660645, "step": 1080 }, { "epoch": 0.5185185185185185, "grad_norm": 0.29967793822288513, "learning_rate": 4.9535731895240254e-05, "loss": 1.3879215240478515, "step": 1085 }, { "epoch": 0.5209080047789725, "grad_norm": 0.24454021453857422, "learning_rate": 4.914887068704375e-05, "loss": 1.3446988105773925, "step": 1090 }, { "epoch": 0.5232974910394266, "grad_norm": 0.2721116542816162, "learning_rate": 4.8762060440213095e-05, "loss": 1.326106357574463, "step": 1095 }, { "epoch": 0.5256869772998806, "grad_norm": 0.2826915383338928, "learning_rate": 4.837532431501098e-05, "loss": 1.3933807373046876, "step": 1100 }, { "epoch": 0.5280764635603346, "grad_norm": 0.2768832743167877, "learning_rate": 4.798868546726212e-05, "loss": 1.3409512519836426, "step": 1105 }, { "epoch": 0.5304659498207885, "grad_norm": 0.2674025893211365, "learning_rate": 4.76021670469667e-05, "loss": 1.349265193939209, "step": 1110 }, { "epoch": 0.5328554360812425, "grad_norm": 0.2236485630273819, "learning_rate": 4.7215792196914345e-05, "loss": 1.3768779754638671, "step": 1115 }, { "epoch": 0.5352449223416965, "grad_norm": 0.30000221729278564, "learning_rate": 4.682958405129837e-05, "loss": 1.3748072624206542, "step": 1120 }, { "epoch": 0.5376344086021505, "grad_norm": 0.250331848859787, "learning_rate": 4.6443565734330715e-05, "loss": 1.333662223815918, "step": 1125 }, { "epoch": 0.5400238948626045, "grad_norm": 0.2686733603477478, "learning_rate": 4.6057760358857274e-05, "loss": 1.3277825355529784, "step": 1130 }, { "epoch": 0.5424133811230586, "grad_norm": 0.29305464029312134, "learning_rate": 4.567219102497412e-05, "loss": 1.3602632522583007, "step": 1135 }, { "epoch": 0.5448028673835126, "grad_norm": 0.30608347058296204, "learning_rate": 4.528688081864433e-05, "loss": 1.3432723999023437, "step": 1140 }, { "epoch": 0.5471923536439666, "grad_norm": 0.27126264572143555, "learning_rate": 4.4901852810315634e-05, "loss": 1.3473407745361328, "step": 1145 }, { "epoch": 0.5495818399044206, "grad_norm": 0.2633585035800934, "learning_rate": 4.4517130053539316e-05, "loss": 1.363102340698242, "step": 1150 }, { "epoch": 0.5519713261648745, "grad_norm": 0.3300692141056061, "learning_rate": 4.413273558358957e-05, "loss": 1.3794038772583008, "step": 1155 }, { "epoch": 0.5543608124253285, "grad_norm": 0.29022032022476196, "learning_rate": 4.3748692416084506e-05, "loss": 1.3189671516418457, "step": 1160 }, { "epoch": 0.5567502986857825, "grad_norm": 0.2611541748046875, "learning_rate": 4.336502354560797e-05, "loss": 1.342616081237793, "step": 1165 }, { "epoch": 0.5591397849462365, "grad_norm": 0.3072931170463562, "learning_rate": 4.298175194433279e-05, "loss": 1.3125935554504395, "step": 1170 }, { "epoch": 0.5615292712066906, "grad_norm": 0.29690274596214294, "learning_rate": 4.259890056064527e-05, "loss": 1.2946578979492187, "step": 1175 }, { "epoch": 0.5639187574671446, "grad_norm": 0.29266008734703064, "learning_rate": 4.221649231777126e-05, "loss": 1.3257877349853515, "step": 1180 }, { "epoch": 0.5663082437275986, "grad_norm": 0.286796897649765, "learning_rate": 4.183455011240342e-05, "loss": 1.3117992401123046, "step": 1185 }, { "epoch": 0.5686977299880526, "grad_norm": 0.2607665956020355, "learning_rate": 4.1453096813330584e-05, "loss": 1.2944452285766601, "step": 1190 }, { "epoch": 0.5710872162485066, "grad_norm": 0.27462026476860046, "learning_rate": 4.107215526006817e-05, "loss": 1.3200563430786132, "step": 1195 }, { "epoch": 0.5734767025089605, "grad_norm": 0.30187466740608215, "learning_rate": 4.0691748261490936e-05, "loss": 1.354791259765625, "step": 1200 }, { "epoch": 0.5758661887694145, "grad_norm": 0.2645605206489563, "learning_rate": 4.0311898594467086e-05, "loss": 1.331553077697754, "step": 1205 }, { "epoch": 0.5782556750298686, "grad_norm": 0.26903611421585083, "learning_rate": 3.993262900249467e-05, "loss": 1.3656150817871093, "step": 1210 }, { "epoch": 0.5806451612903226, "grad_norm": 0.325409471988678, "learning_rate": 3.9553962194339686e-05, "loss": 1.3304656982421874, "step": 1215 }, { "epoch": 0.5830346475507766, "grad_norm": 0.25730133056640625, "learning_rate": 3.9175920842676535e-05, "loss": 1.325532817840576, "step": 1220 }, { "epoch": 0.5854241338112306, "grad_norm": 0.2545807957649231, "learning_rate": 3.8798527582730294e-05, "loss": 1.3666142463684081, "step": 1225 }, { "epoch": 0.5878136200716846, "grad_norm": 0.24386471509933472, "learning_rate": 3.842180501092163e-05, "loss": 1.3361050605773925, "step": 1230 }, { "epoch": 0.5902031063321386, "grad_norm": 0.28684672713279724, "learning_rate": 3.8045775683513785e-05, "loss": 1.2830507278442382, "step": 1235 }, { "epoch": 0.5925925925925926, "grad_norm": 0.30884966254234314, "learning_rate": 3.7670462115261906e-05, "loss": 1.3353318214416503, "step": 1240 }, { "epoch": 0.5949820788530465, "grad_norm": 0.2460509091615677, "learning_rate": 3.729588677806513e-05, "loss": 1.3347097396850587, "step": 1245 }, { "epoch": 0.5973715651135006, "grad_norm": 0.26541683077812195, "learning_rate": 3.6922072099620975e-05, "loss": 1.31211576461792, "step": 1250 }, { "epoch": 0.5997610513739546, "grad_norm": 0.25815343856811523, "learning_rate": 3.6549040462082556e-05, "loss": 1.3165303230285645, "step": 1255 }, { "epoch": 0.6021505376344086, "grad_norm": 0.25208917260169983, "learning_rate": 3.617681420071836e-05, "loss": 1.3751947402954101, "step": 1260 }, { "epoch": 0.6045400238948626, "grad_norm": 0.26578542590141296, "learning_rate": 3.580541560257505e-05, "loss": 1.3281235694885254, "step": 1265 }, { "epoch": 0.6069295101553166, "grad_norm": 0.24992331862449646, "learning_rate": 3.543486690514286e-05, "loss": 1.349024772644043, "step": 1270 }, { "epoch": 0.6093189964157706, "grad_norm": 0.2626526951789856, "learning_rate": 3.506519029502433e-05, "loss": 1.3072402954101563, "step": 1275 }, { "epoch": 0.6117084826762246, "grad_norm": 0.30928000807762146, "learning_rate": 3.469640790660568e-05, "loss": 1.297838020324707, "step": 1280 }, { "epoch": 0.6140979689366786, "grad_norm": 0.2769022285938263, "learning_rate": 3.432854182073166e-05, "loss": 1.3581188201904297, "step": 1285 }, { "epoch": 0.6164874551971327, "grad_norm": 0.2729344964027405, "learning_rate": 3.3961614063383396e-05, "loss": 1.3024734497070312, "step": 1290 }, { "epoch": 0.6188769414575866, "grad_norm": 0.26106736063957214, "learning_rate": 3.3595646604359587e-05, "loss": 1.3699445724487305, "step": 1295 }, { "epoch": 0.6212664277180406, "grad_norm": 0.3085561692714691, "learning_rate": 3.3230661355961076e-05, "loss": 1.3581072807312011, "step": 1300 }, { "epoch": 0.6236559139784946, "grad_norm": 0.2854253351688385, "learning_rate": 3.2866680171678874e-05, "loss": 1.3420246124267579, "step": 1305 }, { "epoch": 0.6260454002389486, "grad_norm": 0.30645105242729187, "learning_rate": 3.250372484488558e-05, "loss": 1.3140718460083007, "step": 1310 }, { "epoch": 0.6284348864994026, "grad_norm": 0.2742772400379181, "learning_rate": 3.214181710753069e-05, "loss": 1.3964643478393555, "step": 1315 }, { "epoch": 0.6308243727598566, "grad_norm": 0.2850540578365326, "learning_rate": 3.178097862883916e-05, "loss": 1.3153679847717286, "step": 1320 }, { "epoch": 0.6332138590203107, "grad_norm": 0.25389060378074646, "learning_rate": 3.1421231014014175e-05, "loss": 1.3177446365356444, "step": 1325 }, { "epoch": 0.6356033452807647, "grad_norm": 0.2736924886703491, "learning_rate": 3.106259580294344e-05, "loss": 1.3437786102294922, "step": 1330 }, { "epoch": 0.6379928315412187, "grad_norm": 0.3225949704647064, "learning_rate": 3.070509446890944e-05, "loss": 1.3047246932983398, "step": 1335 }, { "epoch": 0.6403823178016727, "grad_norm": 0.266668438911438, "learning_rate": 3.0348748417303823e-05, "loss": 1.3629995346069337, "step": 1340 }, { "epoch": 0.6427718040621266, "grad_norm": 0.27443379163742065, "learning_rate": 2.9993578984345672e-05, "loss": 1.377586555480957, "step": 1345 }, { "epoch": 0.6451612903225806, "grad_norm": 0.25409719347953796, "learning_rate": 2.963960743580405e-05, "loss": 1.3082121849060058, "step": 1350 }, { "epoch": 0.6475507765830346, "grad_norm": 0.25010809302330017, "learning_rate": 2.9286854965724686e-05, "loss": 1.2946581840515137, "step": 1355 }, { "epoch": 0.6499402628434886, "grad_norm": 0.2803384065628052, "learning_rate": 2.8935342695160996e-05, "loss": 1.2734054565429687, "step": 1360 }, { "epoch": 0.6523297491039427, "grad_norm": 0.28773632645606995, "learning_rate": 2.8585091670909438e-05, "loss": 1.3738662719726562, "step": 1365 }, { "epoch": 0.6547192353643967, "grad_norm": 0.3194836676120758, "learning_rate": 2.823612286424935e-05, "loss": 1.3134496688842774, "step": 1370 }, { "epoch": 0.6571087216248507, "grad_norm": 0.3230612576007843, "learning_rate": 2.7888457169687298e-05, "loss": 1.3343011856079101, "step": 1375 }, { "epoch": 0.6594982078853047, "grad_norm": 0.2869691848754883, "learning_rate": 2.7542115403706063e-05, "loss": 1.303788185119629, "step": 1380 }, { "epoch": 0.6618876941457587, "grad_norm": 0.26587560772895813, "learning_rate": 2.7197118303518092e-05, "loss": 1.3306425094604493, "step": 1385 }, { "epoch": 0.6642771804062126, "grad_norm": 0.296779066324234, "learning_rate": 2.685348652582409e-05, "loss": 1.2880214691162108, "step": 1390 }, { "epoch": 0.6666666666666666, "grad_norm": 0.2634750008583069, "learning_rate": 2.6511240645576017e-05, "loss": 1.396371078491211, "step": 1395 }, { "epoch": 0.6690561529271206, "grad_norm": 0.311478853225708, "learning_rate": 2.6170401154745227e-05, "loss": 1.312047290802002, "step": 1400 }, { "epoch": 0.6714456391875747, "grad_norm": 0.2735639214515686, "learning_rate": 2.5830988461095508e-05, "loss": 1.3772590637207032, "step": 1405 }, { "epoch": 0.6738351254480287, "grad_norm": 0.28115421533584595, "learning_rate": 2.5493022886961225e-05, "loss": 1.3423656463623046, "step": 1410 }, { "epoch": 0.6762246117084827, "grad_norm": 0.22361046075820923, "learning_rate": 2.51565246680304e-05, "loss": 1.3142578125, "step": 1415 }, { "epoch": 0.6786140979689367, "grad_norm": 0.2917373776435852, "learning_rate": 2.482151395213317e-05, "loss": 1.3070114135742188, "step": 1420 }, { "epoch": 0.6810035842293907, "grad_norm": 0.34901735186576843, "learning_rate": 2.4488010798035504e-05, "loss": 1.3786397933959962, "step": 1425 }, { "epoch": 0.6833930704898447, "grad_norm": 0.30047670006752014, "learning_rate": 2.4156035174237984e-05, "loss": 1.3069480895996093, "step": 1430 }, { "epoch": 0.6857825567502986, "grad_norm": 0.23732812702655792, "learning_rate": 2.3825606957780456e-05, "loss": 1.3747161865234374, "step": 1435 }, { "epoch": 0.6881720430107527, "grad_norm": 0.2703532576560974, "learning_rate": 2.3496745933051655e-05, "loss": 1.3846330642700195, "step": 1440 }, { "epoch": 0.6905615292712067, "grad_norm": 0.27444109320640564, "learning_rate": 2.3169471790604767e-05, "loss": 1.3017520904541016, "step": 1445 }, { "epoch": 0.6929510155316607, "grad_norm": 0.30454879999160767, "learning_rate": 2.2843804125978357e-05, "loss": 1.362583541870117, "step": 1450 }, { "epoch": 0.6953405017921147, "grad_norm": 0.28890475630760193, "learning_rate": 2.2519762438523207e-05, "loss": 1.294306468963623, "step": 1455 }, { "epoch": 0.6977299880525687, "grad_norm": 0.25940513610839844, "learning_rate": 2.2197366130234655e-05, "loss": 1.2947320938110352, "step": 1460 }, { "epoch": 0.7001194743130227, "grad_norm": 0.30465951561927795, "learning_rate": 2.1876634504590987e-05, "loss": 1.3730806350708007, "step": 1465 }, { "epoch": 0.7025089605734767, "grad_norm": 0.30388158559799194, "learning_rate": 2.1557586765397596e-05, "loss": 1.3263800621032715, "step": 1470 }, { "epoch": 0.7048984468339307, "grad_norm": 0.26089686155319214, "learning_rate": 2.1240242015637267e-05, "loss": 1.3501880645751954, "step": 1475 }, { "epoch": 0.7072879330943848, "grad_norm": 0.27901092171669006, "learning_rate": 2.092461925632615e-05, "loss": 1.3349273681640625, "step": 1480 }, { "epoch": 0.7096774193548387, "grad_norm": 0.2533346116542816, "learning_rate": 2.061073738537635e-05, "loss": 1.348337745666504, "step": 1485 }, { "epoch": 0.7120669056152927, "grad_norm": 0.24928832054138184, "learning_rate": 2.0298615196464217e-05, "loss": 1.3316417694091798, "step": 1490 }, { "epoch": 0.7144563918757467, "grad_norm": 0.2750134766101837, "learning_rate": 1.9988271377905166e-05, "loss": 1.3168065071105957, "step": 1495 }, { "epoch": 0.7168458781362007, "grad_norm": 0.3212110996246338, "learning_rate": 1.967972451153466e-05, "loss": 1.3592927932739258, "step": 1500 }, { "epoch": 0.7192353643966547, "grad_norm": 0.2544815242290497, "learning_rate": 1.9372993071595724e-05, "loss": 1.330977439880371, "step": 1505 }, { "epoch": 0.7216248506571087, "grad_norm": 0.30273714661598206, "learning_rate": 1.9068095423632654e-05, "loss": 1.3181983947753906, "step": 1510 }, { "epoch": 0.7240143369175627, "grad_norm": 0.31694936752319336, "learning_rate": 1.8765049823391473e-05, "loss": 1.3054021835327148, "step": 1515 }, { "epoch": 0.7264038231780168, "grad_norm": 0.270060271024704, "learning_rate": 1.8463874415726918e-05, "loss": 1.3121872901916505, "step": 1520 }, { "epoch": 0.7287933094384708, "grad_norm": 0.2667926251888275, "learning_rate": 1.8164587233515823e-05, "loss": 1.329949188232422, "step": 1525 }, { "epoch": 0.7311827956989247, "grad_norm": 0.32022303342819214, "learning_rate": 1.7867206196577645e-05, "loss": 1.297494125366211, "step": 1530 }, { "epoch": 0.7335722819593787, "grad_norm": 0.29288119077682495, "learning_rate": 1.7571749110601338e-05, "loss": 1.3385650634765625, "step": 1535 }, { "epoch": 0.7359617682198327, "grad_norm": 0.32373276352882385, "learning_rate": 1.7278233666079307e-05, "loss": 1.33201904296875, "step": 1540 }, { "epoch": 0.7383512544802867, "grad_norm": 0.2649761438369751, "learning_rate": 1.6986677437248157e-05, "loss": 1.260688018798828, "step": 1545 }, { "epoch": 0.7407407407407407, "grad_norm": 0.287933349609375, "learning_rate": 1.6697097881036538e-05, "loss": 1.2957708358764648, "step": 1550 }, { "epoch": 0.7431302270011948, "grad_norm": 0.28423890471458435, "learning_rate": 1.64095123360197e-05, "loss": 1.376328468322754, "step": 1555 }, { "epoch": 0.7455197132616488, "grad_norm": 0.2617063522338867, "learning_rate": 1.6123938021381606e-05, "loss": 1.3709890365600585, "step": 1560 }, { "epoch": 0.7479091995221028, "grad_norm": 0.24597397446632385, "learning_rate": 1.5840392035883727e-05, "loss": 1.3249582290649413, "step": 1565 }, { "epoch": 0.7502986857825568, "grad_norm": 0.25838223099708557, "learning_rate": 1.555889135684136e-05, "loss": 1.3781761169433593, "step": 1570 }, { "epoch": 0.7526881720430108, "grad_norm": 0.2777562141418457, "learning_rate": 1.527945283910705e-05, "loss": 1.3022884368896483, "step": 1575 }, { "epoch": 0.7550776583034647, "grad_norm": 0.28209856152534485, "learning_rate": 1.5002093214061501e-05, "loss": 1.2888921737670898, "step": 1580 }, { "epoch": 0.7574671445639187, "grad_norm": 0.29105687141418457, "learning_rate": 1.4726829088611666e-05, "loss": 1.3178224563598633, "step": 1585 }, { "epoch": 0.7598566308243727, "grad_norm": 0.25559762120246887, "learning_rate": 1.4453676944196476e-05, "loss": 1.307926082611084, "step": 1590 }, { "epoch": 0.7622461170848268, "grad_norm": 0.3083442747592926, "learning_rate": 1.4182653135799995e-05, "loss": 1.3184553146362306, "step": 1595 }, { "epoch": 0.7646356033452808, "grad_norm": 0.2743801176548004, "learning_rate": 1.391377389097222e-05, "loss": 1.3171270370483399, "step": 1600 }, { "epoch": 0.7670250896057348, "grad_norm": 0.26681846380233765, "learning_rate": 1.3647055308857354e-05, "loss": 1.305533504486084, "step": 1605 }, { "epoch": 0.7694145758661888, "grad_norm": 0.28918692469596863, "learning_rate": 1.338251335922993e-05, "loss": 1.2909788131713866, "step": 1610 }, { "epoch": 0.7718040621266428, "grad_norm": 0.2921421229839325, "learning_rate": 1.3120163881538678e-05, "loss": 1.3338876724243165, "step": 1615 }, { "epoch": 0.7741935483870968, "grad_norm": 0.25906866788864136, "learning_rate": 1.2860022583957954e-05, "loss": 1.2904350280761718, "step": 1620 }, { "epoch": 0.7765830346475507, "grad_norm": 0.27177903056144714, "learning_rate": 1.260210504244747e-05, "loss": 1.2444252014160155, "step": 1625 }, { "epoch": 0.7789725209080047, "grad_norm": 0.3205372393131256, "learning_rate": 1.2346426699819458e-05, "loss": 1.3526622772216796, "step": 1630 }, { "epoch": 0.7813620071684588, "grad_norm": 0.27783483266830444, "learning_rate": 1.209300286481415e-05, "loss": 1.358124351501465, "step": 1635 }, { "epoch": 0.7837514934289128, "grad_norm": 0.2574422359466553, "learning_rate": 1.1841848711183129e-05, "loss": 1.310964298248291, "step": 1640 }, { "epoch": 0.7861409796893668, "grad_norm": 0.2714278995990753, "learning_rate": 1.1592979276780858e-05, "loss": 1.2791467666625977, "step": 1645 }, { "epoch": 0.7885304659498208, "grad_norm": 0.2741813659667969, "learning_rate": 1.1346409462664148e-05, "loss": 1.3601436614990234, "step": 1650 }, { "epoch": 0.7909199522102748, "grad_norm": 0.2800416052341461, "learning_rate": 1.1102154032200147e-05, "loss": 1.3265079498291015, "step": 1655 }, { "epoch": 0.7933094384707288, "grad_norm": 0.2758980095386505, "learning_rate": 1.0860227610182222e-05, "loss": 1.3546831130981445, "step": 1660 }, { "epoch": 0.7956989247311828, "grad_norm": 0.277106374502182, "learning_rate": 1.062064468195439e-05, "loss": 1.2565521240234374, "step": 1665 }, { "epoch": 0.7980884109916367, "grad_norm": 0.2647838592529297, "learning_rate": 1.038341959254397e-05, "loss": 1.2788930892944337, "step": 1670 }, { "epoch": 0.8004778972520908, "grad_norm": 0.29723435640335083, "learning_rate": 1.0148566545802719e-05, "loss": 1.2542386054992676, "step": 1675 }, { "epoch": 0.8028673835125448, "grad_norm": 0.27335965633392334, "learning_rate": 9.916099603556305e-06, "loss": 1.2913834571838378, "step": 1680 }, { "epoch": 0.8052568697729988, "grad_norm": 0.30953383445739746, "learning_rate": 9.686032684762408e-06, "loss": 1.3102952003479005, "step": 1685 }, { "epoch": 0.8076463560334528, "grad_norm": 0.3473983705043793, "learning_rate": 9.458379564677316e-06, "loss": 1.3081298828125, "step": 1690 }, { "epoch": 0.8100358422939068, "grad_norm": 0.27408310770988464, "learning_rate": 9.233153874031102e-06, "loss": 1.3258519172668457, "step": 1695 }, { "epoch": 0.8124253285543608, "grad_norm": 0.369855672121048, "learning_rate": 9.010369098211547e-06, "loss": 1.3175015449523926, "step": 1700 }, { "epoch": 0.8148148148148148, "grad_norm": 0.2810940146446228, "learning_rate": 8.790038576456627e-06, "loss": 1.3398883819580079, "step": 1705 }, { "epoch": 0.8172043010752689, "grad_norm": 0.3095511794090271, "learning_rate": 8.572175501055874e-06, "loss": 1.3127903938293457, "step": 1710 }, { "epoch": 0.8195937873357229, "grad_norm": 0.33327189087867737, "learning_rate": 8.356792916560458e-06, "loss": 1.3538610458374023, "step": 1715 }, { "epoch": 0.8219832735961768, "grad_norm": 0.2976340055465698, "learning_rate": 8.1439037190022e-06, "loss": 1.2965837478637696, "step": 1720 }, { "epoch": 0.8243727598566308, "grad_norm": 0.26589319109916687, "learning_rate": 7.933520655121352e-06, "loss": 1.3339267730712892, "step": 1725 }, { "epoch": 0.8267622461170848, "grad_norm": 0.2981930077075958, "learning_rate": 7.725656321603413e-06, "loss": 1.2726448059082032, "step": 1730 }, { "epoch": 0.8291517323775388, "grad_norm": 0.2475123107433319, "learning_rate": 7.520323164324921e-06, "loss": 1.2942044258117675, "step": 1735 }, { "epoch": 0.8315412186379928, "grad_norm": 0.33591943979263306, "learning_rate": 7.317533477608263e-06, "loss": 1.2391548156738281, "step": 1740 }, { "epoch": 0.8339307048984468, "grad_norm": 0.3285451829433441, "learning_rate": 7.117299403485467e-06, "loss": 1.3259029388427734, "step": 1745 }, { "epoch": 0.8363201911589009, "grad_norm": 0.32050254940986633, "learning_rate": 6.919632930971331e-06, "loss": 1.3052947998046875, "step": 1750 }, { "epoch": 0.8387096774193549, "grad_norm": 0.2564848065376282, "learning_rate": 6.72454589534548e-06, "loss": 1.3109350204467773, "step": 1755 }, { "epoch": 0.8410991636798089, "grad_norm": 0.25423112511634827, "learning_rate": 6.532049977443766e-06, "loss": 1.291252040863037, "step": 1760 }, { "epoch": 0.8434886499402628, "grad_norm": 0.4134671688079834, "learning_rate": 6.342156702958851e-06, "loss": 1.2587574005126954, "step": 1765 }, { "epoch": 0.8458781362007168, "grad_norm": 0.2682115435600281, "learning_rate": 6.154877441750184e-06, "loss": 1.296860122680664, "step": 1770 }, { "epoch": 0.8482676224611708, "grad_norm": 0.2778260409832001, "learning_rate": 5.9702234071631e-06, "loss": 1.2768482208251952, "step": 1775 }, { "epoch": 0.8506571087216248, "grad_norm": 0.2967587411403656, "learning_rate": 5.7882056553575705e-06, "loss": 1.2709309577941894, "step": 1780 }, { "epoch": 0.8530465949820788, "grad_norm": 0.27777722477912903, "learning_rate": 5.6088350846460805e-06, "loss": 1.286030101776123, "step": 1785 }, { "epoch": 0.8554360812425329, "grad_norm": 0.27154114842414856, "learning_rate": 5.432122434841175e-06, "loss": 1.2802083969116211, "step": 1790 }, { "epoch": 0.8578255675029869, "grad_norm": 0.260510116815567, "learning_rate": 5.2580782866124056e-06, "loss": 1.3005708694458007, "step": 1795 }, { "epoch": 0.8602150537634409, "grad_norm": 0.29478469491004944, "learning_rate": 5.086713060852788e-06, "loss": 1.2962477684020997, "step": 1800 }, { "epoch": 0.8626045400238949, "grad_norm": 0.277447909116745, "learning_rate": 4.918037018054844e-06, "loss": 1.2823700904846191, "step": 1805 }, { "epoch": 0.8649940262843488, "grad_norm": 0.28672778606414795, "learning_rate": 4.752060257696278e-06, "loss": 1.288120937347412, "step": 1810 }, { "epoch": 0.8673835125448028, "grad_norm": 0.2634541094303131, "learning_rate": 4.588792717635287e-06, "loss": 1.2693995475769042, "step": 1815 }, { "epoch": 0.8697729988052568, "grad_norm": 0.3322059214115143, "learning_rate": 4.428244173515483e-06, "loss": 1.336508560180664, "step": 1820 }, { "epoch": 0.8721624850657109, "grad_norm": 0.2569097578525543, "learning_rate": 4.270424238180614e-06, "loss": 1.3129013061523438, "step": 1825 }, { "epoch": 0.8745519713261649, "grad_norm": 0.2902193069458008, "learning_rate": 4.1153423610989725e-06, "loss": 1.2740081787109374, "step": 1830 }, { "epoch": 0.8769414575866189, "grad_norm": 0.24282416701316833, "learning_rate": 3.963007827797627e-06, "loss": 1.2854101181030273, "step": 1835 }, { "epoch": 0.8793309438470729, "grad_norm": 0.28557807207107544, "learning_rate": 3.813429759306436e-06, "loss": 1.264516258239746, "step": 1840 }, { "epoch": 0.8817204301075269, "grad_norm": 0.25589361786842346, "learning_rate": 3.6666171116119475e-06, "loss": 1.303445053100586, "step": 1845 }, { "epoch": 0.8841099163679809, "grad_norm": 0.2900035083293915, "learning_rate": 3.5225786751211297e-06, "loss": 1.365576934814453, "step": 1850 }, { "epoch": 0.8864994026284349, "grad_norm": 0.2687821090221405, "learning_rate": 3.381323074135073e-06, "loss": 1.2323289871215821, "step": 1855 }, { "epoch": 0.8888888888888888, "grad_norm": 0.2550598382949829, "learning_rate": 3.242858766332574e-06, "loss": 1.3164402008056642, "step": 1860 }, { "epoch": 0.8912783751493429, "grad_norm": 0.2647612392902374, "learning_rate": 3.1071940422638057e-06, "loss": 1.3278432846069337, "step": 1865 }, { "epoch": 0.8936678614097969, "grad_norm": 0.2806937098503113, "learning_rate": 2.9743370248538017e-06, "loss": 1.3328285217285156, "step": 1870 }, { "epoch": 0.8960573476702509, "grad_norm": 0.27521848678588867, "learning_rate": 2.8442956689162194e-06, "loss": 1.2944007873535157, "step": 1875 }, { "epoch": 0.8984468339307049, "grad_norm": 0.2810365557670593, "learning_rate": 2.717077760676984e-06, "loss": 1.277801513671875, "step": 1880 }, { "epoch": 0.9008363201911589, "grad_norm": 0.23585952818393707, "learning_rate": 2.592690917308066e-06, "loss": 1.3096345901489257, "step": 1885 }, { "epoch": 0.9032258064516129, "grad_norm": 0.28304553031921387, "learning_rate": 2.4711425864714645e-06, "loss": 1.3055765151977539, "step": 1890 }, { "epoch": 0.9056152927120669, "grad_norm": 0.2603612244129181, "learning_rate": 2.3524400458732334e-06, "loss": 1.2935433387756348, "step": 1895 }, { "epoch": 0.9080047789725209, "grad_norm": 0.2704894244670868, "learning_rate": 2.2365904028277376e-06, "loss": 1.3042230606079102, "step": 1900 }, { "epoch": 0.910394265232975, "grad_norm": 0.2532249093055725, "learning_rate": 2.123600593832109e-06, "loss": 1.310106086730957, "step": 1905 }, { "epoch": 0.9127837514934289, "grad_norm": 0.25696516036987305, "learning_rate": 2.0134773841509468e-06, "loss": 1.2937025070190429, "step": 1910 }, { "epoch": 0.9151732377538829, "grad_norm": 0.27187615633010864, "learning_rate": 1.906227367411173e-06, "loss": 1.3083663940429688, "step": 1915 }, { "epoch": 0.9175627240143369, "grad_norm": 0.30270397663116455, "learning_rate": 1.8018569652073381e-06, "loss": 1.2944793701171875, "step": 1920 }, { "epoch": 0.9199522102747909, "grad_norm": 0.2785067856311798, "learning_rate": 1.7003724267170396e-06, "loss": 1.2846717834472656, "step": 1925 }, { "epoch": 0.9223416965352449, "grad_norm": 0.3131451904773712, "learning_rate": 1.6017798283268215e-06, "loss": 1.2391252517700195, "step": 1930 }, { "epoch": 0.9247311827956989, "grad_norm": 0.2623656392097473, "learning_rate": 1.506085073268293e-06, "loss": 1.300617790222168, "step": 1935 }, { "epoch": 0.927120669056153, "grad_norm": 0.3671213388442993, "learning_rate": 1.413293891264722e-06, "loss": 1.33396635055542, "step": 1940 }, { "epoch": 0.929510155316607, "grad_norm": 0.45161667466163635, "learning_rate": 1.3234118381879378e-06, "loss": 1.3078412055969237, "step": 1945 }, { "epoch": 0.931899641577061, "grad_norm": 0.3107600510120392, "learning_rate": 1.236444295725675e-06, "loss": 1.2947958946228026, "step": 1950 }, { "epoch": 0.9342891278375149, "grad_norm": 0.27997028827667236, "learning_rate": 1.1523964710593639e-06, "loss": 1.2905611991882324, "step": 1955 }, { "epoch": 0.9366786140979689, "grad_norm": 0.2577410638332367, "learning_rate": 1.0712733965523246e-06, "loss": 1.2908943176269532, "step": 1960 }, { "epoch": 0.9390681003584229, "grad_norm": 0.2663538455963135, "learning_rate": 9.930799294484705e-07, "loss": 1.290733242034912, "step": 1965 }, { "epoch": 0.9414575866188769, "grad_norm": 0.29762157797813416, "learning_rate": 9.178207515814941e-07, "loss": 1.275669479370117, "step": 1970 }, { "epoch": 0.9438470728793309, "grad_norm": 0.2643646001815796, "learning_rate": 8.455003690945041e-07, "loss": 1.2908592224121094, "step": 1975 }, { "epoch": 0.946236559139785, "grad_norm": 0.2860666811466217, "learning_rate": 7.761231121702517e-07, "loss": 1.3699911117553711, "step": 1980 }, { "epoch": 0.948626045400239, "grad_norm": 0.26261815428733826, "learning_rate": 7.096931347718494e-07, "loss": 1.3525909423828124, "step": 1985 }, { "epoch": 0.951015531660693, "grad_norm": 0.2954842448234558, "learning_rate": 6.462144143940585e-07, "loss": 1.2190754890441895, "step": 1990 }, { "epoch": 0.953405017921147, "grad_norm": 0.3268016576766968, "learning_rate": 5.856907518251298e-07, "loss": 1.3280769348144532, "step": 1995 }, { "epoch": 0.955794504181601, "grad_norm": 0.24760957062244415, "learning_rate": 5.281257709192245e-07, "loss": 1.2637653350830078, "step": 2000 }, { "epoch": 0.9581839904420549, "grad_norm": 0.27675360441207886, "learning_rate": 4.7352291837947094e-07, "loss": 1.2968452453613282, "step": 2005 }, { "epoch": 0.9605734767025089, "grad_norm": 0.3023156225681305, "learning_rate": 4.2188546355153013e-07, "loss": 1.2937963485717774, "step": 2010 }, { "epoch": 0.9629629629629629, "grad_norm": 0.2493332475423813, "learning_rate": 3.7321649822791847e-07, "loss": 1.2865726470947265, "step": 2015 }, { "epoch": 0.965352449223417, "grad_norm": 0.2836077809333801, "learning_rate": 3.2751893646282284e-07, "loss": 1.2656401634216308, "step": 2020 }, { "epoch": 0.967741935483871, "grad_norm": 0.2933541238307953, "learning_rate": 2.8479551439766216e-07, "loss": 1.3273714065551758, "step": 2025 }, { "epoch": 0.970131421744325, "grad_norm": 0.30245599150657654, "learning_rate": 2.4504879009724647e-07, "loss": 1.2473869323730469, "step": 2030 }, { "epoch": 0.972520908004779, "grad_norm": 0.28488799929618835, "learning_rate": 2.082811433966103e-07, "loss": 1.3161026000976563, "step": 2035 }, { "epoch": 0.974910394265233, "grad_norm": 0.25747621059417725, "learning_rate": 1.7449477575851581e-07, "loss": 1.325145149230957, "step": 2040 }, { "epoch": 0.977299880525687, "grad_norm": 0.30804845690727234, "learning_rate": 1.4369171014165794e-07, "loss": 1.2972493171691895, "step": 2045 }, { "epoch": 0.9796893667861409, "grad_norm": 0.3299754858016968, "learning_rate": 1.1587379087952822e-07, "loss": 1.3106929779052734, "step": 2050 }, { "epoch": 0.982078853046595, "grad_norm": 0.28110501170158386, "learning_rate": 9.104268356998624e-08, "loss": 1.319981002807617, "step": 2055 }, { "epoch": 0.984468339307049, "grad_norm": 0.27178576588630676, "learning_rate": 6.919987497552294e-08, "loss": 1.2649413108825684, "step": 2060 }, { "epoch": 0.986857825567503, "grad_norm": 0.2620222568511963, "learning_rate": 5.0346672934270534e-08, "loss": 1.3299697875976562, "step": 2065 }, { "epoch": 0.989247311827957, "grad_norm": 0.27257755398750305, "learning_rate": 3.448420628164306e-08, "loss": 1.3144529342651368, "step": 2070 }, { "epoch": 0.991636798088411, "grad_norm": 0.32942599058151245, "learning_rate": 2.1613424782812586e-08, "loss": 1.29657039642334, "step": 2075 }, { "epoch": 0.994026284348865, "grad_norm": 0.31137213110923767, "learning_rate": 1.173509907579362e-08, "loss": 1.3135198593139648, "step": 2080 }, { "epoch": 0.996415770609319, "grad_norm": 0.2888542413711548, "learning_rate": 4.8498206253133616e-09, "loss": 1.2993997573852538, "step": 2085 }, { "epoch": 0.998805256869773, "grad_norm": 0.26138314604759216, "learning_rate": 9.580016874011223e-10, "loss": 1.3196189880371094, "step": 2090 } ], "logging_steps": 5, "max_steps": 2093, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.1098719100482355e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }