{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 4.36697247706422, "eval_steps": 500, "global_step": 5000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00436871996505024, "grad_norm": 9.839941024780273, "learning_rate": 8e-05, "loss": 2.5246, "step": 5 }, { "epoch": 0.00873743993010048, "grad_norm": 13.773455619812012, "learning_rate": 0.00018, "loss": 1.1343, "step": 10 }, { "epoch": 0.01310615989515072, "grad_norm": 5.6580424308776855, "learning_rate": 0.0001999997582552296, "loss": 0.7712, "step": 15 }, { "epoch": 0.01747487986020096, "grad_norm": 5.294467926025391, "learning_rate": 0.0001999987761691029, "loss": 0.73, "step": 20 }, { "epoch": 0.021843599825251202, "grad_norm": 2.8633503913879395, "learning_rate": 0.00019999703863998527, "loss": 0.7289, "step": 25 }, { "epoch": 0.02621231979030144, "grad_norm": 3.2836177349090576, "learning_rate": 0.00019999454568100293, "loss": 0.4686, "step": 30 }, { "epoch": 0.03058103975535168, "grad_norm": 4.878258228302002, "learning_rate": 0.00019999129731098898, "loss": 0.6629, "step": 35 }, { "epoch": 0.03494975972040192, "grad_norm": 2.899914026260376, "learning_rate": 0.00019998729355448326, "loss": 0.6038, "step": 40 }, { "epoch": 0.039318479685452164, "grad_norm": 3.289844274520874, "learning_rate": 0.00019998253444173235, "loss": 0.4573, "step": 45 }, { "epoch": 0.043687199650502405, "grad_norm": 2.957254648208618, "learning_rate": 0.00019997702000868896, "loss": 0.594, "step": 50 }, { "epoch": 0.048055919615552646, "grad_norm": 3.171276807785034, "learning_rate": 0.00019997075029701207, "loss": 0.5719, "step": 55 }, { "epoch": 0.05242463958060288, "grad_norm": 2.55605149269104, "learning_rate": 0.0001999637253540663, "loss": 0.5971, "step": 60 }, { "epoch": 0.05679335954565312, "grad_norm": 2.127289295196533, "learning_rate": 0.00019995594523292178, "loss": 0.5712, "step": 65 }, { "epoch": 0.06116207951070336, "grad_norm": 3.3928685188293457, "learning_rate": 0.00019994740999235359, "loss": 0.5712, "step": 70 }, { "epoch": 0.0655307994757536, "grad_norm": 2.6700279712677, "learning_rate": 0.00019993811969684142, "loss": 0.427, "step": 75 }, { "epoch": 0.06989951944080385, "grad_norm": 2.6936633586883545, "learning_rate": 0.00019992807441656898, "loss": 0.5321, "step": 80 }, { "epoch": 0.07426823940585409, "grad_norm": 3.9897687435150146, "learning_rate": 0.00019991727422742362, "loss": 0.6025, "step": 85 }, { "epoch": 0.07863695937090433, "grad_norm": 2.3496663570404053, "learning_rate": 0.00019990571921099553, "loss": 0.5975, "step": 90 }, { "epoch": 0.08300567933595457, "grad_norm": 3.3796467781066895, "learning_rate": 0.0001998934094545774, "loss": 0.5255, "step": 95 }, { "epoch": 0.08737439930100481, "grad_norm": 3.1103007793426514, "learning_rate": 0.00019988034505116352, "loss": 0.4946, "step": 100 }, { "epoch": 0.09174311926605505, "grad_norm": 2.002304792404175, "learning_rate": 0.00019986652609944926, "loss": 0.425, "step": 105 }, { "epoch": 0.09611183923110529, "grad_norm": 1.7572168111801147, "learning_rate": 0.00019985195270383018, "loss": 0.6073, "step": 110 }, { "epoch": 0.10048055919615553, "grad_norm": 2.745215654373169, "learning_rate": 0.00019983662497440133, "loss": 0.586, "step": 115 }, { "epoch": 0.10484927916120576, "grad_norm": 1.8170915842056274, "learning_rate": 0.0001998205430269564, "loss": 0.5255, "step": 120 }, { "epoch": 0.109217999126256, "grad_norm": 1.4944056272506714, "learning_rate": 0.00019980370698298677, "loss": 0.4219, "step": 125 }, { "epoch": 0.11358671909130624, "grad_norm": 1.6616989374160767, "learning_rate": 0.00019978611696968074, "loss": 0.4231, "step": 130 }, { "epoch": 0.11795543905635648, "grad_norm": 2.0523645877838135, "learning_rate": 0.00019976777311992247, "loss": 0.5298, "step": 135 }, { "epoch": 0.12232415902140673, "grad_norm": 2.065765619277954, "learning_rate": 0.00019974867557229098, "loss": 0.5228, "step": 140 }, { "epoch": 0.12669287898645698, "grad_norm": 1.7283438444137573, "learning_rate": 0.00019972882447105912, "loss": 0.3452, "step": 145 }, { "epoch": 0.1310615989515072, "grad_norm": 2.655750274658203, "learning_rate": 0.00019970821996619244, "loss": 0.508, "step": 150 }, { "epoch": 0.13543031891655744, "grad_norm": 2.67799973487854, "learning_rate": 0.0001996868622133482, "loss": 0.4359, "step": 155 }, { "epoch": 0.1397990388816077, "grad_norm": 1.6298809051513672, "learning_rate": 0.00019966475137387396, "loss": 0.5447, "step": 160 }, { "epoch": 0.14416775884665792, "grad_norm": 1.4772286415100098, "learning_rate": 0.00019964188761480657, "loss": 0.4105, "step": 165 }, { "epoch": 0.14853647881170817, "grad_norm": 2.2986271381378174, "learning_rate": 0.00019961827110887083, "loss": 0.603, "step": 170 }, { "epoch": 0.1529051987767584, "grad_norm": 2.8261911869049072, "learning_rate": 0.00019959390203447817, "loss": 0.4649, "step": 175 }, { "epoch": 0.15727391874180865, "grad_norm": 1.7771011590957642, "learning_rate": 0.00019956878057572524, "loss": 0.4394, "step": 180 }, { "epoch": 0.16164263870685888, "grad_norm": 1.7315421104431152, "learning_rate": 0.00019954290692239274, "loss": 0.5289, "step": 185 }, { "epoch": 0.16601135867190914, "grad_norm": 1.6124423742294312, "learning_rate": 0.00019951628126994373, "loss": 0.4173, "step": 190 }, { "epoch": 0.17038007863695936, "grad_norm": 1.792577862739563, "learning_rate": 0.00019948890381952232, "loss": 0.4331, "step": 195 }, { "epoch": 0.17474879860200962, "grad_norm": 1.9038774967193604, "learning_rate": 0.000199460774777952, "loss": 0.4247, "step": 200 }, { "epoch": 0.17911751856705985, "grad_norm": 2.457122802734375, "learning_rate": 0.00019943189435773432, "loss": 0.4519, "step": 205 }, { "epoch": 0.1834862385321101, "grad_norm": 1.97683584690094, "learning_rate": 0.00019940226277704706, "loss": 0.4761, "step": 210 }, { "epoch": 0.18785495849716033, "grad_norm": 2.1646862030029297, "learning_rate": 0.0001993718802597426, "loss": 0.5294, "step": 215 }, { "epoch": 0.19222367846221058, "grad_norm": 1.565412998199463, "learning_rate": 0.00019934074703534637, "loss": 0.3999, "step": 220 }, { "epoch": 0.1965923984272608, "grad_norm": 2.4315876960754395, "learning_rate": 0.00019930886333905504, "loss": 0.378, "step": 225 }, { "epoch": 0.20096111839231107, "grad_norm": 2.7567529678344727, "learning_rate": 0.00019927622941173467, "loss": 0.5075, "step": 230 }, { "epoch": 0.2053298383573613, "grad_norm": 1.8640387058258057, "learning_rate": 0.00019924284549991902, "loss": 0.4749, "step": 235 }, { "epoch": 0.20969855832241152, "grad_norm": 2.090924024581909, "learning_rate": 0.00019920871185580757, "loss": 0.4353, "step": 240 }, { "epoch": 0.21406727828746178, "grad_norm": 1.9691081047058105, "learning_rate": 0.00019917382873726376, "loss": 0.4051, "step": 245 }, { "epoch": 0.218435998252512, "grad_norm": 1.8130213022232056, "learning_rate": 0.0001991381964078128, "loss": 0.526, "step": 250 }, { "epoch": 0.22280471821756226, "grad_norm": 2.078805923461914, "learning_rate": 0.00019910181513664, "loss": 0.5654, "step": 255 }, { "epoch": 0.22717343818261249, "grad_norm": 2.0686287879943848, "learning_rate": 0.0001990646851985884, "loss": 0.43, "step": 260 }, { "epoch": 0.23154215814766274, "grad_norm": 1.475821614265442, "learning_rate": 0.00019902680687415705, "loss": 0.355, "step": 265 }, { "epoch": 0.23591087811271297, "grad_norm": 1.901236891746521, "learning_rate": 0.0001989881804494985, "loss": 0.4522, "step": 270 }, { "epoch": 0.24027959807776322, "grad_norm": 1.2583553791046143, "learning_rate": 0.00019894880621641704, "loss": 0.3869, "step": 275 }, { "epoch": 0.24464831804281345, "grad_norm": 1.712336540222168, "learning_rate": 0.00019890868447236613, "loss": 0.454, "step": 280 }, { "epoch": 0.2490170380078637, "grad_norm": 2.3967206478118896, "learning_rate": 0.00019886781552044634, "loss": 0.4074, "step": 285 }, { "epoch": 0.25338575797291396, "grad_norm": 2.0578925609588623, "learning_rate": 0.0001988261996694032, "loss": 0.4268, "step": 290 }, { "epoch": 0.2577544779379642, "grad_norm": 1.7411088943481445, "learning_rate": 0.0001987838372336245, "loss": 0.334, "step": 295 }, { "epoch": 0.2621231979030144, "grad_norm": 1.8145533800125122, "learning_rate": 0.0001987407285331382, "loss": 0.4019, "step": 300 }, { "epoch": 0.26649191786806464, "grad_norm": 1.3501653671264648, "learning_rate": 0.00019869687389361, "loss": 0.32, "step": 305 }, { "epoch": 0.27086063783311487, "grad_norm": 1.208422303199768, "learning_rate": 0.00019865227364634073, "loss": 0.4548, "step": 310 }, { "epoch": 0.27522935779816515, "grad_norm": 1.521690011024475, "learning_rate": 0.00019860692812826396, "loss": 0.3572, "step": 315 }, { "epoch": 0.2795980777632154, "grad_norm": 2.2849714756011963, "learning_rate": 0.0001985608376819434, "loss": 0.4555, "step": 320 }, { "epoch": 0.2839667977282656, "grad_norm": 2.7733798027038574, "learning_rate": 0.00019851400265557037, "loss": 0.4726, "step": 325 }, { "epoch": 0.28833551769331583, "grad_norm": 1.973522424697876, "learning_rate": 0.00019846642340296114, "loss": 0.4585, "step": 330 }, { "epoch": 0.2927042376583661, "grad_norm": 1.7133642435073853, "learning_rate": 0.0001984181002835542, "loss": 0.4679, "step": 335 }, { "epoch": 0.29707295762341634, "grad_norm": 2.8383235931396484, "learning_rate": 0.00019836903366240768, "loss": 0.4119, "step": 340 }, { "epoch": 0.30144167758846657, "grad_norm": 2.798276901245117, "learning_rate": 0.00019831922391019645, "loss": 0.3665, "step": 345 }, { "epoch": 0.3058103975535168, "grad_norm": 2.171276569366455, "learning_rate": 0.00019826867140320938, "loss": 0.5691, "step": 350 }, { "epoch": 0.3101791175185671, "grad_norm": 2.0866177082061768, "learning_rate": 0.00019821737652334653, "loss": 0.4074, "step": 355 }, { "epoch": 0.3145478374836173, "grad_norm": 1.3713918924331665, "learning_rate": 0.0001981653396581162, "loss": 0.3379, "step": 360 }, { "epoch": 0.31891655744866754, "grad_norm": 1.6086684465408325, "learning_rate": 0.0001981125612006321, "loss": 0.3563, "step": 365 }, { "epoch": 0.32328527741371776, "grad_norm": 2.655686378479004, "learning_rate": 0.0001980590415496102, "loss": 0.3988, "step": 370 }, { "epoch": 0.32765399737876805, "grad_norm": 1.5271559953689575, "learning_rate": 0.00019800478110936596, "loss": 0.5784, "step": 375 }, { "epoch": 0.3320227173438183, "grad_norm": 1.3043195009231567, "learning_rate": 0.00019794978028981106, "loss": 0.2637, "step": 380 }, { "epoch": 0.3363914373088685, "grad_norm": 2.539109706878662, "learning_rate": 0.0001978940395064504, "loss": 0.4658, "step": 385 }, { "epoch": 0.34076015727391873, "grad_norm": 1.7521268129348755, "learning_rate": 0.00019783755918037903, "loss": 0.4253, "step": 390 }, { "epoch": 0.34512887723896896, "grad_norm": 1.5679692029953003, "learning_rate": 0.00019778033973827882, "loss": 0.4528, "step": 395 }, { "epoch": 0.34949759720401924, "grad_norm": 1.670640468597412, "learning_rate": 0.00019772238161241528, "loss": 0.3724, "step": 400 }, { "epoch": 0.35386631716906947, "grad_norm": 1.520856261253357, "learning_rate": 0.00019766368524063438, "loss": 0.4141, "step": 405 }, { "epoch": 0.3582350371341197, "grad_norm": 1.0802158117294312, "learning_rate": 0.00019760425106635926, "loss": 0.3268, "step": 410 }, { "epoch": 0.3626037570991699, "grad_norm": 1.7306379079818726, "learning_rate": 0.0001975440795385866, "loss": 0.3654, "step": 415 }, { "epoch": 0.3669724770642202, "grad_norm": 1.5037274360656738, "learning_rate": 0.0001974831711118836, "loss": 0.4285, "step": 420 }, { "epoch": 0.37134119702927043, "grad_norm": 1.4654844999313354, "learning_rate": 0.00019742152624638437, "loss": 0.2548, "step": 425 }, { "epoch": 0.37570991699432066, "grad_norm": 2.6770753860473633, "learning_rate": 0.00019735914540778638, "loss": 0.4238, "step": 430 }, { "epoch": 0.3800786369593709, "grad_norm": 1.1864055395126343, "learning_rate": 0.00019729602906734704, "loss": 0.3959, "step": 435 }, { "epoch": 0.38444735692442117, "grad_norm": 1.904876708984375, "learning_rate": 0.00019723217770188024, "loss": 0.3603, "step": 440 }, { "epoch": 0.3888160768894714, "grad_norm": 1.7086598873138428, "learning_rate": 0.0001971675917937525, "loss": 0.551, "step": 445 }, { "epoch": 0.3931847968545216, "grad_norm": 1.4635995626449585, "learning_rate": 0.00019710227183087947, "loss": 0.3738, "step": 450 }, { "epoch": 0.39755351681957185, "grad_norm": 1.6047295331954956, "learning_rate": 0.00019703621830672238, "loss": 0.475, "step": 455 }, { "epoch": 0.40192223678462213, "grad_norm": 1.4741933345794678, "learning_rate": 0.00019696943172028394, "loss": 0.4021, "step": 460 }, { "epoch": 0.40629095674967236, "grad_norm": 2.8138020038604736, "learning_rate": 0.00019690191257610497, "loss": 0.3665, "step": 465 }, { "epoch": 0.4106596767147226, "grad_norm": 1.6264874935150146, "learning_rate": 0.00019683366138426034, "loss": 0.3598, "step": 470 }, { "epoch": 0.4150283966797728, "grad_norm": 1.6185061931610107, "learning_rate": 0.00019676467866035525, "loss": 0.5003, "step": 475 }, { "epoch": 0.41939711664482304, "grad_norm": 1.8654040098190308, "learning_rate": 0.00019669496492552113, "loss": 0.397, "step": 480 }, { "epoch": 0.4237658366098733, "grad_norm": 1.2525237798690796, "learning_rate": 0.00019662452070641205, "loss": 0.3235, "step": 485 }, { "epoch": 0.42813455657492355, "grad_norm": 1.7755401134490967, "learning_rate": 0.00019655334653520036, "loss": 0.2978, "step": 490 }, { "epoch": 0.4325032765399738, "grad_norm": 1.6025470495224, "learning_rate": 0.00019648144294957297, "loss": 0.4436, "step": 495 }, { "epoch": 0.436871996505024, "grad_norm": 1.085461974143982, "learning_rate": 0.00019640881049272713, "loss": 0.22, "step": 500 }, { "epoch": 0.4412407164700743, "grad_norm": 1.491818904876709, "learning_rate": 0.00019633544971336636, "loss": 0.2714, "step": 505 }, { "epoch": 0.4456094364351245, "grad_norm": 0.9479840993881226, "learning_rate": 0.0001962613611656963, "loss": 0.3735, "step": 510 }, { "epoch": 0.44997815640017474, "grad_norm": 3.0529448986053467, "learning_rate": 0.0001961865454094205, "loss": 0.4779, "step": 515 }, { "epoch": 0.45434687636522497, "grad_norm": 2.831089973449707, "learning_rate": 0.00019611100300973635, "loss": 0.469, "step": 520 }, { "epoch": 0.45871559633027525, "grad_norm": 2.1834311485290527, "learning_rate": 0.00019603473453733052, "loss": 0.4163, "step": 525 }, { "epoch": 0.4630843162953255, "grad_norm": 1.3152204751968384, "learning_rate": 0.00019595774056837493, "loss": 0.3744, "step": 530 }, { "epoch": 0.4674530362603757, "grad_norm": 1.4493387937545776, "learning_rate": 0.00019588002168452223, "loss": 0.3117, "step": 535 }, { "epoch": 0.47182175622542594, "grad_norm": 1.1412076950073242, "learning_rate": 0.00019580157847290147, "loss": 0.3152, "step": 540 }, { "epoch": 0.47619047619047616, "grad_norm": 1.5004645586013794, "learning_rate": 0.00019572241152611365, "loss": 0.3271, "step": 545 }, { "epoch": 0.48055919615552645, "grad_norm": 2.3333992958068848, "learning_rate": 0.0001956425214422272, "loss": 0.3626, "step": 550 }, { "epoch": 0.4849279161205767, "grad_norm": 1.5423107147216797, "learning_rate": 0.0001955619088247736, "loss": 0.4588, "step": 555 }, { "epoch": 0.4892966360856269, "grad_norm": 3.008280038833618, "learning_rate": 0.00019548057428274266, "loss": 0.5275, "step": 560 }, { "epoch": 0.49366535605067713, "grad_norm": 1.0968583822250366, "learning_rate": 0.00019539851843057798, "loss": 0.3233, "step": 565 }, { "epoch": 0.4980340760157274, "grad_norm": 1.265228271484375, "learning_rate": 0.00019531574188817234, "loss": 0.2743, "step": 570 }, { "epoch": 0.5024027959807776, "grad_norm": 1.9382916688919067, "learning_rate": 0.000195232245280863, "loss": 0.3189, "step": 575 }, { "epoch": 0.5067715159458279, "grad_norm": 1.6710058450698853, "learning_rate": 0.00019514802923942687, "loss": 0.345, "step": 580 }, { "epoch": 0.5111402359108781, "grad_norm": 1.8377633094787598, "learning_rate": 0.000195063094400076, "loss": 0.4441, "step": 585 }, { "epoch": 0.5155089558759284, "grad_norm": 1.432173728942871, "learning_rate": 0.0001949774414044525, "loss": 0.3277, "step": 590 }, { "epoch": 0.5198776758409785, "grad_norm": 1.096330165863037, "learning_rate": 0.0001948910708996239, "loss": 0.3821, "step": 595 }, { "epoch": 0.5242463958060288, "grad_norm": 1.1951391696929932, "learning_rate": 0.00019480398353807798, "loss": 0.4303, "step": 600 }, { "epoch": 0.5286151157710791, "grad_norm": 0.9764880537986755, "learning_rate": 0.0001947161799777183, "loss": 0.2693, "step": 605 }, { "epoch": 0.5329838357361293, "grad_norm": 1.2566354274749756, "learning_rate": 0.00019462766088185874, "loss": 0.2851, "step": 610 }, { "epoch": 0.5373525557011796, "grad_norm": 1.494903802871704, "learning_rate": 0.0001945384269192188, "loss": 0.36, "step": 615 }, { "epoch": 0.5417212756662297, "grad_norm": 1.5508995056152344, "learning_rate": 0.00019444847876391844, "loss": 0.3682, "step": 620 }, { "epoch": 0.54608999563128, "grad_norm": 2.227889060974121, "learning_rate": 0.00019435781709547305, "loss": 0.3889, "step": 625 }, { "epoch": 0.5504587155963303, "grad_norm": 0.9221494197845459, "learning_rate": 0.0001942664425987882, "loss": 0.3375, "step": 630 }, { "epoch": 0.5548274355613805, "grad_norm": 1.3386973142623901, "learning_rate": 0.00019417435596415458, "loss": 0.4833, "step": 635 }, { "epoch": 0.5591961555264308, "grad_norm": 1.9686752557754517, "learning_rate": 0.00019408155788724272, "loss": 0.4739, "step": 640 }, { "epoch": 0.563564875491481, "grad_norm": 2.3978073596954346, "learning_rate": 0.00019398804906909777, "loss": 0.4681, "step": 645 }, { "epoch": 0.5679335954565312, "grad_norm": 1.536699652671814, "learning_rate": 0.0001938938302161342, "loss": 0.2684, "step": 650 }, { "epoch": 0.5723023154215815, "grad_norm": 1.691787600517273, "learning_rate": 0.00019379890204013043, "loss": 0.3512, "step": 655 }, { "epoch": 0.5766710353866317, "grad_norm": 1.7557870149612427, "learning_rate": 0.0001937032652582235, "loss": 0.3423, "step": 660 }, { "epoch": 0.581039755351682, "grad_norm": 1.7950220108032227, "learning_rate": 0.0001936069205929036, "loss": 0.2831, "step": 665 }, { "epoch": 0.5854084753167322, "grad_norm": 1.928232192993164, "learning_rate": 0.00019350986877200867, "loss": 0.323, "step": 670 }, { "epoch": 0.5897771952817824, "grad_norm": 1.86429762840271, "learning_rate": 0.00019341211052871887, "loss": 0.4248, "step": 675 }, { "epoch": 0.5941459152468327, "grad_norm": 2.022738456726074, "learning_rate": 0.00019331364660155103, "loss": 0.3411, "step": 680 }, { "epoch": 0.598514635211883, "grad_norm": 1.2337995767593384, "learning_rate": 0.00019321447773435306, "loss": 0.2368, "step": 685 }, { "epoch": 0.6028833551769331, "grad_norm": 2.015075445175171, "learning_rate": 0.00019311460467629843, "loss": 0.5116, "step": 690 }, { "epoch": 0.6072520751419834, "grad_norm": 1.2344030141830444, "learning_rate": 0.00019301402818188036, "loss": 0.3313, "step": 695 }, { "epoch": 0.6116207951070336, "grad_norm": 1.129764437675476, "learning_rate": 0.00019291274901090625, "loss": 0.408, "step": 700 }, { "epoch": 0.6159895150720839, "grad_norm": 1.4350385665893555, "learning_rate": 0.00019281076792849184, "loss": 0.3729, "step": 705 }, { "epoch": 0.6203582350371342, "grad_norm": 1.9586119651794434, "learning_rate": 0.00019270808570505553, "loss": 0.4315, "step": 710 }, { "epoch": 0.6247269550021843, "grad_norm": 1.0157238245010376, "learning_rate": 0.00019260470311631243, "loss": 0.2861, "step": 715 }, { "epoch": 0.6290956749672346, "grad_norm": 1.3841652870178223, "learning_rate": 0.00019250062094326864, "loss": 0.4037, "step": 720 }, { "epoch": 0.6334643949322848, "grad_norm": 1.848821997642517, "learning_rate": 0.00019239583997221525, "loss": 0.3665, "step": 725 }, { "epoch": 0.6378331148973351, "grad_norm": 0.9416481256484985, "learning_rate": 0.0001922903609947225, "loss": 0.339, "step": 730 }, { "epoch": 0.6422018348623854, "grad_norm": 1.0696804523468018, "learning_rate": 0.0001921841848076336, "loss": 0.2783, "step": 735 }, { "epoch": 0.6465705548274355, "grad_norm": 1.9199622869491577, "learning_rate": 0.00019207731221305903, "loss": 0.2904, "step": 740 }, { "epoch": 0.6509392747924858, "grad_norm": 1.347430944442749, "learning_rate": 0.00019196974401837008, "loss": 0.2719, "step": 745 }, { "epoch": 0.6553079947575361, "grad_norm": 0.9743670225143433, "learning_rate": 0.0001918614810361932, "loss": 0.2748, "step": 750 }, { "epoch": 0.6596767147225863, "grad_norm": 1.4043099880218506, "learning_rate": 0.00019175252408440343, "loss": 0.3285, "step": 755 }, { "epoch": 0.6640454346876365, "grad_norm": 2.9343338012695312, "learning_rate": 0.0001916428739861185, "loss": 0.4962, "step": 760 }, { "epoch": 0.6684141546526867, "grad_norm": 2.3201515674591064, "learning_rate": 0.0001915325315696926, "loss": 0.3243, "step": 765 }, { "epoch": 0.672782874617737, "grad_norm": 1.675564169883728, "learning_rate": 0.00019142149766870992, "loss": 0.4596, "step": 770 }, { "epoch": 0.6771515945827873, "grad_norm": 1.664604663848877, "learning_rate": 0.00019130977312197854, "loss": 0.3024, "step": 775 }, { "epoch": 0.6815203145478375, "grad_norm": 1.8358148336410522, "learning_rate": 0.00019119735877352412, "loss": 0.3862, "step": 780 }, { "epoch": 0.6858890345128877, "grad_norm": 1.3632128238677979, "learning_rate": 0.00019108425547258328, "loss": 0.2374, "step": 785 }, { "epoch": 0.6902577544779379, "grad_norm": 2.0279934406280518, "learning_rate": 0.0001909704640735975, "loss": 0.4392, "step": 790 }, { "epoch": 0.6946264744429882, "grad_norm": 1.2824902534484863, "learning_rate": 0.0001908559854362064, "loss": 0.2782, "step": 795 }, { "epoch": 0.6989951944080385, "grad_norm": 1.3477047681808472, "learning_rate": 0.00019074082042524145, "loss": 0.3631, "step": 800 }, { "epoch": 0.7033639143730887, "grad_norm": 1.8478046655654907, "learning_rate": 0.00019062496991071928, "loss": 0.3788, "step": 805 }, { "epoch": 0.7077326343381389, "grad_norm": 1.470382571220398, "learning_rate": 0.0001905084347678352, "loss": 0.3825, "step": 810 }, { "epoch": 0.7121013543031892, "grad_norm": 2.4951813220977783, "learning_rate": 0.00019039121587695652, "loss": 0.3359, "step": 815 }, { "epoch": 0.7164700742682394, "grad_norm": 2.3441359996795654, "learning_rate": 0.000190273314123616, "loss": 0.32, "step": 820 }, { "epoch": 0.7208387942332897, "grad_norm": 2.372884750366211, "learning_rate": 0.00019015473039850513, "loss": 0.3651, "step": 825 }, { "epoch": 0.7252075141983398, "grad_norm": 2.4474101066589355, "learning_rate": 0.0001900354655974672, "loss": 0.4401, "step": 830 }, { "epoch": 0.7295762341633901, "grad_norm": 1.4031054973602295, "learning_rate": 0.0001899155206214909, "loss": 0.308, "step": 835 }, { "epoch": 0.7339449541284404, "grad_norm": 1.6008141040802002, "learning_rate": 0.00018979489637670322, "loss": 0.2937, "step": 840 }, { "epoch": 0.7383136740934906, "grad_norm": 0.9202178120613098, "learning_rate": 0.0001896735937743627, "loss": 0.3157, "step": 845 }, { "epoch": 0.7426823940585409, "grad_norm": 1.024746298789978, "learning_rate": 0.00018955161373085253, "loss": 0.2934, "step": 850 }, { "epoch": 0.747051114023591, "grad_norm": 1.1573566198349, "learning_rate": 0.00018942895716767374, "loss": 0.3617, "step": 855 }, { "epoch": 0.7514198339886413, "grad_norm": 1.227409839630127, "learning_rate": 0.00018930562501143805, "loss": 0.3581, "step": 860 }, { "epoch": 0.7557885539536916, "grad_norm": 1.5460100173950195, "learning_rate": 0.00018918161819386095, "loss": 0.3393, "step": 865 }, { "epoch": 0.7601572739187418, "grad_norm": 1.688852310180664, "learning_rate": 0.0001890569376517548, "loss": 0.4389, "step": 870 }, { "epoch": 0.764525993883792, "grad_norm": 1.5271598100662231, "learning_rate": 0.00018893158432702149, "loss": 0.2915, "step": 875 }, { "epoch": 0.7688947138488423, "grad_norm": 1.695788860321045, "learning_rate": 0.00018880555916664555, "loss": 0.4026, "step": 880 }, { "epoch": 0.7732634338138925, "grad_norm": 1.6879792213439941, "learning_rate": 0.00018867886312268683, "loss": 0.2857, "step": 885 }, { "epoch": 0.7776321537789428, "grad_norm": 2.0718719959259033, "learning_rate": 0.00018855149715227344, "loss": 0.4236, "step": 890 }, { "epoch": 0.782000873743993, "grad_norm": 1.5112775564193726, "learning_rate": 0.00018842346221759448, "loss": 0.325, "step": 895 }, { "epoch": 0.7863695937090432, "grad_norm": 1.2844749689102173, "learning_rate": 0.00018829475928589271, "loss": 0.3782, "step": 900 }, { "epoch": 0.7907383136740935, "grad_norm": 2.150299072265625, "learning_rate": 0.00018816538932945728, "loss": 0.3726, "step": 905 }, { "epoch": 0.7951070336391437, "grad_norm": 1.7050650119781494, "learning_rate": 0.00018803535332561646, "loss": 0.3824, "step": 910 }, { "epoch": 0.799475753604194, "grad_norm": 1.8164982795715332, "learning_rate": 0.00018790465225673012, "loss": 0.3664, "step": 915 }, { "epoch": 0.8038444735692443, "grad_norm": 1.1102941036224365, "learning_rate": 0.00018777328711018244, "loss": 0.3166, "step": 920 }, { "epoch": 0.8082131935342944, "grad_norm": 1.4220764636993408, "learning_rate": 0.0001876412588783743, "loss": 0.3049, "step": 925 }, { "epoch": 0.8125819134993447, "grad_norm": 2.11336088180542, "learning_rate": 0.000187508568558716, "loss": 0.3076, "step": 930 }, { "epoch": 0.8169506334643949, "grad_norm": 1.9948710203170776, "learning_rate": 0.00018737521715361948, "loss": 0.3846, "step": 935 }, { "epoch": 0.8213193534294452, "grad_norm": 1.8913676738739014, "learning_rate": 0.00018724120567049094, "loss": 0.4296, "step": 940 }, { "epoch": 0.8256880733944955, "grad_norm": 1.3633447885513306, "learning_rate": 0.0001871065351217231, "loss": 0.3569, "step": 945 }, { "epoch": 0.8300567933595456, "grad_norm": 1.4957417249679565, "learning_rate": 0.00018697120652468762, "loss": 0.3085, "step": 950 }, { "epoch": 0.8344255133245959, "grad_norm": 2.076399803161621, "learning_rate": 0.0001868352209017275, "loss": 0.3331, "step": 955 }, { "epoch": 0.8387942332896461, "grad_norm": 1.1817855834960938, "learning_rate": 0.00018669857928014906, "loss": 0.3414, "step": 960 }, { "epoch": 0.8431629532546964, "grad_norm": 1.4255414009094238, "learning_rate": 0.00018656128269221454, "loss": 0.2782, "step": 965 }, { "epoch": 0.8475316732197467, "grad_norm": 1.326687216758728, "learning_rate": 0.0001864233321751341, "loss": 0.2998, "step": 970 }, { "epoch": 0.8519003931847968, "grad_norm": 2.222280263900757, "learning_rate": 0.00018628472877105793, "loss": 0.3348, "step": 975 }, { "epoch": 0.8562691131498471, "grad_norm": 1.518401026725769, "learning_rate": 0.00018614547352706863, "loss": 0.3816, "step": 980 }, { "epoch": 0.8606378331148974, "grad_norm": 1.1030207872390747, "learning_rate": 0.00018600556749517305, "loss": 0.3222, "step": 985 }, { "epoch": 0.8650065530799476, "grad_norm": 2.406994104385376, "learning_rate": 0.00018586501173229437, "loss": 0.3754, "step": 990 }, { "epoch": 0.8693752730449978, "grad_norm": 1.2401646375656128, "learning_rate": 0.00018572380730026434, "loss": 0.4402, "step": 995 }, { "epoch": 0.873743993010048, "grad_norm": 2.0233402252197266, "learning_rate": 0.0001855819552658149, "loss": 0.3323, "step": 1000 }, { "epoch": 0.8781127129750983, "grad_norm": 1.5329450368881226, "learning_rate": 0.00018543945670057045, "loss": 0.235, "step": 1005 }, { "epoch": 0.8824814329401486, "grad_norm": 1.8849459886550903, "learning_rate": 0.00018529631268103964, "loss": 0.357, "step": 1010 }, { "epoch": 0.8868501529051988, "grad_norm": 2.016646146774292, "learning_rate": 0.0001851525242886071, "loss": 0.2663, "step": 1015 }, { "epoch": 0.891218872870249, "grad_norm": 2.3272440433502197, "learning_rate": 0.0001850080926095255, "loss": 0.2926, "step": 1020 }, { "epoch": 0.8955875928352992, "grad_norm": 1.7760261297225952, "learning_rate": 0.00018486301873490713, "loss": 0.4155, "step": 1025 }, { "epoch": 0.8999563128003495, "grad_norm": 1.4679979085922241, "learning_rate": 0.0001847173037607159, "loss": 0.2877, "step": 1030 }, { "epoch": 0.9043250327653998, "grad_norm": 1.8398054838180542, "learning_rate": 0.0001845709487877588, "loss": 0.2856, "step": 1035 }, { "epoch": 0.9086937527304499, "grad_norm": 3.05880069732666, "learning_rate": 0.00018442395492167775, "loss": 0.3373, "step": 1040 }, { "epoch": 0.9130624726955002, "grad_norm": 1.2527328729629517, "learning_rate": 0.0001842763232729412, "loss": 0.2412, "step": 1045 }, { "epoch": 0.9174311926605505, "grad_norm": 1.7745814323425293, "learning_rate": 0.00018412805495683575, "loss": 0.3955, "step": 1050 }, { "epoch": 0.9217999126256007, "grad_norm": 3.2864468097686768, "learning_rate": 0.0001839791510934577, "loss": 0.333, "step": 1055 }, { "epoch": 0.926168632590651, "grad_norm": 2.0274927616119385, "learning_rate": 0.0001838296128077046, "loss": 0.4004, "step": 1060 }, { "epoch": 0.9305373525557011, "grad_norm": 1.9851633310317993, "learning_rate": 0.0001836794412292668, "loss": 0.3132, "step": 1065 }, { "epoch": 0.9349060725207514, "grad_norm": 1.3309999704360962, "learning_rate": 0.00018352863749261883, "loss": 0.2645, "step": 1070 }, { "epoch": 0.9392747924858017, "grad_norm": 2.0173072814941406, "learning_rate": 0.00018337720273701088, "loss": 0.4376, "step": 1075 }, { "epoch": 0.9436435124508519, "grad_norm": 1.815408706665039, "learning_rate": 0.00018322513810646024, "loss": 0.2851, "step": 1080 }, { "epoch": 0.9480122324159022, "grad_norm": 1.1190584897994995, "learning_rate": 0.00018307244474974254, "loss": 0.4664, "step": 1085 }, { "epoch": 0.9523809523809523, "grad_norm": 0.9746566414833069, "learning_rate": 0.00018291912382038317, "loss": 0.3816, "step": 1090 }, { "epoch": 0.9567496723460026, "grad_norm": 1.9062715768814087, "learning_rate": 0.0001827651764766485, "loss": 0.3031, "step": 1095 }, { "epoch": 0.9611183923110529, "grad_norm": 1.027502417564392, "learning_rate": 0.00018261060388153718, "loss": 0.2657, "step": 1100 }, { "epoch": 0.9654871122761031, "grad_norm": 2.239164352416992, "learning_rate": 0.00018245540720277135, "loss": 0.3367, "step": 1105 }, { "epoch": 0.9698558322411533, "grad_norm": 1.5922635793685913, "learning_rate": 0.0001822995876127878, "loss": 0.3044, "step": 1110 }, { "epoch": 0.9742245522062036, "grad_norm": 1.9189236164093018, "learning_rate": 0.00018214314628872905, "loss": 0.3326, "step": 1115 }, { "epoch": 0.9785932721712538, "grad_norm": 1.1626375913619995, "learning_rate": 0.00018198608441243467, "loss": 0.2761, "step": 1120 }, { "epoch": 0.9829619921363041, "grad_norm": 1.805367112159729, "learning_rate": 0.00018182840317043202, "loss": 0.3337, "step": 1125 }, { "epoch": 0.9873307121013543, "grad_norm": 1.5879418849945068, "learning_rate": 0.0001816701037539277, "loss": 0.3242, "step": 1130 }, { "epoch": 0.9916994320664045, "grad_norm": 1.3560898303985596, "learning_rate": 0.00018151118735879805, "loss": 0.2794, "step": 1135 }, { "epoch": 0.9960681520314548, "grad_norm": 1.0656763315200806, "learning_rate": 0.0001813516551855806, "loss": 0.3336, "step": 1140 }, { "epoch": 1.0, "grad_norm": 3.2105913162231445, "learning_rate": 0.00018119150843946472, "loss": 0.3753, "step": 1145 }, { "epoch": 1.0043687199650502, "grad_norm": 1.2890548706054688, "learning_rate": 0.00018103074833028258, "loss": 0.2943, "step": 1150 }, { "epoch": 1.0087374399301006, "grad_norm": 1.8480075597763062, "learning_rate": 0.00018086937607250002, "loss": 0.3057, "step": 1155 }, { "epoch": 1.0131061598951507, "grad_norm": 1.3526337146759033, "learning_rate": 0.00018070739288520736, "loss": 0.2328, "step": 1160 }, { "epoch": 1.017474879860201, "grad_norm": 1.0462696552276611, "learning_rate": 0.00018054479999211025, "loss": 0.1896, "step": 1165 }, { "epoch": 1.0218435998252513, "grad_norm": 0.9586630463600159, "learning_rate": 0.00018038159862152027, "loss": 0.2567, "step": 1170 }, { "epoch": 1.0262123197903015, "grad_norm": 1.0181586742401123, "learning_rate": 0.0001802177900063459, "loss": 0.2653, "step": 1175 }, { "epoch": 1.0305810397553516, "grad_norm": 1.2034084796905518, "learning_rate": 0.0001800533753840829, "loss": 0.3953, "step": 1180 }, { "epoch": 1.0349497597204018, "grad_norm": 2.6563191413879395, "learning_rate": 0.0001798883559968053, "loss": 0.185, "step": 1185 }, { "epoch": 1.0393184796854522, "grad_norm": 1.2556034326553345, "learning_rate": 0.00017972273309115568, "loss": 0.2452, "step": 1190 }, { "epoch": 1.0436871996505024, "grad_norm": 1.4751702547073364, "learning_rate": 0.00017955650791833604, "loss": 0.2827, "step": 1195 }, { "epoch": 1.0480559196155526, "grad_norm": 3.8620717525482178, "learning_rate": 0.00017938968173409811, "loss": 0.2953, "step": 1200 }, { "epoch": 1.052424639580603, "grad_norm": 1.2123383283615112, "learning_rate": 0.00017922225579873407, "loss": 0.2165, "step": 1205 }, { "epoch": 1.0567933595456531, "grad_norm": 1.911566972732544, "learning_rate": 0.0001790542313770669, "loss": 0.2444, "step": 1210 }, { "epoch": 1.0611620795107033, "grad_norm": 1.9949162006378174, "learning_rate": 0.00017888560973844083, "loss": 0.255, "step": 1215 }, { "epoch": 1.0655307994757537, "grad_norm": 0.9666941165924072, "learning_rate": 0.0001787163921567118, "loss": 0.1913, "step": 1220 }, { "epoch": 1.0698995194408039, "grad_norm": 0.7195447087287903, "learning_rate": 0.0001785465799102378, "loss": 0.2541, "step": 1225 }, { "epoch": 1.074268239405854, "grad_norm": 1.5414377450942993, "learning_rate": 0.0001783761742818693, "loss": 0.3626, "step": 1230 }, { "epoch": 1.0786369593709044, "grad_norm": 1.2173676490783691, "learning_rate": 0.0001782051765589394, "loss": 0.2913, "step": 1235 }, { "epoch": 1.0830056793359546, "grad_norm": 1.6966580152511597, "learning_rate": 0.00017803358803325416, "loss": 0.2613, "step": 1240 }, { "epoch": 1.0873743993010048, "grad_norm": 1.8033946752548218, "learning_rate": 0.00017786141000108302, "loss": 0.2734, "step": 1245 }, { "epoch": 1.091743119266055, "grad_norm": 1.1834598779678345, "learning_rate": 0.00017768864376314873, "loss": 0.2548, "step": 1250 }, { "epoch": 1.0961118392311053, "grad_norm": 1.441835641860962, "learning_rate": 0.00017751529062461777, "loss": 0.3404, "step": 1255 }, { "epoch": 1.1004805591961555, "grad_norm": 1.443575382232666, "learning_rate": 0.0001773413518950902, "loss": 0.312, "step": 1260 }, { "epoch": 1.1048492791612057, "grad_norm": 1.2344982624053955, "learning_rate": 0.0001771668288885901, "loss": 0.2594, "step": 1265 }, { "epoch": 1.109217999126256, "grad_norm": 1.3508884906768799, "learning_rate": 0.0001769917229235554, "loss": 0.3467, "step": 1270 }, { "epoch": 1.1135867190913062, "grad_norm": 1.3105831146240234, "learning_rate": 0.00017681603532282805, "loss": 0.2393, "step": 1275 }, { "epoch": 1.1179554390563564, "grad_norm": 1.3476370573043823, "learning_rate": 0.00017663976741364394, "loss": 0.3318, "step": 1280 }, { "epoch": 1.1223241590214068, "grad_norm": 1.9715685844421387, "learning_rate": 0.00017646292052762296, "loss": 0.2808, "step": 1285 }, { "epoch": 1.126692878986457, "grad_norm": 1.2339669466018677, "learning_rate": 0.00017628549600075884, "loss": 0.2753, "step": 1290 }, { "epoch": 1.1310615989515072, "grad_norm": 1.3921184539794922, "learning_rate": 0.00017610749517340914, "loss": 0.2096, "step": 1295 }, { "epoch": 1.1354303189165575, "grad_norm": 1.3537594079971313, "learning_rate": 0.0001759289193902851, "loss": 0.2232, "step": 1300 }, { "epoch": 1.1397990388816077, "grad_norm": 2.207932472229004, "learning_rate": 0.00017574977000044147, "loss": 0.4179, "step": 1305 }, { "epoch": 1.144167758846658, "grad_norm": 2.4464988708496094, "learning_rate": 0.0001755700483572663, "loss": 0.3863, "step": 1310 }, { "epoch": 1.148536478811708, "grad_norm": 1.2169779539108276, "learning_rate": 0.00017538975581847077, "loss": 0.2131, "step": 1315 }, { "epoch": 1.1529051987767585, "grad_norm": 1.2162644863128662, "learning_rate": 0.00017520889374607893, "loss": 0.2299, "step": 1320 }, { "epoch": 1.1572739187418086, "grad_norm": 1.7051069736480713, "learning_rate": 0.0001750274635064173, "loss": 0.2994, "step": 1325 }, { "epoch": 1.1616426387068588, "grad_norm": 1.814450740814209, "learning_rate": 0.00017484546647010473, "loss": 0.2948, "step": 1330 }, { "epoch": 1.1660113586719092, "grad_norm": 1.3929287195205688, "learning_rate": 0.00017466290401204186, "loss": 0.3837, "step": 1335 }, { "epoch": 1.1703800786369594, "grad_norm": 1.1973505020141602, "learning_rate": 0.00017447977751140086, "loss": 0.2335, "step": 1340 }, { "epoch": 1.1747487986020095, "grad_norm": 1.4987421035766602, "learning_rate": 0.00017429608835161506, "loss": 0.2484, "step": 1345 }, { "epoch": 1.17911751856706, "grad_norm": 1.3741153478622437, "learning_rate": 0.00017411183792036822, "loss": 0.2475, "step": 1350 }, { "epoch": 1.18348623853211, "grad_norm": 1.8292722702026367, "learning_rate": 0.0001739270276095844, "loss": 0.2898, "step": 1355 }, { "epoch": 1.1878549584971603, "grad_norm": 1.854209065437317, "learning_rate": 0.00017374165881541717, "loss": 0.3992, "step": 1360 }, { "epoch": 1.1922236784622107, "grad_norm": 1.6196980476379395, "learning_rate": 0.0001735557329382393, "loss": 0.3053, "step": 1365 }, { "epoch": 1.1965923984272608, "grad_norm": 1.6935441493988037, "learning_rate": 0.00017336925138263195, "loss": 0.239, "step": 1370 }, { "epoch": 1.200961118392311, "grad_norm": 1.7809889316558838, "learning_rate": 0.00017318221555737422, "loss": 0.2152, "step": 1375 }, { "epoch": 1.2053298383573612, "grad_norm": 2.3215832710266113, "learning_rate": 0.0001729946268754324, "loss": 0.3185, "step": 1380 }, { "epoch": 1.2096985583224116, "grad_norm": 1.347947120666504, "learning_rate": 0.00017280648675394947, "loss": 0.3085, "step": 1385 }, { "epoch": 1.2140672782874617, "grad_norm": 1.3840276002883911, "learning_rate": 0.00017261779661423407, "loss": 0.2016, "step": 1390 }, { "epoch": 1.218435998252512, "grad_norm": 1.2935765981674194, "learning_rate": 0.00017242855788175015, "loss": 0.2063, "step": 1395 }, { "epoch": 1.2228047182175623, "grad_norm": 1.2883801460266113, "learning_rate": 0.00017223877198610591, "loss": 0.2181, "step": 1400 }, { "epoch": 1.2271734381826125, "grad_norm": 1.3021811246871948, "learning_rate": 0.00017204844036104318, "loss": 0.2283, "step": 1405 }, { "epoch": 1.2315421581476627, "grad_norm": 1.1819430589675903, "learning_rate": 0.00017185756444442648, "loss": 0.2652, "step": 1410 }, { "epoch": 1.235910878112713, "grad_norm": 2.0612573623657227, "learning_rate": 0.00017166614567823212, "loss": 0.2977, "step": 1415 }, { "epoch": 1.2402795980777632, "grad_norm": 3.0888679027557373, "learning_rate": 0.00017147418550853756, "loss": 0.2682, "step": 1420 }, { "epoch": 1.2446483180428134, "grad_norm": 2.311062812805176, "learning_rate": 0.0001712816853855101, "loss": 0.3329, "step": 1425 }, { "epoch": 1.2490170380078638, "grad_norm": 1.2064367532730103, "learning_rate": 0.00017108864676339627, "loss": 0.2065, "step": 1430 }, { "epoch": 1.253385757972914, "grad_norm": 1.4042255878448486, "learning_rate": 0.00017089507110051066, "loss": 0.1738, "step": 1435 }, { "epoch": 1.2577544779379641, "grad_norm": 2.3508129119873047, "learning_rate": 0.00017070095985922493, "loss": 0.403, "step": 1440 }, { "epoch": 1.2621231979030143, "grad_norm": 1.2386358976364136, "learning_rate": 0.0001705063145059568, "loss": 0.272, "step": 1445 }, { "epoch": 1.2664919178680647, "grad_norm": 0.806268036365509, "learning_rate": 0.00017031113651115893, "loss": 0.2549, "step": 1450 }, { "epoch": 1.2708606378331149, "grad_norm": 1.6991655826568604, "learning_rate": 0.00017011542734930786, "loss": 0.2331, "step": 1455 }, { "epoch": 1.2752293577981653, "grad_norm": 1.1343746185302734, "learning_rate": 0.00016991918849889283, "loss": 0.3112, "step": 1460 }, { "epoch": 1.2795980777632154, "grad_norm": 1.3381041288375854, "learning_rate": 0.00016972242144240463, "loss": 0.1974, "step": 1465 }, { "epoch": 1.2839667977282656, "grad_norm": 3.6427793502807617, "learning_rate": 0.00016952512766632439, "loss": 0.2315, "step": 1470 }, { "epoch": 1.2883355176933158, "grad_norm": 1.8139313459396362, "learning_rate": 0.0001693273086611123, "loss": 0.1771, "step": 1475 }, { "epoch": 1.2927042376583662, "grad_norm": 1.205609679222107, "learning_rate": 0.00016912896592119654, "loss": 0.2551, "step": 1480 }, { "epoch": 1.2970729576234163, "grad_norm": 1.355162262916565, "learning_rate": 0.00016893010094496172, "loss": 0.2452, "step": 1485 }, { "epoch": 1.3014416775884665, "grad_norm": 1.2561094760894775, "learning_rate": 0.00016873071523473777, "loss": 0.2163, "step": 1490 }, { "epoch": 1.305810397553517, "grad_norm": 1.3165076971054077, "learning_rate": 0.00016853081029678853, "loss": 0.3273, "step": 1495 }, { "epoch": 1.310179117518567, "grad_norm": 1.8802030086517334, "learning_rate": 0.00016833038764130028, "loss": 0.3797, "step": 1500 }, { "epoch": 1.3145478374836173, "grad_norm": 1.7062153816223145, "learning_rate": 0.0001681294487823704, "loss": 0.2989, "step": 1505 }, { "epoch": 1.3189165574486674, "grad_norm": 2.0729176998138428, "learning_rate": 0.00016792799523799613, "loss": 0.2587, "step": 1510 }, { "epoch": 1.3232852774137178, "grad_norm": 1.129841685295105, "learning_rate": 0.00016772602853006268, "loss": 0.2201, "step": 1515 }, { "epoch": 1.327653997378768, "grad_norm": 1.2515584230422974, "learning_rate": 0.00016752355018433206, "loss": 0.2397, "step": 1520 }, { "epoch": 1.3320227173438184, "grad_norm": 1.2597646713256836, "learning_rate": 0.0001673205617304315, "loss": 0.2157, "step": 1525 }, { "epoch": 1.3363914373088686, "grad_norm": 1.8813763856887817, "learning_rate": 0.0001671170647018418, "loss": 0.2765, "step": 1530 }, { "epoch": 1.3407601572739187, "grad_norm": 2.208132266998291, "learning_rate": 0.00016691306063588583, "loss": 0.2258, "step": 1535 }, { "epoch": 1.345128877238969, "grad_norm": 1.9504673480987549, "learning_rate": 0.00016670855107371683, "loss": 0.2779, "step": 1540 }, { "epoch": 1.3494975972040193, "grad_norm": 2.171309471130371, "learning_rate": 0.00016650353756030692, "loss": 0.3031, "step": 1545 }, { "epoch": 1.3538663171690695, "grad_norm": 2.3320510387420654, "learning_rate": 0.00016629802164443519, "loss": 0.3288, "step": 1550 }, { "epoch": 1.3582350371341196, "grad_norm": 1.4883947372436523, "learning_rate": 0.0001660920048786763, "loss": 0.2416, "step": 1555 }, { "epoch": 1.36260375709917, "grad_norm": 1.1198906898498535, "learning_rate": 0.00016588548881938845, "loss": 0.2337, "step": 1560 }, { "epoch": 1.3669724770642202, "grad_norm": 1.4867557287216187, "learning_rate": 0.0001656784750267019, "loss": 0.3154, "step": 1565 }, { "epoch": 1.3713411970292704, "grad_norm": 2.2435972690582275, "learning_rate": 0.0001654709650645069, "loss": 0.245, "step": 1570 }, { "epoch": 1.3757099169943205, "grad_norm": 2.2508065700531006, "learning_rate": 0.00016526296050044215, "loss": 0.3097, "step": 1575 }, { "epoch": 1.380078636959371, "grad_norm": 0.8681387901306152, "learning_rate": 0.00016505446290588277, "loss": 0.295, "step": 1580 }, { "epoch": 1.3844473569244211, "grad_norm": 1.143965244293213, "learning_rate": 0.00016484547385592848, "loss": 0.2534, "step": 1585 }, { "epoch": 1.3888160768894715, "grad_norm": 2.3972761631011963, "learning_rate": 0.00016463599492939177, "loss": 0.2527, "step": 1590 }, { "epoch": 1.3931847968545217, "grad_norm": 2.7718844413757324, "learning_rate": 0.00016442602770878586, "loss": 0.2697, "step": 1595 }, { "epoch": 1.3975535168195719, "grad_norm": 1.7743192911148071, "learning_rate": 0.00016421557378031279, "loss": 0.2784, "step": 1600 }, { "epoch": 1.401922236784622, "grad_norm": 1.879279613494873, "learning_rate": 0.0001640046347338515, "loss": 0.2828, "step": 1605 }, { "epoch": 1.4062909567496724, "grad_norm": 1.5005972385406494, "learning_rate": 0.00016379321216294574, "loss": 0.2161, "step": 1610 }, { "epoch": 1.4106596767147226, "grad_norm": 1.1289541721343994, "learning_rate": 0.00016358130766479202, "loss": 0.1995, "step": 1615 }, { "epoch": 1.4150283966797728, "grad_norm": 0.7074740529060364, "learning_rate": 0.0001633689228402276, "loss": 0.2569, "step": 1620 }, { "epoch": 1.4193971166448232, "grad_norm": 1.4253774881362915, "learning_rate": 0.00016315605929371842, "loss": 0.3133, "step": 1625 }, { "epoch": 1.4237658366098733, "grad_norm": 1.6986387968063354, "learning_rate": 0.0001629427186333469, "loss": 0.2671, "step": 1630 }, { "epoch": 1.4281345565749235, "grad_norm": 1.194734811782837, "learning_rate": 0.0001627289024707998, "loss": 0.2592, "step": 1635 }, { "epoch": 1.4325032765399737, "grad_norm": 1.6335885524749756, "learning_rate": 0.00016251461242135616, "loss": 0.2628, "step": 1640 }, { "epoch": 1.436871996505024, "grad_norm": 1.417205572128296, "learning_rate": 0.0001622998501038749, "loss": 0.2129, "step": 1645 }, { "epoch": 1.4412407164700742, "grad_norm": 1.0214331150054932, "learning_rate": 0.0001620846171407828, "loss": 0.2505, "step": 1650 }, { "epoch": 1.4456094364351246, "grad_norm": 0.9859942197799683, "learning_rate": 0.000161868915158062, "loss": 0.1947, "step": 1655 }, { "epoch": 1.4499781564001748, "grad_norm": 1.9299548864364624, "learning_rate": 0.00016165274578523807, "loss": 0.2716, "step": 1660 }, { "epoch": 1.454346876365225, "grad_norm": 1.4493621587753296, "learning_rate": 0.00016143611065536727, "loss": 0.2066, "step": 1665 }, { "epoch": 1.4587155963302751, "grad_norm": 1.4922335147857666, "learning_rate": 0.00016121901140502456, "loss": 0.291, "step": 1670 }, { "epoch": 1.4630843162953255, "grad_norm": 0.9710771441459656, "learning_rate": 0.00016100144967429113, "loss": 0.207, "step": 1675 }, { "epoch": 1.4674530362603757, "grad_norm": 2.2792580127716064, "learning_rate": 0.0001607834271067419, "loss": 0.1705, "step": 1680 }, { "epoch": 1.4718217562254259, "grad_norm": 2.3811166286468506, "learning_rate": 0.00016056494534943323, "loss": 0.2994, "step": 1685 }, { "epoch": 1.4761904761904763, "grad_norm": 1.6609042882919312, "learning_rate": 0.00016034600605289046, "loss": 0.2, "step": 1690 }, { "epoch": 1.4805591961555264, "grad_norm": 2.835557222366333, "learning_rate": 0.0001601266108710954, "loss": 0.3283, "step": 1695 }, { "epoch": 1.4849279161205766, "grad_norm": 1.5667732954025269, "learning_rate": 0.00015990676146147384, "loss": 0.2685, "step": 1700 }, { "epoch": 1.4892966360856268, "grad_norm": 1.7098650932312012, "learning_rate": 0.0001596864594848831, "loss": 0.2466, "step": 1705 }, { "epoch": 1.4936653560506772, "grad_norm": 1.5239036083221436, "learning_rate": 0.00015946570660559933, "loss": 0.2577, "step": 1710 }, { "epoch": 1.4980340760157274, "grad_norm": 1.0807280540466309, "learning_rate": 0.00015924450449130513, "loss": 0.2017, "step": 1715 }, { "epoch": 1.5024027959807777, "grad_norm": 1.5457370281219482, "learning_rate": 0.0001590228548130768, "loss": 0.2695, "step": 1720 }, { "epoch": 1.506771515945828, "grad_norm": 1.1175668239593506, "learning_rate": 0.00015880075924537185, "loss": 0.2727, "step": 1725 }, { "epoch": 1.511140235910878, "grad_norm": 0.9887433052062988, "learning_rate": 0.00015857821946601615, "loss": 0.2561, "step": 1730 }, { "epoch": 1.5155089558759283, "grad_norm": 1.205710768699646, "learning_rate": 0.00015835523715619144, "loss": 0.2441, "step": 1735 }, { "epoch": 1.5198776758409784, "grad_norm": 1.3492207527160645, "learning_rate": 0.00015813181400042262, "loss": 0.2832, "step": 1740 }, { "epoch": 1.5242463958060288, "grad_norm": 1.2649726867675781, "learning_rate": 0.00015790795168656486, "loss": 0.2393, "step": 1745 }, { "epoch": 1.5286151157710792, "grad_norm": 2.4311208724975586, "learning_rate": 0.00015768365190579103, "loss": 0.2831, "step": 1750 }, { "epoch": 1.5329838357361294, "grad_norm": 1.1982014179229736, "learning_rate": 0.00015745891635257885, "loss": 0.2893, "step": 1755 }, { "epoch": 1.5373525557011796, "grad_norm": 1.842958688735962, "learning_rate": 0.0001572337467246981, "loss": 0.2158, "step": 1760 }, { "epoch": 1.5417212756662297, "grad_norm": 1.9764633178710938, "learning_rate": 0.00015700814472319774, "loss": 0.3043, "step": 1765 }, { "epoch": 1.54608999563128, "grad_norm": 1.0637012720108032, "learning_rate": 0.00015678211205239314, "loss": 0.258, "step": 1770 }, { "epoch": 1.5504587155963303, "grad_norm": 1.0332783460617065, "learning_rate": 0.00015655565041985318, "loss": 0.2373, "step": 1775 }, { "epoch": 1.5548274355613805, "grad_norm": 1.965062141418457, "learning_rate": 0.00015632876153638732, "loss": 0.1725, "step": 1780 }, { "epoch": 1.5591961555264309, "grad_norm": 0.6934140920639038, "learning_rate": 0.00015610144711603272, "loss": 0.2201, "step": 1785 }, { "epoch": 1.563564875491481, "grad_norm": 1.2010079622268677, "learning_rate": 0.00015587370887604123, "loss": 0.1796, "step": 1790 }, { "epoch": 1.5679335954565312, "grad_norm": 1.765724539756775, "learning_rate": 0.00015564554853686645, "loss": 0.2457, "step": 1795 }, { "epoch": 1.5723023154215814, "grad_norm": 1.848937749862671, "learning_rate": 0.00015541696782215084, "loss": 0.2096, "step": 1800 }, { "epoch": 1.5766710353866316, "grad_norm": 1.9904989004135132, "learning_rate": 0.00015518796845871247, "loss": 0.2651, "step": 1805 }, { "epoch": 1.581039755351682, "grad_norm": 0.8786830902099609, "learning_rate": 0.00015495855217653216, "loss": 0.2256, "step": 1810 }, { "epoch": 1.5854084753167323, "grad_norm": 2.0665247440338135, "learning_rate": 0.00015472872070874033, "loss": 0.3413, "step": 1815 }, { "epoch": 1.5897771952817825, "grad_norm": 2.08548903465271, "learning_rate": 0.000154498475791604, "loss": 0.2753, "step": 1820 }, { "epoch": 1.5941459152468327, "grad_norm": 1.4281671047210693, "learning_rate": 0.00015426781916451346, "loss": 0.2527, "step": 1825 }, { "epoch": 1.5985146352118829, "grad_norm": 0.9327197074890137, "learning_rate": 0.00015403675256996942, "loss": 0.1906, "step": 1830 }, { "epoch": 1.602883355176933, "grad_norm": 2.447357177734375, "learning_rate": 0.00015380527775356962, "loss": 0.2377, "step": 1835 }, { "epoch": 1.6072520751419834, "grad_norm": 1.3971562385559082, "learning_rate": 0.00015357339646399578, "loss": 0.2307, "step": 1840 }, { "epoch": 1.6116207951070336, "grad_norm": 1.4695461988449097, "learning_rate": 0.00015334111045300022, "loss": 0.2273, "step": 1845 }, { "epoch": 1.615989515072084, "grad_norm": 2.3516275882720947, "learning_rate": 0.0001531084214753928, "loss": 0.3006, "step": 1850 }, { "epoch": 1.6203582350371342, "grad_norm": 1.7472774982452393, "learning_rate": 0.00015287533128902764, "loss": 0.2943, "step": 1855 }, { "epoch": 1.6247269550021843, "grad_norm": 2.3994486331939697, "learning_rate": 0.00015264184165478977, "loss": 0.2277, "step": 1860 }, { "epoch": 1.6290956749672345, "grad_norm": 2.234163522720337, "learning_rate": 0.00015240795433658187, "loss": 0.2319, "step": 1865 }, { "epoch": 1.6334643949322847, "grad_norm": 1.5536202192306519, "learning_rate": 0.00015217367110131086, "loss": 0.1685, "step": 1870 }, { "epoch": 1.637833114897335, "grad_norm": 1.564252257347107, "learning_rate": 0.0001519389937188747, "loss": 0.2817, "step": 1875 }, { "epoch": 1.6422018348623855, "grad_norm": 1.6615930795669556, "learning_rate": 0.00015170392396214897, "loss": 0.2409, "step": 1880 }, { "epoch": 1.6465705548274356, "grad_norm": 2.7048113346099854, "learning_rate": 0.00015146846360697332, "loss": 0.3594, "step": 1885 }, { "epoch": 1.6509392747924858, "grad_norm": 1.7237365245819092, "learning_rate": 0.00015123261443213837, "loss": 0.2692, "step": 1890 }, { "epoch": 1.655307994757536, "grad_norm": 0.9304201602935791, "learning_rate": 0.00015099637821937192, "loss": 0.2457, "step": 1895 }, { "epoch": 1.6596767147225862, "grad_norm": 1.8771346807479858, "learning_rate": 0.00015075975675332573, "loss": 0.2152, "step": 1900 }, { "epoch": 1.6640454346876365, "grad_norm": 1.0985363721847534, "learning_rate": 0.00015052275182156198, "loss": 0.2191, "step": 1905 }, { "epoch": 1.6684141546526867, "grad_norm": 1.3073694705963135, "learning_rate": 0.00015028536521453968, "loss": 0.3686, "step": 1910 }, { "epoch": 1.6727828746177371, "grad_norm": 0.7887938618659973, "learning_rate": 0.0001500475987256013, "loss": 0.2015, "step": 1915 }, { "epoch": 1.6771515945827873, "grad_norm": 1.5100387334823608, "learning_rate": 0.0001498094541509591, "loss": 0.2545, "step": 1920 }, { "epoch": 1.6815203145478375, "grad_norm": 1.5141887664794922, "learning_rate": 0.00014957093328968156, "loss": 0.2385, "step": 1925 }, { "epoch": 1.6858890345128876, "grad_norm": 1.1331568956375122, "learning_rate": 0.00014933203794367992, "loss": 0.2578, "step": 1930 }, { "epoch": 1.6902577544779378, "grad_norm": 1.6967277526855469, "learning_rate": 0.00014909276991769435, "loss": 0.3004, "step": 1935 }, { "epoch": 1.6946264744429882, "grad_norm": 1.1680165529251099, "learning_rate": 0.00014885313101928055, "loss": 0.2306, "step": 1940 }, { "epoch": 1.6989951944080386, "grad_norm": 0.8564541339874268, "learning_rate": 0.00014861312305879592, "loss": 0.2384, "step": 1945 }, { "epoch": 1.7033639143730888, "grad_norm": 0.8954014778137207, "learning_rate": 0.00014837274784938596, "loss": 0.1804, "step": 1950 }, { "epoch": 1.707732634338139, "grad_norm": 1.9697721004486084, "learning_rate": 0.00014813200720697055, "loss": 0.2337, "step": 1955 }, { "epoch": 1.712101354303189, "grad_norm": 1.3099297285079956, "learning_rate": 0.00014789090295023031, "loss": 0.2387, "step": 1960 }, { "epoch": 1.7164700742682393, "grad_norm": 1.8576496839523315, "learning_rate": 0.00014764943690059269, "loss": 0.2739, "step": 1965 }, { "epoch": 1.7208387942332897, "grad_norm": 1.8436728715896606, "learning_rate": 0.0001474076108822184, "loss": 0.1997, "step": 1970 }, { "epoch": 1.7252075141983398, "grad_norm": 1.0517886877059937, "learning_rate": 0.0001471654267219875, "loss": 0.2432, "step": 1975 }, { "epoch": 1.7295762341633902, "grad_norm": 0.9853880405426025, "learning_rate": 0.00014692288624948557, "loss": 0.3059, "step": 1980 }, { "epoch": 1.7339449541284404, "grad_norm": 2.1775450706481934, "learning_rate": 0.00014667999129699011, "loss": 0.3374, "step": 1985 }, { "epoch": 1.7383136740934906, "grad_norm": 0.7085615396499634, "learning_rate": 0.0001464367436994565, "loss": 0.1907, "step": 1990 }, { "epoch": 1.7426823940585408, "grad_norm": 1.3183834552764893, "learning_rate": 0.00014619314529450405, "loss": 0.2587, "step": 1995 }, { "epoch": 1.747051114023591, "grad_norm": 2.9087324142456055, "learning_rate": 0.00014594919792240246, "loss": 0.301, "step": 2000 }, { "epoch": 1.7514198339886413, "grad_norm": 1.5489122867584229, "learning_rate": 0.00014570490342605751, "loss": 0.2061, "step": 2005 }, { "epoch": 1.7557885539536917, "grad_norm": 1.4719579219818115, "learning_rate": 0.00014546026365099753, "loss": 0.2334, "step": 2010 }, { "epoch": 1.7601572739187419, "grad_norm": 1.499053955078125, "learning_rate": 0.0001452152804453591, "loss": 0.2817, "step": 2015 }, { "epoch": 1.764525993883792, "grad_norm": 1.0219850540161133, "learning_rate": 0.00014496995565987337, "loss": 0.2292, "step": 2020 }, { "epoch": 1.7688947138488422, "grad_norm": 1.4518020153045654, "learning_rate": 0.00014472429114785194, "loss": 0.2321, "step": 2025 }, { "epoch": 1.7732634338138924, "grad_norm": 1.0398428440093994, "learning_rate": 0.00014447828876517277, "loss": 0.2649, "step": 2030 }, { "epoch": 1.7776321537789428, "grad_norm": 0.5719377994537354, "learning_rate": 0.00014423195037026646, "loss": 0.2239, "step": 2035 }, { "epoch": 1.782000873743993, "grad_norm": 1.4201310873031616, "learning_rate": 0.00014398527782410187, "loss": 0.1812, "step": 2040 }, { "epoch": 1.7863695937090434, "grad_norm": 1.380936622619629, "learning_rate": 0.00014373827299017227, "loss": 0.2379, "step": 2045 }, { "epoch": 1.7907383136740935, "grad_norm": 0.9390376210212708, "learning_rate": 0.0001434909377344812, "loss": 0.1879, "step": 2050 }, { "epoch": 1.7951070336391437, "grad_norm": 1.697022557258606, "learning_rate": 0.0001432432739255284, "loss": 0.2644, "step": 2055 }, { "epoch": 1.7994757536041939, "grad_norm": 1.7329260110855103, "learning_rate": 0.00014299528343429566, "loss": 0.4845, "step": 2060 }, { "epoch": 1.8038444735692443, "grad_norm": 1.098286747932434, "learning_rate": 0.00014274696813423269, "loss": 0.2452, "step": 2065 }, { "epoch": 1.8082131935342944, "grad_norm": 0.6370453238487244, "learning_rate": 0.00014249832990124292, "loss": 0.2423, "step": 2070 }, { "epoch": 1.8125819134993448, "grad_norm": 1.1073155403137207, "learning_rate": 0.0001422493706136695, "loss": 0.269, "step": 2075 }, { "epoch": 1.816950633464395, "grad_norm": 1.5754518508911133, "learning_rate": 0.0001420000921522809, "loss": 0.2506, "step": 2080 }, { "epoch": 1.8213193534294452, "grad_norm": 2.118640422821045, "learning_rate": 0.0001417504964002569, "loss": 0.2072, "step": 2085 }, { "epoch": 1.8256880733944953, "grad_norm": 1.0548226833343506, "learning_rate": 0.0001415005852431741, "loss": 0.2509, "step": 2090 }, { "epoch": 1.8300567933595455, "grad_norm": 2.12402606010437, "learning_rate": 0.00014125036056899197, "loss": 0.2599, "step": 2095 }, { "epoch": 1.834425513324596, "grad_norm": 1.251313328742981, "learning_rate": 0.00014099982426803842, "loss": 0.2302, "step": 2100 }, { "epoch": 1.838794233289646, "grad_norm": 1.9044132232666016, "learning_rate": 0.0001407489782329955, "loss": 0.2487, "step": 2105 }, { "epoch": 1.8431629532546965, "grad_norm": 0.7197313904762268, "learning_rate": 0.00014049782435888525, "loss": 0.1864, "step": 2110 }, { "epoch": 1.8475316732197467, "grad_norm": 1.1280083656311035, "learning_rate": 0.00014024636454305515, "loss": 0.2028, "step": 2115 }, { "epoch": 1.8519003931847968, "grad_norm": 2.0197250843048096, "learning_rate": 0.00013999460068516407, "loss": 0.3204, "step": 2120 }, { "epoch": 1.856269113149847, "grad_norm": 1.2510887384414673, "learning_rate": 0.0001397425346871677, "loss": 0.213, "step": 2125 }, { "epoch": 1.8606378331148974, "grad_norm": 1.8764947652816772, "learning_rate": 0.0001394901684533042, "loss": 0.2815, "step": 2130 }, { "epoch": 1.8650065530799476, "grad_norm": 2.152601480484009, "learning_rate": 0.00013923750389007998, "loss": 0.2884, "step": 2135 }, { "epoch": 1.869375273044998, "grad_norm": 1.9770750999450684, "learning_rate": 0.00013898454290625515, "loss": 0.3571, "step": 2140 }, { "epoch": 1.8737439930100481, "grad_norm": 1.4003419876098633, "learning_rate": 0.00013873128741282906, "loss": 0.247, "step": 2145 }, { "epoch": 1.8781127129750983, "grad_norm": 1.565500020980835, "learning_rate": 0.00013847773932302603, "loss": 0.2227, "step": 2150 }, { "epoch": 1.8824814329401485, "grad_norm": 0.588405966758728, "learning_rate": 0.00013822390055228079, "loss": 0.2331, "step": 2155 }, { "epoch": 1.8868501529051986, "grad_norm": 0.9293955564498901, "learning_rate": 0.00013796977301822397, "loss": 0.258, "step": 2160 }, { "epoch": 1.891218872870249, "grad_norm": 3.2338411808013916, "learning_rate": 0.00013771535864066773, "loss": 0.4014, "step": 2165 }, { "epoch": 1.8955875928352992, "grad_norm": 1.0802291631698608, "learning_rate": 0.00013746065934159123, "loss": 0.1996, "step": 2170 }, { "epoch": 1.8999563128003496, "grad_norm": 2.0415797233581543, "learning_rate": 0.00013720567704512593, "loss": 0.2392, "step": 2175 }, { "epoch": 1.9043250327653998, "grad_norm": 1.554219365119934, "learning_rate": 0.00013695041367754133, "loss": 0.2165, "step": 2180 }, { "epoch": 1.90869375273045, "grad_norm": 8.951557159423828, "learning_rate": 0.00013669487116723024, "loss": 0.2473, "step": 2185 }, { "epoch": 1.9130624726955001, "grad_norm": 2.31048321723938, "learning_rate": 0.0001364390514446943, "loss": 0.2333, "step": 2190 }, { "epoch": 1.9174311926605505, "grad_norm": 1.3068311214447021, "learning_rate": 0.0001361829564425293, "loss": 0.2034, "step": 2195 }, { "epoch": 1.9217999126256007, "grad_norm": 0.9649588465690613, "learning_rate": 0.00013592658809541064, "loss": 0.2276, "step": 2200 }, { "epoch": 1.926168632590651, "grad_norm": 2.0399038791656494, "learning_rate": 0.00013566994834007877, "loss": 0.1955, "step": 2205 }, { "epoch": 1.9305373525557012, "grad_norm": 1.4719631671905518, "learning_rate": 0.00013541303911532445, "loss": 0.1883, "step": 2210 }, { "epoch": 1.9349060725207514, "grad_norm": 1.938029170036316, "learning_rate": 0.00013515586236197418, "loss": 0.2386, "step": 2215 }, { "epoch": 1.9392747924858016, "grad_norm": 1.9169769287109375, "learning_rate": 0.00013489842002287542, "loss": 0.2435, "step": 2220 }, { "epoch": 1.9436435124508518, "grad_norm": 1.1561371088027954, "learning_rate": 0.0001346407140428822, "loss": 0.2305, "step": 2225 }, { "epoch": 1.9480122324159022, "grad_norm": 1.9873589277267456, "learning_rate": 0.00013438274636884, "loss": 0.2706, "step": 2230 }, { "epoch": 1.9523809523809523, "grad_norm": 1.9053728580474854, "learning_rate": 0.00013412451894957144, "loss": 0.187, "step": 2235 }, { "epoch": 1.9567496723460027, "grad_norm": 1.3117566108703613, "learning_rate": 0.00013386603373586134, "loss": 0.2317, "step": 2240 }, { "epoch": 1.961118392311053, "grad_norm": 0.7097095251083374, "learning_rate": 0.000133607292680442, "loss": 0.2926, "step": 2245 }, { "epoch": 1.965487112276103, "grad_norm": 1.8591192960739136, "learning_rate": 0.0001333482977379785, "loss": 0.2025, "step": 2250 }, { "epoch": 1.9698558322411532, "grad_norm": 2.633700370788574, "learning_rate": 0.00013308905086505395, "loss": 0.2513, "step": 2255 }, { "epoch": 1.9742245522062036, "grad_norm": 1.4099256992340088, "learning_rate": 0.0001328295540201546, "loss": 0.2447, "step": 2260 }, { "epoch": 1.9785932721712538, "grad_norm": 1.20600163936615, "learning_rate": 0.00013256980916365527, "loss": 0.2449, "step": 2265 }, { "epoch": 1.9829619921363042, "grad_norm": 1.1572972536087036, "learning_rate": 0.0001323098182578042, "loss": 0.1937, "step": 2270 }, { "epoch": 1.9873307121013544, "grad_norm": 1.0929369926452637, "learning_rate": 0.00013204958326670853, "loss": 0.2273, "step": 2275 }, { "epoch": 1.9916994320664045, "grad_norm": 1.674107313156128, "learning_rate": 0.00013178910615631933, "loss": 0.3191, "step": 2280 }, { "epoch": 1.9960681520314547, "grad_norm": 1.306754469871521, "learning_rate": 0.00013152838889441673, "loss": 0.2723, "step": 2285 }, { "epoch": 2.0, "grad_norm": 1.877669334411621, "learning_rate": 0.00013126743345059512, "loss": 0.2246, "step": 2290 }, { "epoch": 2.00436871996505, "grad_norm": 0.8999947905540466, "learning_rate": 0.00013100624179624828, "loss": 0.1528, "step": 2295 }, { "epoch": 2.0087374399301003, "grad_norm": 1.1233611106872559, "learning_rate": 0.00013074481590455433, "loss": 0.215, "step": 2300 }, { "epoch": 2.0131061598951505, "grad_norm": 1.4673850536346436, "learning_rate": 0.00013048315775046108, "loss": 0.1379, "step": 2305 }, { "epoch": 2.017474879860201, "grad_norm": 1.4777660369873047, "learning_rate": 0.0001302212693106709, "loss": 0.1343, "step": 2310 }, { "epoch": 2.0218435998252513, "grad_norm": 1.7447019815444946, "learning_rate": 0.00012995915256362584, "loss": 0.1591, "step": 2315 }, { "epoch": 2.0262123197903015, "grad_norm": 0.8237628936767578, "learning_rate": 0.00012969680948949272, "loss": 0.1182, "step": 2320 }, { "epoch": 2.0305810397553516, "grad_norm": 1.5945308208465576, "learning_rate": 0.00012943424207014818, "loss": 0.1624, "step": 2325 }, { "epoch": 2.034949759720402, "grad_norm": 1.3121789693832397, "learning_rate": 0.00012917145228916367, "loss": 0.1313, "step": 2330 }, { "epoch": 2.039318479685452, "grad_norm": 1.028647780418396, "learning_rate": 0.00012890844213179044, "loss": 0.214, "step": 2335 }, { "epoch": 2.0436871996505026, "grad_norm": 1.0864648818969727, "learning_rate": 0.00012864521358494464, "loss": 0.1929, "step": 2340 }, { "epoch": 2.0480559196155528, "grad_norm": 1.2443722486495972, "learning_rate": 0.00012838176863719217, "loss": 0.1479, "step": 2345 }, { "epoch": 2.052424639580603, "grad_norm": 3.708357095718384, "learning_rate": 0.00012811810927873386, "loss": 0.1897, "step": 2350 }, { "epoch": 2.056793359545653, "grad_norm": 1.483973741531372, "learning_rate": 0.00012785423750139008, "loss": 0.1188, "step": 2355 }, { "epoch": 2.0611620795107033, "grad_norm": 1.0019135475158691, "learning_rate": 0.00012759015529858624, "loss": 0.1683, "step": 2360 }, { "epoch": 2.0655307994757535, "grad_norm": 1.5917549133300781, "learning_rate": 0.00012732586466533715, "loss": 0.2293, "step": 2365 }, { "epoch": 2.0698995194408036, "grad_norm": 1.4390945434570312, "learning_rate": 0.00012706136759823233, "loss": 0.1873, "step": 2370 }, { "epoch": 2.0742682394058543, "grad_norm": 2.951580762863159, "learning_rate": 0.00012679666609542083, "loss": 0.1984, "step": 2375 }, { "epoch": 2.0786369593709044, "grad_norm": 2.2444632053375244, "learning_rate": 0.00012653176215659596, "loss": 0.1709, "step": 2380 }, { "epoch": 2.0830056793359546, "grad_norm": 1.191437840461731, "learning_rate": 0.0001262666577829806, "loss": 0.1926, "step": 2385 }, { "epoch": 2.0873743993010048, "grad_norm": 2.2964091300964355, "learning_rate": 0.00012600135497731156, "loss": 0.1583, "step": 2390 }, { "epoch": 2.091743119266055, "grad_norm": 0.9333243370056152, "learning_rate": 0.00012573585574382487, "loss": 0.1261, "step": 2395 }, { "epoch": 2.096111839231105, "grad_norm": 0.8678691387176514, "learning_rate": 0.00012547016208824038, "loss": 0.1491, "step": 2400 }, { "epoch": 2.1004805591961557, "grad_norm": 1.7139450311660767, "learning_rate": 0.00012520427601774682, "loss": 0.1599, "step": 2405 }, { "epoch": 2.104849279161206, "grad_norm": 1.161952257156372, "learning_rate": 0.0001249381995409864, "loss": 0.1806, "step": 2410 }, { "epoch": 2.109217999126256, "grad_norm": 2.1318793296813965, "learning_rate": 0.00012467193466803982, "loss": 0.273, "step": 2415 }, { "epoch": 2.1135867190913062, "grad_norm": 1.7493711709976196, "learning_rate": 0.00012440548341041108, "loss": 0.192, "step": 2420 }, { "epoch": 2.1179554390563564, "grad_norm": 1.2488105297088623, "learning_rate": 0.00012413884778101207, "loss": 0.1602, "step": 2425 }, { "epoch": 2.1223241590214066, "grad_norm": 1.1086466312408447, "learning_rate": 0.00012387202979414767, "loss": 0.1318, "step": 2430 }, { "epoch": 2.126692878986457, "grad_norm": 1.6508293151855469, "learning_rate": 0.00012360503146550034, "loss": 0.1595, "step": 2435 }, { "epoch": 2.1310615989515074, "grad_norm": 0.8604519367218018, "learning_rate": 0.00012333785481211487, "loss": 0.1409, "step": 2440 }, { "epoch": 2.1354303189165575, "grad_norm": 1.6457815170288086, "learning_rate": 0.00012307050185238333, "loss": 0.1807, "step": 2445 }, { "epoch": 2.1397990388816077, "grad_norm": 1.9465317726135254, "learning_rate": 0.00012280297460602957, "loss": 0.1804, "step": 2450 }, { "epoch": 2.144167758846658, "grad_norm": 1.763515591621399, "learning_rate": 0.00012253527509409418, "loss": 0.152, "step": 2455 }, { "epoch": 2.148536478811708, "grad_norm": 2.8693459033966064, "learning_rate": 0.00012226740533891913, "loss": 0.1469, "step": 2460 }, { "epoch": 2.1529051987767582, "grad_norm": 1.7148571014404297, "learning_rate": 0.00012199936736413246, "loss": 0.1892, "step": 2465 }, { "epoch": 2.157273918741809, "grad_norm": 1.3995704650878906, "learning_rate": 0.00012173116319463306, "loss": 0.2003, "step": 2470 }, { "epoch": 2.161642638706859, "grad_norm": 1.0834511518478394, "learning_rate": 0.00012146279485657532, "loss": 0.1784, "step": 2475 }, { "epoch": 2.166011358671909, "grad_norm": 1.1982789039611816, "learning_rate": 0.00012119426437735384, "loss": 0.1722, "step": 2480 }, { "epoch": 2.1703800786369594, "grad_norm": 2.481656551361084, "learning_rate": 0.0001209255737855881, "loss": 0.1938, "step": 2485 }, { "epoch": 2.1747487986020095, "grad_norm": 0.7617926001548767, "learning_rate": 0.00012065672511110728, "loss": 0.1533, "step": 2490 }, { "epoch": 2.1791175185670597, "grad_norm": 0.9713643789291382, "learning_rate": 0.0001203877203849346, "loss": 0.1241, "step": 2495 }, { "epoch": 2.18348623853211, "grad_norm": 1.348576545715332, "learning_rate": 0.00012011856163927235, "loss": 0.1882, "step": 2500 }, { "epoch": 2.1878549584971605, "grad_norm": 3.1400299072265625, "learning_rate": 0.00011984925090748626, "loss": 0.2369, "step": 2505 }, { "epoch": 2.1922236784622107, "grad_norm": 1.1611807346343994, "learning_rate": 0.00011957979022409027, "loss": 0.1614, "step": 2510 }, { "epoch": 2.196592398427261, "grad_norm": 1.4003978967666626, "learning_rate": 0.00011931018162473117, "loss": 0.1431, "step": 2515 }, { "epoch": 2.200961118392311, "grad_norm": 1.7134933471679688, "learning_rate": 0.00011904042714617311, "loss": 0.1917, "step": 2520 }, { "epoch": 2.205329838357361, "grad_norm": 1.4700738191604614, "learning_rate": 0.00011877052882628237, "loss": 0.1506, "step": 2525 }, { "epoch": 2.2096985583224114, "grad_norm": 0.9795822501182556, "learning_rate": 0.00011850048870401185, "loss": 0.1663, "step": 2530 }, { "epoch": 2.214067278287462, "grad_norm": 1.8647089004516602, "learning_rate": 0.00011823030881938564, "loss": 0.1984, "step": 2535 }, { "epoch": 2.218435998252512, "grad_norm": 1.2616095542907715, "learning_rate": 0.00011795999121348378, "loss": 0.1158, "step": 2540 }, { "epoch": 2.2228047182175623, "grad_norm": 1.105303168296814, "learning_rate": 0.00011768953792842663, "loss": 0.1932, "step": 2545 }, { "epoch": 2.2271734381826125, "grad_norm": 1.0846420526504517, "learning_rate": 0.00011741895100735958, "loss": 0.156, "step": 2550 }, { "epoch": 2.2315421581476627, "grad_norm": 2.285141944885254, "learning_rate": 0.00011714823249443763, "loss": 0.1639, "step": 2555 }, { "epoch": 2.235910878112713, "grad_norm": 1.022466778755188, "learning_rate": 0.00011687738443480975, "loss": 0.1591, "step": 2560 }, { "epoch": 2.2402795980777634, "grad_norm": 2.86263370513916, "learning_rate": 0.00011660640887460377, "loss": 0.24, "step": 2565 }, { "epoch": 2.2446483180428136, "grad_norm": 1.2612568140029907, "learning_rate": 0.00011633530786091051, "loss": 0.1625, "step": 2570 }, { "epoch": 2.249017038007864, "grad_norm": 1.171590805053711, "learning_rate": 0.00011606408344176873, "loss": 0.1702, "step": 2575 }, { "epoch": 2.253385757972914, "grad_norm": 1.118406891822815, "learning_rate": 0.0001157927376661493, "loss": 0.1518, "step": 2580 }, { "epoch": 2.257754477937964, "grad_norm": 0.8476281762123108, "learning_rate": 0.00011552127258394003, "loss": 0.2093, "step": 2585 }, { "epoch": 2.2621231979030143, "grad_norm": 1.7625765800476074, "learning_rate": 0.0001152496902459299, "loss": 0.1764, "step": 2590 }, { "epoch": 2.2664919178680645, "grad_norm": 1.5414633750915527, "learning_rate": 0.00011497799270379374, "loss": 0.1185, "step": 2595 }, { "epoch": 2.270860637833115, "grad_norm": 2.2619729042053223, "learning_rate": 0.00011470618201007677, "loss": 0.1554, "step": 2600 }, { "epoch": 2.2752293577981653, "grad_norm": 0.9168559312820435, "learning_rate": 0.00011443426021817887, "loss": 0.1241, "step": 2605 }, { "epoch": 2.2795980777632154, "grad_norm": 1.3820409774780273, "learning_rate": 0.00011416222938233936, "loss": 0.1387, "step": 2610 }, { "epoch": 2.2839667977282656, "grad_norm": 2.417088270187378, "learning_rate": 0.00011389009155762128, "loss": 0.267, "step": 2615 }, { "epoch": 2.288335517693316, "grad_norm": 1.2400354146957397, "learning_rate": 0.00011361784879989588, "loss": 0.1586, "step": 2620 }, { "epoch": 2.292704237658366, "grad_norm": 1.5232576131820679, "learning_rate": 0.00011334550316582717, "loss": 0.2155, "step": 2625 }, { "epoch": 2.297072957623416, "grad_norm": 1.0652902126312256, "learning_rate": 0.00011307305671285633, "loss": 0.1849, "step": 2630 }, { "epoch": 2.3014416775884667, "grad_norm": 1.0222703218460083, "learning_rate": 0.00011280051149918622, "loss": 0.1351, "step": 2635 }, { "epoch": 2.305810397553517, "grad_norm": 2.64607310295105, "learning_rate": 0.00011252786958376569, "loss": 0.1929, "step": 2640 }, { "epoch": 2.310179117518567, "grad_norm": 1.090824842453003, "learning_rate": 0.00011225513302627422, "loss": 0.1215, "step": 2645 }, { "epoch": 2.3145478374836173, "grad_norm": 1.6671061515808105, "learning_rate": 0.00011198230388710625, "loss": 0.1279, "step": 2650 }, { "epoch": 2.3189165574486674, "grad_norm": 1.3377041816711426, "learning_rate": 0.00011170938422735558, "loss": 0.1527, "step": 2655 }, { "epoch": 2.3232852774137176, "grad_norm": 1.554307460784912, "learning_rate": 0.00011143637610879989, "loss": 0.16, "step": 2660 }, { "epoch": 2.327653997378768, "grad_norm": 1.2246366739273071, "learning_rate": 0.00011116328159388514, "loss": 0.1791, "step": 2665 }, { "epoch": 2.3320227173438184, "grad_norm": 3.7316997051239014, "learning_rate": 0.00011089010274570992, "loss": 0.2552, "step": 2670 }, { "epoch": 2.3363914373088686, "grad_norm": 2.306547164916992, "learning_rate": 0.00011061684162801, "loss": 0.2081, "step": 2675 }, { "epoch": 2.3407601572739187, "grad_norm": 1.9740854501724243, "learning_rate": 0.00011034350030514253, "loss": 0.147, "step": 2680 }, { "epoch": 2.345128877238969, "grad_norm": 1.5787196159362793, "learning_rate": 0.00011007008084207072, "loss": 0.152, "step": 2685 }, { "epoch": 2.349497597204019, "grad_norm": 1.394114375114441, "learning_rate": 0.00010979658530434793, "loss": 0.2236, "step": 2690 }, { "epoch": 2.3538663171690697, "grad_norm": 1.1155906915664673, "learning_rate": 0.0001095230157581024, "loss": 0.1224, "step": 2695 }, { "epoch": 2.35823503713412, "grad_norm": 1.6753724813461304, "learning_rate": 0.00010924937427002136, "loss": 0.1937, "step": 2700 }, { "epoch": 2.36260375709917, "grad_norm": 1.3327598571777344, "learning_rate": 0.00010897566290733552, "loss": 0.2221, "step": 2705 }, { "epoch": 2.36697247706422, "grad_norm": 2.0531933307647705, "learning_rate": 0.00010870188373780352, "loss": 0.146, "step": 2710 }, { "epoch": 2.3713411970292704, "grad_norm": 1.6777331829071045, "learning_rate": 0.00010842803882969623, "loss": 0.1931, "step": 2715 }, { "epoch": 2.3757099169943205, "grad_norm": 2.2374494075775146, "learning_rate": 0.00010815413025178112, "loss": 0.1637, "step": 2720 }, { "epoch": 2.3800786369593707, "grad_norm": 2.244675636291504, "learning_rate": 0.00010788016007330665, "loss": 0.1644, "step": 2725 }, { "epoch": 2.3844473569244213, "grad_norm": 0.7124090194702148, "learning_rate": 0.00010760613036398668, "loss": 0.1246, "step": 2730 }, { "epoch": 2.3888160768894715, "grad_norm": 1.298221230506897, "learning_rate": 0.00010733204319398477, "loss": 0.1604, "step": 2735 }, { "epoch": 2.3931847968545217, "grad_norm": 3.08442759513855, "learning_rate": 0.00010705790063389858, "loss": 0.2225, "step": 2740 }, { "epoch": 2.397553516819572, "grad_norm": 1.3459490537643433, "learning_rate": 0.00010678370475474424, "loss": 0.2159, "step": 2745 }, { "epoch": 2.401922236784622, "grad_norm": 1.9191830158233643, "learning_rate": 0.00010650945762794058, "loss": 0.2479, "step": 2750 }, { "epoch": 2.406290956749672, "grad_norm": 1.102486491203308, "learning_rate": 0.00010623516132529372, "loss": 0.1371, "step": 2755 }, { "epoch": 2.4106596767147224, "grad_norm": 1.311330795288086, "learning_rate": 0.00010596081791898118, "loss": 0.1672, "step": 2760 }, { "epoch": 2.415028396679773, "grad_norm": 2.336740732192993, "learning_rate": 0.00010568642948153636, "loss": 0.2096, "step": 2765 }, { "epoch": 2.419397116644823, "grad_norm": 0.9913440942764282, "learning_rate": 0.00010541199808583286, "loss": 0.2231, "step": 2770 }, { "epoch": 2.4237658366098733, "grad_norm": 1.6791926622390747, "learning_rate": 0.00010513752580506878, "loss": 0.1871, "step": 2775 }, { "epoch": 2.4281345565749235, "grad_norm": 1.459643840789795, "learning_rate": 0.00010486301471275111, "loss": 0.2017, "step": 2780 }, { "epoch": 2.4325032765399737, "grad_norm": 2.661412239074707, "learning_rate": 0.00010458846688268003, "loss": 0.2038, "step": 2785 }, { "epoch": 2.436871996505024, "grad_norm": 2.1982479095458984, "learning_rate": 0.00010431388438893326, "loss": 0.1776, "step": 2790 }, { "epoch": 2.4412407164700745, "grad_norm": 0.9297656416893005, "learning_rate": 0.00010403926930585042, "loss": 0.124, "step": 2795 }, { "epoch": 2.4456094364351246, "grad_norm": 0.9144073128700256, "learning_rate": 0.00010376462370801724, "loss": 0.1728, "step": 2800 }, { "epoch": 2.449978156400175, "grad_norm": 0.8969748616218567, "learning_rate": 0.00010348994967025012, "loss": 0.1268, "step": 2805 }, { "epoch": 2.454346876365225, "grad_norm": 1.952581763267517, "learning_rate": 0.00010321524926758012, "loss": 0.1516, "step": 2810 }, { "epoch": 2.458715596330275, "grad_norm": 1.463630199432373, "learning_rate": 0.00010294052457523766, "loss": 0.1748, "step": 2815 }, { "epoch": 2.4630843162953253, "grad_norm": 0.9995607137680054, "learning_rate": 0.0001026657776686365, "loss": 0.1815, "step": 2820 }, { "epoch": 2.467453036260376, "grad_norm": 1.2662900686264038, "learning_rate": 0.00010239101062335834, "loss": 0.1727, "step": 2825 }, { "epoch": 2.471821756225426, "grad_norm": 1.2032655477523804, "learning_rate": 0.00010211622551513697, "loss": 0.1398, "step": 2830 }, { "epoch": 2.4761904761904763, "grad_norm": 1.363953948020935, "learning_rate": 0.00010184142441984259, "loss": 0.1405, "step": 2835 }, { "epoch": 2.4805591961555264, "grad_norm": 1.4486302137374878, "learning_rate": 0.00010156660941346627, "loss": 0.1877, "step": 2840 }, { "epoch": 2.4849279161205766, "grad_norm": 1.4691563844680786, "learning_rate": 0.00010129178257210413, "loss": 0.1261, "step": 2845 }, { "epoch": 2.489296636085627, "grad_norm": 1.7420620918273926, "learning_rate": 0.0001010169459719416, "loss": 0.1672, "step": 2850 }, { "epoch": 2.493665356050677, "grad_norm": 1.8223110437393188, "learning_rate": 0.00010074210168923804, "loss": 0.1476, "step": 2855 }, { "epoch": 2.4980340760157276, "grad_norm": 1.197109580039978, "learning_rate": 0.00010046725180031062, "loss": 0.1227, "step": 2860 }, { "epoch": 2.5024027959807777, "grad_norm": 1.7598116397857666, "learning_rate": 0.00010019239838151906, "loss": 0.1882, "step": 2865 }, { "epoch": 2.506771515945828, "grad_norm": 1.498608946800232, "learning_rate": 9.99175435092496e-05, "loss": 0.248, "step": 2870 }, { "epoch": 2.511140235910878, "grad_norm": 1.8565483093261719, "learning_rate": 9.964268925989954e-05, "loss": 0.1935, "step": 2875 }, { "epoch": 2.5155089558759283, "grad_norm": 0.8620087504386902, "learning_rate": 9.936783770986145e-05, "loss": 0.126, "step": 2880 }, { "epoch": 2.5198776758409784, "grad_norm": 0.7721559405326843, "learning_rate": 9.909299093550757e-05, "loss": 0.1066, "step": 2885 }, { "epoch": 2.5242463958060286, "grad_norm": 1.5533393621444702, "learning_rate": 9.88181510131739e-05, "loss": 0.2407, "step": 2890 }, { "epoch": 2.5286151157710792, "grad_norm": 1.1294517517089844, "learning_rate": 9.854332001914486e-05, "loss": 0.1601, "step": 2895 }, { "epoch": 2.5329838357361294, "grad_norm": 1.7611277103424072, "learning_rate": 9.826850002963729e-05, "loss": 0.192, "step": 2900 }, { "epoch": 2.5373525557011796, "grad_norm": 1.9960044622421265, "learning_rate": 9.799369312078502e-05, "loss": 0.1327, "step": 2905 }, { "epoch": 2.5417212756662297, "grad_norm": 0.9330713748931885, "learning_rate": 9.771890136862288e-05, "loss": 0.1479, "step": 2910 }, { "epoch": 2.54608999563128, "grad_norm": 1.2665588855743408, "learning_rate": 9.744412684907138e-05, "loss": 0.1906, "step": 2915 }, { "epoch": 2.5504587155963305, "grad_norm": 1.5277937650680542, "learning_rate": 9.716937163792075e-05, "loss": 0.1372, "step": 2920 }, { "epoch": 2.5548274355613803, "grad_norm": 1.4961965084075928, "learning_rate": 9.689463781081542e-05, "loss": 0.1955, "step": 2925 }, { "epoch": 2.559196155526431, "grad_norm": 1.1431431770324707, "learning_rate": 9.661992744323818e-05, "loss": 0.1879, "step": 2930 }, { "epoch": 2.563564875491481, "grad_norm": 1.1940690279006958, "learning_rate": 9.634524261049464e-05, "loss": 0.1565, "step": 2935 }, { "epoch": 2.567933595456531, "grad_norm": 1.741918921470642, "learning_rate": 9.607058538769756e-05, "loss": 0.1563, "step": 2940 }, { "epoch": 2.5723023154215814, "grad_norm": 2.1051597595214844, "learning_rate": 9.579595784975103e-05, "loss": 0.1136, "step": 2945 }, { "epoch": 2.5766710353866316, "grad_norm": 1.361091136932373, "learning_rate": 9.552136207133495e-05, "loss": 0.1554, "step": 2950 }, { "epoch": 2.581039755351682, "grad_norm": 1.1515636444091797, "learning_rate": 9.524680012688928e-05, "loss": 0.1832, "step": 2955 }, { "epoch": 2.5854084753167323, "grad_norm": 1.6642508506774902, "learning_rate": 9.497227409059832e-05, "loss": 0.1716, "step": 2960 }, { "epoch": 2.5897771952817825, "grad_norm": 1.1159862279891968, "learning_rate": 9.469778603637518e-05, "loss": 0.1392, "step": 2965 }, { "epoch": 2.5941459152468327, "grad_norm": 1.3787099123001099, "learning_rate": 9.4423338037846e-05, "loss": 0.1429, "step": 2970 }, { "epoch": 2.598514635211883, "grad_norm": 1.0721005201339722, "learning_rate": 9.414893216833435e-05, "loss": 0.1687, "step": 2975 }, { "epoch": 2.602883355176933, "grad_norm": 1.7461411952972412, "learning_rate": 9.387457050084552e-05, "loss": 0.1757, "step": 2980 }, { "epoch": 2.607252075141983, "grad_norm": 1.4739048480987549, "learning_rate": 9.360025510805078e-05, "loss": 0.1686, "step": 2985 }, { "epoch": 2.611620795107034, "grad_norm": 1.7249925136566162, "learning_rate": 9.332598806227195e-05, "loss": 0.2044, "step": 2990 }, { "epoch": 2.615989515072084, "grad_norm": 0.8930152654647827, "learning_rate": 9.305177143546557e-05, "loss": 0.1248, "step": 2995 }, { "epoch": 2.620358235037134, "grad_norm": 0.8902339935302734, "learning_rate": 9.277760729920728e-05, "loss": 0.1633, "step": 3000 }, { "epoch": 2.6247269550021843, "grad_norm": 2.3700525760650635, "learning_rate": 9.250349772467618e-05, "loss": 0.1805, "step": 3005 }, { "epoch": 2.6290956749672345, "grad_norm": 1.4570140838623047, "learning_rate": 9.222944478263915e-05, "loss": 0.1918, "step": 3010 }, { "epoch": 2.6334643949322847, "grad_norm": 1.3790675401687622, "learning_rate": 9.195545054343529e-05, "loss": 0.1127, "step": 3015 }, { "epoch": 2.637833114897335, "grad_norm": 2.068342924118042, "learning_rate": 9.16815170769602e-05, "loss": 0.2045, "step": 3020 }, { "epoch": 2.6422018348623855, "grad_norm": 2.2792484760284424, "learning_rate": 9.14076464526504e-05, "loss": 0.1942, "step": 3025 }, { "epoch": 2.6465705548274356, "grad_norm": 2.240042209625244, "learning_rate": 9.113384073946765e-05, "loss": 0.1725, "step": 3030 }, { "epoch": 2.650939274792486, "grad_norm": 1.1308009624481201, "learning_rate": 9.086010200588328e-05, "loss": 0.1611, "step": 3035 }, { "epoch": 2.655307994757536, "grad_norm": 1.0229357481002808, "learning_rate": 9.05864323198627e-05, "loss": 0.1892, "step": 3040 }, { "epoch": 2.659676714722586, "grad_norm": 1.2049839496612549, "learning_rate": 9.03128337488497e-05, "loss": 0.1585, "step": 3045 }, { "epoch": 2.6640454346876368, "grad_norm": 1.1866313219070435, "learning_rate": 9.003930835975082e-05, "loss": 0.1224, "step": 3050 }, { "epoch": 2.6684141546526865, "grad_norm": 1.1551754474639893, "learning_rate": 8.976585821891966e-05, "loss": 0.1275, "step": 3055 }, { "epoch": 2.672782874617737, "grad_norm": 1.956389307975769, "learning_rate": 8.949248539214145e-05, "loss": 0.1537, "step": 3060 }, { "epoch": 2.6771515945827873, "grad_norm": 0.9731565117835999, "learning_rate": 8.921919194461735e-05, "loss": 0.1233, "step": 3065 }, { "epoch": 2.6815203145478375, "grad_norm": 0.7945232391357422, "learning_rate": 8.894597994094879e-05, "loss": 0.1518, "step": 3070 }, { "epoch": 2.6858890345128876, "grad_norm": 1.3885526657104492, "learning_rate": 8.867285144512202e-05, "loss": 0.187, "step": 3075 }, { "epoch": 2.690257754477938, "grad_norm": 1.2687430381774902, "learning_rate": 8.839980852049229e-05, "loss": 0.1599, "step": 3080 }, { "epoch": 2.6946264744429884, "grad_norm": 1.0735893249511719, "learning_rate": 8.812685322976851e-05, "loss": 0.1826, "step": 3085 }, { "epoch": 2.6989951944080386, "grad_norm": 1.6104284524917603, "learning_rate": 8.785398763499755e-05, "loss": 0.1791, "step": 3090 }, { "epoch": 2.7033639143730888, "grad_norm": 1.6893842220306396, "learning_rate": 8.758121379754865e-05, "loss": 0.1588, "step": 3095 }, { "epoch": 2.707732634338139, "grad_norm": 0.967934787273407, "learning_rate": 8.730853377809784e-05, "loss": 0.1376, "step": 3100 }, { "epoch": 2.712101354303189, "grad_norm": 1.2705645561218262, "learning_rate": 8.703594963661241e-05, "loss": 0.1689, "step": 3105 }, { "epoch": 2.7164700742682393, "grad_norm": 1.4289767742156982, "learning_rate": 8.67634634323354e-05, "loss": 0.1802, "step": 3110 }, { "epoch": 2.7208387942332894, "grad_norm": 1.4124540090560913, "learning_rate": 8.64910772237699e-05, "loss": 0.1616, "step": 3115 }, { "epoch": 2.72520751419834, "grad_norm": 1.1596726179122925, "learning_rate": 8.62187930686636e-05, "loss": 0.1676, "step": 3120 }, { "epoch": 2.7295762341633902, "grad_norm": 1.243022084236145, "learning_rate": 8.594661302399332e-05, "loss": 0.139, "step": 3125 }, { "epoch": 2.7339449541284404, "grad_norm": 1.3218811750411987, "learning_rate": 8.56745391459492e-05, "loss": 0.2164, "step": 3130 }, { "epoch": 2.7383136740934906, "grad_norm": 2.225620985031128, "learning_rate": 8.540257348991947e-05, "loss": 0.1359, "step": 3135 }, { "epoch": 2.7426823940585408, "grad_norm": 2.3583812713623047, "learning_rate": 8.513071811047478e-05, "loss": 0.2217, "step": 3140 }, { "epoch": 2.747051114023591, "grad_norm": 2.345612049102783, "learning_rate": 8.48589750613527e-05, "loss": 0.1527, "step": 3145 }, { "epoch": 2.751419833988641, "grad_norm": 0.6265102028846741, "learning_rate": 8.458734639544207e-05, "loss": 0.1387, "step": 3150 }, { "epoch": 2.7557885539536917, "grad_norm": 1.4767670631408691, "learning_rate": 8.431583416476779e-05, "loss": 0.1774, "step": 3155 }, { "epoch": 2.760157273918742, "grad_norm": 1.541107416152954, "learning_rate": 8.404444042047507e-05, "loss": 0.1522, "step": 3160 }, { "epoch": 2.764525993883792, "grad_norm": 1.396525263786316, "learning_rate": 8.377316721281402e-05, "loss": 0.1972, "step": 3165 }, { "epoch": 2.7688947138488422, "grad_norm": 1.0470671653747559, "learning_rate": 8.35020165911242e-05, "loss": 0.1808, "step": 3170 }, { "epoch": 2.7732634338138924, "grad_norm": 1.216828465461731, "learning_rate": 8.323099060381896e-05, "loss": 0.1383, "step": 3175 }, { "epoch": 2.777632153778943, "grad_norm": 1.0746128559112549, "learning_rate": 8.296009129837022e-05, "loss": 0.1591, "step": 3180 }, { "epoch": 2.7820008737439927, "grad_norm": 1.3712584972381592, "learning_rate": 8.268932072129287e-05, "loss": 0.161, "step": 3185 }, { "epoch": 2.7863695937090434, "grad_norm": 2.859886646270752, "learning_rate": 8.241868091812924e-05, "loss": 0.1858, "step": 3190 }, { "epoch": 2.7907383136740935, "grad_norm": 1.0394188165664673, "learning_rate": 8.214817393343383e-05, "loss": 0.1196, "step": 3195 }, { "epoch": 2.7951070336391437, "grad_norm": 1.1338419914245605, "learning_rate": 8.187780181075766e-05, "loss": 0.1264, "step": 3200 }, { "epoch": 2.799475753604194, "grad_norm": 0.9405713677406311, "learning_rate": 8.160756659263298e-05, "loss": 0.1177, "step": 3205 }, { "epoch": 2.803844473569244, "grad_norm": 1.4486256837844849, "learning_rate": 8.13374703205578e-05, "loss": 0.1745, "step": 3210 }, { "epoch": 2.8082131935342947, "grad_norm": 1.058084487915039, "learning_rate": 8.106751503498045e-05, "loss": 0.1639, "step": 3215 }, { "epoch": 2.812581913499345, "grad_norm": 1.712044358253479, "learning_rate": 8.079770277528422e-05, "loss": 0.1335, "step": 3220 }, { "epoch": 2.816950633464395, "grad_norm": 1.0125361680984497, "learning_rate": 8.052803557977175e-05, "loss": 0.1197, "step": 3225 }, { "epoch": 2.821319353429445, "grad_norm": 1.248329758644104, "learning_rate": 8.025851548564999e-05, "loss": 0.1143, "step": 3230 }, { "epoch": 2.8256880733944953, "grad_norm": 0.976321280002594, "learning_rate": 7.998914452901447e-05, "loss": 0.1318, "step": 3235 }, { "epoch": 2.8300567933595455, "grad_norm": 1.4181995391845703, "learning_rate": 7.971992474483413e-05, "loss": 0.1309, "step": 3240 }, { "epoch": 2.8344255133245957, "grad_norm": 2.2041068077087402, "learning_rate": 7.945085816693589e-05, "loss": 0.1843, "step": 3245 }, { "epoch": 2.8387942332896463, "grad_norm": 0.7392102479934692, "learning_rate": 7.918194682798914e-05, "loss": 0.1484, "step": 3250 }, { "epoch": 2.8431629532546965, "grad_norm": 1.1940444707870483, "learning_rate": 7.891319275949066e-05, "loss": 0.1984, "step": 3255 }, { "epoch": 2.8475316732197467, "grad_norm": 0.9530165791511536, "learning_rate": 7.864459799174904e-05, "loss": 0.1537, "step": 3260 }, { "epoch": 2.851900393184797, "grad_norm": 1.3266198635101318, "learning_rate": 7.837616455386954e-05, "loss": 0.1721, "step": 3265 }, { "epoch": 2.856269113149847, "grad_norm": 1.8742265701293945, "learning_rate": 7.810789447373846e-05, "loss": 0.1385, "step": 3270 }, { "epoch": 2.8606378331148976, "grad_norm": 0.6404191851615906, "learning_rate": 7.783978977800818e-05, "loss": 0.1598, "step": 3275 }, { "epoch": 2.8650065530799473, "grad_norm": 1.6875426769256592, "learning_rate": 7.757185249208163e-05, "loss": 0.1133, "step": 3280 }, { "epoch": 2.869375273044998, "grad_norm": 2.657672166824341, "learning_rate": 7.730408464009698e-05, "loss": 0.1347, "step": 3285 }, { "epoch": 2.873743993010048, "grad_norm": 1.3153127431869507, "learning_rate": 7.70364882449125e-05, "loss": 0.1183, "step": 3290 }, { "epoch": 2.8781127129750983, "grad_norm": 2.145387649536133, "learning_rate": 7.676906532809115e-05, "loss": 0.1643, "step": 3295 }, { "epoch": 2.8824814329401485, "grad_norm": 1.2466741800308228, "learning_rate": 7.650181790988527e-05, "loss": 0.156, "step": 3300 }, { "epoch": 2.8868501529051986, "grad_norm": 1.4382154941558838, "learning_rate": 7.62347480092215e-05, "loss": 0.1441, "step": 3305 }, { "epoch": 2.8912188728702493, "grad_norm": 1.2591091394424438, "learning_rate": 7.596785764368539e-05, "loss": 0.1966, "step": 3310 }, { "epoch": 2.895587592835299, "grad_norm": 1.8565058708190918, "learning_rate": 7.570114882950619e-05, "loss": 0.1394, "step": 3315 }, { "epoch": 2.8999563128003496, "grad_norm": 1.146375298500061, "learning_rate": 7.543462358154153e-05, "loss": 0.0894, "step": 3320 }, { "epoch": 2.9043250327653998, "grad_norm": 1.0392186641693115, "learning_rate": 7.51682839132624e-05, "loss": 0.1452, "step": 3325 }, { "epoch": 2.90869375273045, "grad_norm": 2.525780439376831, "learning_rate": 7.49021318367378e-05, "loss": 0.1675, "step": 3330 }, { "epoch": 2.9130624726955, "grad_norm": 2.238649368286133, "learning_rate": 7.463616936261952e-05, "loss": 0.1889, "step": 3335 }, { "epoch": 2.9174311926605503, "grad_norm": 1.5493022203445435, "learning_rate": 7.437039850012704e-05, "loss": 0.1021, "step": 3340 }, { "epoch": 2.921799912625601, "grad_norm": 1.0183563232421875, "learning_rate": 7.410482125703225e-05, "loss": 0.1179, "step": 3345 }, { "epoch": 2.926168632590651, "grad_norm": 0.9551679491996765, "learning_rate": 7.383943963964439e-05, "loss": 0.1458, "step": 3350 }, { "epoch": 2.9305373525557012, "grad_norm": 3.2518422603607178, "learning_rate": 7.357425565279483e-05, "loss": 0.1401, "step": 3355 }, { "epoch": 2.9349060725207514, "grad_norm": 2.4154436588287354, "learning_rate": 7.330927129982191e-05, "loss": 0.2123, "step": 3360 }, { "epoch": 2.9392747924858016, "grad_norm": 1.7631040811538696, "learning_rate": 7.304448858255588e-05, "loss": 0.1705, "step": 3365 }, { "epoch": 2.9436435124508518, "grad_norm": 1.741014003753662, "learning_rate": 7.277990950130369e-05, "loss": 0.1277, "step": 3370 }, { "epoch": 2.948012232415902, "grad_norm": 1.7614057064056396, "learning_rate": 7.25155360548339e-05, "loss": 0.147, "step": 3375 }, { "epoch": 2.9523809523809526, "grad_norm": 1.403290867805481, "learning_rate": 7.225137024036164e-05, "loss": 0.172, "step": 3380 }, { "epoch": 2.9567496723460027, "grad_norm": 1.9428198337554932, "learning_rate": 7.19874140535335e-05, "loss": 0.2027, "step": 3385 }, { "epoch": 2.961118392311053, "grad_norm": 2.5754446983337402, "learning_rate": 7.172366948841232e-05, "loss": 0.1601, "step": 3390 }, { "epoch": 2.965487112276103, "grad_norm": 1.068940281867981, "learning_rate": 7.146013853746237e-05, "loss": 0.1444, "step": 3395 }, { "epoch": 2.9698558322411532, "grad_norm": 1.6970683336257935, "learning_rate": 7.119682319153409e-05, "loss": 0.148, "step": 3400 }, { "epoch": 2.974224552206204, "grad_norm": 1.3835630416870117, "learning_rate": 7.093372543984915e-05, "loss": 0.1352, "step": 3405 }, { "epoch": 2.9785932721712536, "grad_norm": 0.9599865078926086, "learning_rate": 7.067084726998548e-05, "loss": 0.1035, "step": 3410 }, { "epoch": 2.982961992136304, "grad_norm": 2.335697889328003, "learning_rate": 7.040819066786195e-05, "loss": 0.1958, "step": 3415 }, { "epoch": 2.9873307121013544, "grad_norm": 1.150230050086975, "learning_rate": 7.014575761772382e-05, "loss": 0.1602, "step": 3420 }, { "epoch": 2.9916994320664045, "grad_norm": 1.7480928897857666, "learning_rate": 6.988355010212742e-05, "loss": 0.1794, "step": 3425 }, { "epoch": 2.9960681520314547, "grad_norm": 1.2962545156478882, "learning_rate": 6.962157010192529e-05, "loss": 0.1637, "step": 3430 }, { "epoch": 3.0, "grad_norm": 0.600903332233429, "learning_rate": 6.935981959625126e-05, "loss": 0.1508, "step": 3435 }, { "epoch": 3.00436871996505, "grad_norm": 0.9637101888656616, "learning_rate": 6.909830056250527e-05, "loss": 0.1208, "step": 3440 }, { "epoch": 3.0087374399301003, "grad_norm": 1.0528173446655273, "learning_rate": 6.883701497633876e-05, "loss": 0.0717, "step": 3445 }, { "epoch": 3.0131061598951505, "grad_norm": 1.6418970823287964, "learning_rate": 6.857596481163957e-05, "loss": 0.0916, "step": 3450 }, { "epoch": 3.017474879860201, "grad_norm": 1.6248525381088257, "learning_rate": 6.831515204051692e-05, "loss": 0.0801, "step": 3455 }, { "epoch": 3.0218435998252513, "grad_norm": 1.015778660774231, "learning_rate": 6.805457863328683e-05, "loss": 0.0989, "step": 3460 }, { "epoch": 3.0262123197903015, "grad_norm": 0.8818852305412292, "learning_rate": 6.779424655845687e-05, "loss": 0.0915, "step": 3465 }, { "epoch": 3.0305810397553516, "grad_norm": 0.8511345982551575, "learning_rate": 6.75341577827115e-05, "loss": 0.0819, "step": 3470 }, { "epoch": 3.034949759720402, "grad_norm": 0.9664357304573059, "learning_rate": 6.727431427089724e-05, "loss": 0.0947, "step": 3475 }, { "epoch": 3.039318479685452, "grad_norm": 1.4519518613815308, "learning_rate": 6.701471798600766e-05, "loss": 0.0688, "step": 3480 }, { "epoch": 3.0436871996505026, "grad_norm": 1.7815632820129395, "learning_rate": 6.675537088916882e-05, "loss": 0.0865, "step": 3485 }, { "epoch": 3.0480559196155528, "grad_norm": 1.5605847835540771, "learning_rate": 6.6496274939624e-05, "loss": 0.1307, "step": 3490 }, { "epoch": 3.052424639580603, "grad_norm": 1.2095025777816772, "learning_rate": 6.623743209471942e-05, "loss": 0.1197, "step": 3495 }, { "epoch": 3.056793359545653, "grad_norm": 1.2208069562911987, "learning_rate": 6.597884430988917e-05, "loss": 0.1115, "step": 3500 }, { "epoch": 3.0611620795107033, "grad_norm": 1.4211591482162476, "learning_rate": 6.572051353864043e-05, "loss": 0.072, "step": 3505 }, { "epoch": 3.0655307994757535, "grad_norm": 1.2047765254974365, "learning_rate": 6.546244173253878e-05, "loss": 0.0544, "step": 3510 }, { "epoch": 3.0698995194408036, "grad_norm": 1.7253801822662354, "learning_rate": 6.520463084119343e-05, "loss": 0.0794, "step": 3515 }, { "epoch": 3.0742682394058543, "grad_norm": 7.309013366699219, "learning_rate": 6.494708281224255e-05, "loss": 0.1807, "step": 3520 }, { "epoch": 3.0786369593709044, "grad_norm": 1.258571743965149, "learning_rate": 6.468979959133852e-05, "loss": 0.0716, "step": 3525 }, { "epoch": 3.0830056793359546, "grad_norm": 1.4243378639221191, "learning_rate": 6.443278312213312e-05, "loss": 0.1365, "step": 3530 }, { "epoch": 3.0873743993010048, "grad_norm": 2.169480562210083, "learning_rate": 6.417603534626306e-05, "loss": 0.1059, "step": 3535 }, { "epoch": 3.091743119266055, "grad_norm": 1.2186070680618286, "learning_rate": 6.391955820333513e-05, "loss": 0.0943, "step": 3540 }, { "epoch": 3.096111839231105, "grad_norm": 2.4399561882019043, "learning_rate": 6.366335363091165e-05, "loss": 0.0943, "step": 3545 }, { "epoch": 3.1004805591961557, "grad_norm": 1.6928596496582031, "learning_rate": 6.340742356449579e-05, "loss": 0.1019, "step": 3550 }, { "epoch": 3.104849279161206, "grad_norm": 2.0251004695892334, "learning_rate": 6.315176993751699e-05, "loss": 0.0932, "step": 3555 }, { "epoch": 3.109217999126256, "grad_norm": 1.7529054880142212, "learning_rate": 6.289639468131622e-05, "loss": 0.0946, "step": 3560 }, { "epoch": 3.1135867190913062, "grad_norm": 1.67819344997406, "learning_rate": 6.264129972513163e-05, "loss": 0.0813, "step": 3565 }, { "epoch": 3.1179554390563564, "grad_norm": 1.2460740804672241, "learning_rate": 6.238648699608375e-05, "loss": 0.1024, "step": 3570 }, { "epoch": 3.1223241590214066, "grad_norm": 1.6184308528900146, "learning_rate": 6.213195841916104e-05, "loss": 0.0847, "step": 3575 }, { "epoch": 3.126692878986457, "grad_norm": 1.5642549991607666, "learning_rate": 6.187771591720536e-05, "loss": 0.1008, "step": 3580 }, { "epoch": 3.1310615989515074, "grad_norm": 1.9698232412338257, "learning_rate": 6.16237614108973e-05, "loss": 0.0994, "step": 3585 }, { "epoch": 3.1354303189165575, "grad_norm": 1.6373701095581055, "learning_rate": 6.137009681874192e-05, "loss": 0.1204, "step": 3590 }, { "epoch": 3.1397990388816077, "grad_norm": 0.9783329963684082, "learning_rate": 6.111672405705402e-05, "loss": 0.0716, "step": 3595 }, { "epoch": 3.144167758846658, "grad_norm": 1.6490020751953125, "learning_rate": 6.086364503994382e-05, "loss": 0.078, "step": 3600 }, { "epoch": 3.148536478811708, "grad_norm": 2.7672507762908936, "learning_rate": 6.061086167930245e-05, "loss": 0.1102, "step": 3605 }, { "epoch": 3.1529051987767582, "grad_norm": 2.6075148582458496, "learning_rate": 6.035837588478737e-05, "loss": 0.0995, "step": 3610 }, { "epoch": 3.157273918741809, "grad_norm": 1.1910531520843506, "learning_rate": 6.010618956380821e-05, "loss": 0.104, "step": 3615 }, { "epoch": 3.161642638706859, "grad_norm": 2.8808867931365967, "learning_rate": 5.985430462151219e-05, "loss": 0.1065, "step": 3620 }, { "epoch": 3.166011358671909, "grad_norm": 2.5939221382141113, "learning_rate": 5.960272296076972e-05, "loss": 0.1232, "step": 3625 }, { "epoch": 3.1703800786369594, "grad_norm": 0.7496421933174133, "learning_rate": 5.935144648216007e-05, "loss": 0.0906, "step": 3630 }, { "epoch": 3.1747487986020095, "grad_norm": 1.836545705795288, "learning_rate": 5.910047708395703e-05, "loss": 0.0794, "step": 3635 }, { "epoch": 3.1791175185670597, "grad_norm": 1.6734864711761475, "learning_rate": 5.884981666211452e-05, "loss": 0.1003, "step": 3640 }, { "epoch": 3.18348623853211, "grad_norm": 2.2766218185424805, "learning_rate": 5.8599467110252305e-05, "loss": 0.1069, "step": 3645 }, { "epoch": 3.1878549584971605, "grad_norm": 1.0550603866577148, "learning_rate": 5.8349430319641716e-05, "loss": 0.0758, "step": 3650 }, { "epoch": 3.1922236784622107, "grad_norm": 1.994390606880188, "learning_rate": 5.809970817919121e-05, "loss": 0.0956, "step": 3655 }, { "epoch": 3.196592398427261, "grad_norm": 1.6109845638275146, "learning_rate": 5.785030257543234e-05, "loss": 0.1103, "step": 3660 }, { "epoch": 3.200961118392311, "grad_norm": 1.7017037868499756, "learning_rate": 5.760121539250532e-05, "loss": 0.0709, "step": 3665 }, { "epoch": 3.205329838357361, "grad_norm": 1.8740450143814087, "learning_rate": 5.7352448512144764e-05, "loss": 0.1109, "step": 3670 }, { "epoch": 3.2096985583224114, "grad_norm": 2.410780191421509, "learning_rate": 5.71040038136657e-05, "loss": 0.1256, "step": 3675 }, { "epoch": 3.214067278287462, "grad_norm": 3.1283838748931885, "learning_rate": 5.6855883173949186e-05, "loss": 0.0943, "step": 3680 }, { "epoch": 3.218435998252512, "grad_norm": 1.051555871963501, "learning_rate": 5.660808846742807e-05, "loss": 0.048, "step": 3685 }, { "epoch": 3.2228047182175623, "grad_norm": 0.789137601852417, "learning_rate": 5.636062156607302e-05, "loss": 0.0728, "step": 3690 }, { "epoch": 3.2271734381826125, "grad_norm": 1.454079031944275, "learning_rate": 5.6113484339378306e-05, "loss": 0.0892, "step": 3695 }, { "epoch": 3.2315421581476627, "grad_norm": 1.1673469543457031, "learning_rate": 5.586667865434766e-05, "loss": 0.0887, "step": 3700 }, { "epoch": 3.235910878112713, "grad_norm": 0.7643448710441589, "learning_rate": 5.5620206375480066e-05, "loss": 0.0947, "step": 3705 }, { "epoch": 3.2402795980777634, "grad_norm": 1.4417402744293213, "learning_rate": 5.537406936475592e-05, "loss": 0.0704, "step": 3710 }, { "epoch": 3.2446483180428136, "grad_norm": 1.4992611408233643, "learning_rate": 5.5128269481622765e-05, "loss": 0.0945, "step": 3715 }, { "epoch": 3.249017038007864, "grad_norm": 1.2667427062988281, "learning_rate": 5.488280858298131e-05, "loss": 0.0671, "step": 3720 }, { "epoch": 3.253385757972914, "grad_norm": 1.1633776426315308, "learning_rate": 5.463768852317146e-05, "loss": 0.0743, "step": 3725 }, { "epoch": 3.257754477937964, "grad_norm": 2.00858211517334, "learning_rate": 5.439291115395808e-05, "loss": 0.0889, "step": 3730 }, { "epoch": 3.2621231979030143, "grad_norm": 0.9733744263648987, "learning_rate": 5.414847832451735e-05, "loss": 0.0445, "step": 3735 }, { "epoch": 3.2664919178680645, "grad_norm": 1.1338526010513306, "learning_rate": 5.390439188142253e-05, "loss": 0.0911, "step": 3740 }, { "epoch": 3.270860637833115, "grad_norm": 0.9580100178718567, "learning_rate": 5.366065366863017e-05, "loss": 0.0815, "step": 3745 }, { "epoch": 3.2752293577981653, "grad_norm": 1.1516432762145996, "learning_rate": 5.341726552746594e-05, "loss": 0.0628, "step": 3750 }, { "epoch": 3.2795980777632154, "grad_norm": 1.5042197704315186, "learning_rate": 5.3174229296611066e-05, "loss": 0.1095, "step": 3755 }, { "epoch": 3.2839667977282656, "grad_norm": 1.399878740310669, "learning_rate": 5.293154681208816e-05, "loss": 0.0836, "step": 3760 }, { "epoch": 3.288335517693316, "grad_norm": 2.0793116092681885, "learning_rate": 5.268921990724751e-05, "loss": 0.1065, "step": 3765 }, { "epoch": 3.292704237658366, "grad_norm": 1.0440312623977661, "learning_rate": 5.2447250412753157e-05, "loss": 0.0904, "step": 3770 }, { "epoch": 3.297072957623416, "grad_norm": 1.3085564374923706, "learning_rate": 5.220564015656899e-05, "loss": 0.0697, "step": 3775 }, { "epoch": 3.3014416775884667, "grad_norm": 1.2351759672164917, "learning_rate": 5.196439096394512e-05, "loss": 0.0894, "step": 3780 }, { "epoch": 3.305810397553517, "grad_norm": 1.8387460708618164, "learning_rate": 5.1723504657403966e-05, "loss": 0.1251, "step": 3785 }, { "epoch": 3.310179117518567, "grad_norm": 2.4653780460357666, "learning_rate": 5.1482983056726544e-05, "loss": 0.061, "step": 3790 }, { "epoch": 3.3145478374836173, "grad_norm": 1.5587716102600098, "learning_rate": 5.124282797893867e-05, "loss": 0.129, "step": 3795 }, { "epoch": 3.3189165574486674, "grad_norm": 0.9583266377449036, "learning_rate": 5.10030412382972e-05, "loss": 0.0981, "step": 3800 }, { "epoch": 3.3232852774137176, "grad_norm": 1.4593812227249146, "learning_rate": 5.0763624646276444e-05, "loss": 0.1017, "step": 3805 }, { "epoch": 3.327653997378768, "grad_norm": 1.0001953840255737, "learning_rate": 5.052458001155442e-05, "loss": 0.1175, "step": 3810 }, { "epoch": 3.3320227173438184, "grad_norm": 2.0613253116607666, "learning_rate": 5.0285909139999175e-05, "loss": 0.0927, "step": 3815 }, { "epoch": 3.3363914373088686, "grad_norm": 1.731712818145752, "learning_rate": 5.004761383465515e-05, "loss": 0.1034, "step": 3820 }, { "epoch": 3.3407601572739187, "grad_norm": 1.1507972478866577, "learning_rate": 4.98096958957295e-05, "loss": 0.0569, "step": 3825 }, { "epoch": 3.345128877238969, "grad_norm": 0.8323259353637695, "learning_rate": 4.9572157120578666e-05, "loss": 0.0931, "step": 3830 }, { "epoch": 3.349497597204019, "grad_norm": 4.5710344314575195, "learning_rate": 4.933499930369466e-05, "loss": 0.0772, "step": 3835 }, { "epoch": 3.3538663171690697, "grad_norm": 1.2585368156433105, "learning_rate": 4.909822423669145e-05, "loss": 0.0881, "step": 3840 }, { "epoch": 3.35823503713412, "grad_norm": 1.0395569801330566, "learning_rate": 4.8861833708291584e-05, "loss": 0.0835, "step": 3845 }, { "epoch": 3.36260375709917, "grad_norm": 1.3017204999923706, "learning_rate": 4.862582950431267e-05, "loss": 0.0804, "step": 3850 }, { "epoch": 3.36697247706422, "grad_norm": 1.038264274597168, "learning_rate": 4.839021340765364e-05, "loss": 0.1365, "step": 3855 }, { "epoch": 3.3713411970292704, "grad_norm": 1.034326434135437, "learning_rate": 4.815498719828163e-05, "loss": 0.082, "step": 3860 }, { "epoch": 3.3757099169943205, "grad_norm": 1.3960469961166382, "learning_rate": 4.7920152653218274e-05, "loss": 0.1569, "step": 3865 }, { "epoch": 3.3800786369593707, "grad_norm": 0.787963330745697, "learning_rate": 4.768571154652645e-05, "loss": 0.0815, "step": 3870 }, { "epoch": 3.3844473569244213, "grad_norm": 0.9106404185295105, "learning_rate": 4.745166564929665e-05, "loss": 0.0726, "step": 3875 }, { "epoch": 3.3888160768894715, "grad_norm": 1.1823543310165405, "learning_rate": 4.7218016729633894e-05, "loss": 0.087, "step": 3880 }, { "epoch": 3.3931847968545217, "grad_norm": 0.7293349504470825, "learning_rate": 4.6984766552644164e-05, "loss": 0.0734, "step": 3885 }, { "epoch": 3.397553516819572, "grad_norm": 4.221632957458496, "learning_rate": 4.675191688042116e-05, "loss": 0.1181, "step": 3890 }, { "epoch": 3.401922236784622, "grad_norm": 1.8997467756271362, "learning_rate": 4.65194694720329e-05, "loss": 0.0869, "step": 3895 }, { "epoch": 3.406290956749672, "grad_norm": 1.2236734628677368, "learning_rate": 4.6287426083508536e-05, "loss": 0.1093, "step": 3900 }, { "epoch": 3.4106596767147224, "grad_norm": 1.8938541412353516, "learning_rate": 4.6055788467825046e-05, "loss": 0.0902, "step": 3905 }, { "epoch": 3.415028396679773, "grad_norm": 1.122361421585083, "learning_rate": 4.5824558374893964e-05, "loss": 0.0838, "step": 3910 }, { "epoch": 3.419397116644823, "grad_norm": 3.0877389907836914, "learning_rate": 4.5593737551548266e-05, "loss": 0.0861, "step": 3915 }, { "epoch": 3.4237658366098733, "grad_norm": 1.2574303150177002, "learning_rate": 4.5363327741528916e-05, "loss": 0.0985, "step": 3920 }, { "epoch": 3.4281345565749235, "grad_norm": 1.3278526067733765, "learning_rate": 4.5133330685472044e-05, "loss": 0.0877, "step": 3925 }, { "epoch": 3.4325032765399737, "grad_norm": 1.873443841934204, "learning_rate": 4.490374812089553e-05, "loss": 0.0861, "step": 3930 }, { "epoch": 3.436871996505024, "grad_norm": 0.9117352962493896, "learning_rate": 4.4674581782186e-05, "loss": 0.0854, "step": 3935 }, { "epoch": 3.4412407164700745, "grad_norm": 0.6332014203071594, "learning_rate": 4.4445833400585724e-05, "loss": 0.0629, "step": 3940 }, { "epoch": 3.4456094364351246, "grad_norm": 1.233370065689087, "learning_rate": 4.42175047041794e-05, "loss": 0.103, "step": 3945 }, { "epoch": 3.449978156400175, "grad_norm": 1.5003916025161743, "learning_rate": 4.398959741788129e-05, "loss": 0.116, "step": 3950 }, { "epoch": 3.454346876365225, "grad_norm": 1.4277344942092896, "learning_rate": 4.37621132634221e-05, "loss": 0.1092, "step": 3955 }, { "epoch": 3.458715596330275, "grad_norm": 1.2345412969589233, "learning_rate": 4.353505395933596e-05, "loss": 0.1223, "step": 3960 }, { "epoch": 3.4630843162953253, "grad_norm": 1.273594617843628, "learning_rate": 4.330842122094749e-05, "loss": 0.1185, "step": 3965 }, { "epoch": 3.467453036260376, "grad_norm": 1.2188774347305298, "learning_rate": 4.308221676035872e-05, "loss": 0.1116, "step": 3970 }, { "epoch": 3.471821756225426, "grad_norm": 1.0663518905639648, "learning_rate": 4.285644228643633e-05, "loss": 0.1012, "step": 3975 }, { "epoch": 3.4761904761904763, "grad_norm": 1.385391116142273, "learning_rate": 4.263109950479863e-05, "loss": 0.0935, "step": 3980 }, { "epoch": 3.4805591961555264, "grad_norm": 1.8860664367675781, "learning_rate": 4.240619011780273e-05, "loss": 0.0949, "step": 3985 }, { "epoch": 3.4849279161205766, "grad_norm": 1.262670636177063, "learning_rate": 4.218171582453163e-05, "loss": 0.061, "step": 3990 }, { "epoch": 3.489296636085627, "grad_norm": 0.7418899536132812, "learning_rate": 4.195767832078134e-05, "loss": 0.0726, "step": 3995 }, { "epoch": 3.493665356050677, "grad_norm": 1.0607564449310303, "learning_rate": 4.1734079299048214e-05, "loss": 0.0992, "step": 4000 }, { "epoch": 3.4980340760157276, "grad_norm": 2.385943651199341, "learning_rate": 4.151092044851612e-05, "loss": 0.1036, "step": 4005 }, { "epoch": 3.5024027959807777, "grad_norm": 1.1264841556549072, "learning_rate": 4.1288203455043504e-05, "loss": 0.0756, "step": 4010 }, { "epoch": 3.506771515945828, "grad_norm": 1.05461585521698, "learning_rate": 4.106593000115091e-05, "loss": 0.1329, "step": 4015 }, { "epoch": 3.511140235910878, "grad_norm": 0.9836155772209167, "learning_rate": 4.084410176600815e-05, "loss": 0.0716, "step": 4020 }, { "epoch": 3.5155089558759283, "grad_norm": 1.2975742816925049, "learning_rate": 4.062272042542151e-05, "loss": 0.0982, "step": 4025 }, { "epoch": 3.5198776758409784, "grad_norm": 1.0592769384384155, "learning_rate": 4.040178765182133e-05, "loss": 0.0805, "step": 4030 }, { "epoch": 3.5242463958060286, "grad_norm": 1.3170608282089233, "learning_rate": 4.0181305114249215e-05, "loss": 0.0898, "step": 4035 }, { "epoch": 3.5286151157710792, "grad_norm": 2.1055305004119873, "learning_rate": 3.996127447834536e-05, "loss": 0.1372, "step": 4040 }, { "epoch": 3.5329838357361294, "grad_norm": 1.4089146852493286, "learning_rate": 3.974169740633614e-05, "loss": 0.1174, "step": 4045 }, { "epoch": 3.5373525557011796, "grad_norm": 0.8465004563331604, "learning_rate": 3.952257555702149e-05, "loss": 0.0694, "step": 4050 }, { "epoch": 3.5417212756662297, "grad_norm": 1.5724999904632568, "learning_rate": 3.9303910585762296e-05, "loss": 0.0866, "step": 4055 }, { "epoch": 3.54608999563128, "grad_norm": 1.2664058208465576, "learning_rate": 3.908570414446802e-05, "loss": 0.08, "step": 4060 }, { "epoch": 3.5504587155963305, "grad_norm": 1.2871352434158325, "learning_rate": 3.8867957881584014e-05, "loss": 0.1048, "step": 4065 }, { "epoch": 3.5548274355613803, "grad_norm": 1.2822154760360718, "learning_rate": 3.8650673442079355e-05, "loss": 0.0611, "step": 4070 }, { "epoch": 3.559196155526431, "grad_norm": 1.6582127809524536, "learning_rate": 3.843385246743417e-05, "loss": 0.0934, "step": 4075 }, { "epoch": 3.563564875491481, "grad_norm": 1.706179141998291, "learning_rate": 3.821749659562739e-05, "loss": 0.0735, "step": 4080 }, { "epoch": 3.567933595456531, "grad_norm": 1.0783063173294067, "learning_rate": 3.8001607461124314e-05, "loss": 0.0953, "step": 4085 }, { "epoch": 3.5723023154215814, "grad_norm": 1.1074267625808716, "learning_rate": 3.7786186694864145e-05, "loss": 0.0873, "step": 4090 }, { "epoch": 3.5766710353866316, "grad_norm": 1.518434762954712, "learning_rate": 3.757123592424794e-05, "loss": 0.0786, "step": 4095 }, { "epoch": 3.581039755351682, "grad_norm": 2.0353386402130127, "learning_rate": 3.735675677312609e-05, "loss": 0.0998, "step": 4100 }, { "epoch": 3.5854084753167323, "grad_norm": 2.324700117111206, "learning_rate": 3.714275086178614e-05, "loss": 0.0869, "step": 4105 }, { "epoch": 3.5897771952817825, "grad_norm": 1.0576910972595215, "learning_rate": 3.692921980694057e-05, "loss": 0.1051, "step": 4110 }, { "epoch": 3.5941459152468327, "grad_norm": 2.1334621906280518, "learning_rate": 3.671616522171445e-05, "loss": 0.0839, "step": 4115 }, { "epoch": 3.598514635211883, "grad_norm": 1.1847120523452759, "learning_rate": 3.650358871563343e-05, "loss": 0.0862, "step": 4120 }, { "epoch": 3.602883355176933, "grad_norm": 1.9714267253875732, "learning_rate": 3.629149189461147e-05, "loss": 0.0978, "step": 4125 }, { "epoch": 3.607252075141983, "grad_norm": 1.4258538484573364, "learning_rate": 3.6079876360938805e-05, "loss": 0.0919, "step": 4130 }, { "epoch": 3.611620795107034, "grad_norm": 3.2902631759643555, "learning_rate": 3.5868743713269635e-05, "loss": 0.0724, "step": 4135 }, { "epoch": 3.615989515072084, "grad_norm": 1.475931167602539, "learning_rate": 3.5658095546610306e-05, "loss": 0.1023, "step": 4140 }, { "epoch": 3.620358235037134, "grad_norm": 1.440937876701355, "learning_rate": 3.54479334523071e-05, "loss": 0.0965, "step": 4145 }, { "epoch": 3.6247269550021843, "grad_norm": 1.6201328039169312, "learning_rate": 3.523825901803427e-05, "loss": 0.0625, "step": 4150 }, { "epoch": 3.6290956749672345, "grad_norm": 1.9484727382659912, "learning_rate": 3.502907382778204e-05, "loss": 0.1019, "step": 4155 }, { "epoch": 3.6334643949322847, "grad_norm": 1.3583996295928955, "learning_rate": 3.482037946184456e-05, "loss": 0.0797, "step": 4160 }, { "epoch": 3.637833114897335, "grad_norm": 1.250138282775879, "learning_rate": 3.461217749680807e-05, "loss": 0.0689, "step": 4165 }, { "epoch": 3.6422018348623855, "grad_norm": 1.7499765157699585, "learning_rate": 3.440446950553904e-05, "loss": 0.0668, "step": 4170 }, { "epoch": 3.6465705548274356, "grad_norm": 1.8903212547302246, "learning_rate": 3.41972570571721e-05, "loss": 0.0834, "step": 4175 }, { "epoch": 3.650939274792486, "grad_norm": 1.5133988857269287, "learning_rate": 3.3990541717098315e-05, "loss": 0.0839, "step": 4180 }, { "epoch": 3.655307994757536, "grad_norm": 0.8838937878608704, "learning_rate": 3.378432504695335e-05, "loss": 0.0721, "step": 4185 }, { "epoch": 3.659676714722586, "grad_norm": 1.4382675886154175, "learning_rate": 3.3578608604605686e-05, "loss": 0.0823, "step": 4190 }, { "epoch": 3.6640454346876368, "grad_norm": 1.3646224737167358, "learning_rate": 3.337339394414473e-05, "loss": 0.1051, "step": 4195 }, { "epoch": 3.6684141546526865, "grad_norm": 1.4034249782562256, "learning_rate": 3.316868261586924e-05, "loss": 0.0734, "step": 4200 }, { "epoch": 3.672782874617737, "grad_norm": 1.422706127166748, "learning_rate": 3.296447616627557e-05, "loss": 0.0803, "step": 4205 }, { "epoch": 3.6771515945827873, "grad_norm": 2.778165817260742, "learning_rate": 3.2760776138045846e-05, "loss": 0.0705, "step": 4210 }, { "epoch": 3.6815203145478375, "grad_norm": 2.6988394260406494, "learning_rate": 3.255758407003652e-05, "loss": 0.1061, "step": 4215 }, { "epoch": 3.6858890345128876, "grad_norm": 1.93489670753479, "learning_rate": 3.2354901497266596e-05, "loss": 0.0971, "step": 4220 }, { "epoch": 3.690257754477938, "grad_norm": 1.7685363292694092, "learning_rate": 3.215272995090612e-05, "loss": 0.0756, "step": 4225 }, { "epoch": 3.6946264744429884, "grad_norm": 1.8435022830963135, "learning_rate": 3.195107095826461e-05, "loss": 0.07, "step": 4230 }, { "epoch": 3.6989951944080386, "grad_norm": 1.3396873474121094, "learning_rate": 3.174992604277932e-05, "loss": 0.0739, "step": 4235 }, { "epoch": 3.7033639143730888, "grad_norm": 0.8422791957855225, "learning_rate": 3.15492967240041e-05, "loss": 0.0737, "step": 4240 }, { "epoch": 3.707732634338139, "grad_norm": 2.0195155143737793, "learning_rate": 3.1349184517597586e-05, "loss": 0.0794, "step": 4245 }, { "epoch": 3.712101354303189, "grad_norm": 1.469390630722046, "learning_rate": 3.1149590935311964e-05, "loss": 0.0589, "step": 4250 }, { "epoch": 3.7164700742682393, "grad_norm": 1.3921406269073486, "learning_rate": 3.0950517484981346e-05, "loss": 0.0893, "step": 4255 }, { "epoch": 3.7208387942332894, "grad_norm": 3.0969502925872803, "learning_rate": 3.075196567051061e-05, "loss": 0.0744, "step": 4260 }, { "epoch": 3.72520751419834, "grad_norm": 1.414971947669983, "learning_rate": 3.055393699186386e-05, "loss": 0.0762, "step": 4265 }, { "epoch": 3.7295762341633902, "grad_norm": 1.7889645099639893, "learning_rate": 3.0356432945053203e-05, "loss": 0.0496, "step": 4270 }, { "epoch": 3.7339449541284404, "grad_norm": 1.676378846168518, "learning_rate": 3.0159455022127382e-05, "loss": 0.1013, "step": 4275 }, { "epoch": 3.7383136740934906, "grad_norm": 1.6016095876693726, "learning_rate": 2.996300471116047e-05, "loss": 0.087, "step": 4280 }, { "epoch": 3.7426823940585408, "grad_norm": 1.4002799987792969, "learning_rate": 2.9767083496240766e-05, "loss": 0.0772, "step": 4285 }, { "epoch": 3.747051114023591, "grad_norm": 1.0628836154937744, "learning_rate": 2.9571692857459467e-05, "loss": 0.0795, "step": 4290 }, { "epoch": 3.751419833988641, "grad_norm": 1.440360426902771, "learning_rate": 2.9376834270899523e-05, "loss": 0.0807, "step": 4295 }, { "epoch": 3.7557885539536917, "grad_norm": 1.8291728496551514, "learning_rate": 2.9182509208624518e-05, "loss": 0.0868, "step": 4300 }, { "epoch": 3.760157273918742, "grad_norm": 2.1751294136047363, "learning_rate": 2.898871913866743e-05, "loss": 0.1253, "step": 4305 }, { "epoch": 3.764525993883792, "grad_norm": 1.6346979141235352, "learning_rate": 2.879546552501968e-05, "loss": 0.1091, "step": 4310 }, { "epoch": 3.7688947138488422, "grad_norm": 1.4858824014663696, "learning_rate": 2.8602749827620033e-05, "loss": 0.0822, "step": 4315 }, { "epoch": 3.7732634338138924, "grad_norm": 1.60005521774292, "learning_rate": 2.841057350234354e-05, "loss": 0.1123, "step": 4320 }, { "epoch": 3.777632153778943, "grad_norm": 2.3461992740631104, "learning_rate": 2.8218938000990557e-05, "loss": 0.0868, "step": 4325 }, { "epoch": 3.7820008737439927, "grad_norm": 1.346450686454773, "learning_rate": 2.8027844771275723e-05, "loss": 0.0947, "step": 4330 }, { "epoch": 3.7863695937090434, "grad_norm": 5.1837897300720215, "learning_rate": 2.7837295256817142e-05, "loss": 0.1054, "step": 4335 }, { "epoch": 3.7907383136740935, "grad_norm": 2.5261070728302, "learning_rate": 2.7647290897125366e-05, "loss": 0.0751, "step": 4340 }, { "epoch": 3.7951070336391437, "grad_norm": 1.8613882064819336, "learning_rate": 2.7457833127592646e-05, "loss": 0.0675, "step": 4345 }, { "epoch": 3.799475753604194, "grad_norm": 1.7511388063430786, "learning_rate": 2.7268923379481838e-05, "loss": 0.0767, "step": 4350 }, { "epoch": 3.803844473569244, "grad_norm": 1.0222926139831543, "learning_rate": 2.708056307991592e-05, "loss": 0.1155, "step": 4355 }, { "epoch": 3.8082131935342947, "grad_norm": 2.234628677368164, "learning_rate": 2.6892753651866997e-05, "loss": 0.1068, "step": 4360 }, { "epoch": 3.812581913499345, "grad_norm": 1.8765060901641846, "learning_rate": 2.6705496514145546e-05, "loss": 0.0806, "step": 4365 }, { "epoch": 3.816950633464395, "grad_norm": 1.1284332275390625, "learning_rate": 2.6518793081389826e-05, "loss": 0.0605, "step": 4370 }, { "epoch": 3.821319353429445, "grad_norm": 1.271816372871399, "learning_rate": 2.6332644764055116e-05, "loss": 0.095, "step": 4375 }, { "epoch": 3.8256880733944953, "grad_norm": 0.6583630442619324, "learning_rate": 2.614705296840301e-05, "loss": 0.0785, "step": 4380 }, { "epoch": 3.8300567933595455, "grad_norm": 0.9801815152168274, "learning_rate": 2.5962019096490875e-05, "loss": 0.0781, "step": 4385 }, { "epoch": 3.8344255133245957, "grad_norm": 0.9474557638168335, "learning_rate": 2.5777544546161246e-05, "loss": 0.0735, "step": 4390 }, { "epoch": 3.8387942332896463, "grad_norm": 1.8281831741333008, "learning_rate": 2.559363071103129e-05, "loss": 0.0916, "step": 4395 }, { "epoch": 3.8431629532546965, "grad_norm": 1.8808575868606567, "learning_rate": 2.5410278980482107e-05, "loss": 0.1045, "step": 4400 }, { "epoch": 3.8475316732197467, "grad_norm": 0.7540014982223511, "learning_rate": 2.5227490739648487e-05, "loss": 0.0826, "step": 4405 }, { "epoch": 3.851900393184797, "grad_norm": 1.4811105728149414, "learning_rate": 2.5045267369408265e-05, "loss": 0.0919, "step": 4410 }, { "epoch": 3.856269113149847, "grad_norm": 1.621551275253296, "learning_rate": 2.486361024637198e-05, "loss": 0.0674, "step": 4415 }, { "epoch": 3.8606378331148976, "grad_norm": 1.2833832502365112, "learning_rate": 2.468252074287245e-05, "loss": 0.1036, "step": 4420 }, { "epoch": 3.8650065530799473, "grad_norm": 1.6416436433792114, "learning_rate": 2.4502000226954337e-05, "loss": 0.1043, "step": 4425 }, { "epoch": 3.869375273044998, "grad_norm": 1.1891071796417236, "learning_rate": 2.4322050062363945e-05, "loss": 0.0864, "step": 4430 }, { "epoch": 3.873743993010048, "grad_norm": 1.4844835996627808, "learning_rate": 2.4142671608538792e-05, "loss": 0.0876, "step": 4435 }, { "epoch": 3.8781127129750983, "grad_norm": 1.3718053102493286, "learning_rate": 2.3963866220597442e-05, "loss": 0.0898, "step": 4440 }, { "epoch": 3.8824814329401485, "grad_norm": 2.1567909717559814, "learning_rate": 2.378563524932922e-05, "loss": 0.0705, "step": 4445 }, { "epoch": 3.8868501529051986, "grad_norm": 2.1048190593719482, "learning_rate": 2.3607980041183918e-05, "loss": 0.0985, "step": 4450 }, { "epoch": 3.8912188728702493, "grad_norm": 1.1503431797027588, "learning_rate": 2.343090193826183e-05, "loss": 0.084, "step": 4455 }, { "epoch": 3.895587592835299, "grad_norm": 0.7724549770355225, "learning_rate": 2.3254402278303423e-05, "loss": 0.0734, "step": 4460 }, { "epoch": 3.8999563128003496, "grad_norm": 0.9084768891334534, "learning_rate": 2.3078482394679325e-05, "loss": 0.086, "step": 4465 }, { "epoch": 3.9043250327653998, "grad_norm": 1.570974588394165, "learning_rate": 2.2903143616380274e-05, "loss": 0.096, "step": 4470 }, { "epoch": 3.90869375273045, "grad_norm": 1.4236451387405396, "learning_rate": 2.2728387268006922e-05, "loss": 0.0946, "step": 4475 }, { "epoch": 3.9130624726955, "grad_norm": 1.0874404907226562, "learning_rate": 2.2554214669760054e-05, "loss": 0.0875, "step": 4480 }, { "epoch": 3.9174311926605503, "grad_norm": 1.8805941343307495, "learning_rate": 2.2380627137430443e-05, "loss": 0.0804, "step": 4485 }, { "epoch": 3.921799912625601, "grad_norm": 2.1116044521331787, "learning_rate": 2.2207625982389025e-05, "loss": 0.1026, "step": 4490 }, { "epoch": 3.926168632590651, "grad_norm": 1.3165743350982666, "learning_rate": 2.203521251157682e-05, "loss": 0.0777, "step": 4495 }, { "epoch": 3.9305373525557012, "grad_norm": 1.7966806888580322, "learning_rate": 2.18633880274953e-05, "loss": 0.0906, "step": 4500 }, { "epoch": 3.9349060725207514, "grad_norm": 1.0090104341506958, "learning_rate": 2.169215382819636e-05, "loss": 0.0713, "step": 4505 }, { "epoch": 3.9392747924858016, "grad_norm": 0.9656439423561096, "learning_rate": 2.1521511207272595e-05, "loss": 0.0654, "step": 4510 }, { "epoch": 3.9436435124508518, "grad_norm": 0.5651183724403381, "learning_rate": 2.135146145384752e-05, "loss": 0.0727, "step": 4515 }, { "epoch": 3.948012232415902, "grad_norm": 2.1474905014038086, "learning_rate": 2.1182005852565755e-05, "loss": 0.0808, "step": 4520 }, { "epoch": 3.9523809523809526, "grad_norm": 1.0668432712554932, "learning_rate": 2.1013145683583446e-05, "loss": 0.1044, "step": 4525 }, { "epoch": 3.9567496723460027, "grad_norm": 0.9837005138397217, "learning_rate": 2.084488222255855e-05, "loss": 0.0537, "step": 4530 }, { "epoch": 3.961118392311053, "grad_norm": 0.860316812992096, "learning_rate": 2.0677216740641082e-05, "loss": 0.0666, "step": 4535 }, { "epoch": 3.965487112276103, "grad_norm": 1.4890446662902832, "learning_rate": 2.051015050446372e-05, "loss": 0.0707, "step": 4540 }, { "epoch": 3.9698558322411532, "grad_norm": 0.803032636642456, "learning_rate": 2.0343684776132098e-05, "loss": 0.0836, "step": 4545 }, { "epoch": 3.974224552206204, "grad_norm": 1.6997323036193848, "learning_rate": 2.017782081321523e-05, "loss": 0.077, "step": 4550 }, { "epoch": 3.9785932721712536, "grad_norm": 0.9332925081253052, "learning_rate": 2.0012559868736157e-05, "loss": 0.0739, "step": 4555 }, { "epoch": 3.982961992136304, "grad_norm": 0.9704071879386902, "learning_rate": 1.984790319116239e-05, "loss": 0.0738, "step": 4560 }, { "epoch": 3.9873307121013544, "grad_norm": 4.083499431610107, "learning_rate": 1.96838520243965e-05, "loss": 0.0937, "step": 4565 }, { "epoch": 3.9916994320664045, "grad_norm": 1.2434906959533691, "learning_rate": 1.9520407607766644e-05, "loss": 0.0514, "step": 4570 }, { "epoch": 3.9960681520314547, "grad_norm": 1.4120557308197021, "learning_rate": 1.9357571176017343e-05, "loss": 0.0881, "step": 4575 }, { "epoch": 4.0, "grad_norm": 2.6155974864959717, "learning_rate": 1.9195343959300073e-05, "loss": 0.079, "step": 4580 }, { "epoch": 4.004368719965051, "grad_norm": 0.6208202838897705, "learning_rate": 1.9033727183163964e-05, "loss": 0.054, "step": 4585 }, { "epoch": 4.0087374399301, "grad_norm": 1.010776162147522, "learning_rate": 1.88727220685466e-05, "loss": 0.0404, "step": 4590 }, { "epoch": 4.013106159895151, "grad_norm": 0.9651484489440918, "learning_rate": 1.8712329831764663e-05, "loss": 0.077, "step": 4595 }, { "epoch": 4.017474879860201, "grad_norm": 0.5756092071533203, "learning_rate": 1.8552551684504904e-05, "loss": 0.0353, "step": 4600 }, { "epoch": 4.021843599825251, "grad_norm": 0.8071658611297607, "learning_rate": 1.839338883381494e-05, "loss": 0.041, "step": 4605 }, { "epoch": 4.026212319790301, "grad_norm": 0.8125472664833069, "learning_rate": 1.8234842482094096e-05, "loss": 0.0558, "step": 4610 }, { "epoch": 4.030581039755352, "grad_norm": 0.6879511475563049, "learning_rate": 1.8076913827084263e-05, "loss": 0.0427, "step": 4615 }, { "epoch": 4.034949759720402, "grad_norm": 0.8721114993095398, "learning_rate": 1.791960406186105e-05, "loss": 0.044, "step": 4620 }, { "epoch": 4.039318479685452, "grad_norm": 0.3937700390815735, "learning_rate": 1.7762914374824592e-05, "loss": 0.0481, "step": 4625 }, { "epoch": 4.043687199650503, "grad_norm": 0.4590921103954315, "learning_rate": 1.7606845949690622e-05, "loss": 0.0354, "step": 4630 }, { "epoch": 4.048055919615552, "grad_norm": 0.786134660243988, "learning_rate": 1.7451399965481575e-05, "loss": 0.0478, "step": 4635 }, { "epoch": 4.052424639580603, "grad_norm": 1.1652159690856934, "learning_rate": 1.7296577596517573e-05, "loss": 0.0504, "step": 4640 }, { "epoch": 4.056793359545653, "grad_norm": 1.1168032884597778, "learning_rate": 1.714238001240769e-05, "loss": 0.0377, "step": 4645 }, { "epoch": 4.061162079510703, "grad_norm": 1.0049817562103271, "learning_rate": 1.698880837804101e-05, "loss": 0.0355, "step": 4650 }, { "epoch": 4.065530799475754, "grad_norm": 0.41330111026763916, "learning_rate": 1.6835863853577894e-05, "loss": 0.0384, "step": 4655 }, { "epoch": 4.069899519440804, "grad_norm": 1.0069596767425537, "learning_rate": 1.6683547594441196e-05, "loss": 0.0346, "step": 4660 }, { "epoch": 4.074268239405854, "grad_norm": 1.251977801322937, "learning_rate": 1.6531860751307437e-05, "loss": 0.0362, "step": 4665 }, { "epoch": 4.078636959370904, "grad_norm": 0.603995144367218, "learning_rate": 1.638080447009832e-05, "loss": 0.0377, "step": 4670 }, { "epoch": 4.083005679335955, "grad_norm": 1.447156548500061, "learning_rate": 1.623037989197187e-05, "loss": 0.0337, "step": 4675 }, { "epoch": 4.087374399301005, "grad_norm": 1.378291368484497, "learning_rate": 1.6080588153313936e-05, "loss": 0.0381, "step": 4680 }, { "epoch": 4.091743119266055, "grad_norm": 1.1817421913146973, "learning_rate": 1.5931430385729585e-05, "loss": 0.0395, "step": 4685 }, { "epoch": 4.0961118392311056, "grad_norm": 0.6046543121337891, "learning_rate": 1.5782907716034467e-05, "loss": 0.0521, "step": 4690 }, { "epoch": 4.100480559196155, "grad_norm": 0.769627571105957, "learning_rate": 1.563502126624642e-05, "loss": 0.0333, "step": 4695 }, { "epoch": 4.104849279161206, "grad_norm": 1.5845779180526733, "learning_rate": 1.548777215357696e-05, "loss": 0.0424, "step": 4700 }, { "epoch": 4.109217999126256, "grad_norm": 2.135448694229126, "learning_rate": 1.5341161490422774e-05, "loss": 0.0444, "step": 4705 }, { "epoch": 4.113586719091306, "grad_norm": 0.9047824144363403, "learning_rate": 1.5195190384357404e-05, "loss": 0.0398, "step": 4710 }, { "epoch": 4.117955439056357, "grad_norm": 1.6250332593917847, "learning_rate": 1.5049859938122856e-05, "loss": 0.0736, "step": 4715 }, { "epoch": 4.122324159021407, "grad_norm": 1.6263458728790283, "learning_rate": 1.4905171249621207e-05, "loss": 0.0458, "step": 4720 }, { "epoch": 4.126692878986457, "grad_norm": 1.2772765159606934, "learning_rate": 1.4761125411906396e-05, "loss": 0.051, "step": 4725 }, { "epoch": 4.131061598951507, "grad_norm": 3.2422568798065186, "learning_rate": 1.4617723513175951e-05, "loss": 0.0454, "step": 4730 }, { "epoch": 4.1354303189165575, "grad_norm": 1.4716161489486694, "learning_rate": 1.4474966636762676e-05, "loss": 0.0475, "step": 4735 }, { "epoch": 4.139799038881607, "grad_norm": 0.8840832114219666, "learning_rate": 1.4332855861126627e-05, "loss": 0.0306, "step": 4740 }, { "epoch": 4.144167758846658, "grad_norm": 0.8200462460517883, "learning_rate": 1.4191392259846804e-05, "loss": 0.0621, "step": 4745 }, { "epoch": 4.1485364788117085, "grad_norm": 0.9062876105308533, "learning_rate": 1.4050576901613177e-05, "loss": 0.0392, "step": 4750 }, { "epoch": 4.152905198776758, "grad_norm": 1.0915573835372925, "learning_rate": 1.3910410850218535e-05, "loss": 0.0398, "step": 4755 }, { "epoch": 4.157273918741809, "grad_norm": 0.7680928707122803, "learning_rate": 1.377089516455039e-05, "loss": 0.0404, "step": 4760 }, { "epoch": 4.161642638706859, "grad_norm": 0.72313392162323, "learning_rate": 1.3632030898583137e-05, "loss": 0.0536, "step": 4765 }, { "epoch": 4.166011358671909, "grad_norm": 0.9777783155441284, "learning_rate": 1.3493819101369987e-05, "loss": 0.0384, "step": 4770 }, { "epoch": 4.17038007863696, "grad_norm": 1.1484917402267456, "learning_rate": 1.3356260817035049e-05, "loss": 0.0559, "step": 4775 }, { "epoch": 4.1747487986020095, "grad_norm": 2.152930498123169, "learning_rate": 1.3219357084765482e-05, "loss": 0.0453, "step": 4780 }, { "epoch": 4.17911751856706, "grad_norm": 0.8922104835510254, "learning_rate": 1.3083108938803545e-05, "loss": 0.0182, "step": 4785 }, { "epoch": 4.18348623853211, "grad_norm": 1.9989736080169678, "learning_rate": 1.2947517408438925e-05, "loss": 0.0379, "step": 4790 }, { "epoch": 4.1878549584971605, "grad_norm": 0.7962925434112549, "learning_rate": 1.2812583518000864e-05, "loss": 0.0493, "step": 4795 }, { "epoch": 4.19222367846221, "grad_norm": 0.9282207489013672, "learning_rate": 1.2678308286850471e-05, "loss": 0.0317, "step": 4800 }, { "epoch": 4.196592398427261, "grad_norm": 1.1934535503387451, "learning_rate": 1.2544692729372998e-05, "loss": 0.0461, "step": 4805 }, { "epoch": 4.2009611183923115, "grad_norm": 0.933497428894043, "learning_rate": 1.2411737854970118e-05, "loss": 0.0419, "step": 4810 }, { "epoch": 4.205329838357361, "grad_norm": 1.0183347463607788, "learning_rate": 1.227944466805242e-05, "loss": 0.0351, "step": 4815 }, { "epoch": 4.209698558322412, "grad_norm": 1.4731885194778442, "learning_rate": 1.2147814168031746e-05, "loss": 0.043, "step": 4820 }, { "epoch": 4.2140672782874615, "grad_norm": 1.1246616840362549, "learning_rate": 1.2016847349313665e-05, "loss": 0.0299, "step": 4825 }, { "epoch": 4.218435998252512, "grad_norm": 0.8816130757331848, "learning_rate": 1.1886545201289933e-05, "loss": 0.0545, "step": 4830 }, { "epoch": 4.222804718217562, "grad_norm": 0.6159882545471191, "learning_rate": 1.1756908708331027e-05, "loss": 0.0314, "step": 4835 }, { "epoch": 4.2271734381826125, "grad_norm": 0.9977670311927795, "learning_rate": 1.162793884977873e-05, "loss": 0.0451, "step": 4840 }, { "epoch": 4.231542158147663, "grad_norm": 1.1949328184127808, "learning_rate": 1.1499636599938735e-05, "loss": 0.0372, "step": 4845 }, { "epoch": 4.235910878112713, "grad_norm": 1.4282013177871704, "learning_rate": 1.1372002928073278e-05, "loss": 0.06, "step": 4850 }, { "epoch": 4.2402795980777634, "grad_norm": 0.7269517779350281, "learning_rate": 1.1245038798393748e-05, "loss": 0.0223, "step": 4855 }, { "epoch": 4.244648318042813, "grad_norm": 1.2948346138000488, "learning_rate": 1.1118745170053513e-05, "loss": 0.0453, "step": 4860 }, { "epoch": 4.249017038007864, "grad_norm": 1.4898144006729126, "learning_rate": 1.0993122997140614e-05, "loss": 0.0398, "step": 4865 }, { "epoch": 4.253385757972914, "grad_norm": 0.6974703669548035, "learning_rate": 1.0868173228670598e-05, "loss": 0.0353, "step": 4870 }, { "epoch": 4.257754477937964, "grad_norm": 1.3846169710159302, "learning_rate": 1.0743896808579235e-05, "loss": 0.0526, "step": 4875 }, { "epoch": 4.262123197903015, "grad_norm": 1.4656192064285278, "learning_rate": 1.0620294675715547e-05, "loss": 0.0586, "step": 4880 }, { "epoch": 4.2664919178680645, "grad_norm": 1.5580966472625732, "learning_rate": 1.049736776383462e-05, "loss": 0.0483, "step": 4885 }, { "epoch": 4.270860637833115, "grad_norm": 0.7030579447746277, "learning_rate": 1.0375117001590528e-05, "loss": 0.0482, "step": 4890 }, { "epoch": 4.275229357798165, "grad_norm": 1.019714593887329, "learning_rate": 1.0253543312529411e-05, "loss": 0.0242, "step": 4895 }, { "epoch": 4.279598077763215, "grad_norm": 0.739118218421936, "learning_rate": 1.0132647615082436e-05, "loss": 0.0307, "step": 4900 }, { "epoch": 4.283966797728265, "grad_norm": 1.2523683309555054, "learning_rate": 1.0012430822558804e-05, "loss": 0.0459, "step": 4905 }, { "epoch": 4.288335517693316, "grad_norm": 1.1019269227981567, "learning_rate": 9.892893843138995e-06, "loss": 0.0425, "step": 4910 }, { "epoch": 4.292704237658366, "grad_norm": 1.9455925226211548, "learning_rate": 9.774037579867801e-06, "loss": 0.0422, "step": 4915 }, { "epoch": 4.297072957623416, "grad_norm": 0.9998888969421387, "learning_rate": 9.655862930647498e-06, "loss": 0.0411, "step": 4920 }, { "epoch": 4.301441677588467, "grad_norm": 0.5358383655548096, "learning_rate": 9.538370788231154e-06, "loss": 0.0953, "step": 4925 }, { "epoch": 4.3058103975535165, "grad_norm": 1.6908740997314453, "learning_rate": 9.421562040215748e-06, "loss": 0.0391, "step": 4930 }, { "epoch": 4.310179117518567, "grad_norm": 2.0013976097106934, "learning_rate": 9.30543756903558e-06, "loss": 0.0309, "step": 4935 }, { "epoch": 4.314547837483618, "grad_norm": 2.235011100769043, "learning_rate": 9.18999825195559e-06, "loss": 0.0359, "step": 4940 }, { "epoch": 4.318916557448667, "grad_norm": 1.4524630308151245, "learning_rate": 9.075244961064656e-06, "loss": 0.0342, "step": 4945 }, { "epoch": 4.323285277413718, "grad_norm": 1.2457258701324463, "learning_rate": 8.961178563269112e-06, "loss": 0.0619, "step": 4950 }, { "epoch": 4.327653997378768, "grad_norm": 1.7082127332687378, "learning_rate": 8.847799920286082e-06, "loss": 0.0348, "step": 4955 }, { "epoch": 4.332022717343818, "grad_norm": 2.379938840866089, "learning_rate": 8.735109888637061e-06, "loss": 0.0512, "step": 4960 }, { "epoch": 4.336391437308868, "grad_norm": 1.533900499343872, "learning_rate": 8.623109319641432e-06, "loss": 0.0434, "step": 4965 }, { "epoch": 4.340760157273919, "grad_norm": 1.5475674867630005, "learning_rate": 8.511799059409997e-06, "loss": 0.0577, "step": 4970 }, { "epoch": 4.345128877238969, "grad_norm": 1.0042622089385986, "learning_rate": 8.401179948838579e-06, "loss": 0.0384, "step": 4975 }, { "epoch": 4.349497597204019, "grad_norm": 0.8465957045555115, "learning_rate": 8.291252823601736e-06, "loss": 0.0315, "step": 4980 }, { "epoch": 4.35386631716907, "grad_norm": 1.8154983520507812, "learning_rate": 8.182018514146405e-06, "loss": 0.0241, "step": 4985 }, { "epoch": 4.358235037134119, "grad_norm": 1.6000292301177979, "learning_rate": 8.073477845685618e-06, "loss": 0.0409, "step": 4990 }, { "epoch": 4.36260375709917, "grad_norm": 1.3662384748458862, "learning_rate": 7.965631638192283e-06, "loss": 0.056, "step": 4995 }, { "epoch": 4.36697247706422, "grad_norm": 1.8090828657150269, "learning_rate": 7.858480706392969e-06, "loss": 0.0497, "step": 5000 } ], "logging_steps": 5, "max_steps": 5725, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 2427767566958592.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }