{"step": 50, "train_loss": 10.734990730285645, "lr": 6.25e-05, "grad_norm": 1.2702687978744507, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0002238468587166e-06, 0.4817124903202057, 0.7315071225166321, 0.8611609935760498, 0.9281138181686401, 0.9627776145935059, 0.9807195067405701, 0.9900127649307251], "elapsed_s": 43.85616445541382} {"step": 100, "train_loss": 9.31834909439087, "lr": 0.000125, "grad_norm": 1.829422116279602, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0009778179664863e-06, 0.4821922481060028, 0.731857419013977, 0.8615373969078064, 0.9283316135406494, 0.9629033803939819, 0.9807791709899902, 0.9900410771369934], "elapsed_s": 86.28551363945007} {"step": 150, "train_loss": 7.084281740188598, "lr": 0.0001875, "grad_norm": 1.1026755571365356, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0002305543821421e-06, 0.4802604913711548, 0.7303768992424011, 0.8608647584915161, 0.928313136100769, 0.9630590081214905, 0.9808744788169861, 0.9900770783424377], "elapsed_s": 128.89940190315247} {"step": 200, "train_loss": 5.92919924736023, "lr": 0.00025, "grad_norm": 0.7145845293998718, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [9.999712347052991e-07, 0.4773836135864258, 0.7281670570373535, 0.8594527840614319, 0.9279407858848572, 0.9631631970405579, 0.9809725284576416, 0.9901391267776489], "elapsed_s": 171.4031205177307} {"step": 250, "train_loss": 5.658856353759766, "lr": 0.0003125, "grad_norm": 2.3270208835601807, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [9.999930625781417e-07, 0.47444576025009155, 0.7256559133529663, 0.8578734397888184, 0.9276237487792969, 0.9634348750114441, 0.9811891317367554, 0.9902123808860779], "elapsed_s": 213.6991572380066} {"step": 300, "train_loss": 5.411606016159058, "lr": 0.000375, "grad_norm": 1.655197024345398, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [9.997880852097296e-07, 0.4711323082447052, 0.7231389880180359, 0.8567068576812744, 0.9284197688102722, 0.9640499949455261, 0.9815149903297424, 0.9903779029846191], "elapsed_s": 256.1730225086212} {"step": 350, "train_loss": 4.938062381744385, "lr": 0.00043750000000000006, "grad_norm": 0.7510020136833191, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [9.998291261581471e-07, 0.46897298097610474, 0.7215477824211121, 0.8559377193450928, 0.9287570714950562, 0.9646110534667969, 0.9818587899208069, 0.9905615448951721], "elapsed_s": 298.6563355922699} {"step": 400, "train_loss": 4.383425388336182, "lr": 0.0005, "grad_norm": 0.8712803721427917, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [9.992352261178894e-07, 0.467578649520874, 0.7205426096916199, 0.8548652529716492, 0.928288996219635, 0.9644635915756226, 0.982043445110321, 0.9907295107841492], "elapsed_s": 340.9387731552124} {"step": 400, "val_loss": 4.223298132419586, "val_ppl": 68.25823815985196} {"step": 450, "train_loss": 4.015942368507385, "lr": 0.0004999487185984905, "grad_norm": 0.8150736689567566, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0006608590629185e-06, 0.46727439761161804, 0.7185104489326477, 0.8524856567382812, 0.926821231842041, 0.9636784791946411, 0.9819425940513611, 0.9908581376075745], "elapsed_s": 390.2024447917938} {"step": 500, "train_loss": 3.791946063041687, "lr": 0.0004997906888557206, "grad_norm": 0.9015093445777893, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0033221542471438e-06, 0.4679517149925232, 0.7169539928436279, 0.8501590490341187, 0.925361156463623, 0.9629070162773132, 0.9816910624504089, 0.9909389019012451], "elapsed_s": 432.7777030467987} {"step": 550, "train_loss": 3.636326537132263, "lr": 0.0004995259569176092, "grad_norm": 0.8536918759346008, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0064098887596629e-06, 0.4687045216560364, 0.7151078581809998, 0.8481104373931885, 0.9241569638252258, 0.962181568145752, 0.9814623594284058, 0.9909554123878479], "elapsed_s": 475.2611827850342} {"step": 600, "train_loss": 3.5342004537582397, "lr": 0.0004991546358687517, "grad_norm": 0.7081570029258728, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0097813856191351e-06, 0.4687873423099518, 0.7131142020225525, 0.8461930751800537, 0.9229856729507446, 0.9614506363868713, 0.9811907410621643, 0.9908850789070129], "elapsed_s": 517.5055453777313} {"step": 650, "train_loss": 3.4384861040115355, "lr": 0.0004986768843250328, "grad_norm": 0.7738450765609741, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.011416998153436e-06, 0.46898776292800903, 0.7107101082801819, 0.843996524810791, 0.9217177629470825, 0.9607320427894592, 0.9809694290161133, 0.9908307790756226], "elapsed_s": 560.0138111114502} {"step": 700, "train_loss": 3.3490319538116453, "lr": 0.0004980929063658709, "grad_norm": 0.8287984132766724, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0136784567293944e-06, 0.46926718950271606, 0.7082386016845703, 0.8417351841926575, 0.9203934669494629, 0.9599849581718445, 0.9806219339370728, 0.9907336235046387], "elapsed_s": 602.5584578514099} {"step": 750, "train_loss": 3.2933116006851195, "lr": 0.0004974029514470429, "grad_norm": 0.7896211743354797, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0140787480850122e-06, 0.4697282314300537, 0.70601886510849, 0.8399713039398193, 0.9192639589309692, 0.959227442741394, 0.9802712798118591, 0.9906533360481262], "elapsed_s": 644.892507314682} {"step": 800, "train_loss": 3.2476093196868896, "lr": 0.0004966073142941239, "grad_norm": 0.6755295395851135, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0169289907935308e-06, 0.47050294280052185, 0.7038367390632629, 0.8381320238113403, 0.9181126356124878, 0.9584382176399231, 0.9799632430076599, 0.990559995174408], "elapsed_s": 687.2594707012177} {"step": 800, "val_loss": 3.2579292237758635, "val_ppl": 25.995650193516358} {"step": 850, "train_loss": 3.199247055053711, "lr": 0.0004957063347765917, "grad_norm": 0.7140491604804993, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0190289003730868e-06, 0.4710964858531952, 0.7014716863632202, 0.8363743424415588, 0.9168205261230469, 0.9576873779296875, 0.9796378016471863, 0.990494966506958], "elapsed_s": 736.653815984726} {"step": 900, "train_loss": 3.1491528654098513, "lr": 0.0004947003977626449, "grad_norm": 0.7625854015350342, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0213103678324842e-06, 0.47173240780830383, 0.6994383931159973, 0.8345409631729126, 0.9155189394950867, 0.9569500684738159, 0.9792838096618652, 0.9904464483261108], "elapsed_s": 779.1607356071472} {"step": 950, "train_loss": 3.1197460317611694, "lr": 0.0004935899329548007, "grad_norm": 0.9000021815299988, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0219689556834055e-06, 0.47231537103652954, 0.6968526840209961, 0.8324554562568665, 0.9143539071083069, 0.9562252163887024, 0.9789531230926514, 0.9903998970985413], "elapsed_s": 821.5044205188751} {"step": 1000, "train_loss": 3.063825559616089, "lr": 0.0004923754147063402, "grad_norm": 0.7314625382423401, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0239825769531308e-06, 0.4732103943824768, 0.6944789290428162, 0.8306148052215576, 0.9133064150810242, 0.9555131793022156, 0.9785862565040588, 0.9903355836868286], "elapsed_s": 864.0523009300232} {"step": 1050, "train_loss": 3.048059973716736, "lr": 0.0004910573618186806, "grad_norm": 0.7288562059402466, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0242229109280743e-06, 0.4739866554737091, 0.6925439834594727, 0.8287705779075623, 0.9122334718704224, 0.9548239707946777, 0.9782277345657349, 0.9902628660202026], "elapsed_s": 906.5086307525635} {"step": 1100, "train_loss": 3.034053015708923, "lr": 0.0004896363373197609, "grad_norm": 0.6297518610954285, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.025352617034514e-06, 0.47500908374786377, 0.690773606300354, 0.8271827697753906, 0.9113901257514954, 0.9542338848114014, 0.9779009222984314, 0.9902053475379944], "elapsed_s": 948.8295276165009} {"step": 1150, "train_loss": 2.99260929107666, "lr": 0.0004881129482235347, "grad_norm": 0.7192388772964478, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0236417438136414e-06, 0.47611045837402344, 0.6892032027244568, 0.825580894947052, 0.9103266000747681, 0.9536670446395874, 0.977605402469635, 0.9901622533798218], "elapsed_s": 991.3601224422455} {"step": 1200, "train_loss": 2.9881377029418945, "lr": 0.0004864878452706749, "grad_norm": 0.7594279050827026, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0228192195427255e-06, 0.47697320580482483, 0.6870102286338806, 0.8238250017166138, 0.9094289541244507, 0.953086256980896, 0.9772968292236328, 0.9901593327522278], "elapsed_s": 1033.8619813919067} {"step": 1200, "val_loss": 3.0253844976425173, "val_ppl": 20.601924579945333} {"step": 1250, "train_loss": 2.9564765644073487, "lr": 0.0004847617226505988, "grad_norm": 0.6386340856552124, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.022395963445888e-06, 0.4777296185493469, 0.6855465173721313, 0.8222944736480713, 0.908504068851471, 0.9525660276412964, 0.9769882559776306, 0.9901034832000732], "elapsed_s": 1083.1503043174744} {"step": 1300, "train_loss": 2.9272817945480347, "lr": 0.00048293531770493347, "grad_norm": 0.9071815013885498, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0231655096504255e-06, 0.4780442714691162, 0.6836577653884888, 0.8206172585487366, 0.9076279401779175, 0.9520626664161682, 0.9767436981201172, 0.9901070594787598], "elapsed_s": 1125.5536270141602} {"step": 1350, "train_loss": 2.903614706993103, "lr": 0.00048100941061254875, "grad_norm": 0.6975362300872803, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0230270390820806e-06, 0.4788578152656555, 0.6820348501205444, 0.8189399242401123, 0.9067961573600769, 0.9515385627746582, 0.976469099521637, 0.9901100993156433], "elapsed_s": 1167.970522403717} {"step": 1400, "train_loss": 2.881771931648254, "lr": 0.0004789848240562905, "grad_norm": 0.7413709759712219, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.022373453452019e-06, 0.47919052839279175, 0.6801131367683411, 0.8174171447753906, 0.9062255024909973, 0.9511388540267944, 0.9762693643569946, 0.9901129007339478], "elapsed_s": 1210.7069098949432} {"step": 1450, "train_loss": 2.845870671272278, "lr": 0.00047686242287155823, "grad_norm": 0.6545770168304443, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0226104905086686e-06, 0.48014888167381287, 0.6786242127418518, 0.8161340951919556, 0.9054211974143982, 0.9506831169128418, 0.9760676026344299, 0.9901042580604553], "elapsed_s": 1253.1822593212128} {"step": 1500, "train_loss": 2.873366665840149, "lr": 0.00047464311367687663, "grad_norm": 0.7443034648895264, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.021077650875668e-06, 0.4802398085594177, 0.6768807768821716, 0.8144398927688599, 0.9047117829322815, 0.9502443671226501, 0.9758262038230896, 0.9900454878807068], "elapsed_s": 1295.599321603775} {"step": 1550, "train_loss": 2.8569049406051636, "lr": 0.0004723278444866184, "grad_norm": 0.6764771342277527, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0209695346929948e-06, 0.4808071553707123, 0.6754607558250427, 0.8129699230194092, 0.9041606187820435, 0.9499630928039551, 0.9755369424819946, 0.9899865984916687], "elapsed_s": 1338.3101603984833} {"step": 1600, "train_loss": 2.8201629400253294, "lr": 0.0004699176043060446, "grad_norm": 0.62758868932724, "gamma": 0.30000001192092896, "gumbel_tau": 1.0, "xi_alphas": [1.0206939577983576e-06, 0.4816034734249115, 0.6741125583648682, 0.8118948340415955, 0.9033462405204773, 0.9495576620101929, 0.9753595590591431, 0.9900047183036804], "elapsed_s": 1380.671379327774} {"step": 1600, "val_loss": 2.8769325852394103, "val_ppl": 17.759713137031618} {"step": 1650, "train_loss": 2.8253145456314086, "lr": 0.00046741342270883437, "grad_norm": 0.661659836769104, "gamma": 0.30000001192092896, "gumbel_tau": 0.9929698705673218, "xi_alphas": [1.020128593154368e-06, 0.48209238052368164, 0.6729132533073425, 0.8109197616577148, 0.9030327200889587, 0.9494014382362366, 0.9752022624015808, 0.9899660348892212], "elapsed_s": 1429.9360041618347} {"step": 1700, "train_loss": 2.777112126350403, "lr": 0.000464816369397286, "grad_norm": 0.666327953338623, "gamma": 0.30000001192092896, "gumbel_tau": 0.9859396815299988, "xi_alphas": [1.0190143484578584e-06, 0.48318013548851013, 0.6717380285263062, 0.8096131086349487, 0.9024155735969543, 0.9490817189216614, 0.9749762415885925, 0.9899915456771851], "elapsed_s": 1472.3905489444733} {"step": 1750, "train_loss": 2.7782770156860352, "lr": 0.00046212755374537594, "grad_norm": 0.5931368470191956, "gamma": 0.30000001192092896, "gumbel_tau": 0.9789095520973206, "xi_alphas": [1.0160478041143506e-06, 0.4838709831237793, 0.6703677177429199, 0.8084045648574829, 0.901915431022644, 0.9487311840057373, 0.9747055172920227, 0.9899275898933411], "elapsed_s": 1515.2212116718292} {"step": 1800, "train_loss": 2.772431139945984, "lr": 0.0004593481243248713, "grad_norm": 0.6457295417785645, "gamma": 0.30000001192092896, "gumbel_tau": 0.9718794226646423, "xi_alphas": [1.013982000586111e-06, 0.48459959030151367, 0.669156014919281, 0.80753493309021, 0.9015550017356873, 0.9485911726951599, 0.9746074080467224, 0.9898820519447327], "elapsed_s": 1557.7190172672272} {"step": 1850, "train_loss": 2.738775553703308, "lr": 0.00045647926841469876, "grad_norm": 0.6292251944541931, "gamma": 0.30000001192092896, "gumbel_tau": 0.9648492336273193, "xi_alphas": [1.0134098147318582e-06, 0.4856792390346527, 0.6683075428009033, 0.8065407276153564, 0.9011130928993225, 0.9483805298805237, 0.9743698239326477, 0.9897868633270264], "elapsed_s": 1600.6029770374298} {"step": 1900, "train_loss": 2.7471803760528566, "lr": 0.000453522211493779, "grad_norm": 0.7223191261291504, "gamma": 0.30000001192092896, "gumbel_tau": 0.9578191041946411, "xi_alphas": [1.012681309475738e-06, 0.4859696626663208, 0.6672297120094299, 0.8056552410125732, 0.9005520343780518, 0.948049008846283, 0.9742001891136169, 0.9897492527961731], "elapsed_s": 1644.2169063091278} {"step": 1950, "train_loss": 2.7289329242706297, "lr": 0.00045047821671754276, "grad_norm": 0.6627303957939148, "gamma": 0.30000001192092896, "gumbel_tau": 0.9507889151573181, "xi_alphas": [1.0122960247826995e-06, 0.4868031144142151, 0.666392982006073, 0.8048924207687378, 0.9001291990280151, 0.9478728175163269, 0.9739952087402344, 0.9896907806396484], "elapsed_s": 1687.1648318767548} {"step": 2000, "train_loss": 2.702504835128784, "lr": 0.0004473485843783537, "grad_norm": 0.6497498750686646, "gamma": 0.30000001192092896, "gumbel_tau": 0.9437587857246399, "xi_alphas": [1.011686094898323e-06, 0.4873221814632416, 0.6654976010322571, 0.8041170239448547, 0.8998672962188721, 0.9479299783706665, 0.9737640023231506, 0.9896008372306824], "elapsed_s": 1730.4733080863953} {"step": 2000, "val_loss": 2.8040360152721404, "val_ppl": 16.51115173385559} {"step": 2050, "train_loss": 2.7027634382247925, "lr": 0.00044413465135006704, "grad_norm": 0.6784384846687317, "gamma": 0.30000001192092896, "gumbel_tau": 0.9367286562919617, "xi_alphas": [1.0108479955306393e-06, 0.4874175488948822, 0.6645726561546326, 0.8033619523048401, 0.8993501663208008, 0.9477128982543945, 0.9736069440841675, 0.9895528554916382], "elapsed_s": 1781.164491891861} {"step": 2100, "train_loss": 2.6849646377563476, "lr": 0.0004408377905169619, "grad_norm": 0.6956356763839722, "gamma": 0.30000001192092896, "gumbel_tau": 0.9296984672546387, "xi_alphas": [1.008410094982537e-06, 0.4876115620136261, 0.663870632648468, 0.8026400804519653, 0.8990164399147034, 0.947594940662384, 0.9735416173934937, 0.9894875288009644], "elapsed_s": 1824.394770860672} {"step": 2150, "train_loss": 2.703203411102295, "lr": 0.0004374594101872913, "grad_norm": 0.6544255018234253, "gamma": 0.30000001192092896, "gumbel_tau": 0.9226683378219604, "xi_alphas": [1.0083438155561453e-06, 0.4882628619670868, 0.6627845168113708, 0.8017413020133972, 0.8985169529914856, 0.9475061297416687, 0.9734689593315125, 0.9894459843635559], "elapsed_s": 1867.1700551509857} {"step": 2200, "train_loss": 2.678543758392334, "lr": 0.00043400095349169996, "grad_norm": 0.6401641368865967, "gamma": 0.30000001192092896, "gumbel_tau": 0.9156382083892822, "xi_alphas": [1.0071444194181822e-06, 0.4883035719394684, 0.6619824171066284, 0.8012616634368896, 0.8983727097511292, 0.9473877549171448, 0.9732435345649719, 0.9893602132797241], "elapsed_s": 1910.0600063800812} {"step": 2250, "train_loss": 2.6841503477096555, "lr": 0.0004304638977667675, "grad_norm": 0.6677794456481934, "gamma": 0.30000001192092896, "gumbel_tau": 0.9086080193519592, "xi_alphas": [1.0044288956123637e-06, 0.48900729417800903, 0.6614410877227783, 0.8005768060684204, 0.8977763652801514, 0.9471811056137085, 0.9730864763259888, 0.9893109798431396], "elapsed_s": 1952.895908355713} {"step": 2300, "train_loss": 2.664648699760437, "lr": 0.0004268497539239402, "grad_norm": 0.9016157984733582, "gamma": 0.30000001192092896, "gumbel_tau": 0.901577889919281, "xi_alphas": [1.003555667011824e-06, 0.48926809430122375, 0.660469651222229, 0.7998034954071045, 0.8972297310829163, 0.9470102190971375, 0.9729001522064209, 0.9892416596412659], "elapsed_s": 1995.0998950004578} {"step": 2350, "train_loss": 2.651437063217163, "lr": 0.0004231600658041194, "grad_norm": 0.6412762403488159, "gamma": 0.30000001192092896, "gumbel_tau": 0.894547700881958, "xi_alphas": [1.0032408681581728e-06, 0.489629328250885, 0.6596353650093079, 0.7992111444473267, 0.8970608711242676, 0.9470024108886719, 0.9727513790130615, 0.9891595244407654], "elapsed_s": 2036.956757068634} {"step": 2400, "train_loss": 2.6589156007766723, "lr": 0.0004193964095181856, "grad_norm": 0.6854519248008728, "gamma": 0.30000001192092896, "gumbel_tau": 0.8875175714492798, "xi_alphas": [1.00259433111205e-06, 0.49038445949554443, 0.6590879559516907, 0.7985088229179382, 0.896687388420105, 0.9469626545906067, 0.9726428389549255, 0.989104151725769], "elapsed_s": 2078.931588411331} {"step": 2400, "val_loss": 2.7511104226112364, "val_ppl": 15.660011463884949} {"step": 2450, "train_loss": 2.63305184841156, "lr": 0.00041556039277373605, "grad_norm": 0.6769775152206421, "gamma": 0.30000001192092896, "gumbel_tau": 0.8804874420166016, "xi_alphas": [1.0003927854995709e-06, 0.4904825985431671, 0.6585503816604614, 0.7977428436279297, 0.8963958024978638, 0.9469143152236938, 0.9725056290626526, 0.9890535473823547], "elapsed_s": 2127.243357181549} {"step": 2500, "train_loss": 2.6504747104644775, "lr": 0.0004116536541883265, "grad_norm": 0.6871638894081116, "gamma": 0.30000001192092896, "gumbel_tau": 0.8734572529792786, "xi_alphas": [9.998776704378542e-07, 0.4911387264728546, 0.658178985118866, 0.7972368001937866, 0.8960312008857727, 0.9467186331748962, 0.9723277688026428, 0.9889979958534241], "elapsed_s": 2168.8604753017426} {"step": 2550, "train_loss": 2.6191725826263426, "lr": 0.0004076778625895085, "grad_norm": 0.6729730367660522, "gamma": 0.30000001192092896, "gumbel_tau": 0.8664271235466003, "xi_alphas": [1.002153680929041e-06, 0.4911564290523529, 0.6575835347175598, 0.796697735786438, 0.895927369594574, 0.9465682506561279, 0.9721391201019287, 0.9889397025108337], "elapsed_s": 2210.7178518772125} {"step": 2600, "train_loss": 2.619506015777588, "lr": 0.0004036347163019624, "grad_norm": 0.7005200386047363, "gamma": 0.30000001192092896, "gumbel_tau": 0.8593969941139221, "xi_alphas": [1.001405507849995e-06, 0.4915800988674164, 0.6571862697601318, 0.7963430285453796, 0.8956236243247986, 0.9464246034622192, 0.9720318913459778, 0.9888757467269897], "elapsed_s": 2252.4622452259064} {"step": 2650, "train_loss": 2.595923981666565, "lr": 0.00039952594242203014, "grad_norm": 0.649885356426239, "gamma": 0.30000001192092896, "gumbel_tau": 0.8523668050765991, "xi_alphas": [1.0001257351177628e-06, 0.4913865923881531, 0.6562029123306274, 0.7956336736679077, 0.8953696489334106, 0.9463444352149963, 0.971863329410553, 0.9887728095054626], "elapsed_s": 2294.072479724884} {"step": 2700, "train_loss": 2.5923688125610354, "lr": 0.0003953532960799577, "grad_norm": 0.6574232578277588, "gamma": 0.30000001192092896, "gumbel_tau": 0.8453366756439209, "xi_alphas": [1.0021411753768916e-06, 0.49154022336006165, 0.6562333106994629, 0.7954502701759338, 0.8952763676643372, 0.946287214756012, 0.971792459487915, 0.9887298941612244], "elapsed_s": 2335.936465024948} {"step": 2750, "train_loss": 2.5814990901947024, "lr": 0.000391118559690162, "grad_norm": 0.6544135808944702, "gamma": 0.30000001192092896, "gumbel_tau": 0.8383064866065979, "xi_alphas": [1.000743964141293e-06, 0.49185219407081604, 0.6558551788330078, 0.7947314977645874, 0.8948419094085693, 0.9462640285491943, 0.9716793894767761, 0.9886896014213562], "elapsed_s": 2377.828433275223} {"step": 2800, "train_loss": 2.5675870609283447, "lr": 0.000386823542189844, "grad_norm": 0.6669807434082031, "gamma": 0.30000001192092896, "gumbel_tau": 0.8312763571739197, "xi_alphas": [9.994658967116266e-07, 0.49219435453414917, 0.6555274724960327, 0.7943354845046997, 0.8947115540504456, 0.9461795091629028, 0.9715828895568848, 0.9886195659637451], "elapsed_s": 2419.712084531784} {"step": 2800, "val_loss": 2.7004212021827696, "val_ppl": 14.88600042045583} {"step": 2850, "train_loss": 2.578838195800781, "lr": 0.0003824700782662709, "grad_norm": 0.6122487783432007, "gamma": 0.30000001192092896, "gumbel_tau": 0.8242462277412415, "xi_alphas": [9.995478649216238e-07, 0.4919794499874115, 0.6547828912734985, 0.7936969995498657, 0.8943763375282288, 0.9460970163345337, 0.9715548753738403, 0.9885751605033875], "elapsed_s": 2468.1272768974304} {"step": 2900, "train_loss": 2.5675631952285767, "lr": 0.00037806002757305945, "grad_norm": 0.6596418023109436, "gamma": 0.30000001192092896, "gumbel_tau": 0.8172160387039185, "xi_alphas": [1.0007563560066046e-06, 0.49221906065940857, 0.6546450853347778, 0.7935342192649841, 0.894315242767334, 0.9460664987564087, 0.9714255928993225, 0.9885280132293701], "elapsed_s": 2510.0288395881653} {"step": 2950, "train_loss": 2.5638837671279906, "lr": 0.0003735952739357946, "grad_norm": 0.6910211443901062, "gamma": 0.30000001192092896, "gumbel_tau": 0.8101859092712402, "xi_alphas": [1.0000064776249928e-06, 0.49200963973999023, 0.65406733751297, 0.7932326197624207, 0.8941636681556702, 0.9460159540176392, 0.9712939262390137, 0.9884777665138245], "elapsed_s": 2551.8794639110565} {"step": 3000, "train_loss": 2.560391945838928, "lr": 0.0003690777245473223, "grad_norm": 0.7150830030441284, "gamma": 0.30000001192092896, "gumbel_tau": 0.803155779838562, "xi_alphas": [1.0000541124099982e-06, 0.49213024973869324, 0.6536248326301575, 0.7929742336273193, 0.8940115571022034, 0.9459211826324463, 0.9710966944694519, 0.9883759021759033], "elapsed_s": 2593.5130236148834} {"step": 3050, "train_loss": 2.524870948791504, "lr": 0.00036450930915306063, "grad_norm": 0.6954441666603088, "gamma": 0.30000001192092896, "gumbel_tau": 0.796125590801239, "xi_alphas": [9.9998067071283e-07, 0.49241358041763306, 0.6531942486763, 0.7928144335746765, 0.8937233686447144, 0.9457643032073975, 0.9708650708198547, 0.9882920384407043], "elapsed_s": 2635.4073917865753} {"step": 3100, "train_loss": 2.5240488862991333, "lr": 0.00035989197922667754, "grad_norm": 0.6489050984382629, "gamma": 0.30000001192092896, "gumbel_tau": 0.7890954613685608, "xi_alphas": [1.0009711104430608e-06, 0.492553174495697, 0.6527701020240784, 0.7925757765769958, 0.8936989903450012, 0.9457046985626221, 0.9707566499710083, 0.9882562756538391], "elapsed_s": 2677.208422899246} {"step": 3150, "train_loss": 2.5216298055648805, "lr": 0.0003552277071364868, "grad_norm": 0.6894673705101013, "gamma": 0.30000001192092896, "gumbel_tau": 0.7820653319358826, "xi_alphas": [9.995696927944664e-07, 0.492703378200531, 0.652773916721344, 0.7926658391952515, 0.8935906291007996, 0.9457374811172485, 0.9707631468772888, 0.9882416129112244], "elapsed_s": 2718.844873189926} {"step": 3200, "train_loss": 2.51953706741333, "lr": 0.00035051848530291826, "grad_norm": 0.6934755444526672, "gamma": 0.30000001192092896, "gumbel_tau": 0.7750351428985596, "xi_alphas": [9.99721351035987e-07, 0.49287140369415283, 0.6524567604064941, 0.7922914624214172, 0.8932865262031555, 0.945716142654419, 0.9707397818565369, 0.9882088899612427], "elapsed_s": 2760.703632593155} {"step": 3200, "val_loss": 2.6281477332115175, "val_ppl": 13.848095763926334} {"step": 3250, "train_loss": 2.516075677871704, "lr": 0.00034576632534742306, "grad_norm": 0.6831614971160889, "gamma": 0.30000001192092896, "gumbel_tau": 0.7680050134658813, "xi_alphas": [1.0013338851422304e-06, 0.4930208623409271, 0.6519848704338074, 0.7918332815170288, 0.8929309844970703, 0.9457318186759949, 0.970752477645874, 0.9881183505058289], "elapsed_s": 2809.151472568512} {"step": 3300, "train_loss": 2.50707368850708, "lr": 0.00034097325723317644, "grad_norm": 0.6176900863647461, "gamma": 0.30000001192092896, "gumbel_tau": 0.7609748244285583, "xi_alphas": [1.0005817330238642e-06, 0.4928469955921173, 0.6515646576881409, 0.7917149066925049, 0.8929391503334045, 0.9457149505615234, 0.9706305861473083, 0.9880260825157166], "elapsed_s": 2851.1112082004547} {"step": 3350, "train_loss": 2.5189144372940064, "lr": 0.00033614132839794537, "grad_norm": 0.647682785987854, "gamma": 0.30000001192092896, "gumbel_tau": 0.7539446949958801, "xi_alphas": [1.000642669168883e-06, 0.4928790330886841, 0.6514319777488708, 0.7913609147071838, 0.8925790786743164, 0.9457030296325684, 0.97055983543396, 0.9879907965660095], "elapsed_s": 2892.8471443653107} {"step": 3400, "train_loss": 2.4821101999282837, "lr": 0.00033127260287949196, "grad_norm": 0.7298967838287354, "gamma": 0.30000001192092896, "gumbel_tau": 0.7469145655632019, "xi_alphas": [1.0001066357290256e-06, 0.4932499825954437, 0.6513258218765259, 0.791205644607544, 0.8924990296363831, 0.9456243515014648, 0.9704422950744629, 0.9879278540611267], "elapsed_s": 2934.920313835144} {"step": 3450, "train_loss": 2.495560326576233, "lr": 0.00032636916043388614, "grad_norm": 0.6516448855400085, "gamma": 0.30000001192092896, "gumbel_tau": 0.7398843765258789, "xi_alphas": [9.996879271056969e-07, 0.49296313524246216, 0.6510049700737, 0.790926992893219, 0.8922106027603149, 0.9454430341720581, 0.9702216386795044, 0.9878647923469543], "elapsed_s": 2977.2428209781647} {"step": 3500, "train_loss": 2.4816325235366823, "lr": 0.00032143309564710345, "grad_norm": 0.7134611010551453, "gamma": 0.30000001192092896, "gumbel_tau": 0.7328542470932007, "xi_alphas": [1.0000999282056e-06, 0.4926743805408478, 0.6508308053016663, 0.7907148003578186, 0.8921062350273132, 0.9453803896903992, 0.970095157623291, 0.9878082871437073], "elapsed_s": 3019.297578573227} {"step": 3550, "train_loss": 2.4953946924209593, "lr": 0.0003164665170402881, "grad_norm": 0.7075076699256897, "gamma": 0.30000001192092896, "gumbel_tau": 0.7258241176605225, "xi_alphas": [9.995840173360193e-07, 0.49265673756599426, 0.6507444977760315, 0.7905413508415222, 0.8920143246650696, 0.9454431533813477, 0.9700918197631836, 0.9877331256866455], "elapsed_s": 3061.157565355301} {"step": 3600, "train_loss": 2.4695998191833497, "lr": 0.0003114715461690636, "grad_norm": 0.7236227989196777, "gamma": 0.30000001192092896, "gumbel_tau": 0.7187939286231995, "xi_alphas": [9.999482699640794e-07, 0.49269792437553406, 0.6505340337753296, 0.7906531095504761, 0.8918806314468384, 0.9453259706497192, 0.9700227975845337, 0.9876937866210938], "elapsed_s": 3103.0709116458893} {"step": 3600, "val_loss": 2.597019702196121, "val_ppl": 13.423671820385946} {"step": 3650, "train_loss": 2.451151142120361, "lr": 0.000306450316717276, "grad_norm": 0.6585278511047363, "gamma": 0.30000001192092896, "gumbel_tau": 0.7117637991905212, "xi_alphas": [1.0001753025790094e-06, 0.49238234758377075, 0.6500630974769592, 0.7904057502746582, 0.8917895555496216, 0.9453954100608826, 0.9700389504432678, 0.9876761436462402], "elapsed_s": 3151.748543739319} {"step": 3700, "train_loss": 2.4797114372253417, "lr": 0.0003014049735855547, "grad_norm": 0.6721935868263245, "gamma": 0.30000001192092896, "gumbel_tau": 0.7047336101531982, "xi_alphas": [9.997633014791063e-07, 0.49250170588493347, 0.6495424509048462, 0.7903251051902771, 0.8915826678276062, 0.9452896118164062, 0.9698702096939087, 0.9875757694244385], "elapsed_s": 3193.4859454631805} {"step": 3750, "train_loss": 2.4741409254074096, "lr": 0.000296337671975084, "grad_norm": 0.8522801399230957, "gamma": 0.30000001192092896, "gumbel_tau": 0.69770348072052, "xi_alphas": [9.999769190471852e-07, 0.4924010932445526, 0.6493362188339233, 0.7902392745018005, 0.8913946747779846, 0.9451977014541626, 0.9698605537414551, 0.9875360131263733], "elapsed_s": 3235.4786739349365} {"step": 3800, "train_loss": 2.4388575315475465, "lr": 0.0002912505764669728, "grad_norm": 0.6468697190284729, "gamma": 0.30000001192092896, "gumbel_tau": 0.6906733512878418, "xi_alphas": [9.999120038628462e-07, 0.49194738268852234, 0.6487563848495483, 0.7900910973548889, 0.8912261128425598, 0.9451736211776733, 0.9698055982589722, 0.987498939037323], "elapsed_s": 3277.4065248966217} {"step": 3850, "train_loss": 2.446234312057495, "lr": 0.0002861458600976186, "grad_norm": 0.6845741271972656, "gamma": 0.30000001192092896, "gumbel_tau": 0.6836431622505188, "xi_alphas": [9.996850849347538e-07, 0.4918505847454071, 0.6481951475143433, 0.7898369431495667, 0.8911628127098083, 0.9451137185096741, 0.969788670539856, 0.987476646900177], "elapsed_s": 3319.179512023926} {"step": 3900, "train_loss": 2.43896906375885, "lr": 0.0002810257034304591, "grad_norm": 0.6733976602554321, "gamma": 0.30000001192092896, "gumbel_tau": 0.6766130328178406, "xi_alphas": [9.99923486233456e-07, 0.4919295012950897, 0.6483253836631775, 0.7900593280792236, 0.8912780284881592, 0.945188581943512, 0.9697652459144592, 0.9874184131622314], "elapsed_s": 3361.0606870651245} {"step": 3950, "train_loss": 2.439266200065613, "lr": 0.0002758922936245088, "grad_norm": 0.670540988445282, "gamma": 0.30000001192092896, "gumbel_tau": 0.6695829033851624, "xi_alphas": [9.996288099500816e-07, 0.49184688925743103, 0.6481252908706665, 0.7898747324943542, 0.8910816311836243, 0.945048451423645, 0.9696606397628784, 0.9873605370521545], "elapsed_s": 3403.2862334251404} {"step": 4000, "train_loss": 2.4230832624435426, "lr": 0.0002707478235000782, "grad_norm": 0.6456267833709717, "gamma": 0.30000001192092896, "gumbel_tau": 0.6625527143478394, "xi_alphas": [9.998309451475507e-07, 0.4920349717140198, 0.6479468941688538, 0.7897244095802307, 0.8909928798675537, 0.9450315833091736, 0.9696220755577087, 0.9873425960540771], "elapsed_s": 3445.06453537941} {"step": 4000, "val_loss": 2.6041493833065035, "val_ppl": 13.519720310484608} {"step": 4050, "train_loss": 2.4236014556884764, "lr": 0.0002655944906020745, "grad_norm": 0.6907345652580261, "gamma": 0.30000001192092896, "gumbel_tau": 0.6555225849151611, "xi_alphas": [1.0000245538321906e-06, 0.4916459619998932, 0.6476824879646301, 0.7896439433097839, 0.8909897804260254, 0.9450522661209106, 0.9695746898651123, 0.9873021841049194], "elapsed_s": 3493.60072183609} {"step": 4100, "train_loss": 2.4245009851455688, "lr": 0.000260434496261284, "grad_norm": 0.672863781452179, "gamma": 0.30000001192092896, "gumbel_tau": 0.6484923958778381, "xi_alphas": [1.0004547448261292e-06, 0.4919453561306, 0.6479862332344055, 0.7895491123199463, 0.8908568024635315, 0.945020854473114, 0.9695667624473572, 0.9872602224349976], "elapsed_s": 3535.6086280345917} {"step": 4150, "train_loss": 2.410296196937561, "lr": 0.0002552700446540379, "grad_norm": 0.6509820818901062, "gamma": 0.30000001192092896, "gumbel_tau": 0.6414622664451599, "xi_alphas": [9.997318102250574e-07, 0.49155405163764954, 0.6477314829826355, 0.7894828915596008, 0.8906516432762146, 0.9449626207351685, 0.9694687724113464, 0.9871930480003357], "elapsed_s": 3577.5937049388885} {"step": 4200, "train_loss": 2.4084894371032717, "lr": 0.0002501033418606619, "grad_norm": 0.6964285373687744, "gamma": 0.30000001192092896, "gumbel_tau": 0.6344321370124817, "xi_alphas": [1.0001609780374565e-06, 0.4915933609008789, 0.6476816534996033, 0.7894876599311829, 0.8903680443763733, 0.9449225068092346, 0.969439685344696, 0.9871383309364319], "elapsed_s": 3619.355444431305} {"step": 4250, "train_loss": 2.4222273063659667, "lr": 0.0002449365949231128, "grad_norm": 0.6520659923553467, "gamma": 0.30000001192092896, "gumbel_tau": 0.6274019479751587, "xi_alphas": [9.999959047490847e-07, 0.4914032220840454, 0.6473397612571716, 0.7894054055213928, 0.8903750777244568, 0.9449379444122314, 0.9694133400917053, 0.98710036277771], "elapsed_s": 3661.5794949531555} {"step": 4300, "train_loss": 2.4163512706756594, "lr": 0.0002397720109022044, "grad_norm": 0.6495473384857178, "gamma": 0.30000001192092896, "gumbel_tau": 0.6203718185424805, "xi_alphas": [1.0004843034039368e-06, 0.4914538562297821, 0.6471801400184631, 0.7892317175865173, 0.8902222514152527, 0.9449529051780701, 0.9693812131881714, 0.9870544075965881], "elapsed_s": 3703.5550191402435} {"step": 4350, "train_loss": 2.4112531757354736, "lr": 0.00023461179593482493, "grad_norm": 0.6522487998008728, "gamma": 0.30000001192092896, "gumbel_tau": 0.6133416891098022, "xi_alphas": [9.999320127462852e-07, 0.49139901995658875, 0.6469477415084839, 0.7892934679985046, 0.8902080059051514, 0.9449576139450073, 0.9693470001220703, 0.9870110750198364], "elapsed_s": 3745.2250764369965} {"step": 4400, "train_loss": 2.3894427824020386, "lr": 0.00022945815429154993, "grad_norm": 0.6633806824684143, "gamma": 0.30000001192092896, "gumbel_tau": 0.6063115000724792, "xi_alphas": [9.998996119975345e-07, 0.4911205470561981, 0.6468325853347778, 0.7891021370887756, 0.8901335597038269, 0.9449583888053894, 0.9693183302879333, 0.9869763255119324], "elapsed_s": 3786.834228038788} {"step": 4400, "val_loss": 2.560677832365036, "val_ppl": 12.944588603583368} {"step": 4450, "train_loss": 2.424029479026794, "lr": 0.00022431328743505147, "grad_norm": 0.6852543354034424, "gamma": 0.30000001192092896, "gumbel_tau": 0.599281370639801, "xi_alphas": [1.0002344197346247e-06, 0.491003155708313, 0.6464546918869019, 0.7889336943626404, 0.8901047706604004, 0.9449513554573059, 0.9692868590354919, 0.9869592785835266], "elapsed_s": 3835.3718452453613} {"step": 4500, "train_loss": 2.395358099937439, "lr": 0.00021917939307970722, "grad_norm": 0.6829625964164734, "gamma": 0.30000001192092896, "gumbel_tau": 0.592251181602478, "xi_alphas": [9.999120038628462e-07, 0.4910127520561218, 0.6460584998130798, 0.7888002395629883, 0.8899078965187073, 0.9449387788772583, 0.9692246317863464, 0.9869197607040405], "elapsed_s": 3877.0966613292694} {"step": 4550, "train_loss": 2.378471641540527, "lr": 0.00021405866425281076, "grad_norm": 0.6616786122322083, "gamma": 0.30000001192092896, "gumbel_tau": 0.5852210521697998, "xi_alphas": [9.999558869822067e-07, 0.4908328950405121, 0.6458994746208191, 0.7887570858001709, 0.8897267580032349, 0.9448848366737366, 0.9690969586372375, 0.9868801832199097], "elapsed_s": 3918.739495754242} {"step": 4600, "train_loss": 2.3819381999969482, "lr": 0.0002089532883577843, "grad_norm": 0.7553941607475281, "gamma": 0.30000001192092896, "gumbel_tau": 0.5781909227371216, "xi_alphas": [1.000151428343088e-06, 0.4904017746448517, 0.645679771900177, 0.7887466549873352, 0.8896822333335876, 0.9448865056037903, 0.9690865278244019, 0.986832857131958], "elapsed_s": 3960.5376613140106} {"step": 4650, "train_loss": 2.3800693368911743, "lr": 0.00020386544623979264, "grad_norm": 0.698921263217926, "gamma": 0.30000001192092896, "gumbel_tau": 0.5711607336997986, "xi_alphas": [1.000003521767212e-06, 0.4904734194278717, 0.6456268429756165, 0.7885639667510986, 0.8893904685974121, 0.9448068141937256, 0.9690258502960205, 0.9867695569992065], "elapsed_s": 4002.334967136383} {"step": 4700, "train_loss": 2.3834323930740355, "lr": 0.00019879731125415983, "grad_norm": 0.6401662230491638, "gamma": 0.30000001192092896, "gumbel_tau": 0.5641306042671204, "xi_alphas": [1.0000007932831068e-06, 0.4903273284435272, 0.6454037427902222, 0.7884418368339539, 0.889251172542572, 0.9447650909423828, 0.9689816832542419, 0.9867319464683533], "elapsed_s": 4044.036673307419} {"step": 4750, "train_loss": 2.3549913644790648, "lr": 0.00019375104833798496, "grad_norm": 0.712135910987854, "gamma": 0.30000001192092896, "gumbel_tau": 0.5571004748344421, "xi_alphas": [9.99755684460979e-07, 0.4903765022754669, 0.6452362537384033, 0.7882261872291565, 0.8890990614891052, 0.9447411298751831, 0.9689207673072815, 0.9866930842399597], "elapsed_s": 4086.2136673927307} {"step": 4800, "train_loss": 2.387789206504822, "lr": 0.000188728813085353, "grad_norm": 0.6925364136695862, "gamma": 0.30000001192092896, "gumbel_tau": 0.5500702857971191, "xi_alphas": [1.0002972885558847e-06, 0.4901595413684845, 0.6450505256652832, 0.78815758228302, 0.8889298439025879, 0.9446451663970947, 0.9688228368759155, 0.9866381883621216], "elapsed_s": 4128.405822277069} {"step": 4800, "val_loss": 2.5163019001483917, "val_ppl": 12.382719353821079} {"step": 4850, "train_loss": 2.3470691347122195, "lr": 0.00018373275082653808, "grad_norm": 0.6916469931602478, "gamma": 0.30000001192092896, "gumbel_tau": 0.5430401563644409, "xi_alphas": [1.0000236443374888e-06, 0.48989778757095337, 0.6449523568153381, 0.7882618308067322, 0.888922393321991, 0.9446659088134766, 0.968815267086029, 0.9866307377815247], "elapsed_s": 4177.51188826561} {"step": 4900, "train_loss": 2.3545164346694945, "lr": 0.00017876499571159, "grad_norm": 0.6822453737258911, "gamma": 0.30000001192092896, "gumbel_tau": 0.5360100269317627, "xi_alphas": [9.99920644062513e-07, 0.48974138498306274, 0.6448699831962585, 0.7883366942405701, 0.8889634013175964, 0.9447195529937744, 0.9688271284103394, 0.9866119623184204], "elapsed_s": 4219.5659120082855} {"step": 4950, "train_loss": 2.3478931093215945, "lr": 0.0001738276697986973, "grad_norm": 0.7245162725448608, "gamma": 0.30000001192092896, "gumbel_tau": 0.5289798378944397, "xi_alphas": [9.999864687415538e-07, 0.4895211160182953, 0.6445471048355103, 0.7883113026618958, 0.8888471722602844, 0.9447153806686401, 0.9687849283218384, 0.9865647554397583], "elapsed_s": 4261.681570529938} {"step": 5000, "train_loss": 2.359030385017395, "lr": 0.00016892288214771488, "grad_norm": 0.7242638468742371, "gamma": 0.30000001192092896, "gumbel_tau": 0.5219497084617615, "xi_alphas": [1.0000875363402884e-06, 0.4893728494644165, 0.6446858048439026, 0.7884078025817871, 0.8887984752655029, 0.9446505308151245, 0.968695342540741, 0.9865109920501709], "elapsed_s": 4304.084185600281} {"step": 5050, "train_loss": 2.342477254867554, "lr": 0.0001640527279192449, "grad_norm": 0.6682330369949341, "gamma": 0.30000001192092896, "gumbel_tau": 0.5149195194244385, "xi_alphas": [9.99870053419727e-07, 0.48940789699554443, 0.6445167660713196, 0.7883467674255371, 0.8887749314308167, 0.9446442127227783, 0.9686791300773621, 0.9864792227745056], "elapsed_s": 4346.106055736542} {"step": 5100, "train_loss": 2.343639736175537, "lr": 0.00015921928747965412, "grad_norm": 0.6734141111373901, "gamma": 0.30000001192092896, "gumbel_tau": 0.5078893899917603, "xi_alphas": [9.996479093388189e-07, 0.4892038404941559, 0.6444762945175171, 0.7884150743484497, 0.8887301087379456, 0.9446228742599487, 0.9686183929443359, 0.986435055732727], "elapsed_s": 4388.145528316498} {"step": 5150, "train_loss": 2.3250454473495483, "lr": 0.00015442462551241122, "grad_norm": 0.6765837073326111, "gamma": 0.30000001192092896, "gumbel_tau": 0.500859260559082, "xi_alphas": [9.99892904474109e-07, 0.48900461196899414, 0.6442641615867615, 0.7883188128471375, 0.8887022137641907, 0.9446041584014893, 0.9685821533203125, 0.9864033460617065], "elapsed_s": 4430.383312702179} {"step": 5200, "train_loss": 2.3407214212417604, "lr": 0.00014967079013612392, "grad_norm": 0.6737232804298401, "gamma": 0.30000001192092896, "gumbel_tau": 0.4938291013240814, "xi_alphas": [1.0000169368140632e-06, 0.48872849345207214, 0.6441112160682678, 0.7882335186004639, 0.8886386156082153, 0.944606363773346, 0.9685529470443726, 0.9863429665565491], "elapsed_s": 4472.547676086426} {"step": 5200, "val_loss": 2.534871429204941, "val_ppl": 12.614808847765126} {"step": 5250, "train_loss": 2.319200191497803, "lr": 0.00014495981202965072, "grad_norm": 0.6924622058868408, "gamma": 0.30000001192092896, "gumbel_tau": 0.4867989420890808, "xi_alphas": [9.998567520597135e-07, 0.48874202370643616, 0.6442390084266663, 0.7882795333862305, 0.8885369300842285, 0.9445995688438416, 0.9685413241386414, 0.9863103628158569], "elapsed_s": 4521.316903114319} {"step": 5300, "train_loss": 2.3443714332580567, "lr": 0.00014029370356466353, "grad_norm": 0.7009459733963013, "gamma": 0.30000001192092896, "gumbel_tau": 0.4797687828540802, "xi_alphas": [9.999482699640794e-07, 0.48810121417045593, 0.6438681483268738, 0.7882598042488098, 0.8885720372200012, 0.9446074366569519, 0.9684737920761108, 0.9862759113311768], "elapsed_s": 4563.320355653763} {"step": 5350, "train_loss": 2.3199344968795774, "lr": 0.00013567445794603018, "grad_norm": 0.7101821899414062, "gamma": 0.30000001192092896, "gumbel_tau": 0.4727386236190796, "xi_alphas": [1.0000388783737435e-06, 0.48777517676353455, 0.6437037587165833, 0.7882404923439026, 0.8884733319282532, 0.9445662498474121, 0.9684044718742371, 0.9862382411956787], "elapsed_s": 4605.552248477936} {"step": 5400, "train_loss": 2.3419530105590822, "lr": 0.00013110404836038365, "grad_norm": 0.6829885840415955, "gamma": 0.30000001192092896, "gumbel_tau": 0.46570849418640137, "xi_alphas": [1.0003021770899068e-06, 0.4875452518463135, 0.643680989742279, 0.7883073091506958, 0.8884539604187012, 0.9445592761039734, 0.9683891534805298, 0.986208438873291], "elapsed_s": 4647.382637023926} {"step": 5450, "train_loss": 2.34338219165802, "lr": 0.00012658442713324347, "grad_norm": 0.6700674295425415, "gamma": 0.30000001192092896, "gumbel_tau": 0.45867833495140076, "xi_alphas": [1.00068473329884e-06, 0.4873881936073303, 0.6435683965682983, 0.7883201241493225, 0.8884710669517517, 0.9445716738700867, 0.9683972001075745, 0.9861932396888733], "elapsed_s": 4689.33279132843} {"step": 5500, "train_loss": 2.32876437664032, "lr": 0.00012211752489504763, "grad_norm": 0.6905581951141357, "gamma": 0.30000001192092896, "gumbel_tau": 0.45164817571640015, "xi_alphas": [1.0000961765399552e-06, 0.4870302975177765, 0.643389880657196, 0.7883015275001526, 0.8884176015853882, 0.9445722103118896, 0.9683950543403625, 0.9861682057380676], "elapsed_s": 4731.423553228378} {"step": 5550, "train_loss": 2.308544545173645, "lr": 0.00011770524975645239, "grad_norm": 0.6735964417457581, "gamma": 0.30000001192092896, "gumbel_tau": 0.44461801648139954, "xi_alphas": [9.99863459583139e-07, 0.4869309067726135, 0.6432676911354065, 0.7882128357887268, 0.8883188366889954, 0.944580614566803, 0.9683910608291626, 0.9861676096916199], "elapsed_s": 4773.452519416809} {"step": 5600, "train_loss": 2.298291506767273, "lr": 0.00011334948649325089, "grad_norm": 0.6831477880477905, "gamma": 0.30000001192092896, "gumbel_tau": 0.4375878870487213, "xi_alphas": [9.99946337287838e-07, 0.4870990812778473, 0.6433655619621277, 0.788322925567627, 0.8883983492851257, 0.9446123242378235, 0.9683867692947388, 0.9861477613449097], "elapsed_s": 4815.324583530426} {"step": 5600, "val_loss": 2.5243636786937715, "val_ppl": 12.482949569092167} {"step": 5650, "train_loss": 2.3194500064849852, "lr": 0.0001090520957412606, "grad_norm": 0.7233587503433228, "gamma": 0.30000001192092896, "gumbel_tau": 0.4305577278137207, "xi_alphas": [9.99960661829391e-07, 0.4868971109390259, 0.6431950926780701, 0.7882030010223389, 0.8882507681846619, 0.944617509841919, 0.9683891534805298, 0.986114501953125], "elapsed_s": 4864.190574169159} {"step": 5700, "train_loss": 2.316688714027405, "lr": 0.00010481491320152237, "grad_norm": 0.7040591835975647, "gamma": 0.30000001192092896, "gumbel_tau": 0.4235275685787201, "xi_alphas": [1.0000188694903045e-06, 0.48642978072166443, 0.6429064869880676, 0.7881137132644653, 0.8882406949996948, 0.9446256160736084, 0.9683458805084229, 0.9860958456993103], "elapsed_s": 4906.219133138657} {"step": 5750, "train_loss": 2.3110457468032837, "lr": 0.00010063974885614999, "grad_norm": 0.6701304316520691, "gamma": 0.30000001192092896, "gumbel_tau": 0.4164974093437195, "xi_alphas": [9.998958603318897e-07, 0.4861501157283783, 0.6427216529846191, 0.7880808115005493, 0.888205885887146, 0.9446287155151367, 0.9683340191841125, 0.9860849380493164], "elapsed_s": 4948.0777707099915} {"step": 5800, "train_loss": 2.283889455795288, "lr": 9.652838619516696e-05, "grad_norm": 0.721359372138977, "gamma": 0.30000001192092896, "gumbel_tau": 0.40946727991104126, "xi_alphas": [9.999969279306242e-07, 0.4859910011291504, 0.642641544342041, 0.7880734205245972, 0.8882020115852356, 0.9446081519126892, 0.9683001637458801, 0.9860534071922302], "elapsed_s": 4990.158578872681} {"step": 5850, "train_loss": 2.2778955698013306, "lr": 9.248258145465977e-05, "grad_norm": 0.6998459696769714, "gamma": 0.30000001192092896, "gumbel_tau": 0.40243712067604065, "xi_alphas": [1.0000741212934372e-06, 0.4861452877521515, 0.6426007747650146, 0.7880829572677612, 0.8881513476371765, 0.9445963501930237, 0.9682609438896179, 0.9860242009162903], "elapsed_s": 5032.238795757294} {"step": 5900, "train_loss": 2.288577423095703, "lr": 8.850406286657198e-05, "grad_norm": 0.6609171032905579, "gamma": 0.30000001192092896, "gumbel_tau": 0.39540696144104004, "xi_alphas": [1.0001475629906054e-06, 0.4859449863433838, 0.642375111579895, 0.7880030870437622, 0.8880743980407715, 0.9445984959602356, 0.9682526588439941, 0.9860055446624756], "elapsed_s": 5074.08855175972} {"step": 5950, "train_loss": 2.293257837295532, "lr": 8.459452992046224e-05, "grad_norm": 0.7034103870391846, "gamma": 0.30000001192092896, "gumbel_tau": 0.38837680220603943, "xi_alphas": [9.998825589718763e-07, 0.48564228415489197, 0.6421820521354675, 0.7879508137702942, 0.8880448937416077, 0.9445952773094177, 0.9682421684265137, 0.9859843850135803], "elapsed_s": 5115.975106239319} {"step": 6000, "train_loss": 2.2848797464370727, "lr": 8.075565263753929e-05, "grad_norm": 0.6808946132659912, "gamma": 0.30000001192092896, "gumbel_tau": 0.3813466727733612, "xi_alphas": [9.999330359278247e-07, 0.48560696840286255, 0.6421452760696411, 0.7879716753959656, 0.888046145439148, 0.9446075558662415, 0.9682236909866333, 0.9859678745269775], "elapsed_s": 5158.067709207535} {"step": 6000, "val_loss": 2.493044066429138, "val_ppl": 12.098047384866735} {"step": 6050, "train_loss": 2.295106348991394, "lr": 7.698907085728466e-05, "grad_norm": 0.6989579796791077, "gamma": 0.30000001192092896, "gumbel_tau": 0.3743165135383606, "xi_alphas": [9.99988401417795e-07, 0.4854675829410553, 0.6421628594398499, 0.7880067229270935, 0.8879995942115784, 0.9446038007736206, 0.9682093858718872, 0.9859528541564941], "elapsed_s": 5206.949064254761} {"step": 6100, "train_loss": 2.279239158630371, "lr": 7.32963935369687e-05, "grad_norm": 0.690148651599884, "gamma": 0.30000001192092896, "gumbel_tau": 0.36728635430336, "xi_alphas": [9.999797612181283e-07, 0.48518702387809753, 0.6420148015022278, 0.7879830002784729, 0.8879450559616089, 0.9446011781692505, 0.96819669008255, 0.985934853553772], "elapsed_s": 5248.817698717117} {"step": 6150, "train_loss": 2.276352467536926, "lr": 6.967919806435793e-05, "grad_norm": 0.7049864530563354, "gamma": 0.30000001192092896, "gumbel_tau": 0.3602561950683594, "xi_alphas": [1.0000388783737435e-06, 0.4849693477153778, 0.6419057846069336, 0.7880445122718811, 0.8879168033599854, 0.9446046352386475, 0.9681994915008545, 0.9859274625778198], "elapsed_s": 5290.902166604996} {"step": 6200, "train_loss": 2.269377131462097, "lr": 6.61390295839077e-05, "grad_norm": 0.6619830131530762, "gamma": 0.30000001192092896, "gumbel_tau": 0.35322606563568115, "xi_alphas": [1.000003521767212e-06, 0.4848330020904541, 0.6418909430503845, 0.7880738377571106, 0.8878930807113647, 0.9445980787277222, 0.968187153339386, 0.9859076142311096], "elapsed_s": 5332.94103717804} {"step": 6250, "train_loss": 2.259030261039734, "lr": 6.267740033672855e-05, "grad_norm": 0.712298572063446, "gamma": 0.30000001192092896, "gumbel_tau": 0.34619590640068054, "xi_alphas": [1.0000284191846731e-06, 0.4847273528575897, 0.6417533755302429, 0.7880222797393799, 0.8878737092018127, 0.9445849061012268, 0.968169629573822, 0.9858931303024292], "elapsed_s": 5374.77995800972} {"step": 6300, "train_loss": 2.2780947303771972, "lr": 5.929578901460789e-05, "grad_norm": 0.6871823668479919, "gamma": 0.30000001192092896, "gumbel_tau": 0.33916574716567993, "xi_alphas": [9.999854455600143e-07, 0.48458659648895264, 0.6416367888450623, 0.7880344390869141, 0.887839138507843, 0.9445793628692627, 0.9681412577629089, 0.9858747124671936], "elapsed_s": 5416.826090812683} {"step": 6350, "train_loss": 2.2732403469085694, "lr": 5.599564012836189e-05, "grad_norm": 0.6682626605033875, "gamma": 0.30000001192092896, "gumbel_tau": 0.3321356177330017, "xi_alphas": [1.0000541124099982e-06, 0.48444393277168274, 0.6415382623672485, 0.7880172729492188, 0.887796938419342, 0.9445658326148987, 0.9681099653244019, 0.9858567714691162], "elapsed_s": 5458.886948347092} {"step": 6400, "train_loss": 2.2879973602294923, "lr": 5.2778363390789426e-05, "grad_norm": 0.6924959421157837, "gamma": 0.30000001192092896, "gumbel_tau": 0.3251054584980011, "xi_alphas": [9.999959047490847e-07, 0.48431193828582764, 0.6414474844932556, 0.7880771160125732, 0.8878232836723328, 0.9445894360542297, 0.9681149125099182, 0.9858432412147522], "elapsed_s": 5500.845377445221} {"step": 6400, "val_loss": 2.4804342806339266, "val_ppl": 11.946451403952656} {"step": 6450, "train_loss": 2.2691568088531495, "lr": 4.964533311448991e-05, "grad_norm": 0.6422522664070129, "gamma": 0.30000001192092896, "gumbel_tau": 0.3180752992630005, "xi_alphas": [1.0001772352552507e-06, 0.4841731786727905, 0.6413607001304626, 0.7880324125289917, 0.8878104090690613, 0.9445914626121521, 0.9681161046028137, 0.9858353734016418], "elapsed_s": 5549.518787145615} {"step": 6500, "train_loss": 2.263563594818115, "lr": 4.659788762480327e-05, "grad_norm": 0.6872193217277527, "gamma": 0.30000001192092896, "gumbel_tau": 0.3110451400279999, "xi_alphas": [1.0004080195358256e-06, 0.48409849405288696, 0.6412529945373535, 0.7879748344421387, 0.8877613544464111, 0.9445912837982178, 0.9681140184402466, 0.9858296513557434], "elapsed_s": 5591.424906015396} {"step": 6550, "train_loss": 2.2809102010726927, "lr": 4.3637328688122065e-05, "grad_norm": 0.7873610258102417, "gamma": 0.30000001192092896, "gumbel_tau": 0.30401501059532166, "xi_alphas": [1.000432803266449e-06, 0.483994722366333, 0.6411702632904053, 0.7879764437675476, 0.8877649307250977, 0.9445748925209045, 0.968097448348999, 0.9858222603797913], "elapsed_s": 5633.5728623867035} {"step": 6600, "train_loss": 2.269444007873535, "lr": 4.0764920955820656e-05, "grad_norm": 0.7070837616920471, "gamma": 0.30000001192092896, "gumbel_tau": 0.29698485136032104, "xi_alphas": [1.0003002444136655e-06, 0.4838281273841858, 0.6410658359527588, 0.7879608869552612, 0.887738049030304, 0.9445841312408447, 0.9680951237678528, 0.9858148097991943], "elapsed_s": 5675.57651758194} {"step": 6650, "train_loss": 2.2563738346099855, "lr": 3.798189142403882e-05, "grad_norm": 0.7014092206954956, "gamma": 0.30000001192092896, "gumbel_tau": 0.28995469212532043, "xi_alphas": [1.0001457440012018e-06, 0.4837263226509094, 0.6410113573074341, 0.787990152835846, 0.8877161741256714, 0.9445787072181702, 0.9680882096290588, 0.9858086109161377], "elapsed_s": 5717.560221195221} {"step": 6700, "train_loss": 2.26067093372345, "lr": 3.5289428909549865e-05, "grad_norm": 0.6757408380508423, "gamma": 0.30000001192092896, "gumbel_tau": 0.2829245328903198, "xi_alphas": [1.0000302381740767e-06, 0.483623206615448, 0.6409977674484253, 0.7880470752716064, 0.8877238035202026, 0.9445897936820984, 0.9680849313735962, 0.9857965111732483], "elapsed_s": 5759.614332675934} {"step": 6750, "train_loss": 2.283302025794983, "lr": 3.268868354193818e-05, "grad_norm": 0.698693037033081, "gamma": 0.30000001192092896, "gumbel_tau": 0.2758944034576416, "xi_alphas": [9.99997837425326e-07, 0.4834996163845062, 0.6409472823143005, 0.7880625128746033, 0.8877245187759399, 0.9445866942405701, 0.9680714011192322, 0.9857865571975708], "elapsed_s": 5801.7009127140045} {"step": 6800, "train_loss": 2.2558958387374877, "lr": 3.018076627230268e-05, "grad_norm": 0.712158739566803, "gamma": 0.30000001192092896, "gumbel_tau": 0.268864244222641, "xi_alphas": [1.0000951533584157e-06, 0.4834335148334503, 0.6408824920654297, 0.7880337834358215, 0.8877047300338745, 0.9445801973342896, 0.9680601358413696, 0.9857820272445679], "elapsed_s": 5843.615223646164} {"step": 6800, "val_loss": 2.4903015434741973, "val_ppl": 12.064913667961124} {"step": 6850, "train_loss": 2.279196619987488, "lr": 2.776674839869539e-05, "grad_norm": 0.663815438747406, "gamma": 0.30000001192092896, "gumbel_tau": 0.2618340849876404, "xi_alphas": [1.0000597967518843e-06, 0.48349282145500183, 0.6408961415290833, 0.7880361080169678, 0.8876832127571106, 0.9445801973342896, 0.9680591225624084, 0.9857768416404724], "elapsed_s": 5892.299204349518} {"step": 6900, "train_loss": 2.2908474159240724, "lr": 2.544766110849911e-05, "grad_norm": 0.6808880567550659, "gamma": 0.30000001192092896, "gumbel_tau": 0.25480392575263977, "xi_alphas": [9.99988401417795e-07, 0.4833686053752899, 0.6408211588859558, 0.7880209684371948, 0.8876679539680481, 0.9445807337760925, 0.9680624008178711, 0.9857692122459412], "elapsed_s": 5934.551500558853} {"step": 6950, "train_loss": 2.282991032600403, "lr": 2.322449503793897e-05, "grad_norm": 0.6912277340888977, "gamma": 0.30000001192092896, "gumbel_tau": 0.24777378141880035, "xi_alphas": [1.0000865131587489e-06, 0.4833112955093384, 0.6407893300056458, 0.788041889667511, 0.8876553773880005, 0.9445803165435791, 0.9680556654930115, 0.985765278339386], "elapsed_s": 5976.469850063324} {"step": 7000, "train_loss": 2.2720803689956663, "lr": 2.1098199848915534e-05, "grad_norm": 0.6867806911468506, "gamma": 0.30000001192092896, "gumbel_tau": 0.24074363708496094, "xi_alphas": [1.0003108172895736e-06, 0.48326653242111206, 0.640739917755127, 0.7880527377128601, 0.8876596093177795, 0.9445871114730835, 0.9680513143539429, 0.9857618808746338], "elapsed_s": 6018.389665842056} {"step": 7050, "train_loss": 2.276384701728821, "lr": 1.9069683823341373e-05, "grad_norm": 0.7064149379730225, "gamma": 0.30000001192092896, "gumbel_tau": 0.23371347784996033, "xi_alphas": [1.0002715953305596e-06, 0.48319733142852783, 0.6406780481338501, 0.7880455851554871, 0.8876548409461975, 0.944589376449585, 0.9680496454238892, 0.9857565760612488], "elapsed_s": 6060.439019680023} {"step": 7100, "train_loss": 2.2738990497589113, "lr": 1.7139813475153476e-05, "grad_norm": 0.6643947958946228, "gamma": 0.30000001192092896, "gumbel_tau": 0.2266833335161209, "xi_alphas": [1.0002611361414893e-06, 0.48315152525901794, 0.6406316161155701, 0.7880399227142334, 0.8876577615737915, 0.9445977210998535, 0.9680500626564026, 0.985751211643219], "elapsed_s": 6102.44086933136} {"step": 7150, "train_loss": 2.2486943435668945, "lr": 1.5309413180167487e-05, "grad_norm": 0.6895858645439148, "gamma": 0.30000001192092896, "gumbel_tau": 0.2196531742811203, "xi_alphas": [1.0002009958043345e-06, 0.48310622572898865, 0.640618622303009, 0.7880449891090393, 0.8876468539237976, 0.9445942640304565, 0.9680473208427429, 0.9857481122016907], "elapsed_s": 6144.306035280228} {"step": 7200, "train_loss": 2.2818832063674925, "lr": 1.3579264823932152e-05, "grad_norm": 0.6781639456748962, "gamma": 0.30000001192092896, "gumbel_tau": 0.21262302994728088, "xi_alphas": [1.000244878923695e-06, 0.4830628037452698, 0.6406031250953674, 0.788067638874054, 0.8876551985740662, 0.9445973038673401, 0.9680450558662415, 0.9857438206672668], "elapsed_s": 6186.314485549927} {"step": 7200, "val_loss": 2.489305871725082, "val_ppl": 12.052906952631966} {"step": 7250, "train_loss": 2.2733489274978638, "lr": 1.1950107467734089e-05, "grad_norm": 0.6713660955429077, "gamma": 0.30000001192092896, "gumbel_tau": 0.20559287071228027, "xi_alphas": [1.0001704140449874e-06, 0.4830351173877716, 0.6405673623085022, 0.7880746126174927, 0.8876688480377197, 0.944604754447937, 0.9680439233779907, 0.9857409596443176], "elapsed_s": 6235.11213350296} {"step": 7300, "train_loss": 2.253528189659119, "lr": 1.0422637032895421e-05, "grad_norm": 0.666466474533081, "gamma": 0.30000001192092896, "gumbel_tau": 0.19856272637844086, "xi_alphas": [1.0001524515246274e-06, 0.482985258102417, 0.6405250430107117, 0.7880667448043823, 0.887671947479248, 0.9446056485176086, 0.9680403470993042, 0.9857388734817505], "elapsed_s": 6276.920036554337} {"step": 7350, "train_loss": 2.240917272567749, "lr": 8.997506003499855e-06, "grad_norm": 0.6927796006202698, "gamma": 0.30000001192092896, "gumbel_tau": 0.19153256714344025, "xi_alphas": [1.0001829195971368e-06, 0.4829595685005188, 0.6405134797096252, 0.7880688905715942, 0.8876718282699585, 0.944607138633728, 0.9680389165878296, 0.9857366681098938], "elapsed_s": 6319.039708137512} {"step": 7400, "train_loss": 2.262943081855774, "lr": 7.675323147673053e-06, "grad_norm": 0.6954596638679504, "gamma": 0.30000001192092896, "gumbel_tau": 0.18450242280960083, "xi_alphas": [1.000122779259982e-06, 0.4829109311103821, 0.6404878497123718, 0.7880603671073914, 0.8876692056655884, 0.9446076154708862, 0.9680348038673401, 0.9857333898544312], "elapsed_s": 6361.056537151337} {"step": 7450, "train_loss": 2.238652057647705, "lr": 6.456653257537665e-06, "grad_norm": 0.675823986530304, "gamma": 0.30000001192092896, "gumbel_tau": 0.17747226357460022, "xi_alphas": [1.0000903785112314e-06, 0.4828631579875946, 0.6404753923416138, 0.7880627512931824, 0.8876636624336243, 0.9446086287498474, 0.9680339694023132, 0.9857315421104431], "elapsed_s": 6402.967420101166} {"step": 7500, "train_loss": 2.2366843318939207, "lr": 5.342016907952857e-06, "grad_norm": 0.694078803062439, "gamma": 0.30000001192092896, "gumbel_tau": 0.1704421192407608, "xi_alphas": [1.000094243863714e-06, 0.48284316062927246, 0.6404492855072021, 0.788048505783081, 0.8876590728759766, 0.9446082711219788, 0.9680339097976685, 0.9857307076454163], "elapsed_s": 6444.865985155106} {"step": 7550, "train_loss": 2.260650157928467, "lr": 4.331890234142266e-06, "grad_norm": 0.7213887572288513, "gamma": 0.30000001192092896, "gumbel_tau": 0.1634119600057602, "xi_alphas": [1.000117094918096e-06, 0.482835590839386, 0.640442430973053, 0.7880492806434631, 0.8876566290855408, 0.9446073174476624, 0.9680321216583252, 0.9857293367385864], "elapsed_s": 6486.929710865021} {"step": 7600, "train_loss": 2.263111515045166, "lr": 3.4267047283051135e-06, "grad_norm": 0.6892427206039429, "gamma": 0.30000001192092896, "gumbel_tau": 0.15638181567192078, "xi_alphas": [1.0001123200709117e-06, 0.48282355070114136, 0.6404368281364441, 0.7880528569221497, 0.8876557946205139, 0.9446080923080444, 0.9680312275886536, 0.9857282042503357], "elapsed_s": 6528.976622343063} {"step": 7600, "val_loss": 2.45281662940979, "val_ppl": 11.62103280809388} {"step": 7650, "train_loss": 2.2420048093795777, "lr": 2.626847055297088e-06, "grad_norm": 0.7101024985313416, "gamma": 0.30000001192092896, "gumbel_tau": 0.14935165643692017, "xi_alphas": [1.0001151622418547e-06, 0.4828246533870697, 0.6404345035552979, 0.7880511283874512, 0.8876543641090393, 0.9446079730987549, 0.9680299758911133, 0.9857273697853088], "elapsed_s": 6577.600989103317} {"step": 7700, "train_loss": 2.2687869834899903, "lr": 1.932658887460165e-06, "grad_norm": 0.6765475273132324, "gamma": 0.30000001192092896, "gumbel_tau": 0.14232151210308075, "xi_alphas": [1.0000923111874727e-06, 0.482817143201828, 0.6404293179512024, 0.7880526781082153, 0.8876530528068542, 0.9446080923080444, 0.9680292010307312, 0.9857266545295715], "elapsed_s": 6619.601394176483} {"step": 7750, "train_loss": 2.2480974435806274, "lr": 1.3444367586716854e-06, "grad_norm": 0.6868705749511719, "gamma": 0.30000001192092896, "gumbel_tau": 0.13529136776924133, "xi_alphas": [1.0000779866459197e-06, 0.48281368613243103, 0.6404280662536621, 0.7880550622940063, 0.8876540660858154, 0.9446092247962952, 0.9680294394493103, 0.9857262372970581], "elapsed_s": 6661.6881539821625} {"step": 7800, "train_loss": 2.270326681137085, "lr": 8.624319376750456e-07, "grad_norm": 0.6724166870117188, "gamma": 0.30000001192092896, "gumbel_tau": 0.12826120853424072, "xi_alphas": [1.0000741212934372e-06, 0.48281148076057434, 0.6404261589050293, 0.7880539894104004, 0.887653112411499, 0.9446091055870056, 0.9680294394493103, 0.9857261180877686], "elapsed_s": 6703.555640697479} {"step": 7850, "train_loss": 2.285753164291382, "lr": 4.868503207459963e-07, "grad_norm": 0.6728518605232239, "gamma": 0.30000001192092896, "gumbel_tau": 0.12123105674982071, "xi_alphas": [1.0000741212934372e-06, 0.48280858993530273, 0.6404241919517517, 0.7880544662475586, 0.8876540064811707, 0.9446096420288086, 0.9680293202400208, 0.9857261180877686], "elapsed_s": 6745.617515802383} {"step": 7900, "train_loss": 2.2460629749298096, "lr": 2.178523437408597e-07, "grad_norm": 0.695744514465332, "gamma": 0.30000001192092896, "gumbel_tau": 0.1142009049654007, "xi_alphas": [1.0000741212934372e-06, 0.4828071892261505, 0.6404235363006592, 0.7880547642707825, 0.8876540660858154, 0.9446097612380981, 0.9680293202400208, 0.9857261180877686], "elapsed_s": 6787.772879123688} {"step": 7950, "train_loss": 2.270185694694519, "lr": 5.555291356351688e-08, "grad_norm": 0.6778465509414673, "gamma": 0.30000001192092896, "gumbel_tau": 0.10717075318098068, "xi_alphas": [1.0000741212934372e-06, 0.482806921005249, 0.6404229998588562, 0.7880547642707825, 0.8876540660858154, 0.9446097612380981, 0.9680293202400208, 0.9857261180877686], "elapsed_s": 6829.640760183334} {"step": 8000, "train_loss": 2.252153687477112, "lr": 2.1359081253669033e-11, "grad_norm": 0.6538051962852478, "gamma": 0.30000001192092896, "gumbel_tau": 0.10014060139656067, "xi_alphas": [1.0000741212934372e-06, 0.4828068017959595, 0.6404229998588562, 0.7880547642707825, 0.8876540660858154, 0.9446097612380981, 0.9680293202400208, 0.9857261180877686], "elapsed_s": 6871.685479640961} {"step": 8000, "val_loss": 2.48194562792778, "val_ppl": 11.9645202916961}