{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 2505, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005993407252022775, "grad_norm": 57.0, "learning_rate": 4.7058823529411767e-07, "loss": 2.9864707946777345, "step": 5 }, { "epoch": 0.01198681450404555, "grad_norm": 56.0, "learning_rate": 1.0588235294117648e-06, "loss": 3.017852783203125, "step": 10 }, { "epoch": 0.017980221756068324, "grad_norm": 46.5, "learning_rate": 1.647058823529412e-06, "loss": 2.9088718414306642, "step": 15 }, { "epoch": 0.0239736290080911, "grad_norm": 44.5, "learning_rate": 2.235294117647059e-06, "loss": 2.8249109268188475, "step": 20 }, { "epoch": 0.029967036260113874, "grad_norm": 36.5, "learning_rate": 2.823529411764706e-06, "loss": 2.9229591369628904, "step": 25 }, { "epoch": 0.03596044351213665, "grad_norm": 20.125, "learning_rate": 3.411764705882353e-06, "loss": 2.6270092010498045, "step": 30 }, { "epoch": 0.041953850764159424, "grad_norm": 13.1875, "learning_rate": 4e-06, "loss": 2.57705078125, "step": 35 }, { "epoch": 0.0479472580161822, "grad_norm": 7.96875, "learning_rate": 4.588235294117647e-06, "loss": 2.5475204467773436, "step": 40 }, { "epoch": 0.05394066526820498, "grad_norm": 8.375, "learning_rate": 5.176470588235295e-06, "loss": 2.3420473098754884, "step": 45 }, { "epoch": 0.05993407252022775, "grad_norm": 5.65625, "learning_rate": 5.764705882352942e-06, "loss": 2.340384864807129, "step": 50 }, { "epoch": 0.06592747977225052, "grad_norm": 5.03125, "learning_rate": 5.9999778749702535e-06, "loss": 2.175357437133789, "step": 55 }, { "epoch": 0.0719208870242733, "grad_norm": 4.25, "learning_rate": 5.999842667636955e-06, "loss": 2.197158432006836, "step": 60 }, { "epoch": 0.07791429427629608, "grad_norm": 3.84375, "learning_rate": 5.9995845501865744e-06, "loss": 2.1903335571289064, "step": 65 }, { "epoch": 0.08390770152831885, "grad_norm": 13.5625, "learning_rate": 5.999203533194751e-06, "loss": 2.1188888549804688, "step": 70 }, { "epoch": 0.08990110878034162, "grad_norm": 9.0, "learning_rate": 5.998699632272582e-06, "loss": 2.0938800811767577, "step": 75 }, { "epoch": 0.0958945160323644, "grad_norm": 4.25, "learning_rate": 5.998072868065993e-06, "loss": 2.0853302001953127, "step": 80 }, { "epoch": 0.10188792328438717, "grad_norm": 3.015625, "learning_rate": 5.997323266254883e-06, "loss": 1.9871812820434571, "step": 85 }, { "epoch": 0.10788133053640996, "grad_norm": 3.265625, "learning_rate": 5.996450857552079e-06, "loss": 1.9602109909057617, "step": 90 }, { "epoch": 0.11387473778843273, "grad_norm": 3.078125, "learning_rate": 5.995455677702073e-06, "loss": 1.9757930755615234, "step": 95 }, { "epoch": 0.1198681450404555, "grad_norm": 4.125, "learning_rate": 5.994337767479563e-06, "loss": 1.9586530685424806, "step": 100 }, { "epoch": 0.12586155229247828, "grad_norm": 2.953125, "learning_rate": 5.993097172687774e-06, "loss": 1.8350349426269532, "step": 105 }, { "epoch": 0.13185495954450105, "grad_norm": 3.265625, "learning_rate": 5.991733944156591e-06, "loss": 1.9099546432495118, "step": 110 }, { "epoch": 0.13784836679652382, "grad_norm": 3.359375, "learning_rate": 5.990248137740468e-06, "loss": 1.8809522628784179, "step": 115 }, { "epoch": 0.1438417740485466, "grad_norm": 2.578125, "learning_rate": 5.988639814316145e-06, "loss": 1.890812873840332, "step": 120 }, { "epoch": 0.1498351813005694, "grad_norm": 2.921875, "learning_rate": 5.986909039780151e-06, "loss": 1.8602405548095704, "step": 125 }, { "epoch": 0.15582858855259216, "grad_norm": 2.953125, "learning_rate": 5.985055885046105e-06, "loss": 1.8328474044799805, "step": 130 }, { "epoch": 0.16182199580461493, "grad_norm": 4.25, "learning_rate": 5.9830804260418115e-06, "loss": 1.7698371887207032, "step": 135 }, { "epoch": 0.1678154030566377, "grad_norm": 2.75, "learning_rate": 5.980982743706147e-06, "loss": 1.7916770935058595, "step": 140 }, { "epoch": 0.17380881030866047, "grad_norm": 4.375, "learning_rate": 5.978762923985747e-06, "loss": 1.804901695251465, "step": 145 }, { "epoch": 0.17980221756068324, "grad_norm": 2.828125, "learning_rate": 5.976421057831481e-06, "loss": 1.8203529357910155, "step": 150 }, { "epoch": 0.18579562481270603, "grad_norm": 3.78125, "learning_rate": 5.97395724119473e-06, "loss": 1.8243545532226562, "step": 155 }, { "epoch": 0.1917890320647288, "grad_norm": 2.78125, "learning_rate": 5.971371575023451e-06, "loss": 1.8442657470703125, "step": 160 }, { "epoch": 0.19778243931675157, "grad_norm": 3.140625, "learning_rate": 5.968664165258043e-06, "loss": 1.8430421829223633, "step": 165 }, { "epoch": 0.20377584656877434, "grad_norm": 3.078125, "learning_rate": 5.96583512282701e-06, "loss": 1.7948385238647462, "step": 170 }, { "epoch": 0.2097692538207971, "grad_norm": 2.6875, "learning_rate": 5.962884563642406e-06, "loss": 1.8050064086914062, "step": 175 }, { "epoch": 0.2157626610728199, "grad_norm": 2.671875, "learning_rate": 5.959812608595099e-06, "loss": 1.7687402725219727, "step": 180 }, { "epoch": 0.22175606832484268, "grad_norm": 2.796875, "learning_rate": 5.956619383549807e-06, "loss": 1.7481760025024413, "step": 185 }, { "epoch": 0.22774947557686545, "grad_norm": 7.78125, "learning_rate": 5.953305019339946e-06, "loss": 1.6539461135864257, "step": 190 }, { "epoch": 0.23374288282888822, "grad_norm": 2.5625, "learning_rate": 5.949869651762269e-06, "loss": 1.760018539428711, "step": 195 }, { "epoch": 0.239736290080911, "grad_norm": 2.46875, "learning_rate": 5.946313421571302e-06, "loss": 1.7840660095214844, "step": 200 }, { "epoch": 0.24572969733293376, "grad_norm": 2.53125, "learning_rate": 5.942636474473575e-06, "loss": 1.7796796798706054, "step": 205 }, { "epoch": 0.25172310458495656, "grad_norm": 2.65625, "learning_rate": 5.938838961121656e-06, "loss": 1.7450885772705078, "step": 210 }, { "epoch": 0.2577165118369793, "grad_norm": 2.734375, "learning_rate": 5.934921037107974e-06, "loss": 1.7745096206665039, "step": 215 }, { "epoch": 0.2637099190890021, "grad_norm": 6.625, "learning_rate": 5.9308828629584476e-06, "loss": 1.679784393310547, "step": 220 }, { "epoch": 0.2697033263410249, "grad_norm": 3.53125, "learning_rate": 5.926724604125908e-06, "loss": 1.7226064682006836, "step": 225 }, { "epoch": 0.27569673359304764, "grad_norm": 2.90625, "learning_rate": 5.922446430983316e-06, "loss": 1.719569206237793, "step": 230 }, { "epoch": 0.28169014084507044, "grad_norm": 2.421875, "learning_rate": 5.9180485188167845e-06, "loss": 1.7614149093627929, "step": 235 }, { "epoch": 0.2876835480970932, "grad_norm": 2.90625, "learning_rate": 5.913531047818396e-06, "loss": 1.7348569869995116, "step": 240 }, { "epoch": 0.293676955349116, "grad_norm": 2.703125, "learning_rate": 5.908894203078821e-06, "loss": 1.703908920288086, "step": 245 }, { "epoch": 0.2996703626011388, "grad_norm": 2.421875, "learning_rate": 5.904138174579734e-06, "loss": 1.732318115234375, "step": 250 }, { "epoch": 0.3056637698531615, "grad_norm": 2.84375, "learning_rate": 5.899263157186027e-06, "loss": 1.7093519210815429, "step": 255 }, { "epoch": 0.3116571771051843, "grad_norm": 2.71875, "learning_rate": 5.894269350637827e-06, "loss": 1.730145263671875, "step": 260 }, { "epoch": 0.31765058435720706, "grad_norm": 2.6875, "learning_rate": 5.889156959542316e-06, "loss": 1.7165464401245116, "step": 265 }, { "epoch": 0.32364399160922985, "grad_norm": 2.515625, "learning_rate": 5.88392619336534e-06, "loss": 1.7485239028930664, "step": 270 }, { "epoch": 0.3296373988612526, "grad_norm": 3.265625, "learning_rate": 5.878577266422833e-06, "loss": 1.709259033203125, "step": 275 }, { "epoch": 0.3356308061132754, "grad_norm": 5.03125, "learning_rate": 5.873110397872036e-06, "loss": 1.6541948318481445, "step": 280 }, { "epoch": 0.3416242133652982, "grad_norm": 3.375, "learning_rate": 5.867525811702511e-06, "loss": 1.712679100036621, "step": 285 }, { "epoch": 0.34761762061732093, "grad_norm": 2.53125, "learning_rate": 5.861823736726973e-06, "loss": 1.7111431121826173, "step": 290 }, { "epoch": 0.35361102786934373, "grad_norm": 3.078125, "learning_rate": 5.856004406571909e-06, "loss": 1.6408029556274415, "step": 295 }, { "epoch": 0.3596044351213665, "grad_norm": 2.203125, "learning_rate": 5.8500680596680045e-06, "loss": 1.7105199813842773, "step": 300 }, { "epoch": 0.36559784237338927, "grad_norm": 3.6875, "learning_rate": 5.844014939240383e-06, "loss": 1.6896808624267579, "step": 305 }, { "epoch": 0.37159124962541207, "grad_norm": 2.96875, "learning_rate": 5.837845293298629e-06, "loss": 1.716895866394043, "step": 310 }, { "epoch": 0.3775846568774348, "grad_norm": 4.125, "learning_rate": 5.8315593746266346e-06, "loss": 1.671182632446289, "step": 315 }, { "epoch": 0.3835780641294576, "grad_norm": 2.421875, "learning_rate": 5.825157440772241e-06, "loss": 1.715642547607422, "step": 320 }, { "epoch": 0.38957147138148035, "grad_norm": 3.453125, "learning_rate": 5.818639754036683e-06, "loss": 1.6915248870849608, "step": 325 }, { "epoch": 0.39556487863350315, "grad_norm": 2.921875, "learning_rate": 5.812006581463847e-06, "loss": 1.6974452972412108, "step": 330 }, { "epoch": 0.40155828588552595, "grad_norm": 2.46875, "learning_rate": 5.805258194829324e-06, "loss": 1.623448371887207, "step": 335 }, { "epoch": 0.4075516931375487, "grad_norm": 2.921875, "learning_rate": 5.798394870629278e-06, "loss": 1.7037521362304688, "step": 340 }, { "epoch": 0.4135451003895715, "grad_norm": 2.703125, "learning_rate": 5.791416890069118e-06, "loss": 1.6795276641845702, "step": 345 }, { "epoch": 0.4195385076415942, "grad_norm": 2.515625, "learning_rate": 5.784324539051974e-06, "loss": 1.713286781311035, "step": 350 }, { "epoch": 0.425531914893617, "grad_norm": 2.46875, "learning_rate": 5.777118108166982e-06, "loss": 1.6925920486450194, "step": 355 }, { "epoch": 0.4315253221456398, "grad_norm": 2.8125, "learning_rate": 5.769797892677382e-06, "loss": 1.6924890518188476, "step": 360 }, { "epoch": 0.43751872939766256, "grad_norm": 3.109375, "learning_rate": 5.7623641925084195e-06, "loss": 1.6399415969848632, "step": 365 }, { "epoch": 0.44351213664968536, "grad_norm": 2.28125, "learning_rate": 5.754817312235051e-06, "loss": 1.6943456649780273, "step": 370 }, { "epoch": 0.4495055439017081, "grad_norm": 2.71875, "learning_rate": 5.747157561069474e-06, "loss": 1.6115127563476563, "step": 375 }, { "epoch": 0.4554989511537309, "grad_norm": 2.515625, "learning_rate": 5.739385252848452e-06, "loss": 1.6767837524414062, "step": 380 }, { "epoch": 0.46149235840575364, "grad_norm": 3.1875, "learning_rate": 5.731500706020455e-06, "loss": 1.6624629974365235, "step": 385 }, { "epoch": 0.46748576565777644, "grad_norm": 2.234375, "learning_rate": 5.7235042436326165e-06, "loss": 1.6880376815795899, "step": 390 }, { "epoch": 0.47347917290979924, "grad_norm": 2.65625, "learning_rate": 5.715396193317496e-06, "loss": 1.6300373077392578, "step": 395 }, { "epoch": 0.479472580161822, "grad_norm": 2.203125, "learning_rate": 5.707176887279654e-06, "loss": 1.6750167846679687, "step": 400 }, { "epoch": 0.4854659874138448, "grad_norm": 2.890625, "learning_rate": 5.6988466622820396e-06, "loss": 1.657137107849121, "step": 405 }, { "epoch": 0.4914593946658675, "grad_norm": 3.8125, "learning_rate": 5.690405859632195e-06, "loss": 1.6478914260864257, "step": 410 }, { "epoch": 0.4974528019178903, "grad_norm": 2.328125, "learning_rate": 5.681854825168274e-06, "loss": 1.6408031463623047, "step": 415 }, { "epoch": 0.5034462091699131, "grad_norm": 3.390625, "learning_rate": 5.673193909244864e-06, "loss": 1.653570556640625, "step": 420 }, { "epoch": 0.5094396164219359, "grad_norm": 6.125, "learning_rate": 5.6644234667186386e-06, "loss": 1.5698701858520507, "step": 425 }, { "epoch": 0.5154330236739586, "grad_norm": 2.59375, "learning_rate": 5.655543856933818e-06, "loss": 1.658770751953125, "step": 430 }, { "epoch": 0.5214264309259814, "grad_norm": 2.859375, "learning_rate": 5.64655544370744e-06, "loss": 1.6563478469848634, "step": 435 }, { "epoch": 0.5274198381780042, "grad_norm": 2.328125, "learning_rate": 5.637458595314458e-06, "loss": 1.6700237274169922, "step": 440 }, { "epoch": 0.533413245430027, "grad_norm": 2.90625, "learning_rate": 5.6282536844726555e-06, "loss": 1.6282310485839844, "step": 445 }, { "epoch": 0.5394066526820498, "grad_norm": 2.359375, "learning_rate": 5.618941088327366e-06, "loss": 1.6337562561035157, "step": 450 }, { "epoch": 0.5454000599340725, "grad_norm": 2.765625, "learning_rate": 5.609521188436027e-06, "loss": 1.5626126289367677, "step": 455 }, { "epoch": 0.5513934671860953, "grad_norm": 3.25, "learning_rate": 5.599994370752548e-06, "loss": 1.6471063613891601, "step": 460 }, { "epoch": 0.5573868744381181, "grad_norm": 2.546875, "learning_rate": 5.5903610256114894e-06, "loss": 1.6376890182495116, "step": 465 }, { "epoch": 0.5633802816901409, "grad_norm": 2.703125, "learning_rate": 5.5806215477120785e-06, "loss": 1.6394289016723633, "step": 470 }, { "epoch": 0.5693736889421637, "grad_norm": 2.796875, "learning_rate": 5.5707763361020325e-06, "loss": 1.6407564163208008, "step": 475 }, { "epoch": 0.5753670961941864, "grad_norm": 2.5625, "learning_rate": 5.560825794161212e-06, "loss": 1.637693214416504, "step": 480 }, { "epoch": 0.5813605034462092, "grad_norm": 2.71875, "learning_rate": 5.550770329585087e-06, "loss": 1.6549390792846679, "step": 485 }, { "epoch": 0.587353910698232, "grad_norm": 2.59375, "learning_rate": 5.540610354368041e-06, "loss": 1.6286939620971679, "step": 490 }, { "epoch": 0.5933473179502547, "grad_norm": 2.765625, "learning_rate": 5.530346284786487e-06, "loss": 1.6594453811645509, "step": 495 }, { "epoch": 0.5993407252022775, "grad_norm": 2.21875, "learning_rate": 5.519978541381812e-06, "loss": 1.6667818069458007, "step": 500 }, { "epoch": 0.6053341324543002, "grad_norm": 2.59375, "learning_rate": 5.509507548943144e-06, "loss": 1.6215856552124024, "step": 505 }, { "epoch": 0.611327539706323, "grad_norm": 2.40625, "learning_rate": 5.498933736489952e-06, "loss": 1.6226997375488281, "step": 510 }, { "epoch": 0.6173209469583458, "grad_norm": 2.484375, "learning_rate": 5.4882575372544665e-06, "loss": 1.627206802368164, "step": 515 }, { "epoch": 0.6233143542103686, "grad_norm": 2.15625, "learning_rate": 5.4774793886639254e-06, "loss": 1.664950180053711, "step": 520 }, { "epoch": 0.6293077614623914, "grad_norm": 2.59375, "learning_rate": 5.466599732322658e-06, "loss": 1.6241949081420899, "step": 525 }, { "epoch": 0.6353011687144141, "grad_norm": 2.875, "learning_rate": 5.455619013993986e-06, "loss": 1.643407440185547, "step": 530 }, { "epoch": 0.6412945759664369, "grad_norm": 3.453125, "learning_rate": 5.444537683581964e-06, "loss": 1.6385822296142578, "step": 535 }, { "epoch": 0.6472879832184597, "grad_norm": 2.421875, "learning_rate": 5.433356195112941e-06, "loss": 1.6456333160400392, "step": 540 }, { "epoch": 0.6532813904704825, "grad_norm": 2.421875, "learning_rate": 5.422075006716963e-06, "loss": 1.6289976119995118, "step": 545 }, { "epoch": 0.6592747977225052, "grad_norm": 3.15625, "learning_rate": 5.410694580608999e-06, "loss": 1.5806748390197753, "step": 550 }, { "epoch": 0.665268204974528, "grad_norm": 2.484375, "learning_rate": 5.399215383070004e-06, "loss": 1.6152603149414062, "step": 555 }, { "epoch": 0.6712616122265508, "grad_norm": 3.609375, "learning_rate": 5.387637884427817e-06, "loss": 1.681402587890625, "step": 560 }, { "epoch": 0.6772550194785736, "grad_norm": 2.4375, "learning_rate": 5.3759625590378845e-06, "loss": 1.6515823364257813, "step": 565 }, { "epoch": 0.6832484267305964, "grad_norm": 3.078125, "learning_rate": 5.364189885263831e-06, "loss": 1.6387393951416016, "step": 570 }, { "epoch": 0.6892418339826191, "grad_norm": 2.828125, "learning_rate": 5.352320345457858e-06, "loss": 1.6449310302734375, "step": 575 }, { "epoch": 0.6952352412346419, "grad_norm": 3.609375, "learning_rate": 5.340354425940977e-06, "loss": 1.6439811706542968, "step": 580 }, { "epoch": 0.7012286484866647, "grad_norm": 3.015625, "learning_rate": 5.328292616983092e-06, "loss": 1.6770418167114258, "step": 585 }, { "epoch": 0.7072220557386875, "grad_norm": 3.21875, "learning_rate": 5.316135412782903e-06, "loss": 1.6499910354614258, "step": 590 }, { "epoch": 0.7132154629907103, "grad_norm": 2.84375, "learning_rate": 5.303883311447664e-06, "loss": 1.6471555709838868, "step": 595 }, { "epoch": 0.719208870242733, "grad_norm": 2.84375, "learning_rate": 5.291536814972772e-06, "loss": 1.6304954528808593, "step": 600 }, { "epoch": 0.7252022774947557, "grad_norm": 2.984375, "learning_rate": 5.279096429221197e-06, "loss": 1.6022422790527344, "step": 605 }, { "epoch": 0.7311956847467785, "grad_norm": 2.4375, "learning_rate": 5.266562663902762e-06, "loss": 1.607522201538086, "step": 610 }, { "epoch": 0.7371890919988013, "grad_norm": 2.609375, "learning_rate": 5.253936032553253e-06, "loss": 1.611679458618164, "step": 615 }, { "epoch": 0.7431824992508241, "grad_norm": 3.0625, "learning_rate": 5.241217052513379e-06, "loss": 1.5275883674621582, "step": 620 }, { "epoch": 0.7491759065028468, "grad_norm": 3.03125, "learning_rate": 5.228406244907581e-06, "loss": 1.615974235534668, "step": 625 }, { "epoch": 0.7551693137548696, "grad_norm": 2.5, "learning_rate": 5.215504134622673e-06, "loss": 1.588763427734375, "step": 630 }, { "epoch": 0.7611627210068924, "grad_norm": 2.296875, "learning_rate": 5.202511250286339e-06, "loss": 1.6228040695190429, "step": 635 }, { "epoch": 0.7671561282589152, "grad_norm": 3.390625, "learning_rate": 5.189428124245477e-06, "loss": 1.6361326217651366, "step": 640 }, { "epoch": 0.773149535510938, "grad_norm": 6.90625, "learning_rate": 5.176255292544384e-06, "loss": 1.5268966674804687, "step": 645 }, { "epoch": 0.7791429427629607, "grad_norm": 2.578125, "learning_rate": 5.162993294902792e-06, "loss": 1.629738998413086, "step": 650 }, { "epoch": 0.7851363500149835, "grad_norm": 2.171875, "learning_rate": 5.149642674693761e-06, "loss": 1.6338333129882812, "step": 655 }, { "epoch": 0.7911297572670063, "grad_norm": 2.5625, "learning_rate": 5.136203978921408e-06, "loss": 1.6386873245239257, "step": 660 }, { "epoch": 0.7971231645190291, "grad_norm": 2.4375, "learning_rate": 5.1226777581985006e-06, "loss": 1.601084327697754, "step": 665 }, { "epoch": 0.8031165717710519, "grad_norm": 2.71875, "learning_rate": 5.109064566723894e-06, "loss": 1.6157960891723633, "step": 670 }, { "epoch": 0.8091099790230746, "grad_norm": 2.515625, "learning_rate": 5.095364962259825e-06, "loss": 1.6251134872436523, "step": 675 }, { "epoch": 0.8151033862750974, "grad_norm": 2.9375, "learning_rate": 5.0815795061090595e-06, "loss": 1.6243366241455077, "step": 680 }, { "epoch": 0.8210967935271202, "grad_norm": 3.171875, "learning_rate": 5.067708763091897e-06, "loss": 1.6381990432739257, "step": 685 }, { "epoch": 0.827090200779143, "grad_norm": 2.484375, "learning_rate": 5.053753301523023e-06, "loss": 1.6106611251831056, "step": 690 }, { "epoch": 0.8330836080311658, "grad_norm": 2.796875, "learning_rate": 5.0397136931882325e-06, "loss": 1.6053501129150392, "step": 695 }, { "epoch": 0.8390770152831885, "grad_norm": 2.53125, "learning_rate": 5.025590513320992e-06, "loss": 1.6138896942138672, "step": 700 }, { "epoch": 0.8450704225352113, "grad_norm": 2.359375, "learning_rate": 5.01138434057888e-06, "loss": 1.64343204498291, "step": 705 }, { "epoch": 0.851063829787234, "grad_norm": 2.453125, "learning_rate": 4.997095757019873e-06, "loss": 1.6422714233398437, "step": 710 }, { "epoch": 0.8570572370392568, "grad_norm": 2.421875, "learning_rate": 4.9827253480785e-06, "loss": 1.6257701873779298, "step": 715 }, { "epoch": 0.8630506442912796, "grad_norm": 3.09375, "learning_rate": 4.968273702541856e-06, "loss": 1.5960515022277832, "step": 720 }, { "epoch": 0.8690440515433023, "grad_norm": 2.5, "learning_rate": 4.953741412525473e-06, "loss": 1.591513442993164, "step": 725 }, { "epoch": 0.8750374587953251, "grad_norm": 2.34375, "learning_rate": 4.939129073449069e-06, "loss": 1.6213167190551758, "step": 730 }, { "epoch": 0.8810308660473479, "grad_norm": 2.6875, "learning_rate": 4.924437284012141e-06, "loss": 1.6189735412597657, "step": 735 }, { "epoch": 0.8870242732993707, "grad_norm": 2.140625, "learning_rate": 4.90966664616945e-06, "loss": 1.598948860168457, "step": 740 }, { "epoch": 0.8930176805513935, "grad_norm": 4.25, "learning_rate": 4.8948177651063375e-06, "loss": 1.6328046798706055, "step": 745 }, { "epoch": 0.8990110878034162, "grad_norm": 2.5625, "learning_rate": 4.879891249213951e-06, "loss": 1.623681640625, "step": 750 }, { "epoch": 0.905004495055439, "grad_norm": 2.375, "learning_rate": 4.8648877100642994e-06, "loss": 1.6063928604125977, "step": 755 }, { "epoch": 0.9109979023074618, "grad_norm": 3.0625, "learning_rate": 4.849807762385208e-06, "loss": 1.5999799728393556, "step": 760 }, { "epoch": 0.9169913095594846, "grad_norm": 2.296875, "learning_rate": 4.834652024035124e-06, "loss": 1.6070215225219726, "step": 765 }, { "epoch": 0.9229847168115073, "grad_norm": 3.78125, "learning_rate": 4.819421115977806e-06, "loss": 1.5875773429870605, "step": 770 }, { "epoch": 0.9289781240635301, "grad_norm": 2.578125, "learning_rate": 4.804115662256878e-06, "loss": 1.600590705871582, "step": 775 }, { "epoch": 0.9349715313155529, "grad_norm": 2.9375, "learning_rate": 4.788736289970265e-06, "loss": 1.5909998893737793, "step": 780 }, { "epoch": 0.9409649385675757, "grad_norm": 3.234375, "learning_rate": 4.773283629244493e-06, "loss": 1.586691665649414, "step": 785 }, { "epoch": 0.9469583458195985, "grad_norm": 2.953125, "learning_rate": 4.757758313208882e-06, "loss": 1.5925468444824218, "step": 790 }, { "epoch": 0.9529517530716212, "grad_norm": 2.328125, "learning_rate": 4.7421609779695955e-06, "loss": 1.6134449005126954, "step": 795 }, { "epoch": 0.958945160323644, "grad_norm": 2.71875, "learning_rate": 4.726492262583582e-06, "loss": 1.5943148612976075, "step": 800 }, { "epoch": 0.9649385675756668, "grad_norm": 2.703125, "learning_rate": 4.710752809032389e-06, "loss": 1.5899864196777345, "step": 805 }, { "epoch": 0.9709319748276896, "grad_norm": 2.453125, "learning_rate": 4.694943262195864e-06, "loss": 1.6044147491455079, "step": 810 }, { "epoch": 0.9769253820797124, "grad_norm": 2.796875, "learning_rate": 4.679064269825728e-06, "loss": 1.6048875808715821, "step": 815 }, { "epoch": 0.982918789331735, "grad_norm": 2.796875, "learning_rate": 4.6631164825190395e-06, "loss": 1.5995728492736816, "step": 820 }, { "epoch": 0.9889121965837578, "grad_norm": 2.640625, "learning_rate": 4.647100553691533e-06, "loss": 1.6007902145385742, "step": 825 }, { "epoch": 0.9949056038357806, "grad_norm": 2.75, "learning_rate": 4.6310171395508524e-06, "loss": 1.581219482421875, "step": 830 }, { "epoch": 1.0, "grad_norm": 12.125, "learning_rate": 4.6148668990696625e-06, "loss": 1.5526494979858398, "step": 835 }, { "epoch": 1.0059934072520227, "grad_norm": 3.671875, "learning_rate": 4.598650493958648e-06, "loss": 1.5961176872253418, "step": 840 }, { "epoch": 1.0119868145040456, "grad_norm": 2.46875, "learning_rate": 4.582368588639405e-06, "loss": 1.6021915435791017, "step": 845 }, { "epoch": 1.0179802217560683, "grad_norm": 2.4375, "learning_rate": 4.566021850217216e-06, "loss": 1.5884843826293946, "step": 850 }, { "epoch": 1.0239736290080912, "grad_norm": 2.421875, "learning_rate": 4.5496109484537165e-06, "loss": 1.6041584014892578, "step": 855 }, { "epoch": 1.0299670362601139, "grad_norm": 5.8125, "learning_rate": 4.533136555739454e-06, "loss": 1.5464808464050293, "step": 860 }, { "epoch": 1.0359604435121366, "grad_norm": 3.28125, "learning_rate": 4.516599347066343e-06, "loss": 1.61636962890625, "step": 865 }, { "epoch": 1.0419538507641595, "grad_norm": 3.03125, "learning_rate": 4.5e-06, "loss": 1.591432476043701, "step": 870 }, { "epoch": 1.0479472580161822, "grad_norm": 2.796875, "learning_rate": 4.483339194651992e-06, "loss": 1.5649962425231934, "step": 875 }, { "epoch": 1.053940665268205, "grad_norm": 2.640625, "learning_rate": 4.466617613651963e-06, "loss": 1.6130435943603516, "step": 880 }, { "epoch": 1.0599340725202278, "grad_norm": 2.328125, "learning_rate": 4.449835942119669e-06, "loss": 1.6055917739868164, "step": 885 }, { "epoch": 1.0659274797722504, "grad_norm": 3.375, "learning_rate": 4.432994867636909e-06, "loss": 1.5505709648132324, "step": 890 }, { "epoch": 1.0719208870242734, "grad_norm": 2.765625, "learning_rate": 4.416095080219348e-06, "loss": 1.573164176940918, "step": 895 }, { "epoch": 1.077914294276296, "grad_norm": 2.265625, "learning_rate": 4.39913727228825e-06, "loss": 1.6401851654052735, "step": 900 }, { "epoch": 1.083907701528319, "grad_norm": 2.859375, "learning_rate": 4.382122138642106e-06, "loss": 1.5508790969848634, "step": 905 }, { "epoch": 1.0899011087803416, "grad_norm": 2.296875, "learning_rate": 4.365050376428166e-06, "loss": 1.5652284622192383, "step": 910 }, { "epoch": 1.0958945160323643, "grad_norm": 2.96875, "learning_rate": 4.347922685113877e-06, "loss": 1.5959260940551758, "step": 915 }, { "epoch": 1.1018879232843872, "grad_norm": 2.421875, "learning_rate": 4.330739766458225e-06, "loss": 1.5954066276550294, "step": 920 }, { "epoch": 1.10788133053641, "grad_norm": 2.375, "learning_rate": 4.313502324482978e-06, "loss": 1.6183170318603515, "step": 925 }, { "epoch": 1.1138747377884328, "grad_norm": 4.71875, "learning_rate": 4.296211065443848e-06, "loss": 1.57733211517334, "step": 930 }, { "epoch": 1.1198681450404555, "grad_norm": 2.359375, "learning_rate": 4.2788666978015435e-06, "loss": 1.6181407928466798, "step": 935 }, { "epoch": 1.1258615522924782, "grad_norm": 4.5, "learning_rate": 4.261469932192756e-06, "loss": 1.6145042419433593, "step": 940 }, { "epoch": 1.131854959544501, "grad_norm": 2.765625, "learning_rate": 4.244021481401032e-06, "loss": 1.6090787887573241, "step": 945 }, { "epoch": 1.1378483667965238, "grad_norm": 3.3125, "learning_rate": 4.2265220603275714e-06, "loss": 1.6005640029907227, "step": 950 }, { "epoch": 1.1438417740485467, "grad_norm": 2.9375, "learning_rate": 4.208972385961943e-06, "loss": 1.5760242462158203, "step": 955 }, { "epoch": 1.1498351813005694, "grad_norm": 2.875, "learning_rate": 4.191373177352699e-06, "loss": 1.616297721862793, "step": 960 }, { "epoch": 1.155828588552592, "grad_norm": 3.09375, "learning_rate": 4.17372515557792e-06, "loss": 1.592572021484375, "step": 965 }, { "epoch": 1.161821995804615, "grad_norm": 2.546875, "learning_rate": 4.1560290437156676e-06, "loss": 1.5869397163391112, "step": 970 }, { "epoch": 1.1678154030566377, "grad_norm": 2.203125, "learning_rate": 4.13828556681436e-06, "loss": 1.5882441520690918, "step": 975 }, { "epoch": 1.1738088103086604, "grad_norm": 3.578125, "learning_rate": 4.120495451863064e-06, "loss": 1.5809892654418944, "step": 980 }, { "epoch": 1.1798022175606833, "grad_norm": 2.40625, "learning_rate": 4.102659427761711e-06, "loss": 1.5863639831542968, "step": 985 }, { "epoch": 1.185795624812706, "grad_norm": 2.3125, "learning_rate": 4.084778225291231e-06, "loss": 1.607373046875, "step": 990 }, { "epoch": 1.1917890320647289, "grad_norm": 3.0, "learning_rate": 4.066852577083606e-06, "loss": 1.5928448677062987, "step": 995 }, { "epoch": 1.1977824393167515, "grad_norm": 2.390625, "learning_rate": 4.048883217591862e-06, "loss": 1.604298210144043, "step": 1000 }, { "epoch": 1.2037758465687745, "grad_norm": 3.0625, "learning_rate": 4.030870883059971e-06, "loss": 1.6035194396972656, "step": 1005 }, { "epoch": 1.2097692538207971, "grad_norm": 1.984375, "learning_rate": 4.012816311492685e-06, "loss": 1.5925480842590332, "step": 1010 }, { "epoch": 1.2157626610728198, "grad_norm": 4.1875, "learning_rate": 3.9947202426253005e-06, "loss": 1.5639928817749023, "step": 1015 }, { "epoch": 1.2217560683248427, "grad_norm": 2.609375, "learning_rate": 3.976583417893348e-06, "loss": 1.588791561126709, "step": 1020 }, { "epoch": 1.2277494755768654, "grad_norm": 2.390625, "learning_rate": 3.958406580402219e-06, "loss": 1.5938244819641114, "step": 1025 }, { "epoch": 1.233742882828888, "grad_norm": 2.71875, "learning_rate": 3.9401904748967105e-06, "loss": 1.5357808113098144, "step": 1030 }, { "epoch": 1.239736290080911, "grad_norm": 2.984375, "learning_rate": 3.9219358477305174e-06, "loss": 1.5924224853515625, "step": 1035 }, { "epoch": 1.2457296973329337, "grad_norm": 2.828125, "learning_rate": 3.903643446835653e-06, "loss": 1.5829044342041017, "step": 1040 }, { "epoch": 1.2517231045849566, "grad_norm": 2.875, "learning_rate": 3.885314021691801e-06, "loss": 1.6062620162963868, "step": 1045 }, { "epoch": 1.2577165118369793, "grad_norm": 2.84375, "learning_rate": 3.866948323295611e-06, "loss": 1.5546697616577148, "step": 1050 }, { "epoch": 1.2637099190890022, "grad_norm": 2.390625, "learning_rate": 3.848547104129925e-06, "loss": 1.5495938301086425, "step": 1055 }, { "epoch": 1.269703326341025, "grad_norm": 2.28125, "learning_rate": 3.830111118132952e-06, "loss": 1.6037557601928711, "step": 1060 }, { "epoch": 1.2756967335930476, "grad_norm": 2.40625, "learning_rate": 3.8116411206673698e-06, "loss": 1.580827808380127, "step": 1065 }, { "epoch": 1.2816901408450705, "grad_norm": 2.390625, "learning_rate": 3.793137868489383e-06, "loss": 1.619793701171875, "step": 1070 }, { "epoch": 1.2876835480970932, "grad_norm": 3.359375, "learning_rate": 3.774602119717716e-06, "loss": 1.570338821411133, "step": 1075 }, { "epoch": 1.2936769553491159, "grad_norm": 3.34375, "learning_rate": 3.756034633802548e-06, "loss": 1.5999541282653809, "step": 1080 }, { "epoch": 1.2996703626011388, "grad_norm": 2.609375, "learning_rate": 3.7374361714943966e-06, "loss": 1.582872486114502, "step": 1085 }, { "epoch": 1.3056637698531615, "grad_norm": 2.734375, "learning_rate": 3.718807494812953e-06, "loss": 1.6128786087036133, "step": 1090 }, { "epoch": 1.3116571771051844, "grad_norm": 3.046875, "learning_rate": 3.700149367015857e-06, "loss": 1.5680095672607421, "step": 1095 }, { "epoch": 1.317650584357207, "grad_norm": 2.390625, "learning_rate": 3.6814625525674226e-06, "loss": 1.5940041542053223, "step": 1100 }, { "epoch": 1.32364399160923, "grad_norm": 2.71875, "learning_rate": 3.6627478171073214e-06, "loss": 1.5805049896240235, "step": 1105 }, { "epoch": 1.3296373988612527, "grad_norm": 2.234375, "learning_rate": 3.644005927419209e-06, "loss": 1.578407382965088, "step": 1110 }, { "epoch": 1.3356308061132753, "grad_norm": 2.375, "learning_rate": 3.625237651399309e-06, "loss": 1.6166658401489258, "step": 1115 }, { "epoch": 1.3416242133652982, "grad_norm": 2.796875, "learning_rate": 3.6064437580249502e-06, "loss": 1.5907697677612305, "step": 1120 }, { "epoch": 1.347617620617321, "grad_norm": 2.15625, "learning_rate": 3.5876250173230607e-06, "loss": 1.5846922874450684, "step": 1125 }, { "epoch": 1.3536110278693436, "grad_norm": 2.265625, "learning_rate": 3.5687822003386186e-06, "loss": 1.551988697052002, "step": 1130 }, { "epoch": 1.3596044351213665, "grad_norm": 3.078125, "learning_rate": 3.5499160791030603e-06, "loss": 1.5860767364501953, "step": 1135 }, { "epoch": 1.3655978423733892, "grad_norm": 2.765625, "learning_rate": 3.5310274266026473e-06, "loss": 1.566511344909668, "step": 1140 }, { "epoch": 1.3715912496254121, "grad_norm": 3.359375, "learning_rate": 3.512117016746796e-06, "loss": 1.5781704902648925, "step": 1145 }, { "epoch": 1.3775846568774348, "grad_norm": 2.265625, "learning_rate": 3.4931856243363714e-06, "loss": 1.5980149269104005, "step": 1150 }, { "epoch": 1.3835780641294577, "grad_norm": 2.796875, "learning_rate": 3.4742340250319376e-06, "loss": 1.604014778137207, "step": 1155 }, { "epoch": 1.3895714713814804, "grad_norm": 2.9375, "learning_rate": 3.4552629953219796e-06, "loss": 1.5630529403686524, "step": 1160 }, { "epoch": 1.395564878633503, "grad_norm": 2.703125, "learning_rate": 3.436273312491089e-06, "loss": 1.6129833221435548, "step": 1165 }, { "epoch": 1.401558285885526, "grad_norm": 2.4375, "learning_rate": 3.4172657545881174e-06, "loss": 1.5282090187072754, "step": 1170 }, { "epoch": 1.4075516931375487, "grad_norm": 2.71875, "learning_rate": 3.3982411003942946e-06, "loss": 1.5788789749145509, "step": 1175 }, { "epoch": 1.4135451003895714, "grad_norm": 2.890625, "learning_rate": 3.379200129391327e-06, "loss": 1.5974101066589355, "step": 1180 }, { "epoch": 1.4195385076415943, "grad_norm": 2.640625, "learning_rate": 3.3601436217294525e-06, "loss": 1.5513522148132324, "step": 1185 }, { "epoch": 1.425531914893617, "grad_norm": 3.015625, "learning_rate": 3.3410723581954804e-06, "loss": 1.5854881286621094, "step": 1190 }, { "epoch": 1.4315253221456399, "grad_norm": 3.09375, "learning_rate": 3.3219871201808043e-06, "loss": 1.6157951354980469, "step": 1195 }, { "epoch": 1.4375187293976626, "grad_norm": 2.46875, "learning_rate": 3.302888689649378e-06, "loss": 1.578996181488037, "step": 1200 }, { "epoch": 1.4435121366496855, "grad_norm": 2.4375, "learning_rate": 3.2837778491056833e-06, "loss": 1.5944060325622558, "step": 1205 }, { "epoch": 1.4495055439017082, "grad_norm": 2.390625, "learning_rate": 3.264655381562668e-06, "loss": 1.5933379173278808, "step": 1210 }, { "epoch": 1.4554989511537308, "grad_norm": 4.71875, "learning_rate": 3.2455220705096634e-06, "loss": 1.568842887878418, "step": 1215 }, { "epoch": 1.4614923584057538, "grad_norm": 2.203125, "learning_rate": 3.2263786998802824e-06, "loss": 1.624056053161621, "step": 1220 }, { "epoch": 1.4674857656577764, "grad_norm": 2.59375, "learning_rate": 3.207226054020301e-06, "loss": 1.5804908752441407, "step": 1225 }, { "epoch": 1.4734791729097991, "grad_norm": 3.515625, "learning_rate": 3.188064917655521e-06, "loss": 1.5621639251708985, "step": 1230 }, { "epoch": 1.479472580161822, "grad_norm": 2.28125, "learning_rate": 3.168896075859622e-06, "loss": 1.6008676528930663, "step": 1235 }, { "epoch": 1.4854659874138447, "grad_norm": 2.609375, "learning_rate": 3.1497203140219867e-06, "loss": 1.5788877487182618, "step": 1240 }, { "epoch": 1.4914593946658674, "grad_norm": 2.078125, "learning_rate": 3.1305384178155325e-06, "loss": 1.569349193572998, "step": 1245 }, { "epoch": 1.4974528019178903, "grad_norm": 2.875, "learning_rate": 3.111351173164511e-06, "loss": 1.5960070610046386, "step": 1250 }, { "epoch": 1.5034462091699132, "grad_norm": 2.34375, "learning_rate": 3.0921593662123126e-06, "loss": 1.5280057907104492, "step": 1255 }, { "epoch": 1.509439616421936, "grad_norm": 2.125, "learning_rate": 3.0729637832892576e-06, "loss": 1.5992347717285156, "step": 1260 }, { "epoch": 1.5154330236739586, "grad_norm": 4.15625, "learning_rate": 3.0537652108803723e-06, "loss": 1.5896804809570313, "step": 1265 }, { "epoch": 1.5214264309259815, "grad_norm": 2.484375, "learning_rate": 3.03456443559317e-06, "loss": 1.5657002449035644, "step": 1270 }, { "epoch": 1.5274198381780042, "grad_norm": 2.578125, "learning_rate": 3.0153622441254238e-06, "loss": 1.6020334243774415, "step": 1275 }, { "epoch": 1.5334132454300269, "grad_norm": 3.3125, "learning_rate": 2.9961594232329253e-06, "loss": 1.586386489868164, "step": 1280 }, { "epoch": 1.5394066526820498, "grad_norm": 2.4375, "learning_rate": 2.97695675969726e-06, "loss": 1.5938687324523926, "step": 1285 }, { "epoch": 1.5454000599340725, "grad_norm": 2.796875, "learning_rate": 2.957755040293562e-06, "loss": 1.613763427734375, "step": 1290 }, { "epoch": 1.5513934671860952, "grad_norm": 3.0, "learning_rate": 2.938555051758287e-06, "loss": 1.585596752166748, "step": 1295 }, { "epoch": 1.557386874438118, "grad_norm": 2.375, "learning_rate": 2.9193575807569694e-06, "loss": 1.547711181640625, "step": 1300 }, { "epoch": 1.563380281690141, "grad_norm": 2.609375, "learning_rate": 2.9001634138519957e-06, "loss": 1.5838136672973633, "step": 1305 }, { "epoch": 1.5693736889421637, "grad_norm": 3.109375, "learning_rate": 2.8809733374703766e-06, "loss": 1.545951747894287, "step": 1310 }, { "epoch": 1.5753670961941864, "grad_norm": 3.359375, "learning_rate": 2.861788137871525e-06, "loss": 1.5860053062438966, "step": 1315 }, { "epoch": 1.5813605034462093, "grad_norm": 2.765625, "learning_rate": 2.842608601115041e-06, "loss": 1.5784396171569823, "step": 1320 }, { "epoch": 1.587353910698232, "grad_norm": 2.640625, "learning_rate": 2.8234355130285076e-06, "loss": 1.6107646942138671, "step": 1325 }, { "epoch": 1.5933473179502546, "grad_norm": 2.109375, "learning_rate": 2.8042696591752894e-06, "loss": 1.5984049797058106, "step": 1330 }, { "epoch": 1.5993407252022775, "grad_norm": 2.34375, "learning_rate": 2.7851118248223482e-06, "loss": 1.553554153442383, "step": 1335 }, { "epoch": 1.6053341324543002, "grad_norm": 2.484375, "learning_rate": 2.765962794908072e-06, "loss": 1.6150043487548829, "step": 1340 }, { "epoch": 1.611327539706323, "grad_norm": 2.453125, "learning_rate": 2.74682335401011e-06, "loss": 1.5275163650512695, "step": 1345 }, { "epoch": 1.6173209469583458, "grad_norm": 2.9375, "learning_rate": 2.727694286313227e-06, "loss": 1.4862093925476074, "step": 1350 }, { "epoch": 1.6233143542103687, "grad_norm": 3.5625, "learning_rate": 2.7085763755771794e-06, "loss": 1.5487674713134765, "step": 1355 }, { "epoch": 1.6293077614623914, "grad_norm": 2.40625, "learning_rate": 2.689470405104594e-06, "loss": 1.594883346557617, "step": 1360 }, { "epoch": 1.635301168714414, "grad_norm": 4.0625, "learning_rate": 2.6703771577088814e-06, "loss": 1.5667915344238281, "step": 1365 }, { "epoch": 1.641294575966437, "grad_norm": 2.5, "learning_rate": 2.651297415682159e-06, "loss": 1.5776025772094726, "step": 1370 }, { "epoch": 1.6472879832184597, "grad_norm": 2.3125, "learning_rate": 2.6322319607631994e-06, "loss": 1.5362964630126954, "step": 1375 }, { "epoch": 1.6532813904704824, "grad_norm": 3.765625, "learning_rate": 2.6131815741054006e-06, "loss": 1.552170181274414, "step": 1380 }, { "epoch": 1.6592747977225053, "grad_norm": 2.8125, "learning_rate": 2.5941470362447847e-06, "loss": 1.5639829635620117, "step": 1385 }, { "epoch": 1.665268204974528, "grad_norm": 2.640625, "learning_rate": 2.5751291270680096e-06, "loss": 1.5860962867736816, "step": 1390 }, { "epoch": 1.6712616122265507, "grad_norm": 2.4375, "learning_rate": 2.5561286257804214e-06, "loss": 1.6001644134521484, "step": 1395 }, { "epoch": 1.6772550194785736, "grad_norm": 3.4375, "learning_rate": 2.537146310874127e-06, "loss": 1.5804731369018554, "step": 1400 }, { "epoch": 1.6832484267305965, "grad_norm": 2.640625, "learning_rate": 2.5181829600960966e-06, "loss": 1.5659726142883301, "step": 1405 }, { "epoch": 1.689241833982619, "grad_norm": 2.25, "learning_rate": 2.499239350416298e-06, "loss": 1.6174665451049806, "step": 1410 }, { "epoch": 1.6952352412346419, "grad_norm": 2.90625, "learning_rate": 2.4803162579958674e-06, "loss": 1.5999381065368652, "step": 1415 }, { "epoch": 1.7012286484866648, "grad_norm": 2.609375, "learning_rate": 2.4614144581552984e-06, "loss": 1.5755634307861328, "step": 1420 }, { "epoch": 1.7072220557386875, "grad_norm": 2.515625, "learning_rate": 2.442534725342684e-06, "loss": 1.6061223983764648, "step": 1425 }, { "epoch": 1.7132154629907101, "grad_norm": 2.78125, "learning_rate": 2.423677833101982e-06, "loss": 1.5202569961547852, "step": 1430 }, { "epoch": 1.719208870242733, "grad_norm": 2.484375, "learning_rate": 2.404844554041323e-06, "loss": 1.581466007232666, "step": 1435 }, { "epoch": 1.7252022774947557, "grad_norm": 2.953125, "learning_rate": 2.386035659801352e-06, "loss": 1.6154771804809571, "step": 1440 }, { "epoch": 1.7311956847467784, "grad_norm": 2.46875, "learning_rate": 2.3672519210236163e-06, "loss": 1.5641371726989746, "step": 1445 }, { "epoch": 1.7371890919988013, "grad_norm": 2.828125, "learning_rate": 2.348494107318988e-06, "loss": 1.5726320266723632, "step": 1450 }, { "epoch": 1.7431824992508242, "grad_norm": 2.640625, "learning_rate": 2.3297629872361327e-06, "loss": 1.5837481498718262, "step": 1455 }, { "epoch": 1.7491759065028467, "grad_norm": 2.75, "learning_rate": 2.31105932823002e-06, "loss": 1.5849193572998046, "step": 1460 }, { "epoch": 1.7551693137548696, "grad_norm": 2.4375, "learning_rate": 2.2923838966304787e-06, "loss": 1.5511338233947753, "step": 1465 }, { "epoch": 1.7611627210068925, "grad_norm": 3.765625, "learning_rate": 2.2737374576107973e-06, "loss": 1.577767562866211, "step": 1470 }, { "epoch": 1.7671561282589152, "grad_norm": 2.15625, "learning_rate": 2.2551207751563795e-06, "loss": 1.6033267974853516, "step": 1475 }, { "epoch": 1.773149535510938, "grad_norm": 2.984375, "learning_rate": 2.2365346120334335e-06, "loss": 1.564914608001709, "step": 1480 }, { "epoch": 1.7791429427629608, "grad_norm": 2.390625, "learning_rate": 2.217979729757725e-06, "loss": 1.5818105697631837, "step": 1485 }, { "epoch": 1.7851363500149835, "grad_norm": 2.171875, "learning_rate": 2.199456888563374e-06, "loss": 1.5914242744445801, "step": 1490 }, { "epoch": 1.7911297572670062, "grad_norm": 3.140625, "learning_rate": 2.180966847371709e-06, "loss": 1.5750107765197754, "step": 1495 }, { "epoch": 1.797123164519029, "grad_norm": 3.046875, "learning_rate": 2.1625103637601677e-06, "loss": 1.602726936340332, "step": 1500 }, { "epoch": 1.803116571771052, "grad_norm": 2.53125, "learning_rate": 2.144088193931266e-06, "loss": 1.5881829261779785, "step": 1505 }, { "epoch": 1.8091099790230745, "grad_norm": 2.4375, "learning_rate": 2.1257010926816023e-06, "loss": 1.584663486480713, "step": 1510 }, { "epoch": 1.8151033862750974, "grad_norm": 2.640625, "learning_rate": 2.107349813370943e-06, "loss": 1.602978515625, "step": 1515 }, { "epoch": 1.8210967935271203, "grad_norm": 3.5, "learning_rate": 2.0890351078913504e-06, "loss": 1.579908847808838, "step": 1520 }, { "epoch": 1.827090200779143, "grad_norm": 2.046875, "learning_rate": 2.0707577266363745e-06, "loss": 1.6223228454589844, "step": 1525 }, { "epoch": 1.8330836080311657, "grad_norm": 2.28125, "learning_rate": 2.0525184184703137e-06, "loss": 1.5527989387512207, "step": 1530 }, { "epoch": 1.8390770152831886, "grad_norm": 2.375, "learning_rate": 2.034317930697527e-06, "loss": 1.555042839050293, "step": 1535 }, { "epoch": 1.8450704225352113, "grad_norm": 2.46875, "learning_rate": 2.0161570090318176e-06, "loss": 1.5828789710998534, "step": 1540 }, { "epoch": 1.851063829787234, "grad_norm": 2.28125, "learning_rate": 1.9980363975658773e-06, "loss": 1.5866991996765136, "step": 1545 }, { "epoch": 1.8570572370392568, "grad_norm": 2.25, "learning_rate": 1.979956838740802e-06, "loss": 1.6043601989746095, "step": 1550 }, { "epoch": 1.8630506442912798, "grad_norm": 2.375, "learning_rate": 1.9619190733156704e-06, "loss": 1.601364517211914, "step": 1555 }, { "epoch": 1.8690440515433022, "grad_norm": 2.484375, "learning_rate": 1.9439238403371956e-06, "loss": 1.595060443878174, "step": 1560 }, { "epoch": 1.8750374587953251, "grad_norm": 2.390625, "learning_rate": 1.92597187710944e-06, "loss": 1.6017419815063476, "step": 1565 }, { "epoch": 1.881030866047348, "grad_norm": 2.40625, "learning_rate": 1.9080639191636142e-06, "loss": 1.5313145637512207, "step": 1570 }, { "epoch": 1.8870242732993707, "grad_norm": 3.296875, "learning_rate": 1.8902007002279344e-06, "loss": 1.5399950981140136, "step": 1575 }, { "epoch": 1.8930176805513934, "grad_norm": 3.671875, "learning_rate": 1.8723829521975593e-06, "loss": 1.5866312026977538, "step": 1580 }, { "epoch": 1.8990110878034163, "grad_norm": 3.578125, "learning_rate": 1.8546114051046077e-06, "loss": 1.5103415489196776, "step": 1585 }, { "epoch": 1.905004495055439, "grad_norm": 2.40625, "learning_rate": 1.8368867870882443e-06, "loss": 1.5101396560668945, "step": 1590 }, { "epoch": 1.9109979023074617, "grad_norm": 3.140625, "learning_rate": 1.8192098243648454e-06, "loss": 1.5957839965820313, "step": 1595 }, { "epoch": 1.9169913095594846, "grad_norm": 2.453125, "learning_rate": 1.8015812411982496e-06, "loss": 1.5403210639953613, "step": 1600 }, { "epoch": 1.9229847168115073, "grad_norm": 3.140625, "learning_rate": 1.7840017598700747e-06, "loss": 1.5802807807922363, "step": 1605 }, { "epoch": 1.92897812406353, "grad_norm": 2.59375, "learning_rate": 1.7664721006501316e-06, "loss": 1.5780931472778321, "step": 1610 }, { "epoch": 1.9349715313155529, "grad_norm": 2.25, "learning_rate": 1.7489929817669087e-06, "loss": 1.6124277114868164, "step": 1615 }, { "epoch": 1.9409649385675758, "grad_norm": 2.140625, "learning_rate": 1.7315651193781476e-06, "loss": 1.5578656196594238, "step": 1620 }, { "epoch": 1.9469583458195985, "grad_norm": 2.875, "learning_rate": 1.7141892275414978e-06, "loss": 1.5834556579589845, "step": 1625 }, { "epoch": 1.9529517530716212, "grad_norm": 2.453125, "learning_rate": 1.696866018185265e-06, "loss": 1.5743704795837403, "step": 1630 }, { "epoch": 1.958945160323644, "grad_norm": 2.90625, "learning_rate": 1.6795962010792332e-06, "loss": 1.5657451629638672, "step": 1635 }, { "epoch": 1.9649385675756668, "grad_norm": 3.046875, "learning_rate": 1.662380483805593e-06, "loss": 1.591666316986084, "step": 1640 }, { "epoch": 1.9709319748276894, "grad_norm": 2.8125, "learning_rate": 1.6452195717299451e-06, "loss": 1.616462516784668, "step": 1645 }, { "epoch": 1.9769253820797124, "grad_norm": 2.359375, "learning_rate": 1.6281141679724012e-06, "loss": 1.5900347709655762, "step": 1650 }, { "epoch": 1.982918789331735, "grad_norm": 2.296875, "learning_rate": 1.6110649733787747e-06, "loss": 1.5767745971679688, "step": 1655 }, { "epoch": 1.9889121965837577, "grad_norm": 2.53125, "learning_rate": 1.5940726864918705e-06, "loss": 1.5806458473205567, "step": 1660 }, { "epoch": 1.9949056038357806, "grad_norm": 2.328125, "learning_rate": 1.577138003522857e-06, "loss": 1.5927390098571776, "step": 1665 }, { "epoch": 2.0, "grad_norm": 8.9375, "learning_rate": 1.5602616183227443e-06, "loss": 1.5510701179504394, "step": 1670 }, { "epoch": 2.005993407252023, "grad_norm": 2.671875, "learning_rate": 1.5434442223539554e-06, "loss": 1.570119285583496, "step": 1675 }, { "epoch": 2.0119868145040454, "grad_norm": 3.390625, "learning_rate": 1.5266865046619966e-06, "loss": 1.5953407287597656, "step": 1680 }, { "epoch": 2.0179802217560683, "grad_norm": 2.84375, "learning_rate": 1.5099891518472226e-06, "loss": 1.5938201904296876, "step": 1685 }, { "epoch": 2.023973629008091, "grad_norm": 2.890625, "learning_rate": 1.4933528480367102e-06, "loss": 1.5878721237182618, "step": 1690 }, { "epoch": 2.0299670362601137, "grad_norm": 3.828125, "learning_rate": 1.4767782748562226e-06, "loss": 1.5573725700378418, "step": 1695 }, { "epoch": 2.0359604435121366, "grad_norm": 2.203125, "learning_rate": 1.4602661114022837e-06, "loss": 1.5768142700195313, "step": 1700 }, { "epoch": 2.0419538507641595, "grad_norm": 3.046875, "learning_rate": 1.4438170342143557e-06, "loss": 1.5976703643798829, "step": 1705 }, { "epoch": 2.0479472580161824, "grad_norm": 2.453125, "learning_rate": 1.4274317172471175e-06, "loss": 1.5801265716552735, "step": 1710 }, { "epoch": 2.053940665268205, "grad_norm": 3.328125, "learning_rate": 1.411110831842852e-06, "loss": 1.4859464645385743, "step": 1715 }, { "epoch": 2.0599340725202278, "grad_norm": 3.21875, "learning_rate": 1.394855046703942e-06, "loss": 1.60047607421875, "step": 1720 }, { "epoch": 2.0659274797722507, "grad_norm": 2.828125, "learning_rate": 1.3786650278654687e-06, "loss": 1.5819260597229003, "step": 1725 }, { "epoch": 2.071920887024273, "grad_norm": 2.484375, "learning_rate": 1.3625414386679239e-06, "loss": 1.5798104286193848, "step": 1730 }, { "epoch": 2.077914294276296, "grad_norm": 2.984375, "learning_rate": 1.3464849397300313e-06, "loss": 1.5598320960998535, "step": 1735 }, { "epoch": 2.083907701528319, "grad_norm": 2.578125, "learning_rate": 1.3304961889216804e-06, "loss": 1.6002256393432617, "step": 1740 }, { "epoch": 2.0899011087803414, "grad_norm": 2.5625, "learning_rate": 1.314575841336971e-06, "loss": 1.6074796676635743, "step": 1745 }, { "epoch": 2.0958945160323643, "grad_norm": 2.265625, "learning_rate": 1.2987245492673759e-06, "loss": 1.5005703926086427, "step": 1750 }, { "epoch": 2.1018879232843872, "grad_norm": 5.34375, "learning_rate": 1.282942962175009e-06, "loss": 1.5408341407775878, "step": 1755 }, { "epoch": 2.10788133053641, "grad_norm": 3.59375, "learning_rate": 1.2672317266660203e-06, "loss": 1.5244135856628418, "step": 1760 }, { "epoch": 2.1138747377884326, "grad_norm": 2.390625, "learning_rate": 1.2515914864641013e-06, "loss": 1.5962490081787108, "step": 1765 }, { "epoch": 2.1198681450404555, "grad_norm": 3.203125, "learning_rate": 1.23602288238411e-06, "loss": 1.5964628219604493, "step": 1770 }, { "epoch": 2.1258615522924784, "grad_norm": 2.375, "learning_rate": 1.220526552305816e-06, "loss": 1.566230583190918, "step": 1775 }, { "epoch": 2.131854959544501, "grad_norm": 3.171875, "learning_rate": 1.2051031311477659e-06, "loss": 1.5923571586608887, "step": 1780 }, { "epoch": 2.137848366796524, "grad_norm": 3.234375, "learning_rate": 1.1897532508412677e-06, "loss": 1.5851450920104981, "step": 1785 }, { "epoch": 2.1438417740485467, "grad_norm": 2.171875, "learning_rate": 1.1744775403044986e-06, "loss": 1.5911012649536134, "step": 1790 }, { "epoch": 2.1498351813005696, "grad_norm": 2.265625, "learning_rate": 1.1592766254167403e-06, "loss": 1.6001272201538086, "step": 1795 }, { "epoch": 2.155828588552592, "grad_norm": 3.546875, "learning_rate": 1.1441511289927315e-06, "loss": 1.5685481071472167, "step": 1800 }, { "epoch": 2.161821995804615, "grad_norm": 2.734375, "learning_rate": 1.1291016707571522e-06, "loss": 1.537570095062256, "step": 1805 }, { "epoch": 2.167815403056638, "grad_norm": 2.421875, "learning_rate": 1.1141288673192298e-06, "loss": 1.6071165084838868, "step": 1810 }, { "epoch": 2.1738088103086604, "grad_norm": 2.6875, "learning_rate": 1.0992333321474803e-06, "loss": 1.5737549781799316, "step": 1815 }, { "epoch": 2.1798022175606833, "grad_norm": 2.890625, "learning_rate": 1.084415675544567e-06, "loss": 1.592972183227539, "step": 1820 }, { "epoch": 2.185795624812706, "grad_norm": 2.359375, "learning_rate": 1.0696765046222984e-06, "loss": 1.6001060485839844, "step": 1825 }, { "epoch": 2.1917890320647286, "grad_norm": 3.28125, "learning_rate": 1.0550164232767532e-06, "loss": 1.5771080017089845, "step": 1830 }, { "epoch": 2.1977824393167515, "grad_norm": 2.625, "learning_rate": 1.0404360321635365e-06, "loss": 1.5672366142272949, "step": 1835 }, { "epoch": 2.2037758465687745, "grad_norm": 2.515625, "learning_rate": 1.025935928673173e-06, "loss": 1.5996353149414062, "step": 1840 }, { "epoch": 2.209769253820797, "grad_norm": 2.265625, "learning_rate": 1.0115167069066236e-06, "loss": 1.570057487487793, "step": 1845 }, { "epoch": 2.21576266107282, "grad_norm": 2.34375, "learning_rate": 9.971789576509511e-07, "loss": 1.6005977630615233, "step": 1850 }, { "epoch": 2.2217560683248427, "grad_norm": 3.0, "learning_rate": 9.8292326835511e-07, "loss": 1.6053207397460938, "step": 1855 }, { "epoch": 2.2277494755768656, "grad_norm": 2.375, "learning_rate": 9.687502231058794e-07, "loss": 1.6059099197387696, "step": 1860 }, { "epoch": 2.233742882828888, "grad_norm": 3.390625, "learning_rate": 9.54660402603929e-07, "loss": 1.5193527221679688, "step": 1865 }, { "epoch": 2.239736290080911, "grad_norm": 2.640625, "learning_rate": 9.406543841400324e-07, "loss": 1.6006130218505858, "step": 1870 }, { "epoch": 2.245729697332934, "grad_norm": 2.140625, "learning_rate": 9.267327415714073e-07, "loss": 1.5676172256469727, "step": 1875 }, { "epoch": 2.2517231045849564, "grad_norm": 2.40625, "learning_rate": 9.12896045298207e-07, "loss": 1.593519401550293, "step": 1880 }, { "epoch": 2.2577165118369793, "grad_norm": 2.359375, "learning_rate": 8.991448622401494e-07, "loss": 1.552088451385498, "step": 1885 }, { "epoch": 2.263709919089002, "grad_norm": 3.140625, "learning_rate": 8.854797558132891e-07, "loss": 1.5782706260681152, "step": 1890 }, { "epoch": 2.269703326341025, "grad_norm": 2.53125, "learning_rate": 8.719012859069315e-07, "loss": 1.5695977210998535, "step": 1895 }, { "epoch": 2.2756967335930476, "grad_norm": 2.203125, "learning_rate": 8.584100088606976e-07, "loss": 1.590079116821289, "step": 1900 }, { "epoch": 2.2816901408450705, "grad_norm": 2.375, "learning_rate": 8.450064774417222e-07, "loss": 1.5943170547485352, "step": 1905 }, { "epoch": 2.2876835480970934, "grad_norm": 3.234375, "learning_rate": 8.316912408220121e-07, "loss": 1.536445426940918, "step": 1910 }, { "epoch": 2.293676955349116, "grad_norm": 3.03125, "learning_rate": 8.184648445559423e-07, "loss": 1.6045770645141602, "step": 1915 }, { "epoch": 2.2996703626011388, "grad_norm": 2.734375, "learning_rate": 8.053278305579042e-07, "loss": 1.5701943397521974, "step": 1920 }, { "epoch": 2.3056637698531617, "grad_norm": 2.40625, "learning_rate": 7.922807370801009e-07, "loss": 1.5690512657165527, "step": 1925 }, { "epoch": 2.311657177105184, "grad_norm": 2.46875, "learning_rate": 7.793240986904986e-07, "loss": 1.5474846839904786, "step": 1930 }, { "epoch": 2.317650584357207, "grad_norm": 4.9375, "learning_rate": 7.664584462509169e-07, "loss": 1.5678466796875, "step": 1935 }, { "epoch": 2.32364399160923, "grad_norm": 3.4375, "learning_rate": 7.536843068952842e-07, "loss": 1.5523689270019532, "step": 1940 }, { "epoch": 2.3296373988612524, "grad_norm": 2.5625, "learning_rate": 7.410022040080365e-07, "loss": 1.5863494873046875, "step": 1945 }, { "epoch": 2.3356308061132753, "grad_norm": 2.8125, "learning_rate": 7.284126572026744e-07, "loss": 1.5466392517089844, "step": 1950 }, { "epoch": 2.3416242133652982, "grad_norm": 3.0625, "learning_rate": 7.159161823004734e-07, "loss": 1.5855929374694824, "step": 1955 }, { "epoch": 2.3476176206173207, "grad_norm": 2.765625, "learning_rate": 7.035132913093509e-07, "loss": 1.5919572830200195, "step": 1960 }, { "epoch": 2.3536110278693436, "grad_norm": 2.765625, "learning_rate": 6.912044924028849e-07, "loss": 1.6009016036987305, "step": 1965 }, { "epoch": 2.3596044351213665, "grad_norm": 2.125, "learning_rate": 6.789902898994955e-07, "loss": 1.5979683876037598, "step": 1970 }, { "epoch": 2.3655978423733894, "grad_norm": 3.453125, "learning_rate": 6.668711842417816e-07, "loss": 1.5708651542663574, "step": 1975 }, { "epoch": 2.371591249625412, "grad_norm": 3.0625, "learning_rate": 6.548476719760146e-07, "loss": 1.5527511596679688, "step": 1980 }, { "epoch": 2.377584656877435, "grad_norm": 2.625, "learning_rate": 6.429202457317982e-07, "loss": 1.5910175323486329, "step": 1985 }, { "epoch": 2.3835780641294577, "grad_norm": 2.671875, "learning_rate": 6.310893942018801e-07, "loss": 1.5702336311340332, "step": 1990 }, { "epoch": 2.38957147138148, "grad_norm": 2.71875, "learning_rate": 6.193556021221307e-07, "loss": 1.5879745483398438, "step": 1995 }, { "epoch": 2.395564878633503, "grad_norm": 2.15625, "learning_rate": 6.077193502516807e-07, "loss": 1.599627685546875, "step": 2000 }, { "epoch": 2.401558285885526, "grad_norm": 3.953125, "learning_rate": 5.961811153532293e-07, "loss": 1.5657157897949219, "step": 2005 }, { "epoch": 2.407551693137549, "grad_norm": 2.546875, "learning_rate": 5.847413701735021e-07, "loss": 1.604018211364746, "step": 2010 }, { "epoch": 2.4135451003895714, "grad_norm": 2.703125, "learning_rate": 5.734005834238864e-07, "loss": 1.58401517868042, "step": 2015 }, { "epoch": 2.4195385076415943, "grad_norm": 3.09375, "learning_rate": 5.621592197612275e-07, "loss": 1.5892996788024902, "step": 2020 }, { "epoch": 2.425531914893617, "grad_norm": 2.328125, "learning_rate": 5.510177397687875e-07, "loss": 1.54070987701416, "step": 2025 }, { "epoch": 2.4315253221456397, "grad_norm": 2.796875, "learning_rate": 5.399765999373772e-07, "loss": 1.573234748840332, "step": 2030 }, { "epoch": 2.4375187293976626, "grad_norm": 2.421875, "learning_rate": 5.29036252646651e-07, "loss": 1.5480667114257813, "step": 2035 }, { "epoch": 2.4435121366496855, "grad_norm": 2.28125, "learning_rate": 5.18197146146572e-07, "loss": 1.5748504638671874, "step": 2040 }, { "epoch": 2.449505543901708, "grad_norm": 5.21875, "learning_rate": 5.074597245390468e-07, "loss": 1.504516315460205, "step": 2045 }, { "epoch": 2.455498951153731, "grad_norm": 2.1875, "learning_rate": 4.968244277597306e-07, "loss": 1.5273450851440429, "step": 2050 }, { "epoch": 2.4614923584057538, "grad_norm": 3.90625, "learning_rate": 4.862916915600001e-07, "loss": 1.6043586730957031, "step": 2055 }, { "epoch": 2.467485765657776, "grad_norm": 2.8125, "learning_rate": 4.7586194748910027e-07, "loss": 1.56980562210083, "step": 2060 }, { "epoch": 2.473479172909799, "grad_norm": 2.9375, "learning_rate": 4.6553562287646366e-07, "loss": 1.624884033203125, "step": 2065 }, { "epoch": 2.479472580161822, "grad_norm": 2.890625, "learning_rate": 4.553131408142016e-07, "loss": 1.5734535217285157, "step": 2070 }, { "epoch": 2.485465987413845, "grad_norm": 2.859375, "learning_rate": 4.4519492013976746e-07, "loss": 1.5830469131469727, "step": 2075 }, { "epoch": 2.4914593946658674, "grad_norm": 2.890625, "learning_rate": 4.3518137541880074e-07, "loss": 1.5716002464294434, "step": 2080 }, { "epoch": 2.4974528019178903, "grad_norm": 2.453125, "learning_rate": 4.252729169281353e-07, "loss": 1.5019952774047851, "step": 2085 }, { "epoch": 2.5034462091699132, "grad_norm": 2.78125, "learning_rate": 4.154699506389924e-07, "loss": 1.6065509796142579, "step": 2090 }, { "epoch": 2.509439616421936, "grad_norm": 2.703125, "learning_rate": 4.0577287820034814e-07, "loss": 1.5870964050292968, "step": 2095 }, { "epoch": 2.5154330236739586, "grad_norm": 2.3125, "learning_rate": 3.9618209692247486e-07, "loss": 1.5819488525390626, "step": 2100 }, { "epoch": 2.5214264309259815, "grad_norm": 3.046875, "learning_rate": 3.866979997606642e-07, "loss": 1.5621922492980957, "step": 2105 }, { "epoch": 2.5274198381780044, "grad_norm": 2.671875, "learning_rate": 3.7732097529912633e-07, "loss": 1.585857105255127, "step": 2110 }, { "epoch": 2.533413245430027, "grad_norm": 2.234375, "learning_rate": 3.680514077350686e-07, "loss": 1.5932723999023437, "step": 2115 }, { "epoch": 2.53940665268205, "grad_norm": 2.328125, "learning_rate": 3.5888967686295293e-07, "loss": 1.5735720634460448, "step": 2120 }, { "epoch": 2.5454000599340727, "grad_norm": 2.328125, "learning_rate": 3.4983615805893745e-07, "loss": 1.5617835998535157, "step": 2125 }, { "epoch": 2.551393467186095, "grad_norm": 2.359375, "learning_rate": 3.4089122226549487e-07, "loss": 1.5272964477539062, "step": 2130 }, { "epoch": 2.557386874438118, "grad_norm": 3.265625, "learning_rate": 3.32055235976213e-07, "loss": 1.5262401580810547, "step": 2135 }, { "epoch": 2.563380281690141, "grad_norm": 2.265625, "learning_rate": 3.233285612207828e-07, "loss": 1.6015663146972656, "step": 2140 }, { "epoch": 2.5693736889421634, "grad_norm": 2.421875, "learning_rate": 3.147115555501604e-07, "loss": 1.5687079429626465, "step": 2145 }, { "epoch": 2.5753670961941864, "grad_norm": 2.890625, "learning_rate": 3.062045720219203e-07, "loss": 1.5640341758728027, "step": 2150 }, { "epoch": 2.5813605034462093, "grad_norm": 2.46875, "learning_rate": 2.9780795918578863e-07, "loss": 1.5695132255554198, "step": 2155 }, { "epoch": 2.5873539106982317, "grad_norm": 3.4375, "learning_rate": 2.8952206106936377e-07, "loss": 1.5747217178344726, "step": 2160 }, { "epoch": 2.5933473179502546, "grad_norm": 3.1875, "learning_rate": 2.813472171640183e-07, "loss": 1.589064598083496, "step": 2165 }, { "epoch": 2.5993407252022775, "grad_norm": 2.609375, "learning_rate": 2.7328376241099316e-07, "loss": 1.5749725341796874, "step": 2170 }, { "epoch": 2.6053341324543, "grad_norm": 2.203125, "learning_rate": 2.6533202718767e-07, "loss": 1.5791637420654296, "step": 2175 }, { "epoch": 2.611327539706323, "grad_norm": 2.828125, "learning_rate": 2.5749233729403755e-07, "loss": 1.5628402709960938, "step": 2180 }, { "epoch": 2.617320946958346, "grad_norm": 2.875, "learning_rate": 2.497650139393419e-07, "loss": 1.6008136749267579, "step": 2185 }, { "epoch": 2.6233143542103687, "grad_norm": 2.953125, "learning_rate": 2.421503737289261e-07, "loss": 1.55454683303833, "step": 2190 }, { "epoch": 2.6293077614623916, "grad_norm": 3.1875, "learning_rate": 2.3464872865126052e-07, "loss": 1.5631563186645507, "step": 2195 }, { "epoch": 2.635301168714414, "grad_norm": 3.453125, "learning_rate": 2.2726038606515452e-07, "loss": 1.5437533378601074, "step": 2200 }, { "epoch": 2.641294575966437, "grad_norm": 2.59375, "learning_rate": 2.1998564868716796e-07, "loss": 1.5831893920898437, "step": 2205 }, { "epoch": 2.64728798321846, "grad_norm": 2.28125, "learning_rate": 2.12824814579207e-07, "loss": 1.574634838104248, "step": 2210 }, { "epoch": 2.6532813904704824, "grad_norm": 2.390625, "learning_rate": 2.0577817713631154e-07, "loss": 1.5991851806640625, "step": 2215 }, { "epoch": 2.6592747977225053, "grad_norm": 2.28125, "learning_rate": 1.9884602507463357e-07, "loss": 1.5547578811645508, "step": 2220 }, { "epoch": 2.665268204974528, "grad_norm": 3.390625, "learning_rate": 1.9202864241960927e-07, "loss": 1.5678149223327638, "step": 2225 }, { "epoch": 2.6712616122265507, "grad_norm": 2.328125, "learning_rate": 1.8532630849432131e-07, "loss": 1.564761257171631, "step": 2230 }, { "epoch": 2.6772550194785736, "grad_norm": 2.65625, "learning_rate": 1.7873929790805367e-07, "loss": 1.6035894393920898, "step": 2235 }, { "epoch": 2.6832484267305965, "grad_norm": 3.90625, "learning_rate": 1.7226788054504096e-07, "loss": 1.5928621292114258, "step": 2240 }, { "epoch": 2.689241833982619, "grad_norm": 2.25, "learning_rate": 1.6591232155341028e-07, "loss": 1.549217700958252, "step": 2245 }, { "epoch": 2.695235241234642, "grad_norm": 2.859375, "learning_rate": 1.5967288133431834e-07, "loss": 1.611344337463379, "step": 2250 }, { "epoch": 2.7012286484866648, "grad_norm": 2.90625, "learning_rate": 1.535498155312811e-07, "loss": 1.585192584991455, "step": 2255 }, { "epoch": 2.7072220557386872, "grad_norm": 3.875, "learning_rate": 1.475433750197004e-07, "loss": 1.5860268592834472, "step": 2260 }, { "epoch": 2.71321546299071, "grad_norm": 2.65625, "learning_rate": 1.416538058965853e-07, "loss": 1.4961959838867187, "step": 2265 }, { "epoch": 2.719208870242733, "grad_norm": 2.328125, "learning_rate": 1.358813494704676e-07, "loss": 1.5699254035949708, "step": 2270 }, { "epoch": 2.7252022774947555, "grad_norm": 2.71875, "learning_rate": 1.3022624225151581e-07, "loss": 1.5856217384338378, "step": 2275 }, { "epoch": 2.7311956847467784, "grad_norm": 2.515625, "learning_rate": 1.2468871594184506e-07, "loss": 1.5691574096679688, "step": 2280 }, { "epoch": 2.7371890919988013, "grad_norm": 2.703125, "learning_rate": 1.1926899742602281e-07, "loss": 1.6150981903076171, "step": 2285 }, { "epoch": 2.7431824992508242, "grad_norm": 2.328125, "learning_rate": 1.1396730876177331e-07, "loss": 1.6000024795532226, "step": 2290 }, { "epoch": 2.7491759065028467, "grad_norm": 2.328125, "learning_rate": 1.0878386717088107e-07, "loss": 1.5612345695495606, "step": 2295 }, { "epoch": 2.7551693137548696, "grad_norm": 3.140625, "learning_rate": 1.0371888503028815e-07, "loss": 1.564107322692871, "step": 2300 }, { "epoch": 2.7611627210068925, "grad_norm": 3.34375, "learning_rate": 9.877256986339334e-08, "loss": 1.576698398590088, "step": 2305 }, { "epoch": 2.7671561282589154, "grad_norm": 2.1875, "learning_rate": 9.394512433155066e-08, "loss": 1.5924538612365722, "step": 2310 }, { "epoch": 2.773149535510938, "grad_norm": 3.015625, "learning_rate": 8.923674622576505e-08, "loss": 1.5212013244628906, "step": 2315 }, { "epoch": 2.779142942762961, "grad_norm": 2.75, "learning_rate": 8.464762845858809e-08, "loss": 1.5764302253723144, "step": 2320 }, { "epoch": 2.7851363500149837, "grad_norm": 2.328125, "learning_rate": 8.01779590562154e-08, "loss": 1.5164117813110352, "step": 2325 }, { "epoch": 2.791129757267006, "grad_norm": 3.234375, "learning_rate": 7.582792115078041e-08, "loss": 1.5848722457885742, "step": 2330 }, { "epoch": 2.797123164519029, "grad_norm": 3.359375, "learning_rate": 7.159769297285313e-08, "loss": 1.6168689727783203, "step": 2335 }, { "epoch": 2.803116571771052, "grad_norm": 2.796875, "learning_rate": 6.748744784413686e-08, "loss": 1.5686025619506836, "step": 2340 }, { "epoch": 2.8091099790230745, "grad_norm": 3.421875, "learning_rate": 6.349735417036661e-08, "loss": 1.5747729301452638, "step": 2345 }, { "epoch": 2.8151033862750974, "grad_norm": 2.65625, "learning_rate": 5.962757543440966e-08, "loss": 1.5890108108520509, "step": 2350 }, { "epoch": 2.8210967935271203, "grad_norm": 2.609375, "learning_rate": 5.587827018956748e-08, "loss": 1.5725753784179688, "step": 2355 }, { "epoch": 2.8270902007791427, "grad_norm": 2.46875, "learning_rate": 5.224959205307833e-08, "loss": 1.570479679107666, "step": 2360 }, { "epoch": 2.8330836080311657, "grad_norm": 2.65625, "learning_rate": 4.874168969982429e-08, "loss": 1.5716233253479004, "step": 2365 }, { "epoch": 2.8390770152831886, "grad_norm": 2.421875, "learning_rate": 4.53547068562391e-08, "loss": 1.5775944709777832, "step": 2370 }, { "epoch": 2.845070422535211, "grad_norm": 2.015625, "learning_rate": 4.208878229441993e-08, "loss": 1.6126760482788085, "step": 2375 }, { "epoch": 2.851063829787234, "grad_norm": 2.984375, "learning_rate": 3.894404982644084e-08, "loss": 1.5892753601074219, "step": 2380 }, { "epoch": 2.857057237039257, "grad_norm": 2.578125, "learning_rate": 3.5920638298871597e-08, "loss": 1.616440773010254, "step": 2385 }, { "epoch": 2.8630506442912798, "grad_norm": 3.640625, "learning_rate": 3.301867158749683e-08, "loss": 1.534116268157959, "step": 2390 }, { "epoch": 2.869044051543302, "grad_norm": 2.9375, "learning_rate": 3.0238268592241435e-08, "loss": 1.5922479629516602, "step": 2395 }, { "epoch": 2.875037458795325, "grad_norm": 3.140625, "learning_rate": 2.7579543232299166e-08, "loss": 1.5355403900146485, "step": 2400 }, { "epoch": 2.881030866047348, "grad_norm": 2.3125, "learning_rate": 2.5042604441464666e-08, "loss": 1.6181652069091796, "step": 2405 }, { "epoch": 2.887024273299371, "grad_norm": 2.359375, "learning_rate": 2.2627556163670403e-08, "loss": 1.6138525009155273, "step": 2410 }, { "epoch": 2.8930176805513934, "grad_norm": 2.203125, "learning_rate": 2.033449734872772e-08, "loss": 1.5659173011779786, "step": 2415 }, { "epoch": 2.8990110878034163, "grad_norm": 3.34375, "learning_rate": 1.816352194827342e-08, "loss": 1.6107234954833984, "step": 2420 }, { "epoch": 2.9050044950554392, "grad_norm": 2.125, "learning_rate": 1.611471891191918e-08, "loss": 1.5589088439941405, "step": 2425 }, { "epoch": 2.9109979023074617, "grad_norm": 2.359375, "learning_rate": 1.4188172183607462e-08, "loss": 1.5950302124023437, "step": 2430 }, { "epoch": 2.9169913095594846, "grad_norm": 3.328125, "learning_rate": 1.2383960698172936e-08, "loss": 1.5727524757385254, "step": 2435 }, { "epoch": 2.9229847168115075, "grad_norm": 2.3125, "learning_rate": 1.070215837810673e-08, "loss": 1.564500904083252, "step": 2440 }, { "epoch": 2.92897812406353, "grad_norm": 2.84375, "learning_rate": 9.142834130529853e-09, "loss": 1.575014305114746, "step": 2445 }, { "epoch": 2.934971531315553, "grad_norm": 2.8125, "learning_rate": 7.706051844368123e-09, "loss": 1.5992634773254395, "step": 2450 }, { "epoch": 2.940964938567576, "grad_norm": 2.8125, "learning_rate": 6.391870387735254e-09, "loss": 1.5753752708435058, "step": 2455 }, { "epoch": 2.9469583458195983, "grad_norm": 2.703125, "learning_rate": 5.200343605520463e-09, "loss": 1.579847240447998, "step": 2460 }, { "epoch": 2.952951753071621, "grad_norm": 3.6875, "learning_rate": 4.1315203171825575e-09, "loss": 1.5922045707702637, "step": 2465 }, { "epoch": 2.958945160323644, "grad_norm": 2.75, "learning_rate": 3.185444314750541e-09, "loss": 1.578541088104248, "step": 2470 }, { "epoch": 2.9649385675756665, "grad_norm": 2.609375, "learning_rate": 2.362154361027713e-09, "loss": 1.580247974395752, "step": 2475 }, { "epoch": 2.9709319748276894, "grad_norm": 2.609375, "learning_rate": 1.6616841880049415e-09, "loss": 1.5620076179504394, "step": 2480 }, { "epoch": 2.9769253820797124, "grad_norm": 2.390625, "learning_rate": 1.0840624954771005e-09, "loss": 1.575712299346924, "step": 2485 }, { "epoch": 2.982918789331735, "grad_norm": 2.109375, "learning_rate": 6.293129498693429e-10, "loss": 1.5367554664611816, "step": 2490 }, { "epoch": 2.9889121965837577, "grad_norm": 3.234375, "learning_rate": 2.9745418326487805e-10, "loss": 1.609522819519043, "step": 2495 }, { "epoch": 2.9949056038357806, "grad_norm": 2.328125, "learning_rate": 8.849979264391461e-11, "loss": 1.609323501586914, "step": 2500 }, { "epoch": 3.0, "grad_norm": 13.5625, "learning_rate": 2.458339324440395e-12, "loss": 1.4981372833251954, "step": 2505 }, { "epoch": 3.0, "step": 2505, "total_flos": 1.2088453569639322e+18, "train_loss": 1.6374940127907638, "train_runtime": 20816.512, "train_samples_per_second": 0.481, "train_steps_per_second": 0.12 } ], "logging_steps": 5, "max_steps": 2505, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.2088453569639322e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }