{ "best_global_step": 52803, "best_metric": 1.1927706003189087, "best_model_checkpoint": "/home/rudenko/multisd/Domain-Aware-SD/outputs/drafter_understanding/checkpoint-52803", "epoch": 4.49961653174265, "eval_steps": 5867, "global_step": 52803, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 8.521516829995739e-05, "grad_norm": 9.25, "learning_rate": 0.0, "loss": 2.8716988563537598, "step": 1 }, { "epoch": 0.0008521516829995739, "grad_norm": 11.3125, "learning_rate": 5.112474437627813e-08, "loss": 3.013946533203125, "step": 10 }, { "epoch": 0.0017043033659991478, "grad_norm": 13.9375, "learning_rate": 1.0793001590547604e-07, "loss": 3.2601852416992188, "step": 20 }, { "epoch": 0.0025564550489987218, "grad_norm": 5.8125, "learning_rate": 1.6473528743467394e-07, "loss": 3.17916202545166, "step": 30 }, { "epoch": 0.0034086067319982955, "grad_norm": 11.4375, "learning_rate": 2.2154055896387187e-07, "loss": 3.127603530883789, "step": 40 }, { "epoch": 0.004260758414997869, "grad_norm": 18.125, "learning_rate": 2.7834583049306975e-07, "loss": 3.599013900756836, "step": 50 }, { "epoch": 0.0051129100979974435, "grad_norm": 6.125, "learning_rate": 3.3515110202226766e-07, "loss": 3.236738586425781, "step": 60 }, { "epoch": 0.005965061780997018, "grad_norm": 9.1875, "learning_rate": 3.919563735514656e-07, "loss": 3.556338882446289, "step": 70 }, { "epoch": 0.006817213463996591, "grad_norm": 12.6875, "learning_rate": 4.487616450806635e-07, "loss": 3.2733489990234377, "step": 80 }, { "epoch": 0.007669365146996165, "grad_norm": 7.40625, "learning_rate": 5.055669166098614e-07, "loss": 3.0490550994873047, "step": 90 }, { "epoch": 0.008521516829995739, "grad_norm": 13.0, "learning_rate": 5.623721881390593e-07, "loss": 3.0611867904663086, "step": 100 }, { "epoch": 0.009373668512995313, "grad_norm": 10.5, "learning_rate": 6.191774596682572e-07, "loss": 3.355224609375, "step": 110 }, { "epoch": 0.010225820195994887, "grad_norm": 9.1875, "learning_rate": 6.759827311974552e-07, "loss": 3.2279945373535157, "step": 120 }, { "epoch": 0.011077971878994461, "grad_norm": 5.84375, "learning_rate": 7.32788002726653e-07, "loss": 3.311595916748047, "step": 130 }, { "epoch": 0.011930123561994035, "grad_norm": 7.90625, "learning_rate": 7.895932742558509e-07, "loss": 3.4494754791259767, "step": 140 }, { "epoch": 0.01278227524499361, "grad_norm": 9.9375, "learning_rate": 8.46398545785049e-07, "loss": 3.3572383880615235, "step": 150 }, { "epoch": 0.013634426927993182, "grad_norm": 29.25, "learning_rate": 9.032038173142469e-07, "loss": 3.3135917663574217, "step": 160 }, { "epoch": 0.014486578610992756, "grad_norm": 8.0625, "learning_rate": 9.600090888434448e-07, "loss": 3.422383117675781, "step": 170 }, { "epoch": 0.01533873029399233, "grad_norm": 12.0, "learning_rate": 1.0168143603726426e-06, "loss": 3.340637969970703, "step": 180 }, { "epoch": 0.016190881976991905, "grad_norm": 12.125, "learning_rate": 1.0736196319018406e-06, "loss": 3.156867790222168, "step": 190 }, { "epoch": 0.017043033659991477, "grad_norm": 14.3125, "learning_rate": 1.1304249034310384e-06, "loss": 3.479510498046875, "step": 200 }, { "epoch": 0.017895185342991053, "grad_norm": 11.6875, "learning_rate": 1.1872301749602364e-06, "loss": 3.1641048431396483, "step": 210 }, { "epoch": 0.018747337025990626, "grad_norm": 19.0, "learning_rate": 1.2440354464894344e-06, "loss": 3.289683532714844, "step": 220 }, { "epoch": 0.0195994887089902, "grad_norm": 14.0, "learning_rate": 1.3008407180186322e-06, "loss": 3.2887969970703126, "step": 230 }, { "epoch": 0.020451640391989774, "grad_norm": 4.15625, "learning_rate": 1.35764598954783e-06, "loss": 2.8173423767089845, "step": 240 }, { "epoch": 0.021303792074989347, "grad_norm": 7.6875, "learning_rate": 1.414451261077028e-06, "loss": 3.5577136993408205, "step": 250 }, { "epoch": 0.022155943757988922, "grad_norm": 15.8125, "learning_rate": 1.471256532606226e-06, "loss": 3.3060630798339843, "step": 260 }, { "epoch": 0.023008095440988495, "grad_norm": 6.21875, "learning_rate": 1.5280618041354238e-06, "loss": 3.1209083557128907, "step": 270 }, { "epoch": 0.02386024712398807, "grad_norm": 21.125, "learning_rate": 1.5848670756646218e-06, "loss": 3.4565467834472656, "step": 280 }, { "epoch": 0.024712398806987643, "grad_norm": 6.40625, "learning_rate": 1.6416723471938196e-06, "loss": 3.513861083984375, "step": 290 }, { "epoch": 0.02556455048998722, "grad_norm": 16.375, "learning_rate": 1.6984776187230176e-06, "loss": 3.2077957153320313, "step": 300 }, { "epoch": 0.026416702172986792, "grad_norm": 7.53125, "learning_rate": 1.7552828902522154e-06, "loss": 3.248636245727539, "step": 310 }, { "epoch": 0.027268853855986364, "grad_norm": 29.5, "learning_rate": 1.8120881617814132e-06, "loss": 3.502519226074219, "step": 320 }, { "epoch": 0.02812100553898594, "grad_norm": 6.25, "learning_rate": 1.8688934333106114e-06, "loss": 3.133312225341797, "step": 330 }, { "epoch": 0.028973157221985513, "grad_norm": 5.59375, "learning_rate": 1.925698704839809e-06, "loss": 3.360553741455078, "step": 340 }, { "epoch": 0.02982530890498509, "grad_norm": 11.5625, "learning_rate": 1.9825039763690075e-06, "loss": 3.4204391479492187, "step": 350 }, { "epoch": 0.03067746058798466, "grad_norm": 10.625, "learning_rate": 2.039309247898205e-06, "loss": 3.517983627319336, "step": 360 }, { "epoch": 0.031529612270984234, "grad_norm": 8.5, "learning_rate": 2.096114519427403e-06, "loss": 3.1697967529296873, "step": 370 }, { "epoch": 0.03238176395398381, "grad_norm": 12.625, "learning_rate": 2.152919790956601e-06, "loss": 3.110427665710449, "step": 380 }, { "epoch": 0.033233915636983385, "grad_norm": 9.125, "learning_rate": 2.2097250624857987e-06, "loss": 3.1533700942993166, "step": 390 }, { "epoch": 0.034086067319982954, "grad_norm": 8.75, "learning_rate": 2.2665303340149967e-06, "loss": 3.4042823791503904, "step": 400 }, { "epoch": 0.03493821900298253, "grad_norm": 6.34375, "learning_rate": 2.3233356055441947e-06, "loss": 3.153865623474121, "step": 410 }, { "epoch": 0.035790370685982106, "grad_norm": 7.375, "learning_rate": 2.3801408770733927e-06, "loss": 3.0755475997924804, "step": 420 }, { "epoch": 0.036642522368981675, "grad_norm": 6.40625, "learning_rate": 2.4369461486025903e-06, "loss": 3.6919315338134764, "step": 430 }, { "epoch": 0.03749467405198125, "grad_norm": 4.59375, "learning_rate": 2.4937514201317883e-06, "loss": 3.340152359008789, "step": 440 }, { "epoch": 0.03834682573498083, "grad_norm": 21.375, "learning_rate": 2.5505566916609863e-06, "loss": 3.1449905395507813, "step": 450 }, { "epoch": 0.0391989774179804, "grad_norm": 7.65625, "learning_rate": 2.607361963190184e-06, "loss": 2.9701929092407227, "step": 460 }, { "epoch": 0.04005112910097997, "grad_norm": 6.40625, "learning_rate": 2.6641672347193823e-06, "loss": 3.0714496612548827, "step": 470 }, { "epoch": 0.04090328078397955, "grad_norm": 5.125, "learning_rate": 2.72097250624858e-06, "loss": 3.1570396423339844, "step": 480 }, { "epoch": 0.041755432466979124, "grad_norm": 9.875, "learning_rate": 2.777777777777778e-06, "loss": 3.134921073913574, "step": 490 }, { "epoch": 0.04260758414997869, "grad_norm": 4.6875, "learning_rate": 2.834583049306976e-06, "loss": 3.0297357559204103, "step": 500 }, { "epoch": 0.04345973583297827, "grad_norm": 7.125, "learning_rate": 2.8913883208361735e-06, "loss": 2.898138236999512, "step": 510 }, { "epoch": 0.044311887515977845, "grad_norm": 9.0, "learning_rate": 2.948193592365372e-06, "loss": 3.1040470123291017, "step": 520 }, { "epoch": 0.04516403919897742, "grad_norm": 11.5625, "learning_rate": 3.0049988638945695e-06, "loss": 3.155125045776367, "step": 530 }, { "epoch": 0.04601619088197699, "grad_norm": 6.09375, "learning_rate": 3.0618041354237675e-06, "loss": 3.0171092987060546, "step": 540 }, { "epoch": 0.046868342564976566, "grad_norm": 6.6875, "learning_rate": 3.1186094069529655e-06, "loss": 3.0729501724243162, "step": 550 }, { "epoch": 0.04772049424797614, "grad_norm": 7.4375, "learning_rate": 3.175414678482163e-06, "loss": 3.117463493347168, "step": 560 }, { "epoch": 0.04857264593097571, "grad_norm": 19.875, "learning_rate": 3.2322199500113616e-06, "loss": 3.194751739501953, "step": 570 }, { "epoch": 0.04942479761397529, "grad_norm": 17.0, "learning_rate": 3.289025221540559e-06, "loss": 3.0364763259887697, "step": 580 }, { "epoch": 0.05027694929697486, "grad_norm": 10.3125, "learning_rate": 3.345830493069757e-06, "loss": 3.1409191131591796, "step": 590 }, { "epoch": 0.05112910097997444, "grad_norm": 12.6875, "learning_rate": 3.4026357645989548e-06, "loss": 3.1454530715942384, "step": 600 }, { "epoch": 0.05198125266297401, "grad_norm": 10.0625, "learning_rate": 3.4594410361281528e-06, "loss": 3.194963073730469, "step": 610 }, { "epoch": 0.052833404345973584, "grad_norm": 9.625, "learning_rate": 3.516246307657351e-06, "loss": 3.0313894271850588, "step": 620 }, { "epoch": 0.05368555602897316, "grad_norm": 13.3125, "learning_rate": 3.5730515791865488e-06, "loss": 2.9854049682617188, "step": 630 }, { "epoch": 0.05453770771197273, "grad_norm": 7.90625, "learning_rate": 3.629856850715747e-06, "loss": 2.8216991424560547, "step": 640 }, { "epoch": 0.055389859394972304, "grad_norm": 9.875, "learning_rate": 3.6866621222449444e-06, "loss": 2.9064985275268556, "step": 650 }, { "epoch": 0.05624201107797188, "grad_norm": 14.3125, "learning_rate": 3.7434673937741424e-06, "loss": 3.093683624267578, "step": 660 }, { "epoch": 0.057094162760971456, "grad_norm": 3.78125, "learning_rate": 3.80027266530334e-06, "loss": 3.085400390625, "step": 670 }, { "epoch": 0.057946314443971025, "grad_norm": 11.4375, "learning_rate": 3.857077936832539e-06, "loss": 2.9046403884887697, "step": 680 }, { "epoch": 0.0587984661269706, "grad_norm": 8.5, "learning_rate": 3.913883208361736e-06, "loss": 3.032069778442383, "step": 690 }, { "epoch": 0.05965061780997018, "grad_norm": 4.40625, "learning_rate": 3.970688479890934e-06, "loss": 3.2747734069824217, "step": 700 }, { "epoch": 0.060502769492969746, "grad_norm": 4.40625, "learning_rate": 4.027493751420132e-06, "loss": 3.0017154693603514, "step": 710 }, { "epoch": 0.06135492117596932, "grad_norm": 7.40625, "learning_rate": 4.08429902294933e-06, "loss": 2.745909309387207, "step": 720 }, { "epoch": 0.0622070728589689, "grad_norm": 4.46875, "learning_rate": 4.141104294478528e-06, "loss": 2.9790897369384766, "step": 730 }, { "epoch": 0.06305922454196847, "grad_norm": 4.625, "learning_rate": 4.197909566007726e-06, "loss": 2.7895235061645507, "step": 740 }, { "epoch": 0.06391137622496805, "grad_norm": 4.3125, "learning_rate": 4.254714837536924e-06, "loss": 2.836431312561035, "step": 750 }, { "epoch": 0.06476352790796762, "grad_norm": 9.5, "learning_rate": 4.311520109066121e-06, "loss": 2.8946548461914063, "step": 760 }, { "epoch": 0.06561567959096719, "grad_norm": 5.90625, "learning_rate": 4.36832538059532e-06, "loss": 2.680698776245117, "step": 770 }, { "epoch": 0.06646783127396677, "grad_norm": 7.625, "learning_rate": 4.425130652124517e-06, "loss": 2.7272478103637696, "step": 780 }, { "epoch": 0.06731998295696634, "grad_norm": 5.53125, "learning_rate": 4.481935923653716e-06, "loss": 2.6610416412353515, "step": 790 }, { "epoch": 0.06817213463996591, "grad_norm": 4.21875, "learning_rate": 4.538741195182913e-06, "loss": 2.708527755737305, "step": 800 }, { "epoch": 0.06902428632296549, "grad_norm": 5.0, "learning_rate": 4.595546466712111e-06, "loss": 2.913991355895996, "step": 810 }, { "epoch": 0.06987643800596506, "grad_norm": 10.875, "learning_rate": 4.652351738241309e-06, "loss": 2.884418487548828, "step": 820 }, { "epoch": 0.07072858968896463, "grad_norm": 3.734375, "learning_rate": 4.709157009770507e-06, "loss": 2.736153411865234, "step": 830 }, { "epoch": 0.07158074137196421, "grad_norm": 5.90625, "learning_rate": 4.7659622812997045e-06, "loss": 2.6700347900390624, "step": 840 }, { "epoch": 0.07243289305496378, "grad_norm": 4.5, "learning_rate": 4.822767552828903e-06, "loss": 2.703378677368164, "step": 850 }, { "epoch": 0.07328504473796335, "grad_norm": 6.15625, "learning_rate": 4.8795728243581005e-06, "loss": 2.880733680725098, "step": 860 }, { "epoch": 0.07413719642096293, "grad_norm": 7.34375, "learning_rate": 4.936378095887299e-06, "loss": 2.8227672576904297, "step": 870 }, { "epoch": 0.0749893481039625, "grad_norm": 9.1875, "learning_rate": 4.9931833674164965e-06, "loss": 2.8054622650146483, "step": 880 }, { "epoch": 0.07584149978696209, "grad_norm": 3.640625, "learning_rate": 5.049988638945694e-06, "loss": 2.5370790481567385, "step": 890 }, { "epoch": 0.07669365146996165, "grad_norm": 4.53125, "learning_rate": 5.106793910474892e-06, "loss": 2.761469268798828, "step": 900 }, { "epoch": 0.07754580315296122, "grad_norm": 4.0625, "learning_rate": 5.16359918200409e-06, "loss": 2.5250988006591797, "step": 910 }, { "epoch": 0.0783979548359608, "grad_norm": 8.125, "learning_rate": 5.2204044535332885e-06, "loss": 2.567232131958008, "step": 920 }, { "epoch": 0.07925010651896038, "grad_norm": 5.0, "learning_rate": 5.277209725062486e-06, "loss": 2.5968038558959963, "step": 930 }, { "epoch": 0.08010225820195994, "grad_norm": 3.609375, "learning_rate": 5.334014996591684e-06, "loss": 2.5281293869018553, "step": 940 }, { "epoch": 0.08095440988495953, "grad_norm": 7.78125, "learning_rate": 5.390820268120881e-06, "loss": 2.7589054107666016, "step": 950 }, { "epoch": 0.0818065615679591, "grad_norm": 9.375, "learning_rate": 5.44762553965008e-06, "loss": 2.593463897705078, "step": 960 }, { "epoch": 0.08265871325095867, "grad_norm": 2.5625, "learning_rate": 5.504430811179278e-06, "loss": 2.383139801025391, "step": 970 }, { "epoch": 0.08351086493395825, "grad_norm": 3.3125, "learning_rate": 5.561236082708476e-06, "loss": 2.755172920227051, "step": 980 }, { "epoch": 0.08436301661695782, "grad_norm": 5.1875, "learning_rate": 5.618041354237673e-06, "loss": 2.5591882705688476, "step": 990 }, { "epoch": 0.08521516829995739, "grad_norm": 4.125, "learning_rate": 5.674846625766871e-06, "loss": 2.5151851654052733, "step": 1000 }, { "epoch": 0.08606731998295697, "grad_norm": 3.828125, "learning_rate": 5.731651897296069e-06, "loss": 2.4544889450073244, "step": 1010 }, { "epoch": 0.08691947166595654, "grad_norm": 5.875, "learning_rate": 5.788457168825268e-06, "loss": 2.53830509185791, "step": 1020 }, { "epoch": 0.08777162334895612, "grad_norm": 4.0625, "learning_rate": 5.845262440354465e-06, "loss": 2.3819211959838866, "step": 1030 }, { "epoch": 0.08862377503195569, "grad_norm": 4.1875, "learning_rate": 5.902067711883663e-06, "loss": 2.369676399230957, "step": 1040 }, { "epoch": 0.08947592671495526, "grad_norm": 5.3125, "learning_rate": 5.9588729834128606e-06, "loss": 2.6290531158447266, "step": 1050 }, { "epoch": 0.09032807839795484, "grad_norm": 3.34375, "learning_rate": 6.015678254942059e-06, "loss": 2.535698127746582, "step": 1060 }, { "epoch": 0.09118023008095441, "grad_norm": 4.125, "learning_rate": 6.0724835264712566e-06, "loss": 2.5194089889526365, "step": 1070 }, { "epoch": 0.09203238176395398, "grad_norm": 4.9375, "learning_rate": 6.129288798000455e-06, "loss": 2.568607711791992, "step": 1080 }, { "epoch": 0.09288453344695356, "grad_norm": 3.71875, "learning_rate": 6.186094069529653e-06, "loss": 2.5331394195556642, "step": 1090 }, { "epoch": 0.09373668512995313, "grad_norm": 3.625, "learning_rate": 6.24289934105885e-06, "loss": 2.5265005111694334, "step": 1100 }, { "epoch": 0.0945888368129527, "grad_norm": 2.859375, "learning_rate": 6.2997046125880495e-06, "loss": 2.39256591796875, "step": 1110 }, { "epoch": 0.09544098849595228, "grad_norm": 2.609375, "learning_rate": 6.356509884117247e-06, "loss": 2.504415512084961, "step": 1120 }, { "epoch": 0.09629314017895185, "grad_norm": 3.28125, "learning_rate": 6.413315155646445e-06, "loss": 2.364087677001953, "step": 1130 }, { "epoch": 0.09714529186195142, "grad_norm": 2.4375, "learning_rate": 6.470120427175642e-06, "loss": 2.458033561706543, "step": 1140 }, { "epoch": 0.097997443544951, "grad_norm": 2.4375, "learning_rate": 6.52692569870484e-06, "loss": 2.42427921295166, "step": 1150 }, { "epoch": 0.09884959522795057, "grad_norm": 3.453125, "learning_rate": 6.583730970234038e-06, "loss": 2.4474485397338865, "step": 1160 }, { "epoch": 0.09970174691095014, "grad_norm": 3.796875, "learning_rate": 6.640536241763236e-06, "loss": 2.469664764404297, "step": 1170 }, { "epoch": 0.10055389859394973, "grad_norm": 4.03125, "learning_rate": 6.697341513292433e-06, "loss": 2.505742645263672, "step": 1180 }, { "epoch": 0.1014060502769493, "grad_norm": 3.109375, "learning_rate": 6.754146784821631e-06, "loss": 2.4006547927856445, "step": 1190 }, { "epoch": 0.10225820195994888, "grad_norm": 2.90625, "learning_rate": 6.8109520563508294e-06, "loss": 2.5564064025878905, "step": 1200 }, { "epoch": 0.10311035364294845, "grad_norm": 4.5625, "learning_rate": 6.867757327880027e-06, "loss": 2.4268138885498045, "step": 1210 }, { "epoch": 0.10396250532594802, "grad_norm": 3.125, "learning_rate": 6.924562599409226e-06, "loss": 2.513792610168457, "step": 1220 }, { "epoch": 0.1048146570089476, "grad_norm": 2.578125, "learning_rate": 6.981367870938424e-06, "loss": 2.2736175537109373, "step": 1230 }, { "epoch": 0.10566680869194717, "grad_norm": 4.375, "learning_rate": 7.0381731424676215e-06, "loss": 2.476356315612793, "step": 1240 }, { "epoch": 0.10651896037494674, "grad_norm": 2.671875, "learning_rate": 7.09497841399682e-06, "loss": 2.4220767974853517, "step": 1250 }, { "epoch": 0.10737111205794632, "grad_norm": 1.90625, "learning_rate": 7.1517836855260175e-06, "loss": 2.377312660217285, "step": 1260 }, { "epoch": 0.10822326374094589, "grad_norm": 3.125, "learning_rate": 7.208588957055215e-06, "loss": 2.303727149963379, "step": 1270 }, { "epoch": 0.10907541542394546, "grad_norm": 3.5, "learning_rate": 7.265394228584413e-06, "loss": 2.3555429458618162, "step": 1280 }, { "epoch": 0.10992756710694504, "grad_norm": 2.671875, "learning_rate": 7.32219950011361e-06, "loss": 2.5191627502441407, "step": 1290 }, { "epoch": 0.11077971878994461, "grad_norm": 2.484375, "learning_rate": 7.379004771642809e-06, "loss": 2.469433403015137, "step": 1300 }, { "epoch": 0.11163187047294418, "grad_norm": 5.15625, "learning_rate": 7.435810043172006e-06, "loss": 2.494273376464844, "step": 1310 }, { "epoch": 0.11248402215594376, "grad_norm": 2.671875, "learning_rate": 7.492615314701204e-06, "loss": 2.3361087799072267, "step": 1320 }, { "epoch": 0.11333617383894333, "grad_norm": 2.53125, "learning_rate": 7.549420586230403e-06, "loss": 2.5023555755615234, "step": 1330 }, { "epoch": 0.11418832552194291, "grad_norm": 4.1875, "learning_rate": 7.606225857759601e-06, "loss": 2.4269634246826173, "step": 1340 }, { "epoch": 0.11504047720494248, "grad_norm": 2.109375, "learning_rate": 7.6630311292888e-06, "loss": 2.434812545776367, "step": 1350 }, { "epoch": 0.11589262888794205, "grad_norm": 2.28125, "learning_rate": 7.719836400817997e-06, "loss": 2.400814247131348, "step": 1360 }, { "epoch": 0.11674478057094163, "grad_norm": 4.84375, "learning_rate": 7.776641672347194e-06, "loss": 2.5161184310913085, "step": 1370 }, { "epoch": 0.1175969322539412, "grad_norm": 3.34375, "learning_rate": 7.833446943876392e-06, "loss": 2.4580944061279295, "step": 1380 }, { "epoch": 0.11844908393694077, "grad_norm": 4.1875, "learning_rate": 7.89025221540559e-06, "loss": 2.296352195739746, "step": 1390 }, { "epoch": 0.11930123561994035, "grad_norm": 4.34375, "learning_rate": 7.947057486934787e-06, "loss": 2.2461688995361326, "step": 1400 }, { "epoch": 0.12015338730293992, "grad_norm": 4.65625, "learning_rate": 8.003862758463986e-06, "loss": 2.352525329589844, "step": 1410 }, { "epoch": 0.12100553898593949, "grad_norm": 2.484375, "learning_rate": 8.060668029993184e-06, "loss": 2.3903362274169924, "step": 1420 }, { "epoch": 0.12185769066893908, "grad_norm": 4.4375, "learning_rate": 8.117473301522382e-06, "loss": 2.301466369628906, "step": 1430 }, { "epoch": 0.12270984235193864, "grad_norm": 4.78125, "learning_rate": 8.174278573051579e-06, "loss": 2.2676326751708986, "step": 1440 }, { "epoch": 0.12356199403493821, "grad_norm": 3.25, "learning_rate": 8.231083844580778e-06, "loss": 2.3397890090942384, "step": 1450 }, { "epoch": 0.1244141457179378, "grad_norm": 6.71875, "learning_rate": 8.287889116109976e-06, "loss": 2.3867767333984373, "step": 1460 }, { "epoch": 0.12526629740093737, "grad_norm": 4.875, "learning_rate": 8.344694387639174e-06, "loss": 2.355057144165039, "step": 1470 }, { "epoch": 0.12611844908393693, "grad_norm": 4.0625, "learning_rate": 8.401499659168371e-06, "loss": 2.2904050827026365, "step": 1480 }, { "epoch": 0.1269706007669365, "grad_norm": 2.484375, "learning_rate": 8.458304930697569e-06, "loss": 2.3388200759887696, "step": 1490 }, { "epoch": 0.1278227524499361, "grad_norm": 4.78125, "learning_rate": 8.515110202226766e-06, "loss": 2.3145654678344725, "step": 1500 }, { "epoch": 0.12867490413293567, "grad_norm": 2.0, "learning_rate": 8.571915473755966e-06, "loss": 2.3789836883544924, "step": 1510 }, { "epoch": 0.12952705581593524, "grad_norm": 4.96875, "learning_rate": 8.628720745285163e-06, "loss": 2.3837446212768554, "step": 1520 }, { "epoch": 0.1303792074989348, "grad_norm": 2.25, "learning_rate": 8.68552601681436e-06, "loss": 2.408094024658203, "step": 1530 }, { "epoch": 0.13123135918193438, "grad_norm": 4.0625, "learning_rate": 8.742331288343558e-06, "loss": 2.289523887634277, "step": 1540 }, { "epoch": 0.13208351086493395, "grad_norm": 3.734375, "learning_rate": 8.799136559872756e-06, "loss": 2.343811798095703, "step": 1550 }, { "epoch": 0.13293566254793354, "grad_norm": 3.171875, "learning_rate": 8.855941831401955e-06, "loss": 2.4515659332275392, "step": 1560 }, { "epoch": 0.1337878142309331, "grad_norm": 6.59375, "learning_rate": 8.912747102931153e-06, "loss": 2.410831832885742, "step": 1570 }, { "epoch": 0.13463996591393268, "grad_norm": 3.765625, "learning_rate": 8.96955237446035e-06, "loss": 2.3500961303710937, "step": 1580 }, { "epoch": 0.13549211759693225, "grad_norm": 3.09375, "learning_rate": 9.026357645989548e-06, "loss": 2.336482620239258, "step": 1590 }, { "epoch": 0.13634426927993182, "grad_norm": 3.9375, "learning_rate": 9.083162917518746e-06, "loss": 2.3367528915405273, "step": 1600 }, { "epoch": 0.13719642096293141, "grad_norm": 4.09375, "learning_rate": 9.139968189047945e-06, "loss": 2.3118993759155275, "step": 1610 }, { "epoch": 0.13804857264593098, "grad_norm": 3.0625, "learning_rate": 9.196773460577142e-06, "loss": 2.2934507369995116, "step": 1620 }, { "epoch": 0.13890072432893055, "grad_norm": 3.53125, "learning_rate": 9.25357873210634e-06, "loss": 2.3574174880981444, "step": 1630 }, { "epoch": 0.13975287601193012, "grad_norm": 4.34375, "learning_rate": 9.310384003635538e-06, "loss": 2.322239303588867, "step": 1640 }, { "epoch": 0.1406050276949297, "grad_norm": 7.34375, "learning_rate": 9.367189275164735e-06, "loss": 2.3842708587646486, "step": 1650 }, { "epoch": 0.14145717937792926, "grad_norm": 4.21875, "learning_rate": 9.423994546693933e-06, "loss": 2.3610225677490235, "step": 1660 }, { "epoch": 0.14230933106092886, "grad_norm": 3.484375, "learning_rate": 9.480799818223132e-06, "loss": 2.3907087326049803, "step": 1670 }, { "epoch": 0.14316148274392843, "grad_norm": 3.796875, "learning_rate": 9.53760508975233e-06, "loss": 2.3129953384399413, "step": 1680 }, { "epoch": 0.144013634426928, "grad_norm": 3.25, "learning_rate": 9.594410361281527e-06, "loss": 2.3283670425415037, "step": 1690 }, { "epoch": 0.14486578610992756, "grad_norm": 4.34375, "learning_rate": 9.651215632810725e-06, "loss": 2.3480297088623048, "step": 1700 }, { "epoch": 0.14571793779292713, "grad_norm": 4.4375, "learning_rate": 9.708020904339924e-06, "loss": 2.2657470703125, "step": 1710 }, { "epoch": 0.1465700894759267, "grad_norm": 2.96875, "learning_rate": 9.764826175869122e-06, "loss": 2.3250925064086916, "step": 1720 }, { "epoch": 0.1474222411589263, "grad_norm": 3.28125, "learning_rate": 9.82163144739832e-06, "loss": 2.3477096557617188, "step": 1730 }, { "epoch": 0.14827439284192587, "grad_norm": 4.0, "learning_rate": 9.878436718927517e-06, "loss": 2.2694833755493162, "step": 1740 }, { "epoch": 0.14912654452492544, "grad_norm": 3.921875, "learning_rate": 9.935241990456714e-06, "loss": 2.2494880676269533, "step": 1750 }, { "epoch": 0.149978696207925, "grad_norm": 2.984375, "learning_rate": 9.992047261985912e-06, "loss": 2.2676013946533202, "step": 1760 }, { "epoch": 0.15083084789092457, "grad_norm": 3.984375, "learning_rate": 1.004885253351511e-05, "loss": 2.3883506774902346, "step": 1770 }, { "epoch": 0.15168299957392417, "grad_norm": 2.0625, "learning_rate": 1.0105657805044307e-05, "loss": 2.2046146392822266, "step": 1780 }, { "epoch": 0.15253515125692374, "grad_norm": 3.6875, "learning_rate": 1.0162463076573507e-05, "loss": 2.338265800476074, "step": 1790 }, { "epoch": 0.1533873029399233, "grad_norm": 3.90625, "learning_rate": 1.0219268348102704e-05, "loss": 2.1492069244384764, "step": 1800 }, { "epoch": 0.15423945462292288, "grad_norm": 2.75, "learning_rate": 1.0276073619631903e-05, "loss": 2.3164609909057616, "step": 1810 }, { "epoch": 0.15509160630592245, "grad_norm": 4.78125, "learning_rate": 1.0332878891161101e-05, "loss": 2.2892261505126954, "step": 1820 }, { "epoch": 0.15594375798892202, "grad_norm": 2.34375, "learning_rate": 1.0389684162690299e-05, "loss": 2.1917913436889647, "step": 1830 }, { "epoch": 0.1567959096719216, "grad_norm": 2.421875, "learning_rate": 1.0446489434219496e-05, "loss": 2.2990839004516603, "step": 1840 }, { "epoch": 0.15764806135492118, "grad_norm": 2.796875, "learning_rate": 1.0503294705748694e-05, "loss": 2.2146095275878905, "step": 1850 }, { "epoch": 0.15850021303792075, "grad_norm": 2.609375, "learning_rate": 1.0560099977277891e-05, "loss": 2.147475814819336, "step": 1860 }, { "epoch": 0.15935236472092032, "grad_norm": 3.328125, "learning_rate": 1.0616905248807089e-05, "loss": 2.2194091796875, "step": 1870 }, { "epoch": 0.1602045164039199, "grad_norm": 2.671875, "learning_rate": 1.0673710520336286e-05, "loss": 2.2038217544555665, "step": 1880 }, { "epoch": 0.16105666808691946, "grad_norm": 2.15625, "learning_rate": 1.0730515791865486e-05, "loss": 2.2554765701293946, "step": 1890 }, { "epoch": 0.16190881976991905, "grad_norm": 2.84375, "learning_rate": 1.0787321063394683e-05, "loss": 2.1494558334350584, "step": 1900 }, { "epoch": 0.16276097145291862, "grad_norm": 2.21875, "learning_rate": 1.0844126334923883e-05, "loss": 2.210985565185547, "step": 1910 }, { "epoch": 0.1636131231359182, "grad_norm": 3.09375, "learning_rate": 1.090093160645308e-05, "loss": 2.18758544921875, "step": 1920 }, { "epoch": 0.16446527481891776, "grad_norm": 2.953125, "learning_rate": 1.0957736877982278e-05, "loss": 2.1813117980957033, "step": 1930 }, { "epoch": 0.16531742650191733, "grad_norm": 3.546875, "learning_rate": 1.1014542149511475e-05, "loss": 2.128385543823242, "step": 1940 }, { "epoch": 0.16616957818491693, "grad_norm": 3.046875, "learning_rate": 1.1071347421040673e-05, "loss": 2.2347604751586916, "step": 1950 }, { "epoch": 0.1670217298679165, "grad_norm": 3.03125, "learning_rate": 1.112815269256987e-05, "loss": 2.149306297302246, "step": 1960 }, { "epoch": 0.16787388155091607, "grad_norm": 7.03125, "learning_rate": 1.1184957964099068e-05, "loss": 2.139059638977051, "step": 1970 }, { "epoch": 0.16872603323391563, "grad_norm": 3.9375, "learning_rate": 1.1241763235628266e-05, "loss": 2.172390365600586, "step": 1980 }, { "epoch": 0.1695781849169152, "grad_norm": 6.96875, "learning_rate": 1.1298568507157465e-05, "loss": 2.182741165161133, "step": 1990 }, { "epoch": 0.17043033659991477, "grad_norm": 4.28125, "learning_rate": 1.1355373778686663e-05, "loss": 2.2214101791381835, "step": 2000 }, { "epoch": 0.17128248828291437, "grad_norm": 3.71875, "learning_rate": 1.1412179050215862e-05, "loss": 2.223723602294922, "step": 2010 }, { "epoch": 0.17213463996591394, "grad_norm": 2.171875, "learning_rate": 1.146898432174506e-05, "loss": 2.146739387512207, "step": 2020 }, { "epoch": 0.1729867916489135, "grad_norm": 3.796875, "learning_rate": 1.1525789593274257e-05, "loss": 2.155897521972656, "step": 2030 }, { "epoch": 0.17383894333191308, "grad_norm": 2.65625, "learning_rate": 1.1582594864803455e-05, "loss": 2.1646808624267577, "step": 2040 }, { "epoch": 0.17469109501491265, "grad_norm": 2.25, "learning_rate": 1.1639400136332652e-05, "loss": 2.1567047119140623, "step": 2050 }, { "epoch": 0.17554324669791224, "grad_norm": 3.546875, "learning_rate": 1.169620540786185e-05, "loss": 2.1691286087036135, "step": 2060 }, { "epoch": 0.1763953983809118, "grad_norm": 3.109375, "learning_rate": 1.1753010679391047e-05, "loss": 2.1327997207641602, "step": 2070 }, { "epoch": 0.17724755006391138, "grad_norm": 1.9140625, "learning_rate": 1.1809815950920245e-05, "loss": 2.093043899536133, "step": 2080 }, { "epoch": 0.17809970174691095, "grad_norm": 3.078125, "learning_rate": 1.1866621222449444e-05, "loss": 2.1950441360473634, "step": 2090 }, { "epoch": 0.17895185342991052, "grad_norm": 6.75, "learning_rate": 1.1923426493978642e-05, "loss": 2.107472229003906, "step": 2100 }, { "epoch": 0.1798040051129101, "grad_norm": 3.546875, "learning_rate": 1.198023176550784e-05, "loss": 2.097618293762207, "step": 2110 }, { "epoch": 0.18065615679590968, "grad_norm": 3.109375, "learning_rate": 1.2037037037037037e-05, "loss": 2.1017202377319335, "step": 2120 }, { "epoch": 0.18150830847890925, "grad_norm": 4.59375, "learning_rate": 1.2093842308566236e-05, "loss": 2.1457643508911133, "step": 2130 }, { "epoch": 0.18236046016190882, "grad_norm": 3.671875, "learning_rate": 1.2150647580095434e-05, "loss": 2.052163314819336, "step": 2140 }, { "epoch": 0.1832126118449084, "grad_norm": 3.265625, "learning_rate": 1.2207452851624632e-05, "loss": 2.2045967102050783, "step": 2150 }, { "epoch": 0.18406476352790796, "grad_norm": 3.765625, "learning_rate": 1.2264258123153829e-05, "loss": 2.1532943725585936, "step": 2160 }, { "epoch": 0.18491691521090753, "grad_norm": 3.78125, "learning_rate": 1.2321063394683027e-05, "loss": 2.114980125427246, "step": 2170 }, { "epoch": 0.18576906689390713, "grad_norm": 3.3125, "learning_rate": 1.2377868666212224e-05, "loss": 2.055535888671875, "step": 2180 }, { "epoch": 0.1866212185769067, "grad_norm": 2.15625, "learning_rate": 1.2434673937741424e-05, "loss": 2.046905517578125, "step": 2190 }, { "epoch": 0.18747337025990626, "grad_norm": 2.1875, "learning_rate": 1.2491479209270621e-05, "loss": 2.054991912841797, "step": 2200 }, { "epoch": 0.18832552194290583, "grad_norm": 2.75, "learning_rate": 1.2548284480799819e-05, "loss": 1.9819568634033202, "step": 2210 }, { "epoch": 0.1891776736259054, "grad_norm": 4.0625, "learning_rate": 1.2605089752329016e-05, "loss": 1.9915426254272461, "step": 2220 }, { "epoch": 0.190029825308905, "grad_norm": 2.578125, "learning_rate": 1.2661895023858214e-05, "loss": 1.9069635391235351, "step": 2230 }, { "epoch": 0.19088197699190457, "grad_norm": 3.203125, "learning_rate": 1.2718700295387411e-05, "loss": 1.9405784606933594, "step": 2240 }, { "epoch": 0.19173412867490414, "grad_norm": 6.34375, "learning_rate": 1.2775505566916609e-05, "loss": 2.051985740661621, "step": 2250 }, { "epoch": 0.1925862803579037, "grad_norm": 3.078125, "learning_rate": 1.2832310838445807e-05, "loss": 2.0342660903930665, "step": 2260 }, { "epoch": 0.19343843204090327, "grad_norm": 2.5, "learning_rate": 1.2889116109975006e-05, "loss": 2.1325550079345703, "step": 2270 }, { "epoch": 0.19429058372390284, "grad_norm": 4.4375, "learning_rate": 1.2945921381504204e-05, "loss": 2.0382659912109373, "step": 2280 }, { "epoch": 0.19514273540690244, "grad_norm": 4.03125, "learning_rate": 1.3002726653033401e-05, "loss": 2.064997100830078, "step": 2290 }, { "epoch": 0.195994887089902, "grad_norm": 2.625, "learning_rate": 1.3059531924562602e-05, "loss": 2.0311309814453127, "step": 2300 }, { "epoch": 0.19684703877290158, "grad_norm": 3.515625, "learning_rate": 1.31163371960918e-05, "loss": 2.0400503158569334, "step": 2310 }, { "epoch": 0.19769919045590115, "grad_norm": 10.375, "learning_rate": 1.3173142467620997e-05, "loss": 2.0415910720825194, "step": 2320 }, { "epoch": 0.19855134213890072, "grad_norm": 2.703125, "learning_rate": 1.3229947739150195e-05, "loss": 2.031826400756836, "step": 2330 }, { "epoch": 0.19940349382190029, "grad_norm": 2.5, "learning_rate": 1.3286753010679392e-05, "loss": 1.9123363494873047, "step": 2340 }, { "epoch": 0.20025564550489988, "grad_norm": 4.625, "learning_rate": 1.334355828220859e-05, "loss": 1.9516468048095703, "step": 2350 }, { "epoch": 0.20110779718789945, "grad_norm": 3.5625, "learning_rate": 1.3400363553737788e-05, "loss": 1.9636569976806642, "step": 2360 }, { "epoch": 0.20195994887089902, "grad_norm": 3.359375, "learning_rate": 1.3457168825266985e-05, "loss": 1.9863510131835938, "step": 2370 }, { "epoch": 0.2028121005538986, "grad_norm": 6.40625, "learning_rate": 1.3513974096796184e-05, "loss": 2.011285591125488, "step": 2380 }, { "epoch": 0.20366425223689816, "grad_norm": 4.25, "learning_rate": 1.3570779368325382e-05, "loss": 2.0010231018066404, "step": 2390 }, { "epoch": 0.20451640391989775, "grad_norm": 4.0, "learning_rate": 1.362758463985458e-05, "loss": 1.9532783508300782, "step": 2400 }, { "epoch": 0.20536855560289732, "grad_norm": 3.75, "learning_rate": 1.3684389911383777e-05, "loss": 1.9184133529663085, "step": 2410 }, { "epoch": 0.2062207072858969, "grad_norm": 3.375, "learning_rate": 1.3741195182912975e-05, "loss": 1.9860279083251953, "step": 2420 }, { "epoch": 0.20707285896889646, "grad_norm": 3.375, "learning_rate": 1.3798000454442172e-05, "loss": 1.938701057434082, "step": 2430 }, { "epoch": 0.20792501065189603, "grad_norm": 6.125, "learning_rate": 1.385480572597137e-05, "loss": 1.8259967803955077, "step": 2440 }, { "epoch": 0.2087771623348956, "grad_norm": 8.0, "learning_rate": 1.3911610997500568e-05, "loss": 1.9804460525512695, "step": 2450 }, { "epoch": 0.2096293140178952, "grad_norm": 4.53125, "learning_rate": 1.3968416269029765e-05, "loss": 1.8732410430908204, "step": 2460 }, { "epoch": 0.21048146570089477, "grad_norm": 3.515625, "learning_rate": 1.4025221540558964e-05, "loss": 1.8008220672607422, "step": 2470 }, { "epoch": 0.21133361738389433, "grad_norm": 3.15625, "learning_rate": 1.4082026812088162e-05, "loss": 1.8646345138549805, "step": 2480 }, { "epoch": 0.2121857690668939, "grad_norm": 3.1875, "learning_rate": 1.413883208361736e-05, "loss": 1.880078125, "step": 2490 }, { "epoch": 0.21303792074989347, "grad_norm": 2.34375, "learning_rate": 1.4195637355146557e-05, "loss": 1.8868989944458008, "step": 2500 }, { "epoch": 0.21389007243289304, "grad_norm": 4.75, "learning_rate": 1.4252442626675755e-05, "loss": 1.8020980834960938, "step": 2510 }, { "epoch": 0.21474222411589264, "grad_norm": 3.140625, "learning_rate": 1.4309247898204956e-05, "loss": 1.8301939010620116, "step": 2520 }, { "epoch": 0.2155943757988922, "grad_norm": 2.765625, "learning_rate": 1.4366053169734153e-05, "loss": 1.877092170715332, "step": 2530 }, { "epoch": 0.21644652748189178, "grad_norm": 4.09375, "learning_rate": 1.4422858441263351e-05, "loss": 1.819587516784668, "step": 2540 }, { "epoch": 0.21729867916489135, "grad_norm": 4.1875, "learning_rate": 1.4479663712792549e-05, "loss": 1.8735994338989257, "step": 2550 }, { "epoch": 0.21815083084789091, "grad_norm": 5.28125, "learning_rate": 1.4536468984321746e-05, "loss": 1.8905181884765625, "step": 2560 }, { "epoch": 0.2190029825308905, "grad_norm": 2.25, "learning_rate": 1.4593274255850944e-05, "loss": 1.8659563064575195, "step": 2570 }, { "epoch": 0.21985513421389008, "grad_norm": 2.390625, "learning_rate": 1.4650079527380143e-05, "loss": 1.9524490356445312, "step": 2580 }, { "epoch": 0.22070728589688965, "grad_norm": 2.875, "learning_rate": 1.470688479890934e-05, "loss": 1.843519401550293, "step": 2590 }, { "epoch": 0.22155943757988922, "grad_norm": 4.6875, "learning_rate": 1.4763690070438538e-05, "loss": 1.812225341796875, "step": 2600 }, { "epoch": 0.2224115892628888, "grad_norm": 4.125, "learning_rate": 1.4820495341967736e-05, "loss": 1.8626331329345702, "step": 2610 }, { "epoch": 0.22326374094588836, "grad_norm": 4.8125, "learning_rate": 1.4877300613496933e-05, "loss": 1.9229364395141602, "step": 2620 }, { "epoch": 0.22411589262888795, "grad_norm": 2.515625, "learning_rate": 1.4934105885026131e-05, "loss": 1.7682260513305663, "step": 2630 }, { "epoch": 0.22496804431188752, "grad_norm": 3.046875, "learning_rate": 1.4990911156555328e-05, "loss": 1.846550750732422, "step": 2640 }, { "epoch": 0.2258201959948871, "grad_norm": 1.8125, "learning_rate": 1.5047716428084526e-05, "loss": 1.8246688842773438, "step": 2650 }, { "epoch": 0.22667234767788666, "grad_norm": 3.453125, "learning_rate": 1.5104521699613725e-05, "loss": 1.8106529235839843, "step": 2660 }, { "epoch": 0.22752449936088623, "grad_norm": 5.21875, "learning_rate": 1.5161326971142923e-05, "loss": 1.6811119079589845, "step": 2670 }, { "epoch": 0.22837665104388583, "grad_norm": 4.875, "learning_rate": 1.521813224267212e-05, "loss": 1.7050310134887696, "step": 2680 }, { "epoch": 0.2292288027268854, "grad_norm": 2.953125, "learning_rate": 1.5274937514201318e-05, "loss": 1.794808578491211, "step": 2690 }, { "epoch": 0.23008095440988496, "grad_norm": 2.734375, "learning_rate": 1.5331742785730516e-05, "loss": 1.8955673217773437, "step": 2700 }, { "epoch": 0.23093310609288453, "grad_norm": 2.15625, "learning_rate": 1.5388548057259713e-05, "loss": 1.7878650665283202, "step": 2710 }, { "epoch": 0.2317852577758841, "grad_norm": 2.90625, "learning_rate": 1.544535332878891e-05, "loss": 1.8762916564941405, "step": 2720 }, { "epoch": 0.23263740945888367, "grad_norm": 4.90625, "learning_rate": 1.550215860031811e-05, "loss": 1.7509052276611328, "step": 2730 }, { "epoch": 0.23348956114188327, "grad_norm": 3.234375, "learning_rate": 1.5558963871847306e-05, "loss": 1.7621074676513673, "step": 2740 }, { "epoch": 0.23434171282488284, "grad_norm": 2.21875, "learning_rate": 1.5615769143376507e-05, "loss": 1.9165668487548828, "step": 2750 }, { "epoch": 0.2351938645078824, "grad_norm": 2.0625, "learning_rate": 1.5672574414905705e-05, "loss": 1.8673730850219727, "step": 2760 }, { "epoch": 0.23604601619088197, "grad_norm": 2.578125, "learning_rate": 1.5729379686434902e-05, "loss": 1.7908060073852539, "step": 2770 }, { "epoch": 0.23689816787388154, "grad_norm": 2.640625, "learning_rate": 1.57861849579641e-05, "loss": 1.837907600402832, "step": 2780 }, { "epoch": 0.2377503195568811, "grad_norm": 2.203125, "learning_rate": 1.5842990229493297e-05, "loss": 1.9105459213256837, "step": 2790 }, { "epoch": 0.2386024712398807, "grad_norm": 3.484375, "learning_rate": 1.5899795501022495e-05, "loss": 1.745107078552246, "step": 2800 }, { "epoch": 0.23945462292288028, "grad_norm": 4.8125, "learning_rate": 1.5956600772551693e-05, "loss": 1.760063362121582, "step": 2810 }, { "epoch": 0.24030677460587985, "grad_norm": 4.6875, "learning_rate": 1.6013406044080894e-05, "loss": 1.8581480026245116, "step": 2820 }, { "epoch": 0.24115892628887942, "grad_norm": 2.421875, "learning_rate": 1.607021131561009e-05, "loss": 1.7769477844238282, "step": 2830 }, { "epoch": 0.24201107797187899, "grad_norm": 2.8125, "learning_rate": 1.612701658713929e-05, "loss": 1.6524904251098633, "step": 2840 }, { "epoch": 0.24286322965487858, "grad_norm": 2.171875, "learning_rate": 1.6183821858668486e-05, "loss": 1.7870126724243165, "step": 2850 }, { "epoch": 0.24371538133787815, "grad_norm": 2.5625, "learning_rate": 1.6240627130197684e-05, "loss": 1.7944723129272462, "step": 2860 }, { "epoch": 0.24456753302087772, "grad_norm": 2.65625, "learning_rate": 1.629743240172688e-05, "loss": 1.7627750396728517, "step": 2870 }, { "epoch": 0.2454196847038773, "grad_norm": 3.71875, "learning_rate": 1.635423767325608e-05, "loss": 1.8314786911010743, "step": 2880 }, { "epoch": 0.24627183638687686, "grad_norm": 2.65625, "learning_rate": 1.6411042944785277e-05, "loss": 1.7635135650634766, "step": 2890 }, { "epoch": 0.24712398806987643, "grad_norm": 2.625, "learning_rate": 1.6467848216314474e-05, "loss": 1.8142375946044922, "step": 2900 }, { "epoch": 0.24797613975287602, "grad_norm": 2.28125, "learning_rate": 1.6524653487843672e-05, "loss": 1.8576541900634767, "step": 2910 }, { "epoch": 0.2488282914358756, "grad_norm": 2.703125, "learning_rate": 1.658145875937287e-05, "loss": 1.7446006774902343, "step": 2920 }, { "epoch": 0.24968044311887516, "grad_norm": 2.734375, "learning_rate": 1.6638264030902067e-05, "loss": 1.8278125762939452, "step": 2930 }, { "epoch": 0.25053259480187473, "grad_norm": 2.8125, "learning_rate": 1.6695069302431265e-05, "loss": 1.710805320739746, "step": 2940 }, { "epoch": 0.2513847464848743, "grad_norm": 1.96875, "learning_rate": 1.6751874573960462e-05, "loss": 1.7502555847167969, "step": 2950 }, { "epoch": 0.25223689816787387, "grad_norm": 1.8671875, "learning_rate": 1.680867984548966e-05, "loss": 1.7655445098876954, "step": 2960 }, { "epoch": 0.25308904985087344, "grad_norm": 3.484375, "learning_rate": 1.6865485117018857e-05, "loss": 1.7399562835693358, "step": 2970 }, { "epoch": 0.253941201533873, "grad_norm": 2.140625, "learning_rate": 1.6922290388548058e-05, "loss": 1.700760841369629, "step": 2980 }, { "epoch": 0.2547933532168726, "grad_norm": 3.171875, "learning_rate": 1.6979095660077256e-05, "loss": 1.7815872192382813, "step": 2990 }, { "epoch": 0.2556455048998722, "grad_norm": 5.125, "learning_rate": 1.7035900931606453e-05, "loss": 1.751510238647461, "step": 3000 }, { "epoch": 0.25649765658287177, "grad_norm": 2.203125, "learning_rate": 1.709270620313565e-05, "loss": 1.7463590621948242, "step": 3010 }, { "epoch": 0.25734980826587134, "grad_norm": 3.078125, "learning_rate": 1.7149511474664852e-05, "loss": 1.8178329467773438, "step": 3020 }, { "epoch": 0.2582019599488709, "grad_norm": 2.3125, "learning_rate": 1.720631674619405e-05, "loss": 1.7225589752197266, "step": 3030 }, { "epoch": 0.2590541116318705, "grad_norm": 3.84375, "learning_rate": 1.7263122017723247e-05, "loss": 1.7269128799438476, "step": 3040 }, { "epoch": 0.25990626331487005, "grad_norm": 2.0625, "learning_rate": 1.7319927289252445e-05, "loss": 1.8231130599975587, "step": 3050 }, { "epoch": 0.2607584149978696, "grad_norm": 2.25, "learning_rate": 1.7376732560781642e-05, "loss": 1.6921531677246093, "step": 3060 }, { "epoch": 0.2616105666808692, "grad_norm": 2.34375, "learning_rate": 1.743353783231084e-05, "loss": 1.7835988998413086, "step": 3070 }, { "epoch": 0.26246271836386875, "grad_norm": 5.59375, "learning_rate": 1.7490343103840038e-05, "loss": 1.679429054260254, "step": 3080 }, { "epoch": 0.2633148700468683, "grad_norm": 2.0625, "learning_rate": 1.7547148375369235e-05, "loss": 1.856195831298828, "step": 3090 }, { "epoch": 0.2641670217298679, "grad_norm": 2.0, "learning_rate": 1.7603953646898433e-05, "loss": 1.7146261215209961, "step": 3100 }, { "epoch": 0.2650191734128675, "grad_norm": 2.296875, "learning_rate": 1.766075891842763e-05, "loss": 1.7138565063476563, "step": 3110 }, { "epoch": 0.2658713250958671, "grad_norm": 3.78125, "learning_rate": 1.7717564189956828e-05, "loss": 1.7057083129882813, "step": 3120 }, { "epoch": 0.26672347677886665, "grad_norm": 2.140625, "learning_rate": 1.7774369461486025e-05, "loss": 1.9075929641723632, "step": 3130 }, { "epoch": 0.2675756284618662, "grad_norm": 2.84375, "learning_rate": 1.7831174733015223e-05, "loss": 1.7495107650756836, "step": 3140 }, { "epoch": 0.2684277801448658, "grad_norm": 2.53125, "learning_rate": 1.788798000454442e-05, "loss": 1.7976697921752929, "step": 3150 }, { "epoch": 0.26927993182786536, "grad_norm": 3.078125, "learning_rate": 1.7944785276073618e-05, "loss": 1.8000255584716798, "step": 3160 }, { "epoch": 0.27013208351086493, "grad_norm": 2.671875, "learning_rate": 1.8001590547602816e-05, "loss": 1.7428268432617187, "step": 3170 }, { "epoch": 0.2709842351938645, "grad_norm": 2.125, "learning_rate": 1.8058395819132017e-05, "loss": 1.7144651412963867, "step": 3180 }, { "epoch": 0.27183638687686407, "grad_norm": 4.4375, "learning_rate": 1.8115201090661214e-05, "loss": 1.7320880889892578, "step": 3190 }, { "epoch": 0.27268853855986364, "grad_norm": 2.28125, "learning_rate": 1.8172006362190412e-05, "loss": 1.862847900390625, "step": 3200 }, { "epoch": 0.2735406902428632, "grad_norm": 1.7578125, "learning_rate": 1.8228811633719613e-05, "loss": 1.8186471939086915, "step": 3210 }, { "epoch": 0.27439284192586283, "grad_norm": 4.875, "learning_rate": 1.828561690524881e-05, "loss": 1.8253974914550781, "step": 3220 }, { "epoch": 0.2752449936088624, "grad_norm": 2.046875, "learning_rate": 1.8342422176778008e-05, "loss": 1.7939125061035157, "step": 3230 }, { "epoch": 0.27609714529186197, "grad_norm": 3.140625, "learning_rate": 1.8399227448307206e-05, "loss": 1.787186050415039, "step": 3240 }, { "epoch": 0.27694929697486154, "grad_norm": 3.515625, "learning_rate": 1.8456032719836403e-05, "loss": 1.7974475860595702, "step": 3250 }, { "epoch": 0.2778014486578611, "grad_norm": 1.7265625, "learning_rate": 1.85128379913656e-05, "loss": 1.6549633026123047, "step": 3260 }, { "epoch": 0.2786536003408607, "grad_norm": 4.03125, "learning_rate": 1.85696432628948e-05, "loss": 1.7429319381713868, "step": 3270 }, { "epoch": 0.27950575202386024, "grad_norm": 1.9609375, "learning_rate": 1.8626448534423996e-05, "loss": 1.700511360168457, "step": 3280 }, { "epoch": 0.2803579037068598, "grad_norm": 2.1875, "learning_rate": 1.8683253805953194e-05, "loss": 1.6811868667602539, "step": 3290 }, { "epoch": 0.2812100553898594, "grad_norm": 1.828125, "learning_rate": 1.874005907748239e-05, "loss": 1.8987232208251954, "step": 3300 }, { "epoch": 0.28206220707285895, "grad_norm": 2.71875, "learning_rate": 1.879686434901159e-05, "loss": 1.7597833633422852, "step": 3310 }, { "epoch": 0.2829143587558585, "grad_norm": 3.6875, "learning_rate": 1.8853669620540786e-05, "loss": 1.5584803581237794, "step": 3320 }, { "epoch": 0.28376651043885814, "grad_norm": 1.5, "learning_rate": 1.8910474892069984e-05, "loss": 1.7714111328125, "step": 3330 }, { "epoch": 0.2846186621218577, "grad_norm": 2.640625, "learning_rate": 1.896728016359918e-05, "loss": 1.732600784301758, "step": 3340 }, { "epoch": 0.2854708138048573, "grad_norm": 3.09375, "learning_rate": 1.902408543512838e-05, "loss": 1.6895940780639649, "step": 3350 }, { "epoch": 0.28632296548785685, "grad_norm": 2.515625, "learning_rate": 1.9080890706657577e-05, "loss": 1.7193382263183594, "step": 3360 }, { "epoch": 0.2871751171708564, "grad_norm": 3.21875, "learning_rate": 1.9137695978186774e-05, "loss": 1.674692153930664, "step": 3370 }, { "epoch": 0.288027268853856, "grad_norm": 3.21875, "learning_rate": 1.9194501249715975e-05, "loss": 1.7754575729370117, "step": 3380 }, { "epoch": 0.28887942053685556, "grad_norm": 3.15625, "learning_rate": 1.9251306521245173e-05, "loss": 1.7507244110107423, "step": 3390 }, { "epoch": 0.2897315722198551, "grad_norm": 3.4375, "learning_rate": 1.930811179277437e-05, "loss": 1.643104362487793, "step": 3400 }, { "epoch": 0.2905837239028547, "grad_norm": 1.953125, "learning_rate": 1.9364917064303568e-05, "loss": 1.6756439208984375, "step": 3410 }, { "epoch": 0.29143587558585426, "grad_norm": 4.1875, "learning_rate": 1.9421722335832766e-05, "loss": 1.7002479553222656, "step": 3420 }, { "epoch": 0.29228802726885383, "grad_norm": 2.34375, "learning_rate": 1.9478527607361967e-05, "loss": 1.5581110954284667, "step": 3430 }, { "epoch": 0.2931401789518534, "grad_norm": 3.75, "learning_rate": 1.9535332878891164e-05, "loss": 1.6264345169067382, "step": 3440 }, { "epoch": 0.293992330634853, "grad_norm": 2.9375, "learning_rate": 1.9592138150420362e-05, "loss": 1.6526121139526366, "step": 3450 }, { "epoch": 0.2948444823178526, "grad_norm": 2.9375, "learning_rate": 1.964894342194956e-05, "loss": 1.7403932571411134, "step": 3460 }, { "epoch": 0.29569663400085217, "grad_norm": 3.390625, "learning_rate": 1.9705748693478757e-05, "loss": 1.7250543594360352, "step": 3470 }, { "epoch": 0.29654878568385173, "grad_norm": 2.03125, "learning_rate": 1.9762553965007955e-05, "loss": 1.7352684020996094, "step": 3480 }, { "epoch": 0.2974009373668513, "grad_norm": 1.9609375, "learning_rate": 1.9819359236537152e-05, "loss": 1.6909282684326172, "step": 3490 }, { "epoch": 0.2982530890498509, "grad_norm": 2.5, "learning_rate": 1.987616450806635e-05, "loss": 1.8747112274169921, "step": 3500 }, { "epoch": 0.29910524073285044, "grad_norm": 1.96875, "learning_rate": 1.9932969779595547e-05, "loss": 1.6911521911621095, "step": 3510 }, { "epoch": 0.29995739241585, "grad_norm": 4.46875, "learning_rate": 1.9989775051124745e-05, "loss": 1.725855255126953, "step": 3520 }, { "epoch": 0.3008095440988496, "grad_norm": 6.15625, "learning_rate": 2.0046580322653943e-05, "loss": 1.5265551567077638, "step": 3530 }, { "epoch": 0.30166169578184915, "grad_norm": 2.5625, "learning_rate": 2.010338559418314e-05, "loss": 1.6919170379638673, "step": 3540 }, { "epoch": 0.3025138474648487, "grad_norm": 3.375, "learning_rate": 2.0160190865712338e-05, "loss": 1.7359495162963867, "step": 3550 }, { "epoch": 0.30336599914784834, "grad_norm": 3.90625, "learning_rate": 2.0216996137241535e-05, "loss": 1.6759765625, "step": 3560 }, { "epoch": 0.3042181508308479, "grad_norm": 3.015625, "learning_rate": 2.0273801408770733e-05, "loss": 1.8724205017089843, "step": 3570 }, { "epoch": 0.3050703025138475, "grad_norm": 3.59375, "learning_rate": 2.0330606680299934e-05, "loss": 1.7565452575683593, "step": 3580 }, { "epoch": 0.30592245419684705, "grad_norm": 2.65625, "learning_rate": 2.038741195182913e-05, "loss": 1.711573028564453, "step": 3590 }, { "epoch": 0.3067746058798466, "grad_norm": 2.9375, "learning_rate": 2.044421722335833e-05, "loss": 1.8155204772949218, "step": 3600 }, { "epoch": 0.3076267575628462, "grad_norm": 2.0, "learning_rate": 2.0501022494887527e-05, "loss": 1.7908649444580078, "step": 3610 }, { "epoch": 0.30847890924584576, "grad_norm": 2.609375, "learning_rate": 2.0557827766416724e-05, "loss": 1.736526107788086, "step": 3620 }, { "epoch": 0.3093310609288453, "grad_norm": 2.0625, "learning_rate": 2.0614633037945922e-05, "loss": 1.8440467834472656, "step": 3630 }, { "epoch": 0.3101832126118449, "grad_norm": 2.34375, "learning_rate": 2.067143830947512e-05, "loss": 1.7345935821533203, "step": 3640 }, { "epoch": 0.31103536429484446, "grad_norm": 2.65625, "learning_rate": 2.0728243581004317e-05, "loss": 1.666162872314453, "step": 3650 }, { "epoch": 0.31188751597784403, "grad_norm": 2.65625, "learning_rate": 2.0785048852533518e-05, "loss": 1.6066532135009766, "step": 3660 }, { "epoch": 0.31273966766084366, "grad_norm": 3.375, "learning_rate": 2.0841854124062715e-05, "loss": 1.7179904937744142, "step": 3670 }, { "epoch": 0.3135918193438432, "grad_norm": 3.6875, "learning_rate": 2.0898659395591913e-05, "loss": 1.6568355560302734, "step": 3680 }, { "epoch": 0.3144439710268428, "grad_norm": 5.09375, "learning_rate": 2.095546466712111e-05, "loss": 1.6793256759643556, "step": 3690 }, { "epoch": 0.31529612270984236, "grad_norm": 3.0, "learning_rate": 2.1012269938650308e-05, "loss": 1.623748207092285, "step": 3700 }, { "epoch": 0.31614827439284193, "grad_norm": 3.5, "learning_rate": 2.1069075210179506e-05, "loss": 1.851422119140625, "step": 3710 }, { "epoch": 0.3170004260758415, "grad_norm": 2.4375, "learning_rate": 2.1125880481708703e-05, "loss": 1.7401918411254882, "step": 3720 }, { "epoch": 0.31785257775884107, "grad_norm": 2.28125, "learning_rate": 2.11826857532379e-05, "loss": 1.719321060180664, "step": 3730 }, { "epoch": 0.31870472944184064, "grad_norm": 2.71875, "learning_rate": 2.12394910247671e-05, "loss": 1.6823297500610352, "step": 3740 }, { "epoch": 0.3195568811248402, "grad_norm": 3.1875, "learning_rate": 2.1296296296296296e-05, "loss": 1.5594244956970216, "step": 3750 }, { "epoch": 0.3204090328078398, "grad_norm": 2.8125, "learning_rate": 2.1353101567825494e-05, "loss": 1.9076438903808595, "step": 3760 }, { "epoch": 0.32126118449083935, "grad_norm": 2.8125, "learning_rate": 2.1409906839354695e-05, "loss": 1.7063247680664062, "step": 3770 }, { "epoch": 0.3221133361738389, "grad_norm": 4.0, "learning_rate": 2.1466712110883892e-05, "loss": 1.750584602355957, "step": 3780 }, { "epoch": 0.32296548785683854, "grad_norm": 2.515625, "learning_rate": 2.152351738241309e-05, "loss": 1.689323616027832, "step": 3790 }, { "epoch": 0.3238176395398381, "grad_norm": 2.703125, "learning_rate": 2.1580322653942288e-05, "loss": 1.7448822021484376, "step": 3800 }, { "epoch": 0.3246697912228377, "grad_norm": 3.65625, "learning_rate": 2.1637127925471485e-05, "loss": 1.6827381134033204, "step": 3810 }, { "epoch": 0.32552194290583725, "grad_norm": 2.90625, "learning_rate": 2.1693933197000683e-05, "loss": 1.6794099807739258, "step": 3820 }, { "epoch": 0.3263740945888368, "grad_norm": 3.34375, "learning_rate": 2.175073846852988e-05, "loss": 1.6809560775756835, "step": 3830 }, { "epoch": 0.3272262462718364, "grad_norm": 1.75, "learning_rate": 2.1807543740059078e-05, "loss": 1.7273214340209961, "step": 3840 }, { "epoch": 0.32807839795483595, "grad_norm": 2.703125, "learning_rate": 2.1864349011588275e-05, "loss": 1.5267117500305176, "step": 3850 }, { "epoch": 0.3289305496378355, "grad_norm": 4.3125, "learning_rate": 2.1921154283117473e-05, "loss": 1.7080059051513672, "step": 3860 }, { "epoch": 0.3297827013208351, "grad_norm": 2.671875, "learning_rate": 2.197795955464667e-05, "loss": 1.7015560150146485, "step": 3870 }, { "epoch": 0.33063485300383466, "grad_norm": 1.7890625, "learning_rate": 2.2034764826175868e-05, "loss": 1.6471054077148437, "step": 3880 }, { "epoch": 0.33148700468683423, "grad_norm": 3.203125, "learning_rate": 2.209157009770507e-05, "loss": 1.6858919143676758, "step": 3890 }, { "epoch": 0.33233915636983385, "grad_norm": 2.828125, "learning_rate": 2.2148375369234267e-05, "loss": 1.5229301452636719, "step": 3900 }, { "epoch": 0.3331913080528334, "grad_norm": 2.9375, "learning_rate": 2.2205180640763464e-05, "loss": 1.7655166625976562, "step": 3910 }, { "epoch": 0.334043459735833, "grad_norm": 3.046875, "learning_rate": 2.2261985912292662e-05, "loss": 1.732417678833008, "step": 3920 }, { "epoch": 0.33489561141883256, "grad_norm": 4.40625, "learning_rate": 2.231879118382186e-05, "loss": 1.600183868408203, "step": 3930 }, { "epoch": 0.33574776310183213, "grad_norm": 1.9453125, "learning_rate": 2.2375596455351057e-05, "loss": 1.6474760055541993, "step": 3940 }, { "epoch": 0.3365999147848317, "grad_norm": 1.9453125, "learning_rate": 2.2432401726880255e-05, "loss": 1.6033010482788086, "step": 3950 }, { "epoch": 0.33745206646783127, "grad_norm": 2.4375, "learning_rate": 2.2489206998409452e-05, "loss": 1.7297969818115235, "step": 3960 }, { "epoch": 0.33830421815083084, "grad_norm": 3.71875, "learning_rate": 2.2546012269938653e-05, "loss": 1.8503284454345703, "step": 3970 }, { "epoch": 0.3391563698338304, "grad_norm": 2.359375, "learning_rate": 2.260281754146785e-05, "loss": 1.7416799545288086, "step": 3980 }, { "epoch": 0.34000852151683, "grad_norm": 1.8046875, "learning_rate": 2.265962281299705e-05, "loss": 1.6267154693603516, "step": 3990 }, { "epoch": 0.34086067319982954, "grad_norm": 2.515625, "learning_rate": 2.2716428084526246e-05, "loss": 1.6257692337036134, "step": 4000 }, { "epoch": 0.34171282488282917, "grad_norm": 3.015625, "learning_rate": 2.2773233356055444e-05, "loss": 1.588251781463623, "step": 4010 }, { "epoch": 0.34256497656582874, "grad_norm": 3.3125, "learning_rate": 2.283003862758464e-05, "loss": 1.6862438201904297, "step": 4020 }, { "epoch": 0.3434171282488283, "grad_norm": 1.96875, "learning_rate": 2.288684389911384e-05, "loss": 1.6842063903808593, "step": 4030 }, { "epoch": 0.3442692799318279, "grad_norm": 3.125, "learning_rate": 2.2943649170643036e-05, "loss": 1.681544876098633, "step": 4040 }, { "epoch": 0.34512143161482745, "grad_norm": 3.0625, "learning_rate": 2.3000454442172234e-05, "loss": 1.5426207542419434, "step": 4050 }, { "epoch": 0.345973583297827, "grad_norm": 3.140625, "learning_rate": 2.305725971370143e-05, "loss": 1.7934782028198242, "step": 4060 }, { "epoch": 0.3468257349808266, "grad_norm": 1.9375, "learning_rate": 2.311406498523063e-05, "loss": 1.7919416427612305, "step": 4070 }, { "epoch": 0.34767788666382615, "grad_norm": 2.453125, "learning_rate": 2.3170870256759827e-05, "loss": 1.7091051101684571, "step": 4080 }, { "epoch": 0.3485300383468257, "grad_norm": 2.84375, "learning_rate": 2.3227675528289024e-05, "loss": 1.6481298446655273, "step": 4090 }, { "epoch": 0.3493821900298253, "grad_norm": 4.03125, "learning_rate": 2.3284480799818222e-05, "loss": 1.74404296875, "step": 4100 }, { "epoch": 0.35023434171282486, "grad_norm": 2.359375, "learning_rate": 2.3341286071347423e-05, "loss": 1.8829521179199218, "step": 4110 }, { "epoch": 0.3510864933958245, "grad_norm": 2.984375, "learning_rate": 2.339809134287662e-05, "loss": 1.5831212997436523, "step": 4120 }, { "epoch": 0.35193864507882405, "grad_norm": 2.03125, "learning_rate": 2.3454896614405818e-05, "loss": 1.642235565185547, "step": 4130 }, { "epoch": 0.3527907967618236, "grad_norm": 2.734375, "learning_rate": 2.3511701885935016e-05, "loss": 1.70123348236084, "step": 4140 }, { "epoch": 0.3536429484448232, "grad_norm": 2.59375, "learning_rate": 2.3568507157464213e-05, "loss": 1.6906259536743165, "step": 4150 }, { "epoch": 0.35449510012782276, "grad_norm": 2.625, "learning_rate": 2.362531242899341e-05, "loss": 1.7254396438598634, "step": 4160 }, { "epoch": 0.35534725181082233, "grad_norm": 2.265625, "learning_rate": 2.3682117700522612e-05, "loss": 1.6922157287597657, "step": 4170 }, { "epoch": 0.3561994034938219, "grad_norm": 3.15625, "learning_rate": 2.373892297205181e-05, "loss": 1.7291748046875, "step": 4180 }, { "epoch": 0.35705155517682147, "grad_norm": 1.9765625, "learning_rate": 2.3795728243581007e-05, "loss": 1.6746204376220704, "step": 4190 }, { "epoch": 0.35790370685982104, "grad_norm": 3.09375, "learning_rate": 2.3852533515110205e-05, "loss": 1.709360694885254, "step": 4200 }, { "epoch": 0.3587558585428206, "grad_norm": 3.296875, "learning_rate": 2.3909338786639402e-05, "loss": 1.748908805847168, "step": 4210 }, { "epoch": 0.3596080102258202, "grad_norm": 3.671875, "learning_rate": 2.39661440581686e-05, "loss": 1.7059072494506835, "step": 4220 }, { "epoch": 0.36046016190881974, "grad_norm": 1.6875, "learning_rate": 2.4022949329697797e-05, "loss": 1.799212646484375, "step": 4230 }, { "epoch": 0.36131231359181937, "grad_norm": 4.75, "learning_rate": 2.4079754601226995e-05, "loss": 1.528904342651367, "step": 4240 }, { "epoch": 0.36216446527481894, "grad_norm": 2.640625, "learning_rate": 2.4136559872756192e-05, "loss": 1.7835922241210938, "step": 4250 }, { "epoch": 0.3630166169578185, "grad_norm": 3.421875, "learning_rate": 2.419336514428539e-05, "loss": 1.6761665344238281, "step": 4260 }, { "epoch": 0.3638687686408181, "grad_norm": 3.40625, "learning_rate": 2.4250170415814588e-05, "loss": 1.8425657272338867, "step": 4270 }, { "epoch": 0.36472092032381764, "grad_norm": 2.984375, "learning_rate": 2.4306975687343785e-05, "loss": 1.702306365966797, "step": 4280 }, { "epoch": 0.3655730720068172, "grad_norm": 2.109375, "learning_rate": 2.4363780958872983e-05, "loss": 1.5865326881408692, "step": 4290 }, { "epoch": 0.3664252236898168, "grad_norm": 4.125, "learning_rate": 2.442058623040218e-05, "loss": 1.6182355880737305, "step": 4300 }, { "epoch": 0.36727737537281635, "grad_norm": 4.0625, "learning_rate": 2.4477391501931378e-05, "loss": 1.658778190612793, "step": 4310 }, { "epoch": 0.3681295270558159, "grad_norm": 4.1875, "learning_rate": 2.4534196773460576e-05, "loss": 1.6500150680541992, "step": 4320 }, { "epoch": 0.3689816787388155, "grad_norm": 1.984375, "learning_rate": 2.4591002044989777e-05, "loss": 1.850885009765625, "step": 4330 }, { "epoch": 0.36983383042181506, "grad_norm": 2.515625, "learning_rate": 2.4647807316518974e-05, "loss": 1.673583984375, "step": 4340 }, { "epoch": 0.3706859821048147, "grad_norm": 2.375, "learning_rate": 2.4704612588048172e-05, "loss": 1.7205814361572265, "step": 4350 }, { "epoch": 0.37153813378781425, "grad_norm": 2.65625, "learning_rate": 2.4761417859577373e-05, "loss": 1.5333928108215331, "step": 4360 }, { "epoch": 0.3723902854708138, "grad_norm": 3.375, "learning_rate": 2.481822313110657e-05, "loss": 1.603163719177246, "step": 4370 }, { "epoch": 0.3732424371538134, "grad_norm": 2.5625, "learning_rate": 2.4875028402635768e-05, "loss": 1.8358612060546875, "step": 4380 }, { "epoch": 0.37409458883681296, "grad_norm": 2.046875, "learning_rate": 2.4931833674164965e-05, "loss": 1.6357099533081054, "step": 4390 }, { "epoch": 0.3749467405198125, "grad_norm": 1.765625, "learning_rate": 2.4988638945694163e-05, "loss": 1.7182041168212892, "step": 4400 }, { "epoch": 0.3757988922028121, "grad_norm": 3.703125, "learning_rate": 2.5045444217223357e-05, "loss": 1.6822734832763673, "step": 4410 }, { "epoch": 0.37665104388581166, "grad_norm": 2.890625, "learning_rate": 2.5102249488752555e-05, "loss": 1.7366880416870116, "step": 4420 }, { "epoch": 0.37750319556881123, "grad_norm": 2.234375, "learning_rate": 2.5159054760281752e-05, "loss": 1.626662254333496, "step": 4430 }, { "epoch": 0.3783553472518108, "grad_norm": 2.8125, "learning_rate": 2.521586003181095e-05, "loss": 1.6350317001342773, "step": 4440 }, { "epoch": 0.37920749893481037, "grad_norm": 3.046875, "learning_rate": 2.5272665303340154e-05, "loss": 1.7188117980957032, "step": 4450 }, { "epoch": 0.38005965061781, "grad_norm": 2.578125, "learning_rate": 2.5329470574869352e-05, "loss": 1.7745798110961915, "step": 4460 }, { "epoch": 0.38091180230080957, "grad_norm": 4.03125, "learning_rate": 2.538627584639855e-05, "loss": 1.683866310119629, "step": 4470 }, { "epoch": 0.38176395398380913, "grad_norm": 2.875, "learning_rate": 2.5443081117927747e-05, "loss": 1.8016012191772461, "step": 4480 }, { "epoch": 0.3826161056668087, "grad_norm": 2.53125, "learning_rate": 2.5499886389456945e-05, "loss": 1.6809507369995118, "step": 4490 }, { "epoch": 0.3834682573498083, "grad_norm": 2.1875, "learning_rate": 2.5556691660986142e-05, "loss": 1.7405269622802735, "step": 4500 }, { "epoch": 0.38432040903280784, "grad_norm": 3.03125, "learning_rate": 2.561349693251534e-05, "loss": 1.6066829681396484, "step": 4510 }, { "epoch": 0.3851725607158074, "grad_norm": 2.078125, "learning_rate": 2.5670302204044537e-05, "loss": 1.689692497253418, "step": 4520 }, { "epoch": 0.386024712398807, "grad_norm": 2.046875, "learning_rate": 2.5727107475573735e-05, "loss": 1.668675422668457, "step": 4530 }, { "epoch": 0.38687686408180655, "grad_norm": 2.390625, "learning_rate": 2.5783912747102933e-05, "loss": 1.7272989273071289, "step": 4540 }, { "epoch": 0.3877290157648061, "grad_norm": 1.8984375, "learning_rate": 2.584071801863213e-05, "loss": 1.7486305236816406, "step": 4550 }, { "epoch": 0.3885811674478057, "grad_norm": 2.125, "learning_rate": 2.589752329016133e-05, "loss": 1.6991172790527345, "step": 4560 }, { "epoch": 0.3894333191308053, "grad_norm": 2.328125, "learning_rate": 2.595432856169053e-05, "loss": 1.6541732788085937, "step": 4570 }, { "epoch": 0.3902854708138049, "grad_norm": 3.78125, "learning_rate": 2.6011133833219726e-05, "loss": 1.514105987548828, "step": 4580 }, { "epoch": 0.39113762249680445, "grad_norm": 3.25, "learning_rate": 2.6067939104748924e-05, "loss": 1.6796003341674806, "step": 4590 }, { "epoch": 0.391989774179804, "grad_norm": 2.375, "learning_rate": 2.612474437627812e-05, "loss": 1.7271780014038085, "step": 4600 }, { "epoch": 0.3928419258628036, "grad_norm": 2.78125, "learning_rate": 2.618154964780732e-05, "loss": 1.6060295104980469, "step": 4610 }, { "epoch": 0.39369407754580316, "grad_norm": 2.28125, "learning_rate": 2.6238354919336517e-05, "loss": 1.6012540817260743, "step": 4620 }, { "epoch": 0.3945462292288027, "grad_norm": 2.640625, "learning_rate": 2.6295160190865714e-05, "loss": 1.7102380752563477, "step": 4630 }, { "epoch": 0.3953983809118023, "grad_norm": 3.859375, "learning_rate": 2.6351965462394912e-05, "loss": 1.7161407470703125, "step": 4640 }, { "epoch": 0.39625053259480186, "grad_norm": 2.328125, "learning_rate": 2.640877073392411e-05, "loss": 1.690864372253418, "step": 4650 }, { "epoch": 0.39710268427780143, "grad_norm": 2.203125, "learning_rate": 2.6465576005453307e-05, "loss": 1.558272361755371, "step": 4660 }, { "epoch": 0.397954835960801, "grad_norm": 3.34375, "learning_rate": 2.6522381276982505e-05, "loss": 1.7281267166137695, "step": 4670 }, { "epoch": 0.39880698764380057, "grad_norm": 2.21875, "learning_rate": 2.6579186548511702e-05, "loss": 1.6394485473632812, "step": 4680 }, { "epoch": 0.3996591393268002, "grad_norm": 2.453125, "learning_rate": 2.66359918200409e-05, "loss": 1.6260711669921875, "step": 4690 }, { "epoch": 0.40051129100979976, "grad_norm": 2.921875, "learning_rate": 2.6692797091570097e-05, "loss": 1.6610958099365234, "step": 4700 }, { "epoch": 0.40136344269279933, "grad_norm": 2.875, "learning_rate": 2.6749602363099295e-05, "loss": 1.7353094100952149, "step": 4710 }, { "epoch": 0.4022155943757989, "grad_norm": 2.28125, "learning_rate": 2.6806407634628493e-05, "loss": 1.654058074951172, "step": 4720 }, { "epoch": 0.40306774605879847, "grad_norm": 3.890625, "learning_rate": 2.6863212906157694e-05, "loss": 1.6033437728881836, "step": 4730 }, { "epoch": 0.40391989774179804, "grad_norm": 3.203125, "learning_rate": 2.692001817768689e-05, "loss": 1.691653060913086, "step": 4740 }, { "epoch": 0.4047720494247976, "grad_norm": 3.21875, "learning_rate": 2.697682344921609e-05, "loss": 1.6990133285522462, "step": 4750 }, { "epoch": 0.4056242011077972, "grad_norm": 2.171875, "learning_rate": 2.7033628720745286e-05, "loss": 1.6851200103759765, "step": 4760 }, { "epoch": 0.40647635279079675, "grad_norm": 2.8125, "learning_rate": 2.7090433992274484e-05, "loss": 1.6349258422851562, "step": 4770 }, { "epoch": 0.4073285044737963, "grad_norm": 2.609375, "learning_rate": 2.714723926380368e-05, "loss": 1.5882797241210938, "step": 4780 }, { "epoch": 0.4081806561567959, "grad_norm": 2.265625, "learning_rate": 2.720404453533288e-05, "loss": 1.6685401916503906, "step": 4790 }, { "epoch": 0.4090328078397955, "grad_norm": 2.609375, "learning_rate": 2.7260849806862077e-05, "loss": 1.7452484130859376, "step": 4800 }, { "epoch": 0.4098849595227951, "grad_norm": 2.640625, "learning_rate": 2.7317655078391274e-05, "loss": 1.6389333724975585, "step": 4810 }, { "epoch": 0.41073711120579465, "grad_norm": 2.28125, "learning_rate": 2.7374460349920472e-05, "loss": 1.6615081787109376, "step": 4820 }, { "epoch": 0.4115892628887942, "grad_norm": 2.125, "learning_rate": 2.743126562144967e-05, "loss": 1.6083765029907227, "step": 4830 }, { "epoch": 0.4124414145717938, "grad_norm": 2.4375, "learning_rate": 2.7488070892978867e-05, "loss": 1.663332748413086, "step": 4840 }, { "epoch": 0.41329356625479335, "grad_norm": 2.5, "learning_rate": 2.7544876164508065e-05, "loss": 1.7092395782470704, "step": 4850 }, { "epoch": 0.4141457179377929, "grad_norm": 2.828125, "learning_rate": 2.7601681436037262e-05, "loss": 1.7002864837646485, "step": 4860 }, { "epoch": 0.4149978696207925, "grad_norm": 2.640625, "learning_rate": 2.765848670756646e-05, "loss": 1.701700782775879, "step": 4870 }, { "epoch": 0.41585002130379206, "grad_norm": 3.65625, "learning_rate": 2.7715291979095657e-05, "loss": 1.5509613037109375, "step": 4880 }, { "epoch": 0.41670217298679163, "grad_norm": 2.703125, "learning_rate": 2.777209725062486e-05, "loss": 1.5213688850402831, "step": 4890 }, { "epoch": 0.4175543246697912, "grad_norm": 1.9609375, "learning_rate": 2.7828902522154056e-05, "loss": 1.4812524795532227, "step": 4900 }, { "epoch": 0.4184064763527908, "grad_norm": 2.625, "learning_rate": 2.7885707793683257e-05, "loss": 1.6381212234497071, "step": 4910 }, { "epoch": 0.4192586280357904, "grad_norm": 2.984375, "learning_rate": 2.7942513065212454e-05, "loss": 1.7476909637451172, "step": 4920 }, { "epoch": 0.42011077971878996, "grad_norm": 2.734375, "learning_rate": 2.7999318336741652e-05, "loss": 1.721877670288086, "step": 4930 }, { "epoch": 0.42096293140178953, "grad_norm": 2.453125, "learning_rate": 2.805612360827085e-05, "loss": 1.6362375259399413, "step": 4940 }, { "epoch": 0.4218150830847891, "grad_norm": 1.9453125, "learning_rate": 2.811292887980005e-05, "loss": 1.6712093353271484, "step": 4950 }, { "epoch": 0.42266723476778867, "grad_norm": 3.34375, "learning_rate": 2.8169734151329248e-05, "loss": 1.6266807556152343, "step": 4960 }, { "epoch": 0.42351938645078824, "grad_norm": 2.046875, "learning_rate": 2.8226539422858446e-05, "loss": 1.64483699798584, "step": 4970 }, { "epoch": 0.4243715381337878, "grad_norm": 1.9296875, "learning_rate": 2.8283344694387643e-05, "loss": 1.65655517578125, "step": 4980 }, { "epoch": 0.4252236898167874, "grad_norm": 2.28125, "learning_rate": 2.834014996591684e-05, "loss": 1.6933605194091796, "step": 4990 }, { "epoch": 0.42607584149978694, "grad_norm": 2.03125, "learning_rate": 2.839695523744604e-05, "loss": 1.6366462707519531, "step": 5000 }, { "epoch": 0.4269279931827865, "grad_norm": 2.296875, "learning_rate": 2.8453760508975236e-05, "loss": 1.5759639739990234, "step": 5010 }, { "epoch": 0.4277801448657861, "grad_norm": 1.75, "learning_rate": 2.8510565780504434e-05, "loss": 1.6916797637939454, "step": 5020 }, { "epoch": 0.4286322965487857, "grad_norm": 5.34375, "learning_rate": 2.856737105203363e-05, "loss": 1.6770757675170898, "step": 5030 }, { "epoch": 0.4294844482317853, "grad_norm": 2.625, "learning_rate": 2.862417632356283e-05, "loss": 1.6593156814575196, "step": 5040 }, { "epoch": 0.43033659991478485, "grad_norm": 2.078125, "learning_rate": 2.8680981595092026e-05, "loss": 1.645838165283203, "step": 5050 }, { "epoch": 0.4311887515977844, "grad_norm": 4.6875, "learning_rate": 2.8737786866621224e-05, "loss": 1.714085578918457, "step": 5060 }, { "epoch": 0.432040903280784, "grad_norm": 2.90625, "learning_rate": 2.879459213815042e-05, "loss": 1.723839569091797, "step": 5070 }, { "epoch": 0.43289305496378355, "grad_norm": 3.1875, "learning_rate": 2.885139740967962e-05, "loss": 1.6333539962768555, "step": 5080 }, { "epoch": 0.4337452066467831, "grad_norm": 3.78125, "learning_rate": 2.8908202681208817e-05, "loss": 1.6218761444091796, "step": 5090 }, { "epoch": 0.4345973583297827, "grad_norm": 1.6640625, "learning_rate": 2.8965007952738014e-05, "loss": 1.7181917190551759, "step": 5100 }, { "epoch": 0.43544951001278226, "grad_norm": 2.484375, "learning_rate": 2.9021813224267212e-05, "loss": 1.7208145141601563, "step": 5110 }, { "epoch": 0.43630166169578183, "grad_norm": 3.828125, "learning_rate": 2.9078618495796413e-05, "loss": 1.6569648742675782, "step": 5120 }, { "epoch": 0.4371538133787814, "grad_norm": 2.234375, "learning_rate": 2.913542376732561e-05, "loss": 1.5916035652160645, "step": 5130 }, { "epoch": 0.438005965061781, "grad_norm": 2.75, "learning_rate": 2.9192229038854808e-05, "loss": 1.576393222808838, "step": 5140 }, { "epoch": 0.4388581167447806, "grad_norm": 2.921875, "learning_rate": 2.9249034310384006e-05, "loss": 1.590585994720459, "step": 5150 }, { "epoch": 0.43971026842778016, "grad_norm": 3.109375, "learning_rate": 2.9305839581913203e-05, "loss": 1.6003120422363282, "step": 5160 }, { "epoch": 0.44056242011077973, "grad_norm": 2.484375, "learning_rate": 2.93626448534424e-05, "loss": 1.6457889556884766, "step": 5170 }, { "epoch": 0.4414145717937793, "grad_norm": 2.9375, "learning_rate": 2.94194501249716e-05, "loss": 1.5815167427062988, "step": 5180 }, { "epoch": 0.44226672347677887, "grad_norm": 2.578125, "learning_rate": 2.9476255396500796e-05, "loss": 1.660608673095703, "step": 5190 }, { "epoch": 0.44311887515977844, "grad_norm": 3.515625, "learning_rate": 2.9533060668029994e-05, "loss": 1.5160191535949707, "step": 5200 }, { "epoch": 0.443971026842778, "grad_norm": 2.453125, "learning_rate": 2.958986593955919e-05, "loss": 1.5215105056762694, "step": 5210 }, { "epoch": 0.4448231785257776, "grad_norm": 2.015625, "learning_rate": 2.964667121108839e-05, "loss": 1.7308181762695312, "step": 5220 }, { "epoch": 0.44567533020877714, "grad_norm": 2.53125, "learning_rate": 2.9703476482617586e-05, "loss": 1.6698080062866212, "step": 5230 }, { "epoch": 0.4465274818917767, "grad_norm": 4.46875, "learning_rate": 2.9760281754146784e-05, "loss": 1.490866470336914, "step": 5240 }, { "epoch": 0.44737963357477634, "grad_norm": 2.515625, "learning_rate": 2.981708702567598e-05, "loss": 1.7061513900756835, "step": 5250 }, { "epoch": 0.4482317852577759, "grad_norm": 2.296875, "learning_rate": 2.987389229720518e-05, "loss": 1.5574399948120117, "step": 5260 }, { "epoch": 0.4490839369407755, "grad_norm": 2.984375, "learning_rate": 2.9930697568734377e-05, "loss": 1.461842918395996, "step": 5270 }, { "epoch": 0.44993608862377504, "grad_norm": 3.34375, "learning_rate": 2.9987502840263574e-05, "loss": 1.5267431259155273, "step": 5280 }, { "epoch": 0.4507882403067746, "grad_norm": 3.9375, "learning_rate": 3.0044308111792775e-05, "loss": 1.6017213821411134, "step": 5290 }, { "epoch": 0.4516403919897742, "grad_norm": 4.4375, "learning_rate": 3.0101113383321973e-05, "loss": 1.5493282318115233, "step": 5300 }, { "epoch": 0.45249254367277375, "grad_norm": 1.9453125, "learning_rate": 3.015791865485117e-05, "loss": 1.5583986282348632, "step": 5310 }, { "epoch": 0.4533446953557733, "grad_norm": 2.6875, "learning_rate": 3.0214723926380368e-05, "loss": 1.546107578277588, "step": 5320 }, { "epoch": 0.4541968470387729, "grad_norm": 2.796875, "learning_rate": 3.0271529197909566e-05, "loss": 1.6106321334838867, "step": 5330 }, { "epoch": 0.45504899872177246, "grad_norm": 2.515625, "learning_rate": 3.0328334469438763e-05, "loss": 1.6307744979858398, "step": 5340 }, { "epoch": 0.455901150404772, "grad_norm": 1.75, "learning_rate": 3.038513974096796e-05, "loss": 1.5628026962280273, "step": 5350 }, { "epoch": 0.45675330208777165, "grad_norm": 2.953125, "learning_rate": 3.0441945012497165e-05, "loss": 1.6138202667236328, "step": 5360 }, { "epoch": 0.4576054537707712, "grad_norm": 4.09375, "learning_rate": 3.0498750284026363e-05, "loss": 1.5373945236206055, "step": 5370 }, { "epoch": 0.4584576054537708, "grad_norm": 1.8828125, "learning_rate": 3.055555555555556e-05, "loss": 1.6871925354003907, "step": 5380 }, { "epoch": 0.45930975713677036, "grad_norm": 2.578125, "learning_rate": 3.0612360827084755e-05, "loss": 1.6597246170043944, "step": 5390 }, { "epoch": 0.4601619088197699, "grad_norm": 3.59375, "learning_rate": 3.0669166098613956e-05, "loss": 1.658172035217285, "step": 5400 }, { "epoch": 0.4610140605027695, "grad_norm": 2.25, "learning_rate": 3.072597137014315e-05, "loss": 1.6962909698486328, "step": 5410 }, { "epoch": 0.46186621218576907, "grad_norm": 2.125, "learning_rate": 3.078277664167235e-05, "loss": 1.5675886154174805, "step": 5420 }, { "epoch": 0.46271836386876863, "grad_norm": 4.8125, "learning_rate": 3.083958191320155e-05, "loss": 1.7222980499267577, "step": 5430 }, { "epoch": 0.4635705155517682, "grad_norm": 2.375, "learning_rate": 3.0896387184730746e-05, "loss": 1.5726357460021974, "step": 5440 }, { "epoch": 0.46442266723476777, "grad_norm": 3.09375, "learning_rate": 3.095319245625995e-05, "loss": 1.5770720481872558, "step": 5450 }, { "epoch": 0.46527481891776734, "grad_norm": 1.703125, "learning_rate": 3.100999772778914e-05, "loss": 1.6962699890136719, "step": 5460 }, { "epoch": 0.4661269706007669, "grad_norm": 3.390625, "learning_rate": 3.106680299931834e-05, "loss": 1.718454360961914, "step": 5470 }, { "epoch": 0.46697912228376653, "grad_norm": 2.484375, "learning_rate": 3.1123608270847536e-05, "loss": 1.7035711288452149, "step": 5480 }, { "epoch": 0.4678312739667661, "grad_norm": 1.9609375, "learning_rate": 3.118041354237674e-05, "loss": 1.6011314392089844, "step": 5490 }, { "epoch": 0.4686834256497657, "grad_norm": 1.8203125, "learning_rate": 3.123721881390593e-05, "loss": 1.7749946594238282, "step": 5500 }, { "epoch": 0.46953557733276524, "grad_norm": 1.75, "learning_rate": 3.129402408543513e-05, "loss": 1.6146448135375977, "step": 5510 }, { "epoch": 0.4703877290157648, "grad_norm": 4.6875, "learning_rate": 3.135082935696433e-05, "loss": 1.7236051559448242, "step": 5520 }, { "epoch": 0.4712398806987644, "grad_norm": 1.65625, "learning_rate": 3.140763462849353e-05, "loss": 1.6478240966796875, "step": 5530 }, { "epoch": 0.47209203238176395, "grad_norm": 2.421875, "learning_rate": 3.146443990002272e-05, "loss": 1.6221492767333985, "step": 5540 }, { "epoch": 0.4729441840647635, "grad_norm": 2.375, "learning_rate": 3.152124517155192e-05, "loss": 1.6969263076782226, "step": 5550 }, { "epoch": 0.4737963357477631, "grad_norm": 2.390625, "learning_rate": 3.157805044308112e-05, "loss": 1.5758830070495606, "step": 5560 }, { "epoch": 0.47464848743076266, "grad_norm": 3.125, "learning_rate": 3.163485571461032e-05, "loss": 1.7200294494628907, "step": 5570 }, { "epoch": 0.4755006391137622, "grad_norm": 2.53125, "learning_rate": 3.169166098613951e-05, "loss": 1.7965293884277345, "step": 5580 }, { "epoch": 0.47635279079676185, "grad_norm": 1.8515625, "learning_rate": 3.174846625766871e-05, "loss": 1.7517217636108398, "step": 5590 }, { "epoch": 0.4772049424797614, "grad_norm": 1.6171875, "learning_rate": 3.1805271529197914e-05, "loss": 1.7771368026733398, "step": 5600 }, { "epoch": 0.478057094162761, "grad_norm": 2.953125, "learning_rate": 3.186207680072711e-05, "loss": 1.667861557006836, "step": 5610 }, { "epoch": 0.47890924584576056, "grad_norm": 3.125, "learning_rate": 3.191888207225631e-05, "loss": 1.5379666328430175, "step": 5620 }, { "epoch": 0.4797613975287601, "grad_norm": 2.125, "learning_rate": 3.1975687343785503e-05, "loss": 1.577298927307129, "step": 5630 }, { "epoch": 0.4806135492117597, "grad_norm": 2.125, "learning_rate": 3.2032492615314704e-05, "loss": 1.7194950103759765, "step": 5640 }, { "epoch": 0.48146570089475926, "grad_norm": 2.171875, "learning_rate": 3.20892978868439e-05, "loss": 1.6017820358276367, "step": 5650 }, { "epoch": 0.48231785257775883, "grad_norm": 3.0, "learning_rate": 3.21461031583731e-05, "loss": 1.674646759033203, "step": 5660 }, { "epoch": 0.4831700042607584, "grad_norm": 2.46875, "learning_rate": 3.2202908429902294e-05, "loss": 1.7410593032836914, "step": 5670 }, { "epoch": 0.48402215594375797, "grad_norm": 3.9375, "learning_rate": 3.2259713701431495e-05, "loss": 1.6946247100830079, "step": 5680 }, { "epoch": 0.48487430762675754, "grad_norm": 2.34375, "learning_rate": 3.231651897296069e-05, "loss": 1.70433349609375, "step": 5690 }, { "epoch": 0.48572645930975716, "grad_norm": 3.0, "learning_rate": 3.237332424448989e-05, "loss": 1.5195603370666504, "step": 5700 }, { "epoch": 0.48657861099275673, "grad_norm": 2.6875, "learning_rate": 3.2430129516019084e-05, "loss": 1.5450196266174316, "step": 5710 }, { "epoch": 0.4874307626757563, "grad_norm": 3.890625, "learning_rate": 3.2486934787548285e-05, "loss": 1.617152214050293, "step": 5720 }, { "epoch": 0.48828291435875587, "grad_norm": 4.28125, "learning_rate": 3.254374005907748e-05, "loss": 1.6053503036499024, "step": 5730 }, { "epoch": 0.48913506604175544, "grad_norm": 2.328125, "learning_rate": 3.260054533060668e-05, "loss": 1.5807092666625977, "step": 5740 }, { "epoch": 0.489987217724755, "grad_norm": 2.15625, "learning_rate": 3.2657350602135875e-05, "loss": 1.666238021850586, "step": 5750 }, { "epoch": 0.4908393694077546, "grad_norm": 1.84375, "learning_rate": 3.2714155873665075e-05, "loss": 1.7004638671875, "step": 5760 }, { "epoch": 0.49169152109075415, "grad_norm": 1.875, "learning_rate": 3.2770961145194276e-05, "loss": 1.5098599433898925, "step": 5770 }, { "epoch": 0.4925436727737537, "grad_norm": 2.265625, "learning_rate": 3.282776641672347e-05, "loss": 1.7283357620239257, "step": 5780 }, { "epoch": 0.4933958244567533, "grad_norm": 2.09375, "learning_rate": 3.288457168825267e-05, "loss": 1.6490198135375977, "step": 5790 }, { "epoch": 0.49424797613975285, "grad_norm": 2.90625, "learning_rate": 3.2941376959781866e-05, "loss": 1.5780851364135742, "step": 5800 }, { "epoch": 0.4951001278227524, "grad_norm": 2.4375, "learning_rate": 3.299818223131107e-05, "loss": 1.6356798171997071, "step": 5810 }, { "epoch": 0.49595227950575205, "grad_norm": 3.359375, "learning_rate": 3.305498750284027e-05, "loss": 1.6724603652954102, "step": 5820 }, { "epoch": 0.4968044311887516, "grad_norm": 3.125, "learning_rate": 3.311179277436947e-05, "loss": 1.646946334838867, "step": 5830 }, { "epoch": 0.4976565828717512, "grad_norm": 1.734375, "learning_rate": 3.316859804589866e-05, "loss": 1.5521369934082032, "step": 5840 }, { "epoch": 0.49850873455475075, "grad_norm": 3.546875, "learning_rate": 3.3225403317427864e-05, "loss": 1.6835994720458984, "step": 5850 }, { "epoch": 0.4993608862377503, "grad_norm": 2.1875, "learning_rate": 3.328220858895706e-05, "loss": 1.5952579498291015, "step": 5860 }, { "epoch": 0.49995739241585, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6196, "eval_loss": 1.8402256965637207, "eval_runtime": 54.3051, "eval_samples_per_second": 363.907, "eval_steps_per_second": 22.76, "step": 5867 }, { "epoch": 0.5002130379207499, "grad_norm": 2.28125, "learning_rate": 3.333901386048626e-05, "loss": 1.6132888793945312, "step": 5870 }, { "epoch": 0.5010651896037495, "grad_norm": 2.640625, "learning_rate": 3.339581913201545e-05, "loss": 1.5635841369628907, "step": 5880 }, { "epoch": 0.501917341286749, "grad_norm": 2.375, "learning_rate": 3.3452624403544654e-05, "loss": 1.5986040115356446, "step": 5890 }, { "epoch": 0.5027694929697486, "grad_norm": 2.75, "learning_rate": 3.350942967507385e-05, "loss": 1.5915512084960937, "step": 5900 }, { "epoch": 0.5036216446527482, "grad_norm": 2.421875, "learning_rate": 3.356623494660305e-05, "loss": 1.6211275100708007, "step": 5910 }, { "epoch": 0.5044737963357477, "grad_norm": 1.9921875, "learning_rate": 3.3623040218132244e-05, "loss": 1.6742708206176757, "step": 5920 }, { "epoch": 0.5053259480187473, "grad_norm": 2.3125, "learning_rate": 3.3679845489661445e-05, "loss": 1.6767108917236329, "step": 5930 }, { "epoch": 0.5061780997017469, "grad_norm": 2.4375, "learning_rate": 3.373665076119064e-05, "loss": 1.7349840164184571, "step": 5940 }, { "epoch": 0.5070302513847464, "grad_norm": 2.40625, "learning_rate": 3.379345603271984e-05, "loss": 1.5701759338378907, "step": 5950 }, { "epoch": 0.507882403067746, "grad_norm": 2.59375, "learning_rate": 3.3850261304249034e-05, "loss": 1.6626012802124024, "step": 5960 }, { "epoch": 0.5087345547507456, "grad_norm": 2.125, "learning_rate": 3.3907066575778235e-05, "loss": 1.4906583786010743, "step": 5970 }, { "epoch": 0.5095867064337452, "grad_norm": 2.71875, "learning_rate": 3.396387184730743e-05, "loss": 1.4974308013916016, "step": 5980 }, { "epoch": 0.5104388581167448, "grad_norm": 3.015625, "learning_rate": 3.402067711883663e-05, "loss": 1.4694526672363282, "step": 5990 }, { "epoch": 0.5112910097997444, "grad_norm": 3.75, "learning_rate": 3.407748239036583e-05, "loss": 1.5988316535949707, "step": 6000 }, { "epoch": 0.512143161482744, "grad_norm": 4.125, "learning_rate": 3.4134287661895025e-05, "loss": 1.7017013549804687, "step": 6010 }, { "epoch": 0.5129953131657435, "grad_norm": 2.453125, "learning_rate": 3.4191092933424226e-05, "loss": 1.723230743408203, "step": 6020 }, { "epoch": 0.5138474648487431, "grad_norm": 3.734375, "learning_rate": 3.424789820495342e-05, "loss": 1.638435173034668, "step": 6030 }, { "epoch": 0.5146996165317427, "grad_norm": 2.8125, "learning_rate": 3.430470347648262e-05, "loss": 1.611750030517578, "step": 6040 }, { "epoch": 0.5155517682147422, "grad_norm": 3.34375, "learning_rate": 3.4361508748011816e-05, "loss": 1.5459184646606445, "step": 6050 }, { "epoch": 0.5164039198977418, "grad_norm": 2.421875, "learning_rate": 3.441831401954102e-05, "loss": 1.5852476119995118, "step": 6060 }, { "epoch": 0.5172560715807414, "grad_norm": 3.6875, "learning_rate": 3.447511929107021e-05, "loss": 1.6705646514892578, "step": 6070 }, { "epoch": 0.518108223263741, "grad_norm": 2.515625, "learning_rate": 3.453192456259941e-05, "loss": 1.6316514968872071, "step": 6080 }, { "epoch": 0.5189603749467405, "grad_norm": 2.265625, "learning_rate": 3.4588729834128606e-05, "loss": 1.6480091094970704, "step": 6090 }, { "epoch": 0.5198125266297401, "grad_norm": 2.578125, "learning_rate": 3.464553510565781e-05, "loss": 1.4838159561157227, "step": 6100 }, { "epoch": 0.5206646783127397, "grad_norm": 3.21875, "learning_rate": 3.4702340377187e-05, "loss": 1.6845561981201171, "step": 6110 }, { "epoch": 0.5215168299957392, "grad_norm": 2.484375, "learning_rate": 3.47591456487162e-05, "loss": 1.493140983581543, "step": 6120 }, { "epoch": 0.5223689816787388, "grad_norm": 2.109375, "learning_rate": 3.4815950920245396e-05, "loss": 1.67198486328125, "step": 6130 }, { "epoch": 0.5232211333617384, "grad_norm": 2.03125, "learning_rate": 3.48727561917746e-05, "loss": 1.6258140563964845, "step": 6140 }, { "epoch": 0.5240732850447379, "grad_norm": 2.0625, "learning_rate": 3.492956146330379e-05, "loss": 1.512892246246338, "step": 6150 }, { "epoch": 0.5249254367277375, "grad_norm": 1.890625, "learning_rate": 3.498636673483299e-05, "loss": 1.7333688735961914, "step": 6160 }, { "epoch": 0.5257775884107371, "grad_norm": 2.65625, "learning_rate": 3.5043172006362193e-05, "loss": 1.4758661270141602, "step": 6170 }, { "epoch": 0.5266297400937366, "grad_norm": 5.59375, "learning_rate": 3.509997727789139e-05, "loss": 1.5940799713134766, "step": 6180 }, { "epoch": 0.5274818917767362, "grad_norm": 2.1875, "learning_rate": 3.515678254942059e-05, "loss": 1.5998306274414062, "step": 6190 }, { "epoch": 0.5283340434597358, "grad_norm": 2.671875, "learning_rate": 3.521358782094978e-05, "loss": 1.7066631317138672, "step": 6200 }, { "epoch": 0.5291861951427355, "grad_norm": 2.9375, "learning_rate": 3.5270393092478984e-05, "loss": 1.5881385803222656, "step": 6210 }, { "epoch": 0.530038346825735, "grad_norm": 1.9140625, "learning_rate": 3.532719836400818e-05, "loss": 1.6089658737182617, "step": 6220 }, { "epoch": 0.5308904985087346, "grad_norm": 3.84375, "learning_rate": 3.538400363553738e-05, "loss": 1.6880929946899415, "step": 6230 }, { "epoch": 0.5317426501917342, "grad_norm": 2.953125, "learning_rate": 3.544080890706657e-05, "loss": 1.5783352851867676, "step": 6240 }, { "epoch": 0.5325948018747337, "grad_norm": 3.828125, "learning_rate": 3.5497614178595774e-05, "loss": 1.489849853515625, "step": 6250 }, { "epoch": 0.5334469535577333, "grad_norm": 2.765625, "learning_rate": 3.555441945012497e-05, "loss": 1.5845025062561036, "step": 6260 }, { "epoch": 0.5342991052407329, "grad_norm": 3.0625, "learning_rate": 3.5611224721654176e-05, "loss": 1.5485843658447265, "step": 6270 }, { "epoch": 0.5351512569237324, "grad_norm": 2.40625, "learning_rate": 3.566802999318337e-05, "loss": 1.658348274230957, "step": 6280 }, { "epoch": 0.536003408606732, "grad_norm": 4.46875, "learning_rate": 3.572483526471257e-05, "loss": 1.5507662773132325, "step": 6290 }, { "epoch": 0.5368555602897316, "grad_norm": 2.5, "learning_rate": 3.5781640536241765e-05, "loss": 1.4487754821777343, "step": 6300 }, { "epoch": 0.5377077119727312, "grad_norm": 2.359375, "learning_rate": 3.5838445807770966e-05, "loss": 1.3841605186462402, "step": 6310 }, { "epoch": 0.5385598636557307, "grad_norm": 3.15625, "learning_rate": 3.589525107930016e-05, "loss": 1.596343421936035, "step": 6320 }, { "epoch": 0.5394120153387303, "grad_norm": 3.984375, "learning_rate": 3.595205635082936e-05, "loss": 1.5664342880249023, "step": 6330 }, { "epoch": 0.5402641670217299, "grad_norm": 2.703125, "learning_rate": 3.6008861622358556e-05, "loss": 1.6195362091064454, "step": 6340 }, { "epoch": 0.5411163187047294, "grad_norm": 1.4140625, "learning_rate": 3.606566689388776e-05, "loss": 1.7268781661987305, "step": 6350 }, { "epoch": 0.541968470387729, "grad_norm": 2.109375, "learning_rate": 3.612247216541695e-05, "loss": 1.664103889465332, "step": 6360 }, { "epoch": 0.5428206220707286, "grad_norm": 2.40625, "learning_rate": 3.617927743694615e-05, "loss": 1.4623477935791016, "step": 6370 }, { "epoch": 0.5436727737537281, "grad_norm": 1.9921875, "learning_rate": 3.623608270847535e-05, "loss": 1.5097785949707032, "step": 6380 }, { "epoch": 0.5445249254367277, "grad_norm": 2.375, "learning_rate": 3.629288798000455e-05, "loss": 1.5764493942260742, "step": 6390 }, { "epoch": 0.5453770771197273, "grad_norm": 3.34375, "learning_rate": 3.634969325153375e-05, "loss": 1.5527497291564942, "step": 6400 }, { "epoch": 0.5462292288027268, "grad_norm": 3.0625, "learning_rate": 3.640649852306294e-05, "loss": 1.5442329406738282, "step": 6410 }, { "epoch": 0.5470813804857264, "grad_norm": 2.5625, "learning_rate": 3.646330379459214e-05, "loss": 1.6177642822265625, "step": 6420 }, { "epoch": 0.547933532168726, "grad_norm": 2.9375, "learning_rate": 3.652010906612134e-05, "loss": 1.526843547821045, "step": 6430 }, { "epoch": 0.5487856838517257, "grad_norm": 3.46875, "learning_rate": 3.657691433765054e-05, "loss": 1.5071699142456054, "step": 6440 }, { "epoch": 0.5496378355347252, "grad_norm": 3.3125, "learning_rate": 3.663371960917973e-05, "loss": 1.6300018310546875, "step": 6450 }, { "epoch": 0.5504899872177248, "grad_norm": 2.015625, "learning_rate": 3.6690524880708934e-05, "loss": 1.6713695526123047, "step": 6460 }, { "epoch": 0.5513421389007244, "grad_norm": 2.296875, "learning_rate": 3.674733015223813e-05, "loss": 1.565483856201172, "step": 6470 }, { "epoch": 0.5521942905837239, "grad_norm": 3.078125, "learning_rate": 3.680413542376733e-05, "loss": 1.5005157470703125, "step": 6480 }, { "epoch": 0.5530464422667235, "grad_norm": 4.3125, "learning_rate": 3.686094069529652e-05, "loss": 1.575784397125244, "step": 6490 }, { "epoch": 0.5538985939497231, "grad_norm": 1.921875, "learning_rate": 3.6917745966825724e-05, "loss": 1.6577960968017578, "step": 6500 }, { "epoch": 0.5547507456327226, "grad_norm": 2.109375, "learning_rate": 3.697455123835492e-05, "loss": 1.6875659942626953, "step": 6510 }, { "epoch": 0.5556028973157222, "grad_norm": 2.59375, "learning_rate": 3.703135650988412e-05, "loss": 1.588006591796875, "step": 6520 }, { "epoch": 0.5564550489987218, "grad_norm": 1.875, "learning_rate": 3.708816178141331e-05, "loss": 1.6177589416503906, "step": 6530 }, { "epoch": 0.5573072006817213, "grad_norm": 2.1875, "learning_rate": 3.7144967052942514e-05, "loss": 1.5093853950500489, "step": 6540 }, { "epoch": 0.5581593523647209, "grad_norm": 2.140625, "learning_rate": 3.7201772324471715e-05, "loss": 1.5467951774597168, "step": 6550 }, { "epoch": 0.5590115040477205, "grad_norm": 2.703125, "learning_rate": 3.725857759600091e-05, "loss": 1.468998336791992, "step": 6560 }, { "epoch": 0.5598636557307201, "grad_norm": 3.515625, "learning_rate": 3.731538286753011e-05, "loss": 1.4736592292785644, "step": 6570 }, { "epoch": 0.5607158074137196, "grad_norm": 2.5, "learning_rate": 3.7372188139059305e-05, "loss": 1.5877131462097167, "step": 6580 }, { "epoch": 0.5615679590967192, "grad_norm": 2.453125, "learning_rate": 3.7428993410588506e-05, "loss": 1.4794707298278809, "step": 6590 }, { "epoch": 0.5624201107797188, "grad_norm": 2.1875, "learning_rate": 3.74857986821177e-05, "loss": 1.5598008155822753, "step": 6600 }, { "epoch": 0.5632722624627183, "grad_norm": 2.328125, "learning_rate": 3.75426039536469e-05, "loss": 1.5837244033813476, "step": 6610 }, { "epoch": 0.5641244141457179, "grad_norm": 2.171875, "learning_rate": 3.7599409225176095e-05, "loss": 1.6200679779052733, "step": 6620 }, { "epoch": 0.5649765658287175, "grad_norm": 2.0625, "learning_rate": 3.7656214496705296e-05, "loss": 1.6108659744262694, "step": 6630 }, { "epoch": 0.565828717511717, "grad_norm": 2.0, "learning_rate": 3.771301976823449e-05, "loss": 1.6593133926391601, "step": 6640 }, { "epoch": 0.5666808691947166, "grad_norm": 3.078125, "learning_rate": 3.776982503976369e-05, "loss": 1.5719993591308594, "step": 6650 }, { "epoch": 0.5675330208777163, "grad_norm": 2.3125, "learning_rate": 3.7826630311292885e-05, "loss": 1.6277524948120117, "step": 6660 }, { "epoch": 0.5683851725607159, "grad_norm": 1.7109375, "learning_rate": 3.7883435582822086e-05, "loss": 1.5893933296203613, "step": 6670 }, { "epoch": 0.5692373242437154, "grad_norm": 2.546875, "learning_rate": 3.794024085435128e-05, "loss": 1.6269601821899413, "step": 6680 }, { "epoch": 0.570089475926715, "grad_norm": 2.453125, "learning_rate": 3.799704612588048e-05, "loss": 1.5729507446289062, "step": 6690 }, { "epoch": 0.5709416276097146, "grad_norm": 2.6875, "learning_rate": 3.8053851397409676e-05, "loss": 1.6154645919799804, "step": 6700 }, { "epoch": 0.5717937792927141, "grad_norm": 1.765625, "learning_rate": 3.811065666893888e-05, "loss": 1.5907581329345704, "step": 6710 }, { "epoch": 0.5726459309757137, "grad_norm": 2.328125, "learning_rate": 3.816746194046808e-05, "loss": 1.638677215576172, "step": 6720 }, { "epoch": 0.5734980826587133, "grad_norm": 1.6953125, "learning_rate": 3.822426721199728e-05, "loss": 1.6257532119750977, "step": 6730 }, { "epoch": 0.5743502343417128, "grad_norm": 2.015625, "learning_rate": 3.828107248352647e-05, "loss": 1.605808639526367, "step": 6740 }, { "epoch": 0.5752023860247124, "grad_norm": 2.96875, "learning_rate": 3.8337877755055674e-05, "loss": 1.5757763862609864, "step": 6750 }, { "epoch": 0.576054537707712, "grad_norm": 2.484375, "learning_rate": 3.839468302658487e-05, "loss": 1.6503097534179687, "step": 6760 }, { "epoch": 0.5769066893907115, "grad_norm": 2.28125, "learning_rate": 3.845148829811407e-05, "loss": 1.5395710945129395, "step": 6770 }, { "epoch": 0.5777588410737111, "grad_norm": 2.09375, "learning_rate": 3.850829356964327e-05, "loss": 1.5377091407775878, "step": 6780 }, { "epoch": 0.5786109927567107, "grad_norm": 1.703125, "learning_rate": 3.8565098841172464e-05, "loss": 1.580072784423828, "step": 6790 }, { "epoch": 0.5794631444397103, "grad_norm": 2.203125, "learning_rate": 3.8621904112701665e-05, "loss": 1.4713946342468263, "step": 6800 }, { "epoch": 0.5803152961227098, "grad_norm": 3.96875, "learning_rate": 3.867870938423086e-05, "loss": 1.445005226135254, "step": 6810 }, { "epoch": 0.5811674478057094, "grad_norm": 4.53125, "learning_rate": 3.873551465576006e-05, "loss": 1.565537452697754, "step": 6820 }, { "epoch": 0.582019599488709, "grad_norm": 2.0625, "learning_rate": 3.8792319927289255e-05, "loss": 1.608312225341797, "step": 6830 }, { "epoch": 0.5828717511717085, "grad_norm": 3.375, "learning_rate": 3.8849125198818455e-05, "loss": 1.652672004699707, "step": 6840 }, { "epoch": 0.5837239028547081, "grad_norm": 2.1875, "learning_rate": 3.890593047034765e-05, "loss": 1.6770214080810546, "step": 6850 }, { "epoch": 0.5845760545377077, "grad_norm": 2.703125, "learning_rate": 3.896273574187685e-05, "loss": 1.7213994979858398, "step": 6860 }, { "epoch": 0.5854282062207072, "grad_norm": 1.765625, "learning_rate": 3.9019541013406045e-05, "loss": 1.4941165924072266, "step": 6870 }, { "epoch": 0.5862803579037068, "grad_norm": 2.90625, "learning_rate": 3.9076346284935246e-05, "loss": 1.625014305114746, "step": 6880 }, { "epoch": 0.5871325095867065, "grad_norm": 2.359375, "learning_rate": 3.913315155646444e-05, "loss": 1.4931087493896484, "step": 6890 }, { "epoch": 0.587984661269706, "grad_norm": 2.84375, "learning_rate": 3.918995682799364e-05, "loss": 1.5034214019775392, "step": 6900 }, { "epoch": 0.5888368129527056, "grad_norm": 3.21875, "learning_rate": 3.9246762099522835e-05, "loss": 1.5171854019165039, "step": 6910 }, { "epoch": 0.5896889646357052, "grad_norm": 2.453125, "learning_rate": 3.9303567371052036e-05, "loss": 1.4951739311218262, "step": 6920 }, { "epoch": 0.5905411163187048, "grad_norm": 2.921875, "learning_rate": 3.936037264258123e-05, "loss": 1.3678095817565918, "step": 6930 }, { "epoch": 0.5913932680017043, "grad_norm": 2.40625, "learning_rate": 3.941717791411043e-05, "loss": 1.4891133308410645, "step": 6940 }, { "epoch": 0.5922454196847039, "grad_norm": 1.421875, "learning_rate": 3.947398318563963e-05, "loss": 1.6572811126708984, "step": 6950 }, { "epoch": 0.5930975713677035, "grad_norm": 3.109375, "learning_rate": 3.9530788457168827e-05, "loss": 1.5185030937194823, "step": 6960 }, { "epoch": 0.593949723050703, "grad_norm": 1.5390625, "learning_rate": 3.958759372869803e-05, "loss": 1.666409683227539, "step": 6970 }, { "epoch": 0.5948018747337026, "grad_norm": 2.1875, "learning_rate": 3.964439900022722e-05, "loss": 1.6802967071533204, "step": 6980 }, { "epoch": 0.5956540264167022, "grad_norm": 2.28125, "learning_rate": 3.970120427175642e-05, "loss": 1.5952302932739257, "step": 6990 }, { "epoch": 0.5965061780997017, "grad_norm": 2.546875, "learning_rate": 3.975800954328562e-05, "loss": 1.6771442413330078, "step": 7000 }, { "epoch": 0.5973583297827013, "grad_norm": 2.0625, "learning_rate": 3.981481481481482e-05, "loss": 1.4942998886108398, "step": 7010 }, { "epoch": 0.5982104814657009, "grad_norm": 3.84375, "learning_rate": 3.987162008634401e-05, "loss": 1.4492556571960449, "step": 7020 }, { "epoch": 0.5990626331487005, "grad_norm": 2.5625, "learning_rate": 3.992842535787321e-05, "loss": 1.5903990745544434, "step": 7030 }, { "epoch": 0.5999147848317, "grad_norm": 1.9453125, "learning_rate": 3.998523062940241e-05, "loss": 1.3911609649658203, "step": 7040 }, { "epoch": 0.6007669365146996, "grad_norm": 4.03125, "learning_rate": 4.004203590093161e-05, "loss": 1.521592903137207, "step": 7050 }, { "epoch": 0.6016190881976992, "grad_norm": 3.625, "learning_rate": 4.00988411724608e-05, "loss": 1.6142061233520508, "step": 7060 }, { "epoch": 0.6024712398806987, "grad_norm": 2.796875, "learning_rate": 4.015564644399e-05, "loss": 1.6196815490722656, "step": 7070 }, { "epoch": 0.6033233915636983, "grad_norm": 2.828125, "learning_rate": 4.02124517155192e-05, "loss": 1.664220428466797, "step": 7080 }, { "epoch": 0.6041755432466979, "grad_norm": 2.4375, "learning_rate": 4.02692569870484e-05, "loss": 1.484408187866211, "step": 7090 }, { "epoch": 0.6050276949296974, "grad_norm": 2.5, "learning_rate": 4.032606225857759e-05, "loss": 1.472763729095459, "step": 7100 }, { "epoch": 0.6058798466126971, "grad_norm": 2.875, "learning_rate": 4.0382867530106794e-05, "loss": 1.4880710601806642, "step": 7110 }, { "epoch": 0.6067319982956967, "grad_norm": 3.125, "learning_rate": 4.0439672801635995e-05, "loss": 1.6559080123901366, "step": 7120 }, { "epoch": 0.6075841499786963, "grad_norm": 2.265625, "learning_rate": 4.049647807316519e-05, "loss": 1.592278289794922, "step": 7130 }, { "epoch": 0.6084363016616958, "grad_norm": 2.28125, "learning_rate": 4.055328334469439e-05, "loss": 1.4581765174865722, "step": 7140 }, { "epoch": 0.6092884533446954, "grad_norm": 2.578125, "learning_rate": 4.0610088616223584e-05, "loss": 1.5346181869506836, "step": 7150 }, { "epoch": 0.610140605027695, "grad_norm": 1.921875, "learning_rate": 4.0666893887752785e-05, "loss": 1.6967073440551759, "step": 7160 }, { "epoch": 0.6109927567106945, "grad_norm": 2.59375, "learning_rate": 4.0723699159281986e-05, "loss": 1.6154413223266602, "step": 7170 }, { "epoch": 0.6118449083936941, "grad_norm": 2.265625, "learning_rate": 4.078050443081119e-05, "loss": 1.4092897415161132, "step": 7180 }, { "epoch": 0.6126970600766937, "grad_norm": 3.078125, "learning_rate": 4.083730970234038e-05, "loss": 1.6090990066528321, "step": 7190 }, { "epoch": 0.6135492117596932, "grad_norm": 1.7109375, "learning_rate": 4.089411497386958e-05, "loss": 1.4968547821044922, "step": 7200 }, { "epoch": 0.6144013634426928, "grad_norm": 2.578125, "learning_rate": 4.0950920245398776e-05, "loss": 1.6127819061279296, "step": 7210 }, { "epoch": 0.6152535151256924, "grad_norm": 2.03125, "learning_rate": 4.100772551692798e-05, "loss": 1.5914322853088378, "step": 7220 }, { "epoch": 0.6161056668086919, "grad_norm": 2.65625, "learning_rate": 4.106453078845717e-05, "loss": 1.57374906539917, "step": 7230 }, { "epoch": 0.6169578184916915, "grad_norm": 2.984375, "learning_rate": 4.112133605998637e-05, "loss": 1.4930489540100098, "step": 7240 }, { "epoch": 0.6178099701746911, "grad_norm": 1.8359375, "learning_rate": 4.117814133151557e-05, "loss": 1.5231494903564453, "step": 7250 }, { "epoch": 0.6186621218576907, "grad_norm": 3.96875, "learning_rate": 4.123494660304477e-05, "loss": 1.7707647323608398, "step": 7260 }, { "epoch": 0.6195142735406902, "grad_norm": 2.125, "learning_rate": 4.129175187457396e-05, "loss": 1.6947399139404298, "step": 7270 }, { "epoch": 0.6203664252236898, "grad_norm": 3.6875, "learning_rate": 4.134855714610316e-05, "loss": 1.3931601524353028, "step": 7280 }, { "epoch": 0.6212185769066894, "grad_norm": 2.890625, "learning_rate": 4.140536241763236e-05, "loss": 1.3306796073913574, "step": 7290 }, { "epoch": 0.6220707285896889, "grad_norm": 2.0, "learning_rate": 4.146216768916156e-05, "loss": 1.52656888961792, "step": 7300 }, { "epoch": 0.6229228802726885, "grad_norm": 2.40625, "learning_rate": 4.151897296069075e-05, "loss": 1.6264070510864257, "step": 7310 }, { "epoch": 0.6237750319556881, "grad_norm": 2.953125, "learning_rate": 4.157577823221995e-05, "loss": 1.4102313995361329, "step": 7320 }, { "epoch": 0.6246271836386876, "grad_norm": 2.703125, "learning_rate": 4.163258350374915e-05, "loss": 1.560826873779297, "step": 7330 }, { "epoch": 0.6254793353216873, "grad_norm": 1.9453125, "learning_rate": 4.168938877527835e-05, "loss": 1.5879319190979004, "step": 7340 }, { "epoch": 0.6263314870046869, "grad_norm": 3.5, "learning_rate": 4.174619404680755e-05, "loss": 1.4431940078735352, "step": 7350 }, { "epoch": 0.6271836386876865, "grad_norm": 1.890625, "learning_rate": 4.1802999318336744e-05, "loss": 1.479648494720459, "step": 7360 }, { "epoch": 0.628035790370686, "grad_norm": 2.515625, "learning_rate": 4.1859804589865945e-05, "loss": 1.4700020790100097, "step": 7370 }, { "epoch": 0.6288879420536856, "grad_norm": 2.359375, "learning_rate": 4.191660986139514e-05, "loss": 1.5747636795043944, "step": 7380 }, { "epoch": 0.6297400937366852, "grad_norm": 3.28125, "learning_rate": 4.197341513292434e-05, "loss": 1.5675525665283203, "step": 7390 }, { "epoch": 0.6305922454196847, "grad_norm": 3.25, "learning_rate": 4.2030220404453534e-05, "loss": 1.504664134979248, "step": 7400 }, { "epoch": 0.6314443971026843, "grad_norm": 2.28125, "learning_rate": 4.2087025675982735e-05, "loss": 1.472569465637207, "step": 7410 }, { "epoch": 0.6322965487856839, "grad_norm": 2.1875, "learning_rate": 4.214383094751193e-05, "loss": 1.5091018676757812, "step": 7420 }, { "epoch": 0.6331487004686834, "grad_norm": 3.75, "learning_rate": 4.220063621904113e-05, "loss": 1.5987555503845214, "step": 7430 }, { "epoch": 0.634000852151683, "grad_norm": 2.265625, "learning_rate": 4.2257441490570324e-05, "loss": 1.4471548080444336, "step": 7440 }, { "epoch": 0.6348530038346826, "grad_norm": 2.875, "learning_rate": 4.2314246762099525e-05, "loss": 1.4750150680541991, "step": 7450 }, { "epoch": 0.6357051555176821, "grad_norm": 3.296875, "learning_rate": 4.237105203362872e-05, "loss": 1.6660371780395509, "step": 7460 }, { "epoch": 0.6365573072006817, "grad_norm": 3.75, "learning_rate": 4.242785730515792e-05, "loss": 1.2341299057006836, "step": 7470 }, { "epoch": 0.6374094588836813, "grad_norm": 2.28125, "learning_rate": 4.2484662576687115e-05, "loss": 1.5386623382568358, "step": 7480 }, { "epoch": 0.6382616105666808, "grad_norm": 3.109375, "learning_rate": 4.2541467848216316e-05, "loss": 1.522665023803711, "step": 7490 }, { "epoch": 0.6391137622496804, "grad_norm": 1.7109375, "learning_rate": 4.2598273119745517e-05, "loss": 1.5357202529907226, "step": 7500 }, { "epoch": 0.63996591393268, "grad_norm": 2.28125, "learning_rate": 4.265507839127471e-05, "loss": 1.5832084655761718, "step": 7510 }, { "epoch": 0.6408180656156796, "grad_norm": 2.890625, "learning_rate": 4.271188366280391e-05, "loss": 1.5495640754699707, "step": 7520 }, { "epoch": 0.6416702172986791, "grad_norm": 1.9296875, "learning_rate": 4.2768688934333106e-05, "loss": 1.6494876861572265, "step": 7530 }, { "epoch": 0.6425223689816787, "grad_norm": 2.453125, "learning_rate": 4.282549420586231e-05, "loss": 1.614303207397461, "step": 7540 }, { "epoch": 0.6433745206646783, "grad_norm": 2.71875, "learning_rate": 4.28822994773915e-05, "loss": 1.4222282409667968, "step": 7550 }, { "epoch": 0.6442266723476778, "grad_norm": 2.828125, "learning_rate": 4.29391047489207e-05, "loss": 1.4683493614196776, "step": 7560 }, { "epoch": 0.6450788240306775, "grad_norm": 2.1875, "learning_rate": 4.2995910020449896e-05, "loss": 1.5414005279541017, "step": 7570 }, { "epoch": 0.6459309757136771, "grad_norm": 1.734375, "learning_rate": 4.30527152919791e-05, "loss": 1.471998691558838, "step": 7580 }, { "epoch": 0.6467831273966766, "grad_norm": 3.234375, "learning_rate": 4.310952056350829e-05, "loss": 1.504387092590332, "step": 7590 }, { "epoch": 0.6476352790796762, "grad_norm": 3.09375, "learning_rate": 4.316632583503749e-05, "loss": 1.4986427307128907, "step": 7600 }, { "epoch": 0.6484874307626758, "grad_norm": 3.484375, "learning_rate": 4.3223131106566687e-05, "loss": 1.594170379638672, "step": 7610 }, { "epoch": 0.6493395824456754, "grad_norm": 2.34375, "learning_rate": 4.327993637809589e-05, "loss": 1.6669710159301758, "step": 7620 }, { "epoch": 0.6501917341286749, "grad_norm": 2.234375, "learning_rate": 4.333674164962509e-05, "loss": 1.4474031448364257, "step": 7630 }, { "epoch": 0.6510438858116745, "grad_norm": 3.78125, "learning_rate": 4.339354692115429e-05, "loss": 1.661581802368164, "step": 7640 }, { "epoch": 0.6518960374946741, "grad_norm": 3.296875, "learning_rate": 4.3450352192683484e-05, "loss": 1.6364736557006836, "step": 7650 }, { "epoch": 0.6527481891776736, "grad_norm": 2.46875, "learning_rate": 4.3507157464212685e-05, "loss": 1.5027189254760742, "step": 7660 }, { "epoch": 0.6536003408606732, "grad_norm": 1.65625, "learning_rate": 4.356396273574188e-05, "loss": 1.5171278953552245, "step": 7670 }, { "epoch": 0.6544524925436728, "grad_norm": 2.703125, "learning_rate": 4.362076800727108e-05, "loss": 1.6341941833496094, "step": 7680 }, { "epoch": 0.6553046442266723, "grad_norm": 2.890625, "learning_rate": 4.3677573278800274e-05, "loss": 1.6163694381713867, "step": 7690 }, { "epoch": 0.6561567959096719, "grad_norm": 1.953125, "learning_rate": 4.3734378550329475e-05, "loss": 1.3673070907592773, "step": 7700 }, { "epoch": 0.6570089475926715, "grad_norm": 2.21875, "learning_rate": 4.379118382185867e-05, "loss": 1.5318901062011718, "step": 7710 }, { "epoch": 0.657861099275671, "grad_norm": 2.296875, "learning_rate": 4.384798909338787e-05, "loss": 1.5446605682373047, "step": 7720 }, { "epoch": 0.6587132509586706, "grad_norm": 1.6953125, "learning_rate": 4.390479436491707e-05, "loss": 1.6056890487670898, "step": 7730 }, { "epoch": 0.6595654026416702, "grad_norm": 2.046875, "learning_rate": 4.3961599636446265e-05, "loss": 1.5144453048706055, "step": 7740 }, { "epoch": 0.6604175543246698, "grad_norm": 1.9453125, "learning_rate": 4.4018404907975466e-05, "loss": 1.4554665565490723, "step": 7750 }, { "epoch": 0.6612697060076693, "grad_norm": 3.515625, "learning_rate": 4.407521017950466e-05, "loss": 1.5115887641906738, "step": 7760 }, { "epoch": 0.6621218576906689, "grad_norm": 2.421875, "learning_rate": 4.413201545103386e-05, "loss": 1.5202644348144532, "step": 7770 }, { "epoch": 0.6629740093736685, "grad_norm": 2.765625, "learning_rate": 4.4188820722563056e-05, "loss": 1.5347124099731446, "step": 7780 }, { "epoch": 0.6638261610566681, "grad_norm": 1.5390625, "learning_rate": 4.424562599409226e-05, "loss": 1.5229363441467285, "step": 7790 }, { "epoch": 0.6646783127396677, "grad_norm": 2.890625, "learning_rate": 4.430243126562145e-05, "loss": 1.4693553924560547, "step": 7800 }, { "epoch": 0.6655304644226673, "grad_norm": 3.109375, "learning_rate": 4.435923653715065e-05, "loss": 1.5325916290283204, "step": 7810 }, { "epoch": 0.6663826161056668, "grad_norm": 2.453125, "learning_rate": 4.4416041808679846e-05, "loss": 1.5399965286254882, "step": 7820 }, { "epoch": 0.6672347677886664, "grad_norm": 1.5390625, "learning_rate": 4.447284708020905e-05, "loss": 1.5229742050170898, "step": 7830 }, { "epoch": 0.668086919471666, "grad_norm": 2.078125, "learning_rate": 4.452965235173824e-05, "loss": 1.6565635681152344, "step": 7840 }, { "epoch": 0.6689390711546656, "grad_norm": 2.25, "learning_rate": 4.458645762326744e-05, "loss": 1.5015705108642579, "step": 7850 }, { "epoch": 0.6697912228376651, "grad_norm": 2.265625, "learning_rate": 4.4643262894796636e-05, "loss": 1.5125876426696778, "step": 7860 }, { "epoch": 0.6706433745206647, "grad_norm": 2.515625, "learning_rate": 4.470006816632584e-05, "loss": 1.4309945106506348, "step": 7870 }, { "epoch": 0.6714955262036643, "grad_norm": 2.765625, "learning_rate": 4.475687343785503e-05, "loss": 1.5023266792297363, "step": 7880 }, { "epoch": 0.6723476778866638, "grad_norm": 1.734375, "learning_rate": 4.481367870938423e-05, "loss": 1.4841794967651367, "step": 7890 }, { "epoch": 0.6731998295696634, "grad_norm": 2.984375, "learning_rate": 4.4870483980913434e-05, "loss": 1.381261157989502, "step": 7900 }, { "epoch": 0.674051981252663, "grad_norm": 1.859375, "learning_rate": 4.492728925244263e-05, "loss": 1.470491600036621, "step": 7910 }, { "epoch": 0.6749041329356625, "grad_norm": 2.625, "learning_rate": 4.498409452397183e-05, "loss": 1.5521530151367187, "step": 7920 }, { "epoch": 0.6757562846186621, "grad_norm": 2.96875, "learning_rate": 4.504089979550102e-05, "loss": 1.5751893997192383, "step": 7930 }, { "epoch": 0.6766084363016617, "grad_norm": 1.8515625, "learning_rate": 4.5097705067030224e-05, "loss": 1.5308305740356445, "step": 7940 }, { "epoch": 0.6774605879846612, "grad_norm": 2.015625, "learning_rate": 4.515451033855942e-05, "loss": 1.3674341201782227, "step": 7950 }, { "epoch": 0.6783127396676608, "grad_norm": 3.140625, "learning_rate": 4.521131561008862e-05, "loss": 1.309333610534668, "step": 7960 }, { "epoch": 0.6791648913506604, "grad_norm": 2.546875, "learning_rate": 4.526812088161781e-05, "loss": 1.4767793655395507, "step": 7970 }, { "epoch": 0.68001704303366, "grad_norm": 3.015625, "learning_rate": 4.5324926153147014e-05, "loss": 1.4995777130126953, "step": 7980 }, { "epoch": 0.6808691947166595, "grad_norm": 2.09375, "learning_rate": 4.538173142467621e-05, "loss": 1.5478962898254394, "step": 7990 }, { "epoch": 0.6817213463996591, "grad_norm": 3.59375, "learning_rate": 4.543853669620541e-05, "loss": 1.5505584716796874, "step": 8000 }, { "epoch": 0.6825734980826587, "grad_norm": 3.375, "learning_rate": 4.5495341967734604e-05, "loss": 1.4325838088989258, "step": 8010 }, { "epoch": 0.6834256497656583, "grad_norm": 3.125, "learning_rate": 4.5552147239263805e-05, "loss": 1.6296140670776367, "step": 8020 }, { "epoch": 0.6842778014486579, "grad_norm": 3.265625, "learning_rate": 4.5608952510793e-05, "loss": 1.3462490081787108, "step": 8030 }, { "epoch": 0.6851299531316575, "grad_norm": 2.734375, "learning_rate": 4.56657577823222e-05, "loss": 1.6294099807739257, "step": 8040 }, { "epoch": 0.685982104814657, "grad_norm": 2.859375, "learning_rate": 4.5722563053851394e-05, "loss": 1.6415676116943358, "step": 8050 }, { "epoch": 0.6868342564976566, "grad_norm": 2.078125, "learning_rate": 4.5779368325380595e-05, "loss": 1.5378034591674805, "step": 8060 }, { "epoch": 0.6876864081806562, "grad_norm": 2.3125, "learning_rate": 4.5836173596909796e-05, "loss": 1.4387834548950196, "step": 8070 }, { "epoch": 0.6885385598636558, "grad_norm": 3.109375, "learning_rate": 4.5892978868439e-05, "loss": 1.4546689987182617, "step": 8080 }, { "epoch": 0.6893907115466553, "grad_norm": 2.265625, "learning_rate": 4.594978413996819e-05, "loss": 1.4492469787597657, "step": 8090 }, { "epoch": 0.6902428632296549, "grad_norm": 2.796875, "learning_rate": 4.600658941149739e-05, "loss": 1.5522924423217774, "step": 8100 }, { "epoch": 0.6910950149126545, "grad_norm": 2.09375, "learning_rate": 4.6063394683026586e-05, "loss": 1.3489866256713867, "step": 8110 }, { "epoch": 0.691947166595654, "grad_norm": 2.453125, "learning_rate": 4.612019995455579e-05, "loss": 1.433194351196289, "step": 8120 }, { "epoch": 0.6927993182786536, "grad_norm": 1.9140625, "learning_rate": 4.617700522608499e-05, "loss": 1.5116780281066895, "step": 8130 }, { "epoch": 0.6936514699616532, "grad_norm": 2.3125, "learning_rate": 4.623381049761418e-05, "loss": 1.4881841659545898, "step": 8140 }, { "epoch": 0.6945036216446527, "grad_norm": 2.703125, "learning_rate": 4.629061576914338e-05, "loss": 1.477534294128418, "step": 8150 }, { "epoch": 0.6953557733276523, "grad_norm": 2.296875, "learning_rate": 4.634742104067258e-05, "loss": 1.5112692832946777, "step": 8160 }, { "epoch": 0.6962079250106519, "grad_norm": 2.578125, "learning_rate": 4.640422631220178e-05, "loss": 1.546157455444336, "step": 8170 }, { "epoch": 0.6970600766936514, "grad_norm": 1.9375, "learning_rate": 4.646103158373097e-05, "loss": 1.6062026977539063, "step": 8180 }, { "epoch": 0.697912228376651, "grad_norm": 2.921875, "learning_rate": 4.6517836855260174e-05, "loss": 1.4311408042907714, "step": 8190 }, { "epoch": 0.6987643800596506, "grad_norm": 2.109375, "learning_rate": 4.657464212678937e-05, "loss": 1.5178155899047852, "step": 8200 }, { "epoch": 0.6996165317426501, "grad_norm": 2.71875, "learning_rate": 4.663144739831857e-05, "loss": 1.4929768562316894, "step": 8210 }, { "epoch": 0.7004686834256497, "grad_norm": 2.890625, "learning_rate": 4.668825266984776e-05, "loss": 1.4565567016601562, "step": 8220 }, { "epoch": 0.7013208351086493, "grad_norm": 2.75, "learning_rate": 4.6745057941376964e-05, "loss": 1.4851869583129882, "step": 8230 }, { "epoch": 0.702172986791649, "grad_norm": 1.703125, "learning_rate": 4.680186321290616e-05, "loss": 1.5175625801086425, "step": 8240 }, { "epoch": 0.7030251384746485, "grad_norm": 3.984375, "learning_rate": 4.685866848443536e-05, "loss": 1.4914751052856445, "step": 8250 }, { "epoch": 0.7038772901576481, "grad_norm": 2.5, "learning_rate": 4.6915473755964553e-05, "loss": 1.425981330871582, "step": 8260 }, { "epoch": 0.7047294418406477, "grad_norm": 1.78125, "learning_rate": 4.6972279027493754e-05, "loss": 1.5601194381713868, "step": 8270 }, { "epoch": 0.7055815935236472, "grad_norm": 1.953125, "learning_rate": 4.702908429902295e-05, "loss": 1.4989837646484374, "step": 8280 }, { "epoch": 0.7064337452066468, "grad_norm": 2.765625, "learning_rate": 4.708588957055215e-05, "loss": 1.4688169479370117, "step": 8290 }, { "epoch": 0.7072858968896464, "grad_norm": 2.78125, "learning_rate": 4.714269484208135e-05, "loss": 1.4816082954406737, "step": 8300 }, { "epoch": 0.708138048572646, "grad_norm": 2.625, "learning_rate": 4.7199500113610545e-05, "loss": 1.500706958770752, "step": 8310 }, { "epoch": 0.7089902002556455, "grad_norm": 2.4375, "learning_rate": 4.7256305385139746e-05, "loss": 1.5168366432189941, "step": 8320 }, { "epoch": 0.7098423519386451, "grad_norm": 2.65625, "learning_rate": 4.731311065666894e-05, "loss": 1.456496810913086, "step": 8330 }, { "epoch": 0.7106945036216447, "grad_norm": 2.421875, "learning_rate": 4.736991592819814e-05, "loss": 1.4631858825683595, "step": 8340 }, { "epoch": 0.7115466553046442, "grad_norm": 1.65625, "learning_rate": 4.7426721199727335e-05, "loss": 1.6604219436645509, "step": 8350 }, { "epoch": 0.7123988069876438, "grad_norm": 2.046875, "learning_rate": 4.7483526471256536e-05, "loss": 1.5539042472839355, "step": 8360 }, { "epoch": 0.7132509586706434, "grad_norm": 2.109375, "learning_rate": 4.754033174278573e-05, "loss": 1.6064273834228515, "step": 8370 }, { "epoch": 0.7141031103536429, "grad_norm": 3.203125, "learning_rate": 4.759713701431493e-05, "loss": 1.5423373222351073, "step": 8380 }, { "epoch": 0.7149552620366425, "grad_norm": 2.96875, "learning_rate": 4.7653942285844125e-05, "loss": 1.3579566955566407, "step": 8390 }, { "epoch": 0.7158074137196421, "grad_norm": 4.59375, "learning_rate": 4.7710747557373326e-05, "loss": 1.590190315246582, "step": 8400 }, { "epoch": 0.7166595654026416, "grad_norm": 2.21875, "learning_rate": 4.776755282890252e-05, "loss": 1.5683403015136719, "step": 8410 }, { "epoch": 0.7175117170856412, "grad_norm": 2.796875, "learning_rate": 4.782435810043172e-05, "loss": 1.515010929107666, "step": 8420 }, { "epoch": 0.7183638687686408, "grad_norm": 3.15625, "learning_rate": 4.7881163371960916e-05, "loss": 1.6502479553222655, "step": 8430 }, { "epoch": 0.7192160204516403, "grad_norm": 2.1875, "learning_rate": 4.793796864349012e-05, "loss": 1.5512897491455078, "step": 8440 }, { "epoch": 0.7200681721346399, "grad_norm": 2.125, "learning_rate": 4.799477391501931e-05, "loss": 1.605518913269043, "step": 8450 }, { "epoch": 0.7209203238176395, "grad_norm": 2.65625, "learning_rate": 4.805157918654851e-05, "loss": 1.4100805282592774, "step": 8460 }, { "epoch": 0.7217724755006392, "grad_norm": 2.296875, "learning_rate": 4.810838445807771e-05, "loss": 1.4636306762695312, "step": 8470 }, { "epoch": 0.7226246271836387, "grad_norm": 3.59375, "learning_rate": 4.816518972960691e-05, "loss": 1.4141928672790527, "step": 8480 }, { "epoch": 0.7234767788666383, "grad_norm": 1.875, "learning_rate": 4.822199500113611e-05, "loss": 1.4545906066894532, "step": 8490 }, { "epoch": 0.7243289305496379, "grad_norm": 4.09375, "learning_rate": 4.82788002726653e-05, "loss": 1.5914154052734375, "step": 8500 }, { "epoch": 0.7251810822326374, "grad_norm": 3.25, "learning_rate": 4.83356055441945e-05, "loss": 1.4246201515197754, "step": 8510 }, { "epoch": 0.726033233915637, "grad_norm": 2.171875, "learning_rate": 4.83924108157237e-05, "loss": 1.5386890411376952, "step": 8520 }, { "epoch": 0.7268853855986366, "grad_norm": 4.25, "learning_rate": 4.84492160872529e-05, "loss": 1.473216438293457, "step": 8530 }, { "epoch": 0.7277375372816361, "grad_norm": 2.875, "learning_rate": 4.85060213587821e-05, "loss": 1.4706079483032226, "step": 8540 }, { "epoch": 0.7285896889646357, "grad_norm": 2.640625, "learning_rate": 4.85628266303113e-05, "loss": 1.44526948928833, "step": 8550 }, { "epoch": 0.7294418406476353, "grad_norm": 2.171875, "learning_rate": 4.8619631901840495e-05, "loss": 1.473697566986084, "step": 8560 }, { "epoch": 0.7302939923306349, "grad_norm": 2.34375, "learning_rate": 4.8676437173369696e-05, "loss": 1.5511651992797852, "step": 8570 }, { "epoch": 0.7311461440136344, "grad_norm": 2.28125, "learning_rate": 4.873324244489889e-05, "loss": 1.4835664749145507, "step": 8580 }, { "epoch": 0.731998295696634, "grad_norm": 2.5625, "learning_rate": 4.879004771642809e-05, "loss": 1.3516793251037598, "step": 8590 }, { "epoch": 0.7328504473796336, "grad_norm": 1.8828125, "learning_rate": 4.8846852987957285e-05, "loss": 1.4311988830566407, "step": 8600 }, { "epoch": 0.7337025990626331, "grad_norm": 2.34375, "learning_rate": 4.8903658259486486e-05, "loss": 1.4285818099975587, "step": 8610 }, { "epoch": 0.7345547507456327, "grad_norm": 2.078125, "learning_rate": 4.896046353101568e-05, "loss": 1.4534584045410157, "step": 8620 }, { "epoch": 0.7354069024286323, "grad_norm": 2.9375, "learning_rate": 4.901726880254488e-05, "loss": 1.631300926208496, "step": 8630 }, { "epoch": 0.7362590541116318, "grad_norm": 1.6953125, "learning_rate": 4.9074074074074075e-05, "loss": 1.503362274169922, "step": 8640 }, { "epoch": 0.7371112057946314, "grad_norm": 2.890625, "learning_rate": 4.9130879345603276e-05, "loss": 1.5459110260009765, "step": 8650 }, { "epoch": 0.737963357477631, "grad_norm": 2.578125, "learning_rate": 4.918768461713247e-05, "loss": 1.4883611679077149, "step": 8660 }, { "epoch": 0.7388155091606305, "grad_norm": 1.921875, "learning_rate": 4.924448988866167e-05, "loss": 1.502527618408203, "step": 8670 }, { "epoch": 0.7396676608436301, "grad_norm": 2.1875, "learning_rate": 4.9301295160190866e-05, "loss": 1.5200899124145508, "step": 8680 }, { "epoch": 0.7405198125266298, "grad_norm": 3.265625, "learning_rate": 4.935810043172007e-05, "loss": 1.4884632110595704, "step": 8690 }, { "epoch": 0.7413719642096294, "grad_norm": 2.765625, "learning_rate": 4.941490570324927e-05, "loss": 1.5431941986083983, "step": 8700 }, { "epoch": 0.7422241158926289, "grad_norm": 3.140625, "learning_rate": 4.947171097477846e-05, "loss": 1.52005033493042, "step": 8710 }, { "epoch": 0.7430762675756285, "grad_norm": 3.75, "learning_rate": 4.952851624630766e-05, "loss": 1.3846478462219238, "step": 8720 }, { "epoch": 0.7439284192586281, "grad_norm": 3.375, "learning_rate": 4.958532151783686e-05, "loss": 1.4402715682983398, "step": 8730 }, { "epoch": 0.7447805709416276, "grad_norm": 1.640625, "learning_rate": 4.964212678936606e-05, "loss": 1.519866943359375, "step": 8740 }, { "epoch": 0.7456327226246272, "grad_norm": 2.59375, "learning_rate": 4.969893206089525e-05, "loss": 1.5270241737365722, "step": 8750 }, { "epoch": 0.7464848743076268, "grad_norm": 3.40625, "learning_rate": 4.975573733242445e-05, "loss": 1.3718441009521485, "step": 8760 }, { "epoch": 0.7473370259906263, "grad_norm": 3.125, "learning_rate": 4.981254260395365e-05, "loss": 1.4556509971618652, "step": 8770 }, { "epoch": 0.7481891776736259, "grad_norm": 2.8125, "learning_rate": 4.986934787548285e-05, "loss": 1.3825878143310546, "step": 8780 }, { "epoch": 0.7490413293566255, "grad_norm": 2.5, "learning_rate": 4.992615314701204e-05, "loss": 1.587350559234619, "step": 8790 }, { "epoch": 0.749893481039625, "grad_norm": 2.984375, "learning_rate": 4.9982958418541243e-05, "loss": 1.489273452758789, "step": 8800 }, { "epoch": 0.7507456327226246, "grad_norm": 2.75, "learning_rate": 4.999999992535195e-05, "loss": 1.4851832389831543, "step": 8810 }, { "epoch": 0.7515977844056242, "grad_norm": 1.6015625, "learning_rate": 4.9999999559728865e-05, "loss": 1.594472885131836, "step": 8820 }, { "epoch": 0.7524499360886238, "grad_norm": 1.78125, "learning_rate": 4.9999998889419875e-05, "loss": 1.5612441062927247, "step": 8830 }, { "epoch": 0.7533020877716233, "grad_norm": 2.625, "learning_rate": 4.999999791442498e-05, "loss": 1.389207935333252, "step": 8840 }, { "epoch": 0.7541542394546229, "grad_norm": 3.140625, "learning_rate": 4.9999996634744215e-05, "loss": 1.4144222259521484, "step": 8850 }, { "epoch": 0.7550063911376225, "grad_norm": 2.046875, "learning_rate": 4.999999505037757e-05, "loss": 1.5845953941345214, "step": 8860 }, { "epoch": 0.755858542820622, "grad_norm": 2.25, "learning_rate": 4.9999993161325085e-05, "loss": 1.6007484436035155, "step": 8870 }, { "epoch": 0.7567106945036216, "grad_norm": 1.8359375, "learning_rate": 4.999999096758677e-05, "loss": 1.515967559814453, "step": 8880 }, { "epoch": 0.7575628461866212, "grad_norm": 2.34375, "learning_rate": 4.999998846916266e-05, "loss": 1.4496795654296875, "step": 8890 }, { "epoch": 0.7584149978696207, "grad_norm": 2.671875, "learning_rate": 4.9999985666052786e-05, "loss": 1.5243459701538087, "step": 8900 }, { "epoch": 0.7592671495526203, "grad_norm": 2.265625, "learning_rate": 4.9999982558257176e-05, "loss": 1.6262691497802735, "step": 8910 }, { "epoch": 0.76011930123562, "grad_norm": 1.8828125, "learning_rate": 4.9999979145775866e-05, "loss": 1.491138744354248, "step": 8920 }, { "epoch": 0.7609714529186196, "grad_norm": 2.265625, "learning_rate": 4.9999975428608904e-05, "loss": 1.5781781196594238, "step": 8930 }, { "epoch": 0.7618236046016191, "grad_norm": 2.125, "learning_rate": 4.999997140675633e-05, "loss": 1.385276699066162, "step": 8940 }, { "epoch": 0.7626757562846187, "grad_norm": 1.890625, "learning_rate": 4.999996708021819e-05, "loss": 1.5406100273132324, "step": 8950 }, { "epoch": 0.7635279079676183, "grad_norm": 3.6875, "learning_rate": 4.999996244899455e-05, "loss": 1.5451981544494628, "step": 8960 }, { "epoch": 0.7643800596506178, "grad_norm": 2.21875, "learning_rate": 4.999995751308546e-05, "loss": 1.4777344703674316, "step": 8970 }, { "epoch": 0.7652322113336174, "grad_norm": 3.65625, "learning_rate": 4.999995227249098e-05, "loss": 1.3854446411132812, "step": 8980 }, { "epoch": 0.766084363016617, "grad_norm": 4.1875, "learning_rate": 4.999994672721118e-05, "loss": 1.4362913131713868, "step": 8990 }, { "epoch": 0.7669365146996165, "grad_norm": 2.515625, "learning_rate": 4.999994087724611e-05, "loss": 1.4521679878234863, "step": 9000 }, { "epoch": 0.7677886663826161, "grad_norm": 2.609375, "learning_rate": 4.9999934722595854e-05, "loss": 1.4544264793395996, "step": 9010 }, { "epoch": 0.7686408180656157, "grad_norm": 2.046875, "learning_rate": 4.999992826326048e-05, "loss": 1.4687441825866698, "step": 9020 }, { "epoch": 0.7694929697486153, "grad_norm": 1.7578125, "learning_rate": 4.9999921499240074e-05, "loss": 1.5336106300354004, "step": 9030 }, { "epoch": 0.7703451214316148, "grad_norm": 2.59375, "learning_rate": 4.9999914430534725e-05, "loss": 1.407618522644043, "step": 9040 }, { "epoch": 0.7711972731146144, "grad_norm": 2.234375, "learning_rate": 4.99999070571445e-05, "loss": 1.596808624267578, "step": 9050 }, { "epoch": 0.772049424797614, "grad_norm": 2.46875, "learning_rate": 4.99998993790695e-05, "loss": 1.5006431579589843, "step": 9060 }, { "epoch": 0.7729015764806135, "grad_norm": 3.546875, "learning_rate": 4.999989139630983e-05, "loss": 1.5537544250488282, "step": 9070 }, { "epoch": 0.7737537281636131, "grad_norm": 1.765625, "learning_rate": 4.999988310886556e-05, "loss": 1.3441713333129883, "step": 9080 }, { "epoch": 0.7746058798466127, "grad_norm": 2.3125, "learning_rate": 4.999987451673681e-05, "loss": 1.4201578140258788, "step": 9090 }, { "epoch": 0.7754580315296122, "grad_norm": 2.796875, "learning_rate": 4.999986561992368e-05, "loss": 1.4762598037719727, "step": 9100 }, { "epoch": 0.7763101832126118, "grad_norm": 2.453125, "learning_rate": 4.999985641842628e-05, "loss": 1.5107603073120117, "step": 9110 }, { "epoch": 0.7771623348956114, "grad_norm": 1.875, "learning_rate": 4.999984691224473e-05, "loss": 1.5191003799438476, "step": 9120 }, { "epoch": 0.7780144865786109, "grad_norm": 1.8203125, "learning_rate": 4.999983710137913e-05, "loss": 1.5581310272216797, "step": 9130 }, { "epoch": 0.7788666382616106, "grad_norm": 5.875, "learning_rate": 4.99998269858296e-05, "loss": 1.3641082763671875, "step": 9140 }, { "epoch": 0.7797187899446102, "grad_norm": 2.546875, "learning_rate": 4.999981656559628e-05, "loss": 1.3604515075683594, "step": 9150 }, { "epoch": 0.7805709416276098, "grad_norm": 2.34375, "learning_rate": 4.999980584067928e-05, "loss": 1.6207145690917968, "step": 9160 }, { "epoch": 0.7814230933106093, "grad_norm": 1.796875, "learning_rate": 4.999979481107875e-05, "loss": 1.5025869369506837, "step": 9170 }, { "epoch": 0.7822752449936089, "grad_norm": 4.5, "learning_rate": 4.99997834767948e-05, "loss": 1.463810348510742, "step": 9180 }, { "epoch": 0.7831273966766085, "grad_norm": 3.359375, "learning_rate": 4.999977183782758e-05, "loss": 1.3960635185241699, "step": 9190 }, { "epoch": 0.783979548359608, "grad_norm": 2.53125, "learning_rate": 4.999975989417723e-05, "loss": 1.4235050201416015, "step": 9200 }, { "epoch": 0.7848317000426076, "grad_norm": 2.3125, "learning_rate": 4.999974764584391e-05, "loss": 1.4450714111328125, "step": 9210 }, { "epoch": 0.7856838517256072, "grad_norm": 2.203125, "learning_rate": 4.999973509282775e-05, "loss": 1.5053540229797364, "step": 9220 }, { "epoch": 0.7865360034086067, "grad_norm": 1.828125, "learning_rate": 4.999972223512891e-05, "loss": 1.6466957092285157, "step": 9230 }, { "epoch": 0.7873881550916063, "grad_norm": 2.09375, "learning_rate": 4.9999709072747546e-05, "loss": 1.5394834518432616, "step": 9240 }, { "epoch": 0.7882403067746059, "grad_norm": 2.453125, "learning_rate": 4.999969560568382e-05, "loss": 1.4335132598876954, "step": 9250 }, { "epoch": 0.7890924584576055, "grad_norm": 2.8125, "learning_rate": 4.9999681833937896e-05, "loss": 1.345384120941162, "step": 9260 }, { "epoch": 0.789944610140605, "grad_norm": 2.703125, "learning_rate": 4.999966775750994e-05, "loss": 1.476162815093994, "step": 9270 }, { "epoch": 0.7907967618236046, "grad_norm": 2.75, "learning_rate": 4.9999653376400123e-05, "loss": 1.4437435150146485, "step": 9280 }, { "epoch": 0.7916489135066042, "grad_norm": 2.96875, "learning_rate": 4.999963869060863e-05, "loss": 1.4113155364990235, "step": 9290 }, { "epoch": 0.7925010651896037, "grad_norm": 2.46875, "learning_rate": 4.9999623700135626e-05, "loss": 1.5722126960754395, "step": 9300 }, { "epoch": 0.7933532168726033, "grad_norm": 2.71875, "learning_rate": 4.999960840498129e-05, "loss": 1.4833497047424316, "step": 9310 }, { "epoch": 0.7942053685556029, "grad_norm": 2.171875, "learning_rate": 4.9999592805145826e-05, "loss": 1.4526105880737306, "step": 9320 }, { "epoch": 0.7950575202386024, "grad_norm": 3.84375, "learning_rate": 4.999957690062942e-05, "loss": 1.5295312881469727, "step": 9330 }, { "epoch": 0.795909671921602, "grad_norm": 1.765625, "learning_rate": 4.9999560691432255e-05, "loss": 1.4862352371215821, "step": 9340 }, { "epoch": 0.7967618236046016, "grad_norm": 2.671875, "learning_rate": 4.999954417755454e-05, "loss": 1.4234679222106934, "step": 9350 }, { "epoch": 0.7976139752876011, "grad_norm": 2.375, "learning_rate": 4.999952735899647e-05, "loss": 1.6995630264282227, "step": 9360 }, { "epoch": 0.7984661269706008, "grad_norm": 1.8125, "learning_rate": 4.999951023575825e-05, "loss": 1.5672082901000977, "step": 9370 }, { "epoch": 0.7993182786536004, "grad_norm": 1.734375, "learning_rate": 4.9999492807840095e-05, "loss": 1.3438551902770997, "step": 9380 }, { "epoch": 0.8001704303366, "grad_norm": 3.140625, "learning_rate": 4.999947507524221e-05, "loss": 1.3921369552612304, "step": 9390 }, { "epoch": 0.8010225820195995, "grad_norm": 2.140625, "learning_rate": 4.999945703796481e-05, "loss": 1.2908531188964845, "step": 9400 }, { "epoch": 0.8018747337025991, "grad_norm": 2.8125, "learning_rate": 4.999943869600813e-05, "loss": 1.4319061279296874, "step": 9410 }, { "epoch": 0.8027268853855987, "grad_norm": 2.390625, "learning_rate": 4.999942004937237e-05, "loss": 1.512883186340332, "step": 9420 }, { "epoch": 0.8035790370685982, "grad_norm": 1.96875, "learning_rate": 4.9999401098057776e-05, "loss": 1.5443568229675293, "step": 9430 }, { "epoch": 0.8044311887515978, "grad_norm": 2.640625, "learning_rate": 4.999938184206457e-05, "loss": 1.5087446212768554, "step": 9440 }, { "epoch": 0.8052833404345974, "grad_norm": 2.25, "learning_rate": 4.999936228139299e-05, "loss": 1.3622276306152343, "step": 9450 }, { "epoch": 0.8061354921175969, "grad_norm": 2.171875, "learning_rate": 4.9999342416043274e-05, "loss": 1.2773728370666504, "step": 9460 }, { "epoch": 0.8069876438005965, "grad_norm": 2.90625, "learning_rate": 4.9999322246015665e-05, "loss": 1.6065923690795898, "step": 9470 }, { "epoch": 0.8078397954835961, "grad_norm": 2.828125, "learning_rate": 4.99993017713104e-05, "loss": 1.5929500579833984, "step": 9480 }, { "epoch": 0.8086919471665956, "grad_norm": 2.421875, "learning_rate": 4.999928099192775e-05, "loss": 1.4375011444091796, "step": 9490 }, { "epoch": 0.8095440988495952, "grad_norm": 1.7421875, "learning_rate": 4.9999259907867947e-05, "loss": 1.5830162048339844, "step": 9500 }, { "epoch": 0.8103962505325948, "grad_norm": 1.6875, "learning_rate": 4.999923851913125e-05, "loss": 1.4180668830871581, "step": 9510 }, { "epoch": 0.8112484022155944, "grad_norm": 1.859375, "learning_rate": 4.999921682571793e-05, "loss": 1.6726062774658204, "step": 9520 }, { "epoch": 0.8121005538985939, "grad_norm": 2.1875, "learning_rate": 4.999919482762825e-05, "loss": 1.4718358039855957, "step": 9530 }, { "epoch": 0.8129527055815935, "grad_norm": 2.5, "learning_rate": 4.999917252486247e-05, "loss": 1.5044471740722656, "step": 9540 }, { "epoch": 0.8138048572645931, "grad_norm": 3.0625, "learning_rate": 4.9999149917420865e-05, "loss": 1.537007999420166, "step": 9550 }, { "epoch": 0.8146570089475926, "grad_norm": 2.1875, "learning_rate": 4.999912700530371e-05, "loss": 1.5938413619995118, "step": 9560 }, { "epoch": 0.8155091606305922, "grad_norm": 2.640625, "learning_rate": 4.999910378851129e-05, "loss": 1.3817507743835449, "step": 9570 }, { "epoch": 0.8163613123135918, "grad_norm": 1.765625, "learning_rate": 4.999908026704388e-05, "loss": 1.6268552780151366, "step": 9580 }, { "epoch": 0.8172134639965913, "grad_norm": 2.8125, "learning_rate": 4.999905644090177e-05, "loss": 1.4298294067382813, "step": 9590 }, { "epoch": 0.818065615679591, "grad_norm": 3.15625, "learning_rate": 4.999903231008526e-05, "loss": 1.5487079620361328, "step": 9600 }, { "epoch": 0.8189177673625906, "grad_norm": 3.640625, "learning_rate": 4.9999007874594626e-05, "loss": 1.505967903137207, "step": 9610 }, { "epoch": 0.8197699190455902, "grad_norm": 3.796875, "learning_rate": 4.999898313443018e-05, "loss": 1.3759042739868164, "step": 9620 }, { "epoch": 0.8206220707285897, "grad_norm": 2.921875, "learning_rate": 4.999895808959222e-05, "loss": 1.5467390060424804, "step": 9630 }, { "epoch": 0.8214742224115893, "grad_norm": 1.75, "learning_rate": 4.999893274008104e-05, "loss": 1.6833316802978515, "step": 9640 }, { "epoch": 0.8223263740945889, "grad_norm": 1.796875, "learning_rate": 4.999890708589696e-05, "loss": 1.3290323257446288, "step": 9650 }, { "epoch": 0.8231785257775884, "grad_norm": 1.984375, "learning_rate": 4.99988811270403e-05, "loss": 1.5140793800354004, "step": 9660 }, { "epoch": 0.824030677460588, "grad_norm": 2.78125, "learning_rate": 4.9998854863511364e-05, "loss": 1.489596176147461, "step": 9670 }, { "epoch": 0.8248828291435876, "grad_norm": 2.5, "learning_rate": 4.999882829531047e-05, "loss": 1.3945554733276366, "step": 9680 }, { "epoch": 0.8257349808265871, "grad_norm": 1.8984375, "learning_rate": 4.999880142243796e-05, "loss": 1.465811824798584, "step": 9690 }, { "epoch": 0.8265871325095867, "grad_norm": 3.703125, "learning_rate": 4.9998774244894134e-05, "loss": 1.3870412826538085, "step": 9700 }, { "epoch": 0.8274392841925863, "grad_norm": 3.46875, "learning_rate": 4.999874676267935e-05, "loss": 1.4775898933410645, "step": 9710 }, { "epoch": 0.8282914358755858, "grad_norm": 3.171875, "learning_rate": 4.9998718975793924e-05, "loss": 1.438868522644043, "step": 9720 }, { "epoch": 0.8291435875585854, "grad_norm": 3.125, "learning_rate": 4.999869088423821e-05, "loss": 1.3746988296508789, "step": 9730 }, { "epoch": 0.829995739241585, "grad_norm": 2.9375, "learning_rate": 4.999866248801253e-05, "loss": 1.3836490631103515, "step": 9740 }, { "epoch": 0.8308478909245846, "grad_norm": 2.359375, "learning_rate": 4.999863378711725e-05, "loss": 1.6386432647705078, "step": 9750 }, { "epoch": 0.8317000426075841, "grad_norm": 2.203125, "learning_rate": 4.999860478155271e-05, "loss": 1.412586784362793, "step": 9760 }, { "epoch": 0.8325521942905837, "grad_norm": 2.59375, "learning_rate": 4.9998575471319263e-05, "loss": 1.401301670074463, "step": 9770 }, { "epoch": 0.8334043459735833, "grad_norm": 2.328125, "learning_rate": 4.9998545856417275e-05, "loss": 1.4127347946166993, "step": 9780 }, { "epoch": 0.8342564976565828, "grad_norm": 3.59375, "learning_rate": 4.9998515936847103e-05, "loss": 1.4943436622619628, "step": 9790 }, { "epoch": 0.8351086493395824, "grad_norm": 2.015625, "learning_rate": 4.99984857126091e-05, "loss": 1.6334787368774415, "step": 9800 }, { "epoch": 0.835960801022582, "grad_norm": 1.9296875, "learning_rate": 4.999845518370365e-05, "loss": 1.3504528999328613, "step": 9810 }, { "epoch": 0.8368129527055816, "grad_norm": 1.953125, "learning_rate": 4.999842435013111e-05, "loss": 1.5480576515197755, "step": 9820 }, { "epoch": 0.8376651043885812, "grad_norm": 2.09375, "learning_rate": 4.999839321189187e-05, "loss": 1.210035514831543, "step": 9830 }, { "epoch": 0.8385172560715808, "grad_norm": 2.03125, "learning_rate": 4.999836176898631e-05, "loss": 1.419936752319336, "step": 9840 }, { "epoch": 0.8393694077545804, "grad_norm": 2.5, "learning_rate": 4.9998330021414796e-05, "loss": 1.4662325859069825, "step": 9850 }, { "epoch": 0.8402215594375799, "grad_norm": 3.859375, "learning_rate": 4.999829796917773e-05, "loss": 1.3699374198913574, "step": 9860 }, { "epoch": 0.8410737111205795, "grad_norm": 2.0625, "learning_rate": 4.9998265612275506e-05, "loss": 1.6149311065673828, "step": 9870 }, { "epoch": 0.8419258628035791, "grad_norm": 2.234375, "learning_rate": 4.9998232950708505e-05, "loss": 1.4971139907836915, "step": 9880 }, { "epoch": 0.8427780144865786, "grad_norm": 2.09375, "learning_rate": 4.999819998447714e-05, "loss": 1.3725492477416992, "step": 9890 }, { "epoch": 0.8436301661695782, "grad_norm": 1.890625, "learning_rate": 4.999816671358179e-05, "loss": 1.5611383438110351, "step": 9900 }, { "epoch": 0.8444823178525778, "grad_norm": 2.25, "learning_rate": 4.999813313802288e-05, "loss": 1.5270634651184083, "step": 9910 }, { "epoch": 0.8453344695355773, "grad_norm": 2.296875, "learning_rate": 4.9998099257800816e-05, "loss": 1.422715663909912, "step": 9920 }, { "epoch": 0.8461866212185769, "grad_norm": 2.078125, "learning_rate": 4.999806507291601e-05, "loss": 1.493323516845703, "step": 9930 }, { "epoch": 0.8470387729015765, "grad_norm": 3.375, "learning_rate": 4.999803058336887e-05, "loss": 1.3951030731201173, "step": 9940 }, { "epoch": 0.847890924584576, "grad_norm": 3.421875, "learning_rate": 4.9997995789159825e-05, "loss": 1.4213690757751465, "step": 9950 }, { "epoch": 0.8487430762675756, "grad_norm": 2.4375, "learning_rate": 4.9997960690289306e-05, "loss": 1.4280331611633301, "step": 9960 }, { "epoch": 0.8495952279505752, "grad_norm": 2.390625, "learning_rate": 4.999792528675772e-05, "loss": 1.4977224349975586, "step": 9970 }, { "epoch": 0.8504473796335748, "grad_norm": 2.84375, "learning_rate": 4.999788957856552e-05, "loss": 1.4229222297668458, "step": 9980 }, { "epoch": 0.8512995313165743, "grad_norm": 2.125, "learning_rate": 4.999785356571313e-05, "loss": 1.5378442764282227, "step": 9990 }, { "epoch": 0.8521516829995739, "grad_norm": 1.625, "learning_rate": 4.9997817248200987e-05, "loss": 1.2707788467407226, "step": 10000 }, { "epoch": 0.8530038346825735, "grad_norm": 1.921875, "learning_rate": 4.9997780626029545e-05, "loss": 1.5098121643066407, "step": 10010 }, { "epoch": 0.853855986365573, "grad_norm": 2.15625, "learning_rate": 4.999774369919923e-05, "loss": 1.5413410186767578, "step": 10020 }, { "epoch": 0.8547081380485726, "grad_norm": 1.9140625, "learning_rate": 4.999770646771051e-05, "loss": 1.4034642219543456, "step": 10030 }, { "epoch": 0.8555602897315722, "grad_norm": 3.046875, "learning_rate": 4.9997668931563844e-05, "loss": 1.3119797706604004, "step": 10040 }, { "epoch": 0.8564124414145718, "grad_norm": 1.8359375, "learning_rate": 4.999763109075967e-05, "loss": 1.5689754486083984, "step": 10050 }, { "epoch": 0.8572645930975714, "grad_norm": 2.5, "learning_rate": 4.999759294529847e-05, "loss": 1.3251471519470215, "step": 10060 }, { "epoch": 0.858116744780571, "grad_norm": 2.53125, "learning_rate": 4.9997554495180683e-05, "loss": 1.4365181922912598, "step": 10070 }, { "epoch": 0.8589688964635706, "grad_norm": 1.7109375, "learning_rate": 4.99975157404068e-05, "loss": 1.5408679962158203, "step": 10080 }, { "epoch": 0.8598210481465701, "grad_norm": 1.9453125, "learning_rate": 4.999747668097727e-05, "loss": 1.461120891571045, "step": 10090 }, { "epoch": 0.8606731998295697, "grad_norm": 2.09375, "learning_rate": 4.99974373168926e-05, "loss": 1.27354097366333, "step": 10100 }, { "epoch": 0.8615253515125693, "grad_norm": 3.125, "learning_rate": 4.999739764815325e-05, "loss": 1.3806052207946777, "step": 10110 }, { "epoch": 0.8623775031955688, "grad_norm": 1.7890625, "learning_rate": 4.999735767475971e-05, "loss": 1.4786602020263673, "step": 10120 }, { "epoch": 0.8632296548785684, "grad_norm": 2.578125, "learning_rate": 4.9997317396712465e-05, "loss": 1.4112396240234375, "step": 10130 }, { "epoch": 0.864081806561568, "grad_norm": 1.6640625, "learning_rate": 4.9997276814011996e-05, "loss": 1.4885991096496582, "step": 10140 }, { "epoch": 0.8649339582445675, "grad_norm": 3.0, "learning_rate": 4.9997235926658813e-05, "loss": 1.4496390342712402, "step": 10150 }, { "epoch": 0.8657861099275671, "grad_norm": 2.46875, "learning_rate": 4.9997194734653407e-05, "loss": 1.4190135955810548, "step": 10160 }, { "epoch": 0.8666382616105667, "grad_norm": 2.75, "learning_rate": 4.999715323799628e-05, "loss": 1.2690008163452149, "step": 10170 }, { "epoch": 0.8674904132935662, "grad_norm": 2.265625, "learning_rate": 4.999711143668795e-05, "loss": 1.556633186340332, "step": 10180 }, { "epoch": 0.8683425649765658, "grad_norm": 2.09375, "learning_rate": 4.9997069330728905e-05, "loss": 1.376850700378418, "step": 10190 }, { "epoch": 0.8691947166595654, "grad_norm": 2.578125, "learning_rate": 4.999702692011967e-05, "loss": 1.5847639083862304, "step": 10200 }, { "epoch": 0.870046868342565, "grad_norm": 2.484375, "learning_rate": 4.9996984204860765e-05, "loss": 1.5807686805725099, "step": 10210 }, { "epoch": 0.8708990200255645, "grad_norm": 2.453125, "learning_rate": 4.99969411849527e-05, "loss": 1.3291110038757323, "step": 10220 }, { "epoch": 0.8717511717085641, "grad_norm": 2.65625, "learning_rate": 4.999689786039601e-05, "loss": 1.5253198623657227, "step": 10230 }, { "epoch": 0.8726033233915637, "grad_norm": 1.890625, "learning_rate": 4.9996854231191216e-05, "loss": 1.43783540725708, "step": 10240 }, { "epoch": 0.8734554750745632, "grad_norm": 1.7578125, "learning_rate": 4.999681029733885e-05, "loss": 1.4768871307373046, "step": 10250 }, { "epoch": 0.8743076267575628, "grad_norm": 1.7421875, "learning_rate": 4.999676605883946e-05, "loss": 1.6311176300048829, "step": 10260 }, { "epoch": 0.8751597784405625, "grad_norm": 3.25, "learning_rate": 4.9996721515693566e-05, "loss": 1.5207758903503419, "step": 10270 }, { "epoch": 0.876011930123562, "grad_norm": 1.546875, "learning_rate": 4.999667666790172e-05, "loss": 1.498174285888672, "step": 10280 }, { "epoch": 0.8768640818065616, "grad_norm": 4.25, "learning_rate": 4.999663151546447e-05, "loss": 1.380601978302002, "step": 10290 }, { "epoch": 0.8777162334895612, "grad_norm": 1.96875, "learning_rate": 4.9996586058382366e-05, "loss": 1.4644041061401367, "step": 10300 }, { "epoch": 0.8785683851725608, "grad_norm": 2.828125, "learning_rate": 4.999654029665596e-05, "loss": 1.529038906097412, "step": 10310 }, { "epoch": 0.8794205368555603, "grad_norm": 2.0, "learning_rate": 4.999649423028581e-05, "loss": 1.4439156532287598, "step": 10320 }, { "epoch": 0.8802726885385599, "grad_norm": 1.3984375, "learning_rate": 4.9996447859272475e-05, "loss": 1.3733504295349122, "step": 10330 }, { "epoch": 0.8811248402215595, "grad_norm": 2.21875, "learning_rate": 4.999640118361652e-05, "loss": 1.5365401268005372, "step": 10340 }, { "epoch": 0.881976991904559, "grad_norm": 2.625, "learning_rate": 4.999635420331852e-05, "loss": 1.4319146156311036, "step": 10350 }, { "epoch": 0.8828291435875586, "grad_norm": 2.5625, "learning_rate": 4.999630691837905e-05, "loss": 1.4634220123291015, "step": 10360 }, { "epoch": 0.8836812952705582, "grad_norm": 1.8046875, "learning_rate": 4.999625932879868e-05, "loss": 1.4397922515869142, "step": 10370 }, { "epoch": 0.8845334469535577, "grad_norm": 2.671875, "learning_rate": 4.9996211434577976e-05, "loss": 1.4071078300476074, "step": 10380 }, { "epoch": 0.8853855986365573, "grad_norm": 1.5859375, "learning_rate": 4.999616323571755e-05, "loss": 1.4970549583435058, "step": 10390 }, { "epoch": 0.8862377503195569, "grad_norm": 2.703125, "learning_rate": 4.9996114732217976e-05, "loss": 1.4055150985717773, "step": 10400 }, { "epoch": 0.8870899020025564, "grad_norm": 4.15625, "learning_rate": 4.9996065924079835e-05, "loss": 1.28329439163208, "step": 10410 }, { "epoch": 0.887942053685556, "grad_norm": 1.671875, "learning_rate": 4.9996016811303747e-05, "loss": 1.5438198089599608, "step": 10420 }, { "epoch": 0.8887942053685556, "grad_norm": 2.046875, "learning_rate": 4.999596739389029e-05, "loss": 1.441487979888916, "step": 10430 }, { "epoch": 0.8896463570515551, "grad_norm": 2.515625, "learning_rate": 4.9995917671840066e-05, "loss": 1.3818248748779296, "step": 10440 }, { "epoch": 0.8904985087345547, "grad_norm": 2.671875, "learning_rate": 4.999586764515368e-05, "loss": 1.5285680770874024, "step": 10450 }, { "epoch": 0.8913506604175543, "grad_norm": 2.609375, "learning_rate": 4.999581731383176e-05, "loss": 1.3479193687438964, "step": 10460 }, { "epoch": 0.8922028121005539, "grad_norm": 2.234375, "learning_rate": 4.9995766677874905e-05, "loss": 1.4904212951660156, "step": 10470 }, { "epoch": 0.8930549637835534, "grad_norm": 2.453125, "learning_rate": 4.999571573728373e-05, "loss": 1.387483787536621, "step": 10480 }, { "epoch": 0.893907115466553, "grad_norm": 1.59375, "learning_rate": 4.999566449205887e-05, "loss": 1.46676607131958, "step": 10490 }, { "epoch": 0.8947592671495527, "grad_norm": 2.40625, "learning_rate": 4.999561294220093e-05, "loss": 1.479750156402588, "step": 10500 }, { "epoch": 0.8956114188325522, "grad_norm": 1.71875, "learning_rate": 4.999556108771056e-05, "loss": 1.54501314163208, "step": 10510 }, { "epoch": 0.8964635705155518, "grad_norm": 1.7265625, "learning_rate": 4.999550892858837e-05, "loss": 1.4615898132324219, "step": 10520 }, { "epoch": 0.8973157221985514, "grad_norm": 2.46875, "learning_rate": 4.999545646483501e-05, "loss": 1.4192378997802735, "step": 10530 }, { "epoch": 0.898167873881551, "grad_norm": 2.0625, "learning_rate": 4.999540369645112e-05, "loss": 1.4304805755615235, "step": 10540 }, { "epoch": 0.8990200255645505, "grad_norm": 2.8125, "learning_rate": 4.999535062343733e-05, "loss": 1.3625181198120118, "step": 10550 }, { "epoch": 0.8998721772475501, "grad_norm": 3.59375, "learning_rate": 4.9995297245794295e-05, "loss": 1.45721378326416, "step": 10560 }, { "epoch": 0.9007243289305497, "grad_norm": 3.515625, "learning_rate": 4.999524356352267e-05, "loss": 1.4093025207519532, "step": 10570 }, { "epoch": 0.9015764806135492, "grad_norm": 2.171875, "learning_rate": 4.99951895766231e-05, "loss": 1.3741091728210448, "step": 10580 }, { "epoch": 0.9024286322965488, "grad_norm": 2.296875, "learning_rate": 4.999513528509626e-05, "loss": 1.5936023712158203, "step": 10590 }, { "epoch": 0.9032807839795484, "grad_norm": 2.015625, "learning_rate": 4.999508068894279e-05, "loss": 1.467106533050537, "step": 10600 }, { "epoch": 0.9041329356625479, "grad_norm": 2.671875, "learning_rate": 4.999502578816337e-05, "loss": 1.4098138809204102, "step": 10610 }, { "epoch": 0.9049850873455475, "grad_norm": 2.5625, "learning_rate": 4.999497058275866e-05, "loss": 1.460271167755127, "step": 10620 }, { "epoch": 0.9058372390285471, "grad_norm": 2.046875, "learning_rate": 4.9994915072729345e-05, "loss": 1.4313092231750488, "step": 10630 }, { "epoch": 0.9066893907115466, "grad_norm": 4.3125, "learning_rate": 4.9994859258076085e-05, "loss": 1.4816787719726563, "step": 10640 }, { "epoch": 0.9075415423945462, "grad_norm": 2.125, "learning_rate": 4.9994803138799575e-05, "loss": 1.40411319732666, "step": 10650 }, { "epoch": 0.9083936940775458, "grad_norm": 2.65625, "learning_rate": 4.9994746714900487e-05, "loss": 1.4498278617858886, "step": 10660 }, { "epoch": 0.9092458457605453, "grad_norm": 2.140625, "learning_rate": 4.999468998637952e-05, "loss": 1.5708395004272462, "step": 10670 }, { "epoch": 0.9100979974435449, "grad_norm": 2.328125, "learning_rate": 4.999463295323735e-05, "loss": 1.5267498016357421, "step": 10680 }, { "epoch": 0.9109501491265445, "grad_norm": 2.203125, "learning_rate": 4.9994575615474694e-05, "loss": 1.4569427490234375, "step": 10690 }, { "epoch": 0.911802300809544, "grad_norm": 1.640625, "learning_rate": 4.999451797309222e-05, "loss": 1.3991607666015624, "step": 10700 }, { "epoch": 0.9126544524925436, "grad_norm": 2.515625, "learning_rate": 4.9994460026090676e-05, "loss": 1.4239755630493165, "step": 10710 }, { "epoch": 0.9135066041755433, "grad_norm": 2.390625, "learning_rate": 4.9994401774470724e-05, "loss": 1.484354019165039, "step": 10720 }, { "epoch": 0.9143587558585429, "grad_norm": 3.5625, "learning_rate": 4.99943432182331e-05, "loss": 1.530380630493164, "step": 10730 }, { "epoch": 0.9152109075415424, "grad_norm": 2.953125, "learning_rate": 4.9994284357378494e-05, "loss": 1.533383846282959, "step": 10740 }, { "epoch": 0.916063059224542, "grad_norm": 1.953125, "learning_rate": 4.999422519190765e-05, "loss": 1.36887845993042, "step": 10750 }, { "epoch": 0.9169152109075416, "grad_norm": 1.8203125, "learning_rate": 4.999416572182128e-05, "loss": 1.607633399963379, "step": 10760 }, { "epoch": 0.9177673625905411, "grad_norm": 2.5625, "learning_rate": 4.9994105947120105e-05, "loss": 1.4593109130859374, "step": 10770 }, { "epoch": 0.9186195142735407, "grad_norm": 2.75, "learning_rate": 4.999404586780485e-05, "loss": 1.4507411956787108, "step": 10780 }, { "epoch": 0.9194716659565403, "grad_norm": 3.0, "learning_rate": 4.999398548387625e-05, "loss": 1.4096275329589845, "step": 10790 }, { "epoch": 0.9203238176395399, "grad_norm": 2.0625, "learning_rate": 4.999392479533505e-05, "loss": 1.3741225242614745, "step": 10800 }, { "epoch": 0.9211759693225394, "grad_norm": 2.234375, "learning_rate": 4.9993863802181985e-05, "loss": 1.4337425231933594, "step": 10810 }, { "epoch": 0.922028121005539, "grad_norm": 2.125, "learning_rate": 4.9993802504417784e-05, "loss": 1.5259113311767578, "step": 10820 }, { "epoch": 0.9228802726885386, "grad_norm": 2.640625, "learning_rate": 4.9993740902043216e-05, "loss": 1.4131948471069335, "step": 10830 }, { "epoch": 0.9237324243715381, "grad_norm": 2.21875, "learning_rate": 4.999367899505902e-05, "loss": 1.4394285202026367, "step": 10840 }, { "epoch": 0.9245845760545377, "grad_norm": 2.5625, "learning_rate": 4.999361678346595e-05, "loss": 1.5096190452575684, "step": 10850 }, { "epoch": 0.9254367277375373, "grad_norm": 2.109375, "learning_rate": 4.999355426726476e-05, "loss": 1.4427218437194824, "step": 10860 }, { "epoch": 0.9262888794205368, "grad_norm": 1.796875, "learning_rate": 4.999349144645623e-05, "loss": 1.5261720657348632, "step": 10870 }, { "epoch": 0.9271410311035364, "grad_norm": 2.21875, "learning_rate": 4.999342832104111e-05, "loss": 1.5477873802185058, "step": 10880 }, { "epoch": 0.927993182786536, "grad_norm": 2.1875, "learning_rate": 4.999336489102017e-05, "loss": 1.6230247497558594, "step": 10890 }, { "epoch": 0.9288453344695355, "grad_norm": 2.0625, "learning_rate": 4.999330115639419e-05, "loss": 1.5138261795043946, "step": 10900 }, { "epoch": 0.9296974861525351, "grad_norm": 1.953125, "learning_rate": 4.9993237117163944e-05, "loss": 1.3247851371765136, "step": 10910 }, { "epoch": 0.9305496378355347, "grad_norm": 2.375, "learning_rate": 4.999317277333021e-05, "loss": 1.553772258758545, "step": 10920 }, { "epoch": 0.9314017895185343, "grad_norm": 2.21875, "learning_rate": 4.999310812489377e-05, "loss": 1.473192024230957, "step": 10930 }, { "epoch": 0.9322539412015338, "grad_norm": 1.75, "learning_rate": 4.999304317185543e-05, "loss": 1.476029396057129, "step": 10940 }, { "epoch": 0.9331060928845335, "grad_norm": 2.1875, "learning_rate": 4.999297791421595e-05, "loss": 1.5537569999694825, "step": 10950 }, { "epoch": 0.9339582445675331, "grad_norm": 1.90625, "learning_rate": 4.999291235197615e-05, "loss": 1.4050898551940918, "step": 10960 }, { "epoch": 0.9348103962505326, "grad_norm": 2.625, "learning_rate": 4.999284648513682e-05, "loss": 1.362996482849121, "step": 10970 }, { "epoch": 0.9356625479335322, "grad_norm": 1.7890625, "learning_rate": 4.999278031369876e-05, "loss": 1.4650721549987793, "step": 10980 }, { "epoch": 0.9365146996165318, "grad_norm": 2.390625, "learning_rate": 4.999271383766279e-05, "loss": 1.4216127395629883, "step": 10990 }, { "epoch": 0.9373668512995313, "grad_norm": 1.65625, "learning_rate": 4.999264705702971e-05, "loss": 1.5031121253967286, "step": 11000 }, { "epoch": 0.9382190029825309, "grad_norm": 3.0, "learning_rate": 4.999257997180033e-05, "loss": 1.4647295951843262, "step": 11010 }, { "epoch": 0.9390711546655305, "grad_norm": 2.328125, "learning_rate": 4.999251258197547e-05, "loss": 1.5924931526184083, "step": 11020 }, { "epoch": 0.93992330634853, "grad_norm": 2.453125, "learning_rate": 4.999244488755596e-05, "loss": 1.2946245193481445, "step": 11030 }, { "epoch": 0.9407754580315296, "grad_norm": 2.3125, "learning_rate": 4.999237688854262e-05, "loss": 1.3412183761596679, "step": 11040 }, { "epoch": 0.9416276097145292, "grad_norm": 3.75, "learning_rate": 4.9992308584936275e-05, "loss": 1.3833059310913085, "step": 11050 }, { "epoch": 0.9424797613975288, "grad_norm": 1.7109375, "learning_rate": 4.999223997673775e-05, "loss": 1.5578954696655274, "step": 11060 }, { "epoch": 0.9433319130805283, "grad_norm": 1.984375, "learning_rate": 4.9992171063947904e-05, "loss": 1.4759401321411132, "step": 11070 }, { "epoch": 0.9441840647635279, "grad_norm": 1.8984375, "learning_rate": 4.999210184656756e-05, "loss": 1.5102104187011718, "step": 11080 }, { "epoch": 0.9450362164465275, "grad_norm": 3.015625, "learning_rate": 4.999203232459757e-05, "loss": 1.3244213104248046, "step": 11090 }, { "epoch": 0.945888368129527, "grad_norm": 1.7890625, "learning_rate": 4.999196249803877e-05, "loss": 1.4628265380859375, "step": 11100 }, { "epoch": 0.9467405198125266, "grad_norm": 2.671875, "learning_rate": 4.9991892366892026e-05, "loss": 1.4253469467163087, "step": 11110 }, { "epoch": 0.9475926714955262, "grad_norm": 2.4375, "learning_rate": 4.999182193115818e-05, "loss": 1.4161304473876952, "step": 11120 }, { "epoch": 0.9484448231785257, "grad_norm": 1.9609375, "learning_rate": 4.999175119083809e-05, "loss": 1.3690985679626464, "step": 11130 }, { "epoch": 0.9492969748615253, "grad_norm": 2.40625, "learning_rate": 4.9991680145932626e-05, "loss": 1.4992414474487306, "step": 11140 }, { "epoch": 0.9501491265445249, "grad_norm": 2.125, "learning_rate": 4.999160879644266e-05, "loss": 1.484964942932129, "step": 11150 }, { "epoch": 0.9510012782275244, "grad_norm": 3.546875, "learning_rate": 4.9991537142369046e-05, "loss": 1.2942465782165526, "step": 11160 }, { "epoch": 0.951853429910524, "grad_norm": 2.28125, "learning_rate": 4.999146518371266e-05, "loss": 1.4547033309936523, "step": 11170 }, { "epoch": 0.9527055815935237, "grad_norm": 2.4375, "learning_rate": 4.999139292047439e-05, "loss": 1.4797821044921875, "step": 11180 }, { "epoch": 0.9535577332765233, "grad_norm": 2.53125, "learning_rate": 4.9991320352655105e-05, "loss": 1.431620979309082, "step": 11190 }, { "epoch": 0.9544098849595228, "grad_norm": 2.609375, "learning_rate": 4.99912474802557e-05, "loss": 1.4281153678894043, "step": 11200 }, { "epoch": 0.9552620366425224, "grad_norm": 2.03125, "learning_rate": 4.999117430327704e-05, "loss": 1.3255415916442872, "step": 11210 }, { "epoch": 0.956114188325522, "grad_norm": 1.921875, "learning_rate": 4.9991100821720055e-05, "loss": 1.5613473892211913, "step": 11220 }, { "epoch": 0.9569663400085215, "grad_norm": 2.015625, "learning_rate": 4.999102703558561e-05, "loss": 1.655480194091797, "step": 11230 }, { "epoch": 0.9578184916915211, "grad_norm": 2.1875, "learning_rate": 4.9990952944874616e-05, "loss": 1.5923831939697266, "step": 11240 }, { "epoch": 0.9586706433745207, "grad_norm": 1.6796875, "learning_rate": 4.999087854958797e-05, "loss": 1.4469969749450684, "step": 11250 }, { "epoch": 0.9595227950575203, "grad_norm": 3.796875, "learning_rate": 4.9990803849726584e-05, "loss": 1.4752691268920899, "step": 11260 }, { "epoch": 0.9603749467405198, "grad_norm": 2.890625, "learning_rate": 4.9990728845291365e-05, "loss": 1.3449110031127929, "step": 11270 }, { "epoch": 0.9612270984235194, "grad_norm": 3.328125, "learning_rate": 4.999065353628324e-05, "loss": 1.4752630233764648, "step": 11280 }, { "epoch": 0.962079250106519, "grad_norm": 3.3125, "learning_rate": 4.999057792270311e-05, "loss": 1.3270380973815918, "step": 11290 }, { "epoch": 0.9629314017895185, "grad_norm": 3.453125, "learning_rate": 4.9990502004551895e-05, "loss": 1.4324079513549806, "step": 11300 }, { "epoch": 0.9637835534725181, "grad_norm": 2.734375, "learning_rate": 4.999042578183053e-05, "loss": 1.4591243743896485, "step": 11310 }, { "epoch": 0.9646357051555177, "grad_norm": 3.640625, "learning_rate": 4.999034925453995e-05, "loss": 1.4659770011901856, "step": 11320 }, { "epoch": 0.9654878568385172, "grad_norm": 3.109375, "learning_rate": 4.9990272422681074e-05, "loss": 1.3392191886901856, "step": 11330 }, { "epoch": 0.9663400085215168, "grad_norm": 1.765625, "learning_rate": 4.999019528625484e-05, "loss": 1.5954577445983886, "step": 11340 }, { "epoch": 0.9671921602045164, "grad_norm": 1.9765625, "learning_rate": 4.9990117845262195e-05, "loss": 1.4787450790405274, "step": 11350 }, { "epoch": 0.9680443118875159, "grad_norm": 2.578125, "learning_rate": 4.9990040099704076e-05, "loss": 1.3555802345275878, "step": 11360 }, { "epoch": 0.9688964635705155, "grad_norm": 2.671875, "learning_rate": 4.998996204958144e-05, "loss": 1.3459888458251954, "step": 11370 }, { "epoch": 0.9697486152535151, "grad_norm": 2.796875, "learning_rate": 4.9989883694895226e-05, "loss": 1.4659692764282226, "step": 11380 }, { "epoch": 0.9706007669365146, "grad_norm": 2.046875, "learning_rate": 4.99898050356464e-05, "loss": 1.4592615127563477, "step": 11390 }, { "epoch": 0.9714529186195143, "grad_norm": 4.34375, "learning_rate": 4.998972607183592e-05, "loss": 1.4380024909973144, "step": 11400 }, { "epoch": 0.9723050703025139, "grad_norm": 2.4375, "learning_rate": 4.9989646803464736e-05, "loss": 1.486290740966797, "step": 11410 }, { "epoch": 0.9731572219855135, "grad_norm": 3.671875, "learning_rate": 4.998956723053382e-05, "loss": 1.4910863876342773, "step": 11420 }, { "epoch": 0.974009373668513, "grad_norm": 2.359375, "learning_rate": 4.998948735304415e-05, "loss": 1.4792920112609864, "step": 11430 }, { "epoch": 0.9748615253515126, "grad_norm": 2.375, "learning_rate": 4.998940717099669e-05, "loss": 1.380869197845459, "step": 11440 }, { "epoch": 0.9757136770345122, "grad_norm": 2.515625, "learning_rate": 4.998932668439243e-05, "loss": 1.3239039421081542, "step": 11450 }, { "epoch": 0.9765658287175117, "grad_norm": 2.609375, "learning_rate": 4.9989245893232335e-05, "loss": 1.3308337211608887, "step": 11460 }, { "epoch": 0.9774179804005113, "grad_norm": 1.8984375, "learning_rate": 4.9989164797517394e-05, "loss": 1.4285831451416016, "step": 11470 }, { "epoch": 0.9782701320835109, "grad_norm": 2.234375, "learning_rate": 4.9989083397248605e-05, "loss": 1.385193157196045, "step": 11480 }, { "epoch": 0.9791222837665104, "grad_norm": 2.234375, "learning_rate": 4.9989001692426945e-05, "loss": 1.326856803894043, "step": 11490 }, { "epoch": 0.97997443544951, "grad_norm": 2.28125, "learning_rate": 4.9988919683053424e-05, "loss": 1.1944941520690917, "step": 11500 }, { "epoch": 0.9808265871325096, "grad_norm": 2.890625, "learning_rate": 4.998883736912903e-05, "loss": 1.4002115249633789, "step": 11510 }, { "epoch": 0.9816787388155092, "grad_norm": 2.234375, "learning_rate": 4.998875475065477e-05, "loss": 1.36177339553833, "step": 11520 }, { "epoch": 0.9825308904985087, "grad_norm": 1.9296875, "learning_rate": 4.998867182763166e-05, "loss": 1.4757869720458985, "step": 11530 }, { "epoch": 0.9833830421815083, "grad_norm": 2.4375, "learning_rate": 4.9988588600060696e-05, "loss": 1.5996501922607422, "step": 11540 }, { "epoch": 0.9842351938645079, "grad_norm": 2.0625, "learning_rate": 4.99885050679429e-05, "loss": 1.4879375457763673, "step": 11550 }, { "epoch": 0.9850873455475074, "grad_norm": 2.65625, "learning_rate": 4.998842123127929e-05, "loss": 1.40621919631958, "step": 11560 }, { "epoch": 0.985939497230507, "grad_norm": 2.4375, "learning_rate": 4.998833709007089e-05, "loss": 1.4782609939575195, "step": 11570 }, { "epoch": 0.9867916489135066, "grad_norm": 2.34375, "learning_rate": 4.998825264431872e-05, "loss": 1.452402687072754, "step": 11580 }, { "epoch": 0.9876438005965061, "grad_norm": 2.015625, "learning_rate": 4.998816789402381e-05, "loss": 1.5723001480102539, "step": 11590 }, { "epoch": 0.9884959522795057, "grad_norm": 2.09375, "learning_rate": 4.998808283918719e-05, "loss": 1.4696362495422364, "step": 11600 }, { "epoch": 0.9893481039625053, "grad_norm": 1.8125, "learning_rate": 4.998799747980991e-05, "loss": 1.4384344100952149, "step": 11610 }, { "epoch": 0.9902002556455048, "grad_norm": 1.875, "learning_rate": 4.9987911815892994e-05, "loss": 1.5518507957458496, "step": 11620 }, { "epoch": 0.9910524073285045, "grad_norm": 2.125, "learning_rate": 4.9987825847437495e-05, "loss": 1.4027294158935546, "step": 11630 }, { "epoch": 0.9919045590115041, "grad_norm": 2.28125, "learning_rate": 4.998773957444447e-05, "loss": 1.380910873413086, "step": 11640 }, { "epoch": 0.9927567106945037, "grad_norm": 2.828125, "learning_rate": 4.9987652996914944e-05, "loss": 1.195546817779541, "step": 11650 }, { "epoch": 0.9936088623775032, "grad_norm": 1.9375, "learning_rate": 4.998756611484999e-05, "loss": 1.3487486839294434, "step": 11660 }, { "epoch": 0.9944610140605028, "grad_norm": 2.171875, "learning_rate": 4.9987478928250676e-05, "loss": 1.3536537170410157, "step": 11670 }, { "epoch": 0.9953131657435024, "grad_norm": 2.65625, "learning_rate": 4.9987391437118045e-05, "loss": 1.3583793640136719, "step": 11680 }, { "epoch": 0.9961653174265019, "grad_norm": 1.6640625, "learning_rate": 4.9987303641453175e-05, "loss": 1.467883586883545, "step": 11690 }, { "epoch": 0.9970174691095015, "grad_norm": 1.890625, "learning_rate": 4.9987215541257125e-05, "loss": 1.5005159378051758, "step": 11700 }, { "epoch": 0.9978696207925011, "grad_norm": 2.375, "learning_rate": 4.9987127136530986e-05, "loss": 1.487344741821289, "step": 11710 }, { "epoch": 0.9987217724755006, "grad_norm": 1.9921875, "learning_rate": 4.9987038427275814e-05, "loss": 1.390496826171875, "step": 11720 }, { "epoch": 0.9995739241585002, "grad_norm": 2.484375, "learning_rate": 4.9986949413492714e-05, "loss": 1.5053525924682618, "step": 11730 }, { "epoch": 0.9999147848317, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6404, "eval_loss": 1.416430115699768, "eval_runtime": 53.4399, "eval_samples_per_second": 369.798, "eval_steps_per_second": 23.129, "step": 11734 }, { "epoch": 1.0004260758414998, "grad_norm": 2.640625, "learning_rate": 4.998686009518275e-05, "loss": 1.3518590927124023, "step": 11740 }, { "epoch": 1.0012782275244994, "grad_norm": 3.546875, "learning_rate": 4.998677047234702e-05, "loss": 1.450896644592285, "step": 11750 }, { "epoch": 1.002130379207499, "grad_norm": 2.59375, "learning_rate": 4.998668054498662e-05, "loss": 1.2943151473999024, "step": 11760 }, { "epoch": 1.0029825308904985, "grad_norm": 2.3125, "learning_rate": 4.9986590313102645e-05, "loss": 1.3893926620483399, "step": 11770 }, { "epoch": 1.003834682573498, "grad_norm": 2.390625, "learning_rate": 4.998649977669618e-05, "loss": 1.3770320892333985, "step": 11780 }, { "epoch": 1.0046868342564976, "grad_norm": 1.8046875, "learning_rate": 4.9986408935768344e-05, "loss": 1.5715555191040038, "step": 11790 }, { "epoch": 1.0055389859394972, "grad_norm": 2.78125, "learning_rate": 4.9986317790320235e-05, "loss": 1.51957368850708, "step": 11800 }, { "epoch": 1.0063911376224968, "grad_norm": 2.15625, "learning_rate": 4.998622634035298e-05, "loss": 1.5366857528686524, "step": 11810 }, { "epoch": 1.0072432893054963, "grad_norm": 2.625, "learning_rate": 4.998613458586767e-05, "loss": 1.494708251953125, "step": 11820 }, { "epoch": 1.008095440988496, "grad_norm": 2.53125, "learning_rate": 4.9986042526865444e-05, "loss": 1.300120735168457, "step": 11830 }, { "epoch": 1.0089475926714955, "grad_norm": 1.859375, "learning_rate": 4.998595016334741e-05, "loss": 1.4294206619262695, "step": 11840 }, { "epoch": 1.009799744354495, "grad_norm": 1.875, "learning_rate": 4.99858574953147e-05, "loss": 1.4553909301757812, "step": 11850 }, { "epoch": 1.0106518960374946, "grad_norm": 2.140625, "learning_rate": 4.998576452276845e-05, "loss": 1.5430313110351563, "step": 11860 }, { "epoch": 1.0115040477204942, "grad_norm": 3.3125, "learning_rate": 4.9985671245709776e-05, "loss": 1.2245972633361817, "step": 11870 }, { "epoch": 1.0123561994034938, "grad_norm": 4.0625, "learning_rate": 4.9985577664139826e-05, "loss": 1.3973017692565919, "step": 11880 }, { "epoch": 1.0132083510864933, "grad_norm": 2.671875, "learning_rate": 4.998548377805974e-05, "loss": 1.4998539924621581, "step": 11890 }, { "epoch": 1.0140605027694929, "grad_norm": 2.96875, "learning_rate": 4.998538958747065e-05, "loss": 1.3996145248413085, "step": 11900 }, { "epoch": 1.0149126544524925, "grad_norm": 2.09375, "learning_rate": 4.998529509237373e-05, "loss": 1.4243392944335938, "step": 11910 }, { "epoch": 1.015764806135492, "grad_norm": 2.984375, "learning_rate": 4.998520029277012e-05, "loss": 1.370992660522461, "step": 11920 }, { "epoch": 1.0166169578184916, "grad_norm": 3.8125, "learning_rate": 4.998510518866096e-05, "loss": 1.5063575744628905, "step": 11930 }, { "epoch": 1.0174691095014912, "grad_norm": 2.46875, "learning_rate": 4.998500978004742e-05, "loss": 1.4521657943725585, "step": 11940 }, { "epoch": 1.0183212611844907, "grad_norm": 2.3125, "learning_rate": 4.998491406693067e-05, "loss": 1.2591962814331055, "step": 11950 }, { "epoch": 1.0191734128674903, "grad_norm": 2.625, "learning_rate": 4.9984818049311865e-05, "loss": 1.431407356262207, "step": 11960 }, { "epoch": 1.02002556455049, "grad_norm": 2.1875, "learning_rate": 4.9984721727192185e-05, "loss": 1.4321168899536132, "step": 11970 }, { "epoch": 1.0208777162334897, "grad_norm": 2.546875, "learning_rate": 4.99846251005728e-05, "loss": 1.3300039291381835, "step": 11980 }, { "epoch": 1.0217298679164892, "grad_norm": 2.984375, "learning_rate": 4.998452816945488e-05, "loss": 1.4713505744934081, "step": 11990 }, { "epoch": 1.0225820195994888, "grad_norm": 1.90625, "learning_rate": 4.998443093383962e-05, "loss": 1.5732263565063476, "step": 12000 }, { "epoch": 1.0234341712824884, "grad_norm": 2.28125, "learning_rate": 4.99843333937282e-05, "loss": 1.4251955032348633, "step": 12010 }, { "epoch": 1.024286322965488, "grad_norm": 3.515625, "learning_rate": 4.998423554912181e-05, "loss": 1.413872241973877, "step": 12020 }, { "epoch": 1.0251384746484875, "grad_norm": 2.046875, "learning_rate": 4.9984137400021625e-05, "loss": 1.3837128639221192, "step": 12030 }, { "epoch": 1.025990626331487, "grad_norm": 2.015625, "learning_rate": 4.998403894642887e-05, "loss": 1.339493179321289, "step": 12040 }, { "epoch": 1.0268427780144866, "grad_norm": 2.578125, "learning_rate": 4.998394018834472e-05, "loss": 1.3920735359191894, "step": 12050 }, { "epoch": 1.0276949296974862, "grad_norm": 2.84375, "learning_rate": 4.998384112577039e-05, "loss": 1.4415093421936036, "step": 12060 }, { "epoch": 1.0285470813804858, "grad_norm": 1.96875, "learning_rate": 4.9983741758707095e-05, "loss": 1.4278323173522949, "step": 12070 }, { "epoch": 1.0293992330634854, "grad_norm": 2.078125, "learning_rate": 4.998364208715603e-05, "loss": 1.5016247749328613, "step": 12080 }, { "epoch": 1.030251384746485, "grad_norm": 2.578125, "learning_rate": 4.9983542111118425e-05, "loss": 1.4637662887573242, "step": 12090 }, { "epoch": 1.0311035364294845, "grad_norm": 2.34375, "learning_rate": 4.998344183059549e-05, "loss": 1.5033452987670899, "step": 12100 }, { "epoch": 1.031955688112484, "grad_norm": 3.25, "learning_rate": 4.9983341245588447e-05, "loss": 1.4359214782714844, "step": 12110 }, { "epoch": 1.0328078397954836, "grad_norm": 2.34375, "learning_rate": 4.9983240356098515e-05, "loss": 1.5469663619995118, "step": 12120 }, { "epoch": 1.0336599914784832, "grad_norm": 2.390625, "learning_rate": 4.998313916212694e-05, "loss": 1.3888364791870118, "step": 12130 }, { "epoch": 1.0345121431614828, "grad_norm": 1.890625, "learning_rate": 4.998303766367494e-05, "loss": 1.4594334602355956, "step": 12140 }, { "epoch": 1.0353642948444823, "grad_norm": 2.09375, "learning_rate": 4.998293586074376e-05, "loss": 1.320540714263916, "step": 12150 }, { "epoch": 1.036216446527482, "grad_norm": 1.9453125, "learning_rate": 4.9982833753334646e-05, "loss": 1.454542350769043, "step": 12160 }, { "epoch": 1.0370685982104815, "grad_norm": 2.203125, "learning_rate": 4.998273134144883e-05, "loss": 1.4811803817749023, "step": 12170 }, { "epoch": 1.037920749893481, "grad_norm": 2.734375, "learning_rate": 4.998262862508757e-05, "loss": 1.5302457809448242, "step": 12180 }, { "epoch": 1.0387729015764806, "grad_norm": 2.484375, "learning_rate": 4.9982525604252115e-05, "loss": 1.4140480995178222, "step": 12190 }, { "epoch": 1.0396250532594802, "grad_norm": 2.40625, "learning_rate": 4.9982422278943716e-05, "loss": 1.4082380294799806, "step": 12200 }, { "epoch": 1.0404772049424797, "grad_norm": 1.890625, "learning_rate": 4.998231864916363e-05, "loss": 1.5216492652893066, "step": 12210 }, { "epoch": 1.0413293566254793, "grad_norm": 1.8359375, "learning_rate": 4.998221471491314e-05, "loss": 1.4953413963317872, "step": 12220 }, { "epoch": 1.0421815083084789, "grad_norm": 2.59375, "learning_rate": 4.998211047619349e-05, "loss": 1.3577397346496582, "step": 12230 }, { "epoch": 1.0430336599914785, "grad_norm": 1.953125, "learning_rate": 4.9982005933005955e-05, "loss": 1.4839506149291992, "step": 12240 }, { "epoch": 1.043885811674478, "grad_norm": 1.9921875, "learning_rate": 4.9981901085351816e-05, "loss": 1.5416820526123047, "step": 12250 }, { "epoch": 1.0447379633574776, "grad_norm": 1.9921875, "learning_rate": 4.998179593323235e-05, "loss": 1.3869638442993164, "step": 12260 }, { "epoch": 1.0455901150404772, "grad_norm": 2.890625, "learning_rate": 4.998169047664883e-05, "loss": 1.4945295333862305, "step": 12270 }, { "epoch": 1.0464422667234767, "grad_norm": 3.078125, "learning_rate": 4.998158471560256e-05, "loss": 1.3314013481140137, "step": 12280 }, { "epoch": 1.0472944184064763, "grad_norm": 1.9140625, "learning_rate": 4.99814786500948e-05, "loss": 1.4593412399291992, "step": 12290 }, { "epoch": 1.0481465700894759, "grad_norm": 2.34375, "learning_rate": 4.998137228012687e-05, "loss": 1.4362794876098632, "step": 12300 }, { "epoch": 1.0489987217724754, "grad_norm": 2.5625, "learning_rate": 4.998126560570005e-05, "loss": 1.346470546722412, "step": 12310 }, { "epoch": 1.049850873455475, "grad_norm": 2.328125, "learning_rate": 4.9981158626815664e-05, "loss": 1.22406005859375, "step": 12320 }, { "epoch": 1.0507030251384746, "grad_norm": 2.859375, "learning_rate": 4.998105134347498e-05, "loss": 1.284235668182373, "step": 12330 }, { "epoch": 1.0515551768214741, "grad_norm": 1.96875, "learning_rate": 4.9980943755679325e-05, "loss": 1.476090145111084, "step": 12340 }, { "epoch": 1.0524073285044737, "grad_norm": 2.5, "learning_rate": 4.998083586343001e-05, "loss": 1.634233283996582, "step": 12350 }, { "epoch": 1.0532594801874733, "grad_norm": 2.015625, "learning_rate": 4.9980727666728345e-05, "loss": 1.481569766998291, "step": 12360 }, { "epoch": 1.0541116318704729, "grad_norm": 2.15625, "learning_rate": 4.998061916557565e-05, "loss": 1.4954610824584962, "step": 12370 }, { "epoch": 1.0549637835534724, "grad_norm": 2.65625, "learning_rate": 4.9980510359973256e-05, "loss": 1.2733585357666015, "step": 12380 }, { "epoch": 1.055815935236472, "grad_norm": 1.84375, "learning_rate": 4.9980401249922475e-05, "loss": 1.3685409545898437, "step": 12390 }, { "epoch": 1.0566680869194716, "grad_norm": 2.546875, "learning_rate": 4.9980291835424654e-05, "loss": 1.386093521118164, "step": 12400 }, { "epoch": 1.0575202386024714, "grad_norm": 2.0, "learning_rate": 4.998018211648111e-05, "loss": 1.3888060569763183, "step": 12410 }, { "epoch": 1.058372390285471, "grad_norm": 2.921875, "learning_rate": 4.9980072093093186e-05, "loss": 1.4281082153320312, "step": 12420 }, { "epoch": 1.0592245419684705, "grad_norm": 2.6875, "learning_rate": 4.9979961765262215e-05, "loss": 1.3911914825439453, "step": 12430 }, { "epoch": 1.06007669365147, "grad_norm": 1.875, "learning_rate": 4.9979851132989565e-05, "loss": 1.4728840827941894, "step": 12440 }, { "epoch": 1.0609288453344696, "grad_norm": 2.828125, "learning_rate": 4.997974019627656e-05, "loss": 1.4332292556762696, "step": 12450 }, { "epoch": 1.0617809970174692, "grad_norm": 1.9453125, "learning_rate": 4.997962895512457e-05, "loss": 1.5307395935058594, "step": 12460 }, { "epoch": 1.0626331487004688, "grad_norm": 2.28125, "learning_rate": 4.997951740953494e-05, "loss": 1.5040672302246094, "step": 12470 }, { "epoch": 1.0634853003834683, "grad_norm": 2.421875, "learning_rate": 4.997940555950902e-05, "loss": 1.4672704696655274, "step": 12480 }, { "epoch": 1.064337452066468, "grad_norm": 2.046875, "learning_rate": 4.99792934050482e-05, "loss": 1.404226303100586, "step": 12490 }, { "epoch": 1.0651896037494675, "grad_norm": 2.390625, "learning_rate": 4.997918094615382e-05, "loss": 1.2676002502441406, "step": 12500 }, { "epoch": 1.066041755432467, "grad_norm": 2.375, "learning_rate": 4.9979068182827274e-05, "loss": 1.3923645973205567, "step": 12510 }, { "epoch": 1.0668939071154666, "grad_norm": 2.0625, "learning_rate": 4.9978955115069914e-05, "loss": 1.4931719779968262, "step": 12520 }, { "epoch": 1.0677460587984662, "grad_norm": 2.828125, "learning_rate": 4.9978841742883133e-05, "loss": 1.325815486907959, "step": 12530 }, { "epoch": 1.0685982104814657, "grad_norm": 2.546875, "learning_rate": 4.997872806626831e-05, "loss": 1.3847946166992187, "step": 12540 }, { "epoch": 1.0694503621644653, "grad_norm": 2.953125, "learning_rate": 4.997861408522683e-05, "loss": 1.4351130485534669, "step": 12550 }, { "epoch": 1.0703025138474649, "grad_norm": 2.171875, "learning_rate": 4.997849979976008e-05, "loss": 1.5130733489990233, "step": 12560 }, { "epoch": 1.0711546655304645, "grad_norm": 1.875, "learning_rate": 4.997838520986945e-05, "loss": 1.5373740196228027, "step": 12570 }, { "epoch": 1.072006817213464, "grad_norm": 2.765625, "learning_rate": 4.9978270315556344e-05, "loss": 1.3545192718505858, "step": 12580 }, { "epoch": 1.0728589688964636, "grad_norm": 1.75, "learning_rate": 4.997815511682215e-05, "loss": 1.4924836158752441, "step": 12590 }, { "epoch": 1.0737111205794632, "grad_norm": 1.671875, "learning_rate": 4.9978039613668284e-05, "loss": 1.2360687255859375, "step": 12600 }, { "epoch": 1.0745632722624627, "grad_norm": 1.59375, "learning_rate": 4.9977923806096154e-05, "loss": 1.3663119316101073, "step": 12610 }, { "epoch": 1.0754154239454623, "grad_norm": 2.546875, "learning_rate": 4.997780769410717e-05, "loss": 1.3977917671203612, "step": 12620 }, { "epoch": 1.0762675756284619, "grad_norm": 3.6875, "learning_rate": 4.997769127770274e-05, "loss": 1.5056680679321288, "step": 12630 }, { "epoch": 1.0771197273114614, "grad_norm": 2.9375, "learning_rate": 4.997757455688429e-05, "loss": 1.3184284210205077, "step": 12640 }, { "epoch": 1.077971878994461, "grad_norm": 2.21875, "learning_rate": 4.9977457531653236e-05, "loss": 1.4398513793945313, "step": 12650 }, { "epoch": 1.0788240306774606, "grad_norm": 2.53125, "learning_rate": 4.997734020201101e-05, "loss": 1.3606810569763184, "step": 12660 }, { "epoch": 1.0796761823604601, "grad_norm": 2.046875, "learning_rate": 4.9977222567959044e-05, "loss": 1.386436653137207, "step": 12670 }, { "epoch": 1.0805283340434597, "grad_norm": 2.15625, "learning_rate": 4.997710462949876e-05, "loss": 1.475816535949707, "step": 12680 }, { "epoch": 1.0813804857264593, "grad_norm": 2.015625, "learning_rate": 4.997698638663161e-05, "loss": 1.3579336166381837, "step": 12690 }, { "epoch": 1.0822326374094589, "grad_norm": 2.1875, "learning_rate": 4.9976867839359034e-05, "loss": 1.4217255592346192, "step": 12700 }, { "epoch": 1.0830847890924584, "grad_norm": 1.8046875, "learning_rate": 4.997674898768246e-05, "loss": 1.3474641799926759, "step": 12710 }, { "epoch": 1.083936940775458, "grad_norm": 2.546875, "learning_rate": 4.997662983160335e-05, "loss": 1.4057387351989745, "step": 12720 }, { "epoch": 1.0847890924584576, "grad_norm": 3.0, "learning_rate": 4.9976510371123155e-05, "loss": 1.2051591873168945, "step": 12730 }, { "epoch": 1.0856412441414571, "grad_norm": 1.7421875, "learning_rate": 4.997639060624333e-05, "loss": 1.4117566108703614, "step": 12740 }, { "epoch": 1.0864933958244567, "grad_norm": 1.765625, "learning_rate": 4.997627053696533e-05, "loss": 1.303009605407715, "step": 12750 }, { "epoch": 1.0873455475074563, "grad_norm": 2.875, "learning_rate": 4.997615016329063e-05, "loss": 1.4416688919067382, "step": 12760 }, { "epoch": 1.0881976991904558, "grad_norm": 1.9765625, "learning_rate": 4.9976029485220686e-05, "loss": 1.478933620452881, "step": 12770 }, { "epoch": 1.0890498508734554, "grad_norm": 2.15625, "learning_rate": 4.9975908502756974e-05, "loss": 1.3329050064086914, "step": 12780 }, { "epoch": 1.089902002556455, "grad_norm": 1.984375, "learning_rate": 4.997578721590097e-05, "loss": 1.3385913848876954, "step": 12790 }, { "epoch": 1.0907541542394545, "grad_norm": 2.90625, "learning_rate": 4.9975665624654145e-05, "loss": 1.3899391174316407, "step": 12800 }, { "epoch": 1.0916063059224541, "grad_norm": 2.171875, "learning_rate": 4.997554372901798e-05, "loss": 1.313408088684082, "step": 12810 }, { "epoch": 1.0924584576054537, "grad_norm": 2.890625, "learning_rate": 4.997542152899397e-05, "loss": 1.3519838333129883, "step": 12820 }, { "epoch": 1.0933106092884533, "grad_norm": 2.5, "learning_rate": 4.9975299024583606e-05, "loss": 1.4008302688598633, "step": 12830 }, { "epoch": 1.0941627609714528, "grad_norm": 1.7578125, "learning_rate": 4.997517621578837e-05, "loss": 1.4138351440429688, "step": 12840 }, { "epoch": 1.0950149126544524, "grad_norm": 3.21875, "learning_rate": 4.997505310260976e-05, "loss": 1.3205681800842286, "step": 12850 }, { "epoch": 1.095867064337452, "grad_norm": 1.8359375, "learning_rate": 4.997492968504929e-05, "loss": 1.4457284927368164, "step": 12860 }, { "epoch": 1.0967192160204515, "grad_norm": 2.578125, "learning_rate": 4.997480596310844e-05, "loss": 1.4243770599365235, "step": 12870 }, { "epoch": 1.0975713677034513, "grad_norm": 2.421875, "learning_rate": 4.997468193678875e-05, "loss": 1.5555892944335938, "step": 12880 }, { "epoch": 1.0984235193864509, "grad_norm": 2.546875, "learning_rate": 4.99745576060917e-05, "loss": 1.4562106132507324, "step": 12890 }, { "epoch": 1.0992756710694505, "grad_norm": 2.1875, "learning_rate": 4.997443297101882e-05, "loss": 1.4111086845397949, "step": 12900 }, { "epoch": 1.10012782275245, "grad_norm": 2.234375, "learning_rate": 4.997430803157163e-05, "loss": 1.278238582611084, "step": 12910 }, { "epoch": 1.1009799744354496, "grad_norm": 2.875, "learning_rate": 4.9974182787751657e-05, "loss": 1.306480598449707, "step": 12920 }, { "epoch": 1.1018321261184492, "grad_norm": 2.4375, "learning_rate": 4.9974057239560415e-05, "loss": 1.3373437881469727, "step": 12930 }, { "epoch": 1.1026842778014487, "grad_norm": 1.8984375, "learning_rate": 4.997393138699944e-05, "loss": 1.4663982391357422, "step": 12940 }, { "epoch": 1.1035364294844483, "grad_norm": 2.265625, "learning_rate": 4.997380523007026e-05, "loss": 1.2609609603881835, "step": 12950 }, { "epoch": 1.1043885811674479, "grad_norm": 3.40625, "learning_rate": 4.997367876877443e-05, "loss": 1.3082695007324219, "step": 12960 }, { "epoch": 1.1052407328504474, "grad_norm": 3.9375, "learning_rate": 4.9973552003113464e-05, "loss": 1.425936508178711, "step": 12970 }, { "epoch": 1.106092884533447, "grad_norm": 2.65625, "learning_rate": 4.9973424933088944e-05, "loss": 1.3133938789367676, "step": 12980 }, { "epoch": 1.1069450362164466, "grad_norm": 1.6171875, "learning_rate": 4.997329755870238e-05, "loss": 1.5252142906188966, "step": 12990 }, { "epoch": 1.1077971878994461, "grad_norm": 3.15625, "learning_rate": 4.997316987995535e-05, "loss": 1.4646465301513671, "step": 13000 }, { "epoch": 1.1086493395824457, "grad_norm": 2.765625, "learning_rate": 4.99730418968494e-05, "loss": 1.4005062103271484, "step": 13010 }, { "epoch": 1.1095014912654453, "grad_norm": 1.71875, "learning_rate": 4.997291360938609e-05, "loss": 1.561762237548828, "step": 13020 }, { "epoch": 1.1103536429484449, "grad_norm": 2.25, "learning_rate": 4.9972785017566985e-05, "loss": 1.4040864944458007, "step": 13030 }, { "epoch": 1.1112057946314444, "grad_norm": 2.84375, "learning_rate": 4.9972656121393655e-05, "loss": 1.354722023010254, "step": 13040 }, { "epoch": 1.112057946314444, "grad_norm": 2.296875, "learning_rate": 4.9972526920867665e-05, "loss": 1.4819647789001464, "step": 13050 }, { "epoch": 1.1129100979974436, "grad_norm": 1.7734375, "learning_rate": 4.9972397415990594e-05, "loss": 1.3606468200683595, "step": 13060 }, { "epoch": 1.1137622496804431, "grad_norm": 2.90625, "learning_rate": 4.997226760676402e-05, "loss": 1.3903170585632325, "step": 13070 }, { "epoch": 1.1146144013634427, "grad_norm": 2.109375, "learning_rate": 4.9972137493189525e-05, "loss": 1.603154182434082, "step": 13080 }, { "epoch": 1.1154665530464423, "grad_norm": 1.859375, "learning_rate": 4.997200707526869e-05, "loss": 1.3675201416015625, "step": 13090 }, { "epoch": 1.1163187047294418, "grad_norm": 2.234375, "learning_rate": 4.997187635300311e-05, "loss": 1.3693686485290528, "step": 13100 }, { "epoch": 1.1171708564124414, "grad_norm": 3.78125, "learning_rate": 4.997174532639438e-05, "loss": 1.2549760818481446, "step": 13110 }, { "epoch": 1.118023008095441, "grad_norm": 2.5, "learning_rate": 4.997161399544409e-05, "loss": 1.4973320007324218, "step": 13120 }, { "epoch": 1.1188751597784405, "grad_norm": 1.921875, "learning_rate": 4.997148236015385e-05, "loss": 1.390463924407959, "step": 13130 }, { "epoch": 1.1197273114614401, "grad_norm": 2.859375, "learning_rate": 4.997135042052525e-05, "loss": 1.2929632186889648, "step": 13140 }, { "epoch": 1.1205794631444397, "grad_norm": 2.359375, "learning_rate": 4.997121817655991e-05, "loss": 1.338559341430664, "step": 13150 }, { "epoch": 1.1214316148274392, "grad_norm": 3.109375, "learning_rate": 4.997108562825944e-05, "loss": 1.2718894004821777, "step": 13160 }, { "epoch": 1.1222837665104388, "grad_norm": 2.21875, "learning_rate": 4.997095277562545e-05, "loss": 1.481546401977539, "step": 13170 }, { "epoch": 1.1231359181934384, "grad_norm": 3.484375, "learning_rate": 4.997081961865956e-05, "loss": 1.2065563201904297, "step": 13180 }, { "epoch": 1.123988069876438, "grad_norm": 2.171875, "learning_rate": 4.99706861573634e-05, "loss": 1.3346307754516602, "step": 13190 }, { "epoch": 1.1248402215594375, "grad_norm": 2.1875, "learning_rate": 4.997055239173859e-05, "loss": 1.329374122619629, "step": 13200 }, { "epoch": 1.125692373242437, "grad_norm": 2.578125, "learning_rate": 4.997041832178677e-05, "loss": 1.3535223007202148, "step": 13210 }, { "epoch": 1.1265445249254367, "grad_norm": 2.890625, "learning_rate": 4.997028394750955e-05, "loss": 1.3633646011352538, "step": 13220 }, { "epoch": 1.1273966766084362, "grad_norm": 2.046875, "learning_rate": 4.997014926890859e-05, "loss": 1.4444130897521972, "step": 13230 }, { "epoch": 1.1282488282914358, "grad_norm": 3.03125, "learning_rate": 4.997001428598553e-05, "loss": 1.4906270027160644, "step": 13240 }, { "epoch": 1.1291009799744354, "grad_norm": 2.609375, "learning_rate": 4.9969878998742014e-05, "loss": 1.5054362297058106, "step": 13250 }, { "epoch": 1.129953131657435, "grad_norm": 1.953125, "learning_rate": 4.996974340717968e-05, "loss": 1.480711078643799, "step": 13260 }, { "epoch": 1.1308052833404345, "grad_norm": 2.578125, "learning_rate": 4.99696075113002e-05, "loss": 1.3347627639770507, "step": 13270 }, { "epoch": 1.131657435023434, "grad_norm": 2.109375, "learning_rate": 4.99694713111052e-05, "loss": 1.35772705078125, "step": 13280 }, { "epoch": 1.1325095867064339, "grad_norm": 3.375, "learning_rate": 4.996933480659637e-05, "loss": 1.3494924545288085, "step": 13290 }, { "epoch": 1.1333617383894334, "grad_norm": 3.203125, "learning_rate": 4.996919799777535e-05, "loss": 1.5030481338500976, "step": 13300 }, { "epoch": 1.134213890072433, "grad_norm": 2.734375, "learning_rate": 4.996906088464383e-05, "loss": 1.3377202987670898, "step": 13310 }, { "epoch": 1.1350660417554326, "grad_norm": 2.5, "learning_rate": 4.996892346720346e-05, "loss": 1.5510427474975585, "step": 13320 }, { "epoch": 1.1359181934384321, "grad_norm": 2.0625, "learning_rate": 4.996878574545594e-05, "loss": 1.4075064659118652, "step": 13330 }, { "epoch": 1.1367703451214317, "grad_norm": 2.40625, "learning_rate": 4.9968647719402914e-05, "loss": 1.230886173248291, "step": 13340 }, { "epoch": 1.1376224968044313, "grad_norm": 2.03125, "learning_rate": 4.99685093890461e-05, "loss": 1.4498579978942872, "step": 13350 }, { "epoch": 1.1384746484874309, "grad_norm": 2.546875, "learning_rate": 4.996837075438716e-05, "loss": 1.3252117156982421, "step": 13360 }, { "epoch": 1.1393268001704304, "grad_norm": 2.484375, "learning_rate": 4.996823181542779e-05, "loss": 1.3897404670715332, "step": 13370 }, { "epoch": 1.14017895185343, "grad_norm": 3.21875, "learning_rate": 4.9968092572169684e-05, "loss": 1.1262738227844238, "step": 13380 }, { "epoch": 1.1410311035364296, "grad_norm": 2.390625, "learning_rate": 4.996795302461454e-05, "loss": 1.3380648612976074, "step": 13390 }, { "epoch": 1.1418832552194291, "grad_norm": 3.546875, "learning_rate": 4.9967813172764054e-05, "loss": 1.319904899597168, "step": 13400 }, { "epoch": 1.1427354069024287, "grad_norm": 1.953125, "learning_rate": 4.9967673016619944e-05, "loss": 1.3360973358154298, "step": 13410 }, { "epoch": 1.1435875585854283, "grad_norm": 2.59375, "learning_rate": 4.9967532556183896e-05, "loss": 1.4211759567260742, "step": 13420 }, { "epoch": 1.1444397102684278, "grad_norm": 3.296875, "learning_rate": 4.996739179145764e-05, "loss": 1.2231721878051758, "step": 13430 }, { "epoch": 1.1452918619514274, "grad_norm": 3.515625, "learning_rate": 4.996725072244289e-05, "loss": 1.3724647521972657, "step": 13440 }, { "epoch": 1.146144013634427, "grad_norm": 1.8828125, "learning_rate": 4.996710934914136e-05, "loss": 1.4725532531738281, "step": 13450 }, { "epoch": 1.1469961653174265, "grad_norm": 3.03125, "learning_rate": 4.996696767155477e-05, "loss": 1.5111541748046875, "step": 13460 }, { "epoch": 1.147848317000426, "grad_norm": 2.4375, "learning_rate": 4.9966825689684853e-05, "loss": 1.4036020278930663, "step": 13470 }, { "epoch": 1.1487004686834257, "grad_norm": 1.7890625, "learning_rate": 4.996668340353334e-05, "loss": 1.3906058311462401, "step": 13480 }, { "epoch": 1.1495526203664252, "grad_norm": 1.9921875, "learning_rate": 4.996654081310196e-05, "loss": 1.4115144729614257, "step": 13490 }, { "epoch": 1.1504047720494248, "grad_norm": 2.09375, "learning_rate": 4.9966397918392446e-05, "loss": 1.4922060012817382, "step": 13500 }, { "epoch": 1.1512569237324244, "grad_norm": 2.203125, "learning_rate": 4.996625471940656e-05, "loss": 1.4177043914794922, "step": 13510 }, { "epoch": 1.152109075415424, "grad_norm": 3.796875, "learning_rate": 4.996611121614602e-05, "loss": 1.5572170257568358, "step": 13520 }, { "epoch": 1.1529612270984235, "grad_norm": 2.09375, "learning_rate": 4.9965967408612596e-05, "loss": 1.4095335960388184, "step": 13530 }, { "epoch": 1.153813378781423, "grad_norm": 2.140625, "learning_rate": 4.9965823296808034e-05, "loss": 1.549162006378174, "step": 13540 }, { "epoch": 1.1546655304644227, "grad_norm": 2.859375, "learning_rate": 4.996567888073409e-05, "loss": 1.37208251953125, "step": 13550 }, { "epoch": 1.1555176821474222, "grad_norm": 2.578125, "learning_rate": 4.996553416039253e-05, "loss": 1.4140872955322266, "step": 13560 }, { "epoch": 1.1563698338304218, "grad_norm": 1.390625, "learning_rate": 4.99653891357851e-05, "loss": 1.541592502593994, "step": 13570 }, { "epoch": 1.1572219855134214, "grad_norm": 2.03125, "learning_rate": 4.9965243806913576e-05, "loss": 1.5492877960205078, "step": 13580 }, { "epoch": 1.158074137196421, "grad_norm": 3.546875, "learning_rate": 4.996509817377974e-05, "loss": 1.3053513526916505, "step": 13590 }, { "epoch": 1.1589262888794205, "grad_norm": 2.6875, "learning_rate": 4.996495223638536e-05, "loss": 1.422947311401367, "step": 13600 }, { "epoch": 1.15977844056242, "grad_norm": 2.765625, "learning_rate": 4.996480599473221e-05, "loss": 1.2517803192138672, "step": 13610 }, { "epoch": 1.1606305922454196, "grad_norm": 2.34375, "learning_rate": 4.9964659448822085e-05, "loss": 1.4619425773620605, "step": 13620 }, { "epoch": 1.1614827439284192, "grad_norm": 2.21875, "learning_rate": 4.9964512598656756e-05, "loss": 1.4445784568786622, "step": 13630 }, { "epoch": 1.1623348956114188, "grad_norm": 1.609375, "learning_rate": 4.996436544423801e-05, "loss": 1.421363639831543, "step": 13640 }, { "epoch": 1.1631870472944184, "grad_norm": 1.796875, "learning_rate": 4.9964217985567664e-05, "loss": 1.3160770416259766, "step": 13650 }, { "epoch": 1.164039198977418, "grad_norm": 2.171875, "learning_rate": 4.996407022264749e-05, "loss": 1.4370339393615723, "step": 13660 }, { "epoch": 1.1648913506604175, "grad_norm": 1.859375, "learning_rate": 4.9963922155479306e-05, "loss": 1.3367527961730956, "step": 13670 }, { "epoch": 1.165743502343417, "grad_norm": 1.796875, "learning_rate": 4.99637737840649e-05, "loss": 1.341475486755371, "step": 13680 }, { "epoch": 1.1665956540264166, "grad_norm": 2.671875, "learning_rate": 4.99636251084061e-05, "loss": 1.5052204132080078, "step": 13690 }, { "epoch": 1.1674478057094162, "grad_norm": 3.5625, "learning_rate": 4.996347612850471e-05, "loss": 1.3885536193847656, "step": 13700 }, { "epoch": 1.1682999573924158, "grad_norm": 2.40625, "learning_rate": 4.996332684436254e-05, "loss": 1.4712907791137695, "step": 13710 }, { "epoch": 1.1691521090754153, "grad_norm": 2.203125, "learning_rate": 4.996317725598141e-05, "loss": 1.379455280303955, "step": 13720 }, { "epoch": 1.170004260758415, "grad_norm": 2.421875, "learning_rate": 4.9963027363363145e-05, "loss": 1.5081066131591796, "step": 13730 }, { "epoch": 1.1708564124414145, "grad_norm": 2.171875, "learning_rate": 4.996287716650958e-05, "loss": 1.4720759391784668, "step": 13740 }, { "epoch": 1.171708564124414, "grad_norm": 2.53125, "learning_rate": 4.996272666542253e-05, "loss": 1.5439623832702636, "step": 13750 }, { "epoch": 1.1725607158074136, "grad_norm": 2.453125, "learning_rate": 4.9962575860103845e-05, "loss": 1.1366774559020996, "step": 13760 }, { "epoch": 1.1734128674904132, "grad_norm": 2.4375, "learning_rate": 4.996242475055535e-05, "loss": 1.3005796432495118, "step": 13770 }, { "epoch": 1.1742650191734127, "grad_norm": 2.546875, "learning_rate": 4.99622733367789e-05, "loss": 1.5651865005493164, "step": 13780 }, { "epoch": 1.1751171708564123, "grad_norm": 1.765625, "learning_rate": 4.996212161877632e-05, "loss": 1.4218174934387207, "step": 13790 }, { "epoch": 1.175969322539412, "grad_norm": 2.984375, "learning_rate": 4.9961969596549484e-05, "loss": 1.4960577011108398, "step": 13800 }, { "epoch": 1.1768214742224117, "grad_norm": 1.8359375, "learning_rate": 4.996181727010023e-05, "loss": 1.578729248046875, "step": 13810 }, { "epoch": 1.1776736259054112, "grad_norm": 2.203125, "learning_rate": 4.996166463943042e-05, "loss": 1.3999414443969727, "step": 13820 }, { "epoch": 1.1785257775884108, "grad_norm": 2.09375, "learning_rate": 4.99615117045419e-05, "loss": 1.4752056121826171, "step": 13830 }, { "epoch": 1.1793779292714104, "grad_norm": 2.921875, "learning_rate": 4.996135846543655e-05, "loss": 1.4078435897827148, "step": 13840 }, { "epoch": 1.18023008095441, "grad_norm": 2.375, "learning_rate": 4.996120492211624e-05, "loss": 1.393674659729004, "step": 13850 }, { "epoch": 1.1810822326374095, "grad_norm": 3.453125, "learning_rate": 4.996105107458283e-05, "loss": 1.4826480865478515, "step": 13860 }, { "epoch": 1.181934384320409, "grad_norm": 2.3125, "learning_rate": 4.9960896922838195e-05, "loss": 1.3742030143737793, "step": 13870 }, { "epoch": 1.1827865360034087, "grad_norm": 2.0, "learning_rate": 4.996074246688422e-05, "loss": 1.5207931518554687, "step": 13880 }, { "epoch": 1.1836386876864082, "grad_norm": 2.640625, "learning_rate": 4.996058770672278e-05, "loss": 1.4547650337219238, "step": 13890 }, { "epoch": 1.1844908393694078, "grad_norm": 3.421875, "learning_rate": 4.9960432642355775e-05, "loss": 1.1624117851257325, "step": 13900 }, { "epoch": 1.1853429910524074, "grad_norm": 1.9296875, "learning_rate": 4.9960277273785084e-05, "loss": 1.5142179489135743, "step": 13910 }, { "epoch": 1.186195142735407, "grad_norm": 1.640625, "learning_rate": 4.99601216010126e-05, "loss": 1.5237648010253906, "step": 13920 }, { "epoch": 1.1870472944184065, "grad_norm": 3.34375, "learning_rate": 4.995996562404022e-05, "loss": 1.3330114364624024, "step": 13930 }, { "epoch": 1.187899446101406, "grad_norm": 1.890625, "learning_rate": 4.995980934286986e-05, "loss": 1.5271039962768556, "step": 13940 }, { "epoch": 1.1887515977844056, "grad_norm": 2.578125, "learning_rate": 4.995965275750339e-05, "loss": 1.5795836448669434, "step": 13950 }, { "epoch": 1.1896037494674052, "grad_norm": 2.984375, "learning_rate": 4.995949586794276e-05, "loss": 1.3432294845581054, "step": 13960 }, { "epoch": 1.1904559011504048, "grad_norm": 4.75, "learning_rate": 4.995933867418986e-05, "loss": 1.2909749031066895, "step": 13970 }, { "epoch": 1.1913080528334044, "grad_norm": 2.109375, "learning_rate": 4.9959181176246607e-05, "loss": 1.503962516784668, "step": 13980 }, { "epoch": 1.192160204516404, "grad_norm": 2.375, "learning_rate": 4.995902337411492e-05, "loss": 1.4833446502685548, "step": 13990 }, { "epoch": 1.1930123561994035, "grad_norm": 1.7578125, "learning_rate": 4.995886526779673e-05, "loss": 1.3071735382080079, "step": 14000 }, { "epoch": 1.193864507882403, "grad_norm": 1.65625, "learning_rate": 4.995870685729395e-05, "loss": 1.499019241333008, "step": 14010 }, { "epoch": 1.1947166595654026, "grad_norm": 3.34375, "learning_rate": 4.995854814260852e-05, "loss": 1.4407891273498534, "step": 14020 }, { "epoch": 1.1955688112484022, "grad_norm": 2.421875, "learning_rate": 4.9958389123742375e-05, "loss": 1.4841028213500977, "step": 14030 }, { "epoch": 1.1964209629314018, "grad_norm": 2.171875, "learning_rate": 4.995822980069745e-05, "loss": 1.412890338897705, "step": 14040 }, { "epoch": 1.1972731146144013, "grad_norm": 3.125, "learning_rate": 4.99580701734757e-05, "loss": 1.4618921279907227, "step": 14050 }, { "epoch": 1.198125266297401, "grad_norm": 1.734375, "learning_rate": 4.995791024207904e-05, "loss": 1.4450358390808105, "step": 14060 }, { "epoch": 1.1989774179804005, "grad_norm": 2.09375, "learning_rate": 4.995775000650945e-05, "loss": 1.3527077674865722, "step": 14070 }, { "epoch": 1.1998295696634, "grad_norm": 2.234375, "learning_rate": 4.995758946676887e-05, "loss": 1.4649301528930665, "step": 14080 }, { "epoch": 1.2006817213463996, "grad_norm": 2.390625, "learning_rate": 4.995742862285925e-05, "loss": 1.5150297164916993, "step": 14090 }, { "epoch": 1.2015338730293992, "grad_norm": 1.7421875, "learning_rate": 4.995726747478257e-05, "loss": 1.3958898544311524, "step": 14100 }, { "epoch": 1.2023860247123987, "grad_norm": 3.71875, "learning_rate": 4.9957106022540766e-05, "loss": 1.382300853729248, "step": 14110 }, { "epoch": 1.2032381763953983, "grad_norm": 1.5703125, "learning_rate": 4.995694426613583e-05, "loss": 1.4001520156860352, "step": 14120 }, { "epoch": 1.2040903280783979, "grad_norm": 1.6328125, "learning_rate": 4.995678220556972e-05, "loss": 1.4953664779663085, "step": 14130 }, { "epoch": 1.2049424797613975, "grad_norm": 2.21875, "learning_rate": 4.9956619840844424e-05, "loss": 1.3739463806152343, "step": 14140 }, { "epoch": 1.205794631444397, "grad_norm": 2.21875, "learning_rate": 4.99564571719619e-05, "loss": 1.3976696968078612, "step": 14150 }, { "epoch": 1.2066467831273966, "grad_norm": 2.234375, "learning_rate": 4.995629419892415e-05, "loss": 1.3068283081054688, "step": 14160 }, { "epoch": 1.2074989348103962, "grad_norm": 2.296875, "learning_rate": 4.9956130921733145e-05, "loss": 1.4145094871520996, "step": 14170 }, { "epoch": 1.2083510864933957, "grad_norm": 2.265625, "learning_rate": 4.9955967340390893e-05, "loss": 1.4580604553222656, "step": 14180 }, { "epoch": 1.2092032381763955, "grad_norm": 2.421875, "learning_rate": 4.995580345489937e-05, "loss": 1.4376260757446289, "step": 14190 }, { "epoch": 1.210055389859395, "grad_norm": 2.703125, "learning_rate": 4.995563926526059e-05, "loss": 1.4739352226257325, "step": 14200 }, { "epoch": 1.2109075415423947, "grad_norm": 4.375, "learning_rate": 4.995547477147654e-05, "loss": 1.281208610534668, "step": 14210 }, { "epoch": 1.2117596932253942, "grad_norm": 2.421875, "learning_rate": 4.9955309973549235e-05, "loss": 1.3492913246154785, "step": 14220 }, { "epoch": 1.2126118449083938, "grad_norm": 1.984375, "learning_rate": 4.995514487148067e-05, "loss": 1.4835445404052734, "step": 14230 }, { "epoch": 1.2134639965913934, "grad_norm": 2.6875, "learning_rate": 4.995497946527287e-05, "loss": 1.2628770828247071, "step": 14240 }, { "epoch": 1.214316148274393, "grad_norm": 1.8046875, "learning_rate": 4.9954813754927845e-05, "loss": 1.4680974006652832, "step": 14250 }, { "epoch": 1.2151682999573925, "grad_norm": 1.578125, "learning_rate": 4.995464774044761e-05, "loss": 1.3382376670837401, "step": 14260 }, { "epoch": 1.216020451640392, "grad_norm": 1.703125, "learning_rate": 4.9954481421834204e-05, "loss": 1.5350460052490233, "step": 14270 }, { "epoch": 1.2168726033233916, "grad_norm": 2.953125, "learning_rate": 4.995431479908964e-05, "loss": 1.280170726776123, "step": 14280 }, { "epoch": 1.2177247550063912, "grad_norm": 2.875, "learning_rate": 4.9954147872215954e-05, "loss": 1.4399465560913085, "step": 14290 }, { "epoch": 1.2185769066893908, "grad_norm": 2.453125, "learning_rate": 4.995398064121518e-05, "loss": 1.3425171852111817, "step": 14300 }, { "epoch": 1.2194290583723904, "grad_norm": 4.3125, "learning_rate": 4.995381310608935e-05, "loss": 1.2472546577453614, "step": 14310 }, { "epoch": 1.22028121005539, "grad_norm": 2.171875, "learning_rate": 4.9953645266840524e-05, "loss": 1.505054759979248, "step": 14320 }, { "epoch": 1.2211333617383895, "grad_norm": 2.15625, "learning_rate": 4.995347712347072e-05, "loss": 1.5722929000854493, "step": 14330 }, { "epoch": 1.221985513421389, "grad_norm": 2.140625, "learning_rate": 4.9953308675982006e-05, "loss": 1.3705035209655763, "step": 14340 }, { "epoch": 1.2228376651043886, "grad_norm": 2.78125, "learning_rate": 4.9953139924376435e-05, "loss": 1.4468297958374023, "step": 14350 }, { "epoch": 1.2236898167873882, "grad_norm": 1.953125, "learning_rate": 4.995297086865606e-05, "loss": 1.4836792945861816, "step": 14360 }, { "epoch": 1.2245419684703878, "grad_norm": 2.03125, "learning_rate": 4.9952801508822934e-05, "loss": 1.4194765090942383, "step": 14370 }, { "epoch": 1.2253941201533873, "grad_norm": 2.484375, "learning_rate": 4.995263184487913e-05, "loss": 1.458961009979248, "step": 14380 }, { "epoch": 1.226246271836387, "grad_norm": 1.75, "learning_rate": 4.9952461876826714e-05, "loss": 1.4324970245361328, "step": 14390 }, { "epoch": 1.2270984235193865, "grad_norm": 2.46875, "learning_rate": 4.995229160466776e-05, "loss": 1.4302024841308594, "step": 14400 }, { "epoch": 1.227950575202386, "grad_norm": 2.8125, "learning_rate": 4.9952121028404345e-05, "loss": 1.4065011024475098, "step": 14410 }, { "epoch": 1.2288027268853856, "grad_norm": 3.0625, "learning_rate": 4.995195014803853e-05, "loss": 1.3956579208374023, "step": 14420 }, { "epoch": 1.2296548785683852, "grad_norm": 2.03125, "learning_rate": 4.995177896357243e-05, "loss": 1.4208772659301758, "step": 14430 }, { "epoch": 1.2305070302513847, "grad_norm": 2.8125, "learning_rate": 4.995160747500809e-05, "loss": 1.4756738662719726, "step": 14440 }, { "epoch": 1.2313591819343843, "grad_norm": 2.59375, "learning_rate": 4.995143568234764e-05, "loss": 1.539161491394043, "step": 14450 }, { "epoch": 1.2322113336173839, "grad_norm": 2.65625, "learning_rate": 4.995126358559315e-05, "loss": 1.552965259552002, "step": 14460 }, { "epoch": 1.2330634853003835, "grad_norm": 2.03125, "learning_rate": 4.995109118474672e-05, "loss": 1.4579837799072266, "step": 14470 }, { "epoch": 1.233915636983383, "grad_norm": 2.0, "learning_rate": 4.995091847981045e-05, "loss": 1.379400634765625, "step": 14480 }, { "epoch": 1.2347677886663826, "grad_norm": 3.078125, "learning_rate": 4.995074547078646e-05, "loss": 1.4334383010864258, "step": 14490 }, { "epoch": 1.2356199403493822, "grad_norm": 2.296875, "learning_rate": 4.995057215767684e-05, "loss": 1.4035685539245606, "step": 14500 }, { "epoch": 1.2364720920323817, "grad_norm": 2.375, "learning_rate": 4.995039854048372e-05, "loss": 1.419433307647705, "step": 14510 }, { "epoch": 1.2373242437153813, "grad_norm": 1.546875, "learning_rate": 4.995022461920919e-05, "loss": 1.420161533355713, "step": 14520 }, { "epoch": 1.2381763953983809, "grad_norm": 4.1875, "learning_rate": 4.9950050393855404e-05, "loss": 1.491496467590332, "step": 14530 }, { "epoch": 1.2390285470813804, "grad_norm": 3.0625, "learning_rate": 4.994987586442446e-05, "loss": 1.3283016204833984, "step": 14540 }, { "epoch": 1.23988069876438, "grad_norm": 2.0625, "learning_rate": 4.994970103091849e-05, "loss": 1.368363380432129, "step": 14550 }, { "epoch": 1.2407328504473796, "grad_norm": 1.6640625, "learning_rate": 4.9949525893339624e-05, "loss": 1.4866290092468262, "step": 14560 }, { "epoch": 1.2415850021303791, "grad_norm": 2.09375, "learning_rate": 4.9949350451690006e-05, "loss": 1.4550883293151855, "step": 14570 }, { "epoch": 1.2424371538133787, "grad_norm": 3.546875, "learning_rate": 4.994917470597176e-05, "loss": 1.2765303611755372, "step": 14580 }, { "epoch": 1.2432893054963783, "grad_norm": 1.6875, "learning_rate": 4.994899865618704e-05, "loss": 1.4653374671936035, "step": 14590 }, { "epoch": 1.2441414571793779, "grad_norm": 2.125, "learning_rate": 4.994882230233799e-05, "loss": 1.4637617111206054, "step": 14600 }, { "epoch": 1.2449936088623774, "grad_norm": 2.71875, "learning_rate": 4.9948645644426754e-05, "loss": 1.5308634757995605, "step": 14610 }, { "epoch": 1.245845760545377, "grad_norm": 2.09375, "learning_rate": 4.9948468682455486e-05, "loss": 1.2706925392150878, "step": 14620 }, { "epoch": 1.2466979122283766, "grad_norm": 2.265625, "learning_rate": 4.994829141642634e-05, "loss": 1.4875590324401855, "step": 14630 }, { "epoch": 1.2475500639113761, "grad_norm": 3.140625, "learning_rate": 4.994811384634149e-05, "loss": 1.2995866775512694, "step": 14640 }, { "epoch": 1.2484022155943757, "grad_norm": 2.890625, "learning_rate": 4.9947935972203084e-05, "loss": 1.343162441253662, "step": 14650 }, { "epoch": 1.2492543672773753, "grad_norm": 3.703125, "learning_rate": 4.994775779401329e-05, "loss": 1.438669204711914, "step": 14660 }, { "epoch": 1.2501065189603748, "grad_norm": 1.75, "learning_rate": 4.99475793117743e-05, "loss": 1.4835090637207031, "step": 14670 }, { "epoch": 1.2509586706433744, "grad_norm": 2.40625, "learning_rate": 4.994740052548827e-05, "loss": 1.249352264404297, "step": 14680 }, { "epoch": 1.251810822326374, "grad_norm": 2.109375, "learning_rate": 4.994722143515738e-05, "loss": 1.5007636070251464, "step": 14690 }, { "epoch": 1.2526629740093735, "grad_norm": 2.0625, "learning_rate": 4.994704204078382e-05, "loss": 1.4570075035095216, "step": 14700 }, { "epoch": 1.2535151256923731, "grad_norm": 2.46875, "learning_rate": 4.994686234236977e-05, "loss": 1.285792064666748, "step": 14710 }, { "epoch": 1.2543672773753727, "grad_norm": 1.7578125, "learning_rate": 4.994668233991743e-05, "loss": 1.4713865280151368, "step": 14720 }, { "epoch": 1.2552194290583725, "grad_norm": 2.375, "learning_rate": 4.994650203342899e-05, "loss": 1.364030647277832, "step": 14730 }, { "epoch": 1.256071580741372, "grad_norm": 2.015625, "learning_rate": 4.9946321422906634e-05, "loss": 1.4608572006225586, "step": 14740 }, { "epoch": 1.2569237324243716, "grad_norm": 2.1875, "learning_rate": 4.994614050835258e-05, "loss": 1.3832966804504394, "step": 14750 }, { "epoch": 1.2577758841073712, "grad_norm": 2.015625, "learning_rate": 4.9945959289769016e-05, "loss": 1.3264688491821288, "step": 14760 }, { "epoch": 1.2586280357903707, "grad_norm": 1.921875, "learning_rate": 4.9945777767158177e-05, "loss": 1.2910219192504884, "step": 14770 }, { "epoch": 1.2594801874733703, "grad_norm": 3.4375, "learning_rate": 4.994559594052225e-05, "loss": 1.4114795684814454, "step": 14780 }, { "epoch": 1.2603323391563699, "grad_norm": 1.7734375, "learning_rate": 4.9945413809863454e-05, "loss": 1.4086143493652343, "step": 14790 }, { "epoch": 1.2611844908393695, "grad_norm": 1.71875, "learning_rate": 4.9945231375184024e-05, "loss": 1.3660019874572753, "step": 14800 }, { "epoch": 1.262036642522369, "grad_norm": 2.21875, "learning_rate": 4.994504863648617e-05, "loss": 1.4913028717041015, "step": 14810 }, { "epoch": 1.2628887942053686, "grad_norm": 2.484375, "learning_rate": 4.994486559377213e-05, "loss": 1.548952007293701, "step": 14820 }, { "epoch": 1.2637409458883682, "grad_norm": 2.4375, "learning_rate": 4.9944682247044126e-05, "loss": 1.4388511657714844, "step": 14830 }, { "epoch": 1.2645930975713677, "grad_norm": 1.8984375, "learning_rate": 4.9944498596304395e-05, "loss": 1.3854342460632325, "step": 14840 }, { "epoch": 1.2654452492543673, "grad_norm": 2.40625, "learning_rate": 4.9944314641555176e-05, "loss": 1.426045799255371, "step": 14850 }, { "epoch": 1.2662974009373669, "grad_norm": 3.546875, "learning_rate": 4.99441303827987e-05, "loss": 1.408604049682617, "step": 14860 }, { "epoch": 1.2671495526203664, "grad_norm": 1.953125, "learning_rate": 4.994394582003724e-05, "loss": 1.5381159782409668, "step": 14870 }, { "epoch": 1.268001704303366, "grad_norm": 1.9140625, "learning_rate": 4.994376095327302e-05, "loss": 1.420274829864502, "step": 14880 }, { "epoch": 1.2688538559863656, "grad_norm": 2.5, "learning_rate": 4.99435757825083e-05, "loss": 1.318889331817627, "step": 14890 }, { "epoch": 1.2697060076693651, "grad_norm": 2.53125, "learning_rate": 4.994339030774534e-05, "loss": 1.5396929740905763, "step": 14900 }, { "epoch": 1.2705581593523647, "grad_norm": 2.453125, "learning_rate": 4.99432045289864e-05, "loss": 1.314832878112793, "step": 14910 }, { "epoch": 1.2714103110353643, "grad_norm": 2.15625, "learning_rate": 4.994301844623374e-05, "loss": 1.3038832664489746, "step": 14920 }, { "epoch": 1.2722624627183639, "grad_norm": 1.640625, "learning_rate": 4.9942832059489624e-05, "loss": 1.3752878189086915, "step": 14930 }, { "epoch": 1.2731146144013634, "grad_norm": 2.125, "learning_rate": 4.994264536875634e-05, "loss": 1.3232406616210937, "step": 14940 }, { "epoch": 1.273966766084363, "grad_norm": 2.78125, "learning_rate": 4.9942458374036146e-05, "loss": 1.429083251953125, "step": 14950 }, { "epoch": 1.2748189177673626, "grad_norm": 2.515625, "learning_rate": 4.9942271075331334e-05, "loss": 1.5557870864868164, "step": 14960 }, { "epoch": 1.2756710694503621, "grad_norm": 2.625, "learning_rate": 4.994208347264417e-05, "loss": 1.503150463104248, "step": 14970 }, { "epoch": 1.2765232211333617, "grad_norm": 1.96875, "learning_rate": 4.994189556597696e-05, "loss": 1.450998592376709, "step": 14980 }, { "epoch": 1.2773753728163613, "grad_norm": 2.15625, "learning_rate": 4.994170735533199e-05, "loss": 1.2565424919128418, "step": 14990 }, { "epoch": 1.2782275244993608, "grad_norm": 3.765625, "learning_rate": 4.9941518840711546e-05, "loss": 1.379530143737793, "step": 15000 }, { "epoch": 1.2790796761823604, "grad_norm": 2.953125, "learning_rate": 4.9941330022117925e-05, "loss": 1.4088248252868651, "step": 15010 }, { "epoch": 1.27993182786536, "grad_norm": 1.921875, "learning_rate": 4.994114089955343e-05, "loss": 1.3337124824523925, "step": 15020 }, { "epoch": 1.2807839795483595, "grad_norm": 1.8671875, "learning_rate": 4.994095147302038e-05, "loss": 1.503870964050293, "step": 15030 }, { "epoch": 1.281636131231359, "grad_norm": 2.453125, "learning_rate": 4.9940761742521057e-05, "loss": 1.3265846252441407, "step": 15040 }, { "epoch": 1.2824882829143587, "grad_norm": 2.828125, "learning_rate": 4.99405717080578e-05, "loss": 1.3617762565612792, "step": 15050 }, { "epoch": 1.2833404345973582, "grad_norm": 1.796875, "learning_rate": 4.99403813696329e-05, "loss": 1.373685073852539, "step": 15060 }, { "epoch": 1.284192586280358, "grad_norm": 2.5625, "learning_rate": 4.9940190727248705e-05, "loss": 1.3941712379455566, "step": 15070 }, { "epoch": 1.2850447379633576, "grad_norm": 1.9375, "learning_rate": 4.9939999780907515e-05, "loss": 1.3488179206848145, "step": 15080 }, { "epoch": 1.2858968896463572, "grad_norm": 2.09375, "learning_rate": 4.993980853061167e-05, "loss": 1.2967330932617187, "step": 15090 }, { "epoch": 1.2867490413293567, "grad_norm": 3.25, "learning_rate": 4.993961697636349e-05, "loss": 1.3902099609375, "step": 15100 }, { "epoch": 1.2876011930123563, "grad_norm": 2.625, "learning_rate": 4.993942511816532e-05, "loss": 1.4284044265747071, "step": 15110 }, { "epoch": 1.2884533446953559, "grad_norm": 2.203125, "learning_rate": 4.993923295601949e-05, "loss": 1.2208783149719238, "step": 15120 }, { "epoch": 1.2893054963783555, "grad_norm": 2.171875, "learning_rate": 4.993904048992834e-05, "loss": 1.393751335144043, "step": 15130 }, { "epoch": 1.290157648061355, "grad_norm": 3.0, "learning_rate": 4.9938847719894236e-05, "loss": 1.2751073837280273, "step": 15140 }, { "epoch": 1.2910097997443546, "grad_norm": 3.296875, "learning_rate": 4.99386546459195e-05, "loss": 1.5040460586547852, "step": 15150 }, { "epoch": 1.2918619514273542, "grad_norm": 2.53125, "learning_rate": 4.9938461268006506e-05, "loss": 1.355534553527832, "step": 15160 }, { "epoch": 1.2927141031103537, "grad_norm": 2.09375, "learning_rate": 4.993826758615759e-05, "loss": 1.2993825912475585, "step": 15170 }, { "epoch": 1.2935662547933533, "grad_norm": 2.46875, "learning_rate": 4.9938073600375145e-05, "loss": 1.4516575813293457, "step": 15180 }, { "epoch": 1.2944184064763529, "grad_norm": 3.625, "learning_rate": 4.993787931066151e-05, "loss": 1.265909767150879, "step": 15190 }, { "epoch": 1.2952705581593524, "grad_norm": 2.296875, "learning_rate": 4.993768471701905e-05, "loss": 1.3825490951538086, "step": 15200 }, { "epoch": 1.296122709842352, "grad_norm": 1.890625, "learning_rate": 4.9937489819450143e-05, "loss": 1.4999930381774902, "step": 15210 }, { "epoch": 1.2969748615253516, "grad_norm": 2.765625, "learning_rate": 4.993729461795717e-05, "loss": 1.507755661010742, "step": 15220 }, { "epoch": 1.2978270132083511, "grad_norm": 1.6171875, "learning_rate": 4.993709911254252e-05, "loss": 1.438334083557129, "step": 15230 }, { "epoch": 1.2986791648913507, "grad_norm": 2.609375, "learning_rate": 4.993690330320855e-05, "loss": 1.393260383605957, "step": 15240 }, { "epoch": 1.2995313165743503, "grad_norm": 3.125, "learning_rate": 4.9936707189957653e-05, "loss": 1.44637451171875, "step": 15250 }, { "epoch": 1.3003834682573499, "grad_norm": 1.9765625, "learning_rate": 4.993651077279223e-05, "loss": 1.448916244506836, "step": 15260 }, { "epoch": 1.3012356199403494, "grad_norm": 1.96875, "learning_rate": 4.993631405171467e-05, "loss": 1.2978431701660156, "step": 15270 }, { "epoch": 1.302087771623349, "grad_norm": 2.1875, "learning_rate": 4.993611702672737e-05, "loss": 1.348383903503418, "step": 15280 }, { "epoch": 1.3029399233063486, "grad_norm": 3.234375, "learning_rate": 4.993591969783273e-05, "loss": 1.3325413703918456, "step": 15290 }, { "epoch": 1.3037920749893481, "grad_norm": 1.8671875, "learning_rate": 4.993572206503316e-05, "loss": 1.3821735382080078, "step": 15300 }, { "epoch": 1.3046442266723477, "grad_norm": 2.25, "learning_rate": 4.993552412833106e-05, "loss": 1.4196990966796874, "step": 15310 }, { "epoch": 1.3054963783553473, "grad_norm": 2.53125, "learning_rate": 4.9935325887728846e-05, "loss": 1.5473015785217286, "step": 15320 }, { "epoch": 1.3063485300383468, "grad_norm": 1.890625, "learning_rate": 4.993512734322894e-05, "loss": 1.4198187828063964, "step": 15330 }, { "epoch": 1.3072006817213464, "grad_norm": 2.34375, "learning_rate": 4.993492849483375e-05, "loss": 1.3523887634277343, "step": 15340 }, { "epoch": 1.308052833404346, "grad_norm": 3.03125, "learning_rate": 4.9934729342545715e-05, "loss": 1.4628597259521485, "step": 15350 }, { "epoch": 1.3089049850873455, "grad_norm": 2.96875, "learning_rate": 4.993452988636724e-05, "loss": 1.3678730010986329, "step": 15360 }, { "epoch": 1.309757136770345, "grad_norm": 2.59375, "learning_rate": 4.993433012630078e-05, "loss": 1.3444419860839845, "step": 15370 }, { "epoch": 1.3106092884533447, "grad_norm": 2.484375, "learning_rate": 4.993413006234876e-05, "loss": 1.3101266860961913, "step": 15380 }, { "epoch": 1.3114614401363442, "grad_norm": 2.21875, "learning_rate": 4.993392969451361e-05, "loss": 1.4063886642456054, "step": 15390 }, { "epoch": 1.3123135918193438, "grad_norm": 2.375, "learning_rate": 4.993372902279778e-05, "loss": 1.4319780349731446, "step": 15400 }, { "epoch": 1.3131657435023434, "grad_norm": 1.7734375, "learning_rate": 4.9933528047203725e-05, "loss": 1.4661645889282227, "step": 15410 }, { "epoch": 1.314017895185343, "grad_norm": 2.109375, "learning_rate": 4.9933326767733875e-05, "loss": 1.372169017791748, "step": 15420 }, { "epoch": 1.3148700468683425, "grad_norm": 2.59375, "learning_rate": 4.99331251843907e-05, "loss": 1.4009513854980469, "step": 15430 }, { "epoch": 1.315722198551342, "grad_norm": 3.15625, "learning_rate": 4.993292329717664e-05, "loss": 1.4361202239990234, "step": 15440 }, { "epoch": 1.3165743502343417, "grad_norm": 2.453125, "learning_rate": 4.993272110609416e-05, "loss": 1.293106746673584, "step": 15450 }, { "epoch": 1.3174265019173412, "grad_norm": 2.84375, "learning_rate": 4.993251861114574e-05, "loss": 1.4290032386779785, "step": 15460 }, { "epoch": 1.3182786536003408, "grad_norm": 3.421875, "learning_rate": 4.9932315812333827e-05, "loss": 1.4665059089660644, "step": 15470 }, { "epoch": 1.3191308052833404, "grad_norm": 2.328125, "learning_rate": 4.99321127096609e-05, "loss": 1.3426048278808593, "step": 15480 }, { "epoch": 1.31998295696634, "grad_norm": 2.109375, "learning_rate": 4.9931909303129456e-05, "loss": 1.2971206665039063, "step": 15490 }, { "epoch": 1.3208351086493395, "grad_norm": 4.71875, "learning_rate": 4.993170559274194e-05, "loss": 1.1577677726745605, "step": 15500 }, { "epoch": 1.321687260332339, "grad_norm": 2.65625, "learning_rate": 4.993150157850085e-05, "loss": 1.4398059844970703, "step": 15510 }, { "epoch": 1.3225394120153386, "grad_norm": 1.8671875, "learning_rate": 4.993129726040867e-05, "loss": 1.448354148864746, "step": 15520 }, { "epoch": 1.3233915636983382, "grad_norm": 2.078125, "learning_rate": 4.9931092638467894e-05, "loss": 1.3897981643676758, "step": 15530 }, { "epoch": 1.3242437153813378, "grad_norm": 2.15625, "learning_rate": 4.993088771268102e-05, "loss": 1.3443238258361816, "step": 15540 }, { "epoch": 1.3250958670643374, "grad_norm": 2.375, "learning_rate": 4.993068248305053e-05, "loss": 1.3508541107177734, "step": 15550 }, { "epoch": 1.325948018747337, "grad_norm": 3.015625, "learning_rate": 4.993047694957893e-05, "loss": 1.2573088645935058, "step": 15560 }, { "epoch": 1.3268001704303365, "grad_norm": 2.1875, "learning_rate": 4.993027111226874e-05, "loss": 1.4446329116821288, "step": 15570 }, { "epoch": 1.327652322113336, "grad_norm": 1.984375, "learning_rate": 4.9930064971122454e-05, "loss": 1.3804923057556153, "step": 15580 }, { "epoch": 1.3285044737963356, "grad_norm": 2.4375, "learning_rate": 4.992985852614259e-05, "loss": 1.431836986541748, "step": 15590 }, { "epoch": 1.3293566254793352, "grad_norm": 2.765625, "learning_rate": 4.992965177733165e-05, "loss": 1.44126558303833, "step": 15600 }, { "epoch": 1.3302087771623348, "grad_norm": 2.484375, "learning_rate": 4.992944472469218e-05, "loss": 1.4906188011169434, "step": 15610 }, { "epoch": 1.3310609288453343, "grad_norm": 2.0625, "learning_rate": 4.992923736822669e-05, "loss": 1.4556781768798828, "step": 15620 }, { "epoch": 1.3319130805283341, "grad_norm": 1.7890625, "learning_rate": 4.9929029707937705e-05, "loss": 1.3230805397033691, "step": 15630 }, { "epoch": 1.3327652322113337, "grad_norm": 1.7578125, "learning_rate": 4.992882174382775e-05, "loss": 1.2990524291992187, "step": 15640 }, { "epoch": 1.3336173838943333, "grad_norm": 2.03125, "learning_rate": 4.992861347589937e-05, "loss": 1.3139657020568847, "step": 15650 }, { "epoch": 1.3344695355773328, "grad_norm": 1.7265625, "learning_rate": 4.99284049041551e-05, "loss": 1.4401927947998048, "step": 15660 }, { "epoch": 1.3353216872603324, "grad_norm": 2.546875, "learning_rate": 4.992819602859748e-05, "loss": 1.3590234756469726, "step": 15670 }, { "epoch": 1.336173838943332, "grad_norm": 4.40625, "learning_rate": 4.9927986849229055e-05, "loss": 1.4534802436828613, "step": 15680 }, { "epoch": 1.3370259906263315, "grad_norm": 2.03125, "learning_rate": 4.992777736605239e-05, "loss": 1.4351642608642579, "step": 15690 }, { "epoch": 1.337878142309331, "grad_norm": 2.25, "learning_rate": 4.992756757907002e-05, "loss": 1.371278667449951, "step": 15700 }, { "epoch": 1.3387302939923307, "grad_norm": 4.03125, "learning_rate": 4.99273574882845e-05, "loss": 1.4183259963989259, "step": 15710 }, { "epoch": 1.3395824456753302, "grad_norm": 2.25, "learning_rate": 4.99271470936984e-05, "loss": 1.4263277053833008, "step": 15720 }, { "epoch": 1.3404345973583298, "grad_norm": 2.5, "learning_rate": 4.9926936395314284e-05, "loss": 1.4635753631591797, "step": 15730 }, { "epoch": 1.3412867490413294, "grad_norm": 2.34375, "learning_rate": 4.992672539313472e-05, "loss": 1.397618865966797, "step": 15740 }, { "epoch": 1.342138900724329, "grad_norm": 3.75, "learning_rate": 4.9926514087162274e-05, "loss": 1.3473658561706543, "step": 15750 }, { "epoch": 1.3429910524073285, "grad_norm": 2.234375, "learning_rate": 4.992630247739952e-05, "loss": 1.4781694412231445, "step": 15760 }, { "epoch": 1.343843204090328, "grad_norm": 1.5078125, "learning_rate": 4.9926090563849046e-05, "loss": 1.5523293495178223, "step": 15770 }, { "epoch": 1.3446953557733277, "grad_norm": 2.328125, "learning_rate": 4.9925878346513436e-05, "loss": 1.4678287506103516, "step": 15780 }, { "epoch": 1.3455475074563272, "grad_norm": 2.0625, "learning_rate": 4.992566582539526e-05, "loss": 1.3307146072387694, "step": 15790 }, { "epoch": 1.3463996591393268, "grad_norm": 1.953125, "learning_rate": 4.992545300049713e-05, "loss": 1.2692235946655273, "step": 15800 }, { "epoch": 1.3472518108223264, "grad_norm": 2.296875, "learning_rate": 4.992523987182161e-05, "loss": 1.3778470993041991, "step": 15810 }, { "epoch": 1.348103962505326, "grad_norm": 1.890625, "learning_rate": 4.992502643937132e-05, "loss": 1.342171859741211, "step": 15820 }, { "epoch": 1.3489561141883255, "grad_norm": 1.828125, "learning_rate": 4.992481270314887e-05, "loss": 1.2687410354614257, "step": 15830 }, { "epoch": 1.349808265871325, "grad_norm": 3.0, "learning_rate": 4.992459866315684e-05, "loss": 1.3367195129394531, "step": 15840 }, { "epoch": 1.3506604175543246, "grad_norm": 2.109375, "learning_rate": 4.992438431939786e-05, "loss": 1.443742275238037, "step": 15850 }, { "epoch": 1.3515125692373242, "grad_norm": 3.203125, "learning_rate": 4.9924169671874524e-05, "loss": 1.2947797775268555, "step": 15860 }, { "epoch": 1.3523647209203238, "grad_norm": 2.171875, "learning_rate": 4.992395472058946e-05, "loss": 1.3448604583740233, "step": 15870 }, { "epoch": 1.3532168726033234, "grad_norm": 2.75, "learning_rate": 4.992373946554529e-05, "loss": 1.5040535926818848, "step": 15880 }, { "epoch": 1.354069024286323, "grad_norm": 1.8671875, "learning_rate": 4.992352390674462e-05, "loss": 1.3975006103515626, "step": 15890 }, { "epoch": 1.3549211759693225, "grad_norm": 1.8203125, "learning_rate": 4.99233080441901e-05, "loss": 1.3388402938842774, "step": 15900 }, { "epoch": 1.355773327652322, "grad_norm": 3.078125, "learning_rate": 4.9923091877884345e-05, "loss": 1.4356613159179688, "step": 15910 }, { "epoch": 1.3566254793353216, "grad_norm": 2.734375, "learning_rate": 4.9922875407829994e-05, "loss": 1.3996021270751953, "step": 15920 }, { "epoch": 1.3574776310183212, "grad_norm": 3.171875, "learning_rate": 4.992265863402968e-05, "loss": 1.3354197502136231, "step": 15930 }, { "epoch": 1.3583297827013208, "grad_norm": 2.109375, "learning_rate": 4.992244155648606e-05, "loss": 1.5402765274047852, "step": 15940 }, { "epoch": 1.3591819343843203, "grad_norm": 3.75, "learning_rate": 4.9922224175201765e-05, "loss": 1.338680362701416, "step": 15950 }, { "epoch": 1.36003408606732, "grad_norm": 2.265625, "learning_rate": 4.9922006490179454e-05, "loss": 1.4441668510437011, "step": 15960 }, { "epoch": 1.3608862377503197, "grad_norm": 2.09375, "learning_rate": 4.992178850142177e-05, "loss": 1.4396280288696288, "step": 15970 }, { "epoch": 1.3617383894333193, "grad_norm": 2.0625, "learning_rate": 4.992157020893138e-05, "loss": 1.482408618927002, "step": 15980 }, { "epoch": 1.3625905411163188, "grad_norm": 2.75, "learning_rate": 4.992135161271093e-05, "loss": 1.4210428237915038, "step": 15990 }, { "epoch": 1.3634426927993184, "grad_norm": 2.0625, "learning_rate": 4.99211327127631e-05, "loss": 1.4267614364624024, "step": 16000 }, { "epoch": 1.364294844482318, "grad_norm": 2.765625, "learning_rate": 4.992091350909055e-05, "loss": 1.4225201606750488, "step": 16010 }, { "epoch": 1.3651469961653175, "grad_norm": 2.359375, "learning_rate": 4.992069400169596e-05, "loss": 1.2785805702209472, "step": 16020 }, { "epoch": 1.365999147848317, "grad_norm": 2.40625, "learning_rate": 4.992047419058199e-05, "loss": 1.460758590698242, "step": 16030 }, { "epoch": 1.3668512995313167, "grad_norm": 2.421875, "learning_rate": 4.992025407575133e-05, "loss": 1.5413223266601563, "step": 16040 }, { "epoch": 1.3677034512143162, "grad_norm": 2.015625, "learning_rate": 4.992003365720666e-05, "loss": 1.3193222999572753, "step": 16050 }, { "epoch": 1.3685556028973158, "grad_norm": 2.890625, "learning_rate": 4.991981293495066e-05, "loss": 1.2139224052429198, "step": 16060 }, { "epoch": 1.3694077545803154, "grad_norm": 2.625, "learning_rate": 4.991959190898603e-05, "loss": 1.2472307205200195, "step": 16070 }, { "epoch": 1.370259906263315, "grad_norm": 2.234375, "learning_rate": 4.991937057931546e-05, "loss": 1.4707404136657716, "step": 16080 }, { "epoch": 1.3711120579463145, "grad_norm": 2.21875, "learning_rate": 4.9919148945941654e-05, "loss": 1.3740017890930176, "step": 16090 }, { "epoch": 1.371964209629314, "grad_norm": 2.21875, "learning_rate": 4.99189270088673e-05, "loss": 1.3063986778259278, "step": 16100 }, { "epoch": 1.3728163613123137, "grad_norm": 1.9453125, "learning_rate": 4.991870476809511e-05, "loss": 1.4068583488464355, "step": 16110 }, { "epoch": 1.3736685129953132, "grad_norm": 2.390625, "learning_rate": 4.9918482223627795e-05, "loss": 1.4492868423461913, "step": 16120 }, { "epoch": 1.3745206646783128, "grad_norm": 2.5, "learning_rate": 4.9918259375468056e-05, "loss": 1.367579746246338, "step": 16130 }, { "epoch": 1.3753728163613124, "grad_norm": 3.5, "learning_rate": 4.9918036223618627e-05, "loss": 1.3593541145324708, "step": 16140 }, { "epoch": 1.376224968044312, "grad_norm": 2.125, "learning_rate": 4.9917812768082203e-05, "loss": 1.4378532409667968, "step": 16150 }, { "epoch": 1.3770771197273115, "grad_norm": 3.90625, "learning_rate": 4.991758900886154e-05, "loss": 1.424837017059326, "step": 16160 }, { "epoch": 1.377929271410311, "grad_norm": 2.0625, "learning_rate": 4.991736494595934e-05, "loss": 1.3699580192565919, "step": 16170 }, { "epoch": 1.3787814230933106, "grad_norm": 3.859375, "learning_rate": 4.991714057937834e-05, "loss": 1.312342071533203, "step": 16180 }, { "epoch": 1.3796335747763102, "grad_norm": 1.6796875, "learning_rate": 4.991691590912127e-05, "loss": 1.3655258178710938, "step": 16190 }, { "epoch": 1.3804857264593098, "grad_norm": 2.4375, "learning_rate": 4.991669093519088e-05, "loss": 1.4168981552124023, "step": 16200 }, { "epoch": 1.3813378781423093, "grad_norm": 2.53125, "learning_rate": 4.991646565758991e-05, "loss": 1.4316397666931153, "step": 16210 }, { "epoch": 1.382190029825309, "grad_norm": 2.765625, "learning_rate": 4.991624007632109e-05, "loss": 1.5553561210632325, "step": 16220 }, { "epoch": 1.3830421815083085, "grad_norm": 2.71875, "learning_rate": 4.991601419138718e-05, "loss": 1.4526914596557616, "step": 16230 }, { "epoch": 1.383894333191308, "grad_norm": 2.234375, "learning_rate": 4.991578800279094e-05, "loss": 1.354128646850586, "step": 16240 }, { "epoch": 1.3847464848743076, "grad_norm": 2.1875, "learning_rate": 4.9915561510535116e-05, "loss": 1.4557415962219238, "step": 16250 }, { "epoch": 1.3855986365573072, "grad_norm": 1.9296875, "learning_rate": 4.991533471462247e-05, "loss": 1.5529701232910156, "step": 16260 }, { "epoch": 1.3864507882403068, "grad_norm": 2.171875, "learning_rate": 4.991510761505578e-05, "loss": 1.3065251350402831, "step": 16270 }, { "epoch": 1.3873029399233063, "grad_norm": 2.921875, "learning_rate": 4.991488021183779e-05, "loss": 1.5089991569519043, "step": 16280 }, { "epoch": 1.388155091606306, "grad_norm": 3.984375, "learning_rate": 4.9914652504971296e-05, "loss": 1.5430168151855468, "step": 16290 }, { "epoch": 1.3890072432893055, "grad_norm": 2.515625, "learning_rate": 4.9914424494459044e-05, "loss": 1.3277385711669922, "step": 16300 }, { "epoch": 1.389859394972305, "grad_norm": 2.15625, "learning_rate": 4.991419618030384e-05, "loss": 1.5390423774719237, "step": 16310 }, { "epoch": 1.3907115466553046, "grad_norm": 2.984375, "learning_rate": 4.9913967562508454e-05, "loss": 1.4054476737976074, "step": 16320 }, { "epoch": 1.3915636983383042, "grad_norm": 2.0625, "learning_rate": 4.9913738641075667e-05, "loss": 1.5128327369689942, "step": 16330 }, { "epoch": 1.3924158500213037, "grad_norm": 3.703125, "learning_rate": 4.9913509416008286e-05, "loss": 1.3690809249877929, "step": 16340 }, { "epoch": 1.3932680017043033, "grad_norm": 4.0, "learning_rate": 4.99132798873091e-05, "loss": 1.212912368774414, "step": 16350 }, { "epoch": 1.3941201533873029, "grad_norm": 3.75, "learning_rate": 4.991305005498088e-05, "loss": 1.4211152076721192, "step": 16360 }, { "epoch": 1.3949723050703025, "grad_norm": 1.7265625, "learning_rate": 4.9912819919026465e-05, "loss": 1.3874286651611327, "step": 16370 }, { "epoch": 1.395824456753302, "grad_norm": 2.140625, "learning_rate": 4.991258947944864e-05, "loss": 1.4531190872192383, "step": 16380 }, { "epoch": 1.3966766084363016, "grad_norm": 3.015625, "learning_rate": 4.991235873625021e-05, "loss": 1.3157031059265136, "step": 16390 }, { "epoch": 1.3975287601193012, "grad_norm": 1.984375, "learning_rate": 4.991212768943401e-05, "loss": 1.3216134071350099, "step": 16400 }, { "epoch": 1.3983809118023007, "grad_norm": 1.9453125, "learning_rate": 4.991189633900282e-05, "loss": 1.4260393142700196, "step": 16410 }, { "epoch": 1.3992330634853003, "grad_norm": 1.90625, "learning_rate": 4.9911664684959495e-05, "loss": 1.3804384231567384, "step": 16420 }, { "epoch": 1.4000852151682999, "grad_norm": 2.515625, "learning_rate": 4.991143272730683e-05, "loss": 1.4018086433410644, "step": 16430 }, { "epoch": 1.4009373668512994, "grad_norm": 2.5, "learning_rate": 4.991120046604767e-05, "loss": 1.4406709671020508, "step": 16440 }, { "epoch": 1.401789518534299, "grad_norm": 4.09375, "learning_rate": 4.991096790118484e-05, "loss": 1.3057876586914063, "step": 16450 }, { "epoch": 1.4026416702172986, "grad_norm": 1.9296875, "learning_rate": 4.991073503272117e-05, "loss": 1.3850075721740722, "step": 16460 }, { "epoch": 1.4034938219002981, "grad_norm": 3.109375, "learning_rate": 4.991050186065951e-05, "loss": 1.3626832008361816, "step": 16470 }, { "epoch": 1.4043459735832977, "grad_norm": 2.484375, "learning_rate": 4.9910268385002686e-05, "loss": 1.3262378692626953, "step": 16480 }, { "epoch": 1.4051981252662973, "grad_norm": 2.640625, "learning_rate": 4.991003460575357e-05, "loss": 1.382601833343506, "step": 16490 }, { "epoch": 1.4060502769492969, "grad_norm": 2.15625, "learning_rate": 4.990980052291497e-05, "loss": 1.3472611427307128, "step": 16500 }, { "epoch": 1.4069024286322964, "grad_norm": 2.125, "learning_rate": 4.990956613648977e-05, "loss": 1.4435226440429687, "step": 16510 }, { "epoch": 1.407754580315296, "grad_norm": 2.390625, "learning_rate": 4.990933144648082e-05, "loss": 1.3799859046936036, "step": 16520 }, { "epoch": 1.4086067319982956, "grad_norm": 3.0, "learning_rate": 4.990909645289098e-05, "loss": 1.3395233154296875, "step": 16530 }, { "epoch": 1.4094588836812953, "grad_norm": 3.078125, "learning_rate": 4.990886115572311e-05, "loss": 1.479389762878418, "step": 16540 }, { "epoch": 1.410311035364295, "grad_norm": 2.46875, "learning_rate": 4.990862555498008e-05, "loss": 1.412367343902588, "step": 16550 }, { "epoch": 1.4111631870472945, "grad_norm": 2.109375, "learning_rate": 4.9908389650664765e-05, "loss": 1.292308235168457, "step": 16560 }, { "epoch": 1.412015338730294, "grad_norm": 1.9609375, "learning_rate": 4.990815344278004e-05, "loss": 1.4566611289978026, "step": 16570 }, { "epoch": 1.4128674904132936, "grad_norm": 2.203125, "learning_rate": 4.9907916931328766e-05, "loss": 1.4230433464050294, "step": 16580 }, { "epoch": 1.4137196420962932, "grad_norm": 4.625, "learning_rate": 4.990768011631385e-05, "loss": 1.3082847595214844, "step": 16590 }, { "epoch": 1.4145717937792928, "grad_norm": 2.65625, "learning_rate": 4.990744299773817e-05, "loss": 1.4198361396789552, "step": 16600 }, { "epoch": 1.4154239454622923, "grad_norm": 1.7890625, "learning_rate": 4.990720557560461e-05, "loss": 1.483175563812256, "step": 16610 }, { "epoch": 1.416276097145292, "grad_norm": 2.015625, "learning_rate": 4.990696784991606e-05, "loss": 1.541404914855957, "step": 16620 }, { "epoch": 1.4171282488282915, "grad_norm": 2.171875, "learning_rate": 4.9906729820675444e-05, "loss": 1.2857479095458983, "step": 16630 }, { "epoch": 1.417980400511291, "grad_norm": 3.890625, "learning_rate": 4.9906491487885634e-05, "loss": 1.468898391723633, "step": 16640 }, { "epoch": 1.4188325521942906, "grad_norm": 1.9296875, "learning_rate": 4.990625285154954e-05, "loss": 1.470689582824707, "step": 16650 }, { "epoch": 1.4196847038772902, "grad_norm": 2.21875, "learning_rate": 4.9906013911670074e-05, "loss": 1.4547731399536132, "step": 16660 }, { "epoch": 1.4205368555602897, "grad_norm": 1.7578125, "learning_rate": 4.990577466825015e-05, "loss": 1.3426969528198243, "step": 16670 }, { "epoch": 1.4213890072432893, "grad_norm": 3.0, "learning_rate": 4.990553512129269e-05, "loss": 1.5220810890197753, "step": 16680 }, { "epoch": 1.4222411589262889, "grad_norm": 1.8046875, "learning_rate": 4.99052952708006e-05, "loss": 1.4956037521362304, "step": 16690 }, { "epoch": 1.4230933106092885, "grad_norm": 2.375, "learning_rate": 4.990505511677681e-05, "loss": 1.372761344909668, "step": 16700 }, { "epoch": 1.423945462292288, "grad_norm": 1.9375, "learning_rate": 4.990481465922425e-05, "loss": 1.3953585624694824, "step": 16710 }, { "epoch": 1.4247976139752876, "grad_norm": 2.640625, "learning_rate": 4.9904573898145845e-05, "loss": 1.3488397598266602, "step": 16720 }, { "epoch": 1.4256497656582872, "grad_norm": 1.9453125, "learning_rate": 4.990433283354453e-05, "loss": 1.4021012306213378, "step": 16730 }, { "epoch": 1.4265019173412867, "grad_norm": 1.9375, "learning_rate": 4.990409146542324e-05, "loss": 1.4549852371215821, "step": 16740 }, { "epoch": 1.4273540690242863, "grad_norm": 3.0, "learning_rate": 4.990384979378492e-05, "loss": 1.3451308250427245, "step": 16750 }, { "epoch": 1.4282062207072859, "grad_norm": 2.171875, "learning_rate": 4.990360781863252e-05, "loss": 1.4056385993957519, "step": 16760 }, { "epoch": 1.4290583723902854, "grad_norm": 1.8515625, "learning_rate": 4.990336553996899e-05, "loss": 1.5739672660827637, "step": 16770 }, { "epoch": 1.429910524073285, "grad_norm": 2.53125, "learning_rate": 4.9903122957797264e-05, "loss": 1.3828129768371582, "step": 16780 }, { "epoch": 1.4307626757562846, "grad_norm": 3.09375, "learning_rate": 4.990288007212033e-05, "loss": 1.4251893997192382, "step": 16790 }, { "epoch": 1.4316148274392841, "grad_norm": 3.59375, "learning_rate": 4.990263688294112e-05, "loss": 1.343691062927246, "step": 16800 }, { "epoch": 1.4324669791222837, "grad_norm": 3.890625, "learning_rate": 4.99023933902626e-05, "loss": 1.1830607414245606, "step": 16810 }, { "epoch": 1.4333191308052833, "grad_norm": 1.9453125, "learning_rate": 4.990214959408776e-05, "loss": 1.435300636291504, "step": 16820 }, { "epoch": 1.4341712824882829, "grad_norm": 2.421875, "learning_rate": 4.990190549441955e-05, "loss": 1.3295042037963867, "step": 16830 }, { "epoch": 1.4350234341712824, "grad_norm": 4.03125, "learning_rate": 4.990166109126095e-05, "loss": 1.3204834938049317, "step": 16840 }, { "epoch": 1.435875585854282, "grad_norm": 2.1875, "learning_rate": 4.9901416384614954e-05, "loss": 1.4136392593383789, "step": 16850 }, { "epoch": 1.4367277375372816, "grad_norm": 2.171875, "learning_rate": 4.990117137448452e-05, "loss": 1.3255990028381348, "step": 16860 }, { "epoch": 1.4375798892202813, "grad_norm": 1.515625, "learning_rate": 4.9900926060872646e-05, "loss": 1.5683833122253419, "step": 16870 }, { "epoch": 1.438432040903281, "grad_norm": 2.46875, "learning_rate": 4.990068044378232e-05, "loss": 1.5142576217651367, "step": 16880 }, { "epoch": 1.4392841925862805, "grad_norm": 3.875, "learning_rate": 4.9900434523216535e-05, "loss": 1.3833659172058106, "step": 16890 }, { "epoch": 1.44013634426928, "grad_norm": 2.765625, "learning_rate": 4.990018829917829e-05, "loss": 1.3520112991333009, "step": 16900 }, { "epoch": 1.4409884959522796, "grad_norm": 1.5703125, "learning_rate": 4.989994177167059e-05, "loss": 1.5121335983276367, "step": 16910 }, { "epoch": 1.4418406476352792, "grad_norm": 1.71875, "learning_rate": 4.989969494069643e-05, "loss": 1.3668981552124024, "step": 16920 }, { "epoch": 1.4426927993182788, "grad_norm": 3.078125, "learning_rate": 4.989944780625883e-05, "loss": 1.3830347061157227, "step": 16930 }, { "epoch": 1.4435449510012783, "grad_norm": 2.25, "learning_rate": 4.9899200368360776e-05, "loss": 1.3925317764282226, "step": 16940 }, { "epoch": 1.444397102684278, "grad_norm": 1.7265625, "learning_rate": 4.989895262700532e-05, "loss": 1.3872329711914062, "step": 16950 }, { "epoch": 1.4452492543672775, "grad_norm": 2.359375, "learning_rate": 4.989870458219546e-05, "loss": 1.4258509635925294, "step": 16960 }, { "epoch": 1.446101406050277, "grad_norm": 1.7734375, "learning_rate": 4.989845623393422e-05, "loss": 1.3485036849975587, "step": 16970 }, { "epoch": 1.4469535577332766, "grad_norm": 4.09375, "learning_rate": 4.989820758222463e-05, "loss": 1.2973434448242187, "step": 16980 }, { "epoch": 1.4478057094162762, "grad_norm": 2.609375, "learning_rate": 4.989795862706973e-05, "loss": 1.3178651809692383, "step": 16990 }, { "epoch": 1.4486578610992757, "grad_norm": 2.1875, "learning_rate": 4.9897709368472525e-05, "loss": 1.3820991516113281, "step": 17000 }, { "epoch": 1.4495100127822753, "grad_norm": 2.09375, "learning_rate": 4.9897459806436084e-05, "loss": 1.3530325889587402, "step": 17010 }, { "epoch": 1.4503621644652749, "grad_norm": 2.265625, "learning_rate": 4.989720994096344e-05, "loss": 1.3475332260131836, "step": 17020 }, { "epoch": 1.4512143161482745, "grad_norm": 2.09375, "learning_rate": 4.989695977205763e-05, "loss": 1.521423625946045, "step": 17030 }, { "epoch": 1.452066467831274, "grad_norm": 2.375, "learning_rate": 4.98967092997217e-05, "loss": 1.4502968788146973, "step": 17040 }, { "epoch": 1.4529186195142736, "grad_norm": 1.9140625, "learning_rate": 4.989645852395872e-05, "loss": 1.296186351776123, "step": 17050 }, { "epoch": 1.4537707711972732, "grad_norm": 1.9140625, "learning_rate": 4.989620744477174e-05, "loss": 1.34154052734375, "step": 17060 }, { "epoch": 1.4546229228802727, "grad_norm": 2.078125, "learning_rate": 4.9895956062163807e-05, "loss": 1.3665245056152344, "step": 17070 }, { "epoch": 1.4554750745632723, "grad_norm": 2.421875, "learning_rate": 4.9895704376138e-05, "loss": 1.2940598487854005, "step": 17080 }, { "epoch": 1.4563272262462719, "grad_norm": 2.015625, "learning_rate": 4.989545238669737e-05, "loss": 1.4055763244628907, "step": 17090 }, { "epoch": 1.4571793779292714, "grad_norm": 2.828125, "learning_rate": 4.989520009384501e-05, "loss": 1.4400612831115722, "step": 17100 }, { "epoch": 1.458031529612271, "grad_norm": 2.859375, "learning_rate": 4.989494749758398e-05, "loss": 1.4512674331665039, "step": 17110 }, { "epoch": 1.4588836812952706, "grad_norm": 2.890625, "learning_rate": 4.989469459791737e-05, "loss": 1.4104082107543945, "step": 17120 }, { "epoch": 1.4597358329782701, "grad_norm": 3.0625, "learning_rate": 4.9894441394848244e-05, "loss": 1.3636759757995605, "step": 17130 }, { "epoch": 1.4605879846612697, "grad_norm": 3.3125, "learning_rate": 4.9894187888379694e-05, "loss": 1.495419979095459, "step": 17140 }, { "epoch": 1.4614401363442693, "grad_norm": 2.28125, "learning_rate": 4.9893934078514824e-05, "loss": 1.3830864906311036, "step": 17150 }, { "epoch": 1.4622922880272688, "grad_norm": 1.6328125, "learning_rate": 4.989367996525671e-05, "loss": 1.5467652320861816, "step": 17160 }, { "epoch": 1.4631444397102684, "grad_norm": 2.609375, "learning_rate": 4.989342554860846e-05, "loss": 1.3091132164001464, "step": 17170 }, { "epoch": 1.463996591393268, "grad_norm": 2.109375, "learning_rate": 4.9893170828573175e-05, "loss": 1.2273611068725585, "step": 17180 }, { "epoch": 1.4648487430762676, "grad_norm": 2.28125, "learning_rate": 4.989291580515395e-05, "loss": 1.367753791809082, "step": 17190 }, { "epoch": 1.4657008947592671, "grad_norm": 2.25, "learning_rate": 4.98926604783539e-05, "loss": 1.4065662384033204, "step": 17200 }, { "epoch": 1.4665530464422667, "grad_norm": 1.6796875, "learning_rate": 4.989240484817613e-05, "loss": 1.3746685981750488, "step": 17210 }, { "epoch": 1.4674051981252663, "grad_norm": 2.671875, "learning_rate": 4.989214891462377e-05, "loss": 1.480102825164795, "step": 17220 }, { "epoch": 1.4682573498082658, "grad_norm": 2.265625, "learning_rate": 4.989189267769993e-05, "loss": 1.3731407165527343, "step": 17230 }, { "epoch": 1.4691095014912654, "grad_norm": 3.296875, "learning_rate": 4.989163613740772e-05, "loss": 1.2703124046325684, "step": 17240 }, { "epoch": 1.469961653174265, "grad_norm": 2.40625, "learning_rate": 4.9891379293750285e-05, "loss": 1.4610873222351075, "step": 17250 }, { "epoch": 1.4708138048572645, "grad_norm": 1.40625, "learning_rate": 4.989112214673075e-05, "loss": 1.4568248748779298, "step": 17260 }, { "epoch": 1.471665956540264, "grad_norm": 3.171875, "learning_rate": 4.9890864696352246e-05, "loss": 1.4603914260864257, "step": 17270 }, { "epoch": 1.4725181082232637, "grad_norm": 1.6875, "learning_rate": 4.989060694261792e-05, "loss": 1.3696999549865723, "step": 17280 }, { "epoch": 1.4733702599062632, "grad_norm": 2.15625, "learning_rate": 4.98903488855309e-05, "loss": 1.4126543045043944, "step": 17290 }, { "epoch": 1.4742224115892628, "grad_norm": 1.765625, "learning_rate": 4.9890090525094336e-05, "loss": 1.414467716217041, "step": 17300 }, { "epoch": 1.4750745632722624, "grad_norm": 2.53125, "learning_rate": 4.9889831861311384e-05, "loss": 1.4281737327575683, "step": 17310 }, { "epoch": 1.475926714955262, "grad_norm": 2.828125, "learning_rate": 4.9889572894185185e-05, "loss": 1.3367189407348632, "step": 17320 }, { "epoch": 1.4767788666382615, "grad_norm": 1.9921875, "learning_rate": 4.98893136237189e-05, "loss": 1.41526460647583, "step": 17330 }, { "epoch": 1.477631018321261, "grad_norm": 1.984375, "learning_rate": 4.9889054049915696e-05, "loss": 1.5706185340881347, "step": 17340 }, { "epoch": 1.4784831700042607, "grad_norm": 2.453125, "learning_rate": 4.988879417277873e-05, "loss": 1.3658226013183594, "step": 17350 }, { "epoch": 1.4793353216872602, "grad_norm": 2.5, "learning_rate": 4.988853399231116e-05, "loss": 1.4650453567504882, "step": 17360 }, { "epoch": 1.4801874733702598, "grad_norm": 2.296875, "learning_rate": 4.9888273508516177e-05, "loss": 1.3058202743530274, "step": 17370 }, { "epoch": 1.4810396250532594, "grad_norm": 2.125, "learning_rate": 4.9888012721396935e-05, "loss": 1.4226384162902832, "step": 17380 }, { "epoch": 1.481891776736259, "grad_norm": 2.171875, "learning_rate": 4.988775163095663e-05, "loss": 1.4347776412963866, "step": 17390 }, { "epoch": 1.4827439284192585, "grad_norm": 1.8046875, "learning_rate": 4.988749023719843e-05, "loss": 1.4277276992797852, "step": 17400 }, { "epoch": 1.483596080102258, "grad_norm": 2.75, "learning_rate": 4.988722854012554e-05, "loss": 1.223691177368164, "step": 17410 }, { "epoch": 1.4844482317852576, "grad_norm": 2.859375, "learning_rate": 4.988696653974112e-05, "loss": 1.3653664588928223, "step": 17420 }, { "epoch": 1.4853003834682572, "grad_norm": 1.9140625, "learning_rate": 4.9886704236048386e-05, "loss": 1.6145305633544922, "step": 17430 }, { "epoch": 1.486152535151257, "grad_norm": 2.6875, "learning_rate": 4.9886441629050525e-05, "loss": 1.4986983299255372, "step": 17440 }, { "epoch": 1.4870046868342566, "grad_norm": 2.859375, "learning_rate": 4.988617871875075e-05, "loss": 1.419572639465332, "step": 17450 }, { "epoch": 1.4878568385172561, "grad_norm": 2.359375, "learning_rate": 4.9885915505152246e-05, "loss": 1.3781362533569337, "step": 17460 }, { "epoch": 1.4887089902002557, "grad_norm": 3.21875, "learning_rate": 4.9885651988258244e-05, "loss": 1.3358088493347169, "step": 17470 }, { "epoch": 1.4895611418832553, "grad_norm": 2.46875, "learning_rate": 4.988538816807193e-05, "loss": 1.4323202133178712, "step": 17480 }, { "epoch": 1.4904132935662548, "grad_norm": 2.78125, "learning_rate": 4.988512404459653e-05, "loss": 1.4348918914794921, "step": 17490 }, { "epoch": 1.4912654452492544, "grad_norm": 2.015625, "learning_rate": 4.988485961783528e-05, "loss": 1.4035324096679687, "step": 17500 }, { "epoch": 1.492117596932254, "grad_norm": 2.046875, "learning_rate": 4.988459488779138e-05, "loss": 1.36165132522583, "step": 17510 }, { "epoch": 1.4929697486152536, "grad_norm": 2.421875, "learning_rate": 4.988432985446806e-05, "loss": 1.3799993515014648, "step": 17520 }, { "epoch": 1.4938219002982531, "grad_norm": 1.921875, "learning_rate": 4.988406451786855e-05, "loss": 1.4808464050292969, "step": 17530 }, { "epoch": 1.4946740519812527, "grad_norm": 1.796875, "learning_rate": 4.9883798877996094e-05, "loss": 1.5201824188232422, "step": 17540 }, { "epoch": 1.4955262036642523, "grad_norm": 3.640625, "learning_rate": 4.9883532934853925e-05, "loss": 1.4357601165771485, "step": 17550 }, { "epoch": 1.4963783553472518, "grad_norm": 2.359375, "learning_rate": 4.988326668844528e-05, "loss": 1.2256975173950195, "step": 17560 }, { "epoch": 1.4972305070302514, "grad_norm": 2.40625, "learning_rate": 4.988300013877341e-05, "loss": 1.3930304527282715, "step": 17570 }, { "epoch": 1.498082658713251, "grad_norm": 3.6875, "learning_rate": 4.988273328584155e-05, "loss": 1.2297667503356933, "step": 17580 }, { "epoch": 1.4989348103962505, "grad_norm": 4.09375, "learning_rate": 4.9882466129652964e-05, "loss": 1.3821701049804687, "step": 17590 }, { "epoch": 1.49978696207925, "grad_norm": 3.703125, "learning_rate": 4.988219867021091e-05, "loss": 1.434330940246582, "step": 17600 }, { "epoch": 1.49987217724755, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6453, "eval_loss": 1.3338326215744019, "eval_runtime": 53.3944, "eval_samples_per_second": 370.114, "eval_steps_per_second": 23.149, "step": 17601 }, { "epoch": 1.5006391137622497, "grad_norm": 2.625, "learning_rate": 4.9881930907518645e-05, "loss": 1.4041208691067166, "step": 17610 }, { "epoch": 1.5014912654452492, "grad_norm": 2.046875, "learning_rate": 4.988166284157943e-05, "loss": 1.420694923400879, "step": 17620 }, { "epoch": 1.5023434171282488, "grad_norm": 1.7109375, "learning_rate": 4.988139447239653e-05, "loss": 1.4241196632385253, "step": 17630 }, { "epoch": 1.5031955688112484, "grad_norm": 2.828125, "learning_rate": 4.988112579997322e-05, "loss": 1.2049378395080566, "step": 17640 }, { "epoch": 1.504047720494248, "grad_norm": 1.90625, "learning_rate": 4.988085682431278e-05, "loss": 1.5493433952331543, "step": 17650 }, { "epoch": 1.5048998721772475, "grad_norm": 2.125, "learning_rate": 4.988058754541847e-05, "loss": 1.3909088134765626, "step": 17660 }, { "epoch": 1.505752023860247, "grad_norm": 2.3125, "learning_rate": 4.988031796329359e-05, "loss": 1.4942003250122071, "step": 17670 }, { "epoch": 1.5066041755432467, "grad_norm": 1.5, "learning_rate": 4.988004807794142e-05, "loss": 1.3605613708496094, "step": 17680 }, { "epoch": 1.5074563272262462, "grad_norm": 2.171875, "learning_rate": 4.987977788936524e-05, "loss": 1.4861967086791992, "step": 17690 }, { "epoch": 1.5083084789092458, "grad_norm": 1.796875, "learning_rate": 4.987950739756836e-05, "loss": 1.4275794982910157, "step": 17700 }, { "epoch": 1.5091606305922454, "grad_norm": 1.9453125, "learning_rate": 4.987923660255406e-05, "loss": 1.4377097129821776, "step": 17710 }, { "epoch": 1.5100127822752452, "grad_norm": 1.7265625, "learning_rate": 4.9878965504325645e-05, "loss": 1.3828031539916992, "step": 17720 }, { "epoch": 1.5108649339582447, "grad_norm": 2.4375, "learning_rate": 4.987869410288643e-05, "loss": 1.4533148765563966, "step": 17730 }, { "epoch": 1.5117170856412443, "grad_norm": 2.234375, "learning_rate": 4.987842239823971e-05, "loss": 1.466334342956543, "step": 17740 }, { "epoch": 1.5125692373242439, "grad_norm": 2.890625, "learning_rate": 4.98781503903888e-05, "loss": 1.1654952049255372, "step": 17750 }, { "epoch": 1.5134213890072434, "grad_norm": 3.328125, "learning_rate": 4.987787807933702e-05, "loss": 1.3493877410888673, "step": 17760 }, { "epoch": 1.514273540690243, "grad_norm": 2.125, "learning_rate": 4.9877605465087684e-05, "loss": 1.3888623237609863, "step": 17770 }, { "epoch": 1.5151256923732426, "grad_norm": 2.46875, "learning_rate": 4.987733254764411e-05, "loss": 1.4413522720336913, "step": 17780 }, { "epoch": 1.5159778440562421, "grad_norm": 2.828125, "learning_rate": 4.987705932700963e-05, "loss": 1.2403035163879395, "step": 17790 }, { "epoch": 1.5168299957392417, "grad_norm": 1.546875, "learning_rate": 4.9876785803187576e-05, "loss": 1.4703388214111328, "step": 17800 }, { "epoch": 1.5176821474222413, "grad_norm": 2.234375, "learning_rate": 4.9876511976181287e-05, "loss": 1.5313583374023438, "step": 17810 }, { "epoch": 1.5185342991052408, "grad_norm": 2.484375, "learning_rate": 4.987623784599408e-05, "loss": 1.273748016357422, "step": 17820 }, { "epoch": 1.5193864507882404, "grad_norm": 1.609375, "learning_rate": 4.987596341262931e-05, "loss": 1.4623674392700194, "step": 17830 }, { "epoch": 1.52023860247124, "grad_norm": 2.015625, "learning_rate": 4.9875688676090316e-05, "loss": 1.3874619483947754, "step": 17840 }, { "epoch": 1.5210907541542396, "grad_norm": 3.265625, "learning_rate": 4.987541363638045e-05, "loss": 1.202446460723877, "step": 17850 }, { "epoch": 1.5219429058372391, "grad_norm": 1.953125, "learning_rate": 4.987513829350307e-05, "loss": 1.412882423400879, "step": 17860 }, { "epoch": 1.5227950575202387, "grad_norm": 1.6953125, "learning_rate": 4.9874862647461524e-05, "loss": 1.4956844329833985, "step": 17870 }, { "epoch": 1.5236472092032383, "grad_norm": 2.53125, "learning_rate": 4.987458669825917e-05, "loss": 1.5040912628173828, "step": 17880 }, { "epoch": 1.5244993608862378, "grad_norm": 2.625, "learning_rate": 4.987431044589937e-05, "loss": 1.3449905395507813, "step": 17890 }, { "epoch": 1.5253515125692374, "grad_norm": 1.9921875, "learning_rate": 4.987403389038551e-05, "loss": 1.4565616607666017, "step": 17900 }, { "epoch": 1.526203664252237, "grad_norm": 2.40625, "learning_rate": 4.9873757031720936e-05, "loss": 1.4705817222595214, "step": 17910 }, { "epoch": 1.5270558159352365, "grad_norm": 2.15625, "learning_rate": 4.987347986990902e-05, "loss": 1.4642358779907227, "step": 17920 }, { "epoch": 1.527907967618236, "grad_norm": 3.296875, "learning_rate": 4.9873202404953165e-05, "loss": 1.334067440032959, "step": 17930 }, { "epoch": 1.5287601193012357, "grad_norm": 2.359375, "learning_rate": 4.9872924636856745e-05, "loss": 1.440049171447754, "step": 17940 }, { "epoch": 1.5296122709842352, "grad_norm": 2.453125, "learning_rate": 4.9872646565623123e-05, "loss": 1.295621681213379, "step": 17950 }, { "epoch": 1.5304644226672348, "grad_norm": 2.8125, "learning_rate": 4.9872368191255716e-05, "loss": 1.495039463043213, "step": 17960 }, { "epoch": 1.5313165743502344, "grad_norm": 3.0, "learning_rate": 4.98720895137579e-05, "loss": 1.3407943725585938, "step": 17970 }, { "epoch": 1.532168726033234, "grad_norm": 1.5234375, "learning_rate": 4.987181053313308e-05, "loss": 1.3959331512451172, "step": 17980 }, { "epoch": 1.5330208777162335, "grad_norm": 1.734375, "learning_rate": 4.987153124938465e-05, "loss": 1.4726980209350586, "step": 17990 }, { "epoch": 1.533873029399233, "grad_norm": 2.109375, "learning_rate": 4.987125166251602e-05, "loss": 1.394070816040039, "step": 18000 }, { "epoch": 1.5347251810822327, "grad_norm": 2.203125, "learning_rate": 4.9870971772530584e-05, "loss": 1.4772534370422363, "step": 18010 }, { "epoch": 1.5355773327652322, "grad_norm": 2.375, "learning_rate": 4.9870691579431765e-05, "loss": 1.4767190933227539, "step": 18020 }, { "epoch": 1.5364294844482318, "grad_norm": 2.34375, "learning_rate": 4.987041108322298e-05, "loss": 1.2551206588745116, "step": 18030 }, { "epoch": 1.5372816361312314, "grad_norm": 1.640625, "learning_rate": 4.987013028390765e-05, "loss": 1.539510726928711, "step": 18040 }, { "epoch": 1.538133787814231, "grad_norm": 2.328125, "learning_rate": 4.986984918148917e-05, "loss": 1.3143300056457519, "step": 18050 }, { "epoch": 1.5389859394972305, "grad_norm": 1.9453125, "learning_rate": 4.986956777597101e-05, "loss": 1.1723329544067382, "step": 18060 }, { "epoch": 1.53983809118023, "grad_norm": 2.546875, "learning_rate": 4.986928606735656e-05, "loss": 1.3447908401489257, "step": 18070 }, { "epoch": 1.5406902428632296, "grad_norm": 2.25, "learning_rate": 4.986900405564927e-05, "loss": 1.4947514533996582, "step": 18080 }, { "epoch": 1.5415423945462292, "grad_norm": 2.5625, "learning_rate": 4.986872174085257e-05, "loss": 1.4295663833618164, "step": 18090 }, { "epoch": 1.5423945462292288, "grad_norm": 1.84375, "learning_rate": 4.9868439122969924e-05, "loss": 1.411195659637451, "step": 18100 }, { "epoch": 1.5432466979122283, "grad_norm": 2.375, "learning_rate": 4.986815620200475e-05, "loss": 1.4260221481323243, "step": 18110 }, { "epoch": 1.544098849595228, "grad_norm": 2.125, "learning_rate": 4.9867872977960504e-05, "loss": 1.522649097442627, "step": 18120 }, { "epoch": 1.5449510012782275, "grad_norm": 2.1875, "learning_rate": 4.9867589450840645e-05, "loss": 1.4565566062927247, "step": 18130 }, { "epoch": 1.545803152961227, "grad_norm": 2.65625, "learning_rate": 4.9867305620648616e-05, "loss": 1.4475955963134766, "step": 18140 }, { "epoch": 1.5466553046442266, "grad_norm": 2.40625, "learning_rate": 4.986702148738789e-05, "loss": 1.295736598968506, "step": 18150 }, { "epoch": 1.5475074563272262, "grad_norm": 3.546875, "learning_rate": 4.986673705106192e-05, "loss": 1.4356660842895508, "step": 18160 }, { "epoch": 1.5483596080102258, "grad_norm": 1.6015625, "learning_rate": 4.986645231167417e-05, "loss": 1.362981414794922, "step": 18170 }, { "epoch": 1.5492117596932253, "grad_norm": 1.96875, "learning_rate": 4.986616726922811e-05, "loss": 1.3472570419311523, "step": 18180 }, { "epoch": 1.550063911376225, "grad_norm": 3.046875, "learning_rate": 4.986588192372723e-05, "loss": 1.4069893836975098, "step": 18190 }, { "epoch": 1.5509160630592245, "grad_norm": 1.765625, "learning_rate": 4.9865596275175e-05, "loss": 1.4669696807861328, "step": 18200 }, { "epoch": 1.551768214742224, "grad_norm": 2.015625, "learning_rate": 4.986531032357489e-05, "loss": 1.4736727714538573, "step": 18210 }, { "epoch": 1.5526203664252236, "grad_norm": 2.265625, "learning_rate": 4.98650240689304e-05, "loss": 1.2985876083374024, "step": 18220 }, { "epoch": 1.5534725181082232, "grad_norm": 2.09375, "learning_rate": 4.986473751124501e-05, "loss": 1.3805397033691407, "step": 18230 }, { "epoch": 1.5543246697912227, "grad_norm": 2.234375, "learning_rate": 4.986445065052221e-05, "loss": 1.460489273071289, "step": 18240 }, { "epoch": 1.5551768214742223, "grad_norm": 1.8203125, "learning_rate": 4.98641634867655e-05, "loss": 1.139877986907959, "step": 18250 }, { "epoch": 1.5560289731572219, "grad_norm": 2.796875, "learning_rate": 4.9863876019978387e-05, "loss": 1.3370166778564454, "step": 18260 }, { "epoch": 1.5568811248402215, "grad_norm": 2.140625, "learning_rate": 4.986358825016436e-05, "loss": 1.4222159385681152, "step": 18270 }, { "epoch": 1.557733276523221, "grad_norm": 2.796875, "learning_rate": 4.986330017732694e-05, "loss": 1.3871906280517579, "step": 18280 }, { "epoch": 1.5585854282062206, "grad_norm": 2.453125, "learning_rate": 4.9863011801469626e-05, "loss": 1.3844464302062989, "step": 18290 }, { "epoch": 1.5594375798892202, "grad_norm": 1.875, "learning_rate": 4.9862723122595946e-05, "loss": 1.4228401184082031, "step": 18300 }, { "epoch": 1.5602897315722197, "grad_norm": 1.75, "learning_rate": 4.98624341407094e-05, "loss": 1.4529316902160645, "step": 18310 }, { "epoch": 1.5611418832552193, "grad_norm": 2.296875, "learning_rate": 4.986214485581353e-05, "loss": 1.3896904945373536, "step": 18320 }, { "epoch": 1.5619940349382189, "grad_norm": 1.8046875, "learning_rate": 4.986185526791185e-05, "loss": 1.4500792503356934, "step": 18330 }, { "epoch": 1.5628461866212184, "grad_norm": 2.140625, "learning_rate": 4.986156537700789e-05, "loss": 1.2554385185241699, "step": 18340 }, { "epoch": 1.563698338304218, "grad_norm": 3.46875, "learning_rate": 4.986127518310518e-05, "loss": 1.4027731895446778, "step": 18350 }, { "epoch": 1.5645504899872176, "grad_norm": 2.09375, "learning_rate": 4.986098468620727e-05, "loss": 1.410761260986328, "step": 18360 }, { "epoch": 1.5654026416702171, "grad_norm": 1.6875, "learning_rate": 4.986069388631769e-05, "loss": 1.368984317779541, "step": 18370 }, { "epoch": 1.5662547933532167, "grad_norm": 1.9921875, "learning_rate": 4.9860402783439984e-05, "loss": 1.5148311614990235, "step": 18380 }, { "epoch": 1.5671069450362163, "grad_norm": 2.765625, "learning_rate": 4.98601113775777e-05, "loss": 1.4773941040039062, "step": 18390 }, { "epoch": 1.567959096719216, "grad_norm": 2.296875, "learning_rate": 4.9859819668734394e-05, "loss": 1.4892706871032715, "step": 18400 }, { "epoch": 1.5688112484022156, "grad_norm": 2.296875, "learning_rate": 4.985952765691362e-05, "loss": 1.5229371070861817, "step": 18410 }, { "epoch": 1.5696634000852152, "grad_norm": 2.40625, "learning_rate": 4.9859235342118926e-05, "loss": 1.4626646041870117, "step": 18420 }, { "epoch": 1.5705155517682148, "grad_norm": 1.9765625, "learning_rate": 4.985894272435389e-05, "loss": 1.4110322952270509, "step": 18430 }, { "epoch": 1.5713677034512143, "grad_norm": 2.078125, "learning_rate": 4.985864980362207e-05, "loss": 1.325582504272461, "step": 18440 }, { "epoch": 1.572219855134214, "grad_norm": 2.515625, "learning_rate": 4.985835657992704e-05, "loss": 1.4149800300598145, "step": 18450 }, { "epoch": 1.5730720068172135, "grad_norm": 2.828125, "learning_rate": 4.985806305327238e-05, "loss": 1.4321595191955567, "step": 18460 }, { "epoch": 1.573924158500213, "grad_norm": 2.71875, "learning_rate": 4.985776922366164e-05, "loss": 1.2287422180175782, "step": 18470 }, { "epoch": 1.5747763101832126, "grad_norm": 2.40625, "learning_rate": 4.9857475091098434e-05, "loss": 1.4304872512817384, "step": 18480 }, { "epoch": 1.5756284618662122, "grad_norm": 2.5, "learning_rate": 4.985718065558632e-05, "loss": 1.3358938217163085, "step": 18490 }, { "epoch": 1.5764806135492118, "grad_norm": 2.40625, "learning_rate": 4.9856885917128914e-05, "loss": 1.2261281967163087, "step": 18500 }, { "epoch": 1.5773327652322113, "grad_norm": 2.4375, "learning_rate": 4.9856590875729786e-05, "loss": 1.2582829475402832, "step": 18510 }, { "epoch": 1.578184916915211, "grad_norm": 1.7109375, "learning_rate": 4.9856295531392535e-05, "loss": 1.3936800003051757, "step": 18520 }, { "epoch": 1.5790370685982105, "grad_norm": 3.046875, "learning_rate": 4.9855999884120775e-05, "loss": 1.5088765144348144, "step": 18530 }, { "epoch": 1.57988922028121, "grad_norm": 1.78125, "learning_rate": 4.985570393391809e-05, "loss": 1.399759864807129, "step": 18540 }, { "epoch": 1.5807413719642096, "grad_norm": 2.625, "learning_rate": 4.9855407680788104e-05, "loss": 1.313217544555664, "step": 18550 }, { "epoch": 1.5815935236472092, "grad_norm": 2.8125, "learning_rate": 4.985511112473441e-05, "loss": 1.3606834411621094, "step": 18560 }, { "epoch": 1.5824456753302087, "grad_norm": 1.609375, "learning_rate": 4.985481426576063e-05, "loss": 1.4913358688354492, "step": 18570 }, { "epoch": 1.5832978270132083, "grad_norm": 1.7578125, "learning_rate": 4.985451710387039e-05, "loss": 1.501854705810547, "step": 18580 }, { "epoch": 1.5841499786962079, "grad_norm": 2.359375, "learning_rate": 4.9854219639067304e-05, "loss": 1.4046009063720704, "step": 18590 }, { "epoch": 1.5850021303792075, "grad_norm": 2.390625, "learning_rate": 4.9853921871354995e-05, "loss": 1.4482380867004394, "step": 18600 }, { "epoch": 1.585854282062207, "grad_norm": 2.828125, "learning_rate": 4.985362380073709e-05, "loss": 1.2223604202270508, "step": 18610 }, { "epoch": 1.5867064337452068, "grad_norm": 2.484375, "learning_rate": 4.985332542721724e-05, "loss": 1.2913620948791504, "step": 18620 }, { "epoch": 1.5875585854282064, "grad_norm": 1.84375, "learning_rate": 4.985302675079906e-05, "loss": 1.4119247436523437, "step": 18630 }, { "epoch": 1.588410737111206, "grad_norm": 2.34375, "learning_rate": 4.9852727771486206e-05, "loss": 1.4272357940673828, "step": 18640 }, { "epoch": 1.5892628887942055, "grad_norm": 1.921875, "learning_rate": 4.985242848928231e-05, "loss": 1.47735538482666, "step": 18650 }, { "epoch": 1.590115040477205, "grad_norm": 2.21875, "learning_rate": 4.985212890419102e-05, "loss": 1.3416533470153809, "step": 18660 }, { "epoch": 1.5909671921602047, "grad_norm": 2.390625, "learning_rate": 4.9851829016215994e-05, "loss": 1.4314695358276368, "step": 18670 }, { "epoch": 1.5918193438432042, "grad_norm": 1.8671875, "learning_rate": 4.985152882536088e-05, "loss": 1.3707731246948243, "step": 18680 }, { "epoch": 1.5926714955262038, "grad_norm": 3.03125, "learning_rate": 4.9851228331629355e-05, "loss": 1.4281984329223634, "step": 18690 }, { "epoch": 1.5935236472092034, "grad_norm": 2.296875, "learning_rate": 4.985092753502505e-05, "loss": 1.3780731201171874, "step": 18700 }, { "epoch": 1.594375798892203, "grad_norm": 1.5, "learning_rate": 4.985062643555165e-05, "loss": 1.5575349807739258, "step": 18710 }, { "epoch": 1.5952279505752025, "grad_norm": 2.5625, "learning_rate": 4.985032503321283e-05, "loss": 1.4265472412109375, "step": 18720 }, { "epoch": 1.596080102258202, "grad_norm": 2.8125, "learning_rate": 4.985002332801225e-05, "loss": 1.498560619354248, "step": 18730 }, { "epoch": 1.5969322539412016, "grad_norm": 1.71875, "learning_rate": 4.984972131995359e-05, "loss": 1.4744964599609376, "step": 18740 }, { "epoch": 1.5977844056242012, "grad_norm": 2.890625, "learning_rate": 4.9849419009040535e-05, "loss": 1.3421407699584962, "step": 18750 }, { "epoch": 1.5986365573072008, "grad_norm": 1.8671875, "learning_rate": 4.9849116395276775e-05, "loss": 1.5655357360839843, "step": 18760 }, { "epoch": 1.5994887089902003, "grad_norm": 2.234375, "learning_rate": 4.984881347866598e-05, "loss": 1.5042695999145508, "step": 18770 }, { "epoch": 1.6003408606732, "grad_norm": 2.25, "learning_rate": 4.9848510259211865e-05, "loss": 1.3637898445129395, "step": 18780 }, { "epoch": 1.6011930123561995, "grad_norm": 3.203125, "learning_rate": 4.98482067369181e-05, "loss": 1.4183788299560547, "step": 18790 }, { "epoch": 1.602045164039199, "grad_norm": 2.796875, "learning_rate": 4.98479029117884e-05, "loss": 1.373438262939453, "step": 18800 }, { "epoch": 1.6028973157221986, "grad_norm": 2.65625, "learning_rate": 4.984759878382647e-05, "loss": 1.499716567993164, "step": 18810 }, { "epoch": 1.6037494674051982, "grad_norm": 4.375, "learning_rate": 4.984729435303601e-05, "loss": 1.3084789276123048, "step": 18820 }, { "epoch": 1.6046016190881978, "grad_norm": 2.203125, "learning_rate": 4.984698961942073e-05, "loss": 1.4494200706481934, "step": 18830 }, { "epoch": 1.6054537707711973, "grad_norm": 3.796875, "learning_rate": 4.9846684582984346e-05, "loss": 1.2650381088256837, "step": 18840 }, { "epoch": 1.606305922454197, "grad_norm": 2.453125, "learning_rate": 4.984637924373058e-05, "loss": 1.369283962249756, "step": 18850 }, { "epoch": 1.6071580741371965, "grad_norm": 1.6015625, "learning_rate": 4.984607360166314e-05, "loss": 1.3645628929138183, "step": 18860 }, { "epoch": 1.608010225820196, "grad_norm": 3.203125, "learning_rate": 4.984576765678576e-05, "loss": 1.3609013557434082, "step": 18870 }, { "epoch": 1.6088623775031956, "grad_norm": 1.9375, "learning_rate": 4.984546140910218e-05, "loss": 1.338018035888672, "step": 18880 }, { "epoch": 1.6097145291861952, "grad_norm": 3.421875, "learning_rate": 4.9845154858616105e-05, "loss": 1.4363526344299316, "step": 18890 }, { "epoch": 1.6105666808691947, "grad_norm": 2.421875, "learning_rate": 4.98448480053313e-05, "loss": 1.3857454299926757, "step": 18900 }, { "epoch": 1.6114188325521943, "grad_norm": 2.859375, "learning_rate": 4.9844540849251485e-05, "loss": 1.4898658752441407, "step": 18910 }, { "epoch": 1.6122709842351939, "grad_norm": 2.390625, "learning_rate": 4.984423339038041e-05, "loss": 1.4333515167236328, "step": 18920 }, { "epoch": 1.6131231359181935, "grad_norm": 2.0, "learning_rate": 4.984392562872182e-05, "loss": 1.4633604049682618, "step": 18930 }, { "epoch": 1.613975287601193, "grad_norm": 2.03125, "learning_rate": 4.9843617564279477e-05, "loss": 1.3485666275024415, "step": 18940 }, { "epoch": 1.6148274392841926, "grad_norm": 2.09375, "learning_rate": 4.984330919705712e-05, "loss": 1.4783991813659667, "step": 18950 }, { "epoch": 1.6156795909671922, "grad_norm": 4.53125, "learning_rate": 4.984300052705851e-05, "loss": 1.3054086685180664, "step": 18960 }, { "epoch": 1.6165317426501917, "grad_norm": 2.609375, "learning_rate": 4.9842691554287415e-05, "loss": 1.4210933685302733, "step": 18970 }, { "epoch": 1.6173838943331913, "grad_norm": 1.9609375, "learning_rate": 4.98423822787476e-05, "loss": 1.4735469818115234, "step": 18980 }, { "epoch": 1.6182360460161909, "grad_norm": 3.25, "learning_rate": 4.9842072700442835e-05, "loss": 1.5212715148925782, "step": 18990 }, { "epoch": 1.6190881976991904, "grad_norm": 2.015625, "learning_rate": 4.984176281937688e-05, "loss": 1.361150360107422, "step": 19000 }, { "epoch": 1.61994034938219, "grad_norm": 2.34375, "learning_rate": 4.984145263555353e-05, "loss": 1.3321546554565429, "step": 19010 }, { "epoch": 1.6207925010651896, "grad_norm": 2.21875, "learning_rate": 4.9841142148976556e-05, "loss": 1.225545883178711, "step": 19020 }, { "epoch": 1.6216446527481891, "grad_norm": 1.796875, "learning_rate": 4.984083135964974e-05, "loss": 1.3111507415771484, "step": 19030 }, { "epoch": 1.6224968044311887, "grad_norm": 2.125, "learning_rate": 4.984052026757688e-05, "loss": 1.3449086189270019, "step": 19040 }, { "epoch": 1.6233489561141883, "grad_norm": 1.46875, "learning_rate": 4.9840208872761755e-05, "loss": 1.3658646583557128, "step": 19050 }, { "epoch": 1.6242011077971878, "grad_norm": 2.34375, "learning_rate": 4.983989717520817e-05, "loss": 1.56469669342041, "step": 19060 }, { "epoch": 1.6250532594801874, "grad_norm": 1.7890625, "learning_rate": 4.983958517491991e-05, "loss": 1.3332048416137696, "step": 19070 }, { "epoch": 1.625905411163187, "grad_norm": 2.046875, "learning_rate": 4.98392728719008e-05, "loss": 1.3321764945983887, "step": 19080 }, { "epoch": 1.6267575628461866, "grad_norm": 2.359375, "learning_rate": 4.9838960266154624e-05, "loss": 1.447859287261963, "step": 19090 }, { "epoch": 1.6276097145291861, "grad_norm": 3.09375, "learning_rate": 4.9838647357685205e-05, "loss": 1.4279949188232421, "step": 19100 }, { "epoch": 1.6284618662121857, "grad_norm": 3.671875, "learning_rate": 4.983833414649635e-05, "loss": 1.346235466003418, "step": 19110 }, { "epoch": 1.6293140178951853, "grad_norm": 2.609375, "learning_rate": 4.9838020632591873e-05, "loss": 1.4151387214660645, "step": 19120 }, { "epoch": 1.6301661695781848, "grad_norm": 1.8515625, "learning_rate": 4.983770681597562e-05, "loss": 1.4365800857543944, "step": 19130 }, { "epoch": 1.6310183212611844, "grad_norm": 2.375, "learning_rate": 4.983739269665138e-05, "loss": 1.364539337158203, "step": 19140 }, { "epoch": 1.631870472944184, "grad_norm": 2.671875, "learning_rate": 4.9837078274622993e-05, "loss": 1.3343406677246095, "step": 19150 }, { "epoch": 1.6327226246271835, "grad_norm": 1.9453125, "learning_rate": 4.983676354989431e-05, "loss": 1.4619688987731934, "step": 19160 }, { "epoch": 1.633574776310183, "grad_norm": 2.640625, "learning_rate": 4.983644852246914e-05, "loss": 1.3817001342773438, "step": 19170 }, { "epoch": 1.6344269279931827, "grad_norm": 1.7109375, "learning_rate": 4.983613319235134e-05, "loss": 1.499324131011963, "step": 19180 }, { "epoch": 1.6352790796761822, "grad_norm": 2.15625, "learning_rate": 4.983581755954475e-05, "loss": 1.4905197143554687, "step": 19190 }, { "epoch": 1.6361312313591818, "grad_norm": 2.03125, "learning_rate": 4.9835501624053215e-05, "loss": 1.358537483215332, "step": 19200 }, { "epoch": 1.6369833830421814, "grad_norm": 1.9140625, "learning_rate": 4.9835185385880576e-05, "loss": 1.3654868125915527, "step": 19210 }, { "epoch": 1.637835534725181, "grad_norm": 1.9140625, "learning_rate": 4.98348688450307e-05, "loss": 1.5099059104919434, "step": 19220 }, { "epoch": 1.6386876864081805, "grad_norm": 1.9921875, "learning_rate": 4.983455200150745e-05, "loss": 1.3197769165039062, "step": 19230 }, { "epoch": 1.63953983809118, "grad_norm": 1.8984375, "learning_rate": 4.983423485531467e-05, "loss": 1.4676291465759277, "step": 19240 }, { "epoch": 1.6403919897741797, "grad_norm": 2.96875, "learning_rate": 4.983391740645623e-05, "loss": 1.2622689247131347, "step": 19250 }, { "epoch": 1.6412441414571792, "grad_norm": 1.6875, "learning_rate": 4.983359965493601e-05, "loss": 1.4669668197631835, "step": 19260 }, { "epoch": 1.6420962931401788, "grad_norm": 1.984375, "learning_rate": 4.983328160075788e-05, "loss": 1.3820133209228516, "step": 19270 }, { "epoch": 1.6429484448231784, "grad_norm": 1.265625, "learning_rate": 4.98329632439257e-05, "loss": 1.395542812347412, "step": 19280 }, { "epoch": 1.643800596506178, "grad_norm": 2.59375, "learning_rate": 4.9832644584443364e-05, "loss": 1.414292335510254, "step": 19290 }, { "epoch": 1.6446527481891777, "grad_norm": 2.46875, "learning_rate": 4.983232562231475e-05, "loss": 1.4571653366088868, "step": 19300 }, { "epoch": 1.6455048998721773, "grad_norm": 2.046875, "learning_rate": 4.983200635754375e-05, "loss": 1.384368896484375, "step": 19310 }, { "epoch": 1.6463570515551769, "grad_norm": 3.859375, "learning_rate": 4.9831686790134255e-05, "loss": 1.3670742988586426, "step": 19320 }, { "epoch": 1.6472092032381764, "grad_norm": 3.15625, "learning_rate": 4.983136692009016e-05, "loss": 1.4137657165527344, "step": 19330 }, { "epoch": 1.648061354921176, "grad_norm": 2.96875, "learning_rate": 4.983104674741536e-05, "loss": 1.399698257446289, "step": 19340 }, { "epoch": 1.6489135066041756, "grad_norm": 1.921875, "learning_rate": 4.983072627211376e-05, "loss": 1.501253318786621, "step": 19350 }, { "epoch": 1.6497656582871751, "grad_norm": 1.96875, "learning_rate": 4.983040549418926e-05, "loss": 1.3349817276000977, "step": 19360 }, { "epoch": 1.6506178099701747, "grad_norm": 2.46875, "learning_rate": 4.9830084413645774e-05, "loss": 1.4522038459777833, "step": 19370 }, { "epoch": 1.6514699616531743, "grad_norm": 2.453125, "learning_rate": 4.982976303048722e-05, "loss": 1.465405559539795, "step": 19380 }, { "epoch": 1.6523221133361738, "grad_norm": 2.0, "learning_rate": 4.98294413447175e-05, "loss": 1.4600700378417968, "step": 19390 }, { "epoch": 1.6531742650191734, "grad_norm": 2.265625, "learning_rate": 4.9829119356340546e-05, "loss": 1.4808540344238281, "step": 19400 }, { "epoch": 1.654026416702173, "grad_norm": 2.03125, "learning_rate": 4.9828797065360286e-05, "loss": 1.463935375213623, "step": 19410 }, { "epoch": 1.6548785683851726, "grad_norm": 2.078125, "learning_rate": 4.982847447178064e-05, "loss": 1.3601481437683105, "step": 19420 }, { "epoch": 1.6557307200681721, "grad_norm": 2.484375, "learning_rate": 4.982815157560555e-05, "loss": 1.3999747276306151, "step": 19430 }, { "epoch": 1.6565828717511717, "grad_norm": 2.234375, "learning_rate": 4.982782837683893e-05, "loss": 1.2857720375061035, "step": 19440 }, { "epoch": 1.6574350234341713, "grad_norm": 1.96875, "learning_rate": 4.9827504875484744e-05, "loss": 1.4512001991271972, "step": 19450 }, { "epoch": 1.6582871751171708, "grad_norm": 1.46875, "learning_rate": 4.982718107154691e-05, "loss": 1.5125085830688476, "step": 19460 }, { "epoch": 1.6591393268001704, "grad_norm": 3.625, "learning_rate": 4.9826856965029396e-05, "loss": 1.4390454292297363, "step": 19470 }, { "epoch": 1.65999147848317, "grad_norm": 1.9609375, "learning_rate": 4.982653255593614e-05, "loss": 1.260635757446289, "step": 19480 }, { "epoch": 1.6608436301661695, "grad_norm": 2.40625, "learning_rate": 4.98262078442711e-05, "loss": 1.3736013412475585, "step": 19490 }, { "epoch": 1.661695781849169, "grad_norm": 4.03125, "learning_rate": 4.982588283003823e-05, "loss": 1.4098575592041016, "step": 19500 }, { "epoch": 1.6625479335321687, "grad_norm": 2.921875, "learning_rate": 4.9825557513241497e-05, "loss": 1.3383109092712402, "step": 19510 }, { "epoch": 1.6634000852151682, "grad_norm": 2.203125, "learning_rate": 4.982523189388486e-05, "loss": 1.4159780502319337, "step": 19520 }, { "epoch": 1.664252236898168, "grad_norm": 2.578125, "learning_rate": 4.9824905971972284e-05, "loss": 1.5545214653015136, "step": 19530 }, { "epoch": 1.6651043885811676, "grad_norm": 2.25, "learning_rate": 4.982457974750775e-05, "loss": 1.316845703125, "step": 19540 }, { "epoch": 1.6659565402641672, "grad_norm": 1.5625, "learning_rate": 4.9824253220495234e-05, "loss": 1.3904570579528808, "step": 19550 }, { "epoch": 1.6668086919471667, "grad_norm": 1.765625, "learning_rate": 4.982392639093871e-05, "loss": 1.3634869575500488, "step": 19560 }, { "epoch": 1.6676608436301663, "grad_norm": 2.546875, "learning_rate": 4.982359925884217e-05, "loss": 1.3542913436889648, "step": 19570 }, { "epoch": 1.6685129953131659, "grad_norm": 2.4375, "learning_rate": 4.9823271824209585e-05, "loss": 1.4533591270446777, "step": 19580 }, { "epoch": 1.6693651469961654, "grad_norm": 2.375, "learning_rate": 4.9822944087044955e-05, "loss": 1.401167869567871, "step": 19590 }, { "epoch": 1.670217298679165, "grad_norm": 1.6484375, "learning_rate": 4.9822616047352276e-05, "loss": 1.2974672317504883, "step": 19600 }, { "epoch": 1.6710694503621646, "grad_norm": 3.65625, "learning_rate": 4.982228770513555e-05, "loss": 1.3336725234985352, "step": 19610 }, { "epoch": 1.6719216020451642, "grad_norm": 2.84375, "learning_rate": 4.982195906039877e-05, "loss": 1.4926626205444335, "step": 19620 }, { "epoch": 1.6727737537281637, "grad_norm": 2.21875, "learning_rate": 4.9821630113145936e-05, "loss": 1.3560620307922364, "step": 19630 }, { "epoch": 1.6736259054111633, "grad_norm": 1.875, "learning_rate": 4.982130086338107e-05, "loss": 1.4872838973999023, "step": 19640 }, { "epoch": 1.6744780570941629, "grad_norm": 2.5625, "learning_rate": 4.982097131110818e-05, "loss": 1.2631665229797364, "step": 19650 }, { "epoch": 1.6753302087771624, "grad_norm": 2.5, "learning_rate": 4.982064145633128e-05, "loss": 1.4465940475463868, "step": 19660 }, { "epoch": 1.676182360460162, "grad_norm": 2.828125, "learning_rate": 4.98203112990544e-05, "loss": 1.2489534378051759, "step": 19670 }, { "epoch": 1.6770345121431616, "grad_norm": 2.21875, "learning_rate": 4.9819980839281545e-05, "loss": 1.5225841522216796, "step": 19680 }, { "epoch": 1.6778866638261611, "grad_norm": 3.421875, "learning_rate": 4.981965007701676e-05, "loss": 1.3734678268432616, "step": 19690 }, { "epoch": 1.6787388155091607, "grad_norm": 2.375, "learning_rate": 4.981931901226407e-05, "loss": 1.3464533805847168, "step": 19700 }, { "epoch": 1.6795909671921603, "grad_norm": 2.234375, "learning_rate": 4.9818987645027505e-05, "loss": 1.3433768272399902, "step": 19710 }, { "epoch": 1.6804431188751598, "grad_norm": 1.875, "learning_rate": 4.9818655975311115e-05, "loss": 1.2834291458129883, "step": 19720 }, { "epoch": 1.6812952705581594, "grad_norm": 2.296875, "learning_rate": 4.9818324003118934e-05, "loss": 1.465505313873291, "step": 19730 }, { "epoch": 1.682147422241159, "grad_norm": 3.421875, "learning_rate": 4.981799172845501e-05, "loss": 1.3777854919433594, "step": 19740 }, { "epoch": 1.6829995739241586, "grad_norm": 2.890625, "learning_rate": 4.981765915132338e-05, "loss": 1.4547465324401856, "step": 19750 }, { "epoch": 1.6838517256071581, "grad_norm": 2.40625, "learning_rate": 4.9817326271728115e-05, "loss": 1.453665542602539, "step": 19760 }, { "epoch": 1.6847038772901577, "grad_norm": 2.6875, "learning_rate": 4.981699308967327e-05, "loss": 1.5053634643554688, "step": 19770 }, { "epoch": 1.6855560289731573, "grad_norm": 2.265625, "learning_rate": 4.9816659605162894e-05, "loss": 1.4271561622619628, "step": 19780 }, { "epoch": 1.6864081806561568, "grad_norm": 2.109375, "learning_rate": 4.981632581820107e-05, "loss": 1.4637690544128419, "step": 19790 }, { "epoch": 1.6872603323391564, "grad_norm": 2.5, "learning_rate": 4.9815991728791845e-05, "loss": 1.4011150360107423, "step": 19800 }, { "epoch": 1.688112484022156, "grad_norm": 2.171875, "learning_rate": 4.9815657336939305e-05, "loss": 1.3375853538513183, "step": 19810 }, { "epoch": 1.6889646357051555, "grad_norm": 1.7578125, "learning_rate": 4.981532264264752e-05, "loss": 1.394768524169922, "step": 19820 }, { "epoch": 1.689816787388155, "grad_norm": 1.3828125, "learning_rate": 4.981498764592057e-05, "loss": 1.4094745635986328, "step": 19830 }, { "epoch": 1.6906689390711547, "grad_norm": 2.296875, "learning_rate": 4.9814652346762534e-05, "loss": 1.381995677947998, "step": 19840 }, { "epoch": 1.6915210907541542, "grad_norm": 1.9140625, "learning_rate": 4.9814316745177505e-05, "loss": 1.3791398048400878, "step": 19850 }, { "epoch": 1.6923732424371538, "grad_norm": 1.78125, "learning_rate": 4.9813980841169575e-05, "loss": 1.4662089347839355, "step": 19860 }, { "epoch": 1.6932253941201534, "grad_norm": 2.15625, "learning_rate": 4.981364463474283e-05, "loss": 1.385054111480713, "step": 19870 }, { "epoch": 1.694077545803153, "grad_norm": 3.140625, "learning_rate": 4.9813308125901366e-05, "loss": 1.2935422897338866, "step": 19880 }, { "epoch": 1.6949296974861525, "grad_norm": 1.8125, "learning_rate": 4.9812971314649294e-05, "loss": 1.4094717025756835, "step": 19890 }, { "epoch": 1.695781849169152, "grad_norm": 2.21875, "learning_rate": 4.9812634200990704e-05, "loss": 1.5178763389587402, "step": 19900 }, { "epoch": 1.6966340008521517, "grad_norm": 1.890625, "learning_rate": 4.9812296784929725e-05, "loss": 1.212514305114746, "step": 19910 }, { "epoch": 1.6974861525351512, "grad_norm": 2.546875, "learning_rate": 4.9811959066470454e-05, "loss": 1.2600293159484863, "step": 19920 }, { "epoch": 1.6983383042181508, "grad_norm": 2.4375, "learning_rate": 4.9811621045617014e-05, "loss": 1.450352668762207, "step": 19930 }, { "epoch": 1.6991904559011504, "grad_norm": 3.734375, "learning_rate": 4.981128272237352e-05, "loss": 1.3151623725891113, "step": 19940 }, { "epoch": 1.70004260758415, "grad_norm": 1.890625, "learning_rate": 4.981094409674409e-05, "loss": 1.307030487060547, "step": 19950 }, { "epoch": 1.7008947592671495, "grad_norm": 3.015625, "learning_rate": 4.9810605168732874e-05, "loss": 1.1804752349853516, "step": 19960 }, { "epoch": 1.701746910950149, "grad_norm": 2.375, "learning_rate": 4.981026593834398e-05, "loss": 1.3662442207336425, "step": 19970 }, { "epoch": 1.7025990626331486, "grad_norm": 2.75, "learning_rate": 4.9809926405581544e-05, "loss": 1.1776183128356934, "step": 19980 }, { "epoch": 1.7034512143161482, "grad_norm": 2.328125, "learning_rate": 4.9809586570449714e-05, "loss": 1.3814671516418457, "step": 19990 }, { "epoch": 1.7043033659991478, "grad_norm": 4.34375, "learning_rate": 4.980924643295262e-05, "loss": 1.4534417152404786, "step": 20000 }, { "epoch": 1.7051555176821473, "grad_norm": 2.234375, "learning_rate": 4.9808905993094425e-05, "loss": 1.4626115798950194, "step": 20010 }, { "epoch": 1.706007669365147, "grad_norm": 2.046875, "learning_rate": 4.980856525087926e-05, "loss": 1.3423304557800293, "step": 20020 }, { "epoch": 1.7068598210481465, "grad_norm": 2.921875, "learning_rate": 4.980822420631129e-05, "loss": 1.37255916595459, "step": 20030 }, { "epoch": 1.707711972731146, "grad_norm": 1.578125, "learning_rate": 4.980788285939466e-05, "loss": 1.4903432846069335, "step": 20040 }, { "epoch": 1.7085641244141456, "grad_norm": 2.859375, "learning_rate": 4.9807541210133546e-05, "loss": 1.3977683067321778, "step": 20050 }, { "epoch": 1.7094162760971452, "grad_norm": 2.15625, "learning_rate": 4.98071992585321e-05, "loss": 1.4219631195068358, "step": 20060 }, { "epoch": 1.7102684277801448, "grad_norm": 1.6953125, "learning_rate": 4.9806857004594484e-05, "loss": 1.4292240142822266, "step": 20070 }, { "epoch": 1.7111205794631443, "grad_norm": 1.7578125, "learning_rate": 4.980651444832488e-05, "loss": 1.4631771087646483, "step": 20080 }, { "epoch": 1.711972731146144, "grad_norm": 1.859375, "learning_rate": 4.980617158972747e-05, "loss": 1.4593969345092774, "step": 20090 }, { "epoch": 1.7128248828291435, "grad_norm": 2.421875, "learning_rate": 4.980582842880641e-05, "loss": 1.3957670211791993, "step": 20100 }, { "epoch": 1.713677034512143, "grad_norm": 2.125, "learning_rate": 4.980548496556591e-05, "loss": 1.3493219375610352, "step": 20110 }, { "epoch": 1.7145291861951426, "grad_norm": 2.125, "learning_rate": 4.980514120001013e-05, "loss": 1.4730380058288575, "step": 20120 }, { "epoch": 1.7153813378781422, "grad_norm": 2.53125, "learning_rate": 4.9804797132143265e-05, "loss": 1.446037769317627, "step": 20130 }, { "epoch": 1.7162334895611417, "grad_norm": 3.40625, "learning_rate": 4.9804452761969524e-05, "loss": 1.2429515838623046, "step": 20140 }, { "epoch": 1.7170856412441413, "grad_norm": 2.890625, "learning_rate": 4.98041080894931e-05, "loss": 1.4888317108154296, "step": 20150 }, { "epoch": 1.7179377929271409, "grad_norm": 2.84375, "learning_rate": 4.980376311471817e-05, "loss": 1.2846237182617188, "step": 20160 }, { "epoch": 1.7187899446101405, "grad_norm": 1.9375, "learning_rate": 4.980341783764897e-05, "loss": 1.492880630493164, "step": 20170 }, { "epoch": 1.71964209629314, "grad_norm": 2.0625, "learning_rate": 4.980307225828969e-05, "loss": 1.2791991233825684, "step": 20180 }, { "epoch": 1.7204942479761396, "grad_norm": 3.90625, "learning_rate": 4.980272637664454e-05, "loss": 1.3616272926330566, "step": 20190 }, { "epoch": 1.7213463996591394, "grad_norm": 2.4375, "learning_rate": 4.9802380192717744e-05, "loss": 1.4045958518981934, "step": 20200 }, { "epoch": 1.722198551342139, "grad_norm": 3.21875, "learning_rate": 4.980203370651352e-05, "loss": 1.2645087242126465, "step": 20210 }, { "epoch": 1.7230507030251385, "grad_norm": 3.078125, "learning_rate": 4.980168691803609e-05, "loss": 1.3572270393371582, "step": 20220 }, { "epoch": 1.723902854708138, "grad_norm": 4.0625, "learning_rate": 4.980133982728968e-05, "loss": 1.3989126205444335, "step": 20230 }, { "epoch": 1.7247550063911377, "grad_norm": 2.859375, "learning_rate": 4.980099243427851e-05, "loss": 1.5402877807617188, "step": 20240 }, { "epoch": 1.7256071580741372, "grad_norm": 2.09375, "learning_rate": 4.9800644739006843e-05, "loss": 1.4351228713989257, "step": 20250 }, { "epoch": 1.7264593097571368, "grad_norm": 2.125, "learning_rate": 4.980029674147888e-05, "loss": 1.483678150177002, "step": 20260 }, { "epoch": 1.7273114614401364, "grad_norm": 2.84375, "learning_rate": 4.9799948441698885e-05, "loss": 1.4575531005859375, "step": 20270 }, { "epoch": 1.728163613123136, "grad_norm": 1.609375, "learning_rate": 4.979959983967109e-05, "loss": 1.4219743728637695, "step": 20280 }, { "epoch": 1.7290157648061355, "grad_norm": 2.65625, "learning_rate": 4.979925093539977e-05, "loss": 1.426582431793213, "step": 20290 }, { "epoch": 1.729867916489135, "grad_norm": 2.390625, "learning_rate": 4.9798901728889135e-05, "loss": 1.3823019981384277, "step": 20300 }, { "epoch": 1.7307200681721346, "grad_norm": 2.890625, "learning_rate": 4.979855222014348e-05, "loss": 1.4528803825378418, "step": 20310 }, { "epoch": 1.7315722198551342, "grad_norm": 2.234375, "learning_rate": 4.979820240916704e-05, "loss": 1.3541604042053224, "step": 20320 }, { "epoch": 1.7324243715381338, "grad_norm": 2.3125, "learning_rate": 4.979785229596409e-05, "loss": 1.378018569946289, "step": 20330 }, { "epoch": 1.7332765232211333, "grad_norm": 1.921875, "learning_rate": 4.979750188053889e-05, "loss": 1.2731481552124024, "step": 20340 }, { "epoch": 1.734128674904133, "grad_norm": 2.546875, "learning_rate": 4.9797151162895716e-05, "loss": 1.4112140655517578, "step": 20350 }, { "epoch": 1.7349808265871325, "grad_norm": 2.15625, "learning_rate": 4.9796800143038844e-05, "loss": 1.4616390228271485, "step": 20360 }, { "epoch": 1.735832978270132, "grad_norm": 1.921875, "learning_rate": 4.979644882097254e-05, "loss": 1.3818855285644531, "step": 20370 }, { "epoch": 1.7366851299531316, "grad_norm": 1.75, "learning_rate": 4.97960971967011e-05, "loss": 1.4312963485717773, "step": 20380 }, { "epoch": 1.7375372816361312, "grad_norm": 2.09375, "learning_rate": 4.979574527022881e-05, "loss": 1.3816766738891602, "step": 20390 }, { "epoch": 1.7383894333191308, "grad_norm": 2.546875, "learning_rate": 4.9795393041559944e-05, "loss": 1.4333789825439454, "step": 20400 }, { "epoch": 1.7392415850021303, "grad_norm": 2.453125, "learning_rate": 4.9795040510698806e-05, "loss": 1.309584617614746, "step": 20410 }, { "epoch": 1.74009373668513, "grad_norm": 1.8828125, "learning_rate": 4.979468767764969e-05, "loss": 1.4809462547302246, "step": 20420 }, { "epoch": 1.7409458883681297, "grad_norm": 2.75, "learning_rate": 4.97943345424169e-05, "loss": 1.5215893745422364, "step": 20430 }, { "epoch": 1.7417980400511293, "grad_norm": 1.828125, "learning_rate": 4.9793981105004736e-05, "loss": 1.3640636444091796, "step": 20440 }, { "epoch": 1.7426501917341288, "grad_norm": 3.6875, "learning_rate": 4.979362736541749e-05, "loss": 1.3521457672119142, "step": 20450 }, { "epoch": 1.7435023434171284, "grad_norm": 2.25, "learning_rate": 4.9793273323659505e-05, "loss": 1.3151159286499023, "step": 20460 }, { "epoch": 1.744354495100128, "grad_norm": 2.1875, "learning_rate": 4.979291897973508e-05, "loss": 1.5370201110839843, "step": 20470 }, { "epoch": 1.7452066467831275, "grad_norm": 1.5703125, "learning_rate": 4.979256433364853e-05, "loss": 1.3770027160644531, "step": 20480 }, { "epoch": 1.746058798466127, "grad_norm": 2.140625, "learning_rate": 4.979220938540418e-05, "loss": 1.360203742980957, "step": 20490 }, { "epoch": 1.7469109501491267, "grad_norm": 1.765625, "learning_rate": 4.979185413500635e-05, "loss": 1.2713259696960448, "step": 20500 }, { "epoch": 1.7477631018321262, "grad_norm": 3.0, "learning_rate": 4.979149858245938e-05, "loss": 1.4564527511596679, "step": 20510 }, { "epoch": 1.7486152535151258, "grad_norm": 2.40625, "learning_rate": 4.9791142727767605e-05, "loss": 1.4364975929260253, "step": 20520 }, { "epoch": 1.7494674051981254, "grad_norm": 1.9765625, "learning_rate": 4.9790786570935354e-05, "loss": 1.2025729179382325, "step": 20530 }, { "epoch": 1.750319556881125, "grad_norm": 1.7734375, "learning_rate": 4.979043011196697e-05, "loss": 1.331564712524414, "step": 20540 }, { "epoch": 1.7511717085641245, "grad_norm": 1.9921875, "learning_rate": 4.979007335086679e-05, "loss": 1.297113037109375, "step": 20550 }, { "epoch": 1.752023860247124, "grad_norm": 1.703125, "learning_rate": 4.9789716287639174e-05, "loss": 1.3798940658569336, "step": 20560 }, { "epoch": 1.7528760119301237, "grad_norm": 2.59375, "learning_rate": 4.978935892228847e-05, "loss": 1.432002067565918, "step": 20570 }, { "epoch": 1.7537281636131232, "grad_norm": 2.421875, "learning_rate": 4.978900125481904e-05, "loss": 1.4409296989440918, "step": 20580 }, { "epoch": 1.7545803152961228, "grad_norm": 2.125, "learning_rate": 4.978864328523521e-05, "loss": 1.3485626220703124, "step": 20590 }, { "epoch": 1.7554324669791224, "grad_norm": 2.046875, "learning_rate": 4.978828501354139e-05, "loss": 1.5070340156555175, "step": 20600 }, { "epoch": 1.756284618662122, "grad_norm": 2.625, "learning_rate": 4.9787926439741914e-05, "loss": 1.5648438453674316, "step": 20610 }, { "epoch": 1.7571367703451215, "grad_norm": 2.375, "learning_rate": 4.978756756384117e-05, "loss": 1.375816822052002, "step": 20620 }, { "epoch": 1.757988922028121, "grad_norm": 2.109375, "learning_rate": 4.978720838584352e-05, "loss": 1.2836458206176757, "step": 20630 }, { "epoch": 1.7588410737111206, "grad_norm": 3.5, "learning_rate": 4.9786848905753344e-05, "loss": 1.4088013648986817, "step": 20640 }, { "epoch": 1.7596932253941202, "grad_norm": 2.671875, "learning_rate": 4.978648912357503e-05, "loss": 1.389228916168213, "step": 20650 }, { "epoch": 1.7605453770771198, "grad_norm": 2.40625, "learning_rate": 4.9786129039312954e-05, "loss": 1.3637378692626954, "step": 20660 }, { "epoch": 1.7613975287601193, "grad_norm": 2.234375, "learning_rate": 4.978576865297151e-05, "loss": 1.3157602310180665, "step": 20670 }, { "epoch": 1.762249680443119, "grad_norm": 2.5, "learning_rate": 4.978540796455509e-05, "loss": 1.4177002906799316, "step": 20680 }, { "epoch": 1.7631018321261185, "grad_norm": 2.34375, "learning_rate": 4.978504697406808e-05, "loss": 1.3880741119384765, "step": 20690 }, { "epoch": 1.763953983809118, "grad_norm": 2.375, "learning_rate": 4.978468568151489e-05, "loss": 1.2587328910827638, "step": 20700 }, { "epoch": 1.7648061354921176, "grad_norm": 2.046875, "learning_rate": 4.9784324086899925e-05, "loss": 1.3628152847290038, "step": 20710 }, { "epoch": 1.7656582871751172, "grad_norm": 2.421875, "learning_rate": 4.978396219022759e-05, "loss": 1.3307220458984375, "step": 20720 }, { "epoch": 1.7665104388581168, "grad_norm": 2.515625, "learning_rate": 4.978359999150228e-05, "loss": 1.348722267150879, "step": 20730 }, { "epoch": 1.7673625905411163, "grad_norm": 2.453125, "learning_rate": 4.978323749072844e-05, "loss": 1.484344482421875, "step": 20740 }, { "epoch": 1.768214742224116, "grad_norm": 2.84375, "learning_rate": 4.9782874687910466e-05, "loss": 1.4644721031188965, "step": 20750 }, { "epoch": 1.7690668939071155, "grad_norm": 1.9609375, "learning_rate": 4.9782511583052774e-05, "loss": 1.4912415504455567, "step": 20760 }, { "epoch": 1.769919045590115, "grad_norm": 1.625, "learning_rate": 4.978214817615981e-05, "loss": 1.5004337310791016, "step": 20770 }, { "epoch": 1.7707711972731146, "grad_norm": 3.078125, "learning_rate": 4.978178446723599e-05, "loss": 1.3332413673400878, "step": 20780 }, { "epoch": 1.7716233489561142, "grad_norm": 2.09375, "learning_rate": 4.978142045628575e-05, "loss": 1.418159294128418, "step": 20790 }, { "epoch": 1.7724755006391137, "grad_norm": 2.078125, "learning_rate": 4.978105614331352e-05, "loss": 1.3767031669616698, "step": 20800 }, { "epoch": 1.7733276523221133, "grad_norm": 2.421875, "learning_rate": 4.9780691528323754e-05, "loss": 1.2692313194274902, "step": 20810 }, { "epoch": 1.7741798040051129, "grad_norm": 2.8125, "learning_rate": 4.978032661132088e-05, "loss": 1.4422119140625, "step": 20820 }, { "epoch": 1.7750319556881125, "grad_norm": 2.734375, "learning_rate": 4.977996139230936e-05, "loss": 1.4642041206359864, "step": 20830 }, { "epoch": 1.775884107371112, "grad_norm": 2.109375, "learning_rate": 4.977959587129364e-05, "loss": 1.3765226364135743, "step": 20840 }, { "epoch": 1.7767362590541116, "grad_norm": 2.234375, "learning_rate": 4.9779230048278155e-05, "loss": 1.2932884216308593, "step": 20850 }, { "epoch": 1.7775884107371112, "grad_norm": 2.140625, "learning_rate": 4.97788639232674e-05, "loss": 1.2529098510742187, "step": 20860 }, { "epoch": 1.7784405624201107, "grad_norm": 2.703125, "learning_rate": 4.9778497496265805e-05, "loss": 1.1797035217285157, "step": 20870 }, { "epoch": 1.7792927141031103, "grad_norm": 4.03125, "learning_rate": 4.977813076727786e-05, "loss": 1.3367047309875488, "step": 20880 }, { "epoch": 1.7801448657861099, "grad_norm": 2.5, "learning_rate": 4.977776373630802e-05, "loss": 1.3998727798461914, "step": 20890 }, { "epoch": 1.7809970174691094, "grad_norm": 2.171875, "learning_rate": 4.977739640336077e-05, "loss": 1.34104585647583, "step": 20900 }, { "epoch": 1.781849169152109, "grad_norm": 2.515625, "learning_rate": 4.977702876844056e-05, "loss": 1.2160417556762695, "step": 20910 }, { "epoch": 1.7827013208351086, "grad_norm": 2.703125, "learning_rate": 4.97766608315519e-05, "loss": 1.4293261528015138, "step": 20920 }, { "epoch": 1.7835534725181081, "grad_norm": 1.703125, "learning_rate": 4.977629259269926e-05, "loss": 1.2650264739990233, "step": 20930 }, { "epoch": 1.7844056242011077, "grad_norm": 2.28125, "learning_rate": 4.9775924051887136e-05, "loss": 1.279607391357422, "step": 20940 }, { "epoch": 1.7852577758841073, "grad_norm": 2.359375, "learning_rate": 4.977555520912001e-05, "loss": 1.3286285400390625, "step": 20950 }, { "epoch": 1.7861099275671068, "grad_norm": 1.9140625, "learning_rate": 4.977518606440238e-05, "loss": 1.4376749038696288, "step": 20960 }, { "epoch": 1.7869620792501064, "grad_norm": 2.4375, "learning_rate": 4.977481661773875e-05, "loss": 1.3472050666809081, "step": 20970 }, { "epoch": 1.787814230933106, "grad_norm": 2.875, "learning_rate": 4.9774446869133624e-05, "loss": 1.3721296310424804, "step": 20980 }, { "epoch": 1.7886663826161056, "grad_norm": 2.25, "learning_rate": 4.9774076818591495e-05, "loss": 1.385002040863037, "step": 20990 }, { "epoch": 1.7895185342991051, "grad_norm": 2.234375, "learning_rate": 4.977370646611689e-05, "loss": 1.3952081680297852, "step": 21000 }, { "epoch": 1.7903706859821047, "grad_norm": 2.546875, "learning_rate": 4.977333581171431e-05, "loss": 1.440528678894043, "step": 21010 }, { "epoch": 1.7912228376651043, "grad_norm": 2.25, "learning_rate": 4.9772964855388284e-05, "loss": 1.3212525367736816, "step": 21020 }, { "epoch": 1.7920749893481038, "grad_norm": 1.984375, "learning_rate": 4.977259359714331e-05, "loss": 1.3821983337402344, "step": 21030 }, { "epoch": 1.7929271410311034, "grad_norm": 2.859375, "learning_rate": 4.977222203698394e-05, "loss": 1.3343013763427733, "step": 21040 }, { "epoch": 1.793779292714103, "grad_norm": 3.765625, "learning_rate": 4.977185017491469e-05, "loss": 1.268205165863037, "step": 21050 }, { "epoch": 1.7946314443971025, "grad_norm": 2.46875, "learning_rate": 4.977147801094009e-05, "loss": 1.4877119064331055, "step": 21060 }, { "epoch": 1.795483596080102, "grad_norm": 1.921875, "learning_rate": 4.977110554506468e-05, "loss": 1.5191075325012207, "step": 21070 }, { "epoch": 1.7963357477631017, "grad_norm": 1.8359375, "learning_rate": 4.9770732777293e-05, "loss": 1.344963836669922, "step": 21080 }, { "epoch": 1.7971878994461012, "grad_norm": 2.609375, "learning_rate": 4.977035970762959e-05, "loss": 1.473278522491455, "step": 21090 }, { "epoch": 1.7980400511291008, "grad_norm": 3.109375, "learning_rate": 4.9769986336079e-05, "loss": 1.4108113288879394, "step": 21100 }, { "epoch": 1.7988922028121006, "grad_norm": 2.78125, "learning_rate": 4.976961266264577e-05, "loss": 1.4069157600402833, "step": 21110 }, { "epoch": 1.7997443544951002, "grad_norm": 2.40625, "learning_rate": 4.976923868733447e-05, "loss": 1.3460987091064454, "step": 21120 }, { "epoch": 1.8005965061780997, "grad_norm": 1.5546875, "learning_rate": 4.976886441014965e-05, "loss": 1.3100528717041016, "step": 21130 }, { "epoch": 1.8014486578610993, "grad_norm": 3.3125, "learning_rate": 4.9768489831095866e-05, "loss": 1.284895420074463, "step": 21140 }, { "epoch": 1.8023008095440989, "grad_norm": 2.828125, "learning_rate": 4.976811495017769e-05, "loss": 1.466740322113037, "step": 21150 }, { "epoch": 1.8031529612270984, "grad_norm": 3.484375, "learning_rate": 4.976773976739969e-05, "loss": 1.2968111038208008, "step": 21160 }, { "epoch": 1.804005112910098, "grad_norm": 2.90625, "learning_rate": 4.976736428276644e-05, "loss": 1.5526827812194823, "step": 21170 }, { "epoch": 1.8048572645930976, "grad_norm": 1.7421875, "learning_rate": 4.976698849628252e-05, "loss": 1.5269868850708008, "step": 21180 }, { "epoch": 1.8057094162760972, "grad_norm": 2.953125, "learning_rate": 4.976661240795249e-05, "loss": 1.3007078170776367, "step": 21190 }, { "epoch": 1.8065615679590967, "grad_norm": 1.7265625, "learning_rate": 4.9766236017780966e-05, "loss": 1.2704954147338867, "step": 21200 }, { "epoch": 1.8074137196420963, "grad_norm": 3.875, "learning_rate": 4.97658593257725e-05, "loss": 1.3088015556335448, "step": 21210 }, { "epoch": 1.8082658713250959, "grad_norm": 2.328125, "learning_rate": 4.9765482331931714e-05, "loss": 1.4553078651428222, "step": 21220 }, { "epoch": 1.8091180230080954, "grad_norm": 1.8125, "learning_rate": 4.9765105036263184e-05, "loss": 1.396735954284668, "step": 21230 }, { "epoch": 1.809970174691095, "grad_norm": 2.390625, "learning_rate": 4.9764727438771505e-05, "loss": 1.5146677017211914, "step": 21240 }, { "epoch": 1.8108223263740946, "grad_norm": 1.9921875, "learning_rate": 4.97643495394613e-05, "loss": 1.4491458892822267, "step": 21250 }, { "epoch": 1.8116744780570941, "grad_norm": 2.296875, "learning_rate": 4.976397133833715e-05, "loss": 1.4962111473083497, "step": 21260 }, { "epoch": 1.8125266297400937, "grad_norm": 2.015625, "learning_rate": 4.976359283540368e-05, "loss": 1.2825480461120606, "step": 21270 }, { "epoch": 1.8133787814230933, "grad_norm": 3.390625, "learning_rate": 4.9763214030665496e-05, "loss": 1.4720367431640624, "step": 21280 }, { "epoch": 1.8142309331060928, "grad_norm": 1.9921875, "learning_rate": 4.976283492412722e-05, "loss": 1.3036224365234375, "step": 21290 }, { "epoch": 1.8150830847890924, "grad_norm": 2.828125, "learning_rate": 4.976245551579347e-05, "loss": 1.411031150817871, "step": 21300 }, { "epoch": 1.815935236472092, "grad_norm": 2.40625, "learning_rate": 4.9762075805668867e-05, "loss": 1.498557186126709, "step": 21310 }, { "epoch": 1.8167873881550916, "grad_norm": 1.453125, "learning_rate": 4.9761695793758046e-05, "loss": 1.2398493766784668, "step": 21320 }, { "epoch": 1.8176395398380913, "grad_norm": 2.28125, "learning_rate": 4.976131548006563e-05, "loss": 1.5654105186462401, "step": 21330 }, { "epoch": 1.818491691521091, "grad_norm": 2.8125, "learning_rate": 4.9760934864596256e-05, "loss": 1.4272167205810546, "step": 21340 }, { "epoch": 1.8193438432040905, "grad_norm": 2.859375, "learning_rate": 4.976055394735457e-05, "loss": 1.2981369018554687, "step": 21350 }, { "epoch": 1.82019599488709, "grad_norm": 2.96875, "learning_rate": 4.976017272834521e-05, "loss": 1.3892311096191405, "step": 21360 }, { "epoch": 1.8210481465700896, "grad_norm": 2.859375, "learning_rate": 4.9759791207572816e-05, "loss": 1.4490299224853516, "step": 21370 }, { "epoch": 1.8219002982530892, "grad_norm": 1.8125, "learning_rate": 4.9759409385042036e-05, "loss": 1.3245162963867188, "step": 21380 }, { "epoch": 1.8227524499360888, "grad_norm": 3.59375, "learning_rate": 4.975902726075754e-05, "loss": 1.3626397132873536, "step": 21390 }, { "epoch": 1.8236046016190883, "grad_norm": 2.34375, "learning_rate": 4.9758644834723975e-05, "loss": 1.292966079711914, "step": 21400 }, { "epoch": 1.824456753302088, "grad_norm": 2.8125, "learning_rate": 4.9758262106946e-05, "loss": 1.3154887199401855, "step": 21410 }, { "epoch": 1.8253089049850875, "grad_norm": 2.84375, "learning_rate": 4.975787907742828e-05, "loss": 1.2448860168457032, "step": 21420 }, { "epoch": 1.826161056668087, "grad_norm": 1.71875, "learning_rate": 4.9757495746175486e-05, "loss": 1.4855997085571289, "step": 21430 }, { "epoch": 1.8270132083510866, "grad_norm": 1.8515625, "learning_rate": 4.97571121131923e-05, "loss": 1.4554905891418457, "step": 21440 }, { "epoch": 1.8278653600340862, "grad_norm": 2.59375, "learning_rate": 4.975672817848337e-05, "loss": 1.4590920448303222, "step": 21450 }, { "epoch": 1.8287175117170857, "grad_norm": 3.390625, "learning_rate": 4.9756343942053396e-05, "loss": 1.2947312355041505, "step": 21460 }, { "epoch": 1.8295696634000853, "grad_norm": 2.890625, "learning_rate": 4.9755959403907055e-05, "loss": 1.3470290184020997, "step": 21470 }, { "epoch": 1.8304218150830849, "grad_norm": 2.8125, "learning_rate": 4.9755574564049035e-05, "loss": 1.481997013092041, "step": 21480 }, { "epoch": 1.8312739667660844, "grad_norm": 2.453125, "learning_rate": 4.9755189422484026e-05, "loss": 1.4030173301696778, "step": 21490 }, { "epoch": 1.832126118449084, "grad_norm": 2.296875, "learning_rate": 4.975480397921672e-05, "loss": 1.2658704757690429, "step": 21500 }, { "epoch": 1.8329782701320836, "grad_norm": 1.8046875, "learning_rate": 4.975441823425182e-05, "loss": 1.5718502044677733, "step": 21510 }, { "epoch": 1.8338304218150832, "grad_norm": 2.328125, "learning_rate": 4.9754032187594026e-05, "loss": 1.3820067405700684, "step": 21520 }, { "epoch": 1.8346825734980827, "grad_norm": 3.078125, "learning_rate": 4.975364583924804e-05, "loss": 1.555393600463867, "step": 21530 }, { "epoch": 1.8355347251810823, "grad_norm": 2.796875, "learning_rate": 4.975325918921856e-05, "loss": 1.2885440826416015, "step": 21540 }, { "epoch": 1.8363868768640819, "grad_norm": 2.015625, "learning_rate": 4.9752872237510315e-05, "loss": 1.502800178527832, "step": 21550 }, { "epoch": 1.8372390285470814, "grad_norm": 2.890625, "learning_rate": 4.975248498412802e-05, "loss": 1.4594696044921875, "step": 21560 }, { "epoch": 1.838091180230081, "grad_norm": 3.3125, "learning_rate": 4.97520974290764e-05, "loss": 1.445231819152832, "step": 21570 }, { "epoch": 1.8389433319130806, "grad_norm": 2.046875, "learning_rate": 4.9751709572360144e-05, "loss": 1.4846868515014648, "step": 21580 }, { "epoch": 1.8397954835960801, "grad_norm": 2.125, "learning_rate": 4.975132141398402e-05, "loss": 1.457199764251709, "step": 21590 }, { "epoch": 1.8406476352790797, "grad_norm": 3.546875, "learning_rate": 4.9750932953952736e-05, "loss": 1.2867145538330078, "step": 21600 }, { "epoch": 1.8414997869620793, "grad_norm": 2.5, "learning_rate": 4.975054419227103e-05, "loss": 1.2111575126647949, "step": 21610 }, { "epoch": 1.8423519386450788, "grad_norm": 3.34375, "learning_rate": 4.975015512894364e-05, "loss": 1.3429107666015625, "step": 21620 }, { "epoch": 1.8432040903280784, "grad_norm": 3.421875, "learning_rate": 4.974976576397532e-05, "loss": 1.5075693130493164, "step": 21630 }, { "epoch": 1.844056242011078, "grad_norm": 1.734375, "learning_rate": 4.97493760973708e-05, "loss": 1.2803661346435546, "step": 21640 }, { "epoch": 1.8449083936940776, "grad_norm": 2.03125, "learning_rate": 4.974898612913483e-05, "loss": 1.4132699012756347, "step": 21650 }, { "epoch": 1.8457605453770771, "grad_norm": 2.53125, "learning_rate": 4.974859585927217e-05, "loss": 1.325773048400879, "step": 21660 }, { "epoch": 1.8466126970600767, "grad_norm": 4.59375, "learning_rate": 4.9748205287787575e-05, "loss": 1.5561424255371095, "step": 21670 }, { "epoch": 1.8474648487430763, "grad_norm": 2.171875, "learning_rate": 4.9747814414685806e-05, "loss": 1.1767801284790038, "step": 21680 }, { "epoch": 1.8483170004260758, "grad_norm": 1.96875, "learning_rate": 4.9747423239971624e-05, "loss": 1.5942469596862794, "step": 21690 }, { "epoch": 1.8491691521090754, "grad_norm": 2.34375, "learning_rate": 4.9747031763649785e-05, "loss": 1.3200909614562988, "step": 21700 }, { "epoch": 1.850021303792075, "grad_norm": 2.046875, "learning_rate": 4.974663998572509e-05, "loss": 1.276998519897461, "step": 21710 }, { "epoch": 1.8508734554750745, "grad_norm": 1.4765625, "learning_rate": 4.974624790620228e-05, "loss": 1.4375268936157226, "step": 21720 }, { "epoch": 1.851725607158074, "grad_norm": 1.9140625, "learning_rate": 4.974585552508616e-05, "loss": 1.4600202560424804, "step": 21730 }, { "epoch": 1.8525777588410737, "grad_norm": 1.7890625, "learning_rate": 4.974546284238149e-05, "loss": 1.482541561126709, "step": 21740 }, { "epoch": 1.8534299105240732, "grad_norm": 1.7890625, "learning_rate": 4.974506985809307e-05, "loss": 1.259676456451416, "step": 21750 }, { "epoch": 1.8542820622070728, "grad_norm": 2.125, "learning_rate": 4.974467657222569e-05, "loss": 1.4758981704711913, "step": 21760 }, { "epoch": 1.8551342138900724, "grad_norm": 3.59375, "learning_rate": 4.974428298478414e-05, "loss": 1.4223085403442384, "step": 21770 }, { "epoch": 1.855986365573072, "grad_norm": 2.03125, "learning_rate": 4.974388909577321e-05, "loss": 1.3271449089050293, "step": 21780 }, { "epoch": 1.8568385172560715, "grad_norm": 2.46875, "learning_rate": 4.974349490519771e-05, "loss": 1.2850196838378907, "step": 21790 }, { "epoch": 1.857690668939071, "grad_norm": 2.28125, "learning_rate": 4.974310041306244e-05, "loss": 1.4448771476745605, "step": 21800 }, { "epoch": 1.8585428206220707, "grad_norm": 2.34375, "learning_rate": 4.974270561937222e-05, "loss": 1.3539199829101562, "step": 21810 }, { "epoch": 1.8593949723050702, "grad_norm": 2.484375, "learning_rate": 4.974231052413184e-05, "loss": 1.3103466987609864, "step": 21820 }, { "epoch": 1.8602471239880698, "grad_norm": 1.7734375, "learning_rate": 4.9741915127346125e-05, "loss": 1.422840404510498, "step": 21830 }, { "epoch": 1.8610992756710694, "grad_norm": 4.03125, "learning_rate": 4.974151942901989e-05, "loss": 1.4040799140930176, "step": 21840 }, { "epoch": 1.861951427354069, "grad_norm": 3.515625, "learning_rate": 4.974112342915797e-05, "loss": 1.2889034271240234, "step": 21850 }, { "epoch": 1.8628035790370685, "grad_norm": 1.7578125, "learning_rate": 4.974072712776518e-05, "loss": 1.284374713897705, "step": 21860 }, { "epoch": 1.863655730720068, "grad_norm": 3.28125, "learning_rate": 4.974033052484636e-05, "loss": 1.3754677772521973, "step": 21870 }, { "epoch": 1.8645078824030676, "grad_norm": 2.40625, "learning_rate": 4.973993362040632e-05, "loss": 1.369448471069336, "step": 21880 }, { "epoch": 1.8653600340860672, "grad_norm": 2.234375, "learning_rate": 4.973953641444993e-05, "loss": 1.2620656967163086, "step": 21890 }, { "epoch": 1.8662121857690668, "grad_norm": 2.578125, "learning_rate": 4.9739138906982006e-05, "loss": 1.3183346748352052, "step": 21900 }, { "epoch": 1.8670643374520663, "grad_norm": 1.9765625, "learning_rate": 4.97387410980074e-05, "loss": 1.4871881484985352, "step": 21910 }, { "epoch": 1.867916489135066, "grad_norm": 2.359375, "learning_rate": 4.973834298753097e-05, "loss": 1.4543505668640138, "step": 21920 }, { "epoch": 1.8687686408180655, "grad_norm": 2.25, "learning_rate": 4.973794457555756e-05, "loss": 1.408955192565918, "step": 21930 }, { "epoch": 1.869620792501065, "grad_norm": 3.609375, "learning_rate": 4.973754586209201e-05, "loss": 1.457133960723877, "step": 21940 }, { "epoch": 1.8704729441840646, "grad_norm": 2.890625, "learning_rate": 4.973714684713921e-05, "loss": 1.303851318359375, "step": 21950 }, { "epoch": 1.8713250958670642, "grad_norm": 1.9375, "learning_rate": 4.9736747530704e-05, "loss": 1.4116884231567384, "step": 21960 }, { "epoch": 1.8721772475500638, "grad_norm": 1.9609375, "learning_rate": 4.973634791279126e-05, "loss": 1.4068601608276368, "step": 21970 }, { "epoch": 1.8730293992330633, "grad_norm": 1.671875, "learning_rate": 4.973594799340585e-05, "loss": 1.4547730445861817, "step": 21980 }, { "epoch": 1.873881550916063, "grad_norm": 1.7421875, "learning_rate": 4.973554777255265e-05, "loss": 1.4476643562316895, "step": 21990 }, { "epoch": 1.8747337025990625, "grad_norm": 2.53125, "learning_rate": 4.973514725023654e-05, "loss": 1.434494686126709, "step": 22000 }, { "epoch": 1.8755858542820623, "grad_norm": 2.234375, "learning_rate": 4.973474642646239e-05, "loss": 1.2982730865478516, "step": 22010 }, { "epoch": 1.8764380059650618, "grad_norm": 1.96875, "learning_rate": 4.973434530123509e-05, "loss": 1.2475129127502442, "step": 22020 }, { "epoch": 1.8772901576480614, "grad_norm": 2.796875, "learning_rate": 4.973394387455954e-05, "loss": 1.4109302520751954, "step": 22030 }, { "epoch": 1.878142309331061, "grad_norm": 2.265625, "learning_rate": 4.9733542146440615e-05, "loss": 1.3582574844360351, "step": 22040 }, { "epoch": 1.8789944610140605, "grad_norm": 2.625, "learning_rate": 4.9733140116883224e-05, "loss": 1.412968063354492, "step": 22050 }, { "epoch": 1.87984661269706, "grad_norm": 3.09375, "learning_rate": 4.9732737785892266e-05, "loss": 1.1701102256774902, "step": 22060 }, { "epoch": 1.8806987643800597, "grad_norm": 2.0625, "learning_rate": 4.973233515347263e-05, "loss": 1.2939085006713866, "step": 22070 }, { "epoch": 1.8815509160630592, "grad_norm": 2.8125, "learning_rate": 4.9731932219629245e-05, "loss": 1.1971869468688965, "step": 22080 }, { "epoch": 1.8824030677460588, "grad_norm": 2.875, "learning_rate": 4.9731528984367007e-05, "loss": 1.3142958641052247, "step": 22090 }, { "epoch": 1.8832552194290584, "grad_norm": 2.015625, "learning_rate": 4.973112544769083e-05, "loss": 1.3665353775024414, "step": 22100 }, { "epoch": 1.884107371112058, "grad_norm": 1.7265625, "learning_rate": 4.973072160960564e-05, "loss": 1.3534710884094239, "step": 22110 }, { "epoch": 1.8849595227950575, "grad_norm": 1.7578125, "learning_rate": 4.973031747011635e-05, "loss": 1.4024744033813477, "step": 22120 }, { "epoch": 1.885811674478057, "grad_norm": 2.484375, "learning_rate": 4.972991302922789e-05, "loss": 1.3242775917053222, "step": 22130 }, { "epoch": 1.8866638261610567, "grad_norm": 2.328125, "learning_rate": 4.97295082869452e-05, "loss": 1.2407726287841796, "step": 22140 }, { "epoch": 1.8875159778440562, "grad_norm": 2.234375, "learning_rate": 4.972910324327319e-05, "loss": 1.495288372039795, "step": 22150 }, { "epoch": 1.8883681295270558, "grad_norm": 2.28125, "learning_rate": 4.972869789821681e-05, "loss": 1.3166879653930663, "step": 22160 }, { "epoch": 1.8892202812100554, "grad_norm": 3.71875, "learning_rate": 4.9728292251781005e-05, "loss": 1.365666961669922, "step": 22170 }, { "epoch": 1.890072432893055, "grad_norm": 1.6640625, "learning_rate": 4.972788630397071e-05, "loss": 1.4262738227844238, "step": 22180 }, { "epoch": 1.8909245845760545, "grad_norm": 3.296875, "learning_rate": 4.9727480054790875e-05, "loss": 1.206150436401367, "step": 22190 }, { "epoch": 1.891776736259054, "grad_norm": 1.9296875, "learning_rate": 4.9727073504246445e-05, "loss": 1.5337434768676759, "step": 22200 }, { "epoch": 1.8926288879420536, "grad_norm": 2.171875, "learning_rate": 4.972666665234239e-05, "loss": 1.1507040023803712, "step": 22210 }, { "epoch": 1.8934810396250532, "grad_norm": 2.140625, "learning_rate": 4.972625949908365e-05, "loss": 1.5025056838989257, "step": 22220 }, { "epoch": 1.894333191308053, "grad_norm": 3.53125, "learning_rate": 4.97258520444752e-05, "loss": 1.3398792266845703, "step": 22230 }, { "epoch": 1.8951853429910526, "grad_norm": 2.046875, "learning_rate": 4.9725444288522004e-05, "loss": 1.4028907775878907, "step": 22240 }, { "epoch": 1.8960374946740521, "grad_norm": 1.53125, "learning_rate": 4.9725036231229025e-05, "loss": 1.3917606353759766, "step": 22250 }, { "epoch": 1.8968896463570517, "grad_norm": 2.65625, "learning_rate": 4.972462787260125e-05, "loss": 1.4165433883666991, "step": 22260 }, { "epoch": 1.8977417980400513, "grad_norm": 2.9375, "learning_rate": 4.9724219212643636e-05, "loss": 1.4212735176086426, "step": 22270 }, { "epoch": 1.8985939497230508, "grad_norm": 3.75, "learning_rate": 4.972381025136118e-05, "loss": 1.3570161819458009, "step": 22280 }, { "epoch": 1.8994461014060504, "grad_norm": 2.9375, "learning_rate": 4.972340098875886e-05, "loss": 1.3355692863464355, "step": 22290 }, { "epoch": 1.90029825308905, "grad_norm": 2.78125, "learning_rate": 4.972299142484166e-05, "loss": 1.3557624816894531, "step": 22300 }, { "epoch": 1.9011504047720496, "grad_norm": 1.9453125, "learning_rate": 4.9722581559614575e-05, "loss": 1.4116636276245118, "step": 22310 }, { "epoch": 1.9020025564550491, "grad_norm": 2.09375, "learning_rate": 4.9722171393082604e-05, "loss": 1.4369298934936523, "step": 22320 }, { "epoch": 1.9028547081380487, "grad_norm": 2.703125, "learning_rate": 4.972176092525075e-05, "loss": 1.4592705726623536, "step": 22330 }, { "epoch": 1.9037068598210483, "grad_norm": 2.265625, "learning_rate": 4.9721350156124e-05, "loss": 1.334829330444336, "step": 22340 }, { "epoch": 1.9045590115040478, "grad_norm": 2.53125, "learning_rate": 4.9720939085707375e-05, "loss": 1.4498994827270508, "step": 22350 }, { "epoch": 1.9054111631870474, "grad_norm": 2.703125, "learning_rate": 4.9720527714005874e-05, "loss": 1.3840163230895997, "step": 22360 }, { "epoch": 1.906263314870047, "grad_norm": 2.984375, "learning_rate": 4.972011604102451e-05, "loss": 1.3680339813232423, "step": 22370 }, { "epoch": 1.9071154665530465, "grad_norm": 2.328125, "learning_rate": 4.9719704066768316e-05, "loss": 1.425376033782959, "step": 22380 }, { "epoch": 1.907967618236046, "grad_norm": 3.265625, "learning_rate": 4.9719291791242294e-05, "loss": 1.3960748672485352, "step": 22390 }, { "epoch": 1.9088197699190457, "grad_norm": 2.53125, "learning_rate": 4.9718879214451475e-05, "loss": 1.4177465438842773, "step": 22400 }, { "epoch": 1.9096719216020452, "grad_norm": 1.8828125, "learning_rate": 4.971846633640089e-05, "loss": 1.46694974899292, "step": 22410 }, { "epoch": 1.9105240732850448, "grad_norm": 2.71875, "learning_rate": 4.971805315709558e-05, "loss": 1.526467514038086, "step": 22420 }, { "epoch": 1.9113762249680444, "grad_norm": 1.65625, "learning_rate": 4.971763967654056e-05, "loss": 1.436185073852539, "step": 22430 }, { "epoch": 1.912228376651044, "grad_norm": 4.78125, "learning_rate": 4.9717225894740886e-05, "loss": 1.345252227783203, "step": 22440 }, { "epoch": 1.9130805283340435, "grad_norm": 2.265625, "learning_rate": 4.971681181170159e-05, "loss": 1.40169677734375, "step": 22450 }, { "epoch": 1.913932680017043, "grad_norm": 2.109375, "learning_rate": 4.971639742742773e-05, "loss": 1.338198184967041, "step": 22460 }, { "epoch": 1.9147848317000427, "grad_norm": 2.8125, "learning_rate": 4.971598274192435e-05, "loss": 1.2337456703186036, "step": 22470 }, { "epoch": 1.9156369833830422, "grad_norm": 3.9375, "learning_rate": 4.9715567755196495e-05, "loss": 1.2537007331848145, "step": 22480 }, { "epoch": 1.9164891350660418, "grad_norm": 2.671875, "learning_rate": 4.9715152467249235e-05, "loss": 1.348224925994873, "step": 22490 }, { "epoch": 1.9173412867490414, "grad_norm": 2.046875, "learning_rate": 4.9714736878087634e-05, "loss": 1.4350197792053223, "step": 22500 }, { "epoch": 1.918193438432041, "grad_norm": 2.640625, "learning_rate": 4.9714320987716745e-05, "loss": 1.4834203720092773, "step": 22510 }, { "epoch": 1.9190455901150405, "grad_norm": 3.984375, "learning_rate": 4.971390479614164e-05, "loss": 1.3520346641540528, "step": 22520 }, { "epoch": 1.91989774179804, "grad_norm": 2.1875, "learning_rate": 4.9713488303367396e-05, "loss": 1.242811679840088, "step": 22530 }, { "epoch": 1.9207498934810396, "grad_norm": 1.8125, "learning_rate": 4.9713071509399086e-05, "loss": 1.3531803131103515, "step": 22540 }, { "epoch": 1.9216020451640392, "grad_norm": 2.0, "learning_rate": 4.971265441424179e-05, "loss": 1.386525821685791, "step": 22550 }, { "epoch": 1.9224541968470388, "grad_norm": 1.9375, "learning_rate": 4.971223701790059e-05, "loss": 1.4529544830322265, "step": 22560 }, { "epoch": 1.9233063485300383, "grad_norm": 2.046875, "learning_rate": 4.9711819320380576e-05, "loss": 1.4859209060668945, "step": 22570 }, { "epoch": 1.924158500213038, "grad_norm": 2.53125, "learning_rate": 4.971140132168683e-05, "loss": 1.3101490974426269, "step": 22580 }, { "epoch": 1.9250106518960375, "grad_norm": 1.5625, "learning_rate": 4.971098302182447e-05, "loss": 1.3672830581665039, "step": 22590 }, { "epoch": 1.925862803579037, "grad_norm": 3.171875, "learning_rate": 4.971056442079857e-05, "loss": 1.381517505645752, "step": 22600 }, { "epoch": 1.9267149552620366, "grad_norm": 2.328125, "learning_rate": 4.9710145518614225e-05, "loss": 1.3430731773376465, "step": 22610 }, { "epoch": 1.9275671069450362, "grad_norm": 2.015625, "learning_rate": 4.970972631527657e-05, "loss": 1.452279281616211, "step": 22620 }, { "epoch": 1.9284192586280358, "grad_norm": 3.4375, "learning_rate": 4.9709306810790696e-05, "loss": 1.424824333190918, "step": 22630 }, { "epoch": 1.9292714103110353, "grad_norm": 2.40625, "learning_rate": 4.970888700516172e-05, "loss": 1.3265213012695312, "step": 22640 }, { "epoch": 1.930123561994035, "grad_norm": 2.78125, "learning_rate": 4.970846689839475e-05, "loss": 1.4786321640014648, "step": 22650 }, { "epoch": 1.9309757136770345, "grad_norm": 2.65625, "learning_rate": 4.970804649049492e-05, "loss": 1.4677753448486328, "step": 22660 }, { "epoch": 1.931827865360034, "grad_norm": 2.265625, "learning_rate": 4.970762578146734e-05, "loss": 1.457444190979004, "step": 22670 }, { "epoch": 1.9326800170430336, "grad_norm": 1.921875, "learning_rate": 4.970720477131714e-05, "loss": 1.4342989921569824, "step": 22680 }, { "epoch": 1.9335321687260332, "grad_norm": 2.078125, "learning_rate": 4.970678346004947e-05, "loss": 1.4489150047302246, "step": 22690 }, { "epoch": 1.9343843204090327, "grad_norm": 2.015625, "learning_rate": 4.970636184766944e-05, "loss": 1.2291653633117676, "step": 22700 }, { "epoch": 1.9352364720920323, "grad_norm": 2.328125, "learning_rate": 4.970593993418219e-05, "loss": 1.4058175086975098, "step": 22710 }, { "epoch": 1.9360886237750319, "grad_norm": 3.046875, "learning_rate": 4.9705517719592876e-05, "loss": 1.3717477798461915, "step": 22720 }, { "epoch": 1.9369407754580314, "grad_norm": 2.484375, "learning_rate": 4.970509520390664e-05, "loss": 1.2957273483276368, "step": 22730 }, { "epoch": 1.937792927141031, "grad_norm": 2.65625, "learning_rate": 4.9704672387128625e-05, "loss": 1.3847367286682128, "step": 22740 }, { "epoch": 1.9386450788240306, "grad_norm": 3.453125, "learning_rate": 4.970424926926399e-05, "loss": 1.3540192604064942, "step": 22750 }, { "epoch": 1.9394972305070302, "grad_norm": 3.40625, "learning_rate": 4.970382585031789e-05, "loss": 1.3698911666870117, "step": 22760 }, { "epoch": 1.9403493821900297, "grad_norm": 2.5625, "learning_rate": 4.970340213029548e-05, "loss": 1.5141324043273925, "step": 22770 }, { "epoch": 1.9412015338730293, "grad_norm": 3.671875, "learning_rate": 4.9702978109201935e-05, "loss": 1.3325658798217774, "step": 22780 }, { "epoch": 1.9420536855560289, "grad_norm": 2.0, "learning_rate": 4.9702553787042414e-05, "loss": 1.452474594116211, "step": 22790 }, { "epoch": 1.9429058372390284, "grad_norm": 2.796875, "learning_rate": 4.97021291638221e-05, "loss": 1.2936758041381835, "step": 22800 }, { "epoch": 1.943757988922028, "grad_norm": 2.53125, "learning_rate": 4.9701704239546145e-05, "loss": 1.3378072738647462, "step": 22810 }, { "epoch": 1.9446101406050276, "grad_norm": 2.96875, "learning_rate": 4.9701279014219746e-05, "loss": 1.3004425048828125, "step": 22820 }, { "epoch": 1.9454622922880271, "grad_norm": 1.8203125, "learning_rate": 4.9700853487848086e-05, "loss": 1.3808444023132325, "step": 22830 }, { "epoch": 1.9463144439710267, "grad_norm": 2.234375, "learning_rate": 4.970042766043634e-05, "loss": 1.362375545501709, "step": 22840 }, { "epoch": 1.9471665956540263, "grad_norm": 2.3125, "learning_rate": 4.970000153198971e-05, "loss": 1.3962382316589355, "step": 22850 }, { "epoch": 1.9480187473370258, "grad_norm": 4.65625, "learning_rate": 4.969957510251339e-05, "loss": 1.3570520401000976, "step": 22860 }, { "epoch": 1.9488708990200254, "grad_norm": 2.078125, "learning_rate": 4.969914837201256e-05, "loss": 1.441751194000244, "step": 22870 }, { "epoch": 1.949723050703025, "grad_norm": 1.703125, "learning_rate": 4.9698721340492435e-05, "loss": 1.279981517791748, "step": 22880 }, { "epoch": 1.9505752023860246, "grad_norm": 3.3125, "learning_rate": 4.9698294007958215e-05, "loss": 1.3655280113220214, "step": 22890 }, { "epoch": 1.9514273540690241, "grad_norm": 2.25, "learning_rate": 4.969786637441511e-05, "loss": 1.3454078674316405, "step": 22900 }, { "epoch": 1.952279505752024, "grad_norm": 2.609375, "learning_rate": 4.969743843986833e-05, "loss": 1.3736822128295898, "step": 22910 }, { "epoch": 1.9531316574350235, "grad_norm": 2.140625, "learning_rate": 4.969701020432309e-05, "loss": 1.365242862701416, "step": 22920 }, { "epoch": 1.953983809118023, "grad_norm": 1.828125, "learning_rate": 4.969658166778461e-05, "loss": 1.349863624572754, "step": 22930 }, { "epoch": 1.9548359608010226, "grad_norm": 2.796875, "learning_rate": 4.969615283025811e-05, "loss": 1.426210594177246, "step": 22940 }, { "epoch": 1.9556881124840222, "grad_norm": 4.09375, "learning_rate": 4.969572369174883e-05, "loss": 1.1607080459594727, "step": 22950 }, { "epoch": 1.9565402641670218, "grad_norm": 2.171875, "learning_rate": 4.969529425226198e-05, "loss": 1.3788677215576173, "step": 22960 }, { "epoch": 1.9573924158500213, "grad_norm": 2.4375, "learning_rate": 4.9694864511802805e-05, "loss": 1.2552536010742188, "step": 22970 }, { "epoch": 1.958244567533021, "grad_norm": 2.171875, "learning_rate": 4.969443447037654e-05, "loss": 1.1421968460083007, "step": 22980 }, { "epoch": 1.9590967192160205, "grad_norm": 2.5625, "learning_rate": 4.969400412798843e-05, "loss": 1.1600841522216796, "step": 22990 }, { "epoch": 1.95994887089902, "grad_norm": 3.390625, "learning_rate": 4.969357348464372e-05, "loss": 1.3974392890930176, "step": 23000 }, { "epoch": 1.9608010225820196, "grad_norm": 2.109375, "learning_rate": 4.969314254034765e-05, "loss": 1.258173942565918, "step": 23010 }, { "epoch": 1.9616531742650192, "grad_norm": 1.8359375, "learning_rate": 4.9692711295105475e-05, "loss": 1.3603211402893067, "step": 23020 }, { "epoch": 1.9625053259480187, "grad_norm": 2.03125, "learning_rate": 4.969227974892246e-05, "loss": 1.4605744361877442, "step": 23030 }, { "epoch": 1.9633574776310183, "grad_norm": 1.671875, "learning_rate": 4.9691847901803845e-05, "loss": 1.4163185119628907, "step": 23040 }, { "epoch": 1.9642096293140179, "grad_norm": 2.28125, "learning_rate": 4.969141575375492e-05, "loss": 1.340340805053711, "step": 23050 }, { "epoch": 1.9650617809970174, "grad_norm": 1.6796875, "learning_rate": 4.969098330478092e-05, "loss": 1.3933826446533204, "step": 23060 }, { "epoch": 1.965913932680017, "grad_norm": 3.71875, "learning_rate": 4.969055055488715e-05, "loss": 1.331510353088379, "step": 23070 }, { "epoch": 1.9667660843630166, "grad_norm": 1.453125, "learning_rate": 4.969011750407886e-05, "loss": 1.3718304634094238, "step": 23080 }, { "epoch": 1.9676182360460162, "grad_norm": 2.140625, "learning_rate": 4.9689684152361335e-05, "loss": 1.3747085571289062, "step": 23090 }, { "epoch": 1.9684703877290157, "grad_norm": 1.796875, "learning_rate": 4.968925049973986e-05, "loss": 1.3516514778137207, "step": 23100 }, { "epoch": 1.9693225394120153, "grad_norm": 2.4375, "learning_rate": 4.968881654621971e-05, "loss": 1.364661979675293, "step": 23110 }, { "epoch": 1.9701746910950149, "grad_norm": 2.75, "learning_rate": 4.968838229180618e-05, "loss": 1.3673133850097656, "step": 23120 }, { "epoch": 1.9710268427780144, "grad_norm": 2.328125, "learning_rate": 4.9687947736504564e-05, "loss": 1.3622868537902832, "step": 23130 }, { "epoch": 1.9718789944610142, "grad_norm": 3.5, "learning_rate": 4.968751288032015e-05, "loss": 1.3601466178894044, "step": 23140 }, { "epoch": 1.9727311461440138, "grad_norm": 1.890625, "learning_rate": 4.9687077723258257e-05, "loss": 1.3621827125549317, "step": 23150 }, { "epoch": 1.9735832978270134, "grad_norm": 2.625, "learning_rate": 4.968664226532417e-05, "loss": 1.3692332267761231, "step": 23160 }, { "epoch": 1.974435449510013, "grad_norm": 2.0625, "learning_rate": 4.9686206506523204e-05, "loss": 1.5044314384460449, "step": 23170 }, { "epoch": 1.9752876011930125, "grad_norm": 3.0, "learning_rate": 4.968577044686066e-05, "loss": 1.3748154640197754, "step": 23180 }, { "epoch": 1.976139752876012, "grad_norm": 1.9921875, "learning_rate": 4.9685334086341866e-05, "loss": 1.4001843452453613, "step": 23190 }, { "epoch": 1.9769919045590116, "grad_norm": 1.8671875, "learning_rate": 4.968489742497213e-05, "loss": 1.4280717849731446, "step": 23200 }, { "epoch": 1.9778440562420112, "grad_norm": 2.921875, "learning_rate": 4.9684460462756785e-05, "loss": 1.3943352699279785, "step": 23210 }, { "epoch": 1.9786962079250108, "grad_norm": 3.765625, "learning_rate": 4.9684023199701147e-05, "loss": 1.3302146911621093, "step": 23220 }, { "epoch": 1.9795483596080103, "grad_norm": 1.4375, "learning_rate": 4.968358563581055e-05, "loss": 1.4008186340332032, "step": 23230 }, { "epoch": 1.98040051129101, "grad_norm": 2.25, "learning_rate": 4.968314777109031e-05, "loss": 1.4806915283203126, "step": 23240 }, { "epoch": 1.9812526629740095, "grad_norm": 2.609375, "learning_rate": 4.9682709605545795e-05, "loss": 1.3543593406677246, "step": 23250 }, { "epoch": 1.982104814657009, "grad_norm": 1.9609375, "learning_rate": 4.968227113918232e-05, "loss": 1.2573406219482421, "step": 23260 }, { "epoch": 1.9829569663400086, "grad_norm": 2.015625, "learning_rate": 4.968183237200523e-05, "loss": 1.213313102722168, "step": 23270 }, { "epoch": 1.9838091180230082, "grad_norm": 2.671875, "learning_rate": 4.9681393304019894e-05, "loss": 1.4247576713562011, "step": 23280 }, { "epoch": 1.9846612697060078, "grad_norm": 2.625, "learning_rate": 4.968095393523164e-05, "loss": 1.464940643310547, "step": 23290 }, { "epoch": 1.9855134213890073, "grad_norm": 2.6875, "learning_rate": 4.968051426564583e-05, "loss": 1.510385227203369, "step": 23300 }, { "epoch": 1.986365573072007, "grad_norm": 1.65625, "learning_rate": 4.9680074295267826e-05, "loss": 1.2876214981079102, "step": 23310 }, { "epoch": 1.9872177247550065, "grad_norm": 3.15625, "learning_rate": 4.9679634024102984e-05, "loss": 1.356037712097168, "step": 23320 }, { "epoch": 1.988069876438006, "grad_norm": 2.703125, "learning_rate": 4.967919345215668e-05, "loss": 1.432675552368164, "step": 23330 }, { "epoch": 1.9889220281210056, "grad_norm": 2.84375, "learning_rate": 4.967875257943427e-05, "loss": 1.3051676750183105, "step": 23340 }, { "epoch": 1.9897741798040052, "grad_norm": 1.9140625, "learning_rate": 4.9678311405941135e-05, "loss": 1.4594215393066405, "step": 23350 }, { "epoch": 1.9906263314870047, "grad_norm": 2.09375, "learning_rate": 4.967786993168265e-05, "loss": 1.432064914703369, "step": 23360 }, { "epoch": 1.9914784831700043, "grad_norm": 2.296875, "learning_rate": 4.967742815666421e-05, "loss": 1.3358946800231934, "step": 23370 }, { "epoch": 1.9923306348530039, "grad_norm": 2.71875, "learning_rate": 4.9676986080891166e-05, "loss": 1.450834369659424, "step": 23380 }, { "epoch": 1.9931827865360034, "grad_norm": 3.171875, "learning_rate": 4.967654370436894e-05, "loss": 1.3284770965576171, "step": 23390 }, { "epoch": 1.994034938219003, "grad_norm": 2.875, "learning_rate": 4.967610102710289e-05, "loss": 1.386086654663086, "step": 23400 }, { "epoch": 1.9948870899020026, "grad_norm": 2.15625, "learning_rate": 4.967565804909844e-05, "loss": 1.2815799713134766, "step": 23410 }, { "epoch": 1.9957392415850022, "grad_norm": 2.34375, "learning_rate": 4.967521477036098e-05, "loss": 1.4414682388305664, "step": 23420 }, { "epoch": 1.9965913932680017, "grad_norm": 2.0625, "learning_rate": 4.967477119089591e-05, "loss": 1.3811742782592773, "step": 23430 }, { "epoch": 1.9974435449510013, "grad_norm": 3.5, "learning_rate": 4.967432731070864e-05, "loss": 1.401411247253418, "step": 23440 }, { "epoch": 1.9982956966340009, "grad_norm": 2.015625, "learning_rate": 4.967388312980458e-05, "loss": 1.3572074890136718, "step": 23450 }, { "epoch": 1.9991478483170004, "grad_norm": 2.484375, "learning_rate": 4.9673438648189125e-05, "loss": 1.448636531829834, "step": 23460 }, { "epoch": 1.9998295696634, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6476, "eval_loss": 1.281622052192688, "eval_runtime": 55.042, "eval_samples_per_second": 359.035, "eval_steps_per_second": 22.456, "step": 23468 }, { "epoch": 2.0, "grad_norm": 1.875, "learning_rate": 4.967299386586771e-05, "loss": 1.3463414192199707, "step": 23470 }, { "epoch": 2.0008521516829996, "grad_norm": 2.296875, "learning_rate": 4.967254878284576e-05, "loss": 1.4144726753234864, "step": 23480 }, { "epoch": 2.001704303365999, "grad_norm": 1.8359375, "learning_rate": 4.967210339912869e-05, "loss": 1.3556209564208985, "step": 23490 }, { "epoch": 2.0025564550489987, "grad_norm": 3.84375, "learning_rate": 4.967165771472193e-05, "loss": 1.4562780380249023, "step": 23500 }, { "epoch": 2.0034086067319983, "grad_norm": 3.796875, "learning_rate": 4.967121172963092e-05, "loss": 1.4202564239501954, "step": 23510 }, { "epoch": 2.004260758414998, "grad_norm": 2.84375, "learning_rate": 4.9670765443861077e-05, "loss": 1.340042781829834, "step": 23520 }, { "epoch": 2.0051129100979974, "grad_norm": 1.3984375, "learning_rate": 4.967031885741785e-05, "loss": 1.3884523391723633, "step": 23530 }, { "epoch": 2.005965061780997, "grad_norm": 2.046875, "learning_rate": 4.9669871970306684e-05, "loss": 1.3646169662475587, "step": 23540 }, { "epoch": 2.0068172134639966, "grad_norm": 2.6875, "learning_rate": 4.966942478253302e-05, "loss": 1.3580389976501466, "step": 23550 }, { "epoch": 2.007669365146996, "grad_norm": 1.90625, "learning_rate": 4.966897729410233e-05, "loss": 1.3718331336975098, "step": 23560 }, { "epoch": 2.0085215168299957, "grad_norm": 3.59375, "learning_rate": 4.966852950502004e-05, "loss": 1.360445785522461, "step": 23570 }, { "epoch": 2.0093736685129953, "grad_norm": 2.78125, "learning_rate": 4.966808141529161e-05, "loss": 1.314264965057373, "step": 23580 }, { "epoch": 2.010225820195995, "grad_norm": 2.71875, "learning_rate": 4.966763302492251e-05, "loss": 1.2909079551696778, "step": 23590 }, { "epoch": 2.0110779718789944, "grad_norm": 2.4375, "learning_rate": 4.96671843339182e-05, "loss": 1.4897342681884767, "step": 23600 }, { "epoch": 2.011930123561994, "grad_norm": 2.34375, "learning_rate": 4.9666735342284156e-05, "loss": 1.3948707580566406, "step": 23610 }, { "epoch": 2.0127822752449935, "grad_norm": 3.40625, "learning_rate": 4.9666286050025845e-05, "loss": 1.410480785369873, "step": 23620 }, { "epoch": 2.013634426927993, "grad_norm": 2.03125, "learning_rate": 4.966583645714874e-05, "loss": 1.4564488410949707, "step": 23630 }, { "epoch": 2.0144865786109927, "grad_norm": 2.109375, "learning_rate": 4.966538656365833e-05, "loss": 1.4781578063964844, "step": 23640 }, { "epoch": 2.0153387302939922, "grad_norm": 3.03125, "learning_rate": 4.9664936369560087e-05, "loss": 1.3559017181396484, "step": 23650 }, { "epoch": 2.016190881976992, "grad_norm": 1.859375, "learning_rate": 4.96644858748595e-05, "loss": 1.3360591888427735, "step": 23660 }, { "epoch": 2.0170430336599914, "grad_norm": 2.4375, "learning_rate": 4.966403507956207e-05, "loss": 1.4178425788879394, "step": 23670 }, { "epoch": 2.017895185342991, "grad_norm": 2.0625, "learning_rate": 4.966358398367327e-05, "loss": 1.2713253021240234, "step": 23680 }, { "epoch": 2.0187473370259905, "grad_norm": 3.359375, "learning_rate": 4.966313258719862e-05, "loss": 1.4044387817382813, "step": 23690 }, { "epoch": 2.01959948870899, "grad_norm": 2.0, "learning_rate": 4.966268089014361e-05, "loss": 1.3328256607055664, "step": 23700 }, { "epoch": 2.0204516403919897, "grad_norm": 2.359375, "learning_rate": 4.966222889251374e-05, "loss": 1.4398807525634765, "step": 23710 }, { "epoch": 2.0213037920749892, "grad_norm": 2.53125, "learning_rate": 4.966177659431453e-05, "loss": 1.3910909652709962, "step": 23720 }, { "epoch": 2.022155943757989, "grad_norm": 2.09375, "learning_rate": 4.966132399555149e-05, "loss": 1.513206386566162, "step": 23730 }, { "epoch": 2.0230080954409884, "grad_norm": 2.75, "learning_rate": 4.9660871096230124e-05, "loss": 1.4796506881713867, "step": 23740 }, { "epoch": 2.023860247123988, "grad_norm": 2.859375, "learning_rate": 4.966041789635597e-05, "loss": 1.3560321807861329, "step": 23750 }, { "epoch": 2.0247123988069875, "grad_norm": 2.9375, "learning_rate": 4.9659964395934544e-05, "loss": 1.3918432235717773, "step": 23760 }, { "epoch": 2.025564550489987, "grad_norm": 2.109375, "learning_rate": 4.965951059497137e-05, "loss": 1.1850905418395996, "step": 23770 }, { "epoch": 2.0264167021729866, "grad_norm": 1.5390625, "learning_rate": 4.9659056493471976e-05, "loss": 1.4927727699279785, "step": 23780 }, { "epoch": 2.027268853855986, "grad_norm": 3.484375, "learning_rate": 4.96586020914419e-05, "loss": 1.4160856246948241, "step": 23790 }, { "epoch": 2.0281210055389858, "grad_norm": 2.359375, "learning_rate": 4.965814738888669e-05, "loss": 1.348812961578369, "step": 23800 }, { "epoch": 2.0289731572219853, "grad_norm": 2.8125, "learning_rate": 4.9657692385811875e-05, "loss": 1.247288417816162, "step": 23810 }, { "epoch": 2.029825308904985, "grad_norm": 1.859375, "learning_rate": 4.9657237082223e-05, "loss": 1.3730144500732422, "step": 23820 }, { "epoch": 2.0306774605879845, "grad_norm": 2.484375, "learning_rate": 4.965678147812561e-05, "loss": 1.338455581665039, "step": 23830 }, { "epoch": 2.031529612270984, "grad_norm": 2.03125, "learning_rate": 4.965632557352527e-05, "loss": 1.3742111206054688, "step": 23840 }, { "epoch": 2.0323817639539836, "grad_norm": 3.046875, "learning_rate": 4.9655869368427547e-05, "loss": 1.2750260353088378, "step": 23850 }, { "epoch": 2.033233915636983, "grad_norm": 3.84375, "learning_rate": 4.965541286283797e-05, "loss": 1.3602775573730468, "step": 23860 }, { "epoch": 2.0340860673199828, "grad_norm": 2.34375, "learning_rate": 4.965495605676213e-05, "loss": 1.3412894248962401, "step": 23870 }, { "epoch": 2.0349382190029823, "grad_norm": 2.734375, "learning_rate": 4.965449895020556e-05, "loss": 1.318477249145508, "step": 23880 }, { "epoch": 2.035790370685982, "grad_norm": 2.484375, "learning_rate": 4.965404154317387e-05, "loss": 1.4854175567626953, "step": 23890 }, { "epoch": 2.0366425223689815, "grad_norm": 2.40625, "learning_rate": 4.9653583835672625e-05, "loss": 1.4985317230224608, "step": 23900 }, { "epoch": 2.037494674051981, "grad_norm": 2.109375, "learning_rate": 4.965312582770738e-05, "loss": 1.4047099113464356, "step": 23910 }, { "epoch": 2.0383468257349806, "grad_norm": 2.640625, "learning_rate": 4.965266751928375e-05, "loss": 1.457968521118164, "step": 23920 }, { "epoch": 2.0391989774179806, "grad_norm": 1.921875, "learning_rate": 4.965220891040729e-05, "loss": 1.5037708282470703, "step": 23930 }, { "epoch": 2.04005112910098, "grad_norm": 2.609375, "learning_rate": 4.965175000108361e-05, "loss": 1.4586959838867188, "step": 23940 }, { "epoch": 2.0409032807839798, "grad_norm": 2.09375, "learning_rate": 4.9651290791318294e-05, "loss": 1.2126375198364259, "step": 23950 }, { "epoch": 2.0417554324669793, "grad_norm": 2.25, "learning_rate": 4.965083128111694e-05, "loss": 1.331644630432129, "step": 23960 }, { "epoch": 2.042607584149979, "grad_norm": 1.9609375, "learning_rate": 4.9650371470485144e-05, "loss": 1.4102209091186524, "step": 23970 }, { "epoch": 2.0434597358329785, "grad_norm": 3.015625, "learning_rate": 4.964991135942852e-05, "loss": 1.145601463317871, "step": 23980 }, { "epoch": 2.044311887515978, "grad_norm": 3.265625, "learning_rate": 4.964945094795267e-05, "loss": 1.3411053657531737, "step": 23990 }, { "epoch": 2.0451640391989776, "grad_norm": 1.75, "learning_rate": 4.964899023606321e-05, "loss": 1.451044464111328, "step": 24000 }, { "epoch": 2.046016190881977, "grad_norm": 1.875, "learning_rate": 4.964852922376575e-05, "loss": 1.459421730041504, "step": 24010 }, { "epoch": 2.0468683425649767, "grad_norm": 2.46875, "learning_rate": 4.96480679110659e-05, "loss": 1.3509794235229493, "step": 24020 }, { "epoch": 2.0477204942479763, "grad_norm": 2.34375, "learning_rate": 4.9647606297969296e-05, "loss": 1.3800724029541016, "step": 24030 }, { "epoch": 2.048572645930976, "grad_norm": 2.796875, "learning_rate": 4.9647144384481565e-05, "loss": 1.3648262023925781, "step": 24040 }, { "epoch": 2.0494247976139754, "grad_norm": 2.015625, "learning_rate": 4.964668217060833e-05, "loss": 1.3178338050842284, "step": 24050 }, { "epoch": 2.050276949296975, "grad_norm": 2.375, "learning_rate": 4.964621965635522e-05, "loss": 1.2613863945007324, "step": 24060 }, { "epoch": 2.0511291009799746, "grad_norm": 2.28125, "learning_rate": 4.9645756841727876e-05, "loss": 1.3065607070922851, "step": 24070 }, { "epoch": 2.051981252662974, "grad_norm": 2.1875, "learning_rate": 4.964529372673194e-05, "loss": 1.4650742530822753, "step": 24080 }, { "epoch": 2.0528334043459737, "grad_norm": 2.359375, "learning_rate": 4.964483031137306e-05, "loss": 1.4301856994628905, "step": 24090 }, { "epoch": 2.0536855560289733, "grad_norm": 2.40625, "learning_rate": 4.9644366595656874e-05, "loss": 1.4286704063415527, "step": 24100 }, { "epoch": 2.054537707711973, "grad_norm": 3.75, "learning_rate": 4.964390257958904e-05, "loss": 1.3800227165222168, "step": 24110 }, { "epoch": 2.0553898593949724, "grad_norm": 3.15625, "learning_rate": 4.9643438263175214e-05, "loss": 1.2089646339416504, "step": 24120 }, { "epoch": 2.056242011077972, "grad_norm": 2.046875, "learning_rate": 4.9642973646421046e-05, "loss": 1.402741050720215, "step": 24130 }, { "epoch": 2.0570941627609716, "grad_norm": 3.71875, "learning_rate": 4.9642508729332217e-05, "loss": 1.4370306015014649, "step": 24140 }, { "epoch": 2.057946314443971, "grad_norm": 2.140625, "learning_rate": 4.964204351191437e-05, "loss": 1.5668046951293946, "step": 24150 }, { "epoch": 2.0587984661269707, "grad_norm": 1.8984375, "learning_rate": 4.9641577994173195e-05, "loss": 1.3536869049072267, "step": 24160 }, { "epoch": 2.0596506178099703, "grad_norm": 1.90625, "learning_rate": 4.964111217611435e-05, "loss": 1.4292938232421875, "step": 24170 }, { "epoch": 2.06050276949297, "grad_norm": 2.109375, "learning_rate": 4.964064605774352e-05, "loss": 1.3815693855285645, "step": 24180 }, { "epoch": 2.0613549211759694, "grad_norm": 2.109375, "learning_rate": 4.964017963906639e-05, "loss": 1.331883144378662, "step": 24190 }, { "epoch": 2.062207072858969, "grad_norm": 2.578125, "learning_rate": 4.9639712920088636e-05, "loss": 1.3306509971618652, "step": 24200 }, { "epoch": 2.0630592245419686, "grad_norm": 3.5625, "learning_rate": 4.9639245900815955e-05, "loss": 1.2720467567443847, "step": 24210 }, { "epoch": 2.063911376224968, "grad_norm": 2.25, "learning_rate": 4.963877858125403e-05, "loss": 1.3642020225524902, "step": 24220 }, { "epoch": 2.0647635279079677, "grad_norm": 1.75, "learning_rate": 4.9638310961408554e-05, "loss": 1.3443536758422852, "step": 24230 }, { "epoch": 2.0656156795909673, "grad_norm": 2.875, "learning_rate": 4.963784304128523e-05, "loss": 1.5100900650024414, "step": 24240 }, { "epoch": 2.066467831273967, "grad_norm": 2.0625, "learning_rate": 4.963737482088977e-05, "loss": 1.3452932357788085, "step": 24250 }, { "epoch": 2.0673199829569664, "grad_norm": 1.5546875, "learning_rate": 4.963690630022787e-05, "loss": 1.4675504684448242, "step": 24260 }, { "epoch": 2.068172134639966, "grad_norm": 2.109375, "learning_rate": 4.963643747930524e-05, "loss": 1.3638057708740234, "step": 24270 }, { "epoch": 2.0690242863229655, "grad_norm": 1.9375, "learning_rate": 4.96359683581276e-05, "loss": 1.3126033782958983, "step": 24280 }, { "epoch": 2.069876438005965, "grad_norm": 1.546875, "learning_rate": 4.9635498936700665e-05, "loss": 1.3331330299377442, "step": 24290 }, { "epoch": 2.0707285896889647, "grad_norm": 2.4375, "learning_rate": 4.963502921503015e-05, "loss": 1.5007944107055664, "step": 24300 }, { "epoch": 2.0715807413719642, "grad_norm": 3.0, "learning_rate": 4.9634559193121786e-05, "loss": 1.464372730255127, "step": 24310 }, { "epoch": 2.072432893054964, "grad_norm": 3.59375, "learning_rate": 4.96340888709813e-05, "loss": 1.4842028617858887, "step": 24320 }, { "epoch": 2.0732850447379634, "grad_norm": 3.75, "learning_rate": 4.9633618248614424e-05, "loss": 1.304439163208008, "step": 24330 }, { "epoch": 2.074137196420963, "grad_norm": 2.140625, "learning_rate": 4.96331473260269e-05, "loss": 1.4162891387939454, "step": 24340 }, { "epoch": 2.0749893481039625, "grad_norm": 2.03125, "learning_rate": 4.9632676103224454e-05, "loss": 1.4842495918273926, "step": 24350 }, { "epoch": 2.075841499786962, "grad_norm": 3.703125, "learning_rate": 4.963220458021284e-05, "loss": 1.4685532569885253, "step": 24360 }, { "epoch": 2.0766936514699617, "grad_norm": 2.734375, "learning_rate": 4.963173275699779e-05, "loss": 1.44152774810791, "step": 24370 }, { "epoch": 2.0775458031529612, "grad_norm": 2.578125, "learning_rate": 4.9631260633585074e-05, "loss": 1.397609329223633, "step": 24380 }, { "epoch": 2.078397954835961, "grad_norm": 2.546875, "learning_rate": 4.9630788209980436e-05, "loss": 1.4619747161865235, "step": 24390 }, { "epoch": 2.0792501065189604, "grad_norm": 2.34375, "learning_rate": 4.963031548618964e-05, "loss": 1.2874589920043946, "step": 24400 }, { "epoch": 2.08010225820196, "grad_norm": 1.609375, "learning_rate": 4.962984246221844e-05, "loss": 1.2877534866333007, "step": 24410 }, { "epoch": 2.0809544098849595, "grad_norm": 2.828125, "learning_rate": 4.962936913807259e-05, "loss": 1.4397937774658203, "step": 24420 }, { "epoch": 2.081806561567959, "grad_norm": 2.171875, "learning_rate": 4.962889551375788e-05, "loss": 1.4848453521728515, "step": 24430 }, { "epoch": 2.0826587132509586, "grad_norm": 3.453125, "learning_rate": 4.9628421589280074e-05, "loss": 1.23794584274292, "step": 24440 }, { "epoch": 2.083510864933958, "grad_norm": 2.125, "learning_rate": 4.962794736464495e-05, "loss": 1.4494899749755858, "step": 24450 }, { "epoch": 2.0843630166169578, "grad_norm": 2.390625, "learning_rate": 4.962747283985828e-05, "loss": 1.367056655883789, "step": 24460 }, { "epoch": 2.0852151682999573, "grad_norm": 2.140625, "learning_rate": 4.962699801492586e-05, "loss": 1.5124821662902832, "step": 24470 }, { "epoch": 2.086067319982957, "grad_norm": 1.78125, "learning_rate": 4.962652288985347e-05, "loss": 1.4180739402770997, "step": 24480 }, { "epoch": 2.0869194716659565, "grad_norm": 2.125, "learning_rate": 4.962604746464689e-05, "loss": 1.4683085441589356, "step": 24490 }, { "epoch": 2.087771623348956, "grad_norm": 2.6875, "learning_rate": 4.962557173931193e-05, "loss": 1.3446252822875977, "step": 24500 }, { "epoch": 2.0886237750319556, "grad_norm": 2.5625, "learning_rate": 4.962509571385438e-05, "loss": 1.3803272247314453, "step": 24510 }, { "epoch": 2.089475926714955, "grad_norm": 2.4375, "learning_rate": 4.962461938828004e-05, "loss": 1.4570124626159668, "step": 24520 }, { "epoch": 2.0903280783979548, "grad_norm": 4.3125, "learning_rate": 4.9624142762594726e-05, "loss": 1.2935152053833008, "step": 24530 }, { "epoch": 2.0911802300809543, "grad_norm": 2.296875, "learning_rate": 4.962366583680423e-05, "loss": 1.413032627105713, "step": 24540 }, { "epoch": 2.092032381763954, "grad_norm": 2.078125, "learning_rate": 4.9623188610914385e-05, "loss": 1.4747846603393555, "step": 24550 }, { "epoch": 2.0928845334469535, "grad_norm": 2.421875, "learning_rate": 4.9622711084930996e-05, "loss": 1.4384450912475586, "step": 24560 }, { "epoch": 2.093736685129953, "grad_norm": 3.4375, "learning_rate": 4.962223325885987e-05, "loss": 1.161112403869629, "step": 24570 }, { "epoch": 2.0945888368129526, "grad_norm": 1.828125, "learning_rate": 4.962175513270686e-05, "loss": 1.30197114944458, "step": 24580 }, { "epoch": 2.095440988495952, "grad_norm": 2.234375, "learning_rate": 4.9621276706477763e-05, "loss": 1.3925385475158691, "step": 24590 }, { "epoch": 2.0962931401789517, "grad_norm": 3.375, "learning_rate": 4.962079798017843e-05, "loss": 1.2516571044921876, "step": 24600 }, { "epoch": 2.0971452918619513, "grad_norm": 1.96875, "learning_rate": 4.962031895381469e-05, "loss": 1.3798511505126954, "step": 24610 }, { "epoch": 2.097997443544951, "grad_norm": 2.671875, "learning_rate": 4.961983962739238e-05, "loss": 1.25577392578125, "step": 24620 }, { "epoch": 2.0988495952279504, "grad_norm": 2.640625, "learning_rate": 4.9619360000917335e-05, "loss": 1.3443109512329101, "step": 24630 }, { "epoch": 2.09970174691095, "grad_norm": 1.890625, "learning_rate": 4.961888007439541e-05, "loss": 1.3943364143371582, "step": 24640 }, { "epoch": 2.1005538985939496, "grad_norm": 2.125, "learning_rate": 4.961839984783246e-05, "loss": 1.310124397277832, "step": 24650 }, { "epoch": 2.101406050276949, "grad_norm": 2.75, "learning_rate": 4.961791932123432e-05, "loss": 1.340707492828369, "step": 24660 }, { "epoch": 2.1022582019599487, "grad_norm": 3.328125, "learning_rate": 4.961743849460686e-05, "loss": 1.3207242965698243, "step": 24670 }, { "epoch": 2.1031103536429483, "grad_norm": 1.8515625, "learning_rate": 4.961695736795593e-05, "loss": 1.3326537132263183, "step": 24680 }, { "epoch": 2.103962505325948, "grad_norm": 2.609375, "learning_rate": 4.96164759412874e-05, "loss": 1.321993064880371, "step": 24690 }, { "epoch": 2.1048146570089474, "grad_norm": 2.140625, "learning_rate": 4.9615994214607145e-05, "loss": 1.3707252502441407, "step": 24700 }, { "epoch": 2.105666808691947, "grad_norm": 2.5625, "learning_rate": 4.961551218792102e-05, "loss": 1.559659481048584, "step": 24710 }, { "epoch": 2.1065189603749466, "grad_norm": 1.9140625, "learning_rate": 4.961502986123491e-05, "loss": 1.2228996276855468, "step": 24720 }, { "epoch": 2.107371112057946, "grad_norm": 2.78125, "learning_rate": 4.9614547234554685e-05, "loss": 1.4288165092468261, "step": 24730 }, { "epoch": 2.1082232637409457, "grad_norm": 1.96875, "learning_rate": 4.9614064307886235e-05, "loss": 1.3574586868286134, "step": 24740 }, { "epoch": 2.1090754154239453, "grad_norm": 1.875, "learning_rate": 4.961358108123544e-05, "loss": 1.3026413917541504, "step": 24750 }, { "epoch": 2.109927567106945, "grad_norm": 2.515625, "learning_rate": 4.96130975546082e-05, "loss": 1.4464306831359863, "step": 24760 }, { "epoch": 2.1107797187899444, "grad_norm": 3.15625, "learning_rate": 4.9612613728010394e-05, "loss": 1.3581218719482422, "step": 24770 }, { "epoch": 2.111631870472944, "grad_norm": 2.4375, "learning_rate": 4.961212960144793e-05, "loss": 1.3187642097473145, "step": 24780 }, { "epoch": 2.1124840221559436, "grad_norm": 2.265625, "learning_rate": 4.9611645174926704e-05, "loss": 1.1702142715454102, "step": 24790 }, { "epoch": 2.113336173838943, "grad_norm": 2.203125, "learning_rate": 4.961116044845261e-05, "loss": 1.3743253707885743, "step": 24800 }, { "epoch": 2.114188325521943, "grad_norm": 2.4375, "learning_rate": 4.961067542203157e-05, "loss": 1.2859460830688476, "step": 24810 }, { "epoch": 2.1150404772049427, "grad_norm": 4.4375, "learning_rate": 4.961019009566949e-05, "loss": 1.125192356109619, "step": 24820 }, { "epoch": 2.1158926288879423, "grad_norm": 2.34375, "learning_rate": 4.960970446937229e-05, "loss": 1.5552944183349608, "step": 24830 }, { "epoch": 2.116744780570942, "grad_norm": 2.921875, "learning_rate": 4.960921854314587e-05, "loss": 1.2766864776611329, "step": 24840 }, { "epoch": 2.1175969322539414, "grad_norm": 3.109375, "learning_rate": 4.960873231699618e-05, "loss": 1.3523138999938964, "step": 24850 }, { "epoch": 2.118449083936941, "grad_norm": 1.71875, "learning_rate": 4.960824579092912e-05, "loss": 1.346316432952881, "step": 24860 }, { "epoch": 2.1193012356199405, "grad_norm": 1.734375, "learning_rate": 4.9607758964950635e-05, "loss": 1.372395133972168, "step": 24870 }, { "epoch": 2.12015338730294, "grad_norm": 2.59375, "learning_rate": 4.960727183906666e-05, "loss": 1.2533565521240235, "step": 24880 }, { "epoch": 2.1210055389859397, "grad_norm": 1.984375, "learning_rate": 4.960678441328312e-05, "loss": 1.4258609771728517, "step": 24890 }, { "epoch": 2.1218576906689393, "grad_norm": 1.8046875, "learning_rate": 4.960629668760596e-05, "loss": 1.4620503425598144, "step": 24900 }, { "epoch": 2.122709842351939, "grad_norm": 2.984375, "learning_rate": 4.960580866204113e-05, "loss": 1.2899831771850585, "step": 24910 }, { "epoch": 2.1235619940349384, "grad_norm": 1.9140625, "learning_rate": 4.960532033659457e-05, "loss": 1.411330795288086, "step": 24920 }, { "epoch": 2.124414145717938, "grad_norm": 1.6328125, "learning_rate": 4.9604831711272236e-05, "loss": 1.3872380256652832, "step": 24930 }, { "epoch": 2.1252662974009375, "grad_norm": 1.7421875, "learning_rate": 4.960434278608008e-05, "loss": 1.4952005386352538, "step": 24940 }, { "epoch": 2.126118449083937, "grad_norm": 2.515625, "learning_rate": 4.960385356102406e-05, "loss": 1.2777822494506836, "step": 24950 }, { "epoch": 2.1269706007669367, "grad_norm": 2.40625, "learning_rate": 4.960336403611015e-05, "loss": 1.4749000549316407, "step": 24960 }, { "epoch": 2.1278227524499362, "grad_norm": 2.375, "learning_rate": 4.9602874211344296e-05, "loss": 1.4158640861511231, "step": 24970 }, { "epoch": 2.128674904132936, "grad_norm": 3.71875, "learning_rate": 4.960238408673249e-05, "loss": 1.5227813720703125, "step": 24980 }, { "epoch": 2.1295270558159354, "grad_norm": 2.984375, "learning_rate": 4.960189366228069e-05, "loss": 1.293776512145996, "step": 24990 }, { "epoch": 2.130379207498935, "grad_norm": 2.171875, "learning_rate": 4.960140293799487e-05, "loss": 1.3697561264038085, "step": 25000 }, { "epoch": 2.1312313591819345, "grad_norm": 2.25, "learning_rate": 4.960091191388102e-05, "loss": 1.3356481552124024, "step": 25010 }, { "epoch": 2.132083510864934, "grad_norm": 1.875, "learning_rate": 4.960042058994513e-05, "loss": 1.274207878112793, "step": 25020 }, { "epoch": 2.1329356625479337, "grad_norm": 2.265625, "learning_rate": 4.959992896619318e-05, "loss": 1.4694486618041993, "step": 25030 }, { "epoch": 2.133787814230933, "grad_norm": 2.46875, "learning_rate": 4.959943704263115e-05, "loss": 1.2522072792053223, "step": 25040 }, { "epoch": 2.134639965913933, "grad_norm": 2.421875, "learning_rate": 4.959894481926505e-05, "loss": 1.2173196792602539, "step": 25050 }, { "epoch": 2.1354921175969324, "grad_norm": 2.59375, "learning_rate": 4.9598452296100883e-05, "loss": 1.3352574348449706, "step": 25060 }, { "epoch": 2.136344269279932, "grad_norm": 2.765625, "learning_rate": 4.959795947314464e-05, "loss": 1.405482006072998, "step": 25070 }, { "epoch": 2.1371964209629315, "grad_norm": 1.8671875, "learning_rate": 4.959746635040233e-05, "loss": 1.3851335525512696, "step": 25080 }, { "epoch": 2.138048572645931, "grad_norm": 2.296875, "learning_rate": 4.959697292787997e-05, "loss": 1.3156540870666504, "step": 25090 }, { "epoch": 2.1389007243289306, "grad_norm": 3.71875, "learning_rate": 4.959647920558357e-05, "loss": 1.3485247611999511, "step": 25100 }, { "epoch": 2.13975287601193, "grad_norm": 1.9609375, "learning_rate": 4.959598518351913e-05, "loss": 1.5079252243041992, "step": 25110 }, { "epoch": 2.1406050276949298, "grad_norm": 2.59375, "learning_rate": 4.9595490861692705e-05, "loss": 1.3806072235107423, "step": 25120 }, { "epoch": 2.1414571793779293, "grad_norm": 2.875, "learning_rate": 4.959499624011029e-05, "loss": 1.3898296356201172, "step": 25130 }, { "epoch": 2.142309331060929, "grad_norm": 2.171875, "learning_rate": 4.9594501318777926e-05, "loss": 1.342735481262207, "step": 25140 }, { "epoch": 2.1431614827439285, "grad_norm": 2.78125, "learning_rate": 4.959400609770164e-05, "loss": 1.3274053573608398, "step": 25150 }, { "epoch": 2.144013634426928, "grad_norm": 2.453125, "learning_rate": 4.959351057688748e-05, "loss": 1.3456432342529296, "step": 25160 }, { "epoch": 2.1448657861099276, "grad_norm": 2.5625, "learning_rate": 4.9593014756341464e-05, "loss": 1.368929100036621, "step": 25170 }, { "epoch": 2.145717937792927, "grad_norm": 2.453125, "learning_rate": 4.9592518636069653e-05, "loss": 1.4163331985473633, "step": 25180 }, { "epoch": 2.1465700894759268, "grad_norm": 2.890625, "learning_rate": 4.959202221607808e-05, "loss": 1.366082000732422, "step": 25190 }, { "epoch": 2.1474222411589263, "grad_norm": 4.375, "learning_rate": 4.959152549637281e-05, "loss": 1.4011985778808593, "step": 25200 }, { "epoch": 2.148274392841926, "grad_norm": 2.65625, "learning_rate": 4.959102847695989e-05, "loss": 1.2879951477050782, "step": 25210 }, { "epoch": 2.1491265445249255, "grad_norm": 2.203125, "learning_rate": 4.9590531157845366e-05, "loss": 1.5333460807800292, "step": 25220 }, { "epoch": 2.149978696207925, "grad_norm": 2.65625, "learning_rate": 4.9590033539035316e-05, "loss": 1.403183650970459, "step": 25230 }, { "epoch": 2.1508308478909246, "grad_norm": 2.984375, "learning_rate": 4.958953562053579e-05, "loss": 1.4477875709533692, "step": 25240 }, { "epoch": 2.151682999573924, "grad_norm": 2.234375, "learning_rate": 4.9589037402352864e-05, "loss": 1.5088629722595215, "step": 25250 }, { "epoch": 2.1525351512569237, "grad_norm": 2.546875, "learning_rate": 4.958853888449262e-05, "loss": 1.4219916343688965, "step": 25260 }, { "epoch": 2.1533873029399233, "grad_norm": 1.796875, "learning_rate": 4.9588040066961115e-05, "loss": 1.4313563346862792, "step": 25270 }, { "epoch": 2.154239454622923, "grad_norm": 2.375, "learning_rate": 4.9587540949764446e-05, "loss": 1.4099774360656738, "step": 25280 }, { "epoch": 2.1550916063059224, "grad_norm": 3.984375, "learning_rate": 4.9587041532908675e-05, "loss": 1.360225772857666, "step": 25290 }, { "epoch": 2.155943757988922, "grad_norm": 2.109375, "learning_rate": 4.9586541816399915e-05, "loss": 1.361496639251709, "step": 25300 }, { "epoch": 2.1567959096719216, "grad_norm": 1.9609375, "learning_rate": 4.9586041800244224e-05, "loss": 1.3201634407043457, "step": 25310 }, { "epoch": 2.157648061354921, "grad_norm": 2.4375, "learning_rate": 4.958554148444772e-05, "loss": 1.3081740379333495, "step": 25320 }, { "epoch": 2.1585002130379207, "grad_norm": 3.984375, "learning_rate": 4.95850408690165e-05, "loss": 1.4933621406555175, "step": 25330 }, { "epoch": 2.1593523647209203, "grad_norm": 3.625, "learning_rate": 4.958453995395665e-05, "loss": 1.349635124206543, "step": 25340 }, { "epoch": 2.16020451640392, "grad_norm": 3.5, "learning_rate": 4.95840387392743e-05, "loss": 1.2903408050537108, "step": 25350 }, { "epoch": 2.1610566680869194, "grad_norm": 2.203125, "learning_rate": 4.958353722497553e-05, "loss": 1.4410484313964844, "step": 25360 }, { "epoch": 2.161908819769919, "grad_norm": 1.78125, "learning_rate": 4.958303541106647e-05, "loss": 1.3749539375305175, "step": 25370 }, { "epoch": 2.1627609714529186, "grad_norm": 2.125, "learning_rate": 4.958253329755323e-05, "loss": 1.3459794044494628, "step": 25380 }, { "epoch": 2.163613123135918, "grad_norm": 2.859375, "learning_rate": 4.958203088444192e-05, "loss": 1.3903809547424317, "step": 25390 }, { "epoch": 2.1644652748189177, "grad_norm": 2.109375, "learning_rate": 4.958152817173869e-05, "loss": 1.3849010467529297, "step": 25400 }, { "epoch": 2.1653174265019173, "grad_norm": 2.34375, "learning_rate": 4.9581025159449636e-05, "loss": 1.4474944114685058, "step": 25410 }, { "epoch": 2.166169578184917, "grad_norm": 2.859375, "learning_rate": 4.9580521847580905e-05, "loss": 1.4317959785461425, "step": 25420 }, { "epoch": 2.1670217298679164, "grad_norm": 2.71875, "learning_rate": 4.9580018236138636e-05, "loss": 1.5042631149291992, "step": 25430 }, { "epoch": 2.167873881550916, "grad_norm": 2.703125, "learning_rate": 4.957951432512896e-05, "loss": 1.3170693397521973, "step": 25440 }, { "epoch": 2.1687260332339156, "grad_norm": 1.90625, "learning_rate": 4.957901011455801e-05, "loss": 1.4053417205810548, "step": 25450 }, { "epoch": 2.169578184916915, "grad_norm": 2.1875, "learning_rate": 4.9578505604431944e-05, "loss": 1.3877450942993164, "step": 25460 }, { "epoch": 2.1704303365999147, "grad_norm": 2.90625, "learning_rate": 4.9578000794756905e-05, "loss": 1.5090109825134277, "step": 25470 }, { "epoch": 2.1712824882829143, "grad_norm": 1.796875, "learning_rate": 4.9577495685539045e-05, "loss": 1.420143985748291, "step": 25480 }, { "epoch": 2.172134639965914, "grad_norm": 1.9140625, "learning_rate": 4.957699027678452e-05, "loss": 1.4032164573669434, "step": 25490 }, { "epoch": 2.1729867916489134, "grad_norm": 2.078125, "learning_rate": 4.957648456849949e-05, "loss": 1.4988727569580078, "step": 25500 }, { "epoch": 2.173838943331913, "grad_norm": 1.7578125, "learning_rate": 4.957597856069012e-05, "loss": 1.3731699943542481, "step": 25510 }, { "epoch": 2.1746910950149125, "grad_norm": 2.109375, "learning_rate": 4.9575472253362574e-05, "loss": 1.3243409156799317, "step": 25520 }, { "epoch": 2.175543246697912, "grad_norm": 2.65625, "learning_rate": 4.957496564652303e-05, "loss": 1.3002074241638184, "step": 25530 }, { "epoch": 2.1763953983809117, "grad_norm": 2.265625, "learning_rate": 4.957445874017766e-05, "loss": 1.4098670959472657, "step": 25540 }, { "epoch": 2.1772475500639112, "grad_norm": 3.15625, "learning_rate": 4.957395153433262e-05, "loss": 1.2663321495056152, "step": 25550 }, { "epoch": 2.178099701746911, "grad_norm": 1.828125, "learning_rate": 4.957344402899412e-05, "loss": 1.370882225036621, "step": 25560 }, { "epoch": 2.1789518534299104, "grad_norm": 1.7265625, "learning_rate": 4.957293622416834e-05, "loss": 1.3089186668395996, "step": 25570 }, { "epoch": 2.17980400511291, "grad_norm": 1.9765625, "learning_rate": 4.957242811986146e-05, "loss": 1.3397855758666992, "step": 25580 }, { "epoch": 2.1806561567959095, "grad_norm": 2.484375, "learning_rate": 4.957191971607967e-05, "loss": 1.2386302947998047, "step": 25590 }, { "epoch": 2.181508308478909, "grad_norm": 1.4296875, "learning_rate": 4.957141101282918e-05, "loss": 1.3772066116333008, "step": 25600 }, { "epoch": 2.1823604601619087, "grad_norm": 3.140625, "learning_rate": 4.9570902010116184e-05, "loss": 1.4674400329589843, "step": 25610 }, { "epoch": 2.1832126118449082, "grad_norm": 2.078125, "learning_rate": 4.957039270794688e-05, "loss": 1.3862972259521484, "step": 25620 }, { "epoch": 2.184064763527908, "grad_norm": 2.578125, "learning_rate": 4.956988310632748e-05, "loss": 1.2357282638549805, "step": 25630 }, { "epoch": 2.1849169152109074, "grad_norm": 3.453125, "learning_rate": 4.956937320526419e-05, "loss": 1.3484744071960448, "step": 25640 }, { "epoch": 2.185769066893907, "grad_norm": 2.765625, "learning_rate": 4.956886300476323e-05, "loss": 1.2194157600402833, "step": 25650 }, { "epoch": 2.1866212185769065, "grad_norm": 2.171875, "learning_rate": 4.956835250483082e-05, "loss": 1.2563741683959961, "step": 25660 }, { "epoch": 2.187473370259906, "grad_norm": 2.46875, "learning_rate": 4.956784170547317e-05, "loss": 1.1922279357910157, "step": 25670 }, { "epoch": 2.1883255219429056, "grad_norm": 2.3125, "learning_rate": 4.956733060669653e-05, "loss": 1.4358983993530274, "step": 25680 }, { "epoch": 2.189177673625905, "grad_norm": 2.171875, "learning_rate": 4.95668192085071e-05, "loss": 1.428616428375244, "step": 25690 }, { "epoch": 2.1900298253089048, "grad_norm": 3.0625, "learning_rate": 4.9566307510911117e-05, "loss": 1.3421480178833007, "step": 25700 }, { "epoch": 2.1908819769919043, "grad_norm": 1.796875, "learning_rate": 4.956579551391484e-05, "loss": 1.3896273612976073, "step": 25710 }, { "epoch": 2.191734128674904, "grad_norm": 1.9140625, "learning_rate": 4.956528321752448e-05, "loss": 1.4832192420959474, "step": 25720 }, { "epoch": 2.1925862803579035, "grad_norm": 3.453125, "learning_rate": 4.95647706217463e-05, "loss": 1.33306941986084, "step": 25730 }, { "epoch": 2.193438432040903, "grad_norm": 1.7890625, "learning_rate": 4.956425772658655e-05, "loss": 1.3360851287841797, "step": 25740 }, { "epoch": 2.1942905837239026, "grad_norm": 3.15625, "learning_rate": 4.956374453205146e-05, "loss": 1.2649789810180665, "step": 25750 }, { "epoch": 2.1951427354069026, "grad_norm": 3.25, "learning_rate": 4.9563231038147306e-05, "loss": 1.395688533782959, "step": 25760 }, { "epoch": 2.195994887089902, "grad_norm": 2.640625, "learning_rate": 4.9562717244880334e-05, "loss": 1.5006726264953614, "step": 25770 }, { "epoch": 2.1968470387729018, "grad_norm": 2.25, "learning_rate": 4.956220315225681e-05, "loss": 1.4342560768127441, "step": 25780 }, { "epoch": 2.1976991904559013, "grad_norm": 1.859375, "learning_rate": 4.9561688760283e-05, "loss": 1.3578615188598633, "step": 25790 }, { "epoch": 2.198551342138901, "grad_norm": 1.9453125, "learning_rate": 4.956117406896516e-05, "loss": 1.3257453918457032, "step": 25800 }, { "epoch": 2.1994034938219005, "grad_norm": 2.078125, "learning_rate": 4.956065907830959e-05, "loss": 1.43723783493042, "step": 25810 }, { "epoch": 2.2002556455049, "grad_norm": 1.796875, "learning_rate": 4.9560143788322543e-05, "loss": 1.4828312873840332, "step": 25820 }, { "epoch": 2.2011077971878996, "grad_norm": 2.09375, "learning_rate": 4.955962819901031e-05, "loss": 1.353468418121338, "step": 25830 }, { "epoch": 2.201959948870899, "grad_norm": 3.859375, "learning_rate": 4.9559112310379175e-05, "loss": 1.446599292755127, "step": 25840 }, { "epoch": 2.2028121005538988, "grad_norm": 2.5625, "learning_rate": 4.955859612243542e-05, "loss": 1.3577800750732423, "step": 25850 }, { "epoch": 2.2036642522368983, "grad_norm": 2.234375, "learning_rate": 4.955807963518533e-05, "loss": 1.3853405952453612, "step": 25860 }, { "epoch": 2.204516403919898, "grad_norm": 2.953125, "learning_rate": 4.955756284863521e-05, "loss": 1.373381996154785, "step": 25870 }, { "epoch": 2.2053685556028975, "grad_norm": 2.46875, "learning_rate": 4.955704576279136e-05, "loss": 1.3827043533325196, "step": 25880 }, { "epoch": 2.206220707285897, "grad_norm": 3.03125, "learning_rate": 4.9556528377660084e-05, "loss": 1.3303184509277344, "step": 25890 }, { "epoch": 2.2070728589688966, "grad_norm": 2.546875, "learning_rate": 4.955601069324767e-05, "loss": 1.35341796875, "step": 25900 }, { "epoch": 2.207925010651896, "grad_norm": 2.34375, "learning_rate": 4.955549270956045e-05, "loss": 1.506661605834961, "step": 25910 }, { "epoch": 2.2087771623348957, "grad_norm": 2.8125, "learning_rate": 4.955497442660472e-05, "loss": 1.3946565628051757, "step": 25920 }, { "epoch": 2.2096293140178953, "grad_norm": 2.15625, "learning_rate": 4.9554455844386794e-05, "loss": 1.3104146003723145, "step": 25930 }, { "epoch": 2.210481465700895, "grad_norm": 3.953125, "learning_rate": 4.955393696291301e-05, "loss": 1.4001965522766113, "step": 25940 }, { "epoch": 2.2113336173838944, "grad_norm": 1.953125, "learning_rate": 4.955341778218968e-05, "loss": 1.370191764831543, "step": 25950 }, { "epoch": 2.212185769066894, "grad_norm": 2.046875, "learning_rate": 4.955289830222314e-05, "loss": 1.4231056213378905, "step": 25960 }, { "epoch": 2.2130379207498936, "grad_norm": 1.7734375, "learning_rate": 4.955237852301971e-05, "loss": 1.4525686264038087, "step": 25970 }, { "epoch": 2.213890072432893, "grad_norm": 2.3125, "learning_rate": 4.955185844458572e-05, "loss": 1.3922402381896972, "step": 25980 }, { "epoch": 2.2147422241158927, "grad_norm": 2.15625, "learning_rate": 4.955133806692753e-05, "loss": 1.3108875274658203, "step": 25990 }, { "epoch": 2.2155943757988923, "grad_norm": 1.8046875, "learning_rate": 4.955081739005146e-05, "loss": 1.3847260475158691, "step": 26000 }, { "epoch": 2.216446527481892, "grad_norm": 3.328125, "learning_rate": 4.955029641396388e-05, "loss": 1.3425674438476562, "step": 26010 }, { "epoch": 2.2172986791648914, "grad_norm": 1.8828125, "learning_rate": 4.9549775138671114e-05, "loss": 1.5259796142578126, "step": 26020 }, { "epoch": 2.218150830847891, "grad_norm": 1.8984375, "learning_rate": 4.954925356417953e-05, "loss": 1.2731609344482422, "step": 26030 }, { "epoch": 2.2190029825308906, "grad_norm": 2.234375, "learning_rate": 4.954873169049548e-05, "loss": 1.293838882446289, "step": 26040 }, { "epoch": 2.21985513421389, "grad_norm": 3.28125, "learning_rate": 4.954820951762533e-05, "loss": 1.4524333000183105, "step": 26050 }, { "epoch": 2.2207072858968897, "grad_norm": 1.4765625, "learning_rate": 4.954768704557543e-05, "loss": 1.3678470611572267, "step": 26060 }, { "epoch": 2.2215594375798893, "grad_norm": 2.71875, "learning_rate": 4.954716427435216e-05, "loss": 1.3932998657226563, "step": 26070 }, { "epoch": 2.222411589262889, "grad_norm": 2.484375, "learning_rate": 4.95466412039619e-05, "loss": 1.3071253776550293, "step": 26080 }, { "epoch": 2.2232637409458884, "grad_norm": 2.640625, "learning_rate": 4.9546117834411e-05, "loss": 1.3177879333496094, "step": 26090 }, { "epoch": 2.224115892628888, "grad_norm": 1.5859375, "learning_rate": 4.954559416570585e-05, "loss": 1.4825329780578613, "step": 26100 }, { "epoch": 2.2249680443118875, "grad_norm": 2.796875, "learning_rate": 4.954507019785284e-05, "loss": 1.3066741943359375, "step": 26110 }, { "epoch": 2.225820195994887, "grad_norm": 2.421875, "learning_rate": 4.9544545930858346e-05, "loss": 1.463277244567871, "step": 26120 }, { "epoch": 2.2266723476778867, "grad_norm": 3.15625, "learning_rate": 4.954402136472876e-05, "loss": 1.2948289871215821, "step": 26130 }, { "epoch": 2.2275244993608863, "grad_norm": 1.84375, "learning_rate": 4.9543496499470474e-05, "loss": 1.2452691078186036, "step": 26140 }, { "epoch": 2.228376651043886, "grad_norm": 2.140625, "learning_rate": 4.95429713350899e-05, "loss": 1.433181381225586, "step": 26150 }, { "epoch": 2.2292288027268854, "grad_norm": 2.203125, "learning_rate": 4.954244587159341e-05, "loss": 1.5320615768432617, "step": 26160 }, { "epoch": 2.230080954409885, "grad_norm": 2.203125, "learning_rate": 4.9541920108987427e-05, "loss": 1.3893924713134767, "step": 26170 }, { "epoch": 2.2309331060928845, "grad_norm": 2.25, "learning_rate": 4.9541394047278353e-05, "loss": 1.3644241333007812, "step": 26180 }, { "epoch": 2.231785257775884, "grad_norm": 2.125, "learning_rate": 4.9540867686472616e-05, "loss": 1.4032806396484374, "step": 26190 }, { "epoch": 2.2326374094588837, "grad_norm": 1.953125, "learning_rate": 4.95403410265766e-05, "loss": 1.511833381652832, "step": 26200 }, { "epoch": 2.2334895611418832, "grad_norm": 1.890625, "learning_rate": 4.953981406759675e-05, "loss": 1.342900848388672, "step": 26210 }, { "epoch": 2.234341712824883, "grad_norm": 2.515625, "learning_rate": 4.9539286809539464e-05, "loss": 1.353493881225586, "step": 26220 }, { "epoch": 2.2351938645078824, "grad_norm": 1.9609375, "learning_rate": 4.953875925241119e-05, "loss": 1.5071919441223145, "step": 26230 }, { "epoch": 2.236046016190882, "grad_norm": 3.265625, "learning_rate": 4.953823139621836e-05, "loss": 1.4069737434387206, "step": 26240 }, { "epoch": 2.2368981678738815, "grad_norm": 3.515625, "learning_rate": 4.953770324096738e-05, "loss": 1.3747960090637208, "step": 26250 }, { "epoch": 2.237750319556881, "grad_norm": 2.78125, "learning_rate": 4.953717478666472e-05, "loss": 1.3499102592468262, "step": 26260 }, { "epoch": 2.2386024712398807, "grad_norm": 3.078125, "learning_rate": 4.953664603331679e-05, "loss": 1.3511683464050293, "step": 26270 }, { "epoch": 2.2394546229228802, "grad_norm": 2.703125, "learning_rate": 4.953611698093006e-05, "loss": 1.367729663848877, "step": 26280 }, { "epoch": 2.24030677460588, "grad_norm": 2.15625, "learning_rate": 4.9535587629510954e-05, "loss": 1.3806740760803222, "step": 26290 }, { "epoch": 2.2411589262888794, "grad_norm": 3.015625, "learning_rate": 4.9535057979065944e-05, "loss": 1.4029370307922364, "step": 26300 }, { "epoch": 2.242011077971879, "grad_norm": 1.734375, "learning_rate": 4.953452802960147e-05, "loss": 1.529551887512207, "step": 26310 }, { "epoch": 2.2428632296548785, "grad_norm": 2.203125, "learning_rate": 4.953399778112401e-05, "loss": 1.2778088569641113, "step": 26320 }, { "epoch": 2.243715381337878, "grad_norm": 2.5, "learning_rate": 4.953346723364e-05, "loss": 1.3863170623779297, "step": 26330 }, { "epoch": 2.2445675330208776, "grad_norm": 3.96875, "learning_rate": 4.9532936387155924e-05, "loss": 1.4259795188903808, "step": 26340 }, { "epoch": 2.245419684703877, "grad_norm": 3.265625, "learning_rate": 4.9532405241678247e-05, "loss": 1.4226335525512694, "step": 26350 }, { "epoch": 2.2462718363868768, "grad_norm": 2.8125, "learning_rate": 4.953187379721344e-05, "loss": 1.3957414627075195, "step": 26360 }, { "epoch": 2.2471239880698763, "grad_norm": 1.890625, "learning_rate": 4.953134205376799e-05, "loss": 1.309096622467041, "step": 26370 }, { "epoch": 2.247976139752876, "grad_norm": 3.125, "learning_rate": 4.953081001134837e-05, "loss": 1.4190630912780762, "step": 26380 }, { "epoch": 2.2488282914358755, "grad_norm": 2.34375, "learning_rate": 4.953027766996107e-05, "loss": 1.3823956489562987, "step": 26390 }, { "epoch": 2.249680443118875, "grad_norm": 2.421875, "learning_rate": 4.9529745029612554e-05, "loss": 1.3466730117797852, "step": 26400 }, { "epoch": 2.2505325948018746, "grad_norm": 2.421875, "learning_rate": 4.9529212090309345e-05, "loss": 1.4090229988098144, "step": 26410 }, { "epoch": 2.251384746484874, "grad_norm": 2.078125, "learning_rate": 4.952867885205792e-05, "loss": 1.4156953811645507, "step": 26420 }, { "epoch": 2.2522368981678738, "grad_norm": 3.09375, "learning_rate": 4.952814531486479e-05, "loss": 1.4080766677856444, "step": 26430 }, { "epoch": 2.2530890498508733, "grad_norm": 4.0625, "learning_rate": 4.952761147873645e-05, "loss": 1.1247954368591309, "step": 26440 }, { "epoch": 2.253941201533873, "grad_norm": 2.703125, "learning_rate": 4.9527077343679396e-05, "loss": 1.3534245491027832, "step": 26450 }, { "epoch": 2.2547933532168725, "grad_norm": 3.4375, "learning_rate": 4.9526542909700154e-05, "loss": 1.375824546813965, "step": 26460 }, { "epoch": 2.255645504899872, "grad_norm": 3.515625, "learning_rate": 4.952600817680523e-05, "loss": 1.3679978370666503, "step": 26470 }, { "epoch": 2.2564976565828716, "grad_norm": 1.90625, "learning_rate": 4.952547314500114e-05, "loss": 1.449778938293457, "step": 26480 }, { "epoch": 2.257349808265871, "grad_norm": 3.875, "learning_rate": 4.952493781429441e-05, "loss": 1.2921728134155273, "step": 26490 }, { "epoch": 2.2582019599488707, "grad_norm": 1.8515625, "learning_rate": 4.9524402184691565e-05, "loss": 1.3556180953979493, "step": 26500 }, { "epoch": 2.2590541116318703, "grad_norm": 2.0, "learning_rate": 4.952386625619913e-05, "loss": 1.4029836654663086, "step": 26510 }, { "epoch": 2.25990626331487, "grad_norm": 2.5625, "learning_rate": 4.952333002882363e-05, "loss": 1.2469037055969239, "step": 26520 }, { "epoch": 2.2607584149978694, "grad_norm": 1.7109375, "learning_rate": 4.952279350257161e-05, "loss": 1.4552250862121583, "step": 26530 }, { "epoch": 2.261610566680869, "grad_norm": 2.09375, "learning_rate": 4.952225667744961e-05, "loss": 1.4341270446777343, "step": 26540 }, { "epoch": 2.2624627183638686, "grad_norm": 2.75, "learning_rate": 4.9521719553464165e-05, "loss": 1.2773492813110352, "step": 26550 }, { "epoch": 2.263314870046868, "grad_norm": 1.828125, "learning_rate": 4.952118213062181e-05, "loss": 1.4093988418579102, "step": 26560 }, { "epoch": 2.2641670217298677, "grad_norm": 2.859375, "learning_rate": 4.952064440892913e-05, "loss": 1.3689388275146483, "step": 26570 }, { "epoch": 2.2650191734128677, "grad_norm": 1.9296875, "learning_rate": 4.9520106388392637e-05, "loss": 1.3856854438781738, "step": 26580 }, { "epoch": 2.2658713250958673, "grad_norm": 1.9140625, "learning_rate": 4.9519568069018926e-05, "loss": 1.4129810333251953, "step": 26590 }, { "epoch": 2.266723476778867, "grad_norm": 1.734375, "learning_rate": 4.9519029450814534e-05, "loss": 1.2604490280151368, "step": 26600 }, { "epoch": 2.2675756284618664, "grad_norm": 3.21875, "learning_rate": 4.9518490533786024e-05, "loss": 1.279830551147461, "step": 26610 }, { "epoch": 2.268427780144866, "grad_norm": 2.1875, "learning_rate": 4.951795131793998e-05, "loss": 1.3572243690490722, "step": 26620 }, { "epoch": 2.2692799318278656, "grad_norm": 2.40625, "learning_rate": 4.951741180328296e-05, "loss": 1.3142014503479005, "step": 26630 }, { "epoch": 2.270132083510865, "grad_norm": 2.40625, "learning_rate": 4.951687198982155e-05, "loss": 1.4526214599609375, "step": 26640 }, { "epoch": 2.2709842351938647, "grad_norm": 2.515625, "learning_rate": 4.9516331877562316e-05, "loss": 1.288248062133789, "step": 26650 }, { "epoch": 2.2718363868768643, "grad_norm": 2.703125, "learning_rate": 4.951579146651185e-05, "loss": 1.3368047714233398, "step": 26660 }, { "epoch": 2.272688538559864, "grad_norm": 2.828125, "learning_rate": 4.951525075667674e-05, "loss": 1.4383451461791992, "step": 26670 }, { "epoch": 2.2735406902428634, "grad_norm": 2.59375, "learning_rate": 4.951470974806357e-05, "loss": 1.3911480903625488, "step": 26680 }, { "epoch": 2.274392841925863, "grad_norm": 1.9140625, "learning_rate": 4.9514168440678933e-05, "loss": 1.4412315368652344, "step": 26690 }, { "epoch": 2.2752449936088626, "grad_norm": 2.203125, "learning_rate": 4.951362683452943e-05, "loss": 1.399741268157959, "step": 26700 }, { "epoch": 2.276097145291862, "grad_norm": 4.1875, "learning_rate": 4.951308492962166e-05, "loss": 1.2776813507080078, "step": 26710 }, { "epoch": 2.2769492969748617, "grad_norm": 2.515625, "learning_rate": 4.951254272596223e-05, "loss": 1.3711941719055176, "step": 26720 }, { "epoch": 2.2778014486578613, "grad_norm": 2.4375, "learning_rate": 4.951200022355774e-05, "loss": 1.1384151458740235, "step": 26730 }, { "epoch": 2.278653600340861, "grad_norm": 3.84375, "learning_rate": 4.951145742241482e-05, "loss": 1.4878551483154296, "step": 26740 }, { "epoch": 2.2795057520238604, "grad_norm": 2.828125, "learning_rate": 4.951091432254006e-05, "loss": 1.4206607818603516, "step": 26750 }, { "epoch": 2.28035790370686, "grad_norm": 1.703125, "learning_rate": 4.9510370923940094e-05, "loss": 1.3350409507751464, "step": 26760 }, { "epoch": 2.2812100553898595, "grad_norm": 2.234375, "learning_rate": 4.950982722662155e-05, "loss": 1.3151317596435548, "step": 26770 }, { "epoch": 2.282062207072859, "grad_norm": 2.140625, "learning_rate": 4.950928323059104e-05, "loss": 1.3929313659667968, "step": 26780 }, { "epoch": 2.2829143587558587, "grad_norm": 1.9140625, "learning_rate": 4.95087389358552e-05, "loss": 1.478129768371582, "step": 26790 }, { "epoch": 2.2837665104388583, "grad_norm": 2.328125, "learning_rate": 4.950819434242068e-05, "loss": 1.3625903129577637, "step": 26800 }, { "epoch": 2.284618662121858, "grad_norm": 1.765625, "learning_rate": 4.950764945029408e-05, "loss": 1.5008699417114257, "step": 26810 }, { "epoch": 2.2854708138048574, "grad_norm": 2.171875, "learning_rate": 4.9507104259482075e-05, "loss": 1.306222915649414, "step": 26820 }, { "epoch": 2.286322965487857, "grad_norm": 1.7578125, "learning_rate": 4.95065587699913e-05, "loss": 1.371852207183838, "step": 26830 }, { "epoch": 2.2871751171708565, "grad_norm": 2.015625, "learning_rate": 4.9506012981828384e-05, "loss": 1.2900089263916015, "step": 26840 }, { "epoch": 2.288027268853856, "grad_norm": 2.875, "learning_rate": 4.9505466895000006e-05, "loss": 1.3037559509277343, "step": 26850 }, { "epoch": 2.2888794205368557, "grad_norm": 2.109375, "learning_rate": 4.950492050951281e-05, "loss": 1.3950721740722656, "step": 26860 }, { "epoch": 2.2897315722198552, "grad_norm": 4.625, "learning_rate": 4.9504373825373454e-05, "loss": 1.1639962196350098, "step": 26870 }, { "epoch": 2.290583723902855, "grad_norm": 2.453125, "learning_rate": 4.9503826842588596e-05, "loss": 1.4430435180664063, "step": 26880 }, { "epoch": 2.2914358755858544, "grad_norm": 2.828125, "learning_rate": 4.950327956116492e-05, "loss": 1.2252471923828125, "step": 26890 }, { "epoch": 2.292288027268854, "grad_norm": 2.6875, "learning_rate": 4.9502731981109075e-05, "loss": 1.1849286079406738, "step": 26900 }, { "epoch": 2.2931401789518535, "grad_norm": 1.921875, "learning_rate": 4.950218410242774e-05, "loss": 1.3316184997558593, "step": 26910 }, { "epoch": 2.293992330634853, "grad_norm": 2.3125, "learning_rate": 4.9501635925127606e-05, "loss": 1.2846985816955567, "step": 26920 }, { "epoch": 2.2948444823178527, "grad_norm": 2.6875, "learning_rate": 4.9501087449215335e-05, "loss": 1.4383767127990723, "step": 26930 }, { "epoch": 2.295696634000852, "grad_norm": 2.25, "learning_rate": 4.9500538674697625e-05, "loss": 1.4209321975708007, "step": 26940 }, { "epoch": 2.296548785683852, "grad_norm": 3.609375, "learning_rate": 4.949998960158116e-05, "loss": 1.183809185028076, "step": 26950 }, { "epoch": 2.2974009373668514, "grad_norm": 2.171875, "learning_rate": 4.949944022987263e-05, "loss": 1.4327253341674804, "step": 26960 }, { "epoch": 2.298253089049851, "grad_norm": 3.1875, "learning_rate": 4.949889055957874e-05, "loss": 1.3026341438293456, "step": 26970 }, { "epoch": 2.2991052407328505, "grad_norm": 1.90625, "learning_rate": 4.949834059070617e-05, "loss": 1.4533546447753907, "step": 26980 }, { "epoch": 2.29995739241585, "grad_norm": 2.109375, "learning_rate": 4.949779032326163e-05, "loss": 1.3805184364318848, "step": 26990 }, { "epoch": 2.3008095440988496, "grad_norm": 1.8984375, "learning_rate": 4.949723975725184e-05, "loss": 1.4291183471679687, "step": 27000 }, { "epoch": 2.301661695781849, "grad_norm": 4.96875, "learning_rate": 4.9496688892683494e-05, "loss": 1.436182403564453, "step": 27010 }, { "epoch": 2.3025138474648488, "grad_norm": 1.8359375, "learning_rate": 4.949613772956332e-05, "loss": 1.284407615661621, "step": 27020 }, { "epoch": 2.3033659991478483, "grad_norm": 2.65625, "learning_rate": 4.949558626789802e-05, "loss": 1.3371740341186524, "step": 27030 }, { "epoch": 2.304218150830848, "grad_norm": 1.2890625, "learning_rate": 4.949503450769432e-05, "loss": 1.1793911933898926, "step": 27040 }, { "epoch": 2.3050703025138475, "grad_norm": 3.859375, "learning_rate": 4.949448244895895e-05, "loss": 1.2726863861083983, "step": 27050 }, { "epoch": 2.305922454196847, "grad_norm": 2.53125, "learning_rate": 4.9493930091698634e-05, "loss": 1.2374646186828613, "step": 27060 }, { "epoch": 2.3067746058798466, "grad_norm": 1.9453125, "learning_rate": 4.94933774359201e-05, "loss": 1.4385141372680663, "step": 27070 }, { "epoch": 2.307626757562846, "grad_norm": 2.640625, "learning_rate": 4.949282448163009e-05, "loss": 1.225803756713867, "step": 27080 }, { "epoch": 2.3084789092458458, "grad_norm": 2.765625, "learning_rate": 4.9492271228835344e-05, "loss": 1.3165093421936036, "step": 27090 }, { "epoch": 2.3093310609288453, "grad_norm": 2.203125, "learning_rate": 4.9491717677542595e-05, "loss": 1.461603832244873, "step": 27100 }, { "epoch": 2.310183212611845, "grad_norm": 1.9609375, "learning_rate": 4.94911638277586e-05, "loss": 1.3244182586669921, "step": 27110 }, { "epoch": 2.3110353642948445, "grad_norm": 2.4375, "learning_rate": 4.949060967949011e-05, "loss": 1.3427942276000977, "step": 27120 }, { "epoch": 2.311887515977844, "grad_norm": 1.9296875, "learning_rate": 4.949005523274386e-05, "loss": 1.3657593727111816, "step": 27130 }, { "epoch": 2.3127396676608436, "grad_norm": 1.6953125, "learning_rate": 4.948950048752663e-05, "loss": 1.5101963996887207, "step": 27140 }, { "epoch": 2.313591819343843, "grad_norm": 2.28125, "learning_rate": 4.948894544384517e-05, "loss": 1.3364906311035156, "step": 27150 }, { "epoch": 2.3144439710268427, "grad_norm": 2.453125, "learning_rate": 4.948839010170625e-05, "loss": 1.4714710235595703, "step": 27160 }, { "epoch": 2.3152961227098423, "grad_norm": 2.28125, "learning_rate": 4.9487834461116636e-05, "loss": 1.423272705078125, "step": 27170 }, { "epoch": 2.316148274392842, "grad_norm": 2.953125, "learning_rate": 4.9487278522083085e-05, "loss": 1.252530574798584, "step": 27180 }, { "epoch": 2.3170004260758414, "grad_norm": 2.796875, "learning_rate": 4.9486722284612396e-05, "loss": 1.4879138946533204, "step": 27190 }, { "epoch": 2.317852577758841, "grad_norm": 2.15625, "learning_rate": 4.948616574871133e-05, "loss": 1.3727804183959962, "step": 27200 }, { "epoch": 2.3187047294418406, "grad_norm": 3.5, "learning_rate": 4.948560891438669e-05, "loss": 1.377587890625, "step": 27210 }, { "epoch": 2.31955688112484, "grad_norm": 2.890625, "learning_rate": 4.948505178164524e-05, "loss": 1.5075556755065918, "step": 27220 }, { "epoch": 2.3204090328078397, "grad_norm": 3.8125, "learning_rate": 4.9484494350493784e-05, "loss": 1.5229333877563476, "step": 27230 }, { "epoch": 2.3212611844908393, "grad_norm": 2.3125, "learning_rate": 4.948393662093911e-05, "loss": 1.2942254066467285, "step": 27240 }, { "epoch": 2.322113336173839, "grad_norm": 2.375, "learning_rate": 4.948337859298802e-05, "loss": 1.347994327545166, "step": 27250 }, { "epoch": 2.3229654878568384, "grad_norm": 1.7890625, "learning_rate": 4.9482820266647315e-05, "loss": 1.3557004928588867, "step": 27260 }, { "epoch": 2.323817639539838, "grad_norm": 2.296875, "learning_rate": 4.948226164192379e-05, "loss": 1.2618526458740233, "step": 27270 }, { "epoch": 2.3246697912228376, "grad_norm": 2.546875, "learning_rate": 4.948170271882426e-05, "loss": 1.268465518951416, "step": 27280 }, { "epoch": 2.325521942905837, "grad_norm": 2.640625, "learning_rate": 4.9481143497355535e-05, "loss": 1.2774321556091308, "step": 27290 }, { "epoch": 2.3263740945888367, "grad_norm": 1.703125, "learning_rate": 4.948058397752443e-05, "loss": 1.3981037139892578, "step": 27300 }, { "epoch": 2.3272262462718363, "grad_norm": 2.359375, "learning_rate": 4.948002415933778e-05, "loss": 1.3606553077697754, "step": 27310 }, { "epoch": 2.328078397954836, "grad_norm": 2.203125, "learning_rate": 4.947946404280238e-05, "loss": 1.4066225051879884, "step": 27320 }, { "epoch": 2.3289305496378354, "grad_norm": 1.4609375, "learning_rate": 4.947890362792508e-05, "loss": 1.3975894927978516, "step": 27330 }, { "epoch": 2.329782701320835, "grad_norm": 2.0625, "learning_rate": 4.9478342914712694e-05, "loss": 1.4622345924377442, "step": 27340 }, { "epoch": 2.3306348530038346, "grad_norm": 2.46875, "learning_rate": 4.947778190317206e-05, "loss": 1.4724958419799805, "step": 27350 }, { "epoch": 2.331487004686834, "grad_norm": 2.75, "learning_rate": 4.947722059331002e-05, "loss": 1.223473358154297, "step": 27360 }, { "epoch": 2.3323391563698337, "grad_norm": 2.234375, "learning_rate": 4.9476658985133414e-05, "loss": 1.4243727684020997, "step": 27370 }, { "epoch": 2.3331913080528333, "grad_norm": 2.515625, "learning_rate": 4.947609707864908e-05, "loss": 1.3431599617004395, "step": 27380 }, { "epoch": 2.334043459735833, "grad_norm": 1.9765625, "learning_rate": 4.947553487386387e-05, "loss": 1.3658555030822754, "step": 27390 }, { "epoch": 2.3348956114188324, "grad_norm": 3.5625, "learning_rate": 4.947497237078465e-05, "loss": 1.3857848167419433, "step": 27400 }, { "epoch": 2.335747763101832, "grad_norm": 2.453125, "learning_rate": 4.9474409569418244e-05, "loss": 1.3120903968811035, "step": 27410 }, { "epoch": 2.3365999147848315, "grad_norm": 1.7734375, "learning_rate": 4.947384646977155e-05, "loss": 1.181531047821045, "step": 27420 }, { "epoch": 2.337452066467831, "grad_norm": 3.46875, "learning_rate": 4.947328307185139e-05, "loss": 1.3525404930114746, "step": 27430 }, { "epoch": 2.3383042181508307, "grad_norm": 2.03125, "learning_rate": 4.947271937566466e-05, "loss": 1.415959644317627, "step": 27440 }, { "epoch": 2.3391563698338302, "grad_norm": 1.8125, "learning_rate": 4.947215538121822e-05, "loss": 1.3788208961486816, "step": 27450 }, { "epoch": 2.34000852151683, "grad_norm": 3.078125, "learning_rate": 4.9471591088518946e-05, "loss": 1.4667165756225586, "step": 27460 }, { "epoch": 2.3408606731998294, "grad_norm": 1.5625, "learning_rate": 4.94710264975737e-05, "loss": 1.3901887893676759, "step": 27470 }, { "epoch": 2.341712824882829, "grad_norm": 2.25, "learning_rate": 4.947046160838939e-05, "loss": 1.3233515739440918, "step": 27480 }, { "epoch": 2.3425649765658285, "grad_norm": 3.3125, "learning_rate": 4.946989642097288e-05, "loss": 1.5190413475036622, "step": 27490 }, { "epoch": 2.343417128248828, "grad_norm": 2.140625, "learning_rate": 4.946933093533107e-05, "loss": 1.3607681274414063, "step": 27500 }, { "epoch": 2.3442692799318277, "grad_norm": 1.625, "learning_rate": 4.946876515147085e-05, "loss": 1.4778873443603515, "step": 27510 }, { "epoch": 2.3451214316148272, "grad_norm": 2.3125, "learning_rate": 4.94681990693991e-05, "loss": 1.2854157447814942, "step": 27520 }, { "epoch": 2.345973583297827, "grad_norm": 2.171875, "learning_rate": 4.9467632689122744e-05, "loss": 1.3032577514648438, "step": 27530 }, { "epoch": 2.3468257349808264, "grad_norm": 2.078125, "learning_rate": 4.946706601064867e-05, "loss": 1.2669963836669922, "step": 27540 }, { "epoch": 2.347677886663826, "grad_norm": 2.1875, "learning_rate": 4.9466499033983784e-05, "loss": 1.2887638092041016, "step": 27550 }, { "epoch": 2.3485300383468255, "grad_norm": 2.03125, "learning_rate": 4.9465931759135e-05, "loss": 1.511556053161621, "step": 27560 }, { "epoch": 2.349382190029825, "grad_norm": 1.890625, "learning_rate": 4.946536418610922e-05, "loss": 1.2696854591369628, "step": 27570 }, { "epoch": 2.3502343417128246, "grad_norm": 2.34375, "learning_rate": 4.946479631491338e-05, "loss": 1.4084112167358398, "step": 27580 }, { "epoch": 2.3510864933958247, "grad_norm": 2.359375, "learning_rate": 4.946422814555439e-05, "loss": 1.4390293121337892, "step": 27590 }, { "epoch": 2.351938645078824, "grad_norm": 2.71875, "learning_rate": 4.946365967803918e-05, "loss": 1.3799168586730957, "step": 27600 }, { "epoch": 2.352790796761824, "grad_norm": 3.171875, "learning_rate": 4.946309091237468e-05, "loss": 1.285257625579834, "step": 27610 }, { "epoch": 2.3536429484448234, "grad_norm": 3.0, "learning_rate": 4.94625218485678e-05, "loss": 1.3764748573303223, "step": 27620 }, { "epoch": 2.354495100127823, "grad_norm": 2.5, "learning_rate": 4.946195248662551e-05, "loss": 1.4261383056640624, "step": 27630 }, { "epoch": 2.3553472518108225, "grad_norm": 2.3125, "learning_rate": 4.946138282655472e-05, "loss": 1.4779726982116699, "step": 27640 }, { "epoch": 2.356199403493822, "grad_norm": 2.46875, "learning_rate": 4.946081286836238e-05, "loss": 1.506078338623047, "step": 27650 }, { "epoch": 2.3570515551768216, "grad_norm": 1.9140625, "learning_rate": 4.946024261205545e-05, "loss": 1.4334749221801757, "step": 27660 }, { "epoch": 2.357903706859821, "grad_norm": 2.109375, "learning_rate": 4.945967205764087e-05, "loss": 1.3695943832397461, "step": 27670 }, { "epoch": 2.3587558585428208, "grad_norm": 2.4375, "learning_rate": 4.945910120512559e-05, "loss": 1.3127439498901368, "step": 27680 }, { "epoch": 2.3596080102258203, "grad_norm": 2.125, "learning_rate": 4.9458530054516575e-05, "loss": 1.3351003646850585, "step": 27690 }, { "epoch": 2.36046016190882, "grad_norm": 3.046875, "learning_rate": 4.945795860582078e-05, "loss": 1.3306647300720216, "step": 27700 }, { "epoch": 2.3613123135918195, "grad_norm": 3.296875, "learning_rate": 4.945738685904517e-05, "loss": 1.3753986358642578, "step": 27710 }, { "epoch": 2.362164465274819, "grad_norm": 2.203125, "learning_rate": 4.9456814814196714e-05, "loss": 1.320256805419922, "step": 27720 }, { "epoch": 2.3630166169578186, "grad_norm": 1.9453125, "learning_rate": 4.945624247128238e-05, "loss": 1.451451873779297, "step": 27730 }, { "epoch": 2.363868768640818, "grad_norm": 2.234375, "learning_rate": 4.945566983030916e-05, "loss": 1.3407296180725097, "step": 27740 }, { "epoch": 2.3647209203238178, "grad_norm": 3.5625, "learning_rate": 4.945509689128401e-05, "loss": 1.388552761077881, "step": 27750 }, { "epoch": 2.3655730720068173, "grad_norm": 2.515625, "learning_rate": 4.945452365421392e-05, "loss": 1.2795190811157227, "step": 27760 }, { "epoch": 2.366425223689817, "grad_norm": 3.546875, "learning_rate": 4.94539501191059e-05, "loss": 1.3808649063110352, "step": 27770 }, { "epoch": 2.3672773753728165, "grad_norm": 2.578125, "learning_rate": 4.94533762859669e-05, "loss": 1.4551251411437989, "step": 27780 }, { "epoch": 2.368129527055816, "grad_norm": 1.9453125, "learning_rate": 4.945280215480393e-05, "loss": 1.4751745223999024, "step": 27790 }, { "epoch": 2.3689816787388156, "grad_norm": 2.046875, "learning_rate": 4.945222772562401e-05, "loss": 1.423712158203125, "step": 27800 }, { "epoch": 2.369833830421815, "grad_norm": 2.5, "learning_rate": 4.9451652998434105e-05, "loss": 1.3565279006958009, "step": 27810 }, { "epoch": 2.3706859821048147, "grad_norm": 3.703125, "learning_rate": 4.9451077973241245e-05, "loss": 1.2542046546936034, "step": 27820 }, { "epoch": 2.3715381337878143, "grad_norm": 3.390625, "learning_rate": 4.945050265005242e-05, "loss": 1.4641481399536134, "step": 27830 }, { "epoch": 2.372390285470814, "grad_norm": 2.828125, "learning_rate": 4.944992702887465e-05, "loss": 1.334609317779541, "step": 27840 }, { "epoch": 2.3732424371538134, "grad_norm": 2.015625, "learning_rate": 4.9449351109714965e-05, "loss": 1.3385556221008301, "step": 27850 }, { "epoch": 2.374094588836813, "grad_norm": 1.9921875, "learning_rate": 4.9448774892580364e-05, "loss": 1.461317253112793, "step": 27860 }, { "epoch": 2.3749467405198126, "grad_norm": 2.671875, "learning_rate": 4.944819837747787e-05, "loss": 1.4992581367492677, "step": 27870 }, { "epoch": 2.375798892202812, "grad_norm": 2.59375, "learning_rate": 4.944762156441452e-05, "loss": 1.4255727767944335, "step": 27880 }, { "epoch": 2.3766510438858117, "grad_norm": 1.96875, "learning_rate": 4.9447044453397326e-05, "loss": 1.3380934715270996, "step": 27890 }, { "epoch": 2.3775031955688113, "grad_norm": 2.65625, "learning_rate": 4.944646704443334e-05, "loss": 1.364833164215088, "step": 27900 }, { "epoch": 2.378355347251811, "grad_norm": 2.171875, "learning_rate": 4.94458893375296e-05, "loss": 1.4218644142150878, "step": 27910 }, { "epoch": 2.3792074989348104, "grad_norm": 2.671875, "learning_rate": 4.944531133269313e-05, "loss": 1.3983062744140624, "step": 27920 }, { "epoch": 2.38005965061781, "grad_norm": 2.375, "learning_rate": 4.9444733029930995e-05, "loss": 1.3194181442260742, "step": 27930 }, { "epoch": 2.3809118023008096, "grad_norm": 2.5, "learning_rate": 4.944415442925022e-05, "loss": 1.4406868934631347, "step": 27940 }, { "epoch": 2.381763953983809, "grad_norm": 2.734375, "learning_rate": 4.944357553065788e-05, "loss": 1.3381430625915527, "step": 27950 }, { "epoch": 2.3826161056668087, "grad_norm": 2.8125, "learning_rate": 4.944299633416102e-05, "loss": 1.2983311653137206, "step": 27960 }, { "epoch": 2.3834682573498083, "grad_norm": 1.6328125, "learning_rate": 4.9442416839766695e-05, "loss": 1.3438676834106444, "step": 27970 }, { "epoch": 2.384320409032808, "grad_norm": 2.5625, "learning_rate": 4.944183704748196e-05, "loss": 1.4719565391540528, "step": 27980 }, { "epoch": 2.3851725607158074, "grad_norm": 1.6015625, "learning_rate": 4.94412569573139e-05, "loss": 1.2117273330688476, "step": 27990 }, { "epoch": 2.386024712398807, "grad_norm": 2.828125, "learning_rate": 4.9440676569269583e-05, "loss": 1.2840847015380858, "step": 28000 }, { "epoch": 2.3868768640818065, "grad_norm": 1.515625, "learning_rate": 4.944009588335607e-05, "loss": 1.5166460037231446, "step": 28010 }, { "epoch": 2.387729015764806, "grad_norm": 1.875, "learning_rate": 4.9439514899580444e-05, "loss": 1.3301438331604003, "step": 28020 }, { "epoch": 2.3885811674478057, "grad_norm": 2.265625, "learning_rate": 4.9438933617949785e-05, "loss": 1.4035649299621582, "step": 28030 }, { "epoch": 2.3894333191308053, "grad_norm": 2.015625, "learning_rate": 4.943835203847118e-05, "loss": 1.43832426071167, "step": 28040 }, { "epoch": 2.390285470813805, "grad_norm": 2.1875, "learning_rate": 4.943777016115171e-05, "loss": 1.2877208709716796, "step": 28050 }, { "epoch": 2.3911376224968044, "grad_norm": 2.421875, "learning_rate": 4.943718798599848e-05, "loss": 1.3216379165649415, "step": 28060 }, { "epoch": 2.391989774179804, "grad_norm": 2.53125, "learning_rate": 4.9436605513018577e-05, "loss": 1.4030476570129395, "step": 28070 }, { "epoch": 2.3928419258628035, "grad_norm": 2.609375, "learning_rate": 4.943602274221909e-05, "loss": 1.2650666236877441, "step": 28080 }, { "epoch": 2.393694077545803, "grad_norm": 1.6484375, "learning_rate": 4.9435439673607144e-05, "loss": 1.4486164093017577, "step": 28090 }, { "epoch": 2.3945462292288027, "grad_norm": 2.5, "learning_rate": 4.943485630718982e-05, "loss": 1.2980013847351075, "step": 28100 }, { "epoch": 2.3953983809118022, "grad_norm": 2.390625, "learning_rate": 4.9434272642974245e-05, "loss": 1.2569592475891114, "step": 28110 }, { "epoch": 2.396250532594802, "grad_norm": 2.53125, "learning_rate": 4.9433688680967525e-05, "loss": 1.310333824157715, "step": 28120 }, { "epoch": 2.3971026842778014, "grad_norm": 2.140625, "learning_rate": 4.943310442117679e-05, "loss": 1.4045312881469727, "step": 28130 }, { "epoch": 2.397954835960801, "grad_norm": 2.765625, "learning_rate": 4.943251986360914e-05, "loss": 1.3877532005310058, "step": 28140 }, { "epoch": 2.3988069876438005, "grad_norm": 2.84375, "learning_rate": 4.943193500827171e-05, "loss": 1.3901968002319336, "step": 28150 }, { "epoch": 2.3996591393268, "grad_norm": 1.7890625, "learning_rate": 4.9431349855171636e-05, "loss": 1.5005313873291015, "step": 28160 }, { "epoch": 2.4005112910097997, "grad_norm": 3.328125, "learning_rate": 4.9430764404316035e-05, "loss": 1.3855514526367188, "step": 28170 }, { "epoch": 2.401363442692799, "grad_norm": 2.46875, "learning_rate": 4.943017865571204e-05, "loss": 1.2618032455444337, "step": 28180 }, { "epoch": 2.402215594375799, "grad_norm": 2.21875, "learning_rate": 4.942959260936681e-05, "loss": 1.5005809783935546, "step": 28190 }, { "epoch": 2.4030677460587984, "grad_norm": 2.625, "learning_rate": 4.942900626528747e-05, "loss": 1.4632654190063477, "step": 28200 }, { "epoch": 2.403919897741798, "grad_norm": 2.609375, "learning_rate": 4.942841962348117e-05, "loss": 1.2449049949645996, "step": 28210 }, { "epoch": 2.4047720494247975, "grad_norm": 2.421875, "learning_rate": 4.9427832683955065e-05, "loss": 1.4216901779174804, "step": 28220 }, { "epoch": 2.405624201107797, "grad_norm": 2.0625, "learning_rate": 4.94272454467163e-05, "loss": 1.4242280006408692, "step": 28230 }, { "epoch": 2.4064763527907966, "grad_norm": 3.75, "learning_rate": 4.942665791177204e-05, "loss": 1.2771838188171387, "step": 28240 }, { "epoch": 2.407328504473796, "grad_norm": 2.484375, "learning_rate": 4.942607007912944e-05, "loss": 1.2662403106689453, "step": 28250 }, { "epoch": 2.4081806561567958, "grad_norm": 2.265625, "learning_rate": 4.942548194879566e-05, "loss": 1.2167242050170899, "step": 28260 }, { "epoch": 2.4090328078397953, "grad_norm": 3.640625, "learning_rate": 4.942489352077788e-05, "loss": 1.317316436767578, "step": 28270 }, { "epoch": 2.409884959522795, "grad_norm": 2.25, "learning_rate": 4.942430479508326e-05, "loss": 1.3105751037597657, "step": 28280 }, { "epoch": 2.4107371112057945, "grad_norm": 2.828125, "learning_rate": 4.942371577171899e-05, "loss": 1.3518381118774414, "step": 28290 }, { "epoch": 2.411589262888794, "grad_norm": 2.078125, "learning_rate": 4.9423126450692234e-05, "loss": 1.3430491447448731, "step": 28300 }, { "epoch": 2.4124414145717936, "grad_norm": 1.734375, "learning_rate": 4.9422536832010176e-05, "loss": 1.2102150917053223, "step": 28310 }, { "epoch": 2.413293566254793, "grad_norm": 2.5, "learning_rate": 4.942194691568001e-05, "loss": 1.2913127899169923, "step": 28320 }, { "epoch": 2.4141457179377928, "grad_norm": 2.0, "learning_rate": 4.942135670170891e-05, "loss": 1.4207167625427246, "step": 28330 }, { "epoch": 2.4149978696207923, "grad_norm": 3.125, "learning_rate": 4.942076619010409e-05, "loss": 1.1604399681091309, "step": 28340 }, { "epoch": 2.415850021303792, "grad_norm": 2.03125, "learning_rate": 4.942017538087273e-05, "loss": 1.3526034355163574, "step": 28350 }, { "epoch": 2.4167021729867915, "grad_norm": 2.296875, "learning_rate": 4.941958427402204e-05, "loss": 1.2986892700195312, "step": 28360 }, { "epoch": 2.417554324669791, "grad_norm": 2.765625, "learning_rate": 4.941899286955922e-05, "loss": 1.2215540885925293, "step": 28370 }, { "epoch": 2.418406476352791, "grad_norm": 1.6640625, "learning_rate": 4.941840116749148e-05, "loss": 1.335383987426758, "step": 28380 }, { "epoch": 2.4192586280357906, "grad_norm": 2.453125, "learning_rate": 4.941780916782604e-05, "loss": 1.2240503311157227, "step": 28390 }, { "epoch": 2.42011077971879, "grad_norm": 2.609375, "learning_rate": 4.941721687057009e-05, "loss": 1.4139941215515137, "step": 28400 }, { "epoch": 2.4209629314017898, "grad_norm": 2.28125, "learning_rate": 4.941662427573087e-05, "loss": 1.4174257278442384, "step": 28410 }, { "epoch": 2.4218150830847893, "grad_norm": 4.03125, "learning_rate": 4.941603138331559e-05, "loss": 1.3021417617797852, "step": 28420 }, { "epoch": 2.422667234767789, "grad_norm": 2.015625, "learning_rate": 4.941543819333149e-05, "loss": 1.4150424957275392, "step": 28430 }, { "epoch": 2.4235193864507885, "grad_norm": 2.875, "learning_rate": 4.941484470578579e-05, "loss": 1.3273483276367188, "step": 28440 }, { "epoch": 2.424371538133788, "grad_norm": 2.421875, "learning_rate": 4.9414250920685724e-05, "loss": 1.410547924041748, "step": 28450 }, { "epoch": 2.4252236898167876, "grad_norm": 2.4375, "learning_rate": 4.9413656838038525e-05, "loss": 1.3072638511657715, "step": 28460 }, { "epoch": 2.426075841499787, "grad_norm": 2.34375, "learning_rate": 4.9413062457851434e-05, "loss": 1.4192524909973145, "step": 28470 }, { "epoch": 2.4269279931827867, "grad_norm": 1.9765625, "learning_rate": 4.941246778013171e-05, "loss": 1.517103099822998, "step": 28480 }, { "epoch": 2.4277801448657863, "grad_norm": 2.453125, "learning_rate": 4.941187280488658e-05, "loss": 1.459243392944336, "step": 28490 }, { "epoch": 2.428632296548786, "grad_norm": 1.96875, "learning_rate": 4.94112775321233e-05, "loss": 1.3601371765136718, "step": 28500 }, { "epoch": 2.4294844482317854, "grad_norm": 1.8125, "learning_rate": 4.941068196184914e-05, "loss": 1.291090965270996, "step": 28510 }, { "epoch": 2.430336599914785, "grad_norm": 2.125, "learning_rate": 4.941008609407134e-05, "loss": 1.4267893791198731, "step": 28520 }, { "epoch": 2.4311887515977846, "grad_norm": 2.875, "learning_rate": 4.940948992879718e-05, "loss": 1.386176586151123, "step": 28530 }, { "epoch": 2.432040903280784, "grad_norm": 3.15625, "learning_rate": 4.9408893466033904e-05, "loss": 1.4786082267761231, "step": 28540 }, { "epoch": 2.4328930549637837, "grad_norm": 3.59375, "learning_rate": 4.9408296705788796e-05, "loss": 1.4363847732543946, "step": 28550 }, { "epoch": 2.4337452066467833, "grad_norm": 1.8828125, "learning_rate": 4.9407699648069125e-05, "loss": 1.3603687286376953, "step": 28560 }, { "epoch": 2.434597358329783, "grad_norm": 2.78125, "learning_rate": 4.940710229288217e-05, "loss": 1.379399871826172, "step": 28570 }, { "epoch": 2.4354495100127824, "grad_norm": 3.78125, "learning_rate": 4.9406504640235206e-05, "loss": 1.171639633178711, "step": 28580 }, { "epoch": 2.436301661695782, "grad_norm": 1.7578125, "learning_rate": 4.940590669013553e-05, "loss": 1.4086972236633302, "step": 28590 }, { "epoch": 2.4371538133787816, "grad_norm": 2.0, "learning_rate": 4.940530844259042e-05, "loss": 1.415414047241211, "step": 28600 }, { "epoch": 2.438005965061781, "grad_norm": 1.6484375, "learning_rate": 4.9404709897607146e-05, "loss": 1.4973505973815917, "step": 28610 }, { "epoch": 2.4388581167447807, "grad_norm": 2.53125, "learning_rate": 4.940411105519304e-05, "loss": 1.3726198196411132, "step": 28620 }, { "epoch": 2.4397102684277803, "grad_norm": 2.0625, "learning_rate": 4.940351191535538e-05, "loss": 1.4269238471984864, "step": 28630 }, { "epoch": 2.44056242011078, "grad_norm": 2.421875, "learning_rate": 4.940291247810148e-05, "loss": 1.2772704124450684, "step": 28640 }, { "epoch": 2.4414145717937794, "grad_norm": 2.828125, "learning_rate": 4.940231274343863e-05, "loss": 1.4830795288085938, "step": 28650 }, { "epoch": 2.442266723476779, "grad_norm": 2.34375, "learning_rate": 4.940171271137415e-05, "loss": 1.3445302963256835, "step": 28660 }, { "epoch": 2.4431188751597785, "grad_norm": 2.953125, "learning_rate": 4.940111238191534e-05, "loss": 1.3368438720703124, "step": 28670 }, { "epoch": 2.443971026842778, "grad_norm": 2.140625, "learning_rate": 4.9400511755069536e-05, "loss": 1.5112857818603516, "step": 28680 }, { "epoch": 2.4448231785257777, "grad_norm": 2.75, "learning_rate": 4.939991083084404e-05, "loss": 1.4388748168945313, "step": 28690 }, { "epoch": 2.4456753302087773, "grad_norm": 2.84375, "learning_rate": 4.939930960924619e-05, "loss": 1.3137739181518555, "step": 28700 }, { "epoch": 2.446527481891777, "grad_norm": 2.421875, "learning_rate": 4.939870809028331e-05, "loss": 1.4002517700195312, "step": 28710 }, { "epoch": 2.4473796335747764, "grad_norm": 2.5625, "learning_rate": 4.939810627396273e-05, "loss": 1.405656337738037, "step": 28720 }, { "epoch": 2.448231785257776, "grad_norm": 2.1875, "learning_rate": 4.9397504160291777e-05, "loss": 1.4414106369018556, "step": 28730 }, { "epoch": 2.4490839369407755, "grad_norm": 2.515625, "learning_rate": 4.9396901749277794e-05, "loss": 1.4093645095825196, "step": 28740 }, { "epoch": 2.449936088623775, "grad_norm": 3.53125, "learning_rate": 4.939629904092813e-05, "loss": 1.2679196357727052, "step": 28750 }, { "epoch": 2.4507882403067747, "grad_norm": 2.21875, "learning_rate": 4.939569603525012e-05, "loss": 1.2849460601806642, "step": 28760 }, { "epoch": 2.4516403919897742, "grad_norm": 2.109375, "learning_rate": 4.9395092732251115e-05, "loss": 1.4199015617370605, "step": 28770 }, { "epoch": 2.452492543672774, "grad_norm": 2.65625, "learning_rate": 4.939448913193847e-05, "loss": 1.2559932708740233, "step": 28780 }, { "epoch": 2.4533446953557734, "grad_norm": 2.15625, "learning_rate": 4.939388523431955e-05, "loss": 1.492990207672119, "step": 28790 }, { "epoch": 2.454196847038773, "grad_norm": 2.53125, "learning_rate": 4.93932810394017e-05, "loss": 1.3471221923828125, "step": 28800 }, { "epoch": 2.4550489987217725, "grad_norm": 4.40625, "learning_rate": 4.939267654719228e-05, "loss": 1.3671221733093262, "step": 28810 }, { "epoch": 2.455901150404772, "grad_norm": 1.9140625, "learning_rate": 4.939207175769869e-05, "loss": 1.4850281715393066, "step": 28820 }, { "epoch": 2.4567533020877717, "grad_norm": 3.59375, "learning_rate": 4.939146667092827e-05, "loss": 1.3802202224731446, "step": 28830 }, { "epoch": 2.457605453770771, "grad_norm": 1.765625, "learning_rate": 4.939086128688839e-05, "loss": 1.2810551643371582, "step": 28840 }, { "epoch": 2.458457605453771, "grad_norm": 1.3828125, "learning_rate": 4.9390255605586454e-05, "loss": 1.4194900512695312, "step": 28850 }, { "epoch": 2.4593097571367704, "grad_norm": 3.234375, "learning_rate": 4.938964962702982e-05, "loss": 1.2744488716125488, "step": 28860 }, { "epoch": 2.46016190881977, "grad_norm": 1.7109375, "learning_rate": 4.938904335122589e-05, "loss": 1.3545391082763671, "step": 28870 }, { "epoch": 2.4610140605027695, "grad_norm": 2.015625, "learning_rate": 4.938843677818205e-05, "loss": 1.3542751312255858, "step": 28880 }, { "epoch": 2.461866212185769, "grad_norm": 1.8984375, "learning_rate": 4.938782990790569e-05, "loss": 1.3741912841796875, "step": 28890 }, { "epoch": 2.4627183638687686, "grad_norm": 2.421875, "learning_rate": 4.93872227404042e-05, "loss": 1.3468076705932617, "step": 28900 }, { "epoch": 2.463570515551768, "grad_norm": 2.4375, "learning_rate": 4.9386615275684986e-05, "loss": 1.2436275482177734, "step": 28910 }, { "epoch": 2.4644226672347678, "grad_norm": 4.1875, "learning_rate": 4.938600751375545e-05, "loss": 1.3338254928588866, "step": 28920 }, { "epoch": 2.4652748189177673, "grad_norm": 1.8359375, "learning_rate": 4.9385399454623e-05, "loss": 1.29441499710083, "step": 28930 }, { "epoch": 2.466126970600767, "grad_norm": 2.9375, "learning_rate": 4.938479109829505e-05, "loss": 1.318662452697754, "step": 28940 }, { "epoch": 2.4669791222837665, "grad_norm": 1.703125, "learning_rate": 4.9384182444779004e-05, "loss": 1.3837244033813476, "step": 28950 }, { "epoch": 2.467831273966766, "grad_norm": 4.0625, "learning_rate": 4.938357349408229e-05, "loss": 1.361196517944336, "step": 28960 }, { "epoch": 2.4686834256497656, "grad_norm": 3.328125, "learning_rate": 4.9382964246212325e-05, "loss": 1.2853667259216308, "step": 28970 }, { "epoch": 2.469535577332765, "grad_norm": 2.71875, "learning_rate": 4.9382354701176534e-05, "loss": 1.2314040184020996, "step": 28980 }, { "epoch": 2.4703877290157648, "grad_norm": 3.3125, "learning_rate": 4.938174485898236e-05, "loss": 1.281636619567871, "step": 28990 }, { "epoch": 2.4712398806987643, "grad_norm": 2.140625, "learning_rate": 4.9381134719637214e-05, "loss": 1.3695212364196778, "step": 29000 }, { "epoch": 2.472092032381764, "grad_norm": 2.265625, "learning_rate": 4.938052428314853e-05, "loss": 1.2959664344787598, "step": 29010 }, { "epoch": 2.4729441840647635, "grad_norm": 2.40625, "learning_rate": 4.937991354952377e-05, "loss": 1.3901883125305177, "step": 29020 }, { "epoch": 2.473796335747763, "grad_norm": 2.875, "learning_rate": 4.9379302518770366e-05, "loss": 1.3966879844665527, "step": 29030 }, { "epoch": 2.4746484874307626, "grad_norm": 2.984375, "learning_rate": 4.937869119089576e-05, "loss": 1.4291330337524415, "step": 29040 }, { "epoch": 2.475500639113762, "grad_norm": 1.953125, "learning_rate": 4.937807956590741e-05, "loss": 1.353662109375, "step": 29050 }, { "epoch": 2.4763527907967617, "grad_norm": 2.15625, "learning_rate": 4.9377467643812764e-05, "loss": 1.444263744354248, "step": 29060 }, { "epoch": 2.4772049424797613, "grad_norm": 1.9921875, "learning_rate": 4.9376855424619285e-05, "loss": 1.2940515518188476, "step": 29070 }, { "epoch": 2.478057094162761, "grad_norm": 2.28125, "learning_rate": 4.937624290833443e-05, "loss": 1.2799890518188477, "step": 29080 }, { "epoch": 2.4789092458457604, "grad_norm": 2.65625, "learning_rate": 4.9375630094965665e-05, "loss": 1.3406128883361816, "step": 29090 }, { "epoch": 2.47976139752876, "grad_norm": 2.03125, "learning_rate": 4.937501698452046e-05, "loss": 1.5718887329101563, "step": 29100 }, { "epoch": 2.4806135492117596, "grad_norm": 1.6796875, "learning_rate": 4.937440357700629e-05, "loss": 1.375706100463867, "step": 29110 }, { "epoch": 2.481465700894759, "grad_norm": 2.1875, "learning_rate": 4.937378987243063e-05, "loss": 1.3936297416687011, "step": 29120 }, { "epoch": 2.4823178525777587, "grad_norm": 2.59375, "learning_rate": 4.937317587080096e-05, "loss": 1.3556442260742188, "step": 29130 }, { "epoch": 2.4831700042607583, "grad_norm": 2.296875, "learning_rate": 4.9372561572124754e-05, "loss": 1.3072365760803222, "step": 29140 }, { "epoch": 2.484022155943758, "grad_norm": 1.875, "learning_rate": 4.9371946976409505e-05, "loss": 1.288632583618164, "step": 29150 }, { "epoch": 2.4848743076267574, "grad_norm": 2.28125, "learning_rate": 4.9371332083662705e-05, "loss": 1.323284912109375, "step": 29160 }, { "epoch": 2.485726459309757, "grad_norm": 2.421875, "learning_rate": 4.937071689389185e-05, "loss": 1.3979221343994142, "step": 29170 }, { "epoch": 2.4865786109927566, "grad_norm": 2.515625, "learning_rate": 4.937010140710443e-05, "loss": 1.361314296722412, "step": 29180 }, { "epoch": 2.487430762675756, "grad_norm": 2.671875, "learning_rate": 4.936948562330795e-05, "loss": 1.287645149230957, "step": 29190 }, { "epoch": 2.4882829143587557, "grad_norm": 2.421875, "learning_rate": 4.936886954250992e-05, "loss": 1.4023371696472169, "step": 29200 }, { "epoch": 2.4891350660417553, "grad_norm": 2.90625, "learning_rate": 4.936825316471784e-05, "loss": 1.4387492179870605, "step": 29210 }, { "epoch": 2.489987217724755, "grad_norm": 2.125, "learning_rate": 4.936763648993923e-05, "loss": 1.404250717163086, "step": 29220 }, { "epoch": 2.4908393694077544, "grad_norm": 3.015625, "learning_rate": 4.936701951818159e-05, "loss": 1.2755617141723632, "step": 29230 }, { "epoch": 2.491691521090754, "grad_norm": 1.609375, "learning_rate": 4.936640224945246e-05, "loss": 1.452579879760742, "step": 29240 }, { "epoch": 2.4925436727737535, "grad_norm": 1.640625, "learning_rate": 4.9365784683759355e-05, "loss": 1.353536033630371, "step": 29250 }, { "epoch": 2.493395824456753, "grad_norm": 3.125, "learning_rate": 4.9365166821109795e-05, "loss": 1.3578160285949707, "step": 29260 }, { "epoch": 2.4942479761397527, "grad_norm": 2.328125, "learning_rate": 4.9364548661511315e-05, "loss": 1.3008554458618165, "step": 29270 }, { "epoch": 2.4951001278227523, "grad_norm": 1.5234375, "learning_rate": 4.9363930204971455e-05, "loss": 1.3821802139282227, "step": 29280 }, { "epoch": 2.495952279505752, "grad_norm": 1.984375, "learning_rate": 4.936331145149774e-05, "loss": 1.4227818489074706, "step": 29290 }, { "epoch": 2.4968044311887514, "grad_norm": 2.296875, "learning_rate": 4.936269240109772e-05, "loss": 1.2677241325378419, "step": 29300 }, { "epoch": 2.497656582871751, "grad_norm": 1.8671875, "learning_rate": 4.936207305377894e-05, "loss": 1.4308663368225099, "step": 29310 }, { "epoch": 2.4985087345547505, "grad_norm": 2.4375, "learning_rate": 4.9361453409548944e-05, "loss": 1.2682315826416015, "step": 29320 }, { "epoch": 2.49936088623775, "grad_norm": 1.859375, "learning_rate": 4.9360833468415285e-05, "loss": 1.4154464721679687, "step": 29330 }, { "epoch": 2.4997869620792503, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6483, "eval_loss": 1.2322964668273926, "eval_runtime": 53.8322, "eval_samples_per_second": 367.104, "eval_steps_per_second": 22.96, "step": 29335 }, { "epoch": 2.5002130379207497, "grad_norm": 2.90625, "learning_rate": 4.936021323038552e-05, "loss": 1.27966251373291, "step": 29340 }, { "epoch": 2.5010651896037492, "grad_norm": 1.9296875, "learning_rate": 4.93595926954672e-05, "loss": 1.4891020774841308, "step": 29350 }, { "epoch": 2.501917341286749, "grad_norm": 2.65625, "learning_rate": 4.93589718636679e-05, "loss": 1.282087802886963, "step": 29360 }, { "epoch": 2.5027694929697484, "grad_norm": 2.046875, "learning_rate": 4.935835073499518e-05, "loss": 1.300413227081299, "step": 29370 }, { "epoch": 2.503621644652748, "grad_norm": 2.34375, "learning_rate": 4.935772930945661e-05, "loss": 1.4406825065612794, "step": 29380 }, { "epoch": 2.5044737963357475, "grad_norm": 2.609375, "learning_rate": 4.935710758705977e-05, "loss": 1.3830117225646972, "step": 29390 }, { "epoch": 2.505325948018747, "grad_norm": 2.828125, "learning_rate": 4.935648556781223e-05, "loss": 1.3100418090820312, "step": 29400 }, { "epoch": 2.5061780997017467, "grad_norm": 3.5625, "learning_rate": 4.935586325172157e-05, "loss": 1.3589463233947754, "step": 29410 }, { "epoch": 2.5070302513847462, "grad_norm": 3.65625, "learning_rate": 4.935524063879538e-05, "loss": 1.4256097793579101, "step": 29420 }, { "epoch": 2.507882403067746, "grad_norm": 2.765625, "learning_rate": 4.935461772904124e-05, "loss": 1.4427206993103028, "step": 29430 }, { "epoch": 2.5087345547507454, "grad_norm": 2.171875, "learning_rate": 4.935399452246675e-05, "loss": 1.3416414260864258, "step": 29440 }, { "epoch": 2.509586706433745, "grad_norm": 2.234375, "learning_rate": 4.93533710190795e-05, "loss": 1.3423335075378418, "step": 29450 }, { "epoch": 2.510438858116745, "grad_norm": 1.921875, "learning_rate": 4.93527472188871e-05, "loss": 1.307940673828125, "step": 29460 }, { "epoch": 2.5112910097997445, "grad_norm": 3.234375, "learning_rate": 4.935212312189713e-05, "loss": 1.360771656036377, "step": 29470 }, { "epoch": 2.512143161482744, "grad_norm": 2.78125, "learning_rate": 4.935149872811722e-05, "loss": 1.3484977722167968, "step": 29480 }, { "epoch": 2.5129953131657436, "grad_norm": 2.65625, "learning_rate": 4.935087403755496e-05, "loss": 1.444387722015381, "step": 29490 }, { "epoch": 2.513847464848743, "grad_norm": 1.5703125, "learning_rate": 4.9350249050217976e-05, "loss": 1.4193211555480958, "step": 29500 }, { "epoch": 2.514699616531743, "grad_norm": 2.234375, "learning_rate": 4.9349623766113886e-05, "loss": 1.2049659729003905, "step": 29510 }, { "epoch": 2.5155517682147424, "grad_norm": 2.234375, "learning_rate": 4.9348998185250306e-05, "loss": 1.3439687728881835, "step": 29520 }, { "epoch": 2.516403919897742, "grad_norm": 1.5390625, "learning_rate": 4.934837230763485e-05, "loss": 1.2886780738830566, "step": 29530 }, { "epoch": 2.5172560715807415, "grad_norm": 2.46875, "learning_rate": 4.934774613327517e-05, "loss": 1.4710840225219726, "step": 29540 }, { "epoch": 2.518108223263741, "grad_norm": 2.640625, "learning_rate": 4.9347119662178875e-05, "loss": 1.3660633087158203, "step": 29550 }, { "epoch": 2.5189603749467406, "grad_norm": 2.5, "learning_rate": 4.934649289435361e-05, "loss": 1.4801499366760253, "step": 29560 }, { "epoch": 2.51981252662974, "grad_norm": 2.078125, "learning_rate": 4.934586582980702e-05, "loss": 1.2702152252197265, "step": 29570 }, { "epoch": 2.5206646783127398, "grad_norm": 1.8125, "learning_rate": 4.934523846854674e-05, "loss": 1.2137346267700195, "step": 29580 }, { "epoch": 2.5215168299957393, "grad_norm": 3.96875, "learning_rate": 4.9344610810580406e-05, "loss": 1.2663664817810059, "step": 29590 }, { "epoch": 2.522368981678739, "grad_norm": 2.234375, "learning_rate": 4.9343982855915685e-05, "loss": 1.3860721588134766, "step": 29600 }, { "epoch": 2.5232211333617385, "grad_norm": 2.875, "learning_rate": 4.934335460456022e-05, "loss": 1.5023865699768066, "step": 29610 }, { "epoch": 2.524073285044738, "grad_norm": 3.078125, "learning_rate": 4.934272605652167e-05, "loss": 1.2806551933288575, "step": 29620 }, { "epoch": 2.5249254367277376, "grad_norm": 2.109375, "learning_rate": 4.934209721180769e-05, "loss": 1.3907343864440918, "step": 29630 }, { "epoch": 2.525777588410737, "grad_norm": 1.7109375, "learning_rate": 4.934146807042596e-05, "loss": 1.23533353805542, "step": 29640 }, { "epoch": 2.5266297400937368, "grad_norm": 2.25, "learning_rate": 4.934083863238414e-05, "loss": 1.438609504699707, "step": 29650 }, { "epoch": 2.5274818917767363, "grad_norm": 3.625, "learning_rate": 4.934020889768988e-05, "loss": 1.4318655014038086, "step": 29660 }, { "epoch": 2.528334043459736, "grad_norm": 2.328125, "learning_rate": 4.9339578866350886e-05, "loss": 1.3392854690551759, "step": 29670 }, { "epoch": 2.5291861951427355, "grad_norm": 1.8203125, "learning_rate": 4.9338948538374827e-05, "loss": 1.368678379058838, "step": 29680 }, { "epoch": 2.530038346825735, "grad_norm": 2.640625, "learning_rate": 4.9338317913769374e-05, "loss": 1.2548916816711426, "step": 29690 }, { "epoch": 2.5308904985087346, "grad_norm": 2.21875, "learning_rate": 4.9337686992542225e-05, "loss": 1.4148241043090821, "step": 29700 }, { "epoch": 2.531742650191734, "grad_norm": 3.75, "learning_rate": 4.9337055774701066e-05, "loss": 1.3113828659057618, "step": 29710 }, { "epoch": 2.5325948018747337, "grad_norm": 1.6796875, "learning_rate": 4.933642426025359e-05, "loss": 1.4722095489501954, "step": 29720 }, { "epoch": 2.5334469535577333, "grad_norm": 1.859375, "learning_rate": 4.9335792449207485e-05, "loss": 1.525057029724121, "step": 29730 }, { "epoch": 2.534299105240733, "grad_norm": 2.34375, "learning_rate": 4.9335160341570466e-05, "loss": 1.450801944732666, "step": 29740 }, { "epoch": 2.5351512569237324, "grad_norm": 2.1875, "learning_rate": 4.933452793735023e-05, "loss": 1.4774545669555663, "step": 29750 }, { "epoch": 2.536003408606732, "grad_norm": 1.65625, "learning_rate": 4.9333895236554474e-05, "loss": 1.3663131713867187, "step": 29760 }, { "epoch": 2.5368555602897316, "grad_norm": 2.359375, "learning_rate": 4.9333262239190924e-05, "loss": 1.3682228088378907, "step": 29770 }, { "epoch": 2.537707711972731, "grad_norm": 2.234375, "learning_rate": 4.933262894526729e-05, "loss": 1.2798317909240722, "step": 29780 }, { "epoch": 2.5385598636557307, "grad_norm": 3.28125, "learning_rate": 4.933199535479129e-05, "loss": 1.3295561790466308, "step": 29790 }, { "epoch": 2.5394120153387303, "grad_norm": 3.296875, "learning_rate": 4.933136146777065e-05, "loss": 1.3105427742004394, "step": 29800 }, { "epoch": 2.54026416702173, "grad_norm": 2.46875, "learning_rate": 4.933072728421309e-05, "loss": 1.497885513305664, "step": 29810 }, { "epoch": 2.5411163187047294, "grad_norm": 4.5625, "learning_rate": 4.933009280412633e-05, "loss": 1.320340061187744, "step": 29820 }, { "epoch": 2.541968470387729, "grad_norm": 1.875, "learning_rate": 4.932945802751812e-05, "loss": 1.3118660926818848, "step": 29830 }, { "epoch": 2.5428206220707286, "grad_norm": 2.1875, "learning_rate": 4.932882295439619e-05, "loss": 1.2551633834838867, "step": 29840 }, { "epoch": 2.543672773753728, "grad_norm": 2.328125, "learning_rate": 4.932818758476828e-05, "loss": 1.1906573295593261, "step": 29850 }, { "epoch": 2.5445249254367277, "grad_norm": 1.765625, "learning_rate": 4.932755191864213e-05, "loss": 1.342491054534912, "step": 29860 }, { "epoch": 2.5453770771197273, "grad_norm": 2.3125, "learning_rate": 4.9326915956025486e-05, "loss": 1.3090352058410644, "step": 29870 }, { "epoch": 2.546229228802727, "grad_norm": 2.09375, "learning_rate": 4.93262796969261e-05, "loss": 1.3764965057373046, "step": 29880 }, { "epoch": 2.5470813804857264, "grad_norm": 3.4375, "learning_rate": 4.932564314135174e-05, "loss": 1.437370204925537, "step": 29890 }, { "epoch": 2.547933532168726, "grad_norm": 1.8203125, "learning_rate": 4.932500628931015e-05, "loss": 1.3897591590881349, "step": 29900 }, { "epoch": 2.5487856838517255, "grad_norm": 2.265625, "learning_rate": 4.932436914080909e-05, "loss": 1.3559428215026856, "step": 29910 }, { "epoch": 2.549637835534725, "grad_norm": 2.90625, "learning_rate": 4.932373169585633e-05, "loss": 1.419397735595703, "step": 29920 }, { "epoch": 2.5504899872177247, "grad_norm": 1.765625, "learning_rate": 4.9323093954459645e-05, "loss": 1.3662492752075195, "step": 29930 }, { "epoch": 2.5513421389007243, "grad_norm": 2.171875, "learning_rate": 4.9322455916626785e-05, "loss": 1.2460050582885742, "step": 29940 }, { "epoch": 2.552194290583724, "grad_norm": 4.5, "learning_rate": 4.932181758236556e-05, "loss": 1.3317615509033203, "step": 29950 }, { "epoch": 2.5530464422667234, "grad_norm": 2.21875, "learning_rate": 4.932117895168372e-05, "loss": 1.4020886421203613, "step": 29960 }, { "epoch": 2.553898593949723, "grad_norm": 2.578125, "learning_rate": 4.932054002458906e-05, "loss": 1.3082941055297852, "step": 29970 }, { "epoch": 2.5547507456327225, "grad_norm": 1.890625, "learning_rate": 4.931990080108937e-05, "loss": 1.5144421577453613, "step": 29980 }, { "epoch": 2.555602897315722, "grad_norm": 2.546875, "learning_rate": 4.9319261281192435e-05, "loss": 1.284346103668213, "step": 29990 }, { "epoch": 2.5564550489987217, "grad_norm": 3.625, "learning_rate": 4.931862146490606e-05, "loss": 1.381810474395752, "step": 30000 }, { "epoch": 2.5573072006817212, "grad_norm": 2.03125, "learning_rate": 4.931798135223802e-05, "loss": 1.5628196716308593, "step": 30010 }, { "epoch": 2.558159352364721, "grad_norm": 2.609375, "learning_rate": 4.9317340943196136e-05, "loss": 1.4118366241455078, "step": 30020 }, { "epoch": 2.5590115040477204, "grad_norm": 2.84375, "learning_rate": 4.9316700237788205e-05, "loss": 1.3554097175598145, "step": 30030 }, { "epoch": 2.55986365573072, "grad_norm": 1.5078125, "learning_rate": 4.9316059236022046e-05, "loss": 1.4095168113708496, "step": 30040 }, { "epoch": 2.5607158074137195, "grad_norm": 3.3125, "learning_rate": 4.931541793790545e-05, "loss": 1.2834081649780273, "step": 30050 }, { "epoch": 2.561567959096719, "grad_norm": 2.9375, "learning_rate": 4.931477634344626e-05, "loss": 1.2743016242980958, "step": 30060 }, { "epoch": 2.5624201107797187, "grad_norm": 2.15625, "learning_rate": 4.9314134452652275e-05, "loss": 1.394196128845215, "step": 30070 }, { "epoch": 2.563272262462718, "grad_norm": 2.765625, "learning_rate": 4.931349226553132e-05, "loss": 1.4462595939636231, "step": 30080 }, { "epoch": 2.564124414145718, "grad_norm": 3.765625, "learning_rate": 4.931284978209123e-05, "loss": 1.0401747703552247, "step": 30090 }, { "epoch": 2.5649765658287174, "grad_norm": 2.734375, "learning_rate": 4.931220700233983e-05, "loss": 1.4441295623779298, "step": 30100 }, { "epoch": 2.565828717511717, "grad_norm": 2.234375, "learning_rate": 4.931156392628495e-05, "loss": 1.269247817993164, "step": 30110 }, { "epoch": 2.5666808691947165, "grad_norm": 2.375, "learning_rate": 4.9310920553934445e-05, "loss": 1.3239253044128418, "step": 30120 }, { "epoch": 2.5675330208777165, "grad_norm": 2.59375, "learning_rate": 4.931027688529613e-05, "loss": 1.350529384613037, "step": 30130 }, { "epoch": 2.568385172560716, "grad_norm": 2.484375, "learning_rate": 4.930963292037787e-05, "loss": 1.4028814315795899, "step": 30140 }, { "epoch": 2.5692373242437156, "grad_norm": 2.375, "learning_rate": 4.9308988659187506e-05, "loss": 1.4341989517211915, "step": 30150 }, { "epoch": 2.570089475926715, "grad_norm": 2.515625, "learning_rate": 4.930834410173288e-05, "loss": 1.3713424682617188, "step": 30160 }, { "epoch": 2.570941627609715, "grad_norm": 4.15625, "learning_rate": 4.930769924802187e-05, "loss": 1.3315113067626954, "step": 30170 }, { "epoch": 2.5717937792927144, "grad_norm": 3.953125, "learning_rate": 4.930705409806232e-05, "loss": 1.4123797416687012, "step": 30180 }, { "epoch": 2.572645930975714, "grad_norm": 2.0625, "learning_rate": 4.930640865186209e-05, "loss": 1.2824299812316895, "step": 30190 }, { "epoch": 2.5734980826587135, "grad_norm": 1.5, "learning_rate": 4.930576290942906e-05, "loss": 1.2757489204406738, "step": 30200 }, { "epoch": 2.574350234341713, "grad_norm": 1.8828125, "learning_rate": 4.930511687077109e-05, "loss": 1.2927535057067872, "step": 30210 }, { "epoch": 2.5752023860247126, "grad_norm": 3.953125, "learning_rate": 4.9304470535896055e-05, "loss": 1.2547629356384278, "step": 30220 }, { "epoch": 2.576054537707712, "grad_norm": 1.953125, "learning_rate": 4.930382390481183e-05, "loss": 1.4209821701049805, "step": 30230 }, { "epoch": 2.5769066893907118, "grad_norm": 5.15625, "learning_rate": 4.9303176977526294e-05, "loss": 1.3510753631591796, "step": 30240 }, { "epoch": 2.5777588410737113, "grad_norm": 1.859375, "learning_rate": 4.930252975404734e-05, "loss": 1.3786381721496581, "step": 30250 }, { "epoch": 2.578610992756711, "grad_norm": 2.015625, "learning_rate": 4.930188223438285e-05, "loss": 1.339045524597168, "step": 30260 }, { "epoch": 2.5794631444397105, "grad_norm": 2.203125, "learning_rate": 4.930123441854072e-05, "loss": 1.4507128715515136, "step": 30270 }, { "epoch": 2.58031529612271, "grad_norm": 2.796875, "learning_rate": 4.930058630652884e-05, "loss": 1.3379980087280274, "step": 30280 }, { "epoch": 2.5811674478057096, "grad_norm": 2.15625, "learning_rate": 4.9299937898355114e-05, "loss": 1.3377238273620606, "step": 30290 }, { "epoch": 2.582019599488709, "grad_norm": 2.1875, "learning_rate": 4.9299289194027446e-05, "loss": 1.280623722076416, "step": 30300 }, { "epoch": 2.5828717511717088, "grad_norm": 3.140625, "learning_rate": 4.929864019355372e-05, "loss": 1.186635971069336, "step": 30310 }, { "epoch": 2.5837239028547083, "grad_norm": 1.8828125, "learning_rate": 4.9297990896941884e-05, "loss": 1.516037940979004, "step": 30320 }, { "epoch": 2.584576054537708, "grad_norm": 2.578125, "learning_rate": 4.929734130419982e-05, "loss": 1.3393630981445312, "step": 30330 }, { "epoch": 2.5854282062207075, "grad_norm": 1.75, "learning_rate": 4.9296691415335464e-05, "loss": 1.421670627593994, "step": 30340 }, { "epoch": 2.586280357903707, "grad_norm": 5.28125, "learning_rate": 4.929604123035672e-05, "loss": 1.3098952293395996, "step": 30350 }, { "epoch": 2.5871325095867066, "grad_norm": 2.625, "learning_rate": 4.9295390749271516e-05, "loss": 1.4080713272094727, "step": 30360 }, { "epoch": 2.587984661269706, "grad_norm": 2.984375, "learning_rate": 4.929473997208779e-05, "loss": 1.4093386650085449, "step": 30370 }, { "epoch": 2.5888368129527057, "grad_norm": 2.296875, "learning_rate": 4.929408889881347e-05, "loss": 1.5019168853759766, "step": 30380 }, { "epoch": 2.5896889646357053, "grad_norm": 2.203125, "learning_rate": 4.929343752945648e-05, "loss": 1.4770524978637696, "step": 30390 }, { "epoch": 2.590541116318705, "grad_norm": 1.828125, "learning_rate": 4.929278586402477e-05, "loss": 1.322902011871338, "step": 30400 }, { "epoch": 2.5913932680017044, "grad_norm": 3.34375, "learning_rate": 4.929213390252628e-05, "loss": 1.3434919357299804, "step": 30410 }, { "epoch": 2.592245419684704, "grad_norm": 2.359375, "learning_rate": 4.929148164496896e-05, "loss": 1.3103975296020507, "step": 30420 }, { "epoch": 2.5930975713677036, "grad_norm": 2.453125, "learning_rate": 4.929082909136075e-05, "loss": 1.3372089385986328, "step": 30430 }, { "epoch": 2.593949723050703, "grad_norm": 2.84375, "learning_rate": 4.92901762417096e-05, "loss": 1.5018144607543946, "step": 30440 }, { "epoch": 2.5948018747337027, "grad_norm": 2.078125, "learning_rate": 4.9289523096023476e-05, "loss": 1.3670042037963868, "step": 30450 }, { "epoch": 2.5956540264167023, "grad_norm": 4.46875, "learning_rate": 4.9288869654310336e-05, "loss": 1.3297239303588868, "step": 30460 }, { "epoch": 2.596506178099702, "grad_norm": 1.8828125, "learning_rate": 4.928821591657815e-05, "loss": 1.535544490814209, "step": 30470 }, { "epoch": 2.5973583297827014, "grad_norm": 2.484375, "learning_rate": 4.928756188283487e-05, "loss": 1.2856406211853026, "step": 30480 }, { "epoch": 2.598210481465701, "grad_norm": 2.765625, "learning_rate": 4.928690755308848e-05, "loss": 1.3494816780090333, "step": 30490 }, { "epoch": 2.5990626331487006, "grad_norm": 2.09375, "learning_rate": 4.928625292734694e-05, "loss": 1.4339337348937988, "step": 30500 }, { "epoch": 2.5999147848317, "grad_norm": 2.84375, "learning_rate": 4.928559800561825e-05, "loss": 1.4152125358581542, "step": 30510 }, { "epoch": 2.6007669365146997, "grad_norm": 2.609375, "learning_rate": 4.928494278791038e-05, "loss": 1.1130412101745606, "step": 30520 }, { "epoch": 2.6016190881976993, "grad_norm": 2.09375, "learning_rate": 4.928428727423131e-05, "loss": 1.4915597915649415, "step": 30530 }, { "epoch": 2.602471239880699, "grad_norm": 4.5, "learning_rate": 4.928363146458904e-05, "loss": 1.3426733016967773, "step": 30540 }, { "epoch": 2.6033233915636984, "grad_norm": 2.09375, "learning_rate": 4.928297535899156e-05, "loss": 1.3834136962890624, "step": 30550 }, { "epoch": 2.604175543246698, "grad_norm": 2.453125, "learning_rate": 4.9282318957446846e-05, "loss": 1.3069187164306642, "step": 30560 }, { "epoch": 2.6050276949296975, "grad_norm": 3.703125, "learning_rate": 4.928166225996294e-05, "loss": 1.2632349014282227, "step": 30570 }, { "epoch": 2.605879846612697, "grad_norm": 3.84375, "learning_rate": 4.92810052665478e-05, "loss": 1.3238301277160645, "step": 30580 }, { "epoch": 2.6067319982956967, "grad_norm": 2.671875, "learning_rate": 4.928034797720946e-05, "loss": 1.3331503868103027, "step": 30590 }, { "epoch": 2.6075841499786963, "grad_norm": 2.25, "learning_rate": 4.927969039195593e-05, "loss": 1.1076024055480957, "step": 30600 }, { "epoch": 2.608436301661696, "grad_norm": 1.6484375, "learning_rate": 4.927903251079522e-05, "loss": 1.316043758392334, "step": 30610 }, { "epoch": 2.6092884533446954, "grad_norm": 2.203125, "learning_rate": 4.9278374333735346e-05, "loss": 1.3158253669738769, "step": 30620 }, { "epoch": 2.610140605027695, "grad_norm": 3.125, "learning_rate": 4.927771586078432e-05, "loss": 1.2350542068481445, "step": 30630 }, { "epoch": 2.6109927567106945, "grad_norm": 3.84375, "learning_rate": 4.927705709195018e-05, "loss": 1.2951837539672852, "step": 30640 }, { "epoch": 2.611844908393694, "grad_norm": 4.96875, "learning_rate": 4.927639802724097e-05, "loss": 1.336087989807129, "step": 30650 }, { "epoch": 2.6126970600766937, "grad_norm": 2.0, "learning_rate": 4.9275738666664685e-05, "loss": 1.2287823677062988, "step": 30660 }, { "epoch": 2.6135492117596932, "grad_norm": 2.3125, "learning_rate": 4.927507901022939e-05, "loss": 1.4634740829467774, "step": 30670 }, { "epoch": 2.614401363442693, "grad_norm": 2.65625, "learning_rate": 4.927441905794311e-05, "loss": 1.331202507019043, "step": 30680 }, { "epoch": 2.6152535151256924, "grad_norm": 3.203125, "learning_rate": 4.927375880981389e-05, "loss": 1.1506818771362304, "step": 30690 }, { "epoch": 2.616105666808692, "grad_norm": 2.1875, "learning_rate": 4.927309826584978e-05, "loss": 1.4087515830993653, "step": 30700 }, { "epoch": 2.6169578184916915, "grad_norm": 2.078125, "learning_rate": 4.927243742605884e-05, "loss": 1.3345422744750977, "step": 30710 }, { "epoch": 2.617809970174691, "grad_norm": 2.59375, "learning_rate": 4.92717762904491e-05, "loss": 1.3153846740722657, "step": 30720 }, { "epoch": 2.6186621218576907, "grad_norm": 1.953125, "learning_rate": 4.927111485902864e-05, "loss": 1.2699180603027345, "step": 30730 }, { "epoch": 2.61951427354069, "grad_norm": 2.421875, "learning_rate": 4.9270453131805516e-05, "loss": 1.3822647094726563, "step": 30740 }, { "epoch": 2.62036642522369, "grad_norm": 2.984375, "learning_rate": 4.9269791108787785e-05, "loss": 1.3856178283691407, "step": 30750 }, { "epoch": 2.6212185769066894, "grad_norm": 2.0, "learning_rate": 4.926912878998352e-05, "loss": 1.3976957321166992, "step": 30760 }, { "epoch": 2.622070728589689, "grad_norm": 1.8359375, "learning_rate": 4.926846617540079e-05, "loss": 1.4839769363403321, "step": 30770 }, { "epoch": 2.6229228802726885, "grad_norm": 2.28125, "learning_rate": 4.926780326504767e-05, "loss": 1.24088716506958, "step": 30780 }, { "epoch": 2.623775031955688, "grad_norm": 1.9921875, "learning_rate": 4.9267140058932254e-05, "loss": 1.4197053909301758, "step": 30790 }, { "epoch": 2.6246271836386876, "grad_norm": 2.0625, "learning_rate": 4.926647655706261e-05, "loss": 1.4368909835815429, "step": 30800 }, { "epoch": 2.625479335321687, "grad_norm": 2.578125, "learning_rate": 4.9265812759446825e-05, "loss": 1.3914610862731933, "step": 30810 }, { "epoch": 2.6263314870046868, "grad_norm": 4.09375, "learning_rate": 4.926514866609299e-05, "loss": 1.3018427848815919, "step": 30820 }, { "epoch": 2.6271836386876863, "grad_norm": 2.59375, "learning_rate": 4.92644842770092e-05, "loss": 1.4160009384155274, "step": 30830 }, { "epoch": 2.628035790370686, "grad_norm": 2.25, "learning_rate": 4.926381959220355e-05, "loss": 1.3657048225402832, "step": 30840 }, { "epoch": 2.6288879420536855, "grad_norm": 1.7109375, "learning_rate": 4.9263154611684145e-05, "loss": 1.234527587890625, "step": 30850 }, { "epoch": 2.629740093736685, "grad_norm": 2.453125, "learning_rate": 4.926248933545909e-05, "loss": 1.5743074417114258, "step": 30860 }, { "epoch": 2.6305922454196846, "grad_norm": 2.484375, "learning_rate": 4.92618237635365e-05, "loss": 1.1575037002563477, "step": 30870 }, { "epoch": 2.631444397102684, "grad_norm": 1.921875, "learning_rate": 4.9261157895924466e-05, "loss": 1.5688873291015626, "step": 30880 }, { "epoch": 2.6322965487856838, "grad_norm": 1.984375, "learning_rate": 4.926049173263112e-05, "loss": 1.2568506240844726, "step": 30890 }, { "epoch": 2.6331487004686833, "grad_norm": 2.640625, "learning_rate": 4.925982527366457e-05, "loss": 1.3693163871765137, "step": 30900 }, { "epoch": 2.634000852151683, "grad_norm": 3.953125, "learning_rate": 4.9259158519032946e-05, "loss": 1.3757155418395997, "step": 30910 }, { "epoch": 2.6348530038346825, "grad_norm": 2.03125, "learning_rate": 4.925849146874437e-05, "loss": 1.4576701164245605, "step": 30920 }, { "epoch": 2.635705155517682, "grad_norm": 2.3125, "learning_rate": 4.925782412280697e-05, "loss": 1.4444282531738282, "step": 30930 }, { "epoch": 2.6365573072006816, "grad_norm": 2.609375, "learning_rate": 4.9257156481228894e-05, "loss": 1.4014348030090331, "step": 30940 }, { "epoch": 2.637409458883681, "grad_norm": 2.421875, "learning_rate": 4.925648854401826e-05, "loss": 1.5031817436218262, "step": 30950 }, { "epoch": 2.6382616105666807, "grad_norm": 2.484375, "learning_rate": 4.925582031118321e-05, "loss": 1.4734247207641602, "step": 30960 }, { "epoch": 2.6391137622496803, "grad_norm": 2.921875, "learning_rate": 4.9255151782731906e-05, "loss": 1.344524097442627, "step": 30970 }, { "epoch": 2.63996591393268, "grad_norm": 2.515625, "learning_rate": 4.925448295867248e-05, "loss": 1.3840701103210449, "step": 30980 }, { "epoch": 2.6408180656156794, "grad_norm": 2.296875, "learning_rate": 4.9253813839013085e-05, "loss": 1.2817405700683593, "step": 30990 }, { "epoch": 2.641670217298679, "grad_norm": 1.3359375, "learning_rate": 4.9253144423761874e-05, "loss": 1.3275320053100585, "step": 31000 }, { "epoch": 2.6425223689816786, "grad_norm": 2.59375, "learning_rate": 4.925247471292702e-05, "loss": 1.3428914070129394, "step": 31010 }, { "epoch": 2.643374520664678, "grad_norm": 2.453125, "learning_rate": 4.9251804706516665e-05, "loss": 1.3752408981323243, "step": 31020 }, { "epoch": 2.6442266723476777, "grad_norm": 2.03125, "learning_rate": 4.9251134404538985e-05, "loss": 1.229560375213623, "step": 31030 }, { "epoch": 2.6450788240306773, "grad_norm": 2.84375, "learning_rate": 4.925046380700215e-05, "loss": 1.220961570739746, "step": 31040 }, { "epoch": 2.645930975713677, "grad_norm": 1.6484375, "learning_rate": 4.9249792913914325e-05, "loss": 1.3560674667358399, "step": 31050 }, { "epoch": 2.6467831273966764, "grad_norm": 2.34375, "learning_rate": 4.92491217252837e-05, "loss": 1.2843482971191407, "step": 31060 }, { "epoch": 2.647635279079676, "grad_norm": 2.15625, "learning_rate": 4.924845024111845e-05, "loss": 1.359001350402832, "step": 31070 }, { "epoch": 2.6484874307626756, "grad_norm": 3.03125, "learning_rate": 4.9247778461426754e-05, "loss": 1.3384191513061523, "step": 31080 }, { "epoch": 2.649339582445675, "grad_norm": 2.71875, "learning_rate": 4.924710638621679e-05, "loss": 1.333875560760498, "step": 31090 }, { "epoch": 2.6501917341286747, "grad_norm": 2.203125, "learning_rate": 4.9246434015496776e-05, "loss": 1.405992603302002, "step": 31100 }, { "epoch": 2.6510438858116743, "grad_norm": 2.234375, "learning_rate": 4.9245761349274885e-05, "loss": 1.5113656997680665, "step": 31110 }, { "epoch": 2.651896037494674, "grad_norm": 2.09375, "learning_rate": 4.9245088387559316e-05, "loss": 1.4828433036804198, "step": 31120 }, { "epoch": 2.6527481891776734, "grad_norm": 3.65625, "learning_rate": 4.9244415130358276e-05, "loss": 1.2607322692871095, "step": 31130 }, { "epoch": 2.653600340860673, "grad_norm": 1.8203125, "learning_rate": 4.924374157767998e-05, "loss": 1.3174310684204102, "step": 31140 }, { "epoch": 2.6544524925436725, "grad_norm": 1.9375, "learning_rate": 4.9243067729532624e-05, "loss": 1.3702322006225587, "step": 31150 }, { "epoch": 2.655304644226672, "grad_norm": 2.125, "learning_rate": 4.924239358592442e-05, "loss": 1.4834062576293945, "step": 31160 }, { "epoch": 2.6561567959096717, "grad_norm": 1.9375, "learning_rate": 4.9241719146863585e-05, "loss": 1.4308369636535645, "step": 31170 }, { "epoch": 2.6570089475926713, "grad_norm": 1.9609375, "learning_rate": 4.924104441235834e-05, "loss": 1.3215059280395507, "step": 31180 }, { "epoch": 2.657861099275671, "grad_norm": 2.328125, "learning_rate": 4.9240369382416916e-05, "loss": 1.3206544876098634, "step": 31190 }, { "epoch": 2.6587132509586704, "grad_norm": 1.7265625, "learning_rate": 4.9239694057047526e-05, "loss": 1.502933883666992, "step": 31200 }, { "epoch": 2.65956540264167, "grad_norm": 2.96875, "learning_rate": 4.9239018436258414e-05, "loss": 1.3732146263122558, "step": 31210 }, { "epoch": 2.6604175543246695, "grad_norm": 1.671875, "learning_rate": 4.92383425200578e-05, "loss": 1.3467529296875, "step": 31220 }, { "epoch": 2.661269706007669, "grad_norm": 2.109375, "learning_rate": 4.923766630845394e-05, "loss": 1.3972507476806642, "step": 31230 }, { "epoch": 2.6621218576906687, "grad_norm": 2.40625, "learning_rate": 4.923698980145506e-05, "loss": 1.315859031677246, "step": 31240 }, { "epoch": 2.6629740093736682, "grad_norm": 1.8046875, "learning_rate": 4.9236312999069406e-05, "loss": 1.3828413009643554, "step": 31250 }, { "epoch": 2.6638261610566683, "grad_norm": 2.234375, "learning_rate": 4.923563590130523e-05, "loss": 1.3287530899047852, "step": 31260 }, { "epoch": 2.664678312739668, "grad_norm": 2.0, "learning_rate": 4.9234958508170793e-05, "loss": 1.3496082305908204, "step": 31270 }, { "epoch": 2.6655304644226674, "grad_norm": 1.875, "learning_rate": 4.9234280819674334e-05, "loss": 1.4853577613830566, "step": 31280 }, { "epoch": 2.666382616105667, "grad_norm": 3.0625, "learning_rate": 4.923360283582412e-05, "loss": 1.288885498046875, "step": 31290 }, { "epoch": 2.6672347677886665, "grad_norm": 2.21875, "learning_rate": 4.923292455662841e-05, "loss": 1.3795795440673828, "step": 31300 }, { "epoch": 2.668086919471666, "grad_norm": 2.8125, "learning_rate": 4.923224598209549e-05, "loss": 1.2927325248718262, "step": 31310 }, { "epoch": 2.6689390711546657, "grad_norm": 2.421875, "learning_rate": 4.92315671122336e-05, "loss": 1.4459827423095704, "step": 31320 }, { "epoch": 2.6697912228376652, "grad_norm": 2.375, "learning_rate": 4.923088794705103e-05, "loss": 1.400082778930664, "step": 31330 }, { "epoch": 2.670643374520665, "grad_norm": 2.328125, "learning_rate": 4.9230208486556064e-05, "loss": 1.2653387069702149, "step": 31340 }, { "epoch": 2.6714955262036644, "grad_norm": 2.203125, "learning_rate": 4.922952873075697e-05, "loss": 1.2847211837768555, "step": 31350 }, { "epoch": 2.672347677886664, "grad_norm": 1.890625, "learning_rate": 4.922884867966204e-05, "loss": 1.4250415802001952, "step": 31360 }, { "epoch": 2.6731998295696635, "grad_norm": 1.9140625, "learning_rate": 4.9228168333279557e-05, "loss": 1.276858139038086, "step": 31370 }, { "epoch": 2.674051981252663, "grad_norm": 4.59375, "learning_rate": 4.9227487691617804e-05, "loss": 1.254532527923584, "step": 31380 }, { "epoch": 2.6749041329356626, "grad_norm": 2.015625, "learning_rate": 4.92268067546851e-05, "loss": 1.3570650100708008, "step": 31390 }, { "epoch": 2.675756284618662, "grad_norm": 2.6875, "learning_rate": 4.922612552248973e-05, "loss": 1.3711977005004883, "step": 31400 }, { "epoch": 2.676608436301662, "grad_norm": 2.21875, "learning_rate": 4.922544399503999e-05, "loss": 1.3269055366516114, "step": 31410 }, { "epoch": 2.6774605879846614, "grad_norm": 2.171875, "learning_rate": 4.9224762172344194e-05, "loss": 1.3112409591674805, "step": 31420 }, { "epoch": 2.678312739667661, "grad_norm": 2.765625, "learning_rate": 4.922408005441066e-05, "loss": 1.461702537536621, "step": 31430 }, { "epoch": 2.6791648913506605, "grad_norm": 1.7890625, "learning_rate": 4.922339764124769e-05, "loss": 1.3669225692749023, "step": 31440 }, { "epoch": 2.68001704303366, "grad_norm": 2.1875, "learning_rate": 4.92227149328636e-05, "loss": 1.499018096923828, "step": 31450 }, { "epoch": 2.6808691947166596, "grad_norm": 2.828125, "learning_rate": 4.922203192926671e-05, "loss": 1.2836081504821777, "step": 31460 }, { "epoch": 2.681721346399659, "grad_norm": 1.90625, "learning_rate": 4.922134863046535e-05, "loss": 1.3146696090698242, "step": 31470 }, { "epoch": 2.6825734980826588, "grad_norm": 2.734375, "learning_rate": 4.922066503646785e-05, "loss": 1.3465105056762696, "step": 31480 }, { "epoch": 2.6834256497656583, "grad_norm": 1.6796875, "learning_rate": 4.921998114728254e-05, "loss": 1.440149688720703, "step": 31490 }, { "epoch": 2.684277801448658, "grad_norm": 1.78125, "learning_rate": 4.921929696291774e-05, "loss": 1.2229995727539062, "step": 31500 }, { "epoch": 2.6851299531316575, "grad_norm": 3.328125, "learning_rate": 4.92186124833818e-05, "loss": 1.3586673736572266, "step": 31510 }, { "epoch": 2.685982104814657, "grad_norm": 2.515625, "learning_rate": 4.921792770868307e-05, "loss": 1.3603469848632812, "step": 31520 }, { "epoch": 2.6868342564976566, "grad_norm": 2.53125, "learning_rate": 4.9217242638829886e-05, "loss": 1.2519837379455567, "step": 31530 }, { "epoch": 2.687686408180656, "grad_norm": 2.078125, "learning_rate": 4.92165572738306e-05, "loss": 1.4362545967102052, "step": 31540 }, { "epoch": 2.6885385598636558, "grad_norm": 2.203125, "learning_rate": 4.9215871613693556e-05, "loss": 1.4155193328857423, "step": 31550 }, { "epoch": 2.6893907115466553, "grad_norm": 2.09375, "learning_rate": 4.921518565842712e-05, "loss": 1.3415151596069337, "step": 31560 }, { "epoch": 2.690242863229655, "grad_norm": 3.65625, "learning_rate": 4.921449940803966e-05, "loss": 1.2958620071411133, "step": 31570 }, { "epoch": 2.6910950149126545, "grad_norm": 2.640625, "learning_rate": 4.9213812862539524e-05, "loss": 1.1113363265991212, "step": 31580 }, { "epoch": 2.691947166595654, "grad_norm": 2.140625, "learning_rate": 4.9213126021935086e-05, "loss": 1.361781692504883, "step": 31590 }, { "epoch": 2.6927993182786536, "grad_norm": 2.453125, "learning_rate": 4.921243888623471e-05, "loss": 1.5411466598510741, "step": 31600 }, { "epoch": 2.693651469961653, "grad_norm": 2.109375, "learning_rate": 4.921175145544678e-05, "loss": 1.3546491622924806, "step": 31610 }, { "epoch": 2.6945036216446527, "grad_norm": 3.328125, "learning_rate": 4.921106372957968e-05, "loss": 1.1581186294555663, "step": 31620 }, { "epoch": 2.6953557733276523, "grad_norm": 5.21875, "learning_rate": 4.921037570864177e-05, "loss": 1.2162455558776855, "step": 31630 }, { "epoch": 2.696207925010652, "grad_norm": 2.78125, "learning_rate": 4.920968739264145e-05, "loss": 1.3951882362365722, "step": 31640 }, { "epoch": 2.6970600766936514, "grad_norm": 2.109375, "learning_rate": 4.92089987815871e-05, "loss": 1.3323251724243164, "step": 31650 }, { "epoch": 2.697912228376651, "grad_norm": 2.046875, "learning_rate": 4.920830987548712e-05, "loss": 1.3539610862731934, "step": 31660 }, { "epoch": 2.6987643800596506, "grad_norm": 1.9140625, "learning_rate": 4.920762067434992e-05, "loss": 1.3427009582519531, "step": 31670 }, { "epoch": 2.69961653174265, "grad_norm": 2.859375, "learning_rate": 4.9206931178183866e-05, "loss": 1.2287185668945313, "step": 31680 }, { "epoch": 2.7004686834256497, "grad_norm": 2.3125, "learning_rate": 4.920624138699739e-05, "loss": 1.31810302734375, "step": 31690 }, { "epoch": 2.7013208351086493, "grad_norm": 2.640625, "learning_rate": 4.920555130079888e-05, "loss": 1.3658460617065429, "step": 31700 }, { "epoch": 2.702172986791649, "grad_norm": 2.140625, "learning_rate": 4.920486091959676e-05, "loss": 1.382988739013672, "step": 31710 }, { "epoch": 2.7030251384746484, "grad_norm": 1.7890625, "learning_rate": 4.920417024339944e-05, "loss": 1.48843412399292, "step": 31720 }, { "epoch": 2.703877290157648, "grad_norm": 2.734375, "learning_rate": 4.920347927221533e-05, "loss": 1.3401480674743653, "step": 31730 }, { "epoch": 2.7047294418406476, "grad_norm": 2.390625, "learning_rate": 4.920278800605286e-05, "loss": 1.447592830657959, "step": 31740 }, { "epoch": 2.705581593523647, "grad_norm": 2.15625, "learning_rate": 4.920209644492045e-05, "loss": 1.3774385452270508, "step": 31750 }, { "epoch": 2.7064337452066467, "grad_norm": 2.28125, "learning_rate": 4.920140458882653e-05, "loss": 1.3300525665283203, "step": 31760 }, { "epoch": 2.7072858968896463, "grad_norm": 2.15625, "learning_rate": 4.920071243777954e-05, "loss": 1.4381416320800782, "step": 31770 }, { "epoch": 2.708138048572646, "grad_norm": 2.828125, "learning_rate": 4.920001999178789e-05, "loss": 1.3946932792663573, "step": 31780 }, { "epoch": 2.7089902002556454, "grad_norm": 2.46875, "learning_rate": 4.919932725086006e-05, "loss": 1.3877840995788575, "step": 31790 }, { "epoch": 2.709842351938645, "grad_norm": 2.234375, "learning_rate": 4.919863421500445e-05, "loss": 1.442970085144043, "step": 31800 }, { "epoch": 2.7106945036216445, "grad_norm": 1.828125, "learning_rate": 4.919794088422954e-05, "loss": 1.368886375427246, "step": 31810 }, { "epoch": 2.711546655304644, "grad_norm": 2.71875, "learning_rate": 4.9197247258543755e-05, "loss": 1.4876214981079101, "step": 31820 }, { "epoch": 2.7123988069876437, "grad_norm": 2.984375, "learning_rate": 4.919655333795556e-05, "loss": 1.3229666709899903, "step": 31830 }, { "epoch": 2.7132509586706433, "grad_norm": 1.90625, "learning_rate": 4.919585912247342e-05, "loss": 1.4024879455566406, "step": 31840 }, { "epoch": 2.714103110353643, "grad_norm": 1.96875, "learning_rate": 4.919516461210578e-05, "loss": 1.2571581840515136, "step": 31850 }, { "epoch": 2.7149552620366424, "grad_norm": 2.1875, "learning_rate": 4.9194469806861115e-05, "loss": 1.3562400817871094, "step": 31860 }, { "epoch": 2.715807413719642, "grad_norm": 1.796875, "learning_rate": 4.9193774706747885e-05, "loss": 1.3544010162353515, "step": 31870 }, { "epoch": 2.7166595654026415, "grad_norm": 2.109375, "learning_rate": 4.919307931177457e-05, "loss": 1.362515926361084, "step": 31880 }, { "epoch": 2.717511717085641, "grad_norm": 2.171875, "learning_rate": 4.919238362194964e-05, "loss": 1.1900020599365235, "step": 31890 }, { "epoch": 2.7183638687686407, "grad_norm": 2.265625, "learning_rate": 4.919168763728157e-05, "loss": 1.4461934089660644, "step": 31900 }, { "epoch": 2.7192160204516402, "grad_norm": 2.234375, "learning_rate": 4.919099135777886e-05, "loss": 1.2622581481933595, "step": 31910 }, { "epoch": 2.72006817213464, "grad_norm": 2.21875, "learning_rate": 4.919029478344997e-05, "loss": 1.3644762992858888, "step": 31920 }, { "epoch": 2.7209203238176394, "grad_norm": 2.625, "learning_rate": 4.918959791430341e-05, "loss": 1.3819028854370117, "step": 31930 }, { "epoch": 2.7217724755006394, "grad_norm": 2.421875, "learning_rate": 4.918890075034765e-05, "loss": 1.3068408966064453, "step": 31940 }, { "epoch": 2.722624627183639, "grad_norm": 2.984375, "learning_rate": 4.918820329159122e-05, "loss": 1.2593417167663574, "step": 31950 }, { "epoch": 2.7234767788666385, "grad_norm": 2.1875, "learning_rate": 4.91875055380426e-05, "loss": 1.245619773864746, "step": 31960 }, { "epoch": 2.724328930549638, "grad_norm": 2.421875, "learning_rate": 4.918680748971028e-05, "loss": 1.216995906829834, "step": 31970 }, { "epoch": 2.7251810822326377, "grad_norm": 1.75, "learning_rate": 4.91861091466028e-05, "loss": 1.2257073402404786, "step": 31980 }, { "epoch": 2.7260332339156372, "grad_norm": 2.171875, "learning_rate": 4.9185410508728644e-05, "loss": 1.468941593170166, "step": 31990 }, { "epoch": 2.726885385598637, "grad_norm": 1.9921875, "learning_rate": 4.918471157609635e-05, "loss": 1.4812886238098144, "step": 32000 }, { "epoch": 2.7277375372816364, "grad_norm": 2.59375, "learning_rate": 4.918401234871441e-05, "loss": 1.2642638206481933, "step": 32010 }, { "epoch": 2.728589688964636, "grad_norm": 4.4375, "learning_rate": 4.9183312826591366e-05, "loss": 1.289449691772461, "step": 32020 }, { "epoch": 2.7294418406476355, "grad_norm": 3.921875, "learning_rate": 4.918261300973574e-05, "loss": 1.2962114334106445, "step": 32030 }, { "epoch": 2.730293992330635, "grad_norm": 2.015625, "learning_rate": 4.9181912898156045e-05, "loss": 1.3649958610534667, "step": 32040 }, { "epoch": 2.7311461440136346, "grad_norm": 1.8515625, "learning_rate": 4.918121249186083e-05, "loss": 1.319959545135498, "step": 32050 }, { "epoch": 2.731998295696634, "grad_norm": 3.03125, "learning_rate": 4.918051179085863e-05, "loss": 1.2991619110107422, "step": 32060 }, { "epoch": 2.732850447379634, "grad_norm": 2.515625, "learning_rate": 4.917981079515798e-05, "loss": 1.3411022186279298, "step": 32070 }, { "epoch": 2.7337025990626334, "grad_norm": 2.34375, "learning_rate": 4.917910950476743e-05, "loss": 1.3806987762451173, "step": 32080 }, { "epoch": 2.734554750745633, "grad_norm": 2.015625, "learning_rate": 4.9178407919695515e-05, "loss": 1.290991973876953, "step": 32090 }, { "epoch": 2.7354069024286325, "grad_norm": 2.625, "learning_rate": 4.91777060399508e-05, "loss": 1.2524175643920898, "step": 32100 }, { "epoch": 2.736259054111632, "grad_norm": 2.265625, "learning_rate": 4.9177003865541825e-05, "loss": 1.4747204780578613, "step": 32110 }, { "epoch": 2.7371112057946316, "grad_norm": 4.25, "learning_rate": 4.917630139647715e-05, "loss": 1.4094782829284669, "step": 32120 }, { "epoch": 2.737963357477631, "grad_norm": 2.09375, "learning_rate": 4.917559863276535e-05, "loss": 1.4665337562561036, "step": 32130 }, { "epoch": 2.7388155091606308, "grad_norm": 1.9453125, "learning_rate": 4.917489557441498e-05, "loss": 1.392624855041504, "step": 32140 }, { "epoch": 2.7396676608436303, "grad_norm": 2.890625, "learning_rate": 4.917419222143461e-05, "loss": 1.216737937927246, "step": 32150 }, { "epoch": 2.74051981252663, "grad_norm": 2.046875, "learning_rate": 4.9173488573832804e-05, "loss": 1.332326602935791, "step": 32160 }, { "epoch": 2.7413719642096295, "grad_norm": 3.453125, "learning_rate": 4.917278463161815e-05, "loss": 1.3574714660644531, "step": 32170 }, { "epoch": 2.742224115892629, "grad_norm": 2.203125, "learning_rate": 4.9172080394799226e-05, "loss": 1.3094283103942872, "step": 32180 }, { "epoch": 2.7430762675756286, "grad_norm": 2.09375, "learning_rate": 4.9171375863384606e-05, "loss": 1.4393661499023438, "step": 32190 }, { "epoch": 2.743928419258628, "grad_norm": 2.203125, "learning_rate": 4.917067103738288e-05, "loss": 1.4511549949645997, "step": 32200 }, { "epoch": 2.7447805709416278, "grad_norm": 2.78125, "learning_rate": 4.916996591680264e-05, "loss": 1.4613842010498046, "step": 32210 }, { "epoch": 2.7456327226246273, "grad_norm": 3.578125, "learning_rate": 4.916926050165248e-05, "loss": 1.3598475456237793, "step": 32220 }, { "epoch": 2.746484874307627, "grad_norm": 3.40625, "learning_rate": 4.916855479194099e-05, "loss": 1.2928400993347169, "step": 32230 }, { "epoch": 2.7473370259906265, "grad_norm": 2.265625, "learning_rate": 4.9167848787676787e-05, "loss": 1.4235535621643067, "step": 32240 }, { "epoch": 2.748189177673626, "grad_norm": 1.859375, "learning_rate": 4.9167142488868454e-05, "loss": 1.3930653572082519, "step": 32250 }, { "epoch": 2.7490413293566256, "grad_norm": 3.296875, "learning_rate": 4.916643589552462e-05, "loss": 1.3665500640869142, "step": 32260 }, { "epoch": 2.749893481039625, "grad_norm": 2.15625, "learning_rate": 4.916572900765388e-05, "loss": 1.3792094230651855, "step": 32270 }, { "epoch": 2.7507456327226247, "grad_norm": 1.8515625, "learning_rate": 4.916502182526487e-05, "loss": 1.373738956451416, "step": 32280 }, { "epoch": 2.7515977844056243, "grad_norm": 4.0, "learning_rate": 4.9164314348366184e-05, "loss": 1.3591911315917968, "step": 32290 }, { "epoch": 2.752449936088624, "grad_norm": 1.875, "learning_rate": 4.916360657696645e-05, "loss": 1.3390961647033692, "step": 32300 }, { "epoch": 2.7533020877716234, "grad_norm": 2.125, "learning_rate": 4.916289851107431e-05, "loss": 1.3489108085632324, "step": 32310 }, { "epoch": 2.754154239454623, "grad_norm": 2.171875, "learning_rate": 4.916219015069837e-05, "loss": 1.3969637870788574, "step": 32320 }, { "epoch": 2.7550063911376226, "grad_norm": 2.625, "learning_rate": 4.916148149584728e-05, "loss": 1.1703693389892578, "step": 32330 }, { "epoch": 2.755858542820622, "grad_norm": 2.0, "learning_rate": 4.9160772546529675e-05, "loss": 1.3687217712402344, "step": 32340 }, { "epoch": 2.7567106945036217, "grad_norm": 2.953125, "learning_rate": 4.91600633027542e-05, "loss": 1.2853832244873047, "step": 32350 }, { "epoch": 2.7575628461866213, "grad_norm": 1.828125, "learning_rate": 4.915935376452948e-05, "loss": 1.3733098983764649, "step": 32360 }, { "epoch": 2.758414997869621, "grad_norm": 2.296875, "learning_rate": 4.9158643931864176e-05, "loss": 1.3069954872131349, "step": 32370 }, { "epoch": 2.7592671495526204, "grad_norm": 1.9296875, "learning_rate": 4.915793380476694e-05, "loss": 1.4107579231262206, "step": 32380 }, { "epoch": 2.76011930123562, "grad_norm": 2.6875, "learning_rate": 4.915722338324642e-05, "loss": 1.4122852325439452, "step": 32390 }, { "epoch": 2.7609714529186196, "grad_norm": 1.78125, "learning_rate": 4.915651266731128e-05, "loss": 1.333594512939453, "step": 32400 }, { "epoch": 2.761823604601619, "grad_norm": 2.0, "learning_rate": 4.915580165697018e-05, "loss": 1.4822110176086425, "step": 32410 }, { "epoch": 2.7626757562846187, "grad_norm": 2.203125, "learning_rate": 4.915509035223179e-05, "loss": 1.079058837890625, "step": 32420 }, { "epoch": 2.7635279079676183, "grad_norm": 3.09375, "learning_rate": 4.915437875310477e-05, "loss": 1.3463369369506837, "step": 32430 }, { "epoch": 2.764380059650618, "grad_norm": 2.34375, "learning_rate": 4.915366685959779e-05, "loss": 1.2804617881774902, "step": 32440 }, { "epoch": 2.7652322113336174, "grad_norm": 2.03125, "learning_rate": 4.915295467171954e-05, "loss": 1.3897995948791504, "step": 32450 }, { "epoch": 2.766084363016617, "grad_norm": 2.1875, "learning_rate": 4.915224218947869e-05, "loss": 1.4060752868652344, "step": 32460 }, { "epoch": 2.7669365146996165, "grad_norm": 2.25, "learning_rate": 4.9151529412883924e-05, "loss": 1.2736166000366211, "step": 32470 }, { "epoch": 2.767788666382616, "grad_norm": 2.421875, "learning_rate": 4.915081634194393e-05, "loss": 1.4346360206604003, "step": 32480 }, { "epoch": 2.7686408180656157, "grad_norm": 2.703125, "learning_rate": 4.91501029766674e-05, "loss": 1.3220163345336915, "step": 32490 }, { "epoch": 2.7694929697486153, "grad_norm": 2.53125, "learning_rate": 4.914938931706303e-05, "loss": 1.4026120185852051, "step": 32500 }, { "epoch": 2.770345121431615, "grad_norm": 1.8515625, "learning_rate": 4.9148675363139515e-05, "loss": 1.3978595733642578, "step": 32510 }, { "epoch": 2.7711972731146144, "grad_norm": 2.90625, "learning_rate": 4.914796111490555e-05, "loss": 1.2756563186645509, "step": 32520 }, { "epoch": 2.772049424797614, "grad_norm": 2.6875, "learning_rate": 4.914724657236985e-05, "loss": 1.1576475143432616, "step": 32530 }, { "epoch": 2.7729015764806135, "grad_norm": 2.734375, "learning_rate": 4.9146531735541124e-05, "loss": 1.4330148696899414, "step": 32540 }, { "epoch": 2.773753728163613, "grad_norm": 3.21875, "learning_rate": 4.914581660442808e-05, "loss": 1.3222329139709472, "step": 32550 }, { "epoch": 2.7746058798466127, "grad_norm": 2.46875, "learning_rate": 4.914510117903943e-05, "loss": 1.3018990516662599, "step": 32560 }, { "epoch": 2.7754580315296122, "grad_norm": 2.25, "learning_rate": 4.914438545938389e-05, "loss": 1.3307975769042968, "step": 32570 }, { "epoch": 2.776310183212612, "grad_norm": 2.6875, "learning_rate": 4.9143669445470196e-05, "loss": 1.3379572868347167, "step": 32580 }, { "epoch": 2.7771623348956114, "grad_norm": 2.15625, "learning_rate": 4.914295313730707e-05, "loss": 1.4238090515136719, "step": 32590 }, { "epoch": 2.778014486578611, "grad_norm": 2.890625, "learning_rate": 4.914223653490323e-05, "loss": 1.369877243041992, "step": 32600 }, { "epoch": 2.7788666382616105, "grad_norm": 2.375, "learning_rate": 4.914151963826743e-05, "loss": 1.4328100204467773, "step": 32610 }, { "epoch": 2.77971878994461, "grad_norm": 2.21875, "learning_rate": 4.914080244740839e-05, "loss": 1.1888471603393556, "step": 32620 }, { "epoch": 2.7805709416276096, "grad_norm": 2.796875, "learning_rate": 4.914008496233486e-05, "loss": 1.4750709533691406, "step": 32630 }, { "epoch": 2.781423093310609, "grad_norm": 2.078125, "learning_rate": 4.913936718305557e-05, "loss": 1.3345293045043944, "step": 32640 }, { "epoch": 2.782275244993609, "grad_norm": 2.71875, "learning_rate": 4.9138649109579295e-05, "loss": 1.4587515830993651, "step": 32650 }, { "epoch": 2.7831273966766084, "grad_norm": 2.71875, "learning_rate": 4.913793074191476e-05, "loss": 1.341120433807373, "step": 32660 }, { "epoch": 2.783979548359608, "grad_norm": 1.8125, "learning_rate": 4.913721208007073e-05, "loss": 1.3066389083862304, "step": 32670 }, { "epoch": 2.7848317000426075, "grad_norm": 2.015625, "learning_rate": 4.913649312405597e-05, "loss": 1.262465000152588, "step": 32680 }, { "epoch": 2.785683851725607, "grad_norm": 1.6171875, "learning_rate": 4.9135773873879234e-05, "loss": 1.2985048294067383, "step": 32690 }, { "epoch": 2.7865360034086066, "grad_norm": 1.921875, "learning_rate": 4.913505432954929e-05, "loss": 1.304604434967041, "step": 32700 }, { "epoch": 2.787388155091606, "grad_norm": 2.140625, "learning_rate": 4.913433449107491e-05, "loss": 1.3872934341430665, "step": 32710 }, { "epoch": 2.7882403067746058, "grad_norm": 1.828125, "learning_rate": 4.913361435846486e-05, "loss": 1.3041987419128418, "step": 32720 }, { "epoch": 2.7890924584576053, "grad_norm": 2.15625, "learning_rate": 4.913289393172793e-05, "loss": 1.265220546722412, "step": 32730 }, { "epoch": 2.789944610140605, "grad_norm": 2.5, "learning_rate": 4.913217321087288e-05, "loss": 1.3304608345031739, "step": 32740 }, { "epoch": 2.7907967618236045, "grad_norm": 2.4375, "learning_rate": 4.913145219590851e-05, "loss": 1.3807100296020507, "step": 32750 }, { "epoch": 2.791648913506604, "grad_norm": 3.09375, "learning_rate": 4.913073088684361e-05, "loss": 1.395268726348877, "step": 32760 }, { "epoch": 2.7925010651896036, "grad_norm": 1.53125, "learning_rate": 4.913000928368695e-05, "loss": 1.4276276588439942, "step": 32770 }, { "epoch": 2.793353216872603, "grad_norm": 2.671875, "learning_rate": 4.9129287386447346e-05, "loss": 1.4594382286071776, "step": 32780 }, { "epoch": 2.7942053685556028, "grad_norm": 2.171875, "learning_rate": 4.912856519513359e-05, "loss": 1.3052129745483398, "step": 32790 }, { "epoch": 2.7950575202386023, "grad_norm": 1.9921875, "learning_rate": 4.9127842709754476e-05, "loss": 1.410041618347168, "step": 32800 }, { "epoch": 2.795909671921602, "grad_norm": 2.046875, "learning_rate": 4.912711993031882e-05, "loss": 1.459219741821289, "step": 32810 }, { "epoch": 2.7967618236046015, "grad_norm": 3.8125, "learning_rate": 4.912639685683542e-05, "loss": 1.2860310554504395, "step": 32820 }, { "epoch": 2.797613975287601, "grad_norm": 2.125, "learning_rate": 4.9125673489313094e-05, "loss": 1.337111473083496, "step": 32830 }, { "epoch": 2.7984661269706006, "grad_norm": 4.09375, "learning_rate": 4.9124949827760666e-05, "loss": 1.2986148834228515, "step": 32840 }, { "epoch": 2.7993182786536, "grad_norm": 2.140625, "learning_rate": 4.9124225872186934e-05, "loss": 1.4516451835632325, "step": 32850 }, { "epoch": 2.8001704303365997, "grad_norm": 1.9140625, "learning_rate": 4.912350162260075e-05, "loss": 1.359345054626465, "step": 32860 }, { "epoch": 2.8010225820195993, "grad_norm": 2.78125, "learning_rate": 4.9122777079010915e-05, "loss": 1.5241430282592774, "step": 32870 }, { "epoch": 2.801874733702599, "grad_norm": 2.046875, "learning_rate": 4.9122052241426274e-05, "loss": 1.398446273803711, "step": 32880 }, { "epoch": 2.8027268853855984, "grad_norm": 2.125, "learning_rate": 4.912132710985566e-05, "loss": 1.345514678955078, "step": 32890 }, { "epoch": 2.803579037068598, "grad_norm": 2.3125, "learning_rate": 4.91206016843079e-05, "loss": 1.122532558441162, "step": 32900 }, { "epoch": 2.8044311887515976, "grad_norm": 2.421875, "learning_rate": 4.911987596479185e-05, "loss": 1.38583402633667, "step": 32910 }, { "epoch": 2.805283340434597, "grad_norm": 2.640625, "learning_rate": 4.911914995131634e-05, "loss": 1.410508918762207, "step": 32920 }, { "epoch": 2.8061354921175967, "grad_norm": 2.921875, "learning_rate": 4.911842364389023e-05, "loss": 1.254329299926758, "step": 32930 }, { "epoch": 2.8069876438005963, "grad_norm": 2.109375, "learning_rate": 4.9117697042522366e-05, "loss": 1.3999207496643067, "step": 32940 }, { "epoch": 2.807839795483596, "grad_norm": 1.6171875, "learning_rate": 4.91169701472216e-05, "loss": 1.4167022705078125, "step": 32950 }, { "epoch": 2.8086919471665954, "grad_norm": 2.828125, "learning_rate": 4.911624295799681e-05, "loss": 1.3479911804199218, "step": 32960 }, { "epoch": 2.809544098849595, "grad_norm": 3.421875, "learning_rate": 4.911551547485682e-05, "loss": 1.4419830322265625, "step": 32970 }, { "epoch": 2.8103962505325946, "grad_norm": 1.78125, "learning_rate": 4.911478769781054e-05, "loss": 1.3678318023681642, "step": 32980 }, { "epoch": 2.811248402215594, "grad_norm": 2.265625, "learning_rate": 4.9114059626866817e-05, "loss": 1.2812980651855468, "step": 32990 }, { "epoch": 2.8121005538985937, "grad_norm": 2.328125, "learning_rate": 4.911333126203452e-05, "loss": 1.3440446853637695, "step": 33000 }, { "epoch": 2.8129527055815933, "grad_norm": 3.59375, "learning_rate": 4.9112602603322534e-05, "loss": 1.2573692321777343, "step": 33010 }, { "epoch": 2.813804857264593, "grad_norm": 1.4609375, "learning_rate": 4.911187365073975e-05, "loss": 1.3444406509399414, "step": 33020 }, { "epoch": 2.8146570089475924, "grad_norm": 2.546875, "learning_rate": 4.911114440429503e-05, "loss": 1.3636661529541017, "step": 33030 }, { "epoch": 2.815509160630592, "grad_norm": 2.296875, "learning_rate": 4.911041486399727e-05, "loss": 1.3856246948242188, "step": 33040 }, { "epoch": 2.8163613123135915, "grad_norm": 3.375, "learning_rate": 4.910968502985537e-05, "loss": 1.34727201461792, "step": 33050 }, { "epoch": 2.817213463996591, "grad_norm": 2.59375, "learning_rate": 4.9108954901878216e-05, "loss": 1.3934877395629883, "step": 33060 }, { "epoch": 2.818065615679591, "grad_norm": 1.78125, "learning_rate": 4.910822448007471e-05, "loss": 1.4011693000793457, "step": 33070 }, { "epoch": 2.8189177673625907, "grad_norm": 2.078125, "learning_rate": 4.910749376445375e-05, "loss": 1.2466876029968261, "step": 33080 }, { "epoch": 2.8197699190455903, "grad_norm": 2.109375, "learning_rate": 4.910676275502425e-05, "loss": 1.3113292694091796, "step": 33090 }, { "epoch": 2.82062207072859, "grad_norm": 2.203125, "learning_rate": 4.910603145179511e-05, "loss": 1.4665118217468263, "step": 33100 }, { "epoch": 2.8214742224115894, "grad_norm": 1.875, "learning_rate": 4.9105299854775245e-05, "loss": 1.378179931640625, "step": 33110 }, { "epoch": 2.822326374094589, "grad_norm": 3.046875, "learning_rate": 4.910456796397358e-05, "loss": 1.4052385330200194, "step": 33120 }, { "epoch": 2.8231785257775885, "grad_norm": 2.015625, "learning_rate": 4.910383577939902e-05, "loss": 1.3601069450378418, "step": 33130 }, { "epoch": 2.824030677460588, "grad_norm": 2.09375, "learning_rate": 4.9103103301060496e-05, "loss": 1.4502280235290528, "step": 33140 }, { "epoch": 2.8248828291435877, "grad_norm": 2.3125, "learning_rate": 4.9102370528966936e-05, "loss": 1.284475326538086, "step": 33150 }, { "epoch": 2.8257349808265873, "grad_norm": 1.890625, "learning_rate": 4.9101637463127263e-05, "loss": 1.3430551528930663, "step": 33160 }, { "epoch": 2.826587132509587, "grad_norm": 3.75, "learning_rate": 4.910090410355043e-05, "loss": 1.2529776573181153, "step": 33170 }, { "epoch": 2.8274392841925864, "grad_norm": 3.359375, "learning_rate": 4.910017045024535e-05, "loss": 1.2052425384521483, "step": 33180 }, { "epoch": 2.828291435875586, "grad_norm": 2.828125, "learning_rate": 4.9099436503220984e-05, "loss": 1.2508377075195312, "step": 33190 }, { "epoch": 2.8291435875585855, "grad_norm": 3.296875, "learning_rate": 4.909870226248627e-05, "loss": 1.3000747680664062, "step": 33200 }, { "epoch": 2.829995739241585, "grad_norm": 2.953125, "learning_rate": 4.9097967728050154e-05, "loss": 1.4623583793640136, "step": 33210 }, { "epoch": 2.8308478909245847, "grad_norm": 2.296875, "learning_rate": 4.909723289992159e-05, "loss": 1.4486165046691895, "step": 33220 }, { "epoch": 2.8317000426075842, "grad_norm": 2.25, "learning_rate": 4.9096497778109536e-05, "loss": 1.3892462730407715, "step": 33230 }, { "epoch": 2.832552194290584, "grad_norm": 2.109375, "learning_rate": 4.909576236262295e-05, "loss": 1.3603233337402343, "step": 33240 }, { "epoch": 2.8334043459735834, "grad_norm": 2.5625, "learning_rate": 4.909502665347079e-05, "loss": 1.4353336334228515, "step": 33250 }, { "epoch": 2.834256497656583, "grad_norm": 3.21875, "learning_rate": 4.909429065066203e-05, "loss": 1.2136691093444825, "step": 33260 }, { "epoch": 2.8351086493395825, "grad_norm": 2.34375, "learning_rate": 4.909355435420563e-05, "loss": 1.5672505378723145, "step": 33270 }, { "epoch": 2.835960801022582, "grad_norm": 2.296875, "learning_rate": 4.9092817764110575e-05, "loss": 1.349999237060547, "step": 33280 }, { "epoch": 2.8368129527055816, "grad_norm": 3.15625, "learning_rate": 4.909208088038584e-05, "loss": 1.362513828277588, "step": 33290 }, { "epoch": 2.837665104388581, "grad_norm": 2.078125, "learning_rate": 4.90913437030404e-05, "loss": 1.3869794845581054, "step": 33300 }, { "epoch": 2.838517256071581, "grad_norm": 3.484375, "learning_rate": 4.909060623208325e-05, "loss": 1.2464408874511719, "step": 33310 }, { "epoch": 2.8393694077545804, "grad_norm": 2.5, "learning_rate": 4.908986846752336e-05, "loss": 1.2381850242614747, "step": 33320 }, { "epoch": 2.84022155943758, "grad_norm": 2.484375, "learning_rate": 4.9089130409369735e-05, "loss": 1.340367603302002, "step": 33330 }, { "epoch": 2.8410737111205795, "grad_norm": 2.640625, "learning_rate": 4.9088392057631363e-05, "loss": 1.3633519172668458, "step": 33340 }, { "epoch": 2.841925862803579, "grad_norm": 2.0, "learning_rate": 4.908765341231725e-05, "loss": 1.4037922859191894, "step": 33350 }, { "epoch": 2.8427780144865786, "grad_norm": 2.53125, "learning_rate": 4.9086914473436385e-05, "loss": 1.3487966537475586, "step": 33360 }, { "epoch": 2.843630166169578, "grad_norm": 2.046875, "learning_rate": 4.908617524099779e-05, "loss": 1.4290482521057128, "step": 33370 }, { "epoch": 2.8444823178525778, "grad_norm": 2.203125, "learning_rate": 4.908543571501047e-05, "loss": 1.3171704292297364, "step": 33380 }, { "epoch": 2.8453344695355773, "grad_norm": 2.609375, "learning_rate": 4.9084695895483436e-05, "loss": 1.3371171951293945, "step": 33390 }, { "epoch": 2.846186621218577, "grad_norm": 2.046875, "learning_rate": 4.90839557824257e-05, "loss": 1.343869972229004, "step": 33400 }, { "epoch": 2.8470387729015765, "grad_norm": 2.4375, "learning_rate": 4.9083215375846285e-05, "loss": 1.3384055137634276, "step": 33410 }, { "epoch": 2.847890924584576, "grad_norm": 2.59375, "learning_rate": 4.908247467575422e-05, "loss": 1.402552890777588, "step": 33420 }, { "epoch": 2.8487430762675756, "grad_norm": 2.21875, "learning_rate": 4.9081733682158524e-05, "loss": 1.4412296295166016, "step": 33430 }, { "epoch": 2.849595227950575, "grad_norm": 3.640625, "learning_rate": 4.908099239506824e-05, "loss": 1.2085922241210938, "step": 33440 }, { "epoch": 2.8504473796335748, "grad_norm": 3.546875, "learning_rate": 4.9080250814492376e-05, "loss": 1.2694511413574219, "step": 33450 }, { "epoch": 2.8512995313165743, "grad_norm": 2.921875, "learning_rate": 4.9079508940440005e-05, "loss": 1.1386459350585938, "step": 33460 }, { "epoch": 2.852151682999574, "grad_norm": 3.46875, "learning_rate": 4.907876677292015e-05, "loss": 1.460175132751465, "step": 33470 }, { "epoch": 2.8530038346825735, "grad_norm": 1.5859375, "learning_rate": 4.907802431194185e-05, "loss": 1.4179560661315918, "step": 33480 }, { "epoch": 2.853855986365573, "grad_norm": 4.25, "learning_rate": 4.9077281557514165e-05, "loss": 1.406212043762207, "step": 33490 }, { "epoch": 2.8547081380485726, "grad_norm": 3.296875, "learning_rate": 4.9076538509646145e-05, "loss": 1.453967571258545, "step": 33500 }, { "epoch": 2.855560289731572, "grad_norm": 2.09375, "learning_rate": 4.9075795168346844e-05, "loss": 1.4471147537231446, "step": 33510 }, { "epoch": 2.8564124414145717, "grad_norm": 2.703125, "learning_rate": 4.907505153362532e-05, "loss": 1.2968905448913575, "step": 33520 }, { "epoch": 2.8572645930975713, "grad_norm": 1.7109375, "learning_rate": 4.9074307605490645e-05, "loss": 1.4941787719726562, "step": 33530 }, { "epoch": 2.858116744780571, "grad_norm": 1.9296875, "learning_rate": 4.907356338395187e-05, "loss": 1.3363382339477539, "step": 33540 }, { "epoch": 2.8589688964635704, "grad_norm": 2.015625, "learning_rate": 4.907281886901808e-05, "loss": 1.369710636138916, "step": 33550 }, { "epoch": 2.85982104814657, "grad_norm": 3.8125, "learning_rate": 4.9072074060698336e-05, "loss": 1.2741218566894532, "step": 33560 }, { "epoch": 2.8606731998295696, "grad_norm": 2.125, "learning_rate": 4.907132895900172e-05, "loss": 1.3626516342163086, "step": 33570 }, { "epoch": 2.861525351512569, "grad_norm": 2.546875, "learning_rate": 4.907058356393732e-05, "loss": 1.4120855331420898, "step": 33580 }, { "epoch": 2.8623775031955687, "grad_norm": 2.203125, "learning_rate": 4.906983787551421e-05, "loss": 1.3656208038330078, "step": 33590 }, { "epoch": 2.8632296548785683, "grad_norm": 2.890625, "learning_rate": 4.906909189374149e-05, "loss": 1.3401921272277832, "step": 33600 }, { "epoch": 2.864081806561568, "grad_norm": 3.71875, "learning_rate": 4.9068345618628245e-05, "loss": 1.2158711433410645, "step": 33610 }, { "epoch": 2.8649339582445674, "grad_norm": 2.59375, "learning_rate": 4.9067599050183566e-05, "loss": 1.4337326049804688, "step": 33620 }, { "epoch": 2.865786109927567, "grad_norm": 2.765625, "learning_rate": 4.9066852188416555e-05, "loss": 1.2204822540283202, "step": 33630 }, { "epoch": 2.8666382616105666, "grad_norm": 1.7265625, "learning_rate": 4.906610503333632e-05, "loss": 1.3338014602661132, "step": 33640 }, { "epoch": 2.867490413293566, "grad_norm": 3.046875, "learning_rate": 4.9065357584951956e-05, "loss": 1.296278953552246, "step": 33650 }, { "epoch": 2.8683425649765657, "grad_norm": 1.8515625, "learning_rate": 4.906460984327258e-05, "loss": 1.37510986328125, "step": 33660 }, { "epoch": 2.8691947166595653, "grad_norm": 2.609375, "learning_rate": 4.906386180830731e-05, "loss": 1.3030195236206055, "step": 33670 }, { "epoch": 2.870046868342565, "grad_norm": 2.296875, "learning_rate": 4.906311348006525e-05, "loss": 1.4424468994140625, "step": 33680 }, { "epoch": 2.8708990200255644, "grad_norm": 2.203125, "learning_rate": 4.9062364858555524e-05, "loss": 1.371271800994873, "step": 33690 }, { "epoch": 2.871751171708564, "grad_norm": 2.109375, "learning_rate": 4.9061615943787257e-05, "loss": 1.2803304672241211, "step": 33700 }, { "epoch": 2.8726033233915635, "grad_norm": 2.0, "learning_rate": 4.906086673576958e-05, "loss": 1.4156993865966796, "step": 33710 }, { "epoch": 2.873455475074563, "grad_norm": 2.6875, "learning_rate": 4.906011723451162e-05, "loss": 1.4752874374389648, "step": 33720 }, { "epoch": 2.8743076267575627, "grad_norm": 1.9296875, "learning_rate": 4.905936744002252e-05, "loss": 1.1580530166625977, "step": 33730 }, { "epoch": 2.8751597784405627, "grad_norm": 2.140625, "learning_rate": 4.9058617352311406e-05, "loss": 1.3409666061401366, "step": 33740 }, { "epoch": 2.8760119301235623, "grad_norm": 3.125, "learning_rate": 4.905786697138742e-05, "loss": 1.2823073387145996, "step": 33750 }, { "epoch": 2.876864081806562, "grad_norm": 2.359375, "learning_rate": 4.905711629725972e-05, "loss": 1.4207749366760254, "step": 33760 }, { "epoch": 2.8777162334895614, "grad_norm": 1.90625, "learning_rate": 4.905636532993744e-05, "loss": 1.2728174209594727, "step": 33770 }, { "epoch": 2.878568385172561, "grad_norm": 3.375, "learning_rate": 4.905561406942974e-05, "loss": 1.401140022277832, "step": 33780 }, { "epoch": 2.8794205368555605, "grad_norm": 2.125, "learning_rate": 4.905486251574577e-05, "loss": 1.3716564178466797, "step": 33790 }, { "epoch": 2.88027268853856, "grad_norm": 2.15625, "learning_rate": 4.9054110668894704e-05, "loss": 1.3517292022705079, "step": 33800 }, { "epoch": 2.8811248402215597, "grad_norm": 2.828125, "learning_rate": 4.905335852888569e-05, "loss": 1.2644742965698241, "step": 33810 }, { "epoch": 2.8819769919045592, "grad_norm": 2.75, "learning_rate": 4.9052606095727905e-05, "loss": 1.2315977096557618, "step": 33820 }, { "epoch": 2.882829143587559, "grad_norm": 2.46875, "learning_rate": 4.905185336943051e-05, "loss": 1.4015048027038575, "step": 33830 }, { "epoch": 2.8836812952705584, "grad_norm": 2.359375, "learning_rate": 4.905110035000268e-05, "loss": 1.4896263122558593, "step": 33840 }, { "epoch": 2.884533446953558, "grad_norm": 2.296875, "learning_rate": 4.9050347037453604e-05, "loss": 1.2405126571655274, "step": 33850 }, { "epoch": 2.8853855986365575, "grad_norm": 1.765625, "learning_rate": 4.9049593431792446e-05, "loss": 1.3546623229980468, "step": 33860 }, { "epoch": 2.886237750319557, "grad_norm": 2.9375, "learning_rate": 4.90488395330284e-05, "loss": 1.2991091728210449, "step": 33870 }, { "epoch": 2.8870899020025567, "grad_norm": 1.90625, "learning_rate": 4.9048085341170656e-05, "loss": 1.3089344024658203, "step": 33880 }, { "epoch": 2.8879420536855562, "grad_norm": 2.421875, "learning_rate": 4.90473308562284e-05, "loss": 1.3975313186645508, "step": 33890 }, { "epoch": 2.888794205368556, "grad_norm": 2.28125, "learning_rate": 4.9046576078210835e-05, "loss": 1.1724771499633788, "step": 33900 }, { "epoch": 2.8896463570515554, "grad_norm": 3.25, "learning_rate": 4.9045821007127153e-05, "loss": 1.3379880905151367, "step": 33910 }, { "epoch": 2.890498508734555, "grad_norm": 1.9140625, "learning_rate": 4.904506564298655e-05, "loss": 1.4229894638061524, "step": 33920 }, { "epoch": 2.8913506604175545, "grad_norm": 4.09375, "learning_rate": 4.9044309985798254e-05, "loss": 1.199350070953369, "step": 33930 }, { "epoch": 2.892202812100554, "grad_norm": 2.59375, "learning_rate": 4.904355403557145e-05, "loss": 1.3446762084960937, "step": 33940 }, { "epoch": 2.8930549637835536, "grad_norm": 1.890625, "learning_rate": 4.904279779231537e-05, "loss": 1.5146955490112304, "step": 33950 }, { "epoch": 2.893907115466553, "grad_norm": 2.890625, "learning_rate": 4.9042041256039216e-05, "loss": 1.4152043342590332, "step": 33960 }, { "epoch": 2.894759267149553, "grad_norm": 2.71875, "learning_rate": 4.904128442675222e-05, "loss": 1.3015361785888673, "step": 33970 }, { "epoch": 2.8956114188325524, "grad_norm": 2.125, "learning_rate": 4.9040527304463594e-05, "loss": 1.5000121116638183, "step": 33980 }, { "epoch": 2.896463570515552, "grad_norm": 1.9453125, "learning_rate": 4.903976988918257e-05, "loss": 1.3803759574890138, "step": 33990 }, { "epoch": 2.8973157221985515, "grad_norm": 2.515625, "learning_rate": 4.9039012180918386e-05, "loss": 1.3156041145324706, "step": 34000 }, { "epoch": 2.898167873881551, "grad_norm": 2.90625, "learning_rate": 4.903825417968027e-05, "loss": 1.2207575798034669, "step": 34010 }, { "epoch": 2.8990200255645506, "grad_norm": 3.453125, "learning_rate": 4.903749588547746e-05, "loss": 1.3249270439147949, "step": 34020 }, { "epoch": 2.89987217724755, "grad_norm": 2.390625, "learning_rate": 4.90367372983192e-05, "loss": 1.3607686042785645, "step": 34030 }, { "epoch": 2.9007243289305498, "grad_norm": 3.140625, "learning_rate": 4.903597841821474e-05, "loss": 1.2973118782043458, "step": 34040 }, { "epoch": 2.9015764806135493, "grad_norm": 3.0625, "learning_rate": 4.9035219245173314e-05, "loss": 1.2944884300231934, "step": 34050 }, { "epoch": 2.902428632296549, "grad_norm": 2.515625, "learning_rate": 4.903445977920419e-05, "loss": 1.338409423828125, "step": 34060 }, { "epoch": 2.9032807839795485, "grad_norm": 2.59375, "learning_rate": 4.903370002031661e-05, "loss": 1.4689344406127929, "step": 34070 }, { "epoch": 2.904132935662548, "grad_norm": 1.953125, "learning_rate": 4.903293996851985e-05, "loss": 1.2271466255187988, "step": 34080 }, { "epoch": 2.9049850873455476, "grad_norm": 2.1875, "learning_rate": 4.903217962382316e-05, "loss": 1.349105167388916, "step": 34090 }, { "epoch": 2.905837239028547, "grad_norm": 4.0625, "learning_rate": 4.903141898623581e-05, "loss": 1.2795624732971191, "step": 34100 }, { "epoch": 2.9066893907115467, "grad_norm": 3.4375, "learning_rate": 4.9030658055767074e-05, "loss": 1.393735122680664, "step": 34110 }, { "epoch": 2.9075415423945463, "grad_norm": 2.421875, "learning_rate": 4.9029896832426215e-05, "loss": 1.3380143165588378, "step": 34120 }, { "epoch": 2.908393694077546, "grad_norm": 2.015625, "learning_rate": 4.902913531622252e-05, "loss": 1.339393424987793, "step": 34130 }, { "epoch": 2.9092458457605455, "grad_norm": 2.375, "learning_rate": 4.902837350716527e-05, "loss": 1.301722526550293, "step": 34140 }, { "epoch": 2.910097997443545, "grad_norm": 2.359375, "learning_rate": 4.902761140526375e-05, "loss": 1.2832578659057616, "step": 34150 }, { "epoch": 2.9109501491265446, "grad_norm": 2.265625, "learning_rate": 4.902684901052724e-05, "loss": 1.4798905372619628, "step": 34160 }, { "epoch": 2.911802300809544, "grad_norm": 3.828125, "learning_rate": 4.9026086322965047e-05, "loss": 1.2674077987670898, "step": 34170 }, { "epoch": 2.9126544524925437, "grad_norm": 2.59375, "learning_rate": 4.902532334258645e-05, "loss": 1.3251222610473632, "step": 34180 }, { "epoch": 2.9135066041755433, "grad_norm": 2.078125, "learning_rate": 4.902456006940075e-05, "loss": 1.5387563705444336, "step": 34190 }, { "epoch": 2.914358755858543, "grad_norm": 2.265625, "learning_rate": 4.902379650341726e-05, "loss": 1.2850091934204102, "step": 34200 }, { "epoch": 2.9152109075415424, "grad_norm": 2.90625, "learning_rate": 4.9023032644645276e-05, "loss": 1.4187431335449219, "step": 34210 }, { "epoch": 2.916063059224542, "grad_norm": 2.140625, "learning_rate": 4.902226849309412e-05, "loss": 1.4002447128295898, "step": 34220 }, { "epoch": 2.9169152109075416, "grad_norm": 2.265625, "learning_rate": 4.902150404877308e-05, "loss": 1.3747098922729493, "step": 34230 }, { "epoch": 2.917767362590541, "grad_norm": 2.40625, "learning_rate": 4.90207393116915e-05, "loss": 1.279073143005371, "step": 34240 }, { "epoch": 2.9186195142735407, "grad_norm": 2.015625, "learning_rate": 4.901997428185868e-05, "loss": 1.362839412689209, "step": 34250 }, { "epoch": 2.9194716659565403, "grad_norm": 2.234375, "learning_rate": 4.901920895928396e-05, "loss": 1.4509453773498535, "step": 34260 }, { "epoch": 2.92032381763954, "grad_norm": 2.8125, "learning_rate": 4.901844334397665e-05, "loss": 1.3598699569702148, "step": 34270 }, { "epoch": 2.9211759693225394, "grad_norm": 2.390625, "learning_rate": 4.90176774359461e-05, "loss": 1.3308717727661132, "step": 34280 }, { "epoch": 2.922028121005539, "grad_norm": 2.703125, "learning_rate": 4.9016911235201634e-05, "loss": 1.3238161087036133, "step": 34290 }, { "epoch": 2.9228802726885386, "grad_norm": 2.859375, "learning_rate": 4.9016144741752586e-05, "loss": 1.3338706970214844, "step": 34300 }, { "epoch": 2.923732424371538, "grad_norm": 2.609375, "learning_rate": 4.901537795560831e-05, "loss": 1.3081178665161133, "step": 34310 }, { "epoch": 2.9245845760545377, "grad_norm": 2.375, "learning_rate": 4.901461087677814e-05, "loss": 1.3372090339660645, "step": 34320 }, { "epoch": 2.9254367277375373, "grad_norm": 3.1875, "learning_rate": 4.901384350527143e-05, "loss": 1.4392790794372559, "step": 34330 }, { "epoch": 2.926288879420537, "grad_norm": 4.4375, "learning_rate": 4.901307584109753e-05, "loss": 1.246976089477539, "step": 34340 }, { "epoch": 2.9271410311035364, "grad_norm": 1.7265625, "learning_rate": 4.9012307884265796e-05, "loss": 1.4104803085327149, "step": 34350 }, { "epoch": 2.927993182786536, "grad_norm": 2.5, "learning_rate": 4.901153963478558e-05, "loss": 1.3554728507995606, "step": 34360 }, { "epoch": 2.9288453344695355, "grad_norm": 3.21875, "learning_rate": 4.901077109266627e-05, "loss": 1.392646026611328, "step": 34370 }, { "epoch": 2.929697486152535, "grad_norm": 3.84375, "learning_rate": 4.901000225791721e-05, "loss": 1.4666799545288085, "step": 34380 }, { "epoch": 2.9305496378355347, "grad_norm": 2.875, "learning_rate": 4.900923313054777e-05, "loss": 1.3394847869873048, "step": 34390 }, { "epoch": 2.9314017895185343, "grad_norm": 2.53125, "learning_rate": 4.900846371056733e-05, "loss": 1.4496477127075196, "step": 34400 }, { "epoch": 2.932253941201534, "grad_norm": 3.59375, "learning_rate": 4.900769399798527e-05, "loss": 1.3702838897705079, "step": 34410 }, { "epoch": 2.9331060928845334, "grad_norm": 2.484375, "learning_rate": 4.900692399281097e-05, "loss": 1.2630003929138183, "step": 34420 }, { "epoch": 2.933958244567533, "grad_norm": 2.359375, "learning_rate": 4.9006153695053806e-05, "loss": 1.2854890823364258, "step": 34430 }, { "epoch": 2.9348103962505325, "grad_norm": 2.78125, "learning_rate": 4.900538310472318e-05, "loss": 1.450594139099121, "step": 34440 }, { "epoch": 2.935662547933532, "grad_norm": 2.90625, "learning_rate": 4.900461222182847e-05, "loss": 1.2844212532043457, "step": 34450 }, { "epoch": 2.9365146996165317, "grad_norm": 2.4375, "learning_rate": 4.900384104637907e-05, "loss": 1.235696792602539, "step": 34460 }, { "epoch": 2.9373668512995312, "grad_norm": 3.125, "learning_rate": 4.90030695783844e-05, "loss": 1.397343349456787, "step": 34470 }, { "epoch": 2.938219002982531, "grad_norm": 3.015625, "learning_rate": 4.9002297817853836e-05, "loss": 1.2885198593139648, "step": 34480 }, { "epoch": 2.9390711546655304, "grad_norm": 2.03125, "learning_rate": 4.90015257647968e-05, "loss": 1.4511228561401368, "step": 34490 }, { "epoch": 2.93992330634853, "grad_norm": 2.640625, "learning_rate": 4.90007534192227e-05, "loss": 1.3295977592468262, "step": 34500 }, { "epoch": 2.9407754580315295, "grad_norm": 2.046875, "learning_rate": 4.8999980781140943e-05, "loss": 1.2875989913940429, "step": 34510 }, { "epoch": 2.941627609714529, "grad_norm": 2.484375, "learning_rate": 4.8999207850560944e-05, "loss": 1.3628891944885253, "step": 34520 }, { "epoch": 2.9424797613975286, "grad_norm": 1.8671875, "learning_rate": 4.899843462749213e-05, "loss": 1.4561284065246582, "step": 34530 }, { "epoch": 2.943331913080528, "grad_norm": 1.8515625, "learning_rate": 4.8997661111943915e-05, "loss": 1.410649299621582, "step": 34540 }, { "epoch": 2.944184064763528, "grad_norm": 3.3125, "learning_rate": 4.8996887303925745e-05, "loss": 1.3450519561767578, "step": 34550 }, { "epoch": 2.9450362164465274, "grad_norm": 2.15625, "learning_rate": 4.8996113203447026e-05, "loss": 1.3718523025512694, "step": 34560 }, { "epoch": 2.945888368129527, "grad_norm": 1.8125, "learning_rate": 4.8995338810517213e-05, "loss": 1.2795862197875976, "step": 34570 }, { "epoch": 2.9467405198125265, "grad_norm": 2.328125, "learning_rate": 4.899456412514573e-05, "loss": 1.4014400482177733, "step": 34580 }, { "epoch": 2.947592671495526, "grad_norm": 2.453125, "learning_rate": 4.8993789147342027e-05, "loss": 1.1852617263793945, "step": 34590 }, { "epoch": 2.9484448231785256, "grad_norm": 2.84375, "learning_rate": 4.8993013877115545e-05, "loss": 1.293740940093994, "step": 34600 }, { "epoch": 2.949296974861525, "grad_norm": 3.90625, "learning_rate": 4.899223831447573e-05, "loss": 1.3530562400817872, "step": 34610 }, { "epoch": 2.9501491265445248, "grad_norm": 1.796875, "learning_rate": 4.8991462459432056e-05, "loss": 1.2468744277954102, "step": 34620 }, { "epoch": 2.9510012782275243, "grad_norm": 2.421875, "learning_rate": 4.899068631199395e-05, "loss": 1.4001447677612304, "step": 34630 }, { "epoch": 2.951853429910524, "grad_norm": 2.109375, "learning_rate": 4.898990987217087e-05, "loss": 1.2903170585632324, "step": 34640 }, { "epoch": 2.9527055815935235, "grad_norm": 3.09375, "learning_rate": 4.89891331399723e-05, "loss": 1.3145498275756835, "step": 34650 }, { "epoch": 2.953557733276523, "grad_norm": 3.28125, "learning_rate": 4.89883561154077e-05, "loss": 1.3774328231811523, "step": 34660 }, { "epoch": 2.9544098849595226, "grad_norm": 2.75, "learning_rate": 4.898757879848653e-05, "loss": 1.2086758613586426, "step": 34670 }, { "epoch": 2.955262036642522, "grad_norm": 3.078125, "learning_rate": 4.8986801189218276e-05, "loss": 1.3756010055541992, "step": 34680 }, { "epoch": 2.9561141883255218, "grad_norm": 2.1875, "learning_rate": 4.898602328761241e-05, "loss": 1.3753944396972657, "step": 34690 }, { "epoch": 2.9569663400085213, "grad_norm": 1.703125, "learning_rate": 4.8985245093678405e-05, "loss": 1.4508041381835937, "step": 34700 }, { "epoch": 2.957818491691521, "grad_norm": 1.9453125, "learning_rate": 4.898446660742576e-05, "loss": 1.5261593818664552, "step": 34710 }, { "epoch": 2.9586706433745205, "grad_norm": 2.484375, "learning_rate": 4.898368782886395e-05, "loss": 1.2918740272521974, "step": 34720 }, { "epoch": 2.95952279505752, "grad_norm": 3.859375, "learning_rate": 4.898290875800247e-05, "loss": 1.528265380859375, "step": 34730 }, { "epoch": 2.9603749467405196, "grad_norm": 2.5, "learning_rate": 4.898212939485082e-05, "loss": 1.2958989143371582, "step": 34740 }, { "epoch": 2.961227098423519, "grad_norm": 3.0, "learning_rate": 4.8981349739418494e-05, "loss": 1.3220001220703126, "step": 34750 }, { "epoch": 2.9620792501065187, "grad_norm": 1.6328125, "learning_rate": 4.898056979171499e-05, "loss": 1.4681740760803224, "step": 34760 }, { "epoch": 2.9629314017895183, "grad_norm": 2.859375, "learning_rate": 4.897978955174982e-05, "loss": 1.2667096138000489, "step": 34770 }, { "epoch": 2.963783553472518, "grad_norm": 1.6953125, "learning_rate": 4.8979009019532496e-05, "loss": 1.4070310592651367, "step": 34780 }, { "epoch": 2.9646357051555174, "grad_norm": 2.734375, "learning_rate": 4.897822819507253e-05, "loss": 1.3178436279296875, "step": 34790 }, { "epoch": 2.965487856838517, "grad_norm": 2.515625, "learning_rate": 4.8977447078379426e-05, "loss": 1.307765007019043, "step": 34800 }, { "epoch": 2.9663400085215166, "grad_norm": 1.8984375, "learning_rate": 4.8976665669462714e-05, "loss": 1.2834477424621582, "step": 34810 }, { "epoch": 2.967192160204516, "grad_norm": 2.609375, "learning_rate": 4.8975883968331914e-05, "loss": 1.485699462890625, "step": 34820 }, { "epoch": 2.9680443118875157, "grad_norm": 2.3125, "learning_rate": 4.897510197499655e-05, "loss": 1.4322772026062012, "step": 34830 }, { "epoch": 2.9688964635705153, "grad_norm": 2.234375, "learning_rate": 4.8974319689466166e-05, "loss": 1.258878517150879, "step": 34840 }, { "epoch": 2.969748615253515, "grad_norm": 2.25, "learning_rate": 4.897353711175028e-05, "loss": 1.4523695945739745, "step": 34850 }, { "epoch": 2.9706007669365144, "grad_norm": 2.125, "learning_rate": 4.897275424185844e-05, "loss": 1.4717657089233398, "step": 34860 }, { "epoch": 2.9714529186195144, "grad_norm": 3.75, "learning_rate": 4.8971971079800186e-05, "loss": 1.3436800956726074, "step": 34870 }, { "epoch": 2.972305070302514, "grad_norm": 1.859375, "learning_rate": 4.897118762558506e-05, "loss": 1.276312255859375, "step": 34880 }, { "epoch": 2.9731572219855136, "grad_norm": 2.59375, "learning_rate": 4.89704038792226e-05, "loss": 1.055103588104248, "step": 34890 }, { "epoch": 2.974009373668513, "grad_norm": 2.71875, "learning_rate": 4.896961984072238e-05, "loss": 1.3430097579956055, "step": 34900 }, { "epoch": 2.9748615253515127, "grad_norm": 2.703125, "learning_rate": 4.896883551009394e-05, "loss": 1.3315566062927247, "step": 34910 }, { "epoch": 2.9757136770345123, "grad_norm": 3.109375, "learning_rate": 4.896805088734684e-05, "loss": 1.2879472732543946, "step": 34920 }, { "epoch": 2.976565828717512, "grad_norm": 2.234375, "learning_rate": 4.896726597249065e-05, "loss": 1.3994495391845703, "step": 34930 }, { "epoch": 2.9774179804005114, "grad_norm": 1.765625, "learning_rate": 4.896648076553494e-05, "loss": 1.4177843093872071, "step": 34940 }, { "epoch": 2.978270132083511, "grad_norm": 2.546875, "learning_rate": 4.8965695266489255e-05, "loss": 1.3502491950988769, "step": 34950 }, { "epoch": 2.9791222837665106, "grad_norm": 1.9453125, "learning_rate": 4.89649094753632e-05, "loss": 1.4128960609436034, "step": 34960 }, { "epoch": 2.97997443544951, "grad_norm": 2.203125, "learning_rate": 4.8964123392166325e-05, "loss": 1.2939409255981444, "step": 34970 }, { "epoch": 2.9808265871325097, "grad_norm": 2.078125, "learning_rate": 4.8963337016908215e-05, "loss": 1.4482420921325683, "step": 34980 }, { "epoch": 2.9816787388155093, "grad_norm": 2.65625, "learning_rate": 4.896255034959848e-05, "loss": 1.289161205291748, "step": 34990 }, { "epoch": 2.982530890498509, "grad_norm": 2.96875, "learning_rate": 4.8961763390246674e-05, "loss": 1.341127586364746, "step": 35000 }, { "epoch": 2.9833830421815084, "grad_norm": 3.078125, "learning_rate": 4.89609761388624e-05, "loss": 1.335605525970459, "step": 35010 }, { "epoch": 2.984235193864508, "grad_norm": 2.0, "learning_rate": 4.8960188595455264e-05, "loss": 1.3580482482910157, "step": 35020 }, { "epoch": 2.9850873455475075, "grad_norm": 3.359375, "learning_rate": 4.8959400760034855e-05, "loss": 1.2916165351867677, "step": 35030 }, { "epoch": 2.985939497230507, "grad_norm": 1.984375, "learning_rate": 4.895861263261077e-05, "loss": 1.3057744026184082, "step": 35040 }, { "epoch": 2.9867916489135067, "grad_norm": 1.890625, "learning_rate": 4.8957824213192616e-05, "loss": 1.4185293197631836, "step": 35050 }, { "epoch": 2.9876438005965062, "grad_norm": 2.578125, "learning_rate": 4.8957035501790004e-05, "loss": 1.4236943244934082, "step": 35060 }, { "epoch": 2.988495952279506, "grad_norm": 2.203125, "learning_rate": 4.895624649841255e-05, "loss": 1.3612574577331542, "step": 35070 }, { "epoch": 2.9893481039625054, "grad_norm": 2.25, "learning_rate": 4.895545720306987e-05, "loss": 1.4583251953125, "step": 35080 }, { "epoch": 2.990200255645505, "grad_norm": 3.59375, "learning_rate": 4.895466761577158e-05, "loss": 1.2996479988098144, "step": 35090 }, { "epoch": 2.9910524073285045, "grad_norm": 1.9765625, "learning_rate": 4.89538777365273e-05, "loss": 1.429866409301758, "step": 35100 }, { "epoch": 2.991904559011504, "grad_norm": 4.25, "learning_rate": 4.8953087565346654e-05, "loss": 1.4731996536254883, "step": 35110 }, { "epoch": 2.9927567106945037, "grad_norm": 1.828125, "learning_rate": 4.895229710223928e-05, "loss": 1.3156149864196778, "step": 35120 }, { "epoch": 2.9936088623775032, "grad_norm": 2.359375, "learning_rate": 4.895150634721483e-05, "loss": 1.3957976341247558, "step": 35130 }, { "epoch": 2.994461014060503, "grad_norm": 2.75, "learning_rate": 4.8950715300282904e-05, "loss": 1.2818201065063477, "step": 35140 }, { "epoch": 2.9953131657435024, "grad_norm": 2.609375, "learning_rate": 4.894992396145316e-05, "loss": 1.4767032623291017, "step": 35150 }, { "epoch": 2.996165317426502, "grad_norm": 2.328125, "learning_rate": 4.894913233073524e-05, "loss": 1.4657261848449707, "step": 35160 }, { "epoch": 2.9970174691095015, "grad_norm": 3.15625, "learning_rate": 4.894834040813879e-05, "loss": 1.4407883644104005, "step": 35170 }, { "epoch": 2.997869620792501, "grad_norm": 1.8671875, "learning_rate": 4.894754819367348e-05, "loss": 1.2496471405029297, "step": 35180 }, { "epoch": 2.9987217724755006, "grad_norm": 2.640625, "learning_rate": 4.894675568734894e-05, "loss": 1.4754337310791015, "step": 35190 }, { "epoch": 2.9995739241585, "grad_norm": 2.578125, "learning_rate": 4.894596288917485e-05, "loss": 1.428765869140625, "step": 35200 }, { "epoch": 2.9997443544951, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6493, "eval_loss": 1.212781548500061, "eval_runtime": 53.7997, "eval_samples_per_second": 367.325, "eval_steps_per_second": 22.974, "step": 35202 }, { "epoch": 3.0004260758415, "grad_norm": 4.03125, "learning_rate": 4.894516979916085e-05, "loss": 1.351663875579834, "step": 35210 }, { "epoch": 3.0012782275244994, "grad_norm": 2.265625, "learning_rate": 4.8944376417316626e-05, "loss": 1.2592275619506836, "step": 35220 }, { "epoch": 3.002130379207499, "grad_norm": 1.71875, "learning_rate": 4.894358274365183e-05, "loss": 1.3806751251220704, "step": 35230 }, { "epoch": 3.0029825308904985, "grad_norm": 2.609375, "learning_rate": 4.894278877817615e-05, "loss": 1.3275663375854492, "step": 35240 }, { "epoch": 3.003834682573498, "grad_norm": 1.953125, "learning_rate": 4.894199452089926e-05, "loss": 1.4783357620239257, "step": 35250 }, { "epoch": 3.0046868342564976, "grad_norm": 1.84375, "learning_rate": 4.8941199971830827e-05, "loss": 1.2589373588562012, "step": 35260 }, { "epoch": 3.005538985939497, "grad_norm": 1.7421875, "learning_rate": 4.8940405130980544e-05, "loss": 1.3920363426208495, "step": 35270 }, { "epoch": 3.0063911376224968, "grad_norm": 2.296875, "learning_rate": 4.89396099983581e-05, "loss": 1.411501121520996, "step": 35280 }, { "epoch": 3.0072432893054963, "grad_norm": 1.96875, "learning_rate": 4.893881457397318e-05, "loss": 1.387434482574463, "step": 35290 }, { "epoch": 3.008095440988496, "grad_norm": 2.40625, "learning_rate": 4.893801885783549e-05, "loss": 1.2976507186889648, "step": 35300 }, { "epoch": 3.0089475926714955, "grad_norm": 2.046875, "learning_rate": 4.893722284995471e-05, "loss": 1.529210662841797, "step": 35310 }, { "epoch": 3.009799744354495, "grad_norm": 1.9140625, "learning_rate": 4.8936426550340554e-05, "loss": 1.441592025756836, "step": 35320 }, { "epoch": 3.0106518960374946, "grad_norm": 2.4375, "learning_rate": 4.893562995900272e-05, "loss": 1.4533265113830567, "step": 35330 }, { "epoch": 3.011504047720494, "grad_norm": 1.7578125, "learning_rate": 4.8934833075950925e-05, "loss": 1.4520505905151366, "step": 35340 }, { "epoch": 3.0123561994034938, "grad_norm": 2.546875, "learning_rate": 4.893403590119487e-05, "loss": 1.359769058227539, "step": 35350 }, { "epoch": 3.0132083510864933, "grad_norm": 1.75, "learning_rate": 4.893323843474428e-05, "loss": 1.4481576919555663, "step": 35360 }, { "epoch": 3.014060502769493, "grad_norm": 2.546875, "learning_rate": 4.893244067660886e-05, "loss": 1.2086164474487304, "step": 35370 }, { "epoch": 3.0149126544524925, "grad_norm": 3.015625, "learning_rate": 4.8931642626798356e-05, "loss": 1.4149481773376464, "step": 35380 }, { "epoch": 3.015764806135492, "grad_norm": 1.8203125, "learning_rate": 4.8930844285322475e-05, "loss": 1.3745408058166504, "step": 35390 }, { "epoch": 3.0166169578184916, "grad_norm": 1.7890625, "learning_rate": 4.893004565219095e-05, "loss": 1.3692413330078126, "step": 35400 }, { "epoch": 3.017469109501491, "grad_norm": 3.546875, "learning_rate": 4.892924672741352e-05, "loss": 1.3091262817382812, "step": 35410 }, { "epoch": 3.0183212611844907, "grad_norm": 1.921875, "learning_rate": 4.892844751099992e-05, "loss": 1.342989444732666, "step": 35420 }, { "epoch": 3.0191734128674903, "grad_norm": 2.375, "learning_rate": 4.892764800295989e-05, "loss": 1.2605403900146483, "step": 35430 }, { "epoch": 3.02002556455049, "grad_norm": 1.8515625, "learning_rate": 4.892684820330317e-05, "loss": 1.3083900451660155, "step": 35440 }, { "epoch": 3.0208777162334894, "grad_norm": 1.8515625, "learning_rate": 4.892604811203951e-05, "loss": 1.3849977493286132, "step": 35450 }, { "epoch": 3.021729867916489, "grad_norm": 2.125, "learning_rate": 4.892524772917867e-05, "loss": 1.3931940078735352, "step": 35460 }, { "epoch": 3.0225820195994886, "grad_norm": 2.25, "learning_rate": 4.892444705473038e-05, "loss": 1.336756992340088, "step": 35470 }, { "epoch": 3.023434171282488, "grad_norm": 3.703125, "learning_rate": 4.8923646088704434e-05, "loss": 1.3725021362304688, "step": 35480 }, { "epoch": 3.0242863229654877, "grad_norm": 2.453125, "learning_rate": 4.8922844831110556e-05, "loss": 1.3397750854492188, "step": 35490 }, { "epoch": 3.0251384746484873, "grad_norm": 3.015625, "learning_rate": 4.892204328195854e-05, "loss": 1.4387834548950196, "step": 35500 }, { "epoch": 3.025990626331487, "grad_norm": 1.7890625, "learning_rate": 4.892124144125815e-05, "loss": 1.4420831680297852, "step": 35510 }, { "epoch": 3.0268427780144864, "grad_norm": 2.375, "learning_rate": 4.892043930901914e-05, "loss": 1.2486224174499512, "step": 35520 }, { "epoch": 3.027694929697486, "grad_norm": 3.046875, "learning_rate": 4.89196368852513e-05, "loss": 1.2169286727905273, "step": 35530 }, { "epoch": 3.0285470813804856, "grad_norm": 2.078125, "learning_rate": 4.8918834169964414e-05, "loss": 1.4186469078063966, "step": 35540 }, { "epoch": 3.029399233063485, "grad_norm": 2.6875, "learning_rate": 4.891803116316825e-05, "loss": 1.28770112991333, "step": 35550 }, { "epoch": 3.0302513847464847, "grad_norm": 2.203125, "learning_rate": 4.8917227864872613e-05, "loss": 1.3163507461547852, "step": 35560 }, { "epoch": 3.0311035364294843, "grad_norm": 2.109375, "learning_rate": 4.891642427508728e-05, "loss": 1.4094456672668456, "step": 35570 }, { "epoch": 3.031955688112484, "grad_norm": 2.796875, "learning_rate": 4.891562039382205e-05, "loss": 1.429908847808838, "step": 35580 }, { "epoch": 3.0328078397954834, "grad_norm": 2.0625, "learning_rate": 4.8914816221086724e-05, "loss": 1.4832121849060058, "step": 35590 }, { "epoch": 3.033659991478483, "grad_norm": 3.296875, "learning_rate": 4.891401175689109e-05, "loss": 1.2217252731323243, "step": 35600 }, { "epoch": 3.0345121431614825, "grad_norm": 3.296875, "learning_rate": 4.891320700124496e-05, "loss": 1.4487842559814452, "step": 35610 }, { "epoch": 3.035364294844482, "grad_norm": 2.515625, "learning_rate": 4.8912401954158145e-05, "loss": 1.3765686988830566, "step": 35620 }, { "epoch": 3.0362164465274817, "grad_norm": 2.671875, "learning_rate": 4.8911596615640456e-05, "loss": 1.2528167724609376, "step": 35630 }, { "epoch": 3.0370685982104813, "grad_norm": 2.4375, "learning_rate": 4.891079098570169e-05, "loss": 1.3745035171508788, "step": 35640 }, { "epoch": 3.0379207498934813, "grad_norm": 4.03125, "learning_rate": 4.89099850643517e-05, "loss": 1.1482791900634766, "step": 35650 }, { "epoch": 3.038772901576481, "grad_norm": 1.703125, "learning_rate": 4.890917885160028e-05, "loss": 1.3587943077087403, "step": 35660 }, { "epoch": 3.0396250532594804, "grad_norm": 2.375, "learning_rate": 4.890837234745726e-05, "loss": 1.4526466369628905, "step": 35670 }, { "epoch": 3.04047720494248, "grad_norm": 2.09375, "learning_rate": 4.8907565551932486e-05, "loss": 1.3553125381469726, "step": 35680 }, { "epoch": 3.0413293566254795, "grad_norm": 2.578125, "learning_rate": 4.8906758465035775e-05, "loss": 1.3514344215393066, "step": 35690 }, { "epoch": 3.042181508308479, "grad_norm": 3.28125, "learning_rate": 4.8905951086776955e-05, "loss": 1.3160470008850098, "step": 35700 }, { "epoch": 3.0430336599914787, "grad_norm": 1.8046875, "learning_rate": 4.8905143417165886e-05, "loss": 1.3134917259216308, "step": 35710 }, { "epoch": 3.0438858116744782, "grad_norm": 2.765625, "learning_rate": 4.8904335456212414e-05, "loss": 1.3659215927124024, "step": 35720 }, { "epoch": 3.044737963357478, "grad_norm": 2.953125, "learning_rate": 4.8903527203926364e-05, "loss": 1.379234218597412, "step": 35730 }, { "epoch": 3.0455901150404774, "grad_norm": 2.390625, "learning_rate": 4.8902718660317596e-05, "loss": 1.2527147293090821, "step": 35740 }, { "epoch": 3.046442266723477, "grad_norm": 5.03125, "learning_rate": 4.8901909825395977e-05, "loss": 1.3165556907653808, "step": 35750 }, { "epoch": 3.0472944184064765, "grad_norm": 2.390625, "learning_rate": 4.8901100699171344e-05, "loss": 1.2738828659057617, "step": 35760 }, { "epoch": 3.048146570089476, "grad_norm": 1.9609375, "learning_rate": 4.8900291281653565e-05, "loss": 1.37349853515625, "step": 35770 }, { "epoch": 3.0489987217724757, "grad_norm": 2.234375, "learning_rate": 4.889948157285251e-05, "loss": 1.4047639846801758, "step": 35780 }, { "epoch": 3.0498508734554752, "grad_norm": 2.578125, "learning_rate": 4.889867157277805e-05, "loss": 1.2845949172973632, "step": 35790 }, { "epoch": 3.050703025138475, "grad_norm": 2.4375, "learning_rate": 4.889786128144005e-05, "loss": 1.2989027976989747, "step": 35800 }, { "epoch": 3.0515551768214744, "grad_norm": 3.21875, "learning_rate": 4.889705069884838e-05, "loss": 1.1811839103698731, "step": 35810 }, { "epoch": 3.052407328504474, "grad_norm": 2.578125, "learning_rate": 4.889623982501294e-05, "loss": 1.3433961868286133, "step": 35820 }, { "epoch": 3.0532594801874735, "grad_norm": 1.78125, "learning_rate": 4.8895428659943586e-05, "loss": 1.3829682350158692, "step": 35830 }, { "epoch": 3.054111631870473, "grad_norm": 2.828125, "learning_rate": 4.889461720365022e-05, "loss": 1.2124544143676759, "step": 35840 }, { "epoch": 3.0549637835534726, "grad_norm": 2.5, "learning_rate": 4.889380545614273e-05, "loss": 1.412430763244629, "step": 35850 }, { "epoch": 3.055815935236472, "grad_norm": 3.234375, "learning_rate": 4.8892993417431e-05, "loss": 1.2702129364013672, "step": 35860 }, { "epoch": 3.056668086919472, "grad_norm": 1.75, "learning_rate": 4.889218108752494e-05, "loss": 1.4570496559143067, "step": 35870 }, { "epoch": 3.0575202386024714, "grad_norm": 1.46875, "learning_rate": 4.889136846643444e-05, "loss": 1.2285653114318849, "step": 35880 }, { "epoch": 3.058372390285471, "grad_norm": 2.4375, "learning_rate": 4.88905555541694e-05, "loss": 1.3810790061950684, "step": 35890 }, { "epoch": 3.0592245419684705, "grad_norm": 2.5, "learning_rate": 4.888974235073975e-05, "loss": 1.3393364906311036, "step": 35900 }, { "epoch": 3.06007669365147, "grad_norm": 2.125, "learning_rate": 4.888892885615538e-05, "loss": 1.307762908935547, "step": 35910 }, { "epoch": 3.0609288453344696, "grad_norm": 3.21875, "learning_rate": 4.8888115070426214e-05, "loss": 1.1849114418029785, "step": 35920 }, { "epoch": 3.061780997017469, "grad_norm": 3.0625, "learning_rate": 4.8887300993562155e-05, "loss": 1.4679147720336914, "step": 35930 }, { "epoch": 3.0626331487004688, "grad_norm": 2.4375, "learning_rate": 4.888648662557315e-05, "loss": 1.268840503692627, "step": 35940 }, { "epoch": 3.0634853003834683, "grad_norm": 2.234375, "learning_rate": 4.88856719664691e-05, "loss": 1.3792448043823242, "step": 35950 }, { "epoch": 3.064337452066468, "grad_norm": 4.125, "learning_rate": 4.888485701625994e-05, "loss": 1.3571955680847168, "step": 35960 }, { "epoch": 3.0651896037494675, "grad_norm": 3.0, "learning_rate": 4.888404177495562e-05, "loss": 1.3884598731994628, "step": 35970 }, { "epoch": 3.066041755432467, "grad_norm": 1.8515625, "learning_rate": 4.888322624256605e-05, "loss": 1.4244738578796388, "step": 35980 }, { "epoch": 3.0668939071154666, "grad_norm": 2.3125, "learning_rate": 4.888241041910119e-05, "loss": 1.2712902069091796, "step": 35990 }, { "epoch": 3.067746058798466, "grad_norm": 1.7734375, "learning_rate": 4.888159430457096e-05, "loss": 1.2776398658752441, "step": 36000 }, { "epoch": 3.0685982104814657, "grad_norm": 2.109375, "learning_rate": 4.888077789898533e-05, "loss": 1.4667505264282226, "step": 36010 }, { "epoch": 3.0694503621644653, "grad_norm": 2.359375, "learning_rate": 4.887996120235423e-05, "loss": 1.2686013221740722, "step": 36020 }, { "epoch": 3.070302513847465, "grad_norm": 2.3125, "learning_rate": 4.8879144214687634e-05, "loss": 1.2413738250732422, "step": 36030 }, { "epoch": 3.0711546655304645, "grad_norm": 2.421875, "learning_rate": 4.8878326935995476e-05, "loss": 1.4213239669799804, "step": 36040 }, { "epoch": 3.072006817213464, "grad_norm": 2.359375, "learning_rate": 4.8877509366287735e-05, "loss": 1.4417719841003418, "step": 36050 }, { "epoch": 3.0728589688964636, "grad_norm": 2.4375, "learning_rate": 4.887669150557437e-05, "loss": 1.207223892211914, "step": 36060 }, { "epoch": 3.073711120579463, "grad_norm": 1.6953125, "learning_rate": 4.887587335386535e-05, "loss": 1.2261752128601073, "step": 36070 }, { "epoch": 3.0745632722624627, "grad_norm": 2.578125, "learning_rate": 4.887505491117064e-05, "loss": 1.3172348976135253, "step": 36080 }, { "epoch": 3.0754154239454623, "grad_norm": 1.953125, "learning_rate": 4.887423617750021e-05, "loss": 1.3811205863952636, "step": 36090 }, { "epoch": 3.076267575628462, "grad_norm": 2.90625, "learning_rate": 4.8873417152864055e-05, "loss": 1.4887721061706543, "step": 36100 }, { "epoch": 3.0771197273114614, "grad_norm": 2.15625, "learning_rate": 4.887259783727215e-05, "loss": 1.281413459777832, "step": 36110 }, { "epoch": 3.077971878994461, "grad_norm": 2.375, "learning_rate": 4.8871778230734474e-05, "loss": 1.2262761116027832, "step": 36120 }, { "epoch": 3.0788240306774606, "grad_norm": 2.90625, "learning_rate": 4.887095833326102e-05, "loss": 1.3897335052490234, "step": 36130 }, { "epoch": 3.07967618236046, "grad_norm": 1.75, "learning_rate": 4.887013814486178e-05, "loss": 1.2455699920654297, "step": 36140 }, { "epoch": 3.0805283340434597, "grad_norm": 1.96875, "learning_rate": 4.8869317665546756e-05, "loss": 1.3121232986450195, "step": 36150 }, { "epoch": 3.0813804857264593, "grad_norm": 1.828125, "learning_rate": 4.886849689532594e-05, "loss": 1.3534335136413573, "step": 36160 }, { "epoch": 3.082232637409459, "grad_norm": 2.53125, "learning_rate": 4.8867675834209334e-05, "loss": 1.3439364433288574, "step": 36170 }, { "epoch": 3.0830847890924584, "grad_norm": 1.7109375, "learning_rate": 4.886685448220695e-05, "loss": 1.4172181129455566, "step": 36180 }, { "epoch": 3.083936940775458, "grad_norm": 2.515625, "learning_rate": 4.88660328393288e-05, "loss": 1.3760550498962403, "step": 36190 }, { "epoch": 3.0847890924584576, "grad_norm": 1.7421875, "learning_rate": 4.886521090558489e-05, "loss": 1.2180950164794921, "step": 36200 }, { "epoch": 3.085641244141457, "grad_norm": 2.765625, "learning_rate": 4.886438868098524e-05, "loss": 1.3893033981323242, "step": 36210 }, { "epoch": 3.0864933958244567, "grad_norm": 2.0625, "learning_rate": 4.8863566165539875e-05, "loss": 1.3688545227050781, "step": 36220 }, { "epoch": 3.0873455475074563, "grad_norm": 2.75, "learning_rate": 4.886274335925882e-05, "loss": 1.245336151123047, "step": 36230 }, { "epoch": 3.088197699190456, "grad_norm": 2.046875, "learning_rate": 4.886192026215209e-05, "loss": 1.3179837226867677, "step": 36240 }, { "epoch": 3.0890498508734554, "grad_norm": 1.7578125, "learning_rate": 4.8861096874229726e-05, "loss": 1.3794519424438476, "step": 36250 }, { "epoch": 3.089902002556455, "grad_norm": 2.90625, "learning_rate": 4.8860273195501774e-05, "loss": 1.436665916442871, "step": 36260 }, { "epoch": 3.0907541542394545, "grad_norm": 2.21875, "learning_rate": 4.8859449225978256e-05, "loss": 1.4353184700012207, "step": 36270 }, { "epoch": 3.091606305922454, "grad_norm": 1.703125, "learning_rate": 4.885862496566922e-05, "loss": 1.4036274909973145, "step": 36280 }, { "epoch": 3.0924584576054537, "grad_norm": 2.4375, "learning_rate": 4.8857800414584706e-05, "loss": 1.3606070518493651, "step": 36290 }, { "epoch": 3.0933106092884533, "grad_norm": 2.71875, "learning_rate": 4.885697557273477e-05, "loss": 1.3729371070861816, "step": 36300 }, { "epoch": 3.094162760971453, "grad_norm": 3.625, "learning_rate": 4.8856150440129464e-05, "loss": 1.4340137481689452, "step": 36310 }, { "epoch": 3.0950149126544524, "grad_norm": 1.953125, "learning_rate": 4.885532501677884e-05, "loss": 1.4299155235290528, "step": 36320 }, { "epoch": 3.095867064337452, "grad_norm": 2.234375, "learning_rate": 4.885449930269297e-05, "loss": 1.384950542449951, "step": 36330 }, { "epoch": 3.0967192160204515, "grad_norm": 2.875, "learning_rate": 4.88536732978819e-05, "loss": 1.3926697731018067, "step": 36340 }, { "epoch": 3.097571367703451, "grad_norm": 2.375, "learning_rate": 4.88528470023557e-05, "loss": 1.398355484008789, "step": 36350 }, { "epoch": 3.0984235193864507, "grad_norm": 3.203125, "learning_rate": 4.885202041612446e-05, "loss": 1.350684928894043, "step": 36360 }, { "epoch": 3.0992756710694502, "grad_norm": 2.609375, "learning_rate": 4.885119353919823e-05, "loss": 1.3959571838378906, "step": 36370 }, { "epoch": 3.10012782275245, "grad_norm": 1.5234375, "learning_rate": 4.8850366371587095e-05, "loss": 1.1630159378051759, "step": 36380 }, { "epoch": 3.1009799744354494, "grad_norm": 2.125, "learning_rate": 4.8849538913301144e-05, "loss": 1.310249423980713, "step": 36390 }, { "epoch": 3.101832126118449, "grad_norm": 2.328125, "learning_rate": 4.8848711164350455e-05, "loss": 1.3971368789672851, "step": 36400 }, { "epoch": 3.1026842778014485, "grad_norm": 2.53125, "learning_rate": 4.884788312474512e-05, "loss": 1.2797934532165527, "step": 36410 }, { "epoch": 3.103536429484448, "grad_norm": 2.21875, "learning_rate": 4.8847054794495217e-05, "loss": 1.4249909400939942, "step": 36420 }, { "epoch": 3.1043885811674476, "grad_norm": 2.578125, "learning_rate": 4.884622617361086e-05, "loss": 1.3400150299072267, "step": 36430 }, { "epoch": 3.105240732850447, "grad_norm": 2.515625, "learning_rate": 4.884539726210214e-05, "loss": 1.4513494491577148, "step": 36440 }, { "epoch": 3.106092884533447, "grad_norm": 1.6640625, "learning_rate": 4.884456805997916e-05, "loss": 1.3222375869750977, "step": 36450 }, { "epoch": 3.1069450362164464, "grad_norm": 1.9921875, "learning_rate": 4.884373856725202e-05, "loss": 1.4538945198059081, "step": 36460 }, { "epoch": 3.107797187899446, "grad_norm": 3.125, "learning_rate": 4.884290878393083e-05, "loss": 1.584315013885498, "step": 36470 }, { "epoch": 3.1086493395824455, "grad_norm": 2.421875, "learning_rate": 4.8842078710025704e-05, "loss": 1.5430667877197266, "step": 36480 }, { "epoch": 3.109501491265445, "grad_norm": 3.09375, "learning_rate": 4.884124834554677e-05, "loss": 1.2467462539672851, "step": 36490 }, { "epoch": 3.1103536429484446, "grad_norm": 1.78125, "learning_rate": 4.884041769050414e-05, "loss": 1.2517412185668946, "step": 36500 }, { "epoch": 3.111205794631444, "grad_norm": 1.7890625, "learning_rate": 4.883958674490794e-05, "loss": 1.4265724182128907, "step": 36510 }, { "epoch": 3.1120579463144438, "grad_norm": 2.703125, "learning_rate": 4.883875550876829e-05, "loss": 1.4492650985717774, "step": 36520 }, { "epoch": 3.1129100979974433, "grad_norm": 1.703125, "learning_rate": 4.8837923982095324e-05, "loss": 1.358853816986084, "step": 36530 }, { "epoch": 3.113762249680443, "grad_norm": 2.125, "learning_rate": 4.883709216489917e-05, "loss": 1.342961311340332, "step": 36540 }, { "epoch": 3.1146144013634425, "grad_norm": 2.234375, "learning_rate": 4.883626005718999e-05, "loss": 1.4081589698791503, "step": 36550 }, { "epoch": 3.115466553046442, "grad_norm": 1.8046875, "learning_rate": 4.883542765897789e-05, "loss": 1.4262554168701171, "step": 36560 }, { "epoch": 3.1163187047294416, "grad_norm": 2.734375, "learning_rate": 4.883459497027304e-05, "loss": 1.2744508743286134, "step": 36570 }, { "epoch": 3.1171708564124416, "grad_norm": 2.265625, "learning_rate": 4.883376199108558e-05, "loss": 1.2825817108154296, "step": 36580 }, { "epoch": 3.118023008095441, "grad_norm": 2.640625, "learning_rate": 4.883292872142567e-05, "loss": 1.264284038543701, "step": 36590 }, { "epoch": 3.1188751597784408, "grad_norm": 3.109375, "learning_rate": 4.883209516130345e-05, "loss": 1.4534035682678224, "step": 36600 }, { "epoch": 3.1197273114614403, "grad_norm": 1.6796875, "learning_rate": 4.8831261310729084e-05, "loss": 1.305465030670166, "step": 36610 }, { "epoch": 3.12057946314444, "grad_norm": 2.203125, "learning_rate": 4.883042716971275e-05, "loss": 1.350407028198242, "step": 36620 }, { "epoch": 3.1214316148274395, "grad_norm": 1.640625, "learning_rate": 4.882959273826459e-05, "loss": 1.4212772369384765, "step": 36630 }, { "epoch": 3.122283766510439, "grad_norm": 2.09375, "learning_rate": 4.882875801639479e-05, "loss": 1.3491463661193848, "step": 36640 }, { "epoch": 3.1231359181934386, "grad_norm": 2.328125, "learning_rate": 4.882792300411352e-05, "loss": 1.3787240982055664, "step": 36650 }, { "epoch": 3.123988069876438, "grad_norm": 3.140625, "learning_rate": 4.882708770143095e-05, "loss": 1.41451997756958, "step": 36660 }, { "epoch": 3.1248402215594377, "grad_norm": 1.8515625, "learning_rate": 4.882625210835727e-05, "loss": 1.289640235900879, "step": 36670 }, { "epoch": 3.1256923732424373, "grad_norm": 3.734375, "learning_rate": 4.882541622490266e-05, "loss": 1.301461410522461, "step": 36680 }, { "epoch": 3.126544524925437, "grad_norm": 1.75, "learning_rate": 4.8824580051077305e-05, "loss": 1.3180501937866211, "step": 36690 }, { "epoch": 3.1273966766084365, "grad_norm": 4.6875, "learning_rate": 4.88237435868914e-05, "loss": 1.2354559898376465, "step": 36700 }, { "epoch": 3.128248828291436, "grad_norm": 4.3125, "learning_rate": 4.882290683235513e-05, "loss": 1.3459080696105956, "step": 36710 }, { "epoch": 3.1291009799744356, "grad_norm": 2.546875, "learning_rate": 4.88220697874787e-05, "loss": 1.0804141998291015, "step": 36720 }, { "epoch": 3.129953131657435, "grad_norm": 2.078125, "learning_rate": 4.8821232452272314e-05, "loss": 1.2994089126586914, "step": 36730 }, { "epoch": 3.1308052833404347, "grad_norm": 3.328125, "learning_rate": 4.882039482674618e-05, "loss": 1.2406600952148437, "step": 36740 }, { "epoch": 3.1316574350234343, "grad_norm": 2.640625, "learning_rate": 4.881955691091049e-05, "loss": 1.2959458351135253, "step": 36750 }, { "epoch": 3.132509586706434, "grad_norm": 2.75, "learning_rate": 4.881871870477547e-05, "loss": 1.2835969924926758, "step": 36760 }, { "epoch": 3.1333617383894334, "grad_norm": 2.65625, "learning_rate": 4.881788020835133e-05, "loss": 1.3359998703002929, "step": 36770 }, { "epoch": 3.134213890072433, "grad_norm": 1.84375, "learning_rate": 4.8817041421648294e-05, "loss": 1.4118128776550294, "step": 36780 }, { "epoch": 3.1350660417554326, "grad_norm": 1.859375, "learning_rate": 4.8816202344676586e-05, "loss": 1.4269683837890625, "step": 36790 }, { "epoch": 3.135918193438432, "grad_norm": 2.828125, "learning_rate": 4.881536297744642e-05, "loss": 1.3406451225280762, "step": 36800 }, { "epoch": 3.1367703451214317, "grad_norm": 3.765625, "learning_rate": 4.881452331996804e-05, "loss": 1.351135540008545, "step": 36810 }, { "epoch": 3.1376224968044313, "grad_norm": 1.8203125, "learning_rate": 4.8813683372251665e-05, "loss": 1.4144777297973632, "step": 36820 }, { "epoch": 3.138474648487431, "grad_norm": 2.25, "learning_rate": 4.881284313430755e-05, "loss": 1.3285079956054688, "step": 36830 }, { "epoch": 3.1393268001704304, "grad_norm": 1.921875, "learning_rate": 4.881200260614591e-05, "loss": 1.3416215896606445, "step": 36840 }, { "epoch": 3.14017895185343, "grad_norm": 2.0625, "learning_rate": 4.8811161787777014e-05, "loss": 1.4600982666015625, "step": 36850 }, { "epoch": 3.1410311035364296, "grad_norm": 1.8046875, "learning_rate": 4.8810320679211106e-05, "loss": 1.328850555419922, "step": 36860 }, { "epoch": 3.141883255219429, "grad_norm": 2.546875, "learning_rate": 4.880947928045842e-05, "loss": 1.3592340469360351, "step": 36870 }, { "epoch": 3.1427354069024287, "grad_norm": 3.296875, "learning_rate": 4.8808637591529226e-05, "loss": 1.30679349899292, "step": 36880 }, { "epoch": 3.1435875585854283, "grad_norm": 2.734375, "learning_rate": 4.880779561243377e-05, "loss": 1.4474233627319335, "step": 36890 }, { "epoch": 3.144439710268428, "grad_norm": 2.015625, "learning_rate": 4.8806953343182324e-05, "loss": 1.2167829513549804, "step": 36900 }, { "epoch": 3.1452918619514274, "grad_norm": 2.40625, "learning_rate": 4.8806110783785154e-05, "loss": 1.348312759399414, "step": 36910 }, { "epoch": 3.146144013634427, "grad_norm": 3.265625, "learning_rate": 4.880526793425253e-05, "loss": 1.402181053161621, "step": 36920 }, { "epoch": 3.1469961653174265, "grad_norm": 3.375, "learning_rate": 4.8804424794594705e-05, "loss": 1.2440653800964356, "step": 36930 }, { "epoch": 3.147848317000426, "grad_norm": 2.796875, "learning_rate": 4.880358136482197e-05, "loss": 1.32386474609375, "step": 36940 }, { "epoch": 3.1487004686834257, "grad_norm": 2.15625, "learning_rate": 4.880273764494461e-05, "loss": 1.333210563659668, "step": 36950 }, { "epoch": 3.1495526203664252, "grad_norm": 2.75, "learning_rate": 4.8801893634972894e-05, "loss": 1.2332324028015136, "step": 36960 }, { "epoch": 3.150404772049425, "grad_norm": 1.9375, "learning_rate": 4.8801049334917115e-05, "loss": 1.4447650909423828, "step": 36970 }, { "epoch": 3.1512569237324244, "grad_norm": 2.609375, "learning_rate": 4.880020474478756e-05, "loss": 1.3622499465942384, "step": 36980 }, { "epoch": 3.152109075415424, "grad_norm": 3.09375, "learning_rate": 4.879935986459453e-05, "loss": 1.3132010459899903, "step": 36990 }, { "epoch": 3.1529612270984235, "grad_norm": 3.109375, "learning_rate": 4.8798514694348315e-05, "loss": 1.3542436599731444, "step": 37000 }, { "epoch": 3.153813378781423, "grad_norm": 2.46875, "learning_rate": 4.8797669234059216e-05, "loss": 1.4185931205749511, "step": 37010 }, { "epoch": 3.1546655304644227, "grad_norm": 2.40625, "learning_rate": 4.879682348373754e-05, "loss": 1.220420742034912, "step": 37020 }, { "epoch": 3.1555176821474222, "grad_norm": 4.40625, "learning_rate": 4.879597744339359e-05, "loss": 1.304822826385498, "step": 37030 }, { "epoch": 3.156369833830422, "grad_norm": 3.40625, "learning_rate": 4.879513111303768e-05, "loss": 1.1375389099121094, "step": 37040 }, { "epoch": 3.1572219855134214, "grad_norm": 2.59375, "learning_rate": 4.879428449268012e-05, "loss": 1.1819942474365235, "step": 37050 }, { "epoch": 3.158074137196421, "grad_norm": 2.671875, "learning_rate": 4.8793437582331235e-05, "loss": 1.3133360862731933, "step": 37060 }, { "epoch": 3.1589262888794205, "grad_norm": 2.578125, "learning_rate": 4.879259038200135e-05, "loss": 1.3947318077087403, "step": 37070 }, { "epoch": 3.15977844056242, "grad_norm": 2.515625, "learning_rate": 4.879174289170078e-05, "loss": 1.3186732292175294, "step": 37080 }, { "epoch": 3.1606305922454196, "grad_norm": 2.109375, "learning_rate": 4.879089511143986e-05, "loss": 1.2697568893432618, "step": 37090 }, { "epoch": 3.161482743928419, "grad_norm": 5.125, "learning_rate": 4.8790047041228916e-05, "loss": 1.3833799362182617, "step": 37100 }, { "epoch": 3.162334895611419, "grad_norm": 3.078125, "learning_rate": 4.878919868107829e-05, "loss": 1.2977768898010253, "step": 37110 }, { "epoch": 3.1631870472944184, "grad_norm": 2.40625, "learning_rate": 4.878835003099832e-05, "loss": 1.3383336067199707, "step": 37120 }, { "epoch": 3.164039198977418, "grad_norm": 1.4140625, "learning_rate": 4.8787501090999354e-05, "loss": 1.5112201690673828, "step": 37130 }, { "epoch": 3.1648913506604175, "grad_norm": 1.9140625, "learning_rate": 4.878665186109173e-05, "loss": 1.4105868339538574, "step": 37140 }, { "epoch": 3.165743502343417, "grad_norm": 2.203125, "learning_rate": 4.878580234128579e-05, "loss": 1.4701683044433593, "step": 37150 }, { "epoch": 3.1665956540264166, "grad_norm": 1.765625, "learning_rate": 4.878495253159191e-05, "loss": 1.3236791610717773, "step": 37160 }, { "epoch": 3.167447805709416, "grad_norm": 2.578125, "learning_rate": 4.878410243202043e-05, "loss": 1.3383379936218263, "step": 37170 }, { "epoch": 3.1682999573924158, "grad_norm": 2.203125, "learning_rate": 4.8783252042581715e-05, "loss": 1.278878116607666, "step": 37180 }, { "epoch": 3.1691521090754153, "grad_norm": 1.859375, "learning_rate": 4.878240136328613e-05, "loss": 1.2710884094238282, "step": 37190 }, { "epoch": 3.170004260758415, "grad_norm": 2.09375, "learning_rate": 4.8781550394144045e-05, "loss": 1.3545422554016113, "step": 37200 }, { "epoch": 3.1708564124414145, "grad_norm": 2.453125, "learning_rate": 4.878069913516583e-05, "loss": 1.3847229957580567, "step": 37210 }, { "epoch": 3.171708564124414, "grad_norm": 2.734375, "learning_rate": 4.877984758636185e-05, "loss": 1.2992165565490723, "step": 37220 }, { "epoch": 3.1725607158074136, "grad_norm": 2.625, "learning_rate": 4.877899574774249e-05, "loss": 1.3041952133178711, "step": 37230 }, { "epoch": 3.173412867490413, "grad_norm": 3.0625, "learning_rate": 4.8778143619318145e-05, "loss": 1.4511687278747558, "step": 37240 }, { "epoch": 3.1742650191734127, "grad_norm": 2.75, "learning_rate": 4.877729120109919e-05, "loss": 1.4999252319335938, "step": 37250 }, { "epoch": 3.1751171708564123, "grad_norm": 2.34375, "learning_rate": 4.8776438493095995e-05, "loss": 1.4200756072998046, "step": 37260 }, { "epoch": 3.175969322539412, "grad_norm": 2.703125, "learning_rate": 4.8775585495318975e-05, "loss": 1.1272132873535157, "step": 37270 }, { "epoch": 3.1768214742224115, "grad_norm": 2.8125, "learning_rate": 4.877473220777853e-05, "loss": 1.3737643241882325, "step": 37280 }, { "epoch": 3.177673625905411, "grad_norm": 1.9609375, "learning_rate": 4.8773878630485034e-05, "loss": 1.3754463195800781, "step": 37290 }, { "epoch": 3.1785257775884106, "grad_norm": 3.109375, "learning_rate": 4.877302476344892e-05, "loss": 1.3223101615905761, "step": 37300 }, { "epoch": 3.17937792927141, "grad_norm": 2.1875, "learning_rate": 4.877217060668058e-05, "loss": 1.4440459251403808, "step": 37310 }, { "epoch": 3.1802300809544097, "grad_norm": 3.125, "learning_rate": 4.877131616019042e-05, "loss": 1.1376382827758789, "step": 37320 }, { "epoch": 3.1810822326374093, "grad_norm": 2.265625, "learning_rate": 4.877046142398885e-05, "loss": 1.3424410820007324, "step": 37330 }, { "epoch": 3.181934384320409, "grad_norm": 2.1875, "learning_rate": 4.87696063980863e-05, "loss": 1.3271432876586915, "step": 37340 }, { "epoch": 3.1827865360034084, "grad_norm": 2.125, "learning_rate": 4.876875108249319e-05, "loss": 1.230632209777832, "step": 37350 }, { "epoch": 3.183638687686408, "grad_norm": 3.234375, "learning_rate": 4.876789547721993e-05, "loss": 1.3184942245483398, "step": 37360 }, { "epoch": 3.1844908393694076, "grad_norm": 1.7734375, "learning_rate": 4.8767039582276966e-05, "loss": 1.4084250450134277, "step": 37370 }, { "epoch": 3.185342991052407, "grad_norm": 2.140625, "learning_rate": 4.876618339767471e-05, "loss": 1.3161922454833985, "step": 37380 }, { "epoch": 3.1861951427354067, "grad_norm": 2.3125, "learning_rate": 4.876532692342361e-05, "loss": 1.4230117797851562, "step": 37390 }, { "epoch": 3.1870472944184063, "grad_norm": 2.65625, "learning_rate": 4.876447015953411e-05, "loss": 1.3349637031555175, "step": 37400 }, { "epoch": 3.187899446101406, "grad_norm": 2.125, "learning_rate": 4.876361310601664e-05, "loss": 1.3570043563842773, "step": 37410 }, { "epoch": 3.1887515977844054, "grad_norm": 1.984375, "learning_rate": 4.876275576288164e-05, "loss": 1.3085145950317383, "step": 37420 }, { "epoch": 3.189603749467405, "grad_norm": 2.5, "learning_rate": 4.876189813013957e-05, "loss": 1.3984167098999023, "step": 37430 }, { "epoch": 3.190455901150405, "grad_norm": 3.421875, "learning_rate": 4.876104020780088e-05, "loss": 1.4718717575073241, "step": 37440 }, { "epoch": 3.1913080528334046, "grad_norm": 1.9296875, "learning_rate": 4.8760181995876026e-05, "loss": 1.4567525863647461, "step": 37450 }, { "epoch": 3.192160204516404, "grad_norm": 2.5, "learning_rate": 4.8759323494375464e-05, "loss": 1.411414337158203, "step": 37460 }, { "epoch": 3.1930123561994037, "grad_norm": 2.921875, "learning_rate": 4.8758464703309656e-05, "loss": 1.339974594116211, "step": 37470 }, { "epoch": 3.1938645078824033, "grad_norm": 2.390625, "learning_rate": 4.8757605622689084e-05, "loss": 1.4057301521301269, "step": 37480 }, { "epoch": 3.194716659565403, "grad_norm": 2.84375, "learning_rate": 4.875674625252419e-05, "loss": 1.3613417625427247, "step": 37490 }, { "epoch": 3.1955688112484024, "grad_norm": 2.203125, "learning_rate": 4.875588659282547e-05, "loss": 1.2609683990478515, "step": 37500 }, { "epoch": 3.196420962931402, "grad_norm": 2.828125, "learning_rate": 4.8755026643603404e-05, "loss": 1.201162052154541, "step": 37510 }, { "epoch": 3.1972731146144016, "grad_norm": 1.9453125, "learning_rate": 4.875416640486845e-05, "loss": 1.4950328826904298, "step": 37520 }, { "epoch": 3.198125266297401, "grad_norm": 2.71875, "learning_rate": 4.875330587663111e-05, "loss": 1.3748868942260741, "step": 37530 }, { "epoch": 3.1989774179804007, "grad_norm": 2.671875, "learning_rate": 4.8752445058901866e-05, "loss": 1.2886775016784668, "step": 37540 }, { "epoch": 3.1998295696634003, "grad_norm": 3.53125, "learning_rate": 4.875158395169121e-05, "loss": 1.3448116302490234, "step": 37550 }, { "epoch": 3.2006817213464, "grad_norm": 3.25, "learning_rate": 4.875072255500964e-05, "loss": 1.3136269569396972, "step": 37560 }, { "epoch": 3.2015338730293994, "grad_norm": 1.953125, "learning_rate": 4.874986086886765e-05, "loss": 1.4987966537475585, "step": 37570 }, { "epoch": 3.202386024712399, "grad_norm": 2.375, "learning_rate": 4.874899889327573e-05, "loss": 1.3435497283935547, "step": 37580 }, { "epoch": 3.2032381763953985, "grad_norm": 2.34375, "learning_rate": 4.874813662824441e-05, "loss": 1.3881525993347168, "step": 37590 }, { "epoch": 3.204090328078398, "grad_norm": 2.0625, "learning_rate": 4.874727407378419e-05, "loss": 1.3980865478515625, "step": 37600 }, { "epoch": 3.2049424797613977, "grad_norm": 2.65625, "learning_rate": 4.874641122990557e-05, "loss": 1.2727723121643066, "step": 37610 }, { "epoch": 3.2057946314443972, "grad_norm": 2.03125, "learning_rate": 4.874554809661908e-05, "loss": 1.3148096084594727, "step": 37620 }, { "epoch": 3.206646783127397, "grad_norm": 1.7421875, "learning_rate": 4.874468467393522e-05, "loss": 1.3960391998291015, "step": 37630 }, { "epoch": 3.2074989348103964, "grad_norm": 3.609375, "learning_rate": 4.874382096186455e-05, "loss": 1.3687189102172852, "step": 37640 }, { "epoch": 3.208351086493396, "grad_norm": 2.890625, "learning_rate": 4.874295696041756e-05, "loss": 1.3963208198547363, "step": 37650 }, { "epoch": 3.2092032381763955, "grad_norm": 2.046875, "learning_rate": 4.874209266960479e-05, "loss": 1.4333647727966308, "step": 37660 }, { "epoch": 3.210055389859395, "grad_norm": 2.03125, "learning_rate": 4.874122808943678e-05, "loss": 1.4089818954467774, "step": 37670 }, { "epoch": 3.2109075415423947, "grad_norm": 1.75, "learning_rate": 4.8740363219924065e-05, "loss": 1.4870546340942383, "step": 37680 }, { "epoch": 3.2117596932253942, "grad_norm": 2.375, "learning_rate": 4.873949806107719e-05, "loss": 1.3106395721435546, "step": 37690 }, { "epoch": 3.212611844908394, "grad_norm": 3.3125, "learning_rate": 4.873863261290669e-05, "loss": 1.1951683044433594, "step": 37700 }, { "epoch": 3.2134639965913934, "grad_norm": 2.484375, "learning_rate": 4.873776687542312e-05, "loss": 1.5004122734069825, "step": 37710 }, { "epoch": 3.214316148274393, "grad_norm": 2.578125, "learning_rate": 4.873690084863702e-05, "loss": 1.1506011962890625, "step": 37720 }, { "epoch": 3.2151682999573925, "grad_norm": 3.109375, "learning_rate": 4.873603453255896e-05, "loss": 1.3822434425354004, "step": 37730 }, { "epoch": 3.216020451640392, "grad_norm": 2.296875, "learning_rate": 4.8735167927199476e-05, "loss": 1.3968093872070313, "step": 37740 }, { "epoch": 3.2168726033233916, "grad_norm": 2.09375, "learning_rate": 4.873430103256915e-05, "loss": 1.488774299621582, "step": 37750 }, { "epoch": 3.217724755006391, "grad_norm": 2.859375, "learning_rate": 4.873343384867854e-05, "loss": 1.4267200469970702, "step": 37760 }, { "epoch": 3.218576906689391, "grad_norm": 2.21875, "learning_rate": 4.8732566375538226e-05, "loss": 1.1799273490905762, "step": 37770 }, { "epoch": 3.2194290583723904, "grad_norm": 3.09375, "learning_rate": 4.873169861315876e-05, "loss": 1.2352582931518554, "step": 37780 }, { "epoch": 3.22028121005539, "grad_norm": 2.40625, "learning_rate": 4.873083056155073e-05, "loss": 1.2806205749511719, "step": 37790 }, { "epoch": 3.2211333617383895, "grad_norm": 2.984375, "learning_rate": 4.8729962220724716e-05, "loss": 1.4318333625793458, "step": 37800 }, { "epoch": 3.221985513421389, "grad_norm": 1.734375, "learning_rate": 4.87290935906913e-05, "loss": 1.3347444534301758, "step": 37810 }, { "epoch": 3.2228376651043886, "grad_norm": 1.7578125, "learning_rate": 4.872822467146107e-05, "loss": 1.2981903076171875, "step": 37820 }, { "epoch": 3.223689816787388, "grad_norm": 2.03125, "learning_rate": 4.87273554630446e-05, "loss": 1.4160927772521972, "step": 37830 }, { "epoch": 3.2245419684703878, "grad_norm": 2.953125, "learning_rate": 4.87264859654525e-05, "loss": 1.2375325202941894, "step": 37840 }, { "epoch": 3.2253941201533873, "grad_norm": 1.6015625, "learning_rate": 4.872561617869535e-05, "loss": 1.2864083290100097, "step": 37850 }, { "epoch": 3.226246271836387, "grad_norm": 2.265625, "learning_rate": 4.872474610278378e-05, "loss": 1.2815088272094726, "step": 37860 }, { "epoch": 3.2270984235193865, "grad_norm": 2.5625, "learning_rate": 4.8723875737728375e-05, "loss": 1.416421890258789, "step": 37870 }, { "epoch": 3.227950575202386, "grad_norm": 2.859375, "learning_rate": 4.8723005083539744e-05, "loss": 1.261652374267578, "step": 37880 }, { "epoch": 3.2288027268853856, "grad_norm": 2.140625, "learning_rate": 4.87221341402285e-05, "loss": 1.4510600090026855, "step": 37890 }, { "epoch": 3.229654878568385, "grad_norm": 2.625, "learning_rate": 4.872126290780526e-05, "loss": 1.2599230766296388, "step": 37900 }, { "epoch": 3.2305070302513847, "grad_norm": 1.8203125, "learning_rate": 4.8720391386280625e-05, "loss": 1.303391933441162, "step": 37910 }, { "epoch": 3.2313591819343843, "grad_norm": 2.421875, "learning_rate": 4.871951957566523e-05, "loss": 1.379041862487793, "step": 37920 }, { "epoch": 3.232211333617384, "grad_norm": 2.96875, "learning_rate": 4.8718647475969714e-05, "loss": 1.4442954063415527, "step": 37930 }, { "epoch": 3.2330634853003835, "grad_norm": 1.8203125, "learning_rate": 4.871777508720468e-05, "loss": 1.405240249633789, "step": 37940 }, { "epoch": 3.233915636983383, "grad_norm": 1.4609375, "learning_rate": 4.8716902409380774e-05, "loss": 1.1842500686645507, "step": 37950 }, { "epoch": 3.2347677886663826, "grad_norm": 2.78125, "learning_rate": 4.871602944250864e-05, "loss": 1.3317380905151368, "step": 37960 }, { "epoch": 3.235619940349382, "grad_norm": 1.8984375, "learning_rate": 4.8715156186598895e-05, "loss": 1.4652162551879884, "step": 37970 }, { "epoch": 3.2364720920323817, "grad_norm": 2.234375, "learning_rate": 4.871428264166219e-05, "loss": 1.2117559432983398, "step": 37980 }, { "epoch": 3.2373242437153813, "grad_norm": 2.078125, "learning_rate": 4.871340880770919e-05, "loss": 1.446175479888916, "step": 37990 }, { "epoch": 3.238176395398381, "grad_norm": 2.890625, "learning_rate": 4.871253468475052e-05, "loss": 1.325773334503174, "step": 38000 }, { "epoch": 3.2390285470813804, "grad_norm": 2.1875, "learning_rate": 4.871166027279684e-05, "loss": 1.2249393463134766, "step": 38010 }, { "epoch": 3.23988069876438, "grad_norm": 3.421875, "learning_rate": 4.871078557185882e-05, "loss": 1.377942180633545, "step": 38020 }, { "epoch": 3.2407328504473796, "grad_norm": 2.0, "learning_rate": 4.87099105819471e-05, "loss": 1.408407211303711, "step": 38030 }, { "epoch": 3.241585002130379, "grad_norm": 4.5625, "learning_rate": 4.870903530307236e-05, "loss": 1.3859939575195312, "step": 38040 }, { "epoch": 3.2424371538133787, "grad_norm": 2.4375, "learning_rate": 4.870815973524526e-05, "loss": 1.2227973937988281, "step": 38050 }, { "epoch": 3.2432893054963783, "grad_norm": 2.796875, "learning_rate": 4.870728387847647e-05, "loss": 1.4326347351074218, "step": 38060 }, { "epoch": 3.244141457179378, "grad_norm": 2.625, "learning_rate": 4.870640773277667e-05, "loss": 1.2572456359863282, "step": 38070 }, { "epoch": 3.2449936088623774, "grad_norm": 2.8125, "learning_rate": 4.8705531298156534e-05, "loss": 1.2440552711486816, "step": 38080 }, { "epoch": 3.245845760545377, "grad_norm": 6.25, "learning_rate": 4.870465457462675e-05, "loss": 1.310968017578125, "step": 38090 }, { "epoch": 3.2466979122283766, "grad_norm": 4.0, "learning_rate": 4.870377756219798e-05, "loss": 1.4077493667602539, "step": 38100 }, { "epoch": 3.247550063911376, "grad_norm": 1.8671875, "learning_rate": 4.870290026088095e-05, "loss": 1.4647851943969727, "step": 38110 }, { "epoch": 3.2484022155943757, "grad_norm": 2.75, "learning_rate": 4.870202267068632e-05, "loss": 1.3276348114013672, "step": 38120 }, { "epoch": 3.2492543672773753, "grad_norm": 2.203125, "learning_rate": 4.8701144791624796e-05, "loss": 1.4025150299072267, "step": 38130 }, { "epoch": 3.250106518960375, "grad_norm": 2.84375, "learning_rate": 4.870026662370709e-05, "loss": 1.326645851135254, "step": 38140 }, { "epoch": 3.2509586706433744, "grad_norm": 2.6875, "learning_rate": 4.869938816694388e-05, "loss": 1.3017867088317872, "step": 38150 }, { "epoch": 3.251810822326374, "grad_norm": 2.84375, "learning_rate": 4.8698509421345893e-05, "loss": 1.2450045585632323, "step": 38160 }, { "epoch": 3.2526629740093735, "grad_norm": 2.0, "learning_rate": 4.869763038692383e-05, "loss": 1.404045009613037, "step": 38170 }, { "epoch": 3.253515125692373, "grad_norm": 2.359375, "learning_rate": 4.8696751063688404e-05, "loss": 1.2480619430541993, "step": 38180 }, { "epoch": 3.2543672773753727, "grad_norm": 2.53125, "learning_rate": 4.869587145165034e-05, "loss": 1.3652132987976073, "step": 38190 }, { "epoch": 3.2552194290583722, "grad_norm": 1.8359375, "learning_rate": 4.8694991550820344e-05, "loss": 1.434619140625, "step": 38200 }, { "epoch": 3.256071580741372, "grad_norm": 1.953125, "learning_rate": 4.869411136120915e-05, "loss": 1.4093022346496582, "step": 38210 }, { "epoch": 3.2569237324243714, "grad_norm": 2.28125, "learning_rate": 4.869323088282748e-05, "loss": 1.3846327781677246, "step": 38220 }, { "epoch": 3.257775884107371, "grad_norm": 2.1875, "learning_rate": 4.869235011568606e-05, "loss": 1.356593132019043, "step": 38230 }, { "epoch": 3.2586280357903705, "grad_norm": 2.015625, "learning_rate": 4.8691469059795635e-05, "loss": 1.3596827507019043, "step": 38240 }, { "epoch": 3.25948018747337, "grad_norm": 2.984375, "learning_rate": 4.869058771516695e-05, "loss": 1.3243318557739259, "step": 38250 }, { "epoch": 3.2603323391563697, "grad_norm": 3.28125, "learning_rate": 4.8689706081810723e-05, "loss": 1.4585325241088867, "step": 38260 }, { "epoch": 3.2611844908393692, "grad_norm": 2.65625, "learning_rate": 4.8688824159737715e-05, "loss": 1.3033634185791017, "step": 38270 }, { "epoch": 3.262036642522369, "grad_norm": 1.8046875, "learning_rate": 4.868794194895867e-05, "loss": 1.3665285110473633, "step": 38280 }, { "epoch": 3.2628887942053684, "grad_norm": 2.359375, "learning_rate": 4.868705944948434e-05, "loss": 1.4202938079833984, "step": 38290 }, { "epoch": 3.263740945888368, "grad_norm": 1.9609375, "learning_rate": 4.8686176661325476e-05, "loss": 1.3026198387145995, "step": 38300 }, { "epoch": 3.2645930975713675, "grad_norm": 2.03125, "learning_rate": 4.868529358449285e-05, "loss": 1.3613317489624024, "step": 38310 }, { "epoch": 3.265445249254367, "grad_norm": 1.7421875, "learning_rate": 4.868441021899721e-05, "loss": 1.576756191253662, "step": 38320 }, { "epoch": 3.2662974009373666, "grad_norm": 2.25, "learning_rate": 4.8683526564849335e-05, "loss": 1.3910984992980957, "step": 38330 }, { "epoch": 3.267149552620366, "grad_norm": 2.46875, "learning_rate": 4.868264262205998e-05, "loss": 1.394943618774414, "step": 38340 }, { "epoch": 3.268001704303366, "grad_norm": 1.9765625, "learning_rate": 4.868175839063993e-05, "loss": 1.4393539428710938, "step": 38350 }, { "epoch": 3.2688538559863654, "grad_norm": 2.390625, "learning_rate": 4.8680873870599955e-05, "loss": 1.292935848236084, "step": 38360 }, { "epoch": 3.269706007669365, "grad_norm": 4.90625, "learning_rate": 4.867998906195084e-05, "loss": 1.4123257637023925, "step": 38370 }, { "epoch": 3.2705581593523645, "grad_norm": 1.75, "learning_rate": 4.867910396470337e-05, "loss": 1.432360553741455, "step": 38380 }, { "epoch": 3.271410311035364, "grad_norm": 2.734375, "learning_rate": 4.867821857886832e-05, "loss": 1.352637481689453, "step": 38390 }, { "epoch": 3.2722624627183636, "grad_norm": 2.75, "learning_rate": 4.8677332904456485e-05, "loss": 1.338908576965332, "step": 38400 }, { "epoch": 3.273114614401363, "grad_norm": 2.640625, "learning_rate": 4.867644694147867e-05, "loss": 1.4696160316467286, "step": 38410 }, { "epoch": 3.2739667660843628, "grad_norm": 2.03125, "learning_rate": 4.867556068994566e-05, "loss": 1.3620471954345703, "step": 38420 }, { "epoch": 3.274818917767363, "grad_norm": 1.8203125, "learning_rate": 4.867467414986826e-05, "loss": 1.35589542388916, "step": 38430 }, { "epoch": 3.2756710694503623, "grad_norm": 2.765625, "learning_rate": 4.867378732125729e-05, "loss": 1.340815258026123, "step": 38440 }, { "epoch": 3.276523221133362, "grad_norm": 2.046875, "learning_rate": 4.867290020412353e-05, "loss": 1.3044601440429688, "step": 38450 }, { "epoch": 3.2773753728163615, "grad_norm": 2.09375, "learning_rate": 4.8672012798477804e-05, "loss": 1.3594209671020507, "step": 38460 }, { "epoch": 3.278227524499361, "grad_norm": 2.984375, "learning_rate": 4.8671125104330936e-05, "loss": 1.3378172874450684, "step": 38470 }, { "epoch": 3.2790796761823606, "grad_norm": 2.453125, "learning_rate": 4.867023712169374e-05, "loss": 1.3302122116088868, "step": 38480 }, { "epoch": 3.27993182786536, "grad_norm": 2.34375, "learning_rate": 4.866934885057703e-05, "loss": 1.3129630088806152, "step": 38490 }, { "epoch": 3.2807839795483598, "grad_norm": 2.578125, "learning_rate": 4.866846029099164e-05, "loss": 1.4444117546081543, "step": 38500 }, { "epoch": 3.2816361312313593, "grad_norm": 2.59375, "learning_rate": 4.8667571442948394e-05, "loss": 1.4223971366882324, "step": 38510 }, { "epoch": 3.282488282914359, "grad_norm": 1.75, "learning_rate": 4.866668230645814e-05, "loss": 1.4142060279846191, "step": 38520 }, { "epoch": 3.2833404345973585, "grad_norm": 2.328125, "learning_rate": 4.866579288153168e-05, "loss": 1.324657917022705, "step": 38530 }, { "epoch": 3.284192586280358, "grad_norm": 2.90625, "learning_rate": 4.8664903168179886e-05, "loss": 1.4191319465637207, "step": 38540 }, { "epoch": 3.2850447379633576, "grad_norm": 2.609375, "learning_rate": 4.866401316641359e-05, "loss": 1.3413072586059571, "step": 38550 }, { "epoch": 3.285896889646357, "grad_norm": 3.234375, "learning_rate": 4.866312287624365e-05, "loss": 1.3494996070861816, "step": 38560 }, { "epoch": 3.2867490413293567, "grad_norm": 2.03125, "learning_rate": 4.86622322976809e-05, "loss": 1.4051873207092285, "step": 38570 }, { "epoch": 3.2876011930123563, "grad_norm": 3.78125, "learning_rate": 4.8661341430736196e-05, "loss": 1.343994140625, "step": 38580 }, { "epoch": 3.288453344695356, "grad_norm": 2.609375, "learning_rate": 4.86604502754204e-05, "loss": 1.372523593902588, "step": 38590 }, { "epoch": 3.2893054963783555, "grad_norm": 2.15625, "learning_rate": 4.865955883174437e-05, "loss": 1.3783529281616211, "step": 38600 }, { "epoch": 3.290157648061355, "grad_norm": 2.234375, "learning_rate": 4.865866709971898e-05, "loss": 1.3274795532226562, "step": 38610 }, { "epoch": 3.2910097997443546, "grad_norm": 2.046875, "learning_rate": 4.8657775079355087e-05, "loss": 1.3304560661315918, "step": 38620 }, { "epoch": 3.291861951427354, "grad_norm": 2.21875, "learning_rate": 4.865688277066357e-05, "loss": 1.3682233810424804, "step": 38630 }, { "epoch": 3.2927141031103537, "grad_norm": 2.109375, "learning_rate": 4.8655990173655296e-05, "loss": 1.2513465881347656, "step": 38640 }, { "epoch": 3.2935662547933533, "grad_norm": 3.265625, "learning_rate": 4.865509728834115e-05, "loss": 1.2948419570922851, "step": 38650 }, { "epoch": 3.294418406476353, "grad_norm": 2.71875, "learning_rate": 4.8654204114732005e-05, "loss": 1.2720263481140137, "step": 38660 }, { "epoch": 3.2952705581593524, "grad_norm": 1.8828125, "learning_rate": 4.8653310652838756e-05, "loss": 1.3978082656860351, "step": 38670 }, { "epoch": 3.296122709842352, "grad_norm": 2.375, "learning_rate": 4.865241690267229e-05, "loss": 1.4126497268676759, "step": 38680 }, { "epoch": 3.2969748615253516, "grad_norm": 2.640625, "learning_rate": 4.8651522864243505e-05, "loss": 1.3265663146972657, "step": 38690 }, { "epoch": 3.297827013208351, "grad_norm": 1.6640625, "learning_rate": 4.865062853756328e-05, "loss": 1.298893356323242, "step": 38700 }, { "epoch": 3.2986791648913507, "grad_norm": 2.46875, "learning_rate": 4.8649733922642537e-05, "loss": 1.341256046295166, "step": 38710 }, { "epoch": 3.2995313165743503, "grad_norm": 2.21875, "learning_rate": 4.864883901949216e-05, "loss": 1.348893928527832, "step": 38720 }, { "epoch": 3.30038346825735, "grad_norm": 2.765625, "learning_rate": 4.864794382812307e-05, "loss": 1.4355583190917969, "step": 38730 }, { "epoch": 3.3012356199403494, "grad_norm": 1.828125, "learning_rate": 4.864704834854615e-05, "loss": 1.3125903129577636, "step": 38740 }, { "epoch": 3.302087771623349, "grad_norm": 1.9296875, "learning_rate": 4.864615258077235e-05, "loss": 1.3571608543395997, "step": 38750 }, { "epoch": 3.3029399233063486, "grad_norm": 2.09375, "learning_rate": 4.8645256524812574e-05, "loss": 1.4200658798217773, "step": 38760 }, { "epoch": 3.303792074989348, "grad_norm": 1.78125, "learning_rate": 4.864436018067773e-05, "loss": 1.4369061470031739, "step": 38770 }, { "epoch": 3.3046442266723477, "grad_norm": 1.96875, "learning_rate": 4.8643463548378754e-05, "loss": 1.3389828681945801, "step": 38780 }, { "epoch": 3.3054963783553473, "grad_norm": 2.96875, "learning_rate": 4.864256662792657e-05, "loss": 1.3470221519470216, "step": 38790 }, { "epoch": 3.306348530038347, "grad_norm": 2.125, "learning_rate": 4.864166941933211e-05, "loss": 1.4178171157836914, "step": 38800 }, { "epoch": 3.3072006817213464, "grad_norm": 2.03125, "learning_rate": 4.864077192260631e-05, "loss": 1.337672519683838, "step": 38810 }, { "epoch": 3.308052833404346, "grad_norm": 1.9453125, "learning_rate": 4.863987413776011e-05, "loss": 1.3904806137084962, "step": 38820 }, { "epoch": 3.3089049850873455, "grad_norm": 2.046875, "learning_rate": 4.8638976064804445e-05, "loss": 1.5340843200683594, "step": 38830 }, { "epoch": 3.309757136770345, "grad_norm": 2.46875, "learning_rate": 4.8638077703750265e-05, "loss": 1.382928466796875, "step": 38840 }, { "epoch": 3.3106092884533447, "grad_norm": 2.625, "learning_rate": 4.863717905460852e-05, "loss": 1.3436033248901367, "step": 38850 }, { "epoch": 3.3114614401363442, "grad_norm": 1.8203125, "learning_rate": 4.863628011739015e-05, "loss": 1.40933198928833, "step": 38860 }, { "epoch": 3.312313591819344, "grad_norm": 2.75, "learning_rate": 4.8635380892106136e-05, "loss": 1.3248291015625, "step": 38870 }, { "epoch": 3.3131657435023434, "grad_norm": 2.96875, "learning_rate": 4.8634481378767416e-05, "loss": 1.369747257232666, "step": 38880 }, { "epoch": 3.314017895185343, "grad_norm": 2.015625, "learning_rate": 4.863358157738496e-05, "loss": 1.3278314590454101, "step": 38890 }, { "epoch": 3.3148700468683425, "grad_norm": 4.6875, "learning_rate": 4.8632681487969725e-05, "loss": 1.3588699340820312, "step": 38900 }, { "epoch": 3.315722198551342, "grad_norm": 2.015625, "learning_rate": 4.8631781110532696e-05, "loss": 1.2837469100952148, "step": 38910 }, { "epoch": 3.3165743502343417, "grad_norm": 3.171875, "learning_rate": 4.8630880445084834e-05, "loss": 1.3742031097412108, "step": 38920 }, { "epoch": 3.3174265019173412, "grad_norm": 2.75, "learning_rate": 4.862997949163712e-05, "loss": 1.411569881439209, "step": 38930 }, { "epoch": 3.318278653600341, "grad_norm": 4.5, "learning_rate": 4.862907825020054e-05, "loss": 1.4205636978149414, "step": 38940 }, { "epoch": 3.3191308052833404, "grad_norm": 1.9296875, "learning_rate": 4.862817672078607e-05, "loss": 1.4205321311950683, "step": 38950 }, { "epoch": 3.31998295696634, "grad_norm": 1.84375, "learning_rate": 4.862727490340471e-05, "loss": 1.3140591621398925, "step": 38960 }, { "epoch": 3.3208351086493395, "grad_norm": 2.5, "learning_rate": 4.8626372798067425e-05, "loss": 1.4283576965332032, "step": 38970 }, { "epoch": 3.321687260332339, "grad_norm": 1.7578125, "learning_rate": 4.8625470404785236e-05, "loss": 1.3601386070251464, "step": 38980 }, { "epoch": 3.3225394120153386, "grad_norm": 1.9921875, "learning_rate": 4.8624567723569125e-05, "loss": 1.2795808792114258, "step": 38990 }, { "epoch": 3.323391563698338, "grad_norm": 3.234375, "learning_rate": 4.8623664754430096e-05, "loss": 1.397195053100586, "step": 39000 }, { "epoch": 3.324243715381338, "grad_norm": 1.7734375, "learning_rate": 4.862276149737917e-05, "loss": 1.5214794158935547, "step": 39010 }, { "epoch": 3.3250958670643374, "grad_norm": 1.7890625, "learning_rate": 4.862185795242733e-05, "loss": 1.4015429496765137, "step": 39020 }, { "epoch": 3.325948018747337, "grad_norm": 1.9375, "learning_rate": 4.8620954119585604e-05, "loss": 1.435603141784668, "step": 39030 }, { "epoch": 3.3268001704303365, "grad_norm": 1.921875, "learning_rate": 4.8620049998865005e-05, "loss": 1.2942501068115235, "step": 39040 }, { "epoch": 3.327652322113336, "grad_norm": 2.0, "learning_rate": 4.861914559027655e-05, "loss": 1.37823543548584, "step": 39050 }, { "epoch": 3.3285044737963356, "grad_norm": 2.984375, "learning_rate": 4.861824089383126e-05, "loss": 1.2795186042785645, "step": 39060 }, { "epoch": 3.329356625479335, "grad_norm": 3.328125, "learning_rate": 4.861733590954016e-05, "loss": 1.4197769165039062, "step": 39070 }, { "epoch": 3.3302087771623348, "grad_norm": 1.921875, "learning_rate": 4.861643063741428e-05, "loss": 1.3618787765502929, "step": 39080 }, { "epoch": 3.3310609288453343, "grad_norm": 1.9609375, "learning_rate": 4.8615525077464664e-05, "loss": 1.4310725212097168, "step": 39090 }, { "epoch": 3.331913080528334, "grad_norm": 2.03125, "learning_rate": 4.861461922970233e-05, "loss": 1.3489150047302245, "step": 39100 }, { "epoch": 3.3327652322113335, "grad_norm": 1.921875, "learning_rate": 4.8613713094138334e-05, "loss": 1.3459013938903808, "step": 39110 }, { "epoch": 3.333617383894333, "grad_norm": 2.140625, "learning_rate": 4.861280667078371e-05, "loss": 1.4347548484802246, "step": 39120 }, { "epoch": 3.3344695355773326, "grad_norm": 2.984375, "learning_rate": 4.8611899959649506e-05, "loss": 1.3298273086547852, "step": 39130 }, { "epoch": 3.335321687260332, "grad_norm": 3.796875, "learning_rate": 4.861099296074678e-05, "loss": 1.2511552810668944, "step": 39140 }, { "epoch": 3.3361738389433317, "grad_norm": 2.671875, "learning_rate": 4.8610085674086576e-05, "loss": 1.4885008811950684, "step": 39150 }, { "epoch": 3.3370259906263313, "grad_norm": 2.90625, "learning_rate": 4.860917809967995e-05, "loss": 1.104481029510498, "step": 39160 }, { "epoch": 3.337878142309331, "grad_norm": 1.9296875, "learning_rate": 4.860827023753798e-05, "loss": 1.2806968688964844, "step": 39170 }, { "epoch": 3.3387302939923305, "grad_norm": 1.9765625, "learning_rate": 4.860736208767171e-05, "loss": 1.3904571533203125, "step": 39180 }, { "epoch": 3.33958244567533, "grad_norm": 3.8125, "learning_rate": 4.860645365009223e-05, "loss": 1.2976606369018555, "step": 39190 }, { "epoch": 3.3404345973583296, "grad_norm": 2.71875, "learning_rate": 4.860554492481058e-05, "loss": 1.3206533432006835, "step": 39200 }, { "epoch": 3.341286749041329, "grad_norm": 2.484375, "learning_rate": 4.860463591183787e-05, "loss": 1.2654123306274414, "step": 39210 }, { "epoch": 3.342138900724329, "grad_norm": 1.78125, "learning_rate": 4.860372661118516e-05, "loss": 1.2261838912963867, "step": 39220 }, { "epoch": 3.3429910524073287, "grad_norm": 3.03125, "learning_rate": 4.8602817022863535e-05, "loss": 1.3316198348999024, "step": 39230 }, { "epoch": 3.3438432040903283, "grad_norm": 2.125, "learning_rate": 4.860190714688408e-05, "loss": 1.2966272354125976, "step": 39240 }, { "epoch": 3.344695355773328, "grad_norm": 4.90625, "learning_rate": 4.8600996983257875e-05, "loss": 1.3728171348571778, "step": 39250 }, { "epoch": 3.3455475074563275, "grad_norm": 1.625, "learning_rate": 4.860008653199603e-05, "loss": 1.3541955947875977, "step": 39260 }, { "epoch": 3.346399659139327, "grad_norm": 3.59375, "learning_rate": 4.859917579310962e-05, "loss": 1.3302648544311524, "step": 39270 }, { "epoch": 3.3472518108223266, "grad_norm": 2.125, "learning_rate": 4.859826476660977e-05, "loss": 1.3713592529296874, "step": 39280 }, { "epoch": 3.348103962505326, "grad_norm": 2.59375, "learning_rate": 4.859735345250757e-05, "loss": 1.289595127105713, "step": 39290 }, { "epoch": 3.3489561141883257, "grad_norm": 1.71875, "learning_rate": 4.859644185081412e-05, "loss": 1.5096066474914551, "step": 39300 }, { "epoch": 3.3498082658713253, "grad_norm": 4.09375, "learning_rate": 4.8595529961540545e-05, "loss": 1.3723441123962403, "step": 39310 }, { "epoch": 3.350660417554325, "grad_norm": 3.078125, "learning_rate": 4.8594617784697947e-05, "loss": 1.2135529518127441, "step": 39320 }, { "epoch": 3.3515125692373244, "grad_norm": 1.9453125, "learning_rate": 4.859370532029745e-05, "loss": 1.394557571411133, "step": 39330 }, { "epoch": 3.352364720920324, "grad_norm": 2.46875, "learning_rate": 4.859279256835017e-05, "loss": 1.4209647178649902, "step": 39340 }, { "epoch": 3.3532168726033236, "grad_norm": 2.234375, "learning_rate": 4.859187952886722e-05, "loss": 1.2475973129272462, "step": 39350 }, { "epoch": 3.354069024286323, "grad_norm": 3.296875, "learning_rate": 4.859096620185975e-05, "loss": 1.198745059967041, "step": 39360 }, { "epoch": 3.3549211759693227, "grad_norm": 2.625, "learning_rate": 4.859005258733888e-05, "loss": 1.334738540649414, "step": 39370 }, { "epoch": 3.3557733276523223, "grad_norm": 2.5, "learning_rate": 4.858913868531575e-05, "loss": 1.3451808929443358, "step": 39380 }, { "epoch": 3.356625479335322, "grad_norm": 2.96875, "learning_rate": 4.858822449580149e-05, "loss": 1.3281976699829101, "step": 39390 }, { "epoch": 3.3574776310183214, "grad_norm": 1.8125, "learning_rate": 4.8587310018807244e-05, "loss": 1.2980805397033692, "step": 39400 }, { "epoch": 3.358329782701321, "grad_norm": 3.546875, "learning_rate": 4.8586395254344166e-05, "loss": 1.2788269996643067, "step": 39410 }, { "epoch": 3.3591819343843206, "grad_norm": 2.359375, "learning_rate": 4.858548020242338e-05, "loss": 1.2841888427734376, "step": 39420 }, { "epoch": 3.36003408606732, "grad_norm": 1.7734375, "learning_rate": 4.858456486305607e-05, "loss": 1.4809195518493652, "step": 39430 }, { "epoch": 3.3608862377503197, "grad_norm": 4.8125, "learning_rate": 4.858364923625338e-05, "loss": 1.2583633422851563, "step": 39440 }, { "epoch": 3.3617383894333193, "grad_norm": 2.28125, "learning_rate": 4.858273332202646e-05, "loss": 1.2813624382019042, "step": 39450 }, { "epoch": 3.362590541116319, "grad_norm": 2.328125, "learning_rate": 4.858181712038647e-05, "loss": 1.3364014625549316, "step": 39460 }, { "epoch": 3.3634426927993184, "grad_norm": 3.109375, "learning_rate": 4.858090063134461e-05, "loss": 1.3255016326904296, "step": 39470 }, { "epoch": 3.364294844482318, "grad_norm": 2.671875, "learning_rate": 4.8579983854912006e-05, "loss": 1.2118115425109863, "step": 39480 }, { "epoch": 3.3651469961653175, "grad_norm": 2.140625, "learning_rate": 4.857906679109985e-05, "loss": 1.4030237197875977, "step": 39490 }, { "epoch": 3.365999147848317, "grad_norm": 1.6328125, "learning_rate": 4.8578149439919316e-05, "loss": 1.2850982666015625, "step": 39500 }, { "epoch": 3.3668512995313167, "grad_norm": 1.65625, "learning_rate": 4.8577231801381596e-05, "loss": 1.4785113334655762, "step": 39510 }, { "epoch": 3.3677034512143162, "grad_norm": 2.453125, "learning_rate": 4.8576313875497855e-05, "loss": 1.3316116333007812, "step": 39520 }, { "epoch": 3.368555602897316, "grad_norm": 3.359375, "learning_rate": 4.85753956622793e-05, "loss": 1.3314924240112305, "step": 39530 }, { "epoch": 3.3694077545803154, "grad_norm": 1.9765625, "learning_rate": 4.85744771617371e-05, "loss": 1.346694278717041, "step": 39540 }, { "epoch": 3.370259906263315, "grad_norm": 2.40625, "learning_rate": 4.857355837388247e-05, "loss": 1.369229507446289, "step": 39550 }, { "epoch": 3.3711120579463145, "grad_norm": 2.1875, "learning_rate": 4.857263929872658e-05, "loss": 1.3152795791625977, "step": 39560 }, { "epoch": 3.371964209629314, "grad_norm": 2.640625, "learning_rate": 4.857171993628067e-05, "loss": 1.2690362930297852, "step": 39570 }, { "epoch": 3.3728163613123137, "grad_norm": 1.890625, "learning_rate": 4.857080028655592e-05, "loss": 1.510484790802002, "step": 39580 }, { "epoch": 3.3736685129953132, "grad_norm": 1.875, "learning_rate": 4.8569880349563534e-05, "loss": 1.3804975509643556, "step": 39590 }, { "epoch": 3.374520664678313, "grad_norm": 1.734375, "learning_rate": 4.8568960125314736e-05, "loss": 1.2655303001403808, "step": 39600 }, { "epoch": 3.3753728163613124, "grad_norm": 1.859375, "learning_rate": 4.8568039613820735e-05, "loss": 1.298169994354248, "step": 39610 }, { "epoch": 3.376224968044312, "grad_norm": 2.859375, "learning_rate": 4.856711881509275e-05, "loss": 1.2969984054565429, "step": 39620 }, { "epoch": 3.3770771197273115, "grad_norm": 2.390625, "learning_rate": 4.8566197729142015e-05, "loss": 1.4190396308898925, "step": 39630 }, { "epoch": 3.377929271410311, "grad_norm": 1.75, "learning_rate": 4.856527635597974e-05, "loss": 1.4088056564331055, "step": 39640 }, { "epoch": 3.3787814230933106, "grad_norm": 1.96875, "learning_rate": 4.8564354695617154e-05, "loss": 1.4739412307739257, "step": 39650 }, { "epoch": 3.37963357477631, "grad_norm": 2.59375, "learning_rate": 4.856343274806549e-05, "loss": 1.4585338592529298, "step": 39660 }, { "epoch": 3.38048572645931, "grad_norm": 2.546875, "learning_rate": 4.8562510513336e-05, "loss": 1.355421543121338, "step": 39670 }, { "epoch": 3.3813378781423093, "grad_norm": 2.53125, "learning_rate": 4.856158799143992e-05, "loss": 1.4885725021362304, "step": 39680 }, { "epoch": 3.382190029825309, "grad_norm": 1.9375, "learning_rate": 4.856066518238847e-05, "loss": 1.3602675437927245, "step": 39690 }, { "epoch": 3.3830421815083085, "grad_norm": 2.34375, "learning_rate": 4.855974208619292e-05, "loss": 1.4589643478393555, "step": 39700 }, { "epoch": 3.383894333191308, "grad_norm": 2.671875, "learning_rate": 4.8558818702864513e-05, "loss": 1.4016035079956055, "step": 39710 }, { "epoch": 3.3847464848743076, "grad_norm": 1.859375, "learning_rate": 4.8557895032414505e-05, "loss": 1.4125651359558105, "step": 39720 }, { "epoch": 3.385598636557307, "grad_norm": 2.4375, "learning_rate": 4.855697107485414e-05, "loss": 1.3104982376098633, "step": 39730 }, { "epoch": 3.3864507882403068, "grad_norm": 2.671875, "learning_rate": 4.8556046830194705e-05, "loss": 1.3600683212280273, "step": 39740 }, { "epoch": 3.3873029399233063, "grad_norm": 1.6953125, "learning_rate": 4.855512229844744e-05, "loss": 1.4089716911315917, "step": 39750 }, { "epoch": 3.388155091606306, "grad_norm": 2.640625, "learning_rate": 4.855419747962362e-05, "loss": 1.2801012992858887, "step": 39760 }, { "epoch": 3.3890072432893055, "grad_norm": 2.671875, "learning_rate": 4.855327237373452e-05, "loss": 1.4323846817016601, "step": 39770 }, { "epoch": 3.389859394972305, "grad_norm": 1.734375, "learning_rate": 4.8552346980791415e-05, "loss": 1.3714764595031739, "step": 39780 }, { "epoch": 3.3907115466553046, "grad_norm": 2.46875, "learning_rate": 4.855142130080558e-05, "loss": 1.4308008193969726, "step": 39790 }, { "epoch": 3.391563698338304, "grad_norm": 2.265625, "learning_rate": 4.855049533378829e-05, "loss": 1.3897066116333008, "step": 39800 }, { "epoch": 3.3924158500213037, "grad_norm": 3.6875, "learning_rate": 4.8549569079750844e-05, "loss": 1.369035816192627, "step": 39810 }, { "epoch": 3.3932680017043033, "grad_norm": 2.40625, "learning_rate": 4.8548642538704524e-05, "loss": 1.2566556930541992, "step": 39820 }, { "epoch": 3.394120153387303, "grad_norm": 2.078125, "learning_rate": 4.854771571066062e-05, "loss": 1.3010330200195312, "step": 39830 }, { "epoch": 3.3949723050703025, "grad_norm": 2.15625, "learning_rate": 4.8546788595630426e-05, "loss": 1.429056167602539, "step": 39840 }, { "epoch": 3.395824456753302, "grad_norm": 2.46875, "learning_rate": 4.854586119362525e-05, "loss": 1.4249690055847168, "step": 39850 }, { "epoch": 3.3966766084363016, "grad_norm": 1.8125, "learning_rate": 4.854493350465639e-05, "loss": 1.431850051879883, "step": 39860 }, { "epoch": 3.397528760119301, "grad_norm": 2.359375, "learning_rate": 4.8544005528735145e-05, "loss": 1.3912315368652344, "step": 39870 }, { "epoch": 3.3983809118023007, "grad_norm": 2.65625, "learning_rate": 4.854307726587284e-05, "loss": 1.3268216133117676, "step": 39880 }, { "epoch": 3.3992330634853003, "grad_norm": 1.796875, "learning_rate": 4.8542148716080775e-05, "loss": 1.3252202033996583, "step": 39890 }, { "epoch": 3.4000852151683, "grad_norm": 3.5625, "learning_rate": 4.854121987937027e-05, "loss": 1.3211967468261718, "step": 39900 }, { "epoch": 3.4009373668512994, "grad_norm": 1.8046875, "learning_rate": 4.854029075575265e-05, "loss": 1.221929931640625, "step": 39910 }, { "epoch": 3.401789518534299, "grad_norm": 3.546875, "learning_rate": 4.853936134523923e-05, "loss": 1.141030502319336, "step": 39920 }, { "epoch": 3.4026416702172986, "grad_norm": 2.65625, "learning_rate": 4.8538431647841346e-05, "loss": 1.3692550659179688, "step": 39930 }, { "epoch": 3.403493821900298, "grad_norm": 2.71875, "learning_rate": 4.853750166357033e-05, "loss": 1.292520236968994, "step": 39940 }, { "epoch": 3.4043459735832977, "grad_norm": 3.0625, "learning_rate": 4.85365713924375e-05, "loss": 1.3041279792785645, "step": 39950 }, { "epoch": 3.4051981252662973, "grad_norm": 2.53125, "learning_rate": 4.853564083445421e-05, "loss": 1.410555076599121, "step": 39960 }, { "epoch": 3.406050276949297, "grad_norm": 3.9375, "learning_rate": 4.853470998963179e-05, "loss": 1.2687252044677735, "step": 39970 }, { "epoch": 3.4069024286322964, "grad_norm": 2.5625, "learning_rate": 4.853377885798159e-05, "loss": 1.4010028839111328, "step": 39980 }, { "epoch": 3.407754580315296, "grad_norm": 2.625, "learning_rate": 4.8532847439514974e-05, "loss": 1.4167840003967285, "step": 39990 }, { "epoch": 3.4086067319982956, "grad_norm": 2.125, "learning_rate": 4.853191573424326e-05, "loss": 1.3881454467773438, "step": 40000 }, { "epoch": 3.409458883681295, "grad_norm": 2.03125, "learning_rate": 4.8530983742177824e-05, "loss": 1.3497108459472655, "step": 40010 }, { "epoch": 3.4103110353642947, "grad_norm": 2.453125, "learning_rate": 4.853005146333003e-05, "loss": 1.360117530822754, "step": 40020 }, { "epoch": 3.4111631870472943, "grad_norm": 2.421875, "learning_rate": 4.8529118897711225e-05, "loss": 1.3710855484008788, "step": 40030 }, { "epoch": 3.412015338730294, "grad_norm": 2.140625, "learning_rate": 4.852818604533278e-05, "loss": 1.3655571937561035, "step": 40040 }, { "epoch": 3.4128674904132934, "grad_norm": 1.8515625, "learning_rate": 4.8527252906206073e-05, "loss": 1.373553466796875, "step": 40050 }, { "epoch": 3.413719642096293, "grad_norm": 1.609375, "learning_rate": 4.8526319480342466e-05, "loss": 1.2748381614685058, "step": 40060 }, { "epoch": 3.4145717937792925, "grad_norm": 1.609375, "learning_rate": 4.852538576775333e-05, "loss": 1.436319065093994, "step": 40070 }, { "epoch": 3.415423945462292, "grad_norm": 1.78125, "learning_rate": 4.852445176845006e-05, "loss": 1.4551606178283691, "step": 40080 }, { "epoch": 3.4162760971452917, "grad_norm": 1.9140625, "learning_rate": 4.8523517482444024e-05, "loss": 1.3450520515441895, "step": 40090 }, { "epoch": 3.4171282488282912, "grad_norm": 1.8359375, "learning_rate": 4.852258290974663e-05, "loss": 1.4995301246643067, "step": 40100 }, { "epoch": 3.417980400511291, "grad_norm": 1.578125, "learning_rate": 4.852164805036924e-05, "loss": 1.304497241973877, "step": 40110 }, { "epoch": 3.4188325521942904, "grad_norm": 2.125, "learning_rate": 4.852071290432327e-05, "loss": 1.3369053840637206, "step": 40120 }, { "epoch": 3.41968470387729, "grad_norm": 2.359375, "learning_rate": 4.851977747162011e-05, "loss": 1.4411649703979492, "step": 40130 }, { "epoch": 3.4205368555602895, "grad_norm": 4.6875, "learning_rate": 4.851884175227115e-05, "loss": 1.44900541305542, "step": 40140 }, { "epoch": 3.421389007243289, "grad_norm": 1.78125, "learning_rate": 4.851790574628781e-05, "loss": 1.3325454711914062, "step": 40150 }, { "epoch": 3.4222411589262887, "grad_norm": 2.4375, "learning_rate": 4.851696945368149e-05, "loss": 1.3345348358154296, "step": 40160 }, { "epoch": 3.4230933106092882, "grad_norm": 2.8125, "learning_rate": 4.85160328744636e-05, "loss": 1.2644622802734375, "step": 40170 }, { "epoch": 3.423945462292288, "grad_norm": 4.125, "learning_rate": 4.851509600864556e-05, "loss": 1.2494346618652343, "step": 40180 }, { "epoch": 3.4247976139752874, "grad_norm": 3.90625, "learning_rate": 4.851415885623878e-05, "loss": 1.3423727035522461, "step": 40190 }, { "epoch": 3.425649765658287, "grad_norm": 4.09375, "learning_rate": 4.851322141725468e-05, "loss": 1.3842280387878418, "step": 40200 }, { "epoch": 3.4265019173412865, "grad_norm": 2.3125, "learning_rate": 4.851228369170471e-05, "loss": 1.2891295433044434, "step": 40210 }, { "epoch": 3.427354069024286, "grad_norm": 1.890625, "learning_rate": 4.8511345679600264e-05, "loss": 1.3883962631225586, "step": 40220 }, { "epoch": 3.428206220707286, "grad_norm": 3.59375, "learning_rate": 4.8510407380952785e-05, "loss": 1.271664524078369, "step": 40230 }, { "epoch": 3.4290583723902857, "grad_norm": 1.7578125, "learning_rate": 4.8509468795773725e-05, "loss": 1.307417392730713, "step": 40240 }, { "epoch": 3.4299105240732852, "grad_norm": 1.9296875, "learning_rate": 4.850852992407451e-05, "loss": 1.2885856628417969, "step": 40250 }, { "epoch": 3.430762675756285, "grad_norm": 2.734375, "learning_rate": 4.850759076586657e-05, "loss": 1.365297222137451, "step": 40260 }, { "epoch": 3.4316148274392844, "grad_norm": 1.8359375, "learning_rate": 4.850665132116138e-05, "loss": 1.2804655075073241, "step": 40270 }, { "epoch": 3.432466979122284, "grad_norm": 4.125, "learning_rate": 4.850571158997036e-05, "loss": 1.1841197967529298, "step": 40280 }, { "epoch": 3.4333191308052835, "grad_norm": 2.109375, "learning_rate": 4.8504771572304984e-05, "loss": 1.4681398391723632, "step": 40290 }, { "epoch": 3.434171282488283, "grad_norm": 2.65625, "learning_rate": 4.8503831268176695e-05, "loss": 1.3165847778320312, "step": 40300 }, { "epoch": 3.4350234341712826, "grad_norm": 2.234375, "learning_rate": 4.850289067759697e-05, "loss": 1.3442906379699706, "step": 40310 }, { "epoch": 3.435875585854282, "grad_norm": 2.140625, "learning_rate": 4.8501949800577254e-05, "loss": 1.4254175186157227, "step": 40320 }, { "epoch": 3.436727737537282, "grad_norm": 2.71875, "learning_rate": 4.850100863712901e-05, "loss": 1.3996311187744142, "step": 40330 }, { "epoch": 3.4375798892202813, "grad_norm": 3.109375, "learning_rate": 4.8500067187263734e-05, "loss": 1.436886692047119, "step": 40340 }, { "epoch": 3.438432040903281, "grad_norm": 2.703125, "learning_rate": 4.849912545099288e-05, "loss": 1.2398691177368164, "step": 40350 }, { "epoch": 3.4392841925862805, "grad_norm": 3.0, "learning_rate": 4.849818342832793e-05, "loss": 1.292905902862549, "step": 40360 }, { "epoch": 3.44013634426928, "grad_norm": 2.171875, "learning_rate": 4.8497241119280365e-05, "loss": 1.311943244934082, "step": 40370 }, { "epoch": 3.4409884959522796, "grad_norm": 2.59375, "learning_rate": 4.8496298523861675e-05, "loss": 1.435760498046875, "step": 40380 }, { "epoch": 3.441840647635279, "grad_norm": 3.21875, "learning_rate": 4.849535564208334e-05, "loss": 1.181520652770996, "step": 40390 }, { "epoch": 3.4426927993182788, "grad_norm": 1.8828125, "learning_rate": 4.8494412473956846e-05, "loss": 1.3499327659606934, "step": 40400 }, { "epoch": 3.4435449510012783, "grad_norm": 3.28125, "learning_rate": 4.84934690194937e-05, "loss": 1.4376476287841797, "step": 40410 }, { "epoch": 3.444397102684278, "grad_norm": 2.046875, "learning_rate": 4.8492525278705395e-05, "loss": 1.4790916442871094, "step": 40420 }, { "epoch": 3.4452492543672775, "grad_norm": 2.484375, "learning_rate": 4.8491581251603435e-05, "loss": 1.2367383003234864, "step": 40430 }, { "epoch": 3.446101406050277, "grad_norm": 2.1875, "learning_rate": 4.849063693819933e-05, "loss": 1.354245662689209, "step": 40440 }, { "epoch": 3.4469535577332766, "grad_norm": 2.09375, "learning_rate": 4.848969233850458e-05, "loss": 1.45693998336792, "step": 40450 }, { "epoch": 3.447805709416276, "grad_norm": 2.265625, "learning_rate": 4.848874745253069e-05, "loss": 1.3308222770690918, "step": 40460 }, { "epoch": 3.4486578610992757, "grad_norm": 3.03125, "learning_rate": 4.84878022802892e-05, "loss": 1.3044435501098632, "step": 40470 }, { "epoch": 3.4495100127822753, "grad_norm": 1.9140625, "learning_rate": 4.84868568217916e-05, "loss": 1.1870835304260254, "step": 40480 }, { "epoch": 3.450362164465275, "grad_norm": 2.03125, "learning_rate": 4.848591107704944e-05, "loss": 1.3109541893005372, "step": 40490 }, { "epoch": 3.4512143161482745, "grad_norm": 2.109375, "learning_rate": 4.848496504607423e-05, "loss": 1.4095161437988282, "step": 40500 }, { "epoch": 3.452066467831274, "grad_norm": 4.125, "learning_rate": 4.84840187288775e-05, "loss": 1.299641227722168, "step": 40510 }, { "epoch": 3.4529186195142736, "grad_norm": 3.640625, "learning_rate": 4.848307212547079e-05, "loss": 1.3208306312561036, "step": 40520 }, { "epoch": 3.453770771197273, "grad_norm": 2.53125, "learning_rate": 4.8482125235865624e-05, "loss": 1.4293401718139649, "step": 40530 }, { "epoch": 3.4546229228802727, "grad_norm": 1.5390625, "learning_rate": 4.848117806007356e-05, "loss": 1.3296449661254883, "step": 40540 }, { "epoch": 3.4554750745632723, "grad_norm": 3.140625, "learning_rate": 4.848023059810614e-05, "loss": 1.2955324172973632, "step": 40550 }, { "epoch": 3.456327226246272, "grad_norm": 4.09375, "learning_rate": 4.847928284997489e-05, "loss": 1.428696346282959, "step": 40560 }, { "epoch": 3.4571793779292714, "grad_norm": 2.0625, "learning_rate": 4.8478334815691386e-05, "loss": 1.3100971221923827, "step": 40570 }, { "epoch": 3.458031529612271, "grad_norm": 2.765625, "learning_rate": 4.8477386495267164e-05, "loss": 1.3354778289794922, "step": 40580 }, { "epoch": 3.4588836812952706, "grad_norm": 1.8046875, "learning_rate": 4.847643788871379e-05, "loss": 1.361093807220459, "step": 40590 }, { "epoch": 3.45973583297827, "grad_norm": 2.25, "learning_rate": 4.847548899604282e-05, "loss": 1.4538472175598145, "step": 40600 }, { "epoch": 3.4605879846612697, "grad_norm": 2.953125, "learning_rate": 4.847453981726583e-05, "loss": 1.2926228523254395, "step": 40610 }, { "epoch": 3.4614401363442693, "grad_norm": 2.96875, "learning_rate": 4.847359035239438e-05, "loss": 1.1971658706665038, "step": 40620 }, { "epoch": 3.462292288027269, "grad_norm": 2.3125, "learning_rate": 4.847264060144003e-05, "loss": 1.4194217681884767, "step": 40630 }, { "epoch": 3.4631444397102684, "grad_norm": 3.53125, "learning_rate": 4.847169056441438e-05, "loss": 1.3102580070495606, "step": 40640 }, { "epoch": 3.463996591393268, "grad_norm": 1.59375, "learning_rate": 4.8470740241329e-05, "loss": 1.2338555335998536, "step": 40650 }, { "epoch": 3.4648487430762676, "grad_norm": 3.421875, "learning_rate": 4.846978963219545e-05, "loss": 1.259634304046631, "step": 40660 }, { "epoch": 3.465700894759267, "grad_norm": 3.453125, "learning_rate": 4.846883873702534e-05, "loss": 1.1922502517700195, "step": 40670 }, { "epoch": 3.4665530464422667, "grad_norm": 2.265625, "learning_rate": 4.846788755583025e-05, "loss": 1.317384624481201, "step": 40680 }, { "epoch": 3.4674051981252663, "grad_norm": 2.234375, "learning_rate": 4.8466936088621785e-05, "loss": 1.3019125938415528, "step": 40690 }, { "epoch": 3.468257349808266, "grad_norm": 3.484375, "learning_rate": 4.846598433541152e-05, "loss": 1.2630435943603515, "step": 40700 }, { "epoch": 3.4691095014912654, "grad_norm": 3.546875, "learning_rate": 4.8465032296211066e-05, "loss": 1.2766209602355958, "step": 40710 }, { "epoch": 3.469961653174265, "grad_norm": 2.765625, "learning_rate": 4.846407997103203e-05, "loss": 1.4000134468078613, "step": 40720 }, { "epoch": 3.4708138048572645, "grad_norm": 3.0625, "learning_rate": 4.8463127359886004e-05, "loss": 1.3739004135131836, "step": 40730 }, { "epoch": 3.471665956540264, "grad_norm": 2.453125, "learning_rate": 4.846217446278461e-05, "loss": 1.377812385559082, "step": 40740 }, { "epoch": 3.4725181082232637, "grad_norm": 3.34375, "learning_rate": 4.846122127973946e-05, "loss": 1.3869999885559081, "step": 40750 }, { "epoch": 3.4733702599062632, "grad_norm": 1.96875, "learning_rate": 4.8460267810762174e-05, "loss": 1.5018574714660644, "step": 40760 }, { "epoch": 3.474222411589263, "grad_norm": 6.125, "learning_rate": 4.845931405586436e-05, "loss": 1.387156867980957, "step": 40770 }, { "epoch": 3.4750745632722624, "grad_norm": 1.921875, "learning_rate": 4.8458360015057644e-05, "loss": 1.1853102684020995, "step": 40780 }, { "epoch": 3.475926714955262, "grad_norm": 1.8359375, "learning_rate": 4.845740568835367e-05, "loss": 1.4707314491271972, "step": 40790 }, { "epoch": 3.4767788666382615, "grad_norm": 2.0625, "learning_rate": 4.845645107576405e-05, "loss": 1.3520751953125, "step": 40800 }, { "epoch": 3.477631018321261, "grad_norm": 1.9140625, "learning_rate": 4.845549617730043e-05, "loss": 1.4046623229980468, "step": 40810 }, { "epoch": 3.4784831700042607, "grad_norm": 1.890625, "learning_rate": 4.8454540992974445e-05, "loss": 1.4266649246215821, "step": 40820 }, { "epoch": 3.4793353216872602, "grad_norm": 3.515625, "learning_rate": 4.8453585522797726e-05, "loss": 1.333553409576416, "step": 40830 }, { "epoch": 3.48018747337026, "grad_norm": 3.15625, "learning_rate": 4.845262976678194e-05, "loss": 1.2682405471801759, "step": 40840 }, { "epoch": 3.4810396250532594, "grad_norm": 2.15625, "learning_rate": 4.845167372493871e-05, "loss": 1.2414907455444335, "step": 40850 }, { "epoch": 3.481891776736259, "grad_norm": 1.546875, "learning_rate": 4.84507173972797e-05, "loss": 1.3369190216064453, "step": 40860 }, { "epoch": 3.4827439284192585, "grad_norm": 1.796875, "learning_rate": 4.844976078381657e-05, "loss": 1.4369956970214843, "step": 40870 }, { "epoch": 3.483596080102258, "grad_norm": 2.484375, "learning_rate": 4.844880388456098e-05, "loss": 1.3760685920715332, "step": 40880 }, { "epoch": 3.4844482317852576, "grad_norm": 2.34375, "learning_rate": 4.844784669952457e-05, "loss": 1.4528277397155762, "step": 40890 }, { "epoch": 3.485300383468257, "grad_norm": 3.546875, "learning_rate": 4.844688922871903e-05, "loss": 1.4328155517578125, "step": 40900 }, { "epoch": 3.486152535151257, "grad_norm": 2.3125, "learning_rate": 4.8445931472156014e-05, "loss": 1.2459001541137695, "step": 40910 }, { "epoch": 3.4870046868342564, "grad_norm": 3.28125, "learning_rate": 4.8444973429847205e-05, "loss": 1.35103759765625, "step": 40920 }, { "epoch": 3.487856838517256, "grad_norm": 1.6171875, "learning_rate": 4.844401510180427e-05, "loss": 1.3202455520629883, "step": 40930 }, { "epoch": 3.4887089902002555, "grad_norm": 2.578125, "learning_rate": 4.844305648803889e-05, "loss": 1.3433338165283204, "step": 40940 }, { "epoch": 3.489561141883255, "grad_norm": 1.9296875, "learning_rate": 4.844209758856277e-05, "loss": 1.1876734733581542, "step": 40950 }, { "epoch": 3.4904132935662546, "grad_norm": 1.890625, "learning_rate": 4.8441138403387566e-05, "loss": 1.346401309967041, "step": 40960 }, { "epoch": 3.491265445249254, "grad_norm": 3.03125, "learning_rate": 4.8440178932524976e-05, "loss": 1.3139327049255372, "step": 40970 }, { "epoch": 3.4921175969322538, "grad_norm": 4.15625, "learning_rate": 4.843921917598671e-05, "loss": 1.3676870346069336, "step": 40980 }, { "epoch": 3.4929697486152533, "grad_norm": 2.265625, "learning_rate": 4.843825913378444e-05, "loss": 1.3434715270996094, "step": 40990 }, { "epoch": 3.493821900298253, "grad_norm": 2.4375, "learning_rate": 4.843729880592987e-05, "loss": 1.28115873336792, "step": 41000 }, { "epoch": 3.4946740519812525, "grad_norm": 1.890625, "learning_rate": 4.843633819243473e-05, "loss": 1.344416618347168, "step": 41010 }, { "epoch": 3.4955262036642525, "grad_norm": 2.359375, "learning_rate": 4.843537729331071e-05, "loss": 1.3301483154296876, "step": 41020 }, { "epoch": 3.496378355347252, "grad_norm": 2.109375, "learning_rate": 4.8434416108569514e-05, "loss": 1.3576518058776856, "step": 41030 }, { "epoch": 3.4972305070302516, "grad_norm": 2.34375, "learning_rate": 4.843345463822286e-05, "loss": 1.4534064292907716, "step": 41040 }, { "epoch": 3.498082658713251, "grad_norm": 4.0625, "learning_rate": 4.843249288228247e-05, "loss": 1.2411938667297364, "step": 41050 }, { "epoch": 3.4989348103962508, "grad_norm": 2.0, "learning_rate": 4.843153084076007e-05, "loss": 1.4226545333862304, "step": 41060 }, { "epoch": 3.49970174691095, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6499, "eval_loss": 1.2019062042236328, "eval_runtime": 53.8816, "eval_samples_per_second": 366.767, "eval_steps_per_second": 22.939, "step": 41069 }, { "epoch": 3.4997869620792503, "grad_norm": 1.75, "learning_rate": 4.8430568513667385e-05, "loss": 1.3086910247802734, "step": 41070 }, { "epoch": 3.50063911376225, "grad_norm": 3.203125, "learning_rate": 4.8429605901016134e-05, "loss": 1.33707857131958, "step": 41080 }, { "epoch": 3.5014912654452495, "grad_norm": 3.515625, "learning_rate": 4.8428643002818055e-05, "loss": 1.4400192260742188, "step": 41090 }, { "epoch": 3.502343417128249, "grad_norm": 2.5625, "learning_rate": 4.842767981908487e-05, "loss": 1.2990756988525392, "step": 41100 }, { "epoch": 3.5031955688112486, "grad_norm": 2.6875, "learning_rate": 4.842671634982834e-05, "loss": 1.3859500885009766, "step": 41110 }, { "epoch": 3.504047720494248, "grad_norm": 2.21875, "learning_rate": 4.842575259506018e-05, "loss": 1.3022441864013672, "step": 41120 }, { "epoch": 3.5048998721772477, "grad_norm": 3.0625, "learning_rate": 4.8424788554792165e-05, "loss": 1.3352688789367675, "step": 41130 }, { "epoch": 3.5057520238602473, "grad_norm": 2.5625, "learning_rate": 4.842382422903603e-05, "loss": 1.3845691680908203, "step": 41140 }, { "epoch": 3.506604175543247, "grad_norm": 2.328125, "learning_rate": 4.8422859617803526e-05, "loss": 1.3369606971740722, "step": 41150 }, { "epoch": 3.5074563272262465, "grad_norm": 2.546875, "learning_rate": 4.842189472110641e-05, "loss": 1.3131857872009278, "step": 41160 }, { "epoch": 3.508308478909246, "grad_norm": 2.90625, "learning_rate": 4.842092953895644e-05, "loss": 1.349133014678955, "step": 41170 }, { "epoch": 3.5091606305922456, "grad_norm": 2.390625, "learning_rate": 4.841996407136539e-05, "loss": 1.2998676300048828, "step": 41180 }, { "epoch": 3.510012782275245, "grad_norm": 2.203125, "learning_rate": 4.8418998318345014e-05, "loss": 1.2281023025512696, "step": 41190 }, { "epoch": 3.5108649339582447, "grad_norm": 3.734375, "learning_rate": 4.841803227990709e-05, "loss": 1.2552600860595704, "step": 41200 }, { "epoch": 3.5117170856412443, "grad_norm": 2.5, "learning_rate": 4.841706595606338e-05, "loss": 1.31652889251709, "step": 41210 }, { "epoch": 3.512569237324244, "grad_norm": 2.265625, "learning_rate": 4.841609934682567e-05, "loss": 1.338723373413086, "step": 41220 }, { "epoch": 3.5134213890072434, "grad_norm": 2.09375, "learning_rate": 4.841513245220575e-05, "loss": 1.4188146591186523, "step": 41230 }, { "epoch": 3.514273540690243, "grad_norm": 2.875, "learning_rate": 4.841416527221538e-05, "loss": 1.2955355644226074, "step": 41240 }, { "epoch": 3.5151256923732426, "grad_norm": 1.9140625, "learning_rate": 4.841319780686637e-05, "loss": 1.3262290000915526, "step": 41250 }, { "epoch": 3.515977844056242, "grad_norm": 1.8671875, "learning_rate": 4.84122300561705e-05, "loss": 1.4766525268554687, "step": 41260 }, { "epoch": 3.5168299957392417, "grad_norm": 2.40625, "learning_rate": 4.8411262020139575e-05, "loss": 1.3151623725891113, "step": 41270 }, { "epoch": 3.5176821474222413, "grad_norm": 3.109375, "learning_rate": 4.8410293698785384e-05, "loss": 1.293250274658203, "step": 41280 }, { "epoch": 3.518534299105241, "grad_norm": 2.671875, "learning_rate": 4.840932509211972e-05, "loss": 1.3080463409423828, "step": 41290 }, { "epoch": 3.5193864507882404, "grad_norm": 2.515625, "learning_rate": 4.8408356200154396e-05, "loss": 1.379481601715088, "step": 41300 }, { "epoch": 3.52023860247124, "grad_norm": 2.375, "learning_rate": 4.840738702290122e-05, "loss": 1.3727097511291504, "step": 41310 }, { "epoch": 3.5210907541542396, "grad_norm": 3.015625, "learning_rate": 4.840641756037201e-05, "loss": 1.3034915924072266, "step": 41320 }, { "epoch": 3.521942905837239, "grad_norm": 3.84375, "learning_rate": 4.840544781257857e-05, "loss": 1.226356315612793, "step": 41330 }, { "epoch": 3.5227950575202387, "grad_norm": 1.8359375, "learning_rate": 4.840447777953273e-05, "loss": 1.205128002166748, "step": 41340 }, { "epoch": 3.5236472092032383, "grad_norm": 4.28125, "learning_rate": 4.8403507461246314e-05, "loss": 1.393295669555664, "step": 41350 }, { "epoch": 3.524499360886238, "grad_norm": 2.375, "learning_rate": 4.840253685773113e-05, "loss": 1.0939643859863282, "step": 41360 }, { "epoch": 3.5253515125692374, "grad_norm": 2.125, "learning_rate": 4.8401565968999015e-05, "loss": 1.2762224197387695, "step": 41370 }, { "epoch": 3.526203664252237, "grad_norm": 2.5, "learning_rate": 4.8400594795061807e-05, "loss": 1.4281756401062011, "step": 41380 }, { "epoch": 3.5270558159352365, "grad_norm": 2.125, "learning_rate": 4.839962333593134e-05, "loss": 1.4324047088623046, "step": 41390 }, { "epoch": 3.527907967618236, "grad_norm": 2.84375, "learning_rate": 4.839865159161946e-05, "loss": 1.3921530723571778, "step": 41400 }, { "epoch": 3.5287601193012357, "grad_norm": 2.265625, "learning_rate": 4.8397679562138e-05, "loss": 1.2327905654907227, "step": 41410 }, { "epoch": 3.5296122709842352, "grad_norm": 2.40625, "learning_rate": 4.839670724749881e-05, "loss": 1.2180965423583985, "step": 41420 }, { "epoch": 3.530464422667235, "grad_norm": 2.3125, "learning_rate": 4.839573464771373e-05, "loss": 1.5274638175964355, "step": 41430 }, { "epoch": 3.5313165743502344, "grad_norm": 1.9375, "learning_rate": 4.839476176279463e-05, "loss": 1.4304553985595703, "step": 41440 }, { "epoch": 3.532168726033234, "grad_norm": 3.28125, "learning_rate": 4.839378859275336e-05, "loss": 1.3346229553222657, "step": 41450 }, { "epoch": 3.5330208777162335, "grad_norm": 1.6953125, "learning_rate": 4.8392815137601786e-05, "loss": 1.319550132751465, "step": 41460 }, { "epoch": 3.533873029399233, "grad_norm": 2.65625, "learning_rate": 4.839184139735176e-05, "loss": 1.1078950881958007, "step": 41470 }, { "epoch": 3.5347251810822327, "grad_norm": 3.328125, "learning_rate": 4.839086737201516e-05, "loss": 1.189216423034668, "step": 41480 }, { "epoch": 3.5355773327652322, "grad_norm": 2.40625, "learning_rate": 4.8389893061603856e-05, "loss": 1.297478199005127, "step": 41490 }, { "epoch": 3.536429484448232, "grad_norm": 1.7578125, "learning_rate": 4.838891846612971e-05, "loss": 1.382310390472412, "step": 41500 }, { "epoch": 3.5372816361312314, "grad_norm": 3.078125, "learning_rate": 4.838794358560461e-05, "loss": 1.3031683921813966, "step": 41510 }, { "epoch": 3.538133787814231, "grad_norm": 1.5390625, "learning_rate": 4.838696842004045e-05, "loss": 1.448872184753418, "step": 41520 }, { "epoch": 3.5389859394972305, "grad_norm": 2.25, "learning_rate": 4.838599296944908e-05, "loss": 1.375063133239746, "step": 41530 }, { "epoch": 3.53983809118023, "grad_norm": 3.59375, "learning_rate": 4.8385017233842424e-05, "loss": 1.3988686561584474, "step": 41540 }, { "epoch": 3.5406902428632296, "grad_norm": 3.015625, "learning_rate": 4.8384041213232354e-05, "loss": 1.310122585296631, "step": 41550 }, { "epoch": 3.541542394546229, "grad_norm": 3.140625, "learning_rate": 4.838306490763077e-05, "loss": 1.286724853515625, "step": 41560 }, { "epoch": 3.542394546229229, "grad_norm": 1.96875, "learning_rate": 4.8382088317049575e-05, "loss": 1.2621440887451172, "step": 41570 }, { "epoch": 3.5432466979122283, "grad_norm": 2.078125, "learning_rate": 4.838111144150066e-05, "loss": 1.3765737533569335, "step": 41580 }, { "epoch": 3.544098849595228, "grad_norm": 2.828125, "learning_rate": 4.8380134280995944e-05, "loss": 1.3347424507141112, "step": 41590 }, { "epoch": 3.5449510012782275, "grad_norm": 3.09375, "learning_rate": 4.837915683554732e-05, "loss": 1.2241161346435547, "step": 41600 }, { "epoch": 3.545803152961227, "grad_norm": 3.703125, "learning_rate": 4.837817910516672e-05, "loss": 1.3311188697814942, "step": 41610 }, { "epoch": 3.5466553046442266, "grad_norm": 1.8671875, "learning_rate": 4.837720108986605e-05, "loss": 1.2988201141357423, "step": 41620 }, { "epoch": 3.547507456327226, "grad_norm": 2.03125, "learning_rate": 4.837622278965723e-05, "loss": 1.341820526123047, "step": 41630 }, { "epoch": 3.5483596080102258, "grad_norm": 3.078125, "learning_rate": 4.8375244204552186e-05, "loss": 1.3586527824401855, "step": 41640 }, { "epoch": 3.5492117596932253, "grad_norm": 2.234375, "learning_rate": 4.837426533456283e-05, "loss": 1.4521449089050293, "step": 41650 }, { "epoch": 3.550063911376225, "grad_norm": 2.84375, "learning_rate": 4.837328617970111e-05, "loss": 1.4157866477966308, "step": 41660 }, { "epoch": 3.5509160630592245, "grad_norm": 3.03125, "learning_rate": 4.8372306739978954e-05, "loss": 1.3911331176757813, "step": 41670 }, { "epoch": 3.551768214742224, "grad_norm": 6.71875, "learning_rate": 4.837132701540829e-05, "loss": 1.3947698593139648, "step": 41680 }, { "epoch": 3.5526203664252236, "grad_norm": 2.53125, "learning_rate": 4.837034700600107e-05, "loss": 1.2974993705749511, "step": 41690 }, { "epoch": 3.553472518108223, "grad_norm": 1.5625, "learning_rate": 4.836936671176924e-05, "loss": 1.4129289627075194, "step": 41700 }, { "epoch": 3.5543246697912227, "grad_norm": 1.96875, "learning_rate": 4.836838613272473e-05, "loss": 1.4109420776367188, "step": 41710 }, { "epoch": 3.5551768214742223, "grad_norm": 2.15625, "learning_rate": 4.8367405268879504e-05, "loss": 1.3458591461181642, "step": 41720 }, { "epoch": 3.556028973157222, "grad_norm": 2.53125, "learning_rate": 4.8366424120245526e-05, "loss": 1.365672779083252, "step": 41730 }, { "epoch": 3.5568811248402215, "grad_norm": 3.03125, "learning_rate": 4.836544268683473e-05, "loss": 1.1606210708618163, "step": 41740 }, { "epoch": 3.557733276523221, "grad_norm": 1.34375, "learning_rate": 4.8364460968659084e-05, "loss": 1.3341217041015625, "step": 41750 }, { "epoch": 3.5585854282062206, "grad_norm": 2.328125, "learning_rate": 4.8363478965730566e-05, "loss": 1.4706729888916015, "step": 41760 }, { "epoch": 3.55943757988922, "grad_norm": 2.25, "learning_rate": 4.836249667806113e-05, "loss": 1.5499217987060547, "step": 41770 }, { "epoch": 3.5602897315722197, "grad_norm": 2.84375, "learning_rate": 4.836151410566275e-05, "loss": 1.391789722442627, "step": 41780 }, { "epoch": 3.5611418832552193, "grad_norm": 2.203125, "learning_rate": 4.8360531248547416e-05, "loss": 1.315841293334961, "step": 41790 }, { "epoch": 3.561994034938219, "grad_norm": 2.234375, "learning_rate": 4.835954810672709e-05, "loss": 1.3592495918273926, "step": 41800 }, { "epoch": 3.5628461866212184, "grad_norm": 1.8359375, "learning_rate": 4.8358564680213755e-05, "loss": 1.387868595123291, "step": 41810 }, { "epoch": 3.563698338304218, "grad_norm": 2.34375, "learning_rate": 4.83575809690194e-05, "loss": 1.4836169242858888, "step": 41820 }, { "epoch": 3.5645504899872176, "grad_norm": 2.359375, "learning_rate": 4.835659697315602e-05, "loss": 1.3696096420288086, "step": 41830 }, { "epoch": 3.565402641670217, "grad_norm": 2.15625, "learning_rate": 4.8355612692635586e-05, "loss": 1.3037410736083985, "step": 41840 }, { "epoch": 3.5662547933532167, "grad_norm": 2.5625, "learning_rate": 4.8354628127470127e-05, "loss": 1.2746944427490234, "step": 41850 }, { "epoch": 3.5671069450362163, "grad_norm": 3.921875, "learning_rate": 4.835364327767161e-05, "loss": 1.4141693115234375, "step": 41860 }, { "epoch": 3.567959096719216, "grad_norm": 2.453125, "learning_rate": 4.835265814325206e-05, "loss": 1.291187858581543, "step": 41870 }, { "epoch": 3.5688112484022154, "grad_norm": 3.125, "learning_rate": 4.835167272422347e-05, "loss": 1.444739532470703, "step": 41880 }, { "epoch": 3.569663400085215, "grad_norm": 1.7890625, "learning_rate": 4.8350687020597865e-05, "loss": 1.3023529052734375, "step": 41890 }, { "epoch": 3.5705155517682146, "grad_norm": 3.0, "learning_rate": 4.834970103238724e-05, "loss": 1.3335810661315919, "step": 41900 }, { "epoch": 3.571367703451214, "grad_norm": 2.015625, "learning_rate": 4.834871475960363e-05, "loss": 1.48587064743042, "step": 41910 }, { "epoch": 3.5722198551342137, "grad_norm": 2.625, "learning_rate": 4.834772820225903e-05, "loss": 1.4671486854553222, "step": 41920 }, { "epoch": 3.5730720068172133, "grad_norm": 2.3125, "learning_rate": 4.8346741360365486e-05, "loss": 1.233396053314209, "step": 41930 }, { "epoch": 3.573924158500213, "grad_norm": 1.90625, "learning_rate": 4.834575423393502e-05, "loss": 1.5029532432556152, "step": 41940 }, { "epoch": 3.5747763101832124, "grad_norm": 4.78125, "learning_rate": 4.8344766822979656e-05, "loss": 1.1016909599304199, "step": 41950 }, { "epoch": 3.575628461866212, "grad_norm": 2.484375, "learning_rate": 4.834377912751144e-05, "loss": 1.373913288116455, "step": 41960 }, { "epoch": 3.5764806135492115, "grad_norm": 2.046875, "learning_rate": 4.8342791147542395e-05, "loss": 1.314387607574463, "step": 41970 }, { "epoch": 3.577332765232211, "grad_norm": 3.125, "learning_rate": 4.8341802883084564e-05, "loss": 1.2242761611938477, "step": 41980 }, { "epoch": 3.5781849169152107, "grad_norm": 2.0625, "learning_rate": 4.8340814334150006e-05, "loss": 1.377543830871582, "step": 41990 }, { "epoch": 3.5790370685982102, "grad_norm": 1.5703125, "learning_rate": 4.8339825500750756e-05, "loss": 1.3786219596862792, "step": 42000 }, { "epoch": 3.57988922028121, "grad_norm": 3.140625, "learning_rate": 4.833883638289887e-05, "loss": 1.3065525054931642, "step": 42010 }, { "epoch": 3.5807413719642094, "grad_norm": 1.9609375, "learning_rate": 4.833784698060639e-05, "loss": 1.4150143623352052, "step": 42020 }, { "epoch": 3.581593523647209, "grad_norm": 2.6875, "learning_rate": 4.83368572938854e-05, "loss": 1.4561026573181153, "step": 42030 }, { "epoch": 3.5824456753302085, "grad_norm": 4.0625, "learning_rate": 4.833586732274794e-05, "loss": 1.3107271194458008, "step": 42040 }, { "epoch": 3.583297827013208, "grad_norm": 1.6796875, "learning_rate": 4.833487706720609e-05, "loss": 1.3316604614257812, "step": 42050 }, { "epoch": 3.5841499786962077, "grad_norm": 1.8046875, "learning_rate": 4.833388652727191e-05, "loss": 1.2825339317321778, "step": 42060 }, { "epoch": 3.5850021303792072, "grad_norm": 2.453125, "learning_rate": 4.833289570295746e-05, "loss": 1.3970711708068848, "step": 42070 }, { "epoch": 3.585854282062207, "grad_norm": 3.15625, "learning_rate": 4.833190459427484e-05, "loss": 1.3323444366455077, "step": 42080 }, { "epoch": 3.586706433745207, "grad_norm": 2.453125, "learning_rate": 4.833091320123612e-05, "loss": 1.3939823150634765, "step": 42090 }, { "epoch": 3.5875585854282064, "grad_norm": 2.359375, "learning_rate": 4.832992152385337e-05, "loss": 1.280727005004883, "step": 42100 }, { "epoch": 3.588410737111206, "grad_norm": 3.28125, "learning_rate": 4.832892956213869e-05, "loss": 1.268071460723877, "step": 42110 }, { "epoch": 3.5892628887942055, "grad_norm": 1.8828125, "learning_rate": 4.8327937316104175e-05, "loss": 1.3528314590454102, "step": 42120 }, { "epoch": 3.590115040477205, "grad_norm": 2.171875, "learning_rate": 4.83269447857619e-05, "loss": 1.3143463134765625, "step": 42130 }, { "epoch": 3.5909671921602047, "grad_norm": 1.8125, "learning_rate": 4.8325951971123976e-05, "loss": 1.2717316627502442, "step": 42140 }, { "epoch": 3.5918193438432042, "grad_norm": 2.1875, "learning_rate": 4.832495887220249e-05, "loss": 1.3085186004638671, "step": 42150 }, { "epoch": 3.592671495526204, "grad_norm": 3.21875, "learning_rate": 4.832396548900956e-05, "loss": 1.2871278762817382, "step": 42160 }, { "epoch": 3.5935236472092034, "grad_norm": 1.9296875, "learning_rate": 4.832297182155728e-05, "loss": 1.3165276527404786, "step": 42170 }, { "epoch": 3.594375798892203, "grad_norm": 1.9375, "learning_rate": 4.832197786985777e-05, "loss": 1.267611312866211, "step": 42180 }, { "epoch": 3.5952279505752025, "grad_norm": 1.609375, "learning_rate": 4.832098363392313e-05, "loss": 1.330553913116455, "step": 42190 }, { "epoch": 3.596080102258202, "grad_norm": 2.984375, "learning_rate": 4.83199891137655e-05, "loss": 1.3761482238769531, "step": 42200 }, { "epoch": 3.5969322539412016, "grad_norm": 2.015625, "learning_rate": 4.831899430939698e-05, "loss": 1.4060404777526856, "step": 42210 }, { "epoch": 3.597784405624201, "grad_norm": 3.375, "learning_rate": 4.831799922082969e-05, "loss": 1.3376286506652832, "step": 42220 }, { "epoch": 3.598636557307201, "grad_norm": 2.828125, "learning_rate": 4.831700384807579e-05, "loss": 1.33874568939209, "step": 42230 }, { "epoch": 3.5994887089902003, "grad_norm": 3.234375, "learning_rate": 4.8316008191147375e-05, "loss": 1.272439193725586, "step": 42240 }, { "epoch": 3.6003408606732, "grad_norm": 1.9296875, "learning_rate": 4.83150122500566e-05, "loss": 1.4580796241760254, "step": 42250 }, { "epoch": 3.6011930123561995, "grad_norm": 2.390625, "learning_rate": 4.83140160248156e-05, "loss": 1.2243328094482422, "step": 42260 }, { "epoch": 3.602045164039199, "grad_norm": 2.875, "learning_rate": 4.831301951543651e-05, "loss": 1.3296566009521484, "step": 42270 }, { "epoch": 3.6028973157221986, "grad_norm": 2.71875, "learning_rate": 4.8312022721931474e-05, "loss": 1.4742637634277345, "step": 42280 }, { "epoch": 3.603749467405198, "grad_norm": 1.4609375, "learning_rate": 4.8311025644312654e-05, "loss": 1.3512228012084961, "step": 42290 }, { "epoch": 3.6046016190881978, "grad_norm": 1.8125, "learning_rate": 4.831002828259219e-05, "loss": 1.3925293922424316, "step": 42300 }, { "epoch": 3.6054537707711973, "grad_norm": 1.6640625, "learning_rate": 4.8309030636782236e-05, "loss": 1.3205390930175782, "step": 42310 }, { "epoch": 3.606305922454197, "grad_norm": 2.546875, "learning_rate": 4.830803270689496e-05, "loss": 1.4286185264587403, "step": 42320 }, { "epoch": 3.6071580741371965, "grad_norm": 2.359375, "learning_rate": 4.8307034492942516e-05, "loss": 1.3910747528076173, "step": 42330 }, { "epoch": 3.608010225820196, "grad_norm": 2.15625, "learning_rate": 4.8306035994937074e-05, "loss": 1.4450172424316405, "step": 42340 }, { "epoch": 3.6088623775031956, "grad_norm": 2.625, "learning_rate": 4.83050372128908e-05, "loss": 1.3190581321716308, "step": 42350 }, { "epoch": 3.609714529186195, "grad_norm": 3.328125, "learning_rate": 4.830403814681587e-05, "loss": 1.3969237327575683, "step": 42360 }, { "epoch": 3.6105666808691947, "grad_norm": 2.96875, "learning_rate": 4.830303879672446e-05, "loss": 1.3599655151367187, "step": 42370 }, { "epoch": 3.6114188325521943, "grad_norm": 1.7109375, "learning_rate": 4.830203916262874e-05, "loss": 1.3948830604553222, "step": 42380 }, { "epoch": 3.612270984235194, "grad_norm": 3.4375, "learning_rate": 4.8301039244540916e-05, "loss": 1.4010393142700195, "step": 42390 }, { "epoch": 3.6131231359181935, "grad_norm": 2.0625, "learning_rate": 4.830003904247315e-05, "loss": 1.3304326057434082, "step": 42400 }, { "epoch": 3.613975287601193, "grad_norm": 2.28125, "learning_rate": 4.829903855643764e-05, "loss": 1.3634976387023925, "step": 42410 }, { "epoch": 3.6148274392841926, "grad_norm": 2.109375, "learning_rate": 4.829803778644658e-05, "loss": 1.3296554565429688, "step": 42420 }, { "epoch": 3.615679590967192, "grad_norm": 2.390625, "learning_rate": 4.8297036732512175e-05, "loss": 1.509408664703369, "step": 42430 }, { "epoch": 3.6165317426501917, "grad_norm": 2.609375, "learning_rate": 4.829603539464661e-05, "loss": 1.3626912117004395, "step": 42440 }, { "epoch": 3.6173838943331913, "grad_norm": 2.171875, "learning_rate": 4.82950337728621e-05, "loss": 1.4506259918212892, "step": 42450 }, { "epoch": 3.618236046016191, "grad_norm": 3.140625, "learning_rate": 4.829403186717085e-05, "loss": 1.3878753662109375, "step": 42460 }, { "epoch": 3.6190881976991904, "grad_norm": 2.78125, "learning_rate": 4.829302967758507e-05, "loss": 1.4015599250793458, "step": 42470 }, { "epoch": 3.61994034938219, "grad_norm": 1.9921875, "learning_rate": 4.829202720411697e-05, "loss": 1.4096206665039062, "step": 42480 }, { "epoch": 3.6207925010651896, "grad_norm": 2.171875, "learning_rate": 4.829102444677878e-05, "loss": 1.432211971282959, "step": 42490 }, { "epoch": 3.621644652748189, "grad_norm": 2.4375, "learning_rate": 4.8290021405582697e-05, "loss": 1.257372760772705, "step": 42500 }, { "epoch": 3.6224968044311887, "grad_norm": 2.234375, "learning_rate": 4.828901808054097e-05, "loss": 1.3654911041259765, "step": 42510 }, { "epoch": 3.6233489561141883, "grad_norm": 1.6796875, "learning_rate": 4.828801447166582e-05, "loss": 1.2438851356506349, "step": 42520 }, { "epoch": 3.624201107797188, "grad_norm": 3.03125, "learning_rate": 4.8287010578969463e-05, "loss": 1.162762451171875, "step": 42530 }, { "epoch": 3.6250532594801874, "grad_norm": 1.8671875, "learning_rate": 4.8286006402464155e-05, "loss": 1.287608528137207, "step": 42540 }, { "epoch": 3.625905411163187, "grad_norm": 2.515625, "learning_rate": 4.828500194216213e-05, "loss": 1.3126714706420899, "step": 42550 }, { "epoch": 3.6267575628461866, "grad_norm": 2.375, "learning_rate": 4.8283997198075625e-05, "loss": 1.4705073356628418, "step": 42560 }, { "epoch": 3.627609714529186, "grad_norm": 3.84375, "learning_rate": 4.8282992170216885e-05, "loss": 1.3930940628051758, "step": 42570 }, { "epoch": 3.6284618662121857, "grad_norm": 1.984375, "learning_rate": 4.8281986858598164e-05, "loss": 1.5216362953186036, "step": 42580 }, { "epoch": 3.6293140178951853, "grad_norm": 2.640625, "learning_rate": 4.82809812632317e-05, "loss": 1.1636028289794922, "step": 42590 }, { "epoch": 3.630166169578185, "grad_norm": 2.796875, "learning_rate": 4.8279975384129764e-05, "loss": 1.2218208312988281, "step": 42600 }, { "epoch": 3.6310183212611844, "grad_norm": 2.0625, "learning_rate": 4.827896922130461e-05, "loss": 1.3174193382263184, "step": 42610 }, { "epoch": 3.631870472944184, "grad_norm": 2.671875, "learning_rate": 4.8277962774768504e-05, "loss": 1.3176143646240235, "step": 42620 }, { "epoch": 3.6327226246271835, "grad_norm": 3.984375, "learning_rate": 4.8276956044533704e-05, "loss": 1.5124584197998048, "step": 42630 }, { "epoch": 3.633574776310183, "grad_norm": 2.546875, "learning_rate": 4.8275949030612486e-05, "loss": 1.4114289283752441, "step": 42640 }, { "epoch": 3.6344269279931827, "grad_norm": 2.6875, "learning_rate": 4.827494173301712e-05, "loss": 1.0939324378967286, "step": 42650 }, { "epoch": 3.6352790796761822, "grad_norm": 1.796875, "learning_rate": 4.827393415175988e-05, "loss": 1.4237726211547852, "step": 42660 }, { "epoch": 3.636131231359182, "grad_norm": 2.8125, "learning_rate": 4.8272926286853055e-05, "loss": 1.3087461471557618, "step": 42670 }, { "epoch": 3.6369833830421814, "grad_norm": 2.34375, "learning_rate": 4.827191813830892e-05, "loss": 1.570412826538086, "step": 42680 }, { "epoch": 3.637835534725181, "grad_norm": 2.078125, "learning_rate": 4.827090970613976e-05, "loss": 1.3939638137817383, "step": 42690 }, { "epoch": 3.6386876864081805, "grad_norm": 2.953125, "learning_rate": 4.826990099035788e-05, "loss": 1.3146188735961915, "step": 42700 }, { "epoch": 3.63953983809118, "grad_norm": 2.75, "learning_rate": 4.826889199097555e-05, "loss": 1.198371124267578, "step": 42710 }, { "epoch": 3.6403919897741797, "grad_norm": 2.3125, "learning_rate": 4.826788270800508e-05, "loss": 1.2866704940795899, "step": 42720 }, { "epoch": 3.6412441414571792, "grad_norm": 1.8984375, "learning_rate": 4.826687314145878e-05, "loss": 1.3548870086669922, "step": 42730 }, { "epoch": 3.642096293140179, "grad_norm": 2.078125, "learning_rate": 4.826586329134894e-05, "loss": 1.4677936553955078, "step": 42740 }, { "epoch": 3.6429484448231784, "grad_norm": 1.84375, "learning_rate": 4.826485315768787e-05, "loss": 1.3517830848693848, "step": 42750 }, { "epoch": 3.643800596506178, "grad_norm": 1.9375, "learning_rate": 4.826384274048789e-05, "loss": 1.3594006538391112, "step": 42760 }, { "epoch": 3.644652748189178, "grad_norm": 2.453125, "learning_rate": 4.826283203976131e-05, "loss": 1.3549465179443358, "step": 42770 }, { "epoch": 3.6455048998721775, "grad_norm": 2.90625, "learning_rate": 4.826182105552043e-05, "loss": 1.4108749389648438, "step": 42780 }, { "epoch": 3.646357051555177, "grad_norm": 2.28125, "learning_rate": 4.82608097877776e-05, "loss": 1.3404871940612793, "step": 42790 }, { "epoch": 3.6472092032381767, "grad_norm": 2.671875, "learning_rate": 4.825979823654513e-05, "loss": 1.4417247772216797, "step": 42800 }, { "epoch": 3.6480613549211762, "grad_norm": 3.1875, "learning_rate": 4.8258786401835345e-05, "loss": 1.3597461700439453, "step": 42810 }, { "epoch": 3.648913506604176, "grad_norm": 1.578125, "learning_rate": 4.8257774283660586e-05, "loss": 1.2508840560913086, "step": 42820 }, { "epoch": 3.6497656582871754, "grad_norm": 2.03125, "learning_rate": 4.825676188203318e-05, "loss": 1.458486843109131, "step": 42830 }, { "epoch": 3.650617809970175, "grad_norm": 2.15625, "learning_rate": 4.8255749196965464e-05, "loss": 1.3795955657958985, "step": 42840 }, { "epoch": 3.6514699616531745, "grad_norm": 1.9296875, "learning_rate": 4.825473622846979e-05, "loss": 1.4807517051696777, "step": 42850 }, { "epoch": 3.652322113336174, "grad_norm": 1.9375, "learning_rate": 4.8253722976558504e-05, "loss": 1.2648712158203126, "step": 42860 }, { "epoch": 3.6531742650191736, "grad_norm": 2.46875, "learning_rate": 4.825270944124394e-05, "loss": 1.329369354248047, "step": 42870 }, { "epoch": 3.654026416702173, "grad_norm": 1.6875, "learning_rate": 4.825169562253846e-05, "loss": 1.4764545440673829, "step": 42880 }, { "epoch": 3.6548785683851728, "grad_norm": 2.15625, "learning_rate": 4.825068152045443e-05, "loss": 1.3299206733703612, "step": 42890 }, { "epoch": 3.6557307200681723, "grad_norm": 2.546875, "learning_rate": 4.8249667135004193e-05, "loss": 1.3443846702575684, "step": 42900 }, { "epoch": 3.656582871751172, "grad_norm": 2.40625, "learning_rate": 4.824865246620012e-05, "loss": 1.3987134933471679, "step": 42910 }, { "epoch": 3.6574350234341715, "grad_norm": 2.734375, "learning_rate": 4.824763751405457e-05, "loss": 1.280120849609375, "step": 42920 }, { "epoch": 3.658287175117171, "grad_norm": 3.078125, "learning_rate": 4.824662227857992e-05, "loss": 1.4423699378967285, "step": 42930 }, { "epoch": 3.6591393268001706, "grad_norm": 2.640625, "learning_rate": 4.824560675978854e-05, "loss": 1.3605942726135254, "step": 42940 }, { "epoch": 3.65999147848317, "grad_norm": 3.5, "learning_rate": 4.824459095769281e-05, "loss": 1.3420199394226073, "step": 42950 }, { "epoch": 3.6608436301661698, "grad_norm": 1.875, "learning_rate": 4.8243574872305106e-05, "loss": 1.4195094108581543, "step": 42960 }, { "epoch": 3.6616957818491693, "grad_norm": 2.375, "learning_rate": 4.824255850363781e-05, "loss": 1.1642986297607423, "step": 42970 }, { "epoch": 3.662547933532169, "grad_norm": 3.046875, "learning_rate": 4.824154185170331e-05, "loss": 1.418912982940674, "step": 42980 }, { "epoch": 3.6634000852151685, "grad_norm": 1.734375, "learning_rate": 4.824052491651401e-05, "loss": 1.3936370849609374, "step": 42990 }, { "epoch": 3.664252236898168, "grad_norm": 2.71875, "learning_rate": 4.823950769808228e-05, "loss": 1.4470015525817872, "step": 43000 }, { "epoch": 3.6651043885811676, "grad_norm": 2.828125, "learning_rate": 4.823849019642053e-05, "loss": 1.2902327537536622, "step": 43010 }, { "epoch": 3.665956540264167, "grad_norm": 2.25, "learning_rate": 4.8237472411541166e-05, "loss": 1.2808645248413086, "step": 43020 }, { "epoch": 3.6668086919471667, "grad_norm": 1.671875, "learning_rate": 4.8236454343456585e-05, "loss": 1.26011323928833, "step": 43030 }, { "epoch": 3.6676608436301663, "grad_norm": 2.796875, "learning_rate": 4.823543599217919e-05, "loss": 1.4844676971435546, "step": 43040 }, { "epoch": 3.668512995313166, "grad_norm": 4.5625, "learning_rate": 4.823441735772141e-05, "loss": 1.1819013595581054, "step": 43050 }, { "epoch": 3.6693651469961654, "grad_norm": 2.25, "learning_rate": 4.823339844009563e-05, "loss": 1.2727543830871582, "step": 43060 }, { "epoch": 3.670217298679165, "grad_norm": 2.171875, "learning_rate": 4.8232379239314304e-05, "loss": 1.353704833984375, "step": 43070 }, { "epoch": 3.6710694503621646, "grad_norm": 2.046875, "learning_rate": 4.8231359755389825e-05, "loss": 1.3387232780456544, "step": 43080 }, { "epoch": 3.671921602045164, "grad_norm": 1.984375, "learning_rate": 4.823033998833463e-05, "loss": 1.083729362487793, "step": 43090 }, { "epoch": 3.6727737537281637, "grad_norm": 1.9140625, "learning_rate": 4.8229319938161144e-05, "loss": 1.3970197677612304, "step": 43100 }, { "epoch": 3.6736259054111633, "grad_norm": 2.9375, "learning_rate": 4.82282996048818e-05, "loss": 1.244295597076416, "step": 43110 }, { "epoch": 3.674478057094163, "grad_norm": 2.265625, "learning_rate": 4.822727898850903e-05, "loss": 1.3656450271606446, "step": 43120 }, { "epoch": 3.6753302087771624, "grad_norm": 2.828125, "learning_rate": 4.822625808905528e-05, "loss": 1.3812041282653809, "step": 43130 }, { "epoch": 3.676182360460162, "grad_norm": 3.75, "learning_rate": 4.8225236906532985e-05, "loss": 1.2421896934509278, "step": 43140 }, { "epoch": 3.6770345121431616, "grad_norm": 1.453125, "learning_rate": 4.82242154409546e-05, "loss": 1.3715524673461914, "step": 43150 }, { "epoch": 3.677886663826161, "grad_norm": 2.234375, "learning_rate": 4.822319369233256e-05, "loss": 1.2899226188659667, "step": 43160 }, { "epoch": 3.6787388155091607, "grad_norm": 2.796875, "learning_rate": 4.822217166067933e-05, "loss": 1.4172628402709961, "step": 43170 }, { "epoch": 3.6795909671921603, "grad_norm": 2.03125, "learning_rate": 4.8221149346007354e-05, "loss": 1.2570703506469727, "step": 43180 }, { "epoch": 3.68044311887516, "grad_norm": 2.59375, "learning_rate": 4.8220126748329106e-05, "loss": 1.4505039215087892, "step": 43190 }, { "epoch": 3.6812952705581594, "grad_norm": 3.03125, "learning_rate": 4.8219103867657036e-05, "loss": 1.3808673858642577, "step": 43200 }, { "epoch": 3.682147422241159, "grad_norm": 1.8984375, "learning_rate": 4.821808070400362e-05, "loss": 1.3955391883850097, "step": 43210 }, { "epoch": 3.6829995739241586, "grad_norm": 2.21875, "learning_rate": 4.821705725738133e-05, "loss": 1.4667595863342284, "step": 43220 }, { "epoch": 3.683851725607158, "grad_norm": 2.125, "learning_rate": 4.821603352780262e-05, "loss": 1.4548219680786132, "step": 43230 }, { "epoch": 3.6847038772901577, "grad_norm": 2.203125, "learning_rate": 4.821500951527999e-05, "loss": 1.3057329177856445, "step": 43240 }, { "epoch": 3.6855560289731573, "grad_norm": 2.453125, "learning_rate": 4.8213985219825895e-05, "loss": 1.4262274742126464, "step": 43250 }, { "epoch": 3.686408180656157, "grad_norm": 3.078125, "learning_rate": 4.8212960641452845e-05, "loss": 1.3477272987365723, "step": 43260 }, { "epoch": 3.6872603323391564, "grad_norm": 5.40625, "learning_rate": 4.8211935780173304e-05, "loss": 1.3042797088623046, "step": 43270 }, { "epoch": 3.688112484022156, "grad_norm": 3.65625, "learning_rate": 4.821091063599978e-05, "loss": 1.33046293258667, "step": 43280 }, { "epoch": 3.6889646357051555, "grad_norm": 2.265625, "learning_rate": 4.820988520894476e-05, "loss": 1.3839834213256836, "step": 43290 }, { "epoch": 3.689816787388155, "grad_norm": 2.328125, "learning_rate": 4.820885949902074e-05, "loss": 1.4055523872375488, "step": 43300 }, { "epoch": 3.6906689390711547, "grad_norm": 1.8515625, "learning_rate": 4.820783350624023e-05, "loss": 1.395895004272461, "step": 43310 }, { "epoch": 3.6915210907541542, "grad_norm": 2.046875, "learning_rate": 4.820680723061572e-05, "loss": 1.318495750427246, "step": 43320 }, { "epoch": 3.692373242437154, "grad_norm": 3.671875, "learning_rate": 4.820578067215972e-05, "loss": 1.4258047103881837, "step": 43330 }, { "epoch": 3.6932253941201534, "grad_norm": 2.71875, "learning_rate": 4.820475383088475e-05, "loss": 1.1886700630187987, "step": 43340 }, { "epoch": 3.694077545803153, "grad_norm": 2.296875, "learning_rate": 4.820372670680332e-05, "loss": 1.2261111259460449, "step": 43350 }, { "epoch": 3.6949296974861525, "grad_norm": 2.390625, "learning_rate": 4.820269929992795e-05, "loss": 1.388859176635742, "step": 43360 }, { "epoch": 3.695781849169152, "grad_norm": 2.296875, "learning_rate": 4.820167161027116e-05, "loss": 1.3032358169555665, "step": 43370 }, { "epoch": 3.6966340008521517, "grad_norm": 2.1875, "learning_rate": 4.820064363784547e-05, "loss": 1.41463623046875, "step": 43380 }, { "epoch": 3.6974861525351512, "grad_norm": 2.125, "learning_rate": 4.819961538266341e-05, "loss": 1.448246955871582, "step": 43390 }, { "epoch": 3.698338304218151, "grad_norm": 1.9453125, "learning_rate": 4.819858684473751e-05, "loss": 1.3785296440124513, "step": 43400 }, { "epoch": 3.6991904559011504, "grad_norm": 3.546875, "learning_rate": 4.8197558024080326e-05, "loss": 1.2923458099365235, "step": 43410 }, { "epoch": 3.70004260758415, "grad_norm": 2.875, "learning_rate": 4.819652892070437e-05, "loss": 1.3094345092773438, "step": 43420 }, { "epoch": 3.7008947592671495, "grad_norm": 2.59375, "learning_rate": 4.81954995346222e-05, "loss": 1.37149019241333, "step": 43430 }, { "epoch": 3.701746910950149, "grad_norm": 2.125, "learning_rate": 4.8194469865846345e-05, "loss": 1.3689780235290527, "step": 43440 }, { "epoch": 3.7025990626331486, "grad_norm": 2.265625, "learning_rate": 4.819343991438937e-05, "loss": 1.437057113647461, "step": 43450 }, { "epoch": 3.703451214316148, "grad_norm": 1.9921875, "learning_rate": 4.819240968026383e-05, "loss": 1.2368280410766601, "step": 43460 }, { "epoch": 3.704303365999148, "grad_norm": 1.9375, "learning_rate": 4.819137916348226e-05, "loss": 1.4327845573425293, "step": 43470 }, { "epoch": 3.7051555176821473, "grad_norm": 1.6875, "learning_rate": 4.8190348364057244e-05, "loss": 1.4253724098205567, "step": 43480 }, { "epoch": 3.706007669365147, "grad_norm": 2.4375, "learning_rate": 4.818931728200133e-05, "loss": 1.231467342376709, "step": 43490 }, { "epoch": 3.7068598210481465, "grad_norm": 3.09375, "learning_rate": 4.8188285917327086e-05, "loss": 1.299738311767578, "step": 43500 }, { "epoch": 3.707711972731146, "grad_norm": 2.15625, "learning_rate": 4.8187254270047086e-05, "loss": 1.339125156402588, "step": 43510 }, { "epoch": 3.7085641244141456, "grad_norm": 2.09375, "learning_rate": 4.8186222340173896e-05, "loss": 1.3400043487548827, "step": 43520 }, { "epoch": 3.709416276097145, "grad_norm": 2.8125, "learning_rate": 4.81851901277201e-05, "loss": 1.327138328552246, "step": 43530 }, { "epoch": 3.7102684277801448, "grad_norm": 2.046875, "learning_rate": 4.818415763269828e-05, "loss": 1.2594643592834474, "step": 43540 }, { "epoch": 3.7111205794631443, "grad_norm": 2.046875, "learning_rate": 4.818312485512101e-05, "loss": 1.356344985961914, "step": 43550 }, { "epoch": 3.711972731146144, "grad_norm": 2.375, "learning_rate": 4.818209179500088e-05, "loss": 1.188958740234375, "step": 43560 }, { "epoch": 3.7128248828291435, "grad_norm": 2.671875, "learning_rate": 4.818105845235048e-05, "loss": 1.3465510368347169, "step": 43570 }, { "epoch": 3.713677034512143, "grad_norm": 1.5859375, "learning_rate": 4.8180024827182414e-05, "loss": 1.2836125373840332, "step": 43580 }, { "epoch": 3.7145291861951426, "grad_norm": 2.0, "learning_rate": 4.817899091950927e-05, "loss": 1.486982250213623, "step": 43590 }, { "epoch": 3.715381337878142, "grad_norm": 4.03125, "learning_rate": 4.817795672934364e-05, "loss": 1.379751491546631, "step": 43600 }, { "epoch": 3.7162334895611417, "grad_norm": 2.03125, "learning_rate": 4.817692225669814e-05, "loss": 1.4169087409973145, "step": 43610 }, { "epoch": 3.7170856412441413, "grad_norm": 2.578125, "learning_rate": 4.8175887501585385e-05, "loss": 1.4764157295227052, "step": 43620 }, { "epoch": 3.717937792927141, "grad_norm": 3.515625, "learning_rate": 4.8174852464017964e-05, "loss": 1.2806448936462402, "step": 43630 }, { "epoch": 3.7187899446101405, "grad_norm": 2.21875, "learning_rate": 4.817381714400851e-05, "loss": 1.383859634399414, "step": 43640 }, { "epoch": 3.71964209629314, "grad_norm": 3.75, "learning_rate": 4.817278154156964e-05, "loss": 1.3323596000671387, "step": 43650 }, { "epoch": 3.7204942479761396, "grad_norm": 1.765625, "learning_rate": 4.817174565671396e-05, "loss": 1.383214569091797, "step": 43660 }, { "epoch": 3.721346399659139, "grad_norm": 2.03125, "learning_rate": 4.817070948945411e-05, "loss": 1.4163935661315918, "step": 43670 }, { "epoch": 3.7221985513421387, "grad_norm": 2.328125, "learning_rate": 4.8169673039802715e-05, "loss": 1.3142024040222169, "step": 43680 }, { "epoch": 3.7230507030251383, "grad_norm": 1.65625, "learning_rate": 4.81686363077724e-05, "loss": 1.464999008178711, "step": 43690 }, { "epoch": 3.723902854708138, "grad_norm": 2.203125, "learning_rate": 4.816759929337581e-05, "loss": 1.4191715240478515, "step": 43700 }, { "epoch": 3.7247550063911374, "grad_norm": 2.109375, "learning_rate": 4.816656199662557e-05, "loss": 1.2923406600952148, "step": 43710 }, { "epoch": 3.725607158074137, "grad_norm": 2.328125, "learning_rate": 4.816552441753434e-05, "loss": 1.4863441467285157, "step": 43720 }, { "epoch": 3.7264593097571366, "grad_norm": 3.859375, "learning_rate": 4.816448655611475e-05, "loss": 1.2379725456237793, "step": 43730 }, { "epoch": 3.727311461440136, "grad_norm": 2.5, "learning_rate": 4.816344841237946e-05, "loss": 1.3530728340148925, "step": 43740 }, { "epoch": 3.7281636131231357, "grad_norm": 2.484375, "learning_rate": 4.816240998634111e-05, "loss": 1.4256890296936036, "step": 43750 }, { "epoch": 3.7290157648061353, "grad_norm": 3.15625, "learning_rate": 4.816137127801237e-05, "loss": 1.117591381072998, "step": 43760 }, { "epoch": 3.729867916489135, "grad_norm": 2.59375, "learning_rate": 4.816033228740588e-05, "loss": 1.2217759132385253, "step": 43770 }, { "epoch": 3.7307200681721344, "grad_norm": 2.953125, "learning_rate": 4.8159293014534325e-05, "loss": 1.1735649108886719, "step": 43780 }, { "epoch": 3.731572219855134, "grad_norm": 2.890625, "learning_rate": 4.815825345941036e-05, "loss": 1.2582507133483887, "step": 43790 }, { "epoch": 3.7324243715381336, "grad_norm": 1.8203125, "learning_rate": 4.8157213622046654e-05, "loss": 1.4721334457397461, "step": 43800 }, { "epoch": 3.733276523221133, "grad_norm": 1.6640625, "learning_rate": 4.815617350245588e-05, "loss": 1.372625732421875, "step": 43810 }, { "epoch": 3.7341286749041327, "grad_norm": 1.6328125, "learning_rate": 4.8155133100650715e-05, "loss": 1.442280387878418, "step": 43820 }, { "epoch": 3.7349808265871323, "grad_norm": 2.65625, "learning_rate": 4.8154092416643836e-05, "loss": 1.4347484588623047, "step": 43830 }, { "epoch": 3.735832978270132, "grad_norm": 2.390625, "learning_rate": 4.815305145044794e-05, "loss": 1.2049548149108886, "step": 43840 }, { "epoch": 3.7366851299531314, "grad_norm": 1.984375, "learning_rate": 4.815201020207569e-05, "loss": 1.5229731559753419, "step": 43850 }, { "epoch": 3.737537281636131, "grad_norm": 1.796875, "learning_rate": 4.8150968671539795e-05, "loss": 1.262214756011963, "step": 43860 }, { "epoch": 3.7383894333191305, "grad_norm": 3.71875, "learning_rate": 4.814992685885294e-05, "loss": 1.4049668312072754, "step": 43870 }, { "epoch": 3.73924158500213, "grad_norm": 2.8125, "learning_rate": 4.814888476402784e-05, "loss": 1.352248191833496, "step": 43880 }, { "epoch": 3.7400937366851297, "grad_norm": 1.8125, "learning_rate": 4.814784238707717e-05, "loss": 1.331105613708496, "step": 43890 }, { "epoch": 3.7409458883681297, "grad_norm": 2.65625, "learning_rate": 4.814679972801364e-05, "loss": 1.4363786697387695, "step": 43900 }, { "epoch": 3.7417980400511293, "grad_norm": 3.09375, "learning_rate": 4.8145756786849964e-05, "loss": 1.2910935401916503, "step": 43910 }, { "epoch": 3.742650191734129, "grad_norm": 1.6953125, "learning_rate": 4.814471356359885e-05, "loss": 1.3157504081726075, "step": 43920 }, { "epoch": 3.7435023434171284, "grad_norm": 2.171875, "learning_rate": 4.8143670058273016e-05, "loss": 1.399945068359375, "step": 43930 }, { "epoch": 3.744354495100128, "grad_norm": 4.0625, "learning_rate": 4.8142626270885175e-05, "loss": 1.4498558044433594, "step": 43940 }, { "epoch": 3.7452066467831275, "grad_norm": 3.328125, "learning_rate": 4.8141582201448054e-05, "loss": 1.302663803100586, "step": 43950 }, { "epoch": 3.746058798466127, "grad_norm": 1.9453125, "learning_rate": 4.8140537849974366e-05, "loss": 1.4146525382995605, "step": 43960 }, { "epoch": 3.7469109501491267, "grad_norm": 2.0, "learning_rate": 4.813949321647685e-05, "loss": 1.2236702919006348, "step": 43970 }, { "epoch": 3.7477631018321262, "grad_norm": 2.5, "learning_rate": 4.813844830096823e-05, "loss": 1.2639739990234375, "step": 43980 }, { "epoch": 3.748615253515126, "grad_norm": 2.40625, "learning_rate": 4.8137403103461245e-05, "loss": 1.3317351341247559, "step": 43990 }, { "epoch": 3.7494674051981254, "grad_norm": 2.234375, "learning_rate": 4.813635762396863e-05, "loss": 1.3974133491516114, "step": 44000 }, { "epoch": 3.750319556881125, "grad_norm": 3.53125, "learning_rate": 4.813531186250314e-05, "loss": 1.374886703491211, "step": 44010 }, { "epoch": 3.7511717085641245, "grad_norm": 1.5390625, "learning_rate": 4.81342658190775e-05, "loss": 1.3412156105041504, "step": 44020 }, { "epoch": 3.752023860247124, "grad_norm": 2.171875, "learning_rate": 4.813321949370447e-05, "loss": 1.4520247459411622, "step": 44030 }, { "epoch": 3.7528760119301237, "grad_norm": 1.7109375, "learning_rate": 4.81321728863968e-05, "loss": 1.3920376777648926, "step": 44040 }, { "epoch": 3.7537281636131232, "grad_norm": 2.046875, "learning_rate": 4.8131125997167246e-05, "loss": 1.3200136184692384, "step": 44050 }, { "epoch": 3.754580315296123, "grad_norm": 2.578125, "learning_rate": 4.813007882602857e-05, "loss": 1.390748882293701, "step": 44060 }, { "epoch": 3.7554324669791224, "grad_norm": 2.171875, "learning_rate": 4.8129031372993524e-05, "loss": 1.4092388153076172, "step": 44070 }, { "epoch": 3.756284618662122, "grad_norm": 2.234375, "learning_rate": 4.8127983638074884e-05, "loss": 1.4034308433532714, "step": 44080 }, { "epoch": 3.7571367703451215, "grad_norm": 2.9375, "learning_rate": 4.812693562128541e-05, "loss": 1.4457584381103517, "step": 44090 }, { "epoch": 3.757988922028121, "grad_norm": 2.765625, "learning_rate": 4.812588732263788e-05, "loss": 1.2753644943237306, "step": 44100 }, { "epoch": 3.7588410737111206, "grad_norm": 3.015625, "learning_rate": 4.812483874214508e-05, "loss": 1.2795486450195312, "step": 44110 }, { "epoch": 3.75969322539412, "grad_norm": 3.359375, "learning_rate": 4.8123789879819775e-05, "loss": 1.3402891159057617, "step": 44120 }, { "epoch": 3.7605453770771198, "grad_norm": 2.421875, "learning_rate": 4.8122740735674754e-05, "loss": 1.2691746711730958, "step": 44130 }, { "epoch": 3.7613975287601193, "grad_norm": 2.421875, "learning_rate": 4.812169130972281e-05, "loss": 1.263369369506836, "step": 44140 }, { "epoch": 3.762249680443119, "grad_norm": 2.71875, "learning_rate": 4.812064160197671e-05, "loss": 1.3779077529907227, "step": 44150 }, { "epoch": 3.7631018321261185, "grad_norm": 2.65625, "learning_rate": 4.811959161244927e-05, "loss": 1.408134651184082, "step": 44160 }, { "epoch": 3.763953983809118, "grad_norm": 2.0625, "learning_rate": 4.8118541341153276e-05, "loss": 1.307836151123047, "step": 44170 }, { "epoch": 3.7648061354921176, "grad_norm": 2.109375, "learning_rate": 4.8117490788101525e-05, "loss": 1.328737163543701, "step": 44180 }, { "epoch": 3.765658287175117, "grad_norm": 2.328125, "learning_rate": 4.8116439953306836e-05, "loss": 1.133058452606201, "step": 44190 }, { "epoch": 3.7665104388581168, "grad_norm": 2.125, "learning_rate": 4.8115388836782e-05, "loss": 1.3836886405944824, "step": 44200 }, { "epoch": 3.7673625905411163, "grad_norm": 2.078125, "learning_rate": 4.811433743853984e-05, "loss": 1.4066531181335449, "step": 44210 }, { "epoch": 3.768214742224116, "grad_norm": 3.09375, "learning_rate": 4.8113285758593164e-05, "loss": 1.3246965408325195, "step": 44220 }, { "epoch": 3.7690668939071155, "grad_norm": 2.828125, "learning_rate": 4.811223379695479e-05, "loss": 1.3602824211120605, "step": 44230 }, { "epoch": 3.769919045590115, "grad_norm": 1.8203125, "learning_rate": 4.811118155363753e-05, "loss": 1.296179962158203, "step": 44240 }, { "epoch": 3.7707711972731146, "grad_norm": 2.015625, "learning_rate": 4.811012902865423e-05, "loss": 1.5014589309692383, "step": 44250 }, { "epoch": 3.771623348956114, "grad_norm": 2.578125, "learning_rate": 4.810907622201769e-05, "loss": 1.3141422271728516, "step": 44260 }, { "epoch": 3.7724755006391137, "grad_norm": 2.453125, "learning_rate": 4.810802313374075e-05, "loss": 1.2839942932128907, "step": 44270 }, { "epoch": 3.7733276523221133, "grad_norm": 2.125, "learning_rate": 4.810696976383626e-05, "loss": 1.5393923759460448, "step": 44280 }, { "epoch": 3.774179804005113, "grad_norm": 2.703125, "learning_rate": 4.810591611231705e-05, "loss": 1.3923325538635254, "step": 44290 }, { "epoch": 3.7750319556881125, "grad_norm": 1.765625, "learning_rate": 4.8104862179195956e-05, "loss": 1.3034440994262695, "step": 44300 }, { "epoch": 3.775884107371112, "grad_norm": 2.65625, "learning_rate": 4.810380796448582e-05, "loss": 1.3930720329284667, "step": 44310 }, { "epoch": 3.7767362590541116, "grad_norm": 1.828125, "learning_rate": 4.8102753468199496e-05, "loss": 1.3802026748657226, "step": 44320 }, { "epoch": 3.777588410737111, "grad_norm": 4.5625, "learning_rate": 4.810169869034984e-05, "loss": 1.4239236831665039, "step": 44330 }, { "epoch": 3.7784405624201107, "grad_norm": 2.296875, "learning_rate": 4.8100643630949695e-05, "loss": 1.3312016487121583, "step": 44340 }, { "epoch": 3.7792927141031103, "grad_norm": 3.40625, "learning_rate": 4.809958829001193e-05, "loss": 1.4441858291625977, "step": 44350 }, { "epoch": 3.78014486578611, "grad_norm": 3.421875, "learning_rate": 4.80985326675494e-05, "loss": 1.4345174789428712, "step": 44360 }, { "epoch": 3.7809970174691094, "grad_norm": 2.078125, "learning_rate": 4.809747676357498e-05, "loss": 1.3860971450805664, "step": 44370 }, { "epoch": 3.781849169152109, "grad_norm": 2.078125, "learning_rate": 4.809642057810152e-05, "loss": 1.469491195678711, "step": 44380 }, { "epoch": 3.7827013208351086, "grad_norm": 1.8359375, "learning_rate": 4.809536411114192e-05, "loss": 1.32332763671875, "step": 44390 }, { "epoch": 3.783553472518108, "grad_norm": 2.9375, "learning_rate": 4.8094307362709026e-05, "loss": 1.4294245719909668, "step": 44400 }, { "epoch": 3.7844056242011077, "grad_norm": 2.28125, "learning_rate": 4.809325033281574e-05, "loss": 1.3111827850341797, "step": 44410 }, { "epoch": 3.7852577758841073, "grad_norm": 1.984375, "learning_rate": 4.809219302147493e-05, "loss": 1.418265438079834, "step": 44420 }, { "epoch": 3.786109927567107, "grad_norm": 2.53125, "learning_rate": 4.809113542869948e-05, "loss": 1.2567232131958008, "step": 44430 }, { "epoch": 3.7869620792501064, "grad_norm": 3.5, "learning_rate": 4.8090077554502305e-05, "loss": 1.4070615768432617, "step": 44440 }, { "epoch": 3.787814230933106, "grad_norm": 4.1875, "learning_rate": 4.8089019398896274e-05, "loss": 1.2462706565856934, "step": 44450 }, { "epoch": 3.7886663826161056, "grad_norm": 2.140625, "learning_rate": 4.808796096189428e-05, "loss": 1.2329950332641602, "step": 44460 }, { "epoch": 3.789518534299105, "grad_norm": 3.328125, "learning_rate": 4.808690224350924e-05, "loss": 1.4807899475097657, "step": 44470 }, { "epoch": 3.7903706859821047, "grad_norm": 3.5625, "learning_rate": 4.808584324375405e-05, "loss": 1.3377829551696778, "step": 44480 }, { "epoch": 3.7912228376651043, "grad_norm": 2.40625, "learning_rate": 4.808478396264161e-05, "loss": 1.3061668395996093, "step": 44490 }, { "epoch": 3.792074989348104, "grad_norm": 1.8828125, "learning_rate": 4.808372440018484e-05, "loss": 1.2666804313659668, "step": 44500 }, { "epoch": 3.7929271410311034, "grad_norm": 2.71875, "learning_rate": 4.808266455639664e-05, "loss": 1.4326581001281737, "step": 44510 }, { "epoch": 3.793779292714103, "grad_norm": 1.9765625, "learning_rate": 4.808160443128994e-05, "loss": 1.0248701095581054, "step": 44520 }, { "epoch": 3.7946314443971025, "grad_norm": 1.96875, "learning_rate": 4.808054402487766e-05, "loss": 1.177776050567627, "step": 44530 }, { "epoch": 3.795483596080102, "grad_norm": 4.5625, "learning_rate": 4.8079483337172706e-05, "loss": 1.2812885284423827, "step": 44540 }, { "epoch": 3.7963357477631017, "grad_norm": 1.9765625, "learning_rate": 4.807842236818803e-05, "loss": 1.4272770881652832, "step": 44550 }, { "epoch": 3.7971878994461012, "grad_norm": 1.609375, "learning_rate": 4.8077361117936546e-05, "loss": 1.2794212341308593, "step": 44560 }, { "epoch": 3.798040051129101, "grad_norm": 2.25, "learning_rate": 4.8076299586431196e-05, "loss": 1.4552925109863282, "step": 44570 }, { "epoch": 3.798892202812101, "grad_norm": 2.203125, "learning_rate": 4.8075237773684906e-05, "loss": 1.311351490020752, "step": 44580 }, { "epoch": 3.7997443544951004, "grad_norm": 2.4375, "learning_rate": 4.8074175679710634e-05, "loss": 1.3263572692871093, "step": 44590 }, { "epoch": 3.8005965061781, "grad_norm": 2.28125, "learning_rate": 4.80731133045213e-05, "loss": 1.2738335609436036, "step": 44600 }, { "epoch": 3.8014486578610995, "grad_norm": 2.484375, "learning_rate": 4.807205064812988e-05, "loss": 1.2291921615600585, "step": 44610 }, { "epoch": 3.802300809544099, "grad_norm": 2.34375, "learning_rate": 4.807098771054931e-05, "loss": 1.397540283203125, "step": 44620 }, { "epoch": 3.8031529612270987, "grad_norm": 3.46875, "learning_rate": 4.806992449179254e-05, "loss": 1.3250956535339355, "step": 44630 }, { "epoch": 3.8040051129100982, "grad_norm": 1.9921875, "learning_rate": 4.806886099187253e-05, "loss": 1.4055066108703613, "step": 44640 }, { "epoch": 3.804857264593098, "grad_norm": 2.546875, "learning_rate": 4.8067797210802254e-05, "loss": 1.3206924438476562, "step": 44650 }, { "epoch": 3.8057094162760974, "grad_norm": 2.671875, "learning_rate": 4.806673314859467e-05, "loss": 1.4249643325805663, "step": 44660 }, { "epoch": 3.806561567959097, "grad_norm": 2.109375, "learning_rate": 4.806566880526274e-05, "loss": 1.2808197021484375, "step": 44670 }, { "epoch": 3.8074137196420965, "grad_norm": 2.484375, "learning_rate": 4.806460418081944e-05, "loss": 1.2351970672607422, "step": 44680 }, { "epoch": 3.808265871325096, "grad_norm": 2.34375, "learning_rate": 4.8063539275277744e-05, "loss": 1.435169792175293, "step": 44690 }, { "epoch": 3.8091180230080957, "grad_norm": 2.515625, "learning_rate": 4.806247408865063e-05, "loss": 1.4004338264465332, "step": 44700 }, { "epoch": 3.8099701746910952, "grad_norm": 1.8984375, "learning_rate": 4.806140862095108e-05, "loss": 1.3533451080322265, "step": 44710 }, { "epoch": 3.810822326374095, "grad_norm": 3.53125, "learning_rate": 4.806034287219208e-05, "loss": 1.2000778198242188, "step": 44720 }, { "epoch": 3.8116744780570944, "grad_norm": 2.9375, "learning_rate": 4.805927684238662e-05, "loss": 1.2328198432922364, "step": 44730 }, { "epoch": 3.812526629740094, "grad_norm": 2.0, "learning_rate": 4.80582105315477e-05, "loss": 1.2602934837341309, "step": 44740 }, { "epoch": 3.8133787814230935, "grad_norm": 2.34375, "learning_rate": 4.80571439396883e-05, "loss": 1.4559345245361328, "step": 44750 }, { "epoch": 3.814230933106093, "grad_norm": 2.125, "learning_rate": 4.8056077066821426e-05, "loss": 1.5134991645812987, "step": 44760 }, { "epoch": 3.8150830847890926, "grad_norm": 2.609375, "learning_rate": 4.805500991296008e-05, "loss": 1.360158920288086, "step": 44770 }, { "epoch": 3.815935236472092, "grad_norm": 1.7109375, "learning_rate": 4.805394247811727e-05, "loss": 1.3965364456176759, "step": 44780 }, { "epoch": 3.8167873881550918, "grad_norm": 2.578125, "learning_rate": 4.8052874762306e-05, "loss": 1.2786026000976562, "step": 44790 }, { "epoch": 3.8176395398380913, "grad_norm": 2.546875, "learning_rate": 4.805180676553929e-05, "loss": 1.2927801132202148, "step": 44800 }, { "epoch": 3.818491691521091, "grad_norm": 2.03125, "learning_rate": 4.805073848783015e-05, "loss": 1.2961416244506836, "step": 44810 }, { "epoch": 3.8193438432040905, "grad_norm": 2.890625, "learning_rate": 4.80496699291916e-05, "loss": 1.352797031402588, "step": 44820 }, { "epoch": 3.82019599488709, "grad_norm": 2.390625, "learning_rate": 4.804860108963667e-05, "loss": 1.4261345863342285, "step": 44830 }, { "epoch": 3.8210481465700896, "grad_norm": 1.671875, "learning_rate": 4.804753196917838e-05, "loss": 1.5462663650512696, "step": 44840 }, { "epoch": 3.821900298253089, "grad_norm": 2.4375, "learning_rate": 4.8046462567829764e-05, "loss": 1.2224921226501464, "step": 44850 }, { "epoch": 3.8227524499360888, "grad_norm": 4.125, "learning_rate": 4.804539288560384e-05, "loss": 1.1949677467346191, "step": 44860 }, { "epoch": 3.8236046016190883, "grad_norm": 1.75, "learning_rate": 4.8044322922513676e-05, "loss": 1.3798480033874512, "step": 44870 }, { "epoch": 3.824456753302088, "grad_norm": 2.34375, "learning_rate": 4.804325267857228e-05, "loss": 1.461019229888916, "step": 44880 }, { "epoch": 3.8253089049850875, "grad_norm": 2.328125, "learning_rate": 4.804218215379271e-05, "loss": 1.4186522483825683, "step": 44890 }, { "epoch": 3.826161056668087, "grad_norm": 1.90625, "learning_rate": 4.804111134818802e-05, "loss": 1.3083230972290039, "step": 44900 }, { "epoch": 3.8270132083510866, "grad_norm": 2.171875, "learning_rate": 4.8040040261771247e-05, "loss": 1.3242752075195312, "step": 44910 }, { "epoch": 3.827865360034086, "grad_norm": 1.9609375, "learning_rate": 4.8038968894555444e-05, "loss": 1.3217623710632325, "step": 44920 }, { "epoch": 3.8287175117170857, "grad_norm": 2.90625, "learning_rate": 4.8037897246553674e-05, "loss": 1.3684128761291503, "step": 44930 }, { "epoch": 3.8295696634000853, "grad_norm": 1.765625, "learning_rate": 4.803682531777901e-05, "loss": 1.3249093055725099, "step": 44940 }, { "epoch": 3.830421815083085, "grad_norm": 2.28125, "learning_rate": 4.80357531082445e-05, "loss": 1.3040740013122558, "step": 44950 }, { "epoch": 3.8312739667660844, "grad_norm": 2.703125, "learning_rate": 4.803468061796321e-05, "loss": 1.3398776054382324, "step": 44960 }, { "epoch": 3.832126118449084, "grad_norm": 2.234375, "learning_rate": 4.8033607846948216e-05, "loss": 1.3560062408447267, "step": 44970 }, { "epoch": 3.8329782701320836, "grad_norm": 2.90625, "learning_rate": 4.8032534795212594e-05, "loss": 1.3470454216003418, "step": 44980 }, { "epoch": 3.833830421815083, "grad_norm": 1.84375, "learning_rate": 4.803146146276942e-05, "loss": 1.3231410026550292, "step": 44990 }, { "epoch": 3.8346825734980827, "grad_norm": 1.6328125, "learning_rate": 4.803038784963178e-05, "loss": 1.320560359954834, "step": 45000 }, { "epoch": 3.8355347251810823, "grad_norm": 2.125, "learning_rate": 4.802931395581275e-05, "loss": 1.3452054977416992, "step": 45010 }, { "epoch": 3.836386876864082, "grad_norm": 1.8515625, "learning_rate": 4.8028239781325426e-05, "loss": 1.3202725410461427, "step": 45020 }, { "epoch": 3.8372390285470814, "grad_norm": 2.109375, "learning_rate": 4.8027165326182885e-05, "loss": 1.3371337890625, "step": 45030 }, { "epoch": 3.838091180230081, "grad_norm": 2.671875, "learning_rate": 4.8026090590398244e-05, "loss": 1.165116024017334, "step": 45040 }, { "epoch": 3.8389433319130806, "grad_norm": 3.25, "learning_rate": 4.802501557398459e-05, "loss": 1.2768226623535157, "step": 45050 }, { "epoch": 3.83979548359608, "grad_norm": 2.453125, "learning_rate": 4.8023940276955015e-05, "loss": 1.3502864837646484, "step": 45060 }, { "epoch": 3.8406476352790797, "grad_norm": 1.6796875, "learning_rate": 4.802286469932264e-05, "loss": 1.4213544845581054, "step": 45070 }, { "epoch": 3.8414997869620793, "grad_norm": 2.5, "learning_rate": 4.802178884110057e-05, "loss": 1.4124296188354493, "step": 45080 }, { "epoch": 3.842351938645079, "grad_norm": 3.03125, "learning_rate": 4.8020712702301903e-05, "loss": 1.4602312088012694, "step": 45090 }, { "epoch": 3.8432040903280784, "grad_norm": 2.84375, "learning_rate": 4.801963628293977e-05, "loss": 1.1873772621154786, "step": 45100 }, { "epoch": 3.844056242011078, "grad_norm": 1.625, "learning_rate": 4.801855958302729e-05, "loss": 1.4325376510620118, "step": 45110 }, { "epoch": 3.8449083936940776, "grad_norm": 2.828125, "learning_rate": 4.801748260257758e-05, "loss": 1.3461426734924316, "step": 45120 }, { "epoch": 3.845760545377077, "grad_norm": 4.125, "learning_rate": 4.8016405341603756e-05, "loss": 1.285301113128662, "step": 45130 }, { "epoch": 3.8466126970600767, "grad_norm": 2.640625, "learning_rate": 4.801532780011897e-05, "loss": 1.316441249847412, "step": 45140 }, { "epoch": 3.8474648487430763, "grad_norm": 3.6875, "learning_rate": 4.801424997813633e-05, "loss": 1.2679938316345214, "step": 45150 }, { "epoch": 3.848317000426076, "grad_norm": 2.46875, "learning_rate": 4.8013171875668994e-05, "loss": 1.3301275253295899, "step": 45160 }, { "epoch": 3.8491691521090754, "grad_norm": 2.21875, "learning_rate": 4.801209349273008e-05, "loss": 1.3409945487976074, "step": 45170 }, { "epoch": 3.850021303792075, "grad_norm": 2.25, "learning_rate": 4.8011014829332754e-05, "loss": 1.3622108459472657, "step": 45180 }, { "epoch": 3.8508734554750745, "grad_norm": 2.78125, "learning_rate": 4.8009935885490145e-05, "loss": 1.2623950958251953, "step": 45190 }, { "epoch": 3.851725607158074, "grad_norm": 1.953125, "learning_rate": 4.800885666121539e-05, "loss": 1.348482894897461, "step": 45200 }, { "epoch": 3.8525777588410737, "grad_norm": 2.5625, "learning_rate": 4.800777715652168e-05, "loss": 1.490289306640625, "step": 45210 }, { "epoch": 3.8534299105240732, "grad_norm": 1.6484375, "learning_rate": 4.800669737142215e-05, "loss": 1.4955899238586425, "step": 45220 }, { "epoch": 3.854282062207073, "grad_norm": 2.359375, "learning_rate": 4.800561730592995e-05, "loss": 1.430125617980957, "step": 45230 }, { "epoch": 3.8551342138900724, "grad_norm": 2.296875, "learning_rate": 4.800453696005826e-05, "loss": 1.3287722587585449, "step": 45240 }, { "epoch": 3.855986365573072, "grad_norm": 2.65625, "learning_rate": 4.800345633382024e-05, "loss": 1.3139598846435547, "step": 45250 }, { "epoch": 3.8568385172560715, "grad_norm": 2.6875, "learning_rate": 4.800237542722906e-05, "loss": 1.3851067543029785, "step": 45260 }, { "epoch": 3.857690668939071, "grad_norm": 2.96875, "learning_rate": 4.800129424029789e-05, "loss": 1.2516231536865234, "step": 45270 }, { "epoch": 3.8585428206220707, "grad_norm": 3.34375, "learning_rate": 4.800021277303991e-05, "loss": 1.211852741241455, "step": 45280 }, { "epoch": 3.8593949723050702, "grad_norm": 1.3515625, "learning_rate": 4.79991310254683e-05, "loss": 1.4581844329833984, "step": 45290 }, { "epoch": 3.86024712398807, "grad_norm": 1.7890625, "learning_rate": 4.7998048997596254e-05, "loss": 1.2819475173950194, "step": 45300 }, { "epoch": 3.8610992756710694, "grad_norm": 2.34375, "learning_rate": 4.799696668943694e-05, "loss": 1.2349209785461426, "step": 45310 }, { "epoch": 3.861951427354069, "grad_norm": 2.8125, "learning_rate": 4.7995884101003565e-05, "loss": 1.40596923828125, "step": 45320 }, { "epoch": 3.8628035790370685, "grad_norm": 3.140625, "learning_rate": 4.7994801232309315e-05, "loss": 1.367288589477539, "step": 45330 }, { "epoch": 3.863655730720068, "grad_norm": 1.984375, "learning_rate": 4.799371808336739e-05, "loss": 1.373025608062744, "step": 45340 }, { "epoch": 3.8645078824030676, "grad_norm": 3.640625, "learning_rate": 4.799263465419098e-05, "loss": 1.3559101104736329, "step": 45350 }, { "epoch": 3.865360034086067, "grad_norm": 3.3125, "learning_rate": 4.799155094479331e-05, "loss": 1.2157044410705566, "step": 45360 }, { "epoch": 3.866212185769067, "grad_norm": 3.5625, "learning_rate": 4.799046695518758e-05, "loss": 1.343609619140625, "step": 45370 }, { "epoch": 3.8670643374520663, "grad_norm": 3.15625, "learning_rate": 4.798938268538698e-05, "loss": 1.3140989303588868, "step": 45380 }, { "epoch": 3.867916489135066, "grad_norm": 2.125, "learning_rate": 4.798829813540475e-05, "loss": 1.4001953125, "step": 45390 }, { "epoch": 3.8687686408180655, "grad_norm": 2.359375, "learning_rate": 4.79872133052541e-05, "loss": 1.453671646118164, "step": 45400 }, { "epoch": 3.869620792501065, "grad_norm": 2.625, "learning_rate": 4.798612819494825e-05, "loss": 1.2649489402770997, "step": 45410 }, { "epoch": 3.8704729441840646, "grad_norm": 1.9921875, "learning_rate": 4.798504280450044e-05, "loss": 1.4497889518737792, "step": 45420 }, { "epoch": 3.871325095867064, "grad_norm": 2.484375, "learning_rate": 4.798395713392386e-05, "loss": 1.3769402503967285, "step": 45430 }, { "epoch": 3.8721772475500638, "grad_norm": 2.390625, "learning_rate": 4.798287118323178e-05, "loss": 1.291226577758789, "step": 45440 }, { "epoch": 3.8730293992330633, "grad_norm": 2.203125, "learning_rate": 4.798178495243742e-05, "loss": 1.3342859268188476, "step": 45450 }, { "epoch": 3.873881550916063, "grad_norm": 2.265625, "learning_rate": 4.7980698441554014e-05, "loss": 1.486951732635498, "step": 45460 }, { "epoch": 3.8747337025990625, "grad_norm": 2.6875, "learning_rate": 4.797961165059481e-05, "loss": 1.391047191619873, "step": 45470 }, { "epoch": 3.875585854282062, "grad_norm": 1.734375, "learning_rate": 4.797852457957306e-05, "loss": 1.4302387237548828, "step": 45480 }, { "epoch": 3.8764380059650616, "grad_norm": 2.796875, "learning_rate": 4.7977437228502e-05, "loss": 1.3522292137145997, "step": 45490 }, { "epoch": 3.877290157648061, "grad_norm": 3.40625, "learning_rate": 4.797634959739488e-05, "loss": 1.225679588317871, "step": 45500 }, { "epoch": 3.8781423093310607, "grad_norm": 2.28125, "learning_rate": 4.7975261686264963e-05, "loss": 1.3950630187988282, "step": 45510 }, { "epoch": 3.8789944610140603, "grad_norm": 2.03125, "learning_rate": 4.7974173495125505e-05, "loss": 1.2703921318054199, "step": 45520 }, { "epoch": 3.87984661269706, "grad_norm": 1.9296875, "learning_rate": 4.7973085023989784e-05, "loss": 1.3929905891418457, "step": 45530 }, { "epoch": 3.8806987643800595, "grad_norm": 2.421875, "learning_rate": 4.797199627287104e-05, "loss": 1.3688940048217773, "step": 45540 }, { "epoch": 3.881550916063059, "grad_norm": 2.625, "learning_rate": 4.797090724178255e-05, "loss": 1.301159381866455, "step": 45550 }, { "epoch": 3.8824030677460586, "grad_norm": 1.8125, "learning_rate": 4.79698179307376e-05, "loss": 1.4166165351867677, "step": 45560 }, { "epoch": 3.883255219429058, "grad_norm": 2.265625, "learning_rate": 4.7968728339749445e-05, "loss": 1.2792539596557617, "step": 45570 }, { "epoch": 3.8841073711120577, "grad_norm": 2.046875, "learning_rate": 4.796763846883139e-05, "loss": 1.4258124351501464, "step": 45580 }, { "epoch": 3.8849595227950573, "grad_norm": 3.375, "learning_rate": 4.7966548317996694e-05, "loss": 1.1888300895690918, "step": 45590 }, { "epoch": 3.885811674478057, "grad_norm": 2.03125, "learning_rate": 4.796545788725865e-05, "loss": 1.5275177955627441, "step": 45600 }, { "epoch": 3.8866638261610564, "grad_norm": 1.859375, "learning_rate": 4.7964367176630555e-05, "loss": 1.4532160758972168, "step": 45610 }, { "epoch": 3.887515977844056, "grad_norm": 1.609375, "learning_rate": 4.7963276186125695e-05, "loss": 1.357813262939453, "step": 45620 }, { "epoch": 3.8883681295270556, "grad_norm": 2.203125, "learning_rate": 4.796218491575737e-05, "loss": 1.341160202026367, "step": 45630 }, { "epoch": 3.889220281210055, "grad_norm": 3.265625, "learning_rate": 4.7961093365538876e-05, "loss": 1.3443613052368164, "step": 45640 }, { "epoch": 3.8900724328930547, "grad_norm": 1.875, "learning_rate": 4.796000153548352e-05, "loss": 1.4692588806152345, "step": 45650 }, { "epoch": 3.8909245845760543, "grad_norm": 2.265625, "learning_rate": 4.795890942560461e-05, "loss": 1.2306757926940919, "step": 45660 }, { "epoch": 3.891776736259054, "grad_norm": 3.03125, "learning_rate": 4.795781703591545e-05, "loss": 1.279228973388672, "step": 45670 }, { "epoch": 3.8926288879420534, "grad_norm": 2.59375, "learning_rate": 4.795672436642935e-05, "loss": 1.479010772705078, "step": 45680 }, { "epoch": 3.893481039625053, "grad_norm": 2.84375, "learning_rate": 4.7955631417159644e-05, "loss": 1.3582714080810547, "step": 45690 }, { "epoch": 3.894333191308053, "grad_norm": 2.265625, "learning_rate": 4.7954538188119635e-05, "loss": 1.2951418876647949, "step": 45700 }, { "epoch": 3.8951853429910526, "grad_norm": 1.8359375, "learning_rate": 4.795344467932265e-05, "loss": 1.3161788940429688, "step": 45710 }, { "epoch": 3.896037494674052, "grad_norm": 2.40625, "learning_rate": 4.7952350890782025e-05, "loss": 1.150360679626465, "step": 45720 }, { "epoch": 3.8968896463570517, "grad_norm": 2.203125, "learning_rate": 4.795125682251108e-05, "loss": 1.3768904685974122, "step": 45730 }, { "epoch": 3.8977417980400513, "grad_norm": 2.09375, "learning_rate": 4.7950162474523154e-05, "loss": 1.4243932723999024, "step": 45740 }, { "epoch": 3.898593949723051, "grad_norm": 2.15625, "learning_rate": 4.7949067846831585e-05, "loss": 1.2191116333007812, "step": 45750 }, { "epoch": 3.8994461014060504, "grad_norm": 1.6171875, "learning_rate": 4.7947972939449716e-05, "loss": 1.4836724281311036, "step": 45760 }, { "epoch": 3.90029825308905, "grad_norm": 1.9375, "learning_rate": 4.7946877752390883e-05, "loss": 1.4848051071166992, "step": 45770 }, { "epoch": 3.9011504047720496, "grad_norm": 2.359375, "learning_rate": 4.7945782285668436e-05, "loss": 1.5294818878173828, "step": 45780 }, { "epoch": 3.902002556455049, "grad_norm": 1.9609375, "learning_rate": 4.794468653929574e-05, "loss": 1.3793517112731934, "step": 45790 }, { "epoch": 3.9028547081380487, "grad_norm": 1.734375, "learning_rate": 4.7943590513286115e-05, "loss": 1.1797484397888183, "step": 45800 }, { "epoch": 3.9037068598210483, "grad_norm": 2.109375, "learning_rate": 4.794249420765296e-05, "loss": 1.363451862335205, "step": 45810 }, { "epoch": 3.904559011504048, "grad_norm": 1.875, "learning_rate": 4.7941397622409615e-05, "loss": 1.3394694328308105, "step": 45820 }, { "epoch": 3.9054111631870474, "grad_norm": 2.40625, "learning_rate": 4.794030075756945e-05, "loss": 1.499141311645508, "step": 45830 }, { "epoch": 3.906263314870047, "grad_norm": 2.078125, "learning_rate": 4.793920361314582e-05, "loss": 1.22957763671875, "step": 45840 }, { "epoch": 3.9071154665530465, "grad_norm": 2.890625, "learning_rate": 4.7938106189152104e-05, "loss": 1.22531156539917, "step": 45850 }, { "epoch": 3.907967618236046, "grad_norm": 2.453125, "learning_rate": 4.7937008485601685e-05, "loss": 1.1344141960144043, "step": 45860 }, { "epoch": 3.9088197699190457, "grad_norm": 1.75, "learning_rate": 4.793591050250794e-05, "loss": 1.267601776123047, "step": 45870 }, { "epoch": 3.9096719216020452, "grad_norm": 2.609375, "learning_rate": 4.793481223988424e-05, "loss": 1.3197052001953125, "step": 45880 }, { "epoch": 3.910524073285045, "grad_norm": 1.6484375, "learning_rate": 4.7933713697743973e-05, "loss": 1.3043647766113282, "step": 45890 }, { "epoch": 3.9113762249680444, "grad_norm": 2.375, "learning_rate": 4.793261487610054e-05, "loss": 1.2173612594604493, "step": 45900 }, { "epoch": 3.912228376651044, "grad_norm": 2.96875, "learning_rate": 4.7931515774967304e-05, "loss": 1.210393524169922, "step": 45910 }, { "epoch": 3.9130805283340435, "grad_norm": 1.859375, "learning_rate": 4.79304163943577e-05, "loss": 1.4172948837280273, "step": 45920 }, { "epoch": 3.913932680017043, "grad_norm": 2.859375, "learning_rate": 4.792931673428509e-05, "loss": 1.2232794761657715, "step": 45930 }, { "epoch": 3.9147848317000427, "grad_norm": 4.25, "learning_rate": 4.792821679476291e-05, "loss": 1.2264112472534179, "step": 45940 }, { "epoch": 3.9156369833830422, "grad_norm": 2.265625, "learning_rate": 4.792711657580453e-05, "loss": 1.3129214286804198, "step": 45950 }, { "epoch": 3.916489135066042, "grad_norm": 2.140625, "learning_rate": 4.7926016077423384e-05, "loss": 1.2824592590332031, "step": 45960 }, { "epoch": 3.9173412867490414, "grad_norm": 2.84375, "learning_rate": 4.792491529963287e-05, "loss": 1.5331445693969727, "step": 45970 }, { "epoch": 3.918193438432041, "grad_norm": 2.328125, "learning_rate": 4.792381424244642e-05, "loss": 1.3110642433166504, "step": 45980 }, { "epoch": 3.9190455901150405, "grad_norm": 2.1875, "learning_rate": 4.792271290587744e-05, "loss": 1.2622344970703125, "step": 45990 }, { "epoch": 3.91989774179804, "grad_norm": 1.75, "learning_rate": 4.792161128993935e-05, "loss": 1.4776005744934082, "step": 46000 }, { "epoch": 3.9207498934810396, "grad_norm": 2.5625, "learning_rate": 4.792050939464559e-05, "loss": 1.284937858581543, "step": 46010 }, { "epoch": 3.921602045164039, "grad_norm": 2.296875, "learning_rate": 4.791940722000957e-05, "loss": 1.459500503540039, "step": 46020 }, { "epoch": 3.9224541968470388, "grad_norm": 1.6953125, "learning_rate": 4.791830476604474e-05, "loss": 1.2751005172729493, "step": 46030 }, { "epoch": 3.9233063485300383, "grad_norm": 2.703125, "learning_rate": 4.791720203276453e-05, "loss": 1.1220364570617676, "step": 46040 }, { "epoch": 3.924158500213038, "grad_norm": 3.578125, "learning_rate": 4.791609902018238e-05, "loss": 1.3800379753112793, "step": 46050 }, { "epoch": 3.9250106518960375, "grad_norm": 3.40625, "learning_rate": 4.791499572831173e-05, "loss": 1.3245964050292969, "step": 46060 }, { "epoch": 3.925862803579037, "grad_norm": 3.34375, "learning_rate": 4.791389215716603e-05, "loss": 1.2231401443481444, "step": 46070 }, { "epoch": 3.9267149552620366, "grad_norm": 1.78125, "learning_rate": 4.791278830675873e-05, "loss": 1.4387187957763672, "step": 46080 }, { "epoch": 3.927567106945036, "grad_norm": 2.75, "learning_rate": 4.791168417710327e-05, "loss": 1.3993596076965331, "step": 46090 }, { "epoch": 3.9284192586280358, "grad_norm": 1.953125, "learning_rate": 4.791057976821313e-05, "loss": 1.3409707069396972, "step": 46100 }, { "epoch": 3.9292714103110353, "grad_norm": 1.9609375, "learning_rate": 4.790947508010175e-05, "loss": 1.4200642585754395, "step": 46110 }, { "epoch": 3.930123561994035, "grad_norm": 2.53125, "learning_rate": 4.79083701127826e-05, "loss": 1.3534883499145507, "step": 46120 }, { "epoch": 3.9309757136770345, "grad_norm": 1.859375, "learning_rate": 4.790726486626915e-05, "loss": 1.3812894821166992, "step": 46130 }, { "epoch": 3.931827865360034, "grad_norm": 2.203125, "learning_rate": 4.790615934057487e-05, "loss": 1.396688175201416, "step": 46140 }, { "epoch": 3.9326800170430336, "grad_norm": 3.140625, "learning_rate": 4.7905053535713226e-05, "loss": 1.28345947265625, "step": 46150 }, { "epoch": 3.933532168726033, "grad_norm": 2.625, "learning_rate": 4.79039474516977e-05, "loss": 1.3235675811767578, "step": 46160 }, { "epoch": 3.9343843204090327, "grad_norm": 3.0, "learning_rate": 4.790284108854178e-05, "loss": 1.4838898658752442, "step": 46170 }, { "epoch": 3.9352364720920323, "grad_norm": 2.5, "learning_rate": 4.790173444625894e-05, "loss": 1.3193022727966308, "step": 46180 }, { "epoch": 3.936088623775032, "grad_norm": 4.21875, "learning_rate": 4.7900627524862665e-05, "loss": 1.2352548599243165, "step": 46190 }, { "epoch": 3.9369407754580314, "grad_norm": 2.1875, "learning_rate": 4.7899520324366455e-05, "loss": 1.3460665702819825, "step": 46200 }, { "epoch": 3.937792927141031, "grad_norm": 1.890625, "learning_rate": 4.78984128447838e-05, "loss": 1.3990689277648927, "step": 46210 }, { "epoch": 3.9386450788240306, "grad_norm": 2.15625, "learning_rate": 4.789730508612819e-05, "loss": 1.3491523742675782, "step": 46220 }, { "epoch": 3.93949723050703, "grad_norm": 2.015625, "learning_rate": 4.789619704841313e-05, "loss": 1.479433536529541, "step": 46230 }, { "epoch": 3.9403493821900297, "grad_norm": 2.21875, "learning_rate": 4.789508873165213e-05, "loss": 1.3077765464782716, "step": 46240 }, { "epoch": 3.9412015338730293, "grad_norm": 1.8203125, "learning_rate": 4.7893980135858695e-05, "loss": 1.1420862197875976, "step": 46250 }, { "epoch": 3.942053685556029, "grad_norm": 3.453125, "learning_rate": 4.789287126104633e-05, "loss": 1.4559080123901367, "step": 46260 }, { "epoch": 3.9429058372390284, "grad_norm": 2.71875, "learning_rate": 4.7891762107228554e-05, "loss": 1.4478361129760742, "step": 46270 }, { "epoch": 3.943757988922028, "grad_norm": 2.015625, "learning_rate": 4.789065267441889e-05, "loss": 1.429545783996582, "step": 46280 }, { "epoch": 3.9446101406050276, "grad_norm": 4.0625, "learning_rate": 4.788954296263085e-05, "loss": 1.148579978942871, "step": 46290 }, { "epoch": 3.945462292288027, "grad_norm": 3.15625, "learning_rate": 4.788843297187796e-05, "loss": 1.2534066200256349, "step": 46300 }, { "epoch": 3.9463144439710267, "grad_norm": 3.640625, "learning_rate": 4.788732270217375e-05, "loss": 1.4438952445983886, "step": 46310 }, { "epoch": 3.9471665956540263, "grad_norm": 2.578125, "learning_rate": 4.788621215353176e-05, "loss": 1.2837782859802247, "step": 46320 }, { "epoch": 3.948018747337026, "grad_norm": 3.375, "learning_rate": 4.788510132596551e-05, "loss": 1.3076828002929688, "step": 46330 }, { "epoch": 3.9488708990200254, "grad_norm": 2.796875, "learning_rate": 4.788399021948854e-05, "loss": 1.4119403839111329, "step": 46340 }, { "epoch": 3.949723050703025, "grad_norm": 3.546875, "learning_rate": 4.78828788341144e-05, "loss": 1.2963871955871582, "step": 46350 }, { "epoch": 3.9505752023860246, "grad_norm": 2.359375, "learning_rate": 4.788176716985663e-05, "loss": 1.2906658172607421, "step": 46360 }, { "epoch": 3.951427354069024, "grad_norm": 3.21875, "learning_rate": 4.7880655226728786e-05, "loss": 1.3980938911437988, "step": 46370 }, { "epoch": 3.952279505752024, "grad_norm": 2.546875, "learning_rate": 4.7879543004744407e-05, "loss": 1.4571193695068358, "step": 46380 }, { "epoch": 3.9531316574350237, "grad_norm": 2.546875, "learning_rate": 4.787843050391705e-05, "loss": 1.346354866027832, "step": 46390 }, { "epoch": 3.9539838091180233, "grad_norm": 3.015625, "learning_rate": 4.787731772426028e-05, "loss": 1.3459988594055177, "step": 46400 }, { "epoch": 3.954835960801023, "grad_norm": 2.0, "learning_rate": 4.787620466578766e-05, "loss": 1.3521272659301757, "step": 46410 }, { "epoch": 3.9556881124840224, "grad_norm": 2.578125, "learning_rate": 4.787509132851276e-05, "loss": 1.382332706451416, "step": 46420 }, { "epoch": 3.956540264167022, "grad_norm": 2.75, "learning_rate": 4.787397771244912e-05, "loss": 1.3268569946289062, "step": 46430 }, { "epoch": 3.9573924158500215, "grad_norm": 2.265625, "learning_rate": 4.787286381761035e-05, "loss": 1.4461989402770996, "step": 46440 }, { "epoch": 3.958244567533021, "grad_norm": 2.078125, "learning_rate": 4.787174964401e-05, "loss": 1.4389985084533692, "step": 46450 }, { "epoch": 3.9590967192160207, "grad_norm": 1.71875, "learning_rate": 4.787063519166166e-05, "loss": 1.4133672714233398, "step": 46460 }, { "epoch": 3.9599488708990203, "grad_norm": 2.265625, "learning_rate": 4.786952046057891e-05, "loss": 1.4002874374389649, "step": 46470 }, { "epoch": 3.96080102258202, "grad_norm": 2.75, "learning_rate": 4.786840545077533e-05, "loss": 1.2474106788635253, "step": 46480 }, { "epoch": 3.9616531742650194, "grad_norm": 2.453125, "learning_rate": 4.786729016226452e-05, "loss": 1.320699977874756, "step": 46490 }, { "epoch": 3.962505325948019, "grad_norm": 2.140625, "learning_rate": 4.786617459506007e-05, "loss": 1.4183341026306153, "step": 46500 }, { "epoch": 3.9633574776310185, "grad_norm": 3.265625, "learning_rate": 4.786505874917556e-05, "loss": 1.3278009414672851, "step": 46510 }, { "epoch": 3.964209629314018, "grad_norm": 1.828125, "learning_rate": 4.786394262462461e-05, "loss": 1.2649542808532714, "step": 46520 }, { "epoch": 3.9650617809970177, "grad_norm": 2.703125, "learning_rate": 4.786282622142081e-05, "loss": 1.311422634124756, "step": 46530 }, { "epoch": 3.9659139326800172, "grad_norm": 3.828125, "learning_rate": 4.7861709539577776e-05, "loss": 1.3868885040283203, "step": 46540 }, { "epoch": 3.966766084363017, "grad_norm": 2.046875, "learning_rate": 4.7860592579109106e-05, "loss": 1.4160380363464355, "step": 46550 }, { "epoch": 3.9676182360460164, "grad_norm": 1.9296875, "learning_rate": 4.7859475340028424e-05, "loss": 1.4216861724853516, "step": 46560 }, { "epoch": 3.968470387729016, "grad_norm": 2.40625, "learning_rate": 4.785835782234934e-05, "loss": 1.3993709564208985, "step": 46570 }, { "epoch": 3.9693225394120155, "grad_norm": 1.953125, "learning_rate": 4.785724002608548e-05, "loss": 1.4857143402099608, "step": 46580 }, { "epoch": 3.970174691095015, "grad_norm": 1.65625, "learning_rate": 4.785612195125046e-05, "loss": 1.2880318641662598, "step": 46590 }, { "epoch": 3.9710268427780147, "grad_norm": 2.203125, "learning_rate": 4.785500359785791e-05, "loss": 1.2496407508850098, "step": 46600 }, { "epoch": 3.9718789944610142, "grad_norm": 3.265625, "learning_rate": 4.785388496592145e-05, "loss": 1.404492473602295, "step": 46610 }, { "epoch": 3.972731146144014, "grad_norm": 2.46875, "learning_rate": 4.785276605545473e-05, "loss": 1.416427230834961, "step": 46620 }, { "epoch": 3.9735832978270134, "grad_norm": 2.671875, "learning_rate": 4.785164686647138e-05, "loss": 1.3758439064025878, "step": 46630 }, { "epoch": 3.974435449510013, "grad_norm": 2.59375, "learning_rate": 4.785052739898503e-05, "loss": 1.274123764038086, "step": 46640 }, { "epoch": 3.9752876011930125, "grad_norm": 3.5, "learning_rate": 4.784940765300934e-05, "loss": 1.2422273635864258, "step": 46650 }, { "epoch": 3.976139752876012, "grad_norm": 3.75, "learning_rate": 4.784828762855794e-05, "loss": 1.2798423767089844, "step": 46660 }, { "epoch": 3.9769919045590116, "grad_norm": 2.546875, "learning_rate": 4.7847167325644495e-05, "loss": 1.2604593276977538, "step": 46670 }, { "epoch": 3.977844056242011, "grad_norm": 2.0625, "learning_rate": 4.784604674428265e-05, "loss": 1.4488536834716796, "step": 46680 }, { "epoch": 3.9786962079250108, "grad_norm": 3.890625, "learning_rate": 4.784492588448606e-05, "loss": 1.34143123626709, "step": 46690 }, { "epoch": 3.9795483596080103, "grad_norm": 1.8359375, "learning_rate": 4.78438047462684e-05, "loss": 1.405916404724121, "step": 46700 }, { "epoch": 3.98040051129101, "grad_norm": 2.53125, "learning_rate": 4.7842683329643324e-05, "loss": 1.2825270652770997, "step": 46710 }, { "epoch": 3.9812526629740095, "grad_norm": 3.15625, "learning_rate": 4.784156163462449e-05, "loss": 1.3758981704711915, "step": 46720 }, { "epoch": 3.982104814657009, "grad_norm": 2.265625, "learning_rate": 4.784043966122558e-05, "loss": 1.4631065368652343, "step": 46730 }, { "epoch": 3.9829569663400086, "grad_norm": 2.421875, "learning_rate": 4.7839317409460275e-05, "loss": 1.3302352905273438, "step": 46740 }, { "epoch": 3.983809118023008, "grad_norm": 2.53125, "learning_rate": 4.783819487934223e-05, "loss": 1.3787496566772461, "step": 46750 }, { "epoch": 3.9846612697060078, "grad_norm": 3.359375, "learning_rate": 4.7837072070885145e-05, "loss": 1.322239589691162, "step": 46760 }, { "epoch": 3.9855134213890073, "grad_norm": 2.453125, "learning_rate": 4.78359489841027e-05, "loss": 1.2797252655029296, "step": 46770 }, { "epoch": 3.986365573072007, "grad_norm": 1.84375, "learning_rate": 4.7834825619008574e-05, "loss": 1.4091143608093262, "step": 46780 }, { "epoch": 3.9872177247550065, "grad_norm": 2.5, "learning_rate": 4.783370197561647e-05, "loss": 1.3491018295288086, "step": 46790 }, { "epoch": 3.988069876438006, "grad_norm": 2.65625, "learning_rate": 4.783257805394007e-05, "loss": 1.279782485961914, "step": 46800 }, { "epoch": 3.9889220281210056, "grad_norm": 2.5625, "learning_rate": 4.783145385399309e-05, "loss": 1.2687750816345216, "step": 46810 }, { "epoch": 3.989774179804005, "grad_norm": 2.703125, "learning_rate": 4.7830329375789204e-05, "loss": 1.3841726303100585, "step": 46820 }, { "epoch": 3.9906263314870047, "grad_norm": 2.046875, "learning_rate": 4.7829204619342146e-05, "loss": 1.3714547157287598, "step": 46830 }, { "epoch": 3.9914784831700043, "grad_norm": 2.5, "learning_rate": 4.7828079584665596e-05, "loss": 1.3050857543945313, "step": 46840 }, { "epoch": 3.992330634853004, "grad_norm": 2.6875, "learning_rate": 4.7826954271773295e-05, "loss": 1.463724136352539, "step": 46850 }, { "epoch": 3.9931827865360034, "grad_norm": 2.09375, "learning_rate": 4.782582868067893e-05, "loss": 1.2991205215454102, "step": 46860 }, { "epoch": 3.994034938219003, "grad_norm": 2.265625, "learning_rate": 4.782470281139624e-05, "loss": 1.1167768478393554, "step": 46870 }, { "epoch": 3.9948870899020026, "grad_norm": 2.453125, "learning_rate": 4.782357666393893e-05, "loss": 1.3260650634765625, "step": 46880 }, { "epoch": 3.995739241585002, "grad_norm": 2.75, "learning_rate": 4.782245023832074e-05, "loss": 1.430446243286133, "step": 46890 }, { "epoch": 3.9965913932680017, "grad_norm": 2.078125, "learning_rate": 4.7821323534555377e-05, "loss": 1.4370346069335938, "step": 46900 }, { "epoch": 3.9974435449510013, "grad_norm": 2.3125, "learning_rate": 4.7820196552656596e-05, "loss": 1.2822793006896973, "step": 46910 }, { "epoch": 3.998295696634001, "grad_norm": 2.71875, "learning_rate": 4.781906929263812e-05, "loss": 1.5082158088684081, "step": 46920 }, { "epoch": 3.9991478483170004, "grad_norm": 3.234375, "learning_rate": 4.781794175451369e-05, "loss": 1.453512954711914, "step": 46930 }, { "epoch": 3.9996591393268, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6502, "eval_loss": 1.1961392164230347, "eval_runtime": 53.6998, "eval_samples_per_second": 368.009, "eval_steps_per_second": 23.017, "step": 46936 }, { "epoch": 4.0, "grad_norm": 2.046875, "learning_rate": 4.781681393829704e-05, "loss": 1.300511074066162, "step": 46940 }, { "epoch": 4.000852151683, "grad_norm": 2.265625, "learning_rate": 4.781568584400194e-05, "loss": 1.2299533843994142, "step": 46950 }, { "epoch": 4.001704303365999, "grad_norm": 2.375, "learning_rate": 4.7814557471642104e-05, "loss": 1.3964510917663575, "step": 46960 }, { "epoch": 4.002556455048999, "grad_norm": 2.421875, "learning_rate": 4.781342882123131e-05, "loss": 1.2541857719421388, "step": 46970 }, { "epoch": 4.003408606731998, "grad_norm": 1.7890625, "learning_rate": 4.7812299892783304e-05, "loss": 1.3152761459350586, "step": 46980 }, { "epoch": 4.004260758414998, "grad_norm": 2.9375, "learning_rate": 4.7811170686311846e-05, "loss": 1.2577550888061524, "step": 46990 }, { "epoch": 4.005112910097997, "grad_norm": 2.3125, "learning_rate": 4.781004120183069e-05, "loss": 1.1592396736145019, "step": 47000 }, { "epoch": 4.005965061780997, "grad_norm": 2.65625, "learning_rate": 4.7808911439353624e-05, "loss": 1.169949436187744, "step": 47010 }, { "epoch": 4.0068172134639966, "grad_norm": 2.515625, "learning_rate": 4.7807781398894386e-05, "loss": 1.4661096572875976, "step": 47020 }, { "epoch": 4.007669365146996, "grad_norm": 1.828125, "learning_rate": 4.780665108046677e-05, "loss": 1.3368958473205566, "step": 47030 }, { "epoch": 4.008521516829996, "grad_norm": 2.375, "learning_rate": 4.780552048408454e-05, "loss": 1.2331791877746583, "step": 47040 }, { "epoch": 4.009373668512995, "grad_norm": 2.953125, "learning_rate": 4.780438960976149e-05, "loss": 1.3676544189453126, "step": 47050 }, { "epoch": 4.010225820195995, "grad_norm": 3.109375, "learning_rate": 4.780325845751139e-05, "loss": 1.2397064208984374, "step": 47060 }, { "epoch": 4.011077971878994, "grad_norm": 2.71875, "learning_rate": 4.780212702734803e-05, "loss": 1.2782855033874512, "step": 47070 }, { "epoch": 4.011930123561994, "grad_norm": 2.546875, "learning_rate": 4.780099531928519e-05, "loss": 1.3840721130371094, "step": 47080 }, { "epoch": 4.0127822752449935, "grad_norm": 2.53125, "learning_rate": 4.779986333333668e-05, "loss": 1.3712411880493165, "step": 47090 }, { "epoch": 4.013634426927993, "grad_norm": 2.3125, "learning_rate": 4.779873106951629e-05, "loss": 1.5240058898925781, "step": 47100 }, { "epoch": 4.014486578610993, "grad_norm": 1.8984375, "learning_rate": 4.77975985278378e-05, "loss": 1.4350605010986328, "step": 47110 }, { "epoch": 4.015338730293992, "grad_norm": 2.5, "learning_rate": 4.779646570831505e-05, "loss": 1.390422248840332, "step": 47120 }, { "epoch": 4.016190881976992, "grad_norm": 2.015625, "learning_rate": 4.7795332610961816e-05, "loss": 1.3409018516540527, "step": 47130 }, { "epoch": 4.017043033659991, "grad_norm": 2.453125, "learning_rate": 4.779419923579192e-05, "loss": 1.5156804084777833, "step": 47140 }, { "epoch": 4.017895185342991, "grad_norm": 1.859375, "learning_rate": 4.7793065582819165e-05, "loss": 1.4086355209350585, "step": 47150 }, { "epoch": 4.0187473370259905, "grad_norm": 2.046875, "learning_rate": 4.779193165205738e-05, "loss": 1.5325950622558593, "step": 47160 }, { "epoch": 4.01959948870899, "grad_norm": 4.0, "learning_rate": 4.779079744352037e-05, "loss": 1.3143775939941407, "step": 47170 }, { "epoch": 4.02045164039199, "grad_norm": 2.53125, "learning_rate": 4.7789662957221976e-05, "loss": 1.29060640335083, "step": 47180 }, { "epoch": 4.021303792074989, "grad_norm": 2.21875, "learning_rate": 4.778852819317601e-05, "loss": 1.3202129364013673, "step": 47190 }, { "epoch": 4.022155943757989, "grad_norm": 2.640625, "learning_rate": 4.778739315139631e-05, "loss": 1.2947892189025878, "step": 47200 }, { "epoch": 4.023008095440988, "grad_norm": 2.8125, "learning_rate": 4.778625783189671e-05, "loss": 1.4063518524169922, "step": 47210 }, { "epoch": 4.023860247123988, "grad_norm": 2.515625, "learning_rate": 4.7785122234691035e-05, "loss": 1.2389161109924316, "step": 47220 }, { "epoch": 4.0247123988069875, "grad_norm": 1.7109375, "learning_rate": 4.778398635979314e-05, "loss": 1.4469123840332032, "step": 47230 }, { "epoch": 4.025564550489987, "grad_norm": 2.015625, "learning_rate": 4.7782850207216856e-05, "loss": 1.36658935546875, "step": 47240 }, { "epoch": 4.026416702172987, "grad_norm": 2.0, "learning_rate": 4.778171377697604e-05, "loss": 1.248394775390625, "step": 47250 }, { "epoch": 4.027268853855986, "grad_norm": 2.6875, "learning_rate": 4.7780577069084534e-05, "loss": 1.3904969215393066, "step": 47260 }, { "epoch": 4.028121005538986, "grad_norm": 2.0625, "learning_rate": 4.777944008355619e-05, "loss": 1.327157974243164, "step": 47270 }, { "epoch": 4.028973157221985, "grad_norm": 2.625, "learning_rate": 4.777830282040488e-05, "loss": 1.2458230018615724, "step": 47280 }, { "epoch": 4.029825308904985, "grad_norm": 1.984375, "learning_rate": 4.7777165279644456e-05, "loss": 1.3221961975097656, "step": 47290 }, { "epoch": 4.0306774605879845, "grad_norm": 2.515625, "learning_rate": 4.7776027461288775e-05, "loss": 1.4812264442443848, "step": 47300 }, { "epoch": 4.031529612270984, "grad_norm": 1.671875, "learning_rate": 4.77748893653517e-05, "loss": 1.2940059661865235, "step": 47310 }, { "epoch": 4.032381763953984, "grad_norm": 3.734375, "learning_rate": 4.7773750991847124e-05, "loss": 1.4195779800415038, "step": 47320 }, { "epoch": 4.033233915636983, "grad_norm": 1.453125, "learning_rate": 4.777261234078891e-05, "loss": 1.3340713500976562, "step": 47330 }, { "epoch": 4.034086067319983, "grad_norm": 4.0625, "learning_rate": 4.7771473412190926e-05, "loss": 1.3656158447265625, "step": 47340 }, { "epoch": 4.034938219002982, "grad_norm": 2.296875, "learning_rate": 4.7770334206067055e-05, "loss": 1.403341007232666, "step": 47350 }, { "epoch": 4.035790370685982, "grad_norm": 2.359375, "learning_rate": 4.776919472243119e-05, "loss": 1.35952730178833, "step": 47360 }, { "epoch": 4.0366425223689815, "grad_norm": 1.8984375, "learning_rate": 4.776805496129721e-05, "loss": 1.2983504295349122, "step": 47370 }, { "epoch": 4.037494674051981, "grad_norm": 2.0, "learning_rate": 4.776691492267902e-05, "loss": 1.3899313926696777, "step": 47380 }, { "epoch": 4.038346825734981, "grad_norm": 1.7578125, "learning_rate": 4.7765774606590493e-05, "loss": 1.3830097198486329, "step": 47390 }, { "epoch": 4.03919897741798, "grad_norm": 1.8203125, "learning_rate": 4.776463401304554e-05, "loss": 1.3430834770202638, "step": 47400 }, { "epoch": 4.04005112910098, "grad_norm": 2.640625, "learning_rate": 4.7763493142058056e-05, "loss": 1.3559734344482421, "step": 47410 }, { "epoch": 4.040903280783979, "grad_norm": 3.359375, "learning_rate": 4.776235199364196e-05, "loss": 1.2423684120178222, "step": 47420 }, { "epoch": 4.041755432466979, "grad_norm": 2.046875, "learning_rate": 4.776121056781114e-05, "loss": 1.1576739311218263, "step": 47430 }, { "epoch": 4.0426075841499785, "grad_norm": 1.90625, "learning_rate": 4.776006886457951e-05, "loss": 1.4650107383728028, "step": 47440 }, { "epoch": 4.043459735832978, "grad_norm": 2.59375, "learning_rate": 4.7758926883960994e-05, "loss": 1.3169390678405761, "step": 47450 }, { "epoch": 4.044311887515978, "grad_norm": 4.96875, "learning_rate": 4.7757784625969506e-05, "loss": 1.4755891799926757, "step": 47460 }, { "epoch": 4.045164039198977, "grad_norm": 3.578125, "learning_rate": 4.775664209061897e-05, "loss": 1.3426870346069335, "step": 47470 }, { "epoch": 4.046016190881977, "grad_norm": 2.765625, "learning_rate": 4.77554992779233e-05, "loss": 1.3452339172363281, "step": 47480 }, { "epoch": 4.046868342564976, "grad_norm": 2.984375, "learning_rate": 4.7754356187896435e-05, "loss": 1.2403080940246582, "step": 47490 }, { "epoch": 4.047720494247976, "grad_norm": 2.28125, "learning_rate": 4.775321282055231e-05, "loss": 1.339348030090332, "step": 47500 }, { "epoch": 4.048572645930975, "grad_norm": 2.78125, "learning_rate": 4.775206917590483e-05, "loss": 1.263771152496338, "step": 47510 }, { "epoch": 4.049424797613975, "grad_norm": 1.65625, "learning_rate": 4.7750925253967975e-05, "loss": 1.3639244079589843, "step": 47520 }, { "epoch": 4.050276949296975, "grad_norm": 1.5625, "learning_rate": 4.774978105475566e-05, "loss": 1.320387840270996, "step": 47530 }, { "epoch": 4.051129100979974, "grad_norm": 2.515625, "learning_rate": 4.774863657828184e-05, "loss": 1.3539169311523438, "step": 47540 }, { "epoch": 4.051981252662974, "grad_norm": 1.8359375, "learning_rate": 4.774749182456046e-05, "loss": 1.3922344207763673, "step": 47550 }, { "epoch": 4.052833404345973, "grad_norm": 2.421875, "learning_rate": 4.774634679360547e-05, "loss": 1.225164031982422, "step": 47560 }, { "epoch": 4.053685556028973, "grad_norm": 2.484375, "learning_rate": 4.774520148543082e-05, "loss": 1.4238046646118163, "step": 47570 }, { "epoch": 4.054537707711972, "grad_norm": 2.75, "learning_rate": 4.774405590005048e-05, "loss": 1.314287281036377, "step": 47580 }, { "epoch": 4.055389859394972, "grad_norm": 2.359375, "learning_rate": 4.77429100374784e-05, "loss": 1.318851375579834, "step": 47590 }, { "epoch": 4.0562420110779716, "grad_norm": 1.7578125, "learning_rate": 4.774176389772856e-05, "loss": 1.256046676635742, "step": 47600 }, { "epoch": 4.057094162760971, "grad_norm": 3.046875, "learning_rate": 4.774061748081492e-05, "loss": 1.324346160888672, "step": 47610 }, { "epoch": 4.057946314443971, "grad_norm": 2.671875, "learning_rate": 4.773947078675145e-05, "loss": 1.3070760726928712, "step": 47620 }, { "epoch": 4.05879846612697, "grad_norm": 2.859375, "learning_rate": 4.7738323815552134e-05, "loss": 1.4984576225280761, "step": 47630 }, { "epoch": 4.05965061780997, "grad_norm": 3.625, "learning_rate": 4.7737176567230936e-05, "loss": 1.3831585884094237, "step": 47640 }, { "epoch": 4.060502769492969, "grad_norm": 1.9765625, "learning_rate": 4.7736029041801846e-05, "loss": 1.4576512336730958, "step": 47650 }, { "epoch": 4.061354921175969, "grad_norm": 1.546875, "learning_rate": 4.773488123927885e-05, "loss": 1.4514892578125, "step": 47660 }, { "epoch": 4.0622070728589685, "grad_norm": 2.078125, "learning_rate": 4.773373315967594e-05, "loss": 1.4810030937194825, "step": 47670 }, { "epoch": 4.063059224541968, "grad_norm": 1.8515625, "learning_rate": 4.7732584803007096e-05, "loss": 1.4243348121643067, "step": 47680 }, { "epoch": 4.063911376224968, "grad_norm": 2.390625, "learning_rate": 4.7731436169286325e-05, "loss": 1.404077911376953, "step": 47690 }, { "epoch": 4.064763527907967, "grad_norm": 2.703125, "learning_rate": 4.7730287258527626e-05, "loss": 1.2444392204284669, "step": 47700 }, { "epoch": 4.065615679590967, "grad_norm": 2.625, "learning_rate": 4.772913807074499e-05, "loss": 1.421997356414795, "step": 47710 }, { "epoch": 4.066467831273966, "grad_norm": 3.046875, "learning_rate": 4.7727988605952444e-05, "loss": 1.3622142791748046, "step": 47720 }, { "epoch": 4.067319982956966, "grad_norm": 3.140625, "learning_rate": 4.7726838864163974e-05, "loss": 1.3980284690856934, "step": 47730 }, { "epoch": 4.0681721346399655, "grad_norm": 2.625, "learning_rate": 4.77256888453936e-05, "loss": 1.2922880172729492, "step": 47740 }, { "epoch": 4.069024286322965, "grad_norm": 2.1875, "learning_rate": 4.772453854965535e-05, "loss": 1.3898727416992187, "step": 47750 }, { "epoch": 4.069876438005965, "grad_norm": 2.03125, "learning_rate": 4.772338797696322e-05, "loss": 1.212886905670166, "step": 47760 }, { "epoch": 4.070728589688964, "grad_norm": 2.15625, "learning_rate": 4.772223712733126e-05, "loss": 1.3963627815246582, "step": 47770 }, { "epoch": 4.071580741371964, "grad_norm": 2.328125, "learning_rate": 4.772108600077347e-05, "loss": 1.2265193939208985, "step": 47780 }, { "epoch": 4.072432893054963, "grad_norm": 3.375, "learning_rate": 4.7719934597303894e-05, "loss": 1.3565152168273926, "step": 47790 }, { "epoch": 4.073285044737963, "grad_norm": 4.03125, "learning_rate": 4.771878291693657e-05, "loss": 1.3068888664245606, "step": 47800 }, { "epoch": 4.0741371964209625, "grad_norm": 1.9609375, "learning_rate": 4.771763095968551e-05, "loss": 1.4305139541625977, "step": 47810 }, { "epoch": 4.074989348103962, "grad_norm": 2.859375, "learning_rate": 4.771647872556478e-05, "loss": 1.3235517501831056, "step": 47820 }, { "epoch": 4.0758414997869625, "grad_norm": 2.328125, "learning_rate": 4.771532621458841e-05, "loss": 1.3988574028015137, "step": 47830 }, { "epoch": 4.076693651469961, "grad_norm": 2.515625, "learning_rate": 4.771417342677044e-05, "loss": 1.3039774894714355, "step": 47840 }, { "epoch": 4.077545803152962, "grad_norm": 2.6875, "learning_rate": 4.771302036212493e-05, "loss": 1.2646217346191406, "step": 47850 }, { "epoch": 4.078397954835961, "grad_norm": 3.671875, "learning_rate": 4.771186702066593e-05, "loss": 1.4632402420043946, "step": 47860 }, { "epoch": 4.079250106518961, "grad_norm": 1.96875, "learning_rate": 4.7710713402407504e-05, "loss": 1.3673407554626464, "step": 47870 }, { "epoch": 4.08010225820196, "grad_norm": 2.109375, "learning_rate": 4.77095595073637e-05, "loss": 1.317457675933838, "step": 47880 }, { "epoch": 4.08095440988496, "grad_norm": 2.546875, "learning_rate": 4.7708405335548584e-05, "loss": 1.3980391502380372, "step": 47890 }, { "epoch": 4.0818065615679595, "grad_norm": 2.34375, "learning_rate": 4.770725088697622e-05, "loss": 1.1955621719360352, "step": 47900 }, { "epoch": 4.082658713250959, "grad_norm": 2.375, "learning_rate": 4.770609616166069e-05, "loss": 1.3510177612304688, "step": 47910 }, { "epoch": 4.083510864933959, "grad_norm": 2.515625, "learning_rate": 4.770494115961605e-05, "loss": 1.2642240524291992, "step": 47920 }, { "epoch": 4.084363016616958, "grad_norm": 2.9375, "learning_rate": 4.770378588085639e-05, "loss": 1.2808422088623046, "step": 47930 }, { "epoch": 4.085215168299958, "grad_norm": 3.203125, "learning_rate": 4.7702630325395786e-05, "loss": 1.2785542488098145, "step": 47940 }, { "epoch": 4.086067319982957, "grad_norm": 3.109375, "learning_rate": 4.770147449324832e-05, "loss": 1.3464609146118165, "step": 47950 }, { "epoch": 4.086919471665957, "grad_norm": 2.578125, "learning_rate": 4.7700318384428075e-05, "loss": 1.3244489669799804, "step": 47960 }, { "epoch": 4.0877716233489565, "grad_norm": 5.125, "learning_rate": 4.769916199894914e-05, "loss": 1.3735685348510742, "step": 47970 }, { "epoch": 4.088623775031956, "grad_norm": 1.8125, "learning_rate": 4.769800533682562e-05, "loss": 1.4667017936706543, "step": 47980 }, { "epoch": 4.089475926714956, "grad_norm": 2.515625, "learning_rate": 4.76968483980716e-05, "loss": 1.4278968811035155, "step": 47990 }, { "epoch": 4.090328078397955, "grad_norm": 2.3125, "learning_rate": 4.76956911827012e-05, "loss": 1.334301471710205, "step": 48000 }, { "epoch": 4.091180230080955, "grad_norm": 2.609375, "learning_rate": 4.7694533690728486e-05, "loss": 1.2901787757873535, "step": 48010 }, { "epoch": 4.092032381763954, "grad_norm": 3.4375, "learning_rate": 4.769337592216761e-05, "loss": 1.3338815689086914, "step": 48020 }, { "epoch": 4.092884533446954, "grad_norm": 2.546875, "learning_rate": 4.769221787703264e-05, "loss": 1.3137399673461914, "step": 48030 }, { "epoch": 4.0937366851299535, "grad_norm": 2.421875, "learning_rate": 4.7691059555337713e-05, "loss": 1.4089853286743164, "step": 48040 }, { "epoch": 4.094588836812953, "grad_norm": 2.46875, "learning_rate": 4.768990095709694e-05, "loss": 1.2792043685913086, "step": 48050 }, { "epoch": 4.095440988495953, "grad_norm": 3.625, "learning_rate": 4.768874208232446e-05, "loss": 1.2361230850219727, "step": 48060 }, { "epoch": 4.096293140178952, "grad_norm": 1.9140625, "learning_rate": 4.768758293103437e-05, "loss": 1.365610408782959, "step": 48070 }, { "epoch": 4.097145291861952, "grad_norm": 1.734375, "learning_rate": 4.76864235032408e-05, "loss": 1.246756649017334, "step": 48080 }, { "epoch": 4.097997443544951, "grad_norm": 2.40625, "learning_rate": 4.76852637989579e-05, "loss": 1.2659334182739257, "step": 48090 }, { "epoch": 4.098849595227951, "grad_norm": 5.5, "learning_rate": 4.7684103818199784e-05, "loss": 1.2163617134094238, "step": 48100 }, { "epoch": 4.0997017469109505, "grad_norm": 2.703125, "learning_rate": 4.76829435609806e-05, "loss": 1.2461376190185547, "step": 48110 }, { "epoch": 4.10055389859395, "grad_norm": 1.8515625, "learning_rate": 4.7681783027314476e-05, "loss": 1.2346834182739257, "step": 48120 }, { "epoch": 4.10140605027695, "grad_norm": 2.859375, "learning_rate": 4.7680622217215574e-05, "loss": 1.3314010620117187, "step": 48130 }, { "epoch": 4.102258201959949, "grad_norm": 2.609375, "learning_rate": 4.7679461130698034e-05, "loss": 1.3150211334228517, "step": 48140 }, { "epoch": 4.103110353642949, "grad_norm": 3.15625, "learning_rate": 4.7678299767776004e-05, "loss": 1.3772512435913087, "step": 48150 }, { "epoch": 4.103962505325948, "grad_norm": 2.265625, "learning_rate": 4.7677138128463634e-05, "loss": 1.360938835144043, "step": 48160 }, { "epoch": 4.104814657008948, "grad_norm": 3.265625, "learning_rate": 4.7675976212775086e-05, "loss": 1.3674981117248535, "step": 48170 }, { "epoch": 4.1056668086919474, "grad_norm": 1.8671875, "learning_rate": 4.767481402072453e-05, "loss": 1.3113393783569336, "step": 48180 }, { "epoch": 4.106518960374947, "grad_norm": 2.8125, "learning_rate": 4.767365155232611e-05, "loss": 1.4360491752624511, "step": 48190 }, { "epoch": 4.107371112057947, "grad_norm": 2.265625, "learning_rate": 4.767248880759402e-05, "loss": 1.26910343170166, "step": 48200 }, { "epoch": 4.108223263740946, "grad_norm": 1.4453125, "learning_rate": 4.76713257865424e-05, "loss": 1.2935510635375977, "step": 48210 }, { "epoch": 4.109075415423946, "grad_norm": 2.21875, "learning_rate": 4.767016248918545e-05, "loss": 1.4740812301635742, "step": 48220 }, { "epoch": 4.109927567106945, "grad_norm": 2.9375, "learning_rate": 4.766899891553734e-05, "loss": 1.2633477210998536, "step": 48230 }, { "epoch": 4.110779718789945, "grad_norm": 1.7890625, "learning_rate": 4.766783506561224e-05, "loss": 1.4012087821960448, "step": 48240 }, { "epoch": 4.111631870472944, "grad_norm": 2.203125, "learning_rate": 4.7666670939424354e-05, "loss": 1.425687313079834, "step": 48250 }, { "epoch": 4.112484022155944, "grad_norm": 1.875, "learning_rate": 4.766550653698785e-05, "loss": 1.3044803619384766, "step": 48260 }, { "epoch": 4.113336173838944, "grad_norm": 3.078125, "learning_rate": 4.766434185831693e-05, "loss": 1.305405044555664, "step": 48270 }, { "epoch": 4.114188325521943, "grad_norm": 2.046875, "learning_rate": 4.766317690342578e-05, "loss": 1.4165460586547851, "step": 48280 }, { "epoch": 4.115040477204943, "grad_norm": 4.03125, "learning_rate": 4.766201167232861e-05, "loss": 1.2899638175964356, "step": 48290 }, { "epoch": 4.115892628887942, "grad_norm": 1.859375, "learning_rate": 4.766084616503962e-05, "loss": 1.4493351936340333, "step": 48300 }, { "epoch": 4.116744780570942, "grad_norm": 2.328125, "learning_rate": 4.7659680381573e-05, "loss": 1.1952903747558594, "step": 48310 }, { "epoch": 4.117596932253941, "grad_norm": 2.09375, "learning_rate": 4.765851432194298e-05, "loss": 1.3408595085144044, "step": 48320 }, { "epoch": 4.118449083936941, "grad_norm": 2.515625, "learning_rate": 4.765734798616375e-05, "loss": 1.3533889770507812, "step": 48330 }, { "epoch": 4.1193012356199405, "grad_norm": 3.625, "learning_rate": 4.7656181374249535e-05, "loss": 1.4102343559265136, "step": 48340 }, { "epoch": 4.12015338730294, "grad_norm": 1.84375, "learning_rate": 4.765501448621455e-05, "loss": 1.3892862319946289, "step": 48350 }, { "epoch": 4.12100553898594, "grad_norm": 2.625, "learning_rate": 4.7653847322073026e-05, "loss": 1.382045364379883, "step": 48360 }, { "epoch": 4.121857690668939, "grad_norm": 2.3125, "learning_rate": 4.765267988183917e-05, "loss": 1.4172561645507813, "step": 48370 }, { "epoch": 4.122709842351939, "grad_norm": 1.5625, "learning_rate": 4.765151216552722e-05, "loss": 1.3300107002258301, "step": 48380 }, { "epoch": 4.123561994034938, "grad_norm": 2.84375, "learning_rate": 4.765034417315142e-05, "loss": 1.390708065032959, "step": 48390 }, { "epoch": 4.124414145717938, "grad_norm": 2.171875, "learning_rate": 4.7649175904725984e-05, "loss": 1.1750479698181153, "step": 48400 }, { "epoch": 4.1252662974009375, "grad_norm": 1.8984375, "learning_rate": 4.764800736026516e-05, "loss": 1.4800661087036133, "step": 48410 }, { "epoch": 4.126118449083937, "grad_norm": 2.375, "learning_rate": 4.764683853978319e-05, "loss": 1.285464096069336, "step": 48420 }, { "epoch": 4.126970600766937, "grad_norm": 2.84375, "learning_rate": 4.7645669443294305e-05, "loss": 1.3720189094543458, "step": 48430 }, { "epoch": 4.127822752449936, "grad_norm": 1.7265625, "learning_rate": 4.7644500070812775e-05, "loss": 1.3477638244628907, "step": 48440 }, { "epoch": 4.128674904132936, "grad_norm": 2.4375, "learning_rate": 4.764333042235284e-05, "loss": 1.4390149116516113, "step": 48450 }, { "epoch": 4.129527055815935, "grad_norm": 3.1875, "learning_rate": 4.764216049792876e-05, "loss": 1.3998882293701171, "step": 48460 }, { "epoch": 4.130379207498935, "grad_norm": 1.5390625, "learning_rate": 4.7640990297554786e-05, "loss": 1.3016554832458496, "step": 48470 }, { "epoch": 4.1312313591819345, "grad_norm": 2.515625, "learning_rate": 4.763981982124518e-05, "loss": 1.4277299880981444, "step": 48480 }, { "epoch": 4.132083510864934, "grad_norm": 2.0625, "learning_rate": 4.763864906901422e-05, "loss": 1.3612414360046388, "step": 48490 }, { "epoch": 4.132935662547934, "grad_norm": 3.25, "learning_rate": 4.763747804087616e-05, "loss": 1.2566667556762696, "step": 48500 }, { "epoch": 4.133787814230933, "grad_norm": 2.515625, "learning_rate": 4.7636306736845284e-05, "loss": 1.3513954162597657, "step": 48510 }, { "epoch": 4.134639965913933, "grad_norm": 2.203125, "learning_rate": 4.763513515693585e-05, "loss": 1.3708138465881348, "step": 48520 }, { "epoch": 4.135492117596932, "grad_norm": 2.203125, "learning_rate": 4.7633963301162155e-05, "loss": 1.366283130645752, "step": 48530 }, { "epoch": 4.136344269279932, "grad_norm": 2.671875, "learning_rate": 4.7632791169538474e-05, "loss": 1.321558380126953, "step": 48540 }, { "epoch": 4.1371964209629315, "grad_norm": 2.265625, "learning_rate": 4.763161876207909e-05, "loss": 1.4204216003417969, "step": 48550 }, { "epoch": 4.138048572645931, "grad_norm": 1.8125, "learning_rate": 4.7630446078798286e-05, "loss": 1.3285573959350585, "step": 48560 }, { "epoch": 4.138900724328931, "grad_norm": 2.6875, "learning_rate": 4.762927311971037e-05, "loss": 1.463331699371338, "step": 48570 }, { "epoch": 4.13975287601193, "grad_norm": 2.03125, "learning_rate": 4.762809988482962e-05, "loss": 1.2648109436035155, "step": 48580 }, { "epoch": 4.14060502769493, "grad_norm": 1.96875, "learning_rate": 4.762692637417035e-05, "loss": 1.3065240859985352, "step": 48590 }, { "epoch": 4.141457179377929, "grad_norm": 3.921875, "learning_rate": 4.7625752587746854e-05, "loss": 1.355554962158203, "step": 48600 }, { "epoch": 4.142309331060929, "grad_norm": 3.046875, "learning_rate": 4.762457852557343e-05, "loss": 1.484284019470215, "step": 48610 }, { "epoch": 4.1431614827439285, "grad_norm": 1.6953125, "learning_rate": 4.7623404187664405e-05, "loss": 1.3965571403503418, "step": 48620 }, { "epoch": 4.144013634426928, "grad_norm": 2.125, "learning_rate": 4.7622229574034085e-05, "loss": 1.3916983604431152, "step": 48630 }, { "epoch": 4.144865786109928, "grad_norm": 2.71875, "learning_rate": 4.762105468469677e-05, "loss": 1.3996733665466308, "step": 48640 }, { "epoch": 4.145717937792927, "grad_norm": 3.71875, "learning_rate": 4.76198795196668e-05, "loss": 1.393204116821289, "step": 48650 }, { "epoch": 4.146570089475927, "grad_norm": 4.09375, "learning_rate": 4.761870407895849e-05, "loss": 1.1848155975341796, "step": 48660 }, { "epoch": 4.147422241158926, "grad_norm": 2.21875, "learning_rate": 4.761752836258615e-05, "loss": 1.3191608428955077, "step": 48670 }, { "epoch": 4.148274392841926, "grad_norm": 1.984375, "learning_rate": 4.761635237056413e-05, "loss": 1.312160110473633, "step": 48680 }, { "epoch": 4.1491265445249255, "grad_norm": 3.0625, "learning_rate": 4.761517610290676e-05, "loss": 1.2967209815979004, "step": 48690 }, { "epoch": 4.149978696207925, "grad_norm": 2.8125, "learning_rate": 4.761399955962837e-05, "loss": 1.383193588256836, "step": 48700 }, { "epoch": 4.150830847890925, "grad_norm": 1.9765625, "learning_rate": 4.76128227407433e-05, "loss": 1.3302936553955078, "step": 48710 }, { "epoch": 4.151682999573924, "grad_norm": 2.515625, "learning_rate": 4.761164564626589e-05, "loss": 1.298797035217285, "step": 48720 }, { "epoch": 4.152535151256924, "grad_norm": 2.171875, "learning_rate": 4.7610468276210485e-05, "loss": 1.516858959197998, "step": 48730 }, { "epoch": 4.153387302939923, "grad_norm": 2.03125, "learning_rate": 4.760929063059144e-05, "loss": 1.2375008583068847, "step": 48740 }, { "epoch": 4.154239454622923, "grad_norm": 2.109375, "learning_rate": 4.76081127094231e-05, "loss": 1.2631479263305665, "step": 48750 }, { "epoch": 4.1550916063059224, "grad_norm": 2.796875, "learning_rate": 4.760693451271984e-05, "loss": 1.3159286499023437, "step": 48760 }, { "epoch": 4.155943757988922, "grad_norm": 2.4375, "learning_rate": 4.760575604049599e-05, "loss": 1.3829811096191407, "step": 48770 }, { "epoch": 4.156795909671922, "grad_norm": 2.53125, "learning_rate": 4.760457729276593e-05, "loss": 1.4201441764831544, "step": 48780 }, { "epoch": 4.157648061354921, "grad_norm": 1.671875, "learning_rate": 4.760339826954403e-05, "loss": 1.4330236434936523, "step": 48790 }, { "epoch": 4.158500213037921, "grad_norm": 4.03125, "learning_rate": 4.7602218970844647e-05, "loss": 1.227365493774414, "step": 48800 }, { "epoch": 4.15935236472092, "grad_norm": 2.046875, "learning_rate": 4.760103939668216e-05, "loss": 1.425423526763916, "step": 48810 }, { "epoch": 4.16020451640392, "grad_norm": 2.59375, "learning_rate": 4.7599859547070946e-05, "loss": 1.390561866760254, "step": 48820 }, { "epoch": 4.161056668086919, "grad_norm": 1.953125, "learning_rate": 4.7598679422025375e-05, "loss": 1.3977280616760255, "step": 48830 }, { "epoch": 4.161908819769919, "grad_norm": 1.703125, "learning_rate": 4.7597499021559845e-05, "loss": 1.4602215766906739, "step": 48840 }, { "epoch": 4.162760971452919, "grad_norm": 2.671875, "learning_rate": 4.759631834568873e-05, "loss": 1.335828971862793, "step": 48850 }, { "epoch": 4.163613123135918, "grad_norm": 2.375, "learning_rate": 4.7595137394426424e-05, "loss": 1.3183299064636231, "step": 48860 }, { "epoch": 4.164465274818918, "grad_norm": 2.28125, "learning_rate": 4.7593956167787325e-05, "loss": 1.4009150505065917, "step": 48870 }, { "epoch": 4.165317426501917, "grad_norm": 2.578125, "learning_rate": 4.759277466578582e-05, "loss": 1.3697755813598633, "step": 48880 }, { "epoch": 4.166169578184917, "grad_norm": 2.1875, "learning_rate": 4.759159288843631e-05, "loss": 1.3098546981811523, "step": 48890 }, { "epoch": 4.167021729867916, "grad_norm": 2.21875, "learning_rate": 4.75904108357532e-05, "loss": 1.4029144287109374, "step": 48900 }, { "epoch": 4.167873881550916, "grad_norm": 2.9375, "learning_rate": 4.7589228507750897e-05, "loss": 1.2851654052734376, "step": 48910 }, { "epoch": 4.1687260332339156, "grad_norm": 2.84375, "learning_rate": 4.7588045904443806e-05, "loss": 1.4302173614501954, "step": 48920 }, { "epoch": 4.169578184916915, "grad_norm": 2.171875, "learning_rate": 4.7586863025846345e-05, "loss": 1.361443328857422, "step": 48930 }, { "epoch": 4.170430336599915, "grad_norm": 2.5625, "learning_rate": 4.7585679871972935e-05, "loss": 1.2974373817443847, "step": 48940 }, { "epoch": 4.171282488282914, "grad_norm": 2.421875, "learning_rate": 4.758449644283798e-05, "loss": 1.3699060440063477, "step": 48950 }, { "epoch": 4.172134639965914, "grad_norm": 2.453125, "learning_rate": 4.7583312738455916e-05, "loss": 1.2762206077575684, "step": 48960 }, { "epoch": 4.172986791648913, "grad_norm": 1.9921875, "learning_rate": 4.758212875884117e-05, "loss": 1.2390380859375, "step": 48970 }, { "epoch": 4.173838943331913, "grad_norm": 1.90625, "learning_rate": 4.758094450400816e-05, "loss": 1.2430787086486816, "step": 48980 }, { "epoch": 4.1746910950149125, "grad_norm": 2.40625, "learning_rate": 4.757975997397133e-05, "loss": 1.3080623626708985, "step": 48990 }, { "epoch": 4.175543246697912, "grad_norm": 4.90625, "learning_rate": 4.757857516874511e-05, "loss": 1.229710006713867, "step": 49000 }, { "epoch": 4.176395398380912, "grad_norm": 2.984375, "learning_rate": 4.757739008834394e-05, "loss": 1.4745704650878906, "step": 49010 }, { "epoch": 4.177247550063911, "grad_norm": 3.265625, "learning_rate": 4.757620473278227e-05, "loss": 1.3261268615722657, "step": 49020 }, { "epoch": 4.178099701746911, "grad_norm": 1.9140625, "learning_rate": 4.757501910207455e-05, "loss": 1.3275458335876464, "step": 49030 }, { "epoch": 4.17895185342991, "grad_norm": 2.5, "learning_rate": 4.7573833196235205e-05, "loss": 1.4990809440612793, "step": 49040 }, { "epoch": 4.17980400511291, "grad_norm": 1.9296875, "learning_rate": 4.7572647015278714e-05, "loss": 1.3691370964050293, "step": 49050 }, { "epoch": 4.1806561567959095, "grad_norm": 2.484375, "learning_rate": 4.757146055921953e-05, "loss": 1.3601126670837402, "step": 49060 }, { "epoch": 4.181508308478909, "grad_norm": 3.3125, "learning_rate": 4.75702738280721e-05, "loss": 1.4400606155395508, "step": 49070 }, { "epoch": 4.182360460161909, "grad_norm": 2.203125, "learning_rate": 4.7569086821850895e-05, "loss": 1.271091842651367, "step": 49080 }, { "epoch": 4.183212611844908, "grad_norm": 2.40625, "learning_rate": 4.756789954057038e-05, "loss": 1.370075511932373, "step": 49090 }, { "epoch": 4.184064763527908, "grad_norm": 2.8125, "learning_rate": 4.7566711984245024e-05, "loss": 1.2563339233398438, "step": 49100 }, { "epoch": 4.184916915210907, "grad_norm": 2.015625, "learning_rate": 4.756552415288931e-05, "loss": 1.25396785736084, "step": 49110 }, { "epoch": 4.185769066893907, "grad_norm": 2.9375, "learning_rate": 4.75643360465177e-05, "loss": 1.3745079040527344, "step": 49120 }, { "epoch": 4.1866212185769065, "grad_norm": 3.390625, "learning_rate": 4.756314766514468e-05, "loss": 1.417917823791504, "step": 49130 }, { "epoch": 4.187473370259906, "grad_norm": 2.21875, "learning_rate": 4.756195900878474e-05, "loss": 1.262403392791748, "step": 49140 }, { "epoch": 4.188325521942906, "grad_norm": 3.03125, "learning_rate": 4.756077007745235e-05, "loss": 1.2370122909545898, "step": 49150 }, { "epoch": 4.189177673625905, "grad_norm": 2.09375, "learning_rate": 4.7559580871162026e-05, "loss": 1.3894992828369142, "step": 49160 }, { "epoch": 4.190029825308905, "grad_norm": 3.046875, "learning_rate": 4.755839138992823e-05, "loss": 1.4386263847351075, "step": 49170 }, { "epoch": 4.190881976991904, "grad_norm": 2.875, "learning_rate": 4.755720163376549e-05, "loss": 1.3980295181274414, "step": 49180 }, { "epoch": 4.191734128674904, "grad_norm": 2.84375, "learning_rate": 4.755601160268829e-05, "loss": 1.3490182876586914, "step": 49190 }, { "epoch": 4.1925862803579035, "grad_norm": 2.625, "learning_rate": 4.755482129671113e-05, "loss": 1.368596363067627, "step": 49200 }, { "epoch": 4.193438432040903, "grad_norm": 2.4375, "learning_rate": 4.7553630715848517e-05, "loss": 1.1886686325073241, "step": 49210 }, { "epoch": 4.194290583723903, "grad_norm": 3.25, "learning_rate": 4.755243986011497e-05, "loss": 1.16476469039917, "step": 49220 }, { "epoch": 4.195142735406902, "grad_norm": 2.96875, "learning_rate": 4.755124872952499e-05, "loss": 1.2332165718078614, "step": 49230 }, { "epoch": 4.195994887089902, "grad_norm": 2.15625, "learning_rate": 4.755005732409311e-05, "loss": 1.3365046501159668, "step": 49240 }, { "epoch": 4.196847038772901, "grad_norm": 3.265625, "learning_rate": 4.754886564383384e-05, "loss": 1.3053548812866211, "step": 49250 }, { "epoch": 4.197699190455901, "grad_norm": 3.265625, "learning_rate": 4.754767368876171e-05, "loss": 1.380927848815918, "step": 49260 }, { "epoch": 4.1985513421389005, "grad_norm": 2.546875, "learning_rate": 4.754648145889124e-05, "loss": 1.3106002807617188, "step": 49270 }, { "epoch": 4.1994034938219, "grad_norm": 2.625, "learning_rate": 4.7545288954236955e-05, "loss": 1.3110767364501954, "step": 49280 }, { "epoch": 4.2002556455049, "grad_norm": 2.171875, "learning_rate": 4.7544096174813404e-05, "loss": 1.2196585655212402, "step": 49290 }, { "epoch": 4.201107797187899, "grad_norm": 2.4375, "learning_rate": 4.7542903120635107e-05, "loss": 1.3031259536743165, "step": 49300 }, { "epoch": 4.201959948870899, "grad_norm": 2.125, "learning_rate": 4.754170979171662e-05, "loss": 1.4532762527465821, "step": 49310 }, { "epoch": 4.202812100553898, "grad_norm": 2.109375, "learning_rate": 4.754051618807248e-05, "loss": 1.3700650215148926, "step": 49320 }, { "epoch": 4.203664252236898, "grad_norm": 1.8828125, "learning_rate": 4.7539322309717225e-05, "loss": 1.4081518173217773, "step": 49330 }, { "epoch": 4.2045164039198974, "grad_norm": 2.5625, "learning_rate": 4.7538128156665426e-05, "loss": 1.4990570068359375, "step": 49340 }, { "epoch": 4.205368555602897, "grad_norm": 1.8984375, "learning_rate": 4.7536933728931624e-05, "loss": 1.387221336364746, "step": 49350 }, { "epoch": 4.206220707285897, "grad_norm": 2.21875, "learning_rate": 4.7535739026530366e-05, "loss": 1.379826545715332, "step": 49360 }, { "epoch": 4.207072858968896, "grad_norm": 2.15625, "learning_rate": 4.753454404947623e-05, "loss": 1.3660501480102538, "step": 49370 }, { "epoch": 4.207925010651896, "grad_norm": 2.265625, "learning_rate": 4.7533348797783774e-05, "loss": 1.362736129760742, "step": 49380 }, { "epoch": 4.208777162334895, "grad_norm": 2.28125, "learning_rate": 4.753215327146756e-05, "loss": 1.262895393371582, "step": 49390 }, { "epoch": 4.209629314017895, "grad_norm": 3.09375, "learning_rate": 4.753095747054217e-05, "loss": 1.4139190673828126, "step": 49400 }, { "epoch": 4.210481465700894, "grad_norm": 2.125, "learning_rate": 4.7529761395022163e-05, "loss": 1.2459911346435546, "step": 49410 }, { "epoch": 4.211333617383894, "grad_norm": 1.8671875, "learning_rate": 4.752856504492212e-05, "loss": 1.3410667419433593, "step": 49420 }, { "epoch": 4.212185769066894, "grad_norm": 2.25, "learning_rate": 4.752736842025664e-05, "loss": 1.1742302894592285, "step": 49430 }, { "epoch": 4.213037920749893, "grad_norm": 3.203125, "learning_rate": 4.7526171521040276e-05, "loss": 1.4090656280517577, "step": 49440 }, { "epoch": 4.213890072432893, "grad_norm": 2.359375, "learning_rate": 4.752497434728764e-05, "loss": 1.4392550468444825, "step": 49450 }, { "epoch": 4.214742224115892, "grad_norm": 2.46875, "learning_rate": 4.752377689901331e-05, "loss": 1.3680381774902344, "step": 49460 }, { "epoch": 4.215594375798892, "grad_norm": 2.0, "learning_rate": 4.752257917623187e-05, "loss": 1.3300402641296387, "step": 49470 }, { "epoch": 4.216446527481891, "grad_norm": 1.9609375, "learning_rate": 4.7521381178957947e-05, "loss": 1.4193572998046875, "step": 49480 }, { "epoch": 4.217298679164891, "grad_norm": 2.5625, "learning_rate": 4.7520182907206124e-05, "loss": 1.3321788787841797, "step": 49490 }, { "epoch": 4.2181508308478906, "grad_norm": 2.390625, "learning_rate": 4.7518984360990995e-05, "loss": 1.3571346282958985, "step": 49500 }, { "epoch": 4.21900298253089, "grad_norm": 3.15625, "learning_rate": 4.751778554032719e-05, "loss": 1.2176753997802734, "step": 49510 }, { "epoch": 4.21985513421389, "grad_norm": 1.875, "learning_rate": 4.7516586445229296e-05, "loss": 1.3061126708984374, "step": 49520 }, { "epoch": 4.220707285896889, "grad_norm": 1.484375, "learning_rate": 4.751538707571194e-05, "loss": 1.4122406005859376, "step": 49530 }, { "epoch": 4.221559437579889, "grad_norm": 2.625, "learning_rate": 4.7514187431789733e-05, "loss": 1.3155564308166503, "step": 49540 }, { "epoch": 4.222411589262888, "grad_norm": 2.1875, "learning_rate": 4.751298751347732e-05, "loss": 1.4394853591918946, "step": 49550 }, { "epoch": 4.223263740945888, "grad_norm": 3.28125, "learning_rate": 4.751178732078928e-05, "loss": 1.344322967529297, "step": 49560 }, { "epoch": 4.2241158926288875, "grad_norm": 1.796875, "learning_rate": 4.751058685374028e-05, "loss": 1.361809539794922, "step": 49570 }, { "epoch": 4.224968044311887, "grad_norm": 2.265625, "learning_rate": 4.7509386112344925e-05, "loss": 1.3801956176757812, "step": 49580 }, { "epoch": 4.225820195994887, "grad_norm": 2.59375, "learning_rate": 4.750818509661787e-05, "loss": 1.4341438293457032, "step": 49590 }, { "epoch": 4.226672347677886, "grad_norm": 1.859375, "learning_rate": 4.750698380657373e-05, "loss": 1.3036595344543458, "step": 49600 }, { "epoch": 4.227524499360886, "grad_norm": 2.109375, "learning_rate": 4.750578224222717e-05, "loss": 1.2360370635986329, "step": 49610 }, { "epoch": 4.228376651043886, "grad_norm": 2.75, "learning_rate": 4.750458040359282e-05, "loss": 1.3356435775756836, "step": 49620 }, { "epoch": 4.229228802726885, "grad_norm": 2.15625, "learning_rate": 4.7503378290685326e-05, "loss": 1.4989697456359863, "step": 49630 }, { "epoch": 4.230080954409885, "grad_norm": 3.65625, "learning_rate": 4.7502175903519344e-05, "loss": 1.3255223274230956, "step": 49640 }, { "epoch": 4.230933106092884, "grad_norm": 2.796875, "learning_rate": 4.750097324210952e-05, "loss": 1.4010430335998536, "step": 49650 }, { "epoch": 4.2317852577758845, "grad_norm": 2.640625, "learning_rate": 4.749977030647051e-05, "loss": 1.4749923706054688, "step": 49660 }, { "epoch": 4.232637409458883, "grad_norm": 1.96875, "learning_rate": 4.7498567096617e-05, "loss": 1.368511962890625, "step": 49670 }, { "epoch": 4.233489561141884, "grad_norm": 3.5, "learning_rate": 4.7497363612563625e-05, "loss": 1.4951513290405274, "step": 49680 }, { "epoch": 4.234341712824883, "grad_norm": 2.578125, "learning_rate": 4.749615985432507e-05, "loss": 1.2186600685119628, "step": 49690 }, { "epoch": 4.235193864507883, "grad_norm": 2.5625, "learning_rate": 4.749495582191599e-05, "loss": 1.401171588897705, "step": 49700 }, { "epoch": 4.236046016190882, "grad_norm": 2.46875, "learning_rate": 4.749375151535107e-05, "loss": 1.2719011306762695, "step": 49710 }, { "epoch": 4.236898167873882, "grad_norm": 2.09375, "learning_rate": 4.7492546934644986e-05, "loss": 1.468121337890625, "step": 49720 }, { "epoch": 4.2377503195568815, "grad_norm": 2.84375, "learning_rate": 4.749134207981242e-05, "loss": 1.3922149658203125, "step": 49730 }, { "epoch": 4.238602471239881, "grad_norm": 2.453125, "learning_rate": 4.749013695086806e-05, "loss": 1.4675384521484376, "step": 49740 }, { "epoch": 4.239454622922881, "grad_norm": 1.7578125, "learning_rate": 4.748893154782658e-05, "loss": 1.483053970336914, "step": 49750 }, { "epoch": 4.24030677460588, "grad_norm": 2.03125, "learning_rate": 4.748772587070267e-05, "loss": 1.3213930130004883, "step": 49760 }, { "epoch": 4.24115892628888, "grad_norm": 2.9375, "learning_rate": 4.748651991951104e-05, "loss": 1.1441027641296386, "step": 49770 }, { "epoch": 4.242011077971879, "grad_norm": 2.28125, "learning_rate": 4.748531369426639e-05, "loss": 1.2551970481872559, "step": 49780 }, { "epoch": 4.242863229654879, "grad_norm": 3.390625, "learning_rate": 4.7484107194983394e-05, "loss": 1.256361484527588, "step": 49790 }, { "epoch": 4.2437153813378785, "grad_norm": 2.953125, "learning_rate": 4.7482900421676785e-05, "loss": 1.4023334503173828, "step": 49800 }, { "epoch": 4.244567533020878, "grad_norm": 2.921875, "learning_rate": 4.748169337436125e-05, "loss": 1.303341770172119, "step": 49810 }, { "epoch": 4.245419684703878, "grad_norm": 2.984375, "learning_rate": 4.748048605305151e-05, "loss": 1.3198208808898926, "step": 49820 }, { "epoch": 4.246271836386877, "grad_norm": 2.234375, "learning_rate": 4.7479278457762284e-05, "loss": 1.2549394607543944, "step": 49830 }, { "epoch": 4.247123988069877, "grad_norm": 1.65625, "learning_rate": 4.747807058850828e-05, "loss": 1.4311066627502442, "step": 49840 }, { "epoch": 4.247976139752876, "grad_norm": 1.96875, "learning_rate": 4.7476862445304216e-05, "loss": 1.3709882736206054, "step": 49850 }, { "epoch": 4.248828291435876, "grad_norm": 2.40625, "learning_rate": 4.7475654028164826e-05, "loss": 1.4230955123901368, "step": 49860 }, { "epoch": 4.2496804431188755, "grad_norm": 2.359375, "learning_rate": 4.747444533710483e-05, "loss": 1.2871713638305664, "step": 49870 }, { "epoch": 4.250532594801875, "grad_norm": 2.0625, "learning_rate": 4.747323637213896e-05, "loss": 1.3640954971313477, "step": 49880 }, { "epoch": 4.251384746484875, "grad_norm": 2.1875, "learning_rate": 4.747202713328196e-05, "loss": 1.3363020896911622, "step": 49890 }, { "epoch": 4.252236898167874, "grad_norm": 3.125, "learning_rate": 4.747081762054856e-05, "loss": 1.247563076019287, "step": 49900 }, { "epoch": 4.253089049850874, "grad_norm": 2.359375, "learning_rate": 4.74696078339535e-05, "loss": 1.4274760246276856, "step": 49910 }, { "epoch": 4.253941201533873, "grad_norm": 2.09375, "learning_rate": 4.746839777351152e-05, "loss": 1.3620950698852539, "step": 49920 }, { "epoch": 4.254793353216873, "grad_norm": 2.359375, "learning_rate": 4.746718743923737e-05, "loss": 1.3057202339172362, "step": 49930 }, { "epoch": 4.2556455048998725, "grad_norm": 2.578125, "learning_rate": 4.746597683114581e-05, "loss": 1.2061559677124023, "step": 49940 }, { "epoch": 4.256497656582872, "grad_norm": 1.5078125, "learning_rate": 4.7464765949251587e-05, "loss": 1.328407096862793, "step": 49950 }, { "epoch": 4.257349808265872, "grad_norm": 2.375, "learning_rate": 4.746355479356945e-05, "loss": 1.3053757667541503, "step": 49960 }, { "epoch": 4.258201959948871, "grad_norm": 2.328125, "learning_rate": 4.746234336411418e-05, "loss": 1.5044212341308594, "step": 49970 }, { "epoch": 4.259054111631871, "grad_norm": 4.8125, "learning_rate": 4.7461131660900523e-05, "loss": 1.2889473915100098, "step": 49980 }, { "epoch": 4.25990626331487, "grad_norm": 3.546875, "learning_rate": 4.745991968394326e-05, "loss": 1.2430792808532716, "step": 49990 }, { "epoch": 4.26075841499787, "grad_norm": 2.015625, "learning_rate": 4.7458707433257155e-05, "loss": 1.2628746032714844, "step": 50000 }, { "epoch": 4.2616105666808695, "grad_norm": 2.21875, "learning_rate": 4.7457494908856975e-05, "loss": 1.4223080635070802, "step": 50010 }, { "epoch": 4.262462718363869, "grad_norm": 1.5, "learning_rate": 4.745628211075751e-05, "loss": 1.2828889846801759, "step": 50020 }, { "epoch": 4.263314870046869, "grad_norm": 2.328125, "learning_rate": 4.745506903897354e-05, "loss": 1.4323182106018066, "step": 50030 }, { "epoch": 4.264167021729868, "grad_norm": 2.421875, "learning_rate": 4.745385569351984e-05, "loss": 1.390472984313965, "step": 50040 }, { "epoch": 4.265019173412868, "grad_norm": 2.625, "learning_rate": 4.74526420744112e-05, "loss": 1.3201151847839356, "step": 50050 }, { "epoch": 4.265871325095867, "grad_norm": 2.96875, "learning_rate": 4.745142818166242e-05, "loss": 1.2218480110168457, "step": 50060 }, { "epoch": 4.266723476778867, "grad_norm": 2.765625, "learning_rate": 4.745021401528829e-05, "loss": 1.3557467460632324, "step": 50070 }, { "epoch": 4.267575628461866, "grad_norm": 2.265625, "learning_rate": 4.74489995753036e-05, "loss": 1.369239616394043, "step": 50080 }, { "epoch": 4.268427780144866, "grad_norm": 1.8359375, "learning_rate": 4.744778486172316e-05, "loss": 1.4584948539733886, "step": 50090 }, { "epoch": 4.269279931827866, "grad_norm": 1.5390625, "learning_rate": 4.744656987456178e-05, "loss": 1.2478509902954102, "step": 50100 }, { "epoch": 4.270132083510865, "grad_norm": 2.078125, "learning_rate": 4.744535461383424e-05, "loss": 1.2384916305541993, "step": 50110 }, { "epoch": 4.270984235193865, "grad_norm": 1.9375, "learning_rate": 4.7444139079555376e-05, "loss": 1.3567386627197267, "step": 50120 }, { "epoch": 4.271836386876864, "grad_norm": 2.40625, "learning_rate": 4.744292327174e-05, "loss": 1.3262225151062013, "step": 50130 }, { "epoch": 4.272688538559864, "grad_norm": 1.8125, "learning_rate": 4.7441707190402926e-05, "loss": 1.4375902175903321, "step": 50140 }, { "epoch": 4.273540690242863, "grad_norm": 3.0625, "learning_rate": 4.744049083555897e-05, "loss": 1.307309055328369, "step": 50150 }, { "epoch": 4.274392841925863, "grad_norm": 2.6875, "learning_rate": 4.743927420722296e-05, "loss": 1.3502630233764648, "step": 50160 }, { "epoch": 4.275244993608863, "grad_norm": 3.46875, "learning_rate": 4.743805730540972e-05, "loss": 1.308204174041748, "step": 50170 }, { "epoch": 4.276097145291862, "grad_norm": 4.375, "learning_rate": 4.74368401301341e-05, "loss": 1.2462507247924806, "step": 50180 }, { "epoch": 4.276949296974862, "grad_norm": 1.65625, "learning_rate": 4.7435622681410905e-05, "loss": 1.4116354942321778, "step": 50190 }, { "epoch": 4.277801448657861, "grad_norm": 2.6875, "learning_rate": 4.743440495925499e-05, "loss": 1.4522403717041015, "step": 50200 }, { "epoch": 4.278653600340861, "grad_norm": 1.5390625, "learning_rate": 4.7433186963681195e-05, "loss": 1.2624896049499512, "step": 50210 }, { "epoch": 4.27950575202386, "grad_norm": 2.0, "learning_rate": 4.7431968694704367e-05, "loss": 1.4514874458312987, "step": 50220 }, { "epoch": 4.28035790370686, "grad_norm": 1.5390625, "learning_rate": 4.743075015233934e-05, "loss": 1.455386257171631, "step": 50230 }, { "epoch": 4.2812100553898595, "grad_norm": 4.25, "learning_rate": 4.7429531336600975e-05, "loss": 1.229531192779541, "step": 50240 }, { "epoch": 4.282062207072859, "grad_norm": 2.078125, "learning_rate": 4.7428312247504126e-05, "loss": 1.4666016578674317, "step": 50250 }, { "epoch": 4.282914358755859, "grad_norm": 2.015625, "learning_rate": 4.7427092885063646e-05, "loss": 1.260392665863037, "step": 50260 }, { "epoch": 4.283766510438858, "grad_norm": 2.546875, "learning_rate": 4.74258732492944e-05, "loss": 1.4338666915893554, "step": 50270 }, { "epoch": 4.284618662121858, "grad_norm": 2.515625, "learning_rate": 4.7424653340211256e-05, "loss": 1.2893310546875, "step": 50280 }, { "epoch": 4.285470813804857, "grad_norm": 2.15625, "learning_rate": 4.7423433157829066e-05, "loss": 1.4298749923706056, "step": 50290 }, { "epoch": 4.286322965487857, "grad_norm": 2.671875, "learning_rate": 4.742221270216273e-05, "loss": 1.267020034790039, "step": 50300 }, { "epoch": 4.2871751171708565, "grad_norm": 2.40625, "learning_rate": 4.7420991973227095e-05, "loss": 1.2611932754516602, "step": 50310 }, { "epoch": 4.288027268853856, "grad_norm": 3.3125, "learning_rate": 4.741977097103703e-05, "loss": 1.3790964126586913, "step": 50320 }, { "epoch": 4.288879420536856, "grad_norm": 2.0, "learning_rate": 4.741854969560746e-05, "loss": 1.3567825317382813, "step": 50330 }, { "epoch": 4.289731572219855, "grad_norm": 2.765625, "learning_rate": 4.741732814695323e-05, "loss": 1.3615248680114747, "step": 50340 }, { "epoch": 4.290583723902855, "grad_norm": 3.578125, "learning_rate": 4.741610632508924e-05, "loss": 1.2442017555236817, "step": 50350 }, { "epoch": 4.291435875585854, "grad_norm": 1.6484375, "learning_rate": 4.7414884230030386e-05, "loss": 1.2584043502807618, "step": 50360 }, { "epoch": 4.292288027268854, "grad_norm": 2.25, "learning_rate": 4.741366186179155e-05, "loss": 1.386863899230957, "step": 50370 }, { "epoch": 4.2931401789518535, "grad_norm": 4.65625, "learning_rate": 4.7412439220387646e-05, "loss": 1.231083297729492, "step": 50380 }, { "epoch": 4.293992330634853, "grad_norm": 1.8984375, "learning_rate": 4.741121630583356e-05, "loss": 1.1295302391052247, "step": 50390 }, { "epoch": 4.294844482317853, "grad_norm": 3.328125, "learning_rate": 4.7409993118144206e-05, "loss": 1.2101696014404297, "step": 50400 }, { "epoch": 4.295696634000852, "grad_norm": 2.5, "learning_rate": 4.740876965733448e-05, "loss": 1.37430419921875, "step": 50410 }, { "epoch": 4.296548785683852, "grad_norm": 3.5, "learning_rate": 4.7407545923419306e-05, "loss": 1.2876007080078125, "step": 50420 }, { "epoch": 4.297400937366851, "grad_norm": 2.3125, "learning_rate": 4.740632191641359e-05, "loss": 1.292160987854004, "step": 50430 }, { "epoch": 4.298253089049851, "grad_norm": 2.109375, "learning_rate": 4.740509763633225e-05, "loss": 1.2395099639892577, "step": 50440 }, { "epoch": 4.2991052407328505, "grad_norm": 2.6875, "learning_rate": 4.740387308319021e-05, "loss": 1.399420166015625, "step": 50450 }, { "epoch": 4.29995739241585, "grad_norm": 1.9921875, "learning_rate": 4.74026482570024e-05, "loss": 1.4046920776367187, "step": 50460 }, { "epoch": 4.30080954409885, "grad_norm": 2.390625, "learning_rate": 4.740142315778373e-05, "loss": 1.361874485015869, "step": 50470 }, { "epoch": 4.301661695781849, "grad_norm": 2.640625, "learning_rate": 4.740019778554915e-05, "loss": 1.4112221717834472, "step": 50480 }, { "epoch": 4.302513847464849, "grad_norm": 1.9765625, "learning_rate": 4.739897214031358e-05, "loss": 1.4817785263061523, "step": 50490 }, { "epoch": 4.303365999147848, "grad_norm": 3.3125, "learning_rate": 4.7397746222091965e-05, "loss": 1.4262489318847655, "step": 50500 }, { "epoch": 4.304218150830848, "grad_norm": 2.3125, "learning_rate": 4.739652003089924e-05, "loss": 1.311110496520996, "step": 50510 }, { "epoch": 4.3050703025138475, "grad_norm": 1.890625, "learning_rate": 4.739529356675035e-05, "loss": 1.2391299247741698, "step": 50520 }, { "epoch": 4.305922454196847, "grad_norm": 2.75, "learning_rate": 4.7394066829660255e-05, "loss": 1.3859442710876464, "step": 50530 }, { "epoch": 4.306774605879847, "grad_norm": 2.6875, "learning_rate": 4.73928398196439e-05, "loss": 1.3971612930297852, "step": 50540 }, { "epoch": 4.307626757562846, "grad_norm": 1.8203125, "learning_rate": 4.739161253671622e-05, "loss": 1.34049711227417, "step": 50550 }, { "epoch": 4.308478909245846, "grad_norm": 2.765625, "learning_rate": 4.73903849808922e-05, "loss": 1.284140396118164, "step": 50560 }, { "epoch": 4.309331060928845, "grad_norm": 2.703125, "learning_rate": 4.738915715218678e-05, "loss": 1.2720701217651367, "step": 50570 }, { "epoch": 4.310183212611845, "grad_norm": 2.90625, "learning_rate": 4.7387929050614935e-05, "loss": 1.3345560073852538, "step": 50580 }, { "epoch": 4.3110353642948445, "grad_norm": 4.21875, "learning_rate": 4.738670067619163e-05, "loss": 1.1046106338500976, "step": 50590 }, { "epoch": 4.311887515977844, "grad_norm": 2.515625, "learning_rate": 4.738547202893183e-05, "loss": 1.3712390899658202, "step": 50600 }, { "epoch": 4.312739667660844, "grad_norm": 2.328125, "learning_rate": 4.738424310885052e-05, "loss": 1.3016518592834472, "step": 50610 }, { "epoch": 4.313591819343843, "grad_norm": 2.515625, "learning_rate": 4.738301391596268e-05, "loss": 1.3429903030395507, "step": 50620 }, { "epoch": 4.314443971026843, "grad_norm": 2.328125, "learning_rate": 4.7381784450283274e-05, "loss": 1.4327373504638672, "step": 50630 }, { "epoch": 4.315296122709842, "grad_norm": 3.40625, "learning_rate": 4.73805547118273e-05, "loss": 1.402329921722412, "step": 50640 }, { "epoch": 4.316148274392842, "grad_norm": 2.234375, "learning_rate": 4.737932470060973e-05, "loss": 1.3560981750488281, "step": 50650 }, { "epoch": 4.3170004260758414, "grad_norm": 1.7109375, "learning_rate": 4.737809441664557e-05, "loss": 1.2822347640991212, "step": 50660 }, { "epoch": 4.317852577758841, "grad_norm": 2.3125, "learning_rate": 4.737686385994981e-05, "loss": 1.372914695739746, "step": 50670 }, { "epoch": 4.318704729441841, "grad_norm": 1.90625, "learning_rate": 4.7375633030537446e-05, "loss": 1.381544589996338, "step": 50680 }, { "epoch": 4.31955688112484, "grad_norm": 2.796875, "learning_rate": 4.737440192842347e-05, "loss": 1.3303171157836915, "step": 50690 }, { "epoch": 4.32040903280784, "grad_norm": 1.9453125, "learning_rate": 4.737317055362291e-05, "loss": 1.4214120864868165, "step": 50700 }, { "epoch": 4.321261184490839, "grad_norm": 3.015625, "learning_rate": 4.737193890615075e-05, "loss": 1.4406256675720215, "step": 50710 }, { "epoch": 4.322113336173839, "grad_norm": 1.828125, "learning_rate": 4.737070698602201e-05, "loss": 1.5361178398132325, "step": 50720 }, { "epoch": 4.322965487856838, "grad_norm": 2.59375, "learning_rate": 4.73694747932517e-05, "loss": 1.4197776794433594, "step": 50730 }, { "epoch": 4.323817639539838, "grad_norm": 2.328125, "learning_rate": 4.736824232785484e-05, "loss": 1.3990236282348634, "step": 50740 }, { "epoch": 4.324669791222838, "grad_norm": 2.5625, "learning_rate": 4.736700958984646e-05, "loss": 1.3582676887512206, "step": 50750 }, { "epoch": 4.325521942905837, "grad_norm": 2.046875, "learning_rate": 4.7365776579241564e-05, "loss": 1.3855535507202148, "step": 50760 }, { "epoch": 4.326374094588837, "grad_norm": 1.734375, "learning_rate": 4.7364543296055194e-05, "loss": 1.3200735092163085, "step": 50770 }, { "epoch": 4.327226246271836, "grad_norm": 3.0625, "learning_rate": 4.736330974030238e-05, "loss": 1.3921517372131347, "step": 50780 }, { "epoch": 4.328078397954836, "grad_norm": 3.28125, "learning_rate": 4.736207591199814e-05, "loss": 1.2619815826416017, "step": 50790 }, { "epoch": 4.328930549637835, "grad_norm": 3.046875, "learning_rate": 4.736084181115753e-05, "loss": 1.424309539794922, "step": 50800 }, { "epoch": 4.329782701320835, "grad_norm": 2.703125, "learning_rate": 4.7359607437795584e-05, "loss": 1.3423463821411132, "step": 50810 }, { "epoch": 4.3306348530038346, "grad_norm": 1.9140625, "learning_rate": 4.735837279192734e-05, "loss": 1.355416488647461, "step": 50820 }, { "epoch": 4.331487004686834, "grad_norm": 1.765625, "learning_rate": 4.7357137873567856e-05, "loss": 1.5033370971679687, "step": 50830 }, { "epoch": 4.332339156369834, "grad_norm": 1.8203125, "learning_rate": 4.735590268273218e-05, "loss": 1.3057289123535156, "step": 50840 }, { "epoch": 4.333191308052833, "grad_norm": 2.78125, "learning_rate": 4.735466721943537e-05, "loss": 1.4246686935424804, "step": 50850 }, { "epoch": 4.334043459735833, "grad_norm": 2.203125, "learning_rate": 4.735343148369246e-05, "loss": 1.3225807189941405, "step": 50860 }, { "epoch": 4.334895611418832, "grad_norm": 2.265625, "learning_rate": 4.7352195475518533e-05, "loss": 1.3866913795471192, "step": 50870 }, { "epoch": 4.335747763101832, "grad_norm": 2.859375, "learning_rate": 4.735095919492864e-05, "loss": 1.414072036743164, "step": 50880 }, { "epoch": 4.3365999147848315, "grad_norm": 2.640625, "learning_rate": 4.734972264193787e-05, "loss": 1.2877133369445801, "step": 50890 }, { "epoch": 4.337452066467831, "grad_norm": 2.5, "learning_rate": 4.734848581656127e-05, "loss": 1.4427194595336914, "step": 50900 }, { "epoch": 4.338304218150831, "grad_norm": 2.671875, "learning_rate": 4.734724871881392e-05, "loss": 1.4052599906921386, "step": 50910 }, { "epoch": 4.33915636983383, "grad_norm": 2.171875, "learning_rate": 4.73460113487109e-05, "loss": 1.2549075126647948, "step": 50920 }, { "epoch": 4.34000852151683, "grad_norm": 1.765625, "learning_rate": 4.7344773706267293e-05, "loss": 1.4091457366943358, "step": 50930 }, { "epoch": 4.340860673199829, "grad_norm": 1.9375, "learning_rate": 4.7343535791498175e-05, "loss": 1.3885202407836914, "step": 50940 }, { "epoch": 4.341712824882829, "grad_norm": 2.703125, "learning_rate": 4.734229760441864e-05, "loss": 1.391251564025879, "step": 50950 }, { "epoch": 4.3425649765658285, "grad_norm": 2.265625, "learning_rate": 4.734105914504376e-05, "loss": 1.2461467742919923, "step": 50960 }, { "epoch": 4.343417128248828, "grad_norm": 2.5625, "learning_rate": 4.733982041338866e-05, "loss": 1.3024754524230957, "step": 50970 }, { "epoch": 4.344269279931828, "grad_norm": 2.03125, "learning_rate": 4.733858140946842e-05, "loss": 1.3537822723388673, "step": 50980 }, { "epoch": 4.345121431614827, "grad_norm": 2.96875, "learning_rate": 4.733734213329814e-05, "loss": 1.3916916847229004, "step": 50990 }, { "epoch": 4.345973583297827, "grad_norm": 2.34375, "learning_rate": 4.733610258489292e-05, "loss": 1.3346473693847656, "step": 51000 }, { "epoch": 4.346825734980826, "grad_norm": 2.765625, "learning_rate": 4.7334862764267876e-05, "loss": 1.3257154464721679, "step": 51010 }, { "epoch": 4.347677886663826, "grad_norm": 1.8203125, "learning_rate": 4.733362267143811e-05, "loss": 1.3731645584106444, "step": 51020 }, { "epoch": 4.3485300383468255, "grad_norm": 1.9921875, "learning_rate": 4.733238230641874e-05, "loss": 1.380575942993164, "step": 51030 }, { "epoch": 4.349382190029825, "grad_norm": 2.546875, "learning_rate": 4.7331141669224886e-05, "loss": 1.3092714309692384, "step": 51040 }, { "epoch": 4.350234341712825, "grad_norm": 2.90625, "learning_rate": 4.732990075987166e-05, "loss": 1.3553893089294433, "step": 51050 }, { "epoch": 4.351086493395824, "grad_norm": 2.21875, "learning_rate": 4.732865957837419e-05, "loss": 1.4099571228027343, "step": 51060 }, { "epoch": 4.351938645078824, "grad_norm": 1.8515625, "learning_rate": 4.7327418124747603e-05, "loss": 1.3616318702697754, "step": 51070 }, { "epoch": 4.352790796761823, "grad_norm": 1.546875, "learning_rate": 4.7326176399007035e-05, "loss": 1.4605762481689453, "step": 51080 }, { "epoch": 4.353642948444823, "grad_norm": 3.203125, "learning_rate": 4.732493440116761e-05, "loss": 1.3118865013122558, "step": 51090 }, { "epoch": 4.3544951001278225, "grad_norm": 1.765625, "learning_rate": 4.732369213124446e-05, "loss": 1.3310431480407714, "step": 51100 }, { "epoch": 4.355347251810822, "grad_norm": 2.0625, "learning_rate": 4.732244958925274e-05, "loss": 1.290602970123291, "step": 51110 }, { "epoch": 4.356199403493822, "grad_norm": 2.75, "learning_rate": 4.732120677520759e-05, "loss": 1.302020263671875, "step": 51120 }, { "epoch": 4.357051555176821, "grad_norm": 2.15625, "learning_rate": 4.7319963689124145e-05, "loss": 1.2885354995727538, "step": 51130 }, { "epoch": 4.357903706859821, "grad_norm": 1.7109375, "learning_rate": 4.7318720331017575e-05, "loss": 1.3284104347229004, "step": 51140 }, { "epoch": 4.35875585854282, "grad_norm": 1.9765625, "learning_rate": 4.731747670090301e-05, "loss": 1.3064316749572753, "step": 51150 }, { "epoch": 4.35960801022582, "grad_norm": 2.46875, "learning_rate": 4.731623279879562e-05, "loss": 1.393303108215332, "step": 51160 }, { "epoch": 4.3604601619088195, "grad_norm": 1.8203125, "learning_rate": 4.731498862471057e-05, "loss": 1.3918696403503419, "step": 51170 }, { "epoch": 4.361312313591819, "grad_norm": 1.9765625, "learning_rate": 4.7313744178663014e-05, "loss": 1.234218215942383, "step": 51180 }, { "epoch": 4.362164465274819, "grad_norm": 2.53125, "learning_rate": 4.731249946066812e-05, "loss": 1.3574142456054688, "step": 51190 }, { "epoch": 4.363016616957818, "grad_norm": 3.90625, "learning_rate": 4.731125447074106e-05, "loss": 1.1485512733459473, "step": 51200 }, { "epoch": 4.363868768640818, "grad_norm": 2.546875, "learning_rate": 4.7310009208897e-05, "loss": 1.3150632858276368, "step": 51210 }, { "epoch": 4.364720920323817, "grad_norm": 1.75, "learning_rate": 4.730876367515113e-05, "loss": 1.480640697479248, "step": 51220 }, { "epoch": 4.365573072006817, "grad_norm": 2.3125, "learning_rate": 4.730751786951862e-05, "loss": 1.2771005630493164, "step": 51230 }, { "epoch": 4.3664252236898164, "grad_norm": 2.203125, "learning_rate": 4.730627179201465e-05, "loss": 1.2767502784729003, "step": 51240 }, { "epoch": 4.367277375372816, "grad_norm": 2.65625, "learning_rate": 4.730502544265443e-05, "loss": 1.4534384727478027, "step": 51250 }, { "epoch": 4.368129527055816, "grad_norm": 3.390625, "learning_rate": 4.730377882145312e-05, "loss": 1.2436548233032227, "step": 51260 }, { "epoch": 4.368981678738815, "grad_norm": 3.890625, "learning_rate": 4.730253192842593e-05, "loss": 1.3371723175048829, "step": 51270 }, { "epoch": 4.369833830421815, "grad_norm": 2.03125, "learning_rate": 4.730128476358805e-05, "loss": 1.2166253089904786, "step": 51280 }, { "epoch": 4.370685982104814, "grad_norm": 1.6953125, "learning_rate": 4.7300037326954676e-05, "loss": 1.33047513961792, "step": 51290 }, { "epoch": 4.371538133787814, "grad_norm": 3.5625, "learning_rate": 4.7298789618541024e-05, "loss": 1.2084044456481933, "step": 51300 }, { "epoch": 4.372390285470813, "grad_norm": 2.1875, "learning_rate": 4.7297541638362286e-05, "loss": 1.4705440521240234, "step": 51310 }, { "epoch": 4.373242437153813, "grad_norm": 3.34375, "learning_rate": 4.729629338643369e-05, "loss": 1.3626379013061523, "step": 51320 }, { "epoch": 4.374094588836813, "grad_norm": 1.96875, "learning_rate": 4.729504486277042e-05, "loss": 1.159682273864746, "step": 51330 }, { "epoch": 4.374946740519812, "grad_norm": 2.203125, "learning_rate": 4.729379606738773e-05, "loss": 1.4414993286132813, "step": 51340 }, { "epoch": 4.375798892202812, "grad_norm": 2.25, "learning_rate": 4.72925470003008e-05, "loss": 1.3296422958374023, "step": 51350 }, { "epoch": 4.376651043885811, "grad_norm": 2.3125, "learning_rate": 4.729129766152489e-05, "loss": 1.3502317428588868, "step": 51360 }, { "epoch": 4.377503195568811, "grad_norm": 2.78125, "learning_rate": 4.72900480510752e-05, "loss": 1.302121639251709, "step": 51370 }, { "epoch": 4.37835534725181, "grad_norm": 3.71875, "learning_rate": 4.728879816896697e-05, "loss": 1.328645896911621, "step": 51380 }, { "epoch": 4.37920749893481, "grad_norm": 2.859375, "learning_rate": 4.728754801521543e-05, "loss": 1.052384090423584, "step": 51390 }, { "epoch": 4.3800596506178096, "grad_norm": 2.140625, "learning_rate": 4.728629758983582e-05, "loss": 1.2132914543151856, "step": 51400 }, { "epoch": 4.38091180230081, "grad_norm": 2.5625, "learning_rate": 4.7285046892843374e-05, "loss": 1.2881938934326171, "step": 51410 }, { "epoch": 4.381763953983809, "grad_norm": 4.4375, "learning_rate": 4.728379592425334e-05, "loss": 1.4086061477661134, "step": 51420 }, { "epoch": 4.382616105666809, "grad_norm": 2.015625, "learning_rate": 4.728254468408095e-05, "loss": 1.4424535751342773, "step": 51430 }, { "epoch": 4.383468257349808, "grad_norm": 1.9609375, "learning_rate": 4.7281293172341476e-05, "loss": 1.3031806945800781, "step": 51440 }, { "epoch": 4.384320409032808, "grad_norm": 1.78125, "learning_rate": 4.728004138905016e-05, "loss": 1.3105274200439454, "step": 51450 }, { "epoch": 4.385172560715807, "grad_norm": 1.9375, "learning_rate": 4.7278789334222256e-05, "loss": 1.2737356185913087, "step": 51460 }, { "epoch": 4.386024712398807, "grad_norm": 3.359375, "learning_rate": 4.727753700787303e-05, "loss": 1.3920160293579102, "step": 51470 }, { "epoch": 4.386876864081806, "grad_norm": 4.15625, "learning_rate": 4.727628441001773e-05, "loss": 1.4093953132629395, "step": 51480 }, { "epoch": 4.387729015764807, "grad_norm": 2.515625, "learning_rate": 4.7275031540671635e-05, "loss": 1.3680298805236817, "step": 51490 }, { "epoch": 4.388581167447805, "grad_norm": 2.09375, "learning_rate": 4.7273778399850015e-05, "loss": 1.2623068809509277, "step": 51500 }, { "epoch": 4.389433319130806, "grad_norm": 4.5, "learning_rate": 4.727252498756814e-05, "loss": 1.3983208656311035, "step": 51510 }, { "epoch": 4.390285470813805, "grad_norm": 2.125, "learning_rate": 4.727127130384128e-05, "loss": 1.2783860206604003, "step": 51520 }, { "epoch": 4.391137622496805, "grad_norm": 2.640625, "learning_rate": 4.727001734868472e-05, "loss": 1.425441074371338, "step": 51530 }, { "epoch": 4.391989774179804, "grad_norm": 2.6875, "learning_rate": 4.7268763122113734e-05, "loss": 1.3890666961669922, "step": 51540 }, { "epoch": 4.392841925862804, "grad_norm": 2.46875, "learning_rate": 4.726750862414363e-05, "loss": 1.3243844032287597, "step": 51550 }, { "epoch": 4.3936940775458035, "grad_norm": 1.515625, "learning_rate": 4.726625385478967e-05, "loss": 1.3592787742614747, "step": 51560 }, { "epoch": 4.394546229228803, "grad_norm": 1.953125, "learning_rate": 4.726499881406716e-05, "loss": 1.4445934295654297, "step": 51570 }, { "epoch": 4.395398380911803, "grad_norm": 2.578125, "learning_rate": 4.72637435019914e-05, "loss": 1.2368955612182617, "step": 51580 }, { "epoch": 4.396250532594802, "grad_norm": 2.21875, "learning_rate": 4.726248791857768e-05, "loss": 1.3715447425842284, "step": 51590 }, { "epoch": 4.397102684277802, "grad_norm": 2.9375, "learning_rate": 4.72612320638413e-05, "loss": 1.3849118232727051, "step": 51600 }, { "epoch": 4.397954835960801, "grad_norm": 2.546875, "learning_rate": 4.7259975937797586e-05, "loss": 1.4091493606567382, "step": 51610 }, { "epoch": 4.398806987643801, "grad_norm": 2.5625, "learning_rate": 4.7258719540461814e-05, "loss": 1.2470948219299316, "step": 51620 }, { "epoch": 4.3996591393268005, "grad_norm": 2.53125, "learning_rate": 4.725746287184933e-05, "loss": 1.3014641761779786, "step": 51630 }, { "epoch": 4.4005112910098, "grad_norm": 3.265625, "learning_rate": 4.725620593197543e-05, "loss": 1.2489782333374024, "step": 51640 }, { "epoch": 4.4013634426928, "grad_norm": 2.125, "learning_rate": 4.725494872085543e-05, "loss": 1.2441680908203125, "step": 51650 }, { "epoch": 4.402215594375799, "grad_norm": 2.6875, "learning_rate": 4.725369123850466e-05, "loss": 1.5057262420654296, "step": 51660 }, { "epoch": 4.403067746058799, "grad_norm": 2.03125, "learning_rate": 4.725243348493845e-05, "loss": 1.2681714057922364, "step": 51670 }, { "epoch": 4.403919897741798, "grad_norm": 2.671875, "learning_rate": 4.725117546017213e-05, "loss": 1.4401752471923828, "step": 51680 }, { "epoch": 4.404772049424798, "grad_norm": 2.3125, "learning_rate": 4.7249917164221016e-05, "loss": 1.380908489227295, "step": 51690 }, { "epoch": 4.4056242011077975, "grad_norm": 1.8984375, "learning_rate": 4.7248658597100445e-05, "loss": 1.542621898651123, "step": 51700 }, { "epoch": 4.406476352790797, "grad_norm": 3.234375, "learning_rate": 4.724739975882577e-05, "loss": 1.3149511337280273, "step": 51710 }, { "epoch": 4.407328504473797, "grad_norm": 2.0, "learning_rate": 4.724614064941233e-05, "loss": 1.3819555282592773, "step": 51720 }, { "epoch": 4.408180656156796, "grad_norm": 3.0, "learning_rate": 4.724488126887547e-05, "loss": 1.383721160888672, "step": 51730 }, { "epoch": 4.409032807839796, "grad_norm": 3.875, "learning_rate": 4.724362161723053e-05, "loss": 1.2156275749206542, "step": 51740 }, { "epoch": 4.409884959522795, "grad_norm": 2.375, "learning_rate": 4.724236169449287e-05, "loss": 1.3051067352294923, "step": 51750 }, { "epoch": 4.410737111205795, "grad_norm": 3.90625, "learning_rate": 4.724110150067784e-05, "loss": 1.339913558959961, "step": 51760 }, { "epoch": 4.4115892628887945, "grad_norm": 2.8125, "learning_rate": 4.72398410358008e-05, "loss": 1.298574733734131, "step": 51770 }, { "epoch": 4.412441414571794, "grad_norm": 1.8203125, "learning_rate": 4.7238580299877125e-05, "loss": 1.404612159729004, "step": 51780 }, { "epoch": 4.413293566254794, "grad_norm": 2.125, "learning_rate": 4.7237319292922155e-05, "loss": 1.4551277160644531, "step": 51790 }, { "epoch": 4.414145717937793, "grad_norm": 1.6640625, "learning_rate": 4.723605801495128e-05, "loss": 1.277263069152832, "step": 51800 }, { "epoch": 4.414997869620793, "grad_norm": 1.9609375, "learning_rate": 4.723479646597986e-05, "loss": 1.3433650970458983, "step": 51810 }, { "epoch": 4.415850021303792, "grad_norm": 2.515625, "learning_rate": 4.723353464602327e-05, "loss": 1.3746132850646973, "step": 51820 }, { "epoch": 4.416702172986792, "grad_norm": 2.140625, "learning_rate": 4.723227255509689e-05, "loss": 1.2627121925354003, "step": 51830 }, { "epoch": 4.4175543246697915, "grad_norm": 2.203125, "learning_rate": 4.7231010193216106e-05, "loss": 1.4529315948486328, "step": 51840 }, { "epoch": 4.418406476352791, "grad_norm": 2.625, "learning_rate": 4.722974756039631e-05, "loss": 1.187460708618164, "step": 51850 }, { "epoch": 4.419258628035791, "grad_norm": 2.984375, "learning_rate": 4.722848465665287e-05, "loss": 1.2811226844787598, "step": 51860 }, { "epoch": 4.42011077971879, "grad_norm": 2.234375, "learning_rate": 4.722722148200119e-05, "loss": 1.320591926574707, "step": 51870 }, { "epoch": 4.42096293140179, "grad_norm": 3.03125, "learning_rate": 4.722595803645666e-05, "loss": 1.3625596046447754, "step": 51880 }, { "epoch": 4.421815083084789, "grad_norm": 1.9140625, "learning_rate": 4.722469432003469e-05, "loss": 1.246577835083008, "step": 51890 }, { "epoch": 4.422667234767789, "grad_norm": 2.515625, "learning_rate": 4.722343033275067e-05, "loss": 1.4002212524414062, "step": 51900 }, { "epoch": 4.4235193864507885, "grad_norm": 2.234375, "learning_rate": 4.722216607462e-05, "loss": 1.456306266784668, "step": 51910 }, { "epoch": 4.424371538133788, "grad_norm": 2.84375, "learning_rate": 4.7220901545658106e-05, "loss": 1.291639804840088, "step": 51920 }, { "epoch": 4.425223689816788, "grad_norm": 2.265625, "learning_rate": 4.721963674588038e-05, "loss": 1.1895892143249511, "step": 51930 }, { "epoch": 4.426075841499787, "grad_norm": 2.265625, "learning_rate": 4.721837167530225e-05, "loss": 1.2824626922607423, "step": 51940 }, { "epoch": 4.426927993182787, "grad_norm": 2.171875, "learning_rate": 4.721710633393913e-05, "loss": 1.4057652473449707, "step": 51950 }, { "epoch": 4.427780144865786, "grad_norm": 2.203125, "learning_rate": 4.7215840721806436e-05, "loss": 1.494624423980713, "step": 51960 }, { "epoch": 4.428632296548786, "grad_norm": 1.7265625, "learning_rate": 4.72145748389196e-05, "loss": 1.3490862846374512, "step": 51970 }, { "epoch": 4.429484448231785, "grad_norm": 2.640625, "learning_rate": 4.721330868529406e-05, "loss": 1.2693161964416504, "step": 51980 }, { "epoch": 4.430336599914785, "grad_norm": 2.625, "learning_rate": 4.721204226094521e-05, "loss": 1.364323043823242, "step": 51990 }, { "epoch": 4.431188751597785, "grad_norm": 4.6875, "learning_rate": 4.7210775565888526e-05, "loss": 1.3372342109680175, "step": 52000 }, { "epoch": 4.432040903280784, "grad_norm": 2.09375, "learning_rate": 4.720950860013942e-05, "loss": 1.4406549453735351, "step": 52010 }, { "epoch": 4.432893054963784, "grad_norm": 2.671875, "learning_rate": 4.720824136371335e-05, "loss": 1.3249446868896484, "step": 52020 }, { "epoch": 4.433745206646783, "grad_norm": 2.109375, "learning_rate": 4.720697385662575e-05, "loss": 1.3559852600097657, "step": 52030 }, { "epoch": 4.434597358329783, "grad_norm": 2.53125, "learning_rate": 4.720570607889206e-05, "loss": 1.285771942138672, "step": 52040 }, { "epoch": 4.435449510012782, "grad_norm": 2.71875, "learning_rate": 4.7204438030527754e-05, "loss": 1.3291612625122071, "step": 52050 }, { "epoch": 4.436301661695782, "grad_norm": 2.359375, "learning_rate": 4.720316971154827e-05, "loss": 1.282005214691162, "step": 52060 }, { "epoch": 4.437153813378782, "grad_norm": 3.140625, "learning_rate": 4.720190112196906e-05, "loss": 1.2889132499694824, "step": 52070 }, { "epoch": 4.438005965061781, "grad_norm": 2.171875, "learning_rate": 4.72006322618056e-05, "loss": 1.2742056846618652, "step": 52080 }, { "epoch": 4.438858116744781, "grad_norm": 2.09375, "learning_rate": 4.719936313107335e-05, "loss": 1.336405372619629, "step": 52090 }, { "epoch": 4.43971026842778, "grad_norm": 2.25, "learning_rate": 4.719809372978777e-05, "loss": 1.3646011352539062, "step": 52100 }, { "epoch": 4.44056242011078, "grad_norm": 5.4375, "learning_rate": 4.719682405796434e-05, "loss": 1.3153329849243165, "step": 52110 }, { "epoch": 4.441414571793779, "grad_norm": 1.984375, "learning_rate": 4.719555411561852e-05, "loss": 1.4329837799072265, "step": 52120 }, { "epoch": 4.442266723476779, "grad_norm": 2.71875, "learning_rate": 4.7194283902765814e-05, "loss": 1.373297882080078, "step": 52130 }, { "epoch": 4.4431188751597785, "grad_norm": 4.28125, "learning_rate": 4.7193013419421675e-05, "loss": 1.38972806930542, "step": 52140 }, { "epoch": 4.443971026842778, "grad_norm": 1.625, "learning_rate": 4.719174266560159e-05, "loss": 1.3506338119506835, "step": 52150 }, { "epoch": 4.444823178525778, "grad_norm": 1.84375, "learning_rate": 4.7190471641321064e-05, "loss": 1.245115375518799, "step": 52160 }, { "epoch": 4.445675330208777, "grad_norm": 2.453125, "learning_rate": 4.7189200346595574e-05, "loss": 1.250307273864746, "step": 52170 }, { "epoch": 4.446527481891777, "grad_norm": 2.734375, "learning_rate": 4.718792878144062e-05, "loss": 1.2771223068237305, "step": 52180 }, { "epoch": 4.447379633574776, "grad_norm": 5.53125, "learning_rate": 4.71866569458717e-05, "loss": 1.4205263137817383, "step": 52190 }, { "epoch": 4.448231785257776, "grad_norm": 2.046875, "learning_rate": 4.71853848399043e-05, "loss": 1.1875314712524414, "step": 52200 }, { "epoch": 4.4490839369407755, "grad_norm": 3.625, "learning_rate": 4.718411246355394e-05, "loss": 1.1972894668579102, "step": 52210 }, { "epoch": 4.449936088623775, "grad_norm": 1.8984375, "learning_rate": 4.718283981683612e-05, "loss": 1.3672625541687011, "step": 52220 }, { "epoch": 4.450788240306775, "grad_norm": 2.453125, "learning_rate": 4.718156689976636e-05, "loss": 1.371175479888916, "step": 52230 }, { "epoch": 4.451640391989774, "grad_norm": 2.28125, "learning_rate": 4.718029371236016e-05, "loss": 1.346265697479248, "step": 52240 }, { "epoch": 4.452492543672774, "grad_norm": 4.3125, "learning_rate": 4.717902025463304e-05, "loss": 1.294471263885498, "step": 52250 }, { "epoch": 4.453344695355773, "grad_norm": 2.8125, "learning_rate": 4.717774652660052e-05, "loss": 1.2498647689819335, "step": 52260 }, { "epoch": 4.454196847038773, "grad_norm": 2.578125, "learning_rate": 4.7176472528278133e-05, "loss": 1.2370388984680176, "step": 52270 }, { "epoch": 4.4550489987217725, "grad_norm": 2.5, "learning_rate": 4.71751982596814e-05, "loss": 1.3271074295043945, "step": 52280 }, { "epoch": 4.455901150404772, "grad_norm": 2.328125, "learning_rate": 4.717392372082584e-05, "loss": 1.3110289573669434, "step": 52290 }, { "epoch": 4.456753302087772, "grad_norm": 1.6015625, "learning_rate": 4.7172648911727e-05, "loss": 1.3674738883972168, "step": 52300 }, { "epoch": 4.457605453770771, "grad_norm": 3.3125, "learning_rate": 4.717137383240041e-05, "loss": 1.3453113555908203, "step": 52310 }, { "epoch": 4.458457605453771, "grad_norm": 1.4765625, "learning_rate": 4.717009848286163e-05, "loss": 1.3489051818847657, "step": 52320 }, { "epoch": 4.45930975713677, "grad_norm": 2.703125, "learning_rate": 4.716882286312617e-05, "loss": 1.2913578987121581, "step": 52330 }, { "epoch": 4.46016190881977, "grad_norm": 1.546875, "learning_rate": 4.716754697320959e-05, "loss": 1.305628490447998, "step": 52340 }, { "epoch": 4.4610140605027695, "grad_norm": 3.921875, "learning_rate": 4.7166270813127446e-05, "loss": 1.384309959411621, "step": 52350 }, { "epoch": 4.461866212185769, "grad_norm": 2.171875, "learning_rate": 4.716499438289529e-05, "loss": 1.3308977127075194, "step": 52360 }, { "epoch": 4.462718363868769, "grad_norm": 4.09375, "learning_rate": 4.7163717682528676e-05, "loss": 1.1653109550476075, "step": 52370 }, { "epoch": 4.463570515551768, "grad_norm": 2.609375, "learning_rate": 4.716244071204317e-05, "loss": 1.2385673522949219, "step": 52380 }, { "epoch": 4.464422667234768, "grad_norm": 1.84375, "learning_rate": 4.7161163471454314e-05, "loss": 1.4105467796325684, "step": 52390 }, { "epoch": 4.465274818917767, "grad_norm": 2.484375, "learning_rate": 4.7159885960777696e-05, "loss": 1.4270364761352539, "step": 52400 }, { "epoch": 4.466126970600767, "grad_norm": 2.3125, "learning_rate": 4.715860818002888e-05, "loss": 1.2516942977905274, "step": 52410 }, { "epoch": 4.4669791222837665, "grad_norm": 2.921875, "learning_rate": 4.715733012922343e-05, "loss": 1.2874069213867188, "step": 52420 }, { "epoch": 4.467831273966766, "grad_norm": 2.421875, "learning_rate": 4.715605180837694e-05, "loss": 1.346622848510742, "step": 52430 }, { "epoch": 4.468683425649766, "grad_norm": 1.75, "learning_rate": 4.715477321750498e-05, "loss": 1.2774553298950195, "step": 52440 }, { "epoch": 4.469535577332765, "grad_norm": 2.421875, "learning_rate": 4.7153494356623116e-05, "loss": 1.402967643737793, "step": 52450 }, { "epoch": 4.470387729015765, "grad_norm": 2.90625, "learning_rate": 4.7152215225746966e-05, "loss": 1.4360246658325195, "step": 52460 }, { "epoch": 4.471239880698764, "grad_norm": 2.90625, "learning_rate": 4.7150935824892094e-05, "loss": 1.4600686073303222, "step": 52470 }, { "epoch": 4.472092032381764, "grad_norm": 2.15625, "learning_rate": 4.714965615407411e-05, "loss": 1.4089665412902832, "step": 52480 }, { "epoch": 4.4729441840647635, "grad_norm": 2.9375, "learning_rate": 4.714837621330859e-05, "loss": 1.1285311698913574, "step": 52490 }, { "epoch": 4.473796335747763, "grad_norm": 2.078125, "learning_rate": 4.714709600261115e-05, "loss": 1.26461181640625, "step": 52500 }, { "epoch": 4.474648487430763, "grad_norm": 2.046875, "learning_rate": 4.714581552199739e-05, "loss": 1.2379318237304688, "step": 52510 }, { "epoch": 4.475500639113762, "grad_norm": 2.703125, "learning_rate": 4.714453477148291e-05, "loss": 1.4417569160461425, "step": 52520 }, { "epoch": 4.476352790796762, "grad_norm": 2.046875, "learning_rate": 4.714325375108332e-05, "loss": 1.1615105628967286, "step": 52530 }, { "epoch": 4.477204942479761, "grad_norm": 2.5625, "learning_rate": 4.714197246081424e-05, "loss": 1.364054298400879, "step": 52540 }, { "epoch": 4.478057094162761, "grad_norm": 3.5, "learning_rate": 4.714069090069128e-05, "loss": 1.1322442054748536, "step": 52550 }, { "epoch": 4.4789092458457604, "grad_norm": 2.578125, "learning_rate": 4.713940907073006e-05, "loss": 1.496981716156006, "step": 52560 }, { "epoch": 4.47976139752876, "grad_norm": 1.6328125, "learning_rate": 4.7138126970946196e-05, "loss": 1.5629865646362304, "step": 52570 }, { "epoch": 4.48061354921176, "grad_norm": 2.078125, "learning_rate": 4.713684460135531e-05, "loss": 1.4211146354675293, "step": 52580 }, { "epoch": 4.481465700894759, "grad_norm": 2.1875, "learning_rate": 4.713556196197305e-05, "loss": 1.2861726760864258, "step": 52590 }, { "epoch": 4.482317852577759, "grad_norm": 2.625, "learning_rate": 4.713427905281504e-05, "loss": 1.4859222412109374, "step": 52600 }, { "epoch": 4.483170004260758, "grad_norm": 2.375, "learning_rate": 4.713299587389691e-05, "loss": 1.3589314460754394, "step": 52610 }, { "epoch": 4.484022155943758, "grad_norm": 1.7109375, "learning_rate": 4.713171242523431e-05, "loss": 1.4134737968444824, "step": 52620 }, { "epoch": 4.484874307626757, "grad_norm": 2.78125, "learning_rate": 4.7130428706842864e-05, "loss": 1.441242790222168, "step": 52630 }, { "epoch": 4.485726459309757, "grad_norm": 2.46875, "learning_rate": 4.712914471873823e-05, "loss": 1.3442601203918456, "step": 52640 }, { "epoch": 4.486578610992757, "grad_norm": 2.71875, "learning_rate": 4.712786046093605e-05, "loss": 1.3221062660217284, "step": 52650 }, { "epoch": 4.487430762675756, "grad_norm": 2.921875, "learning_rate": 4.712657593345198e-05, "loss": 1.3675683975219726, "step": 52660 }, { "epoch": 4.488282914358756, "grad_norm": 2.359375, "learning_rate": 4.712529113630167e-05, "loss": 1.388542366027832, "step": 52670 }, { "epoch": 4.489135066041755, "grad_norm": 1.9453125, "learning_rate": 4.712400606950079e-05, "loss": 1.4359724044799804, "step": 52680 }, { "epoch": 4.489987217724755, "grad_norm": 2.734375, "learning_rate": 4.712272073306499e-05, "loss": 1.2173480987548828, "step": 52690 }, { "epoch": 4.490839369407754, "grad_norm": 2.59375, "learning_rate": 4.712143512700995e-05, "loss": 1.3792642593383788, "step": 52700 }, { "epoch": 4.491691521090754, "grad_norm": 1.6640625, "learning_rate": 4.712014925135132e-05, "loss": 1.4352773666381835, "step": 52710 }, { "epoch": 4.4925436727737535, "grad_norm": 2.9375, "learning_rate": 4.7118863106104774e-05, "loss": 1.2108549118041991, "step": 52720 }, { "epoch": 4.493395824456753, "grad_norm": 2.125, "learning_rate": 4.7117576691285996e-05, "loss": 1.2764716148376465, "step": 52730 }, { "epoch": 4.494247976139753, "grad_norm": 2.9375, "learning_rate": 4.7116290006910665e-05, "loss": 1.3497371673583984, "step": 52740 }, { "epoch": 4.495100127822752, "grad_norm": 1.9921875, "learning_rate": 4.711500305299445e-05, "loss": 1.3921590805053712, "step": 52750 }, { "epoch": 4.495952279505752, "grad_norm": 2.21875, "learning_rate": 4.7113715829553044e-05, "loss": 1.4537745475769044, "step": 52760 }, { "epoch": 4.496804431188751, "grad_norm": 2.578125, "learning_rate": 4.7112428336602135e-05, "loss": 1.372967529296875, "step": 52770 }, { "epoch": 4.497656582871751, "grad_norm": 1.7734375, "learning_rate": 4.711114057415742e-05, "loss": 1.5109396934509278, "step": 52780 }, { "epoch": 4.4985087345547505, "grad_norm": 1.7890625, "learning_rate": 4.7109852542234575e-05, "loss": 1.3316495895385743, "step": 52790 }, { "epoch": 4.49936088623775, "grad_norm": 1.734375, "learning_rate": 4.710856424084932e-05, "loss": 1.2190245628356933, "step": 52800 }, { "epoch": 4.49961653174265, "eval_eval/n_valid_tokens": 450075, "eval_eval/top1_accuracy": 0.6505, "eval_loss": 1.1927706003189087, "eval_runtime": 54.0299, "eval_samples_per_second": 365.761, "eval_steps_per_second": 22.876, "step": 52803 } ], "logging_steps": 10, "max_steps": 293375, "num_input_tokens_seen": 0, "num_train_epochs": 25, "save_steps": 5867, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 6.682424374741893e+17, "train_batch_size": 32, "trial_name": null, "trial_params": null }